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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
cac73f74-0d4d-4f44-a510-983f8453196d | on-the-advantages-of-multiple-stereo-vision | 2105.12691 | null | https://arxiv.org/abs/2105.12691v1 | https://arxiv.org/pdf/2105.12691v1.pdf | On the Advantages of Multiple Stereo Vision Camera Designs for Autonomous Drone Navigation | In this work we showcase the design and assessment of the performance of a multi-camera UAV, when coupled with state-of-the-art planning and mapping algorithms for autonomous navigation. The system leverages state-of-the-art receding horizon exploration techniques for Next-Best-View (NBV) planning with 3D and semantic information, provided by a reconfigurable multi stereo camera system. We employ our approaches in an autonomous drone-based inspection task and evaluate them in an autonomous exploration and mapping scenario. We discuss the advantages and limitations of using multi stereo camera flying systems, and the trade-off between number of cameras and mapping performance. | ['Erdal Kayacan', 'Martim Brandão', 'Jonas Le Fevre', 'Jakob Grimm Hansen', 'Rui Pimentel de Figueiredo'] | 2021-05-26 | null | null | null | null | ['drone-navigation'] | ['computer-vision'] | [-2.24596914e-02 -7.85394758e-02 4.54469591e-01 -2.84977466e-01
-3.41681719e-01 -1.19838345e+00 4.68993902e-01 -1.87766448e-01
-4.58040506e-01 5.77583194e-01 -2.34006032e-01 -3.78779531e-01
-8.16531539e-01 -8.81336689e-01 -5.28701603e-01 -3.40780586e-01
-3.90462577e-01 8.56918633e-01 8.38165998e-01 -1.09587884e+00
3.51860374e-01 8.77318561e-01 -1.79557049e+00 -5.08120805e-02
5.52619278e-01 9.00298297e-01 7.15244353e-01 1.03883088e+00
8.46063852e-01 4.26079422e-01 -2.30820701e-01 1.56078890e-01
8.97928953e-01 6.50057420e-02 -1.05574512e+00 1.27296641e-01
4.96775448e-01 -6.10947371e-01 -2.45030478e-01 7.93374002e-01
2.73965687e-01 2.56379813e-01 6.95596635e-02 -1.03296995e+00
2.58855492e-01 -1.29719198e-01 -1.16159953e-01 4.97449964e-01
7.16315210e-01 2.34840199e-01 6.20674968e-01 -1.60074800e-01
8.01818192e-01 1.23058414e+00 6.51830673e-01 7.34743699e-02
-7.36245692e-01 -2.11144894e-01 -9.41424146e-02 -5.66211902e-02
-1.11754811e+00 -2.69648135e-01 7.35447854e-02 -4.22130287e-01
1.26529455e+00 -1.47358552e-01 1.17199755e+00 5.08922100e-01
9.94632781e-01 5.00787050e-02 1.20694518e+00 -3.10734719e-01
2.40539178e-01 -1.72343105e-01 -4.19544101e-01 1.05359757e+00
3.25486183e-01 9.15382028e-01 -4.65752512e-01 5.41441254e-02
9.86358762e-01 -1.84906706e-01 -2.21110061e-01 -9.30477798e-01
-1.05194914e+00 1.05827808e+00 3.57749432e-01 -3.13372314e-01
-3.79147410e-01 5.90540990e-02 1.77133203e-01 5.10921955e-01
-5.19167185e-02 8.75074387e-01 -5.87252319e-01 5.42467423e-02
-8.60912561e-01 4.32846755e-01 6.93428099e-01 1.27755678e+00
8.07223082e-01 -3.93836536e-02 4.97485429e-01 1.16333038e-01
4.29982543e-01 5.37285924e-01 -4.95180003e-02 -1.54860795e+00
2.67059386e-01 4.32883412e-01 5.47447443e-01 -5.35658777e-01
-6.00588918e-01 -2.55225062e-01 1.07099749e-01 9.20044601e-01
-1.76913619e-01 -4.36912715e-01 -6.17168009e-01 9.24236596e-01
4.95236397e-01 -6.18006051e-01 6.86350882e-01 8.95767748e-01
1.51012734e-01 2.60595113e-01 -7.64042258e-01 -6.72512800e-02
1.21326506e+00 -1.05973840e+00 -2.09309936e-01 -5.83532751e-01
3.99451286e-01 -5.47198892e-01 4.31450129e-01 8.98256779e-01
-8.65462005e-01 -5.08427024e-01 -1.31140149e+00 3.06842536e-01
-3.80384296e-01 2.23232228e-02 5.07856309e-01 2.74337173e-01
-1.55763054e+00 4.99925077e-01 -8.61481905e-01 -8.56576860e-01
-1.94747701e-01 4.60990041e-01 -5.67082584e-01 -3.02896500e-01
-8.99590909e-01 1.66115224e+00 8.11058342e-01 -1.34111598e-01
-1.95559716e+00 -3.03472340e-01 -9.77895141e-01 -4.14275587e-01
6.23866618e-01 -6.91733062e-01 1.33608758e+00 -1.92591146e-01
-1.48833597e+00 5.05713522e-01 5.37480652e-01 -6.68031454e-01
3.51650625e-01 -7.34442234e-01 -2.93704480e-01 7.07384825e-01
2.89513975e-01 8.98075044e-01 3.13018203e-01 -1.12314391e+00
-1.31701708e+00 -4.80338365e-01 9.13162470e-01 7.46117473e-01
2.07873270e-01 -2.44801477e-01 6.35789186e-02 2.75282323e-01
2.49734223e-01 -1.28059757e+00 -6.71997726e-01 -1.20359167e-01
2.63151288e-01 5.70576370e-01 1.23911309e+00 -2.14549765e-01
5.95684111e-01 -1.66543722e+00 3.38836342e-01 -2.36923546e-01
-4.64946985e-01 -1.81145757e-01 2.84576863e-01 8.95459890e-01
6.27023697e-01 -3.84197444e-01 8.98517296e-02 1.28784522e-01
-5.72905242e-01 5.22155464e-01 -3.47531736e-01 6.17079198e-01
-3.31490487e-01 1.75726846e-01 -1.11850011e+00 -2.63148457e-01
6.70942187e-01 -1.97100062e-02 -5.04020452e-01 1.74696609e-01
-2.14765415e-01 6.74615204e-01 -5.86211681e-01 9.47620690e-01
4.79594886e-01 2.95339435e-01 1.02760755e-01 2.33242270e-02
-1.15474010e+00 -1.77200630e-01 -1.06620610e+00 2.31177759e+00
-6.90326154e-01 5.61207652e-01 6.93129897e-01 -2.55247116e-01
9.94141400e-01 3.66014242e-02 1.66071132e-01 -1.98322728e-01
6.11098334e-02 2.91073650e-01 -4.32175189e-01 -4.20198977e-01
1.13779271e+00 -8.43169168e-02 -3.35469663e-01 4.21763249e-02
3.71079415e-01 -1.01188409e+00 2.91344374e-01 -1.15004659e-01
1.30937183e+00 5.64562201e-01 6.86895907e-01 -9.48499739e-01
3.15054148e-01 9.60479379e-01 2.79356092e-01 6.33977473e-01
-8.37991238e-02 2.28526294e-01 -7.87781402e-02 -7.80989945e-01
-9.98262346e-01 -8.18304121e-01 -2.53410846e-01 7.60446668e-01
8.85936558e-01 -1.12726495e-01 -3.32461655e-01 -4.84881222e-01
3.89161631e-02 6.13750041e-01 -4.29564476e-01 3.41938406e-01
-4.42090958e-01 -4.05212641e-01 3.20418000e-01 1.56312779e-01
6.27948582e-01 -4.82770264e-01 -2.31344867e+00 3.86429131e-01
2.13928446e-01 -1.26826966e+00 3.93378437e-01 2.96194702e-01
-9.83488739e-01 -1.46191609e+00 -4.14798409e-02 -3.14882159e-01
4.70769435e-01 9.11561787e-01 8.45937252e-01 -3.93460572e-01
-3.69928509e-01 1.03217876e+00 -7.84544647e-01 -3.76965284e-01
-1.95827991e-01 -4.00235146e-01 3.61196160e-01 -7.95471191e-01
-3.77388239e-01 -4.45668012e-01 -5.62690675e-01 7.63673782e-01
-6.85208321e-01 -4.38929312e-02 8.30342770e-01 6.64726913e-01
6.29945278e-01 3.11403811e-01 -4.67514724e-01 -2.16634780e-01
2.78454691e-01 -2.72903055e-01 -1.56841469e+00 2.05619663e-01
-6.12842679e-01 -1.42347962e-01 4.38240439e-01 2.10155964e-01
-9.12883699e-01 4.87369150e-01 2.32352719e-01 -5.26468456e-01
-2.72368640e-01 4.94212776e-01 3.41087908e-01 -9.37574685e-01
5.78872800e-01 -3.99537198e-02 -3.74038368e-02 -1.85708970e-01
2.46670589e-01 4.29389238e-01 6.38587058e-01 -7.37353787e-02
5.22150576e-01 1.18869340e+00 6.62037790e-01 -8.50443244e-01
-8.58197927e-01 -6.29910052e-01 -1.19853413e+00 -6.52647018e-01
9.26931918e-01 -1.26438653e+00 -3.40023726e-01 1.23402670e-01
-1.02247286e+00 -3.46253425e-01 -2.36780643e-01 5.30900776e-01
-1.00246263e+00 2.11267725e-01 9.33553651e-02 -9.21839833e-01
-2.48282701e-01 -1.53202879e+00 1.32300961e+00 2.14141339e-01
3.29127759e-01 -9.29175079e-01 5.33211827e-01 2.55314499e-01
2.88663059e-01 6.16474092e-01 4.34966460e-02 -2.10364506e-01
-1.04571807e+00 -2.80070677e-02 2.75722831e-01 -2.74793714e-01
-4.89154875e-01 -3.78557771e-01 -5.53501785e-01 -9.76579010e-01
7.05558732e-02 -5.23863733e-01 5.93425751e-01 4.81482655e-01
5.72452135e-02 -3.39171067e-02 -6.22751594e-01 7.94749737e-01
2.14039874e+00 4.55708563e-01 1.57759741e-01 9.45444107e-01
7.65568540e-02 7.90936887e-01 1.93607390e+00 9.05308962e-01
5.33246994e-01 7.90216148e-01 1.54402173e+00 5.46717405e-01
5.94473541e-01 -2.42500469e-01 3.88652027e-01 9.28874239e-02
-4.05026190e-02 -4.43076342e-01 -1.25352728e+00 8.07414770e-01
-1.59491348e+00 -5.23369670e-01 7.63428286e-02 2.09665179e+00
-3.53651255e-01 1.52092442e-01 -2.91371524e-01 -4.56862599e-01
4.21778828e-01 4.20979299e-02 -1.38229460e-01 -5.20456791e-01
1.40495598e-01 -2.83683389e-01 1.19150436e+00 7.57108688e-01
-1.31809866e+00 1.05584240e+00 7.03677416e+00 8.31835996e-03
-5.87481618e-01 1.48195386e-01 -5.03729343e-01 -1.44886538e-01
-6.79081073e-03 7.94711769e-01 -1.30783629e+00 -2.48155653e-01
1.05432546e+00 2.83976793e-01 5.57430506e-01 9.00503695e-01
-4.10287343e-02 -8.88165236e-01 -8.19906771e-01 4.62495387e-01
2.36528426e-01 -1.30820751e+00 -4.70451899e-02 4.40773576e-01
6.84580386e-01 6.14520133e-01 -2.18906164e-01 -1.23380773e-01
7.87590861e-01 -4.80197906e-01 1.16518295e+00 3.09309959e-01
6.35832369e-01 -7.88740098e-01 9.12601113e-01 4.72733915e-01
-1.33050740e+00 -8.89493704e-01 -5.68834364e-01 -2.98173934e-01
7.06938207e-01 -7.37865940e-02 -8.67463589e-01 1.17291939e+00
1.24686253e+00 3.52841288e-01 -3.89696836e-01 1.16385019e+00
-1.74343452e-01 -4.57238793e-01 -5.02458572e-01 9.40360278e-02
1.06795382e+00 -3.84773582e-01 1.03010154e+00 5.76146781e-01
7.48859942e-01 4.01425153e-01 6.96359992e-01 3.05793285e-01
9.20456946e-01 -5.23802042e-01 -1.19399142e+00 4.86842930e-01
4.79958504e-01 1.18127036e+00 -5.19060373e-01 -7.27566332e-02
-3.56715292e-01 5.25488913e-01 -1.39728859e-01 -3.93947154e-01
-5.41932166e-01 -3.18715602e-01 4.35702980e-01 -1.23054674e-03
4.97892022e-01 -6.16328657e-01 1.73163161e-01 -6.55663610e-01
-3.55588824e-01 -2.79641062e-01 3.78807425e-01 -1.09516072e+00
-2.91384131e-01 8.95383894e-01 6.69827044e-01 -1.81725228e+00
-4.20552492e-01 -5.57590663e-01 -3.30408514e-01 5.28387070e-01
-1.82875931e+00 -1.55541265e+00 -7.23300517e-01 2.30439916e-01
8.54578376e-01 -6.00436866e-01 7.86855221e-01 -5.48203170e-01
2.89918363e-01 -5.91686070e-01 8.25283304e-02 -7.25837767e-01
3.90127510e-01 -1.03725994e+00 3.37166429e-01 1.19923651e+00
-4.11493212e-01 7.82548189e-02 1.04492104e+00 -9.63802516e-01
-1.71780491e+00 -9.69611108e-01 -5.15440404e-02 -4.44062144e-01
6.59343064e-01 2.07249634e-02 6.51057288e-02 9.12132740e-01
4.33877617e-01 -2.56079316e-01 9.76483524e-02 -2.01603562e-01
6.63800657e-01 -4.59112488e-02 -1.32921100e+00 2.53369659e-01
8.47025514e-01 -1.75782397e-01 -7.78012931e-01 3.39167237e-01
6.97444499e-01 -8.46923530e-01 -1.03343976e+00 7.09871054e-01
8.20242941e-01 -1.35001314e+00 8.40311646e-01 8.56874213e-02
-8.81072693e-03 -5.00853479e-01 -5.34859717e-01 -1.57065749e+00
-2.33636320e-01 -7.09446847e-01 4.83786404e-01 4.12188441e-01
1.57888219e-01 -4.46581244e-01 6.05121434e-01 -4.28126842e-01
-1.04999077e+00 -3.80832940e-01 -1.16405404e+00 -8.91855538e-01
-4.74459678e-01 2.55018055e-01 2.13214889e-01 2.41745815e-01
-1.32241458e-01 1.18631832e-02 -4.01985973e-01 1.06489146e+00
5.98412573e-01 2.73845732e-01 6.90696061e-01 -1.45840514e+00
-2.04453468e-01 4.68322366e-01 -6.17409408e-01 -7.50418961e-01
-1.60917476e-01 -4.53685261e-02 1.95558041e-01 -1.42934597e+00
-3.71844232e-01 -2.83084303e-01 4.55422014e-01 1.45298511e-01
8.07842314e-01 -3.68598670e-01 1.52197033e-01 3.31084251e-01
-9.13865268e-01 3.53991181e-01 1.14978981e+00 3.03958327e-01
-2.66410530e-01 1.87732764e-02 -1.33713275e-01 7.40411639e-01
6.16020143e-01 -2.60772198e-01 -6.93494976e-01 -8.63468647e-01
2.67028153e-01 8.55351388e-01 4.48896140e-01 -1.65237856e+00
7.36262441e-01 -1.38843536e-01 -6.29823804e-02 -1.03472304e+00
8.56409371e-01 -1.18471837e+00 3.67250919e-01 8.61916244e-01
1.84373140e-01 8.44552696e-01 5.05588710e-01 8.27511966e-01
-4.55801219e-01 -5.42493582e-01 9.90671158e-01 -9.60228622e-01
-1.50009060e+00 2.68158689e-02 -5.81961870e-01 -1.13963768e-01
1.65093875e+00 -3.67457688e-01 -5.42010665e-01 -1.75077636e-02
-5.04856467e-01 6.40170097e-01 1.10293651e+00 2.31196612e-01
1.01451278e+00 -5.40570974e-01 -3.92370880e-01 4.60172564e-01
2.19132975e-01 2.67480254e-01 3.05472404e-01 5.89286864e-01
-1.40539300e+00 8.71389568e-01 -1.03214812e+00 -7.80861676e-01
-1.15378463e+00 4.55417335e-01 4.25359666e-01 -2.13927224e-01
-3.66205573e-01 6.95861161e-01 -4.28365469e-02 -6.39476597e-01
-4.56056088e-01 -3.06069814e-02 -3.43788624e-01 -2.04629704e-01
3.29538018e-01 7.86942959e-01 6.23650923e-02 -4.71867681e-01
-4.33952212e-01 9.22964334e-01 2.07588702e-01 -6.01627409e-01
1.36569762e+00 -6.89124107e-01 -4.92140725e-02 1.14963643e-01
2.36666575e-01 -3.50061059e-01 -1.68621004e+00 5.83263814e-01
-2.80465752e-01 -5.81626892e-01 6.23958886e-01 -7.47303486e-01
-3.66968721e-01 7.62705088e-01 8.92939329e-01 2.02524751e-01
1.05766094e+00 -1.81502804e-01 3.72342467e-01 6.91371977e-01
1.50997245e+00 -1.12172937e+00 -1.71714053e-02 8.44686508e-01
9.06463504e-01 -1.22077119e+00 4.58837211e-01 -3.49506706e-01
-9.85322177e-01 1.38800967e+00 6.51576579e-01 -3.72859120e-01
3.54191154e-01 3.61857921e-01 -6.98157237e-04 -7.76784420e-01
-1.02045274e+00 -3.05367172e-01 -3.29139560e-01 6.91379786e-01
-6.48854733e-01 -1.84727222e-01 4.40722674e-01 -3.40081006e-01
-2.05397218e-01 -3.88915449e-01 1.15052617e+00 1.68643594e+00
-1.11113048e+00 -7.12409317e-01 -4.77018505e-01 9.36782956e-02
-1.48092046e-01 8.76550227e-02 -4.35558856e-02 1.20869946e+00
2.78694212e-01 1.05030787e+00 4.46647517e-02 -5.12299240e-01
6.85186625e-01 -6.69647515e-01 6.30549729e-01 -6.57949686e-01
-6.03813827e-01 -1.68911949e-01 3.93913329e-01 -7.04283774e-01
-4.19757515e-01 -1.13282931e+00 -7.25805461e-01 1.98849112e-01
-6.85374498e-01 2.47458100e-01 7.67105460e-01 7.17144966e-01
3.10349375e-01 4.29730952e-01 6.20716929e-01 -1.08368814e+00
-7.20344424e-01 -6.56792462e-01 -7.00305641e-01 -7.94222593e-01
5.29034853e-01 -1.10610032e+00 -2.13282928e-01 -3.55053127e-01] | [7.347779750823975, -1.901630163192749] |
1beb54cd-710f-40bf-b45a-845cea9e936e | evaluating-the-text-to-sql-capabilities-of-1 | 2204.00498 | null | https://arxiv.org/abs/2204.00498v1 | https://arxiv.org/pdf/2204.00498v1.pdf | Evaluating the Text-to-SQL Capabilities of Large Language Models | We perform an empirical evaluation of Text-to-SQL capabilities of the Codex language model. We find that, without any finetuning, Codex is a strong baseline on the Spider benchmark; we also analyze the failure modes of Codex in this setting. Furthermore, we demonstrate on the GeoQuery and Scholar benchmarks that a small number of in-domain examples provided in the prompt enables Codex to perform better than state-of-the-art models finetuned on such few-shot examples. | ['Dzmitry Bahdanau', 'Raymond Li', 'Nitarshan Rajkumar'] | 2022-03-15 | null | null | null | null | ['text-to-sql'] | ['computer-code'] | [-6.21415079e-01 -1.01846218e-01 -9.04360235e-01 -4.68051553e-01
-1.11700523e+00 -7.89831996e-01 8.36117089e-01 1.25847280e-01
-1.01824112e-01 3.50733161e-01 4.69213307e-01 -8.20769966e-01
-2.43724406e-01 -6.92411780e-01 -9.27182853e-01 2.68401176e-01
-1.76481470e-01 5.42299688e-01 5.52412570e-01 -5.61274171e-01
3.31981868e-01 -6.94832280e-02 -1.74268723e+00 5.70320725e-01
1.06672513e+00 5.20507753e-01 -1.01848535e-01 7.28597522e-01
-5.47353327e-01 1.04990578e+00 -6.76715314e-01 -5.07335067e-01
2.85425037e-01 3.18212360e-01 -9.24071074e-01 -6.47206247e-01
6.82023585e-01 -5.44418573e-01 -3.90060574e-01 8.60986710e-01
2.64172256e-01 3.10237035e-02 2.31307343e-01 -9.69214737e-01
-6.93291128e-01 1.27833939e+00 -5.11358142e-01 3.34427685e-01
4.49728906e-01 6.25649452e-01 1.16250241e+00 -9.88775194e-01
9.26295817e-01 1.00546658e+00 7.25080490e-01 3.42401356e-01
-1.66221440e+00 -4.19527411e-01 -3.18332732e-01 3.45028602e-02
-1.09269357e+00 -5.68211257e-01 2.38345250e-01 -3.87182117e-01
1.83006477e+00 4.26398069e-01 1.22271299e-01 1.40938008e+00
2.96874642e-01 5.82897604e-01 9.30656135e-01 -3.37430328e-01
1.82916090e-01 -4.68712591e-04 5.70793748e-01 7.91545749e-01
-2.23419107e-02 5.00553489e-01 -5.86105764e-01 -6.13348365e-01
9.89255458e-02 2.13564858e-02 2.32997108e-02 -3.16612870e-01
-9.34707105e-01 8.85160685e-01 2.96835572e-01 7.11962134e-02
2.83520040e-03 4.95902896e-01 1.03292358e+00 5.69361925e-01
2.52455890e-01 9.25829530e-01 -8.57920408e-01 -1.01688612e+00
-1.16275442e+00 2.46886820e-01 1.39656067e+00 1.49492669e+00
7.70178258e-01 -2.09151164e-01 -4.03948456e-01 8.36777806e-01
-2.25063145e-01 5.56112289e-01 5.36510110e-01 -1.17467833e+00
7.78314888e-01 8.57875407e-01 -1.87973574e-01 -3.73855710e-01
-1.18361950e-01 -3.40590358e-01 4.86830510e-02 3.58535320e-01
2.51947522e-01 -3.26093175e-02 -7.13503838e-01 1.33053529e+00
-1.53661305e-02 -2.90253103e-01 -1.18978687e-01 3.92424375e-01
7.46758163e-01 5.65888941e-01 1.33295789e-01 -3.21181267e-02
1.02062464e+00 -1.05967569e+00 -2.34176800e-01 -5.21014333e-01
8.68469477e-01 -6.89375579e-01 2.09353042e+00 3.09777051e-01
-1.08074403e+00 -5.04336476e-01 -1.04100370e+00 -1.32164687e-01
-3.44183266e-01 -4.81072739e-02 8.94952893e-01 3.11856091e-01
-7.32055783e-01 6.96684599e-01 -1.09658182e+00 -8.77474248e-01
1.77330598e-01 -1.75137028e-01 1.83952041e-02 -6.64266050e-02
-7.80310631e-01 1.00659049e+00 4.00100201e-01 -1.01202512e+00
-1.20434082e+00 -1.44067264e+00 -4.12640780e-01 4.12043929e-01
1.03197789e+00 -3.88085693e-01 1.68463910e+00 -2.88766682e-01
-1.05359852e+00 7.48500884e-01 1.99780330e-01 -5.49219370e-01
4.03732598e-01 -7.04484224e-01 -3.73614341e-01 1.51720494e-02
1.69507876e-01 2.55657256e-01 4.28632706e-01 -9.99307692e-01
-4.23008412e-01 -1.33953840e-01 5.39814949e-01 -5.02112150e-01
-3.93764138e-01 2.00719282e-01 -7.25539148e-01 -5.44439435e-01
-8.50250959e-01 -5.85537553e-01 -2.00639549e-03 -3.52810100e-02
-5.90117693e-01 -3.26255977e-01 6.67915761e-01 -3.67477119e-01
1.94612384e+00 -2.11883879e+00 -8.30131471e-02 1.04899630e-01
1.51020080e-01 1.40837014e-01 -5.36356390e-01 9.36358154e-01
9.05408291e-04 5.26534796e-01 -5.64674921e-02 1.49224550e-01
3.05080473e-01 3.16302419e-01 -8.23729873e-01 -8.96338001e-03
-8.61971304e-02 1.11633289e+00 -7.98996985e-01 -6.65496767e-01
-3.54126990e-02 -2.62409955e-01 -9.45636868e-01 3.80328536e-01
-1.02827954e+00 -6.03321791e-01 -4.87966180e-01 9.32026267e-01
2.29008779e-01 -5.84417045e-01 6.37241527e-02 8.43853578e-02
-2.30644122e-01 5.33639908e-01 -5.11104524e-01 2.10285234e+00
-7.15716898e-01 2.67076284e-01 -1.49420589e-01 -4.46837425e-01
5.43459237e-01 -1.95203707e-01 1.72621101e-01 -8.96815598e-01
-4.41686898e-01 2.70390540e-01 -1.96280941e-01 -1.01570523e+00
5.04582763e-01 3.63602161e-01 -5.20162582e-01 8.89080942e-01
2.15625793e-01 -2.16839001e-01 6.37944520e-01 6.51490450e-01
1.91854024e+00 3.20865929e-01 4.45460171e-01 -4.88957465e-01
-5.78188635e-02 5.49263537e-01 -7.63922930e-03 1.33145988e+00
2.65340924e-01 3.63918364e-01 7.42498100e-01 -5.11067569e-01
-1.15566468e+00 -1.11531746e+00 -3.81030768e-01 1.92206812e+00
-2.60858715e-01 -1.25924563e+00 -6.98110938e-01 -1.19046831e+00
3.81881714e-01 1.41489613e+00 -7.26684630e-01 -7.02264383e-02
-5.04068375e-01 -5.60499728e-01 7.53099680e-01 7.18114316e-01
1.93155825e-01 -8.73268723e-01 -8.69360864e-01 3.30491424e-01
2.24442020e-01 -6.40715063e-01 -5.74494421e-01 5.53918719e-01
-7.90065348e-01 -1.40717578e+00 2.08949625e-01 -7.57353008e-02
-1.91513062e-01 -1.63503308e-02 2.13281107e+00 5.37540913e-01
-6.41993821e-01 4.94551361e-01 -5.90931475e-01 -1.32715270e-01
-8.43477130e-01 7.68921494e-01 -5.78389406e-01 -1.12567604e+00
7.63918221e-01 -8.15737665e-01 -3.95326465e-01 2.69847363e-01
-1.01482344e+00 -2.93689609e-01 3.98271561e-01 8.19862247e-01
2.77877450e-01 -4.32187378e-01 2.32151613e-01 -1.37385058e+00
8.42170417e-01 -7.15554297e-01 -8.85161638e-01 6.69291139e-01
-1.00608659e+00 4.12107468e-01 6.93666101e-01 -4.01467681e-01
-9.61117506e-01 -2.88303643e-01 -2.41546601e-01 -2.25714713e-01
3.65954116e-02 9.49598908e-01 2.57231712e-01 2.68323459e-02
1.64383113e+00 1.44031674e-01 2.62852088e-02 -8.06333899e-01
8.69381547e-01 4.60370809e-01 6.41223669e-01 -1.13428581e+00
7.92552531e-01 2.44579613e-01 -4.25971180e-01 -3.52555871e-01
-6.98166549e-01 -2.10462779e-01 -1.25216544e-01 2.50449210e-01
3.94667417e-01 -6.60727203e-01 -4.71772045e-01 -3.43281597e-01
-9.08973336e-01 -7.56752610e-01 -6.08704090e-01 -2.15025872e-01
-7.72866488e-01 1.45992383e-01 -7.59195924e-01 -1.46915182e-01
-3.62924039e-01 -1.12899506e+00 1.05542922e+00 -1.05598129e-01
-4.67624336e-01 -8.55337024e-01 5.02409995e-01 -7.19183162e-02
7.92006195e-01 -1.37994513e-01 1.68809724e+00 -1.00587392e+00
-7.32947767e-01 -1.57677144e-01 -2.03810543e-01 -6.26915172e-02
-3.12910080e-01 4.51592028e-01 -5.88116288e-01 -1.67817518e-01
-4.71725389e-02 -7.61656702e-01 9.09971774e-01 -1.00436524e-01
1.22560298e+00 -4.84051645e-01 -5.55168450e-01 8.69199753e-01
1.80335557e+00 -2.79319286e-01 5.02040565e-01 5.33264756e-01
3.03942442e-01 2.07178518e-01 7.58368731e-01 7.43983328e-01
1.71673223e-01 5.77239990e-01 6.51613295e-01 3.81804913e-01
-1.44954130e-01 -3.26363117e-01 3.12705636e-01 4.61534619e-01
4.63956028e-01 6.02894984e-02 -1.51817155e+00 7.36841440e-01
-1.93214715e+00 -9.00437593e-01 -4.37822118e-02 1.96540856e+00
1.33919394e+00 2.20343962e-01 1.38617307e-01 -5.65995455e-01
1.27011403e-01 2.44127423e-01 -6.66379988e-01 -3.36999893e-01
4.50919122e-02 4.98452038e-01 4.41635072e-01 3.23422223e-01
-7.91989446e-01 1.03051019e+00 8.16521835e+00 1.00607324e+00
-9.43731725e-01 1.93520218e-01 1.96428046e-01 -6.04115963e-01
-6.78379655e-01 3.17757964e-01 -9.05729175e-01 6.45884812e-01
1.33661461e+00 -5.45988083e-01 9.00533557e-01 1.59976745e+00
-3.46228480e-01 -1.16356798e-01 -1.55647802e+00 2.92295665e-01
-2.61705458e-01 -1.71303856e+00 -5.09538502e-02 -1.34990141e-01
7.49575078e-01 8.06172431e-01 -1.50851622e-01 1.08972573e+00
9.16117609e-01 -9.55724299e-01 5.19175649e-01 5.30413687e-01
9.59638178e-01 -3.92638952e-01 1.97943911e-01 4.01844770e-01
-5.78378201e-01 -4.39942420e-01 -5.43219030e-01 3.08635354e-01
-1.05495252e-01 4.96121913e-01 -8.32783401e-01 2.00327381e-01
9.24736083e-01 4.75554913e-01 -1.36267328e+00 9.96454835e-01
-3.58159356e-02 8.30977798e-01 -2.74281085e-01 -8.75755474e-02
9.95539594e-03 4.12983000e-01 5.51438272e-01 1.37796462e+00
3.18851024e-01 -2.03460634e-01 2.68001437e-01 1.30063462e+00
-2.20345676e-01 7.26858005e-02 -7.08361864e-01 -2.57312238e-01
7.43775964e-01 1.02595317e+00 1.08616345e-01 -6.32173121e-01
-9.58914220e-01 2.24184170e-01 7.06095815e-01 4.53353763e-01
-8.58250380e-01 -5.54215372e-01 6.37274802e-01 4.59976792e-01
5.10873854e-01 1.23452865e-01 -4.52563643e-01 -1.26581144e+00
1.18537188e-01 -1.41393197e+00 6.25837743e-01 -9.18588877e-01
-1.53471649e+00 4.47724402e-01 3.63596380e-01 -8.26252341e-01
-6.42871439e-01 -4.34262961e-01 -8.28143179e-01 5.22048235e-01
-1.16308451e+00 -9.42310989e-01 -1.42969579e-01 3.09556961e-01
6.10631108e-01 -4.63205397e-01 9.33554053e-01 1.36981830e-01
-1.73681155e-01 5.67033887e-01 4.69254553e-01 -6.57125041e-02
9.05297518e-01 -1.48718333e+00 8.56261969e-01 8.23730350e-01
-2.72573438e-02 1.29294205e+00 9.34900284e-01 -7.92689860e-01
-1.86597621e+00 -1.07591534e+00 4.31645453e-01 -1.04449904e+00
1.22240567e+00 -3.81732345e-01 -1.13349390e+00 9.67706263e-01
5.95779955e-01 1.93869725e-01 4.18371081e-01 7.17265785e-01
-9.29603696e-01 -3.44138443e-01 -7.17680991e-01 4.70407993e-01
9.88133132e-01 -8.45682800e-01 -9.18607652e-01 5.02670348e-01
1.01609254e+00 -4.81747448e-01 -1.19776952e+00 2.84054101e-01
4.41730946e-01 -1.12440276e+00 1.00216472e+00 -1.11136746e+00
1.13371730e+00 5.62452935e-02 -3.71614397e-01 -1.08934414e+00
-1.44785438e-02 -6.41728461e-01 -5.12890100e-01 1.31897938e+00
5.76627970e-01 -1.87992752e-01 6.46294951e-01 7.42128432e-01
-2.45253563e-01 -6.26151681e-01 -7.05110192e-01 -1.12050414e+00
2.95832843e-01 -4.88655776e-01 7.98127532e-01 4.71938729e-01
4.40710813e-01 3.27641696e-01 -7.67019689e-02 -5.14970601e-01
3.60822946e-01 5.62230349e-01 1.17213356e+00 -9.15899396e-01
-9.43366408e-01 -4.12312478e-01 3.50926250e-01 -9.85105217e-01
2.23978534e-01 -9.95526254e-01 -5.03683835e-02 -1.07636905e+00
6.75859153e-01 -3.06164503e-01 -7.30384439e-02 6.78388000e-01
-3.23768169e-01 -3.41177434e-01 1.65886402e-01 2.92087376e-01
-9.30744052e-01 2.47447118e-01 5.98540723e-01 -2.95784652e-01
-2.08768733e-02 -1.56021670e-01 -8.13563108e-01 2.78992027e-01
3.51577491e-01 -6.88497841e-01 -4.44805741e-01 -7.23331153e-01
7.05894172e-01 4.14967179e-01 1.89791098e-01 -8.66851151e-01
3.04231882e-01 -3.23847294e-01 -3.07523906e-01 -4.97398734e-01
-9.59518403e-02 -5.25030315e-01 3.00199492e-03 3.56441051e-01
-6.05399668e-01 3.04131091e-01 4.22305465e-01 3.71379226e-01
-7.94520900e-02 -6.24578059e-01 6.22879446e-01 -2.75910467e-01
-7.38834798e-01 1.14478022e-01 -1.11907296e-01 9.23213303e-01
7.37638533e-01 3.18068057e-01 -1.13365149e+00 -9.60482210e-02
-2.48470485e-01 1.77252606e-01 9.19916153e-01 5.96786916e-01
1.44574508e-01 -1.08726943e+00 -4.46647048e-01 3.71384025e-02
8.14762831e-01 -6.50865138e-01 -2.00372845e-01 6.43603265e-01
-6.37031794e-01 3.97423536e-01 9.08073783e-03 -5.31192839e-01
-1.00665903e+00 1.08961725e+00 3.99062037e-02 -4.19278979e-01
-4.81908202e-01 5.07992804e-01 -1.12925172e-01 -6.41365886e-01
3.42363358e-01 -4.72208619e-01 3.72557223e-01 -3.25356036e-01
4.70820278e-01 3.53526533e-01 1.53875932e-01 4.79680270e-01
-3.55233073e-01 1.28019884e-01 -2.57694930e-01 7.00615719e-02
1.62336934e+00 3.08892906e-01 -2.18583643e-01 4.52309310e-01
1.14554751e+00 3.22841913e-01 -1.13497269e+00 -2.29615599e-01
3.39470685e-01 -5.30564070e-01 -2.75368005e-01 -1.38041985e+00
-7.72526979e-01 8.03604960e-01 1.98910609e-01 4.74730283e-01
8.25776339e-01 1.80360883e-01 4.23776895e-01 8.02087188e-01
4.81926918e-01 -1.18636143e+00 1.39117464e-01 7.01440454e-01
9.91004288e-01 -1.22383332e+00 2.21259087e-01 1.62273441e-02
-3.87381881e-01 1.25794876e+00 9.06640470e-01 -1.77323893e-01
6.79151535e-01 8.32385361e-01 -6.80855587e-02 -4.09245700e-01
-1.55361378e+00 1.44940212e-01 -1.59670129e-01 3.97175968e-01
4.65500772e-01 -2.94094443e-01 -2.85364896e-01 6.65881217e-01
-2.10705101e-01 2.99826890e-01 5.41583180e-01 1.19345176e+00
-4.32473242e-01 -1.28855467e+00 -1.34650812e-01 8.28499734e-01
-6.62149370e-01 -6.08137488e-01 -2.32097447e-01 1.30250943e+00
-5.79108775e-01 5.87879717e-01 -1.52704865e-01 -4.84708488e-01
5.24127722e-01 2.11482748e-01 2.85885125e-01 -9.42412913e-01
-1.27101886e+00 -2.36517519e-01 4.44004118e-01 -1.12716734e+00
5.29969692e-01 -3.92209709e-01 -9.53170359e-01 -6.83397889e-01
-4.19063494e-02 1.97440788e-01 5.22928178e-01 4.31955397e-01
7.77524352e-01 3.84121090e-01 1.85986847e-01 -1.04917817e-01
-1.42261529e+00 -9.60788608e-01 -2.82957673e-01 5.34318030e-01
5.72879873e-02 -2.39823163e-01 -3.56466621e-01 7.76364841e-03] | [9.697219848632812, 7.863587379455566] |
fe79086d-1e63-4138-8f57-83fd9547bef1 | adaptive-services-function-chain | 2304.12853 | null | https://arxiv.org/abs/2304.12853v1 | https://arxiv.org/pdf/2304.12853v1.pdf | Adaptive Services Function Chain Orchestration For Digital Health Twin Use Cases: Heuristic-boosted Q-Learning Approach | Digital Twin (DT) is a prominent technology to utilise and deploy within the healthcare sector. Yet, the main challenges facing such applications are: Strict health data-sharing policies, high-performance network requirements, and possible infrastructure resource limitations. In this paper, we address all the challenges by provisioning adaptive Virtual Network Functions (VNFs) to enforce security policies associated with different data-sharing scenarios. We define a Cloud-Native Network orchestrator on top of a multi-node cluster mesh infrastructure for flexible and dynamic container scheduling. The proposed framework considers the intended data-sharing use case, the policies associated, and infrastructure configurations, then provision Service Function Chaining (SFC) and provides routing configurations accordingly with little to no human intervention. Moreover, what is \textit{optimal} when deploying SFC is dependent on the use case itself, and we tune the hyperparameters to prioritise resource utilisation or latency in an effort to comply with the performance requirements. As a result, we provide an adaptive network orchestration for digital health twin use cases, that is policy-aware, requirements-aware, and resource-aware. | ['Paola Grosso', 'Arie Taal', 'Li Zhong', 'Jamila Alsayed Kassem'] | 2023-04-25 | null | null | null | null | ['q-learning'] | ['methodology'] | [-3.20334822e-01 5.30255213e-02 -3.97338748e-01 -2.71644026e-01
6.35793030e-01 -6.23733282e-01 7.31364638e-02 2.46622413e-01
-1.36714414e-01 5.31471252e-01 -1.49578124e-01 -6.46349192e-01
-9.20931518e-01 -8.10150921e-01 3.43609244e-01 -7.47060299e-01
-2.19865561e-01 1.02014124e+00 4.34089839e-01 -1.71513498e-01
7.32699558e-02 9.06494856e-01 -1.45328593e+00 -4.06221189e-02
5.79760075e-01 1.40662396e+00 1.90907478e-01 4.01258379e-01
-3.06935668e-01 2.92524904e-01 -7.94764280e-01 -6.00661747e-02
4.22657937e-01 1.85226053e-01 -7.20434427e-01 2.55770355e-01
-7.30327964e-01 -4.50164229e-02 4.02598947e-01 4.48619395e-01
6.32342398e-01 -6.40201047e-02 9.69801471e-02 -1.76259172e+00
1.24177530e-01 7.04413056e-01 -7.87768662e-02 5.28349102e-01
5.80699928e-02 1.46943271e-01 3.76255363e-01 1.96116403e-01
9.51166511e-01 6.17145181e-01 7.17680603e-02 4.89534706e-01
-1.06753826e+00 -5.15264332e-01 3.96201044e-01 2.44498283e-01
-1.31029308e+00 -4.51593429e-01 4.87541199e-01 -2.16681838e-01
5.99661052e-01 6.73881292e-01 5.00010431e-01 6.68047726e-01
-6.47381172e-02 -3.25261563e-01 7.03936040e-01 -3.94861162e-01
6.73715115e-01 2.73886055e-01 -3.95753056e-01 -1.34352148e-01
1.95742160e-01 -4.93064493e-01 7.50175118e-02 -5.36889613e-01
8.29535007e-01 -8.16341564e-02 -4.68243301e-01 -3.33604634e-01
-1.06518376e+00 1.36446103e-01 -5.37179746e-02 6.02313876e-01
-1.08284831e+00 -2.65774071e-01 8.69173944e-01 2.15318903e-01
3.34110335e-02 3.43968749e-01 -8.99633467e-01 -1.51463374e-01
-8.59807074e-01 -2.01476350e-01 1.02921641e+00 1.16373789e+00
2.14450315e-01 -2.33198881e-01 -1.59297004e-01 5.20659685e-01
-8.85068849e-02 2.12983221e-01 1.10761911e-01 -1.18233824e+00
-1.30964056e-01 3.93916011e-01 5.52120656e-02 -6.29968405e-01
-9.70447779e-01 -4.80399430e-01 -1.00813043e+00 -2.70435940e-02
-1.05968691e-01 -4.00388330e-01 -3.49910855e-01 1.68833578e+00
1.14724541e+00 2.94027418e-01 -1.48820579e-01 9.38178360e-01
1.16845950e-01 2.60768950e-01 3.40982258e-01 -9.56589818e-01
1.56029046e+00 -4.19008374e-01 -1.03576207e+00 6.68035626e-01
5.41752934e-01 -8.91423285e-01 5.93052328e-01 1.75352737e-01
-9.03678358e-01 -9.84367579e-02 -6.52323246e-01 7.31338739e-01
-2.60319263e-01 -4.33150411e-01 1.58407271e-01 9.83417690e-01
-1.14487791e+00 1.80994540e-01 -6.92388415e-01 -7.93089688e-01
-3.32446069e-01 5.20959437e-01 8.28671753e-02 1.77300751e-01
-9.22742367e-01 6.85789406e-01 4.53166753e-01 -9.86788720e-02
-4.49832350e-01 -8.40099514e-01 -1.06405377e-01 -1.32874744e-02
7.01873064e-01 -9.23232794e-01 9.73414123e-01 -5.77115238e-01
-1.37527311e+00 1.63016826e-01 6.94175780e-01 -6.95601478e-03
5.90088010e-01 5.76026201e-01 -1.09026837e+00 4.52880919e-01
-2.00762168e-01 -3.05267692e-01 1.36060193e-01 -1.20463026e+00
-7.88386643e-01 -9.31557640e-02 4.23745841e-01 1.28785729e-01
-4.67252702e-01 6.20325267e-01 -4.98677582e-01 -1.03225112e-01
-1.89291894e-01 -7.52812743e-01 -4.99489218e-01 -1.08726420e-01
-5.15861869e-01 1.02529362e-01 8.89177144e-01 4.55836914e-02
1.44223535e+00 -2.08775830e+00 -1.46607146e-01 9.34118211e-01
1.24773182e-01 3.08829248e-01 6.96895644e-02 6.54259264e-01
2.99157798e-01 1.49222761e-01 3.38120550e-01 8.22470337e-02
-8.94413888e-02 6.85204089e-01 3.35955560e-01 2.36657664e-01
-5.21990240e-01 -3.25892687e-01 -5.33718348e-01 -6.63233995e-01
2.24849239e-01 5.92728734e-01 -4.94927168e-01 4.65730578e-01
-5.08456767e-01 7.19977677e-01 -8.30590904e-01 7.60215044e-01
9.81908083e-01 -3.73292148e-01 1.17233407e+00 -4.84976977e-01
-5.91584444e-01 -2.11731493e-01 -1.45259070e+00 1.22629297e+00
-7.84661949e-01 -4.20300931e-01 9.81015205e-01 -8.92933369e-01
6.78125501e-01 6.49709582e-01 1.22230053e+00 -4.94077593e-01
3.23770612e-01 3.25190037e-01 -2.64481492e-02 -9.05979753e-01
2.60540336e-01 1.93538710e-01 1.88005716e-01 5.13978720e-01
-5.39713383e-01 7.55738437e-01 6.50915951e-02 2.11304124e-03
1.02824843e+00 -1.52843222e-01 1.11999057e-01 -5.12477756e-01
3.74669254e-01 -2.07712770e-01 9.48335230e-01 2.46596009e-01
-3.42928290e-01 9.05499458e-02 7.56574690e-01 -4.69991863e-01
-8.13756287e-01 -7.60612905e-01 -3.11151475e-01 1.05320823e+00
1.98889226e-01 -2.16512769e-01 -6.61002696e-01 -4.71461087e-01
-2.61928827e-01 6.77882433e-01 -1.80475459e-01 3.13754976e-01
-6.86511934e-01 -6.34159684e-01 3.01063210e-01 -2.92425621e-02
2.20391192e-02 -8.95823181e-01 -1.23208404e+00 7.50545084e-01
-1.35738403e-01 -1.61096728e+00 -6.42367601e-01 -9.15206298e-02
-2.75141269e-01 -1.51399374e+00 -2.16094125e-02 -2.48858973e-01
3.74774128e-01 5.84047250e-02 1.08825529e+00 5.06258905e-01
-5.11822939e-01 5.65010130e-01 -3.70794743e-01 3.15189548e-02
-2.93693215e-01 2.15296410e-02 2.37270921e-01 3.33578944e-01
-2.13292509e-01 -9.00557399e-01 -1.03893745e+00 6.79534316e-01
-1.19483113e+00 -1.81730032e-01 8.32125023e-02 -1.49271740e-02
9.07707036e-01 3.57687563e-01 8.76380265e-01 -1.02230263e+00
5.89607716e-01 -8.63001764e-01 -5.41593373e-01 4.86111313e-01
-9.49105382e-01 -3.58010679e-01 9.93878543e-01 -1.52899429e-01
-7.40909517e-01 -6.45434916e-01 -5.94019284e-03 -6.20164454e-01
-2.29361951e-01 2.24649817e-01 -2.87208021e-01 2.22008601e-01
2.11283773e-01 -2.71132976e-01 1.33110166e-01 -4.24903482e-01
9.92341563e-02 7.92053640e-01 -4.09465209e-02 -7.28918433e-01
3.26271385e-01 6.56645298e-01 2.04993248e-01 -3.50928962e-01
-2.24306077e-01 -3.31537694e-01 -2.25357890e-01 -4.13671225e-01
6.78582788e-01 -2.99359918e-01 -1.27569473e+00 -2.97233641e-01
-8.41536343e-01 -2.45882064e-01 -5.26097380e-02 3.89018804e-01
-3.33961993e-01 2.82115769e-02 -3.50771546e-01 -8.38471532e-01
-7.67867506e-01 -1.20015323e+00 2.53385216e-01 1.37631431e-01
-8.34945962e-02 -8.10531378e-01 -2.95280516e-01 4.07919250e-02
1.21619725e+00 8.05494666e-01 9.07262325e-01 -5.88188887e-01
-3.40523690e-01 4.54771549e-01 -3.58517885e-01 1.17854103e-01
2.80080825e-01 5.55522084e-01 -1.87525928e-01 -6.83085203e-01
-1.59659371e-01 2.65447527e-01 -3.96515697e-01 3.31764609e-01
1.66301167e+00 -7.13303864e-01 -4.21680361e-01 8.52085888e-01
1.74395704e+00 2.79482126e-01 2.65300572e-01 5.11184871e-01
3.06597292e-01 9.51432526e-01 7.52018869e-01 1.40700018e+00
7.47631788e-01 9.58818138e-01 1.17372751e+00 2.78074704e-02
2.61117488e-01 4.48479563e-01 -1.50136366e-01 7.02681541e-01
-8.63086134e-02 -5.44176579e-01 -7.96965778e-01 1.40062049e-01
-1.77313793e+00 -5.21064937e-01 2.33476937e-01 2.18019605e+00
3.02597165e-01 2.70826668e-01 4.11059469e-01 1.18440680e-01
8.00147414e-01 -7.59500191e-02 -5.51023185e-01 -7.54776955e-01
3.53432208e-01 6.26770258e-02 5.21012366e-01 -3.28439593e-01
-3.13529789e-01 3.18815440e-01 4.95288944e+00 4.18762863e-01
-1.39081991e+00 2.98700243e-01 2.12835118e-01 -2.17941284e-01
-3.78328323e-01 -6.18973523e-02 -4.23833020e-02 8.10095668e-01
1.40869033e+00 -1.84952319e-01 5.71775258e-01 6.83939457e-01
9.81516480e-01 3.68051291e-01 -4.42573577e-01 6.13636911e-01
-6.36505544e-01 -1.35096633e+00 -1.74622536e-01 2.64871418e-01
2.55004674e-01 3.71268094e-02 -4.82964039e-01 -4.48545992e-01
-2.24018067e-01 -4.39740062e-01 5.64575076e-01 4.20076936e-01
1.00387549e+00 -1.02700055e+00 6.92328095e-01 1.17011808e-01
-1.14972663e+00 -3.79407644e-01 1.81703512e-02 3.90347809e-01
6.41307354e-01 8.73658359e-01 -5.30200481e-01 8.40782404e-01
8.31037641e-01 -1.82174161e-01 3.09450477e-01 1.05036008e+00
4.16768044e-01 9.59960520e-02 -3.92338067e-01 3.46257776e-01
-1.91531867e-01 -5.29808737e-02 4.10208106e-01 1.02574122e+00
3.01768810e-01 4.21064347e-01 7.64616609e-01 4.84941959e-01
7.16297105e-02 4.06576723e-01 -1.27168810e-02 3.56291622e-01
1.34132290e+00 1.53443956e+00 -8.97523582e-01 -1.10816844e-01
-3.36621672e-01 4.50906873e-01 -1.47243321e-01 3.54871869e-01
-7.29398787e-01 -3.88979226e-01 1.53002954e+00 8.23691726e-01
2.70106420e-02 -7.12869987e-02 -1.04000561e-01 -5.50870717e-01
-7.04691559e-02 -7.84508646e-01 7.52448976e-01 -3.78667861e-01
-9.22636807e-01 1.11937225e+00 -1.22023098e-01 -1.29878020e+00
2.21883938e-01 -2.38293529e-01 -5.50638378e-01 4.69051838e-01
-1.70052564e+00 -9.41804051e-01 -2.67171830e-01 8.87342036e-01
-7.20145255e-02 1.65697753e-01 1.17368829e+00 8.84498060e-01
-8.57349038e-01 5.82297206e-01 -1.12436175e-01 -6.30743146e-01
4.37433958e-01 -6.99595213e-01 -4.10286844e-01 6.94396436e-01
-9.62701738e-01 4.19563562e-01 7.39288688e-01 -1.96057439e-01
-1.24763227e+00 -1.02981579e+00 3.85500461e-01 4.89507109e-01
3.15715909e-01 -1.65503889e-01 -6.68112397e-01 2.81971633e-01
1.86115786e-01 3.99176955e-01 1.11437607e+00 -2.11431906e-01
2.46398430e-02 -5.30637085e-01 -1.88941443e+00 2.30073422e-01
8.64295423e-01 -5.68790408e-03 7.32792079e-01 5.39201677e-01
9.38018978e-01 -2.52539992e-01 -1.29984784e+00 4.64291632e-01
3.04770261e-01 -8.89563322e-01 6.40147686e-01 -6.31536424e-01
-6.48000896e-01 -5.24840057e-01 -3.75273317e-01 -1.07874715e+00
-3.90748799e-01 -8.73448670e-01 2.21775621e-01 1.15304303e+00
2.13359058e-01 -8.63674760e-01 5.87914705e-01 8.18710864e-01
-4.68800396e-01 -1.06502056e+00 -1.26325941e+00 -7.25846231e-01
-6.59416080e-01 -1.49226159e-01 1.47914875e+00 1.29910266e+00
-5.29483259e-02 -4.07443404e-01 -1.39239788e-01 7.05570161e-01
6.67526007e-01 1.86229661e-01 4.51837242e-01 -1.09445059e+00
-3.22658330e-01 -2.69543976e-01 -1.78298146e-01 1.77628081e-02
-2.71424562e-01 -4.40401524e-01 -4.84146714e-01 -1.56627560e+00
-5.20838380e-01 -1.39748347e+00 -6.96209133e-01 4.28674370e-01
5.34476876e-01 -3.75991851e-01 2.59915292e-01 2.55274922e-01
-5.91412365e-01 -1.14796318e-01 1.04259467e+00 5.05456746e-01
-4.26437289e-01 1.54782936e-01 -4.53689098e-01 1.24433339e-01
1.32377291e+00 -3.48115653e-01 -9.72941875e-01 -4.52194601e-01
4.06400524e-02 7.78697789e-01 -1.47270665e-01 -7.93920398e-01
4.09369886e-01 -9.75198805e-01 -4.97813880e-01 -1.12814911e-01
-4.72652987e-02 -1.47179735e+00 1.08489692e+00 7.19374120e-01
-3.94333526e-02 4.73171353e-01 -1.93136543e-01 2.57076800e-01
2.91992456e-01 4.36345965e-01 8.48226130e-01 3.87000442e-02
-2.89089501e-01 7.39419997e-01 -3.60246956e-01 -1.20074935e-01
1.41999424e+00 -1.44880027e-01 -3.93226266e-01 2.12093487e-01
-9.33100462e-01 6.97948337e-01 5.01619935e-01 3.52467060e-01
1.94075093e-01 -6.69672728e-01 -5.57979047e-01 -8.25262368e-02
1.83594838e-01 -2.25038618e-01 7.30804086e-01 9.78591800e-01
-5.17440975e-01 3.13962735e-02 -5.56455493e-01 -2.27132991e-01
-1.05581295e+00 9.19655025e-01 5.34749866e-01 -2.30888948e-01
-4.18306708e-01 2.60580093e-01 -3.63000840e-01 -3.52862328e-01
3.04899782e-01 1.38104260e-01 -3.47903937e-01 3.39575745e-02
4.60081279e-01 8.00760090e-01 2.95947403e-01 -3.61634910e-01
-6.75688088e-01 1.71272919e-01 2.66219616e-01 2.34097466e-01
1.11223686e+00 -6.79535866e-01 -7.10784256e-01 -3.86316061e-01
6.41045213e-01 -2.67046511e-01 -5.42381048e-01 -2.03697428e-01
-6.95026368e-02 -3.19422543e-01 1.28937006e-01 -8.44962597e-01
-1.73508704e+00 -1.11942194e-01 5.97876012e-01 7.69827843e-01
1.64163625e+00 -2.46787056e-01 7.76410580e-01 -4.93298978e-01
7.42997229e-01 -1.23078787e+00 -5.25674105e-01 -2.20756069e-01
2.41337553e-01 -2.99034983e-01 -3.09941947e-01 -7.82631099e-01
-4.74372745e-01 1.29784715e+00 9.48043168e-01 4.05953258e-01
1.10912669e+00 2.78535098e-01 2.54796594e-01 -5.84033608e-01
-1.07979059e+00 7.21298456e-02 -8.22506428e-01 6.86890066e-01
-4.91321413e-03 7.43843973e-01 -6.22468293e-01 3.29786479e-01
8.09828416e-02 -3.63519229e-02 8.00064683e-01 8.71136129e-01
-2.78140724e-01 -1.65791023e+00 -2.35107988e-01 4.20954883e-01
-6.61465168e-01 2.50634432e-01 5.93057334e-01 4.35939044e-01
7.29816020e-01 1.20454657e+00 1.83601320e-01 -2.33886659e-01
6.27923787e-01 -4.09196854e-01 -4.15285081e-01 -4.09841508e-01
-8.51626992e-01 7.79422298e-02 4.67189908e-01 -6.73936009e-01
-3.47752273e-01 -4.10972595e-01 -1.49680769e+00 -6.51477993e-01
-2.03122646e-02 5.09940207e-01 1.10833788e+00 5.11052132e-01
9.21916008e-01 1.03876066e+00 1.22488761e+00 -1.20810054e-01
-3.69033605e-01 -4.69539434e-01 -7.27430105e-01 -4.71525565e-02
2.51748711e-01 -5.80340087e-01 -3.13551933e-01 -5.25168359e-01] | [5.861939907073975, 1.7572933435440063] |
3ad151b1-00d1-4c36-8af3-73368d7428b9 | quantum-natural-language-processing-based | 2305.19383 | null | https://arxiv.org/abs/2305.19383v1 | https://arxiv.org/pdf/2305.19383v1.pdf | Quantum Natural Language Processing based Sentiment Analysis using lambeq Toolkit | Sentiment classification is one the best use case of classical natural language processing (NLP) where we can witness its power in various daily life domains such as banking, business and marketing industry. We already know how classical AI and machine learning can change and improve technology. Quantum natural language processing (QNLP) is a young and gradually emerging technology which has the potential to provide quantum advantage for NLP tasks. In this paper we show the first application of QNLP for sentiment analysis and achieve perfect test set accuracy for three different kinds of simulations and a decent accuracy for experiments ran on a noisy quantum device. We utilize the lambeq QNLP toolkit and $t|ket>$ by Cambridge Quantum (Quantinuum) to bring out the results. | ['Luis Miguel Pozo Coronado', 'Sai Nandan Morapakula', 'Srinjoy Ganguly'] | 2023-05-30 | null | null | null | null | ['marketing', 'sentiment-analysis'] | ['miscellaneous', 'natural-language-processing'] | [ 1.00005746e-01 -9.33567435e-02 1.05518013e-01 -4.74946320e-01
-9.03092921e-01 -6.38768792e-01 6.29708469e-01 4.17144597e-01
-8.04033995e-01 1.03361320e+00 -2.51045138e-01 -6.38708174e-01
2.55126655e-01 -1.05826795e+00 -4.35683489e-01 -7.99914241e-01
1.71309654e-02 5.06652117e-01 4.25127475e-03 -9.42841768e-01
5.80537796e-01 3.05353016e-01 -1.21550643e+00 4.61507589e-01
7.43901491e-01 9.82800841e-01 -4.51671064e-01 1.03564799e+00
-9.10014659e-02 7.83433139e-01 -3.62176180e-01 -1.07640612e+00
4.72265333e-01 -4.46921140e-01 -1.03679252e+00 -9.18813944e-01
-1.25711307e-01 1.84586689e-01 -9.63975862e-02 1.49594617e+00
6.63825750e-01 1.44750133e-01 4.45091009e-01 -8.95913422e-01
-5.07793844e-01 5.97335815e-01 -3.19928080e-01 3.19748640e-01
5.67328572e-01 4.78056282e-01 1.31970215e+00 -4.57025379e-01
5.80205142e-01 1.33574927e+00 3.67297560e-01 6.40943885e-01
-1.01023328e+00 -7.81589270e-01 -1.07009101e+00 4.91847456e-01
-1.11502075e+00 -3.15457314e-01 3.00234348e-01 1.62109271e-01
1.36826956e+00 -1.86889559e-01 4.38320845e-01 7.93320358e-01
1.08229387e+00 5.22185147e-01 1.55393183e+00 -8.50287318e-01
5.07477105e-01 2.59544134e-01 3.79674494e-01 7.66050160e-01
-1.42460570e-01 4.27298576e-01 -9.05408442e-01 -1.76359013e-01
4.12227325e-02 -4.04906422e-01 2.95720726e-01 3.81561339e-01
-1.31133890e+00 1.07030284e+00 3.48352462e-01 3.63064229e-01
-1.43014506e-01 5.76452792e-01 6.65470064e-01 7.49033928e-01
2.13715151e-01 7.49714673e-01 -7.20439315e-01 -6.55095041e-01
-3.94637972e-01 3.07316571e-01 9.39993680e-01 7.32430398e-01
9.00570929e-01 -2.52326339e-01 1.97052389e-01 3.42191845e-01
1.35774612e-01 1.15232110e+00 5.86976767e-01 -9.71403003e-01
-5.70810586e-02 2.10874066e-01 -6.83117062e-02 -5.90206981e-01
-6.42872870e-01 2.99069323e-02 -8.55574250e-01 -8.54641274e-02
2.25076109e-01 -3.02886546e-01 -7.48161793e-01 1.24572074e+00
1.79559290e-01 -1.08466119e-01 6.31040096e-01 7.46803761e-01
7.50366449e-01 1.07212734e+00 -8.09750855e-02 -1.30629450e-01
1.65903437e+00 -4.89448488e-01 -5.66046894e-01 3.56648192e-02
1.26586866e+00 -9.38410878e-01 8.68106365e-01 1.05051613e+00
-8.19471836e-01 -2.46408015e-01 -1.07211912e+00 -4.07871872e-01
-7.46751904e-01 -5.76918125e-01 1.34408331e+00 1.28453922e+00
-8.57334554e-01 1.00129473e+00 -8.01479936e-01 -3.83797735e-01
8.64495784e-02 7.39640176e-01 -3.61080348e-01 -1.45918965e-01
-1.86295807e+00 9.86279547e-01 3.86891156e-01 7.80073851e-02
-2.43055835e-01 -2.70179331e-01 -6.23447061e-01 -1.95969060e-01
1.58493474e-01 -5.41884422e-01 1.61876512e+00 -6.16878569e-01
-2.17908168e+00 1.06842089e+00 -1.62545487e-01 -8.42970192e-01
-1.99216470e-01 4.50610742e-02 -6.41287923e-01 3.00698847e-01
-5.54485843e-02 4.80569601e-01 5.09516656e-01 -3.44621465e-02
-5.36648393e-01 -3.49377722e-01 -3.14134099e-02 4.81593199e-02
1.10415101e-01 2.16487914e-01 3.58945996e-01 2.57214785e-01
5.17967381e-02 -1.07348406e+00 -2.71007001e-01 -7.61943340e-01
-1.22568019e-01 -4.66437548e-01 1.65804282e-01 1.53073668e-02
7.40421236e-01 -2.05168295e+00 -2.74867825e-02 2.94729561e-01
-1.40549123e-01 3.25291425e-01 1.48864388e-01 8.92602861e-01
1.17293276e-01 3.30612004e-01 -2.88231343e-01 -3.76936942e-02
3.66512567e-01 3.03771883e-01 -2.34081894e-01 6.23011112e-01
1.92721665e-01 1.24617493e+00 -1.12466061e+00 -3.74409974e-01
1.09463878e-01 7.54161030e-02 -6.80763483e-01 -4.88567322e-01
-2.13877350e-01 3.66116852e-01 -4.95962441e-01 4.97090936e-01
5.35778463e-01 -1.25665709e-01 -9.22936797e-02 4.82232869e-02
-2.46275589e-01 5.81137240e-01 -9.95726049e-01 1.57724082e+00
-3.32690954e-01 6.30984187e-01 5.47761936e-03 -8.61837804e-01
6.03062510e-01 1.31799966e-01 1.67110618e-02 -1.08584273e+00
4.72967595e-01 4.90920931e-01 3.80571693e-01 -4.86002475e-01
9.92808282e-01 -1.13007641e+00 -6.54541612e-01 6.47624493e-01
3.36104184e-01 -1.12386036e+00 2.60920078e-01 4.38933402e-01
1.11321115e+00 -4.27359790e-01 5.66594481e-01 -5.16037166e-01
6.21357918e-01 2.08548546e-01 2.39092693e-01 9.20264959e-01
-8.66272986e-01 6.11516051e-02 4.33550060e-01 -4.91928697e-01
-1.05391097e+00 -8.74820948e-01 -5.16867816e-01 1.09040058e+00
1.58547431e-01 -5.94425440e-01 -4.59709018e-01 -1.41506627e-01
-3.39798093e-01 9.07264769e-01 -1.56897664e-01 -3.84948641e-01
-2.67106533e-01 -1.26221585e+00 7.24288523e-01 -1.66439340e-01
5.47065914e-01 -1.39438879e+00 -1.37504786e-01 5.34546981e-03
2.50132024e-01 -1.34197330e+00 2.97658741e-01 5.61893821e-01
-6.78583145e-01 -6.02824330e-01 -2.04951949e-02 -6.56835020e-01
2.79060826e-02 -1.31346539e-01 9.59707320e-01 -2.23731950e-01
-2.19401196e-01 7.10375980e-02 -6.96864426e-01 -6.91393435e-01
-1.07258570e+00 2.56015688e-01 6.28558099e-01 -4.02173519e-01
8.86936545e-01 -3.06301147e-01 -4.72373396e-01 -4.30583179e-01
-1.03619027e+00 -2.51363158e-01 5.23055911e-01 9.77164865e-01
1.35260448e-01 4.39662516e-01 5.74533701e-01 -1.16873717e+00
7.97927916e-01 -7.50890225e-02 -5.31344473e-01 -2.05180421e-02
-4.01147604e-01 3.84149730e-01 1.08400893e+00 3.92358631e-01
-8.56537461e-01 -2.34890491e-01 -6.47163868e-01 6.02704525e-01
-1.45105317e-01 6.95703804e-01 3.77876937e-01 -3.83572757e-01
1.03847146e+00 -3.14043835e-02 -3.48954737e-01 2.44288608e-01
6.27848446e-01 9.59443152e-01 2.14419037e-01 -5.85227787e-01
4.08186138e-01 5.84140003e-01 5.35347402e-01 -1.22780561e+00
-9.89493191e-01 -5.75308383e-01 -6.29617929e-01 2.25467727e-01
9.38342750e-01 -5.72541118e-01 -1.45276773e+00 3.99824888e-01
-1.07421625e+00 1.78767927e-02 -2.72152781e-01 7.30640531e-01
-4.59887117e-01 5.13680041e-01 -1.02405953e+00 -1.01148295e+00
-7.50814497e-01 -1.10987222e+00 1.20343041e+00 5.59508383e-01
2.12528601e-01 -8.86496425e-01 1.11938752e-01 3.24845016e-01
1.53405324e-01 -2.31172025e-01 8.63479912e-01 -6.18871093e-01
-3.39016587e-01 -4.38341051e-01 -5.96329533e-02 6.46710515e-01
-4.72141296e-01 1.46744112e-02 -9.83461320e-01 -9.11089852e-02
2.85220087e-01 -7.51411617e-01 7.86239207e-01 -8.05109888e-02
3.66907239e-01 2.92619914e-01 1.65194228e-01 2.39392698e-01
1.46374071e+00 1.52135625e-01 8.20462584e-01 8.78831744e-02
2.91962057e-01 2.58233577e-01 4.84062254e-01 2.41403788e-01
1.29261881e-01 7.93089196e-02 2.47710589e-02 6.11738384e-01
7.42518246e-01 2.33890042e-02 6.68868542e-01 1.48460519e+00
-2.20270157e-02 6.42372072e-02 -1.24140859e+00 -7.26036429e-02
-1.46086538e+00 -1.13511693e+00 -3.08802485e-01 2.08445072e+00
8.25446010e-01 5.89839578e-01 -3.69849235e-01 1.31101266e-01
3.98319513e-01 -2.14345112e-01 -1.82516173e-01 -1.39454424e+00
-1.85673892e-01 1.30599546e+00 7.13854551e-01 4.42681223e-01
-8.75008285e-01 1.44136143e+00 6.19092464e+00 1.03948212e+00
-1.08526957e+00 2.20566928e-01 5.01341283e-01 2.69122332e-01
1.02302223e-01 2.22803816e-01 -7.17968881e-01 9.71387234e-03
1.59287763e+00 -2.60749608e-01 7.52233148e-01 5.68259120e-01
5.12945093e-03 -6.58841789e-01 -1.10257626e+00 1.32859206e+00
-2.54415661e-01 -1.24697852e+00 -1.97628677e-01 -2.24817842e-01
7.99815774e-01 7.25337446e-01 -8.19665492e-02 6.87810123e-01
2.52123386e-01 -1.09762716e+00 1.94067627e-01 3.15084815e-01
6.03186786e-01 -1.08228350e+00 1.24881816e+00 5.61322033e-01
-6.24498010e-01 1.15853185e-02 -7.00447857e-01 -6.75731540e-01
2.39616349e-01 6.59752727e-01 -7.82108128e-01 7.48639107e-01
6.25661790e-01 5.67059457e-01 -2.99542487e-01 4.62753385e-01
-3.25386018e-01 7.80879617e-01 -6.40096188e-01 -9.13340569e-01
5.88129282e-01 -7.40888476e-01 3.37249994e-01 1.05833328e+00
7.61971474e-02 6.23318136e-01 -3.32483709e-01 4.67276961e-01
-3.05458158e-01 2.63793141e-01 -5.64550161e-01 -6.04604661e-01
-5.08498885e-02 1.30422008e+00 -1.00048327e+00 -4.77732569e-01
-2.56752700e-01 1.02858567e+00 -3.39469523e-03 -2.23430261e-01
-5.27353585e-01 -7.06042886e-01 3.33988339e-01 -6.16479814e-01
-1.12732679e-01 -5.60978532e-01 -3.46873820e-01 -1.57174456e+00
-4.64479446e-01 -9.40748811e-01 3.77871394e-02 -6.65373623e-01
-1.49961817e+00 2.91143954e-01 -4.32382345e-01 -6.71876550e-01
-1.31561920e-01 -1.28563082e+00 -2.87595689e-01 8.59599710e-01
-1.29375851e+00 -5.47084272e-01 5.57199240e-01 1.87257260e-01
-1.20761037e-01 2.62710322e-02 1.21084213e+00 1.12778217e-01
-3.72612655e-01 3.34339321e-01 6.24335051e-01 1.05937995e-01
7.33557165e-01 -1.26709104e+00 5.49044907e-01 5.63546002e-01
3.78575474e-01 8.70513320e-01 1.07991266e+00 -1.19224206e-01
-2.02618337e+00 -4.16184157e-01 1.03109705e+00 -6.47748113e-01
1.16289330e+00 -7.45100737e-01 -2.92370737e-01 1.99754223e-01
2.31300071e-01 -2.09090009e-01 8.06360185e-01 2.32726753e-01
-1.39135152e-01 -1.20661840e-01 -1.14202297e+00 6.05248094e-01
4.73838240e-01 -1.19716096e+00 -7.54591465e-01 9.50262845e-01
8.53956640e-01 -2.52194583e-01 -7.30476260e-01 3.43385518e-01
6.32881165e-01 -1.19923472e+00 4.26650554e-01 -5.84829688e-01
3.93863797e-01 -1.25453681e-01 -2.96948403e-01 -1.22842717e+00
3.08178425e-01 -1.18762565e+00 5.59824705e-01 6.97317898e-01
6.45771503e-01 -1.00189972e+00 7.98412919e-01 5.76026440e-01
-3.12048364e-02 -3.45961571e-01 -1.17702568e+00 -4.65168744e-01
6.63588464e-01 -1.17315817e+00 3.31054121e-01 7.31872141e-01
8.32606614e-01 8.47258568e-01 -1.32889688e-01 -1.65880799e-01
4.58329827e-01 5.47403507e-02 5.19675374e-01 -8.10379505e-01
-4.01865840e-01 -2.27586702e-01 -8.60918880e-01 -8.58386517e-01
1.14045553e-02 -1.15348125e+00 1.04939915e-01 -9.48438525e-01
1.27470449e-01 -8.95769745e-02 -3.42633426e-01 -6.42624199e-02
2.77630568e-01 5.95836341e-01 1.41051948e-01 -1.08976528e-01
-8.54561090e-01 4.48143870e-01 1.22841716e+00 1.98632330e-02
-7.01331627e-03 -1.80763870e-01 -5.65457284e-01 5.87341130e-01
9.31696713e-01 -6.97101533e-01 1.96052596e-01 4.74258140e-02
1.41248608e+00 -6.41217232e-02 4.22655679e-02 -9.48275030e-01
3.60644162e-01 6.89934641e-02 -1.60498515e-01 -2.83780277e-01
2.03490585e-01 -3.51269841e-01 -6.36827052e-01 5.39267898e-01
-1.22045465e-01 -4.04238701e-03 -4.76376303e-02 2.79024094e-01
-1.84024185e-01 -5.85046649e-01 8.94464731e-01 -2.69326210e-01
-6.62614763e-01 -4.88428026e-03 -5.07617295e-01 8.54657143e-02
8.27757061e-01 4.48313385e-01 -6.35921836e-01 -3.68025720e-01
-4.71412003e-01 2.97163039e-01 4.79638904e-01 -3.10061965e-02
2.23644793e-01 -6.85443997e-01 -4.04897809e-01 1.09420717e-01
2.99620032e-02 -3.04383248e-01 1.89327121e-01 8.77496362e-01
-1.16893673e+00 9.86665964e-01 -3.88232688e-03 -4.59296733e-01
-5.93520939e-01 5.40500283e-01 2.38276497e-01 -3.05345148e-01
-2.10189328e-01 9.48720992e-01 -3.16165358e-01 -8.10735762e-01
-5.89330077e-01 -5.42929113e-01 2.21035376e-01 -3.73588502e-01
6.34201407e-01 2.74709493e-01 3.75621319e-01 -8.11970055e-01
-3.75468612e-01 3.71917218e-01 -4.74238731e-02 -5.13103843e-01
1.10948014e+00 7.05713704e-02 -6.29344106e-01 8.33256900e-01
1.43439460e+00 -6.68257624e-02 -9.31600779e-02 2.29529776e-02
1.75486639e-01 3.65434319e-01 3.03291455e-02 -6.36870742e-01
-7.58117363e-02 1.36491287e+00 5.17510772e-01 6.65749848e-01
8.32266212e-01 -5.85936978e-02 1.29446113e+00 1.30337691e+00
1.14197850e+00 -1.30845070e+00 -3.79902452e-01 9.73272979e-01
-7.78122246e-02 -1.58196211e+00 2.76457012e-01 -2.07392260e-01
-5.29161036e-01 1.34613001e+00 -2.28549078e-01 -3.72024149e-01
7.67495453e-01 -6.71425238e-02 4.67788838e-02 -4.53450143e-01
-7.40575314e-01 -3.17018479e-01 -1.13469265e-01 1.32762715e-01
6.76541865e-01 3.46655935e-01 -5.67419887e-01 4.42740470e-01
-8.25151145e-01 -2.95862164e-02 8.95036399e-01 1.14076149e+00
-5.56100070e-01 -1.49395680e+00 -3.57991964e-01 5.95818698e-01
-9.03536141e-01 -6.78667486e-01 -3.60473603e-01 4.73404855e-01
9.46561545e-02 1.12819326e+00 -4.03516263e-01 -5.27367532e-01
-9.58763715e-03 4.03085887e-01 8.19088995e-01 -9.59941506e-01
-9.24040616e-01 -5.09387791e-01 8.20544586e-02 -5.11127114e-01
-8.28483403e-01 -5.80551565e-01 -1.93781936e+00 -8.69911373e-01
-6.27030730e-01 7.11122334e-01 9.76585746e-01 1.36302280e+00
1.29657760e-01 1.37734935e-02 4.91301388e-01 -6.45803094e-01
-6.55149519e-01 -7.93027282e-01 -9.92964566e-01 4.34142016e-02
7.50395060e-02 -2.23963156e-01 -5.24056613e-01 -1.26171798e-01] | [5.584643840789795, 4.9522905349731445] |
9797412f-a73b-4fcf-8a2c-5d351dc67ff8 | rotation-invariant-graph-neural-networks | 2106.09575 | null | https://arxiv.org/abs/2106.09575v1 | https://arxiv.org/pdf/2106.09575v1.pdf | Rotation Invariant Graph Neural Networks using Spin Convolutions | Progress towards the energy breakthroughs needed to combat climate change can be significantly accelerated through the efficient simulation of atomic systems. Simulation techniques based on first principles, such as Density Functional Theory (DFT), are limited in their practical use due to their high computational expense. Machine learning approaches have the potential to approximate DFT in a computationally efficient manner, which could dramatically increase the impact of computational simulations on real-world problems. Approximating DFT poses several challenges. These include accurately modeling the subtle changes in the relative positions and angles between atoms, and enforcing constraints such as rotation invariance or energy conservation. We introduce a novel approach to modeling angular information between sets of neighboring atoms in a graph neural network. Rotation invariance is achieved for the network's edge messages through the use of a per-edge local coordinate frame and a novel spin convolution over the remaining degree of freedom. Two model variants are proposed for the applications of structure relaxation and molecular dynamics. State-of-the-art results are demonstrated on the large-scale Open Catalyst 2020 dataset. Comparisons are also performed on the MD17 and QM9 datasets. | ['C. Lawrence Zitnick', 'Zachary Ulissi', 'Anuroop Sriram', 'Aditya Grover', 'Abhishek Das', 'Adeesh Kolluru', 'Muhammed Shuaibi'] | 2021-06-17 | null | null | null | null | ['initial-structure-to-relaxed-energy-is2re'] | ['graphs'] | [ 3.00210208e-01 -2.28439048e-01 -3.35160255e-01 -1.10222675e-01
-3.28466326e-01 -4.27696943e-01 6.30746722e-01 4.30295855e-01
-4.09402966e-01 1.11855710e+00 2.10742261e-02 -6.44833684e-01
7.66312778e-02 -1.01696002e+00 -9.42466795e-01 -1.02615082e+00
-2.60963500e-01 5.20644546e-01 -1.12638548e-01 -3.50586116e-01
3.16099495e-01 9.57494497e-01 -1.16357362e+00 1.65200010e-01
8.81045401e-01 6.76553190e-01 7.88065195e-02 3.68596971e-01
6.79224432e-02 7.21647561e-01 -1.01432420e-01 2.74339110e-01
4.75344993e-02 -4.04034853e-01 -8.77691507e-01 -5.82482100e-01
3.02938759e-01 8.26142281e-02 -6.79809093e-01 9.45619583e-01
6.71797514e-01 8.71538699e-01 5.57461917e-01 -6.50751293e-01
-4.47942644e-01 2.59976029e-01 -4.30349648e-01 -8.89678597e-02
2.20403746e-01 3.77187669e-01 9.70593750e-01 -7.64045417e-01
8.24362636e-01 9.65411127e-01 8.59830499e-01 3.75299484e-01
-1.52802098e+00 -4.58127618e-01 -9.10874084e-02 6.46805584e-01
-1.58903265e+00 -2.84411699e-01 8.94076645e-01 -2.14543134e-01
1.91049874e+00 2.81015337e-01 7.64330685e-01 7.20779955e-01
8.13942194e-01 -1.70696467e-01 7.97613442e-01 -5.14112234e-01
5.43664515e-01 -5.74547470e-01 -8.79354551e-02 8.17862093e-01
3.03181529e-01 4.81160671e-01 -6.71922445e-01 -4.43800628e-01
4.66041863e-01 -4.96460013e-02 -3.08549702e-01 -6.81571841e-01
-1.26213467e+00 9.13834989e-01 1.06156087e+00 -7.53723143e-04
-5.88757396e-01 5.34042120e-01 5.85719943e-01 -1.46335065e-01
4.37415272e-01 7.31700242e-01 -4.56865072e-01 -3.25861871e-02
-8.66579056e-01 4.36834097e-01 6.49991870e-01 4.50621754e-01
7.00723350e-01 6.10331185e-02 6.71387091e-02 1.74917340e-01
2.96225369e-01 1.70925751e-01 -9.79819670e-02 -1.10726464e+00
1.28284082e-01 2.33572364e-01 2.51609534e-01 -8.36728930e-01
-6.38609886e-01 -2.72063076e-01 -1.14818001e+00 3.78687203e-01
-1.46676730e-02 -1.33434147e-01 -7.11766660e-01 1.60712504e+00
7.45908201e-01 -2.07571179e-01 -9.97707993e-02 5.16449809e-01
5.40815115e-01 9.71152484e-01 1.31882692e-03 -5.00876606e-01
9.92207706e-01 -1.00636172e+00 -6.17357612e-01 -6.69052079e-02
9.04145002e-01 -4.95848507e-01 4.82544661e-01 2.99555659e-01
-1.20789218e+00 -4.30825651e-01 -1.07076657e+00 -3.55307221e-01
-5.66750884e-01 -2.25550190e-01 1.02527475e+00 4.68731493e-01
-8.46575856e-01 1.26986051e+00 -1.34677243e+00 -5.33311851e-02
2.55851448e-01 9.05984581e-01 -3.95448506e-01 -1.17972761e-01
-1.13126707e+00 1.17322934e+00 5.57575285e-01 1.62493587e-01
-6.50862753e-01 -9.76354361e-01 -8.13578904e-01 2.69285977e-01
1.84647188e-01 -7.80042470e-01 1.05073261e+00 -5.58936000e-01
-1.48869216e+00 6.62399903e-02 -3.46516043e-01 -5.41127145e-01
1.38375208e-01 4.09920156e-01 -1.95145696e-01 -3.35264578e-02
-2.82421917e-01 7.60755181e-01 1.11732826e-01 -8.13151896e-01
1.19664855e-01 -2.58214265e-01 2.52123941e-02 3.33817571e-01
-7.89139420e-02 -3.13835949e-01 1.07998461e-01 -1.70240447e-01
-2.39845775e-02 -1.44030154e+00 -6.93334103e-01 6.91827908e-02
-3.76765579e-01 -2.00823769e-02 7.56091654e-01 -6.65752351e-01
9.06792521e-01 -1.57370019e+00 5.66478670e-01 4.26772118e-01
2.54248619e-01 2.54510701e-01 2.15912670e-01 8.51708114e-01
-4.66679841e-01 -8.99895728e-02 -2.73446232e-01 -9.40370336e-02
-2.17615649e-01 1.97788164e-01 2.36556068e-01 6.04790390e-01
-9.63710397e-02 7.77323604e-01 -8.62954736e-01 1.89229593e-01
4.12483186e-01 1.07043290e+00 -8.72442603e-01 -2.14829609e-01
-4.44648445e-01 7.65821695e-01 -2.79481113e-01 1.87486634e-01
5.90332747e-01 -3.83738428e-01 7.88925946e-01 -7.65163481e-01
-6.00373447e-01 6.89663827e-01 -7.12500572e-01 1.82911384e+00
-3.22412699e-01 2.65756607e-01 5.87613769e-02 -1.10042787e+00
5.71104467e-01 2.31417969e-01 7.99967170e-01 -8.53859305e-01
-1.42327368e-01 2.54201498e-02 5.43889403e-01 5.36219142e-02
6.27841890e-01 -2.81244099e-01 1.77546844e-01 1.79704070e-01
-2.59586185e-01 -3.35079283e-01 2.77088732e-01 1.96907073e-01
9.13530707e-01 3.24422628e-01 4.30318773e-01 -6.97815955e-01
5.08295596e-01 3.47172827e-01 2.18556404e-01 5.00857294e-01
8.82786512e-02 1.15610704e-01 3.13772708e-01 -8.36493313e-01
-1.36076939e+00 -4.71826524e-01 -1.42419040e-01 6.46095455e-01
-1.69120923e-01 -8.63106847e-01 -7.84872472e-01 -3.73115242e-01
1.13637850e-01 8.63553405e-01 -5.16595185e-01 -6.33850396e-01
-6.54177547e-01 -1.09223425e+00 1.25157818e-01 3.60201746e-01
3.37455660e-01 -7.84203291e-01 -4.89078909e-01 4.57607001e-01
4.55698371e-02 -7.18129575e-01 -4.66620386e-01 6.98292851e-01
-7.85980642e-01 -1.01866007e+00 -1.58067942e-01 -2.22754583e-01
6.67124629e-01 4.91493076e-01 1.03776670e+00 3.04584727e-02
-5.45179307e-01 -1.52207464e-01 1.40304059e-01 8.17390904e-02
-4.42088544e-01 3.04948926e-01 3.72048825e-01 -6.62950635e-01
9.15121809e-02 -7.23079920e-01 -8.64139676e-01 -7.45357126e-02
-5.94247997e-01 2.64524668e-01 2.76638955e-01 1.01023841e+00
9.16231453e-01 2.70783275e-01 1.24906495e-01 -6.72788441e-01
4.38725412e-01 -3.31005037e-01 -6.99321926e-01 1.71473455e-02
-8.71298254e-01 5.30270755e-01 8.30665708e-01 -9.46389511e-02
-1.02150154e+00 3.42723519e-01 -3.11503172e-01 -1.21600293e-01
2.21117422e-01 1.03544307e+00 -6.27139164e-03 -7.75552511e-01
6.45672858e-01 9.88011956e-02 4.30551991e-02 -3.53814572e-01
3.76517236e-01 1.24129802e-01 2.84643590e-01 -9.89004970e-01
4.81686145e-01 3.51038992e-01 9.78932679e-01 -1.01682031e+00
-5.80553889e-01 -1.50371850e-01 -6.88968778e-01 -7.19094202e-02
7.91527927e-01 -7.92440057e-01 -1.14716685e+00 2.91514993e-01
-1.18577945e+00 -2.96327859e-01 -2.33811259e-01 4.37945396e-01
-3.45764101e-01 6.98602855e-01 -3.74724656e-01 -4.26095635e-01
-5.55554807e-01 -1.50661683e+00 7.06925631e-01 1.02755569e-01
-2.04561979e-01 -1.15734065e+00 3.52644295e-01 2.15877384e-01
6.59252524e-01 4.93902445e-01 1.18115532e+00 -1.12581916e-01
-5.70698857e-01 -1.10034890e-01 8.28873366e-02 1.25429079e-01
1.71919540e-01 1.33826211e-01 -4.36067671e-01 -6.88266456e-01
-3.69900793e-01 -2.07226366e-01 9.54956651e-01 5.77458560e-01
1.19462454e+00 -1.90450609e-01 -7.00223923e-01 5.28934896e-01
1.16152954e+00 2.51826406e-01 5.22386730e-01 -8.09744895e-02
1.10060370e+00 1.68788806e-01 1.15405381e-01 2.86271542e-01
2.25361288e-01 8.89612436e-01 5.41256309e-01 -1.01940118e-01
6.89341053e-02 -1.18720457e-01 4.20551986e-01 8.70542347e-01
-6.75801218e-01 -2.37257093e-01 -1.06386960e+00 -1.10714324e-01
-1.68937457e+00 -1.28896475e+00 -3.37606579e-01 2.38315225e+00
8.30023527e-01 1.00101918e-01 -2.16994405e-01 -5.61970705e-03
4.31268960e-01 3.93088400e-01 -8.91286016e-01 -5.58136523e-01
1.33688942e-01 4.56745833e-01 7.91419446e-01 7.66292214e-01
-8.27123940e-01 7.14161634e-01 6.78557873e+00 6.50568128e-01
-1.31344318e+00 8.70141611e-02 6.60188615e-01 -2.52143621e-01
-1.37904376e-01 3.11744094e-01 -6.64408386e-01 3.24726135e-01
1.32572925e+00 -6.21584207e-02 9.72551167e-01 5.65413415e-01
5.76204598e-01 -1.26942769e-02 -8.80528510e-01 5.85844278e-01
-2.57229328e-01 -2.09211302e+00 1.29119465e-02 3.30369622e-01
9.03921962e-01 5.22363544e-01 -1.01938918e-01 6.43838048e-02
1.54415563e-01 -1.15068114e+00 4.64273095e-01 6.28534794e-01
9.73402798e-01 -1.19469059e+00 2.65945226e-01 1.21924005e-01
-1.23917091e+00 5.00171721e-01 -4.16830838e-01 -4.01884258e-01
-2.39832737e-02 6.21749878e-01 -6.82586968e-01 6.97431684e-01
2.65802592e-01 7.73331285e-01 -2.22388301e-02 6.25045180e-01
1.39217898e-01 4.40543175e-01 -5.19827008e-01 -1.12663694e-01
3.13855141e-01 -6.80489361e-01 2.10967541e-01 7.77129591e-01
2.29061216e-01 -3.34464163e-02 3.02380681e-01 9.54812467e-01
-2.25249007e-01 -3.02451342e-01 -3.89483452e-01 -2.70458996e-01
3.60251993e-01 1.25504220e+00 -5.83327889e-01 -1.19600564e-01
-5.21492064e-01 7.81087756e-01 3.72764587e-01 2.31724083e-01
-9.86195922e-01 -1.52112380e-01 8.44919860e-01 1.39878824e-01
3.51057917e-01 -8.32863450e-01 3.75521600e-01 -8.86124074e-01
-2.84984738e-01 -8.98701191e-01 -1.12599358e-01 -5.82628548e-01
-8.52784157e-01 1.34601682e-01 -1.62479475e-01 -3.69606316e-01
8.58352170e-04 -9.11441445e-01 -6.75930858e-01 1.10068655e+00
-1.41811705e+00 -8.95892382e-01 -2.03545809e-01 2.75706530e-01
4.90044728e-02 2.60097414e-01 1.15978587e+00 3.45255554e-01
-6.99743629e-01 1.57320157e-01 7.83819437e-01 -5.20837426e-01
4.10320580e-01 -1.11846972e+00 7.69011199e-01 4.62333739e-01
-1.78213045e-01 8.22459519e-01 9.05752897e-01 -8.24227333e-01
-2.08598804e+00 -1.11028755e+00 6.32225215e-01 -3.56299579e-02
4.10628110e-01 -2.57664502e-01 -9.02015507e-01 5.21945059e-01
6.50817603e-02 4.14773524e-01 5.78285992e-01 1.28534988e-01
-1.18683644e-01 2.29795799e-01 -1.04361105e+00 3.97648811e-01
1.06812835e+00 -7.54389346e-01 1.74200594e-01 9.02909756e-01
5.31082451e-01 -5.47802925e-01 -1.06401658e+00 3.30533624e-01
5.15517831e-01 -7.72560358e-01 1.35757768e+00 -9.96984720e-01
1.36413425e-01 -3.39890838e-01 -1.13967985e-01 -1.33465135e+00
-6.22429192e-01 -9.29935515e-01 -4.76714641e-01 4.89248306e-01
3.26350480e-01 -5.68803012e-01 8.01306129e-01 5.82533538e-01
-3.77639830e-01 -7.60015965e-01 -1.21954453e+00 -6.38153434e-01
3.18377823e-01 3.81480791e-02 4.78533715e-01 1.00986576e+00
1.50648281e-01 4.41558629e-01 -3.49303365e-01 1.80507675e-01
2.75240898e-01 -1.40688077e-01 2.91857898e-01 -1.14232576e+00
-4.31656152e-01 -2.09160298e-01 -1.38877437e-01 -7.12638974e-01
3.24399084e-01 -1.15327287e+00 -2.89644927e-01 -1.42104781e+00
2.62927651e-01 -1.10513233e-01 -3.46478015e-01 3.93217593e-01
1.26842529e-01 -7.16433802e-04 -1.43467709e-01 1.39454752e-01
-3.53014320e-01 9.53636825e-01 1.24448740e+00 -3.43662232e-01
4.51825559e-03 -5.00652254e-01 -2.18918979e-01 3.70299697e-01
8.02833915e-01 -6.23809457e-01 -2.03197658e-01 -7.50635490e-02
5.59314907e-01 9.61826369e-02 4.24974799e-01 -1.14074981e+00
2.91537195e-01 -2.44734064e-01 3.32827330e-01 -4.63193238e-01
5.70318997e-01 -6.62563205e-01 6.96454108e-01 9.70407367e-01
-1.33187667e-01 3.97382468e-01 2.99085736e-01 5.37307203e-01
7.94871598e-02 -2.34183613e-02 9.42188740e-01 -2.95134187e-01
-6.98765218e-02 4.34868127e-01 -3.58361572e-01 -4.21466529e-01
8.28789890e-01 2.57928610e-01 -2.54081428e-01 -7.75937214e-02
-5.50179482e-01 -1.70687869e-01 7.05599964e-01 -2.38195181e-01
-3.80256362e-02 -1.25484133e+00 -1.71656594e-01 -2.77194716e-02
-2.95805931e-01 5.68161719e-02 4.05826747e-01 8.45618188e-01
-9.16566849e-01 7.31084764e-01 -3.77623886e-01 -2.40765631e-01
-1.13650584e+00 6.52182341e-01 8.66909504e-01 -6.02042854e-01
-3.14902961e-01 3.40844601e-01 -1.75427690e-01 -4.83993471e-01
-3.89526635e-01 -3.97183239e-01 4.67735976e-01 -2.48717785e-01
1.95479691e-01 6.03198707e-01 4.02846664e-01 -7.85812855e-01
-5.16736865e-01 2.24846795e-01 -2.80443728e-01 4.29479390e-01
1.54033625e+00 3.60175699e-01 -3.31552774e-01 -1.77748993e-01
1.07804263e+00 -8.79044309e-02 -1.34698069e+00 3.89439501e-02
-4.36888546e-01 4.45406996e-02 7.68180490e-01 -7.32702971e-01
-9.25200462e-01 8.71507168e-01 5.66383421e-01 2.87071727e-02
4.68734682e-01 -5.31578422e-01 7.63458550e-01 1.11138952e+00
3.14025253e-01 -9.76994634e-01 -3.55934560e-01 6.70916200e-01
5.58325052e-01 -1.01001763e+00 5.03725171e-01 -2.09700420e-01
1.56359538e-01 1.37347662e+00 2.28025407e-01 2.88778692e-02
3.20802510e-01 1.63717836e-01 -8.43128860e-01 -2.85706371e-01
-5.73037326e-01 4.25568700e-01 1.44837365e-01 3.53018433e-01
8.80818486e-01 1.12334788e-01 -2.72262633e-01 -1.41250975e-02
1.54887602e-01 -1.97375610e-01 3.95165145e-01 1.15216184e+00
-3.62405807e-01 -1.46656382e+00 -7.45294318e-02 2.04209059e-01
-1.17449142e-01 -3.39822859e-01 -3.36746782e-01 6.82319880e-01
1.51696190e-01 5.17257631e-01 -1.33351907e-01 4.43362147e-02
1.53973818e-01 1.50968522e-01 8.66319358e-01 -2.99065858e-01
-5.03292799e-01 -4.94121909e-01 3.04282486e-01 -5.52716315e-01
-4.04295981e-01 -7.36885905e-01 -1.57213426e+00 -8.86647403e-01
-4.40118581e-01 4.54296142e-01 8.47494483e-01 8.53490949e-01
8.62693727e-01 8.43023062e-01 2.90854126e-01 -1.50624943e+00
-6.53379679e-01 -6.34411991e-01 -1.71435609e-01 -2.94026341e-02
2.27267995e-01 -7.61094451e-01 -7.85029158e-02 -3.19912672e-01] | [5.306546688079834, 5.401623249053955] |
e84068f5-33e1-4197-9365-1b6d1a9aef1a | policy-learning-for-many-outcomes-of-interest | 2212.06312 | null | https://arxiv.org/abs/2212.06312v1 | https://arxiv.org/pdf/2212.06312v1.pdf | Policy learning for many outcomes of interest: Combining optimal policy trees with multi-objective Bayesian optimisation | Methods for learning optimal policies use causal machine learning models to create human-interpretable rules for making choices around the allocation of different policy interventions. However, in realistic policy-making contexts, decision-makers often care about trade-offs between outcomes, not just singlemindedly maximising utility for one outcome. This paper proposes an approach termed Multi-Objective Policy Learning (MOPoL) which combines optimal decision trees for policy learning with a multi-objective Bayesian optimisation approach to explore the trade-off between multiple outcomes. It does this by building a Pareto frontier of non-dominated models for different hyperparameter settings. The key here is that a low-cost surrogate function can be an accurate proxy for the very computationally costly optimal tree in terms of expected regret. This surrogate can be fit many times with different hyperparameter values to proxy the performance of the optimal model. The method is applied to a real-world case-study of conditional cash transfers in Morocco where hybrid (partially optimal, partially greedy) policy trees provide good performance as a surrogate for optimal trees while being computationally cheap enough to feasibly fit a Pareto frontier. | ['Patrick Rehill'] | 2022-12-13 | null | null | null | null | ['bayesian-optimisation'] | ['methodology'] | [ 6.84469864e-02 3.15575540e-01 -8.70675981e-01 -2.81329185e-01
-1.06876910e+00 -4.46646631e-01 5.50261021e-01 4.50031072e-01
-7.20192730e-01 1.16234469e+00 6.57319963e-01 -1.02352428e+00
-8.92983437e-01 -6.63496077e-01 -4.26676780e-01 -7.36725569e-01
-1.25370175e-01 1.07358134e+00 -2.99922854e-01 3.04077029e-01
3.70820254e-01 2.96018124e-01 -1.16631615e+00 2.47481942e-01
9.10085261e-01 7.84057736e-01 1.18404254e-01 6.14974439e-01
5.01854047e-02 5.52353442e-01 -3.76240373e-01 -4.69902247e-01
3.95311892e-01 -2.09679589e-01 -5.98774374e-01 -3.50683033e-01
-1.55816495e-01 -4.96419013e-01 4.67167616e-01 8.87728214e-01
6.73343062e-01 3.86342183e-02 1.07248938e+00 -1.03213930e+00
-1.03422031e-01 8.11383545e-01 -2.10347831e-01 -9.52778757e-02
1.69786483e-01 4.45602924e-01 1.22183692e+00 7.11587667e-02
2.56611168e-01 1.97580898e+00 2.62285739e-01 8.44533294e-02
-1.76743495e+00 -2.97784716e-01 -2.58278176e-02 1.95220187e-01
-6.92602932e-01 -4.31920230e-01 3.52455199e-01 -5.34645200e-01
7.16827273e-01 3.92023295e-01 8.85681391e-01 1.15143669e+00
5.72306097e-01 5.51409781e-01 1.62238336e+00 -5.96133530e-01
6.42188549e-01 1.56405330e-01 -4.96687233e-01 2.01188341e-01
6.14615500e-01 8.87193263e-01 4.92909597e-03 -7.85999775e-01
5.82797348e-01 -1.76568516e-02 7.87191913e-02 -4.04127777e-01
-9.79968429e-01 1.12707305e+00 1.14033997e-01 -8.54874775e-02
-8.88463199e-01 3.67445469e-01 3.01599890e-01 2.10578218e-01
3.58763546e-01 7.84196615e-01 -8.39190722e-01 -2.16487065e-01
-6.60012305e-01 6.57894671e-01 9.05461252e-01 1.24619603e-01
3.50817472e-01 -9.71175432e-02 -8.53951275e-01 5.64954638e-01
6.22758031e-01 4.27452117e-01 1.66604623e-01 -1.31867409e+00
5.62273264e-01 2.22824931e-01 1.02691960e+00 -4.51292068e-01
-3.00336510e-01 -3.26151699e-01 -2.61285812e-01 6.88692749e-01
8.74383986e-01 -8.10057282e-01 -6.92364514e-01 1.53485072e+00
4.73158687e-01 -1.51161551e-01 2.98723262e-02 7.22432494e-01
-1.92694068e-01 6.69933379e-01 6.06293142e-01 -6.05862558e-01
1.23031950e+00 -2.36335799e-01 -4.45879489e-01 -1.43188789e-01
6.95533752e-01 -5.43200135e-01 8.12509298e-01 4.36007112e-01
-1.00025570e+00 8.60978365e-02 -5.57200849e-01 7.37365127e-01
2.09546477e-01 -1.88625917e-01 7.04195917e-01 7.45195866e-01
-6.48516655e-01 8.50361049e-01 -5.32867491e-01 3.46115418e-02
5.40447414e-01 2.75933355e-01 4.33342308e-01 1.39488459e-01
-9.88888502e-01 1.42239296e+00 7.11048841e-01 -2.51655221e-01
-1.11039329e+00 -7.31212735e-01 -1.84892386e-01 3.02532107e-01
8.48122060e-01 -9.19823825e-01 1.47096729e+00 -1.09705162e+00
-1.66393745e+00 2.07782343e-01 6.17730059e-02 -6.38435483e-01
8.53783131e-01 -1.38879478e-01 -1.42427400e-01 -3.23630750e-01
1.09576918e-01 3.56096029e-01 7.30356693e-01 -1.12940061e+00
-8.25066686e-01 -3.40370297e-01 1.35287553e-01 2.91818649e-01
2.17935577e-01 2.29201511e-01 6.29100442e-01 -6.06600046e-01
-5.55252135e-01 -8.30330491e-01 -7.14226127e-01 -5.87536395e-01
-1.57368273e-01 -3.16778302e-01 -4.64203069e-03 -4.65736598e-01
1.37724566e+00 -1.53297973e+00 4.98024300e-02 3.45775008e-01
-4.24724936e-01 -7.02524185e-03 1.20047785e-01 3.24241698e-01
4.72285300e-02 3.68619740e-01 -1.73386559e-01 2.88202256e-01
2.61044323e-01 3.36247236e-01 -2.02138916e-01 4.59427357e-01
-1.16695734e-02 7.34294832e-01 -1.00473821e+00 -4.88112003e-01
3.09976250e-01 -9.84229296e-02 -4.81377244e-01 1.55346826e-01
-5.15616596e-01 4.27043617e-01 -6.83689296e-01 4.92975682e-01
1.37616068e-01 1.27478838e-01 5.94256759e-01 4.55871075e-01
-2.30698779e-01 3.36294681e-01 -1.37129104e+00 9.42585468e-01
-5.11708200e-01 1.43368050e-01 2.16723792e-02 -1.35023463e+00
7.53176391e-01 4.70181942e-01 5.59566140e-01 -6.30231619e-01
3.32702547e-01 4.17423785e-01 1.98471576e-01 -5.56717396e-01
-1.32427976e-01 -7.84697592e-01 -1.31696612e-01 7.01967180e-01
-2.80157596e-01 3.27503979e-02 -1.72080800e-01 -6.25048995e-01
8.97519410e-01 3.17797542e-01 7.73717225e-01 -7.41889715e-01
1.46648601e-01 1.01324975e-01 8.03054452e-01 9.95727420e-01
-2.52002850e-02 -5.05987694e-03 9.51330960e-01 -6.21243298e-01
-1.11039591e+00 -7.53493309e-01 -1.49869218e-01 1.04386902e+00
-5.07647693e-01 3.29043418e-01 -4.61367220e-01 -5.97450614e-01
3.62383366e-01 1.40585899e+00 -5.47458827e-01 6.79033175e-02
-4.01097208e-01 -1.18536425e+00 1.10844009e-01 9.92508754e-02
5.35309725e-02 -1.09323490e+00 -1.31306481e+00 6.53505504e-01
9.22741815e-02 -4.30788487e-01 -6.09760545e-02 2.90190548e-01
-1.20252073e+00 -1.14697134e+00 -7.34836102e-01 1.99983165e-01
3.12542349e-01 -3.27591062e-01 1.07772064e+00 -3.19570184e-01
1.43464178e-01 1.68528602e-01 4.79472306e-04 -1.01242733e+00
-7.19366670e-01 -3.16676110e-01 7.62085617e-02 4.49761860e-02
2.35206038e-01 -4.14111823e-01 -6.12073421e-01 2.25267857e-01
-5.81629038e-01 -2.37466305e-01 7.96042979e-01 9.37679768e-01
3.22185844e-01 7.35969394e-02 8.33414495e-01 -7.85981953e-01
1.00743389e+00 -7.57107794e-01 -1.09639966e+00 7.93505907e-01
-1.05253708e+00 8.20421696e-01 4.31752592e-01 -6.07706964e-01
-1.40821612e+00 -2.47642651e-01 3.99052173e-01 -7.65063688e-02
-1.76724747e-01 4.76996034e-01 -1.26784503e-01 5.55569649e-01
8.00638974e-01 -4.31885600e-01 -1.50391683e-01 -6.63971424e-01
3.56021494e-01 6.32308960e-01 1.33848727e-01 -1.17181134e+00
1.32461503e-01 1.93955868e-01 2.86341697e-01 -9.05667171e-02
-7.17606544e-01 -2.39155795e-02 -2.03897372e-01 -3.22107673e-01
5.55294394e-01 -5.32224357e-01 -1.04503644e+00 -1.47540405e-01
-7.05546737e-01 -6.41608655e-01 -3.50907564e-01 8.13735366e-01
-1.04579484e+00 -1.42723262e-01 3.42912614e-01 -1.42006779e+00
-9.65387374e-02 -1.14052451e+00 4.80531514e-01 1.18766636e-01
-5.19347489e-01 -1.26433384e+00 2.99623460e-01 2.03953639e-01
1.81623966e-01 6.60275519e-01 1.37683547e+00 -5.08806169e-01
-2.98707098e-01 8.38358104e-02 1.06272332e-01 -9.21198204e-02
-4.54898961e-02 -6.45062849e-02 -6.06973350e-01 -4.11390305e-01
2.68526142e-03 -1.82704031e-01 5.32305479e-01 1.03026450e+00
8.11885059e-01 -1.14191401e+00 -2.51977056e-01 2.28456214e-01
1.56379688e+00 6.31246507e-01 3.96397442e-01 7.88456500e-01
1.02190264e-02 8.68333757e-01 8.76565099e-01 7.10463226e-01
3.31849307e-01 8.41745615e-01 4.96259570e-01 3.73196363e-01
5.59767187e-01 -2.75496453e-01 4.43185508e-01 -3.83973151e-01
-3.38274598e-01 -4.20535319e-02 -1.16540337e+00 5.72337389e-01
-2.21001959e+00 -1.04619849e+00 2.02620104e-01 2.72062230e+00
8.35325956e-01 4.76815432e-01 6.28281534e-01 -7.76074231e-02
5.43122530e-01 -2.22101316e-01 -6.05120897e-01 -1.21545041e+00
3.25838119e-01 -3.64199956e-03 1.05612028e+00 6.55009270e-01
-7.92703211e-01 4.26906049e-01 6.74804497e+00 8.87785614e-01
-8.72931719e-01 1.07650105e-02 9.53847051e-01 -3.27384204e-01
-4.49143320e-01 4.66522574e-01 -5.84770918e-01 6.33964777e-01
1.44232726e+00 -2.78787732e-01 5.41282535e-01 5.48360050e-01
9.83364105e-01 -6.23877287e-01 -9.84768927e-01 4.72081304e-01
-9.01879430e-01 -1.39054060e+00 -1.73472688e-01 3.55742753e-01
8.82493258e-01 -1.96105048e-01 -2.49181852e-01 9.24711227e-02
1.22335041e+00 -1.26996195e+00 9.53367174e-01 8.43896925e-01
5.47513247e-01 -1.10718453e+00 7.73665428e-01 7.27743804e-01
-3.22487265e-01 -9.24818337e-01 -2.06471726e-01 -3.20561647e-01
1.08518332e-01 5.05043805e-01 -1.08039725e+00 3.57841223e-01
6.10825777e-01 6.43050224e-02 1.86664667e-02 1.29522371e+00
-3.49172086e-01 7.89236188e-01 -3.72828901e-01 -2.06677318e-01
6.55196726e-01 -3.17570090e-01 6.69173121e-01 1.01587093e+00
4.17795539e-01 1.09442957e-01 7.93533027e-02 7.50633597e-01
5.77696025e-01 -9.34018567e-03 -4.53828633e-01 1.15504816e-01
5.35806417e-01 6.66783810e-01 -5.56616724e-01 -1.52524188e-01
1.84099637e-02 -1.01887606e-01 1.41839415e-01 4.56285238e-01
-4.53063697e-01 2.98546702e-01 7.33782709e-01 -2.73161847e-02
1.33586511e-01 1.86052218e-01 -6.91518664e-01 -5.90193212e-01
-3.55043381e-01 -9.39248621e-01 6.50609195e-01 -3.40020895e-01
-9.50404406e-01 -3.10554415e-01 8.09298873e-01 -8.80483210e-01
-5.76998651e-01 -5.21976650e-01 -7.98493564e-01 1.07683694e+00
-1.20317400e+00 -5.26964664e-01 7.72731066e-01 6.58039562e-03
3.73787314e-01 -8.15676525e-02 7.82004535e-01 -4.87295657e-01
-4.99853015e-01 1.09800152e-01 7.39090025e-01 -5.32782674e-01
2.88975090e-01 -1.34287083e+00 -2.35703886e-01 3.97416800e-01
-3.18410039e-01 1.40820816e-01 1.07746577e+00 -6.46339536e-01
-7.79923439e-01 -7.06376255e-01 7.34935582e-01 -3.04811925e-01
5.42113066e-01 2.20878184e-01 -5.86255670e-01 3.34497809e-01
5.46437204e-02 -6.68269157e-01 4.36049938e-01 3.02316338e-01
2.11231679e-01 -1.38126433e-01 -1.33489645e+00 7.12792516e-01
4.40546274e-01 1.42464250e-01 -7.50492692e-01 1.49215192e-01
2.62086332e-01 5.97116612e-02 -9.43443298e-01 3.48319530e-01
6.76807046e-01 -8.18752289e-01 9.64675665e-01 -1.03266191e+00
1.55417114e-01 1.23309746e-01 -3.69737856e-02 -1.65789151e+00
-3.87237608e-01 -1.04706454e+00 -1.29958734e-01 8.35184455e-01
4.02732670e-01 -6.80123389e-01 3.56903195e-01 8.16130400e-01
5.26715457e-01 -1.10389018e+00 -1.30725873e+00 -7.47355402e-01
2.70608306e-01 -4.41243619e-01 9.92767334e-01 6.93099499e-01
-3.29218149e-01 1.20297082e-01 -3.69440794e-01 -1.71278313e-01
1.09934819e+00 1.60251781e-01 4.94753808e-01 -1.26673198e+00
-4.16706413e-01 -8.54136527e-01 1.79709718e-01 -4.17284578e-01
1.34127755e-02 -5.26727200e-01 -2.23992124e-01 -1.45932245e+00
1.91738158e-01 -8.32630634e-01 -3.13896924e-01 4.57164377e-01
-3.08760464e-01 -1.06160533e+00 3.93591255e-01 1.66584983e-01
-1.33747071e-01 3.71194959e-01 1.05332160e+00 3.53441611e-02
-5.09083509e-01 5.35510540e-01 -9.57678556e-01 7.93283105e-01
1.04353404e+00 -1.06756401e+00 -3.54353935e-01 -1.64674520e-01
4.21428472e-01 7.86229968e-01 5.22237003e-01 -3.45754862e-01
-1.43615976e-01 -1.26817882e+00 2.76917160e-01 -2.50733793e-01
-2.93332487e-01 -8.31357300e-01 5.88025928e-01 7.72069156e-01
-6.57228589e-01 -9.16853473e-02 7.98248351e-02 7.18979478e-01
3.92282039e-01 -6.11213565e-01 7.14040458e-01 -4.31687266e-01
-2.80677319e-01 -1.71597615e-01 -3.79212350e-01 2.09111422e-02
1.08958888e+00 -1.37111813e-01 3.47754993e-02 -5.25441647e-01
-7.52960026e-01 6.04560375e-01 1.44575775e-01 2.75383860e-01
1.25043884e-01 -1.17108035e+00 -1.17380381e+00 -4.19453412e-01
-4.22777623e-01 -2.37077594e-01 -2.20145166e-01 5.24703920e-01
-3.05436581e-01 8.87144983e-01 -2.45252222e-01 -2.80920327e-01
-8.03740382e-01 3.42043161e-01 5.36628723e-01 -8.28726411e-01
-2.37275511e-01 2.76339412e-01 -1.98268756e-01 -2.56509811e-01
1.25625864e-01 -1.35621682e-01 -2.51770943e-01 3.94171000e-01
3.07904959e-01 9.37672913e-01 -3.36742997e-01 -1.25319198e-01
-1.02737784e-01 2.17300192e-01 5.24357080e-01 -6.69104755e-01
1.55460751e+00 -4.98338826e-02 5.52060902e-02 3.96826416e-01
7.49803483e-01 -4.97937888e-01 -1.81555390e+00 -1.19740754e-01
6.18950665e-01 -5.81091821e-01 4.47752297e-01 -1.21882677e+00
-5.42061090e-01 4.59690630e-01 5.93285382e-01 1.87090382e-01
9.93159235e-01 -4.26946193e-01 2.58569862e-03 3.12707841e-01
4.40553486e-01 -1.37792575e+00 -4.25266981e-01 -1.37418225e-01
9.56773281e-01 -1.21648371e+00 4.42183971e-01 5.96157670e-01
-7.78013051e-01 9.53153312e-01 -2.44080480e-02 -3.84254940e-02
5.36588848e-01 -2.71554083e-01 -7.02051669e-02 1.23530738e-01
-1.05549693e+00 -4.63022217e-02 2.11862564e-01 4.93150145e-01
-1.33058131e-01 8.26368988e-01 -6.44057333e-01 1.69779509e-01
2.63333861e-02 1.32408798e-01 1.62573218e-01 8.58574152e-01
-5.21548092e-01 -1.26598823e+00 -9.41029847e-01 7.86663413e-01
-9.28523242e-01 1.16605908e-01 -7.49415765e-03 6.97086394e-01
9.31780636e-02 9.70008850e-01 -1.98049303e-02 2.94660121e-01
2.76628852e-01 1.44441098e-01 3.75427276e-01 -3.26403797e-01
-5.25469184e-01 3.18571627e-01 3.93978357e-01 -5.45558870e-01
-3.76114428e-01 -9.95982587e-01 -7.34847367e-01 -2.30143905e-01
-2.35129111e-02 1.38451338e-01 5.24578512e-01 1.11493218e+00
1.28323779e-01 1.56371564e-01 7.27477670e-01 -7.72002101e-01
-1.30850267e+00 -8.11521292e-01 -2.65303850e-01 -7.25110644e-04
3.83968174e-01 -8.98640692e-01 -2.69852430e-01 -5.50741434e-01] | [4.4045023918151855, 2.823850154876709] |
433b0234-ec21-4830-b3bd-8611e5af9854 | demucs-deep-extractor-for-music-sources-with | 1909.01174 | null | https://arxiv.org/abs/1909.01174v1 | https://arxiv.org/pdf/1909.01174v1.pdf | Demucs: Deep Extractor for Music Sources with extra unlabeled data remixed | We study the problem of source separation for music using deep learning with four known sources: drums, bass, vocals and other accompaniments. State-of-the-art approaches predict soft masks over mixture spectrograms while methods working on the waveform are lagging behind as measured on the standard MusDB benchmark. Our contribution is two fold. (i) We introduce a simple convolutional and recurrent model that outperforms the state-of-the-art model on waveforms, that is, Wave-U-Net, by 1.6 points of SDR (signal to distortion ratio). (ii) We propose a new scheme to leverage unlabeled music. We train a first model to extract parts with at least one source silent in unlabeled tracks, for instance without bass. We remix this extract with a bass line taken from the supervised dataset to form a new weakly supervised training example. Combining our architecture and scheme, we show that waveform methods can play in the same ballpark as spectrogram ones. | ['Francis Bach', 'Léon Bottou', 'Alexandre Défossez', 'Nicolas Usunier'] | 2019-09-03 | null | null | null | null | ['music-source-separation'] | ['music'] | [ 4.62540239e-01 -8.45227540e-02 4.32557054e-02 -3.04632895e-02
-1.60522127e+00 -1.05771446e+00 3.30937386e-01 -2.90060520e-01
-8.06963667e-02 5.75978398e-01 5.10061979e-01 9.45538804e-02
-2.82682329e-01 -1.56828582e-01 -7.98449755e-01 -8.44228625e-01
-2.73073047e-01 1.69285685e-01 1.69025689e-01 -3.49468887e-01
-6.89812824e-02 2.26084039e-01 -1.58770967e+00 6.70332193e-01
4.17335629e-01 1.14124489e+00 -7.87254497e-02 1.04439080e+00
1.06020562e-01 9.72027183e-01 -8.95296454e-01 -2.23332658e-01
4.95641202e-01 -8.56159031e-01 -5.23816407e-01 -1.07993908e-01
5.05528450e-01 -1.58047587e-01 -3.88578713e-01 8.99239719e-01
9.36548769e-01 9.85123590e-03 5.10138512e-01 -9.53850389e-01
-2.19076484e-01 1.66216743e+00 -5.94707131e-01 2.45011702e-01
9.94156152e-02 -1.10801049e-02 1.30776012e+00 -8.66352737e-01
3.57344121e-01 7.76797891e-01 1.03739202e+00 2.66084611e-01
-1.34274936e+00 -8.15587819e-01 -1.73809126e-01 1.62949711e-01
-1.17716706e+00 -9.19126749e-01 1.04818857e+00 -3.04294556e-01
6.73738062e-01 4.86818224e-01 2.85967827e-01 1.24785435e+00
-6.11850083e-01 9.67277467e-01 9.67534006e-01 -4.89130110e-01
1.36688724e-01 -1.71259373e-01 7.81471208e-02 1.14188559e-01
-3.00660133e-01 9.38619822e-02 -9.60837543e-01 -6.99267238e-02
4.62592244e-01 -6.74798548e-01 -3.96374881e-01 1.42109111e-01
-1.18982971e+00 3.26600164e-01 1.39202531e-02 5.55214405e-01
-2.62133837e-01 3.53400230e-01 4.20953959e-01 5.88556051e-01
4.85786766e-01 5.56356370e-01 -5.04673660e-01 -4.58037883e-01
-1.56385565e+00 2.81827480e-01 9.89398837e-01 7.08098888e-01
2.33011082e-01 7.47344077e-01 -2.94909954e-01 1.08472884e+00
-1.81117147e-01 4.13467586e-01 4.57302421e-01 -9.58500147e-01
3.98828208e-01 -3.04003775e-01 5.28210923e-02 -3.20645362e-01
-3.65811199e-01 -1.06660807e+00 -6.12937450e-01 2.19501227e-01
6.21045947e-01 -5.12010217e-01 -8.55988741e-01 1.79165494e+00
-1.45162776e-01 7.98324764e-01 -2.31726002e-02 8.42530489e-01
7.98603714e-01 6.93335176e-01 -7.67692327e-01 -3.14933926e-01
1.13368046e+00 -1.10996592e+00 -6.79125190e-01 -1.39844725e-02
2.89354343e-02 -1.15951812e+00 7.91115701e-01 1.09130681e+00
-1.49110603e+00 -7.22245932e-01 -1.25961173e+00 2.51450390e-01
1.36762664e-01 5.90255082e-01 2.22868696e-01 7.35257745e-01
-9.33804810e-01 1.09659147e+00 -7.28545070e-01 1.53548226e-01
1.14684731e-01 1.76547676e-01 -5.00818007e-02 5.24275839e-01
-1.06684017e+00 5.06422818e-01 7.01692998e-02 -1.17183499e-01
-1.40663433e+00 -1.02698755e+00 -6.11400664e-01 1.73243269e-01
3.35915089e-01 -9.67928171e-02 1.56517410e+00 -1.03586900e+00
-1.89685690e+00 6.90274358e-01 -1.44066382e-02 -8.56808782e-01
4.55587715e-01 -7.35296130e-01 -7.23972261e-01 7.14042336e-02
-2.28577435e-01 2.31189802e-02 1.23481905e+00 -1.36576116e+00
-5.96046925e-01 7.13248551e-02 -2.23383158e-01 -2.71444134e-02
-1.02628522e-01 2.72942990e-01 -2.33253106e-01 -1.09344423e+00
-9.52202827e-02 -7.96908855e-01 7.99323767e-02 -6.65134549e-01
-6.82462096e-01 1.46430194e-01 5.14775574e-01 -7.36531794e-01
1.35153496e+00 -2.34722209e+00 3.98606092e-01 1.02733269e-01
-6.38163276e-03 3.18122983e-01 -2.83377796e-01 4.56759185e-01
-4.14101392e-01 -1.51977316e-01 -2.93779492e-01 -7.29774237e-01
2.49622747e-01 -2.22055480e-01 -8.40828598e-01 3.94370347e-01
1.93176001e-01 5.53196549e-01 -6.95196271e-01 7.38466457e-02
-1.38474151e-01 3.01697910e-01 -4.78437960e-01 2.16150492e-01
-1.41109213e-01 4.68650907e-01 3.03290874e-01 4.43914443e-01
5.71586847e-01 1.66120470e-01 1.04536079e-01 -2.55579621e-01
-3.25392991e-01 8.31893325e-01 -1.47053874e+00 2.18879509e+00
-4.29918587e-01 8.84189725e-01 4.63586748e-01 -6.92202926e-01
9.13087606e-01 7.50086248e-01 5.90966403e-01 -2.10878570e-02
1.83344185e-01 4.98814732e-01 2.90776908e-01 -2.68464863e-01
3.64255220e-01 -3.16316992e-01 4.12338041e-02 6.46069944e-01
6.89988971e-01 -9.73143056e-02 1.26250267e-01 -8.40566903e-02
1.30244231e+00 4.41770881e-01 -5.31468242e-02 -8.08993727e-02
1.79846987e-01 -4.53650385e-01 4.01377082e-01 7.37969398e-01
1.70585930e-01 1.10017216e+00 5.19544423e-01 4.34289798e-02
-8.96249294e-01 -1.19087160e+00 1.17023930e-01 1.43955195e+00
-4.23205316e-01 -6.51611149e-01 -8.13924611e-01 -1.87575459e-01
-2.90532261e-01 6.84785843e-01 -4.15488660e-01 8.85898322e-02
-7.70767272e-01 -5.55950284e-01 1.26040840e+00 5.38090527e-01
1.50427874e-02 -1.00428748e+00 -2.90293783e-01 3.08331192e-01
-2.15651333e-01 -9.39872086e-01 -5.46187103e-01 7.99846470e-01
-3.80391389e-01 -6.44257784e-01 -8.67823243e-01 -5.75781405e-01
-3.73651564e-01 7.57602826e-02 1.15340209e+00 -5.00718176e-01
-1.74824079e-03 1.36980221e-01 -4.04037088e-01 -6.60643578e-01
-5.17082036e-01 2.60099173e-01 2.86376357e-01 3.97005439e-01
-3.10276970e-02 -1.19969845e+00 -3.27030718e-01 1.39876321e-01
-8.32734644e-01 -3.10548306e-01 5.45560479e-01 5.79681277e-01
4.41184402e-01 1.78312421e-01 9.21009779e-01 -6.77411735e-01
5.98863363e-01 -4.95716006e-01 -3.77138644e-01 -1.41595513e-01
-1.18539549e-01 3.23870182e-02 7.45196640e-01 -7.63094127e-01
-8.70316327e-01 3.03931743e-01 -4.84227926e-01 -7.93586493e-01
-2.58678138e-01 3.21847677e-01 -1.31271824e-01 3.50314856e-01
9.98043418e-01 8.65865499e-02 -4.41905379e-01 -1.06426370e+00
4.59834248e-01 7.39525676e-01 1.14562809e+00 -5.14431417e-01
9.96929109e-01 4.71280098e-01 -2.87158072e-01 -8.13061535e-01
-1.00722849e+00 -6.10176980e-01 -6.15271568e-01 -1.07945435e-01
6.38867199e-01 -9.41374063e-01 -4.93893474e-01 4.81893331e-01
-1.12127781e+00 -4.73880500e-01 -6.09552383e-01 7.69262850e-01
-7.81444430e-01 3.38111003e-03 -9.21952963e-01 -1.11149728e+00
-3.77018243e-01 -7.42840886e-01 1.11047399e+00 -9.82838273e-02
-3.55985016e-01 -4.81260389e-01 4.89161402e-01 1.17995478e-02
3.41095626e-01 2.10812122e-01 3.98341298e-01 -1.01332510e+00
-2.17362255e-01 -7.64003396e-02 3.67622644e-01 5.92968047e-01
1.42936125e-01 -8.95672217e-02 -1.89665842e+00 2.51245848e-03
3.45961511e-01 -4.08140659e-01 1.23596931e+00 4.05976027e-01
8.24929953e-01 -1.77312404e-01 1.89183488e-01 8.08278084e-01
9.50158596e-01 1.43300503e-01 6.35777712e-01 -1.19284108e-01
5.37920356e-01 3.14884901e-01 1.75687209e-01 4.46191043e-01
-2.52566040e-01 6.62173271e-01 2.68693686e-01 -5.81967942e-02
-4.82422024e-01 -3.22678149e-01 7.87777424e-01 1.19073319e+00
-2.91300893e-01 -1.43138558e-01 -4.32384968e-01 7.35038221e-01
-1.71864367e+00 -1.17655432e+00 -1.69152722e-01 2.15275717e+00
1.16119158e+00 3.99478197e-01 5.64509928e-01 9.28216696e-01
5.61639071e-01 3.11284482e-01 -3.15641373e-01 -9.28413495e-02
-3.05338919e-01 7.61790574e-01 3.65772128e-01 3.27259451e-01
-1.24254167e+00 5.46520591e-01 6.50986576e+00 1.25567329e+00
-1.18539500e+00 3.13214511e-01 1.09198168e-01 -5.92275202e-01
-2.59481013e-01 -2.79924989e-01 -5.81314445e-01 2.63640910e-01
1.31594813e+00 9.72958207e-02 7.99363375e-01 5.37203431e-01
3.00930366e-02 4.95157063e-01 -1.44748390e+00 1.09273028e+00
2.20008895e-01 -1.22267580e+00 -4.34561551e-01 -2.18602642e-01
6.62300944e-01 1.74464807e-01 2.17795908e-01 3.91720980e-01
3.15937012e-01 -1.17103302e+00 1.28185916e+00 3.62108052e-01
9.66266394e-01 -7.47994602e-01 5.46456337e-01 3.28064919e-01
-1.23086715e+00 8.14818665e-02 -6.76616430e-02 3.25340740e-02
2.11965963e-01 5.09955764e-01 -6.85365081e-01 8.62653017e-01
5.61315238e-01 7.18705714e-01 -1.71348125e-01 1.13389587e+00
-1.75695047e-01 1.47160113e+00 -5.24774492e-01 6.28644705e-01
1.17204733e-01 4.15735394e-02 9.57882702e-01 1.62500882e+00
4.51042086e-01 -2.54827380e-01 -1.01806924e-01 7.53714085e-01
-2.37030253e-01 -2.16664463e-01 -4.44533765e-01 -6.41360208e-02
3.32144797e-01 1.34556615e+00 -6.83632255e-01 -1.17174938e-01
-1.65894255e-01 7.31837809e-01 -1.06406271e-01 2.82572597e-01
-8.99296224e-01 -5.55731833e-01 5.13774335e-01 -1.68820806e-02
6.46281123e-01 5.27175553e-02 -3.16139042e-01 -9.41720784e-01
-1.04968555e-01 -1.13423681e+00 2.15415731e-01 -6.73151612e-01
-1.34024060e+00 8.70275080e-01 -3.52033168e-01 -1.69131768e+00
-5.58033228e-01 -4.30730313e-01 -7.88030863e-01 7.62966216e-01
-1.25541997e+00 -8.95748854e-01 9.10839215e-02 5.51639915e-01
7.26984382e-01 -4.39473003e-01 8.96558464e-01 6.41553760e-01
-7.69665465e-02 5.82729578e-01 1.15649298e-01 3.24132711e-01
9.83441591e-01 -1.45488596e+00 5.97317219e-01 8.66822600e-01
1.03327549e+00 3.04490864e-01 8.76658440e-01 -1.35730669e-01
-1.17317986e+00 -8.81299138e-01 4.78764057e-01 -4.92049575e-01
9.62919176e-01 -6.01704955e-01 -7.53197014e-01 5.59668064e-01
4.34531003e-01 -2.21795380e-01 8.13025296e-01 1.77824378e-01
-5.76290369e-01 -2.87057221e-01 -5.18893898e-01 4.06798333e-01
9.68953669e-01 -7.93021798e-01 -7.42263317e-01 6.65379912e-02
8.61011326e-01 -4.83944684e-01 -4.89813924e-01 2.66399831e-01
7.98348010e-01 -1.15956080e+00 1.05380595e+00 -6.76342964e-01
5.19216239e-01 -3.98031324e-01 -2.37857163e-01 -1.55663061e+00
-1.54030681e-01 -1.56879330e+00 -3.90495688e-01 1.55287671e+00
5.82680106e-01 4.57479358e-02 4.47302520e-01 -4.48119670e-01
-4.60321188e-01 -3.62974346e-01 -8.38042021e-01 -1.02087140e+00
2.79059932e-02 -9.95153308e-01 4.59951043e-01 7.11322725e-01
2.57662311e-02 5.77696919e-01 -9.45525467e-01 8.80473852e-02
4.46147650e-01 2.41820946e-01 7.13758886e-01 -1.12045348e+00
-1.00050330e+00 -5.53866744e-01 -1.07063897e-01 -1.08551550e+00
5.95068280e-03 -9.46859062e-01 2.44195417e-01 -1.08346939e+00
-2.44365841e-01 -1.27982035e-01 -8.51044893e-01 2.26045683e-01
1.09295562e-01 6.22948945e-01 4.66419458e-01 1.13676541e-01
-4.07088637e-01 2.23230943e-01 6.65329456e-01 -1.75533637e-01
-3.61022562e-01 5.48680723e-01 -7.25709379e-01 1.10165060e+00
7.64194012e-01 -6.97855890e-01 -2.35932231e-01 -3.02007020e-01
2.46935368e-01 3.92429084e-01 2.28985414e-01 -1.29097998e+00
1.71187103e-01 4.80136216e-01 1.07993998e-01 -7.29474664e-01
7.27528632e-01 -5.82484424e-01 2.29652390e-01 2.91590635e-02
-7.31626630e-01 -5.25176048e-01 2.64691919e-01 4.82679427e-01
-3.17453891e-01 -4.34531182e-01 6.54649496e-01 1.68641239e-01
9.36353728e-02 -1.65939689e-01 -3.00700456e-01 2.90199190e-01
2.77284026e-01 3.41545910e-01 -2.00969890e-01 -8.49037409e-01
-9.06509995e-01 -4.33736503e-01 -2.02592328e-01 2.95720100e-01
2.22433537e-01 -1.35420513e+00 -1.11909127e+00 4.67220023e-02
-2.50077963e-01 -4.61527199e-01 8.50609913e-02 9.53939617e-01
-9.64487195e-02 1.35731578e-01 -1.31639674e-01 -3.96955162e-01
-1.17374384e+00 4.20552433e-01 3.19923669e-01 -2.05822274e-01
-5.76692998e-01 1.04896593e+00 1.75973892e-01 -6.81825131e-02
7.11084366e-01 -4.86724287e-01 -1.02732137e-01 3.68606240e-01
7.38572061e-01 5.21191120e-01 3.39936048e-01 -5.04026830e-01
-2.53851146e-01 4.08133090e-01 4.03752387e-01 -6.29054904e-01
1.54578948e+00 2.26481751e-01 1.50232658e-01 1.13744283e+00
9.08910871e-01 1.04013121e+00 -1.29878461e+00 -2.66993076e-01
1.31091505e-01 -1.46095946e-01 -1.39518470e-01 -9.51751351e-01
-1.01856685e+00 1.01058006e+00 4.62461084e-01 6.94172144e-01
1.33046877e+00 -2.06200778e-02 8.26808035e-01 2.36160502e-01
8.70975256e-02 -9.58825648e-01 3.64114270e-02 5.21762788e-01
1.01034844e+00 -5.78709781e-01 -2.59745717e-01 4.37717373e-03
-7.56479561e-01 1.16098940e+00 -1.50444299e-01 -4.56709772e-01
6.95501864e-01 9.91471529e-01 2.06330046e-01 5.77792749e-02
-7.50372112e-01 -6.13428235e-01 6.13247991e-01 5.27930856e-01
6.04039192e-01 -6.25451878e-02 2.23606035e-01 1.29742491e+00
-6.93878651e-01 -6.65091053e-02 5.14169455e-01 4.33048576e-01
-3.87725234e-01 -9.84317601e-01 -5.49087882e-01 2.42471933e-01
-1.12955165e+00 -3.81871998e-01 -6.25649035e-01 2.71593988e-01
1.88512221e-01 1.22929084e+00 -1.22125305e-01 -6.46829724e-01
2.62634873e-01 4.97703135e-01 5.10686278e-01 -5.97350597e-01
-1.04882586e+00 9.66255426e-01 3.57139587e-01 -3.13535511e-01
-4.43904221e-01 -5.78582287e-01 -1.06112123e+00 1.67671710e-01
-3.27292651e-01 2.11679742e-01 5.12738407e-01 7.58698404e-01
-9.23661981e-03 1.10934055e+00 5.86070895e-01 -1.30737603e+00
-6.59129679e-01 -1.33676720e+00 -9.55881476e-01 2.27728218e-01
8.23190808e-01 -3.22007209e-01 -5.13519347e-01 5.24760842e-01] | [15.51541519165039, 5.562489032745361] |
1d92f690-e1c3-4793-a5ed-7673b84bcb2d | theres-a-time-and-place-for-reasoning-beyond-1 | null | null | https://aclanthology.org/2022.acl-long.81 | https://aclanthology.org/2022.acl-long.81.pdf | There’s a Time and Place for Reasoning Beyond the Image | Images are often more significant than only the pixels to human eyes, as we can infer, associate, and reason with contextual information from other sources to establish a more complete picture. For example, in Figure 1, we can find a way to identify the news articles related to the picture through segment-wise understandings of the signs, the buildings, the crowds, and more. This reasoning could provide the time and place the image was taken, which will help us in subsequent tasks, such as automatic storyline construction, correction of image source in intended effect photographs, and upper-stream processing such as image clustering for certain location or time.In this work, we formulate this problem and introduce TARA: a dataset with 16k images with their associated news, time, and location, automatically extracted from New York Times, and an additional 61k examples as distant supervision from WIT. On top of the extractions, we present a crowdsourced subset in which we believe it is possible to find the images’ spatio-temporal information for evaluation purpose. We show that there exists a 70% gap between a state-of-the-art joint model and human performance, which is slightly filled by our proposed model that uses segment-wise reasoning, motivating higher-level vision-language joint models that can conduct open-ended reasoning with world knowledge.The data and code are publicly available at https://github.com/zeyofu/TARA. | ['Dan Roth', 'Carl Vondrick', 'Ishaan Chandratreya', 'Ben Zhou', 'Xingyu Fu'] | null | null | null | null | acl-2022-5 | ['image-clustering'] | ['computer-vision'] | [-1.98838606e-01 1.61926553e-01 -2.51632392e-01 -4.20006543e-01
-9.48993564e-01 -5.94276011e-01 8.97297919e-01 2.25580841e-01
-4.53339398e-01 5.99557161e-01 7.72200763e-01 -5.56008108e-02
2.23936029e-02 -6.62535369e-01 -1.02469766e+00 -4.73302901e-01
3.02994490e-01 4.96720344e-01 4.48325992e-01 -1.36532590e-01
2.30126649e-01 1.06352493e-01 -1.47018135e+00 7.27215230e-01
5.25704503e-01 8.51903796e-01 4.10922140e-01 6.11592412e-01
9.87453237e-02 1.31823480e+00 -5.02959609e-01 -9.98101413e-01
1.07172668e-01 -5.77513652e-04 -8.18667531e-01 2.12997556e-01
6.59836054e-01 -6.43520772e-01 -5.31811655e-01 8.78492653e-01
2.01414362e-01 -3.86665352e-02 7.11413145e-01 -1.33175051e+00
-9.06847298e-01 8.05593073e-01 -8.27180147e-01 1.40719756e-01
4.04879242e-01 2.92684793e-01 1.07800376e+00 -9.88437951e-01
9.37783241e-01 1.06936574e+00 6.89825296e-01 -1.27347842e-01
-5.55555880e-01 -4.49019492e-01 2.06805691e-01 6.18234992e-01
-1.42057955e+00 -3.13211501e-01 4.85736817e-01 -5.86524963e-01
6.85993612e-01 3.35366845e-01 7.08523154e-01 1.22153723e+00
-2.43554533e-01 1.22949600e+00 1.06752503e+00 -4.41838712e-01
-1.50431797e-01 2.31701344e-01 4.48027477e-02 8.21835756e-01
-5.91704287e-02 -3.60599279e-01 -6.46868408e-01 1.04349092e-01
6.18223667e-01 1.26719013e-01 -3.98575395e-01 -7.10952431e-02
-1.56375039e+00 5.85705578e-01 8.49040508e-01 1.62426889e-01
-4.41562772e-01 1.42647311e-01 3.55147272e-02 -4.83220965e-01
5.31743586e-01 3.74231011e-01 -3.46491188e-01 -3.48445773e-02
-1.17773986e+00 2.43132353e-01 6.65763855e-01 9.94401336e-01
9.05245304e-01 -6.08402550e-01 -3.84961545e-01 6.54204607e-01
2.06632316e-01 6.67570651e-01 7.64221251e-02 -1.22628212e+00
8.46781731e-01 6.37936652e-01 4.15958941e-01 -1.25136113e+00
-2.54717559e-01 -4.15056618e-03 -4.87767994e-01 -1.40898332e-01
8.20596933e-01 -1.54937565e-01 -8.63635540e-01 1.32570755e+00
3.66122633e-01 3.95875275e-01 -1.07782364e-01 1.11296034e+00
8.77649128e-01 8.45141470e-01 -6.95283189e-02 1.70223281e-01
2.00689387e+00 -1.21289802e+00 -7.00913370e-01 -4.31354433e-01
4.41693574e-01 -7.99420893e-01 8.87515128e-01 2.01130003e-01
-7.22393930e-01 -4.41194892e-01 -7.39347935e-01 -4.38598126e-01
-7.31380284e-01 6.58769548e-01 5.01880407e-01 -5.09393513e-02
-7.84865737e-01 2.06276640e-01 -4.04576600e-01 -7.76848078e-01
5.03560960e-01 -4.43976969e-01 -3.57485145e-01 -4.23792481e-01
-1.30343974e+00 1.13096857e+00 3.40382129e-01 5.60281724e-02
-8.29217494e-01 -6.06083989e-01 -7.63614953e-01 -2.41614208e-01
6.58191860e-01 -5.80250084e-01 1.21607482e+00 -7.10858464e-01
-5.93843758e-01 1.14364469e+00 -1.65504962e-01 -6.53334141e-01
6.50854707e-01 -3.65353465e-01 -3.83204430e-01 3.15143675e-01
7.16817915e-01 9.17971849e-01 7.72076726e-01 -1.34878182e+00
-1.05607915e+00 -3.14956367e-01 3.65520418e-01 2.18929738e-01
-1.17635645e-01 2.34233767e-01 -1.13726079e+00 -5.92892289e-01
-1.80525690e-01 -9.46902931e-01 -1.35137737e-01 1.55830115e-01
-7.01545835e-01 -2.07369119e-01 4.39936131e-01 -1.29690170e+00
1.05356359e+00 -2.07787585e+00 -8.09566975e-02 -9.57991695e-04
2.89072543e-01 -5.13782129e-02 -1.03448257e-02 4.52715844e-01
1.72981083e-01 1.50946183e-02 2.65398994e-03 -3.34956408e-01
-2.44398527e-02 1.58711657e-01 -5.71963310e-01 4.81905878e-01
9.12736356e-02 1.03498626e+00 -1.10943997e+00 -7.06520200e-01
2.12284043e-01 2.98537254e-01 -5.79360835e-02 -6.29603788e-02
-3.78159761e-01 1.58589721e-01 -4.12660599e-01 6.62837267e-01
4.11020666e-01 -4.85419929e-01 -2.51132902e-02 -6.55638695e-01
-1.84966803e-01 -2.44272370e-02 -1.10269296e+00 1.57339990e+00
-1.08206473e-01 1.18561697e+00 -2.35239342e-01 -6.40854359e-01
5.52671552e-01 1.58879369e-01 3.15714419e-01 -7.69385815e-01
3.66166793e-02 -1.53746590e-01 -6.28620803e-01 -9.01940286e-01
8.57603312e-01 5.45978904e-01 -1.54728025e-01 3.33303303e-01
-2.88591087e-01 -9.39054638e-02 3.64281982e-01 5.79611778e-01
9.37596738e-01 2.27304026e-01 2.07871810e-01 2.96484053e-01
8.17396790e-02 7.29847014e-01 3.54113191e-01 7.90727079e-01
-1.32454365e-01 9.21019793e-01 4.08521771e-01 -5.48772693e-01
-1.38008332e+00 -7.25246906e-01 -1.84082519e-02 8.72863173e-01
3.82615477e-01 -4.97315586e-01 -7.44107544e-01 -5.46897173e-01
-1.01388171e-01 9.81041253e-01 -8.48238766e-01 5.49933612e-01
-4.87536103e-01 -3.99912387e-01 7.05954552e-01 5.21632671e-01
7.71893084e-01 -8.19925189e-01 -4.47076470e-01 -1.21725954e-01
-9.72104132e-01 -1.53578854e+00 -4.52741295e-01 -2.46544376e-01
-1.68763787e-01 -1.30177939e+00 -9.38014030e-01 -5.40264070e-01
7.53900230e-01 5.46374321e-01 1.04856539e+00 -5.23418449e-02
-2.47537151e-01 5.21276534e-01 -4.47289944e-01 -4.75588828e-01
-7.61343539e-02 -3.23795378e-01 -2.37214163e-01 1.25293866e-01
4.02557164e-01 -1.16744891e-01 -7.42255390e-01 4.19295192e-01
-6.24404073e-01 5.29633582e-01 5.59962273e-01 3.01063836e-01
5.59485674e-01 2.27807641e-01 -1.52797103e-01 -6.92786634e-01
1.91819966e-01 -7.93880939e-01 -2.96367466e-01 5.53011954e-01
-1.08951233e-01 -1.13865510e-01 1.35392264e-01 -2.90528029e-01
-1.24445868e+00 2.08185345e-01 3.75000894e-01 -4.31662142e-01
-3.03064883e-01 5.87545514e-01 1.49282619e-01 5.50684810e-01
8.19704771e-01 1.72072619e-01 -2.88213551e-01 -2.92955190e-01
8.84282708e-01 8.14107478e-01 8.48170638e-01 -4.34053481e-01
7.82550037e-01 8.98987651e-01 -5.37329197e-01 -6.37836754e-01
-1.50187123e+00 -8.64077508e-01 -7.23683178e-01 -5.48904359e-01
1.22012842e+00 -1.37478256e+00 -3.18449438e-01 3.12421262e-01
-1.46264625e+00 -3.28753650e-01 -4.56236675e-02 4.34710324e-01
-4.56955045e-01 2.79560864e-01 -5.84581733e-01 -7.39359617e-01
8.04682449e-02 -9.71509397e-01 1.15309978e+00 1.83043584e-01
-3.16409171e-01 -7.18385339e-01 -1.92254260e-01 1.21376431e+00
-9.55110863e-02 2.43160784e-01 4.47173238e-01 -4.88465279e-01
-9.26492989e-01 -2.35792562e-01 -7.26244926e-01 1.64114386e-02
-2.31861129e-01 1.97388425e-01 -1.07677937e+00 3.18782210e-01
-3.80007356e-01 -1.89603671e-01 9.33873296e-01 4.54103231e-01
9.24456298e-01 -4.08369780e-01 -5.42628527e-01 2.75565028e-01
1.26580226e+00 -2.44972870e-01 8.48317444e-01 6.19777322e-01
9.63205516e-01 9.00841475e-01 1.02876949e+00 4.34127182e-01
9.69893336e-01 7.29561329e-01 5.77928424e-01 -3.09129208e-01
-4.64165807e-01 -5.16221404e-01 3.47401559e-01 4.97325927e-01
-3.90425354e-01 -3.21066618e-01 -1.08824587e+00 1.01809597e+00
-2.39846349e+00 -1.36301243e+00 -5.79437137e-01 1.69570100e+00
7.30532885e-01 -1.68991253e-01 2.15853173e-02 -3.12969208e-01
8.91360223e-01 2.40522966e-01 -3.85995388e-01 3.26423556e-01
-3.49280179e-01 -6.64559960e-01 9.81992245e-01 3.43251467e-01
-1.14979315e+00 8.92045379e-01 5.62036324e+00 1.12202656e+00
-5.60573041e-01 2.30343297e-01 8.39493334e-01 -1.74246561e-02
-1.83922023e-01 2.90266573e-01 -8.39457214e-01 4.06605542e-01
5.41583240e-01 2.90014327e-01 3.86510551e-01 7.62657762e-01
5.03737509e-01 -5.74070394e-01 -1.01994014e+00 1.06524265e+00
4.52102721e-01 -1.71445107e+00 -1.32815972e-01 4.14770190e-03
9.34144020e-01 2.28256330e-01 -2.53551185e-01 1.00505278e-01
5.08429825e-01 -8.05379033e-01 1.26996517e+00 8.33836019e-01
4.64316249e-01 -2.75140733e-01 6.85652792e-01 5.70650697e-01
-1.10764921e+00 -1.54518306e-01 -2.24408314e-01 -1.51250213e-01
3.26621562e-01 7.42271304e-01 -1.05568862e+00 5.61618209e-01
9.25367236e-01 1.01989663e+00 -9.19609487e-01 1.06315696e+00
-8.30616891e-01 5.40779233e-01 -2.91644663e-01 -1.05599433e-01
3.63723904e-01 -5.23336455e-02 4.06344920e-01 1.17675352e+00
5.47273159e-01 2.41923943e-01 1.11112095e-01 8.73366356e-01
-1.27909154e-01 -1.16909452e-01 -6.13536537e-01 3.26034874e-01
5.09748459e-01 1.35580003e+00 -9.72039878e-01 -6.00553036e-01
-5.33384264e-01 1.04797721e+00 4.69484270e-01 6.73523128e-01
-1.31429935e+00 4.33284231e-03 4.24428493e-01 4.30975467e-01
3.40349495e-01 -1.14108615e-01 -2.88823783e-01 -1.20620799e+00
1.50270075e-01 -5.95069170e-01 4.08540279e-01 -1.80183709e+00
-1.18868279e+00 1.76011249e-01 1.33047745e-01 -1.19297278e+00
1.07895851e-01 -4.98039007e-01 -3.96464795e-01 5.56848645e-01
-1.57045615e+00 -1.50051129e+00 -5.94107032e-01 6.06376112e-01
6.84284389e-01 1.19185753e-01 2.43051305e-01 2.14243919e-01
-5.02223432e-01 -1.47172054e-02 1.61359627e-02 7.42180049e-01
1.01820838e+00 -1.11140287e+00 2.60243088e-01 1.19958460e+00
6.24816597e-01 1.98941365e-01 7.96423674e-01 -7.33276963e-01
-1.01153409e+00 -1.15561974e+00 1.05163133e+00 -1.06566846e+00
1.00554442e+00 -1.30051717e-01 -4.86189038e-01 7.68621504e-01
3.75613570e-01 -2.49422550e-01 3.36912602e-01 7.43789077e-02
-4.50928777e-01 -2.00696327e-02 -7.87338078e-01 9.03726637e-01
9.27985370e-01 -6.18019044e-01 -6.92427397e-01 8.56910706e-01
7.38261640e-01 -5.12792528e-01 -5.76684535e-01 -1.48117900e-01
3.56132805e-01 -8.85306656e-01 1.10617363e+00 -1.63242280e-01
9.74377871e-01 -5.71411550e-01 -1.31040826e-01 -1.07395387e+00
-3.21954280e-01 -9.96772796e-02 1.72536761e-01 1.50766349e+00
6.99764669e-01 -7.28763118e-02 4.50937152e-01 9.95437443e-01
-7.67602678e-03 -3.88123184e-01 -7.16585636e-01 -3.39578867e-01
-5.23526788e-01 -9.03469265e-01 4.61734533e-01 1.09853351e+00
-2.20334008e-02 2.46122658e-01 -6.75740778e-01 6.19012058e-01
4.49971229e-01 2.07982987e-01 9.43680465e-01 -7.16706336e-01
-9.61191282e-02 -2.19148412e-01 -1.87467083e-01 -1.12193036e+00
-1.26256526e-01 -5.86552382e-01 1.75016448e-01 -2.18029904e+00
5.31892419e-01 -4.13186699e-01 1.23291083e-01 8.53983343e-01
-2.31969386e-01 4.18252319e-01 3.08671236e-01 6.06955647e-01
-9.52578664e-01 1.81454584e-01 1.32612741e+00 -4.22358006e-01
2.46736228e-01 -3.16483647e-01 -6.55538559e-01 1.05702329e+00
4.74370986e-01 -5.14145374e-01 -1.50132686e-01 -7.38277078e-01
5.09581149e-01 5.86599968e-02 8.61078084e-01 -9.28526402e-01
6.31051898e-01 -3.35509539e-01 5.55817902e-01 -7.76236773e-01
5.22084773e-01 -9.26867783e-01 1.93877921e-01 9.86649767e-02
-3.25345844e-01 -8.79768878e-02 -1.44431114e-01 7.97803342e-01
-1.91962197e-01 -3.77181917e-01 8.62972289e-02 -3.29327762e-01
-1.26089180e+00 1.15263157e-01 -1.87954232e-01 -3.40179652e-02
1.19733381e+00 -2.87613392e-01 -8.54274094e-01 -6.43439829e-01
-5.37375748e-01 5.02133965e-01 4.75180000e-01 4.08365369e-01
4.50811774e-01 -1.24366331e+00 -9.34937775e-01 -4.82060641e-01
4.15179193e-01 7.11298510e-02 3.91673654e-01 1.02861345e+00
-6.17759466e-01 2.10168093e-01 1.69505551e-01 -5.37053645e-01
-1.17497885e+00 6.80300057e-01 -1.33025780e-01 -1.48915261e-01
-7.16182828e-01 6.76106572e-01 5.21285348e-02 -4.76584733e-02
1.68521792e-01 -5.35055101e-01 -4.03100133e-01 5.10485291e-01
6.90160155e-01 1.41774610e-01 -3.06573749e-01 -8.65817547e-01
-2.74082303e-01 5.30723751e-01 2.55668193e-01 -1.99039608e-01
1.27467036e+00 -4.41078931e-01 -4.40120287e-02 3.56692880e-01
8.06196332e-01 1.59508646e-01 -1.45166636e+00 -5.66014349e-01
-2.19946578e-01 -5.54638922e-01 -9.33051333e-02 -8.99008751e-01
-9.43470597e-01 5.78788340e-01 2.06450194e-01 2.63700217e-01
8.13488066e-01 4.67220396e-01 7.62154996e-01 3.00451577e-01
3.24644268e-01 -1.34954619e+00 1.66509748e-01 3.80817413e-01
1.02550232e+00 -1.63125038e+00 2.33513966e-01 -3.06388378e-01
-1.11666930e+00 9.84012902e-01 4.60107118e-01 8.04235265e-02
5.12521327e-01 4.98512015e-02 3.51531692e-02 -4.20098931e-01
-4.91245389e-01 -6.53792024e-01 3.50532204e-01 6.77599072e-01
-5.28454781e-02 2.43118644e-01 2.80357003e-01 5.70175350e-01
-2.20573917e-01 -1.87753245e-01 5.57782710e-01 4.18598592e-01
-4.87628073e-01 -4.52029675e-01 -6.49704456e-01 2.69129127e-01
-9.28879529e-02 -2.14501485e-01 -3.58682454e-01 7.25006759e-01
4.35962111e-01 1.04218745e+00 9.05548260e-02 -1.32846877e-01
3.23237747e-01 -1.69298813e-01 1.25403568e-01 -3.39689970e-01
-1.17300548e-01 -2.45380357e-01 6.23418748e-01 -5.31997859e-01
-8.92045557e-01 -6.97553396e-01 -1.11082864e+00 -2.68760383e-01
-1.30083844e-01 -2.63688445e-01 6.40649080e-01 1.13082170e+00
2.56222934e-01 3.62435699e-01 1.70639426e-01 -1.01302946e+00
2.21374065e-01 -8.67606699e-01 -2.98167676e-01 7.48472154e-01
4.72117253e-02 -4.19405758e-01 -2.60575622e-01 7.02491879e-01] | [10.693341255187988, 1.264917016029358] |
964bde9e-aec2-49c9-8d3b-ca845ac801e1 | a-stacked-dcnn-to-predict-the-rul-of-a | null | null | http://papers.phmsociety.org/index.php/phmconf/article/view/3110 | http://papers.phmsociety.org/index.php/phmconf/article/download/3110/1838 | A stacked DCNN to predict the RUL of a turbofan engine | This paper presents the data-driven techniques and methodologies used to predict the remaining useful life (RUL) of a fleet of aircraft engines that can suffer failures of diverse nature. The solution presented is based on two Deep Convolutional Neural Networks (DCNN) stacked in two levels. The first DCNN is used to extract a low-dimensional feature vector using the normalized raw data as input. The second DCNN ingests a list of vectors taken from the former DCNN and estimates the RUL. Model selection was carried out by means of Bayesian optimization using a repeated random subsampling validation approach. The proposed methodology was ranked in the third place of the 2021 PHM Conference Data Challenge. | ['Joaquín Borrego-Díaz', 'Juan Galán-Páez', 'David Solís-Martín'] | 2021-11-24 | null | null | null | annual-conference-pf-the-phm-society-2021-11 | ['remaining-useful-lifetime-estimation'] | ['time-series'] | [-8.29327293e-03 -2.27412969e-01 1.64289534e-01 -4.74752992e-01
-2.11850017e-01 -6.75808564e-02 4.19323146e-01 3.33499573e-02
-3.84992421e-01 9.19803977e-01 -6.56506270e-02 -4.14897174e-01
-1.10538208e+00 -8.09131324e-01 -4.02012259e-01 -8.90564322e-01
-4.04321164e-01 6.17849648e-01 4.75473814e-02 -1.39671803e-01
2.27764815e-01 1.26184571e+00 -1.84302878e+00 2.80532122e-01
1.60184771e-01 1.66716576e+00 -2.49263123e-01 7.42914975e-01
4.19749826e-01 8.83147657e-01 -8.56481552e-01 2.16442168e-01
5.22119582e-01 6.32651821e-02 -5.23357034e-01 -2.41695374e-01
-1.07083045e-01 -2.45997444e-01 -3.52968544e-01 4.80871350e-01
4.64975119e-01 5.68294883e-01 8.51422071e-01 -1.49045110e+00
3.09375882e-01 3.69328916e-01 2.58470565e-01 4.71920192e-01
-2.86248684e-01 2.85934564e-02 5.53043902e-01 -9.67335999e-01
2.75073409e-01 1.14054644e+00 6.87824726e-01 1.94291130e-01
-9.17042136e-01 -4.57605958e-01 -5.22698760e-01 2.18632400e-01
-1.31664264e+00 -2.71473765e-01 8.40816855e-01 -7.08797574e-01
1.33227265e+00 1.73734888e-01 7.01418221e-01 9.35837030e-01
1.02060425e+00 1.05120435e-01 9.31526542e-01 -2.61504620e-01
6.71441019e-01 -1.65291533e-01 3.12728822e-01 5.27447581e-01
6.29415512e-01 9.35041547e-01 -5.05997062e-01 -2.20863581e-01
3.96084189e-01 9.29051265e-03 3.33198369e-01 -3.30256343e-01
-7.51459420e-01 8.32802176e-01 2.22467750e-01 2.60825813e-01
-8.12317908e-01 3.37364376e-01 6.46413267e-01 3.36544484e-01
2.09965110e-01 7.48893082e-01 -9.33393419e-01 1.53918087e-01
-1.19742703e+00 3.62207472e-01 8.63860548e-01 5.47312200e-01
3.27119738e-01 4.47720945e-01 -3.54368985e-01 2.69948274e-01
5.43321967e-01 1.40516341e-01 1.98725075e-01 -9.68629241e-01
3.54570225e-02 4.76915210e-01 2.67497033e-01 -8.32544744e-01
-7.31226623e-01 -7.14783490e-01 -8.38471055e-01 7.31000006e-01
-5.89166582e-02 -5.62603116e-01 -1.17091727e+00 8.94754827e-01
-1.32518023e-01 -5.41190445e-01 -7.23320395e-02 5.74857831e-01
5.38860559e-01 6.62234902e-01 -1.26809791e-01 -3.09014291e-01
7.26666451e-01 -4.73343015e-01 -6.21087670e-01 2.37861469e-01
4.52226289e-02 -2.45093465e-01 1.04033306e-01 1.16295338e+00
-9.11039174e-01 -8.02554071e-01 -1.75216782e+00 6.05516315e-01
-6.62350237e-01 3.39583635e-01 1.48248911e-01 6.13807559e-01
-9.84177053e-01 1.10269856e+00 -7.64830470e-01 1.26267061e-01
4.04487789e-01 7.67320991e-01 -2.37665787e-01 -9.42252129e-02
-1.27443552e+00 1.05630279e+00 8.43732774e-01 4.48548734e-01
-1.70389700e+00 -3.59706372e-01 -5.66027641e-01 1.87884104e-02
2.08878443e-01 -4.30033714e-01 8.36773992e-01 -5.12739897e-01
-1.33100057e+00 4.44984324e-02 4.63160783e-01 -6.45575523e-01
5.01375496e-01 -2.61617184e-01 -7.64639676e-01 -1.84270650e-01
-3.68901253e-01 2.67794073e-01 8.29430640e-01 -1.18391514e+00
-7.79961526e-01 -1.68068439e-01 -2.69555897e-01 -4.93398041e-01
-1.19307213e-01 -1.00157268e-01 8.02924633e-02 -4.20951456e-01
-3.86493802e-02 -5.96318245e-01 -2.38315254e-01 -7.78014600e-01
-5.57467580e-01 -3.53227854e-01 6.50690258e-01 -6.84955955e-01
1.43402541e+00 -1.70348322e+00 3.06055486e-01 5.36783755e-01
5.03040403e-02 3.27496789e-02 3.41691792e-01 8.40825200e-01
-3.44655603e-01 3.56101096e-02 -9.06198770e-02 -7.47241685e-03
-6.60669878e-02 3.57168056e-02 2.96354257e-02 5.31664789e-01
4.65135396e-01 1.30450010e-01 -4.76762295e-01 -2.18344748e-01
3.47582042e-01 2.94662535e-01 -1.80222327e-03 4.18674916e-01
-1.26203790e-01 9.63700488e-02 -3.44053447e-01 9.05269444e-01
5.64066768e-01 4.55810487e-01 -3.71143520e-02 -5.08079886e-01
-2.84990937e-01 -2.92274684e-01 -1.03830266e+00 1.02817106e+00
-3.87576848e-01 7.00649142e-01 -3.05122048e-01 -9.47848320e-01
1.37201071e+00 3.58497500e-01 7.81582952e-01 -3.23478103e-01
6.89534068e-01 2.08179921e-01 1.31437391e-01 -5.76783061e-01
4.16110516e-01 -3.11439961e-01 -3.63091290e-01 -1.09905258e-01
4.58395571e-01 8.71295780e-02 3.48777503e-01 -2.74311155e-01
1.40891635e+00 2.52371311e-01 -9.75224376e-03 -4.86869156e-01
5.44150412e-01 2.63051242e-02 9.53717411e-01 6.50674105e-01
-1.93607897e-01 2.40353301e-01 7.98929811e-01 -1.15070236e+00
-1.08054960e+00 -1.04076552e+00 -1.40796751e-01 4.60707694e-01
-5.39350092e-01 8.20161309e-03 -3.56447905e-01 -7.02384949e-01
2.06146896e-01 9.34223533e-01 -8.51301551e-01 -4.82991099e-01
-2.99875438e-01 -7.21965849e-01 4.40046579e-01 6.20313883e-01
9.10665691e-02 -1.08911622e+00 -1.03226483e+00 2.35459253e-01
7.31652558e-01 -5.33491910e-01 7.12469101e-01 8.39830935e-01
-8.06859195e-01 -1.17866039e+00 -1.55605793e-01 -3.21106553e-01
4.47581291e-01 -6.60725236e-01 1.25964367e+00 6.34349091e-03
-6.39722347e-01 -2.34350506e-02 -2.28655025e-01 -9.29338098e-01
-4.42268431e-01 9.62980539e-02 4.92308915e-01 -9.63631123e-02
3.98193032e-01 -2.06951857e-01 -3.59196216e-01 1.94982678e-01
-8.16324770e-01 -6.04036510e-01 8.58832300e-01 9.17826533e-01
5.05860865e-01 1.01868153e+00 6.88552976e-01 -4.16731268e-01
8.65820527e-01 -7.17077971e-01 -8.78333688e-01 1.21684335e-01
-8.85308743e-01 5.00183851e-02 7.70769119e-01 -1.39901843e-02
-7.07876086e-01 1.81411654e-01 -2.08501536e-02 -7.25033820e-01
-3.50055844e-01 6.90881968e-01 -1.84704497e-01 3.24465483e-01
4.22172278e-01 -3.60729426e-01 2.19105836e-02 -5.64563036e-01
-2.28502527e-01 5.66714525e-01 3.73054415e-01 -4.01754797e-01
5.80978751e-01 -9.68127847e-02 6.92901254e-01 -5.84025085e-01
-4.24679637e-01 -2.51819380e-02 -8.35994899e-01 -1.01895404e+00
7.16314554e-01 -4.26793069e-01 -7.18887091e-01 3.79992515e-01
-8.33479047e-01 3.03732697e-02 -4.70409125e-01 6.49588108e-01
-5.72183013e-01 -4.16447282e-01 -2.17045844e-01 -1.18346989e+00
-3.38611782e-01 -1.11982799e+00 4.53220874e-01 1.91238314e-01
6.15638308e-02 -4.83379483e-01 1.10884748e-01 -5.62422648e-02
4.51349765e-01 7.15259850e-01 1.12701356e+00 -1.18536329e+00
-1.36140361e-01 -1.12998998e+00 3.36104572e-01 1.13366139e+00
-9.15347189e-02 3.88859212e-01 -8.80529344e-01 -6.68655813e-01
-1.60727929e-02 -3.16580951e-01 7.56037593e-01 6.98354661e-01
1.30173123e+00 1.95254236e-01 -1.07980445e-01 1.99919060e-01
1.63512039e+00 5.92758775e-01 4.60048199e-01 4.03806150e-01
2.47774318e-01 7.93990254e-01 8.65614414e-01 5.73494613e-01
-6.15897417e-01 1.39097542e-01 1.01877582e+00 2.69231230e-01
3.21261585e-01 3.79365832e-01 2.04212248e-01 6.17107987e-01
-1.57046765e-01 -4.44303215e-01 -1.26844382e+00 5.91921806e-01
-1.57539940e+00 -6.74056411e-01 7.44078085e-02 2.19298506e+00
-2.26656385e-02 8.18211079e-01 1.22511230e-01 5.33816218e-01
3.41391921e-01 -5.29192574e-02 -5.47168851e-01 -6.18490696e-01
1.49367124e-01 3.03745598e-01 8.11126351e-01 1.12889856e-01
-1.16346490e+00 1.00534119e-01 6.75596046e+00 6.74083292e-01
-5.41845798e-01 -2.00946644e-01 7.22764194e-01 -5.35233855e-01
1.75877199e-01 -7.22340196e-02 -8.26619804e-01 3.93422991e-01
1.67506289e+00 1.08239107e-01 2.67340153e-01 6.45445645e-01
2.22619325e-01 -2.64469385e-01 -1.07330430e+00 3.84012192e-01
7.93609023e-03 -1.24787474e+00 -1.10251166e-01 3.00275624e-01
6.81113422e-01 4.33958396e-02 -3.01408637e-02 3.92341048e-01
2.29430899e-01 -1.33442843e+00 6.45752490e-01 1.54256511e+00
4.99477983e-01 -1.61762190e+00 1.58609700e+00 1.00643590e-01
-6.51991785e-01 -8.31954420e-01 -4.11050946e-01 1.18695289e-01
5.80521375e-02 9.78168309e-01 -7.12017417e-01 9.49057996e-01
1.05682278e+00 4.05371666e-01 -5.36183536e-01 1.11280024e+00
4.64764744e-01 5.73710024e-01 -1.17301993e-01 -1.00903556e-01
1.01086564e-01 1.10605434e-01 6.84665620e-01 7.86627054e-01
5.69426537e-01 -6.66385353e-01 3.28366319e-03 6.15291536e-01
3.18807960e-02 -4.64551181e-01 -6.11068189e-01 -1.28202721e-01
6.07312500e-01 1.20967686e+00 -6.15667760e-01 -1.28790185e-01
8.49770755e-02 1.06444426e-01 -2.70653486e-01 1.42705783e-01
-6.01017833e-01 -7.49491155e-01 4.11287218e-01 3.96568030e-02
3.02175671e-01 -2.82776445e-01 -1.01935595e-01 -8.90095234e-02
-2.65360057e-01 -5.66470504e-01 1.92292020e-01 -6.33028805e-01
-1.35523856e+00 1.01943219e+00 4.94585335e-01 -1.39957047e+00
-5.08084059e-01 -1.33240485e+00 -3.77247363e-01 1.03815937e+00
-8.72187912e-01 -6.87083066e-01 -8.70695338e-02 2.62200832e-03
8.19856524e-01 -1.14605689e+00 5.83690941e-01 1.23848349e-01
-8.63468468e-01 -2.51361039e-02 4.35962737e-01 -1.96105540e-01
2.55238682e-01 -1.12015986e+00 5.29877171e-02 7.95345604e-01
-8.02481055e-01 3.31693411e-01 9.67900157e-01 -9.25841093e-01
-1.30884957e+00 -1.26192415e+00 6.46664143e-01 -4.26112175e-01
4.49922442e-01 1.64579879e-02 -3.76112908e-01 2.33659551e-01
1.85962990e-01 -2.44910438e-02 6.02571368e-01 -2.29542062e-01
5.35734951e-01 -5.15065372e-01 -1.31302774e+00 -1.59748569e-01
3.87174487e-02 -4.43235338e-02 -3.07122082e-01 9.35956463e-02
4.32131767e-01 7.50874728e-02 -1.40342319e+00 8.28334093e-01
7.19122350e-01 -1.04150534e+00 9.76427257e-01 -9.68381107e-01
3.55939537e-01 -2.35064134e-01 -4.65522975e-01 -1.19386995e+00
-4.31783617e-01 -1.80283457e-01 -4.52469975e-01 9.92162526e-01
2.29706451e-01 -2.67500699e-01 4.53462213e-01 3.48348588e-01
-2.26655930e-01 -1.30075943e+00 -1.30339026e+00 -1.04681480e+00
-6.63037598e-02 -3.72405231e-01 7.41351962e-01 1.17933549e-01
-9.92283940e-01 -1.78540275e-01 -3.85810554e-01 2.54237264e-01
6.98409319e-01 -6.32134080e-01 6.53396100e-02 -1.84832585e+00
2.31300429e-01 -3.48160237e-01 -5.66338718e-01 3.69999886e-01
-1.22415368e-02 -3.58111799e-01 4.34735507e-01 -1.34327316e+00
-3.34308058e-01 -1.32216364e-01 -1.20232427e+00 -6.91482872e-02
6.21847272e-01 -1.93627238e-01 -1.47108167e-01 -6.66257069e-02
-1.47242457e-01 4.73515064e-01 2.85339177e-01 -1.81646571e-01
1.03419028e-01 3.89117151e-01 -5.55195333e-03 6.10462010e-01
1.10890388e+00 -6.70616508e-01 -4.70945388e-01 7.45154321e-02
4.70362127e-01 1.76935881e-01 5.52357018e-01 -1.69867516e+00
5.95346205e-02 -1.53880447e-01 1.01642334e+00 -1.28628075e+00
2.53193080e-01 -1.10053575e+00 6.68825805e-01 6.81332350e-01
-2.32679367e-01 5.05774379e-01 3.62966716e-01 6.74871683e-01
-1.42805591e-01 -5.66750705e-01 5.67945957e-01 -3.58301885e-02
-5.33335268e-01 4.21993360e-02 -7.22886741e-01 -7.52352178e-01
9.59527969e-01 -2.76642442e-01 1.54555991e-01 2.39407554e-01
-9.40309286e-01 1.65816158e-01 7.30703473e-02 4.75883216e-01
8.67879450e-01 -1.41834211e+00 -6.74632728e-01 2.71336198e-01
-1.35907263e-01 -2.95896769e-01 2.96550632e-01 5.82017064e-01
-6.71365917e-01 7.20940530e-01 -7.56042659e-01 -1.65167704e-01
-7.05703735e-01 5.64541101e-01 7.67968178e-01 -4.58670318e-01
-9.23023373e-02 9.05580640e-01 -8.28498900e-01 -2.73010463e-01
2.83265203e-01 -1.08620010e-01 -5.55343986e-01 3.95865142e-01
1.15958698e-01 1.07881236e+00 7.62145758e-01 -4.86151934e-01
-5.18752635e-01 6.76184818e-02 3.03886116e-01 -2.07511991e-01
1.61060143e+00 3.57464284e-01 -2.82216460e-01 8.44884992e-01
1.19304192e+00 -7.02800155e-01 -1.19437540e+00 4.87889707e-01
2.60812223e-01 -1.41818151e-01 7.58192778e-01 -1.22704244e+00
-1.08815253e+00 7.07641363e-01 1.05369413e+00 4.10678595e-01
1.37844217e+00 -5.79169869e-01 2.84273237e-01 5.29254436e-01
3.11287582e-01 -1.53772950e+00 -2.09335595e-01 3.65493178e-01
1.16261292e+00 -9.00168478e-01 3.69831204e-01 5.49082220e-01
-3.22490513e-01 1.66255522e+00 5.38661897e-01 -3.37905914e-01
1.02603054e+00 2.06512749e-01 -4.33244526e-01 -6.57092690e-01
-1.09890211e+00 2.39144534e-01 4.34585869e-01 4.06841338e-01
1.46456450e-01 1.03772216e-01 -2.82887250e-01 8.43653023e-01
1.56871185e-01 2.31535316e-01 3.56876969e-01 1.23579419e+00
-4.27745491e-01 -8.15658689e-01 -4.74852473e-01 1.05846334e+00
-3.69868785e-01 3.42665792e-01 -2.83686459e-01 8.24848831e-01
5.90481162e-01 1.11416197e+00 4.23612595e-02 -1.04001975e+00
5.27882814e-01 2.23596036e-01 -2.03719120e-02 -1.60072073e-01
-6.60579324e-01 -2.89277047e-01 3.37615579e-01 -5.92701316e-01
-3.18215936e-02 -7.63158143e-01 -6.74399316e-01 6.01181760e-02
-1.41207546e-01 5.69738671e-02 1.13195753e+00 7.69230664e-01
9.84079242e-02 1.38736844e+00 9.44778025e-01 -1.19444942e+00
-6.43874645e-01 -1.17101097e+00 -8.55306268e-01 -3.78273070e-01
6.04781508e-01 -1.12615180e+00 -5.25591195e-01 -2.76896000e-01] | [6.710403919219971, 2.4483323097229004] |
473bf1a9-680d-4e33-a613-9ca84e7fa525 | s-nerf-neural-radiance-fields-for-street | 2303.00749 | null | https://arxiv.org/abs/2303.00749v1 | https://arxiv.org/pdf/2303.00749v1.pdf | S-NeRF: Neural Radiance Fields for Street Views | Neural Radiance Fields (NeRFs) aim to synthesize novel views of objects and scenes, given the object-centric camera views with large overlaps. However, we conjugate that this paradigm does not fit the nature of the street views that are collected by many self-driving cars from the large-scale unbounded scenes. Also, the onboard cameras perceive scenes without much overlapping. Thus, existing NeRFs often produce blurs, 'floaters' and other artifacts on street-view synthesis. In this paper, we propose a new street-view NeRF (S-NeRF) that considers novel view synthesis of both the large-scale background scenes and the foreground moving vehicles jointly. Specifically, we improve the scene parameterization function and the camera poses for learning better neural representations from street views. We also use the the noisy and sparse LiDAR points to boost the training and learn a robust geometry and reprojection based confidence to address the depth outliers. Moreover, we extend our S-NeRF for reconstructing moving vehicles that is impracticable for conventional NeRFs. Thorough experiments on the large-scale driving datasets (e.g., nuScenes and Waymo) demonstrate that our method beats the state-of-the-art rivals by reducing 7% to 40% of the mean-squared error in the street-view synthesis and a 45% PSNR gain for the moving vehicles rendering. | ['Li Zhang', 'Feihu Zhang', 'Wenye Li', 'Junge Zhang', 'Ziyang Xie'] | 2023-03-01 | null | null | null | null | ['self-driving-cars'] | ['computer-vision'] | [ 1.75394729e-01 -1.92791849e-01 2.68848509e-01 -5.34102261e-01
-8.50696266e-01 -6.03447914e-01 5.59468508e-01 -9.64106619e-01
5.82206808e-02 6.27436221e-01 1.79932237e-01 -9.16067362e-02
2.91593254e-01 -8.95395398e-01 -1.25387144e+00 -7.11660862e-01
5.57780623e-01 6.76530749e-02 5.36571503e-01 -2.96118468e-01
1.07256569e-01 5.75270772e-01 -1.84806991e+00 4.32261288e-01
9.29713607e-01 8.93566489e-01 6.61953628e-01 6.78279281e-01
-5.60226142e-02 7.95844436e-01 -2.32532293e-01 -6.83027744e-01
7.17198789e-01 -7.58698583e-02 1.02726229e-01 3.91301215e-01
1.34188020e+00 -8.20521414e-01 -9.04616237e-01 1.06884205e+00
2.24052042e-01 1.78609580e-01 4.53135908e-01 -1.29260111e+00
-7.74886906e-01 -2.48502806e-01 -9.44286287e-01 1.33975685e-01
8.78859684e-02 3.63179207e-01 6.18963480e-01 -1.41622472e+00
7.30580688e-01 1.58273697e+00 6.63859189e-01 4.16726440e-01
-1.14441419e+00 -9.12022352e-01 3.91922206e-01 2.53453642e-01
-1.39011395e+00 -7.75405765e-01 9.21858132e-01 -2.69934654e-01
6.30042076e-01 3.17760080e-01 4.96039152e-01 1.31249547e+00
3.24919641e-01 5.87008774e-01 9.98903751e-01 3.62587571e-01
9.33662876e-02 1.80529520e-01 -3.00530463e-01 5.12490809e-01
5.38665116e-01 4.19801801e-01 -4.75812346e-01 1.20376706e-01
9.11735117e-01 3.79735440e-01 -3.89393389e-01 -5.54361999e-01
-1.26103508e+00 7.24802673e-01 4.73912418e-01 -3.59948695e-01
-2.18599379e-01 3.45032871e-01 -1.09829530e-01 -8.60138610e-02
5.36432624e-01 1.24086346e-02 -3.90661806e-01 2.80650795e-01
-8.31124365e-01 3.68665546e-01 2.33873829e-01 1.34294903e+00
9.14943635e-01 5.89455724e-01 1.72452062e-01 6.63645685e-01
3.48018229e-01 1.14694428e+00 -1.55558780e-01 -1.54291964e+00
9.34277296e-01 1.75791517e-01 3.31818610e-01 -1.31224513e+00
-1.38996527e-01 -4.67367887e-01 -1.08641565e+00 1.90435663e-01
1.85725227e-01 -2.96026152e-02 -8.64965618e-01 1.65291452e+00
5.07357538e-01 6.96374834e-01 1.43516347e-01 1.06213403e+00
9.86536264e-01 9.59369302e-01 -5.78324020e-01 -2.36854047e-01
1.08726120e+00 -1.03112817e+00 -8.32606077e-01 -6.35837317e-01
1.61141038e-01 -8.49364936e-01 9.79077935e-01 4.44378465e-01
-1.05503678e+00 -8.92897487e-01 -9.97844875e-01 -2.90739477e-01
-1.03202231e-01 -3.44061367e-02 4.72099036e-01 5.33525050e-01
-7.70168364e-01 1.61107287e-01 -6.01980269e-01 -2.28421949e-03
4.83592063e-01 -1.90780506e-01 -4.24290717e-01 -6.63269699e-01
-8.76843393e-01 7.96522021e-01 -3.90944071e-02 3.38244647e-01
-9.60652232e-01 -1.03604126e+00 -9.25687015e-01 -1.54339999e-01
6.02153182e-01 -9.16645110e-01 8.07187259e-01 -6.85076356e-01
-1.32146764e+00 5.65223277e-01 -4.40914184e-01 -1.72350243e-01
6.71014249e-01 -2.80308038e-01 -6.38744652e-01 1.52258769e-01
2.71781802e-01 8.73709977e-01 9.30616021e-01 -1.88494444e+00
-7.46089339e-01 -3.84526432e-01 3.23377512e-02 3.14920962e-01
2.89190561e-01 -5.59033811e-01 -5.33774972e-01 -7.26364434e-01
3.22880477e-01 -9.80144799e-01 -2.36471415e-01 2.70933121e-01
-1.38939381e-01 4.80468094e-01 1.03070712e+00 -6.56664371e-01
5.50248563e-01 -2.33586407e+00 -2.53248381e-05 -2.42210194e-01
3.11705321e-01 -7.25516453e-02 -2.94281155e-01 1.18929647e-01
-3.49779949e-02 -2.20065445e-01 -1.92072943e-01 -3.51520061e-01
-1.80931389e-01 4.37561095e-01 -7.74593472e-01 5.74586868e-01
1.33716151e-01 9.48891878e-01 -7.38731682e-01 -1.66190177e-01
6.41892850e-01 6.41713023e-01 -7.40537226e-01 1.49685681e-01
-5.15205003e-02 6.08716011e-01 -2.87112534e-01 4.52763617e-01
1.43030536e+00 -4.77970243e-02 -2.81118631e-01 -5.57619333e-01
-1.05587140e-01 -1.31380782e-01 -1.47165251e+00 1.78594291e+00
-5.38415074e-01 7.27174759e-01 8.82899612e-02 -4.38696623e-01
8.34183574e-01 -2.06353575e-01 3.55244368e-01 -8.07747602e-01
-1.37757093e-01 1.90374002e-01 -4.26857710e-01 -5.89718759e-01
7.53370285e-01 -1.46383554e-01 1.80639103e-01 -1.76494122e-01
-2.17929199e-01 -4.89663243e-01 -1.49335638e-01 2.50706941e-01
7.24222660e-01 2.08858326e-01 -1.21705569e-01 8.74957740e-02
4.20394838e-01 -2.52867103e-01 9.12797332e-01 4.13071781e-01
1.35404214e-01 1.16736281e+00 -5.24081029e-02 -4.92579281e-01
-1.20622611e+00 -1.54090405e+00 -1.52785957e-01 5.79754472e-01
5.08276939e-01 -2.33933926e-01 -5.63114464e-01 -4.06731844e-01
-5.96378073e-02 1.03793645e+00 -2.16088325e-01 -1.21491179e-01
-7.03331113e-01 -5.15129387e-01 1.81412488e-01 4.01995987e-01
8.05099189e-01 -4.46401417e-01 -5.93269646e-01 -1.25008970e-02
-6.81349695e-01 -1.69729233e+00 -6.65073395e-01 -2.66487092e-01
-8.16274047e-01 -9.55981612e-01 -5.66296637e-01 -3.86931330e-01
5.90049088e-01 1.22288024e+00 1.13103652e+00 -4.46898401e-01
2.67210025e-02 1.76161081e-01 -7.78343007e-02 -1.59005746e-01
-9.22460854e-02 -4.83053356e-01 6.08659312e-02 3.66171509e-01
8.99623111e-02 -7.66805112e-01 -6.99125051e-01 5.28906465e-01
-9.26389039e-01 5.36218345e-01 6.36566401e-01 6.64184034e-01
5.81100762e-01 1.09360658e-01 2.13735238e-01 -7.66618073e-01
-2.76350647e-01 -4.54033911e-01 -8.04014564e-01 -5.80195710e-02
-3.11466068e-01 -2.20041513e-01 7.19664693e-01 -2.93069541e-01
-1.39294958e+00 1.57871410e-01 -4.27387469e-03 -1.02990997e+00
-1.99968547e-01 -2.89962113e-01 -5.00651836e-01 -2.74852272e-02
5.24734199e-01 3.40459049e-01 -4.86517102e-01 -3.36254597e-01
6.74190879e-01 3.82061183e-01 7.80588567e-01 -2.78510213e-01
1.31305015e+00 9.95610952e-01 8.28278512e-02 -8.60958397e-01
-8.00709069e-01 -3.57694358e-01 -4.08420384e-01 -2.71457076e-01
1.00132024e+00 -1.51009750e+00 -3.51766825e-01 3.23196739e-01
-1.44561434e+00 -4.92699929e-02 -2.36389220e-01 4.91677284e-01
-6.47358358e-01 3.95193547e-01 -3.03481668e-01 -7.02929199e-01
2.30386168e-01 -1.34061575e+00 1.47155595e+00 9.84322578e-02
4.63622361e-01 -4.42745388e-01 -2.37584904e-01 7.33945668e-01
1.54949352e-01 2.59847105e-01 5.43251455e-01 3.60922277e-01
-1.40208638e+00 4.35940057e-01 -4.90783691e-01 3.31966281e-01
-6.19667545e-02 2.17627734e-02 -1.27563894e+00 -2.69166291e-01
2.74161696e-01 9.13865566e-02 9.92295384e-01 4.54715788e-01
1.07005799e+00 -1.57965377e-01 -1.39164045e-01 1.24624372e+00
1.59861588e+00 2.37165689e-01 7.85795212e-01 4.23881374e-02
1.17150152e+00 8.42738330e-01 6.89000130e-01 3.25570434e-01
6.52639389e-01 7.99858093e-01 7.51679420e-01 -1.19460277e-01
-3.95473242e-01 -5.66735566e-01 5.75145781e-01 7.42077589e-01
1.21564297e-02 -3.60776842e-01 -5.40290296e-01 4.36492443e-01
-1.84509468e+00 -1.06023347e+00 -5.90001047e-01 2.03273988e+00
2.11368144e-01 -8.71308297e-02 -3.98432493e-01 -2.76370049e-01
6.04363799e-01 4.96803284e-01 -8.57648015e-01 7.96368420e-02
-4.63582009e-01 -2.60916173e-01 8.55488777e-01 6.31145537e-01
-7.09510028e-01 1.01804829e+00 5.47146702e+00 7.33896673e-01
-8.99235606e-01 2.45225877e-01 6.60913289e-01 -2.31907502e-01
-6.86132908e-01 -8.07091147e-02 -9.56524789e-01 4.02189404e-01
6.78687274e-01 2.30799451e-01 7.67543316e-01 7.07753003e-01
4.16710496e-01 -4.22761999e-02 -9.78064179e-01 1.50575459e+00
3.94185036e-01 -1.49906433e+00 1.82850212e-01 2.43697479e-01
1.21295178e+00 2.88218528e-01 3.26015055e-01 6.84217829e-03
3.04447562e-01 -7.49606550e-01 1.02230847e+00 6.84013069e-01
9.12509441e-01 -7.12660372e-01 3.75760823e-01 4.38404351e-01
-1.25946391e+00 -1.21785603e-01 -6.83056235e-01 2.37153769e-01
4.77239758e-01 6.37466192e-01 -1.79137483e-01 7.95444429e-01
9.03201699e-01 9.94595587e-01 -6.57830596e-01 5.99917531e-01
1.38719147e-02 2.27549113e-02 -2.96966612e-01 7.00893044e-01
2.28426084e-01 -5.42039990e-01 6.51149154e-01 7.37356722e-01
6.23067915e-01 2.05158770e-01 3.48092653e-02 1.14675856e+00
-9.51886992e-04 -3.83144349e-01 -1.04189837e+00 5.54868937e-01
2.58213460e-01 1.26902223e+00 -3.31301033e-01 -3.12509120e-01
-6.58396304e-01 9.39322233e-01 -5.83134703e-02 8.54231119e-01
-1.24251616e+00 1.04946539e-01 9.77986395e-01 4.99754518e-01
5.92586875e-01 -3.08908343e-01 -1.90250158e-01 -1.68559647e+00
2.86022484e-01 -8.47909510e-01 -2.59639353e-01 -1.40857613e+00
-1.28809011e+00 4.53715593e-01 1.26494750e-01 -1.54909503e+00
-9.44590718e-02 -4.84517813e-01 -3.21308196e-01 5.80132186e-01
-1.79363191e+00 -1.19851315e+00 -6.49580956e-01 7.49614894e-01
1.01093936e+00 -7.82935321e-02 1.65514514e-01 5.40201545e-01
-3.06616992e-01 1.75295338e-01 4.14065629e-01 -3.01625758e-01
7.71257639e-01 -7.92813182e-01 6.95164621e-01 1.19909966e+00
1.85415611e-01 3.17120522e-01 6.93241835e-01 -5.03574193e-01
-1.72732699e+00 -1.58219993e+00 4.26652491e-01 -6.00039065e-01
2.69423515e-01 -6.91265345e-01 -7.57716835e-01 8.01136374e-01
-2.26213150e-02 4.05473292e-01 8.64122957e-02 -4.54991072e-01
-5.53166449e-01 -5.38310528e-01 -1.04266727e+00 8.26808095e-01
1.52330339e+00 -4.56286043e-01 -4.07315463e-01 2.02508584e-01
9.46299911e-01 -5.21552861e-01 -2.97816455e-01 4.74656016e-01
4.86533254e-01 -1.45264578e+00 1.45465839e+00 -1.78932875e-01
5.36743999e-01 -7.03453064e-01 -8.82982850e-01 -1.23106039e+00
-3.48253012e-01 -4.27751780e-01 -1.39692381e-01 1.13453686e+00
-8.90387893e-02 -7.04050183e-01 5.98392189e-01 4.17952985e-01
-4.17356998e-01 -3.76201898e-01 -8.80759299e-01 -6.14003003e-01
-8.05619061e-02 -6.62631214e-01 6.14991605e-01 9.10916090e-01
-1.14346766e+00 5.02185106e-01 -6.55111074e-01 5.36192775e-01
9.76486206e-01 1.97599515e-01 1.26335335e+00 -1.05060244e+00
-2.90198207e-01 4.82098535e-02 -2.39506900e-01 -1.52248526e+00
1.23085655e-01 -6.17575586e-01 1.62857100e-01 -1.29716003e+00
7.29392320e-02 -1.46406904e-01 1.72128245e-01 -2.63788909e-01
-1.08518437e-01 2.57710367e-01 4.95368689e-01 1.13514297e-01
-4.52522933e-01 8.60289931e-01 1.52010548e+00 -8.88635367e-02
8.75305533e-02 -2.10846335e-01 -7.14872122e-01 1.02807069e+00
3.28557283e-01 -4.94592816e-01 -7.68762946e-01 -8.57522011e-01
4.01600361e-01 2.02079490e-01 6.32374167e-01 -1.06006026e+00
4.60414216e-02 -4.00102705e-01 6.31434798e-01 -1.14748859e+00
7.26553679e-01 -1.14296472e+00 6.71045840e-01 1.72581933e-02
9.13685262e-02 1.11895762e-01 -1.49137806e-02 1.06212783e+00
-3.50179598e-02 2.99394637e-01 9.38840628e-01 -2.48014927e-01
-7.27487326e-01 5.42335153e-01 -1.20300744e-02 1.89209104e-01
1.04200804e+00 -4.82499570e-01 -5.37084043e-01 -4.65667605e-01
-1.92558527e-01 1.70262560e-01 6.20684147e-01 7.00900614e-01
9.41347361e-01 -1.55902362e+00 -7.10059285e-01 5.81778646e-01
2.27432415e-01 5.63625455e-01 7.59851396e-01 6.34612083e-01
-5.90184569e-01 3.02920371e-01 -1.38706505e-01 -8.57895911e-01
-1.06672466e+00 6.98300004e-01 2.18346968e-01 1.43844470e-01
-9.92975235e-01 6.75108969e-01 1.24883235e+00 -5.93307734e-01
-1.31616116e-01 -6.12281680e-01 2.51342654e-01 -2.74395347e-01
5.01246274e-01 4.06641960e-01 -1.15977839e-01 -1.03366911e+00
-1.63227797e-01 9.20849681e-01 1.88651294e-01 -8.75936598e-02
1.33782065e+00 -5.58285117e-01 2.28057861e-01 5.81456959e-01
1.27850068e+00 2.59316087e-01 -1.91909742e+00 -2.45848805e-01
-7.69472122e-01 -1.09920418e+00 1.90588966e-01 -2.68191576e-01
-1.36856246e+00 1.11645091e+00 6.31799638e-01 -3.97177756e-01
1.00889218e+00 -2.81077743e-01 1.18893826e+00 4.72870529e-01
4.84457254e-01 -8.18808198e-01 -9.52161662e-03 3.53927046e-01
9.47694778e-01 -1.36168575e+00 -1.07791973e-02 -6.76371753e-01
-8.36487174e-01 9.72277224e-01 7.34785080e-01 -2.64424592e-01
3.94818962e-01 1.46999657e-01 2.03169137e-03 -1.15186125e-01
-8.46358180e-01 -5.66711612e-02 2.29256481e-01 6.98439360e-01
-3.11409146e-01 -1.10705450e-01 5.08378625e-01 2.57049590e-01
-3.04640502e-01 -2.97161758e-01 6.92224324e-01 3.08574408e-01
-3.23629141e-01 -4.37753260e-01 -5.35253525e-01 2.26175919e-01
1.13732712e-02 -1.93515703e-01 -4.25170101e-02 6.53320432e-01
5.72037458e-01 1.03027904e+00 1.36230662e-01 -4.55204815e-01
6.03575170e-01 -2.77580708e-01 4.85416293e-01 -3.54751945e-01
9.25587304e-03 2.90765405e-01 -6.41436726e-02 -9.47764635e-01
-3.96216542e-01 -6.06208742e-01 -9.21170294e-01 -6.22346640e-01
-2.10403591e-01 -5.38233578e-01 4.56554264e-01 6.59425855e-01
4.37304080e-01 5.53865016e-01 9.59025860e-01 -1.14015412e+00
-2.25045890e-01 -4.95840520e-01 -6.26488030e-01 4.08145100e-01
6.11812055e-01 -8.56102407e-01 -5.39825320e-01 1.36265740e-01] | [8.894304275512695, -2.511805772781372] |
278d3086-8b8b-4c1f-bea7-ad2d4e0016f8 | safe-real-world-reinforcement-learning-for | 2209.11789 | null | https://arxiv.org/abs/2209.11789v2 | https://arxiv.org/pdf/2209.11789v2.pdf | SAFER: Safe Collision Avoidance using Focused and Efficient Trajectory Search with Reinforcement Learning | Collision avoidance is key for mobile robots and agents to operate safely in the real world. In this work we present SAFER, an efficient and effective collision avoidance system that is able to improve safety by correcting the control commands sent by an operator. It combines real-world reinforcement learning (RL), search-based online trajectory planning, and automatic emergency intervention, e.g. automatic emergency braking (AEB). The goal of the RL is to learn an effective corrective control action that is used in a focused search for collision-free trajectories, and to reduce the frequency of triggering automatic emergency braking. This novel setup enables the RL policy to learn safely and directly on mobile robots in a real-world indoor environment, minimizing actual crashes even during training. Our real-world experiments show that, when compared with several baselines, our approach enjoys a higher average speed, lower crash rate, less emergency intervention, smaller computation overhead, and smoother overall control. | ['Hugues Thomas', 'Jian Zhang', 'Ali Farhadi', 'Hubert Tsai', 'Mario Srouji'] | 2022-09-23 | null | null | null | null | ['trajectory-planning'] | ['robots'] | [-1.60942838e-01 3.48284930e-01 -1.76188305e-01 -1.37932468e-02
-6.71809316e-01 -3.31377625e-01 5.04720330e-01 3.56619805e-01
-1.03449166e+00 9.08505797e-01 -2.26828545e-01 -7.42190480e-01
-3.93823832e-01 -9.28194106e-01 -1.02146256e+00 -5.89713097e-01
-4.77965444e-01 8.65671158e-01 8.22575212e-01 -8.89522731e-01
2.82679349e-01 8.56690228e-01 -1.72908103e+00 -4.90105867e-01
8.54085803e-01 5.77154338e-01 5.92098653e-01 6.39690578e-01
5.38307846e-01 8.62082243e-01 -2.87040740e-01 1.11064851e-01
2.11314246e-01 2.93842942e-01 -8.82364333e-01 -3.14725518e-01
-3.39485019e-01 -4.31690335e-01 -5.01794100e-01 4.03249353e-01
4.46621150e-01 6.27196133e-01 5.41807175e-01 -1.79709661e+00
4.42417502e-01 3.01511377e-01 -4.22006622e-02 1.49258912e-01
4.80122745e-01 6.50040388e-01 2.89717019e-01 -4.23207164e-01
5.83046913e-01 1.24837601e+00 6.14098608e-01 5.55010974e-01
-1.01092851e+00 -4.86633390e-01 1.77564889e-01 6.90102339e-01
-1.42523444e+00 -3.25684249e-01 9.95899886e-02 1.95065178e-02
1.55999327e+00 -1.61705688e-02 5.38333058e-01 8.86190355e-01
6.46798074e-01 5.89875579e-01 3.06334376e-01 -1.65387854e-01
5.35360336e-01 -2.09222421e-01 -4.33612943e-01 8.22367787e-01
1.75873354e-01 6.22479260e-01 -7.98972622e-02 -5.34520485e-02
1.59430325e-01 1.72262058e-01 -1.50339872e-01 -6.35249853e-01
-1.27313983e+00 7.26411700e-01 5.13205230e-01 -3.84346485e-01
-7.58485794e-01 3.66457313e-01 5.51199257e-01 3.53309959e-01
-2.25497290e-01 3.72561425e-01 -5.44840157e-01 -6.11421406e-01
-5.53212501e-02 6.25624180e-01 7.04670131e-01 1.11719632e+00
9.89020348e-01 5.26385941e-02 8.32232237e-02 5.14770865e-01
6.94424883e-02 9.72206533e-01 8.72411653e-02 -1.44904208e+00
7.67071426e-01 3.67674321e-01 5.88619649e-01 -9.30964351e-01
-9.74561870e-01 4.01168346e-01 -4.77769792e-01 9.50404525e-01
7.80277625e-02 -4.40612257e-01 -4.41173434e-01 1.52952838e+00
5.76350629e-01 3.56696069e-01 3.53005081e-01 7.10416198e-01
-2.87214994e-01 6.31223321e-01 1.19247369e-01 -2.09892586e-01
8.51357281e-01 -8.52579832e-01 -5.84677398e-01 -4.54844594e-01
9.60019588e-01 -3.07303667e-01 9.46884811e-01 6.07983828e-01
-7.73916483e-01 -2.90924966e-01 -1.05120862e+00 2.75920779e-01
-4.15665865e-01 -2.01839209e-01 2.37995967e-01 2.34841987e-01
-8.50842714e-01 8.39197874e-01 -1.10521066e+00 -3.44987690e-01
4.07037772e-02 7.17644989e-01 -5.26903689e-01 -3.40092252e-03
-9.69218314e-01 1.47320020e+00 5.01417100e-01 -4.16436732e-01
-1.30767429e+00 -3.57354403e-01 -1.07883978e+00 -2.24466994e-01
1.02857077e+00 -2.69823313e-01 1.69092107e+00 1.03295194e-02
-1.76090169e+00 1.38105258e-01 -6.99019572e-03 -9.05170619e-01
4.69590813e-01 -6.68572962e-01 -5.41122351e-03 7.70513434e-03
1.83564335e-01 7.50496328e-01 5.40533960e-01 -1.29561424e+00
-1.43255222e+00 2.81209452e-03 -2.21506786e-03 5.00725031e-01
2.17734680e-01 -4.21506912e-01 -2.48100907e-01 -6.18604384e-02
-4.77176398e-01 -1.55551088e+00 -7.43828475e-01 -4.61078197e-01
-5.02496026e-02 -5.46028554e-01 9.60442007e-01 -2.60671586e-01
9.77624774e-01 -2.09363341e+00 1.58109576e-01 3.68879288e-01
-2.27516800e-01 2.69882232e-01 -1.25146568e-01 7.92798162e-01
2.74709433e-01 -3.56940746e-01 -8.37545097e-02 -3.40736896e-01
-1.17771387e-01 6.70320332e-01 -6.33472323e-01 4.98358309e-01
6.77345842e-02 6.36010110e-01 -1.34314179e+00 -1.99814871e-01
5.63555479e-01 1.05655216e-01 -5.82628667e-01 3.92275035e-01
-1.22355364e-01 4.39091176e-01 -2.16657758e-01 2.34078944e-01
2.77142286e-01 6.23740077e-01 1.35888353e-01 5.77614903e-01
-6.55703187e-01 4.13651973e-01 -1.22150135e+00 1.23127270e+00
-6.54537976e-01 6.79475665e-01 1.39432609e-01 -7.92124927e-01
6.46290064e-01 9.91979018e-02 5.72142243e-01 -1.00523257e+00
1.24979824e-01 2.52742320e-01 -4.41671461e-01 -6.70491815e-01
6.57396615e-01 6.46270514e-01 -3.80792588e-01 2.88960040e-01
-3.66832286e-01 -1.95559517e-01 2.24747211e-01 1.10250100e-01
1.44331551e+00 4.66173515e-02 2.16753498e-01 -4.68786471e-02
4.04400110e-01 5.59462249e-01 4.85609323e-01 9.29477751e-01
-3.33121747e-01 -2.55935222e-01 -1.00827385e-02 -4.53896761e-01
-9.56722915e-01 -1.00155032e+00 7.01069534e-01 1.20524561e+00
8.76027048e-01 -4.00197595e-01 -7.13659883e-01 -6.08524144e-01
5.54276444e-02 1.10427630e+00 -2.07512289e-01 -7.08125532e-01
-1.25479650e+00 -8.10140371e-02 5.31454921e-01 3.78878236e-01
3.02695096e-01 -1.28710997e+00 -1.35404372e+00 4.27608877e-01
-3.49147767e-01 -9.80905533e-01 -4.37755376e-01 3.75041217e-01
-4.60714072e-01 -1.41316140e+00 6.10960051e-02 -8.17748368e-01
4.32395220e-01 6.88436449e-01 7.35321760e-01 2.99943238e-01
-4.80737984e-02 5.82077563e-01 -4.65276390e-01 -6.15894437e-01
-5.57611346e-01 9.83085632e-02 5.60702205e-01 -5.32330751e-01
-5.81267364e-02 -2.29102537e-01 -5.79656482e-01 5.72063863e-01
-3.61379713e-01 -2.35450208e-01 5.61461627e-01 6.42517149e-01
4.56624091e-01 3.20286691e-01 4.12106693e-01 -2.32572079e-01
8.13931108e-01 -5.76460898e-01 -8.70915055e-01 1.99667253e-02
-1.01679838e+00 2.11781204e-01 7.25973785e-01 -3.82433623e-01
-8.83687437e-01 2.98328608e-01 -2.75571227e-01 -1.52730746e-02
-4.91022080e-01 9.67646614e-02 1.75456330e-01 -7.61338547e-02
7.70548999e-01 2.26844430e-01 1.96783602e-01 1.96728501e-02
3.12289983e-01 5.66209197e-01 7.00830340e-01 -2.64725596e-01
8.34059417e-01 3.92162055e-01 1.27044588e-01 -6.97838724e-01
3.73441726e-02 -6.05585933e-01 -5.14211833e-01 -3.68225574e-01
6.31368041e-01 -7.78439522e-01 -1.55069065e+00 4.01742220e-01
-1.06170726e+00 -1.14587188e+00 -1.32820755e-01 3.14906180e-01
-1.14750934e+00 2.66581774e-01 -3.27003330e-01 -1.13454092e+00
-1.38945311e-01 -1.23194504e+00 1.07906270e+00 2.77250886e-01
-1.31662980e-01 -5.30966818e-01 2.14855075e-01 -1.62958130e-01
2.92430639e-01 7.03344345e-02 4.58448559e-01 -2.96963006e-01
-7.31885314e-01 -6.75378963e-02 1.93842933e-01 -3.43729734e-01
6.29277974e-02 -2.34460503e-01 -2.19115019e-01 -4.81172085e-01
-5.59128463e-01 -2.15466082e-01 3.99993569e-01 1.16728209e-01
6.21446192e-01 -5.90621948e-01 -8.96010578e-01 1.55297950e-01
1.11007059e+00 7.44718254e-01 7.73213983e-01 8.44763517e-01
2.35775039e-01 6.41679645e-01 1.65343928e+00 4.48669165e-01
9.50777829e-01 1.08747375e+00 1.01534641e+00 8.83630067e-02
4.25460666e-01 -2.59997606e-01 7.51389563e-01 9.46127474e-02
-7.71640986e-03 -2.33685210e-01 -1.09565210e+00 5.53190887e-01
-2.54698086e+00 -1.05504942e+00 -5.07499576e-02 2.50227237e+00
5.74650764e-01 4.66931522e-01 3.68259788e-01 2.55103648e-01
3.95367414e-01 -6.70026243e-01 -4.72332895e-01 -5.56651413e-01
5.46117187e-01 -2.53269911e-01 1.08639121e+00 9.44216847e-01
-1.12742281e+00 1.12390530e+00 6.16555357e+00 6.78883195e-01
-1.02779603e+00 1.93188060e-03 -8.71106982e-02 1.19304650e-01
4.47759479e-01 -1.91797987e-01 -6.87773705e-01 3.34056348e-01
1.27100444e+00 -4.16708179e-02 9.38405275e-01 1.27567613e+00
8.32461715e-01 -4.80157018e-01 -7.96160817e-01 6.72900379e-01
-5.41327775e-01 -1.16480386e+00 -4.33379382e-01 -4.59301062e-02
9.77581143e-02 3.55854481e-01 -2.32515842e-01 6.82957768e-01
6.10374391e-01 -7.76659608e-01 9.76732552e-01 5.48712730e-01
4.64473069e-01 -1.49321508e+00 7.55771935e-01 9.89727199e-01
-1.23234725e+00 -7.51691997e-01 4.27716672e-02 -3.35740775e-01
6.38889790e-01 -1.38579175e-01 -1.47192597e+00 3.39869708e-01
8.75173986e-01 2.45536178e-01 -1.57259643e-01 1.15779233e+00
-3.69954646e-01 1.67866930e-01 -4.09698337e-01 -4.05296206e-01
3.31944942e-01 -9.11280662e-02 7.43673086e-01 1.05867827e+00
1.41445264e-01 2.88406610e-01 6.14850461e-01 9.85759497e-02
7.18134522e-01 -4.04924095e-01 -1.17261028e+00 6.59813166e-01
8.62802863e-01 8.34535599e-01 -5.70180774e-01 -1.31596237e-01
9.35419425e-02 1.05008662e+00 4.17831749e-01 1.86617345e-01
-1.15421796e+00 -7.50877559e-01 1.02057767e+00 9.25896317e-02
1.47277236e-01 -7.93607712e-01 1.08505979e-01 -1.03828177e-01
-3.45711619e-01 -6.32012129e-01 1.45273969e-01 -6.06166303e-01
-4.82720315e-01 4.98942375e-01 2.54387826e-01 -1.32314265e+00
-6.20364487e-01 -5.42455375e-01 -5.33991992e-01 1.88662738e-01
-1.45616794e+00 -4.07641053e-01 -2.28581771e-01 7.38997459e-01
7.76740015e-01 -1.43583924e-01 7.31635869e-01 2.69034803e-01
-4.38430876e-01 3.78178000e-01 -1.11799911e-01 -4.41060990e-01
6.22752905e-01 -1.20968246e+00 4.64651495e-01 6.16221368e-01
-5.21741331e-01 2.85176218e-01 1.06882346e+00 -8.46825540e-01
-1.52591550e+00 -1.25893676e+00 6.54798806e-01 -3.60636532e-01
2.13521957e-01 -9.19470638e-02 -4.96082187e-01 5.95425308e-01
4.95834555e-03 -6.29099011e-01 -5.89445978e-02 -3.48282844e-01
3.39924932e-01 -2.25370333e-01 -1.08431768e+00 1.12495351e+00
1.00178635e+00 1.86013412e-02 -4.57810462e-01 2.53876001e-01
9.98777509e-01 -5.60017407e-01 -1.54640689e-01 4.74792272e-01
1.86810955e-01 -7.64775693e-01 1.07318068e+00 -3.31449449e-01
-3.80501986e-01 -4.92762089e-01 2.02042073e-01 -1.60909593e+00
-2.22873449e-01 -9.78666306e-01 -1.26516491e-01 4.82895553e-01
3.75531375e-01 -7.18414962e-01 4.14282233e-01 4.57990497e-01
-6.02641106e-01 -7.61424422e-01 -1.23867714e+00 -9.16452765e-01
-5.10500014e-01 -7.01463699e-01 3.73808235e-01 2.13378981e-01
2.63548225e-01 1.34430051e-01 -5.27308643e-01 6.52590692e-01
3.98403138e-01 -5.27448356e-01 1.17879224e+00 -8.07498336e-01
3.31247181e-01 -2.28188783e-01 -1.46222651e-01 -8.63378763e-01
2.63184220e-01 -3.61861736e-01 9.12099600e-01 -1.40730298e+00
-4.79823351e-01 -7.57903397e-01 2.26062715e-01 7.11064816e-01
1.12750396e-01 -3.32644820e-01 8.17661732e-02 -1.85494367e-02
-9.97177780e-01 7.49638975e-01 7.32503116e-01 -5.01388833e-02
-6.37117505e-01 3.28780234e-01 1.03990614e-01 7.96458781e-01
1.20582914e+00 -6.26877129e-01 -5.32567203e-01 -1.94995567e-01
1.90345407e-01 2.27011487e-01 2.82739341e-01 -1.27059448e+00
6.89545572e-01 -7.20430195e-01 -5.71876287e-01 -5.95464647e-01
4.21248674e-01 -1.09703350e+00 -9.59215686e-02 1.12079346e+00
-7.76294200e-03 6.19056642e-01 3.86690438e-01 8.71368706e-01
2.73885578e-01 -5.04804850e-02 5.80300331e-01 2.46912971e-01
-1.13298881e+00 1.26478419e-01 -1.28379476e+00 -2.71876514e-01
1.57293987e+00 -5.89729697e-02 -2.73088694e-01 -5.42324483e-01
-3.12536061e-01 8.48775029e-01 3.56624991e-01 6.07775986e-01
9.09870982e-01 -1.04538321e+00 -2.00888842e-01 3.69997509e-02
-9.67708379e-02 6.60371268e-03 -1.02153406e-01 8.67476523e-01
-7.33484328e-01 4.41301882e-01 -2.98249930e-01 -3.92650098e-01
-1.13632441e+00 7.67398477e-01 4.03578609e-01 -1.55513510e-01
-8.92543375e-01 4.39589739e-01 -4.91444290e-01 -6.53386474e-01
6.95120096e-01 -2.72831619e-01 -3.66432041e-01 -4.12877351e-01
4.77172613e-01 1.01561153e+00 1.10283546e-01 -5.08511901e-01
-5.63584566e-01 2.50469267e-01 3.50366861e-01 -2.84127295e-01
1.14784300e+00 -5.06729305e-01 3.50731879e-01 2.16928661e-01
4.28209394e-01 -9.38412175e-02 -1.65591371e+00 3.12385768e-01
3.79030615e-01 -2.28470266e-01 -1.78199470e-01 -6.17216229e-01
-4.47125971e-01 1.99512124e-01 7.75331140e-01 2.53819793e-01
7.48971701e-01 -2.47511566e-01 1.25535321e+00 1.06595373e+00
1.16214180e+00 -1.50569785e+00 2.24697471e-01 1.10841179e+00
8.26186419e-01 -1.17343926e+00 -4.95016903e-01 -3.99028122e-01
-9.67061162e-01 8.08600664e-01 9.35030758e-01 -1.84877098e-01
3.19671869e-01 3.26142639e-01 6.40901551e-02 2.94959247e-02
-9.68769014e-01 -4.39789355e-01 -4.02471572e-01 1.09574950e+00
-8.00860524e-01 1.00417659e-01 -5.19609712e-02 4.56889421e-02
-1.53873622e-01 -2.41917372e-01 5.81616402e-01 1.26074898e+00
-1.05531001e+00 -9.27888393e-01 -3.84398788e-01 -2.13604365e-02
2.53723979e-01 6.00942612e-01 2.03437656e-02 8.27912688e-01
2.00156868e-01 1.45164382e+00 1.33489370e-01 -6.43614650e-01
9.27064419e-01 -3.41320038e-01 8.09917524e-02 -3.52995992e-01
-2.92239219e-01 -6.01783812e-01 2.34204754e-01 -1.20748174e+00
2.10467756e-01 -6.49934292e-01 -2.06971025e+00 -4.46921051e-01
-9.70814228e-02 1.26651719e-01 5.84858835e-01 1.15591621e+00
6.69143498e-01 3.75120014e-01 8.01380098e-01 -1.41759229e+00
-4.21100497e-01 -5.32016635e-01 8.40712711e-02 1.73560530e-01
7.70807087e-01 -1.19368374e+00 -9.89607945e-02 -3.55683565e-01] | [5.030693054199219, 1.336256742477417] |
1eb974ae-8a3a-4a09-a5a8-aa44e4a18ff6 | s-t-a-r-track-latent-motion-models-for-end-to | 2306.17602 | null | https://arxiv.org/abs/2306.17602v1 | https://arxiv.org/pdf/2306.17602v1.pdf | S.T.A.R.-Track: Latent Motion Models for End-to-End 3D Object Tracking with Adaptive Spatio-Temporal Appearance Representations | Following the tracking-by-attention paradigm, this paper introduces an object-centric, transformer-based framework for tracking in 3D. Traditional model-based tracking approaches incorporate the geometric effect of object- and ego motion between frames with a geometric motion model. Inspired by this, we propose S.T.A.R.-Track, which uses a novel latent motion model (LMM) to additionally adjust object queries to account for changes in viewing direction and lighting conditions directly in the latent space, while still modeling the geometric motion explicitly. Combined with a novel learnable track embedding that aids in modeling the existence probability of tracks, this results in a generic tracking framework that can be integrated with any query-based detector. Extensive experiments on the nuScenes benchmark demonstrate the benefits of our approach, showing state-of-the-art performance for DETR3D-based trackers while drastically reducing the number of identity switches of tracks at the same time. | ['Hendrik P. A. Lensch', 'Markus Enzweiler', 'Richard Schulz', 'Lukas Schneider', 'Niklas Hanselmann', 'Simon Doll'] | 2023-06-30 | null | null | null | null | ['object-tracking', '3d-object-tracking'] | ['computer-vision', 'computer-vision'] | [-4.05549049e-01 -4.22513783e-01 -2.68373907e-01 -3.47695351e-02
-5.76767147e-01 -8.57947111e-01 8.10139000e-01 -2.04186976e-01
-2.93654382e-01 1.52634621e-01 2.39427447e-01 5.74304722e-02
1.47236586e-01 -5.77031136e-01 -9.00508523e-01 -6.91691458e-01
1.37786224e-01 5.30780435e-01 7.50060320e-01 1.10239439e-01
-1.08451769e-02 7.89136529e-01 -1.44825447e+00 -1.96289733e-01
3.34223717e-01 7.29710698e-01 1.94195211e-01 5.42075157e-01
-2.88989916e-02 5.45248210e-01 -3.38352561e-01 -4.35837299e-01
4.94261622e-01 2.53891259e-01 -6.17064461e-02 4.02354926e-01
1.17771745e+00 -3.69497329e-01 -7.45701015e-01 9.24789667e-01
4.42511499e-01 -6.86971620e-02 4.10107106e-01 -1.21644616e+00
-6.95395887e-01 1.30937710e-01 -6.12613022e-01 2.56778449e-01
2.21589386e-01 3.59634310e-01 7.47254074e-01 -1.03418398e+00
7.72479951e-01 1.54522157e+00 8.91605794e-01 5.60396552e-01
-1.41696465e+00 -6.99189425e-01 6.06576264e-01 1.14843026e-01
-1.35721481e+00 -3.36953640e-01 6.49007261e-01 -6.48992002e-01
6.80080235e-01 2.33791739e-01 8.31085503e-01 1.06361771e+00
2.62350351e-01 8.69757950e-01 4.63497370e-01 -3.75789106e-01
-2.57629186e-01 1.77759215e-01 1.26609460e-01 6.30395472e-01
5.77560544e-01 4.71657991e-01 -6.89583898e-01 -6.51028603e-02
8.18039417e-01 1.55753180e-01 -1.96101904e-01 -1.34578645e+00
-1.28015590e+00 6.09548807e-01 5.59029818e-01 -2.43902463e-03
-2.27779582e-01 4.70875651e-01 1.47376016e-01 1.64755546e-02
6.38926625e-01 -1.63436651e-01 -2.58432597e-01 1.81162328e-01
-1.04511940e+00 5.28993607e-01 3.16041112e-01 1.36207485e+00
5.37733674e-01 1.66875660e-01 -7.86866665e-01 1.15015514e-01
7.23481953e-01 8.90099466e-01 -1.33691818e-01 -8.80391121e-01
1.08662985e-01 5.58951616e-01 4.98194516e-01 -8.51041198e-01
-1.57766536e-01 -7.39341497e-01 -1.87373742e-01 3.70241553e-01
3.05744022e-01 3.37390691e-01 -1.13262856e+00 1.96677041e+00
9.39201593e-01 5.73312044e-01 -5.48570395e-01 7.51606643e-01
3.69248182e-01 2.99519032e-01 1.34399280e-01 -1.07175037e-01
1.43523443e+00 -1.16824496e+00 -8.70511711e-01 2.43690804e-01
3.80362928e-01 -8.85551691e-01 5.12927592e-01 2.62932628e-02
-1.11549711e+00 -9.09771800e-01 -8.65621507e-01 -8.93955529e-02
-3.93773824e-01 1.33952528e-01 4.12206084e-01 8.15239489e-01
-1.26864743e+00 2.65612990e-01 -1.24151945e+00 -5.80883086e-01
1.96350604e-01 4.66606110e-01 -8.78016204e-02 1.78170372e-02
-6.38763368e-01 9.55674946e-01 2.30374366e-01 2.44428450e-03
-1.01173806e+00 -1.01474428e+00 -7.13054955e-01 -4.50094752e-02
4.94828254e-01 -1.09323013e+00 1.15682256e+00 -2.71269053e-01
-1.40439785e+00 7.82644510e-01 -5.43145657e-01 -4.76288974e-01
8.01722884e-01 -6.31201506e-01 -3.89769077e-01 -2.79629454e-02
1.96342304e-01 8.18545461e-01 1.05791914e+00 -1.24495184e+00
-7.16931283e-01 -3.38427275e-01 3.71218100e-02 1.18211031e-01
-3.47593129e-01 1.50340840e-01 -1.22873580e+00 -6.85133696e-01
-3.48353833e-02 -1.24361062e+00 -1.72878001e-02 7.49394834e-01
-2.00393736e-01 -2.99626857e-01 1.34798753e+00 -3.39436114e-01
1.13039672e+00 -2.10662580e+00 1.59470081e-01 -6.89953193e-02
3.69866133e-01 3.56550992e-01 -1.19547047e-01 1.26458094e-01
1.63940847e-01 -3.66048545e-01 4.91349280e-01 -8.31813157e-01
2.40001798e-01 6.35726796e-03 -3.27642590e-01 6.79901779e-01
2.66754776e-01 1.14833653e+00 -9.52094078e-01 -3.85183603e-01
6.63923144e-01 7.79639125e-01 -5.82756341e-01 4.10740264e-02
-4.31408048e-01 5.65027833e-01 -4.04860348e-01 5.42288184e-01
8.73157620e-01 -2.80325174e-01 -1.58757553e-01 -4.51376349e-01
-6.00357473e-01 1.22089252e-01 -1.09304869e+00 1.78390026e+00
1.22136762e-02 5.83523929e-01 -3.44545260e-04 -1.88097104e-01
6.17667198e-01 2.66149670e-01 6.97406232e-01 -4.17412579e-01
-5.68249002e-02 -1.32384464e-01 -3.04023504e-01 9.58121102e-03
6.49855494e-01 3.40566576e-01 1.88757554e-01 1.72047317e-01
3.22032757e-02 5.90221047e-01 3.40712480e-02 3.44505638e-01
1.00984168e+00 8.55938256e-01 -2.49058425e-01 -1.51448712e-01
3.70194137e-01 1.41904593e-01 5.89857817e-01 8.68305802e-01
-2.91397750e-01 4.16395664e-01 -3.59718859e-01 -5.55085957e-01
-1.04818380e+00 -1.38256133e+00 -1.41350329e-01 7.88066447e-01
4.09071118e-01 -3.57100725e-01 -4.12668824e-01 -5.63213170e-01
3.08946133e-01 5.62253356e-01 -7.52868950e-01 -1.81406498e-01
-6.89474344e-01 -2.63278008e-01 1.75174326e-01 5.59469700e-01
1.55539021e-01 -4.93218422e-01 -6.65505469e-01 4.20063138e-01
-5.21187223e-02 -1.35770380e+00 -8.94679904e-01 -1.17420033e-01
-8.99849713e-01 -8.43989730e-01 -6.40567482e-01 -3.85767192e-01
6.16688788e-01 6.78104341e-01 9.34811771e-01 -5.02013713e-02
-2.96786070e-01 8.13970327e-01 -1.68281302e-01 -3.03150177e-01
-1.89988554e-01 -2.13907972e-01 3.49660963e-01 1.27182707e-01
4.64624941e-01 -6.26642182e-02 -5.71489573e-01 4.69638854e-01
-5.80335855e-01 3.06863873e-03 2.77061373e-01 5.18831313e-01
6.40846193e-01 -3.37718695e-01 -1.08449988e-01 -3.45126390e-01
-2.87404627e-01 -2.45930135e-01 -1.17684007e+00 2.93426037e-01
-4.75367606e-01 2.28207279e-02 -8.17449093e-02 -8.84917438e-01
-1.02606380e+00 3.63223076e-01 3.21689814e-01 -1.32241488e+00
1.74830079e-01 -3.25689733e-01 -3.65778565e-01 -4.77344424e-01
2.07424223e-01 2.62060761e-01 -1.22252591e-01 -6.05398595e-01
6.44808054e-01 2.65136100e-02 6.35066152e-01 -3.17878217e-01
1.53127050e+00 1.03727329e+00 8.42436925e-02 -4.67307299e-01
-9.09359872e-01 -9.04465199e-01 -8.22756350e-01 -4.19863015e-01
9.62212920e-01 -1.26099479e+00 -7.99355805e-01 3.69081944e-01
-1.18818605e+00 -2.45450214e-02 -4.00398731e-01 7.22513735e-01
-4.31229174e-01 3.14922959e-01 -4.30946499e-01 -8.28568578e-01
-7.46267065e-02 -1.08186996e+00 1.45469260e+00 1.75709441e-01
-1.68069694e-02 -1.08581305e+00 3.48066390e-01 8.17995332e-03
4.03877884e-01 2.72460848e-01 4.05601501e-01 -1.90029755e-01
-1.54684997e+00 -3.59282672e-01 -1.69835255e-01 -2.46879339e-01
1.44682243e-01 -8.71682391e-02 -9.78129566e-01 -6.08426571e-01
-7.91121721e-02 3.88230324e-01 8.30378413e-01 7.36020207e-01
4.81019408e-01 4.87060435e-02 -8.56894910e-01 8.68858099e-01
1.43973911e+00 -6.93904310e-02 3.90119702e-01 3.98393363e-01
1.07158780e+00 1.04522370e-01 7.95777559e-01 3.81643087e-01
4.58041131e-01 1.44976711e+00 4.46419090e-01 -1.05574019e-01
-5.86821198e-01 -4.35661077e-01 5.92600346e-01 5.41409969e-01
1.93976104e-01 -3.02630663e-01 -4.38123643e-01 6.57649636e-01
-1.97834957e+00 -1.10151815e+00 -4.07323450e-01 2.27623105e+00
3.33178937e-01 2.46862784e-01 2.92997926e-01 -4.58203197e-01
9.62926865e-01 7.72272944e-02 -7.28824496e-01 3.25060934e-01
-4.32419032e-02 -1.64600447e-01 9.11025047e-01 4.72434044e-01
-1.25163877e+00 1.19136977e+00 6.33182669e+00 5.60132146e-01
-9.98533607e-01 2.98294514e-01 -4.08396900e-01 -3.63172561e-01
-1.77426189e-01 1.56044453e-01 -1.60414898e+00 4.52751487e-01
7.49533296e-01 -4.49065343e-02 9.67082307e-02 5.07263005e-01
2.79497325e-01 4.13278162e-01 -1.09963560e+00 8.27474058e-01
-2.95675807e-02 -1.36424494e+00 1.66895404e-01 3.52035522e-01
5.11759520e-01 3.00774157e-01 2.61278480e-01 2.35920936e-01
3.76903802e-01 -9.71107483e-02 1.33065403e+00 6.40645444e-01
4.84871447e-01 -2.85484046e-01 1.17277063e-01 1.55704424e-01
-1.75357854e+00 2.85846084e-01 -2.60341167e-01 3.17209303e-01
4.83368576e-01 2.35141084e-01 -5.21380067e-01 7.03768075e-01
8.60475361e-01 9.05474663e-01 -6.61035180e-01 1.57089567e+00
-2.31771497e-03 3.42567921e-01 -5.93154013e-01 3.92926157e-01
1.05938971e-01 1.36198014e-01 1.16872883e+00 1.04569304e+00
3.26508939e-01 -4.23511028e-01 3.61987799e-01 1.05202198e+00
1.34659305e-01 -4.55631763e-01 -4.67362881e-01 2.21619308e-01
5.74609995e-01 1.17721379e+00 -6.45913363e-01 -2.85171837e-01
-5.75318277e-01 9.72367406e-01 1.01460218e-01 2.66942263e-01
-1.15385473e+00 3.44126940e-01 6.90352321e-01 4.10173476e-01
1.03714156e+00 -4.12698328e-01 2.70154208e-01 -1.25100064e+00
7.52042234e-02 -3.23006958e-01 2.32785642e-01 -7.82200158e-01
-1.19514120e+00 3.39686215e-01 1.46597058e-01 -1.62789428e+00
1.35635182e-01 -5.10329664e-01 -3.40218425e-01 7.52168179e-01
-1.74526691e+00 -1.70531344e+00 -2.70540446e-01 7.03617990e-01
4.77257878e-01 1.24657959e-01 4.60706443e-01 6.54879570e-01
-3.49250257e-01 6.16852105e-01 1.64385810e-01 1.01394556e-01
9.16665792e-01 -1.17904425e+00 9.71563101e-01 1.08748102e+00
2.74982363e-01 8.16342831e-01 7.28974640e-01 -9.31768179e-01
-1.68327618e+00 -1.46499979e+00 7.72931814e-01 -1.15275121e+00
7.87881255e-01 -5.66376746e-01 -7.71829665e-01 9.81170356e-01
9.69756395e-02 2.97485322e-01 3.52827013e-01 4.73348871e-02
-4.54559207e-01 5.66711016e-02 -8.27381909e-01 6.18646026e-01
1.30108607e+00 -4.31100816e-01 -4.42881674e-01 2.53616363e-01
9.71661568e-01 -6.71235859e-01 -7.22093821e-01 3.53981733e-01
6.86848879e-01 -4.81995940e-01 1.45040190e+00 -4.52711642e-01
-6.51608348e-01 -9.23817098e-01 4.56817858e-02 -7.21337795e-01
-9.45526421e-01 -7.52272904e-01 -7.72389829e-01 1.34271812e+00
-2.57351905e-01 -3.34987402e-01 9.30375576e-01 4.60667759e-01
-2.91898936e-01 -1.17497824e-01 -9.13204074e-01 -1.18175495e+00
-3.55103791e-01 -1.88678131e-01 3.89574319e-01 7.29455233e-01
-8.12184751e-01 4.48508821e-02 -6.30948067e-01 7.56508887e-01
1.24108052e+00 1.19040653e-01 1.03679323e+00 -1.39046562e+00
-3.11545700e-01 -1.18870296e-01 -6.75175130e-01 -1.44202888e+00
-6.71559274e-02 -6.77702129e-01 -5.03853429e-03 -1.09818721e+00
2.76160359e-01 -3.41757774e-01 -4.67335492e-01 2.50176489e-01
-3.12707007e-01 2.11795881e-01 5.89866817e-01 4.28432137e-01
-9.08989012e-01 6.91701055e-01 1.02047944e+00 -1.34096712e-01
-1.60155706e-02 9.25588533e-02 -2.29617998e-01 5.61478674e-01
1.68560416e-01 -9.46921051e-01 -8.00735429e-02 -7.04827309e-01
-2.34189376e-01 -2.94090003e-01 6.82382882e-01 -1.11045265e+00
5.61572075e-01 2.77182221e-01 5.44430315e-01 -1.31886578e+00
6.40713274e-01 -1.26850510e+00 4.46740210e-01 6.21229053e-01
7.43742064e-02 2.02548310e-01 5.20769477e-01 9.90295231e-01
1.35383248e-01 2.57591486e-01 7.01909304e-01 2.46766284e-01
-6.93724632e-01 7.72642493e-01 -5.64985462e-02 -2.88282067e-01
1.10654032e+00 -5.29697478e-01 -2.38826483e-01 3.26035195e-03
-6.36413336e-01 3.92167568e-01 8.24783266e-01 8.12131405e-01
2.43963599e-01 -1.65704477e+00 -5.14928818e-01 1.48196518e-01
3.30830216e-01 -4.14295048e-01 1.75336421e-01 8.77832711e-01
-1.14087239e-01 5.12637734e-01 -1.31984174e-01 -1.14090145e+00
-1.38117182e+00 8.40716779e-01 2.63534993e-01 -2.92770833e-01
-8.28486800e-01 6.06151938e-01 3.92631531e-01 -1.97877556e-01
4.19598728e-01 -2.89480895e-01 6.38276413e-02 -1.99854940e-01
5.11959851e-01 1.78068608e-01 -2.26783484e-01 -8.22309613e-01
-4.93859053e-01 9.67636526e-01 -2.69247502e-01 -6.90983534e-02
9.96474445e-01 -3.84018809e-01 4.36707407e-01 4.66114134e-01
7.39046872e-01 2.84177065e-01 -1.78123975e+00 -4.99825627e-01
1.19802475e-01 -7.25547552e-01 1.77638844e-01 -5.29298782e-01
-9.49288666e-01 7.67365038e-01 1.12594843e+00 -1.50982544e-01
6.79438114e-01 -3.35554816e-02 6.36865616e-01 -5.43027483e-02
5.90368629e-01 -6.23117983e-01 -2.42371764e-02 4.07282323e-01
3.57812315e-01 -1.01806152e+00 6.70091957e-02 -4.11232114e-01
-1.08397983e-01 7.42720664e-01 5.07308185e-01 -8.11861530e-02
4.78886306e-01 2.50424415e-01 -6.08333619e-04 -2.15320662e-01
-7.20487475e-01 -4.21602309e-01 4.68434602e-01 6.33082449e-01
5.00767417e-02 -4.00685102e-01 2.57278085e-01 -2.10366830e-01
4.50848520e-01 4.46676277e-02 -4.02562842e-02 9.54974592e-01
-4.09122854e-01 -1.37370217e+00 -6.21249378e-01 -1.44721687e-01
-1.92451432e-01 1.32532164e-01 3.19636799e-03 8.91082942e-01
1.64196327e-01 6.12036824e-01 1.90115944e-01 -2.07397416e-01
4.22165036e-01 -1.01493776e-01 8.45575094e-01 -5.82628846e-01
-5.84089518e-01 4.93848503e-01 -4.00211841e-01 -7.31210232e-01
-7.20683157e-01 -9.76118565e-01 -8.37064147e-01 -1.50553152e-01
-7.86954105e-01 -1.10054210e-01 5.56709230e-01 5.59553742e-01
6.20058358e-01 5.68824947e-01 3.14054281e-01 -9.74224985e-01
-4.96835887e-01 -6.36882663e-01 -2.72219241e-01 4.56676364e-01
5.28315544e-01 -1.19990754e+00 -7.92108551e-02 2.29256585e-01] | [6.352200984954834, -2.0902817249298096] |
fa6f17a1-9dac-4b77-a48e-13ac5819f645 | adaptive-dithering-using-curved-markov | 2001.06983 | null | https://arxiv.org/abs/2001.06983v1 | https://arxiv.org/pdf/2001.06983v1.pdf | Adaptive Dithering Using Curved Markov-Gaussian Noise in the Quantized Domain for Mapping SDR to HDR Image | High Dynamic Range (HDR) imaging is gaining increased attention due to its realistic content, for not only regular displays but also smartphones. Before sufficient HDR content is distributed, HDR visualization still relies mostly on converting Standard Dynamic Range (SDR) content. SDR images are often quantized, or bit depth reduced, before SDR-to-HDR conversion, e.g. for video transmission. Quantization can easily lead to banding artefacts. In some computing and/or memory I/O limited environment, the traditional solution using spatial neighborhood information is not feasible. Our method includes noise generation (offline) and noise injection (online), and operates on pixels of the quantized image. We vary the magnitude and structure of the noise pattern adaptively based on the luma of the quantized pixel and the slope of the inverse-tone mapping function. Subjective user evaluations confirm the superior performance of our technique. | ['Guan-Ming Su', 'Irene Cheng', 'Subhayan Mukherjee'] | 2020-01-20 | null | null | null | null | ['tone-mapping', 'inverse-tone-mapping'] | ['computer-vision', 'computer-vision'] | [ 6.55028045e-01 -5.31338632e-01 2.13892316e-03 -3.12668383e-02
-4.98791814e-01 -5.91255784e-01 1.20213479e-01 -2.64243949e-02
-3.09761554e-01 6.49290919e-01 1.45929635e-01 -4.43460435e-01
6.75240383e-02 -7.25690365e-01 -2.59172469e-01 -7.22661912e-01
3.49530093e-02 -2.85471916e-01 5.22062778e-01 -2.03759506e-01
2.06560150e-01 5.16273856e-01 -1.68672466e+00 9.15142149e-02
9.68472779e-01 1.28166044e+00 5.16582668e-01 1.00769413e+00
1.94887619e-03 7.74295628e-01 -7.62933731e-01 4.71625105e-02
5.85748911e-01 -5.57763219e-01 -1.78076744e-01 1.75256729e-02
1.53955817e-01 -7.54183948e-01 -6.10296607e-01 1.20280194e+00
5.71929991e-01 1.63599074e-01 -7.54415104e-03 -7.67258346e-01
-5.87741256e-01 1.66081145e-01 -8.02824438e-01 3.37659836e-01
6.55754685e-01 1.70137733e-01 3.50935936e-01 -3.52665126e-01
7.36902177e-01 7.98105836e-01 2.05728799e-01 2.16004223e-01
-1.36862409e+00 -6.54230058e-01 -5.39042413e-01 3.46793354e-01
-1.53968894e+00 -5.59242606e-01 6.97357178e-01 5.23650423e-02
6.54625177e-01 4.31621462e-01 7.72393107e-01 4.98243868e-01
1.92927361e-01 -7.06771314e-02 1.40125322e+00 -4.02941078e-01
2.90606499e-01 3.35278995e-02 -3.50973219e-01 1.00401290e-01
7.19379410e-02 3.05224918e-02 -5.58161795e-01 5.86863458e-02
1.28235364e+00 -9.10788924e-02 -7.75503576e-01 -9.48254243e-02
-1.00415039e+00 2.43094072e-01 2.14815021e-01 1.75928578e-01
-1.68242082e-01 1.93809718e-01 1.22440092e-01 5.92109740e-01
1.83403343e-01 2.58513242e-01 1.10345125e-01 -5.15481114e-01
-1.08425355e+00 -3.09730828e-01 4.05005246e-01 7.05296397e-01
6.21517301e-01 -3.66980471e-02 9.05413702e-02 9.00950253e-01
-3.13585289e-02 6.58358216e-01 5.05807400e-01 -1.26942098e+00
1.90973431e-01 9.63106751e-02 2.85131007e-01 -8.50477040e-01
5.24075143e-02 -2.52184197e-02 -1.10872746e+00 5.17303646e-01
4.06752795e-01 1.32562155e-02 -7.23644912e-01 1.14233005e+00
1.80969596e-01 5.82312047e-02 -1.90129563e-01 1.25992751e+00
5.10196924e-01 8.96928191e-01 -2.43477389e-01 -6.64030612e-01
1.13392568e+00 -1.99409738e-01 -1.09573281e+00 1.05167337e-01
-2.78065279e-02 -1.05853200e+00 1.32485116e+00 6.37068033e-01
-1.09314084e+00 -5.65126956e-01 -1.17544508e+00 -4.08575386e-01
4.71822582e-02 -1.59378111e-01 4.91062105e-02 8.73498976e-01
-1.18453586e+00 5.13645470e-01 -5.94287097e-01 -3.66618973e-03
-8.94735679e-02 1.68803230e-01 -1.88655004e-01 -3.98856193e-01
-1.16147363e+00 5.04927516e-01 -1.78132847e-01 9.00864601e-02
3.01602371e-02 -7.69025028e-01 -4.70298767e-01 1.99866489e-01
2.67072469e-01 -4.66211945e-01 7.87928224e-01 -9.18253720e-01
-1.94481099e+00 7.00714290e-01 -3.15703340e-02 -3.72649491e-01
5.71366727e-01 -4.98079322e-02 -5.25430143e-01 4.56651062e-01
-3.05946678e-01 2.24317119e-01 1.09286356e+00 -1.10252142e+00
-3.62901151e-01 -2.62482643e-01 -2.09513173e-01 3.37916911e-01
-1.99330285e-01 7.14566782e-02 -5.95944285e-01 -5.87708175e-01
4.42977190e-01 -6.25491977e-01 -7.40558356e-02 4.74622637e-01
-1.96053758e-01 5.18103540e-01 1.18200612e+00 -7.75159478e-01
1.53528929e+00 -2.47881222e+00 -4.26356941e-01 3.80805522e-01
1.60299659e-01 2.65242070e-01 2.17977419e-01 1.30394921e-01
-1.84736267e-01 -1.08571097e-01 4.52020112e-03 2.74416149e-01
-3.40438873e-01 -2.18279257e-01 -4.30553854e-01 6.32435977e-01
-4.32175249e-01 3.04175407e-01 -4.86399353e-01 -3.70140076e-01
5.30339837e-01 7.74977982e-01 -2.08063915e-01 1.94923520e-01
1.53687492e-01 6.39216185e-01 5.79797402e-02 4.90438044e-01
9.06828523e-01 -2.63562143e-01 2.80665576e-01 -5.91485083e-01
-4.22762126e-01 2.18038350e-01 -1.36518574e+00 1.20111942e+00
-7.01537013e-01 9.98327196e-01 1.20075233e-01 -1.68505654e-01
1.12846613e+00 -5.06016165e-02 4.16575760e-01 -1.21918881e+00
-1.48176908e-01 3.96418154e-01 -8.70324299e-02 -3.99522245e-01
7.78989732e-01 6.82305843e-02 2.70731002e-01 3.47189218e-01
-7.75738776e-01 -1.56828031e-01 1.47002637e-01 1.17169030e-01
1.09034264e+00 -9.00884494e-02 5.03483295e-01 5.03405333e-02
2.40166292e-01 -3.87255341e-01 5.79696186e-02 6.60606563e-01
-2.28125855e-01 9.04968858e-01 5.38274467e-01 -2.29252316e-02
-1.41951084e+00 -1.03305697e+00 -3.44736665e-01 8.03829968e-01
5.79391658e-01 -2.62421370e-01 -4.89516824e-01 2.30006441e-01
-4.02744055e-01 5.00019133e-01 -3.98285985e-02 1.02444857e-01
-6.57541454e-01 -3.77740085e-01 1.20739602e-01 9.01272241e-03
9.21791911e-01 -7.37782836e-01 -1.13803983e+00 1.89308852e-01
-2.18925595e-01 -1.09139049e+00 -5.51931858e-01 1.30056277e-01
-8.50596905e-01 -5.89689612e-01 -9.10711765e-01 -2.95365453e-01
2.61429816e-01 5.68112671e-01 8.57600152e-01 -7.76751861e-02
-5.42385936e-01 2.73402482e-01 -4.39013094e-01 3.80521864e-01
-3.51025790e-01 -3.37974668e-01 -3.37035954e-01 -7.81478658e-02
-1.16400212e-01 -6.64183021e-01 -1.13088155e+00 5.46000063e-01
-9.60541606e-01 1.86695293e-01 3.93673986e-01 4.87960309e-01
7.82017529e-01 3.23572129e-01 1.23424530e-01 -4.38845098e-01
5.33049703e-01 1.22658079e-02 -8.33340704e-01 -1.09285511e-01
-4.21410233e-01 -3.63331050e-01 7.36538768e-01 -3.95985067e-01
-9.72272694e-01 -2.64135450e-01 -4.67824936e-02 -5.61027110e-01
-2.49005314e-02 -1.15803696e-01 -2.12900490e-01 -8.88707712e-02
6.65702999e-01 2.65452206e-01 1.07480623e-01 -1.52466819e-01
1.84009716e-01 9.73573506e-01 6.81262791e-01 1.33309647e-01
7.26527810e-01 6.67946994e-01 -2.15081628e-02 -1.26608443e+00
-5.82928024e-02 -3.90169382e-01 2.64714807e-02 -4.11120623e-01
7.23146796e-01 -7.79195011e-01 -8.04485679e-01 4.18486148e-01
-6.95383251e-01 -6.69443309e-01 -4.42517519e-01 4.11591321e-01
-3.96876305e-01 3.73044193e-01 -7.78104424e-01 -6.56003714e-01
-1.16974495e-01 -1.06939542e+00 8.21518123e-01 4.75681514e-01
8.27434584e-02 -6.09439731e-01 -1.60297424e-01 1.29024193e-01
8.24309170e-01 3.52878682e-03 7.68129587e-01 4.02182907e-01
-8.67818654e-01 3.16237062e-02 -6.47776008e-01 8.83411095e-02
2.71597564e-01 1.39308749e-02 -8.99451196e-01 -1.99744716e-01
1.78724855e-01 -7.65492916e-02 5.25157511e-01 6.17205977e-01
1.46227467e+00 -1.15058191e-01 -9.08388868e-02 7.76270688e-01
1.73109984e+00 2.86014259e-01 1.35152447e+00 4.43080395e-01
4.80243981e-01 2.23914638e-01 7.28547454e-01 6.37144566e-01
-8.47767368e-02 1.03827488e+00 2.12955058e-01 -3.84857744e-01
-4.32667732e-01 2.26502623e-02 3.51709068e-01 4.21640486e-01
-5.40486425e-02 -3.83411974e-01 -6.26021743e-01 -2.31925361e-02
-1.03615010e+00 -8.77641857e-01 -1.36167437e-01 2.51765203e+00
9.49041426e-01 -2.88976766e-02 -3.52840014e-02 2.13417917e-01
8.47237766e-01 1.98194772e-01 -6.33721948e-01 -3.97105187e-01
-2.70236760e-01 1.63710609e-01 8.01167309e-01 3.48191768e-01
-5.31287491e-01 3.50385487e-01 5.84681559e+00 9.12528813e-01
-1.60343862e+00 -1.10484898e-01 8.36046040e-01 -2.99957603e-01
-4.50924098e-01 -2.60505706e-01 -3.24030191e-01 8.42131436e-01
8.37593675e-01 -1.69680007e-02 8.92770350e-01 6.09148324e-01
6.70944810e-01 -7.13486731e-01 -6.45557642e-01 1.59554636e+00
-1.95649877e-01 -1.00078249e+00 -3.87842506e-01 3.04217488e-01
4.84171897e-01 -4.24209356e-01 5.46649933e-01 -3.38630080e-01
-2.28212774e-01 -8.17806244e-01 5.46950936e-01 1.21952608e-01
1.55707645e+00 -6.53141975e-01 5.16307056e-01 -1.26690775e-01
-1.19140434e+00 -1.80259887e-02 -3.18067133e-01 2.88802236e-01
3.05643380e-01 8.26457918e-01 -4.38560396e-01 -9.31249112e-02
9.90046859e-01 2.71720827e-01 -3.77131581e-01 8.06622028e-01
-1.20384749e-02 3.91127288e-01 -5.27022541e-01 2.84809083e-01
-3.15761328e-01 -5.31846404e-01 5.15520036e-01 8.65278065e-01
5.65363586e-01 4.58374530e-01 -2.51579374e-01 6.35741472e-01
-2.64867228e-02 -2.97908895e-02 -4.22121465e-01 2.11888090e-01
6.30843222e-01 1.11224926e+00 -7.83413291e-01 -9.08442289e-02
-5.22390425e-01 1.31605196e+00 -4.58977997e-01 5.13862610e-01
-8.21793616e-01 -8.44565094e-01 7.24883020e-01 6.79474473e-01
3.36435139e-01 -3.48577619e-01 -3.62744302e-01 -7.93893754e-01
1.26416400e-01 -9.39362884e-01 1.08683325e-01 -1.03403652e+00
-8.80033433e-01 3.94320369e-01 -2.29810551e-01 -1.51387501e+00
-1.42952040e-01 -1.41146794e-01 -3.24580848e-01 8.74370396e-01
-1.56602645e+00 -3.59629631e-01 -5.89264929e-01 6.84295058e-01
3.06452841e-01 3.23190272e-01 3.87914211e-01 5.87946832e-01
-1.54549271e-01 3.82099330e-01 4.31031197e-01 -2.54583567e-01
9.73613262e-01 -1.05197155e+00 -1.63230579e-02 8.65405679e-01
-3.48076403e-01 5.20428061e-01 9.75891709e-01 -4.94161904e-01
-1.53516901e+00 -6.45819962e-01 2.53654391e-01 2.72620767e-01
4.95550364e-01 -3.27924579e-01 -1.15086830e+00 8.75956044e-02
1.63723752e-01 1.03935741e-01 5.80984950e-01 -6.18623495e-01
-2.15672001e-01 -5.52972555e-01 -1.43383455e+00 7.86321163e-01
6.93426788e-01 -6.99433863e-01 2.68191665e-01 -1.68638244e-01
6.90751672e-01 -6.01154029e-01 -8.62837076e-01 -9.54108313e-02
6.91103935e-01 -1.49086225e+00 9.14923549e-01 6.42577648e-01
3.73683095e-01 -7.93001771e-01 -3.37615401e-01 -1.01543677e+00
1.14889042e-02 -9.35127914e-01 1.00869976e-03 9.67673123e-01
4.73863967e-02 -6.04245007e-01 6.14791512e-01 5.89167655e-01
2.33106241e-01 -1.75846562e-01 -9.79412735e-01 -4.70417768e-01
-7.98642457e-01 -2.10905164e-01 2.40864664e-01 6.40916228e-01
-7.64706358e-02 -5.72166555e-02 -4.33765978e-01 1.83767408e-01
7.16268778e-01 1.89285755e-01 5.29258907e-01 -6.19445086e-01
-5.39053142e-01 -1.53477425e-02 -4.03186888e-01 -1.35416794e+00
-6.64119422e-01 -9.22951475e-02 5.46577871e-02 -1.28017116e+00
-6.49837479e-02 -6.08739913e-01 1.18303299e-01 -5.26150204e-02
3.50632332e-02 8.32641006e-01 2.30204329e-01 3.58854920e-01
-3.74345541e-01 1.73888251e-01 1.26369464e+00 3.00340056e-01
-6.13558292e-01 -2.76895314e-01 -1.75777152e-01 4.00529444e-01
7.35225081e-01 -7.45102242e-02 -4.25515890e-01 -1.87909707e-01
4.70253736e-01 5.37755072e-01 4.01546150e-01 -1.10003483e+00
1.20795220e-01 -5.54407984e-02 5.72963595e-01 -3.62239271e-01
3.28675538e-01 -1.03493536e+00 5.75883865e-01 3.43474418e-01
-1.87112853e-01 -1.48920104e-01 -2.21058533e-01 4.40919429e-01
-3.00344229e-01 1.03276864e-01 1.27813053e+00 -1.77572612e-02
-6.13865495e-01 1.64620541e-02 -5.08045495e-01 -2.19759852e-01
7.31048763e-01 -7.62884557e-01 -5.09731114e-01 -7.31899917e-01
-5.48171163e-01 -3.63850802e-01 1.02072287e+00 -7.99349323e-02
8.03215504e-01 -1.01642394e+00 -1.47280097e-01 4.73926723e-01
-2.60432065e-01 -1.96455121e-01 6.92371607e-01 7.89453447e-01
-1.00958753e+00 -6.85222819e-02 -2.30696216e-01 -5.84674060e-01
-1.40862405e+00 3.90401065e-01 2.50728071e-01 -4.62646829e-03
-9.67678666e-01 2.90938109e-01 -1.16526663e-01 7.70066142e-01
6.74138665e-02 -1.61016703e-01 3.60248983e-02 -1.37061924e-01
8.70344222e-01 7.87463725e-01 8.05056375e-03 -2.09226355e-01
-5.30369803e-02 7.56983399e-01 1.77943900e-01 -4.21320289e-01
8.93631995e-01 -8.05332959e-01 -1.01721704e-01 4.51619178e-01
1.19701624e+00 1.99512228e-01 -1.21388781e+00 1.41207799e-01
-5.67056715e-01 -1.06626523e+00 3.03826660e-01 -3.83197576e-01
-9.54710066e-01 8.03760886e-01 1.21269691e+00 6.22630000e-01
1.70959556e+00 -2.16935441e-01 1.00798666e+00 -3.54005061e-02
4.60174590e-01 -1.09118879e+00 1.59678847e-01 -8.21096674e-02
4.30038363e-01 -9.40827250e-01 1.60203978e-01 -5.66985190e-01
-4.30724353e-01 1.29072475e+00 1.65122494e-01 1.53853476e-01
4.59431946e-01 6.38880730e-01 2.75515288e-01 3.81239563e-01
-4.13425267e-01 -4.75010648e-02 -3.26329887e-01 6.88423395e-01
4.16887522e-01 -1.98319167e-01 -2.14208484e-01 -1.62520081e-01
-7.27426559e-02 1.92420676e-01 9.60597396e-01 5.99875808e-01
-6.19459391e-01 -7.87877262e-01 -6.61572158e-01 3.75514001e-01
-5.45772076e-01 -1.22052029e-01 3.17074895e-01 4.39835250e-01
-3.06912195e-02 9.92382467e-01 4.32790428e-01 -4.47499081e-02
2.71263421e-01 -4.87834215e-01 4.04650271e-01 -9.50588062e-02
-1.60396323e-01 4.76142228e-01 -2.44247526e-01 -8.68624389e-01
-3.25966388e-01 -2.72121340e-01 -1.08218122e+00 -6.54947042e-01
-6.72729015e-02 -2.81214207e-01 9.22676265e-01 1.67308867e-01
2.50899285e-01 4.25590187e-01 7.98135996e-01 -5.54214120e-01
-2.40857109e-01 -4.69077945e-01 -1.02423406e+00 2.94608980e-01
5.81058025e-01 -9.60120410e-02 -5.36265552e-01 2.57621855e-01] | [10.87425708770752, -2.3818745613098145] |
88578db4-55be-450e-97a4-621963fd6f04 | mukea-multimodal-knowledge-extraction-and | 2203.09138 | null | https://arxiv.org/abs/2203.09138v1 | https://arxiv.org/pdf/2203.09138v1.pdf | MuKEA: Multimodal Knowledge Extraction and Accumulation for Knowledge-based Visual Question Answering | Knowledge-based visual question answering requires the ability of associating external knowledge for open-ended cross-modal scene understanding. One limitation of existing solutions is that they capture relevant knowledge from text-only knowledge bases, which merely contain facts expressed by first-order predicates or language descriptions while lacking complex but indispensable multimodal knowledge for visual understanding. How to construct vision-relevant and explainable multimodal knowledge for the VQA scenario has been less studied. In this paper, we propose MuKEA to represent multimodal knowledge by an explicit triplet to correlate visual objects and fact answers with implicit relations. To bridge the heterogeneous gap, we propose three objective losses to learn the triplet representations from complementary views: embedding structure, topological relation and semantic space. By adopting a pre-training and fine-tuning learning strategy, both basic and domain-specific multimodal knowledge are progressively accumulated for answer prediction. We outperform the state-of-the-art by 3.35% and 6.08% respectively on two challenging knowledge-required datasets: OK-VQA and KRVQA. Experimental results prove the complementary benefits of the multimodal knowledge with existing knowledge bases and the advantages of our end-to-end framework over the existing pipeline methods. The code is available at https://github.com/AndersonStra/MuKEA. | ['Qi Wu', 'Mingxin Cui', 'Yue Hu', 'Bang Liu', 'Jing Yu', 'Yang Ding'] | 2022-03-17 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Ding_MuKEA_Multimodal_Knowledge_Extraction_and_Accumulation_for_Knowledge-Based_Visual_Question_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Ding_MuKEA_Multimodal_Knowledge_Extraction_and_Accumulation_for_Knowledge-Based_Visual_Question_CVPR_2022_paper.pdf | cvpr-2022-1 | ['implicit-relations'] | ['natural-language-processing'] | [-1.35222495e-01 1.40563682e-01 -1.93202257e-01 -4.16683644e-01
-8.59818578e-01 -7.38710880e-01 4.93450195e-01 1.23949453e-01
-1.88785598e-01 7.03559220e-01 3.10732126e-01 -2.74073780e-01
-2.49733344e-01 -8.22799742e-01 -8.76399636e-01 -3.60680670e-01
4.76536214e-01 5.91088355e-01 3.34872991e-01 -3.62072051e-01
1.85188368e-01 1.90801635e-01 -1.72640979e+00 9.13163424e-01
9.41097796e-01 1.23179197e+00 1.72263995e-01 5.67223847e-01
-6.53470576e-01 1.28005648e+00 -1.46100491e-01 -9.15415943e-01
-1.85946465e-01 -2.52537429e-01 -1.18403125e+00 -7.95816779e-02
7.16495991e-01 -1.56431511e-01 -5.11637151e-01 6.66683674e-01
4.80386585e-01 1.32584721e-01 5.56508124e-01 -1.34589970e+00
-1.46513271e+00 1.76632389e-01 -4.41763848e-01 9.78772119e-02
6.58619463e-01 2.57314414e-01 1.29100478e+00 -1.14542711e+00
7.84914434e-01 1.31977987e+00 3.48410904e-01 5.22603631e-01
-1.10403037e+00 -3.13271970e-01 2.70827949e-01 9.86155093e-01
-1.35460889e+00 -3.13477665e-01 1.04629600e+00 -4.28880870e-01
9.52495515e-01 3.97500753e-01 5.01808345e-01 1.05180979e+00
-4.83615190e-01 1.11020720e+00 1.01764667e+00 -3.20968926e-01
-1.96267709e-01 3.71969342e-01 1.94689929e-01 1.12639081e+00
-6.10746816e-02 -1.59561560e-01 -8.36009085e-01 1.10254241e-02
5.93362093e-01 -9.94202420e-02 -4.02733326e-01 -7.95441985e-01
-1.41158402e+00 8.69236946e-01 8.08772147e-01 4.69842926e-02
-3.34539682e-01 2.02377036e-01 4.56590742e-01 8.68759006e-02
1.19450964e-01 3.31832081e-01 -5.76524734e-01 1.49724975e-01
-3.23172271e-01 -3.71001437e-02 5.89879990e-01 1.03522468e+00
1.07576764e+00 -1.95607588e-01 -3.65358502e-01 7.73350179e-01
3.75677228e-01 6.69624209e-01 1.03838898e-01 -9.78743732e-01
7.25013852e-01 1.09225059e+00 9.97196659e-02 -1.15862548e+00
-2.90061533e-01 -1.98088214e-01 -5.09803236e-01 -2.89470315e-01
5.40926397e-01 1.76447883e-01 -8.95849407e-01 1.51298344e+00
5.68323135e-01 1.42072260e-01 3.44706655e-01 1.17847586e+00
1.68772185e+00 6.25555813e-01 2.01867461e-01 2.05016077e-01
1.73824549e+00 -1.21489906e+00 -7.78313577e-01 -2.18444213e-01
4.60677803e-01 -7.55186379e-01 1.54261708e+00 -1.00366570e-01
-9.58328485e-01 -6.02052152e-01 -5.77641666e-01 -6.76691532e-01
-7.30659068e-01 4.37579840e-01 7.60090768e-01 3.13068569e-01
-8.85959148e-01 -2.87743717e-01 -3.30906421e-01 -3.01187068e-01
7.09811509e-01 8.11904520e-02 -6.02513373e-01 -3.65567744e-01
-1.24232948e+00 9.39851940e-01 4.95894045e-01 2.52468884e-01
-8.35821569e-01 -8.53251398e-01 -1.00194085e+00 1.60089582e-02
7.28307903e-01 -1.34695518e+00 8.98642778e-01 -8.54799569e-01
-1.31259537e+00 9.82332170e-01 -3.89409810e-01 4.43396904e-02
2.40969598e-01 -2.67169446e-01 -2.98848510e-01 6.78319991e-01
4.98516075e-02 7.45722115e-01 6.60592079e-01 -1.89268494e+00
-3.90717715e-01 -5.23067892e-01 6.15449250e-01 4.16327715e-01
-2.39166513e-01 -3.30898285e-01 -9.43910956e-01 -2.62697160e-01
7.22433776e-02 -5.31638741e-01 1.81369409e-01 1.49624869e-01
-4.34849173e-01 -2.30531693e-01 8.23373675e-01 -9.01932776e-01
8.62849355e-01 -1.67525959e+00 4.11509931e-01 9.04042348e-02
3.29437882e-01 2.35265255e-01 -3.46270978e-01 5.36065280e-01
2.39996001e-01 -2.16867507e-01 -1.00112624e-01 -1.33373529e-01
2.24840954e-01 3.56271535e-01 -5.86308837e-01 6.05643615e-02
4.11047339e-01 1.60937083e+00 -9.36259091e-01 -8.04361463e-01
3.24882478e-01 7.42403090e-01 -3.50726396e-01 2.63231695e-01
-5.13502836e-01 5.29563546e-01 -6.44199550e-01 1.11193037e+00
6.71663225e-01 -7.02048659e-01 1.38105825e-01 -8.85120571e-01
2.68194973e-01 -1.40942097e-01 -7.66126752e-01 1.99498558e+00
-5.47487199e-01 5.71237206e-01 -2.55696893e-01 -1.05535793e+00
8.87875378e-01 7.76220560e-02 4.96695423e-03 -9.60614800e-01
1.28525272e-01 -1.20460957e-01 -5.58226168e-01 -1.03667915e+00
4.32059824e-01 -8.66271332e-02 8.21881741e-02 1.52025841e-05
3.28370601e-01 2.06744205e-02 -3.38739320e-03 5.31431317e-01
6.40595138e-01 4.28256691e-01 1.20430835e-01 3.46369624e-01
8.22503150e-01 2.72439867e-01 1.94839284e-01 4.57284361e-01
-1.49361879e-01 4.23916459e-01 5.09948015e-01 -3.22736382e-01
-6.07495129e-01 -1.37036788e+00 9.93843898e-02 1.15240121e+00
5.58089137e-01 -3.36089551e-01 -1.66368172e-01 -8.63745034e-01
1.52482793e-01 6.39628232e-01 -8.18952680e-01 1.96326189e-02
-2.81845659e-01 -2.24875554e-01 4.96290386e-01 6.39302015e-01
5.56075454e-01 -1.09967971e+00 -3.96319866e-01 -3.34905922e-01
-7.26673484e-01 -1.43474996e+00 -1.05978966e-01 -3.57870400e-01
-6.03014171e-01 -1.29656482e+00 -6.14776134e-01 -8.79538178e-01
6.80565834e-01 3.51113111e-01 1.40581667e+00 2.07785577e-01
-2.92147487e-01 1.15654910e+00 -5.55357516e-01 -1.30382687e-01
2.03120381e-01 -2.06712499e-01 -4.89161551e-01 2.73743182e-01
3.63086313e-01 -2.68407524e-01 -8.26015711e-01 1.93598002e-01
-7.43140697e-01 2.73541629e-01 6.60582900e-01 8.49526227e-01
8.58639896e-01 -5.00449121e-01 5.34444213e-01 -6.20206416e-01
2.83600479e-01 -4.88984793e-01 -3.65415394e-01 9.24732625e-01
-9.07184780e-02 2.07322642e-01 3.65674287e-01 -3.01426202e-01
-1.49614787e+00 1.57912761e-01 9.52742249e-02 -6.59529924e-01
-2.41818070e-01 6.37338758e-01 -3.91971201e-01 -1.97043136e-01
3.63238424e-01 2.97973275e-01 -2.06633046e-01 -3.28343749e-01
1.18012965e+00 1.86621279e-01 6.97115958e-01 -7.97826290e-01
7.74167359e-01 7.78762579e-01 -7.76027376e-03 -6.89855933e-01
-1.15310514e+00 -6.64465845e-01 -6.67046785e-01 -3.05508882e-01
1.12784719e+00 -1.03597689e+00 -1.18799889e+00 -6.82601854e-02
-1.33236432e+00 5.99282943e-02 -1.26700193e-01 1.81967333e-01
-6.96302056e-01 7.08660662e-01 -2.67987579e-01 -7.18393624e-01
-3.88959706e-01 -8.28867972e-01 1.10639977e+00 2.97730893e-01
3.02347124e-01 -1.12366414e+00 5.50922714e-02 1.24024093e+00
1.53429270e-01 3.23398083e-01 1.09304440e+00 -2.84042895e-01
-1.01384687e+00 1.02697685e-01 -8.07268381e-01 7.99905211e-02
-2.22121611e-01 -2.60009080e-01 -1.06172764e+00 1.26960844e-01
-5.80390155e-01 -9.03140962e-01 1.25385821e+00 -5.06665409e-02
1.21190977e+00 -2.22679213e-01 -2.48326585e-01 4.97170329e-01
1.58568621e+00 -2.02404588e-01 5.59413910e-01 2.19237491e-01
1.19114459e+00 1.01291597e+00 6.44735217e-01 2.92370886e-01
1.20904672e+00 6.44217849e-01 7.48339176e-01 -1.17877461e-01
-2.48250201e-01 -4.85062331e-01 -3.73518951e-02 7.26280510e-01
-1.59682184e-01 -1.27160624e-01 -1.11962640e+00 8.99704337e-01
-2.14431143e+00 -1.08857274e+00 -3.97726983e-01 1.63264465e+00
8.80355179e-01 -4.94964242e-01 -7.65563399e-02 -3.06659162e-01
2.95907885e-01 5.38387820e-02 -6.21494770e-01 -5.82528263e-02
-4.45750296e-01 -7.55594522e-02 8.68748501e-02 5.02201557e-01
-1.06759846e+00 1.19729400e+00 4.45403337e+00 8.60737503e-01
-6.49189532e-01 1.48129687e-01 2.31875613e-01 -1.63994951e-03
-7.71728516e-01 1.67062700e-01 -4.83399600e-01 -9.50855687e-02
3.86420935e-01 2.24786326e-01 3.61486465e-01 4.74168599e-01
-2.67131090e-01 -1.06434435e-01 -9.31190848e-01 1.33420372e+00
4.33517277e-01 -1.65389907e+00 5.24360836e-01 -4.15615022e-01
6.01621866e-01 -3.54740679e-01 1.49656802e-01 4.57649142e-01
-1.02436366e-02 -1.00729728e+00 6.40195310e-01 1.18391061e+00
6.76498115e-01 -6.59395516e-01 6.72728896e-01 8.46372917e-02
-1.52157307e+00 -8.86958763e-02 -3.84118676e-01 1.34965643e-01
2.77445734e-01 2.76771516e-01 -5.95223904e-01 1.12855923e+00
7.84510255e-01 6.76516712e-01 -9.09516156e-01 7.51061678e-01
-4.68958855e-01 3.51406604e-01 5.74701168e-02 7.06740394e-02
1.53577134e-01 4.29496728e-03 3.34433943e-01 9.45001543e-01
-3.55493347e-03 3.54598254e-01 -4.29261997e-02 1.04429698e+00
-2.48243026e-02 2.14846447e-01 -5.13506651e-01 -8.32379833e-02
2.81067550e-01 1.28584063e+00 -3.35639924e-01 -2.97186732e-01
-7.39030004e-01 1.00784647e+00 7.51174331e-01 7.82061636e-01
-8.84330392e-01 -2.63377160e-01 4.44231689e-01 -2.45392829e-01
6.60103321e-01 -9.02032945e-03 -3.74634534e-01 -1.46766102e+00
2.20214963e-01 -6.20833874e-01 6.43443227e-01 -1.22435260e+00
-1.67066264e+00 5.47448933e-01 -3.96116488e-02 -1.00978541e+00
1.03647098e-01 -9.55486834e-01 -1.91781655e-01 7.80608356e-01
-1.96785378e+00 -2.03741097e+00 -6.46542966e-01 1.07024932e+00
2.01415315e-01 -1.65915936e-01 7.59260118e-01 1.68333709e-01
-3.08184713e-01 5.20721674e-01 -2.23882988e-01 3.59829590e-02
6.76511586e-01 -1.20235968e+00 -2.84405231e-01 3.96625817e-01
4.47291732e-01 4.80261147e-01 4.80930567e-01 -4.05636579e-01
-1.78912961e+00 -8.46058071e-01 8.84835601e-01 -1.09247828e+00
7.60165215e-01 -2.32020512e-01 -1.08699036e+00 6.34770334e-01
4.23083276e-01 2.22666204e-01 9.64567244e-01 2.72331357e-01
-9.12836432e-01 -9.45292562e-02 -7.83024549e-01 5.57881057e-01
1.06219757e+00 -8.71367514e-01 -8.24720860e-01 2.15032369e-01
8.95906150e-01 -3.94342154e-01 -9.15215790e-01 4.92279470e-01
5.89427173e-01 -8.99465680e-01 1.44173622e+00 -9.57720160e-01
6.29134059e-01 -5.90048850e-01 -4.81999874e-01 -8.20830822e-01
-4.79815453e-02 4.76211905e-02 -4.57348078e-01 1.28181696e+00
4.72123802e-01 -2.80939221e-01 6.23668611e-01 3.97171944e-01
2.45359782e-02 -9.27258313e-01 -8.51965308e-01 -3.41428220e-01
-1.22775175e-01 -4.24896657e-01 4.26188111e-01 1.19337428e+00
-3.41384821e-02 6.68258607e-01 -4.37540740e-01 4.41865116e-01
5.49468875e-01 6.48151815e-01 9.43614960e-01 -1.03794301e+00
-1.65068030e-01 -2.81103224e-01 -3.44362319e-01 -1.08923066e+00
3.84802014e-01 -9.43441093e-01 -4.05685842e-01 -2.19439912e+00
4.68172789e-01 2.52659675e-02 -3.18326026e-01 6.67246163e-01
-3.99815381e-01 2.81044245e-01 3.97959262e-01 8.62621964e-05
-1.24450910e+00 1.06081450e+00 1.60574985e+00 -3.08776796e-01
3.92642617e-02 -6.74494028e-01 -7.13153780e-01 6.36866689e-01
6.56440973e-01 1.04169942e-01 -6.69367313e-01 -6.88127756e-01
6.02845490e-01 1.17099717e-01 1.10413921e+00 -4.72446024e-01
2.59439826e-01 -2.09559366e-01 4.04251575e-01 -8.62246335e-01
6.74001217e-01 -9.24686313e-01 -1.03043810e-01 -1.28111392e-01
-2.18996793e-01 -1.35129437e-01 4.74343419e-01 9.15103495e-01
-5.01119316e-01 1.74417198e-01 2.66414434e-01 -2.77813170e-02
-1.30804980e+00 2.50606686e-01 3.35975856e-01 3.06412607e-01
1.08025908e+00 -1.57514021e-01 -9.30849731e-01 -4.70842481e-01
-7.81718433e-01 5.96506536e-01 1.61060825e-01 4.82987314e-01
1.23956275e+00 -1.48813272e+00 -5.67157030e-01 -2.43303940e-01
7.88423121e-01 -2.25894183e-01 9.99789953e-01 9.11270499e-01
-4.17406559e-01 6.23112202e-01 -3.10916334e-01 -5.62384605e-01
-1.29443753e+00 8.38753760e-01 2.55432069e-01 -9.09850225e-02
-3.69658858e-01 1.06710804e+00 4.69800174e-01 -8.15470099e-01
1.49484202e-01 -9.85890776e-02 -4.61695015e-01 1.93716675e-01
2.77041167e-01 3.15122426e-01 -2.76793152e-01 -9.16863501e-01
-6.16853774e-01 8.96644413e-01 2.77102232e-01 5.21201231e-02
9.41351116e-01 -3.83739263e-01 -1.97200388e-01 4.12725657e-01
1.09029889e+00 -1.03442639e-01 -1.04022157e+00 -5.94547033e-01
-2.10047603e-01 -5.72556794e-01 -1.66963458e-01 -1.20767176e+00
-1.04553032e+00 1.15661108e+00 4.67572570e-01 -1.77707955e-01
1.24211085e+00 6.52402282e-01 6.54471338e-01 5.90967238e-01
7.92082474e-02 -7.05650449e-01 6.15291417e-01 5.21595001e-01
1.24340022e+00 -1.60721624e+00 -1.36785090e-01 -6.41976953e-01
-1.22448945e+00 9.77663815e-01 9.06305552e-01 3.42464060e-01
3.67762566e-01 -6.32346809e-01 3.17348808e-01 -6.76376104e-01
-8.51104140e-01 -7.87469089e-01 1.01341867e+00 7.19667137e-01
1.84593171e-01 1.00490581e-02 1.67709291e-02 8.40934813e-01
6.09028898e-02 -1.77192926e-01 -6.89992160e-02 5.93788862e-01
-3.31772447e-01 -7.99872756e-01 -2.93002039e-01 1.02222167e-01
3.49003100e-03 -2.14881465e-01 -6.45605624e-01 8.07166815e-01
2.66029269e-01 1.06108415e+00 -2.21565008e-01 -2.25262135e-01
5.75634599e-01 8.89570788e-02 6.56435549e-01 -2.56868780e-01
-1.23329870e-01 -4.07138556e-01 2.59333611e-01 -6.12378776e-01
-7.05192804e-01 -3.14877510e-01 -1.44355059e+00 -9.97576714e-02
-7.21569732e-02 -1.56958848e-01 3.01347971e-01 9.07059908e-01
5.49409091e-01 5.20446837e-01 9.15037282e-03 -4.33199525e-01
-1.07706450e-02 -5.11598051e-01 -6.83972836e-02 5.47717631e-01
4.10872996e-01 -9.06830072e-01 -7.50630572e-02 2.39182934e-01] | [10.672405242919922, 1.7253015041351318] |
48dc6089-2d21-47c8-a558-dcba837b6f91 | overview-of-the-arabic-sentiment-analysis | 2109.14456 | null | https://arxiv.org/abs/2109.14456v1 | https://arxiv.org/pdf/2109.14456v1.pdf | Overview of the Arabic Sentiment Analysis 2021 Competition at KAUST | This paper provides an overview of the Arabic Sentiment Analysis Challenge organized by King Abdullah University of Science and Technology (KAUST). The task in this challenge is to develop machine learning models to classify a given tweet into one of the three categories Positive, Negative, or Neutral. From our recently released ASAD dataset, we provide the competitors with 55K tweets for training, and 20K tweets for validation, based on which the performance of participating teams are ranked on a leaderboard, https://www.kaggle.com/c/arabic-sentiment-analysis-2021-kaust. The competition received in total 1247 submissions from 74 teams (99 team members). The final winners are determined by another private set of 20K tweets that have the same distribution as the training and validation set. In this paper, we present the main findings in the competition and summarize the methods and tools used by the top ranked teams. The full dataset of 100K labeled tweets is also released for public usage, at https://www.kaggle.com/c/arabic-sentiment-analysis-2021-kaust/data. | ['Xiangliang Zhang', 'Inji Ibrahim Jaber', 'Manal Kalkatawi', 'Zuhair Khayyat', 'Basma Alharbi', 'Manal Alshehri', 'Hind Alamro'] | 2021-09-29 | null | null | null | null | ['arabic-sentiment-analysis'] | ['natural-language-processing'] | [-6.20560050e-01 -1.60401419e-01 -2.33306259e-01 -5.14790177e-01
-9.39397871e-01 -1.05104172e+00 7.24132180e-01 7.98283279e-01
-5.62288046e-01 5.19291699e-01 2.02364385e-01 -2.49765128e-01
2.88603961e-01 -8.10254216e-01 -3.66188705e-01 -4.99117613e-01
-8.91726166e-02 6.31688356e-01 -6.14622980e-02 -1.07664025e+00
6.27825797e-01 8.71740207e-02 -1.38263893e+00 1.01599240e+00
7.31644750e-01 1.26683962e+00 -5.01926363e-01 8.47948551e-01
-1.29328609e-01 8.54021192e-01 -5.78085780e-01 -8.99493814e-01
1.42519936e-01 -3.54616523e-01 -1.08798003e+00 -3.01407307e-01
3.65824089e-03 2.88605928e-01 3.25874120e-01 8.32435489e-01
4.68888819e-01 -2.43279021e-02 6.13322914e-01 -1.53240657e+00
-6.72766805e-01 8.92812192e-01 -5.95543146e-01 -9.39550065e-03
5.42121649e-01 -4.45880711e-01 1.17571414e+00 -1.43081641e+00
4.65932101e-01 8.15973401e-01 5.15331805e-01 3.88704747e-01
-3.47456485e-01 -8.59633267e-01 2.41654217e-01 2.48575121e-01
-1.23076427e+00 -4.24488991e-01 6.26908183e-01 -5.85843027e-01
6.66735291e-01 4.97955948e-01 6.70678020e-01 6.83051467e-01
2.13995129e-01 9.78358924e-01 1.52930045e+00 -7.24873126e-01
3.50387752e-01 6.10596716e-01 6.06671393e-01 6.08829200e-01
8.80404264e-02 -6.95923865e-01 -7.89006114e-01 -3.43544155e-01
-4.42083448e-01 -3.90448958e-01 -1.88507792e-02 2.15901271e-01
-1.11953747e+00 1.32054114e+00 4.39263761e-01 2.09197193e-01
-2.88305908e-01 -5.89858353e-01 7.31392682e-01 7.34046578e-01
9.73352849e-01 6.51187778e-01 -7.67500103e-01 -8.34984705e-02
-6.11974478e-01 4.12832975e-01 8.84978712e-01 6.42555416e-01
7.76912272e-01 -1.97964236e-01 1.96868390e-01 9.07971621e-01
3.58596355e-01 7.77063549e-01 8.87433827e-01 -5.96582353e-01
3.88533711e-01 7.97593117e-01 5.15603244e-01 -1.30358624e+00
-4.63793963e-01 -1.04272947e-01 -3.22603226e-01 6.09799707e-03
3.07588458e-01 -6.97352052e-01 -7.17391074e-01 9.57679212e-01
2.46067956e-01 -5.70462763e-01 4.14391905e-01 8.08017015e-01
1.40056539e+00 8.32593083e-01 -2.03821301e-01 -1.02981791e-01
1.52586472e+00 -1.17452812e+00 -7.75590062e-01 -3.81386787e-01
9.88766789e-01 -1.32127070e+00 1.13208246e+00 9.00917470e-01
-9.59479868e-01 -4.30032723e-02 -1.17588258e+00 3.95743579e-01
-9.75545704e-01 2.37329245e-01 5.72236300e-01 9.66049492e-01
-1.00362825e+00 -1.08761145e-02 -5.16179323e-01 -3.86295199e-01
1.60044089e-01 4.10291523e-01 -3.06391180e-01 1.66447535e-01
-1.62156785e+00 9.82696891e-01 1.53186098e-01 1.55388126e-02
-5.34832060e-01 -2.62126237e-01 -7.80401886e-01 -7.66449690e-01
-8.54901597e-02 3.83783042e-01 1.51091659e+00 -1.28234971e+00
-1.50460339e+00 1.31660008e+00 -1.60484210e-01 -2.78889477e-01
3.74095261e-01 -2.18800083e-01 -7.63856411e-01 -8.39736015e-02
3.94448578e-01 3.23852181e-01 3.82250875e-01 -1.38401759e+00
-1.04328513e+00 -2.37965584e-01 1.61250085e-01 4.79622521e-02
-6.62048876e-01 6.31809890e-01 -4.23169941e-01 -5.62704921e-01
-1.09743603e-01 -1.31473565e+00 2.79953517e-02 -1.17726910e+00
-3.84190738e-01 -3.32810462e-01 5.91327131e-01 -5.66549122e-01
1.36147952e+00 -1.95534134e+00 -2.50203133e-01 5.39488077e-01
-4.15263698e-02 1.18301017e-02 -1.58096001e-01 8.91325116e-01
-2.39345074e-01 2.33703077e-01 3.08322664e-02 -3.40232432e-01
8.78597274e-02 -2.62685150e-01 -5.29161155e-01 5.88569462e-01
1.83582783e-01 5.75116873e-01 -1.10352492e+00 -1.81429207e-01
-1.97666883e-01 8.73515084e-02 -1.04166485e-01 -1.44375473e-01
-3.98006104e-02 1.10606812e-01 -4.67939675e-01 8.93034875e-01
7.84861684e-01 -1.87897846e-01 3.08570087e-01 -1.48913488e-01
-3.39098305e-01 2.86248356e-01 -8.86359572e-01 9.57888424e-01
-3.53278518e-01 6.02733791e-01 1.06754601e-01 -5.26777327e-01
1.13761473e+00 2.23548919e-01 3.23466748e-01 -6.14079475e-01
5.99547327e-01 7.96059966e-01 -2.01322913e-01 8.50157365e-02
9.06324983e-01 -7.20676780e-02 -7.13119209e-01 9.31841373e-01
-3.45151901e-01 -3.46912652e-01 7.68822491e-01 5.18461108e-01
6.01360500e-01 -3.71436715e-01 1.11809127e-01 -5.07649720e-01
8.15214336e-01 4.85222220e-01 1.11825295e-01 2.66087413e-01
-3.58604580e-01 5.10490477e-01 7.67895162e-01 -5.44451356e-01
-6.48597896e-01 -3.21881801e-01 -1.21654354e-01 1.71041191e+00
-1.35578901e-01 -9.78144050e-01 -7.74831116e-01 -8.71832430e-01
-2.81599343e-01 4.96265352e-01 -1.16166317e+00 1.85350761e-01
-2.34783724e-01 -1.16016543e+00 5.18581867e-01 -8.09099376e-02
3.11854213e-01 -1.02580869e+00 -2.03547269e-01 -2.08817959e-01
-6.64642155e-01 -8.63888264e-01 -1.10215895e-01 3.14291269e-01
-1.15451612e-01 -1.04057372e+00 -5.48356831e-01 -9.77550685e-01
5.60346305e-01 1.53713435e-01 1.13699353e+00 1.21866666e-01
5.44322073e-01 1.69586182e-01 -1.38056457e+00 -1.25823593e+00
-5.98448932e-01 3.45086694e-01 9.96052399e-02 1.59208700e-01
7.03463197e-01 4.73657042e-01 -3.54853928e-01 5.13692558e-01
-6.80143774e-01 -2.98420280e-01 8.89879279e-03 5.12896121e-01
1.89333588e-01 -4.82023805e-02 7.76756406e-01 -1.00916231e+00
8.84183645e-01 -6.95624053e-01 -3.00087273e-01 1.15911350e-01
-5.90448380e-01 -6.29103422e-01 5.88861406e-01 1.31186947e-01
-6.34019375e-01 -9.13547177e-04 -1.66272417e-01 5.56570530e-01
2.06386209e-01 1.19966114e+00 2.94544101e-01 -4.79706973e-02
8.62706482e-01 -1.08811319e-01 -1.42188976e-02 -8.23449790e-02
1.87522724e-01 1.29177892e+00 -4.06938083e-02 -4.28012729e-01
5.40912211e-01 5.33326387e-01 -6.45822287e-01 -6.88148916e-01
-1.29725981e+00 -5.51784396e-01 -4.76985216e-01 -4.67685848e-01
4.29682523e-01 -1.25415409e+00 -3.77606571e-01 9.47840273e-01
-7.35473335e-01 -5.07904232e-01 9.98736620e-02 3.26310992e-01
2.79481076e-02 1.57237798e-01 -5.83591878e-01 -5.36853731e-01
-6.22902870e-01 -1.16288006e+00 7.15461850e-01 4.93718497e-02
-5.76356232e-01 -9.84301388e-01 3.02945197e-01 6.39250517e-01
2.50966102e-01 2.58939087e-01 4.30227965e-01 -1.14155972e+00
7.28462696e-01 -8.38917673e-01 2.11736754e-01 5.51318049e-01
8.80431086e-02 6.18338346e-01 -1.04821968e+00 -3.96371812e-01
-3.20557892e-01 -9.56612825e-01 5.80766559e-01 1.45631269e-01
4.79849845e-01 -6.64399415e-02 4.25829552e-02 -3.80091399e-01
9.46165740e-01 -4.47624400e-02 2.17652112e-01 1.13851869e+00
1.59269065e-01 9.44368899e-01 1.37690032e+00 7.53740132e-01
7.29141593e-01 2.04248145e-01 5.15315115e-01 8.45670849e-02
5.66384554e-01 2.07575306e-01 8.21002722e-01 1.33422983e+00
6.79305708e-03 -2.85257369e-01 -1.38391793e+00 6.54441535e-01
-1.74530721e+00 -6.44210219e-01 -5.63738108e-01 1.84693861e+00
9.68911767e-01 1.98104680e-01 5.09855747e-01 4.95733827e-01
5.83803296e-01 1.53834388e-01 1.46820307e-01 -9.35125887e-01
-4.86763805e-01 2.78976828e-01 4.13918972e-01 8.11877429e-01
-1.51893592e+00 1.09394455e+00 5.68767357e+00 8.46676588e-01
-1.37238216e+00 1.69644445e-01 8.33298922e-01 -1.61482468e-01
-2.01330498e-01 -2.13218987e-01 -8.19246948e-01 3.98173690e-01
1.36562026e+00 -3.31400841e-01 -1.56981438e-01 7.05407202e-01
2.62378782e-01 -1.07087508e-01 -3.01117837e-01 5.23857892e-01
3.19346517e-01 -1.27003682e+00 4.30483185e-02 -2.11112618e-01
1.08261228e+00 6.31104589e-01 4.45914239e-01 4.76450384e-01
6.03798091e-01 -1.15049255e+00 1.11831725e+00 -9.61062089e-02
6.00551426e-01 -1.04839909e+00 1.39585829e+00 1.39651284e-01
-8.88703227e-01 -1.54011041e-01 -1.14598855e-01 -3.80810022e-01
-2.61100501e-01 6.38517499e-01 -6.78565502e-01 6.82604432e-01
1.15927589e+00 1.02454293e+00 -8.47253621e-01 5.01011372e-01
-1.82974428e-01 8.60444188e-01 -1.51062682e-01 -5.52529573e-01
7.89768696e-01 -2.31749162e-01 8.82925391e-02 1.56344819e+00
1.95409670e-01 2.48754723e-03 2.69321650e-01 -4.66013819e-01
-1.22195400e-01 8.10271144e-01 -9.84025374e-02 -6.15112185e-02
5.31483114e-01 1.55537736e+00 -1.02012503e+00 -3.23223740e-01
-3.04213047e-01 4.56396729e-01 -1.66781433e-02 -2.44518220e-02
-5.80602288e-01 -6.34462357e-01 2.07331255e-01 -1.06050946e-01
5.06535135e-02 5.84507994e-02 -4.09532130e-01 -1.09912515e+00
-2.47563839e-01 -1.35681665e+00 7.06822336e-01 -7.25156248e-01
-1.48135853e+00 9.72476363e-01 -3.69918704e-01 -1.33565199e+00
9.49478596e-02 -9.70983863e-01 -4.23604429e-01 8.46132517e-01
-1.66978800e+00 -1.41099989e+00 -4.14671212e-01 5.46234787e-01
3.32579672e-01 -5.35093963e-01 9.89245296e-01 2.97073692e-01
-4.11010116e-01 5.88710427e-01 1.94921061e-01 4.65134442e-01
1.38200319e+00 -1.35992086e+00 1.85169026e-01 3.07780176e-01
-1.98071435e-01 5.38090646e-01 8.26808810e-01 -5.59868157e-01
-9.52426910e-01 -1.00262129e+00 1.36665046e+00 -8.98705244e-01
1.20592570e+00 -2.38435656e-01 -4.44337577e-01 6.75940156e-01
5.98203361e-01 -3.81855488e-01 1.35738361e+00 1.90137386e-01
-3.05584371e-01 -3.19537111e-02 -7.51504421e-01 4.25009370e-01
-3.50566566e-01 -2.56253898e-01 -4.11207646e-01 6.91480100e-01
1.45202577e-01 -4.41409469e-01 -8.15487683e-01 1.08866304e-01
6.36139095e-01 -8.17607284e-01 3.62940162e-01 -6.44893467e-01
8.19492638e-01 -3.39921653e-01 -4.57345307e-01 -1.49407732e+00
-4.23972681e-03 -3.72714192e-01 5.21198630e-01 9.50563848e-01
1.05166018e+00 -5.75145543e-01 8.01013231e-01 8.96177217e-02
-1.95103362e-01 -8.25082719e-01 -4.23383772e-01 -1.66129261e-01
6.92496955e-01 -7.13072479e-01 4.54709798e-01 1.38999248e+00
4.79312837e-01 4.23937529e-01 -1.74573824e-01 -7.02247471e-02
8.77705403e-03 1.75397456e-01 8.86481881e-01 -1.04953432e+00
4.54870433e-01 -5.23682833e-01 -8.63510817e-02 -4.45645779e-01
5.96040525e-02 -9.95109499e-01 -1.49243921e-01 -1.54765069e+00
-3.75555940e-02 -4.00008947e-01 -2.60357380e-01 7.66100824e-01
5.12176044e-02 1.04307497e+00 8.19265470e-02 3.52552652e-01
-8.28636885e-01 1.69394016e-01 7.74795949e-01 -2.06145152e-01
-9.81403515e-02 3.79951522e-02 -1.06733739e+00 8.05239677e-01
1.10689867e+00 -3.21433038e-01 3.12100910e-02 -2.02210233e-01
1.14992368e+00 -7.81126618e-01 -2.25851655e-01 -5.02253175e-01
2.02109981e-02 -1.93698287e-01 1.05146609e-01 -6.98658705e-01
2.47394562e-01 -3.63868266e-01 -3.98282319e-01 4.59369361e-01
-2.94029057e-01 4.28581834e-01 2.53061205e-01 -1.65819019e-01
-6.65614009e-01 -2.91102529e-01 7.63046861e-01 -5.25453724e-02
-6.94527984e-01 5.26533648e-02 -8.19662094e-01 1.03647657e-01
1.06737089e+00 -3.01386416e-02 -5.26146710e-01 -6.91833138e-01
-8.75096381e-01 4.24265206e-01 4.25864249e-01 5.22994757e-01
4.21003848e-01 -1.16851389e+00 -1.33277774e+00 -1.24819838e-01
6.75788522e-01 -5.21501660e-01 3.74257453e-02 9.93355453e-01
-1.07135904e+00 3.72615397e-01 -2.58807808e-01 -1.43152699e-01
-1.37548828e+00 -1.52204320e-01 4.55506817e-02 -2.79583335e-01
2.17786297e-01 1.00204217e+00 -4.29987371e-01 -1.01351500e+00
-3.74806046e-01 4.47663784e-01 -8.82239044e-01 8.58958662e-01
7.04949796e-01 2.57762790e-01 8.15835416e-01 -1.17615139e+00
-6.23818755e-01 3.51249188e-01 -3.27879846e-01 -2.75324464e-01
1.36288691e+00 -4.77171466e-02 -7.19278395e-01 7.32881963e-01
1.04995191e+00 6.95282757e-01 1.78000908e-02 -3.17395851e-02
1.15173258e-01 -2.60349900e-01 -5.93644716e-02 -1.16156292e+00
-9.14565623e-01 5.55239558e-01 2.92804748e-01 7.33496785e-01
8.50122035e-01 -1.12977557e-01 7.22402036e-01 3.45233798e-01
7.53795430e-02 -1.41164744e+00 1.39278486e-01 1.22108591e+00
1.09765673e+00 -1.45450091e+00 7.17459470e-02 -3.67368609e-01
-1.30751896e+00 1.18633068e+00 3.98029953e-01 -2.27556840e-01
1.02376378e+00 -1.33792395e-02 1.03492808e+00 -3.45516443e-01
-6.73826456e-01 1.21987842e-01 1.50985658e-01 5.24332941e-01
1.00502932e+00 3.37764591e-01 -5.76809168e-01 9.95629072e-01
-1.00377464e+00 -4.91055787e-01 1.21309030e+00 1.29515803e+00
-4.24806505e-01 -1.16297805e+00 -5.31382322e-01 5.55594742e-01
-7.08942711e-01 -5.90410419e-02 -9.65217352e-01 7.60554850e-01
-2.36233339e-01 1.60100782e+00 -1.43640205e-01 -6.90019906e-01
5.13699174e-01 -7.14171007e-02 -1.87970057e-01 -6.64237916e-01
-1.18481004e+00 8.97169858e-02 5.62001109e-01 -1.32381827e-01
-6.51882648e-01 -7.92428493e-01 -1.29161394e+00 -7.52544463e-01
-3.26894224e-01 1.19346488e+00 8.04034591e-01 6.45434022e-01
1.92880899e-01 -1.25933573e-01 1.01346302e+00 -8.09180081e-01
-2.24427417e-01 -1.05697060e+00 -7.49611378e-01 1.68208137e-01
4.07133043e-01 -3.09083909e-01 -5.12097836e-01 6.64630607e-02] | [11.17921257019043, 6.91127872467041] |
1f1574cc-8c60-44b6-ae2d-ed260fa370f9 | towards-counterfactual-image-manipulation-via | 2207.02812 | null | https://arxiv.org/abs/2207.02812v3 | https://arxiv.org/pdf/2207.02812v3.pdf | Towards Counterfactual Image Manipulation via CLIP | Leveraging StyleGAN's expressivity and its disentangled latent codes, existing methods can achieve realistic editing of different visual attributes such as age and gender of facial images. An intriguing yet challenging problem arises: Can generative models achieve counterfactual editing against their learnt priors? Due to the lack of counterfactual samples in natural datasets, we investigate this problem in a text-driven manner with Contrastive-Language-Image-Pretraining (CLIP), which can offer rich semantic knowledge even for various counterfactual concepts. Different from in-domain manipulation, counterfactual manipulation requires more comprehensive exploitation of semantic knowledge encapsulated in CLIP as well as more delicate handling of editing directions for avoiding being stuck in local minimum or undesired editing. To this end, we design a novel contrastive loss that exploits predefined CLIP-space directions to guide the editing toward desired directions from different perspectives. In addition, we design a simple yet effective scheme that explicitly maps CLIP embeddings (of target text) to the latent space and fuses them with latent codes for effective latent code optimization and accurate editing. Extensive experiments show that our design achieves accurate and realistic editing while driving by target texts with various counterfactual concepts. | ['Chunyan Miao', 'Xian-Sheng Hua', 'Xuansong Xie', 'Miaomiao Cui', 'Shijian Lu', 'Jiahui Zhang', 'Rongliang Wu', 'Fangneng Zhan', 'Yingchen Yu'] | 2022-07-06 | null | null | null | null | ['image-manipulation'] | ['computer-vision'] | [ 4.46359247e-01 2.06016466e-01 -1.89094886e-01 -6.64377034e-01
-5.74596822e-01 -7.14133203e-01 9.66957450e-01 -7.08932042e-01
-1.93163127e-01 7.71177530e-01 5.86369693e-01 -5.44338068e-03
-1.36939272e-01 -6.74384832e-01 -9.23620880e-01 -7.04917312e-01
7.23590255e-02 1.02547176e-01 -5.53569496e-01 -1.19156398e-01
5.50659522e-02 1.96171209e-01 -1.47832775e+00 1.98031530e-01
1.08447611e+00 7.35255778e-01 -1.13694686e-02 3.82541418e-01
-4.04480547e-02 5.24103284e-01 -5.74717700e-01 -1.24376976e+00
6.22837543e-01 -3.42860788e-01 -1.58443391e-01 3.35366338e-01
7.18767643e-01 -3.11488807e-01 -2.17045233e-01 1.28582084e+00
4.78222996e-01 -6.88257581e-03 8.43031228e-01 -1.52149570e+00
-1.26789951e+00 6.68096066e-01 -8.39930654e-01 -3.18797678e-01
1.44946590e-01 5.84705114e-01 8.77423406e-01 -9.58554745e-01
6.71538413e-01 1.58478403e+00 4.59046900e-01 7.75375366e-01
-1.46026385e+00 -1.23342240e+00 4.18153346e-01 -6.88513592e-02
-1.33758116e+00 -5.75481772e-01 1.29809868e+00 -4.05266047e-01
1.58534288e-01 4.31858867e-01 6.82543218e-01 1.65059114e+00
6.37842342e-02 7.60072052e-01 1.32594573e+00 -1.42762989e-01
1.40141934e-01 3.25269222e-01 -7.51045406e-01 7.79902875e-01
1.90873578e-01 3.43371034e-01 -5.34390688e-01 -6.03705086e-02
7.14808941e-01 6.02227338e-02 -4.54427302e-01 -8.12130153e-01
-1.34625554e+00 1.02834797e+00 2.38775566e-01 -1.35185421e-01
-2.33757958e-01 3.20926040e-01 2.67611444e-01 2.27731645e-01
5.84050417e-01 6.69543505e-01 -2.90580094e-01 1.19661108e-01
-1.04358411e+00 3.72182786e-01 4.87267077e-01 1.01919162e+00
7.24113524e-01 3.82284880e-01 -5.52791178e-01 7.50474274e-01
1.32019639e-01 6.23264611e-01 3.23529005e-01 -9.88503754e-01
6.56064153e-01 4.26445931e-01 8.52735862e-02 -1.25285196e+00
2.04715803e-01 -5.66667914e-01 -9.37664866e-01 3.47434759e-01
4.26018268e-01 -3.04638475e-01 -9.03716564e-01 2.22452307e+00
3.24579448e-01 2.39280328e-01 -1.55383483e-01 1.02353907e+00
2.08711192e-01 4.44151282e-01 1.39177963e-01 -3.76746267e-01
1.36222589e+00 -6.63426757e-01 -8.85716498e-01 -2.48664901e-01
2.96468824e-01 -7.48077869e-01 1.39975595e+00 1.55524388e-01
-9.06973541e-01 -3.37996602e-01 -1.01700473e+00 -1.59384519e-01
-2.21182838e-01 3.37346196e-01 1.01824856e+00 8.40338469e-01
-6.97965026e-01 2.85858899e-01 -3.79942715e-01 2.02511445e-01
8.20743859e-01 2.34551564e-01 -3.22115958e-01 -1.33777559e-01
-1.30046082e+00 6.61969543e-01 1.87832683e-01 9.62646827e-02
-1.08299685e+00 -1.18124616e+00 -1.03832090e+00 -5.13392948e-02
6.78469598e-01 -8.95558238e-01 8.62820387e-01 -1.53141487e+00
-1.79869401e+00 1.00463128e+00 1.01263218e-01 -3.49422872e-01
1.05510855e+00 -1.45016879e-01 -3.53831172e-01 -1.17486320e-01
5.87181970e-02 9.10782695e-01 1.64420199e+00 -1.45812893e+00
4.35901619e-02 -5.07461727e-01 3.48698199e-01 4.36480761e-01
-5.93629956e-01 -1.73890784e-01 -3.85328561e-01 -1.12593126e+00
-3.99228185e-01 -9.08746362e-01 -9.11292881e-02 6.13640070e-01
-4.98361170e-01 1.58811450e-01 7.82504976e-01 -5.09092748e-01
1.10366726e+00 -2.25156736e+00 1.99249402e-01 -5.17330877e-02
4.62914616e-01 3.70067544e-02 -1.88689694e-01 -1.96456816e-03
-2.58985132e-01 9.25190225e-02 -3.29870254e-01 -4.36758131e-01
2.93684483e-01 1.03964917e-01 -7.58215606e-01 6.00998163e-01
2.70505697e-01 1.03898561e+00 -9.63978112e-01 -5.11666119e-01
2.05290899e-01 5.63671529e-01 -9.94353235e-01 2.54532527e-02
-3.27598572e-01 3.63316983e-01 -3.05732757e-01 4.11471844e-01
9.75660384e-01 3.35124247e-02 1.82685003e-01 -3.89852941e-01
1.50145255e-02 -2.65904099e-01 -1.01098228e+00 1.80036461e+00
-6.97706521e-01 6.24120653e-01 1.23188764e-01 -1.00542724e+00
9.48965132e-01 9.42958444e-02 9.42831635e-02 -4.30234522e-01
1.87575102e-01 -7.05349222e-02 -2.25281432e-01 -4.85943317e-01
3.74377042e-01 -5.52798808e-01 -2.91115105e-01 3.59061033e-01
5.74791394e-02 -2.73307979e-01 -2.00521410e-01 1.97325632e-01
3.59925807e-01 3.22899759e-01 1.99873298e-01 -2.72327125e-01
2.12367162e-01 -3.67655039e-01 7.55938411e-01 5.45704961e-01
-1.85928613e-01 6.31103873e-01 7.05304205e-01 -2.98159987e-01
-1.08336282e+00 -1.33856153e+00 4.92843613e-02 1.02542281e+00
-5.26267253e-02 -1.03324190e-01 -7.58242011e-01 -8.19666147e-01
1.20284237e-01 8.94455194e-01 -1.03060162e+00 -5.17393351e-01
-3.85090560e-01 -7.12307513e-01 4.11747366e-01 2.82951117e-01
6.45416319e-01 -6.20755434e-01 -2.99646586e-01 -2.29331881e-01
-1.23344347e-01 -9.09653425e-01 -9.79617238e-01 -4.87712592e-01
-5.00033975e-01 -8.36689830e-01 -9.39749062e-01 -2.82041639e-01
8.69364619e-01 9.49842632e-02 9.58648205e-01 -4.54171151e-01
-2.57197887e-01 3.02221775e-01 -2.56912541e-02 -6.04386508e-01
-2.78952628e-01 -4.27561402e-01 8.36059153e-02 5.68960607e-01
1.92994714e-01 -7.39922047e-01 -8.04940939e-01 1.17378503e-01
-9.73216653e-01 5.13608396e-01 5.72981894e-01 9.05474722e-01
1.73911706e-01 -1.61304161e-01 3.67542326e-01 -9.07701969e-01
7.13595152e-01 -4.41632926e-01 -7.50005007e-01 3.16976607e-01
-3.15989256e-01 2.39412025e-01 7.95736134e-01 -8.12780976e-01
-1.43959773e+00 6.87407032e-02 2.98950553e-01 -9.42501187e-01
1.53449669e-01 3.60753350e-02 -4.89020139e-01 2.27974355e-01
6.97055697e-01 3.32389235e-01 2.03407139e-01 -9.72535312e-02
1.00000048e+00 4.70818102e-01 5.83478451e-01 -7.21760571e-01
1.17012525e+00 9.90460277e-01 -1.78463042e-01 -4.47580606e-01
-9.78309155e-01 1.96219146e-01 -3.99389565e-01 -1.97279036e-01
7.46187270e-01 -1.07461894e+00 -7.78070807e-01 3.34870785e-01
-1.06377161e+00 -9.47993696e-02 -3.81870896e-01 3.63045543e-01
-6.94854379e-01 2.48062685e-01 4.26625125e-02 -7.82550395e-01
-4.01440822e-02 -1.00147259e+00 1.19245732e+00 -2.79965326e-02
-1.29791409e-01 -9.95290041e-01 -1.31871581e-01 3.82220656e-01
3.63135725e-01 5.67877710e-01 7.26961851e-01 4.44278456e-02
-6.23977423e-01 8.51539597e-02 -3.07166636e-01 3.21735859e-01
2.04766557e-01 1.12617344e-01 -9.98372555e-01 -2.76622534e-01
4.04454172e-02 -2.19005898e-01 9.63427365e-01 4.05761003e-01
1.45788717e+00 -9.14689541e-01 -2.32104436e-01 1.00029600e+00
1.35786521e+00 -1.81324571e-01 5.65545261e-01 -2.14733660e-01
7.75362253e-01 7.57333398e-01 3.99103850e-01 5.74483573e-01
3.37057650e-01 7.89669871e-01 4.54066932e-01 -7.20752925e-02
3.16284560e-02 -6.42311811e-01 5.58663011e-01 2.27967769e-01
-4.07459140e-02 8.09789896e-02 -4.25691694e-01 7.04233527e-01
-1.63629150e+00 -1.27205074e+00 4.63515937e-01 2.07423210e+00
1.28821909e+00 6.45319223e-02 -1.10683538e-01 -1.78942308e-01
8.46749306e-01 5.46241522e-01 -6.17495656e-01 -4.13495660e-01
-5.03213592e-02 -8.63816440e-02 4.85520840e-01 3.94488901e-01
-1.03999197e+00 9.34264123e-01 5.57448292e+00 1.30107141e+00
-1.33018732e+00 2.51352310e-01 6.33976161e-01 -4.35009688e-01
-9.88349676e-01 -3.20974626e-02 -5.58947682e-01 6.98551178e-01
3.80746871e-01 -4.12613273e-01 6.09254181e-01 8.64030957e-01
3.59263927e-01 2.70654798e-01 -1.01782870e+00 1.14925635e+00
1.67356789e-01 -1.66589379e+00 5.17522097e-01 -8.16104189e-03
1.01890683e+00 -6.66421413e-01 6.79264665e-01 3.68767887e-01
5.59514046e-01 -9.83165026e-01 1.09414971e+00 6.34981275e-01
1.36634302e+00 -8.25483739e-01 5.79669364e-02 1.48217455e-01
-7.37923384e-01 -2.41054133e-01 -3.89795661e-01 -1.45915046e-01
6.99369644e-04 8.06999326e-01 -5.37382185e-01 3.08457106e-01
1.73721120e-01 5.69082558e-01 -2.66661137e-01 3.55201900e-01
-4.82540011e-01 2.99767405e-01 -8.58728662e-02 1.04744628e-01
7.94237033e-02 -1.34889349e-01 6.87782407e-01 1.02856553e+00
3.21908176e-01 -9.25119594e-02 -1.47179723e-01 1.47955048e+00
-2.83558011e-01 -5.36225438e-02 -1.04624951e+00 -1.40760332e-01
5.66018701e-01 1.12577236e+00 -3.14930677e-01 -2.19904765e-01
-3.11690271e-01 1.18207598e+00 3.10648412e-01 6.51438236e-01
-1.23112571e+00 -2.44111687e-01 1.07333469e+00 1.72442824e-01
1.91236347e-01 -2.19008774e-01 -3.35054755e-01 -1.54107451e+00
8.11998472e-02 -8.37508261e-01 1.43507600e-01 -7.29898334e-01
-1.33968937e+00 2.29655713e-01 2.00784743e-01 -1.27156496e+00
-1.26489867e-02 -4.78371203e-01 -6.92005336e-01 6.12640917e-01
-1.53006816e+00 -1.52275968e+00 -1.61664158e-01 8.76660466e-01
4.49312568e-01 -1.87466726e-01 5.19799173e-01 9.16304216e-02
-4.88922805e-01 1.04452217e+00 -5.14369924e-03 1.48877472e-01
8.65551710e-01 -1.13831186e+00 7.09081292e-02 7.85336316e-01
1.92324504e-01 8.61535490e-01 8.83086562e-01 -3.52405965e-01
-1.18352044e+00 -1.15404820e+00 5.00087440e-01 -4.86580133e-01
5.91718316e-01 -8.35912287e-01 -3.56869996e-01 7.64013171e-01
2.22900704e-01 1.56526025e-02 6.29226327e-01 -9.12461523e-03
-6.88440561e-01 -2.93351769e-01 -1.06044054e+00 1.15163338e+00
1.34233642e+00 -6.89520001e-01 -5.44896245e-01 2.31280938e-01
8.48758101e-01 -3.10390085e-01 -3.30823928e-01 2.20101789e-01
6.00326359e-01 -8.60734522e-01 9.75735009e-01 -5.98665476e-01
8.11941922e-01 -2.03772411e-01 -1.50939990e-02 -1.51260686e+00
-1.24712802e-01 -1.07156539e+00 8.58573541e-02 1.33002436e+00
9.84586030e-02 -4.78006005e-01 7.54087746e-01 6.08201802e-01
6.19395152e-02 -7.34604418e-01 -8.86595368e-01 -6.77136540e-01
1.05188951e-01 -6.39372766e-01 8.69513750e-01 1.29481661e+00
-2.44063646e-01 -4.46390323e-02 -9.27954137e-01 4.57131565e-02
8.84567142e-01 3.12121600e-01 9.79490697e-01 -7.58683801e-01
-1.86223134e-01 -3.88279527e-01 -3.17675680e-01 -9.91051853e-01
4.97931004e-01 -8.63472760e-01 -2.65106559e-01 -7.95657575e-01
2.90387571e-01 -1.72605604e-01 6.01164140e-02 3.43269825e-01
-2.00588793e-01 3.63849849e-01 3.90174508e-01 -3.30129676e-02
-3.54001075e-01 1.11136413e+00 1.75027239e+00 -2.45282173e-01
1.33309975e-01 -2.81785935e-01 -1.14326382e+00 8.13207388e-01
4.24824327e-01 -4.09766018e-01 -7.55716681e-01 -4.88460660e-01
3.19023639e-01 -1.64031852e-02 7.02272892e-01 -5.01563132e-01
-1.25710666e-01 -5.52484870e-01 3.41794670e-01 -3.77562121e-02
3.46228272e-01 -7.42999196e-01 2.14909941e-01 9.81304049e-02
-4.46403563e-01 -2.92242408e-01 2.76927054e-02 8.03902566e-01
-1.12849563e-01 1.67544633e-01 8.04827213e-01 -1.94854498e-01
-5.26257515e-01 4.37813491e-01 2.52054445e-02 2.03532249e-01
1.23924625e+00 -4.29551303e-01 -2.64613688e-01 -6.52594388e-01
-6.41912878e-01 2.69409478e-01 5.83748698e-01 6.41617358e-01
5.92013121e-01 -1.59070778e+00 -7.67078400e-01 4.51025099e-01
2.57230550e-01 -2.73265630e-01 4.93884593e-01 7.03124881e-01
-1.98040288e-02 1.84272230e-01 -4.10241425e-01 -3.44393432e-01
-1.02687275e+00 7.09863305e-01 2.36826897e-01 -1.32999361e-01
-3.03470612e-01 9.54060018e-01 8.77146244e-01 -5.18539965e-01
-1.50281027e-01 -6.62519410e-02 3.27275284e-02 1.93305537e-01
3.84278893e-01 5.99541643e-04 -5.74820995e-01 -4.32914019e-01
-7.09524676e-02 5.31186759e-01 -9.20038149e-02 -6.25734925e-02
1.20952427e+00 -3.11475098e-01 1.17892034e-01 9.25474018e-02
1.37465608e+00 4.47044373e-01 -1.99695063e+00 -3.36346477e-02
-4.13164735e-01 -1.03066814e+00 -1.25072390e-01 -6.74223721e-01
-1.19244087e+00 9.54817951e-01 4.44764376e-01 -3.06683570e-01
1.17008507e+00 -1.19674541e-01 4.56744492e-01 -1.38521150e-01
2.87850916e-01 -1.10781848e+00 2.60694534e-01 -2.81380415e-02
1.31193781e+00 -1.22305024e+00 -2.23650988e-02 -4.17182535e-01
-9.90756571e-01 7.06026196e-01 5.15568733e-01 -1.28655225e-01
5.27104676e-01 6.75286576e-02 1.08232377e-02 -2.41054341e-01
-7.31857777e-01 1.84938550e-01 3.98817211e-01 6.60094202e-01
1.51033491e-01 2.61453092e-01 -2.70330578e-01 5.76638877e-01
-4.68262106e-01 3.44294384e-02 4.41871583e-01 2.56723911e-01
1.22117765e-01 -7.41083264e-01 -3.21683168e-01 2.81348467e-01
-4.15210485e-01 -5.48245087e-02 -1.00525588e-01 7.78271019e-01
4.12100464e-01 2.78935522e-01 1.08217850e-01 -2.19420582e-01
1.16310336e-01 1.01540866e-03 6.67013347e-01 -5.63704491e-01
-3.75589244e-02 -1.79289550e-01 -1.19360611e-01 -6.47134006e-01
-4.65726793e-01 -6.94058776e-01 -8.29156995e-01 -4.35233325e-01
-3.31923842e-01 -2.33915597e-02 4.58525240e-01 8.99471521e-01
3.48806977e-01 2.67566919e-01 7.64625788e-01 -7.68642187e-01
-8.33469808e-01 -6.16072834e-01 -5.45575261e-01 7.59752631e-01
3.08365971e-01 -9.92755115e-01 -4.75101918e-01 2.36234441e-01] | [11.877861022949219, -0.26424431800842285] |
7b24b5b5-af7d-48ec-bc96-95a041449451 | ptgb-pre-train-graph-neural-networks-for | 2305.14376 | null | https://arxiv.org/abs/2305.14376v1 | https://arxiv.org/pdf/2305.14376v1.pdf | PTGB: Pre-Train Graph Neural Networks for Brain Network Analysis | The human brain is the central hub of the neurobiological system, controlling behavior and cognition in complex ways. Recent advances in neuroscience and neuroimaging analysis have shown a growing interest in the interactions between brain regions of interest (ROIs) and their impact on neural development and disorder diagnosis. As a powerful deep model for analyzing graph-structured data, Graph Neural Networks (GNNs) have been applied for brain network analysis. However, training deep models requires large amounts of labeled data, which is often scarce in brain network datasets due to the complexities of data acquisition and sharing restrictions. To make the most out of available training data, we propose PTGB, a GNN pre-training framework that captures intrinsic brain network structures, regardless of clinical outcomes, and is easily adaptable to various downstream tasks. PTGB comprises two key components: (1) an unsupervised pre-training technique designed specifically for brain networks, which enables learning from large-scale datasets without task-specific labels; (2) a data-driven parcellation atlas mapping pipeline that facilitates knowledge transfer across datasets with different ROI systems. Extensive evaluations using various GNN models have demonstrated the robust and superior performance of PTGB compared to baseline methods. | ['Carl Yang', 'Hejie Cui', 'Yi Yang'] | 2023-05-20 | null | null | null | null | ['unsupervised-pre-training'] | ['methodology'] | [ 1.33206472e-01 2.17921764e-01 -5.98814152e-02 -4.05267656e-01
2.31391471e-02 -3.04907888e-01 5.69112360e-01 2.95946807e-01
-4.02280658e-01 3.97676021e-01 2.41115302e-01 -1.22203611e-01
-4.16959316e-01 -8.77035797e-01 -5.32984018e-01 -3.97340506e-01
-3.59625667e-01 5.84973693e-01 3.71674806e-01 -3.05768605e-02
-9.80673283e-02 7.56456494e-01 -1.08366001e+00 1.22221515e-01
9.87426996e-01 8.34079385e-01 3.76942456e-01 2.09731132e-01
-2.08026901e-01 7.17723310e-01 -2.14569852e-01 -2.33819038e-01
2.11559579e-01 -4.09668654e-01 -7.40235746e-01 -3.63659948e-01
1.66279405e-01 -5.25730290e-02 -5.21594465e-01 1.20288467e+00
5.97691119e-01 5.10853417e-02 5.28684497e-01 -1.14586556e+00
-6.65368617e-01 7.16967285e-01 -3.12985331e-01 6.66660726e-01
-2.76280522e-01 3.01236778e-01 7.79614687e-01 -5.46317756e-01
8.26317191e-01 1.12430692e+00 5.62528551e-01 5.56459308e-01
-1.39936984e+00 -8.90931904e-01 -1.74110994e-01 2.63463646e-01
-1.28057253e+00 -3.92182291e-01 5.82707703e-01 -6.78187609e-01
9.17941749e-01 -1.75120637e-01 1.14491189e+00 1.10759687e+00
3.11069101e-01 2.81103760e-01 1.15298259e+00 1.89944580e-02
3.22856046e-02 -3.33804071e-01 2.93844957e-02 7.00482488e-01
1.21819243e-01 -9.51187611e-02 -2.94397354e-01 1.36129916e-01
9.83113110e-01 2.49923080e-01 -6.23570345e-02 -3.49221170e-01
-1.44209325e+00 5.80510974e-01 1.02791226e+00 4.34809357e-01
-4.51326638e-01 2.14203894e-01 4.94801402e-01 1.81510583e-01
6.06518805e-01 4.56033975e-01 -2.60273278e-01 2.09075838e-01
-9.90620196e-01 -5.04121631e-02 4.15341944e-01 6.25396371e-01
5.29183030e-01 6.02880539e-03 -3.67619902e-01 1.25143790e+00
2.89855510e-01 2.12922484e-01 5.49541712e-01 -4.82969463e-01
5.16189635e-01 1.08328521e+00 -8.78744125e-01 -1.11918664e+00
-9.50863600e-01 -7.21853375e-01 -1.17811704e+00 -5.93299717e-02
3.33792478e-01 -7.61474296e-02 -8.09717834e-01 2.00102091e+00
3.97568643e-01 2.27391064e-01 -4.92329478e-01 7.40724564e-01
1.07291496e+00 -2.33873483e-02 1.65456951e-01 2.78440207e-01
1.25671434e+00 -8.71907651e-01 -3.86953622e-01 -3.11860859e-01
7.31405616e-01 -1.00292951e-01 8.45302939e-01 1.34542689e-01
-7.84480751e-01 -2.22835064e-01 -7.24832594e-01 -2.41370246e-01
-7.29864061e-01 -1.42706648e-01 7.54068196e-01 4.12848383e-01
-1.46962333e+00 4.72069055e-01 -9.32359576e-01 -4.87544954e-01
1.39475203e+00 5.90526819e-01 -7.22461820e-01 -2.49975681e-01
-1.06815743e+00 9.80223358e-01 4.78581488e-01 3.08252126e-01
-1.10314536e+00 -1.10452521e+00 -7.20240176e-01 3.26863945e-01
2.48819664e-01 -6.50414765e-01 5.67122579e-01 -7.58199692e-01
-1.16068244e+00 9.89125073e-01 3.29700619e-01 -3.85967374e-01
3.44870389e-01 1.53832942e-01 -4.42287028e-01 3.47289711e-01
1.70707792e-01 8.51467907e-01 5.93331993e-01 -4.85371321e-01
-1.71070301e-03 -6.27427518e-01 -3.00363272e-01 6.59210384e-02
-3.58502775e-01 1.87891960e-01 -2.75014609e-01 -5.89648962e-01
2.27025114e-02 -6.68295801e-01 -3.55724484e-01 1.54924646e-01
-5.95068753e-01 -1.84542179e-01 5.79360187e-01 -7.34650195e-01
8.41991305e-01 -1.93898880e+00 2.68155605e-01 4.36301440e-01
9.98378396e-01 3.61723512e-01 -3.12436432e-01 1.84820399e-01
-4.30797726e-01 2.23171741e-01 -2.25874767e-01 3.36453766e-02
-2.29552105e-01 -1.89663947e-01 3.34746480e-01 5.65225065e-01
2.36387312e-01 1.33202302e+00 -1.08423972e+00 -4.42097485e-01
2.01272398e-01 3.85644138e-01 -4.91466254e-01 1.07006639e-01
-7.85595402e-02 7.37285733e-01 -4.62427825e-01 4.32405084e-01
3.35404515e-01 -4.88760829e-01 2.28340939e-01 -1.99438438e-01
1.06476925e-01 3.10573906e-01 -4.78085518e-01 1.83700800e+00
-1.39269128e-01 6.07136130e-01 2.23698467e-01 -1.29721522e+00
8.77969623e-01 3.11403215e-01 6.09920323e-01 -8.27539206e-01
5.20538509e-01 -7.84662589e-02 6.85502350e-01 -4.96662557e-01
-4.71931458e-01 1.71214253e-01 3.28636944e-01 5.44388711e-01
5.04766822e-01 8.81652460e-02 3.69919151e-01 3.46693873e-01
1.65797591e+00 -2.65982598e-01 2.38826096e-01 -5.54268658e-01
3.10163677e-01 -2.22693697e-01 5.88560879e-01 2.78173387e-01
-3.12943786e-01 5.25235176e-01 1.00133812e+00 -2.77413100e-01
-7.90703118e-01 -9.55064118e-01 -1.73311263e-01 1.04380369e+00
-3.78819883e-01 -2.03566775e-01 -8.49613965e-01 -6.37943208e-01
-4.06781174e-02 2.40762904e-01 -9.03244734e-01 -3.80278111e-01
-2.24729255e-01 -8.41768622e-01 8.03902924e-01 4.09357578e-01
5.12661040e-01 -1.43248141e+00 -2.90949047e-01 1.32719278e-02
3.05609912e-01 -1.21790588e+00 -4.30790484e-01 2.83613037e-02
-8.73970449e-01 -1.47907984e+00 -5.84564328e-01 -8.04694057e-01
1.11410332e+00 7.61551708e-02 9.97831762e-01 3.44196200e-01
-5.76535642e-01 1.87120527e-01 -1.24907847e-02 -3.66881371e-01
-2.45998517e-01 3.45621377e-01 -1.79342870e-02 5.67870326e-02
1.37568787e-01 -1.03142619e+00 -8.26854169e-01 2.01821074e-01
-9.92608368e-01 3.30027461e-01 8.74938190e-01 7.34885097e-01
6.03479922e-01 -1.35831490e-01 8.61215830e-01 -1.39048362e+00
7.24404156e-01 -8.91924620e-01 -6.96623623e-01 2.68264979e-01
-5.22845984e-01 -1.73674047e-01 6.64752662e-01 -2.22682491e-01
-7.35328794e-01 -2.44315654e-01 -8.48618001e-02 -3.36816907e-01
-2.66735494e-01 7.81008184e-01 -3.25414032e-01 -1.96313113e-01
5.66459894e-01 -8.10760185e-02 4.03082579e-01 -3.99536967e-01
4.36030954e-01 1.17352135e-01 4.07686859e-01 -3.91136050e-01
4.74653870e-01 2.58085847e-01 3.66686851e-01 -7.23326325e-01
-7.24623382e-01 -2.52941340e-01 -1.10349834e+00 -3.03089440e-01
1.07566285e+00 -5.77846706e-01 -5.64911425e-01 5.56818604e-01
-7.45532930e-01 -7.62097657e-01 -4.53723818e-02 4.39299732e-01
-5.67701943e-02 3.46400626e-02 -5.75153410e-01 8.17940533e-02
-5.92916548e-01 -1.08237529e+00 4.16871816e-01 2.29355648e-01
-2.60833167e-02 -1.24734902e+00 2.47110829e-01 2.64233649e-01
5.63368797e-01 3.84121060e-01 1.37604165e+00 -9.36761081e-01
-2.82526791e-01 -1.25101015e-01 -7.10828125e-01 2.70951897e-01
2.70537853e-01 -1.48787186e-01 -7.61104524e-01 -2.83742324e-02
-5.22964478e-01 -3.68608415e-01 7.08591282e-01 3.55108947e-01
1.51959991e+00 -4.79190499e-02 -4.48311746e-01 8.05828273e-01
1.11235321e+00 -2.31593892e-01 4.74413872e-01 1.94868166e-02
1.18406296e+00 7.97008932e-01 -3.22409570e-01 -1.52798280e-01
5.01017630e-01 4.42274570e-01 5.92543423e-01 -4.01292622e-01
-4.35990602e-01 -1.00700155e-01 -5.96043132e-02 7.73782909e-01
-1.89395342e-02 2.31710505e-02 -1.37747800e+00 7.21526086e-01
-1.73641801e+00 -7.86895990e-01 -1.37722507e-01 1.94958520e+00
8.03609490e-01 4.75605018e-02 1.66742295e-01 -3.62157702e-01
7.00999379e-01 7.69086108e-02 -9.19516146e-01 -2.42779851e-02
-4.37743440e-02 4.36672479e-01 2.39769891e-01 -1.69841275e-01
-7.89237380e-01 1.03689277e+00 6.34515190e+00 5.36912203e-01
-1.20217943e+00 5.03158212e-01 7.76039660e-01 -2.66545743e-01
-9.23955739e-02 -3.18806946e-01 -2.39029691e-01 5.05313337e-01
9.02873993e-01 -7.32677430e-02 7.47193515e-01 5.09964705e-01
3.35075170e-01 2.56504923e-01 -1.16353738e+00 7.50494003e-01
-1.79402962e-01 -1.39487958e+00 -1.20197065e-01 1.60055533e-01
6.74438417e-01 7.08116651e-01 -1.54442221e-01 2.37130001e-01
4.73065048e-01 -1.43362832e+00 2.00754732e-01 6.46820068e-01
9.66699958e-01 -5.36195576e-01 5.13937294e-01 2.02063397e-01
-9.65745509e-01 5.14799841e-02 -4.47897166e-01 2.34260380e-01
-9.06974971e-02 8.99529278e-01 -1.08412111e+00 2.91956604e-01
6.94643021e-01 1.02752435e+00 -8.49885404e-01 1.25712931e+00
-4.06057298e-01 7.99438477e-01 -1.57561362e-01 2.43142083e-01
4.35837694e-02 -4.32502955e-01 1.83718383e-01 1.13256991e+00
2.14518562e-01 3.00175902e-02 6.62535653e-02 1.39724863e+00
-4.85950351e-01 2.18181267e-01 -6.36828005e-01 -3.62468243e-01
3.99181724e-01 1.86788166e+00 -1.22028327e+00 9.72024351e-02
-5.06807446e-01 3.53010476e-01 9.32535172e-01 3.14961910e-01
-5.09530067e-01 -2.38136157e-01 6.37327731e-01 3.79303485e-01
-4.79398295e-02 -2.25296304e-01 -1.52602807e-01 -8.87702405e-01
-4.44709659e-01 -6.20928586e-01 3.70063841e-01 -5.60503781e-01
-1.45959091e+00 6.53563201e-01 -1.38843909e-01 -7.38839090e-01
1.20075874e-01 -5.50303161e-01 -9.42434549e-01 9.39411521e-01
-1.37905729e+00 -1.23010457e+00 -4.41889822e-01 7.48788714e-01
1.32713714e-04 -3.69764507e-01 7.14466870e-01 4.33980525e-01
-1.07962763e+00 3.19443673e-01 -1.18421130e-01 4.92056489e-01
4.74537075e-01 -1.00947440e+00 3.35592270e-01 8.53786469e-01
-3.53808608e-03 7.93277740e-01 -4.82210256e-02 -7.22540855e-01
-1.14179659e+00 -1.41884637e+00 4.13070410e-01 -3.18097711e-01
1.03984630e+00 -7.22803473e-01 -1.09619093e+00 7.90943265e-01
-7.11013824e-02 5.36202610e-01 8.21730316e-01 2.75697142e-01
-3.77659351e-01 -4.00646538e-01 -1.16335142e+00 4.72871602e-01
1.35018218e+00 -5.61561823e-01 -3.86622608e-01 4.66175705e-01
5.04773438e-01 -1.55407220e-01 -1.19962442e+00 2.35232964e-01
3.58709216e-01 -7.00001001e-01 8.70482743e-01 -6.42094493e-01
5.10734737e-01 -4.23997082e-02 4.32508647e-01 -1.57593465e+00
-5.62282324e-01 -1.90003961e-01 3.33880521e-02 1.26048148e+00
4.19275671e-01 -8.08114767e-01 5.55333078e-01 7.42545724e-01
-3.33878011e-01 -8.73803854e-01 -8.67013037e-01 -3.41844559e-01
8.09287503e-02 -3.45956504e-01 7.75150895e-01 1.06829560e+00
1.79822631e-02 4.19900388e-01 1.81256421e-02 -6.65436164e-02
4.15296823e-01 -3.92972499e-01 4.60871637e-01 -1.43368459e+00
1.24573670e-01 -8.77720475e-01 -5.42984366e-01 -2.47198209e-01
3.69992822e-01 -1.69768190e+00 -3.30649793e-01 -1.88506353e+00
2.88188726e-01 -5.84720850e-01 -7.19227850e-01 9.53465044e-01
5.90376966e-02 2.15146139e-01 -1.43152446e-01 4.51593883e-02
-6.36991620e-01 5.44639349e-01 1.49914217e+00 -3.47222015e-02
-4.68257032e-02 -2.38583803e-01 -7.58612990e-01 6.34859324e-01
7.55660951e-01 -5.58451712e-01 -7.19635010e-01 -6.17541909e-01
1.81779861e-01 -2.37106487e-01 5.78619063e-01 -9.80678916e-01
2.59938031e-01 7.62179643e-02 6.20447040e-01 -1.88916057e-01
-7.71437138e-02 -7.17692912e-01 1.51523918e-01 3.58691543e-01
-3.87068391e-01 5.04825823e-02 1.81219995e-01 3.96899581e-01
1.04382008e-01 1.86763614e-01 7.41104126e-01 -1.29056469e-01
-3.95358473e-01 9.11781251e-01 -3.46411526e-01 2.45722800e-01
8.60475600e-01 3.39728706e-02 -5.97592950e-01 -1.82805583e-02
-7.20813811e-01 3.39925319e-01 1.37915865e-01 4.90445256e-01
4.59877461e-01 -1.23344672e+00 -6.28374994e-01 2.68922240e-01
-9.61430818e-02 1.07416645e-01 4.65423942e-01 1.29071200e+00
-5.79838037e-01 3.82914156e-01 -7.09933579e-01 -4.97889429e-01
-9.60680127e-01 2.94479698e-01 3.61233264e-01 -5.27600229e-01
-7.58201480e-01 7.54438758e-01 4.82060969e-01 -7.74474323e-01
-5.21578155e-02 -3.59192282e-01 -6.73300326e-01 -1.32315932e-02
3.55070978e-01 4.13379490e-01 1.80835560e-01 -5.71237266e-01
-3.51319999e-01 -4.25030776e-05 -2.04494506e-01 3.29321414e-01
1.80488276e+00 1.54077619e-01 -6.54965341e-01 1.74854264e-01
9.45714414e-01 -4.11506772e-01 -1.08263779e+00 -3.40560466e-01
1.08735293e-01 -1.61396921e-01 2.02620342e-01 -8.94251466e-01
-1.75287652e+00 9.78792727e-01 5.15829325e-01 1.01884454e-01
1.04532170e+00 6.63605705e-02 5.56025505e-01 1.39616549e-01
3.83917689e-01 -7.47859299e-01 -2.55852509e-02 4.42010880e-01
9.11776960e-01 -9.69793439e-01 -6.18393496e-02 -2.98192054e-01
-2.77843297e-01 9.54927087e-01 8.80038083e-01 -1.72593519e-01
1.06725800e+00 -1.65868923e-02 -1.32396594e-01 -7.70860493e-01
-6.16387367e-01 5.04529290e-03 6.29233718e-01 7.58797228e-01
4.06550705e-01 -8.56710481e-04 -6.90342635e-02 9.24960375e-01
-2.58965939e-01 -7.94023201e-02 1.17798880e-01 2.96591491e-01
-9.92560163e-02 -1.09258902e+00 1.80275172e-01 1.19729793e+00
-4.62779492e-01 -2.95622647e-01 -4.70414251e-01 7.20789492e-01
1.55438915e-01 4.99239773e-01 -8.08260813e-02 -2.17976034e-01
1.27154022e-01 -9.03187692e-02 3.35857391e-01 -1.08370435e+00
-7.52907932e-01 -2.55344748e-01 -1.56586394e-01 -7.08834231e-01
-1.31682798e-01 -6.06110334e-01 -1.44776809e+00 -1.93722382e-01
-1.68672893e-02 -2.94003457e-01 5.42527378e-01 1.18425584e+00
8.18959355e-01 9.50482965e-01 2.41403580e-01 -7.77901053e-01
2.46294156e-01 -1.06730580e+00 -7.21335232e-01 3.61704171e-01
-9.13634673e-02 -7.91990638e-01 2.26744086e-01 -2.22774789e-01] | [12.405475616455078, 3.3552911281585693] |
4a1f42fc-1f5a-4325-9509-b9314c0d2fe5 | cope-end-to-end-trainable-constant-runtime | 2208.08807 | null | https://arxiv.org/abs/2208.08807v2 | https://arxiv.org/pdf/2208.08807v2.pdf | COPE: End-to-end trainable Constant Runtime Object Pose Estimation | State-of-the-art object pose estimation handles multiple instances in a test image by using multi-model formulations: detection as a first stage and then separately trained networks per object for 2D-3D geometric correspondence prediction as a second stage. Poses are subsequently estimated using the Perspective-n-Points algorithm at runtime. Unfortunately, multi-model formulations are slow and do not scale well with the number of object instances involved. Recent approaches show that direct 6D object pose estimation is feasible when derived from the aforementioned geometric correspondences. We present an approach that learns an intermediate geometric representation of multiple objects to directly regress 6D poses of all instances in a test image. The inherent end-to-end trainability overcomes the requirement of separately processing individual object instances. By calculating the mutual Intersection-over-Unions, pose hypotheses are clustered into distinct instances, which achieves negligible runtime overhead with respect to the number of object instances. Results on multiple challenging standard datasets show that the pose estimation performance is superior to single-model state-of-the-art approaches despite being more than ~35 times faster. We additionally provide an analysis showing real-time applicability (>24 fps) for images where more than 90 object instances are present. Further results show the advantage of supervising geometric-correspondence-based object pose estimation with the 6D pose. | ['Markus Vincze', 'Timothy Patten', 'Stefan Thalhammer'] | 2022-08-18 | null | null | null | null | ['6d-pose-estimation'] | ['computer-vision'] | [ 1.91199586e-01 1.44263059e-01 1.60246819e-01 -2.95616955e-01
-1.26806891e+00 -5.62502861e-01 5.72497606e-01 2.05493063e-01
-4.97750074e-01 1.38747655e-02 -5.88634074e-01 3.05256248e-02
-1.55657172e-01 -4.35732573e-01 -1.19994748e+00 -4.79520798e-01
-1.52087286e-01 1.41622221e+00 8.02109778e-01 1.54635474e-01
3.92851651e-01 1.11670673e+00 -1.70181835e+00 1.70378625e-01
1.53879166e-01 1.23315668e+00 2.41244003e-01 9.18203533e-01
1.82365030e-02 7.30681419e-02 -5.84517658e-01 -4.20597047e-01
6.96482897e-01 6.39848039e-02 -4.08723891e-01 4.16776776e-01
1.24308467e+00 -5.66688478e-01 1.18971504e-01 6.53655887e-01
4.84597206e-01 -1.06199764e-01 5.54572403e-01 -1.24164629e+00
2.91151255e-01 -1.96603209e-01 -8.75257671e-01 -3.12544219e-02
3.30975503e-01 1.10200293e-01 8.76591980e-01 -1.28433633e+00
5.91180742e-01 1.31515300e+00 8.32289696e-01 2.28016138e-01
-1.36593020e+00 -4.93107200e-01 3.00386101e-01 -5.25146201e-02
-1.48201072e+00 -1.09387673e-01 6.69570625e-01 -4.37624097e-01
1.28285289e+00 2.02259734e-01 6.81879878e-01 5.44016063e-01
5.37416935e-02 7.63766289e-01 7.56798923e-01 -3.39199692e-01
-8.30285102e-02 6.12317258e-03 -1.51507750e-01 7.88580298e-01
3.89530301e-01 1.15651982e-02 -5.07550240e-01 -1.13664314e-01
1.02828813e+00 1.06720477e-02 2.25564212e-01 -1.01049316e+00
-1.27216280e+00 5.18675148e-01 5.51865160e-01 -3.75032961e-01
-3.02666903e-01 2.84152508e-01 2.72490174e-01 -1.30834207e-01
4.77752924e-01 2.94760317e-01 -7.07090199e-01 7.74075985e-02
-1.04173481e+00 5.74877501e-01 6.17344737e-01 1.33834994e+00
7.60105073e-01 -3.02980840e-01 1.66700333e-01 4.09490407e-01
4.43955421e-01 5.56505799e-01 -1.69698283e-01 -8.79721165e-01
5.18397272e-01 7.60438323e-01 1.60188884e-01 -9.21843946e-01
-4.92527783e-01 -6.91795647e-01 -1.99008167e-01 4.15674239e-01
6.31331623e-01 3.19226354e-01 -9.70254183e-01 1.25144017e+00
9.52993095e-01 1.66916087e-01 -4.24315393e-01 8.62024486e-01
6.94123745e-01 4.07818437e-01 -1.56244546e-01 -3.39382738e-02
1.35729611e+00 -1.05853426e+00 4.99413796e-02 -4.87390935e-01
5.36772370e-01 -9.83205855e-01 6.08148277e-01 4.84194487e-01
-1.34570932e+00 -6.53376698e-01 -9.04935658e-01 -1.48688138e-01
-2.19157070e-01 3.86326164e-01 4.43623871e-01 3.51659924e-01
-6.67554140e-01 4.31110501e-01 -9.61729348e-01 -2.69965231e-01
6.03588462e-01 9.56296086e-01 -4.62694913e-01 6.84319157e-03
-2.11866722e-01 1.20739305e+00 5.92734039e-01 1.56097546e-01
-8.25984240e-01 -1.01645076e+00 -7.10943580e-01 -2.00878397e-01
5.32165885e-01 -6.55176103e-01 1.38954294e+00 -4.72786546e-01
-1.13966084e+00 1.00053096e+00 -1.28851891e-01 -2.83188999e-01
9.15879667e-01 -5.42898595e-01 2.00795844e-01 3.33604813e-01
2.13719327e-02 9.26281512e-01 8.01964462e-01 -1.43834662e+00
-7.36934125e-01 -5.27845681e-01 4.66880091e-02 4.04006332e-01
4.16755117e-02 7.26712309e-03 -9.79470372e-01 -6.86606690e-02
5.99659920e-01 -1.13917232e+00 -3.09156746e-01 5.24224877e-01
-2.01428339e-01 -1.98266715e-01 1.23089814e+00 -4.23863560e-01
4.93163228e-01 -1.88220966e+00 7.23318607e-02 2.30013415e-01
6.28817007e-02 2.13164806e-01 -2.20398650e-01 1.78526849e-01
5.68894781e-02 -3.05925488e-01 1.14275560e-01 -6.76768541e-01
-6.69134781e-02 2.92533617e-02 -6.61543086e-02 7.87222385e-01
4.90885884e-01 9.72824752e-01 -6.35955691e-01 -5.99765182e-01
5.81776023e-01 5.92159927e-01 -6.71313643e-01 2.10590780e-01
-4.70152676e-01 2.93259799e-01 -2.53095329e-01 6.99823201e-01
8.93377602e-01 -3.68996412e-01 1.64044723e-02 -5.93085647e-01
8.57627168e-02 1.80562586e-01 -1.37834752e+00 1.69888413e+00
-3.86971921e-01 3.25547606e-01 -2.60496675e-03 -6.64185047e-01
8.10703874e-01 9.45875794e-02 6.85661435e-01 -8.75412524e-02
1.48334846e-01 3.79145682e-01 2.31700297e-02 -7.20503479e-02
2.20638663e-01 -1.24960415e-01 6.19987920e-02 1.34300038e-01
2.47703716e-01 -5.95355451e-01 1.74540594e-01 -8.71411338e-02
8.11960638e-01 6.55390203e-01 3.27700198e-01 8.45520422e-02
2.95589298e-01 2.23077595e-01 1.68645531e-01 5.44073939e-01
1.19065689e-02 9.96086657e-01 2.51037478e-01 -5.83694518e-01
-1.18832707e+00 -1.20522654e+00 -3.00290018e-01 7.96607018e-01
3.79295170e-01 -4.94039118e-01 -5.68222106e-01 -8.16516399e-01
2.66807288e-01 4.26358938e-01 -2.72601128e-01 2.63291478e-01
-8.43823493e-01 -5.33475935e-01 6.35069609e-02 7.47880161e-01
1.89112559e-01 -7.80295908e-01 -1.13323891e+00 2.38624096e-01
1.29468173e-01 -1.48416996e+00 -5.18109091e-02 3.21617782e-01
-1.12495434e+00 -1.06863630e+00 -5.09452999e-01 -6.37548149e-01
9.18128908e-01 3.64687949e-01 1.11019444e+00 1.22602228e-02
-6.16941631e-01 4.42668378e-01 3.95303145e-02 -6.14417434e-01
-1.39728829e-01 -2.14792266e-02 2.26668045e-02 -2.78886884e-01
3.65982682e-01 -3.85028064e-01 -6.72135532e-01 5.36672592e-01
-4.64774191e-01 2.30815515e-01 8.99767816e-01 4.76267844e-01
1.11470914e+00 -2.26128295e-01 5.01521975e-02 -5.78208804e-01
-4.23259318e-01 3.07441149e-02 -9.51379836e-01 1.82432771e-01
-2.25223988e-01 -2.16558591e-01 1.19589880e-01 -7.03471065e-01
-7.30103970e-01 8.44026923e-01 1.35011375e-01 -8.73374283e-01
-3.35971385e-01 4.74725664e-02 -4.52107191e-02 -4.13092613e-01
4.75181073e-01 -8.53948575e-03 1.92302510e-01 -4.21616554e-01
1.65492982e-01 1.42401457e-01 3.97929072e-01 -4.56244677e-01
1.15660477e+00 4.43488687e-01 4.68545794e-01 -7.20832646e-01
-1.05101204e+00 -7.71295607e-01 -1.16207993e+00 -5.24077713e-01
7.93274403e-01 -9.94370759e-01 -9.73679841e-01 2.04708785e-01
-1.45575655e+00 1.91551493e-03 -1.28837213e-01 6.24537945e-01
-7.95903862e-01 7.87998140e-02 -3.63414973e-01 -8.21696997e-01
-1.40739575e-01 -1.13451803e+00 1.83532262e+00 -6.17419071e-02
-1.80790126e-01 -5.26602030e-01 -4.42173898e-01 6.00949645e-01
-6.06346428e-02 3.83034289e-01 5.51926434e-01 -6.76709890e-01
-1.24797165e+00 -7.41438687e-01 -2.74682462e-01 -9.85106751e-02
-2.83753157e-01 4.29221839e-02 -7.78844178e-01 -4.88808423e-01
-5.77079691e-02 -3.23010921e-01 3.89416337e-01 3.10837656e-01
1.06543827e+00 8.86318237e-02 -4.43494260e-01 3.49587262e-01
1.59857798e+00 -2.73102313e-01 3.06518823e-01 3.61769617e-01
9.54228580e-01 7.11817443e-01 1.05818093e+00 2.95431018e-01
6.68320656e-02 9.34402823e-01 9.02418613e-01 -1.35097191e-01
-1.34394914e-01 -1.72462165e-01 9.86038744e-02 2.66834646e-01
-1.28090233e-01 -2.05523353e-02 -1.09492445e+00 2.75690615e-01
-1.67380714e+00 -7.10199237e-01 -5.22129893e-01 2.37954760e+00
3.41765881e-01 5.68543255e-01 3.44997674e-01 -4.84133773e-02
5.83293319e-01 -1.97781906e-01 -6.35838926e-01 1.23362534e-01
3.23958158e-01 1.21557787e-01 6.84928179e-01 3.50279331e-01
-1.11686468e+00 9.61287200e-01 5.97418261e+00 7.78054416e-01
-8.91483426e-01 -5.10204844e-02 3.34543526e-01 -4.73861247e-01
3.35728496e-01 1.26005501e-01 -1.31630588e+00 -1.19608507e-01
6.21471047e-01 4.25297081e-01 -1.03099123e-01 1.13258481e+00
-9.71618071e-02 -3.30423146e-01 -1.41852856e+00 1.09475267e+00
2.00747862e-01 -1.19132662e+00 1.46404922e-01 2.05916792e-01
5.82205415e-01 2.40288481e-01 -1.49856389e-01 6.17226288e-02
-2.49781907e-01 -6.75238490e-01 1.05676544e+00 2.10295528e-01
7.01877594e-01 -8.23897839e-01 5.21943748e-01 7.51448333e-01
-1.36628973e+00 3.13307904e-02 -4.10732299e-01 9.06564370e-02
2.50705689e-01 3.09003592e-01 -1.34757686e+00 6.21353686e-01
6.03879511e-01 3.08299363e-01 -8.89177203e-01 1.21215332e+00
9.92972925e-02 1.11462407e-01 -8.91717076e-01 5.07423989e-02
1.35392964e-01 8.27700570e-02 6.27673328e-01 1.16484976e+00
3.02560389e-01 -7.21832439e-02 4.57178086e-01 6.88889146e-01
2.70259589e-01 -1.40594453e-01 -3.24370503e-01 3.28948915e-01
2.60771751e-01 1.32359755e+00 -1.17507148e+00 -1.94091767e-01
-3.78541738e-01 7.62390494e-01 2.85631776e-01 -9.70257223e-02
-9.78906810e-01 5.13568521e-02 3.49443287e-01 4.86169159e-01
7.49470413e-01 -5.82022190e-01 -2.82033235e-01 -6.10969424e-01
4.14684236e-01 -5.88209450e-01 3.25673185e-02 -7.59882987e-01
-1.00663543e+00 3.34261417e-01 4.01579797e-01 -1.63788760e+00
-3.11699897e-01 -9.16435778e-01 -1.51857585e-01 7.63197362e-01
-1.16658056e+00 -1.53885925e+00 -3.71282220e-01 8.01684782e-02
8.18618774e-01 1.86174601e-01 8.17641139e-01 2.05863994e-02
-9.13845152e-02 4.85979825e-01 -3.14529091e-01 -1.94709152e-02
4.87833768e-01 -1.08301508e+00 4.49886680e-01 5.69718480e-01
3.83412510e-01 3.91856670e-01 6.59637094e-01 -6.72824562e-01
-1.58519292e+00 -1.03539157e+00 7.59770572e-01 -9.83326316e-01
3.60477775e-01 -7.13386238e-01 -6.90705121e-01 6.75360203e-01
-4.43076283e-01 4.68669623e-01 2.87140161e-01 4.35778126e-03
-2.83002496e-01 -9.81426463e-02 -9.90921795e-01 3.90618056e-01
1.08877015e+00 -3.42374384e-01 -3.43110442e-01 5.55058002e-01
5.13644755e-01 -9.52506065e-01 -7.33525455e-01 7.72730589e-01
6.40555620e-01 -1.00066984e+00 1.24025333e+00 -2.95455664e-01
2.31083602e-01 -6.84290111e-01 -4.24789563e-02 -5.75503409e-01
-4.11087722e-02 -3.20814520e-01 -3.81550640e-01 7.35076904e-01
3.10955286e-01 -1.06915683e-01 1.25910079e+00 4.99868482e-01
-5.81611916e-02 -1.09992635e+00 -9.75295901e-01 -9.22816634e-01
-3.04923594e-01 -5.27171254e-01 1.99626356e-01 3.53653044e-01
-7.54857063e-01 3.52611452e-01 5.08655086e-02 5.92111588e-01
8.04453731e-01 3.82843018e-01 1.30536377e+00 -1.36022413e+00
-4.26631600e-01 -2.14301795e-01 -9.43329573e-01 -1.39503419e+00
-5.40376157e-02 -6.18034959e-01 1.10471301e-01 -1.39489138e+00
2.85524070e-01 -3.11287761e-01 1.06565803e-01 3.34462106e-01
3.04348245e-02 6.04866743e-01 3.73429477e-01 2.14439884e-01
-6.46506190e-01 1.50440946e-01 1.31999838e+00 1.93273947e-01
-8.02459419e-02 1.69358835e-01 -3.27498391e-02 9.00658071e-01
4.19557065e-01 -5.04030883e-01 -2.92703986e-01 -3.82164001e-01
-3.51496041e-02 1.73625544e-01 7.96435595e-01 -1.32104802e+00
2.01484129e-01 2.46983841e-02 8.34227860e-01 -1.43893588e+00
9.44989681e-01 -1.21567380e+00 3.00794423e-01 4.20898706e-01
-1.30522490e-01 2.31671631e-02 3.26034814e-01 6.33204281e-01
1.68445870e-01 -2.22355545e-01 8.72568309e-01 -1.03665739e-01
-5.22625268e-01 5.14255166e-01 1.84129164e-01 -3.61817539e-01
1.41254604e+00 -7.75133610e-01 2.66705751e-01 -7.78468847e-02
-7.08135724e-01 7.05297338e-03 4.61387306e-01 4.15242285e-01
5.89381874e-01 -1.11131048e+00 -6.42867506e-01 2.31910735e-01
1.38914883e-01 7.87820041e-01 5.61884195e-02 9.28486943e-01
-7.09091425e-01 2.44385391e-01 -4.70745238e-03 -1.39374352e+00
-1.70519006e+00 5.09993851e-01 3.27226371e-01 -1.89098433e-01
-6.39441907e-01 1.04993069e+00 2.88577497e-01 -4.14827406e-01
3.26076716e-01 -1.29249886e-01 3.44463229e-01 -2.62334105e-02
2.55007863e-01 2.99812615e-01 4.62948501e-01 -8.18774819e-01
-4.94073719e-01 1.07232952e+00 -3.33621949e-01 -1.36192456e-01
1.43759990e+00 1.78116724e-01 1.72120303e-01 3.18597853e-01
1.42940497e+00 -2.26458296e-01 -1.75568271e+00 -1.22875921e-01
1.14149516e-04 -6.41225874e-01 -2.46153802e-01 -6.82304800e-01
-6.94525242e-01 8.85169923e-01 6.31189585e-01 -3.37765276e-01
7.45100677e-01 3.40483099e-01 4.07564610e-01 5.44465363e-01
6.00820005e-01 -9.20989633e-01 3.53104949e-01 5.90057135e-01
1.00609374e+00 -1.21943808e+00 4.48086202e-01 -8.60439956e-01
-2.45059267e-01 1.36779237e+00 9.85937059e-01 -3.34110230e-01
4.21349317e-01 3.29184592e-01 -2.33719572e-02 -2.57554591e-01
-4.32666123e-01 -1.96629632e-02 6.52775347e-01 3.75499994e-01
1.93395406e-01 -2.66541868e-01 1.76182479e-01 -1.28515646e-01
-2.49485269e-01 -5.29084802e-01 -5.08520789e-02 1.20341635e+00
-4.35098857e-01 -9.62180376e-01 -5.95260084e-01 4.59816337e-01
-2.29629174e-01 3.13338339e-01 -4.18504328e-01 1.11966825e+00
1.74101278e-01 3.97195458e-01 3.36627930e-01 -2.25853726e-01
4.94230807e-01 1.90412924e-02 1.05434597e+00 -9.73086119e-01
-5.28744757e-01 4.37249005e-01 -8.61730799e-02 -7.99507320e-01
-4.78347600e-01 -7.86449015e-01 -1.34255540e+00 1.36475012e-01
-7.55153060e-01 -3.80100220e-01 9.24995720e-01 9.00581777e-01
3.80954355e-01 1.51553676e-01 2.40099624e-01 -1.81123698e+00
-7.64411986e-01 -8.01535785e-01 -1.13381997e-01 2.68316180e-01
2.69888699e-01 -8.65443051e-01 -2.71215767e-01 4.53079417e-02] | [7.567980766296387, -2.6578783988952637] |
ce4312f4-51ed-4c7f-b708-af920ecc887a | beyond-pretrained-features-noisy-image | 2302.01056 | null | https://arxiv.org/abs/2302.01056v2 | https://arxiv.org/pdf/2302.01056v2.pdf | Beyond Pretrained Features: Noisy Image Modeling Provides Adversarial Defense | Recent advancements in masked image modeling (MIM) have made it a prevailing framework for self-supervised visual representation learning. The MIM pretrained models, like most deep neural network methods, are still vulnerable to adversarial attacks, limiting their practical application, and this issue has received little research attention. In this paper, we investigate how this powerful self-supervised learning paradigm can provide adversarial robustness to downstream classifiers. During the exploration, we find that noisy image modeling (NIM), a simple variant of MIM that adopts denoising as the pre-text task, reconstructs noisy images surprisingly well despite severe corruption. Motivated by this observation, we propose an adversarial defense method by exploiting the pretrained decoder for denoising, referred to as De^3, through which NIM is able to enhance adversarial robustness beyond providing pretrained features. Furthermore, we incorporate a simple modification, sampling the noise scale hyperparameter from random distributions, and enable the defense to achieve a better and tunable trade-off between accuracy and robustness. Experimental results demonstrate that, in terms of adversarial robustness, NIM is superior compared to MIM thanks to its effective denoising capability. Moreover, the defense provided by NIM achieves performance on par with adversarial training while offering the extra tunability advantage. Source code and models will be made available. | ['Chang Xu', 'Bohyung Han', 'Daochang Liu', 'Zunzhi You'] | 2023-02-02 | null | null | null | null | ['adversarial-defense'] | ['adversarial'] | [ 3.81382108e-01 2.39363518e-02 -2.61976793e-02 -1.84693411e-01
-6.72289133e-01 -8.42160761e-01 7.72537410e-01 -4.54642564e-01
-3.18615675e-01 4.79945183e-01 1.85784608e-01 -3.95513147e-01
1.74072236e-01 -6.46006882e-01 -8.24447036e-01 -1.09175897e+00
6.31801179e-03 -6.02905691e-01 7.85195827e-02 -3.14533144e-01
5.46991415e-02 5.40011108e-01 -1.21871662e+00 2.44253382e-01
8.30878317e-01 1.04920650e+00 -1.08098157e-01 4.43728507e-01
3.05542707e-01 9.53274071e-01 -7.82232642e-01 -8.05253983e-01
6.58135891e-01 -3.77593219e-01 -4.33226436e-01 1.05274595e-01
3.63219500e-01 -5.74263334e-01 -8.44613135e-01 1.32820594e+00
6.60342932e-01 -4.16252427e-02 4.57920551e-01 -1.20324492e+00
-8.70719135e-01 6.52987540e-01 -3.55629086e-01 1.53246179e-01
-1.22283839e-01 4.73394513e-01 7.97186136e-01 -5.90059578e-01
3.36434394e-01 1.24816120e+00 6.72583878e-01 9.14509952e-01
-1.44315207e+00 -8.97542417e-01 2.79666394e-01 1.53942600e-01
-1.36663771e+00 -6.53290808e-01 1.10081613e+00 -7.20240772e-02
4.75602508e-01 3.22924286e-01 1.32565359e-02 1.63827944e+00
3.04423928e-01 6.24424756e-01 1.36152256e+00 -3.36873561e-01
3.77681345e-01 3.14700246e-01 -3.51166487e-01 4.65571076e-01
4.06874642e-02 6.60540164e-01 -5.45131981e-01 -2.17963845e-01
7.15296865e-01 -1.49641335e-01 -4.78384256e-01 -2.52947837e-01
-6.34883702e-01 8.08078289e-01 8.12792063e-01 1.76708639e-01
1.53020145e-02 3.74401212e-01 4.20147240e-01 7.67544806e-01
4.08861548e-01 4.07728791e-01 -1.38705030e-01 3.44234288e-01
-7.00109720e-01 -7.17923790e-02 5.63418329e-01 5.59672177e-01
5.48470974e-01 7.73325861e-01 -3.59313227e-02 6.53372347e-01
1.56233817e-01 4.47839767e-01 5.39672613e-01 -9.43938255e-01
2.81714916e-01 2.09535956e-01 -2.01779619e-01 -1.07302213e+00
-1.14518799e-01 -8.13507199e-01 -1.22265184e+00 8.21038544e-01
1.91888437e-01 -1.68827891e-01 -9.87477958e-01 2.15549183e+00
6.08849786e-02 1.76206231e-01 4.05928731e-01 9.24245059e-01
5.94246745e-01 4.24603850e-01 1.35957509e-01 -1.28154770e-01
9.99848604e-01 -6.21133685e-01 -6.18591547e-01 -4.82555926e-01
2.20995307e-01 -7.20411003e-01 9.17786181e-01 4.08231527e-01
-8.48153651e-01 -6.19571030e-01 -1.35502231e+00 3.53269368e-01
-2.38914654e-01 -1.81718618e-01 6.40311122e-01 1.15271294e+00
-1.05142939e+00 7.99052417e-01 -8.13401341e-01 1.63943339e-02
6.79198086e-01 3.20982277e-01 -4.77919549e-01 -3.65183502e-02
-1.30185902e+00 7.57205725e-01 7.09984377e-02 1.67033762e-01
-1.29231071e+00 -5.00279725e-01 -8.76963794e-01 -6.87293783e-02
3.40768784e-01 -6.18614912e-01 9.10299361e-01 -1.36653841e+00
-1.61182022e+00 7.73535490e-01 2.63081733e-02 -8.75279963e-01
7.92620599e-01 -9.97307003e-02 -4.45972681e-01 4.08891201e-01
-2.38028586e-01 5.01924396e-01 1.76130569e+00 -1.69975543e+00
8.22449699e-02 -2.97814190e-01 2.75917679e-01 -1.07815206e-01
-7.58950830e-01 1.49735525e-01 -1.65963277e-01 -1.22597539e+00
1.11339010e-01 -7.97953904e-01 -4.26222980e-01 1.83799163e-01
-5.12687683e-01 4.74791884e-01 9.38433707e-01 -4.68964458e-01
1.04434276e+00 -2.52711320e+00 -6.59155324e-02 3.54399294e-01
3.39661479e-01 5.80167770e-01 -1.52427152e-01 3.63364637e-01
-3.13794613e-01 4.60587949e-01 -4.07274157e-01 -5.79644442e-01
-4.39344831e-02 2.74656355e-01 -7.23826647e-01 7.08119452e-01
4.21741754e-01 9.46508944e-01 -5.25831640e-01 -1.34076312e-01
9.93116349e-02 5.61159849e-01 -3.73117119e-01 3.61576617e-01
1.30572729e-02 7.02516437e-01 -2.80787885e-01 6.91447377e-01
9.70666170e-01 7.80967390e-03 1.12786867e-01 -2.01333120e-01
2.62857586e-01 -2.20472634e-01 -1.01608753e+00 1.13283920e+00
-3.70368451e-01 6.37360513e-01 4.44498479e-01 -1.13750744e+00
1.00099063e+00 3.50505590e-01 -7.96800181e-02 -7.33942211e-01
3.68128449e-01 -1.51440548e-02 -7.99703524e-02 -3.46171051e-01
-1.56019544e-02 -2.48928338e-01 2.12662332e-02 2.01585978e-01
-8.17159638e-02 1.19530857e-02 -4.13168758e-01 2.84651875e-01
1.22858834e+00 -5.60536981e-02 1.25737205e-01 -5.31284176e-02
5.01020789e-01 -3.22845578e-01 5.96996844e-01 1.07925749e+00
-3.60390484e-01 6.66280746e-01 4.09089833e-01 -7.02598691e-02
-8.86767626e-01 -1.22291183e+00 -9.02571157e-02 8.84710729e-01
2.56619126e-01 -1.58440933e-01 -9.28085208e-01 -6.54337168e-01
7.26785213e-02 3.03287208e-01 -7.65518785e-01 -6.58620656e-01
-6.86842918e-01 -6.78128779e-01 9.42031503e-01 5.83457649e-01
8.51218879e-01 -8.14830840e-01 -1.52282834e-01 -1.38399273e-01
1.49081096e-01 -1.20625353e+00 -2.82216907e-01 4.53415751e-01
-8.54463100e-01 -8.15563321e-01 -6.04809344e-01 -7.85801113e-01
6.72521949e-01 3.53142381e-01 7.56265640e-01 3.63243818e-01
-1.68580562e-01 2.66703039e-01 -4.01649773e-01 -2.41014183e-01
-9.06857312e-01 -3.00458461e-01 2.46411890e-01 3.40519965e-01
-2.55680442e-01 -1.04498541e+00 -5.51796556e-01 3.48953843e-01
-1.42632425e+00 -3.13266695e-01 6.96509480e-01 1.06585121e+00
2.86825150e-01 3.21944922e-01 7.53780842e-01 -8.09322298e-01
4.17944968e-01 -3.72144789e-01 -4.81164008e-01 1.31119858e-03
-5.43322265e-01 2.50154398e-02 1.11364257e+00 -6.98307097e-01
-1.01467991e+00 -3.91122289e-02 -3.73766720e-01 -7.31852114e-01
-1.71641335e-01 2.56170630e-01 -5.67087531e-01 -7.25089192e-01
8.58694851e-01 2.92615563e-01 3.14095825e-01 -4.57801074e-01
4.55281436e-01 5.33755660e-01 8.80233169e-01 -3.31702024e-01
1.55408823e+00 7.74178922e-01 9.83295869e-03 -6.76150382e-01
-7.99218118e-01 3.88154238e-02 -3.31688970e-01 -6.40758574e-02
6.54015183e-01 -1.04078686e+00 -6.36141241e-01 7.32135355e-01
-8.86457443e-01 -3.71697426e-01 -9.45440829e-02 1.28599107e-01
-4.92948443e-01 6.09215498e-01 -7.06631958e-01 -8.86429906e-01
-2.39922091e-01 -1.02979088e+00 5.43641329e-01 4.26507890e-02
2.12000042e-01 -9.61181223e-01 -5.91539443e-01 3.86839867e-01
7.61662662e-01 5.85762203e-01 9.38457310e-01 -5.62480032e-01
-5.88559806e-01 -2.98121512e-01 -9.24428850e-02 1.02329898e+00
9.39384028e-02 -3.85915220e-01 -1.31443763e+00 -6.32614970e-01
5.89133620e-01 -4.30721164e-01 1.19847012e+00 7.35780299e-02
1.28682888e+00 -5.21359563e-01 3.56431529e-02 1.11354208e+00
1.46357191e+00 -2.60353144e-02 8.35600972e-01 5.53354800e-01
6.18469954e-01 3.89874488e-01 9.44180414e-02 3.04110616e-01
-1.41682819e-01 4.81355220e-01 9.82726097e-01 -2.83145368e-01
-2.60435134e-01 -2.55202949e-01 6.35699093e-01 4.49674457e-01
1.88194305e-01 -1.28908172e-01 -5.34289062e-01 1.95565000e-01
-1.53756702e+00 -9.83206391e-01 4.00889724e-01 2.14270258e+00
8.78137171e-01 4.29105669e-01 -1.87137663e-01 4.56212044e-01
5.74382484e-01 3.81688863e-01 -7.19415545e-01 -2.97040612e-01
-5.84447861e-01 3.03377062e-01 6.49772048e-01 3.83271128e-01
-1.35790002e+00 1.03126276e+00 6.45716619e+00 9.93100941e-01
-1.19744849e+00 1.27161920e-01 7.74833679e-01 1.50088802e-01
-2.66229391e-01 -3.52978222e-02 -4.84461874e-01 5.03568113e-01
7.13517785e-01 -6.39140047e-03 5.59692740e-01 7.69856215e-01
3.02732259e-01 2.55408138e-01 -8.56912374e-01 7.52099156e-01
2.03097805e-01 -1.30298162e+00 2.44091988e-01 -7.11192191e-02
6.84078872e-01 -3.13552320e-01 5.01138210e-01 1.86170325e-01
2.86446482e-01 -1.17486525e+00 8.07093561e-01 2.88392186e-01
6.76907837e-01 -9.06346023e-01 6.50994897e-01 4.08705711e-01
-8.34627569e-01 -5.35294175e-01 -3.59214306e-01 5.89730702e-02
-2.63208169e-02 4.71967310e-01 -2.97835439e-01 5.37545264e-01
7.35723257e-01 3.21505040e-01 -6.56985283e-01 6.67530477e-01
-4.71748590e-01 9.54355001e-01 1.08699329e-01 4.77169812e-01
1.54462904e-01 1.26823410e-01 7.51673758e-01 9.87877727e-01
-1.01392873e-01 -1.83467939e-01 1.46724790e-01 9.00870144e-01
-2.91739702e-01 -2.68744409e-01 -6.89881325e-01 1.47651196e-01
4.66979057e-01 9.09398139e-01 -4.65735525e-01 -5.19414917e-02
-1.32154673e-01 1.23624074e+00 2.63822019e-01 6.59083903e-01
-8.78734589e-01 -2.24633515e-01 9.17544544e-01 -1.48043260e-01
4.20863897e-01 -2.54255086e-01 -5.31725109e-01 -1.02248132e+00
1.22076064e-01 -1.37952232e+00 7.35695213e-02 -4.12420720e-01
-1.56678462e+00 8.29206049e-01 -3.56814742e-01 -1.38632774e+00
2.26127710e-02 -6.06635988e-01 -7.71737576e-01 6.51988566e-01
-1.67677021e+00 -1.29887700e+00 -1.10658959e-01 8.24088931e-01
2.45339185e-01 -2.82086074e-01 7.67045617e-01 1.03021845e-01
-7.83225417e-01 1.22892344e+00 2.42676184e-01 2.88476139e-01
7.60719180e-01 -9.60012257e-01 4.70471740e-01 1.24435115e+00
9.33338553e-02 8.00365746e-01 7.82418609e-01 -4.76958573e-01
-1.47880483e+00 -1.27287805e+00 1.40439317e-01 -2.02102721e-01
7.70220637e-01 -5.48090816e-01 -9.96524751e-01 4.35707152e-01
1.51663497e-01 1.93490341e-01 5.53445995e-01 -4.88735318e-01
-9.31739926e-01 -2.37424031e-01 -1.48498344e+00 8.95418286e-01
1.10095501e+00 -8.05025101e-01 -3.53526205e-01 2.04198956e-01
1.02179360e+00 -2.45095536e-01 -6.23180568e-01 4.46003020e-01
8.29496831e-02 -1.11686444e+00 1.25522435e+00 -4.47878271e-01
3.92403185e-01 -2.33108684e-01 -3.65865022e-01 -1.28708255e+00
-2.46270165e-01 -1.18995440e+00 -3.37512434e-01 1.36804807e+00
1.91186875e-01 -8.49872887e-01 7.93406546e-01 3.55203420e-01
-6.58572465e-03 -6.22768104e-01 -1.05924654e+00 -1.20193088e+00
3.02554786e-01 -5.32551169e-01 3.84331703e-01 7.90450692e-01
-4.56634909e-01 -1.75601438e-01 -7.90819585e-01 6.28223658e-01
8.75251651e-01 -2.00007141e-01 6.88691318e-01 -6.64641619e-01
-5.04088879e-01 -3.13334763e-01 -5.53412795e-01 -9.74691272e-01
3.78151000e-01 -8.63029599e-01 -3.35274935e-02 -8.15698147e-01
-1.29272223e-01 -2.90468603e-01 -4.17463303e-01 5.32932818e-01
-3.44165117e-01 7.62595892e-01 3.36073011e-01 3.32250148e-01
-1.05320312e-01 5.15653133e-01 1.02941132e+00 -2.33429745e-01
8.49448591e-02 5.99201396e-02 -1.10298336e+00 7.28798270e-01
9.77828860e-01 -5.19352555e-01 -2.42691010e-01 -4.94290501e-01
-2.76872724e-01 -3.23859751e-01 6.77590251e-01 -1.12578988e+00
8.63311887e-02 6.66566240e-03 5.17737806e-01 1.86505273e-01
4.59580839e-01 -8.43587577e-01 -1.60290092e-01 4.98861700e-01
-3.97501051e-01 -2.43792787e-01 3.26025158e-01 7.86498964e-01
-2.64767647e-01 -1.93132430e-01 1.08764124e+00 -4.12604026e-02
-5.81887424e-01 1.88287705e-01 -5.45764625e-01 -1.18164174e-01
8.29053521e-01 -2.84537435e-01 -4.10590857e-01 -6.13309562e-01
-7.04568505e-01 -1.48361444e-01 4.92994577e-01 2.57755190e-01
6.42879725e-01 -1.26543558e+00 -6.10998154e-01 3.83172035e-01
-8.69405195e-02 -5.79891264e-01 3.26177359e-01 4.98899251e-01
-1.33708283e-01 -1.18562452e-01 -1.52265346e-02 -3.32353443e-01
-1.22683334e+00 8.56688976e-01 3.46489847e-01 -1.00759901e-01
-6.14546180e-01 9.59104657e-01 2.37770930e-01 -1.77705094e-01
5.57754695e-01 2.93211788e-01 7.57537186e-02 -3.67493689e-01
7.04438388e-01 9.49435984e-04 -3.21354158e-02 -5.36009431e-01
-2.05191776e-01 3.40986222e-01 3.49705629e-02 5.53093329e-02
1.17199576e+00 -2.52002954e-01 -1.37512358e-02 -1.33642673e-01
1.25488007e+00 2.26725891e-01 -1.59050405e+00 -3.81269723e-01
-1.59237072e-01 -6.14580870e-01 2.14463547e-01 -7.00073719e-01
-1.33990192e+00 7.41766393e-01 7.68372893e-01 3.31426024e-01
1.45264530e+00 -2.19317034e-01 6.98091686e-01 2.37554297e-01
2.70200759e-01 -6.72902405e-01 3.71187598e-01 3.27902198e-01
9.97418225e-01 -1.29986191e+00 -1.70959651e-01 -5.76129079e-01
-6.73365593e-01 8.28598678e-01 5.74094236e-01 -3.62259895e-01
6.41474724e-01 5.82896769e-01 5.44370651e-01 2.54886985e-01
-5.28650641e-01 -1.31056011e-01 1.13813348e-01 9.58505869e-01
-1.32588223e-01 -2.47193798e-01 9.80702415e-02 6.40101612e-01
-2.44941011e-01 -4.81859595e-01 3.16673338e-01 9.08642590e-01
-3.45510542e-01 -1.20470786e+00 -6.87544584e-01 2.44890433e-02
-7.70817816e-01 -2.29474440e-01 -4.27338392e-01 5.16286016e-01
7.34697953e-02 1.32948327e+00 -3.09515148e-01 -7.45557904e-01
1.98915765e-01 -1.67710051e-01 1.81164816e-01 -2.72336513e-01
-9.09934759e-01 -2.01536208e-01 -2.08557934e-01 -5.77785969e-01
-1.68599352e-01 -1.46008581e-01 -8.46637309e-01 -4.48498726e-01
-3.28939736e-01 -1.40196994e-01 5.77886581e-01 9.75288212e-01
2.71524101e-01 4.59395021e-01 1.33934081e+00 -9.54838395e-01
-1.17342687e+00 -6.26872957e-01 -3.36030126e-01 5.07094800e-01
6.15415692e-01 -5.48705101e-01 -9.49222863e-01 5.80150224e-02] | [5.545076370239258, 7.932552337646484] |
1326be5c-dee6-4027-be4a-e025e1c86ee3 | tensor-based-multi-view-block-diagonal | null | null | https://ieeexplore.ieee.org/document/9428106 | https://ieeexplore.ieee.org/document/9428106 | Tensor-Based Multi-View Block-Diagonal Structure Diffusion for Clustering Incomplete Multi-View Data | In this paper, we propose a novel incomplete multi-view clustering method, in which a tensor nuclear norm regularizer elegantly diffuses the information of multi-view block-diagonal structure across different views. By exploring the membership between observed and missing samples and that between missing ones in each incomplete view with the guidance of the high-order view consistency, a global block-diagonal structure is well preserved in multiple spectral embedding matrices. Meanwhile, a consensus representation with strong separability is obtained for clustering. An iterative algorithm based on Augmented Lagrange Multiplier (ALM) is designed to solve the resultant model. Experimental results on six benchmark datasets indicate the superiority of the proposed method. | ['En Zhu', 'Wei zhang', 'Xiao Zheng', 'Xinwang Liu', 'Chang Tang', 'Zhenglai Li'] | 2021-06-09 | null | null | null | ieee-international-conference-on-multimedia | ['incomplete-multi-view-clustering'] | ['computer-vision'] | [-2.40004048e-01 -4.13631380e-01 -3.18826407e-01 -2.44053170e-01
-5.60591221e-01 -4.96226460e-01 3.12951773e-01 -5.57616413e-01
1.28496125e-01 3.93904597e-01 5.68959534e-01 3.40889305e-01
-5.83867967e-01 -1.70845404e-01 -2.54010916e-01 -1.22126245e+00
3.28459799e-01 2.74522692e-01 -4.23792869e-01 1.50745586e-01
1.36076465e-01 -1.68382227e-02 -1.11585557e+00 4.90743577e-01
8.36264908e-01 5.72357714e-01 1.23454593e-01 1.66447148e-01
3.22796375e-01 6.51888728e-01 8.61151516e-02 -1.12693034e-01
3.99598479e-01 -3.75432909e-01 -4.49974597e-01 9.35183465e-01
3.81160021e-01 -1.17694728e-01 -3.25372696e-01 1.36250448e+00
3.60082597e-01 -9.26657766e-03 4.74128336e-01 -1.26486814e+00
-9.63112831e-01 4.57965672e-01 -1.28717196e+00 -1.36766866e-01
4.30442005e-01 -8.88852403e-02 1.27222419e+00 -1.49941802e+00
7.56453335e-01 1.26235831e+00 1.71982750e-01 7.34523311e-02
-1.65113604e+00 -4.61152494e-01 3.54368240e-01 3.21170628e-01
-1.65185022e+00 -2.45859161e-01 1.16879511e+00 -6.85856164e-01
2.17248857e-01 3.38721156e-01 6.29546583e-01 9.77582932e-01
4.53156903e-02 7.83166826e-01 1.43417454e+00 9.68094543e-02
-5.97230010e-02 5.79062253e-02 9.86976027e-02 6.99687958e-01
4.51345503e-01 1.34893013e-02 -5.72717905e-01 -4.13782120e-01
4.13454145e-01 4.37184155e-01 -5.69015861e-01 -1.11172509e+00
-1.69641018e+00 8.63483787e-01 1.81594238e-01 2.75845677e-01
-3.84539902e-01 -5.03947139e-01 4.73521918e-01 9.79698896e-02
4.42029178e-01 -2.69121587e-01 -5.52508198e-02 3.19328427e-01
-7.38579690e-01 -2.91015625e-01 4.26247716e-01 9.49450016e-01
9.28810716e-01 1.68284908e-01 2.22338304e-01 8.10799956e-01
7.19164610e-01 6.92931771e-01 3.81561011e-01 -1.26303697e+00
6.12066329e-01 9.58893180e-01 -2.83725951e-02 -1.52015388e+00
-2.25042194e-01 -3.82278204e-01 -1.64636981e+00 -1.18022449e-01
-2.34795152e-03 4.06327769e-02 -2.89885670e-01 1.48190165e+00
5.86118937e-01 2.69670576e-01 1.57264307e-01 1.14997113e+00
7.12796807e-01 6.66270852e-01 -6.51964188e-01 -7.08818972e-01
1.28404903e+00 -7.14586616e-01 -1.00667083e+00 5.99467307e-02
1.42249539e-01 -7.48984754e-01 5.40700972e-01 5.21773100e-01
-1.02756059e+00 -6.41178608e-01 -1.20238650e+00 1.15420319e-01
2.50587076e-01 3.40710104e-01 3.61766040e-01 3.48087788e-01
-7.53257036e-01 -1.57582890e-02 -7.71698475e-01 2.78457850e-02
6.65275231e-02 2.81783134e-01 -8.70012879e-01 -5.08280933e-01
-6.82220221e-01 2.24868387e-01 4.67498094e-01 5.28451622e-01
-5.70837975e-01 -5.70456028e-01 -7.69848347e-01 -1.79841682e-01
4.55013990e-01 -7.78994441e-01 1.18168950e-01 -6.91027284e-01
-1.15601575e+00 8.58752131e-01 -3.62323880e-01 3.56914431e-01
3.42428356e-01 7.79366568e-02 -6.49836183e-01 3.83410424e-01
3.13532293e-01 2.01921716e-01 1.10901558e+00 -1.86864901e+00
-3.74385983e-01 -8.04289877e-01 -2.73656607e-01 4.52694803e-01
-3.92764300e-01 -3.87208045e-01 -5.51069498e-01 -5.40854216e-01
9.69151795e-01 -1.08534479e+00 -2.13312477e-01 -2.36570597e-01
-4.83460605e-01 1.21978022e-01 1.01376748e+00 -7.66480625e-01
1.43592954e+00 -2.32819366e+00 1.07177842e+00 6.15940928e-01
5.61331630e-01 -2.55255044e-01 1.53991014e-01 6.59561992e-01
-4.51384962e-01 -3.05963367e-01 -3.72889578e-01 -1.79114610e-01
-7.79870003e-02 2.65214324e-01 -1.00660853e-01 9.24688280e-01
-4.52871710e-01 4.60523546e-01 -7.77653217e-01 -4.09011394e-01
2.83152670e-01 5.64549088e-01 -5.81922710e-01 1.94346160e-01
5.55335939e-01 8.23310554e-01 -3.64217281e-01 5.53369582e-01
1.11569571e+00 -6.34451210e-01 6.78737998e-01 -7.91574776e-01
-2.62579955e-02 -5.46882093e-01 -1.81882298e+00 1.97367513e+00
2.45369896e-02 -5.65316379e-02 7.28080809e-01 -1.06310916e+00
6.35355830e-01 4.61757064e-01 9.02506053e-01 -2.29520485e-01
1.93430223e-02 1.24468006e-01 -2.08734578e-04 -4.96422321e-01
2.40666419e-01 -1.14023767e-01 1.96171656e-01 5.56818187e-01
-2.19670445e-01 4.78667080e-01 1.47320971e-01 5.15900195e-01
3.53290141e-01 -9.45616141e-02 2.12962195e-01 -6.48375452e-01
1.15217912e+00 -4.72994328e-01 8.95422101e-01 -2.13617440e-02
-1.60809845e-01 6.73501074e-01 3.19871724e-01 -4.65569198e-01
-1.04285979e+00 -1.02356279e+00 -1.06285587e-01 6.28007114e-01
2.64265895e-01 -5.15029967e-01 -4.84749585e-01 -4.50527102e-01
-4.87185605e-02 -1.70794725e-02 -6.69297993e-01 -1.28066018e-01
-3.62555891e-01 -1.00039554e+00 -2.18672886e-01 2.93031037e-01
4.89550114e-01 -2.82808632e-01 8.61685723e-02 6.38434570e-03
-6.30784988e-01 -1.01927483e+00 -8.28803480e-01 -2.81071186e-01
-9.80111241e-01 -1.13797879e+00 -6.09579682e-01 -8.99855793e-01
1.05458486e+00 1.02720153e+00 7.04884350e-01 2.24843938e-02
3.92931476e-02 6.77799046e-01 -1.94755524e-01 5.78140199e-01
-1.52496815e-01 -4.54839140e-01 4.27522063e-01 9.17132556e-01
1.87267259e-01 -6.61276162e-01 -6.49753511e-01 5.88002324e-01
-1.00569069e+00 2.42241561e-01 4.45179015e-01 1.22437036e+00
7.61122525e-01 2.21894965e-01 2.97814161e-01 -7.87710011e-01
3.14024687e-01 -6.09371126e-01 -4.46972430e-01 3.70022982e-01
-7.24405169e-01 -1.58588290e-01 5.20849884e-01 -1.56666100e-01
-1.14235818e+00 1.37007162e-01 3.84067655e-01 -9.21969950e-01
1.83066785e-01 7.41406262e-01 -5.47688007e-01 3.15545499e-02
1.28534194e-02 6.61524236e-01 2.96629548e-01 -4.45002973e-01
6.92701519e-01 3.32878351e-01 3.03121150e-01 -6.79548681e-01
9.68530476e-01 1.16913879e+00 -1.85529459e-02 -7.97804892e-01
-8.16263974e-01 -8.52760017e-01 -1.08945334e+00 -1.66808486e-01
8.95214498e-01 -1.39737821e+00 -8.40237141e-01 3.53174061e-01
-6.13204539e-01 5.89715600e-01 1.38984239e-02 7.92577147e-01
-4.79539782e-01 1.08580887e+00 -7.01153100e-01 -4.65593338e-01
5.34868129e-02 -1.12850428e+00 8.46560121e-01 -3.00480574e-01
2.23198891e-01 -1.20650327e+00 2.30348706e-01 8.37038994e-01
-3.23779047e-01 1.02040730e-01 6.49537504e-01 -5.89146577e-02
-6.48967564e-01 -9.61573422e-02 -2.67129630e-01 4.25619006e-01
3.38822603e-01 1.37379140e-01 -6.41418517e-01 -8.31446528e-01
5.24603546e-01 -6.09801598e-02 6.21285677e-01 4.83738095e-01
8.57398748e-01 -3.91641706e-01 -1.40961811e-01 7.51406670e-01
1.59351265e+00 -1.20810546e-01 3.38574797e-01 8.87662843e-02
1.25606370e+00 6.72099590e-01 4.61319536e-01 7.73500144e-01
7.20321774e-01 3.79560530e-01 3.15358192e-01 1.69571079e-02
3.55166733e-01 -1.38538092e-01 3.54139656e-01 1.95234716e+00
-1.34455130e-01 2.57061541e-01 -5.05650401e-01 4.97535437e-01
-2.09838796e+00 -1.22645032e+00 -6.38818443e-01 1.99081123e+00
3.61759037e-01 -2.86090583e-01 9.87878889e-02 2.23685175e-01
9.59010363e-01 5.60447216e-01 -4.52070773e-01 2.14383528e-01
-4.98593807e-01 -7.88577855e-01 1.54318884e-01 5.93060851e-01
-9.93924081e-01 2.81128824e-01 6.07207251e+00 7.39180326e-01
-5.32418072e-01 7.74738342e-02 4.00728792e-01 7.09429905e-02
-5.81391394e-01 1.58941686e-01 -3.63253951e-01 4.44266468e-01
2.64359623e-01 -1.57235965e-01 7.02697635e-01 4.28250462e-01
4.50375825e-01 8.23456496e-02 -8.27682734e-01 1.43575728e+00
4.23486292e-01 -1.10699868e+00 2.89886951e-01 3.59477043e-01
1.21853125e+00 -2.73700118e-01 2.73288637e-01 -3.14135134e-01
2.54044324e-01 -3.87698561e-01 4.82407421e-01 4.19669420e-01
5.74374378e-01 -8.55480254e-01 3.76177639e-01 4.08820361e-01
-1.56326795e+00 -4.97398674e-02 -5.49402535e-01 1.41324066e-02
2.15615511e-01 7.34748244e-01 -8.35041795e-03 1.24066162e+00
5.64337730e-01 1.59334993e+00 -4.31332409e-01 4.25748020e-01
-5.21267131e-02 2.04756349e-01 7.67966658e-02 8.31508875e-01
2.45125696e-01 -1.37434125e+00 7.37407625e-01 6.24781728e-01
7.62166083e-02 2.95015812e-01 5.43175280e-01 6.34081602e-01
2.65810996e-01 1.12833954e-01 -7.12962687e-01 3.74939024e-01
2.41606981e-01 1.66434681e+00 -6.43263280e-01 -3.01859379e-01
-8.29301715e-01 1.14770281e+00 2.54577219e-01 6.36910856e-01
-5.57556212e-01 5.27684867e-01 5.83673120e-01 -3.94591868e-01
3.56812984e-01 -4.00906056e-01 -3.06749403e-01 -1.77834105e+00
3.48438531e-01 -1.10211968e+00 7.12918878e-01 -6.33748055e-01
-1.76292765e+00 1.54781103e-01 -1.67353168e-01 -1.78256154e+00
2.79181778e-01 -3.32868457e-01 -4.29232866e-01 6.40014112e-01
-1.09944022e+00 -1.29112363e+00 -1.28644615e-01 1.07852757e+00
2.13361591e-01 -2.40283594e-01 4.96294558e-01 5.23470044e-01
-9.52978313e-01 1.51730031e-01 7.43542671e-01 2.64653992e-02
6.88235819e-01 -1.15097547e+00 -5.07060528e-01 8.52764189e-01
1.66634828e-01 9.07572865e-01 3.58310074e-01 -5.84731400e-01
-1.87456322e+00 -8.64819527e-01 3.16579431e-01 -2.44010136e-01
9.26486254e-01 -2.37722158e-01 -9.03081775e-01 9.60913002e-01
5.30843794e-01 5.33286147e-02 1.21131396e+00 1.48524001e-01
-6.94463909e-01 -1.57729879e-01 -8.37964475e-01 3.90444547e-01
7.73218930e-01 -7.86367178e-01 -4.33033824e-01 3.37755442e-01
3.59834999e-01 -2.11230442e-01 -1.31534648e+00 2.84912437e-01
6.48143113e-01 -1.26624012e+00 1.13488495e+00 -5.66544890e-01
3.45932066e-01 -8.74780238e-01 -5.68057120e-01 -1.33804393e+00
-9.59911883e-01 -4.23640490e-01 -2.35841006e-01 1.32471502e+00
2.77571511e-02 -5.24643183e-01 6.32713854e-01 2.89187819e-01
4.15034480e-02 -5.27853549e-01 -9.71068859e-01 -7.63141334e-01
-2.97257394e-01 9.58898589e-02 1.72298342e-01 1.45635366e+00
4.85964827e-02 5.17706752e-01 -9.05961990e-01 4.97918010e-01
1.26470053e+00 5.25109828e-01 5.59042752e-01 -1.10369647e+00
-1.65537283e-01 3.91391478e-03 -2.83486873e-01 -8.67788017e-01
2.36074343e-01 -1.06121111e+00 -4.55525130e-01 -1.31086528e+00
9.33652163e-01 8.89960397e-03 -4.98813719e-01 -1.98540136e-01
-2.98896760e-01 1.55358687e-01 1.42998561e-01 7.79176474e-01
-8.96572649e-01 9.25787210e-01 1.41381681e+00 -1.86695829e-01
1.34144679e-01 -2.81795621e-01 -7.80146837e-01 8.67990315e-01
2.40856305e-01 -2.87803322e-01 -5.53888381e-01 -3.68927687e-01
2.91713238e-01 2.76596278e-01 1.84329420e-01 -6.79228961e-01
2.54265428e-01 -1.18834995e-01 3.38708282e-01 -8.61250579e-01
3.88203025e-01 -1.24395728e+00 5.57307839e-01 4.27980959e-01
7.09300786e-02 1.18611783e-01 -3.81257206e-01 1.19840157e+00
-5.02467632e-01 1.90217599e-01 7.73687303e-01 -7.21426755e-02
-4.53392118e-01 5.97482204e-01 -5.56206852e-02 8.83821249e-02
9.76429701e-01 -2.92538404e-01 -3.13364677e-02 -1.05675325e-01
-1.09517992e+00 4.82189596e-01 6.03618562e-01 3.49422634e-01
8.39899719e-01 -2.12167668e+00 -8.78997982e-01 4.41686094e-01
3.28240931e-01 -2.41714031e-01 9.47474480e-01 1.23001230e+00
-1.57576859e-01 2.16379836e-01 -1.95454091e-01 -9.96148467e-01
-1.33126843e+00 9.95952487e-01 -3.64361145e-02 -2.89480656e-01
-6.83925509e-01 4.51386869e-01 6.48758292e-01 -6.72852576e-01
-1.96964353e-01 3.56848538e-01 -3.10633838e-01 4.70987320e-01
5.58286130e-01 7.72328794e-01 -3.29909176e-01 -1.29427123e+00
-2.88572878e-01 9.91769850e-01 7.15702772e-02 -7.46176243e-02
1.38737106e+00 -9.18039083e-01 -7.72894859e-01 7.78446674e-01
1.56411445e+00 1.52013093e-01 -1.30903935e+00 -4.42264944e-01
-1.21175155e-01 -7.73548067e-01 -4.65665124e-02 -2.86014583e-02
-1.44993067e+00 6.68433547e-01 3.00849408e-01 1.08347826e-01
1.06530380e+00 -2.95628041e-01 4.26699221e-01 1.81415156e-01
2.45332956e-01 -1.04171979e+00 1.95378169e-01 1.17014661e-01
8.92515898e-01 -1.35465980e+00 3.94402921e-01 -6.30548358e-01
-8.36507618e-01 8.98600757e-01 4.99764889e-01 -2.38986492e-01
8.40399206e-01 -3.94058436e-01 -1.06479444e-01 -6.58760726e-01
-6.86350942e-01 2.03101739e-01 4.41175222e-01 3.69802117e-01
8.78751874e-02 1.72389507e-01 -2.40382522e-01 3.03821772e-01
3.67846072e-01 -5.09918749e-01 5.80031157e-01 6.65522814e-01
-1.15964055e-01 -9.48760986e-01 -7.09427774e-01 2.09007874e-01
-2.17176080e-01 1.66367114e-01 -3.31317484e-01 4.26189512e-01
-7.14842230e-02 1.04036975e+00 -2.87634492e-01 -4.40979570e-01
1.18433610e-01 -1.28627539e-01 2.95489401e-01 -4.73930955e-01
-1.06801942e-01 7.96125591e-01 -2.85228252e-01 -6.62239969e-01
-1.03830862e+00 -9.98350918e-01 -9.11136091e-01 -3.62604380e-01
-2.63922572e-01 2.09620982e-01 1.36224061e-01 6.39495194e-01
3.22956622e-01 1.46466538e-01 1.22479570e+00 -6.85772359e-01
-5.06364465e-01 -4.54883456e-01 -1.26401258e+00 7.11802363e-01
2.63666332e-01 -6.17319107e-01 -6.51910424e-01 2.98202187e-01] | [8.262006759643555, 4.622869491577148] |
06b329cf-a8f3-4f2f-9862-c95a09cb0e9a | on-device-scalable-image-based-localization | 1802.03510 | null | http://arxiv.org/abs/1802.03510v2 | http://arxiv.org/pdf/1802.03510v2.pdf | On-device Scalable Image-based Localization via Prioritized Cascade Search and Fast One-Many RANSAC | We present the design of an entire on-device system for large-scale urban
localization using images. The proposed design integrates compact image
retrieval and 2D-3D correspondence search to estimate the location in extensive
city regions. Our design is GPS agnostic and does not require network
connection. In order to overcome the resource constraints of mobile devices, we
propose a system design that leverages the scalability advantage of image
retrieval and accuracy of 3D model-based localization. Furthermore, we propose
a new hashing-based cascade search for fast computation of 2D-3D
correspondences. In addition, we propose a new one-many RANSAC for accurate
pose estimation. The new one-many RANSAC addresses the challenge of repetitive
building structures (e.g. windows, balconies) in urban localization. Extensive
experiments demonstrate that our 2D-3D correspondence search achieves
state-of-the-art localization accuracy on multiple benchmark datasets.
Furthermore, our experiments on a large Google Street View (GSV) image dataset
show the potential of large-scale localization entirely on a typical mobile
device. | ['Ngai-Man Cheung', 'Tuan-Anh Bui', 'Thanh-Toan Do', 'Anh-Dzung Doan', 'Dang-Khoa Le Tan', 'Ngoc-Trung Tran', 'Mengxuan Tan'] | 2018-02-10 | null | null | null | null | ['image-based-localization'] | ['computer-vision'] | [-4.70299989e-01 -6.00178063e-01 -3.02072793e-01 -2.30948702e-01
-1.38848782e+00 -7.44148374e-01 5.43120086e-01 7.68132657e-02
-3.18139523e-01 2.45652080e-01 1.35762721e-01 -4.01875943e-01
9.56655815e-02 -1.01269734e+00 -7.77150810e-01 -3.83264989e-01
1.21866316e-01 5.26981831e-01 6.46090150e-01 -3.26312155e-01
3.33677143e-01 7.56566942e-01 -1.52142274e+00 -2.27853671e-01
5.76725185e-01 1.13137567e+00 1.51565447e-01 6.32643819e-01
1.84349388e-01 9.13508832e-02 -2.32505441e-01 -1.93982467e-01
3.28855157e-01 3.93833756e-01 -4.32229340e-01 -9.48882177e-02
8.75591099e-01 -7.13103235e-01 -5.69561541e-01 6.05698407e-01
9.05716121e-01 -4.91665751e-02 3.25516939e-01 -1.25801754e+00
-4.16543365e-01 -9.30451304e-02 -7.13338256e-01 1.54377729e-01
8.61254513e-01 -7.53751993e-02 7.90424764e-01 -1.28151107e+00
3.48149031e-01 1.09277344e+00 1.27554798e+00 -3.59316111e-01
-1.00786400e+00 -6.10611975e-01 -7.50849023e-02 -9.22522768e-02
-2.34105992e+00 -4.49603170e-01 4.01780993e-01 -2.75848001e-01
1.05952251e+00 3.06587070e-01 6.44746900e-01 6.60960674e-01
-9.37245786e-02 6.21642530e-01 6.29438937e-01 -1.68565720e-01
1.42677873e-01 -1.16109692e-01 -1.32654503e-01 1.04987431e+00
4.78848398e-01 -7.99598470e-02 -5.79015791e-01 -5.63767791e-01
7.19469607e-01 5.13916016e-01 -2.64007486e-02 -7.04216838e-01
-1.30597126e+00 8.59205902e-01 9.19848204e-01 3.21565121e-02
4.77781259e-02 6.75288379e-01 7.61386603e-02 -1.58514857e-01
3.69044632e-01 -7.61347339e-02 -1.32424176e-01 -9.28213447e-02
-1.24871027e+00 8.39611143e-02 5.10708570e-01 1.43452644e+00
1.04575098e+00 -3.92832845e-01 1.23235352e-01 6.39084876e-01
4.84089941e-01 1.34542704e+00 3.90517145e-01 -8.28537524e-01
7.39719689e-01 6.00153029e-01 3.41791868e-01 -1.43616724e+00
-3.24133754e-01 -2.05901042e-01 -5.83648145e-01 -5.16562164e-01
6.86436445e-02 4.44025397e-01 -5.48407912e-01 1.26939368e+00
6.77105427e-01 5.89792609e-01 -2.72181690e-01 6.94865406e-01
9.34877276e-01 4.60686028e-01 -4.59764451e-01 5.09118795e-01
1.44047177e+00 -1.03941381e+00 -1.25358701e-01 -2.81066388e-01
8.84815156e-01 -9.40155506e-01 8.58991086e-01 -3.11519891e-01
-8.23820353e-01 -5.35403311e-01 -9.64845717e-01 -3.12908173e-01
-5.97745299e-01 4.14625466e-01 5.62355518e-01 8.89264107e-01
-1.24140096e+00 -3.81089002e-02 -8.72128785e-01 -8.70623648e-01
1.49829313e-01 5.87192714e-01 -3.93021464e-01 -2.44529173e-01
-8.18041861e-01 5.97973645e-01 3.98360975e-02 -1.23373225e-01
-2.20777124e-01 -3.96867722e-01 -9.75812614e-01 -1.41130574e-02
9.67885256e-02 -7.51941383e-01 9.72079456e-01 2.51448840e-01
-1.08395576e+00 1.01306355e+00 -6.46035075e-01 -5.01601219e-01
3.19730580e-01 -2.90130585e-01 -4.24018443e-01 3.09267640e-01
7.86493540e-01 6.78271353e-01 5.83343387e-01 -1.12291217e+00
-7.93382943e-01 -4.46261466e-01 -2.63963044e-01 1.73241161e-02
-2.32732356e-01 -3.86700809e-01 -9.41429138e-01 -4.25121039e-01
9.24615145e-01 -1.28232455e+00 -1.87865183e-01 5.46020418e-02
-2.89437681e-01 1.19091205e-01 1.11225379e+00 -3.00893843e-01
1.29755306e+00 -2.12834692e+00 -7.26947427e-01 5.32509327e-01
2.81158835e-02 -7.95276612e-02 1.08704440e-01 5.64939439e-01
4.60348517e-01 1.25695616e-01 3.00910592e-01 -6.95719659e-01
1.11036077e-01 6.38711080e-02 -5.74086189e-01 6.98060393e-01
-4.35330480e-01 1.22271466e+00 -8.98272455e-01 -5.74688315e-01
4.86526996e-01 5.68524718e-01 -5.18995821e-01 -2.46448338e-01
6.50364697e-01 2.15091825e-01 -7.38390982e-01 1.24607623e+00
9.27456677e-01 -6.61162794e-01 -2.26391941e-01 -2.92960733e-01
-2.86495179e-01 3.18476528e-01 -1.09159195e+00 1.86592126e+00
-6.44141436e-01 4.25157636e-01 -1.72393203e-01 -4.62947071e-01
8.36383879e-01 6.45313470e-04 4.48382556e-01 -7.03230560e-01
-2.02776909e-01 5.55583894e-01 -1.02293670e+00 -4.43504984e-03
9.74385202e-01 7.48643279e-01 -5.26294172e-01 4.95703191e-01
-2.80070990e-01 -4.44097519e-01 -1.90683082e-01 2.62513518e-01
1.22378910e+00 -1.27197593e-01 4.44625288e-01 -1.03989273e-01
4.96276617e-01 5.48307672e-02 1.23144567e-01 1.03538775e+00
-1.37039810e-01 7.72587299e-01 -4.32922363e-01 -6.29920125e-01
-9.24985409e-01 -1.18410289e+00 -1.24134466e-01 8.38000238e-01
7.61517465e-01 -7.38849044e-01 -3.57743442e-01 -4.49370563e-01
3.72984588e-01 -4.27285843e-02 -1.28814608e-01 3.03408951e-01
-5.95537484e-01 -3.85587454e-01 9.13619041e-01 5.09584308e-01
9.54973698e-01 1.38437167e-01 -7.82689154e-01 2.84426603e-02
-5.36192477e-01 -1.41892207e+00 -7.60480583e-01 -3.31259578e-01
-8.48479271e-01 -8.89353096e-01 -5.88919699e-01 -9.01488543e-01
7.50137210e-01 1.27520001e+00 9.59415972e-01 2.14074433e-01
-5.46185412e-02 8.34599137e-01 -2.09730953e-01 1.71668276e-01
5.04183412e-01 3.09061408e-01 3.08416545e-01 -2.27071032e-01
3.99010062e-01 -5.89018583e-01 -9.81617868e-01 8.72612298e-01
-2.20746815e-01 -3.91560107e-01 3.71128768e-01 6.21056557e-01
9.02680874e-01 -4.07186616e-03 -1.30153582e-01 -5.63452989e-02
1.46428153e-01 -4.51212347e-01 -9.62583482e-01 1.80730268e-01
-4.97994483e-01 -1.64469153e-01 5.24206236e-02 -2.91121453e-01
-3.98693770e-01 5.50223589e-01 -1.25911087e-01 -1.76616326e-01
1.74802803e-02 1.93743244e-01 1.24799535e-01 -7.71661401e-01
4.17836934e-01 4.25443500e-01 -3.55443358e-01 -3.70211989e-01
4.50406581e-01 8.83402169e-01 6.18716657e-01 -3.64058554e-01
1.19869673e+00 1.14204133e+00 4.64301556e-02 -8.75122130e-01
-6.06619358e-01 -1.16698003e+00 -7.93512106e-01 1.58572420e-01
5.58335662e-01 -1.76538062e+00 -9.63831246e-01 9.99185890e-02
-1.10145462e+00 9.37815756e-02 2.42079958e-01 4.09575641e-01
-4.05479074e-01 3.82056981e-01 -3.25242341e-01 -5.15230060e-01
-4.68469381e-01 -1.34256482e+00 1.95144320e+00 1.30747393e-01
-1.87415004e-01 -6.21535957e-01 2.14652061e-01 5.62267303e-01
5.25905669e-01 1.09167591e-01 3.49698327e-02 -2.15224341e-01
-1.17000103e+00 -6.46472037e-01 -3.71117204e-01 -6.92257047e-01
-1.69305429e-02 -3.92877072e-01 -8.61027777e-01 -4.97654617e-01
-5.53427517e-01 -1.12210482e-01 6.92415118e-01 3.00494462e-01
6.88991189e-01 -1.21310875e-01 -9.46678758e-01 7.92216301e-01
1.41401923e+00 -2.24456459e-01 5.45237482e-01 6.37960911e-01
7.44385600e-01 -1.52090356e-01 9.04519796e-01 4.62198555e-01
1.07639658e+00 1.11270332e+00 3.47536087e-01 -9.17637050e-02
-4.78129499e-02 -6.64657176e-01 1.58472687e-01 6.95723355e-01
2.72346884e-01 -1.52742062e-02 -1.01138330e+00 7.81394660e-01
-1.95405746e+00 -8.13910604e-01 -2.30599582e-01 2.25006652e+00
1.27641886e-01 -6.25620484e-02 3.52768251e-03 -2.08857670e-01
7.38522589e-01 2.55375326e-01 -2.16435134e-01 3.09206516e-01
-5.42626483e-03 -9.72878374e-03 1.18630254e+00 5.63549340e-01
-1.46237457e+00 1.13149548e+00 6.51748657e+00 9.97942448e-01
-1.14370942e+00 3.32610697e-01 2.46945471e-01 4.05199751e-02
-6.82862997e-02 -1.15084217e-03 -1.29419911e+00 5.05659819e-01
6.80648804e-01 4.15450543e-01 1.18136667e-01 1.30017948e+00
1.32397667e-01 -4.12409872e-01 -6.42981529e-01 1.53658819e+00
3.17996033e-02 -1.65724289e+00 -1.89312816e-01 4.08759385e-01
7.93855429e-01 6.23696744e-01 9.02086049e-02 -3.47354971e-02
6.70957714e-02 -6.26328766e-01 8.89905751e-01 -9.54913422e-02
1.07594097e+00 -7.32715786e-01 6.11856341e-01 3.32086265e-01
-1.98154414e+00 -1.62519421e-02 -2.44385868e-01 9.99740660e-02
3.67395788e-01 5.83975554e-01 -9.89039361e-01 2.76322722e-01
9.85158861e-01 6.09399855e-01 -7.88825810e-01 1.15654171e+00
-1.61571398e-01 1.25118956e-01 -9.71052110e-01 -8.12277719e-02
3.86143088e-01 5.72330430e-02 2.41449565e-01 1.14255202e+00
1.04654765e+00 2.20327917e-02 3.47452134e-01 3.81038666e-01
-7.17948424e-03 -8.56237561e-02 -1.19393826e+00 5.33621311e-01
1.19737828e+00 1.11951566e+00 -9.89105999e-01 -3.30915481e-01
-2.21090466e-01 1.14285421e+00 4.42736708e-02 1.17785737e-01
-8.55685353e-01 -3.78694624e-01 5.87628603e-01 4.65486288e-01
7.66721904e-01 -7.70407081e-01 -1.42831087e-01 -1.20277131e+00
-3.70890670e-03 -3.28991979e-01 1.07672989e-01 -7.77754843e-01
-7.13287771e-01 3.14378291e-01 -2.01765478e-01 -1.46379864e+00
-9.86869633e-02 -8.24776068e-02 -3.75981361e-01 3.86257023e-01
-1.42353451e+00 -1.55591154e+00 -6.19826078e-01 8.39653969e-01
2.44298443e-01 1.14061628e-02 8.65765989e-01 7.20997751e-01
-1.53377339e-01 5.24037600e-01 4.29928929e-01 2.37462714e-01
6.97258413e-01 -6.93132222e-01 1.05000997e+00 8.45848620e-01
2.19566375e-01 9.07071948e-01 9.84392315e-02 -7.16463804e-01
-1.60715771e+00 -1.07636225e+00 1.10756528e+00 -7.95911908e-01
5.44846833e-01 -6.25998497e-01 6.32294491e-02 7.99960196e-01
-4.13419247e-01 3.28598738e-01 5.42601943e-01 -1.14211537e-01
-4.76289093e-01 -3.31697375e-01 -1.31312251e+00 4.69621599e-01
1.20017719e+00 -8.82613540e-01 -1.64533854e-01 4.97903019e-01
7.15375662e-01 -6.61545753e-01 -6.91404462e-01 3.13274294e-01
6.30580187e-01 -8.67231905e-01 1.73719025e+00 4.19240713e-01
-3.26683253e-01 -8.75185609e-01 -6.98316932e-01 -6.52369201e-01
-2.43678600e-01 -6.32355988e-01 -1.77330971e-01 1.01688802e+00
6.37595216e-03 -8.29572797e-01 9.29455698e-01 4.08717483e-01
1.83037102e-01 -4.07095611e-01 -1.22855890e+00 -7.46840298e-01
-8.78945768e-01 -4.31395799e-01 1.01804256e+00 6.40870333e-01
-4.10478830e-01 2.39550263e-01 -3.31776232e-01 8.80121827e-01
7.00203300e-01 3.87035608e-01 1.30002606e+00 -9.80256319e-01
6.27046824e-02 1.27136961e-01 -7.45603919e-01 -1.80824101e+00
-1.73184380e-01 -6.10816061e-01 -2.20242485e-01 -1.47067022e+00
5.28429821e-02 -7.48470604e-01 2.65794158e-01 2.73752570e-01
3.87590170e-01 9.09343123e-01 -2.63801180e-02 6.35297179e-01
-1.21378803e+00 4.86171246e-01 4.57273573e-01 -2.82598317e-01
-2.31405720e-01 1.34121910e-01 -3.89844030e-01 7.19438910e-01
5.83175123e-01 -5.54011703e-01 -2.85432249e-01 -7.23482370e-01
2.62618929e-01 -2.20500931e-01 6.46518886e-01 -1.23474896e+00
6.97644174e-01 2.88493514e-01 4.37915325e-01 -1.42668581e+00
8.19148540e-01 -1.02932370e+00 2.22867846e-01 6.46549702e-01
4.39628154e-01 4.90351766e-01 -5.12175001e-02 8.24374259e-01
-1.60957783e-01 3.14630121e-01 3.98472011e-01 -1.22191645e-01
-9.02861595e-01 3.09434652e-01 -1.09664813e-01 -3.29301029e-01
1.03782070e+00 -5.54543912e-01 -3.20617676e-01 -6.50944293e-01
-2.05267295e-01 3.35485280e-01 7.92462826e-01 2.01417267e-01
7.29772806e-01 -1.83202040e+00 -7.82382786e-02 2.14089289e-01
4.01680499e-01 -3.40907499e-02 -1.11440912e-01 8.83392751e-01
-9.37697470e-01 8.14104974e-01 3.67488921e-01 -1.10624611e+00
-1.47782385e+00 2.06000417e-01 6.33152351e-02 -2.12732673e-01
-4.88951296e-01 7.55216837e-01 -1.93850875e-01 -8.27732861e-01
1.32197574e-01 -3.94555330e-01 5.03889203e-01 -1.27583042e-01
6.17820621e-01 4.97611076e-01 1.58752427e-01 -1.08499706e+00
-8.21494043e-01 1.34900892e+00 2.71790206e-01 -1.63412124e-01
1.22847509e+00 -8.19137037e-01 -3.34919919e-03 -3.67555432e-02
1.33690500e+00 3.00724447e-01 -7.83787549e-01 -3.46534163e-01
-2.62950510e-02 -7.77705312e-01 4.78629172e-02 -2.00931460e-01
-7.29961038e-01 4.72151399e-01 8.44650149e-01 -8.21079314e-02
6.92357779e-01 5.50476415e-03 1.30568159e+00 8.75650883e-01
1.13530874e+00 -9.83110607e-01 9.38281640e-02 6.98595703e-01
2.60349929e-01 -1.49572587e+00 2.69311666e-01 -4.00574565e-01
-1.29561767e-01 8.59553099e-01 1.18415780e-01 -4.05375175e-02
8.21071506e-01 9.68907624e-02 -1.59998983e-01 -3.48269135e-01
8.20531324e-02 -4.04834330e-01 1.25628352e-01 5.50589740e-01
-8.66008550e-02 -5.10508893e-03 1.78405464e-01 -3.11900284e-02
-2.86515951e-01 -2.42049530e-01 -1.05612598e-01 1.14124537e+00
-6.03537798e-01 -9.22009468e-01 -9.39339936e-01 -7.79745802e-02
4.29036058e-02 -2.39465401e-01 -1.05417676e-01 8.96973968e-01
1.30332839e-02 1.03852010e+00 -1.27277048e-02 -5.24925351e-01
1.43360659e-01 -3.87962282e-01 1.96798578e-01 -2.61970580e-01
-1.93940669e-01 1.48739815e-01 -2.13162333e-01 -1.10884202e+00
-3.10918510e-01 -3.98477316e-01 -1.05587113e+00 -6.62064552e-01
-4.65832651e-01 -2.05565505e-02 9.75031853e-01 6.44791842e-01
1.02054155e+00 -3.76977205e-01 6.64127290e-01 -1.21402621e+00
-2.60227561e-01 -3.68234575e-01 -4.80739802e-01 1.55727146e-02
4.02886689e-01 -7.05383599e-01 -1.06463015e-01 -2.86002815e-01] | [7.694272518157959, -2.194432020187378] |
f1eb1e7f-398b-4b0f-af5f-d6fd5710e667 | indoor-group-activity-recognition-using-multi | 2101.10857 | null | https://arxiv.org/abs/2101.10857v1 | https://arxiv.org/pdf/2101.10857v1.pdf | Indoor Group Activity Recognition using Multi-Layered HMMs | Discovery and recognition of Group Activities (GA) based on imagery data processing have significant applications in persistent surveillance systems, which play an important role in some Internet services. The process is involved with analysis of sequential imagery data with spatiotemporal associations. Discretion of video imagery requires a proper inference system capable of discriminating and differentiating cohesive observations and interlinking them to known ontologies. We propose an Ontology based GAR with a proper inference model that is capable of identifying and classifying a sequence of events in group activities. A multi-layered Hidden Markov Model (HMM) is proposed to recognize different levels of abstract GA. The multi-layered HMM consists of N layers of HMMs where each layer comprises of M number of HMMs running in parallel. The number of layers depends on the order of information to be extracted. At each layer, by matching and correlating attributes of detected group events, the model attempts to associate sensory observations to known ontology perceptions. This paper demonstrates and compares performance of three different implementation of HMM, namely, concatenated N-HMM, cascaded C-HMM and hybrid H-HMM for building effective multi-layered HMM. | ['Vinayak Elangovan'] | 2021-01-23 | null | null | null | null | ['group-activity-recognition'] | ['computer-vision'] | [ 4.93086249e-01 -1.61132082e-01 -1.35835215e-01 -4.43732023e-01
-1.31988212e-01 -4.15344298e-01 8.20897281e-01 3.98854524e-01
-2.32095063e-01 1.94728404e-01 2.61059999e-01 -1.20946459e-01
-3.42888802e-01 -7.21008778e-01 -3.22417378e-01 -7.65583456e-01
-8.22211683e-01 2.77999967e-01 8.91823113e-01 -4.62627597e-02
2.99136281e-01 5.70540130e-01 -2.35437679e+00 9.07430828e-01
1.14306957e-01 1.03847218e+00 5.44895411e-01 1.28649640e+00
1.79953745e-03 1.45357549e+00 -4.48549747e-01 1.38365224e-01
8.31763297e-02 -3.57608497e-01 -8.46166313e-01 6.29347563e-01
1.62748292e-01 -2.26065621e-01 2.09775746e-01 1.03527439e+00
-8.26028064e-02 2.52594650e-01 7.08853841e-01 -1.50721562e+00
3.83329727e-02 8.55693370e-02 1.07717872e-01 4.62838918e-01
7.03977585e-01 -1.52239054e-01 5.48275173e-01 -2.92975187e-01
5.43541968e-01 1.20890689e+00 5.78870893e-01 7.47627988e-02
-7.34460175e-01 -3.20197344e-01 -2.73021124e-02 8.01523685e-01
-1.45067322e+00 -3.94588381e-01 1.96676627e-01 -7.88305461e-01
1.39142239e+00 3.35864693e-01 7.75903821e-01 8.12220156e-01
2.25295782e-01 3.54792684e-01 1.04213333e+00 -6.57569647e-01
3.82419765e-01 1.81831464e-01 4.21157807e-01 6.23827815e-01
1.35437235e-01 -8.77120346e-02 -5.93145549e-01 -3.61459494e-01
6.17317259e-01 1.40068099e-01 3.01181406e-01 1.01323193e-03
-7.99560249e-01 5.67522407e-01 -1.21016957e-01 3.39802563e-01
-1.02854288e+00 -2.95736641e-01 3.20634395e-01 2.06595045e-02
-8.86937976e-02 -8.97959545e-02 -1.60187587e-01 -1.54597044e-01
-6.70167208e-01 2.46789947e-04 7.28492498e-01 8.94058168e-01
7.78996050e-01 -3.92777711e-01 1.84934705e-01 4.13978845e-01
4.63563859e-01 2.60643601e-01 3.74001682e-01 -9.33811128e-01
-7.38856047e-02 9.39904451e-01 1.18368842e-01 -1.06917822e+00
-3.76758784e-01 1.10778503e-01 -4.83593434e-01 2.70382524e-01
-9.44807529e-02 1.73240732e-02 -1.01961184e+00 1.18300223e+00
5.01056254e-01 3.08300227e-01 5.33681870e-01 5.43838382e-01
6.54456258e-01 1.00720751e+00 5.85035384e-01 -2.35091552e-01
1.80851924e+00 -4.12616462e-01 -8.56233895e-01 6.58032298e-02
1.05497383e-01 -6.58227980e-01 1.26476035e-01 5.51382780e-01
-7.11324632e-01 -8.40971649e-01 -8.49593163e-01 2.44889185e-01
-8.89524460e-01 -5.27854338e-02 5.07130921e-01 3.41218293e-01
-1.03989255e+00 1.11621924e-01 -1.02238846e+00 -1.19536150e+00
-1.69651657e-01 4.66598898e-01 -4.32645530e-01 3.80845040e-01
-9.77410436e-01 9.27460670e-01 1.13503277e+00 2.56576270e-01
-1.33280349e+00 5.36986232e-01 -7.25627244e-01 -9.17993039e-02
2.72783279e-01 -2.52046615e-01 9.18256938e-01 -1.07036495e+00
-1.20683205e+00 7.09692776e-01 -4.65000123e-01 -5.75659573e-01
-2.04164952e-01 7.26249954e-03 -1.11283851e+00 5.33897996e-01
1.82213485e-01 6.37643039e-01 8.04934740e-01 -1.07001531e+00
-1.53777492e+00 -3.89853686e-01 2.75760472e-01 4.59809721e-01
-2.37241611e-01 5.53178906e-01 -2.80815333e-01 1.04506403e-01
1.43628448e-01 -6.26040757e-01 -1.38183236e-01 -5.74117601e-01
1.61589637e-01 -1.75687388e-01 1.07690036e+00 -8.02139342e-01
1.30958843e+00 -2.19187069e+00 -2.18082368e-02 3.64895403e-01
-1.93179041e-01 2.01782361e-01 3.23123336e-01 9.79379952e-01
2.18410611e-01 -2.34875292e-01 2.31665298e-02 1.33099139e-01
-2.57637084e-01 8.25556576e-01 -1.51459038e-01 3.51173505e-02
2.43553191e-01 -1.05193548e-01 -7.67830908e-01 -1.05227423e+00
3.98951918e-01 2.93431401e-01 -2.71895200e-01 6.18169129e-01
-2.36043409e-01 3.14732552e-01 -4.30009842e-01 6.66227281e-01
1.69372872e-01 -3.42577025e-02 6.39055610e-01 -3.05788517e-01
-7.42895722e-01 -2.38684611e-03 -1.32866287e+00 1.20874298e+00
2.55454153e-01 5.62276602e-01 -2.48691410e-01 -1.09126699e+00
7.28015959e-01 9.12205517e-01 2.30786145e-01 -1.22085921e-01
-4.22600172e-02 -1.23267248e-01 -1.79390803e-01 -1.47709000e+00
6.62182868e-01 1.83790952e-01 7.47408867e-02 -4.94493134e-02
3.23205382e-01 5.88943958e-01 4.51453179e-01 -1.64290771e-01
1.13070345e+00 1.18200228e-01 8.71123433e-01 9.42893848e-02
6.92439198e-01 4.41437453e-01 6.66161418e-01 7.38591075e-01
-2.33874559e-01 -2.79255688e-01 4.10719126e-01 -7.93562949e-01
-8.02634776e-01 -8.66188347e-01 -2.60892045e-02 1.18749702e+00
1.86107382e-01 -2.70299196e-01 -4.60441858e-01 -2.03966066e-01
-6.32552028e-01 3.93061280e-01 -5.73287010e-01 2.45777398e-01
-2.64973938e-01 -9.21109855e-01 5.34053266e-01 2.42117867e-01
8.28463197e-01 -1.26310480e+00 -1.34938669e+00 3.68262857e-01
-2.28005111e-01 -1.39389300e+00 6.06537521e-01 2.69122690e-01
-9.12065446e-01 -1.15772891e+00 2.65960723e-01 -1.08802557e+00
6.57141209e-01 2.13022366e-01 6.77756965e-01 3.82429734e-02
-6.21432923e-02 4.14220661e-01 -1.05558348e+00 -6.24977171e-01
-4.94620830e-01 -2.00789392e-01 -1.08235136e-01 4.04230833e-01
8.63748133e-01 -4.08031136e-01 -2.12505370e-01 4.31403816e-01
-1.15768933e+00 1.31776646e-01 8.24380636e-01 1.46767035e-01
4.45432812e-01 6.38210416e-01 -1.41434237e-01 -4.10585016e-01
1.27788544e-01 -7.06360519e-01 -6.26088083e-01 6.41856015e-01
1.93533540e-01 -9.80564430e-02 3.08278114e-01 -3.78707796e-01
-1.10867131e+00 3.95117491e-01 1.78398818e-01 -2.14458611e-02
-9.73811924e-01 6.34132981e-01 -1.02840461e-01 1.02959126e-01
3.64528239e-01 2.68733561e-01 -2.30930686e-01 -3.11385423e-01
-1.22174248e-01 1.08181870e+00 7.31674373e-01 -1.46957770e-01
1.86198965e-01 5.24296105e-01 -5.86698875e-02 -1.34171236e+00
-4.08278704e-01 -8.56158793e-01 -7.56498039e-01 -8.70902419e-01
1.73097074e+00 -7.51904488e-01 -7.34764516e-01 5.78204811e-01
-1.08660114e+00 -5.02913333e-02 3.23469639e-01 8.83170664e-01
-3.59893531e-01 3.39089453e-01 -4.02550727e-01 -1.38104832e+00
2.35926405e-01 -9.64324892e-01 8.77110362e-01 3.28164548e-01
-2.25791246e-01 -8.82261813e-01 2.65840083e-01 4.37015116e-01
-5.87660186e-02 3.76597285e-01 6.33012414e-01 -6.99383080e-01
-4.51128691e-01 -2.90990144e-01 1.40385523e-01 2.66834587e-01
2.52902210e-01 1.55315623e-01 -9.51465070e-01 1.54887468e-01
-2.36480944e-02 -2.11007502e-02 3.12359959e-01 3.69161338e-01
5.89582682e-01 -3.88238102e-01 -4.43109483e-01 2.06314892e-01
1.48194706e+00 1.15828431e+00 8.68522406e-01 5.22522628e-01
3.65179271e-01 6.72810316e-01 6.83048546e-01 3.65438014e-01
5.18984914e-01 5.60416281e-01 5.93727767e-01 6.02241978e-02
3.15355957e-01 1.25057071e-01 6.37883067e-01 6.84878170e-01
-7.61498094e-01 -2.57181823e-01 -1.04671514e+00 5.44294834e-01
-2.06149387e+00 -1.61170554e+00 -1.75947458e-01 2.24348688e+00
2.64625400e-01 8.83624107e-02 3.91438574e-01 1.82119116e-01
9.87528324e-01 -4.20320034e-02 5.24301901e-02 -6.56797290e-01
1.37259457e-02 -5.63979484e-02 5.84601700e-01 5.61072409e-01
-1.48930466e+00 8.09435189e-01 6.51667500e+00 4.18088228e-01
-7.56732643e-01 1.49229556e-01 -2.01579660e-01 3.89765978e-01
6.57396555e-01 2.32500166e-01 -9.59739387e-01 3.78829688e-01
1.24549532e+00 5.78243375e-01 2.74053961e-01 5.47327518e-01
4.77643102e-01 -7.66297638e-01 -7.97406554e-01 5.64908922e-01
1.37318999e-01 -1.10467482e+00 3.48810166e-01 1.85548574e-01
5.10926187e-01 -1.26601785e-01 -5.71637630e-01 -1.62303850e-01
4.59349245e-01 -5.76208949e-01 7.43000925e-01 8.85200262e-01
-4.59793434e-02 -5.27570307e-01 8.79311860e-01 4.06996936e-01
-1.71299195e+00 -3.29202354e-01 -2.46960118e-01 -4.71285462e-01
2.96408385e-01 -2.68992245e-01 -9.50543046e-01 8.29562068e-01
1.08999944e+00 6.24217808e-01 -4.92219180e-01 1.10068524e+00
2.59529594e-02 5.84533036e-01 -3.09441835e-01 -2.81647276e-02
5.07697701e-01 -1.92733973e-01 4.81744200e-01 1.40533066e+00
2.49274999e-01 6.25263095e-01 5.74706018e-01 -1.14931781e-02
8.25140297e-01 -2.20105976e-01 -6.37854755e-01 -6.83376789e-02
6.40072823e-01 9.17258441e-01 -7.77770877e-01 -9.65035737e-01
-5.04832029e-01 6.10996425e-01 -6.08250387e-02 2.28336349e-01
-7.28742778e-01 -3.64549786e-01 5.17033100e-01 -1.03015132e-01
4.76573616e-01 -4.55503941e-01 5.37705898e-01 -6.07426345e-01
-2.55485624e-01 -8.38109970e-01 8.67003143e-01 -9.30756986e-01
-8.39190423e-01 8.40601206e-01 6.66978300e-01 -1.54414976e+00
-4.55728978e-01 -4.47100043e-01 -3.50794286e-01 5.56359470e-01
-9.20478523e-01 -1.26532638e+00 -5.17257631e-01 8.22326243e-01
6.21433318e-01 -1.94847107e-01 1.06006193e+00 1.66550249e-01
-6.38045967e-01 -5.05568266e-01 -3.55297416e-01 2.22652897e-01
1.89206883e-01 -9.19705987e-01 -2.56943017e-01 1.28683293e+00
1.13457583e-01 3.21268678e-01 9.75436509e-01 -9.08779919e-01
-9.99044895e-01 -9.99500096e-01 1.17427373e+00 -2.55345136e-01
5.39679289e-01 -3.06310188e-02 -8.40240479e-01 9.99607921e-01
3.70755583e-01 -5.26466012e-01 9.12183166e-01 -2.78736502e-01
7.39387516e-03 -1.43671766e-01 -1.01202846e+00 2.14440003e-01
7.57740557e-01 -7.06325650e-01 -1.08570111e+00 2.83645570e-01
4.96570356e-02 6.48202673e-02 -9.96137023e-01 3.45551908e-01
7.67750442e-01 -1.20208895e+00 7.71172643e-01 -6.50158823e-01
-1.45725086e-01 -7.70917296e-01 -4.88117099e-01 -4.81100082e-01
-4.74834055e-01 -2.13136479e-01 9.99747217e-02 8.99561465e-01
7.72912577e-02 -5.68906009e-01 4.13609385e-01 2.04986364e-01
-9.23378542e-02 1.22227214e-01 -5.76006949e-01 -8.82102787e-01
-1.29658163e+00 -5.19220948e-01 3.46939296e-01 9.27072585e-01
7.37046748e-02 1.60391559e-03 -4.73635197e-01 1.27646732e+00
8.04863513e-01 -3.66970688e-01 5.64557433e-01 -1.32304144e+00
-2.48444855e-01 5.70965111e-02 -1.23067570e+00 -5.09185135e-01
-4.06242073e-01 -1.57641247e-01 9.17077735e-02 -1.68006051e+00
2.00447589e-01 -4.98305112e-02 -1.50768846e-01 6.08727276e-01
2.47131005e-01 2.57024914e-02 -1.10778622e-01 6.08843446e-01
-8.43790352e-01 -2.19435990e-01 1.07192479e-01 3.11209381e-01
-3.25847536e-01 -1.52741075e-01 3.56622159e-01 9.91560042e-01
8.30717623e-01 -5.48128724e-01 -5.35825193e-01 -3.21117520e-01
1.35803502e-02 1.64835736e-01 6.98356986e-01 -1.61775696e+00
6.73450232e-01 -3.59963447e-01 1.56181023e-01 -7.54872203e-01
4.99911994e-01 -1.13977814e+00 9.84781444e-01 5.79968810e-01
-2.81818241e-01 3.97846252e-01 -3.26045826e-02 4.48923767e-01
-5.38417101e-01 -4.21343744e-01 3.74664724e-01 -3.84239495e-01
-1.65584898e+00 -1.78094685e-01 -1.23709965e+00 -8.21165681e-01
1.30379248e+00 -8.51808429e-01 -7.80010670e-02 -2.52263308e-01
-1.16996408e+00 4.40287963e-02 1.48316815e-01 3.00379276e-01
4.50718135e-01 -9.01241004e-01 -2.52476752e-01 3.02114606e-01
3.26846689e-01 -3.68942469e-01 2.04121307e-01 7.52722383e-01
-9.25188959e-01 4.52224284e-01 -7.40574241e-01 -6.63050294e-01
-1.86635089e+00 3.37872148e-01 1.15233347e-01 -1.43254638e-01
-2.21085906e-01 3.78327996e-01 -1.94346845e-01 3.19888115e-01
5.08509815e-01 -3.16137403e-01 -9.83555555e-01 1.43807814e-01
7.33164907e-01 5.34227729e-01 -1.36082813e-01 -1.09602237e+00
-5.14450729e-01 6.63470447e-01 2.49428511e-01 -3.00510436e-01
1.13517463e+00 -1.80283055e-01 -3.05307865e-01 6.54219329e-01
6.70710266e-01 -6.87469184e-01 -1.03055394e+00 7.63888955e-02
4.77736682e-01 -2.05444291e-01 -6.80845901e-02 -5.20099044e-01
-2.58814275e-01 4.75900829e-01 9.53475356e-01 7.82525420e-01
1.48665881e+00 8.73560924e-03 2.51745075e-01 4.81665939e-01
5.50860047e-01 -1.13739967e+00 -2.54380971e-01 1.48641229e-01
4.60973889e-01 -1.03464139e+00 -1.11073472e-01 -2.97653139e-01
-7.87020028e-01 1.18546236e+00 2.73646593e-01 1.67788655e-01
6.30372822e-01 5.35260886e-02 8.22993144e-02 -6.93486333e-01
-9.89612520e-01 -7.34350383e-01 1.47194982e-01 8.80212069e-01
-5.81133626e-02 1.93850100e-01 -8.78632069e-02 -2.41009519e-01
2.33828530e-01 1.76008299e-01 2.45588914e-01 1.44111717e+00
-1.19506812e+00 -9.37901437e-01 -8.13909292e-01 1.34811416e-01
-1.59088403e-01 2.28284985e-01 -2.21887976e-01 8.30943704e-01
9.47709084e-01 1.39993465e+00 4.72352445e-01 -6.80065691e-01
2.00933740e-01 2.03478307e-01 2.47460797e-01 -4.25435096e-01
-6.03279531e-01 2.38078926e-02 4.72522795e-01 -5.12518704e-01
-1.30179632e+00 -9.63998377e-01 -8.88534188e-01 1.10387959e-01
1.38456106e-01 3.65327954e-01 8.21738780e-01 1.05552793e+00
1.18693911e-01 1.13288745e-01 4.05862510e-01 -7.39144444e-01
2.49225348e-01 -8.38576376e-01 -4.35927272e-01 3.90303403e-01
2.33227953e-01 -6.02667212e-01 -1.17574312e-01 9.32048261e-01] | [8.226293563842773, 0.39561864733695984] |
a0f9bd0f-e33a-4c90-aa25-d00966f75391 | multiple-instance-active-learning-for-object | 2104.02324 | null | https://arxiv.org/abs/2104.02324v1 | https://arxiv.org/pdf/2104.02324v1.pdf | Multiple instance active learning for object detection | Despite the substantial progress of active learning for image recognition, there still lacks an instance-level active learning method specified for object detection. In this paper, we propose Multiple Instance Active Object Detection (MI-AOD), to select the most informative images for detector training by observing instance-level uncertainty. MI-AOD defines an instance uncertainty learning module, which leverages the discrepancy of two adversarial instance classifiers trained on the labeled set to predict instance uncertainty of the unlabeled set. MI-AOD treats unlabeled images as instance bags and feature anchors in images as instances, and estimates the image uncertainty by re-weighting instances in a multiple instance learning (MIL) fashion. Iterative instance uncertainty learning and re-weighting facilitate suppressing noisy instances, toward bridging the gap between instance uncertainty and image-level uncertainty. Experiments validate that MI-AOD sets a solid baseline for instance-level active learning. On commonly used object detection datasets, MI-AOD outperforms state-of-the-art methods with significant margins, particularly when the labeled sets are small. Code is available at https://github.com/yuantn/MI-AOD. | ['Qixiang Ye', 'Xiangyang Ji', 'Songcen Xu', 'Jianzhuang Liu', 'Mengying Fu', 'Fang Wan', 'Tianning Yuan'] | 2021-04-06 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Yuan_Multiple_Instance_Active_Learning_for_Object_Detection_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Yuan_Multiple_Instance_Active_Learning_for_Object_Detection_CVPR_2021_paper.pdf | cvpr-2021-1 | ['active-object-detection'] | ['computer-vision'] | [ 3.58445287e-01 5.10524213e-01 -8.91915262e-01 -6.93594396e-01
-1.92382479e+00 -4.69343424e-01 6.68988049e-01 2.38587573e-01
-3.39689642e-01 5.44018328e-01 -1.41426288e-02 6.34752661e-02
-4.49173152e-02 -6.19054496e-01 -8.52518320e-01 -8.95487428e-01
-5.98953441e-02 6.26058519e-01 1.93963125e-01 5.55275083e-01
1.12185664e-01 3.06906730e-01 -1.29584837e+00 5.56318879e-01
8.19957674e-01 1.28837574e+00 -2.77799815e-01 5.50558448e-01
-2.77836055e-01 1.10031974e+00 -5.56835651e-01 -4.39147055e-01
3.97539884e-01 6.13969676e-02 -5.05892158e-01 3.92314225e-01
7.31545508e-01 -4.62260455e-01 -3.23781520e-01 1.03266096e+00
3.03964496e-01 3.15106437e-02 7.76503384e-01 -1.57384992e+00
-6.30872846e-01 8.80187929e-01 -6.01106226e-01 4.09349412e-01
-5.94631024e-02 5.00843167e-01 1.02032781e+00 -1.53326869e+00
4.35081571e-01 1.28797722e+00 3.97960603e-01 5.81230938e-01
-1.31717706e+00 -8.91514421e-01 6.54972017e-01 1.48379430e-01
-1.28095138e+00 -7.10574865e-01 8.05378854e-01 -4.91902024e-01
5.88661313e-01 3.12794626e-01 2.68836975e-01 1.31040883e+00
-2.29916349e-01 1.53446758e+00 9.18620348e-01 -3.97770882e-01
5.84061444e-01 3.41430664e-01 5.75851381e-01 6.13127351e-01
3.64370763e-01 5.08055568e-01 -7.22127616e-01 -4.77145910e-01
1.70456171e-01 1.78208411e-01 4.32050377e-02 -5.61706364e-01
-8.86019289e-01 8.41178656e-01 6.18566275e-01 -4.08208013e-01
-2.46536538e-01 1.53617188e-01 1.67572588e-01 3.03255934e-02
1.03932726e+00 4.98913229e-01 -2.36709893e-01 3.42781901e-01
-7.96409011e-01 -3.54096591e-02 4.72655445e-01 8.82535815e-01
8.00200403e-01 5.33329882e-02 -6.85768366e-01 6.10005498e-01
7.23449945e-01 5.54008663e-01 -2.44785264e-01 -9.22209740e-01
4.38961923e-01 9.06140983e-01 9.79911909e-02 -4.45120454e-01
1.63027704e-01 -4.72690493e-01 -2.38658145e-01 6.69176936e-01
3.22008520e-01 -7.52974078e-02 -1.30540967e+00 1.37725556e+00
4.07744139e-01 6.62944436e-01 -4.41122837e-02 7.67726004e-01
1.15761220e+00 6.92720294e-01 4.87992346e-01 -2.87496984e-01
8.59494090e-01 -9.10247266e-01 -3.20924848e-01 -6.90872252e-01
2.26148814e-01 -3.05584848e-01 7.20078409e-01 2.38718793e-01
-9.02587831e-01 -3.55402380e-01 -1.25718665e+00 3.20710689e-01
-2.51994729e-01 9.41530615e-02 4.25380737e-01 5.52559316e-01
-2.58296102e-01 5.84025169e-03 -9.08600628e-01 4.14831936e-01
1.27143478e+00 2.19686553e-01 -1.52958348e-01 -2.63638169e-01
-9.55383062e-01 6.05819106e-01 5.48843086e-01 2.51622051e-02
-1.59141862e+00 -1.02694941e+00 -9.56114113e-01 -1.72179386e-01
9.04818892e-01 -1.38374016e-01 1.31970680e+00 -1.02478278e+00
-8.82005453e-01 9.75997448e-01 1.30493924e-01 -6.15608573e-01
6.02026820e-01 -2.57075548e-01 -3.33630651e-01 -4.97095957e-02
1.52360443e-02 9.40092146e-01 1.40219271e+00 -1.89491129e+00
-6.48398042e-01 -5.69418073e-01 1.35036930e-01 1.23098165e-01
2.84321774e-02 -1.16056651e-01 -5.81071854e-01 -5.59485018e-01
3.21350187e-01 -8.19382727e-01 -3.74252260e-01 3.79778653e-01
-5.56369364e-01 -3.50655228e-01 1.11509061e+00 -2.61922032e-02
9.10298228e-01 -2.29847789e+00 -1.75622344e-01 2.06888154e-01
5.07227004e-01 4.03390199e-01 -2.20022574e-01 -1.90691948e-01
1.67224891e-02 4.25431877e-02 -4.72134441e-01 -6.78881466e-01
6.77887499e-02 3.10738236e-01 -7.36602664e-01 3.98279816e-01
1.02944160e+00 1.07393754e+00 -1.15639019e+00 -6.41606271e-01
3.25490624e-01 2.58015275e-01 -7.90695101e-02 4.41421777e-01
-7.81543434e-01 5.25376201e-01 -4.53290343e-01 1.19884896e+00
7.69539833e-01 -2.47154176e-01 -2.25166649e-01 -1.19609058e-01
2.36282244e-01 -9.63693634e-02 -1.18330896e+00 1.44675457e+00
-5.81171997e-02 5.48130095e-01 -2.42181107e-01 -1.07316124e+00
8.37982059e-01 1.52567625e-01 5.49845219e-01 -5.53013146e-01
-1.56811457e-02 1.01954684e-01 -2.95036584e-01 -3.53431731e-01
1.84589833e-01 4.60768878e-01 -2.26251900e-01 2.79865772e-01
2.37658158e-01 -7.25800544e-02 1.37873590e-01 6.39050841e-01
1.00841391e+00 1.27939582e-01 2.49215141e-01 5.02695292e-02
1.81494400e-01 3.27295996e-02 8.59795570e-01 1.36549985e+00
-6.23654962e-01 7.54655182e-01 2.90714771e-01 -2.22871408e-01
-7.17321515e-01 -1.61242914e+00 -5.04780471e-01 1.07484460e+00
2.69753754e-01 -7.90869668e-02 -5.00960112e-01 -1.33756518e+00
3.30049008e-01 7.68296957e-01 -6.79786086e-01 -2.35898674e-01
-1.85005128e-01 -9.55506384e-01 3.33734244e-01 5.42535901e-01
3.91912431e-01 -9.00619507e-01 1.78421140e-02 -5.86240813e-02
1.30980343e-01 -9.26417112e-01 -2.45335877e-01 4.30123419e-01
-7.85241127e-01 -1.10661292e+00 -2.57738143e-01 -3.19437742e-01
8.81643772e-01 5.13495617e-02 1.44732296e+00 -5.84077127e-02
-7.18469083e-01 6.69613838e-01 -3.50434095e-01 -9.99548674e-01
-3.89095932e-01 -3.74879509e-01 -6.60631135e-02 1.97871327e-02
5.85476518e-01 1.79887936e-02 -4.39658791e-01 2.10292503e-01
-8.94350410e-01 -4.96509492e-01 6.29377186e-01 9.09529090e-01
1.09490645e+00 -2.44732723e-01 7.31938004e-01 -1.18906474e+00
-7.02607911e-03 -6.87512815e-01 -9.52037871e-01 4.24480587e-01
-8.53752971e-01 3.51471268e-02 -2.27899313e-01 -4.71704453e-01
-1.24066210e+00 5.45404315e-01 2.57309198e-01 -5.31871319e-01
-2.16283128e-02 1.44026548e-01 -3.27180564e-01 -2.02594162e-03
1.07339060e+00 -2.91252315e-01 -2.77853489e-01 -8.00692104e-03
4.42433298e-01 6.69032097e-01 4.91029739e-01 -5.75587392e-01
1.03662074e+00 4.70177919e-01 -4.09482569e-01 -2.72031933e-01
-1.66272986e+00 -7.48000622e-01 -5.32364726e-01 -3.68289411e-01
5.21325707e-01 -1.27153730e+00 -2.22188622e-01 3.06077451e-01
-7.48387992e-01 -1.96123153e-01 -8.66886139e-01 5.43256462e-01
-3.72684270e-01 -3.16035971e-02 -2.10076779e-01 -1.20948279e+00
-2.94566065e-01 -1.39359355e+00 1.33261597e+00 2.04906851e-01
7.19660074e-02 -7.10463881e-01 8.72900859e-02 8.00616264e-01
-5.45442477e-02 1.34717673e-01 3.91387463e-01 -8.11208427e-01
-1.28055167e+00 -5.02278984e-01 -2.40948975e-01 4.65696424e-01
-2.60378331e-01 1.07867911e-01 -1.58987439e+00 -2.08964482e-01
-3.31211329e-01 -9.41909373e-01 1.43109357e+00 5.77597976e-01
1.45386875e+00 -1.83216363e-01 -3.34273279e-01 3.39832813e-01
1.31688416e+00 1.30541861e-01 7.48374999e-01 6.90679252e-02
4.18112934e-01 1.41566366e-01 1.09295809e+00 4.25736874e-01
-1.20393030e-01 3.82831067e-01 8.52024317e-01 -1.89272761e-01
-1.83825061e-01 -1.01796113e-01 3.99113804e-01 1.58845950e-02
4.23857510e-01 -3.25495481e-01 -1.04225957e+00 3.48591536e-01
-1.96547627e+00 -1.13603938e+00 2.20687121e-01 2.36339259e+00
9.88912106e-01 6.64164245e-01 -1.18565178e-02 8.14580619e-02
7.40829527e-01 1.96071103e-01 -1.08340645e+00 2.38463596e-01
-2.65257508e-02 -1.11683890e-01 3.94718498e-01 7.23994374e-01
-1.69960034e+00 5.13124049e-01 5.66717339e+00 9.71168578e-01
-7.04107881e-01 2.62374192e-01 1.05922806e+00 -3.47846895e-01
-1.72106683e-01 1.52544692e-01 -1.06378531e+00 4.20306236e-01
4.13629442e-01 8.93835649e-02 -2.67288387e-01 1.20550382e+00
-1.78951636e-01 -2.02467114e-01 -1.22753239e+00 8.41992080e-01
1.03796214e-01 -1.50609803e+00 -8.40028822e-02 -1.61901221e-01
1.09374797e+00 1.82181165e-01 4.22909856e-01 6.66132629e-01
3.80436033e-01 -8.14216554e-01 6.75307870e-01 6.57011092e-01
5.18205464e-01 -5.31356931e-01 5.87372601e-01 4.35327709e-01
-7.99998879e-01 -4.45152760e-01 -3.36649656e-01 6.15936041e-01
-2.27413833e-01 9.93416190e-01 -1.08315134e+00 1.26974434e-01
5.96817136e-01 5.06294012e-01 -7.05553055e-01 1.22030950e+00
-2.43908808e-01 1.16600811e+00 -2.39973620e-01 3.56507093e-01
8.08384493e-02 1.39148831e-01 9.00623500e-01 1.15501332e+00
-2.22604871e-01 1.90893903e-01 6.10903203e-01 1.15928352e+00
-4.37000155e-01 -3.65896344e-01 -4.08212841e-01 -1.85244039e-01
7.15394557e-01 1.18736088e+00 -5.73127270e-01 -4.07582968e-01
-2.27011338e-01 5.79666555e-01 2.88334548e-01 3.61506790e-01
-8.99550915e-01 7.51037300e-02 4.00904894e-01 -1.81310534e-01
-5.97906895e-02 1.45018309e-01 -3.20713013e-01 -1.11621130e+00
-8.15022588e-02 -6.92311227e-01 6.38602316e-01 -6.22819006e-01
-1.69207323e+00 2.45508298e-01 1.35347709e-01 -1.52945125e+00
-1.48730606e-01 -6.06833220e-01 -7.55804598e-01 4.57851738e-01
-1.29868448e+00 -1.24781370e+00 -4.20769811e-01 2.99952984e-01
9.02465165e-01 -4.34628010e-01 7.97567964e-01 -3.64270690e-03
-8.07843626e-01 8.55211437e-01 2.23681718e-01 4.45348203e-01
7.08780169e-01 -1.39602745e+00 2.95863211e-01 1.03832185e+00
6.70590639e-01 3.15462530e-01 4.19118881e-01 -6.99003398e-01
-1.27149045e+00 -1.25453222e+00 2.32028738e-01 -1.04747701e+00
5.18052340e-01 -5.51797211e-01 -8.33042622e-01 7.52204597e-01
-3.10885757e-01 7.45092332e-01 6.68048441e-01 1.13152705e-01
-8.89208019e-01 -3.57301742e-01 -1.39373815e+00 2.83669919e-01
8.53933632e-01 -7.15245485e-01 -5.01380444e-01 5.91907442e-01
8.34643245e-01 -4.46826845e-01 -6.10464156e-01 7.23370433e-01
1.24987423e-01 -6.84579313e-01 1.24793029e+00 -7.40143061e-01
3.28277260e-01 -2.91618556e-01 -2.55201787e-01 -1.06417561e+00
-3.90840679e-01 -8.88302624e-02 -7.81057298e-01 1.31121111e+00
8.79956305e-01 -3.24222982e-01 1.05849183e+00 8.36361945e-01
-2.57827282e-01 -9.16893065e-01 -9.17983770e-01 -6.62835717e-01
-3.04815829e-01 -7.38129735e-01 2.02841684e-01 7.07702458e-01
-3.29704016e-01 1.05307057e-01 -1.07512459e-01 5.19689858e-01
1.34605217e+00 -2.16061082e-02 4.03917879e-01 -1.29751337e+00
-3.28725696e-01 6.16550595e-02 -5.93029916e-01 -4.90479559e-01
3.43731791e-01 -8.14847469e-01 4.13490504e-01 -1.25222826e+00
5.04889667e-01 -8.72646391e-01 -7.01394558e-01 6.19900346e-01
-5.97243071e-01 5.86955428e-01 3.84863466e-02 4.76681650e-01
-1.09304726e+00 3.84072274e-01 7.70068467e-01 -8.99767995e-01
-9.09076333e-02 4.05215919e-01 -5.35687089e-01 6.30574763e-01
6.66614234e-01 -7.15540588e-01 -6.63627148e-01 -1.09036960e-01
3.18632536e-02 -2.38697723e-01 5.74495196e-01 -8.87058496e-01
2.05138698e-01 -2.14343563e-01 7.47271836e-01 -7.91972578e-01
5.78511119e-01 -8.30498695e-01 -2.07092524e-01 2.48230711e-01
-8.94389570e-01 -8.38957846e-01 9.47768539e-02 1.02696657e+00
-2.02246770e-01 -4.97705847e-01 9.97851372e-01 -1.22730181e-01
-1.14262736e+00 8.46919417e-01 -5.47172762e-02 2.99672157e-01
1.11058998e+00 -1.68603584e-01 -6.50324821e-01 4.37013358e-02
-9.90736127e-01 3.50807130e-01 8.24952405e-03 4.55377847e-01
7.21798182e-01 -1.27906096e+00 -9.20855403e-01 1.38889611e-01
8.60665143e-01 3.55547100e-01 2.12860733e-01 5.43169439e-01
2.15985328e-01 8.11495706e-02 4.04324681e-01 -1.18523109e+00
-1.21284640e+00 4.43027020e-01 3.88104558e-01 1.37067288e-02
-2.74881840e-01 1.31927943e+00 1.34695172e-01 -4.85048115e-01
7.43946970e-01 2.21559688e-01 1.33523971e-01 2.50662088e-01
6.76104844e-01 3.71463925e-01 2.46152773e-01 -2.43273079e-01
-2.85183817e-01 -1.33508131e-01 -3.31574351e-01 7.34497467e-03
1.14542890e+00 1.07935004e-01 2.28211477e-01 6.21096194e-01
1.03118145e+00 -9.47068706e-02 -1.85154116e+00 -6.93464100e-01
1.71854243e-01 -7.01730251e-01 3.85896116e-01 -1.00051749e+00
-1.01736748e+00 5.57612062e-01 1.11302960e+00 -1.37743250e-01
8.79980624e-01 3.04898471e-01 3.45031805e-02 6.62454903e-01
2.51528621e-01 -1.10921896e+00 5.80166876e-01 2.27791995e-01
8.58497679e-01 -1.97822857e+00 2.86761880e-01 -4.76365149e-01
-7.57836938e-01 6.02758110e-01 9.07795310e-01 -1.20385997e-01
7.51028419e-01 6.48530126e-01 2.38032252e-01 -1.38628364e-01
-8.92844379e-01 -2.66411126e-01 5.88721037e-01 6.15153372e-01
1.57333553e-01 1.19490805e-03 4.27234888e-01 3.80902261e-01
8.08953285e-01 -7.23061189e-02 3.46966349e-02 8.94271076e-01
-6.97430968e-01 -8.61309350e-01 -5.40231526e-01 7.46231794e-01
-2.36666039e-01 -2.09925622e-02 -4.29504186e-01 4.45384264e-01
4.10275608e-01 1.02575922e+00 1.96765244e-01 -9.47466493e-02
1.21216089e-01 8.53062496e-02 6.52965844e-01 -9.40000057e-01
-3.94385517e-01 -3.73773664e-01 1.78430721e-01 -4.77540404e-01
-3.16965431e-01 -6.10337496e-01 -1.23677886e+00 3.89449775e-01
-7.25701928e-01 -6.81595597e-03 4.70032543e-01 7.96252012e-01
2.84662247e-01 1.73043340e-01 8.89427900e-01 -7.10665166e-01
-9.90051687e-01 -8.60376239e-01 -3.69242698e-01 3.83066654e-01
4.03374940e-01 -6.73773944e-01 -7.33144701e-01 -8.79617035e-03] | [9.273008346557617, 1.305159330368042] |
0207a487-738b-458d-aabc-74a06338e609 | neural-machine-translation-with-heterogeneous | null | null | https://aclanthology.org/2021.emnlp-main.256 | https://aclanthology.org/2021.emnlp-main.256.pdf | Neural Machine Translation with Heterogeneous Topic Knowledge Embeddings | Neural Machine Translation (NMT) has shown a strong ability to utilize local context to disambiguate the meaning of words. However, it remains a challenge for NMT to leverage broader context information like topics. In this paper, we propose heterogeneous ways of embedding topic information at the sentence level into an NMT model to improve translation performance. Specifically, the topic information can be incorporated as pre-encoder topic embedding, post-encoder topic embedding, and decoder topic embedding to increase the likelihood of selecting target words from the same topic of the source sentence. Experimental results show that NMT models with the proposed topic knowledge embedding outperform the baselines on the English -> German and English -> French translation tasks. | ['Qun Liu', 'Meng Zhang', 'Wei Peng', 'Weixuan Wang'] | null | null | null | null | emnlp-2021-11 | ['topic-models'] | ['natural-language-processing'] | [ 1.95206940e-01 2.12649614e-01 -7.69502759e-01 -4.71472085e-01
-1.23450613e+00 -4.18542355e-01 9.44199264e-01 -7.44862184e-02
-2.15734705e-01 9.79709685e-01 8.73531282e-01 -5.95770419e-01
2.09969729e-01 -6.21734262e-01 -7.51141727e-01 -6.00891292e-01
3.74669671e-01 5.49379945e-01 -3.14841926e-01 -1.28559992e-01
1.24770202e-01 -3.71189803e-01 -4.02635932e-01 4.45586473e-01
1.17097807e+00 4.10818070e-01 5.82022965e-01 -5.56891970e-02
-4.48435366e-01 -5.11720106e-02 -7.38130450e-01 -4.96492237e-01
1.45598292e-01 -6.04445815e-01 -4.57230031e-01 -7.00458214e-02
4.56254244e-01 -7.75977522e-02 -3.29922676e-01 8.79934788e-01
5.63670397e-01 -2.54421774e-02 7.57214785e-01 -8.65154207e-01
-1.20828426e+00 1.10382938e+00 -5.99846423e-01 2.98369855e-01
9.93071347e-02 -3.65715027e-02 1.23438728e+00 -1.32450819e+00
7.88364112e-01 1.53831553e+00 2.00281709e-01 4.84429479e-01
-1.24444497e+00 -6.52731121e-01 3.93472791e-01 1.08004339e-01
-1.06087029e+00 -1.78682774e-01 7.60195434e-01 -1.81807518e-01
1.22317219e+00 -1.24636531e-01 5.29155791e-01 1.58556855e+00
6.67201817e-01 9.70314980e-01 1.19274735e+00 -5.06957591e-01
-5.78878559e-02 3.64098728e-01 1.95255801e-02 2.22775683e-01
1.31317988e-01 -2.18907148e-01 -9.77439642e-01 -2.51208037e-01
5.67716956e-01 -1.25815496e-01 -2.20443711e-01 -1.85094755e-02
-1.80161285e+00 1.15972507e+00 3.61755610e-01 3.05850238e-01
-6.04207337e-01 1.65867761e-01 4.33578640e-01 3.18130016e-01
9.91265714e-01 8.13363075e-01 -8.18097651e-01 -1.81162909e-01
-9.01593268e-01 7.05527738e-02 5.64745069e-01 1.18441951e+00
8.60229790e-01 -2.84906197e-02 -6.93797171e-01 8.70064616e-01
2.69044429e-01 8.80215943e-01 7.23117888e-01 -4.20624852e-01
8.65107536e-01 4.33669657e-01 -1.59839928e-01 -5.84568501e-01
4.36398745e-01 -5.86354315e-01 -3.36161166e-01 -6.73342824e-01
-3.38803351e-01 -5.11792660e-01 -1.00007761e+00 1.97289634e+00
8.21865052e-02 1.71427580e-03 4.36939210e-01 7.72194326e-01
4.92498696e-01 1.35253000e+00 1.27465725e-01 -3.52462292e-01
1.41804612e+00 -1.22261441e+00 -1.11697638e+00 -5.64910948e-01
5.24163306e-01 -1.12864089e+00 1.12409866e+00 -2.87932992e-01
-1.01650059e+00 -3.69467616e-01 -6.90299690e-01 -2.50033468e-01
-3.67452711e-01 4.39839482e-01 5.66352963e-01 2.12118059e-01
-9.55680311e-01 -9.45425555e-02 -7.78522551e-01 -5.71938753e-01
1.79408029e-01 2.14081630e-01 -1.41485140e-01 -2.69232541e-01
-1.63301468e+00 9.71017957e-01 4.73679870e-01 -2.41341665e-01
-7.66714394e-01 -8.76651287e-01 -8.15403879e-01 1.67165786e-01
1.16104126e-01 -1.09710062e+00 1.12098265e+00 -6.99094236e-01
-1.62719893e+00 3.45749497e-01 -7.65670657e-01 -5.21954715e-01
-3.30385678e-02 -4.70718652e-01 -9.34779495e-02 3.78057659e-02
5.87516427e-01 1.03908479e+00 9.06130314e-01 -8.24325860e-01
-6.91709936e-01 -1.55467108e-01 -2.82025784e-01 5.45934975e-01
-6.73617482e-01 1.38517946e-01 -4.59922999e-01 -1.03567779e+00
-1.59628354e-02 -1.11290646e+00 8.45343024e-02 -4.06729966e-01
-4.36795235e-01 -7.35265911e-01 1.10052156e+00 -9.07309651e-01
1.24599922e+00 -2.20044947e+00 5.30137539e-01 -2.83031136e-01
-3.59589368e-01 -6.05858006e-02 -1.23935111e-01 7.52456129e-01
3.00750315e-01 2.40680099e-01 -8.09450150e-02 -2.42185026e-01
1.26838923e-01 2.72800118e-01 -8.99299622e-01 -1.39127225e-01
5.22191763e-01 1.21031821e+00 -9.78440166e-01 -4.41127092e-01
-1.15214966e-01 6.13870680e-01 -3.70072424e-01 -8.76374990e-02
-5.39061904e-01 2.52822936e-01 -6.75763190e-01 3.70490789e-01
1.24786258e-01 -9.06415805e-02 1.26976088e-01 1.29354879e-01
2.02174529e-01 1.25498414e+00 -2.09468514e-01 1.93743944e+00
-6.41864002e-01 9.63952482e-01 -4.07409936e-01 -4.79908407e-01
9.00764346e-01 7.83473611e-01 1.15558036e-01 -3.70202363e-01
-8.36794823e-02 3.72873217e-01 7.41558075e-02 -2.30552569e-01
7.12732017e-01 -3.17814648e-01 -2.19717130e-01 6.93317115e-01
1.80143490e-01 -1.53343290e-01 -6.28894642e-02 1.20109193e-01
6.95009470e-01 1.01718262e-01 3.59137007e-03 -4.59974915e-01
-1.15373328e-01 2.35073790e-01 5.65546930e-01 3.62256885e-01
1.85816005e-01 4.26212728e-01 2.95162976e-01 8.48176554e-02
-1.05875158e+00 -1.16749561e+00 -1.54767111e-01 1.14968002e+00
-8.51268470e-02 -5.52068830e-01 -4.84745890e-01 -8.36287618e-01
1.46425897e-02 1.24555695e+00 -4.82055068e-01 -3.51795524e-01
-8.05804729e-01 -7.25719929e-01 1.32795841e-01 5.25873423e-01
4.04012084e-01 -8.17575455e-01 3.79798040e-02 3.81958514e-01
-7.51548052e-01 -1.08308685e+00 -1.02516258e+00 1.86440602e-01
-1.28372991e+00 -1.99241471e-02 -8.69068265e-01 -1.00411856e+00
6.64990306e-01 4.62108016e-01 9.33119893e-01 -6.41946256e-01
4.03814077e-01 2.81610116e-02 -3.02836686e-01 -4.26034033e-01
-3.23280483e-01 6.81152463e-01 4.04782370e-02 -3.46823633e-01
6.55175626e-01 -4.67211396e-01 -5.57861626e-01 1.90478802e-01
-5.78756750e-01 2.77234286e-01 9.33541298e-01 9.94372487e-01
4.32733357e-01 -4.02280658e-01 8.26507032e-01 -6.23703241e-01
1.02684164e+00 -6.38129115e-01 5.57408333e-02 4.02347803e-01
-7.41747439e-01 1.90187261e-01 2.15588272e-01 -7.42360711e-01
-1.07978141e+00 -5.64758778e-01 2.82040030e-01 -1.65294319e-01
2.55498856e-01 7.88794994e-01 -2.50829518e-01 6.27302051e-01
2.86054909e-01 5.48883140e-01 -2.49821946e-01 -4.00928110e-01
5.50592124e-01 6.61088526e-01 5.03837317e-02 -7.99534857e-01
5.79328060e-01 -2.49540601e-02 -5.68018496e-01 -5.15001476e-01
-6.67408884e-01 -3.91420662e-01 -3.90948981e-01 2.67138958e-01
9.84835327e-01 -1.28886557e+00 5.56290627e-01 -2.20002368e-01
-1.45856404e+00 -4.50108983e-02 -1.13673925e-01 8.64653766e-01
-2.90397584e-01 -4.55239974e-02 -7.12016225e-01 -2.60169685e-01
-5.94247639e-01 -1.34522331e+00 1.48090672e+00 1.26042694e-01
-3.77098501e-01 -1.21212101e+00 -1.28523201e-01 2.38956958e-01
5.37848353e-01 -2.95229286e-01 1.41492474e+00 -6.35190845e-01
-7.83923924e-01 -7.88258687e-02 -1.56339750e-01 1.34933427e-01
4.11119610e-01 -3.45993459e-01 -5.85636735e-01 -4.72902983e-01
1.16578182e-02 9.28470045e-02 1.02306163e+00 4.05683875e-01
2.66619235e-01 -5.45540094e-01 -6.41552567e-01 2.65890568e-01
1.18540120e+00 1.49426147e-01 3.72863650e-01 3.28318059e-01
4.89213228e-01 3.03375036e-01 6.11521065e-01 -2.56618291e-01
5.25441349e-01 8.26052070e-01 -3.47601563e-01 4.56892289e-02
-3.65424067e-01 -5.87058008e-01 9.57148135e-01 1.29732752e+00
3.62901151e-01 -3.43142509e-01 -7.48681784e-01 8.89672697e-01
-1.81819689e+00 -5.61794639e-01 3.40955883e-01 1.72136688e+00
1.21961915e+00 6.84138834e-02 -3.64955306e-01 -7.15998173e-01
8.02281141e-01 1.36725649e-01 -3.55667382e-01 -5.00055313e-01
-1.75738320e-01 -1.45058200e-01 1.64116457e-01 5.27262211e-01
-7.01435566e-01 1.54300010e+00 6.25809002e+00 8.33470523e-01
-1.24502337e+00 6.28082693e-01 3.11310768e-01 -1.01830065e-01
-5.44187486e-01 2.75382280e-01 -1.10801387e+00 5.13588011e-01
9.36301529e-01 -5.88075876e-01 -2.30242573e-02 7.32483208e-01
3.74376744e-01 2.09275723e-01 -1.03886127e+00 4.00385320e-01
2.27824762e-01 -1.09711099e+00 5.69448888e-01 4.22072679e-01
1.18846059e+00 2.11366042e-01 3.61060292e-01 6.19827807e-01
2.61332244e-01 -5.87509573e-01 4.46731418e-01 1.15924083e-01
6.12084210e-01 -4.51930404e-01 6.44993067e-01 2.85789073e-01
-6.71072483e-01 2.42448643e-01 -6.70040429e-01 1.47735119e-01
3.22841138e-01 6.25563622e-01 -1.49087310e+00 5.60007095e-01
2.46315613e-01 7.88860083e-01 -3.25057030e-01 6.38066471e-01
-8.08582664e-01 1.06840837e+00 -1.47056431e-01 -1.62650958e-01
5.80998302e-01 -1.78291172e-01 8.51135910e-01 1.28583682e+00
6.62831724e-01 -2.63036221e-01 2.31854364e-01 7.58653998e-01
-2.46816233e-01 3.51286322e-01 -6.44134879e-01 -2.79149592e-01
4.84612703e-01 7.11124837e-01 -3.26269388e-01 -5.46278417e-01
-5.48842609e-01 1.16385067e+00 2.08422735e-01 6.90489888e-01
-5.72073340e-01 -1.16621047e-01 1.01869977e+00 -1.24597829e-02
4.03457105e-01 -5.74236274e-01 -4.82622594e-01 -1.41351199e+00
1.60195038e-01 -6.71804488e-01 1.35306909e-03 -8.58887672e-01
-1.22151983e+00 6.43413305e-01 1.75686687e-01 -9.42806184e-01
-4.52228963e-01 -3.52578878e-01 -7.00840890e-01 1.29847419e+00
-1.63506138e+00 -1.36580753e+00 6.96030438e-01 7.70165324e-02
1.15697551e+00 -3.67154539e-01 8.07808816e-01 -1.08093023e-01
-4.23574537e-01 5.93703449e-01 4.21322286e-01 4.77965735e-02
1.22789502e+00 -1.23739803e+00 6.28354013e-01 9.21499133e-01
2.61590809e-01 1.14851570e+00 6.77427888e-01 -1.08413374e+00
-1.34367239e+00 -1.31036019e+00 1.67908895e+00 -7.16902614e-01
5.94498217e-01 -5.88909209e-01 -8.06387842e-01 8.25633228e-01
9.34586525e-01 -7.99704313e-01 8.44182968e-01 5.26882231e-01
-3.77317637e-01 1.51568800e-01 -4.57676619e-01 8.44842196e-01
6.38120651e-01 -8.24616253e-01 -1.17344511e+00 4.91613477e-01
1.54942989e+00 -1.20103002e-01 -8.35869253e-01 1.47217706e-01
2.46881247e-01 2.14359730e-01 8.15810144e-01 -6.21939301e-01
7.19184220e-01 -1.18712306e-01 -2.81341404e-01 -1.86550820e+00
-2.02399269e-01 -6.91774607e-01 -2.07465053e-01 1.23076260e+00
1.04592836e+00 -6.09956563e-01 3.59272957e-01 2.99938172e-01
-3.49374533e-01 -8.85833740e-01 -1.15069199e+00 -7.42297471e-01
5.84423661e-01 -2.48264428e-02 6.35205686e-01 9.34962630e-01
3.60570848e-01 1.00127339e+00 -2.58918494e-01 1.32598981e-01
2.95314193e-01 3.35159451e-01 4.55480427e-01 -8.05009842e-01
-9.59581435e-02 -4.20816332e-01 -8.65979791e-02 -1.36866963e+00
3.61610532e-01 -1.16409385e+00 9.95651186e-02 -1.94459987e+00
4.22685057e-01 1.77303981e-02 -2.26857901e-01 5.97449064e-01
-7.54419982e-01 -1.30213618e-01 1.12386189e-01 4.04082298e-01
-3.32571089e-01 1.05212140e+00 1.40615892e+00 -2.20327854e-01
-5.70434183e-02 -1.49210706e-01 -8.09368789e-01 7.04867542e-02
9.21671808e-01 -7.35657036e-01 -3.96591693e-01 -1.10470557e+00
1.75515249e-01 8.89066756e-02 -8.24165493e-02 -3.82886708e-01
1.22979663e-01 -1.73453540e-01 1.89784542e-01 -6.64129734e-01
3.35069805e-01 -5.69949329e-01 -4.58462983e-01 1.81179389e-01
-6.82543576e-01 4.26476479e-01 1.79131672e-01 7.08661973e-01
-4.04136181e-01 9.57528949e-02 1.98758945e-01 7.23131932e-03
-2.66838670e-01 2.05192402e-01 -4.66804504e-01 -4.96225357e-02
6.19402885e-01 1.67243674e-01 -3.00630778e-01 -3.96072894e-01
-3.85233968e-01 3.44586819e-01 4.31985348e-01 1.04702222e+00
5.19984543e-01 -1.57634902e+00 -9.52919543e-01 1.27014861e-01
2.11224124e-01 -3.15762967e-01 -1.97223693e-01 7.56478786e-01
4.39892143e-01 9.77622867e-01 1.22368313e-01 -5.80115736e-01
-9.25515413e-01 4.13827628e-01 -1.42614916e-01 -2.13042155e-01
-5.80427051e-01 6.30162716e-01 4.89725679e-01 -3.86318088e-01
-2.56103054e-02 -4.03653532e-01 1.93971723e-01 -2.02170253e-01
3.33203107e-01 -2.20002919e-01 -1.31039873e-01 -4.69079196e-01
-7.22260028e-02 3.45958829e-01 -5.43804884e-01 -5.48935056e-01
1.11343277e+00 -4.14559931e-01 -2.12840706e-01 6.03083014e-01
1.23835361e+00 -8.30259025e-02 -8.95961702e-01 -7.16425955e-01
2.06861973e-01 -3.43120366e-01 1.82179675e-01 -1.13196456e+00
-4.91150975e-01 1.12948179e+00 1.74457312e-01 -2.13196918e-01
8.65957499e-01 2.56220579e-01 1.13706315e+00 4.33264732e-01
3.02058548e-01 -1.01945031e+00 4.00316389e-03 7.96168089e-01
8.63712609e-01 -1.27056670e+00 -3.06432605e-01 -3.83186817e-01
-7.01123238e-01 1.06864715e+00 6.02548897e-01 1.65560812e-01
4.83994603e-01 -1.56074628e-01 2.17153117e-01 1.42226252e-03
-1.28935516e+00 1.17275611e-01 5.51721692e-01 2.42469788e-01
8.30023229e-01 2.74089277e-01 -7.70662010e-01 5.36123812e-01
-3.00042957e-01 -3.62135828e-01 1.68560103e-01 6.85269773e-01
-4.84758615e-01 -1.40665412e+00 -1.51689127e-01 3.71430248e-01
-5.42468011e-01 -7.17682064e-01 -5.24584115e-01 4.79380012e-01
-9.07878354e-02 8.58376741e-01 -1.85493961e-01 -2.94793934e-01
-4.56849076e-02 7.15962827e-01 1.73018575e-01 -1.17120361e+00
-5.17336369e-01 6.29512131e-01 1.30697349e-02 1.71689019e-02
-1.14239104e-01 -9.55020547e-01 -9.09336567e-01 2.00085863e-01
-5.14971137e-01 5.31126022e-01 1.01610684e+00 1.00946641e+00
7.34929800e-01 4.84329641e-01 4.93937641e-01 -3.29389423e-01
-5.00929832e-01 -1.61111605e+00 3.11282426e-02 -2.54468936e-02
1.02594830e-01 -4.34085250e-01 -2.24421278e-01 8.50889981e-02] | [11.615428924560547, 10.076553344726562] |
c4dc7396-6220-4773-b2c3-47a06a97b4cd | co-existence-of-micro-pico-and-atto-cells-in | 1906.08212 | null | http://arxiv.org/abs/1906.08212v2 | http://arxiv.org/pdf/1906.08212v2.pdf | Co-existence of Micro, Pico and Atto Cells in Optical Wireless Communication | Interference between cells or users can have a significant impact on the
quality of optical wireless communication (OWC) links. This paper studies the
co-existence of infrared based Micro cells with Visible light communication
based Pico and Atto Cells for downlink communication. The signal to noise ratio
(SNR) of each cell and the signal to interference and noise ratio (SINR)
between cells are evaluated when an angle diversity receiver and an imaging
receiver are used. The results show that Atto cell systems provide higher SNR
and data rates compared to the Pico cell systems which provide higher SNR and
data rates compared to Micro cell systems. The Micro cell systems however
provide mobility as they provide a larger coverage area, followed by the Pico
cell systems. | [] | 2019-10-15 | null | null | null | null | ['pico'] | ['natural-language-processing'] | [-1.98199272e-01 1.71842426e-01 1.63654327e-01 5.10244608e-01
1.78915471e-01 -5.58473647e-01 4.60159808e-01 -2.17116252e-01
-5.91096818e-01 1.49282312e+00 -4.59234454e-02 -3.32499921e-01
1.78228930e-01 -8.40175509e-01 8.61332044e-02 -1.45048702e+00
-1.87651575e-01 8.57473612e-02 1.18992202e-01 -9.63771716e-02
3.05764992e-02 6.45294309e-01 -9.53005433e-01 -4.15161282e-01
6.02048576e-01 1.00549042e+00 2.36596808e-01 8.63126040e-01
1.84557021e-01 3.89420092e-01 -8.33331466e-01 1.79211348e-01
8.69140849e-02 -8.11372772e-02 2.59955466e-01 -3.63489181e-01
9.36447158e-02 -2.76849896e-01 -2.37379834e-01 3.75088841e-01
1.02729273e+00 -3.31256270e-01 8.74721825e-01 -1.17577887e+00
-1.60263911e-01 -3.32307294e-02 -5.50631702e-01 1.80162832e-01
6.04432523e-01 1.98680788e-01 1.99464157e-01 -6.71171725e-01
4.31748569e-01 9.33703601e-01 5.80941558e-01 2.67859906e-01
-7.01806605e-01 -9.59850788e-01 -9.17031646e-01 -5.02104126e-02
-1.46324599e+00 -5.37989676e-01 -1.18678555e-01 -1.43545315e-01
9.29366112e-01 3.64342868e-01 1.22418463e+00 3.77810270e-01
1.34411812e+00 -2.88042873e-01 1.07108498e+00 -6.18474960e-01
3.26940536e-01 4.06138003e-01 -1.42685607e-01 6.29338086e-01
1.25854385e+00 2.58316010e-01 -2.36006185e-01 1.66419908e-01
6.41052127e-01 -1.27297223e-01 -6.46476626e-01 3.37718248e-01
-8.77118886e-01 -8.19783881e-02 5.98656714e-01 7.53212810e-01
-3.40634674e-01 6.60371125e-01 -6.47308171e-01 4.98496920e-01
3.61974426e-02 4.98470426e-01 2.00836971e-01 9.51990336e-02
-4.69474941e-01 -6.00580513e-01 9.19977903e-01 1.28359950e+00
2.59094417e-01 -1.76416129e-01 -5.33999801e-01 4.63326395e-01
9.70863998e-01 1.22429752e+00 -1.56580303e-02 -7.87204444e-01
3.07925969e-01 1.54402629e-01 6.60797417e-01 -3.09576362e-01
-6.38458848e-01 -1.21468759e+00 -5.65467477e-01 4.51259375e-01
3.60241294e-01 -9.08234060e-01 -7.27484345e-01 9.36536908e-01
3.18956040e-02 -4.45916504e-02 8.54298413e-01 5.79328001e-01
7.47516632e-01 8.56977046e-01 -4.86393631e-01 -9.46534991e-01
1.35453439e+00 -8.34683180e-01 -9.57917929e-01 -1.98043540e-01
6.06585860e-01 -1.09626138e+00 -1.40558302e-01 3.22370119e-02
-1.09836829e+00 -1.39520718e-02 -1.39158213e+00 2.88281024e-01
-5.08687757e-02 8.01930483e-03 1.06430249e-02 1.10513163e+00
-9.39914942e-01 -3.69567603e-01 -2.29681104e-01 -9.31983769e-01
4.28084612e-01 4.32872862e-01 2.89750844e-01 -3.21047127e-01
-3.04571778e-01 6.48049712e-01 -2.18472287e-01 -3.01811576e-01
-5.86195849e-02 -4.68277067e-01 -1.47775501e-01 5.66490181e-02
1.10667452e-01 -9.95090008e-01 8.07553232e-01 -1.22777693e-01
-1.55169797e+00 3.21178764e-01 -3.49169970e-01 -5.14050007e-01
1.03039995e-01 -3.72547321e-02 -5.80369592e-01 5.06106496e-01
-3.62989515e-01 4.45681036e-01 -2.92579800e-01 -1.58613145e+00
-9.51519191e-01 -2.16299295e-02 4.18506488e-02 2.74278194e-01
2.67860919e-01 -6.49622157e-02 -4.24548723e-02 6.96609020e-02
9.33134630e-02 -1.02044237e+00 -9.19823721e-03 -3.95859256e-02
-3.99908215e-01 -7.14407936e-02 1.24006438e+00 4.06625032e-01
8.26637685e-01 -1.98136556e+00 -5.56120872e-01 5.94691336e-02
1.09521881e-01 3.29475552e-01 -4.20283899e-03 8.92721593e-01
7.99650192e-01 -2.00756356e-01 3.62671435e-01 4.62438576e-02
-9.46537375e-01 1.41993120e-01 5.49221694e-01 6.69607222e-01
-8.21810663e-01 4.90078628e-01 -8.10743034e-01 -1.91625401e-01
4.72532436e-02 5.87919354e-01 -9.75052938e-02 -1.07309885e-01
3.88949186e-01 7.40961611e-01 -5.27746558e-01 8.91769111e-01
8.53829980e-01 3.69285673e-01 -1.89092278e-01 4.36372682e-02
-8.09294522e-01 -2.51321167e-01 -8.11486781e-01 4.38166559e-01
-5.35443366e-01 7.97272265e-01 4.07977462e-01 -2.42879510e-01
9.33105648e-01 8.33594203e-01 6.14586055e-01 -7.46401250e-01
2.09045380e-01 6.55565917e-01 7.38501921e-02 -6.57561123e-01
-2.74906736e-02 -6.86952472e-01 3.72539431e-01 9.91804153e-02
-2.98060834e-01 4.31011468e-02 7.71872774e-02 -5.39001040e-02
9.99446332e-01 -5.81397057e-01 2.23890036e-01 -6.13621533e-01
5.34448445e-01 -5.81911862e-01 4.22789484e-01 7.19188750e-01
3.53141949e-02 2.49752492e-01 -5.84898330e-02 1.16683491e-01
-6.02571845e-01 -1.18694532e+00 -2.86332935e-01 -1.01746336e-01
1.05090415e+00 2.27717206e-01 -2.87318140e-01 3.91573697e-01
1.96835265e-01 8.39568794e-01 4.41162050e-01 3.60734999e-01
-2.71020364e-03 -8.08246136e-01 3.56857985e-01 1.77468434e-02
8.49717617e-01 -3.23102802e-01 -7.32836366e-01 3.53519469e-01
2.01829255e-01 -1.62772560e+00 -1.73408225e-01 -3.06144357e-01
-5.41547418e-01 -1.11801600e+00 -5.45348346e-01 -6.90846264e-01
6.98891759e-01 7.53939331e-01 7.31872261e-01 1.31458819e-01
-1.83461249e-01 1.07897937e+00 -2.48654246e-01 -9.74879682e-01
1.13729827e-01 -4.10845131e-01 4.04860437e-01 4.01769318e-02
3.89102194e-03 -4.09250259e-01 -1.14227378e+00 6.89564347e-01
-4.10042368e-02 -9.26359964e-04 5.66893995e-01 1.31668359e-01
-2.07948536e-02 -6.45544454e-02 7.61899829e-01 -2.93357700e-01
2.01482594e-01 -4.63195890e-01 -9.18035805e-01 3.29345581e-03
-7.43378699e-01 -5.43129027e-01 3.28811675e-01 2.58372068e-01
-1.02574217e+00 -2.46566594e-01 7.97623619e-02 7.60554016e-01
2.87525100e-03 2.66741067e-01 -2.12168515e-01 -4.74280268e-01
4.65396017e-01 -3.09338927e-01 -1.00881942e-01 -2.07745824e-02
-3.74299079e-01 1.12899327e+00 4.63612154e-02 4.47645009e-01
8.95533323e-01 7.70015597e-01 6.27862692e-01 -1.67775774e+00
-3.21093321e-01 -7.43962049e-01 5.97831234e-02 -9.47750986e-01
9.57832098e-01 -1.42667222e+00 -7.67811000e-01 4.57986802e-01
-1.12097728e+00 -3.00987929e-01 4.94709285e-03 1.50403011e+00
3.44707668e-02 2.74239276e-02 -5.85705996e-01 -1.52536774e+00
-2.19656497e-01 -7.71927178e-01 5.86153746e-01 7.66450942e-01
1.63267717e-01 -1.04718196e+00 -2.87524372e-01 5.79260707e-01
7.09146261e-01 3.74941707e-01 4.26731765e-01 6.81115687e-01
-1.10286891e+00 -3.32790464e-01 -1.79245502e-01 -5.91829896e-01
7.35281855e-02 -2.37060502e-01 -8.87174249e-01 -6.56098902e-01
-3.64509940e-01 2.99730361e-01 6.19715273e-01 9.66955066e-01
-5.60058886e-03 -2.73186862e-01 -1.25429440e+00 5.58950305e-01
1.88501537e+00 6.82751358e-01 1.09713256e+00 -4.28543873e-02
2.69173652e-01 2.55070418e-01 5.63249052e-01 2.33766153e-01
-1.00917593e-01 7.16592669e-01 6.29860938e-01 -1.47043377e-01
-3.36135626e-01 7.27462351e-01 6.37054861e-01 6.29239738e-01
-4.35152262e-01 -1.46719563e+00 -5.86453080e-01 1.25223517e-01
-1.31580746e+00 -7.77278423e-01 -1.20779467e+00 2.14161611e+00
1.69006974e-01 -2.08176821e-01 2.85360217e-02 -2.16102228e-01
4.57174182e-01 -4.46647376e-01 7.55019337e-02 -4.26958166e-02
-2.77534902e-01 -5.91204911e-02 1.43163478e+00 1.25929880e+00
-1.04337215e-01 2.20445931e-01 6.46992922e+00 7.00154781e-01
-8.24935794e-01 -1.19143710e-01 1.48395196e-01 -4.56720740e-01
-4.72662836e-01 -7.40442500e-02 -9.78200495e-01 5.75803041e-01
5.95392525e-01 -3.62271932e-03 -6.50492907e-02 -4.83752489e-01
7.65085340e-01 -1.22133398e+00 -6.18966460e-01 1.00710618e+00
-4.71111894e-01 -1.09683406e+00 -2.54539937e-01 9.87087667e-01
6.92093074e-01 2.26421610e-01 -3.93088162e-01 -3.53553265e-01
-2.97627091e-01 -7.73418188e-01 2.42474124e-01 1.10162354e+00
8.40020061e-01 -4.90265936e-01 1.29450488e+00 4.77221131e-01
-1.10386157e+00 -3.81338567e-01 -3.36792707e-01 -5.00076056e-01
4.62104499e-01 1.21804476e+00 -4.82060969e-01 5.03960133e-01
3.48525047e-01 2.33498767e-01 8.22403803e-02 1.61605358e+00
1.12404130e-01 6.22453153e-01 -6.95727229e-01 -7.86942959e-01
7.75481015e-02 -6.95146143e-01 1.08791804e+00 9.31470931e-01
1.12407660e+00 5.02603531e-01 -1.14398777e-01 4.38704044e-01
-9.11766812e-02 -5.89227304e-02 -8.66411984e-01 6.97816193e-01
7.92459726e-01 1.30704379e+00 -7.80433178e-01 -1.15522809e-01
-5.55264652e-01 4.78947282e-01 -6.88163459e-01 9.54767466e-01
-4.01546925e-01 -6.74819291e-01 8.69049668e-01 5.41290164e-01
-3.22229825e-02 -6.17322445e-01 -5.35676837e-01 -4.17357475e-01
-4.52829897e-01 2.85827547e-01 -2.82314450e-01 -7.29367554e-01
-7.68721581e-01 -2.74366736e-01 -2.52217770e-01 -1.46613073e+00
8.90749693e-01 -4.41587061e-01 -6.64704144e-01 9.12766635e-01
-1.57546139e+00 -4.13158089e-01 -6.85324311e-01 -2.51591027e-01
3.43519379e-03 -3.46636355e-01 3.38368952e-01 1.71742082e-01
-4.35415655e-01 1.64626315e-02 1.19257855e+00 -2.03150019e-01
5.59989750e-01 -7.73931324e-01 -9.85085189e-01 5.83201110e-01
-4.20575529e-01 2.16142952e-01 6.40890658e-01 -4.45406973e-01
-1.54864812e+00 -8.04218054e-01 6.49147272e-01 5.67947254e-02
-2.46734872e-01 -2.38552228e-01 2.54073948e-01 -1.09136283e-01
9.67989326e-01 1.09365389e-01 1.10873890e+00 -5.06322742e-01
7.39840448e-01 -5.17533183e-01 -1.41873419e+00 7.84703374e-01
1.04417372e+00 -3.91449109e-02 3.75350535e-01 3.08587521e-01
3.96684714e-04 -2.74024457e-01 -5.95948935e-01 3.25199693e-01
1.25939512e+00 -9.82295454e-01 5.79776883e-01 4.99241531e-01
-5.68475008e-01 -7.50320554e-01 5.40728010e-02 -1.12741578e+00
-5.26324451e-01 -5.38924158e-01 3.91680568e-01 1.20870459e+00
6.85774088e-01 -1.29943371e+00 6.13987207e-01 -5.22535816e-02
1.71335056e-01 -5.85399449e-01 -1.26056182e+00 -1.30955613e+00
-3.28456551e-01 1.85596047e-03 1.59214139e-02 4.08228785e-01
1.90476224e-01 1.12161374e+00 1.04467869e-01 5.78261733e-01
6.56790793e-01 -6.24435425e-01 8.14265251e-01 -1.52194321e+00
1.79395348e-01 -2.14877322e-01 -3.56542736e-01 -1.13094425e+00
-6.76990628e-01 -4.87352163e-01 -3.83630916e-02 -2.30939507e+00
-1.27129793e-01 -3.71576190e-01 3.30810025e-02 -1.80666372e-01
3.42104495e-01 8.62423718e-01 1.94870248e-01 1.43266827e-01
-2.10345164e-01 2.15938717e-01 1.41896820e+00 4.64383326e-02
-4.39784884e-01 3.56986374e-01 -2.70175993e-01 3.21106493e-01
8.62916112e-01 -1.73949942e-01 -3.96508604e-01 -1.55022725e-01
3.40957373e-01 1.63408339e-01 2.20612094e-01 -2.00056434e+00
2.65464395e-01 2.49489304e-02 4.80192691e-01 -1.52004272e-01
2.44975597e-01 -1.24844265e+00 3.05002749e-01 1.21922147e+00
5.14728785e-01 -8.19521666e-01 -2.44580537e-01 1.00477397e+00
2.75750220e-01 -3.01886946e-01 1.21613908e+00 1.97344974e-01
3.13987792e-01 -1.74451843e-01 -1.57802904e+00 -5.89931071e-01
1.78592539e+00 -1.05354846e+00 -9.99148190e-01 -1.00497377e+00
-1.95757881e-01 5.88073850e-01 4.17009532e-01 -4.70134079e-01
1.93737164e-01 -1.07448387e+00 -5.45104980e-01 -1.71594605e-01
1.40358642e-01 -5.04497170e-01 -4.95605171e-01 1.42859912e+00
-9.47997451e-01 9.02791858e-01 -6.95540607e-02 -6.22113585e-01
-1.95378721e+00 -3.85342389e-01 8.59000862e-01 3.37538093e-01
-4.43058968e-01 7.13254213e-01 -1.27171148e-02 6.46201849e-01
6.96183667e-02 -1.69585183e-01 -5.21197140e-01 -4.51930426e-02
9.18948799e-02 7.47017562e-01 -3.91698211e-01 -4.48656678e-01
-4.55981821e-01 1.16837716e+00 8.33702862e-01 -3.10875833e-01
7.74252415e-01 -7.22627938e-01 -1.42528147e-01 3.83906215e-01
6.12845480e-01 8.13129246e-01 -1.03901243e+00 1.68048859e-01
-4.98799086e-01 -3.48351419e-01 2.81392723e-01 -9.76132751e-01
-6.49736285e-01 3.95395875e-01 1.01061928e+00 5.43861568e-01
9.41489816e-01 2.40059346e-02 7.26111293e-01 5.62120318e-01
7.40988255e-01 -1.53790319e+00 -1.43491030e-01 4.55879480e-01
5.55630744e-01 -7.37489462e-01 3.18593383e-01 -9.89327252e-01
7.59642124e-02 1.30004025e+00 5.66899598e-01 1.55824676e-01
9.56165969e-01 6.40188932e-01 1.38382167e-01 -2.07499400e-01
-5.18341899e-01 -6.71069443e-01 -1.65024191e-01 8.54670405e-01
7.85361826e-01 4.03367588e-03 -9.89102721e-01 -1.77975461e-01
1.20121278e-01 -1.70841768e-01 1.05213964e+00 8.18787992e-01
-1.16946328e+00 -9.50228930e-01 -7.37658739e-01 4.43433583e-01
-4.86298651e-01 3.50389957e-01 -1.98227346e-01 6.37193263e-01
3.88724238e-01 1.64694965e+00 1.83341444e-01 -6.17872365e-02
1.81243435e-01 -2.64815569e-01 6.21209204e-01 -6.87535405e-01
1.67422593e-01 4.61938918e-01 5.95417440e-01 2.47341506e-02
-8.40253890e-01 -3.56774479e-01 -1.55980146e+00 -7.32885376e-02
-4.86472100e-01 3.83782893e-01 3.20197642e-01 1.09728253e+00
3.59956950e-01 3.31328332e-01 8.66935909e-01 -3.03154081e-01
8.04173052e-01 -8.49460423e-01 -1.14058316e+00 -6.87249303e-01
6.06869161e-01 -4.05380309e-01 -2.74092317e-01 -4.68172133e-01] | [6.254836559295654, 1.2200568914413452] |
70a93b38-4318-4aec-b77d-8617c5ac91ab | i2edit-towards-multi-turn-interactive-image | 2303.11108 | null | https://arxiv.org/abs/2303.11108v2 | https://arxiv.org/pdf/2303.11108v2.pdf | I2Edit: Towards Multi-turn Interactive Image Editing via Dialogue | Although there have been considerable research efforts on controllable facial image editing, the desirable interactive setting where the users can interact with the system to adjust their requirements dynamically hasn't been well explored. This paper focuses on facial image editing via dialogue and introduces a new benchmark dataset, Multi-turn Interactive Image Editing (I2Edit), for evaluating image editing quality and interaction ability in real-world interactive facial editing scenarios. The dataset is constructed upon the CelebA-HQ dataset with images annotated with a multi-turn dialogue that corresponds to the user editing requirements. I2Edit is challenging, as it needs to 1) track the dynamically updated user requirements and edit the images accordingly, as well as 2) generate the appropriate natural language response to communicate with the user. To address these challenges, we propose a framework consisting of a dialogue module and an image editing module. The former is for user edit requirements tracking and generating the corresponding indicative responses, while the latter edits the images conditioned on the tracked user edit requirements. In contrast to previous works that simply treat multi-turn interaction as a sequence of single-turn interactions, we extract the user edit requirements from the whole dialogue history instead of the current single turn. The extracted global user edit requirements enable us to directly edit the input raw image to avoid error accumulation and attribute forgetting issues. Extensive quantitative and qualitative experiments on the I2Edit dataset demonstrate the advantage of our proposed framework over the previous single-turn methods. We believe our new dataset could serve as a valuable resource to push forward the exploration of real-world, complex interactive image editing. Code and data will be made public. | ['Zhaofeng He', 'Hailin Shi', 'Yibo Hu', 'Peipei Li', 'Zekun Li', 'Xing Cui'] | 2023-03-20 | null | null | null | null | ['facial-editing'] | ['computer-vision'] | [ 7.31414318e-01 2.46629655e-01 1.12415567e-01 -7.45211482e-01
-3.97276551e-01 -6.22359276e-01 5.87036848e-01 -2.97462851e-01
-3.23040366e-01 3.65817547e-01 -7.11922422e-02 1.76940728e-02
1.49059579e-01 -4.39061642e-01 -3.83157849e-01 -3.78648192e-01
3.84249777e-01 4.36089694e-01 1.23069800e-01 -4.53202307e-01
2.74412364e-01 3.86298031e-01 -1.84832120e+00 5.14528096e-01
9.94970798e-01 9.21318114e-01 3.73602003e-01 7.55816340e-01
-1.25288129e-01 5.82395613e-01 -5.82998037e-01 -5.56967318e-01
3.42651427e-01 -5.23246646e-01 -6.82563961e-01 6.11783206e-01
4.01360631e-01 -4.78968829e-01 9.28002298e-02 1.02939236e+00
6.01909697e-01 3.59617025e-01 2.86162168e-01 -1.54684901e+00
-5.90210855e-01 3.38890016e-01 -6.40068054e-01 -2.58222103e-01
8.91036034e-01 3.97083431e-01 5.63087583e-01 -1.01283491e+00
1.05635834e+00 1.43883848e+00 3.12736094e-01 8.86059880e-01
-9.55859065e-01 -8.34791958e-01 4.70303535e-01 3.59582491e-02
-1.22894835e+00 -7.48865724e-01 6.61618948e-01 -5.81601262e-01
4.49345559e-01 4.81233358e-01 6.52455509e-01 9.48922634e-01
-8.21930990e-02 8.28508198e-01 1.11582744e+00 -5.92084527e-01
-1.79809585e-01 5.43228269e-01 -1.61519095e-01 8.74054670e-01
-7.67626047e-01 -1.10166267e-01 -6.03940368e-01 1.89033628e-01
6.64072156e-01 -1.79084390e-01 -3.19691539e-01 -2.69046426e-01
-1.18998659e+00 4.50896710e-01 -9.23235714e-02 -1.05497807e-01
-3.70130301e-01 -1.48119614e-01 4.44984287e-01 6.96360111e-01
4.81773108e-01 1.83024719e-01 -4.19687331e-01 -3.90213430e-01
-8.45142663e-01 2.70398289e-01 6.72941566e-01 1.32577682e+00
8.22305799e-01 -1.82390660e-01 -4.74152446e-01 1.04188287e+00
-7.63214082e-02 4.22077745e-01 -9.96370800e-03 -9.51988757e-01
2.79519528e-01 8.41348588e-01 4.28597450e-01 -1.16109598e+00
-1.09109215e-01 3.97035450e-01 -5.24378181e-01 2.83734202e-01
4.47288573e-01 -4.07250911e-01 -8.53863597e-01 1.83504581e+00
6.08217478e-01 -3.82517248e-01 -2.27506891e-01 1.00271952e+00
9.15751398e-01 5.65536439e-01 -4.04293239e-02 -4.54281062e-01
1.49062002e+00 -8.85841370e-01 -1.04326236e+00 1.17181458e-01
2.61699021e-01 -1.07851064e+00 1.50981641e+00 4.33379024e-01
-1.20211506e+00 -6.84936821e-01 -4.53562200e-01 -6.57456517e-02
-2.98254013e-01 6.59497976e-01 5.76569378e-01 5.53500652e-01
-1.15811265e+00 6.61089197e-02 -4.18270528e-01 -4.83958066e-01
1.85460180e-01 3.33941489e-01 -4.62138057e-01 1.55173719e-01
-1.21131170e+00 6.70776904e-01 3.55211198e-02 2.75095254e-01
-6.56098068e-01 -6.59447610e-01 -7.26645291e-01 -2.78195113e-01
8.34008515e-01 -3.48654240e-01 1.58443999e+00 -1.71081436e+00
-1.87721920e+00 1.13383532e+00 -3.15374345e-01 2.20535368e-01
1.00978327e+00 -1.60413787e-01 -3.76833916e-01 1.68219656e-02
-1.30883157e-02 1.00185537e+00 9.91773784e-01 -1.35457885e+00
-7.38808095e-01 -2.38050520e-01 5.13830960e-01 4.86969352e-01
-3.91342670e-01 3.18018109e-01 -1.23051488e+00 -5.24912357e-01
-5.21381736e-01 -1.25007498e+00 6.90647289e-02 3.00364435e-01
-3.93849075e-01 -1.53969765e-01 9.46876824e-01 -5.85699975e-01
1.31491244e+00 -2.13419318e+00 2.26262122e-01 8.87936279e-02
3.53532210e-02 2.11277872e-01 -1.80758029e-01 5.15424490e-01
-1.40730038e-01 -2.77626198e-02 -4.17754091e-02 -4.21664059e-01
-4.36178371e-02 -9.06941742e-02 -1.40036702e-01 8.93509537e-02
7.72942007e-02 7.07404912e-01 -8.50624204e-01 -6.29725873e-01
2.78048903e-01 3.39471579e-01 -4.98514593e-01 4.08973575e-01
-3.19820493e-01 6.25696898e-01 -3.42336714e-01 7.13799894e-01
6.79046810e-01 7.29747638e-02 1.59265652e-01 -3.97059411e-01
-2.49213845e-01 -5.26301384e-01 -1.17756939e+00 1.60435295e+00
-5.85103869e-01 7.45639443e-01 2.63920367e-01 -3.83964270e-01
9.11713839e-01 2.84501433e-01 5.48000932e-01 -7.65021980e-01
2.10033864e-01 -4.00174797e-01 -2.81568140e-01 -9.11533892e-01
8.24107945e-01 2.91394025e-01 -3.10874373e-01 6.85452104e-01
-2.45650411e-01 -2.18623534e-01 3.78007948e-01 4.15417343e-01
5.88161528e-01 3.75744045e-01 2.85268486e-01 2.40689679e-03
7.36423552e-01 -1.30057633e-01 5.49436629e-01 6.11019611e-01
-1.86226264e-01 5.14841199e-01 5.89207053e-01 -5.17124951e-01
-7.20509350e-01 -7.34650493e-01 2.59934127e-01 1.53204107e+00
2.79668242e-01 -4.41366315e-01 -1.08370054e+00 -6.50009453e-01
-1.87732205e-01 5.15540659e-01 -7.12442636e-01 9.95007455e-02
-4.07764584e-01 -3.24200302e-01 3.65769714e-01 6.30763322e-02
6.72014654e-01 -1.51508570e+00 -1.00857627e+00 -1.22519350e-02
-4.62018490e-01 -1.16542888e+00 -1.19563985e+00 -5.99934638e-01
-1.27114341e-01 -1.16183245e+00 -4.44034368e-01 -6.60229921e-01
1.06526816e+00 7.08743408e-02 1.04320705e+00 3.79649997e-01
-4.75719631e-01 8.28083158e-01 -5.44676065e-01 -5.62989593e-01
-5.57142496e-01 -9.40437168e-02 2.22148523e-02 5.56258261e-01
8.28738604e-03 -2.42161542e-01 -5.34602344e-01 6.82851732e-01
-1.00915456e+00 7.62191653e-01 3.67371708e-01 7.09604979e-01
5.38880706e-01 -3.48936498e-01 6.43506169e-01 -1.51268518e+00
1.02744257e+00 -1.31850445e-03 -5.61851203e-01 6.38999403e-01
-5.40519774e-01 -3.04913878e-01 4.50023741e-01 -6.57328427e-01
-1.58725190e+00 5.42621970e-01 1.41362339e-01 -4.10026908e-01
-8.86476859e-02 2.05925778e-01 -2.39274040e-01 -8.64200443e-02
3.98760557e-01 9.58226994e-02 1.16210155e-01 -6.26057014e-02
4.59196895e-01 9.44461167e-01 6.58435464e-01 -7.06151605e-01
6.38784289e-01 2.62970775e-01 -5.57074308e-01 -8.21375549e-01
-4.51864153e-01 -2.20344588e-01 -7.94283330e-01 -9.88161087e-01
7.58777082e-01 -7.12814748e-01 -1.00441360e+00 6.56723738e-01
-1.22715795e+00 -4.85535234e-01 -8.65250528e-02 -1.54941872e-01
-6.36799276e-01 3.30083787e-01 -3.10660154e-01 -1.03998590e+00
-3.30761075e-01 -1.40143502e+00 1.28425348e+00 4.79269147e-01
-3.84889990e-01 -6.65768743e-01 -3.53107333e-01 3.37528050e-01
3.28452557e-01 4.48311478e-01 6.13216519e-01 -7.84005374e-02
-4.58360285e-01 -1.52514771e-01 -1.77380145e-01 6.77170279e-03
4.38126385e-01 4.84942585e-01 -7.46677577e-01 -2.56229818e-01
-3.62211853e-01 -4.75964218e-01 2.61697024e-01 -1.21562965e-02
1.28321624e+00 -1.95868105e-01 -9.74148959e-02 3.44327331e-01
9.62549686e-01 4.77885723e-01 7.03361869e-01 -7.07125738e-02
4.94811177e-01 7.88871109e-01 1.27090299e+00 7.26191521e-01
5.00097930e-01 8.73873234e-01 2.71651745e-01 -3.04440677e-01
1.23956032e-01 -2.26035029e-01 4.35869247e-01 6.21305883e-01
-1.71379015e-01 -2.38210768e-01 -6.60966516e-01 4.11944062e-01
-1.89854646e+00 -8.26003492e-01 2.37443652e-02 2.03872609e+00
1.05377781e+00 -2.51970917e-01 2.78781299e-02 -2.32832313e-01
9.79044080e-01 -2.35329904e-02 -7.24718988e-01 -6.85201526e-01
2.36100763e-01 7.39548877e-02 -2.14414252e-03 5.03108025e-01
-8.73957813e-01 1.18170333e+00 5.71031046e+00 5.53035557e-01
-1.17629588e+00 2.71738153e-02 6.14414036e-01 -5.58762103e-02
-1.82139084e-01 6.48430828e-03 -6.73358738e-01 5.23297906e-01
4.03603673e-01 -1.90207347e-01 6.51899397e-01 4.96128827e-01
6.72476709e-01 -4.44734514e-01 -1.16687584e+00 1.04188883e+00
2.09567070e-01 -9.94300425e-01 2.40519449e-01 -3.25732589e-01
5.79278350e-01 -6.95378244e-01 1.19985908e-01 3.12863439e-01
1.82353422e-01 -7.34968603e-01 7.79286206e-01 1.04086316e+00
1.50980139e+00 -6.73630774e-01 3.22685361e-01 3.86239260e-01
-1.16767097e+00 5.67581020e-02 1.29001230e-01 4.31538559e-02
3.21574926e-01 2.19380915e-01 -9.42913294e-01 4.36615258e-01
7.36695647e-01 4.96024787e-01 -6.97714329e-01 3.45274955e-01
-6.88973889e-02 -7.22291470e-02 -6.12693280e-02 1.09791346e-01
-3.03749532e-01 -3.99757236e-01 4.21639889e-01 1.29597938e+00
1.84600249e-01 3.32191497e-01 3.14837188e-01 7.10023284e-01
-2.13310033e-01 3.57728899e-01 -4.77313697e-01 -1.47683382e-01
5.25238574e-01 1.54293442e+00 -6.88323975e-01 -3.42687488e-01
-3.18215221e-01 1.41213822e+00 1.67142734e-01 3.43371749e-01
-1.07343209e+00 -5.38180411e-01 6.12035930e-01 1.44812614e-01
3.28436284e-03 4.20563556e-02 2.18362033e-01 -1.01066363e+00
8.83843005e-02 -1.19180274e+00 2.29863003e-01 -1.14096427e+00
-8.71999383e-01 8.02391589e-01 9.67085138e-02 -1.18710768e+00
-2.71785855e-01 -1.95192993e-01 -5.08662999e-01 6.67058587e-01
-1.13343775e+00 -1.21231389e+00 -9.40179288e-01 7.86980927e-01
1.01309621e+00 -2.29446888e-02 7.83224046e-01 3.04072112e-01
-6.63752556e-01 9.92804646e-01 -5.79185367e-01 -8.11511301e-04
1.22168195e+00 -9.36287045e-01 1.15073696e-01 5.62373698e-01
-2.05132782e-01 4.84486848e-01 7.42672384e-01 -6.43839121e-01
-1.37931144e+00 -1.02319944e+00 4.28662270e-01 -5.46541810e-01
2.39687920e-01 -5.36880195e-01 -6.36068702e-01 7.69863069e-01
4.84192073e-01 -1.16936445e-01 3.88229609e-01 -3.77392113e-01
1.21677183e-01 -2.30042428e-01 -1.26871037e+00 8.29252601e-01
1.24902880e+00 -4.21165109e-01 -1.93952486e-01 2.84145176e-01
5.46690583e-01 -9.20004129e-01 -7.38836765e-01 3.49735141e-01
7.70532072e-01 -9.96222913e-01 5.32842159e-01 -3.54991615e-01
3.30518484e-01 -3.21795821e-01 2.74217486e-01 -1.38440609e+00
2.21279785e-01 -1.09193277e+00 5.82974732e-01 1.61508501e+00
2.01378211e-01 -3.07034671e-01 4.59978461e-01 1.08594632e+00
-5.29869273e-02 -7.46545076e-01 -5.83991647e-01 -1.60760999e-01
-7.39408255e-01 -2.35926926e-01 6.92235529e-01 9.69039142e-01
5.11022881e-02 7.63186663e-02 -7.69936621e-01 -4.14557606e-02
1.82386041e-01 1.63622722e-01 1.37466288e+00 -1.05525136e+00
3.92488390e-03 -1.53090432e-01 2.07448145e-03 -9.18136537e-01
2.38375381e-01 -5.51085234e-01 2.78104991e-01 -1.09025168e+00
2.11019143e-01 -4.55382705e-01 3.35702449e-01 6.27347469e-01
-1.99696019e-01 2.18707860e-01 3.61217916e-01 8.52309689e-02
-7.12394774e-01 4.45323199e-01 1.50424016e+00 -1.29128164e-02
-5.99676430e-01 2.06959471e-02 -6.49813473e-01 6.38385236e-01
6.10693157e-01 -1.66063920e-01 -7.22419322e-01 -3.21856230e-01
2.06185982e-01 4.91172016e-01 1.11581855e-01 -6.19517088e-01
3.81493062e-01 -6.88067913e-01 1.62758812e-01 -3.68582219e-01
2.49221712e-01 -9.11276281e-01 5.57615817e-01 -3.11744176e-02
-5.57353020e-01 4.01868284e-01 1.80444613e-01 4.98591214e-01
-2.21380040e-01 -3.34223993e-02 8.17017734e-01 -2.26720646e-01
-1.00517011e+00 4.81613219e-01 -4.94208753e-01 -1.11305654e-01
1.43551195e+00 -4.26242560e-01 -1.78959910e-02 -7.42527783e-01
-9.52281415e-01 5.63051641e-01 7.26187050e-01 7.77096570e-01
8.05126071e-01 -1.25264013e+00 -5.71602106e-01 3.53324145e-01
3.70041907e-01 -1.31833926e-01 4.88879740e-01 7.04220295e-01
-3.00072789e-01 -2.13454232e-01 -4.79780763e-01 -6.33946657e-01
-1.84800327e+00 2.84315854e-01 3.28023732e-01 -3.47931236e-02
-4.28579181e-01 6.38144493e-01 3.43581587e-01 -5.56895614e-01
1.96739107e-01 -5.58892898e-02 -3.30518335e-01 4.14826542e-01
4.48418468e-01 7.21396692e-03 -1.24427646e-01 -7.35340238e-01
-3.50628085e-02 5.06451309e-01 -2.56177634e-01 -1.27572834e-01
1.09166229e+00 -6.87511086e-01 -1.73851088e-01 5.27752578e-01
6.65481865e-01 9.69873741e-02 -1.36715555e+00 -8.69427398e-02
-2.93309897e-01 -7.56118655e-01 -4.40338194e-01 -9.32669938e-01
-1.21239042e+00 6.10283077e-01 6.72988594e-01 -6.01663254e-02
1.55929458e+00 -3.63490164e-01 6.45744741e-01 2.96635598e-01
4.09116805e-01 -1.49942851e+00 3.27615917e-01 3.94177258e-01
1.22669661e+00 -1.17065907e+00 -2.17981249e-01 -6.18600368e-01
-1.26873267e+00 1.14246953e+00 1.20960402e+00 2.98570335e-01
3.58487099e-01 3.31172884e-01 4.27392215e-01 -2.15812102e-01
-8.73205185e-01 2.36711986e-02 1.28663912e-01 4.81884599e-01
3.26866120e-01 1.79971680e-02 -1.64501637e-01 4.83882099e-01
-2.25509092e-01 3.29560250e-01 7.74561465e-01 7.29704797e-01
-8.31158236e-02 -1.18487370e+00 -2.02627450e-01 4.38365430e-01
-7.29863495e-02 1.94307610e-01 -7.61022747e-01 8.99833560e-01
1.82233199e-01 9.80087399e-01 -1.07087366e-01 -4.56024766e-01
7.95541644e-01 1.97134867e-01 3.76213282e-01 -6.95941210e-01
-8.54678512e-01 -2.85471380e-01 4.08160463e-02 -7.50557423e-01
-6.12320960e-01 -6.12343132e-01 -1.23348892e+00 -3.29819024e-01
-3.37398708e-01 -8.05139691e-02 4.67809856e-01 6.75693929e-01
4.89021182e-01 4.50790972e-01 8.93503129e-01 -9.66378450e-01
-1.21942610e-01 -9.12436008e-01 -2.29631871e-01 7.61746109e-01
2.18811169e-01 -7.54941940e-01 -3.53748947e-02 5.88262916e-01] | [12.52230167388916, -0.29158467054367065] |
531afa72-ca8d-44ee-b58c-753209f5360b | on-the-utility-and-protection-of-optimization | 2209.03175 | null | https://arxiv.org/abs/2209.03175v1 | https://arxiv.org/pdf/2209.03175v1.pdf | On the utility and protection of optimization with differential privacy and classic regularization techniques | Nowadays, owners and developers of deep learning models must consider stringent privacy-preservation rules of their training data, usually crowd-sourced and retaining sensitive information. The most widely adopted method to enforce privacy guarantees of a deep learning model nowadays relies on optimization techniques enforcing differential privacy. According to the literature, this approach has proven to be a successful defence against several models' privacy attacks, but its downside is a substantial degradation of the models' performance. In this work, we compare the effectiveness of the differentially-private stochastic gradient descent (DP-SGD) algorithm against standard optimization practices with regularization techniques. We analyze the resulting models' utility, training performance, and the effectiveness of membership inference and model inversion attacks against the learned models. Finally, we discuss differential privacy's flaws and limits and empirically demonstrate the often superior privacy-preserving properties of dropout and l2-regularization. | ['Matteo Matteucci', 'Eugenio Lomurno'] | 2022-09-07 | null | null | null | null | ['l2-regularization'] | ['methodology'] | [ 6.07859418e-02 2.44073480e-01 -1.79952383e-02 -6.68464541e-01
-9.43388879e-01 -8.57805669e-01 4.33879077e-01 4.00951989e-02
-7.44101763e-01 8.76567304e-01 -1.23289280e-01 -4.53766435e-01
-1.17806539e-01 -5.50270319e-01 -9.12791312e-01 -1.03400660e+00
1.81826994e-01 8.08352903e-02 -1.26002878e-01 2.44155034e-01
1.93960130e-01 8.78678799e-01 -1.50978494e+00 1.77213200e-03
6.57679260e-01 1.08260286e+00 -5.11817336e-01 3.07507932e-01
1.75835073e-01 7.16463506e-01 -5.15760303e-01 -1.21102238e+00
7.32609928e-01 -1.58083752e-01 -6.05232179e-01 -4.23320591e-01
6.00316882e-01 -5.85069478e-01 -5.27621329e-01 1.31520033e+00
4.77466732e-01 -1.40473455e-01 2.81357676e-01 -1.61054766e+00
-6.52333975e-01 3.12331587e-01 -3.04371893e-01 -2.29010321e-02
-1.94978252e-01 1.79515302e-01 9.53245878e-01 -4.21410203e-01
5.75241387e-01 8.90185118e-01 7.64653027e-01 9.41480994e-01
-1.62556982e+00 -9.60827172e-01 -6.29165396e-02 -1.03283547e-01
-1.56959522e+00 -8.05896282e-01 5.29357791e-01 -3.90012681e-01
5.00212491e-01 6.56925321e-01 2.71970153e-01 1.33701682e+00
-2.86533497e-02 8.28150511e-01 9.75758076e-01 -1.76704615e-01
5.11279404e-01 9.22035515e-01 1.08881369e-01 6.00405455e-01
6.72013283e-01 2.40083560e-01 -6.49963021e-01 -7.81361520e-01
4.21559155e-01 -1.37345687e-01 -3.32035094e-01 -8.71084809e-01
-1.56889081e-01 8.46565127e-01 -7.02197403e-02 1.61844626e-01
4.70152125e-02 1.83597550e-01 5.37411511e-01 5.82333028e-01
9.18352783e-01 1.36202767e-01 -6.58299387e-01 1.44805610e-01
-7.70731807e-01 5.99241197e-01 9.37427223e-01 1.01219833e+00
6.91164434e-01 -1.82558089e-01 -2.84212418e-02 4.47723478e-01
4.82704043e-01 -2.40782686e-02 4.25192684e-01 -9.48053181e-01
4.05586571e-01 2.01220766e-01 1.18206024e-01 -1.06475401e+00
2.31908113e-01 -4.48840141e-01 -6.22853637e-01 4.73458111e-01
5.09062290e-01 -4.90137875e-01 -1.28790081e-01 2.06850648e+00
4.44098443e-01 -2.97863744e-02 1.70008138e-01 4.53608841e-01
1.72708035e-01 8.30565318e-02 2.40570605e-01 -2.10595243e-02
5.91458857e-01 -4.25767958e-01 -5.05028307e-01 -4.64456709e-04
9.80248451e-01 -4.75017540e-02 8.47833872e-01 9.69034731e-02
-1.10348165e+00 8.89003202e-02 -9.54230607e-01 -4.30713415e-01
-3.35295469e-01 -1.49257809e-01 8.10502470e-01 1.49593282e+00
-1.17555714e+00 9.21583533e-01 -7.82485008e-01 7.70253092e-02
1.23862672e+00 7.27023065e-01 -7.78531551e-01 6.34758100e-02
-1.04247248e+00 5.52683413e-01 8.41209758e-03 -1.90935001e-01
-8.95112514e-01 -1.25531006e+00 -7.43460357e-01 2.13128686e-01
-3.80055420e-02 -5.86174726e-01 1.05620706e+00 -9.20052528e-01
-1.31297505e+00 1.60938561e+00 1.21887870e-01 -1.04683781e+00
1.30782282e+00 -2.22184330e-01 -1.50064786e-03 -2.08392560e-01
-3.28908086e-01 2.86095262e-01 7.58981884e-01 -1.05057466e+00
-7.63803303e-01 -8.85765016e-01 -1.91486366e-02 -2.65748277e-02
-8.27545702e-01 1.00569315e-01 -8.99953395e-02 -4.01301771e-01
-2.65938699e-01 -8.54164004e-01 -2.58151531e-01 6.71485245e-01
-4.59926635e-01 1.01371177e-01 9.24951792e-01 -7.99335480e-01
1.17251766e+00 -2.62068892e+00 -3.27577949e-01 2.64162064e-01
2.55178839e-01 3.74067694e-01 1.43130228e-01 7.37417042e-02
2.24412441e-01 3.75082344e-01 -4.98084128e-01 -1.16404414e+00
4.84029233e-01 1.49653286e-01 -4.86879617e-01 1.13999093e+00
-2.58238584e-01 7.18546271e-01 -3.18218201e-01 -2.22871467e-01
-1.50286714e-02 8.42023313e-01 -7.50410140e-01 1.31825313e-01
-1.44484386e-01 4.03856963e-01 -4.85145241e-01 2.14447752e-01
1.03582573e+00 7.24472627e-02 1.85536608e-01 4.48606580e-01
7.08019286e-02 2.10476130e-01 -8.88147771e-01 1.47396624e+00
-1.87578261e-01 6.77003086e-01 5.05238295e-01 -7.16850042e-01
7.23709702e-01 3.21868300e-01 3.61635655e-01 -3.04102987e-01
1.31578922e-01 3.48983526e-01 -4.45940197e-01 -3.88468921e-01
-5.48641523e-03 -4.74773273e-02 2.27635264e-01 5.02303481e-01
-3.34021188e-02 2.36764282e-01 -6.23325348e-01 -5.33458330e-02
9.56523597e-01 5.09276055e-02 -2.38276482e-01 -6.24973774e-01
5.24229527e-01 -4.16235119e-01 6.90860748e-01 8.06371748e-01
-3.96367759e-01 4.82009977e-01 7.47968137e-01 -2.68873006e-01
-9.10105765e-01 -6.06198967e-01 -3.38558614e-01 1.19408798e+00
-3.50327015e-01 -7.11579323e-02 -1.15745568e+00 -1.02282548e+00
5.39150238e-01 9.68885005e-01 -7.78129816e-01 -5.20090699e-01
-1.10873826e-01 -7.33152330e-01 1.05067694e+00 2.72543393e-02
5.08151233e-01 -4.11984742e-01 -6.14050567e-01 -3.04855466e-01
4.13576514e-01 -8.04187477e-01 -3.96128833e-01 1.34650186e-01
-9.17414904e-01 -8.28661501e-01 -6.42272174e-01 -5.21243453e-01
6.87412858e-01 -3.32895964e-02 6.96004450e-01 2.81602815e-02
-1.32843763e-01 3.70364904e-01 2.50401974e-01 -5.94053984e-01
-5.11698186e-01 1.33590177e-01 5.06570488e-02 5.47390163e-01
5.81337035e-01 -8.23087394e-01 -3.74644101e-01 -9.16902348e-02
-9.74566817e-01 -4.02392417e-01 -1.02582481e-02 4.36794102e-01
5.14214039e-01 1.39488608e-01 1.48683950e-01 -1.42506433e+00
6.14167273e-01 -3.89432520e-01 -9.37048912e-01 2.63731986e-01
-9.97771561e-01 1.72736757e-02 5.35799861e-01 -2.00327307e-01
-1.16365957e+00 1.21222377e-01 -1.45792380e-01 -7.25039542e-01
-3.16281378e-01 3.37945111e-02 -5.69044709e-01 -7.36593127e-01
5.95847189e-01 2.05534652e-01 2.06978828e-01 -9.20885801e-01
1.78898707e-01 5.54372430e-01 2.66769886e-01 -5.59079349e-01
7.05138803e-01 8.13305736e-01 7.41282105e-02 -7.35785306e-01
-6.96141660e-01 -2.63439585e-02 -2.96838790e-01 2.96901375e-01
6.46185100e-01 -8.72639298e-01 -7.64184713e-01 8.05320501e-01
-9.56698179e-01 -1.98871061e-01 -4.57314014e-01 2.49453500e-01
-5.33563375e-01 3.50833029e-01 -5.80794573e-01 -1.12614095e+00
-2.66724139e-01 -8.59540045e-01 7.05841839e-01 -1.39829777e-02
7.59377107e-02 -1.15103233e+00 1.98720917e-02 4.65495318e-01
5.22430360e-01 4.29249287e-01 8.81209373e-01 -1.08543611e+00
-3.42811465e-01 -6.86731458e-01 1.56547487e-01 8.71668458e-01
-2.37621710e-01 -1.45300597e-01 -1.59324360e+00 -5.07190347e-01
7.28186250e-01 -3.02492231e-01 6.24112070e-01 2.42213070e-01
1.52158976e+00 -7.47598827e-01 -1.30454659e-01 1.13972449e+00
1.55104399e+00 -2.13439181e-01 6.22134328e-01 1.85268149e-01
5.68133712e-01 1.08689392e+00 2.04470195e-02 6.52678430e-01
-4.06189077e-02 2.61150479e-01 5.30174315e-01 8.21353421e-02
5.55560946e-01 -5.39561450e-01 3.55933905e-01 1.56781729e-02
3.94350708e-01 9.90684051e-03 -5.60139775e-01 3.16074133e-01
-1.78662932e+00 -8.27512920e-01 -6.54290617e-02 2.46539521e+00
8.55405986e-01 -7.66380280e-02 5.97998761e-02 -3.47055495e-02
5.12007296e-01 1.01365738e-01 -7.00385809e-01 -4.20914710e-01
-4.50947434e-01 9.37641934e-02 9.28332090e-01 3.78491789e-01
-1.08036065e+00 7.43829489e-01 6.47319794e+00 8.90283346e-01
-8.09749305e-01 3.48664403e-01 1.12415516e+00 -5.09732485e-01
-5.41369736e-01 -9.76956934e-02 -7.84196556e-01 5.33566356e-01
9.78748262e-01 -5.18596709e-01 3.91505659e-01 1.18286085e+00
-7.65063316e-02 2.66736925e-01 -1.43127406e+00 1.01700425e+00
-1.42526761e-01 -1.41496980e+00 -3.02991569e-01 5.94392061e-01
6.18556380e-01 2.68438961e-02 5.73653460e-01 7.85744339e-02
3.55425268e-01 -1.03309619e+00 8.20253432e-01 3.31495523e-01
6.86594129e-01 -1.09897029e+00 4.18532550e-01 5.24007976e-01
-2.24492073e-01 -7.48512670e-02 -5.53178608e-01 4.30124812e-02
-3.40386868e-01 4.24789339e-01 -1.44690022e-01 1.25449836e-01
7.66096056e-01 4.39022779e-01 -4.98697340e-01 6.57642961e-01
-1.07622191e-01 6.11266553e-01 -4.03658807e-01 1.24603964e-01
1.46629989e-01 -6.33816779e-01 4.38536137e-01 9.64942396e-01
-1.07578579e-02 -2.22873449e-01 -6.79506421e-01 1.24941123e+00
-3.96143645e-01 -5.94714284e-03 -9.83706951e-01 -1.49956150e-02
3.02148581e-01 7.01571941e-01 8.94697085e-02 1.03396893e-01
-3.70224088e-01 1.16724944e+00 4.32078600e-01 2.88916528e-01
-6.07132196e-01 -1.62679330e-02 1.42012835e+00 7.92680979e-02
1.42447695e-01 1.59809411e-01 -5.23622811e-01 -9.89137173e-01
4.62099731e-01 -9.31130350e-01 5.08662403e-01 -1.09155234e-02
-1.43046069e+00 1.02153264e-01 -3.36665154e-01 -6.07428372e-01
2.16638088e-01 -5.02479196e-01 -2.38990873e-01 1.08833241e+00
-1.26750255e+00 -1.05933416e+00 5.26360273e-01 7.59047449e-01
-1.68053508e-01 -3.14589053e-01 9.85086977e-01 4.13521081e-01
-6.73341870e-01 1.28515971e+00 7.47919083e-01 1.03888676e-01
2.20423102e-01 -9.20609117e-01 4.97629076e-01 7.55259275e-01
1.99487403e-01 7.87529826e-01 7.06717908e-01 -3.00271571e-01
-1.45002735e+00 -1.12050998e+00 1.04806709e+00 -5.47646284e-01
5.73133886e-01 -6.89736843e-01 -1.11111128e+00 1.03136969e+00
-1.23239212e-01 1.62431508e-01 1.04057753e+00 -7.61145428e-02
-6.64139032e-01 -2.35663220e-01 -1.97534788e+00 5.82699299e-01
9.36711907e-01 -1.09553003e+00 6.02451302e-02 4.94555593e-01
7.75143266e-01 -1.36208832e-01 -6.85675323e-01 2.53205933e-02
7.04365075e-01 -1.39480197e+00 7.60554314e-01 -1.13501287e+00
-4.91035245e-02 2.67441422e-01 -5.03903329e-01 -5.25503516e-01
1.48382753e-01 -1.11598194e+00 -1.56065762e-01 1.36257064e+00
5.39012134e-01 -9.61712956e-01 1.40606093e+00 1.75935578e+00
5.55526078e-01 -6.36009037e-01 -1.32154655e+00 -7.59636879e-01
4.94683713e-01 -4.10558164e-01 7.08497524e-01 1.27047002e+00
-3.87079298e-01 -3.95993620e-01 -5.60670078e-01 1.57653749e-01
1.06432307e+00 -4.41665381e-01 7.53995061e-01 -1.06336677e+00
-5.13428390e-01 -3.56067508e-01 -4.12065655e-01 -4.72483754e-01
6.74325347e-01 -8.92349482e-01 -4.94579047e-01 -4.85786051e-01
-7.09959790e-02 -4.16530341e-01 -4.78694439e-01 1.72308132e-01
2.91875869e-01 4.64077480e-02 1.22663617e-01 1.17811300e-01
-1.96029201e-01 4.76057589e-01 6.16121352e-01 7.45357350e-02
-9.94288996e-02 5.57678461e-01 -8.57765496e-01 5.50318599e-01
8.37635040e-01 -9.48169529e-01 -5.19484401e-01 -6.40056610e-01
2.99507707e-01 -4.66195732e-01 6.55106187e-01 -7.71118999e-01
3.44693363e-01 1.78105846e-01 -1.93306077e-02 5.18265404e-02
1.51178151e-01 -1.18425524e+00 5.18731594e-01 5.65349042e-01
-8.73839974e-01 -3.55246514e-01 1.87824741e-01 7.04432309e-01
1.51561752e-01 -3.80287290e-01 1.02679968e+00 -4.99603748e-02
-1.15655251e-01 6.88979328e-01 -9.27780122e-02 3.33150700e-02
1.15583193e+00 -1.56544015e-01 8.96095708e-02 -2.17012048e-01
-5.74844718e-01 7.48918280e-02 8.51121604e-01 9.44999829e-02
4.05980617e-01 -9.18994486e-01 -7.38915622e-01 4.34666216e-01
-4.32246998e-02 -1.87960401e-01 2.97245353e-01 6.30465090e-01
-4.79849100e-01 1.35014951e-01 5.72956093e-02 -1.70430571e-01
-1.50541615e+00 5.99114537e-01 5.66945255e-01 9.97935329e-03
-7.00754106e-01 1.36403275e+00 2.60696113e-01 -4.56610441e-01
7.51649976e-01 1.39018759e-01 4.04477328e-01 -1.83450758e-01
4.84818816e-01 4.89924520e-01 1.73177332e-01 -4.95480895e-01
-3.16959262e-01 -1.84346884e-01 -1.97944894e-01 -2.15396628e-01
1.37352848e+00 -1.05758868e-01 -3.05055022e-01 2.35820279e-01
1.77294934e+00 5.82366697e-02 -1.63951755e+00 -2.43620202e-01
2.07408026e-01 -7.16574073e-01 1.45397604e-01 -3.24135631e-01
-1.23202133e+00 8.45601618e-01 6.82757735e-01 2.31616527e-01
8.15867305e-01 -3.37147236e-01 6.09476686e-01 2.62878925e-01
5.76748073e-01 -1.08207715e+00 -9.88042057e-01 8.40253010e-03
5.30516028e-01 -9.93628323e-01 -8.69117379e-02 -1.54144570e-01
-5.12645543e-01 4.86543059e-01 1.93622380e-01 -1.44359961e-01
9.59972858e-01 2.86843687e-01 5.61941676e-02 1.48737520e-01
-6.37134969e-01 5.44343591e-01 -1.22768328e-01 7.92363703e-01
-2.73319781e-02 -5.29819801e-02 -1.23217866e-01 7.69132793e-01
-1.52500913e-01 -1.70352638e-01 4.70807180e-02 9.69270587e-01
5.91783151e-02 -1.11302733e+00 -5.40334694e-02 3.51554573e-01
-1.08517230e+00 3.80786955e-02 -7.68787682e-01 6.94334328e-01
8.71036649e-02 6.92691982e-01 -2.43032232e-01 -1.92541629e-01
1.48651242e-01 2.40799099e-01 2.52303898e-01 -2.80473113e-01
-1.09409511e+00 -4.78218675e-01 -1.03915356e-01 -7.81993508e-01
-1.57561108e-01 -8.75916541e-01 -6.23962820e-01 -7.90184379e-01
-1.90722480e-01 1.68618768e-01 8.07514906e-01 6.05173230e-01
7.79562414e-01 -2.55441129e-01 8.41397226e-01 -2.14284480e-01
-1.33475804e+00 -2.47309759e-01 -1.15061295e+00 6.06798530e-01
5.86212397e-01 -3.59084457e-02 -9.13675070e-01 -1.38296068e-01] | [5.93593692779541, 6.826268196105957] |
036e6ac4-ceaa-4b5d-bd29-3b2f5f1e31dc | any-shot-sequential-anomaly-detection-in | 2004.02072 | null | https://arxiv.org/abs/2004.02072v1 | https://arxiv.org/pdf/2004.02072v1.pdf | Any-Shot Sequential Anomaly Detection in Surveillance Videos | Anomaly detection in surveillance videos has been recently gaining attention. Even though the performance of state-of-the-art methods on publicly available data sets has been competitive, they demand a massive amount of training data. Also, they lack a concrete approach for continuously updating the trained model once new data is available. Furthermore, online decision making is an important but mostly neglected factor in this domain. Motivated by these research gaps, we propose an online anomaly detection method for surveillance videos using transfer learning and any-shot learning, which in turn significantly reduces the training complexity and provides a mechanism that can detect anomalies using only a few labeled nominal examples. Our proposed algorithm leverages the feature extraction power of neural network-based models for transfer learning and the any-shot learning capability of statistical detection methods. | ['Yasin Yilmaz', 'Keval Doshi'] | 2020-04-05 | null | null | null | null | ['anomaly-detection-in-surveillance-videos', 'anomaly-detection-in-surveillance-videos'] | ['computer-vision', 'methodology'] | [ 2.40803272e-01 -1.07717298e-01 -3.57306272e-01 -3.05206150e-01
-6.75232291e-01 -2.66398877e-01 4.34520900e-01 1.13906279e-01
-3.02612126e-01 4.34629619e-01 -2.38940760e-01 -2.78378129e-01
-1.45294995e-03 -7.07317710e-01 -7.47453928e-01 -7.08642721e-01
-3.86072040e-01 -7.91960657e-02 7.52610147e-01 5.53547032e-02
-1.32695720e-01 2.98650622e-01 -1.85030639e+00 1.83749869e-01
8.58282149e-01 1.44471323e+00 -4.65734839e-01 6.58734858e-01
4.50677238e-02 9.51191604e-01 -4.06759530e-01 -3.82503897e-01
5.68502247e-01 -3.73485893e-01 -2.43001238e-01 3.19784999e-01
6.22820854e-01 -1.00901425e+00 -6.35864079e-01 1.07384551e+00
1.36059240e-01 1.35756150e-01 5.38882732e-01 -1.58021271e+00
-3.46589595e-01 7.97507120e-04 -4.33612198e-01 6.97350025e-01
1.97895348e-01 3.04190934e-01 6.86870694e-01 -7.68915355e-01
3.41610998e-01 7.18341589e-01 7.52212167e-01 6.56537473e-01
-7.65933752e-01 -6.67981088e-01 4.25448060e-01 6.66186273e-01
-9.24458385e-01 -4.15718257e-01 7.88887978e-01 -6.65673316e-01
7.89035141e-01 -1.03506498e-01 9.95103598e-01 1.31667447e+00
2.04445705e-01 9.08590674e-01 5.95741153e-01 -3.13503504e-01
4.66160268e-01 -5.08074574e-02 7.03179017e-02 8.32596600e-01
5.22819161e-01 3.73927891e-01 -3.24588865e-01 -4.13289815e-01
4.74161267e-01 8.54829252e-01 1.32358605e-02 -5.52710056e-01
-5.98596573e-01 8.99829507e-01 7.36541972e-02 7.74480253e-02
-5.61423898e-01 -8.49810466e-02 8.20468485e-01 6.76122129e-01
6.58351481e-01 1.85832590e-01 -6.30932271e-01 -2.93276012e-01
-9.40400124e-01 -5.25659136e-02 6.65187418e-01 8.12815905e-01
5.85952401e-01 3.17711741e-01 -2.26171300e-01 4.57894236e-01
3.42368670e-02 5.48967600e-01 3.74193937e-01 -8.27170253e-01
1.19363181e-01 9.48478341e-01 2.33674627e-02 -1.07963657e+00
-1.55039147e-01 -1.53692916e-01 -7.88523376e-01 2.33415961e-01
4.12433058e-01 -4.23544496e-01 -7.16764390e-01 1.57047629e+00
6.19664729e-01 8.42659950e-01 1.49803581e-02 5.37768483e-01
3.08863223e-01 6.23371184e-01 8.81588273e-03 -4.23805952e-01
1.00172210e+00 -8.66688132e-01 -7.56333292e-01 -2.02216610e-01
6.08507097e-01 -2.14687183e-01 5.24660051e-01 5.45206487e-01
-5.97527802e-01 -3.94092798e-01 -9.53316629e-01 5.19589603e-01
-3.31188917e-01 -1.63854435e-02 7.51644492e-01 5.76020718e-01
-6.44837022e-01 3.96312177e-01 -1.13974023e+00 -7.18185365e-01
8.10896397e-01 7.33028874e-02 -3.71358156e-01 -4.11346108e-01
-9.18284535e-01 6.96660757e-01 3.51732045e-01 1.73274782e-02
-1.26329565e+00 -6.95652127e-01 -9.66507256e-01 -3.16592269e-02
9.60292935e-01 -2.27558181e-01 1.31202769e+00 -1.10212982e+00
-1.35663152e+00 3.96704704e-01 5.20465709e-02 -7.99760878e-01
5.95500588e-01 -5.68096340e-01 -7.34285116e-01 4.10493016e-01
-1.73154905e-01 2.77177453e-01 1.43306637e+00 -7.94641316e-01
-1.07264507e+00 -3.49718601e-01 -4.14031595e-02 -2.71872163e-01
-7.64295101e-01 1.01531714e-01 -5.88385426e-02 -7.72157133e-01
-1.88253611e-01 -7.43150294e-01 -1.66973025e-01 4.31727141e-01
-9.39070247e-03 -1.92353472e-01 1.35207844e+00 -7.17718303e-01
1.37727702e+00 -2.29269314e+00 -2.79073387e-01 1.89542860e-01
9.95693952e-02 6.93176150e-01 -3.82103361e-02 3.35523576e-01
6.12233318e-02 -4.13387388e-01 -3.49205017e-01 1.52786493e-01
-3.21257859e-01 2.33195141e-01 -4.83295590e-01 5.09616315e-01
6.14967942e-01 6.14189446e-01 -1.15996122e+00 -3.89204800e-01
3.76557022e-01 8.44148099e-02 -8.09845805e-01 5.25849283e-01
-1.62829876e-01 3.53385448e-01 -4.55698580e-01 9.14042115e-01
4.22368109e-01 -2.12598443e-01 -4.96211275e-02 1.88892916e-01
-8.77575800e-02 -3.86703312e-01 -9.58208382e-01 1.42830658e+00
1.00414798e-01 5.69200635e-01 -1.75668135e-01 -1.44119895e+00
6.40832722e-01 5.77935636e-01 8.50684822e-01 -3.33073795e-01
1.27579659e-01 7.17387497e-02 1.14633381e-01 -9.60398018e-01
-2.68282536e-02 3.08690906e-01 1.28598273e-01 3.05283666e-01
3.06430250e-01 5.02659500e-01 3.91412288e-01 2.31900308e-02
1.59434128e+00 1.16991065e-01 6.47538602e-01 2.42389634e-01
5.09064436e-01 1.59034058e-01 1.01150763e+00 8.69379818e-01
-6.12514496e-01 1.50839478e-01 5.05626559e-01 -8.45839083e-01
-1.02988017e+00 -7.90985048e-01 -2.63374615e-02 1.37287271e+00
-2.08397150e-01 -1.65789947e-01 -8.08408201e-01 -9.52605605e-01
6.66080639e-02 4.25047815e-01 -8.47798228e-01 -4.88982826e-01
-4.61696804e-01 -6.88400149e-01 3.71034116e-01 5.76025844e-01
5.18927276e-01 -1.03740489e+00 -1.06890929e+00 2.44638950e-01
1.52926162e-01 -1.14733541e+00 -1.87731907e-01 -1.73143104e-01
-1.07786524e+00 -1.62984610e+00 -4.66045856e-01 -4.62327063e-01
8.44974160e-01 4.22039896e-01 7.47600198e-01 9.51197818e-02
-6.56812370e-01 7.80882955e-01 -6.63448095e-01 -8.54174674e-01
-2.57951826e-01 -2.30402783e-01 2.63889372e-01 4.29390132e-01
7.96587229e-01 -3.67711067e-01 -5.31444669e-01 2.45483041e-01
-1.05800605e+00 -4.29661334e-01 7.44660556e-01 9.79210198e-01
2.44559452e-01 -1.09488122e-01 8.17909718e-01 -7.54024565e-01
2.19170541e-01 -7.55445957e-01 -9.87782180e-01 2.64320493e-01
-4.83099431e-01 -2.68090427e-01 6.25269949e-01 -5.52303672e-01
-9.58885074e-01 1.27827395e-02 2.72788137e-01 -8.89284849e-01
-3.81243825e-01 2.58469850e-01 2.25950167e-01 -2.35433914e-02
6.68976724e-01 1.87708765e-01 2.98400164e-01 -1.41385496e-01
-4.96929220e-04 4.45106357e-01 3.60917687e-01 -1.08336568e-01
8.94996345e-01 6.90142453e-01 -1.13342129e-01 -9.15833890e-01
-1.11891329e+00 -6.65744126e-01 -6.58105075e-01 -3.85692269e-01
9.03194010e-01 -9.09791410e-01 -3.81580979e-01 8.39287579e-01
-9.57211375e-01 -4.05058451e-02 -4.73758250e-01 5.20417392e-01
-2.70625114e-01 4.91662234e-01 -4.44041520e-01 -1.10270226e+00
-3.33515346e-01 -7.22451627e-01 6.84730172e-01 7.08110780e-02
7.83551708e-02 -8.09816957e-01 1.52756736e-01 2.14468855e-02
5.36884367e-01 5.00453532e-01 4.64520574e-01 -1.09754109e+00
-6.09356046e-01 -8.73555243e-01 -9.79342014e-02 5.46610177e-01
3.71770829e-01 2.50560671e-01 -1.05542266e+00 -5.17965317e-01
-1.15589714e-02 -2.73528337e-01 8.66732657e-01 5.04737258e-01
1.31879234e+00 -5.05537152e-01 -3.51013333e-01 3.83754045e-01
9.87832546e-01 3.74847263e-01 2.88963884e-01 2.22350597e-01
4.85761046e-01 3.92983317e-01 7.85450518e-01 8.58560324e-01
1.07733451e-01 2.82584518e-01 5.31649888e-01 -2.25996226e-03
3.93463165e-01 -8.20482597e-02 5.66603541e-01 5.87281704e-01
-1.04094565e-01 1.14032567e-01 -8.60597253e-01 5.26217401e-01
-2.22546315e+00 -1.47624743e+00 3.93996894e-01 2.30807877e+00
4.09801662e-01 9.16602314e-02 2.27199599e-01 2.27305949e-01
6.42169595e-01 -4.16432992e-02 -9.26821768e-01 -7.26682320e-03
1.49728492e-01 -1.75038338e-01 2.59036094e-01 -1.58129409e-01
-1.65483379e+00 6.53356254e-01 6.74004126e+00 4.64139044e-01
-1.01279616e+00 4.55668941e-02 4.08440024e-01 -2.02048749e-01
5.61165214e-01 -3.38641077e-01 -4.81297404e-01 6.47366643e-01
9.28229511e-01 -7.23545626e-02 1.33536041e-01 1.41603768e+00
1.22107975e-01 -1.40483126e-01 -1.26440835e+00 9.89176214e-01
3.70089173e-01 -1.08707047e+00 1.38433114e-01 -1.05120659e-01
7.55411685e-01 3.58779654e-02 -2.91341823e-02 5.19184172e-01
7.74397925e-02 -5.69348991e-01 1.51952714e-01 4.77575034e-01
4.74832654e-01 -6.70890987e-01 1.03371632e+00 4.12086159e-01
-9.43117440e-01 -7.71204293e-01 -4.45572764e-01 -1.73811615e-01
-5.73635809e-02 5.64766645e-01 -5.51800489e-01 3.95833880e-01
9.70188022e-01 9.40614164e-01 -5.23516893e-01 1.38630915e+00
-6.22374564e-02 9.39855456e-01 -2.33843401e-01 2.15255156e-01
2.36625716e-01 -1.57858133e-01 5.96375823e-01 9.96455252e-01
5.22721589e-01 1.12719983e-01 6.81918502e-01 2.17269048e-01
1.23006485e-01 -4.59441096e-02 -1.04013777e+00 -8.12746137e-02
3.09228420e-01 1.17985749e+00 -3.09214681e-01 -5.13076782e-01
-9.39049900e-01 7.25525796e-01 8.04083347e-02 2.51774013e-01
-8.24208319e-01 -2.52373070e-01 7.51993179e-01 2.96248086e-02
5.45085967e-01 5.29698133e-02 3.61004293e-01 -1.38639534e+00
1.43559173e-01 -9.86647487e-01 8.09402406e-01 -3.07818372e-02
-1.68614411e+00 2.36414775e-01 1.47923172e-01 -1.70711386e+00
-5.85413396e-01 -7.85351694e-01 -9.13341284e-01 -1.89521033e-02
-1.59345829e+00 -9.03670490e-01 -5.99050522e-01 7.80376256e-01
7.14405477e-01 -6.46979630e-01 8.93276334e-01 3.71294856e-01
-8.45238090e-01 5.08916914e-01 1.95756748e-01 5.24830282e-01
7.30330646e-01 -8.22024584e-01 1.88339472e-01 1.25088525e+00
-9.51397344e-02 1.37960449e-01 4.95605528e-01 -5.93976974e-01
-1.53434539e+00 -1.19845200e+00 1.21312723e-01 -4.58463788e-01
9.34350729e-01 -1.96424022e-01 -1.30316722e+00 7.44237304e-01
-1.18127085e-01 6.02945566e-01 9.56701159e-01 -3.66578326e-02
-4.22919005e-01 -3.34341675e-01 -1.23562717e+00 4.11810428e-01
9.47563469e-01 -3.05724025e-01 -6.76405728e-01 2.69499034e-01
4.84468102e-01 -1.53718725e-01 -4.05223459e-01 5.45978785e-01
4.18164134e-01 -9.58218932e-01 5.70674062e-01 -1.04416323e+00
1.71654209e-01 -1.73371419e-01 5.99715672e-02 -1.16097093e+00
-1.70193896e-01 -5.85267484e-01 -7.88378298e-01 8.42359543e-01
1.92127332e-01 -7.83371925e-01 7.54247665e-01 7.55976498e-01
-1.05411991e-01 -6.43885672e-01 -1.06147313e+00 -1.01833951e+00
-4.94364470e-01 -6.25934720e-01 1.65556490e-01 8.77506375e-01
-4.11726683e-02 -2.31821612e-01 -7.89312720e-01 3.07243943e-01
8.52655947e-01 -1.94127172e-01 8.42738450e-01 -1.54175007e+00
-1.47854894e-01 -2.14489829e-02 -8.11882675e-01 -5.56239426e-01
7.74374604e-02 -4.73808855e-01 -1.70251988e-02 -1.05015182e+00
2.09552765e-01 4.81102392e-02 -7.07418203e-01 5.64342082e-01
-2.79227972e-01 -7.98565224e-02 -2.24900931e-01 6.64859489e-02
-1.00417352e+00 5.78866720e-01 5.50354481e-01 3.18801180e-02
-1.68705076e-01 6.87566996e-02 -1.90443158e-01 1.07487762e+00
7.43490875e-01 -4.40648824e-01 -3.70392710e-01 -1.92986578e-01
-1.63117751e-01 -2.54002929e-01 5.01417935e-01 -1.28116095e+00
3.20523947e-01 -2.07954451e-01 4.81861830e-01 -4.14491773e-01
1.56019375e-01 -1.26297069e+00 -4.31511998e-01 6.48100376e-01
-1.21360056e-01 1.33131698e-01 2.68258154e-01 1.19690716e+00
-1.67651758e-01 1.59891427e-03 6.90697789e-01 2.54285615e-02
-1.05929792e+00 8.38188827e-01 -5.68031847e-01 1.45314112e-01
1.63835084e+00 -3.12961042e-01 -1.17951199e-01 -2.67927855e-01
-3.80214721e-01 2.67431557e-01 2.05125183e-01 6.10948682e-01
6.79684401e-01 -1.14223754e+00 -6.08351886e-01 5.07116556e-01
5.21676064e-01 -1.61766171e-01 4.24267769e-01 8.45797360e-01
-1.52384222e-01 1.51343331e-01 -3.88810128e-01 -6.75335050e-01
-1.14236081e+00 9.95973706e-01 5.21204509e-02 -1.48628075e-02
-8.44905972e-01 4.09183413e-01 -3.46841589e-02 -4.00526524e-02
4.21700388e-01 -1.89135388e-01 -1.58925116e-01 2.78265774e-02
9.33401763e-01 6.26292229e-01 -1.21770240e-01 -1.56244636e-01
-2.13655978e-01 1.57634318e-01 -2.85715520e-01 3.41652691e-01
1.38187480e+00 3.74475002e-01 1.41421452e-01 5.39236128e-01
7.44155884e-01 -5.45844853e-01 -1.66663527e+00 -3.80407274e-01
2.79180467e-01 -7.10744321e-01 -9.14073437e-02 -2.60776281e-01
-9.80064094e-01 8.54528785e-01 7.90456593e-01 3.75477970e-01
1.26219428e+00 -2.21671402e-01 7.92956769e-01 9.03706908e-01
3.24236006e-01 -1.31338191e+00 4.27955240e-01 4.59269553e-01
4.66565818e-01 -1.88292456e+00 -2.39869326e-01 -2.60690480e-01
-4.56018507e-01 1.07358229e+00 8.64310920e-01 -3.23267728e-01
8.78431082e-01 2.11029485e-01 -7.47086033e-02 -5.94197698e-02
-9.18648243e-01 -3.54806632e-02 3.16900045e-01 7.60527790e-01
3.05567756e-02 -3.29265296e-01 2.63661683e-01 3.14879149e-01
5.90676427e-01 2.48136163e-01 3.00647438e-01 1.15965044e+00
-6.90306783e-01 -6.03628695e-01 -2.26013422e-01 8.94051492e-01
-6.29820228e-01 8.07242468e-02 7.52893388e-02 6.04255438e-01
-4.45193873e-04 1.00398529e+00 2.39116043e-01 -6.95898086e-02
3.21903467e-01 3.60032439e-01 1.34617627e-01 -5.55739701e-01
-1.12312846e-01 -2.73605049e-01 -2.93342561e-01 -8.67388368e-01
-4.91584331e-01 -7.06176281e-01 -6.77992404e-01 -9.37520638e-02
-2.91378736e-01 -7.41890147e-02 2.27433428e-01 1.12743497e+00
5.96378982e-01 3.80040795e-01 6.98634982e-01 -5.92847824e-01
-8.14245105e-01 -9.59893644e-01 -3.73809069e-01 4.52882409e-01
5.91720939e-01 -8.50262821e-01 -3.26389134e-01 1.91655785e-01] | [7.877026081085205, 1.571636438369751] |
c01825e7-f7c4-43ba-bccd-378f8248ed1b | mitigating-gender-bias-in-captioning-systems | 2006.08315 | null | https://arxiv.org/abs/2006.08315v7 | https://arxiv.org/pdf/2006.08315v7.pdf | Mitigating Gender Bias in Captioning Systems | Image captioning has made substantial progress with huge supporting image collections sourced from the web. However, recent studies have pointed out that captioning datasets, such as COCO, contain gender bias found in web corpora. As a result, learning models could heavily rely on the learned priors and image context for gender identification, leading to incorrect or even offensive errors. To encourage models to learn correct gender features, we reorganize the COCO dataset and present two new splits COCO-GB V1 and V2 datasets where the train and test sets have different gender-context joint distribution. Models relying on contextual cues will suffer from huge gender prediction errors on the anti-stereotypical test data. Benchmarking experiments reveal that most captioning models learn gender bias, leading to high gender prediction errors, especially for women. To alleviate the unwanted bias, we propose a new Guided Attention Image Captioning model (GAIC) which provides self-guidance on visual attention to encourage the model to capture correct gender visual evidence. Experimental results validate that GAIC can significantly reduce gender prediction errors with a competitive caption quality. Our codes and the designed benchmark datasets are available at https://github.com/datamllab/Mitigating_Gender_Bias_In_Captioning_System. | ['Na Zou', 'Yuening Li', 'Ruixiang Tang', 'Xia Hu', 'Zirui Liu', 'Mengnan Du'] | 2020-06-15 | null | null | null | null | ['gender-prediction'] | ['computer-vision'] | [ 1.42391354e-01 4.23595250e-01 -3.75511348e-01 -8.61243248e-01
-5.31281531e-01 -5.32605648e-01 5.95926046e-01 -9.56722647e-02
-2.87438005e-01 6.20895624e-01 1.12517752e-01 -2.97319461e-02
5.34984291e-01 -5.84408224e-01 -1.06047893e+00 -5.89490652e-01
5.27688980e-01 7.28320479e-01 -2.35732317e-01 -1.62161782e-01
4.57461566e-01 -3.68922174e-01 -1.66632295e+00 5.31763971e-01
9.99661565e-01 1.05973363e+00 4.87271845e-02 3.90839875e-01
-1.99343115e-01 5.29665828e-01 -6.67048037e-01 -1.26177931e+00
1.30223438e-01 -3.53538454e-01 -5.69232166e-01 1.16294669e-03
1.02380145e+00 -1.94072664e-01 -2.05319136e-01 1.29162884e+00
6.97039604e-01 -3.61843079e-01 7.70620108e-01 -1.99719381e+00
-1.27067208e+00 6.11334443e-01 -1.10401714e+00 2.28740629e-02
2.50022441e-01 3.91998589e-01 8.89136970e-01 -1.11121583e+00
4.84589070e-01 1.63689446e+00 4.72518206e-01 1.09376907e+00
-1.18740988e+00 -1.23200727e+00 1.21628493e-01 4.81902778e-01
-1.31992877e+00 -2.91980952e-01 7.50648320e-01 -3.48533213e-01
1.02222227e-01 5.40320575e-01 5.09132922e-01 1.92688537e+00
-2.09080264e-01 9.85883236e-01 1.22345018e+00 -2.07195744e-01
-2.65379906e-01 4.21857387e-01 -1.31317765e-01 4.80199784e-01
4.42733258e-01 -8.64779949e-02 -6.89689875e-01 2.13336289e-01
5.22310197e-01 -3.20529729e-01 -6.56834990e-02 -3.41287822e-01
-1.06715381e+00 8.19179654e-01 4.77580488e-01 -3.02584887e-01
9.86855477e-02 1.64144859e-01 5.85673392e-01 -1.28447190e-01
4.10799146e-01 2.99999893e-01 -1.97633147e-01 -7.59972185e-02
-5.18669307e-01 4.38093126e-01 2.19579697e-01 9.69324112e-01
6.28207505e-01 -1.88751370e-01 -6.84623599e-01 1.14018190e+00
4.06471431e-01 1.04888654e+00 4.44401383e-01 -8.24835420e-01
9.17684376e-01 6.54157460e-01 -4.05800436e-03 -1.42786014e+00
-4.20592502e-02 -1.77142531e-01 -7.99607575e-01 -3.29467088e-01
5.57781696e-01 1.31824583e-01 -1.24901164e+00 2.13284755e+00
2.04947650e-01 -1.83792725e-01 -2.21548066e-01 1.17012298e+00
1.33574665e+00 5.55080116e-01 6.73417866e-01 1.11556813e-01
1.69163692e+00 -9.39551413e-01 -7.50845015e-01 -8.20529699e-01
2.78666437e-01 -8.27695906e-01 1.42937982e+00 7.24697858e-02
-9.41277683e-01 -5.89968324e-01 -7.01234758e-01 -9.39541459e-02
-4.28354263e-01 2.64840871e-01 6.43112779e-01 8.23269129e-01
-7.20112026e-01 -2.34462935e-02 -3.29715580e-01 -3.84272069e-01
8.25906634e-01 3.43120188e-01 -2.43447930e-01 -2.05336154e-01
-1.09007454e+00 5.82228243e-01 3.52192938e-01 2.47514918e-01
-8.13661397e-01 -6.67545497e-01 -1.11766660e+00 -7.34598115e-02
1.62128732e-01 -6.27532065e-01 1.18472815e+00 -1.45057487e+00
-6.95536911e-01 1.38537526e+00 -2.15311930e-01 -3.42307761e-02
6.71471298e-01 5.05014462e-03 -1.88848302e-01 9.21090320e-02
4.42823887e-01 1.58721602e+00 8.79592896e-01 -1.71477675e+00
-3.56827945e-01 -5.84099531e-01 -1.63632065e-01 3.26419443e-01
-6.09006703e-01 5.70977181e-02 -7.03476310e-01 -7.07245111e-01
-9.11470205e-02 -1.08918226e+00 1.15720972e-01 -2.73460835e-01
-6.17677152e-01 -3.04328531e-01 6.20348215e-01 -6.89907432e-01
1.09618962e+00 -2.02526617e+00 -1.05649114e-01 -7.21168816e-02
3.09818745e-01 2.79429145e-02 -2.11019620e-01 1.59866139e-02
-2.11691692e-01 3.37379456e-01 -3.38066928e-02 -4.43362027e-01
1.86700493e-01 2.62039185e-01 -3.37847382e-01 1.77274197e-01
2.56859064e-01 1.03735149e+00 -8.92741919e-01 -1.04183698e+00
-5.49365096e-02 3.30373079e-01 -5.84649026e-01 4.01849836e-01
-1.63383573e-01 6.33445382e-01 -1.46680310e-01 9.01116371e-01
1.10462046e+00 -3.76606137e-01 1.92448631e-01 -5.04633486e-01
3.50043416e-01 -3.36999416e-01 -6.07621849e-01 1.27571368e+00
-2.42980778e-01 6.57462001e-01 -5.36045544e-02 -7.62294471e-01
1.01974511e+00 -2.96424925e-02 1.50252685e-01 -1.15872455e+00
3.92706752e-01 2.74150759e-01 -5.85205890e-02 -8.67250621e-01
6.62773073e-01 1.14037804e-02 -4.05323625e-01 -2.79114451e-02
1.40819177e-01 8.50066245e-02 4.17733669e-01 3.38797539e-01
2.26675451e-01 6.71512485e-02 -4.38659161e-01 -2.64606979e-02
4.26921248e-01 -1.29000589e-01 7.71562517e-01 6.58182919e-01
-6.04562998e-01 1.09051299e+00 9.10762906e-01 -5.05539477e-01
-1.15745234e+00 -9.09828961e-01 -1.89473167e-01 1.58626497e+00
5.28073132e-01 -4.81920242e-01 -9.55452025e-01 -7.45873809e-01
-6.03114665e-02 6.04203403e-01 -1.14645898e+00 -2.54853398e-01
-4.64580357e-01 -1.13440239e+00 4.66958851e-01 5.96217811e-01
7.00377285e-01 -1.18722081e+00 -1.71608374e-01 -4.59438443e-01
-7.33732045e-01 -1.15401995e+00 -6.15312755e-01 -2.92272031e-01
-4.20734644e-01 -1.12166071e+00 -1.06508744e+00 -9.36364710e-01
9.85467196e-01 -1.67280644e-01 1.57110298e+00 1.00265861e-01
-6.37082160e-02 4.04958099e-01 -2.28568867e-01 -7.76836395e-01
-2.69210011e-01 6.59424961e-02 -2.00008616e-01 2.65665144e-01
7.84948587e-01 -2.07460329e-01 -9.40220177e-01 6.06010139e-01
-5.89464664e-01 4.93157297e-01 4.71700281e-01 9.30683196e-01
5.60981572e-01 -8.98848712e-01 6.38726830e-01 -9.39050436e-01
4.38168079e-01 -7.76292980e-01 -3.84008706e-01 2.82613486e-01
-6.96131527e-01 -2.59001076e-01 2.99538583e-01 -4.96184915e-01
-1.06661832e+00 -3.94674763e-02 6.55751303e-02 -7.70245433e-01
-1.07121788e-01 7.56792128e-02 -3.83412421e-01 1.65672943e-01
6.15224600e-01 2.66518146e-01 5.43844420e-03 -2.69573390e-01
6.96370006e-02 7.41924644e-01 6.32884562e-01 -9.42239761e-01
6.25452220e-01 2.05412328e-01 -2.57190853e-01 -1.96186543e-01
-9.96567249e-01 -9.19529200e-02 -1.55879959e-01 -6.46955132e-01
1.06076598e+00 -1.17307389e+00 -8.93554032e-01 4.81728405e-01
-9.88798559e-01 -9.24312770e-02 3.23714286e-01 5.03938459e-02
-5.69093525e-01 5.64329736e-02 -4.64743257e-01 -6.01786852e-01
-3.78296077e-01 -1.28773820e+00 1.35849094e+00 6.64183080e-01
-1.14493534e-01 -6.32761300e-01 -5.60768694e-02 9.78529394e-01
2.28624538e-01 2.44244456e-01 7.22814679e-01 -6.59433484e-01
-3.19017559e-01 1.67577326e-01 -8.43095720e-01 4.19838205e-02
-3.99423301e-01 1.37248054e-01 -1.18647683e+00 -2.17558034e-02
-6.43097281e-01 -7.48288810e-01 8.81878853e-01 2.38289416e-01
1.78819561e+00 -4.47233617e-01 -2.96829402e-01 5.82041144e-01
1.18642712e+00 -1.10870190e-01 7.04037070e-01 4.34040606e-01
1.14537358e+00 8.22863460e-01 8.74470592e-01 3.05921227e-01
8.95423174e-01 5.36494672e-01 7.73191869e-01 -1.45432070e-01
-1.19883887e-01 -5.71099102e-01 -1.19777713e-02 3.52104574e-01
-1.26732618e-01 -2.75956452e-01 -1.19981623e+00 7.82429755e-01
-1.96847951e+00 -8.99121940e-01 -2.42706269e-01 1.97471261e+00
9.73854005e-01 -1.50030807e-01 7.63047859e-02 -3.02288085e-01
1.18194735e+00 1.77168056e-01 -5.25408149e-01 -5.48142076e-01
-1.79311693e-01 -3.07566047e-01 5.26742101e-01 -4.08553779e-02
-1.22664094e+00 8.54674399e-01 5.51571798e+00 1.02249527e+00
-1.15525544e+00 -1.32748736e-02 1.54213595e+00 -1.90263033e-01
-4.37494129e-01 -4.59385246e-01 -7.67461061e-01 1.05977178e+00
7.96775997e-01 -5.49862310e-02 1.08426698e-01 1.01866758e+00
-1.27643004e-01 -1.18589364e-02 -9.09306824e-01 1.62489355e+00
3.42784494e-01 -9.79658544e-01 3.45087945e-01 7.09137693e-03
6.71367705e-01 -3.53531241e-01 6.81591988e-01 6.75407946e-01
-7.82112777e-02 -1.34120333e+00 9.44016218e-01 1.31459519e-01
1.05138385e+00 -8.36499214e-01 8.62086296e-01 -1.33469030e-01
-7.98534870e-01 -2.01921269e-01 -3.30666035e-01 2.37548158e-01
-9.47381705e-02 2.95144260e-01 -7.94344306e-01 1.16034575e-01
1.17058313e+00 5.65185785e-01 -1.30877745e+00 8.81313264e-01
-1.69231594e-01 5.58077574e-01 1.95366573e-02 -1.04944343e-02
-8.38244520e-03 -1.07048683e-01 2.64493644e-01 9.41276670e-01
2.72241324e-01 -2.11149082e-01 -5.71347997e-02 8.20078552e-01
-3.54773283e-01 3.44114989e-01 -4.57698345e-01 -5.94803542e-02
3.95138621e-01 1.40694737e+00 -7.69861341e-01 -3.62104952e-01
-3.38576227e-01 6.66508675e-01 3.36275518e-01 3.27236980e-01
-1.26677597e+00 1.01388738e-01 6.00790083e-01 2.58251578e-01
6.43614456e-02 5.20483255e-01 -5.90930760e-01 -1.10264945e+00
4.00060900e-02 -9.84940469e-01 5.39209723e-01 -1.11239052e+00
-1.34233451e+00 3.72460663e-01 1.47128910e-01 -1.07577848e+00
2.54322719e-02 -6.50988340e-01 -4.95181888e-01 5.09019196e-01
-1.19113660e+00 -1.59310377e+00 -6.34011090e-01 4.44054782e-01
5.45605659e-01 -1.76277697e-01 3.73500407e-01 6.63510203e-01
-8.93116653e-01 1.14774621e+00 -5.49801946e-01 3.23287785e-01
1.39536107e+00 -1.30334818e+00 -3.38536352e-02 4.04291421e-01
-1.87660530e-01 4.38329548e-01 1.06474936e+00 -6.38640821e-01
-1.08115697e+00 -1.07925105e+00 8.51480901e-01 -6.94602251e-01
2.89044112e-01 -6.28325641e-01 -6.92574501e-01 7.35079765e-01
4.57119107e-01 1.17098384e-01 8.11893463e-01 2.48399839e-01
-7.50042379e-01 -1.69579819e-01 -1.19141006e+00 7.01542199e-01
1.08042359e+00 -1.13596290e-01 -2.87426531e-01 3.89270961e-01
4.49081331e-01 -6.37313724e-01 -4.89514172e-01 5.66699982e-01
6.60645247e-01 -1.07092893e+00 1.02176285e+00 -5.14638543e-01
1.28116810e+00 -4.95075174e-02 -4.86812778e-02 -1.00277758e+00
-1.06410548e-01 -7.72604719e-02 -3.43892910e-02 1.75856817e+00
4.84971017e-01 -1.33997977e-01 1.02587891e+00 8.17445815e-01
5.46446070e-02 -7.05402493e-01 -6.41972482e-01 -3.84158760e-01
1.63495883e-01 -1.31453350e-01 7.72734344e-01 9.88859117e-01
-2.79182673e-01 4.86610353e-01 -5.38596928e-01 7.85035938e-02
7.60055125e-01 2.69073457e-01 9.17335689e-01 -9.84174907e-01
1.29271939e-01 -4.75170523e-01 -3.46207917e-01 -4.01872665e-01
2.42169753e-01 -8.08437228e-01 3.61136310e-02 -1.14405584e+00
8.23506951e-01 -2.86003858e-01 -1.45288631e-01 7.20568597e-01
-7.24483609e-01 1.20643461e+00 2.40685031e-01 6.79550096e-02
-1.04170847e+00 7.44490206e-01 1.54783022e+00 -4.40745175e-01
3.57839286e-01 -4.04496908e-01 -1.08400202e+00 6.79854691e-01
7.87745714e-01 -5.47852755e-01 -3.97129774e-01 -5.61059117e-01
6.99695349e-01 -2.34588161e-01 5.17097771e-01 -6.29986107e-01
-1.73651755e-01 -3.44447643e-01 8.88619781e-01 -5.47421873e-01
3.38276148e-01 -4.70200509e-01 -1.60268322e-01 9.26551968e-02
-2.84501731e-01 1.55130625e-01 2.23451659e-01 4.02181625e-01
-3.69292974e-01 -3.57528552e-02 6.65288568e-01 -9.19979811e-03
-7.27059841e-01 4.00113493e-01 4.79341038e-02 3.15040261e-01
9.95038509e-01 -2.52456963e-01 -5.86206973e-01 -4.05355275e-01
-6.84688270e-01 6.69086993e-01 5.76575696e-01 8.35224807e-01
3.33726764e-01 -1.69345415e+00 -9.07671332e-01 2.66696960e-02
7.13097394e-01 -1.71588644e-01 5.61470568e-01 6.79766834e-01
-3.35601926e-01 2.57303059e-01 -6.59206748e-01 -8.86571229e-01
-1.42071497e+00 6.23843372e-01 1.54006049e-01 3.68444622e-02
1.89284280e-01 1.16632009e+00 8.35251808e-01 -6.58297300e-01
5.92926852e-02 2.47896582e-01 -4.81567830e-01 4.03515816e-01
3.82250071e-01 1.10866956e-01 -2.65542507e-01 -8.64093781e-01
-3.53833765e-01 5.49453557e-01 -1.98171496e-01 7.30641335e-02
1.09748673e+00 -2.69163102e-01 -4.29443829e-02 1.68505609e-02
9.62358296e-01 -8.93207341e-02 -1.31827629e+00 3.76173049e-01
-3.41490269e-01 -8.18450689e-01 -4.89288568e-01 -1.02090108e+00
-1.46596563e+00 8.55195343e-01 8.14557731e-01 8.90323063e-05
1.09481311e+00 1.08706310e-01 5.80989540e-01 -2.09406883e-01
5.72169982e-02 -1.19565833e+00 3.09709996e-01 2.23073170e-01
9.85719979e-01 -1.89664960e+00 -2.55567372e-01 -5.45886397e-01
-9.65543568e-01 6.59959733e-01 1.49100149e+00 1.90822542e-01
5.69716655e-02 -2.76837409e-01 2.78243005e-01 -1.28566056e-01
-5.74098170e-01 4.38852981e-02 3.36306512e-01 8.33627999e-01
5.09458244e-01 9.60932747e-02 -4.93092000e-01 1.15094650e+00
-4.82959360e-01 -3.10921043e-01 4.77838516e-01 2.55920798e-01
3.36745419e-02 -1.00617838e+00 -8.07784557e-01 4.51596707e-01
-5.24012029e-01 -1.09756418e-01 -4.51080769e-01 5.90736210e-01
3.94950181e-01 6.65064991e-01 3.69667292e-01 -3.36879075e-01
8.94101933e-02 1.03978947e-01 6.10467196e-01 -2.74620861e-01
-4.26352710e-01 -1.12193443e-01 1.10837966e-01 -5.69649398e-01
-1.99705571e-01 -5.08056879e-01 -7.72264540e-01 -3.00697267e-01
1.94749221e-01 -1.44423665e-02 5.50207496e-01 4.04515773e-01
4.48393136e-01 2.84183443e-01 3.53015572e-01 -9.45898294e-01
-2.85242554e-02 -1.08885694e+00 -2.72031456e-01 9.31438148e-01
-1.44070044e-01 -9.18064296e-01 -2.49122098e-01 2.37338424e-01] | [10.960569381713867, 1.2547354698181152] |
13250c41-09a0-417c-90bd-449606a2e3df | modularized-interaction-network-for-named | null | null | https://aclanthology.org/2021.acl-long.17 | https://aclanthology.org/2021.acl-long.17.pdf | Modularized Interaction Network for Named Entity Recognition | Although the existing Named Entity Recognition (NER) models have achieved promising performance, they suffer from certain drawbacks. The sequence labeling-based NER models do not perform well in recognizing long entities as they focus only on word-level information, while the segment-based NER models which focus on processing segment instead of single word are unable to capture the word-level dependencies within the segment. Moreover, as boundary detection and type prediction may cooperate with each other for the NER task, it is also important for the two sub-tasks to mutually reinforce each other by sharing their information. In this paper, we propose a novel Modularized Interaction Network (MIN) model which utilizes both segment-level information and word-level dependencies, and incorporates an interaction mechanism to support information sharing between boundary detection and type prediction to enhance the performance for the NER task. We have conducted extensive experiments based on three NER benchmark datasets. The performance results have shown that the proposed MIN model has outperformed the current state-of-the-art models. | ['Meihuizi Jia', 'Guoxiu He', 'Jing Xu', 'Dandan song', 'Lejian Liao', 'Siu Cheung Hui', 'Zheng Wang', 'Fei Li'] | 2021-08-01 | null | null | null | acl-2021-5 | ['type-prediction', 'boundary-detection'] | ['computer-code', 'computer-vision'] | [-7.13026151e-02 -5.19616008e-02 -1.69601068e-01 -4.53209370e-01
-3.54981601e-01 -4.97599870e-01 2.55929470e-01 5.01976073e-01
-7.46627212e-01 6.90819442e-01 4.15227711e-01 -2.92326361e-01
7.09441826e-02 -7.93854356e-01 -2.62323529e-01 -3.45522165e-01
2.94888820e-02 1.78504944e-01 5.60082018e-01 -1.99503884e-01
3.24700683e-01 2.23413244e-01 -1.19171941e+00 1.47891209e-01
1.38048565e+00 5.27863503e-01 4.78334844e-01 1.72541544e-01
-8.31664443e-01 5.90225399e-01 -5.05885720e-01 -2.79637367e-01
-9.57799256e-02 -3.95881802e-01 -8.13801706e-01 -4.28148419e-01
-3.41504306e-01 1.55179441e-01 -9.75701958e-02 1.11590159e+00
5.06336808e-01 2.51128942e-01 4.75160927e-01 -7.69677103e-01
-5.67062736e-01 1.00338411e+00 -3.94038171e-01 2.61750013e-01
2.56953448e-01 -4.67194587e-01 1.14594686e+00 -8.44448090e-01
6.22786522e-01 8.28315437e-01 7.68873751e-01 6.39431298e-01
-6.03739262e-01 -5.62000036e-01 3.64723742e-01 1.52135462e-01
-1.37791812e+00 -9.98167768e-02 7.09441543e-01 -2.01661393e-01
1.23839760e+00 -3.69123518e-02 2.36234441e-01 5.61331034e-01
1.48877963e-01 7.20609188e-01 8.85103643e-01 -5.16569674e-01
5.51688857e-02 4.85292040e-02 7.48626888e-01 6.06616437e-01
2.67892808e-01 -2.54689723e-01 -1.29126340e-01 2.40587760e-02
5.02975643e-01 -3.93040404e-02 -2.16912031e-01 1.93305328e-01
-1.08229899e+00 5.75995803e-01 3.50239873e-01 1.01633155e+00
-6.14912391e-01 -3.98264050e-01 4.12298739e-01 -1.42886281e-01
2.52774179e-01 2.82933623e-01 -6.51487768e-01 -1.29627138e-01
-9.49336171e-01 -4.53076094e-01 9.83452737e-01 8.90995979e-01
8.29657435e-01 -1.64218009e-01 -4.09697264e-01 8.19416821e-01
5.01405001e-01 8.48458800e-03 7.02528715e-01 -9.11310241e-02
6.51321590e-01 1.14680505e+00 2.43641689e-01 -9.05970573e-01
-7.83172786e-01 -4.74597901e-01 -9.11935866e-01 -4.02707607e-01
2.36749589e-01 -4.91351902e-01 -1.06668699e+00 1.79507923e+00
3.37321460e-01 3.11165482e-01 3.31108809e-01 6.17433131e-01
1.10771751e+00 9.86980200e-01 6.74453735e-01 -2.61573017e-01
1.52236485e+00 -1.15995717e+00 -9.77823734e-01 -1.87435448e-01
1.01766431e+00 -8.61284196e-01 6.64220631e-01 -1.26033425e-02
-8.14655066e-01 -7.34410763e-01 -1.06286740e+00 8.37161317e-02
-6.82048261e-01 3.29321235e-01 3.33633661e-01 7.77687609e-01
-6.94473863e-01 4.98965025e-01 -7.67841041e-01 -4.68033582e-01
-8.68708268e-02 2.84019709e-01 -3.88410121e-01 9.49624628e-02
-1.66604948e+00 8.98758292e-01 9.10241902e-01 5.90161562e-01
-3.37547600e-01 -3.56897384e-01 -7.79688597e-01 4.02401537e-01
2.56856024e-01 -2.92603016e-01 1.00348914e+00 -6.41388476e-01
-1.18421566e+00 3.46786648e-01 -3.34661752e-01 -1.97214618e-01
7.54167587e-02 -1.34652317e-01 -9.45172608e-01 -1.54159933e-01
9.36135184e-03 5.11215687e-01 -8.44339877e-02 -1.16103613e+00
-9.48696792e-01 -2.65042424e-01 -6.22802135e-03 3.84692848e-01
-6.21599257e-01 2.71486253e-01 -4.59919125e-01 -5.73083162e-01
9.08687152e-03 -5.34885585e-01 -3.37750018e-01 -8.08448613e-01
-5.40760636e-01 -6.33740842e-01 6.69018686e-01 -6.71726406e-01
1.89231479e+00 -1.96295476e+00 -1.09056257e-01 1.50336877e-01
-1.13463871e-01 8.18252265e-01 -2.15153605e-01 8.06299269e-01
-4.39993441e-02 5.62620997e-01 -1.63470775e-01 -2.59519100e-01
-8.48369375e-02 1.64587557e-01 1.87636986e-01 1.69946291e-02
1.81036592e-01 6.52198493e-01 -8.45560431e-01 -7.73640335e-01
-4.35035713e-02 4.83271867e-01 -1.29782036e-01 2.92625457e-01
-9.22254026e-02 3.85264874e-01 -6.11677408e-01 3.14261377e-01
6.72920525e-01 -3.67566198e-02 4.38424200e-01 -1.98888868e-01
-5.01689136e-01 4.23919231e-01 -1.49707782e+00 1.61339021e+00
-3.30118895e-01 -2.93814242e-02 4.60016367e-04 -9.33270752e-01
1.13551962e+00 7.53010452e-01 2.46336952e-01 -4.74899054e-01
6.49320856e-02 2.52716601e-01 8.14915076e-02 -5.38636684e-01
5.93959749e-01 -1.32189214e-01 -1.62511036e-01 3.32290560e-01
3.46967205e-02 6.95224166e-01 4.59046006e-01 3.17016244e-03
1.18038845e+00 5.26266284e-02 4.12666887e-01 -1.27068609e-01
9.53573346e-01 1.31282941e-01 1.25049925e+00 5.18323243e-01
-2.45141461e-01 1.20431542e-01 2.21874088e-01 -1.61522627e-01
-8.50973368e-01 -6.16526306e-01 2.07418595e-02 1.00253284e+00
4.75335151e-01 -5.93554616e-01 -8.99415255e-01 -1.08433044e+00
-7.51856327e-01 7.74641097e-01 -3.82678986e-01 1.07910424e-01
-7.68598318e-01 -7.65190899e-01 9.25545037e-01 7.98374355e-01
8.71286571e-01 -1.24018300e+00 -6.66540489e-02 6.26753509e-01
-5.30222297e-01 -1.05765784e+00 -5.80019057e-01 1.88494265e-01
-7.45371342e-01 -8.50016713e-01 -6.16552234e-01 -1.26795447e+00
6.92227542e-01 8.84655342e-02 7.06580639e-01 3.48620385e-01
2.45534480e-01 2.59862673e-02 -9.01128054e-01 -3.11569452e-01
-3.42960447e-01 6.42682672e-01 -2.10676417e-01 -6.74349815e-02
5.67930758e-01 -3.68120819e-01 -4.27365243e-01 5.82998633e-01
-1.09751797e+00 -4.80740964e-02 8.64136875e-01 9.39012766e-01
3.86665016e-01 3.14007640e-01 9.00231600e-01 -1.03117585e+00
5.31374991e-01 -5.40097535e-01 -2.45606080e-01 7.56594539e-01
-7.14766264e-01 2.00489789e-01 8.51656735e-01 -2.53054142e-01
-1.76755714e+00 1.05268441e-01 -5.62449157e-01 3.03633988e-01
-4.15962487e-01 9.88752723e-01 -4.98530626e-01 4.11455035e-01
2.73386359e-01 2.70182669e-01 -6.26140654e-01 -8.64170313e-01
1.46694258e-01 9.37741399e-01 3.05637121e-01 -4.21930134e-01
5.47222316e-01 -1.07531518e-01 -3.18959564e-01 -7.12526441e-01
-8.78098190e-01 -8.56994867e-01 -8.83637369e-01 8.31954256e-02
1.20575082e+00 -7.15082824e-01 -4.92623150e-01 5.27439713e-01
-1.48706388e+00 1.69483021e-01 6.31670430e-02 6.67095423e-01
1.54765323e-01 7.73470998e-01 -7.91526973e-01 -8.90809178e-01
-5.83632648e-01 -8.39005709e-01 4.96904135e-01 9.12767887e-01
8.00113082e-02 -1.05515492e+00 2.41541922e-01 1.54756591e-01
2.79421300e-01 -2.53540665e-01 1.07046294e+00 -1.40207195e+00
-2.16079667e-01 -2.21758232e-01 -2.43687943e-01 1.94966972e-01
2.23533556e-01 -5.82651980e-02 -5.13883650e-01 1.94667250e-01
-1.73401818e-01 1.30006492e-01 6.67554080e-01 -4.53025103e-03
5.14506996e-01 -4.00150977e-02 -4.91531700e-01 1.29906997e-01
1.41130733e+00 4.84971106e-01 8.18664253e-01 4.69792843e-01
6.56886458e-01 7.36641169e-01 8.58755350e-01 2.55340606e-01
6.12756550e-01 2.97903270e-01 2.19110847e-01 -3.00288618e-01
9.72754359e-02 -4.62815613e-01 2.71207333e-01 1.25060999e+00
-1.38318002e-01 -5.69452941e-01 -9.29507375e-01 6.56301022e-01
-2.04131675e+00 -9.10034180e-01 -4.50676054e-01 1.94862127e+00
8.16207588e-01 1.84226081e-01 -1.95768718e-02 3.66341397e-02
1.23485970e+00 1.29736006e-01 -2.39919960e-01 -3.99043471e-01
-1.09046958e-01 1.19890578e-01 3.50503325e-01 2.32419699e-01
-1.15785611e+00 1.20275426e+00 5.79365730e+00 8.32152843e-01
-7.32552707e-01 5.61642088e-02 2.77744710e-01 8.26255560e-01
-3.26463163e-01 1.96259439e-01 -1.25298190e+00 3.65531474e-01
9.87295866e-01 -1.25102410e-02 -3.38889398e-02 6.72551632e-01
1.37540534e-01 -1.13826528e-01 -6.78512871e-01 5.64287424e-01
-2.07009360e-01 -1.04842865e+00 -5.73531054e-02 -1.81602895e-01
5.76796234e-01 8.93781334e-03 -6.23504400e-01 3.28032851e-01
3.43746275e-01 -8.36390316e-01 3.29320222e-01 5.64496040e-01
2.39479706e-01 -8.11056256e-01 1.31810296e+00 6.44930720e-01
-1.77895117e+00 1.35092586e-01 -3.67649913e-01 8.88029709e-02
4.59498227e-01 5.76356113e-01 -6.00466609e-01 1.17015254e+00
4.32111055e-01 4.44780827e-01 -2.88338572e-01 1.21057069e+00
-6.53938115e-01 8.02847862e-01 -2.46820226e-01 -3.36953342e-01
3.76931548e-01 -2.51339555e-01 3.81015331e-01 1.61967766e+00
3.48773122e-01 2.08438918e-01 2.47334093e-01 5.79567194e-01
-1.14427097e-01 7.18211591e-01 -1.23858191e-01 -3.53279591e-01
7.57918954e-01 1.47937584e+00 -9.42290783e-01 -3.66624355e-01
-7.14142978e-01 8.47662687e-01 4.27092493e-01 1.09436668e-01
-8.89644265e-01 -1.05645502e+00 3.44634920e-01 -3.15694183e-01
6.07644320e-01 -4.99074191e-01 -2.53514469e-01 -1.18985450e+00
-1.01416977e-02 -4.47923541e-01 6.96678996e-01 -4.07450438e-01
-1.24603963e+00 8.72547448e-01 -4.64934856e-01 -9.42947745e-01
2.04398513e-01 -3.51410359e-01 -9.10999060e-01 8.83396149e-01
-1.65028381e+00 -1.18783176e+00 -3.57805169e-03 4.75055963e-01
4.38204736e-01 2.82057345e-01 9.20477390e-01 5.09241700e-01
-1.06228769e+00 6.61737621e-01 3.07134390e-02 6.97491229e-01
6.55798078e-01 -1.04449689e+00 2.51365185e-01 1.16817284e+00
1.41528562e-01 8.81247580e-01 3.88959914e-01 -1.03055227e+00
-9.88631964e-01 -9.21034276e-01 1.55451918e+00 1.89249933e-01
3.92918676e-01 -9.70375389e-02 -9.92700458e-01 5.63505709e-01
3.97260249e-01 -2.82350808e-01 9.95680392e-01 2.49917313e-01
-2.37536296e-01 5.29150777e-02 -1.02149737e+00 4.68988180e-01
1.06572366e+00 -1.11796200e-01 -1.14528382e+00 -2.31576174e-01
6.74636662e-01 -2.44916533e-03 -1.01313436e+00 4.78528410e-01
2.45998263e-01 -7.46135414e-01 5.06748319e-01 -5.66089630e-01
1.28056677e-02 -6.54841721e-01 9.38492939e-02 -1.26300395e+00
-4.62668777e-01 -3.12575102e-01 2.21090928e-01 2.08476758e+00
7.76304901e-01 -6.48219466e-01 4.98832941e-01 6.74132526e-01
-3.44261616e-01 -4.66156632e-01 -6.67546809e-01 -9.06641662e-01
-1.98347464e-01 -2.46045500e-01 7.33416677e-01 9.12042260e-01
4.24374670e-01 7.62978375e-01 -3.88192505e-01 3.12764317e-01
1.44409269e-01 -2.28541456e-02 1.42804608e-01 -1.31573367e+00
6.72918111e-02 -2.12864414e-01 -1.50713399e-01 -1.20995724e+00
1.52673513e-01 -7.80425847e-01 4.06497508e-01 -1.88454711e+00
2.06633255e-01 -6.90086722e-01 -7.58985877e-01 7.21436143e-01
-4.89453524e-01 -1.93572879e-01 3.59662883e-02 2.61129856e-01
-7.60098994e-01 4.61449057e-01 8.23122740e-01 2.15722129e-01
-5.35895288e-01 -2.32938811e-01 -7.04504788e-01 6.55700624e-01
8.13490629e-01 -6.21612906e-01 -1.66470736e-01 -5.25002003e-01
2.08266169e-01 5.46511188e-02 -5.34899116e-01 -9.38619435e-01
6.85145140e-01 -2.13762298e-01 2.63332695e-01 -8.53779674e-01
-1.52682677e-01 -6.96621776e-01 7.72895366e-02 3.32275093e-01
-2.35971674e-01 1.53340772e-01 5.04265390e-02 6.92932904e-01
-3.39625746e-01 -7.12146580e-01 5.63702703e-01 -1.48987308e-01
-1.01849496e+00 2.97310859e-01 -6.19205654e-01 9.58256423e-02
9.80789125e-01 -2.40486801e-01 -3.90721202e-01 6.89352229e-02
-5.96338451e-01 3.67006928e-01 2.92252272e-01 4.44956571e-01
3.06170523e-01 -1.09112656e+00 -5.60224593e-01 -6.77429810e-02
8.31294507e-02 -1.10444397e-01 4.03292030e-01 7.09489107e-01
-3.56299162e-01 5.35449743e-01 -2.05070958e-01 -3.26711772e-04
-1.20144296e+00 6.17979825e-01 1.11278422e-01 -8.99167478e-01
-2.26331189e-01 5.90661764e-01 2.14494903e-05 -6.19470954e-01
2.40606949e-01 -3.92957032e-02 -9.12564218e-01 4.67007309e-02
4.90493476e-01 3.41763556e-01 5.98112196e-02 -8.17117453e-01
-5.33184290e-01 5.99483490e-01 -2.37144887e-01 -5.17734513e-03
1.34037364e+00 -3.83191586e-01 -2.55924761e-01 2.86272317e-01
7.27386177e-01 1.72890902e-01 -5.58426082e-01 -4.27142113e-01
6.18865967e-01 1.33236140e-01 -1.03841893e-01 -8.68694067e-01
-8.49948049e-01 8.04640055e-01 2.19525427e-01 2.72484690e-01
1.13130021e+00 -3.16334903e-01 1.19409704e+00 3.89460474e-01
3.03747445e-01 -1.41467881e+00 -3.15147847e-01 9.17688489e-01
7.97853023e-02 -8.82016361e-01 -3.45945388e-01 -6.81494474e-01
-5.85327804e-01 1.11520708e+00 8.30100417e-01 1.56316280e-01
7.70577371e-01 1.84908271e-01 9.94590372e-02 1.73772693e-01
-4.76733774e-01 -5.19895256e-01 2.34405100e-01 4.59285647e-01
8.46206069e-01 -1.45614922e-01 -1.17890918e+00 1.09883499e+00
4.02945578e-01 -2.98383664e-02 2.16444686e-01 1.02688420e+00
-6.40798807e-01 -1.79540443e+00 -2.00890303e-01 2.03749895e-01
-7.65069127e-01 -2.90082157e-01 -3.63030553e-01 6.14068747e-01
3.11834842e-01 1.42236388e+00 -1.79892898e-01 -5.46948969e-01
3.18277359e-01 3.87861431e-01 -5.17575145e-02 -8.18587899e-01
-9.33851779e-01 5.20213060e-02 5.13690770e-01 -1.50730923e-01
-6.48348212e-01 -3.94099206e-01 -1.74039292e+00 -5.86796850e-02
-7.52480209e-01 8.46667588e-01 6.62480474e-01 1.20037496e+00
5.21386445e-01 4.06025112e-01 6.45285070e-01 -1.94110498e-01
-2.47575521e-01 -1.14585912e+00 -6.40781343e-01 1.23540513e-01
-4.01833475e-01 -4.54152137e-01 -6.63272962e-02 -1.42405614e-01] | [9.716283798217773, 9.596170425415039] |
d9645a1f-0b5f-4455-88d9-b0fc8b6c7efc | accdoa-activity-coupled-cartesian-direction | 2010.15306 | null | https://arxiv.org/abs/2010.15306v2 | https://arxiv.org/pdf/2010.15306v2.pdf | ACCDOA: Activity-Coupled Cartesian Direction of Arrival Representation for Sound Event Localization and Detection | Neural-network (NN)-based methods show high performance in sound event localization and detection (SELD). Conventional NN-based methods use two branches for a sound event detection (SED) target and a direction-of-arrival (DOA) target. The two-branch representation with a single network has to decide how to balance the two objectives during optimization. Using two networks dedicated to each task increases system complexity and network size. To address these problems, we propose an activity-coupled Cartesian DOA (ACCDOA) representation, which assigns a sound event activity to the length of a corresponding Cartesian DOA vector. The ACCDOA representation enables us to solve a SELD task with a single target and has two advantages: avoiding the necessity of balancing the objectives and model size increase. In experimental evaluations with the DCASE 2020 Task 3 dataset, the ACCDOA representation outperformed the two-branch representation in SELD metrics with a smaller network size. The ACCDOA-based SELD system also performed better than state-of-the-art SELD systems in terms of localization and location-dependent detection. | ['Yuki Mitsufuji', 'Shusuke Takahashi', 'Naoya Takahashi', 'Yuichiro Koyama', 'Kazuki Shimada'] | 2020-10-29 | null | null | null | null | ['sound-event-localization-and-detection'] | ['audio'] | [-1.30960420e-01 -3.11351895e-01 2.63877779e-01 -1.43381998e-01
-7.95798898e-01 -2.89621025e-01 4.06404138e-01 3.48198861e-01
-6.04007363e-01 2.81654030e-01 5.26146106e-02 -3.17077219e-01
-5.78980148e-01 -6.36443973e-01 -2.98752040e-01 -7.47963071e-01
-3.68838668e-01 2.93345541e-01 7.32470930e-01 6.45116568e-02
1.20974258e-01 7.82616615e-01 -1.44726622e+00 1.78938746e-01
3.09336603e-01 1.47975779e+00 4.61914450e-01 6.67033017e-01
1.22214733e-02 6.72141671e-01 -9.91077662e-01 2.18052253e-01
2.02763945e-01 -5.86995423e-01 -3.21668983e-01 -6.68843925e-01
3.29783320e-01 2.02536762e-01 -2.99472034e-01 6.93260372e-01
1.10289133e+00 3.77461553e-01 6.40520930e-01 -1.32590055e+00
1.93387896e-01 5.14833450e-01 -3.74459147e-01 8.08091521e-01
3.73117253e-02 -3.08131725e-01 9.88517821e-01 -8.19811463e-01
2.17259824e-01 9.04342771e-01 9.12008762e-01 4.43197861e-02
-9.94112432e-01 -6.30257726e-01 6.36524558e-02 3.00310105e-01
-1.64249730e+00 -2.30840921e-01 8.41082990e-01 -4.06858087e-01
1.15593863e+00 4.01540607e-01 6.08726144e-01 7.76171088e-01
1.50961503e-01 4.56765503e-01 5.74894905e-01 -4.91293669e-01
6.43859744e-01 -2.76282370e-01 5.95759489e-02 4.58698988e-01
1.40689984e-01 2.20344275e-01 -8.15081000e-01 -2.12412149e-01
6.58063173e-01 -2.73773372e-01 -1.97907746e-01 -9.03800428e-02
-1.18256891e+00 8.54067385e-01 4.26719248e-01 6.59182668e-01
-7.07997799e-01 2.62541175e-01 5.70643365e-01 1.63845107e-01
3.68347883e-01 7.92673469e-01 -3.95634294e-01 -1.97785109e-01
-1.27811551e+00 1.90739408e-01 9.12260473e-01 4.53210324e-01
2.94544309e-01 5.82812011e-01 -4.61927772e-01 1.01956737e+00
1.67745352e-01 4.58719969e-01 5.45684934e-01 -6.71577215e-01
5.19638479e-01 1.63225278e-01 -2.50731278e-02 -1.33509064e+00
-1.17286825e+00 -1.17710829e+00 -7.48385668e-01 1.51348218e-01
6.24524832e-01 -4.04560804e-01 -7.52838314e-01 1.65930748e+00
2.60247022e-01 1.97457463e-01 2.20476110e-02 1.00050342e+00
5.65865517e-01 8.78019214e-01 -2.05191493e-01 -2.29396403e-01
1.40599823e+00 -7.89830565e-01 -6.53451562e-01 -4.65445161e-01
5.81516445e-01 -7.00010598e-01 4.70117390e-01 5.65506876e-01
-8.57942462e-01 -5.25187433e-01 -1.11125302e+00 6.40181363e-01
-3.20497572e-01 5.84862828e-01 4.07760650e-01 6.24847412e-01
-8.81643832e-01 2.53303230e-01 -7.73001313e-01 -2.30455220e-01
-2.99342815e-02 2.50741690e-01 -6.23652637e-02 5.16345322e-01
-1.35256207e+00 7.82026052e-01 3.60394448e-01 1.94032624e-01
-9.56250012e-01 -5.37953436e-01 -7.11960137e-01 4.03968036e-01
4.66526479e-01 -8.61999169e-02 1.31536758e+00 -3.89187694e-01
-1.38784945e+00 1.41023546e-01 9.17128101e-02 -8.90191317e-01
1.53984949e-01 -1.53799713e-01 -6.63933694e-01 2.40614712e-01
2.14749992e-01 4.04335409e-01 8.55677366e-01 -7.26401925e-01
-1.02554941e+00 -2.05782913e-02 -1.61988124e-01 1.13472275e-01
-2.57075429e-01 1.35282338e-01 -1.97887897e-01 -8.07616711e-01
5.48874080e-01 -7.22929597e-01 -1.45582616e-01 -9.10158530e-02
-4.28263992e-01 -4.14142042e-01 6.86565101e-01 -2.71990091e-01
1.68224740e+00 -2.23371291e+00 -1.36275679e-01 2.29837954e-01
1.22476757e-01 3.24679613e-01 -1.41243398e-01 4.29858208e-01
-2.42851242e-01 -2.61377275e-01 1.12851247e-01 -3.68467808e-01
-2.23368153e-01 -1.39346957e-01 -1.92495942e-01 3.74051988e-01
-3.30938175e-02 1.30463436e-01 -7.81877041e-01 -3.86630863e-01
4.68288064e-02 4.30598944e-01 -5.00415325e-01 1.69992402e-01
7.50987530e-02 -8.83809328e-02 -3.51014614e-01 4.67447788e-01
4.42881495e-01 7.06705973e-02 7.14600086e-02 -4.68047678e-01
-3.56427610e-01 4.83672827e-01 -1.61796057e+00 1.38726425e+00
-6.07215166e-01 1.01167202e+00 9.83615965e-02 -1.10913956e+00
1.23046851e+00 5.77086568e-01 5.40102899e-01 -8.45175982e-01
1.57829806e-01 3.54056597e-01 3.07812572e-01 -3.56192291e-01
2.24684000e-01 3.51588465e-02 -2.08380222e-01 3.93816501e-01
1.32627442e-01 6.22231364e-02 2.76924431e-01 -8.14410821e-02
1.33465219e+00 -3.96820515e-01 2.12313756e-01 -2.70181477e-01
4.38073635e-01 -3.08892578e-01 7.03901708e-01 1.22954071e+00
-2.32824624e-01 7.64627695e-01 6.50711060e-01 -5.46735704e-01
-2.16771752e-01 -1.06441021e+00 -9.88554060e-02 1.18920541e+00
-5.57552371e-03 -3.99274558e-01 -6.23712301e-01 -6.34407222e-01
-1.47358567e-01 8.78651202e-01 -3.50666076e-01 -5.20686097e-02
-8.47979426e-01 -7.33160794e-01 1.05300903e+00 6.38479054e-01
3.71888697e-01 -9.81434166e-01 -1.04145288e+00 5.41601002e-01
-3.73041958e-01 -1.04001236e+00 -5.42677283e-01 8.58061790e-01
-5.22791386e-01 -7.73834050e-01 -6.93824887e-01 -6.49468601e-01
2.87983388e-01 1.62377611e-01 8.75545800e-01 -4.34143573e-01
-2.45338365e-01 2.30389267e-01 -3.41243774e-01 -7.06236243e-01
-6.63386881e-02 1.15126848e-01 1.52257040e-01 2.41069645e-01
1.40560016e-01 -6.29569769e-01 -4.55875367e-01 5.96815825e-01
-6.51506007e-01 -4.09181386e-01 5.53156793e-01 5.83957970e-01
4.26337957e-01 1.09759539e-01 9.92624283e-01 -7.14275464e-02
8.01067412e-01 -5.96896410e-01 -7.86724329e-01 6.19269907e-03
-6.20690823e-01 -5.18832244e-02 5.36581516e-01 -6.35029852e-01
-6.08189940e-01 2.80416161e-01 -2.82545686e-01 -4.36261028e-01
-2.63930690e-02 6.57706976e-01 1.30884185e-01 1.08066812e-01
8.87387991e-01 2.06902549e-01 -2.11759403e-01 -5.46662450e-01
-1.87786385e-01 7.00612128e-01 5.29554069e-01 -3.47479790e-01
3.45414490e-01 2.14195132e-01 2.69237995e-01 -9.18354630e-01
-8.05118978e-01 -6.91241562e-01 -3.83200884e-01 -3.42237622e-01
8.55042160e-01 -7.67125487e-01 -5.47542989e-01 4.74326551e-01
-1.39219689e+00 1.61650032e-01 -3.80577385e-01 9.41239715e-01
-2.74551362e-01 -1.17865793e-01 -1.63909674e-01 -9.18152809e-01
-2.34438032e-01 -9.48096275e-01 8.35381508e-01 2.15132087e-01
-3.29424679e-01 -5.94281971e-01 2.43933931e-01 -3.27223033e-01
6.75255001e-01 1.17626414e-01 5.85201502e-01 -1.02865779e+00
-1.32679760e-01 -5.55490971e-01 -1.85464062e-02 3.11552823e-01
-7.74104819e-02 -4.64504421e-01 -8.39770973e-01 9.09227356e-02
3.05868745e-01 2.22531334e-01 6.89320922e-01 7.46916413e-01
8.57210517e-01 -3.78398709e-02 -3.04926008e-01 4.33475345e-01
1.05517209e+00 6.72668457e-01 2.82522231e-01 3.49852383e-01
3.92438263e-01 3.82994831e-01 6.09759033e-01 5.33610404e-01
1.70409590e-01 1.05081880e+00 6.50644481e-01 -7.58253410e-02
-3.18673015e-01 4.16957587e-02 2.24370360e-01 6.27977252e-01
1.43183172e-01 -6.55026555e-01 -9.96072173e-01 7.33212888e-01
-1.72214305e+00 -9.34092462e-01 -3.63088459e-01 2.13380599e+00
5.35454214e-01 2.61885017e-01 1.72268480e-01 5.58262646e-01
6.71738029e-01 3.54279935e-01 -1.55435279e-01 -3.50109518e-01
-6.64702579e-02 -6.90536872e-02 3.50289971e-01 1.99261919e-01
-1.15297294e+00 3.46871346e-01 5.98388433e+00 1.13138354e+00
-1.40288246e+00 4.35146481e-01 1.04115851e-01 -4.63338614e-01
5.04121840e-01 -3.48732293e-01 -1.06531334e+00 4.91128147e-01
1.17921293e+00 3.16858143e-01 1.87310874e-01 7.46227980e-01
3.33120584e-01 -3.74449253e-01 -1.08300114e+00 1.12986791e+00
8.37907419e-02 -1.29153156e+00 -4.00778264e-01 -1.75821573e-01
2.26006433e-01 4.04116929e-01 -3.78289312e-01 3.30845356e-01
-4.18645352e-01 -5.68187892e-01 1.12686217e+00 4.12355006e-01
4.49506938e-01 -6.48942947e-01 7.63420403e-01 4.16198134e-01
-1.58690429e+00 -2.86081076e-01 -1.72339991e-01 -3.42715271e-02
4.72049803e-01 8.24803531e-01 -6.20055497e-01 4.74275172e-01
9.51280951e-01 1.13366798e-01 -2.94879436e-01 1.40677667e+00
-9.40331221e-02 9.44569290e-01 -9.04765546e-01 -2.99934804e-01
4.24260557e-01 2.29613394e-01 1.13607025e+00 1.42972958e+00
5.46179295e-01 -2.81149060e-01 1.49224952e-01 6.16646886e-01
2.33592391e-01 -1.13268450e-01 -3.26125860e-01 2.56817043e-01
8.40539694e-01 1.07532263e+00 -9.73448515e-01 -1.08247340e-01
-1.75757244e-01 4.66987878e-01 -4.33645165e-03 2.63497055e-01
-8.64669442e-01 -1.02822578e+00 4.21069026e-01 2.01300859e-01
5.80344558e-01 -8.70312974e-02 -2.38594458e-01 -4.74551320e-01
1.51636060e-02 -6.07918143e-01 4.39940363e-01 -6.99672759e-01
-7.58508801e-01 1.03518856e+00 1.58772424e-01 -1.55298662e+00
-3.18646938e-01 -3.96393627e-01 -7.18734801e-01 7.93272734e-01
-1.17966771e+00 -6.20178521e-01 -1.50622979e-01 3.42971802e-01
5.30377388e-01 -3.05263817e-01 7.08014488e-01 7.19709992e-01
-6.75634742e-01 5.42936087e-01 -1.83387220e-01 3.09742056e-02
5.43753207e-01 -1.05288613e+00 6.68569729e-02 8.91819656e-01
4.02611822e-01 1.32855579e-01 6.07462704e-01 -3.93583089e-01
-9.38155234e-01 -1.00077593e+00 1.11207712e+00 -8.09584633e-02
3.94621998e-01 -2.89029598e-01 -6.24287367e-01 2.24775836e-01
-1.32718131e-01 3.02546114e-01 2.67497838e-01 9.54078361e-02
-1.04747944e-01 -5.58335781e-01 -9.40804780e-01 1.81526795e-01
6.74926341e-01 -4.14546520e-01 -4.19210851e-01 3.27498615e-01
4.25416708e-01 -3.33772063e-01 -6.18528247e-01 3.95869344e-01
4.46645379e-01 -9.52622592e-01 1.04543662e+00 -3.87296490e-02
-1.40639663e-01 -6.33562565e-01 -1.31873995e-01 -1.21346128e+00
-2.84360498e-01 -4.99754369e-01 -3.39849174e-01 9.69524622e-01
4.66450214e-01 -7.41379321e-01 3.45844299e-01 -2.87120134e-01
-3.75035584e-01 -9.19100881e-01 -1.43285155e+00 -1.27666295e+00
-5.24710298e-01 -8.62237453e-01 4.68872339e-01 4.74808216e-01
-2.67698795e-01 5.64427435e-01 -3.17133099e-01 6.35790467e-01
2.25367352e-01 -2.24645898e-01 2.56384730e-01 -1.42190421e+00
-3.00445229e-01 -5.75433254e-01 -5.18418670e-01 -1.06307983e+00
-4.86596197e-01 -6.72159612e-01 4.26635057e-01 -1.49038219e+00
-5.59275091e-01 -4.70197350e-01 -6.28699839e-01 5.48301518e-01
1.93277866e-01 8.28848183e-02 1.75236166e-01 1.67060137e-01
-5.27656138e-01 2.27740541e-01 5.22331655e-01 1.23362564e-01
-5.09586632e-01 4.30653751e-01 -1.72561675e-01 8.30343008e-01
7.45893598e-01 -1.12327123e+00 -3.32079679e-01 -4.01176214e-01
1.87929422e-01 3.18383455e-01 3.77238840e-01 -1.57515860e+00
7.96406806e-01 2.92735606e-01 1.61943078e-01 -1.00938523e+00
5.74571311e-01 -8.10697973e-01 1.12877890e-01 6.14597559e-01
-1.93128064e-01 1.39994189e-01 3.57671380e-01 5.27441800e-01
-4.27755743e-01 -5.15448868e-01 7.23620653e-01 3.60418648e-01
-4.69539672e-01 -2.17325523e-01 -8.11920047e-01 -4.23477627e-02
8.95285487e-01 -1.72167838e-01 -4.61783148e-02 -4.87086445e-01
-7.26110458e-01 -3.98930768e-03 -7.80409157e-01 5.28141975e-01
4.63306457e-01 -1.20062494e+00 -5.87415636e-01 2.42427528e-01
-1.05803646e-01 -1.54657960e-01 2.16313809e-01 1.11814153e+00
-4.24222589e-01 6.48525059e-01 4.51932922e-02 -8.11358154e-01
-1.01405144e+00 2.61113457e-02 8.18213582e-01 -5.07728279e-01
-4.14802849e-01 1.06888402e+00 9.77451429e-02 -4.04075414e-01
7.89072037e-01 -5.46000719e-01 -4.31854993e-01 4.15750474e-01
5.26410520e-01 8.95456254e-01 3.60305041e-01 -4.57055569e-01
-5.91859698e-01 4.16158378e-01 2.61732727e-01 -2.75337398e-01
1.19465685e+00 5.89027591e-02 3.01699996e-01 3.70355159e-01
9.49859917e-01 4.25681472e-02 -8.29067528e-01 -1.23248056e-01
-9.38396621e-03 1.40236523e-02 6.69454753e-01 -1.02993369e+00
-1.28795660e+00 8.98023009e-01 9.49447632e-01 6.86092138e-01
1.33107424e+00 -1.16101727e-01 5.79904318e-01 5.58756530e-01
2.11174726e-01 -1.03108633e+00 2.95707770e-02 7.56346881e-01
1.07914472e+00 -5.72112858e-01 -2.53787398e-01 -1.07032366e-01
-4.28128242e-01 1.22281671e+00 4.26680475e-01 4.84609269e-02
9.79046166e-01 4.48917359e-01 1.17929928e-01 -4.29536194e-01
-4.72400934e-01 -2.25975558e-01 4.34150189e-01 4.41059798e-01
-7.30892420e-02 -5.36298528e-02 -2.83830971e-01 8.19547474e-01
-5.28410338e-02 -3.09523374e-01 1.52949303e-01 9.38957036e-01
-4.68671769e-01 -6.98754191e-01 -6.33128405e-01 5.27944565e-01
-5.37537098e-01 -1.63388580e-01 1.56381987e-02 7.18083501e-01
1.89951479e-01 1.10600066e+00 3.23689461e-01 -4.47067738e-01
7.09499061e-01 -2.16417927e-02 -6.80521876e-03 -5.80349922e-01
-9.26570117e-01 4.30004030e-01 2.84008622e-01 -6.48750842e-01
-1.76638365e-02 -6.97821259e-01 -1.22432077e+00 3.29018861e-01
-8.49197388e-01 3.40773016e-01 9.89071608e-01 8.19953740e-01
6.26857638e-01 1.04554987e+00 5.70087910e-01 -8.86754990e-01
-5.51017582e-01 -1.11232150e+00 -6.13789320e-01 -4.71209109e-01
4.14209336e-01 -7.16876268e-01 -5.44000268e-01 -2.96932995e-01] | [15.173649787902832, 5.237888336181641] |
69799feb-118d-47ea-9e4f-19875da9ec16 | glamr-global-occlusion-aware-human-mesh | 2112.01524 | null | https://arxiv.org/abs/2112.01524v2 | https://arxiv.org/pdf/2112.01524v2.pdf | GLAMR: Global Occlusion-Aware Human Mesh Recovery with Dynamic Cameras | We present an approach for 3D global human mesh recovery from monocular videos recorded with dynamic cameras. Our approach is robust to severe and long-term occlusions and tracks human bodies even when they go outside the camera's field of view. To achieve this, we first propose a deep generative motion infiller, which autoregressively infills the body motions of occluded humans based on visible motions. Additionally, in contrast to prior work, our approach reconstructs human meshes in consistent global coordinates even with dynamic cameras. Since the joint reconstruction of human motions and camera poses is underconstrained, we propose a global trajectory predictor that generates global human trajectories based on local body movements. Using the predicted trajectories as anchors, we present a global optimization framework that refines the predicted trajectories and optimizes the camera poses to match the video evidence such as 2D keypoints. Experiments on challenging indoor and in-the-wild datasets with dynamic cameras demonstrate that the proposed approach outperforms prior methods significantly in terms of motion infilling and global mesh recovery. | ['Jan Kautz', 'Kris Kitani', 'Pavlo Molchanov', 'Umar Iqbal', 'Ye Yuan'] | 2021-12-02 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Yuan_GLAMR_Global_Occlusion-Aware_Human_Mesh_Recovery_With_Dynamic_Cameras_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Yuan_GLAMR_Global_Occlusion-Aware_Human_Mesh_Recovery_With_Dynamic_Cameras_CVPR_2022_paper.pdf | cvpr-2022-1 | ['human-mesh-recovery'] | ['computer-vision'] | [ 3.29575166e-02 -1.29942000e-01 1.01459716e-02 7.48455599e-02
-5.46414554e-01 -5.55774033e-01 3.66558641e-01 -4.43405926e-01
-2.41917193e-01 5.22784293e-01 3.28155190e-01 4.76049900e-01
1.56630784e-01 -6.00860655e-01 -1.11639953e+00 -3.98101121e-01
1.14409111e-01 7.11183131e-01 2.99616337e-01 1.44267157e-01
-1.59784406e-01 5.08804381e-01 -1.53521740e+00 -2.41133422e-01
4.13304538e-01 6.65478170e-01 4.16206494e-02 1.03371656e+00
5.77814877e-01 7.15516329e-01 -2.80963063e-01 -4.51941818e-01
6.01240695e-01 -2.64396399e-01 -4.95814145e-01 7.29862511e-01
9.21193600e-01 -7.29211986e-01 -9.30994213e-01 8.24973047e-01
3.87782365e-01 3.11445922e-01 2.81844884e-01 -1.10352623e+00
-2.36666068e-01 -1.85550675e-01 -6.59247696e-01 -1.93911940e-01
1.03730285e+00 3.38912606e-01 5.51168859e-01 -9.39669549e-01
1.33332336e+00 1.42582595e+00 1.10541117e+00 4.97706324e-01
-1.15630150e+00 -3.61600578e-01 3.02058488e-01 -2.01178327e-01
-1.52621031e+00 -5.19185662e-01 7.97614336e-01 -7.58671463e-01
5.82042634e-01 1.20091133e-01 1.15713191e+00 1.15661895e+00
2.63903648e-01 6.38571560e-01 3.61074179e-01 -2.24067539e-01
-9.52561051e-02 -5.64950466e-01 -4.16536480e-01 1.10403240e+00
2.99907357e-01 1.01369388e-01 -6.76785827e-01 -3.53229016e-01
1.41209006e+00 3.13357085e-01 -4.46989506e-01 -8.04892719e-01
-1.70323324e+00 4.81654882e-01 1.21065035e-01 -2.67112702e-01
-6.22480154e-01 5.48409581e-01 -7.50972852e-02 -1.93075016e-01
5.48387408e-01 2.75840349e-02 -2.82668412e-01 -4.27388623e-02
-1.07827771e+00 6.56517565e-01 6.96289062e-01 1.17932391e+00
6.17700756e-01 6.20239675e-02 1.55811924e-02 5.63740849e-01
4.11324918e-01 9.08885717e-01 -1.29456371e-01 -1.46948659e+00
5.70102334e-01 4.38511103e-01 5.72145045e-01 -1.41798496e+00
-1.77891061e-01 -9.46447924e-02 -7.59042799e-01 -2.11482178e-02
5.00261962e-01 -2.01690897e-01 -7.97446668e-01 1.58536994e+00
9.25599039e-01 6.03697538e-01 -3.53506267e-01 1.18927443e+00
5.89545548e-01 4.81854856e-01 -1.66663826e-01 -1.17101662e-01
1.01654530e+00 -1.09526169e+00 -8.11538160e-01 -3.47320706e-01
5.04463501e-02 -7.01915085e-01 5.63744307e-01 3.39494824e-01
-1.47606802e+00 -5.70219815e-01 -7.03516185e-01 -2.74690002e-01
3.13181221e-01 1.67428061e-01 9.10712406e-02 1.28693804e-01
-8.50910723e-01 6.57708228e-01 -1.32087421e+00 -4.64889050e-01
7.75135085e-02 3.43662441e-01 -5.57656825e-01 -1.72581330e-01
-6.12302959e-01 5.55476606e-01 -5.67543134e-02 3.14493716e-01
-1.07009733e+00 -5.25147915e-01 -1.13557124e+00 -3.86081636e-01
5.16909838e-01 -1.50319684e+00 1.04388618e+00 -5.10162532e-01
-1.55415678e+00 9.21126485e-01 -4.41720605e-01 -2.50445515e-01
1.08970308e+00 -7.24498928e-01 1.50973529e-01 4.43412036e-01
1.23758659e-01 7.29808986e-01 1.02377689e+00 -1.33023751e+00
-5.17096341e-01 -2.71757960e-01 -2.52584308e-01 3.39919001e-01
1.18678883e-01 -3.00996453e-01 -1.17587614e+00 -8.88961434e-01
3.94797921e-01 -1.43522656e+00 -4.77831781e-01 4.49884087e-01
-5.21612465e-01 3.91280830e-01 7.82802999e-01 -1.02250326e+00
1.06184340e+00 -1.74044561e+00 9.03168023e-01 3.19804966e-01
3.23092401e-01 -2.67146528e-01 1.87975094e-01 1.36020109e-01
3.74019891e-01 -4.24054950e-01 -2.07446143e-01 -1.00837123e+00
-3.88620794e-02 3.48001629e-01 -1.44677563e-02 9.22460794e-01
-3.11351866e-01 8.80366147e-01 -1.01925230e+00 -4.38423783e-01
5.56057692e-01 7.81792998e-01 -7.72763908e-01 4.33101267e-01
-1.73696548e-01 1.07668591e+00 -2.66056508e-01 8.60661924e-01
4.92711097e-01 -3.36327821e-01 1.43495724e-01 -2.42838278e-01
5.70765659e-02 -3.36232245e-01 -1.57190752e+00 2.21641469e+00
-8.77343304e-03 2.84114152e-01 4.30004716e-01 -4.00142699e-01
4.69393849e-01 4.28987801e-01 1.01627755e+00 9.35220793e-02
2.26255208e-02 2.43943874e-02 -8.01823735e-01 -5.16408563e-01
5.08131027e-01 5.31637594e-02 2.47863494e-02 -1.48651935e-02
-1.29020512e-01 -5.79088926e-02 -7.74618611e-02 4.53479700e-02
1.15426207e+00 8.01460028e-01 1.87293082e-01 2.28814900e-01
3.81149352e-01 -2.34239697e-02 7.84454465e-01 3.32645059e-01
7.26816654e-02 1.03546333e+00 -4.56667878e-03 -7.65207827e-01
-1.26651311e+00 -1.21727455e+00 3.86316270e-01 4.75651652e-01
4.75228310e-01 -5.72352409e-01 -7.29763389e-01 -3.97901535e-01
9.96534303e-02 -1.65988907e-01 -4.90208685e-01 3.06855947e-01
-1.04453743e+00 -2.97803551e-01 1.36697412e-01 4.93509561e-01
4.02681410e-01 -7.09040523e-01 -8.32204759e-01 3.55573416e-01
-5.59572160e-01 -1.56563878e+00 -8.71243656e-01 -7.42229283e-01
-9.22015369e-01 -1.25507236e+00 -1.08271468e+00 -6.93786740e-01
8.56393635e-01 1.77725971e-01 1.00352526e+00 1.96583703e-01
-3.78806740e-01 8.70455921e-01 1.96673516e-02 2.89960533e-01
-1.80899173e-01 -3.60393882e-01 3.05706024e-01 8.86978656e-02
-2.46004671e-01 -6.54965878e-01 -9.72630322e-01 4.26223606e-01
-5.09680629e-01 1.93017200e-01 6.44268608e-03 6.82952583e-01
9.84679461e-01 -1.12675712e-01 -3.21384698e-01 -5.31477332e-01
-1.68747246e-01 -4.29615617e-01 -6.56245232e-01 -1.22993328e-02
8.45858604e-02 -3.22708309e-01 1.71901613e-01 -6.31336451e-01
-9.30589020e-01 7.18651950e-01 1.15978479e-01 -1.25873184e+00
-1.13758944e-01 9.43148136e-02 -1.48344681e-01 -9.48727578e-02
2.57431686e-01 -1.26226127e-01 -8.27329829e-02 -5.80616832e-01
3.69800031e-01 -2.06296876e-01 1.05316162e+00 -6.41899586e-01
1.13216269e+00 9.62198019e-01 1.40831023e-01 -9.71745014e-01
-5.15798688e-01 -5.39881527e-01 -1.03348660e+00 -4.99412060e-01
1.35332525e+00 -1.27829170e+00 -9.10626531e-01 6.53359592e-01
-1.36405611e+00 -3.86979908e-01 -1.34857371e-01 6.17615104e-01
-7.57607639e-01 7.29492247e-01 -8.61038566e-01 -7.98576176e-01
-2.25570038e-01 -1.16554666e+00 1.72539616e+00 -2.00983271e-01
-5.50371945e-01 -1.03834867e+00 2.38548279e-01 2.99510211e-01
-4.17627037e-01 1.08080685e+00 1.54551223e-01 5.15184879e-01
-1.06674492e+00 -2.93064684e-01 4.22672808e-01 -3.17965336e-02
-4.18641344e-02 8.24472383e-02 -5.36629081e-01 -4.12331611e-01
-1.21071145e-01 1.85492069e-01 4.24269676e-01 7.93000162e-01
7.45840788e-01 -5.61177671e-01 -6.08435571e-01 1.16579616e+00
1.26864398e+00 -2.67971843e-01 5.39982617e-01 3.41255218e-01
1.37027705e+00 5.43080628e-01 5.59306026e-01 7.16593444e-01
5.78215420e-01 9.39483941e-01 3.78391147e-01 3.93914394e-02
-3.01143289e-01 -7.85456121e-01 4.31845844e-01 8.36113989e-01
-4.86937821e-01 -1.71286508e-01 -8.08138907e-01 6.17411673e-01
-2.07179856e+00 -9.18555081e-01 -2.60250986e-01 2.35894346e+00
3.18896741e-01 -2.28324115e-01 3.91961068e-01 -1.77873284e-01
8.39136243e-01 3.08307651e-02 -5.50321579e-01 4.87977624e-01
-1.59259755e-02 -1.18055135e-01 5.57504892e-01 7.96799839e-01
-1.12497699e+00 9.33203816e-01 6.24936962e+00 3.60384807e-02
-6.62132382e-01 1.80382729e-01 6.39787316e-02 -3.54907721e-01
-2.59202858e-03 2.15799920e-02 -7.88978696e-01 4.12365288e-01
3.63953888e-01 2.45051429e-01 4.20988739e-01 6.13042951e-01
2.17735291e-01 -4.57420424e-02 -1.15826774e+00 1.31453431e+00
2.49227732e-01 -1.28400755e+00 -1.21357240e-01 3.42196345e-01
1.14401662e+00 -2.40221456e-01 -1.46817103e-01 -3.49791288e-01
4.09400165e-01 -7.43304491e-01 1.19898748e+00 1.01754963e+00
6.59323871e-01 -6.46778464e-01 2.42725983e-01 5.12902498e-01
-1.43012512e+00 1.15479022e-01 -1.61599696e-01 -1.21154599e-01
8.34344566e-01 4.02486324e-01 -4.73704755e-01 3.97606879e-01
7.71426678e-01 1.06757772e+00 -2.30929688e-01 9.09058809e-01
-2.65102029e-01 1.39188558e-01 -5.85449874e-01 7.97163248e-01
-2.78020740e-01 -4.22911704e-01 9.37965155e-01 8.54713619e-01
6.64307535e-01 4.01546150e-01 6.80087626e-01 7.00599134e-01
-8.99597406e-02 -1.59083068e-01 -7.31355846e-01 3.01620752e-01
4.90060151e-01 8.98969769e-01 -6.57948971e-01 -3.82760674e-01
-2.57244915e-01 1.42625785e+00 1.23616934e-01 5.37375629e-01
-7.97717690e-01 3.73678595e-01 6.31835043e-01 5.28659999e-01
2.62439400e-01 -6.99701309e-01 5.90946823e-02 -1.57103479e+00
3.49046886e-01 -7.18579710e-01 2.44354427e-01 -8.27989638e-01
-1.02842236e+00 3.32596749e-01 1.11063622e-01 -1.32698333e+00
-3.70591551e-01 -1.90640420e-01 -2.33553126e-01 5.19055426e-01
-8.01706135e-01 -1.49571180e+00 -5.59568703e-01 8.68715882e-01
6.35005057e-01 2.17360780e-01 4.32873935e-01 2.86238104e-01
-3.51953357e-01 2.62329757e-01 -2.65110105e-01 1.04330882e-01
6.77794576e-01 -8.96747410e-01 6.14218891e-01 1.01332903e+00
1.29308254e-01 5.52526116e-01 8.80246103e-01 -1.24893045e+00
-1.73370993e+00 -1.11144400e+00 6.18535340e-01 -9.57463980e-01
3.09510380e-01 -3.15921843e-01 -5.92299163e-01 1.09265244e+00
-1.53820261e-01 2.35402465e-01 2.26837128e-01 -3.81905705e-01
8.65918398e-02 2.97719181e-01 -9.74566102e-01 5.76541543e-01
1.58954358e+00 -3.06544423e-01 -3.22345316e-01 5.03808260e-01
4.95754808e-01 -1.21981144e+00 -9.99777079e-01 3.74204338e-01
7.77426779e-01 -6.98118269e-01 1.46657693e+00 -2.55102754e-01
3.51358235e-01 -4.81320441e-01 -2.76815474e-01 -8.88692677e-01
-9.99965146e-02 -1.12586391e+00 -7.05124617e-01 8.18597198e-01
-4.17425185e-01 -6.22399002e-02 1.07143557e+00 7.12402761e-01
-5.98402433e-02 -4.00175452e-01 -8.95790458e-01 -6.60152495e-01
-3.57205331e-01 -2.70968139e-01 3.94423544e-01 7.63083339e-01
-5.17255127e-01 -1.38094440e-01 -1.06674266e+00 5.37208140e-01
1.12563789e+00 -1.00890890e-01 1.43726337e+00 -1.10990167e+00
-3.97234976e-01 1.82006761e-01 -5.18688917e-01 -1.49756718e+00
2.93402761e-01 -4.23450261e-01 3.04318279e-01 -1.43451321e+00
3.16289812e-02 -4.16101739e-02 6.45432770e-01 9.58082229e-02
-2.31962740e-01 4.56013620e-01 3.20973158e-01 4.73345429e-01
-6.36011004e-01 5.22559106e-01 1.36766315e+00 1.00759260e-01
-2.16343686e-01 -4.95917201e-02 2.33460397e-01 1.28554881e+00
4.55300137e-02 -4.19397414e-01 -2.04959840e-01 -6.25488222e-01
1.75184816e-01 4.86824244e-01 8.36468220e-01 -1.19174111e+00
4.75444436e-01 -1.66666210e-01 6.76214755e-01 -8.43026280e-01
7.61453748e-01 -9.03049111e-01 9.85634089e-01 4.20790792e-01
1.74985249e-02 5.21254301e-01 -1.60045475e-01 1.04885185e+00
1.77734822e-01 3.16445470e-01 5.63948333e-01 -3.42853755e-01
-4.83395755e-01 7.83047736e-01 -1.80676699e-01 2.07134355e-02
9.73082900e-01 -4.63457495e-01 1.81839660e-01 -5.59302390e-01
-1.27949417e+00 2.39589348e-01 1.17868733e+00 3.91257256e-01
8.64333510e-01 -1.55677509e+00 -5.77137709e-01 1.94840878e-01
-2.20871985e-01 4.20719385e-01 3.44423592e-01 7.67326295e-01
-1.00265217e+00 1.43836170e-01 8.20260718e-02 -1.03255546e+00
-1.37395120e+00 5.94758630e-01 3.30325514e-01 -5.46272770e-02
-1.18197715e+00 6.56012416e-01 2.60912389e-01 -4.05939221e-01
2.83603281e-01 -3.42020899e-01 3.28239918e-01 -3.81959975e-01
2.63487667e-01 7.93815792e-01 -2.67580122e-01 -1.24269557e+00
-2.00443178e-01 1.21088660e+00 5.41785538e-01 -3.82237196e-01
1.31044412e+00 -5.99425435e-01 3.88778187e-02 3.20601732e-01
1.15009558e+00 3.90118599e-01 -1.91702878e+00 -1.27628461e-01
-3.60752642e-01 -1.00911403e+00 -4.19373602e-01 -1.23861939e-01
-1.12227011e+00 5.66014290e-01 2.59340674e-01 -6.19361401e-01
7.38984466e-01 -1.10707544e-01 1.24073386e+00 1.14366636e-01
6.68757915e-01 -9.25552905e-01 2.22425044e-01 3.24062616e-01
8.13685775e-01 -8.59550953e-01 2.42037952e-01 -7.55618334e-01
-3.48350048e-01 9.22293723e-01 6.03042066e-01 -4.19751078e-01
5.13163149e-01 1.06963173e-01 -1.50592238e-01 -3.23935628e-01
-3.95060748e-01 8.76187384e-02 4.43607718e-01 6.63761735e-01
-8.47421994e-04 -6.15188777e-02 1.66092545e-01 -1.83822755e-02
-2.69545764e-01 4.30157483e-02 3.39394897e-01 9.88451958e-01
-1.18834987e-01 -8.56878281e-01 -9.41273689e-01 -1.15982056e-01
-3.38934600e-01 3.76712918e-01 -1.67229787e-01 7.05856442e-01
3.06018263e-01 7.80195177e-01 -4.25926447e-02 -2.48271182e-01
5.07052541e-01 -1.18957222e-01 7.38036156e-01 -6.19912684e-01
-2.11325645e-01 6.04696870e-01 1.66369863e-02 -9.54024315e-01
-6.50313199e-01 -9.50508356e-01 -1.01981139e+00 -4.19349492e-01
-2.70853173e-02 -3.04713339e-01 2.78622359e-01 7.73664773e-01
2.61845320e-01 3.43361586e-01 1.12584412e-01 -1.48200727e+00
-8.44762474e-02 -4.09499198e-01 -4.28784132e-01 7.70462871e-01
5.97926199e-01 -8.96249950e-01 -1.87099770e-01 7.39305556e-01] | [7.257015705108643, -0.9475335478782654] |
4c420d32-74d6-42d7-ad44-0e0298212ccb | an-end-to-end-pipeline-for-3d-slide-wise | 2305.11968 | null | https://arxiv.org/abs/2305.11968v1 | https://arxiv.org/pdf/2305.11968v1.pdf | An End-to-end Pipeline for 3D Slide-wise Multi-stain Renal Pathology Registration | Tissue examination and quantification in a 3D context on serial section whole slide images (WSIs) were laborintensive and time-consuming tasks. Our previous study proposed a novel registration-based method (Map3D) to automatically align WSIs to the same physical space, reducing the human efforts of screening serial sections from WSIs. However, the registration performance of our Map3D method was only evaluated on single-stain WSIs with large-scale kidney tissue samples. In this paper, we provide a Docker for an end-to-end 3D slide-wise registration pipeline on needle biopsy serial sections in a multi-stain paradigm. The contribution of this study is three-fold: (1) We release a containerized Docker for an end-to-end multi-stain WSI registration. (2) We prove that the Map3D pipeline is capable of sectional registration from multi-stain WSI. (3) We verify that the Map3D pipeline can also be applied to needle biopsy tissue samples. The source code and the Docker have been made publicly available at https://github.com/hrlblab/Map3D. | ['Yuankai Huo', 'Ruining Deng', 'Peize Li'] | 2023-05-19 | null | null | null | null | ['whole-slide-images'] | ['computer-vision'] | [ 1.57858625e-01 -1.71516284e-01 7.26235136e-02 -2.59810805e-01
-1.42685628e+00 -8.61335814e-01 5.85236214e-02 6.34916365e-01
-3.79381746e-01 3.64678502e-01 -1.77869603e-01 -4.32562321e-01
-3.09816182e-01 -5.49456477e-01 -4.79776114e-01 -7.67098844e-01
-5.32579012e-02 7.90958405e-01 5.99339187e-01 1.85026824e-01
1.10062793e-01 1.05818582e+00 -6.95019007e-01 4.79392782e-02
4.57308292e-01 4.38425362e-01 2.23838970e-01 1.26803076e+00
6.78536743e-02 -1.81324258e-02 -2.97907263e-01 -2.95265287e-01
5.58319867e-01 -4.41025943e-01 -1.09843242e+00 -1.92411736e-01
6.64732814e-01 -2.22460374e-01 2.17743628e-02 9.24318671e-01
8.36541235e-01 -4.04327959e-01 5.34059763e-01 -1.05007684e+00
-5.37240170e-02 4.37157182e-03 -6.68394089e-01 3.49530488e-01
1.08392507e-01 3.46516579e-01 5.24346113e-01 -6.30339921e-01
1.28200388e+00 6.11261606e-01 9.81995285e-01 5.75416803e-01
-1.29436445e+00 -5.90904117e-01 -6.30432427e-01 -3.33994120e-01
-1.54436135e+00 -1.34298295e-01 1.65393665e-01 -7.05125272e-01
9.00670171e-01 6.07032657e-01 8.50144565e-01 3.61779094e-01
5.97073436e-01 4.30506378e-01 1.33077216e+00 -2.30928794e-01
1.46114290e-01 -5.04713476e-01 2.37988785e-01 8.84625375e-01
3.96781474e-01 -3.22335511e-01 -2.63854861e-01 -4.31443214e-01
8.46766531e-01 1.78440467e-01 -8.66422281e-02 -2.42029354e-01
-1.60758483e+00 4.54051793e-01 2.88474262e-01 5.31392932e-01
-1.88636556e-01 1.24500640e-01 3.88072550e-01 5.46935983e-02
5.79054117e-01 2.60456562e-01 -5.20889945e-02 -1.94696367e-01
-9.69341397e-01 1.01718776e-01 4.88434792e-01 7.01136172e-01
5.09604812e-01 -9.67919350e-01 -1.13786429e-01 5.05065382e-01
1.97469622e-01 5.44127762e-01 3.75768244e-01 -4.63916600e-01
-8.58959630e-02 6.32916629e-01 -1.50222257e-01 -6.55665994e-01
-9.38229978e-01 -2.02220157e-01 -8.01913142e-01 3.57020766e-01
6.47089243e-01 1.14442654e-01 -8.09186816e-01 1.25149655e+00
8.65207434e-01 2.05853209e-01 -3.21251661e-01 1.11866117e+00
9.27494764e-01 -1.56549420e-02 -7.92277157e-02 8.30451772e-02
1.74717391e+00 -6.45380139e-01 -4.69617456e-01 3.95995885e-01
9.20247376e-01 -1.08605969e+00 8.34598243e-01 2.61633494e-03
-1.07137620e+00 9.20729786e-02 -9.86634433e-01 -1.50268108e-01
-2.34795183e-01 8.73902664e-02 5.01071632e-01 4.17908520e-01
-1.17718399e+00 4.05131042e-01 -1.62515914e+00 -8.63944113e-01
7.28885889e-01 5.56966543e-01 -9.07838106e-01 -9.53194723e-02
-3.80894929e-01 9.46662903e-01 -2.94562012e-01 2.45241955e-01
-4.96458620e-01 -1.03953266e+00 -5.42699039e-01 -6.12061083e-01
-1.46482989e-01 -7.40039706e-01 1.14797318e+00 9.57807824e-02
-1.33534062e+00 1.52718651e+00 -3.00767958e-01 1.16736859e-01
7.60330081e-01 4.66343194e-01 5.96553832e-02 4.88733292e-01
3.37936968e-01 2.45351121e-01 -2.37791255e-01 -8.15917850e-01
-1.64136037e-01 -6.21915758e-01 -5.32303214e-01 -1.38683885e-01
1.47662580e-01 1.65797099e-01 -4.84715194e-01 -2.65165180e-01
2.71043450e-01 -1.19652712e+00 -3.74622792e-01 5.84342778e-01
-4.89784181e-01 8.18921551e-02 4.55760121e-01 -6.59774959e-01
6.92851484e-01 -2.06444645e+00 -1.13193013e-01 4.51606423e-01
4.70364511e-01 2.83657968e-01 -1.62395909e-01 3.74874443e-01
-1.17999457e-01 2.89538115e-01 -1.01491131e-01 -3.41646552e-01
-6.79626390e-02 -3.13927323e-01 4.12291139e-01 1.15727365e+00
-2.35950463e-02 1.16723943e+00 -8.93251061e-01 -8.84647548e-01
-1.12309940e-02 4.78242487e-01 -2.43899986e-01 2.58960187e-01
3.59382778e-01 5.40563822e-01 -3.41086686e-01 9.36362267e-01
9.53244388e-01 -6.49419785e-01 3.20926994e-01 -4.42229271e-01
-1.64425448e-01 -1.27478853e-01 -9.69773293e-01 2.01697230e+00
-4.87049716e-03 3.89260501e-01 3.15704852e-01 -1.97034419e-01
6.30629301e-01 1.78019047e-01 8.07716548e-01 -2.49100149e-01
4.71654274e-02 3.70944470e-01 -7.28591532e-02 -6.63023591e-01
9.60279852e-02 -5.43763876e-01 7.40468800e-02 7.39797175e-01
-2.21387055e-02 -5.28638065e-01 3.12350392e-01 1.67688221e-01
1.61867118e+00 -8.30304176e-02 4.61113751e-01 -4.51182991e-01
2.12065071e-01 4.41612810e-01 3.90063763e-01 3.44906121e-01
-5.15957057e-01 9.13644135e-01 4.94734168e-01 -3.68667483e-01
-1.16200030e+00 -1.28051758e+00 -5.76229632e-01 4.01945204e-01
1.85724288e-01 -2.57123232e-01 -5.56126297e-01 -6.74117208e-01
2.00968489e-01 -2.06724927e-01 -7.56442010e-01 5.37865698e-01
-4.55756128e-01 -1.05439293e+00 8.91529977e-01 2.36398786e-01
-6.46580309e-02 -2.70073116e-01 -6.18929803e-01 6.31650537e-02
-4.64169979e-02 -8.04118335e-01 -7.31405497e-01 2.40390841e-02
-8.39864194e-01 -1.49951601e+00 -9.06297326e-01 -8.04356754e-01
1.18980610e+00 2.25124806e-01 6.42085910e-01 5.98017991e-01
-1.01769364e+00 2.42934659e-01 -1.75551489e-01 -3.95061612e-01
-5.23161292e-01 1.34209692e-01 -1.05288900e-01 -5.22764146e-01
3.49255145e-01 -4.92773443e-01 -9.28209662e-01 4.37220752e-01
-1.00496054e+00 2.62742341e-01 5.32502055e-01 7.26317942e-01
1.22149837e+00 -5.44015527e-01 2.65091360e-01 -1.02677870e+00
3.75662804e-01 -1.26053855e-01 -8.62325788e-01 4.50854421e-01
-2.85378277e-01 -2.26496443e-01 1.64113194e-01 -1.69262346e-02
-5.42836130e-01 1.48925915e-01 -4.67542768e-01 1.96999423e-02
-1.44068941e-01 4.26660657e-01 5.12478232e-01 -4.44222957e-01
6.10762298e-01 -8.96411389e-02 7.88574100e-01 -3.05308729e-01
-3.30896497e-01 6.10515654e-01 3.56499642e-01 -2.91144937e-01
8.34769905e-01 8.89092505e-01 5.55834591e-01 -3.80261064e-01
-5.11309266e-01 -1.02796149e+00 -8.38355243e-01 -2.14264110e-01
8.56468916e-01 -5.04466772e-01 -9.25540328e-01 6.12879574e-01
-7.72309840e-01 -8.16086829e-01 4.58273180e-02 6.80265665e-01
-3.85806590e-01 5.53004563e-01 -1.07214546e+00 -1.41632810e-01
-8.54978204e-01 -1.35768855e+00 1.35129392e+00 1.99920624e-01
-4.26489145e-01 -1.04114079e+00 7.25007117e-01 3.74372214e-01
5.99605918e-01 7.40352869e-01 6.58497274e-01 -6.46467090e-01
-4.88442898e-01 -5.72475791e-01 -1.98700428e-01 -3.66712719e-01
2.83135414e-01 4.88117754e-01 -6.52911127e-01 -3.28952342e-01
-2.83192962e-01 -1.20915771e-01 4.40973729e-01 3.98254454e-01
8.01843405e-01 6.90860078e-02 -5.44448733e-01 8.46826673e-01
1.72154009e+00 -1.52903423e-01 5.65595865e-01 5.16366541e-01
4.17671561e-01 3.58147532e-01 7.53600240e-01 2.34819919e-01
2.21833274e-01 6.01262689e-01 3.28556865e-01 -6.93345666e-01
-3.53537202e-01 1.14272691e-01 -3.78051370e-01 8.57607782e-01
-2.41514251e-01 7.85299912e-02 -1.32982850e+00 6.40008271e-01
-1.48382628e+00 -7.40605950e-01 -4.67225641e-01 1.83069587e+00
1.03322804e+00 -3.79572511e-01 1.38849348e-01 -4.07910973e-01
6.38505161e-01 -3.49893570e-01 -3.91940147e-01 1.37806863e-01
2.93642678e-03 4.29695547e-01 6.67464316e-01 5.93948901e-01
-9.84161258e-01 3.19439441e-01 6.21981668e+00 7.19045341e-01
-1.29332542e+00 2.41951257e-01 5.81891894e-01 -3.38947117e-01
-1.26897380e-01 -1.27321810e-01 -6.62291706e-01 8.92103761e-02
5.50861418e-01 -3.42287451e-01 -2.04600822e-02 4.49699461e-01
9.60100666e-02 -3.15939963e-01 -1.10545707e+00 7.37295270e-01
-2.10435107e-01 -1.81749713e+00 -4.08166856e-01 5.91006637e-01
3.59617114e-01 5.06215632e-01 -2.67902911e-01 -3.98076862e-01
2.71409422e-01 -9.60421920e-01 9.35311168e-02 5.48178494e-01
1.15957797e+00 -2.68164366e-01 1.21999133e+00 -1.28274843e-01
-1.02679980e+00 1.03382874e+00 -1.03732206e-01 5.66396773e-01
1.45130768e-01 9.16967511e-01 -1.70175874e+00 7.38440633e-01
4.40354675e-01 3.89387339e-01 -9.10055935e-01 1.38680601e+00
1.80513248e-01 4.08109099e-01 -2.00944260e-01 -7.87663758e-02
-9.87434760e-02 -2.62790143e-01 3.65651131e-01 1.64383543e+00
4.72690910e-01 1.11970082e-01 -1.38676949e-02 4.91072387e-01
2.47434586e-01 1.85500011e-01 -1.85205847e-01 2.46686675e-02
2.40643755e-01 1.86439562e+00 -1.36638558e+00 3.82884638e-03
-9.68850330e-02 7.44062603e-01 2.16382310e-01 -3.05712316e-02
-6.77923024e-01 -3.84249002e-01 6.65344119e-01 2.99720347e-01
-1.56304042e-03 -9.45506990e-02 -2.42991135e-01 -8.62120926e-01
-8.64952430e-02 -4.36110169e-01 5.56048036e-01 -6.28015161e-01
-1.62159479e+00 4.53191280e-01 -2.09665239e-01 -1.45458019e+00
1.51015088e-01 -8.05779576e-01 -4.46898669e-01 8.64023447e-01
-1.57184911e+00 -1.23055780e+00 -6.11434102e-01 2.73158252e-01
-3.18560541e-01 3.13547760e-01 1.29622650e+00 2.18904749e-01
-3.90333474e-01 5.42369068e-01 2.84909368e-01 4.40426916e-01
9.40684378e-01 -1.26539993e+00 9.45447907e-02 6.12502694e-01
-5.10670364e-01 1.07509160e+00 4.68514055e-01 -6.19213164e-01
-1.77004826e+00 -1.05760252e+00 7.57597089e-01 -5.36645234e-01
8.49264383e-01 -3.62844974e-01 -7.10343063e-01 8.67231190e-01
8.89361426e-02 6.09705567e-01 1.54672945e+00 -7.35459030e-02
-7.93431550e-02 3.73870023e-02 -1.71275020e+00 5.06353378e-01
7.15187311e-01 -3.29043448e-01 -9.84787866e-02 7.03102410e-01
9.17376801e-02 -1.11000812e+00 -1.60889125e+00 8.66502076e-02
7.92813241e-01 -6.01344407e-01 7.42044985e-01 -6.40132353e-02
1.94701403e-01 -7.53011823e-01 2.09188759e-01 -9.48517621e-01
-2.18423858e-01 -4.31210726e-01 5.65991700e-01 8.40659559e-01
3.37895393e-01 -9.64181781e-01 7.60231435e-01 5.46048284e-01
-2.61445194e-01 -8.76765430e-01 -1.45212400e+00 -7.00697541e-01
1.11921422e-01 7.55321011e-02 4.90997195e-01 8.46845627e-01
4.87097651e-01 -3.10659707e-01 4.85112637e-01 1.08576968e-01
6.87687457e-01 3.62444252e-01 9.64038610e-01 -1.09170187e+00
-3.72236788e-01 -4.35238630e-01 -7.78969169e-01 -3.11865032e-01
-3.34438592e-01 -1.43955350e+00 6.00427017e-02 -1.82267606e+00
8.81069005e-01 -5.37115097e-01 -1.16735317e-01 6.62023783e-01
-1.68905482e-01 7.65703142e-01 3.68844904e-02 4.97234553e-01
-7.10824072e-01 -2.28835762e-01 1.63730061e+00 -1.21386021e-01
3.11334759e-01 -2.95602500e-01 -4.81173903e-01 1.43144101e-01
7.83052325e-01 -7.77367234e-01 2.31018260e-01 5.66345640e-02
1.82114139e-01 1.02481708e-01 6.15786314e-01 -8.74260962e-01
4.23149973e-01 -1.69825926e-01 3.94024581e-01 -8.61972690e-01
2.03894619e-02 -6.98576033e-01 6.65921926e-01 8.74397278e-01
-1.52882598e-02 -6.75453469e-02 3.67899477e-01 1.21210545e-01
2.44245343e-02 -2.14385077e-01 7.78528214e-01 -6.64527863e-02
8.39435905e-02 4.48908478e-01 -4.05848831e-01 -2.30116025e-01
1.24238122e+00 -3.94247860e-01 -7.53978968e-01 2.11527571e-01
-5.61126053e-01 2.27491319e-01 1.08203614e+00 -4.68929619e-01
3.62486243e-01 -1.09506929e+00 -9.91650999e-01 -5.15177986e-03
3.18964839e-01 4.92898941e-01 6.22433245e-01 1.58777678e+00
-1.39745510e+00 3.12184542e-01 -2.93244153e-01 -1.05575049e+00
-1.80471766e+00 2.16046363e-01 5.31473041e-01 -7.98396051e-01
-6.25198483e-01 9.47327971e-01 -2.75572143e-06 -8.59705091e-01
-4.58428711e-01 -3.88647974e-01 4.97157633e-01 -3.28737915e-01
5.81069231e-01 4.23209518e-02 5.82211733e-01 -4.64015573e-01
-7.93743253e-01 7.82879233e-01 -2.51358002e-01 7.62704685e-02
1.69909906e+00 1.44531786e-01 -4.52842116e-01 3.57024699e-01
1.35487676e+00 3.57516438e-01 -9.09453034e-01 1.76156983e-02
-2.66662866e-01 -5.00574291e-01 -2.33020559e-01 -7.69371688e-01
-9.06286359e-01 3.21402371e-01 6.42174661e-01 -6.89547583e-02
9.00186002e-01 2.65512437e-01 7.41452813e-01 -4.91080694e-02
4.23092246e-01 -7.57673502e-01 -3.06766868e-01 1.83414862e-01
6.27123773e-01 -1.06650817e+00 5.28377652e-01 -6.75952494e-01
-1.61808491e-01 1.29009736e+00 3.10344070e-01 4.81093004e-02
5.67025959e-01 9.10816193e-01 4.74916428e-01 -7.35767663e-01
-4.92460608e-01 1.76699892e-01 7.13468045e-02 5.53142548e-01
8.89158607e-01 5.79236075e-02 -4.88010466e-01 2.40466312e-01
-6.11090586e-02 2.30291218e-01 5.32507718e-01 1.34166884e+00
9.14743915e-02 -1.01772058e+00 -4.07077640e-01 3.80883038e-01
-5.65714538e-01 1.90413564e-01 -3.37138951e-01 9.17029202e-01
-3.14350963e-01 4.83245730e-01 -1.47369104e-02 -4.79479246e-02
3.17039311e-01 -1.57427922e-01 6.75955653e-01 -5.81568897e-01
-8.37306678e-01 1.70980975e-01 -2.60019928e-01 -5.34093440e-01
-5.13932526e-01 -7.99760461e-01 -1.64553130e+00 -4.29496646e-01
-2.72348613e-01 -5.25360778e-02 8.91905963e-01 5.63962340e-01
4.74798322e-01 5.26149392e-01 3.57238710e-01 -7.76039064e-01
-2.96192672e-02 -8.49100053e-01 -6.94901407e-01 3.85717660e-01
2.76703417e-01 -2.50156730e-01 -2.90785670e-01 2.02706814e-01] | [14.952105522155762, -3.0435307025909424] |
c0aa1dae-5fed-4c1a-83ae-61a1c465e364 | a-human-eye-based-text-color-scheme | 2010.07510 | null | https://arxiv.org/abs/2010.07510v2 | https://arxiv.org/pdf/2010.07510v2.pdf | A Human Eye-based Text Color Scheme Generation Method for Image Synthesis | Synthetic data used for scene text detection and recognition tasks have proven effective. However, there are still two problems: First, the color schemes used for text coloring in the existing methods are relatively fixed color key-value pairs learned from real datasets. The dirty data in real datasets may cause the problem that the colors of text and background are too similar to be distinguished from each other. Second, the generated texts are uniformly limited to the same depth of a picture, while there are special cases in the real world that text may appear across depths. To address these problems, in this paper we design a novel method to generate color schemes, which are consistent with the characteristics of human eyes to observe things. The advantages of our method are as follows: (1) overcomes the color confusion problem between text and background caused by dirty data; (2) the texts generated are allowed to appear in most locations of any image, even across depths; (3) avoids analyzing the depth of background, such that the performance of our method exceeds the state-of-the-art methods; (4) the speed of generating images is fast, nearly one picture generated per three milliseconds. The effectiveness of our method is verified on several public datasets. | ['Xiang Yu Luo', 'Guan Jie Huang', 'Shao Wei Wang'] | 2020-10-15 | null | null | null | null | ['scene-text-detection'] | ['computer-vision'] | [ 1.49399191e-01 -6.58672273e-01 3.91174316e-01 -9.29401815e-02
-1.94546074e-01 -7.17633128e-01 4.37252313e-01 -5.97214326e-03
-3.31858426e-01 6.10606492e-01 -1.19154036e-01 -1.03485100e-01
4.40478325e-01 -8.45769703e-01 -5.19400120e-01 -8.10725331e-01
3.76840293e-01 2.01126158e-01 9.28562582e-01 -2.11646203e-02
5.51537156e-01 4.30667311e-01 -1.70011282e+00 3.71805966e-01
9.33344841e-01 8.45910072e-01 3.75026047e-01 6.91425741e-01
-6.31169796e-01 8.20802391e-01 -1.00338972e+00 -1.69550210e-01
4.50255722e-01 -6.12382531e-01 -3.65762472e-01 4.21435297e-01
4.17383701e-01 -6.31413698e-01 -4.28372175e-01 1.22744703e+00
3.22980583e-01 -1.07096002e-01 5.66413462e-01 -1.48573029e+00
-7.59178817e-01 1.77451968e-01 -1.29926217e+00 -3.93661577e-03
2.89053321e-01 1.02878742e-01 4.24053073e-01 -8.01217437e-01
3.65443945e-01 1.26390123e+00 1.38052315e-01 6.89760268e-01
-8.29036236e-01 -8.44778776e-01 2.83465296e-01 1.32480443e-01
-1.42147124e+00 -2.69350022e-01 6.79236412e-01 -3.94802868e-01
1.83711976e-01 3.49178851e-01 5.89380085e-01 7.90235877e-01
1.56069681e-01 7.90612638e-01 1.25377727e+00 -6.42178357e-01
1.90594748e-01 3.39860737e-01 -5.25523862e-03 8.00410628e-01
5.31873167e-01 -2.40914166e-01 -5.70335805e-01 -7.63590708e-02
7.91300058e-01 2.30133578e-01 -3.57441485e-01 -2.38071501e-01
-1.43075383e+00 4.46415067e-01 7.22489357e-02 2.28899300e-01
1.05943777e-01 1.57279596e-02 1.88296258e-01 -8.01216215e-02
2.45911807e-01 6.39539883e-02 -1.02563657e-01 -1.74732413e-02
-8.28715265e-01 2.99846157e-02 6.71920955e-01 9.89344478e-01
7.89903104e-01 -2.46548951e-02 6.73017930e-03 7.95414567e-01
1.97169349e-01 8.65105152e-01 5.07064581e-01 -5.06022155e-01
7.13605940e-01 8.33845198e-01 3.29942733e-01 -1.22470212e+00
-6.47285655e-02 5.28710306e-01 -8.59727383e-01 4.74378556e-01
8.62686038e-01 -2.58734047e-01 -9.15491283e-01 1.32431316e+00
2.97035307e-01 -8.08032304e-02 -1.50537327e-01 8.75833273e-01
7.05534160e-01 9.62049842e-01 -3.46946865e-01 -1.60590172e-01
1.39429891e+00 -7.90079355e-01 -9.47066963e-01 -2.74667531e-01
1.58549652e-01 -1.13996494e+00 1.38623238e+00 5.64366817e-01
-9.27360594e-01 -5.28637707e-01 -1.21098006e+00 1.59799114e-01
-5.86355805e-01 2.22029403e-01 3.91325742e-01 9.20323551e-01
-8.31262708e-01 -4.19137580e-03 -4.74127501e-01 -3.68833482e-01
6.93194047e-02 6.43074960e-02 -7.46288002e-02 -2.38696337e-01
-9.68341708e-01 5.62624991e-01 4.30622876e-01 1.70185670e-01
-6.16566062e-01 -1.30933121e-01 -4.02061462e-01 -1.19476259e-01
3.29096764e-01 -8.76723751e-02 9.07019377e-01 -1.61251473e+00
-1.38957512e+00 7.30148077e-01 -3.05114895e-01 2.91447163e-01
8.88469994e-01 -8.96688998e-02 -5.99398792e-01 2.48592332e-01
-4.13432531e-02 5.44840753e-01 7.54958928e-01 -1.63516092e+00
-8.44965458e-01 -2.44095027e-01 -3.87551077e-02 2.08033547e-01
-4.82647151e-01 1.17950760e-01 -1.05333531e+00 -6.23031080e-01
1.68759316e-01 -8.04510891e-01 -2.75381543e-02 4.15867180e-01
-5.82105219e-01 5.62493354e-02 1.29795778e+00 -4.17122036e-01
1.15960467e+00 -2.34945440e+00 -4.61016178e-01 8.20033848e-02
1.72297582e-01 3.18202913e-01 -1.04874045e-01 5.21718800e-01
1.86248362e-01 1.49256423e-01 6.73548356e-02 -1.23132188e-02
-2.65388757e-01 7.51955733e-02 -4.00746167e-01 5.28353512e-01
-1.06176712e-01 3.41491461e-01 -8.47887993e-01 -8.48510206e-01
3.51739317e-01 2.75261760e-01 2.46627238e-02 1.72386408e-01
-2.17947960e-01 1.23546578e-01 -4.23762083e-01 5.66324055e-01
1.13441384e+00 -4.03382927e-02 7.83493742e-03 -1.56918645e-01
-3.19084167e-01 -8.23821425e-02 -1.66072130e+00 9.17907238e-01
3.11674289e-02 8.71380925e-01 -2.88748622e-01 -3.74548614e-01
8.66266966e-01 1.43231139e-01 2.67679006e-01 -7.94235408e-01
-4.91091702e-03 -7.50675425e-02 -1.92264602e-01 -5.48266053e-01
6.50047898e-01 2.15453029e-01 1.67000651e-01 7.92096853e-01
-8.32994461e-01 -1.30359814e-01 5.14756978e-01 3.18204284e-01
6.75020158e-01 -8.39445144e-02 -1.48985967e-01 -1.32998556e-01
3.46064746e-01 3.63124758e-02 7.02977538e-01 7.37560987e-01
-1.08094729e-01 8.54563177e-01 7.19529390e-01 -4.37857270e-01
-1.04534459e+00 -9.46080565e-01 1.03935175e-01 8.54021430e-01
8.54766011e-01 -2.00280815e-01 -8.61233950e-01 -5.34774840e-01
-2.35316992e-01 4.11160171e-01 -5.47594428e-01 2.55807847e-01
-4.51021582e-01 -7.02228367e-01 5.38492322e-01 4.47455108e-01
8.77492547e-01 -9.61877048e-01 -8.91399562e-01 -1.11366116e-01
-2.90965736e-01 -1.20848298e+00 -7.77246892e-01 -3.76866162e-01
-7.20452487e-01 -1.36065209e+00 -7.49395907e-01 -7.79120862e-01
1.18149185e+00 1.03377771e+00 8.29185545e-01 3.30081463e-01
-4.56748247e-01 2.66443044e-01 -5.12557983e-01 -4.52026814e-01
-3.83594364e-01 -5.56465030e-01 -1.51219264e-01 2.63839602e-01
3.12314510e-01 2.17913151e-01 -8.39400828e-01 5.82093179e-01
-1.20127749e+00 5.03034294e-01 3.60154778e-01 5.96075594e-01
3.08815360e-01 6.13289714e-01 4.73509021e-02 -9.03339386e-01
5.13484001e-01 1.02690347e-02 -8.69903505e-01 4.71767455e-01
-4.51754779e-01 -1.84173346e-01 7.30045438e-01 -6.87495053e-01
-1.17178440e+00 6.54835552e-02 6.32343709e-01 -1.20421946e-01
-4.63495515e-02 -1.20896414e-01 -2.18254030e-01 7.06462041e-02
4.50944155e-01 3.55569720e-01 -2.06943713e-02 -1.83551490e-01
1.38214245e-01 9.06157851e-01 2.53978342e-01 -5.09280503e-01
9.98920500e-01 7.22086608e-01 -1.05679937e-01 -1.02616274e+00
-3.51374924e-01 -2.59501010e-01 -4.88872379e-01 -3.03633928e-01
7.37012327e-01 -7.02530444e-01 -6.49828553e-01 9.66925681e-01
-1.19153547e+00 -3.72789979e-01 3.32760245e-01 2.19243497e-01
8.20413306e-02 9.14512873e-01 -6.58564210e-01 -1.10347712e+00
-2.04998687e-01 -1.08954132e+00 8.92079175e-01 6.78900778e-01
2.47430593e-01 -7.91496933e-01 -3.47422093e-01 2.36178432e-02
2.83197314e-01 2.49994919e-01 9.94444430e-01 6.25572056e-02
-7.89779961e-01 -1.99617311e-01 -6.28376067e-01 2.52694160e-01
4.15828049e-01 8.13157916e-01 -8.01337540e-01 -2.74468541e-01
-2.93219507e-01 -5.32682613e-02 4.30564433e-01 1.27233818e-01
1.28872943e+00 -1.56111091e-01 -4.06593084e-01 1.49784505e-01
1.62800145e+00 5.39988458e-01 8.67543578e-01 4.38578874e-01
7.72480905e-01 7.09264874e-01 8.40358257e-01 5.11634052e-01
1.51575208e-01 4.68251199e-01 3.00740600e-01 -7.00773299e-01
-1.31771445e-01 -1.76061213e-01 4.42461371e-01 3.96895558e-01
2.36529395e-01 -4.32450145e-01 -8.38117719e-01 4.35807377e-01
-1.70135725e+00 -8.32381845e-01 -6.30205452e-01 2.53777122e+00
7.35985100e-01 1.45623162e-01 1.50104240e-01 2.07784787e-01
1.12045646e+00 -1.05866818e-02 -4.41680253e-01 -1.23630948e-01
-2.75129765e-01 -2.57497966e-01 6.50049150e-01 9.59151387e-02
-8.62428725e-01 6.51578546e-01 6.52589464e+00 5.27299166e-01
-1.34980571e+00 -2.90532500e-01 7.02907801e-01 -8.39078352e-02
-1.33886099e-01 7.47335777e-02 -7.36297011e-01 8.78843367e-01
8.74317884e-02 -4.91763093e-02 4.57218468e-01 4.49951023e-01
2.47738197e-01 -6.06232822e-01 -8.94046783e-01 1.24502516e+00
2.43133262e-01 -8.31295490e-01 1.83608547e-01 -9.16195214e-02
8.29210758e-01 -2.45640531e-01 1.18578546e-01 -2.64413685e-01
3.08747292e-01 -7.47363091e-01 8.71433377e-01 2.17355132e-01
9.73582029e-01 -6.86261773e-01 4.05872017e-01 3.22827816e-01
-1.06503570e+00 1.00321651e-01 -6.63945198e-01 1.86657026e-01
-2.07649931e-01 8.55019331e-01 -6.38152778e-01 2.34335735e-01
7.65198588e-01 3.01921308e-01 -8.48907351e-01 1.13354683e+00
-1.65994912e-01 2.68533587e-01 -2.67125368e-01 -4.42234576e-01
2.03030691e-01 -3.76334876e-01 8.21714327e-02 1.25365603e+00
4.14109558e-01 -1.33825585e-01 5.30955940e-02 8.03640723e-01
1.04299508e-01 2.15102598e-01 -6.07221842e-01 -3.43652740e-02
6.24086559e-01 1.11897862e+00 -1.08806956e+00 -3.07796896e-01
-7.15399146e-01 1.18861556e+00 -3.11942220e-01 5.92239678e-01
-8.81663024e-01 -8.18767011e-01 1.54949129e-01 1.54517770e-01
1.36253163e-01 -1.94881469e-01 -3.46355408e-01 -1.15246964e+00
1.92738891e-01 -1.05597174e+00 3.12871158e-01 -1.07393026e+00
-1.10432720e+00 4.72339869e-01 -2.84306616e-01 -1.61783171e+00
3.55461568e-01 -7.18113482e-01 -7.57532954e-01 9.05521154e-01
-1.29244113e+00 -8.30085576e-01 -8.10205162e-01 7.05173612e-01
6.17985129e-01 2.45714076e-02 6.51748836e-01 2.66264737e-01
-6.95431769e-01 3.92202169e-01 4.54926997e-01 4.91174102e-01
9.72537637e-01 -1.18812990e+00 3.90153229e-01 1.16234469e+00
4.96992394e-02 4.57415402e-01 5.91245830e-01 -6.19501591e-01
-1.54017949e+00 -6.81739926e-01 3.89626563e-01 -2.31208280e-01
3.89012605e-01 -5.87565720e-01 -8.03435028e-01 2.24936336e-01
2.57853001e-01 -1.13432094e-01 4.19825584e-01 -2.03124925e-01
-4.18471873e-01 -4.55937117e-01 -1.05533457e+00 9.61410582e-01
2.93035984e-01 -2.13413715e-01 -9.89367887e-02 2.38217250e-01
1.51916578e-01 -3.97789955e-01 -6.57856017e-02 -7.11160749e-02
6.82651162e-01 -1.27923870e+00 6.03160918e-01 1.10340089e-01
4.31239128e-01 -8.73805046e-01 5.26235253e-02 -8.87624562e-01
1.09984234e-01 -5.13269663e-01 1.81750625e-01 1.38869298e+00
2.44952813e-01 -7.39719629e-01 6.67493939e-01 6.79187000e-01
4.96447265e-01 -2.82476455e-01 -5.84725440e-01 -4.66631472e-01
-1.84957549e-01 1.26084030e-01 5.58696508e-01 9.25465465e-01
-1.13220520e-01 1.06707931e-01 -4.25078213e-01 1.43991679e-01
5.66605985e-01 2.37307817e-01 1.03760040e+00 -1.04173994e+00
-2.51622915e-01 -2.98880517e-01 -1.42856523e-01 -1.19122040e+00
-4.70586866e-01 -6.14003558e-03 2.36304253e-01 -1.59846592e+00
5.82017303e-01 -6.91773355e-01 -1.94708705e-01 3.20932478e-01
-3.96127671e-01 2.28215724e-01 3.00264597e-01 2.16779932e-01
-4.23498750e-01 1.00135773e-01 1.41674161e+00 -2.04603642e-01
-5.15614636e-02 -2.41273388e-01 -6.38674796e-01 7.17172205e-01
6.51790202e-01 -3.83199126e-01 -5.47626555e-01 -6.05574965e-01
1.13975771e-01 -3.76973338e-02 1.10629946e-01 -9.27864134e-01
1.31935760e-01 -5.04223824e-01 8.02931070e-01 -9.16097641e-01
5.67159578e-02 -9.96000111e-01 -1.96509641e-02 5.78472793e-01
1.55557260e-01 1.87483609e-01 1.56145364e-01 5.28924942e-01
-2.57128873e-03 -3.80578995e-01 1.00095975e+00 -1.94802627e-01
-6.88209236e-01 2.75022909e-02 -5.12769043e-01 -4.03313190e-02
1.24419355e+00 -6.23000205e-01 -7.55418718e-01 -4.83440965e-01
3.58478129e-01 1.50862932e-01 9.97528315e-01 4.24603641e-01
5.90783060e-01 -1.07636631e+00 -4.78077561e-01 2.60884225e-01
1.66099116e-01 7.13525042e-02 1.07053161e-01 5.19240856e-01
-1.12852359e+00 2.93609425e-02 -1.41044408e-01 -5.50007761e-01
-1.44540370e+00 6.91349566e-01 2.82424748e-01 2.13929236e-01
-7.31712520e-01 3.70960534e-01 6.78062618e-01 1.74103588e-01
3.85817587e-01 -1.44683063e-01 5.56861386e-02 -1.77429840e-01
8.23486209e-01 4.43834156e-01 -1.99115038e-01 -3.21435273e-01
-1.88445032e-01 7.35539556e-01 -3.02621126e-01 -1.92377076e-01
8.48939717e-01 -1.73214659e-01 -2.03317270e-01 6.26177430e-01
8.17514002e-01 4.59758639e-01 -1.36049378e+00 -1.27519786e-01
-4.04699534e-01 -9.58781540e-01 -2.40035608e-01 -8.09001386e-01
-1.11639392e+00 9.72464323e-01 7.80778766e-01 5.20963907e-01
1.34493816e+00 -4.89200264e-01 8.57715666e-01 1.80075005e-01
3.24681282e-01 -1.45260096e+00 4.47057396e-01 2.25281045e-01
3.60885233e-01 -1.07499671e+00 1.41823277e-01 -4.86883640e-01
-9.29039180e-01 1.40851128e+00 8.35380614e-01 1.10473335e-01
2.48110257e-02 3.73415530e-01 5.39183736e-01 5.53229488e-02
-4.76384372e-01 -8.23680684e-02 1.40950363e-02 5.18949747e-01
4.69314218e-01 -1.17641531e-01 -2.19550803e-01 -8.85739103e-02
3.65303814e-01 -2.94315666e-01 9.50758636e-01 1.08206379e+00
-5.16093433e-01 -1.06775606e+00 -8.32151413e-01 2.44802728e-01
-5.42013228e-01 5.41195925e-03 -6.65439248e-01 8.44545484e-01
-6.51146658e-03 1.22736847e+00 1.32765070e-01 -2.83187497e-02
2.32966229e-01 -2.95107037e-01 4.20518130e-01 -3.53793710e-01
-3.30102481e-02 2.76407093e-01 -2.81174332e-01 -2.40893558e-01
-1.94853142e-01 -4.39688265e-01 -1.37072778e+00 -4.99000400e-01
-5.33388317e-01 5.12078553e-02 6.45411313e-01 5.34440994e-01
8.43048468e-02 3.04873228e-01 8.87748003e-01 -5.45257211e-01
-8.79100636e-02 -7.94560492e-01 -9.27539349e-01 6.72589660e-01
1.35096043e-01 -5.29954374e-01 -5.00688434e-01 3.81386936e-01] | [11.847118377685547, 1.9478427171707153] |
bbf1041d-1aad-4503-ac8b-46b3433afebd | bda-sketret-bi-level-domain-adaptation-for | 2201.06570 | null | https://arxiv.org/abs/2201.06570v1 | https://arxiv.org/pdf/2201.06570v1.pdf | BDA-SketRet: Bi-Level Domain Adaptation for Zero-Shot SBIR | The efficacy of zero-shot sketch-based image retrieval (ZS-SBIR) models is governed by two challenges. The immense distributions-gap between the sketches and the images requires a proper domain alignment. Moreover, the fine-grained nature of the task and the high intra-class variance of many categories necessitates a class-wise discriminative mapping among the sketch, image, and the semantic spaces. Under this premise, we propose BDA-SketRet, a novel ZS-SBIR framework performing a bi-level domain adaptation for aligning the spatial and semantic features of the visual data pairs progressively. In order to highlight the shared features and reduce the effects of any sketch or image-specific artifacts, we propose a novel symmetric loss function based on the notion of information bottleneck for aligning the semantic features while a cross-entropy-based adversarial loss is introduced to align the spatial feature maps. Finally, our CNN-based model confirms the discriminativeness of the shared latent space through a novel topology-preserving semantic projection network. Experimental results on the extended Sketchy, TU-Berlin, and QuickDraw datasets exhibit sharp improvements over the literature. | ['Zeynep Akata', 'Anjan Dutta', 'Biplab Banerjee', 'Ruchika Chavan', 'Ushasi Chaudhuri'] | 2022-01-17 | null | null | null | null | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 1.84951514e-01 -4.00557876e-01 -1.38998404e-01 -2.08405241e-01
-6.74126387e-01 -6.45489097e-01 8.65918040e-01 -3.99100542e-01
-3.12123708e-02 3.94884408e-01 1.93304732e-01 3.37833315e-01
-5.88336766e-01 -6.50130153e-01 -6.14394724e-01 -7.41148829e-01
3.41974974e-01 1.01482227e-01 1.02597915e-01 -2.40801886e-01
5.44855952e-01 5.12475908e-01 -1.25772870e+00 1.04182415e-01
7.02634513e-01 1.22369146e+00 3.34334075e-01 1.77436888e-01
-2.64583498e-01 2.25966856e-01 -1.55913576e-01 -7.02091992e-01
6.99496269e-01 -3.52945715e-01 -5.44105709e-01 4.61644642e-02
9.04625773e-01 -4.69737709e-01 -7.22702622e-01 1.07935214e+00
5.36122501e-01 3.00865918e-01 7.36746907e-01 -1.59402680e+00
-1.05982161e+00 -9.55936909e-02 -3.94519180e-01 -2.85871536e-01
1.87391087e-01 2.75585633e-02 1.28374541e+00 -1.12596774e+00
1.04926491e+00 1.28885520e+00 5.75474143e-01 4.46331918e-01
-1.37808502e+00 -8.51822734e-01 1.68914482e-01 3.03141654e-01
-1.76607800e+00 -3.39778364e-01 1.42929602e+00 -4.76435006e-01
2.72288382e-01 1.58058390e-01 5.06561458e-01 1.36883330e+00
1.08815707e-01 4.90868062e-01 8.50566387e-01 -3.24539155e-01
2.84382880e-01 2.54590541e-01 -3.73498082e-01 6.84553862e-01
-1.23274505e-01 1.64194912e-01 -8.45118344e-01 -1.82448655e-01
1.30162013e+00 5.05314529e-01 -3.16804945e-02 -1.30083871e+00
-1.18772018e+00 7.69359946e-01 5.39932072e-01 4.03706610e-01
-1.56503037e-01 9.96715948e-02 3.75239313e-01 2.89844483e-01
3.65266353e-01 4.35771912e-01 -8.48575756e-02 2.30004176e-01
-1.08559215e+00 2.17446163e-01 4.15292919e-01 1.17234266e+00
5.72320402e-01 -2.33466282e-01 -3.37637901e-01 1.06434989e+00
1.62414610e-01 4.22296226e-01 2.09147617e-01 -7.71506011e-01
4.77399200e-01 5.20092785e-01 3.91045287e-02 -1.57882774e+00
2.58215964e-01 -2.01743335e-01 -9.39051688e-01 5.92196807e-02
1.83275506e-01 6.39656544e-01 -6.26358032e-01 1.83973801e+00
9.36721787e-02 2.85416394e-01 -2.31818467e-01 1.04696906e+00
5.44665694e-01 4.06431824e-01 1.43085718e-01 1.85657740e-01
1.10634303e+00 -8.84073913e-01 -7.18294322e-01 4.29434702e-02
-2.61881083e-01 -7.02489018e-01 1.34493005e+00 -1.16389073e-01
-1.14704323e+00 -6.73664570e-01 -1.27324009e+00 -4.01210636e-01
-6.16723061e-01 1.18355356e-01 2.02514842e-01 2.11909920e-01
-8.40282977e-01 7.99605072e-01 -4.78067607e-01 -4.69817460e-01
5.55702388e-01 6.33167699e-02 -5.09924173e-01 -1.18793651e-01
-1.06871450e+00 6.83040321e-01 7.55816549e-02 -9.03642625e-02
-5.95089734e-01 -9.62042153e-01 -7.56667912e-01 3.09880137e-01
2.82210857e-01 -6.73344195e-01 3.81531835e-01 -7.66822338e-01
-1.55102241e+00 8.51841271e-01 1.14901468e-01 8.22623670e-02
8.82393479e-01 -3.12611572e-02 -2.07166195e-01 2.68633008e-01
1.82778120e-01 7.86547363e-01 1.31208444e+00 -1.36008167e+00
-1.92145124e-01 -5.53719163e-01 -1.89255267e-01 2.43790418e-01
-6.10913754e-01 -4.06802416e-01 -5.93127012e-01 -1.10305893e+00
2.33037397e-01 -8.75086308e-01 3.71929854e-02 7.18184054e-01
-2.25176692e-01 -8.18779543e-02 1.02851605e+00 -7.06786990e-01
1.04538786e+00 -2.56602311e+00 5.15955269e-01 3.23829919e-01
1.66120648e-01 1.81664489e-02 -5.88189483e-01 6.21678650e-01
-7.92881250e-02 -4.10377830e-02 -2.79268831e-01 -3.48728269e-01
2.04666838e-01 4.62368913e-02 -7.72399664e-01 4.53627944e-01
3.53722274e-01 1.09104335e+00 -9.43280458e-01 -4.62609917e-01
5.05317390e-01 6.14952683e-01 -5.26701987e-01 3.04941744e-01
1.16991028e-01 4.56359088e-01 -5.09801149e-01 5.08077204e-01
8.85014713e-01 -1.86411232e-01 7.55015612e-02 -5.09459853e-01
1.29346162e-01 -2.25022957e-02 -9.41991687e-01 2.49431038e+00
-4.74095732e-01 4.87383157e-01 9.25942976e-03 -1.03489327e+00
1.18252587e+00 8.52929503e-02 6.02162182e-01 -9.96501327e-01
-2.44982302e-01 1.76526576e-01 -6.96066499e-01 -8.72525349e-02
4.40136939e-01 -2.68679947e-01 -1.82507977e-01 2.66566813e-01
2.51832277e-01 -3.68536234e-01 -3.51016104e-01 3.44584018e-01
6.02340221e-01 2.80425400e-01 4.98637781e-02 -3.76098871e-01
4.19772327e-01 -4.29035723e-01 3.12083393e-01 5.25802612e-01
-2.12120399e-01 9.06920433e-01 2.26152271e-01 -5.32192409e-01
-1.45358980e+00 -1.43345535e+00 -1.43898398e-01 9.18532848e-01
7.63965547e-01 -3.17239642e-01 -4.30941224e-01 -6.72985852e-01
2.29069635e-01 5.38250268e-01 -6.57685041e-01 -4.72493112e-01
-2.62542844e-01 1.89548373e-01 3.06859165e-01 3.44225138e-01
5.60353994e-01 -8.54148030e-01 -1.79233134e-01 7.49588683e-02
-2.26486728e-01 -1.18803394e+00 -9.13061261e-01 -3.50912392e-01
-6.38455510e-01 -7.62155652e-01 -9.13765788e-01 -7.87502766e-01
7.77893722e-01 4.49930429e-01 7.80384541e-01 -7.35100135e-02
-5.37653804e-01 5.87093055e-01 -1.44379541e-01 1.73021704e-01
1.06580697e-01 -1.05364442e-01 -1.21589795e-01 1.69749916e-01
5.92748448e-02 -8.80618453e-01 -8.89276206e-01 4.50954020e-01
-8.87592912e-01 9.15444642e-02 4.19148862e-01 1.17591965e+00
5.53324044e-01 -1.38982177e-01 5.22218704e-01 -2.36911446e-01
4.76975530e-01 -2.60518730e-01 -4.94646519e-01 6.15369856e-01
-5.32877684e-01 1.48690999e-01 5.69468915e-01 -5.71272135e-01
-9.04428065e-01 8.36652145e-02 3.09965551e-01 -9.59340692e-01
2.49832779e-01 -1.10185154e-01 -3.67460907e-01 -2.16370970e-01
2.46416822e-01 5.05387247e-01 2.29400963e-01 -4.40876931e-01
7.80727267e-01 3.52412283e-01 6.36925220e-01 -7.40149856e-01
1.21715713e+00 6.81441844e-01 1.35363013e-01 -8.03974271e-01
-5.04278064e-01 -5.21203339e-01 -7.73631871e-01 2.03836355e-02
8.57248962e-01 -8.95569086e-01 -4.48134989e-01 2.70763189e-01
-1.11320996e+00 1.74897492e-01 -5.33083320e-01 6.52000681e-03
-8.29625964e-01 6.57184720e-01 -2.59159118e-01 -5.76528609e-01
-4.68306124e-01 -1.08546114e+00 1.29382503e+00 -5.65242283e-02
-7.29482621e-02 -8.53537142e-01 1.71313867e-01 2.19291493e-01
4.70169604e-01 8.96194354e-02 1.00859571e+00 -3.42526853e-01
-8.87798667e-01 -1.23058207e-01 -5.92903674e-01 3.95949662e-01
2.00445056e-01 -8.71666819e-02 -6.81031883e-01 -4.46110666e-01
-3.71048629e-01 -3.08689088e-01 7.78162301e-01 8.41020569e-02
1.36768973e+00 -3.10548037e-01 -2.28773832e-01 7.26726890e-01
1.57311952e+00 -2.67356727e-02 6.31204069e-01 7.37804398e-02
7.81516731e-01 5.84995806e-01 5.70876658e-01 4.84956264e-01
2.12310940e-01 9.95098531e-01 1.93170592e-01 -1.17983699e-01
-3.25997978e-01 -8.89412105e-01 9.40622091e-02 7.41114795e-01
2.69421637e-01 7.30083659e-02 -3.37850064e-01 5.95809758e-01
-1.87322557e+00 -1.03946495e+00 6.06817126e-01 2.30758047e+00
6.71288013e-01 -3.89272898e-01 -1.32290423e-01 -1.08418174e-01
7.04055250e-01 3.68905604e-01 -5.05095899e-01 -6.46421984e-02
-1.06574647e-01 1.93345949e-01 3.53634357e-01 2.85446286e-01
-1.14412904e+00 9.81323659e-01 5.68274260e+00 1.28207397e+00
-1.18494523e+00 -1.68380495e-02 2.82591462e-01 -1.83540449e-01
-4.77998644e-01 -1.14429049e-01 -1.96303502e-01 5.92164934e-01
1.71611398e-01 -2.75334686e-01 7.11912155e-01 7.55550265e-01
-2.20601007e-01 4.80791479e-01 -1.09903765e+00 1.27848387e+00
1.30808815e-01 -1.56787062e+00 3.49258959e-01 8.77755955e-02
7.20840573e-01 -4.37261403e-01 3.96481901e-01 -2.82107834e-02
-3.45089287e-02 -7.30862439e-01 9.63302135e-01 6.83892369e-01
1.24858975e+00 -6.67845130e-01 3.65798064e-02 -1.25416443e-01
-1.29171729e+00 -4.65421602e-02 -5.38195431e-01 3.72666597e-01
-2.26112843e-01 1.19569808e-01 -3.38747412e-01 6.44604564e-01
5.63004494e-01 9.54716206e-01 -4.05411810e-01 6.48993850e-01
6.49276152e-02 -1.97636902e-01 -8.10817182e-02 2.22520322e-01
1.11421958e-01 -5.27358472e-01 7.16751873e-01 9.29900169e-01
3.54193807e-01 -9.30677205e-02 3.43653560e-02 1.33428323e+00
-8.01139548e-02 5.56155965e-02 -9.43511844e-01 -1.98854804e-01
5.57963133e-01 1.29120398e+00 -4.90046501e-01 -1.42658919e-01
-1.60469577e-01 1.48694634e+00 4.00566399e-01 5.29485941e-01
-7.26585507e-01 -3.09619278e-01 8.89619231e-01 3.79597805e-02
4.63941097e-01 -1.86255619e-01 -5.12366652e-01 -1.22754276e+00
3.37748796e-01 -4.71800655e-01 8.09217840e-02 -6.73185170e-01
-1.68950868e+00 2.17928380e-01 -2.16264963e-01 -1.53438210e+00
1.14903294e-01 -4.01884675e-01 -4.78496701e-01 8.68074894e-01
-1.53864992e+00 -1.50972283e+00 -2.69120961e-01 8.53618443e-01
5.87472677e-01 -3.50617111e-01 7.94811428e-01 4.18206036e-01
-3.33346754e-01 9.37564611e-01 3.71137559e-01 1.40880540e-01
9.66925323e-01 -8.03412080e-01 3.42562646e-01 4.54133242e-01
9.49087664e-02 7.15514421e-01 2.60776699e-01 -4.68704313e-01
-1.59964406e+00 -9.83568370e-01 6.66928887e-01 -4.07587498e-01
6.16647422e-01 -6.95863068e-01 -8.43602419e-01 2.89891899e-01
-8.33822563e-02 2.30320722e-01 4.07300770e-01 -9.29787010e-02
-9.64160442e-01 -3.48630369e-01 -1.23800707e+00 7.74768174e-01
1.35645592e+00 -1.14034355e+00 -4.73580569e-01 1.72601044e-01
6.19627774e-01 9.66442153e-02 -9.74021673e-01 1.95657521e-01
1.18291819e+00 -8.48094046e-01 1.50544059e+00 -5.32590628e-01
7.79919505e-01 -9.27762389e-02 -3.97617698e-01 -1.07926393e+00
-5.44320405e-01 -3.81562352e-01 2.74025410e-01 1.29011834e+00
-1.91726729e-01 -4.00367022e-01 5.81128478e-01 7.46412158e-01
3.45667362e-01 -4.50923592e-01 -1.12820137e+00 -9.62525845e-01
9.53394845e-02 3.06629427e-02 6.20392084e-01 1.08446658e+00
4.67181467e-02 1.87468976e-01 -5.95814347e-01 -1.45908222e-01
9.11813617e-01 3.02217692e-01 7.50965476e-01 -1.16162419e+00
3.17797028e-02 -6.44528627e-01 -6.43741190e-01 -8.78149033e-01
2.31698781e-01 -9.10339415e-01 -1.08283423e-01 -1.00967872e+00
4.33130980e-01 -6.10579073e-01 -6.51767850e-01 2.11169675e-01
4.84199189e-02 1.70849040e-01 5.41504800e-01 4.45984215e-01
-6.77440286e-01 1.16164339e+00 1.49298751e+00 -3.37390989e-01
7.25531131e-02 -4.53876644e-01 -3.45492184e-01 2.48560756e-01
2.75561094e-01 -2.01667666e-01 -5.49514532e-01 -2.34419540e-01
-4.76822741e-02 2.77523939e-02 7.57440567e-01 -6.44204021e-01
2.88060755e-01 -2.54914999e-01 3.53109062e-01 -5.05008161e-01
5.51553726e-01 -1.21708763e+00 1.84269682e-01 2.15031192e-01
-6.15382552e-01 -3.00863594e-01 -6.49441555e-02 9.30423677e-01
-2.59574324e-01 2.56763011e-01 9.12585139e-01 5.44484146e-02
-7.02514529e-01 6.52037084e-01 4.98158514e-01 -1.45984188e-01
9.79857802e-01 -4.27305549e-01 -1.24603108e-01 -2.27877378e-01
-3.92628193e-01 -5.93503006e-02 6.88429177e-01 8.26067924e-01
8.62954319e-01 -1.78215075e+00 -4.51972038e-01 6.11039698e-01
5.12009203e-01 -2.62918323e-01 6.88318312e-01 5.30276954e-01
-9.49300453e-02 5.25370955e-01 -7.62710810e-01 -4.44194674e-01
-1.08619320e+00 7.13707805e-01 5.14059216e-02 -2.31102154e-01
-8.00524712e-01 7.23754227e-01 6.93182290e-01 -5.10396481e-01
3.06573927e-01 2.00791657e-01 2.71319866e-01 -6.31389469e-02
2.76780903e-01 3.86222064e-01 -1.59745663e-01 -4.26917970e-01
-4.26347524e-01 7.97501087e-01 2.70026107e-03 -1.32965013e-01
1.28130102e+00 -2.71705627e-01 -7.22246245e-02 2.09886387e-01
1.43662179e+00 -2.57275939e-01 -1.75200009e+00 -4.90848213e-01
-1.75967157e-01 -9.00708795e-01 1.88769568e-02 -6.33346021e-01
-1.04860699e+00 1.01258528e+00 6.63877487e-01 -1.78040981e-01
1.08368099e+00 -8.27388242e-02 8.40345860e-01 8.65643993e-02
4.38162684e-01 -1.20076871e+00 4.94555175e-01 1.38587609e-01
1.37665164e+00 -1.12716341e+00 -5.95326126e-02 -3.38855833e-01
-5.86390018e-01 9.53360021e-01 3.28129768e-01 -4.45316643e-01
6.80159211e-01 -2.53028721e-01 -3.40200394e-01 -1.21296674e-01
-4.23040450e-01 2.34272927e-01 6.23037219e-01 5.96915483e-01
-6.44048303e-02 7.26603493e-02 -2.42155492e-01 5.54083645e-01
1.04587629e-01 9.02200336e-05 -2.06061080e-01 6.18958414e-01
2.00805068e-02 -1.08118153e+00 -8.32859576e-02 -1.40527338e-02
9.67922360e-02 -1.04889788e-01 -5.33999920e-01 6.07968390e-01
-1.20666124e-01 3.54172885e-01 1.50548652e-01 -4.68989521e-01
3.14321548e-01 7.61879906e-02 6.09716058e-01 -9.14419293e-02
-1.14007086e-01 -8.58479887e-02 -5.23862898e-01 -7.25338042e-01
-1.06638148e-01 -3.62665653e-01 -7.27926672e-01 -3.09456259e-01
-1.66000053e-01 -3.71942490e-01 8.30100298e-01 7.05521286e-01
7.17752218e-01 5.59970923e-02 9.19207871e-01 -8.61291289e-01
-8.09598744e-01 -5.26807487e-01 -8.05138886e-01 8.74172628e-01
1.78791359e-01 -1.00968599e+00 -2.74590880e-01 -1.01324424e-01] | [11.602238655090332, 0.7169002890586853] |
ebaeb61b-ba93-40b3-ae82-99a0aeede933 | sana-a-large-scale-multi-genre-multi-dialect | null | null | https://aclanthology.org/L14-1702 | https://aclanthology.org/L14-1702.pdf | SANA: A Large Scale Multi-Genre, Multi-Dialect Lexicon for Arabic Subjectivity and Sentiment Analysis | The computational treatment of subjectivity and sentiment in natural language is usually significantly improved by applying features exploiting lexical resources where entries are tagged with semantic orientation (e.g., positive, negative values). In spite of the fair amount of work on Arabic sentiment analysis over the past few years (e.g., (Abbasi et al., 2008; Abdul-Mageed et al., 2014; Abdul-Mageed et al., 2012; Abdul-Mageed and Diab, 2012a; Abdul-Mageed and Diab, 2012b; Abdul-Mageed et al., 2011a; Abdul-Mageed and Diab, 2011)), the language remains under-resourced as to these polarity repositories compared to the English language. In this paper, we report efforts to build and present SANA, a large-scale, multi-genre, multi-dialect multi-lingual lexicon for the subjectivity and sentiment analysis of the Arabic language and dialects. | ['Muhammad Abdul-Mageed', 'Mona Diab'] | 2014-05-01 | null | null | null | lrec-2014-5 | ['arabic-sentiment-analysis'] | ['natural-language-processing'] | [-4.52966809e-01 2.48670757e-01 -2.74313420e-01 -4.67028916e-01
-4.05877829e-01 -9.42754209e-01 7.01804340e-01 6.51940405e-01
-1.89341664e-01 7.53193080e-01 5.93164980e-01 -2.56662309e-01
-2.84948833e-02 -8.31648409e-01 -1.90261364e-01 -2.01375857e-01
4.29851376e-02 3.45561266e-01 -8.25536400e-02 -1.05137360e+00
5.21380484e-01 2.96542525e-01 -1.68449664e+00 3.11775386e-01
7.54195631e-01 9.47714627e-01 -9.85052809e-03 1.93057328e-01
1.24592390e-02 9.75057840e-01 -3.35659623e-01 -9.67363417e-01
-2.27168575e-01 -4.70599145e-01 -1.19595027e+00 1.29756466e-01
1.68957219e-01 2.37731725e-01 2.37409309e-01 1.16004169e+00
3.93583983e-01 -3.13305348e-01 8.10119331e-01 -9.55265820e-01
-9.17348683e-01 6.13787353e-01 -7.28611112e-01 -1.10442504e-01
7.06095695e-01 -4.36397761e-01 1.09901392e+00 -1.21111262e+00
8.76702666e-01 1.24296105e+00 5.99815905e-01 3.99012566e-01
-5.70455253e-01 -5.48205078e-01 3.35000604e-01 2.07395613e-01
-1.17578793e+00 -3.60102028e-01 5.75062871e-01 -4.47408080e-01
5.79716563e-01 7.40211532e-02 6.03459179e-01 5.63850045e-01
8.15461501e-02 8.54970872e-01 1.53735733e+00 -9.57512558e-01
9.94826779e-02 4.95476753e-01 3.57030451e-01 6.31858408e-01
1.88189432e-01 -6.24244571e-01 -6.79442048e-01 -1.15434289e-01
3.36556360e-02 -1.81528211e-01 1.88949555e-01 -7.39058629e-02
-8.62712860e-01 1.00357127e+00 6.32867292e-02 5.96856654e-01
-6.95279598e-01 -7.56640613e-01 6.61899924e-01 6.41387820e-01
8.18639398e-01 3.30141664e-01 -8.30389261e-01 -4.86665010e-01
-5.98410070e-01 5.74213862e-01 9.48257089e-01 9.76937056e-01
6.41122878e-01 -2.43601706e-02 5.85672379e-01 1.23481894e+00
5.63327909e-01 5.33651888e-01 7.12010503e-01 -4.12919372e-01
2.47239113e-01 9.04610872e-01 1.14314295e-01 -1.24304843e+00
-4.57658619e-01 1.31264657e-01 -3.67342561e-01 -1.49553314e-01
5.93978941e-01 -3.64836305e-01 -6.28719509e-01 1.57417893e+00
3.04385036e-01 -8.62451673e-01 6.77297831e-01 9.68382299e-01
1.04516399e+00 7.57982254e-01 1.42862588e-01 -2.48416767e-01
1.76739240e+00 -8.81510198e-01 -7.64674664e-01 -3.80994350e-01
6.06915653e-01 -1.45898759e+00 1.05439675e+00 7.74769068e-01
-1.29791594e+00 -1.74351651e-02 -8.26117635e-01 1.47764176e-01
-8.73252511e-01 1.41357750e-01 9.41193938e-01 1.01772237e+00
-9.68654037e-01 -1.75702795e-01 -5.52961886e-01 -8.88668537e-01
1.07305288e-01 1.60738081e-01 -4.99387771e-01 -6.47115186e-02
-1.42981851e+00 1.30796397e+00 1.15993701e-01 -5.53497747e-02
-2.55647421e-01 -1.68567821e-01 -9.18552279e-01 -8.32244456e-01
4.12462533e-01 3.99421126e-01 1.08502281e+00 -1.73652685e+00
-1.55179632e+00 1.47314310e+00 -3.03770930e-01 -1.21215507e-02
-1.41522676e-01 -1.87287062e-01 -1.12999821e+00 -8.12678710e-02
4.00496840e-01 3.82692099e-01 3.41985941e-01 -1.07972574e+00
-6.62431180e-01 -7.60481536e-01 6.40136838e-01 4.11621690e-01
-6.25066221e-01 8.40661824e-01 -5.83967492e-02 -9.38249230e-01
2.52215832e-01 -9.68729198e-01 8.09394047e-02 -6.67036712e-01
-2.84951758e-02 -1.20109916e-01 6.35602772e-01 -8.64471376e-01
1.37067652e+00 -1.89748704e+00 1.01603590e-01 1.53083980e-01
-8.95566344e-02 1.15404636e-01 2.46270329e-01 9.97768223e-01
-1.99854612e-01 -7.74670169e-02 -3.14755917e-01 3.28444242e-01
1.05147764e-01 1.62184209e-01 -2.41234705e-01 6.55395806e-01
-1.34360883e-02 6.37045622e-01 -7.92832911e-01 -3.53332520e-01
-1.31726339e-01 2.50604421e-01 -3.93511266e-01 -4.61100250e-01
1.07054211e-01 1.22178540e-01 -3.08674574e-01 1.33531892e+00
4.88977700e-01 -6.97325030e-03 5.87726772e-01 -2.70108461e-01
-6.11439407e-01 5.40051162e-01 -1.29766929e+00 9.72485960e-01
-6.88963532e-01 6.06241763e-01 4.43334490e-01 -1.02121294e+00
9.26602840e-01 6.06402755e-01 3.86059016e-01 -7.40364850e-01
7.05075026e-01 7.95613706e-01 5.41812517e-02 -1.80978522e-01
1.02167153e+00 -4.04881030e-01 -5.55641532e-01 5.67944586e-01
-1.79652452e-01 -1.07685335e-01 7.26792634e-01 1.58427164e-01
9.64443013e-03 1.15111791e-01 7.33936846e-01 -8.98109317e-01
1.12190402e+00 2.74095803e-01 2.19770208e-01 -5.71525842e-03
-4.29794967e-01 4.30051953e-01 5.47755539e-01 -3.76828611e-01
-5.26446223e-01 -5.03475845e-01 -7.52157927e-01 1.45009243e+00
1.09874643e-01 -6.68262839e-01 -7.83681989e-01 -6.25398099e-01
-2.46351920e-02 3.83037925e-01 -7.02622592e-01 2.78304547e-01
-4.16621298e-01 -8.60825717e-01 4.85608906e-01 3.56280096e-02
2.34979182e-01 -1.06068945e+00 -4.34331119e-01 3.45743299e-01
-2.06440032e-01 -6.92349374e-01 -3.31118889e-02 1.50876462e-01
-5.45840442e-01 -9.37053382e-01 -7.01399386e-01 -1.19584167e+00
4.18869793e-01 -1.24244332e-01 1.34823525e+00 -1.97632611e-01
1.98385596e-01 2.68946528e-01 -8.90535116e-01 -9.38744009e-01
-4.02407527e-01 -2.16777250e-01 1.27066061e-01 -1.42440915e-01
9.18626726e-01 6.24000914e-02 -2.72910058e-01 3.03407013e-01
-8.43038023e-01 -4.05208647e-01 2.26700902e-01 7.24179089e-01
4.59593385e-01 -2.20830053e-01 1.12770975e+00 -8.28754544e-01
7.04245031e-01 -6.06325090e-01 -2.80507445e-01 -2.73202173e-02
-6.93578839e-01 -6.17708743e-01 5.87282240e-01 -7.66143054e-02
-1.00386441e+00 -5.30576766e-01 -2.45152816e-01 8.53927732e-01
-4.03959304e-02 9.93832946e-01 -2.22694173e-01 -1.56943351e-01
5.54341733e-01 -1.45711675e-01 -4.47662324e-02 -2.61956125e-01
4.07143593e-01 1.21071529e+00 -1.24355577e-01 -6.00907981e-01
3.08483869e-01 3.89277428e-01 -2.78949291e-01 -9.12856400e-01
-1.05402970e+00 -3.56504738e-01 -6.57078862e-01 -5.85167646e-01
7.00874090e-01 -1.23517931e+00 -2.57366240e-01 8.19687545e-01
-6.16738498e-01 5.33822849e-02 2.68896967e-01 4.01302457e-01
-2.57696509e-01 3.66780013e-01 -7.70537138e-01 -8.67884338e-01
-2.17855021e-01 -1.14230335e+00 5.62628627e-01 6.74949214e-03
-8.21298182e-01 -1.30472374e+00 2.88274735e-02 2.99260229e-01
3.78989190e-01 2.27826625e-01 7.64488161e-01 -5.97746789e-01
3.75388891e-01 -1.12911962e-01 2.40514125e-03 2.77675182e-01
3.37238252e-01 4.51423153e-02 -5.82135081e-01 -2.10697159e-01
3.55463289e-02 -7.75124252e-01 1.77283928e-01 3.39631110e-01
3.56148243e-01 -2.97023714e-01 2.86818385e-01 -2.55465329e-01
1.27096546e+00 3.18896800e-01 5.71274340e-01 1.19862831e+00
2.07592100e-01 9.32476819e-01 1.15619886e+00 4.30733293e-01
1.08782375e+00 3.60613793e-01 7.10272193e-02 1.38827637e-01
2.96351552e-01 2.06684500e-01 3.63882303e-01 1.02261078e+00
-1.30924344e-01 -1.08117916e-01 -9.76794899e-01 1.01968777e+00
-1.54135001e+00 -8.09140384e-01 -3.33654225e-01 1.85351646e+00
1.15176094e+00 -1.15110748e-01 3.71914893e-01 2.96433657e-01
5.02113938e-01 2.15546668e-01 1.75361872e-01 -9.28851068e-01
-5.05566001e-01 2.31536373e-01 3.56631160e-01 5.07806242e-01
-1.04103267e+00 1.19816256e+00 4.22818851e+00 5.44160545e-01
-1.12739992e+00 1.33743286e-01 4.13156450e-01 2.47837856e-01
-2.40161121e-01 -2.70756837e-02 -6.84403181e-01 4.30885613e-01
1.01889610e+00 -2.52151787e-01 3.20958078e-01 8.36164713e-01
-2.02901676e-01 -3.97312880e-01 -4.85494018e-01 7.07698226e-01
5.45773208e-01 -8.56436193e-01 -4.30270940e-01 -1.40423343e-01
9.22821701e-01 1.68420509e-01 1.60834432e-01 3.21980268e-01
1.99905396e-01 -8.41377974e-01 1.34902084e+00 -1.93951204e-01
5.19688308e-01 -1.10023642e+00 9.92062926e-01 1.80991024e-01
-8.58839273e-01 2.53132898e-02 -1.75308779e-01 -5.04340768e-01
1.27135053e-01 5.65714538e-01 5.56908958e-02 5.17584860e-01
1.11972821e+00 7.97068000e-01 -4.16572452e-01 9.04497653e-02
-1.08707495e-01 7.73092151e-01 -2.17588581e-02 -4.21521455e-01
5.67457318e-01 -6.57655001e-01 2.71830052e-01 1.22639370e+00
1.35466799e-01 1.72552481e-01 1.54933289e-01 -5.69336489e-02
1.43043607e-01 9.08817887e-01 -4.41542178e-01 -3.85569513e-01
1.89774275e-01 1.26815808e+00 -1.09260821e+00 -2.38447517e-01
-7.56819308e-01 4.79684293e-01 7.63676580e-05 1.64377302e-01
-2.72845268e-01 -7.23237574e-01 3.75815272e-01 6.75412342e-02
5.29693030e-02 1.54633764e-02 -4.52735126e-01 -9.86950278e-01
7.32113374e-03 -1.42068589e+00 5.24678648e-01 -5.11302948e-01
-1.58939207e+00 7.79816031e-01 3.21098194e-02 -1.19846368e+00
-2.28420407e-01 -9.27280366e-01 2.23572385e-02 1.18776798e+00
-1.35564005e+00 -1.31736159e+00 2.71748722e-01 4.64335024e-01
2.66211092e-01 -5.83776355e-01 1.08101821e+00 5.33695102e-01
-1.88918769e-01 3.25482875e-01 1.79219604e-01 1.00434639e-01
1.00801289e+00 -1.07182586e+00 2.56860666e-02 5.50845385e-01
-2.76561826e-01 7.86915481e-01 7.37815440e-01 -7.06622779e-01
-1.41273677e+00 -5.62793553e-01 1.44784188e+00 -6.36062682e-01
9.86007929e-01 -2.04928666e-01 -7.88663387e-01 8.22483420e-01
8.75707626e-01 -7.21690178e-01 1.03995979e+00 3.02181363e-01
-3.01245779e-01 5.66642322e-02 -1.31292951e+00 8.08821261e-01
2.17227995e-01 -5.37241936e-01 -8.40402067e-01 3.48195404e-01
-1.45696923e-01 -4.66387689e-01 -1.22763669e+00 1.98983341e-01
5.57140768e-01 -9.31745768e-01 4.86075521e-01 -4.18208510e-01
6.38652921e-01 -3.26214433e-01 -4.93195206e-01 -1.26757395e+00
3.27846818e-02 -1.71133086e-01 4.79724795e-01 1.27507186e+00
6.60849035e-01 -8.23049963e-01 3.47645670e-01 2.41927534e-01
-2.69303560e-01 -5.65766931e-01 -7.51348495e-01 -6.15695640e-02
5.11846483e-01 -4.35927451e-01 4.70566839e-01 1.16767967e+00
7.03604460e-01 3.66636783e-01 -2.48220146e-01 -2.76726931e-02
-8.60803353e-04 2.23809391e-01 4.25257474e-01 -1.23222828e+00
2.22038195e-01 -5.78723848e-01 -3.65716994e-01 -4.52154309e-01
1.05135769e-01 -9.22126710e-01 -1.84660435e-01 -1.74678564e+00
-8.59460309e-02 -2.98272461e-01 -5.02023175e-02 4.44534451e-01
-7.51483953e-03 6.01198077e-01 -9.28567499e-02 1.91376209e-01
-4.14342940e-01 1.95067987e-01 9.53461885e-01 2.76117802e-01
-1.96744487e-01 -4.00515139e-01 -1.27519643e+00 1.00664020e+00
8.61167014e-01 -1.77954674e-01 -2.69492745e-01 -2.54080426e-02
9.40575957e-01 -3.72440606e-01 -1.58524349e-01 -1.82881474e-01
-1.33118317e-01 -4.02250350e-01 1.81723297e-01 -6.61993921e-01
1.90703496e-01 -5.83695352e-01 -3.02455664e-01 2.44612321e-01
-1.28373936e-01 5.98154604e-01 3.45343024e-01 -1.63850747e-02
-7.72722125e-01 -5.48150778e-01 8.26932371e-01 -1.58360571e-01
-8.79204631e-01 -2.51096159e-01 -8.22404206e-01 1.84316665e-01
1.02688122e+00 -2.09064975e-01 -2.36809224e-01 -5.86172223e-01
-3.71892512e-01 2.27463543e-01 7.01046109e-01 5.11456907e-01
2.33168080e-01 -1.17183197e+00 -6.46258235e-01 1.63552731e-01
4.86648351e-01 -4.92085844e-01 1.60003945e-01 1.24751258e+00
-5.30713677e-01 4.88850176e-01 -3.88651013e-01 1.22927152e-01
-9.55055475e-01 2.40263820e-01 -9.71666947e-02 3.19852717e-02
5.94705790e-02 5.95682502e-01 -8.15292597e-02 -7.71889448e-01
-3.20798427e-01 2.47949570e-01 -7.43580639e-01 9.50430930e-01
4.88947093e-01 3.33715767e-01 3.59253675e-01 -1.63166261e+00
-6.21779025e-01 5.81795216e-01 -2.23742202e-01 -2.33332857e-01
1.40042543e+00 -4.75275218e-01 -7.86078036e-01 8.15000296e-01
8.71329188e-01 6.66890323e-01 1.05585456e-01 5.01947962e-02
2.11800560e-01 -3.11382800e-01 8.50611702e-02 -1.14962769e+00
-7.66338468e-01 6.10906899e-01 3.32037896e-01 6.12041354e-01
1.10365689e+00 -8.14927891e-02 4.23813015e-01 2.22806007e-01
5.58247387e-01 -1.80278587e+00 -3.48142624e-01 1.12115085e+00
8.76392841e-01 -1.20049572e+00 -5.83307073e-03 -3.92490715e-01
-1.28820825e+00 1.15145409e+00 4.57154274e-01 2.59720951e-01
1.06416368e+00 3.69192585e-02 8.45123589e-01 -4.81616765e-01
-2.51377165e-01 -3.64816993e-01 3.01300704e-01 3.46207738e-01
1.26584804e+00 2.46187881e-01 -1.30604315e+00 8.24336231e-01
-6.57230258e-01 -1.89140096e-01 9.19984818e-01 1.14757931e+00
-3.50637019e-01 -1.10331380e+00 -6.39648855e-01 5.61749458e-01
-9.39872086e-01 -2.06439108e-01 -5.75841844e-01 1.11154687e+00
-1.48600772e-01 1.34792519e+00 -8.92694443e-02 -2.42444687e-03
5.24230123e-01 2.16889996e-02 3.79092664e-01 -7.25372434e-01
-9.25477266e-01 3.46459717e-01 6.53172016e-01 1.62221923e-01
-1.09561861e+00 -8.80954742e-01 -1.25113869e+00 -2.80258805e-01
-3.59471440e-01 4.84587669e-01 6.28254592e-01 1.04878807e+00
-1.23941250e-01 7.44620785e-02 4.44257140e-01 -4.43250239e-01
-2.12914437e-01 -1.22149324e+00 -9.65249538e-01 1.47993222e-01
-4.35915403e-02 -8.11448753e-01 -3.35631043e-01 2.70714015e-01] | [11.100010871887207, 6.919246673583984] |
61642c05-7886-4d61-8ccf-a309690f25e5 | clusterllm-large-language-models-as-a-guide | 2305.14871 | null | https://arxiv.org/abs/2305.14871v1 | https://arxiv.org/pdf/2305.14871v1.pdf | ClusterLLM: Large Language Models as a Guide for Text Clustering | We introduce ClusterLLM, a novel text clustering framework that leverages feedback from an instruction-tuned large language model, such as ChatGPT. Compared with traditional unsupervised methods that builds upon "small" embedders, ClusterLLM exhibits two intriguing advantages: (1) it enjoys the emergent capability of LLM even if its embeddings are inaccessible; and (2) it understands the user's preference on clustering through textual instruction and/or a few annotated data. First, we prompt ChatGPT for insights on clustering perspective by constructing hard triplet questions <does A better correspond to B than C>, where A, B and C are similar data points that belong to different clusters according to small embedder. We empirically show that this strategy is both effective for fine-tuning small embedder and cost-efficient to query ChatGPT. Second, we prompt ChatGPT for helps on clustering granularity by carefully designed pairwise questions <do A and B belong to the same category>, and tune the granularity from cluster hierarchies that is the most consistent with the ChatGPT answers. Extensive experiments on 14 datasets show that ClusterLLM consistently improves clustering quality, at an average cost of ~$0.6 per dataset. | ['Jingbo Shang', 'Zihan Wang', 'Yuwei Zhang'] | 2023-05-24 | null | null | null | null | ['text-clustering'] | ['natural-language-processing'] | [-5.43318391e-01 -5.26521727e-02 -2.77344227e-01 -5.09439230e-01
-1.13701820e+00 -8.04168761e-01 1.96574524e-01 8.02648604e-01
-2.22886086e-01 1.35718584e-01 3.94245446e-01 -5.26142597e-01
-3.66756499e-01 -4.73348141e-01 -5.91196775e-01 -5.98370492e-01
-2.23201662e-01 9.79639411e-01 4.04190689e-01 -2.22885236e-01
6.65491343e-01 -1.96796313e-01 -1.56710958e+00 5.30525804e-01
1.10358644e+00 5.76900125e-01 2.77347684e-01 4.90578562e-01
-2.80866504e-01 8.18731427e-01 -5.49120843e-01 -2.54611075e-01
4.19644900e-02 -3.14493448e-01 -1.16315639e+00 1.29852414e-01
4.29488361e-01 -2.91373972e-02 7.48106604e-03 9.03824925e-01
1.96525410e-01 2.57677197e-01 6.57595992e-01 -1.32450819e+00
-6.89403296e-01 1.32859731e+00 -7.22418010e-01 2.38619342e-01
3.74869198e-01 2.90703595e-01 1.46537519e+00 -7.07849562e-01
5.32771111e-01 1.31397498e+00 5.32936811e-01 2.18583122e-01
-1.35547268e+00 -5.90883493e-01 2.78113842e-01 1.21435255e-01
-1.51208985e+00 -2.91782260e-01 5.24253786e-01 -5.50928712e-01
8.11070323e-01 4.24167097e-01 1.68354794e-01 6.38432086e-01
2.45347619e-01 7.70725250e-01 1.07901275e+00 -2.96346456e-01
5.64636350e-01 1.93540320e-01 6.91241443e-01 6.59552813e-01
-1.44748330e-01 -5.94035804e-01 -7.22466111e-01 -2.93675393e-01
1.17307924e-01 -2.07949299e-02 -1.00390077e-01 -5.13703108e-01
-1.26078033e+00 9.87362027e-01 3.65088582e-01 4.27420437e-01
1.53178528e-01 1.30197585e-01 2.82044947e-01 5.34760773e-01
1.90017328e-01 9.58832562e-01 -5.24012148e-01 -6.40754640e-01
-7.63210654e-01 -1.04385629e-01 1.03279126e+00 1.39887512e+00
1.04552329e+00 -6.15656674e-01 7.03454465e-02 6.78710938e-01
2.53021002e-01 1.10480592e-01 7.97975481e-01 -1.28949821e+00
5.27254403e-01 1.00049150e+00 -1.88273519e-01 -8.80789816e-01
-3.35534573e-01 -1.14631236e-01 -4.65632856e-01 -2.57337123e-01
3.88055474e-01 -9.50041115e-02 -4.20035332e-01 1.50253546e+00
1.11344144e-01 -2.35040396e-01 -1.66701823e-01 5.66606104e-01
6.10760689e-01 5.39570749e-01 -2.00278863e-01 7.84283280e-02
1.48887396e+00 -9.89710510e-01 -5.70315063e-01 -2.79337674e-01
1.30767536e+00 -8.32397461e-01 1.61345148e+00 4.67357457e-01
-6.74806535e-01 -5.14657617e-01 -8.08083296e-01 -1.91553131e-01
-4.30515647e-01 -1.79753900e-01 5.26300192e-01 6.36161804e-01
-1.44703603e+00 4.01853323e-01 -9.12578344e-01 -6.44071400e-01
-2.57170852e-02 4.67632741e-01 -4.97306250e-02 -2.09075168e-01
-7.33046532e-01 4.37760442e-01 5.10260165e-01 -4.56372291e-01
-7.58694589e-01 -5.38375974e-01 -6.71701729e-01 1.74621627e-01
6.84245884e-01 -1.48636997e-01 1.15313494e+00 -3.83238673e-01
-1.28502905e+00 6.22003853e-01 -3.23688298e-01 -1.80733994e-01
1.51019871e-01 -1.69330090e-01 -1.62305504e-01 2.76537389e-01
3.35923314e-01 8.12296212e-01 4.29130584e-01 -1.35546446e+00
-6.00607634e-01 -3.34795356e-01 -1.64178431e-01 3.24671000e-01
-7.59002030e-01 2.54340400e-03 -8.01719904e-01 -1.86819524e-01
1.70371622e-01 -9.22144353e-01 -2.73087472e-01 -5.53256929e-01
-8.85405064e-01 -7.67665565e-01 1.06814802e+00 8.48802850e-02
1.55927455e+00 -2.26190758e+00 -3.76717150e-02 5.06695926e-01
8.86090457e-01 -2.21873388e-01 -1.68771133e-01 8.49954486e-01
1.11521617e-01 6.40946150e-01 1.12825498e-01 -3.06474030e-01
4.43128556e-01 2.27386177e-01 -1.00012764e-01 1.37085378e-01
-7.00433552e-02 7.92757392e-01 -1.18200564e+00 -8.02923918e-01
1.11857601e-01 -1.69064984e-01 -9.92055357e-01 4.03484911e-01
-3.55129600e-01 2.26569891e-01 -3.43351036e-01 4.89841580e-01
2.96149403e-01 -6.95244789e-01 4.64746922e-01 2.00483464e-02
-3.09191912e-01 5.16456723e-01 -1.03298497e+00 1.51856911e+00
7.34458193e-02 6.87757909e-01 1.37638047e-01 -8.43626976e-01
8.88936281e-01 -3.43660032e-03 4.32900131e-01 -2.87590325e-01
1.39865607e-01 1.45505816e-01 8.08209181e-02 -6.22596025e-01
7.37351894e-01 3.80223542e-01 -4.01871771e-01 9.31802809e-01
-4.15880270e-02 -2.61924326e-01 4.35865521e-01 8.39000940e-01
1.46675646e+00 -5.57155073e-01 -6.08220436e-02 -8.51143241e-01
-3.52676176e-02 1.45478576e-01 3.36084694e-01 1.00905788e+00
-3.41132015e-01 2.50397295e-01 9.14516866e-01 -7.46071525e-03
-5.72792888e-01 -1.03480458e+00 -3.05318683e-02 1.68240440e+00
3.27751577e-01 -1.17251384e+00 -9.32625771e-01 -7.14344859e-01
-5.75980172e-02 6.64826393e-01 -7.77471006e-01 -1.62133723e-01
-3.98706466e-01 -4.60273683e-01 2.73307204e-01 5.52840114e-01
2.93302625e-01 -6.37532413e-01 -3.53876323e-01 -1.15524553e-01
-4.70731288e-01 -9.25186515e-01 -1.06529140e+00 8.63247931e-01
-8.46281409e-01 -1.06519616e+00 3.77883874e-02 -1.09124649e+00
7.90777385e-01 4.72321898e-01 1.53609920e+00 4.58692312e-01
1.68804780e-01 4.90342289e-01 -7.24199593e-01 3.18494514e-02
-4.22587067e-01 6.06039286e-01 1.43587261e-01 -4.05623406e-01
9.57711458e-01 -4.80923355e-01 -5.81628680e-01 7.90629625e-01
-7.66996741e-01 -3.28209043e-01 4.97250140e-01 4.76813555e-01
3.63030672e-01 5.25915504e-01 3.21450800e-01 -1.19456387e+00
8.70476484e-01 -7.28310645e-01 -2.02472106e-01 1.66253954e-01
-7.22820282e-01 2.11158171e-01 8.68970215e-01 -5.68515599e-01
-3.71086359e-01 -1.21255554e-01 8.71810466e-02 -2.90762812e-01
-2.62885273e-01 5.69899142e-01 -9.20775384e-02 2.19883069e-01
7.67486334e-01 -4.87068482e-02 -1.51149794e-01 -3.80258232e-01
6.19471848e-01 7.31686413e-01 5.31369746e-01 -1.08270156e+00
8.71253490e-01 1.47052005e-01 -8.27516437e-01 -7.04781890e-01
-5.18739223e-01 -8.56276870e-01 -9.17286336e-01 -1.70260947e-03
9.96674716e-01 -8.23854625e-01 -1.17868888e+00 -8.45803916e-02
-5.06380498e-01 -7.70998836e-01 -1.20205306e-01 2.46975213e-01
-3.29626948e-01 5.42037606e-01 -1.01594424e+00 -2.74162471e-01
1.17107742e-01 -1.60940862e+00 9.56878543e-01 9.54854563e-02
-7.18839884e-01 -1.06974137e+00 -1.94812000e-01 5.23141265e-01
1.93077207e-01 -2.11751103e-01 1.42924571e+00 -1.09753656e+00
-7.42827415e-01 -3.95411327e-02 -9.55066271e-03 -1.31124377e-01
2.40732819e-01 2.17783973e-01 -6.50269389e-01 -5.65439880e-01
-1.01431534e-01 -6.95296049e-01 4.33252186e-01 -2.57540625e-02
1.33722472e+00 -4.45829481e-01 -5.79498649e-01 1.76092178e-01
1.23867798e+00 2.29566783e-01 1.49821043e-01 8.77105668e-02
8.33145142e-01 6.60647273e-01 4.68831331e-01 2.95719326e-01
8.60128224e-01 3.92074943e-01 3.62326264e-01 2.03701273e-01
3.10930222e-01 -3.50802600e-01 5.83690882e-01 1.80612147e+00
6.42059565e-01 -2.75024205e-01 -1.22758496e+00 6.90152109e-01
-1.84308827e+00 -4.38127726e-01 -3.71567130e-01 1.92709935e+00
1.19270253e+00 4.73135263e-01 2.25801110e-01 -2.37540100e-02
4.09812272e-01 -1.26793087e-01 -5.45501351e-01 -4.23353642e-01
2.47955725e-01 -1.04181878e-02 1.99273497e-01 6.07302904e-01
-7.70853579e-01 9.87510085e-01 6.21851063e+00 1.05595350e+00
-6.03989005e-01 3.34641263e-02 7.72587121e-01 5.93122356e-02
-5.78217864e-01 1.99432537e-01 -9.12730098e-01 6.23495460e-01
1.10704041e+00 -1.30825028e-01 4.01998878e-01 7.34250963e-01
7.42340907e-02 -2.34964103e-01 -1.73692679e+00 6.25552416e-01
2.76666805e-02 -1.19209969e+00 3.17039676e-02 2.46983841e-01
7.66963959e-01 7.71318302e-02 1.62469059e-01 6.12060189e-01
1.13034654e+00 -1.03674650e+00 4.90545303e-01 -1.08372763e-01
2.86523074e-01 -8.26737881e-01 5.74093282e-01 6.87870800e-01
-1.40957952e+00 -1.08130641e-01 -3.50202054e-01 6.92359284e-02
-4.75161612e-01 4.17037189e-01 -1.05506015e+00 3.29826772e-02
1.04520309e+00 6.14346504e-01 -1.13225162e+00 6.01727009e-01
6.67443946e-02 8.86629641e-01 -3.19182187e-01 -3.65981728e-01
4.87591267e-01 -2.76545137e-01 1.20365158e-01 1.28713500e+00
4.37407196e-02 2.48028681e-01 7.41784990e-01 8.89777243e-01
-2.89396942e-01 1.05032571e-01 -5.09398043e-01 -1.47704080e-01
1.10278594e+00 1.24760544e+00 -1.02280450e+00 -5.33891499e-01
-1.39534995e-01 6.51748538e-01 5.29873669e-01 2.92387277e-01
-5.02099395e-01 -6.75969899e-01 5.73870540e-01 2.24567160e-01
2.89071023e-01 -3.03217560e-01 -3.09397995e-01 -9.57209051e-01
-3.43088299e-01 -1.20194101e+00 5.35230577e-01 -6.98788762e-01
-1.39787877e+00 6.28475010e-01 -1.21122450e-01 -1.15176451e+00
-9.42134261e-02 -4.38126475e-01 -7.60037005e-01 2.44558603e-01
-6.65680110e-01 -4.93991733e-01 -1.74335182e-01 6.95882380e-01
5.96566677e-01 2.37329498e-01 6.75606608e-01 1.14985526e-01
-7.76925862e-01 8.60035658e-01 4.93909687e-01 3.13505650e-01
1.06866634e+00 -1.76772106e+00 1.75902158e-01 4.29403275e-01
2.48615220e-01 1.19838142e+00 7.68273294e-01 -5.13771176e-01
-1.46929348e+00 -1.02649260e+00 7.75810778e-01 -1.18280602e+00
1.01484013e+00 -7.64918804e-01 -1.02803195e+00 8.14917862e-01
6.23370349e-01 -4.37441140e-01 1.20509851e+00 6.37938380e-01
-4.83754963e-01 4.39715981e-02 -7.42011726e-01 7.00935602e-01
6.77284360e-01 -7.88029611e-01 -5.57362556e-01 5.44078052e-01
1.11337447e+00 -2.65885860e-01 -1.36998022e+00 -1.01652272e-01
-1.92648292e-01 -1.12835085e+00 5.94094396e-01 -4.70867038e-01
4.88866180e-01 -3.92954856e-01 -2.87072301e-01 -1.37383378e+00
-3.63737673e-01 -7.86477804e-01 5.60414605e-02 1.34296834e+00
5.83450556e-01 -3.85365754e-01 8.07180166e-01 5.82649112e-01
-3.68183970e-01 -8.38349581e-01 -5.08478642e-01 -5.71599185e-01
5.06537616e-01 -3.44091147e-01 6.06791198e-01 1.41340542e+00
6.52659833e-01 4.84809011e-01 8.84254724e-02 7.04271197e-02
5.15255988e-01 3.80044729e-01 8.47546160e-01 -1.36346185e+00
-2.52147317e-01 -5.50068319e-01 3.90854925e-02 -1.59287095e+00
-1.26070334e-02 -9.77337122e-01 4.33547139e-01 -1.23863578e+00
4.00860459e-01 -6.78390443e-01 -2.20587865e-01 5.85810483e-01
-2.99401015e-01 -1.30386725e-02 -1.07754655e-02 4.93623227e-01
-1.53635252e+00 1.84235975e-01 8.19425046e-01 -9.32322964e-02
-2.72866130e-01 -4.25306886e-01 -1.14760649e+00 7.67766058e-01
6.56158507e-01 -5.08561373e-01 -5.23992956e-01 -5.49574018e-01
2.42967114e-01 -7.81506300e-03 -3.01197022e-01 -8.90347064e-01
7.33601868e-01 -3.52433100e-02 1.59819499e-02 -6.73553228e-01
-7.24575073e-02 -6.62865222e-01 -4.01918083e-01 2.50645489e-01
-8.36633980e-01 4.16558474e-01 5.86687848e-02 6.83185518e-01
-2.24189609e-01 -6.24122238e-03 5.04103422e-01 -8.15930292e-02
-6.34731352e-01 1.40275627e-01 -8.01842332e-01 4.89259511e-01
7.12688267e-01 -1.43231526e-01 -6.20340109e-01 -4.59884167e-01
-4.50350285e-01 9.78584886e-01 7.83541262e-01 5.35757303e-01
3.10361952e-01 -1.22585893e+00 -2.65630573e-01 -2.81553827e-02
4.80494797e-01 1.90728024e-01 -1.64491937e-01 8.57886970e-01
-2.60383129e-01 3.04841399e-01 3.93846899e-01 -1.12175763e+00
-1.09989762e+00 6.36908174e-01 -1.62292607e-02 -2.77429461e-01
-4.75620955e-01 9.25365984e-01 1.71035811e-01 -8.31334651e-01
4.75176364e-01 -7.48795986e-01 5.45972064e-02 3.85347232e-02
3.11279178e-01 1.08840004e-01 1.38840944e-01 -3.18163246e-01
-3.59916657e-01 4.62101400e-01 -3.57406855e-01 1.75668493e-01
1.01300597e+00 -3.63916218e-01 -4.36443925e-01 9.28322077e-01
1.40840614e+00 2.69405067e-01 -1.02395010e+00 -3.34828943e-01
5.18553734e-01 -2.05302015e-01 -3.67127806e-01 -5.77043951e-01
-8.51178765e-01 8.85650158e-01 2.77640224e-01 7.22120106e-01
7.55198538e-01 4.61416900e-01 7.37398446e-01 7.73921072e-01
4.02308077e-01 -1.28008270e+00 8.00657392e-01 6.18246317e-01
2.49459967e-01 -1.30107796e+00 -1.75497890e-01 -2.73392022e-01
-6.18350327e-01 9.59593594e-01 1.10479105e+00 1.62429392e-01
7.12337911e-01 3.65850002e-01 3.70789230e-01 -6.23731077e-01
-1.20853293e+00 1.27860516e-01 1.29053250e-01 3.57410222e-01
7.19176769e-01 2.28087589e-01 1.42666250e-01 4.34419572e-01
-6.91776812e-01 -6.36919141e-01 5.13411403e-01 8.81140351e-01
-9.11601424e-01 -9.55862701e-01 -3.58464003e-01 7.68365145e-01
-2.02888120e-02 -1.97767332e-01 -7.36425996e-01 7.53187418e-01
-9.74344276e-03 1.52600682e+00 2.73454577e-01 -8.67005110e-01
-9.69834700e-02 4.39108573e-02 -1.47002414e-01 -9.30214226e-01
-8.68260205e-01 1.76471531e-01 -3.03018600e-01 -6.99501991e-01
4.07424420e-02 -4.17050928e-01 -1.50579798e+00 -6.04870379e-01
-3.67862880e-01 9.49210703e-01 6.12314790e-02 9.12165284e-01
5.77588856e-01 3.19291711e-01 8.07624102e-01 -3.30987245e-01
-4.81305569e-01 -8.89945269e-01 -5.26583374e-01 4.57053661e-01
1.36420906e-01 -3.21567625e-01 -5.74468076e-01 2.23343894e-02] | [10.684284210205078, 6.833302974700928] |
31d30994-868a-454c-92d5-5a5fb5dff117 | msrf-net-a-multi-scale-residual-fusion | 2105.07451 | null | https://arxiv.org/abs/2105.07451v2 | https://arxiv.org/pdf/2105.07451v2.pdf | MSRF-Net: A Multi-Scale Residual Fusion Network for Biomedical Image Segmentation | Methods based on convolutional neural networks have improved the performance of biomedical image segmentation. However, most of these methods cannot efficiently segment objects of variable sizes and train on small and biased datasets, which are common for biomedical use cases. While methods exist that incorporate multi-scale fusion approaches to address the challenges arising with variable sizes, they usually use complex models that are more suitable for general semantic segmentation problems. In this paper, we propose a novel architecture called Multi-Scale Residual Fusion Network (MSRF-Net), which is specially designed for medical image segmentation. The proposed MSRF-Net is able to exchange multi-scale features of varying receptive fields using a Dual-Scale Dense Fusion (DSDF) block. Our DSDF block can exchange information rigorously across two different resolution scales, and our MSRF sub-network uses multiple DSDF blocks in sequence to perform multi-scale fusion. This allows the preservation of resolution, improved information flow and propagation of both high- and low-level features to obtain accurate segmentation maps. The proposed MSRF-Net allows to capture object variabilities and provides improved results on different biomedical datasets. Extensive experiments on MSRF-Net demonstrate that the proposed method outperforms the cutting-edge medical image segmentation methods on four publicly available datasets. We achieve the dice coefficient of 0.9217, 0.9420, and 0.9224, 0.8824 on Kvasir-SEG, CVC-ClinicDB, 2018 Data Science Bowl dataset, and ISIC-2018 skin lesion segmentation challenge dataset respectively. We further conducted generalizability tests and achieved a dice coefficient of 0.7921 and 0.7575 on CVC-ClinicDB and Kvasir-SEG, respectively. | ['Pål Halvorsen', 'Sharib Ali', 'Michael A. Riegler', 'Dag Johansen', 'Håvard D. Johansen', 'Umapada Pal', 'Sukalpa Chanda', 'Debesh Jha', 'Abhishek Srivastava'] | 2021-05-16 | null | null | null | null | ['skin-lesion-segmentation'] | ['medical'] | [ 2.42174774e-01 -2.02778414e-01 -1.07696958e-01 -4.14084107e-01
-8.58824193e-01 -3.28667372e-01 6.27116486e-02 1.91896155e-01
-6.28845155e-01 7.48239517e-01 -5.70630990e-02 -8.04568753e-02
-2.23348662e-01 -7.95675039e-01 -5.28743207e-01 -7.97401071e-01
1.12357482e-01 1.55832618e-01 5.91956735e-01 -5.07229706e-03
6.24275655e-02 5.68041503e-01 -1.00147140e+00 4.89132643e-01
1.18004811e+00 1.24803758e+00 5.04343927e-01 6.08947277e-01
-7.44409114e-02 3.06892604e-01 -4.35882539e-01 -2.29265422e-01
1.39583975e-01 -1.78476200e-01 -8.73143733e-01 -2.59087890e-01
3.86449397e-01 -1.36580899e-01 -3.06916386e-01 1.30510509e+00
7.60143995e-01 6.97010255e-04 7.77901113e-01 -8.60947549e-01
-6.14547133e-01 4.78130966e-01 -9.86535609e-01 5.89335263e-01
-1.72446445e-01 1.17084123e-02 4.74215835e-01 -2.97772437e-01
7.17004001e-01 1.11878932e+00 9.25844908e-01 4.40948665e-01
-1.13859463e+00 -8.80196095e-01 -1.46730155e-01 2.30175480e-02
-1.37815559e+00 1.21396303e-01 5.31756580e-01 -2.55392045e-01
6.26463413e-01 4.11412269e-01 3.89682710e-01 8.41311395e-01
5.67361116e-01 8.86294723e-01 1.27850950e+00 -2.96631660e-02
-4.86731119e-02 -1.07062481e-01 3.09015095e-01 5.86498559e-01
3.99488926e-01 -2.83766150e-01 -5.88800758e-02 8.02292824e-02
1.21305430e+00 1.94488466e-01 -4.92702484e-01 1.42948970e-01
-1.25968862e+00 7.04639137e-01 1.00949502e+00 7.75240064e-01
-3.61956060e-01 -1.17536597e-01 5.19691348e-01 -1.11205615e-02
3.96137238e-01 2.16370985e-01 -3.23213279e-01 3.69811922e-01
-1.07635009e+00 5.08866608e-02 2.88457692e-01 6.67524636e-01
4.48754966e-01 -2.91705906e-01 -4.25127864e-01 9.93224323e-01
2.45951593e-01 4.41705406e-01 9.29550231e-01 -8.83651614e-01
2.32454419e-01 6.74458444e-01 -3.45468402e-01 -1.21894109e+00
-8.88019025e-01 -6.86722219e-01 -1.47660136e+00 7.68591976e-03
2.41549432e-01 -3.70318033e-02 -1.43171036e+00 1.55582774e+00
4.65145856e-01 3.13896567e-01 3.64963673e-02 1.08782923e+00
1.33734775e+00 4.78554219e-01 1.23100661e-01 -2.14292377e-01
1.47140038e+00 -9.70746160e-01 -7.96674609e-01 -4.50393325e-03
6.30108476e-01 -7.25258827e-01 6.68818414e-01 3.41016084e-01
-1.15446961e+00 -6.91985548e-01 -9.69277740e-01 -7.27712139e-02
-2.57611841e-01 1.23378031e-01 5.86979806e-01 5.14801264e-01
-1.10250938e+00 5.89719534e-01 -9.43645835e-01 -8.11036080e-02
8.84839058e-01 4.87022519e-01 -4.73311275e-01 -1.94635123e-01
-1.22800124e+00 6.36482894e-01 3.41057956e-01 3.31664741e-01
-6.48788273e-01 -8.98826957e-01 -6.74717128e-01 -1.99767321e-01
9.67880264e-02 -5.60528934e-01 7.94192672e-01 -7.87287295e-01
-9.90995288e-01 8.79038453e-01 1.62356615e-01 -2.79149562e-01
5.80356717e-01 2.65238583e-01 -4.25421774e-01 5.47136426e-01
2.71426022e-01 1.05511534e+00 4.02681857e-01 -1.13381112e+00
-6.25877261e-01 -7.41015077e-01 -1.34414151e-01 2.16748431e-01
-1.14847809e-01 -7.03572258e-02 -4.92819369e-01 -9.31300640e-01
2.50228494e-01 -7.73919046e-01 -4.14503306e-01 -3.09118698e-03
-4.59697306e-01 6.53695390e-02 8.17291200e-01 -8.51763785e-01
1.13629198e+00 -2.04877353e+00 4.90290187e-02 2.04238415e-01
3.41100574e-01 5.71575284e-01 -1.53730065e-01 -4.95099783e-01
6.20107241e-02 3.39958310e-01 -5.43286681e-01 -4.72352579e-02
-5.42681754e-01 4.56081592e-02 4.40804392e-01 5.18975019e-01
-3.10042147e-02 1.03847778e+00 -6.00635886e-01 -8.44706714e-01
3.07395995e-01 7.02255726e-01 -5.25977373e-01 4.35814708e-02
2.95438439e-01 5.63030422e-01 -6.38642490e-01 7.81459749e-01
1.05426073e+00 -3.88489962e-01 -1.54582530e-01 -7.04201221e-01
1.39232054e-01 -5.09798706e-01 -1.19982207e+00 2.00287747e+00
-3.19215536e-01 4.09567118e-01 4.34573233e-01 -1.14542961e+00
8.12009275e-01 3.03370595e-01 8.59757304e-01 -8.89935076e-01
4.08893257e-01 3.00098240e-01 5.01812622e-02 -4.37579751e-01
1.88145071e-01 -2.91402251e-01 7.20478892e-02 -1.26410527e-02
9.57039967e-02 -5.33723198e-02 2.46308073e-01 1.23603754e-01
9.31072056e-01 -3.07183743e-01 -2.44355481e-02 -3.80735964e-01
7.36805141e-01 -2.68488020e-01 7.86927521e-01 6.28866673e-01
-5.17112911e-01 1.00010490e+00 2.35691711e-01 -2.42500633e-01
-6.23113096e-01 -1.05675817e+00 -6.52909577e-01 4.81977850e-01
5.50023675e-01 -4.24389541e-02 -9.35813665e-01 -8.50317121e-01
-1.07404374e-01 1.93762362e-01 -8.03039134e-01 -1.44133434e-01
-5.33122838e-01 -1.07382596e+00 6.80652976e-01 6.18385553e-01
8.61937225e-01 -1.05356669e+00 -5.64550161e-01 2.13319272e-01
-4.44936246e-01 -1.15449345e+00 -4.81861055e-01 -1.06057869e-02
-9.59198356e-01 -1.00672984e+00 -1.24801207e+00 -8.29414606e-01
6.38590038e-01 1.88452587e-01 8.20576549e-01 1.32172883e-01
-7.19201267e-01 -9.98813659e-02 -2.86790997e-01 -1.99530914e-01
-1.52607232e-01 1.88257307e-01 -4.73836809e-01 -8.51850733e-02
-2.87818518e-02 -3.34172159e-01 -9.77091014e-01 5.43267131e-01
-1.30640101e+00 1.85059264e-01 7.17194676e-01 9.88770783e-01
8.73795569e-01 7.93628544e-02 7.54309535e-01 -8.46917331e-01
5.14399767e-01 -3.85736972e-01 -3.06618422e-01 4.19902623e-01
-5.52863955e-01 -3.01861644e-01 4.21977967e-01 -3.83647412e-01
-1.12509584e+00 -1.01977155e-01 -3.36348355e-01 -3.33468825e-01
-1.57626703e-01 3.15380663e-01 6.78699017e-02 -2.69440562e-01
5.39139509e-01 2.67791469e-02 9.19039026e-02 -2.93623120e-01
8.54412466e-02 8.39794755e-01 6.04473114e-01 -2.67233133e-01
9.18596163e-02 5.97447634e-01 6.65909722e-02 -4.62185144e-01
-6.64713502e-01 -4.74895149e-01 -5.19607246e-01 -1.33553922e-01
1.37551606e+00 -8.50264847e-01 -6.25757456e-01 8.10425162e-01
-9.09787595e-01 -1.28509417e-01 -6.51209056e-02 5.53801060e-01
-2.35465586e-01 3.28412175e-01 -8.93906116e-01 -1.69800639e-01
-7.26462126e-01 -1.71138120e+00 1.21186531e+00 8.02748561e-01
8.04061964e-02 -1.02046299e+00 -1.83257163e-01 6.29370332e-01
6.12012804e-01 5.70220530e-01 6.22170269e-01 -5.85278511e-01
-2.53812194e-01 -3.61725055e-02 -6.41259372e-01 4.59341943e-01
4.19032216e-01 -2.56455332e-01 -6.88805640e-01 -4.27002400e-01
-1.04995266e-01 -3.37814063e-01 1.08042157e+00 1.11279368e+00
1.59041929e+00 2.01400667e-01 -5.47214210e-01 9.25214767e-01
1.46767092e+00 3.21123511e-01 6.90340161e-01 2.10246772e-01
8.19958687e-01 3.66289914e-01 5.20029843e-01 -1.58891808e-02
3.55250806e-01 4.43753928e-01 3.05701286e-01 -6.86364710e-01
-3.98214310e-01 3.83100241e-01 -2.73752302e-01 6.46860659e-01
-3.94518226e-02 -9.37930197e-02 -8.29984725e-01 6.65434420e-01
-1.68551779e+00 -5.99083781e-01 -2.38087133e-01 1.66803849e+00
1.03676331e+00 1.61661673e-02 -9.61007252e-02 5.23741764e-04
8.88110220e-01 2.75241006e-02 -6.53258383e-01 -1.58763334e-01
-1.60274446e-01 4.87092555e-01 6.67773604e-01 2.99003750e-01
-1.22241783e+00 6.52130067e-01 5.45962811e+00 1.22340322e+00
-1.25942338e+00 3.23498219e-01 1.07126319e+00 2.16308385e-02
-1.80209726e-01 -6.01224780e-01 -6.53233826e-01 6.08364105e-01
6.13866746e-01 1.74068823e-01 2.63242442e-02 3.36633891e-01
-1.84315331e-02 -2.16204032e-01 -4.55038965e-01 1.09198904e+00
-7.02841803e-02 -1.42768097e+00 -3.92311662e-02 -4.02871668e-02
9.31037366e-01 1.07087046e-01 1.81221187e-01 2.47869976e-02
6.39436543e-02 -1.35649061e+00 1.56520173e-01 5.15767336e-01
9.32883084e-01 -7.59784579e-01 1.05953896e+00 3.58617231e-02
-1.11430156e+00 8.65812004e-02 -3.77654582e-01 7.05412567e-01
1.77174002e-01 8.06000650e-01 -4.18674320e-01 8.60310197e-01
9.13159668e-01 7.10198283e-01 -5.71319282e-01 1.06046534e+00
2.67908841e-01 3.97688419e-01 -2.94228047e-01 2.39626974e-01
3.49018872e-01 -4.90134694e-02 2.29408771e-01 1.40918887e+00
3.52444023e-01 1.29543036e-01 -2.72362586e-02 7.09085524e-01
-1.30066261e-01 1.76055521e-01 4.86531435e-03 2.42360592e-01
9.69945937e-02 1.68025661e+00 -1.17751122e+00 -3.77473623e-01
-2.66694337e-01 7.42653251e-01 -4.46181595e-02 2.64345527e-01
-1.05481207e+00 -4.28057611e-01 3.10335666e-01 -2.67671458e-02
2.70801723e-01 2.21571460e-01 -4.62647706e-01 -1.02875626e+00
-1.96294487e-01 -9.18312073e-01 6.08309150e-01 -6.10398293e-01
-1.29075289e+00 8.24212790e-01 -6.93166777e-02 -7.96471775e-01
4.03109848e-01 -4.97435898e-01 -3.85843694e-01 8.74029934e-01
-1.60634303e+00 -1.27027571e+00 -5.73291063e-01 7.15314388e-01
4.27529752e-01 -6.59068208e-03 4.38228041e-01 5.91017306e-01
-5.96240819e-01 6.70353770e-01 1.34882480e-01 1.27441496e-01
7.83543468e-01 -1.15047050e+00 -7.36552924e-02 5.34296453e-01
-3.19480777e-01 4.86578286e-01 1.82535917e-01 -7.43919373e-01
-8.68902624e-01 -1.19597101e+00 7.14276731e-02 1.38382971e-01
7.47415796e-02 1.74369827e-01 -1.15722656e+00 3.09712172e-01
1.17959164e-01 5.23318827e-01 6.12012208e-01 -3.93539906e-01
7.50565976e-02 -3.08257252e-01 -1.74444199e+00 9.26867127e-02
7.81874180e-01 -5.45345061e-02 -2.94580728e-01 1.74730152e-01
7.35008061e-01 -6.88206971e-01 -1.39306307e+00 9.52487171e-01
6.27696395e-01 -8.62460256e-01 1.11446083e+00 -2.14736998e-01
3.68596196e-01 -2.74893850e-01 -1.35067686e-01 -1.20370686e+00
-4.36840624e-01 -4.81869616e-02 3.13436240e-01 1.08294594e+00
3.19010139e-01 -6.11215472e-01 6.01867259e-01 2.45400593e-01
-3.31412107e-01 -1.04052103e+00 -1.10536790e+00 -3.83405805e-01
3.69640261e-01 -1.60707638e-01 4.97207940e-01 9.52663302e-01
-5.44018686e-01 -1.29927620e-01 4.37390693e-02 -5.91776073e-02
7.28808522e-01 -1.23488888e-01 1.49281755e-01 -1.06937981e+00
1.27052357e-02 -6.29488647e-01 -4.12440509e-01 -7.02422500e-01
-6.61029741e-02 -1.17856443e+00 -1.85726881e-01 -1.93211353e+00
4.27789778e-01 -6.05703115e-01 -7.82517552e-01 4.11998838e-01
-3.56364489e-01 6.62593186e-01 -4.86626849e-02 8.96489471e-02
-4.50866342e-01 1.50090501e-01 1.86195636e+00 -2.84977376e-01
-9.20131207e-02 -1.52518436e-01 -8.14121008e-01 6.84970975e-01
8.38877678e-01 -3.05907160e-01 -3.98705989e-01 -4.02045906e-01
-2.96269774e-01 9.61141288e-02 3.53120029e-01 -1.06209362e+00
2.21731976e-01 1.45881949e-02 9.20497835e-01 -6.61335647e-01
-3.54058631e-02 -6.45191252e-01 2.06398204e-01 5.81550896e-01
-2.90603459e-01 -1.66046619e-01 3.12624991e-01 3.17964166e-01
-3.20292294e-01 -3.22208144e-02 1.20111859e+00 -2.34431013e-01
-3.61839592e-01 5.12936532e-01 -1.52108623e-02 -3.86709459e-02
1.20638049e+00 -2.92871356e-01 -3.17807674e-01 2.00671270e-01
-7.85016358e-01 3.30615938e-01 1.71487108e-01 4.49191958e-01
6.95451021e-01 -1.35245323e+00 -6.95012271e-01 8.44519660e-02
-2.52837926e-01 5.25290489e-01 7.58947492e-01 1.26197982e+00
-7.09248900e-01 3.32418442e-01 -4.84230369e-01 -8.89336169e-01
-1.24668193e+00 1.93389744e-01 5.48761964e-01 -5.96069753e-01
-5.85612059e-01 9.83430266e-01 5.25772095e-01 -4.60522413e-01
-3.13922502e-02 -5.99888742e-01 -2.54850090e-01 2.04759967e-02
5.27061403e-01 4.08215761e-01 3.27662587e-01 -8.33032489e-01
-5.91383815e-01 8.24915946e-01 -3.44759315e-01 2.55191475e-01
1.28309429e+00 -1.56063154e-01 -1.83540940e-01 -7.28174970e-02
1.43759322e+00 -4.91328210e-01 -1.09263265e+00 -1.81231812e-01
-5.16818702e-01 -2.52594531e-01 5.24076462e-01 -1.02899992e+00
-1.67157698e+00 8.32995892e-01 1.19475186e+00 -3.22968997e-02
1.49299884e+00 -7.90785849e-02 1.25773060e+00 -4.04044002e-01
7.91494623e-02 -1.06711161e+00 -2.48953644e-02 3.22630778e-02
7.15132952e-01 -1.24876833e+00 -3.98297943e-02 -3.92573148e-01
-6.96981192e-01 1.01545393e+00 7.29249656e-01 -1.07454937e-02
5.92492878e-01 4.70570654e-01 4.92338054e-02 -1.76746100e-01
-2.34163210e-01 8.83437991e-02 5.56513429e-01 5.31986892e-01
4.35615778e-01 1.72313794e-01 -5.34687579e-01 7.00465679e-01
1.31076887e-01 1.86922699e-01 2.72925138e-01 7.35946774e-01
-3.58847320e-01 -7.47791827e-01 -3.52380514e-01 6.81675196e-01
-8.96946788e-01 1.43962637e-01 9.98280272e-02 7.22933590e-01
4.18966472e-01 8.44008744e-01 8.56966898e-02 -1.92872882e-01
1.50694147e-01 -5.83512247e-01 5.41529596e-01 -1.74179867e-01
-7.69535601e-01 3.74206096e-01 -4.47930664e-01 -6.12735331e-01
-5.56517601e-01 -4.35320050e-01 -1.64509249e+00 1.20878490e-02
-4.35755521e-01 3.17545496e-02 6.95773125e-01 7.39618421e-01
2.62263954e-01 9.75588918e-01 3.83585453e-01 -5.34122527e-01
-1.86139256e-01 -9.44165409e-01 -6.36019409e-01 5.16094029e-01
2.91664243e-01 -6.46134257e-01 4.78148795e-05 -9.73279104e-02] | [14.698726654052734, -2.6127583980560303] |
f6fdebbc-fea9-4559-8aa6-d2ee0f8a1ab8 | conversational-exploratory-search-via | 1709.05298 | null | https://arxiv.org/abs/1709.05298v1 | https://arxiv.org/pdf/1709.05298v1.pdf | Conversational Exploratory Search via Interactive Storytelling | Conversational interfaces are likely to become more efficient, intuitive and engaging way for human-computer interaction than today's text or touch-based interfaces. Current research efforts concerning conversational interfaces focus primarily on question answering functionality, thereby neglecting support for search activities beyond targeted information lookup. Users engage in exploratory search when they are unfamiliar with the domain of their goal, unsure about the ways to achieve their goals, or unsure about their goals in the first place. Exploratory search is often supported by approaches from information visualization. However, such approaches cannot be directly translated to the setting of conversational search. In this paper we investigate the affordances of interactive storytelling as a tool to enable exploratory search within the framework of a conversational interface. Interactive storytelling provides a way to navigate a document collection in the pace and order a user prefers. In our vision, interactive storytelling is to be coupled with a dialogue-based system that provides verbal explanations and responsive design. We discuss challenges and sketch the research agenda required to put this vision into life. | ['Maarten de Rijke', 'Ilya Markov', 'Svitlana Vakulenko'] | 2017-09-15 | null | null | null | null | ['conversational-search'] | ['natural-language-processing'] | [ 1.28793851e-01 4.26383138e-01 -6.27169199e-03 -2.65818506e-01
-1.21995427e-01 -9.18242693e-01 9.77459908e-01 5.30738175e-01
-2.94152796e-01 4.10231620e-01 6.85613453e-01 -8.62624407e-01
-3.81656885e-01 -5.46308517e-01 3.07596087e-01 4.35879640e-02
2.06155166e-01 4.06231523e-01 1.14638291e-01 -5.97999990e-01
6.94948971e-01 2.74385542e-01 -1.57414079e+00 3.48620713e-01
6.68770194e-01 3.56588215e-01 6.60971105e-01 8.33535731e-01
-9.20220613e-01 7.82285452e-01 -6.80301726e-01 1.35732323e-01
-3.59383196e-01 -4.58689481e-01 -1.21788490e+00 -1.08160347e-01
-2.66661140e-04 -5.49675703e-01 1.25983849e-01 5.74828744e-01
3.61081570e-01 3.39209259e-01 1.77350804e-01 -1.08318865e+00
-3.04618418e-01 4.58781302e-01 1.53743373e-02 1.47681773e-01
1.35726583e+00 1.45137310e-01 7.68850505e-01 -5.83366990e-01
7.57665098e-01 1.23056424e+00 2.32223257e-01 5.54041862e-01
-1.25240290e+00 -2.06592917e-01 3.74689043e-01 -2.98831277e-02
-7.96936035e-01 -4.40288752e-01 6.49015844e-01 -6.42035484e-01
9.35647249e-01 9.09784138e-01 8.72120082e-01 7.47037649e-01
-2.83711374e-01 7.44962037e-01 7.76839733e-01 -7.41483510e-01
1.65753290e-01 8.99403274e-01 5.47026634e-01 3.37230355e-01
-8.98065642e-02 -2.52826899e-01 -9.11935449e-01 -3.62521946e-01
7.05567539e-01 6.61785901e-02 -4.62950170e-01 -3.90784919e-01
-1.22436464e+00 5.44566333e-01 9.45998877e-02 8.15578878e-01
-4.18559521e-01 -1.23121433e-01 3.97198409e-01 5.14590442e-01
2.10918352e-01 9.95322108e-01 -1.65689141e-02 -9.74134207e-01
-4.92368013e-01 4.65915293e-01 1.33687675e+00 8.55225265e-01
2.53852606e-01 -5.29024839e-01 -5.82861304e-01 6.01782441e-01
4.62789387e-01 -2.11293940e-02 1.35301471e-01 -9.29750979e-01
5.03210127e-02 1.00751257e+00 6.28315985e-01 -9.07746196e-01
-3.47389430e-01 -1.56986602e-02 5.02321906e-02 4.26379055e-01
5.66060066e-01 -2.90919803e-02 -2.20172424e-02 1.25521612e+00
5.46460211e-01 -9.50713992e-01 -1.78371787e-01 8.92159879e-01
1.04103744e+00 4.20360595e-01 1.60538286e-01 -2.18146130e-01
1.56115508e+00 -5.23506820e-01 -1.12992358e+00 -6.30644560e-02
9.00873184e-01 -8.26745570e-01 1.77802360e+00 3.16625088e-01
-1.03602076e+00 -1.65837899e-01 -8.55365932e-01 -2.62252986e-01
-5.67513049e-01 -2.48445541e-01 5.81744492e-01 6.86441839e-01
-8.62217903e-01 4.33422983e-01 -6.83434904e-01 -8.87952387e-01
-2.00788155e-01 -5.03751300e-02 1.16515737e-02 2.79549003e-01
-8.56646955e-01 1.00785947e+00 1.70740306e-01 -1.44548953e-01
3.22434902e-01 -5.83214998e-01 -4.33621198e-01 2.05454856e-01
3.24115694e-01 -7.21845388e-01 1.64765179e+00 -4.11547452e-01
-1.36911654e+00 5.56424856e-01 -6.70743361e-02 7.25152791e-02
6.15810335e-01 -5.46887517e-01 -1.63693622e-01 -3.79454680e-02
1.53371751e-01 3.90296936e-01 5.49849421e-02 -1.21428013e+00
-4.39241141e-01 -3.58161002e-01 3.30128670e-01 5.62710583e-01
-2.79907733e-01 4.56377327e-01 -2.46927708e-01 -1.27095312e-01
7.33773708e-02 -4.31515127e-01 -3.49926874e-02 4.05752450e-01
-4.89792228e-01 -2.88955271e-01 1.33513725e+00 -4.40090656e-01
1.77518642e+00 -1.81223631e+00 -2.30630696e-01 4.05472815e-02
3.40846628e-01 3.91494066e-01 5.86082458e-01 1.49109566e+00
2.67456234e-01 2.65628397e-01 2.46687382e-01 -2.03810230e-01
5.02167761e-01 -2.51735479e-01 -1.11842290e-01 -2.66317129e-01
-4.86004442e-01 5.92663825e-01 -1.04795897e+00 -5.20220160e-01
4.30098504e-01 5.02578259e-01 -3.15705627e-01 4.14172202e-01
-2.37204567e-01 4.91601825e-01 -6.14835799e-01 4.15220886e-01
1.59978956e-01 -3.65967304e-01 2.89821684e-01 4.00323093e-01
-7.16573000e-01 7.52597690e-01 -1.00191295e+00 1.51911664e+00
-6.53831542e-01 9.11445856e-01 3.97888958e-01 -1.74996391e-01
7.53766477e-01 2.71856070e-01 -8.18447955e-03 -7.85029471e-01
-1.34195879e-01 -6.52522221e-02 -2.50146896e-01 -9.71549928e-01
5.96843660e-01 3.01236629e-01 6.79530427e-02 1.11920106e+00
-6.99901998e-01 -1.91189885e-01 1.17250204e-01 4.83863682e-01
1.05839837e+00 2.72961378e-01 4.19046193e-01 -2.70325035e-01
1.12766728e-01 9.61330980e-02 -3.90633762e-01 9.39008117e-01
2.06630528e-01 1.50139749e-01 3.44186395e-01 -5.39491653e-01
-4.72932130e-01 -6.87469006e-01 1.51727736e-01 1.11500001e+00
1.36503652e-01 -8.34333420e-01 -6.62658632e-01 -2.99471527e-01
-3.26917171e-01 1.16069043e+00 -2.24512443e-01 3.17636639e-01
-2.51862079e-01 1.53663829e-01 -1.21174389e-02 -4.68273982e-02
5.80538273e-01 -1.20136380e+00 -1.62395573e+00 2.79254436e-01
-2.17969641e-01 -5.16233981e-01 -4.54365909e-01 -2.20093161e-01
-7.39650846e-01 -9.98842657e-01 -4.89916801e-01 -5.79692602e-01
4.33281690e-01 4.08225298e-01 9.99545217e-01 3.11613858e-01
-1.23469599e-01 8.17015767e-01 -5.35126686e-01 -3.45669687e-01
-3.86172146e-01 -9.28056464e-02 -5.59042573e-01 -4.86100227e-01
3.16625446e-01 -7.42368042e-01 -9.50084150e-01 3.99160713e-01
-6.63413763e-01 7.65291989e-01 -8.22520107e-02 5.05991697e-01
-2.91182518e-01 -3.94971877e-01 2.79977858e-01 -9.91835356e-01
1.54738784e+00 -4.24984813e-01 9.07777995e-02 2.93348879e-01
-7.31592059e-01 -2.91139353e-02 1.08701371e-01 -4.18366790e-01
-1.20784056e+00 -2.76024848e-01 -3.77839804e-02 2.74207741e-01
-3.69536400e-01 5.00830829e-01 9.04137865e-02 -1.51704196e-02
7.69093454e-01 1.18302554e-01 1.88832998e-01 -6.04160547e-01
4.28287715e-01 8.35754991e-01 3.17428172e-01 -5.64430237e-01
2.83761501e-01 1.84980229e-01 -5.86677969e-01 -8.71974766e-01
7.00434744e-02 -5.99700212e-01 -2.97474384e-01 -7.37343907e-01
5.13728082e-01 2.26190966e-02 -1.33994472e+00 -2.21027732e-01
-1.16264415e+00 -5.28774738e-01 -2.59754270e-01 1.94398612e-01
-5.28666139e-01 3.15625399e-01 -1.13771565e-01 -1.49751592e+00
-3.58975291e-01 -6.50336802e-01 9.12090719e-01 4.28337276e-01
-1.20602584e+00 -9.89380240e-01 3.40452418e-02 4.12599862e-01
8.90553057e-01 4.15978968e-01 1.04439461e+00 -6.84618056e-01
-4.60443169e-01 -4.23633426e-01 -1.54987589e-01 -8.07450891e-01
3.32539111e-01 -2.20295608e-01 -7.56826103e-01 1.23763502e-01
-5.15563749e-02 6.90798229e-03 -1.34478897e-01 -2.60187179e-01
7.10031509e-01 -7.72177160e-01 -6.01124406e-01 -1.22302435e-01
1.05262709e+00 7.20066667e-01 3.86566162e-01 5.88608623e-01
2.49846056e-01 1.04842365e+00 7.50490010e-01 5.19735396e-01
4.27936882e-01 1.00377548e+00 1.44250885e-01 1.57864299e-02
1.51229486e-01 -4.94695574e-01 -2.82145023e-01 7.11044222e-02
2.66327769e-01 -3.91338944e-01 -1.22755706e+00 3.73858005e-01
-1.97539675e+00 -8.09044182e-01 -2.55715996e-01 2.17846060e+00
4.95365560e-01 1.53294140e-02 1.29933566e-01 1.84771314e-01
4.42898840e-01 2.14439910e-02 -3.05609524e-01 -7.54386127e-01
8.02699566e-01 -3.57059449e-01 -3.59070599e-01 6.93098247e-01
-3.13458055e-01 5.86642921e-01 6.11235952e+00 1.20819286e-01
-9.32935476e-01 -1.03652596e-01 3.38331699e-01 -1.09014273e-01
-6.45988166e-01 5.56738190e-02 -2.32383981e-01 2.89190143e-01
4.63973284e-01 -7.20381811e-02 4.56425816e-01 6.22241795e-01
7.34940648e-01 -6.24222994e-01 -1.39009821e+00 9.87214327e-01
-2.91005582e-01 -1.33687294e+00 -3.04691076e-01 -3.18451524e-02
-2.01991692e-01 -8.14395487e-01 -1.27042636e-01 9.76589769e-02
6.92740008e-02 -8.41899693e-01 5.63653529e-01 6.85473561e-01
5.31085491e-01 -4.28216428e-01 1.85683712e-01 7.85626948e-01
-7.66800046e-01 6.63823262e-02 6.68747544e-01 -7.36685514e-01
5.02840102e-01 -1.89259857e-01 -1.12398922e+00 -2.34443918e-01
8.04949462e-01 -2.08468631e-01 -2.56901741e-01 1.09505975e+00
3.31859171e-01 1.02852084e-01 -4.18803841e-01 -7.16198385e-01
-1.04217634e-01 -1.54903233e-01 7.96739876e-01 1.27537918e+00
2.86425054e-01 4.63579983e-01 2.38037780e-02 8.89209509e-01
7.51077175e-01 3.52502316e-01 -9.41551805e-01 -4.58030999e-01
6.92268014e-01 1.00484002e+00 -9.16404843e-01 -3.49377543e-02
-1.60236940e-01 8.22306752e-01 -1.24629900e-01 3.39927673e-01
-9.41592231e-02 -6.05550587e-01 5.70508003e-01 6.86238885e-01
-3.98464531e-01 -4.18464988e-01 -4.63213414e-01 -6.21998668e-01
1.29749402e-01 -8.91389489e-01 5.38530536e-02 -1.16038716e+00
-5.56478024e-01 5.75372458e-01 2.61794209e-01 -7.46408463e-01
-7.05985546e-01 6.66441768e-03 -1.03149867e+00 9.86352086e-01
-4.91561323e-01 -9.11437988e-01 -5.71602523e-01 3.67656387e-02
7.71663964e-01 4.32552457e-01 1.17613018e+00 -8.75734016e-02
3.50933261e-02 -1.02999017e-01 -2.56831914e-01 -4.72769111e-01
4.62408841e-01 -1.21603310e+00 4.39540207e-01 2.92263865e-01
-3.70818973e-02 1.02095115e+00 1.26519239e+00 -4.04685408e-01
-1.55770051e+00 1.87684998e-01 1.23756111e+00 -5.16899049e-01
3.91001761e-01 -5.33086360e-01 -9.64986444e-01 1.77392125e-01
6.77370846e-01 -8.63916934e-01 8.53048265e-01 4.88726467e-01
1.90236807e-01 4.25860107e-01 -1.16614497e+00 1.01196682e+00
1.07188118e+00 -7.66768575e-01 -6.71823382e-01 2.79795319e-01
6.65251732e-01 -3.80030274e-01 -3.19050580e-01 -2.38318935e-01
8.18930745e-01 -1.18542266e+00 7.95139849e-01 -4.32151675e-01
-1.40455842e-01 -1.94001958e-01 1.28477469e-01 -1.04258871e+00
6.85387403e-02 -1.30484712e+00 1.86760381e-01 1.48815787e+00
3.47701192e-01 -6.14654541e-01 5.98763525e-01 1.39890409e+00
-7.95283727e-03 -5.99813461e-01 -4.73756313e-01 2.58645834e-03
-5.43272913e-01 -6.12916052e-01 6.73716903e-01 8.31061244e-01
1.20210910e+00 2.91180670e-01 -2.63795592e-02 -2.51112282e-01
-9.82263964e-03 3.64169590e-02 8.36027920e-01 -1.47214639e+00
-1.62603006e-01 -6.53989553e-01 8.56731981e-02 -1.22631264e+00
-6.70039952e-01 -2.71209151e-01 -3.65116112e-02 -2.01411605e+00
-4.50803377e-02 -2.69020438e-01 5.66080689e-01 1.33580446e-01
2.66435385e-01 -4.65845913e-01 3.84781599e-01 2.75271326e-01
-4.53200817e-01 4.49948460e-02 1.21686423e+00 2.63177246e-01
-9.03704941e-01 1.72051117e-01 -1.15211487e+00 4.79110241e-01
7.20248401e-01 9.04449169e-03 -8.78738821e-01 3.44056636e-02
7.51539469e-01 6.62063122e-01 2.62022495e-01 -6.74440920e-01
7.11010218e-01 -2.94608504e-01 -4.48978879e-02 -5.37753105e-01
3.30509871e-01 -8.83670270e-01 3.81341070e-01 1.80831015e-01
-7.66242325e-01 2.13055551e-01 2.25480914e-01 3.50347012e-01
1.06648700e-02 -1.52838543e-01 -1.50549591e-01 -3.27983387e-02
-2.81278998e-01 -4.57699299e-01 -1.11245537e+00 -2.07957402e-01
6.78287566e-01 -8.51085067e-01 -3.11479092e-01 -9.43881154e-01
-1.07072103e+00 5.18874705e-01 5.07137895e-01 5.19463539e-01
5.38307488e-01 -8.69397163e-01 7.58197531e-02 1.08726948e-01
9.19040442e-02 -6.25212938e-02 -4.16488275e-02 4.06514466e-01
-5.78156829e-01 8.25109422e-01 -8.73310268e-02 -3.49222422e-01
-1.58866179e+00 2.27781147e-01 1.79130122e-01 9.74217504e-02
-7.56748676e-01 5.80677211e-01 1.31820977e-01 -2.73802757e-01
6.11739397e-01 -3.00993025e-01 -5.90853214e-01 1.48881018e-01
8.85741770e-01 4.26357031e-01 -1.71921447e-01 7.09271133e-02
-1.35904983e-01 9.08171237e-02 5.46406619e-02 -6.51131749e-01
9.20803964e-01 -6.37326837e-01 1.09451294e-01 7.96403408e-01
5.60760140e-01 -7.90714175e-02 -8.73118341e-01 1.56491697e-02
3.32159519e-01 -7.34441817e-01 -1.41430005e-01 -1.26227808e+00
1.81792840e-01 7.38452852e-01 4.87644196e-01 1.10985756e+00
6.96960866e-01 1.64276958e-01 2.20046520e-01 7.17505574e-01
2.91760683e-01 -9.58323121e-01 7.15572909e-02 1.67105570e-01
1.22595727e+00 -1.05244160e+00 -3.41039509e-01 -4.39928502e-01
-5.73814452e-01 1.37394190e+00 6.16547465e-01 7.21198022e-01
5.43246031e-01 2.12725922e-01 1.41236961e-01 -7.20259905e-01
-1.00173664e+00 -1.27905399e-01 -1.72762778e-02 7.21787214e-01
9.35996056e-01 -1.39046401e-01 -5.56163609e-01 2.99891472e-01
-3.39986026e-01 4.19361331e-02 1.93126589e-01 1.16077948e+00
-6.95188701e-01 -1.03024316e+00 -3.98497194e-01 3.15415353e-01
-2.21287519e-01 2.91488878e-02 -9.07908797e-01 8.01252782e-01
-4.73632604e-01 1.39841557e+00 4.67887633e-02 -1.07239075e-01
3.90201449e-01 3.45814586e-01 9.11949724e-02 -7.55954087e-01
-9.14199531e-01 -1.69160292e-01 7.21526265e-01 -5.21893919e-01
9.37460810e-02 -6.67148888e-01 -1.14283824e+00 -3.53624582e-01
-2.41840780e-01 6.37615621e-01 9.88238871e-01 7.88191199e-01
4.68922317e-01 1.78541914e-01 -1.44504998e-02 -6.73228741e-01
-1.04607269e-01 -8.80063117e-01 -1.77988745e-02 -7.05760270e-02
3.71434301e-01 -3.15505981e-01 -1.66467708e-02 -3.65069032e-01] | [12.271883010864258, 7.815826416015625] |
43ce6a09-269e-48d3-a97e-b854c3fb8669 | deep-tree-ensembles-for-multi-output | 2011.02829 | null | https://arxiv.org/abs/2011.02829v2 | https://arxiv.org/pdf/2011.02829v2.pdf | Deep tree-ensembles for multi-output prediction | Recently, deep neural networks have expanded the state-of-art in various scientific fields and provided solutions to long standing problems across multiple application domains. Nevertheless, they also suffer from weaknesses since their optimal performance depends on massive amounts of training data and the tuning of an extended number of parameters. As a countermeasure, some deep-forest methods have been recently proposed, as efficient and low-scale solutions. Despite that, these approaches simply employ label classification probabilities as induced features and primarily focus on traditional classification and regression tasks, leaving multi-output prediction under-explored. Moreover, recent work has demonstrated that tree-embeddings are highly representative, especially in structured output prediction. In this direction, we propose a novel deep tree-ensemble (DTE) model, where every layer enriches the original feature set with a representation learning component based on tree-embeddings. In this paper, we specifically focus on two structured output prediction tasks, namely multi-label classification and multi-target regression. We conducted experiments using multiple benchmark datasets and the obtained results confirm that our method provides superior results to state-of-the-art methods in both tasks. | ['Celine Vens', 'Konstantinos Pliakos', 'Felipe Kenji Nakano'] | 2020-11-03 | null | null | null | null | ['multi-target-regression'] | ['miscellaneous'] | [ 6.19172096e-01 -2.61553109e-01 -5.69041431e-01 -5.47949553e-01
-5.43367326e-01 -4.63458858e-02 5.49909651e-01 2.32963115e-01
-3.73779088e-01 8.27168703e-01 -1.80855189e-02 -1.34418353e-01
-2.13077277e-01 -9.28842962e-01 -5.06087840e-01 -9.40957308e-01
2.79553890e-01 3.95328969e-01 1.26411185e-01 2.20521599e-01
1.75960168e-01 2.54640311e-01 -1.72669399e+00 2.23926082e-01
1.12065017e+00 1.34050357e+00 5.19696400e-02 2.33360276e-01
-1.98818848e-01 7.81151533e-01 -3.82980973e-01 -4.96321231e-01
-5.30676953e-02 -1.14292666e-01 -4.83387560e-01 -2.20597491e-01
2.58550406e-01 2.12656669e-02 1.70592628e-02 8.33986461e-01
5.59759617e-01 3.10263131e-02 9.65810657e-01 -1.31406069e+00
-8.11464608e-01 4.72854286e-01 -8.56130540e-01 -1.82644889e-01
-1.26571715e-01 -1.70509502e-01 1.28311455e+00 -1.01056445e+00
2.58923411e-01 1.00233865e+00 9.37479198e-01 3.97354931e-01
-1.45457482e+00 -9.98205543e-01 1.91319734e-01 4.56834614e-01
-1.16141903e+00 -1.02926977e-01 7.49030948e-01 -5.06268620e-01
9.29253101e-01 -4.87146787e-02 3.44036967e-01 1.38935697e+00
3.48455757e-01 8.00522029e-01 1.27924037e+00 -5.97624004e-01
-2.41868906e-02 3.18214864e-01 2.79284298e-01 7.51302242e-01
2.88444072e-01 1.21053346e-01 -4.22983080e-01 -1.01809978e-01
3.18806708e-01 3.76496553e-01 -2.19733834e-01 -5.73374212e-01
-1.10181630e+00 1.11977184e+00 3.32420677e-01 4.82399017e-01
-3.04696292e-01 2.61610188e-02 5.08962452e-01 8.72164741e-02
7.93552876e-01 2.03419685e-01 -6.48074090e-01 1.08111165e-01
-7.72304654e-01 1.44229770e-01 5.69345653e-01 5.58072925e-01
7.21413612e-01 -1.23268194e-01 -2.55922198e-01 1.09012926e+00
1.46509841e-01 1.21695697e-01 7.73893237e-01 -4.95894372e-01
5.88556886e-01 9.18661833e-01 -1.97359666e-01 -1.09560156e+00
-5.98154545e-01 -1.02387595e+00 -1.34078991e+00 1.19335920e-01
2.15613723e-01 1.75528675e-02 -8.63085449e-01 1.67467129e+00
3.57053190e-01 2.28629068e-01 -2.59348117e-02 4.70207095e-01
6.71944082e-01 6.28569305e-01 5.48746824e-01 -1.99031875e-01
1.36561978e+00 -1.47655988e+00 -7.44601190e-01 -1.90728098e-01
9.13520634e-01 -4.90297616e-01 1.03062284e+00 6.63534999e-01
-3.43004823e-01 -8.00609529e-01 -1.01319516e+00 3.42924185e-02
-5.42637587e-01 6.05845511e-01 6.92046106e-01 6.69031084e-01
-6.46529615e-01 8.19305301e-01 -4.28593874e-01 -3.21524769e-01
6.19218767e-01 3.26398253e-01 -4.04680014e-01 -2.24802122e-01
-1.14550412e+00 1.02422917e+00 4.96279895e-01 6.07192069e-02
-7.57229745e-01 -4.65973198e-01 -6.26762569e-01 2.47730806e-01
3.39906633e-01 -6.46889091e-01 1.19395638e+00 -6.93155229e-01
-1.38718069e+00 6.99629784e-01 -8.62163976e-02 -4.16686475e-01
2.08747491e-01 -4.56046939e-01 -5.07288098e-01 -1.64586917e-01
6.38808236e-02 4.97567922e-01 8.50330412e-01 -1.24737096e+00
-8.89179349e-01 -6.66378200e-01 -3.13123941e-01 1.74305812e-02
-1.01616871e+00 -8.27319026e-02 1.47998706e-01 -7.86185682e-01
-9.22637954e-02 -8.17798138e-01 -2.05004096e-01 -1.94183722e-01
-3.49095523e-01 -6.80855870e-01 8.84174585e-01 -4.29090053e-01
1.64910698e+00 -1.90999854e+00 2.57261485e-01 -2.75077373e-01
3.89779001e-01 4.11005765e-01 -1.00090243e-01 4.22154874e-01
-1.45062700e-01 2.07452610e-01 -3.35295022e-01 -4.79852498e-01
-5.20184338e-02 8.56291428e-02 -2.85089374e-01 3.84383649e-01
1.75975263e-01 7.17950404e-01 -6.03545904e-01 -7.51117289e-01
1.16398960e-01 4.41101819e-01 -1.59620687e-01 1.91890284e-01
-4.10467446e-01 3.78738195e-01 -5.13470292e-01 7.16124952e-01
6.86140299e-01 -4.83427614e-01 2.18777671e-01 -1.09377034e-01
1.17236003e-02 3.06879915e-02 -8.06562603e-01 1.66028559e+00
-8.27629149e-01 2.50099748e-01 -2.96235859e-01 -1.46006441e+00
1.13263249e+00 2.69698501e-01 5.05125821e-01 -3.80701661e-01
1.74207255e-01 4.04124826e-01 1.74865173e-03 -4.71451730e-01
2.93021083e-01 -2.36441419e-01 7.58683383e-02 5.37180066e-01
1.53753281e-01 2.65579671e-01 5.06638251e-02 -3.07320714e-01
1.00270009e+00 4.34967309e-01 4.52779830e-01 -4.38486487e-02
7.37997532e-01 -1.99507967e-01 7.70002663e-01 4.36098993e-01
-1.47526696e-01 3.37112546e-01 3.70310605e-01 -4.87142116e-01
-6.91206157e-01 -6.33113146e-01 -3.58590901e-01 1.50738633e+00
6.47262856e-02 -4.85410422e-01 -4.97071326e-01 -1.09792364e+00
1.09191924e-01 7.66880572e-01 -7.47619212e-01 -5.61547950e-02
-5.61755955e-01 -1.10522485e+00 5.61322510e-01 7.80201733e-01
5.79481959e-01 -1.15783501e+00 -4.86996889e-01 3.86232197e-01
-1.73792541e-02 -1.06204474e+00 6.23944178e-02 6.13363683e-01
-1.22050595e+00 -9.96458769e-01 -8.12780976e-01 -9.45543945e-01
3.17844927e-01 1.19167671e-01 1.11329150e+00 -1.17704362e-01
-3.22646439e-01 -3.57378036e-01 -5.07881463e-01 -4.40320134e-01
-1.53156474e-01 5.05023241e-01 -4.13077958e-02 2.45566249e-01
7.04369843e-01 -7.64445782e-01 -3.74950975e-01 3.25211585e-01
-5.82423270e-01 2.12400332e-01 9.67589259e-01 1.16407704e+00
4.60753143e-01 1.54508874e-01 9.73900616e-01 -1.13180625e+00
6.06066942e-01 -6.58534646e-01 -3.07033658e-01 4.94812310e-01
-1.22295427e+00 2.27923572e-01 1.05999112e+00 -3.73604357e-01
-1.09083080e+00 -1.16000744e-02 -1.61720946e-01 -2.89681613e-01
-3.38315934e-01 4.72203195e-01 -1.00617059e-01 1.75814897e-01
5.09457767e-01 4.08017933e-01 -1.78104430e-01 -9.19638634e-01
2.75569707e-01 8.74399304e-01 -6.58763722e-02 -5.95099628e-01
5.29618621e-01 1.18441969e-01 2.12310687e-01 -2.40358084e-01
-1.20518100e+00 -3.34747493e-01 -6.36188984e-01 8.74517951e-03
8.12212884e-01 -7.63299882e-01 -6.39183402e-01 6.65866911e-01
-1.05014932e+00 4.34760638e-02 7.80170709e-02 4.70898271e-01
-2.81014949e-01 3.00852954e-01 -4.45492774e-01 -8.56086254e-01
-5.25198877e-01 -1.20049548e+00 1.04482627e+00 2.61104405e-01
6.90344945e-02 -9.08592641e-01 1.09439723e-01 3.61215621e-01
4.69027221e-01 1.36064216e-01 1.40339255e+00 -7.71709621e-01
-3.46931666e-01 -1.66263029e-01 -5.28697491e-01 4.85529959e-01
1.08299233e-01 -2.34722719e-01 -1.26631212e+00 -1.09040745e-01
-1.40626222e-01 -7.02751935e-01 1.32642102e+00 2.74230957e-01
1.78471076e+00 4.86252233e-02 -7.79993951e-01 5.22812009e-01
1.46078420e+00 1.24458581e-01 1.70925394e-01 5.30091286e-01
6.79843664e-01 6.92813873e-01 6.53782248e-01 3.74665618e-01
3.68429095e-01 7.25304246e-01 5.68887413e-01 2.74256710e-02
1.39545903e-01 -3.46311331e-01 1.00025266e-01 8.86149287e-01
-2.13495299e-01 -4.60269868e-01 -6.74651206e-01 3.73847365e-01
-2.02877545e+00 -7.01646328e-01 -2.47112975e-01 1.99043381e+00
8.48926783e-01 1.17371693e-01 2.88230032e-02 4.21427399e-01
7.12856472e-01 3.84158194e-01 -7.30190217e-01 -3.85539681e-01
6.72241747e-02 4.70275134e-01 3.14196736e-01 -1.03974126e-01
-1.48391104e+00 8.37347865e-01 5.55279732e+00 1.22308958e+00
-1.01732576e+00 3.47824872e-01 7.19487786e-01 1.88858867e-01
-1.35781109e-01 -1.85141563e-01 -1.04140270e+00 3.62012774e-01
9.92869318e-01 1.58103816e-02 -1.16578132e-01 1.06459415e+00
-2.81127065e-01 -2.34468691e-02 -1.08678031e+00 7.71208286e-01
-6.86409473e-02 -1.10906613e+00 -1.66489348e-01 1.64245665e-01
5.42246640e-01 -1.36206105e-01 1.33296445e-01 7.48702466e-01
2.20596194e-01 -1.29216397e+00 2.39031389e-01 3.02690208e-01
8.63465965e-01 -6.62460864e-01 1.00255764e+00 5.65449655e-01
-1.27564311e+00 -5.08802056e-01 -4.18671966e-01 -1.40253857e-01
-4.51704040e-02 9.55107152e-01 -5.12879610e-01 7.83637464e-01
6.62519395e-01 9.87526059e-01 -6.69860303e-01 7.60886252e-01
-3.26472223e-01 8.16900432e-01 -1.01400211e-01 -2.68443465e-01
1.68000877e-01 -3.12683463e-01 -1.31081045e-01 1.18323040e+00
4.26964670e-01 -3.58639568e-01 1.81416363e-01 6.27460897e-01
-2.39750773e-01 4.62564826e-01 -7.57581055e-01 -2.77148932e-02
2.92587221e-01 1.50077128e+00 -4.39477593e-01 -2.27591559e-01
-5.92537344e-01 5.07524550e-01 6.17326498e-01 1.20109871e-01
-9.35407698e-01 -3.63072067e-01 5.77076554e-01 -1.64830714e-01
3.60559940e-01 -3.26348916e-02 -5.35161912e-01 -1.19704103e+00
-3.83163989e-02 -8.18923175e-01 4.03949082e-01 -2.40259141e-01
-1.34182858e+00 5.87452710e-01 -3.38896424e-01 -1.31812060e+00
-1.44701198e-01 -7.70635962e-01 -3.86105925e-01 6.94671810e-01
-1.75329006e+00 -1.35219002e+00 -2.85880387e-01 2.13414595e-01
6.17025793e-01 -4.83179063e-01 1.22770810e+00 4.70580041e-01
-7.92791963e-01 6.86137438e-01 5.81516147e-01 -9.67348143e-02
7.43479967e-01 -1.15204012e+00 -5.44296764e-02 4.11406398e-01
3.78054678e-01 3.43489200e-01 3.03963035e-01 -2.38007665e-01
-1.10660815e+00 -1.15372574e+00 1.00851238e+00 -1.66704491e-01
6.17875636e-01 -4.10297364e-01 -7.61784494e-01 5.21951854e-01
2.06641302e-01 1.41577438e-01 9.48385298e-01 5.43243229e-01
-3.46853822e-01 -3.78813118e-01 -1.01324272e+00 2.49688968e-01
1.03940868e+00 -1.58928066e-01 -3.15201789e-01 3.02096039e-01
7.31702030e-01 1.31062895e-01 -1.03519619e+00 9.21297848e-01
8.04295242e-01 -1.12299168e+00 7.34475791e-01 -4.68051791e-01
9.95110333e-01 -7.54425079e-02 -1.04350962e-01 -1.34570062e+00
-4.87907201e-01 2.57817298e-01 -4.20970857e-01 1.35342360e+00
3.92480344e-01 -5.82738578e-01 9.24067140e-01 7.73014082e-03
-9.87135991e-03 -1.48390663e+00 -7.54707277e-01 -6.73959434e-01
2.01363027e-01 -3.23687494e-01 4.71705467e-01 8.31765532e-01
-4.19692338e-01 7.26924300e-01 -6.19649768e-01 -3.25001806e-01
5.76821327e-01 5.88483334e-01 6.74725354e-01 -1.70731044e+00
-3.10873061e-01 -4.29537803e-01 -1.59132779e-01 -9.50968981e-01
5.30663431e-01 -1.02230561e+00 -2.19629645e-01 -1.56308973e+00
2.59288073e-01 -7.44854450e-01 -7.13167548e-01 6.49170280e-01
-3.21561992e-01 1.44673601e-01 1.02815896e-01 2.47053474e-01
-4.75765496e-01 9.42349195e-01 1.19686365e+00 -2.71712095e-01
1.86287418e-01 2.66367346e-01 -7.09214985e-01 6.85763121e-01
8.65299702e-01 -8.10371637e-01 -3.70679617e-01 -3.64381135e-01
-4.76701222e-02 -1.35805354e-01 4.71448787e-02 -1.10550916e+00
-8.40531588e-02 -1.63059771e-01 4.78507638e-01 -6.66808009e-01
3.20441812e-01 -9.44220960e-01 -1.43758714e-01 4.83187169e-01
-4.64690566e-01 -4.01938893e-02 -8.18403140e-02 9.47511315e-01
-3.94477457e-01 -5.18536806e-01 7.67855167e-01 8.92484412e-02
-7.25657880e-01 2.80827492e-01 -2.03433350e-01 -1.77323326e-01
1.15337813e+00 -1.98083773e-01 -2.42257237e-01 9.58534777e-02
-5.57396352e-01 1.11679539e-01 1.04745314e-01 4.47298616e-01
3.84216905e-01 -1.30828726e+00 -6.35165691e-01 1.51948750e-01
3.34422857e-01 1.50859794e-02 1.21937156e-01 8.10538650e-01
-1.44088045e-01 5.51423609e-01 -9.93215516e-02 -5.77428579e-01
-1.20183361e+00 5.88615060e-01 2.61548348e-02 -9.50552464e-01
-6.12351537e-01 8.50006759e-01 3.08159173e-01 -7.74290681e-01
4.88605708e-01 -2.26765990e-01 -5.93924582e-01 2.87260950e-01
1.75881073e-01 2.45305106e-01 2.99043991e-02 -2.90581495e-01
-3.16441506e-01 8.64389658e-01 -1.63090020e-01 3.86035144e-01
1.59744787e+00 2.61917353e-01 -3.76455486e-02 5.90900600e-01
1.37391400e+00 -3.17940235e-01 -8.01811755e-01 -4.27299768e-01
3.41585189e-01 -3.24034452e-01 9.34875533e-02 -1.00465286e+00
-1.08825207e+00 1.42914164e+00 7.18090713e-01 1.71996430e-01
1.25811803e+00 -3.84035885e-01 9.72030461e-01 4.24117565e-01
4.19777244e-01 -1.08630335e+00 3.10563594e-02 3.49936932e-01
6.23962522e-01 -1.45702362e+00 9.99657139e-02 -3.44730288e-01
-5.18522739e-01 1.25327981e+00 9.58576977e-01 1.25857100e-01
8.32760870e-01 1.67349592e-01 -4.22533937e-02 1.89683929e-01
-9.78688776e-01 -1.08977236e-01 5.83859570e-02 4.58307028e-01
8.39350760e-01 1.74153849e-01 -6.28499389e-01 7.90914178e-01
2.06086800e-01 4.19550568e-01 -5.63068725e-02 6.84521317e-01
-5.22387147e-01 -1.50774264e+00 -1.07424185e-01 7.41327107e-01
-4.89777356e-01 -2.66864691e-02 -8.72094482e-02 6.73598707e-01
3.29334468e-01 8.83038044e-01 -3.57586533e-01 -7.61492491e-01
6.87942579e-02 3.60626519e-01 3.56436253e-01 -6.14231467e-01
-6.84908032e-01 -2.78692275e-01 4.06121984e-02 -1.35484189e-01
-6.11915469e-01 -2.99055725e-01 -7.91407764e-01 -2.96412390e-02
-6.26218677e-01 1.49073154e-01 6.99229836e-01 8.93127024e-01
4.59793091e-01 6.85608268e-01 6.27135992e-01 -6.95333183e-01
-1.04470158e+00 -1.46790063e+00 -6.71087682e-01 1.90836936e-01
-9.67136323e-02 -1.08288288e+00 -1.97331533e-01 -1.46374315e-01] | [9.270298957824707, 4.291067600250244] |
ed6a02ee-da89-4037-ae16-a38344960b36 | error-in-variables-modelling-for-operator | 2204.10909 | null | https://arxiv.org/abs/2204.10909v2 | https://arxiv.org/pdf/2204.10909v2.pdf | Error-in-variables modelling for operator learning | Deep operator learning has emerged as a promising tool for reduced-order modelling and PDE model discovery. Leveraging the expressive power of deep neural networks, especially in high dimensions, such methods learn the mapping between functional state variables. While proposed methods have assumed noise only in the dependent variables, experimental and numerical data for operator learning typically exhibit noise in the independent variables as well, since both variables represent signals that are subject to measurement error. In regression on scalar data, failure to account for noisy independent variables can lead to biased parameter estimates. With noisy independent variables, linear models fitted via ordinary least squares (OLS) will show attenuation bias, wherein the slope will be underestimated. In this work, we derive an analogue of attenuation bias for linear operator regression with white noise in both the independent and dependent variables. In the nonlinear setting, we computationally demonstrate underprediction of the action of the Burgers operator in the presence of noise in the independent variable. We propose error-in-variables (EiV) models for two operator regression methods, MOR-Physics and DeepONet, and demonstrate that these new models reduce bias in the presence of noisy independent variables for a variety of operator learning problems. Considering the Burgers operator in 1D and 2D, we demonstrate that EiV operator learning robustly recovers operators in high-noise regimes that defeat OLS operator learning. We also introduce an EiV model for time-evolving PDE discovery and show that OLS and EiV perform similarly in learning the Kuramoto-Sivashinsky evolution operator from corrupted data, suggesting that the effect of bias in OLS operator learning depends on the regularity of the target operator. | ['Mamikon Gulian', 'Myoungkyu Lee', 'Indu Manickam', 'Ravi G. Patel'] | 2022-04-22 | null | null | null | null | ['model-discovery'] | ['miscellaneous'] | [ 8.92314836e-02 -1.27163038e-01 3.64396483e-01 3.06020111e-01
-6.43435359e-01 -2.49972522e-01 2.59309709e-01 -5.16306385e-02
-4.41097856e-01 9.78015721e-01 -5.40166534e-02 -1.97537586e-01
-4.90260273e-01 -5.44728339e-01 -8.80365014e-01 -1.10488200e+00
-5.13027549e-01 4.70191240e-01 -1.22457176e-01 -4.39722806e-01
-2.03142107e-01 5.34831405e-01 -1.08804405e+00 -3.51077318e-01
7.34561324e-01 1.23028386e+00 -4.29438829e-01 8.13912272e-01
3.33705902e-01 9.67605293e-01 -4.16637421e-01 2.95335889e-01
5.02993524e-01 -4.80407625e-01 -5.66998959e-01 -2.06626683e-01
1.68040529e-01 -1.17224194e-01 -7.59320915e-01 1.01566386e+00
6.56149983e-01 4.37176526e-01 8.77817392e-01 -1.00685811e+00
-5.73501945e-01 4.62105662e-01 -2.95476109e-01 3.01103145e-01
6.25520246e-03 5.12308538e-01 9.00400937e-01 -8.07214677e-01
6.72407448e-01 1.04417872e+00 1.24580014e+00 3.97579640e-01
-1.79738724e+00 -5.15608251e-01 -4.35998887e-01 -2.41184101e-01
-1.28694654e+00 -4.76064533e-02 1.07735181e+00 -9.19437647e-01
6.26452625e-01 1.28252238e-01 6.09067023e-01 1.27873027e+00
6.52645648e-01 4.47167248e-01 1.30275083e+00 -3.50350082e-01
3.57192725e-01 1.18817069e-01 3.26209873e-01 8.99853468e-01
1.67556539e-01 7.05749631e-01 -6.87707961e-01 -5.77341199e-01
1.03836524e+00 -1.78102359e-01 -7.17375398e-01 -4.36258882e-01
-9.86251593e-01 1.07232618e+00 4.04211283e-01 2.44401827e-01
-6.39697194e-01 3.16233724e-01 4.77531403e-01 6.21249259e-01
7.04762042e-01 8.48638177e-01 -5.24972975e-01 -1.34445742e-01
-5.44995129e-01 3.02293122e-01 9.37379122e-01 5.93421340e-01
8.82099926e-01 5.12212873e-01 3.65784094e-02 4.01815146e-01
-2.21220925e-02 7.84121811e-01 2.43951067e-01 -1.25118065e+00
-3.10384966e-02 1.65795550e-01 2.89780587e-01 -8.86638880e-01
-7.01280177e-01 -7.33951271e-01 -1.48306477e+00 5.22907913e-01
5.96720994e-01 -5.20901322e-01 -5.81139684e-01 1.74937475e+00
1.37755319e-01 3.59856814e-01 1.53837770e-01 1.14819980e+00
3.05592477e-01 4.80909258e-01 -3.78894359e-01 -5.69415569e-01
6.44748926e-01 -2.48831391e-01 -7.81216323e-01 1.45991877e-01
9.85825360e-01 -2.42268264e-01 9.41433370e-01 4.83068734e-01
-1.19000578e+00 -4.65680778e-01 -7.99156725e-01 1.28550097e-01
-3.62650380e-02 -4.76462185e-01 5.08375943e-01 1.24494568e-01
-8.82332325e-01 1.37307823e+00 -9.22579288e-01 8.17568153e-02
3.26807261e-01 2.82254696e-01 -3.67346287e-01 4.29130435e-01
-1.34324074e+00 8.57662678e-01 -1.56085044e-01 4.62514400e-01
-8.47204506e-01 -1.30608833e+00 -6.41237855e-01 -1.21256243e-02
1.49841413e-01 -6.58284605e-01 8.19443882e-01 -7.37857819e-01
-1.46724975e+00 4.51748908e-01 5.88238873e-02 -6.85735524e-01
8.25758994e-01 -2.64479965e-01 -1.66016102e-01 9.44538265e-02
-2.55540200e-02 -1.61967546e-01 1.13771510e+00 -1.20835960e+00
1.33608822e-02 -2.13536650e-01 -1.73464194e-01 -3.94365489e-01
6.43417314e-02 -6.00128710e-01 5.89175761e-01 -5.92724919e-01
1.55780017e-01 -1.04898441e+00 -3.91896307e-01 1.60872787e-01
-2.44194210e-01 1.17479354e-01 6.84916914e-01 -6.35863602e-01
9.70978618e-01 -2.12740159e+00 4.49788719e-01 3.55123162e-01
5.95484912e-01 -2.40907241e-02 1.59290209e-01 3.23753983e-01
-2.03885153e-01 8.93946216e-02 -7.72607744e-01 -1.63276628e-01
1.68782920e-02 4.29453492e-01 -5.44833541e-01 9.75522697e-01
6.10202253e-01 6.31349802e-01 -8.85338724e-01 -1.99179456e-01
1.48836553e-01 6.03538215e-01 -7.44119227e-01 2.49036044e-01
1.19097963e-01 1.19177854e+00 -2.86702067e-01 9.96546447e-02
5.71321189e-01 -3.46636862e-01 -4.22716618e-01 -2.54153579e-01
-1.22515596e-01 -8.58795047e-02 -1.40458846e+00 1.21682382e+00
-8.62529278e-01 8.55871081e-01 5.44454694e-01 -1.44546831e+00
7.01534629e-01 4.93314475e-01 8.80724609e-01 -3.79446030e-01
2.26932928e-01 5.31801045e-01 2.60909796e-01 -7.57866740e-01
-8.36621895e-02 -7.56212473e-01 -1.97880678e-02 1.98889896e-01
2.72888392e-01 -4.03605342e-01 -7.70504251e-02 5.21056317e-02
1.39797783e+00 -9.35868621e-02 1.59275800e-01 -5.97137332e-01
5.86559176e-01 -1.37529239e-01 6.40569985e-01 1.13407934e+00
-3.59342694e-01 3.37769836e-01 7.79840946e-01 -5.37262619e-01
-1.07221699e+00 -9.00970101e-01 -6.50289416e-01 6.66958094e-01
-1.14831381e-01 1.13788480e-02 -2.35308304e-01 -2.13530526e-01
3.00664306e-01 4.89523798e-01 -7.48184562e-01 -4.42793012e-01
-9.59274173e-01 -9.39385653e-01 6.51438475e-01 2.00801417e-01
4.01403606e-01 -7.98949301e-01 -4.60857093e-01 3.29868466e-01
3.50046456e-01 -1.09087193e+00 -9.13186446e-02 6.73087776e-01
-6.39645040e-01 -1.10613179e+00 -5.09567499e-01 -3.64541471e-01
2.90070355e-01 -6.49914443e-01 8.39724064e-01 2.51240116e-02
-3.75445575e-01 6.48956597e-01 9.52913240e-02 -2.54934311e-01
-7.72318423e-01 -1.92283809e-01 4.86732423e-01 1.92803770e-01
-3.15517604e-01 -8.56541574e-01 -2.69803852e-01 1.26518831e-02
-8.29515278e-01 -5.24984837e-01 7.94319287e-02 1.34022009e+00
3.07748377e-01 1.17582202e-01 2.07780033e-01 -8.05888116e-01
5.84394634e-01 -5.69747865e-01 -8.03881407e-01 -3.96063119e-01
-4.88231063e-01 6.22389615e-01 1.05178034e+00 -7.46787667e-01
-8.19483817e-01 -5.22096865e-02 -5.51784709e-02 -8.05778325e-01
1.98796257e-01 5.79349399e-01 5.42623639e-01 -6.25060618e-01
9.07842100e-01 -1.94718104e-05 1.35129526e-01 -5.06554544e-01
-2.69624025e-01 4.44524214e-02 4.24667001e-01 -7.81972468e-01
9.20620382e-01 6.81379080e-01 9.41278994e-01 -1.20520449e+00
-6.86077952e-01 -3.63177538e-01 -5.78887820e-01 -1.50984719e-01
6.35419965e-01 -6.11045718e-01 -1.07749712e+00 5.97147882e-01
-1.11255908e+00 -5.90090811e-01 -9.48029459e-01 6.49365008e-01
-6.82597995e-01 2.79031008e-01 -8.96448255e-01 -1.20051897e+00
1.16400290e-02 -1.25617671e+00 9.76555467e-01 -2.21972808e-01
-1.11290902e-01 -1.63831365e+00 6.46320656e-02 -2.92489588e-01
4.90274101e-01 5.47390938e-01 8.55530143e-01 -4.29720879e-01
-1.65670276e-01 -2.75838822e-01 1.62821203e-01 5.60776830e-01
-3.51503760e-01 -5.55338115e-02 -9.29147482e-01 -2.97245115e-01
7.94081330e-01 -3.54503155e-01 1.01272738e+00 6.33052945e-01
8.99746358e-01 -2.91489005e-01 -5.73722646e-02 9.75661993e-01
1.41776907e+00 -2.73334831e-01 -5.37586492e-03 -2.38320842e-01
8.89028490e-01 4.24733430e-01 -2.13926077e-01 4.77845579e-01
-3.94392848e-01 3.78116935e-01 1.59003377e-01 -2.19764873e-01
1.93288207e-01 7.06176385e-02 3.57504576e-01 9.00028110e-01
-3.16854268e-01 6.48651198e-02 -8.82120669e-01 1.41975790e-01
-1.62074816e+00 -8.19833338e-01 -6.41680181e-01 2.15431190e+00
8.61834645e-01 2.44255245e-01 -7.14118034e-02 3.08188945e-01
2.03015745e-01 2.70480122e-02 -8.12910020e-01 -2.36655012e-01
-5.82967997e-01 6.62325263e-01 8.11249077e-01 1.15170908e+00
-9.81750965e-01 3.72539371e-01 6.16236639e+00 2.56535083e-01
-1.45999193e+00 3.74721140e-01 3.16084951e-01 2.12633252e-01
4.62083630e-02 4.38827649e-02 -2.94083327e-01 3.66243869e-01
9.73699093e-01 -4.71505970e-02 4.21996385e-01 3.65533262e-01
4.36900944e-01 7.60148019e-02 -9.91457105e-01 9.15281177e-01
-3.40176076e-01 -1.32992232e+00 -3.75527322e-01 2.78355211e-01
9.31655467e-01 3.18100229e-02 2.15256214e-02 2.89883256e-01
1.95943683e-01 -9.43705976e-01 4.94217753e-01 9.82555032e-01
5.36685824e-01 -3.00198972e-01 8.62456083e-01 4.32639450e-01
-8.35856259e-01 -1.47339329e-01 -1.67230561e-01 -5.03116012e-01
3.53717715e-01 9.21098053e-01 -2.30398312e-01 9.78986621e-02
4.57729101e-01 9.78619814e-01 4.54321466e-02 6.32688940e-01
5.60316443e-02 9.08379197e-01 -5.72363198e-01 2.22460613e-01
2.15197533e-01 -5.90052605e-01 1.20432436e+00 8.36819649e-01
2.58825213e-01 3.06387782e-01 2.77300924e-01 1.25170994e+00
9.92328227e-02 -1.68381467e-01 -6.13758743e-01 1.30037174e-01
-1.20590322e-01 7.95469999e-01 -2.71304876e-01 -9.89411250e-02
-2.00401276e-01 7.86551476e-01 1.49843708e-01 9.33378935e-01
-4.93934751e-01 -3.83385755e-02 7.75237560e-01 2.23155871e-01
3.07552844e-01 -4.66491133e-01 -4.27501202e-01 -1.17243469e+00
1.96948677e-01 -5.57569623e-01 2.45005310e-01 -3.84732157e-01
-1.58394122e+00 2.15413332e-01 -7.64396787e-02 -9.54351723e-01
-1.13069460e-01 -1.10574663e+00 -5.66174448e-01 9.56593335e-01
-1.11871743e+00 -5.11616945e-01 1.08590303e-03 4.39802825e-01
-6.11706078e-02 5.81305400e-02 5.64859986e-01 -2.87860483e-02
-5.56421995e-01 1.57537341e-01 5.93747854e-01 1.81261003e-01
4.16622341e-01 -1.42126906e+00 -4.37792279e-02 5.43729544e-01
-1.90202028e-01 5.26392102e-01 1.23477161e+00 -6.23238504e-01
-1.53126121e+00 -6.51837885e-01 4.19813544e-01 -3.34814817e-01
1.17246151e+00 -4.42991138e-01 -1.25674045e+00 5.28393924e-01
-3.11932981e-01 7.92071402e-01 1.16169907e-01 -1.58065766e-01
1.30555360e-02 7.60333706e-03 -1.00450540e+00 2.60441154e-01
9.39526975e-01 -7.42198408e-01 -2.08109077e-02 5.91324985e-01
5.90022445e-01 -4.48279768e-01 -1.15649402e+00 5.24930537e-01
1.83771983e-01 -8.86389852e-01 8.50060403e-01 -1.04186475e+00
3.41594130e-01 1.07116863e-01 -4.34576496e-02 -1.45480335e+00
-2.62547046e-01 -1.34309745e+00 -5.85518837e-01 3.96894366e-01
2.45708138e-01 -9.13342535e-01 3.90886903e-01 4.10631359e-01
-2.07315639e-01 -9.58691359e-01 -1.15842056e+00 -9.14583981e-01
6.62283838e-01 -6.79041564e-01 -1.85590312e-01 9.60657060e-01
-1.03383780e-01 3.14208537e-01 -3.99015963e-01 4.76881832e-01
7.38837063e-01 -4.05563414e-01 3.87212664e-01 -1.41510963e+00
-5.51107883e-01 -4.28090453e-01 -5.17465532e-01 -8.04876864e-01
5.95925093e-01 -8.01692188e-01 2.52652228e-01 -7.29329407e-01
-4.39991713e-01 -4.51289058e-01 -7.62688071e-02 -2.28474274e-01
-1.44831344e-01 1.47591799e-01 -8.56875330e-02 2.83345759e-01
-4.14662249e-02 6.39456689e-01 1.16925871e+00 -3.11751086e-02
-2.73543388e-01 1.71530604e-01 1.39117077e-01 9.26911235e-01
4.82599437e-01 -4.71470714e-01 2.39585061e-02 8.30255635e-03
3.29550147e-01 1.94662645e-01 8.44265282e-01 -1.28816962e+00
2.36917183e-01 9.07290950e-02 2.05580667e-01 1.32030755e-01
3.93999368e-01 -7.07287192e-01 -5.02378009e-02 6.15277231e-01
-5.50042808e-01 4.98118661e-02 1.43473282e-01 7.27829874e-01
-2.55188458e-02 -1.72741130e-01 9.71385479e-01 -8.73915702e-02
-9.96563956e-02 2.77519077e-01 -6.57353640e-01 6.53538346e-01
3.59603137e-01 2.65238643e-01 1.86124399e-01 -6.01583540e-01
-9.03770804e-01 -6.02171570e-02 1.60744265e-01 -4.12002474e-01
2.77881593e-01 -1.08577418e+00 -8.24952304e-01 5.01404643e-01
-2.97714293e-01 -2.39141081e-02 1.77644044e-01 1.31261599e+00
-5.37159622e-01 -7.26987654e-03 2.19072506e-01 -7.16715753e-01
-5.32081902e-01 4.51557457e-01 8.34754229e-01 -4.22394484e-01
-5.68091333e-01 9.79741871e-01 8.39226693e-02 -4.35984433e-01
1.10206649e-01 -6.23378873e-01 4.97661889e-01 -1.35162234e-01
6.23782873e-02 6.39085710e-01 -1.41471764e-02 -7.81355083e-01
8.40526745e-02 5.62506080e-01 4.55341309e-01 -1.36630073e-01
1.17681253e+00 1.00222588e-01 -1.76861167e-01 1.11088932e+00
1.63843620e+00 1.10624963e-02 -1.46305251e+00 -4.99735564e-01
3.40396329e-03 1.29375532e-01 4.28326428e-01 -1.94005311e-01
-9.73965943e-01 9.86420810e-01 5.04928887e-01 5.88206410e-01
7.93373227e-01 -6.06845953e-02 8.27195108e-01 4.86522883e-01
1.79674625e-02 -1.06592405e+00 -1.68857370e-02 8.25619221e-01
8.44279468e-01 -1.41296756e+00 -1.11467734e-01 -2.03207374e-01
-1.41338989e-01 1.16630030e+00 2.64101505e-01 -6.34223998e-01
1.18629885e+00 4.82492030e-01 8.55948869e-03 -3.17712098e-01
-5.46324551e-01 -2.64549032e-02 3.41176033e-01 4.44146186e-01
3.22424024e-01 -1.42778069e-01 -5.16539291e-02 1.91110268e-01
-1.27757028e-01 -2.93623418e-01 6.38480365e-01 8.37258518e-01
1.69640705e-02 -4.78450716e-01 -3.89153421e-01 5.11748493e-01
-2.62669027e-01 2.42681019e-02 -9.56364870e-02 9.33946311e-01
2.43930787e-01 4.22320217e-01 1.00045346e-01 1.52697954e-02
4.17938679e-01 2.46565640e-01 2.77356714e-01 -1.83783442e-01
-6.16861463e-01 -1.00573249e-01 -1.86318204e-01 -4.81168061e-01
-3.37420672e-01 -6.83382332e-01 -1.23793840e+00 -3.67556244e-01
-3.22201639e-01 1.78918526e-01 1.84480086e-01 1.29576695e+00
1.36567250e-01 6.11601651e-01 3.97999823e-01 -1.11134660e+00
-9.60134745e-01 -9.51110125e-01 -8.83170426e-01 6.32425010e-01
1.22401810e+00 -8.96900535e-01 -1.20085049e+00 -2.20329121e-01] | [6.585322856903076, 3.6106417179107666] |
292b32a7-c246-46df-b567-a6fcf8bf2096 | learning-based-point-cloud-registration-for | 2203.15309 | null | https://arxiv.org/abs/2203.15309v2 | https://arxiv.org/pdf/2203.15309v2.pdf | MatchNorm: Learning-based Point Cloud Registration for 6D Object Pose Estimation in the Real World | In this work, we tackle the task of estimating the 6D pose of an object from point cloud data. While recent learning-based approaches to addressing this task have shown great success on synthetic datasets, we have observed them to fail in the presence of real-world data. We thus analyze the causes of these failures, which we trace back to the difference between the feature distributions of the source and target point clouds, and the sensitivity of the widely-used SVD-based loss function to the range of rotation between the two point clouds. We address the first challenge by introducing a new normalization strategy, Match Normalization, and the second via the use of a loss function based on the negative log likelihood of point correspondences. Our two contributions are general and can be applied to many existing learning-based 3D object registration frameworks, which we illustrate by implementing them in two of them, DCP and IDAM. Our experiments on the real-scene TUD-L, LINEMOD and Occluded-LINEMOD datasets evidence the benefits of our strategies. They allow for the first time learning-based 3D object registration methods to achieve meaningful results on real-world data. We therefore expect them to be key to the future development of point cloud registration methods. | ['Mathieu Salzmann', 'Yu Guo', 'Lizhou Wang', 'Zheng Dang'] | 2022-03-29 | null | null | null | null | ['6d-pose-estimation'] | ['computer-vision'] | [ 5.44664077e-02 -2.26588458e-01 9.69608128e-02 -3.58468413e-01
-9.79429424e-01 -5.54718971e-01 9.53083694e-01 1.77289620e-01
-3.86482924e-01 2.31217563e-01 -2.76299715e-01 -1.86862841e-01
-2.38513887e-01 -5.63894629e-01 -8.22788239e-01 -4.93041009e-01
-2.57641286e-01 9.98533607e-01 5.85691333e-01 -2.03023434e-01
4.25047845e-01 1.09809065e+00 -1.55625451e+00 -1.86230093e-01
4.06252503e-01 8.43520284e-01 -2.91300192e-02 2.07073048e-01
-1.39670104e-01 2.22859710e-01 -3.10684413e-01 -8.38033855e-02
7.12833166e-01 6.03010580e-02 -7.10607946e-01 1.33469582e-01
8.12483072e-01 -8.30468535e-02 -7.68191218e-02 7.33154953e-01
5.77331424e-01 1.08299375e-01 8.41315567e-01 -1.49679494e+00
4.71759625e-02 -2.24545062e-01 -6.91049814e-01 -1.04738168e-01
5.56677878e-01 -1.68359473e-01 7.02725232e-01 -1.05479980e+00
7.52300560e-01 1.20025587e+00 1.13631368e+00 1.51209116e-01
-1.39017904e+00 -4.01491582e-01 -2.25613698e-01 4.57878895e-02
-1.45779049e+00 -5.73131263e-01 8.95774484e-01 -8.45806360e-01
8.94157827e-01 1.08239472e-01 5.07184267e-01 5.84074080e-01
7.48375580e-02 2.67635465e-01 1.07884562e+00 -6.82108879e-01
1.58385590e-01 -3.14605646e-02 -1.00890718e-01 4.42168832e-01
1.54436380e-01 3.35322648e-01 -3.33753109e-01 -4.00175214e-01
9.59525466e-01 -1.17124185e-01 -1.27453938e-01 -1.26595914e+00
-1.38732350e+00 8.25797379e-01 4.76531059e-01 2.29433864e-01
-1.86692029e-01 1.38424024e-01 1.09533407e-01 2.33478218e-01
6.65040076e-01 2.51156926e-01 -4.18279499e-01 -2.86440216e-02
-8.37081254e-01 5.39247334e-01 7.46740520e-01 1.14883184e+00
9.75360513e-01 -3.74225020e-01 4.56926405e-01 5.90824366e-01
6.16980374e-01 5.34736216e-01 5.17718419e-02 -9.62654829e-01
3.86220485e-01 4.79632944e-01 2.18257427e-01 -1.11391759e+00
-4.96750683e-01 -2.98248269e-02 -2.30097800e-01 6.19988263e-01
4.71752405e-01 1.58356681e-01 -7.58944988e-01 1.39771533e+00
6.23229742e-01 5.10377169e-01 -1.35282829e-01 7.10942745e-01
5.48238993e-01 2.37579420e-01 -4.30939823e-01 -2.12369766e-02
9.17189240e-01 -4.02325183e-01 -3.22674453e-01 -1.09228618e-01
5.64428151e-01 -1.26140523e+00 8.09845328e-01 -1.63144991e-02
-9.35680807e-01 -4.34810162e-01 -1.06361556e+00 5.16041666e-02
-2.71687329e-01 -2.11213127e-01 4.63217556e-01 3.83500934e-01
-8.62267792e-01 6.52749479e-01 -1.11784863e+00 -6.52872205e-01
3.38231474e-01 5.63230932e-01 -5.19951522e-01 -3.97168510e-02
-5.60526431e-01 1.14634514e+00 1.37291551e-01 2.20649187e-02
-2.36415908e-01 -7.24412918e-01 -7.75698364e-01 -4.05582786e-01
2.77383357e-01 -6.84243560e-01 1.28987408e+00 -3.76609713e-01
-1.11571407e+00 1.24645889e+00 -3.43644544e-02 -1.18927762e-01
7.17162490e-01 -2.23082006e-01 -2.80674100e-02 -1.35152981e-01
2.31974155e-01 5.06831825e-01 6.16938233e-01 -1.63642585e+00
-4.83871490e-01 -5.65663636e-01 -2.86464006e-01 1.77073181e-01
3.02069128e-01 -9.46412459e-02 -5.00539541e-01 -4.57809865e-01
6.81245089e-01 -1.18815768e+00 -2.78030425e-01 3.29040200e-01
-1.01128682e-01 -1.43813461e-01 9.96688306e-01 -1.79019481e-01
3.26557815e-01 -2.34597373e+00 2.44885646e-02 4.03588533e-01
-8.09753910e-02 6.70364425e-02 -1.56265348e-01 6.12533629e-01
-1.79963946e-01 -2.42582142e-01 -3.06165159e-01 -5.19904733e-01
6.98024631e-02 2.81052500e-01 -4.75437731e-01 9.63058531e-01
1.88502058e-01 6.20801270e-01 -7.70986378e-01 -3.80759537e-01
5.31818926e-01 6.34549558e-01 -2.42683753e-01 2.02083826e-01
-4.98927897e-03 3.76869857e-01 -2.47280642e-01 3.60139668e-01
9.13326144e-01 1.41537488e-01 -3.50011468e-01 -3.85796279e-01
-3.25961500e-01 3.46000910e-01 -1.58114541e+00 1.82882738e+00
-2.10140094e-01 4.37301278e-01 -2.71113664e-02 -7.27019429e-01
1.08269846e+00 2.62559950e-01 9.93446052e-01 -3.91691148e-01
-4.96643931e-02 4.37992632e-01 -1.37677833e-01 -2.00274184e-01
3.81041467e-01 -3.77502650e-01 1.22642688e-01 4.51701045e-01
4.95216697e-02 -8.38727355e-01 -2.73977131e-01 -4.55201454e-02
7.40925968e-01 4.81660903e-01 3.32466006e-01 -1.80064097e-01
4.98701960e-01 2.23123118e-01 3.12915653e-01 6.14016116e-01
1.05648590e-02 1.05212855e+00 1.60566732e-01 -5.19990623e-01
-1.15550840e+00 -1.13636208e+00 -4.09245729e-01 3.90666127e-01
2.18672320e-01 -3.26697677e-01 -2.95454651e-01 -5.24394035e-01
3.70428503e-01 5.12986660e-01 -2.46669948e-01 -8.20478126e-02
-7.45568395e-01 -6.10169470e-01 2.32900783e-01 4.01202351e-01
6.79310486e-02 -5.94922841e-01 -5.57495952e-01 8.02048370e-02
1.09556146e-01 -1.36320817e+00 -1.40006557e-01 1.34925693e-01
-1.10680020e+00 -1.17347324e+00 -3.37801874e-01 -6.52065158e-01
5.67201376e-01 3.48805070e-01 1.15793419e+00 6.38577417e-02
-1.12052210e-01 8.28280151e-01 -1.76778987e-01 -5.69342673e-01
-4.17022139e-01 5.55328690e-02 3.05628389e-01 -2.12554589e-01
4.88674849e-01 -7.20968008e-01 -1.69613779e-01 6.03543937e-01
-7.70108402e-01 -3.29986691e-01 2.57796973e-01 4.92758989e-01
8.07773352e-01 -2.69181132e-01 4.77686487e-02 -6.05419457e-01
3.26954842e-01 -1.94138616e-01 -8.46164584e-01 4.00937460e-02
-4.87961173e-01 -3.42783667e-02 -5.58741689e-02 -3.49248618e-01
-4.78910834e-01 5.23119688e-01 -3.36494684e-01 -5.95035553e-01
-2.29816958e-01 3.21763635e-01 -1.74351130e-02 -7.24964559e-01
6.33957624e-01 -4.89261188e-02 1.99717328e-01 -6.14110112e-01
3.03606510e-01 4.31193560e-01 6.41739786e-01 -7.12709963e-01
1.36225462e+00 8.17243755e-01 4.09216225e-01 -9.15357590e-01
-5.42924762e-01 -8.20598066e-01 -1.13261914e+00 -1.11587711e-01
5.88412464e-01 -8.69469941e-01 -4.66515958e-01 4.70769435e-01
-1.41069734e+00 -1.75956097e-02 -4.87739056e-01 6.41174912e-01
-9.65790272e-01 5.39772272e-01 -1.15526363e-01 -5.17779231e-01
6.86612353e-02 -1.27094424e+00 1.37547874e+00 -2.54474938e-01
-4.23596613e-02 -1.03900981e+00 5.95398784e-01 5.32856211e-02
2.04583168e-01 5.46986878e-01 8.02607059e-01 -6.10517621e-01
-6.79608822e-01 -4.42324728e-01 -6.28998801e-02 1.33504093e-01
2.60361999e-01 4.73448187e-02 -9.78754103e-01 -4.52644348e-01
2.95523584e-01 -9.48866159e-02 5.26179910e-01 2.35380560e-01
4.96352285e-01 3.17270547e-01 -4.86493587e-01 6.45975471e-01
1.61959267e+00 -2.10207075e-01 5.32469988e-01 5.01887619e-01
7.65080273e-01 5.57433605e-01 7.28927910e-01 1.15811087e-01
3.17501843e-01 1.26521623e+00 6.78085029e-01 -1.77262872e-01
-1.54734656e-01 -2.19846547e-01 -1.34630324e-02 8.67343783e-01
-3.33590418e-01 3.25450540e-01 -1.33374405e+00 3.95900011e-01
-1.77339733e+00 -6.30041957e-01 -3.97958875e-01 2.64902377e+00
3.55347037e-01 1.25739664e-01 1.40688941e-01 5.83422109e-02
3.81550074e-01 4.87932861e-02 -3.09038669e-01 4.41149101e-02
1.43946204e-02 2.30266377e-01 6.32213652e-01 6.46742404e-01
-1.21674013e+00 7.89750814e-01 6.64037275e+00 4.14530158e-01
-1.12841642e+00 5.42563424e-02 -1.69198692e-01 2.68845528e-01
-3.92758511e-02 3.02384645e-01 -8.22254062e-01 1.48122460e-01
5.19720495e-01 -1.18116196e-02 1.44611850e-01 8.28911424e-01
9.27369762e-03 4.59690094e-02 -1.37071288e+00 1.04984200e+00
1.76245391e-01 -1.16669190e+00 -1.21159635e-01 2.82528073e-01
4.54156280e-01 5.21751940e-01 -2.40888074e-01 -1.00324973e-01
4.46669348e-02 -7.27084935e-01 7.48061657e-01 3.55069876e-01
5.75429559e-01 -4.43208933e-01 6.81371212e-01 4.43747520e-01
-1.18431187e+00 3.86514872e-01 -4.80906427e-01 -5.69504350e-02
2.80204803e-01 5.56255639e-01 -1.03134751e+00 7.62207329e-01
5.55549324e-01 8.10177088e-01 -6.41427934e-01 1.42219746e+00
-5.82953468e-02 1.18809581e-01 -7.90654063e-01 3.89032930e-01
-3.12264748e-02 -2.40787670e-01 9.04231787e-01 8.27227592e-01
3.59438360e-01 -1.80279359e-01 2.82317132e-01 7.70932972e-01
2.84667611e-01 1.38969496e-02 -7.94263244e-01 4.02195990e-01
3.58847558e-01 1.07353699e+00 -6.69190824e-01 2.97859788e-01
-4.75011259e-01 5.82547724e-01 3.03456008e-01 8.81975591e-02
-5.49190223e-01 -1.50542304e-01 7.40926623e-01 5.01485646e-01
3.78949761e-01 -8.37426960e-01 -2.82204181e-01 -1.08049893e+00
2.52341777e-01 -6.45492554e-01 7.56771713e-02 -6.26524627e-01
-1.50180697e+00 4.53521967e-01 4.11939681e-01 -1.60569358e+00
-2.79801607e-01 -6.63258255e-01 -4.97097552e-01 8.36118162e-01
-1.58562052e+00 -1.16239440e+00 -2.69896179e-01 4.84827399e-01
1.50723428e-01 -1.07865140e-01 7.89941251e-01 4.18070674e-01
9.77512226e-02 1.47791907e-01 6.67423010e-02 -3.97829190e-02
9.02027249e-01 -1.05126560e+00 8.23549092e-01 5.29440999e-01
6.06394172e-01 4.07950193e-01 7.47916460e-01 -5.24897099e-01
-1.35499430e+00 -7.58858562e-01 7.83249557e-01 -9.41177428e-01
5.01191795e-01 -5.52994192e-01 -9.92189348e-01 8.08635294e-01
-3.20356309e-01 2.38141507e-01 3.75758827e-01 1.06451914e-01
-2.54910350e-01 -2.87727900e-02 -1.20402849e+00 1.96833134e-01
9.04222965e-01 -3.51660013e-01 -8.30583334e-01 4.24582124e-01
5.53529799e-01 -7.55035698e-01 -9.65156615e-01 7.54711866e-01
3.95169318e-01 -9.71886694e-01 1.29005837e+00 -4.81704921e-01
-2.24727802e-02 -5.82256913e-01 -2.78079867e-01 -1.20174110e+00
-1.72653556e-01 -4.29143816e-01 1.17198728e-01 1.10851693e+00
1.00047149e-01 -6.91653967e-01 7.54345894e-01 4.44121510e-01
-4.87084948e-02 -5.52865744e-01 -1.36344373e+00 -8.67460549e-01
1.89186305e-01 -6.50601685e-01 6.65863693e-01 1.02884173e+00
-4.89420682e-01 1.33318335e-01 -1.64739992e-02 4.20365453e-01
8.16536605e-01 1.60878405e-01 1.30456710e+00 -1.61646676e+00
-3.45657044e-03 -2.62360811e-01 -1.01796436e+00 -1.06514215e+00
2.99683139e-02 -7.15645790e-01 6.55682683e-02 -1.29877627e+00
-1.83486387e-01 -9.65081751e-01 1.79063663e-01 2.78995305e-01
1.66706011e-01 2.98340708e-01 3.21213454e-01 5.27664602e-01
-1.36548087e-01 4.47459996e-01 8.71459484e-01 1.75336137e-01
-2.30143994e-01 2.30041668e-01 -2.42011417e-02 8.48100424e-01
5.26937842e-01 -6.79967403e-01 -1.07711786e-02 -5.87519705e-01
1.27907142e-01 -3.63781512e-01 6.21907532e-01 -1.14634967e+00
3.93205106e-01 -4.93241586e-02 1.88186705e-01 -9.12799835e-01
5.87025344e-01 -1.31712592e+00 3.42970073e-01 2.23386347e-01
1.23418167e-01 2.35978544e-01 2.57194281e-01 3.77959579e-01
-1.17580764e-01 -2.18030542e-01 8.11028063e-01 -4.32566144e-02
-5.74272931e-01 3.49775672e-01 2.51694649e-01 -6.60581067e-02
1.11507893e+00 -4.66323644e-01 4.80065532e-02 -1.99840531e-01
-3.52106601e-01 -1.03950379e-02 9.67951179e-01 4.61834580e-01
6.03579700e-01 -1.49852979e+00 -8.69082272e-01 4.59841341e-01
2.69237429e-01 3.68034810e-01 -3.64162892e-01 8.40963364e-01
-6.30506039e-01 1.78238317e-01 -1.51136920e-01 -1.20365262e+00
-1.29192424e+00 4.36685652e-01 4.93612766e-01 1.55018702e-01
-5.66402912e-01 4.88468021e-01 -2.21296042e-01 -9.13955390e-01
2.47141704e-01 -2.33519375e-01 1.60286129e-01 -1.92483649e-01
2.76395492e-02 4.37889248e-01 4.12846565e-01 -1.07524216e+00
-5.05667150e-01 1.13792121e+00 7.77234212e-02 -7.24667683e-02
1.60622931e+00 7.53521994e-02 -2.56671011e-01 6.05701983e-01
1.22764099e+00 1.28859505e-01 -1.13251126e+00 -3.64723116e-01
2.41503298e-01 -7.60750115e-01 -1.05211802e-01 -2.32351348e-01
-8.24136257e-01 7.27719843e-01 9.41846490e-01 1.11430407e-01
7.49737203e-01 3.09928685e-01 5.00534773e-01 2.84668386e-01
5.41269660e-01 -6.12661242e-01 -2.41782472e-01 5.14524281e-01
7.87748277e-01 -1.33591080e+00 4.82830256e-01 -7.79857874e-01
2.75447569e-03 1.16045427e+00 2.08348528e-01 -3.76897812e-01
7.34806180e-01 3.51026535e-01 3.15582246e-01 -4.15048808e-01
-1.72747821e-01 -1.64130926e-01 4.52261806e-01 7.89557040e-01
2.99901009e-01 -2.92095602e-01 -4.19086479e-02 -1.89054310e-01
-1.83783218e-01 -2.73758229e-02 2.67073900e-01 1.21542227e+00
-2.44712487e-01 -1.52441144e+00 -7.37799466e-01 3.84461433e-02
-1.65842241e-03 3.38492543e-01 -4.96013790e-01 1.06503499e+00
2.58800369e-02 4.28709537e-01 1.90304369e-01 -2.15664193e-01
8.85760546e-01 4.83171828e-02 7.98018217e-01 -6.35876536e-01
-3.29613864e-01 1.44999221e-01 -2.93086946e-01 -5.29886246e-01
-8.69520307e-01 -1.03812635e+00 -9.56130564e-01 -3.43267098e-02
-4.95581269e-01 -2.41980068e-02 9.21905994e-01 8.91200721e-01
3.19592774e-01 -1.67959318e-01 6.22863233e-01 -1.42557311e+00
-7.92369187e-01 -7.01769888e-01 -4.87139285e-01 5.27992189e-01
4.31059748e-01 -9.88968909e-01 -5.00550985e-01 -2.16668081e-02] | [7.700311660766602, -2.841219186782837] |
0955d98e-056a-42fd-ab6e-323b230f7395 | tackling-online-abuse-a-survey-of-automated | 1908.06024 | null | https://arxiv.org/abs/1908.06024v2 | https://arxiv.org/pdf/1908.06024v2.pdf | Tackling Online Abuse: A Survey of Automated Abuse Detection Methods | Abuse on the Internet represents an important societal problem of our time. Millions of Internet users face harassment, racism, personal attacks, and other types of abuse on online platforms. The psychological effects of such abuse on individuals can be profound and lasting. Consequently, over the past few years, there has been a substantial research effort towards automated abuse detection in the field of natural language processing (NLP). In this paper, we present a comprehensive survey of the methods that have been proposed to date, thus providing a platform for further development of this area. We describe the existing datasets and review the computational approaches to abuse detection, analyzing their strengths and limitations. We discuss the main trends that emerge, highlight the challenges that remain, outline possible solutions, and propose guidelines for ethics and explainability | ['Pushkar Mishra', 'Ekaterina Shutova', 'Helen Yannakoudakis'] | 2019-08-13 | null | null | null | null | ['abuse-detection'] | ['natural-language-processing'] | [-4.60749045e-02 -7.24484846e-02 -4.50245470e-01 -4.08384711e-01
-2.10999325e-01 -8.07855427e-01 2.26224229e-01 4.76571739e-01
-5.68541169e-01 9.16531086e-01 1.72126532e-01 -2.59631693e-01
1.07739614e-02 -5.58090568e-01 -6.39523044e-02 -2.70330422e-02
-2.19733477e-01 5.11933491e-03 -3.40850979e-01 -2.07518488e-01
6.40248656e-01 6.78671718e-01 -9.58528817e-01 7.00522736e-02
9.26514328e-01 3.36590469e-01 -7.80478597e-01 1.57008931e-01
-2.32824162e-01 4.36725676e-01 -8.90785813e-01 -1.63012469e+00
5.99832833e-03 -3.44962776e-01 -9.19928193e-01 8.22554678e-02
3.77176464e-01 -8.98073375e-01 -4.77756500e-01 1.43179357e+00
4.40383136e-01 -1.69090517e-02 1.98851794e-01 -1.58927345e+00
-5.89895904e-01 6.62070274e-01 -7.45710313e-01 5.79796314e-01
1.05382061e+00 1.51136652e-01 4.97536749e-01 -4.80106771e-01
8.92482579e-01 1.19828141e+00 7.44073391e-01 6.08558893e-01
-8.09639752e-01 -1.04333174e+00 -1.26830906e-01 4.60823700e-02
-1.41623759e+00 -4.17489678e-01 9.26487327e-01 -3.38037640e-01
9.74733829e-01 1.95301265e-01 1.15183616e+00 1.60230374e+00
9.19549391e-02 5.90159833e-01 9.20092404e-01 -3.62621784e-01
7.87709504e-02 3.33292246e-01 5.33401191e-01 5.96944749e-01
8.55698287e-01 -5.27793705e-01 -6.53796136e-01 -1.24791741e+00
3.05065542e-01 -1.94388345e-01 4.45104763e-02 6.46097511e-02
-1.37378365e-01 1.09069860e+00 -4.41248640e-02 5.53919971e-01
-3.21249306e-01 -3.10383409e-01 6.44136250e-01 -1.07146442e-01
4.45554346e-01 3.31866860e-01 -3.25644910e-02 -8.16817939e-01
-8.53737831e-01 5.07640362e-01 1.08931231e+00 6.05840087e-01
1.69113040e-01 -1.84997067e-01 7.37073481e-01 9.84456062e-01
2.27475405e-01 5.26627488e-02 3.90911996e-01 -8.95373523e-01
5.77467501e-01 3.99308920e-01 -1.37735844e-01 -1.59953475e+00
-8.14603269e-02 -2.56558694e-02 -4.48426723e-01 -2.82837003e-01
4.15443748e-01 -4.72857088e-01 -1.31583408e-01 1.54574704e+00
3.51123095e-01 3.41801643e-02 -4.48836029e-01 4.59026426e-01
2.38372847e-01 1.84000224e-01 6.58250749e-01 -4.82039064e-01
1.28344524e+00 -2.61097908e-01 -1.32138753e+00 -5.54429531e-01
7.72389233e-01 -6.80094600e-01 5.86160779e-01 6.39785230e-01
-1.30351865e+00 5.15359402e-01 -9.86060917e-01 -2.56115887e-02
-5.69723487e-01 -6.12568796e-01 9.52866554e-01 1.78261971e+00
-6.49179697e-01 7.61607051e-01 -5.98404408e-01 -8.92057061e-01
7.92674899e-01 1.35249659e-01 -4.49315310e-01 -1.67749360e-01
-1.33845603e+00 9.22032177e-01 1.76771492e-01 -5.84757738e-02
1.72618181e-02 -4.31874365e-01 -8.16001117e-01 -3.84367734e-01
3.71153533e-01 -2.81413853e-01 8.96659732e-01 -7.92967916e-01
-9.48890686e-01 1.28320658e+00 -2.82234102e-02 -3.35479051e-01
3.73913229e-01 -4.65468824e-01 -8.33906472e-01 3.01286697e-01
1.47641346e-01 5.22858277e-02 4.66421604e-01 -7.87893772e-01
2.96165012e-02 -9.31657612e-01 9.20005441e-02 1.44026997e-02
-9.02945638e-01 1.16816449e+00 4.62925807e-02 -7.06471324e-01
-2.90291488e-01 -8.09122562e-01 -1.23388454e-01 2.54158854e-01
-5.24071693e-01 -1.19251356e-01 6.83164299e-01 -1.00555825e+00
1.83038592e+00 -2.35643196e+00 -1.13475159e-01 1.56987652e-01
2.52046525e-01 6.93194270e-01 4.12892818e-01 1.04473233e+00
-3.10354680e-02 1.14907598e+00 -4.99081999e-01 -4.84850496e-01
-2.71044016e-01 3.27080488e-01 -3.22618097e-01 6.79738879e-01
-1.24669947e-01 6.81425273e-01 -9.36210811e-01 -5.23368716e-01
-1.03146076e-01 3.03774923e-01 -2.46760935e-01 1.09316424e-01
4.72685516e-01 6.74105659e-02 -5.12190044e-01 9.79754806e-01
1.03579831e+00 4.29436028e-01 4.26791191e-01 6.26434147e-01
-9.13539976e-02 3.95984560e-01 -7.30398059e-01 1.34542942e+00
1.98366299e-01 4.36519921e-01 6.44089580e-01 -4.45380300e-01
5.98713040e-01 2.36622572e-01 3.58077943e-01 -1.24790266e-01
3.32660288e-01 2.99097955e-01 1.14698336e-01 -1.01802790e+00
4.58708376e-01 -3.46964329e-01 -2.84392178e-01 7.64537692e-01
-9.80873033e-02 7.50071853e-02 7.69332796e-02 3.93602312e-01
1.15391660e+00 -3.77299309e-01 8.60030055e-01 3.26212943e-01
1.79536730e-01 -1.75134137e-01 7.20560551e-01 5.25457740e-01
-1.06050897e+00 2.27082912e-02 8.33313763e-01 -5.01914382e-01
-7.53611207e-01 -8.45946491e-01 -2.21161336e-01 6.93051755e-01
-1.57491177e-01 -7.39188790e-01 -9.72991467e-01 -8.53166580e-01
1.94558084e-01 9.13534939e-01 -2.64692247e-01 -1.99805588e-01
-4.73850161e-01 -9.32889938e-01 1.42266464e+00 1.13888897e-01
6.67108893e-01 -1.00209653e+00 -4.16554362e-01 -3.00444756e-02
-4.38316017e-01 -1.52553403e+00 1.18035115e-01 -9.51389492e-01
-9.55529869e-01 -1.10477233e+00 -1.28456742e-01 -1.15520738e-01
4.22406107e-01 2.73110688e-01 6.80191696e-01 5.18986583e-01
-5.51199913e-01 3.35581720e-01 -4.50612456e-01 -3.40622753e-01
-2.78027236e-01 -1.52718633e-01 5.22182941e-01 -2.24014401e-01
9.59893048e-01 -8.97597671e-01 -6.14907630e-02 -3.75545502e-01
-1.13583088e+00 -7.35255957e-01 -3.04386262e-02 5.12497500e-02
-4.38929260e-01 3.90175954e-02 6.11011147e-01 -1.25267637e+00
1.26127720e+00 -1.07519543e+00 1.39988035e-01 -1.01980135e-01
-4.52845007e-01 -9.74636436e-01 2.79616535e-01 -2.24934489e-01
-8.95263851e-01 -5.15988946e-01 -5.03930271e-01 -7.11948648e-02
-5.30165195e-01 6.17123127e-01 -2.93517023e-01 -3.80509794e-01
7.11959541e-01 -2.59122908e-01 8.83063674e-03 -6.70444369e-01
2.22531445e-02 9.58350539e-01 -1.81119531e-01 -3.86137396e-01
7.03201652e-01 4.80234325e-01 -2.64313400e-01 -1.41953242e+00
-7.63094604e-01 -3.20811033e-01 -6.17361188e-01 -3.17836225e-01
5.48860371e-01 -2.89358944e-01 -4.91216481e-01 8.71890366e-01
-1.28525889e+00 3.52463305e-01 3.78871292e-01 2.48970121e-01
7.78283030e-02 1.20555210e+00 -1.09761286e+00 -1.30212522e+00
-2.83577889e-01 -6.74236059e-01 2.25950509e-01 2.00246975e-01
-9.32329535e-01 -9.44179177e-01 1.65080518e-01 9.93464470e-01
2.92095929e-01 5.90232670e-01 9.05033410e-01 -9.59403038e-01
1.60772756e-01 -3.92100960e-01 7.21863508e-02 2.26153895e-01
-9.86860041e-03 3.46332878e-01 -6.83285356e-01 -2.71702647e-01
3.18642080e-01 -6.05851054e-01 8.58556703e-02 -2.23472446e-01
1.09067595e+00 -7.92826474e-01 -3.69327277e-01 1.88117340e-01
1.12354434e+00 2.14855038e-02 7.28373230e-01 3.45765650e-01
2.83392012e-01 8.86579871e-01 4.55845356e-01 6.83559299e-01
-7.36292973e-02 4.02214199e-01 2.70739019e-01 6.05017483e-01
4.35584396e-01 -4.07155424e-01 4.07872289e-01 3.93929094e-01
1.57471257e-03 -1.11451149e-01 -9.54112411e-01 5.05036235e-01
-1.37807500e+00 -9.50460434e-01 -2.42154747e-01 2.14023852e+00
5.46233058e-01 1.19436048e-01 5.50468385e-01 1.83673233e-01
9.50423121e-01 2.37802103e-01 -3.78101975e-01 -1.13409376e+00
2.54072845e-02 1.11921325e-01 -4.39020980e-04 3.22921962e-01
-8.43005836e-01 8.65673244e-01 7.73036671e+00 4.45740879e-01
-6.86931670e-01 8.75697285e-02 4.27941620e-01 -1.92342237e-01
-2.16946751e-02 1.75218489e-02 -3.62978995e-01 6.13060594e-01
1.03700995e+00 -4.09737110e-01 6.46855712e-01 8.78822923e-01
6.39315024e-02 -7.45011196e-02 -4.83877093e-01 9.78177845e-01
6.06567562e-01 -5.84389746e-01 -5.27765267e-02 5.10601163e-01
3.64773944e-02 -1.89852566e-01 8.51218868e-03 -9.59658250e-03
-2.39237264e-01 -9.32410836e-01 1.48581088e-01 -9.00565311e-02
4.68658864e-01 -9.76131082e-01 5.91831565e-01 5.31668007e-01
-1.99049398e-01 -1.67854562e-01 -2.57929951e-01 -7.51685858e-01
6.75212741e-01 7.49827385e-01 -3.89926702e-01 4.08086210e-01
6.68706298e-01 6.99413002e-01 -4.12067235e-01 1.06680965e+00
-5.03914237e-01 7.95052886e-01 -4.91072267e-01 -6.54017776e-02
-2.55720522e-02 -5.52280545e-01 9.34921384e-01 1.16626644e+00
3.46077383e-02 4.61182654e-01 -1.92034304e-01 5.93215227e-01
-3.41379255e-01 2.10455745e-01 -1.17571616e+00 -8.64622891e-01
8.33690345e-01 1.30502224e+00 -5.10174334e-01 4.47186641e-02
-5.30800700e-01 9.91668522e-01 6.22739434e-01 9.00891423e-02
-6.00159943e-01 -3.31729829e-01 1.01016414e+00 3.78809214e-01
-5.37817359e-01 -5.37838459e-01 -2.90186703e-01 -1.38463330e+00
1.83920175e-01 -1.05461168e+00 6.58038259e-01 -4.83077705e-01
-1.57641196e+00 1.49904817e-01 1.95467323e-01 -5.32619834e-01
-3.68035018e-01 -3.97477925e-01 -4.01815027e-01 4.47589695e-01
-7.40712941e-01 -8.88944983e-01 2.87480503e-01 1.01272598e-01
8.28785300e-02 2.14467555e-01 9.79086578e-01 4.25052941e-01
-9.67290044e-01 6.99475646e-01 -5.44712186e-01 4.56863940e-01
6.13494396e-01 -3.18491638e-01 2.44866565e-01 8.86286914e-01
-2.57738113e-01 1.24498141e+00 6.17036045e-01 -1.14540160e+00
-1.30137312e+00 -4.60595906e-01 1.21558702e+00 -4.35290158e-01
1.00980151e+00 -5.90855718e-01 -1.10860419e+00 1.04691708e+00
3.18550169e-01 -3.97196144e-01 1.71666682e+00 4.03019339e-01
-7.13343441e-01 5.19795358e-01 -1.95860386e+00 6.13951862e-01
1.23147810e+00 -6.34826958e-01 -7.23337710e-01 4.31349456e-01
1.93268105e-01 -7.54974633e-02 -8.56114209e-01 -2.10033372e-01
7.24987745e-01 -1.10426712e+00 6.74121737e-01 -1.14921999e+00
3.97034436e-01 6.99686885e-01 4.85464871e-01 -8.22992682e-01
-2.22675994e-01 -9.62896228e-01 -1.10208228e-01 1.60267758e+00
-7.77570754e-02 -8.69654059e-01 6.52367234e-01 1.63291013e+00
3.72063816e-01 -6.19088769e-01 -1.00508857e+00 -4.11252886e-01
3.59242141e-01 -5.81226170e-01 3.19465965e-01 1.76240706e+00
8.70605826e-01 -2.85057537e-02 -5.05831599e-01 -7.23678991e-02
8.30396354e-01 -5.76609433e-01 4.68955338e-01 -1.07759976e+00
4.63321246e-02 -4.12251770e-01 -5.66135526e-01 -2.74488658e-01
4.05235589e-01 -4.88657773e-01 -1.04759097e+00 -1.01652431e+00
5.48838437e-01 9.45037752e-02 2.71380186e-01 5.22659123e-01
4.10708133e-03 2.77302921e-01 3.06007802e-01 4.85332012e-02
-2.65801579e-01 1.50621861e-01 6.35548711e-01 4.15370554e-01
-2.28073988e-02 -1.79058984e-01 -1.13266826e+00 1.07883203e+00
1.05378532e+00 -7.24651039e-01 8.77693295e-02 -2.87359267e-01
7.54785657e-01 -2.45934092e-02 4.42481153e-02 -5.48209488e-01
-5.82489371e-02 -4.77588654e-01 1.87360913e-01 -6.86120838e-02
6.70436025e-01 -6.38081193e-01 -1.92628235e-01 4.73342121e-01
-3.34657989e-02 4.82066646e-02 2.62429863e-01 4.08309281e-01
1.15809098e-01 -9.01542962e-01 6.06654882e-01 -1.81656867e-01
6.97616711e-02 1.82345629e-01 -8.74967217e-01 4.31746431e-02
1.28469968e+00 -8.14403128e-03 -2.81554639e-01 -5.84798694e-01
-5.07007301e-01 5.13337366e-02 6.41191602e-01 4.29721743e-01
7.05287099e-01 -9.73845184e-01 -4.69014347e-01 -7.71592483e-02
-9.42507237e-02 -9.15456414e-01 2.99759388e-01 4.40149397e-01
-3.52969080e-01 8.60190988e-02 -5.01821697e-01 6.00716949e-01
-1.54373968e+00 6.84901834e-01 1.55939264e-02 1.15971044e-02
-1.84277192e-01 4.89398569e-01 -7.08459735e-01 -1.69844717e-01
-1.90613538e-01 6.72283769e-01 -4.06683922e-01 -1.78654082e-02
9.16575551e-01 9.17885423e-01 -2.04280302e-01 -9.83758926e-01
-6.32943451e-01 -6.60258066e-03 -4.20317799e-01 -1.77918494e-01
1.28676820e+00 -2.41097599e-01 -9.11893845e-01 2.71221429e-01
1.13556027e+00 4.36011136e-01 -1.12946674e-01 1.43572643e-01
-4.82556969e-02 -1.11254740e+00 -4.01676089e-01 -5.76983452e-01
-6.85977042e-01 7.43670106e-01 -1.98810488e-01 6.26523376e-01
8.44908416e-01 -1.00730851e-01 1.13306105e+00 1.49166450e-01
5.93745589e-01 -1.00952065e+00 -1.30663365e-01 2.63498098e-01
7.58518338e-01 -7.34541774e-01 4.33445781e-01 -1.20128477e+00
-4.29550946e-01 1.00801003e+00 5.03427744e-01 -1.66687682e-01
7.95345545e-01 9.67245251e-02 -2.75863826e-01 -2.78316855e-01
-5.01314774e-02 4.92625147e-01 -5.40933967e-01 8.16006362e-01
5.01277149e-01 -8.51730257e-02 -1.28861141e+00 8.29895616e-01
-4.03085053e-01 1.05630420e-01 1.01344395e+00 1.32372856e+00
-9.51054022e-02 -1.64827955e+00 -5.53925872e-01 4.92979527e-01
-1.19591725e+00 5.87149076e-02 -1.40581656e+00 8.12087655e-01
-6.05538376e-02 1.16067791e+00 -3.24435800e-01 -3.25392336e-01
4.50198688e-02 3.44274610e-01 4.39348221e-01 -5.72900414e-01
-9.71492410e-01 -5.10857940e-01 6.17231548e-01 -5.93204200e-01
-2.24910542e-01 -1.08850801e+00 -9.11117733e-01 -9.56780076e-01
-4.72105928e-02 1.77726857e-02 5.86708903e-01 9.88598049e-01
5.60228825e-01 -5.11766732e-01 1.74237907e-01 -2.56849051e-01
-5.15248716e-01 -9.53252971e-01 -8.06867957e-01 6.20825112e-01
-6.23840205e-02 -5.02623260e-01 -6.05837584e-01 -4.71642584e-01] | [8.72620964050293, 10.378969192504883] |
8bd42390-c727-44e3-b7c6-fdd66f50616d | breaking-common-sense-whoops-a-vision-and | 2303.07274 | null | https://arxiv.org/abs/2303.07274v2 | https://arxiv.org/pdf/2303.07274v2.pdf | Breaking Common Sense: WHOOPS! A Vision-and-Language Benchmark of Synthetic and Compositional Images | Weird, unusual, and uncanny images pique the curiosity of observers because they challenge commonsense. For example, an image released during the 2022 world cup depicts the famous soccer stars Lionel Messi and Cristiano Ronaldo playing chess, which playfully violates our expectation that their competition should occur on the football field. Humans can easily recognize and interpret these unconventional images, but can AI models do the same? We introduce WHOOPS!, a new dataset and benchmark for visual commonsense. The dataset is comprised of purposefully commonsense-defying images created by designers using publicly-available image generation tools like Midjourney. We consider several tasks posed over the dataset. In addition to image captioning, cross-modal matching, and visual question answering, we introduce a difficult explanation generation task, where models must identify and explain why a given image is unusual. Our results show that state-of-the-art models such as GPT3 and BLIP2 still lag behind human performance on WHOOPS!. We hope our dataset will inspire the development of AI models with stronger visual commonsense reasoning abilities. Data, models and code are available at the project website: whoops-benchmark.github.io | ['Roy Schwartz', 'Gabriel Stanovsky', 'Yuval Elovici', 'Ludwig Schmidt', 'Jack Hessel', 'Yonatan Bitton', 'Nitzan Bitton-Guetta'] | 2023-03-13 | null | null | null | null | ['explanation-generation', 'common-sense-reasoning', 'visual-commonsense-reasoning'] | ['natural-language-processing', 'reasoning', 'reasoning'] | [ 2.66351819e-01 3.82988960e-01 1.52238622e-01 -2.01517493e-01
-4.53099996e-01 -7.72195339e-01 7.23897874e-01 -9.90328938e-02
-1.57002792e-01 6.65255070e-01 3.45895857e-01 -5.38156807e-01
2.01986089e-01 -5.10849655e-01 -1.07161999e+00 -1.59267247e-01
5.16729355e-01 5.30992091e-01 1.16832303e-02 -7.21967578e-01
6.02227867e-01 1.70015424e-01 -1.93405724e+00 9.32587862e-01
8.35898876e-01 9.47323203e-01 2.33934402e-01 7.74132133e-01
4.47258614e-02 1.55337262e+00 -7.43924439e-01 -9.02342558e-01
2.35314965e-01 -7.59187460e-01 -9.79024351e-01 2.71978658e-02
1.04437149e+00 -1.27317756e-01 -4.91384655e-01 1.31519318e+00
1.84199750e-01 1.00574717e-01 5.30881047e-01 -1.96952748e+00
-1.65256262e+00 6.17291629e-01 -4.05917138e-01 5.25362432e-01
5.78220248e-01 9.76526082e-01 1.00667608e+00 -6.19800568e-01
9.87826049e-01 1.40883553e+00 4.41945672e-01 8.70798945e-01
-1.23867774e+00 -9.10854459e-01 -1.52964666e-01 8.78816545e-01
-1.12925744e+00 -4.31335568e-01 7.96241760e-01 -6.90813005e-01
8.59067202e-01 6.94856048e-01 8.75797808e-01 1.47663808e+00
3.80535722e-02 8.55586886e-01 1.35801888e+00 -1.61098510e-01
-5.92598431e-02 1.96624175e-01 -2.15232506e-01 7.07866013e-01
2.88423866e-01 2.72646099e-01 -6.86787188e-01 1.46694973e-01
6.12352729e-01 -2.04864591e-01 -3.17440391e-01 -2.54554719e-01
-1.59017634e+00 7.50042021e-01 7.98544765e-01 1.32636577e-01
-2.78497964e-01 4.94153976e-01 3.14368129e-01 2.49815211e-01
3.26273702e-02 1.03194773e+00 2.28791997e-01 -2.51650631e-01
-6.16056025e-01 7.27389157e-01 3.65553200e-01 9.16782975e-01
4.42522436e-01 6.18180260e-02 -1.23532139e-01 5.34741342e-01
-3.71681228e-02 6.63091481e-01 4.12255645e-01 -1.33015120e+00
3.96595776e-01 6.09089494e-01 2.46471018e-01 -1.65638781e+00
-1.53172523e-01 -4.08703840e-04 -6.47043109e-01 5.82149923e-01
7.37198770e-01 1.70450822e-01 -1.05759978e+00 1.61886132e+00
-1.04231656e-01 -1.20206811e-01 -1.20588124e-01 1.34458387e+00
1.05414009e+00 5.35721540e-01 4.68225062e-01 4.73585457e-01
1.62143672e+00 -7.84118176e-01 -5.16197979e-01 -7.60350764e-01
2.05564469e-01 -7.35630989e-01 1.72987854e+00 4.53415573e-01
-1.20869851e+00 -5.33721805e-01 -1.00728869e+00 -4.10664290e-01
-5.88710904e-01 -1.80147856e-01 7.69738615e-01 3.13488722e-01
-6.90019071e-01 2.92102993e-01 -1.20305955e-01 -4.86094415e-01
7.68200278e-01 -4.25903231e-01 -4.43734378e-01 -1.52784511e-01
-1.23765492e+00 1.26506305e+00 3.71339917e-01 -8.84072930e-02
-8.96355689e-01 -7.22342968e-01 -9.98929203e-01 -9.35298875e-02
3.47506553e-01 -8.16468894e-01 1.21257710e+00 -1.65283024e+00
-5.59335053e-01 1.68797398e+00 1.47652715e-01 -6.20775938e-01
6.40429735e-01 -1.29195094e-01 -4.46834981e-01 2.76407748e-01
3.71822119e-01 1.08366132e+00 7.76188612e-01 -1.54011583e+00
-3.16731483e-01 -2.41715133e-01 4.27682161e-01 -6.11746162e-02
1.92767248e-01 1.03325725e-01 8.28975216e-02 -7.23757088e-01
-2.79336721e-01 -8.57592821e-01 1.71736896e-01 3.31343800e-01
-6.28580391e-01 2.54919175e-02 5.21918356e-01 -8.94297421e-01
9.88033831e-01 -2.10852218e+00 -9.90075395e-02 -1.83591515e-01
3.38750511e-01 -1.25646949e-01 -1.20701842e-01 2.21750796e-01
-3.87208641e-01 3.73082012e-01 -1.26274779e-01 2.47256905e-01
4.68633026e-01 -3.23643759e-02 -6.15783334e-01 1.98014811e-01
1.10468179e-01 1.26527524e+00 -1.20682740e+00 -7.30198503e-01
2.07020313e-01 8.48679468e-02 -4.43495721e-01 -2.07651854e-02
-4.09542352e-01 2.42201790e-01 6.36148378e-02 6.53121054e-01
4.93525386e-01 -5.20168304e-01 -2.22682893e-01 -3.60029161e-01
6.71027750e-02 -1.54620288e-02 -6.21299267e-01 1.50232470e+00
-1.66463852e-01 1.33143318e+00 -5.07143557e-01 -4.34411168e-01
6.57948256e-01 -1.25642225e-01 -1.97875336e-01 -1.21815610e+00
8.14762041e-02 -7.36895949e-02 1.46588996e-01 -9.31158483e-01
8.42475712e-01 -6.08997285e-01 -3.15403491e-01 3.88983220e-01
-3.52418780e-01 -7.95232892e-01 3.44238341e-01 6.73982561e-01
8.18968236e-01 1.63204759e-01 3.63776386e-01 -1.55306300e-02
7.56739751e-02 8.19594800e-01 1.61091298e-01 9.14332628e-01
-4.95767325e-01 9.14372563e-01 7.53538907e-01 -9.07889247e-01
-1.57500625e+00 -1.22891068e+00 3.27412009e-01 1.17664313e+00
5.17946064e-01 -1.70215175e-01 -6.22494876e-01 -3.74724001e-01
1.18716374e-01 1.48546934e+00 -1.06164801e+00 -2.68362254e-01
-1.61184013e-01 -1.66550949e-01 7.49648809e-01 3.25257868e-01
6.13535821e-01 -1.50452173e+00 -1.21489894e+00 -3.45525861e-01
-6.02751195e-01 -1.09950423e+00 -4.36201423e-01 -4.81899828e-01
-3.71904783e-02 -1.40509856e+00 -5.19065738e-01 -5.19541562e-01
5.13456821e-01 3.40311289e-01 1.66991675e+00 3.65215749e-01
-6.71782374e-01 3.34012866e-01 -2.54320145e-01 -6.06075525e-01
-6.13771558e-01 -6.43644869e-01 -4.11382943e-01 -1.91120237e-01
5.59033096e-01 -4.30741102e-01 -7.35592246e-01 2.04097971e-01
-1.04746318e+00 6.87691689e-01 3.47284466e-01 7.05400527e-01
2.46780053e-01 -4.33072567e-01 3.53468925e-01 -7.41754413e-01
5.99978983e-01 -7.13221848e-01 -1.89507112e-01 3.52431566e-01
-1.81861192e-01 1.92768518e-02 3.97173554e-01 -4.54297483e-01
-1.00784326e+00 -1.61824599e-01 4.54619378e-01 -6.47994041e-01
-1.88723922e-01 -4.07409184e-02 1.09443709e-01 1.83056176e-01
1.29481912e+00 2.36294970e-01 -1.93100765e-01 1.46597043e-01
6.96249783e-01 4.66276467e-01 1.08714819e+00 -5.28134644e-01
8.28535378e-01 7.10812867e-01 -3.30630660e-01 -5.31522572e-01
-1.08511400e+00 -1.18077822e-01 4.17636670e-02 -7.54281759e-01
1.08569753e+00 -7.40108311e-01 -1.23143756e+00 2.52522588e-01
-1.54133213e+00 -2.07205534e-01 -4.87940222e-01 -1.11169219e-01
-8.16083789e-01 2.36749992e-01 -1.08516343e-01 -6.71483040e-01
-1.57651350e-01 -7.50212789e-01 7.41703868e-01 3.25482339e-01
-7.45249927e-01 -4.87716317e-01 -1.71068862e-01 9.91146743e-01
4.62884754e-01 7.70696223e-01 1.04917574e+00 -4.83363569e-01
-6.06004179e-01 7.81372637e-02 -8.26309085e-01 3.36443961e-01
-4.54844952e-01 -3.17394920e-02 -7.62335956e-01 4.29696023e-01
-4.07926559e-01 -1.03099310e+00 7.61681139e-01 -2.01791469e-02
1.30930114e+00 -3.45935971e-01 7.55680874e-02 3.57992381e-01
1.37176108e+00 6.76145032e-02 1.11636674e+00 6.92272127e-01
6.62782073e-01 6.57207966e-01 4.58343208e-01 4.32404786e-01
7.52409220e-01 3.65567327e-01 7.63347089e-01 1.05044052e-01
-5.92144608e-01 -6.02759063e-01 1.00459587e-02 -7.16445073e-02
-2.64446676e-01 -2.56251186e-01 -1.39251053e+00 9.31310534e-01
-1.89241910e+00 -1.73685765e+00 -3.12675029e-01 1.71396494e+00
6.77992582e-01 2.57603787e-02 -2.48877984e-02 -2.09510133e-01
7.17530668e-01 2.48731196e-01 -7.07664728e-01 -7.16761112e-01
-5.01772940e-01 -1.10990860e-01 3.33088458e-01 2.89121687e-01
-7.75127769e-01 9.40022886e-01 5.97435093e+00 6.62370741e-01
-7.74733305e-01 1.13694564e-01 8.57524872e-01 -2.37605542e-01
-7.66572058e-01 1.01227574e-01 -4.92945500e-02 4.74364638e-01
5.18069625e-01 -3.59647632e-01 5.97954333e-01 9.16930318e-01
5.12212440e-02 -3.59898955e-01 -9.71810579e-01 1.36338127e+00
8.23781490e-01 -1.57211828e+00 2.59893298e-01 -3.65891397e-01
7.07543850e-01 -2.34052703e-01 3.65576416e-01 3.99041682e-01
4.62855965e-01 -1.53244078e+00 1.16270888e+00 7.74121404e-01
7.00869977e-01 -5.75369120e-01 2.18189627e-01 8.95396024e-02
-3.57228965e-01 -1.84362233e-01 -2.22902209e-01 -2.97228873e-01
1.12490684e-01 1.58999875e-01 -6.52074397e-01 -2.06398278e-01
8.46101642e-01 4.04464424e-01 -9.95859265e-01 8.93133998e-01
-5.16075552e-01 8.13347697e-02 1.41791195e-01 -1.32228136e-01
1.63267732e-01 3.43242258e-01 6.98805451e-01 1.00968063e+00
2.12325528e-01 2.30811134e-01 -5.09119093e-01 1.64852405e+00
-1.62894741e-01 -3.02981228e-01 -6.61514103e-01 -3.42866898e-01
3.63537073e-01 9.54544604e-01 -4.88297135e-01 -5.10500550e-01
-6.90432359e-03 1.25652564e+00 2.75221020e-01 2.71564752e-01
-1.32907212e+00 -4.02490348e-01 6.09613657e-01 4.34194833e-01
-5.33660986e-02 1.84817120e-01 -5.85774183e-01 -1.23785877e+00
-1.77236557e-01 -1.38161826e+00 3.69012833e-01 -2.08832979e+00
-1.43663704e+00 5.77774346e-01 3.92391048e-02 -8.75306249e-01
-3.24190557e-01 -7.53151298e-01 -6.12243295e-01 5.44313133e-01
-1.15384972e+00 -1.40352619e+00 -8.36432159e-01 5.59279203e-01
6.62714720e-01 7.91429877e-02 4.29271698e-01 -1.86495006e-01
-1.76188834e-02 2.77940929e-01 -5.50243318e-01 2.67040581e-02
7.75698304e-01 -1.21445835e+00 6.64902687e-01 7.35801101e-01
2.64915556e-01 3.85270476e-01 1.39670742e+00 -6.34167910e-01
-9.82182384e-01 -5.73368430e-01 8.35489750e-01 -9.33542728e-01
9.90213811e-01 -1.19637214e-01 -7.75571823e-01 8.03324223e-01
5.59364021e-01 -4.81196865e-02 4.47046489e-01 -1.61419123e-01
-9.09411192e-01 4.68245059e-01 -1.09017515e+00 9.82834637e-01
1.18615186e+00 -6.24636710e-01 -1.08731794e+00 5.79118252e-01
7.03092754e-01 -3.24370116e-01 -2.09068149e-01 -9.35434550e-02
8.15614879e-01 -1.26170647e+00 1.00662756e+00 -1.10868812e+00
1.24851871e+00 -3.16464692e-01 -1.32212222e-01 -1.28256702e+00
-2.42714152e-01 -6.07942760e-01 2.92368889e-01 8.59185398e-01
4.37015712e-01 -9.32718962e-02 6.61979556e-01 1.10479665e+00
8.75436980e-03 -7.96423927e-02 -8.84654522e-01 -6.89419329e-01
-9.79691967e-02 -6.31399870e-01 6.19086027e-01 1.11323857e+00
2.67999798e-01 2.72778332e-01 -4.48249519e-01 -2.48888135e-01
7.24503219e-01 2.41456524e-01 9.67163205e-01 -7.12238729e-01
-4.65002954e-01 -5.06362259e-01 -6.52495265e-01 -5.34758925e-01
6.73544854e-02 -8.34793389e-01 -9.26009938e-03 -1.45371556e+00
6.76924109e-01 1.61716435e-02 -5.38299680e-02 5.88651896e-01
-2.14546934e-01 7.53075302e-01 8.94599020e-01 -2.09260993e-02
-8.64836037e-01 1.19219825e-01 1.66683555e+00 -4.24685568e-01
3.81031394e-01 -9.24441040e-01 -1.17934799e+00 7.94024706e-01
8.10093522e-01 -3.06356698e-01 -3.03593695e-01 -5.44417500e-01
8.46821725e-01 -8.99117813e-02 1.35914469e+00 -9.04002249e-01
6.26019016e-02 -5.69957554e-01 4.98004645e-01 -2.24973336e-01
4.36350435e-01 -5.87655544e-01 4.21605229e-01 3.82960677e-01
-4.27475333e-01 2.93940157e-01 4.92285073e-01 2.96443075e-01
-5.57107106e-02 -7.39354864e-02 7.80115306e-01 -4.43153143e-01
-1.23038888e+00 -2.54948825e-01 -9.23945606e-02 4.92699116e-01
1.13885081e+00 -4.14807379e-01 -1.19299757e+00 -8.20405483e-01
-6.67616725e-01 3.75479385e-02 7.42499888e-01 7.23208308e-01
8.57729971e-01 -1.34642160e+00 -9.70956743e-01 -2.53874511e-01
6.96843982e-01 -4.91477579e-01 7.52014697e-01 4.26517457e-01
-8.20631862e-01 8.61104652e-02 -8.48858118e-01 -2.34028980e-01
-1.17230713e+00 9.12043810e-01 2.39088520e-01 2.25526094e-01
-4.61699039e-01 1.00911784e+00 5.22923589e-01 -1.11437544e-01
-4.49250340e-01 -2.02087462e-01 1.87253103e-01 -2.27728918e-01
7.69684553e-01 2.43185848e-01 -6.44217372e-01 -8.62752259e-01
-4.41779047e-01 2.75151748e-02 2.77518690e-01 -1.18609436e-01
1.05513597e+00 -7.01594353e-02 -6.83654249e-02 3.88725191e-01
7.64867425e-01 -2.80094445e-01 -1.02199614e+00 3.96593124e-01
-3.46618354e-01 -8.86143565e-01 -3.47955614e-01 -1.52733815e+00
-6.85141206e-01 9.48243558e-01 3.11442494e-01 2.25625440e-01
9.68084514e-01 3.91823649e-01 7.53298759e-01 1.90953851e-01
1.49199381e-01 -9.18248832e-01 4.47395563e-01 2.05565438e-01
1.73047662e+00 -1.43238676e+00 -2.26192195e-02 4.12472710e-02
-1.38364434e+00 7.34791756e-01 9.89919424e-01 -2.11485460e-01
-1.94037750e-01 -2.61120290e-01 6.39172792e-02 -5.21540761e-01
-9.45312142e-01 -2.85006285e-01 4.44868743e-01 9.16309774e-01
1.90060645e-01 2.56923974e-01 8.48936811e-02 7.16226459e-01
-6.36743069e-01 -6.09041229e-02 7.56131351e-01 4.57023025e-01
-5.46811700e-01 -3.42694938e-01 -7.77069688e-01 2.39015847e-01
-6.34643063e-02 -3.42693150e-01 -8.52433801e-01 1.10791183e+00
5.23259640e-01 7.64698625e-01 1.93562791e-01 -3.16234261e-01
4.20952082e-01 1.15188649e-02 6.73664033e-01 -2.84241676e-01
-5.29055059e-01 -9.75430787e-01 3.62496018e-01 -6.69383764e-01
-2.16027036e-01 -4.64944690e-01 -1.13903344e+00 -7.53435314e-01
2.69293845e-01 -2.22869501e-01 6.95498168e-01 5.33179820e-01
4.05198514e-01 2.51964122e-01 -1.79340988e-01 -7.09365785e-01
-8.12557042e-02 -6.91502929e-01 -2.29780748e-01 1.20583546e+00
1.89558163e-01 -4.83999521e-01 -4.35938179e-01 4.76933271e-01] | [10.780535697937012, 1.8131250143051147] |
d5de4298-f5f5-4eb5-a4a0-9d970992dcd9 | chitransformer-towards-reliable-stereo-from-1 | null | null | http://openaccess.thecvf.com//content/CVPR2022/html/Su_Chitransformer_Towards_Reliable_Stereo_From_Cues_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Su_Chitransformer_Towards_Reliable_Stereo_From_Cues_CVPR_2022_paper.pdf | Chitransformer: Towards Reliable Stereo From Cues | Current stereo matching techniques are challenged by restricted searching space, occluded regions and sheer size. While single image depth estimation is spared from these challenges and can achieve satisfactory results with the extracted monocular cues, the lack of stereoscopic relationship renders the monocular prediction less reliable on its own, especially in highly dynamic or cluttered environments. To address these issues in both scenarios, we present an optic-chiasm-inspired self-supervised binocular depth estimation method, wherein a vision transformer (ViT) with gated positional cross-attention (GPCA) layers is designed to enable feature-sensitive pattern retrieval between views while retaining the extensive context information aggregated through self-attentions. Monocular cues from a single view are thereafter conditionally rectified by a blending layer with the retrieved pattern pairs. This crossover design is biologically analogous to the optic-chasma structure in the human visual system and hence the name, ChiTransformer. Our experiments show that this architecture yields substantial improvements over state-of-the-art self-supervised stereo approaches by 11%, and can be used on both rectilinear and non-rectilinear (e.g., fisheye) images. | ['Shihao Ji', 'Qing Su'] | 2022-01-01 | null | null | null | cvpr-2022-1 | ['stereo-depth-estimation'] | ['computer-vision'] | [ 4.75512058e-01 5.50203770e-02 -6.16178960e-02 -3.26909125e-01
-3.72872204e-01 -5.54243267e-01 6.30971611e-01 -5.10517955e-01
-3.33644152e-01 6.85095310e-01 2.53560632e-01 4.23905589e-02
-5.33306971e-04 -5.00461042e-01 -6.64225698e-01 -9.17466283e-01
5.42731404e-01 2.49131098e-02 6.28726006e-01 1.06026260e-02
7.13941753e-01 4.30257946e-01 -2.31142950e+00 4.98307616e-01
7.80312359e-01 1.23641801e+00 7.68490374e-01 5.63362777e-01
1.50297850e-01 8.12070727e-01 1.29036782e-02 -2.76667744e-01
5.74620008e-01 -3.81987631e-01 -5.65257132e-01 1.79643005e-01
1.09152639e+00 -4.83809531e-01 -7.45285630e-01 8.32378685e-01
5.78636587e-01 -1.67406291e-01 5.62484443e-01 -9.24554646e-01
-4.34694439e-01 -3.04841876e-01 -6.30335450e-01 3.24635208e-01
6.03688300e-01 2.65982687e-01 8.53242636e-01 -1.13770998e+00
8.76666963e-01 1.00734210e+00 2.57724136e-01 5.54056704e-01
-1.15834868e+00 -4.25435424e-01 1.31999508e-01 3.23394835e-01
-1.21651697e+00 -6.98302209e-01 8.31551015e-01 -3.75904799e-01
1.34294963e+00 1.40262455e-01 9.18416917e-01 7.30041504e-01
4.36330348e-01 8.46246839e-01 1.21045530e+00 -3.59030485e-01
8.34919065e-02 2.31968373e-01 -3.83591712e-01 6.75786912e-01
1.48654640e-01 7.06708074e-01 -1.09702766e+00 3.35208207e-01
1.06387186e+00 1.87645689e-01 -6.96850717e-01 -1.05089378e+00
-1.19179940e+00 4.76117253e-01 6.76877975e-01 2.21899241e-01
-2.12158620e-01 -2.03644484e-01 -7.35559836e-02 3.54212075e-01
2.79266357e-01 5.52157998e-01 -3.43993068e-01 1.85476497e-01
-9.96177912e-01 1.24442272e-01 3.40992242e-01 1.20061839e+00
9.52153921e-01 1.33242056e-01 1.25348434e-01 7.69627213e-01
2.29903564e-01 2.83726782e-01 6.01619840e-01 -1.30664659e+00
3.12201351e-01 5.70298851e-01 4.51632440e-02 -8.18450809e-01
-4.09299135e-01 -3.48815084e-01 -6.68705583e-01 6.82482600e-01
4.22178686e-01 3.90678853e-01 -7.96583593e-01 1.50188243e+00
1.87454805e-01 -2.78179914e-01 -1.29763754e-02 1.26378894e+00
9.08082068e-01 1.51203051e-01 -4.45991606e-01 -1.43627152e-01
1.13462639e+00 -1.11448550e+00 -2.97860622e-01 -6.49458528e-01
1.60155982e-01 -7.40359783e-01 6.99408650e-01 5.36602378e-01
-1.35918069e+00 -6.88101590e-01 -9.69740629e-01 -6.16704047e-01
-2.61712492e-01 -1.80751264e-01 5.91603220e-01 3.73717278e-01
-1.38185847e+00 3.47337514e-01 -4.47564512e-01 -6.29953921e-01
4.41655427e-01 4.59206671e-01 -7.18339741e-01 -3.20027620e-01
-7.53291667e-01 8.56934369e-01 1.04572646e-01 -1.01614177e-01
-5.97297072e-01 -5.12462020e-01 -1.07754111e+00 -1.00379213e-04
1.41099662e-01 -1.19210827e+00 8.11884403e-01 -1.02296150e+00
-1.41796482e+00 1.35747814e+00 -6.16475224e-01 -3.64081591e-01
3.17910284e-01 -8.79553854e-02 8.95679593e-02 4.22782809e-01
9.30465683e-02 1.30392492e+00 1.03721321e+00 -1.34490502e+00
-9.31933880e-01 -8.54474962e-01 2.78441585e-03 8.43238175e-01
4.92512621e-02 -1.24280684e-01 -7.24833071e-01 -3.75355929e-01
6.26846313e-01 -7.66834736e-01 2.30766721e-02 4.84678954e-01
-3.80938430e-03 1.84155449e-01 6.88594103e-01 -4.01721984e-01
9.23999369e-01 -1.99505126e+00 2.99625158e-01 -2.70694822e-01
3.11879903e-01 3.79787311e-02 1.04393706e-01 2.74505615e-01
-9.31137428e-02 -6.59466863e-01 -1.41141370e-01 -4.19583946e-01
-5.96924901e-01 1.63430765e-01 -2.31835470e-01 6.17352605e-01
2.32869871e-02 9.04551089e-01 -9.32739496e-01 -4.15161401e-01
6.34469390e-01 2.71193355e-01 -7.51004875e-01 4.31620806e-01
1.07123964e-01 7.35351026e-01 -1.88378468e-02 9.33994055e-01
6.95631623e-01 -2.25553274e-01 4.56298050e-03 -2.21358269e-01
-4.19608802e-01 1.66965216e-01 -8.13776314e-01 2.08011842e+00
-3.74777257e-01 1.07804275e+00 8.93175229e-02 -6.20672107e-01
8.62132013e-01 -1.00009911e-01 1.77262306e-01 -1.27140856e+00
2.86061671e-02 3.40050668e-01 -1.64060518e-01 -3.62770855e-01
5.25371611e-01 -4.17629853e-02 3.70685786e-01 1.82560563e-01
1.96479604e-01 -5.17162383e-01 -1.34299815e-01 -1.08502936e-02
8.39911580e-01 5.29547513e-01 5.49355984e-01 -1.23654880e-01
6.14032328e-01 -1.21633157e-01 5.05687833e-01 4.84875709e-01
-2.96122879e-01 1.28041828e+00 -4.17347997e-02 -6.11937702e-01
-1.17890131e+00 -9.62376297e-01 -2.07295373e-01 7.84903824e-01
6.72477305e-01 -1.52856857e-01 -3.62664610e-01 -2.47819856e-01
7.70895109e-02 2.37438112e-01 -4.58447069e-01 -2.84050871e-02
-3.82690996e-01 -9.07423571e-02 -4.43308577e-02 6.21513486e-01
6.99906230e-01 -1.16508675e+00 -1.16614723e+00 5.53484894e-02
-1.81455642e-01 -1.09931064e+00 -3.46116066e-01 2.58322001e-01
-7.83189952e-01 -1.06764948e+00 -9.66781795e-01 -9.81165230e-01
6.29270673e-01 9.94477093e-01 1.15012097e+00 -3.33984941e-01
-4.24811214e-01 4.16969717e-01 5.54738156e-02 -1.17140621e-01
2.81084567e-01 -3.20742309e-01 -9.01480466e-02 -4.60666604e-02
5.56787312e-01 -7.73891151e-01 -1.28473723e+00 6.44049823e-01
-6.79639995e-01 4.34891492e-01 5.73937297e-01 1.15958798e+00
5.87844014e-01 -3.41241032e-01 6.09129071e-02 -5.14369845e-01
-2.19351903e-01 -2.45905086e-01 -8.67775023e-01 9.68334172e-03
-5.92327416e-01 -3.48059498e-02 4.26986307e-01 -1.21942148e-01
-1.25787532e+00 2.58909911e-01 2.98718214e-01 -5.41562259e-01
-2.31699169e-01 -2.18863696e-01 -1.43722564e-01 -5.53423643e-01
6.21681929e-01 4.94259745e-01 1.98079571e-02 -1.64530113e-01
1.08190335e-01 6.77026570e-01 8.37179005e-01 6.02067374e-02
5.79888523e-01 9.81004238e-01 -1.20783253e-02 -7.06852078e-01
-7.92727947e-01 -8.63631845e-01 -8.59189749e-01 -2.92712152e-01
8.47965360e-01 -1.14891171e+00 -7.58235335e-01 6.76057935e-01
-1.00798833e+00 -1.88625276e-01 -9.89274532e-02 4.91965771e-01
-1.00898659e+00 4.73991811e-01 -2.18665689e-01 -8.53819013e-01
-1.50115818e-01 -8.92132759e-01 1.36718833e+00 3.56366009e-01
3.67511855e-03 -7.51810908e-01 4.43991534e-02 6.14578545e-01
4.12995458e-01 -1.70630008e-01 6.62536263e-01 1.45858098e-02
-1.12412429e+00 1.66436672e-01 -6.03411436e-01 2.09678233e-01
-5.60102016e-02 -4.34267312e-01 -1.50053728e+00 -2.70300090e-01
2.03947574e-01 -4.97838616e-01 1.06983793e+00 5.48461914e-01
1.02799845e+00 8.89359862e-02 -3.64309937e-01 1.04360783e+00
1.59321964e+00 2.68063575e-01 8.78550410e-01 4.35945719e-01
6.01121843e-01 1.03879738e+00 6.82851970e-01 2.75570959e-01
6.03250444e-01 9.03851807e-01 6.19707406e-01 -2.87657499e-01
-3.90468717e-01 -3.32026511e-01 -2.77048964e-02 3.55997056e-01
-3.26382890e-02 -6.46714345e-02 -7.02561915e-01 6.06299579e-01
-1.68387246e+00 -1.07542586e+00 2.84815669e-01 2.40941572e+00
7.09992349e-01 2.24628393e-02 -3.55350137e-01 8.30658004e-02
4.30100262e-01 2.20173344e-01 -7.80194521e-01 -2.72555679e-01
-6.70664310e-01 -6.06343336e-02 5.26284337e-01 4.98511791e-01
-9.30045128e-01 1.06204784e+00 6.54271030e+00 4.23065335e-01
-1.09501076e+00 -3.30515891e-01 4.67584401e-01 -1.76428437e-01
-3.34852189e-01 9.92603004e-02 -6.67289317e-01 2.67348975e-01
2.14444503e-01 1.31565377e-01 5.42069316e-01 6.60717726e-01
-4.62199636e-02 -7.47107923e-01 -1.29661453e+00 1.54564643e+00
5.28827846e-01 -1.18536711e+00 -1.23254955e-01 2.28216767e-01
1.04784191e+00 2.58077234e-01 2.15215161e-01 -2.17350185e-01
-1.12503961e-01 -9.77905869e-01 4.82772440e-01 5.13141394e-01
9.48283195e-01 -3.32671613e-01 4.73897099e-01 2.46004462e-01
-9.73460495e-01 -4.97126281e-01 -6.48245990e-01 -1.49730980e-01
-6.73001632e-02 2.12655157e-01 -3.25823337e-01 4.70120937e-01
9.45050597e-01 9.63574767e-01 -4.44901854e-01 1.36508262e+00
9.36503559e-02 -3.15396637e-01 -1.83584526e-01 3.90078217e-01
1.14103191e-01 -1.59590840e-01 6.17047846e-01 6.84090018e-01
1.77179858e-01 5.73629793e-03 -5.04548788e-01 7.14631617e-01
2.68905908e-01 -1.62890792e-01 -1.10898077e+00 6.10190034e-01
3.26592922e-01 9.86626208e-01 -3.10730040e-01 -1.47996634e-01
-7.22993076e-01 1.23361576e+00 4.55897570e-01 5.10617793e-01
-1.88468426e-01 -9.33434367e-02 5.03860414e-01 4.19229090e-01
7.15386510e-01 5.43349758e-02 -3.92806977e-01 -1.41911638e+00
1.78400770e-01 -6.54925406e-01 2.32537702e-01 -1.41003013e+00
-9.79893625e-01 7.61865914e-01 -3.28211218e-01 -1.66693354e+00
-4.78931695e-01 -7.57288277e-01 -3.42306972e-01 8.95690799e-01
-2.11469984e+00 -1.06252980e+00 -7.57493258e-01 8.65269423e-01
6.59660041e-01 -3.50657821e-01 7.04338372e-01 1.00728013e-02
1.09924458e-01 4.72357601e-01 6.52735755e-02 -4.71974909e-01
9.97146547e-01 -1.17685640e+00 1.42870560e-01 7.76448369e-01
7.04603642e-02 4.49412376e-01 5.69725513e-01 -3.07025135e-01
-1.45232999e+00 -6.94547117e-01 9.50805783e-01 -4.87095028e-01
1.96488768e-01 -3.49026918e-01 -8.59164834e-01 2.46448308e-01
2.08586842e-01 2.20888719e-01 3.65197003e-01 -2.76249528e-01
-5.24398625e-01 -3.46436888e-01 -1.03049242e+00 7.07261980e-01
1.44485271e+00 -6.58706605e-01 -6.43534899e-01 -5.36512174e-02
3.12801421e-01 -3.99586797e-01 -4.92549151e-01 4.05377626e-01
9.35752630e-01 -1.80843782e+00 8.59544575e-01 -1.57761514e-01
4.63680565e-01 -3.87815475e-01 -2.46866524e-01 -9.22635257e-01
-3.15766394e-01 -6.47470891e-01 6.13472797e-02 6.86521113e-01
7.70283192e-02 -6.56042397e-01 9.87235665e-01 5.24789929e-01
-1.63230672e-01 -6.25790477e-01 -1.18809736e+00 -4.65616763e-01
-3.25620949e-01 -1.28054991e-01 1.31162062e-01 5.67006886e-01
1.74541727e-01 2.56440550e-01 -3.73198181e-01 -1.68828666e-01
8.17810655e-01 5.18385291e-01 8.09288442e-01 -1.14432466e+00
-4.10111547e-01 -3.43137681e-01 -8.30639005e-01 -1.57215834e+00
-9.82176811e-02 -4.59861666e-01 3.58579233e-02 -1.27745938e+00
2.53719538e-01 -1.20597064e-01 -2.88199503e-02 5.34677394e-02
7.26964325e-02 5.67661703e-01 9.87639353e-02 4.84750062e-01
-4.54484552e-01 6.49656057e-01 1.45558083e+00 1.62423223e-01
-1.48768306e-01 -8.20150748e-02 -6.10064924e-01 7.08675385e-01
4.02929276e-01 4.97344648e-03 -6.25728965e-01 -5.87099314e-01
1.16043113e-01 3.14519346e-01 4.49981183e-01 -1.06786001e+00
7.96809971e-01 1.52820438e-01 5.85009336e-01 -8.40380907e-01
6.64650440e-01 -8.92634869e-01 -3.38634625e-02 2.19856516e-01
-6.38600290e-02 -8.17011520e-02 3.15936506e-02 6.34325087e-01
-5.83427131e-01 1.48490965e-01 8.24433804e-01 -2.91117609e-01
-1.04037440e+00 2.99314946e-01 -1.80650428e-01 -9.74225849e-02
8.46430838e-01 -1.09754825e+00 -3.89971942e-01 -4.91730571e-01
-4.35989261e-01 9.75972116e-02 9.99354482e-01 2.78344303e-01
1.04392970e+00 -1.11003232e+00 -4.26967233e-01 8.64258826e-01
5.30002952e-01 1.85570493e-01 5.10619462e-01 7.65568078e-01
-4.18196976e-01 1.02718520e+00 -6.37306869e-01 -9.00079131e-01
-1.28450751e+00 6.61125541e-01 4.46885049e-01 6.10288270e-02
-6.77842855e-01 7.88117349e-01 1.10120821e+00 -9.26841050e-02
3.24587643e-01 -1.14904366e-01 -3.66608053e-01 -2.20447391e-01
5.02107441e-01 1.29312336e-01 1.25736371e-03 -5.87603092e-01
-1.85516030e-01 1.16098702e+00 -3.61066908e-02 -6.11223169e-02
1.16460872e+00 -7.30947852e-01 1.91261619e-02 2.52121001e-01
1.06035328e+00 -1.06101885e-01 -1.78783655e+00 -6.20181143e-01
-5.46486437e-01 -9.87801850e-01 1.36138320e-01 -5.77904463e-01
-9.46850538e-01 1.06790471e+00 5.96786916e-01 -1.44396588e-01
1.43697059e+00 9.48671773e-02 5.00755012e-01 1.33718878e-01
6.45288587e-01 -8.38360965e-01 5.97098805e-02 3.60384345e-01
1.04870343e+00 -1.55836105e+00 -1.20812364e-01 -5.42085826e-01
-8.29447210e-01 9.83528733e-01 9.04626489e-01 -6.73767701e-02
5.01739800e-01 -1.61716621e-02 -4.51229438e-02 -2.45507374e-01
-9.98691976e-01 -4.91427630e-01 4.48667735e-01 1.02200055e+00
2.50382543e-01 -6.05264783e-01 1.27490595e-01 -1.21756591e-01
-9.50056911e-02 -2.05441162e-01 3.29003632e-01 8.68730068e-01
-5.24672151e-01 -6.01407349e-01 -1.71696156e-01 2.14520693e-01
-1.21131070e-01 -3.35331053e-01 -2.98823267e-01 6.50222361e-01
4.81757224e-02 8.81934643e-01 4.06000435e-01 -2.51996756e-01
1.90359294e-01 -2.40793169e-01 7.92251527e-01 -6.26811504e-01
-3.99361521e-01 2.88659036e-01 -6.37026504e-02 -9.36856449e-01
-7.35487640e-01 -5.36124229e-01 -6.12607062e-01 -2.17339974e-02
-2.21148118e-01 -4.60111409e-01 3.99173647e-01 6.62871122e-01
4.66502398e-01 3.22193615e-02 7.04099476e-01 -1.14097619e+00
2.14595553e-02 -6.60272241e-01 -5.69241822e-01 1.77086398e-01
4.42404628e-01 -8.52451622e-01 -4.96897280e-01 4.45798114e-02] | [8.766242980957031, -2.4149832725524902] |
7f4259d2-313f-48a3-a2db-c98adf08c204 | weakly-supervised-affordance-detection | null | null | http://openaccess.thecvf.com/content_cvpr_2017/html/Sawatzky_Weakly_Supervised_Affordance_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Sawatzky_Weakly_Supervised_Affordance_CVPR_2017_paper.pdf | Weakly Supervised Affordance Detection | Localizing functional regions of objects or affordances is an important aspect of scene understanding and relevant for many robotics applications. In this work, we introduce a pixel-wise annotated affordance dataset of 3090 images containing 9916 object instances. Since parts of an object can have multiple affordances, we address this by a convo- lutional neural network for multilabel affordance segmen- tation. We also propose an approach to train the network from very few keypoint annotations. Our approach achieves a higher affordance detection accuracy than other weakly supervised methods that also rely on keypoint annotations or image annotations as weak supervision.
| ['Johann Sawatzky', 'Abhilash Srikantha', 'Juergen Gall'] | 2017-07-01 | null | null | null | cvpr-2017-7 | ['affordance-detection'] | ['computer-vision'] | [ 4.47523445e-02 1.93499073e-01 -5.94496787e-01 -5.44184983e-01
-3.52923542e-01 -4.60520893e-01 5.31390131e-01 3.69629264e-01
-7.04994082e-01 6.47304237e-01 3.26938093e-01 -1.45037860e-01
-7.52628893e-02 -4.39793676e-01 -1.15092862e+00 -3.29894602e-01
-1.48225725e-01 2.41138294e-01 5.86767137e-01 -1.69460267e-01
2.84755349e-01 8.55508745e-01 -1.35218215e+00 1.11663714e-01
5.98662734e-01 1.00393403e+00 4.76590276e-01 6.50810659e-01
3.32233161e-01 8.08872402e-01 -3.33669931e-01 1.55669734e-01
4.12338793e-01 8.85565430e-02 -8.94871533e-01 5.52147925e-02
1.04967678e+00 -6.37024045e-01 -4.84394997e-01 1.19412351e+00
1.01203777e-01 2.24060297e-01 7.22865582e-01 -1.40018380e+00
-7.85911500e-01 4.89386797e-01 -3.84504855e-01 3.77976179e-01
5.58797359e-01 -1.55618871e-02 1.45859385e+00 -8.60873640e-01
7.67963707e-01 1.40934503e+00 1.63654923e-01 3.73992831e-01
-1.31701708e+00 -1.78385913e-01 5.76944768e-01 1.91834003e-01
-1.36210430e+00 -2.17299283e-01 7.32689559e-01 -2.81195492e-01
1.25777352e+00 2.54147410e-01 6.54525936e-01 7.20726788e-01
-3.06052342e-02 1.35055995e+00 8.23961198e-01 -4.34202373e-01
-7.24651217e-02 -4.63197947e-01 2.74395555e-01 1.09544241e+00
1.14734545e-01 -9.06806737e-02 -3.60955834e-01 -1.40065402e-01
1.46286809e+00 5.04023314e-01 -2.42014840e-01 -8.02726686e-01
-1.91262746e+00 5.02348721e-01 1.49190295e+00 1.85507521e-01
-8.59637037e-02 6.26398385e-01 5.99024184e-02 1.23127736e-01
2.04517692e-01 8.96525562e-01 -5.12566268e-01 4.58709478e-01
-2.32423648e-01 3.22921813e-01 5.35160601e-01 1.16300642e+00
9.58693385e-01 -3.74652982e-01 -1.09213553e-01 5.84624171e-01
4.00632083e-01 1.56505674e-01 2.95658976e-01 -1.09276319e+00
3.45060527e-01 8.47400427e-01 1.98510826e-01 -6.61938548e-01
-6.52953267e-01 3.55900198e-01 -3.75517398e-01 2.64569163e-01
4.16867405e-01 3.07766050e-01 -1.17383301e+00 1.44634175e+00
3.06481183e-01 4.31994423e-02 -1.67345271e-01 1.34786093e+00
8.65263522e-01 2.57243812e-01 2.12581649e-01 4.28405941e-01
1.34430778e+00 -1.32900047e+00 -4.45666164e-01 -2.05843702e-01
5.44157803e-01 -4.99052018e-01 1.38977098e+00 9.10633430e-02
-4.40711081e-01 -4.62988973e-01 -1.24072194e+00 -7.60013223e-01
-4.36536282e-01 3.74356240e-01 1.22296476e+00 2.31074989e-02
-8.33437324e-01 4.03852731e-01 -9.39664483e-01 -3.99266928e-01
6.78773165e-01 4.23754245e-01 -7.74208605e-01 -5.31761348e-02
-5.46286166e-01 1.02448964e+00 6.61648273e-01 -3.31656896e-02
-1.08918297e+00 -3.80911797e-01 -1.41596997e+00 -1.41200781e-01
7.42731571e-01 -5.90066850e-01 1.15308595e+00 -7.17998743e-01
-1.05679440e+00 7.52588212e-01 -1.34451702e-01 -2.71343499e-01
2.80543178e-01 -5.20453751e-01 5.58490381e-02 1.35365754e-01
3.30875635e-01 1.33559144e+00 7.29720294e-01 -1.20307028e+00
-5.70793331e-01 -4.26190585e-01 7.45846391e-01 3.54664087e-01
5.98243214e-02 1.02341756e-01 -1.64965689e-01 -7.76875079e-01
7.48942912e-01 -8.17941129e-01 -4.36463356e-01 4.85717118e-01
-8.47876251e-01 -8.03404450e-01 9.86335158e-01 -1.14937529e-01
5.05762815e-01 -1.93534338e+00 3.13264459e-01 -2.24446535e-01
2.69665152e-01 -2.88028330e-01 -2.52777517e-01 8.90202597e-02
4.00328822e-02 7.86683697e-04 -2.36694098e-01 -2.11871356e-01
1.96636438e-01 6.20246410e-01 -2.88060129e-01 5.91063261e-01
4.88015801e-01 1.21933329e+00 -1.15836310e+00 -4.55799878e-01
4.60527807e-01 1.88969880e-01 -6.29614472e-01 3.24952066e-01
-5.01447976e-01 4.83323574e-01 -5.68111897e-01 1.11340022e+00
1.22946873e-01 -6.76849112e-02 -2.34832734e-01 -3.97328645e-01
-4.60281707e-02 3.55349809e-01 -1.04386222e+00 2.30415201e+00
-1.99704826e-01 6.89402163e-01 -2.96525210e-01 -8.74655664e-01
6.11275852e-01 5.56281321e-02 3.61811608e-01 -1.43254593e-01
1.13252558e-01 2.22509295e-01 6.57726377e-02 -5.77611744e-01
6.36936724e-01 1.48644909e-01 -3.21249068e-01 3.98595601e-01
5.71987867e-01 -6.21823549e-01 1.29003093e-01 7.53724203e-02
8.08237076e-01 6.23070061e-01 5.98505437e-01 -4.30943429e-01
2.19792783e-01 -6.30753264e-02 3.62413168e-01 7.00532913e-01
-3.91612440e-01 6.41367912e-01 4.77375865e-01 -9.41074133e-01
-1.22299814e+00 -1.03352737e+00 -3.39480311e-01 1.25587404e+00
5.33515036e-01 -3.72770041e-01 -5.11411428e-01 -8.80479932e-01
1.36395231e-01 1.15457214e-01 -6.45865619e-01 6.33800179e-02
-6.07531846e-01 -1.80672273e-01 2.32954130e-01 7.64775097e-01
4.41298455e-01 -1.26768363e+00 -1.08428061e+00 -2.24214599e-01
7.74705559e-02 -1.37357926e+00 -7.57110834e-01 4.89948720e-01
-7.14972198e-01 -1.16372108e+00 -7.29708910e-01 -9.68215942e-01
1.02699089e+00 3.10973316e-01 8.53776693e-01 1.94076419e-01
-4.91846889e-01 2.68929005e-01 -3.53697717e-01 -3.90392393e-01
1.30630970e-01 2.69301295e-01 3.64139289e-01 -2.97346354e-01
4.95985925e-01 -2.33291566e-01 -4.33075726e-01 2.89781243e-01
-8.67376029e-01 9.10805315e-02 5.60486376e-01 8.01540852e-01
1.02531993e+00 -5.37762225e-01 1.90530643e-01 -4.26998466e-01
3.05637062e-01 -1.89029500e-02 -5.56765556e-01 2.43765816e-01
8.06577355e-02 2.64925987e-01 5.15099913e-02 -5.15174389e-01
-5.83809555e-01 6.28959656e-01 -8.00231248e-02 -4.83217180e-01
-6.80743098e-01 1.84906051e-01 -3.61650378e-01 -2.84588546e-01
7.66675949e-01 -4.61949021e-01 -4.20158029e-01 -4.94634062e-01
7.43077934e-01 4.17582214e-01 5.22767782e-01 -7.12952614e-01
4.92366046e-01 5.52486897e-01 4.84135561e-02 -8.78828704e-01
-1.27512920e+00 -6.95721149e-01 -1.53083837e+00 -1.10283248e-01
1.14136267e+00 -8.19058418e-01 -6.93602264e-01 1.33602992e-01
-1.21105313e+00 -5.92830837e-01 -4.82574463e-01 7.46019065e-01
-7.80880570e-01 8.52865800e-02 -4.47200567e-01 -3.36563796e-01
1.42013252e-01 -1.36994672e+00 1.60470235e+00 1.17244832e-01
-2.52221018e-01 -5.87733150e-01 -3.19213152e-01 -6.05949312e-02
-3.30016643e-01 5.43291688e-01 4.83327985e-01 -4.65755701e-01
-8.80547941e-01 -2.31328219e-01 -6.78627968e-01 2.78848335e-02
3.97642463e-01 -4.80726302e-01 -9.11479890e-01 -8.68066624e-02
-2.48105511e-01 -7.61618257e-01 1.04055250e+00 2.51708388e-01
1.13286459e+00 -1.69840187e-01 -3.53680104e-01 5.59982121e-01
1.24356234e+00 -3.93095195e-01 2.06763476e-01 3.02963257e-01
1.36897337e+00 5.69023371e-01 8.34185719e-01 6.13021180e-02
4.65982974e-01 4.29647923e-01 8.51808310e-01 -1.79014988e-02
-1.46218119e-02 -5.56269228e-01 -5.31314202e-02 4.58555892e-02
-3.03927779e-01 2.14326560e-01 -8.48994970e-01 4.98802722e-01
-2.02282834e+00 -4.12992895e-01 -1.21959932e-01 1.72503662e+00
7.10936785e-01 -4.59069526e-03 6.22710288e-02 -8.32419991e-02
3.37926328e-01 2.83049583e-01 -9.32898760e-01 -9.15186554e-02
-1.54158667e-01 -1.65022179e-01 6.74316823e-01 4.10767138e-01
-1.59256577e+00 1.32555509e+00 7.03507376e+00 2.13947535e-01
-7.99968362e-01 1.80754989e-01 2.27218375e-01 3.30789946e-02
-4.90035675e-02 -6.38313890e-02 -6.05251014e-01 -3.72511089e-01
1.39737159e-01 5.62775731e-01 3.89991641e-01 1.05766082e+00
-2.25872502e-01 -4.86733556e-01 -1.50032461e+00 8.56490970e-01
4.06822741e-01 -9.71293151e-01 1.12524770e-01 1.23236030e-01
8.95697176e-01 3.50074172e-01 -2.78388500e-01 -1.02503031e-01
3.86459172e-01 -1.01653838e+00 1.00574303e+00 5.38918674e-01
6.46552622e-01 -4.48345184e-01 3.69471103e-01 4.84923720e-01
-1.23628962e+00 -2.17240244e-01 -7.96436369e-01 -5.12305617e-01
1.63393050e-01 1.15486361e-01 -6.32350862e-01 1.90579146e-01
9.43977773e-01 1.02401674e+00 -5.49054623e-01 9.81092513e-01
-8.76982510e-01 -1.83972985e-01 -5.89074373e-01 -1.52363166e-01
3.41054231e-01 1.15214296e-01 5.02217293e-01 7.36349523e-01
-5.07115088e-02 1.62681431e-01 6.05735242e-01 1.08155310e+00
-2.31658146e-01 4.92830090e-02 -8.04792166e-01 9.11463052e-02
1.80951729e-01 1.21185613e+00 -1.00908852e+00 -4.69274580e-01
-4.17599678e-01 1.22787905e+00 7.11146057e-01 4.14047539e-01
-4.06532347e-01 -2.49622524e-01 8.03377748e-01 -1.85695618e-01
1.29227474e-01 -8.04159045e-01 -2.60515094e-01 -1.46898448e+00
-2.80418363e-03 -2.87615865e-01 2.70850360e-01 -1.09268522e+00
-1.02104783e+00 2.54811376e-01 2.57562727e-01 -9.89126623e-01
1.02924272e-01 -1.09066808e+00 -1.86507478e-01 6.31377697e-01
-1.70105195e+00 -1.66203642e+00 -2.73215115e-01 6.94910944e-01
6.35456622e-01 1.88892558e-01 9.34900820e-01 -1.20865077e-01
2.21589003e-02 1.55372545e-01 -6.99434042e-01 2.96467751e-01
6.58399343e-01 -1.63405859e+00 2.86593527e-01 7.24044383e-01
7.65004516e-01 8.85399759e-01 5.14335871e-01 -4.94730920e-01
-1.33262384e+00 -9.80182350e-01 6.73231006e-01 -9.32870448e-01
6.78716600e-01 -5.41768134e-01 -6.89016879e-01 1.16790247e+00
1.19514704e-01 7.69452214e-01 3.48504454e-01 1.94925472e-01
-4.82125759e-01 3.69944543e-01 -9.65219676e-01 7.54089892e-01
1.53393805e+00 -8.34463716e-01 -9.79135334e-01 5.49626410e-01
1.16374147e+00 -7.63838887e-01 -9.72914934e-01 5.29921114e-01
5.26306272e-01 -5.90350032e-01 1.14960146e+00 -6.72789931e-01
1.30041093e-01 -5.26801407e-01 -4.18029994e-01 -9.91743445e-01
-9.12566632e-02 -1.24199808e-01 -1.38597205e-01 4.75509912e-01
2.93318033e-01 -3.94810677e-01 4.34684575e-01 5.07099569e-01
-2.93632060e-01 -4.06900167e-01 -8.36141586e-01 -5.69727480e-01
-1.51534900e-01 -4.56828266e-01 7.40144789e-01 9.45494652e-01
2.08943710e-01 4.01929408e-01 -9.08648521e-02 2.40538776e-01
4.64809328e-01 2.81974077e-01 8.41273367e-01 -1.21494150e+00
1.48545206e-01 -3.93252879e-01 -7.68629849e-01 -1.48863757e+00
4.54451859e-01 -8.77304375e-01 4.30993378e-01 -1.51802599e+00
1.09365344e-01 -4.77948248e-01 -5.47935486e-01 1.10813272e+00
-2.38396361e-01 7.41215408e-01 1.13879122e-01 4.30434644e-01
-8.83288980e-01 4.23596323e-01 1.73814988e+00 -3.55759948e-01
-2.76828140e-01 -3.57108355e-01 -2.67983794e-01 8.91965628e-01
5.94219029e-01 -2.06385732e-01 -1.70424998e-01 -6.30008578e-01
1.04625627e-01 -3.68228137e-01 7.82852829e-01 -6.20088458e-01
9.77604836e-02 -2.10238799e-01 2.93064386e-01 -5.87186217e-01
3.81920159e-01 -8.69290113e-01 -6.40228271e-01 4.15494323e-01
-4.41461921e-01 -8.37361217e-02 -3.80735584e-02 6.02196336e-01
-1.95288360e-01 -1.69296950e-01 3.20869297e-01 -4.71539438e-01
-1.33266377e+00 6.29241705e-01 -3.99292707e-02 -5.28595567e-01
1.06836998e+00 4.76964079e-02 3.10593117e-02 1.72370344e-01
-8.82795691e-01 1.38356879e-01 6.93633676e-01 8.76481891e-01
6.94987714e-01 -1.41797745e+00 -2.49894619e-01 3.26496929e-01
6.41742229e-01 4.10991818e-01 -2.75929779e-01 6.00302339e-01
-5.93779922e-01 5.00667155e-01 -3.77671182e-01 -9.80487287e-01
-9.37204123e-01 6.97072089e-01 3.27303976e-01 4.60074544e-01
-7.98249304e-01 9.80444491e-01 3.20869803e-01 -1.09640598e+00
2.24404439e-01 -1.13150215e+00 -3.18611264e-01 -2.87062764e-01
2.29606181e-01 -5.01562096e-02 -3.92784357e-01 -1.12365222e+00
-1.75880685e-01 5.64407825e-01 3.22052360e-01 3.11016105e-02
1.28271437e+00 -3.43463011e-02 -5.43506861e-01 9.03020024e-01
1.13678932e+00 -5.87058365e-01 -1.73434889e+00 -4.18327302e-01
2.47639105e-01 -6.55143201e-01 6.68920111e-03 -3.68144006e-01
-5.89624465e-01 9.44529176e-01 3.29495937e-01 -8.01826082e-03
6.22784853e-01 6.65703714e-01 2.84836590e-01 8.98431003e-01
6.98745847e-01 -8.34971905e-01 3.72393459e-01 5.98844290e-01
1.12845647e+00 -1.84084535e+00 8.95451009e-03 -4.16297197e-01
-3.85700256e-01 9.80083108e-01 7.99037814e-01 -6.66932166e-01
6.96470618e-01 -2.33074706e-02 -6.53896257e-02 -2.70316005e-01
-1.75595522e-01 -6.57570541e-01 7.38459527e-01 5.62823176e-01
3.15034449e-01 3.14374894e-01 -2.97451857e-02 1.72185823e-01
-1.43234804e-01 -2.46047437e-01 2.02820435e-01 1.05680907e+00
-7.34947622e-01 -7.69545376e-01 -2.38282606e-01 4.44753826e-01
-3.28874648e-01 -7.90870562e-02 -3.99953693e-01 7.05926478e-01
1.15089372e-01 5.83737493e-01 1.53387353e-01 4.11981530e-02
3.76251817e-01 -1.71728775e-01 8.81587446e-01 -1.08007717e+00
-7.29398802e-02 1.81698114e-01 -1.58627048e-01 -8.72515678e-01
-9.44039345e-01 -7.02322543e-01 -1.48354959e+00 3.57840270e-01
-5.26469886e-01 -3.67909521e-01 6.18104815e-01 1.22367895e+00
-1.69132769e-01 6.56612635e-01 9.03371498e-02 -1.45250189e+00
1.71899311e-02 -1.15895569e+00 -6.02429628e-01 2.29322180e-01
7.40268111e-01 -8.40084672e-01 8.59863609e-02 8.53935927e-02] | [5.164822578430176, -0.11426082998514175] |
08ee943a-a144-4814-be87-1476f56900d1 | sa-cnn-application-to-text-categorization | 2303.07153 | null | https://arxiv.org/abs/2303.07153v1 | https://arxiv.org/pdf/2303.07153v1.pdf | SA-CNN: Application to text categorization issues using simulated annealing-based convolutional neural network optimization | Convolutional neural networks (CNNs) are a representative class of deep learning algorithms including convolutional computation that perform translation-invariant classification of input data based on their hierarchical architecture. However, classical convolutional neural network learning methods use the steepest descent algorithm for training, and the learning performance is greatly influenced by the initial weight settings of the convolutional and fully connected layers, requiring re-tuning to achieve better performance under different model structures and data. Combining the strengths of the simulated annealing algorithm in global search, we propose applying it to the hyperparameter search process in order to increase the effectiveness of convolutional neural networks (CNNs). In this paper, we introduce SA-CNN neural networks for text classification tasks based on Text-CNN neural networks and implement the simulated annealing algorithm for hyperparameter search. Experiments demonstrate that we can achieve greater classification accuracy than earlier models with manual tuning, and the improvement in time and space for exploration relative to human tuning is substantial. | ['Yueying Cao', 'Zihao Guo'] | 2023-03-13 | null | null | null | null | ['text-categorization'] | ['natural-language-processing'] | [ 3.40662301e-01 -5.50293103e-02 -3.57800305e-01 -4.98467177e-01
-3.27335089e-01 -6.23127043e-01 3.40665638e-01 3.45518976e-03
-8.33463371e-01 4.29641247e-01 2.64479360e-03 -6.44263029e-01
-2.76582897e-01 -7.04313576e-01 -5.90641022e-01 -6.14475965e-01
1.03884377e-01 5.21804392e-01 8.39093849e-02 -1.45647034e-01
2.98584908e-01 6.43650174e-01 -1.65225208e+00 3.36242288e-01
6.70760870e-01 1.10005295e+00 4.42410558e-02 8.15170348e-01
-2.73134381e-01 6.17666841e-01 -6.70947194e-01 -2.85567820e-01
1.56865820e-01 -2.61453748e-01 -1.11178863e+00 -2.61175036e-01
1.40287951e-01 -7.60821626e-02 -9.66235474e-02 8.29786181e-01
4.13717210e-01 2.64662176e-01 4.43901777e-01 -1.20779061e+00
-3.00747961e-01 8.15592349e-01 2.44877785e-02 1.33048017e-02
-3.96088213e-01 1.32114440e-01 1.05774808e+00 -6.38001680e-01
1.77121595e-01 1.18120074e+00 9.07434642e-01 7.19044030e-01
-1.10774040e+00 -8.16271603e-01 1.28229186e-01 -6.52829185e-04
-1.41844845e+00 -3.50806177e-01 6.05472207e-01 -1.63095891e-01
1.37929964e+00 2.46727839e-01 9.46869612e-01 7.44038582e-01
2.31091455e-01 8.62369537e-01 4.21222895e-01 -6.41973913e-01
6.84819460e-01 -8.06440637e-02 1.21210121e-01 9.78207707e-01
4.48353469e-01 1.35461211e-01 -4.58748907e-01 -3.09215993e-01
6.21654749e-01 -1.02717765e-01 1.82548026e-03 -4.42370296e-01
-7.81019926e-01 1.03757894e+00 7.07834423e-01 1.29920572e-01
-1.44456536e-01 6.06978297e-01 6.06989443e-01 1.82092682e-01
7.22241700e-02 1.10795629e+00 -9.42300141e-01 -3.00139338e-01
-1.00962043e+00 2.33016282e-01 8.70119274e-01 7.53586769e-01
7.51726270e-01 1.98032886e-01 1.11200750e-01 9.33666468e-01
1.53457224e-01 -1.37018472e-01 8.62710118e-01 -7.77609527e-01
2.99120694e-01 1.09368670e+00 -4.12566155e-01 -7.15033650e-01
-6.89875662e-01 -8.61189485e-01 -9.53280032e-01 3.12511802e-01
1.51970506e-01 -5.23500919e-01 -1.25911796e+00 1.44335449e+00
-2.94709429e-02 -2.18558654e-01 2.73337305e-01 5.15338659e-01
5.41761577e-01 3.72677833e-01 1.68413654e-01 2.09424332e-01
1.15453959e+00 -1.16763902e+00 -4.47174668e-01 -6.04072452e-01
1.03943038e+00 -4.66918141e-01 1.03086460e+00 3.18657756e-01
-1.14976192e+00 -4.57335263e-01 -1.38560688e+00 2.15312969e-02
-6.64119542e-01 3.98770720e-01 8.70858312e-01 9.19784009e-01
-1.11339128e+00 8.13652098e-01 -1.11144209e+00 -1.64105952e-01
7.64192462e-01 8.74701619e-01 1.67009890e-01 1.07814252e-01
-1.02114606e+00 7.87236035e-01 9.76789951e-01 3.13000709e-01
-4.85870361e-01 -6.25726998e-01 -4.84862000e-01 4.75175768e-01
9.49275717e-02 -5.80536425e-01 1.43974698e+00 -1.09730470e+00
-1.62326324e+00 5.09460390e-01 9.86790732e-02 -5.42950809e-01
2.51009017e-01 8.16041753e-02 -3.37183066e-02 -1.33502275e-01
-6.85749769e-01 1.15439248e+00 8.63973856e-01 -8.09286475e-01
-4.55575496e-01 -1.69138595e-01 -1.05720580e-01 2.38421008e-01
-9.03423727e-01 6.77232221e-02 -4.64443415e-01 -5.71052492e-01
5.59727550e-01 -1.40969896e+00 -5.10102749e-01 -1.45650823e-02
-1.70938537e-01 -2.08319440e-01 7.46301532e-01 -4.66151983e-02
1.18141246e+00 -1.80097854e+00 4.33058962e-02 4.80204254e-01
3.31473388e-02 4.53416705e-01 -1.21950343e-01 -2.64035910e-02
-1.01893246e-01 4.01495397e-01 -1.94435343e-01 -2.04707012e-01
-1.47335559e-01 3.59891534e-01 4.78778332e-02 3.73408496e-02
3.97909492e-01 1.25780785e+00 -5.49760759e-01 -2.44999632e-01
-3.10374200e-01 3.83479804e-01 -8.35317135e-01 -1.19219258e-01
-3.77208799e-01 -2.21588120e-01 -3.19750994e-01 4.25234675e-01
2.79475868e-01 -5.20503640e-01 4.49311286e-01 -1.96396679e-01
7.67318457e-02 3.79568636e-01 -1.04904759e+00 1.54939103e+00
-3.93897206e-01 8.90669167e-01 -2.31500089e-01 -7.77604699e-01
1.00667393e+00 1.92590684e-01 1.01280190e-01 -4.69469219e-01
3.84889543e-01 1.71389710e-02 2.66062617e-01 -1.28207579e-01
6.44373000e-01 3.68579209e-01 1.54423058e-01 7.49965429e-01
7.53644109e-02 -4.28108364e-01 4.14861180e-02 -1.48880050e-01
8.45108688e-01 4.56744283e-02 1.21277198e-02 -3.82789165e-01
7.27675781e-02 3.55416954e-01 3.28799367e-01 1.00181377e+00
1.53254792e-01 5.08492112e-01 4.95850861e-01 -1.05787134e+00
-1.19758332e+00 -1.93773657e-01 -3.06872964e-01 1.47811973e+00
-3.93815130e-01 -4.63106960e-01 -1.02247632e+00 -4.41162080e-01
-2.21871346e-01 4.22759563e-01 -7.58730114e-01 -4.93387282e-01
-6.83644593e-01 -1.22879398e+00 9.34924006e-01 8.64452720e-01
7.04698324e-01 -1.17361891e+00 -1.04971182e+00 3.39926004e-01
2.60460168e-01 -8.14915240e-01 -5.24312211e-03 9.67181802e-01
-1.33640599e+00 -9.51475441e-01 -3.07750344e-01 -1.02871263e+00
8.54839325e-01 -2.06015214e-01 1.02609837e+00 5.88878572e-01
-5.49891293e-01 -6.86031505e-02 -1.95553526e-01 -5.84376574e-01
-6.05644345e-01 7.80287862e-01 -2.02663764e-01 -5.97975850e-01
2.55034417e-01 -3.02193910e-01 -3.91729206e-01 2.78940946e-01
-9.62591410e-01 2.94697583e-01 7.01824844e-01 1.20228207e+00
2.30679244e-01 4.90555495e-01 1.35165140e-01 -7.31094420e-01
9.49758351e-01 7.55365426e-03 -7.66188025e-01 3.56411725e-01
-1.00152123e+00 6.27501011e-01 4.53839570e-01 -7.36053169e-01
-8.05938661e-01 3.45966488e-01 5.08009791e-02 -2.72141725e-01
9.25390795e-03 7.16724694e-01 1.99790567e-01 -4.59573120e-01
1.13143873e+00 -1.04728922e-01 2.27849782e-02 -3.38752419e-01
1.77873969e-01 4.76982594e-01 3.34637642e-01 -7.06761181e-01
3.49935979e-01 7.97397047e-02 3.60734127e-02 -3.18856359e-01
-6.11539423e-01 4.37073968e-02 -7.79719412e-01 1.94601412e-03
7.97866762e-01 -6.98932409e-01 -8.40285540e-01 4.30132836e-01
-1.02158201e+00 -6.18708372e-01 1.42114880e-02 4.34614271e-01
-3.41500491e-01 -3.22355598e-01 -5.18318355e-01 -5.81253469e-01
-6.53615952e-01 -1.48418522e+00 5.87446928e-01 2.62384534e-01
-2.83669591e-01 -1.13183916e+00 -2.46514574e-01 1.67513005e-02
7.45093524e-01 -1.22523174e-01 1.32820344e+00 -7.81554639e-01
-3.59010309e-01 -2.42481634e-01 -8.20001215e-03 3.72725844e-01
-1.55327007e-01 6.46182448e-02 -1.02251029e+00 -4.52107370e-01
-3.56372446e-01 -5.89397788e-01 9.01666462e-01 4.33210790e-01
1.53963745e+00 -4.97359067e-01 -3.15105110e-01 8.74267876e-01
1.27877307e+00 2.03186765e-01 5.15210211e-01 8.19733560e-01
5.39316297e-01 3.93340737e-01 1.03404798e-01 3.25439036e-01
-1.03110053e-01 3.59977514e-01 5.78821361e-01 -1.62728399e-01
1.70863181e-01 2.91150268e-02 -1.37713656e-01 4.74972993e-01
1.51423141e-01 -1.89561069e-01 -1.34183383e+00 3.29118431e-01
-1.89910007e+00 -4.35659736e-01 3.75132233e-01 1.86186886e+00
1.01319599e+00 4.82657045e-01 -2.23230973e-01 1.90044507e-01
3.92767847e-01 -1.98037967e-01 -7.73702800e-01 -6.57134175e-01
-3.55121959e-03 5.23967087e-01 8.53183150e-01 7.59450123e-02
-1.05023360e+00 1.13048208e+00 7.31164551e+00 5.47667921e-01
-1.13693297e+00 -5.17150223e-01 7.92634368e-01 -3.37124258e-01
4.01491337e-02 -1.14117272e-01 -8.12811613e-01 1.43443849e-02
1.13476598e+00 2.26315573e-01 6.22286320e-01 1.08480251e+00
-2.29103610e-01 1.45297498e-01 -9.92706120e-01 7.64009535e-01
-2.71062911e-01 -1.76895094e+00 4.96722832e-02 -2.88980633e-01
1.11257648e+00 1.81707680e-01 2.00069845e-01 1.56598747e-01
7.57831395e-01 -1.37394917e+00 7.03304708e-01 1.27591848e-01
5.72012603e-01 -1.09228277e+00 7.97856390e-01 2.33083740e-01
-8.82403493e-01 -5.51199436e-01 -5.56316853e-01 7.11021796e-02
-6.85055792e-01 9.91710573e-02 -1.10629129e+00 -1.47880211e-01
9.83965337e-01 7.71457553e-02 -8.12613368e-01 1.14853442e+00
-1.11879937e-01 7.13591456e-01 -3.35912496e-01 -7.50508785e-01
5.81618428e-01 1.26767486e-01 -2.22911425e-02 1.27804422e+00
2.03485563e-02 -2.50466950e-02 -4.45175134e-02 7.90772319e-01
-3.22827727e-01 -9.15607065e-02 -4.49136719e-02 -1.85716301e-01
6.20227218e-01 1.06562877e+00 -9.49993253e-01 -2.52568841e-01
1.52119875e-01 5.89566290e-01 4.28217083e-01 4.14748996e-01
-4.26087320e-01 -5.90425849e-01 7.95771182e-01 -3.03534061e-01
4.62500274e-01 -3.49361956e-01 -8.39049399e-01 -6.48418367e-01
-2.31958404e-01 -1.05943871e+00 2.61662811e-01 -5.65278471e-01
-4.73170370e-01 8.68021905e-01 -6.87007532e-02 -8.06472838e-01
-4.95889276e-01 -8.74114692e-01 -4.55556959e-01 6.36780679e-01
-1.27589130e+00 -9.82950628e-01 -3.10879558e-01 3.97599250e-01
4.20044035e-01 -4.26041126e-01 1.02010667e+00 -2.67127216e-01
-6.72018886e-01 1.01982832e+00 3.43117684e-01 3.28299016e-01
1.28415018e-01 -9.34743166e-01 8.31721902e-01 4.57219452e-01
-5.01100011e-02 8.89393866e-01 4.36698139e-01 -3.54523629e-01
-1.43116963e+00 -1.17470419e+00 6.16029322e-01 1.01973973e-01
5.03925860e-01 -5.20429313e-01 -9.09864068e-01 6.18830144e-01
7.56194592e-02 -2.29015931e-01 5.71202278e-01 3.89247298e-01
-3.92204881e-01 1.44024745e-01 -9.29730713e-01 8.84020627e-01
9.62156773e-01 -4.76920247e-01 -1.77372530e-01 3.73359650e-01
7.67996788e-01 -7.19827652e-01 -6.67765975e-01 3.33749652e-01
6.73593938e-01 -3.55145842e-01 8.74325454e-01 -1.07803035e+00
2.63212740e-01 6.11201748e-02 1.43080592e-01 -1.29446542e+00
-3.86913687e-01 -6.61676288e-01 1.04455180e-01 5.34556448e-01
8.59939158e-01 -5.35943568e-01 1.24915910e+00 9.61649895e-01
2.02455092e-03 -1.02753627e+00 -6.78102374e-01 -4.97560441e-01
3.12998891e-01 -4.58889991e-01 9.32822883e-01 8.28121781e-01
-2.35486597e-01 2.54103780e-01 1.40972942e-01 -9.83277801e-03
3.00371736e-01 -1.24990292e-01 3.34578991e-01 -1.51475632e+00
-2.54612923e-01 -8.06378961e-01 -2.65084177e-01 -9.37568307e-01
1.79187715e-01 -7.64569104e-01 2.23596841e-01 -1.13520980e+00
7.44021088e-02 -8.71157348e-01 -1.92190930e-01 1.09297788e+00
-4.31884937e-02 2.06822678e-01 4.47044671e-02 2.71669030e-01
-3.74314517e-01 3.05549830e-01 7.58564115e-01 -1.04286030e-01
-5.43618858e-01 6.42309245e-03 -7.08045602e-01 6.06324375e-01
1.14619732e+00 -5.76339900e-01 -4.07373250e-01 -8.04031909e-01
6.68072522e-01 -4.28717881e-01 1.47009254e-01 -1.09280193e+00
8.15182626e-01 1.52008492e-03 8.41681898e-01 -3.82141203e-01
1.81030110e-01 -7.07672417e-01 7.47663248e-03 7.01437414e-01
-1.06505990e+00 3.81627113e-01 6.70299470e-01 2.14406654e-01
9.01288837e-02 -6.74405515e-01 8.22032869e-01 -2.24527493e-01
-4.30012375e-01 2.01763153e-01 -7.09902704e-01 -1.92726582e-01
3.94289136e-01 -3.37324828e-01 -2.26049319e-01 -1.04134575e-01
-5.04210293e-01 1.57514840e-01 2.65684634e-01 4.61915523e-01
6.20384157e-01 -1.06601059e+00 -3.49287003e-01 4.60209429e-01
-9.29371193e-02 2.31448099e-01 -3.36891264e-01 2.01679125e-01
-7.19838321e-01 5.73875010e-01 -2.41941318e-01 -5.48406541e-01
-1.30339241e+00 9.28282365e-02 7.46426404e-01 -2.92349547e-01
-2.84971178e-01 1.10865700e+00 -4.70781684e-01 -7.42398143e-01
7.40582347e-01 -3.85394663e-01 -5.61986305e-02 -2.46076435e-01
4.01978940e-01 1.22662567e-01 5.75098932e-01 2.05030575e-01
-3.07016283e-01 2.85266399e-01 -3.88233960e-01 -7.57375658e-02
1.45940626e+00 5.42608976e-01 -1.06081530e-01 -2.02052802e-01
1.52891779e+00 -8.45364451e-01 -1.30454421e+00 -3.33422989e-01
3.46536428e-01 8.52095187e-02 6.61376297e-01 -9.31357563e-01
-1.09932327e+00 7.86082745e-01 6.80275142e-01 3.17178969e-03
1.19666135e+00 -3.63712698e-01 3.97197992e-01 1.12636161e+00
-1.38222635e-01 -1.35240781e+00 2.55937785e-01 8.72499228e-01
5.68720520e-01 -9.70734358e-01 2.18378544e-01 1.74290970e-01
-5.39362252e-01 1.81068718e+00 6.23695672e-01 6.67226240e-02
6.39071345e-01 6.84674561e-01 1.32802486e-01 -2.69269198e-01
-1.07340980e+00 1.13953002e-01 2.59054482e-01 2.15184808e-01
2.56964833e-01 -1.97821155e-01 1.93619981e-01 2.20726296e-01
-6.07601583e-01 -1.28685415e-01 1.08333699e-01 1.30111372e+00
-5.95449924e-01 -9.22646642e-01 -9.73167643e-02 4.33444709e-01
-2.66874611e-01 -4.14629340e-01 -4.12770361e-01 6.95348263e-01
-1.60970569e-01 7.26564169e-01 4.59213108e-01 -5.31795859e-01
9.81208757e-02 2.45387331e-01 3.56254935e-01 -3.48281622e-01
-1.14039886e+00 -2.78323740e-01 2.77734436e-02 -4.30871785e-01
1.25668512e-03 -4.47593719e-01 -1.52024603e+00 -3.52864623e-01
-6.92466617e-01 2.02030808e-01 1.37450564e+00 9.41591859e-01
4.60684925e-01 7.03933001e-01 4.77718771e-01 -7.59441614e-01
-7.93997049e-01 -6.70540035e-01 8.42989087e-02 -1.95225909e-01
2.19878301e-01 -2.66531587e-01 -2.55912304e-01 -8.50431621e-02] | [8.602965354919434, 3.3916993141174316] |
25676aa9-4782-48d3-a0ff-0e267537a15f | large-raw-emotional-dataset-with-aggregation | 2212.12266 | null | https://arxiv.org/abs/2212.12266v1 | https://arxiv.org/pdf/2212.12266v1.pdf | Large Raw Emotional Dataset with Aggregation Mechanism | We present a new data set for speech emotion recognition (SER) tasks called Dusha. The corpus contains approximately 350 hours of data, more than 300 000 audio recordings with Russian speech and their transcripts. Therefore it is the biggest open bi-modal data collection for SER task nowadays. It is annotated using a crowd-sourcing platform and includes two subsets: acted and real-life. Acted subset has a more balanced class distribution than the unbalanced real-life part consisting of audio podcasts. So the first one is suitable for model pre-training, and the second is elaborated for fine-tuning purposes, model approbation, and validation. This paper describes pre-processing routine, annotation, and experiment with a baseline model to demonstrate some actual metrics which could be obtained with the Dusha data set. | ['Fyodor Minkin', 'Nikita Savushkin', 'Oleg Kutuzov', 'Nikolay Karpov', 'Artem Sokolov', 'Vladimir Kondratenko'] | 2022-12-23 | null | null | null | null | ['speech-emotion-recognition'] | ['speech'] | [-2.23941877e-01 3.89022112e-01 5.03374338e-01 -5.60939550e-01
-1.01138985e+00 -4.25430775e-01 5.86058915e-01 1.10241853e-01
-6.71623766e-01 8.16017389e-01 5.29392540e-01 6.45885617e-02
3.48297298e-01 -5.14824949e-02 -1.79813415e-01 -6.08152568e-01
-1.84818029e-01 7.71598756e-01 5.66454325e-03 -5.29658854e-01
-1.12156123e-01 2.81364322e-01 -1.81703925e+00 7.09423542e-01
4.47980553e-01 1.32554626e+00 3.82567160e-02 7.43582070e-01
-3.13477039e-01 9.89574194e-01 -1.08051789e+00 -6.26650035e-01
-1.58554479e-01 -3.26986194e-01 -1.07879543e+00 1.11343712e-01
-2.06778333e-01 3.00343454e-01 4.60686833e-02 7.65667379e-01
1.11189187e+00 4.46839422e-01 3.62102240e-01 -1.34806681e+00
-2.64558941e-01 9.53551054e-01 3.52602825e-02 3.04475635e-01
6.67234838e-01 -2.73568034e-01 8.99900734e-01 -8.77085745e-01
8.33125174e-01 1.04901230e+00 5.38170397e-01 8.42024922e-01
-6.88859463e-01 -5.26931643e-01 -5.77119924e-02 3.53009105e-01
-1.30723011e+00 -9.07911420e-01 9.57726955e-01 -3.47362906e-01
1.02623498e+00 5.27461946e-01 6.38079524e-01 1.87579179e+00
-4.93901432e-01 1.04873633e+00 1.13489532e+00 -4.57570255e-01
5.34898043e-01 6.43002570e-01 2.41059914e-01 -3.02539133e-02
-7.20233381e-01 -1.24856997e-02 -7.63497710e-01 -4.24496144e-01
8.10584575e-02 -7.58679271e-01 -5.83333492e-01 7.18289614e-02
-1.15105999e+00 5.59349597e-01 -4.59323943e-01 6.78888738e-01
-5.23011446e-01 -6.20042324e-01 1.16310811e+00 9.13456380e-01
9.50325072e-01 3.09785336e-01 -9.97188628e-01 -9.58750367e-01
-6.33790195e-01 3.30814272e-02 1.01378143e+00 7.54927695e-01
2.88072884e-01 3.95335943e-01 9.25543606e-02 1.59210145e+00
-1.20767243e-02 1.28475413e-01 1.05733228e+00 -5.30334115e-01
7.08736002e-01 2.47027487e-01 1.60725355e-01 -5.23347199e-01
-5.13204396e-01 -4.75136973e-02 -8.31916869e-01 -2.39766285e-01
2.56904542e-01 -6.07769549e-01 -4.87864107e-01 1.54016197e+00
3.39971602e-01 3.98487486e-02 5.01737833e-01 7.26131737e-01
1.37492478e+00 9.96885359e-01 -1.09035708e-02 -6.09050274e-01
1.45753801e+00 -1.02638674e+00 -1.17491734e+00 -1.02771252e-01
6.93293750e-01 -6.75188363e-01 1.37875760e+00 9.02399004e-01
-1.03496397e+00 -4.48182195e-01 -8.25065196e-01 1.74438953e-01
-6.51357055e-01 3.05130988e-01 3.84585917e-01 8.83956909e-01
-1.00428653e+00 1.34347290e-01 -3.45493436e-01 -3.16088676e-01
2.72776596e-02 -1.03492796e-01 -9.44454253e-01 5.39453626e-01
-1.38190854e+00 8.50326061e-01 3.92716199e-01 -2.94111222e-02
-6.28507733e-01 -3.82251024e-01 -9.07475471e-01 1.03185855e-01
1.39011458e-01 1.97601356e-02 1.59588194e+00 -1.24008381e+00
-1.98607314e+00 1.19004154e+00 7.94802010e-02 -5.05343795e-01
4.51795042e-01 -2.10899785e-02 -1.07724559e+00 1.87272187e-02
-4.00703430e-01 1.98780581e-01 8.99960399e-01 -1.01414204e+00
-3.76164347e-01 -4.24985737e-01 -4.56512719e-01 8.62186402e-02
-5.44763088e-01 7.69368231e-01 -1.89890608e-01 -8.76104832e-01
-3.00422400e-01 -6.52240574e-01 -5.82436062e-02 -7.97124684e-01
-2.30797544e-01 -3.43965024e-01 6.42407477e-01 -7.16305137e-01
1.42962050e+00 -2.49590492e+00 -1.57565117e-01 5.33472234e-03
-1.49428934e-01 4.17305142e-01 -3.10450613e-01 8.06710243e-01
-6.94470048e-01 8.00529048e-02 -5.43586304e-03 -7.26590931e-01
1.79678276e-02 8.57255831e-02 -3.62491399e-01 2.77373016e-01
8.51969793e-02 3.08075547e-01 -5.18561125e-01 -3.42508316e-01
-1.36099095e-02 4.60890502e-01 -1.78038895e-01 5.08590579e-01
6.08203635e-02 4.91753340e-01 -1.29440382e-01 4.35229629e-01
4.46200520e-01 4.18658823e-01 -1.41432256e-01 5.22431396e-02
-1.04652997e-02 3.67716879e-01 -1.35761726e+00 1.46571338e+00
-6.19901478e-01 8.44478846e-01 4.54324186e-01 -9.95303690e-01
1.31321347e+00 1.02672899e+00 5.71954310e-01 -5.44435501e-01
6.35462105e-01 6.33816198e-02 -2.20769912e-01 -8.68397892e-01
6.82562888e-01 -1.95450455e-01 -2.98825234e-01 3.75767350e-01
5.12664318e-01 -1.62413642e-01 1.82669833e-01 -4.47470993e-02
1.05813253e+00 -4.15352285e-01 3.71503204e-01 -1.12568252e-01
6.76722646e-01 -2.49482423e-01 7.59184361e-01 1.64702266e-01
-5.90505362e-01 7.76932061e-01 6.49532318e-01 -5.42526186e-01
-9.07671213e-01 -5.37109077e-01 -1.11810699e-01 1.17855155e+00
-4.75043833e-01 -5.94009697e-01 -8.70711327e-01 -5.62095284e-01
-6.49473906e-01 5.58408320e-01 -7.48604357e-01 5.38929664e-02
1.10399304e-02 -5.04507184e-01 1.00684571e+00 3.26522619e-01
2.89727837e-01 -1.52565444e+00 -3.16665173e-01 1.60807222e-01
-6.82474911e-01 -1.29927504e+00 -2.42163062e-01 4.84008700e-01
-3.24509621e-01 -9.24886107e-01 -6.81779861e-01 -8.03206086e-01
1.29822090e-01 -2.80941278e-01 1.43887162e+00 -4.93950933e-01
1.14516623e-01 2.94385314e-01 -1.10778499e+00 -9.20225620e-01
-6.24004245e-01 -5.76690026e-02 1.79459721e-01 3.85203391e-01
5.01252413e-01 -4.19700682e-01 2.56885588e-01 5.22265375e-01
-6.36247814e-01 -3.00136656e-01 1.88148245e-01 9.03222382e-01
3.19490403e-01 9.21232849e-02 8.63761961e-01 -7.53078580e-01
1.00742435e+00 -3.85446787e-01 -1.18877023e-01 6.27194569e-02
1.75020128e-01 -4.09417331e-01 4.52742130e-01 -6.00297272e-01
-1.42537141e+00 4.90108598e-03 -7.48240352e-01 -4.53951508e-01
-6.72926009e-01 5.33187091e-01 -6.31267309e-01 4.71537948e-01
8.40531409e-01 -6.61568344e-02 -1.37196362e-01 -8.27659369e-01
-2.71099769e-02 1.48853040e+00 5.31150937e-01 -4.89812613e-01
1.76009778e-02 -4.87639084e-02 -1.11609256e+00 -1.35139227e+00
-5.40824354e-01 -6.69619620e-01 -4.27046746e-01 -3.49411488e-01
7.52959728e-01 -1.11870813e+00 -6.48869038e-01 6.02636218e-01
-1.30907619e+00 -4.36193556e-01 -5.90318322e-01 5.52971125e-01
-5.64370930e-01 -2.80294865e-02 -6.88769400e-01 -1.45156753e+00
-4.32638109e-01 -7.89143026e-01 1.04771638e+00 -2.13531494e-01
-3.59581798e-01 -8.30493093e-01 4.77904230e-01 3.99456650e-01
1.25315398e-01 1.31990895e-01 2.97910899e-01 -1.38079417e+00
8.99108231e-01 -4.55014229e-01 3.24676752e-01 7.05672920e-01
-2.79235035e-01 1.62766740e-01 -1.74914002e+00 -1.31228101e-02
2.61823863e-01 -9.53254640e-01 4.04720008e-01 -8.36831108e-02
1.18902576e+00 -1.10605627e-01 2.99620211e-01 -1.15273193e-01
5.81607163e-01 4.98005390e-01 8.14046860e-01 1.61310717e-01
1.79648161e-01 8.35148454e-01 8.96841705e-01 7.47307539e-01
3.82686019e-01 6.56838238e-01 1.45902395e-01 7.09892213e-02
3.24056089e-01 -4.11168560e-02 6.68735743e-01 1.59438765e+00
-3.53310071e-02 -4.27760482e-01 -1.02768505e+00 7.24908769e-01
-1.65711427e+00 -1.22325504e+00 -7.85837919e-02 2.02692485e+00
8.72555435e-01 -9.29644182e-02 5.58043063e-01 6.78966761e-01
6.38967574e-01 2.13491127e-01 1.29189538e-02 -7.70961285e-01
-4.26010221e-01 1.02041833e-01 -3.79467666e-01 2.38852680e-01
-1.10825789e+00 7.38951266e-01 6.72912169e+00 1.07673132e+00
-9.45308089e-01 3.23837429e-01 8.55533898e-01 -4.01379131e-02
1.28475539e-02 -5.49759805e-01 -4.78040457e-01 4.24743980e-01
1.61115754e+00 -1.05628021e-01 1.38066486e-01 1.03878129e+00
4.30345774e-01 9.23253521e-02 -8.89432371e-01 1.47998428e+00
2.07446650e-01 -9.23333108e-01 -4.10728395e-01 -2.54125476e-01
3.15150321e-01 4.13439840e-01 -5.78887105e-01 7.67208934e-01
1.29545644e-01 -6.91826165e-01 9.66663003e-01 2.28378832e-01
7.71927953e-01 -8.03483546e-01 1.17918384e+00 5.39558589e-01
-9.38399315e-01 -7.46837258e-02 -2.46683151e-01 -8.91674086e-02
2.61852890e-01 5.49201906e-01 -6.02057874e-01 5.28784752e-01
1.10180032e+00 6.12710953e-01 -3.75412315e-01 8.13768268e-01
1.83245763e-01 7.86588967e-01 -3.60182762e-01 -1.33238882e-01
3.50845419e-02 -2.29197115e-01 5.53283334e-01 1.36134410e+00
2.04437777e-01 4.46713567e-02 -4.76185158e-02 1.19681647e-02
-1.54406466e-02 5.87954879e-01 -6.25559807e-01 -2.49141946e-01
5.24304926e-01 1.43860018e+00 -3.78096551e-01 -3.82600039e-01
-2.61420518e-01 8.01001906e-01 3.27633679e-01 1.63255960e-01
-8.25675011e-01 -5.85714996e-01 6.25093997e-01 -2.41816297e-01
4.86423075e-02 2.69592881e-01 6.03027940e-02 -1.37630665e+00
4.84331883e-02 -1.29008186e+00 4.91442174e-01 -1.03172791e+00
-1.43660355e+00 1.53281116e+00 -2.03974038e-01 -1.31504893e+00
-4.64931369e-01 -6.42795384e-01 -6.45807326e-01 5.52579403e-01
-9.34377313e-01 -8.19846034e-01 -2.53845453e-01 7.69760668e-01
8.61082315e-01 -5.99394977e-01 1.15880847e+00 7.49400854e-01
-8.25916588e-01 5.18828511e-01 -1.34191588e-01 2.87648678e-01
7.46115208e-01 -1.14626706e+00 9.64685902e-02 4.23615426e-01
3.62397522e-01 -4.72337715e-02 8.12009096e-01 -8.93323123e-02
-8.22952151e-01 -7.23997235e-01 1.19282913e+00 -4.86785233e-01
8.51914287e-01 -6.21205449e-01 -1.03918040e+00 4.53943551e-01
3.74342173e-01 -5.87180890e-02 1.09335196e+00 2.72800565e-01
-2.26014361e-01 -7.55955428e-02 -1.26714778e+00 2.06783220e-01
6.84198320e-01 -6.23444617e-01 -6.51249409e-01 1.94381014e-01
5.85643947e-01 -3.95500600e-01 -1.11006618e+00 1.40920058e-01
2.75763780e-01 -1.04485464e+00 4.29097742e-01 -8.13768089e-01
1.50275454e-01 1.79968104e-01 -2.45545939e-01 -1.85038948e+00
3.46484691e-01 -1.01065648e+00 1.38154536e-01 1.83075368e+00
6.69724464e-01 -5.10491610e-01 5.09616017e-01 4.27244842e-01
-4.71151352e-01 -5.54984510e-01 -1.16812503e+00 -7.85869956e-01
-1.61744148e-01 -1.00196433e+00 7.46023536e-01 1.23324025e+00
5.83656430e-01 6.98038578e-01 -4.39301580e-01 -2.27859542e-01
-2.18686223e-01 -4.42329168e-01 8.58745158e-01 -1.39265990e+00
-1.05079368e-01 -2.84288615e-01 -4.63995159e-01 -5.02262056e-01
4.32730734e-01 -3.37551504e-01 1.90876454e-01 -9.56058621e-01
-4.42437470e-01 -3.28074098e-01 7.03626499e-02 5.76494575e-01
1.34172276e-01 1.80509329e-01 3.58490050e-02 -1.84319809e-01
-5.68229258e-01 9.16782975e-01 6.75144613e-01 5.87966144e-02
-4.60705161e-01 7.78526962e-02 -3.36168259e-01 7.63321638e-01
8.27503979e-01 -2.84014523e-01 -4.39905286e-01 -2.05063373e-02
2.25655451e-01 3.57536703e-01 -1.16373658e-01 -8.51601899e-01
-1.49442971e-01 1.91368386e-01 -1.94116443e-01 -5.09165823e-01
6.97940171e-01 -8.97829413e-01 2.12707981e-01 -2.84607172e-01
-4.52022702e-01 2.20567182e-01 3.37531120e-01 2.38788724e-01
-8.56942296e-01 -2.27658749e-01 6.96101606e-01 5.97924255e-02
-6.41860008e-01 1.14481203e-01 -7.63462603e-01 4.19775277e-01
9.59790170e-01 -1.06631659e-01 -1.36919215e-01 -7.59961069e-01
-1.29955029e+00 -1.05007393e-02 3.17164920e-02 7.67729878e-01
2.81359881e-01 -1.31084001e+00 -8.71924281e-01 2.58522362e-01
4.96720016e-01 -2.43775368e-01 4.46040660e-01 7.95051157e-01
-1.56654805e-01 2.02176329e-02 -2.08074465e-01 -1.83494836e-01
-1.62145531e+00 4.20008659e-01 2.55865514e-01 -3.00649345e-01
-4.55652237e-01 8.86804879e-01 -2.25219727e-01 -8.20515573e-01
6.11736834e-01 -1.19825706e-01 -6.94743037e-01 6.29608333e-01
9.22176600e-01 3.21522892e-01 6.33437395e-01 -9.96062875e-01
-2.49279872e-01 -2.71871775e-01 3.24680895e-01 -4.41668868e-01
1.52458978e+00 -2.80333549e-01 -1.47108389e-02 1.13935661e+00
1.10859311e+00 6.57153204e-02 -7.62263656e-01 -9.00058597e-02
8.11389238e-02 -1.17314674e-01 1.00917608e-01 -9.17911291e-01
-1.05180275e+00 8.52642775e-01 3.97579908e-01 8.46427500e-01
1.12208045e+00 2.09774706e-03 5.25105655e-01 4.13680553e-01
3.87293190e-01 -1.54109621e+00 4.05273866e-03 8.01170647e-01
1.39929175e+00 -1.10541356e+00 -7.00747550e-01 -2.81001419e-01
-1.42096651e+00 7.20204949e-01 4.97548074e-01 2.95702845e-01
8.10733557e-01 3.89407814e-01 6.56445444e-01 -1.80824250e-01
-1.18587101e+00 -2.18447775e-01 6.13514557e-02 8.10517490e-01
4.35751945e-01 -6.02753740e-03 -1.12592272e-01 1.31226623e+00
-7.15631485e-01 -1.67990685e-01 5.23489892e-01 5.00758708e-01
-1.33279279e-01 -9.19149101e-01 -4.34090972e-01 1.92778304e-01
-6.11014247e-01 2.03163728e-01 -8.10777068e-01 6.61364436e-01
-1.27888739e-01 1.36963773e+00 4.27324772e-02 -5.73416352e-01
6.44462466e-01 4.89710718e-01 2.60794070e-02 -4.33080614e-01
-9.94217932e-01 1.42130956e-01 9.80303407e-01 -5.71494043e-01
-4.25369740e-01 -6.39881432e-01 -8.57876360e-01 -1.32318780e-01
-4.48892415e-01 8.77593338e-01 9.01733577e-01 8.38682771e-01
4.97336447e-01 5.25513887e-01 9.13968444e-01 -1.11696327e+00
-2.03723416e-01 -1.53687751e+00 -8.93157005e-01 5.69314301e-01
1.33400485e-01 -5.48338473e-01 -5.90338051e-01 2.62259513e-01] | [13.582880973815918, 5.8543806076049805] |
3cb32a2e-1b98-48b7-8191-50a92dea6df2 | cross-domain-recommender-systems-via | 2306.13887 | null | https://arxiv.org/abs/2306.13887v2 | https://arxiv.org/pdf/2306.13887v2.pdf | Cross-domain Recommender Systems via Multimodal Domain Adaptation | Collaborative Filtering (CF) has emerged as one of the most prominent implementation strategies for building recommender systems. The key idea is to exploit the usage patterns of individuals to generate personalized recommendations. CF techniques, especially for newly launched platforms, often face a critical issue known as the data sparsity problem, which greatly limits their performance. Several approaches in the literature have been proposed to tackle the problem of data sparsity, among which cross-domain collaborative filtering (CDCF) has gained significant attention in the recent past. In order to compensate for the scarcity of available feedback in a target domain, the CDCF approach utilizes information available in other auxiliary domains. Traditional CDCF approaches primarily focus on finding a common set of entities (users or items) across the domains, which then act as a conduit for knowledge transfer. Nevertheless, most real-world datasets are collected from different domains, so they often lack information about anchor points or reference information for entity alignment. This paper introduces a domain adaptation technique to align the embeddings of entities across the two domains. Our approach first exploits the available textual and visual information to independently learn a multi-view latent representation for each entity in the auxiliary and target domains. The different representations of the entity are then fused to generate the corresponding unified representation. A domain classifier is then trained to learn the embedding for the domain alignment by fixing the unified features as the anchor points. Experiments on two publicly benchmark datasets indicate the effectiveness of our proposed approach. | ['Venkateswara Rao Kagita', 'Vikas Kumar', 'Ramya Kamani'] | 2023-06-24 | null | null | null | null | ['entity-alignment', 'transfer-learning', 'collaborative-filtering', 'entity-alignment'] | ['knowledge-base', 'miscellaneous', 'miscellaneous', 'natural-language-processing'] | [-1.13263436e-01 -4.09851342e-01 -4.87337947e-01 -3.90982300e-01
-3.98775995e-01 -6.57700717e-01 5.80932558e-01 4.21391189e-01
-1.61048606e-01 5.85571527e-01 5.66126525e-01 2.68854171e-01
-2.38091409e-01 -9.62050974e-01 -2.66380876e-01 -5.65522730e-01
2.32044145e-01 2.81470239e-01 2.30871841e-01 -3.15349489e-01
2.15488046e-01 1.37381390e-01 -1.42152619e+00 2.85135716e-01
1.30951416e+00 8.17014158e-01 3.77224296e-01 -6.11156709e-02
-5.83581746e-01 8.96932781e-02 -5.57429671e-01 -6.04919434e-01
4.43459183e-01 -2.34846920e-01 -3.02396059e-01 1.36456802e-01
2.50409365e-01 -1.61525890e-01 -4.39076960e-01 1.01720440e+00
5.27947247e-01 4.19940263e-01 6.86592758e-01 -1.13883305e+00
-9.87080336e-01 4.27466720e-01 -6.70230865e-01 1.90780908e-01
5.07738233e-01 -3.25700015e-01 1.13883042e+00 -9.84363854e-01
6.81067944e-01 1.00974584e+00 4.66819704e-01 4.53460425e-01
-1.07928121e+00 -7.18368948e-01 5.76810777e-01 1.63100123e-01
-1.48561549e+00 -2.78165132e-01 9.24380064e-01 -5.47445238e-01
3.29028964e-01 2.10407265e-02 6.04849100e-01 1.07177222e+00
-2.36408472e-01 6.48717821e-01 9.52968001e-01 -3.71105433e-01
5.91318049e-02 7.39791632e-01 2.11411953e-01 1.92216724e-01
4.90317941e-01 -1.65065706e-01 -5.79074383e-01 -3.76567185e-01
7.36632049e-01 5.97848594e-01 -3.96502733e-01 -5.66692054e-01
-9.86869216e-01 8.59512806e-01 6.34024024e-01 5.70593774e-01
-2.90219426e-01 -8.43867600e-01 2.91191041e-01 3.29503387e-01
5.83939433e-01 4.56606120e-01 -3.17591369e-01 2.52270043e-01
-7.35941470e-01 3.53087544e-01 7.47092187e-01 9.18862283e-01
7.92868793e-01 -2.60341346e-01 -2.73663234e-02 1.01804674e+00
5.49893737e-01 3.41637671e-01 6.75277948e-01 -2.74864942e-01
8.85923386e-01 1.10659873e+00 4.20967162e-01 -1.57402527e+00
3.49679030e-02 -6.64734900e-01 -8.23385954e-01 -3.07897538e-01
4.09298271e-01 -3.03295225e-01 -3.11076820e-01 1.62497211e+00
6.70531750e-01 4.15102482e-01 1.65663600e-01 8.84877384e-01
8.18560898e-01 7.35731423e-01 -1.11166816e-02 2.19476037e-02
1.09061277e+00 -8.28469157e-01 -6.37225628e-01 -2.11690962e-01
5.61056316e-01 -7.74811447e-01 7.01913059e-01 1.89156547e-01
-4.40537751e-01 -7.06024766e-01 -1.10907054e+00 2.10080117e-01
-5.95410645e-01 4.61532027e-01 4.63321120e-01 5.35572171e-01
-3.89186472e-01 4.77013856e-01 -3.55017424e-01 -4.10868734e-01
3.27853024e-01 2.43906528e-01 -5.33458233e-01 -3.73044848e-01
-1.33323002e+00 5.57749808e-01 3.58660161e-01 -9.50989947e-02
-3.31341714e-01 -7.02528417e-01 -6.38290286e-01 1.20764621e-01
2.75779396e-01 -4.74597812e-01 6.43044114e-01 -1.19925976e+00
-1.23395491e+00 4.74601895e-01 -1.34525886e-02 -1.99456111e-01
2.41629824e-01 -3.83327276e-01 -1.03804338e+00 -2.89834052e-01
2.08225533e-01 9.33151990e-02 8.53970468e-01 -1.24057317e+00
-9.41969037e-01 -5.59224486e-01 2.68006504e-01 4.49604541e-01
-9.06387031e-01 -1.44600675e-01 -5.64783633e-01 -8.88535917e-01
2.08943654e-02 -7.75822103e-01 -1.58393845e-01 -1.41801849e-01
-4.17562947e-02 -3.36487800e-01 8.36907864e-01 -6.40334010e-01
1.61768973e+00 -2.10122323e+00 3.54248047e-01 3.92891139e-01
2.99190819e-01 4.03637052e-01 9.70422197e-03 6.33171976e-01
-4.26235721e-02 -3.14761668e-01 1.18848391e-01 -2.87425756e-01
-1.36016190e-01 1.30801693e-01 -3.78634095e-01 3.74037266e-01
-1.15082890e-01 4.10897553e-01 -9.31066692e-01 -4.01597738e-01
4.23557162e-02 3.84856045e-01 -6.71825767e-01 4.19594705e-01
-3.26660611e-02 5.20358562e-01 -6.56069100e-01 3.11004609e-01
7.92738020e-01 -3.77470076e-01 6.56997263e-01 -3.70637655e-01
-1.47931904e-01 1.41602114e-01 -1.69800901e+00 1.71359634e+00
-4.79451984e-01 1.34087116e-01 1.26479678e-02 -1.15231895e+00
1.19193447e+00 2.27588072e-01 6.25060141e-01 -5.90909004e-01
-1.53933257e-01 2.08987907e-01 -1.09600283e-01 -4.33252305e-01
4.38558340e-01 2.70659495e-02 -1.90872580e-01 2.28652716e-01
1.90183908e-01 7.36109555e-01 5.58728315e-02 2.07370088e-01
7.25144565e-01 1.02939300e-01 4.38024819e-01 3.46823186e-02
8.15291643e-01 7.42498487e-02 8.25567961e-01 3.30828160e-01
8.44267085e-02 3.05497050e-01 -1.05617307e-01 -4.64563727e-01
-8.83329630e-01 -8.96252334e-01 6.87291520e-03 1.11998212e+00
4.76760864e-01 -7.42359698e-01 -1.95639849e-01 -1.05249619e+00
2.36571193e-01 2.71297574e-01 -5.56696415e-01 -7.70113021e-02
-3.63077044e-01 -3.69883716e-01 -7.77885085e-03 4.46813077e-01
5.32847762e-01 -5.28667986e-01 1.81390103e-02 3.46129596e-01
-2.04326913e-01 -8.67190838e-01 -5.11388659e-01 -3.59995037e-01
-7.99905717e-01 -1.00814509e+00 -9.43781555e-01 -7.83020735e-01
9.22254443e-01 7.87673533e-01 8.34959686e-01 2.57716677e-03
2.30984181e-01 2.51174033e-01 -5.92373788e-01 -5.65722510e-02
3.73302996e-02 1.32615298e-01 4.05946225e-01 5.38170457e-01
6.61714554e-01 -5.29227853e-01 -7.09748387e-01 6.86345041e-01
-7.82323360e-01 -1.88715488e-01 5.22270977e-01 8.75645578e-01
5.03481209e-01 2.19121397e-01 7.57919431e-01 -1.22478640e+00
7.58163333e-01 -9.22067046e-01 -4.87167686e-01 2.95800418e-01
-5.23149490e-01 -1.96878701e-01 9.69322383e-01 -6.05324030e-01
-1.23679376e+00 -4.29387577e-02 1.36381313e-01 -5.06632388e-01
-2.58969218e-01 7.59557605e-01 -4.91582096e-01 1.57322451e-01
6.20319545e-01 1.14112578e-01 -3.55126172e-01 -8.89054894e-01
3.87521714e-01 8.95870686e-01 2.91131794e-01 -5.55890977e-01
1.02154863e+00 3.47508878e-01 -5.17813504e-01 -6.35684788e-01
-9.25937533e-01 -9.59521294e-01 -7.63669193e-01 3.19794342e-02
5.15435457e-01 -1.11054754e+00 -1.67028040e-01 5.58358431e-02
-8.50471556e-01 4.26623970e-01 -1.75539374e-01 6.00499213e-01
1.16267592e-01 5.69387555e-01 4.36855070e-02 -6.28399014e-01
-2.18139276e-01 -7.77602911e-01 6.57161415e-01 5.88372409e-01
-6.31950647e-02 -1.16777980e+00 3.64000440e-01 3.02587360e-01
3.23273540e-01 2.39486303e-02 7.49698162e-01 -1.09610724e+00
-2.00229272e-01 -4.17959452e-01 -2.39001676e-01 2.12021664e-01
5.94524741e-01 -3.56943220e-01 -5.75486600e-01 -5.39804876e-01
-2.27571100e-01 1.95325628e-01 4.68738317e-01 -3.46935354e-02
8.01330984e-01 -3.91601145e-01 -6.98351920e-01 3.82625282e-01
1.43405724e+00 9.66391489e-02 2.18249783e-01 3.01263571e-01
7.82294691e-01 6.41732395e-01 9.58615839e-01 6.49837196e-01
4.14347619e-01 1.01980102e+00 -5.57998642e-02 7.61448666e-02
1.08339101e-01 -6.45809710e-01 2.97539651e-01 9.55238938e-01
-6.92412257e-02 -1.81812003e-01 -6.17312491e-01 6.04412973e-01
-1.94485247e+00 -1.02054965e+00 2.15539128e-01 2.48587322e+00
4.93747950e-01 -7.57016465e-02 3.64657551e-01 -3.09782289e-02
9.15830851e-01 2.61977762e-02 -5.01350701e-01 9.68835205e-02
3.68211046e-02 -4.33734357e-02 2.77093410e-01 -6.40905127e-02
-1.19688511e+00 5.37553728e-01 4.84130526e+00 6.83052003e-01
-1.02337062e+00 1.15047641e-01 -7.35194534e-02 1.65164128e-01
-2.92322099e-01 -1.50115904e-03 -1.11418855e+00 7.42905378e-01
6.68641686e-01 -3.72404009e-01 1.38057157e-01 1.00097740e+00
1.40098125e-01 3.33106101e-01 -8.92296076e-01 1.05170286e+00
1.98256392e-02 -1.29293084e+00 1.07067175e-01 1.64240971e-01
9.84430134e-01 -3.03468168e-01 1.56515658e-01 4.06099170e-01
2.77877688e-01 -5.67748845e-01 1.73604265e-01 5.77639580e-01
7.06458986e-01 -7.18735576e-01 7.97766924e-01 4.69523400e-01
-1.49976277e+00 -3.22271466e-01 -6.55796230e-01 2.18555648e-02
2.11846232e-02 4.83174771e-01 -7.51839340e-01 9.67648864e-01
7.06423163e-01 1.23332274e+00 -4.32649761e-01 1.26434541e+00
1.03890514e-02 4.21554416e-01 -2.61478186e-01 1.22597791e-01
-9.46281925e-02 -5.73315144e-01 4.28202361e-01 8.35875273e-01
5.92733383e-01 -6.24266006e-02 5.12495160e-01 5.36578476e-01
-2.40506083e-01 6.00842714e-01 -6.99703991e-01 -1.54265478e-01
7.74921298e-01 1.28846455e+00 -3.34355086e-01 -2.58767605e-01
-9.90827680e-01 1.01944184e+00 3.68409008e-01 2.41070166e-01
-6.77942574e-01 -2.98877209e-01 9.16877508e-01 3.20929796e-01
5.02321303e-01 -2.60624945e-01 5.57682291e-02 -1.55822277e+00
1.15443161e-02 -9.34724569e-01 7.64678001e-01 -1.33683503e-01
-1.74990404e+00 2.61912227e-01 -1.67820901e-01 -1.84851539e+00
5.24830371e-02 -2.38233998e-01 -5.65366328e-01 1.04250062e+00
-1.47352374e+00 -1.18905771e+00 -5.05464137e-01 8.42442155e-01
5.37449419e-01 -5.14885604e-01 9.19863999e-01 7.61163414e-01
-8.34991336e-01 7.68585801e-01 6.98121071e-01 2.52154797e-01
1.05371213e+00 -1.12617028e+00 5.90207726e-02 8.91445279e-01
3.64929318e-01 1.14843726e+00 4.84693974e-01 -8.10804605e-01
-1.31673706e+00 -1.18037248e+00 6.32954240e-01 -2.30984867e-01
5.08825421e-01 -2.99715668e-01 -1.06615365e+00 6.19920731e-01
-1.11670913e-02 8.95312205e-02 1.23771095e+00 3.69613439e-01
-4.70947385e-01 -3.62306148e-01 -1.11107910e+00 4.05886143e-01
8.99775445e-01 -3.52683842e-01 -8.06878448e-01 2.05786973e-01
3.73861223e-01 -2.50349283e-01 -1.11955631e+00 -5.07576764e-02
6.42649353e-01 -7.63713956e-01 1.01590657e+00 -7.05127954e-01
3.85058910e-01 -5.80407500e-01 -2.51844108e-01 -1.56448328e+00
-4.78712469e-01 -1.80131570e-01 -2.64328241e-01 1.70053017e+00
1.21992573e-01 -7.00469851e-01 8.34355772e-01 4.39712971e-01
2.00861245e-01 -5.52876294e-01 -5.98506331e-01 -4.37432706e-01
-2.13349760e-01 1.50176167e-01 8.57017159e-01 1.14952099e+00
-9.28337127e-02 4.65415686e-01 -4.90570813e-01 4.26210076e-01
3.77139777e-01 6.16581440e-01 1.10799253e+00 -1.70766056e+00
-3.35997641e-01 1.49346255e-02 -4.36917275e-01 -1.31634891e+00
6.85180398e-03 -9.98296797e-01 -5.40804505e-01 -1.49884999e+00
-3.75532843e-02 -7.76116431e-01 -5.74366927e-01 1.85473427e-01
-7.46091828e-02 -2.13737227e-02 1.10681601e-01 5.55423796e-01
-5.66481173e-01 5.15051484e-01 9.87611532e-01 -7.95489177e-02
-5.07750213e-01 2.78188497e-01 -1.08339441e+00 6.11773014e-01
5.47697484e-01 -3.86132926e-01 -5.76758802e-01 -3.84360701e-01
9.27601159e-02 -2.22916171e-01 -4.05874513e-02 -1.00825644e+00
3.41075271e-01 -2.40139529e-01 7.68115759e-01 -4.55750138e-01
2.86989927e-01 -1.27243507e+00 3.05843115e-01 -2.07195170e-02
-2.45877028e-01 3.17281559e-02 -1.21138051e-01 1.01679182e+00
-4.02276009e-01 -1.35562629e-01 5.70720255e-01 -7.55816102e-02
-9.20253277e-01 4.31691438e-01 1.84982136e-01 -3.98848169e-02
1.07097721e+00 -2.10495502e-01 -2.16384023e-01 -1.64103672e-01
-7.22676277e-01 3.90583277e-01 5.53807676e-01 7.98186123e-01
6.20654523e-01 -1.53131020e+00 -6.21925831e-01 2.65688568e-01
4.45025682e-01 -2.69294113e-01 5.71383476e-01 5.34582198e-01
1.33910313e-01 3.10698092e-01 -3.76320243e-01 -2.99406528e-01
-1.25152850e+00 6.76254392e-01 1.83540523e-01 -4.06406909e-01
-5.33914506e-01 5.74696720e-01 2.01086238e-01 -3.67746234e-01
1.35377236e-02 2.74272233e-01 -8.77874315e-01 5.18278658e-01
6.94698274e-01 2.70505220e-01 4.87007201e-03 -8.73204112e-01
-2.51014650e-01 5.54856837e-01 -3.83451015e-01 2.58212388e-01
1.57278144e+00 -4.44726050e-01 2.44082510e-01 2.55151659e-01
1.05584824e+00 3.84945035e-01 -8.80669594e-01 -8.79601955e-01
-2.67220940e-02 -9.18969214e-01 -1.19378410e-01 -5.39492488e-01
-1.11194694e+00 7.37183154e-01 6.44293725e-01 2.14181989e-01
1.05643642e+00 -2.32629448e-01 8.11880469e-01 9.59105641e-02
4.80770707e-01 -1.11707211e+00 -1.24120854e-01 1.06527701e-01
5.94283700e-01 -1.18892097e+00 1.38663217e-01 -4.31066364e-01
-7.16357470e-01 1.06444132e+00 8.03008199e-01 -1.32158220e-01
7.74020910e-01 -3.33533823e-01 -1.20601796e-01 1.19689032e-02
-3.89738351e-01 -2.04121828e-01 4.52603877e-01 7.26896524e-01
4.63574231e-01 9.65313986e-02 -4.51979905e-01 1.09743440e+00
2.03986958e-01 -2.15087850e-02 1.02231652e-01 8.47604394e-01
-3.33145827e-01 -1.67055559e+00 -3.16583425e-01 6.37039483e-01
-3.22763205e-01 3.47599417e-01 -1.97121650e-01 3.82195532e-01
4.93537009e-01 8.63194108e-01 -1.79049030e-01 -5.36237121e-01
5.85472286e-01 -8.53180513e-03 9.92073864e-02 -8.78998935e-01
-5.38242579e-01 -9.62205157e-02 -1.62714064e-01 -2.69760013e-01
-5.40384650e-01 -6.62053287e-01 -8.68785858e-01 -1.25398651e-01
-4.32140261e-01 5.23825824e-01 3.56065929e-01 7.60441601e-01
6.62981033e-01 1.75149381e-01 9.57650244e-01 -6.63674772e-01
-5.64621389e-01 -6.74331009e-01 -8.38304818e-01 8.21883738e-01
1.87301692e-02 -1.24044943e+00 -1.35924682e-01 -1.45637244e-01] | [10.12230396270752, 5.527244567871094] |
ca06bfd7-4f53-4f13-abeb-96d968598cd2 | data-augmentation-for-skin-lesion-using-self | 1910.11960 | null | https://arxiv.org/abs/1910.11960v1 | https://arxiv.org/pdf/1910.11960v1.pdf | Data Augmentation for Skin Lesion using Self-Attention based Progressive Generative Adversarial Network | Deep Neural Networks (DNNs) show a significant impact on medical imaging. One significant problem with adopting DNNs for skin cancer classification is that the class frequencies in the existing datasets are imbalanced. This problem hinders the training of robust and well-generalizing models. Data Augmentation addresses this by using existing data more effectively. However, standard data augmentation implementations are manually designed and produce only limited reasonably alternative data. Instead, Generative Adversarial Networks (GANs) is utilized to generate a much broader set of augmentations. This paper proposes a novel enhancement for the progressive generative adversarial networks (PGAN) using self-attention mechanism. Self-attention mechanism is used to directly model the long-range dependencies in the feature maps. Accordingly, self-attention complements PGAN to generate fine-grained samples that comprise clinically-meaningful information. Moreover, the stabilization technique was applied to the enhanced generative model. To train the generative models, ISIC 2018 skin lesion challenge dataset was used to synthesize highly realistic skin lesion samples for boosting further the classification result. We achieve an accuracy of 70.1% which is 2.8% better than the non-augmented one of 67.3%. | ['Yousef Bassyouni Mahdy', 'Mamdouh Farouk Mohamed', 'Ibrahim Saad Ali'] | 2019-10-25 | null | null | null | null | ['skin-cancer-classification'] | ['medical'] | [ 6.44766092e-01 4.38816130e-01 -1.32228762e-01 -3.89788508e-01
-9.50227857e-01 -1.03931360e-01 5.45374930e-01 -7.50458315e-02
-2.61418790e-01 1.03375840e+00 3.34999591e-01 -3.01460195e-02
2.26271182e-01 -1.00345314e+00 -7.01427341e-01 -1.04248428e+00
4.37603742e-01 1.14881568e-01 -8.70028585e-02 -2.96235681e-01
-1.55418873e-01 3.88828784e-01 -1.23222816e+00 4.99536365e-01
1.34648561e+00 7.97203720e-01 2.30348036e-02 7.38897681e-01
-1.88554928e-01 6.18808508e-01 -8.27525854e-01 -4.42340165e-01
1.13704093e-01 -6.89362347e-01 -7.15836227e-01 -3.29000875e-02
3.80517036e-01 -4.01280403e-01 -3.64863902e-01 1.18240285e+00
7.16483951e-01 -1.36982858e-01 7.43199706e-01 -1.24106348e+00
-1.16719925e+00 5.54163873e-01 -7.19094515e-01 1.15380511e-01
-1.65188476e-01 3.30784291e-01 3.10689479e-01 -4.47450638e-01
4.87270057e-01 9.28450644e-01 7.32153952e-01 1.09493721e+00
-7.83491135e-01 -6.70760572e-01 -1.58742338e-01 5.94165698e-02
-1.11236608e+00 -1.55972764e-01 8.92262757e-01 -5.83053716e-02
4.35024828e-01 5.98479986e-01 7.11136222e-01 1.61542535e+00
2.84175992e-01 6.56512678e-01 1.34444022e+00 -4.47918475e-01
6.81866631e-02 2.69976109e-01 -5.57342283e-02 3.13172668e-01
2.61621356e-01 -5.68750501e-03 -1.33025795e-01 5.70267029e-02
9.81602073e-01 1.43391445e-01 -1.98384389e-01 2.28241697e-01
-7.35387802e-01 7.94477224e-01 9.64548767e-01 3.38097692e-01
-6.73136652e-01 3.71401906e-02 4.66991752e-01 -1.95374876e-01
4.88993555e-01 4.83209401e-01 -5.34762144e-02 1.16299197e-01
-6.93733037e-01 1.25363857e-01 1.39129460e-01 7.39691675e-01
2.60735571e-01 5.03316760e-01 -4.38895047e-01 1.00471580e+00
1.08589262e-01 4.08443123e-01 9.22015131e-01 -4.54632819e-01
2.40517452e-01 7.60037601e-01 -3.44783276e-01 -7.35882342e-01
-2.61076450e-01 -1.16748118e+00 -1.38210702e+00 7.08713606e-02
3.05040210e-01 -1.99489146e-01 -1.56362355e+00 1.83600390e+00
3.74750704e-01 1.21899240e-01 2.89439887e-01 8.80854309e-01
9.71594393e-01 4.44340974e-01 5.11616170e-01 6.76140487e-02
1.25160468e+00 -1.02233279e+00 -9.34477746e-01 -2.58352011e-01
3.92028242e-01 -6.97997391e-01 1.15933096e+00 1.83028430e-01
-9.44960833e-01 -6.76684558e-01 -1.11607361e+00 -3.11590522e-03
-2.80464798e-01 -7.64018521e-02 7.42845833e-01 7.85039008e-01
-9.37092006e-01 3.53524357e-01 -8.50871384e-01 -2.41246134e-01
9.96783912e-01 1.83763549e-01 -3.28992158e-01 -2.68037647e-01
-1.30769670e+00 6.85764670e-01 2.98179150e-01 2.31625244e-01
-8.53806794e-01 -8.92910898e-01 -7.23265529e-01 -2.35264570e-01
-5.76247089e-03 -7.88643420e-01 8.94675672e-01 -1.45906579e+00
-1.39310014e+00 7.03802049e-01 2.77117997e-01 -4.61626172e-01
6.81817472e-01 -2.09571391e-01 -5.22748113e-01 9.76126119e-02
4.51236330e-02 7.80210674e-01 7.87181497e-01 -1.18546975e+00
-1.44540250e-01 -3.82192671e-01 -1.89091340e-01 3.69653195e-01
-6.35640502e-01 -3.84915024e-01 -2.62839675e-01 -1.05798149e+00
-2.06010655e-01 -8.34610641e-01 -5.00606418e-01 -1.45843983e-01
-8.02390933e-01 2.40189716e-01 6.99374914e-01 -9.47167993e-01
1.02724218e+00 -1.88745296e+00 -2.34281451e-01 1.06032118e-01
1.86499342e-01 5.30081391e-01 -2.84243673e-01 8.76644775e-02
-5.10829568e-01 3.38325500e-01 -4.21202600e-01 -8.63596424e-02
-3.48802477e-01 2.61331886e-01 8.92668515e-02 2.03903511e-01
7.46183038e-01 1.15471852e+00 -7.34443069e-01 -4.10121500e-01
1.54580906e-01 7.97057092e-01 -6.87099993e-01 1.72303766e-01
-3.05605233e-02 6.77300632e-01 -3.64836007e-01 8.07754815e-01
8.96863937e-01 -1.22485273e-01 -1.05106622e-01 -4.62735176e-01
4.36097443e-01 -2.12047517e-01 -5.98795831e-01 1.66444123e+00
-3.32426608e-01 2.56920874e-01 -3.38135779e-01 -9.07055914e-01
8.50582480e-01 1.48130715e-01 4.20015991e-01 -6.77726328e-01
2.42051572e-01 -1.01718441e-01 3.54465753e-01 -5.83705187e-01
2.62171954e-01 -4.09247458e-01 9.13648903e-02 -1.57987662e-02
2.44852304e-02 8.23514387e-02 -1.37098789e-01 1.78551506e-02
1.17526829e+00 1.64431512e-01 1.39105007e-01 -1.41822353e-01
4.12990123e-01 -9.17862169e-03 6.26547873e-01 4.51361775e-01
-2.91251987e-01 8.93591821e-01 2.78816611e-01 -1.97144449e-01
-1.17219400e+00 -1.01399302e+00 -3.28924298e-01 5.91088116e-01
-2.07416847e-01 -3.05024092e-03 -1.08106101e+00 -8.83341134e-01
-3.35583150e-01 6.90586329e-01 -1.19228053e+00 -6.50248706e-01
-2.89736181e-01 -1.52919233e+00 7.39615262e-01 1.03098249e+00
8.44567358e-01 -1.04061425e+00 -1.74262196e-01 6.50110617e-02
-5.09406403e-02 -8.24893594e-01 -3.40297878e-01 2.30470803e-02
-8.42043042e-01 -9.59578335e-01 -9.60307717e-01 -5.77587366e-01
1.05089962e+00 -2.81626076e-01 8.02426517e-01 -2.71764398e-03
-4.95946318e-01 -1.94285139e-01 -4.40311164e-01 -6.82676017e-01
-8.10194194e-01 4.76601161e-02 -2.33906910e-01 5.71770221e-02
2.99434274e-01 -4.49597239e-01 -8.42444599e-01 -1.03835471e-01
-1.19089282e+00 3.68642241e-01 1.03833306e+00 1.33840251e+00
7.03493416e-01 4.30188142e-03 1.16794205e+00 -1.30269873e+00
5.48294306e-01 -5.55544198e-01 6.75419122e-02 1.72582567e-01
-5.35983503e-01 -4.95351963e-02 8.76235604e-01 -6.38315082e-01
-1.29879916e+00 3.06763258e-02 -6.01118922e-01 -3.86930943e-01
-2.48085931e-01 3.78742993e-01 -3.46405953e-01 1.26408217e-02
9.71852899e-01 2.96935946e-01 1.37960374e-01 -1.68936476e-01
2.54414976e-01 5.93983829e-01 7.02225089e-01 -2.74892688e-01
7.39853263e-01 2.52679616e-01 6.24140091e-02 -6.88388765e-01
-8.72625768e-01 8.90080407e-02 -3.93767536e-01 -1.45403882e-02
9.31192756e-01 -7.66532958e-01 1.06758736e-02 9.07778084e-01
-7.22809374e-01 -3.73523831e-01 -5.60682178e-01 3.29631746e-01
-2.64842719e-01 4.50602211e-02 -7.11580217e-01 -6.75280333e-01
-7.08831012e-01 -1.09564161e+00 8.65856767e-01 6.58622205e-01
-1.15643434e-01 -9.55608606e-01 -1.16637833e-01 5.36423743e-01
7.40558684e-01 8.12781155e-01 8.20609212e-01 -7.79249251e-01
-1.46338567e-01 -4.73135769e-01 -2.67714918e-01 7.06339836e-01
5.31848371e-01 5.70660196e-02 -1.27629244e+00 -1.54640511e-01
-2.29223166e-03 -3.19119245e-01 7.17210412e-01 4.75178391e-01
1.59665298e+00 -2.98392564e-01 -1.99489459e-01 7.12250113e-01
1.47270608e+00 3.23258221e-01 1.14623845e+00 1.11625768e-01
9.33196843e-01 3.57550889e-01 3.40872705e-01 1.17856011e-01
1.78436056e-01 1.33355960e-01 5.45008481e-01 -6.98861718e-01
-5.60437858e-01 -3.69335860e-01 -1.52491108e-01 6.91421509e-01
-2.73963541e-01 -2.82391846e-01 -8.84191871e-01 5.07774711e-01
-1.26954222e+00 -6.57640994e-01 -1.56930283e-01 1.81038523e+00
1.19437337e+00 1.15050711e-01 -1.46323487e-01 1.64861739e-01
8.71745408e-01 -1.58753559e-01 -8.86900365e-01 -1.40797853e-01
-2.09897965e-01 6.52743042e-01 5.11687219e-01 2.02159122e-01
-1.14260685e+00 6.52461767e-01 5.79807472e+00 9.16517198e-01
-1.31080019e+00 1.05771974e-01 1.15668738e+00 2.42311545e-02
-3.88391614e-01 -5.59391260e-01 -4.85507816e-01 6.67430937e-01
9.80618775e-01 3.03641204e-02 -5.64432293e-02 8.19643795e-01
-5.30012660e-02 1.71480626e-01 -6.96772158e-01 8.38855624e-01
1.97678596e-01 -1.30381727e+00 3.61679703e-01 1.12543136e-01
9.73087311e-01 -2.87244976e-01 4.07038331e-01 3.14460486e-01
2.61020601e-01 -1.31809247e+00 1.13724589e-01 7.70089865e-01
1.24255800e+00 -9.44466949e-01 1.14060974e+00 7.38170072e-02
-5.32207668e-01 1.67484045e-01 -2.77746081e-01 3.88128221e-01
-9.59933773e-02 5.76171219e-01 -1.11979651e+00 7.16756940e-01
4.99046654e-01 3.77327532e-01 -8.80749166e-01 8.68125796e-01
-3.14148404e-02 7.57529974e-01 -2.11265653e-01 5.07656485e-02
1.60109103e-01 1.21800944e-01 2.08231375e-01 9.13761735e-01
2.31399477e-01 -1.43410817e-01 -2.59931982e-01 7.56952286e-01
-3.23085606e-01 7.66299814e-02 -3.58481973e-01 -5.49607314e-02
2.46179432e-01 1.47971547e+00 -5.84718704e-01 -3.64538848e-01
-9.36345905e-02 1.05854762e+00 5.74295372e-02 2.11023256e-01
-1.06745780e+00 -4.13326532e-01 2.90888339e-01 2.39408582e-01
-2.90199012e-01 4.83895183e-01 -4.18124348e-01 -8.84367645e-01
-1.77172855e-01 -1.07870328e+00 3.07052553e-01 -7.37455249e-01
-1.73052287e+00 9.59901154e-01 -3.64878476e-01 -1.06838107e+00
-2.46838853e-01 -4.18629020e-01 -7.39244640e-01 1.19335222e+00
-1.54191315e+00 -1.65740514e+00 -9.45956230e-01 5.12767494e-01
5.35362422e-01 -1.66583389e-01 1.10487974e+00 4.68075931e-01
-7.97849536e-01 1.08698833e+00 -1.75240055e-01 2.55244792e-01
8.93148065e-01 -1.40363419e+00 2.64825046e-01 9.38326716e-01
-4.00789768e-01 4.92687732e-01 5.09487569e-01 -7.87392437e-01
-9.83279169e-01 -1.58231103e+00 1.01510920e-01 -3.18405211e-01
3.95447671e-01 -1.40603408e-01 -1.19919395e+00 4.60037231e-01
3.89636993e-01 3.05062264e-01 1.03628695e+00 -3.75198543e-01
-1.78516079e-02 -1.40500963e-01 -1.58455968e+00 5.62268794e-01
8.35593164e-01 -2.40021035e-01 -2.81331003e-01 2.79605091e-01
6.74054742e-01 -5.99903226e-01 -1.00066304e+00 7.27097690e-01
3.05552006e-01 -5.92215478e-01 8.90971482e-01 -9.07748401e-01
9.19810951e-01 8.62198789e-03 -5.88108487e-02 -1.49741113e+00
-2.84785956e-01 -2.36087829e-01 5.14729284e-02 1.33012187e+00
2.30179310e-01 -6.60121918e-01 1.08063996e+00 6.25432253e-01
-4.06353772e-01 -1.16497159e+00 -5.39229751e-01 -3.64852935e-01
4.04380023e-01 -1.06035218e-01 7.41322100e-01 1.21080387e+00
-4.46601868e-01 2.16399342e-01 -2.86799729e-01 4.28668819e-02
6.97062910e-01 -4.43639755e-01 4.97420162e-01 -7.37842321e-01
-1.05993345e-01 -2.07080022e-01 -4.14201856e-01 -3.60238910e-01
-1.44005239e-01 -9.44241881e-01 -2.17893854e-01 -1.47172272e+00
3.29448670e-01 -6.81363642e-01 -5.51546514e-01 6.21822596e-01
-6.46684408e-01 6.98603034e-01 -1.84777275e-01 -1.67437539e-01
4.58538160e-02 6.56056464e-01 1.73568964e+00 -1.17477521e-01
-2.27199364e-02 -4.10065055e-02 -1.20179188e+00 6.61720991e-01
1.14278030e+00 -8.53772834e-02 -5.34139633e-01 -1.71810061e-01
-2.33891726e-01 -1.90629423e-01 4.70406413e-01 -1.04499197e+00
-2.35790317e-03 -1.21050790e-01 1.04067063e+00 -5.57077706e-01
2.13965341e-01 -4.84440207e-01 2.24973679e-01 6.02299452e-01
-2.96820313e-01 -2.79387593e-01 4.87849444e-01 4.08676505e-01
-2.23194569e-01 -2.16876060e-01 1.03032231e+00 -7.36095011e-02
-3.78886908e-01 4.87140626e-01 -8.70087370e-03 -7.59698497e-03
1.21660924e+00 -1.35568097e-01 -4.83877182e-01 -1.47547379e-01
-7.39630997e-01 -5.56985475e-02 3.32081169e-01 3.02824110e-01
6.45244062e-01 -1.60564089e+00 -8.98542106e-01 3.80597144e-01
-1.66501720e-02 3.47375929e-01 9.60701346e-01 5.94159067e-01
-6.20172501e-01 2.91999094e-02 -5.96719861e-01 -4.07855481e-01
-9.89660025e-01 4.53907430e-01 4.80224043e-01 -4.88259137e-01
-4.24517125e-01 9.73712862e-01 3.37966233e-01 -2.02582777e-01
-1.52563736e-01 -2.89722562e-01 -1.96653590e-01 -1.80474833e-01
5.59537768e-01 1.69339806e-01 6.53189346e-02 -2.74275988e-01
-1.79122120e-01 1.54072836e-01 -3.34779561e-01 3.12972665e-01
1.44724429e+00 3.05542737e-01 1.11729145e-01 -7.71625564e-02
1.02666879e+00 -3.84365693e-02 -1.32724559e+00 5.06083161e-05
-5.83441675e-01 -2.85049051e-01 1.38516411e-01 -1.29089344e+00
-1.34717035e+00 8.13446701e-01 8.87764037e-01 -3.41566000e-03
1.53926826e+00 -2.51443893e-01 8.81886184e-01 -2.99346298e-01
-2.39492029e-01 -8.77935052e-01 2.76892066e-01 -7.85276070e-02
1.14647341e+00 -1.28450549e+00 -2.32609481e-01 -5.17067075e-01
-7.52676010e-01 9.21267152e-01 1.09045863e+00 -3.36477727e-01
2.29267076e-01 4.77636725e-01 1.34741783e-01 -1.54211093e-02
-4.53450203e-01 8.58168975e-02 2.91943967e-01 8.77825618e-01
3.22902173e-01 2.19374374e-02 -1.46324590e-01 9.69929159e-01
-2.71604478e-01 4.36553769e-02 4.64591593e-01 7.03742206e-01
1.89417168e-01 -1.10044658e+00 -3.81051689e-01 8.06507707e-01
-7.93386221e-01 -2.57128984e-01 -1.62844405e-01 8.93127441e-01
2.02324599e-01 4.51026887e-01 2.55474374e-02 -4.26065385e-01
1.55200720e-01 4.55449969e-02 5.02767384e-01 -5.63995898e-01
-4.93213147e-01 1.05402805e-01 -1.06632434e-01 -2.12208837e-01
-3.33427966e-01 -3.62925649e-01 -1.15832520e+00 3.14324237e-02
-2.81713277e-01 -1.54664427e-01 7.06330478e-01 5.14747977e-01
2.60877699e-01 1.40783274e+00 5.35082757e-01 -2.82979786e-01
-5.32458663e-01 -1.32210386e+00 -2.56075233e-01 6.33172452e-01
1.09685697e-01 -4.43994343e-01 -9.65827182e-02 7.93630928e-02] | [14.198902130126953, -1.990584135055542] |
ec7a83d6-940c-47f5-b290-0a2ebbec288e | radiologist-level-stroke-classification-on | 2003.14287 | null | https://arxiv.org/abs/2003.14287v1 | https://arxiv.org/pdf/2003.14287v1.pdf | Radiologist-level stroke classification on non-contrast CT scans with Deep U-Net | Segmentation of ischemic stroke and intracranial hemorrhage on computed tomography is essential for investigation and treatment of stroke. In this paper, we modified the U-Net CNN architecture for the stroke identification problem using non-contrast CT. We applied the proposed DL model to historical patient data and also conducted clinical experiments involving ten experienced radiologists. Our model achieved strong results on historical data, and significantly outperformed seven radiologist out of ten, while being on par with the remaining three. | ['Vladimir Kokh', 'Dmitry Umerenkov', 'Alex Tuzhilin', 'Manvel Avetisian'] | 2020-03-31 | null | null | null | null | ['stroke-classification'] | ['methodology'] | [-2.65682697e-01 -1.81792319e-01 -1.19985245e-01 -4.37008649e-01
-8.82339060e-01 -4.24117476e-01 3.04878443e-01 1.76908076e-02
-1.04638302e+00 9.15027201e-01 4.00277436e-01 -8.21396589e-01
-2.73561738e-02 -8.35713029e-01 -3.77564698e-01 -4.17987257e-01
-4.53784585e-01 8.68876815e-01 4.52638090e-01 2.62271106e-01
-1.99018627e-01 9.99829531e-01 -7.26507962e-01 3.89138728e-01
5.07472873e-01 8.82430553e-01 -8.67866278e-02 6.58482790e-01
4.12371121e-02 1.04246998e+00 -3.07791680e-01 -2.96316762e-02
4.03302491e-01 -2.84114093e-01 -9.57665682e-01 -7.92972669e-02
2.76063412e-01 -8.30166578e-01 -9.71958101e-01 7.94994295e-01
1.20094395e+00 3.40435565e-01 7.68943131e-01 -6.65642619e-01
-6.00295186e-01 5.91387093e-01 -3.70264739e-01 1.31904304e+00
-8.75046908e-07 -1.73818514e-01 3.73229504e-01 -8.62715542e-01
6.69793248e-01 6.69194162e-01 4.40702647e-01 7.10356712e-01
-2.94566274e-01 -6.70284986e-01 -7.91267976e-02 4.40285712e-01
-1.16336071e+00 -4.39520404e-02 1.45581797e-01 -3.99338186e-01
9.43689525e-01 9.63055715e-02 1.15456688e+00 9.66367185e-01
2.30736792e-01 1.24245048e+00 9.90225017e-01 -1.55746534e-01
3.71117860e-01 -5.30387819e-01 8.48111808e-01 7.00892091e-01
3.26137543e-01 3.76111120e-01 1.84059829e-01 -3.96258205e-01
1.19662535e+00 2.09189087e-01 -7.42856622e-01 -3.67461815e-02
-8.34389210e-01 9.46101546e-01 7.33432353e-01 5.09876966e-01
-5.09328544e-01 -1.32935971e-01 3.21268827e-01 -7.00031668e-02
2.36946061e-01 2.12643091e-02 -2.88044840e-01 -3.28896418e-02
-9.64540780e-01 3.00099440e-02 5.28765082e-01 7.02176452e-01
-4.79477167e-01 1.70440301e-01 -4.57575530e-01 5.74952781e-01
-3.69112760e-01 2.69634485e-01 8.97914290e-01 -5.09331644e-01
1.60630226e-01 3.74952346e-01 -1.36175439e-01 -2.23460495e-01
-1.23676479e+00 -7.51516163e-01 -1.09916008e+00 2.94359505e-01
5.84638298e-01 -6.55635536e-01 -1.72492588e+00 1.06808352e+00
-2.77377158e-01 3.64714146e-01 -1.68369055e-01 1.23527026e+00
1.30191970e+00 -4.50411253e-03 1.79438531e-01 -1.84462607e-01
1.34826922e+00 -8.75129461e-01 -5.49467385e-01 -1.98059365e-01
5.53177476e-01 -2.84583598e-01 6.32279277e-01 1.91034392e-01
-1.09897864e+00 8.08720663e-03 -6.25589430e-01 1.26853138e-01
-8.45596418e-02 2.54063755e-01 5.08111477e-01 6.10290229e-01
-1.35178137e+00 3.98551255e-01 -1.32337987e+00 -5.25427639e-01
1.09551406e+00 3.33505332e-01 -1.77010357e-01 1.74831346e-01
-1.09787917e+00 1.42366087e+00 3.32019448e-01 -2.53199320e-02
-9.51076567e-01 -5.59801698e-01 -3.95411700e-01 1.38140321e-02
9.13586840e-02 -9.05336916e-01 1.43625677e+00 3.76423239e-03
-1.35149479e+00 9.57211614e-01 -1.61398023e-01 -9.15957332e-01
1.18704557e+00 -2.37829819e-01 -3.89594704e-01 6.66857302e-01
-9.56758931e-02 6.35716915e-01 2.12167069e-01 -1.01404464e+00
-7.02040136e-01 -4.74716276e-01 -1.69718459e-01 -3.33753414e-02
1.00170188e-01 1.31199405e-01 -2.18909696e-01 -6.24367535e-01
4.49843138e-01 -7.42243826e-01 -6.74822450e-01 -1.50839955e-01
-1.75653413e-01 2.92842947e-02 8.03133190e-01 -8.09596777e-01
6.86520934e-01 -1.36862993e+00 -2.63690740e-01 3.52615029e-01
7.83461213e-01 4.77167666e-01 2.35371470e-01 -5.39082110e-01
-3.30759317e-01 -2.35024025e-03 -2.32610583e-01 2.01264888e-01
-6.46288753e-01 1.16531901e-01 -1.00091971e-01 6.25648916e-01
-1.36618823e-01 1.32388210e+00 -8.58571291e-01 -5.71394861e-01
4.27268416e-01 4.85257655e-01 -2.15450212e-01 2.57591665e-01
4.77952987e-01 8.73815417e-01 -7.92544842e-01 5.35657048e-01
4.54795182e-01 -3.70356292e-01 -2.53364563e-01 -4.42006737e-02
1.79192014e-02 -3.60089004e-01 -5.76056361e-01 1.28606784e+00
-2.24662125e-02 7.12626517e-01 -2.44220734e-01 -9.68942821e-01
4.51281458e-01 6.05049312e-01 8.96853268e-01 -8.09845686e-01
1.04619730e+00 3.36819999e-02 3.67072165e-01 -5.95496655e-01
-2.59605408e-01 -1.07452720e-01 4.59342629e-01 6.74802601e-01
-3.36365998e-01 7.30622038e-02 -2.69265641e-02 1.70194745e-01
1.29741049e+00 -6.21039748e-01 3.00769180e-01 -3.05139989e-01
3.24664772e-01 1.63721770e-01 1.91787899e-01 1.35390079e+00
-9.38698053e-01 8.93654168e-01 3.48304093e-01 -9.18618083e-01
-8.01280379e-01 -1.41403282e+00 -6.37043834e-01 5.30161083e-01
3.07923555e-03 2.25630939e-01 -9.43053007e-01 -5.32793462e-01
-3.72170746e-01 4.73222703e-01 -6.41377866e-01 2.57665701e-02
-9.90407348e-01 -1.11436725e+00 7.47980297e-01 1.17416263e+00
8.67834032e-01 -1.06042147e+00 -1.26197064e+00 4.73936200e-01
-3.23533893e-01 -1.26131332e+00 -3.05272996e-01 2.75721818e-01
-1.07037437e+00 -1.50587821e+00 -1.36515343e+00 -1.11042070e+00
4.39653099e-01 1.97045654e-01 9.14611638e-01 1.95952341e-01
-6.71041369e-01 3.83757293e-01 -3.79995048e-01 -6.24644756e-01
8.15672204e-02 2.47880593e-02 -8.06301087e-02 -6.02184832e-01
2.82364547e-01 -3.82599324e-01 -1.04486036e+00 -1.37029598e-02
-7.11216569e-01 -2.35369250e-01 5.19753575e-01 7.47152209e-01
6.00908458e-01 -1.73866212e-01 6.20812774e-01 -6.38926506e-01
9.75888371e-01 -2.34709844e-01 -1.82762146e-01 2.17798963e-01
-2.89943159e-01 -3.76092106e-01 4.48286712e-01 -2.65081853e-01
-8.77378404e-01 6.21156842e-02 -1.44281089e-01 -3.20395291e-01
-2.40963832e-01 3.95412534e-01 3.57387215e-01 -4.22337592e-01
7.63825536e-01 -4.33106609e-02 -2.70321995e-01 -3.86800081e-01
1.74905494e-01 6.33321941e-01 1.12035453e+00 -4.00169283e-01
7.45812282e-02 8.53042305e-01 4.15683240e-02 -8.18938076e-01
-7.26236403e-01 -3.52928907e-01 -8.00983250e-01 -2.63735801e-01
1.13197970e+00 -5.94130754e-01 -2.60362715e-01 6.68258965e-01
-1.28076482e+00 -3.68334591e-01 -9.12395865e-02 8.60137641e-01
-2.99112797e-01 4.70632851e-01 -1.01352870e+00 -1.55451104e-01
-7.03147411e-01 -1.50333905e+00 4.03349519e-01 2.44163662e-01
2.33880699e-01 -9.47333336e-01 -1.51512548e-01 5.96307553e-02
6.46944046e-01 5.04723191e-01 1.04118431e+00 -7.45054364e-01
-1.49892598e-01 -6.28336489e-01 -8.12149048e-01 -1.42472312e-02
8.01996142e-02 -4.09353942e-01 -7.99485207e-01 -3.34895402e-01
3.81326079e-02 -5.92194982e-02 1.04900455e+00 1.26974988e+00
1.45669210e+00 6.61363959e-01 -3.93932909e-01 6.81590796e-01
9.28991914e-01 5.91482401e-01 5.18884361e-01 5.63909471e-01
5.97254872e-01 -3.99026461e-02 -4.45915550e-01 4.14469510e-01
2.49668494e-01 7.98858702e-02 2.19171077e-01 -4.51358944e-01
-2.29267448e-01 5.90514719e-01 -3.75133932e-01 3.86218369e-01
-3.56413037e-01 -3.26330662e-01 -1.39264441e+00 5.28318286e-01
-1.70934308e+00 -6.52981520e-01 -3.89764756e-01 1.59718287e+00
3.11907113e-01 3.02862465e-01 1.12994678e-01 -1.79657266e-01
7.79462457e-01 -3.19685757e-01 -6.84948266e-01 3.41693908e-02
-2.31746718e-01 9.07095909e-01 8.77008796e-01 4.06554133e-01
-1.47068298e+00 8.98757875e-01 7.52144289e+00 2.88370490e-01
-1.16746414e+00 4.82153237e-01 6.67397022e-01 -1.19290881e-01
2.86553741e-01 -4.74002719e-01 -1.67282000e-01 5.74421994e-02
5.68304360e-01 -3.98327112e-01 -8.41032788e-02 6.89883888e-01
2.95238644e-01 -2.66662776e-01 -9.33232903e-01 8.47261488e-01
-1.18615918e-01 -1.53600490e+00 1.28109455e-01 -3.15026313e-01
7.21154630e-01 7.41764963e-01 -3.10132980e-01 1.79343283e-01
4.49915707e-01 -1.16420257e+00 3.19664836e-01 3.76338214e-01
5.17906666e-01 -6.11511290e-01 1.09276140e+00 3.25321406e-01
-9.11112905e-01 1.54467404e-01 -2.13175789e-01 7.14285895e-02
6.59410477e-01 1.44039363e-01 -6.77376091e-01 2.46491209e-01
1.03678048e+00 5.26474833e-01 -4.69632179e-01 1.70617735e+00
-4.15900528e-01 6.70590937e-01 -3.17485660e-01 3.38594526e-01
7.62690604e-01 4.78239320e-02 5.06191194e-01 1.29046464e+00
1.10561773e-01 8.38920295e-01 3.89741689e-01 3.65036935e-01
4.00062129e-02 1.92212850e-01 -4.53066885e-01 7.69945204e-01
4.57868576e-02 9.40535843e-01 -1.30871975e+00 -9.02863920e-01
-3.12123775e-01 5.71524858e-01 2.42654562e-01 4.15179521e-01
-8.51761520e-01 -9.61448252e-02 -7.45478049e-02 -7.05474988e-02
3.30440886e-02 -1.00029863e-01 -8.45587969e-01 -1.36495090e+00
-1.55819491e-01 -2.35928640e-01 8.01537216e-01 -8.26541603e-01
-1.21804047e+00 1.31403899e+00 1.26342371e-01 -1.01572239e+00
6.00266084e-02 -7.42088318e-01 -1.09036076e+00 7.57917941e-01
-1.64309335e+00 -6.86559260e-01 -6.75177574e-01 9.13332343e-01
2.84687161e-01 -3.97138357e-01 7.06008255e-01 3.88554513e-01
-7.22514212e-01 3.49038869e-01 -2.62727644e-02 6.78204775e-01
3.46528560e-01 -1.05038881e+00 4.55875874e-01 1.00157547e+00
-4.18020040e-01 4.54257965e-01 3.09653014e-01 -6.73303962e-01
-5.10302007e-01 -1.24002159e+00 4.91815031e-01 -2.72505105e-01
7.99047291e-01 5.72773933e-01 -9.54619706e-01 8.51915896e-01
1.99864835e-01 2.75292337e-01 4.90573317e-01 -8.11005592e-01
2.02741742e-01 8.40869486e-01 -1.30489790e+00 5.30000865e-01
1.06429422e+00 -2.64200658e-01 -9.79864299e-01 6.35439277e-01
3.59838724e-01 -9.35484946e-01 -7.88066924e-01 6.06329918e-01
6.60467029e-01 -4.61747438e-01 1.13461697e+00 -1.30155098e+00
2.89586812e-01 2.56363779e-01 1.61862895e-01 -1.23870885e+00
-4.23557937e-01 1.12058312e-01 1.45468667e-01 9.11421701e-02
2.05751643e-01 -6.91833019e-01 1.20601332e+00 8.76230597e-01
-5.49952209e-01 -7.83540130e-01 -1.05243325e+00 -6.66556656e-01
8.09118629e-01 -6.33442521e-01 2.54727751e-01 6.45839453e-01
-1.82629436e-01 -1.82442114e-01 -4.28335629e-02 2.65948564e-01
7.91842937e-01 -7.78541565e-02 1.42645277e-02 -1.10079634e+00
1.73038259e-01 -9.30280507e-01 -5.00762284e-01 -7.05993950e-01
3.65059316e-01 -1.26894784e+00 -4.16101664e-01 -2.05022812e+00
3.79838079e-01 -1.12607896e-01 -5.43829024e-01 5.42200446e-01
-1.90519542e-01 4.15709764e-01 -5.83463944e-02 3.39149058e-01
6.31280690e-02 2.38993421e-01 1.69243920e+00 -3.29188377e-01
-2.48557717e-01 3.76597375e-01 -3.20638776e-01 1.01376271e+00
1.13071883e+00 -4.14740652e-01 -1.61277741e-01 -6.72392905e-01
-6.78030849e-01 3.27496797e-01 6.75526977e-01 -1.15171623e+00
3.97220463e-01 2.35560134e-01 7.71512151e-01 -8.46275508e-01
-1.67877406e-01 -7.19231963e-01 -5.50329447e-01 9.36822474e-01
-2.89695054e-01 2.71406829e-01 3.07785213e-01 1.49822757e-01
-1.32075444e-01 -3.60892713e-02 1.10368836e+00 -4.35252488e-01
-8.07597697e-01 7.98789501e-01 -9.73522425e-01 3.27694982e-01
9.78003740e-01 -1.54340584e-02 -5.19065857e-01 -4.66226935e-01
-1.40275621e+00 4.38389868e-01 -2.43602812e-01 1.03399642e-01
9.38098729e-01 -1.07440424e+00 -8.61950636e-01 6.49788380e-02
-2.07109407e-01 7.57032037e-02 8.37008059e-02 1.11046922e+00
-1.19104433e+00 7.79993415e-01 -7.02009618e-01 -4.18528944e-01
-8.92060637e-01 -7.09386216e-03 8.28372061e-01 -2.73533434e-01
-1.33677864e+00 7.53300607e-01 -7.87480697e-02 -1.32689461e-01
4.67452139e-01 -7.07782507e-01 -5.06462336e-01 -1.10071994e-01
5.93244433e-01 5.38165748e-01 4.94597822e-01 -5.27360737e-01
-5.59552908e-01 3.00835967e-01 -4.29645181e-01 -1.35134265e-01
1.41339147e+00 2.92494804e-01 -9.16398596e-04 -4.08985883e-01
1.18322349e+00 -1.00719488e+00 -7.90126562e-01 -3.93381327e-01
1.28442913e-01 4.23948728e-02 7.66952038e-01 -7.88774669e-01
-1.66566277e+00 9.12830591e-01 1.13238466e+00 -3.09837192e-01
7.89225161e-01 -1.26223182e-02 1.31949914e+00 5.83779216e-01
5.67453921e-01 -7.82126248e-01 -2.04578474e-01 5.57128429e-01
5.04816592e-01 -1.38698852e+00 3.82219926e-02 -7.79206753e-02
-6.47921443e-01 1.37013924e+00 5.74660182e-01 -4.87936348e-01
1.15048397e+00 4.81038451e-01 4.86232638e-01 -4.90714878e-01
-1.16883971e-01 -4.28909868e-01 4.05780882e-01 4.91198182e-01
3.63488406e-01 3.61077517e-01 -5.85370719e-01 7.32766330e-01
-1.96765631e-01 5.45771658e-01 7.80927479e-01 1.43158865e+00
-6.43981636e-01 -6.49822652e-01 -1.00177653e-01 9.85723138e-01
-6.15533352e-01 -1.77117288e-01 -3.68168086e-01 9.44461405e-01
-1.05355509e-01 8.15909684e-01 -6.72920942e-02 -7.75638446e-02
6.25333250e-01 -8.18819925e-02 5.47531366e-01 -3.22077930e-01
-6.79820716e-01 1.44328341e-01 -1.95249036e-01 -4.06406283e-01
-2.16834053e-01 -7.83790290e-01 -1.65922225e+00 5.57451546e-02
-7.58164823e-02 -8.23754631e-03 2.35085279e-01 1.11596525e+00
-1.63927987e-01 6.94437981e-01 3.16918463e-01 -5.13048589e-01
-2.67142713e-01 -8.93403590e-01 -5.96704662e-01 1.06680036e-01
4.02494431e-01 -7.53935695e-01 -3.74544472e-01 -7.46251196e-02] | [14.327607154846191, -2.092195749282837] |
13d95f5e-c502-40ca-b72c-717f83fa4d1f | cross-field-transformer-for-diabetic | 2211.14552 | null | https://arxiv.org/abs/2211.14552v2 | https://arxiv.org/pdf/2211.14552v2.pdf | Cross-Field Transformer for Diabetic Retinopathy Grading on Two-field Fundus Images | Automatic diabetic retinopathy (DR) grading based on fundus photography has been widely explored to benefit the routine screening and early treatment. Existing researches generally focus on single-field fundus images, which have limited field of view for precise eye examinations. In clinical applications, ophthalmologists adopt two-field fundus photography as the dominating tool, where the information from each field (i.e.,macula-centric and optic disc-centric) is highly correlated and complementary, and benefits comprehensive decisions. However, automatic DR grading based on two-field fundus photography remains a challenging task due to the lack of publicly available datasets and effective fusion strategies. In this work, we first construct a new benchmark dataset (DRTiD) for DR grading, consisting of 3,100 two-field fundus images. To the best of our knowledge, it is the largest public DR dataset with diverse and high-quality two-field images. Then, we propose a novel DR grading approach, namely Cross-Field Transformer (CrossFiT), to capture the correspondence between two fields as well as the long-range spatial correlations within each field. Considering the inherent two-field geometric constraints, we particularly define aligned position embeddings to preserve relative consistent position in fundus. Besides, we perform masked cross-field attention during interaction to flter the noisy relations between fields. Extensive experiments on our DRTiD dataset and a public DeepDRiD dataset demonstrate the effectiveness of our CrossFiT network. The new dataset and the source code of CrossFiT will be publicly available at https://github.com/FDU-VTS/DRTiD. | ['Rui Feng', 'Wenwen Xue', 'Lina Lu', 'Haidong Zou', 'Yuejie Zhang', 'Rui-Wei Zhao', 'Fan Xiao', 'Jilan Xu', 'Junlin Hou'] | 2022-11-26 | null | null | null | null | ['diabetic-retinopathy-grading'] | ['medical'] | [-2.22338557e-01 -3.90665203e-01 -2.83756088e-02 -5.88332355e-01
-5.03945231e-01 -3.60557616e-01 1.49605647e-01 -4.08344090e-01
-1.09418325e-01 7.40431428e-01 4.42649186e-01 -2.76685625e-01
-3.97147954e-01 -6.81388140e-01 -3.51141781e-01 -7.22411752e-01
4.15993571e-01 -2.04999462e-01 2.11130187e-01 -2.02389464e-01
2.32762128e-01 3.56683075e-01 -1.54081285e+00 2.26977304e-01
1.43380415e+00 1.04344738e+00 3.32300246e-01 3.10951442e-01
1.92425415e-01 5.81565976e-01 -2.83506453e-01 -4.94557261e-01
4.20189410e-01 -2.82720208e-01 -5.03114998e-01 1.87751010e-01
1.02092969e+00 -6.75293863e-01 -7.24072516e-01 1.31736326e+00
1.02990830e+00 -6.13304228e-02 3.06966662e-01 -7.39425123e-01
-1.10169017e+00 1.69723108e-02 -8.59109640e-01 6.31104290e-01
4.00036424e-02 5.72139025e-01 6.41918063e-01 -5.45819700e-01
4.39570546e-01 9.81724799e-01 3.15530658e-01 2.41965920e-01
-1.00201321e+00 -5.72286069e-01 2.09933594e-02 3.22493285e-01
-1.37607396e+00 -4.62519854e-01 5.99264324e-01 -6.81688428e-01
4.59173709e-01 6.53021038e-02 8.24065208e-01 5.69397211e-01
2.72942513e-01 4.40579027e-01 1.50285375e+00 -1.59146324e-01
-2.54963487e-01 -1.34560406e-01 2.26484179e-01 5.35903454e-01
4.61001128e-01 2.89770484e-01 9.65966508e-02 9.86400694e-02
9.16765988e-01 1.51244774e-01 -8.93752396e-01 -9.02673528e-02
-1.05229926e+00 5.63315928e-01 6.98755682e-01 1.07198507e-01
-2.51874357e-01 -1.65600792e-01 1.37081653e-01 1.27428874e-01
4.70238239e-01 1.78469136e-01 -2.47179702e-01 1.21866845e-01
-4.44516957e-01 1.12569556e-01 3.68202142e-02 7.61691093e-01
6.45844221e-01 -3.48259777e-01 -4.70920652e-01 1.01467955e+00
3.28262269e-01 4.92971927e-01 4.22518432e-01 -8.16125274e-01
4.86558110e-01 9.56697881e-01 9.75641832e-02 -9.18654263e-01
-4.63207155e-01 -4.28819507e-01 -9.69007194e-01 4.30420995e-01
4.14462954e-01 -2.67860413e-01 -9.58795130e-01 1.26967752e+00
2.99066603e-01 3.21825325e-01 -2.40593567e-01 1.40165365e+00
1.15836537e+00 1.20914295e-01 -3.36661488e-01 -1.56260669e-01
1.37548339e+00 -7.65860200e-01 -6.87394977e-01 -1.31161019e-01
6.23759985e-01 -9.37046170e-01 9.37842488e-01 3.51315618e-01
-9.98716235e-01 -6.39492512e-01 -7.86372125e-01 -5.79345405e-01
-1.06431190e-02 9.34926867e-01 7.77861416e-01 2.61569381e-01
-1.07148838e+00 1.86461121e-01 -7.15872228e-01 -2.80715466e-01
7.70892560e-01 2.66475230e-01 -6.16882563e-01 -6.07525647e-01
-9.70296621e-01 7.99613237e-01 -1.24144390e-01 4.38267350e-01
-2.51293797e-02 -8.22340906e-01 -8.41727257e-01 -2.85590410e-01
8.45374689e-02 -9.02834177e-01 9.50883985e-01 -5.68649828e-01
-1.21853840e+00 1.01279426e+00 -2.92976737e-01 -4.36352976e-02
4.43907619e-01 -1.60614789e-01 -7.43747652e-01 2.40766615e-01
3.59690599e-02 3.64394128e-01 2.90663600e-01 -8.55880201e-01
-6.88146591e-01 -8.31230879e-01 2.37967238e-01 1.81271851e-01
-1.96867739e-03 1.63942710e-01 -5.32603920e-01 -4.80972946e-01
8.99980739e-02 -6.99574590e-01 -4.38376628e-02 2.72343129e-01
-6.26768410e-01 -2.03433737e-01 5.31967461e-01 -7.49868512e-01
1.37361538e+00 -2.15221786e+00 -1.95786774e-01 -1.16359480e-01
8.40204656e-01 6.68465078e-01 -5.76395281e-02 -1.54117167e-01
-2.71073788e-01 -1.15425669e-01 1.49500400e-01 -1.27832130e-01
-4.90158588e-01 -1.40404359e-01 3.20897810e-02 7.26996720e-01
2.17988431e-01 8.95969510e-01 -8.27127516e-01 -4.21778977e-01
3.21498960e-01 6.26185834e-01 -3.59196991e-01 9.54336151e-02
3.03339064e-01 5.13559878e-01 -6.50549531e-01 7.35492826e-01
1.09242630e+00 -5.34962535e-01 -1.84256673e-01 -8.00808668e-01
-3.84238213e-01 -1.03278644e-03 -8.59845817e-01 1.56598496e+00
-9.14761275e-02 7.79156923e-01 -3.94476175e-01 -4.74936128e-01
9.98995066e-01 -2.70060659e-03 3.51674676e-01 -9.18380558e-01
3.43850225e-01 2.50184000e-01 2.92168021e-01 -8.58537614e-01
2.46417020e-02 2.26678386e-01 6.24819100e-01 -2.13562902e-02
-1.32926717e-01 3.00229102e-01 3.92930657e-01 -1.68414190e-01
8.42687309e-01 4.65228446e-02 3.86010110e-01 1.43547440e-02
5.22144854e-01 -2.62941420e-01 7.37931013e-01 1.19339548e-01
-4.40895170e-01 1.18771839e+00 5.96807718e-01 -5.84858239e-01
-7.60358691e-01 -8.27669859e-01 -6.58091307e-01 2.81363949e-02
5.23277164e-01 -2.15829596e-01 -3.95488083e-01 -5.69092870e-01
1.98505327e-01 8.83346498e-02 -7.59106040e-01 5.08641973e-02
-4.11096811e-01 -8.68782401e-01 2.55211562e-01 4.99056846e-01
8.01544845e-01 -4.36497569e-01 -1.33232906e-01 -1.44056618e-01
-9.87819135e-02 -8.98863912e-01 -7.38674164e-01 -8.43798220e-01
-6.62577152e-01 -1.61202407e+00 -1.14021397e+00 -7.08177209e-01
7.28956819e-01 6.81463957e-01 9.16016936e-01 1.44273251e-01
-6.32395744e-01 -2.45463118e-01 -1.61345527e-01 -1.16073750e-01
2.75786996e-01 -3.76614332e-01 -1.32498518e-01 4.37902868e-01
6.62465394e-01 -7.24092782e-01 -1.36511636e+00 5.57429075e-01
-5.70071042e-01 1.56798371e-04 9.68215227e-01 8.75337183e-01
7.92501569e-01 -6.21643253e-02 2.30696276e-01 -5.98582685e-01
5.07402003e-01 -3.33924264e-01 -9.89251673e-01 4.09840941e-01
-4.91073668e-01 -3.79320711e-01 4.00869042e-01 -1.99492782e-01
-1.13892663e+00 -1.78622112e-01 1.81470037e-01 -6.52969837e-01
-2.60020077e-01 3.54412884e-01 -2.11428225e-01 -4.78458822e-01
8.45476389e-01 4.68181968e-02 1.98266566e-01 -6.63728595e-01
1.90156773e-01 9.39842045e-01 6.65489912e-01 -7.33510554e-02
3.42911005e-01 6.30205810e-01 4.67423461e-02 -4.63395655e-01
-1.01183093e+00 -6.38001978e-01 -4.24995929e-01 -7.78399333e-02
9.08296525e-01 -1.06943023e+00 -8.87124777e-01 7.92931080e-01
-1.02478385e+00 -6.39474951e-03 1.45532787e-01 9.34464693e-01
-1.89006254e-01 5.54030240e-01 -3.31423432e-01 -3.29716027e-01
-2.19522297e-01 -1.40869963e+00 1.06838799e+00 7.78474092e-01
5.57390094e-01 -8.89601171e-01 2.14073598e-01 6.20112240e-01
2.69811124e-01 9.00991186e-02 6.71789467e-01 1.06955752e-01
-7.87374079e-01 -6.16626814e-02 -9.14244473e-01 6.53567076e-01
5.11821270e-01 9.28842798e-02 -9.05373931e-01 -1.42942607e-01
-3.73957396e-01 -6.44169301e-02 9.67151523e-01 8.61378729e-01
1.15284014e+00 8.64126012e-02 -3.32737029e-01 1.06926751e+00
1.46826291e+00 1.86984301e-01 1.13483059e+00 1.54912338e-01
7.89410293e-01 5.00578463e-01 5.82924724e-01 2.96825320e-01
6.46898568e-01 8.25035453e-01 1.81764007e-01 -5.29547334e-01
-6.45841420e-01 3.40599157e-02 -1.87113136e-01 4.42168683e-01
-5.21740794e-01 -2.82668263e-01 -8.25272977e-01 5.55009425e-01
-1.71067917e+00 -8.40553641e-01 -4.70282614e-01 2.22176933e+00
8.63098323e-01 -2.81046510e-01 -8.27000216e-02 -3.33959907e-01
9.12012815e-01 -4.83920164e-02 -7.01467931e-01 4.10540193e-01
-5.30893028e-01 5.05997613e-02 5.83653331e-01 3.42563689e-01
-1.15968895e+00 4.79481488e-01 5.08861017e+00 6.64959967e-01
-1.35738039e+00 -7.82125816e-02 7.55217671e-01 -3.47536981e-01
-1.46928549e-01 -6.84117600e-02 -8.59781504e-01 8.09168756e-01
-7.55433785e-03 -9.37487707e-02 3.21038514e-01 3.73474807e-01
5.29805243e-01 -1.49146393e-01 -6.34809613e-01 1.31494117e+00
4.69713397e-02 -1.37285304e+00 -1.67157277e-01 2.73507446e-01
7.83829451e-01 1.87885538e-01 1.49030000e-01 -2.30867505e-01
1.90246496e-02 -1.17081356e+00 -2.30717659e-01 1.05940926e+00
1.29091215e+00 -5.32247484e-01 1.01315701e+00 -3.08570951e-01
-8.91550958e-01 -2.37956131e-03 -4.63569403e-01 1.93346769e-01
8.82712081e-02 8.94224942e-01 -1.86280042e-01 7.05207169e-01
8.45585525e-01 1.33463216e+00 -8.23184013e-01 1.72021747e+00
-1.92979857e-01 2.94606000e-01 -1.02105523e-02 5.93964934e-01
-1.73218280e-01 -6.84987724e-01 5.09483516e-01 5.71715295e-01
5.15050590e-01 3.23151439e-01 -8.57814401e-02 8.71908069e-01
2.22026378e-01 8.19814503e-02 -4.44576919e-01 1.69386327e-01
3.46740872e-01 1.27757561e+00 4.49768975e-02 1.15669332e-01
-8.19401145e-01 5.27698576e-01 1.24302611e-01 6.29251778e-01
-5.19467115e-01 -5.88127851e-01 1.11905909e+00 3.83227915e-01
1.64620921e-01 1.67548016e-01 -1.88915178e-01 -1.44283831e+00
4.99447882e-01 -6.97691083e-01 3.58038425e-01 -1.16295385e+00
-1.55315661e+00 7.08558738e-01 -3.22083801e-01 -1.63327980e+00
3.65101069e-01 -7.13928580e-01 -7.39946127e-01 1.32592344e+00
-2.02649117e+00 -1.24075472e+00 -9.48530078e-01 7.40076780e-01
-2.12316230e-01 -3.52745622e-01 3.32921833e-01 6.44749105e-01
-8.82827580e-01 6.90143108e-01 1.42149746e-01 2.58719534e-01
1.08514547e+00 -1.01701331e+00 8.23050141e-02 8.07038665e-01
-5.50675154e-01 7.47175395e-01 1.22147456e-01 -5.67263842e-01
-9.95448232e-01 -1.29522586e+00 6.39511287e-01 -3.81959975e-01
4.96982217e-01 2.80472368e-01 -9.38946784e-01 6.08289957e-01
2.83606946e-02 7.07455635e-01 4.75534856e-01 -1.11335799e-01
-2.14924529e-01 -5.33804893e-01 -9.76267397e-01 5.79107583e-01
1.14389551e+00 -2.83952117e-01 -3.36875826e-01 4.86513317e-01
4.77660567e-01 -6.90231800e-01 -1.06128991e+00 5.51144660e-01
4.96913761e-01 -1.37516356e+00 7.84929395e-01 -4.12654400e-01
4.62157845e-01 -6.59788907e-01 1.05937123e-01 -1.22317040e+00
-2.31631726e-01 -7.15279281e-01 1.56059310e-01 1.13503122e+00
-1.61435194e-02 -1.13250601e+00 4.08175856e-01 2.10444361e-01
-3.72304469e-01 -1.00501418e+00 -5.93725443e-01 -4.21252459e-01
-1.51470810e-01 8.47241431e-02 6.35964870e-01 7.93088198e-01
-4.63327348e-01 7.36747459e-02 -2.36839756e-01 4.20210063e-01
5.19916594e-01 3.98299068e-01 6.89475238e-01 -1.15734911e+00
-2.19212413e-01 -5.19254088e-01 -1.02889240e+00 -9.98689234e-01
-2.96494335e-01 -6.73882425e-01 -4.39456463e-01 -1.77930510e+00
1.37783766e-01 -3.40097100e-01 -1.87290981e-01 3.91474009e-01
-3.79785568e-01 3.69830459e-01 -2.43637636e-01 2.40303040e-01
-2.24740863e-01 5.20723701e-01 1.95340729e+00 6.35530278e-02
-3.17017496e-01 1.25464320e-01 -1.13583875e+00 6.47521615e-01
8.07782352e-01 7.44248480e-02 -6.45144522e-01 -6.59767747e-01
1.13753878e-01 -2.17149290e-03 7.04706430e-01 -7.95615435e-01
3.06918323e-01 -1.98243469e-01 3.90304297e-01 -4.70838875e-01
1.97029516e-01 -1.59072965e-01 -7.81500190e-02 -2.68084377e-01
6.93583265e-02 -2.45715961e-01 4.99522500e-02 3.87407392e-01
-5.71451068e-01 3.50307763e-01 1.01643932e+00 5.46123646e-02
-6.48307264e-01 8.00339937e-01 3.74920577e-01 2.55711168e-01
9.68078077e-01 -3.06262285e-01 -1.15956151e+00 -7.22369477e-02
-4.80375141e-01 4.43158358e-01 5.36067486e-01 2.23655522e-01
7.44676352e-01 -1.25976622e+00 -1.02881217e+00 3.11806232e-01
4.25547987e-01 2.51898408e-01 1.02217913e+00 1.44144464e+00
-5.68859220e-01 4.30585504e-01 -4.48868811e-01 -5.26612520e-01
-1.25965405e+00 2.81149030e-01 8.35305929e-01 2.51727939e-01
-9.30905342e-01 7.40957022e-01 6.63597286e-01 2.89726611e-02
-7.13318884e-02 -5.71505785e-01 -6.35693789e-01 -1.94908261e-01
1.06016564e+00 3.89113754e-01 2.25396827e-01 -5.85220337e-01
-1.23324536e-01 1.09403694e+00 -3.83349985e-01 5.92841744e-01
1.08663070e+00 -3.71429682e-01 -3.33210796e-01 -1.49710730e-01
1.11910570e+00 1.69732973e-01 -1.23782647e+00 -4.40566748e-01
-7.73591459e-01 -1.07384551e+00 6.12831235e-01 -8.88502002e-01
-1.57903314e+00 1.02029610e+00 9.27330315e-01 -1.91187456e-01
1.48535097e+00 -1.63199723e-01 8.41024637e-01 1.77841820e-02
8.45265090e-02 -5.27720153e-01 -3.42033803e-01 2.55168905e-03
1.00252736e+00 -1.38582015e+00 1.14208594e-01 -6.35759890e-01
-6.70267463e-01 1.12487900e+00 8.69238853e-01 -2.63188146e-02
7.02116549e-01 -2.25019500e-01 4.90781546e-01 -3.03931028e-01
-3.95523131e-01 -6.19146585e-01 7.92839587e-01 8.40239346e-01
4.34017122e-01 -4.91647720e-02 -3.96394581e-01 6.10247374e-01
-1.31757185e-01 5.72250903e-01 5.64266205e-01 3.89922619e-01
-1.73841506e-01 -9.67041135e-01 6.18731119e-02 7.71059155e-01
-3.46854657e-01 -1.98073640e-01 -5.38630709e-02 6.77932680e-01
4.96929765e-01 9.82339442e-01 6.50525093e-02 -3.01863998e-01
4.90410537e-01 -6.91195250e-01 5.48827648e-01 -5.21663964e-01
-3.67227830e-02 1.03057861e-01 -3.50938663e-02 -7.28575647e-01
-4.90114897e-01 -5.39574802e-01 -8.64011228e-01 -2.65799582e-01
-1.86092138e-01 -4.76330638e-01 1.63343400e-01 6.78962409e-01
7.78015316e-01 2.97396094e-01 6.56545579e-01 -3.55255306e-01
-4.26826000e-01 -9.14197624e-01 -8.74837279e-01 1.70743451e-01
6.06339693e-01 -9.70528960e-01 -2.42468208e-01 -9.17136297e-02] | [15.784130096435547, -3.9609992504119873] |
a56b20d5-ea8b-4583-90a3-6ca18fc54b27 | mega-tts-zero-shot-text-to-speech-at-scale | 2306.03509 | null | https://arxiv.org/abs/2306.03509v1 | https://arxiv.org/pdf/2306.03509v1.pdf | Mega-TTS: Zero-Shot Text-to-Speech at Scale with Intrinsic Inductive Bias | Scaling text-to-speech to a large and wild dataset has been proven to be highly effective in achieving timbre and speech style generalization, particularly in zero-shot TTS. However, previous works usually encode speech into latent using audio codec and use autoregressive language models or diffusion models to generate it, which ignores the intrinsic nature of speech and may lead to inferior or uncontrollable results. We argue that speech can be decomposed into several attributes (e.g., content, timbre, prosody, and phase) and each of them should be modeled using a module with appropriate inductive biases. From this perspective, we carefully design a novel and large zero-shot TTS system called Mega-TTS, which is trained with large-scale wild data and models different attributes in different ways: 1) Instead of using latent encoded by audio codec as the intermediate feature, we still choose spectrogram as it separates the phase and other attributes very well. Phase can be appropriately constructed by the GAN-based vocoder and does not need to be modeled by the language model. 2) We model the timbre using global vectors since timbre is a global attribute that changes slowly over time. 3) We further use a VQGAN-based acoustic model to generate the spectrogram and a latent code language model to fit the distribution of prosody, since prosody changes quickly over time in a sentence, and language models can capture both local and long-range dependencies. We scale Mega-TTS to multi-domain datasets with 20K hours of speech and evaluate its performance on unseen speakers. Experimental results demonstrate that Mega-TTS surpasses state-of-the-art TTS systems on zero-shot TTS, speech editing, and cross-lingual TTS tasks, with superior naturalness, robustness, and speaker similarity due to the proper inductive bias of each module. Audio samples are available at https://mega-tts.github.io/demo-page. | ['Zhou Zhao', 'Zejun Ma', 'Xiang Yin', 'Chunfeng Wang', 'Rongjie Huang', 'Shengpeng Ji', 'Qian Yang', 'Chen Zhang', 'Jinglin Liu', 'Zhenhui Ye', 'Yi Ren', 'Ziyue Jiang'] | 2023-06-06 | null | null | null | null | ['style-generalization'] | ['computer-vision'] | [-1.27084926e-01 -8.99486151e-03 1.40648186e-02 -4.51206386e-01
-1.10042250e+00 -6.59945428e-01 3.80479604e-01 -4.17979866e-01
1.34718232e-02 3.94824415e-01 4.62085307e-01 -2.69497544e-01
4.71387893e-01 -6.24108493e-01 -7.34535754e-01 -9.00426209e-01
1.77621752e-01 2.76439130e-01 -2.83350539e-03 -2.95770884e-01
-5.16681552e-01 -5.65117076e-02 -1.46495688e+00 1.68378502e-01
8.61503661e-01 1.02141953e+00 1.72239318e-01 1.00967848e+00
-2.31288165e-01 7.43101120e-01 -1.00988817e+00 -1.98587105e-01
1.59447014e-01 -8.69787574e-01 -4.21074331e-01 -2.15655833e-01
1.72855347e-01 -1.81622475e-01 -2.76255280e-01 9.27943766e-01
8.91121149e-01 1.30110323e-01 6.13785088e-01 -9.74696934e-01
-6.84048176e-01 8.55875432e-01 -8.29622895e-02 -1.32584304e-01
2.80521452e-01 3.38841558e-01 9.24002171e-01 -6.77823544e-01
1.37406558e-01 1.38119745e+00 5.87308466e-01 8.50219846e-01
-1.46723187e+00 -1.08627200e+00 -2.59985812e-02 3.55832204e-02
-1.30412900e+00 -9.06254649e-01 1.07245100e+00 -3.87562543e-01
7.52263784e-01 4.91135687e-01 6.32110536e-01 1.85473657e+00
2.70493571e-02 7.62020469e-01 1.07868552e+00 -5.02124131e-01
5.04251063e-01 2.43379232e-02 -1.74987346e-01 1.51979417e-01
-8.59764934e-01 2.95055747e-01 -7.40910053e-01 -1.15737677e-01
3.19792449e-01 -3.90780985e-01 -3.69484752e-01 9.57323685e-02
-9.65239882e-01 7.99051583e-01 8.99461284e-02 2.68912405e-01
-2.39117488e-01 3.31153721e-01 4.67474729e-01 5.70196450e-01
6.66793227e-01 1.57056034e-01 -5.11830389e-01 -5.80966532e-01
-9.95276749e-01 -1.37952104e-01 8.17415714e-01 9.31863427e-01
5.25678754e-01 8.40248048e-01 -2.26023227e-01 1.32235897e+00
1.78987965e-01 9.27018225e-01 8.58618855e-01 -8.40824485e-01
3.91978651e-01 -2.87364453e-01 -2.77752280e-01 -4.78767604e-01
-4.27990146e-02 -4.49297279e-01 -9.71342206e-01 -5.33949099e-02
8.81989971e-02 -3.57555062e-01 -1.15583694e+00 2.05641770e+00
1.05306052e-01 2.87600249e-01 1.13143757e-01 6.46216929e-01
7.23234892e-01 1.07860327e+00 -7.92150870e-02 -3.77662510e-01
1.17443800e+00 -9.17854071e-01 -1.16287589e+00 -3.02813649e-01
3.53594720e-01 -8.27394307e-01 1.39408624e+00 4.25345927e-01
-9.61218238e-01 -1.01163781e+00 -1.03272533e+00 7.18902051e-02
-2.70423889e-01 -9.74928737e-02 5.91572337e-02 9.70900059e-01
-9.26550090e-01 3.79644603e-01 -7.55926788e-01 -9.98907536e-03
-1.94173515e-01 1.15062863e-01 1.35357216e-01 2.20309168e-01
-1.71602583e+00 5.13925374e-01 1.91935867e-01 -5.81530575e-03
-1.12185061e+00 -7.07439363e-01 -9.41497386e-01 2.69573797e-02
1.55294627e-01 -4.99029338e-01 1.40144598e+00 -1.07883430e+00
-2.25422287e+00 2.10978746e-01 -3.58266801e-01 -4.40747738e-01
2.44944930e-01 -9.97261852e-02 -9.63027358e-01 -3.08063142e-02
-1.47879332e-01 5.24196088e-01 1.38080931e+00 -9.82254744e-01
-2.69959897e-01 -1.49523206e-02 -3.52398932e-01 3.02563366e-02
-6.84903443e-01 -6.11607917e-02 -4.29423362e-01 -1.05882764e+00
-9.25779641e-02 -1.01861525e+00 2.04839274e-01 -6.18505955e-01
-3.90673429e-01 -8.16760212e-02 8.16085398e-01 -7.76837409e-01
1.36263120e+00 -2.52960300e+00 9.22848508e-02 -1.00487515e-01
-1.65680006e-01 2.97004282e-01 -1.73158735e-01 3.73355389e-01
-9.37381461e-02 1.95286721e-01 -2.42213428e-01 -6.49317205e-01
1.74296275e-01 2.62938261e-01 -7.08201230e-01 1.98740438e-01
-7.16634560e-03 8.08773756e-01 -8.84163558e-01 -2.54384637e-01
3.11215371e-01 8.44531775e-01 -4.25184816e-01 4.07822728e-01
-1.21668167e-01 6.84293747e-01 -1.72209650e-01 3.36858869e-01
4.91238564e-01 3.32760125e-01 3.18394527e-02 -1.88807145e-01
2.05013026e-02 7.00907528e-01 -9.36303914e-01 1.74030781e+00
-9.14695859e-01 7.06846595e-01 1.23404734e-01 -6.41207039e-01
1.18882871e+00 8.03742647e-01 3.36748153e-01 -7.30523407e-01
3.81991640e-02 2.79785901e-01 1.84284002e-02 -3.56670350e-01
1.67721391e-01 -4.38596636e-01 -1.13872804e-01 2.77445346e-01
4.28199708e-01 -5.90962768e-01 -3.50926489e-01 -1.02152407e-01
8.33325148e-01 -5.01026548e-02 -7.14133009e-02 3.88537012e-02
3.07256043e-01 -5.75644851e-01 7.12487876e-01 5.47372401e-01
1.16152465e-02 7.86821485e-01 2.24245593e-01 1.29431661e-03
-1.01540947e+00 -1.22838557e+00 -1.64045691e-01 1.26962447e+00
-3.06213200e-01 -5.38601458e-01 -9.32785094e-01 -2.65395582e-01
-3.06123793e-01 1.08852375e+00 -4.74607080e-01 -3.91329348e-01
-5.34315646e-01 -5.05121469e-01 8.95969748e-01 3.76720846e-01
2.05022648e-01 -1.04679012e+00 5.62670007e-02 5.33269763e-01
-3.93511385e-01 -1.05386496e+00 -8.66987288e-01 4.31566954e-01
-5.08171618e-01 -2.36879185e-01 -7.64998674e-01 -4.69653338e-01
2.30402627e-04 -1.17531955e-01 8.30298185e-01 -4.96651351e-01
1.92753762e-01 2.40448147e-01 -6.83922529e-01 -5.18802762e-01
-9.45252120e-01 2.08335340e-01 4.18661088e-01 4.71055031e-01
2.09368721e-01 -8.24908197e-01 -4.30954307e-01 3.39601755e-01
-7.93641031e-01 -5.41617014e-02 2.00858533e-01 8.79341483e-01
3.99809301e-01 2.36523356e-02 7.57389843e-01 -6.02392256e-01
6.93908632e-01 -3.94221365e-01 -2.85119921e-01 3.75413336e-02
-4.19242531e-01 -2.74311323e-02 1.04701638e+00 -8.29441130e-01
-1.16160750e+00 -1.25244036e-01 -5.92477441e-01 -6.72361672e-01
-1.63157254e-01 1.94361091e-01 -4.18443561e-01 4.00218934e-01
7.13113368e-01 3.64522099e-01 -1.16006725e-01 -5.27455628e-01
4.97026086e-01 1.28152657e+00 5.17822742e-01 -5.34648001e-01
9.17343497e-01 1.79678276e-01 -5.79629779e-01 -1.08921134e+00
-7.26534009e-01 -2.80284375e-01 -4.07314271e-01 -1.15394697e-01
7.38291323e-01 -1.10252273e+00 -5.11180758e-01 6.62571013e-01
-1.10646784e+00 -6.28946781e-01 -4.82294887e-01 7.27881968e-01
-5.89509487e-01 1.70531929e-01 -8.10817957e-01 -9.92252588e-01
-4.43175912e-01 -1.11427808e+00 9.84841645e-01 -1.94441751e-01
-3.50958079e-01 -9.05225515e-01 1.82067230e-01 2.26066798e-01
6.61094248e-01 -4.13280837e-02 6.91774189e-01 -4.50626940e-01
-8.23304281e-02 2.77888384e-02 6.17004335e-01 8.36639762e-01
3.57874453e-01 8.61896127e-02 -1.72139883e+00 -3.28911006e-01
4.62379187e-01 -3.12917888e-01 6.42217517e-01 5.15972197e-01
1.04003358e+00 -4.55628186e-01 1.28302559e-01 9.29809928e-01
8.12121093e-01 4.67906594e-01 6.68261468e-01 -1.86934948e-01
7.99182653e-01 5.47635138e-01 3.07451904e-01 2.53370434e-01
2.23870441e-01 8.81155372e-01 8.00853968e-02 -2.09659904e-01
-4.84476119e-01 -4.82012779e-01 9.70297933e-01 1.89236534e+00
2.10284635e-01 -4.05658454e-01 -6.43871248e-01 4.18874711e-01
-1.41989875e+00 -9.49887097e-01 6.83450326e-02 2.27286887e+00
1.22946870e+00 1.71104968e-01 1.13204159e-01 2.51710713e-01
6.19170845e-01 3.68690342e-01 -6.06376410e-01 -6.35935366e-01
-1.65101990e-01 3.54308486e-01 2.31803223e-01 6.24773502e-01
-7.36956537e-01 1.03714287e+00 5.79456234e+00 1.28862560e+00
-1.56074083e+00 4.60696161e-01 6.44313335e-01 -2.42856756e-01
-5.49787521e-01 -1.11223638e-01 -6.36969507e-01 7.37207532e-01
1.53285754e+00 -1.54702002e-02 7.85387814e-01 5.49189568e-01
3.54534864e-01 5.62470078e-01 -1.10781837e+00 8.85952652e-01
-3.25084701e-02 -7.39909470e-01 -1.61194295e-01 -6.50517642e-02
5.17160416e-01 1.15803182e-01 3.26449066e-01 6.50994480e-01
3.12730312e-01 -9.99947250e-01 1.19659925e+00 2.70553827e-01
1.30035520e+00 -5.57080090e-01 3.62581342e-01 3.49666566e-01
-1.32562912e+00 -1.25397965e-02 -1.45599291e-01 2.02742219e-01
2.52631992e-01 5.32469869e-01 -7.93345690e-01 3.45618486e-01
6.96382105e-01 4.16719824e-01 -7.23871663e-02 4.30134386e-01
-4.57911223e-01 1.36149764e+00 -2.83215672e-01 1.19023860e-01
3.15545239e-02 -4.94764075e-02 6.03139460e-01 1.33385634e+00
4.59383547e-01 -1.34291708e-01 -3.95963341e-02 8.85314941e-01
-6.59825578e-02 5.97942807e-02 -4.41367388e-01 -9.52344462e-02
8.41746211e-01 9.46706116e-01 -6.48804232e-02 -3.29035312e-01
-2.89612263e-01 9.73586977e-01 -1.39980093e-01 7.75340080e-01
-1.03563428e+00 -4.58584279e-01 6.91109002e-01 1.12373673e-03
3.86911482e-01 -2.38132924e-01 6.85559958e-02 -1.18691730e+00
-1.42853241e-04 -1.07845354e+00 -2.50508804e-02 -8.42992961e-01
-1.21532416e+00 1.05560446e+00 -2.72508651e-01 -1.42284143e+00
-7.35749543e-01 -2.11018652e-01 -5.30568302e-01 1.11664975e+00
-1.40341604e+00 -1.10793471e+00 -5.75537495e-02 6.60956085e-01
9.10232246e-01 -1.49600565e-01 1.10876274e+00 2.42274463e-01
-4.45161015e-01 7.57834852e-01 2.73625433e-01 1.53038248e-01
1.07065964e+00 -1.19713902e+00 8.41851711e-01 7.02330530e-01
3.01618516e-01 4.32972968e-01 7.88700461e-01 -3.63418698e-01
-1.26827097e+00 -1.14826059e+00 6.94310606e-01 -4.01605934e-01
7.64208972e-01 -1.01267111e+00 -1.08000171e+00 4.56127912e-01
3.93524826e-01 4.50460650e-02 7.22659945e-01 1.53555095e-01
-4.84028667e-01 -5.01262903e-01 -6.57989383e-01 4.87981617e-01
7.24789441e-01 -1.14790726e+00 -5.40328324e-01 8.01129192e-02
1.21326447e+00 -3.65249574e-01 -8.17491114e-01 1.33755758e-01
4.69718277e-01 -8.51686478e-01 8.11048806e-01 -1.00514732e-01
5.94342453e-03 -2.74809092e-01 -3.79788667e-01 -1.84465778e+00
-2.13865966e-01 -1.18509233e+00 -1.39939904e-01 1.82807434e+00
4.76498872e-01 -7.32379675e-01 1.54604167e-01 3.95154744e-01
-3.32908928e-01 -3.83950621e-01 -9.76491988e-01 -1.14209425e+00
1.32821679e-01 -9.62450206e-01 9.92155194e-01 9.71104681e-01
-5.88979907e-02 5.91934800e-01 -7.77982056e-01 1.30182117e-01
4.13634986e-01 -7.75643215e-02 8.00566196e-01 -8.80275071e-01
-5.77733994e-01 -2.38316879e-01 3.94576639e-02 -1.21049869e+00
2.54519850e-01 -6.79159582e-01 3.96373272e-01 -9.27643657e-01
-4.85414088e-01 -4.35724884e-01 -3.48322600e-01 4.23090816e-01
-4.45928648e-02 2.05003154e-02 2.77791768e-01 1.40675023e-01
2.47090589e-02 1.04519880e+00 1.02065635e+00 -3.48184377e-01
-3.83186460e-01 9.53300819e-02 -3.45741302e-01 5.38895369e-01
8.32503021e-01 -6.43609405e-01 -5.26344657e-01 -4.23994094e-01
-1.84682906e-01 3.85283649e-01 1.13460243e-01 -1.03352678e+00
2.01636866e-01 -4.52552177e-02 -9.82654188e-03 -1.11798488e-01
8.66616905e-01 -6.15749121e-01 3.01613986e-01 1.67194888e-01
-3.55805933e-01 -4.16068375e-01 1.11402564e-01 3.82575095e-01
-4.83035386e-01 3.54774408e-02 6.59771979e-01 1.05884746e-01
-2.34356433e-01 3.58308434e-01 -4.45044458e-01 1.57935962e-01
4.51611429e-01 -6.66051731e-02 -1.50386989e-01 -8.00330281e-01
-5.88098347e-01 -1.74342588e-01 1.46087587e-01 6.44558489e-01
3.98198366e-01 -1.47234368e+00 -7.80288696e-01 4.91906375e-01
8.88481811e-02 -1.99226737e-01 3.58605027e-01 5.30621588e-01
8.19722712e-02 2.26106137e-01 3.05838823e-01 -7.10734487e-01
-9.02457535e-01 5.05173981e-01 3.47022116e-01 2.41703585e-01
-5.32003880e-01 8.23996544e-01 4.90068674e-01 -4.67051297e-01
3.40312093e-01 -4.10870194e-01 8.12047645e-02 1.33884877e-01
3.93518388e-01 1.15187634e-02 5.25234267e-02 -7.64004171e-01
-2.19708860e-01 7.28385389e-01 3.31680238e-01 -4.71113831e-01
1.16753960e+00 -3.50563496e-01 2.58138329e-01 1.08636987e+00
1.39639604e+00 4.73209560e-01 -1.41391230e+00 -2.06467122e-01
-5.12840390e-01 -1.62854612e-01 1.98337480e-01 -6.03238761e-01
-1.05669391e+00 1.28141654e+00 5.84507704e-01 5.60012400e-01
1.19758141e+00 -1.04711995e-01 1.11914659e+00 -1.31007448e-01
2.50807673e-01 -1.26652086e+00 1.13516554e-01 6.73308790e-01
9.28168893e-01 -9.31382418e-01 -7.15077162e-01 -3.52752917e-02
-8.69618237e-01 1.00332451e+00 2.95820594e-01 3.26206803e-01
6.83189511e-01 5.98384678e-01 5.65330923e-01 2.75867254e-01
-8.96610379e-01 2.35919412e-02 2.90665776e-01 5.23145556e-01
4.09229398e-01 3.73091310e-01 4.27165210e-01 6.02179050e-01
-7.94692278e-01 -3.35141033e-01 2.41588995e-01 2.91881055e-01
-1.75382823e-01 -1.20369732e+00 -5.34923971e-01 -6.25793030e-03
-3.78560454e-01 -3.51678044e-01 -1.06699780e-01 2.59884298e-01
1.26534645e-02 1.28278553e+00 -2.73248944e-02 -8.02609921e-01
3.63598585e-01 3.74734968e-01 1.37466878e-01 -4.80370224e-01
-6.45884752e-01 6.69724405e-01 5.44821769e-02 -3.60261798e-01
-9.56389233e-02 -4.44913775e-01 -1.13516271e+00 -2.36551031e-01
-3.62640738e-01 2.84389168e-01 8.62052202e-01 7.31346548e-01
2.73648411e-01 8.94930482e-01 1.10129416e+00 -6.72350585e-01
-6.53286397e-01 -1.31672299e+00 -6.76789045e-01 3.97396952e-01
7.42441833e-01 -2.06692979e-01 -7.47466028e-01 3.09373975e-01] | [14.98084545135498, 6.462255001068115] |
5f062c20-bec9-4749-9ec6-12a5f8c0d81c | towards-precise-weakly-supervised-object | 2304.14114 | null | https://arxiv.org/abs/2304.14114v2 | https://arxiv.org/pdf/2304.14114v2.pdf | Towards Precise Weakly Supervised Object Detection via Interactive Contrastive Learning of Context Information | Weakly supervised object detection (WSOD) aims at learning precise object detectors with only image-level tags. In spite of intensive research on deep learning (DL) approaches over the past few years, there is still a significant performance gap between WSOD and fully supervised object detection. In fact, most existing WSOD methods only consider the visual appearance of each region proposal but ignore employing the useful context information in the image. To this end, this paper proposes an interactive end-to-end WSDO framework called JLWSOD with two innovations: i) two types of WSOD-specific context information (i.e., instance-wise correlation andsemantic-wise correlation) are proposed and introduced into WSOD framework; ii) an interactive graph contrastive learning (iGCL) mechanism is designed to jointly optimize the visual appearance and context information for better WSOD performance. Specifically, the iGCL mechanism takes full advantage of the complementary interpretations of the WSOD, namely instance-wise detection and semantic-wise prediction tasks, forming a more comprehensive solution. Extensive experiments on the widely used PASCAL VOC and MS COCO benchmarks verify the superiority of JLWSOD over alternative state-of-the-art approaches and baseline models (improvement of 3.6%~23.3% on mAP and 3.4%~19.7% on CorLoc, respectively). | ['ChiMan Vong', 'Qi Lai'] | 2023-04-27 | null | null | null | null | ['weakly-supervised-object-detection'] | ['computer-vision'] | [-1.67992674e-02 6.28790259e-03 -3.88336003e-01 -3.49038601e-01
-5.46242177e-01 -2.27286965e-01 7.36341953e-01 2.45389596e-01
-4.87247944e-01 3.49986583e-01 -6.98379725e-02 -4.16074879e-02
9.31960531e-03 -5.69327712e-01 -5.99622548e-01 -8.18969548e-01
-2.77368294e-04 5.05898744e-02 8.21464181e-01 -1.34862721e-01
9.40825790e-03 5.55284798e-01 -1.53801346e+00 2.40358040e-01
7.81800449e-01 1.40280199e+00 6.11120999e-01 4.63418752e-01
-2.48433292e-01 8.72884214e-01 -4.21701431e-01 -2.58263111e-01
2.73241699e-01 -1.98734030e-01 -4.78130251e-01 3.09437156e-01
7.55994439e-01 -1.29915476e-01 -3.15992504e-01 1.25284898e+00
5.79977334e-01 3.37622724e-02 3.92534196e-01 -1.34733510e+00
-7.21660912e-01 4.09494042e-01 -8.92301559e-01 4.40474540e-01
-1.47208586e-01 3.14153939e-01 1.28742576e+00 -1.17323220e+00
5.95037341e-01 1.31021345e+00 3.73851925e-01 2.90244609e-01
-1.16290212e+00 -4.79590535e-01 6.93744004e-01 3.47099423e-01
-1.55080032e+00 -1.01302110e-01 8.54526877e-01 -4.26757395e-01
9.12191808e-01 2.50291616e-01 4.87879485e-01 7.19059646e-01
8.36013705e-02 1.25242472e+00 1.29251862e+00 -4.28115338e-01
1.63014531e-02 3.33727330e-01 3.15123230e-01 8.97507489e-01
5.14453948e-01 1.38103455e-01 -5.00921428e-01 1.65916786e-01
4.17195916e-01 1.41068026e-01 -2.81898119e-02 -7.65036643e-01
-1.17136300e+00 5.02483070e-01 9.34614897e-01 1.58135310e-01
-2.01752365e-01 8.32678452e-02 4.35229599e-01 -7.08799586e-02
5.73116839e-01 9.88831520e-02 -2.93197691e-01 5.06164074e-01
-7.45394647e-01 1.92755967e-01 4.38286275e-01 1.13099742e+00
7.83577025e-01 8.41862112e-02 -4.31637675e-01 9.39292431e-01
4.06369656e-01 5.39165020e-01 1.44389361e-01 -2.43995726e-01
5.17734885e-01 7.91286170e-01 7.26744160e-02 -1.14281309e+00
-5.11946619e-01 -8.79172623e-01 -4.88989294e-01 1.84968114e-01
1.99226648e-01 7.27275461e-02 -1.04787397e+00 1.56933248e+00
5.00851333e-01 2.07383081e-01 -2.05409989e-01 1.22590709e+00
1.15388060e+00 5.63652635e-01 4.41261202e-01 5.79388924e-02
1.33675444e+00 -1.36768413e+00 -6.07275903e-01 -5.62989354e-01
5.04707277e-01 -7.21902728e-01 1.11451602e+00 1.78460836e-01
-9.22914565e-01 -7.85655975e-01 -1.23078442e+00 -1.16831757e-01
-5.25363207e-01 5.96456170e-01 6.22584581e-01 2.38805622e-01
-8.63257706e-01 1.58408042e-02 -7.26367831e-01 -4.87369686e-01
5.51384568e-01 2.59077221e-01 -1.62755236e-01 -1.90000027e-01
-7.89960742e-01 6.84920907e-01 6.56601965e-01 2.48062298e-01
-1.07204223e+00 -5.31832159e-01 -6.21826231e-01 -1.23136587e-01
8.96991789e-01 -5.00361025e-01 1.13117409e+00 -9.52969253e-01
-9.23669636e-01 1.07057714e+00 -1.42425217e-03 -3.32073808e-01
6.66930377e-01 -4.35243875e-01 -4.89217520e-01 1.14050210e-01
1.16030633e-01 6.71496630e-01 6.25165164e-01 -1.48395419e+00
-1.05911171e+00 -5.01896799e-01 2.50972062e-01 3.37788969e-01
-3.15340668e-01 7.39114592e-03 -1.08148944e+00 -7.29630530e-01
1.87010512e-01 -9.02453959e-01 -1.92088500e-01 4.85122681e-01
-7.54955769e-01 -5.74672580e-01 1.00550520e+00 -4.08366740e-01
1.16872859e+00 -2.21596885e+00 4.38366979e-02 -1.26404047e-01
4.51760381e-01 5.29956758e-01 -3.11146319e-01 1.97889999e-01
1.39068529e-01 -3.70308846e-01 -6.87569827e-02 -4.93962854e-01
1.22819848e-01 7.69358724e-02 7.00611100e-02 4.77370232e-01
3.67837578e-01 9.93882537e-01 -9.62099493e-01 -6.49593234e-01
3.16204578e-01 2.55172104e-01 -2.89705306e-01 3.12417001e-01
-4.50159937e-01 -5.73245750e-04 -4.76497829e-01 9.92127478e-01
8.02164257e-01 -3.01323771e-01 9.16852336e-03 -4.32534248e-01
-2.33651668e-01 2.58445796e-02 -1.29465616e+00 1.61012614e+00
-2.02309430e-01 4.52321708e-01 3.30099501e-02 -1.01682246e+00
9.18096006e-01 -2.16194287e-01 2.99265265e-01 -8.40675116e-01
1.42570749e-01 1.80318117e-01 -8.85996744e-02 -5.32221973e-01
4.31900144e-01 8.61799940e-02 1.04524858e-01 -2.08751690e-02
2.40111589e-01 3.49070340e-01 2.10545719e-01 3.16133887e-01
6.88434899e-01 1.75732702e-01 4.84380662e-01 -5.50430894e-01
6.52520061e-01 4.93271053e-02 7.46331573e-01 8.08077693e-01
-4.14511591e-01 6.49952590e-01 4.38192368e-01 -2.93283254e-01
-7.08612144e-01 -1.04518425e+00 -8.26568231e-02 1.42098618e+00
7.48952329e-01 -2.89105177e-01 -5.94646275e-01 -1.02191556e+00
1.84027612e-01 5.82359374e-01 -8.01645458e-01 -5.11221364e-02
-4.14952606e-01 -8.49248350e-01 2.90077418e-01 8.33583653e-01
8.18746269e-01 -8.91315460e-01 -1.90051854e-01 2.11840086e-02
1.74894080e-01 -1.43978870e+00 -6.70037210e-01 1.69653088e-01
-6.28895521e-01 -1.01387227e+00 -5.55866063e-01 -1.01639068e+00
6.82370484e-01 9.02885079e-01 1.05078769e+00 1.96635455e-01
-5.80203176e-01 2.69851387e-01 -3.07209373e-01 -6.37893438e-01
7.69899115e-02 -3.16602476e-02 -1.55895844e-01 2.67501622e-01
5.08859873e-01 1.10812075e-02 -8.60341132e-01 4.17895108e-01
-7.25390136e-01 1.92911357e-01 9.70436454e-01 8.05891514e-01
8.51074338e-01 -2.07651675e-01 3.37358356e-01 -1.07929003e+00
1.74659267e-01 -4.36071634e-01 -8.45078588e-01 4.55130309e-01
-9.38616395e-01 -7.87673667e-02 3.23507458e-01 -2.38557339e-01
-1.20456433e+00 1.69315606e-01 -7.94917624e-03 -3.64198476e-01
-1.01347409e-01 2.19443232e-01 -4.17763144e-01 -8.91961604e-02
5.87287784e-01 2.77264714e-01 -3.05815101e-01 -5.19432247e-01
3.86354864e-01 5.50732970e-01 5.68543017e-01 -3.59570950e-01
7.94012964e-01 6.36090696e-01 -1.96391493e-01 -6.19819283e-01
-1.18508542e+00 -8.92432392e-01 -6.51970208e-01 -2.05617249e-01
1.00709343e+00 -1.05290782e+00 -4.11282599e-01 5.35055041e-01
-9.01178896e-01 -2.45675728e-01 -1.51935056e-01 4.63031024e-01
-1.48239762e-01 2.31943801e-01 -4.12351966e-01 -7.60808051e-01
-3.56898695e-01 -1.09775341e+00 1.22377884e+00 2.73330718e-01
4.77676928e-01 -8.86637330e-01 -3.37016493e-01 3.19024563e-01
1.07239656e-01 2.15442196e-01 5.99992573e-01 -4.81888354e-01
-8.30982924e-01 -3.28074962e-01 -9.17754352e-01 4.66991276e-01
-1.22017980e-01 -9.79704857e-02 -1.18382740e+00 -2.87753046e-01
-4.76054341e-01 -2.64115185e-01 1.06475127e+00 3.16391647e-01
1.09507942e+00 -1.28480941e-02 -5.55979729e-01 4.70598400e-01
1.72997260e+00 7.07077459e-02 2.66664475e-01 2.56224066e-01
1.03576350e+00 5.23345947e-01 1.01629305e+00 3.01469296e-01
5.95956087e-01 9.81635690e-01 6.90323114e-01 -4.00263727e-01
-6.66997790e-01 -4.28523272e-01 3.42903197e-01 5.12104452e-01
4.14544577e-03 -3.39940429e-01 -7.97056675e-01 5.53225636e-01
-2.01721692e+00 -5.21628976e-01 -3.22983176e-01 1.97240400e+00
3.69100660e-01 4.73512888e-01 2.64048785e-01 -1.17787175e-01
8.60345900e-01 4.84971642e-01 -6.61059916e-01 1.20380800e-03
-2.24682719e-01 -3.21785688e-01 5.99060237e-01 2.28967205e-01
-1.36134338e+00 9.87357140e-01 4.79152441e+00 9.79910016e-01
-1.17938817e+00 2.48579115e-01 5.18608928e-01 -2.93933954e-02
8.90698433e-02 -1.20468333e-01 -1.17305183e+00 3.54963839e-01
7.91713074e-02 1.23495884e-01 3.29570100e-02 1.21396911e+00
9.58834067e-02 -2.36199409e-01 -9.04594481e-01 1.03528845e+00
2.42710739e-01 -1.25416625e+00 2.77039483e-02 -1.11420050e-01
7.66540289e-01 2.77944505e-01 2.63020638e-02 3.49140465e-01
1.53487340e-01 -5.13943374e-01 1.11987257e+00 2.98695087e-01
5.40610015e-01 -5.44860125e-01 7.99830496e-01 4.02286947e-01
-1.57189918e+00 -2.54110962e-01 -4.47340012e-01 2.34247610e-01
-1.58852994e-01 7.07867086e-01 -8.29599738e-01 5.69901526e-01
8.85744393e-01 8.63390148e-01 -8.97535622e-01 1.17935681e+00
-4.86601561e-01 6.22820735e-01 -8.97591263e-02 -2.40894288e-01
6.01692855e-01 1.13095194e-01 5.58704138e-01 1.58600533e+00
-1.05843760e-01 5.63304946e-02 4.86413687e-01 8.43964934e-01
-7.15843588e-02 7.40268752e-02 -2.78602451e-01 2.33528748e-01
3.01574320e-01 1.49382246e+00 -8.32913518e-01 -2.91478902e-01
-7.27246642e-01 8.77152920e-01 4.35146600e-01 4.33450371e-01
-7.43409932e-01 -2.32313246e-01 4.17268544e-01 1.76377267e-01
5.78131676e-01 -2.74001956e-01 -3.08507472e-01 -1.03530335e+00
1.87234476e-01 -4.37343657e-01 6.40436471e-01 -6.42812192e-01
-1.30240750e+00 5.22729814e-01 -7.03687295e-02 -1.27077723e+00
4.97757941e-01 -7.35266805e-01 -6.17908120e-01 6.22468114e-01
-2.04700732e+00 -1.46972179e+00 -7.36025393e-01 3.57238799e-01
7.71990120e-01 -2.76957583e-02 3.38716626e-01 3.67978513e-01
-8.52068722e-01 6.67760313e-01 6.83116838e-02 2.16492683e-01
4.92120862e-01 -1.49784756e+00 3.06711555e-01 9.43207860e-01
3.03964257e-01 3.45563799e-01 4.74522978e-01 -4.79586214e-01
-1.42070699e+00 -1.39487123e+00 8.04480612e-01 -1.82427943e-01
5.74088037e-01 -6.71349943e-01 -8.58551800e-01 3.80347550e-01
7.56787211e-02 7.15424657e-01 1.87484980e-01 1.04192376e-01
-4.67985183e-01 -5.27033091e-01 -9.35539186e-01 6.29960001e-01
1.33651984e+00 -4.17300493e-01 -3.98047566e-01 3.95915180e-01
7.41276145e-01 -3.62748504e-01 -4.74268198e-01 5.51489413e-01
3.66794616e-01 -8.84398937e-01 9.93726313e-01 -4.35842454e-01
4.59310412e-02 -7.70201564e-01 -2.45134369e-01 -8.83287549e-01
-4.28915530e-01 -2.34586596e-01 -2.62999505e-01 1.25512898e+00
3.07883680e-01 -3.75991225e-01 6.54967546e-01 2.28640437e-01
-3.81627232e-01 -1.00164640e+00 -5.92653990e-01 -9.60587740e-01
-3.49528372e-01 -4.59863484e-01 3.26928407e-01 7.83911765e-01
-3.84657592e-01 3.39420170e-01 -2.95617402e-01 4.35180783e-01
7.42467761e-01 3.77353013e-01 7.99476326e-01 -1.05051005e+00
-3.32142055e-01 -4.58228022e-01 -6.51876926e-01 -1.31265366e+00
-2.70226508e-01 -9.32025850e-01 2.91817844e-01 -1.54511023e+00
5.36278069e-01 -6.37808144e-01 -8.41456413e-01 5.03098011e-01
-6.11938894e-01 4.30300623e-01 4.20058489e-01 1.84900060e-01
-1.17101347e+00 5.49219370e-01 1.32456744e+00 -2.15719298e-01
7.08316714e-02 -1.78049684e-01 -5.98340750e-01 7.51536548e-01
5.57879865e-01 -4.85101372e-01 -4.71734673e-01 -5.05701780e-01
-2.23175079e-01 -2.74880946e-01 7.40524530e-01 -8.95261943e-01
2.60763109e-01 -1.27066895e-01 2.54075319e-01 -8.12253416e-01
1.20638289e-01 -6.74266279e-01 -4.23831731e-01 5.40842414e-01
-2.66529679e-01 -2.83109009e-01 1.92147925e-01 9.67204154e-01
-2.50319541e-01 5.18389307e-02 9.23058808e-01 1.79733690e-02
-1.47770476e+00 4.17172998e-01 1.62288040e-01 5.50518334e-02
1.28974438e+00 -8.32528472e-02 -3.30962360e-01 1.34658054e-01
-6.13840580e-01 3.47290486e-01 1.06973857e-01 6.54218256e-01
5.61364233e-01 -1.26950634e+00 -7.45169103e-01 1.72063801e-02
7.77044833e-01 5.15556186e-02 2.92660177e-01 9.88345921e-01
-2.88306326e-01 4.42479551e-01 1.43989623e-01 -9.05688226e-01
-1.27853084e+00 6.12517774e-01 3.24111938e-01 -1.87124059e-01
-6.48858368e-01 1.19619071e+00 6.20683968e-01 -2.92982638e-01
5.35296142e-01 -2.91800082e-01 -1.30242869e-01 1.05953120e-01
4.77791667e-01 3.05376053e-01 1.49035588e-01 -5.90082049e-01
-5.24763823e-01 5.26815414e-01 -3.54146570e-01 4.64682043e-01
1.23739100e+00 -1.92602158e-01 2.20977157e-01 3.35861653e-01
1.07326853e+00 -1.85146794e-01 -1.65341234e+00 -7.75937736e-01
2.37996548e-01 -5.02648175e-01 3.02198827e-01 -1.04897106e+00
-1.34218037e+00 9.21243250e-01 8.03810596e-01 7.65929073e-02
1.14457476e+00 2.22005442e-01 6.71565294e-01 2.36982301e-01
3.74819338e-01 -1.15029287e+00 2.47478738e-01 5.16651087e-02
8.63718688e-01 -1.48175204e+00 2.28537038e-01 -6.78729057e-01
-8.10920119e-01 7.97441006e-01 9.89551246e-01 -1.17192440e-01
5.49003541e-01 1.72699898e-01 -5.92614524e-02 -2.56009281e-01
-6.43738329e-01 -6.46609247e-01 7.93918073e-01 3.83482099e-01
2.45520011e-01 1.07049927e-01 -4.39234912e-01 5.50987363e-01
5.69870651e-01 -1.34815782e-01 -1.11135088e-01 8.84599745e-01
-6.75014138e-01 -7.86038399e-01 -1.97564825e-01 3.86506945e-01
-1.21645972e-01 6.04784023e-03 -3.94736141e-01 9.37967837e-01
5.29128253e-01 9.25422728e-01 -2.18176663e-01 -3.92038584e-01
5.84051490e-01 -4.11564022e-01 2.22169191e-01 -7.47213662e-01
-4.21908647e-01 2.26940066e-01 4.96158600e-02 -6.37199998e-01
-4.72907305e-01 -6.30784333e-01 -1.35886312e+00 8.48657340e-02
-6.75504744e-01 -1.55796483e-01 6.55922413e-01 7.93165207e-01
2.02538610e-01 6.26217842e-01 5.47179401e-01 -6.93881691e-01
-4.33083564e-01 -7.93399096e-01 -6.93613529e-01 4.31633085e-01
3.01186174e-01 -8.89026582e-01 -9.19346809e-02 -1.09508030e-01] | [9.33964729309082, 1.1931759119033813] |
721d4cb9-94f3-4d9e-b7af-7b59054fa859 | towards-an-lstm-based-predictive-framework | 1907.09395 | null | https://arxiv.org/abs/1907.09395v2 | https://arxiv.org/pdf/1907.09395v2.pdf | Mining Temporal Evolution of Knowledge Graph and Genealogical Features for Literature-based Discovery Prediction | Literature-based knowledge discovery process identifies the important but implicit relations among information embedded in published literature. Existing techniques from Information Retrieval and Natural Language Processing attempt to identify the hidden or unpublished connections between information concepts within published literature, however, these techniques undermine the concept of predicting the future and emerging relations among scientific knowledge components encapsulated within the literature. Keyword Co-occurrence Network (KCN), built upon author selected keywords (i.e., knowledge entities), is considered as a knowledge graph that focuses both on these knowledge components and knowledge structure of a scientific domain by examining the relationships between knowledge entities. Using data from two multidisciplinary research domains other than the medical domain, capitalizing on bibliometrics, the dynamicity of temporal KCNs, and a Long Short Term Memory recurrent neural network, this study proposed a framework to successfully predict the future literature-based discoveries - the emerging connections among knowledge units. Framing the problem as a dynamic supervised link prediction task, the proposed framework integrates some novel node and edge-level features. Temporal importance of keywords computed from both bipartite and unipartite networks, communities of keywords, built upon genealogical relations, and relative importance of temporal citation counts used in the feature construction process. Both node and edge-level features were input into an LSTM network to forecast the feature values for positive and negatively labeled non-connected keyword pairs and classify them accurately. High classification performance rates suggest that these features are supportive both in predicting the emerging connections between scientific knowledge units and emerging trend analysis. | ['Fahim Faisal', 'Nazim Choudhury', 'Matloob Khushi'] | 2019-07-22 | null | null | null | null | ['implicit-relations'] | ['natural-language-processing'] | [-3.95326048e-01 -1.38612073e-02 -7.42260873e-01 2.12737963e-01
2.88630128e-01 -5.45197070e-01 6.98329926e-01 6.32709026e-01
-3.17005247e-01 9.99286652e-01 2.77335018e-01 -6.11771643e-01
-1.08657277e+00 -1.15059686e+00 -5.06904304e-01 -3.88062775e-01
-6.98172867e-01 1.69494241e-01 7.32406648e-03 1.54464245e-01
7.74999201e-01 5.90457618e-01 -1.17920756e+00 2.23383531e-01
7.49309361e-01 8.05476665e-01 2.07309753e-01 2.73850441e-01
-8.31410229e-01 1.02292538e+00 -3.77448708e-01 -3.00360799e-01
-3.17942590e-01 -9.99491587e-02 -1.08982563e+00 -6.12147689e-01
-4.94298428e-01 6.24386668e-01 -7.02604592e-01 7.90501893e-01
1.68351904e-02 4.69491519e-02 5.97921550e-01 -1.19717586e+00
-1.24986613e+00 1.07528925e+00 -4.85016376e-01 9.19356108e-01
3.34648132e-01 -3.04007828e-01 1.27111256e+00 -8.80000949e-01
1.16815710e+00 9.02518213e-01 9.21453357e-01 -3.35066468e-01
-7.81884789e-01 -8.19474697e-01 2.32161745e-01 7.64876306e-01
-1.50127602e+00 1.20492183e-01 9.52019155e-01 -8.89512718e-01
1.17142832e+00 -6.61770925e-02 9.15738046e-01 1.10694015e+00
7.33322918e-01 6.13129772e-02 9.03895855e-01 -3.93454850e-01
-1.67537525e-01 3.11758965e-01 6.46711588e-01 5.83131850e-01
4.68105972e-01 -2.25188117e-02 -9.95912850e-01 -3.83985847e-01
5.35502493e-01 1.66284055e-01 -1.79448649e-01 3.61945271e-01
-1.43469882e+00 6.14879131e-01 6.98223889e-01 1.22723401e+00
-8.49171698e-01 -2.44289175e-01 4.42152411e-01 5.63390911e-01
6.64638758e-01 5.69634736e-01 -6.90155923e-01 -1.34753957e-02
-9.90743339e-01 -3.39541823e-01 1.05966270e+00 6.17882967e-01
8.88521910e-01 -3.35793719e-02 -9.02503282e-02 4.51223582e-01
2.72976220e-01 -5.52282706e-02 8.38440835e-01 -2.14909106e-01
1.12714641e-01 1.23433280e+00 -4.05069858e-01 -1.86238074e+00
-7.00519323e-01 -8.01692486e-01 -7.77641475e-01 -7.54516721e-01
-2.20554754e-01 -1.86223224e-01 -6.21026397e-01 1.47566044e+00
-5.97781762e-02 5.21049500e-01 2.47554835e-02 3.64015579e-01
1.30229151e+00 7.80920327e-01 3.12384218e-01 -7.20608532e-01
1.21414804e+00 -4.53142554e-01 -1.16931844e+00 5.35975993e-01
5.15393734e-01 -4.91243958e-01 -6.11184770e-03 -1.72431067e-01
-8.56265604e-01 -2.57522285e-01 -7.09695518e-01 2.42512882e-01
-1.31599081e+00 -9.45819393e-02 9.20580626e-01 -2.80362844e-01
-1.44169009e+00 8.19357634e-01 -2.26100877e-01 -5.65603495e-01
3.13240111e-01 3.33471984e-01 -2.81860739e-01 3.87259573e-01
-1.82692313e+00 1.11197412e+00 9.52588320e-01 3.75444114e-01
-2.02439114e-01 -9.19838727e-01 -4.06015277e-01 3.48221391e-01
3.01565379e-01 -8.23371947e-01 3.04480612e-01 -8.22419941e-01
-9.99276221e-01 5.98708868e-01 -3.28178704e-01 -4.14118946e-01
-2.06995636e-01 4.12212640e-01 -1.17272282e+00 8.16138908e-02
3.65851283e-01 8.47230181e-02 2.77892500e-01 -1.08491814e+00
-7.07878530e-01 -3.52495849e-01 -4.40449059e-01 1.26620904e-01
-9.56601501e-01 -1.80313483e-01 -2.99602598e-01 -6.58585727e-01
1.42506272e-01 -4.73904550e-01 1.03140518e-01 -6.94375813e-01
-5.53481758e-01 -8.45100999e-01 8.62107098e-01 -7.66282737e-01
1.69096017e+00 -1.68443882e+00 2.24009916e-01 7.72347808e-01
5.49399555e-01 -5.11311591e-02 4.83285785e-02 6.50192022e-01
-2.68061697e-01 4.50312763e-01 2.99634516e-01 5.37511408e-01
-6.45969212e-01 1.51849568e-01 -2.88282603e-01 2.25696266e-01
1.71411783e-02 1.34146082e+00 -1.09773564e+00 -5.10849297e-01
-4.37709749e-01 2.89368838e-01 3.18989396e-01 -1.82589412e-01
-9.62745771e-02 1.21507592e-01 -8.79145801e-01 7.30939209e-01
9.17806774e-02 -5.66080868e-01 4.38318759e-01 -1.91187575e-01
-4.35375512e-01 2.23117068e-01 -5.92443943e-01 1.32615817e+00
-3.07622671e-01 9.20114040e-01 -3.52953136e-01 -1.49096203e+00
1.13690531e+00 6.60626352e-01 8.65033388e-01 -6.17814124e-01
7.94014335e-02 4.11805004e-01 2.69721568e-01 -7.10191250e-01
3.61181080e-01 1.95647046e-01 3.36703897e-01 2.10063994e-01
4.15289849e-01 8.43073010e-01 2.89017200e-01 5.85242450e-01
1.11636055e+00 -2.04945758e-01 2.53655791e-01 -5.00112653e-01
6.39377177e-01 3.93815525e-02 4.80658293e-01 5.79368830e-01
2.31011286e-01 -4.89096880e-01 6.71443880e-01 -4.05455142e-01
-6.55324757e-01 -6.98278904e-01 -3.67703378e-01 7.81713486e-01
-1.27971441e-01 -1.98114991e-01 1.56643763e-01 -2.25117981e-01
4.79938716e-01 4.97890919e-01 -9.52685535e-01 -2.54387081e-01
-1.76467955e-01 -5.44854164e-01 6.13525927e-01 1.96632877e-01
1.79068536e-01 -1.24875164e+00 1.33744851e-01 3.62808645e-01
1.12665743e-02 -8.96378696e-01 3.51813599e-03 3.19268823e-01
-1.13435578e+00 -1.35570848e+00 -5.24360299e-01 -1.11521435e+00
5.86736262e-01 5.07872105e-02 1.09063220e+00 9.16961655e-02
-3.05533171e-01 6.07990503e-01 -2.80328989e-01 -3.11183184e-01
-4.74194549e-02 3.68030608e-01 2.93808550e-01 -8.31898153e-02
7.75147200e-01 -1.05448127e+00 -3.16867799e-01 -8.42751339e-02
-5.13144553e-01 -3.06662291e-01 7.27693856e-01 8.79284501e-01
1.68137431e-01 4.71512079e-01 1.30010664e+00 -5.67133963e-01
1.01465106e+00 -1.32942629e+00 -2.60766834e-01 4.32714254e-01
-1.39441323e+00 3.98598611e-02 3.48236322e-01 -4.28898871e-01
-9.81480777e-01 -7.64732182e-01 7.00532019e-01 -4.13545847e-01
2.15845302e-01 1.62963963e+00 3.81200999e-01 -1.15778781e-01
2.59543717e-01 5.66199124e-01 -3.52847248e-01 -3.88491720e-01
4.25279409e-01 6.09357774e-01 4.00466710e-01 -6.29551947e-01
6.02035046e-01 -4.97545190e-02 1.79771602e-01 -7.49929011e-01
-5.26825428e-01 -7.77454853e-01 -8.61787498e-01 -3.16569358e-01
5.80655575e-01 -8.14590335e-01 -1.31310952e+00 -1.26470253e-01
-1.47569120e+00 7.48541772e-01 -6.46323292e-03 7.46842146e-01
3.08240503e-01 5.72560318e-02 -7.02008724e-01 -8.10318470e-01
-6.38150632e-01 -4.93673414e-01 1.82719946e-01 3.57870936e-01
-2.95952678e-01 -1.69479895e+00 2.83009827e-01 1.94304083e-02
4.26969916e-01 2.14035437e-01 1.38902700e+00 -9.74665105e-01
-3.86103153e-01 -1.37348026e-01 -4.94500965e-01 -4.25292522e-01
4.45456564e-01 2.42754430e-01 -4.60886717e-01 1.36358827e-01
-2.98305571e-01 2.41610780e-01 1.15398669e+00 5.16236365e-01
9.10384417e-01 -6.99055076e-01 -9.35098350e-01 2.19239354e-01
1.26161945e+00 7.34691262e-01 5.86747192e-02 6.34921849e-01
8.20791543e-01 7.80225873e-01 -1.29901528e-01 3.09190750e-01
5.33633292e-01 1.93760946e-01 -1.35487527e-01 2.39905834e-01
3.11488003e-01 -2.40709022e-01 -5.32744303e-02 1.62241137e+00
-5.60010612e-01 -1.66793853e-01 -1.16197217e+00 1.00568771e+00
-1.83635294e+00 -1.20742822e+00 -3.77112985e-01 1.67687750e+00
1.32290232e+00 2.00863853e-01 -1.85764432e-01 -2.39831299e-01
8.49077225e-01 -5.66073740e-03 -6.73785150e-01 -4.17864799e-01
-4.68230903e-01 -3.96520793e-02 4.24180686e-01 2.50714213e-01
-6.22808397e-01 9.16205764e-01 5.64605665e+00 5.85867703e-01
-1.08412647e+00 -1.70106925e-02 5.25798440e-01 7.30520040e-02
-6.71240985e-01 1.11847565e-01 -5.82362235e-01 5.82526684e-01
1.27161324e+00 -1.00119746e+00 2.25476459e-01 5.53503215e-01
2.44498849e-01 -3.85147110e-02 -8.67389679e-01 7.57968903e-01
-1.28981471e-01 -1.98578680e+00 3.50809693e-01 1.24620058e-01
8.52884173e-01 4.94036674e-02 9.06163827e-03 1.47081971e-01
4.66514677e-01 -9.20247316e-01 -7.39781326e-03 1.19488537e+00
5.02166331e-01 -5.67514837e-01 8.63291502e-01 1.32917896e-01
-1.36687088e+00 -1.09680228e-01 -1.66434005e-01 -1.56070963e-01
-1.12366199e-01 9.39142942e-01 -9.33388352e-01 1.17333114e+00
9.29427505e-01 1.48232675e+00 -5.43922901e-01 8.62289608e-01
4.84575704e-02 6.93118453e-01 1.71896785e-01 -2.77820975e-01
3.94298643e-01 -3.23768646e-01 5.85271418e-01 1.38419926e+00
4.50651079e-01 1.54536158e-01 -2.32915625e-01 1.17553508e+00
-5.20111501e-01 3.83690745e-01 -9.96053994e-01 -6.63425624e-01
7.21956372e-01 1.38873959e+00 -1.00509202e+00 -2.27750272e-01
-3.02537471e-01 3.03036064e-01 3.47969294e-01 7.38830745e-01
-3.20605218e-01 -4.71603304e-01 2.29792148e-01 -2.57458314e-02
1.12744644e-01 -1.43522412e-01 -5.73295236e-01 -1.01248908e+00
-2.51960695e-01 4.16368358e-02 5.43075860e-01 -6.54717803e-01
-1.72831404e+00 6.18909359e-01 -2.47084916e-01 -8.46017957e-01
1.13656966e-03 -3.59423935e-01 -8.66115868e-01 1.22557056e+00
-1.47806311e+00 -1.06376791e+00 1.14806987e-01 5.97357988e-01
1.33394763e-01 -6.02893472e-01 6.90735519e-01 1.81935608e-01
-7.25534558e-01 7.67270625e-02 3.94051611e-01 1.81987166e-01
2.99975395e-01 -9.76808727e-01 -8.57325867e-02 3.86972934e-01
2.43441701e-01 1.07838583e+00 2.29993373e-01 -1.27137291e+00
-1.37640536e+00 -8.52238715e-01 1.32926714e+00 -3.49268615e-01
1.60929585e+00 2.82177657e-01 -1.14652050e+00 6.17532253e-01
5.22581995e-01 -1.89559191e-01 9.57291663e-01 6.03773832e-01
-9.85149145e-02 1.11028619e-01 -5.26243985e-01 3.58312368e-01
1.05721903e+00 -6.72425568e-01 -9.33836877e-01 4.37031299e-01
8.88447940e-01 3.78069311e-01 -1.43913233e+00 3.98143530e-01
5.14389038e-01 1.62394773e-02 9.21316624e-01 -9.70538497e-01
5.39999485e-01 1.33458495e-01 4.59802240e-01 -1.32648146e+00
-8.53946090e-01 -3.19849610e-01 -5.59280634e-01 1.33851588e+00
7.68769085e-01 -7.30406702e-01 5.49823046e-01 2.22847342e-01
2.04258934e-01 -9.34962332e-01 -1.05779994e+00 -5.99199235e-01
1.19031720e-01 3.37985232e-02 3.06581985e-02 1.91154075e+00
6.73581362e-01 7.26293385e-01 -1.56546891e-01 -7.30573684e-02
3.40211064e-01 1.82385996e-01 -2.48567030e-01 -1.94407427e+00
3.83782297e-01 -8.54618251e-01 -4.86250639e-01 -9.97152478e-02
7.03356504e-01 -1.50665903e+00 -7.55708456e-01 -1.72394443e+00
1.72346935e-01 -3.95124674e-01 -9.05034125e-01 5.98707795e-01
5.97836226e-02 -5.20305991e-01 -3.39112610e-01 6.33787632e-01
-3.59709024e-01 5.06700695e-01 9.67580616e-01 -3.28839660e-01
-6.03247941e-01 -1.93804637e-01 -5.18660367e-01 6.99502468e-01
4.85722095e-01 -4.48312372e-01 -3.06412697e-01 -4.91062030e-02
7.25207865e-01 2.69633591e-01 1.45671889e-01 -5.40409565e-01
1.06520665e+00 -2.53294796e-01 4.91762310e-01 -7.05557287e-01
-2.04202995e-01 -8.88951480e-01 2.88993448e-01 4.30914968e-01
-5.61336339e-01 2.66957968e-01 3.94174188e-01 9.43351567e-01
-4.25157845e-01 2.16240343e-02 -1.35081515e-01 -1.81058526e-01
-8.16966832e-01 2.92619944e-01 -5.69772720e-01 -2.49006316e-01
9.88223314e-01 -1.77515984e-01 -4.62499529e-01 -8.15297961e-02
-9.84299839e-01 3.54524612e-01 -4.67803061e-01 7.82504857e-01
7.95871675e-01 -1.24969363e+00 -5.71520329e-01 -3.14431399e-01
-5.52217066e-02 -5.98695338e-01 9.31526646e-02 9.73031878e-01
-6.38852641e-02 1.01433909e+00 -1.54644668e-01 -1.55837193e-01
-1.00213921e+00 7.89420784e-01 1.33795021e-02 -4.81822580e-01
-4.41711217e-01 7.63131499e-01 -3.13476950e-01 -1.04484200e-01
2.10602686e-01 -3.13482769e-02 -1.06128597e+00 8.04610550e-01
-4.42952588e-02 4.72817868e-01 -1.07926369e-01 -7.39470005e-01
-5.03498495e-01 5.34171641e-01 -1.76671222e-01 2.93354154e-01
1.61674595e+00 -1.23466216e-01 -9.91368055e-01 1.10214996e+00
1.16076899e+00 -3.67132723e-01 8.81726891e-02 -8.22990716e-01
8.05719256e-01 1.74677759e-01 2.51088172e-01 -9.83280599e-01
-1.12854993e+00 4.17329103e-01 2.53877848e-01 5.77906370e-01
6.19135857e-01 4.58580367e-02 5.33429325e-01 5.72486699e-01
1.71102166e-01 -9.66068864e-01 -1.32015735e-01 5.89639843e-01
7.35135853e-01 -8.73818040e-01 2.53825277e-01 -1.91330254e-01
-1.13559283e-01 1.57136130e+00 5.25996029e-01 3.94778013e-01
1.38156033e+00 5.01963263e-03 -4.47192937e-01 -7.96379447e-01
-9.08469975e-01 3.02186068e-02 7.95837283e-01 4.28479584e-03
5.35608470e-01 -1.05910249e-01 -7.83261061e-01 6.95967317e-01
1.68143567e-02 2.92542011e-01 3.15629452e-01 5.72536588e-01
-2.59606719e-01 -6.07021928e-01 -1.88446954e-01 9.24989581e-01
-5.26117802e-01 -4.80370581e-01 -4.76509184e-01 4.39424813e-01
2.29089543e-01 7.16360509e-01 1.72523275e-01 -6.81863904e-01
-9.12172347e-03 2.47158527e-01 -3.49611849e-01 -3.12260628e-01
-7.20677197e-01 -4.41984385e-01 -3.04130372e-02 4.75366451e-02
-6.41216695e-01 -2.46827200e-01 -1.13758349e+00 -4.19173360e-01
-4.76037592e-01 6.63158298e-01 5.38550377e-01 1.25361311e+00
8.26500654e-01 1.14265442e+00 6.01961851e-01 -4.08133209e-01
3.89966249e-01 -1.14540994e+00 -7.22719431e-01 6.82563260e-02
-6.10840060e-02 -8.25096250e-01 -3.86196434e-01 7.70214722e-02] | [9.469139099121094, 8.184218406677246] |
e9bc7c1f-62fd-4dc3-b684-89700ef92fb5 | mobilevig-graph-based-sparse-attention-for | 2307.00395 | null | https://arxiv.org/abs/2307.00395v1 | https://arxiv.org/pdf/2307.00395v1.pdf | MobileViG: Graph-Based Sparse Attention for Mobile Vision Applications | Traditionally, convolutional neural networks (CNN) and vision transformers (ViT) have dominated computer vision. However, recently proposed vision graph neural networks (ViG) provide a new avenue for exploration. Unfortunately, for mobile applications, ViGs are computationally expensive due to the overhead of representing images as graph structures. In this work, we propose a new graph-based sparse attention mechanism, Sparse Vision Graph Attention (SVGA), that is designed for ViGs running on mobile devices. Additionally, we propose the first hybrid CNN-GNN architecture for vision tasks on mobile devices, MobileViG, which uses SVGA. Extensive experiments show that MobileViG beats existing ViG models and existing mobile CNN and ViT architectures in terms of accuracy and/or speed on image classification, object detection, and instance segmentation tasks. Our fastest model, MobileViG-Ti, achieves 75.7% top-1 accuracy on ImageNet-1K with 0.78 ms inference latency on iPhone 13 Mini NPU (compiled with CoreML), which is faster than MobileNetV2x1.4 (1.02 ms, 74.7% top-1) and MobileNetV2x1.0 (0.81 ms, 71.8% top-1). Our largest model, MobileViG-B obtains 82.6% top-1 accuracy with only 2.30 ms latency, which is faster and more accurate than the similarly sized EfficientFormer-L3 model (2.77 ms, 82.4%). Our work proves that well designed hybrid CNN-GNN architectures can be a new avenue of exploration for designing models that are extremely fast and accurate on mobile devices. Our code is publicly available at https://github.com/SLDGroup/MobileViG. | ['Radu Marculescu', 'William Avery', 'Mustafa Munir'] | 2023-07-01 | null | null | null | null | ['instance-segmentation', 'graph-attention'] | ['computer-vision', 'graphs'] | [-2.19494164e-01 2.16395289e-01 -4.97396111e-01 3.57218832e-02
-3.36843163e-01 -3.85846406e-01 1.98140010e-01 -4.27633196e-01
-3.71799290e-01 3.17623824e-01 -2.39359379e-01 -9.55669701e-01
3.45912963e-01 -8.25800359e-01 -1.06737232e+00 -2.01796085e-01
1.39649175e-02 3.58019292e-01 2.37120405e-01 -4.41219583e-02
-4.34774496e-02 7.52108991e-02 -1.27159309e+00 4.67423573e-02
5.70439219e-01 1.10508120e+00 4.64558184e-01 1.24783337e+00
-1.50240108e-01 9.38507080e-01 -3.98464829e-01 -5.03814757e-01
1.01193517e-01 -1.25484690e-01 -9.85168040e-01 -2.00024903e-01
9.02620316e-01 -3.11829031e-01 -7.87149906e-01 1.20193672e+00
3.01326364e-01 -1.39083207e-01 1.42552212e-01 -1.49472952e+00
-1.04152966e+00 7.81019032e-01 -5.98940551e-01 2.81547099e-01
-2.97745857e-02 5.97591937e-01 1.01309645e+00 -8.41414392e-01
4.61839795e-01 1.25711524e+00 8.70648503e-01 7.71853089e-01
-6.65113330e-01 -6.07090771e-01 4.25220191e-01 3.91965628e-01
-1.41679192e+00 -3.87015492e-01 3.61574322e-01 -4.28109355e-02
1.37853825e+00 8.92979875e-02 9.47685778e-01 1.11364186e+00
2.48821110e-01 1.04882932e+00 6.62827134e-01 -1.20428793e-01
1.29927933e-01 -2.60256290e-01 3.92196238e-01 1.33314788e+00
5.35538435e-01 -2.49209806e-01 -3.47028077e-01 1.79135725e-01
8.92776310e-01 7.16215149e-02 -3.06048274e-01 9.78001207e-02
-9.39702332e-01 6.58738315e-01 1.06036270e+00 3.36674899e-02
-3.40519428e-01 1.14803720e+00 3.71735930e-01 1.54696316e-01
2.35398993e-01 2.27891952e-01 -3.56860936e-01 -3.54444742e-01
-7.71188974e-01 -8.63538980e-02 8.41053784e-01 1.43691897e+00
6.18898809e-01 6.45227194e-01 3.84592712e-02 6.07468188e-01
4.39997137e-01 7.65552580e-01 4.12806809e-01 -8.82273912e-01
4.14097250e-01 5.79859018e-01 -4.82008696e-01 -1.04096091e+00
-4.11029935e-01 -7.76182950e-01 -1.00363195e+00 -1.60021797e-01
2.68223464e-01 -9.42584425e-02 -1.37013137e+00 1.39599395e+00
-1.09838903e-01 4.85234737e-01 1.51473396e-02 7.96045482e-01
1.43181086e+00 7.26964533e-01 6.20636754e-02 4.11029279e-01
1.36044478e+00 -1.59672058e+00 -1.72116414e-01 -6.53848231e-01
6.30093217e-01 -3.92275602e-01 1.49340153e+00 2.58075655e-01
-1.19508278e+00 -7.43163347e-01 -9.90825534e-01 -2.23684624e-01
-3.26780945e-01 1.87410221e-01 1.02030993e+00 6.50858641e-01
-1.76760197e+00 2.72378325e-01 -9.18692231e-01 -5.28304636e-01
8.79871309e-01 6.75553143e-01 6.76041394e-02 -2.37987816e-01
-6.72408104e-01 2.48818919e-01 1.12925604e-01 2.20933989e-01
-1.21388936e+00 -5.12796879e-01 -8.88105929e-01 3.04740399e-01
5.83844662e-01 -9.44157600e-01 1.40034199e+00 -9.83671665e-01
-1.32379103e+00 7.57134676e-01 -4.75214757e-02 -8.71826589e-01
1.57722667e-01 -9.55786556e-02 -2.37937078e-01 1.50185183e-01
1.52237872e-02 1.03320611e+00 7.93036461e-01 -9.79948401e-01
-4.67135787e-01 -1.87021226e-01 3.79671544e-01 3.94531563e-02
-2.43446037e-01 -3.23367566e-01 -1.31557226e+00 -1.95112735e-01
-1.14540368e-01 -1.25516438e+00 -4.09194022e-01 -6.43937439e-02
-6.06925726e-01 -9.62332860e-02 1.00140452e+00 -4.99272048e-01
1.08665502e+00 -1.92567909e+00 -1.95185632e-01 7.69051686e-02
8.57592404e-01 8.10920060e-01 -3.92470151e-01 -1.59415856e-01
3.65677983e-01 3.18307757e-01 -1.37469480e-02 -5.05891979e-01
-2.41937280e-01 2.73177236e-01 9.21821147e-02 1.40125901e-01
-1.10238111e-02 1.85073972e+00 -9.11765635e-01 -2.34828219e-01
1.93191484e-01 4.92413700e-01 -6.80860519e-01 -1.23382390e-01
-3.86217147e-01 -1.20297834e-01 -3.03281724e-01 1.18429303e+00
5.46829402e-01 -8.57115388e-01 2.87713587e-01 -2.43356630e-01
1.13851801e-01 1.70483775e-02 -5.94284356e-01 1.68775606e+00
-5.02480328e-01 1.02501619e+00 1.24454819e-01 -9.30642903e-01
7.34037340e-01 -1.86781019e-01 4.24705446e-02 -9.13722575e-01
4.48999912e-01 5.01455255e-02 -1.20193668e-01 -2.70800620e-01
6.75541520e-01 8.11372161e-01 1.43201187e-01 1.19865581e-01
2.60787606e-01 9.18160006e-02 6.56564012e-02 5.87814808e-01
1.34836602e+00 -1.46653444e-01 -2.55272016e-02 -2.43431300e-01
1.24258056e-01 4.18743603e-02 2.63863355e-01 1.11321282e+00
-3.58857155e-01 5.06393611e-01 6.81435466e-01 -6.05015695e-01
-7.92834103e-01 -8.45005214e-01 5.88560343e-01 8.73281896e-01
3.10699224e-01 -8.64928186e-01 -9.49837863e-01 -7.48504817e-01
-3.19027156e-01 2.32834071e-01 -3.54046345e-01 -1.30992264e-01
-6.08650625e-01 -5.15375733e-01 8.37977469e-01 9.06641960e-01
9.55659866e-01 -1.19784784e+00 -4.51651871e-01 1.78015158e-01
1.62139013e-02 -1.37647152e+00 -5.64864099e-01 -2.10788757e-01
-9.79713678e-01 -1.23928165e+00 -7.00111270e-01 -1.01682103e+00
4.98803020e-01 7.09969699e-01 1.49447298e+00 6.32056653e-01
-7.77296349e-02 7.12729335e-01 -3.51932645e-01 -2.77575523e-01
2.12594792e-02 5.93317688e-01 -2.44089931e-01 -3.38310540e-01
2.13843837e-01 -3.86321485e-01 -7.16882527e-01 -9.19575095e-02
-5.41355669e-01 3.37184340e-01 5.13498306e-01 8.18671346e-01
8.48113179e-01 -3.33997726e-01 2.18794107e-01 -9.72964048e-01
3.99363995e-01 -5.21160781e-01 -7.74680555e-01 2.54353315e-01
-8.02667141e-01 -2.62500882e-01 6.40200496e-01 -3.91180754e-01
-5.07774115e-01 -7.66719785e-03 -3.10352147e-01 -8.85883927e-01
1.55273125e-01 6.50659442e-01 2.88996063e-02 -4.26063836e-01
5.63660502e-01 2.63141394e-01 4.22870107e-02 -2.25441575e-01
5.08641958e-01 5.46068251e-01 7.43774235e-01 -1.42714322e-01
5.07276952e-01 2.63589054e-01 -6.32846355e-02 -1.07528245e+00
-5.59016526e-01 -2.65313774e-01 2.50295573e-03 -1.83644742e-01
8.61646354e-01 -1.15658772e+00 -1.01083004e+00 6.82116985e-01
-9.29441214e-01 -1.07739329e+00 6.31405413e-02 1.48026869e-01
-2.80935496e-01 3.11278731e-01 -8.38540256e-01 -5.47619164e-01
-9.66576278e-01 -1.34703612e+00 1.09036565e+00 4.66262609e-01
8.94542187e-02 -1.03792560e+00 -5.94968319e-01 4.29145873e-01
6.56948328e-01 -9.70317125e-02 5.29149771e-01 -1.01633534e-01
-1.13608098e+00 6.26351163e-02 -7.88640857e-01 2.53862381e-01
-2.61342078e-01 1.20277517e-01 -9.46769476e-01 -4.14406061e-01
-5.62771022e-01 -2.68308848e-01 1.07145703e+00 8.78068149e-01
1.67076015e+00 -4.66381013e-01 -3.93438071e-01 1.23448265e+00
1.63919842e+00 2.61374652e-01 8.84484351e-01 2.36541376e-01
1.59032273e+00 -2.51319200e-01 1.71668544e-01 -1.58265475e-02
7.91491151e-01 4.12352026e-01 1.00837910e+00 -3.49364936e-01
-5.53660214e-01 -2.52613872e-01 3.16039324e-01 9.31596637e-01
-2.46336356e-01 -5.10770142e-01 -1.09635687e+00 6.76103890e-01
-2.00082850e+00 -6.00211263e-01 -4.73801762e-01 1.71685278e+00
1.20887361e-01 3.86000186e-01 1.90626144e-01 -2.63087064e-01
5.09977341e-01 2.43889645e-01 -6.51852429e-01 -5.55144846e-01
8.42112228e-02 5.31541407e-01 9.04277265e-01 5.47160625e-01
-1.00224781e+00 1.47328579e+00 5.41057396e+00 7.83551395e-01
-1.28904450e+00 1.01591773e-01 8.27771127e-01 -1.47716608e-02
-2.86278039e-01 -6.02359325e-02 -8.66039991e-01 4.45244670e-01
1.18143535e+00 1.06968746e-01 6.76203489e-01 1.30525637e+00
-1.99551195e-01 1.87728599e-01 -6.47797942e-01 1.46998215e+00
-1.74551196e-02 -1.83492053e+00 1.62697539e-01 9.13050547e-02
6.05209053e-01 6.48214519e-01 1.75986022e-01 5.22802114e-01
2.94776142e-01 -1.19100499e+00 6.81160033e-01 2.63435334e-01
1.15565264e+00 -6.94667637e-01 6.85687542e-01 -1.16475888e-01
-1.54854631e+00 1.08650789e-01 -4.28687274e-01 -1.24361381e-01
-6.07264228e-02 2.55469412e-01 -8.68121147e-01 2.21124873e-01
1.05636132e+00 9.79251504e-01 -8.60542595e-01 9.19995546e-01
-1.58929482e-01 1.06567097e+00 -2.02922836e-01 -3.18537891e-01
7.04354942e-01 1.11056775e-01 4.04925913e-01 1.21341419e+00
2.61629671e-01 -1.53262109e-01 2.06999362e-01 8.45087588e-01
-5.40357590e-01 -2.76950985e-01 -6.94068313e-01 -2.45106146e-01
3.96494478e-01 1.23658001e+00 -1.00489402e+00 -6.76061273e-01
-5.17724514e-01 1.15671575e+00 3.83320063e-01 5.31503677e-01
-1.15640759e+00 -4.29973334e-01 7.72410989e-01 -1.77284718e-01
4.88318712e-01 -3.77746999e-01 -2.57586479e-01 -1.23567080e+00
2.26331875e-02 -9.08893645e-01 2.58733332e-01 -9.85235751e-01
-7.50496209e-01 8.79813433e-01 -4.35002983e-01 -6.76008344e-01
8.38560984e-02 -8.71440947e-01 -6.17801547e-01 4.72850680e-01
-1.39506996e+00 -1.44422805e+00 -8.37085068e-01 6.48207903e-01
8.32825541e-01 -1.90432131e-01 5.74939907e-01 3.08026999e-01
-7.45081782e-01 1.02656364e+00 -2.45131001e-01 2.86047935e-01
-1.01779640e-01 -1.24480808e+00 1.39978027e+00 8.87417078e-01
3.71417373e-01 5.56090176e-01 9.08043906e-02 -5.87727010e-01
-1.96063817e+00 -1.51920843e+00 6.52676821e-01 -2.61891901e-01
5.56546509e-01 -3.47746193e-01 -8.59916866e-01 9.96201038e-01
9.75423232e-02 1.69446737e-01 1.56192303e-01 1.57431006e-01
-4.47275728e-01 -4.01148461e-02 -7.99403489e-01 9.46152329e-01
1.55331457e+00 -6.29686713e-01 2.31245667e-01 4.21580911e-01
1.29021716e+00 -6.81007802e-01 -4.82523859e-01 4.44406755e-02
5.13079822e-01 -6.72522664e-01 1.02709246e+00 -4.48947817e-01
2.61260539e-01 -2.54107624e-01 -1.71646446e-01 -9.64213133e-01
-4.46631372e-01 -6.67866349e-01 -6.51694596e-01 6.95682526e-01
4.46724951e-01 -7.00799704e-01 1.11946571e+00 2.55519629e-01
-4.23973799e-01 -1.11845744e+00 -4.68777031e-01 -7.87471294e-01
-3.39514673e-01 -8.94440830e-01 6.05660200e-01 6.39795721e-01
-6.92125440e-01 5.16194701e-01 -4.19608057e-01 7.88434893e-02
5.87147236e-01 -1.41044065e-01 8.99269998e-01 -8.32202137e-01
-5.26928425e-01 -5.31952262e-01 -5.99040806e-01 -1.33827543e+00
-3.09620164e-02 -8.94315422e-01 -3.42596680e-01 -1.89290392e+00
1.00360446e-01 -4.77418065e-01 -1.20283656e-01 7.97863126e-01
-8.22016597e-02 7.05618501e-01 4.10414785e-01 4.34628315e-02
-9.66015995e-01 3.29986691e-01 1.16195691e+00 -5.58814764e-01
-2.77765304e-01 -1.48152662e-02 -1.11168253e+00 7.18199015e-01
1.10597110e+00 -9.73019078e-02 -6.34940863e-01 -8.26874793e-01
2.31548011e-01 -2.04639465e-01 7.90527821e-01 -1.21041048e+00
3.12090367e-01 2.10471958e-01 4.36423533e-02 -5.34699380e-01
2.44181320e-01 -3.14091116e-01 -1.94176298e-03 7.05934286e-01
1.92023233e-01 3.44040692e-01 4.01605576e-01 5.29051602e-01
2.65235510e-02 1.68033689e-01 3.90214622e-01 -3.96505892e-01
-1.44766009e+00 8.09125721e-01 -3.01375687e-01 1.02545463e-01
6.41939759e-01 -4.04802203e-01 -6.29154503e-01 -5.60249329e-01
-5.32259166e-01 3.04854810e-01 3.04210156e-01 4.53651339e-01
9.02771652e-01 -1.12817729e+00 -3.81628364e-01 2.31046136e-02
2.23699473e-02 1.92876056e-01 3.18978041e-01 6.86134219e-01
-9.08980191e-01 5.66653013e-01 -1.13329642e-01 -8.48382652e-01
-1.28994286e+00 4.41733658e-01 3.84272873e-01 -8.87400284e-02
-7.63835549e-01 1.26048148e+00 1.99681416e-01 -1.31853744e-01
3.99878621e-01 -7.23066807e-01 4.45978940e-02 -4.77759063e-01
4.73731220e-01 4.90184724e-01 8.69547725e-02 -3.80939513e-01
-5.01580656e-01 5.23194492e-01 -1.52244657e-01 3.90965998e-01
1.06778920e+00 1.31159760e-02 -4.28475365e-02 1.10400934e-02
1.24194682e+00 -4.96356338e-01 -1.21774483e+00 -7.29198009e-03
-3.84302735e-01 -1.59626156e-01 2.64970273e-01 -5.21106660e-01
-1.66590178e+00 7.79949248e-01 6.78293347e-01 1.11569412e-01
1.13293540e+00 1.31758526e-01 1.18077493e+00 5.34671187e-01
5.43021262e-01 -7.14911401e-01 1.90455735e-01 8.78538787e-01
6.23721957e-01 -1.20916522e+00 -2.27380186e-01 -3.87905449e-01
-4.88124847e-01 7.78109431e-01 1.09529066e+00 -2.27601171e-01
6.17456436e-01 8.73259604e-02 -1.31912222e-02 -6.88519418e-01
-6.45015657e-01 -4.15813625e-01 3.58042359e-01 7.82265782e-01
2.90373504e-01 4.51315939e-01 1.66155368e-01 3.62185955e-01
-2.68584639e-01 -2.86894441e-02 5.43109834e-01 6.87080443e-01
-1.84453830e-01 -6.45931482e-01 1.11075573e-01 5.81656396e-01
-4.20194149e-01 -5.42363167e-01 -3.96561325e-01 8.77264082e-01
-7.61403367e-02 9.57837105e-01 1.65336624e-01 -8.36720884e-01
8.63065571e-02 -4.87317950e-01 3.68102610e-01 -6.00406528e-01
-5.70853472e-01 -2.31267899e-01 1.79955289e-01 -1.03071558e+00
-5.98907880e-02 1.03118025e-01 -1.22143900e+00 -7.48189926e-01
-2.30919629e-01 -2.24705562e-01 7.35371232e-01 5.15188158e-01
7.50355601e-01 8.41995776e-01 1.96516011e-02 -8.42597127e-01
-1.84816439e-02 -5.20011067e-01 -2.82871187e-01 -2.22756654e-01
2.03984916e-01 -2.41280362e-01 2.98723672e-02 -1.39969632e-01] | [9.287571907043457, 1.3897967338562012] |
f5b62a30-8d4c-4f15-9342-a8a7eeee5122 | trustworthy-deep-learning-for-medical-image | 2305.17456 | null | https://arxiv.org/abs/2305.17456v1 | https://arxiv.org/pdf/2305.17456v1.pdf | Trustworthy Deep Learning for Medical Image Segmentation | Despite the recent success of deep learning methods at achieving new state-of-the-art accuracy for medical image segmentation, some major limitations are still restricting their deployment into clinics. One major limitation of deep learning-based segmentation methods is their lack of robustness to variability in the image acquisition protocol and in the imaged anatomy that were not represented or were underrepresented in the training dataset. This suggests adding new manually segmented images to the training dataset to better cover the image variability. However, in most cases, the manual segmentation of medical images requires highly skilled raters and is time-consuming, making this solution prohibitively expensive. Even when manually segmented images from different sources are available, they are rarely annotated for exactly the same regions of interest. This poses an additional challenge for current state-of-the-art deep learning segmentation methods that rely on supervised learning and therefore require all the regions of interest to be segmented for all the images to be used for training. This thesis introduces new mathematical and optimization methods to mitigate those limitations. | ['Lucas Fidon'] | 2023-05-27 | null | null | null | null | ['anatomy'] | ['miscellaneous'] | [ 1.94833621e-01 2.51360893e-01 -2.79364526e-01 -5.38174748e-01
-8.83025885e-01 -5.53417981e-01 -7.94382244e-02 4.47509348e-01
-6.45999372e-01 6.80449724e-01 -3.15149039e-01 -4.89189118e-01
-4.42701671e-03 -5.68182290e-01 -5.29668570e-01 -6.60225630e-01
7.64380917e-02 7.58542180e-01 2.06308201e-01 1.34349838e-01
7.62534663e-02 7.35017776e-01 -1.08261645e+00 1.77041993e-01
8.49073827e-01 7.14053273e-01 3.32115680e-01 5.47645032e-01
-3.41490597e-01 3.98641437e-01 -7.93202162e-01 -2.20081463e-01
4.14599359e-01 -6.69092715e-01 -9.64202583e-01 4.19978917e-01
5.58144331e-01 -5.74907422e-01 -1.11264892e-01 8.14580083e-01
6.56504035e-01 4.32959720e-02 4.72086608e-01 -8.41610074e-01
-2.53422797e-01 3.67534816e-01 -5.51225960e-01 3.77882540e-01
-7.82345831e-02 7.86224082e-02 4.48583543e-01 -2.92717129e-01
7.42495179e-01 5.89900970e-01 8.14281046e-01 6.36184096e-01
-1.32716656e+00 -5.86496174e-01 -1.10372871e-01 -3.64861250e-01
-1.15931189e+00 -1.63723126e-01 5.32649577e-01 -6.11492217e-01
5.92330873e-01 1.87374189e-01 9.85192299e-01 5.72724223e-01
2.74493039e-01 7.08007276e-01 1.13588905e+00 -2.68483430e-01
1.67184934e-01 2.68091500e-01 -8.42156187e-02 8.29159796e-01
3.70753735e-01 -2.34905899e-01 1.05896071e-01 1.01011116e-02
1.10111296e+00 -5.10382559e-03 -9.21439677e-02 -5.72471321e-01
-1.01891196e+00 9.51140940e-01 5.87172508e-01 7.52837121e-01
-3.32558692e-01 -1.19953148e-01 4.27442372e-01 -1.62928134e-01
2.92735219e-01 6.25123978e-01 -4.91074502e-01 -5.69943935e-02
-1.69833505e+00 2.44714506e-02 4.92809236e-01 6.36153638e-01
5.99703372e-01 -2.22018771e-02 1.46187544e-01 7.73168206e-01
1.59033090e-01 5.89903258e-02 7.74381042e-01 -1.03402793e+00
1.20269805e-01 8.78035247e-01 -2.35936940e-01 -9.83260274e-01
-6.88245058e-01 -7.44101405e-01 -6.61961019e-01 3.68226051e-01
9.60513353e-01 -7.88008720e-02 -1.55652344e+00 1.15081859e+00
3.71314734e-01 -2.41022587e-01 -2.71563202e-01 9.97396469e-01
9.45574582e-01 2.50358075e-01 1.85746878e-01 -7.50773698e-02
1.05805838e+00 -9.05579805e-01 -5.83837688e-01 -3.07192981e-01
8.31850827e-01 -8.14688802e-01 9.85752106e-01 3.95872504e-01
-1.17977655e+00 -3.45016897e-01 -1.08788860e+00 -2.63123512e-02
-3.42790484e-01 1.08113676e-01 8.63263309e-01 8.85550916e-01
-8.26649308e-01 7.68297791e-01 -1.14980698e+00 -1.45076603e-01
1.08200920e+00 7.31262565e-01 -4.77731556e-01 1.75057054e-02
-7.59775579e-01 1.00163102e+00 2.95321047e-01 2.58196652e-01
-7.35777676e-01 -1.08588231e+00 -8.46372128e-01 -3.01660299e-01
2.33000740e-01 -2.48425111e-01 1.26440024e+00 -1.28719246e+00
-1.22617638e+00 1.37669158e+00 6.44478993e-03 -3.78387094e-01
9.19651389e-01 4.34422027e-03 1.74714271e-02 3.74992549e-01
1.73079103e-01 7.97998488e-01 5.88107646e-01 -1.46072257e+00
-4.58086342e-01 -3.44603688e-01 -8.87184665e-02 4.33185659e-02
1.95514292e-01 -3.28337327e-02 -5.51659107e-01 -6.48510873e-01
3.40061367e-01 -9.48407888e-01 -6.40176773e-01 1.60016239e-01
-3.42967421e-01 3.10247958e-01 8.55026662e-01 -8.87680292e-01
8.80229354e-01 -2.04370594e+00 -1.50850207e-01 3.02555352e-01
4.93760705e-01 6.33938313e-01 1.84534758e-01 -8.45694169e-02
-1.26448860e-02 4.29713339e-01 -4.91636127e-01 -2.57661730e-01
-5.63752770e-01 3.86542797e-01 3.83194298e-01 7.13218272e-01
-1.07732259e-01 8.10025573e-01 -8.74882102e-01 -9.21639621e-01
7.09223628e-01 5.19774139e-01 -2.08943069e-01 9.36150327e-02
-1.44442121e-04 1.04993713e+00 -4.21633542e-01 6.11791492e-01
5.89838147e-01 -1.63165018e-01 5.47059067e-02 -1.21760711e-01
1.31266832e-01 1.93137079e-02 -1.16223264e+00 1.77115750e+00
-3.31859976e-01 6.00787282e-01 3.17452639e-01 -1.51355648e+00
5.59394419e-01 4.68963832e-01 1.26288664e+00 -5.54434240e-01
3.94199401e-01 4.68935639e-01 3.78373384e-01 -5.16631424e-01
1.72637463e-01 -4.63175714e-01 1.90302655e-01 3.92351449e-01
1.30636036e-01 -4.63735968e-01 3.52918029e-01 -9.19852033e-02
5.29006541e-01 6.70980662e-02 5.87194264e-02 -2.45151430e-01
3.50709796e-01 4.71332073e-01 7.98840582e-01 6.07929885e-01
-5.70352733e-01 1.07633054e+00 4.28836286e-01 -6.45085454e-01
-1.16185546e+00 -8.21290076e-01 -5.34472406e-01 5.46919763e-01
-1.05314448e-01 1.38857961e-01 -1.18789208e+00 -8.15735281e-01
-1.52894899e-01 2.75809169e-01 -6.91054761e-01 1.33831710e-01
-6.20962143e-01 -7.14780927e-01 5.94690681e-01 5.79503417e-01
3.83741945e-01 -1.02895677e+00 -1.10311246e+00 4.19802040e-01
-1.13023154e-01 -1.20003128e+00 -1.11836269e-01 1.11896336e-01
-1.33218312e+00 -1.24100995e+00 -1.26390886e+00 -8.59468460e-01
1.12923872e+00 -1.79195344e-01 1.13062882e+00 4.23078895e-01
-8.02657366e-01 2.73239702e-01 -8.96049961e-02 -7.11710811e-01
-5.56890368e-01 2.61130959e-01 -5.72647810e-01 -3.28480899e-01
-3.26686762e-02 -1.25820756e-01 -7.35386074e-01 1.90088078e-01
-8.52006137e-01 -3.62100564e-02 5.42675197e-01 7.85472691e-01
9.87709224e-01 1.87513828e-01 3.09412867e-01 -1.25429201e+00
3.30899984e-01 -8.27120543e-02 -5.49385965e-01 1.43974662e-01
-3.08023453e-01 -3.75529885e-01 3.54571730e-01 -4.14366633e-01
-8.37467194e-01 4.25918579e-01 -1.05094314e-01 -2.25693583e-01
-5.11485040e-01 6.43276632e-01 2.66371071e-01 -3.32219630e-01
4.99450475e-01 -1.12813756e-01 4.41835344e-01 -1.44289136e-01
-1.40045777e-01 3.90867889e-01 3.58752578e-01 -1.57929599e-01
3.50478232e-01 5.47496617e-01 9.85231921e-02 -9.10311043e-01
-9.67042744e-01 -4.47105080e-01 -1.09336698e+00 -2.58573055e-01
1.15768301e+00 -4.85213280e-01 -1.49515465e-01 3.70584935e-01
-7.29185820e-01 -6.12360418e-01 -3.06764364e-01 6.66900933e-01
-4.33178425e-01 2.56744027e-01 -5.75321019e-01 -4.12550032e-01
-2.82553554e-01 -1.85535204e+00 8.31059754e-01 4.52759683e-01
-4.60307091e-01 -1.30519331e+00 -2.86483407e-01 7.67980158e-01
4.61854428e-01 6.97504878e-01 7.76051044e-01 -5.05396008e-01
-3.65619034e-01 -6.44741356e-01 -1.53088635e-02 4.12776440e-01
4.34429556e-01 7.23685026e-02 -6.85393155e-01 -2.14001521e-01
3.87132168e-02 -2.64953613e-01 5.18394828e-01 1.11514592e+00
1.40331566e+00 3.17278147e-01 -4.44111258e-01 6.24338865e-01
1.39755785e+00 3.58230919e-01 4.64967012e-01 3.68025988e-01
7.04173088e-01 6.62180603e-01 4.89524573e-01 -1.02656014e-01
9.91516858e-02 4.05588061e-01 1.98327333e-01 -9.69889283e-01
-6.90360218e-02 1.88127771e-01 -4.66935307e-01 3.33583444e-01
-1.32583782e-01 2.20093891e-01 -1.10850239e+00 8.13919425e-01
-1.45166159e+00 -5.94705820e-01 -3.45429063e-01 2.18577695e+00
1.00348461e+00 1.03986323e-01 2.44273037e-01 1.31135300e-01
5.56188524e-01 -2.43805841e-01 -6.53444767e-01 -3.87802273e-01
2.63482153e-01 4.08422172e-01 7.81640828e-01 3.83705467e-01
-1.32285798e+00 8.13016891e-01 7.19388819e+00 4.37505275e-01
-1.46957266e+00 6.49735034e-02 1.03497720e+00 -1.36492988e-02
2.32606620e-01 -2.30848625e-01 -3.88942659e-01 3.40537101e-01
6.16584361e-01 2.38689259e-01 9.88536254e-02 7.63156712e-01
3.51200819e-01 -5.69405854e-01 -8.46071839e-01 8.06364954e-01
-4.36544940e-02 -1.27078378e+00 -4.42495733e-01 1.53373122e-01
8.57917905e-01 -7.67163038e-02 1.66711658e-02 -1.14368826e-01
2.25246940e-02 -1.37246835e+00 2.82947212e-01 2.27515206e-01
6.98444784e-01 -6.77694023e-01 1.01472938e+00 1.14823096e-01
-6.53468072e-01 3.17953318e-01 -3.19177210e-01 3.62919867e-01
1.97785094e-01 7.09639370e-01 -8.31608891e-01 3.19471776e-01
6.49818242e-01 2.35279560e-01 -4.49254602e-01 1.20086026e+00
-3.61353047e-02 6.22775435e-01 -3.51467013e-01 3.02576452e-01
5.35331249e-01 -6.07634932e-02 1.68292075e-01 1.12139428e+00
-9.05020162e-02 -1.04266226e-01 4.80727494e-01 8.51056635e-01
8.28505754e-02 1.00916184e-01 -3.26915056e-01 -1.03762649e-01
2.93868091e-02 1.34868670e+00 -1.49724901e+00 -3.22038531e-01
-4.48480397e-01 8.48558247e-01 -2.81546731e-02 2.08314508e-01
-7.78827071e-01 -1.62078828e-01 2.68484294e-01 3.84696335e-01
8.07877481e-02 -2.77319819e-01 -7.02059567e-01 -7.02254355e-01
-1.40379027e-01 -9.85135138e-01 4.69496280e-01 -2.94029742e-01
-9.81938362e-01 3.82845730e-01 1.74265906e-01 -9.06496048e-01
-8.75930116e-02 -6.04819536e-01 -3.98767769e-01 8.57677996e-01
-1.14597893e+00 -1.15566885e+00 -3.80121887e-01 4.83335048e-01
4.90006685e-01 -1.09232194e-03 8.59825015e-01 5.79969466e-01
-4.89944518e-01 5.90848327e-01 -4.33028750e-02 5.74543238e-01
5.34163237e-01 -1.38985538e+00 -8.66094604e-02 5.80052316e-01
1.03921026e-01 6.83853507e-01 4.65492815e-01 -6.22367084e-01
-8.54342103e-01 -7.84122348e-01 4.73876625e-01 -1.96717307e-01
2.48397827e-01 -3.62641998e-02 -9.13866878e-01 7.07257509e-01
-8.22678134e-02 3.42751861e-01 9.72988665e-01 -1.84426501e-01
4.28080112e-01 6.81343749e-02 -1.49430895e+00 4.63405043e-01
3.68748277e-01 -1.49428055e-01 -3.59371871e-01 3.99915516e-01
1.58476412e-01 -1.13448322e+00 -9.89936352e-01 3.26802462e-01
5.46181381e-01 -7.97707558e-01 8.41010571e-01 -4.93768692e-01
5.19570589e-01 -7.79532716e-02 5.15953064e-01 -1.11182141e+00
5.18931523e-02 -1.93333134e-01 3.61172140e-01 9.88325298e-01
6.79684997e-01 -4.49024916e-01 1.35812688e+00 1.11276567e+00
-1.37732536e-01 -1.08717024e+00 -9.04143155e-01 -5.42955279e-01
5.19042432e-01 -2.36225128e-01 2.45531961e-01 1.18764758e+00
-3.63670558e-01 -2.39960238e-01 8.07201117e-02 -1.98165596e-01
6.03204191e-01 5.73811308e-02 5.70906937e-01 -1.11524343e+00
-7.29196146e-03 -6.35768533e-01 -5.87031364e-01 -3.75029564e-01
-1.51228979e-01 -8.20659578e-01 1.07588120e-01 -2.07221222e+00
-3.28688994e-02 -7.12436855e-01 -7.64917433e-02 5.76844811e-01
-1.26755973e-02 5.63578844e-01 -9.39580612e-03 1.14203125e-01
-3.87618020e-02 -1.78223923e-01 1.68101883e+00 -1.60481185e-01
-4.44740474e-01 1.79010347e-01 -6.44740105e-01 8.93085837e-01
1.09401262e+00 -5.89580357e-01 -4.35558259e-01 -4.88199532e-01
-4.65564042e-01 -6.14010058e-02 2.36662179e-01 -1.04559243e+00
-3.17139365e-02 -4.91155013e-02 7.66807377e-01 -5.85928857e-01
4.75626700e-02 -1.02937603e+00 5.79932593e-02 7.78932989e-01
-3.46698195e-01 -1.31281853e-01 4.96873558e-01 -3.63284349e-02
-2.20084161e-01 -6.51330173e-01 1.19124985e+00 -7.52628148e-01
-3.38047743e-01 3.72950822e-01 -5.59680223e-01 2.51085967e-01
1.24294984e+00 -6.40538514e-01 3.24075013e-01 -2.32553884e-01
-1.08010876e+00 6.46267533e-02 4.61495161e-01 -3.71932164e-02
3.40851605e-01 -7.67095149e-01 -5.19495130e-01 7.59254908e-03
-3.43031287e-01 6.18545234e-01 3.24437201e-01 1.04615021e+00
-1.15240479e+00 3.21491510e-01 -3.84415388e-01 -8.04912150e-01
-1.21473014e+00 2.84215987e-01 8.25278401e-01 -4.82231170e-01
-9.29161251e-01 7.48741210e-01 -1.52450874e-01 -5.35240054e-01
3.23109150e-01 -3.96305054e-01 -1.06924571e-01 -4.78662364e-02
1.50328189e-01 1.97645664e-01 2.79060632e-01 -6.31114841e-01
-3.36535841e-01 5.57769358e-01 -2.72163630e-01 -3.86489332e-02
1.18854594e+00 1.86264664e-01 1.47624211e-02 2.98819274e-01
1.10060430e+00 -1.67722449e-01 -1.09851360e+00 2.31194973e-01
-6.25322610e-02 -4.39952463e-01 4.36913192e-01 -8.59832048e-01
-1.59232271e+00 1.03718746e+00 8.17985117e-01 2.22917479e-02
8.75079691e-01 -2.01971397e-01 9.17120218e-01 1.77672487e-02
2.48075098e-01 -1.33252251e+00 -2.77980089e-01 3.27137037e-04
5.12457848e-01 -1.52977610e+00 4.08134729e-01 -5.13546467e-01
-7.22428381e-01 1.19469631e+00 5.58892012e-01 -2.01702267e-01
5.96896946e-01 3.92625988e-01 6.38530552e-01 -3.90493929e-01
4.10344332e-01 -3.76878455e-02 4.26238716e-01 7.80928433e-01
8.92707050e-01 -1.98104419e-02 -6.33569241e-01 1.70110747e-01
-1.89959675e-01 7.03312308e-02 4.99648124e-01 1.13897836e+00
-4.12451625e-02 -1.19031727e+00 -2.82821506e-01 6.47500455e-01
-1.11004448e+00 3.03369135e-01 -2.91071028e-01 1.06607950e+00
2.63001204e-01 8.16210926e-01 -1.03099588e-02 2.52363563e-01
2.12983683e-01 -1.75999656e-01 6.26219511e-01 -7.96700954e-01
-8.64080071e-01 1.72372758e-01 -2.34227806e-01 -5.44762671e-01
-4.87930715e-01 -6.82208836e-01 -1.72343969e+00 1.89358182e-02
-3.23878139e-01 -8.66039917e-02 8.27407777e-01 1.06709063e+00
6.05529882e-02 7.49766588e-01 4.34769504e-02 -9.24705386e-01
-1.82900921e-01 -6.80938065e-01 -4.24062431e-01 5.52687049e-01
2.25929156e-01 -6.68018341e-01 6.81879595e-02 2.64303684e-01] | [14.603623390197754, -2.3876285552978516] |
1c6ef58e-0a1e-41ce-8b75-11fd2a7b037c | personalized-federated-learning-under-mixture | 2305.01068 | null | https://arxiv.org/abs/2305.01068v1 | https://arxiv.org/pdf/2305.01068v1.pdf | Personalized Federated Learning under Mixture of Distributions | The recent trend towards Personalized Federated Learning (PFL) has garnered significant attention as it allows for the training of models that are tailored to each client while maintaining data privacy. However, current PFL techniques primarily focus on modeling the conditional distribution heterogeneity (i.e. concept shift), which can result in suboptimal performance when the distribution of input data across clients diverges (i.e. covariate shift). Additionally, these techniques often lack the ability to adapt to unseen data, further limiting their effectiveness in real-world scenarios. To address these limitations, we propose a novel approach, FedGMM, which utilizes Gaussian mixture models (GMM) to effectively fit the input data distributions across diverse clients. The model parameters are estimated by maximum likelihood estimation utilizing a federated Expectation-Maximization algorithm, which is solved in closed form and does not assume gradient similarity. Furthermore, FedGMM possesses an additional advantage of adapting to new clients with minimal overhead, and it also enables uncertainty quantification. Empirical evaluations on synthetic and benchmark datasets demonstrate the superior performance of our method in both PFL classification and novel sample detection. | ['Wei Cheng', 'Haifeng Chen', 'Dawei Zhou', 'Quanquan Gu', 'Yanchi Liu', 'Wenchao Yu', 'Shuaicheng Zhang', 'Yue Wu'] | 2023-05-01 | null | null | null | null | ['personalized-federated-learning'] | ['methodology'] | [-2.86943078e-01 -3.26996803e-01 -3.48201036e-01 -6.41295433e-01
-7.81182230e-01 -6.54067397e-01 4.10943896e-01 2.14727465e-02
-2.50916362e-01 6.48532450e-01 -3.68076377e-03 -3.19344550e-01
-7.37449676e-02 -5.34417331e-01 -6.88966155e-01 -8.12749267e-01
-1.41483620e-01 3.60477418e-01 -2.03261096e-02 5.83447158e-01
6.86958358e-02 4.59557325e-01 -1.33942604e+00 1.13975674e-01
1.04470098e+00 1.10340965e+00 -7.86566064e-02 2.37053514e-01
-4.03206557e-01 6.12210929e-01 -5.69353163e-01 -8.05601776e-01
5.00037611e-01 -1.09401636e-01 -4.11640257e-01 1.13039471e-01
1.75849527e-01 -7.02284038e-01 -4.33628201e-01 1.06781590e+00
3.81979287e-01 2.14306816e-01 5.34926891e-01 -1.59150398e+00
-5.62093675e-01 4.63135630e-01 -5.85367560e-01 -3.90588082e-02
3.08810491e-02 1.20235253e-02 7.87360609e-01 -6.64393425e-01
2.40887880e-01 1.23029411e+00 6.61671937e-01 5.32666564e-01
-1.47491395e+00 -7.71824777e-01 4.71840411e-01 1.06042154e-01
-1.43852830e+00 -3.97698790e-01 5.40074348e-01 -2.21787527e-01
4.13188368e-01 3.13699394e-01 6.68470282e-03 9.92831230e-01
-3.46657969e-02 1.09570563e+00 8.37511659e-01 3.54304574e-02
6.69897735e-01 6.82153761e-01 -9.84869152e-02 1.02537408e-01
3.75351101e-01 -5.89564666e-02 -5.38010120e-01 -8.71901870e-01
4.46333736e-01 5.36261320e-01 -3.94297183e-01 -9.27074850e-01
-5.29864132e-01 8.51047099e-01 1.25277981e-01 -2.08284527e-01
-4.72525686e-01 -1.86849028e-01 4.26115364e-01 1.30029887e-01
2.70122021e-01 -3.05128604e-01 -5.16730249e-01 -2.14770921e-02
-8.86000037e-01 3.29894364e-01 9.68884647e-01 1.30879760e+00
9.63358223e-01 -3.18330288e-01 -1.85622007e-01 8.44374597e-01
4.66183454e-01 4.11433816e-01 6.66304529e-01 -1.04071903e+00
5.48246741e-01 5.94572246e-01 2.81447232e-01 -7.37282455e-01
9.73013267e-02 -4.96364325e-01 -6.55240595e-01 2.37818900e-03
4.34971005e-01 -2.72143185e-01 -4.83156681e-01 1.94085884e+00
7.08893478e-01 2.14006156e-01 1.59233391e-01 5.83787084e-01
2.81117912e-02 3.36767495e-01 1.22008480e-01 -2.97736734e-01
8.86774600e-01 -5.42408645e-01 -6.08390450e-01 1.49770245e-01
4.53406364e-01 -3.93247932e-01 9.13486660e-01 4.93046999e-01
-7.18039036e-01 2.19842538e-01 -7.16677308e-01 5.09121656e-01
-1.52540758e-01 -5.35643399e-01 6.82838857e-01 9.74972844e-01
-7.03093588e-01 4.66125995e-01 -8.53865564e-01 -3.79128635e-01
8.43735516e-01 3.90791178e-01 -3.07932585e-01 -3.25909317e-01
-7.90306926e-01 1.75511345e-01 2.61717588e-01 -4.08312321e-01
-7.81234741e-01 -8.77497256e-01 -5.50272167e-01 4.04634416e-01
5.93883157e-01 -5.91927707e-01 1.55705595e+00 -6.96605980e-01
-1.46697831e+00 3.06019485e-01 -1.09164670e-01 -5.23102880e-01
1.00149190e+00 -1.44822262e-02 -4.30319071e-01 -1.39035404e-01
-2.08378911e-01 4.56850268e-02 8.87430370e-01 -1.38522375e+00
-1.08539581e+00 -5.97258985e-01 -2.57578343e-01 -3.13541517e-02
-8.59401822e-01 -7.14213103e-02 -6.63894057e-01 -5.14488041e-01
1.01731770e-01 -5.40168405e-01 -1.87652513e-01 1.85043573e-01
-1.38078704e-01 -1.18648224e-01 1.17194474e+00 -3.17609638e-01
1.43677998e+00 -2.24615169e+00 -5.76412618e-01 3.82137686e-01
9.60939005e-02 1.90513819e-01 1.96210340e-01 5.89970410e-01
5.19687474e-01 4.97956797e-02 -2.25194857e-01 -6.18351698e-01
2.57140905e-01 3.41787428e-01 -3.32006007e-01 5.32722890e-01
-3.37521791e-01 4.59125489e-01 -6.70139730e-01 -5.99423230e-01
5.05708940e-02 3.99918437e-01 -9.01174664e-01 4.46332842e-01
-3.29089850e-01 4.38483685e-01 -6.19632602e-01 7.30309725e-01
1.21161246e+00 -4.95208979e-01 5.01159072e-01 3.94803733e-02
1.39442548e-01 -4.97629307e-02 -1.48936510e+00 1.22604251e+00
-4.90615159e-01 -1.03707902e-01 3.93402427e-01 -8.10106397e-01
8.34700882e-01 3.37487221e-01 7.03314424e-01 -4.02960390e-01
6.76397830e-02 1.22928657e-01 -4.94813114e-01 -2.34613582e-01
1.81724846e-01 2.98191644e-02 8.84145051e-02 7.62981951e-01
-1.06330454e-01 6.05195165e-01 -2.28587732e-01 2.34502137e-01
9.76324081e-01 -1.82006747e-01 1.09939694e-01 -7.13962072e-05
2.96981961e-01 -5.62624216e-01 1.03451014e+00 9.43627477e-01
-4.62732315e-01 4.28220600e-01 2.16306642e-01 -1.50102451e-01
-5.57455778e-01 -1.12765706e+00 -2.66942322e-01 1.00343537e+00
2.56778926e-01 -2.36748502e-01 -8.20572197e-01 -9.70204711e-01
7.20436156e-01 9.29947913e-01 -3.32128793e-01 -1.52682215e-01
-2.21125513e-01 -7.75803268e-01 2.79770464e-01 2.65352219e-01
4.92482007e-01 -4.24615949e-01 -5.81283510e-01 4.57639694e-01
-6.18616976e-02 -9.32122469e-01 -6.92192852e-01 5.37256477e-04
-9.26187515e-01 -1.03826404e+00 -5.01860261e-01 -1.83872893e-01
5.68566263e-01 4.82514769e-01 6.36786222e-01 -4.15014356e-01
-2.27440774e-01 7.49515772e-01 -1.29095525e-01 -2.37316281e-01
-2.88801938e-01 -4.19266075e-02 1.95445001e-01 6.52571619e-01
6.53062880e-01 -7.47496605e-01 -7.64530897e-01 5.95116556e-01
-1.17021942e+00 -6.07212782e-01 3.49204987e-01 8.49328816e-01
4.96836036e-01 1.78919688e-01 8.48609746e-01 -1.12664545e+00
6.75792634e-01 -9.76447225e-01 -6.03074312e-01 5.53276598e-01
-1.23233163e+00 -1.07517056e-01 7.64969707e-01 -6.26498222e-01
-1.55052531e+00 -7.31099471e-02 3.44035953e-01 -5.77430606e-01
-3.90716642e-02 1.39596045e-01 -6.71639502e-01 4.30308878e-02
4.34686601e-01 2.69287407e-01 2.63238251e-01 -1.00854373e+00
4.16393459e-01 1.23159885e+00 6.07854605e-01 -7.92042315e-01
5.17321110e-01 6.46629572e-01 -3.08626145e-01 -2.66856432e-01
-3.36499572e-01 -4.23570842e-01 -1.38824716e-01 -1.00616356e-02
-6.87186942e-02 -7.85889328e-01 -1.09721899e+00 5.17084658e-01
-5.23296475e-01 6.04713522e-02 -2.30390534e-01 5.30371785e-01
-4.40437168e-01 8.47258449e-01 -4.58218455e-01 -1.24410093e+00
-5.37843466e-01 -9.04845655e-01 3.78926426e-01 4.64606822e-01
-1.78704094e-02 -9.98653650e-01 -2.45153934e-01 3.35506290e-01
6.97986066e-01 -8.14127456e-03 9.07861948e-01 -1.06902516e+00
-6.85196698e-01 -5.67781568e-01 -1.19543202e-01 4.49885041e-01
3.74286264e-01 -2.64189810e-01 -1.08562994e+00 -7.84223616e-01
1.70166627e-01 -1.85570523e-01 2.84545839e-01 5.15598506e-02
1.44930661e+00 -7.34988451e-01 -4.59484756e-01 7.76394665e-01
1.37640107e+00 1.46480545e-01 1.54717341e-01 2.43876055e-01
2.23637909e-01 5.25289476e-01 6.76368475e-01 1.25166988e+00
4.92680103e-01 4.99598622e-01 6.55967593e-01 4.70521480e-01
4.40023333e-01 -5.38493633e-01 1.60073519e-01 7.00460002e-02
8.23665738e-01 -3.74954402e-01 -6.13389611e-01 5.08089364e-01
-2.02121282e+00 -6.86357915e-01 2.63268858e-01 2.74155545e+00
6.78733706e-01 -3.04472119e-01 3.02863687e-01 -2.28874817e-01
9.13676023e-01 -2.98303545e-01 -1.35882175e+00 -8.79214928e-02
3.08320019e-02 -2.57662743e-01 8.30046535e-01 -8.38327631e-02
-6.94311857e-01 4.70581234e-01 5.88628197e+00 7.37610757e-01
-9.73409474e-01 2.04241544e-01 6.54309034e-01 -2.44398415e-01
-4.88084704e-01 6.55732900e-02 -7.66053617e-01 8.26606393e-01
7.67440975e-01 -6.02889419e-01 5.68924785e-01 1.19458067e+00
-1.73909217e-02 1.97696853e-02 -1.27856791e+00 1.16720104e+00
-1.95962027e-01 -8.79180133e-01 -4.53794487e-02 2.68149704e-01
6.68402851e-01 2.15286892e-02 2.82182753e-01 3.33604336e-01
5.39528728e-01 -4.68794614e-01 3.85966539e-01 4.80248481e-01
5.76853931e-01 -9.72498596e-01 4.26279545e-01 6.95521474e-01
-7.23667979e-01 -5.29882669e-01 -3.80055130e-01 4.70118225e-01
-4.12313677e-02 4.92447495e-01 -9.05667126e-01 5.43236554e-01
8.96650493e-01 2.63981149e-02 -1.16775230e-01 1.35074687e+00
4.55247521e-01 6.57394230e-01 -6.16885900e-01 2.49238208e-01
-2.29153305e-01 -1.37721077e-01 4.95939046e-01 1.09128678e+00
4.28618729e-01 -3.19383025e-01 1.52740613e-01 7.54808366e-01
-2.46688291e-01 3.77532959e-01 -6.53967038e-02 6.37447909e-02
9.77824867e-01 1.08484662e+00 -1.05148487e-01 -9.10483897e-02
-6.37121439e-01 9.60097253e-01 3.88416797e-01 5.41911125e-01
-5.24758160e-01 -1.39946982e-01 1.13806713e+00 7.89361671e-02
4.18824494e-01 2.35093057e-01 -6.92321435e-02 -1.26764297e+00
2.82984167e-01 -9.29895461e-01 9.49069560e-01 2.92076170e-01
-1.90075231e+00 2.79389262e-01 -1.38483092e-01 -1.14452994e+00
-3.95973742e-01 -1.73880368e-01 -5.65584660e-01 9.00056422e-01
-1.41550112e+00 -9.11877692e-01 -1.22801110e-01 9.43902314e-01
-6.11149296e-02 -1.60557806e-01 6.99029088e-01 5.77598870e-01
-7.59099483e-01 1.46568894e+00 8.15646052e-01 -3.94501656e-01
1.02614725e+00 -9.45086896e-01 7.67800435e-02 5.93950391e-01
-4.17590648e-01 9.20655370e-01 5.76514184e-01 -4.81231749e-01
-1.56741917e+00 -1.39536929e+00 3.65473449e-01 -7.79229403e-02
2.94955939e-01 -4.07546520e-01 -1.29070437e+00 7.23268509e-01
-4.10762399e-01 2.60487854e-01 1.20498836e+00 1.45508964e-02
-7.01730371e-01 -5.07957697e-01 -1.95396566e+00 3.86673927e-01
7.47190058e-01 -3.64482790e-01 9.49549824e-02 1.77062079e-01
5.41967094e-01 -1.84418201e-01 -9.78621244e-01 2.01208368e-01
5.43862641e-01 -1.11046863e+00 6.64640784e-01 -8.14443409e-01
-5.89619339e-01 -1.71036482e-01 -5.23739815e-01 -9.01040435e-01
-2.41039023e-01 -1.08071053e+00 -7.57873654e-01 1.89499462e+00
1.14635430e-01 -1.21845961e+00 9.67687309e-01 1.52384090e+00
3.76549751e-01 -7.22968519e-01 -1.01604426e+00 -1.00300467e+00
-1.39747530e-01 -4.12053406e-01 1.36626542e+00 8.57554495e-01
9.83520672e-02 -4.51942474e-01 -5.15824795e-01 4.12349582e-01
1.04403698e+00 2.08856314e-01 9.94043529e-01 -1.11908329e+00
-6.36127174e-01 -1.79350749e-01 -2.73687631e-01 -1.21739995e+00
1.05237253e-01 -7.77740777e-01 -2.62469411e-01 -1.03437996e+00
4.82415408e-01 -6.37989461e-01 -6.92768037e-01 3.56337875e-01
-2.51676023e-01 -4.40180153e-01 1.25941083e-01 4.54785615e-01
-6.98503554e-01 7.62248039e-01 5.86512387e-01 2.53191501e-01
-2.67373651e-01 7.41711259e-01 -1.04186571e+00 3.41799259e-01
7.85176098e-01 -4.25725996e-01 -4.78394389e-01 -1.36063039e-01
-4.12513882e-01 1.94470268e-02 1.55069217e-01 -7.06781924e-01
3.30787152e-01 -4.03333038e-01 2.68237919e-01 -4.78086531e-01
-3.81212458e-02 -1.20074344e+00 4.18035388e-01 1.32086620e-01
-1.14260137e-01 -3.81143063e-01 1.16028134e-02 9.97080445e-01
-3.49269435e-03 -8.98389444e-02 8.08830619e-01 1.38804302e-01
-2.04073846e-01 9.09666419e-01 -1.53352454e-01 -7.23906383e-02
1.14991307e+00 -9.34583098e-02 -9.62983221e-02 -7.26662397e-01
-3.06112856e-01 7.21963644e-01 8.15746486e-01 3.90949190e-01
2.57955343e-01 -1.25854146e+00 -3.96913737e-01 1.82288334e-01
1.83984771e-01 5.46687506e-02 5.46262383e-01 4.30381119e-01
6.01708181e-02 2.86395848e-01 3.68835181e-01 -5.16914845e-01
-1.06603110e+00 9.60112929e-01 3.61448199e-01 2.91240849e-02
-6.07883036e-01 5.53980589e-01 3.55241865e-01 -5.60518146e-01
7.67752290e-01 1.81625739e-01 3.65266323e-01 -2.90070832e-01
7.61143148e-01 6.26241922e-01 9.55583155e-02 -2.73430794e-01
-3.13074768e-01 -1.75305381e-01 -5.23488045e-01 7.27843568e-02
1.25908351e+00 -5.33107102e-01 1.14781640e-01 1.46735147e-01
1.36869454e+00 -5.74687272e-02 -1.80108809e+00 -9.27855372e-01
1.18991032e-01 -1.07327950e+00 -8.45191702e-02 -7.21297622e-01
-1.05583394e+00 5.70423603e-01 7.59855151e-01 9.94009078e-02
1.14903414e+00 -2.03176215e-01 8.99165630e-01 2.32828528e-01
6.84356868e-01 -1.01455700e+00 -2.86593854e-01 -5.36115579e-02
3.22510093e-01 -1.15549791e+00 -2.23726466e-01 -8.09815824e-02
-5.91804087e-01 8.37375224e-01 6.69650316e-01 5.34457266e-01
6.56724930e-01 1.56591341e-01 1.37060672e-01 3.76663446e-01
-7.85983801e-01 4.43475485e-01 -2.29105785e-01 6.74110293e-01
-4.46945168e-02 9.49385762e-02 -5.20833656e-02 9.99999344e-01
2.44522780e-01 8.46363083e-02 2.22701952e-01 1.15076017e+00
-2.16919616e-01 -1.42192948e+00 -4.68719274e-01 6.05258524e-01
-7.29133308e-01 4.15843517e-01 -8.41224194e-02 2.85015762e-01
-1.67256922e-01 1.13271916e+00 -5.74868880e-02 -1.55155778e-01
1.72953099e-01 3.25897306e-01 6.13256022e-02 -3.30503017e-01
-3.86355221e-01 9.87872332e-02 -6.03463948e-01 -5.12242138e-01
3.30030203e-01 -1.02632630e+00 -1.22510755e+00 -4.98454213e-01
-4.04173017e-01 4.38909352e-01 8.58075619e-01 5.37239015e-01
9.28344786e-01 -5.84607683e-02 1.25150537e+00 -2.28718579e-01
-1.55043066e+00 -6.57990694e-01 -9.60645735e-01 4.81045276e-01
1.79497316e-01 -5.26243448e-01 -5.22748649e-01 -3.10220957e-01] | [5.808900833129883, 6.320919513702393] |
7d318165-712d-4064-9937-66c13070dbbb | show-me-your-face-and-i-ll-tell-you-how-you | 2206.14009 | null | https://arxiv.org/abs/2206.14009v1 | https://arxiv.org/pdf/2206.14009v1.pdf | Show Me Your Face, And I'll Tell You How You Speak | When we speak, the prosody and content of the speech can be inferred from the movement of our lips. In this work, we explore the task of lip to speech synthesis, i.e., learning to generate speech given only the lip movements of a speaker where we focus on learning accurate lip to speech mappings for multiple speakers in unconstrained, large vocabulary settings. We capture the speaker's voice identity through their facial characteristics, i.e., age, gender, ethnicity and condition them along with the lip movements to generate speaker identity aware speech. To this end, we present a novel method "Lip2Speech", with key design choices to achieve accurate lip to speech synthesis in unconstrained scenarios. We also perform various experiments and extensive evaluation using quantitative, qualitative metrics and human evaluation. | ['Timon Ulrich', 'Lotfy Abdel Khaliq', 'Christen Millerdurai'] | 2022-06-28 | null | null | null | null | ['lip-to-speech-synthesis'] | ['computer-vision'] | [ 7.79844895e-02 3.88887644e-01 -6.39167786e-01 -4.78641719e-01
-9.78789687e-01 -5.58392704e-01 7.13478625e-01 -6.08759284e-01
5.29952012e-02 7.45471358e-01 8.34675193e-01 3.63484509e-02
5.80534756e-01 -4.35278118e-02 -4.78816360e-01 -6.07194066e-01
5.61235428e-01 2.98777997e-01 -4.74637806e-01 7.38187656e-02
1.03073366e-01 4.55366343e-01 -1.89337647e+00 2.71400869e-01
4.18378711e-01 8.78370941e-01 3.08995284e-02 9.29231048e-01
-2.92380750e-01 4.68020320e-01 -6.45538807e-01 -4.69593316e-01
1.44210905e-01 -4.61300313e-01 -6.59237862e-01 1.15132898e-01
6.12963200e-01 -4.48263705e-01 -5.42214550e-02 1.00994897e+00
1.01344371e+00 -8.38490427e-02 7.78399408e-01 -1.49159861e+00
-5.08788109e-01 9.42544043e-01 -3.11902255e-01 -3.52564424e-01
7.05868900e-01 3.35674584e-01 8.49804044e-01 -1.18514121e+00
6.56004071e-01 1.66376209e+00 5.28374910e-01 1.41507947e+00
-1.13877273e+00 -1.00640392e+00 6.07985854e-02 -1.54658198e-01
-1.73011768e+00 -1.72797656e+00 9.35446799e-01 -4.40752387e-01
2.20545486e-01 3.47579718e-01 3.32544744e-01 1.50293446e+00
-5.32936692e-01 8.42345238e-01 7.57469833e-01 -6.48684502e-01
-2.38895733e-02 4.15628195e-01 -5.24695277e-01 3.99334252e-01
-3.62428784e-01 1.16373762e-01 -8.82747948e-01 -1.55935409e-02
5.06421208e-01 -7.21986771e-01 -1.89056143e-01 -1.34493127e-01
-1.39987767e+00 6.94667339e-01 -2.42760792e-01 -4.99550602e-04
-2.74025023e-01 2.66753584e-01 2.39489704e-01 -3.93805979e-03
4.25019354e-01 -2.51847804e-02 -1.70558766e-01 -9.39960331e-02
-9.85318005e-01 3.62698466e-01 8.27559292e-01 1.32474816e+00
5.61855316e-01 3.39657933e-01 -2.00647444e-01 1.27694643e+00
8.29105437e-01 8.72337580e-01 6.85037076e-01 -1.10349131e+00
4.04211491e-01 -1.48682475e-01 9.23815072e-02 -3.95989418e-01
-7.83915538e-03 4.41768676e-01 -5.93701243e-01 1.46948472e-01
3.38436157e-01 -6.03284776e-01 -9.48771119e-01 2.31897855e+00
4.29690540e-01 2.96848595e-01 4.31912810e-01 5.02510309e-01
1.08058786e+00 3.89270842e-01 2.24767223e-01 -5.53877175e-01
1.29976749e+00 -8.24112535e-01 -1.12722647e+00 -2.82866091e-01
1.83444291e-01 -1.06189382e+00 1.19843435e+00 -1.78017959e-01
-1.39016211e+00 -5.93803287e-01 -6.53044283e-01 2.94244401e-02
-1.43115222e-01 5.71471035e-01 2.93176413e-01 9.21162844e-01
-1.32331932e+00 -1.29222451e-02 -3.60803157e-01 -3.53650004e-01
3.52922916e-01 5.32774806e-01 -3.50203484e-01 4.14744079e-01
-1.06667233e+00 5.05644739e-01 -8.36057290e-02 -3.65814507e-01
-8.42966855e-01 -7.23994970e-01 -1.03253639e+00 -2.40427598e-01
8.29570591e-02 -6.37003958e-01 1.63699305e+00 -1.05459428e+00
-2.11844158e+00 1.17230129e+00 -7.86884665e-01 -1.78384811e-01
5.86455941e-01 3.20011973e-01 -4.54491228e-01 -5.13017774e-02
2.55534425e-02 1.26950574e+00 1.16749823e+00 -1.57833672e+00
-6.80174291e-01 -1.90871105e-01 -5.13863385e-01 3.79225701e-01
-1.41900182e-01 5.39554775e-01 -4.01265502e-01 -6.35671556e-01
-2.90829331e-01 -1.00888813e+00 3.20712715e-01 3.49999368e-01
-5.80483675e-01 -3.33461583e-01 7.63255119e-01 -8.24042022e-01
9.93348897e-01 -2.39229536e+00 -1.58626929e-01 -6.59679025e-02
2.86196601e-02 1.63009480e-01 -1.14601135e-01 1.60094485e-01
8.62758905e-02 4.39609468e-01 1.71990275e-01 -9.99132514e-01
1.75355881e-01 -1.36073977e-01 -4.73261446e-01 3.10615093e-01
-3.84666622e-02 9.75726962e-01 -5.61836243e-01 -9.62648928e-01
1.02643125e-01 6.85119987e-01 -5.32433987e-01 3.86809170e-01
-1.46984696e-01 3.79685581e-01 -1.71788305e-01 9.17724013e-01
5.46010196e-01 3.79149765e-01 -5.51574193e-02 -2.05780461e-01
-1.32452011e-01 3.88606280e-01 -1.09084952e+00 1.59207833e+00
-7.59837389e-01 8.68687153e-01 8.39966416e-01 -2.51607418e-01
1.09003818e+00 9.51766312e-01 4.01132435e-01 -1.45839229e-01
1.93895712e-01 3.00337881e-01 -1.18078411e-01 -5.16615689e-01
2.12378219e-01 -3.27582330e-01 9.34205428e-02 4.46089178e-01
-2.73664575e-02 -3.21214825e-01 -1.70333818e-01 -3.99903089e-01
6.65190890e-02 -1.37431189e-01 3.46794844e-01 -3.56068462e-01
7.50148535e-01 -6.85685337e-01 2.30344862e-01 2.45209947e-01
-4.53343272e-01 6.99536979e-01 4.90353778e-02 1.42619118e-01
-1.04663634e+00 -1.28485179e+00 1.85895222e-03 1.20759344e+00
-6.70791417e-02 -1.51669264e-01 -1.22012079e+00 -2.85106987e-01
-5.32165505e-02 6.78511739e-01 -4.36074346e-01 4.23718104e-03
-6.30338609e-01 -9.12540779e-02 9.67799008e-01 3.36781979e-01
1.33198470e-01 -1.26270497e+00 2.23626822e-01 -1.12189807e-01
-4.39465642e-01 -1.32808185e+00 -1.35979581e+00 -7.02408612e-01
-5.38310111e-02 -5.39692700e-01 -9.53686535e-01 -1.29730737e+00
3.82762045e-01 -3.33265930e-01 7.03466356e-01 -2.65046239e-01
-1.57081231e-01 3.25273275e-01 1.72696963e-01 -8.39081645e-01
-1.03917778e+00 1.39016643e-01 7.95929909e-01 5.52089036e-01
4.07849103e-02 -5.31209886e-01 -4.22571540e-01 4.84993726e-01
-3.46334606e-01 2.85057575e-01 3.16666871e-01 4.36321229e-01
4.34701800e-01 -3.68321747e-01 1.16159916e+00 -2.80756086e-01
9.15541947e-01 -2.38008812e-01 -2.24967122e-01 3.84553134e-01
-2.86197245e-01 -2.30473340e-01 4.06918645e-01 -1.03367603e+00
-1.18690741e+00 2.32247442e-01 -4.29536998e-01 -5.44194520e-01
-4.68736351e-01 -1.78276286e-01 -7.62700260e-01 3.80169712e-02
4.84424859e-01 3.83475929e-01 5.56043983e-01 -4.92674738e-01
6.19046986e-01 1.59689045e+00 9.31476176e-01 -6.59192622e-01
5.72654009e-01 4.07168567e-01 -4.04954433e-01 -1.23995364e+00
-2.90657908e-01 -3.35234165e-01 -6.29935205e-01 -2.62690037e-01
7.34378815e-01 -1.08501589e+00 -1.17445576e+00 7.49120593e-01
-1.19305480e+00 -5.56811213e-01 -3.36034983e-01 5.02173305e-01
-1.11899722e+00 -3.33461707e-04 -1.88286275e-01 -1.29706514e+00
-5.77660501e-01 -1.21686077e+00 1.45067084e+00 3.00726235e-01
-5.93985200e-01 -1.00759494e+00 -7.58161098e-02 5.53326964e-01
3.91268671e-01 -1.53483972e-01 5.69552124e-01 -6.03931904e-01
-3.16613048e-01 2.58497000e-01 -1.05372600e-01 2.90400714e-01
7.98565507e-01 2.00099632e-01 -1.18633890e+00 -9.75526050e-02
-4.42510962e-01 -2.87878573e-01 2.35754713e-01 5.91301978e-01
7.95915425e-01 -9.00607109e-01 -4.48651701e-01 7.88094044e-01
8.67237329e-01 1.34421855e-01 3.00191641e-01 -6.24594629e-01
6.32684469e-01 1.03887320e+00 3.22532386e-01 3.74949336e-01
5.02503395e-01 9.21182096e-01 -6.81289956e-02 4.20798771e-02
-1.09122312e+00 -8.15541327e-01 7.11807489e-01 7.12695599e-01
1.65705502e-01 -3.43930691e-01 -8.32165658e-01 8.16649795e-01
-1.24564254e+00 -1.02562141e+00 5.37479937e-01 2.06039190e+00
1.37807369e+00 -3.28794837e-01 4.69801515e-01 -5.02117947e-02
1.21701026e+00 1.72026843e-01 -5.92423737e-01 -3.07678878e-01
-1.38763875e-01 4.08482924e-02 2.39101857e-01 1.01209545e+00
-7.98445165e-01 1.40010130e+00 7.35449886e+00 9.01784182e-01
-1.50280607e+00 -1.23892032e-01 5.67436814e-01 -1.97084755e-01
-5.22701740e-01 -4.72066700e-01 -1.19974828e+00 4.06655461e-01
9.71937120e-01 -6.72088265e-01 9.17842865e-01 4.74608392e-01
5.93277335e-01 4.71750528e-01 -1.24122226e+00 1.24433410e+00
3.42883945e-01 -1.33015597e+00 1.95960850e-01 -2.02810876e-02
6.17216051e-01 -1.01900376e-01 3.87534350e-01 1.06456326e-02
1.98771536e-01 -1.34041226e+00 1.11757600e+00 4.60866600e-01
1.68007386e+00 -3.51748526e-01 5.16164489e-02 3.81877154e-01
-1.25528920e+00 -8.12334567e-02 3.42157036e-01 4.90312666e-01
4.36736166e-01 -3.80590886e-01 -1.36469769e+00 -8.26514736e-02
2.46394634e-01 3.84760112e-01 -4.50839549e-02 5.21676004e-01
-1.16264246e-01 5.38508058e-01 -3.72062773e-01 -1.24678075e-01
-3.98572236e-01 3.37177366e-01 6.41163826e-01 1.05555308e+00
5.63152254e-01 -7.45627359e-02 -6.87585250e-02 1.18820775e+00
-3.72831553e-01 5.86375952e-01 -5.85488856e-01 -1.33456632e-01
8.20296645e-01 8.92626464e-01 3.44939679e-02 -1.19019397e-01
-2.21310347e-01 5.31897783e-01 -2.05150247e-01 4.73850429e-01
-3.16345394e-01 -1.15875624e-01 1.40298903e+00 4.63985980e-01
-8.76255259e-02 1.25628874e-01 -2.53949136e-01 -8.79827678e-01
-1.90054372e-01 -1.08066308e+00 -2.28512123e-01 -8.80288422e-01
-1.05543351e+00 5.90076804e-01 2.80984286e-02 -6.58351421e-01
-9.25289154e-01 -3.82408082e-01 -7.73675025e-01 1.27209401e+00
-1.42030168e+00 -1.71235907e+00 -4.16225605e-02 7.78894246e-01
9.77765977e-01 -5.54607332e-01 8.51886928e-01 1.19558349e-01
-3.82966638e-01 1.26692319e+00 -3.80213022e-01 3.44301075e-01
9.91750956e-01 -7.81815529e-01 6.06930375e-01 2.82972723e-01
7.24694282e-02 5.38539052e-01 7.79619873e-01 -5.60069382e-01
-1.03654885e+00 -9.21473622e-01 1.28244662e+00 -2.37107217e-01
4.11825418e-01 -7.70282984e-01 -4.23863679e-01 6.72980547e-01
2.14062870e-01 -2.35187970e-02 7.77801394e-01 -3.00768882e-01
-1.52427465e-01 -1.22918807e-01 -1.40370262e+00 1.04521203e+00
1.16195774e+00 -7.75432765e-01 -1.59652844e-01 3.75435203e-02
8.13454926e-01 -3.16752851e-01 -7.91602373e-01 3.73157293e-01
8.76299858e-01 -4.59995806e-01 1.17308366e+00 -5.42856514e-01
-1.78149655e-01 -1.12354182e-01 -3.08711022e-01 -1.12360382e+00
3.46301019e-01 -1.28037536e+00 5.84903397e-02 1.94278347e+00
6.05593204e-01 -5.60308158e-01 9.84713733e-01 6.01358712e-01
9.22482088e-02 -5.83775461e-01 -1.02172232e+00 -4.62945968e-01
3.29285026e-01 -2.67157644e-01 1.19961476e+00 6.36929572e-01
-1.71323679e-02 1.83803514e-01 -6.08375132e-01 1.51436672e-01
6.31158173e-01 -9.53035057e-02 1.01866484e+00 -1.10000443e+00
2.56308645e-01 -6.74825847e-01 3.20272259e-02 -1.00425351e+00
8.86403739e-01 -7.70155430e-01 3.78456622e-01 -1.04440236e+00
-6.59412220e-02 -4.84652311e-01 2.23004058e-01 3.56686324e-01
-7.78991878e-02 8.21693614e-02 1.38402686e-01 1.30870089e-01
1.26721025e-01 5.06092846e-01 1.23358011e+00 -2.34918773e-01
-3.94412071e-01 5.70554078e-01 -7.47097313e-01 7.40043998e-01
1.05760944e+00 -1.81066304e-01 -6.00978971e-01 -1.60198808e-01
-4.32805926e-01 3.03133100e-01 9.25070420e-02 -5.96018493e-01
4.36958939e-01 -5.55708051e-01 -1.07014887e-01 -2.12327003e-01
7.53264010e-01 -3.87637824e-01 -1.00934759e-01 6.88105077e-02
-8.86270821e-01 -2.43661776e-01 1.14227459e-01 7.39925876e-02
-1.39439538e-01 -1.62502810e-01 8.89765978e-01 1.79760978e-01
-2.71208614e-01 6.47108555e-01 -3.07723522e-01 2.89702266e-01
8.80666196e-01 -2.96193480e-01 2.58210842e-02 -9.54853535e-01
-8.19391668e-01 -5.54559231e-02 5.88039875e-01 7.16643453e-01
5.90724468e-01 -1.62910151e+00 -1.06433535e+00 5.98062992e-01
1.28060848e-01 -2.93355882e-01 -3.86958905e-02 4.30262417e-01
-1.09662928e-01 4.11486804e-01 -1.24589957e-01 -4.30885166e-01
-1.91113627e+00 4.64445323e-01 5.80540955e-01 5.11019945e-01
-3.43196504e-02 8.10017109e-01 2.30579868e-01 -5.21748662e-01
5.96667528e-01 1.48349479e-02 -2.99067050e-01 5.57846688e-02
5.07433951e-01 1.11843728e-01 -5.53868115e-01 -1.36261630e+00
-3.22158009e-01 7.48758137e-01 3.42878222e-01 -6.25206053e-01
6.23293042e-01 -5.71467698e-01 2.52050668e-01 4.14290428e-01
1.27627683e+00 8.36804152e-01 -1.20615494e+00 -1.58109412e-01
-3.73089582e-01 -4.01810586e-01 -2.60319501e-01 -6.48276448e-01
-9.75597203e-01 8.81733000e-01 4.04463261e-01 -1.42094672e-01
8.13296676e-01 3.71726990e-01 8.03089380e-01 -1.21970795e-01
4.76168431e-02 -1.15420914e+00 -1.24974146e-01 8.12864602e-02
1.01509631e+00 -1.27876627e+00 -5.52378356e-01 -6.03030801e-01
-7.97262609e-01 8.15682471e-01 4.05044168e-01 6.50143921e-01
7.42608666e-01 4.80967164e-01 7.27625012e-01 5.82750678e-01
-7.83050537e-01 -3.44616354e-01 4.47334796e-01 1.08475339e+00
4.38280195e-01 2.55967408e-01 2.25670904e-01 6.60392165e-01
-9.00862217e-01 -9.49651971e-02 1.14698932e-01 1.11952797e-01
-3.37014109e-01 -1.30358446e+00 -4.07259911e-01 -3.83933559e-02
-5.48093259e-01 -1.05508201e-01 -6.93525255e-01 4.21299666e-01
2.74578452e-01 1.07251382e+00 3.15601029e-03 -2.83418834e-01
1.74350947e-01 4.87928540e-01 3.22345704e-01 -6.38567030e-01
-1.93279460e-01 -5.89410104e-02 2.88081318e-01 -1.25029087e-01
-1.72154665e-01 -1.08714938e+00 -1.32462299e+00 -3.84654939e-01
-1.19286910e-01 -1.27624214e-01 1.14886665e+00 6.18030727e-01
2.90409833e-01 -4.44080532e-02 8.49073827e-01 -8.28505516e-01
-7.87556827e-01 -9.76198971e-01 -4.45964456e-01 4.61892456e-01
8.43855023e-01 -4.31695461e-01 -6.84190750e-01 4.01746273e-01] | [14.325013160705566, 4.9457688331604] |
144b69e1-2e1a-4e9e-ac5b-0dfe5d9562cb | multi-camera-multiple-3d-object-tracking-on | 2204.09151 | null | https://arxiv.org/abs/2204.09151v1 | https://arxiv.org/pdf/2204.09151v1.pdf | Multi-Camera Multiple 3D Object Tracking on the Move for Autonomous Vehicles | The development of autonomous vehicles provides an opportunity to have a complete set of camera sensors capturing the environment around the car. Thus, it is important for object detection and tracking to address new challenges, such as achieving consistent results across views of cameras. To address these challenges, this work presents a new Global Association Graph Model with Link Prediction approach to predict existing tracklets location and link detections with tracklets via cross-attention motion modeling and appearance re-identification. This approach aims at solving issues caused by inconsistent 3D object detection. Moreover, our model exploits to improve the detection accuracy of a standard 3D object detector in the nuScenes detection challenge. The experimental results on the nuScenes dataset demonstrate the benefits of the proposed method to produce SOTA performance on the existing vision-based tracking dataset. | ['Khoa Luu', 'Xuan-Bac Nguyen', 'Ngan Le', 'Chi Nhan Duong', 'Kha Gia Quach', 'Pha Nguyen'] | 2022-04-19 | null | null | null | null | ['3d-object-tracking'] | ['computer-vision'] | [-2.67088473e-01 -3.31877559e-01 -4.04813915e-01 -1.58086419e-02
-3.39073002e-01 -5.70899069e-01 6.88110709e-01 1.00082727e-02
-3.33361700e-02 3.41653824e-01 -3.02848190e-01 -1.23430192e-01
2.95762531e-03 -5.63564777e-01 -8.22816432e-01 -4.27520454e-01
-1.25835538e-02 4.94826376e-01 1.10428894e+00 -5.13972752e-02
2.00649902e-01 9.27902818e-01 -1.92431104e+00 -8.09352621e-02
5.30113816e-01 8.93879771e-01 5.40397763e-01 5.80171049e-01
-1.59832299e-01 5.08097112e-01 -1.62017569e-01 -4.21727538e-01
4.11719412e-01 2.09736556e-01 -8.30985010e-02 3.08509320e-01
1.12813652e+00 -3.78891647e-01 -3.07924211e-01 1.35325420e+00
-4.20342572e-02 -3.50906588e-02 5.65159678e-01 -1.76680171e+00
-5.08075535e-01 1.03558131e-01 -1.04996181e+00 5.00062168e-01
1.42172471e-01 1.31032825e-01 7.87658334e-01 -9.58801448e-01
7.46620059e-01 1.26248550e+00 8.97667825e-01 4.21665549e-01
-8.12516868e-01 -7.92086422e-01 4.33472335e-01 6.96409523e-01
-1.44861221e+00 -3.57145071e-01 6.11248493e-01 -5.96186340e-01
7.90200114e-01 9.72998366e-02 7.94194520e-01 7.09440172e-01
1.46911845e-01 6.32467210e-01 6.79449677e-01 -2.91186482e-01
-2.53523737e-01 3.46357077e-01 6.78732276e-01 9.43424940e-01
1.05354118e+00 4.00938660e-01 -5.52850842e-01 6.26028478e-02
5.10846615e-01 1.27722308e-01 1.15491070e-01 -8.80838811e-01
-9.08796728e-01 5.00762343e-01 2.71436185e-01 9.27448496e-02
-2.77350426e-01 2.36275598e-01 1.29353628e-01 -4.36549410e-02
2.66224831e-01 -1.60688490e-01 -2.30734974e-01 2.64614493e-01
-5.64970672e-01 2.36690208e-01 3.50014627e-01 1.50642800e+00
6.62415564e-01 3.98501828e-02 -1.34913266e-01 3.49038512e-01
7.30408370e-01 8.41393590e-01 -1.64023861e-01 -7.24221110e-01
2.79588819e-01 9.10015404e-01 4.06723261e-01 -1.35330868e+00
-4.18614030e-01 -4.20518100e-01 -3.76391232e-01 6.67948008e-01
2.15557069e-01 3.05014521e-01 -8.02874744e-01 1.50126100e+00
8.72083962e-01 6.40128672e-01 -1.24531113e-01 7.72187591e-01
8.85363281e-01 4.93723899e-01 -3.89509276e-02 -2.64351517e-01
1.31453860e+00 -1.01320744e+00 -8.44692826e-01 -6.08611032e-02
4.05946434e-01 -9.45705712e-01 1.41609758e-01 -1.29872579e-02
-8.37055862e-01 -1.12345314e+00 -1.15751398e+00 3.29592764e-01
-4.22134697e-01 2.41517991e-01 5.06603003e-01 6.30683005e-01
-9.62683380e-01 4.74442132e-02 -9.02452350e-01 -7.81977534e-01
2.52536476e-01 3.52556854e-01 -4.29054767e-01 -5.56307957e-02
-5.19280672e-01 1.23097432e+00 7.31982112e-01 9.78173390e-02
-8.33112240e-01 -5.75851798e-01 -6.03641093e-01 -3.07548612e-01
5.14250159e-01 -7.32555211e-01 8.33728492e-01 -4.79714245e-01
-8.05572987e-01 8.16375136e-01 -2.61772662e-01 -5.17616987e-01
5.07817686e-01 -3.70707929e-01 -7.41232157e-01 -2.67322689e-01
2.76337087e-01 5.00970960e-01 7.95884967e-01 -1.52957869e+00
-1.34027445e+00 -4.87611383e-01 -3.30476195e-01 -3.39338928e-02
-1.40260115e-01 3.04017309e-02 -9.58153009e-01 -2.04977810e-01
1.84789926e-01 -1.11427736e+00 1.36739481e-02 1.69565350e-01
-2.05275878e-01 -4.64062333e-01 1.62054372e+00 -3.55655730e-01
8.63105416e-01 -2.01351833e+00 -6.04501627e-02 1.99330732e-01
2.39654109e-01 5.66154063e-01 -1.45465404e-01 1.19197868e-01
1.20573387e-01 -2.26873308e-01 5.71836293e-01 -3.28922242e-01
-3.22825938e-01 7.01795220e-02 -3.66094001e-02 6.42893791e-01
2.78634846e-01 7.78195083e-01 -7.28226185e-01 -6.71525240e-01
5.76474130e-01 3.67373228e-01 -1.93195552e-01 7.20653236e-02
-2.08254606e-01 3.28937113e-01 -5.99974155e-01 8.71834874e-01
9.73035336e-01 -1.72104955e-01 -8.89399946e-02 -5.55309117e-01
-5.39639115e-01 -4.48418170e-01 -1.42327559e+00 1.40049887e+00
2.62259096e-01 6.99620605e-01 -2.59477422e-02 -6.42546356e-01
9.31386232e-01 8.05725977e-02 7.08151102e-01 -4.52759653e-01
2.16855168e-01 3.44533548e-02 -1.99080572e-01 -4.98698741e-01
6.49547279e-01 4.41569269e-01 2.65273869e-01 -9.48886424e-02
-1.93863928e-01 3.98583442e-01 2.52065867e-01 3.09294790e-01
1.03195906e+00 2.56061852e-01 1.13568276e-01 2.63619684e-02
7.74001598e-01 6.79370105e-01 6.13938570e-01 7.12568760e-01
-5.06949663e-01 1.05686948e-01 -4.09664840e-01 -5.56902647e-01
-1.04957080e+00 -1.11818624e+00 -1.31187961e-01 3.70640785e-01
7.99686790e-01 -1.84673622e-01 -1.54742822e-01 -8.08850586e-01
3.56782347e-01 5.05628467e-01 -4.80556697e-01 -1.75710902e-01
-5.34907520e-01 -3.25877547e-01 7.81590566e-02 6.15149379e-01
5.26892602e-01 -4.16991621e-01 -5.38271904e-01 1.11069612e-01
1.11614466e-01 -1.45714331e+00 -3.42353851e-01 -2.61081666e-01
-7.65839458e-01 -1.38480473e+00 -1.62064448e-01 -9.41021919e-01
5.65719485e-01 1.09316885e+00 7.37688839e-01 2.21507996e-01
-4.49322343e-01 7.29255438e-01 -4.42693114e-01 -5.92644930e-01
-5.63510418e-01 -3.25860113e-01 2.90268958e-01 1.52885735e-01
5.31029165e-01 4.43461537e-02 -4.43153620e-01 6.02843523e-01
-2.56580085e-01 -2.39488650e-02 5.45618951e-01 4.27078664e-01
7.54718959e-01 1.94744825e-01 4.89916414e-01 -4.29344505e-01
-6.30383715e-02 -4.61970657e-01 -1.32815528e+00 4.53456134e-01
-7.06914604e-01 -1.75949097e-01 1.50154792e-02 -5.55345118e-01
-9.88715053e-01 5.70369244e-01 4.55863148e-01 -8.87824893e-01
-6.33356944e-02 3.66305932e-02 -2.90696193e-02 -3.99080068e-01
2.72989511e-01 7.95698017e-02 1.47701427e-01 -3.93757284e-01
5.71927071e-01 3.39004934e-01 7.60632694e-01 8.11871812e-02
1.32646227e+00 7.19465077e-01 3.53697121e-01 -7.09392965e-01
-6.16171062e-01 -1.20741582e+00 -8.36787164e-01 -8.50253642e-01
1.02576482e+00 -1.14589155e+00 -7.99807072e-01 2.79568344e-01
-1.41765952e+00 5.08620441e-01 6.32168651e-02 6.90608978e-01
-2.78284162e-01 5.35262942e-01 -2.35093534e-02 -1.00783885e+00
-1.82671309e-01 -8.87589335e-01 1.04899931e+00 4.38566118e-01
1.16155045e-02 -7.77165651e-01 1.90831080e-01 4.43647772e-01
9.57638845e-02 2.39713758e-01 3.82849008e-01 -4.25212801e-01
-1.33640349e+00 -5.09271204e-01 -4.70073253e-01 -7.08111748e-02
2.03958005e-01 2.62875199e-01 -6.74612939e-01 -3.29704642e-01
-4.40601647e-01 2.48915955e-01 7.90170193e-01 4.02665645e-01
3.70404154e-01 3.61935437e-01 -1.06989872e+00 2.88386106e-01
1.51060653e+00 3.39642853e-01 1.85754374e-01 4.52512950e-01
9.53949571e-01 3.66259277e-01 1.21448290e+00 2.27262631e-01
4.00373369e-01 1.08046949e+00 7.99307108e-01 1.01383425e-01
-6.63975358e-01 -2.04963133e-01 2.60554403e-01 5.47318339e-01
-1.48038536e-01 -1.20333284e-01 -7.67205238e-01 6.63654089e-01
-2.20726109e+00 -1.24935269e+00 -8.77174735e-01 2.10118222e+00
-1.81179345e-02 1.65260717e-01 2.16281295e-01 -2.63624519e-01
1.17629015e+00 -2.72070646e-01 -5.16743839e-01 3.85924309e-01
-2.35677026e-02 -5.74931443e-01 8.89969528e-01 2.56893158e-01
-1.22097981e+00 1.02541721e+00 6.23824549e+00 5.16207516e-01
-7.63979912e-01 1.49001345e-01 -2.34461606e-01 1.87636778e-01
2.49354541e-01 -4.23627682e-02 -1.72265458e+00 3.15933734e-01
5.00039876e-01 -1.34646684e-01 -6.90033566e-03 8.98529351e-01
3.26459676e-01 -2.98106689e-02 -9.52043593e-01 9.95468438e-01
2.76386797e-01 -1.49340594e+00 8.80354196e-02 -6.41627684e-02
7.83999026e-01 1.74223796e-01 4.10895832e-02 7.49736428e-02
5.36705375e-01 -3.39759022e-01 8.15075219e-01 5.88329911e-01
2.19199896e-01 -5.33793449e-01 5.59160054e-01 4.06918585e-01
-1.83615887e+00 -9.08137038e-02 -4.76983100e-01 2.90933490e-01
3.61415356e-01 7.95790553e-02 -1.34715199e+00 6.80180311e-01
8.23969483e-01 9.80075836e-01 -8.12287807e-01 1.71201658e+00
2.58667320e-01 2.20550254e-01 -3.02138418e-01 -8.46655443e-02
-1.45601258e-01 -2.26026285e-03 9.89954591e-01 8.82699311e-01
5.15681744e-01 -1.93217859e-01 4.07286733e-01 7.16101408e-01
2.25915536e-01 -1.82582423e-01 -1.08353651e+00 3.85184497e-01
6.45767510e-01 1.35066295e+00 -8.10411930e-01 -2.69748479e-01
-8.51055443e-01 4.34449762e-01 2.58272767e-01 1.88578889e-01
-1.26424110e+00 3.68800282e-01 6.77869916e-01 2.60860950e-01
6.68884933e-01 -5.13875186e-01 -1.40857445e-02 -8.36846471e-01
-9.83647257e-02 -3.23271334e-01 4.69510376e-01 -8.93292189e-01
-1.25080907e+00 3.04695904e-01 1.58634737e-01 -1.59210825e+00
1.67678580e-01 -6.44074082e-01 -4.59609509e-01 4.86943632e-01
-1.63148797e+00 -1.74410498e+00 -6.04356587e-01 6.61112070e-01
6.68196023e-01 -3.97476375e-01 1.31416813e-01 5.30912101e-01
-4.68518168e-01 2.86626518e-01 -6.18499182e-02 -1.26722708e-01
6.15180731e-01 -7.18448699e-01 2.98262805e-01 1.18854320e+00
2.76770949e-01 2.47132599e-01 7.53420591e-01 -1.12945855e+00
-1.82489550e+00 -1.60547578e+00 5.43732166e-01 -8.64610732e-01
6.86902642e-01 7.16067478e-02 -7.93808103e-01 8.00801694e-01
1.07105136e-01 1.19813032e-01 1.85165927e-01 -1.54860327e-02
-2.38526508e-01 -2.86365569e-01 -7.75112927e-01 4.22386169e-01
1.37883246e+00 -4.84140776e-03 -6.48156226e-01 1.74878061e-01
4.94427800e-01 -2.71794528e-01 -4.99007940e-01 7.10638106e-01
4.98849958e-01 -6.02409244e-01 1.25420332e+00 -5.51589608e-01
-5.09093285e-01 -1.03848302e+00 -1.77893564e-01 -8.47911954e-01
-6.27284288e-01 -2.31275901e-01 -3.20496559e-01 1.38283765e+00
1.89276353e-01 -3.36246401e-01 9.37746406e-01 4.11428928e-01
-3.32026690e-01 1.15782611e-01 -1.02733254e+00 -1.04207289e+00
-6.49522126e-01 -3.70390505e-01 2.54889250e-01 7.42241383e-01
-5.44611335e-01 3.74363244e-01 -5.25475800e-01 9.82232034e-01
1.14269376e+00 3.20675164e-01 1.19136274e+00 -1.68361235e+00
-3.84795368e-02 -3.14309657e-01 -9.83100832e-01 -9.09156382e-01
5.22463173e-02 -9.18279588e-01 -3.81298140e-02 -1.50228930e+00
3.78231227e-01 -4.78476286e-01 -2.76125968e-01 1.15844145e-01
-1.71718970e-01 2.69038081e-01 3.43371809e-01 2.97188044e-01
-1.14815152e+00 3.75011086e-01 1.03202951e+00 -4.08880621e-01
-6.45204782e-02 1.72373444e-01 -1.06259055e-01 7.00915337e-01
4.51735795e-01 -5.70196331e-01 -3.55493009e-01 -4.11864370e-01
-2.72278464e-03 -5.42456806e-02 7.91772425e-01 -1.22351038e+00
7.40714967e-01 2.02027312e-03 4.28207785e-01 -1.57226360e+00
3.90366167e-01 -1.42333591e+00 7.25910604e-01 6.97716117e-01
3.15516263e-01 3.84712398e-01 5.46473980e-01 1.15937138e+00
-3.30857909e-03 -1.14444308e-01 6.72131956e-01 1.75155088e-01
-1.50780225e+00 4.04776543e-01 -2.17475966e-01 -4.71817642e-01
1.81473291e+00 -5.58728397e-01 -4.70344245e-01 1.53113559e-01
-5.66460252e-01 5.96772015e-01 5.09730935e-01 9.47997928e-01
6.56665385e-01 -1.57306492e+00 -6.43158793e-01 1.64121419e-01
4.54195529e-01 -2.61019498e-01 8.96954238e-02 7.62187481e-01
-3.20259899e-01 4.62174177e-01 -3.69457960e-01 -1.10588753e+00
-1.96865177e+00 8.31983030e-01 1.64276004e-01 1.37896448e-01
-5.55144668e-01 5.71531713e-01 2.69228444e-02 1.20066859e-01
3.17386895e-01 -9.84235629e-02 -5.40382266e-01 -7.03608617e-03
4.86579746e-01 5.54669201e-01 -4.87518236e-02 -9.22552943e-01
-4.30064648e-01 1.04707015e+00 -1.65046334e-01 4.07776028e-01
1.00909686e+00 -4.83742654e-01 1.69681847e-01 2.14121848e-01
7.09585845e-01 2.23603286e-02 -1.37799037e+00 -1.78113952e-01
1.14420645e-01 -6.11068130e-01 1.07664876e-01 -4.67934549e-01
-1.03075588e+00 3.04383099e-01 1.16663754e+00 1.69870108e-01
5.68015933e-01 2.90033430e-01 4.14122105e-01 3.26550633e-01
6.74681604e-01 -8.32466245e-01 1.09217770e-01 3.26642394e-01
4.69281316e-01 -1.53092897e+00 3.67490575e-02 -8.14784288e-01
-1.99776992e-01 9.25136507e-01 8.29419613e-01 6.91655874e-02
7.11513817e-01 2.00868860e-01 -4.44477089e-02 -3.82164448e-01
-6.82109952e-01 -6.01168036e-01 4.73108470e-01 8.82845938e-01
-2.01978028e-01 -1.51218429e-01 -9.54730511e-02 8.22469313e-03
5.38842022e-01 -1.43769115e-01 1.43205896e-01 7.51418948e-01
-6.85397625e-01 -1.08187592e+00 -7.24500477e-01 1.96929380e-01
-3.13744284e-02 4.13962334e-01 -3.56375426e-01 1.05357933e+00
3.17391813e-01 1.16475487e+00 3.58503312e-02 -5.17821610e-01
5.48305035e-01 -1.18600763e-01 4.11352366e-01 -4.00137514e-01
-1.02633990e-01 2.53468808e-02 1.38568670e-01 -3.58789265e-01
-8.27801704e-01 -8.40832233e-01 -1.19675910e+00 -1.40790790e-01
-8.87620270e-01 -1.50782928e-01 7.70828843e-01 5.58399737e-01
5.66230178e-01 5.39099038e-01 3.82043004e-01 -7.06600428e-01
-4.05014038e-01 -4.70671952e-01 -4.28083777e-01 5.57036340e-01
2.40219995e-01 -1.25848365e+00 4.89025824e-02 2.44403392e-01] | [6.636973857879639, -2.1424431800842285] |
a632eb23-8800-4ee9-92f6-6978ed2feded | a-complex-network-based-graph-embedding | 2209.04884 | null | https://arxiv.org/abs/2209.04884v1 | https://arxiv.org/pdf/2209.04884v1.pdf | A Complex Network based Graph Embedding Method for Link Prediction | Graph embedding methods aim at finding useful graph representations by mapping nodes to a low-dimensional vector space. It is a task with important downstream applications, such as link prediction, graph reconstruction, data visualization, node classification, and language modeling. In recent years, the field of graph embedding has witnessed a shift from linear algebraic approaches towards local, gradient-based optimization methods combined with random walks and deep neural networks to tackle the problem of embedding large graphs. However, despite this improvement in the optimization tools, graph embedding methods are still generically designed in a way that is oblivious to the particularities of real-life networks. Indeed, there has been significant progress in understanding and modeling complex real-life networks in recent years. However, the obtained results have had a minor influence on the development of graph embedding algorithms. This paper aims to remedy this by designing a graph embedding method that takes advantage of recent valuable insights from the field of network science. More precisely, we present a novel graph embedding approach based on the popularity-similarity and local attraction paradigms. We evaluate the performance of the proposed approach on the link prediction task on a large number of real-life networks. We show, using extensive experimental analysis, that the proposed method outperforms state-of-the-art graph embedding algorithms. We also demonstrate its robustness to data scarcity and the choice of embedding dimensionality. | ['Hafida Benhidour', 'Said Kerrache'] | 2022-09-11 | null | null | null | null | ['graph-reconstruction'] | ['graphs'] | [-1.59500867e-01 2.09841266e-01 -4.15283918e-01 3.07411095e-03
-7.41594806e-02 -4.67658669e-01 7.26223826e-01 7.30326533e-01
-2.27246925e-01 3.08519810e-01 2.32300967e-01 -4.13548887e-01
-3.96287978e-01 -1.00025058e+00 -2.61402637e-01 -5.06064892e-01
-5.82484186e-01 3.86063069e-01 1.12693891e-01 -3.41321081e-01
1.55333132e-01 6.48298979e-01 -9.75892901e-01 -3.55841696e-01
4.88632411e-01 6.26155853e-01 -1.96483180e-01 4.73655432e-01
-1.74230233e-01 3.63596588e-01 -7.16847479e-02 -7.30031610e-01
1.88378111e-01 -2.48214558e-01 -5.78048706e-01 -2.75286138e-01
3.61295901e-02 1.62945598e-01 -9.53795969e-01 1.12487113e+00
5.23419142e-01 7.76648521e-02 3.98147404e-01 -1.43408358e+00
-8.06436598e-01 5.26824296e-01 -3.21235031e-01 3.32819283e-01
2.67860204e-01 -8.89480580e-03 1.48337996e+00 -6.09377325e-01
6.77429616e-01 1.22401321e+00 7.52170861e-01 2.29176536e-01
-1.38151693e+00 -3.77921849e-01 1.28073275e-01 3.22410792e-01
-1.36439860e+00 -1.65972356e-02 1.27003717e+00 -5.30826867e-01
7.37464488e-01 2.35915691e-01 9.14177835e-01 7.81529665e-01
2.45202169e-01 4.17607456e-01 7.26150393e-01 -3.90441567e-01
1.00895666e-01 2.63057917e-01 1.98015913e-01 9.83033359e-01
3.06200445e-01 4.52333763e-02 -1.64129749e-01 -2.92157143e-01
4.30903703e-01 1.41345993e-01 -2.75698751e-01 -9.12856281e-01
-1.18025684e+00 1.17219663e+00 9.38624144e-01 6.71606362e-01
-2.56384879e-01 2.55778342e-01 6.42759860e-01 5.06450891e-01
6.78597867e-01 4.81840372e-01 1.04537442e-01 1.42190516e-01
-6.07780874e-01 7.54779428e-02 1.05981195e+00 5.15331864e-01
7.02852786e-01 4.56292881e-03 2.54811049e-02 5.28213084e-01
4.57865506e-01 1.94345921e-01 4.31090146e-01 -1.63694784e-01
4.14153576e-01 9.90499139e-01 -3.40978295e-01 -1.99584937e+00
-4.66876566e-01 -5.95208704e-01 -1.02540255e+00 -1.59962997e-01
2.05071181e-01 2.46164978e-01 -2.70246655e-01 1.57966340e+00
3.86527389e-01 2.40438834e-01 -1.92893282e-01 7.87001848e-01
6.29344940e-01 5.38870156e-01 -1.20855525e-01 1.33583426e-01
9.23797190e-01 -8.84769976e-01 -6.08851850e-01 -3.69615294e-02
7.94727981e-01 -3.43338519e-01 9.69131351e-01 -1.62643462e-01
-6.16140485e-01 -2.35077456e-01 -1.12112176e+00 1.51664257e-01
-7.22367764e-01 -2.29168177e-01 8.63178909e-01 6.93977237e-01
-1.19852662e+00 7.06271052e-01 -7.04448521e-01 -7.56770611e-01
3.44036311e-01 3.72534811e-01 -6.04949534e-01 -1.30459920e-01
-1.23945165e+00 6.83180392e-01 3.60861808e-01 2.19514742e-01
-3.81241918e-01 -4.47742611e-01 -9.37735021e-01 4.49607521e-01
3.86581928e-01 -6.19135737e-01 3.34369302e-01 -6.56067908e-01
-1.18251073e+00 7.02120841e-01 1.96171507e-01 -6.88357711e-01
3.25888813e-01 1.39849678e-01 -4.96512979e-01 4.92964387e-02
-2.39952058e-01 2.00076908e-01 7.28421152e-01 -7.85003662e-01
-8.58426932e-03 -4.88993973e-01 1.39611676e-01 -6.80431277e-02
-9.46128547e-01 -2.54847318e-01 -2.65909225e-01 -6.52278304e-01
4.93887812e-02 -1.02306199e+00 -3.28830481e-01 2.19968975e-01
-4.55159664e-01 -2.69465744e-01 7.95682788e-01 -3.86643738e-01
1.45487046e+00 -2.14914346e+00 5.89088321e-01 5.03593504e-01
7.61490881e-01 4.48024958e-01 -2.53177285e-01 1.05496585e+00
-2.48195618e-01 2.98250735e-01 -1.75654590e-01 -3.03460717e-01
7.96041712e-02 7.43644461e-02 -2.31391996e-01 7.28848457e-01
6.09279424e-02 1.19378531e+00 -1.17401767e+00 -4.65028554e-01
3.57712448e-01 6.82857335e-01 -5.65370500e-01 1.06219754e-01
5.15635982e-02 1.42724529e-01 -4.26029354e-01 3.75041962e-01
2.52369195e-01 -3.77624840e-01 5.17820239e-01 -2.90614724e-01
1.25244424e-01 4.17927466e-02 -1.00405097e+00 1.59049153e+00
-2.94945389e-01 8.72528851e-01 -2.69403737e-02 -1.57453287e+00
1.01687288e+00 1.85035199e-01 7.52744973e-01 -4.91951555e-01
1.26563579e-01 1.22775197e-01 1.19612299e-01 -1.88091382e-01
3.28972965e-01 1.23939337e-02 1.87835559e-01 5.01885414e-01
-6.86649233e-02 2.57374853e-01 3.04364145e-01 5.75138867e-01
1.31239212e+00 -5.07533491e-01 3.28402817e-01 -2.00002670e-01
5.60148001e-01 -2.97466427e-01 2.84299403e-02 2.78621763e-01
-2.11412847e-01 1.43618733e-01 8.47525835e-01 -5.07046700e-01
-9.26872849e-01 -9.02181685e-01 8.99644569e-02 6.70391619e-01
7.93802664e-02 -8.06395352e-01 -3.84663254e-01 -7.37816393e-01
3.64169925e-01 1.41571268e-01 -6.71646178e-01 -5.56760371e-01
-4.84774619e-01 -6.77493572e-01 3.92108172e-01 1.70160219e-01
2.87007131e-02 -9.85284090e-01 -5.15360711e-03 3.14747602e-01
2.65479326e-01 -1.08612096e+00 -3.91060412e-01 -1.89755321e-01
-9.94703531e-01 -1.11853075e+00 -6.45357132e-01 -7.85046816e-01
7.23140001e-01 4.09244448e-01 1.17582941e+00 4.44339633e-01
-3.79417926e-01 6.68784440e-01 -3.82607162e-01 2.26900607e-01
-4.33342665e-01 5.38066983e-01 -2.70000529e-02 3.55585873e-01
2.20934138e-01 -9.69706059e-01 -5.56755364e-01 1.27598807e-01
-1.05199122e+00 -1.76833779e-01 6.24762893e-01 8.43912303e-01
2.40832418e-01 1.83561400e-01 6.56363785e-01 -9.45094407e-01
1.22341895e+00 -7.60571241e-01 -6.27572238e-01 2.38919303e-01
-1.09457684e+00 3.47238302e-01 7.23589122e-01 -2.57274508e-01
-7.14765564e-02 -3.09860498e-01 1.53372750e-01 -4.52082336e-01
3.49664778e-01 1.06613934e+00 -1.61611903e-02 -3.81341040e-01
3.96444440e-01 2.26144478e-01 3.14389259e-01 -3.74683738e-01
6.15199447e-01 3.63979042e-01 1.22451365e-01 -1.36100575e-01
1.09322321e+00 4.14984643e-01 4.35285836e-01 -9.76071119e-01
-2.23957002e-01 -4.48376477e-01 -6.98889732e-01 -1.38247862e-01
5.03978729e-01 -5.44258296e-01 -6.36844158e-01 -9.20619518e-02
-9.18325961e-01 -8.62585157e-02 -1.81992829e-01 3.10555905e-01
-2.50996828e-01 7.67433584e-01 -4.20206517e-01 -5.01341939e-01
-3.73195589e-01 -1.02414966e+00 7.13578939e-01 -4.20165844e-02
-4.25451621e-02 -1.62359262e+00 5.66696346e-01 -8.06710199e-02
5.03358960e-01 3.39292049e-01 1.23294246e+00 -6.19188726e-01
-6.18963778e-01 -7.12475657e-01 -3.89383465e-01 1.66628048e-01
2.06752747e-01 -4.22707200e-02 -4.29784000e-01 -5.46031475e-01
-5.70840180e-01 1.23691559e-01 7.95328498e-01 4.15098341e-03
8.07734489e-01 -2.59451717e-01 -5.93584836e-01 5.68906784e-01
1.59638917e+00 -4.23610598e-01 3.59340876e-01 2.44193226e-01
9.82392907e-01 6.69566333e-01 1.99357450e-01 2.55838126e-01
4.67183977e-01 8.60506833e-01 7.67932415e-01 -4.53158468e-02
-1.18269093e-01 -4.22191381e-01 9.78665054e-02 1.01716077e+00
1.78508759e-01 -2.33057067e-01 -1.02839148e+00 7.44722962e-01
-1.97655737e+00 -9.27465081e-01 2.32129227e-02 2.25462484e+00
2.76114374e-01 2.05805928e-01 3.11607301e-01 2.68454045e-01
7.06274450e-01 6.26871347e-01 -3.29298228e-01 -5.03478348e-01
1.62636302e-02 -1.09678827e-01 3.94701123e-01 4.86389339e-01
-9.37846482e-01 8.06226730e-01 5.42863989e+00 5.61064780e-01
-1.32045257e+00 -9.92051288e-02 2.74844915e-01 3.12078118e-01
-5.75523555e-01 5.23975044e-02 -2.67120004e-01 4.82570171e-01
8.85701418e-01 -5.08962393e-01 7.20550895e-01 6.93238974e-01
2.39993334e-02 4.92902100e-01 -1.24908364e+00 1.01480711e+00
-4.34546880e-02 -1.47280931e+00 1.22459255e-01 3.17401528e-01
4.09792036e-01 1.32736817e-01 2.83356130e-01 2.47435465e-01
-7.00689554e-02 -1.08532071e+00 8.24995190e-02 3.77039254e-01
5.37485063e-01 -7.56428182e-01 6.41542852e-01 1.91517428e-01
-1.59175837e+00 -1.53136507e-01 -3.49390984e-01 -1.14427537e-01
1.42372966e-01 8.57456207e-01 -9.82418001e-01 7.08593190e-01
2.44263425e-01 1.10577881e+00 -7.58663476e-01 1.06063688e+00
-7.75732622e-02 4.93396014e-01 -1.64092064e-01 -2.18230456e-01
4.00218755e-01 -5.90419769e-01 8.77107084e-01 9.62890983e-01
7.83386603e-02 -4.02936637e-01 5.09517491e-02 9.14832890e-01
-2.64453292e-01 3.76911372e-01 -1.17791367e+00 -7.51813591e-01
3.28263998e-01 1.51170862e+00 -8.64805460e-01 6.02110140e-02
-5.10276377e-01 7.88442016e-01 6.89749181e-01 2.26477623e-01
-6.17484450e-01 -4.30079490e-01 5.81570208e-01 3.69599283e-01
2.21478462e-01 -6.36947870e-01 2.28435636e-01 -1.04922771e+00
1.30686179e-01 -5.75705409e-01 3.01149279e-01 -2.32374907e-01
-1.43660784e+00 6.55698776e-01 -1.43835142e-01 -9.55425382e-01
-2.80974895e-01 -5.24892628e-01 -8.10267508e-01 6.07370079e-01
-1.53967083e+00 -1.05651319e+00 -2.05563232e-01 3.66169274e-01
-3.87696619e-03 -1.85212091e-01 8.51146042e-01 5.42564452e-01
-6.04943693e-01 6.35262370e-01 5.32450855e-01 3.14492375e-01
3.74855280e-01 -1.24467278e+00 6.15114927e-01 5.72992563e-01
5.58605134e-01 5.36987960e-01 5.80352485e-01 -4.52585816e-01
-1.81729746e+00 -9.52087402e-01 8.66373241e-01 -1.99653089e-01
1.17520094e+00 -6.78186119e-01 -8.17099988e-01 6.09335482e-01
-7.47616887e-02 5.20401537e-01 5.70681393e-01 3.52489948e-01
-3.60144228e-01 -2.68863648e-01 -9.37654138e-01 7.54985809e-01
1.01930451e+00 -8.42208922e-01 -3.26055363e-02 4.41895127e-01
6.60439432e-01 1.71197593e-01 -9.88236248e-01 1.82102948e-01
6.55037224e-01 -7.30226159e-01 1.21710837e+00 -8.19684744e-01
8.90241042e-02 -1.69080608e-02 4.93815057e-02 -1.48603129e+00
-3.55356365e-01 -7.13311493e-01 -3.95932317e-01 9.66281652e-01
1.44217387e-01 -8.72964740e-01 9.51930702e-01 2.77523220e-01
2.87686706e-01 -1.07840741e+00 -9.09584761e-01 -7.15131164e-01
-1.54894115e-02 2.54299864e-03 3.66853535e-01 1.12352633e+00
2.23990664e-01 6.22074783e-01 -3.32779378e-01 -1.82281211e-02
5.67511141e-01 2.44630292e-01 8.87814999e-01 -1.56916690e+00
-1.68468162e-01 -7.58604646e-01 -1.18723142e+00 -7.33045340e-01
4.36557561e-01 -1.31188679e+00 -7.15986609e-01 -1.49672616e+00
-1.96481124e-02 -4.60182995e-01 -3.64553064e-01 1.16248988e-01
-1.69336915e-01 2.16856942e-01 2.95599073e-01 1.43080607e-01
-5.06244659e-01 7.53946662e-01 9.68991458e-01 -3.23456764e-01
-2.09599167e-01 -2.15557292e-02 -5.28635859e-01 3.29786599e-01
7.34530210e-01 -5.16806424e-01 -5.75583041e-01 -2.22081795e-01
6.97816730e-01 -1.21577658e-01 4.76142824e-01 -9.31696415e-01
2.00253904e-01 2.62070417e-01 -3.21479231e-01 -5.50673530e-02
2.07759053e-01 -1.10602403e+00 2.25566160e-02 6.34318650e-01
-2.61872858e-01 3.74508381e-01 -6.65118918e-02 1.09069395e+00
-2.46374533e-01 4.17662226e-02 4.42021489e-01 2.17978671e-01
-5.62998414e-01 5.11146247e-01 3.22839953e-02 -1.37929216e-01
1.03588247e+00 -1.70437008e-01 -1.73552155e-01 -5.19961059e-01
-7.88850307e-01 1.10952362e-01 4.56479192e-01 5.12937486e-01
6.38189971e-01 -1.47865009e+00 -6.15665078e-01 9.15849209e-02
2.53496021e-01 -5.30151486e-01 -6.37885407e-02 1.06050706e+00
-4.25389379e-01 6.22404635e-01 -2.27048434e-02 -4.89266098e-01
-1.15654397e+00 7.64208496e-01 1.43990800e-01 -7.79249310e-01
-8.44941616e-01 3.73751104e-01 -6.53490350e-02 -4.89557207e-01
1.94147766e-01 -1.25470266e-01 -2.83843815e-01 8.90160576e-02
8.68816227e-02 4.44325089e-01 -2.70362068e-02 -5.93211651e-01
-3.64376962e-01 5.57191551e-01 5.66591546e-02 2.62213498e-01
1.44101143e+00 -2.72384211e-02 -2.23737314e-01 4.96290207e-01
1.56702065e+00 -1.56201953e-02 -6.32963777e-01 -3.85503560e-01
1.63022086e-01 -4.61443037e-01 1.09265916e-01 -1.94242522e-01
-1.28813720e+00 9.94195998e-01 5.68057001e-01 8.71259272e-01
6.04826868e-01 2.60743070e-02 7.51290143e-01 4.79487956e-01
2.91347355e-01 -4.97443050e-01 2.37817749e-01 2.14357749e-01
7.68824458e-01 -1.36434519e+00 1.17165141e-01 -3.56849670e-01
-7.99367577e-02 1.02161264e+00 1.33899137e-01 -2.68156439e-01
1.11593163e+00 -3.32065433e-01 -4.02648181e-01 -5.45892775e-01
-4.94197756e-01 -7.99054578e-02 4.58711863e-01 4.59003150e-01
3.74418676e-01 1.24902472e-01 -3.99901032e-01 1.88850492e-01
5.19779436e-02 -3.58838201e-01 2.79312760e-01 6.47798657e-01
-2.62239009e-01 -1.33179986e+00 1.28207490e-01 4.29028302e-01
-1.10001825e-01 -1.23554647e-01 -5.90738356e-01 9.45391119e-01
-6.30663335e-01 5.96778154e-01 -1.71112955e-01 -3.79824013e-01
5.31421639e-02 -1.01251662e-01 2.49742076e-01 -5.45237422e-01
-4.16074932e-01 -5.89049935e-01 -1.57140836e-01 -5.12196004e-01
-2.43530169e-01 -2.77083784e-01 -9.35707033e-01 -5.55606246e-01
-4.35282886e-01 3.57436419e-01 7.52401888e-01 6.25747681e-01
7.05250978e-01 4.74967003e-01 8.03557813e-01 -8.29294026e-01
-5.82235932e-01 -6.81705773e-01 -6.62816584e-01 3.69581252e-01
1.99668348e-01 -6.95944786e-01 -4.70631957e-01 -6.34453654e-01] | [7.0891432762146, 6.01719331741333] |
a972da5b-cc8e-4a9c-b42d-5c1ba0100690 | image-based-virtual-try-on-system-with | 2302.14197 | null | https://arxiv.org/abs/2302.14197v1 | https://arxiv.org/pdf/2302.14197v1.pdf | Image-Based Virtual Try-on System With Clothing-Size Adjustment | The conventional image-based virtual try-on method cannot generate fitting images that correspond to the clothing size because the system cannot accurately reflect the body information of a person. In this study, an image-based virtual try-on system that could adjust the clothing size was proposed. The size information of the person and clothing were used as the input for the proposed method to visualize the fitting of various clothing sizes in a virtual space. First, the distance between the shoulder width and height of the clothing in the person image is calculated based on the coordinate information of the key points detected by OpenPose. Then, the system changes the size of only the clothing area of the segmentation map, whose layout is estimated using the size of the person measured in the person image based on the ratio of the person and clothing sizes. If the size of the clothing area increases during the drawing, the details in the collar and overlapping areas are corrected to improve visual appearance. | ['Nobuo Funabiki', 'Koki Nakai', 'Minoru Kuribayashi'] | 2023-02-27 | null | null | null | null | ['virtual-try-on'] | ['computer-vision'] | [-6.32426143e-02 -8.36216211e-02 3.65841001e-01 -1.42949253e-01
4.87468213e-01 -4.44465846e-01 -2.81445801e-01 2.03221709e-01
-3.84467572e-01 1.34028584e-01 -2.16320544e-01 3.25583629e-02
1.88009337e-01 -9.55118239e-01 -3.67780983e-01 -1.83276758e-01
6.13730967e-01 4.08179909e-01 5.72582424e-01 -3.46601158e-01
2.78723508e-01 4.97485489e-01 -1.21670544e+00 -7.30405748e-02
6.82176828e-01 6.67729676e-01 3.10174793e-01 4.37851787e-01
-2.62020171e-01 -3.76282215e-01 -7.81214118e-01 -4.34902489e-01
4.81085688e-01 -4.09170300e-01 7.76161030e-02 6.71583235e-01
4.93418068e-01 -3.55629683e-01 -5.39570414e-02 1.10595286e+00
4.00721312e-01 -7.50925466e-02 3.38615626e-01 -6.16293192e-01
-5.17457783e-01 3.29059988e-01 -1.15272796e+00 -2.82336622e-01
5.04954934e-01 -2.32694820e-01 8.00881609e-02 -5.40853798e-01
7.02475190e-01 1.12268198e+00 6.31890059e-01 2.50832051e-01
-1.12544501e+00 -6.51408911e-01 1.94017515e-01 -1.52462229e-01
-1.54634035e+00 2.67970890e-01 1.08331811e+00 -4.03900474e-01
-6.80028461e-03 6.29413784e-01 1.23288286e+00 -7.63914585e-02
1.74522772e-01 9.23107788e-02 1.05795109e+00 -8.39044034e-01
-2.36827627e-01 4.84529436e-01 8.71395413e-03 9.31169987e-01
3.35865587e-01 1.58642400e-02 5.91138261e-04 6.18297122e-02
1.28492177e+00 1.18995480e-01 -1.84725881e-01 -4.52042937e-01
-1.04042625e+00 3.28207850e-01 5.43470919e-01 2.80090600e-01
-8.10856372e-02 -1.73950791e-01 1.86801836e-01 -2.58654863e-01
3.41382265e-01 3.48602146e-01 -2.32397065e-01 2.61474609e-01
-1.17729127e+00 6.48747757e-02 2.87040740e-01 6.25366807e-01
7.32419968e-01 -3.28424275e-02 6.36504218e-02 5.42806804e-01
3.20395797e-01 4.87026364e-01 1.45068079e-01 -7.36715972e-01
3.82054925e-01 1.26153409e+00 2.56668597e-01 -1.33968389e+00
-4.48982209e-01 -3.68979163e-02 -4.92098272e-01 2.89404273e-01
8.42408955e-01 -2.67685145e-01 -1.04586196e+00 8.36303174e-01
8.59232545e-01 -3.49932402e-01 -6.50537133e-01 1.16507792e+00
9.88411129e-01 4.09853905e-01 -6.22266307e-02 -1.85829207e-01
1.58920050e+00 -5.81974506e-01 -8.83165717e-01 -1.81291848e-01
1.65763229e-01 -1.02296567e+00 1.06464279e+00 -5.68978712e-02
-1.12185013e+00 -9.04121280e-01 -1.02433038e+00 2.13813096e-01
-1.87083736e-01 7.33347476e-01 2.04477116e-01 5.21207988e-01
-4.77131337e-01 4.95264888e-01 -5.47801912e-01 -4.32972223e-01
-4.15664315e-01 6.46568239e-02 -1.12250380e-01 2.73435056e-01
-8.56866896e-01 8.12841713e-01 1.71581641e-01 6.48610115e-01
-1.37178734e-01 -5.27264476e-01 -8.33828092e-01 -5.56973554e-02
2.58158982e-01 -7.50422120e-01 4.68368381e-01 -1.27487266e+00
-1.46809614e+00 1.05084527e+00 3.27070266e-01 1.33480147e-01
9.56923425e-01 -5.37676439e-02 -3.41847658e-01 2.90122665e-02
3.23112309e-02 2.71003693e-01 7.42577195e-01 -1.21351945e+00
-1.12810746e-01 -7.38317192e-01 1.22978501e-01 4.19342041e-01
3.42306733e-01 8.22266638e-02 -8.24082077e-01 -6.11291230e-01
6.20447099e-01 -7.87139714e-01 -2.86889046e-01 4.46986973e-01
-7.23829329e-01 2.85525650e-01 8.66886854e-01 -1.18065715e+00
1.36329198e+00 -2.30721307e+00 -1.67948499e-01 5.86516738e-01
3.22869793e-02 1.39251620e-01 3.38057607e-01 4.89534624e-02
8.73771310e-02 -1.89291185e-03 -1.20367162e-01 9.98852700e-02
-3.27756733e-01 -2.28212759e-01 6.04000330e-01 4.19707388e-01
-6.79713011e-01 2.65357524e-01 -4.51813579e-01 -8.80278528e-01
5.52070677e-01 4.34078425e-01 5.66940457e-02 2.30400890e-01
9.87558588e-02 1.97683200e-01 -2.22819641e-01 3.92672718e-01
1.03824794e+00 5.03289640e-01 5.14921486e-01 -7.77083814e-01
-1.99183494e-01 -5.31276345e-01 -1.55257630e+00 1.06560636e+00
-1.80710554e-01 3.90570998e-01 2.87063688e-01 -7.60952681e-02
1.40292275e+00 1.20005943e-01 3.47457498e-01 -4.79056805e-01
1.97189897e-01 -1.93507329e-01 -1.00710623e-01 -5.04740775e-01
4.48978513e-01 1.26376092e-01 9.93076414e-02 2.14132622e-01
-8.58723402e-01 -3.45416635e-01 3.97930562e-01 -1.33419678e-01
-1.78026363e-01 5.15546083e-01 2.93268859e-01 -5.75169362e-02
3.05132359e-01 7.39798099e-02 3.47277284e-01 1.21325970e-01
2.98610181e-01 6.63090348e-01 2.67628640e-01 -7.07641423e-01
-1.27189660e+00 -7.58398592e-01 -3.23876828e-01 7.25249529e-01
7.08544135e-01 -4.40952480e-01 -1.29184890e+00 -3.26610833e-01
1.28532657e-02 3.46697420e-01 -6.20815873e-01 2.37780064e-01
-5.22740901e-01 -4.70656395e-01 -1.71779037e-01 2.89007246e-01
8.04030657e-01 -7.39076972e-01 -8.07799816e-01 -5.26508465e-02
-4.60138358e-02 -7.52893150e-01 -7.57454276e-01 -6.94134057e-01
-8.54179263e-01 -1.26148188e+00 -8.72837901e-01 -7.67363369e-01
1.29465795e+00 1.08320862e-01 4.97799218e-01 3.03847015e-01
-3.79524261e-01 -8.50002654e-03 -2.00857952e-01 -3.85462642e-01
-2.97778398e-01 -3.77443999e-01 -1.57468468e-01 7.41967112e-02
-1.30257398e-01 8.19898769e-02 -7.60610521e-01 7.08336473e-01
-2.91565627e-01 6.68403506e-01 -2.10104331e-01 -2.62273271e-02
7.77263045e-01 1.68027103e-01 -4.58328635e-01 -6.99386060e-01
6.93918288e-01 2.32326090e-01 -6.80025756e-01 4.28779960e-01
-3.36490542e-01 -4.31578070e-01 3.34128857e-01 -6.86941743e-01
-8.75277460e-01 2.49452866e-03 2.12876245e-01 -3.49862546e-01
-1.35611668e-01 -1.09233260e-02 -3.21805716e-01 6.62311539e-02
3.01257581e-01 -2.29075458e-02 1.04771040e-01 -3.49960685e-01
2.04491228e-01 5.33899367e-01 3.94645810e-01 -8.01810175e-02
5.25216222e-01 2.82816350e-01 -7.19271228e-02 -6.00025296e-01
-2.95702517e-01 -1.80676337e-02 -1.11364233e+00 -7.01641798e-01
9.51808095e-01 -3.83617818e-01 -6.66047871e-01 3.49565476e-01
-1.20471525e+00 -6.24825433e-02 -2.47376263e-01 4.92492437e-01
8.95486586e-03 4.06422287e-01 -6.26372516e-01 -6.97820544e-01
-4.84783471e-01 -9.44062889e-01 7.69428015e-01 6.72795177e-01
-1.46391526e-01 -6.46915615e-01 -1.98174357e-01 1.97381422e-01
-1.03333421e-01 4.18686539e-01 7.94766963e-01 3.27495903e-01
-2.75047988e-01 -5.42408586e-01 -1.55945746e-02 2.23200433e-02
1.30349368e-01 7.26112068e-01 -4.49878722e-01 3.00001223e-02
-3.22784036e-01 7.58798957e-01 2.26915896e-01 7.86230624e-01
1.04587531e+00 -1.39207184e-01 -4.83855397e-01 6.56299114e-01
1.52697849e+00 2.66927123e-01 5.98212302e-01 2.67877579e-01
8.64315867e-01 7.36954272e-01 8.52837205e-01 3.04050684e-01
2.13454515e-01 9.31505442e-01 1.61471158e-01 -6.83065295e-01
-2.10960135e-01 -4.05185014e-01 1.64349061e-02 5.79923153e-01
-5.85298717e-01 2.81595320e-01 -4.75493371e-01 5.66790849e-02
-1.22678638e+00 -5.98355412e-01 -7.66978979e-01 2.57094908e+00
4.31667238e-01 5.34837246e-02 2.38585606e-01 3.87772806e-02
1.09508872e+00 -2.37426996e-01 -3.89978290e-01 -8.82021546e-01
2.39637226e-01 -3.52465391e-01 6.42501175e-01 5.04814565e-01
-6.33333802e-01 5.13567984e-01 6.45093727e+00 3.44644517e-01
-1.16649187e+00 -2.62297839e-01 4.14178669e-01 -1.69757511e-02
-1.79682389e-01 -7.33985230e-02 -6.15897715e-01 7.69981086e-01
-2.51334868e-02 -6.10023970e-03 3.85571778e-01 6.77450299e-01
4.53094631e-01 -5.88350236e-01 -5.91658115e-01 1.01120329e+00
1.23345748e-01 -7.07658350e-01 -1.33552074e-01 -7.13339970e-02
4.47026968e-01 -8.53072345e-01 -1.24830902e-01 -3.51539105e-01
-5.21368980e-01 -3.33827913e-01 1.04757190e+00 7.72549450e-01
1.16710722e+00 -5.23114800e-01 7.91226864e-01 2.07782507e-01
-1.43536794e+00 3.84359628e-01 -5.59250116e-01 -9.43799764e-02
2.94592142e-01 4.13232267e-01 -8.40557098e-01 2.82863289e-01
5.99495590e-01 -6.61459491e-02 -7.18873799e-01 9.30948794e-01
-1.77859098e-01 7.77047426e-02 -3.86785448e-01 1.46709025e-01
-4.09333587e-01 -8.74033213e-01 4.30932015e-01 9.26269352e-01
4.04280663e-01 -8.20223242e-02 2.64101774e-01 1.11489975e+00
2.83180743e-01 4.99214381e-01 -3.06779921e-01 3.91578466e-01
4.34725225e-01 1.26416433e+00 -9.39841449e-01 -2.81749755e-01
-5.53133599e-02 1.07016170e+00 -1.74611911e-01 4.29528236e-01
-9.16160882e-01 -2.78773189e-01 -9.00530517e-02 1.01250982e+00
3.62633690e-02 -1.92638054e-01 -2.84814000e-01 -5.01942635e-01
1.68889984e-01 -3.62697035e-01 1.61195248e-01 -1.14242852e+00
-5.54605782e-01 5.15264571e-01 6.17455021e-02 -9.95873511e-01
2.72593647e-01 -1.93364263e-01 -5.88979900e-01 1.03431177e+00
-2.00612411e-01 -1.33529758e+00 -5.99928200e-01 1.07567154e-01
3.97173762e-01 4.18417156e-01 7.81992018e-01 1.42581895e-01
-6.54688656e-01 5.29805064e-01 -5.68863675e-02 2.37750322e-01
5.06574690e-01 -1.02305567e+00 1.17620736e-01 5.66950917e-01
-2.07118094e-01 4.39910024e-01 6.82718635e-01 -1.05569625e+00
-7.35237420e-01 -4.73861963e-01 6.75450504e-01 -2.00109869e-01
-6.00599721e-02 -5.01835585e-01 -6.18623257e-01 4.76018012e-01
-7.80204833e-02 -2.23484457e-01 4.24000561e-01 -3.50452843e-03
2.32642829e-01 -4.16605771e-01 -1.24016345e+00 5.36321104e-01
4.61829811e-01 -1.01983964e-01 -1.98456094e-01 8.85375217e-03
3.29380780e-01 -8.63304555e-01 -9.43842411e-01 1.25843883e-01
1.05296791e+00 -9.59218204e-01 1.11286116e+00 6.68509752e-02
8.40153694e-02 -6.59688354e-01 4.18180615e-01 -1.15902257e+00
-3.22417855e-01 6.56612739e-02 5.11001408e-01 1.07860148e+00
2.75076419e-01 -1.21918373e-01 6.30251050e-01 9.60081279e-01
3.73193830e-01 -7.01110542e-01 -4.92905706e-01 1.05969258e-01
-7.55034626e-01 1.30036801e-01 6.07731223e-01 5.71070015e-01
-2.19406292e-01 5.82292071e-03 -1.88218519e-01 2.00960249e-01
5.06439865e-01 3.34098697e-01 8.01994264e-01 -1.06098688e+00
1.30971838e-02 6.09311610e-02 -3.78945827e-01 -6.93374813e-01
-6.84406996e-01 -1.77708536e-01 -4.41019416e-01 -1.77860916e+00
3.43654215e-01 -3.77080381e-01 2.65634388e-01 6.81656823e-02
-3.32659692e-01 2.22224787e-01 6.37174487e-01 2.08133504e-01
2.79820394e-02 -1.06864996e-01 1.92707956e+00 4.90515903e-02
-5.42801440e-01 1.78976014e-01 -1.73168078e-01 1.14209652e+00
5.49582601e-01 -9.79051441e-02 -1.39369667e-01 -4.27611262e-01
-1.16362244e-01 1.67827234e-01 2.13515773e-01 -7.37983644e-01
5.46884239e-02 -1.66452974e-01 1.05367064e+00 -8.35851550e-01
2.80078232e-01 -1.06861842e+00 6.91284478e-01 5.98575056e-01
-5.05304895e-03 2.37411618e-01 2.27535203e-01 1.01029344e-01
5.05451798e-01 -2.83387840e-01 6.96677983e-01 -5.90701699e-01
-2.18975261e-01 -1.37249947e-01 -3.72029506e-02 -7.48860300e-01
1.23990941e+00 -8.98109794e-01 8.33998099e-02 -2.52742082e-01
-9.61408854e-01 -9.58268065e-03 9.62868989e-01 -1.36498734e-01
7.78009713e-01 -1.04520655e+00 -5.08602083e-01 5.28507292e-01
-2.97647655e-01 -6.80777151e-03 5.65012813e-01 7.78844893e-01
-1.10926425e+00 -1.69369638e-01 -3.81692618e-01 -3.42888802e-01
-1.95385683e+00 6.85325444e-01 9.38959718e-01 2.13128030e-01
-6.60643578e-01 3.91037792e-01 2.54897237e-01 -3.63795429e-01
5.92279546e-02 -3.63421708e-01 -3.79086643e-01 -1.81563273e-01
4.60750997e-01 6.66591048e-01 -4.00445670e-01 -8.21763217e-01
-1.57907575e-01 1.25388312e+00 3.15410703e-01 2.20262874e-02
6.94038570e-01 -5.45402348e-01 -2.28745788e-01 4.11843061e-01
3.79955947e-01 5.59084654e-01 -1.12131858e+00 9.51133221e-02
-7.99560726e-01 -8.87474716e-01 -1.97464824e-01 -8.33678901e-01
-1.05033624e+00 7.35414505e-01 1.04506707e+00 1.47761047e-01
9.75348175e-01 -1.03212148e-01 5.57221591e-01 -7.21956253e-01
3.96708757e-01 -1.35520434e+00 -5.21993697e-01 -3.56961876e-01
7.78284192e-01 -6.32484257e-01 4.38975632e-01 -9.36316669e-01
-6.43887281e-01 1.39586985e+00 1.09289980e+00 -3.07669509e-02
3.45662981e-01 2.07625151e-01 2.12584496e-01 -2.40598902e-01
4.84630078e-01 1.76985618e-02 5.67857325e-01 5.07023036e-01
2.76343852e-01 4.67578262e-01 -8.74592423e-01 4.25676078e-01
-1.90393493e-01 -8.92565474e-02 3.70007873e-01 3.91707361e-01
-4.96273190e-01 -6.76473558e-01 -9.96202052e-01 2.73106515e-01
-2.57932007e-01 3.26776147e-01 -4.39964980e-01 7.11713672e-01
8.86969268e-01 7.08651721e-01 4.98539954e-01 -2.64491200e-01
5.98109305e-01 -5.52802265e-01 8.89040589e-01 -4.39529270e-01
-4.41494435e-01 4.05895531e-01 -8.49708617e-02 -1.77200124e-01
-1.53246075e-01 -5.86647280e-02 -1.35542381e+00 -5.18572927e-01
-4.51092511e-01 -4.59840484e-02 9.32960927e-01 4.25218463e-01
-1.00484580e-01 3.40644807e-01 3.34757268e-01 -7.77620494e-01
1.62957415e-01 -9.37485397e-01 -9.55354214e-01 6.13454223e-01
-3.70634943e-01 -3.69405061e-01 -4.86389361e-02 -3.11440621e-02] | [11.35093879699707, -1.1047451496124268] |
df8af7d0-047e-428b-973c-df78291100d4 | abusive-and-threatening-language-detection-in | 2111.14830 | null | https://arxiv.org/abs/2111.14830v1 | https://arxiv.org/pdf/2111.14830v1.pdf | Abusive and Threatening Language Detection in Urdu using Boosting based and BERT based models: A Comparative Approach | Online hatred is a growing concern on many social media platforms. To address this issue, different social media platforms have introduced moderation policies for such content. They also employ moderators who can check the posts violating moderation policies and take appropriate action. Academicians in the abusive language research domain also perform various studies to detect such content better. Although there is extensive research in abusive language detection in English, there is a lacuna in abusive language detection in low resource languages like Hindi, Urdu etc. In this FIRE 2021 shared task - "HASOC- Abusive and Threatening language detection in Urdu" the organizers propose an abusive language detection dataset in Urdu along with threatening language detection. In this paper, we explored several machine learning models such as XGboost, LGBM, m-BERT based models for abusive and threatening content detection in Urdu based on the shared task. We observed the Transformer model specifically trained on abusive language dataset in Arabic helps in getting the best performance. Our model came First for both abusive and threatening content detection with an F1scoreof 0.88 and 0.54, respectively. | ['Punyajoy Saha', 'Somnath Banerjee', 'Mithun Das'] | 2021-11-27 | null | null | null | null | ['abusive-language'] | ['natural-language-processing'] | [-6.75750494e-01 -8.64014477e-02 -3.14229488e-01 -3.55050229e-02
-7.71610618e-01 -8.54831100e-01 6.75209522e-01 3.38702559e-01
-3.95033896e-01 8.27939987e-01 4.52878475e-01 -2.98220336e-01
2.56988913e-01 -5.90448439e-01 -1.68907136e-01 -4.96201664e-01
3.79696578e-01 1.43700778e-01 8.32450092e-02 -8.77258778e-01
7.23009169e-01 3.61778557e-01 -7.89016008e-01 4.22217816e-01
1.06811965e+00 2.79453784e-01 -1.32564336e-01 7.30328441e-01
-2.09741965e-01 1.48063385e+00 -1.15545595e+00 -7.57566571e-01
5.02595529e-02 -4.94056433e-01 -9.71477449e-01 -1.09515734e-01
5.78495026e-01 -4.60296690e-01 -2.27082968e-01 1.31452632e+00
7.09401369e-01 -1.85386494e-01 5.49943447e-01 -1.15987873e+00
-9.90329623e-01 1.05309284e+00 -1.20451880e+00 8.63607764e-01
4.96710211e-01 -2.09017843e-01 9.32299137e-01 -6.68756127e-01
7.67782331e-01 1.57999277e+00 4.95453328e-01 5.51792383e-01
-6.58207715e-01 -1.00608420e+00 -5.19761741e-02 1.78910330e-01
-1.12226760e+00 4.97999825e-02 8.19427609e-01 -8.12572479e-01
6.47875547e-01 3.65351558e-01 3.09252858e-01 1.45484114e+00
3.90194237e-01 8.52661848e-01 1.26360607e+00 -3.70057344e-01
-3.38594049e-01 6.42354846e-01 5.92652738e-01 6.75135851e-01
3.85364830e-01 -7.12050140e-01 -6.43287003e-01 -5.39227307e-01
-1.33547530e-01 -8.12221766e-02 8.46056715e-02 8.36134613e-01
-5.01687825e-01 1.63872194e+00 2.56886840e-01 7.72627413e-01
-6.38081059e-02 -1.79602563e-01 1.01590979e+00 5.22870958e-01
1.23618340e+00 5.57865441e-01 5.09093776e-02 6.01852834e-02
-8.52624536e-01 2.94004798e-01 6.56498790e-01 5.36060333e-01
4.27737027e-01 -1.07677191e-01 -6.82943463e-02 1.32519996e+00
5.22433281e-01 4.77085531e-01 5.97431839e-01 -3.12861294e-01
5.47002912e-01 6.49344027e-01 -8.32237229e-02 -1.36997175e+00
-3.12901497e-01 -2.75485098e-01 -3.33708465e-01 -1.71942208e-02
3.37218791e-01 -6.65595293e-01 -5.91884851e-01 1.24011385e+00
4.30168539e-01 -4.26869661e-01 -5.17876267e-01 9.80953097e-01
6.50662184e-01 9.82605100e-01 3.45443100e-01 -3.35182279e-01
1.55017626e+00 -7.32368350e-01 -1.24657238e+00 1.28465265e-01
1.04348397e+00 -1.38800097e+00 1.10025609e+00 8.83514881e-01
-7.26010859e-01 9.94562432e-02 -1.15705609e+00 -3.74585152e-01
-6.90147519e-01 -4.41221774e-01 3.92869711e-01 1.33677793e+00
-6.27384126e-01 3.71649981e-01 -1.56136407e-02 -6.56600356e-01
3.04568976e-01 -1.96862489e-01 -2.26293340e-01 1.58870772e-01
-1.51504421e+00 1.20103347e+00 1.43144339e-01 -4.29703146e-01
-1.05895007e+00 -3.73508096e-01 -3.29722524e-01 -8.21800470e-01
4.53099936e-01 2.75450379e-01 9.74725544e-01 -1.17987597e+00
-9.66586471e-01 1.46308172e+00 3.92005920e-01 -3.33076924e-01
8.63075078e-01 -9.09725904e-01 -6.12013698e-01 8.64517763e-02
4.48648959e-01 1.09992787e-01 9.17236269e-01 -8.46723497e-01
-5.72781086e-01 -6.31169200e-01 1.92152143e-01 2.53301244e-02
-8.18552315e-01 1.35986185e+00 5.42418361e-01 -9.08942342e-01
-1.90265477e-01 -6.59283757e-01 3.72095883e-01 -6.36010408e-01
-7.25027204e-01 -6.52612567e-01 1.46123135e+00 -1.33539844e+00
1.77390385e+00 -1.91504765e+00 -1.84134543e-01 -1.30632550e-01
5.64653695e-01 2.74450123e-01 6.09051645e-01 6.08444035e-01
2.71759808e-01 6.90328479e-01 1.93189174e-01 -1.48489207e-01
-5.33506945e-02 -5.01620304e-03 -4.45473850e-01 9.56166327e-01
-8.45604613e-02 1.88242733e-01 -9.30717528e-01 -7.71232367e-01
-2.51860112e-01 2.43994936e-01 -2.82560110e-01 1.92224890e-01
5.48480731e-03 4.15769480e-02 -3.28973174e-01 1.03592348e+00
6.70333326e-01 2.97598869e-01 -5.67256659e-02 4.04866785e-01
-4.36370075e-01 2.75156677e-01 -5.46018064e-01 9.16705668e-01
-8.24378282e-02 1.03056908e+00 2.56601930e-01 -3.36328059e-01
1.16957700e+00 3.82870436e-01 2.51216054e-01 -3.23012143e-01
7.14164257e-01 4.21516567e-01 1.96073920e-01 -8.06598127e-01
8.17126453e-01 -2.21188560e-01 -3.11482430e-01 7.56820977e-01
-5.68829551e-02 2.22078934e-01 1.75799727e-01 6.82642758e-01
9.24333036e-01 3.26028909e-03 5.14188886e-01 -6.25045896e-01
6.20776236e-01 -8.36993456e-02 2.65564293e-01 6.36492193e-01
-9.07838583e-01 3.65663111e-01 8.02856326e-01 -5.68269491e-01
-9.87976670e-01 -3.61526042e-01 -3.61225754e-01 1.87944603e+00
-2.19214544e-01 -4.93049681e-01 -8.13353539e-01 -1.19976795e+00
-3.91348377e-02 9.03184116e-01 -6.75182462e-01 -5.42507730e-02
-4.55071688e-01 -1.12978411e+00 9.24772859e-01 -2.97604233e-01
7.89267778e-01 -9.97150898e-01 -2.82507896e-01 7.12023154e-02
-2.18671724e-01 -8.19458544e-01 -3.33048940e-01 1.85905714e-02
-3.50533634e-01 -9.64309037e-01 -6.53307319e-01 -4.82552052e-01
9.84917358e-02 1.76946253e-01 5.54438233e-01 2.93945134e-01
-3.55436057e-01 -3.22946340e-01 -7.50817180e-01 -9.56652820e-01
-7.41364062e-01 1.82328254e-01 -2.46656425e-02 1.02767803e-01
8.76926959e-01 9.00784414e-03 -2.20254824e-01 -1.73624083e-02
-6.96962655e-01 -4.74845797e-01 -1.43565401e-01 2.99313456e-01
-4.96176988e-01 -3.56878102e-01 5.88460803e-01 -1.32574260e+00
9.19801533e-01 -1.25235236e+00 -2.96474900e-02 -2.31306404e-01
-4.96702850e-01 -6.28872156e-01 2.86518931e-01 -4.49124664e-01
-9.30616319e-01 -8.31785440e-01 -1.36641562e-01 5.93368448e-02
-1.43913656e-01 3.44688654e-01 9.46367607e-02 1.53713018e-01
1.16010988e+00 -5.22250891e-01 -2.52353221e-01 -7.13578761e-01
-1.69140488e-01 1.22366214e+00 -4.23729092e-01 -2.11848646e-01
6.17228687e-01 2.77834535e-01 -5.40738404e-01 -9.94465172e-01
-1.33399928e+00 -8.99468720e-01 -3.36000413e-01 -6.01018250e-01
1.11338031e+00 -8.64779294e-01 -3.39735359e-01 6.33087039e-01
-1.36337268e+00 1.15339793e-01 7.61579454e-01 1.27282754e-01
1.94206566e-01 4.76161420e-01 -1.24694443e+00 -1.29029763e+00
-6.76678240e-01 -6.68686211e-01 4.98294204e-01 -1.75771594e-01
-6.44823432e-01 -9.71896291e-01 4.07539606e-01 7.40569293e-01
2.69027144e-01 6.63298547e-01 8.00459743e-01 -1.15505040e+00
2.65009493e-01 -2.18551978e-01 -1.73793703e-01 4.44174409e-01
1.67232990e-01 4.26323503e-01 -9.74906743e-01 -1.71139032e-01
1.87905163e-01 -6.94184184e-01 5.29470503e-01 -9.72993672e-02
4.84883398e-01 -6.45857275e-01 -7.33317137e-02 -2.34931007e-01
1.36240482e+00 1.96325973e-01 4.96287525e-01 8.60298097e-01
8.17018867e-01 7.12995172e-01 6.82908058e-01 7.08864510e-01
2.21233219e-02 3.57746273e-01 4.03474867e-01 3.04280847e-01
9.40589458e-02 -2.27145940e-01 7.88995326e-01 7.95192659e-01
5.62114716e-02 -3.00224841e-01 -1.20547497e+00 4.46266919e-01
-1.68783879e+00 -1.00600576e+00 -7.57197678e-01 1.54029012e+00
8.84385645e-01 1.80921271e-01 7.19354033e-01 3.50705683e-01
1.03221965e+00 2.60164708e-01 1.01719782e-01 -1.15925741e+00
-3.46808106e-01 -3.38323265e-01 4.04750079e-01 8.20193291e-01
-1.23802888e+00 1.15452695e+00 5.60895109e+00 7.88286865e-01
-1.24740148e+00 8.22392046e-01 8.65087330e-01 -1.90051883e-01
1.07184902e-01 -1.98184550e-01 -1.08641934e+00 8.11696112e-01
8.85982215e-01 -3.95049863e-02 -8.98192897e-02 9.82078314e-01
4.50509876e-01 -2.17307463e-01 -1.76496387e-01 7.58990824e-01
7.17670679e-01 -7.73637831e-01 2.75708325e-02 2.19589442e-01
8.43330860e-01 7.89586753e-02 2.74127088e-02 4.41627145e-01
2.97326595e-01 -8.62035334e-01 7.51888752e-01 -4.93782312e-02
6.48392811e-02 -1.03863633e+00 9.56733942e-01 5.22325933e-01
4.09952961e-02 -2.66755223e-02 -5.30307889e-01 -3.94629568e-01
-3.05078682e-02 5.89991450e-01 -8.65678370e-01 -2.43862830e-02
9.90367711e-01 5.84474802e-01 -8.07827234e-01 6.36149287e-01
-4.26959664e-01 1.08271015e+00 1.06366843e-01 -5.91208935e-01
6.21254981e-01 -2.62707382e-01 8.47699583e-01 1.48822951e+00
5.69085032e-02 -1.03363022e-01 -1.31821856e-02 3.78199399e-01
-9.74960774e-02 7.72586346e-01 -9.36459720e-01 -2.30931848e-01
1.66059986e-01 1.42742431e+00 -6.79733098e-01 -1.69642940e-01
-3.05131078e-01 8.76509368e-01 5.39000034e-01 -1.66821346e-01
-1.09669280e+00 -4.16020334e-01 3.12258005e-01 5.25296867e-01
-4.92126852e-01 -4.68901657e-02 -3.29493701e-01 -8.89743328e-01
-6.17725253e-01 -1.35059988e+00 9.08745110e-01 -4.66459900e-01
-1.49219120e+00 5.16202748e-01 -9.98307914e-02 -6.22139871e-01
1.08396567e-01 -7.24676073e-01 -4.32407945e-01 6.36256099e-01
-9.62971509e-01 -1.24066663e+00 5.16926646e-02 4.98452336e-01
7.59203911e-01 -4.35334295e-01 3.14978302e-01 4.21381384e-01
-1.00746441e+00 3.52966726e-01 -4.10195917e-01 3.71550679e-01
1.19761395e+00 -1.21884310e+00 -3.57865840e-01 9.25675273e-01
-2.77095824e-01 5.83278537e-01 1.09913218e+00 -1.15154862e+00
-7.52459586e-01 -8.26027751e-01 1.24785912e+00 -9.55653906e-01
1.32937872e+00 -4.62745756e-01 -7.99992442e-01 7.68490434e-01
8.36715817e-01 -5.45045435e-01 1.07005501e+00 2.39748269e-01
-6.30707085e-01 3.86654824e-01 -1.47915614e+00 6.16285384e-01
7.52136350e-01 -3.60652357e-01 -6.50144339e-01 1.12015319e+00
4.90024298e-01 6.88156709e-02 -4.12816495e-01 -3.49366307e-01
1.16342790e-01 -9.92359340e-01 1.92903057e-01 -1.08913600e+00
1.08324051e+00 1.52807787e-01 -2.71528155e-01 -9.56353068e-01
-3.82123441e-01 -9.77984786e-01 -1.45540386e-01 1.50359321e+00
4.24823612e-01 -4.62436348e-01 5.30004621e-01 3.51554871e-01
5.54127395e-02 -2.59126127e-01 -6.12837255e-01 -3.92883867e-01
6.79335713e-01 -3.37007314e-01 -6.37023570e-03 1.62152970e+00
2.44438380e-01 7.52479494e-01 -8.43422115e-01 -4.91349958e-02
4.43423927e-01 -4.58768904e-01 6.82376504e-01 -9.75259781e-01
9.07648429e-02 -3.54880869e-01 -2.28707939e-01 -1.86586931e-01
3.44767310e-02 -8.54331851e-01 -5.58058202e-01 -1.04945087e+00
5.92975199e-01 3.61443982e-02 -1.29269632e-02 3.14117372e-01
-3.26292992e-01 7.50904202e-01 4.17612344e-01 2.95467198e-01
-6.53427660e-01 -1.71041608e-01 9.65606868e-01 -2.06227973e-01
-1.49013251e-01 -4.96711493e-01 -9.37937856e-01 8.89550149e-01
9.75052357e-01 -7.02741385e-01 1.25062004e-01 -1.23323582e-01
8.10214400e-01 -5.78355253e-01 -7.31972221e-04 -5.17222106e-01
-3.36863846e-01 -1.86360002e-01 1.13009624e-01 -6.00290477e-01
-7.24502727e-02 -3.95532787e-01 -5.65121949e-01 6.77625299e-01
-2.97622681e-01 3.79772276e-01 -1.61049634e-01 3.07334084e-02
-1.63597632e-02 -7.76378393e-01 1.12834895e+00 -3.00821930e-01
-3.45976681e-01 8.18382669e-03 -9.45110261e-01 1.88870817e-01
1.29988563e+00 1.23314979e-02 -7.63493240e-01 -7.21646845e-01
-6.01204395e-01 -1.07161462e-01 7.14983568e-02 5.36983073e-01
2.42598936e-01 -9.96761024e-01 -1.34184682e+00 -4.53785449e-01
7.17224926e-02 -9.40813363e-01 -1.25825197e-01 9.51485217e-01
-9.34834659e-01 1.28524661e-01 -2.19353259e-01 -9.69129428e-02
-1.59522784e+00 5.63067138e-01 1.58694401e-01 -1.80164859e-01
-9.40945074e-02 9.01822090e-01 -2.18964055e-01 -2.26832464e-01
-9.99382231e-03 6.66994154e-01 -6.12601340e-01 7.71753132e-01
8.34817052e-01 9.99256313e-01 -1.12463683e-01 -1.13036203e+00
-2.50747710e-01 -4.85812388e-02 -7.54644454e-01 4.88346443e-03
8.62890065e-01 -1.91629305e-01 -7.16892302e-01 8.03548098e-01
1.33912051e+00 7.62345970e-01 -2.81197339e-01 1.34205863e-01
5.06813943e-01 -7.53584266e-01 1.61812812e-01 -9.60351110e-01
-6.02608860e-01 5.15720069e-01 3.59723389e-01 9.66820836e-01
3.63648951e-01 -6.78581521e-02 8.05120647e-01 -3.16599198e-02
2.23702475e-01 -1.55487478e+00 5.50940514e-01 6.92899823e-01
1.25998509e+00 -1.47885346e+00 2.63189375e-01 -4.86578584e-01
-7.00003326e-01 9.95401919e-01 8.82177889e-01 -4.57620472e-01
5.61167419e-01 1.31793752e-01 6.32648647e-01 -3.78876776e-01
-4.10833180e-01 2.86673665e-01 5.77267725e-04 3.72315794e-01
1.09104180e+00 1.20779559e-01 -1.19074154e+00 3.94774675e-01
-4.46162939e-01 -7.89837837e-01 8.49009991e-01 9.18161750e-01
-7.56053686e-01 -7.75657296e-01 -8.82881224e-01 3.65472645e-01
-1.44084024e+00 1.06648132e-02 -1.36893058e+00 1.12172925e+00
2.62088537e-01 1.23225391e+00 -2.91832745e-01 -3.59654725e-01
-1.02686003e-01 1.06214404e-01 1.01836137e-01 -7.69205034e-01
-1.35841596e+00 1.29194424e-01 6.07250869e-01 -1.20093301e-01
-1.77569315e-01 -6.68183446e-01 -7.96273410e-01 -8.62408340e-01
-4.40135658e-01 1.01077810e-01 6.64975703e-01 8.76926482e-01
-2.25964934e-01 -5.34037985e-02 7.03190148e-01 -2.19424129e-01
-5.60678542e-01 -1.58342171e+00 -7.53177047e-01 7.90892005e-01
3.23058635e-01 -4.16468143e-01 -5.64706028e-01 -4.66298819e-01] | [8.827747344970703, 10.554503440856934] |
6072ef16-4bba-4b02-af8a-1400e96bf5ed | continuous-implicit-authentication-for-mobile | 1705.06715 | null | http://arxiv.org/abs/1705.06715v1 | http://arxiv.org/pdf/1705.06715v1.pdf | Continuous Implicit Authentication for Mobile Devices based on Adaptive Neuro-Fuzzy Inference System | As mobile devices have become indispensable in modern life, mobile security
is becoming much more important. Traditional password or PIN-like
point-of-entry security measures score low on usability and are vulnerable to
brute force and other types of attacks. In order to improve mobile security, an
adaptive neuro-fuzzy inference system(ANFIS)-based implicit authentication
system is proposed in this paper to provide authentication in a continuous and
transparent manner.To illustrate the applicability and capability of ANFIS in
our implicit authentication system, experiments were conducted on behavioural
data collected for up to 12 weeks from different Android users. The ability of
the ANFIS-based system to detect an adversary is also tested with scenarios
involving an attacker with varying levels of knowledge. The results demonstrate
that ANFIS is a feasible and efficient approach for implicit authentication
with an average of 95% user recognition rate. Moreover, the use of ANFIS-based
system for implicit authentication significantly reduces manual tuning and
configuration tasks due to its selflearning capability. | ['Suleiman Y. Yerima', 'Sakir Sezer', 'Feng Yao', 'BooJoong Kang'] | 2017-05-18 | null | null | null | null | ['mobile-security'] | ['miscellaneous'] | [ 4.30509374e-02 -1.53500766e-01 6.88524991e-02 8.06814805e-02
2.18116418e-01 -6.80967987e-01 4.09240544e-01 4.46669191e-01
-7.91725338e-01 7.09308565e-01 -5.91034412e-01 -7.49310672e-01
-5.39392650e-01 -6.23671651e-01 -2.42326204e-02 -3.53315651e-01
-2.31157154e-01 -6.27538040e-02 2.11827844e-01 -4.31192786e-01
6.54371500e-01 7.43274271e-01 -1.52719343e+00 5.63338101e-02
8.18679452e-01 7.62019932e-01 -1.39053628e-01 7.72652149e-01
4.04318720e-01 4.07856852e-01 -1.06795549e+00 -2.69415170e-01
5.75914532e-02 1.86527655e-01 -5.51448524e-01 -4.53191519e-01
-3.85363102e-01 -5.68696141e-01 2.08332902e-03 7.91096807e-01
5.15504122e-01 2.66135395e-01 4.63597834e-01 -1.18792367e+00
-3.28584462e-01 4.46974039e-02 -1.16965920e-01 2.45213494e-01
8.55075300e-01 1.10424250e-01 1.70513600e-01 -7.74890840e-01
-3.45081091e-01 6.85456932e-01 1.06682694e+00 5.02579451e-01
-1.19541490e+00 -8.77375245e-01 -2.24043235e-01 2.28308305e-01
-1.73535407e+00 -2.16573432e-01 4.33082163e-01 -2.34967902e-01
1.18460286e+00 4.29289788e-01 6.19431138e-01 6.82350278e-01
2.09621429e-01 1.40726715e-01 1.09721971e+00 -7.00094104e-01
3.39170933e-01 6.48694515e-01 4.91012216e-01 3.48656267e-01
9.04384136e-01 1.65146798e-01 -6.96519464e-02 -4.51852232e-01
6.19651079e-01 3.64488393e-01 -1.28366292e-01 5.08133359e-02
-6.38929129e-01 7.32709289e-01 2.99083646e-02 6.92836881e-01
-4.90499705e-01 -3.11042070e-01 2.35687912e-01 1.71599075e-01
-3.57235134e-01 7.04150200e-01 -5.60244381e-01 -5.95823407e-01
-5.41388333e-01 -1.20821665e-03 1.19972694e+00 2.22065419e-01
3.06827903e-01 4.47682023e-01 5.70086896e-01 4.90776300e-01
5.17440021e-01 6.24788284e-01 8.46753001e-01 -6.37135506e-01
-8.33347291e-02 8.71504724e-01 4.18398827e-01 -1.24646699e+00
-3.60467464e-01 -2.26312459e-01 -7.11835563e-01 6.51778638e-01
4.31096643e-01 -2.36983255e-01 -7.18133390e-01 1.45196569e+00
1.30677119e-01 4.94216740e-01 1.60237119e-01 3.28985870e-01
2.83345640e-01 4.92432445e-01 3.75532418e-01 -2.29203925e-01
1.36960149e+00 5.13100252e-02 -9.91472542e-01 2.34653071e-01
2.75750160e-01 -4.38455403e-01 1.18253577e+00 6.48411274e-01
-7.60606170e-01 -6.00263774e-01 -1.40468085e+00 6.50625408e-01
-7.22848713e-01 6.31886795e-02 4.80334997e-01 1.54163551e+00
-9.07827377e-01 4.37292516e-01 -8.32599878e-01 -3.55797470e-01
-1.69450551e-01 1.26636541e+00 -5.14944673e-01 8.54891956e-01
-1.58592474e+00 1.00992393e+00 4.71769750e-01 2.14246824e-01
-1.74946725e-01 -1.88648179e-01 -8.61460030e-01 6.19888343e-02
5.01258448e-02 -4.68098298e-02 1.00780666e+00 -8.27153862e-01
-1.57630038e+00 1.92068443e-01 7.24396855e-02 -7.66761839e-01
-2.60981265e-02 -1.78626537e-01 -8.01715970e-01 3.90515476e-01
-3.61861348e-01 -9.62329432e-02 8.29301715e-01 -9.85478282e-01
-2.99858063e-01 -2.59813249e-01 2.47054368e-01 3.93435694e-02
-8.26847076e-01 -3.08345705e-01 2.04949766e-01 -4.02941763e-01
-3.33664447e-01 -9.72150922e-01 -5.60658164e-02 -5.23318827e-01
3.60032588e-01 1.63656220e-01 1.25651562e+00 -9.48642313e-01
1.94191051e+00 -2.02808404e+00 -7.01076090e-01 9.89890993e-01
4.34892066e-02 1.43650043e+00 4.17760491e-01 4.76580739e-01
-1.67986972e-03 3.52623433e-01 -6.51280582e-02 3.11255425e-01
-2.01892868e-01 2.97770590e-01 -6.87150005e-03 1.04506582e-01
8.19970621e-04 4.79175657e-01 -6.01595104e-01 2.58331615e-02
4.97145563e-01 8.79073262e-01 -3.13111246e-01 -2.77174637e-02
7.66035080e-01 1.04577370e-01 -5.39793670e-01 5.27923524e-01
2.69099176e-01 4.22280803e-02 5.02745688e-01 1.10067971e-01
-7.28336647e-02 9.83632654e-02 -1.31268108e+00 6.99294329e-01
-4.01710421e-01 2.70880610e-01 -7.71372160e-03 -8.42702031e-01
8.50068271e-01 1.00575292e+00 2.23503694e-01 -8.00690234e-01
4.38733727e-01 1.01596676e-01 2.83815086e-01 -5.33959448e-01
3.30204576e-01 2.40094796e-01 -3.56357284e-02 5.53602576e-01
-1.14804387e-01 3.23525012e-01 -6.44720718e-02 1.00714669e-01
9.11929190e-01 -2.36885875e-01 8.97809446e-01 -2.54376411e-01
1.26628423e+00 -5.58327198e-01 1.46908864e-01 7.26065636e-01
-3.72154266e-01 -4.03455585e-01 -1.76131979e-01 -4.04750615e-01
-4.75630999e-01 -7.72566140e-01 1.22114621e-01 3.77368480e-01
-1.66610386e-02 -4.36234236e-01 -1.09834683e+00 -4.64114696e-01
-1.62506700e-02 6.08495355e-01 -4.50806350e-01 -3.82503420e-01
-2.76160389e-01 -2.19858825e-01 9.70780551e-01 5.49451709e-01
6.49489462e-01 -1.17652977e+00 -1.04681861e+00 3.46345067e-01
7.04675019e-02 -8.05929005e-01 3.06977108e-02 -9.84492451e-02
-1.01286519e+00 -9.42309022e-01 -3.05735707e-01 -5.90213835e-01
6.90483689e-01 3.37880433e-01 3.71228486e-01 6.82010233e-01
-3.43860015e-02 7.76325345e-01 -1.43628046e-01 -6.56807899e-01
-4.03586537e-01 4.97149788e-02 5.77887595e-01 5.60535155e-02
7.16842711e-01 -5.30046880e-01 -6.64080203e-01 6.23321712e-01
-8.56488585e-01 -7.97999561e-01 3.86956990e-01 7.50611126e-01
8.54233373e-03 6.38902724e-01 1.02953041e+00 -7.04491496e-01
1.38440812e+00 -3.70896757e-01 -4.50387686e-01 1.58253610e-01
-1.14928830e+00 -1.85939118e-01 4.97363329e-01 -9.76579666e-01
-1.02442789e+00 3.98275703e-02 -2.75055677e-01 1.88753098e-01
-3.55722964e-01 7.90962279e-01 -1.66746318e-01 -7.31585622e-01
8.94068062e-01 1.33295313e-01 4.45809901e-01 -1.81324527e-01
-2.10939482e-01 1.15975404e+00 3.95265669e-01 -3.96934003e-01
9.15341079e-01 8.62034261e-02 -2.02245504e-01 -1.26874387e+00
-8.36573541e-02 -4.36022282e-01 -3.83123189e-01 -3.01777333e-01
5.75070739e-01 -6.01999104e-01 -1.70307481e+00 6.97187424e-01
-7.05733657e-01 -1.43863380e-01 3.05325240e-01 4.30013061e-01
1.25685528e-01 7.19839752e-01 -6.09618306e-01 -1.56172681e+00
-7.00609505e-01 -9.21915650e-01 1.65179119e-01 6.58459067e-01
-8.46493363e-01 -1.09386277e+00 -9.22967568e-02 3.44770491e-01
6.63557053e-01 -2.51833946e-02 4.82348263e-01 -8.30020905e-01
1.62625551e-01 -8.98040533e-01 4.49543327e-01 4.70956147e-01
4.99073356e-01 -5.56525625e-02 -8.68555427e-01 -3.56677800e-01
4.13703084e-01 5.18615060e-02 -2.65537381e-01 -1.55801687e-03
6.49964690e-01 -6.85500622e-01 -7.51586407e-02 -9.43883285e-02
1.27209055e+00 1.08166432e+00 9.42022860e-01 6.80470586e-01
2.26679325e-01 4.60077167e-01 6.53265119e-01 3.84453923e-01
1.27464712e-01 4.61014718e-01 2.33195931e-01 -1.66823119e-01
7.85421252e-01 1.48241892e-01 4.30975378e-01 3.11338365e-01
-6.43798053e-01 8.96377340e-02 -7.95303643e-01 8.81836861e-02
-1.46022117e+00 -1.19263434e+00 -6.32895976e-02 2.75758243e+00
6.40888870e-01 3.95035863e-01 3.55102956e-01 1.31597650e+00
6.51588142e-01 -8.85133445e-01 -8.93739313e-02 -9.08826351e-01
4.69832182e-01 8.63126576e-01 2.70158231e-01 5.58244646e-01
-9.80671048e-01 4.49100852e-01 5.87453508e+00 3.31211179e-01
-1.39655817e+00 -1.75213307e-01 1.51167110e-01 5.50773561e-01
3.77092302e-01 -1.40721917e-01 -4.15548921e-01 5.01338720e-01
1.34563076e+00 -7.34731406e-02 4.78662342e-01 7.06763089e-01
2.66605854e-01 -3.30590278e-01 -2.22508460e-01 8.39704216e-01
-2.04802513e-01 -8.32898617e-01 -4.22500186e-02 2.69898444e-01
1.38359651e-01 -8.45764220e-01 9.87230241e-02 2.63156503e-01
-4.13993120e-01 -8.15622687e-01 2.57048100e-01 4.37851578e-01
6.10393405e-01 -1.07150102e+00 9.95158494e-01 3.86141837e-01
-1.24642015e+00 -3.48903835e-01 2.73003846e-01 -7.04702079e-01
-4.45780940e-02 -2.95699269e-01 -9.75754857e-01 2.40610838e-01
5.32612324e-01 -7.45338947e-02 -5.83165169e-01 8.91844630e-01
-1.94062397e-01 8.53690445e-01 -4.42352921e-01 -3.48102808e-01
-8.86573046e-02 -6.64062276e-02 9.30258036e-02 1.06400025e+00
3.55970681e-01 3.91337991e-01 -2.79269069e-01 5.90783022e-02
7.63357401e-01 -5.35880439e-02 -7.28709400e-01 -4.10318047e-01
6.58410907e-01 1.32359433e+00 -8.77820492e-01 -2.30242670e-01
-2.55340010e-01 7.35571444e-01 -3.90472561e-01 3.72406691e-01
-6.03482962e-01 -9.05543864e-01 4.36672688e-01 3.92447203e-01
-1.77655518e-02 -6.15075648e-01 -2.34848306e-01 -8.12730253e-01
-4.05896664e-01 -1.21819699e+00 4.00536120e-01 -4.95116800e-01
-6.44531488e-01 4.71329451e-01 -1.37925893e-01 -1.13827217e+00
-4.90329444e-01 -6.65722251e-01 -4.60350841e-01 1.08058321e+00
-7.31672287e-01 -8.45679939e-01 -2.33101100e-01 9.85905468e-01
-2.33140253e-02 -3.14666122e-01 1.49626243e+00 3.48093629e-01
-1.77907974e-01 9.02507067e-01 -5.30692577e-01 2.67286777e-01
1.79963380e-01 -9.32786167e-01 1.32778510e-01 9.62285459e-01
-1.69270113e-01 1.76317430e+00 6.15431190e-01 -7.64807642e-01
-1.24411559e+00 -2.60760546e-01 6.10003471e-01 -3.65551323e-01
4.92585063e-01 -1.82231978e-01 -1.16018820e+00 3.39424908e-01
1.64556220e-01 -6.26426160e-01 1.13747549e+00 -1.90565288e-01
-1.40602076e-02 -6.19339347e-02 -1.53554833e+00 7.59918690e-01
3.49753276e-02 -7.90485501e-01 -6.34467483e-01 -6.72779024e-01
-1.16203139e-02 -3.56010310e-02 -9.45910096e-01 5.17467737e-01
9.78926539e-01 -8.05195689e-01 1.17156649e+00 -1.34939045e-01
-6.94041669e-01 -5.01488388e-01 2.99053878e-01 -5.11767924e-01
-2.16090024e-01 -1.00003600e+00 -4.11105961e-01 1.21236825e+00
4.54008013e-01 -1.28292298e+00 5.96280515e-01 1.03190982e+00
5.84064484e-01 -5.49770653e-01 -4.85342741e-01 -4.63487178e-01
-8.08122635e-01 -2.93169260e-01 3.66358161e-01 9.08109426e-01
4.29835647e-01 4.12277073e-01 -2.76012987e-01 3.93345922e-01
2.14766636e-01 -1.04583967e+00 7.15050459e-01 -1.44618261e+00
-2.66465515e-01 -1.57235190e-01 -8.24836910e-01 -1.56973124e-01
-2.91970462e-01 -2.19988570e-01 -4.74399388e-01 -9.75526512e-01
-4.50606525e-01 -3.54195356e-01 -7.73794532e-01 7.56460965e-01
-5.88495173e-02 5.31688988e-01 -6.90042824e-02 -2.22831052e-02
-4.97052912e-03 -1.20590150e-01 2.76123852e-01 4.86613847e-02
-5.96601367e-01 6.49306476e-01 -7.38379240e-01 6.82394028e-01
1.05180812e+00 -2.32303441e-01 -9.50780272e-01 4.32352692e-01
3.08572829e-01 4.52982560e-02 6.83229342e-02 -1.26225662e+00
3.80025804e-02 1.62627831e-01 5.24971366e-01 -8.21273848e-02
4.94810432e-01 -1.27481365e+00 3.67182195e-01 1.05788851e+00
2.42928848e-01 4.33020055e-01 5.88189304e-01 3.36633742e-01
-6.55311719e-02 -4.15807873e-01 2.92845160e-01 3.45138401e-01
-5.40374279e-01 -2.51376331e-01 -1.20075476e+00 -9.14290011e-01
1.16399777e+00 -1.04211473e+00 3.21533419e-02 -6.06078744e-01
-8.27844739e-01 -2.61390626e-01 3.42094928e-01 1.15109690e-01
8.45506966e-01 -7.20599234e-01 2.68248975e-01 6.36203766e-01
-2.71724254e-01 -8.95529151e-01 1.31478891e-01 7.63219535e-01
-7.37456679e-01 3.71896029e-01 -5.86842954e-01 -2.41813928e-01
-1.90471375e+00 4.72055137e-01 2.88711786e-01 -1.38512298e-01
-1.69241637e-01 1.74009800e-01 -5.40590823e-01 -1.19854413e-01
2.21376896e-01 8.91762152e-02 -8.11843812e-01 -1.80634350e-01
8.69164884e-01 4.50724959e-01 1.43566728e-01 -5.42719185e-01
-5.30591249e-01 3.50946695e-01 -4.70941737e-02 -5.61940551e-01
7.91223109e-01 5.42926900e-02 1.00418188e-01 2.20075756e-01
2.78844625e-01 3.71767402e-01 -7.25885570e-01 3.08582932e-01
2.94847488e-01 -2.04248860e-01 -3.22974116e-01 -9.91406024e-01
-4.91527617e-01 7.95095384e-01 1.01623356e+00 6.50965691e-01
1.25401711e+00 -1.21909785e+00 4.23280180e-01 7.16523767e-01
4.45174187e-01 -1.04124367e+00 -3.74910682e-02 4.72830296e-01
2.56936491e-01 -8.34954739e-01 -2.50300188e-02 1.67499483e-02
-6.45593762e-01 1.31833899e+00 4.41497713e-01 -6.30388185e-02
9.98035312e-01 4.81985897e-01 2.59677857e-01 1.20335072e-01
-1.70130685e-01 1.26455471e-01 3.33391428e-01 9.86000359e-01
3.87119055e-01 5.74382581e-03 -7.95634389e-01 9.09484684e-01
6.62721917e-02 2.64654756e-01 5.68082929e-01 1.38929927e+00
-3.46859574e-01 -1.31931198e+00 -4.36818868e-01 1.75642997e-01
-1.01546192e+00 2.66916484e-01 -4.08938259e-01 1.05538821e+00
-1.56887710e-01 1.34271145e+00 -4.65933651e-01 -8.43038261e-01
1.98763847e-01 3.74981426e-02 2.08504334e-01 -3.16214383e-01
-9.54014063e-01 -1.70967773e-01 -4.21767123e-02 -2.50878602e-01
-3.50784063e-01 -3.35780859e-01 -1.55452895e+00 -4.10017848e-01
-6.43371403e-01 6.81540549e-01 1.13975418e+00 1.03545117e+00
3.52217555e-01 2.43183658e-01 3.17282170e-01 -6.16002738e-01
-3.63843590e-01 -9.03426230e-01 -2.65645891e-01 2.17447668e-01
-6.53614476e-02 -7.18509793e-01 -2.13056058e-01 -3.44615504e-02] | [13.542731285095215, 1.3425278663635254] |
7dae8d8e-535c-448e-b785-695356856316 | sgl-pt-a-strong-graph-learner-with-graph | 2302.12449 | null | https://arxiv.org/abs/2302.12449v1 | https://arxiv.org/pdf/2302.12449v1.pdf | SGL-PT: A Strong Graph Learner with Graph Prompt Tuning | Recently, much exertion has been paid to design graph self-supervised methods to obtain generalized pre-trained models, and adapt pre-trained models onto downstream tasks through fine-tuning. However, there exists an inherent gap between pretext and downstream graph tasks, which insufficiently exerts the ability of pre-trained models and even leads to negative transfer. Meanwhile, prompt tuning has seen emerging success in natural language processing by aligning pre-training and fine-tuning with consistent training objectives. In this paper, we identify the challenges for graph prompt tuning: The first is the lack of a strong and universal pre-training task across sundry pre-training methods in graph domain. The second challenge lies in the difficulty of designing a consistent training objective for both pre-training and downstream tasks. To overcome above obstacles, we propose a novel framework named SGL-PT which follows the learning strategy ``Pre-train, Prompt, and Predict''. Specifically, we raise a strong and universal pre-training task coined as SGL that acquires the complementary merits of generative and contrastive self-supervised graph learning. And aiming for graph classification task, we unify pre-training and fine-tuning by designing a novel verbalizer-free prompting function, which reformulates the downstream task in a similar format as pretext task. Empirical results show that our method surpasses other baselines under unsupervised setting, and our prompt tuning method can greatly facilitate models on biological datasets over fine-tuning methods. | ['Siliang Tang', 'Jianhao Guo', 'Yun Zhu'] | 2023-02-24 | null | null | null | null | ['graph-classification'] | ['graphs'] | [ 4.93222386e-01 4.34663922e-01 -2.81338602e-01 -5.03618300e-01
-4.24046904e-01 -5.47842741e-01 7.87630260e-01 8.09873790e-02
-2.82172054e-01 5.99512815e-01 3.53278726e-01 -3.93364668e-01
-5.49551323e-02 -9.60036039e-01 -9.04169559e-01 -4.76704895e-01
1.93519622e-01 3.85285020e-01 2.49892268e-02 -5.34182310e-01
-3.25011574e-02 -2.94585377e-02 -8.38045299e-01 2.65196443e-01
1.22055113e+00 5.30473292e-01 4.20306206e-01 4.89239931e-01
-3.11697245e-01 6.39666796e-01 -4.23988909e-01 -6.53943956e-01
4.09145327e-03 -8.15779388e-01 -1.02791131e+00 -2.36721113e-02
3.09511095e-01 -7.41367862e-02 -2.72605330e-01 9.79427338e-01
5.74136913e-01 1.16286755e-01 8.38090658e-01 -1.32226777e+00
-1.39883184e+00 9.51031923e-01 -3.68638366e-01 5.04735149e-02
1.35760084e-01 4.06979471e-01 1.16999352e+00 -5.44521034e-01
5.01551807e-01 1.20048404e+00 9.45916116e-01 9.25237834e-01
-1.40752256e+00 -4.49608296e-01 4.81579185e-01 -8.59958827e-02
-1.22963667e+00 -3.05990934e-01 9.31320369e-01 -4.23236132e-01
8.81897867e-01 -7.88074508e-02 6.79179430e-01 1.55703712e+00
2.02901095e-01 6.90659881e-01 1.07850742e+00 -3.54213715e-01
-1.72652956e-02 -2.62543540e-02 9.74785537e-03 9.02327478e-01
1.32452816e-01 2.71599114e-01 -4.90642041e-01 4.93698753e-02
9.39922631e-01 -1.48050308e-01 -3.24498385e-01 -1.79745823e-01
-1.09978449e+00 8.04033458e-01 5.95188200e-01 2.34725550e-01
-1.24119706e-01 2.65402466e-01 4.58926350e-01 4.52493608e-01
6.40103281e-01 4.56786156e-01 -6.63492680e-01 1.47145391e-01
-5.92255056e-01 -6.73566610e-02 5.43965995e-01 1.16283536e+00
8.94290507e-01 2.17301324e-01 -5.00296116e-01 7.19668210e-01
5.45324028e-01 2.56968230e-01 5.99848866e-01 -3.61287773e-01
6.63325965e-01 7.32714653e-01 -6.29125834e-01 -7.17583418e-01
-2.67480761e-01 -9.05946910e-01 -1.13431966e+00 -2.08688855e-01
3.64579529e-01 -2.92533308e-01 -1.02173436e+00 2.33833599e+00
2.53120899e-01 2.06102952e-01 1.88284367e-01 6.69633031e-01
1.02182126e+00 5.67863107e-01 5.85664630e-01 -1.60160214e-01
1.38783085e+00 -1.27566981e+00 -6.83140218e-01 -5.56215703e-01
7.74454296e-01 -3.09421390e-01 1.68895805e+00 8.04550201e-02
-7.13731766e-01 -8.39691460e-01 -1.01900196e+00 -1.84294328e-01
-4.05733854e-01 -1.40824124e-01 8.39894891e-01 4.89614725e-01
-1.22795665e+00 7.79502273e-01 -5.47051728e-01 -4.28318411e-01
4.30947900e-01 3.15834790e-01 -3.93014461e-01 2.24506054e-02
-1.33827651e+00 7.88780391e-01 6.20478988e-01 -3.92512046e-02
-9.26914632e-01 -8.31485689e-01 -8.44064713e-01 5.30790649e-02
5.16626060e-01 -1.33813703e+00 1.05082285e+00 -9.84204948e-01
-1.83362496e+00 9.24462557e-01 1.79242507e-01 -2.28991210e-01
4.10814852e-01 -1.44852936e-01 -2.50582755e-01 -3.87188017e-01
8.35678130e-02 5.79463959e-01 1.09531748e+00 -1.15306127e+00
-2.96015497e-02 -3.06457698e-01 8.95298272e-03 3.42949152e-01
-5.76578021e-01 -3.86880517e-01 -4.36742634e-01 -8.39987814e-01
-2.98190743e-01 -6.46324933e-01 -4.02929515e-01 -5.67494869e-01
-5.37102759e-01 -5.47359884e-01 6.01245999e-01 -1.76671371e-01
1.39211535e+00 -2.01567888e+00 1.52147308e-01 -6.94443658e-02
5.64824879e-01 3.93061280e-01 -5.68311214e-01 7.02864707e-01
-1.86375722e-01 3.93225878e-01 -3.38218689e-01 -3.76945019e-01
1.28119171e-01 4.67267781e-01 -3.85708094e-01 7.20213428e-02
4.52006727e-01 1.53184927e+00 -1.28407872e+00 -5.92702091e-01
-1.91588625e-01 2.33755410e-01 -6.92001820e-01 6.17376029e-01
-5.18042743e-01 6.71594799e-01 -5.96825898e-01 1.98614433e-01
4.42080647e-01 -6.63941681e-01 3.00502807e-01 -2.78799355e-01
2.63172299e-01 4.22391653e-01 -7.80566692e-01 2.00179744e+00
-3.47261816e-01 -3.75307873e-02 -1.96307287e-01 -1.37030113e+00
1.03096485e+00 2.09863544e-01 3.26394588e-02 -7.29841411e-01
5.89925684e-02 -7.56661147e-02 -1.49911687e-01 -6.04423940e-01
2.18876213e-01 -3.81480604e-01 -1.61086038e-01 5.33650219e-01
5.78638077e-01 -3.28857660e-01 1.07655048e-01 4.07774836e-01
1.10037529e+00 5.60721576e-01 4.35646504e-01 -2.33892813e-01
4.10179287e-01 -1.81131557e-01 4.49912578e-01 6.33871436e-01
-2.16660902e-01 4.08092886e-01 6.85520291e-01 -1.17961660e-01
-6.54235363e-01 -9.99262989e-01 4.41905230e-01 1.68470907e+00
-3.66743244e-02 -8.68374050e-01 -8.40835929e-01 -1.25120890e+00
-1.04870960e-01 5.88545144e-01 -8.62486780e-01 -5.55463493e-01
-4.71913546e-01 -8.53221178e-01 7.33849704e-01 6.78245544e-01
6.46601439e-01 -1.17921066e+00 2.26519838e-01 1.97318420e-01
2.77248621e-02 -1.04082847e+00 -8.67658496e-01 4.90452468e-01
-1.00214350e+00 -7.93531239e-01 -4.04028058e-01 -1.11917591e+00
8.57400119e-01 6.90212473e-02 1.44285262e+00 2.82318711e-01
3.09809923e-01 1.05162479e-01 -3.51056755e-01 -1.74015164e-01
-5.66425800e-01 4.74751860e-01 -1.36749417e-01 -1.20465055e-01
6.99536130e-02 -7.89928138e-01 -4.01665688e-01 1.37777403e-01
-9.10109043e-01 5.43878555e-01 8.18756700e-01 9.22155499e-01
5.42878211e-01 -3.25702906e-01 1.07721972e+00 -1.44929063e+00
1.10743153e+00 -4.72608328e-01 -3.55090797e-01 4.59421605e-01
-1.18900144e+00 5.29251158e-01 1.00265789e+00 -4.11310047e-01
-9.73902643e-01 -7.06607699e-02 -2.49187395e-01 -2.64943033e-01
-1.26527160e-01 7.81245053e-01 -3.63014072e-01 6.26261979e-02
9.23935235e-01 1.58046812e-01 9.30989999e-03 -3.62164706e-01
9.54449832e-01 2.88246214e-01 6.23819411e-01 -9.00361836e-01
1.20151997e+00 -3.81566510e-02 -1.09228484e-01 -3.06892961e-01
-1.20508659e+00 -1.79984823e-01 -6.14834845e-01 4.28184904e-02
9.04409707e-01 -8.55360150e-01 -5.16463339e-01 3.48823220e-01
-1.06332052e+00 -9.07845140e-01 -3.75653833e-01 8.76503363e-02
-5.21047354e-01 4.58221525e-01 -6.02001369e-01 -3.26497316e-01
-5.70878625e-01 -8.95294785e-01 9.63233173e-01 9.30323005e-02
-1.70184895e-01 -1.52408218e+00 2.72744924e-01 2.88965732e-01
6.91718757e-01 1.86473519e-01 1.05257308e+00 -7.08577335e-01
-2.94934839e-01 3.35384935e-01 -3.39963943e-01 3.29612970e-01
3.15412074e-01 -3.56821239e-01 -8.85066390e-01 -1.25412717e-01
-1.78435326e-01 -6.93161905e-01 8.60385299e-01 7.03684241e-02
1.05798614e+00 -5.01236975e-01 -2.99202204e-01 9.32411611e-01
1.29732203e+00 -2.55797327e-01 3.89540792e-01 1.07222959e-01
1.09965944e+00 5.14415324e-01 2.25328639e-01 -1.04410872e-01
6.58694029e-01 4.21875656e-01 2.13330820e-01 -3.17815483e-01
-2.86401927e-01 -9.73551810e-01 4.38533336e-01 1.03638422e+00
-9.78930965e-02 -4.97383028e-01 -7.46302128e-01 3.25247139e-01
-1.98683810e+00 -7.44600296e-01 -6.47413209e-02 1.88623047e+00
1.28670800e+00 6.17158525e-02 9.44190249e-02 -2.25839660e-01
5.65221727e-01 2.74834841e-01 -5.45328081e-01 -2.55362451e-01
-2.05712888e-04 2.79021561e-01 2.76910454e-01 5.73726416e-01
-9.43318427e-01 1.57237542e+00 5.93625021e+00 9.90314305e-01
-1.25534904e+00 2.03865454e-01 6.08803153e-01 5.53804457e-01
-6.43329859e-01 1.67166904e-01 -7.53104150e-01 2.60666519e-01
9.96330678e-01 -3.05822670e-01 4.90381628e-01 6.98775768e-01
1.77923918e-01 6.54689431e-01 -1.32223618e+00 8.06582808e-01
1.12588368e-01 -1.19245887e+00 3.32714319e-01 -8.58683661e-02
6.36593461e-01 3.36256139e-02 -4.99367155e-02 9.21253204e-01
6.26365244e-01 -1.14673209e+00 5.53797364e-01 1.32212535e-01
9.63916361e-01 -1.65963411e-01 3.48564535e-01 4.99581099e-01
-1.19400966e+00 2.68983841e-01 -2.66613960e-01 -2.12018058e-01
1.40235052e-01 4.66968149e-01 -1.02725601e+00 7.95918941e-01
1.97350442e-01 8.18793774e-01 -7.19456613e-01 5.12436330e-01
-9.10141289e-01 8.74335766e-01 1.19311988e-01 -7.16859326e-02
1.93085477e-01 -3.12772542e-01 1.56926110e-01 1.36739933e+00
5.53841889e-02 1.62745327e-01 4.75470781e-01 1.01914585e+00
-3.57515544e-01 2.58476406e-01 -6.89776659e-01 -5.05873203e-01
2.46054605e-01 1.35541356e+00 -5.49072742e-01 -4.72337633e-01
-3.05567205e-01 1.00570714e+00 8.89262855e-01 6.23298168e-01
-8.83286536e-01 -1.90450817e-01 1.27669409e-01 1.80227876e-01
-1.69372167e-02 -1.94457158e-01 -3.88601214e-01 -1.30500638e+00
-3.22295040e-01 -8.82381916e-01 4.83874977e-01 -7.25260496e-01
-1.58853757e+00 6.38931751e-01 -3.86310704e-02 -8.54682326e-01
-3.31146598e-01 -5.96143544e-01 -8.26216459e-01 7.11224914e-01
-1.62111056e+00 -1.60320890e+00 -4.53012854e-01 7.22133279e-01
3.84765118e-01 -6.92800134e-02 7.65734732e-01 1.59111381e-01
-5.69231331e-01 8.02232385e-01 -4.55551654e-01 1.82977393e-01
9.05685127e-01 -1.51387060e+00 6.82543218e-01 7.42552698e-01
1.31316885e-01 7.78250039e-01 5.15592039e-01 -8.32134008e-01
-1.47219121e+00 -1.49054480e+00 9.58883405e-01 -4.84808773e-01
7.82594025e-01 -6.79456294e-01 -1.03603804e+00 9.06251431e-01
3.27550441e-01 2.55286843e-02 6.37451887e-01 4.01703924e-01
-4.85286325e-01 -5.42229041e-02 -5.67551434e-01 5.90845227e-01
1.62783277e+00 -5.61156511e-01 -7.36494958e-01 5.49494922e-01
1.14902782e+00 -3.15013796e-01 -7.94581175e-01 4.66698885e-01
-7.15191588e-02 -5.86875677e-01 8.83884847e-01 -9.17341113e-01
5.15926957e-01 1.12912105e-02 2.65701413e-01 -1.75405002e+00
-6.79912567e-01 -1.08204186e+00 3.03502311e-03 1.60440075e+00
5.88306725e-01 -7.86093533e-01 8.20153713e-01 1.06136724e-01
-6.20834708e-01 -8.21781456e-01 -3.56283128e-01 -8.04127932e-01
1.84958532e-01 -3.30377340e-01 5.16933441e-01 1.24258745e+00
1.05031617e-01 1.23948777e+00 -2.42726520e-01 -9.72644016e-02
4.68422502e-01 1.92818329e-01 1.03007329e+00 -1.17123330e+00
-7.14874744e-01 -4.37637717e-01 1.26699030e-01 -1.27431905e+00
3.28510582e-01 -1.53163266e+00 8.24841335e-02 -1.67689908e+00
2.62242496e-01 -5.18124640e-01 -3.08255494e-01 1.08006203e+00
-6.97295427e-01 -8.80804495e-04 -1.32252142e-01 1.05873048e-01
-6.20472431e-01 8.75440419e-01 1.65178835e+00 -1.71544030e-01
-3.16762358e-01 -2.09463790e-01 -1.25675321e+00 4.35900033e-01
7.45403707e-01 -4.27175730e-01 -9.01601195e-01 -6.32570505e-01
4.96188164e-01 -9.63633209e-02 2.18337908e-01 -4.69516963e-01
1.25145078e-01 -3.11763495e-01 8.02870393e-02 6.09541312e-02
-1.54227242e-01 -2.83112973e-01 -4.71242368e-02 3.15088987e-01
-3.67373765e-01 1.40423372e-01 9.10210386e-02 7.51933038e-01
-6.38939887e-02 -9.38626565e-03 5.30982792e-01 -1.40305489e-01
-7.08736420e-01 6.84372067e-01 9.47140232e-02 5.65139949e-01
6.71570837e-01 -7.77396187e-02 -6.33785546e-01 -2.24946052e-01
-6.31169558e-01 3.63503456e-01 2.58573622e-01 4.63486910e-01
4.32989120e-01 -1.30154395e+00 -8.37999046e-01 4.09434408e-01
2.91673928e-01 4.31273878e-02 1.35752708e-01 8.33871067e-01
-1.47889987e-01 1.52725965e-01 2.59784218e-02 -4.14998919e-01
-8.52724671e-01 7.93866873e-01 7.13627189e-02 -7.10858405e-01
-4.97750342e-01 9.54153538e-01 6.38328552e-01 -7.86164522e-01
-5.63595956e-03 -5.06637514e-01 -2.97946185e-01 -1.35472029e-01
1.29873576e-02 -2.44205445e-01 -8.89067352e-03 -2.07708836e-01
-1.79577678e-01 4.84924018e-01 1.65378898e-02 2.16804206e-01
1.31829166e+00 -7.41073862e-02 2.14723381e-03 2.23228648e-01
1.08338010e+00 4.64400239e-02 -1.16983056e+00 -4.12205249e-01
-1.05254523e-01 1.57165483e-01 -1.61689386e-01 -8.27226639e-01
-9.98027980e-01 8.08906734e-01 3.89878452e-02 2.60028571e-01
1.06091690e+00 1.12246461e-01 7.89320648e-01 3.20533663e-01
9.73276123e-02 -8.21897984e-01 4.61008370e-01 7.39448190e-01
8.35896730e-01 -1.14432108e+00 -1.77382931e-01 -6.25278592e-01
-6.56123996e-01 8.28799784e-01 9.15036440e-01 -1.17203057e-01
4.53261822e-01 1.18130222e-01 5.68367615e-02 -4.26400274e-01
-9.02157426e-01 -2.22990185e-01 4.87430662e-01 7.30840087e-01
7.51089156e-01 -3.30163389e-02 -3.55198860e-01 9.43111002e-01
-3.86632442e-01 2.99846120e-02 1.36436103e-02 5.73107958e-01
-1.93612024e-01 -1.28296435e+00 7.67117962e-02 2.82256335e-01
-2.11924195e-01 -4.73538041e-01 -5.02730787e-01 8.03929687e-01
1.01625368e-01 8.29118252e-01 -4.23093796e-01 -6.11769736e-01
2.72941589e-01 1.37534231e-01 5.09779334e-01 -9.77017462e-01
-8.01527798e-01 -1.22649753e-02 1.19241200e-01 -4.14284289e-01
-1.78647652e-01 7.39409104e-02 -1.19514871e+00 -9.76114348e-02
-4.12770152e-01 4.27572489e-01 1.75240308e-01 1.11675334e+00
5.19713521e-01 8.70093107e-01 3.30400646e-01 -6.67961478e-01
-9.31553483e-01 -1.06663501e+00 -3.12102079e-01 6.54402792e-01
6.75962633e-03 -4.23033565e-01 -2.64619917e-01 3.12871397e-01] | [9.956206321716309, 8.12522029876709] |
7dc24e7e-de66-4f7d-8e05-265d459bf07f | graf-generative-radiance-fields-for-3d-aware | 2007.02442 | null | https://arxiv.org/abs/2007.02442v4 | https://arxiv.org/pdf/2007.02442v4.pdf | GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis | While 2D generative adversarial networks have enabled high-resolution image synthesis, they largely lack an understanding of the 3D world and the image formation process. Thus, they do not provide precise control over camera viewpoint or object pose. To address this problem, several recent approaches leverage intermediate voxel-based representations in combination with differentiable rendering. However, existing methods either produce low image resolution or fall short in disentangling camera and scene properties, e.g., the object identity may vary with the viewpoint. In this paper, we propose a generative model for radiance fields which have recently proven successful for novel view synthesis of a single scene. In contrast to voxel-based representations, radiance fields are not confined to a coarse discretization of the 3D space, yet allow for disentangling camera and scene properties while degrading gracefully in the presence of reconstruction ambiguity. By introducing a multi-scale patch-based discriminator, we demonstrate synthesis of high-resolution images while training our model from unposed 2D images alone. We systematically analyze our approach on several challenging synthetic and real-world datasets. Our experiments reveal that radiance fields are a powerful representation for generative image synthesis, leading to 3D consistent models that render with high fidelity. | ['Michael Niemeyer', 'Katja Schwarz', 'Yiyi Liao', 'Andreas Geiger'] | 2020-07-05 | null | http://proceedings.neurips.cc/paper/2020/hash/e92e1b476bb5262d793fd40931e0ed53-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/e92e1b476bb5262d793fd40931e0ed53-Paper.pdf | neurips-2020-12 | ['scene-generation', '3d-aware-image-synthesis'] | ['computer-vision', 'computer-vision'] | [ 6.14465952e-01 -7.03448206e-02 1.55675247e-01 4.73931469e-02
-7.83962131e-01 -1.06914210e+00 9.52280998e-01 -4.44102734e-01
1.74485862e-01 7.05084085e-01 8.27966556e-02 -1.10327981e-01
-2.63265092e-02 -1.10248041e+00 -9.56155837e-01 -8.70613635e-01
3.02317113e-01 3.21141452e-01 -1.83073338e-02 -1.86206400e-01
1.39186950e-02 8.70892107e-01 -1.49923968e+00 1.89454615e-01
7.94391334e-01 8.05780828e-01 9.85609666e-02 8.30132902e-01
2.09274203e-01 7.34665751e-01 -7.74406910e-01 -1.82306781e-01
7.09701657e-01 -6.44695103e-01 -2.50940621e-01 4.65128660e-01
7.65387654e-01 -7.20067799e-01 -6.75032854e-01 1.04368687e+00
1.98459595e-01 -3.91454883e-02 7.78570294e-01 -1.13085544e+00
-8.29568267e-01 5.11911176e-02 -3.42423648e-01 -1.91926196e-01
5.18085778e-01 3.38908494e-01 7.32146800e-01 -5.89457810e-01
7.31929123e-01 1.18014574e+00 4.55589116e-01 4.18222666e-01
-1.71846449e+00 -4.38192308e-01 6.99436516e-02 -3.82349759e-01
-1.33577275e+00 -3.48359555e-01 1.17836213e+00 -4.89723206e-01
5.04870713e-01 5.47517836e-01 6.02837741e-01 1.40053499e+00
3.44093889e-01 4.05666500e-01 1.45865953e+00 -2.93225318e-01
1.80746824e-01 2.30742898e-02 -6.37706220e-01 5.23168623e-01
1.64926648e-01 5.53626955e-01 -3.97275478e-01 -8.09024721e-02
1.62938333e+00 8.84527564e-02 -6.02028489e-01 -8.47635746e-01
-1.36876631e+00 8.52751911e-01 6.40271604e-01 2.58932561e-02
-3.31477344e-01 2.50662833e-01 -2.85991192e-01 2.33633891e-01
3.77887368e-01 7.03706503e-01 6.33015409e-02 3.01066846e-01
-7.69523144e-01 5.70451081e-01 6.00074172e-01 9.50925112e-01
7.94944167e-01 5.01198053e-01 -1.89700499e-01 4.09522891e-01
3.08939312e-02 7.71553397e-01 -1.18896766e-02 -1.22328913e+00
2.20467821e-01 2.18337983e-01 3.15256923e-01 -1.13611197e+00
-3.71311530e-02 -4.27492112e-01 -9.94534910e-01 8.29244196e-01
4.39710945e-01 1.09393872e-01 -1.08083057e+00 1.79078162e+00
2.64680922e-01 1.69106469e-01 1.31923974e-01 1.04270744e+00
7.57626832e-01 6.31899297e-01 -3.39488268e-01 -3.18684466e-02
1.05323720e+00 -6.27211094e-01 -4.92834717e-01 -2.17709601e-01
-2.41923690e-01 -7.80597389e-01 8.84685457e-01 3.92470956e-01
-1.43465245e+00 -5.90352595e-01 -1.15079665e+00 -1.19706415e-01
-1.28007621e-01 -4.45976317e-01 5.54233611e-01 6.82324708e-01
-1.05074787e+00 3.38476330e-01 -6.71353638e-01 1.77245736e-01
3.53260279e-01 2.67457664e-02 -4.17168766e-01 -4.27660465e-01
-9.82174397e-01 7.31851101e-01 1.18807629e-02 -1.61959082e-01
-1.09844530e+00 -8.48488867e-01 -8.73521686e-01 -6.24797158e-02
3.05874705e-01 -1.15724981e+00 9.42551970e-01 -8.10990334e-01
-1.57342553e+00 7.90645063e-01 2.23828048e-01 -2.53348947e-01
7.77930260e-01 1.61556154e-01 -1.85331479e-01 2.94913530e-01
-1.09779071e-02 5.80174983e-01 1.21459484e+00 -1.84308469e+00
-2.29225740e-01 -1.90392300e-01 6.04423225e-01 3.51792961e-01
3.05952340e-01 -5.30529201e-01 -1.77184850e-01 -7.49794364e-01
2.84200311e-01 -7.33474374e-01 -3.56220841e-01 2.80705154e-01
-4.57007527e-01 7.46282101e-01 7.62283921e-01 -3.86704028e-01
4.42262977e-01 -2.02602720e+00 3.60736877e-01 -6.68074563e-02
4.49526101e-01 -6.95563853e-02 -1.70665950e-01 4.15813744e-01
7.29434751e-03 1.48044690e-01 -2.88675904e-01 -2.59100407e-01
1.96569897e-02 3.66084129e-01 -6.14571452e-01 5.67111135e-01
3.24681312e-01 1.07168853e+00 -8.78538728e-01 -1.19090743e-01
6.19378984e-01 9.72660303e-01 -8.28193665e-01 3.07077169e-01
-4.26568300e-01 1.03980827e+00 -5.31461120e-01 4.24291760e-01
8.89115930e-01 -2.10477665e-01 -6.41744733e-02 -2.69430578e-01
-2.47911196e-02 1.14712149e-01 -1.07043874e+00 1.77627492e+00
-6.53471589e-01 4.83381301e-01 2.18041524e-01 -6.87851071e-01
8.16386938e-01 1.89016059e-01 5.12636483e-01 -8.23711395e-01
2.80009154e-02 -3.63110527e-02 -2.73451358e-01 -1.00880854e-01
5.79217613e-01 -3.83758396e-01 -1.58418104e-01 2.71424174e-01
-1.90403983e-01 -9.31646943e-01 -3.75155061e-01 3.72750871e-02
9.97100413e-01 2.95799077e-01 1.61480561e-01 1.43676937e-01
3.53062637e-02 -2.04070196e-01 3.59596491e-01 8.23467731e-01
3.29544038e-01 1.36486399e+00 3.35409075e-01 -2.35337511e-01
-1.40020728e+00 -1.60446167e+00 -2.47548610e-01 1.46591857e-01
4.28394914e-01 -1.04573280e-01 -5.93508244e-01 -2.83661962e-01
-9.44137871e-02 7.71583617e-01 -6.04920387e-01 -1.81015030e-01
-5.85065663e-01 -5.03391802e-01 3.08688998e-01 2.06218317e-01
5.29185712e-01 -6.05803132e-01 -8.20057213e-01 1.33695364e-01
-1.23874307e-01 -1.48650515e+00 -3.42597961e-01 -1.23762116e-01
-6.86253250e-01 -9.89525199e-01 -6.37535512e-01 -2.89207906e-01
6.70245290e-01 4.30271596e-01 1.37231517e+00 -2.40670264e-01
-4.14600819e-01 5.55404305e-01 -7.01357797e-02 -7.16279307e-03
-7.99655378e-01 -3.45184088e-01 -1.42719820e-01 6.36994019e-02
-5.24018288e-01 -1.05265915e+00 -7.90690243e-01 3.29490840e-01
-1.19460499e+00 4.30887371e-01 4.35234308e-01 8.44303966e-01
6.41984582e-01 1.40053645e-01 1.57718137e-01 -8.73468280e-01
2.64152408e-01 -3.08214754e-01 -7.51458943e-01 9.59824398e-03
-3.05809140e-01 -1.02021620e-02 7.61615098e-01 -5.69730401e-01
-1.08163500e+00 6.85001537e-02 -1.55784213e-03 -7.83042669e-01
-1.96366504e-01 -3.33935507e-02 -3.20201784e-01 -2.99210638e-01
7.60056734e-01 4.02933896e-01 -5.25759123e-02 -3.55178624e-01
5.21429181e-01 -1.89622175e-02 8.04870307e-01 -6.91973984e-01
1.27523220e+00 7.93651640e-01 3.96074355e-01 -5.88197470e-01
-5.55971682e-01 2.60736495e-01 -4.66694564e-01 -2.78427731e-02
9.64692891e-01 -1.01110506e+00 -4.66370165e-01 3.57279450e-01
-1.09624350e+00 -3.07583213e-01 -5.54101706e-01 2.51657128e-01
-8.49375904e-01 1.65507913e-01 -3.85138720e-01 -5.98430634e-01
1.23175696e-01 -1.31524503e+00 1.28626657e+00 8.67060423e-02
8.36110339e-02 -9.55940962e-01 -1.72189355e-01 2.71776944e-01
5.21492004e-01 9.49584961e-01 9.47070658e-01 4.10579085e-01
-1.37755454e+00 -7.90470541e-02 -1.31383926e-01 2.72960365e-01
2.77412146e-01 -1.58994794e-01 -1.12860322e+00 -2.60390341e-01
2.54223764e-01 -1.25061333e-01 5.96585929e-01 3.61306548e-01
9.92089391e-01 -4.41077352e-01 3.42156217e-02 1.21091652e+00
1.56647384e+00 6.72027469e-02 8.21427107e-01 3.67647707e-02
9.30406928e-01 3.28062475e-01 1.87903330e-01 3.01997155e-01
8.72993618e-02 9.62413609e-01 8.62900674e-01 -2.51221418e-01
-5.11552691e-01 -6.29273653e-01 1.43104360e-01 1.74628660e-01
-1.74557343e-01 -5.11018515e-01 -5.25902450e-01 9.84577611e-02
-1.38802552e+00 -9.50227320e-01 8.05987343e-02 2.18881607e+00
5.49452305e-01 3.30879241e-02 -1.88284844e-01 4.10704054e-02
3.86281848e-01 4.82297421e-01 -7.81081975e-01 -2.26347707e-02
-4.83574808e-01 2.00423867e-01 5.33488512e-01 6.85761034e-01
-7.12525249e-01 7.33987510e-01 6.46542835e+00 5.39570510e-01
-1.26520443e+00 3.63775901e-02 4.31702107e-01 -1.89658463e-01
-9.65615630e-01 1.85445230e-02 -4.19604897e-01 2.68456846e-01
2.97285825e-01 -1.10911913e-01 6.76207066e-01 4.85925555e-01
-6.14814311e-02 1.30331740e-01 -1.16899061e+00 1.10547650e+00
1.84031233e-01 -1.54658830e+00 3.82347465e-01 2.35010311e-01
1.01749897e+00 -2.30360478e-01 5.30250072e-01 -1.54407054e-01
3.34966362e-01 -1.29625499e+00 8.95390809e-01 5.74097693e-01
1.10719860e+00 -6.79628193e-01 -8.69993418e-02 4.44619119e-01
-7.32132733e-01 2.18598142e-01 -2.12041274e-01 9.47561562e-02
3.16699117e-01 4.31483835e-01 -6.04421735e-01 6.86285377e-01
4.49737817e-01 4.82836664e-01 -3.08908731e-01 5.01956165e-01
-3.89202416e-01 9.26971883e-02 -1.97638631e-01 5.55225432e-01
1.55691117e-01 -3.76027644e-01 8.81480992e-01 5.09725988e-01
4.49349612e-01 2.96400338e-01 1.07367598e-01 1.55015969e+00
-8.52715448e-02 -5.07108629e-01 -9.43420291e-01 1.72942340e-01
2.73766398e-01 1.03822744e+00 -4.61586386e-01 1.61583394e-01
-2.76325583e-01 1.03514040e+00 -1.81666017e-03 7.47042000e-01
-8.44922841e-01 1.75865754e-01 9.15888011e-01 5.12429833e-01
3.40667844e-01 -5.37130892e-01 -3.01940262e-01 -1.47925222e+00
5.09601757e-02 -9.97868896e-01 -3.38504940e-01 -1.12885368e+00
-1.14607441e+00 6.76702142e-01 8.36091638e-02 -1.40632701e+00
-6.11401021e-01 -5.22171617e-01 -2.89406866e-01 1.06196177e+00
-1.52636504e+00 -1.41832292e+00 -3.50608230e-01 6.97697997e-01
3.85653406e-01 5.14320172e-02 8.83861661e-01 6.90998435e-02
-5.52790053e-02 3.29109609e-01 1.30133376e-01 -4.91367131e-02
4.44517165e-01 -1.05690873e+00 5.10892451e-01 9.24703717e-01
3.21459562e-01 4.60460216e-01 7.77366519e-01 -2.80255914e-01
-1.71738160e+00 -9.56600487e-01 7.93149322e-02 -7.79759407e-01
1.79934874e-01 -6.27949297e-01 -8.22479248e-01 6.35465920e-01
9.41728503e-02 3.53847653e-01 2.06474081e-01 -3.69032800e-01
-5.77496588e-01 -1.48293618e-02 -1.35405958e+00 7.74865329e-01
1.18912768e+00 -7.02696800e-01 -1.98063567e-01 1.03220649e-01
8.29482079e-01 -7.45798051e-01 -8.90499294e-01 3.52193534e-01
3.89981866e-01 -1.36426222e+00 1.54214549e+00 -2.54495353e-01
7.88941979e-01 -4.83324647e-01 -4.69234794e-01 -1.39707649e+00
-3.23288262e-01 -6.68857753e-01 -9.37607586e-02 1.04784286e+00
-3.17265908e-03 -6.60739362e-01 6.20007515e-01 5.48079491e-01
-1.78047031e-01 -4.29457337e-01 -7.72971928e-01 -6.78150952e-01
2.60784805e-01 -2.91861355e-01 7.96540976e-01 8.98229718e-01
-8.46808434e-01 2.36476541e-01 -5.27890205e-01 5.31419277e-01
9.33150291e-01 5.74599862e-01 9.12620425e-01 -8.40435028e-01
-6.58966184e-01 -4.23344195e-01 -5.29940307e-01 -1.20793760e+00
9.12596956e-02 -6.48424685e-01 1.08698551e-02 -1.33397484e+00
9.16708086e-04 -5.82757831e-01 1.39861912e-01 -1.38570219e-02
9.19439346e-02 5.76015174e-01 2.08217055e-01 2.21197128e-01
1.18661106e-01 4.87479627e-01 1.78777087e+00 -1.35694802e-01
3.26832347e-02 -1.09428629e-01 -7.40311444e-01 5.28251410e-01
5.28777003e-01 -1.65444329e-01 -7.85924613e-01 -8.35467994e-01
1.97272494e-01 4.51496601e-01 9.78420079e-01 -8.76963854e-01
-9.54719856e-02 -3.54966909e-01 6.87978804e-01 -2.40604192e-01
7.72212982e-01 -8.23178947e-01 7.27839053e-01 1.00980200e-01
-2.03820035e-01 -2.78031558e-01 8.89320523e-02 6.53115869e-01
-2.03532532e-01 2.29135379e-01 9.78839219e-01 -2.53639251e-01
-4.09690380e-01 5.16386867e-01 4.23904434e-02 1.34042222e-02
8.13701689e-01 -3.38381320e-01 -3.23874682e-01 -6.02032125e-01
-4.31581616e-01 -4.01579559e-01 1.12686884e+00 4.53410864e-01
5.62182605e-01 -1.43200529e+00 -7.24733531e-01 7.06687331e-01
2.43721772e-02 4.74102646e-01 4.46760684e-01 1.37149528e-01
-6.50166035e-01 8.24254677e-02 -3.85713428e-01 -8.61248791e-01
-8.64919066e-01 7.43282020e-01 5.67639053e-01 -6.98543862e-02
-8.32260132e-01 5.32260358e-01 8.66528451e-01 -3.14490229e-01
-2.30212077e-01 -2.23418951e-01 3.30735117e-01 -4.14749831e-01
3.49191636e-01 -1.78802490e-01 -1.78484768e-01 -6.59793317e-01
3.97057608e-02 8.62308323e-01 3.25270563e-01 -3.13779622e-01
1.21264052e+00 -1.74896240e-01 2.02730462e-01 2.49589309e-01
1.19768250e+00 3.29618186e-01 -1.86539567e+00 -2.03488350e-01
-9.54105735e-01 -8.81386399e-01 2.22154170e-01 -5.87672412e-01
-1.21877217e+00 8.27158630e-01 2.88172752e-01 2.80469805e-01
1.20714998e+00 -5.30483872e-02 5.64452767e-01 -1.07186377e-01
5.80704093e-01 -3.28463346e-01 1.75895080e-01 1.77509174e-01
1.03326249e+00 -1.12889147e+00 7.38619640e-03 -5.87811112e-01
-3.14667255e-01 8.26655865e-01 3.57186526e-01 -3.34835857e-01
3.13505888e-01 5.26985347e-01 2.03770269e-02 -1.36920158e-02
-5.64061821e-01 8.87131914e-02 3.41592759e-01 8.43456089e-01
9.79886502e-02 -1.75035577e-02 4.62282270e-01 8.37533176e-02
-4.25034702e-01 -3.84114414e-01 6.11996412e-01 6.44544899e-01
5.28457994e-03 -9.93036091e-01 -5.43388486e-01 -6.76009804e-02
-2.59584039e-01 -1.29070738e-02 -1.74125314e-01 8.64899814e-01
6.74998090e-02 6.85944378e-01 1.48858711e-01 -2.88556337e-01
3.74735683e-01 -4.37708318e-01 1.04019582e+00 -5.42205453e-01
-1.33299977e-01 1.37969613e-01 -1.59651577e-01 -6.45129144e-01
-3.24347794e-01 -5.86441576e-01 -6.96424842e-01 -4.12064493e-01
8.56528580e-02 -2.51390874e-01 5.90378761e-01 5.40862679e-01
2.33892262e-01 6.66256189e-01 9.03574467e-01 -1.12346697e+00
-5.33076465e-01 -3.45769137e-01 -6.40980184e-01 4.44942117e-01
7.38655508e-01 -6.27722979e-01 -4.46926296e-01 1.09111890e-01] | [9.313028335571289, -3.194117307662964] |
815603ef-e69a-461d-96b9-5aceac187ffe | a-transformer-based-contrastive-learning | 2204.02803 | null | https://arxiv.org/abs/2204.02803v1 | https://arxiv.org/pdf/2204.02803v1.pdf | A Transformer-Based Contrastive Learning Approach for Few-Shot Sign Language Recognition | Sign language recognition from sequences of monocular images or 2D poses is a challenging field, not only due to the difficulty to infer 3D information from 2D data, but also due to the temporal relationship between the sequences of information. Additionally, the wide variety of signs and the constant need to add new ones on production environments makes it infeasible to use traditional classification techniques. We propose a novel Contrastive Transformer-based model, which demonstrate to learn rich representations from body key points sequences, allowing better comparison between vector embedding. This allows us to apply these techniques to perform one-shot or few-shot tasks, such as classification and translation. The experiments showed that the model could generalize well and achieved competitive results for sign classes never seen in the training process. | ['Jampierre Rocha', 'Márcio Dahia', 'Esdras Costa', 'Silvan Ferreira'] | 2022-04-05 | null | null | null | null | ['sign-language-recognition'] | ['computer-vision'] | [ 1.54911175e-01 -5.63951552e-01 -3.03943545e-01 -2.47377738e-01
-2.59710044e-01 -5.02710760e-01 6.91266954e-01 -7.13813961e-01
-2.97221243e-01 6.09047949e-01 2.33851418e-01 1.21281169e-01
-9.97262746e-02 -4.10039783e-01 -5.20686150e-01 -6.56439066e-01
-4.46004346e-02 3.16034615e-01 5.12432754e-01 -2.90771723e-01
1.26193330e-01 6.53401792e-01 -2.01330996e+00 2.19252467e-01
5.96670032e-01 8.00168335e-01 3.07366371e-01 5.38898647e-01
-2.59551197e-01 8.68427932e-01 -4.38623667e-01 -3.85387152e-01
3.83748919e-01 -6.58069670e-01 -2.79642105e-01 6.40564039e-02
8.58846843e-01 -7.42009997e-01 -8.82539213e-01 8.85468543e-01
7.23319709e-01 1.06757954e-01 6.98787987e-01 -1.39811051e+00
-6.49635375e-01 -2.78419852e-02 -3.11114430e-01 8.72945413e-02
7.07162023e-01 2.54401475e-01 7.97073364e-01 -7.90961981e-01
9.72738445e-01 1.12374341e+00 4.31981862e-01 8.71150017e-01
-7.00366378e-01 -6.33969247e-01 1.82228759e-01 8.05924177e-01
-8.97632301e-01 -2.57113725e-01 8.46668005e-01 -5.91510773e-01
5.68966329e-01 4.08489406e-02 1.21337128e+00 1.49696219e+00
6.76318482e-02 1.24438465e+00 1.21823716e+00 -5.22837639e-01
6.35034814e-02 -1.48257285e-01 1.24550797e-01 7.18844652e-01
2.35045865e-01 3.75800043e-01 -7.77995467e-01 2.52584517e-01
8.79521012e-01 2.69093961e-01 -4.31813061e-01 -8.21902335e-01
-1.26259398e+00 4.20190543e-01 4.57731456e-01 4.33285475e-01
-2.95868874e-01 1.65146753e-01 2.33188957e-01 5.16455472e-01
-8.24603885e-02 1.49111226e-01 -2.97679335e-01 -4.34052914e-01
-6.26393080e-01 2.02064544e-01 4.83928114e-01 9.09128010e-01
2.06572235e-01 8.84734839e-02 -1.11030862e-02 6.93707705e-01
2.78894573e-01 7.08904624e-01 8.02319825e-01 -6.22885585e-01
5.39047599e-01 5.18432915e-01 -4.94565330e-02 -8.32833529e-01
-2.50462264e-01 -1.53687477e-01 -5.84664822e-01 5.89886904e-01
7.00728595e-01 1.93523824e-01 -1.43158841e+00 1.69464076e+00
1.39929995e-01 -2.73381583e-02 -1.60697415e-01 1.16931248e+00
6.89640999e-01 2.28375584e-01 -2.60381252e-01 -2.58243755e-02
1.18438482e+00 -7.77203798e-01 -8.49024475e-01 -1.75521627e-01
4.58792299e-01 -5.80183268e-01 1.07828355e+00 4.15434182e-01
-6.82208300e-01 -5.67369521e-01 -1.05218983e+00 -9.68607962e-02
-6.39180422e-01 8.13215896e-02 7.53598690e-01 4.78280246e-01
-6.27105236e-01 5.89062035e-01 -1.07720733e+00 -5.10334790e-01
2.29071826e-01 2.43795887e-01 -6.44365370e-01 -3.02890390e-01
-9.35538888e-01 1.17574811e+00 9.53688696e-02 2.47918025e-01
-4.07309353e-01 -3.87292445e-01 -9.93257403e-01 -2.56672084e-01
1.33426562e-01 -4.74385440e-01 1.04465008e+00 -5.99948883e-01
-1.68912363e+00 8.82209957e-01 -2.04584002e-01 -1.00531034e-01
9.06853676e-01 -4.49687421e-01 -2.71004587e-01 6.17072023e-02
-2.05697358e-01 4.63674903e-01 8.85366201e-01 -9.70476329e-01
-4.89646554e-01 -5.85956037e-01 8.86314735e-02 2.03255579e-01
-1.15810171e-01 -1.49327755e-01 -2.66294390e-01 -5.12406766e-01
5.38880765e-01 -1.05151272e+00 9.33685526e-02 5.53881705e-01
-5.69699034e-02 -6.14386648e-02 1.00596464e+00 -7.86870837e-01
5.90801954e-01 -2.13764954e+00 4.81891364e-01 -4.57374454e-02
-9.60769728e-02 4.47319865e-01 -1.07506745e-01 4.33718622e-01
-1.60813946e-02 -3.75614405e-01 1.20400548e-01 5.84607795e-02
1.73868574e-02 4.90744799e-01 -3.44703883e-01 4.32517052e-01
9.50253084e-02 1.04980469e+00 -1.10948002e+00 -3.86207163e-01
5.19834518e-01 4.58175302e-01 -2.55604804e-01 1.70971379e-01
-4.02306579e-02 7.05629766e-01 -2.62273967e-01 8.51433337e-01
4.70933616e-01 -9.43571627e-02 -7.59302527e-02 -3.48501205e-01
6.29157126e-02 1.90073669e-01 -1.05574977e+00 1.88011515e+00
-2.46100113e-01 9.14380550e-01 -4.18577105e-01 -1.03022540e+00
7.52239823e-01 3.76322091e-01 4.67824489e-01 -9.06254113e-01
1.89336151e-01 3.02921385e-01 2.38013025e-02 -9.72027361e-01
1.33131370e-01 -1.80710375e-01 1.48239015e-02 2.54432648e-01
1.89825267e-01 -3.59346181e-01 2.30819881e-01 -3.00668001e-01
9.20704842e-01 6.52692258e-01 2.82095224e-01 4.45756316e-01
3.15382659e-01 -2.50450391e-02 5.65677941e-01 3.94081384e-01
-3.41929793e-01 6.26761615e-01 9.15990993e-02 -5.20309269e-01
-8.20528626e-01 -1.31160688e+00 7.18101263e-02 6.63756073e-01
3.24554175e-01 -5.32252006e-02 3.51820029e-02 -7.70031393e-01
1.04294136e-01 3.93354475e-01 -5.47544003e-01 -1.88519120e-01
-7.86467135e-01 -2.42204174e-01 3.32079053e-01 8.20294261e-01
5.75304508e-01 -6.87991798e-01 -9.35754120e-01 -8.31842721e-02
-1.98244378e-01 -1.06266439e+00 -3.89613509e-01 -2.48416867e-02
-9.93264318e-01 -1.18033135e+00 -1.12533391e+00 -1.01983082e+00
5.68691313e-01 2.03112677e-01 4.30290014e-01 -2.76756555e-01
-4.53282237e-01 4.95242000e-01 -6.38355851e-01 -3.57461691e-01
-2.57976830e-01 -3.39167565e-01 1.61063120e-01 1.61043778e-02
4.55517054e-01 -7.16890872e-01 -3.74252230e-01 4.66007620e-01
-5.73593616e-01 1.21806137e-01 6.50294483e-01 1.25871575e+00
2.86555707e-01 -6.39107883e-01 1.94063917e-01 -6.97093783e-03
2.87261218e-01 1.74668923e-01 -5.78457475e-01 4.49796081e-01
-2.79187202e-01 2.60721058e-01 2.51594245e-01 -8.79701674e-01
-8.23569715e-01 1.91965297e-01 4.44527855e-03 -6.45730197e-01
-1.22264802e-01 1.68851539e-01 -1.09845005e-01 -2.49747857e-01
4.23940927e-01 5.82426310e-01 4.35253769e-01 -5.82973301e-01
4.57412034e-01 6.41787291e-01 6.38009727e-01 -1.36723861e-01
8.38491619e-01 4.36846375e-01 1.52512208e-01 -8.51406217e-01
-5.02116144e-01 -3.07518840e-01 -1.10842633e+00 -4.80874747e-01
6.40576303e-01 -7.63172507e-01 -6.81303561e-01 7.39849389e-01
-1.12005210e+00 -6.46616220e-02 -3.70248973e-01 1.06260157e+00
-7.63265550e-01 5.81753671e-01 -5.84515929e-01 -6.89681053e-01
1.19619496e-01 -9.85823095e-01 9.28390026e-01 2.47207910e-01
-1.47385165e-01 -4.65589046e-01 1.20584950e-01 3.15177381e-01
1.61848158e-01 3.29941779e-01 8.50226700e-01 -3.13515484e-01
-8.05613995e-01 -5.67979932e-01 -8.77870992e-02 5.38530171e-01
4.37611312e-01 -1.77679718e-01 -8.48669887e-01 -2.52006620e-01
-1.94947153e-01 -5.39233565e-01 6.61608458e-01 1.58288404e-01
7.22871423e-01 2.80352011e-02 -3.08439285e-01 5.80540299e-01
1.07548130e+00 3.26890796e-01 7.87512004e-01 1.71448767e-01
5.81750214e-01 5.07785261e-01 5.95433772e-01 1.46050707e-01
1.62224323e-01 1.08011079e+00 1.05024174e-01 1.77127644e-01
-5.40988982e-01 -5.84165990e-01 3.43846887e-01 1.11285603e+00
-4.33063924e-01 1.41981140e-01 -6.54492438e-01 4.89896774e-01
-1.89712763e+00 -1.20258832e+00 1.87078014e-01 2.06711316e+00
6.09066010e-01 -2.17551380e-01 -8.85233954e-02 4.70731944e-01
4.23701763e-01 -4.14604833e-03 -6.06883526e-01 -5.23569137e-02
-1.94329411e-01 1.46178454e-01 2.36757770e-01 1.32950231e-01
-9.36837077e-01 8.91769290e-01 7.01898050e+00 3.70599896e-01
-1.54302323e+00 -2.99672391e-02 -3.22398990e-01 -2.79548407e-01
2.09483370e-01 -2.96638310e-01 -5.15925229e-01 4.97541040e-01
2.51789331e-01 -1.11319125e-01 1.92569852e-01 7.33129561e-01
-2.33956873e-02 1.84247270e-02 -1.41334021e+00 1.37404716e+00
4.39685583e-01 -9.44076717e-01 2.13425651e-01 1.96880754e-02
5.30110121e-01 4.69751582e-02 -8.33069012e-02 3.18538606e-01
2.98963431e-02 -6.95124388e-01 6.66841805e-01 6.57891095e-01
8.09980154e-01 -9.38008279e-02 4.98461217e-01 4.84362394e-01
-1.04457915e+00 -1.77847698e-01 -1.70746148e-01 -3.41612071e-01
3.85318220e-01 -4.40471172e-02 -7.59149611e-01 4.20138836e-01
5.12375593e-01 8.95141602e-01 -3.02375048e-01 1.42000198e+00
-5.82613707e-01 2.76915252e-01 -4.52817082e-01 -4.78041112e-01
-1.25875935e-01 -1.66528240e-01 5.94769120e-01 8.01325083e-01
4.62892860e-01 -8.80585909e-02 -7.64460862e-02 5.51642418e-01
3.42990637e-01 1.05732977e-01 -1.04074359e+00 -4.22888920e-02
7.88807869e-03 5.73901892e-01 -3.80768418e-01 -2.63576329e-01
-5.65702617e-01 1.36020231e+00 -3.06720827e-02 3.85408700e-01
-7.64736593e-01 -5.02319098e-01 8.14034641e-01 -7.16322809e-02
4.88708109e-01 -7.06488729e-01 -5.22726849e-02 -1.67849231e+00
5.50508559e-01 -7.60978281e-01 1.84552953e-01 -9.16267216e-01
-1.26661885e+00 3.21504653e-01 1.51862413e-01 -1.92697847e+00
-6.64833724e-01 -9.46701527e-01 -2.78954983e-01 6.08008206e-01
-1.52022982e+00 -1.16943312e+00 -5.00345767e-01 7.63803661e-01
4.68213588e-01 -2.50679463e-01 8.05930793e-01 4.16823894e-01
3.65002453e-02 7.38194644e-01 1.10379949e-01 4.30136174e-01
7.28990316e-01 -8.26018870e-01 2.31431544e-01 7.23059952e-01
4.53764468e-01 3.10498506e-01 5.24508715e-01 -4.00758147e-01
-1.55837655e+00 -3.83613259e-01 9.59820926e-01 -5.66118062e-01
4.02287215e-01 -3.77203256e-01 -6.11797154e-01 4.73021269e-01
-3.39248121e-01 -2.16825604e-02 4.66397256e-01 -7.32719749e-02
-5.78221738e-01 -2.33835414e-01 -1.03632367e+00 7.73589015e-01
1.65243137e+00 -5.60426593e-01 -1.26235688e+00 1.30014494e-01
3.77267987e-01 -4.88846332e-01 -4.84207541e-01 5.68885744e-01
1.25602651e+00 -6.36769235e-01 9.36673224e-01 -8.27743530e-01
3.98930609e-01 -1.87951788e-01 -3.40900064e-01 -1.28905666e+00
-1.61725059e-01 -1.08145133e-01 -3.14712346e-01 7.27392972e-01
1.78403422e-01 -7.63658464e-01 7.68354297e-01 3.60854268e-01
1.45829618e-01 -4.45658863e-01 -1.26367056e+00 -1.32053268e+00
-1.20873891e-01 -2.05760568e-01 4.78485018e-01 6.68824792e-01
1.52552113e-01 1.10128663e-01 -6.41081750e-01 -1.48034558e-01
6.41259670e-01 5.10945320e-01 1.01717865e+00 -1.26620686e+00
-4.55142707e-01 -3.89419466e-01 -1.25948393e+00 -1.47871494e+00
-1.28298208e-01 -7.26182878e-01 1.13885894e-01 -1.48447788e+00
-1.04521610e-01 -7.80555010e-02 -2.06876263e-01 4.89644974e-01
1.15412965e-01 3.46997947e-01 5.13887405e-01 2.04755053e-01
-5.72601147e-02 8.78290117e-01 1.73072946e+00 -4.15386170e-01
-1.09204195e-01 1.74731523e-01 -3.02192438e-02 5.73349714e-01
4.20710951e-01 -1.51554734e-01 -4.52869028e-01 -5.00229299e-01
-1.22156978e-01 -9.77968890e-03 5.08064091e-01 -1.14460671e+00
1.56117454e-01 -1.77027017e-01 4.50028449e-01 -7.86862910e-01
7.55354524e-01 -7.72873402e-01 -4.93543558e-02 5.65007031e-01
-6.54715821e-02 -1.77712619e-01 6.04553968e-02 4.97210681e-01
-4.50787127e-01 -6.39218464e-02 5.42795777e-01 -1.70652300e-01
-1.21236694e+00 2.26972967e-01 -2.05977023e-01 -1.45120531e-01
9.64927018e-01 -8.07424307e-01 -5.56512401e-02 -4.62963015e-01
-8.63990307e-01 6.99627176e-02 3.48123252e-01 9.52504098e-01
7.78820038e-01 -1.60791409e+00 -4.53099459e-01 5.58545053e-01
3.80568564e-01 -3.60656500e-01 1.43722832e-01 7.62823284e-01
-4.66521025e-01 4.00071681e-01 -7.68523455e-01 -9.03071582e-01
-1.57588089e+00 3.87553990e-01 2.46960908e-01 9.99233052e-02
-1.09842741e+00 5.64800262e-01 -6.60486817e-02 -3.98677558e-01
6.10742867e-01 -4.81548429e-01 -2.40376815e-02 8.08934197e-02
6.13575935e-01 1.55314952e-01 -1.47351265e-01 -5.77739298e-01
-3.69942516e-01 1.13773715e+00 1.95534617e-01 -2.96097338e-01
1.26872885e+00 1.02155149e-01 3.84874761e-01 7.93105721e-01
1.15363145e+00 -4.13612008e-01 -1.21907318e+00 -3.45321387e-01
-7.95313939e-02 -9.45009291e-01 -3.48819703e-01 -9.11869407e-01
-8.40186715e-01 1.10115623e+00 8.93015921e-01 -5.32602727e-01
9.23775971e-01 -3.00891250e-02 8.43556941e-01 6.38919115e-01
8.79031122e-01 -1.00735569e+00 1.90787569e-01 5.66132486e-01
1.01752496e+00 -1.18924868e+00 -1.55744463e-01 -2.65445590e-01
-3.92009050e-01 1.22966707e+00 4.06645328e-01 -3.70306224e-02
6.47531569e-01 2.59767170e-03 4.09515768e-01 1.71791892e-02
-5.34809113e-01 -4.25077260e-01 4.75682586e-01 8.83146107e-01
2.33318210e-01 -2.55014692e-02 -4.89919424e-01 1.53796166e-01
-2.25066662e-01 5.95167696e-01 3.40379715e-01 1.25578225e+00
-1.14737414e-01 -1.30853534e+00 -2.58937657e-01 3.03558588e-01
2.05946758e-01 4.46044415e-01 -3.12341541e-01 9.20124531e-01
3.36336136e-01 4.87481803e-01 -9.90624055e-02 -6.12971783e-01
6.46349549e-01 4.88625824e-01 1.17981839e+00 -2.73643464e-01
1.07325636e-01 -1.91215038e-01 -1.87081296e-03 -5.62531710e-01
-4.63672221e-01 -8.85422945e-01 -9.84562516e-01 1.08459994e-01
-4.63580370e-01 -4.26297605e-01 8.49476337e-01 1.01726449e+00
3.66120636e-02 2.78319925e-01 3.61243337e-01 -8.85143995e-01
-8.99304450e-01 -9.87899303e-01 -5.38922906e-01 7.95495510e-01
3.19850534e-01 -1.16257107e+00 -1.65186092e-01 1.82955146e-01] | [9.064603805541992, -6.3711981773376465] |
3ef0d588-dd0e-41a4-a20c-faf20bc1496f | rethinking-gradient-operator-for-exposing-ai | 2205.00767 | null | https://arxiv.org/abs/2205.00767v1 | https://arxiv.org/pdf/2205.00767v1.pdf | Rethinking Gradient Operator for Exposing AI-enabled Face Forgeries | For image forensics, convolutional neural networks (CNNs) tend to learn content features rather than subtle manipulation traces, which limits forensic performance. Existing methods predominantly solve the above challenges by following a general pipeline, that is, subtracting the original pixel value from the predicted pixel value to make CNNs pay attention to the manipulation traces. However, due to the complicated learning mechanism, these methods may bring some unnecessary performance losses. In this work, we rethink the advantages of gradient operator in exposing face forgery, and design two plug-and-play modules by combining gradient operator with CNNs, namely tensor pre-processing (TP) and manipulation trace attention (MTA) module. Specifically, TP module refines the feature tensor of each channel in the network by gradient operator to highlight the manipulation traces and improve the feature representation. Moreover, MTA module considers two dimensions, namely channel and manipulation traces, to force the network to learn the distribution of manipulation traces. These two modules can be seamlessly integrated into CNNs for end-to-end training. Experiments show that the proposed network achieves better results than prior works on five public datasets. Especially, TP module greatly improves the accuracy by at least 4.60% compared with the existing pre-processing module only via simple tensor refinement. The code is available at: https://github.com/EricGzq/GocNet-pytorch. | ['Ming Xia', 'Dengyong Zhang', 'Gaobo Yang', 'Zhiqing Guo'] | 2022-05-02 | null | null | null | null | ['image-forensics'] | ['computer-vision'] | [ 1.0285518e-02 -3.8612968e-01 6.6623330e-02 -2.4824123e-01
-2.5050735e-01 -3.5540524e-01 2.1692738e-01 -2.8505310e-01
-3.6307770e-01 1.8623536e-02 -5.3078663e-02 -3.7454224e-01
1.3639842e-01 -8.9406675e-01 -7.8653896e-01 -6.9605917e-01
7.8618256e-03 -3.5661694e-01 2.4768420e-01 -5.1473353e-02
5.2292609e-01 6.5780807e-01 -1.0445092e+00 5.5430895e-01
6.1907440e-01 1.2134646e+00 -1.2676869e-01 3.0770621e-01
-1.2079631e-01 9.8064995e-01 -4.8880067e-01 -8.6014020e-01
3.0384579e-01 -3.4356799e-02 -7.0097387e-01 -1.1881318e-01
4.0547544e-01 -9.4052935e-01 -7.8433615e-01 1.3169643e+00
5.3774977e-01 -8.8615172e-02 1.5653902e-01 -1.3274901e+00
-9.0478134e-01 4.1729900e-01 -1.0670705e+00 4.7069237e-01
-6.1382413e-02 6.0195303e-01 6.8009627e-01 -8.3959675e-01
3.6247459e-01 1.3310348e+00 8.8607252e-01 5.8847237e-01
-9.4862288e-01 -1.1900804e+00 -1.0126963e-02 8.1739926e-01
-1.2073971e+00 -4.9118644e-01 1.2745723e+00 -3.6177701e-01
7.1458364e-01 1.5542172e-01 2.9452738e-01 1.1533233e+00
-1.7061666e-01 9.6911514e-01 8.8204074e-01 2.2439332e-01
-1.0745170e-01 -2.7280208e-01 3.0371112e-01 7.1849132e-01
2.7893537e-01 -5.6420811e-02 -4.3589982e-01 -3.6982521e-02
7.9159451e-01 4.6695262e-01 -3.6217943e-01 2.4447618e-01
-7.5639689e-01 6.4004320e-01 6.4766008e-01 8.1350386e-02
-3.6119869e-01 2.7051494e-01 8.3452100e-01 6.9003068e-02
3.2041091e-01 1.3960330e-01 -4.1672504e-01 -1.0403187e-01
-8.2072937e-01 2.4245532e-01 2.5291210e-01 7.8597575e-01
1.0709574e+00 3.9927933e-02 -3.5740495e-01 7.0606303e-01
1.1570444e-01 2.0454909e-01 2.0276953e-01 -8.8893217e-01
6.5509105e-01 6.8911332e-01 -4.8276949e-01 -1.4512893e+00
-1.0132852e-01 -1.8726647e-01 -8.9700031e-01 4.2376745e-02
4.6891525e-01 -7.1441583e-02 -8.4515798e-01 1.3225224e+00
5.0361872e-01 6.2441283e-01 -4.8603711e-01 9.5314378e-01
7.6059192e-01 4.4590569e-01 1.0787250e-01 4.6217520e-02
1.4686109e+00 -9.3248498e-01 -4.6175388e-01 5.9355553e-02
7.5076556e-01 -7.0975691e-01 1.2547385e+00 6.6096556e-01
-7.7862382e-01 -5.4165441e-01 -8.0174506e-01 -4.7541699e-01
-1.7214723e-01 2.3278587e-01 7.2354883e-01 7.6325524e-01
-5.9740371e-01 1.0514109e+00 -8.9145911e-01 1.6825332e-01
1.1482916e+00 2.7071047e-01 -3.1851497e-01 -1.5213229e-01
-9.7689134e-01 1.4036795e-01 3.3659500e-01 6.5267956e-01
-9.8866266e-01 -9.1360283e-01 -6.8381023e-01 1.8769328e-01
4.5101571e-01 -5.3027712e-02 1.0747520e+00 -6.9203997e-01
-1.4631257e+00 6.7323393e-01 1.3658239e-02 -2.1328157e-02
5.5424136e-01 -4.1829953e-01 -2.2744630e-01 4.2224011e-01
2.7393334e-02 1.8719395e-01 9.6097147e-01 -1.1299937e+00
-2.6824415e-01 -4.4081438e-01 2.0273088e-01 -5.2382129e-01
-7.8360903e-01 3.7084451e-01 -8.5855609e-01 -6.7716974e-01
-8.2951128e-02 -5.0845641e-01 2.4011986e-01 2.5532383e-01
-7.6999241e-01 -2.0976511e-01 1.1522211e+00 -1.1352013e+00
1.1663520e+00 -2.4336364e+00 -2.4805160e-01 1.2108456e-02
6.1554998e-01 5.7005280e-01 -2.0160171e-01 8.3004221e-02
-4.0122417e-01 2.4688564e-01 -2.0892753e-01 -6.4805281e-01
1.1886628e-02 -9.7758226e-02 -5.4182488e-01 7.7514619e-01
5.1296139e-01 9.7752869e-01 -8.3603954e-01 -5.7675594e-01
2.0070699e-01 4.6290436e-01 -6.5113991e-01 1.2607172e-01
-1.2351224e-01 5.4862052e-01 -4.3388823e-01 9.4089919e-01
1.4000660e+00 -6.8480350e-02 -1.4540911e-01 -6.0927302e-01
3.8916066e-03 2.3352830e-01 -9.2436689e-01 1.5541210e+00
-1.7214590e-01 5.6160492e-01 1.7019543e-01 -1.0530595e+00
7.1811855e-01 5.3151641e-02 3.5717186e-01 -8.3783954e-01
5.0987166e-01 -5.4361284e-02 4.7585271e-02 -9.2689264e-01
2.8041226e-01 5.8723066e-02 3.0403063e-01 5.9899724e-01
4.6728015e-02 5.5929047e-01 -7.9828016e-02 3.0321720e-01
1.1240897e+00 2.4017827e-01 -4.8471174e-01 1.4426312e-01
3.9223415e-01 -3.2248726e-01 9.5506018e-01 3.6257616e-01
-3.1914306e-01 6.5406954e-01 9.5907527e-01 -5.6391597e-01
-1.0635648e+00 -6.5206361e-01 2.2040233e-02 9.7435087e-01
2.4797256e-01 -5.9381831e-01 -1.0170242e+00 -9.5319670e-01
-7.5989671e-02 3.8545203e-01 -8.7210959e-01 -3.0949470e-01
-1.0023954e+00 -8.6737937e-01 9.4195163e-01 5.8521450e-01
1.0022804e+00 -1.1237073e+00 -2.6817584e-01 -1.2868854e-01
-1.6304646e-01 -1.2154833e+00 -7.1180785e-01 -4.8641953e-01
-7.4793965e-01 -1.3265846e+00 -2.1987627e-01 -2.8580061e-01
5.3970253e-01 6.1062175e-01 4.0132907e-01 7.2634751e-01
-4.3369022e-01 -1.3754949e-01 -5.0589198e-01 -1.0093331e-01
2.9507583e-01 -2.8353615e-02 -3.2687303e-01 4.8251575e-01
4.6863797e-01 -8.9290941e-01 -9.5263237e-01 1.3818268e-01
-1.1411865e+00 2.6386946e-02 6.1354977e-01 7.0750654e-01
2.2289833e-01 1.6054493e-01 4.7175609e-02 -9.1697413e-01
4.9840602e-01 -7.2597748e-01 -2.3777069e-01 6.3989107e-03
-4.2807475e-01 -1.9664636e-01 7.7275962e-01 -7.6076895e-01
-1.0178970e+00 -1.9468649e-01 -6.4491719e-02 -1.0763117e+00
-1.0964131e-01 1.8335851e-01 -5.1621526e-01 -1.7806354e-01
4.1866264e-01 3.6732736e-01 -1.2820072e-01 -8.0876124e-01
1.1157278e-01 4.8008406e-01 5.5523354e-01 -8.7734205e-01
1.0053784e+00 5.6608635e-01 -5.6670707e-02 -5.7781667e-01
-6.0826641e-01 -2.1432677e-01 -6.3072288e-01 -1.7323703e-01
6.5747333e-01 -5.5770653e-01 -1.2375671e+00 1.0551610e+00
-1.4479455e+00 -3.8123888e-01 1.5192899e-01 1.3449258e-01
-3.5460286e-02 1.0264343e+00 -1.0658044e+00 -7.0893514e-01
-5.6910235e-01 -1.2801058e+00 1.0681601e+00 3.5394493e-01
3.5704213e-01 -6.0632718e-01 -5.2037740e-01 5.2720153e-01
3.4819952e-01 3.7623528e-01 7.7662033e-01 -4.9962449e-01
-6.4080578e-01 -2.3741998e-01 -8.8719898e-01 6.2372303e-01
-1.4974658e-01 2.9030822e-02 -1.3909985e+00 -2.3850541e-01
3.0090341e-01 -2.1796742e-01 1.0483536e+00 -1.5209168e-01
1.9296659e+00 -4.3893522e-01 -8.0947138e-02 1.1600642e+00
1.2162004e+00 -1.0074568e-01 1.1898142e+00 4.9507779e-01
1.3662701e+00 5.8534068e-01 2.1831237e-01 4.5332390e-01
2.4256131e-01 4.6054780e-01 8.2653290e-01 8.9755073e-02
-2.4056785e-01 -2.3310477e-01 3.8323510e-01 4.7127509e-01
-2.2489285e-01 1.7030051e-01 -8.3902824e-01 4.2757189e-01
-1.6491771e+00 -1.0510906e+00 -4.3903294e-01 1.9026115e+00
8.3736962e-01 1.3601021e-01 5.9925582e-02 3.3629864e-01
8.5682839e-01 3.4827748e-01 -6.3839090e-01 -1.7331185e-02
1.6390337e-01 4.2993766e-01 5.7267207e-01 1.8546943e-01
-9.5453227e-01 1.0441563e+00 4.5188742e+00 1.3439622e+00
-1.5228924e+00 4.5505664e-01 6.7698944e-01 -1.4261484e-01
-5.9553310e-02 1.4427769e-01 -7.1693182e-01 8.9845037e-01
6.3146454e-01 4.6453175e-01 5.3996378e-01 7.6740587e-01
3.2132235e-01 2.6028058e-01 -7.4947065e-01 1.0782731e+00
5.8578234e-02 -1.3159621e+00 5.8462527e-02 2.8709170e-01
1.3565286e-01 -2.6810437e-01 1.4386605e-01 2.9429013e-01
-9.3721487e-02 -9.9016839e-01 7.5533402e-01 4.5802921e-01
8.3567291e-01 -8.2843608e-01 6.7152607e-01 1.4935274e-01
-1.0829840e+00 -1.8616170e-01 -5.3075993e-01 -7.4558362e-02
-5.3116906e-02 8.4801382e-01 -4.8753056e-01 5.9179324e-01
1.0018140e+00 7.2692323e-01 -6.5145713e-01 1.0200112e+00
-4.2862344e-01 9.7665352e-01 -1.5280007e-01 3.4805381e-01
1.8778698e-01 -1.0389340e-01 3.1715003e-01 1.3293698e+00
1.7010105e-01 1.3009042e-03 -1.7392617e-01 1.1487906e+00
-3.6336437e-01 -8.5871056e-02 -1.5015601e-01 8.5059166e-02
5.3909832e-01 1.6276855e+00 -5.9595382e-01 -1.9918782e-01
-3.4926984e-01 1.1644280e+00 5.2287996e-01 3.7129882e-01
-1.0426506e+00 -7.8772366e-01 7.1842539e-01 3.3120003e-01
4.2690715e-01 -1.2374799e-01 -5.6446892e-01 -1.3023475e+00
4.3670544e-01 -8.4728193e-01 2.3724301e-01 -6.1647379e-01
-1.2634184e+00 4.5542017e-01 -1.4596763e-01 -1.0828434e+00
4.9451655e-01 -9.4957113e-01 -1.1109142e+00 8.7660551e-01
-1.5833336e+00 -1.5327914e+00 -6.1365229e-01 7.5503904e-01
1.9871505e-01 7.9476513e-02 3.0528301e-01 7.5951642e-01
-1.2877935e+00 1.0095097e+00 -2.3693705e-01 8.3536851e-01
6.6880542e-01 -8.6494190e-01 5.0601792e-01 1.0687826e+00
-3.8489982e-01 1.0712602e+00 2.9113388e-01 -7.3351568e-01
-1.4946618e+00 -1.2867391e+00 3.1644553e-01 -3.7462032e-01
8.2261080e-01 -5.0466913e-01 -1.2941622e+00 6.7272234e-01
-2.1544790e-04 3.4306642e-01 4.5484781e-01 -5.6686215e-02
-1.1311162e+00 -1.9851908e-01 -1.1627980e+00 4.4721299e-01
1.0305727e+00 -7.4469000e-01 -1.8445846e-01 2.3983139e-01
6.9007099e-01 -3.2984000e-01 -7.0176589e-01 2.9547083e-01
4.8488879e-01 -1.1156745e+00 1.0012487e+00 -4.8825073e-01
7.1726340e-01 -3.9821079e-01 9.1792449e-02 -6.6769296e-01
-3.4937087e-01 -8.1669372e-01 -4.0409625e-01 1.4658617e+00
-1.4742362e-01 -5.8574277e-01 8.4155428e-01 3.8400623e-01
-3.1351882e-01 -9.6536416e-01 -8.2026005e-01 -6.6913748e-01
1.9029224e-01 -7.5719452e-01 9.6742421e-01 1.1156198e+00
-2.4658337e-01 1.5121334e-02 -5.7751852e-01 3.2750204e-01
8.2102287e-01 -1.2116397e-01 8.8077044e-01 -8.5109317e-01
-3.7371498e-01 -6.8161213e-01 -4.2665169e-01 -1.0602793e+00
8.2608111e-02 -7.3081160e-01 -2.7304858e-01 -8.9835089e-01
4.9305993e-01 -3.5136271e-01 -1.4092696e-01 8.4489357e-01
-4.1348320e-01 5.7182771e-01 2.8338352e-01 4.0687978e-01
-3.8561112e-01 4.6932518e-01 1.3310416e+00 -1.4000711e-01
1.8350188e-01 -3.9743865e-01 -8.6202782e-01 6.1037517e-01
8.5326546e-01 -5.6506646e-01 -1.1696799e-01 -7.7840883e-01
3.1734478e-02 -5.2236331e-01 8.6424893e-01 -8.8801283e-01
3.1358248e-01 -7.2585002e-02 5.5256009e-01 -4.8421818e-01
2.4953046e-01 -4.6594155e-01 -1.1971275e-01 3.8972190e-01
1.3357626e-01 -1.3943046e-01 3.6409101e-01 3.2666540e-01
-5.4889582e-03 -3.0369288e-01 7.3286313e-01 1.8738059e-02
-5.3033942e-01 7.2882175e-01 1.6862296e-01 -2.9338717e-01
7.7487445e-01 -3.6725995e-01 -7.1527797e-01 -2.8746143e-02
-3.3381665e-01 3.8649593e-02 3.7690747e-01 2.2767842e-01
6.7313766e-01 -1.2865044e+00 -5.5363774e-01 3.0541009e-01
-8.2099535e-02 2.3423582e-01 8.4179753e-01 1.1326047e+00
-7.1806794e-01 -2.3436628e-01 -2.5364003e-01 -3.4779117e-01
-1.0584526e+00 7.2283036e-01 2.6914346e-01 -7.6250978e-02
-7.0766455e-01 9.7261983e-01 2.7843636e-01 -3.7160441e-01
2.1572264e-01 -2.0317996e-02 -1.9102281e-02 -1.0250093e-01
9.0659869e-01 5.9905064e-01 -1.6232629e-01 -6.5074223e-01
-1.6751117e-01 5.6670219e-01 -4.4683817e-01 3.2072496e-01
1.5410111e+00 2.2899894e-01 -5.1890224e-01 -2.2748579e-01
1.4758846e+00 6.8396136e-02 -1.4650412e+00 -1.8948649e-01
-1.7287537e-01 -7.8410608e-01 1.6936624e-01 -4.0039951e-01
-1.7750460e+00 1.4120553e+00 4.6253693e-01 -5.5936337e-02
1.3345332e+00 -2.6164737e-01 1.2643536e+00 9.2481472e-02
4.9689502e-02 -6.8524313e-01 3.1369826e-01 3.8365075e-01
6.7540228e-01 -9.2913288e-01 3.2934215e-02 -6.0577554e-01
-4.4407585e-01 1.2664479e+00 8.9242250e-01 -3.9748305e-01
6.2455523e-01 1.4771578e-01 -1.9939415e-01 -3.9843065e-01
-2.3236245e-01 1.6788235e-01 -5.2450411e-02 4.8734587e-01
1.9673775e-01 -1.8384188e-01 -6.3930191e-02 8.5455966e-01
-9.1035374e-02 -1.2353398e-01 2.4345778e-01 8.1719202e-01
-4.7218330e-02 -9.4606268e-01 -3.9276254e-01 4.2615744e-01
-6.8759251e-01 -2.5365859e-01 -3.2367173e-01 5.3690362e-01
6.5382653e-01 8.1758875e-01 -6.3950956e-02 -9.1912144e-01
3.2399333e-01 -2.5073633e-01 2.3677042e-01 -3.7967026e-01
-7.4279755e-01 -7.8873515e-02 -2.2124256e-01 -9.4887125e-01
8.0698080e-02 -8.3290774e-01 -1.1090388e+00 -8.6085773e-01
-4.1011775e-01 -1.2175686e-01 3.8408732e-01 9.2318779e-01
5.3893173e-01 4.0733659e-01 7.0771396e-01 -1.0091851e+00
-4.7603488e-01 -1.2580389e+00 -4.4489622e-01 6.5074652e-01
2.4758193e-01 -6.5553123e-01 -3.7631533e-01 2.9138021e-02] | [12.410929679870605, 0.957928478717804] |
76bf607e-c259-4b80-803c-7f38581069b9 | detecting-frames-in-news-headlines-and-lead | null | null | https://aclanthology.org/2021.findings-emnlp.339 | https://aclanthology.org/2021.findings-emnlp.339.pdf | Detecting Frames in News Headlines and Lead Images in U.S. Gun Violence Coverage | News media structure their reporting of events or issues using certain perspectives. When describing an incident involving gun violence, for example, some journalists may focus on mental health or gun regulation, while others may emphasize the discussion of gun rights. Such perspectives are called “frames” in communication research.We study, for the first time, the value of combining lead images and their contextual information with text to identify the frame of a given news article. We observe that using multiple modes of information(article- and image-derived features) improves prediction of news frames over any single mode of information when the images are relevant to the frames of the headlines.We also observe that frame image relevance is related to the ease of conveying frames via images, which we call frame concreteness. Additionally, we release the first multimodal news framing dataset related to gun violence in the U.S., curated and annotated by communication researchers. The dataset will allow researchers to further examine the use of multiple information modalities for studying media framing. | ['Derry Tanti Wijaya', 'Prakash Ishwar', 'Margrit Betke', 'Hengchang Hu', 'Sha Lai', 'Boqi Chen', 'Mona Jalal', 'Edward Edberg Halim', 'Fabian Zhafransyah', 'Taufiq Husada Daryanto', 'Lei Guo', 'Isidora Tourni'] | null | null | null | null | findings-emnlp-2021-11 | ['multimodal-text-and-image-classification', 'news-classification', 'news-annotation'] | ['methodology', 'natural-language-processing', 'natural-language-processing'] | [ 6.02687597e-01 2.73317188e-01 -6.16780102e-01 -4.07240033e-01
-7.65664339e-01 -5.94573200e-01 1.20963502e+00 9.67020750e-01
-4.31432843e-01 5.16335130e-01 1.48807180e+00 -1.35068446e-01
8.88456702e-02 -7.55540490e-01 -6.33492410e-01 -2.33287469e-01
2.14143530e-01 -7.86595270e-02 3.53847304e-03 -2.26765946e-01
5.04440248e-01 1.23464867e-01 -1.18295658e+00 1.06286466e+00
9.69786197e-02 6.41598880e-01 3.42523992e-01 3.45084161e-01
-3.23697418e-01 1.34509218e+00 -7.09501088e-01 -6.75074637e-01
-2.75567889e-01 -5.02436042e-01 -9.96337056e-01 7.11019635e-02
5.38198411e-01 -5.90392590e-01 -4.09414291e-01 8.59861493e-01
3.53442848e-01 -1.26566321e-01 4.69465107e-01 -9.89792764e-01
-5.56389987e-01 1.11561584e+00 -5.84026039e-01 8.23614299e-01
1.08883154e+00 -2.77272344e-01 8.53457153e-01 -6.33654356e-01
1.50747097e+00 1.72516942e+00 5.78161836e-01 2.68705606e-01
-9.98294175e-01 -4.26846594e-01 3.33130628e-01 2.18977213e-01
-7.94630468e-01 -6.83128059e-01 7.97459424e-01 -9.04026568e-01
5.10465801e-01 3.31781119e-01 5.67646801e-01 1.65470946e+00
1.77577674e-01 5.24349391e-01 1.03488648e+00 -4.36001122e-01
-2.50754684e-01 -9.01413634e-02 2.99396425e-01 5.93397677e-01
1.87079296e-01 -1.83932394e-01 -7.17223525e-01 -3.53330433e-01
3.85581136e-01 2.83430070e-01 -1.71344340e-01 5.10469198e-01
-1.36200488e+00 1.18448353e+00 2.26815134e-01 5.15681624e-01
-4.50245291e-01 -1.61405671e-02 7.97658086e-01 -1.07491687e-02
9.90691960e-01 2.13481307e-01 2.88711220e-01 -4.39044952e-01
-5.71865022e-01 1.64103389e-01 5.97683012e-01 4.82361764e-01
4.42950934e-01 -5.13334870e-01 -5.13803422e-01 6.96622670e-01
2.33840525e-01 4.18480068e-01 -3.56289297e-01 -8.15445423e-01
1.01182568e+00 4.68098074e-01 1.14434637e-01 -1.76257968e+00
-7.16902316e-01 2.54904091e-01 -1.91944048e-01 -4.34513092e-01
5.19480705e-01 -4.40225452e-01 -3.28366756e-01 1.72020316e+00
4.57723379e-01 1.32527202e-02 -1.81527868e-01 8.33691478e-01
1.20414078e+00 8.37422371e-01 5.20712316e-01 -4.73247766e-01
2.01212263e+00 -2.57207960e-01 -1.06069219e+00 -2.49655813e-01
5.43219388e-01 -9.69033659e-01 5.67965806e-01 -1.51901454e-01
-1.05808067e+00 -1.15307599e-01 -5.59562743e-01 -8.78513306e-02
-2.43999332e-01 -2.50159681e-01 2.09321082e-01 2.65758902e-01
-4.05900627e-01 2.61710614e-01 -4.96176749e-01 -9.27134812e-01
6.53632045e-01 -6.59740806e-01 -4.74819809e-01 -9.55373570e-02
-1.26907218e+00 1.15567124e+00 1.99483931e-01 -4.07937139e-01
-6.40134037e-01 -5.86258173e-01 -9.99818265e-01 -4.84878570e-01
5.34824669e-01 -2.85158008e-01 9.70155239e-01 -1.02103066e+00
-7.78340816e-01 1.03438282e+00 -4.10390079e-01 -3.79348904e-01
4.12454575e-01 -3.52690816e-01 -6.65369391e-01 8.69214892e-01
6.09098971e-01 4.15602505e-01 9.53361690e-01 -1.31602049e+00
-6.66901410e-01 -4.20784622e-01 8.01282108e-01 1.53459966e-01
-2.94506311e-01 9.78560925e-01 -2.37164777e-02 -7.73601294e-01
-2.28881985e-01 -7.22078383e-01 1.69889495e-01 -3.07031721e-01
-5.32293737e-01 1.21120811e-01 9.92466331e-01 -1.03705537e+00
1.44257128e+00 -2.23056293e+00 9.01922882e-02 -1.29609406e-01
4.13741976e-01 -3.75392079e-01 1.74871221e-01 1.08288479e+00
8.41698423e-03 4.24823076e-01 8.55144784e-02 -2.25856468e-01
-2.07950249e-01 9.79861841e-02 -5.05433977e-01 8.04985404e-01
-4.35189903e-02 6.52280986e-01 -1.05926383e+00 -7.20327616e-01
4.12767157e-02 6.30759716e-01 -2.62539148e-01 -3.39378089e-01
1.55078527e-02 7.30881333e-01 -4.99980211e-01 3.76089424e-01
3.26311529e-01 -2.57778049e-01 2.15975553e-01 -5.07391036e-01
-5.48200250e-01 1.97454154e-01 -6.13658369e-01 1.13299429e+00
-1.80948347e-01 1.22508812e+00 8.09465572e-02 -7.30580926e-01
4.23345894e-01 6.01179361e-01 4.24172819e-01 -6.38234437e-01
1.92977786e-01 -5.09704232e-01 -4.21082586e-01 -8.29101264e-01
7.77500689e-01 -1.51951835e-01 -3.33748996e-01 7.05927610e-01
-3.49650443e-01 3.88701200e-01 3.23081523e-01 4.84764397e-01
7.94834375e-01 -3.32548022e-01 3.97054046e-01 -4.62397411e-02
-5.83761223e-02 1.82126701e-01 1.56400412e-01 8.14715207e-01
-1.60723999e-01 5.97125888e-01 8.63617837e-01 -6.14150643e-01
-1.13165057e+00 -4.95656580e-01 -2.48493642e-01 1.29051673e+00
3.11112940e-01 -8.62789750e-01 -8.81458759e-01 -5.63706338e-01
-3.73994380e-01 8.21497142e-01 -9.53351498e-01 2.38649070e-01
-6.83616936e-01 -3.96889627e-01 4.45171982e-01 2.17238232e-01
1.99776635e-01 -7.75578976e-01 -1.05211627e+00 1.69196859e-01
-8.01360607e-01 -1.50239444e+00 -3.55864197e-01 -6.03803575e-01
-2.56824404e-01 -1.26263416e+00 -4.70174372e-01 -2.04024777e-01
6.82684958e-01 3.47989500e-01 6.66242480e-01 2.82655805e-01
-5.00700995e-03 8.64212573e-01 -8.69292319e-01 -5.11021614e-01
-5.71442187e-01 -2.13805243e-01 -2.50770628e-01 2.40172595e-01
-9.37328339e-02 -3.56474668e-02 -3.72022808e-01 5.06915636e-02
-1.24531138e+00 3.76766920e-01 -1.97910234e-01 4.23811942e-01
6.03210330e-02 -4.30048257e-01 1.59274712e-01 -1.07512724e+00
8.22041094e-01 -1.02545130e+00 9.17560905e-02 1.84459463e-01
2.03535229e-01 -4.82793838e-01 3.48374955e-02 -3.07423234e-01
-1.18755269e+00 -5.20348907e-01 1.32864967e-01 1.58406213e-01
-2.61170149e-01 7.74303377e-01 4.69222456e-01 4.61850554e-01
8.28875124e-01 -3.69230002e-01 -1.99538767e-01 -1.96666628e-01
2.60042101e-01 2.66338885e-01 2.14255422e-01 -6.08469069e-01
3.94075811e-01 9.40486848e-01 -1.33499563e-01 -1.08633840e+00
-1.22150791e+00 -3.17905128e-01 -3.33681375e-01 -1.01261127e+00
1.27463102e+00 -8.11601460e-01 -5.25413692e-01 1.05820410e-01
-1.64399612e+00 1.92084432e-01 8.27170983e-02 5.75945854e-01
-3.70370597e-01 4.04207498e-01 -7.25504935e-01 -7.71693289e-01
1.48686633e-01 -1.07253134e+00 1.07307863e+00 -8.03452358e-02
-5.73155224e-01 -1.04490507e+00 -7.00271204e-02 5.73090851e-01
1.63727000e-01 1.01291013e+00 6.26202583e-01 -6.30835116e-01
2.94576809e-02 -1.04763910e-01 -3.34188193e-01 -5.07110000e-01
1.17296778e-01 -1.22389067e-02 -9.16622519e-01 8.95465687e-02
-1.00419551e-01 -1.52982324e-01 7.53616095e-01 4.91037816e-01
5.33734262e-01 -8.40980828e-01 -4.74265367e-01 -1.83363482e-01
1.20604897e+00 2.34800681e-01 5.14046311e-01 6.63207114e-01
6.03551388e-01 9.10481453e-01 4.25543159e-01 7.70504475e-01
6.30279362e-01 8.16288650e-01 2.06695423e-01 -1.21423759e-01
-9.90155432e-03 -4.26935226e-01 3.80394578e-01 2.11933300e-01
-3.66304398e-01 -4.32168186e-01 -1.25637829e+00 2.89207011e-01
-1.82915747e+00 -1.76783144e+00 -3.00185412e-01 1.43809998e+00
4.44668859e-01 -3.82786170e-02 2.83649981e-01 -7.73019865e-02
1.24941802e+00 6.59617245e-01 2.33841091e-01 -3.07909846e-01
-2.17744827e-01 -5.87615013e-01 3.29283029e-01 7.86626160e-01
-1.34309256e+00 6.65566742e-01 6.59962988e+00 6.38763726e-01
-8.30104172e-01 6.41042173e-01 9.25247252e-01 8.99171755e-02
-4.24848735e-01 1.36553496e-01 -7.31265187e-01 4.78813320e-01
8.25691581e-01 6.28327439e-03 8.78215507e-02 3.78483206e-01
5.70575953e-01 -5.92976987e-01 -1.01902819e+00 6.75365508e-01
3.47018003e-01 -1.91712952e+00 -6.57570511e-02 -7.50600453e-03
4.23203111e-01 -4.07760173e-01 -1.54469684e-02 -4.54819918e-01
-2.84167439e-01 -7.50588238e-01 1.59928966e+00 8.41967404e-01
5.97127438e-01 -5.16795754e-01 6.01554811e-01 1.21829852e-01
-9.99848664e-01 -1.04705215e-01 8.83589238e-02 -4.42721874e-01
9.54138875e-01 4.81998980e-01 -8.12229097e-01 3.26112628e-01
4.69233453e-01 8.81704390e-01 -4.79347318e-01 4.42879140e-01
-1.10397547e-01 6.61050677e-01 -8.32386240e-02 2.12111119e-02
2.89518595e-01 1.74488977e-01 9.47038531e-01 1.63475037e+00
2.31662259e-01 5.44116020e-01 3.30862015e-01 4.97757405e-01
2.06214219e-01 9.69146490e-02 -9.51993883e-01 -1.77769393e-01
5.31290889e-01 9.29028511e-01 -1.06805289e+00 -3.60273033e-01
-5.56002796e-01 4.14923429e-01 1.79847300e-01 2.24941820e-01
-8.31215739e-01 2.28774369e-01 3.05163950e-01 5.01563847e-01
-1.75088659e-01 -1.66567653e-01 -2.44879633e-01 -9.34916914e-01
-3.83160233e-01 -5.69610536e-01 5.16098320e-01 -9.11060572e-01
-1.12874031e+00 5.39868772e-01 7.61707842e-01 -9.57720041e-01
1.29344156e-02 -6.76372498e-02 -4.63628590e-01 2.43533358e-01
-1.04488504e+00 -1.22476649e+00 -1.01429723e-01 4.38090950e-01
5.97300589e-01 2.74537235e-01 4.80252355e-01 1.21031664e-01
-5.57008564e-01 -3.22613493e-02 -4.04180408e-01 3.43087912e-01
6.07346535e-01 -5.21228254e-01 6.08913302e-02 7.49803126e-01
3.88436094e-02 4.21759874e-01 1.15907383e+00 -1.06371856e+00
-1.19572353e+00 -7.19694614e-01 1.20046234e+00 -5.89651585e-01
9.30305541e-01 -1.70957521e-01 -3.94660264e-01 7.93362617e-01
6.55625582e-01 -5.22649169e-01 8.86052847e-01 1.25466019e-01
-3.46552819e-01 4.65670675e-01 -1.32762301e+00 6.73768163e-01
1.14270818e+00 -6.62531078e-01 -8.95304859e-01 6.71478033e-01
6.87125504e-01 -5.32787025e-01 -8.47034097e-01 -1.26622915e-02
4.56801027e-01 -7.94988751e-01 9.15939391e-01 -5.70394158e-01
1.06259990e+00 1.27647623e-01 -3.07464868e-01 -9.94736314e-01
-2.78896362e-01 -2.38872185e-01 2.30939820e-01 1.52608943e+00
5.33159435e-01 -4.00864780e-01 3.22300829e-02 7.40960300e-01
-9.73517448e-02 -8.04755837e-02 -9.47205663e-01 -6.91290572e-02
-4.07712162e-01 -8.51218283e-01 3.65891486e-01 1.34642828e+00
6.25278175e-01 2.14217842e-01 -5.87228656e-01 1.25259861e-01
2.40833148e-01 -1.44491881e-01 4.72812384e-01 -9.82142806e-01
2.25237697e-01 -2.13392675e-01 -4.81728435e-01 -4.10635591e-01
1.64811656e-01 -6.84563935e-01 -4.95913714e-01 -1.77532744e+00
6.09888017e-01 8.41329172e-02 4.36853707e-01 3.43944311e-01
7.63563737e-02 2.81886071e-01 6.64967716e-01 2.27542013e-01
-7.19286442e-01 -2.11760834e-01 1.30157924e+00 -6.68001622e-02
2.09435523e-01 -6.33643210e-01 -8.72818530e-01 1.04951692e+00
6.28085017e-01 -4.58745688e-01 -8.87394920e-02 -5.59525192e-01
5.90994656e-01 5.49647510e-01 6.62741721e-01 -5.56337416e-01
1.52177602e-01 -6.31114602e-01 3.91508341e-01 -4.09876734e-01
3.81900698e-01 -7.55623162e-01 5.36477685e-01 3.17968547e-01
-7.74461567e-01 3.42606127e-01 2.41407350e-01 6.98381305e-01
-2.12592065e-01 -2.82453418e-01 4.69155282e-01 -1.96568802e-01
-5.57790101e-01 -9.95004103e-02 -9.46282983e-01 2.39560142e-01
1.16801012e+00 -2.22777128e-01 -1.00928092e+00 -9.24299657e-01
-9.55931187e-01 -2.34312654e-01 3.24306637e-01 5.32063723e-01
6.55663729e-01 -1.42386055e+00 -9.21878219e-01 -4.40054417e-01
1.61131322e-01 -6.95123017e-01 4.39220041e-01 1.11284244e+00
-3.67286295e-01 1.82243332e-01 -2.63143051e-02 -4.51244473e-01
-1.33053362e+00 4.16811258e-01 -2.86917984e-01 1.83180720e-01
-9.65340853e-01 4.72493410e-01 2.02238500e-01 3.72311980e-01
-1.05315752e-01 -1.30137220e-01 -8.22063684e-01 9.06586170e-01
1.11923361e+00 4.79990274e-01 -4.96329725e-01 -1.61142087e+00
-3.05786848e-01 3.14300001e-01 1.62856102e-01 -3.81954342e-01
1.56045318e+00 -5.68062365e-01 -1.90474957e-01 6.24473810e-01
1.01111257e+00 2.53662080e-01 -9.17882502e-01 -2.23364145e-01
1.07328244e-01 -7.43794918e-01 -1.80711597e-01 -5.97256780e-01
-7.45620549e-01 4.30237830e-01 1.72760725e-01 7.47051716e-01
6.06926024e-01 5.68661869e-01 5.45174301e-01 -1.14932984e-01
2.40978613e-01 -1.10039473e+00 7.95223787e-02 5.33151746e-01
1.35451317e+00 -1.12389660e+00 1.58598110e-01 -5.15199840e-01
-9.90374744e-01 1.50876522e+00 1.65161043e-01 1.82187587e-01
7.32217252e-01 3.07601690e-01 -6.63105398e-02 -6.45370007e-01
-5.50492644e-01 -7.71178156e-02 2.98510641e-01 4.00094539e-01
5.08688927e-01 1.68148234e-01 -7.18192458e-01 5.35156250e-01
-8.52510780e-02 -3.54024202e-01 6.60219073e-01 9.82447624e-01
-5.88701367e-01 -5.99632263e-01 -9.41562355e-01 3.76749188e-01
-1.02140737e+00 -1.12779960e-02 -5.79493523e-01 5.96225679e-01
1.98662698e-01 1.31745315e+00 2.01576531e-01 -2.11921930e-01
2.34647587e-01 -6.43156245e-02 2.65694648e-01 -5.23722053e-01
-6.40675843e-01 -1.83813814e-02 9.20681953e-01 -4.16341513e-01
-1.17263758e+00 -1.02654421e+00 -1.01070058e+00 -4.30998862e-01
8.25441815e-03 -7.66837001e-02 6.09389842e-01 1.41760099e+00
2.51346976e-01 2.33331725e-01 2.68509150e-01 -9.47294116e-01
5.04039705e-01 -7.47256935e-01 -5.44530191e-02 5.85408509e-01
5.36623478e-01 -7.27780282e-01 -7.51769310e-03 4.41911221e-01] | [8.73755168914795, 9.890167236328125] |
e092c8ab-67b7-4d57-ac9e-99c3c36da380 | reading-chinese-in-natural-scenes-with-a-bag | 2210.02576 | null | https://arxiv.org/abs/2210.02576v1 | https://arxiv.org/pdf/2210.02576v1.pdf | Reading Chinese in Natural Scenes with a Bag-of-Radicals Prior | Scene text recognition (STR) on Latin datasets has been extensively studied in recent years, and state-of-the-art (SOTA) models often reach high accuracy. However, the performance on non-Latin transcripts, such as Chinese, is not satisfactory. In this paper, we collect six open-source Chinese STR datasets and evaluate a series of classic methods performing well on Latin datasets, finding a significant performance drop. To improve the performance on Chinese datasets, we propose a novel radical-embedding (RE) representation to utilize the ideographic descriptions of Chinese characters. The ideographic descriptions of Chinese characters are firstly converted to bags of radicals and then fused with learnable character embeddings by a character-vector-fusion-module (CVFM). In addition, we utilize a bag of radicals as supervision signals for multi-task training to improve the ideographic structure perception of our model. Experiments show performance of the model with RE + CVFM + multi-task training is superior compared with the baseline on six Chinese STR datasets. In addition, we utilize a bag of radicals as supervision signals for multi-task training to improve the ideographic structure perception of our model. Experiments show performance of the model with RE + CVFM + multi-task training is superior compared with the baseline on six Chinese STR datasets. | ['Wang Yunhong', 'Chen Jiaxin', 'Liu Qingjie', 'Liu Yongbin'] | 2022-10-05 | null | null | null | null | ['scene-text-recognition'] | ['computer-vision'] | [ 5.36318421e-02 -6.68935716e-01 -1.71400785e-01 -3.60680282e-01
-9.72199976e-01 -4.04102087e-01 6.51326835e-01 -1.67212367e-01
-5.26990831e-01 3.64582688e-02 6.45708263e-01 1.37533452e-02
6.33275807e-01 -5.40563822e-01 -7.13509142e-01 -7.32556283e-01
6.21637583e-01 1.68854788e-01 1.24371171e-01 -2.80555159e-01
5.77175617e-01 1.28407374e-01 -1.05799246e+00 9.19448018e-01
1.10035872e+00 7.10252345e-01 3.26003343e-01 5.48908830e-01
-4.40389037e-01 7.70793319e-01 -6.73918903e-01 -4.82931107e-01
1.57297939e-01 -1.50465190e-01 -3.60299706e-01 2.32028097e-01
8.05372536e-01 -1.67641759e-01 -3.12560797e-01 7.51963079e-01
7.62401938e-01 -2.57027298e-01 7.61825085e-01 -8.11851263e-01
-1.47400570e+00 8.41392875e-01 -8.88729692e-01 -1.62244335e-01
2.38883823e-01 -1.80274785e-01 1.15444410e+00 -1.50284088e+00
5.83925426e-01 1.51188433e+00 7.51922488e-01 5.36453724e-01
-9.87136841e-01 -8.53936732e-01 4.41137701e-01 6.20533526e-01
-1.49495184e+00 -2.67257750e-01 1.10960484e+00 -3.25676978e-01
9.23370659e-01 1.18406206e-01 2.49148771e-01 1.56577289e+00
2.03671426e-01 1.37166786e+00 8.85889232e-01 -5.70053697e-01
-1.82796165e-01 9.08245966e-02 2.94001818e-01 8.44921470e-01
9.23175812e-02 -5.80849648e-01 -8.01113665e-01 2.98273653e-01
4.19546664e-01 -2.60921232e-02 -1.34912238e-01 2.76614930e-02
-1.38649035e+00 8.61372888e-01 3.86914819e-01 4.24045622e-01
3.91702279e-02 1.19599337e-02 5.70446968e-01 7.97519386e-02
5.71858287e-01 1.64687157e-01 -1.25326037e-01 4.22452725e-02
-6.73425555e-01 8.74357000e-02 4.00951087e-01 9.68195736e-01
6.37673438e-01 3.77265871e-01 -4.25596654e-01 1.39264870e+00
2.15661928e-01 8.15516233e-01 8.81839156e-01 -2.90755004e-01
1.21082306e+00 6.59792721e-01 -4.75717485e-01 -1.08585691e+00
-2.31308147e-01 -4.45327520e-01 -9.23806906e-01 -4.16730464e-01
1.37793139e-01 3.34355682e-02 -8.80396068e-01 1.38101423e+00
-3.94725353e-02 -5.08091822e-02 2.83745378e-01 7.45998979e-01
9.03710306e-01 1.05000925e+00 -1.58336952e-01 2.46356264e-01
1.30073428e+00 -1.49377477e+00 -7.07084775e-01 -5.20783603e-01
1.09311593e+00 -1.04102099e+00 1.40001118e+00 4.99844402e-01
-4.02259052e-01 -9.57470536e-01 -9.89130914e-01 -5.79648674e-01
-4.10686314e-01 1.00403130e+00 3.21873009e-01 8.61976743e-01
-6.04205132e-01 4.01625447e-02 -5.15583098e-01 -3.25216562e-01
4.28705156e-01 -1.17880277e-01 -3.77907306e-01 -4.31479424e-01
-8.01631331e-01 6.90276802e-01 2.47889146e-01 1.80322528e-01
-6.65503204e-01 -5.19658327e-01 -1.07082593e+00 -1.01215966e-01
1.48218006e-01 -7.36738788e-03 8.82083833e-01 -7.69059896e-01
-1.42473602e+00 6.83231592e-01 -1.37908906e-01 5.02360798e-02
4.80241209e-01 -4.09739614e-01 -5.42294025e-01 1.48872465e-01
2.65297592e-01 7.88684487e-01 9.05592084e-01 -1.36894810e+00
-5.00881910e-01 -3.21023464e-01 -5.36425591e-01 3.54773134e-01
-9.23153102e-01 1.29415274e-01 -7.98370898e-01 -1.06109667e+00
2.26941198e-01 -9.67361152e-01 1.05931543e-01 -1.07333057e-01
-6.71077609e-01 -4.54914510e-01 1.16956258e+00 -9.24930096e-01
1.36946225e+00 -2.47203898e+00 8.72655958e-02 -3.26825052e-01
-7.74045363e-02 2.91796416e-01 -6.27859354e-01 3.76866549e-01
1.63578335e-02 2.34825969e-01 -9.26981270e-02 -8.09854567e-01
1.64577469e-01 -1.20975506e-02 -4.99557078e-01 4.09320563e-01
4.85940874e-01 8.96562040e-01 -5.10507882e-01 -6.40659094e-01
1.71154633e-01 4.31994379e-01 -5.03062248e-01 -3.30632217e-02
-1.77046657e-01 1.17981531e-01 -6.31911755e-01 7.99761951e-01
9.52416241e-01 -9.33018178e-02 1.35507494e-01 -3.58331054e-01
-3.57562900e-01 4.82762866e-02 -1.01121747e+00 1.69088447e+00
-4.56595093e-01 8.06155026e-01 -6.53810620e-01 -7.32985318e-01
1.22102845e+00 -4.44462933e-02 6.76599145e-02 -9.28441346e-01
-5.80435358e-02 2.22765312e-01 -6.91207424e-02 -6.99174583e-01
7.73847163e-01 1.71068966e-01 -3.56489360e-01 4.20495659e-01
-3.40196818e-01 -7.24119740e-03 1.68311074e-01 2.96841145e-01
9.29277956e-01 1.27651304e-01 -1.90952539e-01 3.01075801e-02
6.56849146e-01 -4.23549339e-02 6.19182110e-01 4.82147902e-01
9.94124785e-02 9.20709908e-01 5.08152962e-01 -4.32583272e-01
-1.16818702e+00 -8.44302416e-01 -1.78966403e-01 1.21715665e+00
1.69743244e-02 -6.66525304e-01 -6.40004635e-01 -7.53808022e-01
3.59098837e-02 6.46393239e-01 -6.54449105e-01 5.99419549e-02
-8.62103760e-01 -9.03423011e-01 9.79819298e-01 7.79880762e-01
8.94930482e-01 -1.04441786e+00 -2.59491742e-01 1.94434538e-01
-4.58658785e-01 -1.56630075e+00 -6.18545055e-01 -1.50427118e-01
-5.47960877e-01 -7.04770923e-01 -1.01469791e+00 -1.06976068e+00
5.46686709e-01 6.07197762e-01 3.95508170e-01 -1.37748316e-01
-1.81691661e-01 1.97539181e-02 -7.53138423e-01 -3.27935249e-01
-1.51909336e-01 1.19753256e-01 -1.41122848e-01 3.16465259e-01
3.69882345e-01 2.25362346e-01 -1.30466536e-01 2.49647781e-01
-8.71838570e-01 4.22463030e-01 7.25609541e-01 7.40688205e-01
2.99105436e-01 -2.52109885e-01 2.30538920e-01 -8.25627625e-01
3.78547549e-01 -1.31906480e-01 -2.25445196e-01 5.27737617e-01
-7.76130632e-02 -2.93056271e-03 9.77309227e-01 -6.88014209e-01
-1.29519343e+00 -1.43868014e-01 4.23413701e-02 -4.29872870e-01
1.17929444e-01 6.98064983e-01 -2.03854382e-01 1.36970669e-01
5.23469806e-01 4.93207633e-01 -4.22669858e-01 -8.26120913e-01
1.40576825e-01 1.11651099e+00 3.56288284e-01 -6.98380053e-01
6.33610070e-01 4.91831124e-01 -2.30414107e-01 -1.25925219e+00
-9.21863794e-01 -3.15971196e-01 -5.99737406e-01 4.48336303e-02
9.12350833e-01 -1.22625983e+00 -1.72363490e-01 1.07319832e+00
-1.21469319e+00 -4.40917283e-01 3.35927486e-01 4.41026717e-01
-1.59035191e-01 9.42227185e-01 -7.48905540e-01 -5.75501382e-01
-2.42153749e-01 -1.07382250e+00 1.44453037e+00 2.54622139e-02
2.58773625e-01 -9.48440433e-01 1.19644061e-01 7.24880576e-01
1.19251586e-01 -7.47826025e-02 1.13621092e+00 -5.29078484e-01
-4.91404921e-01 -1.75555810e-01 -7.01188028e-01 4.72105533e-01
4.10155542e-02 2.11957380e-01 -9.46291208e-01 -2.39796177e-01
-3.99318129e-01 -5.34756958e-01 1.23533320e+00 3.43707204e-02
1.35538435e+00 1.67562664e-02 -2.77386636e-01 7.08323956e-01
1.37872314e+00 -1.82867661e-01 6.77300513e-01 5.14795005e-01
1.35788083e+00 3.95564288e-01 6.32199705e-01 5.92255890e-01
9.42364812e-01 9.07952964e-01 -7.51879811e-02 2.18608957e-02
-5.26102841e-01 -4.03948128e-01 9.30172622e-01 1.29884803e+00
1.88688040e-01 -4.06006545e-01 -1.13486993e+00 4.09319788e-01
-1.89202631e+00 -9.10998642e-01 -5.19015670e-01 1.81046605e+00
7.45130420e-01 -9.62295290e-03 -1.41193300e-01 1.47796914e-01
7.56777883e-01 4.92123455e-01 -1.90348640e-01 -5.61810017e-01
-7.54082739e-01 -8.83265436e-02 3.06607604e-01 3.87814716e-02
-1.18587375e+00 1.45011342e+00 5.39131308e+00 1.18732584e+00
-1.51700878e+00 1.39345497e-01 3.64966810e-01 3.05548757e-01
-3.77492517e-01 -2.99326293e-02 -1.26520455e+00 4.73241061e-01
4.89794135e-01 6.68898821e-02 8.35159123e-02 7.43442416e-01
7.91549236e-02 6.59940764e-02 -9.60540712e-01 1.15593219e+00
9.92526531e-01 -1.24712968e+00 3.73901546e-01 7.58809224e-02
9.68926311e-01 8.78543593e-03 3.18197936e-01 5.52713513e-01
-4.00369391e-02 -8.48654449e-01 1.07053423e+00 3.37662041e-01
9.87565041e-01 -4.05577451e-01 5.58866739e-01 2.87408829e-01
-1.30282271e+00 -3.79337937e-01 -7.00724185e-01 1.55257255e-01
9.19826478e-02 4.12748396e-01 -5.09982944e-01 6.96081698e-01
4.41161066e-01 1.21588826e+00 -1.27121401e+00 8.41944277e-01
-4.42446381e-01 6.36546135e-01 -1.16718620e-01 -3.46901625e-01
5.85076630e-01 -1.80726156e-01 1.10905573e-01 1.70788848e+00
3.85262519e-01 -3.39962512e-01 3.45087737e-01 6.25317752e-01
-1.24726005e-01 5.16419709e-01 -6.09672487e-01 -1.95932508e-01
2.75507390e-01 1.20304489e+00 -4.10790354e-01 -4.33714539e-01
-6.31629407e-01 1.09306145e+00 5.64647913e-01 3.21953386e-01
-1.01616144e+00 -4.95279372e-01 1.39235750e-01 -4.28275853e-01
5.74622631e-01 -3.92503709e-01 -4.95259553e-01 -1.58969712e+00
2.96515733e-01 -1.10401082e+00 3.95837158e-01 -8.75967979e-01
-1.27709699e+00 4.47593391e-01 -3.01419854e-01 -1.37545502e+00
3.62016261e-01 -8.76371145e-01 -8.27060938e-01 6.71486259e-01
-1.55493057e+00 -1.82738256e+00 -3.88996661e-01 5.00866115e-01
1.10200059e+00 -5.39801836e-01 5.13495684e-01 3.52066904e-01
-1.17906427e+00 9.86592591e-01 3.64799321e-01 5.11887312e-01
1.16917777e+00 -9.45834637e-01 3.78134251e-01 1.03614998e+00
3.36439312e-01 3.46338451e-01 6.27219751e-02 -8.08144569e-01
-1.64601731e+00 -1.25002182e+00 1.19418180e+00 -3.84457946e-01
6.39993012e-01 -6.84021294e-01 -9.51971233e-01 5.81919312e-01
4.63969260e-02 -2.36036330e-01 7.36778200e-01 1.77569106e-01
-1.00359261e+00 -1.37930349e-01 -4.93771464e-01 8.78895819e-01
8.74500394e-01 -7.07647264e-01 -5.99091172e-01 3.40662420e-01
4.81809646e-01 -1.74551815e-01 -3.96180123e-01 2.41582811e-01
5.84246874e-01 -7.05341339e-01 7.56972790e-01 -2.84569651e-01
8.79895449e-01 -3.15384030e-01 -5.29759407e-01 -1.16182876e+00
-4.18115973e-01 -5.35103567e-02 2.65457630e-01 1.35857487e+00
3.08775574e-01 -3.02711666e-01 6.41547382e-01 -9.70200971e-02
-5.76774478e-01 -3.97024453e-01 -8.09646606e-01 -9.08187330e-01
3.54606807e-01 -5.04836917e-01 2.31516838e-01 1.18239081e+00
-6.10784255e-02 5.70945144e-01 -4.81000841e-01 9.92678404e-02
6.10074937e-01 1.17778324e-01 9.54051316e-01 -7.21186638e-01
-4.85068448e-02 -5.64151406e-01 -1.77195966e-01 -1.11839831e+00
4.47987944e-01 -1.13482761e+00 -1.82902366e-01 -1.55391169e+00
4.72506046e-01 -3.24789345e-01 -5.22447452e-02 6.07678175e-01
-3.03670496e-01 3.08698237e-01 6.39838517e-01 1.77376628e-01
-7.59226620e-01 9.03312325e-01 1.19687510e+00 -3.67241949e-01
7.75873661e-04 -4.60343033e-01 -5.09773672e-01 5.26011407e-01
5.43236494e-01 -3.21338803e-01 1.07960381e-01 -1.10090411e+00
2.54982114e-01 -3.91679406e-01 2.15561062e-01 -1.08389616e+00
2.62085378e-01 4.82575595e-02 6.91050053e-01 -1.01154125e+00
4.73938733e-01 -3.35170805e-01 -3.38187814e-01 3.76126319e-01
-2.56603301e-01 1.19879976e-01 1.51369929e-01 3.32876921e-01
-2.06656232e-01 -2.51974583e-01 5.89784026e-01 2.77265042e-01
-8.08008969e-01 7.52398893e-02 -3.85783881e-01 -6.24208301e-02
7.85076857e-01 -5.39745927e-01 -8.18547666e-01 -1.11027882e-02
-6.83315322e-02 2.75025994e-01 4.05786365e-01 7.86855698e-01
8.11928749e-01 -1.49867666e+00 -1.04880750e+00 2.90567875e-01
5.12939155e-01 -2.01419771e-01 4.70555335e-01 7.32512474e-01
-7.37567961e-01 4.21742588e-01 -2.40522668e-01 -5.84067464e-01
-1.30421400e+00 4.50563550e-01 -2.36511722e-01 -3.88173550e-01
-6.16644561e-01 6.82820737e-01 2.75166512e-01 -5.27824163e-01
1.58548385e-01 -1.44204944e-01 -3.44318271e-01 2.29506746e-01
5.01663864e-01 4.53160256e-01 -1.75842017e-01 -8.34580958e-01
-3.79190922e-01 1.17075133e+00 -4.10734504e-01 -1.97808027e-01
1.41308343e+00 1.35284811e-01 -3.44783161e-03 4.88495678e-01
1.32789838e+00 4.50665921e-01 -1.04545200e+00 -4.25626516e-01
4.47713733e-02 -5.74044943e-01 -1.20074525e-01 -5.44580638e-01
-8.25746894e-01 1.43259835e+00 3.81516606e-01 -4.58532572e-01
8.99497867e-01 -3.08716357e-01 7.78991580e-01 6.95658684e-01
1.12189956e-01 -1.36075211e+00 7.41014481e-01 9.38531935e-01
9.76684928e-01 -1.27574503e+00 -6.76235631e-02 -5.59865952e-01
-1.30182672e+00 1.40305007e+00 6.67777419e-01 -1.51355132e-01
3.54386717e-01 1.69679429e-02 3.62167954e-01 2.01715350e-01
-7.86325753e-01 -5.96862473e-02 3.50875288e-01 4.31847602e-01
5.89354515e-01 4.05353755e-02 -2.63205975e-01 6.36332154e-01
-1.32795215e-01 -3.50506306e-01 7.10252643e-01 7.63323307e-01
-3.92442793e-01 -1.27278781e+00 -4.62542355e-01 1.86188459e-01
-2.03185797e-01 -3.08212101e-01 -4.45454568e-01 5.74798405e-01
6.36405572e-02 8.12321186e-01 1.97004214e-01 -5.43327212e-01
2.10790545e-01 1.60949584e-02 4.07769680e-01 -8.18485260e-01
-6.93046212e-01 3.54979783e-01 1.06160179e-01 -1.29750028e-01
4.18334529e-02 -7.54111290e-01 -1.02028906e+00 -1.75229967e-01
-2.06220388e-01 -2.68033624e-01 8.41239095e-01 9.01183724e-01
2.00559884e-01 2.85450459e-01 9.97560084e-01 -7.28713930e-01
-5.92856288e-01 -1.11327922e+00 -4.65325207e-01 4.57594007e-01
-7.13161752e-02 -3.29644680e-01 -2.06854075e-01 -4.94595291e-03] | [11.767202377319336, 2.0599093437194824] |
7c881c29-ace5-41fe-ac15-d15839c232f7 | the-sweet-home-speech-and-multimodal-corpus | null | null | https://aclanthology.org/L14-1125 | https://aclanthology.org/L14-1125.pdf | The Sweet-Home speech and multimodal corpus for home automation interaction | Ambient Assisted Living aims at enhancing the quality of life of older and disabled people at home thanks to Smart Homes and Home Automation. However, many studies do not include tests in real settings, because data collection in this domain is very expensive and challenging and because of the few available data sets. The S WEET-H OME multimodal corpus is a dataset recorded in realistic conditions in D OMUS, a fully equipped Smart Home with microphones and home automation sensors, in which participants performed Activities of Daily living (ADL). This corpus is made of a multimodal subset, a French home automation speech subset recorded in Distant Speech conditions, and two interaction subsets, the first one being recorded by 16 persons without disabilities and the second one by 6 seniors and 5 visually impaired people. This corpus was used in studies related to ADL recognition, context aware interaction and distant speech recognition applied to home automation controled through voice. | ['Fran{\\c{c}}ois Portet', 'Brigitte Meillon', 'Pedro Chahuara', 'Nicolas Bonnefond', 'Michel Vacher', 'Benjamin Lecouteux'] | 2014-05-01 | null | null | null | lrec-2014-5 | ['distant-speech-recognition'] | ['speech'] | [ 4.14400101e-01 4.48634118e-01 3.79677057e-01 -3.23649079e-01
-8.34176540e-01 -3.68866950e-01 8.32363546e-01 -2.50430077e-01
-6.88304484e-01 1.27304912e+00 1.11455476e+00 6.45036250e-02
-1.96655318e-01 -4.92688745e-01 3.91742885e-02 -7.08825290e-01
-1.80661947e-01 3.66520137e-01 3.37503329e-02 -1.81758463e-01
-5.08607030e-01 3.30360651e-01 -2.16871190e+00 4.46375430e-01
8.44030857e-01 8.66918206e-01 5.45253336e-01 7.42711604e-01
2.40608260e-01 7.00074792e-01 -6.79585278e-01 2.01582424e-02
-2.10080534e-01 -2.04540476e-01 -5.60513675e-01 1.42035127e-01
2.22057462e-01 -7.00030684e-01 -5.49881756e-01 6.79698169e-01
1.57316196e+00 3.42326641e-01 3.68316203e-01 -1.27333617e+00
-2.02457681e-01 3.26096892e-01 8.02796483e-01 -1.26475945e-01
1.23298514e+00 1.16354436e-01 5.63228548e-01 -6.61746919e-01
3.45670193e-01 1.24928379e+00 6.62948310e-01 8.90557230e-01
-1.10116172e+00 -3.39883342e-02 -2.55003814e-02 4.09113258e-01
-1.17524624e+00 -1.31876194e+00 3.28671306e-01 -3.91265541e-01
1.38408709e+00 4.60383743e-01 1.08071983e+00 1.93445134e+00
-5.03876090e-01 7.73583949e-01 8.84923041e-01 -7.41924465e-01
9.31899667e-01 2.61786431e-01 3.24324220e-01 1.47782464e-03
2.25082606e-01 9.87199321e-02 -6.88970923e-01 -4.81986254e-01
1.47132516e-01 1.78834289e-01 -9.29591715e-01 -2.08968058e-01
-1.77185094e+00 4.05819029e-01 -1.06580846e-01 8.01389277e-01
-7.16606200e-01 -6.54661775e-01 3.79823446e-01 5.78237057e-01
9.11563784e-02 -1.56321943e-01 -6.29438043e-01 -8.15313756e-01
-2.05161050e-01 -8.60689953e-02 1.32079732e+00 1.21083510e+00
4.86085638e-02 -5.01923680e-01 -2.81353116e-01 1.09585881e+00
4.65127110e-01 9.26400065e-01 7.84744740e-01 -9.53423381e-01
7.01446176e-01 6.98770344e-01 3.57609302e-01 -2.25861266e-01
-5.58388472e-01 6.47384226e-01 -8.39797616e-01 3.40313137e-01
5.04691899e-01 -6.38317764e-01 -5.70359528e-01 1.47338140e+00
7.73866028e-02 -6.48169160e-01 3.09422493e-01 4.56375688e-01
8.24083388e-01 1.18097432e-01 1.74623013e-01 -7.10663021e-01
1.42733097e+00 -7.11008132e-01 -1.43417013e+00 -3.52299064e-01
7.76801586e-01 -1.72697499e-01 1.39425862e+00 5.51119268e-01
-7.50743270e-01 -3.57893556e-01 -1.04905415e+00 7.11117079e-03
-7.03069925e-01 -9.58896503e-02 2.06478059e-01 1.03374743e+00
-1.44220507e+00 1.08689643e-01 -6.13831699e-01 -1.12304688e+00
3.50831658e-01 5.44951499e-01 -8.76945317e-01 5.21358028e-02
-1.27626705e+00 7.89112747e-01 7.44011626e-02 -2.55549699e-01
-2.40660384e-01 -1.82501256e-01 -1.06137681e+00 -4.14391309e-01
-1.33378312e-01 -6.71938181e-01 1.32950032e+00 -5.44529200e-01
-1.69032288e+00 9.70369399e-01 -2.12486058e-01 -2.99725652e-01
7.86902964e-01 -6.40604258e-01 -9.69902694e-01 -1.08304687e-01
2.37487689e-01 1.34677947e-01 4.88568276e-01 -8.05218577e-01
-7.64278054e-01 -1.22591913e+00 -2.88689643e-01 3.28953236e-01
-5.42677820e-01 -2.87891645e-02 2.61654109e-01 -4.09413457e-01
-3.15487832e-01 -6.25553012e-01 1.53676987e-01 1.07579075e-01
-3.14388245e-01 4.12625112e-02 8.46396267e-01 -1.10290790e+00
1.20003259e+00 -2.42345214e+00 -3.78407203e-02 -1.36021405e-01
2.71081775e-02 3.25447768e-01 2.49877885e-01 4.38134521e-01
1.26126304e-01 -2.36651376e-01 -3.44824344e-01 -6.76830709e-01
6.16567254e-01 3.51184279e-01 3.55687559e-01 2.56706655e-01
-6.72211945e-01 5.15524566e-01 -8.15664530e-01 -3.25956583e-01
7.95523465e-01 9.00997579e-01 -6.45857379e-02 4.66025472e-01
3.34566414e-01 6.34970605e-01 -1.33602813e-01 4.77311462e-01
-7.74379224e-02 4.97284025e-01 4.20251563e-02 7.70177320e-02
-3.58115762e-01 1.89357504e-01 -1.08358014e+00 1.77619731e+00
-6.07673883e-01 6.28929734e-01 3.58322054e-01 -3.92312825e-01
7.47265756e-01 1.00163913e+00 2.97126055e-01 -9.90147293e-01
1.67891562e-01 2.74316370e-01 -4.42294896e-01 -1.28348601e+00
-1.76761091e-01 2.93506056e-01 2.14418367e-01 1.71224251e-02
-2.31138438e-01 2.78546005e-01 -1.01055831e-01 -4.65056032e-01
1.76075733e+00 -3.69371861e-01 8.66007507e-01 -5.45804538e-02
6.86192155e-01 -6.89407527e-01 6.43081069e-02 6.53147221e-01
-7.71420777e-01 5.47939718e-01 -2.37800226e-01 -1.62483126e-01
-7.75938690e-01 -1.10641026e+00 6.48595840e-02 1.09534395e+00
-5.78070939e-01 -3.93612146e-01 -1.00670552e+00 -3.99008036e-01
-3.87944460e-01 9.47474539e-01 -1.90583453e-01 -2.07945593e-02
-2.27214962e-01 -1.43590391e-01 4.04468864e-01 5.00214696e-01
1.17255688e+00 -1.54450428e+00 -7.93239772e-01 1.77290007e-01
-7.24644542e-01 -1.32435298e+00 -3.06476623e-01 2.47739419e-01
-5.55339754e-01 -9.84073579e-01 -1.22202218e+00 -1.02940822e+00
3.28517258e-01 2.15199925e-02 7.99929261e-01 -7.45496154e-01
-8.68463591e-02 1.19463527e+00 -6.44027054e-01 -3.44604522e-01
-2.60642558e-01 -3.33648808e-02 4.12809342e-01 -4.83262427e-02
8.32830250e-01 -1.05847704e+00 -4.04454887e-01 5.48800409e-01
-1.49436459e-01 -4.57714081e-01 3.52141559e-01 6.99209034e-01
9.74070504e-02 9.48527381e-02 3.39865237e-01 1.46748394e-01
7.18822062e-01 -1.88751012e-01 4.95863616e-01 2.39241317e-01
-2.09392048e-02 -3.80806506e-01 3.36403221e-01 -5.09187520e-01
-1.06414449e+00 3.09440434e-01 -4.80313599e-01 4.80302483e-01
-1.34140682e+00 -2.53951013e-01 -1.54245961e+00 3.92388850e-01
7.24966109e-01 1.90976396e-01 5.44345379e-02 -6.51735663e-01
1.06253482e-01 1.78030729e+00 8.99829865e-01 -8.26803893e-02
8.69361609e-02 2.66568601e-01 -7.59319901e-01 -1.73706186e+00
1.43834114e-01 -4.08676356e-01 -9.87569571e-01 -4.22204018e-01
1.10163915e+00 -7.20687807e-01 -6.63967550e-01 9.71413493e-01
-1.09748852e+00 -8.07851434e-01 -7.15860963e-01 6.82920218e-01
-1.07917750e+00 1.58671156e-01 5.65818809e-02 -1.42711973e+00
-3.64031255e-01 -8.38677466e-01 1.35128295e+00 -8.00421461e-02
-9.14209127e-01 -8.02450776e-01 2.19206270e-02 4.65239346e-01
3.35973769e-01 4.66304600e-01 6.30151629e-01 -6.58543289e-01
3.75607520e-01 -5.28368771e-01 2.16997996e-01 4.42880452e-01
7.97865808e-01 -9.81794953e-01 -1.23051369e+00 -2.47107610e-01
3.43439102e-01 -1.69864282e-01 -9.77819562e-02 3.45293492e-01
2.09309578e-01 -5.34162164e-01 -3.38031530e-01 -9.80723649e-02
6.82930112e-01 7.07437515e-01 1.17927849e+00 4.42827851e-01
3.97262603e-01 8.14167500e-01 1.99685797e-01 6.26883686e-01
5.34134030e-01 1.01300144e+00 6.55162930e-02 3.61724406e-01
-2.38423601e-01 6.78723454e-02 7.76122689e-01 5.41342258e-01
-4.31143135e-01 -3.38920087e-01 -8.21128488e-01 5.06116211e-01
-2.03303695e+00 -1.07915545e+00 -1.54567480e-01 2.36108160e+00
3.93612087e-01 -3.90913993e-01 6.34695411e-01 1.31153512e+00
8.89340580e-01 -1.72144458e-01 -3.98914903e-01 2.93336868e-01
-4.05414760e-01 -1.14764258e-01 -3.77606153e-02 4.84254926e-01
-1.04978216e+00 1.06628627e-01 5.48439837e+00 4.00797389e-02
-3.74567270e-01 2.32160792e-01 3.65280956e-02 5.15062585e-02
4.57512200e-01 -8.85967076e-01 -3.67015928e-01 6.85635567e-01
1.17377889e+00 3.61387163e-01 6.60018623e-01 8.41868877e-01
6.87496066e-01 -2.49316424e-01 -1.59669125e+00 1.66644156e+00
5.84969223e-02 -1.92064196e-01 -4.18121010e-01 2.83037364e-01
-5.68184480e-02 -1.00786477e-01 -4.97565478e-01 1.09163925e-01
-1.62463889e-01 -7.52383828e-01 7.39340007e-01 1.03279150e+00
8.87851357e-01 -4.71924752e-01 7.37282276e-01 5.04051626e-01
-1.17992270e+00 -5.60877502e-01 4.72548932e-01 -3.02309215e-01
4.54949081e-01 1.08700112e-01 -3.22582066e-01 -9.60279927e-02
1.15532660e+00 6.09225273e-01 -3.61195117e-01 8.81023645e-01
-1.74659446e-01 2.24374816e-01 -3.94839853e-01 -3.91722381e-01
-4.84479576e-01 6.17139749e-02 7.48366475e-01 9.36379969e-01
7.50598073e-01 2.41946101e-01 -4.86836404e-01 1.42808557e-01
6.97986424e-01 -6.80589154e-02 -1.26640475e+00 2.38196537e-01
6.96046054e-01 5.90672791e-01 -7.07100630e-02 4.73467149e-02
-5.50292373e-01 1.59346402e+00 -4.62618411e-01 5.30104101e-01
-2.00703666e-01 -6.12370729e-01 7.92124569e-01 3.61840785e-01
-1.36593103e-01 -4.64080602e-01 -2.12459654e-01 -8.61998260e-01
4.75838095e-01 -1.04237032e+00 1.73384786e-01 -9.43607271e-01
-1.39064348e+00 5.54648221e-01 -3.87739837e-02 -1.09364116e+00
-4.43330348e-01 -7.24607527e-01 -2.10704103e-01 5.68438113e-01
-3.63229841e-01 -7.18553185e-01 -9.26295340e-01 1.15614331e+00
6.71790123e-01 -5.49846470e-01 1.66349375e+00 7.39514887e-01
-3.44536215e-01 4.02515978e-01 1.64304718e-01 -1.95052609e-01
5.53804815e-01 -1.35937512e+00 1.75139233e-01 1.54610515e-01
-6.95107639e-01 5.14240921e-01 6.05862021e-01 -3.98510486e-01
-1.16709089e+00 -7.51256883e-01 1.37052107e+00 -6.32327199e-01
4.05382156e-01 -5.98571301e-01 -3.46464068e-01 4.25347865e-01
1.63600966e-01 -3.04879367e-01 7.71276951e-01 -5.12427866e-01
4.68413144e-01 2.42340248e-02 -1.82521594e+00 7.33768284e-01
2.00799155e+00 -8.73564780e-01 -9.15339530e-01 2.71980524e-01
6.05379879e-01 3.16973716e-01 -1.06295097e+00 3.13437641e-01
8.82975280e-01 -9.43812966e-01 9.30823267e-01 1.01155564e-01
-4.58537549e-01 -4.16476242e-02 -1.16113758e+00 -1.25716019e+00
7.04237297e-02 -8.12634170e-01 -4.01469946e-01 1.55526519e+00
1.37521457e-02 -1.21725750e+00 2.13074893e-01 1.18005252e+00
-8.27824846e-02 5.68315275e-02 -1.35022962e+00 -8.20707321e-01
-6.74273312e-01 -6.57070100e-01 9.82455850e-01 4.81316566e-01
4.16186839e-01 4.27371413e-01 1.27732828e-02 -6.03593551e-02
5.58842719e-01 -1.40367723e+00 7.03763485e-01 -1.81845045e+00
2.36896053e-01 -2.76243120e-01 -9.59445179e-01 -7.49224365e-01
1.02599621e-01 -1.17898993e-01 2.58716047e-01 -1.84390187e+00
-4.08018827e-01 1.47942528e-01 4.97741371e-01 5.03271461e-01
4.61994678e-01 -2.53176808e-01 -2.52633005e-01 -1.34901460e-02
-2.64487416e-01 7.80473232e-01 5.76511860e-01 -4.45151448e-01
-4.69906032e-01 3.26739401e-01 -2.97467083e-01 7.61024833e-01
6.74816787e-01 4.42972928e-01 -6.49205148e-01 5.58408536e-02
-4.04576093e-01 -3.70837413e-02 7.85604894e-01 -1.62007475e+00
1.98087275e-01 4.91218984e-01 1.62600413e-01 -3.78928483e-01
5.90914369e-01 -1.38771522e+00 3.41935366e-01 2.14931652e-01
-3.35078128e-02 -4.62791622e-01 -2.27671355e-01 3.24473321e-01
1.04065001e-01 1.34989679e-01 5.08111596e-01 -6.51023686e-02
-7.92340457e-01 -1.32127836e-01 -9.47114050e-01 -2.04968035e-01
9.70385194e-01 -5.78572452e-01 -3.92090678e-01 -6.25025570e-01
-1.60913515e+00 -9.31583792e-02 4.76004571e-01 5.60999155e-01
5.10891438e-01 -1.65645754e+00 -9.28362757e-02 6.12944782e-01
2.35660717e-01 -4.44103256e-02 -2.17770096e-02 7.51292825e-01
-5.51361814e-02 5.96240163e-01 -2.84307778e-01 -1.40345648e-01
-1.51386762e+00 4.82931957e-02 5.28142869e-01 4.45613533e-01
-1.13721859e+00 8.97326693e-02 -3.52485210e-01 -6.91113174e-01
1.26521122e+00 -2.26792082e-01 -4.59902793e-01 1.35412514e-01
1.11697066e+00 1.13367593e+00 3.21324050e-01 -8.51077557e-01
-5.46346843e-01 3.03488851e-01 7.48006225e-01 -6.38614237e-01
1.20673382e+00 -7.48941362e-01 1.89494211e-02 9.56720412e-01
1.09944057e+00 -4.74921554e-01 -7.62812555e-01 -7.17000440e-02
1.73296198e-01 1.17909409e-01 -1.16474047e-01 -9.61206436e-01
-2.55642354e-01 4.04415756e-01 1.54819715e+00 6.13181293e-01
1.24064445e+00 3.14210691e-02 6.97976470e-01 1.03762984e+00
1.04215765e+00 -1.15071368e+00 -3.55910599e-01 4.52212512e-01
1.37793684e+00 -1.13240588e+00 -7.82961726e-01 -9.55710635e-02
-4.07738566e-01 8.02007735e-01 9.95047688e-02 6.43998682e-01
9.18877482e-01 5.01167297e-01 2.87146457e-02 1.34841606e-01
1.59393419e-02 -5.75353384e-01 -1.38857409e-01 1.63237333e+00
2.94430554e-01 2.34561697e-01 -5.70398308e-02 1.16369712e+00
-6.63800776e-01 3.23669523e-01 -1.81378901e-01 1.27990746e+00
-3.81763637e-01 -7.78716385e-01 -6.29588246e-01 4.34400052e-01
1.45178884e-01 3.19034606e-01 -7.31281459e-01 6.56074584e-01
4.18226808e-01 1.80939686e+00 -2.19754428e-01 -4.72703636e-01
1.10201645e+00 5.01867115e-01 3.87023598e-01 -2.68171370e-01
-1.58393532e-01 -6.06264293e-01 6.65545225e-01 -6.40257776e-01
-6.14939630e-01 -1.11325800e+00 -1.00726366e+00 -5.34335412e-02
2.85187334e-01 -1.55340418e-01 6.57927990e-01 1.13141716e+00
1.38974085e-01 6.35819495e-01 3.10330570e-01 -1.31514347e+00
1.10572502e-01 -1.38509047e+00 -1.02302539e+00 4.52408232e-02
7.75987864e-01 -4.51275170e-01 -6.03119910e-01 1.69415325e-01] | [7.121118068695068, 0.5146183967590332] |
3186814f-32c2-4aee-bdfb-a9b91b67fa4a | towards-unpaired-depth-enhancement-and-super | 2105.12038 | null | https://arxiv.org/abs/2105.12038v4 | https://arxiv.org/pdf/2105.12038v4.pdf | Unpaired Depth Super-Resolution in the Wild | Depth maps captured with commodity sensors are often of low quality and resolution; these maps need to be enhanced to be used in many applications. State-of-the-art data-driven methods of depth map super-resolution rely on registered pairs of low- and high-resolution depth maps of the same scenes. Acquisition of real-world paired data requires specialized setups. Another alternative, generating low-resolution maps from high-resolution maps by subsampling, adding noise and other artificial degradation methods, does not fully capture the characteristics of real-world low-resolution images. As a consequence, supervised learning methods trained on such artificial paired data may not perform well on real-world low-resolution inputs. We consider an approach to depth super-resolution based on learning from unpaired data. While many techniques for unpaired image-to-image translation have been proposed, most fail to deliver effective hole-filling or reconstruct accurate surfaces using depth maps. We propose an unpaired learning method for depth super-resolution, which is based on a learnable degradation model, enhancement component and surface normal estimates as features to produce more accurate depth maps. We propose a benchmark for unpaired depth SR and demonstrate that our method outperforms existing unpaired methods and performs on par with paired. | ['Evgeny Burnaev', 'Denis Zorin', 'Alexander Filippov', 'Alexey Artemov', 'Oleg Voynov', 'Nikita Drobyshev', 'Maxim Kan', 'Aleksandr Safin'] | 2021-05-25 | null | null | null | null | ['depth-map-super-resolution'] | ['computer-vision'] | [ 1.06045902e+00 1.11943446e-01 1.57237187e-01 -4.29670155e-01
-1.39372289e+00 -1.90959081e-01 4.67407376e-01 -1.46431047e-02
-1.08451739e-01 9.59836781e-01 1.59604445e-01 3.52562308e-01
-1.47237912e-01 -1.24663258e+00 -8.72960925e-01 -6.47448719e-01
1.73539951e-01 6.01592004e-01 6.98479891e-01 -2.92009920e-01
5.20587385e-01 5.17717838e-01 -2.24259520e+00 5.93602538e-01
9.07040954e-01 1.00189710e+00 6.96372032e-01 7.63443112e-01
-1.39707774e-01 4.76904958e-01 -1.77925572e-01 1.30150929e-01
4.62516099e-01 -5.23920774e-01 -6.08298421e-01 -6.23237155e-02
7.87170768e-01 -7.64106333e-01 -9.21646878e-02 8.73481154e-01
4.81599241e-01 -2.75721729e-01 5.72115481e-01 -4.83901203e-01
-4.56714928e-01 -1.19704910e-01 -5.98222673e-01 3.90896313e-02
8.17444801e-01 -2.43843228e-01 3.99013788e-01 -9.96762693e-01
8.35155904e-01 1.02395046e+00 4.76365477e-01 7.51421630e-01
-1.52664185e+00 -3.57013643e-01 -3.57184708e-01 1.06177643e-01
-9.75042760e-01 -5.56141973e-01 8.61319959e-01 -3.01440686e-01
6.52165234e-01 2.45240018e-01 4.59725797e-01 8.44377160e-01
1.37711301e-01 1.38705626e-01 1.71059310e+00 -5.61073244e-01
3.83350164e-01 1.15384832e-01 -5.63318133e-01 4.08590198e-01
2.31009811e-01 3.92057002e-01 -1.08081317e+00 1.00075208e-01
1.54804325e+00 -2.41697907e-01 -6.66516900e-01 -5.56824446e-01
-1.14030147e+00 3.80255252e-01 4.33103889e-01 3.04326236e-01
-5.19647360e-01 -2.43877992e-01 -2.36070439e-01 3.17367584e-01
7.88331151e-01 5.54991305e-01 -2.68242240e-01 -1.15823165e-01
-9.36198473e-01 1.62651837e-01 9.30050686e-02 7.50587761e-01
1.28879642e+00 -2.85649180e-01 3.13062072e-01 6.66274548e-01
-6.64342642e-02 4.85132188e-01 4.22443688e-01 -1.44712627e+00
3.46951306e-01 4.77009565e-01 3.90953600e-01 -6.46988392e-01
-1.40906230e-01 -1.46391258e-01 -8.41214836e-01 7.89665759e-01
4.30974305e-01 6.59384131e-01 -9.21309352e-01 1.31025565e+00
2.34824508e-01 2.19387516e-01 2.01677337e-01 1.00620341e+00
6.75598741e-01 5.64604700e-01 -7.72289932e-01 -4.82172608e-01
7.87074149e-01 -5.40355206e-01 -7.16355979e-01 -4.14541721e-01
5.04894927e-02 -5.65866590e-01 1.25289309e+00 7.46669292e-01
-1.57421923e+00 -4.81830388e-01 -1.22242582e+00 -4.57202882e-01
-8.23238418e-02 -3.72095555e-01 9.91306454e-02 3.25844556e-01
-1.17324877e+00 9.35762882e-01 -7.51040936e-01 -1.30449623e-01
3.64104301e-01 1.15124725e-01 -5.95470726e-01 -5.86065888e-01
-8.68990541e-01 7.00975239e-01 3.88451666e-02 -2.29313657e-01
-7.69269586e-01 -1.04810727e+00 -8.38523090e-01 -5.11220157e-01
9.29645076e-02 -9.00163829e-01 1.07268870e+00 -5.02510011e-01
-1.67941809e+00 1.14095485e+00 -3.56535465e-01 -1.33838683e-01
6.00032985e-01 -2.63062268e-01 -4.02899794e-02 6.17884040e-01
1.45785779e-01 4.87472415e-01 1.03498924e+00 -1.81986356e+00
-6.26232564e-01 -8.25598001e-01 -2.10021392e-01 2.65683591e-01
-6.18301928e-02 -3.41366380e-01 -6.74788207e-02 -1.41620398e-01
8.43501806e-01 -2.00619083e-02 -8.36935192e-02 4.58306998e-01
1.73093751e-01 7.56276488e-01 8.89276981e-01 -7.73005366e-01
7.01625228e-01 -1.78221560e+00 2.74416298e-01 -2.02902138e-01
3.37805122e-01 -8.33800435e-02 -4.03863899e-02 1.41256526e-01
2.16850311e-01 -7.55189359e-02 -5.62021494e-01 -7.14771628e-01
-7.28024960e-01 2.27860272e-01 -3.32766920e-02 5.99801242e-01
1.01036668e-01 5.22341669e-01 -9.86474276e-01 -4.29105908e-01
6.75824523e-01 1.00903654e+00 -3.99568915e-01 5.70340455e-01
-2.18367994e-01 1.12704539e+00 7.37711927e-03 6.69475615e-01
8.94942224e-01 -4.78759333e-02 -2.26649508e-01 -4.06825364e-01
-4.25582141e-01 2.38010377e-01 -1.08054590e+00 2.01179051e+00
-7.41913319e-01 4.43503737e-01 2.23431364e-01 -5.10322213e-01
1.26655614e+00 7.38498420e-02 4.95008916e-01 -1.38479030e+00
-2.25137204e-01 6.92423582e-01 -7.92460442e-01 -4.49290574e-01
6.37770534e-01 -5.98919928e-01 3.86219621e-01 2.55231440e-01
-2.53169090e-01 -8.31143916e-01 -4.38640982e-01 -1.75151095e-01
1.07020700e+00 3.40565562e-01 3.02897356e-02 1.43387884e-01
5.46965182e-01 -2.57729441e-01 4.73911792e-01 3.79763782e-01
2.22935840e-01 1.46722639e+00 5.40300533e-02 -2.98496962e-01
-1.74661660e+00 -1.31698990e+00 -4.52012628e-01 4.11229968e-01
3.72904718e-01 3.92464735e-02 -7.56049514e-01 6.73777461e-02
-4.00141150e-01 2.99776614e-01 -6.80598140e-01 7.50464350e-02
-5.90517938e-01 -5.31415820e-01 -6.70645908e-02 4.79474902e-01
7.44214237e-01 -7.87145495e-01 -9.29486811e-01 2.41344184e-01
-4.16509032e-01 -1.49182332e+00 1.17198780e-01 5.48313297e-02
-1.27781177e+00 -9.47270751e-01 -9.34224248e-01 -5.90741932e-01
6.38116956e-01 3.45913649e-01 1.23862529e+00 -3.10397893e-01
-2.50590503e-01 3.18530887e-01 -3.94969285e-01 5.95873483e-02
-4.82036144e-01 -3.84592891e-01 5.45278192e-02 1.73694178e-01
-2.46856824e-01 -1.02477956e+00 -7.19886184e-01 4.31140304e-01
-1.00357938e+00 4.37866688e-01 7.18810856e-01 8.09629619e-01
1.19805789e+00 1.75042257e-01 3.97716969e-01 -6.38919771e-01
-8.74797814e-03 -8.53469409e-03 -6.47330582e-01 -2.78127253e-01
-6.71145916e-01 2.62913615e-01 4.90389466e-01 -2.24106625e-01
-1.41511226e+00 5.28702475e-02 -5.86663149e-02 -4.30008233e-01
-2.24442229e-01 7.46349096e-02 -3.47064763e-01 -2.94837296e-01
1.18908489e+00 4.02948439e-01 1.17579028e-01 -7.21719384e-01
6.38536587e-02 5.94612777e-01 1.05580831e+00 -3.44235718e-01
8.94424558e-01 1.15830922e+00 2.21177012e-01 -1.03376174e+00
-8.17435503e-01 -5.16034484e-01 -8.25445712e-01 -1.14394829e-01
7.71628499e-01 -9.70083117e-01 -7.14732334e-02 7.01047957e-01
-1.06555176e+00 -5.50024569e-01 -4.88485098e-01 4.19025421e-01
-1.05986309e+00 3.20682645e-01 -4.74188089e-01 -8.55269790e-01
-7.42020831e-02 -7.54455388e-01 1.64317691e+00 1.62562862e-01
2.75608093e-01 -7.69161522e-01 1.54948115e-01 6.11549318e-01
6.80137217e-01 6.26174688e-01 7.77679622e-01 6.97744489e-01
-1.02587247e+00 1.53125241e-01 -3.65337014e-01 4.89631712e-01
4.82714951e-01 -3.80038023e-01 -1.26722562e+00 -2.70145744e-01
4.08193350e-01 -4.36111689e-01 7.24816024e-01 5.80011845e-01
1.00735080e+00 -1.73139032e-02 -1.16663091e-01 9.15387809e-01
2.00037956e+00 -1.98669657e-01 1.09830737e+00 6.01152658e-01
7.31766224e-01 9.29111063e-01 9.74688828e-01 3.61702353e-01
2.44823918e-01 8.73037815e-01 8.51278901e-01 -5.81461266e-02
-3.81697237e-01 -4.54244018e-01 1.65219128e-01 4.02188599e-01
-4.31153595e-01 2.82514334e-01 -7.40073919e-01 6.00712776e-01
-1.35910475e+00 -8.27072203e-01 -3.86323452e-01 2.63935876e+00
8.83353770e-01 1.83713421e-01 -1.05388202e-01 5.42211652e-01
5.51087141e-01 6.43461719e-02 -6.59665167e-01 5.42894229e-02
-5.54871440e-01 4.10331398e-01 4.83885050e-01 9.39785302e-01
-5.55756330e-01 6.06777549e-01 6.19724607e+00 3.15297693e-01
-1.03702462e+00 2.44532704e-01 4.74920034e-01 -3.71671133e-02
-7.76020527e-01 -2.30282575e-01 -6.11926496e-01 1.74877241e-01
8.65075350e-01 1.93568580e-02 4.20260012e-01 4.68837678e-01
3.22487354e-01 -5.40809929e-01 -1.21776557e+00 1.58324826e+00
1.16768301e-01 -1.33810461e+00 -7.39680827e-02 1.69854432e-01
1.13012516e+00 -1.84407204e-01 7.05963373e-02 -5.25194347e-01
-1.92107424e-01 -1.21461356e+00 5.54460227e-01 8.43193948e-01
1.28293502e+00 -5.62492847e-01 6.49327278e-01 4.84702379e-01
-1.11930871e+00 2.68280972e-03 -5.67919374e-01 -1.22779548e-01
4.02595103e-01 9.80006814e-01 -4.43286330e-01 6.67349756e-01
9.71004784e-01 8.16835701e-01 -5.39493918e-01 8.46727133e-01
-1.59138098e-01 -1.25014096e-01 -2.45331258e-01 7.19334185e-01
-4.94742930e-01 -7.40291923e-02 3.55657727e-01 5.10725617e-01
5.70767641e-01 1.51208743e-01 -4.57335770e-01 1.12825561e+00
2.52279818e-01 -1.03915744e-01 -8.49588275e-01 4.29129064e-01
3.63540173e-01 8.52776468e-01 -3.15068841e-01 -6.59698993e-02
-3.53386372e-01 1.12408638e+00 1.31018043e-01 1.45486951e-01
-3.44318420e-01 1.46792352e-01 5.01339793e-01 9.25835073e-01
-1.14915907e-01 -3.02363783e-01 -6.88297987e-01 -1.06168509e+00
4.59636956e-01 -5.71672738e-01 -7.92701766e-02 -1.18895996e+00
-1.17628074e+00 7.10285544e-01 -7.31480569e-02 -1.50498676e+00
-3.61615390e-01 -3.93504739e-01 -7.22465962e-02 1.09056222e+00
-1.97264564e+00 -1.01134527e+00 -1.06482172e+00 5.56461453e-01
4.76006597e-01 3.69456470e-01 7.92145669e-01 2.85604298e-01
2.08238542e-01 3.85085456e-02 7.12123960e-02 -5.49351692e-01
7.93292642e-01 -1.21095097e+00 2.71406800e-01 7.57376075e-01
-3.61063480e-01 -2.13471666e-01 8.42392445e-01 -6.00598812e-01
-1.27720225e+00 -9.32046652e-01 5.79610169e-01 -6.62213385e-01
-9.02446061e-02 -3.18759948e-01 -1.29381180e+00 3.24957430e-01
-3.75868946e-01 4.13200676e-01 5.76905683e-02 -5.02728224e-01
-2.06857458e-01 -2.13739216e-01 -1.59801841e+00 8.38014185e-02
1.29396904e+00 -8.39055777e-01 -5.03330112e-01 -1.30287081e-01
7.13298202e-01 -6.47965431e-01 -9.34417009e-01 7.31341183e-01
5.54321885e-01 -1.69621992e+00 1.12492716e+00 3.21864873e-01
9.14676845e-01 -4.65764880e-01 -4.25229400e-01 -1.19138062e+00
-1.08596392e-01 -2.32981294e-01 -1.47224024e-01 9.16317284e-01
1.69991300e-01 -6.39775991e-01 1.03805423e+00 3.63127798e-01
-1.72076240e-01 -6.12477660e-01 -9.48937237e-01 -8.70134592e-01
3.13432552e-02 -2.46319532e-01 5.63005030e-01 8.75261068e-01
-5.42944431e-01 9.20899957e-02 -2.38437667e-01 4.07733589e-01
1.31787229e+00 1.64990127e-01 7.43230820e-01 -1.40009594e+00
-2.05899835e-01 -1.19712055e-01 -5.25678337e-01 -9.85488296e-01
-2.86699235e-01 -2.41391450e-01 1.57040477e-01 -1.85849583e+00
9.86967050e-03 -5.54775715e-01 2.23230809e-01 -1.19357202e-02
1.88441589e-01 7.56191969e-01 -5.57730615e-01 3.52938533e-01
-8.39979760e-03 6.40860856e-01 1.44053662e+00 1.39051542e-01
-4.07760561e-01 -3.09882790e-01 -4.09540623e-01 6.02619052e-01
6.24873877e-01 -1.77056521e-01 -4.58682984e-01 -4.42064762e-01
4.04164851e-01 3.21709722e-01 5.30104816e-01 -1.31644797e+00
1.14030421e-01 -1.33612469e-01 4.61259186e-01 -5.99393308e-01
8.53871226e-01 -7.55447388e-01 3.17976028e-01 3.03481966e-02
-1.32444128e-02 -4.14799929e-01 -1.39606684e-01 4.53352273e-01
-4.10899460e-01 3.21586579e-02 1.13296497e+00 -3.95651728e-01
-6.94849193e-01 3.57605070e-01 2.31947333e-01 -1.60642132e-01
7.98437357e-01 -8.26586008e-01 -4.23721492e-01 -3.89958411e-01
-5.99763811e-01 -4.84335452e-01 1.30859280e+00 2.35069826e-01
1.20809829e+00 -1.37000608e+00 -8.80528510e-01 3.91316742e-01
3.30587924e-01 7.89304793e-01 4.44640875e-01 4.76212919e-01
-7.53336489e-01 2.76114177e-02 -5.96982956e-01 -8.66592646e-01
-1.09816694e+00 3.97111297e-01 5.69837749e-01 -5.52101135e-02
-8.91958594e-01 6.86196446e-01 1.41692743e-01 -2.96782315e-01
-1.05701111e-01 -3.00404757e-01 -2.95540076e-02 -2.12396175e-01
9.25716698e-01 4.25008088e-01 3.79722953e-01 -5.57203472e-01
-3.16964537e-02 1.29116607e+00 3.22047591e-01 -4.02720332e-01
1.64638865e+00 -5.21347046e-01 -1.04673795e-01 7.15562522e-01
1.10170460e+00 -9.91681367e-02 -1.52754390e+00 -4.18937117e-01
-4.61860478e-01 -1.20092869e+00 3.53210211e-01 -4.95128095e-01
-9.75087762e-01 9.65770245e-01 9.39569831e-01 -5.73864244e-02
1.56145859e+00 1.22909546e-01 6.27339423e-01 -2.90654540e-01
1.17385340e+00 -9.39736307e-01 4.86464709e-01 -1.08202687e-02
1.05547023e+00 -1.49225497e+00 3.42042185e-02 -7.62523949e-01
1.41099868e-02 1.03741038e+00 6.49683475e-01 5.55005632e-02
2.98207760e-01 4.69388843e-01 4.18734141e-02 -1.51586488e-01
-5.61080635e-01 -5.28448038e-02 -2.29861140e-02 1.10309327e+00
1.29676193e-01 -3.10849071e-01 -5.70234284e-02 -1.29328128e-02
-2.34587595e-01 1.31777838e-01 8.69637251e-01 1.00123858e+00
-6.67971790e-01 -1.17292237e+00 -8.49464834e-01 3.40360850e-01
1.72757264e-03 1.52490944e-01 -2.72650644e-02 2.94554681e-01
1.00795388e-01 6.79904640e-01 2.27192059e-01 -4.21341777e-01
4.56829727e-01 -3.39073986e-01 1.04540741e+00 -7.78621554e-01
3.10822606e-01 -2.31868312e-01 -1.01994231e-01 -1.02143323e+00
-7.92627573e-01 -6.81315780e-01 -1.10050833e+00 -2.11262852e-01
2.92152874e-02 -3.74995917e-01 7.29794443e-01 6.12557054e-01
2.52113789e-01 4.23870027e-01 6.75530910e-01 -1.48703289e+00
1.60870507e-01 -8.52807760e-01 -8.33926976e-01 3.78946096e-01
8.25947285e-01 -7.35126615e-01 -7.58544147e-01 7.14815706e-02] | [9.526644706726074, -2.4783198833465576] |
ea77395a-2af0-43ac-931f-6f8fea8cbd77 | socialgym-2-0-simulator-for-multi-agent | 2303.05584 | null | https://arxiv.org/abs/2303.05584v1 | https://arxiv.org/pdf/2303.05584v1.pdf | SOCIALGYM 2.0: Simulator for Multi-Agent Social Robot Navigation in Shared Human Spaces | We present SocialGym 2, a multi-agent navigation simulator for social robot research. Our simulator models multiple autonomous agents, replicating real-world dynamics in complex environments, including doorways, hallways, intersections, and roundabouts. Unlike traditional simulators that concentrate on single robots with basic kinematic constraints in open spaces, SocialGym 2 employs multi-agent reinforcement learning (MARL) to develop optimal navigation policies for multiple robots with diverse, dynamic constraints in complex environments. Built on the PettingZoo MARL library and Stable Baselines3 API, SocialGym 2 offers an accessible python interface that integrates with a navigation stack through ROS messaging. SocialGym 2 can be easily installed and is packaged in a docker container, and it provides the capability to swap and evaluate different MARL algorithms, as well as customize observation and reward functions. We also provide scripts to allow users to create their own environments and have conducted benchmarks using various social navigation algorithms, reporting a broad range of social navigation metrics. Projected hosted at: https://amrl.cs.utexas.edu/social_gym/index.html | ['Joydeep Biswas', 'Jarrett Holtz', 'Rohan Chandra', 'Zayne Sprague'] | 2023-03-09 | null | null | null | null | ['social-navigation', 'robot-navigation'] | ['robots', 'robots'] | [-8.11155021e-01 3.96996699e-02 3.71391289e-02 -2.43788719e-01
-4.46022332e-01 -6.95608795e-01 5.31395853e-01 3.40720385e-01
-8.23684275e-01 1.02324522e+00 -1.83076069e-01 -6.15127444e-01
-3.27648848e-01 -9.61315274e-01 -6.01589203e-01 -4.42586750e-01
-7.42414474e-01 9.01698411e-01 5.76999664e-01 -9.62480426e-01
2.54353821e-01 1.42489389e-01 -1.75148654e+00 -7.44470537e-01
6.72473192e-01 1.90339852e-02 6.50314152e-01 1.03762472e+00
5.02494335e-01 6.81637883e-01 -5.27678311e-01 2.93147475e-01
2.97938854e-01 -9.49757844e-02 -7.22764015e-01 -5.91319442e-01
-2.60780245e-01 -3.52096111e-01 -2.37070039e-01 6.37218535e-01
6.88076019e-01 6.40630782e-01 2.04759374e-01 -2.09739065e+00
-1.15329757e-01 4.69504893e-01 -2.69980609e-01 9.33365971e-02
8.84486914e-01 5.77726543e-01 7.35971570e-01 -3.22279096e-01
6.47540569e-01 1.54041839e+00 6.01409733e-01 2.88806319e-01
-1.00907815e+00 -6.10291719e-01 2.01887414e-01 1.31497070e-01
-1.02527380e+00 -2.03286529e-01 -1.64427280e-01 -1.33526400e-01
1.13768959e+00 5.62810600e-02 6.57489061e-01 1.20953012e+00
6.30420387e-01 4.66740578e-01 1.03802741e+00 2.35166494e-02
6.50385499e-01 -3.49722862e-01 9.18551832e-02 8.28055978e-01
3.63272488e-01 1.29635796e-01 -2.07277492e-01 -4.27757353e-01
5.64385831e-01 -1.38755113e-01 2.14384899e-01 -6.58990562e-01
-1.37656832e+00 8.99616778e-01 4.43073571e-01 -3.75917226e-01
-5.53586841e-01 6.61606133e-01 2.17761502e-01 3.87383133e-01
-3.99874806e-01 5.71002364e-01 -4.11403358e-01 -6.82130098e-01
3.70120496e-01 9.59027052e-01 1.25760043e+00 1.10092723e+00
8.79255712e-01 -1.02477111e-01 5.88817477e-01 7.10836649e-01
5.72391868e-01 6.54035568e-01 4.08201814e-01 -1.98759508e+00
1.33456022e-01 3.03463697e-01 6.21753633e-01 -7.96664476e-01
-1.32196653e+00 -1.47451069e-02 -1.40081182e-01 1.03254032e+00
4.15629774e-01 -7.25224316e-01 -2.65627116e-01 1.63294876e+00
7.67210186e-01 -8.44528303e-02 5.22909284e-01 8.57965350e-01
6.43609166e-01 4.63903219e-01 2.45636590e-02 4.19598937e-01
1.09013867e+00 -1.30210972e+00 -3.16813588e-01 -4.47789937e-01
1.01637518e+00 -4.97121364e-01 1.08422434e+00 1.16427995e-01
-9.09744442e-01 2.65841633e-02 -1.17098248e+00 2.10553199e-01
-5.92362165e-01 -6.36070132e-01 7.27021813e-01 2.38453090e-01
-1.50894058e+00 6.02850139e-01 -1.28016818e+00 -9.19958651e-01
-2.22254679e-01 4.62277770e-01 -2.49528587e-01 -5.26183238e-03
-9.79404151e-01 1.24826050e+00 9.54085067e-02 -3.70353281e-01
-1.27030253e+00 -3.07303783e-03 -1.08984149e+00 -4.65101182e-01
5.67327023e-01 -7.86795735e-01 1.72972560e+00 -3.61654833e-02
-1.72414625e+00 1.94868892e-01 3.90571445e-01 -3.42649192e-01
5.32550812e-01 -7.29302317e-02 -4.14543487e-02 -2.05120564e-01
7.93749750e-01 7.63204038e-01 -8.42514187e-02 -1.27880979e+00
-7.65149832e-01 -1.00917974e-03 7.08273828e-01 7.71705866e-01
4.96737868e-01 -7.19003305e-02 -9.51702446e-02 4.86340299e-02
-1.46055624e-01 -1.53075433e+00 -9.90810513e-01 -3.14317495e-01
-1.11382663e-01 -6.86185136e-02 5.50838590e-01 9.90745351e-02
3.81410152e-01 -1.83832335e+00 2.01402366e-01 1.46498993e-01
-1.35794580e-01 -4.04672980e-01 -4.69195455e-01 9.78051424e-01
5.53499401e-01 -1.48503050e-01 1.11593895e-01 -5.28410733e-01
5.33250332e-01 5.46193242e-01 4.35497105e-01 4.46594507e-01
-5.27654231e-01 5.40524960e-01 -1.55788100e+00 -3.01583976e-01
4.57093269e-01 7.98684508e-02 -7.32930720e-01 -1.94874346e-01
-1.32640183e-01 6.72787666e-01 -4.72273260e-01 5.57562411e-01
5.29393911e-01 -2.12570384e-01 4.55174446e-01 1.08368301e+00
-5.67566693e-01 4.19327974e-01 -1.51775348e+00 1.77377141e+00
-6.82298362e-01 3.73324662e-01 4.75742072e-01 -2.70100176e-01
4.89133716e-01 -1.65487066e-01 4.97702271e-01 -6.76925540e-01
2.48234957e-01 2.89567918e-01 -6.45456538e-02 -2.06053481e-01
1.12571859e+00 6.67756200e-01 -5.10014117e-01 7.67748058e-01
-1.91122875e-01 -3.64861578e-01 5.27435184e-01 2.31052935e-01
1.59846735e+00 4.68624473e-01 3.10469031e-01 -3.73193860e-01
-5.20524904e-02 4.99653548e-01 5.06376982e-01 1.12515795e+00
-5.61569929e-01 1.60752647e-02 3.97036791e-01 -1.49521381e-01
-8.15832973e-01 -1.40537274e+00 1.56169668e-01 1.71475339e+00
9.28463638e-01 -5.24943948e-01 -3.48090500e-01 -1.40767708e-01
3.55803430e-01 9.69722033e-01 -4.01278496e-01 1.91430077e-01
-4.20030206e-01 -8.56941521e-01 4.81601208e-01 1.94034614e-02
4.41427231e-01 -1.21676886e+00 -1.02124250e+00 2.83427745e-01
-1.33908000e-02 -8.18938792e-01 -1.64101869e-02 3.27311903e-01
-1.81640312e-01 -1.32069016e+00 -6.64650127e-02 -9.53031719e-01
2.95252383e-01 6.48330092e-01 1.03632653e+00 2.66019851e-01
-1.78245977e-01 8.06213260e-01 -5.55607259e-01 -3.81095201e-01
-5.57145596e-01 3.44449639e-01 2.99016535e-01 -1.04076695e+00
-7.77281895e-02 -7.76117265e-01 -6.25540555e-01 7.15169489e-01
-2.62399733e-01 4.77196127e-02 2.10904796e-02 3.96146178e-01
8.45529661e-02 -2.47573987e-01 7.22045660e-01 -1.60535291e-01
9.26889420e-01 -1.06053662e+00 -1.05195951e+00 -4.12345737e-01
-4.95165586e-01 -3.24667126e-01 2.65767753e-01 1.55940447e-02
-4.82140452e-01 -9.01485011e-02 -1.03913255e-01 3.82937700e-01
-1.93560570e-01 4.64915425e-01 2.62491614e-01 -1.23016901e-01
7.45418370e-01 -7.23692477e-02 4.28112417e-01 4.13886942e-02
4.00830448e-01 5.05527079e-01 2.34287545e-01 -7.17071652e-01
6.03125989e-01 3.98206443e-01 -2.86102016e-02 -6.45495772e-01
9.20955315e-02 -2.62685061e-01 -8.20670724e-02 -3.01140040e-01
6.98123634e-01 -9.73267674e-01 -1.53196275e+00 6.05307281e-01
-4.97228324e-01 -1.43868005e+00 2.21568998e-02 5.79376400e-01
-8.95910800e-01 7.35220015e-02 -7.02061355e-01 -6.49886012e-01
2.22039387e-01 -1.48992479e+00 7.29105294e-01 6.40913546e-01
-3.83538425e-01 -9.53233302e-01 4.75532144e-01 -1.56925526e-02
4.27942634e-01 1.93752661e-01 2.21733287e-01 -4.68856424e-01
-7.10884988e-01 4.54353429e-02 2.05874965e-01 -5.02283275e-01
-1.34724095e-01 2.52121910e-02 -3.25045943e-01 -5.45408547e-01
-8.19120288e-01 -4.95730042e-01 -4.11763154e-02 3.48035693e-01
1.50116310e-01 1.69043630e-01 -6.51438951e-01 3.38640869e-01
1.09298754e+00 4.85353589e-01 4.02865171e-01 1.51457322e+00
2.32968569e-01 3.53876740e-01 1.08556426e+00 6.88888550e-01
1.54933369e+00 6.38223886e-01 1.16554511e+00 2.64978647e-01
5.00207722e-01 -3.01969107e-02 6.37828827e-01 4.61218834e-01
1.27456248e-01 -2.05952168e-01 -1.11098731e+00 3.72930169e-01
-2.43851542e+00 -7.77262568e-01 -3.40501726e-01 2.05321097e+00
2.64582753e-01 7.91504085e-02 3.70186657e-01 -5.77314019e-01
4.86670673e-01 9.23131332e-02 -8.02562296e-01 -4.09987092e-01
2.71329228e-02 -4.69530612e-01 9.78451848e-01 1.16438603e+00
-1.02622509e+00 1.04527092e+00 6.25096083e+00 2.29909658e-01
-5.23060322e-01 2.21712962e-01 -1.17458023e-01 -2.30436698e-01
8.16008728e-03 3.08899164e-01 -7.77189434e-01 2.68480688e-01
1.10103214e+00 -4.82330769e-01 9.60455835e-01 1.07152939e+00
5.85629582e-01 -7.68083036e-01 -6.48455024e-01 4.66353625e-01
-5.25463343e-01 -1.04209423e+00 -8.14868629e-01 3.22906166e-01
6.26937389e-01 1.05574274e+00 -2.09175110e-01 8.23218405e-01
1.63424587e+00 -7.96912611e-01 8.20069134e-01 2.33445898e-01
-4.67661209e-02 -9.04376626e-01 5.87721169e-01 5.29558480e-01
-1.07400107e+00 -3.55002344e-01 -8.38026106e-02 -7.20584869e-01
5.02582669e-01 -3.15896392e-01 -1.07275176e+00 4.22852933e-01
1.06346953e+00 5.71167529e-01 -5.45603931e-01 1.33404922e+00
-1.90974027e-01 -1.38425231e-01 -5.81664085e-01 -6.36604667e-01
4.59021091e-01 -6.39523625e-01 7.45749950e-01 5.87574959e-01
3.24203938e-01 -2.40323663e-01 6.70216441e-01 2.49101192e-01
4.03374970e-01 -2.40164809e-02 -7.94418275e-01 4.45578724e-01
8.88144016e-01 1.44533551e+00 -8.78144503e-01 -5.02990186e-02
-1.88195318e-01 6.81472719e-01 2.23488361e-01 3.90934229e-01
-1.15437531e+00 -6.34695411e-01 1.32829332e+00 -3.04685771e-01
-2.71273553e-01 -9.32121575e-01 -8.09639809e-04 -6.76213503e-01
-4.00462598e-01 -7.86650538e-01 1.93506032e-01 -1.15044749e+00
-6.73962653e-01 4.21338320e-01 -4.82013598e-02 -1.18134713e+00
-4.66161102e-01 -3.90173644e-01 -5.14549375e-01 4.20080185e-01
-1.35605669e+00 -7.54111886e-01 -6.29620671e-01 3.50409120e-01
4.01935339e-01 -1.62314609e-01 1.02280569e+00 1.26990423e-01
-7.91730583e-01 2.23411509e-04 3.63845795e-01 -2.68826962e-01
7.25725412e-01 -1.58570206e+00 8.46820474e-01 2.94266224e-01
-8.29778552e-01 5.82532883e-01 1.15778005e+00 -7.13470995e-01
-1.51587129e+00 -9.23092544e-01 5.51224612e-02 -5.44630885e-01
1.10270035e+00 -1.93002492e-01 -3.59674215e-01 1.25008404e+00
5.09640276e-01 -3.22751552e-01 5.56098223e-01 2.15123773e-01
2.72153974e-01 2.75418907e-01 -9.41691399e-01 1.36217999e+00
1.16452205e+00 1.81877926e-01 -1.11305699e-01 5.97622216e-01
7.71255672e-01 -8.79792690e-01 -7.16837287e-01 -4.25222218e-02
5.19945741e-01 -1.04303277e+00 7.49881089e-01 -7.02987537e-02
-6.60958365e-02 -6.93763435e-01 -4.09194499e-01 -1.86498868e+00
-1.91699311e-01 -8.26142669e-01 4.09378976e-01 6.72077894e-01
5.75671732e-01 -1.36667752e+00 5.14576256e-01 3.04164588e-01
-4.71137583e-01 -2.63864636e-01 -9.08087969e-01 -8.43384564e-01
5.90883009e-03 -4.00544345e-01 6.97860420e-01 6.18083119e-01
3.55824590e-01 3.05213362e-01 -7.38926828e-02 6.87690854e-01
7.10920334e-01 -3.11923862e-01 1.44396734e+00 -8.89896274e-01
-3.19618672e-01 -4.87422496e-01 -2.76796609e-01 -7.72133768e-01
2.42541701e-01 -7.58329570e-01 3.54205996e-01 -1.89419436e+00
-3.61308008e-01 -1.07072222e+00 1.40709579e-01 6.26108825e-01
3.53257596e-01 1.49219725e-02 1.85430467e-01 3.84171382e-02
-1.31694818e+00 7.29749620e-01 1.05085123e+00 1.92845523e-01
-5.93308806e-01 -1.31724412e-02 -5.73237658e-01 8.81086290e-01
1.15105307e+00 -3.82644385e-01 -3.35203886e-01 -4.03399110e-01
4.21268016e-01 1.22627176e-01 5.23966789e-01 -1.48387253e+00
5.16162813e-01 -5.42266548e-01 -2.39449814e-01 -3.99569422e-01
6.36649072e-01 -5.23782909e-01 1.65674016e-01 7.77672887e-01
4.84378263e-02 7.56560624e-01 3.04056376e-01 5.35627246e-01
4.43817794e-01 -2.62257367e-01 4.87182617e-01 -3.42776179e-01
-9.40548062e-01 -1.00881726e-01 -1.22415161e+00 3.48527469e-02
1.57637811e+00 1.71025507e-02 -9.15249288e-01 -7.01545537e-01
-7.54695117e-01 1.32680058e+00 1.20156562e+00 6.67958975e-01
2.16147527e-01 -1.02323878e+00 -2.98175365e-01 -2.14204550e-01
2.11928532e-01 -5.91662712e-02 2.13724673e-01 7.01624811e-01
-1.17529035e+00 -1.45278245e-01 -4.49134916e-01 -4.67303306e-01
-1.02618051e+00 1.88547552e-01 6.07555985e-01 1.75053537e-01
-6.00773752e-01 3.28579873e-01 -3.61221373e-01 -1.37674046e+00
1.33835822e-01 -2.17143565e-01 -1.47369310e-01 -2.78330654e-01
3.46437782e-01 7.08658397e-01 -3.48648608e-01 -4.53955650e-01
-5.89457393e-01 1.27994224e-01 5.55230260e-01 -4.88599569e-01
1.42928791e+00 -6.30164623e-01 3.33296228e-03 5.97548187e-01
5.09575069e-01 -8.32937583e-02 -1.41694736e+00 4.22753483e-01
6.72376752e-02 -1.08199485e-01 -3.94445211e-01 -6.50088727e-01
-1.87117875e-01 -5.10752611e-02 5.76948106e-01 2.68341303e-01
2.86001563e-01 -1.34259030e-01 4.81679022e-01 8.38483512e-01
1.24147010e+00 -1.26279461e+00 1.17467865e-01 1.05768442e+00
7.86029875e-01 -1.28436589e+00 -3.46185341e-02 -3.25109996e-02
-7.32871771e-01 6.74451828e-01 9.07336831e-01 -4.20183510e-01
5.62696338e-01 3.25920582e-01 5.39916515e-01 -1.96431667e-01
-1.00479758e+00 -4.37269062e-01 -1.11926317e+00 1.05844164e+00
-2.36562118e-01 2.69805640e-01 -5.37760742e-02 9.48076323e-02
-6.46828592e-01 -2.49269679e-01 1.56003690e+00 1.57303143e+00
-7.03554332e-01 -1.09835601e+00 -6.57279730e-01 1.05755530e-01
1.21170439e-01 3.86341184e-01 1.96031198e-01 1.03671384e+00
-2.75436133e-01 1.22084987e+00 1.04307406e-01 -3.67307395e-01
3.16521347e-01 -5.53533077e-01 5.86085133e-02 -5.75493217e-01
-3.78071785e-01 -3.68565172e-01 4.73595083e-01 -8.42774332e-01
-2.59770919e-02 -1.00461125e+00 -1.96645319e+00 -5.64186454e-01
2.08263397e-01 3.47543836e-01 9.68808889e-01 5.14061928e-01
6.21045172e-01 6.20558202e-01 4.28788275e-01 -1.52330577e+00
-2.61930466e-01 -8.05331111e-01 -5.14945805e-01 -2.75589108e-01
3.37736905e-01 -1.12438726e+00 -2.87554890e-01 -8.20109785e-01] | [4.627251625061035, 1.121848464012146] |
afb7aa7d-76bd-4359-99b1-b85f8b76bdb7 | a-convnet-for-the-2020s | 2201.03545 | null | https://arxiv.org/abs/2201.03545v2 | https://arxiv.org/pdf/2201.03545v2.pdf | A ConvNet for the 2020s | The "Roaring 20s" of visual recognition began with the introduction of Vision Transformers (ViTs), which quickly superseded ConvNets as the state-of-the-art image classification model. A vanilla ViT, on the other hand, faces difficulties when applied to general computer vision tasks such as object detection and semantic segmentation. It is the hierarchical Transformers (e.g., Swin Transformers) that reintroduced several ConvNet priors, making Transformers practically viable as a generic vision backbone and demonstrating remarkable performance on a wide variety of vision tasks. However, the effectiveness of such hybrid approaches is still largely credited to the intrinsic superiority of Transformers, rather than the inherent inductive biases of convolutions. In this work, we reexamine the design spaces and test the limits of what a pure ConvNet can achieve. We gradually "modernize" a standard ResNet toward the design of a vision Transformer, and discover several key components that contribute to the performance difference along the way. The outcome of this exploration is a family of pure ConvNet models dubbed ConvNeXt. Constructed entirely from standard ConvNet modules, ConvNeXts compete favorably with Transformers in terms of accuracy and scalability, achieving 87.8% ImageNet top-1 accuracy and outperforming Swin Transformers on COCO detection and ADE20K segmentation, while maintaining the simplicity and efficiency of standard ConvNets. | ['Saining Xie', 'Trevor Darrell', 'Christoph Feichtenhofer', 'Chao-yuan Wu', 'Hanzi Mao', 'Zhuang Liu'] | 2022-01-10 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Liu_A_ConvNet_for_the_2020s_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Liu_A_ConvNet_for_the_2020s_CVPR_2022_paper.pdf | cvpr-2022-1 | ['real-time-object-detection'] | ['computer-vision'] | [ 8.56948346e-02 7.45704994e-02 1.36790618e-01 -1.59373343e-01
-3.04416902e-02 -5.31184852e-01 9.66342270e-01 -3.54594409e-01
-6.42467320e-01 2.20600933e-01 -2.88151130e-02 -3.97873104e-01
-2.00207774e-02 -7.08760440e-01 -5.65901816e-01 -6.76460385e-01
3.05548310e-01 1.77717939e-01 5.56784570e-01 -3.60669076e-01
9.84434783e-02 3.94555688e-01 -1.66208601e+00 4.95439097e-02
6.63168311e-01 1.24044955e+00 4.04890925e-01 4.18114930e-01
-1.90447167e-01 9.75402296e-01 -4.71396923e-01 -7.07204759e-01
2.06898540e-01 -1.49433697e-02 -7.73847997e-01 -5.08473665e-02
6.47912860e-01 -1.82031885e-01 -6.59337103e-01 1.08289695e+00
1.93438306e-01 -3.07832211e-01 3.84646088e-01 -1.29333031e+00
-8.39444637e-01 6.33870304e-01 -3.92142445e-01 2.79347777e-01
-4.61851619e-02 3.38913769e-01 1.23067808e+00 -9.15085912e-01
3.70083809e-01 1.09977007e+00 1.10947001e+00 4.97676671e-01
-1.24128640e+00 -4.78467673e-01 3.92990500e-01 2.34431863e-01
-1.32340515e+00 -4.57422465e-01 3.83717924e-01 -4.81701940e-01
1.21851861e+00 6.17271438e-02 9.81920600e-01 1.17124665e+00
3.14050883e-01 9.39254463e-01 1.18017566e+00 -1.74622744e-01
1.45090654e-01 6.70812353e-02 2.06860498e-01 7.34680593e-01
2.58045822e-01 9.15280059e-02 -3.43028694e-01 3.11179131e-01
6.68232918e-01 9.02432390e-03 -4.13072288e-01 -3.16491783e-01
-1.06312060e+00 6.17372751e-01 8.83854687e-01 4.27009583e-01
-3.48075926e-01 2.92168975e-01 3.74434292e-01 1.78372920e-01
1.81661591e-01 4.81240392e-01 -3.01224887e-01 5.21059185e-02
-9.71680760e-01 2.67200097e-02 5.82465768e-01 7.85871565e-01
6.97185040e-01 5.20178020e-01 4.16318327e-03 7.16783643e-01
3.00211191e-01 3.53570968e-01 4.60631251e-01 -8.14588964e-01
-1.47293717e-01 8.16698313e-01 -3.39327455e-01 -5.98997951e-01
-3.23537886e-01 -9.22551274e-01 -8.92110705e-01 4.19204652e-01
5.47184587e-01 8.97939056e-02 -1.43182588e+00 1.61629009e+00
-1.29746541e-01 5.34745678e-02 2.83317417e-02 6.91070437e-01
9.18315053e-01 4.09750044e-01 3.24378699e-01 4.04291362e-01
1.53789878e+00 -9.54576612e-01 -1.13583393e-01 -3.98646444e-01
8.51046667e-02 -6.94740772e-01 9.53170240e-01 4.33783442e-01
-9.20981526e-01 -7.67612338e-01 -1.18805778e+00 2.53846962e-02
-4.75684136e-01 -9.25336778e-02 9.39046323e-01 7.90789843e-01
-1.53779352e+00 4.59467202e-01 -8.23815763e-01 -4.69526678e-01
7.68352389e-01 5.78348339e-01 -1.50323629e-01 5.92132993e-02
-8.46533239e-01 1.07096744e+00 3.15012664e-01 2.57525951e-01
-1.24984825e+00 -9.16707456e-01 -6.80546105e-01 1.79164559e-01
3.76231641e-01 -9.01638508e-01 1.32615280e+00 -1.17574251e+00
-1.28229368e+00 8.95394862e-01 9.17414129e-02 -8.07318747e-01
6.09776974e-01 -9.66116861e-02 -2.47345731e-01 7.73436874e-02
2.81116925e-02 1.12727547e+00 1.01696527e+00 -1.07594264e+00
-8.54892254e-01 -1.36120141e-01 3.20359290e-01 6.52599856e-02
-3.96164775e-01 -1.24648996e-01 -5.93096793e-01 -2.98485607e-01
-1.78826544e-02 -8.10018182e-01 -3.87630522e-01 -1.10353820e-01
-3.73368770e-01 -4.64775145e-01 8.88622403e-01 -1.68027371e-01
4.73362982e-01 -2.23854303e+00 -1.93940178e-02 1.78644806e-02
8.31760705e-01 5.65610230e-01 -2.36979514e-01 1.19817287e-01
-1.91898849e-02 -3.46924104e-02 -7.00026676e-02 -2.45085582e-01
-4.07310873e-02 2.26530537e-01 -3.95842880e-01 4.63977188e-01
1.76856101e-01 1.34031069e+00 -8.09795856e-01 -2.33952403e-01
4.97354269e-01 7.03408182e-01 -2.66591907e-01 -1.66474730e-01
-1.10810272e-01 4.74905409e-02 -3.41324568e-01 5.74713826e-01
5.02287924e-01 -3.98337722e-01 5.11998795e-02 -4.90465522e-01
-4.26244974e-01 2.21145868e-01 -6.95174873e-01 1.47420287e+00
-1.35883495e-01 9.87155318e-01 1.69637129e-01 -1.02033556e+00
7.93290794e-01 1.58373266e-01 4.70140278e-01 -9.93864238e-01
3.35662782e-01 1.94728866e-01 2.56007642e-01 -1.07897669e-01
6.68571413e-01 -1.25901729e-01 3.04883808e-01 -1.81092709e-01
4.30702746e-01 -2.54024655e-01 3.42563055e-02 2.04848588e-01
1.07635093e+00 1.20226391e-01 6.83607832e-02 -6.10898733e-01
2.60840714e-01 1.22313969e-01 5.16435325e-01 8.66099596e-01
-3.19467217e-01 6.54660225e-01 4.19817805e-01 -6.09089613e-01
-1.05779028e+00 -1.32862782e+00 -2.54474789e-01 8.70335102e-01
1.86395183e-01 -5.37850201e-01 -8.18825722e-01 -5.13088226e-01
-1.33131489e-01 3.12637508e-01 -7.27263331e-01 8.57570022e-02
-2.55670369e-01 -9.54838336e-01 9.99497890e-01 6.80938840e-01
1.04624426e+00 -9.21600223e-01 -8.83473337e-01 1.38170078e-01
1.53359890e-01 -1.32319021e+00 3.46399867e-03 3.81033480e-01
-6.12366259e-01 -1.14491832e+00 -7.96668291e-01 -8.39215934e-01
4.28495795e-01 4.42853600e-01 1.19496083e+00 1.10212609e-01
-4.08070296e-01 5.97585142e-01 -1.01051509e-01 -4.88600940e-01
-1.94389492e-01 1.98559329e-01 -6.17019013e-02 -6.64438009e-02
1.74311861e-01 -6.79804087e-01 -8.22007060e-01 1.84924528e-01
-7.91621089e-01 7.71927387e-02 6.83524013e-01 8.14149201e-01
2.74410278e-01 -4.93208244e-02 3.81494820e-01 -8.07724774e-01
3.29601586e-01 -2.02839032e-01 -6.47677422e-01 2.54444301e-01
-7.56439209e-01 -2.16048062e-01 6.40890539e-01 -2.60023266e-01
-9.59172487e-01 -2.61189658e-02 -4.33967829e-01 -5.04589438e-01
8.79665837e-03 2.64343411e-01 1.27329990e-01 -3.02649647e-01
5.78333199e-01 4.61078554e-01 -9.59802493e-02 -3.41728747e-01
4.16597456e-01 3.35562676e-01 8.02147031e-01 -3.89934093e-01
6.96456552e-01 5.22486448e-01 -2.43279904e-01 -1.19254816e+00
-7.31326342e-01 -2.46839479e-01 -2.19531119e-01 -1.28908589e-01
1.10771537e+00 -1.05948567e+00 -9.17417943e-01 9.51287210e-01
-9.92580354e-01 -3.45032007e-01 -4.68159676e-01 2.01277256e-01
-1.87498063e-01 2.15291306e-01 -7.33552694e-01 -4.99786913e-01
-2.92040348e-01 -1.36677670e+00 6.53002143e-01 5.53278208e-01
9.04565901e-02 -1.01395357e+00 -2.79294223e-01 3.59097868e-01
7.89610267e-01 -1.08879007e-01 8.08238804e-01 -4.37771320e-01
-6.44582748e-01 7.18248487e-02 -6.59023881e-01 7.11353362e-01
-1.72654495e-01 9.29858163e-02 -1.49254501e+00 -2.65652597e-01
3.92470658e-02 -3.59234422e-01 1.19433212e+00 4.01093960e-01
1.01013994e+00 1.66654199e-01 -2.60638863e-01 9.47072566e-01
1.60306287e+00 1.79557130e-01 8.85175884e-01 3.63869309e-01
8.17872941e-01 2.86809713e-01 -2.65479565e-01 -5.88188171e-02
4.58139718e-01 3.96238625e-01 7.23864079e-01 -3.76937270e-01
-4.75845069e-01 -3.08215797e-01 1.69082165e-01 7.08848655e-01
-2.76037872e-01 -1.56951398e-01 -1.02957022e+00 6.15788877e-01
-1.55613673e+00 -7.03062475e-01 -1.24978222e-01 1.79135990e+00
3.24726373e-01 5.10252893e-01 -1.24577500e-01 3.55203710e-02
5.04252195e-01 2.62665987e-01 -6.45713329e-01 -1.72600403e-01
-3.35607380e-01 5.42955279e-01 7.20906138e-01 1.12446547e-01
-1.09118676e+00 1.19099736e+00 7.03261709e+00 8.12130332e-01
-1.30979609e+00 -1.97701305e-02 6.40237391e-01 4.12575006e-01
-1.46302894e-01 9.13301855e-02 -8.74716759e-01 1.75628126e-01
6.41218424e-01 1.61730781e-01 3.76243621e-01 9.22572494e-01
-3.14980537e-01 -5.66647425e-02 -9.81356919e-01 1.01946139e+00
8.96479748e-03 -1.57627416e+00 1.60600796e-01 5.39216138e-02
5.86339474e-01 7.54880130e-01 3.09307724e-01 4.19457436e-01
5.14287770e-01 -1.40354264e+00 9.98817921e-01 4.33756858e-02
8.05303037e-01 -1.68745548e-01 6.16789520e-01 -6.03153743e-02
-1.24779499e+00 -1.49392903e-01 -3.74050319e-01 -1.20899200e-01
-1.97690830e-01 3.05356175e-01 -6.27603292e-01 2.92266399e-01
1.01862478e+00 8.56197059e-01 -7.82380462e-01 1.20750701e+00
-2.45227516e-01 7.27160513e-01 -2.84823865e-01 1.69675410e-01
7.92591512e-01 1.13382284e-02 4.00334299e-01 1.26553845e+00
-2.37588212e-02 -1.35910168e-01 6.39011860e-02 9.67871070e-01
-2.88116872e-01 -4.02750701e-01 -5.17356873e-01 1.05138689e-01
3.76335621e-01 1.43876374e+00 -1.16112924e+00 -4.10414219e-01
-5.55557549e-01 6.91398919e-01 2.16306880e-01 4.27316815e-01
-8.47474337e-01 -2.29887828e-01 8.44838798e-01 -1.18843213e-01
6.37179017e-01 -1.79006666e-01 -2.99445450e-01 -1.19683039e+00
-1.56662121e-01 -8.87430906e-01 2.05360092e-02 -6.50335133e-01
-1.18215728e+00 9.13961649e-01 -3.08667243e-01 -8.14231932e-01
2.68557638e-01 -8.99430037e-01 -7.61573613e-01 5.30708790e-01
-1.63288677e+00 -1.37191260e+00 -5.35211742e-01 8.12648714e-01
5.12132347e-01 -1.77707151e-01 6.35482907e-01 2.57842034e-01
-5.38406074e-01 5.85812509e-01 -2.68335547e-03 3.67940456e-01
2.15285048e-01 -1.28816497e+00 7.01922655e-01 8.80091369e-01
3.94673437e-01 5.72943926e-01 5.77670693e-01 -1.70656040e-01
-1.38544762e+00 -8.51892173e-01 4.06047970e-01 -4.05423760e-01
8.37739170e-01 -4.65143293e-01 -6.54107153e-01 7.67205000e-01
4.72066998e-01 -5.06907068e-02 1.05797350e-02 1.56969681e-01
-7.40514040e-01 -2.98326045e-01 -9.03737843e-01 7.87679732e-01
1.16180241e+00 -5.95112920e-01 -7.38035500e-01 -1.44363344e-01
5.88171899e-01 -1.30588204e-01 -6.62566662e-01 4.41142291e-01
6.54606938e-01 -1.23848438e+00 1.18210447e+00 -3.55815321e-01
3.04040253e-01 -2.21825644e-01 -1.87184140e-01 -1.22889304e+00
-4.64416057e-01 -6.22241974e-01 4.69414026e-01 1.19372344e+00
3.74637336e-01 -8.40973139e-01 8.81560922e-01 3.39060485e-01
-4.25237954e-01 -7.52652645e-01 -8.75258029e-01 -6.24699891e-01
2.27330588e-02 -6.61983490e-01 4.33940321e-01 7.43533909e-01
-4.41448033e-01 5.50904930e-01 -3.59640904e-02 -2.14288861e-01
7.24935353e-01 -2.06088424e-01 4.50641572e-01 -1.24905658e+00
-3.02495211e-01 -8.02291930e-01 -8.32882762e-01 -1.21061850e+00
-1.59705818e-01 -9.37060535e-01 4.58486155e-02 -1.47561562e+00
2.51237482e-01 -5.81674039e-01 -5.25448203e-01 6.97001100e-01
7.31145293e-02 7.19026685e-01 6.41078889e-01 2.09518701e-01
-5.69235206e-01 3.82493079e-01 1.06729305e+00 -3.63219678e-01
-1.63267888e-02 -1.52238697e-01 -1.03282893e+00 9.86009717e-01
6.90011144e-01 -2.76731029e-02 -5.86548567e-01 -7.52869904e-01
1.54500216e-01 -3.54623169e-01 7.59827912e-01 -1.39328372e+00
3.28981787e-01 2.65791982e-01 4.89629805e-01 -2.57907927e-01
4.44578469e-01 -5.66999435e-01 1.16071798e-01 5.53979456e-01
-6.95902184e-02 5.10100722e-02 3.53179663e-01 3.31809342e-01
-1.58625066e-01 3.74010578e-02 9.89255428e-01 -2.76922673e-01
-1.22891450e+00 2.41340876e-01 -4.74147707e-01 2.12598905e-01
8.47896576e-01 -6.14733458e-01 -6.71714723e-01 4.90311086e-02
-5.14919579e-01 1.12143494e-01 4.66515899e-01 5.30766666e-01
7.05688477e-01 -6.28624260e-01 -6.20086312e-01 1.71006739e-01
-6.47824630e-02 -2.79548485e-02 1.05135635e-01 8.68009031e-01
-4.84100223e-01 4.62177247e-01 -5.19023836e-01 -9.15170431e-01
-9.93762970e-01 2.37982243e-01 6.83736682e-01 -1.29750848e-01
-1.05064464e+00 1.05807388e+00 7.49209762e-01 -9.51890275e-02
2.79470086e-01 -3.50927919e-01 -1.06684975e-01 2.76036821e-02
3.97271127e-01 1.82896256e-01 1.64637402e-01 -3.22502613e-01
-3.90166253e-01 4.46352690e-01 -2.68594086e-01 9.25502405e-02
1.50043368e+00 1.36877745e-01 -1.13313571e-01 8.55754539e-02
7.30253994e-01 -3.42583060e-01 -1.50835252e+00 -5.90299852e-02
-6.45992607e-02 -8.26573521e-02 1.67859510e-01 -8.94280612e-01
-1.42544687e+00 8.28454137e-01 5.57474434e-01 2.34322548e-01
1.07621658e+00 -7.20228709e-04 6.85745299e-01 3.20166469e-01
3.79451036e-01 -7.80626893e-01 -9.01604220e-02 6.58786654e-01
4.02847350e-01 -9.54093874e-01 -2.54534006e-01 -1.95411548e-01
-5.63338101e-01 9.28071260e-01 6.29253328e-01 -4.07509983e-01
5.87534964e-01 4.56342876e-01 4.18672673e-02 -3.77189517e-01
-6.72356367e-01 -4.34527606e-01 4.55138162e-02 7.55150974e-01
3.69487524e-01 1.09444723e-01 9.08444747e-02 2.24389225e-01
-3.76223207e-01 7.36592039e-02 4.02656078e-01 5.75991452e-01
-4.70435560e-01 -8.10852766e-01 -1.58356816e-01 4.94178623e-01
-4.73555118e-01 -3.30207348e-01 -2.94587433e-01 1.05273378e+00
5.10020614e-01 9.52769041e-01 3.81406695e-02 -6.13356054e-01
3.11468720e-01 -1.12336814e-01 5.25822222e-01 -2.27074802e-01
-1.08437932e+00 -1.69376269e-01 -1.16789378e-01 -4.52366650e-01
-3.12484950e-01 -3.57866973e-01 -9.96159017e-01 -5.43893099e-01
-9.88011584e-02 8.43669921e-02 6.89016342e-01 1.06435561e+00
5.41148297e-02 6.93151355e-01 2.22204164e-01 -8.21890891e-01
-5.87108016e-01 -6.75773740e-01 -4.12563145e-01 8.93686190e-02
3.43146265e-01 -5.30574620e-01 2.71872487e-02 -1.29856899e-01] | [9.412857055664062, 1.5904340744018555] |
a833d4fb-34c1-47ab-8697-b4ee289b9110 | domain-adaptation-for-question-answering-via | 2209.04998 | null | https://arxiv.org/abs/2209.04998v2 | https://arxiv.org/pdf/2209.04998v2.pdf | Domain Adaptation for Question Answering via Question Classification | Question answering (QA) has demonstrated impressive progress in answering questions from customized domains. Nevertheless, domain adaptation remains one of the most elusive challenges for QA systems, especially when QA systems are trained in a source domain but deployed in a different target domain. In this work, we investigate the potential benefits of question classification for QA domain adaptation. We propose a novel framework: Question Classification for Question Answering (QC4QA). Specifically, a question classifier is adopted to assign question classes to both the source and target data. Then, we perform joint training in a self-supervised fashion via pseudo-labeling. For optimization, inter-domain discrepancy between the source and target domain is reduced via maximum mean discrepancy (MMD) distance. We additionally minimize intra-class discrepancy among QA samples of the same question class for fine-grained adaptation performance. To the best of our knowledge, this is the first work in QA domain adaptation to leverage question classification with self-supervised adaptation. We demonstrate the effectiveness of the proposed QC4QA with consistent improvements against the state-of-the-art baselines on multiple datasets. | ['Dong Wang', 'Lanyu Shang', 'Ziyi Kou', 'Huimin Zeng', 'Zhenrui Yue'] | 2022-09-12 | null | https://aclanthology.org/2022.coling-1.153 | https://aclanthology.org/2022.coling-1.153.pdf | coling-2022-10 | ['classification'] | ['methodology'] | [ 4.19610262e-01 8.58971551e-02 -6.99137850e-03 -7.92947352e-01
-1.57150364e+00 -9.86783862e-01 3.64624739e-01 1.95455506e-01
-2.87009597e-01 8.30390811e-01 1.68678939e-01 -3.84884477e-01
1.14459349e-02 -6.49766207e-01 -6.41463220e-01 -3.61432672e-01
7.19475448e-01 8.23450446e-01 4.97297198e-01 -5.50396740e-01
3.37351412e-02 9.43439305e-02 -1.26534748e+00 5.65091014e-01
1.54935098e+00 1.14377785e+00 8.25666562e-02 6.33831739e-01
-3.76818359e-01 8.24545503e-01 -7.97784090e-01 -7.53744602e-01
8.00572187e-02 -8.09715390e-01 -1.30713964e+00 8.13893825e-02
9.11600709e-01 -1.39298648e-01 -4.15424369e-02 8.68773520e-01
5.16710222e-01 2.80282050e-01 6.80824578e-01 -9.68117774e-01
-9.90055084e-01 -9.05056149e-02 -1.48935631e-01 2.17299968e-01
4.34710681e-01 8.52641985e-02 1.16200864e+00 -7.80799329e-01
4.81790304e-01 1.03868771e+00 3.52744192e-01 8.07294786e-01
-1.21164322e+00 -3.83643478e-01 4.44688313e-02 5.11938214e-01
-9.07574177e-01 -5.45633078e-01 7.00805128e-01 -2.16213480e-01
7.11247027e-01 1.52107656e-01 -2.07754195e-01 8.58043671e-01
-6.02340810e-02 6.83620691e-01 1.04380000e+00 -6.98106349e-01
4.53979343e-01 2.98101038e-01 2.91700423e-01 4.25816208e-01
-2.16825962e-01 -4.29700911e-01 -3.48332644e-01 -3.13410878e-01
1.18251234e-01 -4.26103324e-01 -3.03944707e-01 -4.55468059e-01
-8.42601538e-01 9.60758448e-01 2.20027730e-01 1.91801846e-01
-2.94937611e-01 -3.18054616e-01 2.93627888e-01 6.37534082e-01
3.47767830e-01 6.99080348e-01 -1.00704002e+00 -9.60065871e-02
-5.96837044e-01 2.86963493e-01 9.30018008e-01 8.33685994e-01
7.95327842e-01 -4.18703169e-01 -4.85904843e-01 1.24405169e+00
1.77593842e-01 5.31526148e-01 6.67548776e-01 -1.10771120e+00
8.19115996e-01 9.66319323e-01 2.10630372e-01 -4.22420174e-01
-1.42395973e-01 -4.32201982e-01 -4.79542345e-01 -9.58543867e-02
7.79411733e-01 -1.60816580e-01 -9.38343287e-01 1.81168437e+00
6.56464636e-01 -1.67543560e-01 4.65612769e-01 6.86105013e-01
8.32126617e-01 5.13796926e-01 3.39164138e-01 1.29304126e-01
1.64178002e+00 -1.34346819e+00 -7.12850392e-01 -4.99134153e-01
6.55954063e-01 -8.93817604e-01 1.48623157e+00 -4.52193581e-02
-9.61869538e-01 -6.93915784e-01 -7.69435227e-01 -3.77020806e-01
-2.45547026e-01 2.17697874e-01 2.76013613e-02 7.82037795e-01
-6.44685268e-01 -7.15946034e-02 -4.25562918e-01 -4.79475349e-01
3.58370960e-01 1.30724788e-01 -3.05293322e-01 -4.36440587e-01
-1.40578246e+00 1.02802241e+00 1.51999533e-01 -4.85956132e-01
-8.15009296e-01 -7.53212988e-01 -8.00260484e-01 1.27547095e-02
3.94931644e-01 -8.72893453e-01 1.73836827e+00 -1.14429271e+00
-1.69290173e+00 1.01691031e+00 -6.23006344e-01 -5.07978678e-01
3.49673666e-02 -2.92880416e-01 -6.37119830e-01 2.99461097e-01
3.89437646e-01 4.28987503e-01 8.38991463e-01 -1.24250162e+00
-7.40977764e-01 -5.30798554e-01 3.97705227e-01 4.26628977e-01
-3.40314776e-01 -2.06628487e-01 -2.86785632e-01 -4.00109231e-01
1.06205001e-01 -6.89613163e-01 -9.26989538e-04 -2.03857958e-01
1.22112624e-01 -5.58411837e-01 6.13563657e-01 -1.03263187e+00
1.28353596e+00 -1.96855044e+00 -4.33764094e-03 -1.78267956e-01
-4.40220647e-02 4.80274975e-01 -3.78055215e-01 3.40747029e-01
2.00317621e-01 -2.84407675e-01 -7.40532219e-01 -2.72912413e-01
2.96330154e-02 2.84999788e-01 -4.88970906e-01 4.32953164e-02
5.40734828e-01 9.90363836e-01 -8.16214919e-01 -4.72124904e-01
-2.52879918e-01 -5.32427765e-02 -5.28332829e-01 8.27108800e-01
-6.17660999e-01 6.80832803e-01 -5.17812371e-01 6.30737245e-01
7.25416899e-01 -4.58678871e-01 4.55540083e-02 -2.77945459e-01
5.16784608e-01 6.04175746e-01 -7.76203036e-01 2.00260925e+00
-6.31360233e-01 3.38742435e-01 1.76012411e-03 -1.13522553e+00
1.16495430e+00 4.53369290e-01 5.66220060e-02 -1.05234313e+00
-1.33494571e-01 4.11177635e-01 -2.97087915e-02 -7.14256823e-01
5.62400639e-01 -3.60914499e-01 -1.95366979e-01 4.46137488e-01
4.49008435e-01 -3.46775770e-01 1.26204938e-01 -2.83320597e-03
1.07596684e+00 1.26105770e-01 3.64885449e-01 1.63418204e-02
9.85717654e-01 3.49465847e-01 2.94646472e-01 6.79847360e-01
-5.99674404e-01 8.16736042e-01 2.28960052e-01 -5.44440076e-02
-8.96769106e-01 -1.30284369e+00 -8.73670354e-02 1.59037113e+00
-2.02410407e-02 1.42846957e-01 -9.20221984e-01 -1.31928718e+00
-9.18164104e-03 8.47184420e-01 -5.55374324e-01 -3.28268200e-01
-7.13995278e-01 -3.56644183e-01 7.24519968e-01 3.54270935e-01
6.28484070e-01 -8.98981810e-01 -2.13068083e-01 3.06121051e-01
-5.62121034e-01 -1.14356589e+00 -6.49917185e-01 2.04061586e-02
-9.72053111e-01 -1.15111923e+00 -7.75201499e-01 -1.10100806e+00
4.99670327e-01 1.81118816e-01 1.61238110e+00 -2.39213884e-01
3.21129501e-01 6.37713611e-01 -5.26210070e-01 -8.73294026e-02
-5.53529620e-01 5.67930996e-01 -3.40882361e-01 1.62404597e-01
6.89706862e-01 -2.04315111e-01 -7.22650588e-01 5.63668787e-01
-1.01966214e+00 -5.30397177e-01 3.69217664e-01 9.46405053e-01
5.45281529e-01 -4.42098498e-01 1.22248662e+00 -1.05446696e+00
9.49253917e-01 -7.87551761e-01 -3.28195214e-01 8.80511463e-01
-4.99772429e-01 1.56420693e-01 7.97995687e-01 -2.43918881e-01
-1.55190372e+00 -1.56687140e-01 -4.14071947e-01 2.45400637e-01
-5.51092267e-01 3.71311098e-01 -4.64811444e-01 -1.81968454e-02
1.11255598e+00 1.74000278e-01 -2.00984567e-01 -5.71000040e-01
6.59546196e-01 9.70275998e-01 6.14270329e-01 -7.48121321e-01
6.40807033e-01 3.91693175e-01 -4.71327096e-01 -3.99290800e-01
-1.31088984e+00 -7.66718447e-01 -7.28309751e-01 1.97579190e-01
8.96142721e-01 -8.32062185e-01 -3.24536920e-01 2.73465902e-01
-1.13054073e+00 -2.92733014e-01 -4.75623816e-01 1.18101522e-01
-4.17782903e-01 4.69486386e-01 -2.14708194e-01 -6.02830350e-01
-5.08668900e-01 -9.38630700e-01 9.52680647e-01 4.91245419e-01
-2.01533705e-01 -1.19009113e+00 5.92698455e-01 1.15690994e+00
4.68027383e-01 -1.44733518e-01 9.43611205e-01 -1.08241451e+00
-1.93915844e-01 -6.00740425e-02 -2.07631707e-01 6.62405312e-01
5.09151459e-01 -7.09688067e-01 -1.01323581e+00 -2.36305654e-01
2.85256028e-01 -8.02617073e-01 6.42790258e-01 -5.83107881e-02
8.33601356e-01 -9.56868604e-02 1.22286230e-01 3.41230989e-01
1.14518702e+00 2.06208766e-01 5.33248484e-01 6.72267914e-01
4.53949183e-01 8.24344218e-01 9.19873714e-01 -9.18859709e-03
9.56160545e-01 6.87145770e-01 1.51894957e-01 1.45226061e-01
-3.84301782e-01 -1.89266801e-01 2.47276545e-01 6.63818359e-01
6.58169329e-01 -3.90648961e-01 -1.08403456e+00 9.58081901e-01
-1.63216388e+00 -4.98274446e-01 -8.16101953e-02 2.10042143e+00
1.18655109e+00 -2.36869708e-01 -5.51090054e-02 -2.71741152e-01
5.71894228e-01 -3.58170867e-02 -9.47598577e-01 -2.66975969e-01
-1.77240729e-01 6.20978236e-01 3.02455891e-02 5.29618919e-01
-1.10288644e+00 8.36628616e-01 5.54561043e+00 6.84312880e-01
-5.57403147e-01 3.94523859e-01 5.68848073e-01 3.44042391e-01
-3.13496351e-01 1.41482547e-01 -7.81989992e-01 3.27466547e-01
1.16950595e+00 9.80873033e-02 1.02944866e-01 8.04388881e-01
-2.98194259e-01 -1.64319780e-02 -9.65542257e-01 6.04609311e-01
3.85427922e-01 -8.63147140e-01 1.57302558e-01 -4.27918792e-01
1.04774189e+00 -1.45490468e-01 8.87398198e-02 5.67774415e-01
2.35242337e-01 -8.00237954e-01 2.56035507e-01 2.97059029e-01
7.23174930e-01 -5.04404068e-01 9.06008065e-01 3.29600096e-01
-9.06984627e-01 -2.74812933e-02 -3.91667753e-01 2.81009138e-01
3.49170119e-01 3.18729550e-01 -8.99623096e-01 7.13347852e-01
6.30695701e-01 2.01886058e-01 -8.01944315e-01 8.83824944e-01
-3.38233292e-01 9.68159080e-01 6.73428327e-02 3.10545236e-01
8.43503997e-02 -1.75615966e-01 2.53417701e-01 9.31185186e-01
1.89236984e-01 1.56610012e-01 -1.17159165e-01 5.21332860e-01
-3.19327116e-01 2.50741929e-01 -5.06226607e-02 1.74789801e-01
5.40785015e-01 9.39862192e-01 1.10016257e-01 -4.55946863e-01
-6.19058788e-01 1.29603076e+00 6.65851116e-01 4.22784567e-01
-5.64238429e-01 -4.98257279e-01 7.10255325e-01 -1.45625830e-01
3.59996080e-01 1.19708432e-02 -4.76482570e-01 -1.47881639e+00
2.99507737e-01 -1.31563163e+00 1.00472569e+00 -5.89663267e-01
-1.76088428e+00 6.68037832e-01 -2.74139345e-01 -1.24121130e+00
-2.59411931e-01 -4.99208003e-01 -3.80166441e-01 1.06356287e+00
-2.01839972e+00 -1.15257287e+00 -4.26955968e-01 8.41833830e-01
7.07533598e-01 -4.03709561e-01 1.03268862e+00 4.17240173e-01
-2.06930101e-01 1.02562904e+00 4.98459965e-01 9.72946584e-02
1.37708342e+00 -1.41277397e+00 5.07161021e-01 6.25229359e-01
2.64735203e-02 5.24751008e-01 4.43881124e-01 -4.41193104e-01
-1.02106249e+00 -1.04362106e+00 1.05230308e+00 -8.10582578e-01
5.07377505e-01 -1.36027098e-01 -1.56357002e+00 5.85098863e-01
4.05409783e-01 -1.57896295e-01 1.07184875e+00 6.47000372e-02
-6.61703527e-01 -2.95681387e-01 -1.37553859e+00 3.88585508e-01
4.69708502e-01 -7.91195989e-01 -1.07634151e+00 3.60666186e-01
9.61407721e-01 -4.45166588e-01 -1.09479618e+00 3.62312943e-01
2.55305350e-01 -7.64066815e-01 9.48622346e-01 -8.64248216e-01
2.70713627e-01 -4.57040071e-01 -2.94219375e-01 -1.29205871e+00
-1.08976051e-01 -1.67161584e-01 -2.62025326e-01 1.48719764e+00
5.42934000e-01 -6.80434585e-01 8.80147398e-01 5.71974039e-01
-1.50769055e-01 -5.21816969e-01 -1.06445158e+00 -6.57405674e-01
5.75518668e-01 -8.81048664e-02 7.30613828e-01 9.23181295e-01
-3.17961484e-01 7.98469305e-01 -9.13924798e-02 3.73873383e-01
4.84870732e-01 2.63799548e-01 8.64226222e-01 -9.48900998e-01
-4.66529548e-01 3.23475301e-02 8.76364410e-02 -1.54937172e+00
1.57838464e-01 -7.19375074e-01 1.98383287e-01 -1.63517308e+00
-6.73711374e-02 -4.07752484e-01 -2.31912822e-01 1.34227633e-01
-7.75809169e-01 2.68130153e-02 -8.01246390e-02 2.80959636e-01
-8.57624531e-01 8.31478655e-01 1.21250939e+00 -1.12305634e-01
-1.75612003e-01 1.15932651e-01 -8.00257802e-01 2.54390985e-01
1.18378222e+00 -5.54030180e-01 -4.94757622e-01 -8.12922478e-01
-7.03267977e-02 3.65121476e-02 2.06016749e-02 -8.97283435e-01
2.44688451e-01 -7.92724341e-02 3.32224136e-03 -3.81556213e-01
2.53837883e-01 -8.11630309e-01 -5.00070393e-01 1.21717408e-01
-4.96558547e-01 -6.26410171e-02 2.51721084e-01 5.81772208e-01
-5.82813203e-01 -5.86623609e-01 8.80784571e-01 -2.08536461e-02
-6.23473287e-01 2.03526150e-02 -5.41120060e-02 7.40230322e-01
8.19142461e-01 9.00574997e-02 -6.13174558e-01 -5.38752139e-01
-4.39082026e-01 4.84704256e-01 3.23218793e-01 3.70779395e-01
3.84114504e-01 -1.25896490e+00 -9.77147937e-01 -8.00660402e-02
6.57324135e-01 4.94967252e-02 4.96784806e-01 4.24420774e-01
-2.53240258e-01 6.10745013e-01 -5.59794419e-02 -5.29039681e-01
-1.09692633e+00 2.67439246e-01 4.70449001e-01 -7.46424675e-01
2.13985875e-01 9.32816565e-01 3.33619595e-01 -1.25955021e+00
-2.80883852e-02 1.01993009e-02 -3.82425070e-01 -5.07872701e-02
5.11050463e-01 3.70355159e-01 3.22840065e-01 -5.26325405e-01
-2.87554055e-01 5.51149130e-01 -3.78056139e-01 -2.13317290e-01
9.28247690e-01 -4.83976275e-01 -5.94245531e-02 2.52478868e-01
1.18589747e+00 -7.88345858e-02 -1.27008796e+00 -7.06728816e-01
2.50504047e-01 -3.10612112e-01 -4.77714777e-01 -1.32592964e+00
-5.48944116e-01 9.49168503e-01 7.73198903e-01 1.24173321e-01
1.33758283e+00 2.66030729e-01 1.10804641e+00 4.86312807e-01
4.00816537e-02 -1.06599867e+00 4.10184026e-01 8.09063137e-01
8.93390954e-01 -1.62998724e+00 -3.49817157e-01 -1.54520422e-01
-9.20851350e-01 8.94505799e-01 9.82789457e-01 1.37272300e-02
2.50233620e-01 -6.68092787e-01 6.04325175e-01 -6.19848408e-02
-6.21377647e-01 -3.76975238e-01 7.14514911e-01 7.96697974e-01
5.83827972e-01 -1.02660798e-01 -2.41473690e-01 7.59323180e-01
-4.68368866e-02 -1.26670867e-01 1.34995773e-01 1.02004337e+00
-5.76680660e-01 -1.56630719e+00 -5.07818937e-01 4.20553535e-01
-5.08505583e-01 -1.86815426e-01 -3.93658340e-01 3.88358057e-01
-6.98746890e-02 1.44264698e+00 -2.54247636e-01 -1.53577656e-01
8.12865376e-01 5.26980937e-01 3.27059686e-01 -8.78452182e-01
-8.29915941e-01 -5.27562499e-01 1.01291038e-01 -1.47495568e-01
-3.57588172e-01 -5.14089882e-01 -1.04579842e+00 3.49426478e-01
-3.07389200e-01 5.66540718e-01 4.35811877e-01 1.05779636e+00
7.14791656e-01 3.26468080e-01 5.54326713e-01 1.50412470e-01
-9.41274166e-01 -1.13334644e+00 -7.75502995e-02 7.22568810e-01
5.39255619e-01 -3.82532448e-01 -3.21860194e-01 2.79682666e-01] | [11.264833450317383, 8.06272029876709] |
7764481e-1f8c-448b-b63f-a4ff40567025 | acr-pose-adversarial-canonical-representation | 2111.10524 | null | https://arxiv.org/abs/2111.10524v1 | https://arxiv.org/pdf/2111.10524v1.pdf | ACR-Pose: Adversarial Canonical Representation Reconstruction Network for Category Level 6D Object Pose Estimation | Recently, category-level 6D object pose estimation has achieved significant improvements with the development of reconstructing canonical 3D representations. However, the reconstruction quality of existing methods is still far from excellent. In this paper, we propose a novel Adversarial Canonical Representation Reconstruction Network named ACR-Pose. ACR-Pose consists of a Reconstructor and a Discriminator. The Reconstructor is primarily composed of two novel sub-modules: Pose-Irrelevant Module (PIM) and Relational Reconstruction Module (RRM). PIM tends to learn canonical-related features to make the Reconstructor insensitive to rotation and translation, while RRM explores essential relational information between different input modalities to generate high-quality features. Subsequently, a Discriminator is employed to guide the Reconstructor to generate realistic canonical representations. The Reconstructor and the Discriminator learn to optimize through adversarial training. Experimental results on the prevalent NOCS-CAMERA and NOCS-REAL datasets demonstrate that our method achieves state-of-the-art performance. | ['Jun He', 'Hongyan Liu', 'Kejian Wu', 'Zhicheng Wang', 'Jian Xu', 'Zhengbo Song', 'Zhaoxin Fan'] | 2021-11-20 | null | null | null | null | ['6d-pose-estimation'] | ['computer-vision'] | [ 2.59849668e-01 2.09987149e-01 6.07191436e-02 -3.09809148e-01
-1.03356373e+00 -4.31801498e-01 5.56449234e-01 -5.34431815e-01
-1.00867143e-02 3.92667323e-01 8.55186805e-02 1.20561630e-01
2.37239286e-01 -6.98457658e-01 -1.21608138e+00 -6.65209353e-01
1.07272260e-01 4.67184156e-01 2.07530022e-01 -5.34529425e-02
-3.15431952e-02 7.39255488e-01 -1.21123028e+00 4.83547300e-01
6.15134537e-01 1.32430744e+00 2.96744734e-01 3.34368348e-01
2.51255065e-01 7.22233534e-01 -5.74372649e-01 -2.97152549e-01
5.35036087e-01 -3.72638673e-01 -3.82539272e-01 -7.39707984e-03
3.59977067e-01 -4.60364789e-01 -8.38779509e-01 7.69169390e-01
5.79666317e-01 -1.28609449e-01 6.94710493e-01 -1.21249223e+00
-5.76656818e-01 3.44211668e-01 -4.91416126e-01 -3.05664808e-01
4.27315593e-01 2.15731755e-01 7.18391538e-01 -9.64779198e-01
7.64739335e-01 1.57080901e+00 6.16165400e-01 7.68752038e-01
-1.35996246e+00 -9.81340110e-01 2.21000969e-01 -2.38475446e-02
-1.46894801e+00 -3.64380836e-01 1.02206135e+00 -1.95575163e-01
6.92849517e-01 1.42472148e-01 6.38950050e-01 1.49901676e+00
4.80403513e-01 8.66657138e-01 9.99878824e-01 -1.24231510e-01
8.73588547e-02 7.17652813e-02 -6.79198623e-01 6.60974145e-01
9.00210068e-02 3.99783641e-01 -4.47552174e-01 5.08607291e-02
1.25341332e+00 5.63701661e-03 -2.97948211e-01 -9.22071040e-01
-1.18053579e+00 5.91990232e-01 8.84062946e-01 -1.85504138e-01
-2.99201339e-01 4.15338695e-01 2.78355360e-01 2.60318458e-01
1.06905118e-01 4.36031044e-01 -1.33636802e-01 2.55630732e-01
-4.82476652e-01 3.57178301e-01 6.11755550e-01 1.18565822e+00
5.01396418e-01 2.95961261e-01 -1.22064278e-01 9.23085213e-01
4.86457020e-01 6.08002305e-01 3.07805508e-01 -1.10135877e+00
7.04465508e-01 5.50217271e-01 -1.65261194e-01 -1.15911961e+00
-1.44440815e-01 -6.70964241e-01 -9.89073217e-01 4.34114993e-01
8.02633911e-02 3.40716541e-01 -9.00982738e-01 1.74963880e+00
4.43307370e-01 2.41807744e-01 -5.58525026e-02 1.11046314e+00
7.44708180e-01 5.40249407e-01 -1.76312461e-01 2.20170662e-01
9.33457732e-01 -8.20322037e-01 -4.00472879e-01 -1.87286988e-01
7.83740019e-04 -8.06288540e-01 7.13649750e-01 3.53512287e-01
-1.18975294e+00 -7.05988467e-01 -1.38646090e+00 -1.38932660e-01
-2.18108103e-01 3.31167467e-02 4.37344074e-01 2.96752542e-01
-6.76503062e-01 5.93452036e-01 -8.91719282e-01 6.41634539e-02
6.54000640e-01 4.80227649e-01 -6.87078357e-01 -4.27355975e-01
-1.02589881e+00 8.50856066e-01 1.94963425e-01 1.33076413e-02
-1.60729551e+00 -6.87763870e-01 -9.59797263e-01 -3.49931926e-01
2.54502177e-01 -1.01840699e+00 1.03650892e+00 -9.05057371e-01
-1.82692945e+00 6.62680566e-01 2.60623842e-01 -1.51711911e-01
7.61639595e-01 -8.28599632e-02 -4.35685843e-01 4.03726906e-01
1.47669822e-01 7.99416900e-01 1.29530811e+00 -1.90554798e+00
-2.77023315e-01 -4.21016872e-01 8.69516432e-02 2.56539673e-01
1.16877273e-01 -4.41180706e-01 -8.46811295e-01 -1.06105185e+00
5.28760970e-01 -9.77734268e-01 -1.61778420e-01 5.43839514e-01
-4.92230684e-01 2.23634601e-01 8.76441240e-01 -5.23029208e-01
5.44383645e-01 -2.25122762e+00 5.34393311e-01 1.64754704e-01
2.69678622e-01 1.06145084e-01 -5.16569853e-01 2.22040236e-01
-3.76416862e-01 -2.28432074e-01 4.68114112e-03 -6.52326584e-01
-1.06070489e-01 4.24762905e-01 -3.74304324e-01 7.67729580e-01
5.14058948e-01 9.49090004e-01 -9.90745425e-01 -3.09055984e-01
4.32481706e-01 8.54888976e-01 -7.49133050e-01 5.48226297e-01
-9.39892083e-02 7.28391290e-01 -7.32795119e-01 9.08989429e-01
9.38374162e-01 -1.09799929e-01 1.26658782e-01 -5.76105237e-01
2.80618817e-01 2.18301713e-01 -1.05557072e+00 2.03495026e+00
-5.46405375e-01 2.29513958e-01 1.24528475e-01 -9.78207409e-01
8.36946249e-01 1.08283438e-01 5.11266410e-01 -7.90285528e-01
2.99546242e-01 7.63744637e-02 -2.80315101e-01 -1.54485449e-01
3.41412127e-01 -1.38234153e-01 -2.87556410e-01 3.56437117e-02
1.40676484e-01 -4.64882553e-01 -5.12487590e-01 2.53091693e-01
1.08028793e+00 5.20734429e-01 5.05492687e-02 1.48855522e-01
5.85847259e-01 -4.22929287e-01 6.43609345e-01 3.93206239e-01
-2.14959569e-02 1.04526639e+00 4.97947574e-01 -2.03083605e-01
-1.07432234e+00 -1.70736957e+00 -9.46844295e-02 3.28999013e-01
5.59611559e-01 -2.71017611e-01 -4.89958495e-01 -9.10394490e-01
2.55897224e-01 6.18954718e-01 -8.29705298e-01 -5.40282309e-01
-6.28072917e-01 -4.47310023e-02 5.93533933e-01 7.01880157e-01
4.74971861e-01 -7.82074153e-01 -3.64276260e-01 1.50782719e-01
-1.90890636e-02 -1.18781006e+00 -5.50664246e-01 2.22537637e-01
-9.10248935e-01 -1.01910543e+00 -4.70483512e-01 -3.88227403e-01
9.49805558e-01 4.24383521e-01 1.02710891e+00 -2.63825744e-01
-2.96152711e-01 2.51543790e-01 -4.03912723e-01 -1.20056257e-01
-6.02828622e-01 -1.26706347e-01 3.21025811e-02 7.41839483e-02
-3.81201774e-01 -9.43771660e-01 -7.11264312e-01 6.66081369e-01
-1.00569403e+00 1.34951994e-01 1.04021585e+00 9.54145312e-01
1.12418199e+00 -3.09126556e-01 3.66174579e-01 -6.91279829e-01
-1.24621682e-01 -4.67182457e-01 -5.09650588e-01 -1.39890581e-01
-4.30463046e-01 2.50731289e-01 8.72422636e-01 -5.61261594e-01
-9.86854374e-01 3.74424577e-01 -1.61889672e-01 -1.05787349e+00
5.56926504e-02 9.61712562e-03 -7.33903885e-01 -1.99371547e-01
4.54981625e-01 2.43830055e-01 3.42079997e-02 -5.72323978e-01
3.45852077e-01 3.10820878e-01 7.67870307e-01 -6.57660127e-01
1.37829196e+00 7.10874796e-01 1.64862335e-01 -3.80534649e-01
-7.84155071e-01 -8.31999108e-02 -5.31353712e-01 -2.56750882e-01
6.10315740e-01 -1.24544609e+00 -4.35752213e-01 3.71539205e-01
-1.05208886e+00 -1.23094283e-01 -6.07815266e-01 4.07133430e-01
-8.12528253e-01 7.37783536e-02 -2.70683140e-01 -2.38426626e-01
-9.19515342e-02 -1.34874666e+00 1.33308995e+00 6.95661977e-02
8.08061808e-02 -5.12430012e-01 -8.40348303e-02 5.29471159e-01
2.23196745e-01 7.06287503e-01 7.84110129e-01 -1.35629073e-01
-1.17359865e+00 -3.59273165e-01 -1.50291011e-01 5.67747414e-01
2.18491964e-02 -4.49772269e-01 -9.50761616e-01 -5.47955036e-01
-1.44992113e-01 -4.95312870e-01 7.06988513e-01 -5.21565340e-02
1.44049025e+00 -1.81427613e-01 -2.81530619e-01 1.21590507e+00
1.38760698e+00 -3.50867137e-02 9.97002423e-01 1.95186213e-01
9.24582243e-01 1.88957438e-01 5.78955948e-01 1.40482634e-01
3.61442387e-01 8.43494296e-01 9.62812722e-01 1.86027989e-01
-5.23664892e-01 -6.99072778e-01 6.18849516e-01 8.32107186e-01
-9.10311267e-02 -1.80165954e-02 -4.27397996e-01 2.32700139e-01
-1.57585490e+00 -7.62037039e-01 3.69601965e-01 2.01987600e+00
6.28279626e-01 2.05104351e-01 -2.78181612e-01 5.82160838e-02
3.75509739e-01 2.89263338e-01 -9.17758644e-01 -2.53766086e-02
-4.82664444e-02 1.86503649e-01 3.34822178e-01 8.54963139e-02
-8.77232730e-01 8.03712249e-01 5.46820927e+00 8.79476070e-01
-1.05985618e+00 9.59017277e-02 3.46731156e-01 -1.31213844e-01
-3.42467278e-01 -1.13894314e-01 -5.39038062e-01 2.64577717e-01
4.98133570e-01 2.96177864e-01 5.75016260e-01 1.02021003e+00
-1.96998864e-01 2.03816235e-01 -1.32192171e+00 9.88908827e-01
2.28262380e-01 -1.22097015e+00 4.89455342e-01 -1.15288785e-02
7.33574033e-01 3.41302082e-02 2.50389606e-01 4.68381286e-01
5.06736450e-02 -1.15095878e+00 1.14926481e+00 7.09577739e-01
1.17797613e+00 -9.86727118e-01 4.35725272e-01 1.86105505e-01
-1.33618152e+00 8.59549344e-02 -4.43869710e-01 5.24755895e-01
-8.63942206e-02 2.92864144e-01 -5.58560073e-01 1.02783036e+00
5.26191652e-01 9.58667874e-01 -4.77643073e-01 6.35770261e-01
-3.81031781e-01 -3.56545709e-02 -1.87175348e-01 5.21107733e-01
1.71155599e-03 9.06815007e-02 8.50142956e-01 7.87678897e-01
5.92826828e-02 -1.32604137e-01 5.99689484e-02 9.58487034e-01
-4.37367141e-01 -3.53344828e-01 -5.51361978e-01 1.85278922e-01
5.71197450e-01 1.16069233e+00 -3.46048146e-01 1.00384496e-01
-2.70223737e-01 1.27219784e+00 4.88447279e-01 1.89890847e-01
-9.55476522e-01 3.52955908e-02 8.62312078e-01 3.22661221e-01
4.55983102e-01 -2.02743337e-01 -1.68823466e-01 -1.14545333e+00
1.29858285e-01 -1.05661690e+00 9.83314067e-02 -8.45286131e-01
-1.30766726e+00 6.79688692e-01 -2.09821519e-02 -1.75166857e+00
-2.78294116e-01 -5.81433952e-01 -3.02159309e-01 9.52400863e-01
-1.50601578e+00 -1.46594834e+00 -5.33983588e-01 7.78613925e-01
5.07372141e-01 -3.70138288e-01 8.98461044e-01 2.92612791e-01
-4.96650189e-01 9.64640856e-01 5.53357713e-02 8.56779888e-02
6.40616894e-01 -9.95044589e-01 1.99578926e-01 5.18785000e-01
-1.70374662e-01 5.37989020e-01 4.53981817e-01 -5.08595705e-01
-1.93074417e+00 -1.44716299e+00 4.50987183e-03 -5.19161880e-01
3.00968081e-01 -5.42889416e-01 -5.37577331e-01 6.72618270e-01
-2.23901659e-01 5.82975924e-01 3.63643974e-01 -4.46943164e-01
-9.47638571e-01 -4.32450444e-01 -1.18394840e+00 5.60050547e-01
1.19964933e+00 -7.55098760e-01 -5.23940265e-01 1.09974392e-01
7.36846745e-01 -8.38084519e-01 -1.03231668e+00 3.77449989e-01
8.29263031e-01 -9.77674365e-01 1.43076372e+00 -2.04435199e-01
6.76562130e-01 -4.68732566e-01 -5.14599144e-01 -1.35058606e+00
-1.77069291e-01 -4.55717027e-01 -4.93095279e-01 9.15424347e-01
-7.79675692e-02 -3.74723077e-01 7.54843116e-01 5.67362085e-03
-4.19072092e-01 -1.06485116e+00 -1.18682957e+00 -1.00055838e+00
2.16907058e-02 -5.11528492e-01 6.98814869e-01 7.12864578e-01
-6.03873193e-01 4.22508210e-01 -3.91899973e-01 3.20502430e-01
8.55111241e-01 1.75249994e-01 1.07757282e+00 -7.93494582e-01
-5.47561347e-01 -1.20449312e-01 -5.48370004e-01 -1.40451181e+00
2.54745424e-01 -9.73927975e-01 4.51926254e-02 -1.22859120e+00
4.43254635e-02 -5.17270207e-01 -3.43214154e-01 2.43865594e-01
1.15799755e-01 3.23642403e-01 2.24770993e-01 3.00202131e-01
-5.57677984e-01 1.00762498e+00 1.80795085e+00 -3.00997883e-01
1.08493410e-01 -2.86106616e-02 -6.49281800e-01 6.42598569e-01
4.43420857e-01 -6.67893767e-01 -4.37969387e-01 -4.62754667e-01
-8.77130926e-02 1.66269600e-01 7.31921971e-01 -1.00039160e+00
-1.49280727e-02 1.10638730e-01 9.17771041e-01 -8.39749038e-01
7.23880351e-01 -1.11431074e+00 4.53937292e-01 4.11536723e-01
1.42005933e-02 -4.39618528e-02 1.26462638e-01 8.47049534e-01
-2.13705808e-01 2.06867158e-01 1.09502566e+00 -2.79844403e-02
-4.18605596e-01 5.75190365e-01 1.79841310e-01 2.13675052e-02
1.04843771e+00 -2.84957796e-01 -2.09155418e-02 -2.19725922e-01
-3.74060035e-01 1.68908685e-01 5.67845821e-01 6.50168777e-01
1.03260744e+00 -1.67673159e+00 -6.42399669e-01 5.03782451e-01
2.19771579e-01 4.59266931e-01 3.02568465e-01 4.62147206e-01
-5.26914299e-01 1.64945588e-01 -3.98488879e-01 -7.14404106e-01
-7.82225132e-01 5.42165697e-01 4.64873672e-01 -4.58035439e-01
-7.46923327e-01 7.59428561e-01 3.55488271e-01 -5.48650265e-01
4.39039916e-01 -1.96926743e-01 1.12968929e-01 -2.31967762e-01
2.90322155e-01 1.87482163e-01 -2.01324418e-01 -8.14461350e-01
-4.32178855e-01 5.77575862e-01 -1.02817252e-01 6.43522665e-02
1.51381290e+00 -1.75205562e-02 1.23269483e-01 1.86825350e-01
1.55955255e+00 -1.49466977e-01 -1.60055983e+00 -1.48710683e-01
-6.39444590e-01 -6.65623665e-01 9.19164196e-02 -7.11303234e-01
-1.44134390e+00 6.94785476e-01 6.60141826e-01 -3.27201188e-01
1.22749448e+00 1.89028364e-02 7.32655525e-01 1.73665747e-01
6.38372600e-01 -7.95151591e-01 5.83014190e-01 3.86118233e-01
1.49727440e+00 -1.03326154e+00 1.33506849e-01 -5.26533365e-01
-3.47958386e-01 9.54324961e-01 8.50731254e-01 -4.76617247e-01
6.10286772e-01 2.45972022e-01 -1.38649389e-01 -7.37065524e-02
-5.59039295e-01 2.88548708e-01 4.38077569e-01 8.13719094e-01
-9.87907499e-02 -4.75743078e-02 2.70997107e-01 7.33356476e-01
-7.75044262e-02 -3.23628336e-01 6.27490357e-02 7.80385435e-01
1.33835450e-01 -1.22909725e+00 -4.38737333e-01 1.92381829e-01
-1.82174757e-01 2.46053293e-01 -1.73130259e-01 1.00510371e+00
2.30795026e-01 4.54251975e-01 8.32068995e-02 -8.02793920e-01
6.85825348e-01 -4.64527249e-01 7.53138781e-01 -5.27757466e-01
-4.65084821e-01 2.96836696e-03 -7.25536272e-02 -1.24890745e+00
-2.38452002e-01 -7.00587869e-01 -1.12233567e+00 -1.32847890e-01
-1.69203311e-01 -1.70296893e-01 6.55631244e-01 6.19843006e-01
2.49453679e-01 1.00944471e+00 9.88772273e-01 -1.19360399e+00
-8.55900407e-01 -7.24111557e-01 -4.90320235e-01 4.13710386e-01
3.05860341e-01 -9.49415743e-01 -1.74389049e-01 -3.02110128e-02] | [8.526199340820312, -3.2519333362579346] |
34fc8b85-961a-4a90-a402-e7e9646e3b70 | acnet-approaching-and-centralizing-network | 2111.12757 | null | https://arxiv.org/abs/2111.12757v4 | https://arxiv.org/pdf/2111.12757v4.pdf | ACNet: Approaching-and-Centralizing Network for Zero-Shot Sketch-Based Image Retrieval | The huge domain gap between sketches and photos and the highly abstract sketch representations pose challenges for sketch-based image retrieval (\underline{SBIR}). The zero-shot sketch-based image retrieval (\underline{ZS-SBIR}) is more generic and practical but poses an even greater challenge because of the additional knowledge gap between the seen and unseen categories. To simultaneously mitigate both gaps, we propose an \textbf{A}pproaching-and-\textbf{C}entralizing \textbf{Net}work (termed "\textbf{ACNet}") to jointly optimize sketch-to-photo synthesis and the image retrieval. The retrieval module guides the synthesis module to generate large amounts of diverse photo-like images which gradually approach the photo domain, and thus better serve the retrieval module than ever to learn domain-agnostic representations and category-agnostic common knowledge for generalizing to unseen categories. These diverse images generated with retrieval guidance can effectively alleviate the overfitting problem troubling concrete category-specific training samples with high gradients. We also discover the use of proxy-based NormSoftmax loss is effective in the zero-shot setting because its centralizing effect can stabilize our joint training and promote the generalization ability to unseen categories. Our approach is simple yet effective, which achieves state-of-the-art performance on two widely used ZS-SBIR datasets and surpasses previous methods by a large margin. | ['Sai-Kit Yeung', 'Ying Shan', 'Yang Yang', 'Hong Lu', 'Yang Wu', 'Ziqiang Zheng', 'Hao Ren'] | 2021-11-24 | null | null | null | null | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 1.97005436e-01 -6.39583245e-02 -3.27615589e-01 -1.53999403e-01
-1.20252037e+00 -8.12043428e-01 9.89142597e-01 -4.77533013e-01
-5.22549562e-02 6.12351477e-01 1.18691161e-01 2.55285561e-01
-4.13802356e-01 -7.25723445e-01 -7.50822544e-01 -6.83467031e-01
3.99050534e-01 4.63845313e-01 5.13862446e-02 -4.18255031e-01
2.81240553e-01 5.40988564e-01 -1.86853266e+00 3.12002063e-01
6.50317669e-01 1.16187489e+00 3.32602680e-01 5.02527177e-01
-2.57755190e-01 4.69993025e-01 -4.16844219e-01 -7.02757061e-01
5.89866996e-01 -1.57415450e-01 -5.54847896e-01 3.31041247e-01
1.19379246e+00 -5.35103619e-01 -7.15049088e-01 7.69670725e-01
5.83338141e-01 4.80563134e-01 1.02675033e+00 -1.44534242e+00
-1.30326498e+00 3.32256518e-02 -4.35993612e-01 -1.26116842e-01
2.49913886e-01 2.71670043e-01 1.19723392e+00 -1.47220325e+00
9.81143475e-01 1.43145227e+00 4.66548234e-01 9.64575648e-01
-1.33033133e+00 -8.20061684e-01 3.90191019e-01 1.02966979e-01
-1.72743118e+00 -4.55297112e-01 9.53560948e-01 -3.86031181e-01
4.72543716e-01 3.51033449e-01 4.11516100e-01 1.38151872e+00
-5.15098035e-01 1.09933329e+00 7.49351442e-01 -1.73093349e-01
2.13418260e-01 2.22536936e-01 -2.17202395e-01 4.63990659e-01
5.90366079e-03 2.01651618e-01 -5.84603608e-01 -9.47352797e-02
1.11018789e+00 5.10208189e-01 -2.49041021e-01 -6.75345719e-01
-1.12147951e+00 8.21075261e-01 7.71099627e-01 1.68174312e-01
-2.13920370e-01 3.89803320e-01 2.27497101e-01 5.19338429e-01
4.36474919e-01 6.19044721e-01 -5.37689403e-02 2.76924938e-01
-1.17676127e+00 3.83728743e-01 5.15116215e-01 1.32362020e+00
9.24425542e-01 2.65444458e-01 -3.11892569e-01 1.44881415e+00
-7.43831545e-02 7.20062375e-01 2.23623365e-01 -1.10792756e+00
4.32222366e-01 4.09502357e-01 1.16733931e-01 -9.09210205e-01
3.74357611e-01 -2.75814980e-01 -9.48550105e-01 1.60446048e-01
2.98882186e-01 2.91006833e-01 -1.27389634e+00 1.74282110e+00
-6.40745759e-02 1.04750350e-01 4.75703506e-03 1.09688640e+00
1.01875997e+00 7.70912468e-01 4.01234239e-01 3.35773587e-01
1.09620607e+00 -9.84484732e-01 -2.92183876e-01 -3.53434443e-01
-8.30673352e-02 -7.60309458e-01 1.47678041e+00 1.62102759e-01
-1.12834382e+00 -5.69328785e-01 -1.06231594e+00 -2.98855960e-01
-6.55572653e-01 2.36098379e-01 5.83265007e-01 3.69736880e-01
-1.23515046e+00 6.27852619e-01 -9.51435976e-03 -5.53494513e-01
8.42872977e-01 1.59095451e-01 -4.79466170e-01 -6.54510856e-01
-9.37191188e-01 5.68731368e-01 1.58996478e-01 -2.41491854e-01
-1.25377893e+00 -9.38044071e-01 -8.13678801e-01 1.36787683e-01
6.44326389e-01 -8.23259473e-01 7.53706872e-01 -9.81503129e-01
-1.25679064e+00 1.03461850e+00 2.09079742e-01 4.07329313e-02
5.84499538e-01 6.84396103e-02 -2.17864126e-01 4.44713563e-01
1.15631431e-01 1.33389115e+00 1.41273594e+00 -1.58541667e+00
-2.86874682e-01 -2.38722771e-01 1.08747043e-01 3.72223258e-01
-5.71983159e-01 -4.08305258e-01 -7.19856739e-01 -1.17060018e+00
1.37833208e-01 -8.76895010e-01 3.20342928e-02 7.81661808e-01
-9.10936669e-02 -3.65703166e-01 8.07466269e-01 -3.45244199e-01
8.47597539e-01 -2.14838433e+00 2.31597900e-01 -1.05065424e-02
2.04475701e-01 3.69145930e-01 -7.35113263e-01 7.64850855e-01
-2.64676455e-02 1.12016656e-01 -1.49230123e-01 -2.54769236e-01
4.61555831e-02 3.86498898e-01 -7.84748137e-01 1.49397314e-01
4.89425719e-01 1.17826319e+00 -1.11470735e+00 -3.49538803e-01
3.67617965e-01 4.46805775e-01 -5.03318131e-01 2.30552718e-01
-3.99324208e-01 4.80277948e-02 -3.64314586e-01 9.81298983e-01
7.25444853e-01 -4.35872167e-01 -4.10291851e-02 -7.52200186e-02
3.68532002e-01 -3.81880194e-01 -1.06377912e+00 1.94394243e+00
-4.95969117e-01 5.19396722e-01 9.99983866e-03 -9.21981931e-01
1.05848944e+00 1.77905694e-01 3.33393723e-01 -5.68123162e-01
-3.06153595e-01 1.53945744e-01 -7.43610978e-01 -2.02339232e-01
5.89262068e-01 -3.64197284e-01 -9.20021832e-02 3.20480704e-01
3.53923887e-01 -5.58994651e-01 2.81236321e-02 4.47106987e-01
6.49818361e-01 1.94391057e-01 -1.31967857e-01 -3.14849734e-01
3.86621654e-01 -7.08248615e-02 1.36894271e-01 9.87346530e-01
-2.39772089e-02 9.73615110e-01 1.14054523e-01 -5.44639111e-01
-1.40137887e+00 -1.52971184e+00 5.19925775e-03 1.35128021e+00
4.96186316e-01 -1.41910836e-01 -2.98827469e-01 -6.61522508e-01
4.12814140e-01 4.49039549e-01 -5.21354675e-01 -3.92872512e-01
-2.43753746e-01 -1.56750664e-01 4.73536193e-01 6.05128586e-01
4.30525154e-01 -1.03842044e+00 1.84513599e-01 1.42941456e-02
-1.92456082e-01 -8.31269145e-01 -6.95853710e-01 -3.88956517e-01
-6.62846625e-01 -8.62403691e-01 -1.48829114e+00 -8.56897235e-01
8.22779953e-01 6.71068966e-01 1.32875752e+00 2.44557768e-01
-6.04281902e-01 8.52194071e-01 -2.31360778e-01 -2.05944568e-01
-2.68601049e-02 -1.74469024e-01 -1.50387838e-01 7.26850554e-02
1.60676211e-01 -5.97856522e-01 -9.74076807e-01 5.19055545e-01
-1.17552149e+00 -2.28468642e-01 6.30921543e-01 1.14377201e+00
5.30198872e-01 -4.36583966e-01 8.11482787e-01 -4.37017769e-01
6.35818779e-01 -4.32578951e-01 -4.10078734e-01 5.23373544e-01
-6.54540837e-01 -8.00231285e-03 5.56451678e-01 -9.13892925e-01
-9.96085882e-01 -2.40947619e-01 2.03488678e-01 -1.11215258e+00
9.66036841e-02 -7.10173100e-02 -7.86890537e-02 -1.76699907e-01
7.79329419e-01 3.80479634e-01 2.96951104e-02 -4.99078572e-01
8.51236403e-01 5.74179351e-01 5.97415447e-01 -9.05499697e-01
9.85993266e-01 6.40029728e-01 -1.15461640e-01 -9.36564147e-01
-7.62934327e-01 -6.12748981e-01 -2.41432890e-01 -2.09727868e-01
6.06310725e-01 -1.21302962e+00 -4.97425914e-01 2.45900214e-01
-1.01634669e+00 -1.79714918e-01 -6.32045865e-01 -1.05913460e-01
-7.39719927e-01 4.47810709e-01 -4.28630143e-01 -8.29995811e-01
-4.98086900e-01 -8.78603220e-01 1.54765749e+00 1.62952006e-01
9.71545279e-02 -7.56388187e-01 -2.86152124e-01 2.77115732e-01
3.70431095e-01 -1.78584959e-02 9.44970846e-01 -3.51949930e-01
-1.11380172e+00 -3.67169857e-01 -7.67330945e-01 6.34071290e-01
-2.95246154e-01 -1.88347533e-01 -1.12092495e+00 -4.79469061e-01
-4.76157546e-01 -7.87962675e-01 1.08232427e+00 -8.37935228e-03
1.37476432e+00 -3.15211535e-01 -1.98516637e-01 5.20356357e-01
1.67493606e+00 -1.49860859e-01 9.02551115e-01 -1.47125423e-01
6.11088514e-01 5.28149307e-01 6.12054169e-01 3.13365430e-01
1.94191769e-01 6.59820437e-01 2.47519210e-01 -1.06035702e-01
-6.80618882e-01 -6.23417556e-01 2.82714833e-02 3.91383708e-01
-6.77784160e-03 -3.19294959e-01 -5.64425468e-01 9.26211536e-01
-1.72409689e+00 -1.21295369e+00 5.35583019e-01 2.30714393e+00
7.47074127e-01 -3.89625102e-01 1.66293383e-02 -2.29484960e-01
7.66244531e-01 4.37902749e-01 -5.47309518e-01 -2.63240132e-02
-1.51563019e-01 3.68857592e-01 2.03479663e-01 2.32143298e-01
-8.58810604e-01 1.29755092e+00 6.21924734e+00 1.53991842e+00
-1.02273440e+00 4.36924696e-02 5.15831470e-01 -1.95433006e-01
-5.78316689e-01 -7.61107262e-03 -5.87315381e-01 4.14558500e-01
1.90345332e-01 -1.34040222e-01 8.28825772e-01 1.14506888e+00
-4.55872983e-01 2.54558235e-01 -1.18047500e+00 1.41831493e+00
2.04748049e-01 -1.66507649e+00 6.28860176e-01 -2.17473432e-02
9.37464833e-01 -1.06614195e-01 4.34777051e-01 5.17339349e-01
3.09554070e-01 -9.09010530e-01 6.95306540e-01 7.26906359e-01
1.49207592e+00 -4.06117022e-01 8.58379155e-03 6.85666949e-02
-1.21015739e+00 -3.65103424e-01 -8.42268050e-01 3.32895786e-01
-1.96087748e-01 3.30799907e-01 -6.09990299e-01 4.48495626e-01
5.49969077e-01 7.33493328e-01 -4.23458606e-01 7.70145655e-01
-1.52738944e-01 -8.60811323e-02 -1.51498660e-01 -6.01949170e-02
2.98609942e-01 -2.38662615e-01 6.07920647e-01 9.01665866e-01
3.72413129e-01 2.42985100e-01 2.89051414e-01 1.12923956e+00
-4.07138944e-01 -1.07661784e-01 -9.41734076e-01 -3.02380174e-01
5.67159057e-01 1.15645289e+00 -4.05398309e-01 -6.28631473e-01
-9.17087942e-02 1.06511736e+00 3.50050360e-01 7.07201481e-01
-3.54236841e-01 -3.91647220e-01 7.49242604e-01 1.86569795e-01
3.84962887e-01 -5.72237447e-02 -1.99045420e-01 -1.40703046e+00
7.50340223e-02 -7.48425186e-01 4.01797980e-01 -1.12500560e+00
-1.83064604e+00 4.16688591e-01 3.77408154e-02 -1.45695662e+00
-1.43655285e-01 -5.47417760e-01 -4.66141611e-01 8.93167436e-01
-1.63791716e+00 -1.50060081e+00 -4.81511354e-01 7.99658895e-01
9.62390900e-01 -3.95654857e-01 7.73466229e-01 3.93387705e-01
-9.14390162e-02 8.60846937e-01 2.53576934e-01 -7.54065765e-03
9.35405612e-01 -1.05657268e+00 2.54637867e-01 4.77861971e-01
2.02583164e-01 6.41177952e-01 3.82082343e-01 -6.61994040e-01
-1.51163018e+00 -1.09421241e+00 5.11574745e-01 -6.21215940e-01
6.05653763e-01 -5.29073477e-01 -7.90678024e-01 2.32342452e-01
-1.41397610e-01 9.19762552e-02 2.86997765e-01 -3.46126296e-02
-9.62270796e-01 -2.97943234e-01 -1.29221845e+00 9.00853038e-01
1.41198266e+00 -9.79116797e-01 -5.68625867e-01 5.07943869e-01
6.02658987e-01 5.71079999e-02 -8.59619319e-01 2.27212116e-01
7.28836656e-01 -6.79034591e-01 1.60505581e+00 -5.49703896e-01
7.47359097e-01 3.71431559e-02 -3.43212545e-01 -9.90046442e-01
-3.28268409e-01 -5.42867601e-01 6.11131415e-02 1.19010997e+00
9.11057740e-02 -3.52685273e-01 9.20274258e-01 7.28525043e-01
1.18925966e-01 -6.25870645e-01 -7.08214760e-01 -1.08814418e+00
2.72442043e-01 -8.27307403e-02 6.18781507e-01 9.33357477e-01
-3.01769286e-01 1.28931433e-01 -5.11541903e-01 -2.43565366e-01
8.71087909e-01 2.55973041e-01 7.90108800e-01 -1.18356490e+00
-6.76957592e-02 -6.08277082e-01 -3.85279745e-01 -1.17269838e+00
7.73161724e-02 -1.09007621e+00 -3.75693180e-02 -1.35264289e+00
3.13610971e-01 -7.05779314e-01 -3.40433270e-01 5.22997379e-01
-1.42580912e-01 6.48461819e-01 5.94681084e-01 5.79601049e-01
-6.42396450e-01 7.26148248e-01 1.55114102e+00 -4.65466797e-01
1.38573036e-01 -2.26216555e-01 -7.99133897e-01 2.02763975e-01
1.73535645e-01 -2.61886507e-01 -7.06579864e-01 -3.83076906e-01
1.56711843e-02 8.03822428e-02 7.87884414e-01 -7.46086955e-01
9.73578170e-02 -2.35608608e-01 4.93695498e-01 -5.39145589e-01
8.58336091e-01 -7.24516630e-01 7.07700998e-02 -2.65631732e-02
-4.57820058e-01 -3.93340230e-01 -7.64778703e-02 9.65599239e-01
-7.79179484e-02 -1.31432265e-01 8.45258176e-01 -4.93429124e-01
-8.60266089e-01 6.93456709e-01 9.30896550e-02 9.30174291e-02
8.72935832e-01 -4.62751389e-01 -5.40072381e-01 -6.61691010e-01
-6.05621517e-01 2.42596552e-01 7.80284882e-01 6.93299651e-01
9.29509461e-01 -1.67669129e+00 -5.01464486e-01 3.31373096e-01
5.39674580e-01 -1.99243337e-01 5.12080133e-01 2.18048453e-01
-1.27137840e-01 3.07470500e-01 -2.61746377e-01 -6.33038878e-01
-7.55511045e-01 8.70391011e-01 2.20692344e-02 7.44684264e-02
-5.85083723e-01 1.07806039e+00 6.48525715e-01 -4.30432111e-01
3.95267159e-01 4.14040446e-01 3.01506609e-01 1.09117471e-01
4.91956890e-01 3.76423776e-01 -2.13960052e-01 -3.17710310e-01
-1.64579630e-01 6.65911436e-01 -1.94416672e-01 -7.24594370e-02
1.27887857e+00 -1.94204487e-02 1.50095835e-01 -1.36293704e-02
1.34652948e+00 -4.68806982e-01 -1.71072733e+00 -3.28357995e-01
-1.84552267e-01 -9.09425259e-01 -1.30111113e-01 -1.06575835e+00
-9.68816876e-01 1.00356734e+00 5.56225717e-01 -5.35872877e-02
9.15438414e-01 2.05540270e-01 8.03369284e-01 4.49416459e-01
4.71139401e-01 -1.15486372e+00 7.78474331e-01 1.31322592e-01
1.35858965e+00 -1.31934381e+00 8.10061172e-02 -4.01957840e-01
-8.09483886e-01 9.20383990e-01 5.39219141e-01 -4.68124539e-01
3.61944228e-01 -4.36420679e-01 -4.88483496e-02 -1.93942115e-01
-6.24380291e-01 -3.55668128e-01 6.82187259e-01 8.40160549e-01
-1.98261186e-01 -6.33258596e-02 -4.58389781e-02 3.63178492e-01
2.67628014e-01 1.57639328e-02 9.19647142e-02 7.15770900e-01
-3.19467098e-01 -1.19889581e+00 -2.59263396e-01 4.47678417e-01
8.16999525e-02 -1.46634594e-01 -5.61548591e-01 8.17066610e-01
-1.64863840e-01 7.44611740e-01 -3.09352931e-02 -1.80042267e-01
2.43916646e-01 1.09364986e-01 6.23345494e-01 -5.67488790e-01
-3.32341611e-01 -1.58016216e-02 -1.98331952e-01 -5.35100520e-01
-7.39984289e-02 -2.79569656e-01 -6.69451833e-01 -3.24877441e-01
-2.09133044e-01 -2.93965608e-01 6.38877988e-01 5.35784483e-01
5.46718597e-01 -8.44377652e-02 7.30098009e-01 -1.20197380e+00
-8.97151232e-01 -5.55180907e-01 -6.42086029e-01 9.02554572e-01
1.81142345e-01 -8.89760613e-01 -4.88523275e-01 -1.11232705e-01] | [11.617443084716797, 0.645307719707489] |
07c2eb57-98b6-403c-8942-8c635ad55e96 | towards-real-time-dnn-inference-on-mobile | 2004.11250 | null | https://arxiv.org/abs/2004.11250v1 | https://arxiv.org/pdf/2004.11250v1.pdf | Towards Real-Time DNN Inference on Mobile Platforms with Model Pruning and Compiler Optimization | High-end mobile platforms rapidly serve as primary computing devices for a wide range of Deep Neural Network (DNN) applications. However, the constrained computation and storage resources on these devices still pose significant challenges for real-time DNN inference executions. To address this problem, we propose a set of hardware-friendly structured model pruning and compiler optimization techniques to accelerate DNN executions on mobile devices. This demo shows that these optimizations can enable real-time mobile execution of multiple DNN applications, including style transfer, DNN coloring and super resolution. | ['Yanzhi Wang', 'Wei Niu', 'Pu Zhao', 'Zheng Zhan', 'Xue Lin', 'Bin Ren'] | 2020-04-22 | null | null | null | null | ['compiler-optimization'] | ['computer-code'] | [ 1.99639335e-01 -2.51574039e-01 -5.32041371e-01 -4.48055208e-01
-1.12912275e-01 -4.47384238e-01 3.36571783e-01 -5.34214795e-01
-5.07295191e-01 6.62499785e-01 -2.78520644e-01 -1.02186966e+00
1.99683398e-01 -8.58058691e-01 -6.51225805e-01 -3.68713319e-01
2.85014570e-01 5.73496997e-01 4.12118077e-01 -1.66555196e-01
-1.33505628e-01 8.60421956e-01 -1.81145668e+00 2.85212517e-01
7.28151560e-01 1.03798378e+00 4.10013765e-01 9.17773128e-01
-2.46858954e-01 8.76236260e-01 -6.13680959e-01 -4.83102202e-01
2.94175297e-01 4.20603268e-02 -4.97145057e-01 -5.21862686e-01
7.62175143e-01 -8.48442554e-01 -5.23951828e-01 1.29852533e+00
5.37055552e-01 -4.95819673e-02 4.26417496e-03 -1.37049901e+00
-3.19370739e-02 8.23769450e-01 -1.27774179e-01 5.26628375e-01
-3.23634714e-01 2.49572415e-02 4.89239901e-01 -6.15743220e-01
6.04299784e-01 1.23931265e+00 6.18097782e-01 1.08461475e+00
-8.88690591e-01 -9.06669140e-01 8.95912349e-02 2.81363487e-01
-1.11724472e+00 -8.32663715e-01 5.60488880e-01 1.62510604e-01
1.36950052e+00 2.77136713e-01 1.00257492e+00 1.30140948e+00
6.03821427e-02 9.32188272e-01 6.53791249e-01 -8.92613381e-02
6.22927904e-01 -3.85152340e-01 1.18295908e-01 7.13916659e-01
4.75652993e-01 -8.84326920e-02 -7.26401806e-01 2.32725739e-02
9.21296895e-01 1.09420143e-01 2.15870813e-01 2.27210343e-01
-7.59335101e-01 4.82900739e-01 2.91001052e-01 -2.18700543e-01
-1.06495038e-01 6.43931568e-01 8.03014815e-01 -3.38385031e-02
1.35577917e-01 -2.22348515e-02 -7.54979372e-01 -6.18095100e-01
-1.14064312e+00 3.68118584e-01 8.06023121e-01 1.29134607e+00
5.16049027e-01 4.99157250e-01 3.61028373e-01 5.34168243e-01
3.45294446e-01 7.83323228e-01 2.58837223e-01 -1.22747278e+00
6.29381835e-01 5.71633101e-01 -6.34523690e-01 -5.28180659e-01
-3.81330192e-01 -3.01332086e-01 -1.03916419e+00 2.48085320e-01
2.48331413e-01 -2.15893313e-01 -9.34501171e-01 1.53238893e+00
4.62523162e-01 4.41925317e-01 9.26156119e-02 5.80859423e-01
7.80741215e-01 5.93913853e-01 1.82524756e-01 6.90414533e-02
1.23610353e+00 -1.09719706e+00 -5.87637961e-01 -5.16158521e-01
8.12877774e-01 -4.41088080e-01 8.68928492e-01 4.10277635e-01
-1.22296774e+00 -4.85041231e-01 -1.13819814e+00 -6.17170513e-01
-2.53739208e-01 -2.37581022e-02 1.16407299e+00 7.49592423e-01
-1.28048253e+00 5.08320749e-01 -1.50837386e+00 2.76743863e-02
8.90909851e-01 9.64098155e-01 1.00576438e-01 -1.27309039e-01
-8.39163125e-01 5.14616609e-01 3.67851436e-01 3.48975182e-01
-8.25728536e-01 -9.38142955e-01 -5.38025081e-01 -5.84739074e-03
1.08560078e-01 -7.75953293e-01 1.53843248e+00 -8.69418919e-01
-1.84088647e+00 8.74149263e-01 -3.88126194e-01 -7.16171563e-01
3.16943258e-01 -2.96045512e-01 -5.73116720e-01 -8.53959247e-02
-4.16624129e-01 8.66202712e-01 4.48651046e-01 -5.53046823e-01
-8.72262418e-01 -5.65599918e-01 1.51184425e-01 1.80588081e-01
-3.83478194e-01 1.27571061e-01 -6.66375518e-01 -4.77208346e-01
-8.99850801e-02 -9.52052772e-01 -3.13466758e-01 1.66402698e-01
-4.05470908e-01 -6.86387196e-02 1.42765057e+00 -3.86626929e-01
1.05453801e+00 -2.07110953e+00 4.81609851e-02 5.19839525e-02
5.37444592e-01 7.35145628e-01 1.96034640e-01 -6.62818849e-01
5.00265837e-01 -5.28528262e-03 2.21867472e-01 -5.02403140e-01
3.09268925e-02 7.89642751e-01 -4.33445901e-01 -7.92573243e-02
-3.45239162e-01 1.02065516e+00 -5.90183973e-01 -4.45634693e-01
1.95095405e-01 5.81395030e-01 -7.91558504e-01 -2.55986959e-01
-7.11446047e-01 -1.43178622e-03 -2.96673894e-01 8.21527004e-01
7.72777259e-01 -1.49412761e-02 4.41376120e-01 -2.21684292e-01
6.26074448e-02 6.04751170e-01 -7.31625140e-01 1.64188039e+00
-4.54962999e-01 9.80076134e-01 3.61003935e-01 -6.01796925e-01
5.74068248e-01 -1.46051809e-01 -7.43851438e-02 -9.45941806e-01
1.75867081e-01 3.38308603e-01 -7.56000429e-02 1.02104312e-02
6.86632931e-01 3.90802652e-01 1.33910537e-01 -2.08508950e-02
-2.26679277e-02 3.13245445e-01 1.50667965e-01 5.65653443e-02
1.08958089e+00 -2.71780193e-01 -2.46440679e-01 -5.51409841e-01
5.53735122e-02 -4.67221588e-02 9.51984286e-01 6.56466246e-01
-2.13329166e-01 -1.96353436e-01 5.89084387e-01 -6.96660757e-01
-1.17265260e+00 -1.12150812e+00 -5.27937263e-02 1.30876720e+00
1.24342050e-02 -4.28573996e-01 -1.20770979e+00 -4.77056593e-01
-3.71140122e-01 5.26975214e-01 -5.96933626e-03 -3.29654105e-02
-9.38074887e-01 -6.94469690e-01 1.06167984e+00 1.00743806e+00
9.19383824e-01 -7.11830318e-01 -1.01224446e+00 1.71975046e-01
3.57958198e-01 -1.64779997e+00 -2.76316851e-01 3.83275032e-01
-1.58331442e+00 -6.20643675e-01 -1.81319132e-01 -9.70632553e-01
4.97170776e-01 1.89780667e-01 1.15472233e+00 1.42970324e-01
-1.94667503e-01 -4.62973386e-01 1.87736467e-01 -3.85235220e-01
-2.83967167e-01 3.81195068e-01 3.33174497e-01 -6.28028095e-01
4.20136452e-01 -9.41664159e-01 -4.06395614e-01 1.49896652e-01
-8.20636928e-01 4.98232037e-01 2.45607257e-01 5.75066328e-01
9.59588110e-01 1.76523834e-01 6.53728247e-02 -1.06387293e+00
2.87164748e-01 1.99693488e-03 -1.28723395e+00 2.35624209e-01
-4.43308085e-01 2.53290117e-01 1.15414488e+00 -6.99120581e-01
-8.63965809e-01 9.07789841e-02 -5.37263036e-01 -4.61742699e-01
-1.40252545e-01 2.66188562e-01 -7.99928665e-01 -9.91541520e-02
4.65528280e-01 -4.67245243e-02 -4.15944695e-01 -4.75587189e-01
5.52755855e-02 4.45134014e-01 8.72354567e-01 -4.59465921e-01
3.26184511e-01 4.32807297e-01 2.97433883e-01 -8.32576156e-01
-6.82604671e-01 2.62459844e-01 -1.65210277e-01 3.94667536e-02
7.14574933e-01 -1.04455912e+00 -7.42670298e-01 7.28047132e-01
-1.27058864e+00 -1.07302916e+00 6.29868433e-02 2.89417237e-01
-7.43584335e-02 -2.89615333e-01 -7.25826621e-01 -5.14780998e-01
-5.87897480e-01 -1.62285769e+00 9.18471575e-01 4.69705939e-01
-7.00028241e-02 -9.32964563e-01 -3.67605388e-01 2.61890382e-01
4.86779839e-01 -1.20046020e-01 1.10853755e+00 -3.86498839e-01
-1.01967621e+00 2.71224171e-01 -4.27347034e-01 1.54594466e-01
-3.15603167e-01 1.36357889e-01 -1.03786039e+00 3.55318114e-02
-1.92501158e-01 -3.00462209e-02 6.59153819e-01 4.43341851e-01
1.64700651e+00 -4.56099033e-01 -5.12822628e-01 1.54901230e+00
1.40084743e+00 3.92010272e-01 6.44219697e-01 3.90171111e-02
1.32331586e+00 -4.15355653e-01 -5.76523431e-02 2.20559373e-01
4.15853977e-01 7.33662486e-01 5.57559013e-01 7.47202486e-02
-2.47860670e-01 -2.37567350e-01 4.09761786e-01 1.06865036e+00
-3.60661256e-03 -1.37730896e-01 -1.13869774e+00 2.40956113e-01
-1.68786013e+00 -6.48530960e-01 -9.04879794e-02 1.76354969e+00
7.95441687e-01 5.58372438e-01 6.59310073e-02 1.21830486e-01
4.84284580e-01 6.52627796e-02 -1.13642335e+00 -5.31459451e-01
-2.14281678e-01 4.64741856e-01 9.19868886e-01 4.28633273e-01
-7.13735402e-01 1.33274281e+00 6.90437078e+00 1.07734609e+00
-1.34832859e+00 2.20911428e-01 7.97211051e-01 -3.79364491e-01
-3.38252336e-01 -2.52684355e-01 -1.61258507e+00 4.67147321e-01
1.41130435e+00 4.07758839e-02 8.47843587e-01 1.44554305e+00
-1.50369434e-02 2.45180372e-02 -1.16110551e+00 1.28207099e+00
-3.31178337e-01 -1.95442402e+00 1.84137821e-01 1.74872696e-01
6.59843504e-01 4.67051774e-01 3.49338800e-02 2.33446211e-01
3.81744385e-01 -1.00885582e+00 7.16098130e-01 1.17319465e-01
1.10963702e+00 -1.22137105e+00 5.49946308e-01 2.14664698e-01
-8.92383039e-01 1.19544275e-01 -4.68673766e-01 -2.99466938e-01
9.11727026e-02 7.25857079e-01 -6.05198920e-01 -3.57859731e-01
8.48285139e-01 3.90885532e-01 -4.72588211e-01 5.53163946e-01
7.08573556e-04 6.56842411e-01 -6.85142100e-01 -2.69178391e-01
-1.98558882e-01 1.16487689e-01 1.45506546e-01 1.05240166e+00
2.73969322e-01 4.62052226e-02 -2.74351209e-01 7.40588784e-01
-5.84020913e-01 -5.79444945e-01 -4.73904163e-01 3.44581380e-02
8.39471519e-01 1.01040781e+00 -1.17663550e+00 -3.76110524e-01
-1.98672354e-01 1.04135907e+00 3.67084593e-01 2.19897896e-01
-1.17152834e+00 -2.64039576e-01 1.29811418e+00 1.96841322e-02
1.61420360e-01 -6.66001081e-01 -9.29782212e-01 -1.16689074e+00
1.20468453e-01 -9.04312611e-01 -1.56344473e-01 -5.45333683e-01
-4.66680497e-01 7.25368202e-01 -1.81817338e-01 -5.39339304e-01
-2.21178848e-02 -1.10652530e+00 -3.12229455e-01 3.37062299e-01
-1.23611188e+00 -7.83945441e-01 -2.88046032e-01 7.13430882e-01
5.53761661e-01 -4.66158003e-01 8.97525966e-01 6.44765615e-01
-9.72349823e-01 9.49699819e-01 2.57215470e-01 2.54129469e-01
-1.81513578e-02 -8.35750520e-01 1.11206317e+00 1.07464147e+00
1.91308543e-01 6.63825035e-01 4.89815176e-01 -5.91580093e-01
-1.90122080e+00 -1.36099970e+00 6.39927745e-01 -5.84474131e-02
4.39653188e-01 -8.65509093e-01 -5.67976236e-01 7.43823647e-01
-2.81551093e-01 2.67505437e-01 5.26766121e-01 1.00236446e-01
-3.05691004e-01 -6.36376202e-01 -9.95747328e-01 1.29102552e+00
1.57805037e+00 -6.99168921e-01 2.01430485e-01 4.17960912e-01
8.78307045e-01 -1.07382178e+00 -4.18222457e-01 2.17178315e-01
5.02333641e-01 -8.72351289e-01 9.23662901e-01 -4.63953793e-01
4.37577933e-01 -8.68989304e-02 -3.57303470e-01 -5.75905263e-01
2.31865183e-01 -9.02736604e-01 -9.59883153e-01 9.35909331e-01
2.93732524e-01 -4.60867792e-01 1.33988631e+00 9.00434494e-01
-2.25390717e-01 -8.21360707e-01 -1.04495215e+00 -5.84063351e-01
-3.58484685e-01 -9.38552856e-01 8.10947061e-01 4.31565762e-01
-5.23000598e-01 2.93337494e-01 -5.12269437e-02 1.32923022e-01
7.07905650e-01 -2.54417032e-01 6.03141367e-01 -1.10546708e+00
-2.01467618e-01 -5.93459904e-01 -5.08543491e-01 -1.42648554e+00
4.38230634e-01 -7.15057671e-01 -2.07393721e-01 -1.20243657e+00
-9.17294398e-02 -5.65712810e-01 9.53391343e-02 4.96042997e-01
3.55671167e-01 2.33016893e-01 6.87449202e-02 -1.16595641e-01
-8.88633132e-01 2.06480864e-02 8.30339730e-01 8.74147490e-02
-1.29874751e-01 3.48625965e-02 -5.99758506e-01 1.08932877e+00
1.02157748e+00 -4.45076346e-01 -6.71001554e-01 -1.18454409e+00
5.51943123e-01 -6.73316121e-02 5.71432650e-01 -1.27734101e+00
4.77220953e-01 -1.47201180e-01 3.24142605e-01 -6.72125161e-01
2.76555866e-01 -7.35436797e-01 4.66857515e-02 5.64773858e-01
-7.88095146e-02 4.73130763e-01 6.43306017e-01 2.07924038e-01
8.22085962e-02 1.83386624e-01 8.50222766e-01 1.48939565e-01
-1.07454431e+00 5.33821881e-01 -5.54603398e-01 1.95147648e-01
6.93315983e-01 -4.12849158e-01 -4.42695558e-01 1.24800436e-01
-3.92879933e-01 -1.58113930e-02 4.92175400e-01 8.82026255e-02
7.09672391e-01 -1.26319015e+00 -7.08263740e-02 4.78060067e-01
-6.62863612e-01 6.97810054e-01 3.23053449e-01 4.23780859e-01
-1.26988542e+00 4.22599852e-01 -4.65229452e-01 -5.13563812e-01
-1.45758498e+00 2.63216823e-01 4.14636433e-01 -3.16824138e-01
-6.98611915e-01 1.17350113e+00 -1.51281208e-01 -2.00913638e-01
5.61771810e-01 -7.32191086e-01 5.51418245e-01 -5.05622268e-01
8.19393754e-01 6.16409302e-01 2.07539141e-01 -2.01140732e-01
-4.98115808e-01 1.21814050e-01 -2.17207253e-01 -1.13566607e-01
1.12785685e+00 9.27299485e-02 -1.68183923e-01 9.83100906e-02
1.16885555e+00 -3.18425357e-01 -1.41579926e+00 6.84354901e-02
-5.49486056e-02 1.17825225e-01 5.03283679e-01 -6.08974576e-01
-1.45534682e+00 1.00993872e+00 5.79208076e-01 -4.07228559e-01
1.38007855e+00 -3.26966822e-01 1.36055768e+00 9.00891483e-01
5.42098105e-01 -1.30506778e+00 -4.88522291e-01 9.99210119e-01
2.92934515e-02 -8.33595753e-01 -1.43748969e-01 -2.49938697e-01
-1.40907615e-01 1.10622370e+00 8.72878194e-01 1.52319055e-02
6.72245681e-01 1.19963515e+00 -2.03587059e-02 1.16280876e-01
-7.57261276e-01 4.34406698e-01 -2.33720064e-01 6.42784834e-01
-6.84133917e-02 1.43530682e-01 3.19906533e-01 5.50419927e-01
-4.75570321e-01 2.24958465e-01 1.88990727e-01 8.30357552e-01
-1.68938205e-01 -1.03571498e+00 1.33617604e-02 4.32522982e-01
-5.61802983e-01 -4.40496534e-01 -3.28068808e-02 6.11767948e-01
2.57362962e-01 5.08475363e-01 3.03729773e-01 -6.87154412e-01
-1.82130355e-02 -1.12291127e-01 6.08240187e-01 -3.01473886e-01
-5.40924907e-01 -1.58077702e-01 9.58809108e-02 -7.49027371e-01
-2.45751105e-02 -2.89035410e-01 -1.48641598e+00 -1.01347446e+00
6.16097786e-02 -5.22347569e-01 8.69090915e-01 1.04246807e+00
6.74165189e-01 6.55219972e-01 -1.15123674e-01 -9.61086810e-01
-1.66846886e-01 -2.60850757e-01 -4.06221986e-01 -1.54720336e-01
3.63982618e-01 -1.35853022e-01 8.72577354e-02 2.16080882e-02] | [8.46019172668457, 3.0226731300354004] |
ad871dd8-644a-4b9a-9925-e5dfc9d1d79b | learnable-front-ends-based-on-temporal | 2211.15254 | null | https://arxiv.org/abs/2211.15254v1 | https://arxiv.org/pdf/2211.15254v1.pdf | Learnable Front Ends Based on Temporal Modulation for Music Tagging | While end-to-end systems are becoming popular in auditory signal processing including automatic music tagging, models using raw audio as input needs a large amount of data and computational resources without domain knowledge. Inspired by the fact that temporal modulation is regarded as an essential component in auditory perception, we introduce the Temporal Modulation Neural Network (TMNN) that combines Mel-like data-driven front ends and temporal modulation filters with a simple ResNet back end. The structure includes a set of temporal modulation filters to capture long-term patterns in all frequency channels. Experimental results show that the proposed front ends surpass state-of-the-art (SOTA) methods on the MagnaTagATune dataset in automatic music tagging, and they are also helpful for keyword spotting on speech commands. Moreover, the model performance for each tag suggests that genre or instrument tags with complex rhythm and mood tags can especially be improved with temporal modulation. | ['Richard M. Stern', 'Yinghao Ma'] | 2022-11-28 | null | null | null | null | ['keyword-spotting'] | ['speech'] | [ 1.26437008e-01 -4.13760215e-01 -2.50631366e-02 -1.65639803e-01
-6.08534753e-01 -4.99235570e-01 3.04325223e-01 -3.64395455e-02
-6.58499777e-01 2.40494102e-01 4.66137111e-01 -7.90564045e-02
-4.40039515e-01 -3.87949586e-01 -2.01401010e-01 -4.48221654e-01
-3.38390052e-01 2.19056513e-02 2.37945825e-01 -4.69977558e-01
-1.08490465e-02 -3.64501439e-02 -1.96016634e+00 5.32321513e-01
4.99215454e-01 1.41357064e+00 4.60239679e-01 8.53160858e-01
-1.89111471e-01 7.52891719e-01 -6.67399406e-01 1.38669573e-02
-6.89092800e-02 -6.19226396e-01 -6.16193950e-01 -4.59774703e-01
2.15676695e-01 2.05639854e-01 -2.61288702e-01 1.01971471e+00
1.16162205e+00 4.74188060e-01 5.36434412e-01 -7.42911816e-01
-3.94625366e-01 1.55554366e+00 -2.58632824e-02 4.03646141e-01
3.81224096e-01 -1.23905502e-01 1.19649482e+00 -1.02428472e+00
3.34536359e-02 9.34414566e-01 1.09127927e+00 3.30769062e-01
-9.72961187e-01 -8.69642019e-01 5.02266875e-03 8.02775443e-01
-1.46346545e+00 -7.64687300e-01 1.09620595e+00 -5.41515708e-01
1.11615992e+00 4.59754080e-01 8.36938500e-01 9.23774540e-01
-4.97599728e-02 6.16016269e-01 8.49193454e-01 -6.07054532e-01
2.58263331e-02 -3.09435606e-01 7.59018734e-02 9.13080871e-02
-7.78991103e-01 2.81347662e-01 -1.18141615e+00 1.06175028e-01
5.42696416e-01 -3.39105934e-01 -2.52240747e-01 3.70512933e-01
-1.24780607e+00 5.72470129e-01 2.64737695e-01 7.39254177e-01
-4.55201149e-01 1.55346632e-01 6.67878449e-01 4.47125822e-01
3.84199113e-01 6.32703900e-01 -6.54773533e-01 -6.01852000e-01
-1.29058230e+00 3.26063558e-02 5.43320417e-01 5.00632882e-01
1.81457311e-01 7.17601895e-01 -3.46851170e-01 1.53821313e+00
-7.93944374e-02 2.62419641e-01 1.01968408e+00 -6.29568458e-01
3.35392356e-02 7.22723678e-02 -2.06208423e-01 -7.07569003e-01
-8.25693011e-01 -1.03599370e+00 -9.70718801e-01 -1.65022954e-01
3.84960204e-01 1.41957566e-01 -6.36401713e-01 1.84239733e+00
-1.00185931e-01 6.56233072e-01 -3.02219093e-01 9.70149040e-01
8.64759922e-01 6.71716869e-01 9.50559899e-02 -3.57836246e-01
1.72477567e+00 -8.15258384e-01 -1.11810637e+00 2.06587669e-02
1.85687035e-01 -1.09955990e+00 1.34039509e+00 8.40511203e-01
-1.14659750e+00 -1.34683084e+00 -1.01848090e+00 -8.99195969e-02
-4.08378243e-01 2.76836008e-01 5.56212485e-01 6.40136957e-01
-9.00083780e-01 8.81041169e-01 -4.90630597e-01 -1.33853704e-01
1.58783421e-02 3.47919226e-01 1.72983900e-01 7.40786493e-01
-1.58246791e+00 4.70229596e-01 6.64035141e-01 1.08221404e-01
-7.63989925e-01 -8.38633358e-01 -4.41792876e-01 1.76909447e-01
1.40402570e-01 -4.69158858e-01 1.56723106e+00 -7.04606354e-01
-1.96122122e+00 5.39168239e-01 -1.74737386e-02 -7.70680010e-01
6.75034849e-03 -4.24818635e-01 -1.01557112e+00 6.53615408e-03
-2.70436585e-01 6.04551256e-01 1.14930499e+00 -4.95259374e-01
-8.91241789e-01 -2.38956753e-02 -3.46834421e-01 2.31830165e-01
-8.51615608e-01 1.95869163e-01 -7.36218095e-02 -1.23052621e+00
1.92445740e-01 -6.66726589e-01 7.90423304e-02 -6.36458278e-01
-4.34848189e-01 -3.93099904e-01 4.88515824e-01 -7.80778229e-01
2.00506973e+00 -2.43946624e+00 -2.95643043e-03 6.96738511e-02
-2.08840743e-01 2.88499653e-01 -2.67163366e-01 2.17386201e-01
-2.08962396e-01 -1.54401049e-01 2.49703482e-01 -6.44961968e-02
3.64737719e-01 -2.88151354e-01 -6.06492162e-01 1.55256167e-02
-3.12288880e-01 7.38126993e-01 -7.50890791e-01 -1.44125298e-01
2.88781643e-01 5.38523436e-01 -6.12280488e-01 1.39971018e-01
-2.35981375e-01 4.77042377e-01 2.29657948e-01 3.59471381e-01
3.04775864e-01 1.38485402e-01 -1.60026876e-03 -2.90784031e-01
-4.09299463e-01 1.08172917e+00 -1.35356343e+00 2.04028130e+00
-5.71402788e-01 4.70829487e-01 -1.70476418e-02 -6.14613533e-01
8.51325274e-01 7.22831428e-01 6.03823423e-01 -6.73874974e-01
2.59900838e-01 2.74394631e-01 3.80020976e-01 -4.04216528e-01
6.21812761e-01 -2.02002183e-01 -1.77073136e-01 2.55855858e-01
3.49472940e-01 -1.53551966e-01 -1.38179185e-02 -4.69124138e-01
8.67064595e-01 -5.61332479e-02 1.43340901e-01 -2.79415008e-02
5.26014984e-01 -3.62568259e-01 6.24270022e-01 6.53120041e-01
1.60080776e-01 5.74015260e-01 -2.27364659e-01 -3.36586833e-01
-6.90156341e-01 -8.60055804e-01 -2.11170912e-01 1.93469095e+00
-4.01760876e-01 -7.80465722e-01 -6.96430266e-01 1.13611765e-01
-2.23047704e-01 4.05774415e-01 -2.55766332e-01 -1.51680127e-01
-4.46521789e-01 -5.03354490e-01 1.05431199e+00 5.64604163e-01
3.09022188e-01 -1.42671990e+00 -1.94785282e-01 7.58440256e-01
-5.09872198e-01 -8.05100322e-01 -6.60266936e-01 6.86505437e-01
-5.77488482e-01 -5.95066607e-01 -6.02857709e-01 -8.28411102e-01
-5.27351201e-01 -1.59881432e-02 1.02419841e+00 -3.75176847e-01
-9.07281414e-03 2.38082763e-02 -6.95554495e-01 -7.18395352e-01
-1.44014075e-01 5.43427229e-01 4.96305257e-01 1.11534715e-01
3.93777460e-01 -1.14571249e+00 -5.21527290e-01 2.96137810e-01
-7.21867740e-01 -6.92608301e-03 3.20548296e-01 8.25042367e-01
5.51034153e-01 3.72417957e-01 1.02909410e+00 -4.46507066e-01
7.87926435e-01 -5.25962226e-02 -3.12116176e-01 -3.53917211e-01
-4.74537700e-01 -3.56121093e-01 6.83648050e-01 -1.02059305e+00
-6.55624568e-01 6.54736087e-02 -7.09997714e-01 -4.01845813e-01
-9.54295769e-02 6.95968032e-01 -1.17437169e-01 3.01126480e-01
8.41683388e-01 3.07453781e-01 -3.69700253e-01 -1.09231174e+00
3.40012014e-01 9.59074140e-01 8.12970221e-01 -2.51750916e-01
6.48035109e-01 -9.53097362e-03 -2.89469123e-01 -9.41220105e-01
-1.03138757e+00 -6.57770336e-01 -5.34668863e-01 -3.24059367e-01
7.78310597e-01 -1.01237094e+00 -8.85832846e-01 5.82234561e-01
-9.60209548e-01 -1.71803445e-01 -5.83514571e-01 8.44559431e-01
-6.89794958e-01 -9.52880830e-02 -9.09929276e-01 -1.01110148e+00
-5.31808197e-01 -5.75485170e-01 6.44478738e-01 2.57138051e-02
-5.44071317e-01 -5.94383359e-01 1.19053066e-01 1.23498835e-01
8.02917898e-01 -5.55606127e-01 9.42688167e-01 -5.41247249e-01
-2.97589023e-02 1.24204025e-01 4.69135761e-01 4.79826391e-01
1.56981463e-03 -4.24778610e-01 -1.74365890e+00 3.55871543e-02
1.31438663e-02 -4.17916887e-02 9.16363358e-01 7.47500181e-01
1.43509734e+00 -1.61471739e-01 1.76646858e-01 7.56117463e-01
6.75875247e-01 4.92157072e-01 5.50726593e-01 1.34051472e-01
6.11744165e-01 3.18246186e-01 5.80156624e-01 5.95711529e-01
1.48419604e-01 9.42198575e-01 1.45236254e-01 -8.26173574e-02
-5.38515091e-01 -3.90253752e-01 5.03939927e-01 1.82717454e+00
-1.69699803e-01 -1.11061603e-01 -6.28755748e-01 5.92871070e-01
-1.72927165e+00 -1.09935927e+00 -7.09798560e-02 2.17586684e+00
1.22431636e+00 5.52019775e-02 5.47283471e-01 1.00914800e+00
6.03234589e-01 2.44515151e-01 -3.65936637e-01 -2.57451415e-01
-3.42112243e-01 8.10406804e-01 -1.01131216e-01 1.49367452e-01
-1.18263483e+00 1.04660153e+00 6.23075581e+00 1.52787745e+00
-1.24922597e+00 4.57632542e-01 2.90025584e-02 -3.48916888e-01
1.32171050e-01 -2.87934452e-01 -5.71582675e-01 3.98109466e-01
1.34394908e+00 1.75337762e-01 8.41130435e-01 5.65899014e-01
4.75946486e-01 3.63697529e-01 -9.27327991e-01 1.22867000e+00
-3.64427388e-01 -1.07385969e+00 -2.54951984e-01 -2.30634570e-01
3.56187820e-01 -5.25186956e-02 3.53602320e-01 5.79688668e-01
-2.00998619e-01 -9.90732074e-01 1.07467818e+00 5.73204994e-01
1.08321452e+00 -9.33879972e-01 6.23234510e-01 1.77194670e-01
-1.78459370e+00 -3.44853461e-01 -3.65847766e-01 -4.39833194e-01
6.42066970e-02 7.22158968e-01 -9.36944604e-01 2.69071519e-01
8.00766826e-01 7.55461097e-01 -3.96148056e-01 1.28983164e+00
-1.66198403e-01 1.16539991e+00 -3.07222992e-01 -2.14398745e-03
-2.60042995e-01 2.06210569e-01 6.35861099e-01 1.40717673e+00
5.08116543e-01 -2.13068932e-01 1.61076918e-01 3.43561828e-01
4.80319038e-02 4.25490499e-01 -1.88955709e-01 -3.19370747e-01
6.95214689e-01 1.19508123e+00 -5.32393813e-01 -6.21410049e-02
1.29860584e-02 6.72398329e-01 -2.14896291e-01 2.45964676e-01
-6.98102295e-01 -5.20315886e-01 8.81794214e-01 1.08804695e-01
3.39496642e-01 -1.32341161e-01 -2.62294888e-01 -7.30595291e-01
-3.06023687e-01 -8.27267110e-01 3.28264832e-01 -8.99690390e-01
-1.31345713e+00 9.94769812e-01 -5.88251948e-01 -1.91160059e+00
-4.43929315e-01 -5.55593193e-01 -1.85636446e-01 7.30287850e-01
-1.38614523e+00 -1.04937506e+00 2.60722935e-01 8.22945893e-01
6.30542219e-01 -2.72070140e-01 1.24147785e+00 8.18980634e-01
-9.20509994e-02 7.09598005e-01 -2.79349029e-01 1.59882009e-01
8.19301546e-01 -1.21548629e+00 1.11347377e-01 5.71571529e-01
8.37078273e-01 5.50994694e-01 4.18744147e-01 -2.51230866e-01
-9.64939415e-01 -1.14002848e+00 1.05213726e+00 -8.09453800e-02
7.94707894e-01 -5.35519183e-01 -7.25410581e-01 8.64386633e-02
3.03811640e-01 -4.01973248e-01 9.15142000e-01 6.63316071e-01
-3.36049885e-01 -3.56893599e-01 -3.53453219e-01 4.67664450e-01
1.27397680e+00 -1.09821546e+00 -6.57799721e-01 4.76409718e-02
7.71500707e-01 -2.89782673e-01 -8.41286302e-01 5.17117381e-01
7.61213124e-01 -7.24714756e-01 1.02851248e+00 -3.24441075e-01
-1.07943334e-01 -5.74297905e-01 -2.14978024e-01 -1.56267786e+00
-6.86608851e-01 -1.15752327e+00 -4.31657471e-02 1.38721919e+00
3.79342616e-01 -1.36891827e-01 2.47628614e-01 -4.65128183e-01
-5.25322676e-01 -1.84671521e-01 -1.05731547e+00 -9.42094624e-01
-3.73867720e-01 -1.33884335e+00 5.90651512e-01 1.07293069e+00
3.42961311e-01 7.45904684e-01 -7.23713040e-01 -9.48032141e-02
9.09091458e-02 9.28222761e-03 3.61576438e-01 -1.63139880e+00
-5.52409232e-01 -5.99894285e-01 -3.97124052e-01 -1.05392313e+00
-3.85471061e-03 -1.00556409e+00 1.74204364e-01 -9.38193321e-01
-2.68155843e-01 -3.78183246e-01 -1.09152901e+00 6.83506846e-01
5.80480322e-02 6.11924291e-01 2.77960420e-01 -5.00546135e-02
-5.48827946e-01 5.97187161e-01 9.35501158e-01 -1.87926829e-01
-5.55517137e-01 4.35506016e-01 -4.49480951e-01 1.00575435e+00
6.38626158e-01 -5.31888902e-01 -3.82169664e-01 -1.14614539e-01
5.21325529e-01 2.87032366e-01 3.05037424e-02 -1.60275817e+00
4.42230016e-01 2.01219693e-01 2.25026563e-01 -7.80905545e-01
8.86670828e-01 -6.91357076e-01 3.74754280e-01 1.15749545e-01
-5.15323520e-01 -5.98943643e-02 3.70753109e-01 2.04556301e-01
-5.95668495e-01 -2.33562570e-02 5.92711926e-01 1.28737912e-01
-5.77321708e-01 3.95358056e-02 -6.68748975e-01 -3.12262867e-02
-1.28823835e-02 -1.60792559e-01 3.98429297e-03 -4.66641247e-01
-1.21611083e+00 -4.56441998e-01 -3.19094896e-01 6.66543067e-01
4.45023477e-01 -1.61785889e+00 -6.42244220e-01 3.30016792e-01
-3.67186265e-03 -6.05221868e-01 6.25997066e-01 7.67760396e-01
2.24182874e-01 4.18524653e-01 -2.53392279e-01 -5.83778441e-01
-1.24090660e+00 2.83325732e-01 2.35974252e-01 -2.84095258e-02
-4.46138531e-01 1.28430569e+00 4.37944633e-04 -2.44134456e-01
6.83502138e-01 -6.23609364e-01 -6.94507837e-01 5.78750193e-01
6.81076527e-01 3.75365704e-01 3.16857547e-01 -5.49983263e-01
-2.75848091e-01 6.00490570e-01 3.16884607e-01 -5.47464490e-01
1.43819153e+00 -1.59941733e-01 2.65074428e-03 1.11210966e+00
6.16246819e-01 4.10185784e-01 -7.37525582e-01 -5.04342973e-01
1.44257829e-01 -5.81918545e-02 3.58705431e-01 -1.09617603e+00
-9.37924027e-01 1.09303677e+00 1.00148642e+00 5.93858063e-01
1.59397614e+00 -3.23863298e-01 9.80738282e-01 2.67216653e-01
3.27250123e-01 -1.42327678e+00 6.45972863e-02 9.80308414e-01
9.01891828e-01 -5.29769957e-01 -4.24725503e-01 -1.40556440e-01
-4.86231834e-01 1.02338552e+00 1.18183464e-01 1.20811276e-01
1.01060021e+00 3.94341707e-01 3.42396945e-01 8.83623138e-02
-7.79159546e-01 -5.92805445e-01 7.44809270e-01 4.08974946e-01
6.79625750e-01 2.23188683e-01 -1.71803817e-01 1.31970775e+00
-1.08401620e+00 -1.40052915e-01 7.84763545e-02 3.73822421e-01
-4.99348670e-01 -1.25256181e+00 -5.57194173e-01 3.34705323e-01
-8.40526521e-01 -5.21293998e-01 -1.80614695e-01 2.86059767e-01
5.15901566e-01 1.38970912e+00 -5.14319018e-02 -1.05898964e+00
3.87105167e-01 6.63083613e-01 3.00642431e-01 -6.52577937e-01
-1.15360236e+00 9.93329823e-01 -3.27060930e-03 -2.91982263e-01
-3.80406737e-01 -4.40805465e-01 -1.16372073e+00 2.71037310e-01
-4.69813496e-01 2.71922052e-01 7.11824298e-01 7.42792070e-01
3.47636163e-01 1.37152135e+00 5.49130142e-01 -9.79082346e-01
-3.01963359e-01 -1.87388432e+00 -9.90012348e-01 1.97087303e-01
3.11923087e-01 -5.21363437e-01 -9.32745412e-02 4.38820660e-01] | [15.688530921936035, 5.313606262207031] |
36fbaf60-0494-4625-b1fc-d89d7f27f95e | sarg-a-novel-semi-autoregressive-generator | 2008.01474 | null | https://arxiv.org/abs/2008.01474v3 | https://arxiv.org/pdf/2008.01474v3.pdf | SARG: A Novel Semi Autoregressive Generator for Multi-turn Incomplete Utterance Restoration | Dialogue systems in open domain have achieved great success due to the easily obtained single-turn corpus and the development of deep learning, but the multi-turn scenario is still a challenge because of the frequent coreference and information omission. In this paper, we investigate the incomplete utterance restoration which has brought general improvement over multi-turn dialogue systems in recent studies. Meanwhile, jointly inspired by the autoregression for text generation and the sequence labeling for text editing, we propose a novel semi autoregressive generator (SARG) with the high efficiency and flexibility. Moreover, experiments on two benchmarks show that our proposed model significantly outperforms the state-of-the-art models in terms of quality and inference speed. | ['Weidong Zhang', 'Feng Li', 'Mengzuo Huang', 'Wuhe Zou'] | 2020-08-04 | null | null | null | null | ['dialogue-rewriting'] | ['natural-language-processing'] | [ 1.12071030e-01 5.36494970e-01 2.68877447e-01 -3.58493328e-01
-1.08074188e+00 -1.59605622e-01 9.45828915e-01 -3.99122775e-01
-1.86833039e-01 1.23382127e+00 8.02427053e-01 -1.26679376e-01
7.93163031e-02 -4.82638448e-01 -3.95614266e-01 -6.23497128e-01
3.54540944e-01 8.49613607e-01 -1.65036526e-02 -8.55734825e-01
4.88085672e-02 -3.39543462e-01 -1.09293222e+00 2.58404851e-01
1.35191250e+00 6.32736981e-01 4.40234751e-01 6.17331505e-01
-4.48018342e-01 9.05368447e-01 -6.95114434e-01 -4.80271637e-01
-5.55937141e-02 -6.58738613e-01 -1.07373667e+00 6.02253564e-02
-1.68856859e-01 -4.92253333e-01 -4.15607601e-01 8.60752225e-01
8.87397289e-01 6.22308016e-01 4.86631244e-01 -9.80211735e-01
-6.97172642e-01 8.89116704e-01 -3.87712091e-01 -1.49907947e-01
7.16528893e-01 1.83044933e-02 1.04126894e+00 -8.07435274e-01
6.54478550e-01 1.62566090e+00 3.58614355e-01 7.27077425e-01
-7.85547554e-01 -3.84963393e-01 3.71007800e-01 2.56853253e-01
-7.98051715e-01 -7.04040408e-01 7.22134829e-01 -5.98204061e-02
1.20478952e+00 1.59180120e-01 1.28161967e-01 1.36231947e+00
7.55079985e-02 9.78666365e-01 9.40212786e-01 -5.34980834e-01
-1.81585222e-01 1.05318069e-01 9.78431031e-02 6.02442324e-01
-4.09809113e-01 -1.97006419e-01 -2.16143027e-01 1.57697335e-01
5.78183174e-01 -3.88018548e-01 -2.21936554e-01 1.57871485e-01
-1.15228903e+00 9.51861441e-01 -4.16897237e-02 1.39137298e-01
-5.69459915e-01 -5.47803164e-01 7.55886197e-01 3.91418070e-01
5.52924871e-01 3.62880379e-01 -3.49540979e-01 -5.86184144e-01
-5.21981061e-01 4.27203268e-01 1.16860259e+00 1.20276618e+00
1.91540152e-01 3.72348726e-01 -5.26243985e-01 1.36041558e+00
2.08705902e-01 3.19678456e-01 8.11946809e-01 -6.94749892e-01
7.92730689e-01 5.20077288e-01 1.31307438e-01 -8.53460670e-01
-5.90172946e-01 -1.31199449e-01 -1.35142231e+00 -4.05673981e-01
3.72067869e-01 -7.27115750e-01 -4.15040821e-01 1.69086576e+00
3.28029007e-01 -2.03100033e-02 6.38992548e-01 8.62027466e-01
1.14571488e+00 9.05630946e-01 -1.01631790e-01 -6.25845969e-01
1.33146822e+00 -1.46468961e+00 -1.23961878e+00 -1.86060339e-01
5.67898929e-01 -8.28847408e-01 9.55917358e-01 5.32473743e-01
-1.02636778e+00 -6.62259459e-01 -8.65404069e-01 -3.02215040e-01
-1.03619300e-01 2.79547215e-01 6.17718637e-01 4.04776067e-01
-8.03366005e-01 1.46527678e-01 -4.43499535e-01 -4.11430687e-01
-1.49470568e-01 9.22622755e-02 -3.81459266e-01 1.53754801e-01
-1.64331317e+00 1.07298863e+00 4.80628163e-01 3.24808270e-01
-6.41370714e-01 -1.96365312e-01 -8.04925919e-01 7.87411407e-02
6.73434913e-01 -5.82908988e-01 1.68612778e+00 -5.51777542e-01
-2.39040208e+00 4.10763264e-01 -1.21957667e-01 -8.25264871e-01
6.69670939e-01 -6.22859180e-01 -3.76957417e-01 -1.39887542e-01
-8.81443694e-02 4.49821144e-01 5.04739165e-01 -8.40453625e-01
-6.70746505e-01 -3.57621610e-01 2.66111642e-01 6.88760221e-01
-1.65759668e-01 1.02229774e-01 -1.75614879e-01 -4.25812572e-01
-1.12052672e-01 -9.71320152e-01 -4.08916026e-01 -1.10147011e+00
-6.65829360e-01 -6.44094884e-01 6.55553341e-01 -1.09452891e+00
1.39738810e+00 -1.62569666e+00 4.86649007e-01 -6.45949483e-01
-1.58209249e-01 5.27138650e-01 1.43195866e-02 7.94019997e-01
2.24198159e-02 -4.72400896e-02 -2.31994361e-01 -4.98196632e-01
5.40008880e-02 9.42876711e-02 -5.75895131e-01 5.43200821e-02
5.76898977e-02 6.41494334e-01 -9.10773158e-01 -5.71606755e-01
2.87110865e-01 1.61469206e-01 -3.52542013e-01 7.19764590e-01
-6.25784993e-01 6.46078169e-01 -5.35720170e-01 1.33004338e-01
5.57421088e-01 1.29334247e-02 3.72506112e-01 8.16792101e-02
-1.38521761e-01 5.77614248e-01 -1.12465489e+00 1.96460354e+00
-6.65591419e-01 2.84566581e-01 9.69626307e-02 -9.50227141e-01
1.29096353e+00 7.54505157e-01 4.96330187e-02 -6.72306478e-01
2.22318813e-01 1.85631085e-02 1.44070655e-01 -6.96361721e-01
1.05752587e+00 -1.23133384e-01 -2.73518950e-01 6.16842568e-01
2.12695986e-01 -2.50667959e-01 1.88633576e-01 4.43376988e-01
5.81381023e-01 4.31640923e-01 5.16359031e-01 -6.17019599e-04
8.85407150e-01 -2.75167733e-01 6.71903431e-01 5.96956432e-01
-1.31266471e-02 4.76191372e-01 6.22415245e-01 -1.24361843e-01
-8.67323399e-01 -4.99065071e-01 1.23015210e-01 1.30329287e+00
-1.42785951e-01 -3.53808969e-01 -9.75846887e-01 -6.72704041e-01
-6.01185799e-01 9.63777840e-01 -2.42216080e-01 6.55299500e-02
-7.01022387e-01 -7.66357303e-01 7.42573202e-01 3.48620176e-01
7.24740684e-01 -1.28640306e+00 -1.34768948e-01 5.16345561e-01
-7.65194893e-01 -1.31515872e+00 -4.05687451e-01 -3.09457809e-01
-5.12646556e-01 -9.02577877e-01 -7.05607355e-01 -7.46335328e-01
1.70351490e-01 2.14647651e-01 8.73825312e-01 -2.42788643e-01
2.59353787e-01 -8.54588374e-02 -7.64117002e-01 -4.51855153e-01
-8.02215457e-01 4.99206930e-01 -1.85803678e-02 3.70012224e-02
8.02338347e-02 -3.73881698e-01 -6.63354918e-02 1.38959110e-01
-6.64740443e-01 3.95307213e-01 5.32092929e-01 1.19671810e+00
8.01251307e-02 -2.91846752e-01 1.08895743e+00 -1.17218912e+00
1.27303529e+00 -5.18255413e-01 -3.38411480e-01 3.59116077e-01
-5.80268025e-01 1.89746380e-01 8.15693438e-01 5.11223860e-02
-2.11862755e+00 -3.00084263e-01 -4.97515380e-01 1.95647970e-01
-2.44333625e-01 5.01888514e-01 -3.98679078e-01 6.30783856e-01
4.30938244e-01 3.74131173e-01 1.53067291e-01 -7.11945891e-01
5.65831840e-01 1.09595537e+00 5.12985647e-01 -6.35375440e-01
2.18175009e-01 4.14006598e-02 -3.39284450e-01 -1.06256378e+00
-8.08348000e-01 -2.20029935e-01 -6.40170813e-01 8.72131884e-02
8.29232633e-01 -8.87954354e-01 -6.31088912e-01 7.66802430e-01
-1.68956959e+00 -1.12193309e-01 2.34629169e-01 3.86627465e-01
-7.83240974e-01 7.66153872e-01 -7.56211936e-01 -1.11304188e+00
-7.92033613e-01 -1.17324328e+00 7.48031378e-01 3.49747807e-01
-1.76998034e-01 -9.52993989e-01 1.79468051e-01 5.38432837e-01
3.37369353e-01 -6.87941164e-02 7.61173844e-01 -1.13782167e+00
-3.24629754e-01 -9.00847930e-03 -6.53393865e-02 3.98916483e-01
-3.56499553e-02 -1.48463324e-01 -8.35895121e-01 -9.69113186e-02
1.67694181e-01 -5.71527779e-01 5.44633567e-01 2.05319747e-01
6.47188842e-01 -4.52655852e-01 -4.24023345e-02 3.68585587e-01
8.32865357e-01 3.03183377e-01 7.05235124e-01 1.27702802e-01
5.61637223e-01 8.77877474e-01 1.06026065e+00 6.55520976e-01
6.49273217e-01 6.65307164e-01 2.23595724e-01 2.49822214e-01
7.34107941e-02 -4.67495829e-01 2.84377426e-01 1.35895252e+00
-1.12086885e-01 -4.57936466e-01 -6.12477303e-01 4.85012978e-01
-2.29688120e+00 -9.36579466e-01 -4.39704478e-01 1.92418480e+00
9.41699028e-01 1.32694483e-01 2.68478654e-02 -3.31039041e-01
7.83699453e-01 4.98293400e-01 -2.64310688e-01 -6.35939240e-01
-3.00143749e-01 -1.43735126e-01 -8.61538872e-02 6.90155327e-01
-8.84816945e-01 1.28723359e+00 6.00669384e+00 8.71044695e-01
-6.04315579e-01 7.27304220e-02 5.71604729e-01 1.72621533e-01
5.38831986e-02 -8.23040828e-02 -1.19789112e+00 3.95090252e-01
1.07386196e+00 -3.27013552e-01 3.72972041e-01 6.28900826e-01
3.18872750e-01 1.38670653e-02 -8.28171670e-01 7.22656906e-01
2.64357775e-01 -9.95500386e-01 9.29205492e-02 -2.73317516e-01
7.92868614e-01 -2.19175875e-01 -4.22288775e-01 9.37900305e-01
6.40910387e-01 -9.23876286e-01 1.52012989e-01 7.06193447e-01
5.06815314e-01 -8.01547587e-01 1.03563476e+00 7.64757872e-01
-7.85569012e-01 -7.58619756e-02 -4.49815661e-01 -3.69122922e-01
4.46994305e-01 2.26121575e-01 -1.13713753e+00 9.42973316e-01
9.55148414e-02 4.95601207e-01 -9.86029357e-02 5.49688637e-01
-5.76924026e-01 4.83344585e-01 -9.16878507e-03 -2.04194024e-01
3.11775357e-01 -5.40311456e-01 7.80759275e-01 1.23623669e+00
6.04287162e-02 2.51297295e-01 3.24355543e-01 6.00467563e-01
-2.00906038e-01 4.49543148e-01 -5.32890975e-01 5.52366264e-02
6.04610562e-01 1.32810998e+00 -4.20978479e-02 -3.22460324e-01
-4.03332621e-01 8.71190190e-01 3.95760715e-01 2.87138641e-01
-6.85906947e-01 -5.41045904e-01 3.91778618e-01 -4.47124958e-01
5.36205072e-04 -9.50893983e-02 -4.87852469e-02 -1.18128026e+00
3.83022577e-02 -1.19189060e+00 2.49479398e-01 -5.66511989e-01
-1.05451477e+00 9.16549683e-01 -2.03883737e-01 -1.03754234e+00
-1.01488078e+00 -1.46316305e-01 -5.84625006e-01 8.94777656e-01
-1.22998488e+00 -1.22516692e+00 -7.65874162e-02 3.88691127e-01
1.22175944e+00 -4.11778867e-01 1.12118518e+00 5.16560003e-02
-8.04363668e-01 5.94690025e-01 3.32651943e-01 3.10114592e-01
1.00198829e+00 -1.18424571e+00 5.21582425e-01 6.64218366e-01
-1.83100834e-01 4.07112330e-01 8.19614410e-01 -4.70621824e-01
-1.37159050e+00 -7.33256996e-01 9.30025697e-01 -1.66479975e-01
5.46106160e-01 -2.86280483e-01 -9.64362562e-01 8.38898957e-01
7.31781006e-01 -7.70359159e-01 3.32353204e-01 4.16330516e-01
1.08554013e-01 7.55816028e-02 -8.64023566e-01 6.99167550e-01
8.74470830e-01 -3.05779099e-01 -1.01375687e+00 4.71570373e-01
1.15305233e+00 -8.69074821e-01 -7.82527983e-01 2.27442190e-01
4.21922803e-01 -1.01987243e+00 5.80705345e-01 -8.67648065e-01
6.06545031e-01 2.27753386e-01 -1.39192408e-02 -1.71316481e+00
-1.24867685e-01 -1.05474412e+00 -3.62928510e-02 1.74385607e+00
3.82684052e-01 -4.50082719e-01 3.96826267e-01 5.80575168e-01
-5.27903736e-01 -5.30903399e-01 -8.65292370e-01 -3.46376628e-01
6.82546794e-02 -1.35748899e-02 6.54016197e-01 7.50997841e-01
5.09256124e-01 1.19030523e+00 -1.08043063e+00 -3.28465700e-01
1.49574459e-01 1.36177793e-01 1.12123048e+00 -1.09710193e+00
-5.04182220e-01 -1.43048912e-01 2.01332331e-01 -1.60876775e+00
4.63628709e-01 -4.99440104e-01 4.34151232e-01 -1.60306180e+00
-2.61169914e-02 -4.37393710e-02 2.53192127e-01 -5.47063760e-02
-4.68983233e-01 -5.34667850e-01 2.57785738e-01 -9.44964513e-02
-8.75065744e-01 1.09312761e+00 1.34951437e+00 9.69678983e-02
-4.60664809e-01 2.52892196e-01 -8.36972713e-01 7.43236244e-01
6.31841063e-01 -2.70104483e-02 -3.41092974e-01 -4.06148314e-01
-4.98106405e-02 9.60438848e-01 -3.69568825e-01 -7.04096556e-01
3.98470998e-01 4.88258041e-02 -2.76153713e-01 -7.90395498e-01
5.33735454e-01 -3.47600877e-01 -3.18208367e-01 3.74222808e-02
-6.93028927e-01 -5.51323742e-02 -1.61097869e-02 7.51037657e-01
-4.15670246e-01 -3.80039871e-01 5.71914613e-01 -3.04990977e-01
-6.29284799e-01 2.41630208e-02 -3.60370696e-01 3.59569401e-01
6.29977703e-01 2.48671934e-01 -3.39141846e-01 -7.71266580e-01
-5.94784558e-01 5.12866199e-01 -7.44553283e-02 7.87852526e-01
3.97468567e-01 -8.90440941e-01 -1.19982207e+00 2.59247292e-02
-1.98707789e-01 3.27909172e-01 6.51505709e-01 8.28965187e-01
-5.06341644e-02 8.28887939e-01 -1.67397171e-01 -3.04815769e-01
-1.12399995e+00 1.65485114e-01 1.51668519e-01 -9.04902220e-01
-5.92041910e-01 6.49504840e-01 1.89155802e-01 -8.91044080e-01
3.14346194e-01 5.38096763e-02 -5.49421847e-01 1.33472234e-01
6.11564934e-01 5.41361868e-01 7.89825991e-02 -6.77340508e-01
2.23206282e-01 -7.03143477e-02 -5.56575239e-01 -1.20244406e-01
1.25377142e+00 -4.66768861e-01 -6.77537769e-02 4.75386053e-01
7.79636085e-01 -1.42288595e-01 -1.17864275e+00 -4.79987979e-01
1.88068934e-02 -1.02535218e-01 -2.65323490e-01 -8.72921228e-01
-5.01269102e-01 1.01961720e+00 -3.47691961e-02 3.31679344e-01
6.65563226e-01 -4.09187943e-01 1.06585407e+00 7.26035357e-01
3.60631049e-01 -1.30096352e+00 -9.48424712e-02 1.16030800e+00
1.11025369e+00 -1.35273039e+00 -1.91966653e-01 -3.44452560e-01
-1.23052096e+00 9.61121500e-01 8.67898643e-01 -4.26769219e-02
2.89707422e-01 -1.78579930e-02 2.61673834e-02 1.70559958e-01
-1.16822290e+00 1.37699291e-01 -1.84576094e-01 6.71513498e-01
8.14914823e-01 1.29818365e-01 -6.90762639e-01 1.05329978e+00
-4.22343671e-01 7.33427564e-03 7.86315978e-01 3.43846560e-01
-4.51807112e-01 -1.18574965e+00 -2.15075761e-01 4.25329268e-01
-3.84225339e-01 -1.16372548e-01 -4.83408362e-01 6.63286388e-01
-5.79417527e-01 1.38412130e+00 -2.32985556e-01 -2.29687229e-01
5.16971707e-01 3.99826378e-01 2.04253808e-01 -6.87035680e-01
-6.35137558e-01 1.68492153e-01 7.39665389e-01 -1.32258773e-01
-2.85423279e-01 -6.34307623e-01 -1.41405952e+00 -5.76448701e-02
-5.06792903e-01 5.28677702e-01 5.37408948e-01 1.12718332e+00
4.08135265e-01 7.25082099e-01 7.60680616e-01 -6.11938238e-01
-1.21786809e+00 -1.91468596e+00 -5.53767920e-01 2.96798497e-01
1.98307633e-03 -5.52082539e-01 -3.47263627e-02 -1.26857340e-01] | [12.569392204284668, 8.239862442016602] |
e3561950-72cb-4745-854b-34e97898a344 | side-window-filtering | 1905.07177 | null | https://arxiv.org/abs/1905.07177v1 | https://arxiv.org/pdf/1905.07177v1.pdf | Side Window Filtering | Local windows are routinely used in computer vision and almost without exception the center of the window is aligned with the pixels being processed. We show that this conventional wisdom is not universally applicable. When a pixel is on an edge, placing the center of the window on the pixel is one of the fundamental reasons that cause many filtering algorithms to blur the edges. Based on this insight, we propose a new Side Window Filtering (SWF) technique which aligns the window's side or corner with the pixel being processed. The SWF technique is surprisingly simple yet theoretically rooted and very effective in practice. We show that many traditional linear and nonlinear filters can be easily implemented under the SWF framework. Extensive analysis and experiments show that implementing the SWF principle can significantly improve their edge preserving capabilities and achieve state of the art performances in applications such as image smoothing, denoising, enhancement, structure-preserving texture-removing, mutual-structure extraction, and HDR tone mapping. In addition to image filtering, we further show that the SWF principle can be extended to other applications involving the use of a local window. Using colorization by optimization as an example, we demonstrate that implementing the SWF principle can effectively prevent artifacts such as color leakage associated with the conventional implementation. Given the ubiquity of window based operations in computer vision, the new SWF technique is likely to benefit many more applications. | ['Guoping Qiu', 'Yuanhao Gong', 'Hui Yin'] | 2019-05-17 | side-window-filtering-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Yin_Side_Window_Filtering_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Yin_Side_Window_Filtering_CVPR_2019_paper.pdf | cvpr-2019-6 | ['image-smoothing', 'tone-mapping', 'point-interactive-image-colorization'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 6.82688177e-01 -3.26336831e-01 1.78292140e-01 -6.40794560e-02
-6.91023283e-03 -2.08158493e-01 3.81900936e-01 -1.00139752e-01
-4.15195495e-01 4.48425114e-01 5.76693937e-03 -3.34610283e-01
8.32067430e-02 -8.21327865e-01 -5.52911222e-01 -9.60816324e-01
3.66226025e-02 -6.63489759e-01 6.90303087e-01 -4.26934749e-01
5.06666601e-01 5.87036014e-01 -1.63386261e+00 2.95812905e-01
9.30587828e-01 1.05887198e+00 3.70918453e-01 7.25064278e-01
-1.75317675e-01 2.72868812e-01 -5.34854949e-01 -1.36627987e-01
5.53487301e-01 -5.94269037e-01 -4.44779128e-01 2.31310070e-01
5.53878248e-01 -4.01443094e-01 -2.26388332e-02 1.27760184e+00
3.72815847e-01 2.69375950e-01 2.55947381e-01 -6.97602570e-01
-6.53596759e-01 3.37273069e-02 -9.38134670e-01 -1.08464556e-02
7.52300024e-02 -1.88122436e-01 6.69233561e-01 -9.38949883e-01
5.43878198e-01 1.03536201e+00 9.56726611e-01 3.25653344e-01
-1.29926765e+00 -2.69408017e-01 1.83673818e-02 2.74700075e-01
-1.23152399e+00 -4.16544110e-01 9.12277699e-01 -7.78667629e-02
5.35762668e-01 6.67710245e-01 6.04888260e-01 7.28180707e-02
5.40777862e-01 4.23359096e-01 1.46184921e+00 -1.05839264e+00
2.47265380e-02 9.90127400e-02 2.94829994e-01 6.88350618e-01
2.23491877e-01 2.06187800e-01 -5.36423624e-01 6.46795109e-02
9.25873339e-01 -3.90798077e-02 -6.81154966e-01 -2.38006487e-01
-1.08771348e+00 6.66393101e-01 3.36550713e-01 2.89328992e-01
-5.18022887e-02 3.41294929e-02 1.36737734e-01 3.52951169e-01
6.45204365e-01 2.83003837e-01 -1.93414301e-01 3.83036405e-01
-1.09288359e+00 1.02635682e-01 5.34630239e-01 5.32447875e-01
8.76336455e-01 -1.15248986e-01 -7.69427866e-02 9.09779608e-01
1.19478658e-01 4.72394496e-01 1.64764330e-01 -1.09809387e+00
-1.77271754e-01 4.31188822e-01 2.13431790e-01 -1.05975389e+00
-2.46981785e-01 -2.15457097e-01 -7.48254120e-01 1.01737118e+00
4.76591259e-01 8.39615464e-02 -8.08124006e-01 1.33422518e+00
3.82546008e-01 2.93864816e-01 -1.99571416e-01 7.56996989e-01
4.89420384e-01 6.65310085e-01 -3.68703902e-01 -5.04082978e-01
1.51295686e+00 -6.81817412e-01 -9.27807093e-01 -4.16253433e-02
5.34359850e-02 -1.35966623e+00 1.15448081e+00 5.27099073e-01
-9.65357840e-01 -5.90738595e-01 -1.20395708e+00 -3.61566097e-01
-4.36080933e-01 1.82929128e-01 5.68817794e-01 9.28339005e-01
-1.12044501e+00 7.82248259e-01 -6.08041465e-01 -4.76767033e-01
7.86174536e-02 1.23664334e-01 -2.29251489e-01 -7.38760158e-02
-7.33735204e-01 9.53610837e-01 -1.70393288e-03 3.32242668e-01
1.49552986e-01 -6.89155996e-01 -6.01153970e-01 2.70106867e-02
3.85513306e-01 -4.78877217e-01 7.48444676e-01 -9.69979763e-01
-1.72662163e+00 9.28644836e-01 -5.47300577e-01 -4.18133259e-01
5.58639526e-01 -3.46840262e-01 -4.39728558e-01 1.01603217e-01
-3.29647034e-01 1.62878036e-01 1.46031725e+00 -1.19775152e+00
-5.84305048e-01 -2.31094852e-01 -3.34061950e-01 1.17679592e-02
-4.55518961e-01 2.03279540e-01 -4.07762170e-01 -8.84120762e-01
3.34648371e-01 -6.60259783e-01 1.70278344e-02 5.21645248e-01
-3.16446841e-01 2.04529494e-01 1.32408941e+00 -7.56873488e-01
1.52094758e+00 -2.48912621e+00 -2.01093316e-01 3.57750207e-01
6.26069754e-02 4.83733594e-01 -4.16925214e-02 4.37196434e-01
-1.70907065e-01 -2.75348812e-01 -3.95000398e-01 -1.18484251e-01
-2.48496711e-01 2.69337222e-02 -3.83298308e-01 6.74582064e-01
4.92674746e-02 5.72330773e-01 -3.78233701e-01 -3.06075454e-01
5.23631275e-01 8.10633123e-01 -4.08622980e-01 -1.75205663e-01
3.26924890e-01 -1.31302029e-02 5.04858978e-02 4.00120974e-01
1.16058135e+00 1.00356042e-01 -3.84967066e-02 -5.21688163e-01
-6.31052256e-01 -1.77780300e-01 -1.59384751e+00 1.10969543e+00
-3.02808136e-01 9.94637191e-01 4.46956247e-01 -8.23286176e-01
9.11131024e-01 4.03569415e-02 2.35342577e-01 -6.87701106e-01
-2.26053447e-01 2.40047082e-01 -2.10425079e-01 -3.59696865e-01
6.34616077e-01 -2.04707816e-01 5.59837282e-01 1.74871787e-01
-4.88184512e-01 -1.87380522e-01 1.38817444e-01 -5.32353446e-02
8.71898115e-01 -6.51553739e-03 5.53331375e-01 -6.70813799e-01
8.29114437e-01 -1.93966702e-01 4.73400116e-01 7.47079372e-01
-3.43192965e-01 9.03091490e-01 1.50174648e-01 -4.37290341e-01
-9.38952088e-01 -1.07411957e+00 -4.76594955e-01 1.04947984e+00
3.63829851e-01 -3.08768123e-01 -9.76195633e-01 -1.54161215e-01
-1.00071222e-01 3.36825430e-01 -5.98528862e-01 1.20971255e-01
-6.31090820e-01 -7.23911345e-01 1.20432917e-02 1.75195798e-01
9.15134490e-01 -7.71415293e-01 -9.61210608e-01 1.95105761e-01
5.42252064e-02 -9.43567395e-01 -7.77591884e-01 7.07935989e-02
-9.74104345e-01 -9.55749750e-01 -8.47910404e-01 -7.04141736e-01
8.54122281e-01 8.83833349e-01 8.54415536e-01 4.31654334e-01
-5.00816047e-01 3.00302446e-01 -3.52259725e-01 -2.91502148e-01
-1.38366342e-01 -5.96079767e-01 -4.15415436e-01 3.87680203e-01
9.95329991e-02 -4.23344910e-01 -8.79021347e-01 4.43803519e-01
-1.08644748e+00 1.57841355e-01 3.04618359e-01 9.46040869e-01
3.68542731e-01 4.87356961e-01 1.14656650e-01 -9.55582500e-01
5.58157563e-01 5.07364213e-01 -7.57979691e-01 1.95520550e-01
-5.81172228e-01 -5.98518215e-02 4.94151741e-01 -3.81973326e-01
-1.32397485e+00 2.63020433e-02 5.86856194e-02 1.53032482e-01
-6.21384494e-02 9.06309858e-02 -5.90344779e-02 -5.27338326e-01
5.66662192e-01 1.66850924e-01 2.56208122e-01 -4.99087900e-01
2.64950722e-01 3.84054035e-01 6.98837578e-01 -1.17695682e-01
9.00147855e-01 1.09757614e+00 3.88653010e-01 -1.38728654e+00
-3.49947274e-01 -5.06010115e-01 -3.45383584e-01 -2.82681465e-01
7.99771845e-01 -5.65273762e-01 -7.39076912e-01 7.70428061e-01
-1.04530990e+00 -2.71972001e-01 -3.03591847e-01 2.11206183e-01
-3.24346006e-01 6.44211173e-01 -5.16033411e-01 -8.72027457e-01
-2.45594069e-01 -1.03062630e+00 6.96825206e-01 5.37082493e-01
3.87271345e-02 -1.08179545e+00 -2.77434021e-01 -2.66893879e-02
8.02581728e-01 7.95796141e-02 9.23903823e-01 3.09917182e-01
-5.04384696e-01 -1.40552551e-01 -3.34843159e-01 5.94432056e-01
3.06205004e-01 3.45858991e-01 -1.04137671e+00 -1.98495328e-01
3.11007351e-01 5.70923507e-01 1.13897955e+00 8.25681567e-01
7.76266038e-01 5.48870489e-02 -1.63899913e-01 7.67564654e-01
1.86972678e+00 1.27206698e-01 1.28987706e+00 5.10376632e-01
2.41344079e-01 6.47948146e-01 4.56539154e-01 2.46787384e-01
-2.47263685e-01 8.81674290e-01 1.77898675e-01 -6.44681931e-01
-5.63531458e-01 3.56955737e-01 5.32851636e-01 3.61875862e-01
-3.21175158e-01 1.81617066e-01 -3.21608841e-01 2.04960555e-01
-1.51349449e+00 -1.12406278e+00 -5.03542900e-01 2.50077462e+00
7.28459001e-01 -1.87624618e-02 -1.95228264e-01 4.39546555e-01
9.04715061e-01 2.23856702e-01 -1.35382339e-01 -6.90443516e-01
-4.48023826e-01 5.29093504e-01 7.78559804e-01 8.36254537e-01
-1.19404125e+00 6.22776628e-01 6.55029869e+00 9.52157557e-01
-1.31163931e+00 -2.54178653e-03 5.32612026e-01 2.93728977e-01
-8.06584582e-02 2.92270742e-02 -6.20060980e-01 3.64743024e-01
2.14090213e-01 1.38215408e-01 5.37313879e-01 3.39912951e-01
5.88843584e-01 -7.48336375e-01 -6.54535651e-01 1.04386640e+00
-1.55269699e-02 -1.21088970e+00 -3.10515910e-01 2.72049997e-02
7.04795301e-01 -5.67992508e-01 3.11569065e-01 -4.86149579e-01
-4.45484817e-01 -8.10613930e-01 5.18824518e-01 3.74750882e-01
5.95540762e-01 -5.68678200e-01 4.78958130e-01 -1.25598079e-02
-1.31030977e+00 1.60983671e-02 -3.78626853e-01 -2.19674781e-01
7.79923648e-02 1.11520481e+00 -2.94797987e-01 4.44711119e-01
8.53267908e-01 4.71664488e-01 -4.91518110e-01 1.22754121e+00
-2.13466257e-01 3.32793146e-01 -3.73375565e-01 3.41502160e-01
-1.32971004e-01 -7.13303447e-01 6.77117288e-01 1.24132204e+00
3.63306314e-01 -8.03898349e-02 -2.89061338e-01 7.25382268e-01
2.89270192e-01 1.38103887e-01 -3.77228290e-01 4.32702869e-01
1.68012798e-01 1.24311960e+00 -1.16617095e+00 -2.82616705e-01
-8.14953029e-01 1.27867889e+00 -3.99478614e-01 6.41853631e-01
-5.64981937e-01 -8.63537192e-01 7.71336198e-01 1.95138156e-01
5.62773585e-01 -2.21497923e-01 -8.12560976e-01 -8.33833337e-01
2.22742707e-01 -7.84963429e-01 -1.27291698e-02 -7.06268907e-01
-1.04695320e+00 2.69462496e-01 -4.02564883e-01 -1.28830993e+00
4.94677782e-01 -8.67924035e-01 -9.61437225e-01 1.00863862e+00
-1.78739250e+00 -8.09952617e-01 -3.24856848e-01 7.99590290e-01
3.63912255e-01 2.98656315e-01 6.14918053e-01 3.48502040e-01
-3.18686306e-01 3.00772071e-01 3.57326895e-01 -7.05903694e-02
9.16815281e-01 -1.15742421e+00 1.52127773e-01 1.37218118e+00
1.33761624e-02 8.19934607e-01 1.05062830e+00 -4.07166898e-01
-1.35817254e+00 -5.56517780e-01 7.80570388e-01 1.74442992e-01
3.57819766e-01 -2.18411043e-01 -1.11881030e+00 2.10416228e-01
4.47042525e-01 -1.46833928e-02 3.47057641e-01 -1.65193453e-01
-2.49397084e-01 -4.34808135e-01 -1.10803962e+00 6.62888527e-01
7.05672145e-01 -4.71245557e-01 -4.94011611e-01 3.52737606e-02
2.33968630e-01 -3.32481176e-01 -5.04438102e-01 2.97657758e-01
7.51847327e-01 -1.56082404e+00 1.16761196e+00 1.44062251e-01
2.22048402e-01 -6.37148440e-01 -5.74460812e-02 -1.09226084e+00
-2.45389804e-01 -1.06705821e+00 2.31984049e-01 1.14594042e+00
1.96390674e-01 -9.77977872e-01 5.56826174e-01 5.23686469e-01
-1.17388010e-01 -4.09786224e-01 -9.80467677e-01 -6.89752221e-01
-2.34961495e-01 -2.20620856e-01 1.92194462e-01 7.35822320e-01
-1.62937492e-01 -1.69030368e-01 -4.06157047e-01 1.89129099e-01
8.39937866e-01 2.73375213e-01 7.01373339e-01 -1.10186195e+00
-2.70830244e-01 -5.87382853e-01 -2.74052799e-01 -1.16563559e+00
-5.08834720e-01 -2.26283818e-01 1.87128663e-01 -1.43728232e+00
-7.21257925e-02 -9.36362967e-02 -2.16767207e-01 7.07719252e-02
-4.33319867e-01 6.09460056e-01 3.58633369e-01 -3.29148546e-02
1.71030372e-01 3.97783630e-02 1.51206732e+00 1.55238003e-01
-4.02516752e-01 8.21220279e-02 -5.90472639e-01 8.63896310e-01
5.73686182e-01 -5.92821166e-02 -1.34196088e-01 -2.20504820e-01
9.19926465e-02 -5.52229047e-01 4.82602566e-01 -9.11114991e-01
2.76026964e-01 -9.72635821e-02 5.01622140e-01 -3.61369371e-01
3.89170557e-01 -1.04249275e+00 1.69785902e-01 5.83289444e-01
9.27414224e-02 -1.77325904e-01 1.19495556e-01 3.92437011e-01
-2.59203583e-01 -2.73490191e-01 1.16253901e+00 2.02052630e-02
-9.38908458e-01 -2.70237714e-01 -3.56465638e-01 -4.13150132e-01
1.02150047e+00 -7.45182037e-01 -4.92137164e-01 -2.59087056e-01
-5.14660716e-01 -3.07137072e-01 6.57469392e-01 -6.64234161e-02
7.67715752e-01 -9.61871624e-01 -5.56494534e-01 6.82084382e-01
-3.85157734e-01 -5.25627136e-01 3.69145513e-01 1.19112158e+00
-9.10029411e-01 9.66436639e-02 -1.58185497e-01 -4.95465219e-01
-1.70855546e+00 5.02977490e-01 2.81422764e-01 9.65275168e-02
-1.04019225e+00 8.36087465e-01 4.02801543e-01 3.93402785e-01
3.96305099e-02 -5.32932162e-01 1.31764650e-01 -1.09776951e-01
8.67800951e-01 6.28801048e-01 7.89097250e-02 -3.58649194e-01
-1.40860379e-01 1.11063421e+00 3.33886072e-02 -1.34697929e-01
1.24524415e+00 -4.59324211e-01 -5.14765620e-01 2.07722634e-01
9.24177825e-01 7.42258012e-01 -1.35274327e+00 -5.19476607e-02
-2.15672404e-01 -9.10409749e-01 2.34540865e-01 -4.71938372e-01
-1.16092551e+00 9.20375109e-01 7.75190890e-01 5.76053917e-01
1.86559439e+00 -4.39814597e-01 6.76078379e-01 7.26799655e-04
6.40335307e-02 -1.29931295e+00 -2.01444492e-01 2.20576048e-01
7.21915662e-01 -9.21216130e-01 3.20635289e-01 -9.56906319e-01
-1.54842108e-01 1.42287290e+00 -9.36502998e-04 -2.23349363e-01
7.95577526e-01 3.98569286e-01 2.51989603e-01 2.57757306e-01
-2.73078948e-01 -3.23548973e-01 3.39578062e-01 6.37064040e-01
4.63305980e-01 -4.80757765e-02 -6.94168806e-01 -1.48785546e-01
2.27020174e-01 -3.87682877e-02 6.85858250e-01 9.19972718e-01
-9.59729373e-01 -1.20482206e+00 -8.73776615e-01 1.32589296e-01
-5.27054846e-01 -2.54207671e-01 -6.93946099e-03 6.41773820e-01
1.92943513e-01 9.48546469e-01 -7.21332133e-02 7.79890865e-02
3.24913770e-01 6.59349263e-02 5.04291177e-01 -1.10778280e-01
-5.25557518e-01 4.57971334e-01 -2.46226341e-01 -6.50226951e-01
-6.80423856e-01 -3.92421693e-01 -1.07740939e+00 -4.30385590e-01
-5.11680841e-01 -2.27638364e-01 6.67223454e-01 6.11983895e-01
1.56837702e-01 3.96843433e-01 3.77711922e-01 -6.45756304e-01
-2.88729817e-01 -6.30132735e-01 -9.25597906e-01 5.96571147e-01
6.11700535e-01 -4.78256941e-01 -5.07171035e-01 5.09166718e-01] | [11.013176918029785, -2.4692118167877197] |
f8c3e7a4-0cfb-4d6f-b834-6b5d75492822 | unsupervised-procedure-learning-via-joint | null | null | http://openaccess.thecvf.com/content_ICCV_2019/html/Elhamifar_Unsupervised_Procedure_Learning_via_Joint_Dynamic_Summarization_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Elhamifar_Unsupervised_Procedure_Learning_via_Joint_Dynamic_Summarization_ICCV_2019_paper.pdf | Unsupervised Procedure Learning via Joint Dynamic Summarization | We address the problem of unsupervised procedure learning from unconstrained instructional videos. Our goal is to produce a summary of the procedure key-steps and their ordering needed to perform a given task, as well as localization of the key-steps in videos. We develop a collaborative sequential subset selection framework, where we build a dynamic model on videos by learning states and transitions between them, where states correspond to different subactivities, including background and procedure steps. To extract procedure key-steps, we develop an optimization framework that finds a sequence of a small number of states that well represents all videos and is compatible with the state transition model. Given that our proposed optimization is non-convex and NP-hard, we develop a fast greedy algorithm whose complexity is linear in the length of the videos and the number of states of the dynamic model, hence, scales to large datasets. Under appropriate conditions on the transition model, our proposed formulation is approximately submodular, hence, comes with performance guarantees. We also present ProceL, a new multimodal dataset of 47.3 hours of videos and their transcripts from diverse tasks, for procedure learning evaluation. By extensive experiments, we show that our framework significantly improves the state of the art performance.
| [' Zwe Naing', 'Ehsan Elhamifar'] | 2019-10-01 | null | null | null | iccv-2019-10 | ['procedure-learning'] | ['computer-vision'] | [ 5.77941418e-01 -2.13847026e-01 -4.50770557e-01 -2.00994670e-01
-1.01675200e+00 -1.04715574e+00 2.75074631e-01 1.58661172e-01
-4.06712025e-01 3.53440255e-01 3.46970677e-01 -2.93897420e-01
-2.15157598e-01 -1.14001907e-01 -1.26951504e+00 -8.32601309e-01
-3.06384206e-01 2.94553995e-01 4.21075262e-02 4.23130155e-01
2.04026893e-01 1.33422494e-01 -1.69266677e+00 4.25427645e-01
8.14493120e-01 8.93902361e-01 3.46685380e-01 1.02870202e+00
2.60994673e-01 1.13037086e+00 -3.04217130e-01 -6.14393428e-02
3.06625187e-01 -5.83532870e-01 -9.38192785e-01 8.16318512e-01
9.30736601e-01 -4.52191770e-01 -6.46529019e-01 1.06607676e+00
2.67616093e-01 6.83294952e-01 6.11559153e-01 -1.25168979e+00
-3.72635736e-03 7.12054253e-01 -3.65432590e-01 1.21921763e-01
8.20012748e-01 8.29199329e-02 9.89715040e-01 -8.85930896e-01
8.51401210e-01 9.47349131e-01 3.46602023e-01 6.35088146e-01
-1.06058073e+00 -2.65723765e-01 5.01125276e-01 3.52038652e-01
-1.12214506e+00 -6.65567756e-01 4.58361804e-01 -4.39012051e-01
6.79361463e-01 4.96759862e-01 9.34186399e-01 1.20137382e+00
-2.96870217e-04 1.40301585e+00 9.41864252e-01 -3.55858445e-01
4.57058579e-01 -1.09168619e-01 2.12578699e-01 1.22063303e+00
5.08172177e-02 -4.72408175e-01 -9.67209160e-01 4.18305695e-02
4.91236269e-01 1.65104255e-01 -4.74762261e-01 -9.48081374e-01
-1.39848650e+00 4.72565442e-01 -2.77445555e-01 -1.18402012e-01
-2.23959386e-01 1.38381645e-01 2.55298138e-01 3.13091904e-01
-1.56691059e-01 2.91466624e-01 -3.86868060e-01 -4.91869122e-01
-9.64783192e-01 2.67400891e-01 1.07724404e+00 1.49119318e+00
7.85510600e-01 -2.98468292e-01 -4.52227235e-01 5.44825137e-01
-2.41770044e-01 5.82927525e-01 2.23662674e-01 -1.45526552e+00
6.63325071e-01 6.18656516e-01 2.36890987e-01 -7.89256632e-01
-2.97410369e-01 -9.64269042e-02 -3.77241969e-01 -5.70262372e-01
6.04185879e-01 -2.57982612e-01 -9.91283953e-01 1.76238012e+00
4.80759770e-01 3.74527276e-01 2.18105558e-02 4.82848674e-01
7.50594735e-01 8.19017351e-01 -2.95609534e-01 -7.64422774e-01
1.25065732e+00 -1.29490709e+00 -9.90339160e-01 -4.02223706e-01
8.20487082e-01 -5.02086937e-01 9.05982256e-01 4.81791884e-01
-1.32238460e+00 -2.21480235e-01 -5.61485767e-01 1.27084315e-01
1.49077654e-01 4.36339170e-01 7.19502687e-01 3.82381976e-01
-8.19934547e-01 3.26259196e-01 -1.10440385e+00 -3.81006598e-01
2.71974772e-01 5.12914836e-01 -5.52131951e-01 -3.75777185e-01
-4.15802509e-01 6.01292491e-01 6.14442348e-01 1.91191494e-01
-1.59165084e+00 -5.02715826e-01 -1.36286128e+00 2.67303973e-01
1.09394920e+00 -5.85184336e-01 1.46114075e+00 -8.83559287e-01
-1.49535036e+00 6.29232287e-01 -5.02985656e-01 -2.32251585e-01
2.75988102e-01 -4.04717863e-01 -2.79090437e-03 7.11192012e-01
-3.46495844e-02 4.12090003e-01 1.01727033e+00 -1.11608732e+00
-9.52137530e-01 -2.00694412e-01 3.48910302e-01 6.03299201e-01
-5.13418913e-01 4.24694642e-02 -1.27314985e+00 -1.78389877e-01
1.47693947e-01 -1.43961442e+00 -2.19360888e-01 -2.72944391e-01
-5.06409764e-01 -1.69516638e-01 3.59290421e-01 -6.49523795e-01
1.61851454e+00 -2.17520809e+00 7.18408763e-01 2.29048789e-01
1.19092710e-01 -1.52148515e-01 -7.66202956e-02 5.24625838e-01
-3.17156054e-02 -4.07431871e-01 -3.76653820e-01 -5.48255801e-01
8.71649384e-02 2.75285691e-01 -1.98015869e-01 8.22914362e-01
-3.18088233e-01 6.59412801e-01 -1.00265312e+00 -6.86770201e-01
2.16597199e-01 -1.76904663e-01 -8.10979307e-01 4.84898359e-01
-2.16838911e-01 2.82890350e-01 -4.65009183e-01 9.27248895e-01
3.57050210e-01 -2.67380506e-01 5.16944289e-01 -8.79500657e-02
3.03783338e-03 8.18804204e-02 -1.67995262e+00 2.20315957e+00
-2.34007433e-01 4.05086070e-01 1.68674842e-01 -1.13022864e+00
3.53475548e-02 2.20475778e-01 7.51801670e-01 -6.45834208e-02
7.31382445e-02 2.06169970e-02 -1.92381546e-01 -8.39222133e-01
3.55769098e-01 3.83260608e-01 -3.04969341e-01 3.04063588e-01
7.32241392e-01 2.60436479e-02 8.98042262e-01 5.75355053e-01
1.38507199e+00 4.14651960e-01 4.53045189e-01 -1.88554332e-01
4.99184400e-01 -1.41582176e-01 5.72862208e-01 1.05661106e+00
-1.90176442e-01 2.16964155e-01 8.86959314e-01 -1.52293697e-01
-3.58131796e-01 -8.45550835e-01 3.03828031e-01 1.26862597e+00
1.34700835e-01 -7.82016695e-01 -1.00019836e+00 -9.08873737e-01
-4.10804749e-01 5.18729091e-01 -6.52965009e-01 -9.48224142e-02
-7.37488925e-01 -2.39721194e-01 9.38145258e-03 3.97527665e-01
1.18182138e-01 -9.13912654e-01 -4.73070502e-01 3.67110707e-02
-3.12108129e-01 -1.55712080e+00 -1.01341939e+00 2.90831417e-01
-8.23861897e-01 -1.28543890e+00 -6.20062888e-01 -1.23762488e+00
1.11913633e+00 4.58999008e-01 1.02612126e+00 -2.41347298e-01
5.64840585e-02 1.02724481e+00 -1.25358596e-01 -2.86501460e-02
-1.46345988e-01 -5.06901667e-02 3.65493745e-02 2.26614431e-01
-1.44936800e-01 -1.35638118e-01 -4.43934709e-01 7.97473043e-02
-8.91851008e-01 2.71868348e-01 5.72459757e-01 6.52249575e-01
8.71524870e-01 2.03276813e-01 -2.66224027e-01 -9.09673452e-01
8.79770517e-02 -2.79752791e-01 -6.67694986e-01 6.50283992e-01
-2.26052981e-02 8.88309330e-02 5.92029333e-01 -5.63275099e-01
-1.13459325e+00 8.18262041e-01 4.77035016e-01 -6.59780204e-01
-1.78189829e-01 6.01230204e-01 -2.35324234e-01 6.35997206e-02
2.81999618e-01 6.05278432e-01 -3.82402122e-01 -1.71741709e-01
2.30470002e-01 3.11173707e-01 5.77572107e-01 -8.23700666e-01
7.40869343e-01 3.45396757e-01 6.98694959e-02 -9.13416982e-01
-1.29544342e+00 -9.02444720e-01 -4.56122220e-01 -4.41592753e-01
6.93648577e-01 -1.13285255e+00 -1.10648584e+00 3.99663895e-01
-7.61893392e-01 -5.40917814e-01 -1.71631396e-01 6.48379564e-01
-9.31527913e-01 6.41132951e-01 -7.97734857e-01 -9.00797248e-01
9.17862728e-02 -1.40779293e+00 1.02160239e+00 2.91253954e-01
1.52263362e-02 -8.11178088e-01 -4.00092825e-02 5.24755538e-01
-4.03922588e-01 9.06086564e-02 6.01679683e-01 -5.33908725e-01
-8.56084108e-01 -1.17004655e-01 4.33275312e-01 2.89531827e-01
5.79342283e-02 6.26311004e-02 -4.51967895e-01 -4.54454869e-01
1.98705643e-02 -4.60421771e-01 9.25388515e-01 4.56487507e-01
1.45232105e+00 -5.26870191e-01 -2.88857907e-01 6.97951913e-01
1.16901338e+00 1.92646742e-01 1.62100539e-01 1.66054517e-01
9.72132683e-01 6.28610253e-01 9.40763295e-01 5.25394261e-01
3.45152646e-01 4.05334383e-01 2.33735949e-01 4.35157329e-01
3.20390850e-01 -2.98218846e-01 9.37119126e-01 1.10883021e+00
-1.11597657e-01 -4.12429214e-01 -7.04610229e-01 7.21573472e-01
-2.10232735e+00 -1.23205543e+00 2.18574867e-01 2.31440306e+00
8.60849977e-01 2.78338529e-02 2.99021751e-01 -5.09085096e-02
6.09681308e-01 1.86994448e-01 -4.59047556e-01 -2.89665282e-01
2.19921529e-01 1.49972066e-01 4.85921651e-01 3.64312500e-01
-1.43054652e+00 9.29959953e-01 6.35606909e+00 8.28886628e-01
-6.00843251e-01 -3.44497055e-01 1.98933795e-01 -4.89958167e-01
-3.62715609e-02 -9.67395082e-02 -8.16977859e-01 3.48213375e-01
9.56133127e-01 -6.60018474e-02 7.97748089e-01 8.15948009e-01
1.06575102e-01 -2.31729656e-01 -1.60249877e+00 1.05507147e+00
3.87159228e-01 -1.14670360e+00 4.96534780e-02 -2.53662646e-01
1.18903017e+00 -5.17489851e-01 -2.06726659e-02 4.58422184e-01
3.22190255e-01 -5.83058238e-01 8.21965158e-01 1.33145422e-01
5.80529749e-01 -8.10132504e-01 2.10517570e-01 5.03539741e-01
-1.28829277e+00 -3.92076015e-01 5.05241267e-02 2.02737734e-01
-3.49557661e-02 5.90630844e-02 -5.66580355e-01 5.46291173e-01
5.72862983e-01 9.89898145e-01 -3.61495763e-01 1.05726922e+00
-4.69297409e-01 8.29416037e-01 -2.25515053e-01 3.23354118e-02
2.20142618e-01 -2.97891706e-01 5.39374590e-01 1.33691430e+00
2.95004725e-01 1.61133394e-01 5.98095834e-01 1.45341277e-01
-1.91418320e-01 1.78181663e-01 -6.31048918e-01 -3.05658400e-01
4.75050867e-01 1.28036177e+00 -7.39325404e-01 -5.01294971e-01
-6.02926552e-01 1.11663389e+00 4.46238071e-01 6.06215298e-01
-1.14981282e+00 -1.01914629e-01 4.79873151e-01 -3.77693743e-01
6.25454485e-01 -3.75088364e-01 2.89167166e-01 -1.56452501e+00
2.73705810e-01 -1.28737581e+00 7.15421557e-01 -3.99271876e-01
-5.27712464e-01 1.27344802e-01 1.47480011e-01 -1.35424125e+00
-2.06782773e-01 -4.35600936e-01 -3.19805771e-01 1.03849933e-01
-1.16682041e+00 -7.83196032e-01 -2.16959625e-01 8.91624928e-01
8.84348392e-01 1.21922992e-01 7.19495714e-01 1.50867343e-01
-9.96309459e-01 4.32051748e-01 4.13665213e-02 -6.87996950e-03
5.72044909e-01 -1.47541761e+00 -2.73267210e-01 1.37372041e+00
3.37634504e-01 8.59588146e-01 6.93866968e-01 -4.47933614e-01
-2.21636915e+00 -8.64069939e-01 6.13530159e-01 -4.70763981e-01
6.80488527e-01 -4.47822571e-01 -3.80940765e-01 1.16185701e+00
1.22764841e-01 -1.09641574e-01 8.48856330e-01 1.37838215e-01
5.60526140e-02 -3.12987380e-02 -5.23339391e-01 7.18335271e-01
1.35678506e+00 -5.72643340e-01 -5.91235399e-01 7.24141717e-01
5.50691485e-01 -1.09128213e+00 -6.18507504e-01 1.21128701e-01
3.60777110e-01 -5.16928077e-01 6.99124277e-01 -7.57600248e-01
6.15615487e-01 -3.11242163e-01 -1.32239893e-01 -1.04771388e+00
-1.32434905e-01 -1.11255217e+00 -1.02533591e+00 9.08893406e-01
3.07919234e-01 1.33645266e-01 8.19935799e-01 6.79157615e-01
-4.78100210e-01 -9.41562057e-01 -6.31540000e-01 -6.98226452e-01
-7.04218090e-01 -4.02806252e-01 -7.32492879e-02 6.89516604e-01
2.29875762e-02 3.28167409e-01 -7.90708184e-01 3.73059779e-01
5.74843764e-01 4.70036387e-01 8.28557849e-01 -5.62723398e-01
-5.34646213e-01 3.00139585e-03 -6.70360262e-03 -1.51051056e+00
4.56679434e-01 -6.69924080e-01 3.01126838e-01 -1.52674186e+00
7.71554887e-01 2.44906694e-01 -3.45528722e-01 6.06173575e-01
-4.12202537e-01 -3.47241759e-01 2.92324424e-01 5.80574246e-03
-1.51993883e+00 3.95760715e-01 1.42761636e+00 -1.99559480e-01
-4.11891758e-01 1.16056502e-01 -4.56644207e-01 8.79334629e-01
1.99477613e-01 -4.55263972e-01 -4.61787343e-01 -4.72674131e-01
1.48597404e-01 4.51074660e-01 -1.32002443e-01 -9.58466947e-01
4.11253780e-01 -3.74174803e-01 9.87191945e-02 -5.51492393e-01
3.91224921e-01 -9.62062597e-01 -1.25247344e-01 4.40921247e-01
-6.23236001e-01 -7.86258578e-02 4.22375128e-02 8.25937510e-01
-1.77202582e-01 -3.62311125e-01 3.50424796e-01 -7.54886791e-02
-1.02691901e+00 5.37898183e-01 -5.38066447e-01 6.05501905e-02
1.35577095e+00 -2.22695962e-01 7.15719443e-03 -4.95906085e-01
-8.87908757e-01 5.41925371e-01 4.48762476e-01 1.87865362e-01
4.46371794e-01 -1.03451896e+00 -3.03180337e-01 -7.13540688e-02
1.86104029e-01 1.23653553e-01 5.32586753e-01 1.07986999e+00
-4.14832860e-01 3.02332103e-01 1.60406187e-01 -7.04091012e-01
-1.64277744e+00 7.31651664e-01 -9.82402861e-02 -2.87640840e-01
-3.99261922e-01 9.57013011e-01 3.24077070e-01 -2.45061621e-01
7.46571422e-01 -5.98885715e-01 -3.41519952e-01 2.35716328e-01
5.14073014e-01 3.88296276e-01 -1.44086406e-01 -3.91432375e-01
-4.14488256e-01 2.88513571e-01 -7.85215646e-02 1.00559801e-01
1.22934413e+00 1.23031531e-02 -4.91879843e-02 5.65386772e-01
1.17938519e+00 1.48938820e-01 -1.56571317e+00 -2.53768712e-01
-1.20942503e-01 -3.19308877e-01 -1.49783939e-01 -4.25227046e-01
-9.42123890e-01 3.23651969e-01 1.26176462e-01 -1.24421388e-01
1.34387231e+00 -3.57565433e-02 6.06636643e-01 1.01455843e+00
3.44794184e-01 -1.20549285e+00 2.07160562e-01 6.62248075e-01
4.78713423e-01 -1.07475400e+00 5.86550348e-02 -6.94654107e-01
-7.78412879e-01 1.12766743e+00 6.39682055e-01 1.99168637e-01
1.28569245e-01 9.27410871e-02 -4.92179722e-01 -9.52848643e-02
-1.03328288e+00 -2.06012577e-01 4.24096882e-01 4.49564680e-02
1.80737704e-01 1.10933959e-01 -4.86485586e-02 4.07722443e-01
1.23953477e-01 1.75304431e-02 6.13531351e-01 1.46230578e+00
-4.08538371e-01 -8.10108304e-01 -1.34448349e-01 4.88110453e-01
-4.56762582e-01 -4.30588238e-02 -1.57710403e-01 4.91127819e-01
-2.83097439e-02 8.43045652e-01 -8.32471102e-02 -1.65648848e-01
4.53401208e-01 2.91632246e-02 9.46525753e-01 -9.21415806e-01
-2.37172768e-01 1.42263636e-01 2.73099780e-01 -9.54436302e-01
-6.73449636e-01 -9.72729683e-01 -1.10086536e+00 -2.02387363e-01
-3.37679952e-01 1.56863138e-01 2.78855175e-01 9.72039342e-01
4.83927839e-02 4.68652248e-01 7.18196452e-01 -7.73187220e-01
-7.04921842e-01 -7.28911936e-01 -5.61686754e-01 3.93353879e-01
2.63410866e-01 -4.66940016e-01 -2.88411677e-01 5.07653832e-01] | [8.71296501159668, 0.6173815727233887] |
f616b498-704b-40ca-9aa8-499e0de09374 | inference-in-linear-dyadic-data-models-with | 2203.03497 | null | https://arxiv.org/abs/2203.03497v5 | https://arxiv.org/pdf/2203.03497v5.pdf | Inference in Linear Dyadic Data Models with Network Spillovers | When using dyadic data (i.e., data indexed by pairs of units), researchers typically assume a linear model, estimate it using Ordinary Least Squares and conduct inference using ``dyadic-robust" variance estimators. The latter assumes that dyads are uncorrelated if they do not share a common unit (e.g., if the same individual is not present in both pairs of data). We show that this assumption does not hold in many empirical applications because indirect links may exist due to network connections, generating correlated outcomes. Hence, ``dyadic-robust'' estimators can be biased in such situations. We develop a consistent variance estimator for such contexts by leveraging results in network statistics. Our estimator has good finite sample properties in simulations, while allowing for decay in spillover effects. We illustrate our message with an application to politicians' voting behavior when they are seating neighbors in the European Parliament. | ['Ko Sugiura', 'Nathan Canen'] | 2022-03-07 | null | null | null | null | ['econometrics'] | ['miscellaneous'] | [-7.91894794e-02 3.27081889e-01 -8.21149528e-01 -2.42899895e-01
-3.21538210e-01 -6.34665608e-01 7.06322193e-01 2.61781335e-01
-4.98926461e-01 1.08006036e+00 6.24570370e-01 -8.38088751e-01
-6.51597679e-01 -1.11825681e+00 -8.78042459e-01 -5.06612301e-01
-5.00454962e-01 3.31690788e-01 -2.44422525e-01 1.15410425e-01
9.89162326e-02 1.97933301e-01 -6.86308026e-01 -1.71016678e-01
8.13909173e-01 3.63586128e-01 -1.82944253e-01 3.55443209e-01
2.79660344e-01 8.11770856e-01 -5.08337319e-01 -5.41822195e-01
4.85742092e-01 -4.96802837e-01 -5.55321932e-01 -4.54238523e-03
3.20506573e-01 -3.74175549e-01 -5.03197551e-01 8.99123847e-01
4.37895387e-01 1.27661377e-01 1.03014886e+00 -1.39340067e+00
-7.23430097e-01 1.25020003e+00 -9.63114023e-01 3.31200570e-01
3.65926802e-01 -1.66429311e-01 1.42627168e+00 -4.11885709e-01
7.24720657e-01 1.42937887e+00 1.00054276e+00 -1.19724579e-01
-2.01142669e+00 -7.45413303e-01 2.97037721e-01 -3.45265746e-01
-1.16769004e+00 -4.78445619e-01 3.87005270e-01 -4.74056959e-01
3.92431915e-01 1.82514787e-01 5.09760499e-01 1.30845964e+00
-1.64693799e-02 3.24255854e-01 1.38081348e+00 -2.38681689e-01
5.75348102e-02 -3.11605837e-02 4.16712344e-01 1.06612533e-01
9.50583696e-01 2.75174141e-01 -3.37043792e-01 -6.86628759e-01
1.06724405e+00 1.76529378e-01 3.64854047e-03 -2.55891204e-01
-1.18376923e+00 1.16886270e+00 5.13341010e-01 3.34443808e-01
-7.68742383e-01 2.76923120e-01 3.55127007e-02 6.74447715e-01
3.12451124e-01 2.42147088e-01 -4.84405197e-02 2.14095246e-02
-6.59350157e-01 2.79155105e-01 1.00359452e+00 8.12000036e-01
5.58507442e-01 -4.79446977e-01 -1.10644372e-02 6.77801549e-01
6.20343611e-02 6.21361494e-01 -1.03184640e-01 -1.32512820e+00
7.43178010e-01 2.07331285e-01 3.75410855e-01 -1.30803299e+00
-5.92567623e-01 -3.28857809e-01 -1.14328194e+00 -1.61488995e-01
1.08234072e+00 -7.84841776e-01 -1.52244061e-01 2.21733975e+00
2.01912910e-01 1.45911902e-01 -3.24510634e-01 4.80977029e-01
2.21479774e-01 4.17417228e-01 4.12635654e-02 -7.79724598e-01
1.09094608e+00 -3.73549014e-01 -6.32261992e-01 -3.70426988e-03
7.33596563e-01 -3.96585792e-01 5.10057390e-01 1.30593941e-01
-1.14305532e+00 -2.31775008e-02 -2.28919163e-01 3.73193085e-01
6.22972399e-02 -5.50154924e-01 7.72380829e-01 5.71673572e-01
-9.07266378e-01 5.64524949e-01 -5.38193703e-01 -4.32285041e-01
4.05337334e-01 1.39960140e-01 -2.75448769e-01 -9.35476571e-02
-1.23117733e+00 4.83159363e-01 -5.94494760e-01 -4.12414335e-02
-5.35083473e-01 -7.66009152e-01 -4.28918272e-01 2.15003237e-01
4.50452298e-01 -9.30927813e-01 1.01628256e+00 -1.06250370e+00
-8.86110544e-01 3.72090310e-01 -1.21934861e-01 -3.25033098e-01
6.71487153e-01 1.70880809e-01 -3.26524109e-01 -2.33263314e-01
6.35469496e-01 -9.88290161e-02 3.12729120e-01 -1.08166802e+00
-4.64065641e-01 -5.29147148e-01 3.53581965e-01 2.44358137e-01
-2.62423247e-01 1.87183723e-01 1.97243169e-01 -4.31146562e-01
9.68942698e-03 -1.03471446e+00 -5.33388376e-01 -1.89933985e-01
-5.53217590e-01 -1.03309982e-01 1.51328921e-01 -2.53359467e-01
1.48717082e+00 -1.79604840e+00 -1.39484555e-01 8.38646472e-01
2.43820816e-01 -4.50874865e-01 -1.53582036e-01 9.32566941e-01
-2.05160901e-01 3.22654426e-01 -6.44646818e-03 -1.94849521e-01
-1.70407612e-02 1.21604390e-01 4.73533273e-02 9.41030681e-01
-6.56801462e-01 5.88848650e-01 -8.98612142e-01 -3.27510774e-01
-2.32042238e-01 -5.05478233e-02 -6.23990715e-01 -4.15262222e-01
5.62964141e-01 3.22722822e-01 -4.27960634e-01 2.96936810e-01
5.38103342e-01 -4.57604140e-01 7.69405007e-01 8.67788643e-02
-2.88799822e-01 3.99665624e-01 -1.41428673e+00 1.04897797e+00
-6.36160851e-01 8.29929113e-01 4.54616696e-01 -1.46727061e+00
3.17940444e-01 3.44632983e-01 4.36999798e-01 -5.24188280e-01
8.26361328e-02 1.49056494e-01 4.51390952e-01 -2.31567234e-01
9.24052298e-02 -2.23294154e-01 -2.18301520e-01 8.99665892e-01
-3.20277393e-01 3.17227989e-01 9.81660932e-02 3.63891035e-01
1.51411533e+00 -7.54033923e-01 4.54716414e-01 -5.88409364e-01
-7.70129412e-02 -1.60014570e-01 7.41038442e-01 1.47850311e+00
3.11392010e-03 8.30345675e-02 1.04882693e+00 1.70104191e-01
-1.31618440e+00 -1.29508591e+00 -4.98100489e-01 7.08395422e-01
-1.71379950e-02 -5.41070402e-02 -4.30022866e-01 -4.87460703e-01
5.34250677e-01 5.87879956e-01 -7.35694706e-01 2.71147072e-01
-1.94496647e-01 -8.17472875e-01 3.71281832e-01 4.06646609e-01
2.21391946e-01 -2.29603529e-01 1.76618367e-01 2.02532604e-01
-1.03743315e-01 -7.10333765e-01 -2.79386789e-01 -2.43510827e-01
-9.51020360e-01 -1.16207218e+00 -7.29022026e-01 -3.10350657e-01
7.97078192e-01 5.95494032e-01 1.06371677e+00 -1.33533344e-01
1.72201067e-01 3.89603317e-01 1.16662988e-02 3.64097431e-02
-9.46908519e-02 -1.32752538e-01 4.18570131e-01 -1.62225023e-01
3.73079956e-01 -8.99935961e-01 -8.72396350e-01 4.08842564e-01
-5.19067287e-01 -3.67112190e-01 4.03244823e-01 8.58176827e-01
2.20231581e-02 -3.10469288e-02 9.74354625e-01 -1.24442315e+00
1.08926380e+00 -1.14199948e+00 -7.38498390e-01 1.14161916e-01
-6.13092005e-01 -4.43965316e-01 5.53268611e-01 -4.75870311e-01
-8.99735808e-01 -5.97305715e-01 6.69902146e-01 3.27557594e-01
1.82947684e-02 8.38515580e-01 1.32060826e-01 3.15587401e-01
7.98699200e-01 -7.94300079e-01 1.77751139e-01 -4.12871093e-01
5.02351105e-01 9.60901976e-01 2.38728430e-02 -5.76979280e-01
7.82726169e-01 6.99135125e-01 2.50186145e-01 -7.56628335e-01
-4.81572598e-01 -3.67820770e-01 -2.19986364e-01 3.68504226e-02
4.67805833e-01 -1.32130218e+00 -1.17935228e+00 2.36938149e-01
-8.14431965e-01 -4.59701955e-01 -1.51390120e-01 1.23045874e+00
-3.69390070e-01 3.74986306e-02 -7.22548485e-01 -1.09520149e+00
3.72652829e-01 -6.88907564e-01 3.33780557e-01 -8.85536661e-04
-6.27742529e-01 -1.44978225e+00 3.01969469e-01 2.49105483e-01
2.40407601e-01 1.17030360e-01 6.07012033e-01 -6.10347807e-01
-3.86136383e-01 -2.05825076e-01 -2.40038291e-01 -3.96656506e-02
3.99456114e-01 1.51186526e-01 -3.07671040e-01 -2.34358191e-01
-4.89693016e-01 1.05892494e-01 5.42823732e-01 1.04034436e+00
7.55065441e-01 -7.75256455e-01 -7.63038576e-01 6.40307367e-02
1.19635665e+00 -3.78547274e-02 2.07447678e-01 -6.16063625e-02
4.75382239e-01 8.80834103e-01 2.46446982e-01 6.67911768e-01
5.92371881e-01 5.59376657e-01 -6.76210150e-02 -1.90942958e-01
3.30085784e-01 -6.13037825e-01 3.30928117e-01 5.25602341e-01
-1.46713302e-01 -5.02737999e-01 -6.82277918e-01 8.82069170e-01
-2.00537181e+00 -1.35146868e+00 -6.94769859e-01 2.47841311e+00
6.79798603e-01 -3.92980054e-02 6.07730627e-01 -2.58104920e-01
1.08743417e+00 -1.08076930e-02 -2.10442677e-01 -4.85613912e-01
-6.95880204e-02 4.00369009e-03 1.09603441e+00 8.71375740e-01
-5.51216066e-01 3.73477310e-01 7.32350206e+00 5.16434848e-01
-3.61267835e-01 1.31796420e-01 7.63228774e-01 -3.35744977e-01
-6.20445967e-01 5.49451053e-01 -5.74341893e-01 5.61464429e-01
1.09433722e+00 -4.83967632e-01 2.62118042e-01 3.66944104e-01
8.03911150e-01 -4.81200844e-01 -9.48894143e-01 5.02736151e-01
-3.27362955e-01 -1.03534770e+00 -6.77872837e-01 7.05683887e-01
1.28611040e+00 -1.57601148e-01 6.08781204e-02 -1.73869163e-01
1.41811204e+00 -1.05817044e+00 1.94437668e-01 4.30937499e-01
4.99001890e-01 -8.01306963e-01 3.31859767e-01 5.04125834e-01
-6.34736478e-01 -3.81068438e-01 -3.48896384e-01 -6.95503831e-01
6.46323919e-01 1.22566438e+00 -4.41641659e-01 3.14730018e-01
4.67911065e-01 7.43046463e-01 7.80444667e-02 1.11799526e+00
-2.05370933e-01 8.21892083e-01 -7.55350709e-01 1.96815088e-01
1.73872754e-01 -7.26051688e-01 4.72231567e-01 6.08133733e-01
6.56150997e-01 2.29935974e-01 -3.04161936e-01 8.27589393e-01
-3.37524980e-01 5.35309166e-02 -1.12215936e+00 1.31921351e-01
9.76559222e-01 1.00920749e+00 -5.56084216e-01 -3.07934940e-01
-9.29031968e-01 3.90763789e-01 1.95043549e-01 5.70040047e-01
-5.88841796e-01 -1.19644880e-01 7.72074103e-01 4.24550354e-01
-3.17917839e-02 -3.11550051e-01 -3.16796809e-01 -1.06665981e+00
-3.65640968e-01 -6.86659396e-01 3.95630866e-01 -4.41516489e-01
-1.67262959e+00 -3.43871593e-01 2.76794076e-01 -8.12184215e-01
-2.79720992e-01 -1.73568279e-01 -6.98895156e-01 8.42234552e-01
-7.56100893e-01 -6.27409577e-01 2.48174012e-01 4.80976194e-01
-2.39087597e-01 3.24130327e-01 3.07418883e-01 4.63121742e-01
-7.10698724e-01 3.81971568e-01 5.19762218e-01 1.14333250e-01
7.30064869e-01 -1.16818142e+00 2.49664292e-01 8.20989132e-01
1.33752868e-01 9.41840529e-01 7.41867006e-01 -6.40780330e-01
-9.75209773e-01 -5.69853842e-01 1.07056546e+00 -3.73754650e-01
1.16497862e+00 -5.50140321e-01 -2.97468901e-01 1.20720124e+00
1.96639776e-01 -1.27778709e-01 9.60363388e-01 8.79478097e-01
-1.71043336e-01 -1.74335316e-01 -1.00879288e+00 9.15806711e-01
1.63105881e+00 -5.60359061e-01 -2.70719737e-01 4.60415900e-01
3.29868376e-01 4.02530134e-01 -1.23668206e+00 1.92228526e-01
6.06488645e-01 -1.12975812e+00 8.88057590e-01 -7.33153164e-01
3.63957405e-01 8.01759064e-02 3.17639709e-02 -1.65448785e+00
-4.95452791e-01 -6.74863756e-01 3.55876982e-01 1.52279544e+00
6.49249196e-01 -9.12962735e-01 3.87890905e-01 7.51089513e-01
5.01139343e-01 -7.63225779e-02 -9.72054064e-01 -9.77108717e-01
4.92464185e-01 -1.42602995e-01 6.07725203e-01 1.45416892e+00
5.03434598e-01 4.10456210e-01 -4.32524443e-01 -4.11244817e-02
7.85258889e-01 9.74052250e-02 9.77511823e-01 -1.39192235e+00
-4.29960698e-01 -4.47318137e-01 -2.06530765e-01 -1.36459625e+00
4.09276664e-01 -4.92258817e-01 -3.05215746e-01 -1.38026047e+00
5.48156023e-01 -9.00518894e-01 1.96265355e-01 6.81061000e-02
8.89354050e-02 7.12161362e-02 -1.53821468e-01 1.44621268e-01
-1.75936282e-01 8.63740072e-02 1.02733493e+00 2.54724026e-01
-3.14703643e-01 4.40840691e-01 -1.04094613e+00 6.53816044e-01
6.48250759e-01 -5.48529148e-01 -3.19459915e-01 -8.11298192e-02
6.29272461e-01 6.67427897e-01 4.66138303e-01 -2.76707947e-01
2.71049529e-01 -5.60422957e-01 3.37881893e-02 4.62421328e-02
2.33082138e-02 -9.69232082e-01 6.69512272e-01 3.33014190e-01
-5.43767691e-01 2.25766152e-01 -4.69614774e-01 8.15760374e-01
3.31691206e-02 -6.90930113e-02 2.88505673e-01 -2.49616038e-02
3.57356608e-01 1.85550317e-01 -5.79829574e-01 2.36254521e-02
9.23359156e-01 -2.69083437e-02 -7.30381429e-01 -1.00546348e+00
-7.09479988e-01 3.44738394e-01 6.15315974e-01 -1.14695154e-01
-9.72343907e-02 -1.56271899e+00 -7.55222857e-01 -3.78037423e-01
-1.15580432e-01 -4.88075227e-01 3.38396132e-02 1.47286618e+00
9.32604149e-02 1.80471495e-01 2.40328893e-01 -1.39319956e-01
-1.25503850e+00 5.47179997e-01 1.75529361e-01 -3.14644650e-02
-2.00667322e-01 4.15652603e-01 5.31372249e-01 -2.53011972e-01
-1.68648645e-01 -1.04190074e-01 -5.29368147e-02 4.55243021e-01
2.25128174e-01 8.17648172e-01 -6.71817958e-01 -6.78381205e-01
-1.64223403e-01 2.82546818e-01 8.02685246e-02 -4.72997397e-01
1.29600048e+00 -7.54836440e-01 -3.50537598e-01 7.59269655e-01
1.23360753e+00 5.07039011e-01 -9.33178604e-01 -4.40816700e-01
-1.12279862e-01 -6.77932918e-01 -2.27478191e-01 -1.72869474e-01
-1.02347362e+00 3.07116508e-01 -1.10575438e-01 8.62278581e-01
4.50375855e-01 1.70064438e-02 5.24923950e-02 1.76477730e-01
2.72383988e-01 -1.12842894e+00 -3.82904023e-01 1.03950299e-01
3.73315662e-01 -9.21578407e-01 2.65882701e-01 -3.97675574e-01
-2.66451716e-01 4.54721302e-01 1.82169359e-02 -3.41422349e-01
9.07034457e-01 -1.02116294e-01 -3.26203585e-01 -1.18971229e-01
-8.48098516e-01 -2.81240139e-02 -1.83983907e-01 5.08282304e-01
6.59855604e-01 4.35931116e-01 -1.09833920e+00 3.33445042e-01
-2.95509905e-01 -4.12877053e-01 9.76389825e-01 5.06127656e-01
-8.84038210e-02 -1.11482704e+00 -4.45471913e-01 8.91356111e-01
-6.96237981e-01 -3.27126086e-01 -4.40418005e-01 1.00085235e+00
-2.22724065e-01 1.22636425e+00 5.86933374e-01 1.89824864e-01
3.33491206e-01 -3.66332203e-01 3.02430004e-01 -4.49901730e-01
-3.96727711e-01 1.65610820e-01 3.91117483e-01 -4.05886978e-01
-7.46709466e-01 -1.14205801e+00 -3.99884462e-01 -1.31302464e+00
-7.03243017e-01 1.97282761e-01 1.25809908e-01 6.22751117e-01
2.53663868e-01 6.82793185e-02 1.11819935e+00 -2.18243599e-01
-5.63619137e-01 -9.54461515e-01 -1.10556746e+00 4.94362772e-01
2.64389783e-01 -6.39192164e-01 -8.69944990e-01 -4.04019564e-01] | [7.708118915557861, 5.231941223144531] |
54f31da5-0e17-4580-b3c7-cd09a68f3be1 | overfitting-at-semeval-2016-task-3-detecting | null | null | https://aclanthology.org/S16-1133 | https://aclanthology.org/S16-1133.pdf | Overfitting at SemEval-2016 Task 3: Detecting Semantically Similar Questions in Community Question Answering Forums with Word Embeddings | null | ['Pascal Poupart', 'Hujie Wang'] | 2016-06-01 | null | null | null | semeval-2016-6 | ['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] | [-7.362209320068359, 3.6909050941467285] |
167fab57-bb72-488c-bc5c-7f0259c2ae2b | automated-segmentation-and-connectivity | null | null | https://downloads.spj.sciencemag.org/bmef/2022/9783128.pdf | https://downloads.spj.sciencemag.org/bmef/2022/9783128.pdf | Automated Segmentation and Connectivity Analysis for Normal Pressure Hydrocephalus | We propose an automated method of predicting Normal Pressure Hydrocephalus (NPH) from CT scans. A deep convolutional network segments regions of interest from the scans. These regions are then combined with MRI information to predict NPH. To our knowledge, this is the first method which automatically predicts NPH from CT scans and incorporates diffusion tractography information for prediction. Introduction. Due to their low cost and high versatility, CT scans are often used in NPH diagnosis. No well-defined and effective protocol currently exists for analysis of CT scans for NPH. Evans’ index, an approximation of the ventricle to brain volume using one 2D image slice, has been proposed but is not robust. The proposed approach is an effective way to quantify regions of interest and offers a computational method for predicting NPH. Methods. We propose a novel method to predict NPH by combining regions of interest segmented from CT scans with connectome data to compute features which capture the impact of enlarged ventricles by excluding fiber tracts passing through these regions. The segmentation and network features are used to train a model for NPH prediction. Results. Our method outperforms the current state-of-the-art by 9 precision points and 29 recall points. Our segmentation model outperforms the current state-of-the-art in segmenting the ventricle, gray-white matter, and subarachnoid space in CT scans. Conclusion. Our experimental results demonstrate that fast and accurate volumetric segmentation of CT brain scans can help improve the NPH diagnosis process, and network properties can increase NPH prediction accuracy. | ['B. S. Manjunath', 'Jefferson Chen', 'Ashutosh Shelat', 'Vikram Iyer', 'Peter Tran', 'Christopher Nguyen', 'Judy Pham', 'Saisidharth Majeti', 'Amil Khan', 'Angela Zhang'] | 2022-01-10 | null | null | null | bme-frontiers-2022-1 | ['3d-medical-imaging-segmentation'] | ['medical'] | [-4.07499857e-02 3.43148708e-01 -8.34994838e-02 -5.70526719e-01
-2.70936817e-01 -1.97837442e-01 3.41872983e-02 3.46007913e-01
-7.19364583e-01 6.41648293e-01 2.68121034e-01 -3.24035555e-01
-2.03622743e-01 -1.04986584e+00 -3.86147618e-01 -4.82446283e-01
-6.93330169e-01 1.26923573e+00 4.99102086e-01 4.10966501e-02
3.39965224e-01 8.08609307e-01 -9.43680406e-01 8.84489529e-03
1.23239648e+00 9.68308091e-01 8.27370346e-01 4.51731652e-01
-4.61298048e-01 4.44836557e-01 -2.42753789e-01 8.05049092e-02
3.98982108e-01 -5.97239673e-01 -1.17071724e+00 -2.36317664e-01
1.31511077e-01 -6.95525169e-01 -1.16965964e-01 9.74994540e-01
8.14900041e-01 -2.76748180e-01 6.98042154e-01 -5.87341964e-01
-3.10657233e-01 8.82232189e-01 -4.20300156e-01 7.69633949e-01
-1.84410557e-01 2.18828365e-01 6.13067329e-01 -6.06598914e-01
8.39535832e-01 7.15533257e-01 7.54322171e-01 6.08288169e-01
-9.50738966e-01 -6.90249979e-01 -4.16789383e-01 6.49924129e-02
-8.31749558e-01 2.97321647e-01 1.21939242e-01 -8.96578848e-01
1.12918258e+00 6.96411133e-02 1.58568323e+00 1.74513817e-01
4.88360733e-01 6.41575217e-01 1.37888467e+00 2.75036152e-02
8.54357183e-02 -5.48948646e-01 2.06381455e-01 1.01900887e+00
1.34854242e-01 -2.60071810e-02 9.25462097e-02 -6.61017746e-02
9.87820327e-01 1.69787183e-01 -4.79907095e-01 -2.85350025e-01
-1.28465092e+00 8.58718991e-01 7.68267810e-01 5.18452168e-01
-5.20822883e-01 -1.58848882e-01 5.88246942e-01 -7.64795467e-02
4.18176711e-01 5.47354162e-01 -4.48741436e-01 1.25889763e-01
-1.36335909e+00 2.95813769e-01 4.75064546e-01 3.35998237e-01
2.60273606e-01 -2.29968101e-01 -2.32658684e-01 1.03888774e+00
1.96373209e-01 5.05070388e-01 1.01652288e+00 -8.32827270e-01
4.45530623e-01 7.13762343e-01 -5.51279485e-01 -2.81145036e-01
-1.29483271e+00 -3.68289322e-01 -7.08144605e-01 2.20873713e-01
3.59812021e-01 -1.54813215e-01 -1.34746921e+00 1.04784310e+00
2.25557566e-01 -2.61843681e-01 -4.41481948e-01 1.26790702e+00
9.07139778e-01 1.53564796e-01 9.40388814e-02 -1.58388868e-01
1.09117043e+00 -9.02017057e-01 -3.56596231e-01 1.42978683e-01
9.59025264e-01 -3.56478035e-01 7.73417294e-01 1.23518564e-01
-1.17682433e+00 1.36210814e-01 -7.45489657e-01 1.40283797e-02
-1.07414663e-01 -1.22748084e-01 6.43450141e-01 4.40332025e-01
-1.11827981e+00 1.06555390e+00 -1.33555436e+00 -3.05917174e-01
1.01396143e+00 7.12288201e-01 -4.35135812e-01 -2.95005701e-02
-8.63568246e-01 1.34475768e+00 4.52025086e-01 -3.25434625e-01
-6.11983061e-01 -1.02867138e+00 -5.17968714e-01 1.04135938e-01
-7.97316581e-02 -7.43260682e-01 1.09734118e+00 -2.90842950e-01
-1.13376415e+00 9.98371780e-01 7.75621086e-02 -9.90724206e-01
7.05634415e-01 2.17965886e-01 2.45322343e-02 8.05590510e-01
2.25982085e-01 8.51305425e-01 3.33987594e-01 -7.41572559e-01
-5.31524777e-01 -8.63735855e-01 -5.18112123e-01 1.94245905e-01
4.07476053e-02 1.24599360e-01 -5.35029918e-03 -3.63908708e-01
8.25014770e-01 -8.07873666e-01 -5.59276760e-01 1.51161462e-01
-4.40060556e-01 -1.25014544e-01 6.54968679e-01 -1.22080588e+00
6.73437774e-01 -1.34041286e+00 -8.65274593e-02 5.08423686e-01
1.05457973e+00 4.66568202e-01 1.71998948e-01 -2.30076492e-01
-3.84577811e-02 4.78132188e-01 -7.09047258e-01 -5.15235998e-02
-3.52549434e-01 1.09436072e-01 2.48842001e-01 7.61473536e-01
-2.36427456e-01 1.12475979e+00 -7.73400307e-01 -6.31610453e-01
3.98923934e-01 3.83354634e-01 -5.11596441e-01 1.10213891e-01
9.60318670e-02 9.11519587e-01 -2.57031858e-01 5.02002239e-01
7.18529940e-01 -1.56655535e-01 -2.42861249e-02 2.42462054e-01
-1.24088459e-01 2.43533164e-01 -2.74508297e-01 1.43382394e+00
-7.02641159e-02 3.65131497e-01 -1.19113542e-01 -1.16163206e+00
9.90821242e-01 3.18985224e-01 8.98343205e-01 -7.86958456e-01
3.89134407e-01 7.23553240e-01 6.80141151e-01 -7.12863743e-01
-3.32711130e-01 -4.45748657e-01 6.40065789e-01 4.33878779e-01
1.50037155e-01 -5.83353639e-01 3.43618929e-01 -1.54714167e-01
1.28206456e+00 -1.37203455e-01 4.06576931e-01 -4.37067807e-01
3.44793767e-01 3.89083587e-02 5.79674602e-01 6.54178560e-01
-5.47277093e-01 8.06598365e-01 5.41852951e-01 -6.75567448e-01
-1.40868616e+00 -1.09421754e+00 -6.80370748e-01 2.72487760e-01
-1.11845575e-01 1.19594693e-01 -9.96995807e-01 -6.15826309e-01
1.08604960e-01 8.39238316e-02 -4.77025896e-01 5.44540524e-01
-8.38098943e-01 -9.49994802e-01 5.56239367e-01 6.74884140e-01
7.44235575e-01 -1.10431838e+00 -6.46993518e-01 3.00148249e-01
-1.50478944e-01 -8.63032043e-01 -2.01228335e-01 1.21694822e-02
-1.52103555e+00 -1.23873162e+00 -1.39920211e+00 -1.04580164e+00
9.00052190e-01 -3.15448433e-01 7.98885167e-01 4.57575917e-01
-5.86060941e-01 -1.15229338e-01 -2.21496783e-02 -2.97151029e-01
-3.18543851e-01 -5.44058974e-04 -1.78476013e-02 -7.86168516e-01
1.76438019e-01 -8.68193984e-01 -1.15803504e+00 7.24532679e-02
-5.59578419e-01 1.89002097e-01 6.08191073e-01 5.99153101e-01
1.06423986e+00 -1.83345526e-01 4.08248693e-01 -1.05747831e+00
5.49716294e-01 -6.03906035e-01 -4.68867034e-01 -1.17376901e-01
-1.01610541e+00 -2.22638026e-01 4.87763286e-01 -7.41743669e-02
-5.69254398e-01 4.28200401e-02 -3.62898260e-01 -3.32276702e-01
-2.31163517e-01 5.21213114e-01 5.68901837e-01 -4.92034137e-01
4.04744238e-01 1.98192447e-01 3.17859530e-01 -5.30847013e-01
-1.29179865e-01 3.75911027e-01 4.61451322e-01 -1.13580987e-01
2.88905114e-01 5.98487079e-01 3.83863926e-01 -5.42025030e-01
-3.94941717e-01 -7.99097717e-01 -1.09728253e+00 -8.37740377e-02
1.06235564e+00 -3.35549325e-01 -3.27987850e-01 2.92206287e-01
-9.20733154e-01 -3.35127324e-01 -2.30136365e-01 8.22408259e-01
-7.07669199e-01 5.61651289e-01 -9.22365546e-01 -3.17096859e-01
-9.35728610e-01 -1.44721603e+00 5.23514569e-01 8.15583020e-02
-9.07532275e-02 -8.87575090e-01 9.18426737e-02 6.62499666e-01
6.50713623e-01 2.28470415e-01 1.26062930e+00 -8.79148364e-01
-3.77645165e-01 4.81050871e-02 -4.63041246e-01 3.77368867e-01
-2.00906172e-01 -4.44870591e-01 -5.85162640e-01 -5.24937548e-02
1.19315676e-01 -4.95639537e-03 1.14369106e+00 9.82096970e-01
1.43801630e+00 -9.42570493e-02 -5.19954860e-01 8.11959803e-01
1.50673318e+00 3.29601198e-01 5.85184813e-01 4.23749179e-01
6.74752951e-01 5.58957100e-01 1.23536780e-01 3.29149455e-01
5.59259057e-01 4.32679765e-02 5.72740138e-01 -3.02818894e-01
-3.95293504e-01 2.09746674e-01 -5.13267875e-01 1.05801141e+00
-6.36879086e-01 1.84899151e-01 -1.60895467e+00 7.70777166e-01
-1.28692985e+00 -6.94429815e-01 -5.07234454e-01 2.01918411e+00
6.84257805e-01 9.72098708e-02 1.39277205e-01 1.28334146e-02
8.42906654e-01 -2.91858047e-01 -6.11281812e-01 -1.12761594e-01
2.33050603e-02 6.58149898e-01 9.48944986e-01 2.37865224e-01
-9.82618451e-01 6.02914691e-01 6.42755365e+00 3.14374894e-01
-1.39010239e+00 5.18623233e-01 5.52190244e-01 -8.57903510e-02
3.59765552e-02 -2.26812094e-01 -5.06976724e-01 5.08604705e-01
6.99126780e-01 4.83735800e-02 3.80730569e-01 4.98589575e-01
3.16618085e-01 -1.02919959e-01 -6.46679401e-01 6.73545241e-01
-7.23703802e-02 -1.45933425e+00 -2.29720429e-01 2.41312221e-01
7.39884794e-01 9.82535899e-01 -1.43341690e-01 3.95800099e-02
-4.10445891e-02 -1.17449069e+00 2.85447687e-01 6.04733586e-01
7.13348567e-01 -4.90328848e-01 1.06910515e+00 2.73928344e-01
-8.84086013e-01 -1.27276391e-01 -3.43790531e-01 1.80923175e-02
2.69350260e-01 6.72971606e-01 -1.54228210e+00 -7.15362504e-02
8.24962914e-01 3.02829087e-01 -2.99994916e-01 1.92594731e+00
-1.78075418e-01 5.82661450e-01 -4.22538310e-01 1.91037178e-01
3.00768137e-01 -2.50804126e-01 5.46093643e-01 1.05995023e+00
6.20661736e-01 4.23817903e-01 2.81566381e-01 1.21086097e+00
-2.28391409e-01 7.70254016e-01 -4.90947127e-01 8.84710476e-02
1.47508472e-01 1.07657743e+00 -1.46245062e+00 -3.51146489e-01
-4.68289882e-01 5.20689428e-01 3.14135760e-01 -6.92529902e-02
-3.58278900e-01 -1.51004568e-01 1.54270604e-02 5.70995986e-01
1.17724784e-01 9.47336853e-03 -7.31855333e-01 -8.68576050e-01
-1.51440427e-01 -2.88164973e-01 1.92258343e-01 -5.18808246e-01
-1.21791863e+00 5.89265049e-01 -3.16116333e-01 -9.40866530e-01
1.47852227e-01 -5.14695704e-01 -6.69539332e-01 1.07662797e+00
-1.64433682e+00 -8.67523611e-01 -1.57435268e-01 1.94792792e-01
3.14030677e-01 -4.50510569e-02 8.37776959e-01 -3.44975176e-03
-1.03432260e-01 -1.28736570e-01 1.84296042e-01 4.22769159e-01
2.05291063e-01 -1.60742652e+00 2.05852747e-01 5.22521496e-01
-7.08929360e-01 5.16805768e-01 3.22217315e-01 -1.15694606e+00
-5.14365017e-01 -1.10804188e+00 9.50694144e-01 -1.78041812e-02
6.22999012e-01 3.33564132e-01 -9.56046760e-01 5.12748063e-01
-2.62907714e-01 4.05425429e-01 6.17063940e-01 -4.43926692e-01
1.57949746e-01 3.06089401e-01 -1.57151330e+00 1.75695181e-01
8.82254481e-01 -1.40277803e-01 -9.83814359e-01 5.77992976e-01
5.31060755e-01 -8.55213284e-01 -1.38591838e+00 7.53659606e-01
4.67595667e-01 -8.79006505e-01 8.45471799e-01 -1.66706502e-01
6.48811042e-01 1.90368965e-01 3.67519230e-01 -1.35903728e+00
-1.08170018e-01 9.87117067e-02 1.84344932e-01 4.23788160e-01
4.80851829e-01 -7.88893223e-01 9.04620826e-01 8.11179459e-01
-4.53614563e-01 -1.21010745e+00 -9.48880732e-01 -5.34477711e-01
7.50751495e-01 -2.17452422e-01 6.87765121e-01 8.76021385e-01
9.24186185e-02 -3.74461830e-01 3.85050982e-01 3.60616371e-02
5.86456716e-01 1.63874224e-01 -1.27057791e-01 -1.65940154e+00
5.78252785e-02 -8.53594542e-01 -8.43562901e-01 -4.10328805e-01
7.43034407e-02 -1.53579760e+00 -2.95848668e-01 -2.25223136e+00
4.91381317e-01 -6.20538712e-01 -1.57944754e-01 4.94896024e-01
3.14553171e-01 3.79650295e-01 1.45633951e-01 4.73502815e-01
9.34624150e-02 2.19193384e-01 2.11574221e+00 -1.48812637e-01
-3.26267719e-01 -1.77527778e-03 4.12823297e-02 9.77736831e-01
1.16895926e+00 -5.02482593e-01 -3.67513865e-01 -1.59031078e-01
-3.36536556e-01 3.90955061e-01 1.69769853e-01 -1.07458639e+00
1.51330441e-01 2.14144856e-01 7.42894113e-01 -7.13650405e-01
1.20491674e-02 -5.20476043e-01 -5.07686317e-01 8.84888172e-01
-1.67673185e-01 -2.14764327e-01 -1.43897131e-01 1.50228217e-02
-1.43001555e-02 -4.16525811e-01 1.00641000e+00 -7.60034919e-01
-3.49718511e-01 7.80627608e-01 -6.18256867e-01 1.15202464e-01
8.87897968e-01 -2.21573755e-01 -3.82357925e-01 -4.93249483e-02
-1.16677523e+00 3.20146650e-01 2.04248160e-01 -1.37565151e-01
8.53160381e-01 -7.29708076e-01 -6.56097233e-01 1.67133749e-01
-3.40563506e-01 1.98610693e-01 3.09629500e-01 1.49495578e+00
-1.33018255e+00 6.38351738e-01 -6.97645724e-01 -7.86538899e-01
-1.19298673e+00 2.01413572e-01 7.79578328e-01 -4.38736290e-01
-1.36559057e+00 7.48673260e-01 2.19291989e-02 -8.09284151e-01
4.16923501e-02 -8.57814491e-01 -6.88757241e-01 -2.24320561e-01
3.34723294e-01 3.55967462e-01 3.06427747e-01 -7.55478859e-01
-5.76076023e-02 5.83831310e-01 -4.82790376e-04 1.98544919e-01
1.64846122e+00 -2.26918384e-01 -6.39444590e-01 9.66512505e-03
1.01114011e+00 -5.82141280e-01 -7.11653709e-01 -1.14686780e-01
-6.05883710e-02 -2.53848702e-01 4.46504056e-01 -1.12327003e+00
-1.46894598e+00 1.04769456e+00 1.08337069e+00 5.79822995e-02
9.13856685e-01 8.30241740e-02 1.25321400e+00 5.37894182e-02
5.92359960e-01 -9.77975726e-01 -5.16106606e-01 5.61428547e-01
6.73584759e-01 -1.24830532e+00 -1.98918507e-01 -4.61003423e-01
-4.33255851e-01 1.25939643e+00 7.97013283e-01 -2.84180492e-01
6.65496230e-01 3.55742395e-01 1.05045110e-01 -5.60811877e-01
-5.37070855e-02 -1.47092506e-01 1.48043975e-01 6.41445756e-01
4.16284621e-01 4.17701513e-01 -8.58635604e-01 6.36420369e-01
-5.70778430e-01 1.01086564e-01 4.83896106e-01 6.91212118e-01
-8.66113603e-01 -9.67162430e-01 -2.36670524e-01 1.42923343e+00
-8.39618385e-01 -2.98579395e-01 -9.67404991e-02 7.45532572e-01
4.49483782e-01 5.93668483e-02 -9.19261202e-02 1.34783253e-01
3.87724228e-02 1.08551688e-01 6.17295563e-01 -7.23423958e-01
-5.96084893e-01 -1.34135792e-02 -1.83969140e-01 -5.43628335e-01
-3.07049364e-01 -5.36130071e-01 -2.23855352e+00 -8.99651572e-02
-2.20487744e-01 -1.32081062e-01 8.95205259e-01 1.28117073e+00
-2.52684634e-02 4.46731955e-01 1.60857767e-01 -9.26524997e-01
-1.71015590e-01 -1.01691902e+00 -1.07712793e+00 2.32144430e-01
3.96371931e-01 -7.10163593e-01 -1.61876023e-01 -3.85637611e-01] | [14.217615127563477, -2.3587467670440674] |
73c03e82-6a47-4f45-84f0-ade9865658ac | progressive-multi-scale-fusion-network-for | 2106.03941 | null | https://arxiv.org/abs/2106.03941v1 | https://arxiv.org/pdf/2106.03941v1.pdf | Progressive Multi-scale Fusion Network for RGB-D Salient Object Detection | Salient object detection(SOD) aims at locating the most significant object within a given image. In recent years, great progress has been made in applying SOD on many vision tasks. The depth map could provide additional spatial prior and boundary cues to boost the performance. Combining the depth information with image data obtained from standard visual cameras has been widely used in recent SOD works, however, introducing depth information in a suboptimal fusion strategy may have negative influence in the performance of SOD. In this paper, we discuss about the advantages of the so-called progressive multi-scale fusion method and propose a mask-guided feature aggregation module(MGFA). The proposed framework can effectively combine the two features of different modalities and, furthermore, alleviate the impact of erroneous depth features, which are inevitably caused by the variation of depth quality. We further introduce a mask-guided refinement module(MGRM) to complement the high-level semantic features and reduce the irrelevant features from multi-scale fusion, leading to an overall refinement of detection. Experiments on five challenging benchmarks demonstrate that the proposed method outperforms 11 state-of-the-art methods under different evaluation metrics. | ['Tania Stathaki', 'Tianhong Dai', 'Yanchu Xie', 'Guangyu Ren'] | 2021-06-07 | null | null | null | null | ['rgb-d-salient-object-detection'] | ['computer-vision'] | [ 3.16690505e-01 -3.02832425e-01 1.63888231e-01 -3.30665261e-01
-5.48476815e-01 -7.12372921e-03 5.01817405e-01 2.30538532e-01
-5.73315442e-01 5.86239636e-01 2.33288348e-01 4.06528413e-01
-1.57516617e-02 -7.19523907e-01 -5.19782424e-01 -9.20438409e-01
4.34898704e-01 -3.82817835e-01 1.05622160e+00 -6.90627247e-02
4.69058365e-01 3.75448197e-01 -1.87143493e+00 1.78220570e-01
1.16368663e+00 1.31355238e+00 8.61655116e-01 5.78547036e-03
-1.63386568e-01 7.31474638e-01 -4.19373840e-01 -1.50495157e-01
4.36216831e-01 -1.81873918e-01 -4.06333923e-01 3.73299807e-01
4.20522511e-01 -4.93008047e-01 -1.44331887e-01 1.39240062e+00
5.63430369e-01 1.28240168e-01 2.47324631e-01 -1.05197823e+00
-1.87064290e-01 1.25475243e-01 -1.13341451e+00 5.16461730e-01
1.36484191e-01 7.48150274e-02 6.76259637e-01 -1.20782030e+00
4.68746513e-01 1.32896423e+00 3.17721248e-01 1.47418395e-01
-8.50113511e-01 -4.57323909e-01 3.60273391e-01 4.87495244e-01
-1.46354675e+00 -3.10178995e-01 1.05728042e+00 -3.24555576e-01
4.96919513e-01 6.02579415e-02 5.85984349e-01 3.57530504e-01
9.93435159e-02 8.24459791e-01 1.21364200e+00 -3.43204021e-01
6.31010458e-02 3.71553786e-02 -1.67635232e-02 8.96694720e-01
4.77869332e-01 -3.84745263e-02 -6.34509146e-01 5.52902482e-02
7.75425911e-01 2.96918631e-01 -5.28634250e-01 -5.97918689e-01
-1.13904440e+00 5.43270528e-01 8.23572457e-01 2.21248463e-01
-5.18819392e-01 -2.10450456e-01 1.96453005e-01 -2.10229203e-01
5.04421830e-01 2.27014180e-02 -3.12633991e-01 2.66709000e-01
-7.67822146e-01 2.15792105e-01 -7.05343187e-02 5.64551771e-01
1.03039944e+00 -1.81160122e-01 -4.57911521e-01 8.97938132e-01
3.05572897e-01 2.73057789e-01 4.26241279e-01 -8.77490580e-01
5.83637834e-01 9.50337708e-01 3.58273268e-01 -1.09714031e+00
-5.26883483e-01 -6.10610485e-01 -7.58458078e-01 4.06272918e-01
3.09991539e-01 7.95764625e-02 -9.67738092e-01 1.39616716e+00
6.96686983e-01 2.49028832e-01 -1.67635694e-01 1.42902064e+00
1.02585697e+00 4.65342700e-01 -6.30897582e-02 -3.45781684e-01
1.24260890e+00 -1.00693929e+00 -6.23666346e-01 -4.34497863e-01
2.73886859e-01 -8.74803841e-01 6.83502853e-01 3.32905114e-01
-1.04017210e+00 -7.63069212e-01 -1.18041015e+00 -4.97486144e-02
-1.52794883e-01 2.46002138e-01 5.36379457e-01 4.10364509e-01
-8.40124547e-01 2.45739132e-01 -7.84614146e-01 -2.21351191e-01
5.67416787e-01 1.95961371e-01 -3.69096696e-01 -4.95036274e-01
-1.01307511e+00 8.35216880e-01 5.03552139e-01 4.37841445e-01
-7.36348987e-01 -5.04597068e-01 -8.44700217e-01 -1.70287535e-01
7.11088419e-01 -7.11759448e-01 8.61540675e-01 -7.13673294e-01
-1.00634909e+00 4.75794762e-01 -4.51657653e-01 -2.82539696e-01
4.96693313e-01 -2.42083281e-01 -1.59617051e-01 3.00291508e-01
3.28454912e-01 8.11243713e-01 8.33175898e-01 -1.37802207e+00
-1.26251829e+00 -5.35406947e-01 2.25067630e-01 6.03518903e-01
-2.81515718e-01 2.86469795e-02 -8.05124104e-01 -5.71008265e-01
5.80523729e-01 -6.16841793e-01 -2.97287315e-01 1.00383945e-01
-1.79341003e-01 -2.20630974e-01 8.66620541e-01 -5.11378109e-01
1.08312500e+00 -2.06198716e+00 2.77989477e-01 -1.44268438e-01
2.27891743e-01 2.37410426e-01 9.06332433e-02 -7.58449221e-03
3.65989149e-01 -2.83725798e-01 -2.78861791e-01 -4.51292753e-01
-5.56370497e-01 1.27015233e-01 8.46673399e-02 5.75605571e-01
3.60624045e-01 6.01289570e-01 -8.59978259e-01 -7.78098822e-01
6.54904842e-01 4.02214587e-01 -3.89207393e-01 1.45245403e-01
-4.38674055e-02 6.24178290e-01 -5.06381631e-01 7.99603403e-01
1.03085554e+00 -1.12515531e-01 -4.81277406e-01 -4.24576283e-01
-4.04476881e-01 -8.47549830e-03 -1.37920499e+00 1.78843558e+00
-2.08702803e-01 3.08477491e-01 2.19181731e-01 -7.87096620e-01
9.17929411e-01 -1.35551319e-01 4.25388128e-01 -7.94824302e-01
1.79405883e-01 1.88117325e-01 7.06115887e-02 -3.73236597e-01
5.91444612e-01 5.36162406e-02 2.17108130e-01 -1.88619033e-01
-1.93561658e-01 1.12077378e-01 1.13166243e-01 5.46526276e-02
7.00259626e-01 -1.54113141e-03 3.00034583e-01 -8.82078335e-02
1.06632185e+00 -2.23250911e-01 1.08149719e+00 5.97456217e-01
-5.10010779e-01 7.37881899e-01 1.01755194e-01 -2.79519707e-01
-6.17562413e-01 -8.15100133e-01 -1.59472108e-01 8.21066320e-01
8.69389713e-01 -2.98830092e-01 -5.03587246e-01 -5.49549699e-01
-2.51029003e-02 1.64247736e-01 -6.40854537e-01 -1.45085424e-01
-3.96486491e-01 -8.11903238e-01 3.39741968e-02 5.00155091e-01
1.01556587e+00 -8.46567512e-01 -9.06088173e-01 1.81405291e-01
-4.56593096e-01 -1.30386329e+00 -4.19417679e-01 -4.94153053e-02
-8.86565208e-01 -8.99923742e-01 -9.82827425e-01 -6.35118365e-01
6.75959408e-01 1.01536393e+00 5.90270340e-01 1.45424664e-01
-3.19396198e-01 -5.25205582e-02 -5.16035736e-01 -4.25365925e-01
1.45192549e-01 -1.20722920e-01 -1.61895022e-01 4.81046975e-01
2.11342528e-01 -2.37134397e-01 -1.01068556e+00 4.38942760e-01
-9.74647522e-01 3.61727595e-01 8.07834983e-01 8.00016284e-01
6.13569021e-01 2.77365863e-01 4.95723814e-01 -4.25315320e-01
1.29155695e-01 -2.10596457e-01 -6.28624618e-01 6.21749051e-02
-4.93237674e-01 -2.15714108e-02 3.07029665e-01 -8.39768648e-02
-1.35917509e+00 1.25420794e-01 1.10767715e-01 -5.08031905e-01
-9.44139343e-03 2.82978982e-01 -4.65004563e-01 -3.36900204e-01
2.98595548e-01 4.13006037e-01 -1.09614588e-01 -5.92909575e-01
1.27888829e-01 6.07149780e-01 5.05539656e-01 -2.64782786e-01
7.01446414e-01 9.00876760e-01 1.73498482e-01 -6.02027774e-01
-1.02481890e+00 -7.75385916e-01 -5.37618101e-01 -3.05299819e-01
6.84674025e-01 -1.16543603e+00 -3.14515084e-01 7.89207876e-01
-1.15752184e+00 2.60635734e-01 9.87562612e-02 4.10670906e-01
-1.66610017e-01 6.31594360e-01 -2.77183622e-01 -8.54538620e-01
-2.63516158e-01 -1.30915189e+00 1.26227498e+00 7.13771343e-01
4.76997375e-01 -4.83656287e-01 -4.25877631e-01 5.06385028e-01
2.91974455e-01 1.86653227e-01 4.03706223e-01 4.30639600e-03
-9.56989646e-01 1.05475351e-01 -6.69063687e-01 4.11526650e-01
2.76856631e-01 -3.35194707e-01 -9.38211262e-01 -1.96080804e-01
1.56721115e-01 2.38389261e-02 1.21580040e+00 4.98826087e-01
9.53297853e-01 3.19803774e-01 -4.45789814e-01 5.77409625e-01
1.61066294e+00 1.32073626e-01 4.47645098e-01 5.63846409e-01
9.48735714e-01 6.06989264e-01 1.28568459e+00 6.11821532e-01
4.39193428e-01 8.61778021e-01 6.85843349e-01 -2.15946242e-01
-3.03813845e-01 7.62675777e-02 2.84996212e-01 2.33433276e-01
-1.08135812e-01 1.71071544e-01 -7.01062799e-01 6.86567724e-01
-2.03554583e+00 -6.05413735e-01 -2.03638375e-01 2.06709599e+00
6.07898951e-01 2.95481563e-01 4.40293401e-02 2.58083999e-01
9.75359976e-01 2.07096636e-01 -5.44528842e-01 2.09123448e-01
-4.28715318e-01 -1.92850217e-01 6.38961315e-01 2.51081198e-01
-1.25152361e+00 7.41224766e-01 4.72250366e+00 9.25862789e-01
-1.03451800e+00 1.40236259e-01 3.64751905e-01 -1.12021796e-01
8.15403461e-03 -3.97671163e-02 -9.86388922e-01 5.70489585e-01
3.08958534e-02 -9.17031616e-02 -1.01401201e-02 7.27920949e-01
4.33857203e-01 -6.84306622e-01 -6.81686878e-01 1.21714139e+00
2.03012764e-01 -1.10639215e+00 -8.88118818e-02 -3.68049741e-02
8.82768333e-01 7.80408457e-02 -2.60632653e-02 -1.53362960e-01
-2.87003309e-01 -4.36403513e-01 8.23826969e-01 4.87331778e-01
3.67150217e-01 -8.14495981e-01 1.07653391e+00 3.55812997e-01
-1.49505591e+00 -4.16192114e-01 -4.68033671e-01 -6.33597896e-02
1.58473268e-01 7.94154584e-01 -4.73529786e-01 9.52830076e-01
9.83330727e-01 9.64498162e-01 -8.93640041e-01 1.54003835e+00
-1.35131657e-01 1.64052341e-02 -3.23303759e-01 2.75302112e-01
2.12199703e-01 -7.19720945e-02 7.33423710e-01 8.36708307e-01
1.66333020e-01 1.39335558e-01 1.91232353e-01 6.49947464e-01
3.20155025e-01 1.01396613e-01 -1.42143294e-01 6.16067052e-01
4.44998711e-01 1.36262488e+00 -9.13857579e-01 -2.90109694e-01
-7.20387638e-01 9.71329391e-01 1.72013432e-01 1.35844126e-01
-6.55325770e-01 -1.15938842e-01 5.72482944e-01 2.11270645e-01
6.79584742e-01 -1.01114608e-01 -2.99204767e-01 -1.17336261e+00
3.51900339e-01 -5.35828412e-01 4.44422603e-01 -8.35694611e-01
-9.71066058e-01 3.83850396e-01 -1.43100200e-02 -1.48648190e+00
2.55800873e-01 -2.34569848e-01 -4.26446170e-01 9.82559204e-01
-1.94080913e+00 -9.98163521e-01 -7.67367840e-01 5.48567712e-01
7.91161776e-01 1.53049469e-01 3.47896852e-02 3.87153417e-01
-7.08847165e-01 2.84714371e-01 -1.37728468e-01 -1.18183330e-01
6.89547598e-01 -8.76965165e-01 -1.01489879e-01 1.21967781e+00
-1.97639316e-01 3.60946208e-01 6.07404053e-01 -6.28302455e-01
-1.04041028e+00 -1.22654402e+00 4.34823692e-01 -5.61292171e-02
1.32111847e-01 3.06551643e-02 -9.49432909e-01 2.46262364e-02
-2.16300428e-01 3.29915166e-01 -2.75303572e-02 -3.72955263e-01
-4.31713741e-03 -4.19118375e-01 -1.10696828e+00 4.12145883e-01
1.06292081e+00 -2.08682775e-01 -6.43323779e-01 -1.98983878e-01
6.60513878e-01 -4.61988181e-01 -5.54298043e-01 8.17718625e-01
4.38323855e-01 -1.43314600e+00 1.04932928e+00 3.08501959e-01
2.20701620e-01 -9.19240296e-01 -6.72461912e-02 -1.02732015e+00
-2.99323440e-01 -1.06445178e-01 -1.14306107e-01 1.34515798e+00
-2.92101175e-01 -5.56542993e-01 6.67124569e-01 3.79440218e-01
-2.07711041e-01 -7.91587174e-01 -9.23090339e-01 -5.42601705e-01
-5.22899330e-01 -1.38371721e-01 4.82526511e-01 5.15456259e-01
-4.50180113e-01 9.86004546e-02 -3.00293267e-01 4.29268003e-01
8.22738767e-01 3.04670274e-01 6.66391134e-01 -1.41111755e+00
-6.93214163e-02 -3.51492018e-01 -7.41063058e-01 -1.19394457e+00
-3.10859680e-01 -4.19294119e-01 1.72714099e-01 -1.75849032e+00
4.04728204e-01 -3.72106999e-01 -5.76945245e-01 2.80101657e-01
-7.21564472e-01 4.64776129e-01 3.05579007e-01 2.75859863e-01
-7.85772622e-01 8.04408550e-01 1.45092654e+00 6.05666675e-02
-1.83150545e-01 -1.96652874e-01 -7.75821745e-01 8.58457327e-01
5.22802651e-01 -4.11421180e-01 -4.00184304e-01 -3.25272977e-01
-1.28905177e-01 -8.28683898e-02 5.39141417e-01 -1.22096229e+00
3.64015460e-01 -1.08704917e-01 6.31837785e-01 -9.89009976e-01
4.56429601e-01 -8.78816545e-01 -2.36876443e-01 3.91597033e-01
1.99184984e-01 -3.06910157e-01 2.56840676e-01 8.38464320e-01
-5.10238171e-01 -6.82336232e-03 9.85912681e-01 -1.28501635e-02
-1.15255439e+00 2.34147459e-01 7.47211501e-02 -2.53537059e-01
1.11535823e+00 -4.74559397e-01 -3.05226713e-01 1.21773062e-02
-3.72118205e-01 3.41107607e-01 6.88872039e-01 6.70833766e-01
8.14879417e-01 -1.18374765e+00 -6.60810113e-01 2.54275143e-01
4.18322116e-01 2.84827530e-01 5.50485373e-01 1.04025447e+00
-2.89131045e-01 4.13939416e-01 -3.97260427e-01 -8.88405383e-01
-1.37701404e+00 4.30361301e-01 2.68080980e-01 -8.01811740e-02
-6.25703692e-01 9.18998063e-01 6.91557050e-01 3.25069666e-01
1.62193388e-01 -4.35223490e-01 -4.81396437e-01 9.51239094e-02
6.39024496e-01 3.68807733e-01 2.44363412e-01 -8.54407907e-01
-6.53894722e-01 9.00671184e-01 -2.86184549e-01 4.73331027e-02
1.16362369e+00 -6.94916248e-01 -1.85111407e-02 3.44235063e-01
7.82607377e-01 -1.40792280e-01 -1.56587684e+00 -5.76496184e-01
-2.43232742e-01 -8.67303431e-01 4.61377859e-01 -5.28196216e-01
-1.20884943e+00 7.71238685e-01 7.95022070e-01 -1.04359716e-01
1.41707039e+00 -9.13737863e-02 8.10371280e-01 -1.37759998e-01
6.47952020e-01 -9.79084373e-01 1.26292259e-01 4.85204672e-03
7.44047225e-01 -1.53846931e+00 2.47729927e-01 -8.46235156e-01
-5.88898659e-01 8.18982601e-01 8.11815381e-01 -1.21963471e-02
4.03431416e-01 -2.79843453e-02 -1.27953261e-01 3.55670936e-02
-3.89396459e-01 -6.48051262e-01 4.15801138e-01 4.33430374e-01
2.93588880e-02 -3.17213386e-01 -4.79449958e-01 5.63044369e-01
4.25786078e-01 6.03193827e-02 4.37699974e-01 1.00645447e+00
-8.01268339e-01 -8.98196042e-01 -6.84055984e-01 3.51404846e-01
-4.18824792e-01 -4.59289700e-02 -8.82232562e-02 5.62209010e-01
6.44019485e-01 1.17838585e+00 -7.27893710e-02 -2.51925737e-01
2.58777738e-01 -3.03706735e-01 5.42218745e-01 -5.40360391e-01
-2.80896306e-01 2.21192509e-01 -1.90361723e-01 -5.98803878e-01
-8.16739917e-01 -7.76640117e-01 -1.23531592e+00 1.45423383e-01
-5.68156242e-01 -1.39466166e-01 4.13791269e-01 9.56319571e-01
3.38469326e-01 5.52686751e-01 5.27905166e-01 -1.08447528e+00
-2.00580493e-01 -1.00711906e+00 -4.89708036e-01 3.57882410e-01
6.03427827e-01 -1.19705331e+00 -3.99367690e-01 -1.03387430e-01] | [9.695062637329102, -0.719302773475647] |
93a089d9-967d-4334-9dc4-2d6f5cb1bf99 | constraint-based-causal-structure-learning | null | null | http://papers.nips.cc/paper/9573-constraint-based-causal-structure-learning-with-consistent-separating-sets | http://papers.nips.cc/paper/9573-constraint-based-causal-structure-learning-with-consistent-separating-sets.pdf | Constraint-based Causal Structure Learning with Consistent Separating Sets | We consider constraint-based methods for causal structure learning, such as the PC algorithm or any PC-derived algorithms whose first step consists in pruning a complete graph to obtain an undirected graph skeleton, which is subsequently oriented. All constraint-based methods perform this first step of removing dispensable edges, iteratively, whenever a separating set and corresponding conditional independence can be found. Yet, constraint-based methods lack robustness over sampling noise and are prone to uncover spurious conditional independences in finite datasets. In particular, there is no guarantee that the separating sets identified during the iterative pruning step remain consistent with the final graph. In this paper, we propose a simple modification of PC and PC-derived algorithms so as to ensure that all separating sets identified to remove dispensable edges are consistent with the final graph,thus enhancing the explainability of constraint-basedmethods. It is achieved by repeating the constraint-based causal structure learning scheme, iteratively, while searching for separating sets that are consistent with the graph obtained at the previous iteration. Ensuring the consistency of separating sets can be done at a limited complexity cost, through the use of block-cut tree decomposition of graph skeletons, and is found to increase their validity in terms of actual d-separation. It also significantly improves the sensitivity of constraint-based methods while retaining good overall structure learning performance. Finally and foremost, ensuring sepset consistency improves the interpretability of constraint-based models for real-life applications. | ['Honghao Li', 'Nadir Sella', 'Herve Isambert', 'Vincent Cabeli'] | 2019-12-01 | null | null | null | neurips-2019-12 | ['tree-decomposition'] | ['graphs'] | [ 2.97285736e-01 2.85356492e-01 -7.34780371e-01 8.05665404e-02
-2.90478796e-01 -4.50955033e-01 3.52835566e-01 3.76478881e-01
1.04574084e-01 9.88214433e-01 -3.65579873e-02 -3.99826407e-01
-8.92054617e-01 -9.97175038e-01 -6.40123010e-01 -9.59743619e-01
-4.24359500e-01 5.26083350e-01 3.90992165e-01 3.24521989e-01
1.96002461e-02 6.59264863e-01 -1.57796943e+00 7.00629726e-02
8.63932550e-01 5.05330265e-01 -2.58563496e-02 3.01067919e-01
-3.65392566e-02 5.77484548e-01 -3.16115975e-01 -2.41974875e-01
-4.90292981e-02 -6.40659809e-01 -7.51321852e-01 4.41989660e-01
5.97998835e-02 -1.12568088e-01 -4.06032950e-02 8.71105492e-01
-1.00617841e-01 -8.58159289e-02 6.44192696e-01 -1.51775932e+00
-7.63041154e-02 6.92721546e-01 -6.74198925e-01 7.31008640e-03
2.83538043e-01 -2.54138768e-01 1.28074205e+00 -4.80930656e-01
6.36444092e-01 1.22738707e+00 6.36599541e-01 2.09297881e-01
-1.77448308e+00 -8.07646811e-01 5.04995346e-01 1.40600607e-01
-1.44839239e+00 -1.87913641e-01 9.66154099e-01 -2.72170514e-01
5.94592988e-01 5.58493912e-01 8.49565685e-01 8.59872341e-01
3.35537344e-01 7.19307482e-01 8.29201460e-01 -5.58606684e-01
5.39765477e-01 -1.24402642e-01 2.42564991e-01 8.36040139e-01
8.67061496e-01 2.63633192e-01 -7.46167481e-01 -3.48989576e-01
5.51244378e-01 -1.61529392e-01 -1.49908394e-01 -7.25893736e-01
-8.21209729e-01 8.40854168e-01 2.61252701e-01 1.95187718e-01
-1.92627370e-01 1.97769612e-01 1.05919145e-01 2.24688038e-01
2.35503435e-01 2.64868319e-01 -5.10270894e-01 4.08231914e-01
-9.45334077e-01 1.47037476e-01 6.80331528e-01 1.03114605e+00
8.82254660e-01 -3.06698792e-02 1.87155738e-01 3.79170984e-01
6.14080787e-01 2.30086774e-01 -3.16147171e-02 -7.79545426e-01
2.11378783e-01 1.02991104e+00 -2.67519683e-01 -1.33658600e+00
-5.11240602e-01 -6.43555760e-01 -1.05143332e+00 -5.85998520e-02
5.65443814e-01 -1.10076077e-01 -9.29875255e-01 1.71802235e+00
4.92668778e-01 2.38412350e-01 -1.46355212e-03 4.84365433e-01
4.89762992e-01 4.22459185e-01 2.60537770e-02 -6.86089039e-01
1.04392135e+00 -3.93464983e-01 -7.41486609e-01 -2.00825572e-01
6.19432092e-01 -3.50927740e-01 6.11461043e-01 5.60912848e-01
-6.63158715e-01 -2.46012345e-01 -1.29730868e+00 5.15148163e-01
1.29604623e-01 -1.83128148e-01 1.00165939e+00 6.29146278e-01
-5.38294256e-01 6.21642530e-01 -9.90703821e-01 -8.32247734e-02
3.29689294e-01 5.05620837e-01 -2.90998220e-01 -1.48077965e-01
-9.41548288e-01 5.07571936e-01 3.69191498e-01 -3.93433422e-02
-8.85663688e-01 -7.42440164e-01 -7.66845584e-01 1.30163446e-01
9.28789735e-01 -5.42165399e-01 6.93489492e-01 -7.23495841e-01
-7.94793844e-01 4.09787893e-01 -4.31833953e-01 -2.41357550e-01
4.18920100e-01 2.62408733e-01 -3.37669194e-01 1.22828700e-01
4.18948606e-02 2.11334065e-01 7.64606237e-01 -1.77596712e+00
-6.83228374e-01 -5.46670616e-01 8.86417776e-02 -9.39301550e-02
-2.98742950e-01 -2.26628423e-01 -7.41075277e-01 -4.26900029e-01
8.76583040e-01 -1.18641627e+00 -3.71700406e-01 -2.63538957e-01
-8.67749214e-01 -1.62722751e-01 7.06615567e-01 -3.50444853e-01
1.74909651e+00 -2.14012146e+00 2.67235130e-01 8.40317845e-01
3.23977649e-01 -1.03153609e-01 1.48059586e-02 5.94489217e-01
-2.58222640e-01 4.75274444e-01 -7.76986182e-01 2.32715949e-01
-3.55677754e-01 4.21904206e-01 3.08582857e-02 5.41740537e-01
4.17546779e-01 5.45289934e-01 -9.02150035e-01 -6.76217020e-01
2.52810985e-01 6.76278919e-02 -6.39923573e-01 1.16745867e-01
-3.47175360e-01 3.32286805e-01 -4.26003277e-01 4.55079705e-01
5.79267859e-01 -1.66481301e-01 9.97038305e-01 1.17983960e-01
7.74685070e-02 1.40671596e-01 -1.72777200e+00 9.84943032e-01
-1.54564455e-01 2.62626529e-01 1.38883129e-01 -1.28840983e+00
1.07035160e+00 4.06101137e-01 8.77628386e-01 -3.23564261e-01
-1.21865384e-01 8.82125348e-02 2.20300123e-01 -3.25735658e-01
-1.91058382e-01 -2.26801217e-01 9.36332420e-02 3.27999026e-01
-2.36537382e-01 -5.80576761e-03 3.41248095e-01 3.52928042e-02
1.06796896e+00 -2.33836379e-02 4.55373406e-01 -4.65311736e-01
6.86318636e-01 9.93295461e-02 9.06546533e-01 5.71755886e-01
1.82769477e-01 2.84922928e-01 9.70616639e-01 -1.69826820e-01
-5.42223573e-01 -1.15325165e+00 -3.31164926e-01 4.26507950e-01
-2.02068202e-02 -7.49284625e-01 -3.58264804e-01 -9.29568291e-01
1.31730288e-01 6.99064791e-01 -5.93857765e-01 -2.81801909e-01
-5.55292070e-01 -9.06405747e-01 1.84916705e-01 3.40084374e-01
2.93518811e-01 -5.42814970e-01 -3.96954030e-01 2.91958243e-01
-1.52559787e-01 -7.05737591e-01 4.62616943e-02 6.34502530e-01
-1.24511051e+00 -1.70931673e+00 1.77394107e-01 -7.33500540e-01
9.42121506e-01 2.08480105e-01 8.07837009e-01 5.41404426e-01
7.85726607e-02 -4.79963943e-02 -2.15876535e-01 -1.88823491e-01
-3.64120424e-01 -2.24529982e-01 -1.69123530e-01 -8.77540559e-02
1.35306492e-01 -8.48205864e-01 -1.01591311e-01 6.80649817e-01
-9.53390896e-01 -7.06848130e-03 4.97145921e-01 7.94167459e-01
7.60888398e-01 7.87797153e-01 4.74157393e-01 -8.64669144e-01
4.54743177e-01 -4.42721456e-01 -6.60728931e-01 4.06187922e-01
-1.10854948e+00 3.03032041e-01 7.12047517e-01 -2.45436937e-01
-9.10721660e-01 2.37539828e-01 2.96988457e-01 -3.48509699e-01
4.83508818e-02 9.70807493e-01 -4.14914280e-01 3.88447344e-01
5.19592583e-01 -2.36327411e-03 2.15973966e-02 -3.29048991e-01
1.90094277e-01 1.70717150e-01 3.66596073e-01 -7.36238658e-01
7.68509865e-01 4.50153738e-01 6.97344542e-01 -9.15001929e-01
-4.88964885e-01 -3.63503724e-01 -9.34644401e-01 -2.20105410e-01
5.57871580e-01 -5.61546206e-01 -6.06701851e-01 -3.93573046e-02
-9.47821259e-01 8.44557658e-02 2.77958829e-02 6.24914944e-01
-3.47634286e-01 5.73351800e-01 -2.66260177e-01 -1.02028239e+00
2.40834355e-01 -9.60292757e-01 5.09485960e-01 -1.76925272e-01
-4.45677340e-01 -1.13545728e+00 5.36597110e-02 1.55219167e-01
-4.45585579e-01 4.42573875e-01 1.45432603e+00 -5.30283630e-01
-6.11648917e-01 -2.11770296e-01 -3.81688960e-02 4.87472154e-02
3.68186057e-01 3.62494826e-01 -4.28639531e-01 -2.37556666e-01
-2.05581903e-01 1.85201928e-01 7.81373143e-01 6.21916294e-01
8.49512994e-01 -3.62318754e-01 -6.99571252e-01 3.99766833e-01
1.53347492e+00 2.82670915e-01 2.59599477e-01 1.71648011e-01
7.01800644e-01 8.81256104e-01 6.58112407e-01 2.97295064e-01
4.63146679e-02 6.83032274e-01 3.49789798e-01 -1.51524663e-01
-1.46847159e-01 -2.35115260e-01 7.12147504e-02 5.89944363e-01
-2.25759894e-01 -1.67382360e-01 -1.01663518e+00 5.13315976e-01
-2.32571268e+00 -9.20135260e-01 -7.44764149e-01 2.42791247e+00
8.22063386e-01 3.02883625e-01 1.04857348e-01 7.28554130e-01
7.91960418e-01 -7.58484900e-02 -3.42118979e-01 -3.45264703e-01
1.85007498e-01 1.29713446e-01 3.20164084e-01 5.68258166e-01
-8.96886349e-01 5.76917171e-01 6.09659910e+00 5.71522236e-01
-5.91034651e-01 -4.58468825e-01 3.19565415e-01 1.16891325e-01
-5.61314225e-01 4.41187829e-01 -7.25878179e-01 1.39430135e-01
7.80796468e-01 -3.96840721e-01 1.39892653e-01 6.48828447e-01
6.32366896e-01 -4.27602082e-01 -1.36138332e+00 4.39108372e-01
-1.28961608e-01 -1.12257266e+00 1.34043470e-01 1.99993372e-01
6.48446262e-01 -6.50429547e-01 -5.00180960e-01 -1.46308661e-01
4.64124650e-01 -8.51475358e-01 5.57421446e-01 1.10286020e-01
4.98734742e-01 -1.27593124e+00 6.53566599e-01 3.75430942e-01
-1.21123695e+00 -1.50447175e-01 -1.04112700e-01 -1.60911188e-01
6.87524602e-02 1.10105419e+00 -7.95194030e-01 9.30514634e-01
6.12475097e-01 7.13871956e-01 -4.18082744e-01 9.27171290e-01
-6.94438636e-01 8.67933512e-01 -2.72168666e-01 8.43316317e-02
7.73658752e-02 -4.49134618e-01 6.60640001e-01 1.07128286e+00
-1.07853808e-01 8.55067447e-02 3.75948668e-01 8.92678857e-01
2.07765073e-01 -8.41628686e-02 -6.63591981e-01 2.59675980e-01
5.58641851e-01 8.79934072e-01 -1.03370619e+00 -5.33315130e-02
-4.01213378e-01 3.11211646e-01 2.05183581e-01 3.10147643e-01
-8.13128710e-01 -3.06665972e-02 3.92489105e-01 1.61521137e-01
3.06355178e-01 -4.55400884e-01 -5.66843271e-01 -7.96788812e-01
-5.85888773e-02 -1.03015387e+00 7.61363089e-01 -9.65645388e-02
-9.23711956e-01 1.25357345e-01 4.76491004e-01 -1.10016584e+00
-1.29808888e-01 -2.74566442e-01 -4.25198853e-01 3.49981040e-01
-1.32585657e+00 -1.03539324e+00 -3.21323015e-02 6.12475395e-01
3.15769285e-01 4.22195673e-01 6.99909270e-01 -6.64720312e-02
-9.81371880e-01 2.22322062e-01 1.47863077e-02 -2.25121174e-02
3.84360969e-01 -1.14458990e+00 -3.71379703e-01 1.25851011e+00
1.68650344e-01 6.77144468e-01 6.24505699e-01 -1.18333638e+00
-1.37740600e+00 -9.49295998e-01 8.91663015e-01 -2.23674700e-01
5.64138055e-01 -4.22163427e-01 -8.18282008e-01 6.47456825e-01
-2.61764675e-01 -3.12379301e-01 6.58120811e-01 6.66216969e-01
-2.76076406e-01 -2.09161699e-01 -8.65410030e-01 6.71301723e-01
1.27056229e+00 -1.51191846e-01 -6.77567899e-01 2.68586069e-01
3.69515240e-01 1.41655698e-01 -8.36033344e-01 7.72824824e-01
4.52330500e-01 -8.86855960e-01 9.59344268e-01 -7.61823356e-01
5.62307477e-01 -5.22258461e-01 3.76510955e-02 -9.42502141e-01
-6.08857572e-01 -4.25828606e-01 -4.01227713e-01 1.21450007e+00
7.53895879e-01 -3.83489370e-01 8.83365631e-01 5.53063512e-01
6.11222908e-02 -6.48039043e-01 -1.05281937e+00 -1.18970621e+00
-1.50064752e-01 -7.55408347e-01 4.35687184e-01 1.04689717e+00
2.55758941e-01 5.24266541e-01 -2.47126088e-01 4.16780353e-01
9.44712341e-01 2.32928991e-01 7.20874846e-01 -1.60870767e+00
-3.32637757e-01 -3.22077572e-01 -9.76407230e-02 -4.00129914e-01
1.02686703e-01 -9.11161780e-01 -1.07069269e-01 -1.53402734e+00
3.52352649e-01 -6.58910573e-01 -1.04344927e-01 6.29677892e-01
-1.50810508e-02 -7.31633930e-03 -1.54644668e-01 4.05002713e-01
-5.83234988e-02 2.52477616e-01 1.27731657e+00 -1.78514868e-01
-6.00979745e-01 1.75655603e-01 -5.75370550e-01 7.93998361e-01
6.45521522e-01 -8.40673208e-01 -7.40102530e-01 8.00000653e-02
2.02980429e-01 2.06624374e-01 2.55394936e-01 -6.49753809e-01
1.37670293e-01 -6.12638354e-01 4.25054640e-01 -5.18892765e-01
-3.45448852e-01 -1.02910876e+00 7.06828177e-01 8.67347360e-01
-1.77173495e-01 -2.44317204e-01 2.79669315e-01 7.70292103e-01
-2.56764446e-03 -3.93643469e-01 6.68475211e-01 -1.26260994e-02
-2.65279502e-01 7.71121606e-02 -5.09438157e-01 -3.35083753e-01
1.14312136e+00 -3.85412246e-01 9.60866734e-02 -6.42283335e-02
-7.46453643e-01 1.87923044e-01 2.66845554e-01 3.32611650e-01
3.90948683e-01 -1.02278531e+00 -5.60549796e-01 1.06433563e-01
-5.42570874e-02 2.36888781e-01 -2.44415015e-01 9.53426838e-01
-8.24968070e-02 5.44635832e-01 3.78118418e-02 -6.01733506e-01
-1.40797710e+00 8.89385343e-01 -4.12871912e-02 -4.83141303e-01
-6.78664804e-01 3.47275108e-01 3.28271799e-02 -6.39829040e-02
2.21968293e-01 -3.54928523e-01 -2.06539392e-01 8.00780430e-02
7.35989884e-02 5.47615945e-01 -2.06789896e-02 -1.83215335e-01
-8.27418864e-01 5.03032625e-01 1.11015350e-01 2.47565731e-02
1.39312792e+00 -1.27072185e-02 -3.44385177e-01 1.77570120e-01
8.58547091e-01 4.04922336e-01 -9.85406578e-01 1.51216626e-01
4.66867536e-01 -3.74952376e-01 1.21663667e-01 -7.78506279e-01
-1.08269954e+00 2.06843510e-01 -7.78320059e-02 3.04591566e-01
1.18653011e+00 5.64104952e-02 1.57141890e-02 2.70383418e-01
2.64289916e-01 -8.82662654e-01 -1.21193258e-02 -1.38702067e-02
8.58068347e-01 -7.69656062e-01 6.55588448e-01 -1.16018033e+00
-5.13326712e-02 1.16723979e+00 6.63724720e-01 1.31841645e-01
6.68259323e-01 3.10218424e-01 -4.44382876e-01 -3.07557613e-01
-7.87222505e-01 -2.11399078e-01 3.41238230e-01 6.52053535e-01
3.32415223e-01 -3.59884650e-02 -8.03248227e-01 6.55856729e-01
-1.43317461e-01 -4.32224721e-01 4.38905865e-01 9.84007418e-01
-3.87274086e-01 -1.23596406e+00 -3.62553865e-01 3.76193196e-01
-3.36039104e-02 2.19693467e-01 -6.44141674e-01 1.20444930e+00
2.08583802e-01 1.51996636e+00 -1.04521997e-01 -2.99380869e-01
3.75181317e-01 -1.89079046e-02 5.90440333e-01 -5.62299192e-01
-2.32163206e-01 3.25400084e-01 4.99753535e-01 -4.78072345e-01
-5.56755543e-01 -9.04447436e-01 -1.58362830e+00 -3.66836399e-01
-6.81209564e-01 2.49287203e-01 2.90243328e-01 1.02259398e+00
1.28322214e-01 5.47326744e-01 6.97808385e-01 -1.04564413e-01
-9.90997553e-02 -3.22761625e-01 -4.91802871e-01 2.35637799e-01
-1.18575003e-02 -7.83407152e-01 -6.19243026e-01 1.55801317e-02] | [7.771693706512451, 5.327712535858154] |
1ab0fd6b-a708-4574-983f-aaad32f8523c | lake-net-topology-aware-point-cloud | 2203.16771 | null | https://arxiv.org/abs/2203.16771v1 | https://arxiv.org/pdf/2203.16771v1.pdf | LAKe-Net: Topology-Aware Point Cloud Completion by Localizing Aligned Keypoints | Point cloud completion aims at completing geometric and topological shapes from a partial observation. However, some topology of the original shape is missing, existing methods directly predict the location of complete points, without predicting structured and topological information of the complete shape, which leads to inferior performance. To better tackle the missing topology part, we propose LAKe-Net, a novel topology-aware point cloud completion model by localizing aligned keypoints, with a novel Keypoints-Skeleton-Shape prediction manner. Specifically, our method completes missing topology using three steps: 1) Aligned Keypoint Localization. An asymmetric keypoint locator, including an unsupervised multi-scale keypoint detector and a complete keypoint generator, is proposed for localizing aligned keypoints from complete and partial point clouds. We theoretically prove that the detector can capture aligned keypoints for objects within a sub-category. 2) Surface-skeleton Generation. A new type of skeleton, named Surface-skeleton, is generated from keypoints based on geometric priors to fully represent the topological information captured from keypoints and better recover the local details. 3) Shape Refinement. We design a refinement subnet where multi-scale surface-skeletons are fed into each recursive skeleton-assisted refinement module to assist the completion process. Experimental results show that our method achieves the state-of-the-art performance on point cloud completion. | ['Lizhuang Ma', 'Yuan Xie', 'Ran Yi', 'Zhijun Gong', 'Junshu Tang'] | 2022-03-31 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Tang_LAKe-Net_Topology-Aware_Point_Cloud_Completion_by_Localizing_Aligned_Keypoints_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Tang_LAKe-Net_Topology-Aware_Point_Cloud_Completion_by_Localizing_Aligned_Keypoints_CVPR_2022_paper.pdf | cvpr-2022-1 | ['point-cloud-completion'] | ['computer-vision'] | [-5.22824414e-02 7.98745826e-02 -5.80820851e-02 2.85538007e-02
-9.18168843e-01 -5.85327387e-01 5.80448687e-01 3.07438463e-01
5.56698479e-02 1.35263607e-01 -1.32652730e-01 1.49475217e-01
-7.44994134e-02 -1.06172943e+00 -9.31526065e-01 -3.11283976e-01
1.10304601e-01 9.85461712e-01 6.46961212e-01 4.52703945e-02
4.98541504e-01 9.59993005e-01 -1.34814560e+00 -5.30389808e-02
1.05141926e+00 9.27364588e-01 5.43248475e-01 4.68517065e-01
-4.11068499e-01 -1.42597020e-01 -1.57073155e-01 -2.53535155e-02
4.94960785e-01 2.34163985e-01 -4.67866838e-01 1.89495906e-01
5.64808369e-01 -2.75329500e-01 -6.59675747e-02 8.41487885e-01
7.63287991e-02 -2.25472644e-01 5.94313562e-01 -1.28518605e+00
-7.93927535e-02 5.55514731e-02 -8.33358407e-01 -3.84156585e-01
2.69659400e-01 1.15991309e-02 9.61679935e-01 -1.52173972e+00
7.37383127e-01 1.18697798e+00 8.27905893e-01 1.03813477e-01
-1.08652043e+00 -8.61393094e-01 1.42014503e-01 -2.01938108e-01
-1.75644100e+00 -2.11048737e-01 1.16906917e+00 -3.35714340e-01
5.00601768e-01 2.38109931e-01 7.74811804e-01 2.62752652e-01
-1.31535664e-01 6.16599381e-01 5.65620184e-01 -6.24655783e-02
1.81822211e-01 -4.60162878e-01 -2.53685921e-01 8.24702561e-01
1.48655042e-01 -1.59085274e-01 -4.26863283e-01 -5.87989926e-01
1.52036476e+00 5.85060120e-01 -1.27441376e-01 -9.84002411e-01
-1.60461605e+00 4.31678444e-01 7.05400944e-01 -8.94230008e-02
-6.66144073e-01 2.33918756e-01 -1.50415540e-01 -3.84983681e-02
5.10953426e-01 1.13792829e-01 -6.55951262e-01 1.02100074e-01
-9.67146575e-01 3.49599898e-01 4.43488389e-01 1.43551660e+00
1.39773285e+00 -4.64820683e-01 1.26199648e-01 6.25432611e-01
4.69522685e-01 8.43625128e-01 -1.63280040e-01 -9.89422381e-01
7.55292416e-01 1.35419774e+00 1.06332056e-01 -1.21979022e+00
-4.36557084e-01 -3.79052162e-01 -9.13480461e-01 1.51074499e-01
1.20964497e-01 3.93622160e-01 -9.13405359e-01 1.09199953e+00
8.76483560e-01 6.34829640e-01 -3.62034351e-01 6.57450199e-01
6.53008878e-01 5.21121025e-01 -3.08774620e-01 -5.63627854e-02
1.38722384e+00 -7.22683847e-01 4.88358289e-02 5.82119487e-02
2.11921722e-01 -8.13057840e-01 7.37249374e-01 2.29480967e-01
-1.08370697e+00 -5.11772633e-01 -7.64959574e-01 -1.24140389e-01
1.46070555e-01 2.42737576e-01 3.70330572e-01 -9.96043086e-02
-7.81338394e-01 6.35554075e-01 -9.57412720e-01 2.39681602e-02
5.77598691e-01 3.47529918e-01 -4.30445403e-01 -1.22018702e-01
-2.69902289e-01 2.03975350e-01 1.41111225e-01 -1.34063497e-01
-7.67978907e-01 -1.25676036e+00 -6.81115210e-01 9.79023576e-02
5.66908538e-01 -1.15317464e+00 9.16827857e-01 -2.50910491e-01
-1.04962301e+00 7.08827555e-01 -5.15209079e-01 1.83736440e-03
5.44900298e-01 -1.02153711e-01 1.10019118e-01 3.41338903e-01
4.62919295e-01 7.24303484e-01 1.07185543e+00 -1.66105628e+00
-8.91965210e-01 -5.95299363e-01 -4.00223881e-01 2.12367341e-01
1.77101523e-01 -4.64626580e-01 -9.92367268e-01 -7.72644579e-01
1.00221384e+00 -7.84380436e-01 -4.02154773e-01 3.31728727e-01
-5.41345894e-01 -5.17557383e-01 1.06903887e+00 -4.81483579e-01
9.10342872e-01 -2.05319071e+00 9.91956592e-02 6.83721542e-01
5.83106339e-01 -6.93518370e-02 -1.95708394e-01 5.54478049e-01
5.27744703e-02 8.71105641e-02 -4.72310126e-01 -7.16143429e-01
-1.88062951e-01 2.00741172e-01 -5.04515052e-01 4.17599171e-01
2.61673212e-01 8.58714879e-01 -9.77110445e-01 -5.22721887e-01
4.66658920e-01 5.04686773e-01 -5.78294575e-01 2.74679828e-02
-2.69467503e-01 5.74257731e-01 -9.57212985e-01 1.12480795e+00
1.24610364e+00 -2.26394236e-01 -3.79746318e-01 -4.30879861e-01
-2.97960669e-01 -1.01181030e-01 -1.60601377e+00 2.12325382e+00
-2.60346144e-01 -1.46630958e-01 4.24012095e-01 -3.49597156e-01
1.13233316e+00 1.51594564e-01 7.80863941e-01 -2.18387976e-01
-3.08088005e-01 3.46904308e-01 -6.44134223e-01 6.89309239e-02
4.92273569e-01 1.62214823e-02 1.67248443e-01 2.05104768e-01
-2.82850146e-01 -4.96618032e-01 -2.50606298e-01 3.37007254e-01
1.25660086e+00 1.91407815e-01 6.03787601e-02 -5.84447347e-02
6.36534810e-01 1.52207792e-01 8.68427992e-01 2.56366849e-01
2.37547740e-01 1.10968733e+00 9.67964679e-02 -4.80222583e-01
-1.33071887e+00 -1.29531717e+00 -1.21574067e-01 4.75401372e-01
5.92630506e-01 -8.53827357e-01 -4.99181300e-01 -5.81506252e-01
3.27160448e-01 1.34100080e-01 -2.72908747e-01 2.83547908e-01
-8.18445206e-01 -4.31144126e-02 1.49916679e-01 4.75455403e-01
2.87525058e-01 -8.88255954e-01 -4.43547070e-01 1.80655524e-01
-2.51955271e-01 -9.86156225e-01 -6.74687028e-01 -3.36237341e-01
-1.46926045e+00 -1.37921238e+00 -7.89292514e-01 -7.35779285e-01
1.15778875e+00 6.50119245e-01 8.95314515e-01 5.92836201e-01
2.37114802e-02 4.49537665e-01 -3.29865515e-01 -9.98541564e-02
-8.07443112e-02 9.29639786e-02 5.07517830e-02 1.60428792e-01
-5.03830127e-02 -1.01526105e+00 -8.77816498e-01 5.45039356e-01
-7.24225521e-01 3.34174424e-01 9.91139352e-01 4.50319231e-01
1.30260062e+00 8.28866810e-02 1.99894145e-01 -3.59052688e-01
-3.64929400e-02 -3.55443120e-01 -5.77990413e-01 2.01424249e-02
-2.32833400e-01 3.55497710e-02 5.00022352e-01 -2.00815648e-01
-5.87231815e-01 5.61455965e-01 -6.26698732e-02 -9.67656195e-01
-2.78585702e-01 2.54025131e-01 -3.45964551e-01 -2.62895256e-01
1.78948820e-01 5.10368943e-01 -5.69649152e-02 -1.01411843e+00
3.93792123e-01 3.52742881e-01 6.86074317e-01 -8.14435422e-01
1.52131438e+00 1.04942584e+00 2.85501570e-01 -9.01538074e-01
-3.02854151e-01 -9.43241358e-01 -1.15543401e+00 -2.59387735e-02
5.08588910e-01 -1.07191718e+00 -7.13585258e-01 3.91567290e-01
-1.47320497e+00 1.56800523e-02 -4.07470942e-01 2.38625810e-01
-6.09916747e-01 6.19615197e-01 -2.80747563e-01 -5.63912809e-01
-4.75733638e-01 -8.84093344e-01 1.76924086e+00 4.38784854e-03
2.11025283e-01 -5.94986320e-01 2.85404325e-01 2.39620656e-01
-2.16473192e-01 4.81664419e-01 6.38950467e-01 -2.45861366e-01
-1.29316616e+00 -2.10346535e-01 -2.90950954e-01 -1.21632323e-01
1.73391029e-02 9.25206691e-02 -5.95003843e-01 -3.27956557e-01
-1.14564478e-01 2.80088663e-01 6.42178595e-01 1.57466099e-01
1.07288802e+00 -1.62431404e-01 -8.50983024e-01 9.14511681e-01
1.46326494e+00 -4.30128276e-01 4.86305296e-01 8.02230388e-02
1.14791334e+00 3.23385537e-01 7.36447215e-01 6.56307399e-01
7.35818386e-01 5.81036985e-01 8.06470871e-01 -1.30066529e-01
-1.68578938e-01 -9.09345090e-01 5.85335642e-02 9.92867649e-01
-1.64343968e-01 4.22876596e-01 -1.20069838e+00 7.32906461e-01
-1.87028456e+00 -4.88477021e-01 -5.89852273e-01 2.27954769e+00
5.45408070e-01 -2.99240556e-02 9.85473022e-02 1.30606785e-01
9.58403647e-01 -5.82002066e-02 -6.49602354e-01 6.07593179e-01
1.00206412e-01 2.22928882e-01 5.31838179e-01 4.12435472e-01
-8.16881120e-01 1.13170528e+00 4.92998362e+00 1.14195073e+00
-6.59165502e-01 -1.30573492e-02 -3.69077697e-02 4.30463642e-01
-4.06773210e-01 6.02971792e-01 -7.28256881e-01 2.83372134e-01
3.94473411e-02 1.38505757e-01 1.23200163e-01 1.07820141e+00
2.66406447e-01 4.11900766e-02 -8.53544116e-01 1.19409740e+00
-1.09325707e-01 -1.49043322e+00 3.74933183e-01 3.90771449e-01
7.29593396e-01 2.60707855e-01 -3.88871521e-01 -1.57194212e-01
1.06535852e-01 -5.23069680e-01 8.03985059e-01 8.32000375e-01
1.04349613e+00 -9.33654606e-01 1.33241013e-01 8.65673065e-01
-1.96614254e+00 5.52296229e-02 -5.62252939e-01 1.50658965e-01
2.94888109e-01 8.23923469e-01 -8.49109352e-01 9.21229482e-01
5.30273914e-01 9.21183109e-01 -5.65368950e-01 1.47099578e+00
-2.30950236e-01 3.00462246e-01 -7.24966466e-01 5.44619203e-01
-5.45847118e-02 -4.11579758e-01 1.00632370e+00 7.41026282e-01
4.74062741e-01 3.79909635e-01 4.89673495e-01 1.07438433e+00
4.88934154e-03 1.35970637e-01 -4.43565845e-01 4.95784312e-01
9.63189721e-01 1.64997339e+00 -1.16132832e+00 -2.62579829e-01
-2.52668142e-01 7.69334197e-01 2.76017904e-01 7.01623410e-02
-4.24173594e-01 -2.13586360e-01 6.28737867e-01 6.14541233e-01
4.26175177e-01 -6.53859854e-01 -6.19165063e-01 -1.23030412e+00
3.51336420e-01 -2.96832800e-01 7.11335540e-02 -9.82300937e-01
-9.95259225e-01 4.86715743e-03 -7.75920078e-02 -1.54047441e+00
3.84972155e-01 -1.56374872e-01 -9.60852683e-01 7.18536854e-01
-1.46890819e+00 -1.69806361e+00 -7.38916099e-01 7.12921441e-01
6.25604570e-01 1.58019379e-01 3.99218619e-01 4.10553515e-02
-1.04421131e-01 1.26252592e-01 -1.00594521e-01 1.03722528e-01
3.45201612e-01 -1.13947701e+00 6.85872197e-01 6.93523288e-01
1.68190733e-01 7.76596844e-01 1.15919635e-01 -1.28159535e+00
-1.66037321e+00 -1.23627603e+00 5.33925235e-01 -6.52548492e-01
3.60730201e-01 -3.98017704e-01 -9.88888323e-01 5.15441239e-01
-7.39053309e-01 5.24979942e-02 -2.13450715e-02 -1.56793952e-01
-1.51121199e-01 -9.59163085e-02 -1.04226255e+00 5.33949792e-01
1.25629544e+00 -3.58225048e-01 -7.31146395e-01 2.35193864e-01
8.10007930e-01 -4.74503696e-01 -9.94497776e-01 6.44611776e-01
3.10046971e-01 -6.08046889e-01 1.36387384e+00 1.53098762e-01
1.59080192e-01 -9.20073152e-01 8.08241069e-02 -1.03199542e+00
-3.56974602e-01 -7.06279397e-01 -2.58729219e-01 1.34234262e+00
-6.95371926e-02 -3.10327530e-01 1.23756242e+00 1.97642922e-01
-5.03860593e-01 -7.24215150e-01 -1.10488427e+00 -5.94257474e-01
-1.70961916e-01 -6.60850346e-01 1.09190953e+00 8.42652142e-01
-3.66058230e-01 2.56088048e-01 2.37936433e-02 5.31456351e-01
9.75012183e-01 2.95127302e-01 1.41187871e+00 -1.79544139e+00
2.83194184e-01 -2.43766397e-01 -5.01632154e-01 -1.62378120e+00
-1.42401859e-01 -9.41559374e-01 -1.54103160e-01 -1.64371347e+00
1.09905422e-01 -1.07751083e+00 2.84042060e-01 4.50412720e-01
-1.43609997e-02 2.29999974e-01 1.13640562e-01 8.14732850e-01
-5.42001247e-01 8.41674566e-01 1.49698436e+00 1.49668097e-01
-4.20901805e-01 1.59146130e-01 -3.39658916e-01 1.04907560e+00
3.93282413e-01 -5.56959152e-01 -1.07417025e-01 -3.47666621e-01
2.81226963e-01 3.18591058e-01 6.98066294e-01 -1.26275563e+00
6.43757403e-01 -3.10694307e-01 3.55950147e-01 -1.52775824e+00
5.11993170e-01 -1.10619378e+00 3.35857600e-01 2.79313475e-01
3.69107306e-01 1.60982966e-01 -2.24200830e-01 9.91177738e-01
7.64892325e-02 4.39213105e-02 4.51918423e-01 -1.84624448e-01
-4.07392114e-01 1.06764472e+00 5.11660099e-01 -2.00589538e-01
1.10193992e+00 -5.04822910e-01 6.50468171e-02 1.28139660e-01
-6.52902663e-01 5.35420835e-01 1.10268831e+00 3.10491800e-01
1.18525779e+00 -1.49461615e+00 -8.10907245e-01 4.60326672e-01
2.93145299e-01 1.08190727e+00 2.27101877e-01 7.96360016e-01
-7.79383719e-01 9.30326581e-02 1.91060707e-01 -1.14534664e+00
-1.12784600e+00 3.74704361e-01 1.52627546e-02 8.62654857e-03
-1.19919610e+00 4.37252313e-01 3.68620515e-01 -6.22529209e-01
-9.41121131e-02 -6.45299673e-01 6.33020326e-02 -2.16854379e-01
3.61012638e-01 5.33413589e-01 7.92132020e-02 -9.18361068e-01
-3.48105550e-01 1.37409282e+00 1.86173797e-01 2.26638769e-03
1.42263901e+00 -1.05669796e-01 -3.26582044e-01 1.58501983e-01
9.61864471e-01 3.96112889e-01 -1.49967599e+00 -4.59797651e-01
-1.10291563e-01 -6.34301603e-01 -1.73375577e-01 -2.72458643e-01
-9.42847669e-01 7.12817252e-01 4.87853400e-02 -1.73908398e-01
7.73518622e-01 2.46012241e-01 1.14558268e+00 1.65127411e-01
8.23485970e-01 -8.49251330e-01 3.33965793e-02 4.48717356e-01
1.01193714e+00 -8.85601103e-01 1.19159289e-01 -9.53252912e-01
-1.10719740e-01 1.25965548e+00 5.33586144e-01 -4.57323700e-01
7.60069430e-01 -1.44313782e-01 -4.90487725e-01 -6.29363537e-01
-4.17148679e-01 -4.13456373e-02 3.81858498e-01 6.56480014e-01
-5.09163558e-01 1.21135041e-01 1.04842253e-01 5.14923453e-01
-2.92222798e-01 -1.96305439e-01 1.67884827e-01 8.61675799e-01
-7.37330735e-01 -1.12685525e+00 -9.09908414e-01 4.09582973e-01
1.40792742e-01 7.09149390e-02 -3.95511061e-01 5.48933029e-01
2.26059228e-01 4.94866550e-01 3.71155232e-01 -4.50544089e-01
4.85923082e-01 -2.72653937e-01 2.09720675e-02 -9.05855775e-01
-2.16975376e-01 2.94556290e-01 -5.72903693e-01 -7.19737649e-01
-4.24595438e-02 -7.15368032e-01 -1.60925269e+00 -2.48514906e-01
-3.48938674e-01 2.46564582e-01 7.68803120e-01 7.34271109e-01
7.27466762e-01 3.40553932e-02 7.56340504e-01 -1.39848053e+00
-1.91633329e-01 -7.50150323e-01 -5.08437812e-01 4.25965935e-01
3.15526932e-01 -7.69052327e-01 -3.04191321e-01 -7.13857934e-02] | [8.330403327941895, -3.4848885536193848] |
d3637e33-7591-4278-92a7-08f4a9ad0614 | unsupervised-rewriter-for-multi-sentence | null | null | https://aclanthology.org/P19-1216 | https://aclanthology.org/P19-1216.pdf | Unsupervised Rewriter for Multi-Sentence Compression | Multi-sentence compression (MSC) aims to generate a grammatical but reduced compression from multiple input sentences while retaining their key information. Previous dominating approach for MSC is the extraction-based word graph approach. A few variants further leveraged lexical substitution to yield more abstractive compression. However, two limitations exist. First, the word graph approach that simply concatenates fragments from multiple sentences may yield non-fluent or ungrammatical compression. Second, lexical substitution is often inappropriate without the consideration of context information. To tackle the above-mentioned issues, we present a neural rewriter for multi-sentence compression that does not need any parallel corpus. Empirical studies have shown that our approach achieves comparable results upon automatic evaluation and improves the grammaticality of compression based on human evaluation. A parallel corpus with more than 140,000 (sentence group, compression) pairs is also constructed as a by-product for future research. | ['Akiko Aizawa', 'Xiaoyu Shen', 'Yang Zhao', 'Wei Bi'] | 2019-07-01 | null | null | null | acl-2019-7 | ['sentence-compression'] | ['natural-language-processing'] | [ 6.18969619e-01 2.25464910e-01 1.85753610e-02 -2.77961135e-01
-8.85977268e-01 -1.60383791e-01 5.07601917e-01 5.76497614e-01
-5.69461763e-01 8.55914712e-01 5.37326336e-01 -4.97882634e-01
1.42425597e-01 -1.06493652e+00 -5.43010473e-01 -1.74367994e-01
2.88024008e-01 2.35583395e-01 1.68357223e-01 -5.04835904e-01
6.27034843e-01 8.11260715e-02 -1.58182621e+00 4.95797962e-01
1.07360649e+00 3.67790043e-01 6.66839242e-01 7.96525717e-01
-4.22734469e-01 6.15288675e-01 -8.29428017e-01 -8.12713444e-01
1.67895302e-01 -6.42399132e-01 -9.66712058e-01 -1.70261934e-01
4.37882811e-01 -1.69408798e-01 -7.17488527e-02 1.06584942e+00
5.24259567e-01 4.66967300e-02 4.48807895e-01 -5.01750648e-01
-4.70966667e-01 9.80734348e-01 -4.09749627e-01 2.90752947e-01
7.62376666e-01 -8.06553140e-02 1.09889913e+00 -7.31752098e-01
5.93930602e-01 1.27837718e+00 4.61427212e-01 3.03265095e-01
-1.05039907e+00 -3.67299914e-01 -3.25842798e-01 1.30742297e-01
-1.38833094e+00 -5.03827214e-01 4.40964967e-01 2.84791470e-01
1.81437862e+00 6.79894388e-01 7.12748289e-01 7.26644158e-01
5.35404027e-01 4.17640775e-01 8.42228830e-01 -1.14782190e+00
-5.92505373e-02 -1.09691471e-01 2.61148632e-01 6.12158835e-01
5.55240512e-01 -3.16293657e-01 -3.22506517e-01 2.42083520e-02
1.01184979e-01 -3.85539562e-01 -1.06149442e-01 4.05270576e-01
-8.20526123e-01 8.41635585e-01 -2.55133063e-01 6.20944083e-01
-2.75344342e-01 -7.32855499e-02 7.75144696e-01 5.79515040e-01
3.88663322e-01 6.05266333e-01 -1.40395433e-01 -5.02938747e-01
-1.29465020e+00 4.54870373e-01 9.26776409e-01 1.23291719e+00
6.26493514e-01 1.63082674e-01 -2.15342179e-01 8.20968211e-01
1.40991107e-01 4.10328120e-01 8.14776599e-01 -8.45481634e-01
1.03583992e+00 6.71896517e-01 -5.11279583e-01 -1.09023774e+00
-1.02614790e-01 -2.76417106e-01 -9.15039122e-01 -3.31026524e-01
-1.73522606e-01 6.72956035e-02 -7.87076116e-01 1.57727349e+00
-1.88994899e-01 -2.64251560e-01 2.57898003e-01 4.46721494e-01
8.54504764e-01 7.57825553e-01 -8.07809532e-02 -7.10300505e-01
1.28466225e+00 -9.82395470e-01 -8.61896873e-01 -1.61805317e-01
8.41064095e-01 -1.04940271e+00 1.27766919e+00 3.32771420e-01
-1.70972574e+00 -4.59491760e-01 -1.32153320e+00 -2.22631797e-01
-2.31835708e-01 -2.85700202e-01 2.34062955e-01 9.01013911e-01
-1.17725074e+00 8.11020255e-01 -5.50442934e-01 -3.63013387e-01
-9.12727192e-02 3.90040040e-01 -2.83723086e-01 -1.22171968e-01
-1.16913283e+00 1.10411680e+00 7.40154207e-01 -4.32737917e-01
-1.06619738e-01 -2.84791380e-01 -1.03488028e+00 2.73974478e-01
3.10832024e-01 -7.52602100e-01 1.34077775e+00 -5.69922447e-01
-1.42826796e+00 5.44715822e-01 -2.88607299e-01 -8.14042032e-01
7.64933154e-02 -1.15682572e-01 -3.87240827e-01 3.60778064e-01
1.44245597e-02 5.56787193e-01 5.00134706e-01 -6.81740344e-01
-4.98196244e-01 5.19618113e-03 2.15152483e-02 5.61065733e-01
-4.73736435e-01 9.88272652e-02 -4.79847580e-01 -8.14635038e-01
-1.33386627e-01 -6.57016695e-01 -1.30871475e-01 -9.55213606e-01
-4.28632379e-01 -1.19005814e-01 5.77856123e-01 -8.13785255e-01
1.95365858e+00 -1.78944194e+00 2.36445412e-01 -9.49136540e-02
-4.89246771e-02 5.99185109e-01 -3.74665648e-01 9.63885486e-01
4.03712243e-02 5.59287012e-01 -4.99201596e-01 -5.62950671e-01
-1.73542827e-01 2.67952502e-01 -1.60603046e-01 7.48014376e-02
-1.10355876e-02 8.68843317e-01 -6.62750423e-01 -9.13613975e-01
1.31301761e-01 1.80808187e-01 -6.78611755e-01 8.13871026e-02
-1.86298832e-01 -3.00693214e-01 3.10911145e-02 4.46545959e-01
4.99487251e-01 3.70619416e-01 2.23165110e-01 -9.35958326e-02
6.95404771e-04 6.86886191e-01 -9.07724679e-01 1.84644878e+00
-5.56513250e-01 3.76727372e-01 -1.90092593e-01 -9.45650220e-01
8.95426214e-01 3.04682255e-01 2.09867924e-01 -9.32196498e-01
1.32957265e-01 3.18565428e-01 1.80610921e-02 -5.37917435e-01
1.38173187e+00 -3.19163829e-01 -2.58619100e-01 6.45527959e-01
7.38822892e-02 -4.37993050e-01 9.66143906e-01 4.39698428e-01
1.10962737e+00 2.00669300e-02 6.06762886e-01 -4.00148109e-02
6.66609347e-01 -8.94935727e-02 4.46181029e-01 7.16579556e-01
1.07488632e-01 8.36438119e-01 3.65040451e-01 -8.46084207e-02
-1.42603731e+00 -5.65637767e-01 1.66845009e-01 6.83903515e-01
-1.39745161e-01 -1.07705104e+00 -1.03989255e+00 -4.51156110e-01
-4.62762982e-01 1.19568825e+00 1.27558067e-01 -2.28442922e-01
-1.09801388e+00 -5.99151075e-01 9.21697259e-01 2.82607317e-01
2.44682312e-01 -1.30714881e+00 -7.76858985e-01 3.22316527e-01
-6.04425907e-01 -1.06973767e+00 -5.01157224e-01 -7.50088412e-03
-1.01601422e+00 -7.96201766e-01 -3.89179349e-01 -7.98232377e-01
4.85297471e-01 2.81048059e-01 1.17242181e+00 3.93297315e-01
-1.19695090e-01 -4.58816066e-02 -7.00731575e-01 -2.77699381e-01
-1.03663337e+00 4.84623909e-01 -6.79582879e-02 -7.56801903e-01
3.80506158e-01 -6.65370345e-01 -1.51236624e-01 -3.56052071e-01
-1.20184684e+00 1.72903910e-01 7.29470372e-01 6.76127017e-01
5.12202978e-01 6.18728474e-02 6.32648110e-01 -9.97793913e-01
1.32919812e+00 -1.44744754e-01 -1.27905518e-01 4.74549174e-01
-9.88504767e-01 2.70891160e-01 9.86957014e-01 1.11458357e-02
-9.58842337e-01 -3.09286833e-01 -4.24372166e-01 -9.20689199e-03
-1.18737742e-01 8.44156981e-01 2.07634866e-02 3.31033260e-01
5.82243025e-01 4.68832821e-01 9.69797596e-02 -3.84332806e-01
4.42559034e-01 8.88868034e-01 6.11262679e-01 -4.70277637e-01
4.37504619e-01 -2.55778223e-01 3.85294706e-02 -8.01784158e-01
-3.31566423e-01 -2.54921138e-01 -5.38535237e-01 3.73500846e-02
5.26212871e-01 -6.23677909e-01 -1.52091876e-01 -5.30055016e-02
-1.37140679e+00 2.00997010e-01 -3.96597654e-01 3.77623498e-01
-6.08154356e-01 1.09085870e+00 -8.25735748e-01 -6.55922592e-01
-1.04456115e+00 -1.07185066e+00 8.64648402e-01 -6.15705550e-02
-6.62411451e-01 -5.84447801e-01 8.65241960e-02 3.72686505e-01
3.17317843e-01 2.56212093e-02 1.25508893e+00 -6.94916427e-01
-2.18219295e-01 -5.31521618e-01 7.61414170e-02 4.09516960e-01
1.13135345e-01 -2.76670307e-02 -3.24876875e-01 -4.58665311e-01
2.38445073e-01 -3.92317474e-01 6.40840530e-01 1.30767927e-01
7.74297833e-01 -6.37719572e-01 1.42093956e-01 3.54268640e-01
1.62593460e+00 2.13179410e-01 1.03198481e+00 3.40909392e-01
3.78525108e-01 4.38749462e-01 6.59497619e-01 4.03362364e-01
3.09999704e-01 4.32599366e-01 2.69849628e-01 3.65835130e-01
-3.65356445e-01 -3.10329437e-01 4.18849409e-01 1.74590826e+00
-4.90176491e-02 -6.76231802e-01 -6.39701188e-01 4.14053619e-01
-1.36739993e+00 -1.10313296e+00 -2.76673734e-01 2.16083145e+00
9.16253984e-01 4.26196814e-01 -2.74109449e-02 5.19645929e-01
6.42431080e-01 2.53120244e-01 1.44060746e-01 -1.25685227e+00
-1.33705527e-01 6.95220590e-01 4.09249604e-01 7.01921880e-01
-5.76330602e-01 9.79822457e-01 6.54852629e+00 1.01513493e+00
-8.12937498e-01 2.26298682e-02 3.66005182e-01 -2.30881959e-01
-6.13948107e-01 1.33505493e-01 -8.32358539e-01 3.55632663e-01
1.36731589e+00 -6.70161843e-01 3.71290028e-01 3.83870006e-01
1.58867151e-01 -3.31728280e-01 -7.33627975e-01 6.78245604e-01
3.84201288e-01 -1.35176957e+00 4.01643604e-01 -6.80410340e-02
3.29414755e-01 -2.26276189e-01 -3.00755858e-01 4.64197725e-01
-1.36786684e-01 -8.91761184e-01 7.13236868e-01 3.46419185e-01
1.01431930e+00 -1.07786322e+00 9.16939378e-01 6.41890645e-01
-1.09291971e+00 1.67440653e-01 -5.49678147e-01 -4.15560037e-01
5.32117248e-01 3.55499297e-01 -7.97389925e-01 9.90701377e-01
1.59737557e-01 3.82159293e-01 -7.56805837e-01 7.41178095e-01
-9.02518481e-02 5.54003596e-01 -2.78816074e-01 -3.49772483e-01
1.67474687e-01 -2.95494586e-01 7.13963449e-01 1.64892089e+00
5.27176440e-01 8.48944671e-03 -1.99840337e-01 4.52322274e-01
8.08782354e-02 5.81859648e-01 -8.30370247e-01 -1.22554608e-01
4.24419999e-01 1.04496634e+00 -6.76993251e-01 -4.20341104e-01
-4.77105111e-01 1.04330695e+00 4.60794419e-01 -6.81984276e-02
-5.47129333e-01 -6.28481388e-01 -1.25061363e-01 1.05224885e-01
2.07101136e-01 -2.64808565e-01 -4.99868721e-01 -9.90939438e-01
2.95114398e-01 -1.14440513e+00 2.54595190e-01 -4.42873776e-01
-6.54657066e-01 8.44775796e-01 4.03606951e-01 -1.16163170e+00
-7.14093387e-01 1.96756683e-02 -5.24388611e-01 8.47783744e-01
-1.13477933e+00 -9.11657333e-01 1.56739727e-01 7.84642473e-02
1.01651251e+00 -3.35602969e-01 1.07622540e+00 3.82470340e-01
-5.31989038e-01 7.99014747e-01 -2.82313049e-01 -3.90132219e-01
3.31254840e-01 -1.04803455e+00 4.72157657e-01 1.27911258e+00
1.47764668e-01 7.67513990e-01 8.14210773e-01 -8.44089210e-01
-1.26870179e+00 -1.00171435e+00 1.77110243e+00 2.02075914e-01
2.86115199e-01 -1.83942884e-01 -8.81921470e-01 2.89575249e-01
8.56344700e-01 -8.63641918e-01 6.84489250e-01 -2.13450164e-01
7.96053857e-02 1.78067796e-02 -1.08142412e+00 8.38064671e-01
9.29829299e-01 -3.81882310e-01 -9.23380613e-01 2.40349263e-01
1.14028442e+00 -3.74056429e-01 -6.57868505e-01 4.08506721e-01
1.60645410e-01 -1.13903320e+00 5.81835091e-01 -3.26549143e-01
9.70105171e-01 1.83458477e-01 -3.59172404e-01 -1.11035573e+00
-1.93662331e-01 -6.63984954e-01 -2.27046549e-01 1.37260270e+00
5.09434223e-01 -2.23888382e-01 6.42783642e-01 5.47860563e-01
-5.44808507e-01 -9.73267972e-01 -8.89521360e-01 -8.72224987e-01
2.06450999e-01 -4.87178922e-01 6.68349802e-01 4.25197601e-01
3.91751230e-01 7.66357899e-01 -4.07982111e-01 -5.50835013e-01
2.30584279e-01 1.99461952e-02 5.30545831e-01 -6.35994554e-01
-3.48763049e-01 -5.46369970e-01 -1.68064237e-01 -7.50538766e-01
2.38293409e-01 -1.20666885e+00 -9.19225067e-02 -1.57367516e+00
3.06834280e-01 3.83072533e-02 2.00420648e-01 1.74254611e-01
-5.90405427e-02 1.39274538e-01 4.26173776e-01 1.03686750e-01
-3.95479232e-01 5.56363285e-01 1.18322468e+00 2.16523036e-02
-1.58506751e-01 -1.50006518e-01 -8.07079852e-01 3.48950118e-01
1.04382420e+00 -5.69674134e-01 -6.23194456e-01 -5.96441746e-01
1.95352808e-01 2.70920217e-01 -3.71517718e-01 -9.92240131e-01
3.25919807e-01 -3.11698336e-02 -4.07324821e-01 -8.30026329e-01
1.27242461e-01 -5.69432378e-01 2.09180474e-01 6.52351558e-01
-3.95124495e-01 8.30700696e-01 9.58056897e-02 3.34345311e-01
-3.89343768e-01 -9.28300440e-01 6.08581483e-01 -4.59012240e-01
-1.97102636e-01 -1.93432689e-01 -5.63978672e-01 -7.51441903e-03
8.04312766e-01 -5.45127094e-01 -1.65684536e-01 -3.77281427e-01
-1.46983549e-01 -2.05319285e-01 3.75546634e-01 2.56361902e-01
8.53490651e-01 -1.12240386e+00 -8.54604959e-01 2.33721927e-01
-2.02975303e-01 -6.43562376e-02 -8.00862834e-02 5.35742939e-01
-1.05417109e+00 8.11079502e-01 -1.42476350e-01 -1.22763120e-01
-1.66235137e+00 4.73934889e-01 -1.35914564e-01 -6.13106549e-01
-8.42377186e-01 3.92409325e-01 -6.75032914e-01 -1.83722481e-01
7.99012259e-02 -3.54215264e-01 -3.15200686e-01 -6.38912395e-02
4.30805922e-01 5.22057772e-01 4.86961454e-01 -7.25698292e-01
-4.94684500e-04 2.28996471e-01 -3.28966528e-01 -4.65427905e-01
1.27568960e+00 -2.76850402e-01 -4.46280092e-01 1.55913293e-01
1.38125789e+00 4.13145982e-02 -2.36233577e-01 -5.37314899e-02
1.42804787e-01 -4.80100989e-01 -2.14915618e-01 -3.81371289e-01
-7.51451671e-01 6.47732913e-01 7.75281712e-02 3.58313352e-01
1.40100658e+00 -5.14600754e-01 1.24121106e+00 6.80783391e-01
2.02789575e-01 -1.31557906e+00 -1.43087596e-01 7.73421764e-01
8.31750393e-01 -6.38598502e-01 2.71690547e-01 -4.91302907e-01
-4.53891963e-01 1.33630586e+00 5.32120943e-01 -7.02616014e-03
2.01481119e-01 3.75738025e-01 -2.99895734e-01 -4.45853658e-02
-8.42368901e-01 -3.24271172e-02 7.51214428e-03 3.60347331e-01
7.51097798e-01 -1.24270879e-01 -1.45759344e+00 3.68205696e-01
-6.90481603e-01 -7.98288807e-02 8.79649222e-01 1.20994270e+00
-5.20000815e-01 -1.68511879e+00 -1.49987698e-01 7.09582090e-01
-7.74389267e-01 -6.03785992e-01 -3.22641134e-01 5.65280437e-01
-1.28471777e-01 1.14376390e+00 -8.80057588e-02 -6.05778813e-01
3.93613935e-01 1.57251582e-01 6.27586901e-01 -9.12820756e-01
-9.38999712e-01 1.19524084e-01 3.96147251e-01 -2.92247981e-01
-3.32181692e-01 -6.19945645e-01 -1.14979231e+00 -6.22738659e-01
-5.06934047e-01 2.13627934e-01 5.58578491e-01 9.73369896e-01
3.00617039e-01 4.43146676e-01 4.21867549e-01 -4.58272785e-01
-6.41942441e-01 -1.20594263e+00 -3.62133473e-01 3.81579101e-01
-2.47936510e-02 1.44402847e-01 -1.08758487e-01 9.50497761e-02] | [12.24853515625, 9.333260536193848] |
3a44f3b1-b799-44a8-86ca-61a39edf2023 | chatgpt-a-study-on-its-utility-for-ubiquitous | 2305.16837 | null | https://arxiv.org/abs/2305.16837v1 | https://arxiv.org/pdf/2305.16837v1.pdf | ChatGPT: A Study on its Utility for Ubiquitous Software Engineering Tasks | ChatGPT (Chat Generative Pre-trained Transformer) is a chatbot launched by OpenAI on November 30, 2022. OpenAI's GPT-3 family of large language models serve as the foundation for ChatGPT. ChatGPT is fine-tuned with both supervised and reinforcement learning techniques and has received widespread attention for its articulate responses across diverse domains of knowledge. In this study, we explore how ChatGPT can be used to help with common software engineering tasks. Many of the ubiquitous tasks covering the breadth of software engineering such as ambiguity resolution in software requirements, method name suggestion, test case prioritization, code review, log summarization can potentially be performed using ChatGPT. In this study, we explore fifteen common software engineering tasks using ChatGPT. We juxtapose and analyze ChatGPT's answers with the respective state of the art outputs (where available) and/or human expert ground truth. Our experiments suggest that for many tasks, ChatGPT does perform credibly and the response from it is detailed and often better than the human expert output or the state of the art output. However, for a few other tasks, ChatGPT in its present form provides incorrect answers and hence is not suited for such tasks. | ['Sourav Mazumdar', 'Ranjani H. G.', 'Giriprasad Sridhara'] | 2023-05-26 | null | null | null | null | ['chatbot', 'chatbot'] | ['methodology', 'natural-language-processing'] | [ 2.95554614e-03 5.21441519e-01 -1.98560059e-02 -2.25587338e-01
-1.18535542e+00 -7.72593796e-01 2.82360017e-01 -1.56740602e-02
2.07708567e-01 8.03613007e-01 3.37546259e-01 -7.15897858e-01
-1.88098475e-01 -4.82161611e-01 -1.42398775e-01 2.75363717e-02
4.70739514e-01 7.44435608e-01 1.08060129e-01 -5.88553369e-01
6.99494183e-01 -3.19401264e-01 -1.09317720e+00 9.44880247e-01
1.00162590e+00 5.26183069e-01 3.98723304e-01 8.55077505e-01
-5.29877663e-01 1.42047620e+00 -9.79344487e-01 -6.78011596e-01
1.75749455e-02 -4.12089974e-01 -1.33771920e+00 -2.80881912e-01
-2.44738981e-02 -8.50491673e-02 2.58345872e-01 7.82329500e-01
6.15278661e-01 -9.14099738e-02 3.76530975e-01 -1.55083025e+00
-8.37287426e-01 1.23014832e+00 -1.76872134e-01 1.20865434e-01
9.04596090e-01 3.66779685e-01 1.29735684e+00 -8.47194552e-01
6.41883612e-01 1.20012915e+00 8.40094030e-01 7.15483606e-01
-9.29566145e-01 -5.02540410e-01 3.16898674e-02 1.90333530e-01
-1.10614538e+00 -9.22496244e-02 5.38202047e-01 -6.73128128e-01
1.60849679e+00 4.52098936e-01 2.81688005e-01 1.10708964e+00
3.88438880e-01 9.08047080e-01 8.17418993e-01 -5.10239542e-01
1.87107146e-01 2.11408183e-01 -1.25645146e-01 5.25082111e-01
-3.86910230e-01 -4.53948468e-01 -4.65383321e-01 -6.02912903e-01
6.39121473e-01 -3.64713550e-01 6.86127916e-02 2.62039214e-01
-1.18610764e+00 8.54383171e-01 4.96206768e-02 3.76009107e-01
-3.42754781e-01 3.10489416e-01 4.07451421e-01 9.03746247e-01
4.67589527e-01 1.16963625e+00 -8.22491407e-01 -9.29347992e-01
-5.01639128e-01 3.66072595e-01 1.61893570e+00 1.36681485e+00
7.98096895e-01 -2.54930649e-02 -2.60068953e-01 9.60010946e-01
2.15610832e-01 6.73129931e-02 3.78595322e-01 -1.36659753e+00
8.23768020e-01 8.60926509e-01 2.65833050e-01 -7.27209151e-01
-7.07233623e-02 -6.86788559e-02 -1.12551421e-01 5.30839823e-02
3.10689270e-01 -6.98128700e-01 -3.54552239e-01 1.24656177e+00
-2.92555630e-01 -3.88954312e-01 1.65075511e-01 5.30112624e-01
9.97602284e-01 7.28109241e-01 4.33867052e-02 2.31161378e-02
1.10909319e+00 -1.10411823e+00 -6.47513330e-01 -6.72495186e-01
9.53798056e-01 -1.21803093e+00 1.21851254e+00 4.46937114e-01
-9.77378428e-01 -2.73916572e-01 -6.41693413e-01 1.29177406e-01
-1.42160490e-01 1.39508590e-01 5.82905591e-01 4.61800754e-01
-1.12358367e+00 4.78183240e-01 -4.96851563e-01 -5.02367020e-01
-7.08843544e-02 4.34297062e-02 1.21457011e-01 -3.04282486e-01
-1.24529421e+00 1.34475005e+00 8.98791291e-03 -2.26996109e-01
-5.55490017e-01 -7.20303893e-01 -6.89450860e-01 1.73089564e-01
9.57544327e-01 -5.39594352e-01 2.20314503e+00 -5.90099216e-01
-1.81566858e+00 5.41094482e-01 3.17477942e-01 -1.85794070e-01
4.88237143e-01 -3.10155600e-01 -1.98574141e-01 -2.98122764e-01
4.91667926e-01 4.55303788e-01 4.22263324e-01 -5.86911678e-01
-5.58551252e-01 3.47403735e-01 4.49208140e-01 7.76350796e-02
9.01211426e-02 5.25553882e-01 -7.66573846e-02 -3.13003123e-01
-5.17335415e-01 -8.68936181e-01 -3.87678415e-01 -5.05124509e-01
-2.62687713e-01 -8.24624121e-01 5.58217824e-01 -5.87382078e-01
1.39632058e+00 -1.79272902e+00 -8.56599733e-02 -1.85437068e-01
3.13772351e-01 6.91539943e-02 -1.97249889e-01 1.32128692e+00
1.69289768e-01 5.15894651e-01 -1.60477608e-02 2.58744329e-01
4.67127830e-01 1.84635535e-01 -1.39889508e-01 -3.05098593e-01
5.20911515e-01 1.11370838e+00 -1.12245798e+00 -4.65848505e-01
1.76021814e-01 -1.26827821e-01 -8.70685875e-01 3.45561832e-01
-7.24896789e-01 2.29965866e-01 -7.50545681e-01 6.06628239e-01
-5.71289510e-02 -6.28564954e-01 2.53409326e-01 4.66923892e-01
-1.61557510e-01 6.27466679e-01 -7.83483207e-01 1.68568063e+00
-8.46259356e-01 6.35743976e-01 -1.80380538e-01 -4.91247982e-01
1.23914707e+00 6.80927396e-01 9.68973413e-02 -3.78523707e-01
-8.35023727e-03 4.37831074e-01 3.73878866e-01 -1.00923800e+00
4.29037541e-01 1.85021926e-02 -5.61593354e-01 1.17755520e+00
2.57993668e-01 -6.78658247e-01 2.96175420e-01 1.97721407e-01
1.57657230e+00 9.42862779e-02 6.98027730e-01 -2.24181011e-01
3.59175026e-01 4.04417932e-01 3.39221299e-01 9.13299441e-01
5.88128567e-02 4.61525738e-01 8.01356673e-01 -4.45396096e-01
-7.83504605e-01 -3.63766998e-01 4.49810922e-01 1.55189514e+00
-3.39776725e-01 -9.05892313e-01 -7.04619646e-01 -8.43339980e-01
-1.91591874e-01 1.17475951e+00 -2.98039079e-01 -8.75205770e-02
-4.97881174e-01 -1.82801545e-01 6.71575606e-01 5.60197234e-01
3.38759422e-01 -1.70018780e+00 -7.13830352e-01 6.22612476e-01
-4.83899146e-01 -1.02270961e+00 -6.75380945e-01 3.05342972e-01
-4.00305539e-01 -1.13176775e+00 -5.56961179e-01 -7.85318673e-01
1.50025159e-01 -5.97785600e-02 1.66264606e+00 2.13005692e-01
5.03402110e-03 4.51305360e-01 -7.98092902e-01 -4.78441536e-01
-1.21392393e+00 3.57898325e-01 -6.01455808e-01 -7.22582340e-01
5.98783493e-01 -5.92755854e-01 -9.43968520e-02 5.87040544e-01
-3.37486804e-01 1.72652602e-01 7.68993556e-01 6.72001600e-01
-2.51825422e-01 -4.32943553e-01 9.90260005e-01 -1.19859588e+00
1.42622519e+00 -5.96211135e-01 -2.13242531e-01 5.45188785e-01
-6.21404767e-01 -3.18078953e-03 5.37159145e-01 -4.55360591e-01
-1.04842508e+00 -5.75159967e-01 -2.50634849e-01 7.33749494e-02
-1.53263852e-01 1.14503121e+00 1.22693129e-01 -1.27888434e-02
1.09017837e+00 -4.08341810e-02 -2.59260625e-01 -4.22742069e-01
1.83058977e-01 1.08059645e+00 2.27730066e-01 -9.76961017e-01
3.88847411e-01 -5.54974735e-01 -8.90371382e-01 -2.97623217e-01
-7.27511048e-01 -3.53542447e-01 4.12093215e-02 -2.24079639e-01
4.61013556e-01 -4.85722780e-01 -7.18436301e-01 -6.89392835e-02
-1.40505350e+00 -7.70315468e-01 -2.52072513e-01 8.04162994e-02
-7.77688265e-01 8.79736841e-02 -8.82340848e-01 -9.72321212e-01
-4.98394132e-01 -1.17021739e+00 8.59780312e-01 2.37889573e-01
-1.07144678e+00 -9.38235879e-01 1.10735528e-01 5.43013632e-01
7.59278655e-01 -4.98060696e-02 1.21096790e+00 -1.10671437e+00
-3.52066457e-01 -4.13488358e-01 1.95264909e-03 4.15025562e-01
2.35529557e-01 1.57961488e-01 -6.13891363e-01 2.31692895e-01
-6.96087927e-02 -6.09649837e-01 -2.62907427e-02 -2.97352802e-02
6.84414923e-01 -4.74390507e-01 7.25716427e-02 -2.29414970e-01
1.01535261e+00 5.16642690e-01 4.34532881e-01 4.12820369e-01
3.79465908e-01 8.00374329e-01 7.26970017e-01 5.52501321e-01
3.01429480e-01 4.47784722e-01 2.40012378e-01 4.72602874e-01
1.57678545e-01 -3.30286831e-01 3.67335588e-01 9.78808820e-01
-2.86498182e-02 -4.20733094e-01 -1.39347827e+00 5.23020566e-01
-2.13244271e+00 -7.15833247e-01 -1.56015400e-02 1.53051388e+00
1.03668559e+00 3.50087255e-01 -1.67150453e-01 -2.58720011e-01
6.51161849e-01 -1.13875523e-01 -4.32297587e-01 -9.80028689e-01
3.11798424e-01 2.41633251e-01 -2.21155137e-01 2.42804945e-01
-3.03058028e-01 8.11501265e-01 6.87075901e+00 7.30535984e-01
-5.84340572e-01 1.25057817e-01 2.29178071e-01 4.60229367e-01
-2.91688591e-01 2.52184361e-01 -3.44815195e-01 1.71930656e-01
9.49341774e-01 -7.76447713e-01 5.16993642e-01 1.18474174e+00
1.21208020e-02 -1.21469260e-03 -1.45462489e+00 8.12780261e-01
-1.22225992e-01 -1.55843127e+00 -1.73452958e-01 -3.17458034e-01
8.48997295e-01 1.70807610e-03 -3.02438378e-01 1.00727320e+00
9.16672528e-01 -1.04121089e+00 5.67905784e-01 2.59152681e-01
5.35865068e-01 -6.13613904e-01 1.20271027e+00 9.25586164e-01
-7.19522059e-01 -2.06495047e-01 -1.71633348e-01 -5.75243890e-01
1.22477084e-01 1.11097850e-01 -1.48423159e+00 5.24914205e-01
4.55617577e-01 5.99001110e-01 -4.81056839e-01 9.32112217e-01
-5.55730402e-01 7.29167283e-01 1.69948135e-02 -5.77800393e-01
3.13305676e-01 1.14070348e-01 4.45914507e-01 1.26686168e+00
3.47554475e-01 6.74277693e-02 5.73155344e-01 1.23219454e+00
-5.40176108e-02 -2.39720959e-02 -4.85759228e-01 -4.75596130e-01
6.44047141e-01 1.14363778e+00 -3.30095619e-01 -4.10389960e-01
-5.09751141e-01 4.13373023e-01 2.57718563e-03 3.79098803e-01
-5.24864554e-01 -5.91404796e-01 4.72884297e-01 -7.34484941e-02
2.10934713e-01 2.52244115e-01 -1.64024472e-01 -7.64111102e-01
-1.18386699e-03 -1.41895676e+00 3.29116881e-01 -1.35876215e+00
-1.58168435e+00 9.32676435e-01 3.28086037e-03 -1.34917498e+00
-9.34698820e-01 -3.50509495e-01 -9.72859919e-01 9.78385627e-01
-9.05021131e-01 -8.06948185e-01 -1.84882939e-01 2.47450262e-01
9.94262516e-01 -2.38349810e-01 1.09123886e+00 -5.41411228e-02
-2.37325475e-01 3.28477859e-01 -4.85030532e-01 2.14134663e-01
7.99438715e-01 -1.46722257e+00 8.13142538e-01 3.81218046e-01
-2.67107617e-02 7.94538438e-01 1.13086128e+00 -7.15552509e-01
-1.20962954e+00 -7.37651348e-01 1.09703004e+00 -7.85098374e-01
1.04661512e+00 -7.35237449e-02 -9.48647320e-01 8.54575157e-01
5.74626029e-01 -4.88913268e-01 6.67534769e-01 2.54624367e-01
-1.82975039e-01 3.48598927e-01 -9.09143627e-01 4.89966542e-01
6.72165453e-01 -7.48622775e-01 -1.04176140e+00 6.57560706e-01
8.64296973e-01 -7.53002822e-01 -9.10579979e-01 -2.54416242e-02
2.39395082e-01 -7.73215890e-01 2.21782386e-01 -7.77573287e-01
7.81406522e-01 -2.03900505e-02 -6.95324168e-02 -1.58264673e+00
-2.76859373e-01 -1.28662145e+00 1.48359329e-01 1.19877660e+00
8.29765320e-01 -6.10972106e-01 6.74826205e-01 8.34132910e-01
-5.37939608e-01 -7.51383007e-01 -5.90064108e-01 -6.12764060e-01
1.81122054e-03 -6.33844197e-01 5.41648448e-01 8.83579612e-01
7.27935910e-01 7.24161446e-01 -2.78401762e-01 -3.59152049e-01
2.67077275e-02 2.30270207e-01 8.82351160e-01 -1.25644946e+00
-4.73232269e-01 -4.42976981e-01 -1.72922406e-02 -1.00758445e+00
-1.54771386e-02 -8.39927793e-01 4.26666141e-01 -1.84928775e+00
-4.48536128e-02 -1.79901004e-01 1.49935946e-01 8.39649141e-01
-2.44454406e-02 -3.39835197e-01 2.63558477e-01 9.43996757e-02
-7.04360425e-01 -3.04739159e-02 1.25557578e+00 -2.31911600e-01
-4.78029788e-01 4.66580331e-01 -1.27801299e+00 7.06540108e-01
8.73389959e-01 -6.63288116e-01 -3.78774852e-01 -4.17612463e-01
8.68065357e-01 6.58734500e-01 1.57287315e-01 -7.78593063e-01
4.47218776e-01 -4.71467435e-01 -4.78420258e-01 -1.89786300e-01
-2.55731314e-01 -4.22146022e-01 2.11636752e-01 1.53467208e-01
-6.96146548e-01 3.23555112e-01 2.32603431e-01 2.20779970e-01
-3.16562414e-01 -6.11393511e-01 1.19912058e-01 -5.47041237e-01
-7.37608850e-01 -2.03095168e-01 -1.06072330e+00 4.00331557e-01
5.94931185e-01 -2.46304497e-01 -6.64668381e-01 -7.10002244e-01
-5.44139087e-01 5.38384378e-01 1.33524552e-01 6.72147870e-01
5.15682459e-01 -1.00134635e+00 -9.51474965e-01 -3.15730423e-01
4.64357197e-01 -1.32338539e-01 2.27380637e-02 9.01356161e-01
-3.39390010e-01 5.00536442e-01 -3.20712805e-01 -4.57557142e-01
-9.76329446e-01 2.43261188e-01 1.97158992e-01 -6.02803349e-01
-7.00161636e-01 8.12081456e-01 1.08762063e-01 -6.79106116e-01
7.04129273e-03 -5.51751137e-01 -1.50281444e-01 -1.93018958e-01
5.59666872e-01 1.50901914e-01 3.05809975e-01 2.22020283e-01
-3.39445889e-01 1.00691199e-01 -2.50680596e-01 -1.94489896e-01
1.52721846e+00 1.15914389e-01 -9.39052552e-02 7.16123939e-01
6.31072164e-01 -1.69126794e-01 -9.00824666e-01 -2.02382222e-01
4.88852650e-01 -2.74310231e-01 -4.79946673e-01 -1.45176721e+00
-5.49210131e-01 7.65106916e-01 -3.43885541e-01 6.83660865e-01
6.56726182e-01 2.77272522e-01 5.42575479e-01 8.41407776e-01
7.46756673e-01 -1.05704665e+00 4.11368638e-01 8.85372281e-01
1.38594556e+00 -1.06552947e+00 -3.67544264e-01 -3.80801648e-01
-1.19365346e+00 1.41640663e+00 1.04368222e+00 -1.10589735e-01
1.75830983e-02 5.15931249e-01 3.59710276e-01 -3.15664202e-01
-1.53262889e+00 1.97482690e-01 -1.73368473e-02 6.56481981e-01
7.12033033e-01 -3.59619930e-02 -9.85801071e-02 8.68760407e-01
-3.58232439e-01 2.20527351e-01 9.43376839e-01 1.33418572e+00
-4.87863094e-01 -1.14276791e+00 -2.52913952e-01 6.17571950e-01
-2.91787416e-01 -3.67033988e-01 -8.20800006e-01 7.59473622e-01
-2.33204201e-01 1.23628402e+00 -6.09868944e-01 -5.51483810e-01
5.62811136e-01 3.74709696e-01 2.08776504e-01 -1.40738297e+00
-1.23262119e+00 -2.38129515e-02 7.30381727e-01 -4.30089682e-01
-2.00444087e-01 -4.07328993e-01 -1.02253079e+00 -2.24718824e-01
-4.77502495e-01 7.34461010e-01 2.59420663e-01 1.16027832e+00
2.48865873e-01 6.64192200e-01 1.91739157e-01 -5.03068030e-01
-8.31093431e-01 -1.44646180e+00 -1.12728700e-01 -1.18885487e-01
-1.07412301e-01 -3.93358409e-01 -2.44544789e-01 -2.74369912e-03] | [11.966883659362793, 8.211235046386719] |
3718d3ea-65c8-42e7-ad65-00fd17f40560 | softctc-unicode-x2013-semi-supervised | 2212.02135 | null | https://arxiv.org/abs/2212.02135v2 | https://arxiv.org/pdf/2212.02135v2.pdf | SoftCTC -- Semi-Supervised Learning for Text Recognition using Soft Pseudo-Labels | This paper explores semi-supervised training for sequence tasks, such as Optical Character Recognition or Automatic Speech Recognition. We propose a novel loss function $\unicode{x2013}$ SoftCTC $\unicode{x2013}$ which is an extension of CTC allowing to consider multiple transcription variants at the same time. This allows to omit the confidence based filtering step which is otherwise a crucial component of pseudo-labeling approaches to semi-supervised learning. We demonstrate the effectiveness of our method on a challenging handwriting recognition task and conclude that SoftCTC matches the performance of a finely-tuned filtering based pipeline. We also evaluated SoftCTC in terms of computational efficiency, concluding that it is significantly more efficient than a na\"ive CTC-based approach for training on multiple transcription variants, and we make our GPU implementation public. | ['Michal Kula', 'Petr Buchal', 'Karel Beneš', 'Michal Hradiš', 'Martin Kišš'] | 2022-12-05 | null | null | null | null | ['handwriting-recognition'] | ['computer-vision'] | [ 5.57137907e-01 5.61709926e-02 1.74548626e-01 -5.38835943e-01
-1.19646060e+00 -8.33234429e-01 5.44741690e-01 2.20430531e-02
-8.43917727e-01 9.47269619e-01 -4.12597768e-02 -5.89861453e-01
9.19155255e-02 -4.28787857e-01 -8.24341536e-01 -7.99306273e-01
2.91559309e-01 5.93865633e-01 4.43751395e-01 2.20720410e-01
3.62463236e-01 5.29911101e-01 -1.44216490e+00 3.71233821e-01
5.72258711e-01 8.19741488e-01 2.59828925e-01 9.33605492e-01
-2.40490526e-01 5.76779366e-01 -5.91609120e-01 -4.95008707e-01
1.43664656e-02 -3.85854423e-01 -9.89908636e-01 4.38552856e-01
6.98037207e-01 -4.50766832e-02 1.19269781e-01 9.01797950e-01
5.79879403e-01 1.50609851e-01 5.39648116e-01 -5.44170499e-01
9.30366591e-02 4.94288236e-01 -4.51736122e-01 2.34640941e-01
1.56259432e-01 1.77547142e-01 8.66602421e-01 -7.79003084e-01
6.27195537e-01 8.01325858e-01 7.52933025e-01 3.36515278e-01
-1.17865193e+00 -3.52551609e-01 1.04334541e-01 -2.39460170e-02
-1.18882275e+00 -4.86986965e-01 2.31820643e-01 -3.70799005e-01
1.07594883e+00 3.85803044e-01 3.02589417e-01 9.83245373e-01
-1.63406402e-01 9.72161829e-01 1.43221986e+00 -7.61713088e-01
2.84491181e-01 -4.15613689e-02 4.67184722e-01 1.09884834e+00
-8.06405768e-02 -6.02325015e-02 -4.49270904e-01 -1.61304027e-01
6.48712635e-01 -4.99139726e-01 -1.02248155e-01 5.39747514e-02
-1.18225038e+00 6.32665098e-01 -4.70860690e-01 1.83097765e-01
1.72170877e-01 3.67543519e-01 4.88429964e-01 3.08598876e-01
3.96027327e-01 1.87644079e-01 -3.95964593e-01 -6.70607686e-01
-1.18634284e+00 -1.26377717e-01 9.28574085e-01 1.19394565e+00
7.88335204e-01 1.52220726e-01 -3.93145919e-01 8.97833407e-01
1.59936041e-01 3.97881955e-01 2.16737956e-01 -1.08615339e+00
2.90352553e-01 2.39215150e-01 9.66200158e-02 8.52977410e-02
-3.03950846e-01 -6.67858481e-01 -5.04368961e-01 1.56672493e-01
7.24176526e-01 -3.39547276e-01 -9.85444307e-01 1.28268039e+00
2.58988887e-02 2.34991401e-01 -3.05803746e-01 7.40161598e-01
3.62535655e-01 4.17716026e-01 -1.43627733e-01 -3.36376339e-01
1.21659124e+00 -1.05932748e+00 -5.93626916e-01 4.03516740e-02
6.07534170e-01 -1.07454646e+00 1.20759153e+00 8.17225635e-01
-1.00381172e+00 -4.96376097e-01 -1.13939452e+00 -6.44991025e-02
-2.42828056e-02 5.90516388e-01 7.96794534e-01 1.22438884e+00
-1.02097082e+00 8.22703362e-01 -1.20950711e+00 -1.72552511e-01
3.55924606e-01 4.99441743e-01 -1.07659064e-01 -6.95675611e-02
-5.63357294e-01 6.66061878e-01 2.84335166e-01 1.86974630e-02
-8.64361286e-01 -1.22214772e-01 -6.93150461e-01 -8.99558514e-02
4.64320928e-01 8.63052085e-02 1.27761149e+00 -7.13277042e-01
-1.75147021e+00 9.46845233e-01 -5.15359282e-01 -4.27038342e-01
6.98054552e-01 -8.07074159e-02 -2.74147063e-01 2.28232667e-01
-9.16884914e-02 5.01908422e-01 8.33701074e-01 -9.58751976e-01
-6.28676534e-01 -5.13228253e-02 -4.12617400e-02 -9.92278159e-02
-2.69760311e-01 1.09764256e-01 -9.03665245e-01 -7.85861254e-01
-1.47835270e-01 -1.16327286e+00 -9.22911391e-02 -1.22065187e-01
-4.23585892e-01 -2.56128341e-01 6.57971144e-01 -4.15632308e-01
9.24100637e-01 -1.98123407e+00 -7.52626508e-02 3.12242419e-01
-1.79650038e-01 6.11714482e-01 -7.03557730e-02 3.62354875e-01
1.35961473e-01 5.20211719e-02 -6.43593073e-01 -8.16348016e-01
9.76270661e-02 3.51273924e-01 -2.25555196e-01 4.71815258e-01
2.83821970e-01 6.85332596e-01 -8.52722347e-01 -4.92082328e-01
1.67932749e-01 4.56067204e-01 -2.28538677e-01 -1.56024843e-02
-5.80015063e-01 3.25907290e-01 -9.20198858e-02 6.75239325e-01
5.18102765e-01 -3.18226278e-01 3.24207813e-01 1.17281809e-01
-2.66596317e-01 2.56061584e-01 -1.31561363e+00 1.90734434e+00
-3.57830286e-01 6.67947710e-01 -3.23321931e-02 -8.22791755e-01
1.00676155e+00 4.24378037e-01 9.15726945e-02 -2.30076492e-01
1.34702057e-01 3.70096385e-01 -2.66334653e-01 -1.56483561e-01
4.52135265e-01 -1.39512867e-01 9.63500887e-02 7.81640947e-01
2.47711957e-01 -2.41767168e-01 4.42465067e-01 8.40609986e-03
1.32120752e+00 7.34532773e-01 -1.06468350e-01 -5.28380454e-01
6.50972009e-01 7.01054931e-02 3.62172544e-01 9.50032949e-01
-9.67812613e-02 8.95407319e-01 5.43699026e-01 -2.03938439e-01
-1.05074418e+00 -9.60483730e-01 -2.89867431e-01 1.17281306e+00
-3.19539696e-01 -4.92044032e-01 -9.14143384e-01 -9.32071567e-01
-3.59619468e-01 6.15657210e-01 -3.99426401e-01 3.25197488e-01
-6.54206872e-01 -9.31140602e-01 9.16535139e-01 5.63136458e-01
4.48185295e-01 -9.64318812e-01 -4.74064291e-01 2.54953176e-01
-5.94574474e-02 -1.24777055e+00 -6.19041324e-01 8.26797009e-01
-7.88452864e-01 -7.51259565e-01 -6.44862890e-01 -8.20836544e-01
6.02442026e-01 -2.12623864e-01 7.89596379e-01 1.55809999e-01
-3.93031090e-01 2.98424333e-01 -4.87414062e-01 -2.41283596e-01
-4.70101744e-01 2.61650950e-01 -1.00084476e-01 -1.01574846e-01
3.71483088e-01 -2.64277339e-01 -2.09696680e-01 2.65594661e-01
-8.32738578e-01 -1.72147080e-01 4.25464064e-01 9.95021641e-01
6.05901301e-01 -2.04883456e-01 1.96101040e-01 -1.21690214e+00
4.20800030e-01 3.09751391e-01 -7.18880057e-01 3.86255622e-01
-4.66824949e-01 3.76677752e-01 6.40139818e-01 -3.00397336e-01
-9.55514610e-01 5.11813939e-01 -6.80225015e-01 -1.24773182e-01
-3.02054137e-01 3.90423566e-01 1.03620745e-01 -1.35225698e-01
5.21186113e-01 4.78738427e-01 -2.58949352e-03 -6.75190508e-01
7.26113515e-03 7.26309836e-01 6.23153210e-01 -8.28645766e-01
5.47215998e-01 5.11741877e-01 -3.41252275e-02 -1.02580321e+00
-7.63183415e-01 -5.46014845e-01 -7.36622691e-01 2.31221765e-02
6.85370624e-01 -6.38855517e-01 -9.49313462e-01 5.34426153e-01
-1.20761538e+00 -5.18036187e-01 -2.63382196e-01 6.02065444e-01
-7.26305187e-01 8.81681204e-01 -9.02660668e-01 -9.72481966e-01
-2.12973088e-01 -1.17564487e+00 1.27312875e+00 -3.65140922e-02
-1.19132668e-01 -8.24747324e-01 2.70006619e-02 4.85021472e-01
1.11129940e-01 -7.70514905e-02 7.18433380e-01 -7.35774636e-01
-4.98696685e-01 -1.69538841e-01 -2.73955792e-01 5.53620875e-01
5.26780076e-02 1.63618714e-01 -1.27762902e+00 -4.85695839e-01
-1.26650497e-01 -6.05403423e-01 1.14013648e+00 4.31182161e-02
1.12929690e+00 1.22187145e-01 -3.84372175e-02 4.52879041e-01
1.33197188e+00 1.14831008e-01 9.08477426e-01 1.06858455e-01
5.73159635e-01 3.75960231e-01 6.13937676e-01 4.80531812e-01
2.46521309e-02 7.04904974e-01 7.24672824e-02 -5.10743037e-02
-1.62422508e-01 3.73295099e-02 3.98740411e-01 7.38671005e-01
-2.09606200e-01 -2.95901567e-01 -8.91049385e-01 4.97780293e-01
-1.69146931e+00 -7.93105721e-01 -3.13594043e-01 2.26759052e+00
1.07014787e+00 4.54007030e-01 1.07739680e-01 3.23172331e-01
6.60016954e-01 -6.30142987e-02 -2.51844138e-01 -6.68750584e-01
-2.20050558e-01 9.94022727e-01 6.35431886e-01 8.22332978e-01
-1.32637036e+00 1.14265680e+00 6.32445860e+00 1.17453206e+00
-1.12422347e+00 1.48605078e-01 4.49444383e-01 6.22099191e-02
-1.36510506e-01 2.33217608e-02 -1.02875948e+00 4.22403991e-01
9.95903969e-01 6.40751779e-01 2.35116273e-01 4.28162307e-01
5.21919839e-02 -4.45402712e-01 -1.08689475e+00 7.31444418e-01
1.19033948e-01 -1.40787911e+00 -2.35867634e-01 9.96274594e-03
5.07571042e-01 7.37927631e-02 -2.45860711e-01 7.73360431e-02
5.29506087e-01 -9.97123778e-01 6.70292258e-01 2.42665470e-01
9.10046637e-01 -5.65900385e-01 6.19597673e-01 3.76984000e-01
-1.11113179e+00 3.22088420e-01 -2.61620075e-01 1.73804667e-02
2.01596528e-01 6.66280866e-01 -9.13623154e-01 4.24370497e-01
3.66698086e-01 4.08863872e-01 -5.34206569e-01 1.03014290e+00
-4.06479567e-01 1.07828259e+00 -3.89574856e-01 -1.74693614e-01
3.76281261e-01 -1.32979855e-01 3.64599377e-01 1.87933242e+00
2.32809171e-01 -1.69510935e-02 1.46349043e-01 5.14061630e-01
1.05244018e-01 -5.58919497e-02 1.81568787e-02 -1.77239910e-01
2.24525139e-01 1.06795096e+00 -1.21250308e+00 -4.42207754e-01
-2.06581160e-01 1.30625045e+00 3.96000057e-01 5.65319359e-02
-7.30682790e-01 -4.70564634e-01 2.01423779e-01 -1.80151790e-01
7.96001434e-01 -5.55310547e-01 -2.95605689e-01 -1.16786027e+00
1.08664498e-01 -6.98585093e-01 1.35742605e-01 -3.84834528e-01
-8.59136164e-01 6.53490841e-01 -3.28098834e-01 -1.12552369e+00
-2.45043352e-01 -8.34664047e-01 -2.54521370e-01 9.06767368e-01
-1.45508611e+00 -9.96362746e-01 2.07767338e-02 4.64244992e-01
5.20577669e-01 -3.08078397e-02 1.04029047e+00 2.95628965e-01
-4.36947167e-01 7.49146819e-01 2.32068434e-01 2.18760386e-01
6.75647199e-01 -1.33429718e+00 4.24406618e-01 1.02513385e+00
3.66821349e-01 5.71720481e-01 5.73584139e-01 -6.09340370e-01
-1.28895473e+00 -9.50276017e-01 9.85506058e-01 -2.93322533e-01
3.60474288e-01 -6.30400479e-01 -7.90862858e-01 5.10297716e-01
2.52873331e-01 -3.78179625e-02 7.92813063e-01 -4.17071618e-02
-3.61011088e-01 1.44291013e-01 -1.05892420e+00 3.19061816e-01
1.14493120e+00 -7.20450282e-01 -3.21400046e-01 4.70989078e-01
4.58494097e-01 -4.28007483e-01 -8.39475393e-01 3.03894609e-01
5.36679268e-01 -9.78020132e-01 5.63364804e-01 -1.76310375e-01
1.42260507e-01 -3.45264077e-01 -1.30486488e-01 -7.43905485e-01
-2.64226198e-02 -7.71119058e-01 3.46063748e-02 1.04292870e+00
5.69777250e-01 -5.30429363e-01 1.11323476e+00 2.94261694e-01
-3.88195366e-01 -4.62665051e-01 -9.74127054e-01 -9.36349154e-01
-9.31602437e-03 -6.65886402e-01 9.00673717e-02 8.26922834e-01
-3.30165513e-02 4.16794010e-02 -4.63993371e-01 -2.25769710e-02
7.42287338e-01 2.10046738e-01 4.67732847e-01 -8.49932253e-01
-8.74057412e-01 -1.91663116e-01 -2.24968776e-01 -1.34629285e+00
4.83104475e-02 -8.24005663e-01 2.02016160e-01 -1.20116210e+00
1.51026413e-01 -6.35941565e-01 -7.17811212e-02 7.26229846e-01
5.34886308e-02 7.45823801e-01 7.91784972e-02 1.14934742e-02
-6.20963573e-01 1.76080778e-01 9.69711602e-01 -1.58015937e-01
3.76628116e-02 -3.07126474e-02 -2.44330585e-01 5.80276370e-01
6.19553030e-01 -3.63759428e-01 -1.39448196e-01 -5.64929485e-01
-4.18737642e-02 5.16382419e-02 1.99640945e-01 -9.54624772e-01
2.91329414e-01 5.05823903e-02 1.74917117e-01 -6.04958057e-01
4.19426829e-01 -4.67484236e-01 -9.60697308e-02 5.84882140e-01
-3.84895235e-01 -1.89543277e-01 3.37544322e-01 3.78218710e-01
7.86595105e-04 -6.79555476e-01 7.14576721e-01 -1.37601182e-01
-6.63706660e-01 -8.64998437e-03 -6.65914297e-01 -4.36315946e-02
7.13947892e-01 -3.81652147e-01 -2.69184917e-01 -1.72814727e-01
-8.67586851e-01 -8.60763341e-03 4.24513519e-01 -2.19289158e-02
3.01866740e-01 -7.47447133e-01 -6.79531634e-01 7.66038895e-02
-1.46748754e-03 -1.13527402e-01 -2.01877758e-01 7.51666486e-01
-7.31791973e-01 5.65200269e-01 -4.33950983e-02 -7.90636420e-01
-1.45177794e+00 7.90690780e-02 7.47980177e-02 -3.84947687e-01
-3.88086617e-01 9.95730221e-01 -5.64106941e-01 -4.95774806e-01
4.52561110e-01 -3.06197226e-01 1.41977757e-01 -1.41546875e-01
3.51069272e-01 3.94663960e-01 5.44346392e-01 -3.14802617e-01
-4.33719426e-01 5.10598838e-01 -2.17711434e-01 -5.29503345e-01
1.23358655e+00 1.24649771e-01 -4.39399108e-02 4.01872128e-01
1.13940287e+00 1.22921579e-01 -1.49435401e+00 -3.20178211e-01
4.12297934e-01 -3.33289385e-01 -1.51147038e-01 -1.08418405e+00
-7.41479695e-01 1.00420380e+00 6.33075953e-01 -3.37830484e-01
8.56112778e-01 -1.99849591e-01 5.25769472e-01 6.80239141e-01
6.05765700e-01 -1.28630602e+00 -1.64809637e-02 8.06027770e-01
3.69329691e-01 -1.01770508e+00 1.17542952e-01 -6.11973763e-01
-5.32738805e-01 1.22939634e+00 2.40655392e-01 -7.09824190e-02
3.64472210e-01 7.71893382e-01 -1.12064905e-01 1.44236058e-01
-6.47143304e-01 -4.13768351e-01 5.31713627e-02 4.46115166e-01
8.18531573e-01 -1.05327338e-01 -6.02856994e-01 1.60552159e-01
3.20699811e-02 1.99330866e-01 6.34976089e-01 1.34664893e+00
-3.02858740e-01 -1.73655427e+00 -1.61907583e-01 4.81346548e-01
-5.51499963e-01 -3.69637609e-01 -3.29456627e-01 4.57926005e-01
1.53876424e-01 9.46587026e-01 4.22489606e-02 -3.06169748e-01
-1.78117920e-02 3.73065442e-01 8.58831465e-01 -8.62900972e-01
-8.79425228e-01 4.39339757e-01 4.33372825e-01 -2.20805958e-01
-5.66200554e-01 -9.01494741e-01 -1.46060252e+00 -4.56800349e-02
-3.14829886e-01 2.80326575e-01 7.74405062e-01 1.18036926e+00
1.11284651e-01 3.04064602e-01 3.65419447e-01 -7.07221150e-01
-5.44179857e-01 -8.20332527e-01 -4.30001169e-01 1.35752022e-01
1.47818014e-01 -4.38779384e-01 -2.76907552e-02 2.52287269e-01] | [11.86205005645752, 2.75075364112854] |
94079529-7240-4365-8db3-aa4327535c78 | dspoint-dual-scale-point-cloud-recognition | 2111.10332 | null | https://arxiv.org/abs/2111.10332v4 | https://arxiv.org/pdf/2111.10332v4.pdf | DSPoint: Dual-scale Point Cloud Recognition with High-frequency Fusion | Point cloud processing is a challenging task due to its sparsity and irregularity. Prior works introduce delicate designs on either local feature aggregator or global geometric architecture, but few combine both advantages. We propose Dual-Scale Point Cloud Recognition with High-frequency Fusion (DSPoint) to extract local-global features by concurrently operating on voxels and points. We reverse the conventional design of applying convolution on voxels and attention to points. Specifically, we disentangle point features through channel dimension for dual-scale processing: one by point-wise convolution for fine-grained geometry parsing, the other by voxel-wise global attention for long-range structural exploration. We design a co-attention fusion module for feature alignment to blend local-global modalities, which conducts inter-scale cross-modality interaction by communicating high-frequency coordinates information. Experiments and ablations on widely-adopted ModelNet40, ShapeNet, and S3DIS demonstrate the state-of-the-art performance of our DSPoint. | ['Jianbo Shi', 'Kexue Fu', 'Xinben Gao', 'Ziyu Guo', 'Ziyao Zeng', 'Renrui Zhang'] | 2021-11-19 | null | null | null | null | ['3d-shape-retrieval', 'scene-segmentation', '3d-part-segmentation'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 4.52852212e-02 -3.05112749e-01 3.52887005e-01 -4.73335028e-01
-9.45522428e-01 -6.15362287e-01 5.85668445e-01 1.08650230e-01
-3.26394439e-01 -1.54211500e-03 8.37941915e-02 -2.13182315e-01
-2.00873315e-01 -1.01506054e+00 -9.14589882e-01 -3.96602929e-01
-2.01142907e-01 4.29988176e-01 1.65716290e-01 -2.38549218e-01
4.98034179e-01 1.15784752e+00 -1.83553922e+00 5.83362699e-01
8.47995639e-01 1.38664031e+00 1.74935162e-01 6.06540740e-01
-6.43058717e-01 -2.73247231e-02 -3.29859465e-01 -1.14482239e-01
5.25250256e-01 5.83270788e-01 -6.35282636e-01 -9.15669352e-02
8.66034687e-01 -1.89871266e-01 -1.53858423e-01 8.77659559e-01
4.38666016e-01 -7.42820576e-02 4.39336151e-01 -1.20355153e+00
-7.29210556e-01 2.16515765e-01 -9.11934733e-01 3.66383903e-02
2.83733249e-01 4.04544353e-01 9.32992339e-01 -1.31574595e+00
3.27542365e-01 1.44715500e+00 7.27611065e-01 1.25646114e-01
-1.07708335e+00 -8.81304801e-01 3.59744728e-01 -9.11793560e-02
-1.50863659e+00 -1.79165304e-01 1.03129506e+00 -5.19329071e-01
1.50630116e+00 4.81209695e-01 7.00518191e-01 7.31778920e-01
1.95358619e-01 3.64133924e-01 8.01347613e-01 -7.13173253e-03
1.65633827e-01 -5.97555697e-01 -1.26615530e-02 6.20706975e-01
5.84429801e-02 1.69926181e-01 -6.80743933e-01 -2.96873838e-01
1.39504540e+00 3.25100332e-01 8.01370852e-03 -3.00881088e-01
-1.54286087e+00 6.64526403e-01 9.65774357e-01 8.56185183e-02
-5.33894002e-01 4.18722779e-01 1.90803900e-01 1.99392110e-01
3.43831122e-01 6.42047703e-01 -7.39602089e-01 -4.73151356e-02
-8.76776278e-01 2.38694847e-01 3.03483784e-01 1.09507930e+00
8.95497978e-01 -1.91809341e-01 -2.21436098e-01 5.69084942e-01
4.70151275e-01 6.52793109e-01 1.47461131e-01 -7.04943061e-01
7.28609264e-01 8.84984910e-01 -2.92513102e-01 -1.04980934e+00
-5.16013324e-01 -5.24238229e-01 -6.95074022e-01 6.33783042e-01
-8.74232221e-03 4.23612259e-02 -1.15598047e+00 1.40467799e+00
5.17264664e-01 4.61029500e-01 -4.08022583e-01 1.16939759e+00
1.05765128e+00 3.80258024e-01 3.61276537e-01 6.02293193e-01
1.64916587e+00 -8.53120625e-01 -9.37535334e-03 -1.43108651e-01
3.44953477e-01 -6.75581932e-01 1.14575696e+00 2.57897943e-01
-1.12249660e+00 -6.97788358e-01 -1.07568526e+00 -5.80669641e-01
-5.65237582e-01 -5.77400587e-02 9.88954127e-01 3.64176124e-01
-8.46964300e-01 5.73093593e-01 -1.08053792e+00 2.88858134e-02
7.86573350e-01 7.69321382e-01 -4.36812967e-01 9.15803686e-02
-5.80326974e-01 2.33531296e-01 6.38241172e-02 1.32901460e-01
-1.81393459e-01 -1.24521744e+00 -8.59137297e-01 3.53048772e-01
1.32057175e-01 -1.12173367e+00 9.19411600e-01 -5.11317730e-01
-1.26526737e+00 7.53703058e-01 -1.19962595e-01 7.67970979e-02
2.06989631e-01 -3.80108476e-01 -2.37755656e-01 1.87318966e-01
2.18138069e-01 1.02326977e+00 8.81601095e-01 -1.19784236e+00
-6.62761092e-01 -8.31200540e-01 -8.16739798e-02 4.02951568e-01
-1.63067430e-01 -1.32717356e-01 -7.59585559e-01 -7.49417603e-01
6.66435599e-01 -4.66476291e-01 -3.40365559e-01 3.10884029e-01
-3.31915081e-01 -2.31369376e-01 1.02502048e+00 -3.95195007e-01
8.05145025e-01 -2.36851645e+00 -1.80272143e-02 5.40812969e-01
5.69530010e-01 -1.85738236e-01 -2.80694246e-01 1.56165838e-01
-2.69938052e-01 1.87702134e-01 -1.80923268e-01 -7.03755736e-01
-8.17633525e-04 -1.61193199e-02 -4.79785025e-01 4.43723977e-01
7.71781027e-01 1.25432467e+00 -7.40745485e-01 -3.55410367e-01
5.07946312e-01 7.22636759e-01 -9.28357124e-01 -2.19154656e-01
-1.12956449e-01 3.60618740e-01 -8.56309772e-01 1.20647252e+00
1.04930806e+00 -4.37258810e-01 -4.71976370e-01 -6.78812742e-01
-3.19919318e-01 9.83542427e-02 -1.15487194e+00 2.26905847e+00
-4.47388977e-01 1.89415559e-01 3.35666925e-01 -3.65107775e-01
8.42949271e-01 -1.89906731e-01 7.16388345e-01 -6.99150443e-01
2.33895957e-01 1.83497369e-01 -2.31744066e-01 2.35504806e-02
7.58283734e-01 1.12992212e-01 -2.35071555e-01 -3.56309190e-02
-2.91108750e-02 -3.92906666e-01 -5.53268671e-01 4.05333303e-02
1.13533640e+00 1.36608213e-01 -3.26027386e-02 -2.71523803e-01
2.63794631e-01 1.28698209e-02 1.39404848e-01 4.68574941e-01
2.52291471e-01 1.07643974e+00 6.30096421e-02 -4.16898280e-01
-7.50377715e-01 -1.27313161e+00 -2.62729108e-01 9.34236705e-01
3.18577915e-01 -6.16366327e-01 -3.31856757e-01 -4.70113993e-01
4.74721372e-01 2.10623294e-01 -5.14237106e-01 1.23800091e-01
-7.17776477e-01 -4.03313011e-01 2.28868544e-01 8.88596356e-01
4.10080194e-01 -9.34077740e-01 -8.51881742e-01 2.05094606e-01
3.93152922e-01 -1.12294829e+00 -4.99965847e-01 3.44695359e-01
-9.40708578e-01 -8.47615480e-01 -3.27206969e-01 -4.43519622e-01
6.79398477e-01 4.37532365e-01 1.24636471e+00 1.71843186e-01
-3.21934015e-01 4.02253181e-01 -2.41964683e-01 -2.85724610e-01
5.28974414e-01 2.75378585e-01 -1.97506219e-01 -1.81674272e-01
3.14027041e-01 -1.04272246e+00 -8.17077160e-01 1.81417972e-01
-7.28135824e-01 1.81296006e-01 8.93914878e-01 5.56641698e-01
1.00728071e+00 -3.74755949e-01 1.77574098e-01 -5.25457382e-01
3.34012866e-01 -4.37208563e-01 -6.33095801e-01 -1.28424829e-02
-1.10410199e-01 1.57913305e-02 3.26113492e-01 -1.52897641e-01
-6.07578218e-01 2.51424581e-01 -2.70849109e-01 -1.06247008e+00
-3.71877640e-01 4.15698111e-01 -3.73237431e-01 -5.31605005e-01
3.56259048e-01 -1.38362078e-02 -3.14894080e-01 -6.85496628e-01
6.34614825e-01 3.54789138e-01 7.99632788e-01 -1.11106634e+00
8.87682915e-01 8.09345484e-01 3.65878753e-02 -7.63699949e-01
-1.51178479e-01 -4.37072277e-01 -8.23419094e-01 1.24083869e-01
1.15883875e+00 -1.01213348e+00 -1.04457998e+00 4.32133436e-01
-1.39099753e+00 4.94940542e-02 -5.74686229e-01 2.21511066e-01
-4.22821641e-01 9.94004235e-02 -3.27996522e-01 -3.41923714e-01
-5.44790924e-01 -1.27545094e+00 1.97211635e+00 8.35284591e-02
-3.20084542e-02 -4.44748819e-01 -1.76741257e-01 1.19875841e-01
4.22362596e-01 3.59451592e-01 7.53014565e-01 -2.88025022e-01
-1.16086030e+00 -2.02411786e-02 -7.87204266e-01 -2.28831410e-01
-1.12081312e-01 -2.74316464e-02 -1.08104742e+00 -1.09666161e-01
-1.52945757e-01 -1.92062706e-02 6.76001370e-01 3.05435568e-01
1.77790511e+00 1.80921972e-01 -4.81559575e-01 1.35903990e+00
1.54506040e+00 -2.00307488e-01 4.89175618e-01 1.97378710e-01
1.36257541e+00 2.11490631e-01 2.17501685e-01 4.63818192e-01
5.68538010e-01 6.15432680e-01 7.46790349e-01 -3.63316834e-01
-1.70917273e-01 -1.17647760e-01 -1.55894160e-01 7.22670019e-01
-1.83816567e-01 2.10280687e-01 -1.11701834e+00 3.91167104e-01
-1.61229503e+00 -5.48321366e-01 -1.45996705e-01 1.78859174e+00
3.00116122e-01 3.96850407e-02 -2.16832876e-01 -2.82950759e-01
3.40252757e-01 1.42375141e-01 -6.13167465e-01 -1.82807803e-01
-1.23213656e-01 6.57558620e-01 7.61925399e-01 2.21073374e-01
-1.16589081e+00 8.98761034e-01 5.56766462e+00 1.09854877e+00
-1.55286431e+00 1.25256136e-01 5.74544013e-01 -4.30347294e-01
-4.53388691e-01 -1.68528825e-01 -6.07308626e-01 2.58543551e-01
3.75843704e-01 4.11834568e-01 2.55162716e-01 9.09302831e-01
-2.53838122e-01 2.56428659e-01 -1.15555358e+00 1.54955542e+00
-2.24156290e-01 -1.60516131e+00 2.34001745e-02 3.15962493e-01
5.03199041e-01 6.52410626e-01 1.78845564e-03 1.13843709e-01
8.16185176e-02 -1.16707134e+00 1.13988543e+00 4.97101545e-01
1.13648295e+00 -6.06995463e-01 1.47854432e-01 -2.08581612e-02
-1.77782142e+00 -3.02130841e-02 -2.66212542e-02 4.41238768e-02
2.16811568e-01 5.02055764e-01 -1.96573049e-01 7.99035609e-01
9.87118781e-01 6.67099893e-01 -5.91500759e-01 9.78494942e-01
3.12561989e-01 2.60218251e-02 -9.33975399e-01 3.62178355e-01
2.88509846e-01 -1.27401143e-01 5.07243812e-01 1.10143924e+00
4.37198550e-01 1.49183050e-01 3.39456886e-01 1.21456122e+00
3.27433273e-02 -2.00914100e-01 -3.24644059e-01 3.13387096e-01
6.75126791e-01 1.42128456e+00 -9.94681776e-01 -3.28122407e-01
-5.87747574e-01 9.35200632e-01 4.33403254e-01 1.78751498e-01
-7.32936621e-01 -2.93966532e-01 1.22686601e+00 1.79323494e-01
6.70378685e-01 -6.74493909e-01 -1.24332273e+00 -1.11360097e+00
2.71428823e-01 -4.36268866e-01 1.42566353e-01 -7.03787327e-01
-1.52918172e+00 7.53435552e-01 -2.40431111e-02 -1.32699752e+00
2.62946546e-01 -6.11455679e-01 -7.47046292e-01 1.21861231e+00
-1.66846967e+00 -1.61672461e+00 -5.60584962e-01 8.05460989e-01
4.74007100e-01 1.85864553e-01 6.17825508e-01 4.41570878e-01
-1.17438249e-01 5.53794086e-01 -4.07572240e-01 5.84049970e-02
2.02923670e-01 -9.85534430e-01 8.95952404e-01 5.48901200e-01
4.70477939e-02 7.41977870e-01 6.34721369e-02 -8.29923451e-01
-1.88381171e+00 -1.08036864e+00 3.13286692e-01 -5.79294026e-01
3.77726227e-01 -7.63953924e-01 -9.09735501e-01 3.31691086e-01
-2.63926357e-01 5.01448750e-01 2.95050323e-01 2.22133398e-01
-5.83518386e-01 1.38978601e-01 -1.25485051e+00 5.65623820e-01
1.53355122e+00 -6.98088944e-01 -3.88479292e-01 1.00087024e-01
1.13582301e+00 -7.29434848e-01 -9.61163044e-01 6.91736937e-01
5.31235218e-01 -7.67972589e-01 1.45809770e+00 -3.44577461e-01
4.73881423e-01 -4.53797758e-01 -4.73893255e-01 -8.70179653e-01
-6.43242538e-01 -3.68587375e-01 -1.02501765e-01 9.36536729e-01
8.93718973e-02 -5.65333188e-01 1.02290392e+00 6.12805128e-01
-6.81812465e-01 -1.09088814e+00 -1.13588464e+00 -4.99124050e-01
1.01100681e-02 -9.01963711e-01 1.38029408e+00 1.00798678e+00
-3.78294289e-01 1.40666142e-01 4.93090838e-01 6.61335707e-01
4.27762091e-01 4.15865958e-01 7.51907170e-01 -1.29792595e+00
-1.29111379e-01 -8.55747640e-01 -5.92763186e-01 -1.36887062e+00
-1.25564173e-01 -9.16133702e-01 -2.61599839e-01 -1.23227394e+00
-3.37108344e-01 -7.72704780e-01 -1.47742748e-01 5.87712884e-01
7.66654238e-02 3.87987643e-01 3.10904592e-01 2.37476617e-01
-3.37435097e-01 7.85157561e-01 1.29486799e+00 -6.48958459e-02
-3.43242317e-01 -6.92768037e-01 -8.62337530e-01 5.94322741e-01
3.74822706e-01 -1.13883786e-01 -2.13908002e-01 -1.01834691e+00
1.19641505e-01 -7.23060295e-02 7.73698092e-01 -1.36227524e+00
5.01450956e-01 -1.88136294e-01 6.43234849e-01 -1.08589303e+00
8.04053724e-01 -1.14680851e+00 1.80254698e-01 -2.71374099e-02
2.44620264e-01 4.34471518e-01 4.64809507e-01 4.39706057e-01
-1.61466762e-01 5.96106648e-01 3.12867880e-01 -9.45734158e-02
-8.93488944e-01 7.98555791e-01 4.24266636e-01 -4.76458251e-01
8.70671928e-01 -6.68192327e-01 -2.85878807e-01 2.86826193e-01
-6.86264038e-01 2.96111345e-01 6.62625432e-01 7.05878258e-01
9.49784696e-01 -1.38045621e+00 -5.11261582e-01 8.55974555e-01
7.97150880e-02 6.37848973e-01 6.08501315e-01 6.77567482e-01
-6.11810982e-01 2.35354662e-01 -2.67851055e-01 -1.00017023e+00
-1.05990028e+00 4.17030305e-01 1.28442004e-01 -7.28650833e-04
-9.64762926e-01 1.13419187e+00 5.16462803e-01 -5.25448084e-01
-1.11489393e-01 -7.45700419e-01 2.30142638e-01 -1.64752871e-01
2.80434251e-01 -3.96973314e-03 5.89300036e-01 -6.53937340e-01
-6.20779634e-01 1.15683138e+00 2.38347575e-02 3.79685536e-02
1.52252102e+00 5.28131910e-02 -1.94086999e-01 -1.12574026e-02
1.29672337e+00 4.55912873e-02 -1.33079207e+00 -1.27617210e-01
-5.07352233e-01 -8.07429194e-01 3.15994591e-01 -7.10157812e-01
-1.28324389e+00 7.79353380e-01 5.49681127e-01 -1.12781785e-01
1.05374873e+00 2.16086999e-01 8.58945191e-01 1.33797422e-01
5.38899004e-01 -7.37221956e-01 -2.85077006e-01 6.01953208e-01
1.21875322e+00 -1.13562608e+00 -3.43702398e-02 -9.29973364e-01
-1.02225199e-01 1.08288181e+00 8.60382557e-01 -4.99917358e-01
1.02120090e+00 4.83580321e-01 -2.65200257e-01 -8.80792737e-01
-5.29720426e-01 1.17545333e-02 6.58685684e-01 4.93410856e-01
1.66846603e-01 1.98169410e-01 2.37776175e-01 7.67758369e-01
-5.67203879e-01 -2.37950847e-01 -3.33564162e-01 1.16396153e+00
-5.17277494e-02 -9.25617933e-01 -6.73453450e-01 6.76195741e-01
1.30828889e-02 -2.93875188e-01 -1.49126336e-01 5.35000801e-01
3.42990845e-01 4.33366686e-01 7.16572464e-01 -7.34097719e-01
4.42893982e-01 -1.52526408e-01 4.50003982e-01 -6.42900467e-01
-9.56665576e-01 1.23354375e-01 -3.19516063e-01 -1.09631968e+00
-2.03690827e-02 -6.99044585e-01 -1.35902297e+00 -3.48780900e-01
-6.73911795e-02 -4.15687859e-01 1.06361091e+00 7.63348281e-01
1.15610123e+00 8.85893703e-01 3.93116832e-01 -1.65253794e+00
-2.14785486e-01 -5.98655701e-01 -3.12381476e-01 2.65307128e-01
4.33588803e-01 -7.33215034e-01 5.93175702e-02 -3.59305054e-01] | [7.867537975311279, -3.6124649047851562] |
4a629735-413b-4a2d-aa7f-517a355019f6 | deep-learning-based-group-wise-registration | 2306.10611 | null | https://arxiv.org/abs/2306.10611v1 | https://arxiv.org/pdf/2306.10611v1.pdf | Deep learning-based group-wise registration for longitudinal MRI analysis in glioma | Glioma growth may be quantified with longitudinal image registration. However, the large mass-effects and tissue changes across images pose an added challenge. Here, we propose a longitudinal, learning-based, and groupwise registration method for the accurate and unbiased registration of glioma MRI. We evaluate on a dataset from the Glioma Longitudinal AnalySiS consortium and compare it to classical registration methods. We achieve comparable Dice coefficients, with more detailed registrations, while significantly reducing the runtime to under a minute. The proposed methods may serve as an alternative to classical toolboxes, to provide further insight into glioma growth. | ['Bo Li', 'Esther Bron', 'Frans Vos', 'Roel Verhaak', 'Mathilde Kouwenhoven', 'Pim French', 'Bart Westerman', 'Pieter Wesseling', 'Marion Smits', 'Karin van Garderen', 'Claudia Chinea Hammecher'] | 2023-06-18 | null | null | null | null | ['image-registration'] | ['computer-vision'] | [ 1.13281958e-01 4.87358682e-03 -5.64666688e-02 -5.05317628e-01
-1.12057137e+00 -4.38429654e-01 7.86730111e-01 6.00940764e-01
-7.28161752e-01 4.86284673e-01 4.15056229e-01 -1.73493728e-01
-2.64577657e-01 -3.38618845e-01 -1.13907427e-01 -9.86963987e-01
-6.51950121e-01 5.04826427e-01 2.90430754e-01 -1.46342233e-01
2.33931005e-01 3.62956405e-01 -7.38257527e-01 -1.43329248e-01
1.10380232e+00 7.04588950e-01 2.48082533e-01 8.45644921e-02
3.67998034e-01 2.14604497e-01 -1.47157162e-01 -2.92459995e-01
1.25137880e-01 -1.04663156e-01 -9.62719500e-01 -4.42831106e-02
6.05791390e-01 -2.70147473e-01 -4.39544559e-01 1.27720702e+00
7.05327392e-01 7.48988017e-02 6.83928132e-01 -7.97940910e-01
-3.87729377e-01 5.62686980e-01 -7.79689729e-01 4.80157882e-01
2.33616903e-01 1.50599137e-01 3.23033482e-01 -6.26182675e-01
7.49146819e-01 4.57236558e-01 1.20107412e+00 4.42538381e-01
-1.20335865e+00 -6.52942002e-01 1.18185677e-01 -1.34146363e-01
-1.41908407e+00 -6.06667221e-01 2.99588501e-01 -8.70157599e-01
7.03820944e-01 3.53344947e-01 9.37433600e-01 7.68571377e-01
8.07317317e-01 2.43081212e-01 1.39968133e+00 1.05054244e-01
-2.86787655e-03 -7.12647557e-01 4.22871858e-01 7.08401799e-01
4.08519298e-01 1.82345629e-01 -2.08653465e-01 -1.59659952e-01
6.08150959e-01 2.36600451e-02 -6.20318294e-01 -9.06970277e-02
-1.69121337e+00 5.17133534e-01 6.32736802e-01 6.69719756e-01
-4.59998369e-01 2.52810687e-01 5.23413420e-01 1.06320314e-01
9.81791854e-01 1.91317767e-01 5.18557569e-03 -4.89358753e-02
-1.24637127e+00 7.52433166e-02 2.29539037e-01 8.36634636e-01
1.89775333e-01 -2.60896653e-01 -2.62362599e-01 5.74908257e-01
2.58674681e-01 1.78264096e-01 1.08722210e+00 -7.71112084e-01
1.90266684e-01 4.12831217e-01 -4.15619671e-01 -8.70436609e-01
-1.02666855e+00 -7.36586452e-01 -1.27640700e+00 4.75625582e-02
5.42737126e-01 1.43652424e-01 -7.77366519e-01 1.66867959e+00
1.15926132e-01 1.98643699e-01 -4.26073343e-01 7.01786339e-01
7.87548840e-01 -4.02432680e-01 -7.97157809e-02 -4.54876244e-01
1.40886629e+00 -8.95375609e-01 -9.52268243e-01 -7.86202103e-02
1.15269685e+00 -7.79437721e-01 6.05074286e-01 3.03656254e-02
-1.37553823e+00 2.59206444e-01 -6.82278454e-01 -1.40798921e-02
-1.43839464e-01 -1.26497313e-01 8.69774342e-01 6.72083437e-01
-1.58513451e+00 6.67246699e-01 -1.55182064e+00 -5.81706703e-01
6.48161054e-01 5.46980917e-01 -8.00263882e-01 -2.35115495e-02
-7.37397492e-01 1.09265888e+00 1.96367446e-02 1.70957133e-01
-7.72872806e-01 -1.22823608e+00 -8.97755563e-01 -5.63488841e-01
-4.43886250e-01 -7.74276078e-01 8.76482427e-01 -2.12955415e-01
-1.09075129e+00 1.30541015e+00 -3.36039752e-01 -5.26176453e-01
9.04034257e-01 2.95873672e-01 -1.39831319e-01 -1.02819249e-01
3.14694285e-01 7.43963778e-01 4.60725278e-02 -8.39809597e-01
1.57541819e-02 -1.02004969e+00 -6.35467291e-01 4.43574607e-01
-2.10093617e-01 3.11365873e-01 -2.86972463e-01 -8.08318257e-01
6.58413231e-01 -1.03551531e+00 -8.06082249e-01 2.69800872e-02
-1.14071280e-01 2.32361779e-01 1.21275067e-01 -1.25825298e+00
6.94711328e-01 -1.83012557e+00 5.14628505e-03 8.92913118e-02
7.42122471e-01 -4.09891605e-01 -3.40496562e-02 -2.95992136e-01
-3.20230722e-01 1.07678622e-01 -5.90473235e-01 -6.56897485e-01
-5.73454142e-01 -2.76105553e-01 4.20182943e-01 1.29133940e+00
-4.08465952e-01 1.18184602e+00 -1.11418486e+00 -2.32277766e-01
-1.51122659e-01 5.43623984e-01 -2.81492740e-01 -2.61100549e-02
8.06108892e-01 1.11366487e+00 2.65851337e-02 5.08454442e-01
7.99176872e-01 1.61080748e-01 -7.77827129e-02 -2.92095304e-01
-2.03728721e-01 1.70747057e-01 -3.63475859e-01 2.28752232e+00
-2.84976184e-01 7.71902442e-01 1.14489459e-01 -9.68110800e-01
9.43720698e-01 3.02304566e-01 1.13997889e+00 -5.22340298e-01
4.13818985e-01 3.63353193e-01 4.05695379e-01 -1.54263869e-01
4.46781218e-02 -2.97898173e-01 2.68600821e-01 3.92695278e-01
-7.68955350e-02 -2.88880527e-01 1.92770436e-01 -7.24450648e-02
1.35806370e+00 -2.45129704e-01 4.35969859e-01 -8.11372161e-01
3.02677423e-01 -1.35637730e-01 4.70154852e-01 3.78997356e-01
-6.59341514e-01 8.20428908e-01 2.08382457e-01 -3.25441837e-01
-5.66825449e-01 -1.01427317e+00 -5.63941896e-01 2.18495518e-01
6.77734241e-02 -1.98845431e-01 -9.23168778e-01 -6.22681975e-01
-1.76497534e-01 1.01432510e-01 -7.45840549e-01 -2.37184867e-01
-4.82343435e-01 -1.54551482e+00 8.16883385e-01 4.89878953e-01
3.08768630e-01 -2.67652839e-01 -1.32360727e-01 1.10455841e-01
-3.47833663e-01 -1.14380276e+00 -9.74725783e-01 -2.45430060e-02
-1.48758662e+00 -1.08203042e+00 -1.25703704e+00 -9.09469008e-01
1.25190711e+00 1.58930987e-01 8.37622583e-01 2.19571800e-03
-4.12988186e-01 1.90579727e-01 -3.82120162e-02 -8.54880437e-02
-2.60415524e-01 2.55550593e-01 3.15837920e-01 -2.69893020e-01
5.81642985e-02 -8.19592953e-01 -8.89760792e-01 4.63855982e-01
-5.27677596e-01 6.49233684e-02 4.44718927e-01 8.94650102e-01
7.64845788e-01 -1.29146934e-01 5.42116106e-01 -7.98139274e-01
8.59107971e-01 -4.00181651e-01 -6.34454846e-01 -4.57005948e-02
-1.04116058e+00 -2.28569686e-01 -2.55119558e-02 -2.55531102e-01
-7.80523956e-01 -5.34325950e-02 -1.10932790e-01 1.60344079e-01
2.53555719e-02 7.00682580e-01 2.81538546e-01 -9.29104149e-01
6.61000013e-01 2.44729698e-01 7.11167276e-01 -2.64962167e-01
1.42280713e-01 1.48967966e-01 6.33832216e-01 -2.47487515e-01
6.50706828e-01 7.64858186e-01 2.85480022e-01 -3.38443428e-01
-3.24980617e-01 -8.91822040e-01 -1.17901480e+00 -2.32782662e-01
6.91358387e-01 -1.06021369e+00 -3.80126357e-01 7.37948775e-01
-9.99734938e-01 -5.61994493e-01 -1.23083822e-01 8.41310322e-01
-6.52144432e-01 4.15882945e-01 -6.83232188e-01 -5.41227125e-03
-8.96292627e-01 -1.62793744e+00 1.01180077e+00 -1.86185941e-01
-4.00962114e-01 -1.52878511e+00 3.68541986e-01 2.80071050e-01
7.70444453e-01 5.82448900e-01 3.57864082e-01 -5.97831309e-01
-1.44006923e-01 -5.55431426e-01 -2.08220124e-01 -1.16807468e-01
5.97226739e-01 -4.68692869e-01 -4.88723367e-01 -6.51528895e-01
2.01163262e-01 2.36763746e-01 8.21503639e-01 7.45411217e-01
1.14071560e+00 -1.48167193e-01 -5.91455340e-01 1.09234726e+00
1.15870821e+00 5.42099401e-02 6.10220134e-01 2.81486243e-01
7.17527807e-01 7.32517719e-01 2.96156704e-01 -7.55441040e-02
5.55660486e-01 8.21372926e-01 1.76375061e-01 -3.92285883e-01
-6.28634632e-01 3.32043856e-01 5.30388812e-03 1.56795764e+00
-6.75957501e-01 5.50129056e-01 -1.35132658e+00 6.46062493e-01
-1.54391217e+00 -6.45295680e-01 -6.52160883e-01 2.37550282e+00
1.03331125e+00 -2.99472779e-01 -9.63950083e-02 -2.23338753e-01
5.73982298e-01 3.74847502e-02 -2.22932503e-01 3.45204949e-01
-1.39195368e-01 1.95725337e-02 1.02846658e+00 7.18263805e-01
-1.00422132e+00 6.02876484e-01 8.13967323e+00 5.82772315e-01
-1.09596229e+00 8.31467509e-01 7.23517716e-01 4.74387184e-02
-1.95696712e-01 -2.77700555e-02 -4.43209440e-01 1.27487123e-01
7.96021402e-01 -5.60521066e-01 4.14453030e-01 1.16118498e-01
5.71424425e-01 -2.04938635e-01 -1.04583573e+00 9.10116434e-01
1.98743194e-01 -1.31595361e+00 -3.68071973e-01 4.95392233e-01
9.77121711e-01 6.44836843e-01 -4.74170707e-02 -3.81000042e-01
8.71280655e-02 -1.04310668e+00 1.89418137e-01 7.23821998e-01
1.03356504e+00 -2.98698097e-01 1.03716612e+00 -4.88319807e-02
-1.21635175e+00 5.94008088e-01 -2.13830918e-01 1.51718467e-01
2.18605578e-01 8.94931018e-01 -7.64261305e-01 7.27181911e-01
3.80296797e-01 1.19413745e+00 -9.65842605e-01 1.59649611e+00
6.12748750e-02 3.38672161e-01 9.02390108e-02 4.88847643e-01
-1.26588210e-01 -5.42782128e-01 5.63144207e-01 1.10527241e+00
5.63289821e-01 1.62719432e-02 5.64316399e-02 7.24911451e-01
3.44680175e-02 3.97593200e-01 -6.86989129e-01 4.33015436e-01
-9.52965673e-03 1.48813593e+00 -1.05065572e+00 7.24823624e-02
-4.77508277e-01 8.78077805e-01 3.00238878e-01 1.69316888e-01
-4.65553999e-01 -4.32956219e-02 4.50724483e-01 3.05084944e-01
-7.41107225e-01 -5.03835201e-01 -7.00267076e-01 -1.19689918e+00
-3.88424993e-02 -6.74395204e-01 2.62717992e-01 -3.61768186e-01
-1.22436666e+00 6.46969020e-01 -3.67980488e-02 -1.12643421e+00
7.64918327e-03 -4.02691126e-01 -7.34058499e-01 1.03816986e+00
-1.32936049e+00 -1.01509809e+00 -7.98544049e-01 4.98640925e-01
1.96694165e-01 -1.77689552e-01 9.55797851e-01 4.72685248e-01
-6.18220031e-01 7.43121147e-01 1.01109715e-02 1.33907855e-01
8.41822207e-01 -1.35758209e+00 4.34618413e-01 8.25265408e-01
-2.09069982e-01 7.64069796e-01 3.65496874e-01 -7.12546110e-01
-9.90621150e-01 -1.36360919e+00 7.57600188e-01 -5.54945409e-01
1.10672057e+00 -1.55380383e-01 -9.27082956e-01 8.72582793e-01
3.24561089e-01 2.74483591e-01 8.82129848e-01 -2.13864143e-03
3.47703546e-02 -2.11309239e-01 -1.17877614e+00 5.49920201e-01
1.27299178e+00 -1.94841132e-01 -4.51059222e-01 6.81432068e-01
4.95444089e-01 -6.93655491e-01 -1.63410962e+00 5.77865720e-01
4.76164311e-01 -4.68648016e-01 9.03155804e-01 -1.67289570e-01
-1.34200573e-01 -7.13337511e-02 2.59738147e-01 -1.71442544e+00
-4.11507040e-01 -7.70585120e-01 4.87107128e-01 1.04207206e+00
3.43444705e-01 -1.08280277e+00 5.88238955e-01 6.78995848e-01
-6.09194815e-01 -7.11110830e-01 -1.39070189e+00 -1.03724313e+00
4.90838081e-01 -2.46021867e-01 4.39738333e-01 1.14286160e+00
4.45071906e-01 -4.64725703e-01 2.20515266e-01 1.15595967e-01
8.86953890e-01 -5.14835775e-01 2.14589521e-01 -1.17667437e+00
3.27404171e-01 -9.98169720e-01 -8.94665420e-01 -3.77901763e-01
3.47772837e-01 -1.69920373e+00 6.36209249e-02 -1.56811011e+00
4.97314572e-01 -5.75574338e-01 -3.17407280e-01 4.15616572e-01
-9.80154052e-02 6.30059302e-01 -4.53925967e-01 5.35526872e-01
-1.16457149e-01 4.41187054e-01 1.45606685e+00 -3.79070163e-01
-1.24603789e-02 -9.00107548e-02 -5.84833384e-01 7.20942378e-01
8.79557967e-01 -4.70138103e-01 -1.33830726e-01 -4.56031144e-01
-2.61533678e-01 -1.36532813e-01 2.49007672e-01 -9.98283148e-01
5.63796818e-01 1.53565601e-01 1.55138239e-01 -2.24317253e-01
-1.26963094e-01 -5.78184485e-01 2.37350881e-01 8.29172432e-01
-1.94694489e-01 4.93425071e-01 2.28896186e-01 1.72923341e-01
-3.77855033e-01 -3.63813527e-02 7.05259800e-01 4.67051156e-02
5.77498600e-02 8.95631850e-01 -3.33444625e-01 -2.86381431e-02
9.17669237e-01 -5.55707887e-02 -4.26829755e-01 -1.53765753e-01
-9.58683610e-01 1.64990276e-01 5.87361097e-01 -2.25748625e-02
4.88487244e-01 -1.64814901e+00 -9.29350257e-01 6.06746338e-02
2.44817004e-01 1.07700072e-01 2.36341074e-01 2.19196582e+00
-6.93233252e-01 4.78939056e-01 -2.65537769e-01 -7.17351377e-01
-1.23358297e+00 1.73718974e-01 9.01583076e-01 -6.00310266e-01
-7.77800918e-01 6.82439506e-01 3.17027986e-01 -5.38324893e-01
-9.09068361e-02 -3.81746292e-01 -2.92783529e-01 1.49746686e-01
5.46701193e-01 4.48210448e-01 6.41969085e-01 -9.00894940e-01
-5.36519945e-01 8.08660209e-01 1.60515867e-02 -8.46069306e-02
1.18853402e+00 -2.01681852e-01 -6.75591171e-01 2.97342062e-01
1.19790363e+00 -6.17263578e-02 -9.40010250e-01 -1.45570740e-01
1.04866445e-01 -2.80442297e-01 5.89105785e-01 -5.66816032e-01
-1.44026542e+00 5.53919256e-01 1.07151699e+00 -1.18794441e-01
1.14458191e+00 3.74459964e-03 5.23939610e-01 -1.99080110e-01
6.57770753e-01 -4.41980273e-01 -4.45591807e-01 2.75609881e-01
1.19922936e+00 -1.34342849e+00 5.94736673e-02 -6.27236843e-01
-1.33748800e-01 9.41224933e-01 2.37916932e-01 -9.64289755e-02
7.87868261e-01 5.97663343e-01 2.18067169e-01 -4.66687858e-01
-2.75741994e-01 2.07969695e-02 7.63674855e-01 1.05610836e+00
9.62868154e-01 2.39594802e-01 -8.96874130e-01 4.57103342e-01
-4.29326206e-01 -1.23996377e-01 3.93089324e-01 3.66692632e-01
7.73138851e-02 -1.13717639e+00 -1.80043757e-01 5.98561943e-01
-4.29206252e-01 -3.69510204e-01 -4.79853675e-02 7.14258075e-01
-2.57497132e-01 5.59362173e-01 1.87715843e-01 5.58696175e-03
2.33015224e-01 -2.62634128e-01 7.15403795e-01 -5.28704882e-01
-5.72538316e-01 1.12882346e-01 -8.18659216e-02 -7.57210374e-01
-6.63523912e-01 -1.18902075e+00 -1.14082623e+00 6.47071982e-03
-2.94300348e-01 -2.50901967e-01 6.68941259e-01 9.35165823e-01
1.96332574e-01 6.13972306e-01 5.81311107e-01 -8.08094084e-01
-6.21285476e-02 -1.18312311e+00 -4.60872710e-01 4.28400725e-01
-6.44577444e-02 -7.10810602e-01 -3.51630837e-01 2.49754861e-02] | [14.000199317932129, -2.563905715942383] |
7157a9db-acf5-428b-97f6-0baf7bed0468 | team-hub-lt-edi-eacl2021-hope-speech | null | null | https://aclanthology.org/2021.ltedi-1.17 | https://aclanthology.org/2021.ltedi-1.17.pdf | TEAM HUB@LT-EDI-EACL2021: Hope Speech Detection Based On Pre-trained Language Model | This article introduces the system description of TEAM_HUB team participating in LT-EDI 2021: Hope Speech Detection. This shared task is the first task related to the desired voice detection. The data set in the shared task consists of three different languages (English, Tamil, and Malayalam). The task type is text classification. Based on the analysis and understanding of the task description and data set, we designed a system based on a pre-trained language model to complete this shared task. In this system, we use methods and models that combine the XLM-RoBERTa pre-trained language model and the Tf-Idf algorithm. In the final result ranking announced by the task organizer, our system obtained F1 scores of 0.93, 0.84, 0.59 on the English dataset, Malayalam dataset, and Tamil dataset. Our submission results are ranked 1, 2, and 3 respectively. | ['Yang Bai', 'Bo Huang'] | null | null | null | null | eacl-ltedi-2021-4 | ['hope-speech-detection'] | ['natural-language-processing'] | [-3.59710127e-01 -2.52902687e-01 -2.01662555e-01 -4.08874661e-01
-1.22754455e+00 -4.72906530e-01 7.91783035e-01 -1.20825365e-01
-5.06064236e-01 5.57459235e-01 6.88920557e-01 -5.23332655e-01
-7.96311051e-02 -1.26523525e-02 2.21531522e-02 -2.77775496e-01
4.02852520e-02 6.90393388e-01 1.30541101e-01 -3.92274857e-01
3.24647933e-01 -1.29941842e-02 -1.45092475e+00 1.12891197e+00
7.49431968e-01 8.01008940e-01 5.77904522e-01 9.89691436e-01
-2.43697256e-01 1.16722047e+00 -8.21916938e-01 -1.62737782e-03
-3.09110314e-01 -1.59950301e-01 -1.18064868e+00 -3.59259725e-01
1.78862333e-01 2.22920142e-02 -2.97976822e-01 6.01317585e-01
9.33023512e-01 2.63385683e-01 5.16582727e-01 -1.14562500e+00
-3.49129111e-01 1.17400324e+00 -5.28386608e-02 1.55321673e-01
5.55200040e-01 -2.46767849e-01 1.10908735e+00 -1.43660104e+00
5.56302905e-01 1.35770023e+00 5.66784203e-01 7.35955179e-01
-6.01424932e-01 -5.98327339e-01 -5.35183074e-03 1.51235715e-01
-1.70206475e+00 -9.63333547e-01 3.63513410e-01 -5.65167069e-01
1.41372395e+00 6.89808607e-01 5.87886907e-02 1.13930464e+00
5.42054437e-02 1.04991746e+00 9.92640555e-01 -7.34909117e-01
2.90739778e-02 6.39391243e-01 2.16272369e-01 4.56787348e-01
-3.65610808e-01 -2.73857266e-01 -1.15777206e+00 -1.51185900e-01
-2.67621100e-01 -6.20219231e-01 -2.93522179e-01 5.32326579e-01
-1.31677210e+00 6.04173422e-01 -3.82523626e-01 7.70226777e-01
-8.07024613e-02 -4.12901878e-01 8.15700591e-01 3.45395625e-01
9.22909141e-01 2.09001794e-01 -8.61838698e-01 -3.65058273e-01
-1.26223016e+00 9.89399478e-02 1.03445780e+00 1.00440359e+00
4.76959720e-02 -1.17512411e-02 -4.44266677e-01 1.39428008e+00
5.24957120e-01 5.40410638e-01 8.55211914e-01 -4.52206522e-01
7.71319568e-01 2.82761365e-01 -4.53224480e-02 -3.56525987e-01
-5.45124054e-01 -5.30056000e-01 -5.29866099e-01 -3.42897832e-01
1.55294448e-01 -3.06882769e-01 -9.34119880e-01 1.29470909e+00
-1.67641878e-01 -5.22248566e-01 3.72092038e-01 5.21087587e-01
1.33759713e+00 6.82712555e-01 1.18836258e-02 -3.22276056e-01
1.28859997e+00 -1.35351551e+00 -1.13161767e+00 -1.33896060e-02
1.02030873e+00 -1.33129847e+00 1.09350550e+00 4.95722979e-01
-9.78026390e-01 -7.07157314e-01 -6.30869567e-01 2.40742322e-02
-4.88076657e-01 7.50481725e-01 2.33990908e-01 8.89619350e-01
-1.10045290e+00 7.14085251e-02 -3.57108146e-01 -7.14518607e-01
-2.72696882e-01 2.27234945e-01 -1.88162833e-01 4.13130701e-01
-1.13922238e+00 7.04769135e-01 3.23369861e-01 -2.91923583e-01
-1.06586826e+00 -5.70747793e-01 -3.48931283e-01 -1.47528440e-01
2.13680565e-01 -2.44316444e-01 1.63085616e+00 -3.32025439e-01
-1.51455283e+00 9.75589454e-01 -3.97127867e-01 -4.73588228e-01
4.21415269e-01 -3.37238640e-01 -8.85832608e-01 -3.77221167e-01
2.65143842e-01 1.85373917e-01 5.77048779e-01 -8.49057674e-01
-1.03362787e+00 -3.06097746e-01 -5.87572217e-01 2.68570393e-01
-4.72281009e-01 8.03708673e-01 -4.98016924e-01 -5.56764424e-01
-1.19026385e-01 -6.51230276e-01 8.49023759e-02 -1.09949231e+00
-8.00185382e-01 -6.39443457e-01 9.41604555e-01 -1.03814054e+00
1.67888033e+00 -2.45334816e+00 -2.15002686e-01 -1.15455709e-01
1.07175395e-01 1.99411780e-01 8.06910098e-02 7.24913597e-01
-8.29921141e-02 4.92784619e-01 3.69026929e-01 -7.29449749e-01
3.05446684e-02 -2.25050315e-01 -3.53670925e-01 2.52144691e-02
-2.82556504e-01 3.67593020e-01 -7.09600091e-01 -4.24713731e-01
2.10474938e-01 1.22723885e-01 -1.03829846e-01 4.25396025e-01
6.26253933e-02 3.83057028e-01 -1.90164179e-01 4.96297300e-01
3.26033413e-01 2.06664175e-01 9.31506604e-02 -2.25345314e-01
-7.11740851e-01 1.06808197e+00 -9.52335238e-01 1.57749224e+00
-6.39116526e-01 7.79301167e-01 2.57667392e-01 -6.51411653e-01
1.03342950e+00 9.48374987e-01 4.69259053e-01 -4.23882127e-01
2.57493462e-02 6.56552672e-01 1.22234955e-01 -5.77728689e-01
6.45847678e-01 2.54962683e-01 -1.58142969e-01 2.10581481e-01
3.33308458e-01 -2.29956746e-01 2.40785852e-01 2.29969069e-01
9.51290011e-01 -2.86527455e-01 1.68343812e-01 -4.90153372e-01
9.36893940e-01 -9.88008156e-02 4.00111735e-01 9.66225684e-01
-2.54527837e-01 6.60774946e-01 -1.64835416e-02 -3.85467768e-01
-8.49435449e-01 -6.30506575e-01 -2.42368922e-01 1.54972899e+00
-7.79756725e-01 -1.19355845e+00 -7.05939353e-01 -6.64732754e-01
-3.03022504e-01 8.25397432e-01 -1.50503829e-01 1.92032848e-02
-3.58194947e-01 -5.03048539e-01 1.09453928e+00 1.11068912e-01
6.55752599e-01 -9.77903247e-01 1.25461474e-01 2.95502752e-01
-7.33403862e-01 -1.35679042e+00 -7.96578944e-01 3.92656565e-01
-2.67018735e-01 -4.40178990e-01 -6.43355370e-01 -9.84258056e-01
-9.58596449e-03 1.33703858e-01 8.58820200e-01 -1.90849721e-01
-1.37086868e-01 3.47335935e-01 -5.79843879e-01 -5.57522357e-01
-7.14011967e-01 5.22897124e-01 2.97615021e-01 -7.01292977e-02
3.03760886e-01 1.66213647e-01 5.14075980e-02 3.78262579e-01
-8.68555754e-02 1.21431321e-01 2.35761270e-01 6.97213948e-01
8.53723437e-02 1.86842069e-01 5.86696744e-01 -5.34286082e-01
1.03661609e+00 -1.99300677e-01 -9.87119824e-02 4.44579691e-01
-5.28224051e-01 -1.53471246e-01 2.11232170e-01 -2.03202978e-01
-1.12716639e+00 2.24627033e-01 -4.98620152e-01 -1.28453169e-02
-1.05200626e-01 6.70824885e-01 -3.13535511e-01 3.39131713e-01
4.65335578e-01 4.03164208e-01 -2.38385439e-01 -9.36800361e-01
6.88383281e-02 1.59098053e+00 3.90451550e-01 -3.31528574e-01
2.78511643e-01 -1.34457141e-01 -7.18528152e-01 -1.18297994e+00
-8.01447928e-01 -8.75184000e-01 -4.76805359e-01 -2.63389379e-01
8.63325596e-01 -1.16647220e+00 -5.12487829e-01 8.09416771e-01
-1.32973528e+00 -2.89774716e-01 -1.20282218e-01 7.36361742e-01
-3.74849439e-01 1.09149732e-01 -5.92377007e-01 -1.32475305e+00
-7.77625144e-01 -1.29634905e+00 1.20606589e+00 -3.63368660e-01
-5.36882818e-01 -1.04914045e+00 7.62766302e-02 6.94598913e-01
5.88215709e-01 -6.02088094e-01 7.55027235e-01 -1.22341251e+00
2.63031214e-01 3.74630801e-02 1.32266760e-01 5.59488058e-01
2.86672145e-01 1.68266669e-01 -1.25029635e+00 -2.96372086e-01
6.80833384e-02 -3.43691856e-01 8.83862555e-01 3.63591224e-01
1.19308746e+00 -2.81238947e-02 -2.69189477e-01 1.16337659e-02
6.02990985e-01 3.63699943e-01 2.06003964e-01 2.01230198e-01
5.38708925e-01 8.64770293e-01 5.83749056e-01 5.04214883e-01
4.86366153e-01 9.65170026e-01 -1.44745708e-01 1.36126012e-01
-5.08765638e-01 -3.24317396e-01 9.27092493e-01 1.47479522e+00
2.02569425e-01 -5.36407769e-01 -1.36177289e+00 6.66693985e-01
-1.73424840e+00 -7.24104941e-01 -6.04619682e-01 2.23358488e+00
7.86766887e-01 1.99824423e-01 4.01925802e-01 3.10242832e-01
6.51039422e-01 -2.09171642e-02 3.02326143e-01 -6.90489352e-01
-1.99975684e-01 5.79697564e-02 2.93095857e-01 7.92278051e-01
-1.27809834e+00 1.18649840e+00 6.59020758e+00 1.24272728e+00
-1.00370288e+00 4.25361276e-01 3.54368389e-01 -5.38103655e-02
1.09297615e-02 -1.46292895e-01 -1.29841745e+00 4.69723642e-01
1.64081049e+00 -6.23715222e-01 4.00302976e-01 6.39730871e-01
3.64114404e-01 1.23654492e-01 -9.36576962e-01 9.69823241e-01
1.24409340e-01 -1.09169352e+00 3.60414162e-02 1.15317568e-01
3.54388326e-01 4.70181525e-01 1.38573915e-01 7.97498584e-01
2.28491858e-01 -1.17157280e+00 8.25921655e-01 3.71315449e-01
7.06237614e-01 -6.62500143e-01 9.08172131e-01 7.70454466e-01
-1.27865851e+00 5.29515333e-02 8.23275838e-03 8.24496076e-02
-6.84053153e-02 6.44972384e-01 -1.25695646e+00 6.61576986e-01
7.41636097e-01 5.37617683e-01 -3.63962948e-01 7.28816807e-01
-3.91167961e-02 1.00954771e+00 -2.40793124e-01 -1.33857921e-01
1.41107095e-02 5.25823474e-01 7.33321428e-01 1.61073732e+00
4.33664143e-01 -5.46141565e-01 5.74240029e-01 2.90750533e-01
-9.48197544e-02 6.79127157e-01 -6.00936055e-01 -7.00502098e-02
5.56825876e-01 1.16751993e+00 -2.14279771e-01 -3.37477922e-01
-3.78190279e-01 8.70436609e-01 -2.26119868e-02 7.00196251e-02
-3.65968287e-01 -4.40977126e-01 4.69205141e-01 -2.57325888e-01
-1.59057841e-01 -2.31443033e-01 -1.94737181e-01 -8.61209631e-01
-1.34200677e-01 -1.27042627e+00 2.78124630e-01 -5.36546528e-01
-1.08964276e+00 9.89183903e-01 -2.65004098e-01 -1.04569411e+00
-4.40843046e-01 -6.78610623e-01 -5.29681742e-01 1.22978270e+00
-9.68593836e-01 -1.15906048e+00 5.14090471e-02 6.49007022e-01
1.09335363e+00 -9.56610382e-01 1.15675783e+00 6.93398058e-01
-7.54287004e-01 7.79485166e-01 1.59047827e-01 2.12478831e-01
1.07849717e+00 -1.21996176e+00 4.30634588e-01 4.53045905e-01
2.43207827e-01 5.40098250e-01 6.86843216e-01 -5.98470151e-01
-1.20377791e+00 -1.09431934e+00 1.76423430e+00 -6.69258654e-01
7.61719525e-01 -6.46147549e-01 -4.74307328e-01 4.64659959e-01
5.36410928e-01 -2.87155181e-01 8.15884709e-01 4.48888421e-01
-6.59912750e-02 -6.87772110e-02 -9.16326046e-01 1.51630580e-01
7.88122356e-01 -9.68152285e-01 -3.94540370e-01 7.97410905e-01
8.03172588e-01 -4.55501497e-01 -8.38963270e-01 2.48206168e-01
4.69984531e-01 -4.28395569e-01 5.33372104e-01 -6.97339356e-01
1.22603709e-02 -5.22532947e-02 -7.05416381e-01 -1.43103683e+00
-5.22926077e-02 -6.24854207e-01 2.26828083e-01 1.57001591e+00
1.05360651e+00 -3.56238961e-01 3.23356539e-01 8.01396817e-02
-4.56135064e-01 -3.99178147e-01 -1.13775265e+00 -1.04544532e+00
3.78267057e-02 -8.95353138e-01 3.32662612e-01 7.05979884e-01
2.34276935e-01 7.27631271e-01 -5.94329894e-01 8.48044176e-03
2.56776690e-01 -4.92451280e-01 7.14617491e-01 -1.31501901e+00
-2.33738590e-02 -2.40567386e-01 7.64279021e-03 -8.83834422e-01
1.53951690e-01 -1.19563508e+00 3.53677645e-02 -1.40938544e+00
1.58442885e-01 -4.02540594e-01 -3.01166087e-01 6.20809734e-01
1.99514896e-01 -2.03441605e-01 2.45128378e-01 4.80743974e-01
-5.80430508e-01 2.78846413e-01 7.28518248e-01 -2.88477659e-01
-3.83225054e-01 3.66559684e-01 -5.94009697e-01 7.20731258e-01
1.08443403e+00 -4.62471455e-01 -2.99740911e-01 -4.84773844e-01
-7.79285878e-02 -5.96795119e-02 -8.12536180e-02 -1.11049616e+00
4.34962153e-01 8.93379524e-02 -1.30170226e-01 -1.07247484e+00
3.17884207e-01 -3.89082819e-01 -2.02133700e-01 3.99835080e-01
-6.99543715e-01 -1.18066818e-01 2.13299960e-01 -2.33868539e-01
-4.37335551e-01 -2.32910559e-01 3.49464327e-01 2.31735911e-02
-6.79126441e-01 -7.86159784e-02 -9.27277863e-01 9.29426029e-03
5.06747246e-01 3.61781389e-01 -4.09249842e-01 -4.98353004e-01
-1.15798020e+00 2.54320562e-01 -4.67311114e-01 9.75358903e-01
6.27687156e-01 -1.16566288e+00 -1.20686138e+00 2.41827995e-01
3.01702142e-01 -7.22928166e-01 -1.21559121e-01 9.81837988e-01
1.47104189e-01 9.08758521e-01 3.31266433e-01 -4.32455599e-01
-1.70612574e+00 1.95846539e-02 4.43275899e-01 -5.52817523e-01
-1.32045969e-01 7.54140615e-01 -1.80519000e-01 -1.01100612e+00
5.48015416e-01 -2.60972589e-01 -4.30060327e-01 6.56131729e-02
7.82062173e-01 4.76548851e-01 6.84571922e-01 -6.80413127e-01
-7.73068249e-01 1.12091742e-01 -4.39737648e-01 -6.19109631e-01
9.58774209e-01 -1.34425014e-01 -2.16163799e-01 8.11566472e-01
1.06687498e+00 5.41022420e-01 1.54186681e-01 -2.00460225e-01
5.03097117e-01 6.27024770e-02 6.07737064e-01 -1.20735812e+00
-6.15723372e-01 8.09219778e-01 7.66399443e-01 4.57422733e-01
8.60700667e-01 5.83622903e-02 5.26196837e-01 2.75071561e-01
3.68397743e-01 -1.42486989e+00 -1.68903530e-01 1.28994453e+00
1.18997419e+00 -1.12662160e+00 -4.64131534e-01 -4.34254378e-01
-6.51486039e-01 1.05103886e+00 4.76876497e-01 8.03660274e-01
7.53057539e-01 3.71589541e-01 2.99690962e-01 -2.81507568e-03
-1.25437713e+00 -2.85867155e-01 5.55650830e-01 4.65350360e-01
1.20651090e+00 3.97576779e-01 -6.47429645e-01 9.94598866e-01
-4.87052798e-01 -1.89434484e-01 6.98234662e-02 7.33564138e-01
-6.28549099e-01 -1.41620803e+00 -3.26597512e-01 3.13987941e-01
-7.02881098e-01 -3.58439773e-01 -8.09458792e-01 5.44968784e-01
3.10087744e-02 1.63942385e+00 -2.94090629e-01 -9.29710269e-01
4.58613843e-01 4.72796530e-01 -1.93228215e-01 -8.86666000e-01
-9.35250878e-01 3.05686533e-01 8.04456353e-01 -1.95643082e-01
1.54865822e-02 -6.11273706e-01 -1.14352477e+00 -1.36124790e-01
-3.06115627e-01 5.47238350e-01 1.10157514e+00 8.73353064e-01
3.57008517e-01 6.55440807e-01 8.71448636e-01 -2.89390087e-01
-5.86388052e-01 -1.40704727e+00 -4.56086159e-01 -1.29492879e-01
2.96916634e-01 -2.48222664e-01 -5.55048287e-01 -1.32202655e-01] | [9.524809837341309, 10.670978546142578] |
9032c677-366f-41a8-a85b-ea02deda4948 | histruct-improving-extractive-text-1 | 2203.09629 | null | https://arxiv.org/abs/2203.09629v1 | https://arxiv.org/pdf/2203.09629v1.pdf | HiStruct+: Improving Extractive Text Summarization with Hierarchical Structure Information | Transformer-based language models usually treat texts as linear sequences. However, most texts also have an inherent hierarchical structure, i.e., parts of a text can be identified using their position in this hierarchy. In addition, section titles usually indicate the common topic of their respective sentences. We propose a novel approach to formulate, extract, encode and inject hierarchical structure information explicitly into an extractive summarization model based on a pre-trained, encoder-only Transformer language model (HiStruct+ model), which improves SOTA ROUGEs for extractive summarization on PubMed and arXiv substantially. Using various experimental settings on three datasets (i.e., CNN/DailyMail, PubMed and arXiv), our HiStruct+ model outperforms a strong baseline collectively, which differs from our model only in that the hierarchical structure information is not injected. It is also observed that the more conspicuous hierarchical structure the dataset has, the larger improvements our method gains. The ablation study demonstrates that the hierarchical position information is the main contributor to our model's SOTA performance. | ['Georg Rehm', 'Malte Ostendorff', 'Qian Ruan'] | 2022-03-17 | null | https://aclanthology.org/2022.findings-acl.102 | https://aclanthology.org/2022.findings-acl.102.pdf | findings-acl-2022-5 | ['extractive-summarization', 'extractive-document-summarization'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.13093653e-01 4.61472690e-01 -4.48352575e-01 -7.72634670e-02
-1.12832355e+00 -6.88172579e-01 6.85986698e-01 5.06389558e-01
-4.19214368e-01 9.16484714e-01 1.19412243e+00 -2.67084092e-01
2.38558650e-01 -5.52792370e-01 -9.89378572e-01 -3.73719513e-01
3.83585423e-01 3.73889625e-01 2.58224234e-02 -1.87865078e-01
4.29843336e-01 -1.19807199e-01 -9.84889567e-01 6.63581610e-01
1.27732635e+00 5.99072695e-01 2.12887868e-01 6.00919247e-01
-3.49329859e-01 1.13396013e+00 -7.67899752e-01 -5.04408360e-01
-1.25364482e-01 -6.51148200e-01 -1.01757526e+00 -1.28020987e-01
7.36019373e-01 -3.39767247e-01 -6.34126484e-01 9.12771881e-01
4.01130915e-01 -3.60713005e-01 8.19726646e-01 -7.51847446e-01
-5.84261477e-01 1.33114886e+00 -7.36358523e-01 1.99941427e-01
3.99601609e-01 1.41298501e-02 1.56138957e+00 -7.32564688e-01
8.49282384e-01 1.12990558e+00 6.38162613e-01 2.89323241e-01
-1.15029442e+00 -5.13326824e-01 1.58395752e-01 -2.20950078e-02
-1.27707601e+00 -4.93802965e-01 6.15545452e-01 -3.14970940e-01
1.03205812e+00 3.65322024e-01 4.24815834e-01 1.22473145e+00
7.05715477e-01 1.22594428e+00 9.05125856e-01 -2.93056428e-01
2.64771581e-02 2.56323740e-02 3.95364940e-01 7.68487453e-01
3.94728214e-01 -4.89006221e-01 -8.74218166e-01 -1.91431418e-01
1.83953837e-01 -1.22443996e-01 -2.47192696e-01 1.07742786e-01
-1.28362310e+00 6.87328160e-01 2.56868571e-01 4.60399568e-01
-3.91587347e-01 7.53364861e-02 7.97980547e-01 1.72679394e-01
6.11371398e-01 7.83452511e-01 -4.06022817e-01 -1.05142467e-01
-1.48113012e+00 2.46263564e-01 9.83387887e-01 1.21283686e+00
6.06675863e-01 -2.58342028e-01 -6.76813424e-01 6.36616766e-01
-4.48987372e-02 2.39470616e-01 5.51267803e-01 -7.86245823e-01
8.06171775e-01 7.40308702e-01 -2.39348352e-01 -7.84751534e-01
-2.71528542e-01 -9.44194198e-01 -1.07034791e+00 -9.74433959e-01
-1.67157110e-02 -1.24486573e-01 -8.16754043e-01 1.61235106e+00
-2.61443436e-01 -7.82712325e-02 1.51061714e-01 3.24585140e-01
1.38649344e+00 8.25337291e-01 2.22918205e-02 -2.77811080e-01
1.44720125e+00 -1.04489815e+00 -9.63150799e-01 -2.81255722e-01
8.67695749e-01 -6.30589902e-01 9.79629755e-01 2.10015997e-01
-1.18450844e+00 -1.30984202e-01 -1.12468207e+00 -6.89102948e-01
-1.83349147e-01 3.37146670e-01 3.75050485e-01 1.65881559e-01
-1.18786395e+00 5.68892717e-01 -7.94226706e-01 -3.04146469e-01
5.65002024e-01 -8.89182836e-02 -2.64342487e-01 1.94048762e-01
-1.30085897e+00 6.92717373e-01 5.10189712e-01 -4.30768341e-01
-6.44659936e-01 -1.05881035e+00 -8.48276258e-01 4.94631380e-01
2.67424434e-01 -1.06489789e+00 1.47078419e+00 -3.71068060e-01
-1.19318020e+00 7.12090373e-01 -6.21695220e-01 -8.54657590e-01
4.80407953e-01 -3.13533157e-01 2.54063427e-01 3.32138568e-01
3.70865315e-01 6.10419691e-01 4.66580778e-01 -1.09522283e+00
-5.89657605e-01 -2.14763373e-01 1.91309135e-02 2.03233972e-01
-6.39036000e-01 -4.17232327e-03 -5.57686567e-01 -6.86646760e-01
-9.17864516e-02 -5.38999915e-01 -1.22635193e-01 -6.47761345e-01
-1.14735579e+00 -5.68405032e-01 4.85513330e-01 -9.00738299e-01
1.85865831e+00 -1.78792405e+00 3.99503917e-01 -1.97297871e-01
6.64471626e-01 -1.24020986e-01 2.43138941e-03 9.67558384e-01
1.72815740e-01 5.80797434e-01 -2.50158191e-01 -4.67905343e-01
2.10573271e-01 -1.00233488e-01 -7.04253733e-01 2.29234491e-02
-2.35097557e-02 1.30503738e+00 -9.32669342e-01 -8.26198876e-01
-3.83473933e-01 1.09283745e-01 -6.82630837e-01 -2.52814647e-02
-3.69393200e-01 1.22114383e-01 -5.65334320e-01 3.47199023e-01
4.32716131e-01 -5.17673433e-01 1.49114579e-01 -3.91710371e-01
-2.05927730e-01 8.57936621e-01 -2.73691893e-01 1.68619883e+00
-4.12367016e-01 8.34449291e-01 -1.97252825e-01 -7.72096753e-01
4.25920546e-01 5.25464833e-01 5.46941698e-01 -7.02981830e-01
-7.21901655e-02 9.06892940e-02 -1.81534022e-01 -1.77734897e-01
1.07488477e+00 1.95572287e-01 -3.69036704e-01 6.48022473e-01
1.08079940e-01 -1.04422852e-01 4.89933461e-01 1.09099066e+00
1.43034708e+00 -2.44112238e-01 4.46092099e-01 -4.02818501e-01
2.92631179e-01 -6.41437769e-02 4.55183893e-01 1.03712022e+00
1.81116536e-01 5.79357803e-01 1.08443260e+00 1.27762437e-01
-1.14026296e+00 -7.42795885e-01 -6.34750053e-02 9.48244572e-01
-1.79547146e-01 -9.97500360e-01 -9.43565249e-01 -8.59163225e-01
-8.86464771e-03 8.87856424e-01 -6.08413041e-01 -3.32669020e-01
-5.90835929e-01 -4.58927900e-01 8.53317499e-01 3.62485528e-01
4.04977411e-01 -7.93209314e-01 -2.39899144e-01 1.35115921e-01
-5.99954069e-01 -1.33618271e+00 -9.69015777e-01 -5.28072715e-02
-9.28767562e-01 -5.33906579e-01 -5.73648691e-01 -5.76758325e-01
5.90779126e-01 2.87566334e-01 1.37569463e+00 -6.62272274e-02
1.28162324e-01 3.79598618e-01 -4.00703758e-01 -5.75369179e-01
-6.06273651e-01 8.77670348e-01 -4.08957988e-01 -3.29664975e-01
1.83311328e-01 -5.43307126e-01 -5.45384228e-01 -2.87925005e-01
-1.01172268e+00 4.05456185e-01 7.85972774e-01 8.61419559e-01
3.63024235e-01 -6.49670586e-02 7.22299695e-01 -1.32467330e+00
9.21546578e-01 -4.60663289e-01 -3.17735113e-02 3.27875733e-01
-5.74869692e-01 3.19374204e-01 7.66438127e-01 -1.20398859e-02
-9.28887725e-01 -3.56756985e-01 -2.14784741e-01 -5.55456914e-02
1.64161876e-01 1.02543700e+00 -7.80042037e-02 4.99704331e-01
4.59997863e-01 7.27903426e-01 -2.20451102e-01 -5.16050279e-01
3.21397215e-01 6.71742678e-01 4.64899480e-01 -4.92836267e-01
7.60116518e-01 2.17352375e-01 -2.09903538e-01 -9.76895452e-01
-1.31342411e+00 -6.01795375e-01 -6.57884717e-01 1.13684520e-01
6.92990303e-01 -1.07888412e+00 -3.73265803e-01 2.70329267e-01
-1.50561917e+00 6.40076213e-03 -2.48389214e-01 1.06914751e-01
-2.89226979e-01 7.00202763e-01 -1.02702272e+00 -3.97281229e-01
-1.05376208e+00 -9.13973033e-01 1.44367337e+00 1.63990632e-02
-5.65997124e-01 -1.01368833e+00 -1.59980610e-01 4.41943198e-01
2.35755339e-01 3.42493989e-02 1.28718531e+00 -7.67084360e-01
-5.83652318e-01 -9.31716040e-02 -3.04061741e-01 1.17153578e-01
9.54840630e-02 2.37335153e-02 -6.56876385e-01 -2.27479696e-01
-1.60346076e-01 -2.47644737e-01 1.28784955e+00 3.24323922e-01
1.29652917e+00 -8.38471115e-01 -3.59480411e-01 2.98411965e-01
1.02961028e+00 -2.03371421e-01 7.79696107e-01 1.54449746e-01
7.84965694e-01 4.29624498e-01 2.37657994e-01 2.94208348e-01
7.11598456e-01 4.04485255e-01 5.09363832e-03 -2.26721242e-01
-2.15083644e-01 -6.82885170e-01 7.06485569e-01 1.42797244e+00
3.50348562e-01 -4.75159854e-01 -6.74018681e-01 4.48503315e-01
-1.63988614e+00 -1.03569007e+00 -1.46841258e-01 1.85297132e+00
1.62445605e+00 2.48009339e-01 -7.94899762e-02 -1.82677180e-01
2.64931679e-01 3.79914463e-01 -4.27717775e-01 -2.48074010e-01
-2.48874232e-01 3.05835009e-01 5.41638434e-01 3.83899659e-01
-9.02236104e-01 1.06062376e+00 6.11706591e+00 1.04640830e+00
-1.04119480e+00 2.59969030e-02 4.69394982e-01 -1.29589438e-01
-4.85056162e-01 -1.31441459e-01 -1.01689827e+00 5.96172333e-01
1.00497365e+00 -9.13262427e-01 -8.12560394e-02 3.97275358e-01
2.49013364e-01 2.40702704e-02 -1.31397402e+00 4.34773147e-01
1.15452215e-01 -1.67346358e+00 5.05580544e-01 1.46102488e-01
6.81083024e-01 6.79750517e-02 6.64426386e-02 5.03975928e-01
3.13167304e-01 -8.98194253e-01 8.29516530e-01 3.77848834e-01
7.19237208e-01 -5.50598979e-01 6.90944910e-01 6.16065085e-01
-8.34307671e-01 2.84996450e-01 -2.48504892e-01 2.69647717e-01
4.59391391e-03 7.70025849e-01 -9.25028026e-01 1.01152050e+00
3.96343052e-01 1.06076682e+00 -7.52987146e-01 7.34668553e-01
-3.40104073e-01 1.14233732e+00 -4.15287279e-02 -1.91779241e-01
4.01386708e-01 6.54170243e-03 6.89333141e-01 1.67875707e+00
2.95571595e-01 2.73678056e-03 2.20246628e-01 8.26960564e-01
-7.62055278e-01 2.17292041e-01 -5.44924021e-01 -4.44068670e-01
5.99749088e-01 1.20012677e+00 -5.40475965e-01 -5.80944836e-01
-3.09404045e-01 8.56467128e-01 3.19953471e-01 2.90955216e-01
-5.97235918e-01 -5.39866567e-01 1.62181914e-01 1.89031526e-01
2.61706948e-01 -2.16071501e-01 -4.84801322e-01 -1.48613036e+00
1.95875019e-02 -1.13313520e+00 3.96599054e-01 -5.83164036e-01
-1.06541181e+00 4.86651480e-01 -6.46250099e-02 -1.03743088e+00
-2.33901769e-01 1.89460404e-02 -6.16756260e-01 7.71516025e-01
-1.41808832e+00 -1.10721481e+00 1.32934958e-01 1.52021842e-02
7.81379044e-01 5.57927927e-03 5.21359205e-01 2.16657922e-01
-7.60040760e-01 8.86901379e-01 2.08932757e-01 4.05271679e-01
7.56298006e-01 -1.41835713e+00 4.35369045e-01 8.92197728e-01
4.82703792e-03 1.09833157e+00 9.17888939e-01 -8.13022733e-01
-1.49383545e+00 -1.28182232e+00 1.38955951e+00 -6.21172309e-01
8.38319182e-01 -5.69120467e-01 -1.00746751e+00 8.60443473e-01
9.68786001e-01 -8.94819498e-01 7.19854832e-01 2.07079902e-01
-4.27532881e-01 1.28020057e-02 -6.37849033e-01 7.06535578e-01
1.04180181e+00 -5.89392066e-01 -8.34121168e-01 5.32458365e-01
1.26682198e+00 -4.88408417e-01 -9.73811686e-01 1.91387489e-01
3.80012244e-01 -5.80259204e-01 6.60360515e-01 -7.12081969e-01
1.28647351e+00 1.27697095e-01 2.98770189e-01 -1.37564468e+00
-3.46582770e-01 -7.31727958e-01 -4.53687608e-01 1.46995366e+00
6.89117551e-01 -5.77829659e-01 6.51006877e-01 1.92263573e-01
-4.12770361e-01 -8.74198794e-01 -8.02815735e-01 -6.35938525e-01
5.42379618e-01 1.06287993e-01 4.02143389e-01 7.56895721e-01
2.82154322e-01 1.05097628e+00 -2.69865245e-01 -2.95408338e-01
5.81132352e-01 2.18137190e-01 6.65675938e-01 -1.08574283e+00
1.09668165e-01 -5.75533092e-01 5.03713489e-02 -1.52750266e+00
3.73235464e-01 -1.31314635e+00 7.00901374e-02 -1.93007421e+00
8.40415776e-01 1.82337053e-02 -1.23821609e-01 4.90610391e-01
-4.80131805e-01 -3.02185982e-01 6.83271699e-03 3.19022149e-01
-9.04579699e-01 8.17465484e-01 1.23919392e+00 -3.97090197e-01
-8.90412554e-03 -1.49285406e-01 -1.33437812e+00 4.00639266e-01
5.67810059e-01 -4.75272387e-01 -3.40538204e-01 -4.74336416e-01
3.75382990e-01 2.34931156e-01 6.55323118e-02 -5.83156168e-01
6.95384204e-01 1.99850291e-01 1.30917847e-01 -9.50105309e-01
-1.19346358e-01 -2.16897339e-01 -3.25275391e-01 4.69028503e-01
-9.58566844e-01 1.11249916e-01 2.28194609e-01 3.72539461e-01
-3.67898822e-01 -2.12614074e-01 4.21233922e-01 -1.44863933e-01
-1.30982608e-01 2.90097833e-01 -6.33801162e-01 4.41194266e-01
2.76663393e-01 1.55003563e-01 -6.74935520e-01 -6.36139214e-01
-2.02475622e-01 3.77818227e-01 3.05596918e-01 2.21646100e-01
4.87345755e-01 -1.03396475e+00 -1.09085846e+00 -2.91233778e-01
5.53660840e-02 3.87224518e-02 1.28605053e-01 1.12814927e+00
-2.03602791e-01 9.89223242e-01 2.98624903e-01 -4.59770858e-01
-1.25978446e+00 2.88257569e-01 -3.85850444e-02 -9.75323021e-01
-7.00900733e-01 6.20678902e-01 4.65581447e-01 -2.49406353e-01
1.96687415e-01 -5.26106775e-01 -2.73891300e-01 1.82444528e-01
4.38551992e-01 1.36082008e-01 1.70577407e-01 -3.55865091e-01
-2.95587513e-03 8.94986615e-02 -7.26724923e-01 -1.79266278e-02
1.27811694e+00 -6.33766949e-02 -6.33581638e-01 3.88826072e-01
1.32120097e+00 3.87390256e-01 -6.41189098e-01 -5.62856972e-01
2.73220301e-01 1.35995254e-01 1.78705066e-01 -7.63795614e-01
-7.22358823e-01 7.70318329e-01 -4.34140146e-01 2.72937030e-01
9.72549915e-01 1.62705123e-01 1.17690647e+00 3.75151962e-01
1.59139395e-01 -9.03127372e-01 1.09451801e-01 9.75953281e-01
1.07732856e+00 -7.37671316e-01 3.03529620e-01 -5.20999372e-01
-6.68619454e-01 8.88498962e-01 4.13044453e-01 3.12176675e-01
2.14591980e-01 3.72176766e-01 -3.83660734e-01 -2.57796258e-01
-1.10641968e+00 5.38130030e-02 4.51651394e-01 -2.02181283e-02
8.87548506e-01 -2.38218173e-01 -5.66063106e-01 8.77860904e-01
-6.71210229e-01 -7.11486712e-02 7.70040572e-01 7.32545614e-01
-5.24431646e-01 -8.59618127e-01 6.33878782e-02 8.44229281e-01
-8.29138756e-01 -5.55088878e-01 -7.87282169e-01 4.98452812e-01
-5.21981359e-01 8.57776582e-01 -4.70156260e-02 -3.76958340e-01
2.53450990e-01 3.87325808e-02 3.42960060e-01 -9.71752822e-01
-7.76264369e-01 4.55147512e-02 2.18027875e-01 -2.67956078e-01
-7.02830404e-02 -7.21468449e-01 -1.27149141e+00 -4.79756653e-01
-1.35880532e-02 6.93246722e-01 2.82003731e-01 8.60956967e-01
7.09234536e-01 9.13285136e-01 5.88871539e-01 -3.83677244e-01
-6.87237382e-01 -1.24816728e+00 -2.61350155e-01 6.55860007e-02
3.46447468e-01 -5.32967560e-02 -3.28518867e-01 1.73632473e-01] | [12.285308837890625, 9.279956817626953] |
96a87aab-a6c3-42d4-ad8e-1a8db988e12c | ensemble-learning-for-spectral-clustering | null | null | https://ieeexplore.ieee.org/document/9338349 | https://scholar.google.com/scholar_url?url=https://www.computer.org/csdl/pds/api/csdl/proceedings/download-article/1r54GbKzRdu/pdf%3Fcasa_token%3Dx1Nb_lJqLlcAAAAA:DPz3ReM6KGDlssLERhLHXS-KoqLClMeS5iiK0o2QhAvmKpsl2Umkk2oeWpQH5xmaagT0ALi43N4&hl=en&sa=T&oi=gsb-gga&ct=res&cd=0&d=13812841914647698097&ei=drqOYL62HILUyATd4pf4Dg&scisig=AAGBfm2wg99cdm46BHPq6pB_ZGwaThq2Fw | Ensemble Learning for Spectral Clustering | Ensemble clustering has attracted much attention in machine learning and data mining for the high performance in the task of clustering. Spectral clustering is one of the most popular clustering methods and has superior performance compared with the traditional clustering methods. Existing ensemble clustering methods usually directly use the clustering results of the base clustering algorithms for ensemble learning, which cannot make good use of the intrinsic data structures explored by the graph Laplacians in spectral clustering, thus cannot obtain the desired clustering result. In this paper, we propose a new ensemble learning method for spectral clustering-based clustering algorithms. Instead of directly using the clustering results obtained from each base spectral clustering algorithm, the proposed method learns a robust presentation of graph Laplacian by ensemble learning from the spectral embedding of each base spectral clustering algorithm. Finally, the proposed method applies k-means on the spectral embedding obtain from the learned graph Laplacian to get clusters. Experimental results on both synthetic and real-world datasets show that the proposed method outperforms other existing ensemble clustering methods. | ['Tetsuya Sakurai', 'Akira Imakura', 'Xiucai Ye', 'Hongmin Li'] | 2020-11-20 | null | null | null | null | ['imagedocument-clustering'] | ['computer-vision'] | [-8.47344007e-03 -5.48257113e-01 8.97260606e-02 -5.90690598e-02
-5.15039146e-01 -5.26702821e-01 1.48008242e-01 8.29466805e-02
-2.27248698e-01 2.78863996e-01 6.62145540e-02 -3.29587758e-02
-6.15306497e-01 -7.05540717e-01 -2.05628186e-01 -1.29878080e+00
-1.74154863e-01 2.81409502e-01 3.58899236e-02 1.61343351e-01
2.95831949e-01 9.36419703e-03 -1.42824018e+00 9.25617740e-02
1.26862788e+00 6.10821009e-01 2.58397520e-01 1.47506192e-01
-2.58382231e-01 2.28631616e-01 -2.67250150e-01 1.12207823e-01
1.16718665e-01 -7.86906064e-01 -2.18951941e-01 3.59564751e-01
-2.65756547e-01 3.39824498e-01 -2.56416589e-01 1.13434505e+00
5.20423353e-01 4.77630138e-01 8.79049003e-01 -1.34679711e+00
-4.36479300e-01 7.03364909e-01 -1.02460182e+00 -4.92200889e-02
-2.31577232e-02 -1.80335835e-01 8.69704425e-01 -7.68622041e-01
1.92444891e-01 1.27051139e+00 4.74021316e-01 2.51358122e-01
-1.42096710e+00 -1.01830339e+00 2.40994543e-01 4.54652101e-01
-1.92067289e+00 -1.90535754e-01 1.24387944e+00 -4.26089168e-01
2.50638336e-01 6.00518286e-02 3.91965181e-01 4.68943149e-01
-3.97896171e-01 5.82049489e-01 1.32780194e+00 -2.40656972e-01
2.37460524e-01 1.13382429e-01 3.43617916e-01 7.38883615e-01
6.68782517e-02 -2.69613177e-01 -6.39741123e-02 -3.10798407e-01
2.25482151e-01 2.57634103e-01 -5.55228114e-01 -6.55483067e-01
-1.34371567e+00 7.76133358e-01 6.53282106e-01 5.82576990e-01
-3.71258229e-01 1.61581449e-02 2.84354061e-01 8.46651495e-02
3.73103440e-01 -1.13753594e-01 -1.10805042e-01 3.03737819e-01
-9.27231669e-01 -3.15067321e-01 4.69339073e-01 7.35702574e-01
1.19702661e+00 -4.00126316e-02 3.63133192e-01 8.23868215e-01
5.37778795e-01 5.54805994e-01 5.90556204e-01 -7.87147224e-01
2.22891137e-01 1.10272849e+00 -7.77194425e-02 -1.51382291e+00
-3.06568056e-01 -3.18093210e-01 -1.28859770e+00 -1.15906723e-01
2.26707563e-01 -2.67391652e-01 -5.17162800e-01 1.45471525e+00
5.21781623e-01 7.82523453e-01 1.58228353e-01 6.61677182e-01
6.13467932e-01 9.00176048e-01 -1.36867195e-01 -8.08639765e-01
6.68813825e-01 -4.89728570e-01 -7.03572690e-01 3.90707314e-01
5.12600303e-01 -6.44726813e-01 6.78281367e-01 4.86970842e-01
-4.03124481e-01 -6.48109794e-01 -1.00044990e+00 7.07100630e-01
-2.44547069e-01 1.50859043e-01 2.19566777e-01 6.09023750e-01
-8.26220930e-01 2.57978439e-01 -8.50479186e-01 -2.71398216e-01
9.82126221e-02 4.94020522e-01 -2.77410775e-01 -2.07991183e-01
-8.01710546e-01 2.14707837e-01 9.55734491e-01 7.49716014e-02
-2.31265917e-01 -2.72051007e-01 -4.30654287e-01 -8.04733261e-02
6.30187571e-01 -3.35187614e-01 5.01933873e-01 -1.20297396e+00
-1.13040543e+00 2.40787566e-01 -3.60747010e-01 1.04633205e-01
1.19979821e-01 1.38821676e-01 -5.61351538e-01 1.57719269e-01
-1.72519889e-02 -1.47932619e-02 8.98082376e-01 -1.65170884e+00
-4.19266880e-01 -6.29732549e-01 -7.66866505e-01 2.18167186e-01
-5.81136167e-01 -5.32495797e-01 -4.26421881e-01 -4.00299430e-01
6.49775565e-01 -9.37824130e-01 -3.65448952e-01 -6.39395714e-01
-4.35906291e-01 -4.90723044e-01 1.43274474e+00 -4.24448937e-01
1.67905867e+00 -2.52394843e+00 5.52862346e-01 8.86714697e-01
1.79354116e-01 -1.40426783e-02 5.31778485e-02 7.32330263e-01
-2.22694397e-01 1.55031383e-01 -4.62581843e-01 1.48179829e-01
-2.93986559e-01 1.02161035e-01 3.64577360e-02 4.79975343e-01
-5.50557613e-01 2.73510516e-01 -1.06617427e+00 -8.09126318e-01
4.77765560e-01 4.81503844e-01 -2.22551450e-01 2.00301200e-01
4.54046987e-02 6.23269618e-01 -5.01538098e-01 -5.43766022e-02
5.77913344e-01 -2.99789280e-01 7.65182912e-01 -5.99779010e-01
7.69295692e-02 -5.75367630e-01 -1.72747540e+00 1.58851302e+00
-1.70055494e-01 -5.29884845e-02 7.46853352e-02 -1.42253876e+00
9.69056010e-01 3.71508449e-01 9.67845142e-01 5.36293630e-03
-6.86342828e-03 2.20312536e-01 3.18917722e-01 -3.68941784e-01
-1.10831156e-01 8.45187679e-02 1.47915438e-01 7.06853151e-01
-1.23568512e-01 3.26830357e-01 3.34763229e-01 4.56360996e-01
6.48942709e-01 -3.01571935e-01 2.04420760e-01 -3.79341960e-01
1.06617343e+00 -1.20216079e-01 7.29903102e-01 2.28832453e-01
1.82673112e-01 2.41332144e-01 -3.68652157e-02 -1.26446381e-01
-8.41113031e-01 -1.07719457e+00 1.31970957e-01 9.32665706e-01
4.40601379e-01 -5.95415771e-01 -9.21281159e-01 -8.06764781e-01
-9.07067806e-02 4.51104194e-01 -2.91727513e-01 -2.31793836e-01
-3.59297484e-01 -1.06944215e+00 2.03213364e-01 2.47184917e-01
7.32353806e-01 -7.40255952e-01 1.84189111e-01 3.47434849e-01
-5.02100170e-01 -6.97339952e-01 -6.36713862e-01 -2.63558596e-01
-9.81815279e-01 -1.31642127e+00 -2.26374075e-01 -9.45663452e-01
8.37671936e-01 8.50399256e-01 3.64105880e-01 5.41205704e-01
6.20882958e-02 5.87096751e-01 -5.64484239e-01 -4.09811325e-02
-3.74998033e-01 1.72260612e-01 3.59872043e-01 7.92802572e-01
4.95843828e-01 -9.13615406e-01 -6.11722589e-01 4.94119495e-01
-9.54249680e-01 -7.94275552e-02 6.30697966e-01 7.71251023e-01
5.96602738e-01 1.04818058e+00 7.77651429e-01 -7.93600798e-01
6.52228236e-01 -6.97445750e-01 -4.11483765e-01 2.95593232e-01
-8.24026108e-01 7.69969001e-02 7.59155929e-01 -4.16584700e-01
-1.01356459e+00 3.81939083e-01 4.54275668e-01 -5.11867106e-01
-1.35773286e-01 6.80526674e-01 -5.29078960e-01 1.30867511e-01
4.60244834e-01 7.49312937e-01 2.53356665e-01 -4.68326747e-01
5.00343978e-01 9.66769040e-01 3.91110241e-01 -5.13635933e-01
9.82700229e-01 5.77259064e-01 -1.08586393e-01 -9.99795139e-01
-3.47533941e-01 -8.70235801e-01 -1.00041878e+00 -5.06094277e-01
1.10490048e+00 -6.10017598e-01 -8.14520240e-01 4.24800932e-01
-7.13478267e-01 2.47951448e-01 5.73090672e-01 5.38837492e-01
-2.97766656e-01 7.55618572e-01 -1.63296133e-01 -1.04417455e+00
-4.14196067e-02 -1.08734667e+00 6.53656304e-01 1.24622382e-01
1.41903371e-01 -1.23714411e+00 4.73259157e-03 2.84680277e-01
-4.60270718e-02 3.83489519e-01 1.09743953e+00 -6.79083824e-01
-5.00439763e-01 -5.70427440e-02 -7.42801949e-02 2.60396302e-01
6.57623053e-01 1.83773786e-01 -5.93136251e-01 -6.30584002e-01
-9.29404348e-02 2.13974118e-01 9.37026739e-01 2.58605719e-01
1.41302085e+00 -2.22518906e-01 -7.17532218e-01 4.41660613e-01
1.62827349e+00 4.46784556e-01 2.42443696e-01 -1.60995126e-02
1.12798965e+00 5.69688678e-01 3.49666119e-01 3.00343156e-01
2.32288495e-01 2.99114376e-01 1.29754230e-01 7.05588758e-02
4.72494245e-01 -2.29392260e-01 4.97842610e-01 1.52087867e+00
-2.20507652e-01 3.21631171e-02 -1.03387940e+00 4.37795460e-01
-2.31855655e+00 -1.22958028e+00 -5.27940631e-01 2.27975440e+00
5.16251504e-01 -2.56510615e-01 3.58426690e-01 5.64920843e-01
1.24378610e+00 -1.13853188e-02 -4.72117037e-01 1.76971138e-01
6.44852072e-02 -1.45943880e-01 2.34927684e-01 2.56440043e-01
-1.09464359e+00 6.16092324e-01 5.16949987e+00 9.36765850e-01
-7.60149658e-01 -2.48840842e-02 3.19679350e-01 2.01039553e-01
-7.16224462e-02 9.72200856e-02 -2.39701629e-01 9.04495776e-01
6.26961827e-01 -3.61243993e-01 6.77932739e-01 5.81650257e-01
4.41562951e-01 -4.18122560e-02 -9.42053139e-01 1.17997622e+00
-6.14654012e-02 -9.12043571e-01 3.01801357e-02 1.87934607e-01
1.04587317e+00 -4.03862596e-01 1.74206614e-01 -7.98735172e-02
5.30015469e-01 -9.33440626e-01 -1.16764382e-01 6.97152078e-01
3.28352153e-01 -1.25388539e+00 6.31424665e-01 7.69576192e-01
-1.42340183e+00 -1.48452520e-01 -2.40249440e-01 2.58095711e-01
-1.78224474e-01 8.74183059e-01 -6.50356710e-01 9.01651382e-01
7.33607650e-01 1.23007143e+00 -6.36988938e-01 8.82308006e-01
5.60990945e-02 9.81470525e-01 -3.41952294e-01 4.99668233e-02
2.14895859e-01 -9.89448130e-01 4.74735439e-01 9.94689882e-01
3.73144150e-01 3.41478348e-01 7.00840175e-01 6.07523739e-01
1.01215601e-01 5.30272782e-01 -6.77601337e-01 -3.71907875e-02
8.44917953e-01 1.39034271e+00 -1.00079429e+00 -3.77235442e-01
-4.15935576e-01 7.98163176e-01 5.10825515e-02 5.14690161e-01
-5.46444774e-01 -2.24227205e-01 2.34608874e-01 -1.19442046e-01
2.10531503e-01 -5.40262520e-01 2.41557937e-02 -8.64839673e-01
-2.63751537e-01 -9.08631623e-01 6.35159671e-01 -5.38873732e-01
-1.62964177e+00 1.99797064e-01 -2.50359662e-02 -1.45974398e+00
6.07170314e-02 -3.03195655e-01 -8.27914834e-01 7.06944764e-01
-7.35957682e-01 -1.03553128e+00 -5.46110570e-01 9.68097687e-01
2.45562196e-01 -2.39423424e-01 6.07664466e-01 7.89233595e-02
-8.19539964e-01 1.94642603e-01 8.34703028e-01 3.29971552e-01
9.27180290e-01 -1.38232923e+00 -5.12797177e-01 7.67108083e-01
2.37825438e-01 8.01844358e-01 3.97586465e-01 -6.45174742e-01
-1.36553574e+00 -1.14017832e+00 -3.71451746e-03 -1.78437546e-01
6.54284060e-01 -1.29550472e-02 -1.17622197e+00 4.42107260e-01
3.58705193e-01 -4.24223274e-01 1.09145010e+00 1.75169453e-01
-3.57329488e-01 -5.11196136e-01 -9.02300537e-01 3.86183590e-01
7.72741079e-01 -2.60926574e-01 -4.05559987e-01 1.80750966e-01
3.02862167e-01 2.51832813e-01 -1.04959917e+00 3.37546468e-01
2.79928327e-01 -1.16082096e+00 7.96019495e-01 -8.46679211e-02
-1.19318031e-02 -1.01193547e+00 -1.10788375e-01 -1.70774651e+00
-4.85486269e-01 -1.48704246e-01 2.27448374e-01 1.50122476e+00
1.47862092e-01 -6.65367782e-01 4.98384923e-01 -2.07468774e-02
2.15592131e-01 -3.05896282e-01 -5.66964567e-01 -6.72586083e-01
-1.44166902e-01 -3.06212485e-01 6.26148999e-01 1.34386289e+00
-4.41022730e-03 5.70147634e-01 -1.79876104e-01 4.27371293e-01
1.25329113e+00 3.50860268e-01 8.99848759e-01 -1.41786933e+00
-1.87349245e-01 -4.82829720e-01 -4.59126413e-01 -4.61717963e-01
3.12475353e-01 -1.25194478e+00 -7.27505013e-02 -1.46397531e+00
4.34090942e-01 -5.59443295e-01 -4.98019993e-01 5.72507419e-02
-5.76911390e-01 3.76640470e-03 1.42727256e-01 5.29563189e-01
-6.79847002e-01 3.95119935e-01 9.22334313e-01 -2.30095342e-01
-4.62260395e-01 -8.86524320e-02 -5.87399900e-01 7.53728688e-01
7.48772085e-01 -3.17616314e-01 -7.68872142e-01 -8.23109373e-02
-1.34649858e-01 -2.24810187e-02 6.93349913e-02 -1.13923728e+00
4.44461882e-01 -1.98086366e-01 3.09569448e-01 -5.83947599e-01
-8.82423595e-02 -1.23187172e+00 5.26643634e-01 2.99534172e-01
2.02964200e-03 -9.71273407e-02 -1.82896465e-01 1.07367349e+00
-2.70231992e-01 1.39149889e-01 8.78449202e-01 -4.16331217e-02
-8.04736972e-01 4.00787830e-01 -3.36191982e-01 -1.32555947e-01
1.25504184e+00 -3.29199702e-01 1.67959869e-01 -3.52759570e-01
-1.00081563e+00 5.64397812e-01 5.35760939e-01 2.33627319e-01
6.63470864e-01 -1.46925128e+00 -7.58399844e-01 5.60487658e-02
1.09650679e-01 3.54426354e-02 2.49959812e-01 1.12433779e+00
6.24010339e-02 -1.23604879e-01 2.10399806e-01 -9.30708826e-01
-1.48240125e+00 1.02663743e+00 1.55866459e-01 5.03120534e-02
-3.58526528e-01 1.40723452e-01 1.01804271e-01 -3.86069715e-01
-5.04754968e-02 5.32283008e-01 -3.62950802e-01 1.00719959e-01
2.63009638e-01 6.74867630e-01 -1.82928383e-01 -8.82771969e-01
-3.26705515e-01 8.29024851e-01 1.78141981e-01 -1.34871498e-01
1.15212166e+00 -2.41756514e-01 -5.89086592e-01 7.70264149e-01
1.36290348e+00 1.01617593e-02 -5.74594021e-01 -3.60591084e-01
2.87821323e-01 -2.72060215e-01 -1.07264649e-02 -4.06787395e-01
-1.12551975e+00 8.64723384e-01 6.16079271e-01 3.71903956e-01
1.53053594e+00 -2.81696647e-01 5.58255136e-01 4.49590594e-01
4.02322084e-01 -1.21497190e+00 1.86019346e-01 1.44502148e-02
2.76823580e-01 -1.11668587e+00 5.95663227e-02 -5.16717672e-01
-4.97711033e-01 1.00604844e+00 6.00999117e-01 -3.56276840e-01
9.35707808e-01 -2.83469379e-01 1.42510027e-01 -4.06775810e-02
-4.09726888e-01 -2.99214423e-01 3.98236483e-01 6.28429294e-01
4.21656162e-01 1.58115938e-01 -3.53151172e-01 4.09888715e-01
3.53477411e-02 -6.39030397e-01 2.17386842e-01 3.67101967e-01
-6.80854201e-01 -1.31399250e+00 -7.37935841e-01 5.85077882e-01
-6.13792874e-02 2.26933405e-01 -7.76460290e-01 6.53636456e-01
2.46702969e-01 1.21902156e+00 -6.02788888e-02 -8.34258020e-01
4.83854190e-02 3.55626255e-01 2.10935190e-01 -4.96092975e-01
-1.42884687e-01 3.22915792e-01 -5.09133577e-01 -2.63359874e-01
-1.02319109e+00 -7.58727610e-01 -1.50062823e+00 -3.40734154e-01
-5.53936124e-01 7.44249701e-01 3.92158926e-01 8.68691802e-01
2.50046045e-01 3.77918035e-01 1.17988503e+00 -6.32644713e-01
-9.62720439e-02 -8.09586287e-01 -7.36532450e-01 7.14410245e-01
-8.75877496e-03 -7.22452164e-01 -6.09919548e-01 3.17271203e-01] | [7.915320873260498, 4.648407936096191] |
12ae4b91-9214-49b4-88db-766646d5e5af | gocor-bringing-globally-optimized | 2009.07823 | null | https://arxiv.org/abs/2009.07823v4 | https://arxiv.org/pdf/2009.07823v4.pdf | GOCor: Bringing Globally Optimized Correspondence Volumes into Your Neural Network | The feature correlation layer serves as a key neural network module in numerous computer vision problems that involve dense correspondences between image pairs. It predicts a correspondence volume by evaluating dense scalar products between feature vectors extracted from pairs of locations in two images. However, this point-to-point feature comparison is insufficient when disambiguating multiple similar regions in an image, severely affecting the performance of the end task. We propose GOCor, a fully differentiable dense matching module, acting as a direct replacement to the feature correlation layer. The correspondence volume generated by our module is the result of an internal optimization procedure that explicitly accounts for similar regions in the scene. Moreover, our approach is capable of effectively learning spatial matching priors to resolve further matching ambiguities. We analyze our GOCor module in extensive ablative experiments. When integrated into state-of-the-art networks, our approach significantly outperforms the feature correlation layer for the tasks of geometric matching, optical flow, and dense semantic matching. The code and trained models will be made available at github.com/PruneTruong/GOCor. | ['Luc van Gool', 'Radu Timofte', 'Martin Danelljan', 'Prune Truong'] | 2020-09-16 | null | http://proceedings.neurips.cc/paper/2020/hash/a4a8a31750a23de2da88ef6a491dfd5c-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/a4a8a31750a23de2da88ef6a491dfd5c-Paper.pdf | neurips-2020-12 | ['geometric-matching', 'dense-pixel-correspondence-estimation'] | ['computer-vision', 'computer-vision'] | [-1.54628351e-01 -1.55471386e-02 6.55926391e-02 -2.34434202e-01
-5.95921099e-01 -4.22813654e-01 8.15369129e-01 1.19667456e-01
-5.19490778e-01 2.39607766e-01 1.80057138e-01 -9.22904070e-03
-3.78088765e-02 -6.74171269e-01 -6.59950495e-01 -4.35582697e-01
8.62870440e-02 3.77329856e-01 2.18864709e-01 -9.68596935e-02
4.25001502e-01 8.51654172e-01 -1.48373127e+00 2.15078861e-01
6.03689969e-01 1.13357890e+00 3.32903981e-01 5.08619189e-01
-8.35265964e-03 6.23382986e-01 -1.01911329e-01 -3.41689736e-01
7.03834295e-01 -2.94747621e-01 -9.42511499e-01 -1.68793835e-02
1.27631795e+00 -3.91624331e-01 -6.33731842e-01 1.09803426e+00
3.38292569e-01 3.80674452e-01 4.15343106e-01 -1.18017423e+00
-4.58117515e-01 1.58429861e-01 -5.73065639e-01 4.46140140e-01
2.80959159e-01 3.23308229e-01 1.30861616e+00 -1.24175346e+00
7.47839153e-01 1.38240504e+00 6.94310784e-01 1.65743962e-01
-1.22828555e+00 -2.87603676e-01 6.58475002e-03 2.39014417e-01
-1.32895029e+00 -7.22561836e-01 8.29831898e-01 -5.67429304e-01
1.06185460e+00 -9.00774635e-03 5.95491469e-01 5.00395536e-01
1.44892305e-01 6.62276745e-01 4.95905817e-01 -1.60237253e-01
-2.07294710e-02 -2.90711731e-01 -1.27789602e-01 9.49715614e-01
1.15128674e-01 6.23480439e-01 -5.97736001e-01 -3.26721966e-02
1.21165442e+00 1.50630593e-01 -4.36194986e-01 -7.25352347e-01
-1.35858488e+00 8.57381105e-01 1.07177174e+00 3.58208954e-01
-3.11257541e-01 4.13118362e-01 5.57963699e-02 2.05983117e-01
2.95319915e-01 7.34810233e-01 -9.20372829e-02 1.18629053e-01
-1.01172113e+00 3.69598389e-01 4.99073923e-01 8.29359233e-01
1.14671135e+00 -1.56360850e-01 -1.53740823e-01 5.36303461e-01
3.00964385e-01 2.95806676e-01 1.97475761e-01 -1.38301754e+00
4.42786574e-01 5.96143425e-01 -6.05410570e-03 -1.36753511e+00
-4.23746616e-01 -5.41623354e-01 -7.12273717e-01 2.71573216e-01
6.02473021e-01 6.08639941e-02 -5.93580663e-01 1.55450845e+00
2.87353516e-01 5.55411756e-01 -2.52556115e-01 1.31638825e+00
9.47429538e-01 2.96070725e-01 -1.17315039e-01 4.31705236e-01
1.11844242e+00 -1.21845329e+00 -2.83488214e-01 -3.69454861e-01
5.59109867e-01 -1.04521298e+00 7.75861442e-01 -1.68722615e-01
-1.35422385e+00 -7.41119862e-01 -8.41619670e-01 -5.23473799e-01
-3.54685217e-01 5.34441434e-02 7.79254138e-01 -1.09774508e-01
-1.28565025e+00 9.13157046e-01 -8.34097862e-01 -2.88260996e-01
5.77833951e-01 2.87077695e-01 -5.20882368e-01 -1.29048482e-01
-8.99478197e-01 1.04623663e+00 2.17222571e-02 3.13143164e-01
-4.05764461e-01 -9.12734449e-01 -1.34625196e+00 2.06782416e-01
-1.42194973e-02 -1.02330017e+00 9.13149476e-01 -6.11512005e-01
-1.09109199e+00 1.23220193e+00 -3.82503450e-01 -3.29517454e-01
6.36782885e-01 -2.42627650e-01 2.15587735e-01 3.71318191e-01
1.61970705e-01 1.06183362e+00 7.56872952e-01 -9.40669894e-01
-5.24825692e-01 -3.49357605e-01 3.42053697e-02 1.42367199e-01
3.99181187e-01 -1.20715119e-01 -3.76851588e-01 -5.49228489e-01
3.48286867e-01 -7.52510965e-01 -4.36061889e-01 4.37477231e-01
-2.85431951e-01 6.73312321e-03 6.03686690e-01 -4.30480957e-01
6.58087075e-01 -2.18923664e+00 1.09126054e-01 1.77967265e-01
3.36795628e-01 1.96169719e-01 -5.60352385e-01 5.07776998e-02
-1.40929639e-01 -2.11084843e-01 -2.59990007e-01 -6.09355152e-01
-8.17684829e-02 -1.60766110e-01 -3.06116730e-01 8.56622100e-01
5.36808550e-01 1.28687727e+00 -9.44397867e-01 -3.27991128e-01
6.94314361e-01 7.00673461e-01 -6.87474966e-01 3.67765933e-01
-9.52920038e-03 6.35012269e-01 -2.26457223e-01 4.80323076e-01
7.43006468e-01 -4.32878107e-01 -3.08480054e-01 -5.32135308e-01
-9.65526327e-02 2.97627568e-01 -1.15061963e+00 2.16131949e+00
-5.09456694e-01 9.09417987e-01 -8.07705075e-02 -1.04112816e+00
8.55330229e-01 -5.76465093e-02 7.06850350e-01 -7.48652518e-01
3.53176057e-01 1.62551552e-01 1.68119483e-02 -2.82661885e-01
4.59511340e-01 2.76361942e-01 2.65361279e-01 2.67571330e-01
2.02226967e-01 -2.54033834e-01 2.64983803e-01 1.60562441e-01
8.73877764e-01 3.03977609e-01 2.54321903e-01 -2.16528654e-01
7.38070488e-01 -7.36248195e-02 4.89576131e-01 7.10199714e-01
-2.83745557e-01 1.07884347e+00 3.05297285e-01 -7.20156968e-01
-1.13025010e+00 -1.15200746e+00 -3.31928730e-01 4.80983615e-01
5.55496454e-01 -3.47588122e-01 -3.69084239e-01 -3.53313893e-01
2.31159091e-01 4.93822545e-02 -5.57746768e-01 -1.55077996e-02
-7.39837110e-01 -1.16636366e-01 2.45567724e-01 5.26632667e-01
7.28451610e-01 -9.99748707e-01 -7.90069163e-01 7.36648738e-02
-3.13445866e-01 -1.42303932e+00 -7.98235297e-01 -2.44057834e-01
-7.46406496e-01 -1.44293237e+00 -5.16991138e-01 -6.79819822e-01
8.12820196e-01 4.03227359e-01 1.31438255e+00 5.94451904e-01
-6.73925579e-01 2.30904624e-01 -1.62796360e-02 1.79520786e-01
-1.26063928e-01 2.92663975e-03 -3.58121157e-01 2.93916706e-02
3.83709610e-01 -7.06996739e-01 -9.57524896e-01 4.46586877e-01
-5.84813774e-01 1.32371426e-01 2.72826076e-01 7.52511322e-01
5.54871500e-01 -7.08653808e-01 -1.40373809e-02 -3.47026646e-01
2.41268352e-01 -1.38355419e-01 -1.01843524e+00 6.15127943e-02
-1.32043824e-01 3.29324961e-01 3.10395002e-01 -8.62421989e-02
-5.59022725e-01 3.38034153e-01 -1.16768323e-01 -6.83544695e-01
-1.88648909e-01 1.59110039e-01 1.42183855e-01 -4.09601837e-01
4.27765518e-01 -1.49946302e-01 2.75967658e-01 -4.14414793e-01
5.41772366e-01 1.43576130e-01 8.57861638e-01 -4.91126478e-01
9.51939344e-01 7.68578708e-01 3.38273853e-01 -5.69768131e-01
-9.13385987e-01 -7.85176694e-01 -1.04987323e+00 -5.93093298e-02
9.01968002e-01 -9.22580659e-01 -8.21236730e-01 4.27811086e-01
-1.38757801e+00 -2.47561723e-01 -3.04157525e-01 6.19829893e-01
-7.73103058e-01 3.04530293e-01 -5.16536653e-01 -1.16788775e-01
-2.84727395e-01 -1.30874503e+00 1.28599405e+00 3.16227794e-01
-2.69451648e-01 -1.34598863e+00 1.19399734e-01 2.74390817e-01
4.80939746e-01 1.77877665e-01 5.42687535e-01 -4.08207864e-01
-9.99882936e-01 -1.97531462e-01 -5.36773384e-01 7.63830096e-02
8.53043124e-02 5.24743423e-02 -1.06328416e+00 -2.77433842e-01
-2.85173476e-01 -1.30389690e-01 1.09351587e+00 5.34124672e-01
9.69632685e-01 2.49418132e-02 -1.80405185e-01 1.23918772e+00
1.45774484e+00 -3.74010980e-01 6.42491758e-01 3.36395681e-01
8.45844507e-01 6.67202830e-01 4.86525238e-01 1.98362023e-01
3.01619440e-01 8.81356120e-01 6.02105916e-01 -4.85062897e-01
-5.38712740e-01 -2.50816911e-01 -2.52880957e-02 2.86328286e-01
6.95250481e-02 1.26071364e-01 -7.71339118e-01 5.64206481e-01
-2.04603863e+00 -1.02981770e+00 -1.98484194e-02 2.25576973e+00
4.38467771e-01 -2.54649818e-01 -1.17471494e-01 -3.02412748e-01
6.78100467e-01 4.07762229e-01 -5.05865097e-01 -1.84027091e-01
-1.38899326e-01 2.08672270e-01 3.48182648e-01 9.46178317e-01
-1.14192176e+00 1.13352346e+00 5.76753759e+00 3.15659016e-01
-1.24687958e+00 -1.51891068e-01 5.31623781e-01 -3.38306762e-02
-2.06622362e-01 2.07788900e-01 -6.51849985e-01 2.30051786e-01
1.97715387e-01 5.23050278e-02 3.29750627e-01 4.58551645e-01
3.13130677e-01 -2.77631164e-01 -1.49192381e+00 1.22723472e+00
8.44550356e-02 -1.79517448e+00 7.40858242e-02 5.69498986e-02
6.87747896e-01 4.10875976e-01 1.62335098e-01 -2.80656219e-01
1.52606398e-01 -1.13877845e+00 4.88926470e-01 5.51859081e-01
5.81607640e-01 -4.09072042e-01 6.53390765e-01 -1.40356913e-03
-1.22303557e+00 1.52572036e-01 -5.33817470e-01 -1.89215615e-02
2.27239981e-01 5.83973467e-01 -5.18320262e-01 3.32479089e-01
5.33002436e-01 1.09544647e+00 -6.28631353e-01 1.42072606e+00
-2.77451202e-02 -2.12080896e-01 -3.34015608e-01 5.82616866e-01
4.22109783e-01 -3.74635279e-01 5.23463190e-01 1.00274730e+00
1.78294033e-01 -1.78883180e-01 1.96526706e-01 1.36441171e+00
-8.32245648e-02 -3.90760712e-02 -6.66239262e-01 3.53360265e-01
4.64736313e-01 1.33205795e+00 -4.98559803e-01 -1.71917439e-01
-4.86591101e-01 8.99128973e-01 6.55994177e-01 4.48291808e-01
-5.43147326e-01 -2.88876772e-01 1.08904016e+00 2.03269452e-01
2.88398772e-01 -3.26174140e-01 -3.64070207e-01 -1.37183690e+00
5.79348803e-02 -3.67003202e-01 1.57713398e-01 -9.16778147e-01
-1.26957488e+00 5.40863931e-01 -3.34310859e-01 -1.36558342e+00
-4.20371056e-01 -6.25910580e-01 -8.40965033e-01 1.17531216e+00
-1.82793486e+00 -9.65671957e-01 -6.83920443e-01 7.74572551e-01
2.80013323e-01 -3.20494287e-02 6.11258924e-01 2.96715170e-01
-3.46302718e-01 5.42539597e-01 -2.38305956e-01 5.01554310e-01
6.91415250e-01 -9.72311676e-01 7.53908873e-01 1.07811582e+00
3.45587075e-01 7.03456342e-01 3.65946621e-01 -2.64059424e-01
-1.08591044e+00 -9.82022166e-01 1.03272295e+00 -4.48074907e-01
5.75889826e-01 -2.91234195e-01 -9.57412958e-01 5.41892648e-01
1.47867858e-01 5.48748732e-01 1.70665473e-01 -1.39215305e-01
-5.51855803e-01 3.81032452e-02 -1.02471614e+00 5.20721495e-01
1.17518711e+00 -6.54373467e-01 -4.41138804e-01 2.45980173e-01
5.35550416e-01 -6.69074059e-01 -8.75575900e-01 2.46223524e-01
3.37954938e-01 -1.26502061e+00 1.25250041e+00 -4.45601225e-01
6.67119265e-01 -3.56970131e-01 6.38315603e-02 -1.18110275e+00
-2.80679345e-01 -7.03289568e-01 5.23105785e-02 7.08007514e-01
2.55794019e-01 -6.61874592e-01 7.82333016e-01 6.94678783e-01
-3.63481268e-02 -6.03690028e-01 -8.31486762e-01 -6.35836244e-01
7.01567456e-02 -2.34062269e-01 5.45099616e-01 9.30113614e-01
-9.11424160e-02 2.29120031e-01 -2.31107641e-02 1.44571811e-01
7.73685277e-01 3.80619675e-01 8.03599000e-01 -1.13431716e+00
-2.81276703e-01 -7.80630887e-01 -7.60444283e-01 -1.45952606e+00
4.33040738e-01 -1.03297591e+00 1.44498022e-02 -1.40346432e+00
-7.27570383e-03 -5.45140564e-01 1.29461080e-01 4.23770100e-01
-8.48257318e-02 4.51720834e-01 5.47976911e-01 3.17340612e-01
-3.01559687e-01 4.88252133e-01 1.51741362e+00 -6.46969229e-02
-1.34523183e-01 -4.39946055e-02 -4.35759902e-01 5.98313928e-01
7.37258971e-01 -9.68217775e-02 -1.40564844e-01 -5.95681727e-01
-8.24715346e-02 -7.35760061e-03 9.05333757e-01 -9.71962988e-01
4.29737985e-01 -4.61654924e-02 3.84682089e-01 -4.45250750e-01
4.32282418e-01 -8.43550444e-01 -1.02087460e-01 3.53539884e-01
-4.01042610e-01 3.14414233e-01 5.75680286e-02 5.11109866e-02
-5.09063601e-01 -5.88028505e-02 9.30312872e-01 -2.28748903e-01
-8.85449886e-01 7.93954313e-01 1.76315188e-01 2.91350067e-01
5.88935137e-01 -3.93552661e-01 -4.50343758e-01 -3.35883766e-01
-7.30132222e-01 2.46849135e-01 7.09353983e-01 5.68703830e-01
8.26226532e-01 -1.33959079e+00 -7.70326793e-01 5.58362007e-01
4.21948060e-02 3.02545547e-01 6.59683049e-02 1.03127432e+00
-6.64806902e-01 5.19175649e-01 -3.85963321e-01 -7.92850196e-01
-1.02103889e+00 3.45652789e-01 8.81257832e-01 -6.68647364e-02
-9.20125425e-01 8.05222750e-01 4.29649740e-01 -3.59253138e-01
2.56081492e-01 -2.85087317e-01 -8.93493071e-02 -2.54466206e-01
5.13961315e-01 2.36573353e-01 4.51099537e-02 -9.59816456e-01
-3.49880457e-01 1.03052211e+00 1.40666336e-01 -6.83953166e-02
1.16640186e+00 -2.11994007e-01 -3.02541673e-01 -1.87290534e-02
1.72543883e+00 -3.19079638e-01 -1.59575701e+00 -6.40127063e-01
-1.85090274e-01 -8.78458977e-01 1.17759138e-01 -4.08840597e-01
-1.43431103e+00 1.03683054e+00 4.49037194e-01 -3.52298975e-01
8.36258471e-01 5.00715412e-02 7.95114100e-01 1.98390305e-01
1.29971102e-01 -6.69680178e-01 6.10358380e-02 5.78211069e-01
9.94590938e-01 -1.54504561e+00 -6.75845668e-02 -4.43757236e-01
-1.84912980e-01 1.17163312e+00 7.60830998e-01 -4.65061516e-01
7.52436757e-01 1.64636433e-01 1.42021000e-01 -3.36996883e-01
-5.07118881e-01 -3.97600085e-01 6.89055860e-01 6.00587606e-01
5.01639307e-01 -3.72125804e-01 3.53207812e-02 -4.55563039e-01
-2.99259931e-01 -2.96195865e-01 4.02161419e-01 5.97032964e-01
-3.48561376e-01 -9.61091101e-01 -2.36470237e-01 1.42309263e-01
-1.04235865e-01 -3.43672216e-01 -1.74767137e-01 6.55653715e-01
-5.05103320e-02 7.40022719e-01 7.38566458e-01 -1.44050807e-01
3.06681126e-01 -4.51781899e-01 5.78464508e-01 -4.43999887e-01
-6.32217765e-01 -8.83338526e-02 -2.84279406e-01 -1.21225655e+00
-5.55199504e-01 -6.03216946e-01 -1.33709824e+00 -4.76692051e-01
3.76841798e-02 -6.42443374e-02 4.28423017e-01 8.88292253e-01
4.58375365e-01 5.72443157e-02 5.62404156e-01 -1.10670900e+00
-2.88102299e-01 -5.96855760e-01 -2.36564144e-01 7.44335592e-01
7.87827194e-01 -7.32764482e-01 -4.02490914e-01 -3.30140814e-02] | [8.602540016174316, -2.1729979515075684] |
45114b57-704d-4d75-9665-e36e19726e30 | unsupervised-activity-segmentation-by-joint | 2105.13353 | null | https://arxiv.org/abs/2105.13353v6 | https://arxiv.org/pdf/2105.13353v6.pdf | Unsupervised Action Segmentation by Joint Representation Learning and Online Clustering | We present a novel approach for unsupervised activity segmentation which uses video frame clustering as a pretext task and simultaneously performs representation learning and online clustering. This is in contrast with prior works where representation learning and clustering are often performed sequentially. We leverage temporal information in videos by employing temporal optimal transport. In particular, we incorporate a temporal regularization term which preserves the temporal order of the activity into the standard optimal transport module for computing pseudo-label cluster assignments. The temporal optimal transport module enables our approach to learn effective representations for unsupervised activity segmentation. Furthermore, previous methods require storing learned features for the entire dataset before clustering them in an offline manner, whereas our approach processes one mini-batch at a time in an online manner. Extensive evaluations on three public datasets, i.e. 50-Salads, YouTube Instructions, and Breakfast, and our dataset, i.e., Desktop Assembly, show that our approach performs on par with or better than previous methods, despite having significantly less memory constraints. | ['Quoc-Huy Tran', 'M. Zeeshan Zia', 'Andrey Konin', 'Awais Ahmed', 'Sanjay Haresh', 'Sateesh Kumar'] | 2021-05-27 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Kumar_Unsupervised_Action_Segmentation_by_Joint_Representation_Learning_and_Online_Clustering_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Kumar_Unsupervised_Action_Segmentation_by_Joint_Representation_Learning_and_Online_Clustering_CVPR_2022_paper.pdf | cvpr-2022-1 | ['online-clustering'] | ['computer-vision'] | [ 2.24373534e-01 -2.43633762e-01 -6.01695955e-01 -3.99727017e-01
-8.54405999e-01 -8.69571865e-01 4.16015625e-01 3.40609580e-01
-5.78740060e-01 2.95090288e-01 2.03707442e-01 -2.29562804e-01
-1.10199740e-02 -3.79290164e-01 -7.47165084e-01 -6.83744907e-01
-3.62180680e-01 4.31468666e-01 3.56803179e-01 5.03326774e-01
4.73452806e-01 2.63541490e-01 -1.47002470e+00 4.61580455e-01
7.30355084e-01 1.03204739e+00 1.51611120e-01 8.06948066e-01
-1.21966556e-01 1.29827571e+00 -3.23774487e-01 8.06143209e-02
3.09666038e-01 -6.50450051e-01 -1.11645615e+00 7.81148911e-01
1.69893727e-01 -2.01237187e-01 -5.36069334e-01 5.51157176e-01
-3.55472509e-03 8.23603868e-01 6.65783525e-01 -1.14834917e+00
-2.34419197e-01 4.63693887e-01 -7.47097492e-01 4.89104152e-01
4.83662635e-01 4.65871803e-02 1.02967060e+00 -4.96200144e-01
7.32438028e-01 8.81416261e-01 4.70905542e-01 3.78075629e-01
-1.44436705e+00 -4.10211146e-01 5.84311306e-01 2.32908666e-01
-1.52425015e+00 -5.20568728e-01 5.77943087e-01 -6.26382470e-01
9.03795302e-01 4.37678136e-02 7.91800082e-01 9.21152949e-01
-2.26967856e-01 1.19704366e+00 8.11208963e-01 -2.80620962e-01
5.31592965e-01 -4.12402838e-01 3.53812397e-01 7.88660944e-01
-2.22926229e-01 -5.29477477e-01 -6.61548078e-01 -5.50285615e-02
7.46372283e-01 3.20987552e-01 -1.43923387e-01 -6.67554915e-01
-1.34782994e+00 5.96378267e-01 -1.17916579e-03 2.08093390e-01
-2.28996396e-01 5.52415371e-01 5.19213080e-01 1.41040117e-01
4.65519547e-01 1.33185998e-01 -4.99559432e-01 -7.61443853e-01
-1.22247577e+00 -8.86952206e-02 8.36348295e-01 1.05448949e+00
1.02639675e+00 -1.85258269e-01 -9.30107683e-02 6.34878457e-01
1.98123589e-01 -1.37864456e-01 5.75841308e-01 -1.61658168e+00
4.11029130e-01 5.27353525e-01 1.13693923e-01 -6.73113406e-01
-2.37887576e-01 1.27660528e-01 -3.96530718e-01 -3.00626904e-01
6.33933425e-01 -1.04640186e-01 -9.39562976e-01 1.46481466e+00
4.27905589e-01 8.69337797e-01 -1.49276376e-01 6.96470976e-01
2.45583802e-01 7.83977509e-01 5.28882891e-02 -5.74865878e-01
9.62823331e-01 -1.31021488e+00 -7.49443352e-01 1.43131480e-01
8.99757683e-01 -5.54160595e-01 1.01522171e+00 6.21653795e-01
-1.09099507e+00 -4.05039072e-01 -8.76991451e-01 -2.26787001e-01
-2.44904056e-01 7.32153207e-02 9.51677501e-01 6.77636802e-01
-9.60816801e-01 8.81710649e-01 -1.70080137e+00 -4.61908370e-01
6.78087771e-01 5.07979035e-01 -3.06510061e-01 -7.24125877e-02
-2.88210094e-01 3.74876969e-02 3.90761465e-01 -2.90662408e-01
-1.00728357e+00 -5.62495291e-01 -1.01073873e+00 2.19214670e-02
6.91664279e-01 -1.74604237e-01 1.40361023e+00 -1.21547055e+00
-1.72540712e+00 5.38524687e-01 -4.26534206e-01 -5.54540038e-01
3.43126446e-01 -3.84364396e-01 6.62086019e-03 5.71136713e-01
-2.11418774e-02 7.06172109e-01 8.64692926e-01 -1.02526569e+00
-6.39543176e-01 -5.92827648e-02 2.73780882e-01 4.77805048e-01
-4.51359212e-01 -1.60696730e-02 -1.12932205e+00 -6.21168375e-01
1.98812813e-01 -1.00948811e+00 -4.73954350e-01 -1.01498127e-01
-1.95290372e-01 -3.49260569e-01 8.30340266e-01 -5.55467546e-01
1.40773189e+00 -2.39626598e+00 2.97232747e-01 4.56742287e-01
2.08132520e-01 -3.77316654e-01 1.14683449e-01 1.93999767e-01
1.03512523e-03 1.23500936e-01 -2.73415834e-01 -6.45944774e-01
-7.10418224e-02 5.10417998e-01 1.56818777e-01 8.71443629e-01
-6.48062006e-02 5.59690952e-01 -1.08046615e+00 -7.55372941e-01
2.30916217e-01 3.60853583e-01 -9.63392794e-01 9.68697146e-02
-2.60567755e-01 6.72863185e-01 -3.11473131e-01 5.75511456e-01
1.71353400e-01 -5.74287117e-01 5.89875996e-01 1.65738791e-01
-2.09776416e-01 4.87122178e-01 -1.26816225e+00 2.34527779e+00
-2.51318604e-01 7.60030389e-01 -6.88401759e-02 -1.27500796e+00
3.13183665e-01 2.66638190e-01 1.34994411e+00 -3.93596262e-01
-2.39938311e-02 -1.93248421e-01 -3.90749693e-01 -3.52658361e-01
6.23291731e-01 6.33644581e-01 -1.29823461e-01 8.97717714e-01
1.51950449e-01 4.18103874e-01 7.13646472e-01 5.64409494e-01
1.37590313e+00 6.69131100e-01 1.56185171e-02 -1.99206829e-01
1.01684615e-01 1.68773860e-01 8.00072014e-01 6.09180212e-01
-2.20388010e-01 6.28247976e-01 6.98078632e-01 -3.11613202e-01
-7.31425762e-01 -9.20516491e-01 3.49538237e-01 1.46807742e+00
2.80145288e-01 -9.51994479e-01 -1.06421316e+00 -1.02931547e+00
-3.30788374e-01 2.89257109e-01 -5.92313349e-01 1.21466391e-01
-8.21957111e-01 -5.22824526e-01 2.33684927e-01 8.06644738e-01
2.77011961e-01 -7.07476199e-01 -7.09258974e-01 1.99745104e-01
-2.84648716e-01 -1.06813288e+00 -8.38211060e-01 2.67419368e-01
-1.07343388e+00 -1.24287641e+00 -3.45286757e-01 -6.23210251e-01
9.35884953e-01 5.74580789e-01 9.47343111e-01 1.75196350e-01
-3.55760068e-01 7.95692623e-01 -5.20582020e-01 2.78449714e-01
1.14236772e-01 3.43179882e-01 -9.48317125e-02 9.17786211e-02
4.39001858e-01 -4.07950401e-01 -7.69520581e-01 3.99336308e-01
-9.54482317e-01 -3.11588887e-02 1.76213324e-01 5.71492195e-01
9.20890212e-01 1.36822268e-01 -6.75875973e-03 -1.03805423e+00
1.31771371e-01 -8.59301627e-01 -6.14546657e-01 7.81227574e-02
-4.07352000e-01 -4.82793376e-02 4.56979036e-01 -6.09757543e-01
-1.00674820e+00 6.26797557e-01 3.23934674e-01 -6.75389647e-01
-1.62986904e-01 2.46081129e-01 -3.65666375e-02 2.34954223e-01
3.93398672e-01 1.49024099e-01 -4.52262498e-02 -5.63888907e-01
2.62954891e-01 3.85108262e-01 4.37508225e-01 -7.77093172e-01
5.32324076e-01 7.11077273e-01 -3.91414732e-01 -7.72226155e-01
-6.16181135e-01 -1.08314729e+00 -1.18716490e+00 -2.40964666e-01
9.58439350e-01 -1.06845939e+00 -9.41496193e-01 2.20922425e-01
-7.54293680e-01 -8.35349739e-01 -3.92988741e-01 6.10447228e-01
-9.17416096e-01 6.83718801e-01 -7.72486329e-01 -6.23482525e-01
2.14349940e-01 -1.02953124e+00 1.04272282e+00 9.07316208e-02
-3.32501531e-01 -1.02803695e+00 -4.24533635e-02 6.01129949e-01
-1.44203514e-01 1.62302479e-01 4.98356551e-01 -5.49360573e-01
-8.63741517e-01 1.12508059e-01 -3.89971919e-02 1.50714070e-01
2.09931523e-01 1.98202997e-01 -6.88959956e-01 -4.31489795e-01
-3.46058190e-01 -2.80567974e-01 9.48332608e-01 3.94673228e-01
1.79151630e+00 -3.10294181e-01 -4.53613877e-01 7.61205614e-01
1.15322483e+00 2.83193409e-01 7.02352881e-01 2.09391326e-01
8.29425037e-01 3.97345036e-01 1.00058568e+00 8.06131124e-01
5.35392225e-01 6.07528806e-01 -2.67722178e-02 1.51478857e-01
9.45989415e-02 -6.80452362e-02 7.15068460e-01 9.41968560e-01
-3.75540763e-01 -2.14656144e-01 -8.07170689e-01 6.36096418e-01
-2.24514318e+00 -1.13840652e+00 6.39651716e-02 2.17100239e+00
6.94326341e-01 -9.47866812e-02 4.61375773e-01 3.07131618e-01
3.80786598e-01 2.31970668e-01 -5.46028733e-01 -1.20783545e-01
3.94591987e-01 1.84458926e-01 6.96701407e-01 2.82456815e-01
-1.47547090e+00 1.06794524e+00 6.50612688e+00 7.26626039e-01
-7.39146888e-01 2.75813550e-01 6.09406590e-01 -5.32167494e-01
8.95561948e-02 2.42348105e-01 -5.06761372e-01 6.34019673e-01
1.10961390e+00 1.09506734e-01 7.60372221e-01 8.48596632e-01
4.36975390e-01 -4.53073591e-01 -1.21683967e+00 1.08418870e+00
8.52487013e-02 -1.32227600e+00 -3.64195853e-01 1.59925103e-01
8.29261065e-01 -8.32626745e-02 -1.31469876e-01 1.78092137e-01
6.83847129e-01 -7.69435048e-01 5.94906747e-01 4.61177707e-01
5.20745277e-01 -8.73172998e-01 1.72593340e-01 2.42753282e-01
-1.53717029e+00 -1.05535604e-01 -3.31740314e-03 -5.91710098e-02
9.87487808e-02 3.26589942e-01 -5.93703687e-01 3.47007453e-01
1.02721167e+00 1.22556019e+00 -5.76071918e-01 9.23857570e-01
-4.96635921e-02 8.31654906e-01 -3.70561004e-01 3.76538306e-01
4.21000898e-01 -3.83841872e-01 7.02699274e-02 1.41526341e+00
1.59861639e-01 1.61364138e-01 8.23877573e-01 2.47978732e-01
-2.53512472e-01 1.49738982e-01 -4.71072704e-01 -4.19238448e-01
3.46129596e-01 1.20432007e+00 -1.33283544e+00 -6.09740138e-01
-5.42035997e-01 1.29991961e+00 2.78153539e-01 4.67693508e-01
-1.06067717e+00 -2.58624852e-01 6.69677973e-01 1.46810666e-01
5.87363183e-01 -8.54491889e-01 1.25686741e-02 -1.30087519e+00
-1.59174398e-01 -5.93180180e-01 7.13756263e-01 -2.61770487e-01
-7.87775040e-01 3.52933891e-02 7.32954293e-02 -1.26148868e+00
-3.75398368e-01 -2.89481014e-01 -3.38338405e-01 -7.60035738e-02
-1.30864727e+00 -7.92794824e-01 -2.64245331e-01 9.39410448e-01
1.01308262e+00 2.28347659e-01 5.07958233e-01 4.58480328e-01
-9.30028379e-01 4.26503688e-01 2.05423787e-01 2.73646533e-01
5.60161889e-01 -1.34600711e+00 5.98614961e-02 8.87029409e-01
5.02541602e-01 6.83840930e-01 2.13255659e-01 -5.72720647e-01
-1.46251738e+00 -1.08890522e+00 3.47572714e-01 -4.26520765e-01
6.28470302e-01 -6.19997501e-01 -7.69375682e-01 1.03437102e+00
4.01959717e-01 1.80165961e-01 1.19605923e+00 3.65192257e-02
-2.05462039e-01 2.35385671e-01 -7.97575176e-01 4.35763747e-01
1.43812895e+00 -4.35455889e-01 -2.19685003e-01 5.95249355e-01
6.02070034e-01 -4.08746600e-01 -9.61627722e-01 -7.62345642e-02
4.56179202e-01 -7.15318382e-01 7.37832129e-01 -5.27076960e-01
1.02327339e-01 -6.04395390e-01 -1.84180751e-01 -8.47479224e-01
-1.93666026e-01 -1.09195888e+00 -5.93032122e-01 1.05882430e+00
2.44356990e-01 -2.97957715e-02 1.11584973e+00 6.55709684e-01
-2.30121464e-01 -5.85570097e-01 -6.89355910e-01 -7.66190946e-01
-5.22556961e-01 -6.14879847e-01 2.24431515e-01 1.10570562e+00
1.11174621e-01 -4.38087173e-02 -2.63718843e-01 3.43907103e-02
5.64585924e-01 1.89803943e-01 8.40198040e-01 -1.05108917e+00
-4.07481909e-01 -1.49362445e-01 -3.53732497e-01 -1.52019060e+00
4.68594640e-01 -8.15014780e-01 1.98725104e-01 -1.37645602e+00
2.89023608e-01 -4.40008938e-01 -4.83068496e-01 6.51185393e-01
2.49824822e-01 2.20897406e-01 3.04582175e-02 5.07601857e-01
-1.56911528e+00 2.86305785e-01 7.75697410e-01 -1.26162305e-01
-5.86282134e-01 -7.79856667e-02 -4.80912238e-01 9.20221090e-01
7.37028480e-01 -5.10142803e-01 -6.44115806e-01 -4.61460471e-01
-1.68805525e-01 -2.28844151e-01 -1.93337798e-02 -1.12732375e+00
4.28162545e-01 -3.37695420e-01 3.95239413e-01 -5.81365526e-01
4.47127998e-01 -7.70676851e-01 -8.05644765e-02 1.18575439e-01
-3.64902735e-01 1.02035582e-01 -1.22822516e-01 9.95955884e-01
-1.82185456e-01 -5.01769930e-02 6.13867879e-01 -2.36725122e-01
-1.02853155e+00 4.81419981e-01 -8.85581911e-01 -4.71393690e-02
1.33291662e+00 -6.05955064e-01 5.84967360e-02 -2.26196975e-01
-9.35613394e-01 4.22137141e-01 7.53915191e-01 3.61000657e-01
5.56516945e-01 -1.07423794e+00 -1.78918503e-02 4.59607504e-02
-8.49248245e-02 1.72673374e-01 2.21326366e-01 1.14502013e+00
-5.44215679e-01 -3.89974238e-03 3.13732214e-02 -8.57294142e-01
-1.30867159e+00 5.49184978e-01 -1.44658864e-01 -3.00970465e-01
-6.51535511e-01 5.85240066e-01 1.48113482e-02 -3.88931744e-02
4.61638987e-01 -3.27614635e-01 -2.35375479e-01 3.08024019e-01
4.22406554e-01 6.44682229e-01 -1.76822454e-01 -5.88954926e-01
-2.81846285e-01 4.14778203e-01 -2.88274527e-01 -1.85485348e-01
1.29388738e+00 -2.98251390e-01 2.06491873e-02 6.98453963e-01
1.30430543e+00 -3.64611037e-02 -1.73463440e+00 -1.15897186e-01
4.64886367e-01 -6.45918131e-01 -9.73778740e-02 -6.27120584e-02
-1.17370820e+00 6.14615023e-01 3.24011087e-01 -1.71371680e-02
1.12357056e+00 8.14257637e-02 8.64161670e-01 6.28841877e-01
3.17400932e-01 -1.70673895e+00 6.07212782e-01 4.33281243e-01
-7.64294788e-02 -1.11274970e+00 1.35314822e-01 -4.80533361e-01
-7.53096163e-01 8.39756191e-01 5.65999508e-01 -1.31181166e-01
5.14488399e-01 2.20618054e-01 -2.57502317e-01 -9.52099264e-02
-8.73466492e-01 -2.37444952e-01 -1.75384283e-02 4.26086962e-01
5.14019012e-01 -1.16522931e-01 -6.76870719e-02 3.65403473e-01
3.25517088e-01 4.94711250e-02 5.20211816e-01 1.61091518e+00
-4.58647966e-01 -1.11938298e+00 -1.53731331e-01 4.46858048e-01
-6.81260526e-01 2.80082464e-01 -2.34659955e-01 5.54952800e-01
-4.74786060e-03 1.03680253e+00 5.32327592e-01 -2.22533524e-01
-4.14054543e-02 2.53919363e-01 3.98574144e-01 -8.57377291e-01
-5.36574662e-01 3.96262139e-01 -3.38578001e-02 -1.19710302e+00
-9.18348670e-01 -9.71519709e-01 -1.74323583e+00 -1.46224499e-01
-1.06434450e-01 3.21235090e-01 5.12904286e-01 8.85327578e-01
5.17701387e-01 4.79276389e-01 7.87292182e-01 -9.48970318e-01
1.34620920e-01 -6.45581484e-01 -5.43691456e-01 5.69396734e-01
9.37190801e-02 -7.55957663e-01 -1.82618871e-01 7.41755843e-01] | [8.534173011779785, 0.5103325247764587] |
4c7b65f9-e78d-47b8-8ca7-793c2f797c6c | boosted-cascaded-convnets-for-multilabel | 1711.08760 | null | http://arxiv.org/abs/1711.08760v1 | http://arxiv.org/pdf/1711.08760v1.pdf | Boosted Cascaded Convnets for Multilabel Classification of Thoracic Diseases in Chest Radiographs | Chest X-ray is one of the most accessible medical imaging technique for
diagnosis of multiple diseases. With the availability of ChestX-ray14, which is
a massive dataset of chest X-ray images and provides annotations for 14
thoracic diseases; it is possible to train Deep Convolutional Neural Networks
(DCNN) to build Computer Aided Diagnosis (CAD) systems. In this work, we
experiment a set of deep learning models and present a cascaded deep neural
network that can diagnose all 14 pathologies better than the baseline and is
competitive with other published methods. Our work provides the quantitative
results to answer following research questions for the dataset: 1) What loss
functions to use for training DCNN from scratch on ChestX-ray14 dataset that
demonstrates high class imbalance and label co occurrence? 2) How to use
cascading to model label dependency and to improve accuracy of the deep
learning model? | ['Monika Grewal', 'Muktabh Mayank Srivastava', 'Pulkit Kumar'] | 2017-11-23 | null | null | null | null | ['lung-disease-classification'] | ['medical'] | [ 1.52762264e-01 1.00237548e-01 -5.19889772e-01 -6.53744340e-01
-1.00166798e+00 -1.72871232e-01 -1.02825463e-01 1.57940000e-01
-2.80684441e-01 6.84380531e-01 2.15335023e-02 -8.43223810e-01
-2.03845292e-01 -7.60348558e-01 -4.36096847e-01 -5.54626107e-01
2.75943838e-02 7.82136023e-01 2.35899806e-01 1.07093424e-01
-4.06278104e-01 5.12489974e-01 -1.11190927e+00 7.68431664e-01
4.40395564e-01 1.25723946e+00 3.24400961e-02 1.09041214e+00
8.34927186e-02 1.47200680e+00 -4.87293899e-01 -3.74615252e-01
1.81676358e-01 -6.26408577e-01 -1.14233315e+00 1.47241652e-01
4.16203529e-01 -8.58418524e-01 -3.85061324e-01 6.43341541e-01
8.37250352e-01 -3.24520230e-01 6.92837715e-01 -6.82227850e-01
-6.42890692e-01 6.01537168e-01 -5.97216845e-01 6.51020348e-01
-2.15574160e-01 6.62731677e-02 7.98482239e-01 -6.08986914e-01
5.66478729e-01 7.23945916e-01 1.11457407e+00 8.07946324e-01
-4.24724042e-01 -5.49001396e-01 -4.75105941e-01 3.87176126e-01
-9.22765613e-01 2.89196849e-01 2.88432598e-01 -2.38233104e-01
7.60792613e-01 5.17519116e-01 3.68222207e-01 1.31091964e+00
2.33796999e-01 6.42953992e-01 9.76753473e-01 -4.25867856e-01
-1.38575569e-01 -2.21937776e-01 5.55883706e-01 1.11715531e+00
1.96308330e-01 1.22648939e-01 -5.99019118e-02 -1.93516359e-01
7.09857941e-01 5.85419297e-01 -1.50909826e-01 1.26909330e-01
-1.09016871e+00 9.00468707e-01 9.54406559e-01 3.60409707e-01
-4.51740682e-01 4.33199346e-01 6.52429938e-01 1.82489336e-01
3.25988203e-01 5.72244763e-01 -6.35821700e-01 2.91940242e-01
-7.44467854e-01 1.60132438e-01 4.12376881e-01 4.05691415e-01
5.27658015e-02 -3.42844874e-01 -3.17528427e-01 1.16115606e+00
1.50453866e-01 1.51404843e-01 6.63190603e-01 -9.16923285e-01
4.19050574e-01 6.74381554e-01 -5.27486622e-01 -2.46446311e-01
-8.99268270e-01 -8.02088857e-01 -1.10225546e+00 1.17476530e-01
4.09135282e-01 -4.06496197e-01 -1.22205198e+00 1.30432224e+00
2.62550741e-01 5.86206745e-03 -3.99388462e-01 9.76152301e-01
1.20890081e+00 2.04686418e-01 2.26799265e-01 7.95312747e-02
1.66543281e+00 -9.22465026e-01 -6.92171574e-01 1.99137226e-01
1.17420447e+00 -5.66908300e-01 9.84312236e-01 3.54229629e-01
-1.00258887e+00 -5.89641809e-01 -1.06302774e+00 -3.36647719e-01
-2.68046141e-01 5.67652464e-01 7.42701709e-01 6.54475987e-01
-9.31792855e-01 5.42152405e-01 -9.78032887e-01 -4.09033060e-01
8.30662072e-01 4.26921636e-01 -1.24889635e-01 -3.62797797e-01
-1.20774543e+00 1.01277542e+00 1.97065532e-01 -1.75394163e-01
-8.36292267e-01 -9.54229176e-01 -3.03366482e-01 1.01619065e-01
4.79786158e-01 -1.15718031e+00 1.45244038e+00 -6.77957058e-01
-1.15020370e+00 1.21080112e+00 5.32138467e-01 -5.23662090e-01
5.09218216e-01 -1.49082959e-01 -2.30580568e-01 4.51673597e-01
-1.86224729e-02 7.54451275e-01 2.80699104e-01 -1.05992782e+00
-7.30695665e-01 -4.08746570e-01 2.96668001e-02 -1.08126298e-01
-3.11004519e-01 1.25171021e-01 -4.37267840e-01 -7.68422723e-01
8.12156051e-02 -9.41928506e-01 -3.12998921e-01 3.12840641e-01
-5.80094397e-01 -3.66126269e-01 7.17329025e-01 -7.78704345e-01
1.21512163e+00 -1.78424847e+00 -3.68730187e-01 -4.37786020e-02
7.25152493e-01 4.00460899e-01 2.97176510e-01 1.09456703e-01
-5.02781212e-01 2.57407010e-01 -2.56494313e-01 -2.26660192e-01
-2.78555930e-01 3.96541923e-01 1.59898270e-02 3.33827198e-01
3.32963467e-01 1.06612968e+00 -5.62668324e-01 -7.54053831e-01
3.60556513e-01 3.51442635e-01 -4.29368615e-01 4.17741001e-01
-1.92214340e-01 3.23548168e-01 -3.65071088e-01 9.67052758e-01
4.58014965e-01 -1.02011001e+00 -1.06231853e-01 -2.61653811e-01
5.60069263e-01 1.24931768e-01 -5.15609145e-01 1.32469118e+00
-2.68611372e-01 3.80519360e-01 -9.37114656e-02 -8.82757664e-01
4.30169106e-01 4.96912539e-01 7.08762407e-01 -5.21420777e-01
5.98601520e-01 2.10598275e-01 2.10486934e-01 -1.04746485e+00
-2.85852432e-01 -3.83780956e-01 1.91305816e-01 6.22422874e-01
9.22355950e-02 9.21020508e-02 -2.45300248e-01 -1.62914649e-01
1.71806073e+00 -6.11670554e-01 2.29122654e-01 9.26837623e-02
4.34287637e-01 8.28013569e-02 2.14925736e-01 8.09811890e-01
-3.79897028e-01 1.11178124e+00 2.64365166e-01 -9.57451761e-01
-9.52785492e-01 -1.00047851e+00 -4.94287014e-01 1.00150061e+00
-3.43405157e-01 4.30883877e-02 -5.32176375e-01 -9.60007906e-01
-1.10111915e-01 1.87941596e-01 -9.92885768e-01 2.07053404e-02
-7.05406666e-01 -1.19397008e+00 8.78020048e-01 1.04608858e+00
7.11539328e-01 -9.37567949e-01 -5.91172695e-01 -8.17481726e-02
-1.11514941e-01 -7.86514342e-01 -8.83188173e-02 5.32710850e-01
-8.95076931e-01 -1.64026964e+00 -9.79215026e-01 -7.84265876e-01
5.57344437e-01 -1.88106582e-01 1.38310349e+00 7.25245178e-01
-8.12627077e-01 1.96191043e-01 -3.98260504e-01 -6.19171917e-01
-5.72925270e-01 4.25300986e-01 -5.50784349e-01 -5.38721383e-01
3.07599515e-01 -1.50398433e-01 -1.00613320e+00 -5.89017048e-02
-9.55951095e-01 3.85302748e-03 7.66985357e-01 9.79144096e-01
7.21710801e-01 8.86435155e-03 6.29357100e-01 -1.65758848e+00
6.09137654e-01 -7.04075873e-01 -9.26739443e-03 3.78020465e-01
-5.71437955e-01 -4.69870180e-01 5.04781246e-01 -4.20612060e-02
-8.43564153e-01 -1.92355514e-02 -9.84833002e-01 -5.48333049e-01
-3.64048362e-01 4.94551241e-01 4.58959490e-01 6.21311292e-02
8.61093640e-01 -3.36691886e-01 -2.91212916e-01 -4.91351634e-01
3.74769419e-02 6.47364914e-01 7.17240930e-01 -1.99739426e-01
2.49558643e-01 3.93857539e-01 3.70401919e-01 -2.28761494e-01
-1.42798126e+00 -4.20834124e-01 -6.22590959e-01 -2.13352531e-01
1.33542442e+00 -8.21208000e-01 -5.86122215e-01 3.85596663e-01
-8.04588735e-01 -2.79475152e-01 -3.20806324e-01 4.67793882e-01
-1.20758347e-01 1.30462036e-01 -1.17131758e+00 -3.43849719e-01
-6.79838359e-01 -1.11261451e+00 1.17749918e+00 -5.03765084e-02
-3.03356200e-01 -1.15926862e+00 1.71657935e-01 6.72001660e-01
4.19886708e-01 6.29239976e-01 1.45297122e+00 -7.44670093e-01
-3.96197438e-01 -3.66117924e-01 -4.36304569e-01 4.89765137e-01
1.19639769e-01 -6.78856904e-03 -1.01690316e+00 -2.30343401e-01
3.08529526e-01 -6.81252420e-01 1.13134909e+00 7.56596565e-01
2.01827931e+00 1.36352748e-01 -5.28564513e-01 9.61438656e-01
1.44510329e+00 1.05831213e-01 5.06799698e-01 2.67174631e-01
9.57967103e-01 3.30094844e-01 7.63662755e-02 1.19013660e-01
4.53799963e-01 7.66147673e-02 5.85586965e-01 -5.97885251e-01
-5.20601869e-01 2.24202648e-01 -6.33113265e-01 9.11739886e-01
1.45050988e-01 -3.10306102e-01 -1.31338537e+00 3.58556718e-01
-1.35115409e+00 -3.65826488e-01 -5.33942103e-01 1.40182292e+00
8.76038373e-01 7.65000135e-02 -2.85735875e-02 3.10886145e-01
4.86120611e-01 -2.20394626e-01 -5.33889055e-01 -1.11385398e-01
1.68577015e-01 7.60712087e-01 4.84791905e-01 -1.02157111e-03
-1.34474266e+00 5.88129126e-02 7.59989882e+00 5.93416810e-01
-1.39146411e+00 4.95784789e-01 1.17031014e+00 -8.85399580e-02
2.30732843e-01 -6.25640512e-01 -5.30324817e-01 4.61393118e-01
8.60704839e-01 5.16696572e-01 -3.32646191e-01 8.21046114e-01
-1.47991553e-01 3.92773636e-02 -1.11343491e+00 8.03353429e-01
-3.48658264e-02 -1.58572066e+00 -1.02700539e-01 -8.71265829e-02
8.35758388e-01 4.82052505e-01 3.17123413e-01 2.87824661e-01
4.66790110e-01 -1.21669376e+00 -1.69369459e-01 2.45338932e-01
1.07636571e+00 -5.27630389e-01 1.31085467e+00 1.53201580e-01
-5.75723469e-01 -2.13410452e-01 -1.21178187e-01 2.62863189e-01
-1.79819107e-01 6.22865081e-01 -1.17495298e+00 4.06515181e-01
9.40696776e-01 5.17941654e-01 -9.22975123e-01 9.90646303e-01
-6.37985915e-02 1.15576649e+00 -1.09175377e-01 2.10148871e-01
1.67485595e-01 4.39808011e-01 -1.59646660e-01 1.18752396e+00
2.49680161e-01 3.69352430e-01 9.67746228e-02 6.92054689e-01
-6.29782438e-01 -7.88497999e-02 -3.22774678e-01 3.38203907e-01
2.43780643e-01 1.30080414e+00 -8.64395380e-01 -6.25205457e-01
-4.37280685e-01 5.88663220e-01 1.88064948e-01 -7.28094205e-02
-9.88321960e-01 -1.78319309e-02 -3.39414575e-03 3.12447578e-01
1.18642353e-01 4.16603208e-01 -7.92167366e-01 -8.74303758e-01
-4.19148862e-01 -8.67376268e-01 8.70179236e-01 -8.93025696e-01
-1.71187627e+00 6.04466081e-01 -3.40504616e-01 -1.03269386e+00
-1.72172591e-01 -8.57349634e-01 -6.94716692e-01 5.54817259e-01
-1.62785900e+00 -1.31828392e+00 -6.41484380e-01 6.33719742e-01
4.45650220e-01 -1.16717860e-01 1.00453889e+00 8.57542813e-01
-5.46009660e-01 7.06630707e-01 1.94979142e-02 4.47247654e-01
8.45296979e-01 -1.59987664e+00 7.19334558e-02 4.33905214e-01
-3.41646343e-01 1.95783839e-01 1.59849182e-01 -3.96479636e-01
-7.93949306e-01 -1.63630354e+00 6.68509245e-01 -8.25789630e-01
1.93508908e-01 2.25227252e-01 -1.09495652e+00 6.53397739e-01
1.52263165e-01 4.21663284e-01 1.17986035e+00 -1.05847307e-01
-1.98915303e-01 5.05775623e-02 -1.27023876e+00 -1.71027593e-02
6.98898673e-01 -4.20268565e-01 -4.76518840e-01 7.27483273e-01
7.46671319e-01 -7.52055526e-01 -1.20514643e+00 8.61594677e-01
2.77924001e-01 -1.02983344e+00 1.06159246e+00 -5.60514092e-01
9.60838199e-01 7.52905905e-02 -3.90573367e-02 -8.94707918e-01
-3.02410722e-01 2.77546227e-01 -2.03885347e-01 7.93573558e-01
2.48031095e-01 -4.25780833e-01 9.96493936e-01 3.34664911e-01
-2.47006729e-01 -1.42448473e+00 -7.13603497e-01 -6.64190948e-02
4.23748910e-01 -3.08376402e-01 4.75817591e-01 1.05625153e+00
-7.50667334e-01 3.00337702e-01 -8.73560235e-02 -6.69430336e-03
3.41971964e-01 -1.00207530e-01 3.60641062e-01 -1.18280339e+00
-4.51417625e-01 -3.43486071e-01 -7.95010924e-02 -5.51505327e-01
-1.93109840e-01 -1.08527720e+00 -3.90770763e-01 -1.70706761e+00
4.62836832e-01 -5.23845255e-01 -6.88304842e-01 6.83540285e-01
-5.24775088e-01 5.75998962e-01 -4.08791974e-02 3.38585407e-01
-5.30472875e-01 -6.85731992e-02 1.52589989e+00 -1.34920910e-01
4.50532377e-01 2.57215112e-01 -6.35304570e-01 7.34008253e-01
5.24297774e-01 -5.98905683e-01 -3.35549086e-01 -7.54052103e-01
2.48352349e-01 3.59119773e-01 3.78317803e-01 -1.14506626e+00
7.15229437e-02 3.18478733e-01 1.01815772e+00 -1.18859160e+00
-1.76713392e-02 -9.23363268e-01 -2.78966818e-02 8.80306959e-01
-6.88106716e-01 1.19647235e-01 4.77957167e-02 3.27598304e-01
-2.99958557e-01 -2.26841033e-01 1.01086640e+00 -5.15795290e-01
-2.57106870e-01 5.50858676e-01 -4.11877126e-01 1.12097141e-05
9.74457026e-01 1.95899323e-01 -5.32931805e-01 -2.09218804e-02
-5.85782409e-01 3.81732672e-01 -4.59293649e-03 2.05847979e-01
5.55192113e-01 -1.31317401e+00 -7.43209660e-01 -2.97372118e-02
2.46182773e-02 3.45292598e-01 2.74316490e-01 7.26359487e-01
-1.12964332e+00 4.25848484e-01 -2.13815823e-01 -8.67872000e-01
-1.30378473e+00 4.95304525e-01 8.97261560e-01 -9.12593067e-01
-6.35492384e-01 1.26382983e+00 2.20371455e-01 -6.69282794e-01
3.77200186e-01 -7.37420321e-01 -1.72423780e-01 -2.87540346e-01
3.60639453e-01 2.55836427e-01 4.86574769e-01 -3.26700211e-02
-3.38939190e-01 3.06254297e-01 -3.20606321e-01 4.65883583e-01
1.50608301e+00 8.25725570e-02 -1.13556251e-01 1.98126107e-01
1.49039924e+00 -7.72801697e-01 -6.62478149e-01 -4.17303406e-02
-4.11862209e-02 -3.77924368e-02 2.59332120e-01 -1.33165121e+00
-1.45029426e+00 1.22883046e+00 1.20784545e+00 2.10490882e-01
1.26962352e+00 1.27057761e-01 1.20767689e+00 5.46639800e-01
-2.51846582e-01 -8.04692507e-01 3.36095244e-01 2.95109779e-01
3.94301802e-01 -1.42156196e+00 8.60035941e-02 -3.50880861e-01
-4.00578916e-01 1.31253684e+00 7.76319623e-01 -2.70356566e-01
9.12055314e-01 2.78682739e-01 4.57306266e-01 -8.95418823e-01
-9.49403942e-01 -1.05734035e-01 1.71662346e-01 4.41227794e-01
7.21714020e-01 1.72740325e-01 -1.77829728e-01 7.64498889e-01
-1.12977792e-02 3.89214128e-01 1.83485687e-01 8.86971176e-01
-3.13155651e-01 -1.03637779e+00 -4.07026947e-01 1.08164155e+00
-1.18592286e+00 1.07838430e-01 -5.93248487e-01 9.10646796e-01
6.79858565e-01 6.38454437e-01 -1.55865535e-01 -3.21314871e-01
2.77822405e-01 -8.27965792e-03 4.69440609e-01 -7.46685803e-01
-1.00348210e+00 -7.26812258e-02 -5.87458313e-02 -2.30128706e-01
-3.82933050e-01 -1.30514011e-01 -1.37628686e+00 -2.45167702e-01
-6.11556411e-01 -4.16431069e-01 4.30335253e-01 1.00874901e+00
3.06060035e-02 1.44908869e+00 5.20749986e-01 -1.74071744e-01
-5.59424102e-01 -9.53274250e-01 -3.74149054e-01 5.54411411e-01
5.85265875e-01 -4.84612405e-01 -1.09229438e-01 2.27989152e-01] | [15.193033218383789, -2.0208754539489746] |
90f209cc-107d-43f0-9e88-f499e837ced3 | contrast-stylize-and-adapt-unsupervised | 2306.09098 | null | https://arxiv.org/abs/2306.09098v1 | https://arxiv.org/pdf/2306.09098v1.pdf | Contrast, Stylize and Adapt: Unsupervised Contrastive Learning Framework for Domain Adaptive Semantic Segmentation | To overcome the domain gap between synthetic and real-world datasets, unsupervised domain adaptation methods have been proposed for semantic segmentation. Majority of the previous approaches have attempted to reduce the gap either at the pixel or feature level, disregarding the fact that the two components interact positively. To address this, we present CONtrastive FEaTure and pIxel alignment (CONFETI) for bridging the domain gap at both the pixel and feature levels using a unique contrastive formulation. We introduce well-estimated prototypes by including category-wise cross-domain information to link the two alignments: the pixel-level alignment is achieved using the jointly trained style transfer module with the prototypical semantic consistency, while the feature-level alignment is enforced to cross-domain features with the \textbf{pixel-to-prototype contrast}. Our extensive experiments demonstrate that our method outperforms existing state-of-the-art methods using DeepLabV2. Our code is available at https://github.com/cxa9264/CONFETI | ['Stephane Lathuiliere', 'Hongtao Lu', 'Huayi Zhou', 'Subhankar Roy', 'Tianyu Li'] | 2023-06-15 | null | null | null | null | ['style-transfer', 'contrastive-learning', 'contrastive-learning', 'unsupervised-domain-adaptation'] | ['computer-vision', 'computer-vision', 'methodology', 'methodology'] | [ 4.13311690e-01 1.67186171e-01 -7.69117549e-02 -5.66686213e-01
-8.52435589e-01 -5.55530190e-01 7.19291329e-01 -1.79996878e-01
-2.65587777e-01 5.57071984e-01 -1.16785608e-01 4.32642214e-02
5.23434617e-02 -6.29060447e-01 -7.02564776e-01 -5.99769473e-01
4.85354811e-01 3.38064730e-01 7.14181006e-01 -1.90536439e-01
3.25667202e-01 2.09716961e-01 -1.56865144e+00 4.15556341e-01
1.34367764e+00 1.15332460e+00 1.69842809e-01 1.69232801e-01
-3.56902122e-01 2.58553058e-01 -3.57985258e-01 -3.16114128e-01
5.97225964e-01 -5.46662152e-01 -8.13597918e-01 3.11715990e-01
7.10466146e-01 -6.20272234e-02 1.01432487e-01 1.31861973e+00
4.09570664e-01 -1.02342457e-01 5.80359936e-01 -1.49115527e+00
-6.53060675e-01 9.78187695e-02 -8.45015109e-01 -1.90200955e-01
1.29671976e-01 2.14720070e-01 9.82551932e-01 -9.63209212e-01
7.97199786e-01 1.13847244e+00 6.29397094e-01 3.21713477e-01
-1.42760551e+00 -5.88784933e-01 3.59233469e-01 1.86237246e-01
-1.39831102e+00 -1.47084340e-01 1.18040109e+00 -6.29823983e-01
8.04156303e-01 -1.09608822e-01 5.77953637e-01 1.12423837e+00
-6.32072613e-02 8.24826300e-01 1.48855281e+00 -6.33673608e-01
2.50851750e-01 3.05339098e-01 -1.07814431e-01 4.92758930e-01
1.40398622e-01 6.45553321e-02 -4.31994081e-01 2.52955724e-02
9.63184178e-01 -4.32261974e-01 -1.76452383e-01 -8.15926373e-01
-1.06162965e+00 7.71045625e-01 4.11089331e-01 3.86380821e-01
-1.28597707e-01 -2.32402354e-01 4.07255948e-01 1.55281141e-01
5.03729165e-01 3.95056874e-01 -7.25902915e-01 9.35872644e-02
-9.08054531e-01 2.36701906e-01 5.32655716e-01 1.07130206e+00
9.89571035e-01 -1.04334511e-01 -1.47199601e-01 1.10716212e+00
2.91189671e-01 2.77050495e-01 4.54579264e-01 -1.16177380e+00
4.80045766e-01 6.99124157e-01 1.65439680e-01 -8.19143593e-01
-2.73345381e-01 -3.26678813e-01 -4.33124334e-01 2.94334233e-01
6.41520202e-01 7.38271475e-02 -1.07269478e+00 1.83401847e+00
6.96827769e-01 2.57863075e-01 -1.21958628e-01 9.41382229e-01
7.11657047e-01 3.11843455e-01 2.25821674e-01 2.15753764e-01
1.42940748e+00 -1.22376657e+00 -4.84635949e-01 -4.63892043e-01
4.61918592e-01 -9.18861866e-01 1.51415765e+00 2.14804098e-01
-9.56782222e-01 -7.74798274e-01 -1.12871563e+00 -2.34757245e-01
-4.84090567e-01 3.03764611e-01 1.73282146e-01 4.97109890e-01
-9.42999363e-01 4.69623953e-01 -7.51865745e-01 -4.08943027e-01
4.63893145e-01 3.89426239e-02 -2.29245648e-01 1.08166046e-01
-1.17937219e+00 7.08918095e-01 4.57823664e-01 -1.40285328e-01
-2.66185790e-01 -7.90942669e-01 -9.31955397e-01 -1.45174712e-01
2.33297825e-01 -7.46545494e-01 1.11690676e+00 -1.52597797e+00
-1.47366869e+00 1.33632410e+00 2.69248281e-02 -1.30727768e-01
7.32537150e-01 -2.21102029e-01 -2.35011220e-01 1.12438604e-01
3.53186488e-01 1.16294885e+00 7.84210265e-01 -1.55970478e+00
-6.83823168e-01 -3.30788225e-01 -4.47257832e-02 3.39779615e-01
-3.22201639e-01 -9.78186876e-02 -6.58290863e-01 -8.63182485e-01
2.08529085e-01 -8.86870742e-01 -1.69903114e-02 4.25331235e-01
-2.88137525e-01 -1.35075808e-01 1.03058696e+00 -5.86113155e-01
9.85413015e-01 -2.26854730e+00 1.04846396e-01 5.64907826e-02
-1.08993575e-01 4.10773486e-01 -2.24976182e-01 2.39061460e-01
-1.06623054e-01 -2.55687207e-01 -8.04619074e-01 -4.77706492e-01
-5.14495037e-02 7.52131194e-02 -1.78059861e-01 3.68748724e-01
5.26250780e-01 7.77685642e-01 -8.79251659e-01 -8.28129292e-01
3.00357968e-01 5.53838074e-01 -5.24692774e-01 1.44842759e-01
-5.11895418e-01 7.20493495e-01 -4.73155737e-01 5.09567142e-01
1.10893416e+00 -2.27286413e-01 2.28658259e-01 -2.98780739e-01
-1.62227094e-01 3.18572521e-01 -1.24275625e+00 2.04676390e+00
-1.61226869e-01 3.54308486e-01 2.51565605e-01 -1.12445867e+00
9.38850224e-01 4.79484610e-02 5.45773268e-01 -9.95669425e-01
6.10373579e-02 3.39145452e-01 -7.06695542e-02 -2.57652491e-01
3.95156175e-01 -8.01067054e-02 -2.34851651e-02 3.81644592e-02
1.04626134e-01 -5.01398563e-01 1.34169415e-01 -2.12648753e-02
6.53897941e-01 8.72430801e-01 2.38047719e-01 -4.64502513e-01
5.77237964e-01 2.98042178e-01 8.48758936e-01 3.98561329e-01
-5.07664859e-01 1.11907148e+00 5.24614394e-01 -1.14147261e-01
-1.07819927e+00 -1.14434588e+00 -3.89423817e-01 7.79549837e-01
5.55761576e-01 -6.51336685e-02 -1.04297471e+00 -7.68132210e-01
7.11574554e-02 6.79366052e-01 -5.33912897e-01 3.38384882e-02
-5.26608109e-01 -3.60364854e-01 3.04739714e-01 6.79888427e-01
1.00527692e+00 -9.84583437e-01 -5.42410910e-01 8.72404203e-02
-1.87934250e-01 -1.35535169e+00 -4.83505815e-01 4.89188395e-02
-7.48186171e-01 -8.20790231e-01 -6.78428292e-01 -9.00463820e-01
7.35178053e-01 9.14566219e-02 1.20258105e+00 8.13539512e-03
-3.31837684e-01 1.42724797e-01 -4.48951215e-01 -1.62549630e-01
-2.29307309e-01 1.43773649e-02 -3.59272689e-01 -1.16434202e-01
2.86191136e-01 -6.26341164e-01 -7.75265872e-01 4.82731313e-01
-1.01047003e+00 4.94406283e-01 4.14153308e-01 8.09294105e-01
9.13063943e-01 -5.08476794e-01 5.76973677e-01 -9.80461895e-01
2.44033471e-01 -2.06883743e-01 -6.73205078e-01 2.55353659e-01
-5.19299150e-01 1.66186783e-02 4.58783656e-01 -3.90724927e-01
-1.23632002e+00 3.69461834e-01 -2.60735542e-01 -3.48728597e-01
-4.71197575e-01 9.97999832e-02 -5.38601398e-01 9.28473994e-02
6.43741012e-01 1.36858657e-01 -6.21334463e-02 -3.75299364e-01
4.76889282e-01 6.19742990e-01 4.88287508e-01 -8.71152520e-01
5.61931193e-01 6.85333014e-01 -2.76553154e-01 -4.56162274e-01
-8.43536258e-01 -4.16632652e-01 -1.08415663e+00 -7.98301995e-02
9.38143790e-01 -8.68581235e-01 1.93479121e-01 6.63225234e-01
-1.16156888e+00 -5.46647310e-01 -4.65943992e-01 1.93692192e-01
-8.27499688e-01 4.23435628e-01 -4.47145134e-01 -2.62898773e-01
-1.18136011e-01 -1.24645698e+00 1.33215070e+00 2.53067702e-01
-1.44191265e-01 -9.44073737e-01 1.56466603e-01 3.70031446e-01
2.17754528e-01 4.07455385e-01 8.05287540e-01 -5.43105543e-01
-2.76833177e-01 2.50178128e-01 -6.40762985e-01 5.01513183e-01
3.03511500e-01 1.49082663e-02 -1.10657036e+00 -1.40426293e-01
-3.19706276e-02 -1.69668883e-01 7.51090944e-01 4.43786472e-01
1.24310064e+00 2.04960704e-01 -3.79604250e-01 6.34798229e-01
1.25215447e+00 1.32539630e-01 7.05651522e-01 5.79366148e-01
7.60752797e-01 7.26811588e-01 8.18823934e-01 2.34117717e-01
5.61699212e-01 9.25046444e-01 1.92693621e-01 -5.13952494e-01
-5.46397388e-01 -2.09095985e-01 -2.08379235e-02 3.71250391e-01
3.34017515e-01 -1.23388492e-01 -9.28307295e-01 6.70469880e-01
-1.97375739e+00 -5.74538231e-01 -1.14021152e-01 2.11705375e+00
9.77808118e-01 1.92161798e-01 1.88154578e-01 -8.81186426e-02
8.07367027e-01 5.48346937e-02 -5.78797519e-01 -3.40880215e-01
-2.26649895e-01 9.24987998e-03 2.66561270e-01 4.70058501e-01
-1.23676372e+00 1.16076005e+00 5.67121410e+00 9.94375467e-01
-1.28851700e+00 1.89823210e-01 7.32967734e-01 2.52934754e-01
-3.91386539e-01 1.65219858e-01 -6.14325225e-01 5.21037877e-01
2.83161283e-01 3.18760633e-01 6.29655346e-02 9.21813667e-01
4.24579158e-03 -1.89631879e-01 -1.07557249e+00 7.69946277e-01
-1.71280373e-03 -1.05667341e+00 -1.28099158e-01 -5.88988662e-02
1.02874875e+00 -1.01236172e-01 1.96881756e-01 4.24695276e-02
1.44868866e-01 -6.52178645e-01 9.78120208e-01 2.67560959e-01
7.95363247e-01 -3.67727309e-01 3.36715341e-01 6.48738295e-02
-1.29117012e+00 1.78611651e-01 2.00010487e-03 1.48361936e-01
1.64355636e-01 6.39431000e-01 -5.68512201e-01 6.89634025e-01
8.62284839e-01 7.34478056e-01 -5.70496976e-01 7.42839515e-01
-3.15616399e-01 3.63768101e-01 -3.05438668e-01 6.09790325e-01
2.31488690e-01 -5.53900778e-01 4.68523026e-01 1.10960746e+00
2.68047810e-01 -2.31806874e-01 1.68152332e-01 1.28056073e+00
1.28078297e-01 1.44130709e-02 -3.91799033e-01 2.21689522e-01
5.39381564e-01 1.18093991e+00 -1.03574955e+00 -4.22264278e-01
-6.55461431e-01 1.27647328e+00 3.30305099e-01 2.96195835e-01
-9.59648013e-01 -3.73648316e-01 7.06081808e-01 3.47760051e-01
5.99494398e-01 -1.68895587e-01 -8.41906726e-01 -1.13738906e+00
3.22751850e-01 -7.64501691e-01 3.88140470e-01 -8.11195552e-01
-1.38591528e+00 5.81444323e-01 1.91433966e-01 -1.45662832e+00
-5.31211384e-02 -6.42872453e-01 -5.40977895e-01 8.41701806e-01
-1.74768686e+00 -1.46917224e+00 -3.27938050e-01 3.94355357e-01
6.07777417e-01 1.57452494e-01 5.31321585e-01 3.12408954e-01
-4.17289436e-01 6.61186755e-01 1.42984064e-02 -4.19437028e-02
9.41459477e-01 -1.16415668e+00 4.71755832e-01 9.07474697e-01
-1.93503395e-01 3.86916846e-01 7.44093299e-01 -6.30629838e-01
-8.21623504e-01 -1.07057047e+00 6.29815280e-01 -4.52401578e-01
4.86428887e-01 -4.33633238e-01 -1.03625929e+00 5.04622102e-01
2.44748577e-01 2.61603475e-01 2.86852628e-01 -2.63606250e-01
-5.64664781e-01 -6.05965406e-02 -1.47739434e+00 5.23982227e-01
1.15592182e+00 -3.40100557e-01 -5.14628232e-01 4.23780717e-02
7.27270603e-01 -4.53123868e-01 -7.48959661e-01 6.48411095e-01
3.93468916e-01 -1.10633016e+00 9.23987329e-01 -2.64921281e-02
5.98892272e-01 -6.18493795e-01 -3.62054147e-02 -1.27762675e+00
-9.32905897e-02 -1.35199547e-01 3.35321844e-01 1.49699473e+00
4.06430006e-01 -6.59355640e-01 6.43244326e-01 4.96024162e-01
-3.29674542e-01 -6.87608659e-01 -1.09852052e+00 -9.13425922e-01
4.49960649e-01 -1.76517725e-01 5.63237906e-01 1.06404006e+00
-1.96461037e-01 1.55077651e-01 -1.89657547e-02 1.60979971e-01
5.05715609e-01 3.36760789e-01 6.12860143e-01 -1.15280581e+00
-2.15072259e-01 -6.88944638e-01 -2.82701433e-01 -1.12556791e+00
1.41694188e-01 -7.74391651e-01 1.67243466e-01 -1.57899189e+00
2.01741382e-01 -5.58502197e-01 -1.70512512e-01 4.69657689e-01
-1.33506477e-01 2.65426844e-01 2.81554330e-02 1.45592734e-01
-4.66336250e-01 5.97297013e-01 1.41052485e+00 -7.42490031e-03
-2.32623443e-01 -4.55782771e-01 -7.39747703e-01 7.68721223e-01
8.14651906e-01 -2.92098194e-01 -4.47036266e-01 -5.87816894e-01
-1.89198002e-01 -4.27859366e-01 2.63880640e-01 -1.03344059e+00
-3.33485380e-02 -2.57833898e-01 1.85923994e-01 -4.16463464e-01
4.11847472e-01 -7.98272491e-01 -2.79572811e-02 2.34605685e-01
-3.16955149e-01 -1.99504167e-01 3.42528701e-01 3.45311970e-01
-3.36628377e-01 3.41456681e-02 1.19824994e+00 4.42514196e-02
-8.75528812e-01 4.36087251e-02 1.47476986e-01 2.17977554e-01
1.21634638e+00 -5.29791653e-01 -3.68919879e-01 4.09093574e-02
-3.91566396e-01 2.69443184e-01 9.21640754e-01 4.74895209e-01
3.88652474e-01 -1.28345048e+00 -5.34415841e-01 3.07630271e-01
4.73001808e-01 2.15495825e-01 2.49623507e-01 9.12337422e-01
-4.19798464e-01 8.95312950e-02 -4.91660178e-01 -8.65167141e-01
-1.06731904e+00 3.85026455e-01 5.11009812e-01 -1.84730902e-01
-5.36204815e-01 9.57575619e-01 5.88072419e-01 -7.60406375e-01
1.17877461e-01 -2.74357826e-01 1.36131391e-01 -1.33826062e-01
9.31400806e-02 -1.48297576e-02 1.02418482e-01 -7.42594898e-01
-4.49193329e-01 8.42001975e-01 -7.75424168e-02 -1.08599722e-01
1.16110647e+00 -2.75136381e-01 1.76995751e-02 2.71185219e-01
1.10829651e+00 -2.85146236e-01 -1.78457308e+00 -3.42263460e-01
-1.14037856e-01 -6.19200051e-01 -1.35466218e-01 -9.05336678e-01
-1.03931212e+00 8.77479672e-01 8.87455106e-01 -5.61927930e-02
1.22867095e+00 7.38670230e-02 8.31868410e-01 -3.33730996e-01
1.35829985e-01 -1.42171550e+00 1.84199587e-01 4.14027452e-01
8.22360516e-01 -1.45498788e+00 -1.06644914e-01 -8.84638965e-01
-9.55823541e-01 7.35857546e-01 1.04653966e+00 -2.58698285e-01
5.40897906e-01 3.28134865e-01 1.75315186e-01 -3.16183791e-02
-2.60392308e-01 -3.10154408e-01 3.73078138e-01 8.35511446e-01
5.73500097e-01 -8.02138299e-02 -6.43625021e-01 4.85957652e-01
3.01265921e-02 -9.04544517e-02 1.03531964e-02 1.09223485e+00
-2.66405106e-01 -1.41823030e+00 -3.57307166e-01 5.82137555e-02
-1.31320134e-01 -7.58479983e-02 -4.52238739e-01 8.56386662e-01
3.67510021e-01 6.85519457e-01 3.15144897e-01 -1.87622383e-01
3.95622522e-01 8.45692214e-03 5.60889006e-01 -4.79811788e-01
-3.30016643e-01 2.73783445e-01 -4.33690734e-02 -5.79935133e-01
-5.54746389e-01 -8.19843650e-01 -1.37155938e+00 2.12335378e-01
-9.80895981e-02 -2.80511051e-01 6.23559535e-01 9.12544906e-01
6.59672022e-01 3.85818928e-01 3.89867455e-01 -9.50824261e-01
-2.33422667e-01 -8.45948756e-01 -3.28947634e-01 7.58332253e-01
3.18047963e-02 -9.01646495e-01 -2.13074222e-01 2.55396008e-01] | [9.6831693649292, 1.2855991125106812] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.