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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
7f59e3a0-c940-49e9-8868-b7092d8af1b2 | object-aware-multi-branch-relation-networks | 2008.06941 | null | https://arxiv.org/abs/2008.06941v2 | https://arxiv.org/pdf/2008.06941v2.pdf | Object-Aware Multi-Branch Relation Networks for Spatio-Temporal Video Grounding | Spatio-temporal video grounding aims to retrieve the spatio-temporal tube of a queried object according to the given sentence. Currently, most existing grounding methods are restricted to well-aligned segment-sentence pairs. In this paper, we explore spatio-temporal video grounding on unaligned data and multi-form sentences. This challenging task requires to capture critical object relations to identify the queried target. However, existing approaches cannot distinguish notable objects and remain in ineffective relation modeling between unnecessary objects. Thus, we propose a novel object-aware multi-branch relation network for object-aware relation discovery. Concretely, we first devise multiple branches to develop object-aware region modeling, where each branch focuses on a crucial object mentioned in the sentence. We then propose multi-branch relation reasoning to capture critical object relationships between the main branch and auxiliary branches. Moreover, we apply a diversity loss to make each branch only pay attention to its corresponding object and boost multi-branch learning. The extensive experiments show the effectiveness of our proposed method. | ['Zhijie Lin', 'Zhou Zhao', 'Zhu Zhang', 'Nicholas Jing Yuan', 'Baoxing Huai'] | 2020-08-16 | null | null | null | null | ['video-grounding', 'spatio-temporal-video-grounding'] | ['computer-vision', 'computer-vision'] | [ 2.48397827e-01 2.98098265e-03 -6.54532313e-01 -3.88072491e-01
-8.87437105e-01 -4.34290409e-01 3.35646987e-01 4.36722308e-01
-1.43049255e-01 4.98281598e-01 3.23553145e-01 -7.57644400e-02
-4.68181968e-01 -7.71451414e-01 -8.26308608e-01 -4.91198301e-01
-3.91311832e-02 4.71029669e-01 8.72879446e-01 4.88170236e-03
1.18679263e-01 2.95356363e-01 -1.11822367e+00 6.51054025e-01
8.06007445e-01 1.02920687e+00 3.77127707e-01 3.58926892e-01
-2.66954392e-01 1.07889640e+00 -6.34035110e-01 -3.08734536e-01
-6.84050992e-02 -7.01642692e-01 -1.20128322e+00 3.28809619e-01
5.29308915e-01 -3.23356330e-01 -3.90110165e-01 9.80639577e-01
7.38952085e-02 2.48720080e-01 3.45599324e-01 -1.43138230e+00
-3.17396909e-01 8.61688614e-01 -8.35978866e-01 8.98563623e-01
4.55424577e-01 -2.33030893e-04 1.32792068e+00 -5.52624941e-01
9.02016222e-01 1.14858305e+00 2.54755735e-01 1.35759369e-01
-7.07811534e-01 -6.14872932e-01 8.23502779e-01 6.83273375e-01
-1.43331778e+00 -9.30892825e-02 1.10202849e+00 -2.70349115e-01
7.14804053e-01 3.18214208e-01 9.29759085e-01 7.14108706e-01
2.21586991e-02 1.13056278e+00 7.77107775e-01 -2.19632253e-01
-2.97292341e-02 -3.05314869e-01 2.09482685e-01 8.60229611e-01
-3.93738896e-02 -5.09295464e-01 -8.07535291e-01 7.27451444e-02
7.42663085e-01 4.67745401e-02 -2.49454275e-01 -4.08797115e-01
-1.58454311e+00 4.92361009e-01 6.44351482e-01 5.51626027e-01
-4.33024049e-01 2.40951970e-01 3.92177910e-01 -8.21832865e-02
5.06687403e-01 6.30624443e-02 -4.99075502e-01 3.78319472e-02
-1.12391090e+00 3.08213741e-01 2.77841687e-01 1.29366577e+00
8.51022065e-01 -6.19071007e-01 -6.51487052e-01 3.64602596e-01
2.25275964e-01 -8.84089842e-02 -4.19683456e-02 -8.01208377e-01
1.14124262e+00 1.05821204e+00 -1.31831408e-01 -1.50198424e+00
-3.91707569e-01 -4.38102514e-01 -6.36706233e-01 -5.75749695e-01
1.75583482e-01 2.18258098e-01 -6.97387993e-01 1.32088280e+00
8.00118029e-01 6.53471172e-01 -1.71319366e-01 1.04896307e+00
9.90838468e-01 7.45667517e-01 1.52123317e-01 -6.32655561e-01
1.54057586e+00 -1.18475235e+00 -6.82426512e-01 -1.10882431e-01
6.88609123e-01 -4.36985910e-01 9.09132838e-01 2.84570754e-02
-9.75379348e-01 -3.77180874e-01 -8.20318758e-01 -1.16892315e-01
-3.51038426e-02 -1.73809066e-01 5.07345557e-01 -1.44686401e-01
-3.82557690e-01 1.49109647e-01 -7.87528753e-01 -2.52029896e-01
6.25060141e-01 2.21844077e-01 -2.84885287e-01 -2.69742906e-01
-1.25253141e+00 5.84594846e-01 8.45571160e-01 8.45521465e-02
-8.89195621e-01 -3.92390996e-01 -7.86857843e-01 -1.14092037e-01
1.10178685e+00 -6.58487141e-01 1.01895845e+00 -7.71419287e-01
-7.02334881e-01 7.94408917e-01 -2.95455277e-01 -4.99534458e-01
1.30265519e-01 -1.78621039e-01 -5.37170351e-01 7.82350898e-01
4.55643326e-01 6.27751529e-01 5.26116610e-01 -1.22550082e+00
-1.09511542e+00 -5.11403620e-01 5.32932997e-01 4.78961408e-01
-3.43399912e-01 2.48446181e-01 -1.05584526e+00 -6.80659711e-01
6.40050709e-01 -4.20325518e-01 -1.93174817e-02 -1.69118688e-01
-5.79991341e-01 -4.25500512e-01 9.57498431e-01 -7.48031974e-01
1.66146767e+00 -1.92263532e+00 2.88746357e-01 1.02496697e-02
3.22969347e-01 -1.46815300e-01 -4.49531749e-02 2.07786411e-01
7.93775097e-02 7.31193870e-02 -1.19099736e-01 4.29113843e-02
-4.23150986e-01 4.29326952e-01 -1.63465738e-01 4.65684175e-01
2.09324330e-01 1.07395923e+00 -1.04651701e+00 -1.33047009e+00
-5.81508167e-02 1.17769822e-01 -3.64825726e-01 6.96243867e-02
-5.03619730e-01 4.57555145e-01 -8.25780511e-01 8.00765097e-01
5.15168488e-01 -3.57542485e-01 9.28386301e-03 -5.54278791e-01
1.21933244e-01 3.66345555e-01 -1.02047515e+00 2.13872671e+00
-7.54400641e-02 3.39212567e-01 -9.92971882e-02 -1.19931042e+00
8.05814981e-01 1.72062978e-01 7.48394728e-01 -8.67387176e-01
-2.11601090e-02 4.49869968e-03 -2.07829922e-01 -9.87668455e-01
3.43776226e-01 -1.12655506e-01 1.34617453e-02 1.41392052e-01
-2.09975004e-01 1.58548668e-01 2.71927774e-01 4.16234702e-01
1.04126263e+00 4.12172943e-01 2.69851089e-01 3.45271304e-02
6.99156106e-01 2.58322626e-01 8.12178910e-01 3.49977434e-01
-2.42073655e-01 5.48964202e-01 7.10303903e-01 -3.86582553e-01
-5.25671721e-01 -7.12341905e-01 2.36650944e-01 1.10233963e+00
1.00980830e+00 -5.25852919e-01 -5.82199395e-01 -1.07033086e+00
-4.16752130e-01 5.23797095e-01 -5.44879615e-01 -4.51666042e-02
-1.04128158e+00 -5.31922400e-01 1.26460522e-01 6.25276268e-01
6.34058714e-01 -1.06516981e+00 -7.66770840e-01 1.48852646e-01
-9.50608671e-01 -1.29796076e+00 -6.70541644e-01 -1.66372970e-01
-9.20062721e-01 -1.25822139e+00 -5.72115898e-01 -1.13591981e+00
5.75455308e-01 4.29565817e-01 1.21104383e+00 3.96594375e-01
9.49708298e-02 7.09036589e-02 -5.56667864e-01 1.42210191e-02
6.61682412e-02 3.64289284e-01 -3.81422490e-01 1.97659627e-01
4.00423408e-01 -3.12765568e-01 -7.01136708e-01 5.61218381e-01
-7.12869287e-01 2.67691463e-01 4.88124818e-01 4.34526414e-01
1.05935705e+00 4.45156366e-01 3.22888136e-01 -4.51526165e-01
1.62445202e-01 -5.31287909e-01 -5.26182830e-01 6.45150781e-01
-1.59247875e-01 -1.21586874e-01 8.22085962e-02 -4.49258208e-01
-8.92941892e-01 -4.47307415e-02 2.74929732e-01 -5.19506454e-01
-1.00680180e-01 7.73963332e-01 -5.38517833e-01 5.62319815e-01
2.37448081e-01 1.72489092e-01 -5.30312777e-01 -2.82385170e-01
2.48068213e-01 3.39750171e-01 6.19770467e-01 -6.88272834e-01
5.56301355e-01 5.57398856e-01 1.57104507e-01 -2.82913178e-01
-1.29565251e+00 -8.08018208e-01 -8.81388426e-01 -4.14963871e-01
1.17282510e+00 -8.18191111e-01 -5.67507863e-01 -1.77234396e-01
-1.33364630e+00 -1.27912253e-01 -1.70718163e-01 4.14362878e-01
-4.39088136e-01 3.96591842e-01 -3.03622931e-01 -6.38935387e-01
-7.01859146e-02 -9.88386631e-01 1.33140492e+00 5.20831868e-02
-3.54002863e-02 -6.72274292e-01 -5.53964563e-02 7.44380832e-01
-4.30124074e-01 2.22073436e-01 8.34279239e-01 -5.66798687e-01
-1.01420808e+00 -8.79240409e-02 -4.70131427e-01 -3.57163429e-01
6.73616305e-02 -6.24546334e-02 -3.66391927e-01 1.23244546e-01
-6.50469586e-02 8.58284906e-03 9.15910184e-01 3.09959918e-01
1.15189862e+00 -5.67477882e-01 -7.13573754e-01 4.31059659e-01
1.03651881e+00 2.54875779e-01 4.07616794e-01 5.09667158e-01
1.02373421e+00 9.06576157e-01 1.28065956e+00 1.72240555e-01
6.24786496e-01 9.31501985e-01 7.26392031e-01 7.00061470e-02
8.53088964e-03 -2.81772494e-01 3.09103783e-02 6.91652298e-01
-8.91101658e-02 -3.61483037e-01 -1.01019859e+00 8.22907507e-01
-2.18130136e+00 -1.18766236e+00 -4.39335227e-01 1.81884718e+00
8.12558711e-01 3.68117154e-01 3.39297026e-01 3.31633613e-02
8.65154743e-01 2.28227675e-01 -3.89429659e-01 5.91339171e-01
-1.45573318e-01 -4.37379956e-01 1.61261554e-03 2.88872659e-01
-1.19263697e+00 9.20453608e-01 5.00419235e+00 1.02313757e+00
-7.01004028e-01 3.30945551e-01 7.31070697e-01 -2.54143029e-01
-4.51679587e-01 2.27937043e-01 -8.49064708e-01 2.70664930e-01
3.83341581e-01 -4.45820764e-02 -8.10005069e-02 6.05202615e-01
1.65931270e-01 -3.13116193e-01 -1.00438821e+00 8.94989610e-01
2.22174719e-01 -1.42101514e+00 1.28026336e-01 -1.14264309e-01
6.37031019e-01 -4.16039348e-01 -4.21134651e-01 8.16246867e-02
-4.31792557e-01 -7.07689464e-01 9.04587507e-01 4.84583348e-01
5.62062085e-01 -8.73004436e-01 5.90063334e-01 5.36409497e-01
-1.67279756e+00 7.95993805e-02 -1.38022304e-01 1.77995250e-01
3.18381399e-01 4.18373764e-01 -5.40422499e-01 9.47767735e-01
7.22073376e-01 8.73507738e-01 -4.93771434e-01 1.03854573e+00
-2.03956872e-01 6.07021570e-01 -1.70078248e-01 1.28779095e-02
3.80088359e-01 -7.25229979e-02 6.46534741e-01 9.44146872e-01
1.05508529e-01 4.62283254e-01 5.87285817e-01 5.95011830e-01
7.74212703e-02 2.94791698e-01 -3.35345954e-01 3.03815871e-01
3.37909222e-01 1.14305186e+00 -1.24401391e+00 -4.16860521e-01
-5.05307257e-01 8.93893957e-01 4.06477600e-01 4.21428680e-01
-1.23912573e+00 -1.37160689e-01 9.41103995e-02 4.13599700e-01
4.65272635e-01 -8.82386267e-02 -2.08813235e-01 -1.18429196e+00
4.75029200e-01 -6.24341488e-01 7.81318128e-01 -8.83985102e-01
-9.57702398e-01 6.19535029e-01 4.76037204e-01 -1.38092351e+00
1.63219750e-01 -5.76707646e-02 -4.30992335e-01 4.42542553e-01
-1.40379083e+00 -1.60316515e+00 -4.41127479e-01 7.85578251e-01
7.94369996e-01 1.61921188e-01 2.37585053e-01 3.07560593e-01
-7.57028222e-01 2.74122119e-01 -6.30463243e-01 2.37009183e-01
3.06248367e-01 -8.16292584e-01 -9.83213186e-02 1.11291289e+00
4.80032772e-01 5.23620784e-01 5.50594449e-01 -9.93690431e-01
-9.96446133e-01 -1.12228346e+00 9.53521311e-01 -4.95187283e-01
5.95408976e-01 -2.30294615e-01 -1.07609951e+00 7.93761969e-01
9.43899229e-02 2.80688256e-01 2.78278530e-01 1.41094953e-01
-3.32877576e-01 -3.67930919e-01 -7.66170382e-01 4.67638433e-01
1.38604271e+00 -4.53335017e-01 -7.72975385e-01 5.55321038e-01
1.16004574e+00 -5.35685599e-01 -6.83728635e-01 5.80629647e-01
2.72596508e-01 -7.69918382e-01 9.83725309e-01 -8.44121158e-01
6.92736804e-01 -5.69386303e-01 -1.49666712e-01 -5.67139566e-01
-2.22060934e-01 -4.80182797e-01 -6.44562066e-01 1.49058342e+00
3.34252715e-01 5.05131073e-02 9.01591182e-01 1.82911754e-01
-2.74502754e-01 -1.08739066e+00 -9.99227643e-01 -8.67390096e-01
-3.45315397e-01 -5.16727149e-01 7.79942453e-01 9.67162967e-01
1.38387874e-01 5.71642280e-01 -2.18995631e-01 4.61235344e-01
3.39260966e-01 7.11123645e-01 5.98837197e-01 -9.37453032e-01
-2.98892498e-01 -3.00652087e-01 -3.86447787e-01 -1.33064067e+00
1.11353070e-01 -6.91076934e-01 1.70565292e-01 -1.94829047e+00
5.41499257e-01 -5.50387621e-01 -4.24038529e-01 4.48271722e-01
-5.31311274e-01 1.37908220e-01 5.25359213e-02 4.42299783e-01
-1.33338857e+00 4.06773090e-01 1.54152381e+00 -3.83142024e-01
-1.45241305e-01 3.96699123e-02 -5.09801567e-01 7.17110455e-01
6.10012889e-01 -6.07004344e-01 -7.09918797e-01 -5.13989210e-01
4.58206683e-01 4.21929359e-01 6.29732847e-01 -9.39744294e-01
3.69280934e-01 -6.10339344e-01 4.84559983e-02 -1.14404607e+00
1.80258945e-01 -1.02924323e+00 4.36636470e-02 2.17426106e-01
-4.23490316e-01 3.86442654e-02 -8.14830065e-02 7.97773719e-01
-3.85922104e-01 -2.41887018e-01 2.71747142e-01 -6.18517324e-02
-9.98436630e-01 6.63710773e-01 1.58101559e-01 2.68555820e-01
1.44819510e+00 -2.08769009e-01 -3.12442958e-01 -4.73198555e-02
-7.60031223e-01 4.45438772e-01 2.00952768e-01 4.07113314e-01
6.83285952e-01 -1.36514628e+00 -5.49081266e-01 -3.16899896e-01
4.41035926e-01 5.38553715e-01 4.69022691e-01 1.07737470e+00
-4.30827051e-01 2.80969501e-01 1.41084939e-01 -7.39224195e-01
-1.60459054e+00 9.94375169e-01 1.44773051e-01 -3.17582846e-01
-7.49817491e-01 1.07926416e+00 3.44860703e-01 1.59500211e-01
2.56122619e-01 -4.57519799e-01 -4.47782636e-01 2.87396491e-01
4.08949047e-01 2.07187667e-01 -1.33157879e-01 -1.00619602e+00
-6.98244929e-01 7.31348395e-01 -6.87316805e-02 -4.29680012e-02
9.98664081e-01 -2.46178284e-01 -3.44425261e-01 5.35611928e-01
1.06019914e+00 -1.45783722e-01 -1.19554853e+00 -3.95640135e-01
3.13915797e-02 -5.26807606e-01 -2.33820770e-02 -4.33902353e-01
-1.14018691e+00 7.56174147e-01 1.15983933e-01 2.20998868e-01
1.36187744e+00 6.09162152e-01 8.80846977e-01 1.56683207e-01
4.43239421e-01 -9.24273551e-01 4.17275846e-01 1.64928406e-01
8.33774388e-01 -9.95136023e-01 3.42857659e-01 -9.61022377e-01
-6.36945248e-01 6.75608814e-01 1.02453196e+00 3.72280657e-01
3.70198131e-01 -1.46921813e-01 -3.03963393e-01 -6.75986588e-01
-5.99701285e-01 -3.07244003e-01 5.64702332e-01 3.52813274e-01
1.23085201e-01 -2.06056893e-01 -2.45430827e-01 4.07814175e-01
2.27334827e-01 -3.79468873e-02 5.51769473e-02 1.07083571e+00
-4.97717798e-01 -7.83383012e-01 -3.90382290e-01 4.09453064e-01
-3.42045158e-01 -7.34255537e-02 -2.19053209e-01 6.68125033e-01
5.51214755e-01 8.82413030e-01 2.11375996e-01 -3.45362991e-01
2.99902022e-01 -1.20763890e-01 4.59680021e-01 -6.69280946e-01
-4.86230075e-01 3.69008243e-01 -4.23598066e-02 -4.07646477e-01
-9.71265078e-01 -8.47874820e-01 -1.63569283e+00 9.70869139e-02
-5.65838933e-01 1.08830206e-01 5.45054255e-03 1.16209888e+00
3.14183474e-01 6.82028472e-01 3.81661892e-01 -3.58813584e-01
1.95772827e-01 -7.18135238e-01 -3.69243950e-01 4.01013851e-01
3.10200721e-01 -7.34773934e-01 1.48254737e-01 3.37187737e-01] | [9.6812105178833, 0.6747910976409912] |
43c1b055-77f7-4770-a517-83a74d392d7e | learning-camera-aware-noise-models | 2008.09370 | null | https://arxiv.org/abs/2008.09370v1 | https://arxiv.org/pdf/2008.09370v1.pdf | Learning Camera-Aware Noise Models | Modeling imaging sensor noise is a fundamental problem for image processing and computer vision applications. While most previous works adopt statistical noise models, real-world noise is far more complicated and beyond what these models can describe. To tackle this issue, we propose a data-driven approach, where a generative noise model is learned from real-world noise. The proposed noise model is camera-aware, that is, different noise characteristics of different camera sensors can be learned simultaneously, and a single learned noise model can generate different noise for different camera sensors. Experimental results show that our method quantitatively and qualitatively outperforms existing statistical noise models and learning-based methods. | ['Hwann-Tzong Chen', 'Yu-Lin Chang', 'Chia-Ping Chen', 'Yu-Lun Liu', 'Hung-Jin Lin', 'Ke-Chi Chang', 'Ren Wang'] | 2020-08-21 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/4617_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123690341.pdf | eccv-2020-8 | ['noise-estimation'] | ['medical'] | [ 1.52385801e-01 -7.03221560e-01 3.59156728e-01 -3.27860653e-01
-8.30201864e-01 -5.52988172e-01 4.31334615e-01 -2.50371873e-01
-3.68792832e-01 2.35267103e-01 2.03868300e-01 6.94702193e-02
-2.15162426e-01 -6.58073604e-01 -6.03958428e-01 -9.29968834e-01
5.26646435e-01 6.31090105e-02 5.35782337e-01 1.66316628e-01
-7.85325170e-02 3.23114842e-01 -1.35279655e+00 -1.59318626e-01
5.24844408e-01 1.03059936e+00 7.23657668e-01 9.54358935e-01
-1.81703448e-01 9.52743769e-01 -6.77843928e-01 5.27699590e-02
4.18715864e-01 -4.02695298e-01 1.95477769e-01 5.10492444e-01
3.35664123e-01 -1.08301148e-01 -8.59387279e-01 1.78957033e+00
4.85340536e-01 -1.66733526e-02 5.02538085e-01 -9.95278299e-01
-7.30738163e-01 6.88546538e-01 -5.88249981e-01 1.86372221e-01
-5.41518778e-02 5.70096850e-01 3.76711786e-01 -4.60059494e-01
2.72533238e-01 1.35010362e+00 6.31530285e-01 7.36587882e-01
-1.30053830e+00 -6.17493689e-01 3.05197537e-01 -1.51514232e-01
-1.28075767e+00 -2.18204930e-01 1.18116355e+00 -5.26946366e-01
1.94392294e-01 -1.04278415e-01 3.85398209e-01 1.39285326e+00
3.13063353e-01 6.06900573e-01 1.40600467e+00 -1.43174529e-01
5.09110689e-01 -1.20681264e-01 9.24615115e-02 2.24690050e-01
7.03447402e-01 2.94173121e-01 -3.39691222e-01 -6.94887787e-02
1.04543757e+00 2.38122970e-01 -3.02538246e-01 -2.07365483e-01
-1.06212676e+00 6.19204879e-01 1.94783181e-01 3.29734325e-01
-4.57561016e-01 6.46064401e-01 8.06531981e-02 6.34210557e-02
3.04966152e-01 3.25891599e-02 -2.46233493e-01 -2.92586774e-01
-7.71318436e-01 2.05347419e-01 7.32195377e-01 1.12463295e+00
8.74260664e-01 5.09413898e-01 4.25845794e-02 7.82041669e-01
4.79191810e-01 1.21212482e+00 4.25293326e-01 -1.17789805e+00
5.06582223e-02 1.68638229e-01 3.04455999e-02 -1.20200181e+00
-3.74465704e-01 -3.39994192e-01 -1.32292616e+00 3.21168244e-01
1.75806239e-01 -4.47642207e-01 -1.22344553e+00 1.51216555e+00
-1.93861052e-01 7.28683352e-01 1.19185699e-02 8.33945155e-01
9.72052872e-01 5.26952386e-01 -1.17917724e-01 -3.60916644e-01
8.51631105e-01 -6.59457564e-01 -1.09426475e+00 -6.38797641e-01
-2.29280233e-01 -9.15528178e-01 7.36128569e-01 7.77812839e-01
-8.21506917e-01 -5.68388224e-01 -7.73457646e-01 4.46632147e-01
1.92133754e-01 1.21025201e-02 2.00936168e-01 7.56209910e-01
-9.35155749e-01 5.33896238e-02 -1.06412280e+00 -3.37840557e-01
3.13497484e-01 -2.15875223e-01 7.83287734e-02 -4.43079412e-01
-5.11623144e-01 4.63991821e-01 1.29536167e-01 2.80856252e-01
-1.14344501e+00 -4.27537560e-01 -7.72641063e-01 -2.34497011e-01
8.64052415e-01 -7.68955350e-01 1.29337502e+00 -6.91298306e-01
-1.49206305e+00 2.21451432e-01 -3.09090167e-01 -3.01389754e-01
4.53726292e-01 -3.34276736e-01 -5.58060586e-01 -9.37645733e-02
-1.18015438e-01 -1.37758832e-02 1.28262019e+00 -2.02648807e+00
-3.66766542e-01 -1.92428216e-01 -3.23815465e-01 -1.43642336e-01
3.36967632e-02 1.75731070e-03 -8.95780206e-01 -6.88601792e-01
4.45102930e-01 -8.51026952e-01 -9.76682484e-01 8.33841227e-03
-6.19991839e-01 2.55308777e-01 8.89736056e-01 -6.89659566e-02
1.16231859e+00 -2.08062792e+00 -1.56740621e-01 5.42922854e-01
4.65131074e-01 2.63211936e-01 -3.30073893e-01 5.27551323e-02
3.15367669e-01 6.17546663e-02 -2.86782146e-01 -3.20089817e-01
-2.37653956e-01 5.63566208e-01 -6.73270375e-02 1.84916705e-01
3.17884497e-02 7.10281074e-01 -9.98442948e-01 -5.82988672e-02
6.28931820e-01 5.21136045e-01 -2.36116871e-01 2.38635808e-01
-2.28718057e-01 4.91480321e-01 -4.86235529e-01 6.27482593e-01
1.02893662e+00 -1.06191307e-01 9.97065157e-02 -5.23938119e-01
1.34201795e-01 -4.90652978e-01 -1.87229943e+00 1.57781601e+00
-3.46835673e-01 5.21358848e-01 2.90337443e-01 -7.40410805e-01
1.07689118e+00 -7.40883201e-02 3.39313388e-01 -4.15492594e-01
3.38153780e-01 1.41662270e-01 -3.37308459e-02 -8.21712792e-01
1.99912637e-01 -1.60124451e-01 -2.82205790e-02 3.44052687e-02
5.69068827e-03 -6.58685863e-01 -3.47667634e-02 5.52934706e-02
1.50052834e+00 -3.02071661e-01 1.97095096e-01 -9.42649245e-02
2.69567400e-01 -3.32424432e-01 9.20590937e-01 1.43848467e+00
-4.47547764e-01 1.13788545e+00 1.00502044e-01 -3.07670146e-01
-8.67653728e-01 -1.24012613e+00 2.40872324e-01 2.74919868e-01
4.28094506e-01 -3.52994144e-01 -8.51058781e-01 -4.14106041e-01
-2.27875873e-01 3.23579341e-01 -3.81954491e-01 6.47242963e-02
-3.83212239e-01 -7.95381010e-01 3.03924859e-01 4.38343465e-01
4.87090111e-01 -7.64985263e-01 -3.58543754e-01 2.87298709e-01
-1.11196555e-01 -1.55073225e+00 -3.81804913e-01 1.78964317e-01
-6.93521321e-01 -1.23085189e+00 -1.15311056e-01 -3.69163126e-01
5.60779393e-01 7.93072820e-01 1.28573263e+00 -1.33133642e-02
-5.46085775e-01 9.93461132e-01 -2.88222432e-01 -8.42251539e-01
-5.22307992e-01 -6.43054008e-01 2.93835402e-01 1.50773942e-01
7.01805592e-01 -7.13665605e-01 -4.63683009e-01 1.11310937e-01
-1.51145601e+00 -4.99224782e-01 8.03464770e-01 7.87217319e-01
9.46702242e-01 7.68163443e-01 -9.85722803e-03 -9.04273331e-01
7.31606126e-01 -4.23989594e-01 -9.01393414e-01 8.25383738e-02
-3.63457292e-01 3.35628837e-02 5.99379718e-01 -8.18452418e-01
-1.08167648e+00 4.88600880e-01 -1.89980157e-02 -8.48058343e-01
-4.92200196e-01 5.20974576e-01 -4.90622044e-01 -1.49176836e-01
8.28441620e-01 2.98781157e-01 -4.58053686e-02 -5.04882634e-01
3.35416973e-01 4.15015697e-01 7.57641733e-01 -5.07580578e-01
1.35150909e+00 6.60188437e-01 5.72396889e-02 -1.14205289e+00
-7.59775400e-01 -5.22839546e-01 -3.30594838e-01 -2.41920918e-01
7.54621625e-01 -1.03001022e+00 -4.28130925e-01 9.44239497e-01
-1.26254749e+00 -1.65941596e-01 -4.21121001e-01 5.80810785e-01
-3.94170225e-01 3.52492988e-01 -4.90888596e-01 -1.20020747e+00
4.06130761e-01 -1.53847098e+00 1.02554798e+00 5.18489242e-01
2.59306878e-01 -1.15764213e+00 1.18482694e-01 -7.88142979e-02
5.59912086e-01 5.41496500e-02 3.60328704e-01 -2.26201028e-01
-9.90546465e-01 -2.17715219e-01 -3.66233915e-01 7.66330957e-01
3.22798967e-01 9.89394635e-02 -9.61706161e-01 4.93914746e-02
7.14551747e-01 -4.53001149e-02 9.95575368e-01 8.83148313e-01
1.36685300e+00 1.85673330e-02 -1.46621048e-01 5.17179608e-01
1.94416142e+00 2.69453555e-01 7.94564247e-01 1.71495259e-01
9.13101256e-01 9.33299586e-02 6.74340799e-02 5.73659241e-01
9.99809131e-02 3.22309494e-01 5.75378478e-01 -6.20443933e-02
-2.71452248e-01 -2.01573640e-01 2.73074627e-01 8.55277061e-01
2.48976797e-01 -4.26388323e-01 -8.44823956e-01 4.25389826e-01
-1.86552596e+00 -9.43087339e-01 -3.45843136e-01 1.83179045e+00
4.00157332e-01 1.14657104e-01 -3.41284275e-01 -1.08177200e-01
6.36797965e-01 2.23851308e-01 -7.98165202e-01 3.79180789e-01
-5.53479910e-01 4.62620109e-02 6.54744864e-01 5.15230119e-01
-1.08489645e+00 6.98822737e-01 7.92522001e+00 8.17939520e-01
-9.54600573e-01 8.88225362e-02 4.26326931e-01 -8.06979015e-02
-3.97261322e-01 -1.24027029e-01 -7.34317839e-01 6.22776389e-01
4.63778079e-01 -5.84478676e-02 4.84020025e-01 7.93383956e-01
4.91305262e-01 -3.93605232e-01 -7.68109322e-01 1.51995611e+00
2.75395453e-01 -1.10955524e+00 1.77051663e-01 2.30692234e-03
8.16616058e-01 2.62677789e-01 9.47965235e-02 -1.14694819e-01
8.77438307e-01 -9.21789885e-01 4.40312803e-01 1.08117378e+00
8.25504661e-02 -4.33255255e-01 8.80894542e-01 6.08081222e-01
-8.92043948e-01 -1.94158286e-01 -7.94453979e-01 -1.28960935e-04
1.35513827e-01 1.41662431e+00 1.05139790e-02 3.69208574e-01
9.87924337e-01 6.84291899e-01 -5.96104801e-01 1.26550901e+00
-3.31033796e-01 9.35949981e-01 -4.25906420e-01 5.86809814e-02
-4.80012745e-02 -3.97737622e-01 8.63944054e-01 1.20446944e+00
6.62870288e-01 1.66553408e-01 5.31357169e-01 8.37086916e-01
-3.77279706e-02 -4.38122272e-01 -8.72287035e-01 2.68048674e-01
5.09389997e-01 1.41501188e+00 -5.86818576e-01 3.40414234e-04
-7.72016883e-01 7.45565057e-01 -2.57228553e-01 7.17195749e-01
-5.28734386e-01 4.27084789e-02 9.73120332e-01 1.25144040e-02
3.00668538e-01 -5.32022238e-01 -3.50372791e-01 -1.36481285e+00
6.16153032e-02 -8.84216905e-01 -1.81624919e-01 -8.75353456e-01
-1.99351943e+00 6.16013050e-01 -2.37319529e-01 -1.51182628e+00
-7.20984861e-03 -7.76357770e-01 -7.47562647e-01 6.32594347e-01
-1.20261621e+00 -1.23311603e+00 -7.15160370e-01 4.95249510e-01
5.74669480e-01 -3.32046688e-01 4.59181190e-01 2.40367800e-02
-4.64690864e-01 1.92102239e-01 4.08251941e-01 2.48906821e-01
5.79116642e-01 -1.15100086e+00 2.09515065e-01 1.43310094e+00
3.17506582e-01 4.78798389e-01 9.15115833e-01 -7.36469507e-01
-1.56589508e+00 -1.25894356e+00 5.15717342e-02 -6.47307515e-01
1.01433587e+00 -4.39620018e-01 -9.39142108e-01 4.25427824e-01
7.02207834e-02 5.26895821e-01 4.57174838e-01 -1.69483900e-01
-4.18312103e-01 -3.25965375e-01 -9.59697127e-01 4.79507565e-01
1.02175188e+00 -4.79751408e-01 -1.08039990e-01 2.44351216e-02
7.02669621e-01 -3.79267812e-01 -3.99757802e-01 3.34089160e-01
1.96984168e-02 -1.06974769e+00 8.64261866e-01 -7.74840340e-02
1.96994513e-01 -8.38978827e-01 -5.70552468e-01 -1.66083336e+00
-3.66422892e-01 -5.41045725e-01 -1.03803433e-01 1.17311692e+00
1.21117681e-01 -2.06591323e-01 6.99765563e-01 6.27290070e-01
-8.86818394e-03 -9.54097062e-02 -8.00726295e-01 -1.10404754e+00
2.57877745e-02 -1.24243593e+00 3.49236459e-01 5.43601215e-01
-8.66809726e-01 8.65375325e-02 -5.88351309e-01 5.75475037e-01
1.07112527e+00 -5.63691854e-01 8.46627355e-01 -1.27211392e+00
-4.43358421e-01 -3.85724068e-01 -7.49689102e-01 -1.17320514e+00
-1.02903381e-01 -1.65482789e-01 7.34781265e-01 -1.53810215e+00
3.46801400e-01 -2.22937018e-01 -2.58603781e-01 -1.73309729e-01
-4.73048091e-01 2.49572173e-01 2.16903985e-01 1.44616947e-01
-7.05625951e-01 3.76793087e-01 9.70691681e-01 -2.83817530e-01
5.14203720e-02 4.59510200e-02 -8.97806883e-01 1.15471363e+00
4.55522895e-01 -6.15949154e-01 -6.21655345e-01 -7.68649697e-01
2.00585663e-01 -3.16661298e-01 6.03217006e-01 -1.31642103e+00
7.83156991e-01 -4.32368159e-01 3.58963877e-01 -4.28125143e-01
3.86740975e-02 -1.22984087e+00 2.36314461e-01 2.00191855e-01
1.50768518e-01 -1.86277494e-01 1.08489275e-01 1.22107816e+00
-4.22203332e-01 -3.27983469e-01 7.66474009e-01 -3.03544164e-01
-7.75788307e-01 5.45454383e-01 -8.29318643e-01 1.55677572e-01
7.20945895e-01 -2.04285420e-02 -2.48430029e-01 -7.28142321e-01
-7.14140236e-01 -1.36414543e-01 3.86552364e-01 3.98110151e-01
8.65423203e-01 -1.34403753e+00 -6.51965976e-01 2.51231492e-01
2.74904519e-01 3.72965217e-01 3.98025453e-01 2.50407428e-01
-8.84081796e-02 -3.31241131e-01 4.00250226e-01 -9.95750368e-01
-9.25660014e-01 5.03150582e-01 5.03638506e-01 -7.18561038e-02
-4.25696164e-01 7.79974937e-01 2.87928581e-01 -5.64139962e-01
3.57752770e-01 -4.91614431e-01 2.22408533e-01 -6.53893471e-01
8.75181556e-01 3.37944567e-01 -1.70080662e-01 -4.38434511e-01
6.59461245e-02 8.58709753e-01 2.59659886e-01 -1.18502215e-01
1.19337535e+00 -3.87653708e-01 5.69872092e-03 8.59809875e-01
7.69850612e-01 2.08184659e-01 -1.49254620e+00 -9.01854634e-01
-3.49616855e-02 -6.26608908e-01 3.00232232e-01 -7.48764932e-01
-1.29059601e+00 6.95585489e-01 9.16845262e-01 2.74522901e-01
1.53954494e+00 3.24381292e-02 5.19737124e-01 4.26900297e-01
4.38016087e-01 -1.22401869e+00 3.17463517e-01 6.28244638e-01
5.49461842e-01 -1.63943160e+00 -1.83435678e-01 -7.34341681e-01
-4.69154894e-01 9.32190180e-01 7.76978552e-01 -2.49780431e-01
1.25427699e+00 1.11641622e+00 5.67022026e-01 -1.02305144e-01
-6.18620872e-01 -3.75641495e-01 9.38714743e-02 1.09797001e+00
1.50807817e-02 1.57704316e-02 3.75047296e-01 9.83878195e-01
1.45796299e-01 1.23396464e-01 7.77753174e-01 8.21323395e-01
-5.26892126e-01 -1.15566635e+00 -6.66626334e-01 5.57865918e-01
-3.19654226e-01 1.53197162e-03 -2.07616687e-01 3.49428535e-01
3.34509969e-01 1.35528302e+00 -9.00764018e-03 -6.60037279e-01
5.75094461e-01 -4.62573588e-01 3.22085053e-01 -6.57614052e-01
-6.33305907e-02 5.88442445e-01 -5.01490295e-01 -5.92770636e-01
-6.77882493e-01 -5.71792483e-01 -7.51477003e-01 -1.22347049e-01
-2.61025667e-01 -2.08702624e-01 8.01542819e-01 9.92065191e-01
1.10967405e-01 7.08551168e-01 6.83644414e-01 -7.20522761e-01
-5.00832856e-01 -7.83641577e-01 -8.42975020e-01 4.36399460e-01
4.10235345e-01 -4.77398366e-01 -7.75937676e-01 3.10426295e-01] | [11.387425422668457, -2.479482889175415] |
d723ca85-c86b-424d-ae97-f813eb5a0c5f | the-weltmodell-a-data-driven-commonsense | null | null | https://aclanthology.org/L14-1351 | https://aclanthology.org/L14-1351.pdf | The Weltmodell: A Data-Driven Commonsense Knowledge Base | We present the Weltmodell, a commonsense knowledge base that was automatically generated from aggregated dependency parse fragments gathered from over 3.5 million English language books. We leverage the magnitude and diversity of this dataset to arrive at close to ten million distinct N-ary commonsense facts using techniques from open-domain Information Extraction (IE). Furthermore, we compute a range of measures of association and distributional similarity on this data. We present the results of our efforts using a browsable web demonstrator and publicly release all generated data for use and discussion by the research community. In this paper, we give an overview of our knowledge acquisition method and representation model, and present our web demonstrator. | ['Alan Akbik', 'Thilo Michael'] | 2014-05-01 | null | null | null | lrec-2014-5 | ['open-information-extraction'] | ['natural-language-processing'] | [ 1.92141578e-01 3.62122357e-01 -5.29021621e-01 -3.55010033e-01
-9.55071867e-01 -1.01751363e+00 8.34732592e-01 6.07372522e-01
-5.19758224e-01 1.29541183e+00 7.52943575e-01 -2.07564667e-01
-2.66169637e-01 -6.90558016e-01 -5.65117955e-01 4.90029827e-02
2.15667635e-01 6.61935031e-01 1.80655375e-01 -5.44323981e-01
7.75018930e-01 2.87731975e-01 -1.41903961e+00 4.09788102e-01
1.25439930e+00 7.76988149e-01 1.28625602e-01 2.69597858e-01
-4.65791553e-01 1.41197956e+00 -8.87289047e-01 -1.00008559e+00
-3.17361593e-01 -3.47455889e-01 -1.39057982e+00 -5.43455124e-01
3.74237746e-01 -6.06966671e-04 -2.36128584e-01 1.15502000e+00
3.09000701e-01 1.88993961e-01 9.33482230e-01 -1.15138328e+00
-1.18136334e+00 1.44840705e+00 3.15649472e-02 8.00497770e-01
6.72817945e-01 -2.95500517e-01 1.45813692e+00 -5.91736138e-01
1.33501780e+00 1.04047108e+00 4.90537494e-01 5.31220257e-01
-9.15081382e-01 -6.59153402e-01 -3.86850715e-01 7.27796018e-01
-1.15763593e+00 -5.94847322e-01 9.83417332e-01 -3.37312788e-01
1.34026265e+00 1.06913395e-01 7.23780215e-01 1.64029527e+00
-2.83168461e-02 8.73493493e-01 1.23840201e+00 -6.75205171e-01
1.44144863e-01 1.03585243e-01 6.56574428e-01 4.46368784e-01
6.14136755e-01 -4.01591301e-01 -7.57445872e-01 -4.34406221e-01
4.91242856e-01 -6.40865922e-01 -1.14773899e-01 3.79639864e-02
-1.12755263e+00 9.57217157e-01 1.37148499e-01 6.32305324e-01
-3.03844333e-01 -1.54017404e-01 5.48554301e-01 9.36865881e-02
5.40312529e-01 6.78003788e-01 -5.76868355e-01 -4.93380308e-01
-8.76870275e-01 5.31043589e-01 1.28253412e+00 1.06301618e+00
2.20817432e-01 -2.70462900e-01 2.83965856e-01 9.92036819e-01
-1.69606060e-02 4.92552519e-01 7.38255084e-01 -1.01092565e+00
6.23039424e-01 4.47520792e-01 1.91688575e-02 -1.32306123e+00
-3.01457644e-01 1.34002138e-02 -2.17305660e-01 -4.69720662e-01
1.40256673e-01 1.68190207e-02 -5.77423930e-01 1.67885363e+00
8.17331895e-02 -3.22059482e-01 2.90401071e-01 4.00660396e-01
1.28713286e+00 1.54294506e-01 2.17016295e-01 -1.66904792e-01
1.23846841e+00 -1.46359950e-01 -1.03021717e+00 -5.11594675e-02
6.55652523e-01 -4.63072985e-01 1.20504093e+00 4.73426163e-01
-1.05718696e+00 1.78240225e-01 -1.16461349e+00 -6.94791317e-01
-1.00027120e+00 -5.07486820e-01 6.67354941e-01 3.47659379e-01
-4.41394985e-01 5.96872032e-01 -5.36185086e-01 -4.38562840e-01
7.03235269e-01 -4.24749941e-01 -1.26002282e-01 -9.21816938e-03
-1.59619761e+00 1.63321483e+00 9.22127306e-01 -6.19733036e-01
-3.11934829e-01 -8.36059868e-01 -8.49437118e-01 -4.44258153e-01
6.04417324e-01 -4.02704328e-01 1.03340125e+00 -4.00852770e-01
-9.96867537e-01 1.37724078e+00 -4.67204861e-02 -6.77131057e-01
9.84475017e-02 -4.97766674e-01 -8.53102803e-01 2.03758717e-01
4.17242825e-01 1.59613654e-01 7.58102238e-02 -1.14690447e+00
-6.36691213e-01 -3.84838909e-01 9.14569944e-02 -1.60678700e-02
-2.31987968e-01 4.77094561e-01 3.78649712e-01 -8.87155592e-01
-2.92842329e-01 -4.89228904e-01 4.87890542e-01 -5.95598340e-01
-5.81604362e-01 -3.54754895e-01 4.45690542e-01 -1.05339026e+00
1.54434109e+00 -1.82221496e+00 1.47958115e-01 4.20181677e-02
2.90699869e-01 -2.19108939e-01 3.34513754e-01 4.67087746e-01
1.03139222e-01 5.48439264e-01 -5.14091730e-01 2.29991913e-01
3.06771845e-01 5.28141081e-01 -5.43122113e-01 8.49097446e-02
1.33501971e-02 1.12742984e+00 -1.31787157e+00 -9.68497455e-01
-8.72743279e-02 1.68221936e-01 -3.46535951e-01 -2.81972438e-01
-3.83502752e-01 -1.52403355e-01 -2.85501331e-01 7.12915540e-01
1.54447518e-02 -6.91560283e-02 9.33668017e-02 -1.95580408e-01
1.82619523e-02 1.21883512e+00 -5.62329769e-01 1.64627874e+00
-4.51889992e-01 8.96205425e-01 -3.75977337e-01 -5.70941448e-01
7.83152044e-01 1.84077006e-02 -1.84876815e-04 -3.34743112e-01
2.64627725e-01 2.86885858e-01 -2.09062640e-02 -4.12641346e-01
7.22796440e-01 -5.04034638e-01 -6.00427568e-01 4.57945704e-01
4.32636410e-01 -6.52260482e-01 7.09156513e-01 6.20438337e-01
1.24812818e+00 7.58932605e-02 9.71707940e-01 -5.11456668e-01
1.88750058e-01 6.08758926e-01 5.02806485e-01 4.54058319e-01
-9.51193720e-02 1.45899076e-02 5.70903182e-01 -2.89258271e-01
-1.09747875e+00 -1.30479133e+00 -6.69061482e-01 8.55012834e-01
-1.02523036e-01 -7.96185017e-01 -4.28810477e-01 -7.32687831e-01
1.42042097e-02 1.74884653e+00 -7.74310589e-01 1.25080526e-01
-4.02480006e-01 -4.79946792e-01 1.16993690e+00 4.51686114e-01
4.40423459e-01 -1.47420442e+00 -8.72187555e-01 7.30225444e-02
-7.98305929e-01 -1.11827636e+00 1.74772993e-01 3.43401551e-01
-3.72037977e-01 -1.28060544e+00 1.25848368e-01 -5.28844833e-01
-1.27211913e-01 -5.68203449e-01 1.77608466e+00 -3.46604198e-01
-3.05685073e-01 1.91618763e-02 -7.50756383e-01 -6.95796847e-01
-6.94019258e-01 -7.55542219e-02 1.38640091e-01 -9.79330361e-01
9.66718435e-01 -9.88072991e-01 3.37932855e-01 -5.06647587e-01
-9.36323762e-01 -1.98069096e-01 1.09022081e-01 6.35228217e-01
3.75866830e-01 1.03402182e-01 7.22324252e-01 -1.18602824e+00
1.08706820e+00 -9.64781404e-01 -4.96236905e-02 3.35096776e-01
-3.53548467e-01 1.80439234e-01 4.54544187e-01 -3.74100022e-02
-1.09031379e+00 -5.60458302e-01 2.45337058e-02 -8.13382771e-03
-1.46654397e-01 8.47694099e-01 -2.62926131e-01 4.95066911e-01
9.27256823e-01 4.31763902e-02 -4.78416443e-01 -3.68939668e-01
1.00310445e+00 9.53662157e-01 9.85306442e-01 -9.37697291e-01
4.35133278e-01 1.96100831e-01 -3.76963079e-01 -8.96959484e-01
-1.35486507e+00 -1.02317125e-01 -7.91184783e-01 1.34205237e-01
7.38495111e-01 -5.25051117e-01 -3.15605581e-01 1.08407550e-01
-1.16799200e+00 -1.94301039e-01 -5.57126999e-01 9.28791910e-02
-6.70887887e-01 1.87953815e-01 -6.40992761e-01 -3.94471347e-01
-4.33213562e-01 -7.32087418e-02 2.89874375e-01 9.32616517e-02
-1.02020836e+00 -1.24201488e+00 2.74769694e-01 4.13438827e-01
1.45748600e-01 6.97819591e-01 1.27038717e+00 -1.27353632e+00
2.11860046e-01 1.19460896e-01 -7.66720027e-02 2.45541424e-01
2.35950202e-01 1.77338824e-01 -6.49924099e-01 6.00227773e-01
-4.25939038e-02 -1.05862439e+00 6.45086944e-01 5.94671145e-02
9.59454834e-01 -6.61378086e-01 -1.84047103e-01 1.69618070e-01
1.38613117e+00 -5.91264591e-02 5.58710575e-01 6.22692704e-01
5.11978090e-01 5.17323971e-01 3.46361518e-01 4.69872326e-01
6.87318444e-01 4.88943197e-02 -1.55564576e-01 9.12673533e-01
-2.54859358e-01 -5.28091550e-01 8.62428695e-02 1.09282625e+00
-2.93594211e-01 1.49078406e-02 -1.24251783e+00 1.00676525e+00
-1.52417374e+00 -1.48759401e+00 1.13680236e-01 1.54341066e+00
1.64409542e+00 4.67070073e-01 1.75791048e-02 3.69173586e-02
5.60384393e-01 4.93826717e-02 -4.64954853e-01 -5.54052353e-01
-6.02060080e-01 6.14679039e-01 3.36449593e-01 4.99992430e-01
-7.24411070e-01 1.26745617e+00 7.06766415e+00 1.10658753e+00
-1.48827463e-01 8.94254223e-02 7.50197768e-02 -3.18313122e-01
-6.93882585e-01 1.34402793e-02 -6.01812124e-01 4.38694626e-01
1.25605333e+00 -9.61715996e-01 4.39066976e-01 8.55037987e-01
-4.24231887e-01 -2.98238844e-01 -1.01589406e+00 8.63432646e-01
4.47189331e-01 -1.45168698e+00 2.76859343e-01 -1.14976548e-01
6.62032604e-01 3.13015997e-01 -4.20786053e-01 3.31033677e-01
1.13290358e+00 -8.79961014e-01 8.16533089e-01 4.31908190e-01
7.48893619e-01 -1.11679900e+00 5.49285889e-01 3.10440123e-01
-5.78077376e-01 3.28485551e-03 -2.10963994e-01 -2.82989830e-01
3.34197611e-01 9.47924197e-01 -3.67400020e-01 5.13785124e-01
5.68198502e-01 8.66843879e-01 -6.62270963e-01 2.91619837e-01
-7.56375372e-01 7.01801658e-01 -3.57382268e-01 -4.62206393e-01
-1.69352084e-01 1.50158659e-01 7.24124849e-01 1.71490884e+00
-2.41378412e-01 6.14438713e-01 -1.56063348e-01 1.24585664e+00
-4.17110652e-01 2.89870519e-02 -7.01720119e-01 -4.87473398e-01
1.04225314e+00 1.07183409e+00 -5.01090467e-01 -7.08236516e-01
-9.99874845e-02 8.10929477e-01 9.44467664e-01 -9.92037803e-02
-8.17192495e-01 -1.07940745e+00 6.39330149e-01 -3.32347691e-01
1.54392377e-01 -2.70025223e-01 -7.18528330e-01 -1.38843715e+00
7.27936551e-02 -7.21116841e-01 5.84553838e-01 -9.78147805e-01
-2.06212640e+00 5.00708699e-01 5.93091249e-01 -3.40037882e-01
-6.29923701e-01 -6.18121207e-01 -2.69114166e-01 6.07751846e-01
-9.87632215e-01 -8.78658175e-01 1.36611134e-01 4.68718946e-01
2.52356261e-01 5.35456464e-02 9.98168111e-01 -2.34168410e-01
-2.22501785e-01 2.96940237e-01 -1.66578796e-02 3.87492269e-01
5.86464584e-01 -1.36948943e+00 5.66015005e-01 5.90953410e-01
3.81491601e-01 9.89010751e-01 8.94349337e-01 -1.19129860e+00
-8.76445949e-01 -5.53779066e-01 1.37556136e+00 -1.18542802e+00
1.53631639e+00 -9.55534652e-02 -8.31820011e-01 9.56016243e-01
4.51467484e-01 -3.73762012e-01 1.21157801e+00 4.50405777e-01
-7.82473385e-01 6.30115390e-01 -1.48438013e+00 5.30699193e-01
1.48961139e+00 -8.03727806e-01 -2.01287508e+00 6.92235008e-02
6.28799736e-01 -2.28276163e-01 -1.10721910e+00 1.74847960e-01
4.14901614e-01 -5.77074051e-01 4.95657533e-01 -9.28543329e-01
1.27547157e+00 1.23104282e-01 -6.39251828e-01 -1.57459974e+00
-2.67411262e-01 -3.86947602e-01 -4.95852351e-01 1.47503793e+00
6.64231658e-01 -3.18318069e-01 4.69092056e-02 8.51117671e-01
1.44958701e-02 -4.37228024e-01 -9.97794330e-01 -7.85558999e-01
6.35940909e-01 -7.61283278e-01 2.88216025e-01 1.46523428e+00
9.61773753e-01 7.57212698e-01 2.76048839e-01 -4.89977121e-01
8.98999453e-01 1.44590706e-01 1.56258598e-01 -1.33354795e+00
-3.64753865e-02 -2.80903518e-01 -3.97638232e-01 -4.46102291e-01
5.91005743e-01 -1.31410944e+00 -2.10044488e-01 -1.67044055e+00
6.46281123e-01 -1.44875169e-01 -1.99011534e-01 6.57952726e-01
-1.13231214e-02 1.95627615e-01 3.32798585e-02 2.07188219e-01
-4.81088132e-01 3.36390108e-01 7.37674534e-01 1.17848597e-01
-1.12063110e-01 -9.47352886e-01 -1.12647712e+00 1.07102132e+00
1.04853499e+00 -5.04884005e-01 -2.32771888e-01 -2.93209136e-01
6.74433827e-01 -5.34554601e-01 5.18068254e-01 -8.12373519e-01
-4.40756567e-02 -5.03109336e-01 4.41802174e-01 -6.73418343e-01
1.60265967e-01 -3.79732519e-01 -2.47190744e-01 -4.38426584e-02
-6.65629208e-01 -1.43436715e-02 3.55054140e-01 1.96831092e-01
9.07714367e-02 -3.65946561e-01 5.11401653e-01 -3.35374862e-01
-8.13797951e-01 -4.74115223e-01 -1.82361767e-01 1.14772570e+00
7.27123857e-01 1.06698461e-01 -7.01294363e-01 -1.48431614e-01
-6.02827668e-01 -1.82636186e-01 5.57740986e-01 4.52166975e-01
4.68172312e-01 -1.29781580e+00 -9.59312737e-01 -5.15724242e-01
3.15186650e-01 -4.83192027e-01 -2.54633784e-01 1.81833610e-01
-1.94131851e-01 4.16025966e-01 -5.49230218e-01 3.83492351e-01
-7.66546786e-01 7.42374301e-01 -3.01093936e-01 -1.33146450e-01
-7.47607052e-01 5.92707396e-01 -7.52512157e-01 -3.18855762e-01
-3.78615648e-01 -3.74283850e-01 -3.23332191e-01 8.15949738e-02
7.40081251e-01 6.07144356e-01 -2.01651663e-01 -7.75075138e-01
-6.45692766e-01 -4.38619480e-02 -7.66751468e-02 -6.29397929e-01
1.60857606e+00 -1.55692071e-01 -4.34800088e-01 1.10612261e+00
9.43557739e-01 5.12429655e-01 -2.95695096e-01 -1.01908199e-01
3.36271435e-01 -1.41689733e-01 -6.44088238e-02 -1.41162968e+00
-2.56372273e-01 1.84273809e-01 -5.50719678e-01 3.29342276e-01
6.70378327e-01 5.11154711e-01 9.03998017e-01 6.19096875e-01
5.26679933e-01 -1.47018564e+00 -4.13812935e-01 9.82589960e-01
9.31036294e-01 -8.63932192e-01 2.11394697e-01 -3.74195039e-01
-7.47670352e-01 9.77015078e-01 2.42080525e-01 -2.16834366e-01
5.82755685e-01 4.86954927e-01 -1.99055299e-01 -6.59636676e-01
-7.04595387e-01 -2.25662768e-01 1.58323631e-01 8.31928670e-01
7.00930536e-01 2.55146295e-01 -7.44808674e-01 1.22831106e+00
-1.12426913e+00 1.07535601e-01 7.23612607e-01 1.10891402e+00
-6.47258401e-01 -7.36761510e-01 -4.50735614e-02 6.87765002e-01
-5.85799396e-01 -7.58512974e-01 -1.12541521e+00 9.76257443e-01
4.97997701e-02 1.08901799e+00 -2.08135769e-01 -3.11179161e-01
8.49521011e-02 4.45284963e-01 8.50359082e-01 -5.73150098e-01
-1.89847201e-01 -8.85654390e-01 8.23918104e-01 -4.16691005e-01
-7.19027877e-01 -7.27729619e-01 -1.50874496e+00 -5.86402059e-01
-1.44133836e-01 1.55616552e-01 5.38667500e-01 1.30408967e+00
1.41450375e-01 2.08129525e-01 -2.65842706e-01 -3.48651350e-01
-3.06812435e-01 -1.15081966e+00 -6.90375745e-01 5.14744699e-01
-1.25327498e-01 -8.83192897e-01 -5.19940972e-01 3.36678237e-01] | [9.90664005279541, 8.15768814086914] |
13d4ed29-40be-4cd2-8e0b-cbd63a957126 | covid-19-ct-cxr-a-freely-accessible-and | 2006.06177 | null | https://arxiv.org/abs/2006.06177v2 | https://arxiv.org/pdf/2006.06177v2.pdf | COVID-19-CT-CXR: a freely accessible and weakly labeled chest X-ray and CT image collection on COVID-19 from biomedical literature | The latest threat to global health is the COVID-19 outbreak. Although there exist large datasets of chest X-rays (CXR) and computed tomography (CT) scans, few COVID-19 image collections are currently available due to patient privacy. At the same time, there is a rapid growth of COVID-19-relevant articles in the biomedical literature. Here, we present COVID-19-CT-CXR, a public database of COVID-19 CXR and CT images, which are automatically extracted from COVID-19-relevant articles from the PubMed Central Open Access (PMC-OA) Subset. We extracted figures, associated captions, and relevant figure descriptions in the article and separated compound figures into subfigures. We also designed a deep-learning model to distinguish them from other figure types and to classify them accordingly. The final database includes 1,327 CT and 263 CXR images (as of May 9, 2020) with their relevant text. To demonstrate the utility of COVID-19-CT-CXR, we conducted four case studies. (1) We show that COVID-19-CT-CXR, when used as additional training data, is able to contribute to improved DL performance for the classification of COVID-19 and non-COVID-19 CT. (2) We collected CT images of influenza and trained a DL baseline to distinguish a diagnosis of COVID-19, influenza, or normal or other types of diseases on CT. (3) We trained an unsupervised one-class classifier from non-COVID-19 CXR and performed anomaly detection to detect COVID-19 CXR. (4) From text-mined captions and figure descriptions, we compared clinical symptoms and clinical findings of COVID-19 vs. those of influenza to demonstrate the disease differences in the scientific publications. We believe that our work is complementary to existing resources and hope that it will contribute to medical image analysis of the COVID-19 pandemic. The dataset, code, and DL models are publicly available at https://github.com/ncbi-nlp/COVID-19-CT-CXR. | ['Sung-Won Lee', 'Yu-Xing Tang', 'Yingying Zhu', 'Yifan Peng', 'Ronald M. Summers', 'Zhiyong Lu'] | 2020-06-11 | null | null | null | null | ['one-class-classifier'] | ['methodology'] | [ 1.62166387e-01 -3.30462515e-01 -1.05082415e-01 -1.62072986e-01
-1.04260767e+00 -6.65730476e-01 2.22722605e-01 6.55578494e-01
-5.65092564e-01 6.79902077e-01 1.75543740e-01 -8.24447334e-01
1.02318013e-02 -7.73848951e-01 -9.37883437e-01 -6.26019478e-01
-3.91660482e-01 9.08916354e-01 3.21903229e-02 3.83990794e-01
-1.37529552e-01 7.05497622e-01 -9.00853336e-01 5.96269548e-01
6.22505665e-01 9.50964153e-01 6.67324364e-01 1.04775357e+00
2.72682309e-01 4.82275367e-01 -5.74023366e-01 -8.27760622e-02
1.51494831e-01 -4.59779203e-01 -7.33849466e-01 -4.47929889e-01
2.91340828e-01 -6.45134747e-01 -2.34022379e-01 3.61613780e-01
6.40207350e-01 -4.13558543e-01 1.09644341e+00 -1.15722692e+00
-4.80990171e-01 -1.32620379e-01 -7.69698679e-01 1.08928275e+00
3.93158883e-01 4.13503766e-01 4.86598283e-01 -8.12996745e-01
1.17269969e+00 7.02398360e-01 9.45633292e-01 5.97761810e-01
-6.61714911e-01 -8.21830451e-01 -4.50673491e-01 5.89787811e-02
-1.31772840e+00 3.48235935e-01 -4.43267403e-03 -9.68259633e-01
1.11870468e+00 3.94807220e-01 9.90710855e-01 1.47216904e+00
7.62105048e-01 3.98090810e-01 9.85148787e-01 -2.98802834e-02
-7.73312449e-02 -3.52374852e-01 -1.70623809e-01 9.09797490e-01
5.14291584e-01 8.84848908e-02 9.75845158e-02 -5.23899436e-01
5.68155050e-01 6.47276461e-01 -4.62208629e-01 1.60941482e-01
-1.28824067e+00 8.51662159e-01 4.76180941e-01 2.49952614e-01
-4.64152008e-01 -3.80805612e-01 6.79248035e-01 5.50321527e-02
3.95878822e-01 3.73493135e-01 -5.02773643e-01 1.68204159e-01
-8.41143131e-01 1.47746041e-01 2.21309841e-01 7.23136306e-01
1.20667927e-01 -2.62252539e-01 -1.05296910e-01 6.32700741e-01
7.70897567e-02 1.22381580e+00 6.07366204e-01 -3.96648943e-01
4.37484264e-01 3.69135678e-01 -2.45753318e-01 -1.00636876e+00
-9.09878910e-01 -1.45767987e-01 -1.05156446e+00 -3.71307790e-01
1.47247523e-01 -4.60436374e-01 -1.18417108e+00 1.39865887e+00
1.87119946e-01 4.92059216e-02 -1.49914190e-01 8.65978122e-01
1.23133886e+00 7.06247926e-01 3.85815613e-02 -2.48872951e-01
1.81542718e+00 -4.82309073e-01 -5.80164552e-01 4.14972186e-01
7.92689025e-01 -6.43623531e-01 8.72386217e-01 2.23835334e-01
-7.46047199e-01 -8.92154425e-02 -8.46851587e-01 2.46909603e-01
-7.85933495e-01 -2.25769501e-04 3.77823971e-02 2.49291673e-01
-7.62085140e-01 7.38997087e-02 -1.06812525e+00 -6.70465827e-01
7.92495430e-01 -3.89771163e-02 -3.58426094e-01 -2.24357948e-01
-1.27437747e+00 1.05935013e+00 1.27623618e-01 -2.80344903e-01
-1.18781221e+00 -1.04942393e+00 -7.28939354e-01 -4.70891654e-01
4.70195800e-01 -1.06859803e+00 1.01622105e+00 2.37109349e-03
-3.41199070e-01 1.32214165e+00 -2.36188188e-01 -3.29021633e-01
5.02737880e-01 1.06192324e-02 -6.66267276e-01 7.93715596e-01
3.91487062e-01 5.88538766e-01 6.36434078e-01 -1.29425597e+00
-6.23092771e-01 -5.47274351e-01 -2.89799571e-01 -1.70086592e-01
2.19708905e-01 3.34744155e-01 -4.61693853e-01 -9.30127263e-01
-4.43345308e-01 -1.02956474e+00 -1.79815724e-01 4.33321334e-02
-4.91114229e-01 6.70915842e-02 9.49917912e-01 -9.27594304e-01
1.04395747e+00 -1.92343640e+00 -6.10795140e-01 -5.84001513e-03
7.43578732e-01 2.84199685e-01 2.04724640e-01 2.19961494e-01
-8.04290920e-02 5.41337550e-01 -5.60941041e-01 1.11499935e-01
-6.09981775e-01 2.50124037e-01 -1.38914630e-01 8.16932082e-01
3.99978459e-01 9.96083081e-01 -8.26772034e-01 -7.22416341e-01
2.75691785e-02 5.54745853e-01 -3.73502851e-01 4.09009635e-01
-1.58059508e-01 7.78369367e-01 -4.95396793e-01 8.65789354e-01
6.96946263e-01 -7.62782812e-01 -2.71519832e-02 -2.13715360e-01
4.85728495e-03 -2.23533958e-02 -2.61471093e-01 1.11779726e+00
-2.59512693e-01 6.63033128e-01 1.51688755e-01 -7.07750201e-01
2.51787722e-01 7.13325620e-01 8.83513689e-01 -5.83922267e-01
2.10331380e-01 1.77677080e-01 6.01599589e-02 -8.27496707e-01
1.52659789e-01 -2.07536429e-01 2.01876871e-02 9.06472445e-01
-2.97220975e-01 -1.99371427e-01 1.19204536e-01 3.21887434e-01
1.04431045e+00 -4.98442858e-01 4.97136772e-01 -6.22048266e-02
1.19437620e-01 5.24827123e-01 2.86810338e-01 9.46937263e-01
-2.49708101e-01 1.04804122e+00 2.85198331e-01 -6.19538724e-01
-1.14015472e+00 -1.41567802e+00 -7.72972226e-01 4.59799975e-01
-3.73366207e-01 -3.91671151e-01 -5.66686869e-01 -7.00687170e-01
-2.14522049e-01 3.45601112e-01 -8.67329657e-01 3.76027584e-01
-8.06986749e-01 -1.01009512e+00 9.80535686e-01 5.72014749e-01
1.13794014e-01 -1.13026392e+00 -1.22755992e+00 -1.66296382e-02
-3.97438586e-01 -1.02596259e+00 -5.81473827e-01 9.82487500e-02
-5.35507143e-01 -1.64014995e+00 -1.25293982e+00 -6.10916495e-01
8.04555655e-01 2.18997747e-02 1.13343024e+00 4.73123163e-01
-1.01006758e+00 6.40030205e-01 -3.38080615e-01 -8.28063846e-01
-5.93926370e-01 -2.22302750e-01 1.83476970e-01 -6.74608946e-01
2.89634258e-01 1.36072859e-01 -7.91596234e-01 1.12896673e-01
-1.06640220e+00 -2.38654241e-02 3.53559643e-01 7.65687406e-01
1.05507541e+00 -1.69078603e-01 2.70125091e-01 -9.58497584e-01
5.53838491e-01 -9.56889689e-01 -3.10439229e-01 1.09471865e-01
-7.10392237e-01 -5.67251980e-01 6.43267930e-01 -3.30876410e-02
-7.56197393e-01 -3.39562982e-01 -3.45143169e-01 -6.17745340e-01
-3.99449021e-01 5.94554543e-01 6.10636532e-01 7.20854640e-01
5.31903863e-01 2.08528504e-01 8.19857642e-02 -5.25313377e-01
-8.50038454e-02 8.99024606e-01 6.96968377e-01 -3.21851999e-01
5.23932993e-01 7.57230878e-01 -3.96997407e-02 -9.29798841e-01
-6.66994512e-01 -6.60845995e-01 -5.09584427e-01 -1.13862365e-01
1.56845593e+00 -9.08189833e-01 -5.11821032e-01 4.71384406e-01
-1.07755566e+00 -1.52541935e-01 1.72504019e-02 7.06111073e-01
-2.00580776e-01 2.08546162e-01 -9.24820900e-01 -1.85957551e-01
-7.45213389e-01 -1.54482698e+00 1.08580399e+00 -6.35300875e-02
-3.56563628e-01 -9.38415825e-01 3.32072854e-01 1.85354471e-01
3.30384433e-01 6.71562135e-01 1.25492799e+00 -1.02708590e+00
-2.69999415e-01 -3.79019859e-03 -3.38634044e-01 -6.52461648e-02
2.97961265e-01 2.51039624e-01 -6.22332692e-01 -3.87747198e-01
1.27548918e-01 -3.62880468e-01 9.27002013e-01 6.96740925e-01
1.41268647e+00 -5.14808178e-01 -6.88339353e-01 8.41999650e-01
1.28519058e+00 8.18999290e-01 1.43813834e-01 1.51449904e-01
7.96818912e-01 2.18770653e-01 4.02465075e-01 4.85278010e-01
1.85440645e-01 4.88745064e-01 2.85644144e-01 -5.14534771e-01
1.31007329e-01 -1.22497424e-01 -2.17760891e-01 7.15908587e-01
-2.65549988e-01 -4.14085209e-01 -1.48354387e+00 5.44381559e-01
-9.12519753e-01 -8.22368443e-01 -3.65407437e-01 1.59335089e+00
6.82475328e-01 -1.38358608e-01 1.90021724e-01 -3.89678955e-01
6.97012246e-01 -2.83998977e-02 -6.67981386e-01 -3.51836264e-01
-1.33017108e-01 4.08726186e-01 4.30058450e-01 1.56783849e-01
-1.08331418e+00 4.49066944e-02 6.70657063e+00 5.16444564e-01
-1.42146063e+00 3.80318642e-01 8.41503084e-01 -1.69130847e-01
-1.01018861e-01 -6.85772598e-01 -4.82105225e-01 5.90438485e-01
8.53588641e-01 -1.23221852e-01 -1.95545122e-01 6.80186212e-01
2.99201965e-01 1.03911320e-02 -8.59264433e-01 7.76083708e-01
1.50497586e-01 -1.88615930e+00 1.53600767e-01 2.44953081e-01
5.52295923e-01 6.24854028e-01 7.66106844e-02 -6.08591214e-02
-7.80180022e-02 -1.22483838e+00 2.53324151e-01 3.10319334e-01
1.63452721e+00 -4.24783826e-01 1.01853216e+00 7.19668949e-03
-1.23175168e+00 4.46633488e-01 -1.39293015e-01 6.25130475e-01
1.94548294e-01 3.24584275e-01 -1.34681070e+00 7.95824230e-01
1.05462492e+00 7.35733867e-01 -5.47624171e-01 8.37179482e-01
9.97955278e-02 8.10333669e-01 -3.06168765e-01 3.09667379e-01
2.17321754e-01 2.00888544e-01 6.14922106e-01 1.73236084e+00
3.60113114e-01 5.06684542e-01 9.51678455e-02 9.14807677e-01
-1.29145607e-01 6.68595657e-02 -9.53547478e-01 -1.27190188e-01
3.52827430e-01 1.00467801e+00 -1.08325636e+00 -6.52666092e-01
-6.83463931e-01 5.86085558e-01 -3.93334746e-01 9.64825079e-02
-1.07577741e+00 -3.99316549e-02 3.09273839e-01 5.18260896e-01
4.01067644e-01 2.27449015e-01 -7.15380311e-02 -9.79450285e-01
-3.94156367e-01 -9.52812672e-01 9.16609466e-01 -1.05350614e+00
-1.59847093e+00 9.84347641e-01 4.26222175e-01 -1.30560839e+00
-3.28170329e-01 -7.98600614e-01 -4.09246624e-01 6.93411291e-01
-1.11380661e+00 -8.54379833e-01 -2.74796069e-01 5.65166116e-01
3.72730732e-01 -1.47683606e-01 1.10180628e+00 2.37364680e-01
-4.90460575e-01 5.06124437e-01 3.19265693e-01 4.89833653e-01
4.83488113e-01 -1.01389933e+00 2.73666859e-01 5.68134427e-01
-4.36263621e-01 7.99022794e-01 1.96886078e-01 -1.00220358e+00
-9.18962896e-01 -1.33110547e+00 5.79821885e-01 -7.25246668e-01
4.06488836e-01 -1.93498507e-01 -8.39054406e-01 8.69630814e-01
2.06080392e-01 1.07702754e-01 9.54557121e-01 -7.64827907e-01
-1.92851812e-01 4.86096501e-01 -1.39904535e+00 2.79738426e-01
6.94327116e-01 -4.62141514e-01 -9.16506112e-01 4.11785990e-01
7.83112526e-01 -6.67177439e-01 -1.08773661e+00 4.90102202e-01
5.68562746e-01 -6.56225324e-01 1.10653353e+00 -7.59966850e-01
7.07386255e-01 -1.25508243e-02 -9.09647867e-02 -1.10635328e+00
1.42658025e-01 -6.11109994e-02 2.13357165e-01 3.13744485e-01
4.76616949e-01 -6.63186610e-01 3.90548289e-01 6.78569218e-03
-2.16071501e-01 -1.19642210e+00 -8.58520389e-01 -5.49372852e-01
2.73959428e-01 -4.27528650e-01 5.56169927e-01 1.14769924e+00
-4.14030880e-01 -1.55480742e-01 7.91932493e-02 2.07698401e-02
2.74836630e-01 2.59921521e-01 4.41163540e-01 -9.34369445e-01
-1.57612905e-01 -2.74398923e-01 2.20492296e-03 -5.76763690e-01
-4.02563959e-01 -1.04639244e+00 -1.61048904e-01 -1.93482757e+00
6.08279824e-01 -5.04379869e-01 -2.21552804e-01 6.09638393e-01
-1.35657370e-01 4.19912010e-01 4.69067432e-02 5.63364387e-01
-1.76908314e-01 -1.31439954e-01 1.57434189e+00 -1.99261263e-01
1.11283086e-01 -1.92809820e-01 -2.28932485e-01 7.01599538e-01
7.36078203e-01 -9.12742555e-01 -1.61003441e-01 -1.14979424e-01
-1.68853447e-01 2.70662397e-01 4.03038383e-01 -7.04694092e-01
-3.02786022e-01 -7.38593489e-02 7.54862249e-01 -1.33555233e+00
4.84565869e-02 -7.30022550e-01 3.05153877e-01 1.05532897e+00
-1.37535734e-02 8.20039630e-01 3.50188941e-01 4.87706184e-01
4.09444347e-02 9.57244169e-03 7.32401371e-01 -4.87911314e-01
-1.96591392e-01 5.21416008e-01 -1.08872724e+00 6.78007126e-01
1.03542757e+00 -5.94598874e-02 -6.80197358e-01 -1.20239712e-01
-5.48087478e-01 4.49812002e-02 5.04360259e-01 1.88389093e-01
8.42126250e-01 -8.66968095e-01 -9.26191270e-01 3.13638657e-01
2.80575782e-01 4.04298991e-01 4.52857137e-01 1.29049659e+00
-1.15807843e+00 7.96280742e-01 -3.88955146e-01 -9.97472763e-01
-1.24814677e+00 9.35327113e-01 2.68850386e-01 -5.28345942e-01
-9.90870237e-01 4.57696855e-01 5.76537311e-01 -3.07840973e-01
1.89816542e-02 -5.80756068e-01 1.28675532e-03 1.76797695e-02
6.59856021e-01 -8.20000563e-03 2.60726839e-01 -8.38612854e-01
-8.48260343e-01 4.81696129e-01 -3.24903399e-01 1.97982639e-01
1.37432766e+00 4.82664667e-02 4.50524427e-02 3.76618117e-01
1.49751258e+00 1.39864489e-01 -6.27534270e-01 2.43125245e-01
-4.85228449e-01 -1.50835395e-01 -5.01768529e-01 -8.85270774e-01
-1.11578000e+00 6.85147822e-01 8.21065605e-01 -8.55310634e-02
1.05922019e+00 4.00814861e-01 9.76828754e-01 -1.15202200e-02
3.88997011e-02 -3.82309854e-01 4.78242412e-02 3.52154613e-01
7.86137640e-01 -1.10783851e+00 2.81184018e-01 -2.61094756e-02
-6.79447591e-01 8.20841312e-01 4.59080279e-01 2.41070449e-01
8.31924379e-01 6.38761282e-01 3.91387880e-01 -9.51323688e-01
-7.32419014e-01 2.25542858e-01 1.28869608e-01 8.64185870e-01
3.58564287e-01 3.10581028e-01 -1.62488371e-01 7.12235034e-01
-1.78062543e-01 -3.92912887e-02 5.19142807e-01 1.12245548e+00
4.29329500e-02 -6.15346849e-01 -6.13505661e-01 1.10639358e+00
-9.81125653e-01 -3.63071233e-01 -2.49608740e-01 1.21103835e+00
4.27121758e-01 6.98964298e-01 3.41536522e-01 -1.99823141e-01
1.02234453e-01 -5.35112992e-02 3.98047507e-01 -5.71482182e-01
-5.67831039e-01 2.88484953e-02 -4.65573259e-02 -3.18120480e-01
-4.32646304e-01 -5.45656085e-01 -1.56871402e+00 -2.56777018e-01
8.44756514e-02 5.34116291e-02 4.20146793e-01 7.82346010e-01
3.08869362e-01 6.20507896e-01 1.82348341e-01 -2.76113540e-01
1.32641166e-01 -7.36301482e-01 -4.34301019e-01 5.97400546e-01
6.11407399e-01 -6.04883134e-01 -4.27784771e-01 3.60899270e-01] | [15.487689018249512, -1.7840631008148193] |
c191a847-f468-4dd2-a01f-5c7863a2a8c1 | monet-unsupervised-scene-decomposition-and | 1901.11390 | null | http://arxiv.org/abs/1901.11390v1 | http://arxiv.org/pdf/1901.11390v1.pdf | MONet: Unsupervised Scene Decomposition and Representation | The ability to decompose scenes in terms of abstract building blocks is
crucial for general intelligence. Where those basic building blocks share
meaningful properties, interactions and other regularities across scenes, such
decompositions can simplify reasoning and facilitate imagination of novel
scenarios. In particular, representing perceptual observations in terms of
entities should improve data efficiency and transfer performance on a wide
range of tasks. Thus we need models capable of discovering useful
decompositions of scenes by identifying units with such regularities and
representing them in a common format. To address this problem, we have
developed the Multi-Object Network (MONet). In this model, a VAE is trained
end-to-end together with a recurrent attention network -- in a purely
unsupervised manner -- to provide attention masks around, and reconstructions
of, regions of images. We show that this model is capable of learning to
decompose and represent challenging 3D scenes into semantically meaningful
components, such as objects and background elements. | ['Alexander Lerchner', 'Irina Higgins', 'Nicholas Watters', 'Christopher P. Burgess', 'Rishabh Kabra', 'Matt Botvinick', 'Loic Matthey'] | 2019-01-22 | null | null | null | null | ['unsupervised-object-segmentation'] | ['computer-vision'] | [ 2.57503808e-01 1.88774541e-01 2.44916752e-01 -4.58057463e-01
-3.49220157e-01 -5.39440393e-01 7.19282448e-01 2.18901873e-01
9.52448249e-02 4.05883789e-01 6.25970304e-01 -2.08658904e-01
-1.91714779e-01 -7.17322052e-01 -9.09388542e-01 -5.12345076e-01
-2.37975288e-02 4.54134673e-01 -1.19483434e-02 2.11835280e-03
-1.14466054e-02 7.94442296e-01 -1.72335637e+00 8.12145352e-01
6.37626410e-01 8.08361888e-01 5.83309174e-01 5.17070651e-01
-9.45306346e-02 1.10402882e+00 -5.30029714e-01 -3.79187375e-01
2.15263799e-01 -3.63913804e-01 -9.67921376e-01 4.67296779e-01
5.30497193e-01 -3.01211745e-01 -2.84004092e-01 8.86655629e-01
-1.16272960e-02 6.19071245e-01 6.55500829e-01 -8.26088607e-01
-8.58494997e-01 4.40885067e-01 -3.90599161e-01 8.09038952e-02
3.62245262e-01 1.89976364e-01 1.17969060e+00 -8.64887416e-01
7.96525776e-01 1.52845192e+00 4.45895374e-01 6.03467524e-01
-1.74474669e+00 -3.86930071e-02 4.48323190e-01 1.59675151e-01
-1.03610778e+00 -5.60642123e-01 9.12413597e-01 -3.46108735e-01
1.47097301e+00 4.89386797e-01 7.33043194e-01 1.12134790e+00
1.28794104e-01 9.32470977e-01 8.01244259e-01 -3.18124205e-01
2.65097290e-01 1.58428356e-01 1.87078044e-01 7.17914820e-01
3.04451168e-01 -3.77695084e-01 -4.46554095e-01 3.56529206e-01
8.80342185e-01 2.35766470e-01 -1.24937944e-01 -7.07483590e-01
-1.41617513e+00 4.93047118e-01 8.85288835e-01 3.55840355e-01
-7.26503134e-01 3.61875832e-01 1.68689832e-01 -4.41253521e-02
4.06540781e-01 1.00232542e+00 -2.42062271e-01 3.95062417e-01
-5.04987001e-01 1.89612940e-01 3.64341706e-01 7.19990849e-01
6.65832460e-01 1.55965596e-01 -1.03274398e-01 7.46009171e-01
-1.32256132e-02 1.87768340e-02 3.54565322e-01 -1.18115079e+00
4.19313431e-01 8.00314069e-01 8.04948732e-02 -1.02600467e+00
-5.49095392e-01 -5.74974418e-01 -9.24607992e-01 2.86509216e-01
9.41439345e-03 1.97803199e-01 -1.14994776e+00 2.07858562e+00
2.98901260e-01 -5.97292557e-02 3.30693483e-01 7.53909290e-01
8.96828830e-01 9.07865822e-01 2.19090760e-01 1.39114320e-01
1.49805391e+00 -9.91510689e-01 -4.67899352e-01 -4.67598200e-01
2.47254491e-01 -3.04695398e-01 1.27809274e+00 3.84417623e-01
-1.49430406e+00 -1.00307345e+00 -9.48601186e-01 -6.61489487e-01
-5.24805009e-01 -5.52232042e-02 8.02434325e-01 1.19326189e-01
-1.06464255e+00 5.22950351e-01 -9.03137863e-01 -6.58106208e-01
7.39816546e-01 8.72598216e-02 -4.49417979e-01 -7.04469457e-02
-5.63462794e-01 1.10658944e+00 5.98277509e-01 1.01727113e-01
-1.24939847e+00 -6.46217287e-01 -1.00633562e+00 4.26059097e-01
2.94886410e-01 -1.35964155e+00 1.01377320e+00 -1.09359992e+00
-1.09120810e+00 8.18216801e-01 -2.43058011e-01 -4.78544533e-01
3.63982394e-02 -1.97784245e-01 -2.35169142e-01 3.07610422e-01
1.39710128e-01 8.17956567e-01 8.67150486e-01 -1.67970312e+00
-3.51961792e-01 -3.36168766e-01 5.03680706e-01 4.35084969e-01
-1.62240595e-01 -9.37622227e-03 -2.90897399e-01 -6.96729064e-01
3.78109306e-01 -5.45561969e-01 -4.78085011e-01 1.66043714e-02
-4.76941168e-01 6.41696975e-02 6.10282302e-01 -7.55399227e-01
5.80369473e-01 -2.26702070e+00 6.29749000e-01 7.34653994e-02
6.22201383e-01 -5.55360615e-02 -2.78733015e-01 2.04508230e-01
-4.06477243e-01 1.05463676e-01 -4.37775962e-02 -5.94669223e-01
8.89487565e-02 2.76261389e-01 -5.36391795e-01 6.93423152e-02
6.11996353e-01 1.22515297e+00 -8.81192207e-01 1.86164712e-03
4.35920864e-01 4.48120147e-01 -6.38848841e-01 3.47830713e-01
-6.60186708e-01 4.58134443e-01 -3.30946207e-01 4.71170634e-01
4.14339006e-01 -5.67145169e-01 1.82469249e-01 -3.69022548e-01
2.33222276e-01 5.64114869e-01 -1.04634762e+00 2.00645161e+00
-6.70641661e-01 7.44491279e-01 -1.06131911e-01 -1.28513312e+00
6.80949509e-01 9.29793194e-02 2.48426422e-01 -5.75237811e-01
7.33227208e-02 -2.76828140e-01 -4.15242091e-02 -6.77164316e-01
4.75577950e-01 -4.09501433e-01 -8.24235454e-02 5.32093465e-01
2.38272503e-01 -2.98574805e-01 3.00436951e-02 2.25262463e-01
1.06305432e+00 6.45924732e-02 3.50349218e-01 -1.58147380e-01
1.61877215e-01 -4.38402221e-02 3.90771449e-01 7.80894518e-01
2.05484599e-01 6.42433882e-01 4.22465473e-01 -8.16779912e-01
-1.03752041e+00 -1.58968747e+00 9.80092958e-02 9.36405838e-01
2.26169482e-01 -4.20385659e-01 -4.16907102e-01 -3.42309177e-01
-7.85388425e-02 9.83511031e-01 -7.71168828e-01 -2.72271693e-01
-2.98950881e-01 -4.13762480e-01 1.32656535e-02 7.61617959e-01
5.08777618e-01 -1.29576850e+00 -1.07144821e+00 1.86180741e-01
-2.18178481e-01 -1.18046069e+00 3.10932428e-01 4.97912586e-01
-7.67247438e-01 -7.80030966e-01 -3.48109454e-01 -7.73125768e-01
1.03100502e+00 7.90853262e-01 1.49568951e+00 -4.08301726e-02
-4.43812549e-01 4.37938005e-01 -3.03910732e-01 -3.36592585e-01
-3.03185254e-01 -4.16651607e-01 -7.58934692e-02 1.55836135e-01
5.06339930e-02 -8.51939559e-01 -2.16791287e-01 -1.09254546e-01
-1.16808999e+00 8.12601805e-01 5.87460756e-01 6.45987451e-01
5.67513645e-01 2.77503133e-01 2.06217259e-01 -7.23404408e-01
5.31287670e-01 -4.16379005e-01 -2.94345826e-01 5.03795505e-01
2.55285144e-01 2.63004273e-01 8.53378236e-01 -3.06202680e-01
-1.29169154e+00 -9.20016095e-02 2.18370482e-01 -6.78085744e-01
-4.46344495e-01 6.24327958e-01 -5.47420502e-01 3.31905454e-01
7.00028896e-01 1.23045877e-01 -3.83951724e-01 -5.62665343e-01
7.60396659e-01 1.83099598e-01 7.90852666e-01 -9.16327178e-01
6.11885488e-01 5.67889571e-01 8.59553739e-02 -8.24218452e-01
-1.07797742e+00 -2.26321340e-01 -7.98850477e-01 2.28326488e-02
1.09715796e+00 -1.07619739e+00 -4.56053555e-01 1.27703264e-01
-1.34522283e+00 -2.91490018e-01 -7.66997993e-01 2.52224416e-01
-7.12265253e-01 8.38881657e-02 -3.66784394e-01 -6.45288408e-01
2.20897809e-01 -1.17856729e+00 8.87795150e-01 2.10107356e-01
-4.54122335e-01 -9.34320748e-01 -2.06537843e-01 4.43359524e-01
1.13179229e-01 5.76109946e-01 1.23265362e+00 -3.47809672e-01
-9.89885926e-01 3.03611681e-02 -4.30564433e-01 2.21052974e-01
1.69364065e-01 -2.13464782e-01 -1.23687685e+00 -1.63909674e-01
2.93352425e-01 -5.57888329e-01 1.15056002e+00 3.26444775e-01
1.61597586e+00 -4.25049663e-01 -1.66541472e-01 7.94887424e-01
1.37773120e+00 1.38639465e-01 7.30724216e-01 3.45623940e-02
6.90557837e-01 8.30910981e-01 -1.79136708e-01 3.28201920e-01
4.17284191e-01 5.01273036e-01 4.73886609e-01 -1.90125316e-01
-3.81285340e-01 -3.38386983e-01 2.51446366e-01 3.89937550e-01
-2.17417136e-01 -2.87765384e-01 -8.29798520e-01 6.87358737e-01
-1.75542378e+00 -1.24735129e+00 2.72828192e-01 1.83639705e+00
5.34363449e-01 2.45365977e-01 -2.90046595e-02 -2.75190264e-01
3.76757324e-01 3.18670154e-01 -7.28683054e-01 -4.59617615e-01
-3.08315545e-01 2.24276543e-01 -2.20406339e-01 3.51701617e-01
-1.02768803e+00 9.88479376e-01 6.20151329e+00 3.42986435e-01
-9.26655471e-01 -1.30084112e-01 7.59589314e-01 -1.58877671e-01
-5.71229041e-01 3.78415585e-02 -3.66412699e-01 -1.18034519e-01
6.55400038e-01 2.92515624e-02 7.37822235e-01 6.52249515e-01
-6.81571290e-02 2.55810912e-03 -1.42141759e+00 9.41596866e-01
3.25446337e-01 -1.73748231e+00 6.66543186e-01 -2.30770767e-01
7.38633156e-01 -2.31751546e-01 2.48466060e-01 1.23783514e-01
6.95458233e-01 -1.33014727e+00 9.51684177e-01 6.41733646e-01
5.97408891e-01 -3.83120865e-01 2.21728504e-01 3.36554319e-01
-1.10849833e+00 -1.58941701e-01 -5.21393299e-01 -2.67987490e-01
2.17840187e-02 3.84032816e-01 -6.40393734e-01 6.75747275e-01
6.44404054e-01 9.66420412e-01 -8.12403679e-01 9.69710410e-01
-3.57105583e-01 1.01336256e-01 -2.17706889e-01 1.62155852e-01
2.26770177e-01 -1.73569024e-02 4.69582111e-01 1.02471364e+00
2.56776869e-01 2.24281430e-01 -4.69567962e-02 1.45784295e+00
-1.77626625e-01 -4.59427595e-01 -7.73746729e-01 -1.17580317e-01
-3.41990478e-02 1.31616664e+00 -9.58623409e-01 -5.13219297e-01
-4.29087520e-01 9.01542127e-01 6.59771323e-01 6.75553739e-01
-7.54094303e-01 -2.13557422e-01 8.60479236e-01 -1.20038465e-01
3.78390253e-01 -3.86152565e-01 -6.04813337e-01 -1.62009275e+00
1.40185997e-01 -9.70022380e-01 2.76904106e-01 -1.38839078e+00
-1.26869726e+00 6.34168327e-01 1.04032934e-01 -1.02937841e+00
2.63047051e-02 -7.38687217e-01 -7.76133597e-01 7.07262933e-01
-1.07110572e+00 -1.53018904e+00 -4.56878275e-01 6.64107978e-01
7.34778047e-01 -1.00426441e-02 9.05079484e-01 -2.69395828e-01
-5.67409039e-01 7.70679414e-02 -4.99364361e-02 -4.40361165e-02
1.12598263e-01 -1.25076330e+00 6.94940865e-01 1.11729062e+00
8.01750839e-01 8.30663383e-01 8.03371668e-01 -3.84564191e-01
-1.20879936e+00 -1.11050701e+00 7.06386685e-01 -8.15971315e-01
4.14560735e-01 -8.66002321e-01 -1.08989942e+00 1.06246674e+00
2.14132756e-01 -1.62267029e-01 6.54053152e-01 4.44227993e-01
-7.58044481e-01 -7.29302093e-02 -9.54909563e-01 9.77664649e-01
1.23713803e+00 -7.57528424e-01 -1.09355927e+00 3.96901906e-01
8.75676751e-01 -1.63546637e-01 -5.21246016e-01 2.08222955e-01
3.86439294e-01 -1.07789767e+00 1.55385470e+00 -1.23965752e+00
7.44700789e-01 -2.39163503e-01 -4.14353043e-01 -1.43141794e+00
-6.79422081e-01 -5.62808037e-01 -2.33365610e-01 7.99814343e-01
2.73176551e-01 -3.29390675e-01 5.76188624e-01 9.35732722e-01
-3.41828346e-01 -4.86061603e-01 -6.63543284e-01 -6.57503307e-01
-9.04938951e-02 -6.08421862e-01 7.17361629e-01 8.40875089e-01
-2.19791732e-03 7.15173066e-01 -1.19748719e-01 2.56288826e-01
5.47458351e-01 4.98646468e-01 8.39102626e-01 -1.15250194e+00
-4.08323437e-01 -5.56478202e-01 -4.28499877e-01 -1.16178322e+00
3.36055338e-01 -8.91979933e-01 -1.02945395e-01 -1.93021226e+00
4.34307277e-01 -2.27221966e-01 -4.31454808e-01 5.35457253e-01
-7.55262449e-02 -2.91572511e-02 4.75899816e-01 1.73118904e-01
-9.06382680e-01 5.75160980e-01 1.28455865e+00 -3.14232141e-01
-1.74888909e-01 -4.36439961e-01 -9.91307259e-01 9.39456820e-01
6.37385726e-01 -9.83813852e-02 -5.43497682e-01 -8.98335457e-01
3.35754007e-01 -2.16369610e-02 8.28565001e-01 -1.19403863e+00
1.19255355e-03 -3.77439976e-01 8.66468489e-01 -5.00687420e-01
7.76898623e-01 -9.93471503e-01 3.84220809e-01 1.35899648e-01
-5.58856130e-01 -6.83356076e-02 3.85766953e-01 5.31860173e-01
-1.95551604e-01 -4.59469557e-02 6.39605463e-01 -5.42096436e-01
-9.26278770e-01 -4.67489520e-03 -2.78353840e-01 -8.86290371e-02
1.02513385e+00 -4.04180169e-01 -3.52787405e-01 -2.87504852e-01
-9.89758790e-01 2.22179685e-02 5.72175324e-01 4.02747899e-01
8.41209710e-01 -1.27076089e+00 -5.01585722e-01 5.12854815e-01
2.52556026e-01 3.11536342e-01 4.00950283e-01 2.17670009e-01
-5.73752582e-01 4.13666397e-01 -6.69539630e-01 -4.04133260e-01
-9.07192051e-01 8.53050113e-01 2.74874389e-01 -2.93821394e-01
-7.45712817e-01 1.12557149e+00 9.74837065e-01 -2.88021356e-01
1.34783581e-01 -7.96998560e-01 -1.01021953e-01 -1.12941064e-01
5.23053169e-01 -9.10946429e-02 -2.32919738e-01 -5.33005416e-01
-1.74474388e-01 3.88629854e-01 -1.98214762e-02 5.07123768e-02
1.74807465e+00 -1.71498999e-01 -3.09666067e-01 6.16731226e-01
1.04598892e+00 -1.99756280e-01 -1.38105464e+00 -2.81529784e-01
-1.03226818e-01 -5.88161826e-01 -1.62893444e-01 -8.50924730e-01
-6.80888236e-01 1.10696161e+00 -8.70626606e-03 3.23370546e-01
1.28139234e+00 3.65166157e-01 5.10632098e-01 7.51157284e-01
1.84894964e-01 -5.91847777e-01 5.38392007e-01 3.82826954e-01
1.24316442e+00 -1.07606578e+00 4.95593697e-02 -1.09569281e-01
-6.51350021e-01 1.12407517e+00 6.66212499e-01 -1.00631997e-01
4.31608438e-01 -9.18613970e-02 -3.42823803e-01 -4.44317400e-01
-1.07105958e+00 -3.13882917e-01 6.49586380e-01 7.87950695e-01
3.95916104e-02 6.16791211e-02 7.66224623e-01 6.76466942e-01
-1.29148215e-01 -4.52907592e-01 4.47514117e-01 6.25662029e-01
-4.77701783e-01 -5.35716534e-01 -2.09488511e-01 4.58348781e-01
-3.33374576e-03 -1.22447386e-02 -4.24185872e-01 4.94860858e-01
3.46335053e-01 6.85365438e-01 2.73232639e-01 -3.49173009e-01
2.50602633e-01 1.18219927e-01 5.53048790e-01 -9.72465456e-01
-4.66777891e-01 -1.65856630e-01 1.29640430e-01 -5.67700386e-01
-5.65972447e-01 -6.67530596e-01 -1.00315976e+00 -7.10605383e-02
1.75359040e-01 -2.56742656e-01 3.84467065e-01 8.80551934e-01
4.58590031e-01 9.10020292e-01 3.43896925e-01 -9.15882707e-01
-1.72193557e-01 -5.94020963e-01 -4.79128301e-01 8.79433036e-01
4.68470156e-01 -6.20125115e-01 -8.99548531e-02 3.96754414e-01] | [10.054214477539062, 1.0241361856460571] |
147a3d16-dc61-4131-aae6-c5258c7894f4 | cooperative-audio-source-separation-and | 1912.05038 | null | http://arxiv.org/abs/1912.05038v1 | http://arxiv.org/pdf/1912.05038v1.pdf | Cooperative Audio Source Separation and Enhancement Using Distributed Microphone Arrays and Wearable Devices | Augmented listening devices such as hearing aids often perform poorly in
noisy and reverberant environments with many competing sound sources. Large
distributed microphone arrays can improve performance, but data from remote
microphones often cannot be used for delay-constrained real-time processing. We
present a cooperative audio source separation and enhancement system that
leverages wearable listening devices and other microphone arrays spread around
a room. The full distributed array is used to separate sound sources and
estimate their statistics. Each listening device uses these statistics to
design real-time binaural audio enhancement filters using its own local
microphones. The system is demonstrated experimentally using 10 speech sources
and 160 microphones in a large, reverberant room. | [] | 2019-12-10 | null | null | null | null | ['audio-source-separation'] | ['audio'] | [ 1.86698914e-01 -6.67592347e-01 7.71737635e-01 -2.07644445e-03
-1.52699053e+00 -6.68465555e-01 -3.24457496e-01 1.42094493e-01
-2.78151840e-01 2.98004448e-01 8.26687813e-01 -3.64902824e-01
-5.92741743e-02 -2.58747995e-01 -8.50486532e-02 -6.00001216e-01
-5.24141788e-01 -3.73327911e-01 3.31828564e-01 -3.47244088e-03
-3.15569401e-01 1.04887793e-02 -1.74743521e+00 2.35264152e-01
5.27947903e-01 1.15805411e+00 3.57149690e-01 1.73988056e+00
3.55842322e-01 4.89815801e-01 -1.43038273e+00 4.37986583e-01
3.88401508e-01 -3.91192794e-01 6.92923605e-01 -3.97257298e-01
6.31438792e-01 -5.87207735e-01 -6.75957724e-02 7.95368671e-01
1.84061384e+00 5.84713459e-01 2.18980342e-01 -7.93031752e-01
1.36869341e-01 2.38961592e-01 -3.02749306e-01 5.87679029e-01
8.14666867e-01 -1.29676476e-01 6.88393414e-01 -1.11308622e+00
-5.10036290e-01 1.07140577e+00 1.00980699e+00 2.80233562e-01
-8.45766902e-01 -1.03590548e+00 -1.32856861e-01 -2.17508376e-01
-1.39224660e+00 -1.28658462e+00 8.54093969e-01 1.83133800e-02
1.12224841e+00 7.10510015e-01 6.00206196e-01 7.15038002e-01
-1.03248976e-01 3.23200881e-01 1.02185047e+00 -4.53798056e-01
6.94420457e-01 -1.81609035e-01 -3.46109481e-03 9.66836289e-02
-1.24177322e-01 2.94192433e-01 -1.14168847e+00 -6.48490310e-01
4.71408248e-01 -2.66343117e-01 -6.12274170e-01 5.40132940e-01
-1.08377874e+00 8.50608945e-02 1.36502134e-02 2.48812914e-01
-3.15596342e-01 7.30657652e-02 1.17433205e-01 3.93845707e-01
7.57514119e-01 2.40708366e-01 -3.50242853e-01 -3.11916202e-01
-8.08029532e-01 -1.28167719e-01 1.06191266e+00 6.01378381e-01
-4.79966495e-03 4.75907147e-01 -8.77032727e-02 1.45158410e+00
7.35998511e-01 9.24444437e-01 1.86885953e-01 -1.12809300e+00
4.46943671e-01 -5.93742430e-01 2.82506078e-01 -9.26393807e-01
-3.01149905e-01 -7.28526175e-01 -5.49872518e-01 6.77054644e-01
1.81468606e-01 -8.64850163e-01 -4.77972507e-01 1.44789350e+00
6.36390865e-01 7.54881382e-01 -3.29734944e-02 1.05467618e+00
4.66337442e-01 6.39325261e-01 -3.44813347e-01 -4.28312778e-01
1.20715666e+00 -8.61638486e-01 -1.04490232e+00 -4.54976469e-01
-6.52326941e-01 -1.12338519e+00 9.04452145e-01 9.97093737e-01
-1.50884151e+00 -7.07556903e-01 -1.14781189e+00 2.80179352e-01
1.81177750e-01 -3.80939275e-01 1.28941894e-01 1.49246705e+00
-1.52560771e+00 -3.82427394e-01 -9.39746380e-01 3.64289463e-01
-5.34606874e-02 2.91784942e-01 2.85961270e-01 -1.70294440e-03
-6.54060960e-01 8.85448977e-03 -8.80850017e-01 2.84936726e-01
-9.89178419e-01 -9.63081479e-01 -6.03267491e-01 1.33048490e-01
-4.85757850e-02 -7.61691511e-01 1.81993437e+00 -4.90798503e-01
-1.61485052e+00 1.28499582e-01 -5.97681940e-01 -6.12191744e-02
1.07716642e-01 -4.80345249e-01 -1.13786221e+00 -3.60334776e-02
1.38099743e-02 -3.42816502e-01 9.61052954e-01 -1.33922446e+00
-6.42976880e-01 -2.73191452e-01 -5.78681827e-01 5.38270712e-01
-6.93494499e-01 5.09310424e-01 -2.02092435e-02 -8.43443215e-01
4.00375307e-01 -1.58145607e-01 -4.34061587e-01 3.22145447e-02
-2.18714803e-01 4.74935889e-01 3.82933617e-01 -9.70603704e-01
1.34029639e+00 -2.44017148e+00 -6.39128149e-01 3.83734614e-01
3.25363994e-01 1.76796809e-01 4.22550179e-02 7.60542527e-02
2.01389760e-01 -4.25815910e-01 3.42179060e-01 -8.47874343e-01
-9.00701806e-02 -3.95836860e-01 -2.81702906e-01 4.26771313e-01
-7.06806898e-01 -1.41341493e-01 -8.63974214e-01 -2.34110460e-01
8.40843618e-02 8.39494765e-01 -8.58567476e-01 5.19008815e-01
5.59549093e-01 5.68419218e-01 -1.38895750e-01 8.12489927e-01
7.49425232e-01 5.48756599e-01 -1.24389291e-01 1.75900191e-01
-1.21116139e-01 6.17638052e-01 -1.76158834e+00 1.48677647e+00
-1.03812134e+00 7.35802472e-01 1.48574376e+00 -1.25743538e-01
1.01282024e+00 7.95526505e-01 1.17992252e-01 -4.65482533e-01
1.70727428e-02 6.59941614e-01 -6.57668263e-02 -3.27035099e-01
5.39092481e-01 -1.94920018e-01 1.85303152e-01 5.27688086e-01
-2.09444597e-01 -1.95810735e-01 -4.75736439e-01 -1.53428793e-01
1.73711383e+00 -9.38831687e-01 1.83245122e-01 -4.42476720e-02
2.72991687e-01 -9.31837797e-01 3.69996011e-01 8.93875957e-01
-5.11200190e-01 5.48046649e-01 -8.31726313e-01 5.51445901e-01
-3.45449090e-01 -1.87990201e+00 2.70511061e-01 1.86939383e+00
-1.60588935e-01 -4.87885863e-01 -4.06495869e-01 1.86309725e-01
-2.57243574e-01 4.52317625e-01 3.42919886e-01 9.30211693e-02
-3.27458262e-01 -5.87086380e-03 7.23866761e-01 5.28841019e-01
2.14473248e-01 -4.13778722e-01 -3.45340788e-01 6.30633652e-01
-2.88553864e-01 -5.78301251e-01 -9.08830643e-01 3.76080394e-01
-4.66055721e-01 -3.73694211e-01 -8.49552631e-01 -8.09851825e-01
2.99915165e-01 9.44512308e-01 9.82725382e-01 -3.06963831e-01
-3.09151709e-01 1.24286926e+00 1.89242363e-02 -1.13523710e+00
-2.66434520e-01 -7.16724336e-01 6.33813083e-01 1.68347396e-02
-6.13193773e-02 -1.25917447e+00 -9.97891605e-01 6.53824985e-01
-1.64276063e-01 -8.35297287e-01 4.02673241e-03 3.16248178e-01
2.44076639e-01 7.05956399e-01 8.72571766e-01 3.97400558e-01
1.22575974e+00 -5.14787078e-01 -3.41344416e-01 -2.63746679e-01
-3.43871280e-03 -1.09203219e+00 4.52263653e-01 -8.47082913e-01
-1.65588927e+00 -1.92884341e-01 -5.08282900e-01 7.07190856e-02
-1.84991613e-01 3.55188772e-02 -5.77775061e-01 6.04400411e-02
1.04635048e+00 -1.97590426e-01 -2.44959563e-01 -7.59715140e-01
-3.53764445e-02 1.48517764e+00 8.51447165e-01 -2.30630085e-01
4.19500887e-01 2.34338179e-01 -6.44573569e-01 -1.08420396e+00
-1.49675593e-01 -9.85748053e-01 2.61810809e-01 -4.57955360e-01
3.28453660e-01 -1.38506413e+00 -7.40106523e-01 6.38486207e-01
-1.05594647e+00 -6.45767212e-01 -2.12434381e-01 9.96865213e-01
-1.79856330e-01 -2.83644021e-01 -6.57834947e-01 -1.81048393e+00
-3.74564230e-01 -6.77016854e-01 9.77909923e-01 1.87466770e-01
-4.11838263e-01 -6.74457848e-01 3.54576916e-01 4.89505589e-01
9.56681609e-01 -4.10860240e-01 -9.66887921e-02 -3.01463246e-01
-2.04991415e-01 -3.21207434e-01 6.29042864e-01 5.29056847e-01
6.14712000e-01 -4.74402040e-01 -1.75784969e+00 -2.95080781e-01
6.10530853e-01 8.47312659e-02 3.47316563e-01 1.14171851e+00
5.79933941e-01 -2.05237493e-01 -2.57645398e-01 1.98177397e-01
8.41577590e-01 6.54910088e-01 3.54819357e-01 -5.63691020e-01
1.53720453e-01 2.88800895e-01 3.27973992e-01 5.90452135e-01
-1.45518314e-02 3.14070672e-01 1.84001867e-02 -2.79487491e-01
-5.06769419e-01 2.28747819e-02 6.48885965e-01 1.35151649e+00
2.19796807e-01 -6.44733012e-01 -7.86349297e-01 6.27915859e-01
-9.05475259e-01 -7.46649802e-01 3.52062355e-03 2.36250114e+00
8.82261157e-01 -5.99237308e-02 1.08466543e-01 5.78296661e-01
7.97275841e-01 1.15299352e-01 -4.84409600e-01 -2.34002978e-01
-3.24205942e-02 6.34434342e-01 2.46963114e-01 9.74967599e-01
-6.40199661e-01 -9.04327184e-02 7.43152666e+00 6.22887075e-01
-8.24263632e-01 6.38155282e-01 3.76022667e-01 -7.16141164e-01
-1.81569189e-01 -8.92907083e-01 -4.16387737e-01 1.26824796e-01
1.12837124e+00 -1.95825592e-01 5.53740680e-01 7.62345731e-01
6.24883115e-01 -5.06880105e-01 -1.00782073e+00 1.22367632e+00
1.28778011e-01 -6.90516174e-01 -1.09513628e+00 5.94636686e-02
3.93014759e-01 1.70726046e-01 4.82805103e-01 -6.79171905e-02
5.32825470e-01 -8.14948022e-01 6.84275746e-01 2.81619489e-01
6.40846252e-01 -4.61099267e-01 5.07008731e-02 3.83756995e-01
-1.52215958e+00 -4.30224180e-01 2.19436139e-02 -5.00553310e-01
4.32120562e-01 9.73071992e-01 -8.90138268e-01 -1.24344118e-01
1.09538507e+00 -1.87731072e-01 -1.28121063e-01 1.63700485e+00
-6.59294501e-02 1.20549858e+00 -7.60084152e-01 -1.32426351e-01
-7.46548831e-01 3.37946296e-01 1.09148216e+00 1.12781048e+00
7.96774983e-01 5.02751112e-01 1.21928729e-01 1.37098312e-01
7.69689530e-02 -4.06988785e-02 -5.47050595e-01 7.40186810e-01
9.67604816e-01 9.99784410e-01 -4.41809237e-01 -8.45718011e-02
-1.46587804e-01 6.17251158e-01 -8.33351851e-01 6.74665868e-01
-5.55166841e-01 -6.65100992e-01 1.20081711e+00 8.60114023e-03
7.76216537e-02 -7.44272530e-01 -2.15026110e-01 -5.16339600e-01
5.22804819e-02 -1.03765488e+00 1.16340861e-01 -1.17002797e+00
-1.37311518e+00 5.59286594e-01 -5.46305120e-01 -1.34858572e+00
-9.39052701e-02 -3.16821247e-01 -8.97336066e-01 1.22847915e+00
-8.54210138e-01 -2.53386736e-01 -1.96585149e-01 1.19771159e+00
5.39828598e-01 -1.43542305e-01 1.04230225e+00 8.90370667e-01
-4.38139066e-02 8.00797641e-01 4.15625662e-01 -8.65081325e-03
9.96969223e-01 -1.10795772e+00 6.51615411e-02 8.52791786e-01
-9.56893712e-02 7.48400509e-01 9.80772793e-01 -3.20950389e-01
-1.26250601e+00 -6.67354941e-01 5.42006552e-01 -4.33132708e-01
4.55479860e-01 -9.35468376e-01 -5.88141680e-01 1.76121250e-01
3.13987851e-01 2.25924641e-01 1.61564064e+00 2.94900864e-01
-2.46069148e-01 -6.11085415e-01 -1.07530081e+00 5.08372128e-01
1.04596591e+00 -6.88292146e-01 -3.37957531e-01 2.67765671e-01
8.57307255e-01 -3.74995351e-01 -4.97022539e-01 -4.05186862e-01
6.38543785e-01 -9.58917618e-01 1.30533409e+00 3.78311425e-01
-4.45027173e-01 -3.48097682e-01 -6.08112931e-01 -1.57967603e+00
-9.54574347e-02 -1.42130351e+00 -3.03413957e-01 1.64108300e+00
3.05843115e-01 -1.10701394e+00 3.78662586e-01 5.29923141e-01
-4.41123456e-01 7.91158974e-02 -1.24454153e+00 -9.47404623e-01
-6.15762353e-01 -1.28086722e+00 4.81248617e-01 2.31880471e-01
5.47513813e-02 4.43493485e-01 -1.04516484e-01 6.66642845e-01
8.10446143e-01 -6.70322120e-01 6.73481584e-01 -8.92202318e-01
-8.04971814e-01 -1.74980834e-01 -1.82285830e-01 -1.55136395e+00
-3.69210660e-01 -6.23355620e-02 7.19068110e-01 -1.36806834e+00
-9.53452051e-01 -7.23844826e-01 -5.25281608e-01 -4.07497697e-02
-2.68247515e-01 3.47678870e-01 -9.47520062e-02 -3.30084711e-01
-4.63974565e-01 3.66500586e-01 6.43912375e-01 -2.15553418e-01
-7.10427284e-01 7.57688344e-01 -8.84773374e-01 9.72209334e-01
5.73214352e-01 -2.85846502e-01 -6.58132613e-01 -6.90039814e-01
2.05406740e-01 3.16355944e-01 3.01315665e-01 -1.58258164e+00
8.86384726e-01 2.77863503e-01 3.13400954e-01 -4.03923094e-01
8.12044561e-01 -1.06433463e+00 1.12113982e-01 1.01466253e-01
-2.51953244e-01 -2.32799888e-01 3.76832038e-01 8.31024647e-01
-2.19153553e-01 5.63525677e-01 2.74870604e-01 1.75910980e-01
7.42049739e-02 -1.89452544e-01 -9.70950484e-01 -1.58952042e-01
4.21713501e-01 -1.44966602e-01 6.83632568e-02 -1.12125254e+00
-7.95049250e-01 -8.46787766e-02 -2.84220427e-01 1.34028509e-01
6.99091196e-01 -1.21428251e+00 -6.37951493e-01 5.33627391e-01
-5.51873922e-01 4.72691767e-02 5.34295499e-01 5.94233274e-01
1.87381789e-01 -1.11239336e-01 2.54416347e-01 -7.37722754e-01
-1.87616980e+00 1.32546261e-01 6.84199810e-01 3.96062762e-01
-2.65854269e-01 1.53382874e+00 1.02806212e-02 -1.21248484e-01
7.10806966e-01 -3.72798860e-01 7.46100768e-02 -1.20444618e-01
1.29258215e+00 1.00991559e+00 4.12790656e-01 -1.15965694e-01
-3.67349416e-01 3.70958686e-01 7.93711066e-01 -9.87544417e-01
9.54683840e-01 -9.37331736e-01 1.72802225e-01 6.13322854e-01
1.18758023e+00 1.29353642e+00 -9.98833716e-01 -3.31433654e-01
-6.83380008e-01 -8.47367644e-01 3.65272015e-01 -1.01302779e+00
-7.48973131e-01 7.99863398e-01 1.09000087e+00 4.21638995e-01
1.94173288e+00 -1.67737424e-01 8.91028345e-01 2.27457628e-01
6.69106424e-01 -6.76609099e-01 3.96167368e-01 2.65688509e-01
7.19196141e-01 -6.43265903e-01 -4.69585299e-01 -2.37629429e-01
-1.28068462e-01 6.81499779e-01 1.87394068e-01 1.30166426e-01
1.17389083e+00 1.27129269e+00 7.16685951e-01 1.71498805e-01
-4.88851458e-01 2.46028509e-02 1.51910469e-01 1.19586027e+00
4.81258601e-01 1.03507424e-02 4.49506938e-01 1.12535393e+00
-6.33668423e-01 -3.36013496e-01 7.82880187e-02 9.99102056e-01
-6.96628869e-01 -7.09590614e-01 -1.40218806e+00 2.92574078e-01
-7.06005275e-01 -4.20691937e-01 -9.71866399e-02 -5.06580889e-01
8.27865973e-02 2.12234354e+00 2.52155900e-01 -4.66869414e-01
6.53781354e-01 -1.73364542e-02 1.34648636e-01 -5.22687316e-01
-1.03652465e+00 1.07156456e+00 2.57124662e-01 -4.23228800e-01
-2.52853215e-01 -6.28375947e-01 -1.03577077e+00 -1.82526574e-01
-3.86073470e-01 2.98050761e-01 9.71173823e-01 2.34577402e-01
4.43695664e-01 1.34448528e+00 9.14138615e-01 -1.12975836e+00
-3.94217104e-01 -7.79078364e-01 -9.70615327e-01 -5.14021754e-01
1.01937294e+00 -3.78082879e-02 -6.60164416e-01 1.37537330e-01] | [15.045682907104492, 5.8096842765808105] |
b95d889c-b01b-40b4-b22e-ce60b4d979f0 | hsr-diff-hyperspectral-image-super-resolution | 2306.12085 | null | https://arxiv.org/abs/2306.12085v1 | https://arxiv.org/pdf/2306.12085v1.pdf | HSR-Diff:Hyperspectral Image Super-Resolution via Conditional Diffusion Models | Despite the proven significance of hyperspectral images (HSIs) in performing various computer vision tasks, its potential is adversely affected by the low-resolution (LR) property in the spatial domain, resulting from multiple physical factors. Inspired by recent advancements in deep generative models, we propose an HSI Super-resolution (SR) approach with Conditional Diffusion Models (HSR-Diff) that merges a high-resolution (HR) multispectral image (MSI) with the corresponding LR-HSI. HSR-Diff generates an HR-HSI via repeated refinement, in which the HR-HSI is initialized with pure Gaussian noise and iteratively refined. At each iteration, the noise is removed with a Conditional Denoising Transformer (CDF ormer) that is trained on denoising at different noise levels, conditioned on the hierarchical feature maps of HR-MSI and LR-HSI. In addition, a progressive learning strategy is employed to exploit the global information of full-resolution images. Systematic experiments have been conducted on four public datasets, demonstrating that HSR-Diff outperforms state-of-the-art methods. | ['Ying Li', 'Hanyu Mao', 'Dong Wang', 'Chanyue Wu'] | 2023-06-21 | null | null | null | null | ['image-super-resolution', 'super-resolution'] | ['computer-vision', 'computer-vision'] | [ 8.55163515e-01 -2.26811737e-01 4.25184309e-01 -5.72119839e-02
-1.26551545e+00 -2.88132995e-01 7.07490683e-01 -4.65788096e-01
-4.77590635e-02 9.15762067e-01 3.38584810e-01 1.60833433e-01
-4.32014823e-01 -1.13120973e+00 -3.89034927e-01 -1.36938131e+00
2.38476127e-01 -2.82439664e-02 1.89600453e-01 -1.28936216e-01
8.24493729e-03 5.51312268e-01 -1.70721531e+00 1.96909755e-01
1.31440902e+00 1.07820880e+00 6.45346224e-01 6.77341521e-01
2.83462048e-01 6.32294595e-01 -2.12552845e-01 1.12784699e-01
4.34121490e-01 -6.74618125e-01 -5.13847053e-01 3.53910029e-01
3.68906111e-01 -2.58420110e-01 -5.29914856e-01 1.40966821e+00
3.30428839e-01 3.76517415e-01 6.87477171e-01 -6.35397851e-01
-1.02482510e+00 3.52875739e-01 -1.03029823e+00 9.03311558e-03
-1.12491772e-01 1.65986806e-01 8.23285580e-01 -8.51263344e-01
4.11013693e-01 1.14332533e+00 6.73722446e-01 4.62356135e-02
-1.53385615e+00 -5.40985942e-01 -1.24238670e-01 5.50437020e-03
-1.69128489e+00 -1.25809312e-01 7.16055036e-01 -3.80917579e-01
5.80845058e-01 1.92953765e-01 4.15765166e-01 9.00016963e-01
3.44093353e-01 7.07256421e-02 1.53765535e+00 -1.84895605e-01
3.03100139e-01 -2.22047076e-01 -1.73735470e-01 2.02563614e-01
2.12883621e-01 3.99816960e-01 -4.69377220e-01 -5.95436916e-02
9.67332184e-01 -7.21803904e-02 -4.98101950e-01 1.83694556e-01
-7.05955386e-01 6.62272096e-01 8.58149469e-01 4.30832744e-01
-9.86300647e-01 -3.62290204e-01 -3.75713080e-01 -1.28080517e-01
7.87384570e-01 1.02242313e-01 5.37245572e-02 6.22115374e-01
-1.27482522e+00 -2.05401890e-02 5.20025007e-02 5.16946137e-01
1.01094437e+00 2.44551152e-01 -2.38007739e-01 1.05763161e+00
2.14547858e-01 5.30974329e-01 1.03094354e-02 -1.10487533e+00
-5.82164489e-02 2.61943102e-01 1.93467155e-01 -7.67168939e-01
-1.59571141e-01 -9.06510115e-01 -1.69322145e+00 2.83880174e-01
-2.53792733e-01 4.02200259e-02 -1.14831996e+00 1.46834326e+00
1.98821574e-01 4.02831614e-01 3.55235398e-01 8.52981985e-01
7.47436106e-01 9.71590102e-01 2.72841185e-01 -3.46187383e-01
8.86358023e-01 -6.26137614e-01 -4.70525295e-01 -3.64291638e-01
-3.34213465e-01 -4.32442397e-01 5.86003661e-01 5.12283623e-01
-9.16153073e-01 -9.40509737e-01 -9.10423458e-01 1.54112035e-03
-2.40581810e-01 2.71652520e-01 2.33840793e-01 4.36327636e-01
-1.23024261e+00 7.01684892e-01 -6.31589413e-01 -2.46330366e-01
5.79661548e-01 -1.77714854e-01 -2.21997395e-01 -4.42710489e-01
-1.06338644e+00 6.83042347e-01 4.03649569e-01 5.61328709e-01
-1.18999910e+00 -8.98326159e-01 -7.25041211e-01 -5.11665717e-02
1.44191533e-01 -5.50502896e-01 4.04576689e-01 -7.07726240e-01
-1.36252153e+00 5.32351971e-01 -2.36013189e-01 -1.73259661e-01
2.82933325e-01 -1.15107432e-01 -5.45602560e-01 4.34001744e-01
2.14598477e-01 4.85500693e-01 1.23102403e+00 -1.97928107e+00
-8.12844872e-01 -5.74519813e-01 -3.31606567e-01 3.17987531e-01
-4.55566635e-03 -1.79414719e-01 -1.46557584e-01 -4.77919906e-01
5.92683971e-01 -7.16109514e-01 -4.14967000e-01 -4.38857824e-01
-4.97974455e-01 2.30114490e-01 6.99011505e-01 -1.16725016e+00
9.92700875e-01 -2.41450572e+00 5.32218099e-01 1.21231183e-01
1.64588124e-01 2.54024565e-01 -3.32172394e-01 1.09657012e-01
-2.53498912e-01 2.01267123e-01 -8.26622128e-01 -1.56080842e-01
-3.53638530e-01 1.14756629e-01 -3.62963170e-01 5.66174090e-01
4.80895609e-01 6.48277164e-01 -8.88485849e-01 -2.04242200e-01
3.51008922e-01 1.07808328e+00 6.77174404e-02 4.15201098e-01
-7.04570785e-02 1.02830434e+00 -1.81981906e-01 5.66533327e-01
1.28786302e+00 -2.32112542e-01 5.56671843e-02 -6.08950317e-01
-6.09279156e-01 -3.58862221e-01 -1.15481472e+00 1.43646824e+00
-2.17976049e-01 1.70855016e-01 3.82411361e-01 -6.67727232e-01
1.04146123e+00 1.54530749e-01 4.01432335e-01 -7.49076366e-01
-3.47019404e-01 1.26770943e-01 -4.22776580e-01 -1.15523860e-01
5.83413959e-01 -4.62746561e-01 4.37012136e-01 -4.93771918e-02
3.86297107e-02 -4.33418542e-01 -1.24027446e-01 1.66725889e-01
6.43497944e-01 3.94053102e-01 6.15353659e-02 -1.51171267e-01
6.90986931e-01 -4.90184426e-02 4.40670192e-01 9.88123238e-01
9.47628915e-02 9.97171462e-01 -8.10538232e-03 6.60425331e-03
-1.24203587e+00 -1.32629824e+00 -3.87205958e-01 8.23743701e-01
6.81786537e-02 1.15345299e-01 -4.71931100e-01 -1.10735834e-01
-3.53162318e-01 1.01064205e+00 -7.17204034e-01 -1.14167109e-01
-7.86470398e-02 -1.46981788e+00 2.25269303e-01 1.59345448e-01
1.05974376e+00 -7.81012654e-01 -1.84444025e-01 1.66482374e-01
-2.73715526e-01 -1.07725692e+00 2.36406818e-01 1.98197559e-01
-7.51543999e-01 -7.17741370e-01 -7.40650058e-01 -2.27526441e-01
3.21728945e-01 6.60857677e-01 8.81593764e-01 -4.32448328e-01
-4.77495551e-01 2.19963372e-01 -4.10632074e-01 9.89095196e-02
-5.54247916e-01 -1.54797897e-01 -5.11102557e-01 4.39156830e-01
-1.84684515e-01 -7.70922780e-01 -6.40735626e-01 -1.39176464e-02
-1.36759353e+00 2.85388768e-01 9.13716137e-01 9.10347879e-01
9.30346966e-01 1.04137456e+00 4.45574403e-01 -6.54713571e-01
1.06987357e-01 -6.36828959e-01 -6.94621146e-01 2.41469368e-01
-5.56040704e-01 -3.01812202e-01 2.94649839e-01 5.34712709e-03
-1.90876126e+00 6.49964064e-02 -2.91575044e-02 -2.71792948e-01
-5.25202155e-01 6.34921014e-01 -1.27036437e-01 -1.52148575e-01
5.72631001e-01 7.45477855e-01 -2.84804493e-01 -5.19648910e-01
4.24872220e-01 6.80839717e-01 9.11927521e-01 -1.84951052e-01
1.27480173e+00 7.18056262e-01 2.19648391e-01 -1.32214797e+00
-1.41162753e+00 -3.20789337e-01 -7.85280287e-01 -2.44085655e-01
1.18477702e+00 -1.23481441e+00 1.02317028e-01 8.17873716e-01
-6.56205893e-01 -3.60897839e-01 -1.71030179e-01 2.87542760e-01
-1.84757024e-01 3.79599631e-01 -5.73994815e-01 -1.15690029e+00
-2.37984121e-01 -8.80345941e-01 1.23957658e+00 6.52189136e-01
5.40985167e-01 -7.76881695e-01 -2.07387377e-02 5.11683881e-01
7.68515110e-01 3.76139343e-01 8.84657145e-01 2.02956185e-01
-8.35650086e-01 9.54146013e-02 -7.70132840e-01 8.50880563e-01
3.18760902e-01 6.81614801e-02 -1.19155383e+00 -2.90177882e-01
2.04310507e-01 -2.95618922e-01 1.25088274e+00 7.68048346e-01
1.22858202e+00 1.37427673e-01 7.59742483e-02 9.20344055e-01
2.17173314e+00 -6.14537974e-04 1.14420843e+00 2.87687778e-01
7.27617383e-01 4.98449594e-01 5.15793204e-01 5.52561462e-01
1.38373345e-01 1.22938782e-01 7.23764241e-01 -4.56983149e-01
-4.38999355e-01 -1.35561014e-02 6.98301196e-02 7.10303336e-02
-5.14356494e-01 -1.19239010e-01 -4.72850025e-01 3.63191009e-01
-1.46222532e+00 -1.16496611e+00 -3.54163706e-01 2.24304938e+00
7.66361952e-01 -2.93491781e-01 -3.70121419e-01 -7.16773048e-02
8.14504802e-01 6.38612807e-01 -7.35934675e-01 6.19432509e-01
-8.04021597e-01 4.42947894e-01 4.89541829e-01 5.54026544e-01
-1.14254928e+00 9.02764976e-01 5.73406649e+00 6.80717468e-01
-9.56310093e-01 1.05274685e-01 8.18875551e-01 3.82913619e-01
-2.61190534e-01 -1.53031677e-01 -6.80274963e-01 5.25828525e-02
7.67441928e-01 3.48879620e-02 7.41501033e-01 2.77560502e-01
3.09387803e-01 -5.22333503e-01 -2.29199022e-01 7.32723653e-01
-3.90690453e-02 -1.08682191e+00 1.36694551e-01 1.23965563e-02
1.14882851e+00 1.95245549e-01 3.77799332e-01 1.36292562e-01
4.41018045e-01 -1.00726163e+00 3.31651509e-01 9.37528312e-01
1.06020224e+00 -8.28957438e-01 7.09222436e-01 2.50416160e-01
-1.33410931e+00 -2.66560435e-01 -7.66277790e-01 5.79038382e-01
-8.33810121e-02 9.16606307e-01 -1.68343350e-01 1.35925817e+00
1.02025843e+00 9.38597918e-01 -6.71065331e-01 7.08712339e-01
-3.93788993e-01 6.48483217e-01 -5.40103987e-02 1.08583951e+00
1.38790563e-01 -8.71475101e-01 6.03474140e-01 1.09346890e+00
6.95195675e-01 5.70263922e-01 -7.81871974e-02 1.35368276e+00
2.10429624e-01 -4.54334646e-01 -2.74512947e-01 1.21443542e-02
2.33710244e-01 1.68960536e+00 -5.25639057e-01 -1.89719915e-01
-2.80419886e-01 1.09156549e+00 -1.74035087e-01 8.23390365e-01
-6.42624855e-01 6.51564077e-02 4.67730820e-01 -5.06059751e-02
4.97222722e-01 -1.48194224e-01 -1.26102939e-01 -1.02988946e+00
-4.17289525e-01 -7.85933495e-01 3.72349083e-01 -1.35795784e+00
-1.39687097e+00 9.06936228e-01 -1.85568094e-01 -1.12198734e+00
2.01639086e-01 -9.04600844e-02 -3.44300210e-01 1.44656014e+00
-2.05097246e+00 -1.44885135e+00 -9.24223542e-01 5.10294735e-01
4.16721970e-01 1.13824792e-01 7.87430227e-01 -6.11980483e-02
-7.04753399e-01 -3.55710596e-01 4.12536830e-01 -2.54657686e-01
4.20204222e-01 -1.17447889e+00 7.94878881e-03 1.28390253e+00
-3.71146053e-01 1.66553453e-01 7.72319734e-01 -8.97422850e-01
-1.12543881e+00 -1.66059911e+00 3.10347825e-01 7.95487985e-02
6.21478140e-01 2.16902450e-01 -1.22752655e+00 4.70997125e-01
2.24849150e-01 8.51330683e-02 4.98542398e-01 -4.10249680e-01
-4.16192919e-01 -3.50245386e-01 -1.27942085e+00 3.38042974e-01
8.23087454e-01 -5.81497729e-01 -2.17743620e-01 2.67968178e-01
6.77435696e-01 -1.11995175e-01 -1.04288411e+00 7.18422353e-01
7.17711076e-02 -1.41346598e+00 1.10995579e+00 7.45425820e-02
6.36735618e-01 -6.67577982e-01 -6.08966470e-01 -1.52046871e+00
-9.83827472e-01 -2.18309119e-01 2.58961231e-01 1.24192011e+00
7.04674870e-02 -4.96998072e-01 2.50070751e-01 3.61904770e-01
-2.83959627e-01 -3.77064906e-02 -8.43823075e-01 -5.92688918e-01
-8.30272362e-02 -3.27294506e-02 4.78589416e-01 6.99028373e-01
-1.02163780e+00 2.48405635e-01 -4.06441033e-01 1.13940156e+00
1.41582835e+00 2.35623211e-01 2.24480838e-01 -1.48945463e+00
-1.15945987e-01 -3.48302811e-01 3.21219712e-01 -4.11238849e-01
2.00778786e-02 -4.75341022e-01 2.70344496e-01 -1.87382615e+00
4.53645915e-01 -1.58866927e-01 -4.22851145e-01 2.17715248e-01
-3.86809021e-01 6.51943564e-01 -1.08479120e-01 2.99195856e-01
-2.64267594e-01 7.88613260e-01 1.23744845e+00 -1.02483943e-01
-2.13237837e-01 -1.29855290e-01 -6.78888083e-01 5.09251535e-01
5.89169860e-01 -3.56766522e-01 -1.50081009e-01 -1.82660818e-01
4.87646423e-02 1.95882156e-01 7.53844678e-01 -1.19280720e+00
1.00923389e-01 -2.59706527e-01 7.54712760e-01 -8.73015583e-01
4.31780964e-01 -6.22434020e-01 6.99024200e-01 6.00213371e-02
-2.50589669e-01 -6.87690437e-01 -3.60017866e-02 6.70149803e-01
-2.47985944e-01 -8.14816505e-02 1.40741551e+00 -2.16628641e-01
-6.57338381e-01 4.70907480e-01 -2.10146815e-01 -4.98171240e-01
5.92776239e-01 -5.81596121e-02 -4.48006690e-01 -2.39080220e-01
-6.82464659e-01 -2.59935051e-01 4.47077602e-01 -1.70904230e-02
7.27483451e-01 -8.70264709e-01 -1.21308649e+00 3.25439662e-01
3.80013809e-02 3.79811168e-01 9.23274517e-01 6.42940223e-01
-1.34460479e-01 6.30128607e-02 7.56704481e-03 -6.25727832e-01
-8.41396689e-01 3.44327271e-01 5.03798723e-01 -2.95706093e-01
-7.59903014e-01 7.48072565e-01 3.06561291e-01 -2.21022099e-01
-1.61629260e-01 1.49288625e-01 -3.56330067e-01 1.41597548e-02
6.72547758e-01 5.32312751e-01 -8.28581899e-02 -8.15290987e-01
1.32534251e-01 5.42174935e-01 4.00068134e-01 -2.18029127e-01
1.93515670e+00 -5.56763887e-01 -6.63441718e-01 2.98301607e-01
6.64812565e-01 -1.72906667e-01 -1.71460545e+00 -5.20734310e-01
-3.29550087e-01 -5.96506834e-01 7.84153163e-01 -9.51086938e-01
-1.16053784e+00 5.57820976e-01 7.17068911e-01 2.84481317e-01
1.67782187e+00 -9.39764902e-02 2.96269864e-01 -1.60487950e-01
3.40711296e-01 -8.73851299e-01 -1.76419057e-02 4.30211186e-01
1.01985013e+00 -1.22343612e+00 1.93747088e-01 -3.92058611e-01
-6.30342662e-01 9.61726964e-01 2.34387249e-01 -5.59295267e-02
6.17592275e-01 5.03248200e-02 -1.89432785e-01 -2.56118029e-02
-3.65234226e-01 -4.83511388e-01 2.06440613e-01 8.64310086e-01
1.87481865e-01 -1.95888523e-02 3.79185647e-01 3.54031116e-01
1.68332145e-01 -7.13291764e-03 4.14108336e-01 3.76993865e-01
-6.47464037e-01 -5.83008647e-01 -6.71858907e-01 3.71288568e-01
-7.66688511e-02 -3.28838408e-01 7.86220655e-02 3.49596709e-01
2.13439032e-01 1.27614760e+00 -1.20857656e-01 -1.94287330e-01
-5.10149188e-02 -9.68379825e-02 2.46445566e-01 -5.45526624e-01
-2.11667255e-01 5.43507755e-01 -3.89923573e-01 -5.10583043e-01
-7.54296064e-01 -8.29047859e-01 -7.50307143e-01 -2.37821694e-02
-8.31407532e-02 -1.08784465e-02 5.81000149e-01 7.76246071e-01
1.86646372e-01 7.02849090e-01 8.22792590e-01 -1.20744848e+00
-5.39526761e-01 -1.17314982e+00 -1.34171617e+00 -8.65113549e-03
5.56987047e-01 -3.51427644e-01 -8.47269952e-01 2.73672887e-03] | [10.227842330932617, -1.9520615339279175] |
1cedfe03-7490-43e6-9fa8-ed3190cc6827 | hierarchical-pretraining-for-biomedical-term | 2307.00266 | null | https://arxiv.org/abs/2307.00266v1 | https://arxiv.org/pdf/2307.00266v1.pdf | Hierarchical Pretraining for Biomedical Term Embeddings | Electronic health records (EHR) contain narrative notes that provide extensive details on the medical condition and management of patients. Natural language processing (NLP) of clinical notes can use observed frequencies of clinical terms as predictive features for downstream applications such as clinical decision making and patient trajectory prediction. However, due to the vast number of highly similar and related clinical concepts, a more effective modeling strategy is to represent clinical terms as semantic embeddings via representation learning and use the low dimensional embeddings as feature vectors for predictive modeling. To achieve efficient representation, fine-tuning pretrained language models with biomedical knowledge graphs may generate better embeddings for biomedical terms than those from standard language models alone. These embeddings can effectively discriminate synonymous pairs of from those that are unrelated. However, they often fail to capture different degrees of similarity or relatedness for concepts that are hierarchical in nature. To overcome this limitation, we propose HiPrBERT, a novel biomedical term representation model trained on additionally complied data that contains hierarchical structures for various biomedical terms. We modify an existing contrastive loss function to extract information from these hierarchies. Our numerical experiments demonstrate that HiPrBERT effectively learns the pair-wise distance from hierarchical information, resulting in a substantially more informative embeddings for further biomedical applications | ['Lu Tian', 'Doudou Zhou', 'Zheng Yuan', 'Yucong Lin', 'Sihang Zeng', 'Bryan Cai'] | 2023-07-01 | null | null | null | null | ['trajectory-prediction', 'knowledge-graphs', 'management', 'decision-making'] | ['computer-vision', 'knowledge-base', 'miscellaneous', 'reasoning'] | [ 6.89305887e-02 1.49804473e-01 -5.75675488e-01 -3.79048049e-01
-7.73239613e-01 -3.00269902e-01 2.28515029e-01 1.24853981e+00
-2.82665253e-01 5.84969699e-01 9.17609394e-01 -3.94474328e-01
-5.61732054e-01 -8.86509895e-01 -1.01409018e-01 -7.34704673e-01
-5.51361561e-01 5.26520908e-01 -3.50996107e-01 2.54773982e-02
-9.49841663e-02 4.40744191e-01 -9.38122094e-01 3.16995889e-01
1.01093650e+00 7.81452239e-01 -3.59808095e-02 5.20067692e-01
-3.64378750e-01 6.73472822e-01 -3.50234061e-01 -3.36334884e-01
-1.55243546e-01 -3.32065225e-01 -7.80503213e-01 -3.60840440e-01
-1.62533015e-01 5.07132448e-02 -6.64570391e-01 1.03663051e+00
6.28431201e-01 -1.22347856e-02 1.06495237e+00 -9.22416091e-01
-1.26101637e+00 4.25814062e-01 -2.22079724e-01 1.58086345e-01
4.09060627e-01 -2.49272764e-01 1.41634560e+00 -6.99307561e-01
6.51745379e-01 1.06062806e+00 8.98204625e-01 6.28962159e-01
-1.21533501e+00 -5.30066609e-01 -3.58629487e-02 2.43039981e-01
-1.55686259e+00 -5.69302216e-02 4.98586625e-01 -4.94052231e-01
1.02808523e+00 4.32229728e-01 6.44460857e-01 1.08320940e+00
6.08611941e-01 3.54643852e-01 3.80186647e-01 -1.66597143e-01
8.54634717e-02 9.63745564e-02 3.96127224e-01 8.58335614e-01
5.24731517e-01 -1.66560069e-01 -2.31447995e-01 -8.87223661e-01
4.28177625e-01 7.28325248e-01 -6.25880182e-01 -1.86792120e-01
-1.21542156e+00 1.11434865e+00 7.06378222e-01 5.42063117e-01
-4.67091531e-01 -2.51384526e-02 6.34637296e-01 -5.68056963e-02
3.51544112e-01 8.25003505e-01 -4.47270989e-01 1.18109316e-01
-5.93985617e-01 1.04562044e-02 6.50212467e-01 8.64808857e-01
3.12561721e-01 -4.42019284e-01 -4.50133085e-01 1.02041054e+00
4.18805361e-01 -7.16918856e-02 1.12756205e+00 -4.08699721e-01
3.29108417e-01 9.81675923e-01 -2.98528314e-01 -1.20650792e+00
-4.62846637e-01 -3.71379852e-01 -1.13680553e+00 -8.46068084e-01
-2.88179487e-01 1.92585498e-01 -9.13192570e-01 1.57255518e+00
1.73052490e-01 2.12644905e-01 3.99910420e-01 5.63466072e-01
1.10740924e+00 8.19033146e-01 5.38737714e-01 -1.85147777e-01
1.87020612e+00 -5.54250360e-01 -1.01949501e+00 1.82382632e-02
1.19822764e+00 -3.82364780e-01 7.36895442e-01 -2.83984393e-01
-6.81083441e-01 4.35426310e-02 -9.40144479e-01 -3.99242282e-01
-5.50882161e-01 -1.87766269e-01 7.30470300e-01 2.35765487e-01
-7.20320046e-01 5.92699885e-01 -9.03794706e-01 -3.89796972e-01
7.09896147e-01 2.39217415e-01 -4.55854863e-01 -4.96035635e-01
-1.57127380e+00 7.80629396e-01 3.53242427e-01 -1.33018345e-01
-2.28792474e-01 -1.19359171e+00 -1.30294001e+00 4.27764535e-01
-2.21925735e-01 -9.51541901e-01 6.58090532e-01 -7.33779073e-02
-7.86246777e-01 8.44539285e-01 -1.88812435e-01 -3.35431159e-01
-1.35898143e-01 2.90984288e-02 -6.12120092e-01 2.58834243e-01
1.79086119e-01 3.68423909e-01 8.13290551e-02 -7.40379453e-01
-2.90030420e-01 -4.72132772e-01 -3.51688862e-01 7.43666142e-02
-9.16562676e-01 -2.43822068e-01 -2.31315896e-01 -9.15384233e-01
2.52051521e-02 -6.89771950e-01 -6.20051920e-01 2.51757085e-01
-4.53091055e-01 -4.39932853e-01 2.60643780e-01 -5.90057254e-01
1.66015267e+00 -2.12616587e+00 4.97720689e-02 2.38836948e-02
6.15304172e-01 2.65711427e-01 -6.38697073e-02 6.56118453e-01
-1.64637640e-01 4.86003995e-01 -3.40021402e-01 9.98145193e-02
-2.25738183e-01 3.02046001e-01 -1.40345693e-01 4.11019444e-01
4.16541278e-01 1.05286300e+00 -1.23347223e+00 -7.12260664e-01
-1.11861132e-01 6.56921983e-01 -7.12223768e-01 4.35711443e-01
5.12136780e-02 -6.76776692e-02 -8.80924761e-01 7.26336718e-01
3.36421162e-01 -7.23506510e-01 4.23224747e-01 -4.11704570e-01
5.15568376e-01 4.51667845e-01 -2.92327374e-01 1.61467481e+00
-3.47087413e-01 3.65541458e-01 -4.66943145e-01 -1.21729314e+00
7.37113953e-01 5.36917150e-01 8.92256141e-01 -2.41411030e-01
3.44434730e-03 4.35045883e-02 5.02776802e-02 -7.61697590e-01
1.61413908e-01 -4.44432855e-01 -2.67529547e-01 3.14046107e-02
-1.22613244e-01 2.53729343e-01 -2.50652015e-01 2.56394267e-01
1.53514564e+00 -5.50939977e-01 8.74586046e-01 -1.71497628e-01
4.19544131e-01 1.57214060e-01 9.39743519e-01 1.68773621e-01
-3.24825585e-01 5.49195468e-01 4.96840179e-01 -3.83767813e-01
-6.46030366e-01 -1.24264252e+00 -7.62374520e-01 7.18774438e-01
-9.74623859e-03 -8.89610410e-01 -4.94254986e-03 -6.45894587e-01
3.70366871e-01 5.03584683e-01 -7.44949520e-01 -6.72613025e-01
-3.75522733e-01 -1.14260268e+00 6.65451944e-01 7.93299794e-01
-2.39075482e-01 -7.01052904e-01 -2.42340237e-01 4.65612978e-01
-1.09627023e-01 -9.21934783e-01 -8.10675025e-01 3.45578551e-01
-1.01894355e+00 -1.26029038e+00 -7.23149300e-01 -1.06313360e+00
8.17335784e-01 -1.05383970e-01 1.02266860e+00 1.44380108e-01
-9.11281645e-01 2.12189883e-01 -5.01171887e-01 -2.16079265e-01
-3.45689207e-01 -1.97920218e-01 9.87189114e-02 -3.25138599e-01
7.69720852e-01 -4.26527321e-01 -9.76647019e-01 -2.27043778e-01
-1.08642006e+00 -2.89484173e-01 5.46404243e-01 1.33726954e+00
6.20162129e-01 -2.22558334e-01 7.34393358e-01 -1.03047776e+00
9.41704273e-01 -8.50368917e-01 1.06823429e-01 4.73708957e-01
-8.61979425e-01 3.87708962e-01 7.40540564e-01 -3.81943613e-01
-5.29314697e-01 -3.21783066e-01 -2.13039771e-01 -4.92123634e-01
1.59481943e-01 9.25752282e-01 1.46696687e-01 4.41692889e-01
4.63653356e-01 2.79066503e-01 -1.68338150e-01 -3.82208467e-01
3.15035939e-01 9.76857483e-01 2.05514163e-01 -3.15295637e-01
2.78846174e-01 2.16293335e-01 8.01052004e-02 -7.06936002e-01
-7.49641061e-01 -7.98158884e-01 -2.98871011e-01 6.61339283e-01
1.25542212e+00 -9.17733133e-01 -5.16521096e-01 -4.09940302e-01
-1.08998156e+00 4.46180493e-01 -7.78102279e-02 7.41979480e-01
-2.50358582e-01 4.21033680e-01 -1.08753729e+00 -4.22201127e-01
-5.87026775e-01 -8.78000736e-01 9.90455687e-01 -6.96245283e-02
-6.18366122e-01 -1.41597033e+00 3.85463566e-01 1.25924289e-01
7.03791156e-02 3.43227386e-01 1.94010246e+00 -1.04050374e+00
-3.00164334e-02 -5.35955846e-01 -2.57613987e-01 1.28966002e-02
7.88686335e-01 -1.98574588e-01 -6.43955410e-01 -2.44392708e-01
-1.77005485e-01 -1.31547421e-01 9.09119666e-01 3.29342604e-01
1.44699967e+00 -5.87889373e-01 -8.91506910e-01 8.75070393e-01
1.46246743e+00 3.44016761e-01 4.41116959e-01 2.76182983e-02
8.84413779e-01 6.18654191e-01 3.48443508e-01 4.61259007e-01
4.90734845e-01 4.45918411e-01 2.89534274e-02 -2.76006199e-02
2.94927299e-01 -3.25128704e-01 9.60532352e-02 1.23282635e+00
3.02998096e-01 -1.93205595e-01 -1.14362741e+00 8.04469347e-01
-1.68673694e+00 -7.43455291e-01 8.80154893e-02 1.90142715e+00
1.32064116e+00 -3.77950937e-01 -3.81958455e-01 -1.83690831e-01
6.46436870e-01 -1.88491380e-04 -4.40444142e-01 -5.18295825e-01
3.31106216e-01 2.07656652e-01 3.64664972e-01 1.04475744e-01
-8.62603068e-01 5.24781406e-01 6.07607651e+00 6.80295587e-01
-8.58194232e-01 -1.08938534e-02 5.89177370e-01 -2.70808805e-02
-5.28262377e-01 -4.01582092e-01 -5.90731621e-01 5.35266519e-01
1.04834962e+00 -5.84828734e-01 -2.24798858e-01 7.08168805e-01
1.26683205e-01 7.31967330e-01 -1.48050547e+00 1.06392956e+00
1.55025348e-02 -1.62674749e+00 5.45249045e-01 1.59753591e-01
5.22047579e-01 -2.39730984e-01 2.45683752e-02 2.46441081e-01
2.53143281e-01 -1.44336462e+00 -3.26127589e-01 4.70453322e-01
1.04024470e+00 -5.76974213e-01 9.36050355e-01 5.44770211e-02
-1.34236586e+00 -1.37380227e-01 -6.39416814e-01 5.00737369e-01
8.24537426e-02 6.63575709e-01 -1.10863721e+00 7.88827717e-01
4.34635639e-01 1.10710394e+00 -2.27467299e-01 9.49255109e-01
1.48073420e-01 5.09873867e-01 1.82931677e-01 -5.81141226e-02
2.31920868e-01 -9.25522596e-02 2.20835820e-01 1.50527799e+00
5.69763482e-01 4.91335213e-01 2.37973109e-01 7.56209850e-01
-3.15528512e-01 4.45865691e-01 -8.63065124e-01 -6.46095514e-01
6.51754022e-01 1.06913233e+00 -3.71797472e-01 -3.58296722e-01
-4.74411160e-01 8.31771612e-01 4.45700109e-01 1.81562454e-01
-6.81362867e-01 -5.74739516e-01 1.22457039e+00 1.45323828e-01
1.27368188e-03 1.69135779e-01 -1.29831225e-01 -1.31363881e+00
-3.09459984e-01 -6.18824422e-01 7.88879871e-01 -3.59497368e-01
-1.91666746e+00 7.40702033e-01 -2.86596894e-01 -1.47198951e+00
-2.85434037e-01 -7.30782688e-01 -4.52501327e-01 7.99647570e-01
-1.60339856e+00 -8.54962707e-01 3.31694521e-02 4.10325795e-01
2.62281954e-01 -1.83564857e-01 1.50783539e+00 3.31749976e-01
-5.27351975e-01 7.54854679e-01 2.98873335e-01 4.46781039e-01
8.03376615e-01 -1.34420979e+00 -8.59529823e-02 -3.35782096e-02
-1.00113980e-01 1.22715056e+00 3.53409111e-01 -5.51884651e-01
-1.17097068e+00 -1.57000732e+00 1.20830023e+00 -2.56238222e-01
7.05271244e-01 -1.70822889e-01 -1.33200717e+00 5.87605238e-01
-4.25775684e-02 3.34415734e-01 1.71409380e+00 1.52226999e-01
-4.84520614e-01 1.02973040e-02 -9.98802543e-01 7.36040294e-01
1.03992569e+00 -7.11150527e-01 -1.03921676e+00 5.21585107e-01
1.01524341e+00 2.25734532e-01 -1.55207121e+00 4.64304179e-01
3.56347382e-01 -2.17385113e-01 1.19600415e+00 -1.29495811e+00
7.26929069e-01 6.46289811e-02 -1.63814887e-01 -1.41490626e+00
-5.25157928e-01 -2.20464781e-01 -1.54159427e-01 9.15344357e-01
4.20994997e-01 -7.50996470e-01 4.97403681e-01 5.91053367e-01
-6.20833151e-02 -1.35362995e+00 -8.73140931e-01 -5.38293421e-01
3.00049126e-01 3.82920578e-02 5.71321189e-01 1.36607933e+00
5.98232985e-01 3.65327239e-01 6.40778318e-02 2.03622416e-01
3.06717783e-01 3.24077964e-01 -9.02357399e-02 -1.43943107e+00
-6.33618608e-02 -4.06399280e-01 -9.44129050e-01 -8.00849855e-01
2.61645347e-01 -1.48279965e+00 -2.38325130e-02 -1.90771115e+00
5.61534703e-01 -6.01005971e-01 -8.81971717e-01 4.29025620e-01
-6.22774959e-01 -1.17222689e-01 -3.59884918e-01 3.26579422e-01
-2.87499309e-01 7.90862262e-01 1.04595554e+00 -4.26313907e-01
-7.86978602e-02 -3.35224658e-01 -9.33719277e-01 5.60170889e-01
4.64402884e-01 -7.51938760e-01 -3.96450102e-01 -2.13047341e-01
6.65965080e-02 2.78021932e-01 9.10622105e-02 -5.03340364e-01
2.01052010e-01 -2.20531061e-01 3.87477309e-01 -1.73928812e-01
1.95927083e-01 -7.47691810e-01 -1.89489588e-01 7.57662416e-01
-8.28144014e-01 1.82446778e-01 3.96356406e-03 9.33018804e-01
-4.73230988e-01 -1.59835353e-01 3.80592346e-01 5.34872152e-02
-3.00212979e-01 6.40398622e-01 -3.56927782e-01 2.04381287e-01
1.01032829e+00 1.95413195e-02 -1.55107826e-01 -9.11470801e-02
-8.73492122e-01 3.76601577e-01 2.48726040e-01 3.94424528e-01
9.32142437e-01 -1.66199052e+00 -6.47790134e-01 -2.14549527e-02
6.10529661e-01 -6.05605207e-02 1.35727003e-01 6.93178535e-01
-5.91493547e-01 6.25656128e-01 1.14839025e-01 -3.45299602e-01
-1.34864151e+00 8.38598788e-01 1.48002893e-01 -6.68963075e-01
-7.84283519e-01 7.50774086e-01 5.61092198e-01 -4.19852912e-01
9.96709540e-02 -5.99339068e-01 -4.89084512e-01 2.14394644e-01
5.40636778e-01 -1.32974029e-01 2.37735771e-02 -3.30791086e-01
-7.65349984e-01 5.61550617e-01 -2.50395894e-01 4.91204798e-01
1.65129709e+00 2.42745608e-01 -2.61282325e-01 3.95651042e-01
1.85918641e+00 -5.24517745e-02 -4.50697929e-01 -2.34180033e-01
2.56363362e-01 -3.42988700e-01 6.63890243e-02 -4.07089502e-01
-7.55469561e-01 1.00543797e+00 3.31121236e-01 9.42639783e-02
9.89627540e-01 1.69037908e-01 9.14226353e-01 4.51274365e-01
-1.74191706e-02 -5.71593106e-01 -2.68846611e-03 2.24610344e-01
6.65747166e-01 -9.24358666e-01 8.13809931e-02 -5.78233004e-01
-5.87152600e-01 1.21069503e+00 1.93102747e-01 9.23562124e-02
9.07090306e-01 1.51593536e-01 2.85241730e-03 -4.97071743e-01
-9.49127078e-01 -4.21707751e-03 4.73975062e-01 5.94618618e-01
9.38359380e-01 2.17909634e-01 -6.18290126e-01 1.04581606e+00
1.20555736e-01 -1.02120183e-01 1.92125663e-01 8.43655348e-01
-1.38132766e-01 -1.15329039e+00 -3.86450626e-02 9.72790480e-01
-6.86388969e-01 -5.80384552e-01 -9.46114808e-02 3.71390492e-01
-5.34088649e-02 6.97522402e-01 3.79017182e-02 -3.11556280e-01
1.96233526e-01 2.77283281e-01 8.57467949e-02 -1.06899250e+00
-4.23296064e-01 -2.46270046e-01 -5.62093407e-02 -3.18979859e-01
-9.99492556e-02 -3.06628615e-01 -1.63216567e+00 1.86767448e-02
-2.21926078e-01 4.30371255e-01 9.23257228e-03 6.85289562e-01
6.89678371e-01 8.22551250e-01 4.53970373e-01 -6.31037429e-02
-8.07636380e-01 -6.32310688e-01 -6.46947801e-01 7.87218571e-01
4.12856996e-01 -5.12952566e-01 -2.36621529e-01 -9.81550850e-03] | [7.90706729888916, 6.853460311889648] |
c6661b00-0a1e-450b-bbeb-df72f97f56de | authorship-verification-in-the-absence-of | null | null | https://link.springer.com/chapter/10.1007/978-3-319-76941-7_34 | https://link.springer.com/chapter/10.1007/978-3-319-76941-7_34 | Authorship verification in the absence of explicit features and thresholds | Enhancing information retrieval systems with the ability to take the writing style of people into account opens the door for a number of applications. For example, one can link articles by authorships that can help identifying authors who generate hoaxes and deliberate misinformation in news stories, distributed across different platforms. Authorship verification (AV) is a technique that can be used for this purpose. AV deals with the task to judge, whether two or more documents stem from the same author. The majority of existing AV approaches relies on machine learning concepts based on explicitly defined stylistic features and complex models that involve a fair amount of parameters. Moreover, many existing AV methods are based on explicit thresholds (needed to accept or reject a stated authorship), which are determined on training corpora. We propose a novel parameter-free AV approach, which derives its thresholds for each verification case individually and enables AV in the absence of explicit features and training corpora. In an experimental setup based on eight evaluation corpora (each one from another language) we show that our approach yields competitive results against the current state of the art and other noteworthy AV baselines. | ['Inna Vogel', 'Lukas Graner', 'Oren Halvani'] | 2018-03-01 | null | null | null | null | ['authorship-verification'] | ['natural-language-processing'] | [ 8.39071944e-02 -1.19299062e-01 -3.72270525e-01 -1.18451901e-01
-6.55887306e-01 -9.12036061e-01 1.27050078e+00 5.49758077e-01
-5.69905102e-01 5.48414290e-01 8.47856775e-02 -3.23426783e-01
-1.76444560e-01 -6.09122336e-01 -2.93567747e-01 -2.60035068e-01
3.90725076e-01 7.69484222e-01 2.54401416e-01 -1.93394825e-01
8.80820155e-01 5.37288249e-01 -1.41587615e+00 2.11647853e-01
8.34101558e-01 8.70007634e-01 -7.04813674e-02 5.57082772e-01
-4.07786369e-01 6.86675072e-01 -1.06581533e+00 -8.16842675e-01
5.70500493e-02 -3.18774521e-01 -8.73687446e-01 -7.56099224e-02
4.60962296e-01 1.01990536e-01 1.57478929e-01 9.87643421e-01
1.89499691e-01 -3.65516186e-01 1.20596957e+00 -1.14798439e+00
-5.07862389e-01 8.38087499e-01 -5.72630346e-01 3.42785835e-01
3.74472767e-01 -1.39720902e-01 1.19872010e+00 -6.76579118e-01
7.90718615e-01 1.12031353e+00 5.22659123e-01 2.12357476e-01
-1.08691847e+00 -4.90135342e-01 -1.21643201e-01 3.50709200e-01
-1.08531344e+00 -2.08842009e-01 8.80249918e-01 -8.86300504e-01
3.49610001e-01 2.22642049e-01 2.77538270e-01 1.30663466e+00
-7.87255317e-02 6.72319293e-01 1.24879932e+00 -7.46983528e-01
1.52292982e-01 5.53367138e-01 6.16101205e-01 4.68330592e-01
5.27896583e-01 -3.69888455e-01 -5.25820553e-01 -3.35173547e-01
2.64815509e-01 -4.67167109e-01 -2.02104405e-01 -1.85567543e-01
-1.30220568e+00 9.31993127e-01 -1.16210617e-01 7.10322797e-01
-2.03971967e-01 -1.95671335e-01 7.82381356e-01 2.96399593e-01
2.12914601e-01 1.05499279e+00 -2.48768732e-01 -2.30538100e-01
-1.30010593e+00 2.78590679e-01 1.08460438e+00 7.66372025e-01
5.51329494e-01 -4.30098355e-01 -5.24199903e-01 6.30873263e-01
2.31349915e-01 4.01060730e-01 6.53483808e-01 -7.34551549e-01
4.46104467e-01 1.02371514e+00 3.24712366e-01 -1.07835305e+00
-1.53158158e-01 -5.30938208e-01 -5.77839434e-01 -5.25088571e-02
6.65323853e-01 -1.54191488e-02 -4.96569365e-01 1.34164715e+00
6.67562038e-02 -2.96910137e-01 -1.90877561e-02 6.04693294e-01
8.37220609e-01 3.99620593e-01 -3.10642183e-01 -3.04469019e-01
1.69253123e+00 -6.24508739e-01 -7.10268080e-01 1.47750407e-01
5.59559226e-01 -1.07585728e+00 1.20194900e+00 8.04562390e-01
-7.27594852e-01 -2.60473192e-01 -1.05964065e+00 -9.60597489e-03
-6.10095978e-01 5.86245000e-01 1.78687707e-01 9.56778169e-01
-6.45670056e-01 5.88434160e-01 -2.47816041e-01 -3.21715415e-01
1.43024340e-01 7.08338171e-02 -2.20218018e-01 4.43932950e-01
-1.15138328e+00 8.89477611e-01 1.81883693e-01 -1.62297897e-02
-3.63428861e-01 -4.84244406e-01 -5.14465511e-01 1.11710653e-02
6.07405543e-01 -4.05695647e-01 1.16644311e+00 -1.14961612e+00
-1.58014321e+00 1.07318079e+00 -1.24821939e-01 -4.76669043e-01
1.10639882e+00 -2.95095980e-01 -2.80162156e-01 3.21135998e-01
1.78581282e-01 9.38376859e-02 9.63803113e-01 -1.12262464e+00
-3.22901785e-01 -3.04108769e-01 -1.14527516e-01 -2.72605002e-01
-6.04533613e-01 5.38049877e-01 -5.36997199e-01 -9.15295601e-01
-2.91436136e-01 -8.58237565e-01 3.43012869e-01 -1.38037294e-01
-7.45449126e-01 -5.73419154e-01 6.84312761e-01 -8.18416238e-01
1.57195950e+00 -1.80087185e+00 2.65752643e-01 5.77836454e-01
2.89438903e-01 4.08497185e-01 1.92036465e-01 4.84359115e-01
5.91220975e-01 4.37772423e-01 7.27835521e-02 -3.37526560e-01
2.89401591e-01 -2.58834094e-01 -2.41537005e-01 2.68554747e-01
2.18075309e-02 6.31912351e-01 -6.43112481e-01 -7.60832250e-01
-1.97247848e-01 2.89200008e-01 -1.50651962e-01 6.59850612e-02
-1.15935124e-01 1.17825389e-01 -3.05198699e-01 3.59302372e-01
3.78247976e-01 -4.08133179e-01 2.98359811e-01 5.99222854e-02
-1.75096676e-01 4.08994406e-01 -1.29319060e+00 1.05729973e+00
-2.39589617e-01 8.36250305e-01 -2.88117319e-01 -6.93032801e-01
1.10958207e+00 1.78535938e-01 -7.31399581e-02 -3.15327317e-01
4.78436619e-01 4.84532684e-01 -5.80228530e-02 -2.56428301e-01
7.79348791e-01 3.09692383e-01 -3.65151823e-01 8.75330806e-01
-8.08822587e-02 2.53475726e-01 7.31545150e-01 4.53645051e-01
9.49003041e-01 -2.23209247e-01 3.46599877e-01 -2.84625202e-01
9.83514845e-01 -2.17177942e-01 3.56475532e-01 9.73022759e-01
2.58701686e-02 5.75510919e-01 9.82231319e-01 -2.22948298e-01
-1.18526626e+00 -5.62735975e-01 -1.73736468e-01 8.05602491e-01
7.05498680e-02 -4.68251079e-01 -9.39528763e-01 -8.09219658e-01
1.36538660e-02 6.14860117e-01 -7.98204541e-01 2.73369014e-01
-5.26001513e-01 -3.68489206e-01 7.58047223e-01 2.22979471e-01
4.05234009e-01 -1.00645900e+00 -7.72043645e-01 7.03039914e-02
-2.04732522e-01 -1.07403672e+00 -3.73504579e-01 -1.56463072e-01
-4.61295038e-01 -1.29375350e+00 -9.22349274e-01 -3.98001045e-01
4.30550069e-01 -9.90768522e-02 9.70588326e-01 3.42045724e-01
-6.46855459e-02 1.50313959e-01 -5.54915071e-01 -5.27543128e-01
-9.06276047e-01 6.37598395e-01 -1.04226574e-01 2.35303894e-01
4.86194938e-01 -1.21832214e-01 -1.28665492e-01 3.38940203e-01
-7.93226242e-01 -9.27376896e-02 6.63867116e-01 8.43969226e-01
3.24980691e-02 -3.12891006e-01 5.81771433e-01 -1.36184525e+00
8.96296978e-01 -3.07280391e-01 -6.93326831e-01 6.61776423e-01
-8.99850965e-01 4.65948015e-01 5.90889513e-01 -5.90582788e-01
-7.93385208e-01 -3.80378306e-01 2.45043516e-01 -1.49132118e-01
-1.58612177e-01 6.37608588e-01 -1.82209864e-01 6.45822063e-02
6.01059914e-01 1.01622507e-01 9.35343653e-02 -7.06825912e-01
2.64092118e-01 9.39197004e-01 5.16106308e-01 -5.61811984e-01
8.32575500e-01 8.51823613e-02 -1.31579727e-01 -6.42689407e-01
-7.30994761e-01 -7.20351815e-01 -8.16294968e-01 -2.13614270e-01
3.62981230e-01 -4.51058388e-01 -6.41640723e-01 5.10263622e-01
-1.40016937e+00 5.09535037e-02 1.33756220e-01 3.03375036e-01
-2.30893701e-01 6.85747147e-01 -5.48407912e-01 -7.04128623e-01
-4.53152567e-01 -1.01759171e+00 7.63609588e-01 1.08842291e-01
-8.09447765e-01 -8.43247056e-01 1.02716431e-01 5.66032290e-01
2.19619259e-01 1.91568539e-01 1.04403341e+00 -1.19697571e+00
-2.43650988e-01 -4.54007894e-01 -2.21477747e-01 2.31698588e-01
-1.23413920e-01 5.06034434e-01 -7.60824203e-01 5.66189513e-02
-3.37141693e-01 -1.44641727e-01 8.23181808e-01 -5.03566004e-02
8.35405350e-01 -6.22699201e-01 -2.53054887e-01 4.24120352e-02
1.09128106e+00 -2.70041436e-01 6.02002800e-01 7.08345413e-01
5.91305852e-01 7.54420936e-01 3.18964154e-01 5.66110790e-01
2.85393685e-01 9.54943657e-01 7.55999237e-02 2.60795444e-01
-4.90748957e-02 -1.10951118e-01 2.22074166e-01 5.20919859e-01
-1.28040463e-01 -4.12201852e-01 -8.97230744e-01 4.46277082e-01
-1.64052188e+00 -1.08954668e+00 -4.58637565e-01 2.23417115e+00
1.00261331e+00 3.64859641e-01 4.39586639e-01 7.22964764e-01
8.01156163e-01 -4.28948738e-02 -4.54801247e-02 -6.58857882e-01
-2.85841256e-01 -3.37297693e-02 3.98205101e-01 3.96901995e-01
-1.03858638e+00 6.90589607e-01 5.88146257e+00 7.04095006e-01
-1.16861522e+00 1.64505970e-02 4.03213292e-01 2.78261214e-01
-3.61984670e-01 -9.33490172e-02 -1.01871932e+00 5.73953509e-01
7.20037341e-01 -2.33560011e-01 -6.13320945e-03 5.78054488e-01
8.24028328e-02 -1.46527976e-01 -1.05840898e+00 9.70017552e-01
4.55427229e-01 -1.35091782e+00 1.84202433e-01 1.72148809e-01
5.69874287e-01 -5.17159581e-01 9.51285735e-02 7.31329769e-02
5.64510636e-02 -7.19720542e-01 1.14326656e+00 4.90257651e-01
5.02960324e-01 -5.97263455e-01 9.95532453e-01 4.34913963e-01
-5.46220005e-01 1.78836696e-02 -7.47185946e-03 1.07881069e-01
-2.16871072e-02 6.14082456e-01 -8.84217918e-01 5.08425474e-01
3.58472258e-01 5.60513318e-01 -8.95176888e-01 1.04241621e+00
-5.77898026e-01 6.86215758e-01 -6.89272434e-02 -6.09366357e-01
1.41417366e-02 -9.14459117e-03 7.33629644e-01 1.44808030e+00
2.34338403e-01 -3.83300006e-01 -3.30761850e-01 9.80129540e-01
-3.12121779e-01 6.24367416e-01 -2.55925149e-01 -2.17636839e-01
5.61922669e-01 1.34813380e+00 -9.37204778e-01 -4.11178738e-01
-2.56369501e-01 9.53205049e-01 1.95062190e-01 5.85754775e-02
-4.91036475e-01 -5.01810849e-01 5.70589900e-02 4.36344475e-01
4.12596911e-01 -7.44614825e-02 -4.97405469e-01 -1.21842194e+00
1.87344238e-01 -1.17065752e+00 4.59043592e-01 -4.86252159e-01
-1.26960945e+00 6.69015586e-01 -1.24278605e-01 -1.23056126e+00
-3.89770240e-01 -6.44654453e-01 -4.90133345e-01 7.47478068e-01
-1.51600111e+00 -1.08555925e+00 -2.64618486e-01 1.05849870e-01
2.77106851e-01 -4.43743169e-01 4.78614748e-01 1.30354077e-01
-5.82163155e-01 9.15372372e-01 8.40134248e-02 4.45559561e-01
1.03363276e+00 -1.33612990e+00 1.45095840e-01 9.56107438e-01
4.31886822e-01 6.69341981e-01 1.01159203e+00 -6.23658836e-01
-1.02473378e+00 -5.88271201e-01 1.51883447e+00 -7.30373859e-01
8.10274422e-01 -3.31371844e-01 -9.97844279e-01 2.60309994e-01
1.90550476e-01 -5.09357274e-01 7.46885002e-01 3.46799523e-01
-7.66157568e-01 1.98415861e-01 -9.00983810e-01 3.64268810e-01
4.78156358e-01 -3.70793551e-01 -8.38880539e-01 1.04158774e-01
5.50823472e-02 -1.40014187e-01 -6.70314789e-01 3.31259891e-03
6.77309155e-01 -8.81764829e-01 5.36102235e-01 -3.91115457e-01
6.55492663e-01 -3.97570789e-01 3.64441603e-01 -8.43146384e-01
-1.96529627e-02 -7.49421954e-01 -1.28940985e-01 1.53318059e+00
5.91378391e-01 -5.07042527e-01 4.44055706e-01 4.63894039e-01
2.44368210e-01 -4.39939290e-01 -6.83350027e-01 -9.70627785e-01
8.67688563e-03 -1.27380475e-01 4.90705967e-01 9.96603370e-01
1.06524006e-01 5.40313244e-01 -2.52881944e-01 -1.46928504e-01
4.01272207e-01 2.52464056e-01 1.01283157e+00 -1.84681487e+00
-2.30679139e-01 -9.78396416e-01 -4.82018679e-01 -6.01929963e-01
2.25423068e-01 -7.43266165e-01 -1.82825729e-01 -1.32987487e+00
4.44415152e-01 -4.30542141e-01 -4.56953831e-02 3.09328258e-01
-2.73407191e-01 1.92103848e-01 1.58388853e-01 6.42851174e-01
-5.79749942e-01 1.47834182e-01 6.44415379e-01 -1.33312210e-01
-2.22558573e-01 3.06480043e-02 -7.40111351e-01 6.62005305e-01
6.48929000e-01 -4.64896828e-01 -1.70090441e-02 4.78561642e-03
5.87440014e-01 -4.31002796e-01 4.26703393e-01 -7.63638198e-01
2.59728700e-01 -4.04460216e-03 1.01428859e-01 -3.69319856e-01
-1.52230144e-01 -3.74742448e-01 -5.15356250e-02 3.83569956e-01
-5.28235376e-01 2.13032007e-01 -1.96419641e-01 5.14859796e-01
-2.74633318e-01 -7.97181964e-01 5.94082654e-01 5.95022738e-02
-3.50426048e-01 -1.69588923e-01 -5.05378425e-01 1.86244577e-01
8.56349111e-01 -2.32187718e-01 -4.57865983e-01 -3.60543251e-01
-2.49999359e-01 -8.91271457e-02 6.76440656e-01 3.93265665e-01
2.00181291e-01 -1.03275883e+00 -8.64665568e-01 -1.18645363e-01
3.12886715e-01 -6.40282631e-01 -3.66420805e-01 8.72242928e-01
-6.77996337e-01 3.84403795e-01 -9.63008925e-02 -2.96283513e-01
-1.57311916e+00 5.95404148e-01 3.57891084e-03 -5.62200844e-01
-3.82610321e-01 4.22851503e-01 -3.44657123e-01 1.70536973e-02
2.34301016e-01 -1.75087318e-01 -7.53023028e-01 5.94112098e-01
7.02046275e-01 4.17056620e-01 2.43632331e-01 -8.39799881e-01
-2.32182473e-01 3.23666185e-01 -3.05758327e-01 -4.56895381e-01
1.04560935e+00 5.59967346e-02 -2.33940743e-02 6.40821874e-01
8.62965167e-01 5.40132165e-01 -6.59771919e-01 -3.96425426e-01
4.74964887e-01 -2.99334526e-01 -5.79271205e-02 -8.58721137e-01
-6.61884367e-01 9.69393194e-01 1.36633828e-01 4.89155114e-01
6.72198951e-01 6.59809709e-02 5.42510033e-01 2.87358910e-01
2.18119308e-01 -1.39490533e+00 2.42317542e-01 4.73448426e-01
9.06625628e-01 -1.22099841e+00 2.34841421e-01 -4.06538278e-01
-6.62750781e-01 1.37175286e+00 1.75124064e-01 1.68979436e-01
4.08617765e-01 6.45216331e-02 2.11122662e-01 -1.29856795e-01
-3.86632502e-01 1.01174049e-01 7.16650426e-01 1.57980666e-01
4.80832368e-01 -1.83997471e-02 -1.01060486e+00 9.16578174e-01
-3.21507931e-01 -1.55333236e-01 7.56635249e-01 7.16829717e-01
-3.61643404e-01 -1.47913516e+00 -3.64437073e-01 5.34651637e-01
-6.75414860e-01 -4.52923588e-02 -9.94395792e-01 6.75857127e-01
-5.86294718e-02 7.01313496e-01 -1.41269043e-01 -1.97569847e-01
7.06265271e-02 1.14161424e-01 4.43581969e-01 -5.08827746e-01
-9.92216229e-01 -7.10251555e-02 2.46533707e-01 8.02131072e-02
-4.41425860e-01 -1.11041951e+00 -7.70686209e-01 -4.65496689e-01
-4.65996593e-01 2.77438819e-01 6.53729439e-01 1.10736012e+00
2.32242629e-01 2.50639375e-02 4.71579224e-01 -3.99189532e-01
-6.50228083e-01 -1.09247065e+00 -4.94396210e-01 5.54803073e-01
-1.62459072e-02 -7.66211927e-01 -5.42797029e-01 2.05569297e-01] | [9.556314468383789, 10.589239120483398] |
88bc4ebb-40f6-4341-93ed-e0a06188a5fe | unifying-identification-and-context-learning | 1806.03084 | null | http://arxiv.org/abs/1806.03084v1 | http://arxiv.org/pdf/1806.03084v1.pdf | Unifying Identification and Context Learning for Person Recognition | Despite the great success of face recognition techniques, recognizing persons
under unconstrained settings remains challenging. Issues like profile views,
unfavorable lighting, and occlusions can cause substantial difficulties.
Previous works have attempted to tackle this problem by exploiting the context,
e.g. clothes and social relations. While showing promising improvement, they
are usually limited in two important aspects, relying on simple heuristics to
combine different cues and separating the construction of context from people
identities. In this work, we aim to move beyond such limitations and propose a
new framework to leverage context for person recognition. In particular, we
propose a Region Attention Network, which is learned to adaptively combine
visual cues with instance-dependent weights. We also develop a unified
formulation, where the social contexts are learned along with the reasoning of
people identities. These models substantially improve the robustness when
working with the complex contextual relations in unconstrained environments. On
two large datasets, PIPA and Cast In Movies (CIM), a new dataset proposed in
this work, our method consistently achieves state-of-the-art performance under
multiple evaluation policies. | ['Yu Xiong', 'Qingqiu Huang', 'Dahua Lin'] | 2018-06-08 | unifying-identification-and-context-learning-1 | http://openaccess.thecvf.com/content_cvpr_2018/html/Huang_Unifying_Identification_and_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Huang_Unifying_Identification_and_CVPR_2018_paper.pdf | cvpr-2018-6 | ['person-recognition'] | ['computer-vision'] | [ 1.86896235e-01 -3.83537829e-01 -1.38540208e-01 -6.40008271e-01
-3.28597963e-01 -3.59119564e-01 8.50967884e-01 -2.72752881e-01
-3.76951784e-01 6.15362704e-01 3.57142031e-01 7.23309293e-02
-1.34378701e-01 -4.98482853e-01 -5.40348232e-01 -6.10232174e-01
2.24015176e-01 1.99948594e-01 7.22544044e-02 -7.37009645e-02
1.78587452e-01 5.63495696e-01 -1.73967171e+00 1.86644912e-01
7.44135916e-01 1.00483131e+00 -1.12543553e-01 3.46134573e-01
-8.10626596e-02 7.36493766e-01 -5.15931070e-01 -7.87068367e-01
2.97041029e-01 -3.43880594e-01 -5.81578553e-01 4.86731142e-01
8.89603317e-01 -3.99005771e-01 -3.52646947e-01 8.48694265e-01
6.46937430e-01 1.97105199e-01 4.82502788e-01 -1.25608540e+00
-9.39642847e-01 1.92883119e-01 -7.33478487e-01 -2.26916559e-02
5.77648580e-01 1.49155229e-01 9.54699278e-01 -1.00685430e+00
5.33622980e-01 1.40324616e+00 7.13389635e-01 7.64727056e-01
-1.16472888e+00 -5.94285131e-01 8.48110914e-01 5.09507596e-01
-1.44767213e+00 -7.87193835e-01 8.52575541e-01 -3.61454248e-01
6.41778231e-01 3.07910234e-01 4.58825648e-01 1.54301775e+00
-4.14225638e-01 7.28551507e-01 1.26240265e+00 -5.19088328e-01
7.89170712e-02 2.15331063e-01 2.35699892e-01 6.72710419e-01
1.51965201e-01 -1.70833856e-01 -6.07494950e-01 -2.55528927e-01
7.20170856e-01 3.50072861e-01 -3.29937518e-01 -3.99312437e-01
-8.71872067e-01 5.76235235e-01 3.16864371e-01 1.21602468e-01
-2.40709901e-01 -1.75076127e-01 8.04013088e-02 -2.08897591e-02
4.95549649e-01 1.81910425e-01 -2.47379869e-01 1.07174292e-01
-6.92428470e-01 2.32672840e-01 8.18937421e-01 9.02391732e-01
5.34087598e-01 -3.16311866e-01 -3.93522680e-01 1.30635798e+00
4.02050436e-01 4.27393317e-01 2.13408377e-02 -9.19469357e-01
4.77340639e-01 5.08328080e-01 1.68264195e-01 -1.19052911e+00
-3.18423957e-01 -3.48794341e-01 -5.74535549e-01 3.53898257e-02
7.58470595e-01 -2.05555588e-01 -9.58571613e-01 2.01339459e+00
4.98638064e-01 4.74420518e-01 -2.11921722e-01 1.08735394e+00
7.25627184e-01 1.88346580e-01 2.43302792e-01 -2.80352235e-02
1.56550062e+00 -1.17272115e+00 -7.26380646e-01 -3.61514449e-01
-1.24570303e-01 -6.76832855e-01 9.70221400e-01 3.09417218e-01
-1.06373441e+00 -5.64177573e-01 -7.82249212e-01 -8.66304114e-02
-3.22726339e-01 2.12267280e-01 6.77942157e-01 7.47274697e-01
-1.16347420e+00 4.22423124e-01 -5.69530189e-01 -8.11762512e-01
5.26347458e-01 3.40615779e-01 -4.39067513e-01 -3.40362579e-01
-9.29233134e-01 6.87257826e-01 -3.39180946e-01 3.40934217e-01
-6.60478413e-01 -2.49866143e-01 -7.89993167e-01 4.71829437e-02
8.72128785e-01 -7.06614733e-01 8.72330308e-01 -1.08865678e+00
-1.45331097e+00 8.94154370e-01 -4.33157384e-01 -1.18993171e-01
5.72315514e-01 -4.70594049e-01 -3.99698168e-01 1.37377471e-01
-9.37789232e-02 4.07531977e-01 8.52595627e-01 -1.52213144e+00
-4.57988411e-01 -6.55133128e-01 4.13498640e-01 2.14209452e-01
-7.73776531e-01 4.47743088e-01 -1.18929517e+00 -6.89894199e-01
-1.33372620e-01 -1.05241048e+00 -2.04756409e-01 2.23093972e-01
-2.96905190e-01 -2.51865506e-01 6.24999404e-01 -6.18459642e-01
1.04615700e+00 -1.93481898e+00 2.63721347e-01 6.50710613e-02
1.55462012e-01 3.47421557e-01 -8.19680169e-02 1.98438853e-01
2.09975258e-01 4.26772423e-02 -4.74631786e-02 -8.86225641e-01
7.03009069e-02 2.77941674e-01 -1.52093068e-01 3.35652083e-01
3.95937979e-01 6.73471272e-01 -7.90826499e-01 -4.79960501e-01
-3.48887779e-02 8.51636291e-01 -6.33470118e-01 3.27272683e-01
5.02944365e-02 5.75531244e-01 -4.40094113e-01 9.67137396e-01
6.87712133e-01 -2.71732658e-01 5.43420494e-01 -1.23532608e-01
1.71160802e-01 -1.29119053e-01 -1.30575395e+00 1.45130050e+00
-2.96080649e-01 3.85474771e-01 3.73587310e-01 -8.43170822e-01
7.11852610e-01 2.65110284e-01 3.62786859e-01 -4.65834945e-01
1.29779458e-01 -2.25223988e-01 -2.01097384e-01 -7.30123878e-01
4.81956333e-01 5.77222258e-02 1.91217750e-01 1.60319909e-01
-3.71403433e-02 6.27352834e-01 5.83574772e-02 3.77264363e-03
8.14887583e-01 4.76416588e-01 1.38403401e-01 -1.22265599e-03
6.16874516e-01 -6.92367792e-01 9.10014689e-01 8.19230616e-01
-5.02946496e-01 8.16017032e-01 5.84557831e-01 -4.56652164e-01
-6.44282997e-01 -8.59147012e-01 -3.92863527e-02 1.38679516e+00
2.57028282e-01 -4.59324032e-01 -7.03138411e-01 -9.12376821e-01
3.52875032e-02 1.45639956e-01 -7.75286078e-01 1.67425483e-01
-7.26645827e-01 -7.87624478e-01 2.52220511e-01 7.76401520e-01
4.62324768e-01 -8.57836902e-01 -1.45305008e-01 3.09807751e-02
-1.18535571e-01 -1.53072965e+00 -5.69653451e-01 -4.89867449e-01
-4.73171860e-01 -1.11564898e+00 -7.90455103e-01 -6.11809611e-01
7.94210553e-01 5.10888159e-01 1.17823100e+00 3.15393835e-01
-2.46974349e-01 5.70442438e-01 -2.13957891e-01 -2.78412461e-01
3.10334921e-01 -6.25749677e-02 1.58699617e-01 6.66671276e-01
4.91078317e-01 -6.90018773e-01 -7.10130274e-01 6.56091213e-01
-6.40765965e-01 -5.89515716e-02 3.91927630e-01 8.61747205e-01
3.73892665e-01 -3.40386689e-01 5.82192123e-01 -1.00894213e+00
3.69441390e-01 -5.19413412e-01 -2.82824963e-01 5.90756357e-01
-3.13686490e-01 -2.94267952e-01 5.64020038e-01 -5.11072397e-01
-1.23270321e+00 9.03659407e-03 -4.64124270e-02 -5.90983689e-01
-4.88825470e-01 5.79196140e-02 -5.23385108e-01 -1.93213657e-01
3.05210829e-01 -1.09636687e-01 -2.47562632e-01 -5.77345431e-01
2.05740005e-01 6.10650539e-01 4.50589210e-01 -1.02834737e+00
6.64484441e-01 6.24515951e-01 -1.25110000e-01 -7.72741973e-01
-1.06483424e+00 -4.17482316e-01 -6.78038955e-01 -2.77659208e-01
7.87132502e-01 -9.76581812e-01 -7.60830164e-01 4.26171035e-01
-9.19131041e-01 -1.02895260e-01 2.08588243e-01 2.24251539e-01
-1.10465713e-01 5.54729998e-01 -6.16903186e-01 -1.19571304e+00
-3.08258031e-02 -1.15454519e+00 9.81204808e-01 4.87294644e-01
1.14265606e-02 -8.58859956e-01 -3.25509101e-01 6.92156792e-01
4.64490116e-01 4.31668192e-01 3.66054982e-01 -4.02599663e-01
-6.56258225e-01 -2.26592589e-02 -4.79032069e-01 1.82285681e-01
3.37861240e-01 -4.18405747e-04 -1.29423344e+00 -2.56190389e-01
-1.49552658e-01 -3.71952474e-01 9.97374058e-01 4.54096757e-02
1.35863316e+00 -2.78009236e-01 -4.11557704e-01 7.94699371e-01
1.12071812e+00 1.17468476e-01 4.34615225e-01 6.25234321e-02
9.58929002e-01 9.21824872e-01 3.46983343e-01 4.78847563e-01
6.75509095e-01 1.02716076e+00 2.26733968e-01 -7.12814480e-02
-2.43827432e-01 -4.67347614e-02 1.82980493e-01 4.22474772e-01
-5.70842206e-01 -2.05183730e-01 -7.95393527e-01 4.77591217e-01
-2.04166269e+00 -1.07068121e+00 3.16724300e-01 2.06893015e+00
6.86670184e-01 -9.10400227e-02 3.04153830e-01 -1.79298371e-01
7.37245500e-01 3.31843615e-01 -4.78424698e-01 9.73830670e-02
-1.99949831e-01 -4.14230153e-02 4.26277258e-02 4.88353848e-01
-1.28239346e+00 8.55184197e-01 6.06348038e+00 3.51154417e-01
-9.89848316e-01 3.66772339e-02 6.82766199e-01 -3.39989245e-01
-3.69183868e-02 -2.32369885e-01 -1.07384408e+00 5.27520776e-01
4.91490990e-01 5.22244692e-01 6.25755847e-01 6.47190154e-01
-6.47253096e-02 1.39260471e-01 -1.14156055e+00 1.07972002e+00
5.96343935e-01 -6.74903929e-01 -1.73768520e-01 1.55204698e-01
7.12444901e-01 -3.31581444e-01 2.19859496e-01 3.57032239e-01
8.00648034e-02 -1.01549077e+00 6.27311647e-01 6.89289808e-01
6.11336648e-01 -5.22784948e-01 5.05349219e-01 -3.27203013e-02
-1.12228322e+00 -2.87612259e-01 -2.94149965e-01 -1.72965407e-01
4.95890118e-02 3.01845312e-01 -4.13200825e-01 6.02054596e-01
7.03620970e-01 7.28269637e-01 -7.15925753e-01 8.82426441e-01
-2.76297808e-01 5.47478378e-01 -2.48904660e-01 2.08710492e-01
-2.35314947e-02 -2.08705142e-01 2.81371623e-01 1.20858622e+00
5.53909615e-02 2.29561135e-01 5.21999419e-01 6.46002293e-01
-2.06477314e-01 1.26592249e-01 -4.83081758e-01 2.16385633e-01
3.09076607e-01 1.41508973e+00 -6.26583934e-01 -2.26528853e-01
-8.46616268e-01 9.21960533e-01 7.63393521e-01 6.47327363e-01
-9.35764611e-01 8.20985809e-02 9.18592513e-01 1.10831894e-01
3.84151787e-01 -2.89079875e-01 -1.16922788e-01 -1.54327333e+00
4.15600479e-01 -9.41435575e-01 4.89012033e-01 -4.22052920e-01
-1.68069875e+00 7.03937769e-01 -4.44732383e-02 -8.67008924e-01
4.38894741e-02 -7.87464201e-01 -5.79794466e-01 7.83817291e-01
-1.74807274e+00 -1.44884658e+00 -5.25664389e-01 7.43236899e-01
4.76913452e-01 -1.77331790e-01 6.90736115e-01 5.87658823e-01
-1.01558745e+00 9.11565244e-01 -2.27021471e-01 3.66377085e-01
1.09673727e+00 -1.09708333e+00 2.27928355e-01 8.82782757e-01
1.66160330e-01 9.58745599e-01 4.16592360e-01 -4.64088082e-01
-1.38716781e+00 -8.23313415e-01 7.21240878e-01 -5.97419620e-01
4.27143902e-01 -7.31420517e-01 -9.32286561e-01 6.42565191e-01
1.95485488e-01 3.48340780e-01 9.25490677e-01 6.96733654e-01
-7.26845324e-01 -1.74682647e-01 -1.01779389e+00 8.74796331e-01
1.57384408e+00 -5.68141878e-01 -3.85590017e-01 2.24880874e-01
3.22159946e-01 -3.24459761e-01 -7.38886058e-01 3.51932555e-01
8.61733079e-01 -1.11066568e+00 1.13713694e+00 -6.94409311e-01
2.75440335e-01 -3.10793012e-01 -2.63358653e-01 -9.26226974e-01
-3.33594203e-01 -6.24025166e-01 -3.36530149e-01 1.65986538e+00
4.44352580e-03 -7.68930972e-01 7.90851176e-01 1.09316039e+00
1.39033824e-01 -8.98528934e-01 -7.14385986e-01 -5.86900353e-01
-2.79593319e-01 -2.48032182e-01 8.17817271e-01 1.04766405e+00
-2.33532757e-01 3.73414934e-01 -8.56035113e-01 2.60188550e-01
6.38423502e-01 1.57882378e-01 8.78806889e-01 -1.24432886e+00
-4.60540414e-01 -4.24906313e-01 -2.58371890e-01 -9.50820982e-01
2.89416850e-01 -4.14620906e-01 -1.23400584e-01 -1.27652800e+00
4.42757398e-01 -5.69668114e-01 -5.78630805e-01 5.89434981e-01
-7.20103681e-01 3.28607470e-01 5.91397882e-01 1.58823520e-01
-8.65376174e-01 4.74438787e-01 1.08666563e+00 -1.63195938e-01
2.49215290e-02 -4.27650176e-02 -9.84812915e-01 7.93103755e-01
7.36415088e-01 -3.32839675e-02 -3.26167196e-01 -7.84871697e-01
-8.16752836e-02 -1.41937196e-01 5.28299451e-01 -9.57142413e-01
2.67061383e-01 -2.83576548e-01 7.16177821e-01 -1.36161909e-01
7.64370441e-01 -9.16776299e-01 -8.34905431e-02 -1.45493940e-01
-2.84403592e-01 -4.49538603e-02 2.92927660e-02 7.34277844e-01
-2.80904397e-02 7.29480907e-02 5.68780780e-01 -1.34101227e-01
-6.19661927e-01 4.82916534e-01 1.60638586e-01 -2.76340358e-02
8.86606574e-01 -1.77239567e-01 -3.67418230e-01 -4.14891660e-01
-8.31825674e-01 3.74101222e-01 6.42217517e-01 6.72721624e-01
4.22819138e-01 -1.39317799e+00 -7.12459922e-01 2.43599296e-01
1.81991607e-01 -2.95224667e-01 4.36984420e-01 7.42577016e-01
5.26228100e-02 1.54323846e-01 -1.03802465e-01 -4.80034322e-01
-1.49016821e+00 6.91298962e-01 2.76508987e-01 -2.01508954e-01
-4.28982407e-01 7.78340340e-01 4.29645568e-01 -2.59571046e-01
5.97586811e-01 2.69152373e-02 -4.80527550e-01 1.06879570e-01
9.07241583e-01 2.14653075e-01 -1.82666898e-01 -9.35433030e-01
-4.27753419e-01 8.81327212e-01 -2.08813339e-01 5.48479334e-02
1.33067799e+00 -2.76964158e-01 3.60485129e-02 1.19861081e-01
8.63465607e-01 2.75919527e-01 -1.64037466e+00 -5.45052052e-01
2.62407400e-02 -8.69157434e-01 -3.63880426e-01 -8.61636579e-01
-1.30941868e+00 8.16460311e-01 5.46002030e-01 1.36679545e-01
1.14758837e+00 -6.59591854e-02 6.72858238e-01 2.62500733e-01
4.84264284e-01 -1.00153279e+00 1.09219208e-01 3.69705260e-01
7.73800015e-01 -1.51630831e+00 6.83417171e-02 -6.09978259e-01
-6.31248772e-01 9.58890676e-01 8.94629359e-01 1.74953267e-01
4.85148668e-01 8.62952620e-02 1.16721250e-01 -4.85661067e-02
-7.03604639e-01 -4.30741698e-01 4.37810630e-01 6.50596321e-01
4.92091119e-01 -1.47681637e-02 -7.31799006e-02 8.00632119e-01
2.77693689e-01 -1.03937067e-01 9.72655043e-02 9.49995458e-01
-1.04730412e-01 -1.23584151e+00 -6.44567788e-01 1.84197247e-01
-8.15474391e-01 1.74076349e-01 -5.89062274e-01 6.03086233e-01
3.81865829e-01 1.18942475e+00 -1.22871131e-01 -2.60019481e-01
5.25454700e-01 1.73214778e-01 5.76758325e-01 -4.87047911e-01
-6.99698985e-01 6.37622550e-02 1.59604311e-01 -6.39581442e-01
-6.63708091e-01 -8.76524985e-01 -5.92933536e-01 -1.82180747e-01
-1.91428885e-01 -2.05102324e-01 4.39135313e-01 9.32296276e-01
4.80702609e-01 3.32981020e-01 6.27673268e-01 -9.49163020e-01
-3.33919972e-01 -8.58375728e-01 -3.82854313e-01 8.88178349e-01
1.96179718e-01 -9.10963714e-01 -7.69191459e-02 5.49618565e-02] | [14.581668853759766, 0.9546076059341431] |
d7e49f4f-44cc-427e-8b13-3c46b459ca0d | diffclip-leveraging-stable-diffusion-for | 2305.15957 | null | https://arxiv.org/abs/2305.15957v2 | https://arxiv.org/pdf/2305.15957v2.pdf | DiffCLIP: Leveraging Stable Diffusion for Language Grounded 3D Classification | Large pre-trained models have had a significant impact on computer vision by enabling multi-modal learning, where the CLIP model has achieved impressive results in image classification, object detection, and semantic segmentation. However, the model's performance on 3D point cloud processing tasks is limited due to the domain gap between depth maps from 3D projection and training images of CLIP. This paper proposes DiffCLIP, a new pre-training framework that incorporates stable diffusion with ControlNet to minimize the domain gap in the visual branch. Additionally, a style-prompt generation module is introduced for few-shot tasks in the textual branch. Extensive experiments on the ModelNet10, ModelNet40, and ScanObjectNN datasets show that DiffCLIP has strong abilities for 3D understanding. By using stable diffusion and style-prompt generation, DiffCLIP achieves an accuracy of 43.2\% for zero-shot classification on OBJ\_BG of ScanObjectNN, which is state-of-the-art performance, and an accuracy of 80.6\% for zero-shot classification on ModelNet10, which is comparable to state-of-the-art performance. | ['Xinxiao wu', 'Harry Zhang', 'Linqian Fan', 'Zilin Zhu', 'Sitian Shen'] | 2023-05-25 | null | null | null | null | ['3d-classification'] | ['computer-vision'] | [-3.59332298e-05 7.63817653e-02 -2.27044687e-01 -3.89516532e-01
-8.46160829e-01 -2.86832213e-01 4.14218426e-01 -2.33142093e-01
-3.12084824e-01 -3.58520299e-02 -4.53841388e-01 -1.84886396e-01
2.90088564e-01 -8.17798197e-01 -6.94133222e-01 -3.16921622e-01
4.07643944e-01 7.11408496e-01 1.01129961e+00 -8.79357830e-02
2.88461298e-01 5.54388165e-01 -1.83497596e+00 4.43954319e-01
7.83407152e-01 1.34532571e+00 5.48976541e-01 5.78864276e-01
-6.14441574e-01 5.78541338e-01 -5.24883747e-01 -2.38058433e-01
4.15781170e-01 -7.77727813e-02 -5.90632379e-01 2.91348755e-01
9.10915792e-01 -4.97257113e-01 -1.65873751e-01 9.79857147e-01
5.84324121e-01 -2.13940106e-02 8.82981181e-01 -1.33337307e+00
-4.74534065e-01 8.79776925e-02 -7.40349054e-01 2.15991631e-01
1.19190335e-01 3.60712916e-01 7.77300358e-01 -1.02289343e+00
9.45931256e-01 1.26047432e+00 6.94837868e-01 7.33446479e-01
-1.06835258e+00 -9.57307220e-01 5.10067157e-02 3.57171983e-01
-1.23391318e+00 -6.20791763e-02 6.98569059e-01 -6.32924557e-01
1.30845523e+00 -8.68753120e-02 9.29657876e-01 9.30312395e-01
3.43302935e-01 9.81444418e-01 9.45334911e-01 -3.09871078e-01
2.80738413e-01 1.26236439e-01 1.33927539e-01 7.82418847e-01
-3.10931891e-01 9.68685076e-02 -5.40351093e-01 2.86218703e-01
9.50291038e-01 -1.15506016e-01 5.91387637e-02 -4.88190204e-01
-7.74574876e-01 7.96387672e-01 5.81927598e-01 5.43016158e-02
-1.62524104e-01 4.11239266e-02 2.09633350e-01 5.75166848e-03
8.57158720e-01 2.70959437e-01 -3.24371308e-01 3.93407047e-02
-9.54764128e-01 2.95640409e-01 5.98615885e-01 1.42204654e+00
6.68207049e-01 1.73735768e-01 -1.30265966e-01 1.04543293e+00
1.52150184e-01 7.49841630e-01 1.81533307e-01 -1.30022955e+00
3.51287782e-01 7.96156883e-01 -2.99408942e-01 -6.80507064e-01
-3.39778572e-01 -2.92020887e-01 -4.81241822e-01 6.09634876e-01
2.21015826e-01 2.66617477e-01 -1.53524053e+00 1.08645868e+00
3.54509205e-01 4.05387580e-01 -9.54481214e-02 8.15897703e-01
1.34956324e+00 7.92588532e-01 1.58830106e-01 1.51214942e-01
1.01772881e+00 -1.17581844e+00 -2.30740249e-01 -5.15632212e-01
3.32264870e-01 -7.34463096e-01 1.20602608e+00 3.28470737e-01
-1.13854921e+00 -8.36521983e-01 -1.05856025e+00 -2.41774380e-01
-3.03391993e-01 -1.16253451e-01 5.54898620e-01 4.38838869e-01
-9.04408276e-01 3.93436462e-01 -6.82102263e-01 -4.97862339e-01
9.01006281e-01 -1.01983085e-01 -1.73521712e-01 -4.50111091e-01
-5.66704929e-01 8.52639079e-01 5.44251025e-01 -6.50847077e-01
-1.19537187e+00 -1.38031590e+00 -8.84777248e-01 -2.10940853e-01
3.28572512e-01 -6.56171024e-01 1.43990612e+00 -4.60703671e-01
-1.23351359e+00 1.27700782e+00 5.49777634e-02 -3.38512868e-01
6.43944263e-01 8.89504515e-03 -2.21278280e-01 5.26144207e-01
4.16572899e-01 1.37243128e+00 1.05909503e+00 -1.23198628e+00
-9.79616463e-01 -4.39736247e-01 3.18125300e-02 2.42660522e-01
-7.05290437e-02 -9.68696550e-02 -9.89942014e-01 -3.20360363e-01
3.60620558e-01 -8.61346781e-01 -3.74035835e-01 3.69853526e-01
-1.92013741e-01 -5.10092437e-01 1.32307887e+00 -3.14058602e-01
7.42308140e-01 -2.11511588e+00 -2.03258470e-01 -8.87436643e-02
5.26443496e-02 4.47789997e-01 -3.62160712e-01 -1.82618946e-01
1.25645295e-01 7.87598193e-02 -1.59411177e-01 -4.77211326e-01
-3.74556899e-01 3.44160557e-01 -1.59088269e-01 2.10564792e-01
2.76130050e-01 8.83808136e-01 -6.40687048e-01 -7.53305674e-01
9.03175414e-01 1.97370216e-01 -6.23917460e-01 1.61928236e-01
-8.08349490e-01 1.63801670e-01 -3.11725467e-01 1.02956831e+00
7.57481337e-01 -3.68328214e-01 -5.05070150e-01 -2.24556208e-01
-1.81160465e-01 -3.71540010e-01 -1.01271510e+00 2.11225581e+00
-3.77560467e-01 5.49752116e-01 -1.86407343e-01 -5.68734229e-01
9.16657805e-01 2.27235649e-02 5.58295429e-01 -8.14527690e-01
1.91285148e-01 3.46582569e-02 -3.02029908e-01 -5.64163148e-01
4.12948012e-01 8.38649049e-02 1.20511577e-01 2.81888604e-01
2.56525874e-01 -9.94881868e-01 1.94068074e-01 2.97592670e-01
8.23504448e-01 1.86407700e-01 -1.26992285e-01 -1.22548074e-01
1.36639833e-01 4.59009826e-01 5.41419327e-01 7.96891510e-01
-2.21348956e-01 1.11219442e+00 2.34227911e-01 -2.73104399e-01
-1.19524240e+00 -1.17726874e+00 -3.07045788e-01 8.95503759e-01
4.93350446e-01 -2.91850924e-01 -7.41782486e-01 -5.80827415e-01
5.34098446e-02 9.51791048e-01 -4.51547772e-01 -3.51526700e-02
-2.10812807e-01 -4.40214515e-01 4.41137493e-01 8.47358644e-01
8.38579237e-01 -9.79349792e-01 -7.87382662e-01 6.67661279e-02
-8.47321078e-02 -1.52245569e+00 -2.32554987e-01 8.01889449e-02
-1.02008164e+00 -1.21544933e+00 -6.87537313e-01 -8.01760197e-01
4.82916057e-01 6.31254852e-01 1.13421082e+00 -1.59333542e-01
-6.71999276e-01 4.58941787e-01 -3.04356486e-01 -8.67022336e-01
-1.95662245e-01 6.61740676e-02 -3.97697687e-01 -5.72523773e-01
3.95824999e-01 -3.89971614e-01 -5.21883130e-01 6.12711906e-01
-9.07569110e-01 3.99846017e-01 4.71665591e-01 4.16418970e-01
9.27280903e-01 -3.29977959e-01 3.05598319e-01 -6.46575749e-01
-7.82078039e-03 -1.99600592e-01 -5.45851827e-01 6.47564456e-02
-5.73789597e-01 -4.48885649e-01 3.30380090e-02 -4.89642411e-01
-1.23719692e+00 2.20134422e-01 -4.27302271e-01 -1.03454280e+00
-4.57836837e-01 -1.28293335e-01 -4.13220078e-02 -1.79877311e-01
8.20418894e-01 7.88505469e-03 9.17746350e-02 -4.45924610e-01
4.38375086e-01 6.69726729e-01 4.67128336e-01 -2.33783230e-01
5.11200964e-01 7.41108954e-01 -9.16189551e-02 -1.27913821e+00
-1.18933129e+00 -8.22033763e-01 -7.73298204e-01 -4.59601104e-01
1.35021687e+00 -1.02184820e+00 -1.78795949e-01 7.99057722e-01
-1.30635083e+00 -5.80352068e-01 -3.31239343e-01 2.30596736e-01
-7.40280092e-01 -7.21540079e-02 -6.11371875e-01 -4.29155707e-01
-3.69774461e-01 -1.12262774e+00 1.17148435e+00 3.83142352e-01
-3.46852769e-03 -7.60203719e-01 -3.52191865e-01 6.13487601e-01
1.28040612e-01 1.51409745e-01 1.13050902e+00 -3.99192125e-01
-7.84303665e-01 -1.62108138e-01 -6.70931101e-01 6.48343265e-01
-4.47767854e-01 -7.30035678e-02 -1.21018660e+00 9.02241543e-02
-1.91657215e-01 -4.69247758e-01 8.27452660e-01 6.95428848e-01
1.38620341e+00 5.88185310e-01 -5.79146683e-01 9.52749610e-01
1.49088788e+00 4.36116785e-01 7.20909953e-01 2.69449383e-01
8.60942781e-01 2.74287194e-01 8.10105979e-01 2.24194631e-01
4.38698500e-01 6.69794977e-01 6.62194192e-01 -1.38979539e-01
-6.36855781e-01 -2.34072253e-01 -9.40145627e-02 4.48875785e-01
-4.51795943e-02 -8.58549401e-03 -1.15686715e+00 5.04311025e-01
-1.63678730e+00 -8.34063172e-01 -3.28578085e-01 1.62158513e+00
4.72369254e-01 4.49630350e-01 -2.56954189e-02 -1.78422123e-01
5.60286939e-01 2.62674809e-01 -8.92358541e-01 -1.95729136e-01
-4.40459438e-02 2.19261825e-01 5.74116170e-01 1.95635125e-01
-1.13406849e+00 1.31473744e+00 6.38263655e+00 1.17810035e+00
-1.06461847e+00 2.95019239e-01 6.35936081e-01 -2.65130281e-01
3.91195677e-02 -3.74398053e-01 -1.21052742e+00 2.06394985e-01
4.62044448e-01 -8.19038153e-02 -1.74137294e-01 1.39723265e+00
9.43787396e-02 -3.73343021e-01 -1.03565228e+00 1.31380451e+00
4.63708997e-01 -1.60701418e+00 4.11180668e-02 8.80156178e-03
1.08177400e+00 4.89674449e-01 -9.08447728e-02 4.93847877e-01
5.60957566e-02 -8.60711813e-01 8.56027961e-01 3.91633570e-01
9.94328022e-01 -7.95540273e-01 4.28886354e-01 6.57255590e-01
-1.20378613e+00 -5.40309073e-03 -6.93854809e-01 1.80693984e-01
4.60386425e-01 5.68099022e-01 -8.51633608e-01 2.70754427e-01
1.04143310e+00 9.01538491e-01 -7.49147534e-01 1.17910028e+00
-2.90216982e-01 2.72181571e-01 -2.79279739e-01 1.23889834e-01
3.47612888e-01 -5.28988205e-02 6.13073587e-01 1.13759172e+00
2.29124859e-01 3.95671099e-01 4.73487020e-01 1.04166603e+00
3.73019697e-03 -1.08445920e-01 -6.57141805e-01 2.12980434e-01
2.80849010e-01 1.12408102e+00 -9.61112142e-01 -4.87781346e-01
-4.29361552e-01 8.83333623e-01 4.86835167e-02 1.43954977e-01
-9.08087134e-01 -4.61982876e-01 6.19409800e-01 2.58583844e-01
5.21112084e-01 -2.72655606e-01 -6.37575746e-01 -8.49176526e-01
-1.52370304e-01 -3.83662492e-01 1.80367231e-01 -1.06336427e+00
-1.24478185e+00 6.07408762e-01 3.28773052e-01 -1.28393245e+00
-1.93050504e-01 -8.62466156e-01 -6.39260650e-01 5.58074892e-01
-1.32073796e+00 -1.45299387e+00 -6.45134270e-01 6.38116241e-01
1.21499503e+00 -1.50660843e-01 6.33048475e-01 2.69780844e-01
-3.27977926e-01 2.55255908e-01 -2.58968502e-01 -8.01520720e-02
4.46742088e-01 -1.01864755e+00 5.15984058e-01 5.74663103e-01
-4.78389487e-02 -1.36163503e-01 3.96112293e-01 -7.25701392e-01
-1.04741263e+00 -1.25876713e+00 3.92553449e-01 -4.89164591e-01
3.80939066e-01 -2.76480973e-01 -9.67331827e-01 3.74081612e-01
2.41076406e-02 8.38504359e-02 3.17608654e-01 -2.12322116e-01
-3.51884186e-01 -2.49251630e-02 -1.26784861e+00 4.27631974e-01
1.39482248e+00 -3.88123602e-01 -6.93597436e-01 4.27858114e-01
9.24313188e-01 -8.12554419e-01 -7.29341149e-01 5.51971197e-01
4.79182035e-01 -1.13679171e+00 1.20462203e+00 -3.16302508e-01
7.32666790e-01 -1.98423192e-02 -1.67353749e-01 -1.24702775e+00
-1.10180795e-01 3.24677616e-01 -1.15706652e-01 1.07508743e+00
3.48757207e-01 -1.09210357e-01 1.13167751e+00 5.39471865e-01
-6.17796540e-01 -8.13542902e-01 -7.62628615e-01 -9.50212717e-01
8.96375775e-02 -1.10476625e+00 1.88516170e-01 6.76449239e-01
-6.72168136e-01 2.64310986e-01 4.96853366e-02 -1.22447006e-01
9.09591556e-01 1.44526809e-01 9.59883332e-01 -1.34061348e+00
-1.41683724e-02 -4.74337727e-01 -5.12742937e-01 -1.18126822e+00
2.32022986e-01 -8.85435998e-01 2.06566826e-02 -1.93004513e+00
1.38883978e-01 -5.69766641e-01 1.29452944e-01 3.65348786e-01
2.64524311e-01 6.14957094e-01 6.00006580e-01 1.51236981e-01
-5.44832408e-01 3.61439794e-01 1.58438647e+00 -3.58341247e-01
-3.69745970e-01 6.25052750e-02 -1.79907382e-01 1.10425973e+00
6.35092199e-01 -3.13471109e-01 -4.97679710e-01 -5.88871419e-01
-4.03886378e-01 -3.97183374e-02 4.25918072e-01 -1.30451083e+00
2.71110266e-01 -2.24280372e-01 5.48744798e-01 -1.32851827e+00
7.64273047e-01 -6.04623675e-01 -7.56361037e-02 3.52349013e-01
1.63957030e-02 -2.99924612e-01 3.93152177e-01 5.21472812e-01
1.38132907e-02 -2.65107274e-01 1.19894886e+00 -5.53311586e-01
-1.50899732e+00 5.87588370e-01 -1.09201096e-01 1.94655657e-01
1.42641938e+00 -6.04093850e-01 -3.54767561e-01 6.99885115e-02
-7.21079409e-01 3.65604490e-01 5.08073032e-01 7.97238231e-01
1.07153058e+00 -1.08890843e+00 -4.68984067e-01 5.21404207e-01
3.91480654e-01 6.04485333e-01 5.30328393e-01 6.05019927e-01
-6.70224309e-01 3.04541290e-01 -3.56787175e-01 -1.25704801e+00
-1.39498270e+00 1.69134095e-01 3.62736076e-01 1.57202736e-01
-9.80674446e-01 1.09437013e+00 2.22916350e-01 -3.10432523e-01
4.78083819e-01 -3.77261370e-01 1.12601496e-01 1.84152737e-01
5.29174387e-01 5.78162909e-01 1.08682789e-01 -3.81654054e-01
-1.98394880e-01 9.26794052e-01 -3.50108504e-01 -1.42533675e-01
1.22609460e+00 8.74182060e-02 2.05533683e-01 4.49735940e-01
1.16133976e+00 -7.72721052e-01 -1.47046614e+00 -7.94778466e-02
-3.68572384e-01 -6.54336810e-01 4.11812901e-01 -7.76163697e-01
-1.05927193e+00 1.11743665e+00 8.81137669e-01 -2.52932757e-01
7.19233692e-01 4.79874790e-01 8.56890619e-01 2.09620461e-01
6.86108947e-01 -1.29791009e+00 5.62867224e-01 8.64485860e-01
7.36289740e-01 -1.49315679e+00 -2.00701281e-01 -6.46137834e-01
-7.26622224e-01 9.27257717e-01 1.21707404e+00 -1.55144304e-01
7.51018524e-01 1.27227157e-01 2.22679004e-01 -3.00417334e-01
-5.34694672e-01 -3.06208909e-01 4.35171574e-01 9.25372303e-01
-1.49094477e-01 -1.82662487e-01 1.32577717e-01 3.20069611e-01
-8.27866867e-02 6.29663169e-02 3.03714633e-01 8.85765851e-01
-7.93443799e-01 -6.09030366e-01 -2.96121418e-01 6.03121758e-01
-4.88056019e-02 1.79245606e-01 -1.35109559e-01 9.16344941e-01
3.52118045e-01 8.46121669e-01 4.09557313e-01 -4.04815435e-01
6.18499219e-01 1.23185683e-02 5.87090433e-01 -9.98389423e-01
-2.11285979e-01 1.07771739e-01 -1.75148532e-01 -8.50850582e-01
-2.50599474e-01 -4.75517243e-01 -1.30481160e+00 -1.27972737e-01
-3.04418027e-01 -3.73678595e-01 8.93597424e-01 8.79254282e-01
1.98689401e-01 6.80673957e-01 2.60475934e-01 -1.27828705e+00
-3.34378690e-01 -9.19477344e-01 -6.44755244e-01 2.66875058e-01
-1.01381361e-01 -7.92669177e-01 -2.06048310e-01 9.00726765e-03] | [8.115486145019531, -3.056816339492798] |
f943e9db-da78-4c3e-a2ff-ac2fbeff40b1 | graph-structure-learning-with-variational | 2112.08903 | null | https://arxiv.org/abs/2112.08903v1 | https://arxiv.org/pdf/2112.08903v1.pdf | Graph Structure Learning with Variational Information Bottleneck | Graph Neural Networks (GNNs) have shown promising results on a broad spectrum of applications. Most empirical studies of GNNs directly take the observed graph as input, assuming the observed structure perfectly depicts the accurate and complete relations between nodes. However, graphs in the real world are inevitably noisy or incomplete, which could even exacerbate the quality of graph representations. In this work, we propose a novel Variational Information Bottleneck guided Graph Structure Learning framework, namely VIB-GSL, in the perspective of information theory. VIB-GSL advances the Information Bottleneck (IB) principle for graph structure learning, providing a more elegant and universal framework for mining underlying task-relevant relations. VIB-GSL learns an informative and compressive graph structure to distill the actionable information for specific downstream tasks. VIB-GSL deduces a variational approximation for irregular graph data to form a tractable IB objective function, which facilitates training stability. Extensive experimental results demonstrate that the superior effectiveness and robustness of VIB-GSL. | ['Philip S. Yu', 'Cheng Ji', 'Xingcheng Fu', 'Jia Wu', 'Hao Peng', 'JianXin Li', 'Qingyun Sun'] | 2021-12-16 | null | null | null | null | ['graph-structure-learning'] | ['graphs'] | [ 0.25698164 0.6176557 -0.423129 -0.06739556 -0.26276866 -0.24198711
0.2512225 0.23558508 0.2022323 0.621987 0.29472816 -0.48560205
-0.7837894 -1.0096892 -0.85222685 -0.8486445 -0.2867816 0.39566886
-0.1289176 -0.14327145 -0.21983816 0.24850719 -0.9650507 -0.08355199
0.88315886 0.8659638 0.5148341 0.41041642 -0.17018108 1.0509062
-0.18414319 -0.38443553 0.2632931 -0.5919458 -0.74770796 0.03624694
0.12727165 -0.05461852 -1.1492853 1.2226139 0.2401212 0.3023492
0.5790159 -1.2722257 -0.87873304 1.0829178 -0.5344826 0.23748823
0.07174447 -0.12262802 1.6222852 -0.48699036 0.61138904 1.2032043
0.53643644 0.5008997 -1.2928013 -0.48034474 0.4006845 0.2899924
-1.4007708 -0.1879662 1.1037502 -0.16795775 0.7209273 0.09086067
0.8340193 1.1061566 -0.10062034 1.0621591 0.36802545 -0.34499863
0.12929896 -0.29586634 0.1863354 1.0114092 0.7298789 0.12054189
-0.7638978 0.02722424 1.0160972 0.10023061 -0.46934915 -0.6125584
-0.96364045 0.91271305 1.0080327 0.2446 -0.3092202 0.21909454
0.44590577 0.48858336 0.37936485 0.27953783 -0.18545158 0.37655976
-0.5234525 -0.17157723 0.67378783 1.248755 0.82342726 0.19489071
-0.296413 0.62184376 0.52725005 0.38258377 -0.06233908 -0.80464554
0.7140479 0.87050885 -0.63042897 -1.4026544 -0.33940467 -0.77890193
-1.5504935 -0.5150973 0.29370508 0.03864726 -0.7613003 1.8371549
0.20592028 0.1481359 -0.01061163 0.82884955 1.0557044 0.69997424
-0.3595755 -0.2019761 0.7933819 -0.6354563 -0.7032323 -0.32200915
0.6415064 0.15436995 0.9904432 0.31967983 -0.88365006 -0.34038192
-1.0549613 -0.08803409 0.11186372 -0.35972744 1.0473098 0.33732733
-1.2390687 0.76735 -0.7558159 -0.10804038 0.7304616 0.22436188
-0.4108447 -0.4475515 -1.1528462 0.17684846 0.66003305 0.45878544
-0.9952451 -0.56336355 -1.3043408 0.379174 0.9333618 -0.78397757
0.80024034 -0.39707738 -1.1057183 0.46535978 -0.16043521 -0.45260814
0.14828303 0.09448988 -0.09354672 0.29875016 -0.02001633 0.20766565
0.77024376 -1.1645864 0.00677867 -0.44516993 0.1375382 0.1100005
-0.33418974 -0.84203273 -0.55705017 -0.56478715 0.55012983 -0.4648169
-0.40291974 -0.19225878 -0.797429 -0.42292202 0.5381241 -0.40254107
1.4763285 -1.9724157 0.40858543 0.46051362 0.9821807 0.1728277
-0.26175013 0.7118089 -0.14258048 0.17951216 -0.22831227 -0.258276
-0.02641804 0.6165063 -0.2955798 0.403263 0.03362641 1.3769999
-1.1664658 -0.43889657 0.03220018 0.37832892 -0.6630246 0.23127107
-0.34509015 0.59169763 -0.7920656 0.55403525 0.4261438 -0.979129
0.54252064 -0.06942504 0.6076754 0.28544882 -0.93658215 1.7977759
-0.24524702 0.7940301 0.12309515 -1.6828179 0.9935657 0.0856391
0.5411326 -0.7918272 0.08341851 -0.18946525 0.05270476 -0.4358497
0.05018518 0.01431635 0.03041668 0.36873037 0.22940172 0.20394295
0.2986212 0.89387494 1.2970694 -0.06922208 0.43307453 -0.38097823
0.32697394 -0.5901564 0.66171855 0.9914069 -0.1397498 0.38848686
0.9590886 -0.47640362 -0.84992665 -1.2230489 0.30370525 0.8226779
0.23922625 -0.6905897 -0.59962404 -0.73554265 -0.16133344 0.44253293
-0.47269574 -0.50870717 -0.41753343 -0.5882123 0.28854343 0.28067866
0.4503034 -0.993491 0.21196055 0.16759016 -0.36343914 -1.1378641
-0.5286115 0.04433951 -1.0014397 -1.2120123 -0.38042945 -0.82649577
0.8217044 0.5547301 1.3720697 0.5108724 -0.20659038 0.23345113
-0.40269715 -0.11534569 -0.3933119 0.22969583 -0.11435843 -0.03504733
0.08546019 -1.0676391 -0.54732716 -0.09850274 -0.8493878 0.20649625
0.736576 0.96566284 0.57097703 0.28502923 0.51234674 -1.175145
0.6861572 -0.56096286 -0.579485 0.3041303 -0.8162712 0.5303816
0.75285447 0.05968408 -0.68612367 -0.33751294 -0.02013629 -0.70341986
0.30896822 1.0693798 -0.41832703 0.15038529 0.5320909 0.3869309
0.04139971 -0.26016137 0.5057024 0.05217567 0.4158189 -0.6372101
0.7953644 0.37086004 0.47041422 -1.0235443 -1.1862559 -0.49168104
-0.5773949 -0.22761731 0.37564498 -0.73354614 -1.07575 0.1017794
-0.96366197 -0.5167075 -0.30049792 0.27244177 -0.35393476 0.7220501
-0.74499005 -0.86038256 -0.30346477 -0.824016 0.8651796 0.08432778
0.2495229 -1.3859007 -0.04456595 0.27077645 -0.12158652 0.2677513
1.1507579 -0.4247021 -1.1580355 0.02368119 -0.5265779 0.14637953
0.03429878 -0.38262078 -0.69591236 -0.46852612 -0.18951418 -0.31733853
1.2708049 0.67628676 1.2548037 -0.5261837 -0.4106678 0.93349
1.4585907 -0.20674399 0.5000648 -0.38124138 1.1983582 0.47187042
0.05545463 0.3730723 0.42327583 0.08197226 0.68394667 0.00619652
-0.07426272 -0.7580336 0.11190408 1.3873285 -0.02498321 -0.60902107
-0.7615637 0.39390376 -2.0169795 -0.8690183 -0.15353265 1.9594097
0.47820455 0.14328292 -0.03289301 0.18856275 0.8424902 0.5699424
-0.7993398 0.15858604 -0.08576484 0.00657922 0.39335304 0.5195635
-0.7434805 0.8091163 5.9888725 0.84053683 -0.62834275 -0.13835916
0.613688 0.06593908 -0.70351774 0.04627486 -0.4605044 0.0917874
0.7005874 -0.4275328 0.84933674 0.6657851 0.06172498 0.27262807
-1.1007237 1.202015 -0.1580174 -1.6386819 0.2725928 0.33194143
0.6588863 0.20584182 -0.02845519 0.3967875 0.5802874 -1.1515006
0.35691264 0.53106004 0.7998689 -0.7226862 0.38937387 0.5395273
-1.4991263 -0.04266401 -0.6626949 -0.29199156 0.01549795 0.8476862
-0.79798245 1.0489551 0.4495736 1.2211696 -0.34735632 0.60038066
-0.41157442 0.8390621 -0.19971791 -0.27427623 0.36977378 -0.6538426
0.6798161 0.709847 0.06508883 0.09474304 0.31711328 1.2024624
-0.41593155 -0.03363736 -1.0854279 -0.59723157 0.5191327 1.0732585
-0.7829277 0.01763705 -0.28989592 0.7018171 0.9279741 0.67365205
-0.4374264 -0.21053232 0.42948815 -0.08292563 0.34148827 -0.24412262
-0.03327378 -1.2305586 0.23307645 -0.7510434 0.7071576 -0.37595108
-1.4631122 0.4821939 -0.10862689 -0.87616134 -0.05957222 -0.6904697
-0.5941446 0.5941447 -1.2534075 -1.0799369 -0.29709357 0.6193203
0.1656652 -0.2129592 0.4944345 0.13318773 -0.8207152 0.6027049
0.2362779 0.43694046 -0.06273481 -1.2574633 0.56569946 0.951503
0.63361424 0.70649415 0.44017816 -0.69274795 -1.7459065 -1.1754225
0.39573586 -0.12850487 0.76420355 -0.59516484 -0.9780058 0.7468163
-0.20553555 0.18480442 0.5317724 0.47213677 -0.49963257 -0.13515753
-0.6078231 0.5326584 1.74027 -0.805226 -0.1641302 0.5327673
1.0673442 -0.19876698 -0.80400026 0.3288753 0.19288814 -1.0214115
0.96296275 -0.7316607 0.49461266 0.02463914 -0.16676982 -1.2951131
-0.43374902 -0.9565617 -0.6342371 0.8660073 0.36886552 -0.682594
1.0237285 0.11930099 -0.14997531 -1.0215139 -0.7478032 -0.7606337
-0.09700669 -0.5803015 0.46658218 0.8663912 -0.01275557 0.7830205
-0.44261733 0.250812 0.96136445 0.2309016 0.60325015 -1.4475298
-0.58314365 -0.54282856 -0.54649556 -1.4399648 0.39177862 -1.5321081
-0.11354181 -1.9069208 0.3722518 -0.34967467 -0.39313334 0.1879311
-0.38555655 -0.22425735 0.03126336 0.13998462 -0.7954463 0.860057
1.5662928 -0.30206355 -0.08320435 0.08638567 -1.0093837 0.40993163
0.5696085 -0.3085978 -1.1412507 -0.35797572 0.71035784 0.15740414
0.30311185 -0.5701249 0.25909233 -0.01497114 0.07045647 -0.45577106
0.0299936 -0.5175399 -0.02295497 0.3982139 -0.36002365 -0.4078868
-0.2126754 1.1976528 -0.13791925 -0.07016013 0.32236338 -0.24224621
-0.59220684 1.0595052 0.22294277 0.49559882 0.48950902 -0.11805343
-0.19521281 -0.82199335 -0.6476683 0.36627886 0.05700353 0.042213
0.89575404 -1.3090599 -0.62520105 0.40828076 0.06767961 0.6679167
0.34814858 0.65475583 -0.25786865 0.4495653 0.34444886 -0.65045595
-0.69867647 0.89236826 0.17059533 -0.5323524 -1.1875997 0.99518466
0.62482756 -0.38644934 0.28503773 -0.2656389 0.07339668 -0.19562687
0.39477727 0.14107543 -0.15107888 -0.33051035 0.02512637 0.03606054
-0.07960472 0.19882435 1.3626962 -0.25282094 -0.24168453 0.49140224
1.2074589 -0.33083373 -1.2804672 -0.7474212 0.32952267 -0.32368222
0.1600771 0.018012 -1.3201294 0.90003973 -0.37103435 0.61847454
1.1363274 0.42862508 0.82796735 0.8669295 0.47600374 -0.5892857
0.23885311 0.5362334 0.8607354 -1.1485407 0.07498162 -0.63367414
-0.28587615 0.92153996 0.3487707 -0.04562449 0.80147445 -0.09837507
-0.464323 -0.6300104 -0.7664158 -0.28876182 0.42464408 0.7544394
0.1427709 0.067313 0.29473382 0.58065176 -0.05310576 -0.5223912
0.30586898 0.48612782 -0.20201774 -0.7904687 0.21771872 0.6888374
-0.11321706 -0.3047435 -0.47047883 0.79847914 -0.46701518 0.8301688
-0.14679293 -0.36975217 -0.05171385 -0.32840562 0.67489827 -0.61298597
0.25848565 0.05441896 -0.07531435 -0.6519192 -0.2475063 -0.35611764
-1.1940604 -0.46682334 -0.37592566 0.21423256 0.08446791 1.0264359
0.44622862 0.8049851 0.65563935 -0.6399706 -0.44222263 -0.8421396
-0.8625438 0.36146864 0.5367674 -0.6009751 -0.31291908 -0.36228183] | [7.1199493408203125, 6.221828460693359] |
0441cb95-3d0d-4c42-9e48-474230563fa7 | rgb-d-salient-object-detection-with | 2109.03425 | null | https://arxiv.org/abs/2109.03425v1 | https://arxiv.org/pdf/2109.03425v1.pdf | RGB-D Salient Object Detection with Ubiquitous Target Awareness | Conventional RGB-D salient object detection methods aim to leverage depth as complementary information to find the salient regions in both modalities. However, the salient object detection results heavily rely on the quality of captured depth data which sometimes are unavailable. In this work, we make the first attempt to solve the RGB-D salient object detection problem with a novel depth-awareness framework. This framework only relies on RGB data in the testing phase, utilizing captured depth data as supervision for representation learning. To construct our framework as well as achieving accurate salient detection results, we propose a Ubiquitous Target Awareness (UTA) network to solve three important challenges in RGB-D SOD task: 1) a depth awareness module to excavate depth information and to mine ambiguous regions via adaptive depth-error weights, 2) a spatial-aware cross-modal interaction and a channel-aware cross-level interaction, exploiting the low-level boundary cues and amplifying high-level salient channels, and 3) a gated multi-scale predictor module to perceive the object saliency in different contextual scales. Besides its high performance, our proposed UTA network is depth-free for inference and runs in real-time with 43 FPS. Experimental evidence demonstrates that our proposed network not only surpasses the state-of-the-art methods on five public RGB-D SOD benchmarks by a large margin, but also verifies its extensibility on five public RGB SOD benchmarks. | ['Xiaowu Chen', 'Jia Li', 'Jiawei Zhao', 'Yifan Zhao'] | 2021-09-08 | null | null | null | null | ['rgb-d-salient-object-detection', 'thermal-image-segmentation'] | ['computer-vision', 'computer-vision'] | [ 3.78270775e-01 7.58379623e-02 -2.37529948e-01 -2.72470325e-01
-9.17073190e-01 -6.93840832e-02 1.26255915e-01 1.03875510e-01
-3.72387081e-01 3.41174871e-01 2.79555678e-01 2.74566002e-02
5.27952723e-02 -7.20646262e-01 -7.82436848e-01 -7.55105138e-01
1.03971161e-01 -2.57993072e-01 1.09290540e+00 -3.98859650e-01
2.71540225e-01 4.34413224e-01 -2.04436755e+00 2.90307760e-01
9.23077464e-01 1.64418411e+00 6.66581213e-01 4.85072345e-01
7.00508952e-02 9.31875288e-01 -1.21198140e-01 3.17435488e-02
5.10748982e-01 -5.54058254e-02 -4.93520916e-01 1.43871143e-01
6.58655703e-01 -5.46754122e-01 -3.50810766e-01 1.09270966e+00
6.92934155e-01 1.23394176e-01 3.93666886e-02 -1.28560007e+00
-4.93216872e-01 3.32239658e-01 -1.03689706e+00 5.53729117e-01
3.05717766e-01 4.31642383e-01 9.61691856e-01 -1.25829387e+00
3.91189903e-01 1.31400657e+00 4.54140365e-01 5.54046333e-01
-7.95062959e-01 -6.52886689e-01 6.13637030e-01 2.74341315e-01
-1.13196039e+00 -2.31138229e-01 1.26794589e+00 5.77470195e-03
8.96799624e-01 1.42049000e-01 9.18760896e-01 9.30954635e-01
-4.62461673e-02 1.31251156e+00 1.23077011e+00 -1.32961437e-01
3.06421518e-01 -1.71160102e-01 -1.07479177e-01 7.93654740e-01
1.53016731e-01 1.55521289e-01 -1.10372531e+00 2.89098889e-01
9.69087303e-01 1.36284277e-01 -1.68923721e-01 -6.43721163e-01
-1.27944946e+00 6.07210934e-01 1.08310890e+00 3.67172509e-02
-4.35510665e-01 1.90528661e-01 2.14321211e-01 -2.47582451e-01
3.74784261e-01 1.88049719e-01 -4.74783629e-01 -5.06582521e-02
-6.49105549e-01 9.24371257e-02 1.60276634e-03 9.21852648e-01
9.71772432e-01 5.67420088e-02 -1.87456891e-01 3.33217263e-01
3.91985267e-01 6.29673183e-01 4.06000197e-01 -6.76339030e-01
8.46121371e-01 1.00863457e+00 1.55048460e-01 -1.09955764e+00
-6.34545147e-01 -3.37768942e-01 -6.46537304e-01 3.18094403e-01
3.81335258e-01 2.18661666e-01 -1.03244984e+00 1.58448696e+00
7.21068799e-01 3.28810364e-01 -4.75239269e-02 1.54408026e+00
1.02096772e+00 3.49454284e-01 1.65000752e-01 -6.26815259e-02
1.28355432e+00 -9.41038549e-01 -3.96936029e-01 -6.94895446e-01
2.57546514e-01 -4.27928954e-01 1.46326220e+00 2.65915871e-01
-1.12944090e+00 -6.55811608e-01 -1.15443802e+00 -5.95021844e-01
-2.91952431e-01 1.79280430e-01 1.15858948e+00 4.06137973e-01
-8.23722541e-01 1.96510211e-01 -9.40684140e-01 -8.65211710e-02
6.51836157e-01 2.85131872e-01 -2.41328418e-01 -1.75488666e-01
-1.19141674e+00 6.27495110e-01 3.89789253e-01 4.03556496e-01
-1.01016152e+00 -7.10312724e-01 -9.90081668e-01 -6.58125952e-02
5.78597009e-01 -6.00896060e-01 9.13323402e-01 -8.29948545e-01
-1.26774120e+00 8.54263127e-01 -1.49148509e-01 -1.88960090e-01
4.66606557e-01 -3.55907440e-01 -1.51074320e-01 4.80843067e-01
2.82157689e-01 8.53245378e-01 8.07607710e-01 -1.31833911e+00
-1.26693594e+00 -7.26427019e-01 2.99685538e-01 4.91103709e-01
-3.22265178e-01 -3.98556560e-01 -6.39542341e-01 -5.38817942e-01
8.28694701e-01 -4.89572018e-01 -1.81551322e-01 4.70459521e-01
-4.33195740e-01 -1.66880712e-01 8.14546466e-01 -4.03088540e-01
8.36279333e-01 -2.04311967e+00 2.94844508e-01 -7.90998936e-02
4.91982102e-01 -6.23344630e-02 1.60121247e-01 -2.40245953e-01
2.14732841e-01 -3.61762345e-01 -1.90993562e-01 -5.26980877e-01
-8.73969272e-02 3.56234983e-02 -4.67620522e-01 5.07574797e-01
5.28945625e-01 1.03561091e+00 -1.29227555e+00 -8.22123885e-01
5.66262662e-01 4.34244305e-01 -4.77243721e-01 1.61845416e-01
-1.31363884e-01 2.62370914e-01 -8.04636538e-01 1.49331844e+00
7.72333682e-01 -1.88738376e-01 -3.47951263e-01 -5.48203826e-01
-3.30250323e-01 1.82902575e-01 -1.33995807e+00 2.14014840e+00
-3.13384593e-01 4.34113353e-01 4.46894541e-02 -6.64006650e-01
9.45444226e-01 -3.34004104e-01 5.07460058e-01 -1.15852869e+00
1.67048529e-01 4.03787196e-01 -4.59793627e-01 -4.31903988e-01
7.77620077e-01 1.06667370e-01 -2.05944896e-01 6.17828853e-02
-1.30579695e-01 -9.65997279e-02 -2.96267509e-01 2.07603261e-01
9.93643343e-01 1.55281469e-01 -1.18016014e-02 3.14094052e-02
2.69607395e-01 -1.33985311e-01 8.25371444e-01 6.93448901e-01
-6.43548667e-01 6.85325086e-01 3.43714654e-01 -3.51995707e-01
-5.97670674e-01 -1.32746506e+00 2.05726791e-02 1.09421158e+00
1.22422159e+00 -7.55683472e-03 -2.50119746e-01 -7.18954384e-01
1.58738315e-01 2.27450117e-01 -9.45201397e-01 -4.99890238e-01
-3.78830820e-01 -5.02112031e-01 1.40439421e-01 8.36765468e-01
8.87477577e-01 -1.00760043e+00 -1.34244072e+00 4.62920889e-02
-3.36056918e-01 -1.22851801e+00 -2.59007990e-01 6.07332826e-01
-1.01371276e+00 -9.33918297e-01 -6.67419732e-01 -6.29025221e-01
5.05271792e-01 8.78358841e-01 8.77574086e-01 -1.33190900e-01
-3.25997859e-01 1.24411836e-01 -3.43630672e-01 -4.89499629e-01
4.11024272e-01 1.28472984e-01 -7.60345086e-02 -6.63505569e-02
2.80928940e-01 -5.61620653e-01 -1.10891700e+00 4.03096318e-01
-8.54171216e-01 4.13127780e-01 9.89507139e-01 5.77768028e-01
8.22385252e-01 -2.30083659e-01 5.10033727e-01 -1.49359882e-01
-1.32143289e-01 -2.75435537e-01 -5.69705665e-01 3.97290327e-02
-4.16223973e-01 -1.99323550e-01 1.30908694e-02 -3.85010511e-01
-9.47594881e-01 4.88617688e-01 2.29853876e-02 -6.98648155e-01
1.09100863e-01 7.96165019e-02 -4.60459441e-01 -2.65731782e-01
5.51496506e-01 2.59317249e-01 -3.62906784e-01 -3.67442966e-01
3.43143821e-01 4.21598464e-01 6.67524338e-01 -3.79658967e-01
8.03397059e-01 8.96795511e-01 1.69427730e-02 -4.73201901e-01
-1.29394472e+00 -7.14071095e-01 -6.50806069e-01 -4.21796083e-01
8.73429835e-01 -1.39645016e+00 -7.69219339e-01 5.79986453e-01
-8.18040252e-01 -3.72393698e-01 -4.26193386e-01 3.10918003e-01
-4.78469789e-01 1.94804072e-01 -4.60554808e-01 -8.14908504e-01
-3.17142338e-01 -1.05282331e+00 1.80965889e+00 6.23722374e-01
4.62330341e-01 -3.90338629e-01 -4.53945369e-01 3.50607753e-01
2.07147464e-01 4.42089200e-01 4.82397050e-01 2.48318091e-01
-1.07347608e+00 9.86048132e-02 -8.09045553e-01 -3.18763070e-02
1.66536421e-01 -4.69778150e-01 -1.25601256e+00 -5.22891097e-02
-1.95790455e-01 -5.89909196e-01 1.08478320e+00 4.25506949e-01
1.26987278e+00 1.53001919e-01 -3.10779363e-01 8.21796119e-01
1.50295615e+00 -4.00456280e-01 3.97169799e-01 5.85521758e-01
1.07305360e+00 4.13158059e-01 1.14492524e+00 6.31296396e-01
6.89518988e-01 6.73675895e-01 1.12678158e+00 -4.86857146e-01
-2.43909240e-01 -5.39963841e-01 1.88059554e-01 3.01833361e-01
3.38058807e-02 9.08246413e-02 -7.61534274e-01 8.15297246e-01
-2.05081177e+00 -7.30735004e-01 -2.37879269e-02 1.92262161e+00
9.47067022e-01 4.53061283e-01 2.78331995e-01 2.48842806e-01
3.29294533e-01 2.21656531e-01 -1.08106923e+00 2.49328509e-01
-5.28734684e-01 -4.79366742e-02 6.69703722e-01 2.07255051e-01
-1.32119811e+00 1.01151657e+00 4.83518457e+00 6.95189178e-01
-1.34018183e+00 8.98971260e-02 4.74351466e-01 -5.37736416e-01
-3.80351096e-01 -1.66419744e-01 -9.08095360e-01 3.36126298e-01
8.69781245e-03 2.86127627e-01 -6.22584634e-02 1.14945185e+00
2.20923170e-01 -6.33368909e-01 -8.68748426e-01 1.23681188e+00
3.84836346e-02 -1.29170585e+00 -2.26329729e-01 -6.36849403e-02
7.68009245e-01 1.35861188e-01 2.16907784e-01 1.80383712e-01
3.71512142e-03 -6.20155513e-01 1.19168794e+00 4.09359634e-01
4.96931762e-01 -7.57384002e-01 5.78926861e-01 1.44054681e-01
-1.52157438e+00 -4.91875082e-01 -5.55949569e-01 -1.04316711e-01
1.06165551e-01 7.55946338e-01 -3.65666568e-01 4.58277911e-01
1.15481853e+00 1.16212511e+00 -8.53426218e-01 1.09341991e+00
-3.15878063e-01 1.72388613e-01 -4.31910425e-01 -4.87997644e-02
4.09987241e-01 4.29065883e-01 5.16327560e-01 9.46206331e-01
1.64844841e-01 2.89580673e-01 2.43725821e-01 8.43943954e-01
1.64097935e-01 -2.03149185e-01 -1.61299616e-01 5.28512359e-01
3.93038422e-01 1.27024424e+00 -9.41619635e-01 -6.64501041e-02
-2.42034182e-01 1.04159415e+00 4.28693593e-01 2.34020561e-01
-1.03715897e+00 -2.39132971e-01 7.21102774e-01 1.68386862e-01
5.59198320e-01 -1.81074142e-01 -5.78987420e-01 -1.13491070e+00
3.85690331e-01 -5.80544889e-01 4.33678418e-01 -1.12433231e+00
-9.63754654e-01 3.20249557e-01 -2.17959881e-01 -1.60178828e+00
2.33250573e-01 -5.33778012e-01 -1.95192218e-01 7.90336311e-01
-2.21296501e+00 -1.53248394e+00 -8.72372568e-01 8.70121062e-01
4.38565791e-01 2.37714842e-01 1.51768193e-01 2.57169217e-01
-4.59465176e-01 4.54686433e-01 -5.12778938e-01 -8.25622901e-02
5.17971992e-01 -1.18468034e+00 3.22016403e-02 1.14658797e+00
-7.95355886e-02 2.03461066e-01 5.41583300e-01 -5.12864709e-01
-1.74092245e+00 -1.06435919e+00 3.84989500e-01 -4.18893605e-01
4.32917148e-01 -5.31846464e-01 -7.86713541e-01 1.83384478e-01
-5.63316882e-01 5.05702138e-01 1.43733798e-02 -1.14037953e-01
-3.40953141e-01 -4.12514806e-01 -9.95244741e-01 5.20083904e-01
1.45841146e+00 -5.89306056e-01 -4.76669103e-01 5.52643463e-02
1.12709987e+00 -8.37500095e-01 -6.46170199e-01 5.70195079e-01
4.66020823e-01 -1.30034995e+00 1.35728860e+00 -1.04481585e-01
7.20593750e-01 -7.13647664e-01 -4.07589972e-01 -6.35810971e-01
-5.62446304e-02 -1.70105442e-01 -4.06865507e-01 1.11254311e+00
1.36259673e-02 -2.63851076e-01 1.07295763e+00 6.06316268e-01
-4.36323017e-01 -1.09134734e+00 -1.26282871e+00 -3.91791910e-01
-6.32633805e-01 -7.39316404e-01 5.55817187e-01 6.67475104e-01
-1.85241058e-01 5.33862785e-02 -2.54548222e-01 5.90059280e-01
8.07439148e-01 6.45741105e-01 6.81350946e-01 -1.06585312e+00
-7.81895518e-02 -4.54298198e-01 -7.14893520e-01 -1.42971265e+00
-2.63481557e-01 -4.87660795e-01 3.71360809e-01 -1.51566350e+00
6.45274594e-02 -6.56445980e-01 -6.38037860e-01 6.76921248e-01
-5.47883332e-01 4.61274058e-01 1.52322605e-01 8.55761096e-02
-9.80149269e-01 8.75856400e-01 1.36281538e+00 -1.62446782e-01
-3.28859419e-01 -2.56932974e-01 -9.30266678e-01 7.31996000e-01
4.16993737e-01 -2.28397712e-01 -4.86521751e-01 -4.37132776e-01
3.06722641e-01 -8.24293792e-02 9.20259714e-01 -1.27307069e+00
3.28702241e-01 -2.80843139e-01 7.51959801e-01 -1.03020704e+00
5.49295425e-01 -7.92990685e-01 -7.01345801e-01 2.61496991e-01
-3.63241658e-02 -2.26851612e-01 3.36974323e-01 7.98363507e-01
-1.77269503e-01 4.50498253e-01 8.48775506e-01 1.11328093e-02
-1.59665954e+00 5.05749762e-01 2.59394646e-01 1.41394019e-01
1.04405320e+00 -7.02805698e-01 -4.43114281e-01 -2.35215519e-02
-2.95780241e-01 4.25513923e-01 5.96165836e-01 6.83877230e-01
1.07178998e+00 -1.28605354e+00 -2.80388772e-01 3.78182381e-01
5.46383381e-01 5.30583918e-01 5.71820974e-01 1.00533414e+00
-1.21266782e-01 7.74901584e-02 -3.49092305e-01 -1.03724456e+00
-9.17945564e-01 4.30983931e-01 3.16721201e-01 3.34587283e-02
-6.89710140e-01 1.30015790e+00 4.58158404e-01 4.75757010e-02
4.79387432e-01 -6.86069131e-01 -2.22691353e-02 -5.27572818e-03
5.61283588e-01 6.22205064e-02 4.48766090e-02 -5.83113432e-01
-6.12638891e-01 7.72319138e-01 2.51794308e-01 1.61169350e-01
1.28915060e+00 -5.86802721e-01 2.84826010e-01 5.87962806e-01
9.37286615e-01 -2.40753442e-01 -1.95255888e+00 -4.54209566e-01
-3.28527302e-01 -8.60618711e-01 4.86733705e-01 -6.04138017e-01
-1.19022202e+00 9.78750527e-01 9.03320849e-01 -2.25151807e-01
1.52171922e+00 2.08066851e-01 8.77581537e-01 1.07096501e-01
6.03950500e-01 -1.23111892e+00 6.50262058e-01 1.57164633e-01
7.73270309e-01 -1.79077697e+00 1.65580809e-01 -5.62344968e-01
-7.04779208e-01 8.08929026e-01 1.10761452e+00 6.31578639e-02
5.01830399e-01 5.61628677e-02 -4.77622338e-02 -2.99076110e-01
-5.72859108e-01 -6.53732240e-01 3.30650449e-01 7.44864047e-01
-1.48610264e-01 -2.40308151e-01 3.87871593e-01 7.60808051e-01
8.68825614e-02 -7.74664804e-02 2.28490368e-01 9.35395896e-01
-6.79574311e-01 -2.87937909e-01 -1.83738112e-01 2.17444599e-01
-4.01074365e-02 -4.74403873e-02 -2.49974370e-01 8.70870411e-01
3.76357615e-01 8.87590766e-01 -4.01782282e-02 -4.82865453e-01
6.19901419e-01 -5.85005343e-01 3.61483574e-01 -3.76573116e-01
-3.42858851e-01 -3.78844552e-02 -2.85585552e-01 -1.09941196e+00
-6.98645711e-01 -5.72498500e-01 -1.41581917e+00 3.36989537e-02
-4.00052518e-01 -4.55496609e-01 4.88810807e-01 8.64381969e-01
3.89757991e-01 5.86691916e-01 6.77874804e-01 -1.37286770e+00
-2.50225157e-01 -6.52211726e-01 -6.24461114e-01 2.38527492e-01
6.52064204e-01 -1.03848696e+00 -3.13778937e-01 -2.42502987e-01] | [9.731304168701172, -0.7387523055076599] |
43a16598-08e1-4024-a22d-68a9a4da03af | a-lightweight-neural-network-for-monocular | 2007.12577 | null | https://arxiv.org/abs/2007.12577v1 | https://arxiv.org/pdf/2007.12577v1.pdf | A Lightweight Neural Network for Monocular View Generation with Occlusion Handling | In this article, we present a very lightweight neural network architecture, trained on stereo data pairs, which performs view synthesis from one single image. With the growing success of multi-view formats, this problem is indeed increasingly relevant. The network returns a prediction built from disparity estimation, which fills in wrongly predicted regions using a occlusion handling technique. To do so, during training, the network learns to estimate the left-right consistency structural constraint on the pair of stereo input images, to be able to replicate it at test time from one single image. The method is built upon the idea of blending two predictions: a prediction based on disparity estimation, and a prediction based on direct minimization in occluded regions. The network is also able to identify these occluded areas at training and at test time by checking the pixelwise left-right consistency of the produced disparity maps. At test time, the approach can thus generate a left-side and a right-side view from one input image, as well as a depth map and a pixelwise confidence measure in the prediction. The work outperforms visually and metric-wise state-of-the-art approaches on the challenging KITTI dataset, all while reducing by a very significant order of magnitude (5 or 10 times) the required number of parameters (6.5 M). | ['Simon Evain', 'Christine Guillemot'] | 2020-07-24 | null | null | null | null | ['occlusion-handling'] | ['computer-vision'] | [ 5.53446472e-01 3.31622720e-01 2.81242043e-01 -4.81652945e-01
-7.29439855e-01 -4.09807473e-01 5.88101566e-01 -8.20811167e-02
-4.51857656e-01 8.00615191e-01 -1.49888858e-01 -2.31270209e-01
2.38762513e-01 -9.53508437e-01 -1.04300368e+00 -6.14060283e-01
1.53759435e-01 6.53761744e-01 4.37615305e-01 3.20087783e-02
5.01935482e-01 6.58051789e-01 -2.09514070e+00 5.49794734e-01
6.20348752e-01 1.42004788e+00 3.17337334e-01 7.74764955e-01
1.92992955e-01 6.07850790e-01 -2.51157135e-01 -2.81948924e-01
5.47740817e-01 -2.44977877e-01 -6.24230981e-01 2.15243205e-01
1.21985137e+00 -6.09813511e-01 6.66138157e-02 8.97519886e-01
4.32180583e-01 -1.25697032e-01 5.20171940e-01 -9.30914283e-01
8.71940255e-02 4.30744477e-02 -5.81063509e-01 -1.30737528e-01
4.95865911e-01 3.05820912e-01 8.11853230e-01 -1.05345488e+00
9.50040758e-01 8.36085558e-01 6.10266089e-01 2.71943033e-01
-1.34525263e+00 -2.99261510e-01 -7.51910433e-02 8.96771848e-02
-1.09025240e+00 -4.42448556e-01 6.62400603e-01 -5.83016634e-01
1.09903061e+00 1.13332547e-01 7.74817109e-01 6.48830056e-01
2.67318100e-01 3.25794309e-01 1.36371982e+00 -4.15124118e-01
3.12960625e-01 2.04564422e-01 -6.45137906e-01 7.10941315e-01
-4.54525724e-02 5.78073382e-01 -5.22955060e-01 3.05189908e-01
8.96145582e-01 -2.03730121e-01 -4.71521437e-01 -8.73825967e-01
-1.02350688e+00 5.48029423e-01 5.79824626e-01 2.52139196e-02
-5.56438565e-01 -3.86418179e-02 1.58369482e-01 1.80337042e-01
4.96179193e-01 2.07666561e-01 -5.04892230e-01 8.23844895e-02
-1.18775237e+00 2.75367051e-01 5.25743008e-01 4.47303742e-01
1.09184372e+00 2.92136874e-02 1.91159919e-01 6.06919527e-01
1.49063215e-01 4.71830487e-01 2.88270473e-01 -1.02050781e+00
7.04824090e-01 7.05408156e-01 1.06513895e-01 -9.18285906e-01
-3.82124186e-01 -3.84397238e-01 -8.75755310e-01 1.14352369e+00
6.33464754e-01 2.84114154e-03 -1.00817895e+00 1.57755399e+00
5.31633496e-01 -7.41639659e-02 -1.04146069e-02 8.78841639e-01
3.74904543e-01 3.95915687e-01 -4.74773079e-01 -1.59963313e-02
1.02257514e+00 -7.86333084e-01 4.28060489e-03 -6.09256029e-01
2.88402557e-01 -9.06154215e-01 6.55301213e-01 7.54054844e-01
-1.60150099e+00 -8.43345523e-01 -1.16876936e+00 8.70141946e-03
-3.19765091e-01 2.16362059e-01 5.61723001e-02 4.60478991e-01
-1.37801099e+00 9.43693459e-01 -4.32882816e-01 -4.27973606e-02
1.32912338e-01 4.29566294e-01 -7.10051477e-01 -1.94619223e-01
-8.59086037e-01 8.68038774e-01 4.76268739e-01 2.64764726e-01
-5.66471756e-01 -7.14604557e-01 -1.09981596e+00 3.97548117e-02
1.53938875e-01 -7.33101070e-01 9.41771805e-01 -1.20063508e+00
-1.39640522e+00 1.26519525e+00 -7.77318552e-02 -5.91559887e-01
9.82013166e-01 -3.04575041e-02 -1.70347262e-02 2.63808846e-01
2.25750118e-01 1.24955750e+00 9.70796168e-01 -1.46759343e+00
-1.03273094e+00 -5.11566281e-01 -1.18548665e-02 3.19053501e-01
3.98788959e-01 -5.73030770e-01 -5.08420110e-01 -3.35286558e-01
5.94034612e-01 -7.12081671e-01 -9.73975211e-02 3.75894874e-01
-3.51861954e-01 3.88172716e-01 5.61980009e-01 -8.87592793e-01
5.84338367e-01 -1.89790046e+00 2.26180717e-01 3.71682495e-01
1.40715882e-01 1.35919213e-01 1.05544375e-02 1.58219770e-01
-3.71203274e-01 -3.38795662e-01 -4.10716325e-01 -5.70698500e-01
-4.49767828e-01 -1.28666490e-01 -2.75754720e-01 5.47824919e-01
2.19967544e-01 5.10381997e-01 -6.79738522e-01 -2.00874358e-01
6.70434117e-01 4.82066780e-01 -6.88988328e-01 4.44613516e-01
-2.49227673e-01 5.53236961e-01 2.76392817e-01 3.60524714e-01
9.98103559e-01 -7.51269469e-03 2.14402109e-01 -2.97285497e-01
-3.46550643e-01 2.17255712e-01 -1.40300405e+00 1.66193497e+00
-7.71457732e-01 8.91975343e-01 -2.10498795e-01 -9.98049498e-01
1.03373170e+00 1.32794559e-01 3.00191790e-01 -9.65188324e-01
5.79451881e-02 4.64596957e-01 -1.60522580e-01 -2.87829638e-01
3.65305543e-01 -8.23036730e-02 3.07412505e-01 3.73354048e-01
-1.92850545e-01 -2.84746766e-01 1.39066294e-01 -2.28619248e-01
7.83745170e-01 5.92136264e-01 3.51643175e-01 1.32588714e-01
9.40944314e-01 -2.45621249e-01 2.93934256e-01 3.70882243e-01
1.63545266e-01 1.20005500e+00 6.12630188e-01 -1.01221514e+00
-1.33684313e+00 -1.08779967e+00 -1.11775217e-03 3.52880180e-01
1.46853462e-01 1.17775261e-01 -7.16550529e-01 -6.19689643e-01
-9.86748412e-02 5.01348674e-01 -7.94296801e-01 1.27174154e-01
-8.26161265e-01 5.98045513e-02 -8.99013877e-03 4.57079291e-01
7.22597539e-01 -1.21941745e+00 -1.21413147e+00 1.27632484e-01
-3.74488011e-02 -1.05691695e+00 -6.84085265e-02 3.98761451e-01
-9.93220747e-01 -1.07736611e+00 -8.32424760e-01 -9.03451085e-01
8.48719478e-01 7.77041391e-02 1.42394412e+00 1.13455087e-01
-2.57097512e-01 -1.95717722e-01 1.15238898e-01 6.05816469e-02
-5.34952581e-01 -2.50060588e-01 -1.46467134e-01 1.10401161e-01
-2.21390992e-01 -8.08117032e-01 -9.28052962e-01 4.76059705e-01
-9.07638371e-01 5.39216340e-01 5.48608541e-01 9.47266400e-01
8.01847160e-01 -2.77017981e-01 -5.20408228e-02 -6.69773698e-01
-1.59378514e-01 1.53769003e-02 -1.04306614e+00 7.65106007e-02
-6.25475049e-01 1.27007008e-01 7.71440148e-01 -1.65522508e-02
-8.46485555e-01 4.16856885e-01 -2.72240847e-01 -5.62709391e-01
-4.25347298e-01 7.06733838e-02 6.83680698e-02 -1.15269539e-03
6.35606527e-01 2.85128027e-01 -5.65691339e-03 -2.39354804e-01
2.76801437e-01 4.06543821e-01 6.47000372e-01 -1.07249850e-02
7.13401496e-01 6.97338998e-01 9.06983018e-02 -2.59879440e-01
-6.02244020e-01 -2.23680317e-01 -8.58924448e-01 -5.05531609e-01
9.29698110e-01 -9.14794505e-01 -5.38777232e-01 4.42295641e-01
-1.31629086e+00 -3.90175462e-01 -3.91302943e-01 3.25694323e-01
-7.81951666e-01 7.88094401e-02 -1.35483280e-01 -5.44600070e-01
-1.74323529e-01 -1.22212136e+00 1.25075471e+00 -3.51306843e-03
2.13695206e-02 -7.40913272e-01 1.07893847e-01 3.28097135e-01
1.16843626e-01 2.65797853e-01 9.24252987e-01 -1.59522802e-01
-9.20482039e-01 -2.53872514e-01 -3.44683617e-01 5.39122283e-01
-1.51752576e-01 -4.09561805e-02 -1.23381364e+00 -2.96239614e-01
3.43246311e-02 -3.24936837e-01 7.96888471e-01 5.67565560e-01
9.28458929e-01 1.55885946e-02 -1.30911022e-01 7.30491757e-01
1.75669825e+00 1.32599935e-01 9.40825880e-01 4.97962296e-01
4.79798645e-01 8.25003326e-01 5.05265236e-01 2.90334046e-01
1.30209982e-01 9.49737787e-01 8.48341346e-01 -2.16820806e-01
-2.99241364e-01 -2.47252554e-01 1.04668051e-01 1.00262403e-01
-3.15017365e-02 -1.57308444e-01 -9.27892566e-01 4.66564149e-01
-1.47652888e+00 -8.65743041e-01 -4.00753170e-02 2.67963338e+00
4.79423851e-01 4.76308137e-01 1.32896695e-02 3.58715117e-01
6.48613453e-01 1.59349471e-01 -4.93053168e-01 -6.01827562e-01
1.59265082e-02 1.96433872e-01 4.38295871e-01 7.32329726e-01
-8.47590983e-01 5.89912176e-01 5.43034458e+00 5.55824578e-01
-1.51562345e+00 -4.27983940e-01 1.13954413e+00 -1.67232174e-02
-2.17442378e-01 1.62353590e-02 -5.85494757e-01 3.05251956e-01
5.39823234e-01 4.00353491e-01 3.81857753e-01 8.59380901e-01
7.60183036e-02 -7.51026452e-01 -1.24052024e+00 1.08852828e+00
3.02169263e-01 -1.66198707e+00 -1.02455482e-01 1.58924758e-01
9.36378419e-01 2.64503434e-02 7.73031712e-02 -1.89978197e-01
-2.19948888e-01 -9.98076141e-01 8.86361122e-01 4.07825887e-01
1.06832457e+00 -7.50435770e-01 7.87425101e-01 4.40373272e-01
-1.05672002e+00 5.60813546e-02 -1.57080740e-01 -1.07222475e-01
2.81926215e-01 5.64945757e-01 -7.40769207e-01 5.69081724e-01
6.70510650e-01 5.82141459e-01 -4.36708659e-01 1.04270983e+00
-4.50394034e-01 -2.08659649e-01 -2.09112495e-01 4.76877838e-01
1.75864935e-01 -3.14739883e-01 3.37327987e-01 7.28467524e-01
4.26042497e-01 -5.09454250e-01 -2.34349310e-01 8.50822449e-01
1.36476398e-01 -1.34789795e-01 -8.31797600e-01 9.12924051e-01
1.14606075e-01 1.11425018e+00 -7.61541247e-01 -4.58717883e-01
-1.72795773e-01 1.26604664e+00 5.84684372e-01 9.77219120e-02
-6.93260491e-01 -1.01189941e-01 3.12007189e-01 3.09081167e-01
7.07693756e-01 1.62246719e-01 -5.51128626e-01 -8.17048490e-01
5.36332846e-01 -7.56372392e-01 9.62696911e-04 -1.08714926e+00
-8.61601055e-01 9.74719226e-01 -3.33336890e-01 -1.68898654e+00
-8.88024092e-01 -7.10205913e-01 -6.12789035e-01 1.13463736e+00
-1.51251054e+00 -7.42694795e-01 -4.54331726e-01 1.83287546e-01
5.81213713e-01 -3.95725183e-02 7.13024437e-01 3.04853380e-01
-1.75663888e-01 1.46521181e-01 -4.80208136e-02 -1.61178216e-01
5.62586010e-01 -1.27081895e+00 6.24408901e-01 7.61149049e-01
1.81732461e-01 5.85518719e-04 7.14369476e-01 -3.78096968e-01
-6.77802563e-01 -8.24983418e-01 1.12415171e+00 -1.06064014e-01
1.13892004e-01 -1.89938620e-01 -7.49136031e-01 3.28880250e-01
1.62672073e-01 2.04943120e-01 -1.06917255e-01 -3.53883952e-01
-2.60565370e-01 -2.79992640e-01 -1.31052089e+00 4.71622229e-01
7.84098327e-01 -4.38597649e-01 -2.90974289e-01 9.01244953e-02
1.97713077e-01 -8.03517580e-01 -5.29904366e-01 3.93193007e-01
8.76847148e-01 -2.16198635e+00 9.99025643e-01 -7.54023716e-02
1.10346746e+00 -3.85152340e-01 -4.59786467e-02 -1.19920599e+00
2.01547757e-01 -2.01392666e-01 1.78964689e-01 6.91614509e-01
4.92000878e-01 -5.00913560e-01 1.18077087e+00 4.18721944e-01
5.30431569e-02 -1.14438570e+00 -1.30811441e+00 -6.12051189e-01
-1.29917353e-01 -3.26913893e-01 2.06028000e-01 3.64868462e-01
-4.69966799e-01 1.47126779e-01 -3.23423207e-01 2.16793925e-01
4.92246985e-01 3.52560222e-01 9.80174601e-01 -1.06816685e+00
-4.92055416e-01 -4.14391905e-01 -6.26931429e-01 -1.16633606e+00
-1.99971929e-01 -4.16141093e-01 2.12822691e-01 -1.50115120e+00
-2.85187513e-01 -3.12450022e-01 1.44896477e-01 3.93855534e-02
1.46739170e-01 7.68117189e-01 1.64849877e-01 -8.79175309e-03
-1.44600999e-02 1.81426361e-01 1.05869508e+00 1.80368081e-01
-1.95145994e-01 1.15294531e-01 -1.07683614e-01 1.02272308e+00
7.06739485e-01 -3.65795374e-01 -2.97607124e-01 -3.60011399e-01
4.39527899e-01 4.77556050e-01 6.94659889e-01 -1.51184893e+00
9.21317264e-02 1.87911153e-01 9.01462317e-01 -9.35067892e-01
6.75886691e-01 -8.62962842e-01 2.47288048e-01 6.50200009e-01
-2.29744971e-01 2.79845625e-01 1.81041941e-01 3.19558620e-01
-2.20615655e-01 -1.55622259e-01 9.80539799e-01 -2.47629672e-01
-8.22450459e-01 9.05079320e-02 1.08569510e-01 -1.06494389e-01
9.47356462e-01 -8.31142247e-01 -5.01487069e-02 -3.67534816e-01
-6.81288004e-01 -1.36260703e-01 7.25430906e-01 1.24841206e-01
8.75805199e-01 -1.10017669e+00 -5.72762430e-01 7.46514499e-01
6.64843544e-02 2.46908277e-01 4.20527190e-01 6.48582518e-01
-8.61379266e-01 2.92881012e-01 -4.98247772e-01 -9.21235859e-01
-1.19062340e+00 4.10748750e-01 6.57101393e-01 -4.79490578e-01
-6.66907668e-01 7.85911679e-01 3.17538381e-01 -3.22366357e-01
2.65535623e-01 -3.45071256e-01 -1.38488621e-01 -1.41317979e-01
4.64653611e-01 1.02129810e-01 2.96378344e-01 -5.71182311e-01
-1.47963136e-01 9.79710281e-01 1.67205587e-01 -5.32539129e-01
1.30303085e+00 -7.84080476e-02 7.21888617e-02 2.85046577e-01
1.29170036e+00 1.05846062e-01 -1.81716001e+00 1.82928331e-02
-3.33682567e-01 -7.47983634e-01 3.75963072e-03 -1.02262032e+00
-1.24099386e+00 1.08050561e+00 9.60918844e-01 1.12849332e-01
1.18199265e+00 -2.91531712e-01 5.29006302e-01 5.66366576e-02
3.64987820e-01 -9.94273722e-01 -5.08481590e-03 2.94377387e-01
1.05288863e+00 -1.39919066e+00 1.34794474e-01 -3.85293454e-01
-4.51477975e-01 1.32206774e+00 7.55983055e-01 -2.70755351e-01
2.56640255e-01 1.34571552e-01 1.57305390e-01 -1.98656097e-01
-6.29275024e-01 -1.49990439e-01 4.92487162e-01 6.37480319e-01
3.20455074e-01 -2.19251961e-01 1.93015590e-01 -5.86015821e-01
-2.34528482e-01 -1.35313064e-01 3.96751851e-01 5.46563745e-01
-5.15016019e-01 -1.10855389e+00 -3.56136143e-01 2.89977074e-01
-1.11600846e-01 -2.57443160e-01 -8.65429714e-02 8.37552428e-01
4.90562528e-01 5.17943740e-01 5.12943268e-01 -3.33763599e-01
3.80349547e-01 -1.47213802e-01 4.68309939e-01 -4.23016131e-01
-4.69072044e-01 -1.09113097e-01 1.04718089e-01 -8.40430558e-01
-3.61953318e-01 -5.33308327e-01 -8.75216603e-01 -1.52087271e-01
-1.75186228e-02 -4.28709805e-01 8.19117427e-01 9.08728957e-01
1.82883412e-01 3.17483485e-01 8.52955878e-01 -1.48418033e+00
-2.58636534e-01 -6.26713455e-01 -2.94866085e-01 4.18692350e-01
4.19688553e-01 -4.63607162e-01 -4.18945730e-01 3.08823641e-02] | [8.891796112060547, -2.582298517227173] |
60a01982-36cf-49cf-b84d-298905cb34aa | a-control-centric-benchmark-for-video | 2304.13723 | null | https://arxiv.org/abs/2304.13723v1 | https://arxiv.org/pdf/2304.13723v1.pdf | A Control-Centric Benchmark for Video Prediction | Video is a promising source of knowledge for embodied agents to learn models of the world's dynamics. Large deep networks have become increasingly effective at modeling complex video data in a self-supervised manner, as evaluated by metrics based on human perceptual similarity or pixel-wise comparison. However, it remains unclear whether current metrics are accurate indicators of performance on downstream tasks. We find empirically that for planning robotic manipulation, existing metrics can be unreliable at predicting execution success. To address this, we propose a benchmark for action-conditioned video prediction in the form of a control benchmark that evaluates a given model for simulated robotic manipulation through sampling-based planning. Our benchmark, Video Prediction for Visual Planning ($VP^2$), includes simulated environments with 11 task categories and 310 task instance definitions, a full planning implementation, and training datasets containing scripted interaction trajectories for each task category. A central design goal of our benchmark is to expose a simple interface -- a single forward prediction call -- so it is straightforward to evaluate almost any action-conditioned video prediction model. We then leverage our benchmark to study the effects of scaling model size, quantity of training data, and model ensembling by analyzing five highly-performant video prediction models, finding that while scale can improve perceptual quality when modeling visually diverse settings, other attributes such as uncertainty awareness can also aid planning performance. | ['Jiajun Wu', 'Chelsea Finn', 'Stephen Tian'] | 2023-04-26 | null | null | null | null | ['video-prediction'] | ['computer-vision'] | [ 5.43266758e-02 -4.76556178e-03 -2.59809673e-01 -2.64486492e-01
-5.18752277e-01 -5.91804504e-01 7.78732657e-01 1.79582089e-02
-5.68303287e-01 5.36489248e-01 5.93191504e-01 -2.14133888e-01
-6.34307861e-02 -4.37070668e-01 -9.61308658e-01 -1.86210856e-01
-6.38018310e-01 4.50069696e-01 3.24429125e-01 -1.73299789e-01
2.30709106e-01 4.44695950e-01 -1.62508953e+00 5.97189248e-01
6.73690796e-01 9.18414950e-01 4.73808527e-01 8.22728038e-01
5.01348019e-01 1.27984929e+00 -3.74081105e-01 -4.58430052e-02
5.48176348e-01 -1.76741153e-01 -9.15704072e-01 -4.67245281e-02
3.12588692e-01 -6.75194800e-01 -6.85475647e-01 6.24409318e-01
2.43643448e-01 5.40250361e-01 7.36074686e-01 -1.63026929e+00
-2.73767769e-01 7.04123795e-01 8.24991316e-02 2.44318843e-01
6.17877841e-01 1.19247103e+00 9.21499789e-01 -4.02939290e-01
1.13375711e+00 1.24282300e+00 7.54380703e-01 3.63412350e-01
-1.34500837e+00 -4.79610115e-01 2.85771519e-01 3.95374238e-01
-1.08826447e+00 -6.85360253e-01 2.33862042e-01 -7.89456844e-01
1.57357311e+00 -8.30627903e-02 9.32755888e-01 1.38276255e+00
3.24221402e-01 8.45739603e-01 8.57235849e-01 1.82241738e-01
4.53429699e-01 -2.99951136e-01 -1.75521657e-01 8.41633201e-01
-1.03523597e-01 5.97428203e-01 -6.37162268e-01 4.14765477e-02
8.92502487e-01 -3.00628722e-01 -5.02216876e-01 -7.92815506e-01
-1.64033306e+00 3.63704711e-01 4.98796791e-01 -1.64567709e-01
-5.36873996e-01 5.62747598e-01 5.33587337e-01 4.51984406e-01
-6.04225439e-04 1.05374467e+00 -5.88996887e-01 -8.08835149e-01
-6.10371590e-01 7.00873673e-01 9.70206439e-01 1.11314905e+00
5.54351211e-01 1.32908136e-01 -3.63555819e-01 3.25553745e-01
3.86093184e-02 2.14225009e-01 3.32191586e-01 -1.82842577e+00
3.68532419e-01 4.44182336e-01 4.88755554e-01 -8.83112729e-01
-7.69930720e-01 8.04216042e-02 -4.91485111e-02 5.89040101e-01
6.87736034e-01 -3.20470750e-01 -8.86984944e-01 1.80293608e+00
-7.05125928e-02 8.00465941e-02 1.17239088e-01 1.05814326e+00
6.47870004e-01 7.25467026e-01 4.29523855e-01 -3.00506502e-02
8.23469520e-01 -1.33933222e+00 -1.72281936e-01 -4.42572504e-01
8.02825451e-01 -3.94275308e-01 1.48371601e+00 4.06318516e-01
-1.07143962e+00 -5.16015053e-01 -8.39285433e-01 8.34559798e-02
-9.25460011e-02 -8.90720710e-02 6.52584314e-01 1.51571020e-01
-1.22132587e+00 9.70342994e-01 -1.13301241e+00 -6.21697605e-01
3.37650895e-01 2.59252608e-01 -5.81838906e-01 -2.08019450e-01
-5.35896659e-01 1.23004127e+00 4.44997162e-01 -3.34310532e-01
-1.62082791e+00 -6.43238664e-01 -8.44338894e-01 2.86570102e-01
3.50614250e-01 -7.23302841e-01 1.48161757e+00 -1.03770733e+00
-1.47922492e+00 3.82218599e-01 1.23941481e-01 -4.67754871e-01
6.11583471e-01 -3.57905291e-02 -6.94458485e-02 6.54314607e-02
1.91731134e-03 1.17783976e+00 6.83519185e-01 -1.38904643e+00
-4.60846633e-01 1.63726322e-02 4.75105405e-01 5.07626712e-01
-3.72242518e-02 -1.53608546e-01 -4.08657342e-01 -3.69239122e-01
-3.31218719e-01 -1.08076537e+00 -2.63559967e-01 8.88940692e-02
5.56870503e-03 -4.06198353e-02 3.62767398e-01 -6.36194289e-01
7.36849427e-01 -2.06697798e+00 3.48528028e-01 -3.93371619e-02
1.91383585e-01 -8.15724283e-02 -5.94166458e-01 4.99129474e-01
1.74263388e-01 1.51571289e-01 7.70790428e-02 -3.18268359e-01
1.98370904e-01 2.76778173e-02 1.36332624e-02 2.52636552e-01
1.94042996e-01 1.04523194e+00 -1.05367374e+00 -4.72206503e-01
2.71277308e-01 2.33505860e-01 -9.25335765e-01 1.38485819e-01
-7.24910557e-01 4.27863598e-01 -1.24748707e-01 5.36268234e-01
-4.01374213e-02 -2.68049002e-01 2.54869878e-01 -1.89916760e-01
-8.31198096e-02 2.09643260e-01 -7.98711002e-01 1.89376545e+00
-2.52626359e-01 9.90208030e-01 -3.84330563e-02 -5.63710988e-01
4.29058880e-01 7.24195614e-02 5.12674987e-01 -6.18596733e-01
7.82207474e-02 -2.73614168e-01 3.08270335e-01 -7.26315081e-01
7.27788270e-01 2.96890616e-01 8.44596624e-02 3.21902126e-01
-4.13828949e-03 -3.72749805e-01 2.91942567e-01 1.19516313e-01
1.81753087e+00 5.20826578e-01 -1.07890032e-02 -1.72985673e-01
-4.98414427e-01 8.52222025e-01 3.52370948e-01 9.36052322e-01
-6.74539089e-01 4.96596396e-01 6.87716901e-01 -5.55673957e-01
-1.34356272e+00 -1.04018068e+00 3.03874701e-01 1.24530327e+00
2.92461723e-01 -7.22057700e-01 -5.95969141e-01 -5.01145601e-01
-5.27043752e-02 8.76050293e-01 -5.43299258e-01 -2.13805139e-01
-3.85283649e-01 -4.13633525e-01 3.60667050e-01 7.33555079e-01
2.70903468e-01 -1.40494132e+00 -1.07371235e+00 5.53179830e-02
-1.47242740e-01 -1.17543924e+00 -2.89898485e-01 1.46937728e-01
-7.15042114e-01 -1.15622759e+00 -3.07074219e-01 -7.46923923e-01
4.06380981e-01 2.58407027e-01 1.42999887e+00 2.59215292e-02
2.28225645e-02 9.26464081e-01 -5.08946240e-01 -1.47003442e-01
-6.37239218e-01 -1.04251906e-01 2.67650038e-01 -7.15824723e-01
5.73322549e-02 -4.92612958e-01 -6.43752277e-01 4.35231298e-01
-4.64215517e-01 3.44920218e-01 5.97555697e-01 7.22855091e-01
3.49025041e-01 -1.95462644e-01 8.91814083e-02 -2.72566646e-01
7.73514152e-01 -4.49092805e-01 -5.20050526e-01 2.36287475e-01
-3.63626122e-01 3.33126844e-03 5.36333859e-01 -6.11084998e-01
-8.06557596e-01 2.26540029e-01 2.48268127e-01 -6.63026690e-01
-3.52503002e-01 6.55840576e-01 1.62822664e-01 -1.37929037e-01
9.68015134e-01 6.60597682e-02 2.67550856e-01 4.42797430e-02
3.26534152e-01 9.45680141e-02 4.55811977e-01 -7.26047039e-01
4.15153742e-01 2.13428661e-01 -2.76703894e-01 -7.26952374e-01
-3.28998230e-02 -5.72725683e-02 -3.66574466e-01 -5.04844129e-01
9.97675300e-01 -1.04041314e+00 -1.18292022e+00 3.20885420e-01
-1.06741810e+00 -1.40021205e+00 -2.09639356e-01 6.88491344e-01
-1.05301702e+00 1.91093404e-02 -7.77685165e-01 -4.10943747e-01
3.19957674e-01 -1.71020222e+00 8.32336426e-01 -3.24046202e-02
-5.08141637e-01 -5.56331813e-01 6.09750627e-03 2.12141201e-01
5.11164606e-01 1.24702804e-01 9.76070166e-01 -5.17004550e-01
-9.27636504e-01 6.50316477e-02 -3.52083206e-01 -5.80999954e-03
-3.91235352e-01 5.83502688e-02 -7.38851011e-01 -2.26058990e-01
-3.80445361e-01 -7.49204695e-01 7.05265820e-01 3.98524463e-01
1.05758476e+00 -1.70365721e-01 -3.62103134e-01 5.28728127e-01
1.16152310e+00 2.09354877e-01 6.23015583e-01 4.67902899e-01
6.13914013e-01 3.59309286e-01 7.06939757e-01 4.09354866e-01
5.86859226e-01 7.21529484e-01 7.09418774e-01 3.82940650e-01
2.89440677e-02 -3.02927405e-01 6.89045668e-01 2.69603223e-01
-2.09868312e-01 -4.05919075e-01 -1.33761895e+00 2.92111874e-01
-1.78355885e+00 -1.19961071e+00 4.34307754e-01 2.06135893e+00
4.14371252e-01 3.53445768e-01 4.10300642e-01 -3.65347862e-01
2.48521462e-01 1.06721960e-01 -6.92593753e-01 -7.30251446e-02
2.80695051e-01 -3.07346314e-01 4.49427813e-01 4.08033431e-01
-1.00705302e+00 1.00668001e+00 7.17598963e+00 5.62121809e-01
-1.00898790e+00 -5.57998568e-02 6.28428698e-01 -4.84346956e-01
-8.95066932e-02 -1.06865287e-01 -2.79913634e-01 3.53185028e-01
8.84391665e-01 4.11763601e-03 8.14567447e-01 1.02607417e+00
5.71883500e-01 -4.27679569e-01 -1.55390453e+00 8.63377035e-01
-9.97070596e-02 -1.59500396e+00 -1.88143048e-02 5.82108973e-03
6.42547369e-01 5.57009518e-01 9.93298367e-03 7.70460725e-01
6.55029893e-01 -1.36567700e+00 9.15683329e-01 4.66903597e-01
5.74780524e-01 -2.47511923e-01 3.04238945e-01 4.14243549e-01
-9.34710681e-01 -3.27398092e-01 -1.71271041e-01 -5.01745820e-01
2.16482759e-01 -4.14596796e-01 -9.09993112e-01 -1.18842937e-01
8.31590831e-01 6.58109128e-01 -5.04351318e-01 1.27088547e+00
2.27823779e-01 4.17321146e-01 -3.63000154e-01 -1.34270728e-01
3.88069212e-01 1.64233014e-01 5.46951234e-01 9.33366477e-01
1.78593546e-01 2.32454479e-01 4.93442595e-01 7.05392063e-01
1.34299636e-01 -1.86655298e-01 -8.64547074e-01 -2.40514502e-01
6.33367777e-01 7.94507444e-01 -9.14336264e-01 -1.56613320e-01
-4.20824587e-01 7.51146436e-01 5.29069364e-01 4.89823520e-01
-9.78009522e-01 2.80107379e-01 9.75458264e-01 5.11735752e-02
1.29384577e-01 -6.89005911e-01 -1.65419698e-01 -1.04364586e+00
-7.63289332e-02 -1.14685798e+00 -7.51729161e-02 -1.18564916e+00
-7.27572680e-01 4.51914430e-01 1.24228843e-01 -1.08503413e+00
-2.21204311e-01 -6.86724603e-01 -4.95023966e-01 3.63719106e-01
-9.18684363e-01 -9.38835204e-01 -6.65316105e-01 3.80374402e-01
7.13391483e-01 -1.37094811e-01 7.25525022e-01 -1.58190653e-01
-5.98709881e-01 3.39618802e-01 -3.95660521e-03 -1.11069001e-01
6.45080686e-01 -1.01580608e+00 6.27039433e-01 6.87415719e-01
7.76180997e-02 3.45678896e-01 9.97633338e-01 -7.52818227e-01
-1.57656407e+00 -9.10925031e-01 7.40108639e-02 -9.33822155e-01
6.32643461e-01 -1.16634794e-01 -5.31516552e-01 1.03151023e+00
1.51855513e-01 -1.68597564e-01 1.79749459e-01 8.96779448e-02
-1.44634590e-01 3.01221311e-01 -9.39442337e-01 9.52325344e-01
1.61807501e+00 -3.61871481e-01 -4.33539212e-01 4.82673198e-01
9.22342718e-01 -6.35921836e-01 -1.11420536e+00 3.76605421e-01
7.18228281e-01 -1.09345889e+00 9.35041368e-01 -7.88709879e-01
9.10540164e-01 -1.37450829e-01 -2.55131155e-01 -1.59546101e+00
-5.50189734e-01 -5.02012253e-01 4.73316871e-02 5.86164832e-01
5.98210812e-01 -8.08748752e-02 1.01202726e+00 1.08412218e+00
-4.45243090e-01 -7.27688432e-01 -5.08824766e-01 -8.65382016e-01
-1.71154752e-01 -7.26225376e-01 4.55845416e-01 6.97764158e-01
4.55696344e-01 8.86570662e-02 -2.45814636e-01 4.70507927e-02
1.00846007e-01 -3.42146039e-01 1.05212045e+00 -7.42626429e-01
-4.43353474e-01 -7.69609690e-01 -5.76973498e-01 -1.14915919e+00
2.88399518e-01 -7.02503860e-01 3.37063938e-01 -1.64417756e+00
1.29001483e-01 -5.56420803e-01 6.98726997e-02 4.76966977e-01
1.02087162e-01 -1.48540094e-01 6.15340233e-01 3.43088597e-01
-1.00901592e+00 6.09974444e-01 1.22923076e+00 -4.69351225e-02
-3.67060393e-01 -3.67380768e-01 -1.16885595e-01 6.98303699e-01
7.52411664e-01 2.99997684e-02 -5.97718716e-01 -8.32228243e-01
9.51459110e-02 3.13277364e-01 5.68051100e-01 -1.56566691e+00
1.48491666e-01 -4.74038363e-01 3.82591695e-01 -1.04810931e-02
6.43458664e-01 -8.07490528e-01 2.38736346e-01 5.13107479e-01
-6.05002224e-01 4.48030710e-01 3.71537387e-01 7.01633096e-01
1.56513825e-01 7.52165820e-03 3.84009421e-01 -2.67036736e-01
-1.43710709e+00 2.43089169e-01 -6.66689932e-01 2.08882332e-01
1.21849132e+00 -3.51737201e-01 -5.53370059e-01 -6.23509645e-01
-7.59424210e-01 4.18723345e-01 1.09709263e+00 4.73974288e-01
4.83322471e-01 -9.07933950e-01 -4.05224681e-01 -3.12370318e-03
2.69800723e-01 -2.51232922e-01 8.20936784e-02 7.68237352e-01
-9.08611774e-01 1.72945723e-01 -5.92152894e-01 -6.12657785e-01
-9.26511109e-01 6.37051165e-01 4.08636928e-01 1.31233871e-01
-5.21385789e-01 9.52491403e-01 1.34112611e-01 -3.89131755e-01
5.40331304e-01 -7.20221341e-01 4.54015359e-02 -3.96915019e-01
1.96271196e-01 3.41785759e-01 -4.04227376e-01 -3.93350542e-01
-1.79367661e-01 -2.62136348e-02 1.99436978e-01 -1.67682290e-01
1.23974001e+00 4.21354687e-03 3.57951611e-01 3.16242725e-01
8.43686581e-01 -6.55476332e-01 -2.03406835e+00 2.34174445e-01
-2.67006278e-01 -2.81441569e-01 -9.50134620e-02 -9.60503876e-01
-7.67221808e-01 5.10497272e-01 4.63347614e-01 -2.14563459e-02
6.68238461e-01 -1.26915663e-01 3.49574238e-01 8.70737195e-01
6.70221686e-01 -9.84800100e-01 3.57321739e-01 7.11309791e-01
9.90289927e-01 -1.60259354e+00 -1.53819025e-01 -5.43183237e-02
-1.03765583e+00 8.19530070e-01 1.16812336e+00 1.61879547e-02
3.58310461e-01 2.16684684e-01 -1.99405476e-01 -1.59084350e-01
-1.09193707e+00 -9.18985009e-02 -2.60814428e-02 7.88641930e-01
8.54658112e-02 6.45361915e-02 2.67843962e-01 4.00900215e-01
-4.27402824e-01 2.17788503e-01 5.55967093e-01 9.71806705e-01
-4.39618737e-01 -4.35594559e-01 2.04768628e-02 8.80673409e-01
1.72567993e-01 2.38547903e-02 -2.17922255e-01 9.90967155e-01
-2.81760275e-01 8.30822468e-01 1.90670371e-01 -8.73693228e-01
1.40416145e-01 1.02563552e-03 6.83535397e-01 -6.12104893e-01
-5.33870518e-01 -5.19030690e-01 5.36040902e-01 -1.16545963e+00
-2.22708747e-01 -8.65094781e-01 -1.00270748e+00 -6.29542291e-01
8.66319537e-02 -2.50504464e-01 4.66412127e-01 8.84797335e-01
5.29204547e-01 4.46338773e-01 -1.93452969e-01 -1.57930505e+00
-5.07311761e-01 -8.90472293e-01 -7.63751119e-02 5.67707181e-01
8.99689049e-02 -9.32396293e-01 -1.67281911e-01 9.71496254e-02] | [4.529105186462402, 0.7856602668762207] |
9e8bde04-f746-42f6-a207-c8d1eb1f671e | a-gold-anaphora-annotation-layer-on-an-eye | null | null | https://aclanthology.org/L18-1555 | https://aclanthology.org/L18-1555.pdf | A Gold Anaphora Annotation Layer on an Eye Movement Corpus | null | ['Pascal Amsili', 'Olga Seminck'] | 2018-05-01 | a-gold-anaphora-annotation-layer-on-an-eye-1 | https://aclanthology.org/L18-1555 | https://aclanthology.org/L18-1555.pdf | lrec-2018-5 | ['abstract-anaphora-resolution'] | ['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] | [-7.439014434814453, 3.6454989910125732] |
a4148790-8d84-440b-b963-f0ecc1baa98a | semantic-enhanced-image-clustering | 2208.09849 | null | https://arxiv.org/abs/2208.09849v2 | https://arxiv.org/pdf/2208.09849v2.pdf | Semantic-Enhanced Image Clustering | Image clustering is an important and open-challenging task in computer vision. Although many methods have been proposed to solve the image clustering task, they only explore images and uncover clusters according to the image features, thus being unable to distinguish visually similar but semantically different images. In this paper, we propose to investigate the task of image clustering with the help of a visual-language pre-training model. Different from the zero-shot setting, in which the class names are known, we only know the number of clusters in this setting. Therefore, how to map images to a proper semantic space and how to cluster images from both image and semantic spaces are two key problems. To solve the above problems, we propose a novel image clustering method guided by the visual-language pre-training model CLIP, named \textbf{Semantic-Enhanced Image Clustering (SIC)}. In this new method, we propose a method to map the given images to a proper semantic space first and efficient methods to generate pseudo-labels according to the relationships between images and semantics. Finally, we propose performing clustering with consistency learning in both image space and semantic space, in a self-supervised learning fashion. The theoretical result of convergence analysis shows that our proposed method can converge at a sublinear speed. Theoretical analysis of expectation risk also shows that we can reduce the expected risk by improving neighborhood consistency, increasing prediction confidence, or reducing neighborhood imbalance. Experimental results on five benchmark datasets clearly show the superiority of our new method. | ['Longteng Chen', 'Qin Zhang', 'Xiaojun Chen', 'Liping Qiu', 'Shaotian Cai'] | 2022-08-21 | null | null | null | null | ['image-clustering'] | ['computer-vision'] | [ 3.09890658e-01 -9.86817107e-02 -1.80831820e-01 -5.19167721e-01
-6.04508162e-01 -4.09055978e-01 2.52628148e-01 1.91325963e-01
-5.01263499e-01 1.05432764e-01 -2.31856525e-01 -1.46702483e-01
-3.84683549e-01 -6.31546557e-01 -7.52933681e-01 -8.50007296e-01
3.48789155e-01 3.95046264e-01 2.17627838e-01 2.95674324e-01
3.66096407e-01 1.55541338e-02 -1.75906253e+00 2.13576764e-01
1.08808947e+00 9.38258648e-01 5.79949856e-01 3.50680113e-01
-1.61524266e-01 6.05277658e-01 -2.53223479e-01 -1.27970800e-01
3.41655135e-01 -6.53900445e-01 -8.77870262e-01 8.19947600e-01
1.21487945e-01 2.33864471e-01 1.30120307e-01 1.58351266e+00
2.35666975e-01 2.70352691e-01 6.42782390e-01 -1.42857659e+00
-6.09105468e-01 5.08643627e-01 -7.71076202e-01 -9.28477719e-02
5.20751476e-02 -1.81287169e-01 7.95865178e-01 -7.16403961e-01
4.67903912e-01 1.03858900e+00 5.20232201e-01 6.05976582e-01
-1.27308404e+00 -5.43414354e-01 2.20150948e-01 5.00263810e-01
-1.63233769e+00 -1.32442757e-01 8.31644893e-01 -5.89152932e-01
4.58352193e-02 2.84916610e-01 3.48815143e-01 4.87952203e-01
-4.15917218e-01 7.63366818e-01 1.22944486e+00 -6.39988959e-01
5.71125209e-01 6.47376478e-01 1.74137339e-01 6.35475814e-01
1.69032127e-01 -3.49842101e-01 -5.26993312e-02 3.45101058e-02
3.45202386e-01 3.54888141e-01 -3.79250795e-01 -7.78608024e-01
-1.05859482e+00 9.11342859e-01 5.84632456e-01 3.71300936e-01
-1.24531627e-01 -1.46746933e-01 2.61890680e-01 -3.11159678e-02
3.69028687e-01 2.04633191e-01 -2.56015897e-01 5.21717668e-01
-8.64376783e-01 -1.87020600e-01 5.52944183e-01 8.68820488e-01
9.06583250e-01 -3.99473488e-01 7.22229108e-02 8.92886579e-01
1.99212432e-01 3.35566223e-01 5.50251424e-01 -1.01095581e+00
2.56625861e-01 6.95916533e-01 -7.03349859e-02 -1.34708297e+00
-3.59656513e-01 -2.65746742e-01 -1.19116807e+00 7.51015842e-02
5.26821375e-01 2.18069240e-01 -8.58930647e-01 1.85605311e+00
4.99676794e-01 3.19174558e-01 1.96663216e-01 9.44398820e-01
4.66836929e-01 6.88062787e-01 6.62220940e-02 -4.01403874e-01
1.28836656e+00 -9.00180578e-01 -6.30102158e-01 -9.79449376e-02
5.75598061e-01 -5.31454563e-01 1.09634876e+00 3.89169276e-01
-7.61194587e-01 -6.76641285e-01 -8.82870436e-01 3.40751231e-01
-3.20139378e-01 2.77391583e-01 3.42885703e-01 7.45605886e-01
-8.93553674e-01 4.01720285e-01 -6.93910122e-01 -4.56618071e-01
4.93439287e-01 3.06407958e-01 -2.76470870e-01 -2.43333757e-01
-8.43511224e-01 5.42714179e-01 9.09221053e-01 -2.88116410e-02
-5.25587142e-01 -4.46349531e-01 -8.82703304e-01 1.08345583e-01
7.15654492e-01 -4.15230423e-01 6.59058928e-01 -1.45977175e+00
-9.59045708e-01 8.84888113e-01 -2.70990431e-01 -3.53997558e-01
3.99715036e-01 2.34031215e-01 -1.40573472e-01 3.79186690e-01
3.29298586e-01 7.45104611e-01 9.63349283e-01 -1.73570859e+00
-8.89542758e-01 -4.80219275e-01 -2.55659461e-01 2.72375822e-01
-5.74374080e-01 -1.79249734e-01 -1.02173340e+00 -4.62005913e-01
2.42761850e-01 -1.04398835e+00 -6.06763124e-01 -1.53830349e-01
-6.07231796e-01 -2.20820427e-01 7.34242141e-01 -5.72693765e-01
1.09469724e+00 -2.33549976e+00 4.20805775e-02 6.90208733e-01
1.94352731e-01 2.43801609e-01 1.14256479e-01 -1.35063067e-01
-1.22824341e-01 2.00355932e-01 -6.02609396e-01 -3.01269829e-01
-2.06482649e-01 1.85421973e-01 -6.27267212e-02 4.90759820e-01
-1.81477770e-01 6.57518089e-01 -8.03565741e-01 -8.73522162e-01
3.70929867e-01 2.99988002e-01 -5.61961770e-01 8.59882683e-02
1.24924257e-02 5.07824063e-01 -4.06499237e-01 3.47475410e-01
7.31603861e-01 -6.80336833e-01 3.75511527e-01 -2.16093644e-01
2.46251866e-01 -5.54252684e-01 -1.46276700e+00 1.43155670e+00
-3.48199993e-01 2.08852738e-01 -1.76779568e-01 -1.60236692e+00
8.42027068e-01 3.51749174e-03 5.14330804e-01 -6.51395380e-01
1.63685307e-01 -2.43741516e-02 -2.39929020e-01 -4.90735501e-01
1.48723453e-01 -2.30104715e-01 -3.13976258e-02 3.87554675e-01
-1.95213377e-01 2.43670642e-01 2.39410087e-01 2.88930148e-01
5.56571722e-01 -3.61169964e-01 1.60427317e-01 -1.68823063e-01
7.08068192e-01 1.04296915e-01 6.56159461e-01 7.45028734e-01
-2.04500467e-01 8.51580560e-01 3.46374422e-01 -1.64719135e-01
-9.33685899e-01 -8.81506920e-01 -1.54911369e-01 7.36327529e-01
6.91331267e-01 -3.75251055e-01 -1.21731353e+00 -7.83937812e-01
-2.41542861e-01 4.96117651e-01 -6.64920092e-01 -2.48331770e-01
-2.61341989e-01 -9.25695419e-01 -9.71777085e-03 2.90168226e-01
5.75306714e-01 -9.25013006e-01 -4.79016393e-01 -7.43417367e-02
-4.74244654e-01 -1.18256330e+00 -5.36441505e-01 5.94636537e-02
-7.00689852e-01 -1.19067883e+00 -6.61426961e-01 -1.20118296e+00
1.14415193e+00 5.77101886e-01 7.15745926e-01 2.04113200e-01
-3.98725212e-01 3.65792245e-01 -4.72717285e-01 -8.90393108e-02
-3.78140599e-01 -1.49684817e-01 -7.22089335e-02 6.45085812e-01
3.16645145e-01 -2.44527474e-01 -6.53809726e-01 4.32258815e-01
-1.08623207e+00 1.65554762e-01 6.13611042e-01 6.98685825e-01
9.01246369e-01 7.70877540e-01 3.50338221e-01 -9.93731797e-01
1.78629443e-01 -5.37056863e-01 -6.14143431e-01 5.28130233e-01
-8.69940639e-01 1.94486603e-01 7.29933083e-01 -3.38211745e-01
-9.20751810e-01 7.58055747e-01 3.48014712e-01 -7.03470767e-01
-2.59166002e-01 3.28442097e-01 -4.78955984e-01 1.33229628e-01
3.21082413e-01 4.51103359e-01 2.57890433e-01 -3.44992638e-01
6.56871617e-01 6.96971238e-01 7.44691074e-01 -3.32419336e-01
8.62362027e-01 8.71485591e-01 -2.12405950e-01 -6.24748707e-01
-9.15893555e-01 -9.37313020e-01 -7.07953155e-01 -2.62638718e-01
1.21043205e+00 -8.36121202e-01 -6.96402013e-01 1.60003915e-01
-7.67020047e-01 4.99826781e-02 -9.82878357e-02 3.84398669e-01
-6.49788916e-01 7.45781898e-01 -1.28523618e-01 -7.64646649e-01
-7.46878758e-02 -1.18555748e+00 7.87335932e-01 3.06407899e-01
1.62490577e-01 -1.00892711e+00 -2.34883353e-01 5.10970175e-01
-8.42384920e-02 2.27494314e-01 9.27380681e-01 -5.52932918e-01
-4.93378252e-01 1.33108767e-02 -5.09329140e-01 5.54992080e-01
2.49940604e-01 -3.34906101e-01 -6.79182351e-01 -1.80158243e-01
2.90740430e-01 -4.39278409e-02 9.07121599e-01 5.60072482e-01
1.56537557e+00 -3.77122313e-01 -4.96092886e-01 6.50217235e-01
1.73216665e+00 3.65324050e-01 5.71051657e-01 3.35753679e-01
8.90539169e-01 1.00978947e+00 7.40527749e-01 3.31321508e-01
3.94170940e-01 6.05778456e-01 2.82950491e-01 -3.31277728e-01
2.09171083e-02 -1.82330862e-01 1.40100494e-01 7.01402903e-01
2.80961156e-01 -3.33353765e-02 -7.85163701e-01 6.27899885e-01
-2.09913111e+00 -9.52245295e-01 -2.09524900e-01 2.26401258e+00
7.35750675e-01 -9.80997086e-02 1.01897739e-01 2.65107602e-01
1.30720735e+00 -3.54590088e-01 -5.90160191e-01 3.05319410e-02
-7.15126395e-02 -1.90487415e-01 5.79489052e-01 4.13368374e-01
-1.30673504e+00 8.20396066e-01 5.26199341e+00 1.12787890e+00
-9.15513217e-01 2.12373927e-01 1.10906363e+00 2.24739105e-01
2.66150162e-02 1.52298898e-01 -6.45238996e-01 7.45156229e-01
5.06146193e-01 -1.80335790e-01 4.85610545e-01 9.96657789e-01
1.63966373e-01 -2.08040297e-01 -9.66338217e-01 1.34843910e+00
4.45673496e-01 -1.08172119e+00 1.59162387e-01 -8.39050859e-03
8.40149641e-01 -4.76734042e-01 1.25008538e-01 -1.09524071e-01
2.47556776e-01 -9.06726837e-01 6.83394492e-01 3.86556029e-01
5.83672464e-01 -9.31234479e-01 5.56850731e-01 5.39707959e-01
-1.13535154e+00 -2.56188929e-01 -4.84942138e-01 3.08558881e-01
-1.48461461e-02 6.10968769e-01 -6.53816760e-01 5.56355000e-01
9.43392754e-01 8.55059505e-01 -7.70687699e-01 1.02374434e+00
-9.61932261e-03 5.15441060e-01 -1.20249450e-01 2.31044129e-01
2.30868250e-01 -4.75545079e-01 7.72249848e-02 9.44635987e-01
2.34049186e-01 2.05426708e-01 4.09633666e-01 8.45460713e-01
1.43250942e-01 2.71610260e-01 -4.06937718e-01 2.75978386e-01
2.30454415e-01 1.27897370e+00 -1.41494024e+00 -4.70372766e-01
-2.66413242e-01 1.16338563e+00 1.54238909e-01 3.33189547e-01
-8.18129957e-01 -3.69951993e-01 2.37659946e-01 1.30141582e-02
4.35614884e-01 1.15869842e-01 -4.64530796e-01 -1.04540980e+00
1.89578354e-01 -6.16771936e-01 6.17527008e-01 -5.56874275e-01
-1.23877048e+00 4.71968472e-01 3.65925245e-02 -1.40396452e+00
5.13499826e-02 -3.29647034e-01 -3.52483779e-01 3.86384636e-01
-1.39396679e+00 -8.18603694e-01 -3.71750355e-01 9.41768169e-01
5.17979145e-01 3.13912630e-02 4.60627973e-01 2.92621076e-01
-5.29495060e-01 4.69816685e-01 4.73449260e-01 2.51356184e-01
5.59407949e-01 -1.25457120e+00 -1.91830114e-01 8.27736497e-01
3.00753027e-01 4.19068664e-01 5.02599299e-01 -3.95949781e-01
-9.63520050e-01 -1.42124569e+00 7.21944809e-01 -3.54520649e-01
2.47197986e-01 -4.53249484e-01 -1.01753414e+00 2.65279979e-01
-6.24044053e-03 5.71846925e-02 5.40914237e-01 -2.04757020e-01
-2.97424942e-01 -2.48064995e-01 -1.10460961e+00 4.84679371e-01
8.44910681e-01 -3.09064656e-01 -4.87747401e-01 5.08627057e-01
6.70753419e-01 8.33095163e-02 -5.31822979e-01 2.93483615e-01
8.55633244e-02 -8.96503747e-01 1.03648734e+00 -3.46636742e-01
2.56983906e-01 -6.82614565e-01 -1.25445470e-01 -1.14782894e+00
-1.76884100e-01 -1.69177935e-01 4.09276187e-01 1.39608443e+00
1.74816087e-01 -3.83995891e-01 8.05817783e-01 5.04428864e-01
1.92825422e-01 -4.24714208e-01 -6.96599841e-01 -9.76167023e-01
-1.94821581e-01 -4.50098068e-01 2.78792918e-01 1.11239135e+00
-9.16986242e-02 2.91187346e-01 -3.75764310e-01 3.79481941e-01
9.56952572e-01 2.96860546e-01 5.80822289e-01 -1.15235376e+00
-2.66025096e-01 -4.90370274e-01 -5.32065988e-01 -6.53128028e-01
3.71974498e-01 -1.07571554e+00 3.71586800e-01 -1.49759912e+00
8.44386756e-01 -7.44215786e-01 -4.18485492e-01 3.50124866e-01
-4.27690089e-01 4.37814802e-01 3.00853699e-01 3.66493016e-01
-1.03084207e+00 3.86427581e-01 7.80806243e-01 -2.82088786e-01
-2.21569449e-01 -4.51838709e-02 -8.31306934e-01 7.11847901e-01
7.92206228e-01 -6.43922389e-01 -5.96801877e-01 -7.71500915e-02
-6.59128353e-02 -1.77394763e-01 4.26059633e-01 -1.03617263e+00
3.89199436e-01 -6.99105933e-02 2.89141715e-01 -4.26489919e-01
-7.40979463e-02 -1.02257895e+00 1.26400143e-01 5.23709536e-01
-5.00066280e-01 -2.97222614e-01 -1.73604980e-01 9.35687602e-01
-3.56249750e-01 -3.19889426e-01 1.06815255e+00 -1.42878920e-01
-9.42460120e-01 2.56749302e-01 -2.31307417e-01 1.11999944e-01
1.41544557e+00 -2.34135136e-01 9.65000391e-02 -3.69960338e-01
-9.68610406e-01 3.87632549e-01 5.60289085e-01 4.16752756e-01
6.88517749e-01 -1.26480210e+00 -5.14245987e-01 8.08253065e-02
3.43672991e-01 4.96066995e-02 4.34379429e-01 8.22879732e-01
-2.41177961e-01 2.23748103e-01 2.86524352e-02 -1.02097535e+00
-1.46816480e+00 1.20004094e+00 1.27928600e-01 -4.26010694e-03
-6.31090820e-01 5.05404770e-01 7.16449499e-01 -4.29278433e-01
3.86241585e-01 -8.31752941e-02 -3.59833181e-01 1.63748837e-03
5.56948662e-01 1.41514108e-01 -9.36841220e-02 -7.77794242e-01
-2.92906940e-01 8.98790181e-01 2.26364024e-02 1.44116636e-02
1.04750645e+00 -5.61313212e-01 -2.09271833e-01 3.56584698e-01
1.49862587e+00 -3.35514694e-01 -1.03910446e+00 -4.00792301e-01
3.77871305e-01 -4.84208286e-01 -1.09288275e-01 -4.01328683e-01
-1.37283826e+00 7.35002458e-01 9.09639060e-01 2.07281128e-01
1.45951045e+00 3.31531078e-01 6.65168524e-01 1.80558339e-01
1.99518502e-01 -1.23677862e+00 2.21132979e-01 2.81085586e-03
2.59591222e-01 -1.53272462e+00 -2.06438541e-01 -5.37273407e-01
-9.92561102e-01 7.38657475e-01 4.92291659e-01 -2.76193209e-02
7.92127132e-01 -2.35656276e-01 9.94435847e-02 -8.75210613e-02
-3.54624718e-01 -5.42128205e-01 3.56470585e-01 5.90373755e-01
-1.79127544e-01 4.78840955e-02 -1.61999702e-01 5.29115379e-01
4.31122109e-02 -2.67521441e-01 3.40712994e-01 5.81077933e-01
-5.33873081e-01 -9.77522254e-01 -5.90111315e-01 3.05812567e-01
-2.96961427e-01 1.25300154e-01 -3.02324682e-01 5.58355689e-01
3.07894886e-01 1.22796881e+00 1.93966359e-01 -4.60828036e-01
-1.77611057e-02 -1.51592895e-01 1.57667249e-01 -4.70759660e-01
-7.25760087e-02 1.17548913e-01 -5.61346591e-01 -4.27930504e-01
-7.28991210e-01 -5.84999323e-01 -1.52744722e+00 1.33116931e-01
-4.98108357e-01 4.33008701e-01 7.41407573e-01 9.67989087e-01
2.37048611e-01 2.05830574e-01 1.11408806e+00 -5.29149890e-01
-2.17041016e-01 -3.38507235e-01 -6.98516607e-01 8.31513882e-01
5.77547476e-02 -4.42171633e-01 -5.78552544e-01 4.97288346e-01] | [9.42833423614502, 2.9814491271972656] |
793ae31d-305d-4b67-bb93-c64c6a7dd61c | devolutionary-genetic-algorithms-with | 2004.10048 | null | https://arxiv.org/abs/2004.10048v1 | https://arxiv.org/pdf/2004.10048v1.pdf | Devolutionary genetic algorithms with application to the minimum labeling Steiner tree problem | This paper characterizes and discusses devolutionary genetic algorithms and evaluates their performances in solving the minimum labeling Steiner tree (MLST) problem. We define devolutionary algorithms as the process of reaching a feasible solution by devolving a population of super-optimal unfeasible solutions over time. We claim that distinguishing them from the widely used evolutionary algorithms is relevant. The most important distinction lies in the fact that in the former type of processes, the value function decreases over successive generation of solutions, thus providing a natural stopping condition for the computation process. We show how classical evolutionary concepts, such as crossing, mutation and fitness can be adapted to aim at reaching an optimal or close-to-optimal solution among the first generations of feasible solutions. We additionally introduce a novel integer linear programming formulation of the MLST problem and a valid constraint used for speeding up the devolutionary process. Finally, we conduct an experiment comparing the performances of devolutionary algorithms to those of state of the art approaches used for solving randomly generated instances of the MLST problem. Results of this experiment support the use of devolutionary algorithms for the MLST problem and their development for other NP-hard combinatorial optimization problems. | ['Nassim Dehouche'] | 2020-04-18 | null | null | null | null | ['steiner-tree-problem'] | ['graphs'] | [ 6.67530358e-01 3.39922398e-01 -7.70753771e-02 -1.98988989e-01
-1.43552363e-01 -5.27801037e-01 1.19899269e-02 4.91383404e-01
-5.19317150e-01 1.11423886e+00 -4.75433499e-01 -2.77662545e-01
-1.00894094e+00 -1.09718549e+00 -3.83647859e-01 -8.61999631e-01
-1.90944657e-01 9.93362248e-01 2.41673633e-01 -3.98344815e-01
4.90218729e-01 6.08349323e-01 -1.86970687e+00 1.81749035e-02
1.25484943e+00 6.67232275e-01 2.79634327e-01 3.08877766e-01
-5.16549289e-01 -2.78286189e-01 -9.26235020e-01 -4.94049430e-01
3.54984015e-01 -5.23726463e-01 -8.73846173e-01 2.33253390e-01
-2.53680527e-01 6.22075140e-01 6.39224648e-01 1.01633108e+00
4.31816310e-01 5.10495119e-02 5.97215950e-01 -1.68543768e+00
-2.59352267e-01 8.00491214e-01 -7.29102015e-01 1.78364992e-01
4.99396145e-01 -1.05136499e-01 8.55271220e-01 -2.24705473e-01
9.13895071e-01 1.01272869e+00 5.01572013e-01 4.74962384e-01
-1.24854100e+00 -8.84781927e-02 2.10513547e-01 2.70291597e-01
-1.43195045e+00 -2.61254460e-02 4.92797643e-01 -4.85469699e-02
1.06817508e+00 1.02353430e+00 1.01488066e+00 3.44487190e-01
2.44063601e-01 5.83387077e-01 8.83889139e-01 -8.47682059e-01
7.05820441e-01 1.05350792e-01 -5.03149964e-02 5.04317462e-01
9.47319567e-01 9.43795890e-02 -7.47509906e-03 -2.91343212e-01
4.26641032e-02 -5.11869729e-01 -3.57115865e-01 -7.87682891e-01
-7.78246343e-01 1.03284073e+00 1.87394202e-01 6.23424828e-01
-5.13162494e-01 -6.46271035e-02 1.12474777e-01 2.68289030e-01
6.34202361e-02 9.02140141e-01 -3.75185490e-01 -7.74069205e-02
-9.58823085e-01 1.96694270e-01 1.01301444e+00 7.27450132e-01
3.18048000e-01 -1.59189790e-01 -7.06504434e-02 5.68697453e-01
1.46944359e-01 6.11415468e-02 1.91480324e-01 -8.62129033e-01
3.77911717e-01 9.12802875e-01 6.04584925e-02 -8.97358239e-01
-2.22000942e-01 -6.77479446e-01 -4.08442527e-01 4.98534918e-01
1.43943205e-01 -1.87877730e-01 -6.52090847e-01 1.39563465e+00
5.00481367e-01 -1.98989049e-01 1.26091838e-01 5.45979083e-01
2.34269172e-01 7.92873442e-01 -3.01537365e-01 -1.21920538e+00
7.39349186e-01 -9.97378409e-01 -5.10586083e-01 2.55901217e-01
5.63344777e-01 -6.13662899e-01 2.86642671e-01 6.41974211e-01
-1.44945383e+00 -6.73540682e-02 -1.03217173e+00 7.35800385e-01
-3.54321480e-01 -3.02326351e-01 6.07375383e-01 1.20039320e+00
-1.15109468e+00 4.93421733e-01 -4.81865108e-01 -4.15317267e-01
1.18543535e-01 7.50731766e-01 1.22903846e-01 1.34206573e-02
-4.65153098e-01 8.55978012e-01 7.38571882e-01 2.20623285e-01
-1.17298551e-01 -3.95646691e-01 -4.55860376e-01 2.95798212e-01
8.47242534e-01 -7.99407303e-01 7.68782914e-01 -1.08195746e+00
-1.15503061e+00 8.13784122e-01 -1.68839097e-01 -3.86648178e-01
5.51417053e-01 7.19486833e-01 -2.52643794e-01 -3.12571079e-02
-1.61043465e-01 5.37939548e-01 5.48428416e-01 -1.39117801e+00
-9.08951223e-01 -1.32280573e-01 1.34950936e-01 1.06552586e-01
-2.74382532e-01 -1.91594195e-02 -2.09110498e-01 -4.01558369e-01
2.47985587e-01 -8.20943236e-01 -6.91306829e-01 -4.09155816e-01
-1.69419393e-01 -3.00515950e-01 5.08959711e-01 -1.07899472e-01
1.58934677e+00 -1.65724325e+00 8.72986972e-01 7.54492044e-01
-1.60757303e-01 3.46191913e-01 -2.99029469e-01 8.18499684e-01
-1.91969872e-01 2.81511724e-01 -5.56122601e-01 6.16106912e-02
1.25480309e-01 5.79927325e-01 1.84232920e-01 1.33365646e-01
-2.64793728e-02 7.67552257e-01 -8.78509879e-01 -4.33702946e-01
-5.14959060e-02 1.09110847e-02 -5.69362104e-01 -4.38181847e-01
-4.99610484e-01 2.85565630e-02 -4.04509038e-01 6.19788468e-01
7.67836511e-01 3.41021538e-01 4.00904298e-01 2.61319876e-01
-4.73477066e-01 -3.45121384e-01 -1.42361009e+00 1.35686958e+00
-1.48831427e-01 2.14120179e-01 -2.40328219e-02 -1.35545862e+00
1.08276320e+00 6.11454062e-02 7.91633189e-01 -6.64001882e-01
4.44855243e-01 3.56332093e-01 1.41811728e-01 -4.79035109e-01
6.56918466e-01 1.62935499e-02 7.57793412e-02 6.05925560e-01
-4.01638806e-01 -3.44684213e-01 1.04923105e+00 -1.47391930e-01
8.25754642e-01 -5.07363640e-02 1.77358523e-01 -3.58685315e-01
7.94805110e-01 2.54785597e-01 7.33655870e-01 4.72330540e-01
2.01175198e-01 1.89987749e-01 7.27826715e-01 -3.50375533e-01
-9.54413414e-01 -7.82974184e-01 -4.50108014e-02 4.11523908e-01
1.78881228e-01 -2.72384971e-01 -8.04701805e-01 -6.58659756e-01
-1.00217022e-01 1.05787337e+00 -4.22632426e-01 -2.00195849e-01
-6.31856382e-01 -1.29591334e+00 1.18431337e-02 -1.01961546e-01
1.27066419e-01 -7.96169043e-01 -1.30353343e+00 5.74106276e-01
-2.79823318e-02 -5.17528296e-01 3.29835713e-01 2.71095991e-01
-9.96321797e-01 -1.01359582e+00 -5.20235658e-01 -7.93663561e-01
1.04590130e+00 3.22051048e-02 1.00875187e+00 4.99411851e-01
-6.66561484e-01 1.05014831e-01 -5.59377193e-01 -3.56880873e-01
-4.95282412e-01 1.74816817e-01 -3.39272380e-01 -2.66936868e-01
1.45685896e-01 -5.60322225e-01 -1.00494754e-02 4.65364695e-01
-8.00368190e-01 -2.13731036e-01 3.45312864e-01 4.87048239e-01
7.80287921e-01 9.55984294e-01 3.26929212e-01 -4.58792061e-01
9.73426998e-01 -3.30795169e-01 -1.09024835e+00 8.11689436e-01
-8.48160505e-01 4.13820922e-01 2.24871144e-01 -3.75972658e-01
-7.67090976e-01 3.94209987e-03 -2.16566492e-02 -6.25419840e-02
3.00637722e-01 4.51894015e-01 -9.70021710e-02 -4.67657447e-01
2.50803471e-01 6.87215626e-02 -1.45728469e-01 -2.15143755e-01
2.54569035e-02 3.13410431e-01 1.94635749e-01 -7.30468035e-01
6.34966791e-01 2.13809967e-01 5.41024983e-01 -3.75940502e-01
-2.64389932e-01 -2.03685403e-01 -3.95080388e-01 -3.76891106e-01
5.25072157e-01 3.25839788e-01 -7.94008553e-01 1.08765505e-01
-1.01477838e+00 2.08335370e-01 -8.62969577e-01 -4.07980308e-02
-6.65427685e-01 3.44367325e-01 2.67232686e-01 -1.13294363e+00
-1.86269686e-01 -1.09204936e+00 5.49651146e-01 3.44543517e-01
-3.12669486e-01 -7.96531975e-01 1.70275331e-01 2.01674566e-01
1.80001095e-01 8.94281209e-01 1.14168382e+00 -1.74120262e-01
-3.92139375e-01 -1.06256343e-01 3.76781434e-01 -2.22347662e-01
-1.69770122e-01 4.66361910e-01 -8.54053274e-02 -4.42489564e-01
1.28429696e-01 2.70677656e-01 5.15652120e-01 4.15822536e-01
5.67027330e-01 -3.43927920e-01 -7.81335354e-01 4.86720473e-01
1.93742526e+00 8.32528353e-01 6.04118466e-01 6.82312429e-01
-2.03006729e-01 9.02885854e-01 1.07273161e+00 5.17732561e-01
4.95407917e-02 8.16833496e-01 7.19733536e-01 2.36078486e-01
3.39383125e-01 2.88991690e-01 -1.64754018e-01 2.98189908e-01
-3.05512667e-01 -8.13082218e-01 -7.92429566e-01 5.79338670e-01
-1.90037262e+00 -9.51928079e-01 -5.00369728e-01 2.24411726e+00
3.32458317e-01 1.48379058e-01 2.65315801e-01 8.48134339e-01
1.00030422e+00 -3.79035741e-01 -2.45602235e-01 -1.41082954e+00
-1.76167995e-01 4.61981773e-01 3.97436917e-01 5.39911151e-01
-5.71843207e-01 6.20091081e-01 6.35152102e+00 6.82965040e-01
-1.00968421e+00 5.24537787e-02 3.31744581e-01 -4.59834933e-01
-5.54902554e-01 -2.16570646e-02 -5.54442346e-01 5.53262651e-01
6.76228881e-01 -5.59183657e-01 6.01082921e-01 2.97564179e-01
1.30863890e-01 -5.40870130e-01 -6.89565063e-01 5.36179900e-01
3.52236703e-02 -1.26198971e+00 -8.74939933e-02 3.17617655e-01
1.26361990e+00 -6.82197869e-01 1.08943991e-02 -2.69545764e-01
2.24512070e-01 -8.75686228e-01 7.70151436e-01 2.15830445e-01
2.27055341e-01 -1.27223957e+00 7.07521439e-01 3.69913816e-01
-9.68523204e-01 -5.94735324e-01 -4.02737260e-01 1.32816091e-01
6.61064804e-01 4.95912671e-01 -5.31355619e-01 1.21779978e+00
6.59309685e-01 -1.13244727e-01 -2.45711267e-01 1.77939975e+00
-2.63014942e-01 -8.59510377e-02 -5.65170467e-01 -3.45252156e-01
6.13155663e-01 -5.15197814e-01 8.68371904e-01 6.68015540e-01
6.85305595e-01 2.68041436e-02 -3.08002055e-01 9.33464408e-01
6.80837095e-01 3.01317573e-01 -2.57672578e-01 -1.21827260e-01
4.59184259e-01 9.81144547e-01 -1.29265296e+00 1.70504749e-01
2.16442719e-01 7.58906066e-01 -8.94936845e-02 2.93503515e-02
-9.75858748e-01 -3.63876343e-01 2.06044421e-01 -1.36811510e-01
5.92314243e-01 -2.68319827e-02 -4.57210690e-01 -4.79449838e-01
8.08142647e-02 -6.64628804e-01 6.23360097e-01 -6.35290623e-01
-5.66310287e-01 9.00649965e-01 2.98480749e-01 -8.12863529e-01
-4.10711318e-01 -3.49944115e-01 -7.28135467e-01 7.10316420e-01
-1.20859635e+00 -6.29547894e-01 -1.15660943e-01 2.84752309e-01
4.45941806e-01 -1.75838545e-01 5.63358665e-01 5.03784642e-02
-6.75123155e-01 2.65666485e-01 1.29304856e-01 -1.00534904e+00
-2.17682436e-01 -9.22713041e-01 -2.14356906e-03 1.00537121e+00
-2.18372241e-01 3.25131893e-01 1.07998669e+00 -6.03032887e-01
-1.18053567e+00 -4.41896230e-01 1.10382712e+00 2.00092241e-01
3.36953193e-01 6.55442700e-02 -3.38134199e-01 5.40084690e-02
2.19854802e-01 -7.48588741e-01 4.31963652e-01 -1.27006620e-01
5.40400624e-01 -3.80782858e-02 -1.66472507e+00 6.04271770e-01
1.19404936e+00 7.20498264e-01 -4.89799917e-01 9.42842811e-02
4.34398562e-01 -1.08008154e-01 -4.83081549e-01 5.43794215e-01
2.72127479e-01 -1.09004259e+00 9.50986981e-01 -2.84978658e-01
1.73297554e-01 -2.82199323e-01 1.60486937e-01 -1.33592987e+00
-3.55772108e-01 -9.22461927e-01 9.63734388e-02 1.19113910e+00
3.65868956e-01 -8.26047182e-01 8.30405772e-01 2.94878811e-01
-8.62690881e-02 -1.10609329e+00 -1.16093671e+00 -1.12410045e+00
1.20586455e-02 3.56482826e-02 9.52906430e-01 8.00665021e-01
-6.42943829e-02 -1.52288541e-01 1.38831764e-01 -4.65025455e-02
6.62810385e-01 4.67809767e-01 2.48334989e-01 -1.45135689e+00
-4.99929994e-01 -6.60803437e-01 -4.54574376e-01 -1.05484419e-01
-2.31668539e-02 -7.82746732e-01 -2.72405833e-01 -1.83453870e+00
-1.24002956e-01 -8.23828518e-01 7.00293155e-03 2.59172380e-01
2.84314603e-01 1.13117501e-01 3.81409436e-01 -3.18307668e-01
-2.86961675e-01 1.51328415e-01 1.11079931e+00 -7.39031881e-02
-4.06251043e-01 2.00373128e-01 -6.84648216e-01 4.41407204e-01
8.05948853e-01 -8.22181940e-01 -6.57469094e-01 -3.72745603e-01
6.92281127e-01 9.02436152e-02 -2.14696646e-01 -8.34758699e-01
1.28603563e-01 -4.45256770e-01 -2.14080751e-01 -5.60739756e-01
1.68586850e-01 -1.14319599e+00 8.61926675e-01 1.15513361e+00
4.89346869e-02 3.25556636e-01 1.61531240e-01 1.60579458e-01
-1.61236823e-02 -1.20904052e+00 6.25366211e-01 -5.27912043e-02
-6.79806411e-01 -1.45657778e-01 -5.16769648e-01 -2.56441474e-01
1.87423873e+00 -1.24304390e+00 -1.94073066e-01 4.69060838e-02
-8.35782409e-01 4.66372877e-01 5.81316888e-01 2.86393285e-01
4.56902921e-01 -7.86801159e-01 -6.53831542e-01 2.73696929e-02
-3.24759901e-01 -1.98054209e-01 -8.21892694e-02 8.62290859e-01
-8.30380499e-01 3.82143080e-01 -5.88966787e-01 -3.67958993e-01
-1.67585194e+00 9.62364078e-01 3.30302060e-01 -2.62074590e-01
-1.96117714e-01 8.81954849e-01 -5.99075735e-01 -2.25869957e-02
1.44959673e-01 -1.47578180e-01 -2.31836736e-01 2.67861903e-01
-5.03137670e-02 8.08225751e-01 6.28367215e-02 -3.27904284e-01
-4.58973765e-01 7.64275491e-01 3.68597567e-01 -1.98400229e-01
1.68816757e+00 -7.24083483e-02 -6.35808885e-01 -1.46025315e-01
6.24889433e-01 5.83791770e-02 -5.23281574e-01 4.08985794e-01
2.33608603e-01 -5.02243936e-01 -1.12709455e-01 -9.19572532e-01
-1.19243979e+00 2.64769822e-01 4.16008800e-01 6.13790989e-01
1.73315430e+00 -2.38278657e-01 5.69063962e-01 2.47952268e-01
8.02043676e-01 -1.14186549e+00 -3.80806506e-01 1.55520111e-01
9.09107625e-01 -3.32007378e-01 1.41924620e-01 -1.00619459e+00
-2.73983389e-01 1.27596986e+00 2.77122706e-01 6.40638173e-02
-4.35651951e-02 6.32951081e-01 -4.78682518e-01 -4.30045277e-02
-7.85985053e-01 -2.33079687e-01 -4.81557809e-02 5.61556101e-01
-1.30966783e-01 -1.66179672e-01 -1.53768456e+00 1.47730052e-01
-3.53086531e-01 -6.24042042e-02 5.66655576e-01 1.33071494e+00
-6.39235258e-01 -1.70511937e+00 -6.70710266e-01 -2.95247789e-02
-1.52190641e-01 2.05272913e-01 -7.70373344e-01 8.66599917e-01
7.47304082e-01 1.09022367e+00 1.35975450e-01 7.93077052e-02
2.95209795e-01 -2.71902680e-01 1.09547448e+00 -3.38104367e-01
-8.59164298e-01 -9.59341303e-02 3.21788818e-01 -3.19435596e-01
-4.77529436e-01 -8.82495821e-01 -1.19589210e+00 -3.58937562e-01
-5.42679071e-01 7.72904694e-01 8.25747013e-01 6.72556520e-01
2.55482737e-02 7.21613169e-01 6.32323086e-01 -4.38917339e-01
-2.55027652e-01 -3.96847278e-02 -2.10578471e-01 -3.22823882e-01
-4.30691510e-01 -6.40588105e-01 -3.78771722e-02 -3.20288181e-01] | [5.746589183807373, 3.6707987785339355] |
5df6ede4-bfb4-4cbf-8478-eaab69c31305 | location-free-human-pose-estimation | 2205.12619 | null | https://arxiv.org/abs/2205.12619v1 | https://arxiv.org/pdf/2205.12619v1.pdf | Location-free Human Pose Estimation | Human pose estimation (HPE) usually requires large-scale training data to reach high performance. However, it is rather time-consuming to collect high-quality and fine-grained annotations for human body. To alleviate this issue, we revisit HPE and propose a location-free framework without supervision of keypoint locations. We reformulate the regression-based HPE from the perspective of classification. Inspired by the CAM-based weakly-supervised object localization, we observe that the coarse keypoint locations can be acquired through the part-aware CAMs but unsatisfactory due to the gap between the fine-grained HPE and the object-level localization. To this end, we propose a customized transformer framework to mine the fine-grained representation of human context, equipped with the structural relation to capture subtle differences among keypoints. Concretely, we design a Multi-scale Spatial-guided Context Encoder to fully capture the global human context while focusing on the part-aware regions and a Relation-encoded Pose Prototype Generation module to encode the structural relations. All these works together for strengthening the weak supervision from image-level category labels on locations. Our model achieves competitive performance on three datasets when only supervised at a category-level and importantly, it can achieve comparable results with fully-supervised methods with only 25\% location labels on MS-COCO and MPII. | ['Qi Zou', 'Xue Lin', 'Ke Yan', 'Yingguo Gao', 'Xixia Xu'] | 2022-05-25 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Xu_Location-Free_Human_Pose_Estimation_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Xu_Location-Free_Human_Pose_Estimation_CVPR_2022_paper.pdf | cvpr-2022-1 | ['weakly-supervised-object-localization'] | ['computer-vision'] | [ 5.96055463e-02 -2.14298274e-02 -2.80066520e-01 -4.07063872e-01
-8.77554119e-01 -2.96463609e-01 3.20287645e-01 -5.35384566e-02
-4.56678867e-01 4.57252413e-01 3.79581749e-01 3.74761641e-01
-3.47666293e-02 -6.19465292e-01 -9.14090753e-01 -4.52494323e-01
2.83613540e-02 4.74492580e-01 5.59930444e-01 -3.89468700e-01
-8.11679885e-02 2.33222112e-01 -1.65098071e+00 2.90326387e-01
6.59571946e-01 1.27960122e+00 3.56951237e-01 3.85469884e-01
3.31577063e-01 5.94001532e-01 -4.55477268e-01 -2.87819535e-01
1.41252324e-01 -2.70740002e-01 -7.29193509e-01 2.76665092e-01
8.59301507e-01 -2.30953753e-01 -3.75388294e-01 8.46420646e-01
4.78387177e-01 -7.00146258e-02 3.97508472e-01 -1.31345618e+00
-3.34858119e-01 2.44070724e-01 -7.23815322e-01 5.43167489e-03
4.50885773e-01 1.88668296e-01 1.01581621e+00 -1.11484957e+00
6.35796070e-01 1.22947657e+00 7.72125065e-01 2.80124605e-01
-8.46734643e-01 -6.14205539e-01 5.32434762e-01 4.01739478e-01
-1.64529932e+00 -1.40010938e-01 8.69743943e-01 -3.12855065e-01
5.18893480e-01 1.43186599e-01 8.98667455e-01 1.21670675e+00
1.68764040e-01 9.74515498e-01 1.03052890e+00 -1.93847388e-01
-1.05857402e-02 -2.21756041e-01 -7.77714252e-02 9.70491469e-01
7.35424012e-02 1.08417980e-01 -7.92762160e-01 1.96368232e-01
1.00667179e+00 2.47672349e-01 -2.60453165e-01 -6.51688099e-01
-1.62625241e+00 5.21287262e-01 9.40788865e-01 9.54495147e-02
-4.00138378e-01 3.00404340e-01 3.19598705e-01 -1.71687007e-01
2.57156760e-01 2.89561331e-01 -6.94455445e-01 4.24331091e-02
-8.85409057e-01 3.21423590e-01 3.05264205e-01 1.27066553e+00
7.07957029e-01 -4.15730417e-01 -5.08983195e-01 7.06689835e-01
2.46584401e-01 5.14471650e-01 3.68672252e-01 -8.35122585e-01
6.56877518e-01 7.74587929e-01 1.18815273e-01 -1.22140098e+00
-5.19141436e-01 -8.92121792e-01 -8.20121050e-01 -1.72383532e-01
4.63599920e-01 1.71891034e-01 -1.03863502e+00 1.78583908e+00
5.93439996e-01 2.95400620e-01 -5.91610014e-01 1.33973694e+00
6.53089762e-01 2.98421234e-01 1.62032351e-01 2.22213209e-01
1.67272866e+00 -1.52738142e+00 -5.14206648e-01 -4.17165309e-01
5.80569506e-01 -2.91447610e-01 1.25664818e+00 2.38010168e-01
-6.83899283e-01 -9.95198488e-01 -1.05631340e+00 -2.12828621e-01
-3.05576146e-01 6.02517903e-01 5.64388156e-01 2.40732819e-01
-6.42816305e-01 2.56017625e-01 -8.79021049e-01 -1.91548973e-01
3.16120923e-01 1.53765231e-01 -6.77816093e-01 -2.47062579e-01
-1.17423427e+00 6.99901521e-01 3.45387042e-01 5.12978256e-01
-7.92555690e-01 -5.83690882e-01 -1.23932040e+00 -1.80686921e-01
9.18813229e-01 -8.28055382e-01 9.45781052e-01 -5.51473022e-01
-1.10349786e+00 6.88283622e-01 -9.85986516e-02 -2.24792421e-01
6.45366490e-01 -5.48212409e-01 -1.25034258e-01 2.79081196e-01
4.52715904e-01 9.13090467e-01 7.22426713e-01 -1.26774883e+00
-9.89499390e-01 -5.30119419e-01 3.14394943e-02 2.61294633e-01
-2.26660911e-02 -4.16555345e-01 -1.05945730e+00 -1.01245224e+00
3.29954237e-01 -1.16126633e+00 -2.59872198e-01 3.24166238e-01
-4.90067691e-01 -2.41108224e-01 6.61703706e-01 -7.48340607e-01
1.13572681e+00 -1.94620228e+00 4.67741817e-01 7.35684186e-02
1.42635033e-01 -3.33685428e-02 -9.04729515e-02 8.68901536e-02
1.05663188e-01 -4.03703094e-01 -2.25169405e-01 -4.78757024e-01
1.06237568e-01 3.28727156e-01 -9.61290151e-02 5.42217374e-01
2.60376900e-01 1.29342544e+00 -1.07309413e+00 -7.61109948e-01
4.89211768e-01 5.00407755e-01 -7.80730069e-01 1.72534078e-01
-2.46098816e-01 7.29556084e-01 -4.81431216e-01 9.52933192e-01
4.21196222e-01 -3.99018824e-01 3.51455919e-02 -7.25509226e-01
1.88932061e-01 -1.02938917e-02 -1.33666527e+00 2.41463280e+00
-4.55182880e-01 1.03194363e-01 3.75224352e-02 -7.91435122e-01
4.95968819e-01 3.80066596e-02 5.00237525e-01 -7.15686321e-01
-1.00535877e-01 1.48561090e-01 -2.63549149e-01 -2.29632020e-01
5.45248330e-01 1.28597803e-02 -5.53730786e-01 -7.73670971e-02
3.27858150e-01 2.94940054e-01 -1.95725486e-01 1.20460123e-01
9.72958982e-01 6.82610452e-01 2.71110356e-01 3.64387082e-03
5.24807751e-01 2.49448083e-02 6.60233676e-01 4.92568135e-01
-4.04845268e-01 8.78863811e-01 2.54319478e-02 -3.04922819e-01
-6.56664550e-01 -1.03266060e+00 8.79841819e-02 1.50922489e+00
6.08547390e-01 -8.73279691e-01 -7.11763144e-01 -9.79586303e-01
-2.30283625e-02 -7.93579966e-02 -8.79665136e-01 -1.57967299e-01
-8.27193379e-01 -3.36884022e-01 4.35929745e-01 9.22984838e-01
8.05888176e-01 -9.84740436e-01 -6.60602331e-01 1.02404080e-01
-6.79415405e-01 -1.25993598e+00 -8.22117627e-01 1.60818785e-01
-5.65565646e-01 -1.20590711e+00 -9.89235997e-01 -7.93489337e-01
5.73370993e-01 2.73956433e-02 1.13841271e+00 7.37163574e-02
-4.74039614e-01 3.09737623e-01 -5.17293215e-01 -4.85251360e-02
3.86037976e-01 2.14474961e-01 4.46395464e-02 -1.21780507e-01
7.79692009e-02 -4.29301560e-01 -1.02645707e+00 7.17996120e-01
-4.65065747e-01 1.03419051e-01 1.00446200e+00 9.08500254e-01
1.06911278e+00 1.22920917e-02 3.06064010e-01 -5.68992257e-01
-2.05472887e-01 -2.26453930e-01 -1.22947596e-01 4.13671732e-01
-2.16226101e-01 9.70249623e-02 2.59593934e-01 -5.31309843e-01
-9.00561810e-01 4.80920941e-01 -2.49465391e-01 -5.49882054e-01
-3.15682262e-01 2.70083010e-01 -4.87580240e-01 7.45247072e-03
5.38212717e-01 2.78496355e-01 -4.13641036e-01 -6.69096649e-01
5.23688078e-01 3.20763588e-01 1.03167212e+00 -8.85322928e-01
8.45041454e-01 5.47301710e-01 -3.55531499e-02 -2.72303075e-01
-1.30223525e+00 -7.02211678e-01 -9.85963464e-01 -2.21631601e-01
9.50715125e-01 -1.30294704e+00 -4.80664074e-01 3.50254625e-01
-8.96787584e-01 -3.29263419e-01 -4.24593776e-01 4.00822669e-01
-6.96482956e-01 3.53464931e-01 -6.16788566e-01 -4.39079434e-01
-1.22639544e-01 -1.09324503e+00 1.92906475e+00 -6.08858615e-02
-1.41174763e-01 -4.93110508e-01 -4.29137908e-02 4.47413206e-01
-7.74188414e-02 4.09218401e-01 3.40357095e-01 -2.03582853e-01
-6.01604521e-01 -1.96719259e-01 -2.99878418e-01 1.19799510e-01
-3.68340909e-02 -7.11926341e-01 -8.15862358e-01 -3.48297536e-01
-1.57845438e-01 -4.65963811e-01 7.95577168e-01 2.09599182e-01
1.34234202e+00 -1.60760313e-01 -6.28026724e-01 7.56700218e-01
1.13152921e+00 -5.72771728e-01 2.92290181e-01 2.50369310e-01
1.21643448e+00 7.27370262e-01 1.23283088e+00 4.14270222e-01
7.23388672e-01 1.08972871e+00 3.21967661e-01 -1.30540341e-01
-4.33150798e-01 -8.46311688e-01 8.83492082e-02 6.36854708e-01
-2.25848556e-01 1.76781610e-01 -6.88841522e-01 5.80641806e-01
-2.06187606e+00 -7.48021841e-01 6.89905509e-02 1.89210832e+00
9.48061049e-01 1.47383317e-01 3.80122244e-01 1.39397740e-01
7.05334246e-01 1.72214746e-01 -3.62155199e-01 6.06818259e-01
1.21129557e-01 6.25028927e-03 5.24702728e-01 2.64240265e-01
-1.37745583e+00 9.80965197e-01 5.40225935e+00 1.05581200e+00
-8.65808904e-01 3.52817088e-01 2.49736860e-01 -1.49936199e-01
2.04612896e-01 -2.61504710e-01 -9.35973048e-01 6.13284230e-01
3.70088190e-01 5.52411020e-01 -1.18634388e-01 1.16114819e+00
-1.06468633e-01 -9.07400250e-02 -1.26470709e+00 1.34928167e+00
1.57485649e-01 -9.46725905e-01 -1.52947471e-01 1.41837806e-01
5.59782088e-01 -1.54003218e-01 -7.67195895e-02 5.36775351e-01
-1.42449141e-01 -9.34381604e-01 1.22001934e+00 4.75546330e-01
9.18745458e-01 -7.02795565e-01 7.21215546e-01 6.37972295e-01
-1.84294236e+00 -1.55422956e-01 -2.10711122e-01 -5.88360317e-02
1.76306963e-01 3.08644027e-01 -3.15882176e-01 7.52787650e-01
1.03289425e+00 8.03081393e-01 -9.07838762e-01 9.18726265e-01
-4.99526918e-01 2.03867793e-01 -4.49960530e-01 3.92215490e-01
1.04257263e-01 2.81026453e-01 1.80833310e-01 1.22145450e+00
1.29332751e-01 1.09092921e-01 7.10843205e-01 6.20947123e-01
1.53123379e-01 -2.02584967e-01 -6.25669658e-02 6.02121949e-01
4.43489045e-01 1.31184387e+00 -7.41560638e-01 -3.41098398e-01
-2.38738820e-01 1.33964705e+00 4.19854641e-01 1.12057418e-01
-1.13393044e+00 -2.06787914e-01 4.09340709e-01 3.88842911e-01
4.42814916e-01 -3.51034820e-01 -1.14204183e-01 -1.32166123e+00
4.65850115e-01 -8.53762209e-01 5.76429904e-01 -6.95514798e-01
-1.22605968e+00 4.85626668e-01 2.18228638e-01 -1.37933123e+00
-2.67758638e-01 -5.97534776e-01 -1.32355615e-01 5.86416960e-01
-1.42672896e+00 -1.70908976e+00 -7.08398998e-01 9.10065055e-01
6.50361001e-01 3.31941664e-01 5.51777840e-01 4.86391902e-01
-3.05407256e-01 9.51209307e-01 -7.30726302e-01 4.69483644e-01
9.97388959e-01 -1.26693451e+00 1.43180490e-01 7.31441140e-01
3.28069270e-01 7.52676487e-01 4.43040520e-01 -8.13289523e-01
-1.22530687e+00 -1.30098104e+00 6.99453831e-01 -9.31661904e-01
2.40229115e-01 -6.49569511e-01 -6.01615369e-01 6.60134077e-01
-4.79223490e-01 6.07325613e-01 3.29045057e-01 2.91689813e-01
-4.96356189e-01 -6.10358715e-02 -8.95675838e-01 4.83744591e-01
1.73467505e+00 -6.63947046e-01 -8.72011125e-01 2.60525525e-01
7.90196359e-01 -6.59847677e-01 -1.03984392e+00 7.68106759e-01
6.28276169e-01 -7.29640722e-01 1.25668538e+00 -1.10178165e-01
3.48877221e-01 -7.45058000e-01 -3.13841134e-01 -9.47088480e-01
-3.05639446e-01 -5.68926856e-02 -3.16874802e-01 1.04358363e+00
1.87537018e-02 -5.65730780e-02 1.02665567e+00 2.78923839e-01
-2.31648743e-01 -9.61715519e-01 -8.58064413e-01 -7.15149164e-01
-3.17236662e-01 -5.52083671e-01 4.21011955e-01 6.23356938e-01
-1.41191423e-01 3.41866672e-01 -6.15209401e-01 2.99386859e-01
7.44866073e-01 3.24141294e-01 9.49634075e-01 -1.01982450e+00
-5.38907111e-01 -2.30385344e-02 -7.02429473e-01 -1.55925190e+00
-2.98228115e-02 -4.74160314e-01 4.92153555e-01 -1.30413485e+00
2.38963798e-01 -6.45275116e-01 -3.72968853e-01 6.40185177e-01
-3.29699904e-01 7.89588213e-01 2.69629061e-01 4.23038512e-01
-1.18568432e+00 6.26424491e-01 1.38822353e+00 -9.35969055e-02
1.05204873e-01 -1.79289088e-01 -4.72407818e-01 8.44770193e-01
2.25568458e-01 -3.62443656e-01 -4.45041150e-01 -2.36858606e-01
1.10162809e-01 -1.77636649e-02 9.33220685e-01 -1.39339137e+00
4.13399965e-01 -7.86013156e-02 7.24061370e-01 -1.04746461e+00
4.43596601e-01 -8.62559080e-01 3.63294035e-02 2.72581011e-01
-3.60579610e-01 -1.99723929e-01 -2.94672489e-01 1.03124940e+00
-3.80726844e-01 3.54764611e-01 5.42722106e-01 -2.76032805e-01
-1.17121089e+00 6.74367607e-01 3.72566670e-01 1.72433719e-01
9.55839217e-01 -1.44396603e-01 1.08291700e-01 -6.90473467e-02
-8.85389924e-01 3.63581151e-01 6.39434218e-01 6.35645747e-01
5.34433424e-01 -1.51630855e+00 -2.36253381e-01 2.66302645e-01
5.81356108e-01 5.47188222e-01 5.13731599e-01 9.92158651e-01
-1.78620145e-01 4.98503149e-01 -1.78486511e-01 -9.36419785e-01
-1.05515575e+00 6.28531516e-01 3.23394984e-01 -3.64282906e-01
-8.38309228e-01 9.40023124e-01 5.11649132e-01 -5.15374660e-01
4.74661976e-01 -5.17663121e-01 -1.03857061e-02 -1.22692965e-01
5.19825161e-01 1.53237179e-01 3.27938199e-02 -9.94035065e-01
-6.53598726e-01 1.00601387e+00 2.20138103e-01 -7.56762549e-02
9.55199957e-01 -3.19118261e-01 2.24097386e-01 2.77024806e-01
1.26732492e+00 1.44779123e-02 -1.68106771e+00 -4.23011124e-01
-6.53950348e-02 -4.71702158e-01 -1.92134783e-01 -9.49390709e-01
-1.10469949e+00 9.49546099e-01 6.05554461e-01 -6.10395133e-01
1.04581738e+00 3.09867293e-01 1.01415086e+00 3.20574015e-01
9.36733544e-01 -1.23934424e+00 4.07556295e-01 2.03206360e-01
9.26844418e-01 -1.36486197e+00 7.92057291e-02 -7.82641172e-01
-6.98444903e-01 6.33689821e-01 9.90245163e-01 -2.09414959e-01
6.57101035e-01 6.17008209e-02 -2.21470803e-01 -3.63632381e-01
-3.22310358e-01 -4.98617500e-01 8.14364254e-01 7.28761435e-01
2.84575313e-01 7.79486969e-02 -8.05854350e-02 1.03080261e+00
-1.98181480e-01 7.15742409e-02 -3.50961059e-01 9.83087957e-01
-3.44924659e-01 -9.31691527e-01 -4.27959800e-01 1.24905005e-01
-2.08410069e-01 1.29007161e-01 -1.46624982e-01 8.82217050e-01
6.85831249e-01 6.99278653e-01 -5.03330901e-02 -6.45247817e-01
5.39969563e-01 -9.41092372e-02 5.91707408e-01 -7.66519725e-01
-4.25277054e-01 1.32574201e-01 -2.92966887e-02 -1.03756964e+00
-5.62833130e-01 -3.42508197e-01 -1.23992872e+00 4.12652530e-02
-2.80378878e-01 -2.19469313e-02 2.50652254e-01 1.03966510e+00
3.93570960e-01 5.60652971e-01 1.39000788e-01 -1.17698014e+00
-3.47399890e-01 -9.46429193e-01 -5.73692143e-01 6.02463126e-01
2.09719777e-01 -1.11386693e+00 -1.82390735e-02 -7.60956481e-02] | [7.232062339782715, -0.7080006003379822] |
cb08b3fe-6bfb-4e79-84d0-848090be76d1 | inferring-fluid-dynamics-via-inverse | 2304.04446 | null | https://arxiv.org/abs/2304.04446v1 | https://arxiv.org/pdf/2304.04446v1.pdf | Inferring Fluid Dynamics via Inverse Rendering | Humans have a strong intuitive understanding of physical processes such as fluid falling by just a glimpse of such a scene picture, i.e., quickly derived from our immersive visual experiences in memory. This work achieves such a photo-to-fluid-dynamics reconstruction functionality learned from unannotated videos, without any supervision of ground-truth fluid dynamics. In a nutshell, a differentiable Euler simulator modeled with a ConvNet-based pressure projection solver, is integrated with a volumetric renderer, supporting end-to-end/coherent differentiable dynamic simulation and rendering. By endowing each sampled point with a fluid volume value, we derive a NeRF-like differentiable renderer dedicated from fluid data; and thanks to this volume-augmented representation, fluid dynamics could be inversely inferred from the error signal between the rendered result and ground-truth video frame (i.e., inverse rendering). Experiments on our generated Fluid Fall datasets and DPI Dam Break dataset are conducted to demonstrate both effectiveness and generalization ability of our method. | ['Zhenbo Yu', 'Jiyao Mao', 'Bingbing Ni', 'Ye Chen', 'Jinxian Liu'] | 2023-04-10 | null | null | null | null | ['inverse-rendering'] | ['computer-vision'] | [-1.60483688e-01 1.87493354e-01 7.90285110e-01 -2.21387461e-01
-2.23184884e-01 -3.12375516e-01 6.70640349e-01 -1.67892277e-01
-2.41798654e-01 7.12083817e-01 1.40992299e-01 8.95856172e-02
3.33281964e-01 -1.07515216e+00 -1.12082052e+00 -4.33271259e-01
-3.43084872e-01 5.91274679e-01 -6.16921186e-02 -2.83297330e-01
1.11627512e-01 3.47027719e-01 -1.61225128e+00 1.29011035e-01
9.44161475e-01 1.14480698e+00 2.75581181e-01 1.08774102e+00
-1.41313612e-01 1.23222661e+00 -1.34281904e-01 -2.14788809e-01
4.35920209e-01 -2.72666961e-01 -4.94607568e-01 1.58792973e-01
5.62339842e-01 -1.06687880e+00 -6.83239698e-01 4.98925269e-01
1.38672799e-01 4.37004864e-01 4.25513685e-01 -8.60156715e-01
-3.36706609e-01 -1.32433906e-01 -2.73028105e-01 -1.22707807e-01
9.07530129e-01 7.41241753e-01 5.70579171e-01 -1.19657624e+00
1.08894539e+00 1.46655071e+00 6.02502882e-01 5.81044018e-01
-1.15748239e+00 -3.71329576e-01 2.26166040e-01 -3.80204231e-01
-1.01875353e+00 -7.32315481e-02 6.73287332e-01 -8.51449430e-01
7.35701501e-01 4.59729075e-01 1.48482311e+00 9.91096199e-01
2.78388739e-01 5.72081149e-01 1.00142598e+00 2.28347242e-01
4.74264681e-01 -9.87024754e-02 -1.60634935e-01 9.34551954e-01
-7.98295736e-02 3.39332283e-01 -8.17128837e-01 -3.33195031e-01
1.40208173e+00 3.27917755e-01 -7.80776918e-01 -7.20734894e-01
-1.01536345e+00 3.57270896e-01 6.63940012e-01 -3.09744000e-01
-4.71581936e-01 6.09380364e-01 3.67897689e-01 8.63600448e-02
5.90278268e-01 1.32230327e-01 -1.12059481e-01 -4.92670506e-01
-6.76617682e-01 9.36670244e-01 1.04759336e+00 6.64661944e-01
8.65919054e-01 2.89219528e-01 -2.08146758e-02 2.58457232e-02
4.08601463e-01 9.10284400e-01 3.49611863e-02 -1.26367652e+00
3.68799716e-01 4.28618014e-01 5.40548027e-01 -1.05592608e+00
-2.56014496e-01 1.82249367e-01 -8.54771912e-01 6.44287109e-01
5.22694230e-01 -1.69667467e-01 -6.48669958e-01 1.51844907e+00
6.42950952e-01 7.89293170e-01 -1.02202088e-01 1.53811884e+00
5.22528052e-01 8.03048074e-01 7.71410093e-02 3.18034348e-04
9.81165290e-01 -5.36775410e-01 -4.35143679e-01 -5.90250231e-02
2.85140783e-01 2.87538450e-02 1.25980496e+00 3.80180061e-01
-1.16209543e+00 -4.85780507e-01 -8.11062455e-01 -2.89602816e-01
2.50188291e-01 -5.93436360e-01 4.61364061e-01 -9.19977501e-02
-8.45139146e-01 1.19112897e+00 -1.31592059e+00 1.75439090e-01
2.86125392e-01 -2.67643481e-01 -4.25550520e-01 2.30416596e-01
-1.00634348e+00 6.72970653e-01 -1.49709880e-01 3.95296097e-01
-1.09588587e+00 -1.53453422e+00 -8.65528107e-01 -8.25645626e-02
7.13079870e-02 -1.37632871e+00 1.10713172e+00 -7.08277166e-01
-1.64749384e+00 6.76656008e-01 1.08114053e-02 -4.86934811e-01
1.44391000e+00 -8.67855489e-01 3.82009327e-01 5.46727359e-01
-1.76273242e-01 3.36602688e-01 6.87498093e-01 -1.71182632e+00
-1.49581000e-01 -2.39347354e-01 8.35299194e-02 5.34504592e-01
1.31900758e-01 -6.06708050e-01 -2.75396742e-02 -1.80361927e-01
3.01230773e-02 -6.72936738e-01 -2.75212497e-01 1.09158850e+00
-2.79676169e-01 5.77508628e-01 8.00534189e-01 -1.07177556e+00
7.93653846e-01 -1.84046555e+00 4.99863684e-01 6.57349601e-02
7.61702895e-01 -2.84319315e-02 2.77585149e-01 3.46815050e-01
2.35861644e-01 -1.07661933e-01 -3.94094050e-01 -6.81232750e-01
1.51556954e-01 8.28104094e-02 -8.35960448e-01 7.58358717e-01
7.73045123e-02 9.36444581e-01 -1.30019498e+00 -1.97181776e-01
7.94885933e-01 8.80489469e-01 -7.23128438e-01 7.25247145e-01
-3.76079410e-01 1.09929085e+00 -6.65550768e-01 2.52714932e-01
1.03654480e+00 -6.09149188e-02 3.45019437e-02 1.01455539e-01
-2.91928947e-01 5.77933155e-02 -1.13724184e+00 1.95739043e+00
-7.56166339e-01 5.16451716e-01 4.21440452e-01 -4.97243524e-01
7.87564516e-01 2.00437039e-01 4.28793013e-01 -6.83392584e-01
2.40372643e-01 1.30347526e-02 -5.21948755e-01 -7.17594445e-01
6.72416389e-01 -4.60781723e-01 9.84748676e-02 3.26790124e-01
-3.71022344e-01 -6.05519950e-01 -4.28201020e-01 3.94840747e-01
1.17569947e+00 8.96796346e-01 -1.63769349e-01 -1.29289240e-01
2.41443768e-01 -2.11552054e-01 2.27014512e-01 4.19519752e-01
9.83011723e-02 8.33462596e-01 3.11966151e-01 -7.53891587e-01
-1.48984253e+00 -1.62925959e+00 -2.86722686e-02 4.58298713e-01
5.97395301e-01 -4.18014884e-01 -7.69102693e-01 9.60596874e-02
4.79159832e-01 3.30941141e-01 -7.94010699e-01 -7.01699182e-02
-9.72403884e-01 2.82037072e-02 2.05564424e-01 5.02712786e-01
5.64871252e-01 -1.14522934e+00 -1.56554854e+00 3.62529099e-01
-5.34690358e-02 -9.10917401e-01 -3.07593755e-02 -3.28650028e-01
-8.35392952e-01 -1.09131229e+00 -7.13416934e-01 -3.36325690e-02
3.57092053e-01 -1.35967927e-02 1.20988548e+00 3.86318684e-01
-2.72174180e-01 4.04092968e-01 -3.20099667e-02 2.55999058e-01
-4.03183103e-01 -8.53857458e-01 -1.29511088e-01 6.99949339e-02
-5.25166929e-01 -1.00796199e+00 -1.13080668e+00 -1.64604872e-01
-6.52683139e-01 5.96083879e-01 -2.42293835e-01 6.02338612e-01
5.22657812e-01 -8.13556075e-01 -2.25828886e-01 -5.41059375e-01
4.61392671e-01 -4.03580368e-01 -5.18552423e-01 -8.45795199e-02
2.14670688e-01 -3.57996598e-02 9.20652509e-01 -6.30386651e-01
-1.18570518e+00 -1.34064499e-02 3.38579640e-02 -1.01417947e+00
1.81795776e-01 2.07818210e-01 1.64052039e-01 9.47712734e-02
7.17100561e-01 1.98172465e-01 4.38986011e-02 -3.58654082e-01
5.83648622e-01 1.57941923e-01 8.25343966e-01 -1.00861168e+00
9.42252815e-01 9.32061732e-01 5.16926385e-02 -7.54240453e-01
-3.89842391e-01 -9.26120430e-02 -5.13320029e-01 -7.18068957e-01
9.61154640e-01 -9.32258189e-01 -1.36194432e+00 7.07265735e-01
-1.16475928e+00 -8.72308850e-01 -5.31878948e-01 4.00654644e-01
-9.55546021e-01 3.54766518e-01 -1.05052972e+00 -1.10841966e+00
-3.62947762e-01 -9.40826476e-01 1.37286568e+00 1.92119852e-01
-3.37933034e-01 -9.97201324e-01 2.88091332e-01 2.15533879e-02
2.06580743e-01 1.24235940e+00 1.65952623e-01 5.08128941e-01
-1.11793697e+00 1.33099094e-01 -1.54091762e-02 1.16220474e-01
-2.79385358e-01 1.34325519e-01 -1.15015590e+00 -8.72698724e-02
2.93138802e-01 -3.53350222e-01 6.72169924e-01 2.69359916e-01
9.88420486e-01 -2.87653238e-01 -2.44507328e-01 9.81219530e-01
1.56241930e+00 -2.72984475e-01 6.59706831e-01 -1.19912460e-01
1.07288969e+00 3.71968061e-01 4.60032463e-01 9.98381317e-01
3.78285110e-01 4.77955282e-01 6.17244422e-01 -9.67886820e-02
-5.80917150e-02 -6.14598632e-01 1.61731869e-01 5.28306663e-01
-4.26811576e-01 -2.62409151e-01 -8.93577039e-01 5.06900549e-02
-1.81862307e+00 -8.21955025e-01 -3.57537419e-01 2.06286478e+00
6.13300920e-01 -5.34386560e-02 -1.48679450e-01 -2.51837462e-01
2.17519447e-01 2.44978815e-01 -9.30679202e-01 -8.66298154e-02
2.54751623e-01 2.36223519e-01 -4.76645008e-02 8.42087090e-01
-6.31175637e-01 6.86799705e-01 4.84896660e+00 1.77099541e-01
-1.30769920e+00 3.55549785e-03 4.96189922e-01 -3.03820670e-01
-6.45554602e-01 9.48762372e-02 5.46709262e-02 4.08909529e-01
7.36915708e-01 -3.64344895e-01 5.76048076e-01 6.55091822e-01
5.87926626e-01 -3.32344264e-01 -1.22838414e+00 9.53606367e-01
-2.38501817e-01 -1.58898866e+00 3.05231437e-02 4.93023768e-02
3.66657555e-01 -3.17758262e-01 -9.43942815e-02 2.28697043e-02
3.66310090e-01 -9.16337073e-01 1.42073584e+00 1.18382525e+00
9.57554162e-01 -3.92753541e-01 -7.02509880e-02 5.11868179e-01
-1.32379699e+00 3.09399337e-01 -1.11996025e-01 -5.80238283e-01
8.59702051e-01 5.79016447e-01 -1.77839354e-01 6.22159123e-01
5.50648034e-01 7.73153543e-01 1.22195974e-01 6.88095987e-01
1.19061610e-02 1.78386867e-01 -3.80296528e-01 1.03315778e-01
-4.47577797e-02 -7.29342878e-01 7.66102850e-01 9.36854243e-01
2.81886518e-01 4.76775676e-01 2.86477596e-01 1.35327148e+00
3.40976939e-02 -2.60070950e-01 -5.33591926e-01 3.06942165e-01
1.78017706e-01 1.10309231e+00 -4.88541126e-01 -4.89827931e-01
1.28024630e-02 1.02835834e+00 5.05271375e-01 4.80357021e-01
-1.07777274e+00 2.11305648e-01 1.15659845e+00 6.14169657e-01
-1.80913836e-01 -3.44995797e-01 -2.60567367e-01 -1.44622827e+00
3.91735882e-01 -1.47272274e-01 -3.56933236e-01 -1.37692285e+00
-1.15752971e+00 4.83655632e-01 -1.03362486e-01 -1.41450298e+00
-2.91768491e-01 -3.38004470e-01 -5.32444358e-01 1.15203643e+00
-1.17981195e+00 -7.47895241e-01 -9.67218518e-01 3.48382145e-01
2.72425376e-02 7.25790679e-01 5.11417031e-01 -1.64130814e-02
1.25429975e-02 -3.05897705e-02 -2.09198549e-01 6.24162555e-02
2.87340790e-01 -1.20835197e+00 5.70040941e-01 5.94024599e-01
-4.96406704e-01 2.98892558e-01 9.55366552e-01 -8.55616033e-01
-1.70666444e+00 -1.03981793e+00 5.13177784e-03 -6.21182501e-01
7.79387116e-01 -6.10374212e-01 -1.48284352e+00 6.13730729e-01
-3.46791923e-01 4.78364587e-01 -2.81828105e-01 -5.01293123e-01
-1.34577766e-01 2.72502005e-01 -1.24254549e+00 6.99401557e-01
1.43043447e+00 -6.28905535e-01 -3.28437656e-01 1.08721256e-01
6.83987379e-01 -1.32436419e+00 -7.36761987e-01 4.90606837e-02
6.43799365e-01 -1.16261876e+00 1.10646999e+00 -3.04887801e-01
1.21238816e+00 -3.78534973e-01 -3.67275923e-02 -1.28880954e+00
2.72657096e-01 -7.52020180e-01 -4.95769054e-01 6.68729424e-01
-3.71430248e-01 -4.46337998e-01 5.85886419e-01 1.08412194e+00
-2.88540035e-01 -9.45452034e-01 -8.32904279e-01 -5.44990599e-01
3.04861367e-01 -6.05766118e-01 6.41524076e-01 6.90064192e-01
-1.11421749e-01 -2.48926163e-01 -5.25646150e-01 9.90418121e-02
7.48849034e-01 1.56962469e-01 1.04619348e+00 -9.60447073e-01
-3.95807564e-01 2.02240031e-02 -5.06068289e-01 -1.40147591e+00
2.58755833e-01 -5.65989494e-01 3.36935252e-01 -1.46670556e+00
4.50334884e-02 -4.40813690e-01 5.41262746e-01 -2.01763302e-01
-2.34925121e-01 1.13763221e-01 5.42220414e-01 5.44359758e-02
-2.01990098e-01 8.66500199e-01 1.93655074e+00 2.89791614e-01
-3.23794663e-01 -6.12347364e-01 -5.40314522e-03 8.09490919e-01
2.04201803e-01 -2.02005759e-01 -3.87993276e-01 -4.63825673e-01
1.59391880e-01 1.05301094e+00 1.13891578e+00 -9.72065687e-01
-3.82458940e-02 -2.24406868e-01 3.90293866e-01 -3.33442867e-01
7.74320066e-01 -5.30783117e-01 6.26713514e-01 6.32681489e-01
-3.05785239e-01 -2.16043472e-01 9.32380483e-02 7.66854525e-01
2.31364623e-01 4.16606307e-01 6.37829661e-01 -3.59123409e-01
-6.56695247e-01 6.26764178e-01 3.55709121e-02 3.43001425e-01
7.57782638e-01 -3.78864795e-01 -1.51986226e-01 -4.42080528e-01
-9.47897375e-01 2.09905013e-01 9.61534798e-01 6.60624355e-02
8.68923187e-01 -1.14167523e+00 -6.87329829e-01 3.02005321e-01
-2.16256067e-01 7.07236946e-01 8.17464888e-01 3.68708014e-01
-1.17409909e+00 -3.83315116e-01 -1.80757463e-01 -8.52648020e-01
-6.31999850e-01 2.89800376e-01 7.92519808e-01 -1.24428146e-01
-1.72462773e+00 7.86721766e-01 6.53477371e-01 -4.12127703e-01
-1.31136149e-01 -7.06879616e-01 4.49077398e-01 -4.40875053e-01
7.43702650e-01 3.47772539e-01 -1.82974324e-01 -4.48386729e-01
-1.50853366e-01 6.32109880e-01 6.61139429e-01 -4.45151925e-01
1.36267149e+00 -1.58026237e-02 2.27621794e-01 8.16601872e-01
1.06568599e+00 -2.42999554e-01 -2.37126374e+00 2.71865547e-01
-7.18578875e-01 -8.68990242e-01 -2.31067911e-01 -4.40829337e-01
-1.07950795e+00 9.72519815e-01 2.92491466e-01 -2.28423104e-01
6.08522475e-01 -2.08526686e-01 9.15018916e-01 1.86812952e-01
7.25910783e-01 -6.07523620e-01 1.21665366e-01 3.22974414e-01
1.20076609e+00 -7.82567620e-01 -2.53588259e-02 -6.99926078e-01
-3.40703607e-01 1.08637667e+00 7.19333291e-01 -8.66589129e-01
6.66080773e-01 6.63659632e-01 8.94587301e-03 -2.86517739e-01
-7.80553937e-01 4.56645072e-01 -2.25058228e-01 5.24755955e-01
5.31866997e-02 1.76035911e-01 2.08077982e-01 2.45522335e-01
-5.82575440e-01 4.35625881e-01 7.32931793e-01 1.01168871e+00
-3.52809489e-01 -2.51162022e-01 -2.92843163e-01 2.71817744e-01
4.37210351e-01 1.98408067e-01 1.06391430e-01 7.71766782e-01
-9.78014618e-02 2.32400745e-01 6.57261789e-01 -1.25703484e-01
4.40259129e-01 -2.74186730e-01 7.66387224e-01 -4.09579337e-01
-6.52342439e-01 -2.67706782e-01 -2.16231570e-01 -1.09902656e+00
-1.31984740e-01 -5.24468303e-01 -1.87247396e+00 -8.86571884e-01
3.33431751e-01 -1.27545148e-01 2.02272892e-01 7.83901751e-01
1.00217968e-01 6.27263010e-01 4.46458519e-01 -1.74372649e+00
-2.84674466e-01 -7.06611931e-01 -7.81672537e-01 1.03343630e+00
5.37062407e-01 -7.48870134e-01 -5.76410592e-01 2.64145553e-01] | [9.092000961303711, -3.068535327911377] |
7693d80f-d344-4193-a0b4-2a7d39ae4d77 | mudpt-multi-modal-deep-symphysis-prompt | 2306.11400 | null | https://arxiv.org/abs/2306.11400v1 | https://arxiv.org/pdf/2306.11400v1.pdf | MuDPT: Multi-modal Deep-symphysis Prompt Tuning for Large Pre-trained Vision-Language Models | Prompt tuning, like CoOp, has recently shown promising vision recognizing and transfer learning ability on various downstream tasks with the emergence of large pre-trained vision-language models like CLIP. However, we identify that existing uni-modal prompt tuning approaches may result in sub-optimal performance since this uni-modal design breaks the original alignment of textual and visual representations in the pre-trained model. Inspired by the nature of pre-trained vision-language models, we aim to achieve completeness in prompt tuning and propose a novel approach called Multi-modal Deep-symphysis Prompt Tuning, dubbed as MuDPT, which extends independent multi-modal prompt tuning by additionally learning a model-agnostic transformative network to allow deep hierarchical bi-directional prompt fusion. We evaluate the effectiveness of MuDPT on few-shot vision recognition and out-of-domain generalization tasks. Compared with the state-of-the-art methods, MuDPT achieves better recognition and generalization ability with an apparent margin thanks to synergistic alignment of textual and visual representations. Our code is available at: https://github.com/Mechrev0/MuDPT. | ['Ting Wang', 'Jintao Tang', 'Shasha Li', 'Yongzhu Miao'] | 2023-06-20 | null | null | null | null | ['domain-generalization'] | ['methodology'] | [ 1.11566484e-01 -9.90238786e-02 -1.70979053e-01 -2.56204516e-01
-1.09510231e+00 -7.85329461e-01 1.03611898e+00 -1.79062083e-01
-4.83447462e-01 1.41726211e-01 4.42908883e-01 -2.94119835e-01
-1.14866480e-01 -4.45524991e-01 -8.09102118e-01 -5.55852830e-01
4.75608945e-01 3.70410889e-01 3.98659378e-01 -3.67905289e-01
2.38745257e-01 3.63047451e-01 -1.68762016e+00 8.10702682e-01
6.65407121e-01 1.03044963e+00 4.91056800e-01 7.95404613e-01
-1.87450245e-01 9.97692287e-01 -2.37435505e-01 -4.04035568e-01
3.72536182e-01 -2.52508819e-01 -8.18972647e-01 3.62209305e-02
8.69222641e-01 -5.89802027e-01 -5.33538818e-01 7.06858695e-01
6.97136819e-01 2.24994555e-01 8.17188203e-01 -1.24663484e+00
-1.35140908e+00 5.41897774e-01 -6.42584920e-01 4.05219346e-01
5.65280095e-02 6.54557109e-01 1.08912623e+00 -1.01518631e+00
3.80410939e-01 1.19616592e+00 6.89786017e-01 8.70440066e-01
-1.38384414e+00 -5.14886439e-01 2.25672320e-01 4.36215192e-01
-1.23201740e+00 -6.36659563e-01 4.80251104e-01 -6.27913892e-01
1.28247976e+00 -1.45710438e-01 4.18993533e-01 1.41252756e+00
-1.19782768e-01 9.94651735e-01 1.12339330e+00 -4.12511677e-01
-1.63567998e-02 6.40378073e-02 1.68787181e-01 7.63699710e-01
-1.78470045e-01 5.16566813e-01 -7.68313706e-01 1.37779251e-01
5.87143123e-01 -5.31742387e-02 -1.06040619e-01 -3.42720330e-01
-1.16603374e+00 7.03961015e-01 6.73441947e-01 2.81685889e-01
-3.37300003e-01 3.42284888e-01 5.65584183e-01 1.42467037e-01
2.77564913e-01 4.12067682e-01 -4.84759599e-01 6.72830315e-03
-1.16389871e+00 -2.23826896e-02 4.12724346e-01 8.61369848e-01
5.86326838e-01 1.29815608e-01 -8.76634359e-01 9.57561553e-01
2.13856846e-01 3.62693369e-01 5.41900694e-01 -9.49475586e-01
2.10174665e-01 5.38393676e-01 -1.13941498e-01 -2.84957677e-01
-4.54864383e-01 -4.51053262e-01 -6.13406122e-01 2.73407072e-01
2.77552664e-01 4.13996167e-02 -1.27118003e+00 1.80811656e+00
-1.48956841e-02 2.71385998e-01 1.78592876e-01 8.21375608e-01
1.21631491e+00 6.68318987e-01 3.90582293e-01 1.62611336e-01
1.36019170e+00 -1.41391671e+00 -1.20707594e-01 -3.80427629e-01
3.77228767e-01 -6.94954872e-01 1.42855573e+00 1.17583908e-01
-9.71782207e-01 -7.79705882e-01 -8.10555875e-01 -3.26444119e-01
-5.45310736e-01 3.22165817e-01 5.93935490e-01 3.13965589e-01
-1.35220659e+00 2.71084458e-01 -7.17842638e-01 -6.70842409e-01
7.06571996e-01 1.12007551e-01 -2.35295013e-01 -2.28380069e-01
-8.53414178e-01 1.13721275e+00 4.10458148e-01 -1.81994751e-01
-1.41181135e+00 -1.09042037e+00 -6.11767113e-01 -3.67636830e-02
3.33927780e-01 -1.13746595e+00 1.54641199e+00 -8.87124121e-01
-1.35332251e+00 1.13904464e+00 -1.43363610e-01 -5.67647636e-01
4.36913490e-01 -3.06454271e-01 -4.82652709e-02 3.02229792e-01
6.56399056e-02 1.26896000e+00 1.07131577e+00 -1.18967688e+00
-5.08381009e-01 -2.12964371e-01 4.29751039e-01 2.11944580e-01
-4.77731049e-01 1.71702250e-03 -5.59037387e-01 -3.86693120e-01
-5.46740413e-01 -8.04045200e-01 -2.22049281e-01 1.73329413e-01
-2.70661741e-01 -3.14403445e-01 6.98148668e-01 -4.30446148e-01
9.54031289e-01 -2.17076325e+00 1.33991361e-01 -5.74263275e-01
2.12868959e-01 6.81560040e-01 -6.32991254e-01 6.82258129e-01
-1.07716419e-01 -2.40382016e-01 -1.39399394e-01 -4.94461626e-01
7.28682652e-02 2.53151029e-01 -5.22089779e-01 2.41779178e-01
3.43918502e-01 1.30769563e+00 -7.30723441e-01 -2.85498619e-01
5.34417748e-01 5.04634023e-01 -4.78000790e-01 2.26180583e-01
-5.34237564e-01 3.41236532e-01 -2.33526662e-01 7.97253072e-01
3.98333937e-01 -4.98738289e-01 -5.09533346e-01 -3.40099812e-01
-2.43109271e-01 -1.53131321e-01 -4.74618018e-01 2.07630062e+00
-4.00372416e-01 7.08565474e-01 -1.73338681e-01 -1.10304654e+00
7.38181591e-01 2.01040134e-01 2.57370412e-01 -1.01358163e+00
1.07199103e-01 -7.63896629e-02 -2.57560760e-01 -5.83854198e-01
6.00598633e-01 -1.69258684e-01 8.29628576e-03 1.71612069e-01
4.89821464e-01 -4.47256753e-05 1.17612667e-01 5.02533972e-01
1.06300700e+00 3.28414917e-01 1.98962063e-01 -6.98483437e-02
3.62368047e-01 3.65851820e-01 -2.52093822e-02 9.77086604e-01
-4.35001791e-01 7.43225515e-01 1.45973310e-01 -1.07320875e-01
-1.04656422e+00 -1.31232023e+00 -2.05591358e-02 1.64068735e+00
-6.82723671e-02 -4.26367223e-01 -5.93251824e-01 -4.99365956e-01
-4.26812433e-02 7.83626080e-01 -8.56375754e-01 -3.25493902e-01
2.54230388e-02 -5.12033701e-01 8.18806171e-01 8.47521603e-01
4.98613685e-01 -1.10405612e+00 -5.94471574e-01 -6.61089644e-02
-2.13305037e-02 -1.28614497e+00 -3.62649500e-01 3.42145264e-01
-4.71005261e-01 -9.16831315e-01 -1.00630522e+00 -7.51734197e-01
4.32447702e-01 5.39396286e-01 1.08245265e+00 -2.28023499e-01
-4.85645384e-01 9.86308932e-01 -3.51910084e-01 -3.57500792e-01
-2.31735170e-01 9.02768597e-02 -2.98980381e-02 -1.56233937e-01
4.42079484e-01 -5.11475801e-01 -7.90118754e-01 4.67503369e-02
-1.01981735e+00 3.83215696e-01 7.45485902e-01 1.03412056e+00
4.60339844e-01 -6.98667824e-01 5.49305201e-01 -4.64268684e-01
5.24680018e-01 -3.51736039e-01 -5.13303101e-01 4.87208128e-01
-5.09221077e-01 9.48357657e-02 4.69095081e-01 -6.41865909e-01
-1.10506356e+00 1.30162984e-01 -9.44434851e-02 -9.18963492e-01
-3.02644700e-01 4.21684921e-01 2.78062999e-01 -1.55157954e-01
1.09520423e+00 4.94400501e-01 -1.73387900e-01 -2.96142250e-01
9.96072352e-01 5.62123477e-01 9.37936962e-01 -5.55858135e-01
7.75424540e-01 5.36003828e-01 -2.58239836e-01 -7.29147613e-01
-1.24359953e+00 -6.65678740e-01 -6.61268830e-01 -2.64389068e-01
1.11252165e+00 -1.15815318e+00 -6.57042444e-01 6.09109223e-01
-1.19023597e+00 -6.84001923e-01 -2.22592205e-01 1.58853471e-01
-8.00041318e-01 2.01703444e-01 -4.76357847e-01 -5.40705562e-01
-5.31445205e-01 -9.19073045e-01 1.24331152e+00 3.91585082e-01
-6.97364286e-02 -8.74674916e-01 1.40657604e-01 6.09675109e-01
5.67098498e-01 -8.67536813e-02 8.16119015e-01 -6.45542026e-01
-5.50295889e-01 6.10011294e-02 -7.27584600e-01 4.75278229e-01
-2.27806643e-01 -1.31138740e-02 -1.41110206e+00 -3.10754150e-01
-4.85025913e-01 -8.53089809e-01 1.41432810e+00 5.84180176e-01
8.87164891e-01 5.03318124e-02 -1.52740881e-01 9.41587269e-01
1.48447120e+00 -1.28944054e-01 4.26474363e-01 5.36706150e-01
8.00451517e-01 6.17200911e-01 3.74987900e-01 4.32371229e-01
4.75777298e-01 5.73764920e-01 5.21419704e-01 -1.19761549e-01
-4.90735739e-01 -3.68254304e-01 5.73244691e-01 3.51635039e-01
1.26336038e-01 -4.00386415e-02 -1.08499050e+00 7.31134713e-01
-1.94944072e+00 -1.02311575e+00 1.78123206e-01 1.78165162e+00
6.48575544e-01 -7.70398229e-02 2.03969061e-01 -4.74649340e-01
4.04241592e-01 2.65811861e-01 -7.69662321e-01 -3.88565928e-01
-2.79067576e-01 2.29930431e-02 4.14923012e-01 2.16578171e-01
-9.59207773e-01 1.25192499e+00 6.03880692e+00 9.27421808e-01
-1.26168084e+00 3.51463556e-01 3.87857288e-01 -3.58330399e-01
-2.32256681e-01 -5.84899001e-02 -9.94976938e-01 7.08193332e-02
9.07344580e-01 -1.55626222e-01 4.45871204e-01 7.18960047e-01
3.17885354e-02 4.80625927e-02 -1.22279763e+00 1.06672502e+00
3.42713535e-01 -1.67501295e+00 2.56389260e-01 -2.58221835e-01
7.79155135e-01 7.19782114e-01 3.31304550e-01 7.25579798e-01
2.89989054e-01 -1.05276656e+00 7.73433149e-01 5.60593247e-01
9.45843577e-01 -3.20628673e-01 1.06284536e-01 3.31161827e-01
-1.10517967e+00 -3.63253385e-01 -4.46945280e-01 8.52983817e-02
3.00455149e-02 -1.25048012e-02 -6.81434214e-01 4.54672128e-01
8.05175841e-01 7.93019772e-01 -8.63935351e-01 1.10587180e+00
-1.25295758e-01 4.55164701e-01 -1.71742700e-02 2.01461360e-01
5.87465465e-01 2.39154100e-01 5.44719338e-01 1.36594307e+00
1.26878038e-01 4.30479050e-02 2.35380486e-01 1.00573480e+00
1.19505860e-01 -1.65825844e-01 -6.74903691e-01 -3.15191478e-01
3.20825726e-01 1.24354851e+00 -4.05103028e-01 -4.34868753e-01
-7.67084777e-01 9.40069675e-01 5.62232077e-01 5.57217479e-01
-8.33833098e-01 2.08903074e-01 4.70486790e-01 -1.38466805e-01
5.69860935e-01 -2.11300448e-01 -3.22275490e-01 -1.07668281e+00
-2.36746684e-01 -6.37230277e-01 6.04025126e-01 -1.27811134e+00
-1.50517166e+00 7.52888680e-01 4.53337468e-02 -1.18758631e+00
-1.70883685e-01 -9.33462799e-01 -7.01340437e-01 7.72575617e-01
-1.60507905e+00 -1.93742287e+00 -4.76336867e-01 1.05446374e+00
7.99969673e-01 -3.87749523e-01 9.05109227e-01 -4.95664068e-02
-4.82517391e-01 7.49539077e-01 5.39657474e-02 -3.65346670e-02
9.05651450e-01 -1.06532180e+00 2.36166924e-01 9.28739071e-01
2.87583143e-01 4.01086390e-01 6.07924640e-01 -2.82503873e-01
-1.39409149e+00 -1.12943280e+00 3.91456306e-01 -6.95371270e-01
1.00055051e+00 -1.52046442e-01 -7.81545520e-01 7.27464139e-01
5.59045434e-01 1.36512563e-01 5.07171810e-01 9.64481011e-02
-9.24701989e-01 -1.21527761e-01 -6.09832644e-01 7.48236120e-01
1.04490519e+00 -9.05142426e-01 -9.22114491e-01 3.36221904e-01
8.74311626e-01 -1.94786072e-01 -5.35937667e-01 3.19464833e-01
4.31950241e-01 -9.97670770e-01 1.17760670e+00 -6.09671175e-01
6.82100594e-01 -1.12199694e-01 -3.89145076e-01 -1.20851326e+00
-5.78554273e-01 -4.43641484e-01 -9.48135555e-02 1.32949436e+00
3.92499924e-01 -3.61336321e-01 5.64034998e-01 3.74650925e-01
-4.34079796e-01 -7.66750336e-01 -6.73051178e-01 -9.75852489e-01
1.17481261e-01 -5.22127151e-01 -3.74285057e-02 7.63549566e-01
-1.77357078e-01 7.43261695e-01 -4.09936756e-01 1.77202240e-01
5.22517622e-01 1.84790179e-01 7.31124878e-01 -9.20602322e-01
-5.38489044e-01 -7.19445407e-01 -1.49858415e-01 -9.60683048e-01
1.08975403e-01 -1.12535346e+00 -2.59241499e-02 -1.86407840e+00
5.83640158e-01 2.86703315e-02 -5.41773438e-01 9.54636335e-01
3.96643505e-02 3.30305755e-01 4.72200334e-01 2.27415368e-01
-8.54713619e-01 7.65268326e-01 1.25024545e+00 -3.16047609e-01
-1.77400634e-01 -2.95576751e-01 -1.01440132e+00 4.79609311e-01
6.85903192e-01 -7.98965767e-02 -5.89389026e-01 -6.51337326e-01
-1.19345859e-01 -5.64136170e-02 7.55389392e-01 -9.21245813e-01
4.62069213e-01 -5.76154739e-02 3.09826285e-01 -4.99799937e-01
5.44601083e-01 -4.88233656e-01 -2.71133542e-01 2.39306942e-01
-5.09008884e-01 -1.95958540e-01 6.68827534e-01 4.10931349e-01
-1.57452524e-01 -1.06067555e-02 9.36896026e-01 -1.91340476e-01
-1.39086390e+00 4.65598166e-01 -2.40225419e-01 7.81676024e-02
8.99316251e-01 -2.81408042e-01 -1.04186440e+00 -1.68868855e-01
-7.07502544e-01 2.83116549e-01 3.55311155e-01 7.87997663e-01
6.45552158e-01 -1.05686963e+00 -8.23852301e-01 -1.41769305e-01
7.10271120e-01 -4.03767318e-01 6.64482474e-01 1.11936605e+00
-1.25586027e-02 5.52706718e-01 -5.44612408e-01 -8.56929421e-01
-1.26979864e+00 8.39003146e-01 3.97519469e-01 8.69569741e-03
-6.82183743e-01 1.19777930e+00 7.36327052e-01 -3.16173673e-01
2.23050430e-01 -1.97309721e-02 -1.15197480e-01 1.28322557e-01
7.02110529e-01 4.09944467e-02 -1.07891232e-01 -4.38236475e-01
-3.54024827e-01 5.47290504e-01 -3.50226939e-01 -1.56878218e-01
1.29746687e+00 -1.41392827e-01 1.54221639e-01 4.45567876e-01
1.01033056e+00 -4.64624703e-01 -1.66371584e+00 -4.93015230e-01
-2.10481912e-01 -9.01007801e-02 1.40250340e-01 -1.20563126e+00
-8.75712097e-01 1.08459187e+00 7.16765165e-01 -1.97180584e-01
1.32565892e+00 3.82958084e-01 5.00033081e-01 4.58968222e-01
1.31101981e-01 -7.34626353e-01 4.87613231e-01 6.48663402e-01
1.02469397e+00 -1.44648302e+00 -3.36024761e-01 1.40664071e-01
-9.50031281e-01 7.92155862e-01 1.01765466e+00 -1.83379009e-01
2.87995845e-01 1.64052621e-01 2.34987453e-01 -1.41359866e-01
-1.12167358e+00 -6.67778373e-01 6.11270785e-01 9.02760565e-01
2.39649385e-01 -1.53875142e-01 2.34036446e-01 6.43538833e-01
-3.68646644e-02 5.48709147e-02 2.19899163e-01 5.98710239e-01
-5.72746634e-01 -7.65574694e-01 -1.76519319e-01 3.04049402e-01
-1.00085519e-01 -4.40357566e-01 -3.95694673e-01 6.86796725e-01
3.87545489e-03 8.99321735e-01 -1.60145879e-01 -4.39082891e-01
2.83179283e-01 6.74153194e-02 6.40877128e-01 -6.37699962e-01
-6.09582961e-01 -4.35630158e-02 -3.82643677e-02 -7.43839800e-01
-3.85270476e-01 -3.49758476e-01 -1.00788176e+00 -9.02511477e-02
1.41399592e-01 -3.83423150e-01 3.95611167e-01 1.08901429e+00
4.52937305e-01 6.32678211e-01 1.51207238e-01 -1.07801878e+00
-6.56660736e-01 -9.78022635e-01 -2.29636580e-01 3.17629933e-01
5.17345607e-01 -6.80299580e-01 -3.94164175e-02 2.15085655e-01] | [10.306297302246094, 1.923471212387085] |
061f0519-268b-4743-83f1-441c5e39dcc7 | generalizable-person-re-identification-by | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Song_Generalizable_Person_Re-Identification_by_Domain-Invariant_Mapping_Network_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Song_Generalizable_Person_Re-Identification_by_Domain-Invariant_Mapping_Network_CVPR_2019_paper.pdf | Generalizable Person Re-Identification by Domain-Invariant Mapping Network | We aim to learn a domain generalizable person re-identification (ReID) model. When such a model is trained on a set of source domains (ReID datasets collected from different camera networks), it can be directly applied to any new unseen dataset for effective ReID without any model updating. Despite its practical value in real-world deployments, generalizable ReID has seldom been studied. In this work, a novel deep ReID model termed Domain-Invariant Mapping Network (DIMN) is proposed. DIMN is designed to learn a mapping between a person image and its identity classifier, i.e., it produces a classifier using a single shot. To make the model domain-invariant, we follow a meta-learning pipeline and sample a subset of source domain training tasks during each training episode. However, the model is significantly different from conventional meta-learning methods in that: (1) no model updating is required for the target domain, (2) different training tasks share a memory bank for maintaining both scalability and discrimination ability, and (3) it can be used to match an arbitrary number of identities in a target domain. Extensive experiments on a newly proposed large-scale ReID domain generalization benchmark show that our DIMN significantly outperforms alternative domain generalization or meta-learning methods.
| [' Timothy M. Hospedales', ' Tao Xiang', ' Yi-Zhe Song', ' Yongxin Yang', 'Jifei Song'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['generalizable-person-re-identification'] | ['computer-vision'] | [ 1.88699067e-01 -3.33923548e-01 -2.51412541e-01 -6.85929954e-01
-2.40972862e-01 -6.63525641e-01 7.63989747e-01 2.98945364e-02
-5.37226915e-01 7.70308137e-01 -1.85593925e-02 1.99616402e-01
1.20470226e-02 -8.18260193e-01 -7.69957602e-01 -4.10576820e-01
2.43777528e-01 8.21834266e-01 2.40886196e-01 -2.00266510e-01
-2.68430468e-02 5.34075975e-01 -1.64940679e+00 1.17358871e-01
9.05891240e-01 6.70054376e-01 1.17166094e-01 2.38711596e-01
-5.99026121e-02 4.88378644e-01 -8.90329182e-01 -8.91257942e-01
5.98418713e-01 -4.88783062e-01 -8.80463004e-01 -1.61196545e-01
7.88936496e-01 -7.59407401e-01 -7.51934528e-01 1.08314526e+00
5.94443023e-01 3.53501797e-01 7.04031408e-01 -1.52907121e+00
-1.21966374e+00 2.97949910e-01 -3.30744475e-01 2.37466305e-01
3.83983344e-01 1.91355488e-04 2.77348936e-01 -5.47723651e-01
7.58890450e-01 1.29241800e+00 6.95813477e-01 1.06162798e+00
-1.19038188e+00 -1.30173075e+00 3.10166955e-01 3.66990298e-01
-1.51436102e+00 -4.03485090e-01 4.82800692e-01 -2.45391250e-01
7.97659874e-01 -1.43696815e-01 4.32004392e-01 1.67082727e+00
-6.73434958e-02 6.56911433e-01 9.83663976e-01 -9.58075598e-02
2.57125139e-01 3.77740592e-01 2.01010674e-01 9.06729102e-02
4.53152180e-01 2.02992365e-01 -6.66715860e-01 -1.48592874e-01
7.18289435e-01 2.81431764e-01 -1.73069146e-02 -5.70779026e-01
-1.10465205e+00 4.63660806e-01 4.60420907e-01 3.23833339e-02
-7.54031613e-02 -2.23120093e-01 6.70882404e-01 5.87550759e-01
4.66824830e-01 3.17647368e-01 -2.13269383e-01 1.12060130e-01
-5.47868192e-01 4.29168254e-01 6.67700887e-01 1.38017094e+00
9.32048857e-01 -2.60733157e-01 -2.66575497e-02 9.63025391e-01
-3.87189925e-01 6.05368793e-01 7.06135094e-01 -5.81779420e-01
3.89658779e-01 7.23166883e-01 2.18185067e-01 -6.54130697e-01
-2.02097893e-01 -3.67293000e-01 -1.10903037e+00 -1.05592832e-01
3.00353348e-01 -2.44275555e-02 -8.26834023e-01 1.93672252e+00
3.35084945e-01 8.62324357e-01 2.44652689e-01 6.35037124e-01
1.03268707e+00 3.25930983e-01 2.62057424e-01 2.79915899e-01
1.16960287e+00 -7.32106388e-01 -3.44983518e-01 -4.95674103e-01
3.67111564e-01 -2.27554470e-01 5.50622225e-01 -3.92243080e-02
-7.96068490e-01 -9.83451664e-01 -1.19068480e+00 -1.08859636e-01
-7.99807191e-01 -1.69372350e-01 4.33536947e-01 6.73576176e-01
-1.00594223e+00 3.23063105e-01 -3.72817248e-01 -8.66017044e-01
4.65760052e-01 6.60404384e-01 -8.49249482e-01 -4.76919323e-01
-1.16307163e+00 8.34083557e-01 7.33730912e-01 -3.27312917e-01
-9.04527783e-01 -8.88330936e-01 -1.01286316e+00 2.16701580e-03
-6.34978786e-02 -9.83950615e-01 1.24542785e+00 -1.14316797e+00
-1.32061672e+00 1.33271241e+00 -4.06980067e-01 -4.95340109e-01
5.98094165e-01 -1.06207214e-01 -9.07775342e-01 -8.31687823e-02
3.27743351e-01 8.79195988e-01 9.42547023e-01 -1.29599094e+00
-9.53654230e-01 -4.56853062e-01 1.99182287e-01 1.45936489e-01
-6.81634665e-01 -2.86689885e-02 -4.69555110e-01 -7.13582277e-01
-1.23375386e-01 -1.02534056e+00 3.01350892e-01 -1.36736915e-01
-3.63937467e-02 -3.44470233e-01 7.91144609e-01 -5.50687551e-01
7.02788591e-01 -2.16044188e+00 -1.66977253e-02 2.98310276e-02
2.40608364e-01 5.65978408e-01 -3.69852096e-01 2.86027670e-01
-2.65802950e-01 -1.79802239e-01 3.95374082e-04 -4.91082221e-01
-1.61067158e-01 2.29847446e-01 -3.39142233e-01 4.73891258e-01
4.47208323e-02 9.84590709e-01 -1.12807131e+00 -2.01041415e-01
1.86942443e-01 2.57518768e-01 -2.77758062e-01 4.00821209e-01
3.18216801e-01 6.08617902e-01 -8.66828188e-02 5.26539028e-01
1.16950715e+00 -8.02137628e-02 2.02396318e-01 -4.44325991e-02
1.90793067e-01 -8.75640586e-02 -1.21798050e+00 1.79554057e+00
-3.99160922e-01 4.98140752e-01 -3.27867091e-01 -1.12684536e+00
1.19297445e+00 3.80838476e-02 2.13311687e-01 -1.08995247e+00
-1.56014934e-01 1.35468006e-01 -2.17558160e-01 -1.95011366e-02
5.03723323e-01 7.15589523e-02 -2.45908052e-01 5.40215611e-01
3.95110309e-01 6.83425844e-01 1.66836813e-01 9.67985764e-02
9.87343252e-01 -7.54371956e-02 3.61131966e-01 -1.33088201e-01
6.73878133e-01 1.01699859e-01 8.57045472e-01 1.13648200e+00
-4.09211338e-01 4.46672648e-01 -9.76317376e-02 -7.88381994e-01
-1.14945567e+00 -1.37383008e+00 -2.16741428e-01 1.44185984e+00
8.40970516e-01 -1.73164736e-02 -6.91249847e-01 -7.49086201e-01
4.10194814e-01 4.28901523e-01 -7.73066461e-01 -5.99058509e-01
-7.34678507e-01 -5.73511600e-01 6.78897679e-01 7.36600637e-01
1.18817472e+00 -6.95791066e-01 -6.07601404e-02 1.57798678e-01
-7.19913170e-02 -1.18637741e+00 -7.30540633e-01 -3.21072996e-01
-6.68779552e-01 -1.20348537e+00 -9.44868684e-01 -1.23473585e+00
8.64832044e-01 8.32695723e-01 1.05029440e+00 -1.09147541e-02
-7.85846263e-02 5.62350452e-01 -3.12204331e-01 -4.92820770e-01
-4.82557118e-01 1.59179375e-01 6.17162824e-01 2.04312444e-01
1.03180218e+00 -5.51793754e-01 -5.93596458e-01 7.31743634e-01
-6.41671538e-01 -7.03046173e-02 5.56828678e-01 1.06746066e+00
2.84069955e-01 4.98250388e-02 1.11278105e+00 -8.94315720e-01
6.00465894e-01 -6.61960065e-01 -5.10478199e-01 5.55282652e-01
-4.36967880e-01 -2.72533655e-01 6.78450167e-01 -8.70074332e-01
-1.40245342e+00 -3.75512871e-03 2.50433534e-01 -3.59507054e-01
-3.32175851e-01 -1.95420444e-01 -3.88058245e-01 -1.73271716e-01
7.61516333e-01 5.50068915e-01 7.49929249e-02 -5.82527041e-01
2.99409151e-01 9.70060945e-01 1.18872786e+00 -6.13520026e-01
1.21489048e+00 6.39261484e-01 -3.54305208e-01 -4.51304287e-01
-5.84266663e-01 -6.96448326e-01 -1.19410598e+00 -6.12453222e-02
5.87988019e-01 -1.37030470e+00 -7.83468723e-01 8.52229655e-01
-1.03100777e+00 -1.69481799e-01 1.47840893e-02 4.07291353e-01
-2.49578312e-01 3.51606637e-01 -3.59991342e-01 -3.93082649e-01
-3.56825233e-01 -6.72815084e-01 8.80543590e-01 6.68546557e-01
-1.96489245e-01 -1.06359076e+00 3.31447981e-02 2.26220503e-01
4.70378816e-01 1.01988511e-02 7.24910259e-01 -1.05491149e+00
-2.61729628e-01 -4.35240984e-01 -5.06009877e-01 3.42342913e-01
3.33624750e-01 -7.16755211e-01 -1.16407931e+00 -8.22775185e-01
-4.29446638e-01 -3.28921914e-01 6.38830066e-01 -1.41446352e-01
1.12573659e+00 -1.57613114e-01 -7.32037961e-01 8.83756757e-01
1.08930469e+00 2.34215200e-01 4.20484781e-01 7.23954499e-01
4.52823550e-01 3.93308103e-01 4.10731554e-01 2.28519633e-01
6.48488581e-01 8.05348814e-01 6.21055514e-02 -1.15514979e-01
-2.79263139e-01 -6.24071777e-01 2.22336859e-01 6.15201853e-02
5.12515344e-02 -5.09281978e-02 -7.92212427e-01 7.42844403e-01
-1.92302465e+00 -1.15042150e+00 4.61001635e-01 2.44799137e+00
5.42668104e-01 -5.61794378e-02 4.42846268e-01 -4.56164330e-01
1.21909702e+00 -2.94941097e-01 -1.27449977e+00 -1.28843755e-01
-2.36272559e-01 1.38024643e-01 5.57351172e-01 -7.74901882e-02
-1.20427644e+00 8.93456340e-01 6.23195934e+00 5.17220438e-01
-9.30769563e-01 3.15202653e-01 1.46882117e-01 -1.13046914e-01
2.26638913e-01 -2.61767149e-01 -1.06326258e+00 6.25590384e-01
7.06318796e-01 -7.91050553e-01 4.01486635e-01 1.14203584e+00
-4.90537852e-01 2.80575901e-01 -1.50019240e+00 1.40703762e+00
2.84661740e-01 -1.24075854e+00 3.92472267e-01 -6.21904209e-02
7.76024759e-01 -9.80866253e-02 1.56489104e-01 8.21510553e-01
4.34301585e-01 -9.35833693e-01 4.89224523e-01 2.32808188e-01
9.99434769e-01 -9.08203900e-01 9.28723454e-01 4.69421536e-01
-1.15926683e+00 -3.75816226e-01 -8.85248780e-01 7.79165551e-02
-1.24299429e-01 -1.89052716e-01 -7.86825359e-01 6.80519640e-01
9.82893407e-01 9.48122621e-01 -7.87001669e-01 1.11164439e+00
1.16081923e-01 -9.26149078e-03 -9.00446102e-02 4.90966320e-01
-1.41699970e-01 6.54225498e-02 2.82340795e-01 1.21838474e+00
2.46024236e-01 5.54619469e-02 1.04454115e-01 7.91628540e-01
-4.57475334e-01 -2.58350819e-01 -6.97023332e-01 5.37706494e-01
1.00675595e+00 9.31025147e-01 -2.05416620e-01 -4.43199426e-01
-6.02874577e-01 1.48594165e+00 4.97396290e-01 5.65154016e-01
-7.25441873e-01 -3.16996902e-01 1.29801154e+00 2.51342729e-02
6.66571632e-02 1.30192414e-01 1.16982944e-01 -1.30132687e+00
4.72713113e-02 -7.00057566e-01 7.56315351e-01 -4.28319126e-01
-1.90955901e+00 5.66844046e-01 3.33274037e-01 -1.37687433e+00
-3.76241982e-01 -5.94205856e-01 -6.48895919e-01 1.17634714e+00
-1.63709068e+00 -1.48468924e+00 -7.96031177e-01 1.07890975e+00
3.89779776e-01 -7.44456947e-01 9.31732178e-01 4.51590002e-01
-7.73202240e-01 1.27918172e+00 3.93000752e-01 6.72811925e-01
1.26470017e+00 -1.01463389e+00 9.47266459e-01 9.00733888e-01
-3.19072157e-01 1.00573111e+00 2.68140912e-01 -7.73354948e-01
-1.38106048e+00 -1.51058662e+00 6.02599621e-01 -6.67949080e-01
1.47628963e-01 -5.72132528e-01 -1.09771645e+00 1.06355846e+00
-6.99480847e-02 1.10529244e-01 8.88693511e-01 2.10744604e-01
-8.41319144e-01 -4.80469525e-01 -1.48391068e+00 3.24284226e-01
1.62878239e+00 -6.36511803e-01 -7.73055315e-01 3.47558230e-01
3.95133764e-01 -5.88099480e-01 -8.81564558e-01 2.55580366e-01
6.44223273e-01 -8.72981131e-01 1.31384420e+00 -9.21467721e-01
-9.13971961e-02 -1.96110979e-01 1.76817462e-01 -1.35268092e+00
-7.10042000e-01 -2.09200293e-01 -1.76034123e-01 1.49530292e+00
-1.41497910e-01 -1.07046092e+00 6.87122464e-01 8.97985816e-01
1.43188596e-01 -7.62595097e-03 -1.04923844e+00 -1.51730657e+00
2.62809068e-01 5.62301800e-02 1.28382301e+00 1.08872783e+00
-2.12021053e-01 9.62762088e-02 -5.02369821e-01 5.41478574e-01
9.34701085e-01 1.09019473e-01 1.28798962e+00 -1.59542286e+00
2.87071038e-02 -2.84925550e-01 -6.92206800e-01 -1.18725538e+00
6.75956547e-01 -1.13025641e+00 -5.16505361e-01 -1.25595927e+00
5.91502905e-01 -5.86544454e-01 -4.36283141e-01 5.15580714e-01
-1.70427516e-01 1.37936309e-01 1.42632097e-01 6.22678459e-01
-6.09228194e-01 3.19627702e-01 7.13623881e-01 -4.62295890e-01
-1.94576845e-01 1.02306031e-01 -9.52880919e-01 4.63575840e-01
6.01354837e-01 -3.99106145e-01 -5.45656800e-01 -5.76170027e-01
-3.43127847e-01 -4.89709526e-01 7.05402553e-01 -1.19564807e+00
6.37012422e-01 -4.47173640e-02 9.54956055e-01 -2.83168167e-01
1.97208315e-01 -6.81142867e-01 3.79951984e-01 2.85866976e-01
-6.74492493e-02 3.80506143e-02 3.21570069e-01 6.64866984e-01
-7.43484795e-02 -7.64992237e-02 9.39090490e-01 -9.97225642e-02
-1.55933774e+00 5.83766580e-01 2.87204385e-01 -3.96890529e-02
1.21758652e+00 -5.31366944e-01 -6.94506705e-01 -2.24693790e-01
-3.68585914e-01 3.79158586e-01 9.97192025e-01 1.03778243e+00
4.83518749e-01 -1.47127712e+00 -8.70450318e-01 3.85279328e-01
8.01316798e-01 -1.14005394e-01 5.48059464e-01 1.63581967e-02
-1.29833043e-01 1.59454554e-01 -5.81329465e-01 -6.16529524e-01
-1.14594245e+00 8.72734487e-01 6.11948133e-01 1.29615545e-01
-8.11672270e-01 7.06125021e-01 5.42679012e-01 -7.24000990e-01
1.88490033e-01 4.18935299e-01 -3.30248512e-02 -1.77294895e-01
1.20048928e+00 4.25994933e-01 -7.05728233e-02 -8.23673725e-01
-4.81625438e-01 4.57428336e-01 -5.44479370e-01 3.44614387e-01
1.10837674e+00 -1.50913373e-01 3.31843123e-02 -8.31784755e-02
1.13746190e+00 -5.98332405e-01 -1.35106552e+00 -6.51895940e-01
-1.33297041e-01 -7.41616964e-01 -4.88246948e-01 -7.75532961e-01
-6.72985911e-01 4.11291629e-01 8.78969014e-01 -3.28091711e-01
1.15880775e+00 -2.94626635e-02 1.00329757e+00 6.61022246e-01
7.44693995e-01 -1.19388211e+00 -7.29786754e-02 3.17884088e-01
5.70226073e-01 -1.40811169e+00 -1.72594771e-01 -8.93675536e-02
-5.57818234e-01 9.39668417e-01 1.04969990e+00 3.96471433e-02
4.00964379e-01 -1.84124470e-01 -5.58012724e-02 3.33659112e-01
-2.69678056e-01 -7.07205310e-02 5.76603152e-02 1.42439282e+00
-1.47153273e-01 -3.16036381e-02 2.48661354e-01 7.90766954e-01
-2.31490701e-01 2.02602431e-01 2.68960208e-01 7.35362053e-01
-1.00819968e-01 -1.30822897e+00 -2.40851134e-01 2.84580410e-01
1.63923621e-01 2.73016840e-01 -4.14106846e-01 8.98882329e-01
3.74762237e-01 9.75233436e-01 2.47860119e-01 -4.70378608e-01
6.58354819e-01 1.65249243e-01 4.30752933e-01 -7.48699844e-01
-4.30246592e-01 -9.86379743e-01 -3.98089468e-01 -3.40294093e-01
-3.34643483e-01 -6.94053233e-01 -8.24854732e-01 -8.31261754e-01
2.01126352e-01 -8.47705901e-02 3.47286075e-01 8.70356798e-01
7.23022699e-01 -7.83554930e-03 5.39975405e-01 -7.24297464e-01
-5.09191751e-01 -7.42650390e-01 -6.77784383e-01 9.30554688e-01
2.35842139e-01 -9.63336885e-01 -7.50344917e-02 1.16354115e-01] | [14.7474946975708, 1.0804717540740967] |
5178dfe1-2c4d-47d9-991f-a1e93fc6bd13 | balpa-a-balanced-primal-dual-algorithm-for | 2212.02835 | null | https://arxiv.org/abs/2212.02835v1 | https://arxiv.org/pdf/2212.02835v1.pdf | BALPA: A Balanced Primal-Dual Algorithm for Nonsmooth Optimization with Application to Distributed Optimization | In this paper, we propose a novel primal-dual proximal splitting algorithm (PD-PSA), named BALPA, for the composite optimization problem with equality constraints, where the loss function consists of a smooth term and a nonsmooth term composed with a linear mapping. In BALPA, the dual update is designed as a proximal point for a time-varying quadratic function, which balances the implementation of primal and dual update and retains the proximity-induced feature of classic PD-PSAs. In addition, by this balance, BALPA eliminates the inefficiency of classic PD-PSAs for composite optimization problems in which the Euclidean norm of the linear mapping or the equality constraint mapping is large. Therefore, BALPA not only inherits the advantages of simple structure and easy implementation of classic PD-PSAs but also ensures a fast convergence when these norms are large. Moreover, we propose a stochastic version of BALPA (S-BALPA) and apply the developed BALPA to distributed optimization to devise a new distributed optimization algorithm. Furthermore, a comprehensive convergence analysis for BALPA and S-BALPA is conducted, respectively. Finally, numerical experiments demonstrate the efficiency of the proposed algorithms. | ['Shaofu Yang', 'Xinli Shi', 'Jinde Cao', 'Luyao Guo'] | 2022-12-06 | null | null | null | null | ['distributed-optimization'] | ['methodology'] | [-4.29386616e-01 -2.01365709e-01 1.34711564e-01 6.90281158e-04
-9.57668483e-01 -2.82641172e-01 3.09735853e-02 8.90214071e-02
-2.82799125e-01 1.07992744e+00 5.70030995e-02 -1.23276472e-01
-5.53672016e-01 -6.91308558e-01 -8.40757251e-01 -1.20061493e+00
1.28702059e-01 2.55644977e-01 6.64432943e-02 -4.51429754e-01
-3.41519602e-02 1.92579374e-01 -9.15567458e-01 -4.43011820e-01
1.37894559e+00 1.21749640e+00 2.22091407e-01 1.22507520e-01
1.77372828e-01 5.43127954e-01 -1.67718589e-01 -1.84678227e-01
6.22768998e-01 -4.86816406e-01 -3.04283470e-01 2.50327528e-01
9.18337107e-02 -2.97411293e-01 -2.16204524e-01 1.09428215e+00
8.04628491e-01 4.22155410e-01 3.20567101e-01 -1.46085036e+00
-3.37812543e-01 3.22217345e-01 -1.28174698e+00 9.65248942e-02
7.36642331e-02 -6.10437877e-02 1.04472959e+00 -1.13460445e+00
2.04424232e-01 1.22026789e+00 9.09236372e-01 -8.19043890e-02
-1.25774467e+00 -4.65592057e-01 2.33040839e-01 4.22906354e-02
-1.56739306e+00 -2.01806396e-01 7.98789442e-01 -3.66266280e-01
2.64936894e-01 4.31854308e-01 5.39729297e-01 1.90371096e-01
3.30907553e-02 9.30991471e-01 9.30469692e-01 -2.62357503e-01
2.55804896e-01 1.66378561e-02 -5.95765486e-02 7.74651766e-01
4.04438943e-01 -2.42612928e-01 -4.24961895e-02 -5.57852387e-01
5.86351454e-01 5.44901825e-02 -6.11721277e-01 -6.85312688e-01
-1.12424171e+00 8.37913096e-01 2.77474761e-01 1.29898801e-01
-6.09312534e-01 -2.10016102e-01 5.60046732e-01 1.30081445e-01
5.66428602e-01 -2.05769047e-01 -9.16999876e-02 -4.07782383e-02
-7.81937480e-01 5.09603083e-01 8.51673186e-01 9.48250473e-01
8.07367086e-01 2.39986315e-01 -3.12328756e-01 1.05149293e+00
4.23168302e-01 6.39507771e-01 3.95355344e-01 -9.58845437e-01
9.55370784e-01 3.77185851e-01 2.93182760e-01 -1.50490916e+00
-2.86694378e-01 -7.01263189e-01 -1.05388010e+00 1.09414890e-01
3.87523979e-01 -5.04523933e-01 1.25845566e-01 1.74136722e+00
7.43451774e-01 2.30294093e-01 -7.48843104e-02 1.23049223e+00
1.02902465e-02 1.12908816e+00 -4.26765084e-01 -7.41783857e-01
8.62457156e-01 -1.13848305e+00 -6.84434474e-01 6.20630104e-03
2.93447912e-01 -6.71281993e-01 8.65090966e-01 3.00758898e-01
-1.27724302e+00 -3.46421808e-01 -1.03174233e+00 7.61561692e-02
3.76018494e-01 2.45736316e-01 2.36971095e-01 4.59128410e-01
-8.42424989e-01 4.26406354e-01 -5.55269837e-01 2.39030793e-02
1.01013042e-01 3.67276520e-01 -1.42735288e-01 1.32305428e-01
-1.11693072e+00 3.78863603e-01 3.48169178e-01 5.86813867e-01
-6.05763078e-01 -9.00956035e-01 -6.14992142e-01 2.11434394e-01
4.46305335e-01 -6.46493912e-01 8.25978816e-01 -1.00131047e+00
-1.71112800e+00 3.30414176e-01 -1.00268692e-01 -1.63881794e-01
9.48782504e-01 -8.01260248e-02 -4.98886257e-02 -1.43443746e-02
3.78225863e-01 -2.80181348e-01 9.09455538e-01 -1.17511749e+00
-7.75371909e-01 -3.83115888e-01 5.52625060e-02 5.24545729e-01
-5.41317463e-01 -3.20695698e-01 -3.76937687e-01 -7.55906641e-01
-3.48562971e-02 -7.35772789e-01 -3.68577778e-01 1.97758213e-01
-2.53056109e-01 -8.91451985e-02 8.74713063e-01 -8.69387984e-01
1.49077475e+00 -2.50456429e+00 5.16322017e-01 4.03916568e-01
1.80037096e-01 1.75960749e-01 -3.32985446e-02 6.69658661e-01
-1.86718558e-03 -3.08820009e-01 -5.00350595e-01 -3.66886556e-01
1.78506538e-01 2.36866012e-01 -2.75153816e-01 9.20642674e-01
-2.91309446e-01 2.63124257e-01 -1.01185870e+00 -4.81800109e-01
3.18571180e-02 1.73109174e-01 -6.03209734e-01 -8.45149253e-03
6.53094798e-02 2.93384820e-01 -7.55090773e-01 4.12668854e-01
1.11420941e+00 -4.07120064e-02 6.67731762e-02 -1.90335348e-01
-5.03262460e-01 -2.89590180e-01 -1.72401273e+00 1.46006608e+00
-5.55700302e-01 -8.59056339e-02 1.01426339e+00 -1.48348045e+00
1.00792193e+00 2.73237884e-01 9.07355905e-01 -4.89511669e-01
3.80771048e-02 5.60996771e-01 -3.75149339e-01 -3.23720902e-01
1.23696044e-01 -2.99989402e-01 2.39477366e-01 3.50154519e-01
-3.95310521e-01 2.62248039e-01 3.83212805e-01 1.03352793e-01
7.46257246e-01 -3.49433743e-03 3.43476951e-01 -7.36944854e-01
1.36349916e+00 -2.95459330e-01 1.30585217e+00 3.13561797e-01
-4.19167668e-01 4.95583862e-01 6.33247554e-01 -1.36729673e-01
-8.18132877e-01 -8.89064908e-01 -1.63295150e-01 9.15769756e-01
6.17512226e-01 -2.85937697e-01 -3.77852440e-01 -6.84558570e-01
2.83975691e-01 2.75089145e-01 -1.31288931e-01 7.24421367e-02
-6.64427102e-01 -9.48883057e-01 1.43174931e-01 2.17480391e-01
8.38313520e-01 -2.21513033e-01 1.80192962e-01 3.92427772e-01
-3.37872952e-01 -7.55200386e-01 -1.01065791e+00 -2.39180982e-01
-7.73984432e-01 -8.76677692e-01 -1.15907991e+00 -1.14454806e+00
8.15595865e-01 4.91317928e-01 5.22492230e-01 -1.85318947e-01
1.62339419e-01 2.70571202e-01 -1.25849828e-01 -1.47387609e-01
1.27876222e-01 -2.00744152e-01 1.21515885e-01 7.61676431e-01
-4.13859099e-01 -6.67177081e-01 -7.00084388e-01 6.50701106e-01
-7.04372108e-01 -1.93575432e-03 4.29371685e-01 1.23609257e+00
8.33483219e-01 1.97981983e-01 7.28286922e-01 -5.02216220e-01
8.41373742e-01 -8.52504313e-01 -8.87609661e-01 1.40170842e-01
-6.51974857e-01 -1.38734549e-01 1.12239850e+00 -3.26230556e-01
-1.20969009e+00 2.43112128e-02 -9.49544832e-02 -5.49321532e-01
6.64968193e-01 6.87854707e-01 -3.20096135e-01 -3.61514300e-01
3.66519064e-01 4.74214166e-01 3.51802617e-01 -4.55401361e-01
2.24071026e-01 5.16511798e-01 2.90441126e-01 -9.62888122e-01
8.58664393e-01 7.15484500e-01 1.27681002e-01 -8.00920963e-01
-6.09304368e-01 -6.46930873e-01 -8.92761722e-02 -2.54934523e-02
2.10908413e-01 -8.30746472e-01 -8.25678468e-01 6.20155871e-01
-9.51682925e-01 -9.12203640e-03 -4.10559058e-01 5.00016153e-01
-7.08872676e-01 8.44564199e-01 -5.08729637e-01 -8.29663813e-01
-4.64491189e-01 -9.27066863e-01 7.05782950e-01 2.35748351e-01
3.43971223e-01 -1.21619248e+00 3.40211481e-01 1.53745458e-01
3.19099396e-01 4.48658824e-01 3.95888209e-01 -1.09090105e-01
-3.42555732e-01 -3.53231058e-02 -3.64685863e-01 6.47875905e-01
1.82801872e-01 -1.70480818e-01 -2.22608879e-01 -8.25834990e-01
4.44143683e-01 -1.33732527e-01 8.71613622e-02 4.63600695e-01
9.01361823e-01 -6.33007586e-01 -2.88248241e-01 8.43911350e-01
1.90281928e+00 1.76814541e-01 1.72044784e-01 4.66585368e-01
5.39666057e-01 3.19121033e-01 8.26952755e-01 9.96374786e-01
3.89926046e-01 6.41388714e-01 4.00554597e-01 -2.37228423e-01
4.93602008e-01 -5.09998091e-02 5.10342717e-01 1.22736502e+00
-6.88472763e-02 2.51335278e-02 -5.83954751e-01 7.35842645e-01
-2.40422940e+00 -7.26464808e-01 -3.39337468e-01 2.33535647e+00
8.44498754e-01 -3.95554990e-01 2.96746820e-01 1.31417736e-01
1.12116945e+00 1.72504172e-01 -5.25027514e-01 -4.18772340e-01
-1.01437069e-01 -1.57037660e-01 6.73985720e-01 4.29733276e-01
-9.62226629e-01 5.03470562e-02 5.20666456e+00 1.28155875e+00
-1.08405089e+00 3.02429706e-01 4.03453052e-01 3.82181741e-02
-2.58807510e-01 9.07688141e-02 -5.73936164e-01 9.12154078e-01
2.80339897e-01 -6.50812626e-01 5.95896184e-01 9.13768053e-01
7.14059353e-01 3.89323905e-02 -5.90415001e-01 1.15856946e+00
-1.04757413e-01 -9.39339519e-01 -4.01390821e-01 6.60379231e-02
8.74653876e-01 -2.71472126e-01 -1.94208268e-02 2.07264408e-01
-8.88640732e-02 -2.91260600e-01 7.57509887e-01 1.44593254e-01
2.85846800e-01 -1.02190542e+00 7.74782658e-01 4.77838427e-01
-1.45474386e+00 -2.83557832e-01 -5.42176008e-01 -9.43082049e-02
3.79895777e-01 8.66613626e-01 8.90883878e-02 1.18769097e+00
4.32524860e-01 8.33535790e-01 1.07364498e-01 1.35773718e+00
-1.37787923e-01 4.08888578e-01 -7.08988845e-01 1.96454003e-01
4.84260291e-01 -1.07792830e+00 9.44262266e-01 8.73462081e-01
4.36059982e-01 1.04037918e-01 6.42294645e-01 7.13941991e-01
1.91889137e-01 7.26340771e-01 -1.95490241e-01 3.39629799e-01
5.31746149e-01 1.24974918e+00 -1.53341055e-01 -1.71746001e-01
-5.23544490e-01 7.55415916e-01 3.34282190e-01 5.39538264e-01
-1.04985547e+00 -7.50570953e-01 6.57524705e-01 -1.49559528e-01
4.29618686e-01 -1.96591839e-01 -3.32306534e-01 -1.11322451e+00
6.25521481e-01 -8.95656824e-01 4.73366946e-01 -2.27193460e-01
-1.41549158e+00 2.57763237e-01 -2.17683434e-01 -1.46807933e+00
2.10355848e-01 -2.90065110e-02 -9.53204274e-01 1.08287871e+00
-1.47246408e+00 -1.06433976e+00 -2.20045432e-01 8.93057287e-01
1.76682770e-01 2.14163847e-02 1.27361953e-01 8.73576164e-01
-1.06597245e+00 4.93798971e-01 1.02915645e+00 -2.04034150e-01
6.27222657e-01 -1.14834249e+00 -5.16356409e-01 9.98138487e-01
-5.75013876e-01 5.63794494e-01 6.78919733e-01 -2.76402116e-01
-1.49286449e+00 -9.86304879e-01 5.89055419e-01 5.39691031e-01
8.60058665e-01 -1.28115937e-01 -8.49890649e-01 4.26848173e-01
2.64982618e-02 1.55317962e-01 3.37513715e-01 -8.74758139e-02
2.65555620e-01 -8.64604115e-01 -1.17602694e+00 3.09227169e-01
5.71316123e-01 -1.38013154e-01 -3.55016798e-01 5.54932177e-01
4.07460243e-01 -4.41242486e-01 -9.69226360e-01 3.15887958e-01
3.64192724e-01 -9.55091178e-01 1.00916278e+00 6.72393292e-03
2.06765264e-01 -6.11941397e-01 8.37383196e-02 -1.35100615e+00
-4.18972701e-01 -1.01622069e+00 3.98403639e-03 1.45998621e+00
-4.34281565e-02 -1.17664003e+00 5.04807234e-01 2.68117100e-01
-2.47453034e-01 -9.64881599e-01 -1.21374667e+00 -1.21755540e+00
8.80464613e-02 2.59842455e-01 2.96824455e-01 1.04427767e+00
1.35245413e-01 5.44329993e-02 -4.85135436e-01 3.92856419e-01
8.87522578e-01 2.18392447e-01 7.04163790e-01 -6.07950568e-01
-7.60850132e-01 -3.15017879e-01 -7.37703443e-02 -1.18226647e+00
-1.08168095e-01 -7.62437105e-01 1.62379026e-01 -1.34979677e+00
-1.33209854e-01 -7.99232364e-01 -3.65667045e-01 2.37561747e-01
-3.08341950e-01 -1.66536823e-01 2.50087172e-01 4.41786766e-01
-3.16488415e-01 1.10509646e+00 1.37256503e+00 -1.27388313e-01
-4.90483761e-01 1.06564842e-01 -5.37688673e-01 6.30286038e-01
4.34060335e-01 -3.26811522e-01 -6.20528758e-01 -4.63180423e-01
4.02721763e-02 3.20835829e-01 -4.68150191e-02 -8.33055973e-01
2.77401537e-01 -3.07817787e-01 -3.48342270e-01 -4.26285297e-01
1.06330156e-01 -9.87618566e-01 1.33211628e-01 6.64190590e-01
1.60754755e-01 -7.61905089e-02 7.03650434e-03 7.33873308e-01
-6.25121057e-01 -2.38643199e-01 9.37377095e-01 2.95322269e-01
-4.55158830e-01 5.21316350e-01 -1.65926382e-01 1.96064651e-01
1.38116074e+00 -7.84000978e-02 -7.32160434e-02 -3.22658598e-01
-5.59758425e-01 8.99760187e-01 1.59367383e-01 -5.12027591e-02
3.95546257e-01 -1.53721714e+00 -7.71147251e-01 -4.85692360e-02
-4.61306363e-01 2.21427754e-01 4.37937230e-01 1.57407701e+00
-8.61204147e-01 1.24638416e-01 2.34348420e-02 -4.16860342e-01
-1.00076890e+00 5.63510418e-01 3.00660700e-01 -3.16140234e-01
-6.59304142e-01 7.43126929e-01 4.76338416e-01 -4.15320098e-01
6.29275888e-02 -2.85872556e-02 2.62594759e-01 2.03455798e-02
2.74115443e-01 9.32164550e-01 -8.21485668e-02 -4.54347998e-01
-3.14172536e-01 4.70013916e-01 3.16669345e-01 7.93534070e-02
1.39405751e+00 -5.52712917e-01 -6.28620088e-01 1.85342446e-01
1.41387403e+00 3.99516076e-01 -1.33899975e+00 -3.11535567e-01
-4.78706211e-01 -6.05013907e-01 1.27867669e-01 -1.48839474e-01
-1.31421542e+00 4.83053744e-01 3.02265763e-01 4.49509919e-02
1.36486518e+00 -7.54676402e-01 1.23880661e+00 -1.19917234e-02
4.49787974e-01 -1.10293996e+00 -2.23107904e-01 4.28771585e-01
1.12534547e+00 -9.08175588e-01 2.73061007e-01 -5.10878265e-01
-3.79853249e-01 1.06830657e+00 6.48238540e-01 -4.76435393e-01
5.58603227e-01 -1.11275278e-02 -2.79556096e-01 2.89493501e-01
-2.96788424e-01 8.29303190e-02 -9.03083533e-02 7.65648186e-02
-6.50679320e-02 -2.51219630e-01 -1.28522372e+00 6.48607373e-01
3.06235701e-01 5.24504110e-02 3.23529989e-01 9.91329312e-01
-2.28130803e-01 -1.09358633e+00 -6.65456772e-01 -8.62220360e-04
-2.42533371e-01 1.74647808e-01 1.41397163e-01 6.18932605e-01
1.78217113e-01 6.51410162e-01 -1.95978597e-01 -1.68295149e-02
3.96432042e-01 -3.67561191e-01 5.07930629e-02 -2.16774791e-01
-4.62221086e-01 3.19805622e-01 -7.94662163e-02 -4.69521999e-01
-1.78900495e-01 -6.16403222e-01 -1.14577878e+00 -4.91606861e-01
-5.02719760e-01 5.93990684e-01 4.52449530e-01 9.33153152e-01
3.15815449e-01 2.82025903e-01 1.25891697e+00 -5.84894180e-01
-1.07807064e+00 -5.28130889e-01 -1.03369272e+00 2.71991998e-01
2.49170989e-01 -5.84626138e-01 -5.12436032e-01 -2.58423924e-01] | [6.424151420593262, 4.81345796585083] |
df2b65d7-118e-476f-b3f2-ee29ef49b9b4 | brain-tumor-segmentation-using-enhanced-u-net | 2210.13336 | null | https://arxiv.org/abs/2210.13336v2 | https://arxiv.org/pdf/2210.13336v2.pdf | Brain Tumor Segmentation using Enhanced U-Net Model with Empirical Analysis | Cancer of the brain is deadly and requires careful surgical segmentation. The brain tumors were segmented using U-Net using a Convolutional Neural Network (CNN). When looking for overlaps of necrotic, edematous, growing, and healthy tissue, it might be hard to get relevant information from the images. The 2D U-Net network was improved and trained with the BraTS datasets to find these four areas. U-Net can set up many encoder and decoder routes that can be used to get information from images that can be used in different ways. To reduce computational time, we use image segmentation to exclude insignificant background details. Experiments on the BraTS datasets show that our proposed model for segmenting brain tumors from MRI (MRI) works well. In this study, we demonstrate that the BraTS datasets for 2017, 2018, 2019, and 2020 do not significantly differ from the BraTS 2019 dataset's attained dice scores of 0.8717 (necrotic), 0.9506 (edema), and 0.9427 (enhancing). | ['Faisal Muhammad Shah', 'Md. Mahim Anjum Haque', 'Md Aminul Haque Palash', 'Maksuda Islam', 'Abdullah Al Munem', 'MD Abdullah Al Nasim'] | 2022-10-24 | null | null | null | null | ['tumor-segmentation', 'brain-tumor-segmentation'] | ['computer-vision', 'medical'] | [ 5.30369096e-02 3.32904786e-01 -1.59048080e-01 -2.06302568e-01
-4.54098344e-01 -2.58677274e-01 4.32087958e-01 3.90828013e-01
-3.92558277e-01 8.09340000e-01 2.51887292e-01 -5.84159791e-01
1.43904418e-01 -9.58946347e-01 -4.09759462e-01 -7.54315138e-01
-4.20693517e-01 1.03468381e-01 3.12679797e-01 3.66823003e-02
9.06191766e-02 4.72390503e-01 -9.74308431e-01 3.98424506e-01
8.48352492e-01 1.14702523e+00 2.62980789e-01 6.29822969e-01
-2.74985254e-01 9.37503397e-01 -5.05643010e-01 -2.22262144e-02
4.06528443e-01 -4.99141723e-01 -9.45970178e-01 -2.80317187e-01
3.33434641e-01 -7.07870364e-01 -4.45332706e-01 1.30330610e+00
3.99115682e-01 -3.47119361e-01 8.54270995e-01 -9.56239223e-01
-5.76783061e-01 6.97905421e-01 -6.87333286e-01 4.15426582e-01
-8.86147842e-02 1.18288338e-01 4.78621602e-01 -5.02077460e-01
1.09960771e+00 9.59449410e-01 5.79917908e-01 7.61479437e-01
-5.73199689e-01 -8.77128363e-01 -3.85754764e-01 1.62384793e-01
-1.09475374e+00 2.63264822e-03 1.57178313e-01 -6.78114235e-01
6.40576601e-01 5.23356140e-01 9.31177080e-01 9.59186673e-01
9.07124400e-01 9.32386935e-01 1.16083097e+00 -6.38297275e-02
1.57926500e-01 -1.92774594e-01 3.15291286e-01 8.72094870e-01
2.09199980e-01 7.23247826e-02 7.52709731e-02 1.08810253e-01
8.43146145e-01 3.19085509e-01 -5.79703212e-01 -6.86206818e-02
-1.50653076e+00 7.05784917e-01 9.94475782e-01 4.10140723e-01
-3.56001973e-01 1.89541131e-02 6.51275337e-01 1.84420794e-01
5.12615442e-01 2.36029759e-01 -3.43117982e-01 1.22221164e-01
-1.19355667e+00 -4.20169413e-01 3.52036029e-01 9.22235489e-01
2.32335448e-01 -1.40049890e-01 -3.56056154e-01 5.96513391e-01
1.01645529e-01 8.51474851e-02 8.59425724e-01 -6.41702294e-01
-5.74065605e-04 3.41667682e-01 -4.01829004e-01 -4.90289688e-01
-6.27813637e-01 -5.87366164e-01 -1.38717330e+00 3.22752625e-01
3.67356986e-01 -3.97199988e-01 -1.65035641e+00 1.14150190e+00
6.21977821e-02 6.81996020e-03 1.53643474e-01 7.75116026e-01
1.35173821e+00 4.54343826e-01 -4.59778309e-02 -9.30250138e-02
1.12939668e+00 -1.06794405e+00 -8.80710602e-01 -2.52162963e-02
8.30669463e-01 -5.45961261e-01 5.34882784e-01 1.28072768e-01
-9.72436547e-01 -5.05205058e-02 -1.05324352e+00 -1.62229931e-03
-5.15523195e-01 -6.61568120e-02 5.54379761e-01 2.80674785e-01
-1.35977626e+00 5.95707953e-01 -9.48345602e-01 -4.86532986e-01
1.01475739e+00 1.50237337e-01 -5.13596117e-01 -9.69272703e-02
-1.18470871e+00 1.08882773e+00 6.27122104e-01 -1.59948066e-01
-1.18555534e+00 -5.28829634e-01 -6.92060411e-01 -8.32661018e-02
9.59904641e-02 -4.23780799e-01 9.75548565e-01 -6.57211602e-01
-9.27001536e-01 7.86322832e-01 1.96305498e-01 -7.68502116e-01
8.97334993e-01 3.10373098e-01 -3.85203362e-02 4.09322858e-01
4.61513288e-02 1.16548157e+00 1.77709892e-01 -1.04445887e+00
-5.42519629e-01 -3.89539689e-01 -3.33645374e-01 -7.14142574e-04
-5.25111184e-02 6.68973848e-02 -3.22970927e-01 -5.68498194e-01
2.97438174e-01 -6.12480283e-01 -2.20139533e-01 3.70820969e-01
-5.92081368e-01 3.65394354e-01 8.31208110e-01 -1.19211149e+00
7.84808397e-01 -1.83306408e+00 -4.03628111e-01 2.86106527e-01
5.58435917e-01 1.45894587e-01 7.26685897e-02 -3.36511552e-01
-2.64576435e-01 3.59210819e-01 -5.31500041e-01 2.78441548e-01
-5.28401077e-01 3.09683055e-01 5.39091825e-01 6.06499851e-01
-1.56675696e-01 1.10070634e+00 -1.04310513e+00 -7.62183070e-01
6.89870417e-02 3.40909004e-01 -3.39021057e-01 -1.44024312e-01
2.64738798e-01 4.56695288e-01 -2.02449769e-01 1.07394075e+00
6.77379966e-01 6.61473274e-02 -1.93533361e-01 -1.82667986e-01
-8.92101899e-02 -2.96967566e-01 -3.92813504e-01 1.38123846e+00
-1.05244778e-01 1.07386732e+00 2.86719173e-01 -1.10517824e+00
7.53013909e-01 4.54127073e-01 8.97629559e-01 -8.55979919e-01
5.80640256e-01 2.21184090e-01 5.33319354e-01 -6.49677932e-01
2.72063538e-02 -4.49792184e-02 3.30900997e-01 2.93425977e-01
5.02489135e-02 -2.60036707e-01 2.68654466e-01 1.42953619e-01
1.25543356e+00 -4.48818773e-01 3.22145402e-01 -4.66556728e-01
1.72628939e-01 4.39190567e-02 5.04216373e-01 4.44880247e-01
-8.78957212e-01 6.43702209e-01 7.45705366e-01 -4.71270382e-01
-8.91770959e-01 -1.06300545e+00 -5.29209495e-01 1.84531108e-01
4.19225171e-02 -7.29343295e-02 -1.05123591e+00 -7.92014062e-01
-3.49868208e-01 5.15114903e-01 -1.03457654e+00 -1.21644393e-01
-2.91950613e-01 -9.05189872e-01 3.37957114e-01 5.44183314e-01
7.66075492e-01 -1.04423058e+00 -9.06853735e-01 6.25181720e-02
-2.70625949e-01 -8.91635358e-01 -4.84690160e-01 5.12120664e-01
-9.86706793e-01 -1.29602110e+00 -1.29244542e+00 -9.76110101e-01
1.15917778e+00 1.88921466e-01 1.01291335e+00 9.23884511e-02
-5.99775672e-01 -1.90146919e-02 -3.16254407e-01 -5.27358115e-01
-5.07177889e-01 -2.71878570e-01 -2.56489754e-01 -7.24462986e-01
9.19714663e-03 -1.63817316e-01 -6.44490838e-01 2.14309603e-01
-1.18518472e+00 3.54499370e-01 8.19252729e-01 8.81475925e-01
7.69699872e-01 1.08257614e-01 2.50014007e-01 -6.06796205e-01
7.21055567e-01 -6.35933220e-01 -1.72189444e-01 2.66509622e-01
-5.96780837e-01 -2.39725113e-01 3.81215781e-01 -3.70496809e-01
-5.67779183e-01 -2.52603423e-02 -5.20131513e-02 -4.26586986e-01
-1.73355773e-01 4.42820370e-01 2.86500245e-01 -2.01327860e-01
5.03399253e-01 4.26586643e-02 1.64447188e-01 3.35897245e-02
1.52482763e-01 8.26996863e-01 7.11410522e-01 3.33899185e-02
3.87642473e-01 5.29803514e-01 -1.57122612e-01 -5.81664205e-01
-4.49677080e-01 -3.12476546e-01 -7.56962538e-01 -4.54709470e-01
1.28151250e+00 -5.66929340e-01 4.88252081e-02 4.70102668e-01
-1.04297423e+00 -4.36840683e-01 -2.80219346e-01 5.81975877e-01
-1.98950157e-01 2.29992732e-01 -9.40088987e-01 -2.55136222e-01
-8.42164934e-01 -1.55509794e+00 5.41379273e-01 5.15142918e-01
-2.16780245e-01 -7.85138428e-01 -2.86476552e-01 2.31222257e-01
4.62744325e-01 6.44638181e-01 8.92159641e-01 -6.71054184e-01
-2.31074944e-01 -3.26708078e-01 -5.89399815e-01 3.90410244e-01
3.62783879e-01 2.58448362e-01 -6.09209955e-01 -2.02424452e-01
-1.32056773e-01 -6.25865608e-02 9.40077960e-01 9.12605822e-01
1.54575193e+00 -3.95685822e-01 -5.30935347e-01 8.00971448e-01
1.20905304e+00 8.71041358e-01 9.26499248e-01 3.53976011e-01
4.47052211e-01 3.96831721e-01 2.65975654e-01 -2.42829230e-02
1.16712563e-01 -1.59535512e-01 9.13415134e-01 -6.55061364e-01
-3.52448523e-01 4.01599884e-01 1.99557558e-01 7.51580119e-01
9.92750451e-02 8.03427696e-02 -1.20806575e+00 8.80736887e-01
-1.35890198e+00 -8.31236541e-01 -1.54290423e-01 1.86035359e+00
8.96842301e-01 2.56427556e-01 -3.74175966e-01 -1.74223587e-01
7.07059085e-01 -2.32967868e-01 -5.47673225e-01 -4.27583843e-01
-3.41502950e-02 2.75634497e-01 8.59581470e-01 2.38818273e-01
-1.11929083e+00 6.27421200e-01 6.81784153e+00 6.11019135e-01
-1.40898156e+00 1.72422126e-01 1.37545490e+00 -7.75837451e-02
-3.87232751e-02 -4.35191542e-01 -4.31324467e-02 5.46248019e-01
7.80802608e-01 -1.55355275e-01 -5.46624660e-02 4.92705882e-01
1.89179510e-01 -4.60620940e-01 -7.72328854e-01 9.01374817e-01
-4.35564704e-02 -1.35831928e+00 -1.71130881e-01 2.20715925e-01
9.45572257e-01 5.60097337e-01 -2.06796244e-01 8.45017806e-02
2.58262992e-01 -1.33954537e+00 5.36430240e-01 6.44108891e-01
8.42279375e-01 -5.76685727e-01 1.25204408e+00 1.74489766e-01
-7.51359582e-01 1.82538390e-01 -3.24628770e-01 4.28799301e-01
-1.76517382e-01 7.97181606e-01 -1.12219954e+00 4.39414978e-01
9.37429965e-01 5.86084723e-01 -5.51393628e-01 1.44170022e+00
6.90144952e-03 4.56545055e-01 -1.08987048e-01 5.24858683e-02
5.28994799e-01 -1.01632982e-01 3.30525845e-01 1.18488789e+00
6.55208051e-01 -2.51254756e-02 -4.83857282e-03 8.70572805e-01
-2.12775826e-01 2.27524579e-01 -6.76015973e-01 7.84124509e-02
2.19792258e-02 1.31995571e+00 -1.39647448e+00 -6.41325057e-01
-3.00103426e-01 8.82295370e-01 -2.75523216e-01 7.35384524e-02
-7.92379022e-01 -3.58500242e-01 3.78234625e-01 1.84317380e-02
-1.26153201e-01 -1.47693634e-01 -6.00700796e-01 -9.85538185e-01
-5.17472386e-01 -4.30307388e-01 2.35520646e-01 -1.07199383e+00
-9.72496092e-01 8.12753737e-01 -6.76419362e-02 -1.22765779e+00
1.42375767e-01 -6.11404419e-01 -9.65855181e-01 7.63437569e-01
-1.30060565e+00 -6.47500515e-01 -4.40504193e-01 4.41116184e-01
4.54260945e-01 -3.16808955e-03 5.94216287e-01 4.07711752e-02
-5.31609297e-01 1.51440799e-01 1.97281063e-01 5.67640901e-01
6.57820761e-01 -1.32340312e+00 6.39868453e-02 9.24333811e-01
-5.17855048e-01 3.48222733e-01 5.22161245e-01 -9.30507123e-01
-6.34814382e-01 -1.23790944e+00 3.90145570e-01 2.41589576e-01
5.65871000e-01 9.42145213e-02 -8.26351702e-01 5.55880368e-01
6.97489262e-01 3.06026816e-01 6.33589566e-01 -6.17926300e-01
1.21350922e-01 3.42912674e-01 -1.53013980e+00 6.04591489e-01
5.45429826e-01 -3.60413827e-02 -3.09973598e-01 4.52328652e-01
3.93321723e-01 -6.09544039e-01 -1.07740843e+00 4.27156299e-01
5.09794593e-01 -8.01246762e-01 8.82669806e-01 -2.91260034e-01
8.74974132e-01 8.71642306e-02 1.88151568e-01 -1.69297683e+00
-1.60096586e-01 -2.34026043e-03 1.04586892e-01 4.47556674e-01
5.14203191e-01 -5.15506625e-01 6.22362137e-01 3.80289704e-01
-4.53577220e-01 -1.14172161e+00 -9.15122330e-01 -5.62519848e-01
6.81051910e-01 -1.07938059e-01 5.31816900e-01 1.08824241e+00
-3.23091969e-02 -2.76284605e-01 9.87708867e-02 -3.24551016e-01
5.84353387e-01 -3.21693987e-01 1.20961599e-01 -1.30356157e+00
6.30834937e-01 -9.40110505e-01 -4.46568727e-01 -4.11602080e-01
-1.75664842e-01 -1.26264203e+00 3.81946377e-02 -2.25530410e+00
5.04368067e-01 -2.77707040e-01 -4.25224245e-01 8.71094167e-01
-3.06331664e-02 1.98642448e-01 -4.25373614e-02 2.95323223e-01
-3.30671631e-02 2.92582184e-01 2.04206276e+00 -7.21346200e-01
2.86356211e-01 -4.13265914e-01 -5.60756445e-01 6.51111066e-01
1.24543452e+00 -2.11684540e-01 -2.33218372e-01 -3.39833617e-01
-3.69540662e-01 1.55645058e-01 2.64366478e-01 -1.22669315e+00
1.95561826e-01 -2.04120781e-02 8.81375492e-01 -8.20776045e-01
-2.47271627e-01 -8.36603224e-01 -1.05729155e-01 1.02267182e+00
-2.86163181e-01 3.26284617e-02 3.33601505e-01 3.85133480e-03
-2.30398431e-01 -2.92243630e-01 1.08192253e+00 -5.17441809e-01
-4.84791011e-01 7.85875976e-01 -7.12683201e-01 -3.36142093e-01
1.38536048e+00 -4.58746940e-01 -2.82568336e-01 -4.49298888e-01
-7.94555664e-01 4.00817662e-01 3.90718400e-01 1.51171654e-01
7.90687263e-01 -1.25995862e+00 -6.01957381e-01 4.25738133e-02
-7.83062801e-02 2.11744905e-01 2.42224231e-01 1.27930546e+00
-1.08911061e+00 4.52395886e-01 -8.30651581e-01 -5.62718987e-01
-1.21177304e+00 2.72960931e-01 6.84382856e-01 -3.77243310e-01
-9.19253886e-01 9.01276171e-01 1.99292749e-01 -3.79496902e-01
2.34133333e-01 -8.17430258e-01 -3.98671120e-01 1.69784814e-01
5.05560637e-01 1.70206636e-01 1.14478864e-01 -5.80824196e-01
-2.31491178e-01 1.05620116e-01 -1.86906889e-01 1.39089361e-01
1.38750029e+00 7.71315396e-02 -4.27734584e-01 -5.57297878e-02
1.47965837e+00 -4.45570558e-01 -1.01078427e+00 5.55864424e-02
-2.21256062e-01 -8.57349485e-03 6.02211714e-01 -1.08994699e+00
-1.91310060e+00 8.87851954e-01 1.12197733e+00 3.05134207e-01
1.04745376e+00 -1.80935040e-01 1.18572772e+00 -1.91035241e-01
1.16547108e-01 -1.03106713e+00 -2.88685381e-01 6.64411962e-01
7.27882087e-01 -1.34075367e+00 -2.01656207e-01 -1.38675556e-01
-4.23754722e-01 1.54287779e+00 6.39977336e-01 1.65268406e-01
8.01024854e-01 5.56614161e-01 3.30158651e-01 -4.12155718e-01
-4.35018599e-01 -9.80879441e-02 1.35761529e-01 6.54410183e-01
5.34663498e-01 3.59452575e-01 -5.09754479e-01 2.36131832e-01
-3.53248090e-01 -1.08421957e-02 9.15059805e-01 9.20402527e-01
-5.18767059e-01 -5.72026849e-01 -4.97513950e-01 1.19795549e+00
-7.37800598e-01 -2.07904845e-01 -5.11536002e-01 7.69779325e-01
3.46162915e-01 4.96322274e-01 1.15285940e-01 -2.89397299e-01
-3.10611725e-01 -1.58040583e-01 3.24727565e-01 -1.83315173e-01
-3.29130858e-01 1.29600450e-01 -2.65081286e-01 -5.81293762e-01
-3.31858963e-01 -4.95263785e-01 -1.47966397e+00 -6.15416616e-02
-1.46650225e-01 -3.06326710e-02 7.55342245e-01 7.84754276e-01
-1.56075582e-01 1.17954922e+00 6.24014623e-02 -4.95679438e-01
1.52852118e-01 -1.02796769e+00 -4.49643403e-01 7.25581795e-02
4.16175574e-01 -4.94355440e-01 -1.59745887e-01 -2.85188183e-02] | [14.590232849121094, -2.453660726547241] |
ba0e38f0-cbb7-463a-8e14-cc408ff82e6f | an-efficient-network-design-for-face-video | 2109.13626 | null | https://arxiv.org/abs/2109.13626v1 | https://arxiv.org/pdf/2109.13626v1.pdf | An Efficient Network Design for Face Video Super-resolution | Face video super-resolution algorithm aims to reconstruct realistic face details through continuous input video sequences. However, existing video processing algorithms usually contain redundant parameters to guarantee different super-resolution scenes. In this work, we focus on super-resolution of face areas in original video scenes, while rest areas are interpolated. This specific super-resolved task makes it possible to cut redundant parameters in general video super-resolution networks. We construct a dataset consisting entirely of face video sequences for network training and evaluation, and conduct hyper-parameter optimization in our experiments. We use three combined strategies to optimize the network parameters with a simultaneous train-evaluation method to accelerate optimization process. Results show that simultaneous train-evaluation method improves the training speed and facilitates the generation of efficient networks. The generated network can reduce at least 52.4% parameters and 20.7% FLOPs, achieve better performance on PSNR, SSIM compared with state-of-art video super-resolution algorithms. When processing 36x36x1x3 input video frame sequences, the efficient network provides 47.62 FPS real-time processing performance. We name our proposal as hyper-parameter optimization for face Video Super-Resolution (HO-FVSR), which is open-sourced at https://github.com/yphone/efficient-network-for-face-VSR. | ['Yongming Tang', 'Sige Bian', 'He Li', 'Feng Yu'] | 2021-09-28 | null | null | null | null | ['video-super-resolution'] | ['computer-vision'] | [ 3.01385522e-01 -3.89803290e-01 -1.53505981e-01 -3.50416601e-01
-5.57108402e-01 -1.65234078e-02 -6.27704263e-02 -9.73978102e-01
-4.08940703e-01 8.55945587e-01 1.13608994e-01 1.39099389e-01
3.80058140e-02 -7.36001313e-01 -7.57135570e-01 -7.09819019e-01
-1.08626612e-01 5.40578663e-02 3.67446840e-01 -2.06011668e-01
3.05109788e-02 6.87728107e-01 -1.87623763e+00 5.31805217e-01
5.95080912e-01 8.13350677e-01 4.66446012e-01 9.03465986e-01
-1.14980750e-01 7.83240557e-01 -6.57612503e-01 -4.90856767e-01
3.78798813e-01 -3.62465531e-01 -6.72230303e-01 -6.15492910e-02
7.18287528e-01 -9.20709848e-01 -4.32897568e-01 1.22789657e+00
7.78209925e-01 1.24620274e-01 6.79544359e-02 -9.39713299e-01
-6.23272419e-01 6.81135654e-01 -9.16419864e-01 8.13995004e-01
3.14294606e-01 6.27506990e-03 3.46510768e-01 -1.08482409e+00
6.49962544e-01 1.54688120e+00 5.13633132e-01 1.02965772e+00
-9.69871581e-01 -1.10762203e+00 -2.16337502e-01 4.89499003e-01
-1.83700800e+00 -9.02113736e-01 5.19283652e-01 2.08082289e-01
7.33782589e-01 2.49541506e-01 4.38278854e-01 1.02672863e+00
5.04901297e-02 3.17630619e-01 7.52804518e-01 2.52963565e-02
-5.14672697e-02 -2.21795455e-01 -2.67518342e-01 6.60494149e-01
2.43692130e-01 1.32548198e-01 -8.16211939e-01 1.96230650e-01
1.88844395e+00 -7.11187050e-02 -6.14789069e-01 4.45701063e-01
-8.51152778e-01 5.21066189e-01 2.00941667e-01 3.54738325e-01
-3.57828975e-01 -7.35338330e-02 2.63193011e-01 3.48644078e-01
4.60561991e-01 -1.84618741e-01 -3.60403568e-01 -4.46157195e-02
-1.11467767e+00 3.10699463e-01 3.06925893e-01 9.65797365e-01
5.42665124e-01 6.57833219e-01 -1.79075167e-01 1.14398646e+00
-7.34649822e-02 4.37525392e-01 3.90745074e-01 -1.63870263e+00
5.44026256e-01 3.39902788e-02 5.79522774e-02 -9.23498392e-01
-1.74900189e-01 -2.26278797e-01 -1.25468838e+00 1.78671837e-01
1.79549456e-01 -2.59223133e-01 -8.63286495e-01 1.48622072e+00
5.91100872e-01 7.83665955e-01 9.60710272e-02 1.24325299e+00
1.17781413e+00 9.03653026e-01 -1.92068994e-01 -9.17149365e-01
1.31034124e+00 -9.32723343e-01 -9.34273362e-01 2.89365858e-01
-2.28288382e-01 -7.72859693e-01 9.29439783e-01 4.47273612e-01
-1.74743760e+00 -1.04508269e+00 -9.48098302e-01 -1.50135905e-01
3.73915046e-01 4.02501673e-02 2.42077574e-01 4.67921853e-01
-1.52278769e+00 8.57117295e-01 -6.62306845e-01 1.63611829e-01
7.32245147e-01 8.26455891e-01 -3.77434433e-01 -2.78135426e-02
-1.11335969e+00 3.81872743e-01 5.83144605e-01 1.40648872e-01
-8.60720754e-01 -1.01641262e+00 -5.99786997e-01 1.68016553e-01
4.85513687e-01 -6.96338415e-01 1.19064796e+00 -1.24353766e+00
-1.68106198e+00 5.93431890e-01 -3.86514306e-01 -2.06447631e-01
3.53669703e-01 -7.98881873e-02 -7.09678829e-01 5.54613531e-01
-2.83750325e-01 7.22802937e-01 1.25328577e+00 -1.09990501e+00
-7.88029552e-01 -4.10081446e-01 -1.22140080e-01 2.98373491e-01
-4.67854619e-01 5.84902644e-01 -9.11539257e-01 -7.36913502e-01
-1.38429910e-01 -4.76466060e-01 -5.58510385e-02 7.59718791e-02
1.76992208e-01 1.56753823e-01 9.02868211e-01 -9.29494858e-01
1.41838038e+00 -1.97347867e+00 1.91437513e-01 -1.95099205e-01
4.48920608e-01 6.79638386e-01 -2.66248882e-01 -4.60126370e-01
-2.37845823e-01 1.82402819e-01 -4.36833389e-02 -3.91591221e-01
-6.95142984e-01 1.70625031e-01 1.11374915e-01 2.82896072e-01
7.36285299e-02 4.77578372e-01 -3.91272932e-01 -1.02612484e+00
7.25810826e-02 1.37841058e+00 -7.00633347e-01 2.62169927e-01
1.67007387e-01 4.75982785e-01 -3.66081417e-01 7.67102957e-01
1.25525928e+00 -3.14621001e-01 1.53663218e-01 -5.10818243e-01
-9.06639248e-02 -3.31327200e-01 -1.57452750e+00 1.59924209e+00
-3.22088212e-01 4.49858516e-01 5.77952683e-01 -4.64692265e-01
7.09301949e-01 4.07052636e-01 4.64838237e-01 -7.68002331e-01
2.83336282e-01 -5.90686016e-02 -3.58778447e-01 -6.15555286e-01
6.77468181e-01 8.15673918e-02 7.84650207e-01 7.60481209e-02
8.12928379e-02 5.11391044e-01 3.42840344e-01 -1.40070111e-01
6.42591238e-01 1.74604118e-01 -8.41903873e-03 -6.14596382e-02
9.19119060e-01 -5.87442577e-01 8.33250225e-01 1.79411918e-01
-1.83822662e-01 7.88104355e-01 3.41863930e-01 -5.88711143e-01
-1.41827822e+00 -9.93627727e-01 -2.46679425e-01 1.29441404e+00
1.55270353e-01 -2.46081978e-01 -1.14246702e+00 -5.23227341e-02
-5.68568349e-01 1.35024726e-01 -3.26166034e-01 3.29588145e-01
-1.16048431e+00 -8.00587833e-01 4.60268289e-01 4.76112932e-01
9.97942984e-01 -1.20003605e+00 -4.73351032e-01 -8.27129111e-02
-3.24722558e-01 -1.50135887e+00 -6.35101438e-01 -6.37072384e-01
-1.14241219e+00 -8.66345823e-01 -8.81859303e-01 -9.61866617e-01
6.05970860e-01 4.49127853e-01 1.12303662e+00 3.59643877e-01
-3.63162756e-01 -1.72707796e-01 -5.94479069e-02 3.02552432e-01
-2.76765317e-01 -1.69044659e-01 2.81328529e-01 3.06383893e-02
3.32292281e-02 -8.57762396e-01 -7.42384613e-01 2.79285669e-01
-9.36921954e-01 1.58206016e-01 4.31902766e-01 5.89548409e-01
9.81848180e-01 2.42809877e-01 2.51698494e-01 -5.99397421e-01
3.94645512e-01 -1.76787734e-01 -8.38117778e-01 1.37140438e-01
-3.01305801e-01 1.05553502e-02 8.29371035e-01 -6.35242760e-01
-1.43247855e+00 -7.09249154e-02 -2.20465168e-01 -1.14589894e+00
1.84567794e-02 -1.53046817e-01 -2.18950033e-01 -3.66577357e-01
5.55313468e-01 3.51659030e-01 1.24680445e-01 -3.63762110e-01
1.29993947e-03 5.19026577e-01 7.52725780e-01 -3.30713868e-01
8.89986694e-01 3.94773602e-01 4.48354520e-02 -1.01816463e+00
-4.18781579e-01 -4.20982949e-02 -4.14781272e-01 -3.24936569e-01
9.11084533e-01 -1.14883244e+00 -9.49264646e-01 3.89452636e-01
-1.09621358e+00 -3.00747156e-01 1.11570455e-01 4.16369319e-01
-3.99756312e-01 3.55354130e-01 -9.59461093e-01 -5.63717246e-01
-6.81228697e-01 -1.18402183e+00 8.26603055e-01 6.73743010e-01
4.95334715e-01 -5.91054857e-01 -3.46368164e-01 4.51306403e-01
6.85464323e-01 1.07490718e-01 1.05752528e-01 1.42657042e-01
-8.13572347e-01 6.09868824e-01 -6.06178999e-01 4.41004723e-01
1.03439558e-02 2.77670652e-01 -8.88802588e-01 -4.54105020e-01
1.93200052e-01 -2.83912532e-02 5.51483214e-01 6.59031332e-01
1.61477315e+00 -6.22382104e-01 6.36727214e-02 1.34434593e+00
1.72346926e+00 1.34580925e-01 8.79056811e-01 1.29767478e-01
7.88097084e-01 3.26186031e-01 5.18789530e-01 6.50707662e-01
1.41090378e-01 7.77947426e-01 4.07935888e-01 2.67187841e-02
-3.15686136e-01 1.38703063e-01 3.78897786e-01 7.45828569e-01
-8.44713449e-01 -6.39727619e-03 -2.91807503e-01 1.46788850e-01
-1.39542949e+00 -1.35459888e+00 4.64449525e-02 2.12687063e+00
1.00664771e+00 -6.52602091e-02 2.62883186e-01 -3.89251709e-02
1.22121906e+00 3.31513107e-01 -5.97467899e-01 -1.02297917e-01
-2.15468198e-01 3.90258163e-01 4.42789167e-01 7.38834023e-01
-8.16004813e-01 1.04751003e+00 5.69081831e+00 1.20901227e+00
-1.11749399e+00 3.57103348e-01 8.85276318e-01 -7.24827647e-01
5.81685230e-02 -5.98988235e-01 -1.17265403e+00 5.33339679e-01
1.23619711e+00 -1.48016796e-01 9.29952145e-01 6.97903156e-01
4.41033751e-01 2.98468888e-01 -6.86356008e-01 1.65514481e+00
2.16548383e-01 -1.56381893e+00 4.40118462e-01 -1.84536234e-01
6.96153283e-01 -3.04632366e-01 1.92725688e-01 2.94911712e-02
-1.88984394e-01 -1.28207958e+00 2.37587035e-01 3.33623230e-01
1.46256590e+00 -1.17419970e+00 5.46752453e-01 -1.14179835e-01
-1.61331201e+00 -2.30495200e-01 -8.29545140e-01 3.23901594e-01
2.38746867e-01 1.95304945e-01 -3.40936095e-01 5.11545241e-01
1.09663355e+00 4.26258117e-01 -3.50074977e-01 6.40700877e-01
2.61167288e-01 3.08954626e-01 -1.76337302e-01 4.74237829e-01
-3.54662597e-01 -1.37820393e-01 4.77997214e-01 1.10253012e+00
3.99798483e-01 6.91355228e-01 -2.33572274e-01 6.84143960e-01
-3.21879894e-01 3.02592479e-02 1.36086799e-03 3.23467016e-01
6.99617922e-01 1.33033359e+00 -5.73626757e-01 -3.85547340e-01
-3.72501433e-01 1.07403135e+00 1.67959824e-01 4.16637480e-01
-1.34918809e+00 -1.37087822e-01 8.28168213e-01 1.46235883e-01
4.31525677e-01 -4.61681597e-02 5.64152226e-02 -1.20522070e+00
9.12113711e-02 -1.14907873e+00 3.39981019e-01 -8.93324971e-01
-6.13438547e-01 1.09562230e+00 1.46453544e-01 -1.09515202e+00
-2.21624121e-01 -4.88688350e-01 -2.54683822e-01 6.83905065e-01
-1.74222398e+00 -8.97562385e-01 -7.57319212e-01 1.16509140e+00
1.11439967e+00 -2.46662885e-01 2.40988657e-01 7.21878290e-01
-9.10311759e-01 9.25270975e-01 -1.76985458e-01 -3.42330448e-02
6.15794301e-01 -5.12998462e-01 2.27742076e-01 9.51567709e-01
-2.99265444e-01 3.84987324e-01 7.26156890e-01 -4.54910070e-01
-1.45049608e+00 -1.15595019e+00 4.56815898e-01 3.71599272e-02
2.16502309e-01 3.34431455e-02 -1.15480030e+00 3.84968042e-01
4.07506675e-02 2.46789813e-01 2.31229872e-01 -5.11071861e-01
-1.78625822e-01 -4.59823877e-01 -1.43929970e+00 6.27307534e-01
1.31540930e+00 -2.79998511e-01 -9.56981480e-02 1.33007810e-01
1.05322134e+00 -7.41164804e-01 -1.16399431e+00 4.15303528e-01
5.03530025e-01 -1.29319894e+00 1.37529290e+00 -1.93760455e-01
8.01465631e-01 -3.57567400e-01 -9.89472121e-02 -5.56904078e-01
-4.82439458e-01 -7.01981246e-01 -4.84845072e-01 1.12252808e+00
9.65128317e-02 -3.25166017e-01 9.53300536e-01 3.62534821e-01
1.18635498e-01 -7.62736022e-01 -8.99392366e-01 -5.19745052e-01
-4.90903199e-01 7.90254995e-02 8.74404490e-01 8.10659587e-01
-7.09028304e-01 1.07064366e-01 -6.08780801e-01 5.15956581e-01
1.03998208e+00 -2.80646890e-01 5.42350292e-01 -8.89508665e-01
-1.81513578e-01 -3.85142803e-01 -1.25897169e-01 -9.43938434e-01
6.90986291e-02 -2.41844952e-01 -4.33599085e-01 -1.07369161e+00
2.80174077e-01 1.09532863e-01 -7.99493566e-02 1.93048552e-01
-1.85858309e-01 6.68382049e-01 1.76639766e-01 2.12555483e-01
-5.44690073e-01 3.42995405e-01 1.65168953e+00 4.18041438e-01
-4.02113378e-01 -1.28008321e-01 -5.14211416e-01 6.77573204e-01
1.02083886e+00 -6.40332326e-02 -5.84381998e-01 -7.58444786e-01
-1.67041942e-01 6.93381429e-01 3.01850975e-01 -1.04635513e+00
2.40756243e-01 -4.49493647e-01 7.89620817e-01 -4.63218480e-01
7.01247275e-01 -6.84928477e-01 4.77563947e-01 2.60424882e-01
-9.75963995e-02 3.36205959e-01 3.21410835e-01 1.11907981e-01
-1.16957612e-01 -8.20484683e-02 1.19174623e+00 -2.65867710e-01
-8.45590770e-01 8.26339185e-01 1.44521788e-01 -1.86303798e-02
7.99631476e-01 -6.29525185e-01 -4.10078555e-01 -2.69219548e-01
-7.02965200e-01 -4.81471717e-02 4.07359660e-01 3.46501023e-01
1.15478098e+00 -1.24150777e+00 -1.14510262e+00 3.80633891e-01
-8.76716614e-01 1.80838004e-01 9.59829509e-01 4.64896828e-01
-9.91600633e-01 5.85214533e-02 -7.43990183e-01 -4.45338219e-01
-1.98065007e+00 5.75003505e-01 4.42745268e-01 -8.05164408e-03
-6.93299353e-01 1.00947404e+00 1.31007120e-01 2.54544705e-01
2.56804138e-01 7.46129155e-02 -6.35534048e-01 -2.43214011e-01
1.20124507e+00 7.05284536e-01 -3.51990640e-01 -9.11452830e-01
-7.24033415e-02 8.12235832e-01 -2.47340351e-01 4.91584167e-02
1.44686437e+00 -2.18478322e-01 -2.57475644e-01 -2.20302612e-01
1.16900849e+00 -2.33946368e-01 -1.55363226e+00 -8.39118958e-02
-7.84470856e-01 -8.29240799e-01 2.03617349e-01 -2.65780091e-01
-1.74577808e+00 5.02436399e-01 9.50918615e-01 -2.87434071e-01
1.68775439e+00 -3.28491658e-01 8.45331371e-01 7.60590956e-02
3.99341375e-01 -9.79362905e-01 1.24830067e-01 2.57162958e-01
8.55624318e-01 -9.72827792e-01 2.38414571e-01 -7.43959546e-01
-5.45693815e-01 1.19540226e+00 1.08839977e+00 -2.14098930e-01
3.05720866e-01 5.80351233e-01 -1.98782533e-01 1.42096370e-01
-8.51230204e-01 2.08584324e-01 -8.92972276e-02 5.02378285e-01
3.41916293e-01 -3.52916867e-01 -1.03751980e-01 3.61751646e-01
-2.82878458e-01 3.60214174e-01 6.99303091e-01 4.70738769e-01
-5.42771339e-01 -7.45354295e-01 -6.15916967e-01 1.83296740e-01
-9.54600513e-01 -1.62089363e-01 5.06491244e-01 6.50636315e-01
1.47643626e-01 7.67506182e-01 2.80681670e-01 -3.23492944e-01
1.52557641e-01 -4.31015462e-01 6.89655483e-01 -3.98508795e-02
-4.28131312e-01 2.23016441e-01 -1.04360469e-01 -9.15763140e-01
-6.33808136e-01 -2.41021961e-01 -1.18743134e+00 -9.46829557e-01
-5.56737743e-02 8.70640203e-02 4.36792314e-01 3.47446978e-01
3.65894735e-01 7.25117624e-01 7.52190888e-01 -1.17041445e+00
-3.34201515e-01 -6.99617505e-01 -4.47518885e-01 -9.90587380e-03
4.60712463e-01 -4.11915451e-01 -3.51976991e-01 4.81227189e-01] | [11.066776275634766, -1.891029715538025] |
319262ad-f0fb-452f-bb7a-36c650105e53 | modeling-dialogue-in-conversational-cognitive | null | null | https://aclanthology.org/2020.lrec-1.147 | https://aclanthology.org/2020.lrec-1.147.pdf | Modeling Dialogue in Conversational Cognitive Health Screening Interviews | Automating straightforward clinical tasks can reduce workload for healthcare professionals, increase accessibility for geographically-isolated patients, and alleviate some of the economic burdens associated with healthcare. A variety of preliminary screening procedures are potentially suitable for automation, and one such domain that has remained underexplored to date is that of structured clinical interviews. A task-specific dialogue agent is needed to automate the collection of conversational speech for further (either manual or automated) analysis, and to build such an agent, a dialogue manager must be trained to respond to patient utterances in a manner similar to a human interviewer. To facilitate the development of such an agent, we propose an annotation schema for assigning dialogue act labels to utterances in patient-interviewer conversations collected as part of a clinically-validated cognitive health screening task. We build a labeled corpus using the schema, and show that it is characterized by high inter-annotator agreement. We establish a benchmark dialogue act classification model for the corpus, thereby providing a proof of concept for the proposed annotation schema. The resulting dialogue act corpus is the first such corpus specifically designed to facilitate automated cognitive health screening, and lays the groundwork for future exploration in this area. | ['Natalie Parde', 'Mina Valizadeh', 'Shahla Farzana'] | 2020-05-01 | null | null | null | lrec-2020-5 | ['dialogue-act-classification'] | ['natural-language-processing'] | [ 6.57864928e-01 1.11123288e+00 5.60856014e-02 -8.33303690e-01
-8.78432095e-01 -6.16272628e-01 4.68329132e-01 5.89155972e-01
-4.95254755e-01 8.21412981e-01 5.95019519e-01 -6.43015444e-01
-8.13390091e-02 -4.05772567e-01 1.21523649e-01 -5.26159585e-01
1.81319565e-01 1.33676100e+00 4.72679548e-02 -4.55203801e-02
-2.28792012e-01 2.95047283e-01 -8.46515536e-01 6.30377889e-01
8.71764898e-01 5.44543326e-01 3.53643924e-01 6.75741494e-01
-9.63912904e-02 1.15910518e+00 -9.18128908e-01 -3.72476280e-01
-1.49000123e-01 -9.15877938e-01 -1.54201996e+00 3.89133275e-01
-5.56913316e-01 -2.81307131e-01 3.67276818e-01 6.09980404e-01
6.54104471e-01 3.17076361e-03 5.34774780e-01 -7.07889020e-01
3.49946842e-02 5.96094191e-01 3.03444743e-01 -2.61332929e-01
6.80977345e-01 4.53647487e-02 7.90433466e-01 -2.63262451e-01
8.68767798e-01 8.65839958e-01 5.16790390e-01 1.01492405e+00
-1.00546229e+00 -7.50000123e-03 -3.72106880e-01 -2.39660710e-01
-9.07907486e-01 -5.38546622e-01 4.25166219e-01 -4.51331079e-01
8.52188587e-01 5.70056438e-01 7.60051668e-01 1.29085279e+00
-9.01412219e-02 8.25360358e-01 9.68741000e-01 -6.05112672e-01
5.24915755e-01 5.45621932e-01 2.18287870e-01 6.31276309e-01
-1.78618535e-01 -6.83024943e-01 -7.40472749e-02 -6.03022218e-01
3.79730493e-01 -2.15414166e-01 -3.48679751e-01 -2.57629484e-01
-1.09614885e+00 9.86095309e-01 -1.00543555e-02 6.07778668e-01
-4.43188936e-01 -6.63692415e-01 8.70532632e-01 2.35923544e-01
5.40883303e-01 8.40848923e-01 -2.71450460e-01 -4.31628525e-01
-4.31897730e-01 -2.61671394e-02 1.40731001e+00 7.38170862e-01
2.33686745e-01 -6.35836482e-01 -2.60366321e-01 1.04418039e+00
-6.05732959e-04 9.37060192e-02 5.25402844e-01 -9.00145411e-01
2.50161797e-01 9.17909920e-01 3.19384605e-01 -4.43540812e-01
-9.75657761e-01 3.87848765e-02 -7.91054726e-01 -4.61473703e-01
5.43984830e-01 -5.34674346e-01 -2.71622717e-01 1.39806056e+00
6.20271564e-01 -5.12004972e-01 3.36586773e-01 7.84605801e-01
7.90886641e-01 2.31386244e-01 2.04197541e-01 -6.84100211e-01
1.86385620e+00 -8.56358111e-01 -8.88932824e-01 -1.47870421e-01
1.19222903e+00 -6.14568889e-01 8.39770019e-01 2.65641272e-01
-1.04483747e+00 -7.91138783e-02 -5.39661169e-01 8.78736377e-02
-4.93716169e-03 1.11606613e-01 5.98606348e-01 8.73791277e-01
-9.81624424e-01 1.14753723e-01 -8.63609910e-01 -7.39789307e-01
9.66068581e-02 3.75057459e-01 -4.96662587e-01 4.15234901e-02
-1.23753667e+00 1.10261905e+00 2.53836989e-01 1.88752681e-01
-5.31442702e-01 -2.67076015e-01 -8.67559314e-01 -1.89959742e-02
4.21845227e-01 -7.37526119e-01 1.73983741e+00 -7.14873254e-01
-1.54770863e+00 1.17375505e+00 -2.42193807e-02 -4.77929384e-01
5.82563758e-01 1.99475899e-01 -2.02885911e-01 4.49992597e-01
1.59220204e-01 4.30211365e-01 2.50066012e-01 -8.31429482e-01
-5.54284036e-01 -3.89425546e-01 7.42789134e-02 3.86480182e-01
-3.55925560e-01 3.44086796e-01 -2.55593151e-01 -2.21611932e-01
-4.29817975e-01 -1.07116187e+00 -4.64484274e-01 -3.78902912e-01
-4.73353833e-01 -3.69237423e-01 3.06605130e-01 -7.62850523e-01
1.05179560e+00 -1.98600101e+00 -3.36446683e-04 8.73939842e-02
4.70441878e-01 4.06161427e-01 1.85289264e-01 6.28394246e-01
7.91853592e-02 -5.08195013e-02 -3.94235432e-01 -4.07672048e-01
-9.04569328e-02 2.27801055e-01 3.40425104e-01 4.40954655e-01
-1.35684684e-02 5.75898230e-01 -1.13467538e+00 -6.72384501e-01
2.16336414e-01 2.73321956e-01 -4.13099736e-01 7.13997781e-01
-2.45099321e-01 7.73868859e-01 -7.52783239e-01 3.08633983e-01
2.28537377e-02 -5.06212354e-01 7.96554923e-01 1.28602520e-01
6.49396479e-02 5.01631618e-01 -5.39341629e-01 1.55287385e+00
-3.97297382e-01 2.85836965e-01 4.45395976e-01 -1.00904095e+00
9.85448778e-01 8.66784871e-01 8.96282911e-01 -2.48941898e-01
3.44836056e-01 -8.57430790e-03 2.65615493e-01 -9.40904498e-01
2.21232146e-01 -1.77158624e-01 -2.67764181e-01 9.40660477e-01
-7.85787702e-02 -9.23721865e-02 1.09863520e-01 8.62138197e-02
1.55805492e+00 -3.84598047e-01 9.11142111e-01 -1.50797054e-01
7.55356610e-01 3.33929688e-01 3.06844234e-01 7.04250395e-01
-5.01104832e-01 3.03370953e-01 4.84851986e-01 -6.86153829e-01
-9.09468174e-01 -2.62097001e-01 -2.78061777e-01 9.32876050e-01
-3.63878816e-01 -3.37767303e-01 -1.25063181e+00 -8.42673659e-01
-5.85456133e-01 5.82004845e-01 -3.76152962e-01 1.61490738e-01
-4.78159040e-01 -7.51072288e-01 6.71621621e-01 2.72758216e-01
3.63999814e-01 -1.31182587e+00 -9.50846970e-01 6.30984545e-01
-5.60659468e-01 -1.41883409e+00 -3.05911064e-01 2.48183623e-01
-3.65558535e-01 -1.36361313e+00 -6.97792888e-01 -9.84761000e-01
6.79927647e-01 -1.38304278e-01 8.96372795e-01 3.33610848e-02
-3.97435755e-01 5.13492346e-01 -6.42512381e-01 -5.10574162e-01
-1.26761305e+00 3.86717528e-01 -1.45500243e-01 -8.73560533e-02
5.60748398e-01 -3.94187719e-02 -6.15458250e-01 6.04064584e-01
-7.13330567e-01 3.12388420e-01 3.79085988e-01 1.10642195e+00
2.12963764e-02 -3.22061926e-01 6.55977190e-01 -1.34233928e+00
1.27141774e+00 -3.14788461e-01 -1.79172263e-01 4.09968406e-01
-2.17529356e-01 -1.00107290e-01 6.35982931e-01 -1.34229928e-01
-1.36478400e+00 4.26086694e-01 -5.13364911e-01 5.64760625e-01
-5.48194766e-01 5.26006043e-01 9.37746167e-02 2.96486646e-01
8.38814616e-01 2.14698315e-02 4.45315212e-01 -3.71179909e-01
3.61630693e-03 1.34370291e+00 3.49939495e-01 -4.01900887e-01
4.58518714e-02 2.42108211e-01 -3.38504791e-01 -1.01125133e+00
-6.16159439e-01 -7.58962810e-01 -4.85015363e-01 -1.66730732e-01
1.31699145e+00 -5.87154031e-01 -9.90434945e-01 -9.55685377e-02
-1.33100677e+00 -4.96047020e-01 -1.29368201e-01 4.63394046e-01
-5.97934246e-01 2.45266408e-01 -8.12897861e-01 -1.04733324e+00
-5.17939806e-01 -1.22464609e+00 1.11894166e+00 -2.19993845e-01
-1.14453328e+00 -1.12086093e+00 3.24279010e-01 8.16093206e-01
1.91304758e-01 1.42370775e-01 1.03102243e+00 -1.36217797e+00
1.00820400e-01 -4.62547570e-01 3.47540341e-02 1.23552665e-01
3.69466186e-01 -5.07537544e-01 -9.62038338e-01 -1.08358160e-01
6.73852444e-01 -5.97899258e-01 1.09840460e-01 1.48070589e-01
7.15137124e-01 -3.85221571e-01 -4.04306531e-01 -2.49984965e-01
6.00987792e-01 5.72582185e-01 3.82329166e-01 6.93230331e-02
2.47488856e-01 1.09793222e+00 6.08149529e-01 5.67619443e-01
5.60853183e-01 7.56498635e-01 -1.37240335e-01 -4.52859282e-01
3.31467956e-01 2.87601769e-01 5.80995518e-04 8.94716680e-01
-1.43356174e-01 -1.45543292e-01 -1.18819118e+00 3.89714479e-01
-1.86886299e+00 -7.26415813e-01 4.76430468e-02 1.86467874e+00
1.10358822e+00 -2.90438473e-01 4.39635783e-01 1.42526507e-01
6.20868206e-01 -2.35490367e-01 -1.03820771e-01 -5.17929673e-01
5.10667503e-01 6.09950386e-02 -8.40972140e-02 6.35388851e-01
-1.01870906e+00 5.64037263e-01 6.30662441e+00 9.58179012e-02
-6.41525149e-01 1.63498506e-01 8.16769838e-01 3.80536735e-01
1.81760311e-01 -3.49347413e-01 -3.73870283e-01 2.00892314e-01
1.14665592e+00 -1.77895259e-02 1.89555794e-01 9.69776034e-01
5.12466967e-01 -2.21129149e-01 -1.54313385e+00 5.87729692e-01
8.71816203e-02 -1.22415388e+00 -5.20359814e-01 1.13236658e-01
2.06278339e-01 -3.35674822e-01 -5.92158556e-01 1.16216354e-01
4.68476057e-01 -8.17304075e-01 -8.00668669e-04 1.73188180e-01
4.81491506e-01 -3.30633670e-01 1.16660094e+00 6.06448948e-01
-8.21664631e-01 1.06229924e-01 3.10636088e-02 1.31570613e-02
3.97081465e-01 3.87896299e-01 -1.99867237e+00 3.59703362e-01
1.76584437e-01 1.33059630e-02 -2.90872216e-01 7.48637497e-01
6.75556883e-02 4.11374152e-01 -6.27906397e-02 -3.95026654e-01
1.49959177e-01 -1.46736056e-01 2.74716616e-01 1.38921320e+00
-4.59619798e-02 3.89649659e-01 4.18495208e-01 3.43847781e-01
-1.32464394e-02 4.38867837e-01 -6.63855791e-01 -2.02916697e-01
4.60584283e-01 1.23910034e+00 -8.69266331e-01 -5.32803178e-01
-3.59740257e-01 9.46729898e-01 3.49802494e-01 2.02845242e-02
-5.07696033e-01 -1.15720510e-01 2.11951509e-01 1.96040962e-02
-3.40877175e-01 2.50679761e-01 -1.55511737e-01 -7.02439845e-01
3.55836973e-02 -1.30315518e+00 5.33013105e-01 -4.63687032e-01
-9.82526600e-01 9.77254033e-01 -4.60288338e-02 -1.10029447e+00
-8.88409436e-01 -4.15614545e-01 -1.63017690e-01 6.20115876e-01
-7.43006706e-01 -7.09931314e-01 -4.56522912e-01 4.85850483e-01
5.40904820e-01 -2.23770171e-01 1.63570857e+00 1.95049375e-01
-4.95781839e-01 2.15648726e-01 -3.79275590e-01 3.19884866e-01
8.24105144e-01 -1.18443346e+00 6.54876083e-02 1.00247443e-01
-3.18235010e-01 7.00180590e-01 5.71304321e-01 -5.57225347e-01
-1.24115729e+00 -9.62734997e-01 1.25999093e+00 -5.43532491e-01
4.60172057e-01 -4.43879187e-01 -8.56680334e-01 5.38839519e-01
2.05166593e-01 -5.95312417e-01 1.17369473e+00 2.26329461e-01
2.46760115e-01 1.42693564e-01 -1.33418262e+00 5.70162714e-01
7.79530287e-01 -6.95934772e-01 -6.47803426e-01 9.16398227e-01
5.59421480e-01 -5.84769011e-01 -1.13657236e+00 3.85715067e-02
3.98707986e-01 -9.66638386e-01 5.06224513e-01 -5.69093287e-01
1.54026061e-01 1.24618366e-01 4.19014335e-01 -1.26341772e+00
3.92088704e-02 -9.90502596e-01 4.04081196e-01 8.79797161e-01
5.69315374e-01 -6.48449123e-01 8.44008684e-01 1.30580115e+00
-2.54202396e-01 -6.96494162e-01 -6.71757042e-01 -3.81721616e-01
-4.09708709e-01 -2.50073373e-01 5.10259628e-01 1.04854727e+00
9.78040934e-01 7.84629047e-01 -2.80431360e-01 -2.23240137e-01
1.36835724e-01 -1.84303924e-01 7.35744774e-01 -1.24388862e+00
-4.49792176e-01 -1.13218561e-01 -1.46577626e-01 -7.71706641e-01
1.70364436e-02 -7.06994176e-01 3.56439918e-01 -1.65442860e+00
3.05987269e-01 -3.97640765e-01 5.58189511e-01 6.46198630e-01
-1.78274497e-01 -2.65133172e-01 -1.38938278e-01 2.00498119e-01
-6.62332654e-01 4.26067226e-02 1.08813322e+00 -4.89508398e-02
-4.91425574e-01 2.47855112e-01 -7.72626221e-01 6.67012274e-01
8.06506515e-01 -4.60192949e-01 -5.28363764e-01 1.16319522e-01
-1.69569001e-01 6.49251461e-01 -2.34157532e-01 -5.63156962e-01
3.71626854e-01 -7.95220211e-02 -9.26467329e-02 4.08204086e-02
2.58650482e-01 -9.30622518e-01 2.46117413e-01 7.15426147e-01
-7.61745691e-01 -9.10506770e-02 -2.82695383e-01 2.70296693e-01
-2.68643528e-01 -4.31872904e-01 5.63560307e-01 -3.84118736e-01
-1.13167487e-01 -2.13751301e-01 -9.50763226e-01 4.78608422e-02
1.10162616e+00 -2.20849402e-02 -3.48883450e-01 -6.26280189e-01
-1.00552428e+00 2.16009796e-01 2.78601408e-01 -5.66452257e-02
2.24827781e-01 -8.06900561e-01 -6.56016707e-01 -7.85936490e-02
2.54404843e-01 7.84493312e-02 1.30369961e-01 9.36130583e-01
-6.74888313e-01 6.54234529e-01 -1.40299853e-02 -6.46997571e-01
-1.54435432e+00 4.02739912e-01 2.38264546e-01 -6.69301510e-01
-6.52994275e-01 3.86523783e-01 2.00114951e-01 -5.03014863e-01
4.35112000e-01 -3.48112643e-01 -3.07823300e-01 4.78209294e-02
7.81763434e-01 9.66611430e-02 3.34434867e-01 -3.45102340e-01
-3.21299702e-01 -3.19313705e-01 -1.41734302e-01 -1.37512565e-01
1.26885629e+00 -3.93629782e-02 -2.26831034e-01 1.88622639e-01
9.10062850e-01 -3.25442672e-01 -6.02059126e-01 -1.14540264e-01
1.97370902e-01 5.41219003e-02 -4.50277686e-01 -8.16064119e-01
-3.48443031e-01 5.11258781e-01 2.24773243e-01 8.82934332e-01
9.88509953e-01 8.37078542e-02 5.08295894e-01 8.38700712e-01
4.35936868e-01 -1.12958825e+00 5.40216528e-02 3.66427809e-01
8.02687585e-01 -1.31030977e+00 -2.01371267e-01 -6.52041018e-01
-1.12807095e+00 9.83550787e-01 3.15725595e-01 6.67828619e-01
2.95915812e-01 3.71419281e-01 3.38083148e-01 -5.58317244e-01
-6.65725589e-01 -1.35772927e-02 -3.00448164e-02 7.08053410e-01
8.51773322e-01 2.44992703e-01 -5.11724770e-01 5.41674495e-01
-1.04873106e-01 4.80008982e-02 4.66606021e-01 1.01554608e+00
-2.87797481e-01 -1.33834279e+00 -2.73122847e-01 4.64343488e-01
-3.38273764e-01 4.99674678e-02 -9.37829256e-01 5.93266308e-01
-3.19050640e-01 1.15958858e+00 -1.36635691e-01 2.61168797e-02
4.18111116e-01 4.79739606e-01 2.99664795e-01 -1.02196705e+00
-9.82090533e-01 1.85952500e-01 1.09784472e+00 -2.86500186e-01
-5.14132082e-01 -5.32859325e-01 -1.21596313e+00 1.37301590e-02
-5.84407263e-02 5.32048821e-01 4.88926440e-01 1.05944908e+00
1.94616511e-01 4.89170939e-01 5.11378169e-01 -6.01827562e-01
-4.22078162e-01 -1.07729578e+00 -3.18894327e-01 6.15873694e-01
1.57518119e-01 -1.81575015e-01 2.80219484e-02 1.47399575e-01] | [12.53559398651123, 8.314465522766113] |
86b828ef-f4fa-4b30-8678-3fd0b25238f3 | jointly-learning-convolutional | null | null | https://drive.google.com/file/d/1i2jl5M0ddr-STAma0a2Bsr5rOMtcCSyB/view | https://drive.google.com/file/d/1i2jl5M0ddr-STAma0a2Bsr5rOMtcCSyB/view | Jointly Learning Convolutional Representations to Compress Radiological Images and Classify Thoracic Diseases in the Compressed Domain | Deep learning models trained in natural images are commonly used for different classification tasks in the medical domain. Generally, very high dimensional medical images are down-sampled by us- ing interpolation techniques before feeding them to deep learning models that are ImageNet compliant and accept only low-resolution images of size 224 × 224 px. This popular technique may lead to the loss of key information thus hampering the classification. Signifi- cant pathological features in medical images typically being small sized and highly affected. To combat this problem, we introduce a convolutional neural network (CNN) based classification approach which learns to reduce the resolution of the image using an autoen- coder and at the same time classify it using another network, while both the tasks are trained jointly. This algorithm guides the model to learn essential representations from high-resolution images for classification along with reconstruction. We have used the publicly available dataset of chest x-rays to evaluate this approach and have outperformed state-of-the-art on test data. Besides, we have experi- mented with the effects of different augmentation approaches in this dataset and report baselines using some well known ImageNet class of CNNs. | ['Soumava Paul', 'Ramanathan Sethuraman', 'Ekagra Ranjan', 'Siddharth Kapoor', 'Debdoot Sheet', 'Aupendu Kar'] | 2018-12-18 | null | null | null | icvgip-2018-2018-12 | ['thoracic-disease-classification', 'pneumonia-detection'] | ['computer-vision', 'medical'] | [ 4.05862540e-01 3.95095825e-01 -1.00583375e-01 -4.68873888e-01
-8.61053586e-01 2.11940140e-01 3.74927878e-01 1.35112911e-01
-7.26784706e-01 7.99269080e-01 2.19878972e-01 -1.89087972e-01
5.61567508e-02 -1.02268243e+00 -9.31110740e-01 -6.12727880e-01
-2.51050174e-01 5.24379313e-01 1.55295908e-01 -7.25432038e-02
-5.13931662e-02 5.84475994e-01 -1.27273846e+00 1.01548874e+00
4.27234590e-01 1.10222363e+00 1.49197698e-01 8.79049838e-01
1.32577956e-01 1.28316212e+00 -4.79326069e-01 -1.66752059e-02
3.55228364e-01 -1.99685603e-01 -9.49438214e-01 1.05260521e-01
4.33667243e-01 -4.83063400e-01 -5.36642671e-01 8.18970501e-01
5.36009431e-01 -2.65224166e-02 6.91379488e-01 -4.39686060e-01
-6.44404113e-01 3.46600682e-01 -5.94324827e-01 5.54911315e-01
8.76933858e-02 -1.91539375e-03 3.90437007e-01 -7.99477518e-01
6.74419463e-01 1.16567600e+00 9.71755803e-01 7.52283990e-01
-1.20035124e+00 -4.57117498e-01 -5.81061482e-01 2.13584900e-01
-1.09264767e+00 -1.95587486e-01 6.99678719e-01 -2.35568836e-01
7.96520591e-01 3.67982447e-01 3.19818348e-01 1.09592223e+00
4.81145054e-01 5.24139225e-01 1.23371935e+00 -3.73334378e-01
4.56548929e-02 1.17387280e-01 -9.32880715e-02 8.68019283e-01
-1.56387120e-01 2.15778083e-01 -1.65840507e-01 -1.07465304e-01
1.09157920e+00 1.79134697e-01 -2.70500153e-01 4.92045507e-02
-1.19531560e+00 9.02687371e-01 9.13672328e-01 5.14547408e-01
-7.16648996e-01 3.28157446e-03 5.79998970e-01 3.22510034e-01
4.62842226e-01 4.17261928e-01 -4.10291493e-01 2.44609147e-01
-9.73663032e-01 7.47770816e-02 4.36452240e-01 3.38953048e-01
3.55859816e-01 -1.05571754e-01 -2.04903632e-01 9.98481035e-01
-1.65291905e-01 -2.80313760e-01 9.43223357e-01 -6.99878395e-01
3.82347316e-01 5.27871668e-01 -3.04714590e-01 -8.19442570e-01
-7.76424944e-01 -5.39775312e-01 -1.55751812e+00 4.99346107e-01
4.23419505e-01 1.09104179e-01 -1.47281849e+00 1.28597164e+00
1.27862558e-01 3.35497558e-01 8.70492607e-02 8.98383558e-01
1.05406499e+00 6.09845519e-01 1.46684274e-01 -4.12310883e-02
1.39453757e+00 -9.52262580e-01 -6.04983211e-01 1.41679421e-01
5.39900839e-01 -6.49293125e-01 1.22091365e+00 6.89449847e-01
-1.24802804e+00 -9.26531136e-01 -1.29735076e+00 -2.24300086e-01
-3.42147857e-01 2.03306377e-01 4.37020659e-01 3.44986647e-01
-1.01023066e+00 1.05006075e+00 -1.05602860e+00 -8.42034891e-02
8.76663804e-01 4.92059797e-01 -5.88486016e-01 -2.91724145e-01
-1.05147779e+00 9.85737681e-01 4.75557387e-01 -4.55057099e-02
-8.22717428e-01 -9.43652034e-01 -6.45666659e-01 -8.94485116e-02
-1.68664753e-01 -5.66234291e-01 9.95103359e-01 -1.14190066e+00
-1.19399357e+00 1.09427190e+00 5.35830855e-01 -8.95708144e-01
7.00691938e-01 -9.13300738e-02 -5.04654586e-01 3.48085850e-01
7.73425549e-02 6.87964559e-01 9.42986488e-01 -1.06502104e+00
-4.91924018e-01 -2.98069686e-01 -1.40992999e-01 1.17069080e-01
-3.20077002e-01 -2.26299480e-01 -2.22034618e-01 -9.48092580e-01
1.15974598e-01 -7.72594213e-01 -4.48680580e-01 2.03475848e-01
-5.31174183e-01 1.79945469e-01 9.19940054e-01 -7.78496087e-01
7.40171492e-01 -2.15008616e+00 -7.16327652e-02 -1.78733040e-02
3.66719872e-01 3.00535619e-01 1.44010168e-02 -2.04821497e-01
-6.11877859e-01 -3.07001937e-02 -3.46204191e-01 -3.22334200e-01
-8.47297609e-01 3.93768042e-01 1.20942548e-01 6.18071079e-01
3.68226975e-01 6.89335465e-01 -4.55773503e-01 -7.05617905e-01
5.62983215e-01 8.39544952e-01 -4.88727897e-01 4.93728101e-01
2.09851433e-02 7.31057703e-01 -2.64049381e-01 4.78063285e-01
7.14996338e-01 -4.40795124e-01 -2.98926741e-01 -5.46377122e-01
2.22578391e-01 4.56566922e-02 -8.72816622e-01 1.80216157e+00
-6.69141710e-01 4.64390814e-01 -6.76510483e-02 -1.51890087e+00
5.50289214e-01 4.61646318e-01 7.53177762e-01 -8.09888661e-01
2.19801158e-01 5.47966957e-02 8.53068456e-02 -7.11057603e-01
9.04701129e-02 -3.55371982e-01 3.37866694e-01 2.40375504e-01
4.22207192e-02 1.24633849e-01 -2.22932607e-01 -1.78284317e-01
1.21502578e+00 -1.54300392e-01 3.36557239e-01 -1.49507046e-01
6.69466197e-01 1.39938906e-01 1.87013298e-01 6.82495236e-01
-1.42089846e-02 1.11377072e+00 2.99882561e-01 -1.04981685e+00
-1.31767106e+00 -9.20867741e-01 -5.81889510e-01 6.82153165e-01
-4.26348120e-01 -6.37316704e-03 -7.31306911e-01 -7.52208054e-01
-3.46451342e-01 4.08006847e-01 -1.21905518e+00 -1.49298444e-01
-8.73451412e-01 -1.04333866e+00 5.40876269e-01 6.01911902e-01
6.44146025e-01 -1.32319736e+00 -8.54345858e-01 4.00249064e-01
9.53837112e-02 -1.02707875e+00 8.11082646e-02 4.07327503e-01
-1.07804835e+00 -1.23116469e+00 -9.98183548e-01 -7.39781678e-01
8.44005346e-01 -3.09612423e-01 1.37111342e+00 4.55367833e-01
-8.30470979e-01 -9.44253132e-02 -3.02334934e-01 -4.34724301e-01
-7.12742209e-01 1.91376716e-01 -2.40633398e-01 -1.36761352e-01
8.01263377e-02 -5.41768670e-01 -8.64158809e-01 -2.90556438e-03
-1.07318068e+00 3.31879199e-01 9.53290582e-01 1.20680296e+00
8.83456230e-01 1.98072761e-01 3.95347208e-01 -1.26617038e+00
4.76279080e-01 -4.66919154e-01 -2.17639655e-01 -1.24037929e-01
-3.83857608e-01 1.47982568e-01 1.02039194e+00 -3.70577306e-01
-8.09162021e-01 2.03717262e-01 -6.59778178e-01 -4.47674394e-01
-2.91715920e-01 3.35266650e-01 4.00097251e-01 -1.22655265e-01
1.02209902e+00 1.44680947e-01 1.59204960e-01 -5.49299598e-01
7.66936131e-03 5.13835192e-01 7.64733434e-01 -6.65134043e-02
4.35551673e-01 7.37223089e-01 3.36551189e-01 -7.56843388e-01
-1.00738537e+00 -6.92698136e-02 -6.44181252e-01 5.01710400e-02
1.22144914e+00 -8.92940044e-01 -2.62109011e-01 2.54229695e-01
-8.45927238e-01 -2.34603480e-01 -4.89866406e-01 6.20998919e-01
-5.72184384e-01 7.95899630e-02 -1.07107008e+00 -1.34723648e-01
-5.66558421e-01 -1.30772698e+00 8.71324480e-01 9.48616415e-02
5.10457046e-02 -8.44488263e-01 1.09938584e-01 2.80381352e-01
6.14449084e-01 6.44366026e-01 1.11170208e+00 -5.70476294e-01
-3.11640948e-01 -2.93836087e-01 -4.47117329e-01 5.94640017e-01
2.02385142e-01 -4.54233408e-01 -1.12779939e+00 -2.64225602e-01
3.69554907e-01 -6.34701729e-01 1.05567288e+00 7.21744776e-01
1.99932361e+00 -2.84918964e-01 -1.69218346e-01 9.11176383e-01
1.51322663e+00 -2.23961800e-01 1.00218511e+00 5.21672368e-01
6.73528075e-01 3.55862945e-01 4.80384022e-01 3.05224419e-01
5.53570949e-02 5.00363529e-01 5.32694578e-01 -8.16359222e-01
-3.81068230e-01 2.25126058e-01 -4.00273830e-01 5.68087280e-01
-3.10759783e-01 2.76596129e-01 -8.22526097e-01 2.90603787e-01
-1.45407879e+00 -6.62184834e-01 5.48748448e-02 1.86321127e+00
1.14203537e+00 2.56312251e-01 -1.43643796e-01 3.84255379e-01
3.38007033e-01 -4.09893468e-02 -3.72327626e-01 -3.50330919e-01
1.85062185e-01 7.12181270e-01 7.15947747e-01 2.19768018e-01
-1.32241130e+00 3.94104660e-01 6.29273939e+00 6.89972818e-01
-1.52562118e+00 3.03056836e-01 1.26052845e+00 -5.14047258e-02
4.33486164e-01 -7.11791217e-01 -2.72284150e-01 3.22653651e-01
1.05368328e+00 4.13968652e-01 1.58344030e-01 9.87611771e-01
1.18439898e-01 -7.62142986e-02 -1.14851058e+00 1.15208447e+00
-1.65392447e-03 -1.68864071e+00 4.12237048e-02 -8.17322060e-02
6.76976383e-01 2.48649627e-01 2.19144404e-01 2.62327969e-01
-1.03263922e-01 -1.67255688e+00 1.09387403e-02 6.04787052e-01
1.11859298e+00 -9.24566686e-01 9.22126710e-01 2.21707895e-01
-8.08685124e-01 3.07837129e-02 -7.65674353e-01 2.08122522e-01
-2.26815805e-01 5.60698032e-01 -1.11623359e+00 4.37527448e-01
9.13654923e-01 5.20491242e-01 -6.16523564e-01 7.72396386e-01
3.38213146e-01 4.20393854e-01 -1.84042841e-01 5.75509667e-01
3.41039538e-01 2.17447132e-01 7.99449682e-02 1.15948462e+00
1.94017634e-01 2.41108045e-01 5.35563938e-02 7.28114307e-01
-3.36949289e-01 8.02893043e-02 -4.72109348e-01 3.92822951e-01
-2.39526212e-01 1.21382046e+00 -7.07371652e-01 -5.73939323e-01
-6.10272527e-01 1.01546526e+00 2.80801117e-01 -5.94797283e-02
-7.52455354e-01 -1.14525944e-01 2.54988641e-01 4.86011416e-01
1.83988959e-01 1.98549137e-01 -4.63450581e-01 -9.09926951e-01
-1.93461016e-01 -9.03606176e-01 5.58838129e-01 -5.86590827e-01
-1.17470860e+00 1.05046821e+00 -2.31042027e-01 -1.32782292e+00
-3.02864313e-01 -6.47297919e-01 -5.01915395e-01 9.00796473e-01
-1.65581417e+00 -9.60131109e-01 -5.20327032e-01 9.22665179e-01
7.39286125e-01 -3.08577299e-01 1.11845171e+00 6.89996064e-01
-1.79591343e-01 4.22669709e-01 -4.34281975e-02 3.61758828e-01
5.87559521e-01 -1.42774069e+00 1.79124102e-01 3.62587065e-01
-1.05952188e-01 1.84839562e-01 6.03314519e-01 -4.34607327e-01
-1.08301902e+00 -1.38401830e+00 3.03450108e-01 -3.27732772e-01
1.74426496e-01 -1.05772801e-01 -1.21110225e+00 5.02476335e-01
1.35637581e-01 7.92206287e-01 5.70637882e-01 -3.93268287e-01
1.08276650e-01 -7.53781945e-03 -1.58124352e+00 2.04511940e-01
3.36765379e-01 -3.42511475e-01 -4.79042262e-01 6.06263161e-01
4.49043691e-01 -7.17588186e-01 -1.20747077e+00 6.18383408e-01
4.11426246e-01 -1.15337408e+00 1.31030035e+00 -6.77897215e-01
9.74673808e-01 9.90608633e-02 -3.01273467e-05 -1.22437882e+00
-2.50263602e-01 2.43785173e-01 2.61267126e-02 4.69069421e-01
3.92707884e-01 -2.76221901e-01 1.17065299e+00 2.82824367e-01
-2.76263244e-02 -1.17736006e+00 -9.93986964e-01 -2.69964904e-01
2.95483053e-01 -1.65820330e-01 2.11959213e-01 1.03062880e+00
-5.94629526e-01 1.39384925e-01 -5.18399358e-01 1.69557072e-02
7.23508835e-01 -2.81879574e-01 4.20023471e-01 -9.78857577e-01
-3.03144962e-01 -6.71468675e-03 -5.08260846e-01 -3.58477026e-01
-2.22158447e-01 -6.97473824e-01 -1.44854620e-01 -1.40616155e+00
1.57349378e-01 -5.35481989e-01 -4.27915126e-01 3.82551014e-01
1.48948198e-02 7.37861037e-01 -1.55552357e-01 3.38454157e-01
-9.32096019e-02 1.05534405e-01 1.53011394e+00 -1.16888158e-01
-4.89056334e-02 1.19211122e-01 -3.52397233e-01 8.39909077e-01
7.42732644e-01 -5.32520533e-01 -3.06302369e-01 -3.25499237e-01
-1.67013139e-01 2.67829448e-01 3.26428145e-01 -1.44078481e+00
-6.04279265e-02 3.01369429e-01 1.21404910e+00 -7.09122360e-01
3.86502355e-01 -1.00966775e+00 1.95360985e-02 7.59675622e-01
-6.07799947e-01 3.72959897e-02 1.59795433e-01 2.18520239e-01
-3.38937044e-01 -7.68824220e-02 1.42352605e+00 -7.05103338e-01
-3.75473350e-01 4.57838923e-01 -9.71283540e-02 -1.98958009e-01
9.69725490e-01 -9.90493074e-02 1.00958146e-01 -6.65784031e-02
-1.08068752e+00 -3.65908265e-01 1.59804940e-01 2.17720821e-01
7.97755182e-01 -1.23919213e+00 -9.28263843e-01 3.63213748e-01
-9.60615352e-02 2.52858490e-01 2.77925819e-01 8.30296099e-01
-1.01339960e+00 2.44149536e-01 -7.23855019e-01 -7.44543493e-01
-1.24840808e+00 6.69854999e-01 6.60978615e-01 -6.65558636e-01
-1.14929712e+00 7.28558004e-01 2.20571727e-01 -2.91689336e-01
2.40416303e-01 -5.35886288e-01 -4.64941770e-01 -3.45657736e-01
8.24204028e-01 -6.48096800e-02 5.26719511e-01 -4.15091932e-01
-1.65352285e-01 2.59270132e-01 -4.57184672e-01 2.40969077e-01
1.67275584e+00 2.71098584e-01 3.42638418e-02 1.25337645e-01
1.64517069e+00 -6.15574777e-01 -1.08201051e+00 -2.93525815e-01
-2.25432470e-01 -4.34272707e-01 3.38430941e-01 -7.53698349e-01
-1.43321550e+00 9.40402806e-01 1.25313675e+00 1.10113822e-01
1.38374805e+00 -1.01458512e-01 7.36042261e-01 3.14096212e-01
-1.25329033e-01 -1.04547024e+00 2.78611213e-01 1.61914989e-01
1.00731349e+00 -1.51389956e+00 2.56641507e-01 -1.93179563e-01
-5.58786571e-01 1.34061718e+00 5.55844665e-01 -6.62298322e-01
8.49291265e-01 5.34596980e-01 1.77760348e-01 -3.89010310e-01
-6.45891666e-01 2.31817722e-01 2.33146563e-01 6.40019953e-01
6.73656344e-01 -7.49917403e-02 -1.88450187e-01 3.74879926e-01
-1.67397410e-01 3.08902532e-01 4.98538584e-01 9.15453374e-01
-1.26142785e-01 -9.04453516e-01 -5.39502323e-01 9.37659085e-01
-1.18539047e+00 -1.52946472e-01 1.74087167e-01 8.81765664e-01
2.69131064e-01 2.59589851e-01 1.01294674e-01 -2.20478699e-01
3.52400720e-01 -3.46652299e-01 4.94404048e-01 -5.93771517e-01
-6.93996727e-01 -5.14218993e-02 -1.00589551e-01 -6.80313587e-01
-3.86510611e-01 -3.02277356e-01 -1.18623662e+00 -1.19442545e-01
8.27341974e-02 -2.75884181e-01 7.09718406e-01 7.80410111e-01
-8.86788294e-02 1.04856110e+00 5.45025826e-01 -8.42934132e-01
-5.96447170e-01 -1.15172136e+00 -4.32902724e-01 7.02825427e-01
6.18400276e-01 -4.99019146e-01 7.06658289e-02 2.01658472e-01] | [14.91383171081543, -2.373023748397827] |
a7b535fc-99a9-4089-a57a-3f4facb2d6b7 | immune-defense-a-novel-adversarial-defense | 2303.04502 | null | https://arxiv.org/abs/2303.04502v1 | https://arxiv.org/pdf/2303.04502v1.pdf | Immune Defense: A Novel Adversarial Defense Mechanism for Preventing the Generation of Adversarial Examples | The vulnerability of Deep Neural Networks (DNNs) to adversarial examples has been confirmed. Existing adversarial defenses primarily aim at preventing adversarial examples from attacking DNNs successfully, rather than preventing their generation. If the generation of adversarial examples is unregulated, images within reach are no longer secure and pose a threat to non-robust DNNs. Although gradient obfuscation attempts to address this issue, it has been shown to be circumventable. Therefore, we propose a novel adversarial defense mechanism, which is referred to as immune defense and is the example-based pre-defense. This mechanism applies carefully designed quasi-imperceptible perturbations to the raw images to prevent the generation of adversarial examples for the raw images, and thereby protecting both images and DNNs. These perturbed images are referred to as Immune Examples (IEs). In the white-box immune defense, we provide a gradient-based and an optimization-based approach, respectively. Additionally, the more complex black-box immune defense is taken into consideration. We propose Masked Gradient Sign Descent (MGSD) to reduce approximation error and stabilize the update to improve the transferability of IEs and thereby ensure their effectiveness against black-box adversarial attacks. The experimental results demonstrate that the optimization-based approach has superior performance and better visual quality in white-box immune defense. In contrast, the gradient-based approach has stronger transferability and the proposed MGSD significantly improve the transferability of baselines. | ['Bin Ma', 'Xiangyang Luo', 'Jiawei Zhang', 'Haihua Wang', 'Hao Wu', 'Jinwei Wang'] | 2023-03-08 | null | null | null | null | ['adversarial-defense'] | ['adversarial'] | [ 3.88771236e-01 -8.37159976e-02 3.34259450e-01 -1.75322905e-01
-5.34347057e-01 -9.16945636e-01 6.30117536e-01 -3.71369749e-01
-5.12821615e-01 5.69831014e-01 -8.95746425e-02 -5.63637853e-01
3.13659400e-01 -8.52137029e-01 -9.40535307e-01 -1.06840873e+00
-2.36512674e-03 -5.46060443e-01 2.64570117e-01 -3.61665964e-01
9.50138345e-02 7.92020798e-01 -1.01489365e+00 1.16974697e-01
1.06561887e+00 8.79608035e-01 -2.35561997e-01 7.46818781e-01
4.06883508e-02 8.15901101e-01 -1.14152813e+00 -6.24841928e-01
8.34198296e-01 -4.71106231e-01 -2.78061002e-01 -2.84850568e-01
7.95943141e-01 -6.58228993e-01 -7.10993409e-01 1.61703253e+00
5.93380988e-01 8.18047151e-02 3.89506370e-01 -1.45209134e+00
-1.08101428e+00 2.71674424e-01 -5.73703706e-01 3.16879719e-01
-8.43208879e-02 4.40372229e-01 4.08251017e-01 -6.62076831e-01
1.90583751e-01 1.52218294e+00 3.90424401e-01 1.21607542e+00
-1.04122078e+00 -1.17464685e+00 4.45834458e-01 1.17971182e-01
-1.13510096e+00 -3.64865601e-01 7.20820963e-01 -1.04479678e-01
3.84420723e-01 4.76850629e-01 2.35197976e-01 1.23722827e+00
2.55641460e-01 6.06160998e-01 1.21012235e+00 -2.29112193e-01
2.41765812e-01 1.39206052e-01 3.24198194e-02 6.36430323e-01
3.99489492e-01 7.97459304e-01 -9.01478380e-02 -3.13297600e-01
7.39673793e-01 9.42961313e-03 -7.54084945e-01 -2.30261683e-02
-8.79873335e-01 8.49837601e-01 9.14187312e-01 -5.26198074e-02
-2.18736887e-01 1.35667413e-01 4.02833641e-01 3.81908834e-01
2.47185901e-01 1.64496973e-01 -1.46358415e-01 5.56232154e-01
-5.20831287e-01 1.89245597e-01 5.35296798e-01 6.01937652e-01
5.67145586e-01 8.03498626e-01 -2.01638609e-01 6.16674364e-01
3.10349733e-01 8.78873289e-01 4.29017544e-01 -5.57628095e-01
5.86102009e-01 4.95732129e-01 -1.20247066e-01 -1.14326620e+00
-1.80973194e-03 -4.28050429e-01 -1.01354206e+00 1.05337286e+00
3.93787563e-01 -5.97733200e-01 -1.26149714e+00 2.14240789e+00
4.43965346e-01 3.94946784e-01 5.76625407e-01 9.18030918e-01
6.48621142e-01 7.60247588e-01 1.54555142e-02 1.52812479e-02
9.53898966e-01 -8.76877785e-01 -6.66725814e-01 -3.84138674e-01
1.19604707e-01 -6.65643156e-01 8.93272400e-01 1.59284279e-01
-7.47539043e-01 -5.50605297e-01 -1.51420832e+00 3.40701908e-01
-4.53340858e-01 -2.22890541e-01 1.42948702e-01 1.15692270e+00
-8.57454658e-01 5.44238746e-01 -7.32651949e-01 1.68881029e-01
6.57147288e-01 3.80354404e-01 -4.57657367e-01 3.07677183e-02
-1.34116697e+00 8.53139818e-01 2.62874901e-01 1.61382139e-01
-1.32960939e+00 -6.84939682e-01 -7.88321078e-01 -7.26595968e-02
9.85794738e-02 -3.67243737e-01 7.59737432e-01 -1.24679530e+00
-1.35904837e+00 6.04290903e-01 3.51912826e-01 -7.59919286e-01
8.42075467e-01 -2.59170771e-01 -6.25843346e-01 3.40061903e-01
-3.85828048e-01 5.37524879e-01 1.41116369e+00 -1.50148869e+00
-3.81253451e-01 -2.45940596e-01 2.86258906e-01 -8.63898247e-02
-7.74994910e-01 1.30655006e-01 1.20161893e-02 -1.18306708e+00
-2.01367334e-01 -1.01528358e+00 -1.57923713e-01 3.23489785e-01
-4.90143836e-01 4.68950063e-01 1.29242897e+00 -5.26351750e-01
1.06913316e+00 -2.36374831e+00 -2.47325495e-01 1.56390727e-01
2.30165362e-01 1.03266931e+00 -4.78761703e-01 1.24371402e-01
-4.43236589e-01 3.17686737e-01 -3.76644850e-01 4.09043953e-02
-9.00007412e-02 3.30710918e-01 -9.07628417e-01 6.10104382e-01
4.50166017e-01 7.18482733e-01 -6.95946038e-01 -3.12503055e-02
1.18795130e-02 7.05333889e-01 -4.12938148e-01 3.22381407e-01
1.44833446e-01 2.81711698e-01 -3.67092580e-01 5.22861123e-01
1.22712576e+00 3.23382735e-01 -2.65523970e-01 -1.69313669e-01
2.63161570e-01 -2.84975529e-01 -1.07661450e+00 5.19642472e-01
-1.53504193e-01 5.56215584e-01 1.69072151e-01 -6.33298755e-01
1.03356695e+00 2.63250321e-01 -2.75015950e-01 -3.81459355e-01
2.21442848e-01 1.52468801e-01 3.40277582e-01 -1.07413851e-01
1.37277231e-01 -1.13131993e-01 1.98828071e-01 3.10182393e-01
-3.62472117e-01 2.58531094e-01 -2.98494816e-01 4.11292732e-01
1.15598392e+00 -1.12259597e-01 -4.17400375e-02 -1.91096678e-01
9.26960289e-01 -2.84759313e-01 8.37592542e-01 7.56890714e-01
-4.52351660e-01 5.38467765e-01 3.06459516e-01 -3.50544333e-01
-7.88065732e-01 -1.27016735e+00 4.53442894e-02 7.06577361e-01
3.42788130e-01 7.31509775e-02 -1.04625010e+00 -1.20208919e+00
1.60176963e-01 4.15563941e-01 -6.78890347e-01 -6.91817999e-01
-6.86287642e-01 -8.35780144e-01 1.10209727e+00 5.50004125e-01
1.06733453e+00 -8.36439073e-01 -2.74701267e-01 -1.29141480e-01
2.04489067e-01 -9.38803077e-01 -6.69094801e-01 -8.39970931e-02
-6.48114145e-01 -1.08315086e+00 -8.77086222e-01 -5.37570536e-01
1.09286451e+00 3.88303280e-01 4.53634858e-01 3.06027532e-01
-1.75939336e-01 3.95809785e-02 -2.84582108e-01 -5.93853354e-01
-8.12649786e-01 -6.59631550e-01 4.63189006e-01 3.06136280e-01
5.30662611e-02 -4.89990950e-01 -7.12724030e-01 4.10093367e-01
-1.28228080e+00 -6.20586157e-01 6.52110636e-01 9.52937484e-01
3.56445879e-01 1.18638158e-01 6.21646345e-01 -8.36608887e-01
6.82071745e-01 -1.66006878e-01 -7.88445830e-01 2.33943492e-01
-5.39413512e-01 -3.87064442e-02 1.21579111e+00 -1.00929165e+00
-8.49614918e-01 -1.44926831e-01 -2.83179134e-01 -9.09085989e-01
1.81423612e-02 -1.27988204e-01 -5.15154660e-01 -9.06149387e-01
8.59873056e-01 2.56941319e-01 -1.68807022e-02 -2.61880279e-01
2.26542234e-01 3.97504836e-01 7.63234913e-01 -3.89523655e-01
1.66862345e+00 5.81897378e-01 -8.61329772e-03 -5.71177006e-01
-5.50345659e-01 2.09364250e-01 1.13197537e-02 -2.56075054e-01
6.99591219e-01 -7.34262943e-01 -5.38092434e-01 9.85002816e-01
-1.01163745e+00 -2.06843659e-01 6.40375316e-02 2.08380684e-01
-5.00233434e-02 6.71511471e-01 -6.20029092e-01 -9.47186291e-01
-5.16355872e-01 -1.16476393e+00 4.47174370e-01 4.51478750e-01
3.39756548e-01 -8.47310781e-01 -2.30313763e-01 1.22784339e-01
4.88913327e-01 8.06008577e-01 8.38097036e-01 -8.37230146e-01
-5.86345673e-01 -5.20648658e-01 -2.83260643e-01 1.06689060e+00
2.10208640e-01 1.56197727e-01 -1.09334481e+00 -6.02797687e-01
2.65418917e-01 -1.90752178e-01 7.37262487e-01 -2.97616739e-02
9.11842704e-01 -7.71961451e-01 -6.88832253e-02 8.92629147e-01
1.46986043e+00 4.18260902e-01 9.04864788e-01 5.71743786e-01
8.19028020e-01 4.93610948e-01 4.82808232e-01 9.71889049e-02
-1.54420942e-01 2.92711765e-01 8.60419750e-01 -4.51858610e-01
-2.47461237e-02 -9.24997106e-02 7.49510288e-01 3.54148120e-01
2.26169392e-01 -4.55227435e-01 -6.38249040e-01 1.83016106e-01
-1.53614450e+00 -1.00247467e+00 -5.88376224e-02 2.15909314e+00
7.90135086e-01 3.84196907e-01 -1.18944012e-01 2.46698648e-01
9.22486663e-01 3.94419730e-01 -8.69387209e-01 -5.25603235e-01
-4.17636245e-01 5.35614677e-02 7.40206897e-01 5.06121576e-01
-1.16216934e+00 8.40440452e-01 5.89872885e+00 8.32750380e-01
-1.29523492e+00 -4.53307014e-03 6.91661119e-01 -9.64033455e-02
-1.29317760e-01 -1.53374150e-01 -8.02429378e-01 6.16407037e-01
6.39664233e-01 -1.87121868e-01 4.75129396e-01 8.83497179e-01
9.44479853e-02 3.78959566e-01 -7.74714351e-01 6.47661269e-01
-3.29907499e-02 -1.15254045e+00 4.11154687e-01 1.64578974e-01
7.41817117e-01 -2.38975376e-01 4.63301450e-01 3.03295195e-01
4.60969806e-01 -8.13170135e-01 7.49237776e-01 1.86262354e-01
6.19983315e-01 -1.02324116e+00 6.38621092e-01 2.44156495e-01
-9.45169091e-01 -2.35350892e-01 -4.95284468e-01 3.16802680e-01
-1.19360767e-01 2.17698008e-01 -2.22262308e-01 4.04481471e-01
6.36777163e-01 9.29627009e-03 -5.14880717e-01 6.40769958e-01
-5.50973773e-01 6.46384001e-01 -5.32470047e-02 1.71540067e-01
4.72163111e-01 -2.31692493e-02 8.79863024e-01 9.53174889e-01
-1.19994786e-02 8.30670148e-02 1.57922823e-02 8.85576010e-01
-1.90748289e-01 -2.27247462e-01 -5.22467554e-01 -1.61836110e-02
5.36542892e-01 1.16048956e+00 -3.57594818e-01 -2.02548847e-01
-2.00302094e-01 1.10109174e+00 9.87983495e-02 5.90604663e-01
-1.09782290e+00 -7.49020755e-01 1.16336453e+00 -4.08441693e-01
2.78841436e-01 -2.17714217e-02 -1.32227212e-01 -1.08056962e+00
6.34566247e-02 -1.43745708e+00 3.32530648e-01 -4.27065432e-01
-1.42488694e+00 8.99456322e-01 -2.32097358e-01 -1.38073635e+00
9.31182280e-02 -7.36365199e-01 -9.84832883e-01 1.02120638e+00
-1.63286817e+00 -8.36040914e-01 -2.33605966e-01 9.50926304e-01
1.78186521e-01 -3.88101697e-01 6.45660520e-01 2.13525146e-01
-8.34663749e-01 1.23513377e+00 2.04903275e-01 4.98683810e-01
6.13206863e-01 -1.11043417e+00 6.15933836e-01 1.64062250e+00
-2.98041347e-02 9.24931645e-01 6.65617347e-01 -6.45988941e-01
-1.24811447e+00 -1.42586613e+00 3.89709584e-02 -6.03369661e-02
6.55144274e-01 -3.86839569e-01 -1.20520651e+00 3.81421804e-01
9.49482899e-03 2.39652067e-01 5.53642392e-01 -8.14308405e-01
-7.87291586e-01 -3.14865023e-01 -1.56877911e+00 9.42357481e-01
7.01582015e-01 -5.52914202e-01 -5.60011625e-01 1.83304071e-01
7.83957899e-01 -3.91744614e-01 -4.66196328e-01 5.30671477e-01
3.37066203e-01 -9.52661633e-01 1.22962415e+00 -5.08483648e-01
2.67095685e-01 -6.03602171e-01 -1.63766712e-01 -1.13475609e+00
-1.34267718e-01 -9.09566522e-01 -2.05737025e-01 1.43349755e+00
1.97983399e-01 -1.08783305e+00 6.57237709e-01 4.84195888e-01
6.18602522e-02 -6.88125193e-01 -8.61119270e-01 -1.28879881e+00
2.68398821e-01 -4.22850586e-02 6.80529416e-01 9.37905610e-01
-5.87262928e-01 -2.62764454e-01 -4.82862681e-01 8.77394438e-01
8.66415262e-01 -3.22937638e-01 8.02759051e-01 -5.48438311e-01
-1.82414785e-01 -3.69969606e-01 -5.78889787e-01 -7.98139513e-01
1.47926375e-01 -5.35637379e-01 4.18659970e-02 -8.31768632e-01
-3.30009013e-01 -1.58919513e-01 -5.08829355e-01 5.49857736e-01
-7.40731001e-01 6.47502065e-01 3.78280789e-01 1.55356109e-01
1.50199279e-01 4.41484123e-01 1.07042921e+00 -1.62075087e-01
9.75306481e-02 1.88507065e-02 -7.90525496e-01 7.46181905e-01
8.94427121e-01 -6.40579104e-01 -4.87437278e-01 -4.37832832e-01
-3.60349894e-01 -4.24894750e-01 5.71380436e-01 -1.10567570e+00
-2.27221400e-02 -5.40672168e-02 3.82697880e-01 -2.16479495e-01
1.45327166e-01 -8.53730083e-01 -9.85357612e-02 9.30189073e-01
-2.44936794e-01 1.16731040e-01 4.49072182e-01 7.94491589e-01
-1.72590092e-01 -2.13501468e-01 1.30390179e+00 9.28615406e-02
-4.88024503e-01 5.48711181e-01 -3.00408572e-01 4.13479768e-02
1.20248604e+00 -3.29437643e-01 -7.48894334e-01 -3.45743090e-01
-3.58687401e-01 1.13063239e-01 4.84519333e-01 3.46538991e-01
8.38536441e-01 -1.40839374e+00 -6.86368167e-01 3.39164197e-01
-1.71270087e-01 -4.16097701e-01 2.68012911e-01 3.13368976e-01
-6.74896121e-01 -1.45263895e-01 -2.47199982e-01 -1.23368353e-01
-1.61115050e+00 9.44670320e-01 5.16150415e-01 -4.61634807e-02
-4.99365568e-01 1.16929996e+00 6.44772589e-01 -1.67287007e-01
4.78353024e-01 5.50111234e-02 -1.24447152e-01 -3.93518507e-01
9.79376853e-01 2.65328705e-01 -2.64591217e-01 -6.62788451e-01
-4.63428974e-01 4.28380966e-01 -3.43055010e-01 1.19055480e-01
1.01068580e+00 7.53140599e-02 -6.68870136e-02 -2.67153144e-01
1.24382019e+00 2.42733136e-01 -1.63838196e+00 -1.89462870e-01
-3.25158238e-01 -6.84735954e-01 3.22810300e-02 -8.49402308e-01
-1.42971921e+00 8.97181511e-01 8.69718730e-01 3.49935412e-01
1.49055028e+00 -7.85022676e-01 1.06298125e+00 2.36155584e-01
1.26783457e-02 -5.12128234e-01 2.09024489e-01 3.65793854e-01
8.78997505e-01 -1.10738611e+00 -1.08442552e-01 -4.07972366e-01
-5.88671148e-01 1.08757734e+00 9.05633092e-01 -4.73814905e-01
3.64301443e-01 4.29663211e-01 5.40453732e-01 3.16666275e-01
-4.93938118e-01 1.48510203e-01 2.99004585e-01 8.63949478e-01
-3.82735431e-01 -3.44350249e-01 -1.74704660e-02 4.41653967e-01
-1.42457768e-01 -6.51551723e-01 4.03699875e-01 9.73586082e-01
-3.66222769e-01 -1.10283184e+00 -8.49774301e-01 -3.72035429e-02
-8.50060821e-01 -2.80267239e-01 -6.10511899e-01 6.78198576e-01
1.10425919e-01 1.09002304e+00 -3.34204942e-01 -6.83591247e-01
3.47792238e-01 -2.56538123e-01 2.41448447e-01 -3.27804983e-01
-8.97876740e-01 -2.28128791e-01 -2.71020859e-01 -4.43361729e-01
-4.81638238e-02 -8.43778178e-02 -1.14511871e+00 -5.38974524e-01
-4.75969791e-01 -7.25475326e-02 5.42072117e-01 7.30999768e-01
1.82304546e-01 5.15332222e-01 1.16697180e+00 -6.67336404e-01
-1.01360464e+00 -4.80769932e-01 -2.17334345e-01 3.63553226e-01
7.39511371e-01 -2.85423160e-01 -8.17605734e-01 -1.14425153e-01] | [5.531124591827393, 7.918909549713135] |
7fd34156-61e6-4b3d-82f7-b772603869e1 | ra-v-net-deep-learning-network-for-automated | 2112.08232 | null | https://arxiv.org/abs/2112.08232v2 | https://arxiv.org/pdf/2112.08232v2.pdf | RA V-Net: Deep learning network for automated liver segmentation | Accurate segmentation of the liver is a prerequisite for the diagnosis of disease. Automated segmentation is an important application of computer-aided detection and diagnosis of liver disease. In recent years, automated processing of medical images has gained breakthroughs. However, the low contrast of abdominal scan CT images and the complexity of liver morphology make accurate automatic segmentation challenging. In this paper, we propose RA V-Net, which is an improved medical image automatic segmentation model based on U-Net. It has the following three main innovations. CofRes Module (Composite Original Feature Residual Module) is proposed. With more complex convolution layers and skip connections to make it obtain a higher level of image feature extraction capability and prevent gradient disappearance or explosion. AR Module (Attention Recovery Module) is proposed to reduce the computational effort of the model. In addition, the spatial features between the data pixels of the encoding and decoding modules are sensed by adjusting the channels and LSTM convolution. Finally, the image features are effectively retained. CA Module (Channel Attention Module) is introduced, which used to extract relevant channels with dependencies and strengthen them by matrix dot product, while weakening irrelevant channels without dependencies. The purpose of channel attention is achieved. The attention mechanism provided by LSTM convolution and CA Module are strong guarantees for the performance of the neural network. The accuracy of U-Net network: 0.9862, precision: 0.9118, DSC: 0.8547, JSC: 0.82. The evaluation metrics of RA V-Net, accuracy: 0.9968, precision: 0.9597, DSC: 0.9654, JSC: 0.9414. The most representative metric for the segmentation effect is DSC, which improves 0.1107 over U-Net, and JSC improves 0.1214. | ['Ziwei Xie', 'Chongchong Fan', 'Sumin Qi', 'Zhiqi Lee'] | 2021-12-15 | null | null | null | null | ['liver-segmentation'] | ['medical'] | [-1.56433682e-03 -2.39976943e-01 -9.72552747e-02 -1.02125831e-01
-2.42469206e-01 -1.07823744e-01 1.99199036e-01 1.74601182e-01
-6.94852471e-01 5.26633084e-01 6.76839240e-03 -2.57612199e-01
6.40462562e-02 -8.59769285e-01 -3.74339521e-01 -1.00399601e+00
-3.36487114e-01 -2.44552046e-01 1.58616662e-01 5.59643991e-02
3.62750031e-02 6.00933909e-01 -8.90474975e-01 2.72098899e-01
1.00256336e+00 1.26566756e+00 3.43191564e-01 6.34714544e-01
-3.31574827e-01 7.79515922e-01 -4.60535407e-01 1.57431871e-01
1.33698553e-01 -6.46571696e-01 -6.50628924e-01 1.32850274e-01
-4.87788618e-01 -3.00775409e-01 -3.80180627e-01 1.30103815e+00
6.41743660e-01 -1.77774921e-01 4.65170920e-01 -6.34257019e-01
-5.24876177e-01 7.34244227e-01 -7.65006602e-01 5.21507382e-01
-5.46383150e-02 2.41106629e-01 2.57114381e-01 -6.92551017e-01
1.45039663e-01 9.25179660e-01 6.66787565e-01 4.22989309e-01
-9.42466080e-01 -5.21749258e-01 -1.44742191e-01 1.34262785e-01
-1.41922641e+00 9.80287716e-02 4.11137968e-01 -3.02505672e-01
7.63854146e-01 4.77806866e-01 9.53002870e-01 2.93689281e-01
6.23386621e-01 9.39078987e-01 9.06686962e-01 -4.14028883e-01
-2.44360581e-01 1.56446636e-01 2.09885061e-01 9.20589805e-01
1.42671749e-01 4.36864533e-02 1.48597553e-01 3.87686044e-01
1.33848119e+00 4.61333066e-01 -6.16981924e-01 2.43403413e-03
-1.40155721e+00 7.12048113e-01 1.11678815e+00 8.34079266e-01
-6.00848794e-01 1.12130260e-02 5.18832862e-01 3.05549353e-01
-9.96485054e-02 4.05793667e-01 -3.74598920e-01 2.86189079e-01
-6.91495240e-01 -3.58351201e-01 3.84742171e-01 8.31902802e-01
4.66333419e-01 3.33021820e-01 -5.48903763e-01 5.93987465e-01
2.74092138e-01 4.10399646e-01 1.09825087e+00 -3.20287436e-01
1.70619059e-02 8.60248864e-01 -4.35652971e-01 -8.44981551e-01
-7.30501592e-01 -8.61780345e-01 -1.57628107e+00 -8.83397982e-02
8.49805772e-02 -2.38873735e-01 -1.27363479e+00 1.34170771e+00
8.26791953e-03 6.31532725e-03 1.41578317e-01 1.02083528e+00
1.22208309e+00 7.97943890e-01 3.62052023e-01 -3.95438910e-01
1.58922875e+00 -1.09144926e+00 -9.10168469e-01 1.01086833e-01
5.34047663e-01 -8.42564106e-01 8.52599442e-01 8.34897682e-02
-1.11500800e+00 -5.85428834e-01 -1.08796918e+00 1.20534420e-01
-2.21734688e-01 5.10062456e-01 6.50003016e-01 4.14950907e-01
-8.93168569e-01 4.10707533e-01 -8.70865524e-01 -2.29666352e-01
4.08117354e-01 7.54626155e-01 -1.27786368e-01 3.00992448e-02
-1.28456509e+00 8.36365223e-01 6.58605635e-01 3.47927302e-01
-5.19172072e-01 -5.71465254e-01 -8.61107171e-01 3.18097264e-01
1.06544290e-02 -6.10440433e-01 1.03186154e+00 -1.28074265e+00
-1.41029501e+00 5.84610224e-01 2.21914962e-01 -5.46193182e-01
4.12440211e-01 2.40731269e-01 -3.65421414e-01 3.21886629e-01
-1.71220705e-01 8.44335616e-01 4.64547306e-01 -7.44645476e-01
-6.89503551e-01 -3.05984169e-01 -4.26528901e-01 3.49641830e-01
-2.01370180e-01 -4.13967557e-02 -6.87130451e-01 -8.60476494e-01
3.68046552e-01 -5.97337961e-01 -5.42804301e-01 -8.13839734e-02
-2.71191329e-01 1.19541034e-01 7.07242966e-01 -1.00336885e+00
1.29216385e+00 -2.18102145e+00 -2.26381630e-01 5.10661185e-01
1.41350999e-01 5.07095814e-01 1.06295794e-01 -3.33809704e-01
-1.50893196e-01 1.45167917e-01 -4.91764396e-01 4.20166731e-01
-5.33307016e-01 1.54773057e-01 3.09792221e-01 4.80626196e-01
2.17769891e-01 1.21074617e+00 -5.15558124e-01 -7.45761335e-01
5.55819333e-01 6.43160939e-01 -3.79787326e-01 1.70592144e-01
2.47439861e-01 5.53744972e-01 -4.60762680e-01 5.52574992e-01
6.79449022e-01 -3.01382273e-01 2.13280115e-02 -4.27823097e-01
-2.36983582e-01 -2.07686454e-01 -1.02168810e+00 1.65120387e+00
-3.37382644e-01 3.19165319e-01 1.55566633e-01 -9.58445251e-01
8.85889828e-01 4.77413327e-01 7.10454047e-01 -9.84688520e-01
6.64344430e-01 5.04009910e-02 3.74206513e-01 -8.40695918e-01
1.83437839e-02 -2.01726649e-02 1.30039796e-01 2.59666070e-02
-9.66610536e-02 1.61110997e-01 1.25398606e-01 -2.14848965e-02
6.26230001e-01 -2.23397449e-01 5.64789832e-01 -6.21201873e-01
9.45745826e-01 -1.92323159e-02 5.61860800e-01 4.14606631e-01
-2.80458748e-01 2.73369998e-01 2.36447945e-01 -5.33788085e-01
-7.40746737e-01 -6.81194305e-01 -3.87899995e-01 5.24712980e-01
3.90340507e-01 1.06090076e-01 -8.77920747e-01 -6.16074681e-01
-1.73092112e-01 1.73140347e-01 -6.32893741e-01 -2.64648557e-01
-8.10042977e-01 -1.08312929e+00 3.45466167e-01 6.98128998e-01
1.11023319e+00 -1.21422875e+00 -7.60508001e-01 3.02231640e-01
-9.41120312e-02 -7.54175663e-01 -4.60462064e-01 1.26649573e-01
-1.15333295e+00 -1.02243686e+00 -1.07054543e+00 -1.21512449e+00
1.06066382e+00 1.56416550e-01 8.07777643e-01 5.03658772e-01
-6.14978433e-01 -1.81301087e-01 -1.06013805e-01 -8.53356495e-02
-1.53734431e-01 -7.69394124e-03 -3.74739945e-01 -1.56220496e-01
1.27283886e-01 -7.98074305e-02 -8.82874846e-01 1.29267052e-01
-9.50850546e-01 1.51225537e-01 1.18807721e+00 1.14902210e+00
7.05716908e-01 -1.06901199e-01 3.59780639e-01 -7.13500619e-01
4.43760663e-01 -2.12784857e-01 -5.45323372e-01 2.67925207e-02
-5.86785495e-01 -1.33340269e-01 8.23549509e-01 -3.47608894e-01
-9.00166035e-01 2.12603971e-01 -2.12853834e-01 -2.45362848e-01
3.07566510e-03 4.68594193e-01 -1.17286146e-02 -2.23057032e-01
3.03145587e-01 5.33493161e-01 2.13412523e-01 -2.96201617e-01
-4.15390395e-02 5.20383716e-01 6.73704147e-01 5.30789830e-02
1.91538557e-01 5.73346019e-02 -1.76581264e-01 -7.60322452e-01
-9.57056731e-02 -3.49444717e-01 -3.75129670e-01 -9.70834717e-02
9.43417907e-01 -8.04760098e-01 -8.58591914e-01 5.39694667e-01
-7.93456256e-01 -2.04668894e-01 -1.78689539e-01 8.55199516e-01
-7.93330669e-02 4.23633724e-01 -1.19892156e+00 -2.75717527e-01
-1.03395188e+00 -1.57134748e+00 2.28646144e-01 7.18585432e-01
1.32975385e-01 -8.44346821e-01 -6.65562391e-01 -2.31214836e-01
6.95364594e-01 2.13370666e-01 9.32925701e-01 -5.19425333e-01
-5.43863416e-01 -1.83071360e-01 -6.16196573e-01 3.98843050e-01
2.23869175e-01 -2.96372268e-02 -6.07553303e-01 -3.86483192e-01
1.87397659e-01 2.89130360e-01 1.04885113e+00 9.61341977e-01
1.49363959e+00 -3.07193160e-01 -3.09586436e-01 8.63211274e-01
1.53007627e+00 7.39976883e-01 7.47278810e-01 3.22754085e-01
4.46642697e-01 7.02588782e-02 3.19092989e-01 2.21438363e-01
4.57606018e-02 1.06245704e-01 4.15258080e-01 -7.57595301e-01
-2.68978149e-01 2.64630139e-01 3.44675384e-04 9.57930446e-01
1.05448462e-01 1.12944253e-01 -8.69216800e-01 5.94495773e-01
-1.48309255e+00 -5.91559529e-01 -3.11989635e-01 2.01993012e+00
8.11096549e-01 5.85859418e-02 -2.68578529e-01 1.65051475e-01
8.79357755e-01 -2.50319690e-01 -3.11608374e-01 -2.71117657e-01
1.19552901e-02 2.69763499e-01 6.38894618e-01 4.71891582e-01
-1.30795443e+00 6.17018998e-01 5.14992046e+00 8.17356825e-01
-1.51995635e+00 -4.62615974e-02 9.37115133e-01 2.59444177e-01
1.51046410e-01 -5.52147448e-01 -3.32064778e-01 7.22425461e-01
2.49367163e-01 1.18443556e-01 1.94586426e-01 5.89361429e-01
7.08837807e-02 -2.91781962e-01 -7.48300195e-01 1.20698631e+00
5.40172830e-02 -1.19694626e+00 9.73212048e-02 -2.38039821e-01
7.17027664e-01 -1.29706696e-01 1.29521206e-01 2.43672162e-01
-6.96650520e-02 -1.11572719e+00 1.91125408e-01 3.92386287e-01
9.84787047e-01 -8.45340431e-01 1.38316655e+00 1.75076246e-01
-1.20106316e+00 -2.97479369e-02 -3.17587286e-01 1.51533678e-01
-1.78993240e-01 7.34093249e-01 -8.94417524e-01 5.58046579e-01
6.39720619e-01 4.75144774e-01 -3.87462616e-01 1.33589196e+00
-1.49636656e-01 3.97703290e-01 -3.05054456e-01 -1.57281801e-01
5.26464820e-01 -1.16270833e-01 2.29617983e-01 1.60809207e+00
2.69492030e-01 3.85821611e-01 2.49610081e-01 6.36640847e-01
-1.69504285e-01 4.58865851e-01 -1.36672258e-01 4.12160844e-01
2.25911215e-01 1.47571504e+00 -1.00323391e+00 -5.67318082e-01
-1.64487675e-01 9.17247891e-01 -3.12753677e-01 1.57379881e-01
-9.08935606e-01 -6.31823838e-01 5.14870770e-02 -3.05395812e-01
1.31009892e-01 2.87043065e-01 -5.21632254e-01 -9.77650166e-01
-2.86836475e-01 -6.97959065e-01 6.19625449e-01 -3.50676984e-01
-7.86961615e-01 7.20110834e-01 -6.85574651e-01 -1.11957872e+00
2.69719422e-01 -4.46273476e-01 -6.84838831e-01 1.02180934e+00
-1.66854465e+00 -9.37380612e-01 -5.87800443e-01 6.70958161e-01
5.74612200e-01 -2.00334191e-02 7.32220232e-01 4.53695267e-01
-8.19234431e-01 5.40433049e-01 6.56556413e-02 5.95131397e-01
4.20314252e-01 -1.07989395e+00 -7.25688785e-02 9.43947971e-01
-4.88254458e-01 5.11913121e-01 1.57684654e-01 -5.06141067e-01
-1.13586926e+00 -1.23118353e+00 5.84334671e-01 5.39365053e-01
5.76876588e-02 2.98669547e-01 -9.38670695e-01 4.71882999e-01
2.13195801e-01 2.89128751e-01 4.46146756e-01 -6.49911344e-01
3.27305824e-01 -2.03948259e-01 -1.36980987e+00 3.75744879e-01
1.90882146e-01 2.52987221e-02 -3.19761544e-01 5.00617996e-02
5.39664149e-01 -7.22809434e-01 -1.14455879e+00 6.05504215e-01
4.57203656e-01 -6.24967039e-01 9.15346563e-01 -1.74849615e-01
3.21451515e-01 -4.32362854e-01 2.60391593e-01 -1.06541777e+00
-6.28530562e-01 -1.06145926e-01 1.55159771e-01 6.87781811e-01
5.25528610e-01 -6.56118035e-01 5.24277925e-01 3.31585288e-01
-3.45353127e-01 -1.02366388e+00 -6.70487344e-01 -3.01138341e-01
-5.81016839e-02 5.99667914e-02 6.24923348e-01 1.05882132e+00
9.65582430e-02 1.48294538e-01 7.77379051e-02 4.27728146e-02
3.42276037e-01 6.64884374e-02 1.53351858e-01 -8.63018692e-01
1.24311028e-02 -7.64501929e-01 -2.65248954e-01 -1.13474798e+00
-6.20296359e-01 -8.68679881e-01 -2.14028969e-01 -1.72578323e+00
2.90939659e-01 -5.77700734e-01 -6.67460978e-01 6.31725252e-01
-3.37054163e-01 2.25803435e-01 1.61168069e-01 2.76299030e-01
-2.07739681e-01 3.11392099e-01 1.74804640e+00 -2.98401535e-01
-4.00483608e-01 1.56751443e-02 -4.89798725e-01 7.38619089e-01
1.14191651e+00 -3.38079110e-02 8.52351077e-03 -5.68740785e-01
-4.27531451e-01 2.82937109e-01 1.38822481e-01 -9.48990107e-01
4.38903898e-01 2.60337234e-01 1.19967246e+00 -6.11251652e-01
-1.28059372e-01 -1.00716996e+00 4.70845625e-02 1.39983749e+00
-2.10797787e-01 1.20052181e-01 2.65230417e-01 -3.74303795e-02
-3.29125404e-01 -1.26906767e-01 1.07946038e+00 -4.68028039e-01
-7.81247497e-01 4.07441825e-01 -4.48957294e-01 -3.54859740e-01
1.11859500e+00 -2.60099232e-01 1.24838695e-01 2.36947797e-02
-8.35588336e-01 3.79787534e-01 -4.58332002e-02 -5.78644201e-02
8.08053434e-01 -1.29753363e+00 -6.11143827e-01 6.80453598e-01
-4.19178009e-01 3.13332498e-01 4.69600886e-01 1.33479536e+00
-9.99579370e-01 5.50855398e-01 -3.11206132e-01 -6.36559248e-01
-1.25883567e+00 5.48857629e-01 6.63310945e-01 -4.09081250e-01
-8.16462576e-01 7.89523363e-01 2.33773768e-01 -4.79020691e-03
2.31069356e-01 -7.89670527e-01 -4.93768364e-01 -1.06396236e-01
6.26194358e-01 3.28806907e-01 1.81905895e-01 -3.58820081e-01
-2.88129747e-01 3.97684813e-01 -7.49649331e-02 3.86046767e-01
1.20334268e+00 -5.42457588e-02 -4.89338905e-01 -2.66089857e-01
9.87286806e-01 -3.17909569e-01 -8.58377457e-01 -2.28595197e-01
-3.82197499e-01 -1.40526146e-01 3.63699973e-01 -1.23321581e+00
-1.56818163e+00 9.72085893e-01 9.08583045e-01 9.57995579e-02
1.47358954e+00 -4.25403118e-01 9.87078547e-01 -7.75131956e-02
-1.25777110e-01 -7.74360895e-01 -2.28638992e-01 4.49883789e-01
7.45703459e-01 -9.76721525e-01 -6.13214038e-02 -3.76284868e-01
-5.90982914e-01 1.29018688e+00 7.64119864e-01 -4.16134708e-02
4.78578329e-01 4.99281794e-01 1.69655547e-01 -1.71515550e-02
-7.60399997e-02 -1.45470068e-01 1.30033284e-01 1.86387613e-01
5.15203416e-01 2.29852289e-01 -6.22584701e-01 8.61533821e-01
5.07032312e-02 -2.69134678e-02 3.28202337e-01 6.94661081e-01
-6.31150484e-01 -4.32851851e-01 -5.69805741e-01 5.13370574e-01
-8.22981000e-01 -2.65580624e-01 1.28679976e-01 8.57596457e-01
1.85798019e-01 6.04081273e-01 1.58466384e-01 -1.67406976e-01
2.01545104e-01 -2.62431234e-01 3.32128376e-01 -1.29934728e-01
-9.53389049e-01 5.78422844e-01 -4.27138567e-01 -2.92217642e-01
-3.29989165e-01 -1.14120111e-01 -1.73290765e+00 -2.10083619e-01
-4.89646167e-01 3.09578627e-01 5.77309906e-01 5.50539672e-01
1.07090801e-01 1.04937708e+00 5.86161494e-01 -4.46056545e-01
-3.66863519e-01 -9.27117348e-01 -3.76017660e-01 2.56294787e-01
3.85689646e-01 -1.36505708e-01 8.14132616e-02 1.19678438e-01] | [14.536794662475586, -2.612531900405884] |
d1074c89-e8ca-4748-a3e3-c2e3461d3792 | fully-automated-pancreas-segmentation-with | 1906.01795 | null | https://arxiv.org/abs/1906.01795v2 | https://arxiv.org/pdf/1906.01795v2.pdf | Fully Automated Pancreas Segmentation with Two-stage 3D Convolutional Neural Networks | Due to the fact that pancreas is an abdominal organ with very large variations in shape and size, automatic and accurate pancreas segmentation can be challenging for medical image analysis. In this work, we proposed a fully automated two stage framework for pancreas segmentation based on convolutional neural networks (CNN). In the first stage, a U-Net is trained for the down-sampled 3D volume segmentation. Then a candidate region covering the pancreas is extracted from the estimated labels. Motivated by the superior performance reported by renowned region based CNN, in the second stage, another 3D U-Net is trained on the candidate region generated in the first stage. We evaluated the performance of the proposed method on the NIH computed tomography (CT) dataset, and verified its superiority over other state-of-the-art 2D and 3D approaches for pancreas segmentation in terms of dice-sorensen coefficient (DSC) accuracy in testing. The mean DSC of the proposed method is 85.99%. | ['Ningning Zhao', 'Dan Ruan', 'Nuo Tong', 'Ke Sheng'] | 2019-06-05 | null | null | null | null | ['pancreas-segmentation', 'automated-pancreas-segmentation'] | ['medical', 'medical'] | [-6.63024262e-02 1.61391363e-01 -5.64756542e-02 -4.84795421e-01
-6.11455083e-01 -5.08648038e-01 2.42851570e-01 5.07412553e-01
-3.63806605e-01 5.97849548e-01 1.16518037e-02 -3.16908479e-01
-3.51390727e-02 -7.31763959e-01 -7.51379490e-01 -6.70029223e-01
-4.23769474e-01 7.61764288e-01 1.82465658e-01 3.74053031e-01
3.39925219e-03 7.68503785e-01 -7.62883782e-01 -1.66938826e-01
1.18191469e+00 1.28637683e+00 -9.08789039e-02 6.13257289e-01
-3.22850525e-01 4.61537659e-01 -1.50547028e-01 -2.05390174e-02
6.15849376e-01 -7.00818837e-01 -4.55880195e-01 3.05804491e-01
2.37423927e-01 -5.34351945e-01 -1.00877158e-01 1.02592850e+00
4.03414547e-01 -1.59935609e-01 7.45154738e-01 -8.68939340e-01
-2.50049233e-01 7.62232006e-01 -3.17257643e-01 1.10486783e-01
1.70331389e-01 1.00607485e-01 3.79387856e-01 -5.64139068e-01
6.59749985e-01 7.81947136e-01 7.85936952e-01 2.79084474e-01
-1.08041966e+00 -6.10319674e-01 -1.79179356e-01 -5.90146363e-01
-1.11021912e+00 2.80458331e-01 6.60783947e-01 -5.46729624e-01
3.32234681e-01 -9.75931808e-02 1.09526527e+00 3.78636092e-01
5.52786648e-01 8.07334781e-01 1.22542977e+00 -2.45247111e-01
2.67837107e-01 -6.24451600e-02 1.35791197e-01 1.01345706e+00
5.12155533e-01 7.64761195e-02 4.11115229e-01 -9.06682611e-02
1.18567455e+00 8.81511644e-02 -2.17161164e-01 -7.73522317e-01
-1.44611979e+00 8.75520766e-01 8.50417078e-01 3.01950783e-01
-7.92367458e-01 -1.55397668e-01 5.10234475e-01 -1.36817321e-01
2.83377826e-01 3.62813205e-01 -4.03446615e-01 2.82980770e-01
-1.15104866e+00 5.17234430e-02 1.05267882e+00 8.58789206e-01
1.02435164e-01 -1.39881313e-01 -2.52732158e-01 4.10894424e-01
7.21342146e-01 6.12742752e-02 4.96161461e-01 -5.61324596e-01
1.30217195e-01 9.03425813e-01 -1.14910774e-01 -3.15323710e-01
-8.34017277e-01 -7.06644654e-01 -1.02611732e+00 3.74811679e-01
8.32770586e-01 -3.79635274e-01 -1.41277230e+00 1.17926383e+00
6.01406217e-01 -2.96080299e-02 1.48039699e-01 1.38332260e+00
1.19922590e+00 3.29302996e-01 3.12707752e-01 -9.59053636e-02
1.28883076e+00 -1.03761089e+00 -3.70267540e-01 4.29986656e-01
4.99723345e-01 -5.69274962e-01 2.77358860e-01 3.18462312e-01
-1.05120909e+00 -1.89549968e-01 -1.03727686e+00 2.48460591e-01
-9.13290540e-04 2.61392713e-01 5.96239388e-01 6.11084819e-01
-8.33954751e-01 6.28741324e-01 -1.02245188e+00 -6.34058654e-01
6.70032620e-01 3.42391670e-01 -3.45705926e-01 -7.74944946e-02
-8.43885422e-01 8.55487227e-01 6.46608114e-01 2.42528200e-01
-9.69838679e-01 -5.99374354e-01 -8.74018371e-01 1.90792575e-01
-4.68598641e-02 -6.45156384e-01 1.13831544e+00 -9.13097799e-01
-1.75094962e+00 8.79744947e-01 5.47257543e-01 -7.07309067e-01
1.08400452e+00 2.35919431e-01 9.85267088e-02 4.47610319e-01
-2.59839863e-01 8.76573980e-01 3.40508699e-01 -1.31327653e+00
-3.57512891e-01 -4.68842864e-01 -2.46745870e-01 1.59134746e-01
4.40848082e-01 -1.64487675e-01 -2.31022432e-01 -5.31991720e-01
6.88852668e-01 -1.00662625e+00 -4.47448552e-01 4.96213228e-01
-4.41080153e-01 -2.79520615e-03 4.06809062e-01 -9.34499621e-01
5.44833183e-01 -1.80633581e+00 -1.32914782e-01 5.26807666e-01
2.67916352e-01 2.74493605e-01 2.78108865e-01 -2.92793900e-01
4.85946126e-02 -9.46520343e-02 -5.44069588e-01 1.16591081e-01
-3.12883221e-02 5.71536906e-02 4.81234819e-01 8.05354059e-01
7.78028816e-02 9.07856524e-01 -9.33483005e-01 -8.03976536e-01
4.49579746e-01 4.84786987e-01 -2.88607121e-01 4.20913160e-01
-2.22700611e-01 8.11640084e-01 -4.95000064e-01 9.35370803e-01
9.70031738e-01 -2.11582541e-01 3.61174732e-01 -2.16799259e-01
-2.25187480e-01 -1.83276787e-01 -9.42703009e-01 1.67936242e+00
-1.26403794e-01 2.05217510e-01 1.66483104e-01 -8.50058079e-01
1.09384310e+00 6.69739664e-01 9.79296684e-01 -4.92972225e-01
6.03051841e-01 5.51870048e-01 3.63941640e-01 -5.14203846e-01
-1.59532964e-01 -2.98766404e-01 8.87242630e-02 2.88835049e-01
2.80654013e-01 -1.36002809e-01 2.42239207e-01 -1.60085559e-01
6.69743538e-01 4.29947734e-01 4.62379277e-01 -6.52012050e-01
5.37643492e-01 3.35404664e-01 4.47008371e-01 3.73476416e-01
-7.68092334e-01 7.32682824e-01 5.35516143e-01 -8.30179334e-01
-9.95621562e-01 -1.08034778e+00 -5.19403875e-01 1.84170693e-01
3.66770983e-01 4.25648510e-01 -9.69687402e-01 -9.81581867e-01
1.55401602e-01 3.76073420e-01 -5.31973124e-01 4.66327876e-01
-5.80743670e-01 -8.38716269e-01 5.53401232e-01 5.84399879e-01
6.40128791e-01 -1.04820430e+00 -8.94170582e-01 3.85732412e-01
2.25641638e-01 -9.21887934e-01 -3.49662900e-01 4.24812198e-01
-1.56058371e+00 -1.26592028e+00 -1.29637480e+00 -1.12808275e+00
1.07988310e+00 -3.64121646e-01 1.08675647e+00 -3.04463059e-02
-1.15210637e-01 -7.38278553e-02 -2.31711790e-01 -3.05709451e-01
-6.31539166e-01 3.00642708e-03 -3.79186958e-01 -3.47530216e-01
1.55734062e-01 -2.17464507e-01 -9.95109439e-01 1.53227374e-01
-1.02202427e+00 -4.95887501e-03 8.21578920e-01 6.12016857e-01
8.00801873e-01 -3.33021849e-01 4.97554868e-01 -7.18250930e-01
4.34249192e-01 -5.37142515e-01 -9.40644085e-01 1.00227490e-01
-5.97670078e-01 2.33720560e-02 6.08750522e-01 -4.14827675e-01
-8.23691726e-01 6.29756331e-01 -1.91023856e-01 -3.04767817e-01
-5.16185582e-01 4.69318986e-01 3.74462306e-01 -3.45903873e-01
3.56964111e-01 9.66687128e-02 4.34783518e-01 -4.00793672e-01
9.27226096e-02 3.73593390e-01 3.60837936e-01 -2.69596308e-01
3.54768544e-01 3.05690616e-01 2.28151157e-01 -2.23710224e-01
-3.19790632e-01 -2.28098616e-01 -9.07699645e-01 -3.44879240e-01
1.22005761e+00 -7.05275893e-01 -5.53737223e-01 3.78513575e-01
-9.19787049e-01 -3.10600370e-01 -1.59710646e-01 1.09436750e+00
-3.71595919e-01 3.38230699e-01 -8.98633659e-01 -4.78224635e-01
-8.29610825e-01 -1.62522912e+00 6.24388933e-01 5.22690177e-01
4.53550275e-03 -1.10108042e+00 -2.70701181e-02 8.80478323e-02
5.73463082e-01 8.92613530e-01 9.26197767e-01 -9.92274463e-01
-5.53477705e-01 -3.74064147e-01 -5.20422101e-01 1.63963839e-01
-6.82211667e-02 -6.38655797e-02 -4.86654341e-01 -9.84897241e-02
-1.26894787e-01 -8.66231769e-02 5.54420412e-01 1.02002823e+00
9.06889081e-01 6.62524551e-02 -1.46525383e-01 8.23348820e-01
1.77885294e+00 5.01142859e-01 1.64059624e-01 3.21095407e-01
4.32370812e-01 5.74352406e-02 4.00694072e-01 4.26276058e-01
1.92070067e-01 4.96524805e-03 7.06877828e-01 -4.79464322e-01
-9.43321511e-02 -5.11666387e-02 -1.18163146e-01 6.76267028e-01
1.24637976e-01 3.76956202e-02 -8.41091752e-01 5.69347262e-01
-1.45186579e+00 -2.96129465e-01 -1.72049657e-01 2.07526207e+00
5.69985151e-01 1.43192202e-01 3.11980039e-01 -3.78989369e-01
8.56282473e-01 -3.50997508e-01 -6.00679636e-01 -4.65643406e-01
2.64661670e-01 1.22698359e-01 6.38557792e-01 2.59089708e-01
-1.26196837e+00 3.99949163e-01 6.01864529e+00 -5.74572943e-02
-1.30375326e+00 -1.71488196e-01 7.95349598e-01 2.58528054e-01
-4.33504872e-04 -2.21154526e-01 -2.75417149e-01 4.92923677e-01
4.84775186e-01 1.06847927e-01 7.76169822e-02 7.97306716e-01
-6.36963127e-03 -3.41357142e-01 -1.13272035e+00 5.16645908e-01
3.77132744e-02 -1.04525638e+00 -1.44276813e-01 -1.07624708e-02
9.13852990e-01 3.29892367e-01 -4.69294786e-01 1.74376920e-01
7.30336085e-02 -9.22423303e-01 4.68304187e-01 4.61809009e-01
7.13895380e-01 -6.50123954e-01 1.39698184e+00 2.94864506e-01
-1.04442883e+00 2.37589926e-01 -9.46767777e-02 3.55307013e-01
2.12537810e-01 7.51934826e-01 -1.17983508e+00 3.98767531e-01
3.74683589e-01 2.34892607e-01 -1.36143044e-01 1.76737082e+00
-9.52029601e-02 4.49277014e-01 -5.85146427e-01 -1.13440931e-01
6.08608246e-01 -6.18592322e-01 4.99339312e-01 1.29884470e+00
6.02312386e-01 3.31466794e-02 3.16545069e-01 1.16104162e+00
-3.85738820e-01 4.33882624e-01 -3.73089731e-01 1.98511779e-01
-3.54601182e-02 1.50009787e+00 -1.36816776e+00 -6.08590603e-01
-2.85562128e-01 6.32161260e-01 -2.35180587e-01 4.70816798e-04
-9.21838820e-01 -1.72022983e-01 -1.25006750e-01 -1.21683992e-01
2.69537896e-01 3.42568941e-02 -4.19433147e-01 -1.11224329e+00
-1.56979218e-01 -3.47417206e-01 5.69686472e-01 -4.17930156e-01
-1.23759389e+00 6.74592137e-01 -1.34759128e-01 -1.43761468e+00
-2.54622567e-02 -5.84181428e-01 -6.31918192e-01 8.67619872e-01
-1.72146940e+00 -9.74240184e-01 -5.95149696e-01 1.42042458e-01
3.72747213e-01 1.86065733e-01 7.14800358e-01 2.21548721e-01
-3.16362292e-01 2.78353602e-01 1.30800083e-01 5.21705508e-01
3.93402934e-01 -1.52014875e+00 -9.25604813e-03 6.55866444e-01
-4.88222897e-01 2.92739183e-01 3.14774930e-01 -7.73911953e-01
-1.29056573e+00 -1.19741619e+00 5.22059619e-01 1.99551508e-01
2.59970635e-01 1.64633945e-01 -6.74350441e-01 7.01382101e-01
2.15214759e-01 4.57457513e-01 5.49553931e-01 -8.17607999e-01
1.70688555e-01 1.83668762e-01 -1.89523423e+00 1.98406756e-01
2.89603204e-01 4.11984712e-01 -5.55347085e-01 1.73319578e-01
4.24950689e-01 -8.82953823e-01 -1.54174018e+00 5.51364958e-01
9.05120552e-01 -8.05650175e-01 7.41168678e-01 -1.85876697e-01
5.99791348e-01 -2.10851088e-01 1.72057346e-01 -1.24998093e+00
-1.37587830e-01 -2.72624403e-01 -2.28335839e-02 6.78915024e-01
4.27625299e-01 -5.88771105e-01 9.18436170e-01 6.24595225e-01
-4.12405968e-01 -1.00746644e+00 -7.33844519e-01 -4.21291023e-01
3.27217609e-01 1.67850047e-01 4.18134302e-01 9.20850039e-01
-1.06405534e-01 -2.68684030e-01 3.84718388e-01 1.47185758e-01
1.07639682e+00 4.78019446e-01 3.23118120e-01 -1.47113717e+00
2.66657919e-01 -3.82987946e-01 -3.77652198e-01 -7.91605294e-01
-3.39954317e-01 -1.16132045e+00 2.20879808e-01 -1.89788425e+00
2.84167618e-01 -4.37288284e-01 -5.48140466e-01 1.12329058e-01
-4.30882461e-02 1.72847167e-01 2.01445460e-01 1.02209404e-01
-3.11107934e-01 1.34919912e-01 1.88568759e+00 4.34937291e-02
-4.40309584e-01 1.55771792e-01 -2.30195895e-01 6.47969067e-01
9.36359644e-01 -3.62412930e-01 1.00912452e-02 -1.19350873e-01
-5.11575043e-01 5.54138005e-01 1.91438526e-01 -1.06996751e+00
7.95329586e-02 2.45126218e-01 9.25586045e-01 -8.09823811e-01
-2.95100600e-01 -1.18526018e+00 -2.10078694e-02 1.02230036e+00
-3.36920053e-01 -1.04498871e-01 1.48296356e-01 6.47354200e-02
-2.18020856e-01 -2.96114832e-01 1.11519694e+00 -5.90276957e-01
-2.34619200e-01 5.77650726e-01 -2.54604131e-01 -1.93054244e-01
1.21481311e+00 -2.58891582e-01 1.79424301e-01 -1.32582104e-02
-6.90515041e-01 2.75669575e-01 3.64292771e-01 -2.24078417e-01
6.25549138e-01 -1.27254105e+00 -7.40369260e-01 3.29567134e-01
-5.11320904e-02 5.17258644e-01 9.98213589e-02 1.39070368e+00
-1.41955662e+00 4.18985516e-01 -5.42539418e-01 -8.06965411e-01
-9.15552318e-01 3.58870000e-01 6.26614273e-01 -4.44290698e-01
-1.03226733e+00 5.03279269e-01 -8.95798579e-02 -6.96409464e-01
2.78407872e-01 -1.09038734e+00 -1.93108454e-01 -2.70399660e-01
-4.66970261e-03 2.31588662e-01 9.24217608e-03 -5.74886322e-01
-3.96003693e-01 4.66158330e-01 2.09731370e-01 1.84885398e-01
1.41910326e+00 2.31114089e-01 -1.62593052e-01 -1.12597741e-01
1.16499865e+00 -5.14043272e-01 -1.37163758e+00 -4.80838791e-02
-4.82128672e-02 -8.77885446e-02 2.64468700e-01 -1.11431479e+00
-1.36902320e+00 7.06967831e-01 9.23160136e-01 1.56823292e-01
9.93570864e-01 -2.35673934e-01 1.21469510e+00 -1.84737757e-01
4.20592688e-02 -6.73892558e-01 -5.15225530e-01 2.71146536e-01
4.75378752e-01 -1.49418783e+00 3.14533636e-02 -3.07287157e-01
-5.29568791e-01 1.56005609e+00 6.03226602e-01 -6.05175853e-01
6.64358616e-01 3.13443244e-01 5.03584504e-01 -2.01836824e-01
-2.26635672e-02 1.02686666e-01 3.45093399e-01 2.42864400e-01
7.67774701e-01 2.42565632e-01 -5.75838029e-01 3.44688803e-01
2.65657932e-01 4.26922530e-01 4.11178946e-01 9.11030829e-01
-3.77451301e-01 -5.69249034e-01 -5.82965851e-01 6.08065605e-01
-6.40645027e-01 2.33754233e-01 -1.05133839e-01 1.03733873e+00
1.40993476e-01 3.92622530e-01 7.05309138e-02 2.20148012e-01
3.05694006e-02 1.66884556e-01 6.20314300e-01 -1.92392126e-01
-1.01120758e+00 4.08353180e-01 -2.16309592e-01 -3.94165963e-01
-4.24552888e-01 -4.90844876e-01 -1.66763997e+00 2.48240024e-01
-2.32511699e-01 1.75695330e-01 1.07676232e+00 7.89204657e-01
-5.53698651e-02 6.36748970e-01 4.62652534e-01 -9.56017971e-01
-5.91451466e-01 -1.00894451e+00 -6.23695672e-01 6.44208312e-01
2.65025735e-01 -5.06098926e-01 -1.96813166e-01 -1.10961854e-01] | [14.476716995239258, -2.6888554096221924] |
86a29ae1-8613-46b4-a6b9-bb07c1517065 | a-comparative-study-of-different-machine | 2104.07469 | null | https://arxiv.org/abs/2104.07469v1 | https://arxiv.org/pdf/2104.07469v1.pdf | A comparative study of Different Machine Learning Regressors For Stock Market Prediction | For the development of successful share trading strategies, forecasting the course of action of the stock market index is important. Effective prediction of closing stock prices could guarantee investors attractive benefits. Machine learning algorithms have the ability to process and forecast almost reliable closing prices for historical stock patterns. In this article, we intensively studied NASDAQ stock market and targeted to choose the portfolio of ten different companies belongs to different sectors. The objective is to compute opening price of next day stock using historical data. To fulfill this task nine different Machine Learning regressor applied on this data and evaluated using MSE and R2 as performance metric. | ['Muhammad Ilyas', 'Zubair Nawaz', 'Nazish Ashfaq'] | 2021-04-14 | null | null | null | null | ['stock-market-prediction'] | ['time-series'] | [-6.85200334e-01 -2.98035920e-01 -1.79354027e-01 -2.99811065e-01
-4.60354716e-01 -7.44490027e-01 7.73535252e-01 -1.86937213e-01
-4.47637349e-01 1.18027329e+00 -9.05236900e-02 -6.14985526e-01
-2.97343612e-01 -9.72549081e-01 -2.23669205e-02 -6.25267386e-01
-2.08296731e-01 4.64840680e-01 4.07378599e-02 -3.44277769e-01
9.02985752e-01 9.96297479e-01 -1.06469476e+00 2.74939127e-02
2.23786592e-01 1.55805552e+00 1.80724997e-03 3.59414250e-01
-3.76625568e-01 1.12884319e+00 -6.22164249e-01 -5.20398021e-01
1.09806609e+00 -1.94081008e-01 -1.71702340e-01 -2.11412400e-01
-2.34611303e-01 -3.25351775e-01 -4.47370410e-02 9.55056608e-01
-5.60561493e-02 1.62223242e-02 6.27444625e-01 -1.03905642e+00
-2.94622719e-01 1.01829708e+00 -4.44054037e-01 1.04482579e+00
-4.63572085e-01 -1.02856129e-01 1.41992533e+00 -1.06623566e+00
2.47197375e-01 3.72105449e-01 2.47709185e-01 5.42697869e-02
-9.42629457e-01 -8.46910954e-01 -9.29456130e-02 1.55612126e-01
-9.86839533e-01 3.18957642e-02 7.46302485e-01 -6.35562539e-01
8.94788682e-01 3.31063658e-01 8.97722185e-01 2.50685334e-01
8.70334268e-01 2.75914371e-01 1.35144341e+00 -4.79707643e-02
2.95338809e-01 3.58168721e-01 -4.07413095e-02 -8.05391297e-02
6.78081632e-01 3.36098731e-01 -2.54468471e-01 -1.05608873e-01
7.03249514e-01 2.84770340e-01 3.43524618e-03 2.13260353e-01
-1.07086658e+00 1.20361376e+00 1.20147906e-01 6.24655187e-01
-8.92181754e-01 -1.50289610e-01 1.07678317e-01 9.92307365e-01
6.06542885e-01 6.15009725e-01 -1.06773782e+00 -2.09170982e-01
-1.14903009e+00 4.89676982e-01 1.13839567e+00 2.91788489e-01
1.60246119e-01 6.99940562e-01 1.97959676e-01 -5.56851970e-04
2.21265107e-01 6.08068526e-01 9.32005286e-01 -6.65562212e-01
5.81832945e-01 4.93638337e-01 4.28428859e-01 -8.39929044e-01
-5.02715886e-01 -7.60666907e-01 -6.11513734e-01 8.31306458e-01
4.07696575e-01 -3.19275379e-01 -3.95547658e-01 7.93289542e-01
-2.85435468e-01 7.04496726e-02 2.55391449e-01 4.80355829e-01
-6.85124695e-02 9.95524108e-01 -1.41142413e-01 -8.47474694e-01
9.49353933e-01 -5.57264864e-01 -7.91808486e-01 2.19432458e-01
1.04424603e-01 -8.62268865e-01 5.84837869e-02 7.65072882e-01
-1.10193574e+00 -3.81185740e-01 -9.60508883e-01 7.90329337e-01
-5.19335508e-01 1.19722880e-01 5.15827417e-01 1.67162970e-01
-7.26129293e-01 1.05884254e+00 -5.62109530e-01 9.13914025e-01
1.00668939e-02 4.14458215e-01 -1.41210817e-02 1.15746224e+00
-1.17864287e+00 1.31046999e+00 8.96694720e-01 3.87850553e-01
-2.02844888e-01 -7.74531126e-01 -1.98618755e-01 1.94752052e-01
1.15137316e-01 8.22057873e-02 1.21697760e+00 -1.05966496e+00
-1.34465992e+00 6.18102789e-01 6.39960229e-01 -1.20704603e+00
7.25911856e-01 -5.53380372e-03 -8.29272985e-01 -2.26284176e-01
-1.85730174e-01 -2.10270748e-01 7.67078221e-01 -4.18981463e-01
-1.25900173e+00 -3.67915481e-01 -5.79851449e-01 -1.50023505e-01
-3.55192162e-02 1.99364242e-03 6.03884101e-01 -1.21802926e+00
2.27660835e-01 -7.51838565e-01 -1.74585477e-01 -8.97325277e-01
9.50038731e-02 -1.48946017e-01 4.78621930e-01 -1.07058191e+00
1.50902331e+00 -1.74979711e+00 -3.81051630e-01 6.24726772e-01
-2.82945707e-02 3.39735136e-03 4.86785948e-01 5.41038036e-01
-5.75774789e-01 5.79549782e-02 4.09213230e-02 5.83965719e-01
1.03336409e-01 -1.84084296e-01 -1.05108678e+00 4.09156233e-01
2.28635862e-01 7.90083110e-01 1.24053946e-02 2.47215983e-02
1.71429843e-01 -3.15519989e-01 1.63587227e-01 1.06578104e-01
-4.65101600e-01 -6.13330714e-02 -5.18803060e-01 5.20198286e-01
5.75214803e-01 -1.12838231e-01 -1.33155689e-01 4.45289537e-02
-7.64929056e-01 1.57782212e-01 -1.31674147e+00 4.03245896e-01
-1.14081949e-01 6.28391743e-01 -4.85910207e-01 -1.01237762e+00
1.46384037e+00 4.76051390e-01 7.46166050e-01 -8.99873435e-01
1.32620513e-01 8.87190580e-01 1.82866409e-01 -7.19975829e-02
4.95498419e-01 -6.51488900e-01 5.51523194e-02 5.16034365e-01
-4.98600930e-01 2.05073599e-02 2.39455715e-01 -5.33319831e-01
5.12281895e-01 -2.75250524e-01 6.37562096e-01 -3.56325448e-01
5.58028877e-01 9.71652791e-02 4.44491595e-01 1.64276719e-01
7.00491741e-02 -2.94303149e-02 5.94143510e-01 -1.06159878e+00
-1.27863848e+00 -6.77372694e-01 -2.63850629e-01 6.14702940e-01
-6.53573334e-01 7.08852410e-01 -1.08423188e-01 -2.92687535e-01
4.81102616e-01 9.76674557e-01 -4.85008925e-01 3.79982114e-01
-5.43813884e-01 -9.43425298e-01 -3.89742404e-02 3.12047750e-01
4.64626670e-01 -1.36057508e+00 -8.21377814e-01 6.61892533e-01
6.63061082e-01 -6.30562305e-01 -2.00562701e-02 3.92728150e-01
-1.04640710e+00 -1.15029967e+00 -9.74470794e-01 -4.24303174e-01
3.64408195e-02 -2.65147805e-01 1.14441776e+00 -2.89574683e-01
1.47856757e-01 -2.87964761e-01 -5.16204312e-02 -1.10797787e+00
-2.99093753e-01 1.30195946e-01 -1.46396980e-01 3.15052807e-01
4.95277017e-01 -3.76815856e-01 -3.93107355e-01 9.38326642e-02
-7.24448979e-01 -3.89247864e-01 6.43750668e-01 4.80401397e-01
3.76117319e-01 5.27844429e-01 8.82327318e-01 -5.78788519e-01
6.53803885e-01 -5.83522499e-01 -1.79680872e+00 3.13295662e-01
-1.26882720e+00 1.56395957e-01 3.56751144e-01 -1.67594384e-03
-1.13482594e+00 -3.50409001e-01 2.26446509e-01 -1.42957062e-01
5.65831125e-01 8.48443091e-01 4.41699207e-01 1.90710470e-01
-9.62757617e-02 4.27541107e-01 -1.57167260e-02 -6.60553277e-01
-2.64634669e-01 2.68089712e-01 4.09273416e-01 1.24005005e-01
1.08370650e+00 1.37461886e-01 4.00202721e-01 -2.83324450e-01
-7.64338672e-01 -3.69500458e-01 -7.07853973e-01 -1.69012487e-01
4.56299067e-01 -7.90611029e-01 -6.90919757e-01 5.59661865e-01
-7.81849802e-01 6.67457506e-02 -3.27859402e-01 9.56766844e-01
-4.76765275e-01 -4.01991159e-01 -6.55359030e-01 -1.18519139e+00
-7.11961567e-01 -5.67391038e-01 1.25557870e-01 2.48999089e-01
-4.39052954e-02 -1.24617302e+00 5.25586307e-01 3.19445021e-02
5.02431333e-01 5.74699163e-01 5.87143064e-01 -1.54838765e+00
-7.40377843e-01 -6.22074962e-01 -2.63337456e-02 5.36957860e-01
2.14551017e-01 1.88946158e-01 -5.59984684e-01 -2.09812045e-01
6.67936265e-01 1.30728140e-01 7.62002409e-01 5.63758254e-01
4.60480034e-01 -4.44096744e-01 2.79100716e-01 4.38227177e-01
1.79166722e+00 9.16313052e-01 5.18136501e-01 1.08637261e+00
-2.47536041e-03 6.50676608e-01 9.36136842e-01 8.22991967e-01
-2.38655761e-01 2.37387955e-01 1.20400995e-01 4.30621237e-01
8.88159454e-01 8.29515755e-02 4.08547997e-01 9.10997629e-01
-3.84987384e-01 2.79291362e-01 -8.15013170e-01 6.77985027e-02
-1.50311470e+00 -1.46230853e+00 4.36525866e-02 2.00924444e+00
5.56640267e-01 8.23195517e-01 2.52623677e-01 4.53201681e-01
4.09308016e-01 3.97166789e-01 -4.64987785e-01 -2.73634046e-01
-3.16069722e-01 3.21638525e-01 1.02315211e+00 3.13417137e-01
-1.06615722e+00 4.00970459e-01 6.11205578e+00 6.54913366e-01
-1.39186633e+00 -4.42034930e-01 9.75184381e-01 -1.30075574e-01
-4.72839773e-01 -2.06587151e-01 -1.20075119e+00 7.05978274e-01
1.32704473e+00 -7.01103747e-01 1.79050937e-01 8.64717007e-01
5.33632755e-01 3.76854911e-02 -4.97139424e-01 7.69430339e-01
-3.64990801e-01 -1.77958429e+00 1.48673716e-03 4.74538714e-01
7.45074153e-01 6.08647661e-03 2.73277134e-01 2.58563697e-01
2.88067013e-02 -8.52333903e-01 9.36300933e-01 1.27521372e+00
3.77010293e-02 -1.10540223e+00 1.18678188e+00 3.25489521e-01
-1.11800158e+00 -4.23669964e-01 -4.19903457e-01 -2.54369497e-01
7.98278898e-02 5.51385701e-01 -6.56814098e-01 3.90472919e-01
2.77119190e-01 4.72822964e-01 -1.97960302e-01 9.28401887e-01
2.60443121e-01 4.75385129e-01 -1.53506204e-01 -3.25997829e-01
6.19666576e-01 -9.17825103e-01 2.24250227e-01 5.51321805e-01
6.59174681e-01 2.94885337e-01 -2.14474708e-01 6.63622081e-01
1.59657717e-01 3.27042878e-01 -5.62438786e-01 -2.66926855e-01
4.02500272e-01 7.65735388e-01 -7.62665570e-01 -9.24684256e-02
-5.92954159e-01 3.93820435e-01 -2.56000400e-01 7.91038871e-02
-5.65504849e-01 -1.80456847e-01 5.20950079e-01 3.25455993e-01
5.32926738e-01 -3.44142884e-01 -6.09689295e-01 -1.02121818e+00
1.25108868e-01 -8.38204384e-01 5.40884137e-01 -2.67175078e-01
-1.08826005e+00 4.92332727e-01 -1.00176178e-01 -1.50816584e+00
-5.25962591e-01 -1.08285940e+00 -9.42152441e-01 1.19520116e+00
-1.84915960e+00 -4.80890721e-01 5.63950479e-01 4.66850042e-01
7.24435210e-01 -1.16814542e+00 4.24147248e-01 -1.68099925e-01
-5.36820829e-01 -1.99021071e-01 7.60845363e-01 3.76966447e-01
3.00831329e-02 -1.42945337e+00 6.40013278e-01 6.44841075e-01
4.79094982e-01 1.53566271e-01 6.95075095e-01 -1.03594530e+00
-8.49531054e-01 -7.26096749e-01 1.10472810e+00 -1.92468584e-01
1.44681144e+00 6.18297279e-01 -9.14518058e-01 8.37760031e-01
1.98192716e-01 -2.88568914e-01 5.77372551e-01 -6.43403232e-01
8.45918357e-02 -5.99199772e-01 -8.38417113e-01 1.63754687e-01
-3.57820302e-01 -4.84283715e-02 -9.12517965e-01 -1.78363085e-01
3.57731968e-01 3.14954296e-02 -1.12165892e+00 1.38491109e-01
6.29456460e-01 -1.03512979e+00 7.39039302e-01 -7.80597627e-01
-3.64078842e-02 1.15187153e-01 -7.60299563e-02 -1.07494617e+00
-3.49509120e-01 -6.14577472e-01 -1.31518424e-01 1.01412928e+00
8.34980190e-01 -1.02517784e+00 8.70963931e-01 9.94082510e-01
4.56158757e-01 -6.84181988e-01 -9.83685374e-01 -6.56581819e-01
1.62463889e-01 -6.56876490e-02 8.87153983e-01 7.35739529e-01
-3.33076000e-01 -2.09469810e-01 -4.49684858e-01 -8.25328901e-02
7.15962291e-01 5.88517964e-01 2.93540716e-01 -1.55165493e+00
-2.95512043e-02 -9.96686399e-01 -1.01837762e-01 -1.12459436e-01
1.51031449e-01 -5.41729927e-01 -8.37851942e-01 -7.55304635e-01
-4.61975217e-01 -2.02271774e-01 -9.84689355e-01 -1.44871145e-01
1.83843240e-01 -1.03355132e-01 3.91130805e-01 5.84301889e-01
2.88119137e-01 2.49177203e-01 9.34696913e-01 2.44819410e-02
-1.44458309e-01 9.01458383e-01 -1.83232233e-01 7.46834934e-01
1.30025423e+00 -2.93178082e-01 -1.46320403e-01 3.12468618e-01
6.77978694e-01 4.81877118e-01 1.90738756e-02 -7.04430521e-01
1.12529784e-01 -7.45954931e-01 7.25693524e-01 -1.15230751e+00
-2.41531674e-02 -1.03030455e+00 6.42484605e-01 1.00677800e+00
-4.26134318e-01 8.30155313e-01 -1.18334994e-01 2.68842489e-01
-7.02103913e-01 -6.73489392e-01 6.22577190e-01 -3.63510698e-01
-8.90081286e-01 3.48900974e-01 -3.63381177e-01 -4.42210585e-02
1.42854631e+00 -1.24940619e-01 1.30292624e-01 -3.00261259e-01
-6.11346066e-01 9.59727019e-02 -1.32672504e-01 2.63909847e-01
5.97945273e-01 -1.30397463e+00 -1.08030033e+00 1.66116849e-01
-3.92063081e-01 -7.78846979e-01 -4.29858148e-01 4.61828858e-01
-1.18611872e+00 9.38364148e-01 -4.55365300e-01 3.74522358e-01
-7.24950671e-01 4.16441619e-01 6.68452144e-01 -6.34584904e-01
-3.67818922e-01 6.45507932e-01 -6.08672202e-01 4.10220504e-01
-8.08442831e-02 -4.25928593e-01 -7.94681370e-01 9.83412623e-01
7.93270171e-01 6.30030155e-01 2.27089241e-01 -4.49777573e-01
1.14725225e-01 4.74630445e-01 -3.40281799e-02 -3.04965407e-01
1.79437530e+00 1.50702849e-01 -2.62430161e-01 8.78545403e-01
1.08433163e+00 6.68200552e-02 -1.20396495e+00 -1.46393880e-01
1.27206111e+00 -4.24733102e-01 3.02077793e-02 -5.75767756e-01
-1.43710434e+00 3.11633676e-01 4.17831242e-01 8.32102776e-01
8.71982992e-01 -7.00398445e-01 8.59052122e-01 5.04656255e-01
3.61588836e-01 -1.41608667e+00 -4.09413308e-01 1.65151253e-01
1.00071418e+00 -1.29795372e+00 1.77062467e-01 2.72317290e-01
-7.52795935e-01 1.68725336e+00 -1.32356212e-01 -4.59378034e-01
1.32750857e+00 3.21606129e-01 3.85819286e-01 -1.67923689e-01
-9.83070254e-01 1.96517810e-01 4.57104087e-01 -1.23601317e-01
3.26715112e-01 2.41353258e-01 -6.79947019e-01 6.96670294e-01
-6.36783898e-01 -7.63427541e-02 6.03033900e-01 7.45383978e-01
-9.10491347e-01 -8.86376560e-01 -6.48848355e-01 8.44247401e-01
-1.13959515e+00 -1.50027737e-01 -2.09371895e-02 1.10145938e+00
-3.53115141e-01 3.97704214e-01 4.62898493e-01 1.52513921e-01
4.15965885e-01 2.15850234e-01 -1.10170573e-01 -1.17564701e-01
-7.78572261e-01 1.97016537e-01 -6.41558915e-02 -1.21875383e-01
-2.60881573e-01 -9.61195648e-01 -1.09362912e+00 -3.91894162e-01
-1.53642103e-01 5.91498017e-01 6.87161446e-01 8.85214686e-01
-4.61546481e-01 1.89106137e-01 1.44806838e+00 -5.39852202e-01
-1.60649788e+00 -8.80145729e-01 -1.48958004e+00 2.22380921e-01
4.04691845e-01 -3.23534429e-01 -6.44513249e-01 4.65944074e-02] | [4.547225475311279, 4.192534446716309] |
501f21e7-4b11-4c39-b3c1-eecd3299980f | word-level-fine-grained-story-visualization | 2208.02341 | null | https://arxiv.org/abs/2208.02341v3 | https://arxiv.org/pdf/2208.02341v3.pdf | Word-Level Fine-Grained Story Visualization | Story visualization aims to generate a sequence of images to narrate each sentence in a multi-sentence story with a global consistency across dynamic scenes and characters. Current works still struggle with output images' quality and consistency, and rely on additional semantic information or auxiliary captioning networks. To address these challenges, we first introduce a new sentence representation, which incorporates word information from all story sentences to mitigate the inconsistency problem. Then, we propose a new discriminator with fusion features and further extend the spatial attention to improve image quality and story consistency. Extensive experiments on different datasets and human evaluation demonstrate the superior performance of our approach, compared to state-of-the-art methods, neither using segmentation masks nor auxiliary captioning networks. | ['Thomas Lukasiewicz', 'Bowen Li'] | 2022-08-03 | null | null | null | null | ['story-visualization'] | ['computer-vision'] | [ 4.39168245e-01 -1.33969948e-01 1.36231869e-01 -3.67073357e-01
-4.93443847e-01 -3.41795117e-01 7.80706942e-01 -1.44306734e-01
-8.96329656e-02 8.99491310e-01 5.65861404e-01 1.30373865e-01
2.20375612e-01 -6.29118443e-01 -6.23564005e-01 -5.37493229e-01
5.75071812e-01 -1.00425370e-01 5.59479535e-01 -3.82911891e-01
3.43645602e-01 5.53945862e-02 -1.43524885e+00 6.90568030e-01
1.05735314e+00 7.68269956e-01 5.52692831e-01 6.27793550e-01
-3.99787992e-01 1.08874333e+00 -9.86995578e-01 -4.37235922e-01
-1.53594792e-01 -8.77222896e-01 -5.66438317e-01 3.62699896e-01
4.50654745e-01 -3.75082493e-01 -4.01149124e-01 1.06773698e+00
6.67774022e-01 1.12481199e-01 4.23194468e-01 -1.28848147e+00
-1.25090766e+00 6.25123262e-01 -8.39021564e-01 3.26461434e-01
8.15909266e-01 5.69872379e-01 5.51815629e-01 -7.94794381e-01
9.62507904e-01 1.38619900e+00 5.22264898e-01 5.27917385e-01
-1.17348254e+00 -5.13337851e-01 4.54642475e-01 3.71965587e-01
-1.22912323e+00 -3.69524390e-01 1.04127884e+00 -2.53523797e-01
7.33501017e-01 3.85363251e-01 7.92341888e-01 1.50000441e+00
4.13901061e-02 1.01072431e+00 1.23054063e+00 -1.22823790e-01
-6.83522373e-02 1.29464507e-01 -1.51696846e-01 6.13316774e-01
7.15488046e-02 -2.03981131e-01 -6.82043016e-01 3.74078363e-01
1.01581395e+00 -1.11419603e-01 -5.53773761e-01 -3.35613668e-01
-1.43906462e+00 5.71804464e-01 3.52814287e-01 4.19136345e-01
-2.42154434e-01 2.13770613e-01 3.34151387e-01 -1.07859537e-01
5.26650965e-01 4.03726012e-01 1.79698423e-01 -7.68677220e-02
-1.08556175e+00 2.99824655e-01 2.45258689e-01 9.62270081e-01
2.18321979e-01 2.04863012e-01 -8.76792848e-01 7.36944020e-01
6.69578761e-02 5.47275007e-01 2.83524424e-01 -7.84012556e-01
7.47183025e-01 5.57511508e-01 1.65229395e-01 -1.36221206e+00
-4.18707877e-01 -4.67739910e-01 -1.02042747e+00 7.72917569e-02
1.90287337e-01 -7.00491741e-02 -1.01560080e+00 1.67254901e+00
1.77979127e-01 3.24892223e-01 6.89626709e-02 1.24258614e+00
1.23973441e+00 8.26201439e-01 1.17829643e-01 -2.02477649e-01
1.35919356e+00 -1.44225752e+00 -1.17152488e+00 -3.11283022e-01
1.21555030e-01 -7.36029565e-01 1.36236620e+00 8.52837116e-02
-1.31329310e+00 -8.15808952e-01 -1.21600926e+00 -1.60244077e-01
-3.21167618e-01 1.01310417e-01 4.38234091e-01 4.16589111e-01
-1.05751324e+00 3.18172038e-01 -3.90578389e-01 -3.06758672e-01
5.85121334e-01 -2.12496191e-01 -2.78524965e-01 3.26242521e-02
-1.17473304e+00 9.41781163e-01 5.08630037e-01 7.29594752e-02
-5.72149217e-01 -5.51908672e-01 -1.08267212e+00 3.78271341e-02
2.81556308e-01 -1.10617065e+00 1.02238786e+00 -1.08884072e+00
-1.54156017e+00 7.06539333e-01 1.56718884e-02 -1.68624386e-01
7.43834019e-01 -1.69285968e-01 -4.43228483e-01 4.69637185e-01
2.72207260e-01 1.06568062e+00 6.23483002e-01 -1.57808983e+00
-3.52127939e-01 2.10470594e-02 7.64456242e-02 4.82540011e-01
-2.75205851e-01 2.45090142e-01 -7.60472178e-01 -9.83712971e-01
-2.32169684e-02 -4.48760480e-01 -2.87924767e-01 6.57161176e-02
-6.58799291e-01 -7.03927595e-03 1.08633864e+00 -7.38525093e-01
1.26177728e+00 -1.97529030e+00 2.04538286e-01 -4.21305746e-01
1.52286813e-01 7.73198605e-02 -4.01824862e-01 2.49202862e-01
1.54041965e-02 1.55169949e-01 -6.09707117e-01 -7.46278524e-01
-1.74541429e-01 5.12096100e-02 -3.29539001e-01 1.62621722e-01
4.43959981e-01 1.20442641e+00 -9.66201186e-01 -9.71171916e-01
5.31219184e-01 5.88286161e-01 -1.52509749e-01 2.11476609e-01
-3.11040878e-01 7.30750382e-01 -2.46925488e-01 4.22162890e-01
8.01963210e-01 -4.18997318e-01 -1.31234795e-01 -1.95975795e-01
1.30706534e-01 -3.74600403e-02 -1.18813562e+00 2.21055460e+00
-2.51876324e-01 7.46796668e-01 -1.53209835e-01 -7.01856732e-01
9.31862652e-01 1.29040182e-01 1.86486840e-01 -1.03730261e+00
2.08162263e-01 -2.29690596e-01 -4.19742346e-01 -1.00353646e+00
7.58056641e-01 1.58927798e-01 -1.81387112e-01 4.44975972e-01
-2.04240054e-01 -1.81284592e-01 3.30743313e-01 2.56554395e-01
6.32730484e-01 5.40850282e-01 -1.89814791e-02 3.37855681e-03
4.44126159e-01 3.82294506e-02 4.14551616e-01 7.52601624e-01
-9.82379243e-02 1.42404544e+00 6.11629963e-01 -3.55213791e-01
-1.20748830e+00 -1.11127317e+00 9.07799229e-02 8.08369040e-01
7.80314982e-01 -2.17624351e-01 -8.15173030e-01 -6.77845776e-01
-4.56377715e-01 9.51847315e-01 -6.52324736e-01 5.23309857e-02
-6.57510996e-01 -7.92796433e-01 2.98358589e-01 7.12546647e-01
9.65280414e-01 -1.28042519e+00 -9.53502834e-01 3.85146737e-02
-5.43214679e-01 -1.33278692e+00 -7.23506510e-01 -4.74176109e-01
-4.18057233e-01 -8.47145736e-01 -1.02180123e+00 -6.87002480e-01
7.27017403e-01 3.71857494e-01 1.14590430e+00 4.10368256e-02
-2.52215832e-01 -3.86080630e-02 -4.06328946e-01 -2.12310255e-02
-1.66170776e-01 -1.43722072e-01 -4.15049285e-01 4.75858189e-02
-3.29689115e-01 -5.40746868e-01 -8.15717340e-01 3.42672542e-02
-1.04789639e+00 9.04389560e-01 4.72463071e-01 8.05596232e-01
3.78945500e-01 -2.17868209e-01 5.22772670e-01 -8.06323588e-01
9.86386359e-01 -3.56647581e-01 -5.14357165e-03 4.86022025e-01
-2.15925589e-01 -3.16624418e-02 7.16814399e-01 -4.41127032e-01
-1.32198322e+00 -3.17207798e-02 -1.55929383e-02 -5.25851429e-01
-1.32872522e-01 1.64395079e-01 -3.80843550e-01 3.66631061e-01
2.64672160e-01 5.57116330e-01 -9.97390151e-02 -2.70291269e-01
7.36906946e-01 4.70653832e-01 1.06148279e+00 -3.00741434e-01
4.89050776e-01 5.63442051e-01 -3.07361871e-01 -3.01109016e-01
-7.81560957e-01 4.46382947e-02 -6.63201809e-01 -5.33495247e-01
1.10690081e+00 -8.12516570e-01 -4.28822935e-01 4.45946366e-01
-1.67813551e+00 -5.02089448e-02 -3.66030127e-01 1.68582007e-01
-7.20109701e-01 4.33802485e-01 -4.88484949e-01 -6.86829627e-01
-3.84114981e-01 -1.08335197e+00 1.25359499e+00 5.16608953e-01
-2.24728301e-01 -7.37280190e-01 -1.62656099e-01 3.84421349e-01
4.16486621e-01 7.44638741e-01 5.54015279e-01 -1.33725956e-01
-6.58373892e-01 1.05935320e-01 -7.04775572e-01 -4.24352959e-02
5.85366711e-02 3.26835252e-02 -7.40733385e-01 3.46098207e-02
-1.07881781e-02 -1.71144053e-01 1.01236522e+00 2.51456976e-01
1.20468223e+00 -4.19554085e-01 -3.85882467e-01 5.74974358e-01
1.41707599e+00 1.44491926e-01 7.77810037e-01 3.77458781e-01
9.19791520e-01 6.85123920e-01 5.90983152e-01 3.42186868e-01
5.88021159e-01 6.49470210e-01 3.84461462e-01 -4.82740641e-01
-6.03563547e-01 -6.18514001e-01 1.26565292e-01 5.82811415e-01
2.59256721e-01 -6.20752454e-01 -6.13880575e-01 6.24651253e-01
-2.26757812e+00 -1.13510466e+00 -3.50714058e-01 1.52775002e+00
6.63084090e-01 2.17924520e-01 5.45777306e-02 -7.87729844e-02
9.49427366e-01 6.12030864e-01 -5.83271801e-01 -3.64198536e-01
-7.47231543e-01 -2.42730796e-01 1.09878175e-01 2.30506167e-01
-1.02531087e+00 9.78597522e-01 7.04969740e+00 9.35654223e-01
-1.09110868e+00 2.07577422e-01 9.67574954e-01 -1.35301575e-01
-7.01192439e-01 -1.74371660e-01 -3.15329373e-01 7.09554374e-01
3.69871557e-01 5.48675749e-03 -4.56918329e-02 5.70304275e-01
4.84264046e-01 -2.79262930e-01 -7.16626525e-01 1.19978166e+00
7.31387794e-01 -1.69977891e+00 2.96645433e-01 -5.78702271e-01
9.75431502e-01 -7.02744782e-01 2.61341095e-01 -1.78697743e-02
1.36024445e-01 -1.14182317e+00 1.17962265e+00 9.61880326e-01
8.64600062e-01 -6.78580046e-01 6.69548690e-01 2.33013019e-01
-1.21313870e+00 1.50946721e-01 -3.21075544e-02 -2.04928622e-01
7.14373231e-01 1.72740981e-01 -4.19109106e-01 6.68572962e-01
5.53982794e-01 7.80184388e-01 -9.92288172e-01 8.90599489e-01
-4.84113008e-01 4.78706732e-02 5.32658584e-03 -2.92039871e-01
2.66617537e-01 -6.07559383e-02 4.79269624e-01 1.37963307e+00
2.52626061e-01 1.88714251e-01 -2.36749053e-02 1.25330520e+00
1.06663279e-01 1.79781973e-01 -6.43415928e-01 1.01963952e-01
3.46011937e-01 1.19196200e+00 -1.10581505e+00 -6.27308249e-01
-2.85443455e-01 1.57320464e+00 2.28197157e-01 5.08610845e-01
-1.11914957e+00 -3.59508693e-01 3.13631177e-01 1.22672856e-01
2.25489169e-01 -2.55141318e-01 -7.96631455e-01 -1.10445440e+00
3.49318504e-01 -6.41927302e-01 1.41765282e-01 -1.41392148e+00
-1.14573395e+00 8.12883317e-01 1.45919546e-01 -1.28402185e+00
-1.43668473e-01 -8.44858289e-02 -9.53725576e-01 6.24589980e-01
-1.34673178e+00 -1.40635645e+00 -5.17529130e-01 3.92077297e-01
8.48743737e-01 7.01728016e-02 3.87038499e-01 5.62794805e-02
-6.74431443e-01 5.13872862e-01 -1.20665364e-01 8.81852303e-03
5.49270451e-01 -1.03143728e+00 5.48760951e-01 1.06828320e+00
1.07302018e-01 2.54109800e-01 8.25460851e-01 -6.70308828e-01
-9.15711582e-01 -1.08198166e+00 6.67782784e-01 -4.28606212e-01
3.13172340e-01 -5.07380724e-01 -8.11970770e-01 2.19487593e-01
9.30069089e-01 -1.77752987e-01 3.11522722e-01 -4.55421358e-01
-1.71695247e-01 1.11991599e-01 -1.01862204e+00 9.44658041e-01
1.38463795e+00 -2.87152976e-01 -6.47106171e-01 2.35948920e-01
1.18757224e+00 -5.90463042e-01 -3.96032423e-01 4.86652911e-01
3.04309100e-01 -1.23342621e+00 1.02291512e+00 -3.66726667e-01
1.02264464e+00 -5.55881679e-01 2.62753844e-01 -1.07163715e+00
-2.36989394e-01 -4.89683598e-01 2.85916161e-02 1.53000867e+00
2.25168705e-01 -1.60740882e-01 6.55580997e-01 4.24882680e-01
-2.23566562e-01 -8.09622526e-01 -8.08450699e-01 -5.16663194e-01
-2.99062729e-01 -2.55383044e-01 7.83498466e-01 8.89631629e-01
-1.01178207e-01 6.23797119e-01 -7.98193574e-01 -1.83642402e-01
4.11458760e-01 4.11062509e-01 7.12978244e-01 -4.94120598e-01
-1.63726006e-02 -6.59906209e-01 -3.14376742e-01 -1.12018907e+00
-5.83385527e-02 -5.13819873e-01 1.45428956e-01 -2.19124651e+00
5.13452590e-01 5.49874790e-02 -2.19080094e-02 2.92579949e-01
-6.81720912e-01 5.76748669e-01 5.63264072e-01 2.13885605e-02
-1.07834339e+00 9.30888057e-01 1.84382272e+00 -2.87206143e-01
-1.62187055e-01 -6.54904902e-01 -8.46355855e-01 6.07260466e-01
5.60293853e-01 -2.30797410e-01 -5.76484561e-01 -7.32575774e-01
-3.18318489e-03 3.46808881e-01 6.42757475e-01 -1.17778051e+00
3.96316499e-01 -3.10457855e-01 8.08262050e-01 -8.92035842e-01
5.11959791e-01 -4.41699803e-01 2.76849478e-01 3.43172312e-01
-2.90710956e-01 2.69325972e-01 1.25477806e-01 6.68518364e-01
-2.90262341e-01 5.15246093e-02 7.61779070e-01 -1.05965994e-01
-7.74278283e-01 5.00849262e-02 -5.77700809e-02 -7.17403144e-02
1.32990861e+00 -4.69065756e-01 -4.99670237e-01 -5.14880538e-01
-5.57602286e-01 5.30421257e-01 5.44146776e-01 8.37727547e-01
1.03021920e+00 -1.59158683e+00 -1.03944111e+00 8.75211433e-02
1.95007101e-02 1.98062398e-02 7.20657885e-01 6.48950875e-01
-6.07498050e-01 2.43268073e-01 -3.61369431e-01 -7.14529157e-01
-1.19796026e+00 6.37891769e-01 3.44942398e-02 -1.78996861e-01
-8.90857518e-01 8.62303972e-01 3.18271369e-01 1.06300287e-01
5.81501238e-02 -1.29595518e-01 -3.74643534e-01 -7.91981257e-03
5.06720722e-01 -6.57585338e-02 -5.38534760e-01 -6.97643042e-01
-1.90317005e-01 6.94654107e-01 4.93806154e-02 -2.66621798e-01
1.14930832e+00 -5.57414889e-01 2.21377060e-01 4.63933825e-01
8.44994307e-01 -3.26468945e-01 -1.58247447e+00 -7.24412501e-02
-2.70532608e-01 -8.04777622e-01 -4.38843697e-01 -7.61338770e-01
-1.09128022e+00 8.94986093e-01 5.10341346e-01 1.47769481e-01
1.28910100e+00 -7.82005861e-03 9.78757083e-01 -2.69185066e-01
8.71157087e-03 -1.04015851e+00 6.31791055e-01 1.32872567e-01
1.35184979e+00 -1.36456323e+00 -3.72318849e-02 -4.75509077e-01
-1.13131166e+00 1.01644576e+00 1.02341259e+00 -7.02468008e-02
4.85598063e-03 3.02416652e-01 1.41828492e-01 -1.06833503e-01
-6.87587678e-01 -2.07447022e-01 1.60161212e-01 6.14862561e-01
2.93658942e-01 -2.67910957e-01 -5.86142004e-01 8.46109271e-01
-1.64992094e-01 -1.80762872e-01 4.58813369e-01 7.14302421e-01
-3.32941443e-01 -7.91928947e-01 -4.85574216e-01 1.98404267e-01
-1.14204980e-01 -2.55042702e-01 -4.66523349e-01 6.53849840e-01
3.05875540e-01 1.04858065e+00 2.11129889e-01 -3.33946228e-01
3.58639121e-01 -1.24343611e-01 4.97795016e-01 -4.60492700e-01
-5.06874502e-01 4.09452841e-02 2.77577359e-02 -4.56589520e-01
-6.54619992e-01 -5.16233385e-01 -1.17501581e+00 -1.76618248e-01
-3.25499065e-02 -2.52089292e-01 4.46169078e-01 7.03827739e-01
4.40815836e-01 1.05744648e+00 5.88886261e-01 -7.31651664e-01
8.02700967e-02 -1.10813046e+00 -2.89267808e-01 7.77508676e-01
1.33777440e-01 -5.25015771e-01 -6.65296540e-02 3.05236489e-01] | [11.129158020019531, 0.5939027070999146] |
15b89722-2550-4146-b098-c741e0f32f96 | perception-test-a-diagnostic-benchmark-for-1 | 2305.13786 | null | https://arxiv.org/abs/2305.13786v1 | https://arxiv.org/pdf/2305.13786v1.pdf | Perception Test: A Diagnostic Benchmark for Multimodal Video Models | We propose a novel multimodal video benchmark - the Perception Test - to evaluate the perception and reasoning skills of pre-trained multimodal models (e.g. Flamingo, BEiT-3, or GPT-4). Compared to existing benchmarks that focus on computational tasks (e.g. classification, detection or tracking), the Perception Test focuses on skills (Memory, Abstraction, Physics, Semantics) and types of reasoning (descriptive, explanatory, predictive, counterfactual) across video, audio, and text modalities, to provide a comprehensive and efficient evaluation tool. The benchmark probes pre-trained models for their transfer capabilities, in a zero-shot / few-shot or limited finetuning regime. For these purposes, the Perception Test introduces 11.6k real-world videos, 23s average length, designed to show perceptually interesting situations, filmed by around 100 participants worldwide. The videos are densely annotated with six types of labels (multiple-choice and grounded video question-answers, object and point tracks, temporal action and sound segments), enabling both language and non-language evaluations. The fine-tuning and validation splits of the benchmark are publicly available (CC-BY license), in addition to a challenge server with a held-out test split. Human baseline results compared to state-of-the-art video QA models show a significant gap in performance (91.4% vs 43.6%), suggesting that there is significant room for improvement in multimodal video understanding. Dataset, baselines code, and challenge server are available at https://github.com/deepmind/perception_test | ['João Carreira', 'Andrew Zisserman', 'Dima Damen', 'Simon Osindero', 'Yusuf Aytar', 'Stephanie Winkler', 'Junlin Zhang', 'Raphael Koster', 'Hanna Klimczak', 'Alex Frechette', 'Antoine Miech', 'Yury Sulsky', 'Tatiana Matejovicova', 'Carl Doersch', 'Yi Yang', 'Mateusz Malinowski', 'Joseph Heyward', 'Skanda Koppula', 'Dylan Banarse', 'Larisa Markeeva', 'Adrià Recasens Continente', 'Ankush Gupta', 'Lucas Smaira', 'Viorica Pătrăucean'] | 2023-05-23 | null | null | null | null | ['video-understanding'] | ['computer-vision'] | [ 1.13780543e-01 -5.01448587e-02 -2.43827164e-01 -1.60441145e-01
-1.21857417e+00 -9.48870897e-01 7.78125346e-01 1.00114331e-01
-4.01212782e-01 4.31767642e-01 5.62291384e-01 -2.11521447e-01
1.05909906e-01 -4.28574145e-01 -1.02256095e+00 -4.20154721e-01
-1.20067939e-01 4.01020944e-01 2.08204851e-01 -2.00484842e-01
1.27996713e-01 -1.73586369e-01 -1.83352458e+00 1.24486959e+00
6.16651595e-01 1.31399167e+00 1.34015322e-01 1.17493284e+00
2.71756232e-01 1.27356839e+00 -4.52516854e-01 -8.50742757e-01
-7.45865703e-02 -4.24629271e-01 -1.11348307e+00 -1.50858805e-01
8.75479519e-01 -5.18231690e-01 -6.27695441e-01 7.12897122e-01
5.85358143e-01 4.42093045e-01 5.22950292e-01 -1.65934777e+00
-6.36813343e-01 3.76601696e-01 -6.66081533e-03 1.88209221e-01
1.25954759e+00 7.13287532e-01 9.46374655e-01 -8.06660175e-01
9.61294055e-01 1.56106043e+00 5.65680981e-01 7.57197142e-01
-1.01208973e+00 -7.41750002e-01 -1.75936930e-02 7.63921559e-01
-1.15363169e+00 -6.55217767e-01 2.82471627e-01 -5.90812266e-01
9.94688451e-01 4.92274165e-01 6.15359664e-01 2.09586787e+00
-9.68476292e-03 9.17950630e-01 1.15644109e+00 -3.56093310e-02
2.38540471e-01 2.86870301e-01 -1.93733260e-01 7.12906182e-01
-4.44466859e-01 1.16625708e-02 -1.11391628e+00 -7.75876865e-02
4.60392624e-01 -2.62815803e-01 -6.32594347e-01 -2.57744282e-01
-1.64298213e+00 5.97206235e-01 2.77117878e-01 -1.39249608e-01
-2.98391223e-01 2.07411125e-02 7.87247419e-01 3.38714749e-01
5.43025769e-02 3.02744508e-01 -4.15545970e-01 -7.76453257e-01
-7.79027998e-01 4.35047120e-01 8.81256282e-01 1.00385201e+00
2.97107369e-01 -2.89289623e-01 -7.20779300e-01 6.02341175e-01
2.65440792e-01 6.49071395e-01 3.21847796e-01 -1.52604389e+00
7.36395299e-01 3.58309180e-01 9.56453606e-02 -9.20319319e-01
-2.73091376e-01 1.69852123e-01 -5.16021073e-01 -1.62126333e-01
4.27142531e-01 1.32137286e-02 -9.63557422e-01 1.75329149e+00
2.02424213e-01 2.86487728e-01 1.30566388e-01 1.16963768e+00
1.57136762e+00 9.39464808e-01 4.71796453e-01 3.12288627e-02
1.71437478e+00 -1.26677310e+00 -6.34219408e-01 9.78399720e-03
5.12404203e-01 -7.19798446e-01 1.63338447e+00 6.65317774e-01
-1.31231737e+00 -8.23559999e-01 -5.25828779e-01 -2.26012528e-01
-4.91703779e-01 -2.15603769e-01 2.93963730e-01 3.66257697e-01
-1.17907524e+00 3.08946669e-01 -5.91337621e-01 -6.46410286e-01
3.16862762e-01 -1.42768130e-01 -6.46884978e-01 -4.69051838e-01
-1.40775490e+00 8.87673855e-01 3.30271363e-01 -3.57268751e-01
-1.58562315e+00 -8.22028339e-01 -1.06724572e+00 -1.51441693e-02
4.98125076e-01 -6.65709734e-01 1.57068837e+00 -1.16601431e+00
-1.38431644e+00 1.10322189e+00 -1.24066748e-01 -3.35378230e-01
7.46151149e-01 -3.16118836e-01 -6.21980011e-01 8.30795169e-01
1.15656778e-01 1.22027075e+00 6.51911676e-01 -1.22631383e+00
-6.14467621e-01 1.51126176e-01 7.01971948e-01 3.44756871e-01
-1.30649224e-01 1.69705898e-01 -6.76461339e-01 -4.49102670e-01
-5.05383611e-01 -8.33119214e-01 4.27792758e-01 -1.29293367e-01
-3.02062809e-01 -1.45324126e-01 6.11121714e-01 -1.06123757e+00
1.01543057e+00 -2.26324606e+00 3.27688783e-01 -3.31705868e-01
-3.06306109e-02 -1.43032804e-01 -3.73205483e-01 6.96033478e-01
8.64799134e-03 1.38643101e-01 3.07903197e-02 -4.85563457e-01
3.22125584e-01 1.46312770e-02 -3.33988637e-01 1.61957726e-01
1.65602341e-01 8.30530226e-01 -9.95716214e-01 -6.50579870e-01
3.65352958e-01 5.73202193e-01 -6.71019793e-01 3.40892404e-01
-5.35475791e-01 6.90351546e-01 -2.42664907e-02 9.85191762e-01
4.19447839e-01 -3.30159456e-01 -1.51586365e-02 -5.05397379e-01
1.18862711e-01 3.12332898e-01 -8.74798059e-01 2.12643766e+00
-3.28671306e-01 9.48671639e-01 5.98647594e-02 -6.55292809e-01
1.41680136e-01 6.54968143e-01 1.57966197e-01 -1.00427580e+00
7.92985503e-03 -1.66174293e-01 -3.02948147e-01 -1.08466065e+00
5.69254279e-01 2.69301236e-01 -2.18510240e-01 -4.27114628e-02
5.82790375e-01 1.70754448e-01 4.73350376e-01 5.75931966e-01
1.22916663e+00 4.07062501e-01 -1.62746817e-01 9.22396779e-02
3.39617759e-01 4.06055927e-01 2.63864845e-01 8.18142116e-01
-3.57554674e-01 8.50152493e-01 7.48826742e-01 -1.29711898e-02
-8.14757824e-01 -1.18027222e+00 1.42729908e-01 1.60124302e+00
4.03926283e-01 -7.94583738e-01 -9.12518144e-01 -5.18020988e-01
-2.17289224e-01 9.12520349e-01 -6.62365735e-01 -1.75571814e-01
-3.03599805e-01 -5.91477714e-02 7.50226438e-01 5.06481171e-01
7.44094372e-01 -1.27530229e+00 -8.07992876e-01 -3.03057522e-01
-9.54299569e-01 -1.49130428e+00 -1.75795048e-01 -4.60482419e-01
-2.81802654e-01 -1.22448111e+00 -5.88677227e-01 -4.44003731e-01
6.33481005e-03 1.49788797e-01 1.56200957e+00 -2.34816410e-02
5.50899319e-02 1.24190283e+00 -7.59663641e-01 -7.91751742e-02
-3.59221816e-01 -3.74600679e-01 -1.70259625e-01 -1.03161335e-01
1.81495979e-01 -2.19252724e-02 -9.14192677e-01 6.22865856e-01
-7.76769698e-01 3.42336565e-01 3.78353417e-01 7.28981674e-01
3.73381257e-01 -3.94459248e-01 -9.55615565e-02 -1.47064537e-01
4.17190731e-01 -7.37776816e-01 -1.13211535e-01 3.89136493e-01
1.00883573e-01 -5.93735337e-01 2.27994859e-01 -6.65184796e-01
-1.12638831e+00 -4.23133433e-01 -4.45380993e-02 -7.69461095e-01
-6.34220004e-01 4.66064364e-01 -2.43720055e-01 1.85787067e-01
6.20048344e-01 1.64980680e-01 -2.32231691e-01 -1.95616603e-01
4.75210130e-01 3.71294349e-01 8.55274618e-01 -8.01943839e-01
3.18457514e-01 3.95670325e-01 -4.41278756e-01 -6.27533436e-01
-6.97617292e-01 -4.40290242e-01 -1.99994504e-01 -7.09449768e-01
1.28146148e+00 -1.34485996e+00 -1.23586345e+00 1.51930735e-01
-1.13404322e+00 -7.28083134e-01 -4.65452932e-02 4.73214507e-01
-8.45481277e-01 2.38282397e-01 -7.62117147e-01 -7.35437214e-01
2.54151993e-03 -1.31138563e+00 1.21482563e+00 1.80592120e-01
-4.24732149e-01 -7.32414842e-01 -1.25024870e-01 1.04179752e+00
1.17043138e-01 4.33074087e-01 5.38167894e-01 -6.10348403e-01
-6.28103018e-01 1.14524975e-01 -2.86998600e-01 2.15210006e-01
-6.79103494e-01 7.78821483e-02 -1.19826388e+00 -2.85282582e-01
-3.02818149e-01 -1.18074989e+00 1.02523458e+00 2.65719146e-01
1.14759159e+00 -4.05525148e-01 -2.05672756e-01 3.88609916e-01
9.71492946e-01 1.34593427e-01 8.42682064e-01 3.23534757e-01
4.26542878e-01 6.81759477e-01 8.56123507e-01 3.46357733e-01
8.72374475e-01 6.34626150e-01 6.47725940e-01 1.99058115e-01
-2.02147484e-01 -4.11864907e-01 7.97199667e-01 4.11517799e-01
-2.74761051e-01 -6.13421857e-01 -1.16639924e+00 4.84600216e-01
-1.94075382e+00 -1.29632759e+00 6.78235106e-03 2.20109320e+00
6.14866674e-01 5.02292216e-02 3.62000138e-01 -2.76952162e-02
4.80886638e-01 1.09796420e-01 -3.26577544e-01 -3.39220434e-01
-1.51309311e-01 -1.29599392e-01 -3.96570773e-04 2.58468807e-01
-1.13143396e+00 7.04274595e-01 5.85175514e+00 9.94791389e-01
-1.00222826e+00 4.06805247e-01 6.19939864e-01 -3.04070562e-01
-9.03552845e-02 -1.78392947e-01 -4.41436768e-01 6.65391266e-01
1.32184482e+00 4.68880028e-01 5.43022752e-01 5.69847226e-01
7.10775554e-02 -3.67362916e-01 -1.27365839e+00 1.17672873e+00
3.61637563e-01 -1.41153610e+00 1.32637531e-01 -2.95124173e-01
4.17611301e-01 4.34248745e-02 2.88422376e-01 7.99414337e-01
-9.68786925e-02 -1.24941885e+00 1.27951193e+00 6.15387380e-01
1.00797272e+00 -3.57423753e-01 6.34112954e-01 1.33556157e-01
-1.02657270e+00 -2.19008297e-01 3.27741541e-02 -9.50353593e-02
3.40222836e-01 -2.14473814e-01 -3.86819154e-01 6.03938639e-01
1.36890614e+00 6.28852606e-01 -6.24096990e-01 8.97125781e-01
-1.24621943e-01 7.88971663e-01 -6.13767654e-02 3.99595723e-02
2.56267369e-01 2.14942619e-01 5.05101025e-01 1.57564044e+00
2.84776270e-01 2.03002900e-01 1.87283322e-01 5.81982613e-01
-1.35548055e-01 -1.72077522e-01 -2.73248732e-01 -1.16713196e-01
4.87885267e-01 1.21927488e+00 -5.04828513e-01 -5.81303298e-01
-5.47889948e-01 9.96370614e-01 4.94092107e-02 5.57693779e-01
-1.31099474e+00 1.58902269e-03 6.37348652e-01 1.29514322e-01
2.19783902e-01 -1.15137016e-02 2.39432603e-01 -1.22838163e+00
-1.43513620e-01 -1.37465465e+00 8.21598232e-01 -1.38836086e+00
-1.21483767e+00 4.58853394e-01 4.48518038e-01 -1.18820357e+00
-1.77151442e-01 -8.52336049e-01 -4.62614328e-01 3.54636967e-01
-1.15181637e+00 -1.31012821e+00 -8.89627039e-01 8.50987613e-01
7.58945525e-01 -6.43785670e-02 7.81313419e-01 2.71816999e-01
-3.26583117e-01 5.59005439e-01 -3.40515554e-01 1.56727538e-01
1.05672562e+00 -1.10270298e+00 -4.86689918e-02 4.51599211e-01
4.80808280e-02 2.71917582e-01 9.36904609e-01 -4.35017467e-01
-1.47967935e+00 -8.08002412e-01 5.50899684e-01 -9.02396023e-01
7.69721508e-01 -5.50487578e-01 -7.65580118e-01 5.82229257e-01
7.87948132e-01 -2.44811699e-01 7.81834304e-01 1.43577242e-02
-6.86058342e-01 1.82773650e-01 -1.14647877e+00 5.15792608e-01
1.23571920e+00 -6.90427125e-01 -5.58614671e-01 5.78718841e-01
9.43173647e-01 -7.31046319e-01 -1.00339842e+00 5.12560546e-01
7.71584511e-01 -1.30780375e+00 1.11516023e+00 -8.10305119e-01
1.10449743e+00 -2.28802469e-02 -4.93836075e-01 -1.05935085e+00
-1.13682181e-01 -5.65329552e-01 -3.15549940e-01 1.23828983e+00
2.65403956e-01 -1.87366605e-01 3.47142279e-01 6.22258723e-01
-2.09518462e-01 -5.64590335e-01 -8.10667098e-01 -5.53553879e-01
-2.00904340e-01 -6.77563071e-01 2.08560050e-01 9.98042762e-01
1.61141708e-01 2.55012661e-01 -2.73813933e-01 8.71312618e-02
4.14382190e-01 -1.33741260e-01 8.02775681e-01 -7.24749446e-01
-3.71886939e-01 -5.76646805e-01 -3.63373339e-01 -9.07162905e-01
1.44080177e-01 -5.54977298e-01 -2.18529463e-01 -1.59318697e+00
3.89033109e-01 1.66319922e-01 -2.45371863e-01 4.16082978e-01
3.22891325e-02 4.81859028e-01 4.48199838e-01 7.19123706e-02
-1.27500403e+00 3.70788217e-01 1.16097295e+00 -2.31184423e-01
1.73966840e-01 -5.27966738e-01 -2.72568285e-01 5.73343694e-01
5.56954384e-01 3.13186310e-02 -3.58708143e-01 -5.16961217e-01
2.28909522e-01 3.09155613e-01 1.10000122e+00 -1.11152411e+00
1.65003166e-01 -1.13719963e-01 4.36318219e-01 -5.87929249e-01
9.41847503e-01 -4.83375639e-01 1.10103495e-01 1.99417666e-01
-5.62217295e-01 1.95466086e-01 5.07160127e-01 5.40457904e-01
-4.42074984e-01 3.21585573e-02 4.09123182e-01 -2.19056427e-01
-1.13848150e+00 4.01909240e-02 -3.48774016e-01 2.85030693e-01
9.46170270e-01 -2.75725782e-01 -8.23349655e-01 -8.83102119e-01
-8.47495973e-01 5.16297221e-01 4.11614150e-01 7.35818624e-01
6.61274910e-01 -1.28739941e+00 -8.29151094e-01 -4.12488014e-01
4.99580234e-01 -3.30046475e-01 9.07994986e-01 1.07594633e+00
-4.66557473e-01 4.10625786e-01 -3.74856323e-01 -8.69092584e-01
-1.32999611e+00 6.29585207e-01 1.19212039e-01 8.75190049e-02
-3.65997791e-01 9.51917470e-01 3.03420603e-01 -3.70088995e-01
5.30050874e-01 -2.29376078e-01 -1.06253080e-01 1.91834271e-01
6.54211581e-01 5.00362992e-01 -2.57631212e-01 -7.90118575e-01
-4.61235672e-01 3.32600713e-01 4.37778682e-01 -3.31746399e-01
7.87026942e-01 -9.10880342e-02 1.97113767e-01 5.30846298e-01
1.07570648e+00 -3.01271617e-01 -1.29119110e+00 6.50077537e-02
-3.60833824e-01 -3.64205778e-01 -2.91211933e-01 -1.28419983e+00
-5.49308240e-01 1.11087203e+00 6.72573388e-01 2.59382874e-01
1.13163638e+00 3.68137211e-01 5.01866221e-01 2.11056188e-01
1.40243992e-01 -9.75177050e-01 4.20076013e-01 4.02899861e-01
1.28919613e+00 -1.46727908e+00 -4.62383568e-01 -2.55876094e-01
-1.18557799e+00 8.01252723e-01 9.44284558e-01 3.17875236e-01
1.76085472e-01 -1.90962151e-01 -1.97870843e-02 -8.31082389e-02
-1.20189738e+00 -9.86153856e-02 6.01338208e-01 6.08325720e-01
4.03373808e-01 3.05927508e-02 3.29638541e-01 7.01959610e-01
-2.50806212e-01 -1.37144715e-01 3.09745163e-01 6.96284354e-01
-6.20054044e-02 -2.03258783e-01 -4.37362134e-01 1.06644705e-01
-2.03047335e-01 2.59124879e-02 -3.17087322e-01 9.97525692e-01
2.76531100e-01 1.23699629e+00 2.39406019e-01 -5.97939849e-01
4.65035260e-01 1.05254330e-01 6.76179051e-01 -1.71145841e-01
-6.72131538e-01 -9.28229615e-02 4.98081356e-01 -1.11952472e+00
-6.20980263e-01 -7.15298891e-01 -9.61222768e-01 -4.70116436e-01
2.23531067e-01 -2.48407312e-02 4.87867057e-01 7.47398853e-01
4.52830821e-01 5.15360653e-01 -1.07688196e-01 -1.03688884e+00
-1.16875537e-01 -1.12850535e+00 1.58572376e-01 8.14416230e-01
2.21617281e-01 -8.92667115e-01 -3.35945398e-01 3.85876417e-01] | [10.448195457458496, 1.051189661026001] |
671e0a82-06a3-4fc6-909c-4b9d22aa34b5 | satellite-image-classification-methods-and | 1308.1801 | null | http://arxiv.org/abs/1308.1801v1 | http://arxiv.org/pdf/1308.1801v1.pdf | Satellite image classification methods and Landsat 5TM bands | This paper attempts to find the most accurate classification method among
parallelepiped, minimum distance and chain methods. Moreover, this study also
challenges to find the suitable combination of bands, which can lead to better
results in case combinations of bands occur. After comparing these three
methods, the chain method over perform the other methods with 79% overall
accuracy. Hence, it is more accurate than minimum distance with 67% and
parallelepiped with 65%. On the other hand, based on bands features, and also
by combining several researchers' findings, a table was created which includes
the main objects on the land and the suitable combination of the bands for
accurately detecting of landcover objects. During this process, it was observed
that band 4 (out of 7 bands of Landsat 5TM) is the band, which can be used for
increasing the accuracy of the combined bands in detecting objects on the land. | ['Nasser Lotfi', 'Jamshid Tamouk', 'Mina Farmanbar'] | 2013-08-08 | null | null | null | null | ['satellite-image-classification'] | ['computer-vision'] | [ 5.72887994e-02 -4.78939235e-01 -2.81990826e-01 -3.49481893e-03
-1.66964695e-01 -3.25952560e-01 9.99282300e-02 2.54977882e-01
-3.14016104e-01 6.91394031e-01 -1.14483982e-01 -5.07548630e-01
-6.76446438e-01 -1.26380706e+00 -1.63169846e-01 -9.63555634e-01
-9.60045606e-02 4.02068421e-02 2.39377558e-01 -3.98420155e-01
4.11653996e-01 7.31034160e-01 -1.97183979e+00 2.67694265e-01
9.47258174e-01 8.56594086e-01 5.21625280e-01 1.60930023e-01
-1.95766538e-01 1.79431960e-01 -3.47776532e-01 8.95686224e-02
4.18089449e-01 -4.46166068e-01 -6.20423436e-01 8.00105184e-02
2.13331208e-01 -5.91678143e-01 5.22242904e-01 1.03088307e+00
2.91678190e-01 -1.69574711e-02 8.44506502e-01 -7.43071854e-01
-9.49866399e-02 7.56459534e-01 -9.18137014e-01 1.04573719e-01
2.12848201e-01 -1.94829732e-01 9.17238355e-01 -7.78952181e-01
3.40755075e-01 1.07148409e+00 7.76930571e-01 -3.78517658e-01
-1.07917047e+00 -8.48833144e-01 -1.09529577e-01 3.85090590e-01
-1.62384200e+00 4.32497412e-02 6.36355698e-01 -6.05107129e-01
7.58514583e-01 6.35593832e-01 1.11873376e+00 -4.52796072e-02
3.35999757e-01 2.71450609e-01 1.45559502e+00 -8.88812780e-01
-1.12718649e-01 1.18640557e-01 3.12102109e-01 4.23028737e-01
8.83422017e-01 6.98004216e-02 1.83692887e-01 -2.93249339e-01
3.43299180e-01 2.57431567e-01 -2.90924668e-01 -7.29002133e-02
-7.48688400e-01 6.38634205e-01 5.33321977e-01 7.96323657e-01
-7.01583803e-01 -4.37444001e-01 3.79443020e-02 2.94416435e-02
3.77332896e-01 3.88829350e-01 -2.48207957e-01 2.62616694e-01
-1.09710503e+00 1.66183680e-01 3.08882028e-01 2.65248567e-01
1.10383987e+00 -1.99942529e-01 1.32238805e-01 7.98685789e-01
6.03183210e-01 7.46523619e-01 -1.20859198e-01 -2.83590287e-01
2.76653707e-01 1.25497222e+00 1.40158847e-01 -1.22252762e+00
-4.95652765e-01 -6.44600570e-01 -5.52579701e-01 3.56891900e-01
3.13215256e-01 -5.96456863e-02 -9.84130681e-01 8.61779034e-01
2.44547993e-01 -4.79228556e-01 1.62902728e-01 8.85647237e-01
4.55596387e-01 1.00695813e+00 7.43154585e-02 -3.05278718e-01
1.62943256e+00 -2.89442718e-01 -8.18966687e-01 -2.13266164e-01
7.18928039e-01 -9.79935527e-01 6.80481970e-01 2.44568929e-01
-2.74450153e-01 -4.70706165e-01 -1.36474538e+00 6.88754976e-01
-5.74302375e-01 5.32934427e-01 5.34230292e-01 6.01179481e-01
-7.10618317e-01 3.60291541e-01 -5.80622017e-01 -7.86240160e-01
-1.22327637e-02 -8.03850871e-03 -1.78553924e-01 -1.67540997e-01
-9.90335226e-01 1.20364547e+00 6.60722673e-01 4.51713353e-01
-2.77715344e-02 -2.84217864e-01 -4.14223135e-01 -1.09018557e-01
2.63952911e-02 7.58025721e-02 4.09814566e-01 -8.66564751e-01
-8.67434442e-01 7.52625883e-01 1.91625342e-01 -3.52183700e-01
3.01084012e-01 -1.59157798e-01 -4.80228692e-01 2.03489572e-01
2.48125017e-01 1.65682405e-01 2.25227609e-01 -1.27777350e+00
-1.21294534e+00 -6.64213121e-01 -1.05946817e-01 1.92091405e-01
-3.92838210e-01 2.39398866e-03 2.87310928e-01 -1.53480068e-01
7.02976704e-01 -9.75381732e-01 -3.91801074e-02 -2.58182168e-01
-7.11494237e-02 -5.14835343e-02 9.45200682e-01 -9.46906090e-01
1.33989024e+00 -2.33934832e+00 -4.23678428e-01 6.82144105e-01
-3.67859155e-01 4.91190076e-01 3.37881982e-01 7.60222197e-01
-1.63684290e-05 2.67638505e-01 -2.96797425e-01 7.12228835e-01
-2.94050753e-01 2.35201269e-01 7.33836144e-02 5.59180617e-01
-1.64661303e-01 -6.49059936e-02 -6.12833440e-01 -7.03187406e-01
4.39566642e-01 3.21290791e-01 1.59560651e-01 -2.14361355e-01
2.63249874e-01 -1.76143974e-01 -4.68492627e-01 8.92749965e-01
1.28510559e+00 3.17736834e-01 2.18296617e-01 -4.02743608e-01
-6.29126430e-01 -1.62708551e-01 -1.30291045e+00 6.97049260e-01
-1.90399468e-01 5.80727994e-01 -1.60328954e-01 -8.98869753e-01
1.50676739e+00 3.53013545e-01 5.76474845e-01 -5.08175194e-01
-6.63450547e-03 5.76470077e-01 1.44321442e-01 -9.08369899e-01
4.21735168e-01 4.58467044e-02 4.02627110e-01 -5.92959300e-02
-5.64848900e-01 -1.33635163e-01 4.04062599e-01 -4.21316624e-01
3.70542347e-01 3.69087122e-02 5.51363409e-01 -6.15038812e-01
4.15295780e-01 4.05570894e-01 2.46520638e-01 5.63397467e-01
3.11706755e-02 9.05300379e-02 2.44313180e-01 -6.74422622e-01
-7.26057589e-01 -5.90695202e-01 -6.32431984e-01 7.71026194e-01
2.39486828e-01 1.98814899e-01 -5.08347571e-01 -1.89673111e-01
8.31133276e-02 6.21861517e-01 -4.60753709e-01 3.08980018e-01
-2.14929152e-02 -1.21052945e+00 3.62868160e-01 3.50651145e-02
1.09221518e+00 -6.92814231e-01 -9.19932663e-01 2.10132807e-01
-4.38991040e-01 -4.34060514e-01 4.44555908e-01 2.34679326e-01
-1.23126066e+00 -1.13229787e+00 -7.02953279e-01 -6.46889329e-01
7.77287126e-01 8.14096987e-01 5.72996259e-01 4.82189953e-01
-1.11084081e-01 -4.10838366e-01 -8.66647482e-01 -5.24236679e-01
-2.83020437e-01 2.26018488e-01 -6.76774561e-01 5.01152687e-02
5.70310712e-01 -3.22050601e-01 -4.11076009e-01 3.23605895e-01
-7.37820923e-01 6.07159175e-02 9.07815039e-01 4.62448388e-01
4.55134571e-01 6.66842401e-01 8.83410051e-02 -5.27198732e-01
3.44205141e-01 -5.24121583e-01 -6.78248286e-01 3.64109248e-01
-6.60558462e-01 -1.73973471e-01 5.28585650e-02 1.21898971e-01
-8.74426425e-01 1.25563174e-01 -7.45214596e-02 5.41195750e-01
-4.40715373e-01 8.97267461e-01 5.13645932e-02 -3.27863038e-01
8.35037053e-01 2.26865053e-01 1.56835914e-01 -5.50532043e-01
-4.66842234e-01 1.00811648e+00 -9.21677053e-02 -1.00689761e-01
5.50353467e-01 4.48630005e-01 8.94304663e-02 -1.14106739e+00
-5.72877467e-01 -7.54110456e-01 -7.64894187e-01 -5.42402804e-01
9.32166219e-01 -6.81474447e-01 -2.67659456e-01 5.66265643e-01
-8.14081252e-01 1.97773337e-01 4.67176735e-01 9.26447868e-01
1.07841790e-01 1.86584383e-01 -1.55428633e-01 -1.25969684e+00
-2.85239249e-01 -9.99183834e-01 7.27698743e-01 2.16474384e-01
-1.06391788e-01 -4.28390503e-01 -1.92503825e-01 3.55784893e-01
4.19780314e-01 5.00708580e-01 9.80614066e-01 -3.18398744e-01
-1.81030065e-01 -4.33473706e-01 -3.34725142e-01 1.71272665e-01
5.70510507e-01 5.24510741e-01 -5.77707052e-01 -3.09486054e-02
-1.44479543e-01 5.36637723e-01 8.77153158e-01 5.28636754e-01
3.42453748e-01 -1.36796668e-01 -5.41440010e-01 4.68260437e-01
2.01292181e+00 7.94669747e-01 8.89619589e-01 1.10369372e+00
1.83721200e-01 8.36480737e-01 1.16670918e+00 3.50190133e-01
1.89066693e-01 5.22477388e-01 8.25980306e-01 -5.34896612e-01
1.17631920e-01 2.50825971e-01 2.35311523e-01 2.09131151e-01
-8.64504516e-01 6.56166598e-02 -1.19274127e+00 5.73683560e-01
-1.65786815e+00 -1.26146221e+00 -7.75034487e-01 2.12762022e+00
2.58282095e-01 -7.18459021e-03 2.24144444e-01 8.80443156e-01
8.75267804e-01 8.56547654e-02 1.24629043e-01 -3.02855134e-01
-3.10952455e-01 -1.52144089e-01 1.07717967e+00 3.59194279e-01
-1.10575747e+00 4.74902600e-01 6.53479290e+00 4.49323475e-01
-1.20377016e+00 -3.45267713e-01 1.40882656e-01 4.15248781e-01
-1.75749734e-02 1.04975216e-01 -8.41981649e-01 4.38966364e-01
4.03542280e-01 1.65966347e-01 -3.10367402e-02 6.23428106e-01
4.50046867e-01 -8.09698403e-01 -1.28073394e-01 4.04479444e-01
-2.49563619e-01 -9.03417706e-01 1.70360267e-01 2.79372960e-01
7.04728603e-01 -2.25812912e-01 -4.59907025e-01 -2.80768454e-01
3.70739289e-02 -5.17241061e-01 6.54495776e-01 4.33164716e-01
1.51343286e-01 -7.48352289e-01 1.50519598e+00 4.10570025e-01
-1.26790142e+00 -2.27547407e-01 -3.18300515e-01 -5.33765674e-01
-2.56622016e-01 6.01539791e-01 -7.72115588e-01 9.34189379e-01
9.88099098e-01 4.26194817e-01 -6.79517925e-01 1.45035148e+00
-1.82371646e-01 7.74448931e-01 -5.58354020e-01 -2.79723495e-01
4.12029356e-01 -4.11166161e-01 2.16396540e-01 1.23587418e+00
9.42459524e-01 1.12960607e-01 1.75749987e-01 4.39428598e-01
9.18280125e-01 6.18827045e-01 -7.13347554e-01 3.81415218e-01
7.20387042e-01 9.33247387e-01 -1.11125934e+00 -1.86814070e-01
-3.64504576e-01 2.70902425e-01 -5.12612641e-01 3.58602822e-01
-4.53220785e-01 -4.75842029e-01 2.81868368e-01 3.93528312e-01
1.57934651e-01 -2.11699978e-01 -2.86511153e-01 -3.27416480e-01
7.04780445e-02 -5.81703126e-01 4.68248397e-01 -7.11362958e-01
-6.48398519e-01 3.60175163e-01 4.95642245e-01 -1.44266677e+00
2.47307330e-01 -5.56743979e-01 -1.76446274e-01 9.84004557e-01
-1.38054609e+00 -1.20880830e+00 -4.96716350e-01 2.05482900e-01
1.28223762e-01 1.34273231e-01 8.34827423e-01 2.70297617e-01
-4.02659178e-01 -1.36170074e-01 3.90866697e-01 3.89367864e-02
2.03159630e-01 -8.04104686e-01 -3.51079255e-01 1.00495541e+00
-3.41751099e-01 9.86608416e-02 8.88455451e-01 -7.39703178e-01
-9.34356868e-01 -6.26980543e-01 9.38908577e-01 5.23124278e-01
4.01428223e-01 2.58035928e-01 -7.58022666e-01 2.99564272e-01
5.41620143e-02 -4.89452630e-01 6.00069582e-01 7.57446289e-02
1.47237316e-01 -4.94952232e-01 -1.17429507e+00 3.70597214e-01
4.26602781e-01 -4.46650991e-03 -3.41347873e-01 1.22195944e-01
-3.09210494e-02 -1.02232613e-01 -9.81134176e-01 5.15155971e-01
9.66707468e-01 -1.26610851e+00 8.10050428e-01 4.57025141e-01
2.35319048e-01 -5.29932499e-01 -2.80618548e-01 -1.16798854e+00
-5.49934745e-01 3.97025853e-01 8.76173675e-01 1.35521543e+00
7.35850692e-01 -1.06889176e+00 4.56268489e-01 -4.77811508e-02
2.57208478e-02 -4.30535734e-01 -5.55452883e-01 -8.88661027e-01
-3.25817287e-01 -1.41220883e-01 8.76209915e-01 6.38660371e-01
-1.66355267e-01 -3.39419305e-01 -3.27105969e-01 4.63108093e-01
6.38289809e-01 3.01581860e-01 6.42362535e-01 -1.59766638e+00
4.21627581e-01 -4.99878854e-01 -6.37256026e-01 -3.94120038e-01
-3.64862561e-01 -5.15582383e-01 8.37910771e-02 -2.00934172e+00
1.44159272e-01 -7.40488708e-01 -9.39941183e-02 7.05052853e-01
-2.40689412e-01 2.96393186e-01 2.05959782e-01 5.86163700e-01
4.47203398e-01 -2.26692155e-01 1.10686159e+00 -2.48349383e-01
-5.50451934e-01 3.19974720e-01 -5.40172517e-01 6.81307018e-01
9.46390450e-01 -5.33043325e-01 -4.72931489e-02 -4.16572660e-01
3.17515343e-01 -3.01967729e-02 2.16896951e-01 -1.13336182e+00
-7.23259524e-02 -6.35307550e-01 3.75392318e-01 -1.07261074e+00
6.99724928e-02 -1.35192311e+00 6.45977378e-01 9.52660263e-01
1.37616321e-01 -7.10660517e-02 3.24209332e-01 -1.10857897e-01
-2.42725879e-01 -5.95973551e-01 6.78475499e-01 -1.05640158e-01
-1.09810638e+00 -3.66284847e-01 -7.46885359e-01 -9.17032361e-01
1.07369113e+00 -8.77496362e-01 -2.54936248e-01 -1.34207770e-01
-5.08894742e-01 4.32861131e-03 3.39077979e-01 1.33562442e-02
2.57520378e-01 -1.15836966e+00 -9.11633134e-01 1.08774185e-01
2.19392985e-01 -3.50325495e-01 1.43959865e-01 9.00577128e-01
-1.34273899e+00 2.49250367e-01 -7.70392418e-01 -7.21005738e-01
-1.56724346e+00 -1.20344423e-02 3.35629761e-01 1.16553176e-02
-3.68623197e-01 3.14879149e-01 -4.15574551e-01 8.71782303e-02
-1.27594871e-02 -3.67947310e-01 -7.85349369e-01 6.11386120e-01
3.90378833e-01 6.76251054e-01 2.28800356e-01 -9.49338078e-01
-6.43352211e-01 1.03344595e+00 3.29593182e-01 -2.71962322e-02
1.37963986e+00 -1.86372817e-01 -4.74598765e-01 5.31629145e-01
6.63689911e-01 1.08841047e-01 -7.60320783e-01 6.34312555e-02
1.80038452e-01 -6.96716547e-01 1.16708145e-01 -7.94382751e-01
-8.23881269e-01 4.50408816e-01 9.90203857e-01 7.32848883e-01
1.65442407e+00 -4.89661843e-01 2.39919052e-01 2.01332256e-01
5.64464033e-01 -1.03373849e+00 -9.11185145e-01 1.58346415e-01
7.78685570e-01 -1.14241338e+00 3.46328229e-01 -6.01964712e-01
-2.05238894e-01 1.50837052e+00 2.54581779e-01 3.11090029e-03
8.05292845e-01 1.27986640e-01 1.80319324e-01 -1.73830196e-01
3.77562176e-03 -6.12713218e-01 5.81805669e-02 5.49883544e-01
6.53393745e-01 4.90864307e-01 -1.14771473e+00 1.39541621e-03
-2.17251509e-01 1.41436011e-01 1.99094564e-01 9.69730616e-01
-1.16353047e+00 -8.78308415e-01 -1.12638342e+00 6.80107951e-01
-4.73509759e-01 1.87025890e-01 -3.23290110e-01 1.15358293e+00
5.31136155e-01 1.16656005e+00 -6.72252988e-03 -4.67354506e-01
4.55479592e-01 -1.41478553e-01 1.43386602e-01 -1.60086170e-01
-4.78900433e-01 6.12627044e-02 1.29025295e-01 1.03077866e-01
-8.84118497e-01 -7.30662048e-01 -9.07511592e-01 -4.74726856e-01
-9.07267272e-01 4.85956609e-01 8.55920732e-01 1.06535053e+00
-1.18342504e-01 1.22208074e-01 6.50334656e-01 -7.59636104e-01
-1.71312779e-01 -1.23028672e+00 -9.08928335e-01 8.64721462e-02
1.88426167e-01 -7.17014372e-01 -4.88569945e-01 -1.19656309e-01] | [9.595450401306152, -1.7379859685897827] |
7be6ca3a-07cc-4f5d-9c66-7415d6605f38 | neural-speech-synthesis-in-german | null | null | https://www.iaria.org/conferences2021/filesCENTRIC21/30009_centric.pdf | https://www.thinkmind.org/articles/centric_2021_2_30_30009.pdf | Neural Speech Synthesis in German | While many speech synthesis systems based on deep neural networks are thoroughly evaluated and released for free use in English, models for languages with far less active speakers like German are scarcely trained and most often not published for common use. This work covers specific challenges in training text to speech models for the German language, including dataset selection and data preprocessing, and presents the training process for multiple models of an end-to-end text to speech system based on a combination of Tacotron 2 and Multi-Band MelGAN. All model compositions were evaluated against the mean opinion score, which revealed comparable results to models in literature that are trained and evaluated on English datasets. In addition, empirical analyses identified distinct aspects influencing the quality of such systems, based on subjective user experience. All trained models are released for public use. | ['René Peinl', 'Pascal Puchtler', 'Johannes Wirth'] | 2021-10-03 | null | null | null | 14th-international-conference-on-advances-in | ['text-to-speech-synthesis'] | ['speech'] | [-3.23947333e-02 2.00895965e-01 -3.26399982e-01 -2.75125444e-01
-8.26670408e-01 -6.10693455e-01 6.64688587e-01 -2.71261573e-01
-2.47973979e-01 5.89196920e-01 4.19032156e-01 -6.16196036e-01
2.16120422e-01 -4.85071063e-01 -3.25022459e-01 -6.13651156e-01
3.30175608e-01 2.74593383e-01 -4.87911463e-01 -6.36849761e-01
-3.82803649e-01 2.59224236e-01 -1.35124683e+00 4.29630131e-01
4.38516498e-01 8.93234611e-01 2.04839289e-01 9.98486340e-01
9.23480690e-02 7.94318855e-01 -1.10197830e+00 -5.17715573e-01
1.81791708e-01 -7.17155337e-01 -5.46379626e-01 1.34083614e-01
3.12054068e-01 -4.40819830e-01 -4.09803510e-01 8.92406106e-01
1.39650548e+00 1.87503010e-01 2.55620629e-01 -8.73472869e-01
-7.77961493e-01 1.23348868e+00 3.48031312e-01 6.89601749e-02
2.84844279e-01 2.42514431e-01 1.18375957e+00 -9.18475688e-01
4.99409348e-01 1.07795382e+00 5.16927421e-01 7.63595283e-01
-1.07360208e+00 -7.62258291e-01 -1.78690031e-01 -1.04514554e-01
-1.07810259e+00 -1.16982794e+00 8.80928695e-01 -3.11646819e-01
1.39978325e+00 4.65194523e-01 5.49929202e-01 1.70683539e+00
-1.08099811e-01 7.09047616e-01 9.06909347e-01 -5.51542282e-01
2.65988708e-02 3.85531008e-01 -2.88374722e-01 3.75425458e-01
-1.06240757e-01 2.87758946e-01 -8.59447837e-01 2.42058441e-01
3.20504427e-01 -7.54026234e-01 -2.60533184e-01 2.92510986e-01
-1.20954216e+00 1.01745152e+00 -8.63441452e-02 5.97679257e-01
-5.49599946e-01 -2.25204565e-02 8.49010885e-01 6.42651856e-01
8.58590007e-01 2.66121864e-01 -6.81753159e-01 -4.87613171e-01
-1.25310647e+00 3.65137160e-01 8.38331878e-01 1.04562271e+00
1.06265999e-01 9.75013375e-01 -1.92348927e-01 1.37271726e+00
2.00916097e-01 9.10468578e-01 9.94905353e-01 -5.75994134e-01
5.74197292e-01 3.74368615e-02 -3.49987864e-01 -7.49910414e-01
-4.57206994e-01 -7.69546390e-01 -8.39440227e-01 -1.27122268e-01
-2.65466142e-02 -7.92423785e-01 -8.17727387e-01 1.41419923e+00
-2.61891633e-01 -4.89695460e-01 6.95826828e-01 7.13442147e-01
1.73637557e+00 9.99296784e-01 -1.52063727e-01 -2.60601819e-01
1.15812075e+00 -1.38409185e+00 -1.41849399e+00 -2.84820825e-01
5.04928410e-01 -9.95060861e-01 1.09108150e+00 5.62490463e-01
-1.56219709e+00 -8.44770730e-01 -1.15325403e+00 -1.99033901e-01
-6.08686924e-01 6.86792552e-01 1.17570944e-01 1.28362823e+00
-1.33552885e+00 4.58204776e-01 -4.46188658e-01 -3.07844877e-01
5.50135523e-02 3.33906382e-01 -1.83014661e-01 6.43411934e-01
-1.54048800e+00 1.17055166e+00 5.02523661e-01 1.49889842e-01
-7.53866494e-01 -4.05576319e-01 -8.14897597e-01 -5.38907759e-02
1.44009784e-01 -4.34173524e-01 1.78287840e+00 -1.26248324e+00
-2.55485129e+00 7.22866118e-01 4.45277020e-02 -9.23203647e-01
4.74863440e-01 -1.26708582e-01 -1.08537591e+00 -9.30946991e-02
-3.40749860e-01 5.27623594e-01 9.25550878e-01 -9.98800933e-01
-5.15692294e-01 2.88828731e-01 -3.25804064e-03 2.02340260e-01
-3.95040125e-01 2.75336921e-01 -2.45534092e-01 -1.01766908e+00
-2.77334392e-01 -6.67396426e-01 8.72666389e-02 -5.68292677e-01
-5.10026455e-01 -3.95792909e-02 9.01061237e-01 -1.13337231e+00
1.43811393e+00 -2.04219103e+00 -1.86003987e-02 -1.27725452e-02
-2.41584286e-01 6.30280972e-01 -1.32635608e-01 6.84470415e-01
-1.83014080e-01 3.31599385e-01 1.11069746e-01 -6.71407819e-01
3.16060513e-01 -8.77204910e-02 -5.49104691e-01 2.68181473e-01
6.62032142e-02 6.69403136e-01 -5.38944304e-01 -4.56048548e-02
6.20197177e-01 5.95173776e-01 -4.95335281e-01 3.34847987e-01
-2.01877728e-01 2.05901325e-01 9.12551880e-02 4.23179805e-01
3.10923308e-01 1.13790676e-01 3.90888602e-02 -1.28328189e-01
-2.47368023e-01 7.95455575e-01 -7.77758658e-01 1.45683527e+00
-9.66528535e-01 1.12377083e+00 1.63994372e-01 -7.61982501e-01
1.02826798e+00 1.18037176e+00 2.40554616e-01 -6.79452121e-01
5.08594215e-01 5.15429854e-01 2.89393872e-01 -4.72280324e-01
6.74435556e-01 -8.90917331e-02 1.80185810e-01 6.96038753e-02
6.92158759e-01 -7.73296416e-01 3.01146507e-01 -2.73950368e-01
4.88291383e-01 -9.80226472e-02 1.73624650e-01 -1.51917711e-01
5.73975503e-01 -2.83758044e-01 -1.28569929e-02 5.27615607e-01
-1.39408678e-01 8.67520392e-01 1.77261785e-01 -1.11190386e-01
-1.18878543e+00 -5.81793785e-01 -5.99482842e-02 1.20085299e+00
-6.04430676e-01 -7.34968603e-01 -1.17711353e+00 -2.05031931e-01
-6.14507198e-01 8.75757277e-01 -2.72136599e-01 -6.04904778e-02
-2.17297077e-01 -4.87420261e-01 9.92212355e-01 1.61982805e-01
4.36597109e-01 -1.48940682e+00 -2.64259964e-01 4.63225901e-01
-3.76770347e-01 -1.43891191e+00 -3.24882388e-01 3.02058518e-01
-4.59774852e-01 -2.28865489e-01 -9.86194730e-01 -6.37050390e-01
-7.02202022e-02 -2.35626638e-01 9.90899742e-01 -1.81164891e-01
4.51188445e-01 1.63395137e-01 -5.61221004e-01 -9.53962922e-01
-1.03722668e+00 5.21550655e-01 1.73508465e-01 1.92240831e-02
1.52947962e-01 -2.25499094e-01 -3.41288149e-01 9.46904439e-03
-7.01701641e-01 1.69506893e-01 5.42338371e-01 8.58210266e-01
8.99435580e-02 -8.11124817e-02 8.90490294e-01 -5.56073904e-01
1.05543792e+00 -2.24967033e-01 -3.56385082e-01 -4.56725806e-02
-3.82027566e-01 -2.59459972e-01 7.55433142e-01 -4.50219154e-01
-1.03380668e+00 -2.37870961e-01 -1.00011718e+00 -2.44730815e-01
-1.84355512e-01 8.02280962e-01 -1.09962545e-01 1.94022283e-01
7.89279103e-01 5.85110150e-02 -9.16034058e-02 -2.66206056e-01
3.44469130e-01 1.24271739e+00 2.84260869e-01 -1.76735684e-01
5.09022295e-01 -3.85880508e-02 -6.49125457e-01 -1.41732359e+00
-5.24742007e-01 -7.13597611e-02 -2.92744815e-01 -2.96591133e-01
7.90674448e-01 -1.24104011e+00 -3.59304368e-01 7.13727593e-01
-1.09060252e+00 -5.62760413e-01 -4.53705102e-01 6.60085678e-01
-6.21240377e-01 -1.15878306e-01 -5.68007827e-01 -8.15519333e-01
-7.76040435e-01 -1.27598143e+00 1.04839957e+00 8.99600051e-03
-4.03830647e-01 -1.07128823e+00 -8.54969583e-03 6.49114728e-01
8.91322732e-01 -2.33505055e-01 5.12877882e-01 -9.07987058e-01
2.79438555e-01 -3.08650464e-01 4.48679239e-01 8.83320093e-01
1.29928932e-01 3.99898589e-01 -1.52764308e+00 -3.77086967e-01
-4.21320088e-02 -4.47561979e-01 4.38472509e-01 6.29484415e-01
9.42303479e-01 -3.98831695e-01 3.33590716e-01 3.61006528e-01
6.39907956e-01 3.50812316e-01 4.17869747e-01 3.51280123e-02
4.13295239e-01 6.79130495e-01 8.20430741e-02 3.94117892e-01
-2.55418122e-02 6.24672592e-01 1.77750681e-02 -3.28263968e-01
-4.92043495e-01 -1.61720961e-01 8.19123924e-01 1.72076261e+00
-8.29309970e-02 -8.09926748e-01 -6.93023562e-01 4.78464097e-01
-1.19888473e+00 -9.63947356e-01 8.05370882e-02 1.79746807e+00
8.35472405e-01 2.98112243e-01 3.03007811e-01 3.54065508e-01
5.96580982e-01 5.69992065e-01 -2.36381859e-01 -7.63868093e-01
-6.31813526e-01 5.43738782e-01 3.35495472e-01 4.55003917e-01
-1.03714824e+00 1.06286597e+00 7.14366341e+00 1.15556288e+00
-1.52488816e+00 3.60655308e-01 7.71063328e-01 -2.48913661e-01
-1.52292535e-01 -2.36713603e-01 -6.55503750e-01 2.78186560e-01
1.71432984e+00 -1.92131847e-01 5.11422873e-01 8.18736672e-01
7.69975543e-01 4.71872836e-01 -7.91082799e-01 1.00641525e+00
-3.55343036e-02 -1.33137596e+00 5.48792109e-02 -3.84578966e-02
1.00533557e+00 2.61934966e-01 4.06993866e-01 4.97321814e-01
1.36490181e-01 -1.07770383e+00 1.29818249e+00 5.88814020e-02
9.33438897e-01 -6.68931663e-01 8.42378378e-01 2.00384468e-01
-7.65290082e-01 -3.70844416e-02 3.84123288e-02 -2.83339117e-02
1.45062983e-01 4.13857341e-01 -8.91329348e-01 5.59937656e-01
4.74497706e-01 5.07894576e-01 -2.26722717e-01 3.78062159e-01
-1.50262117e-01 1.06242621e+00 -1.25051036e-01 -4.12120193e-01
3.54617268e-01 5.22490367e-02 5.11452496e-01 1.43028796e+00
5.72272599e-01 -2.78528005e-01 -1.58411190e-01 7.55636215e-01
-2.15292171e-01 5.13614237e-01 -6.14737511e-01 -6.48025692e-01
3.41727465e-01 1.07106984e+00 -3.34776610e-01 -3.46848369e-01
-5.46811759e-01 6.96929574e-01 -3.06250900e-01 4.92815048e-01
-7.97688603e-01 -3.81649196e-01 4.82564062e-01 -1.13170929e-01
5.90460040e-02 -1.92646027e-01 -3.23938370e-01 -1.01670849e+00
-5.70005365e-02 -1.44764245e+00 -1.73963070e-01 -8.83921146e-01
-1.08202076e+00 1.07889760e+00 -1.40840977e-01 -1.17200637e+00
-5.18887043e-01 -7.46067286e-01 -4.36867148e-01 1.11192489e+00
-1.12568986e+00 -1.27236748e+00 1.33862212e-01 3.80133063e-01
1.04411972e+00 -8.64069641e-01 1.12139714e+00 5.30964255e-01
-6.88179314e-01 7.71961749e-01 2.93335676e-01 9.37879831e-02
5.86146295e-01 -9.22890723e-01 8.37863684e-01 8.73804927e-01
2.70423621e-01 4.51586217e-01 7.96689093e-01 -2.78297484e-01
-1.11277938e+00 -8.71851146e-01 1.00803947e+00 7.59592187e-03
5.70749462e-01 -6.12844288e-01 -4.45105553e-01 4.58745927e-01
1.06237590e+00 -4.35884029e-01 5.93090475e-01 -1.35260880e-01
1.36146963e-01 -4.48119864e-02 -8.90095115e-01 8.07330072e-01
6.63460910e-01 -8.55767250e-01 -2.10702360e-01 2.49495521e-01
8.72398853e-01 -7.26657629e-01 -7.89865255e-01 2.73870826e-01
4.51560020e-01 -9.11014795e-01 7.00935483e-01 -3.70593429e-01
4.59043235e-01 4.97024097e-02 -4.94260907e-01 -1.69837165e+00
6.91862106e-02 -1.13227713e+00 -5.09540290e-02 1.26579952e+00
9.00543332e-01 -5.90548754e-01 3.99680376e-01 4.90616448e-02
-4.99487519e-01 -4.64181721e-01 -9.24610555e-01 -5.51144004e-01
2.50885129e-01 -6.03182256e-01 6.35066211e-01 9.08130705e-01
-1.66267484e-01 5.68407357e-01 -5.85893333e-01 5.71690127e-03
-1.90700158e-01 -5.11235714e-01 7.69274473e-01 -8.46562326e-01
-2.19499424e-01 -8.83383453e-01 -1.12153888e-01 -8.40326667e-01
2.62961358e-01 -8.31492960e-01 1.09183468e-01 -1.36319494e+00
-5.22392392e-01 4.85585444e-02 8.28888491e-02 3.57564121e-01
2.49459624e-01 1.30144045e-01 1.58038899e-01 -2.93910831e-01
2.30723634e-01 6.36147916e-01 1.10036314e+00 -3.33029658e-01
-2.49165624e-01 7.62468353e-02 -5.27899265e-01 6.05374098e-01
1.06615400e+00 -7.22854957e-02 -5.68874717e-01 -3.32072675e-01
7.54480511e-02 7.10092857e-02 2.72909831e-02 -1.28476405e+00
-7.22912746e-03 1.43630505e-01 6.82587549e-02 -5.67568660e-01
6.39177501e-01 -6.99438035e-01 2.33983874e-01 1.68745682e-01
-6.02595270e-01 6.58195093e-02 3.84275883e-01 -2.62799889e-01
-5.61992109e-01 -2.42639050e-01 8.77001464e-01 5.53821176e-02
-3.34484875e-01 3.91095355e-02 -7.78810024e-01 -7.41435811e-02
4.63523954e-01 -6.85026720e-02 -2.71016449e-01 -1.06048739e+00
-8.34190726e-01 -4.28563207e-01 -3.18578631e-01 6.87811315e-01
5.12418091e-01 -1.32962823e+00 -9.82055545e-01 4.90716070e-01
-7.14627132e-02 -4.98925567e-01 4.03226405e-01 5.84205508e-01
-5.90726614e-01 7.59201467e-01 -3.06735300e-02 -2.44329050e-01
-1.21616995e+00 1.45236045e-01 6.81087852e-01 -1.02469940e-02
-3.35147142e-01 7.19426453e-01 -3.01076263e-01 -9.36587274e-01
5.60988069e-01 -5.55570662e-01 -2.04830587e-01 2.14454696e-01
2.18258500e-01 3.05808365e-01 6.80926919e-01 -1.06731224e+00
1.58941988e-02 6.29812405e-02 3.10389340e-01 -7.48187244e-01
1.21001863e+00 -1.11085869e-01 2.36651599e-01 7.45261550e-01
1.17991042e+00 4.55703996e-02 -6.52575672e-01 -7.58303627e-02
-4.87203807e-01 1.73488006e-01 4.49824393e-01 -1.09816051e+00
-1.34747887e+00 1.15562689e+00 6.94831729e-01 5.95806360e-01
1.11306834e+00 -4.71420616e-01 8.20492208e-01 5.82833469e-01
-2.00763986e-01 -1.45082533e+00 -2.14923546e-02 8.85017514e-01
1.19925034e+00 -1.10644436e+00 -3.51652890e-01 -1.35932583e-03
-7.96238542e-01 1.13198495e+00 2.82210022e-01 3.30354720e-01
8.25023532e-01 4.54956323e-01 5.61930537e-01 1.42868370e-01
-7.91901946e-01 -4.29419465e-02 3.72291893e-01 4.31593239e-01
1.00925422e+00 1.40866920e-01 -2.77187198e-01 8.39663684e-01
-1.05545640e+00 -2.48037249e-01 5.48800826e-01 6.23190463e-01
5.01734577e-02 -1.05239749e+00 -2.63482779e-01 2.22693861e-01
-9.43270564e-01 -3.01132292e-01 -5.73303223e-01 5.85998654e-01
2.95178648e-02 1.50261414e+00 -1.52400449e-01 -5.83959401e-01
6.66652560e-01 2.22673342e-01 5.31512238e-02 -4.36001241e-01
-1.18626559e+00 3.90584916e-01 6.91533744e-01 9.99283716e-02
-5.39906323e-01 -5.04362762e-01 -7.93254495e-01 -4.93973732e-01
-6.59333944e-01 5.55978641e-02 1.13135028e+00 8.53931367e-01
1.43690482e-01 9.66860652e-01 6.14111841e-01 -1.03658891e+00
-4.95800734e-01 -1.40768707e+00 -5.25249124e-01 4.40914966e-02
7.11253881e-01 -1.50680900e-01 -3.31139117e-01 3.06594074e-01] | [14.794325828552246, 6.660373210906982] |
77025c89-9ad7-469b-9fb5-4bf5a1fd3347 | a-proposal-project-for-a-blind-image-quality | 1512.04354 | null | http://arxiv.org/abs/1512.04354v1 | http://arxiv.org/pdf/1512.04354v1.pdf | A proposal project for a blind image quality assessment by learning distortions from the full reference image quality assessments | This short paper presents a perspective plan to build a null reference image
quality assessment. Its main goal is to deliver both the objective score and
the distortion map for a given distorted image without the knowledge of its
reference image. | ['Stéfane Paris'] | 2015-11-04 | null | null | null | null | ['blind-image-quality-assessment'] | ['computer-vision'] | [ 5.97138882e-01 8.27038381e-03 4.98362303e-01 -5.06957948e-01
-5.96932948e-01 -2.76110321e-01 6.76090419e-01 -3.23564857e-01
-1.50732517e-01 5.18211603e-01 3.71281296e-01 9.30728987e-02
-3.29896152e-01 -6.15089357e-01 -1.00004092e-01 -7.28417814e-01
-9.80460942e-02 -2.90370196e-01 1.23155646e-01 -1.53399572e-01
6.30778491e-01 5.43229699e-01 -1.48650932e+00 -3.24577689e-02
9.17602360e-01 9.20336127e-01 5.74569523e-01 9.81251121e-01
7.27562010e-01 4.73517448e-01 -9.27275598e-01 -4.41615850e-01
4.51714486e-01 -6.49794757e-01 -6.00333333e-01 4.92987812e-01
6.45619392e-01 -4.82500792e-01 -4.37475175e-01 1.52617538e+00
8.79002213e-01 3.48139033e-02 8.30773175e-01 -9.18872178e-01
-5.42124569e-01 6.12452254e-02 -1.75787210e-01 5.40639579e-01
5.90749323e-01 2.95706898e-01 5.08416295e-01 -7.72746444e-01
7.85883784e-01 6.80307567e-01 8.81488398e-02 1.27469197e-01
-9.17251348e-01 -9.47751701e-02 -4.56230491e-01 4.94999439e-01
-1.34860396e+00 -5.33842742e-01 6.13274217e-01 -2.16656789e-01
6.63553476e-01 3.63893300e-01 4.04306024e-01 4.22672570e-01
6.05956614e-01 -6.49643457e-03 9.90736663e-01 -5.27895868e-01
7.06526339e-02 6.09558001e-02 -5.75687513e-02 1.80058494e-01
2.22086295e-01 4.93090451e-01 -3.39172274e-01 5.37738562e-01
6.62542880e-01 -6.43390894e-01 -8.60679805e-01 -4.30418193e-01
-1.08680558e+00 1.27356991e-01 4.33500081e-01 6.86412215e-01
-3.07359159e-01 -1.23814307e-01 -1.61826938e-01 5.86416483e-01
1.78296074e-01 6.26108110e-01 1.71310529e-01 -1.42504305e-01
-1.03353024e+00 3.12963203e-02 2.83486992e-01 6.43824279e-01
3.53690237e-01 4.07947034e-01 -2.26577416e-01 4.20812875e-01
1.47967681e-01 9.91158187e-01 1.80315554e-01 -1.49998605e+00
2.63084739e-01 1.16329215e-01 3.68768662e-01 -1.19196093e+00
-1.49106637e-01 -9.76411760e-01 -6.87442422e-01 1.07596636e+00
1.05793923e-01 -6.39698878e-02 -5.71417153e-01 1.17729950e+00
-1.57744691e-01 -2.79189408e-01 2.48230249e-01 1.02867854e+00
9.97291327e-01 8.72392178e-01 -8.06550264e-01 -7.63819695e-01
5.02358973e-01 -8.49685848e-01 -9.85592067e-01 2.35296503e-01
3.03423149e-03 -1.06805325e+00 6.26872361e-01 8.80928814e-01
-1.65808105e+00 -9.36279535e-01 -1.58385897e+00 1.18736461e-01
1.22678548e-01 -1.84971020e-02 -2.39406019e-01 9.23759043e-01
-1.47621155e+00 2.94252157e-01 4.42254823e-03 7.15556890e-02
-1.91281021e-01 4.77228053e-02 -3.05382341e-01 -2.28551015e-01
-7.22822726e-01 1.36972857e+00 1.36890501e-01 -6.07887171e-02
-1.05184007e+00 -5.26094139e-01 -5.54152548e-01 -6.47070818e-03
5.42751960e-02 -5.31180322e-01 8.30215812e-01 -8.48018050e-01
-1.36960101e+00 9.34173584e-01 -1.48476034e-01 -1.95051491e-01
6.34289324e-01 1.15979344e-01 -7.07320094e-01 3.41264039e-01
-5.23956865e-02 2.71654159e-01 8.35694730e-01 -1.56340110e+00
-8.32040310e-01 -2.96105951e-01 7.23380893e-02 7.77912676e-01
2.78209984e-01 6.20412864e-02 -4.43775684e-01 -6.02004826e-01
2.01645657e-01 -2.58191288e-01 1.20366439e-01 -1.36135861e-01
-4.30408344e-02 4.33318615e-01 7.31648862e-01 -7.53886759e-01
1.11756015e+00 -2.03349853e+00 3.28728497e-01 1.92112967e-01
1.22744583e-01 4.07715440e-01 -3.28950465e-01 2.20507085e-01
-2.70111054e-01 -1.29174381e-01 -1.53554305e-01 1.61237419e-01
-3.51103336e-01 -2.05924779e-01 -7.63747543e-02 7.14617729e-01
-1.45962849e-01 4.80162472e-01 -7.42967129e-01 -3.20898265e-01
6.12259328e-01 6.56405747e-01 -2.42899954e-01 3.25070500e-01
6.18275940e-01 3.87803823e-01 3.55023712e-01 3.96591991e-01
1.03648829e+00 3.69169027e-01 -1.64589614e-01 -5.89340925e-01
-3.82126451e-01 -2.57444561e-01 -1.46089816e+00 1.33313966e+00
-2.30533257e-01 1.05261528e+00 1.63753122e-01 -5.65604925e-01
8.88784468e-01 4.38069403e-01 4.59609956e-01 -1.27234781e+00
-1.17874369e-01 1.31311476e-01 9.90811363e-02 -5.53489923e-01
4.69507039e-01 -7.83561990e-02 3.59000891e-01 1.85552090e-01
3.74479085e-01 -6.98547959e-01 2.30106309e-01 1.92433119e-01
6.43183589e-01 -1.14614563e-03 4.27642643e-01 -3.11175376e-01
1.08964539e+00 -4.47806388e-01 1.58945218e-01 5.09447813e-01
-3.07060629e-01 9.94749129e-01 2.13070065e-01 -3.27572197e-01
-1.30846465e+00 -1.46250987e+00 -3.92440975e-01 9.26488489e-02
7.16423213e-01 7.20939264e-02 -7.76251733e-01 -1.49901882e-01
-8.27008367e-01 7.68634737e-01 -1.22577734e-01 1.11945227e-01
-3.38561893e-01 -4.78237480e-01 7.62803108e-02 -3.50864083e-01
7.76760101e-01 -7.41433740e-01 -8.06258202e-01 -1.64014429e-01
-5.36993921e-01 -6.82409763e-01 -4.26428735e-01 -3.43451083e-01
-6.59550965e-01 -1.02400923e+00 -1.02246130e+00 -7.95314372e-01
7.88445532e-01 5.30063689e-01 1.14200413e+00 1.17135115e-01
-1.88168630e-01 3.60501170e-01 -2.36405060e-01 -7.38526806e-02
-4.64899570e-01 -6.78888977e-01 -1.08098812e-01 -8.99198349e-04
-4.23028350e-01 -4.63333458e-01 -8.17305863e-01 6.94253445e-01
-8.79092574e-01 -1.53351679e-01 4.20369983e-01 4.27034587e-01
5.65938234e-01 7.08048820e-01 7.45019391e-02 2.93595940e-01
5.28944016e-01 1.59406707e-01 -7.33528554e-01 3.41517299e-01
-1.01272440e+00 -3.19419503e-01 9.51214507e-02 1.99454874e-01
-1.25069714e+00 -4.50010806e-01 -1.13181949e-01 3.13950211e-01
-1.37270108e-01 1.17351934e-01 -5.71195483e-01 -4.62574124e-01
8.44043374e-01 5.25119364e-01 -1.49380594e-01 -2.53209293e-01
-1.14221796e-02 4.54022080e-01 1.26744974e+00 3.30160260e-02
8.73423755e-01 2.99222440e-01 3.26893568e-01 -9.08046782e-01
-3.02149534e-01 -5.44943273e-01 -7.44119108e-01 -8.48995388e-01
6.41376734e-01 -7.70601511e-01 -3.97952110e-01 4.55692977e-01
-1.08391535e+00 1.60091475e-01 -1.92604333e-01 5.74024677e-01
-7.37507284e-01 6.23820007e-01 1.78819045e-01 -5.97061753e-01
-2.01581016e-01 -1.18129277e+00 5.90064049e-01 3.03214133e-01
2.45403826e-01 -8.67079735e-01 1.66493163e-01 2.97751397e-01
5.75678408e-01 1.86938554e-01 3.20710361e-01 2.34475374e-01
-7.15233386e-01 -2.97698051e-01 -4.35041904e-01 7.49281824e-01
1.44235402e-01 -6.74397349e-02 -7.96112478e-01 -4.41366911e-01
8.60821605e-01 3.23010415e-01 3.25014889e-01 6.93966746e-01
5.58816791e-01 -1.36458322e-01 2.72384167e-01 1.00530028e+00
2.03766179e+00 6.70086682e-01 1.23082900e+00 2.39759579e-01
1.80610139e-02 3.92742544e-01 7.26318538e-01 8.86405632e-03
-1.40363038e-01 1.02721560e+00 5.71135819e-01 -2.60768205e-01
-7.29939699e-01 1.84640780e-01 1.69599921e-01 5.74351251e-01
-2.99796939e-01 -6.90499723e-01 -7.28805900e-01 4.05940384e-01
-9.84987140e-01 -1.35241306e+00 -2.41497427e-01 2.15347004e+00
-3.17707807e-02 7.15598837e-02 -3.22000742e-01 6.85981154e-01
5.52565038e-01 2.42209807e-02 -2.23080769e-01 -4.06807542e-01
-4.67386961e-01 -1.63498223e-01 3.63997638e-01 9.96688187e-01
-7.43086755e-01 -1.78272072e-02 8.64081860e+00 7.88017809e-01
-1.24394560e+00 -1.07786350e-01 5.29035926e-01 2.48303283e-02
-1.82077020e-01 -1.40480638e-01 -1.78280085e-01 3.40993553e-01
9.51778352e-01 -8.46832633e-01 3.25329989e-01 3.25041831e-01
5.33705592e-01 -7.13820815e-01 -5.71481824e-01 1.16707611e+00
6.58559144e-01 -1.01228905e+00 2.44529620e-01 3.55078489e-01
9.57481921e-01 -2.55835593e-01 3.55011374e-01 -4.43568110e-01
-4.57004666e-01 -1.12561560e+00 6.21269464e-01 9.59990799e-01
1.02464950e+00 -9.44890022e-01 1.08429790e+00 1.86249092e-01
-5.59228659e-01 2.62096286e-01 -2.63696253e-01 -1.36266202e-01
2.96750337e-01 4.63089198e-01 -5.37685931e-01 1.04900694e+00
3.36007744e-01 3.58236790e-01 -7.73488939e-01 1.60197806e+00
-3.07728112e-01 1.61013380e-01 2.80829489e-01 5.73784888e-01
-6.88276365e-02 -3.48268420e-01 1.01713526e+00 8.65598202e-01
7.65652955e-01 2.27761805e-01 -3.23636681e-01 5.19135058e-01
5.25257468e-01 1.28510460e-01 -6.99864209e-01 6.06348813e-01
-9.43042561e-02 8.31530333e-01 -3.77860755e-01 -1.38429910e-01
-1.42804384e-01 1.17560649e+00 -7.16789901e-01 4.40105855e-01
-5.56614041e-01 -6.77470922e-01 2.67588794e-01 1.78323790e-01
1.35884374e-01 -7.79397115e-02 -5.67619324e-01 -1.02287626e+00
3.63635980e-02 -8.01030457e-01 4.54281159e-02 -1.47078586e+00
-5.34324050e-01 6.61109865e-01 1.97812229e-01 -1.65327585e+00
-5.40709436e-01 -4.19422239e-01 -6.06045783e-01 1.29780650e+00
-1.30032837e+00 -7.41514325e-01 -5.91226220e-01 4.64809626e-01
4.30013955e-01 -5.83061755e-01 6.97905123e-01 2.50865608e-01
3.88145261e-02 4.96533334e-01 2.22062260e-01 -2.28823870e-01
5.02602577e-01 -1.21736133e+00 -4.60100137e-02 1.57356179e+00
-7.59219751e-02 -2.72177961e-02 1.44812179e+00 -2.39006847e-01
-1.05220163e+00 -4.79598075e-01 9.34827685e-01 -3.85240644e-01
-4.75739762e-02 5.24789751e-01 -4.46903825e-01 -7.65372515e-02
4.72752035e-01 -1.60377696e-01 1.44485787e-01 -4.82345253e-01
2.86777794e-01 -5.25180161e-01 -1.56761885e+00 2.91734517e-01
8.53229940e-01 -3.42416734e-01 -5.72188795e-01 -3.50342304e-01
5.03482044e-01 -4.53801990e-01 -8.11464846e-01 5.70435107e-01
4.73080367e-01 -1.66223502e+00 1.07329345e+00 6.09064758e-01
4.38115597e-01 -7.01767206e-01 -4.10763472e-01 -1.45279384e+00
-4.65358466e-01 -4.73558515e-01 3.66044313e-01 9.68714416e-01
2.64615208e-01 -2.48633623e-01 2.58010328e-01 8.70347545e-02
-2.39555314e-01 -9.40047354e-02 -1.04097438e+00 -8.01925063e-01
-3.91835690e-01 -4.98823971e-01 3.01024109e-01 3.82661045e-01
-3.87621336e-02 9.11748707e-02 -4.93784338e-01 3.52137715e-01
1.23782218e+00 -2.84850419e-01 6.72332585e-01 -6.88256502e-01
9.50154960e-02 -4.36680436e-01 -1.18517590e+00 -6.23406768e-01
-6.34516299e-01 -5.00356257e-01 -6.04366288e-02 -1.71260679e+00
2.14825168e-01 3.81949604e-01 -1.27951458e-01 -6.47227466e-01
2.31032982e-01 6.66312993e-01 2.26778641e-01 -3.68201477e-03
-4.35530990e-01 1.28856421e-01 1.57054305e+00 -1.87314272e-01
1.44760653e-01 2.11204067e-01 -7.39662170e-01 5.43008745e-01
5.77340186e-01 -1.82818756e-01 -8.95001292e-01 -5.71031034e-01
2.49706984e-01 3.88891757e-01 4.73762333e-01 -1.58303261e+00
1.84257418e-01 -7.93925151e-02 4.43848461e-01 -8.80795896e-01
1.59772649e-01 -9.07763183e-01 5.97966313e-01 5.15791297e-01
-1.67056829e-01 8.66320506e-02 -1.81727424e-01 1.84778228e-01
-6.38663948e-01 -5.37320554e-01 1.23835802e+00 1.30245149e-01
-7.40397334e-01 -1.83841839e-01 -1.01220407e-01 -1.78696573e-01
1.14905965e+00 -6.43802881e-01 -3.91165495e-01 -7.04009056e-01
-7.49454260e-01 -2.13478550e-01 8.46798539e-01 2.35051975e-01
1.02833688e+00 -1.21859980e+00 -1.05052555e+00 2.47656778e-01
1.02511473e-01 -5.74331641e-01 3.52895766e-01 2.56302893e-01
-8.34931433e-01 5.65276802e-01 -8.51134241e-01 -5.90988994e-01
-1.71279788e+00 5.87013483e-01 8.43579531e-01 -6.78621158e-02
-6.40343547e-01 5.02056003e-01 1.88615009e-01 3.37982059e-01
3.67033601e-01 1.03094280e-01 -6.25993490e-01 -3.36369932e-01
1.27645802e+00 7.08266795e-01 2.25337163e-01 -9.41017985e-01
-1.97772570e-02 8.71632814e-01 7.74301469e-01 -6.81600571e-01
9.80231464e-01 -8.13778579e-01 -1.06984533e-01 2.97654569e-01
1.11171162e+00 1.37242243e-01 -9.72191632e-01 1.41384393e-01
-4.19148773e-01 -1.17135787e+00 4.70565379e-01 -1.34967995e+00
-8.68508399e-01 6.68502569e-01 1.47155166e+00 1.38644785e-01
1.72260451e+00 -3.43326628e-01 -7.72004128e-02 3.41951698e-02
4.73464787e-01 -8.97776306e-01 1.41618311e-01 1.63471326e-01
1.47972441e+00 -1.05088449e+00 3.27808946e-01 -3.32186073e-01
-4.38718379e-01 1.13177443e+00 2.49926582e-01 3.00881863e-02
5.20990312e-01 6.39572814e-02 2.61958182e-01 -2.41262689e-01
-4.43353772e-01 -3.20090711e-01 7.71423399e-01 1.41212726e+00
4.03959960e-01 -2.99597323e-01 -6.83286428e-01 -2.17194825e-01
-2.25155979e-01 -1.98863354e-03 1.04777634e+00 1.61011949e-01
-6.54321909e-01 -7.49687731e-01 -6.59501553e-01 -6.09814003e-02
-6.65347397e-01 1.46480769e-01 3.36090364e-02 2.85071552e-01
5.05876243e-01 1.30146599e+00 -6.11694530e-02 -2.66980916e-01
7.25684881e-01 -5.52473009e-01 6.34707272e-01 -1.67012170e-01
-8.23998004e-02 -9.33190212e-02 -3.05850781e-03 -6.29265904e-01
-7.04189539e-01 -4.17651743e-01 -6.42751217e-01 -3.57751310e-01
-6.53408691e-02 1.55024305e-01 1.01784456e+00 4.25409228e-01
-2.15492159e-01 6.82135701e-01 9.74837065e-01 -1.03427625e+00
-2.34997988e-01 -8.49833071e-01 -5.48304856e-01 2.99537122e-01
6.24669611e-01 -1.72615647e-01 -5.91522336e-01 4.13749725e-01] | [11.737582206726074, -1.983135461807251] |
574b4c9f-6ecd-4ce3-9e02-bfe920d2b1c9 | bridging-textual-and-tabular-data-for-cross | 2012.12627 | null | https://arxiv.org/abs/2012.12627v2 | https://arxiv.org/pdf/2012.12627v2.pdf | Bridging Textual and Tabular Data for Cross-Domain Text-to-SQL Semantic Parsing | We present BRIDGE, a powerful sequential architecture for modeling dependencies between natural language questions and relational databases in cross-DB semantic parsing. BRIDGE represents the question and DB schema in a tagged sequence where a subset of the fields are augmented with cell values mentioned in the question. The hybrid sequence is encoded by BERT with minimal subsequent layers and the text-DB contextualization is realized via the fine-tuned deep attention in BERT. Combined with a pointer-generator decoder with schema-consistency driven search space pruning, BRIDGE attained state-of-the-art performance on popular cross-DB text-to-SQL benchmarks, Spider (71.1\% dev, 67.5\% test with ensemble model) and WikiSQL (92.6\% dev, 91.9\% test). Our analysis shows that BRIDGE effectively captures the desired cross-modal dependencies and has the potential to generalize to more text-DB related tasks. Our implementation is available at \url{https://github.com/salesforce/TabularSemanticParsing}. | ['Caiming Xiong', 'Richard Socher', 'Xi Victoria Lin'] | 2020-12-23 | null | https://aclanthology.org/2020.findings-emnlp.438 | https://aclanthology.org/2020.findings-emnlp.438.pdf | findings-of-the-association-for-computational | ['deep-attention', 'deep-attention'] | ['computer-vision', 'natural-language-processing'] | [-1.02998704e-01 5.18009424e-01 -2.78546989e-01 -7.05227077e-01
-1.38907003e+00 -7.90078461e-01 3.91782939e-01 5.12536228e-01
-3.18802804e-01 6.30666375e-01 2.38760069e-01 -5.32960176e-01
-2.47595191e-01 -9.46771979e-01 -1.18824828e+00 5.80253303e-02
6.59684688e-02 1.01970410e+00 4.95274842e-01 -4.84025180e-01
2.21815798e-02 6.35717660e-02 -1.47495699e+00 1.04871523e+00
8.05163741e-01 1.27805531e+00 1.17660083e-01 5.83120823e-01
-8.25372875e-01 1.14508832e+00 -6.07822597e-01 -1.06036234e+00
2.19386984e-02 1.37299910e-01 -1.25044632e+00 -5.61129868e-01
7.21072912e-01 2.94112843e-02 -5.74273132e-02 7.93138921e-01
3.44352663e-01 -2.22364053e-01 -5.37568815e-02 -1.02046549e+00
-5.23479521e-01 1.21387053e+00 -3.84884387e-01 2.37437040e-01
4.79257762e-01 -1.46453232e-01 1.69741738e+00 -7.36373484e-01
9.83717203e-01 1.32847631e+00 6.21780097e-01 4.88777786e-01
-1.44517839e+00 -4.43715423e-01 1.04611613e-01 2.09853306e-01
-1.03432167e+00 -4.29862410e-01 4.16083336e-01 -2.59161413e-01
1.60002553e+00 5.78258932e-01 1.21620730e-01 8.96399140e-01
-1.26317353e-03 8.16363573e-01 8.52488458e-01 -3.57813001e-01
-6.74191192e-02 2.89022595e-01 6.72730327e-01 7.62593746e-01
5.26278988e-02 -1.96261555e-01 -6.16972446e-01 -2.39473939e-01
3.86562455e-03 -4.90208805e-01 2.00501829e-01 -3.09055507e-01
-7.57165849e-01 7.78933465e-01 3.14885706e-01 5.19165397e-02
-2.05427080e-01 2.85809308e-01 6.85018241e-01 3.53361070e-01
1.59368455e-01 5.63392520e-01 -9.61146772e-01 -1.93311915e-01
-4.36538696e-01 5.72515965e-01 1.30090475e+00 1.52500010e+00
6.94058657e-01 -4.27726716e-01 -1.23371579e-01 1.13819408e+00
1.05099179e-01 4.35646087e-01 1.26718748e-02 -1.13166118e+00
1.00829625e+00 8.73928428e-01 -1.19575910e-01 -4.03870612e-01
-4.20011610e-01 -4.84029144e-01 -1.93435237e-01 -3.38749588e-01
4.25400347e-01 1.06604487e-01 -6.87624156e-01 1.86604762e+00
3.69185269e-01 -5.87598860e-01 2.00722635e-01 3.81812721e-01
1.17245317e+00 5.21921933e-01 4.88551348e-01 1.22762226e-01
1.87594712e+00 -7.11984158e-01 -5.06061614e-01 -5.14037728e-01
8.30787599e-01 -6.86475456e-01 1.37476492e+00 2.99884707e-01
-1.58216321e+00 -5.06002724e-01 -8.63754749e-01 -6.86539769e-01
-6.48514569e-01 -3.52088064e-01 5.22646725e-01 4.25735027e-01
-8.37615967e-01 1.65319517e-01 -6.96622014e-01 -5.39336026e-01
3.53241146e-01 2.08083212e-01 -2.97004431e-01 -1.81645378e-01
-1.28564227e+00 7.70666420e-01 5.53849399e-01 -2.81929076e-01
-5.40527880e-01 -1.05795467e+00 -8.40236306e-01 1.48688272e-01
8.75589550e-01 -9.12517726e-01 1.48350978e+00 -2.33714432e-01
-8.40658307e-01 1.25119972e+00 -2.54854053e-01 -6.49528682e-01
2.76750445e-01 -4.77802932e-01 -5.53952515e-01 -1.09983332e-01
3.65456372e-01 7.39748776e-01 7.55578373e-03 -1.08296371e+00
-8.09116304e-01 -7.69916594e-01 3.35271180e-01 -1.65429320e-02
1.75902724e-01 2.34868973e-01 -1.09172237e+00 -4.08918321e-01
5.61731122e-02 -5.92696905e-01 1.29054546e-01 -4.80232298e-01
-6.46827102e-01 -4.01152432e-01 3.17829758e-01 -1.01626241e+00
1.50282061e+00 -1.94060373e+00 5.90667017e-02 4.51359600e-02
1.75403222e-01 1.32060554e-02 -6.18626736e-02 8.35967183e-01
-3.38328518e-02 2.83549577e-01 -1.92502707e-01 -1.15068182e-01
3.45528364e-01 3.30027670e-01 -4.37984169e-01 -3.34253818e-01
2.28307977e-01 1.10034740e+00 -4.41076040e-01 -4.99485582e-01
-2.46986911e-01 7.91094080e-02 -1.01767290e+00 3.64575237e-01
-1.15156627e+00 -2.09966287e-01 -4.09874290e-01 6.94329977e-01
5.73570371e-01 -3.86681557e-01 5.89631140e-01 -3.98711145e-01
1.49824426e-01 1.06618631e+00 -1.02156544e+00 1.98899412e+00
-5.55727720e-01 1.56202301e-01 3.08631510e-01 -7.54385889e-01
9.37171459e-01 -1.30636409e-01 2.91440845e-01 -1.22057605e+00
-3.51648301e-01 3.08966070e-01 -2.40454480e-01 -7.73834765e-01
5.79497516e-01 2.35965759e-01 -5.75307250e-01 2.33034939e-01
2.04293281e-01 -1.51809096e-01 4.88051862e-01 5.03927052e-01
1.21294403e+00 2.73636281e-01 4.38168012e-02 -4.04920727e-01
4.11978364e-01 4.55098450e-01 5.92001438e-01 8.20154309e-01
5.32407820e-01 1.12235196e-01 1.06064153e+00 -4.78421092e-01
-9.31052506e-01 -1.17997706e+00 -2.81959802e-01 1.56871045e+00
-1.51866928e-01 -8.72384727e-01 -8.51325929e-01 -9.19348419e-01
3.67314696e-01 1.28739429e+00 -5.79317033e-01 2.12230720e-02
-9.49760318e-01 -5.22352695e-01 6.22470021e-01 7.22543478e-01
3.52850497e-01 -1.20092356e+00 -3.64474714e-01 3.69787008e-01
-3.27712774e-01 -1.26605594e+00 -2.26343215e-01 5.96713960e-01
-6.87076688e-01 -1.20352733e+00 3.51366311e-01 -7.09413469e-01
-1.06316423e-02 -3.25223655e-01 1.94727731e+00 -7.66027570e-02
-2.84999132e-01 1.96704298e-01 -2.55977064e-01 -3.54282886e-01
-4.72939968e-01 4.68368739e-01 -9.10223305e-01 -6.20487452e-01
8.37642670e-01 -3.61772209e-01 -4.86833066e-01 2.67116487e-01
-9.68742490e-01 1.37471231e-02 2.11351216e-01 8.45384955e-01
8.38243604e-01 -4.38571185e-01 4.69762743e-01 -1.47516942e+00
3.45616519e-01 -7.43152380e-01 -9.07938242e-01 6.00787103e-01
-6.46045864e-01 3.91038954e-01 3.97317737e-01 2.02612698e-01
-1.03418136e+00 -3.95237148e-01 -5.51908553e-01 2.49639958e-01
-1.62136629e-01 6.95634663e-01 -6.40036285e-01 7.79390514e-01
7.50825882e-01 -1.71007112e-01 -1.36506349e-01 -8.73039544e-01
6.63302362e-01 4.51378256e-01 6.97295427e-01 -1.12757230e+00
1.73929840e-01 2.20222831e-01 -3.70826006e-01 -1.28362805e-01
-1.02456367e+00 -2.44639248e-01 -2.78962940e-01 4.39124882e-01
1.04901004e+00 -7.78873265e-01 -9.58841980e-01 2.23811455e-02
-1.17199874e+00 -4.80480075e-01 -5.12427151e-01 -1.42181724e-01
-5.39479554e-01 -1.39213920e-01 -7.46811390e-01 -3.77030969e-01
-5.67182183e-01 -9.92429435e-01 1.11056387e+00 -7.02761337e-02
-3.55867833e-01 -6.88489914e-01 -2.13325500e-01 9.21167970e-01
2.03215569e-01 -1.27473116e-01 1.63715076e+00 -1.12764442e+00
-7.14669228e-01 2.09598038e-02 -3.83480132e-01 2.55651414e-01
-4.45313662e-01 -2.15349406e-01 -7.96591222e-01 4.94078025e-02
-2.55896389e-01 -4.88349378e-01 7.26729631e-01 9.06058867e-03
1.45409012e+00 -2.83069044e-01 -3.23477179e-01 7.27961719e-01
1.55937779e+00 4.83152121e-01 5.09109974e-01 5.39600909e-01
4.92885470e-01 8.58540773e-01 5.09725928e-01 3.72093648e-01
9.63329554e-01 7.10676610e-01 6.36848330e-01 2.55827844e-01
-1.91492438e-01 -4.81918007e-01 -5.80768660e-03 5.06529212e-01
7.30437875e-01 -3.92404288e-01 -1.38799131e+00 5.08688569e-01
-1.63582432e+00 -7.43574858e-01 -2.22406477e-01 1.96438539e+00
1.05472159e+00 4.29682374e-01 1.12049803e-01 -3.85249287e-01
4.73268151e-01 4.06979509e-02 -7.11525083e-01 -6.31306708e-01
-3.86300355e-01 5.13137758e-01 4.09907371e-01 6.52755797e-01
-7.83007383e-01 1.17475748e+00 5.16450644e+00 6.51612878e-01
-4.57095295e-01 2.76730567e-01 7.44734168e-01 -3.34806442e-01
-7.34346211e-01 1.66247680e-03 -1.36289728e+00 3.63737851e-01
1.23502672e+00 -1.04846340e-02 4.21354830e-01 7.62596011e-01
-5.07293999e-01 -3.75969969e-02 -1.40094960e+00 5.84757745e-01
-2.51347423e-01 -1.63160849e+00 -1.27161115e-01 -1.93037853e-01
1.93159074e-01 2.85595059e-01 -9.27506462e-02 6.85449421e-01
7.12481916e-01 -9.17290807e-01 8.57505858e-01 4.21566099e-01
8.98616612e-01 -6.30032361e-01 5.76079488e-01 1.49680436e-01
-9.51428175e-01 -2.76586294e-01 -1.76829606e-01 6.73533440e-01
1.38454422e-01 3.68861139e-01 -7.06927478e-01 7.10819364e-01
1.11569273e+00 2.93320447e-01 -6.06268108e-01 3.20384026e-01
3.25912125e-02 4.53038096e-01 -4.18900251e-01 1.39923664e-02
1.36487037e-01 4.68705269e-03 3.98077488e-01 1.46977687e+00
1.12749346e-01 6.27064854e-02 -5.67670390e-02 9.46217597e-01
-2.94947356e-01 1.38193116e-01 -2.49366328e-01 -5.55213355e-02
7.17368126e-01 8.35269511e-01 1.09668830e-02 -4.85740423e-01
-6.08135819e-01 4.57636893e-01 7.46894658e-01 2.70137668e-01
-9.27154243e-01 -4.03930157e-01 8.02726924e-01 4.40253198e-01
5.88587105e-01 2.23330393e-01 -5.64233959e-01 -9.08945978e-01
4.79482472e-01 -1.26691973e+00 1.20265603e+00 -9.14115846e-01
-1.25181901e+00 8.46857369e-01 1.61163270e-01 -2.86494374e-01
-3.51791739e-01 -5.78054845e-01 -9.16210562e-03 9.62369859e-01
-1.19465065e+00 -1.13317430e+00 -2.29062214e-01 4.79819924e-01
3.99925560e-01 -1.14217706e-01 8.69908750e-01 5.34086585e-01
-6.12225592e-01 8.00138116e-01 -1.18322678e-01 1.64325595e-01
5.52950859e-01 -1.43300688e+00 9.05743182e-01 3.93061489e-01
2.22810395e-02 8.26176226e-01 5.23404896e-01 -6.28493905e-01
-1.63493049e+00 -9.56094205e-01 1.28850520e+00 -6.34639919e-01
8.55833232e-01 -7.85266817e-01 -1.17455208e+00 1.02649248e+00
4.57318097e-01 -2.98335478e-02 6.24719977e-01 4.99136746e-01
-1.00326467e+00 -6.23011172e-01 -1.16558623e+00 4.24205661e-01
1.29214406e+00 -6.55083835e-01 -8.11460435e-01 3.99272412e-01
1.12007439e+00 -6.96084738e-01 -1.25362718e+00 4.10265326e-01
4.28109556e-01 -1.03465641e+00 9.65505123e-01 -1.15324783e+00
5.08267164e-01 1.85471326e-01 -6.39558017e-01 -6.08665943e-01
-8.23034495e-02 -4.71003860e-01 -3.34181756e-01 1.55852878e+00
8.97634387e-01 -6.61588609e-01 8.95166636e-01 9.20613050e-01
-3.35453361e-01 -9.03213859e-01 -8.06802571e-01 -5.62164783e-01
2.25829959e-01 -7.92992353e-01 9.74257648e-01 6.81625128e-01
-9.17501226e-02 4.73757803e-01 2.62650371e-01 1.37675673e-01
5.33399403e-01 3.95119518e-01 6.24767125e-01 -1.01896501e+00
-4.96157944e-01 -4.18277025e-01 1.57175034e-01 -1.21278262e+00
7.89419934e-02 -1.10969198e+00 -3.84211272e-01 -1.54638958e+00
7.17382655e-02 -7.27538764e-01 -2.47907475e-01 4.83999819e-01
-2.96050049e-02 -4.08921093e-01 2.08294652e-02 -2.01890171e-01
-6.26460493e-01 1.48888305e-01 6.83000565e-01 -1.21927805e-01
8.21766630e-02 -2.50631899e-01 -9.41419303e-01 1.76059887e-01
9.14748371e-01 -6.52903199e-01 -3.48617494e-01 -8.80981684e-01
6.45948827e-01 5.38018942e-01 2.49529779e-01 -5.76812327e-01
1.78097129e-01 -1.93750784e-02 -1.12336174e-01 -7.86026239e-01
3.97348076e-01 -6.07598960e-01 2.91107357e-01 2.37337708e-01
-7.90694475e-01 5.43108523e-01 5.39081335e-01 2.43345395e-01
-3.55060041e-01 -2.40757748e-01 5.94083250e-01 -3.20352197e-01
-6.17755949e-01 1.39689133e-01 2.84227163e-01 9.29372072e-01
7.00152159e-01 1.84955597e-01 -9.26402271e-01 1.38203008e-02
-7.80915141e-01 4.86558139e-01 3.01336236e-02 5.21083713e-01
9.23387706e-02 -9.28521693e-01 -6.36704981e-01 2.57201999e-01
3.48692238e-01 1.98850602e-01 2.21065894e-01 5.51944673e-01
-6.09346867e-01 7.78200448e-01 -2.10812651e-02 -4.84247863e-01
-1.04601657e+00 6.32750571e-01 3.39858025e-01 -5.91595232e-01
-4.01674777e-01 1.23036718e+00 1.58536658e-01 -7.74405003e-01
3.66112560e-01 -3.49451065e-01 2.20663533e-01 6.35781512e-02
1.75689593e-01 2.21530601e-01 5.35585999e-01 -8.74183550e-02
-5.39173067e-01 1.67055860e-01 -4.49768215e-01 -1.50977477e-01
1.48469377e+00 5.37029244e-02 -5.28904796e-01 2.38882676e-01
1.27831793e+00 1.39055282e-01 -8.22123885e-01 -5.19738734e-01
6.47388041e-01 -8.70848447e-02 -3.71684194e-01 -1.53323805e+00
-1.02356231e+00 7.44567573e-01 1.64532855e-01 3.40218902e-01
9.93630111e-01 5.71769893e-01 8.45076025e-01 4.68469858e-01
2.64929175e-01 -9.31429505e-01 -2.05508307e-01 6.01217389e-01
1.15978670e+00 -1.12364948e+00 -3.37316900e-01 -5.43617368e-01
-6.53358817e-01 7.60297239e-01 1.02890456e+00 3.00501108e-01
5.74297905e-01 6.81769311e-01 4.95303981e-02 -5.80154598e-01
-1.32064033e+00 -1.33718863e-01 1.16894998e-01 3.16485018e-01
6.42987847e-01 -1.60560489e-01 -2.22886086e-01 9.33203936e-01
-4.20179933e-01 -3.84641051e-01 -9.11267698e-02 9.31839049e-01
-1.85107246e-01 -1.43069279e+00 1.06317550e-01 5.64255059e-01
-7.37275839e-01 -5.36626220e-01 -1.94414079e-01 1.10678446e+00
1.76795502e-03 7.82572508e-01 3.58693004e-01 -2.05060020e-01
7.99695075e-01 5.67859054e-01 3.34078670e-01 -7.01644063e-01
-1.16819191e+00 -3.85062806e-02 9.53281462e-01 -1.01268208e+00
3.61089081e-01 -7.31781900e-01 -1.40549242e+00 -4.26529646e-01
1.57880470e-01 3.38087887e-01 7.33722508e-01 3.47916216e-01
8.98139834e-01 4.13037062e-01 -3.73473987e-02 5.52924395e-01
-6.69825554e-01 -9.08018172e-01 -1.54504374e-01 5.91987669e-01
-4.39452976e-02 -2.43583292e-01 7.70470574e-02 -9.64929014e-02] | [9.880845069885254, 7.855257511138916] |
1d6eaa8c-11ec-40d6-a201-030ad48bc420 | ground-roll-suppression-using-convolutional | 2010.15209 | null | https://arxiv.org/abs/2010.15209v1 | https://arxiv.org/pdf/2010.15209v1.pdf | Ground Roll Suppression using Convolutional Neural Networks | Seismic data processing plays a major role in seismic exploration as it conditions much of the seismic interpretation performance. In this context, generating reliable post-stack seismic data depends also on disposing of an efficient pre-stack noise attenuation tool. Here we tackle ground roll noise, one of the most challenging and common noises observed in pre-stack seismic data. Since ground roll is characterized by relative low frequencies and high amplitudes, most commonly used approaches for its suppression are based on frequency-amplitude filters for ground roll characteristic bands. However, when signal and noise share the same frequency ranges, these methods usually deliver also signal suppression or residual noise. In this paper we take advantage of the highly non-linear features of convolutional neural networks, and propose to use different architectures to detect ground roll in shot gathers and ultimately to suppress them using conditional generative adversarial networks. Additionally, we propose metrics to evaluate ground roll suppression, and report strong results compared to expert filtering. Finally, we discuss generalization of trained models for similar and different geologies to better understand the feasibility of our proposal in real applications. | ['Semen Zaytsev', 'Daniil Semin', 'Dario Augusto Borges Oliveira'] | 2020-10-28 | null | null | null | null | ['seismic-interpretation'] | ['miscellaneous'] | [ 2.24243507e-01 2.01777369e-02 5.64333975e-01 -6.69495314e-02
-1.00711036e+00 -4.96276170e-01 5.46381831e-01 4.24580544e-01
-6.14173651e-01 6.47801697e-01 5.02872169e-01 -1.01034351e-01
-3.56579840e-01 -1.19202471e+00 -7.81941831e-01 -9.14463162e-01
-4.13824528e-01 1.87589571e-01 6.83149397e-01 -6.76084042e-01
1.47390798e-01 7.77729332e-01 -1.49812734e+00 3.94111156e-01
5.80009520e-01 8.48728299e-01 7.01138228e-02 8.73211443e-01
1.24577411e-01 8.21824431e-01 -8.96112442e-01 -8.73605236e-02
1.84766829e-01 -3.16369772e-01 -6.05775476e-01 -2.64662027e-01
2.14042827e-01 -2.42030218e-01 -2.62094259e-01 1.03899741e+00
8.00375044e-01 4.80472445e-01 7.34841645e-01 -4.80997562e-01
1.27807677e-01 9.44094777e-01 -3.01580995e-01 4.85364795e-01
4.86279786e-01 1.77387133e-01 6.01043761e-01 -7.27457583e-01
4.31196302e-01 8.94973874e-01 1.14344847e+00 2.04011694e-01
-1.28320479e+00 -3.64725530e-01 -1.39285535e-01 1.81123465e-01
-1.08470249e+00 -4.62409139e-01 1.05600905e+00 -5.60807586e-01
6.11063123e-01 3.17260146e-01 4.84567285e-01 1.20712578e+00
-6.81650732e-03 5.34988880e-01 8.54492605e-01 -4.17838842e-01
5.37933171e-01 -4.15746987e-01 3.16198799e-04 -5.09607345e-02
2.11857647e-01 -1.47086605e-01 -6.08061135e-01 -1.08200714e-01
3.47674847e-01 -2.85770863e-01 -6.69638097e-01 4.43280429e-01
-7.36966431e-01 6.76680624e-01 4.19187129e-01 6.63907468e-01
-5.79200029e-01 4.98076558e-01 4.18315291e-01 1.31689936e-01
5.55994093e-01 6.33873641e-01 -8.32168572e-03 -1.05747528e-01
-1.27878284e+00 6.11383379e-01 7.08962739e-01 3.11879665e-01
7.72727609e-01 3.72148067e-01 -1.03679948e-01 5.01376569e-01
1.52315915e-01 6.14062965e-01 4.29794759e-01 -5.09889543e-01
4.50098991e-01 6.12800941e-02 -3.76101322e-02 -1.30275512e+00
-7.49754608e-01 -7.07710505e-01 -9.94300723e-01 2.93966740e-01
4.82612401e-01 -2.16168493e-01 -9.44386661e-01 1.43244052e+00
-4.48329821e-02 2.63534516e-01 2.05873817e-01 8.35388303e-01
6.15045726e-01 5.97842216e-01 1.67472094e-01 4.73140851e-02
1.22868109e+00 -1.90123007e-01 -7.40264893e-01 -4.08126056e-01
2.26461709e-01 -6.73231542e-01 8.91242564e-01 4.89910305e-01
-1.01658893e+00 -4.20111835e-01 -1.15793085e+00 2.02611402e-01
-3.20477277e-01 -1.95449606e-01 2.88149148e-01 7.69961059e-01
-7.90237129e-01 1.13586569e+00 -1.15364158e+00 1.62635833e-01
2.19388485e-01 6.37823120e-02 -1.49025679e-01 2.63092816e-01
-1.46436489e+00 7.69102871e-01 2.52321422e-01 4.91777390e-01
-8.17586482e-01 -8.24025989e-01 -6.01567328e-01 1.93065628e-01
3.75000060e-01 -6.60534799e-02 1.04933858e+00 -3.82908702e-01
-1.19563591e+00 4.36137706e-01 3.84672463e-01 -9.32273507e-01
8.90750825e-01 -5.98121643e-01 -5.51533580e-01 3.52089286e-01
-7.16268132e-03 -2.27355093e-01 9.33406770e-01 -1.15413225e+00
-3.30235332e-01 2.56929602e-02 -1.32030576e-01 -3.62319171e-01
-3.64922613e-01 -9.61335897e-02 3.96548361e-02 -9.81090486e-01
4.72504705e-01 -4.71859455e-01 -4.32385653e-01 -5.42964637e-01
-4.36122924e-01 3.80760312e-01 6.37421846e-01 -9.65976834e-01
1.30090153e+00 -2.08506918e+00 -5.55376783e-02 6.03253424e-01
1.22291259e-01 3.30535650e-01 2.66510218e-01 6.05526686e-01
-1.84000343e-01 8.41454268e-02 -7.39174128e-01 -1.82256162e-01
-1.99679390e-01 -1.21458687e-01 -6.72805786e-01 5.11396050e-01
4.48914945e-01 4.33473080e-01 -6.77844763e-01 3.10380217e-02
3.01730275e-01 5.75744987e-01 -7.16023088e-01 1.02890521e-01
-1.43371090e-01 4.53664362e-01 -4.99136329e-01 2.97640681e-01
7.88945198e-01 4.52594280e-01 -3.00099105e-01 -5.19348860e-01
-2.62805045e-01 1.29204497e-01 -1.36183763e+00 1.24633896e+00
-3.04401070e-01 5.93025446e-01 1.73459068e-01 -1.17462993e+00
1.12137938e+00 2.09466517e-01 2.93372333e-01 -6.89028382e-01
7.72688463e-02 6.76866889e-01 -1.23782516e-01 -7.96520114e-01
5.98982155e-01 -4.13264692e-01 -6.82420358e-02 8.56072083e-02
-6.36330321e-02 -5.90526640e-01 2.34189048e-01 -8.75035971e-02
1.38963127e+00 -3.06488741e-02 -4.65925075e-02 -3.91295254e-01
5.07915735e-01 -3.40712249e-01 1.80736870e-01 9.31326210e-01
2.62671202e-01 1.17397690e+00 6.00980580e-01 -3.86046797e-01
-9.03873563e-01 -8.25198114e-01 9.69077498e-02 7.39693105e-01
-2.06153736e-01 -2.32373197e-02 -7.05943406e-01 -2.12703839e-01
-2.00974390e-01 7.44509220e-01 -6.61283374e-01 -3.42601925e-01
-1.20105147e+00 -1.08773208e+00 1.01919377e+00 4.98096317e-01
2.78487593e-01 -1.09251153e+00 -9.73551333e-01 7.56809413e-01
-2.97680944e-01 -8.66193593e-01 1.21395223e-01 7.35182047e-01
-6.40881479e-01 -8.96155059e-01 -8.38769078e-01 -2.00787798e-01
3.05014034e-03 -1.18322179e-01 1.12554860e+00 1.26999795e-01
-1.83125034e-01 -2.01040525e-02 -7.30550230e-01 -4.47533071e-01
-6.15387440e-01 8.22683498e-02 -1.15693338e-01 1.87066942e-01
-1.24502622e-01 -9.99201059e-01 -6.21394455e-01 9.78355557e-02
-1.25501549e+00 -4.44893271e-01 3.87112260e-01 6.11068666e-01
1.79969773e-01 2.76881039e-01 5.26662469e-01 -6.65462852e-01
6.74652636e-01 -3.65230888e-01 -4.07943398e-01 -1.57100275e-01
2.14988396e-01 1.08122028e-01 8.29591036e-01 -3.21603268e-01
-9.64725435e-01 -1.69166952e-01 -8.46223772e-01 -5.28809801e-02
-2.84306794e-01 7.10734963e-01 -5.34364283e-02 -4.07731310e-02
1.10554671e+00 1.14403434e-01 -4.35421020e-01 -5.83209276e-01
-5.62071279e-02 3.01769286e-01 8.32839251e-01 -4.76593792e-01
9.99167919e-01 6.92451835e-01 1.81211770e-01 -1.35381746e+00
-6.16187811e-01 -4.01771545e-01 -1.75979644e-01 -3.78268540e-01
8.28549266e-01 -6.44487977e-01 -4.62721944e-01 6.41768694e-01
-1.05238152e+00 -2.72936344e-01 -5.42004883e-01 4.89180773e-01
-3.90243471e-01 4.35822189e-01 -6.32800877e-01 -1.21150541e+00
-3.08591157e-01 -1.09280956e+00 8.49631906e-01 1.15881585e-01
-1.43684387e-01 -7.91038752e-01 1.20494388e-01 -1.36853248e-01
6.86787724e-01 6.98766470e-01 5.19723177e-01 -7.18513489e-01
-2.92064190e-01 -3.89635116e-01 2.82057166e-01 3.09688210e-01
-1.32897258e-01 -1.00103626e-02 -1.33450425e+00 1.82902031e-02
4.34338689e-01 -6.17240779e-02 1.29711342e+00 6.44095600e-01
7.92439461e-01 -4.03247550e-02 6.39679730e-02 6.90841973e-01
1.40141165e+00 8.98605436e-02 9.96500552e-01 6.21045947e-01
4.95639354e-01 8.82358372e-01 2.95568168e-01 6.24453485e-01
-4.27882403e-01 4.79312927e-01 4.93273318e-01 -6.35921434e-02
-2.12908760e-01 -3.45282257e-02 1.43635809e-01 6.04895234e-01
-4.33973879e-01 -4.11871821e-01 -1.14337826e+00 6.58558726e-01
-1.52061939e+00 -1.14081645e+00 -6.08400941e-01 1.93923414e+00
6.29693747e-01 4.31750327e-01 -8.69417936e-03 7.95616090e-01
5.78937173e-01 2.80637383e-01 1.11923583e-01 1.58488795e-01
-6.07990384e-01 6.21386170e-01 8.51323128e-01 3.69745016e-01
-1.29698527e+00 4.22438174e-01 5.62201500e+00 9.04674411e-01
-1.17226684e+00 -1.05620304e-03 4.44556564e-01 2.76106834e-01
-5.49614906e-01 -2.87956506e-01 -4.72146958e-01 5.00420809e-01
8.37867796e-01 1.97879598e-01 -7.32484646e-03 5.77909112e-01
3.72562289e-01 -2.29047716e-01 -6.60306036e-01 5.08905470e-01
-1.89167634e-01 -1.40987360e+00 -3.17504466e-01 -2.76067793e-01
4.74138707e-01 -2.53549486e-01 -1.38201743e-01 4.52858806e-02
4.75794263e-03 -9.71107543e-01 1.27038634e+00 8.21653903e-01
2.66251236e-01 -8.80144298e-01 9.78715241e-01 1.65176526e-01
-1.22278714e+00 1.06827850e-02 -2.29120061e-01 4.94765788e-02
5.47256172e-01 1.00117075e+00 -6.08710051e-01 7.22152829e-01
7.45898306e-01 2.83802599e-01 -4.18869495e-01 1.25248373e+00
-3.63501847e-01 9.31422114e-01 -6.89395010e-01 1.43595725e-01
3.00608903e-01 8.18914324e-02 7.38849342e-01 1.53476882e+00
5.87142766e-01 9.81269591e-03 -6.17941245e-02 6.55870378e-01
3.20054591e-01 -5.99827580e-02 -3.99028599e-01 2.10428178e-01
2.24988475e-01 7.35542595e-01 -1.04232478e+00 -7.54898861e-02
-1.58264190e-01 4.53154683e-01 -2.32666641e-01 3.03351671e-01
-6.95216954e-01 -4.62413788e-01 3.02013993e-01 5.13785064e-01
2.74342418e-01 -1.21238112e-01 -3.70769978e-01 -7.45184362e-01
1.91830561e-01 -6.55378342e-01 1.87878951e-01 -4.13713008e-01
-1.03763914e+00 6.31554127e-01 5.68724424e-02 -1.45284879e+00
-2.62247264e-01 -3.76859039e-01 -8.39763045e-01 7.72546053e-01
-1.26574087e+00 -9.39104021e-01 -4.11150664e-01 4.01576191e-01
3.86484057e-01 1.13375820e-01 6.35521650e-01 6.49575472e-01
-8.09855685e-02 1.54433042e-01 -3.57031152e-02 1.61774203e-01
4.01479751e-01 -1.17242241e+00 6.45409405e-01 1.26519144e+00
-1.32667214e-01 1.60682186e-01 1.51255202e+00 -9.30589855e-01
-1.12800205e+00 -9.93837297e-01 4.01507765e-01 1.51129857e-01
9.38352525e-01 -1.92818269e-01 -1.52192223e+00 2.99234807e-01
-1.14609413e-01 -7.92795569e-02 1.89313427e-01 -2.96256602e-01
7.12716579e-02 -6.07144982e-02 -9.31590796e-01 5.29044867e-01
5.84286809e-01 -3.58038664e-01 -8.12678277e-01 6.45485967e-02
5.33674598e-01 -3.35090071e-01 -4.39411491e-01 5.31027079e-01
2.16503739e-01 -1.29045665e+00 1.13027692e+00 -1.20731018e-01
6.97590172e-01 -4.52751309e-01 -1.60307348e-01 -1.26964688e+00
-8.07799995e-02 -6.09870017e-01 2.18316764e-01 1.10417402e+00
2.69426674e-01 -4.13904846e-01 9.54866350e-01 1.29072979e-01
-5.43680787e-01 -1.61138132e-01 -1.08615506e+00 -5.95414937e-01
-4.06508595e-02 -8.40376616e-01 3.83918822e-01 5.61372519e-01
-4.63438898e-01 -2.61234611e-01 -5.12436867e-01 4.90619004e-01
6.66572809e-01 -3.38132143e-01 4.50680733e-01 -1.10482705e+00
-4.94901896e-01 -5.38934767e-01 -4.74784851e-01 -4.30376738e-01
-4.18689698e-01 -3.57738674e-01 5.86401224e-01 -1.19122148e+00
-5.74330091e-01 -1.81996897e-01 -1.29807845e-01 1.40154853e-01
-1.90963879e-01 6.48227632e-01 -3.11502945e-02 -3.70968990e-02
1.14895932e-01 5.23357093e-01 6.85741901e-01 -2.40393177e-01
-1.84409529e-01 2.42191091e-01 6.33354159e-03 1.09582150e+00
7.67290056e-01 -6.01691484e-01 -1.98743585e-02 -3.44512790e-01
6.65426195e-01 2.65440568e-02 4.56725150e-01 -1.65670800e+00
2.58685231e-01 8.60878304e-02 1.16741061e-01 -7.06331432e-01
2.73702741e-01 -7.98210919e-01 4.45121974e-01 6.85803771e-01
-1.52685598e-01 -2.24501386e-01 1.57269329e-01 4.23615843e-01
-5.00853181e-01 -6.04818404e-01 9.46354568e-01 -9.70016867e-02
-7.21500099e-01 -1.88347459e-01 -8.76934767e-01 -9.04926378e-03
6.05065584e-01 -1.46657825e-01 -1.92472916e-02 -3.96433175e-01
-9.64314580e-01 -2.20002562e-01 1.69507973e-02 -5.88116683e-02
6.53501689e-01 -7.42842495e-01 -8.78297150e-01 2.23216891e-01
-1.59665719e-01 5.14319018e-02 7.94450939e-01 8.75817835e-01
-1.13709497e+00 -3.25000167e-01 -5.78394122e-02 -4.31223035e-01
-8.27216387e-01 1.33749783e-01 4.36227769e-01 -3.27390432e-01
-8.96093607e-01 1.06805968e+00 -1.41667172e-01 2.04030484e-01
5.05811609e-02 -4.33392584e-01 -4.87814456e-01 4.95936394e-01
7.83279955e-01 6.46217585e-01 7.19803393e-01 -4.86708581e-01
-3.43403876e-01 5.15015662e-01 3.71867597e-01 -2.37137362e-01
1.73280096e+00 1.97924733e-01 1.11659043e-01 3.93217415e-01
8.60871255e-01 3.88908386e-01 -1.15294755e+00 1.27920732e-01
3.30271870e-01 -1.76468357e-01 5.32311350e-02 -2.37080559e-01
-1.07915318e+00 8.17908287e-01 5.24347961e-01 6.82811797e-01
1.27967942e+00 -1.31657183e-01 4.83440220e-01 4.01021093e-01
1.97582483e-01 -1.03367043e+00 1.03152946e-01 5.24354100e-01
1.11481941e+00 -6.44104242e-01 1.18305661e-01 -3.74740124e-01
-2.19786435e-01 1.39751267e+00 -6.78469939e-03 -6.21152818e-01
5.81878006e-01 8.54895055e-01 1.35655135e-01 -1.34234011e-01
-1.01911806e-01 -2.45018512e-01 -4.18320149e-02 5.42331338e-01
2.60464996e-01 -1.01491429e-01 -3.17625791e-01 7.50724733e-01
-4.20778722e-01 -3.16540629e-01 7.12510288e-01 1.04204786e+00
-6.76598489e-01 -7.69304395e-01 -8.82088125e-01 3.94826114e-01
-7.61251748e-01 -6.25033677e-02 1.40113935e-01 5.80016911e-01
5.31039014e-02 1.00069225e+00 -1.10361017e-01 -3.81058782e-01
7.70368099e-01 -7.18954802e-02 3.91574092e-02 -3.39129031e-01
-1.01546276e+00 5.28537452e-01 1.95736721e-01 -4.33762342e-01
-4.16653544e-01 -5.72825551e-01 -1.15881205e+00 -1.73604697e-01
-2.47165516e-01 1.48734480e-01 6.24038041e-01 8.06119978e-01
-4.27899867e-01 1.16660571e+00 3.46801817e-01 -1.30864131e+00
-6.07558310e-01 -1.27515066e+00 -7.27911234e-01 7.49095559e-01
5.44444323e-01 -6.41189098e-01 -7.71967769e-01 1.97033554e-01] | [6.949085235595703, 2.491682529449463] |
d7ffcb24-9865-45cf-a9a4-679957af177b | scrabblegan-semi-supervised-varying-length | 2003.10557 | null | https://arxiv.org/abs/2003.10557v1 | https://arxiv.org/pdf/2003.10557v1.pdf | ScrabbleGAN: Semi-Supervised Varying Length Handwritten Text Generation | Optical character recognition (OCR) systems performance have improved significantly in the deep learning era. This is especially true for handwritten text recognition (HTR), where each author has a unique style, unlike printed text, where the variation is smaller by design. That said, deep learning based HTR is limited, as in every other task, by the number of training examples. Gathering data is a challenging and costly task, and even more so, the labeling task that follows, of which we focus here. One possible approach to reduce the burden of data annotation is semi-supervised learning. Semi supervised methods use, in addition to labeled data, some unlabeled samples to improve performance, compared to fully supervised ones. Consequently, such methods may adapt to unseen images during test time. We present ScrabbleGAN, a semi-supervised approach to synthesize handwritten text images that are versatile both in style and lexicon. ScrabbleGAN relies on a novel generative model which can generate images of words with an arbitrary length. We show how to operate our approach in a semi-supervised manner, enjoying the aforementioned benefits such as performance boost over state of the art supervised HTR. Furthermore, our generator can manipulate the resulting text style. This allows us to change, for instance, whether the text is cursive, or how thin is the pen stroke. | ['Roee Litman', 'Hadar Averbuch-Elor', 'Sharon Fogel', 'Shai Mazor', 'Sarel Cohen'] | 2020-03-23 | scrabblegan-semi-supervised-varying-length-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Fogel_ScrabbleGAN_Semi-Supervised_Varying_Length_Handwritten_Text_Generation_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Fogel_ScrabbleGAN_Semi-Supervised_Varying_Length_Handwritten_Text_Generation_CVPR_2020_paper.pdf | cvpr-2020-6 | ['handwriting-generation'] | ['computer-vision'] | [ 4.77794617e-01 -1.25984356e-01 6.94405884e-02 -1.47510618e-01
-3.89524728e-01 -1.05520749e+00 5.80373466e-01 -2.09486604e-01
-4.46614355e-01 7.93297231e-01 -2.33852684e-01 -2.74208993e-01
1.52814209e-01 -8.86894166e-01 -8.36434722e-01 -7.46993005e-01
7.06516802e-01 7.65045643e-01 8.95417929e-02 -2.13240668e-01
3.51186633e-01 8.41452301e-01 -1.33427453e+00 2.52719074e-01
8.53557348e-01 6.77605629e-01 4.01445597e-01 8.45095217e-01
-2.46789679e-01 7.43187785e-01 -9.64123607e-01 -6.47003233e-01
1.67347193e-01 -5.72957218e-01 -6.06934726e-01 7.06676900e-01
3.64413917e-01 -3.22092175e-01 -2.92764515e-01 8.72504950e-01
3.43428940e-01 -1.48854107e-01 7.98565626e-01 -8.60763609e-01
-8.46355557e-01 7.75349498e-01 -3.45429003e-01 -4.28355634e-01
1.45510450e-01 1.08762816e-01 6.96904778e-01 -8.37456107e-01
6.91824436e-01 9.03513908e-01 1.83407679e-01 7.84638286e-01
-1.21335578e+00 -2.52143174e-01 -1.73164885e-02 -1.90067798e-01
-1.20181048e+00 -2.50802934e-01 8.20512235e-01 -5.28782725e-01
4.64928091e-01 1.62095398e-01 4.83151495e-01 1.37883413e+00
-1.69992492e-01 9.47556019e-01 1.19053519e+00 -8.93364251e-01
3.65014881e-01 4.54976916e-01 1.66004933e-02 7.89527178e-01
3.47461075e-01 -2.76708901e-01 -3.62662464e-01 2.99092889e-01
9.13199842e-01 -3.74113992e-02 -3.34522784e-01 -3.69737715e-01
-1.17013454e+00 6.37854218e-01 -3.46277654e-02 2.90191829e-01
1.14753760e-01 -1.23102576e-01 1.20907180e-01 3.49171251e-01
1.20623939e-01 8.16074014e-01 -1.78313926e-01 -2.03364193e-01
-1.23308611e+00 -1.27536386e-01 9.99759197e-01 1.20099080e+00
8.15573692e-01 3.51434797e-01 -9.40539613e-02 1.03331745e+00
-3.86394304e-03 5.82622290e-01 7.23179996e-01 -3.41659933e-01
3.62257063e-01 6.13582850e-01 8.90110508e-02 -6.59763217e-01
1.42273493e-02 -3.72144371e-01 -9.72799897e-01 3.87452960e-01
5.46654642e-01 -2.12133184e-01 -1.19879889e+00 1.32060671e+00
-2.67378092e-01 -6.55806541e-01 9.13906097e-03 7.67513275e-01
5.27269840e-01 5.88124454e-01 -4.35940176e-01 -1.12667736e-02
1.13231790e+00 -1.12263632e+00 -6.16073608e-01 -3.08450192e-01
3.57314557e-01 -9.58025694e-01 1.28340256e+00 7.42210209e-01
-7.14352310e-01 -4.61656332e-01 -1.24597549e+00 -6.66786656e-02
-5.74466825e-01 7.10437477e-01 4.10671890e-01 9.12925482e-01
-7.45969892e-01 5.94176471e-01 -6.53846443e-01 -3.01691979e-01
3.77854913e-01 2.69146055e-01 -3.62508476e-01 -6.39469028e-02
-7.45285451e-01 7.03979552e-01 4.26632345e-01 1.10051051e-01
-6.27343595e-01 -2.35947832e-01 -7.42101789e-01 3.66748087e-02
4.50337350e-01 -1.99664608e-01 1.08437681e+00 -1.16737163e+00
-2.10895491e+00 9.53129232e-01 -1.18818888e-02 -3.18462849e-01
1.10858810e+00 -5.51909879e-02 -1.97036371e-01 3.28914821e-02
-3.97723734e-01 5.57572186e-01 1.29474878e+00 -1.32597101e+00
-1.74791187e-01 -2.25154385e-01 -1.26494244e-01 -7.94243142e-02
-4.65198666e-01 -2.03802332e-01 -5.68942070e-01 -8.12895417e-01
-1.10729434e-01 -9.86125231e-01 -2.54035816e-02 5.35032079e-02
-7.18478441e-01 5.60355047e-03 9.39626515e-01 -4.54342365e-01
9.44243729e-01 -2.11066699e+00 2.21938670e-01 2.55816102e-01
7.94623271e-02 4.77608830e-01 -1.10847630e-01 4.20712620e-01
2.74674416e-01 2.02620015e-01 -4.03667539e-01 -5.15023768e-01
6.14444315e-02 2.54486918e-01 -4.94269401e-01 1.64161816e-01
3.28235030e-01 8.60821187e-01 -6.42143071e-01 -4.75636631e-01
1.34867877e-01 2.44934082e-01 -2.40913257e-01 1.83329999e-01
-4.28964972e-01 4.28794473e-01 -2.35151798e-01 7.13970184e-01
4.51315790e-01 -2.40582958e-01 1.60579443e-01 -5.51563241e-02
-2.57532090e-01 -1.92899540e-01 -1.07789195e+00 1.39691162e+00
-6.92141831e-01 9.14674520e-01 -3.63469243e-01 -8.64769578e-01
1.45336258e+00 9.55943689e-02 3.52396979e-03 -3.29049736e-01
1.57154813e-01 4.90907878e-01 -1.87548667e-01 -2.74155110e-01
7.08332241e-01 2.13781804e-01 6.70267418e-02 6.75534844e-01
-8.56290236e-02 -4.21279222e-01 5.41031301e-01 7.82483350e-03
6.85395420e-01 2.97903538e-01 1.18480965e-01 -1.21227847e-02
4.06371802e-01 6.44832328e-02 1.62683204e-01 9.01472211e-01
2.95966506e-01 9.79592919e-01 6.17077053e-01 -3.71160865e-01
-1.35850942e+00 -7.95523763e-01 -1.03285708e-01 7.35452950e-01
-3.36765125e-02 -6.23907894e-03 -9.28354621e-01 -6.38296962e-01
-2.36212507e-01 4.66420650e-01 -5.43179154e-01 1.07603930e-01
-6.57357752e-01 -5.44778049e-01 5.93204141e-01 4.47967768e-01
6.07990861e-01 -1.18731797e+00 -5.70305765e-01 1.85147241e-01
2.14384690e-01 -1.21786928e+00 -4.83476311e-01 3.01086336e-01
-7.59642839e-01 -7.30223596e-01 -1.26517725e+00 -9.03490424e-01
1.08375764e+00 -9.42761302e-02 8.36369216e-01 -6.56104460e-02
-3.72537911e-01 2.81438708e-01 -6.59310460e-01 -3.81556869e-01
-8.13637018e-01 4.33974087e-01 -2.61363149e-01 3.65107059e-01
1.10490555e-02 -2.73325026e-01 -3.35205466e-01 4.79556471e-01
-1.19987464e+00 2.25152999e-01 8.55632842e-01 9.65497792e-01
5.92516184e-01 5.70650073e-03 2.53423959e-01 -1.07455659e+00
5.28230667e-01 1.95757791e-01 -7.67951608e-01 6.00638688e-01
-3.49908561e-01 2.22643197e-01 1.08926344e+00 -7.51466215e-01
-8.76574397e-01 4.08808202e-01 7.39596970e-03 -2.26380795e-01
-2.65677661e-01 3.23151857e-01 -2.73741663e-01 -2.22002521e-01
6.84039354e-01 5.77148974e-01 -1.06998133e-02 -4.28364098e-01
1.85628399e-01 9.58418429e-01 5.16903698e-01 -6.04019105e-01
9.85857725e-01 3.14598948e-01 -7.59987682e-02 -1.19567168e+00
-5.78670025e-01 1.57911867e-01 -9.31754351e-01 -1.89944878e-01
6.64739728e-01 -4.04133499e-01 -4.24539894e-01 9.41758811e-01
-1.09360337e+00 -7.97594130e-01 -3.64974260e-01 2.21611634e-01
-4.88417238e-01 4.50152069e-01 -5.44125319e-01 -7.19477475e-01
-1.43741816e-01 -1.32262361e+00 1.03327227e+00 3.36349607e-01
-1.09761186e-01 -9.63285446e-01 -1.62674755e-01 2.26580068e-01
4.20945197e-01 1.98142514e-01 9.48483884e-01 -6.50580645e-01
-7.21467316e-01 -4.80566442e-01 -1.90062910e-01 5.53980112e-01
3.48959446e-01 2.61597812e-01 -7.58108377e-01 -3.33444774e-01
-3.36105973e-01 -4.61288393e-01 7.76629329e-01 -1.76729634e-01
1.24614334e+00 -1.96689487e-01 2.11525671e-02 4.91657853e-01
1.30552661e+00 2.55861044e-01 8.26342821e-01 2.94524878e-01
9.48760450e-01 4.80465382e-01 2.08162367e-01 4.29023147e-01
-5.04412390e-02 8.28834713e-01 -6.29545823e-02 -2.14817420e-01
-1.90579817e-01 -3.06437552e-01 3.95199180e-01 7.56755650e-01
-5.19792438e-02 -4.90099847e-01 -8.07545960e-01 1.99626967e-01
-1.45644617e+00 -7.06057012e-01 -4.50784527e-02 2.33304310e+00
1.12530136e+00 1.26576215e-01 -4.91385944e-02 4.44080055e-01
8.16346228e-01 3.86484079e-02 -5.93570113e-01 -5.62490582e-01
-4.61487055e-01 3.46548796e-01 5.07911742e-01 3.38349521e-01
-9.12591755e-01 1.09682906e+00 5.59716129e+00 7.50819027e-01
-1.55962753e+00 -3.08645785e-01 6.03052914e-01 2.04409510e-01
-1.99511826e-01 -1.92996904e-01 -9.13001895e-01 4.60826337e-01
3.71418417e-01 3.41185331e-02 6.78708851e-01 7.89517105e-01
3.40982080e-02 -1.16240777e-01 -1.39321244e+00 1.08164442e+00
4.47279096e-01 -1.24574471e+00 4.42802936e-01 5.10305427e-02
9.17074680e-01 -6.23468637e-01 1.63556606e-01 6.60494342e-02
8.02799836e-02 -9.84057665e-01 7.67037630e-01 4.78565425e-01
1.07304859e+00 -5.72070777e-01 5.03006816e-01 3.46895933e-01
-6.92681968e-01 1.88773721e-01 -3.72025043e-01 3.22659045e-01
3.39814797e-02 7.34283268e-01 -7.36577690e-01 3.12830776e-01
1.47737056e-01 4.22008067e-01 -8.17131042e-01 9.27466094e-01
-5.45934319e-01 4.92390513e-01 -9.96511057e-02 -3.54534209e-01
1.67282760e-01 -3.12611520e-01 3.89764965e-01 1.32186007e+00
4.44184870e-01 -2.37858370e-01 9.99483392e-02 9.80789661e-01
-3.49621236e-01 1.29961476e-01 -5.68998933e-01 -4.52317864e-01
2.42784292e-01 1.11714303e+00 -9.39796984e-01 -3.39586079e-01
-8.10668096e-02 1.44814265e+00 6.15983009e-02 3.39766592e-01
-5.55780292e-01 -8.42059553e-01 -1.70839190e-01 4.61024456e-02
3.94950479e-01 -3.99063855e-01 -3.38394105e-01 -1.28752851e+00
1.88419312e-01 -1.07301998e+00 -1.75188333e-01 -7.27878809e-01
-1.03241897e+00 7.38766789e-01 -4.09102052e-01 -1.20284534e+00
-1.75993711e-01 -9.82867360e-01 -3.08092564e-01 7.28199482e-01
-1.34849083e+00 -1.11105466e+00 -4.76368666e-01 3.84714901e-01
8.10512543e-01 -3.15078408e-01 7.16341436e-01 1.51975885e-01
-6.35136962e-01 9.92779315e-01 2.46956125e-01 6.28765881e-01
8.52703929e-01 -1.34533751e+00 3.50315094e-01 8.96812677e-01
4.92429852e-01 4.81275052e-01 5.17268598e-01 -5.13005316e-01
-1.56562710e+00 -1.02794135e+00 6.09495044e-01 -2.94662625e-01
4.62339073e-01 -8.05293262e-01 -8.65041614e-01 4.49233890e-01
-3.72842997e-02 -1.92471936e-01 4.48736101e-01 -2.94769973e-01
-4.20459628e-01 -9.21911150e-02 -9.18220222e-01 8.67664814e-01
6.95908427e-01 -5.40112317e-01 -3.72986585e-01 4.01433140e-01
3.17176133e-01 -5.80100000e-01 -5.76157451e-01 3.73762585e-02
7.49210477e-01 -8.41086268e-01 3.55675131e-01 -2.72725999e-01
7.74738908e-01 -3.02382201e-01 5.58495522e-02 -1.31308258e+00
2.80887373e-02 -8.03617299e-01 -7.98083283e-03 1.31447411e+00
7.37882197e-01 -6.21265829e-01 9.56759155e-01 4.67732787e-01
2.01961342e-02 -5.76509833e-01 -3.31877917e-01 -1.04762328e+00
6.47639036e-02 4.33988608e-02 5.86220801e-01 8.29413712e-01
-2.28603214e-01 2.39821598e-01 -5.68409324e-01 1.07455915e-02
4.02863562e-01 2.44637042e-01 9.10735548e-01 -1.14104438e+00
-6.32597744e-01 -4.77334619e-01 -3.87606442e-01 -1.28339398e+00
6.73942566e-02 -7.17340469e-01 1.96902543e-01 -1.24972081e+00
1.72577444e-02 -4.02977377e-01 2.36088797e-01 5.02617240e-01
4.56963293e-02 5.18574119e-01 3.18049312e-01 3.01407605e-01
-3.24267924e-01 2.81633586e-01 1.46365702e+00 -3.53916019e-01
-3.58047754e-01 1.87602583e-02 -3.07547987e-01 4.37145233e-01
8.80680740e-01 -1.58566028e-01 -2.25656584e-01 -6.19138479e-01
2.08111748e-01 -1.23769276e-01 2.17424467e-01 -9.46483254e-01
2.23270774e-01 -1.98352132e-02 5.02948582e-01 -3.08741063e-01
5.99399321e-02 -5.37432909e-01 -8.94000828e-02 3.21457088e-01
-5.33127189e-01 -9.22182277e-02 -4.00820896e-02 3.38328987e-01
-2.24212632e-01 -7.30139911e-01 9.79063153e-01 -2.73723781e-01
-4.33271289e-01 2.02693224e-01 -6.23047650e-01 -1.55315220e-01
7.51572132e-01 -4.97697055e-01 -4.41809416e-01 -2.98199475e-01
-6.38503194e-01 -3.02680522e-01 6.31888390e-01 2.72383600e-01
4.93014485e-01 -9.96365368e-01 -6.15252614e-01 1.88218266e-01
1.81612208e-01 1.22201420e-01 -7.27961510e-02 4.54463035e-01
-7.80055702e-01 3.00477684e-01 -2.27575779e-01 -4.99198675e-01
-9.94868815e-01 5.17637432e-01 2.36468673e-01 -2.54628122e-01
-4.48531330e-01 4.06873584e-01 -8.74923989e-02 -1.61408722e-01
3.64976645e-01 -1.35964021e-01 -9.07411575e-02 7.51527548e-02
4.81751382e-01 1.64785400e-01 2.84362078e-01 -2.11459577e-01
2.13117972e-01 6.88568115e-01 -4.15963590e-01 -1.64717674e-01
1.14855146e+00 2.13451982e-01 -1.74950987e-01 5.05741358e-01
9.35631037e-01 1.70627296e-01 -1.24745023e+00 -1.25772640e-01
-6.79096729e-02 -3.30776364e-01 -3.03033918e-01 -9.42687333e-01
-9.65780079e-01 9.16425109e-01 3.87015641e-01 2.42963076e-01
9.14624691e-01 -2.70121932e-01 5.36489964e-01 9.02698755e-01
4.89319086e-01 -1.18876648e+00 3.27414423e-01 6.27450168e-01
9.33999002e-01 -1.18239737e+00 -5.92808351e-02 -2.55675077e-01
-6.81954920e-01 1.62155008e+00 4.00830060e-01 -1.19089998e-01
1.52523130e-01 4.17704642e-01 1.23149678e-01 1.72054380e-01
-3.12075883e-01 -4.54797447e-02 1.03623495e-01 4.67329264e-01
4.24507290e-01 5.29201590e-02 -1.87897086e-01 2.56636351e-01
-3.43119740e-01 -3.49304527e-02 9.33565497e-01 1.06007981e+00
-2.61854202e-01 -1.48408079e+00 -4.61337537e-01 3.96618873e-01
-1.88078612e-01 -5.31682000e-03 -7.92401254e-01 6.17276013e-01
-1.74815163e-01 7.05492079e-01 -5.60485274e-02 -1.93721563e-01
2.43864059e-01 1.25576869e-01 6.56969368e-01 -7.15056539e-01
-2.56798893e-01 -6.63972721e-02 -2.16371909e-01 1.15820326e-01
-1.20076872e-01 -4.28230524e-01 -8.82411182e-01 -5.61104938e-02
-3.86923701e-01 1.73738282e-02 8.92066479e-01 9.42964077e-01
1.12506105e-02 3.75874192e-01 9.38866675e-01 -8.89275253e-01
-5.81753552e-01 -7.55052149e-01 -5.80655932e-01 2.95218378e-01
2.09388837e-01 -2.92792648e-01 -2.60254025e-01 4.66538876e-01] | [11.778715133666992, 2.339946985244751] |
3f83ef04-1a87-4baf-a565-062ca3c8ebea | dual-path-style-learning-for-end-to-end-noise | 2203.14838 | null | https://arxiv.org/abs/2203.14838v3 | https://arxiv.org/pdf/2203.14838v3.pdf | Dual-Path Style Learning for End-to-End Noise-Robust Speech Recognition | Automatic speech recognition (ASR) systems degrade significantly under noisy conditions. Recently, speech enhancement (SE) is introduced as front-end to reduce noise for ASR, but it also suppresses some important speech information, i.e., over-suppression. To alleviate this, we propose a dual-path style learning approach for end-to-end noise-robust speech recognition (DPSL-ASR). Specifically, we first introduce clean speech feature along with the fused feature from IFF-Net as dual-path inputs to recover the suppressed information. Then, we propose style learning to map the fused feature close to clean feature, in order to learn latent speech information from the latter, i.e., clean "speech style". Furthermore, we also minimize the distance of final ASR outputs in two paths to improve noise-robustness. Experiments show that the proposed approach achieves relative word error rate (WER) reductions of 10.6% and 8.6% over the best IFF-Net baseline, on RATS and CHiME-4 datasets respectively. | ['Eng Siong Chng', 'Chen Chen', 'Nana Hou', 'Yuchen Hu'] | 2022-03-28 | null | null | null | null | ['robust-speech-recognition'] | ['speech'] | [ 4.61122692e-01 -3.59021053e-02 3.25753391e-01 -4.31804836e-01
-1.18605328e+00 -2.78156042e-01 3.07638526e-01 -1.55104294e-01
-4.56127793e-01 3.46901685e-01 7.90926278e-01 -3.45911384e-01
1.89191505e-01 -2.81836718e-01 -6.45908713e-01 -9.13247645e-01
4.84656930e-01 -4.34545189e-01 -5.25809228e-02 -3.29776794e-01
-8.29204023e-02 3.38292807e-01 -1.14283502e+00 3.55112702e-01
1.04335546e+00 8.64955187e-01 6.54495358e-01 5.67259729e-01
-7.11237565e-02 6.65612459e-01 -7.56349206e-01 -6.49267808e-02
1.82273045e-01 -3.95031512e-01 -1.92043960e-01 2.06875101e-01
1.04567207e-01 -1.73456296e-01 -9.10930753e-01 1.39428878e+00
8.52162540e-01 4.77448434e-01 3.57083350e-01 -6.20849311e-01
-6.70752466e-01 7.04574168e-01 -6.09576523e-01 2.39375219e-01
1.10843249e-01 3.52047890e-01 7.39272833e-01 -1.11689889e+00
3.64527814e-02 1.63774979e+00 2.56027818e-01 8.92540514e-01
-1.13916576e+00 -8.05491149e-01 3.06950539e-01 8.60463232e-02
-1.34801400e+00 -1.36389053e+00 8.59764040e-01 1.61985606e-01
8.81788254e-01 4.39945072e-01 -1.48108089e-02 1.21487987e+00
-2.24276125e-01 1.04354405e+00 1.01601493e+00 -4.69867766e-01
6.85440227e-02 -3.02484911e-02 7.83345401e-02 2.88823187e-01
-2.56242961e-01 1.55241653e-01 -6.83199406e-01 2.63896763e-01
3.76080632e-01 -2.33955219e-01 -5.30438721e-01 4.34671134e-01
-1.02653491e+00 1.94569096e-01 3.11048478e-01 2.17155948e-01
-3.47187787e-01 -3.86295319e-02 4.18845594e-01 4.57917958e-01
5.83009064e-01 -5.55298366e-02 -5.48955381e-01 -1.38076663e-01
-7.58514762e-01 -6.76555280e-03 3.42481822e-01 1.04386902e+00
4.40355569e-01 5.81117153e-01 -6.39319003e-01 1.60651946e+00
6.85061574e-01 7.83785522e-01 5.46238959e-01 -5.93826234e-01
1.06315529e+00 3.98826636e-02 -6.25500530e-02 -5.39518476e-01
1.06418036e-01 -8.83519888e-01 -1.22029197e+00 -1.57945957e-02
-9.00283754e-02 -2.98992395e-01 -1.43475318e+00 1.84546185e+00
-3.37871239e-02 2.64922827e-01 4.65938538e-01 8.66494358e-01
6.61844671e-01 1.16454720e+00 7.56898820e-02 -4.64704454e-01
1.15313828e+00 -1.21218061e+00 -1.05004966e+00 -6.17204189e-01
4.05077785e-01 -9.91105556e-01 1.16102147e+00 5.56924582e-01
-1.08932233e+00 -6.30456507e-01 -1.06172431e+00 4.52425182e-02
-7.28048831e-02 5.24761260e-01 -4.07605410e-01 8.06888282e-01
-8.95655870e-01 3.89267504e-01 -7.83134818e-01 -4.18772874e-03
3.79763722e-01 1.29442096e-01 -2.67957866e-01 -2.33634263e-01
-1.16042054e+00 6.14611149e-01 2.32813731e-01 3.31581801e-01
-9.84498858e-01 -6.84514701e-01 -9.13650811e-01 2.08442107e-01
5.10066450e-01 -2.51979053e-01 1.26631927e+00 -7.32781589e-01
-1.89221585e+00 3.78388643e-01 -5.70596576e-01 -4.50192720e-01
3.24098438e-01 -4.81225133e-01 -8.36710453e-01 -2.54961073e-01
-2.69862711e-01 3.11190665e-01 9.49232876e-01 -1.36345696e+00
-5.03984153e-01 -3.74415725e-01 -4.60276783e-01 4.99969035e-01
-3.44742745e-01 3.81591529e-01 -5.18711388e-01 -1.23556519e+00
3.63418639e-01 -6.72824442e-01 -2.60661721e-01 -4.79523778e-01
-4.72506195e-01 -9.16953385e-02 1.04185379e+00 -1.33973801e+00
1.44336498e+00 -2.51779151e+00 2.83026658e-02 1.56534687e-01
-1.68653354e-01 9.06736076e-01 -5.15051484e-01 7.92613551e-02
-2.75274992e-01 1.65610343e-01 -3.77028614e-01 -7.15756893e-01
-1.56776533e-01 1.38402194e-01 -3.05867553e-01 3.01178038e-01
3.81071985e-01 4.88864511e-01 -7.70494163e-01 6.05335413e-03
4.04147834e-01 6.61117136e-01 -3.24728161e-01 2.99334913e-01
1.07575841e-01 4.70893025e-01 -1.90048352e-01 4.13810879e-01
9.16912198e-01 6.69872761e-01 -2.05275849e-01 -2.12659225e-01
-1.06398515e-01 7.42683768e-01 -1.23027050e+00 1.64395010e+00
-8.72069180e-01 3.35810125e-01 6.09668672e-01 -8.69663239e-01
1.24920595e+00 5.46443045e-01 -2.06592545e-01 -9.38537240e-01
7.37306057e-03 2.88207829e-01 1.04837425e-01 -1.79459006e-01
1.60101905e-01 -6.83391169e-02 1.84448436e-01 -1.35169804e-01
2.34281532e-02 1.72705337e-01 -3.35815638e-01 1.66654438e-02
1.16238928e+00 -2.27504179e-01 1.54797733e-01 -4.76937629e-02
1.04740763e+00 -9.00219381e-01 8.77034783e-01 5.82089841e-01
-2.57427573e-01 7.85653710e-01 1.41334653e-01 3.41860443e-01
-9.70000148e-01 -1.21984720e+00 3.21228385e-01 9.32107866e-01
-2.96261199e-02 -2.89243251e-01 -9.10381854e-01 -5.83296716e-01
-3.39517593e-01 1.03102207e+00 1.95814017e-02 -3.46324772e-01
-7.99840569e-01 -3.96458387e-01 7.03855038e-01 4.41071510e-01
5.44550180e-01 -1.03228009e+00 5.20048738e-01 2.74507403e-01
-1.92592993e-01 -1.05706465e+00 -9.40159380e-01 4.11176145e-01
-5.41991711e-01 -1.46200553e-01 -8.10179114e-01 -7.83467889e-01
5.15703678e-01 6.85246944e-01 3.90643060e-01 4.49181236e-02
2.06916496e-01 -2.49582157e-01 -5.76464176e-01 -2.03386303e-02
-7.25339770e-01 -1.17306761e-01 3.21717829e-01 3.89373630e-01
1.44561782e-01 -6.58384383e-01 -4.31111336e-01 3.45918059e-01
-7.76197672e-01 -1.72313824e-01 8.61233890e-01 1.10552359e+00
6.24219954e-01 2.37443134e-01 9.31172431e-01 -4.70746189e-01
7.59799659e-01 -2.83281505e-01 -3.32008094e-01 2.98635930e-01
-3.74471456e-01 1.18539773e-01 9.78745162e-01 -3.61080736e-01
-1.60245407e+00 1.17875382e-01 -7.49613464e-01 -5.27593672e-01
-4.18822169e-01 3.12224776e-01 -9.63848770e-01 1.81612119e-01
5.56316495e-01 7.06301928e-01 -2.76470572e-01 -9.44993198e-01
3.25081229e-01 1.46761024e+00 6.35029078e-01 -3.01725239e-01
9.83735681e-01 -1.44049466e-01 -4.70927596e-01 -1.29017615e+00
-7.37760127e-01 -7.14102149e-01 -1.84722114e-02 2.78241277e-01
4.60669965e-01 -1.18792546e+00 -2.23981962e-01 6.71042562e-01
-1.27459323e+00 -1.41905174e-01 5.90080246e-02 5.91006994e-01
-3.59176695e-01 5.02026320e-01 -6.07627928e-01 -1.20520020e+00
-7.62645721e-01 -1.25224161e+00 9.70050812e-01 2.46488750e-01
1.87766939e-01 -3.12848061e-01 -3.09307665e-01 4.18879449e-01
4.73785549e-01 -6.40557051e-01 5.86240828e-01 -7.03122020e-01
-2.68200815e-01 1.97250083e-01 -3.56630206e-01 1.14143884e+00
4.27936286e-01 -5.29835820e-01 -1.24397504e+00 -3.72024775e-01
2.42336556e-01 1.19937912e-01 9.83906150e-01 2.03388423e-01
1.16626930e+00 -4.45811331e-01 -1.49543688e-01 6.71787679e-01
1.04156446e+00 6.34714484e-01 1.01265252e+00 -9.31191593e-02
7.05846012e-01 3.58866334e-01 5.84071338e-01 1.02246724e-01
-2.82522768e-01 6.99050069e-01 -1.45991728e-01 1.29145272e-02
-8.78611982e-01 -4.42354798e-01 8.09253156e-01 1.42367876e+00
4.99441415e-01 -5.20676911e-01 -6.57552958e-01 4.81540680e-01
-1.61357713e+00 -6.61494374e-01 1.16758205e-01 2.18631697e+00
9.59693789e-01 2.58199841e-01 -1.83722630e-01 2.97205031e-01
1.01552343e+00 3.64592731e-01 -3.87320518e-01 -1.91008225e-01
-3.08600187e-01 3.07528883e-01 4.56941664e-01 6.70938253e-01
-8.60686123e-01 1.05041647e+00 4.83100843e+00 1.51252186e+00
-9.31127191e-01 2.70073771e-01 6.80089176e-01 8.62923115e-02
-3.00501704e-01 -2.18457207e-01 -6.83285117e-01 5.72021484e-01
1.01688457e+00 -9.96486023e-02 7.55609870e-01 6.82547808e-01
9.08931017e-01 3.06871206e-01 -7.15653956e-01 1.15178728e+00
-2.19629034e-02 -6.14753485e-01 3.52886058e-02 -2.66514182e-01
3.98982406e-01 -4.67439331e-02 1.06558464e-01 4.12569046e-01
3.92114371e-02 -9.32595849e-01 8.16124499e-01 3.19682568e-01
8.42322946e-01 -1.02438176e+00 5.70581913e-01 4.29807544e-01
-1.22981572e+00 -1.58018112e-01 -3.10459346e-01 3.79635423e-01
1.71165109e-01 7.42635012e-01 -6.50993645e-01 5.41596651e-01
4.10581172e-01 4.50104266e-01 -2.77369499e-01 1.10052776e+00
-5.91901243e-01 1.09771156e+00 -3.30758274e-01 1.96626797e-01
3.06875501e-02 -8.97304192e-02 1.04838574e+00 1.45494676e+00
4.09404010e-01 1.72898293e-01 -1.09055787e-01 7.04644322e-01
-3.18953753e-01 6.25974759e-02 -3.01488936e-01 -2.75745858e-02
7.59957075e-01 9.35528815e-01 -1.61318332e-01 -4.42863703e-02
-2.31524959e-01 1.18152523e+00 1.06069617e-01 6.06601059e-01
-5.31616449e-01 -8.18365335e-01 8.83694112e-01 -1.55072585e-01
2.25854322e-01 -2.89392799e-01 -3.34395856e-01 -1.08386612e+00
2.44778708e-01 -1.18938303e+00 -2.40028724e-01 -7.82747686e-01
-1.12768805e+00 6.30834579e-01 -6.03416324e-01 -1.10560548e+00
2.14133650e-01 -3.98278058e-01 -6.93038404e-01 1.36073780e+00
-1.63326561e+00 -1.03972042e+00 1.77509151e-02 3.93450499e-01
1.24184048e+00 -3.34569216e-01 4.26344454e-01 6.88925266e-01
-7.86892772e-01 1.12095344e+00 3.04795086e-01 8.00108910e-02
7.54819095e-01 -9.75567937e-01 9.18332815e-01 1.34872770e+00
3.23064737e-02 6.73421383e-01 6.83647156e-01 -7.71816254e-01
-1.28487706e+00 -1.41435134e+00 6.39351428e-01 2.80959278e-01
3.59999925e-01 -5.63824832e-01 -1.07434070e+00 4.37175691e-01
1.20976552e-01 -7.48546273e-02 1.85328260e-01 -8.29779580e-02
-4.87001151e-01 -5.13358414e-01 -9.87334669e-01 8.45508575e-01
1.19573724e+00 -6.42020285e-01 -7.06526101e-01 -6.90876767e-02
1.35366118e+00 -3.05286616e-01 -2.79434890e-01 3.90321374e-01
1.21800616e-01 -5.20294845e-01 8.82962286e-01 -2.18569562e-01
-1.06223747e-02 -6.95384562e-01 -4.67976302e-01 -1.92343044e+00
-2.75347263e-01 -1.05397058e+00 -7.80306682e-02 1.79029715e+00
6.28921092e-01 -3.96384060e-01 4.80599970e-01 -6.21801876e-02
-6.91834152e-01 -3.02811742e-01 -8.28215539e-01 -9.19296682e-01
-2.33412847e-01 -5.01989067e-01 4.33822900e-01 4.48913217e-01
-3.56753945e-01 3.92121106e-01 -6.19661748e-01 4.76075143e-01
6.79365873e-01 -7.79716492e-01 4.51247722e-01 -4.95518476e-01
-2.75042683e-01 -1.51723161e-01 -4.12434787e-02 -1.49247718e+00
1.57956436e-01 -7.52038538e-01 6.18725538e-01 -1.34989393e+00
-1.88277319e-01 -2.71012485e-01 -5.43806434e-01 2.98495233e-01
-4.71717596e-01 -2.84601808e-01 2.44274884e-01 -2.13030741e-01
-4.37433511e-01 9.55948710e-01 1.07131290e+00 -2.21452430e-01
-4.47849423e-01 2.47261688e-01 -7.72481263e-01 5.62718809e-01
8.87584805e-01 -5.14054239e-01 -3.59662503e-01 -5.82245588e-01
-5.40485680e-01 1.76826343e-01 -7.17011001e-03 -1.10044527e+00
2.17655107e-01 -2.22107973e-02 1.13530792e-01 -6.50432646e-01
3.54276508e-01 -6.42364919e-01 -1.91471636e-01 1.60788253e-01
-6.35210395e-01 -5.74881256e-01 2.14094386e-01 7.70992398e-01
-4.24146324e-01 -2.34260947e-01 1.08989549e+00 9.89803672e-02
-3.65283608e-01 1.01745306e-02 -4.20114934e-01 -1.15447856e-01
4.21778440e-01 1.75035089e-01 -2.97517926e-01 -4.23471034e-01
-7.23883152e-01 2.03301877e-01 -1.09742217e-01 5.71719587e-01
7.98957705e-01 -1.33612049e+00 -1.02896249e+00 2.69346118e-01
-1.96005832e-02 -6.70782700e-02 5.66081345e-01 4.71487284e-01
-3.83922830e-02 2.33885944e-01 3.14385027e-01 -1.75807953e-01
-1.46300113e+00 4.06931490e-01 2.94874340e-01 -2.81530954e-02
-5.51187515e-01 9.61794794e-01 3.05462152e-01 -4.62025106e-01
6.67885184e-01 -2.14805380e-01 2.75510759e-03 -3.95143092e-01
8.04019868e-01 4.67785448e-01 4.47261304e-01 -6.17533505e-01
-3.04465443e-01 2.13449404e-01 -3.70637685e-01 -3.82434040e-01
1.25147080e+00 -6.62258089e-01 9.94562581e-02 6.20449781e-02
1.37378001e+00 3.40844840e-01 -9.70527291e-01 -5.71692228e-01
-9.60535370e-03 -4.56063539e-01 4.21669215e-01 -1.07678688e+00
-1.02825725e+00 9.90585029e-01 7.85948038e-01 -6.17218651e-02
1.42083943e+00 -3.28428686e-01 1.16875243e+00 3.25262219e-01
-1.49011612e-01 -1.12928140e+00 -1.78976804e-02 6.87071919e-01
1.13035691e+00 -1.07193696e+00 -5.77998221e-01 -5.61635673e-01
-5.30844033e-01 8.32637429e-01 5.56214392e-01 1.39545694e-01
4.91562039e-01 4.24653858e-01 1.32411271e-01 4.91797656e-01
-6.19045377e-01 -3.35686445e-01 1.53133512e-01 7.36768484e-01
3.23619902e-01 1.69861719e-01 -1.88595131e-01 8.96661997e-01
-1.14151612e-01 -3.29180181e-01 2.74476171e-01 6.45745754e-01
-6.90064490e-01 -1.12466645e+00 -5.62538743e-01 2.92539090e-01
-5.41235507e-01 -4.74384129e-01 -1.55752301e-01 -1.18560806e-01
-8.54332149e-02 1.54509366e+00 -3.79075110e-01 -7.13843167e-01
7.99378574e-01 1.49969488e-01 3.20533141e-02 -7.31545627e-01
-4.78065431e-01 5.48291087e-01 2.27834240e-01 -2.48979375e-01
2.29986861e-01 -3.99185807e-01 -1.02762330e+00 1.13918550e-01
-4.86405790e-01 4.75887284e-02 7.11997211e-01 9.67879236e-01
3.08397114e-01 9.25703406e-01 1.00575316e+00 -4.55816448e-01
-8.48266721e-01 -1.22745717e+00 -5.52475393e-01 3.04492176e-01
6.60198450e-01 -1.42146498e-01 -6.55105472e-01 -2.78716944e-02] | [14.797005653381348, 6.136204719543457] |
0e2d4644-d714-41d7-8498-0164a7606fc3 | harmonic-networks-deep-translation-and | 1612.04642 | null | http://arxiv.org/abs/1612.04642v2 | http://arxiv.org/pdf/1612.04642v2.pdf | Harmonic Networks: Deep Translation and Rotation Equivariance | Translating or rotating an input image should not affect the results of many
computer vision tasks. Convolutional neural networks (CNNs) are already
translation equivariant: input image translations produce proportionate feature
map translations. This is not the case for rotations. Global rotation
equivariance is typically sought through data augmentation, but patch-wise
equivariance is more difficult. We present Harmonic Networks or H-Nets, a CNN
exhibiting equivariance to patch-wise translation and 360-rotation. We achieve
this by replacing regular CNN filters with circular harmonics, returning a
maximal response and orientation for every receptive field patch.
H-Nets use a rich, parameter-efficient and low computational complexity
representation, and we show that deep feature maps within the network encode
complicated rotational invariants. We demonstrate that our layers are general
enough to be used in conjunction with the latest architectures and techniques,
such as deep supervision and batch normalization. We also achieve
state-of-the-art classification on rotated-MNIST, and competitive results on
other benchmark challenges. | ['Gabriel J. Brostow', 'Stephan J. Garbin', 'Daniyar Turmukhambetov', 'Daniel E. Worrall'] | 2016-12-14 | harmonic-networks-deep-translation-and-1 | http://openaccess.thecvf.com/content_cvpr_2017/html/Worrall_Harmonic_Networks_Deep_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Worrall_Harmonic_Networks_Deep_CVPR_2017_paper.pdf | cvpr-2017-7 | ['rotated-mnist'] | ['computer-vision'] | [ 7.58039504e-02 1.24834433e-01 -1.30834132e-01 -4.22394007e-01
-8.68361741e-02 -6.96627438e-01 8.60911369e-01 -4.23498124e-01
-6.05345428e-01 9.33673456e-02 4.02125031e-01 -7.18551725e-02
1.39957175e-01 -5.78405261e-01 -1.16032541e+00 -7.74203539e-01
-1.15672825e-02 2.43583158e-01 5.89788426e-03 -6.43896520e-01
-4.57908679e-03 1.14846241e+00 -1.26706851e+00 3.28877270e-01
-1.07861713e-01 5.55386424e-01 -3.64443660e-01 7.24572182e-01
6.74651086e-01 4.52177078e-01 -3.26243937e-01 -3.23939711e-01
6.89727366e-01 -1.63570851e-01 -9.57504034e-01 -1.00579292e-01
8.39701533e-01 -3.45860302e-01 -6.86523557e-01 8.67518365e-01
5.26127398e-01 1.34577483e-01 6.45256937e-01 -9.49351728e-01
-1.15446198e+00 7.02657998e-01 -3.86272848e-01 2.42766395e-01
-9.05435067e-03 -1.84148252e-01 9.83844042e-01 -1.04902101e+00
9.67947960e-01 1.34471190e+00 6.84963942e-01 3.66385162e-01
-1.27678204e+00 -2.37714484e-01 1.60419285e-01 4.94588017e-02
-1.29530501e+00 -4.75574553e-01 6.91031456e-01 -1.67114422e-01
1.43836915e+00 2.68012822e-01 7.69877255e-01 1.16982341e+00
5.82649231e-01 4.52322662e-01 6.87703669e-01 -2.90086985e-01
-7.20393658e-02 -3.94313335e-01 -3.24582189e-01 6.11578286e-01
2.74435848e-01 -2.35623553e-01 -3.09751183e-01 4.51725781e-01
1.39604795e+00 4.28644866e-01 -2.28883505e-01 -6.15687549e-01
-1.98285604e+00 6.46341860e-01 1.10064745e+00 2.30838776e-01
-2.90240020e-01 3.65248382e-01 2.39239410e-01 5.93164265e-01
1.66435182e-01 9.06202137e-01 -5.44399023e-01 2.81743377e-01
-5.28325140e-01 3.76365691e-01 4.83112156e-01 1.01459694e+00
7.13159561e-01 2.78065443e-01 -5.47352917e-02 5.88899195e-01
-1.84553668e-01 5.26533186e-01 7.27575362e-01 -9.77414608e-01
1.08907387e-01 2.28701219e-01 -2.30882823e-01 -1.05343616e+00
-9.65209067e-01 -8.92607331e-01 -1.25449967e+00 1.14262938e-01
2.43556380e-01 1.22875862e-01 -1.08325934e+00 1.89601505e+00
2.34947875e-02 -2.27419123e-01 -3.76230180e-02 9.61316764e-01
6.07008219e-01 2.36178428e-01 -5.39788783e-01 2.02549562e-01
1.42424619e+00 -9.33912873e-01 -5.14542580e-01 -1.47789016e-01
5.91444790e-01 -1.06622386e+00 9.93906498e-01 1.05847940e-01
-1.10818779e+00 -5.69795370e-01 -1.44624436e+00 -5.98866999e-01
-5.40719986e-01 2.06175789e-01 6.68351710e-01 1.79560557e-01
-1.36989117e+00 6.38884008e-01 -1.11171556e+00 -4.55536991e-01
2.80518860e-01 6.16437614e-01 -8.35458279e-01 1.35810480e-01
-8.70841146e-01 9.41860139e-01 -1.25967011e-01 2.32819438e-01
-6.60618901e-01 -7.10468948e-01 -1.13979781e+00 3.47497426e-02
-4.43418473e-01 -7.99239099e-01 1.29375875e+00 -9.05111194e-01
-1.56254256e+00 8.72683465e-01 -1.00903213e-01 -5.31749666e-01
5.26127875e-01 -1.83912054e-01 -1.70401827e-01 1.67708602e-02
-7.96157867e-02 1.13469076e+00 1.12273359e+00 -8.49099696e-01
-1.44525468e-01 -4.32929546e-01 7.70746619e-02 3.06727648e-01
-3.22201103e-01 2.09343508e-02 -2.58929938e-01 -8.72084081e-01
8.07848930e-01 -1.26583004e+00 -1.81233689e-01 -1.83006838e-01
-3.67506444e-01 -4.40248139e-02 9.15361404e-01 -3.80625397e-01
4.00465071e-01 -1.94798481e+00 2.47932374e-01 1.36678293e-01
1.66656345e-01 -9.65460315e-02 -4.21857148e-01 8.28794986e-02
-6.95415914e-01 8.96017402e-02 2.10508242e-01 -2.08525643e-01
9.13930386e-02 1.73371732e-01 -6.26447499e-01 1.05707145e+00
7.06452131e-01 1.19748640e+00 -4.90943104e-01 1.89249545e-01
2.69736379e-01 1.01574302e+00 -9.40794289e-01 -2.44721606e-01
3.27242136e-01 4.04527009e-01 2.18480378e-01 5.25448501e-01
6.04510128e-01 -2.28282154e-01 -3.07956114e-02 -7.19202280e-01
-2.34978020e-01 4.53225732e-01 -1.01895893e+00 1.71381271e+00
-4.38466519e-01 9.63313997e-01 -2.96286196e-01 -1.24291253e+00
7.83423424e-01 1.11414671e-01 4.84888792e-01 -8.12754512e-01
3.90815884e-01 4.77460437e-02 3.58057410e-01 1.30811594e-02
6.75147355e-01 2.57794470e-01 4.48352955e-02 3.32918108e-01
4.80771333e-01 -2.59730995e-01 -3.60857770e-02 -1.28669724e-01
8.15481544e-01 2.34655440e-01 2.69453615e-01 -3.90059352e-01
1.19568169e-01 -3.68526429e-01 3.34344685e-01 6.46900833e-01
5.71940877e-02 1.14656627e+00 4.40975606e-01 -1.13726211e+00
-1.52819562e+00 -8.88041139e-01 -4.87469286e-01 1.22298634e+00
-6.04127347e-02 -1.21908046e-01 -7.00689077e-01 -1.49198219e-01
-3.16850185e-01 -1.02745362e-01 -8.94244134e-01 -2.51090199e-01
-1.03881884e+00 -8.27837408e-01 6.18684173e-01 7.92437196e-01
5.11350513e-01 -8.55380595e-01 -4.63224918e-01 1.12426773e-01
1.26651540e-01 -1.35859990e+00 -4.41753507e-01 5.76876283e-01
-7.74799705e-01 -7.60707617e-01 -8.58914912e-01 -1.13627112e+00
1.09088552e+00 4.79630262e-01 1.00694168e+00 -3.82663399e-01
-3.19199771e-01 1.93441451e-01 -2.10953981e-01 -2.94928163e-01
8.21445435e-02 5.65010607e-01 4.08412218e-01 -1.74339771e-01
-2.30838303e-02 -9.26230133e-01 -7.92879760e-01 6.66346669e-01
-9.27752137e-01 -8.40886608e-02 8.20202112e-01 8.53170574e-01
8.02917182e-01 -3.90465707e-01 1.71472073e-01 -5.58992982e-01
3.62243354e-01 2.40886614e-01 -4.30206597e-01 -8.86944905e-02
-1.66369095e-01 3.63586605e-01 8.53501856e-01 -5.37200212e-01
-4.30232525e-01 3.56516629e-01 -8.78642723e-02 -5.04623473e-01
-6.93810284e-02 1.87276050e-01 3.89692076e-02 -3.78719747e-01
1.12688732e+00 -1.58751123e-02 -1.17678523e-01 -2.70736337e-01
7.63047814e-01 -1.88263543e-02 1.12489319e+00 -3.69191736e-01
1.18904650e+00 1.07778370e+00 4.10388350e-01 -8.17161798e-01
-4.77760375e-01 -1.50690213e-01 -9.95235384e-01 2.22430855e-01
9.14741457e-01 -9.98561203e-01 -6.99788868e-01 4.59278047e-01
-1.27601993e+00 -1.36161506e-01 -4.18783844e-01 6.78006053e-01
-5.57054222e-01 -1.08708672e-01 -5.49872518e-01 2.17344210e-01
-2.91273087e-01 -1.40102983e+00 1.10968959e+00 1.04511335e-01
-5.59435859e-02 -7.63931930e-01 3.66075784e-02 -2.18044296e-01
6.76005304e-01 1.13323368e-01 5.93289733e-01 -6.68032289e-01
-4.66254711e-01 -5.33214748e-01 -1.54931694e-01 3.05892438e-01
-6.39268458e-02 2.78262466e-01 -1.06679857e+00 -6.11969650e-01
-3.05696636e-01 -3.69142622e-01 1.09185159e+00 4.75656271e-01
1.18942547e+00 -4.86668706e-01 2.21707910e-01 1.31472242e+00
9.34179783e-01 -3.34490150e-01 6.71591580e-01 5.78425944e-01
9.99956369e-01 3.54273945e-01 -7.20776021e-02 5.56016006e-02
3.64726245e-01 7.06504881e-01 6.99139059e-01 -5.53299248e-01
-2.71564782e-01 9.34336800e-03 3.33494604e-01 9.13698673e-01
-7.46477604e-01 2.82767773e-01 -7.59093344e-01 4.15264100e-01
-1.62399113e+00 -1.00786948e+00 1.48473337e-01 2.12648153e+00
6.08543038e-01 1.24194264e-01 -1.29378825e-01 1.87078461e-01
3.91042054e-01 4.26730841e-01 -3.72400820e-01 -5.06005764e-01
-5.76224208e-01 5.77278435e-01 9.29329038e-01 4.87485409e-01
-1.52056742e+00 9.94277596e-01 6.70015097e+00 2.22019017e-01
-1.76867449e+00 -9.98341292e-02 4.22900677e-01 6.26411475e-03
-1.47269145e-01 -2.37743527e-01 -6.19565845e-01 -5.36341488e-01
5.11545300e-01 1.46983027e-01 7.40223110e-01 1.00128591e+00
-5.91734722e-02 7.66691327e-01 -1.39250851e+00 8.30664337e-01
2.94505924e-01 -1.57924294e+00 1.84859633e-01 7.31413439e-02
1.11270893e+00 6.69312060e-01 5.12072563e-01 2.61479169e-01
2.27130264e-01 -1.25848949e+00 7.93331087e-01 2.18811855e-01
8.66651535e-01 -8.54982078e-01 7.24689603e-01 -2.86836535e-01
-1.15519190e+00 2.23385453e-01 -9.65171933e-01 -1.13872765e-02
-2.66346335e-01 2.08675236e-01 -8.24685931e-01 3.89604062e-01
9.31847155e-01 9.90120530e-01 -7.36496985e-01 5.95872402e-01
-2.47449234e-01 3.48197334e-02 -3.49934876e-01 2.84622639e-01
2.21542686e-01 3.94145958e-02 3.34660858e-01 1.27388012e+00
2.04313591e-01 -4.42494243e-01 -1.97274029e-01 6.50847673e-01
-5.46042383e-01 2.70575546e-02 -7.30124176e-01 2.59291381e-01
1.35554254e-01 1.52387714e+00 -8.42135012e-01 -1.75921500e-01
-3.79849911e-01 1.02181315e+00 3.56197029e-01 6.07786596e-01
-7.15406299e-01 -6.45210326e-01 9.09402668e-01 -9.75633860e-02
4.76533234e-01 -5.48672080e-01 -1.59010231e-01 -1.39932752e+00
7.63218775e-02 -8.98581624e-01 -1.14194825e-01 -6.49338782e-01
-7.96641767e-01 7.04550266e-01 -1.83858261e-01 -1.42838883e+00
-3.26503128e-01 -1.20893264e+00 -5.38476646e-01 5.97084343e-01
-1.40491450e+00 -1.35977471e+00 -4.32242841e-01 7.97048271e-01
3.07810307e-01 -1.07139289e-01 9.10945773e-01 2.37403736e-01
-3.91630054e-01 8.25680196e-01 -1.47873729e-01 6.17385328e-01
9.77663457e-01 -1.21699727e+00 9.07899559e-01 9.73530769e-01
4.95299816e-01 1.20078325e+00 7.00891316e-01 2.08232015e-01
-1.75111520e+00 -1.30976319e+00 7.01456547e-01 -4.59951162e-01
5.79415560e-01 -5.89598596e-01 -5.98720431e-01 1.13179946e+00
2.31551349e-01 6.66417122e-01 2.00019389e-01 5.99995293e-02
-1.05627906e+00 -2.30986431e-01 -6.61083758e-01 8.48755479e-01
1.14330018e+00 -7.83242702e-01 -4.19007540e-01 5.18470645e-01
9.35537159e-01 -8.47864926e-01 -8.25761318e-01 6.70784056e-01
8.92025054e-01 -7.44288027e-01 1.29945815e+00 -1.07206798e+00
4.16965663e-01 -3.87712449e-01 -1.96023628e-01 -1.34104741e+00
-7.49577820e-01 -6.26996756e-01 2.99412370e-01 4.18112427e-01
6.00489259e-01 -7.78050125e-01 3.47217113e-01 3.04228496e-02
-4.70985740e-01 -5.35408974e-01 -8.87156069e-01 -8.24018478e-01
3.82626951e-01 -3.19100678e-01 6.65868521e-01 1.37954605e+00
-1.57556131e-01 2.34387368e-01 -2.30847254e-01 3.48107487e-01
2.08371207e-01 -7.23724291e-02 8.82670939e-01 -8.21273148e-01
-1.94177811e-03 -8.13065290e-01 -9.51670647e-01 -1.07957172e+00
2.06069916e-01 -1.12427604e+00 -2.89490104e-01 -1.10723746e+00
-9.79772396e-03 3.34270984e-01 -2.30143949e-01 8.39205444e-01
4.05077010e-01 9.39807236e-01 4.63528410e-02 3.01668912e-01
-3.65408629e-01 4.15917546e-01 1.58057117e+00 -2.00305864e-01
-1.58188462e-01 -2.51521260e-01 -5.45510113e-01 7.94178486e-01
9.35219169e-01 -2.27032930e-01 -3.21156442e-01 -8.13057244e-01
6.63456082e-01 -6.66045725e-01 4.67460483e-01 -9.95260596e-01
2.20727459e-01 7.11157843e-02 8.33688974e-01 -3.77600223e-01
1.50640532e-01 -6.06611133e-01 -5.68060167e-02 4.21104074e-01
-4.46517438e-01 6.53273404e-01 1.31499069e-02 8.85642245e-02
-2.15541139e-01 4.64302868e-01 1.00828421e+00 5.94870783e-02
-1.89456031e-01 5.98534763e-01 -1.92811728e-01 -1.88879147e-01
3.58492732e-01 7.15671480e-02 -7.93644488e-01 -3.77632886e-01
-6.84702575e-01 -3.63395870e-01 5.80064416e-01 7.06632435e-01
4.13421333e-01 -1.54660153e+00 -6.40858769e-01 5.28269589e-01
1.19801141e-01 3.45318317e-01 -1.08088456e-01 1.02019882e+00
-1.04915643e+00 5.68713188e-01 -6.69248581e-01 -8.25984836e-01
-8.82103205e-01 4.07796443e-01 6.29021287e-01 4.77701128e-02
-7.13652849e-01 7.27232158e-01 3.07098389e-01 -7.55938768e-01
7.41672888e-02 -8.76621723e-01 -2.06614092e-01 -1.43193841e-01
3.84163350e-01 -1.78015009e-02 5.96595109e-01 -8.82377684e-01
-4.03324574e-01 7.86388636e-01 -1.79286376e-01 -2.97000766e-01
1.54923642e+00 3.30395669e-01 -3.43658090e-01 -9.26062912e-02
1.64096010e+00 -1.81115314e-01 -1.03552175e+00 -3.14161867e-01
-5.15844941e-01 5.27365459e-03 -6.96021542e-02 -1.50372460e-01
-1.24921012e+00 9.31784630e-01 4.56112921e-01 8.24374035e-02
8.97098958e-01 -7.05079362e-02 7.88856447e-02 1.25588143e+00
-8.49782526e-02 -8.45953465e-01 8.41706321e-02 9.46063161e-01
1.43856680e+00 -1.07088661e+00 1.06911480e-01 6.41928762e-02
-1.71156064e-01 1.36543345e+00 5.47593653e-01 -7.05208480e-01
6.71002567e-01 2.92217076e-01 2.30476141e-01 -1.55097499e-01
-5.09397805e-01 -4.82411459e-02 5.67889929e-01 5.06409049e-01
6.89402521e-01 7.38193914e-02 -1.30060822e-01 1.91085890e-01
-9.41667855e-01 -7.57366419e-01 5.21839738e-01 7.63870001e-01
-1.32222354e-01 -8.01566839e-01 -3.75589907e-01 -3.85459065e-02
-5.10593534e-01 4.68836464e-02 -3.81117582e-01 1.01729965e+00
-1.12934463e-01 2.30468675e-01 4.24899876e-01 -4.64950234e-01
4.37257499e-01 -2.39591286e-01 6.83887959e-01 -3.32900196e-01
-2.59766132e-01 2.69650906e-01 -4.05016124e-01 -8.24391067e-01
-6.28706098e-01 -4.13001657e-01 -1.22313762e+00 -3.96152556e-01
2.07236230e-01 -5.37118196e-01 1.01111162e+00 6.81848466e-01
1.55771270e-01 7.47187197e-01 8.52796674e-01 -1.39472961e+00
-5.28617799e-01 -1.02208781e+00 -1.22742616e-01 4.48836982e-01
6.28252327e-01 -3.55745196e-01 -4.65382665e-01 2.64982164e-01] | [8.909558296203613, 2.32755970954895] |
f5480e98-965a-4958-b5d5-a89c65338713 | cocatt-a-cognitive-conditioned-driver | 2111.10014 | null | https://arxiv.org/abs/2111.10014v2 | https://arxiv.org/pdf/2111.10014v2.pdf | CoCAtt: A Cognitive-Conditioned Driver Attention Dataset | The task of driver attention prediction has drawn considerable interest among researchers in robotics and the autonomous vehicle industry. Driver attention prediction can play an instrumental role in mitigating and preventing high-risk events, like collisions and casualties. However, existing driver attention prediction models neglect the distraction state and intention of the driver, which can significantly influence how they observe their surroundings. To address these issues, we present a new driver attention dataset, CoCAtt (Cognitive-Conditioned Attention). Unlike previous driver attention datasets, CoCAtt includes per-frame annotations that describe the distraction state and intention of the driver. In addition, the attention data in our dataset is captured in both manual and autopilot modes using eye-tracking devices of different resolutions. Our results demonstrate that incorporating the above two driver states into attention modeling can improve the performance of driver attention prediction. To the best of our knowledge, this work is the first to provide autopilot attention data. Furthermore, CoCAtt is currently the largest and the most diverse driver attention dataset in terms of autonomy levels, eye tracker resolutions, and driving scenarios. | ['Katie Driggs-Campbell', 'Peter Du', 'Aamir Hasan', 'Pranav Sriram', 'Niviru Wijayaratne', 'Yuan Shen'] | 2021-11-19 | null | null | null | null | ['driver-attention-monitoring'] | ['computer-vision'] | [-2.97187239e-01 7.60718249e-04 -4.48613793e-01 -2.22384974e-01
-1.09686181e-01 -9.49289128e-02 5.03513217e-01 5.83265536e-02
-4.82323676e-01 2.56749123e-01 2.14798361e-01 -5.28230667e-01
-1.14665702e-01 -2.26118803e-01 -3.09801728e-01 -3.50660354e-01
6.38205588e-01 5.08825528e-03 5.22359431e-01 -4.71270502e-01
3.55482072e-01 5.02209246e-01 -2.25147343e+00 -2.04966411e-01
7.44537234e-01 8.01361382e-01 5.99297702e-01 8.52406919e-01
2.21787766e-01 1.06308520e+00 -3.96707863e-01 -2.05543190e-01
-1.76421478e-02 -2.26995140e-01 -3.83200556e-01 -3.58488411e-02
4.20615673e-01 -1.62973478e-01 -6.14218771e-01 7.19705284e-01
5.79740405e-01 2.85529137e-01 2.61793196e-01 -1.81336761e+00
-7.84538746e-01 -2.28635252e-01 -3.94013435e-01 8.33908796e-01
2.03885496e-01 5.96137106e-01 4.56945449e-01 -7.09576964e-01
1.69801593e-01 1.16760802e+00 3.37889791e-01 6.39861941e-01
-8.31115007e-01 -8.92006576e-01 2.97355294e-01 9.27890182e-01
-1.22758484e+00 -6.30318582e-01 7.43507266e-01 -6.78962588e-01
9.89328504e-01 3.13762784e-01 7.40431488e-01 1.01714373e+00
5.89941323e-01 9.84495223e-01 7.39123166e-01 -2.47456864e-01
7.97075182e-02 2.69115359e-01 5.75217843e-01 2.46457040e-01
1.13053486e-01 5.37832856e-01 -7.04436004e-01 4.13103223e-01
2.02510938e-01 2.68377960e-01 -9.81215462e-02 -8.01161751e-02
-9.42744553e-01 5.34750223e-01 2.92960554e-01 -1.51424721e-01
-6.21609211e-01 -8.73757526e-02 2.01830193e-01 7.75356889e-02
3.96211952e-01 1.79366723e-01 -1.68121755e-01 -5.26151896e-01
-3.17235678e-01 7.78231695e-02 4.37187672e-01 1.38650286e+00
7.80215204e-01 6.38152510e-02 -7.23944128e-01 6.07097089e-01
2.35424072e-01 6.24263108e-01 1.54180840e-01 -9.47833002e-01
1.72716439e-01 5.55334091e-01 2.24675268e-01 -7.91186154e-01
-5.69476187e-01 -3.40189427e-01 -1.61350429e-01 4.09659058e-01
1.19610645e-01 -1.71820983e-01 -7.37899840e-01 1.60826516e+00
2.07262158e-01 3.44425082e-01 -1.83700532e-01 1.24010134e+00
9.06886220e-01 2.07123190e-01 5.06377101e-01 -1.30744621e-01
1.71450233e+00 -1.02754188e+00 -1.32459593e+00 -8.70357454e-01
4.52001929e-01 -8.79302263e-01 1.02423728e+00 2.36521646e-01
-1.06010711e+00 -1.04897857e+00 -8.23458314e-01 -4.27574426e-01
-2.40197927e-01 1.04127236e-01 6.63552165e-01 4.98268813e-01
-7.79467702e-01 -3.70818764e-01 -6.80492520e-01 -5.13589025e-01
4.26688373e-01 1.14115968e-01 -2.13981554e-01 -1.08696803e-01
-1.08241487e+00 1.54036438e+00 -1.15577698e-01 1.04298778e-01
-8.02064478e-01 -8.06481361e-01 -9.45832491e-01 6.33415207e-02
7.28035331e-01 -3.81523967e-01 1.59603882e+00 -5.37418067e-01
-1.12694156e+00 5.98468482e-01 -6.62112057e-01 -6.15501165e-01
2.16204301e-01 -6.19330466e-01 -8.35600853e-01 -4.61627573e-01
5.78065850e-02 8.08885992e-01 7.38730550e-01 -7.61137366e-01
-1.06555343e+00 -2.40436748e-01 1.03862956e-01 7.14619830e-02
1.51633648e-02 2.60709018e-01 -7.65409112e-01 -1.91511121e-02
-5.13957143e-01 -1.11913049e+00 -7.51282414e-03 -6.63505867e-02
-1.22647569e-01 -5.26579738e-01 1.14993393e+00 -3.25807124e-01
1.47035205e+00 -2.49437690e+00 -1.25473320e-01 -5.15420318e-01
6.18045330e-01 5.80983520e-01 -2.65888926e-02 -1.14525133e-03
5.21212481e-02 -2.34011307e-01 6.35116458e-01 -5.63722610e-01
4.86227125e-02 4.94398773e-01 -2.08909452e-01 3.70193332e-01
4.08079922e-01 1.01848865e+00 -7.88429320e-01 -5.53597331e-01
8.23919833e-01 5.77083945e-01 -6.63296640e-01 4.70845729e-01
1.33290917e-01 5.93328476e-01 -1.82482585e-01 3.64032626e-01
4.19846147e-01 3.39853406e-01 -7.18142748e-01 -1.03129245e-01
-6.87151432e-01 3.78035367e-01 -5.65221786e-01 1.16968119e+00
-4.61170256e-01 1.34486282e+00 -1.28378958e-01 -2.82950670e-01
8.85351479e-01 2.43563935e-01 1.72752246e-01 -1.14753640e+00
5.11485100e-01 -4.52068865e-01 5.05218387e-01 -8.39723647e-01
9.93085742e-01 3.15433770e-01 -3.33451629e-02 2.81514347e-01
-2.32588634e-01 3.19173843e-01 2.91006237e-01 -7.58150592e-02
7.65769958e-01 -3.16747695e-01 2.61715412e-01 -1.45093396e-01
5.25415003e-01 -8.14851448e-02 6.98240519e-01 8.01447988e-01
-1.00254405e+00 2.84254074e-01 3.61008137e-01 -4.77857620e-01
-7.15943694e-01 -6.20824397e-01 4.89316732e-02 1.52757668e+00
3.92737508e-01 -3.39103162e-01 -8.64915788e-01 -1.35944471e-01
-7.26290792e-02 1.18619514e+00 -7.30150402e-01 -7.61286557e-01
-3.18333060e-01 -1.61383543e-02 1.95530266e-01 7.76775181e-01
2.75125265e-01 -1.22126269e+00 -1.01404297e+00 -1.60978824e-01
-3.47292513e-01 -1.26559985e+00 -8.38455737e-01 -1.25779491e-02
-1.10625155e-01 -1.27372980e+00 -1.08286217e-01 -5.71135163e-01
3.51750016e-01 8.28871131e-01 9.67107654e-01 1.13602482e-01
-3.49910110e-01 5.62078357e-01 -2.20626257e-02 -1.32515252e+00
-2.08963305e-01 1.56543836e-01 1.17896937e-01 1.26154587e-01
1.34967864e+00 1.84452251e-01 -6.78030193e-01 6.65976584e-01
-2.10581571e-01 2.31502712e-01 5.43062627e-01 2.79402196e-01
4.79506195e-01 -6.72093555e-02 4.55481052e-01 -3.26407611e-01
7.44793236e-01 -5.54713130e-01 -6.41535044e-01 -1.27051175e-01
-6.60126448e-01 -9.30183157e-02 5.67055605e-02 -5.19870400e-01
-1.07730496e+00 -8.59420598e-02 9.89089385e-02 -6.52078450e-01
-6.43160045e-01 6.58338517e-02 -1.49636507e-01 1.79412454e-01
5.15848577e-01 -5.21786287e-02 2.56811261e-01 -3.02132636e-01
3.72604914e-02 7.83506453e-01 6.90450728e-01 1.08067013e-01
2.06046984e-01 1.47421315e-01 -2.03079190e-02 -9.09510195e-01
-1.00810552e+00 -8.38634670e-01 -5.71586609e-01 -6.83735251e-01
1.10941398e+00 -8.96629572e-01 -1.22791469e+00 2.85595983e-01
-1.01938236e+00 -1.83832914e-01 -3.15634698e-01 7.83409774e-01
-5.39659142e-01 -1.53478950e-01 -1.72192946e-01 -1.17208076e+00
3.71952057e-02 -1.46539795e+00 9.87518311e-01 3.67050648e-01
-3.87056231e-01 -5.42568862e-01 -2.65210003e-01 5.36428392e-01
5.85056543e-01 -2.68889666e-01 1.52515829e-01 -2.30151922e-01
-6.40418231e-01 -3.03290248e-01 -2.05792412e-01 9.54762995e-02
1.34586006e-01 -9.55403969e-02 -1.12601161e+00 3.93954925e-02
8.15172214e-03 7.45982528e-02 6.67978942e-01 5.84814131e-01
9.50801671e-01 3.35812807e-01 -4.56897289e-01 2.23008648e-01
5.59718966e-01 3.68862033e-01 9.65318143e-01 2.95994937e-01
5.94781578e-01 6.69189990e-01 1.05886972e+00 3.59095991e-01
8.43274593e-01 8.33472013e-01 7.30504513e-01 -1.87369660e-01
-5.36332846e-01 -3.92412208e-02 2.32663170e-01 3.42098117e-01
-1.70194238e-01 -3.63321155e-01 -9.91550684e-01 7.82702923e-01
-1.93208408e+00 -1.14923465e+00 -5.27689517e-01 2.05406737e+00
2.55290329e-01 2.26152316e-01 1.99587747e-01 9.23335552e-02
7.89838016e-01 -1.19250692e-01 -7.84131408e-01 -6.28840864e-01
1.73880711e-01 -3.56570333e-01 6.13759995e-01 6.36109233e-01
-8.88616562e-01 8.59190524e-01 6.38860893e+00 3.74547809e-01
-9.98404682e-01 3.20438862e-01 7.70588219e-02 -5.77702522e-01
2.99388021e-01 -1.74954548e-01 -1.22736609e+00 6.51065648e-01
1.22050166e+00 -2.37594366e-01 2.76138127e-01 9.45221663e-01
5.73110104e-01 -3.92866164e-01 -1.14295363e+00 1.06348479e+00
2.52968967e-01 -6.28278255e-01 -6.89690113e-01 3.53724748e-01
-3.90342809e-02 2.28693083e-01 3.22262406e-01 7.34540880e-01
-5.26365712e-02 -9.08474922e-01 9.77728307e-01 9.59195495e-01
4.16948825e-01 -9.52674210e-01 8.06345224e-01 6.05086029e-01
-1.02183068e+00 -3.46702784e-01 -2.70412177e-01 -5.28324485e-01
5.17319143e-01 -1.39910243e-02 -6.61747217e-01 -2.93378979e-01
7.39657402e-01 7.79543400e-01 -7.95910835e-01 1.05079281e+00
-2.20006466e-01 6.42021298e-01 5.46512082e-02 1.45227101e-03
1.26875639e-02 1.68422163e-01 5.82430422e-01 1.05498159e+00
1.65679589e-01 2.00111657e-01 1.07523397e-01 7.73353934e-01
3.38021964e-01 -5.39716244e-01 -6.66691124e-01 2.51759827e-01
4.77335513e-01 1.05211711e+00 -2.33076084e-02 -4.47432399e-02
-8.05972993e-01 3.47021759e-01 1.65284514e-01 4.23733503e-01
-1.24715710e+00 -2.56300062e-01 1.55807340e+00 5.30952930e-01
2.28937030e-01 -2.65182465e-01 -2.44991615e-01 -5.51640213e-01
-1.58355221e-01 -2.45486528e-01 1.33897573e-01 -1.20829344e+00
-6.85771406e-01 5.31670988e-01 1.00777738e-01 -1.31852353e+00
-1.50604516e-01 -6.07368529e-01 -4.43799019e-01 1.12807012e+00
-1.89945805e+00 -8.95191669e-01 -8.30965221e-01 5.71038544e-01
8.70656133e-01 -3.07368279e-01 3.38330597e-01 4.30959284e-01
-1.02981400e+00 6.37263536e-01 -7.88818717e-01 -2.81665593e-01
9.22621846e-01 -7.85651088e-01 3.78066629e-01 8.91665280e-01
-4.92013991e-01 3.33174288e-01 1.02348363e+00 -4.34148371e-01
-1.40154707e+00 -1.15053797e+00 9.98520851e-01 -1.03021884e+00
4.07814264e-01 -1.71237141e-01 -1.02868819e+00 9.49605882e-01
4.88702416e-01 5.93123399e-02 6.44331336e-01 1.40231848e-01
1.35129899e-01 -6.60200641e-02 -7.06632018e-01 8.55960965e-01
9.95385110e-01 -6.55545175e-01 -7.07234263e-01 1.64644737e-02
6.72961712e-01 -6.43147588e-01 -5.13502955e-01 8.83719474e-02
5.20166934e-01 -1.00526083e+00 9.10740018e-01 -5.73753595e-01
3.32487226e-02 -3.64937216e-01 2.59385079e-01 -1.17601097e+00
-8.44095230e-01 -4.45433736e-01 -3.87695789e-01 1.01718593e+00
6.62843287e-02 -7.33509958e-01 3.51721674e-01 1.13546443e+00
-6.40453815e-01 -3.87754321e-01 -6.26346946e-01 -5.88184178e-01
-4.59470332e-01 -1.02304101e+00 8.63091290e-01 3.29532593e-01
7.15823621e-02 5.14722228e-01 -6.55013919e-01 2.86050558e-01
3.17309797e-01 -3.21102887e-01 1.14152145e+00 -1.41032732e+00
3.84252578e-01 -4.43655252e-01 -6.81667089e-01 -1.16341150e+00
3.32960278e-01 -3.52683872e-01 3.71150225e-01 -1.24824572e+00
-2.90542003e-02 -1.39702871e-01 -4.36090946e-01 4.78230536e-01
-3.88926327e-01 -7.70990690e-03 1.27374098e-01 1.68270737e-01
-6.86397672e-01 4.44688767e-01 1.28827071e+00 2.05604166e-01
-1.47232160e-01 1.97305940e-02 -1.01646149e+00 6.08906150e-01
1.15713108e+00 -2.46703595e-01 -5.97343922e-01 -6.53700292e-01
-1.05761737e-01 -8.17997605e-02 5.36367595e-01 -1.03280127e+00
8.67545009e-01 -5.41687071e-01 -4.68690172e-02 -9.03140306e-01
6.38991475e-01 -9.54273045e-01 7.52364770e-02 1.81008905e-01
-2.79637337e-01 1.04002289e-01 7.94618547e-01 5.73593915e-01
-1.23920687e-01 9.35798064e-02 6.32931471e-01 5.01332104e-01
-1.16100073e+00 3.80838364e-01 -9.17257249e-01 2.74842605e-03
1.41725481e+00 -3.84108812e-01 -7.17716873e-01 -3.72099638e-01
-5.52947462e-01 5.58412015e-01 2.37530932e-01 1.23164546e+00
5.06313264e-01 -1.26774180e+00 -5.27330518e-01 6.55488193e-01
6.16395712e-01 -9.07563046e-02 3.65648776e-01 1.36933208e+00
1.62879959e-01 9.25363600e-01 -3.11899155e-01 -9.08084571e-01
-1.33304667e+00 1.02906775e+00 2.25617260e-01 5.54200411e-01
-5.85100293e-01 3.40659738e-01 5.29367805e-01 1.14958018e-01
1.98436961e-01 -3.18819314e-01 -6.08392119e-01 -1.06803596e-01
8.99760365e-01 6.46202326e-01 8.82357582e-02 -1.22597051e+00
-3.60425532e-01 2.10539177e-01 -2.16896579e-01 3.65583509e-01
6.36199832e-01 -5.90576053e-01 4.38695848e-01 3.55576605e-01
5.27076483e-01 -2.06751734e-01 -1.64158392e+00 -1.20822594e-01
-2.43674040e-01 -6.06976807e-01 6.75380349e-01 -6.28529727e-01
-1.03948331e+00 1.04838347e+00 7.02897787e-01 1.93511233e-01
1.03620732e+00 2.45852217e-01 7.97022700e-01 1.22355759e-01
2.09568933e-01 -1.08539808e+00 -2.33708024e-01 6.12023890e-01
9.53342676e-01 -1.56015968e+00 -4.05212551e-01 -3.51727962e-01
-1.01379669e+00 3.53253961e-01 1.23316050e+00 3.33677381e-01
6.42007649e-01 2.45243952e-01 2.77478009e-01 -2.73557574e-01
-1.16542244e+00 -8.22558463e-01 6.68387294e-01 7.74262846e-01
3.18356931e-01 2.76723076e-02 8.46661553e-02 5.89769185e-01
-3.45016748e-01 1.92606390e-01 3.67653877e-01 8.89522851e-01
-6.29346490e-01 -5.15124917e-01 -3.96197915e-01 3.56331080e-01
-9.74447429e-02 1.59576491e-01 -1.69576675e-01 7.60465026e-01
3.61156315e-01 1.49624467e+00 5.19400299e-01 -4.33973879e-01
9.40888286e-01 1.98891703e-02 2.12449022e-02 -5.14913738e-01
-3.24490458e-01 -2.89746314e-01 -9.64879319e-02 -7.15439975e-01
-3.26528579e-01 -6.72059417e-01 -8.63473713e-01 -5.99110126e-01
-4.47228462e-01 -1.12423539e-01 5.60331106e-01 9.82314646e-01
8.18973184e-01 1.11607206e+00 1.32425129e-01 -9.63931143e-01
1.40220210e-01 -1.01837957e+00 -2.52780169e-01 1.79852009e-01
7.29126096e-01 -1.26977038e+00 -2.28908017e-01 -4.97629195e-02] | [7.5804338455200195, -0.061216700822114944] |
494002ed-80ce-4a2f-9940-f5a2409e9101 | license-plate-recognition-system-based-on | 1506.03128 | null | http://arxiv.org/abs/1506.03128v1 | http://arxiv.org/pdf/1506.03128v1.pdf | License Plate Recognition System Based on Color Coding Of License Plates | License Plate Recognition Systems are used to determine the license plate
number of a vehicle. The current system mainly uses Optical Character
Recognition to recognize the number plate. There are several problems to this
system. Some of them include interchanging of several letters or numbers
(letter O with digit 0), difficulty in localizing the license plate, high error
rate, use of different fonts in license plates etc. So a new system to
recognize the license plate number using color coding of license plates is
proposed in this paper. Easier localization of license plate can be done by
searching for the start or stop patters of license plates. An eight segment
display system along with traditional numbering with the first and last
segments left for start or stop patterns is proposed in this paper. Practical
applications include several areas under Internet of Things (IoT). | ['Jani Biju Babjan'] | 2015-06-08 | null | null | null | null | ['license-plate-recognition'] | ['computer-vision'] | [-1.56677678e-01 -9.85540926e-01 -6.13775700e-02 -1.26588970e-01
5.65863699e-02 -1.17434156e+00 2.20336869e-01 -6.48505926e-01
-2.35233143e-01 5.83109081e-01 -4.17941540e-01 -6.42123282e-01
3.18145484e-01 -6.89314663e-01 -2.07753062e-01 -4.74248230e-01
6.74105406e-01 4.81000155e-01 9.20714438e-01 -2.29593277e-01
1.12838519e+00 1.06773281e+00 -1.15402639e+00 8.16621631e-02
5.26457191e-01 5.91860414e-01 2.92222589e-01 6.95899606e-01
-4.64498490e-01 5.21495998e-01 -5.74735165e-01 -3.36473405e-01
8.30078483e-01 -2.85314322e-01 -1.12798519e-01 4.74874645e-01
1.54976547e-01 -7.01906681e-01 -4.68947172e-01 1.25071943e+00
6.46347851e-02 -1.19701616e-01 1.08072388e+00 -1.46374738e+00
-6.72858179e-01 -2.55908549e-01 -7.51402795e-01 2.35127136e-01
2.89202094e-01 -7.94240162e-02 1.71245068e-01 -1.11511719e+00
3.89560431e-01 7.79168189e-01 9.31202888e-01 3.69432598e-01
-2.46631488e-01 -8.86107504e-01 -8.84940982e-01 5.45742631e-01
-1.87162435e+00 -3.63389224e-01 4.99663949e-01 -6.23523772e-01
9.43171740e-01 3.69027436e-01 2.05615774e-01 -4.78033692e-01
5.09360135e-01 2.91568607e-01 1.30898499e+00 -7.28258550e-01
-1.80747986e-01 6.41812265e-01 5.90893686e-01 6.28772974e-01
5.25219798e-01 -3.54242176e-01 3.92377794e-01 4.02271122e-01
1.17711735e+00 6.51316285e-01 3.74454498e-01 1.38250574e-01
-6.23113453e-01 5.92667162e-01 -2.55181372e-01 3.97785962e-01
8.20111409e-02 -3.20079148e-01 -8.64224732e-02 9.24830213e-02
-4.93666947e-01 6.62027597e-02 -1.87308323e-02 -3.09116006e-01
-9.61277306e-01 -2.27324605e-01 6.19414389e-01 1.44587886e+00
8.39002967e-01 1.43394396e-01 3.77848536e-01 8.59559119e-01
5.41143119e-01 9.14748788e-01 2.74723977e-01 -4.00947005e-01
8.98396790e-01 6.27601326e-01 4.30710256e-01 -8.88775647e-01
-3.52139682e-01 6.09888852e-01 -1.90639466e-01 9.52021778e-01
5.02845466e-01 -5.83024442e-01 -1.00207245e+00 1.30800277e-01
-4.46481556e-01 -1.61139164e-02 1.32657677e-01 5.48101962e-01
7.57865667e-01 1.24739099e+00 -4.92709368e-01 2.20300093e-01
1.66812098e+00 -1.10098004e+00 -8.32176507e-01 -3.54189396e-01
2.59144396e-01 -1.31582594e+00 4.53468978e-01 1.01297930e-01
-7.30572462e-01 -4.09148574e-01 -1.17970920e+00 1.44817486e-01
-9.43519592e-01 7.72877157e-01 -4.16694395e-02 1.13861251e+00
-9.20830309e-01 -1.09533563e-01 -1.75507620e-01 -4.13556933e-01
-7.36197457e-02 7.07161367e-01 -3.01535994e-01 -1.77908421e-01
-7.66917050e-01 1.25011837e+00 -3.13321361e-03 -9.37557220e-03
4.52250183e-01 1.86257094e-01 -4.16721582e-01 -1.67676121e-01
-6.26977980e-02 3.34699035e-01 7.97325015e-01 -1.04627967e+00
-1.79364157e+00 7.96714306e-01 -1.65048584e-01 3.85081142e-01
4.04910952e-01 3.62968385e-01 -1.41452396e+00 1.61226690e-01
5.08239418e-02 3.04993033e-01 5.47324598e-01 -7.21763551e-01
-1.05308974e+00 -2.50892550e-01 -4.91778374e-01 2.99042076e-01
3.63398045e-01 7.44736731e-01 -8.37338567e-01 -8.97864848e-02
1.95083827e-01 -1.11486113e+00 2.97639459e-01 -2.95237482e-01
-4.35490519e-01 -3.78075898e-01 1.37687922e+00 -1.00739157e+00
1.03234470e+00 -2.25878716e+00 -1.32344699e+00 6.76815927e-01
-1.80692494e-01 7.38915026e-01 3.63600552e-01 5.64690769e-01
1.58316150e-01 6.98454902e-02 2.77589202e-01 5.50431132e-01
-2.91006826e-02 -1.54598087e-01 3.04658003e-02 6.95473909e-01
3.84226181e-02 5.76699436e-01 1.72478855e-01 -4.30634379e-01
4.15578783e-01 3.38002920e-01 2.20422626e-01 -4.11834776e-01
8.73129487e-01 -2.46900231e-01 -3.94941956e-01 9.17659283e-01
1.44350219e+00 3.33533585e-01 -2.58294642e-01 1.04197478e-02
-9.59763467e-01 -1.97325677e-01 -1.46781838e+00 1.15283161e-01
1.34977221e-01 1.25019681e+00 -2.47165546e-01 -3.61945063e-01
1.35242856e+00 3.56498212e-01 8.03078562e-02 -6.23601317e-01
6.82959333e-02 4.82885599e-01 1.32046090e-02 -8.81037176e-01
6.92609668e-01 -1.95542350e-01 1.02016263e-01 7.32052606e-03
-5.64242542e-01 -4.61341739e-02 3.31707746e-01 -4.74124640e-01
4.83251035e-01 -3.07936966e-01 9.59667936e-02 -1.82907030e-01
7.76918709e-01 3.37426096e-01 3.36214006e-01 4.15318966e-01
-4.95572656e-01 5.02363801e-01 2.27963924e-01 -3.66187215e-01
-1.28906393e+00 -6.47495389e-01 -3.49074125e-01 5.02781570e-01
6.89397812e-01 3.74527603e-01 -4.53234971e-01 -5.04564382e-02
-8.28385428e-02 5.35716295e-01 8.50912780e-02 5.68025351e-01
-9.46617126e-01 -1.45484999e-01 7.06919491e-01 5.85182488e-01
8.77528548e-01 -7.69787788e-01 -3.14499736e-01 3.10462434e-02
3.92707914e-01 -1.08696425e+00 -7.20450044e-01 -3.59848350e-01
-7.01928318e-01 -8.55183244e-01 -1.00676966e+00 -1.56368911e+00
1.04702771e+00 7.82899618e-01 4.14035976e-01 3.83777060e-02
1.27749056e-01 1.32230401e-01 -4.70047295e-02 -6.42540276e-01
-4.21182394e-01 -5.17964840e-01 -1.11362562e-01 1.86898541e-02
8.90432358e-01 1.42319158e-01 -4.58562613e-01 9.27072525e-01
-6.14306808e-01 -2.56280750e-01 7.96919286e-01 -2.77495682e-01
2.98428945e-02 4.06305671e-01 1.12427123e-01 -3.23169410e-01
5.84026694e-01 -2.47641921e-01 -1.09109449e+00 2.30148017e-01
-4.92970556e-01 -4.99102950e-01 7.72045612e-01 1.38041511e-01
-6.88997746e-01 2.12094113e-01 -6.99775666e-02 6.82133734e-02
-4.06627655e-01 -7.45064020e-02 -3.47822130e-01 -6.01027846e-01
-2.73910891e-02 7.43235290e-01 3.98922712e-01 -3.98364931e-01
-3.41170013e-01 1.25726044e+00 6.57167733e-01 3.98690820e-01
1.01260066e+00 1.32130057e-01 4.49441262e-02 -1.00611913e+00
8.50904346e-01 -1.09133446e+00 -6.54349387e-01 -6.90542996e-01
9.10268664e-01 -5.92754304e-01 -9.10535157e-01 9.12301421e-01
-1.16538978e+00 4.60478723e-01 6.09699965e-01 7.15875566e-01
1.40487060e-01 4.52893347e-01 -5.24110198e-01 -7.55687296e-01
1.08861782e-01 -1.16532052e+00 4.51598346e-01 8.17207992e-01
4.20464486e-01 -8.56142879e-01 2.64777392e-02 2.81283498e-01
4.87826049e-01 -2.65248537e-01 6.08794510e-01 -7.55244792e-01
-6.30450845e-01 -9.84226823e-01 -5.65637589e-01 1.11008354e-01
2.52380878e-01 5.30768394e-01 -3.04554015e-01 3.50160837e-01
-3.22800547e-01 5.13441265e-01 4.86363888e-01 3.40842694e-01
1.40351966e-01 -3.53090942e-01 -5.59125543e-01 3.12918246e-01
1.99242043e+00 1.32833886e+00 1.43093932e+00 8.81748438e-01
6.98241591e-01 1.43094301e-01 3.29613745e-01 1.52095869e-01
2.21164420e-01 6.63661659e-01 -1.77736029e-01 1.44474611e-01
6.26200661e-02 4.54019234e-02 6.18133068e-01 6.83944166e-01
-5.02127588e-01 -2.14424357e-01 -1.15951371e+00 -1.04201078e-01
-9.13543344e-01 -1.26356006e+00 -7.75888622e-01 2.16267014e+00
3.05734694e-01 9.43540968e-03 3.97896379e-01 2.51436561e-01
1.52017152e+00 -5.51077485e-01 -1.34971231e-01 -8.06103706e-01
-9.89159048e-02 -4.84320313e-01 1.45195556e+00 7.14595079e-01
-8.99567187e-01 8.29171240e-01 6.56630707e+00 7.45633304e-01
-1.64676106e+00 -4.17991787e-01 3.54320079e-01 7.54413784e-01
2.33158991e-01 -2.44670007e-02 -1.42927861e+00 8.37467432e-01
5.47193348e-01 -1.30403414e-02 4.15511310e-01 5.07355630e-01
1.33221284e-01 -2.87190914e-01 -5.00815332e-01 1.25774491e+00
3.12919170e-01 -1.31869078e+00 -2.16681004e-01 1.67148173e-01
6.09445512e-01 -1.08334377e-01 1.19455103e-02 4.49893810e-02
-1.91136420e-01 -7.51441836e-01 8.14301550e-01 5.28535426e-01
9.80038464e-01 -6.49681032e-01 8.40102434e-01 2.04828903e-01
-1.14085603e+00 1.24222651e-01 -7.14517891e-01 -1.50674090e-01
-5.90315312e-02 -3.71887356e-01 -1.25477040e+00 -2.12271765e-01
4.21617068e-02 2.39691362e-01 -5.30420780e-01 1.63417470e+00
3.59793961e-01 4.82683301e-01 -5.83201408e-01 -5.67017496e-01
5.74875712e-01 -8.80301297e-01 4.09613878e-01 1.48146021e+00
9.15942729e-01 1.39809430e-01 -4.20043200e-01 5.09226143e-01
2.22567990e-01 3.10930699e-01 -6.77403152e-01 4.38252501e-02
6.99959993e-01 9.73299205e-01 -1.29277897e+00 -4.24763978e-01
-8.75244021e-01 8.25095296e-01 -8.59193087e-01 3.28399926e-01
-8.10185254e-01 -1.14117944e+00 -1.24013824e-02 4.70000565e-01
5.60507238e-01 -5.99784255e-01 -4.61776495e-01 -5.11684477e-01
-1.53745517e-01 -3.67789567e-01 3.68486196e-02 -8.17662060e-01
-5.38695693e-01 2.30215147e-01 -1.80721715e-01 -1.92926836e+00
3.38949338e-02 -1.23663807e+00 -1.04785538e+00 9.83376682e-01
-1.01856351e+00 -6.98334217e-01 1.24462545e-02 7.65525937e-01
7.65716493e-01 -9.18291867e-01 3.30020338e-01 3.45873296e-01
-6.19653285e-01 5.25430262e-01 1.14469123e+00 7.11456120e-01
5.48581719e-01 -9.68523860e-01 1.49929255e-01 9.89077449e-01
-3.84804040e-01 5.07863581e-01 4.65471655e-01 -8.68884146e-01
-1.33726561e+00 -4.96562868e-01 1.01112986e+00 -2.55925328e-01
2.70563275e-01 -6.00981414e-02 -4.09474373e-01 7.44024396e-01
4.04786825e-01 -3.03188801e-01 5.79245389e-01 -7.97083855e-01
-6.74380660e-02 -3.80156875e-01 -1.39506650e+00 6.50850296e-01
-1.78833827e-01 -1.41548365e-01 -4.42403942e-01 2.59272337e-01
-4.39116240e-01 -6.72791228e-02 -1.39958978e-01 -4.15192753e-01
8.14287841e-01 -8.39064658e-01 6.96044564e-01 1.70364678e-01
-6.77490830e-02 -9.91619587e-01 -1.29411146e-02 -6.11351609e-01
-5.92499912e-01 -3.92634600e-01 1.10015166e+00 1.34785998e+00
6.62875712e-01 -1.00190747e+00 7.86590397e-01 1.16261268e+00
-2.42242515e-01 1.51135206e-01 -8.65147412e-01 -8.74728441e-01
-2.15536028e-01 1.24497198e-01 4.01318014e-01 6.81179881e-01
2.02196717e-01 1.52322143e-01 -5.20753682e-01 3.55573237e-01
2.65208960e-01 -2.79706836e-01 4.45854753e-01 -1.02769268e+00
2.15234965e-01 -6.22081757e-01 -1.03148043e+00 -9.44636703e-01
-4.13347274e-01 -3.45822304e-01 1.82330422e-02 -1.58094418e+00
2.28317874e-03 -3.07411373e-01 -1.22809438e-02 2.11796954e-01
3.09721559e-01 6.77403212e-01 5.52363873e-01 7.65134215e-01
-4.80550319e-01 -6.37976229e-01 9.82621908e-01 -2.80047119e-01
-2.23095968e-01 5.38007915e-01 -1.38448924e-01 7.12592423e-01
1.10950887e+00 -3.45944405e-01 -9.98547394e-03 -2.36633852e-01
1.68999448e-01 8.69222730e-02 4.73371893e-03 -1.10476458e+00
7.83847511e-01 -8.23378041e-02 6.45906746e-01 -1.02420962e+00
1.45761356e-01 -1.26058829e+00 1.65569395e-01 4.56262618e-01
5.15803993e-01 7.67492056e-01 4.12565589e-01 5.27299754e-02
-2.48306230e-01 -1.14448869e+00 7.55574942e-01 -1.10874474e-01
-1.54934072e+00 -2.88985848e-01 -1.43077505e+00 -7.39165545e-01
1.36378074e+00 -1.43652248e+00 -3.66518706e-01 -2.88807184e-01
-9.03898925e-02 -3.51204485e-01 6.04054630e-01 1.92285508e-01
5.45643449e-01 -1.37495494e+00 -5.99768698e-01 3.79398942e-01
-2.94799451e-02 -9.86302733e-01 1.31709546e-01 5.59734762e-01
-1.82853293e+00 1.02341247e+00 -1.01413977e+00 -1.96006939e-01
-1.63174212e+00 1.85931429e-01 4.54985917e-01 3.96404594e-01
-3.53418440e-01 4.50014412e-01 -3.96630526e-01 -2.02234630e-02
-3.25023711e-01 -1.61937907e-01 -7.45814323e-01 -2.42114097e-01
6.55542552e-01 1.03933430e+00 -1.80487093e-02 -1.03330851e+00
-7.41783977e-01 1.50436842e+00 -4.47552651e-02 -2.94832706e-01
8.81170690e-01 -1.96050972e-01 -2.22903013e-01 1.53604433e-01
1.10871267e+00 7.52232492e-01 -1.24003565e+00 2.71802962e-01
-2.21403018e-02 -5.84330022e-01 -3.79098833e-01 -8.89281571e-01
-7.36681879e-01 4.31821793e-01 8.58148575e-01 2.61303127e-01
6.62956953e-01 -4.55888689e-01 5.60436964e-01 5.97791016e-01
2.46508867e-01 -1.56350303e+00 -6.68753862e-01 6.62858605e-01
5.83023906e-01 -1.06517947e+00 -2.34561086e-01 -3.62469882e-01
-8.99318933e-01 1.99401283e+00 4.22658712e-01 -3.18317354e-01
5.53110063e-01 6.28353417e-01 3.53192061e-01 1.33737698e-01
1.43582866e-01 3.65526937e-02 1.95835486e-01 6.96458220e-01
2.93484569e-01 1.42733589e-01 -5.91928899e-01 2.52042472e-01
8.33572149e-02 1.39394045e-01 1.06750798e+00 1.14749455e+00
-1.01520598e+00 -1.00912941e+00 -1.35493243e+00 4.11178231e-01
-6.22241735e-01 -1.13720328e-01 -4.12662357e-01 1.02828467e+00
2.56419271e-01 1.13546467e+00 3.21479052e-01 -4.52164054e-01
2.02934608e-01 3.57007176e-01 2.51062512e-01 1.53457314e-01
5.89281209e-02 9.19175744e-02 -2.87430175e-02 3.45052838e-01
-2.82200966e-02 -8.21184099e-01 -1.52588665e+00 -7.07545280e-01
-2.50752419e-01 1.73846349e-01 1.04420638e+00 9.05730367e-01
2.05666590e-02 -3.52169186e-01 6.09626234e-01 -3.57789367e-01
1.04855951e-02 -6.95828378e-01 -1.19675481e+00 -1.96263883e-02
8.79230723e-02 -2.15547815e-01 -8.56005102e-02 3.77622724e-01] | [9.797786712646484, -4.995074272155762] |
ea2a5173-cff4-449c-9641-530e6bb3b493 | self-knowledge-distillation-adversarial | null | null | https://openreview.net/forum?id=rJejta4KDS | https://openreview.net/pdf?id=rJejta4KDS | SELF-KNOWLEDGE DISTILLATION ADVERSARIAL ATTACK | Neural networks show great vulnerability under the threat of adversarial examples.
By adding small perturbation to a clean image, neural networks with high classification accuracy can be completely fooled.
One intriguing property of the adversarial examples is transferability. This property allows adversarial examples to transfer to networks of unknown structure, which is harmful even to the physical world.
The current way of generating adversarial examples is mainly divided into optimization based and gradient based methods.
Liu et al. (2017) conjecture that gradient based methods can hardly produce transferable targeted adversarial examples in black-box-attack.
However, in this paper, we use a simple technique to improve the transferability and success rate of targeted attacks with gradient based methods.
We prove that gradient based methods can also generate transferable adversarial examples in targeted attacks.
Specifically, we use knowledge distillation for gradient based methods, and show that the transferability can be improved by effectively utilizing different classes of information.
Unlike the usual applications of knowledge distillation, we did not train a student network to generate adversarial examples.
We take advantage of the fact that knowledge distillation can soften the target and obtain higher information, and combine the soft target and hard target of the same network as the loss function.
Our method is generally applicable to most gradient based attack methods. | ['Duan Zhenhua', 'Dong Zeqian', 'Tian Cong', 'Wang Renzhi[1]', 'Ma Xiaoxiong[1]'] | 2019-09-25 | null | null | null | null | ['self-knowledge-distillation'] | ['computer-vision'] | [ 3.87925774e-01 5.06468356e-01 1.04879461e-01 5.07708006e-02
-6.73424542e-01 -1.13510215e+00 5.15429974e-01 -4.89297420e-01
-3.66360426e-01 1.14787722e+00 -3.42885852e-01 -5.38433909e-01
-3.55532691e-02 -1.36207557e+00 -1.22506642e+00 -9.24067438e-01
-6.41059056e-02 -6.17845505e-02 2.39043623e-01 -5.46991229e-01
-2.86376141e-02 7.69533753e-01 -8.58720779e-01 1.37879089e-01
1.02273893e+00 5.59422493e-01 -4.08432811e-01 9.07869935e-01
1.62692845e-01 8.99238825e-01 -1.08859158e+00 -8.36729825e-01
8.19074512e-01 -7.33463883e-01 -7.68505394e-01 -5.40886581e-01
6.71447515e-01 -2.76129842e-01 -8.96855414e-01 1.55480659e+00
5.98163962e-01 2.14469120e-01 5.76245606e-01 -1.51548505e+00
-8.94243240e-01 8.79334509e-01 -6.21292815e-02 2.45985575e-02
1.68862417e-01 3.75071257e-01 5.69000363e-01 -5.35652816e-01
4.42974061e-01 1.40792942e+00 5.95693052e-01 1.14042056e+00
-1.04327965e+00 -1.08968735e+00 3.69865559e-02 1.85694829e-01
-1.07698441e+00 -1.05003059e-01 9.07099068e-01 -2.25144088e-01
3.93235177e-01 5.32810986e-01 3.89776111e-01 1.43772840e+00
7.93827921e-02 7.59463072e-01 1.30821788e+00 -3.78418833e-01
2.60213464e-01 4.32611823e-01 -2.69372016e-01 7.15581656e-01
4.02729601e-01 8.10796678e-01 -1.72767863e-01 -2.94920325e-01
5.57624280e-01 -3.85764763e-02 -6.98388278e-01 -1.89593256e-01
-1.02499485e+00 1.02847457e+00 1.00965023e+00 1.04799777e-01
1.42962903e-01 3.75558108e-01 2.14095414e-01 7.15624213e-01
1.83588460e-01 1.08954787e+00 -4.04372841e-01 2.94617921e-01
-4.45733398e-01 3.24726731e-01 1.12543058e+00 7.18208790e-01
7.81594813e-01 5.99922895e-01 -1.64191172e-01 3.59352618e-01
-1.21253185e-01 9.34073806e-01 4.72735107e-01 -7.92687356e-01
5.11866510e-01 9.57625136e-02 -1.95379734e-01 -9.45564091e-01
1.84783503e-01 -5.12511492e-01 -9.00559604e-01 8.85628104e-01
5.76534450e-01 -7.62818515e-01 -1.02400124e+00 1.98244750e+00
2.43288025e-01 4.25148547e-01 5.18974125e-01 5.69274783e-01
5.67920923e-01 6.81447327e-01 -3.21265072e-01 1.66573003e-01
6.43517673e-01 -8.73258948e-01 -4.03596967e-01 -2.44268015e-01
6.32305920e-01 -3.67996991e-01 8.70283842e-01 1.99375793e-01
-1.05871439e+00 -1.91689879e-01 -1.25864184e+00 5.27996957e-01
-7.14260638e-01 -6.62955046e-01 5.50726295e-01 1.16973543e+00
-6.94763303e-01 9.88390863e-01 -5.67857206e-01 1.97611779e-01
7.92689443e-01 4.39055741e-01 -5.56594193e-01 -1.78815182e-02
-1.71585095e+00 1.01282668e+00 6.66093469e-01 -7.17646554e-02
-1.27134538e+00 -9.34298992e-01 -9.25161183e-01 2.85008941e-02
3.86872709e-01 -6.61933780e-01 8.83783162e-01 -1.18389940e+00
-1.77839386e+00 4.20914054e-01 4.76889521e-01 -7.27298915e-01
8.26067090e-01 -1.11803673e-01 -1.69398069e-01 1.85650498e-01
-3.96072626e-01 3.58860940e-01 1.12099779e+00 -1.25652432e+00
-3.02129507e-01 -6.27252534e-02 5.47175348e-01 -5.26963621e-02
-6.42696083e-01 -2.42867395e-01 3.50184739e-01 -7.63035238e-01
-4.58857208e-01 -1.00828111e+00 -3.54988515e-01 1.43639758e-01
-7.78728604e-01 5.06473839e-01 9.96879876e-01 -3.12804997e-01
6.69517279e-01 -2.10042930e+00 1.32509902e-01 3.92258197e-01
3.35515738e-01 9.17878807e-01 -2.84861654e-01 1.22260801e-01
-2.95200944e-01 6.12768650e-01 -4.59748924e-01 3.32904696e-01
9.22596529e-02 2.40939826e-01 -9.27516222e-01 3.66761059e-01
3.42160046e-01 1.23610842e+00 -1.16577625e+00 -5.83906099e-02
3.09634116e-02 5.12387753e-01 -5.87830305e-01 2.58378416e-01
-3.24713178e-02 3.74329835e-01 -4.98197854e-01 3.82749647e-01
7.71556795e-01 3.21089208e-01 -2.92859435e-01 7.69045800e-02
6.68814480e-01 -9.64759588e-02 -9.00205433e-01 8.65567982e-01
-4.43298101e-01 7.82408178e-01 -2.41620354e-02 -1.21092427e+00
7.76880741e-01 2.61887729e-01 -2.94373989e-01 -1.51246414e-01
9.87067446e-02 2.22787201e-01 3.30481440e-01 -1.82838097e-01
-4.14853133e-02 -4.68135595e-01 -1.66889712e-01 2.33510256e-01
9.09920484e-02 -5.27725041e-01 -3.98737937e-01 2.57257283e-01
1.26223671e+00 -4.23954129e-02 5.59202302e-03 6.34267926e-02
4.87984747e-01 -1.01637565e-01 5.80601633e-01 1.15207565e+00
-1.36015370e-01 4.35542881e-01 6.10130131e-01 -2.02088654e-01
-1.05165076e+00 -1.33215678e+00 1.41338393e-01 7.10061550e-01
2.68211290e-02 -4.62679937e-03 -8.93956780e-01 -1.33426487e+00
2.00387567e-01 6.55875921e-01 -9.26304400e-01 -9.61549222e-01
-5.38011312e-01 -8.00177872e-01 1.24015963e+00 4.16189343e-01
8.49103153e-01 -9.81183887e-01 7.12792054e-02 -1.16738670e-01
2.88914204e-01 -8.78461242e-01 -3.53477806e-01 1.38898179e-01
-6.47783399e-01 -1.08827841e+00 -7.82545447e-01 -5.65295935e-01
9.09818470e-01 -5.42232506e-02 7.02861845e-01 1.45173311e-01
-2.07332253e-01 2.70938486e-01 -1.46258816e-01 -7.19784141e-01
-1.02545810e+00 -7.81717710e-03 1.76270366e-01 -1.03044845e-01
-2.30592728e-01 -7.33336985e-01 -2.79804647e-01 2.43824854e-01
-1.08460808e+00 -4.15774375e-01 5.79696596e-01 1.14011765e+00
1.23325251e-01 3.20233107e-01 7.10155845e-01 -1.13588071e+00
6.26799762e-01 -3.96322817e-01 -6.45011425e-01 1.69417769e-01
-4.09295291e-01 2.42937729e-01 1.32218587e+00 -1.00615394e+00
-7.18209624e-01 -7.68647045e-02 -1.42920449e-01 -6.97945356e-01
3.29725519e-02 1.20207757e-01 -4.45907861e-01 -9.22844052e-01
1.22966838e+00 2.16452748e-01 -1.37561820e-02 -8.29761755e-03
5.50855100e-01 1.01984091e-01 5.93337357e-01 -6.98953032e-01
1.87011075e+00 3.55333477e-01 3.51093590e-01 -3.76838475e-01
-1.07650793e+00 4.48275685e-01 -2.30615050e-01 3.18006314e-02
5.15707016e-01 -4.64406908e-01 -7.04872131e-01 7.20737398e-01
-9.18545723e-01 -6.56356990e-01 -6.45187914e-01 4.01068807e-01
-4.55632180e-01 3.53436947e-01 -5.14473081e-01 -6.09091580e-01
-1.98757768e-01 -9.55934107e-01 2.60430396e-01 2.67685622e-01
3.90334338e-01 -1.37758040e+00 -7.92460814e-02 7.52301812e-02
6.05860591e-01 7.87312865e-01 7.10820258e-01 -1.11313105e+00
-5.44190168e-01 -5.38129032e-01 1.16149962e-01 8.91229391e-01
9.09724012e-02 -1.00338578e-01 -1.09272337e+00 -3.86276573e-01
1.90071940e-01 -5.66684663e-01 1.02268147e+00 -1.50291905e-01
1.22236848e+00 -9.26343739e-01 -2.29774445e-01 9.43026364e-01
1.31366074e+00 -1.58827496e-03 8.82588685e-01 4.48175848e-01
1.02059221e+00 3.19317788e-01 2.19833151e-01 -2.46580601e-01
-3.12527955e-01 3.55996519e-01 7.22710311e-01 -1.47841543e-01
-2.78922096e-02 -2.53230184e-01 7.53493547e-01 2.46476904e-01
8.71009082e-02 -1.56911388e-01 -6.83801472e-01 3.81899685e-01
-1.50704169e+00 -1.37882948e+00 2.81288892e-01 2.37491941e+00
1.28682148e+00 4.12009865e-01 -1.12496607e-01 2.09040269e-01
9.34627354e-01 -1.12784714e-01 -6.24302685e-01 -6.39672458e-01
-2.52181768e-01 5.31776547e-01 7.83924937e-01 8.20263505e-01
-1.09043646e+00 9.81315672e-01 6.67249632e+00 1.06403589e+00
-1.13074827e+00 1.80561990e-02 6.04663849e-01 -1.42646581e-01
-3.65961820e-01 1.75226051e-02 -6.68412209e-01 5.12142539e-01
8.94947827e-01 -6.31028354e-01 6.21822357e-01 1.08672893e+00
-4.57355320e-01 5.99214375e-01 -1.14627361e+00 3.98590475e-01
-3.08635295e-03 -1.33632004e+00 2.15250641e-01 1.03044823e-01
1.06714094e+00 -3.43692958e-01 4.49216753e-01 6.58107817e-01
7.41604924e-01 -1.42608523e+00 2.79558539e-01 3.07658315e-01
6.10405982e-01 -1.02585447e+00 7.41557479e-01 3.53751600e-01
-5.88116646e-01 -1.97123349e-01 -5.35486341e-01 3.49982455e-02
-2.19675496e-01 6.12234712e-01 -9.21775639e-01 4.89103049e-01
2.42199764e-01 1.75187781e-01 -4.19135123e-01 9.30138528e-01
-7.43529141e-01 8.87649298e-01 -3.28461617e-01 -1.37686462e-03
3.46776456e-01 6.43152222e-02 8.63834381e-01 1.03868222e+00
1.45386517e-01 -1.94257841e-01 -2.41966825e-02 1.12279999e+00
-4.06383842e-01 -4.16447699e-01 -1.11644733e+00 3.18184979e-02
4.80967432e-01 1.15479016e+00 -1.36955857e-01 -4.62009668e-01
2.55932361e-02 1.08860958e+00 3.09498459e-01 4.08971161e-01
-9.38656807e-01 -1.10722208e+00 5.86810231e-01 -2.86763340e-01
2.53354579e-01 1.47578210e-01 -1.13803059e-01 -1.19896913e+00
-2.98401769e-02 -1.14994919e+00 1.25616029e-01 -5.64199865e-01
-1.49475276e+00 4.69382286e-01 -9.98765007e-02 -1.10464168e+00
-1.63339496e-01 -7.95977414e-01 -1.14465261e+00 9.70529497e-01
-1.47005761e+00 -1.00954092e+00 -1.37499571e-01 9.35629964e-01
-8.93668681e-02 -4.27508324e-01 8.82830322e-01 -3.56679522e-02
-4.72162038e-01 1.27050161e+00 2.96052694e-01 6.80940866e-01
6.66926682e-01 -1.58521748e+00 4.91332561e-01 1.14679480e+00
3.52956690e-02 6.12628520e-01 7.36460090e-01 -4.65591997e-01
-1.30095387e+00 -1.51750076e+00 2.00123638e-01 -5.95929086e-01
1.00490201e+00 -3.82794470e-01 -9.51425135e-01 7.61931658e-01
-5.93344085e-02 3.24728757e-01 5.34430921e-01 -2.88918227e-01
-8.01891088e-01 -8.06332007e-02 -1.67129624e+00 9.76682723e-01
8.25003088e-01 -6.33365214e-01 -7.64055729e-01 5.92184603e-01
1.05451345e+00 -4.95354652e-01 -8.07202160e-01 4.08868045e-01
2.57544011e-01 -6.58657193e-01 1.20526695e+00 -1.04629135e+00
5.63669980e-01 -1.04156911e-01 8.15057755e-02 -1.82365692e+00
2.94503495e-02 -1.07126260e+00 -3.92456859e-01 1.03405106e+00
6.22366667e-01 -1.23545742e+00 7.88473547e-01 4.76840287e-01
4.16157320e-02 -6.00268185e-01 -8.61105800e-01 -1.37175930e+00
9.06911314e-01 -1.13060482e-01 5.84015071e-01 1.09458208e+00
5.98727167e-02 -1.11923277e-01 -3.93461943e-01 3.83955359e-01
7.43749499e-01 -2.87931591e-01 9.22733128e-01 -8.91608238e-01
-6.82978213e-01 -4.93378818e-01 -8.00052702e-01 -5.45775414e-01
4.96596247e-01 -1.12795925e+00 -1.06634274e-02 -8.91150892e-01
-1.03362449e-01 -3.76539141e-01 -3.50811958e-01 6.93069875e-01
-6.59942806e-01 4.96565223e-01 4.45859879e-01 -1.15452185e-01
-3.83983105e-02 4.45699185e-01 1.51447570e+00 -3.27429920e-01
4.32624929e-02 2.88078725e-01 -9.07498837e-01 7.02987015e-01
1.12625039e+00 -7.85753250e-01 -5.24267793e-01 -1.86753005e-03
2.24899709e-01 -3.36217731e-01 6.68590605e-01 -1.08007956e+00
4.96129431e-02 -2.97678471e-01 2.98736691e-01 3.53247434e-01
2.62174606e-01 -6.85679913e-01 -2.61966854e-01 9.28388178e-01
-4.30065095e-01 -6.24245405e-01 3.19325238e-01 5.99565685e-01
-7.21095949e-02 -5.25134623e-01 1.07380581e+00 -2.01113790e-01
-1.20230056e-01 4.73439127e-01 -2.37215698e-01 4.09406960e-01
1.15230405e+00 4.16181162e-02 -6.92075253e-01 -5.02225697e-01
-7.37216413e-01 2.05466319e-02 3.44443232e-01 3.73278074e-02
6.48884833e-01 -1.42175269e+00 -8.66087675e-01 1.32435113e-01
-3.60237718e-01 -2.10644733e-02 -9.63637680e-02 3.55921060e-01
-4.54862684e-01 -1.15994237e-01 -2.97841668e-01 -3.75661813e-02
-9.51512694e-01 9.23170805e-01 8.14442396e-01 -1.98683262e-01
-4.36311126e-01 1.02519524e+00 4.50914532e-01 -6.80124462e-01
1.28535897e-01 2.39049733e-01 2.12515905e-01 -5.46741843e-01
6.84077740e-01 3.06969017e-01 -2.83247232e-01 -4.10600230e-02
-1.05281994e-01 3.70586544e-01 -2.19853744e-01 -1.48849145e-01
1.01435578e+00 5.07209837e-01 1.04541853e-01 -3.68199237e-02
1.28023040e+00 2.87227005e-01 -1.17232597e+00 -1.25111073e-01
-5.77443779e-01 -4.66800809e-01 -2.09801301e-01 -8.11309457e-01
-1.42734516e+00 1.00796497e+00 3.76224995e-01 5.44237018e-01
1.00261354e+00 -3.49903464e-01 7.04477847e-01 9.95738268e-01
1.52569577e-01 -6.43981993e-01 1.11778475e-01 5.45407593e-01
8.28331888e-01 -1.12049460e+00 -1.57516211e-01 -4.34458107e-01
-3.73195887e-01 1.03723788e+00 7.50344396e-01 -5.91260731e-01
5.73076427e-01 3.78378093e-01 -2.89696939e-02 2.31472537e-01
-4.77726400e-01 1.59458399e-01 1.78128079e-01 1.00274730e+00
-2.81921744e-01 -1.02203883e-01 1.13473654e-01 3.40084165e-01
-5.72650909e-01 -3.73012304e-01 9.27722633e-01 8.47472608e-01
-4.39768225e-01 -1.09133124e+00 -6.86587214e-01 1.99218020e-01
-8.15620005e-01 -2.98238903e-01 -6.70896053e-01 7.82199442e-01
-8.39490443e-02 8.31428230e-01 -6.48487329e-01 -5.87275803e-01
1.39105186e-01 5.89871518e-02 5.81296921e-01 -5.20087242e-01
-8.08484674e-01 -8.84147346e-01 -1.33844271e-01 -3.88732433e-01
6.25014771e-03 -2.88973916e-02 -9.64067876e-01 -6.25311375e-01
-5.65586984e-01 3.05991769e-01 5.23637652e-01 8.34004283e-01
1.47572204e-01 6.17355645e-01 1.17427301e+00 -7.44709849e-01
-1.08084178e+00 -5.50055683e-01 -3.78432244e-01 4.17162657e-01
5.60434759e-01 -3.78925294e-01 -1.16099584e+00 -9.55505967e-02] | [5.676266193389893, 7.86157751083374] |
274a8dd5-3804-4c84-8d9a-775480512c57 | bi-directional-loop-closure-for-visual-slam | 2204.01524 | null | https://arxiv.org/abs/2204.01524v1 | https://arxiv.org/pdf/2204.01524v1.pdf | Bi-directional Loop Closure for Visual SLAM | A key functional block of visual navigation system for intelligent autonomous vehicles is Loop Closure detection and subsequent relocalisation. State-of-the-Art methods still approach the problem as uni-directional along the direction of the previous motion. As a result, most of the methods fail in the absence of a significantly similar overlap of perspectives. In this study, we propose an approach for bi-directional loop closure. This will, for the first time, provide us with the capability to relocalize to a location even when traveling in the opposite direction, thus significantly reducing long-term odometry drift in the absence of a direct loop. We present a technique to select training data from large datasets in order to make them usable for the bi-directional problem. The data is used to train and validate two different CNN architectures for loop closure detection and subsequent regression of 6-DOF camera pose between the views in an end-to-end manner. The outcome packs a considerable impact and aids significantly to real-world scenarios that do not offer direct loop closure opportunities. We provide a rigorous empirical comparison against other established approaches and evaluate our method on both outdoor and indoor data from the FinnForest dataset and PennCOSYVIO dataset. | ['Atanas Gotchev', 'Sari Peltonen', 'Ihtisham Ali'] | 2022-04-01 | null | null | null | null | ['loop-closure-detection'] | ['computer-vision'] | [-7.95276240e-02 -8.88552964e-02 -2.27806360e-01 -5.47163785e-01
-1.80535883e-01 -8.02183330e-01 9.09000635e-01 -7.93811586e-03
-6.99076355e-01 6.89196527e-01 -1.66071624e-01 -7.56181121e-01
-8.88222083e-02 -7.02947438e-01 -9.44726825e-01 -4.37796324e-01
-2.29233071e-01 5.33091605e-01 5.85707366e-01 -6.54471517e-01
3.05164099e-01 7.84784257e-01 -1.88742673e+00 -2.89456934e-01
6.82069302e-01 8.49907756e-01 2.95809835e-01 7.34267592e-01
2.82416016e-01 5.08709311e-01 -7.84538910e-02 1.22406088e-01
5.33330142e-01 6.85621351e-02 -4.63260502e-01 -1.30707175e-01
1.10244548e+00 -4.57504630e-01 -4.76111740e-01 7.16074049e-01
3.93194050e-01 2.23521411e-01 4.73750412e-01 -1.45420825e+00
8.85832235e-02 -3.99072021e-01 -3.68289351e-01 9.46099088e-02
6.48659527e-01 3.42719704e-01 7.04309762e-01 -8.59482527e-01
1.18667948e+00 1.02988482e+00 1.04356742e+00 3.01319063e-01
-1.11921716e+00 -5.22781372e-01 2.00209096e-01 1.37345344e-01
-1.16984260e+00 -6.46160364e-01 5.33343792e-01 -5.16871572e-01
1.26856470e+00 -5.82883880e-02 6.58048391e-01 1.10001123e+00
3.45399350e-01 4.85553920e-01 6.75948858e-01 -1.51528701e-01
6.31041303e-02 -1.53614245e-02 -4.32183407e-02 8.42379391e-01
4.97726738e-01 6.08555377e-01 -4.50188518e-01 3.14459652e-01
3.68538052e-01 2.84978617e-02 -3.95248979e-01 -1.33192921e+00
-1.39287615e+00 7.79629230e-01 7.32242405e-01 4.51048929e-03
-1.38049766e-01 9.34358388e-02 4.26096320e-01 4.95500565e-01
1.42096207e-02 1.70310482e-01 -4.71852392e-01 -4.84076291e-02
-7.63317347e-01 5.94994187e-01 6.80271924e-01 1.16211307e+00
1.01228976e+00 -2.92542040e-01 2.33729556e-01 1.83473557e-01
5.27636766e-01 6.74050212e-01 3.01960826e-01 -8.59195054e-01
8.54477465e-01 5.55861056e-01 4.58187997e-01 -1.11578536e+00
-9.21906769e-01 -2.48031765e-01 -5.26651382e-01 9.32353437e-01
6.41492963e-01 -2.13759109e-01 -9.84474778e-01 1.75984073e+00
5.10343432e-01 -1.35884225e-01 2.75707245e-01 8.28908622e-01
5.88105679e-01 4.81183439e-01 -4.38122481e-01 2.41834611e-01
9.23465431e-01 -1.13252234e+00 -5.70809126e-01 -7.14744389e-01
1.00839877e+00 -5.83681524e-01 7.93140173e-01 1.71756655e-01
-3.47684324e-01 -6.07363045e-01 -1.52562618e+00 -1.67436108e-01
-6.56523108e-01 1.19505711e-01 4.37803119e-01 4.34588075e-01
-1.25316453e+00 3.25148076e-01 -7.64293969e-01 -9.42228377e-01
8.04452375e-02 1.85352266e-01 -9.14421201e-01 -1.34583414e-01
-1.09156156e+00 1.27841926e+00 2.60440081e-01 3.07736695e-01
-7.51697302e-01 -4.70379621e-01 -1.12673509e+00 -5.69989145e-01
2.54407585e-01 -6.59181237e-01 1.14862204e+00 -6.72249615e-01
-1.10766780e+00 9.26058829e-01 -4.18003589e-01 -9.41724598e-01
1.09866893e+00 -4.18008059e-01 -2.54966170e-01 -2.49118984e-01
4.32130516e-01 9.90423799e-01 5.81176639e-01 -1.45752621e+00
-1.23659468e+00 -4.65996504e-01 1.99342251e-01 4.50956911e-01
4.78161037e-01 -8.39863420e-01 -4.63537276e-01 1.52785936e-02
4.14046735e-01 -1.34073508e+00 -2.61129737e-01 3.72783780e-01
-2.15625912e-01 1.11797228e-01 1.13969815e+00 -1.70432031e-01
7.38541126e-01 -1.87066424e+00 4.50150669e-02 2.49604657e-02
1.77101687e-01 2.47659773e-01 1.21414199e-01 6.02455318e-01
1.48165554e-01 -3.52866828e-01 -1.43371280e-02 -5.52691519e-01
-1.70707211e-01 3.36755008e-01 -2.77206540e-01 9.22553420e-01
-1.22118793e-01 6.43407881e-01 -1.03701353e+00 4.48005684e-02
5.48806965e-01 4.01017100e-01 -3.91459733e-01 5.77667654e-02
-8.23771954e-03 5.24145246e-01 -6.63935170e-02 3.13727349e-01
8.10367346e-01 4.20906693e-01 -1.28655285e-01 -1.02929264e-01
-6.17219746e-01 3.27099383e-01 -1.21207011e+00 1.81054771e+00
-5.55865884e-01 1.31468630e+00 -1.33880466e-01 -7.31716871e-01
9.37858462e-01 -1.18211210e-01 2.08104074e-01 -1.21817183e+00
8.00537132e-03 4.72528189e-01 -2.41973847e-01 -4.11429435e-01
7.97169089e-01 1.93013594e-01 -8.73376802e-02 -1.02128044e-01
-2.77700424e-01 -2.44946871e-02 3.61961663e-01 2.50081513e-02
9.96479154e-01 2.77967185e-01 3.21966857e-01 -2.71545410e-01
5.96437931e-01 4.09165919e-01 3.49473596e-01 8.64885747e-01
-5.79013526e-01 6.76236808e-01 2.52636105e-01 -8.44668984e-01
-1.08532786e+00 -9.00914073e-01 3.82163539e-03 5.89674354e-01
7.50619233e-01 -6.37283102e-02 -2.76147306e-01 -7.52875924e-01
3.35992128e-01 4.83122528e-01 -7.71298230e-01 -1.61533251e-01
-5.56196511e-01 -5.51945642e-02 6.02178633e-01 4.05540228e-01
7.19878614e-01 -6.28067255e-01 -1.14518714e+00 5.59435273e-03
-1.53269380e-01 -1.17111552e+00 -2.20628381e-01 3.76530141e-01
-8.12518179e-01 -1.17618537e+00 -3.97728562e-01 -8.43988836e-01
5.08045733e-01 8.19365740e-01 9.08130884e-01 -2.09858000e-01
1.53430730e-01 2.29223073e-01 -4.34039272e-02 -2.19599292e-01
-2.52298892e-01 1.22478023e-01 1.70490786e-01 -3.34882051e-01
4.09950793e-01 -2.34446302e-01 -7.51264036e-01 5.47738552e-01
-2.83531934e-01 6.77879527e-02 4.00465786e-01 7.36439884e-01
3.37848693e-01 -3.03082138e-01 1.51404485e-01 -5.40592074e-01
1.71437010e-01 -3.74675870e-01 -9.77820277e-01 -1.92629918e-01
-9.07213390e-01 9.24733356e-02 4.18116212e-01 -8.35809633e-02
-8.80216300e-01 4.10330981e-01 -1.35708138e-01 -2.27068782e-01
-3.20487350e-01 4.70925033e-01 7.83452392e-02 -2.72509247e-01
9.13990617e-01 -3.66524644e-02 2.41333321e-01 -1.26246229e-01
4.49093878e-01 4.16317701e-01 6.45631731e-01 2.05855176e-01
8.21887255e-01 1.08863294e+00 2.93654531e-01 -8.31909597e-01
-3.22907418e-01 -8.22917104e-01 -9.04242754e-01 -3.01202655e-01
6.22513235e-01 -1.03650296e+00 -5.37302673e-01 3.65583718e-01
-1.07844436e+00 -4.72908437e-01 -4.41525988e-02 4.14969236e-01
-6.19158864e-01 3.46014589e-01 2.19620004e-01 -6.02900863e-01
6.65388480e-02 -1.15708590e+00 9.86357510e-01 2.26662725e-01
-5.47587350e-02 -9.75345492e-01 5.70828199e-01 6.23015426e-02
3.35897058e-01 3.97773027e-01 2.68742461e-02 -1.52076215e-01
-6.65245414e-01 -5.84648132e-01 -1.61589801e-01 -1.20208032e-01
-1.03454940e-01 -1.94099635e-01 -8.13626170e-01 -6.04734540e-01
-3.80074114e-01 -1.28403246e-01 1.14573491e+00 3.24049920e-01
-1.39114514e-01 1.72579676e-01 -7.23017931e-01 7.94932604e-01
1.52454579e+00 2.38117039e-01 6.63134217e-01 1.01924980e+00
8.61444652e-01 6.16704762e-01 8.97023678e-01 2.66952123e-02
8.25197458e-01 9.56809759e-01 8.54804754e-01 -2.00072154e-01
8.44317488e-03 -4.67694014e-01 3.44489127e-01 -4.01386879e-02
1.35272220e-01 -3.74747515e-01 -1.09279120e+00 8.65163147e-01
-2.18601155e+00 -8.44240904e-01 -5.09515166e-01 2.47058249e+00
-3.01260259e-02 4.79294717e-01 -1.95686426e-02 -7.67570212e-02
3.60560685e-01 2.47675121e-01 -4.14625615e-01 -3.40148151e-01
3.88164371e-02 -7.71995068e-01 1.30006754e+00 7.81549215e-01
-1.29849958e+00 9.24658716e-01 5.73209572e+00 6.16410486e-02
-1.25638437e+00 -1.40481815e-01 -1.71420779e-02 1.88776270e-01
3.74108180e-02 3.13517660e-01 -1.02011216e+00 4.52064537e-02
8.05598676e-01 4.02795345e-01 1.09754279e-01 8.78105223e-01
3.84157926e-01 -6.82806551e-01 -1.12706089e+00 8.37917328e-01
-3.38886082e-02 -1.19581735e+00 -3.26544702e-01 9.54973474e-02
7.39875436e-01 6.91697717e-01 -1.09037161e-01 1.75510406e-01
1.00525379e-01 -7.57296145e-01 1.03185201e+00 2.80477077e-01
6.14009976e-01 -6.75246835e-01 8.45099211e-01 5.14091372e-01
-1.33451152e+00 -2.16965109e-01 -1.30905882e-01 -2.63690442e-01
3.67640108e-01 8.87277499e-02 -1.00404561e+00 6.19147897e-01
7.24115431e-01 7.86654592e-01 -6.56900287e-01 1.13139343e+00
-9.48404223e-02 -2.78871786e-02 -5.66590190e-01 9.72437859e-02
6.43873513e-01 3.59071493e-02 7.70692766e-01 1.06670702e+00
3.73278826e-01 -6.85695112e-01 4.11367044e-02 3.97111475e-01
4.09965575e-01 -2.90133744e-01 -1.40181828e+00 5.89643896e-01
4.44617689e-01 7.92417884e-01 -6.79413974e-01 -1.16867743e-01
-5.91324866e-01 7.60716736e-01 3.97032768e-01 4.10543025e-01
-7.72251010e-01 -4.80212897e-01 8.83708358e-01 2.57725328e-01
4.82708395e-01 -6.22334003e-01 -7.52613768e-02 -9.04164732e-01
3.23639482e-01 -4.58592981e-01 -3.46147232e-02 -6.44133747e-01
-5.19196212e-01 4.76458907e-01 -8.24055597e-02 -1.57881045e+00
-4.94089127e-01 -5.82676172e-01 -3.21640253e-01 8.13582480e-01
-1.80210853e+00 -1.19783223e+00 -7.39572287e-01 3.77631962e-01
5.09543836e-01 -4.29994278e-02 5.22985995e-01 3.93583983e-01
-1.25930876e-01 3.37467819e-01 3.77249926e-01 -9.64884758e-02
8.47248673e-01 -1.12471437e+00 7.95856774e-01 1.22002184e+00
-1.97464898e-02 5.40683091e-01 1.21454990e+00 -6.24372661e-01
-1.45514750e+00 -1.01949143e+00 1.16006052e+00 -7.50069737e-01
3.79386097e-01 -4.67710763e-01 -5.14371812e-01 8.52232397e-01
-1.54231489e-01 8.67047235e-02 -2.89851874e-01 -6.91554416e-03
-2.44366258e-01 -3.18420708e-01 -9.63825047e-01 8.13571334e-01
1.18703687e+00 -4.31386262e-01 -4.74532902e-01 -6.41301796e-02
4.31182176e-01 -7.66657889e-01 -4.30518053e-02 4.90314901e-01
9.59311962e-01 -1.43753552e+00 9.71142650e-01 -1.10161960e-01
8.82268474e-02 -7.15956151e-01 -1.63614139e-01 -1.18832743e+00
4.25145179e-02 -5.04551649e-01 8.61356035e-02 7.90964305e-01
5.46085238e-01 -8.64770830e-01 7.95815766e-01 1.09373137e-01
-1.66361168e-01 -4.44277227e-01 -1.10528433e+00 -8.78479004e-01
-3.55535448e-01 -6.53461874e-01 3.29943091e-01 4.17825818e-01
-2.62200624e-01 3.84577513e-01 -4.95258331e-01 3.81159931e-01
5.13355672e-01 5.19918911e-02 1.46365476e+00 -1.18380511e+00
3.65425974e-01 -3.32486153e-01 -9.69769239e-01 -1.56386483e+00
-1.33457273e-01 -5.50030589e-01 5.54423034e-01 -1.65946817e+00
-5.54222584e-01 -4.48558629e-01 1.42694786e-01 2.20068730e-02
2.80966789e-01 2.59707659e-01 -1.62884563e-01 1.89089462e-01
-5.80889344e-01 4.74282861e-01 8.34930539e-01 -1.24441497e-02
-4.12070721e-01 1.45227492e-01 -2.30393499e-01 5.19596815e-01
7.27051020e-01 -3.08579981e-01 -7.72108734e-01 -4.44449544e-01
4.49926317e-01 -5.89236505e-02 5.48897684e-01 -1.37874329e+00
4.96486485e-01 1.60248235e-01 1.39942229e-01 -1.07228827e+00
2.29829997e-01 -1.12295902e+00 1.32256299e-01 5.99488795e-01
1.29739696e-03 3.74612510e-01 3.03997964e-01 9.25354600e-01
-9.69113484e-02 1.95781793e-02 6.03173554e-01 1.32667601e-01
-1.38423204e+00 1.07981391e-01 -6.43478870e-01 -1.53431192e-01
1.13906813e+00 -7.14012265e-01 -2.71931946e-01 -4.48745430e-01
-2.97377050e-01 6.23261213e-01 8.99547577e-01 8.42528760e-01
4.00506675e-01 -1.18447399e+00 -2.07830682e-01 6.62744462e-01
7.04768658e-01 1.85705930e-01 5.79832606e-02 1.04816365e+00
-1.01073337e+00 9.07012761e-01 -4.17769998e-01 -1.04699671e+00
-1.18390095e+00 5.75192332e-01 5.64472258e-01 1.96329318e-02
-9.07323837e-01 5.51705897e-01 1.67932838e-01 -8.59264374e-01
2.53034323e-01 -4.82642263e-01 -3.45428169e-01 1.76086247e-01
2.80813783e-01 1.60323143e-01 4.37261641e-01 -1.14214051e+00
-5.25106728e-01 6.63878679e-01 8.40888843e-02 -2.69919991e-01
8.93710673e-01 -8.94801617e-01 4.63265628e-01 4.99139369e-01
1.25810707e+00 -1.06149904e-01 -1.77794611e+00 5.01589142e-02
-1.51465284e-02 -4.79545891e-01 3.53694595e-02 -3.90097082e-01
-7.02916920e-01 6.64783716e-01 1.18344343e+00 4.47303392e-02
4.95457917e-01 -3.34998429e-01 5.10082543e-01 7.80402124e-01
5.55469871e-01 -1.08653331e+00 -3.65367323e-01 8.31494749e-01
6.27568781e-01 -1.77239227e+00 -1.18475862e-01 2.87447665e-02
-3.70049447e-01 1.18332887e+00 6.43729746e-01 -1.65301830e-01
3.98600191e-01 8.34543109e-02 4.36185449e-01 -2.83532590e-01
-5.71809351e-01 -3.55756700e-01 6.34597018e-02 9.79427338e-01
1.78719118e-01 -1.75427258e-01 2.33414955e-03 -3.57700050e-01
-2.95271650e-02 -8.44771042e-02 4.13392693e-01 1.28051412e+00
-5.79636812e-01 -6.62024081e-01 -3.73367041e-01 1.49808815e-02
3.17597926e-01 3.72940958e-01 -2.48389363e-01 1.26610899e+00
2.54892230e-01 9.63017642e-01 9.56900492e-02 -3.68034929e-01
7.14032829e-01 -1.44571617e-01 2.21644729e-01 -3.02962642e-02
9.20757157e-05 -3.86719853e-01 3.33300799e-01 -9.18443143e-01
-2.91184813e-01 -7.72448361e-01 -1.11397672e+00 -3.19955409e-01
-2.69465119e-01 -3.04315329e-01 7.53624499e-01 9.79714513e-01
4.04447496e-01 3.40522677e-01 3.92298609e-01 -1.55020976e+00
-1.51264399e-01 -5.75495303e-01 -1.01533951e-02 2.88058549e-01
1.13239455e+00 -1.00086558e+00 -3.05519462e-01 -1.90146983e-01] | [7.51947021484375, -1.9667763710021973] |
23b4b6aa-0646-4fc8-8ed6-df65f40cd1d0 | non-autoregressive-neural-dialogue-generation | 2002.04250 | null | https://arxiv.org/abs/2002.04250v2 | https://arxiv.org/pdf/2002.04250v2.pdf | Non-Autoregressive Neural Dialogue Generation | Maximum Mutual information (MMI), which models the bidirectional dependency between responses ($y$) and contexts ($x$), i.e., the forward probability $\log p(y|x)$ and the backward probability $\log p(x|y)$, has been widely used as the objective in the \sts model to address the dull-response issue in open-domain dialog generation. Unfortunately, under the framework of the \sts model, direct decoding from $\log p(y|x) + \log p(x|y)$ is infeasible since the second part (i.e., $p(x|y)$) requires the completion of target generation before it can be computed, and the search space for $y$ is enormous. Empirically, an N-best list is first generated given $p(y|x)$, and $p(x|y)$ is then used to rerank the N-best list, which inevitably results in non-globally-optimal solutions. In this paper, we propose to use non-autoregressive (non-AR) generation model to address this non-global optimality issue. Since target tokens are generated independently in non-AR generation, $p(x|y)$ for each target word can be computed as soon as it's generated, and does not have to wait for the completion of the whole sequence. This naturally resolves the non-global optimal issue in decoding. Experimental results demonstrate that the proposed non-AR strategy produces more diverse, coherent, and appropriate responses, yielding substantive gains in BLEU scores and in human evaluations. | ['Qinghong Han', 'Yuxian Meng', 'Jiwei Li', 'Fei Wu'] | 2020-02-11 | null | null | null | null | ['open-domain-dialog'] | ['natural-language-processing'] | [ 2.29365602e-01 7.27458745e-02 -3.72866690e-02 -5.12430966e-01
-1.27017832e+00 -5.59689999e-01 1.69442907e-01 -1.89918727e-02
-6.10586286e-01 1.07266212e+00 2.46597156e-01 -4.17490929e-01
-7.60509819e-02 -7.48967171e-01 -4.63577628e-01 -6.15939558e-01
1.06079973e-01 5.22714078e-01 4.06638198e-02 -3.42539370e-01
2.04136997e-01 -3.27778697e-01 -1.31839895e+00 -1.00813182e-02
1.08845210e+00 7.67791569e-01 7.86043167e-01 7.92167246e-01
-4.70193535e-01 2.15238392e-01 -8.78087103e-01 -6.54948831e-01
-5.44224195e-02 -9.92207348e-01 -8.74824226e-01 -5.17222047e-01
-1.38615340e-01 -4.94338483e-01 -1.61265358e-01 1.17918897e+00
4.82101202e-01 3.90259862e-01 6.92678213e-01 -1.01303947e+00
-5.94360352e-01 9.78747606e-01 -7.12276578e-01 -2.23096609e-02
6.77088261e-01 1.55790240e-01 1.47108531e+00 -6.38412237e-01
4.66482222e-01 1.45675242e+00 3.24130774e-01 6.92744017e-01
-1.00356805e+00 -9.10054564e-01 2.46023044e-01 -4.61274870e-02
-1.38535750e+00 -2.48891875e-01 5.48970699e-01 -4.25868258e-02
9.43288207e-01 4.80209678e-01 1.56819552e-01 9.70633507e-01
1.60507590e-01 8.50906253e-01 1.03781533e+00 -6.14004016e-01
1.13033868e-01 1.81491207e-02 2.32168585e-01 3.56701940e-01
-2.74436265e-01 8.25254619e-02 -6.96076095e-01 -2.31681734e-01
4.12635118e-01 -2.57401526e-01 -2.07756788e-01 3.77338439e-01
-9.05756891e-01 8.34856212e-01 1.36490464e-01 3.02353889e-01
-3.66822660e-01 1.65245861e-01 -6.92968220e-02 2.04717189e-01
1.06982589e-01 2.58473158e-01 -7.35421181e-01 -7.06734896e-01
-7.95212805e-01 4.83613044e-01 9.06650424e-01 1.05186927e+00
9.62330401e-01 2.67135873e-02 -2.78606176e-01 1.17085218e+00
6.92103207e-01 7.48189807e-01 5.22026241e-01 -8.49256039e-01
7.85903692e-01 4.54002410e-01 3.87637138e-01 -7.54693270e-01
-5.49215600e-02 -1.75219283e-01 -7.20823944e-01 -3.40024531e-01
6.65044904e-01 -4.41058159e-01 -7.05582082e-01 2.42532182e+00
1.40995711e-01 -1.82827652e-01 2.78894871e-01 1.01164401e+00
5.35252988e-01 1.13946915e+00 3.12050998e-01 -6.37667716e-01
1.38627803e+00 -4.33957368e-01 -7.17076838e-01 -4.90940034e-01
5.48291504e-01 -1.00010979e+00 1.36388850e+00 5.94136082e-02
-1.24708331e+00 -3.35564166e-01 -7.00538576e-01 5.87532297e-02
3.15567590e-02 8.84664953e-02 4.24776584e-01 4.90284681e-01
-9.77588832e-01 2.85242021e-01 -2.76105314e-01 -2.12996587e-01
-3.07580173e-01 4.54718709e-01 2.12308746e-02 -2.52539128e-01
-1.71612346e+00 6.77571774e-01 2.11035714e-01 2.49867588e-01
-6.50467932e-01 -2.54676491e-01 -6.11089826e-01 2.02356383e-01
5.70720613e-01 -4.81171548e-01 1.35013437e+00 -8.82673979e-01
-1.71588838e+00 3.07752430e-01 -8.17982614e-01 -6.99209720e-02
3.25216800e-01 -2.72418916e-01 -1.70133263e-01 -2.46445104e-01
2.22921282e-01 7.88770556e-01 6.72884881e-01 -1.18397450e+00
-6.91977501e-01 -4.59588677e-01 -1.69960782e-02 6.44672453e-01
-1.63081005e-01 3.29657137e-01 -4.88053769e-01 -5.70838809e-01
4.33265537e-01 -1.00980031e+00 -1.29159555e-01 -8.40777159e-01
-5.79733670e-01 -4.54139858e-01 3.16770256e-01 -1.03233480e+00
1.62186265e+00 -2.14022040e+00 9.88105685e-02 2.95850247e-01
-1.48476828e-02 1.22998625e-01 -9.92671028e-02 6.12568259e-01
1.34410113e-01 2.25730866e-01 -4.20987517e-01 -1.87083274e-01
1.28560916e-01 2.48557687e-01 -1.95760861e-01 -2.02786043e-01
-1.31180108e-01 6.58693492e-01 -7.57580519e-01 -2.98569500e-01
-1.77274212e-01 1.60885468e-01 -6.45075321e-01 3.82191360e-01
-4.70098853e-01 2.18788117e-01 -4.22490984e-01 3.50462437e-01
5.67434728e-01 -1.43851936e-01 4.69699055e-01 2.67829329e-01
-8.59516710e-02 5.26001275e-01 -1.29381514e+00 1.38576424e+00
-4.44729805e-01 2.13018909e-01 1.06705315e-01 -7.33231604e-01
1.08486378e+00 2.80148357e-01 1.61074847e-01 -7.87268519e-01
2.78148707e-02 2.67509073e-01 2.96890855e-01 -1.11093976e-01
7.82865644e-01 -1.93820313e-01 -5.20899534e-01 7.51309097e-01
-1.07341014e-01 4.66646701e-02 2.43577033e-01 3.49063724e-01
9.33457196e-01 2.60330271e-02 1.80344075e-01 3.30706500e-03
4.27591801e-01 -2.13917390e-01 8.17247331e-01 1.08863986e+00
-6.82714283e-02 4.52443957e-01 6.79892302e-01 1.68763325e-01
-8.72652113e-01 -1.08155656e+00 3.30482721e-01 1.43341839e+00
3.48293394e-01 -3.59289289e-01 -9.12689209e-01 -5.75913966e-01
-4.88072276e-01 1.06390584e+00 -1.84259102e-01 2.09380202e-02
-8.15541744e-01 -6.69609249e-01 5.40689766e-01 3.26755106e-01
5.27838707e-01 -1.24470878e+00 -3.59984070e-01 5.42279005e-01
-7.64838636e-01 -6.34280503e-01 -7.93429732e-01 2.37329006e-01
-6.34039223e-01 -7.75160372e-01 -7.09703088e-01 -6.52231276e-01
6.05833232e-01 7.26957992e-02 7.85474658e-01 7.22743198e-02
1.16211340e-01 1.08354941e-01 -5.56761682e-01 -1.49342567e-01
-3.75015140e-01 -2.01571267e-02 -7.81182274e-02 -1.50230721e-01
3.47889483e-01 -2.21775800e-01 -5.38494527e-01 4.02186275e-01
-7.99387276e-01 -1.52214661e-01 6.45737708e-01 1.07475924e+00
4.38256055e-01 8.59194249e-02 8.30975592e-01 -7.82432020e-01
1.10233641e+00 -3.77999574e-01 -4.47177470e-01 4.78210449e-01
-6.14063323e-01 1.00237332e-01 6.71363890e-01 -5.10738611e-01
-1.36510944e+00 -1.70425683e-01 -4.59243923e-01 3.81508050e-03
5.02897687e-02 7.93477595e-01 -2.94535905e-01 6.94488525e-01
3.09177518e-01 5.55178165e-01 -1.24342337e-01 -4.04868692e-01
3.67057383e-01 7.66256452e-01 3.65635097e-01 -4.92811799e-01
4.41454142e-01 -3.08948815e-01 -5.22788465e-01 -5.95287323e-01
-3.34574491e-01 -2.27077156e-01 -1.56999275e-01 -5.30960318e-03
7.68459737e-01 -7.62210369e-01 -8.65586281e-01 4.02244419e-01
-1.30977440e+00 -1.12144180e-01 1.57079902e-02 7.42012382e-01
-3.54425281e-01 4.78375167e-01 -5.35330355e-01 -1.31424189e+00
-4.17230964e-01 -1.26321185e+00 7.86178648e-01 4.97713864e-01
-7.28295386e-01 -6.29428923e-01 -2.10517958e-01 4.21338528e-01
2.49988228e-01 -5.73315859e-01 1.38783765e+00 -6.78252459e-01
-6.09050751e-01 -2.08678618e-01 -2.22862169e-01 2.44675204e-01
-3.82734910e-02 -2.67631054e-01 -5.82963049e-01 -1.25948057e-01
7.12710395e-02 -9.20290872e-02 4.88456398e-01 3.77045274e-01
5.98100305e-01 -7.46313870e-01 -1.71553776e-01 -6.49790019e-02
1.08274519e+00 6.89639330e-01 6.15686059e-01 -2.77176380e-01
2.79958665e-01 4.87523139e-01 8.34842086e-01 7.26620376e-01
7.04619586e-01 7.69050419e-01 6.39148727e-02 3.86333525e-01
1.59203380e-01 -4.94016677e-01 7.37843096e-01 9.53917742e-01
2.67825037e-01 -6.64168656e-01 -7.29221284e-01 4.39230829e-01
-1.65978014e+00 -7.23529696e-01 -5.79813048e-02 2.41533470e+00
1.18519711e+00 6.60476787e-03 -1.46245569e-01 -1.34842634e-01
8.17591786e-01 1.92990899e-01 -4.71323580e-01 -5.91971934e-01
-1.66249335e-01 3.82499844e-01 7.55971894e-02 9.03861284e-01
-3.24286819e-01 1.24067438e+00 6.21197414e+00 1.15771735e+00
-9.29711461e-01 1.42392844e-01 7.09615588e-01 -4.61853668e-02
-6.48485065e-01 4.39430028e-01 -1.17900288e+00 9.08608735e-01
8.94575298e-01 -1.46914601e-01 6.29622936e-01 6.97347760e-01
2.87325889e-01 -6.08538032e-01 -8.57099175e-01 9.18957949e-01
-1.65245198e-02 -6.76988780e-01 -4.58921567e-02 2.20449548e-02
3.71744424e-01 -5.06157935e-01 7.92783499e-03 6.45924687e-01
7.18906999e-01 -7.69322395e-01 5.52616537e-01 1.10973291e-01
8.33493292e-01 -8.57082963e-01 6.38123572e-01 7.07101405e-01
-1.12467813e+00 -7.63455480e-02 -3.47428352e-01 -5.41424900e-02
4.04797316e-01 4.58224386e-01 -9.51775432e-01 5.48657119e-01
6.21291637e-01 -4.51843329e-02 1.04513898e-01 3.21500301e-01
-3.28812420e-01 5.52050173e-01 -4.36094940e-01 -5.15806198e-01
2.74922013e-01 -2.76121289e-01 6.73836708e-01 1.03293550e+00
7.05823660e-01 5.89360595e-01 1.87675476e-01 7.86006868e-01
-4.91374843e-02 3.12622339e-01 -2.39689693e-01 -2.46239584e-02
7.06374824e-01 6.94182098e-01 -4.95941967e-01 -3.03475827e-01
-8.73721614e-02 9.93160665e-01 2.39493623e-01 5.60701728e-01
-8.87514114e-01 -4.05431390e-01 5.92180014e-01 -1.82139739e-01
1.45082533e-01 -1.20825872e-01 -2.26997077e-01 -7.72924721e-01
5.13764843e-02 -9.85104084e-01 5.47457874e-01 -7.29407728e-01
-1.08658147e+00 6.02325439e-01 2.67887842e-02 -1.00697649e+00
-7.86367834e-01 -5.87578677e-03 -3.81892532e-01 1.33988619e+00
-1.37433600e+00 -6.24470532e-01 3.30108136e-01 5.43294787e-01
7.34362900e-01 7.04734623e-02 9.99893308e-01 1.30291581e-01
-4.86419231e-01 1.06096292e+00 1.06849208e-01 -1.26910523e-01
6.94958746e-01 -1.02719998e+00 -5.83054088e-02 8.61574233e-01
7.95284007e-03 8.37006867e-01 6.93817675e-01 -9.63436902e-01
-1.28363705e+00 -5.07470787e-01 1.52845013e+00 -1.19142361e-01
1.97544113e-01 -7.72488937e-02 -8.36021066e-01 3.95922929e-01
2.01476887e-01 -8.52046847e-01 8.17105532e-01 1.13421530e-01
-1.16800435e-01 7.72211924e-02 -9.95374918e-01 9.21516657e-01
7.79292583e-01 -3.72453421e-01 -3.86163622e-01 4.24419455e-02
8.10049832e-01 -3.79919320e-01 -6.31459713e-01 3.42523247e-01
5.35404623e-01 -8.36891592e-01 6.10443115e-01 -1.65221781e-01
2.54252911e-01 -5.40605597e-02 -3.72427791e-01 -1.22278857e+00
-3.08067841e-03 -9.57833648e-01 1.76627174e-01 1.38724959e+00
9.28382754e-01 -5.51636219e-01 5.32188475e-01 9.98521924e-01
-3.87453027e-02 -5.63098907e-01 -1.29637444e+00 -4.01632100e-01
-7.38531053e-02 -5.22471249e-01 7.26941884e-01 5.31365812e-01
1.03717543e-01 3.62933725e-01 -6.56882703e-01 3.91900055e-02
2.68262923e-01 2.37788096e-01 6.59166276e-01 -7.72661686e-01
-5.56117952e-01 -2.50440121e-01 4.02961612e-01 -1.79315853e+00
3.26567329e-02 -5.67764163e-01 5.85251331e-01 -1.34477985e+00
2.31965542e-01 -8.11228096e-01 -1.76476479e-01 3.40693533e-01
-6.79596305e-01 -3.07840616e-01 3.65321964e-01 4.42680031e-01
-4.52417046e-01 5.54439127e-01 1.10957265e+00 1.25194177e-01
-4.68645871e-01 3.51793580e-02 -8.26442122e-01 5.33455670e-01
6.64389849e-01 -5.76632857e-01 -4.37243372e-01 -5.34020722e-01
1.72550946e-01 7.90087759e-01 -2.28745356e-01 -3.63620222e-01
3.00610751e-01 -2.91939914e-01 1.12106726e-01 -6.42987907e-01
6.57993138e-01 -4.91318375e-01 1.12678044e-01 2.98314929e-01
-5.07621288e-01 2.26703897e-01 -1.91540465e-01 5.65954983e-01
-1.76696658e-01 -5.74252069e-01 5.09538174e-01 -3.36413234e-01
-4.05350506e-01 3.55274305e-02 -2.71106005e-01 5.57927601e-02
7.41215527e-01 -2.26886049e-01 -1.10694729e-01 -7.52919853e-01
-7.06381559e-01 2.22651422e-01 8.94499496e-02 3.59868020e-01
7.28630543e-01 -1.00308359e+00 -6.50862515e-01 1.87514499e-01
-2.37936690e-01 -5.53066435e-04 5.82210839e-01 5.11501849e-01
-3.38634662e-02 2.90592283e-01 3.15029413e-01 -4.99225795e-01
-1.24376845e+00 -3.97846512e-02 -1.52263820e-01 -8.24268341e-01
-1.91034507e-02 1.02740228e+00 2.02506498e-01 -3.99731040e-01
2.50289023e-01 1.52544994e-02 -1.36276767e-01 7.19654411e-02
3.18414599e-01 4.94842641e-02 -3.59318346e-01 -6.30801201e-01
-2.71215379e-01 2.31075481e-01 -4.15338069e-01 -8.33552361e-01
8.53473127e-01 -5.23908615e-01 -1.96612880e-01 4.25266832e-01
1.10930896e+00 5.17585911e-02 -9.84270573e-01 -4.06138659e-01
-1.21318817e-01 -6.48661852e-01 -3.98396701e-01 -9.94864643e-01
-6.56482160e-01 6.42064452e-01 3.57559294e-01 1.15748029e-02
9.96535420e-01 -6.85975850e-02 1.11663508e+00 2.50139654e-01
4.36620623e-01 -1.30134869e+00 1.53797448e-01 9.27648842e-01
5.50362587e-01 -1.08559692e+00 -4.06867474e-01 -3.90631795e-01
-1.03229964e+00 6.19148850e-01 7.93969810e-01 4.11536068e-01
4.42107618e-01 -2.55016647e-02 1.04541883e-01 1.63692281e-01
-7.40033984e-01 7.58395530e-03 -1.90028802e-01 3.41664165e-01
5.18124521e-01 2.87469089e-01 -7.39606440e-01 8.77219915e-01
-4.78911757e-01 -3.58648896e-01 2.45773807e-01 9.27393377e-01
-5.40935695e-01 -1.60614121e+00 -4.24519658e-01 4.55480039e-01
-5.52026272e-01 -4.01873648e-01 -3.14246327e-01 3.96221071e-01
-1.66431129e-01 1.36439788e+00 -1.43808201e-01 -4.17041987e-01
1.26843184e-01 3.40679288e-01 1.08515039e-01 -5.25980711e-01
-6.22092843e-01 5.18645227e-01 1.23031057e-01 -2.38918096e-01
1.21203758e-01 -6.49446487e-01 -1.58892930e+00 -3.83014619e-01
-6.22843444e-01 4.97801483e-01 7.35820472e-01 9.55340505e-01
2.74797350e-01 1.77076206e-01 8.19096982e-01 -2.79093623e-01
-7.96588242e-01 -1.07070172e+00 -5.52100241e-01 2.81359076e-01
-2.11240843e-01 -4.34363097e-01 -3.73201370e-01 -1.46613166e-01] | [12.48780345916748, 8.38443660736084] |
5c1614dd-87cc-4dac-896b-b08f7cc3af9c | flow-guided-video-inpainting-with-scene | 2108.12845 | null | https://arxiv.org/abs/2108.12845v1 | https://arxiv.org/pdf/2108.12845v1.pdf | Flow-Guided Video Inpainting with Scene Templates | We consider the problem of filling in missing spatio-temporal regions of a video. We provide a novel flow-based solution by introducing a generative model of images in relation to the scene (without missing regions) and mappings from the scene to images. We use the model to jointly infer the scene template, a 2D representation of the scene, and the mappings. This ensures consistency of the frame-to-frame flows generated to the underlying scene, reducing geometric distortions in flow based inpainting. The template is mapped to the missing regions in the video by a new L2-L1 interpolation scheme, creating crisp inpaintings and reducing common blur and distortion artifacts. We show on two benchmark datasets that our approach out-performs state-of-the-art quantitatively and in user studies. | ['Ganesh Sundaramoorthi', 'Peter Wonka', 'Peihao Zhu', 'Dong Lao'] | 2021-08-29 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Lao_Flow-Guided_Video_Inpainting_With_Scene_Templates_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Lao_Flow-Guided_Video_Inpainting_With_Scene_Templates_ICCV_2021_paper.pdf | iccv-2021-1 | ['video-inpainting'] | ['computer-vision'] | [ 3.29646379e-01 4.55546007e-02 1.82777405e-01 -1.43376738e-01
-3.71940941e-01 -7.63574183e-01 6.87360048e-01 -4.53273356e-01
-4.09071669e-02 8.35632324e-01 5.68633080e-01 1.54826835e-01
4.69863042e-02 -7.16242373e-01 -9.21881139e-01 -3.27240497e-01
7.88277313e-02 1.93470955e-01 3.20131510e-01 -5.00777252e-02
2.32227191e-01 6.14445090e-01 -1.49383700e+00 7.24083364e-01
7.49825954e-01 6.96908534e-01 3.23789954e-01 1.07256520e+00
-7.85017535e-02 1.34408283e+00 -3.33531260e-01 -2.83858597e-01
6.94444180e-01 -8.84960890e-01 -8.67337286e-01 7.52512217e-01
8.57971370e-01 -6.68431103e-01 -8.14077675e-01 8.48894298e-01
-1.69302911e-01 2.45370895e-01 4.44521666e-01 -1.02628040e+00
-5.95486760e-01 -1.01221956e-01 -7.08988309e-01 6.08672053e-02
8.15328717e-01 2.98723400e-01 4.09067512e-01 -9.24655318e-01
1.29301786e+00 1.38230538e+00 4.48202163e-01 4.48509783e-01
-1.52976704e+00 -5.55867441e-02 -1.18728593e-01 5.84811382e-02
-1.28272355e+00 -8.48277032e-01 8.10778797e-01 -6.58515275e-01
7.52952337e-01 3.01841497e-01 6.64114177e-01 6.82648242e-01
3.69286060e-01 4.05518889e-01 9.50129569e-01 -5.72504938e-01
2.74641067e-01 -7.39561170e-02 -5.67840219e-01 6.83190227e-01
-1.25552654e-01 2.42991671e-01 -8.25038373e-01 7.71675557e-02
1.38958275e+00 -4.34624180e-02 -5.08391023e-01 -4.49814409e-01
-1.15204799e+00 3.79263312e-01 2.33223766e-01 1.39403597e-01
-5.36115408e-01 3.68809938e-01 3.40504907e-02 2.07475245e-01
7.84776211e-01 2.46796757e-01 -1.20019265e-01 -3.29797938e-02
-1.45657802e+00 4.80556339e-01 6.43031478e-01 1.09617317e+00
9.45670962e-01 1.18326008e-01 -3.36970508e-01 4.75868285e-01
6.82155415e-02 2.78714895e-01 -8.29454437e-02 -1.92588222e+00
4.57849145e-01 1.08681684e-02 4.65211749e-01 -9.79488015e-01
1.98852614e-01 -1.62094146e-01 -7.37127364e-01 4.15071994e-01
4.81451988e-01 1.37173817e-01 -8.14741433e-01 1.49194884e+00
2.95562923e-01 8.20963085e-01 -3.01803678e-01 8.72510910e-01
2.64839083e-01 9.38782692e-01 -2.72191942e-01 -4.64771181e-01
1.00363886e+00 -1.08334219e+00 -9.65093136e-01 -3.86128157e-01
2.44864285e-01 -9.40003276e-01 7.04568565e-01 4.38024908e-01
-1.79701793e+00 -9.18070674e-01 -5.91867507e-01 -4.43908960e-01
3.07420552e-01 -1.91994041e-01 1.51740223e-01 3.98443669e-01
-1.39761508e+00 7.26076841e-01 -7.59395540e-01 -6.43596575e-02
4.40382510e-01 -1.43319607e-01 -5.71749270e-01 -3.82606834e-01
-8.16912889e-01 9.68924999e-01 1.80613160e-01 -1.05563268e-01
-1.11493838e+00 -1.15358639e+00 -1.14649439e+00 1.55232409e-02
1.40801400e-01 -1.18496764e+00 1.02592051e+00 -1.12007987e+00
-1.29221559e+00 8.83690357e-01 -6.63821936e-01 -3.91243428e-01
9.90964234e-01 -3.65769506e-01 -4.31805179e-02 4.18532759e-01
2.00706482e-01 9.20415223e-01 1.13275456e+00 -1.59175658e+00
-6.60398304e-01 5.23962677e-02 7.16551691e-02 2.78520286e-01
2.63429701e-01 -6.58518448e-02 -6.97417200e-01 -8.80508542e-01
-1.01891339e-01 -6.00052416e-01 -2.24771738e-01 4.13577616e-01
-1.86435297e-01 5.81062078e-01 1.06511986e+00 -1.19254982e+00
1.10000455e+00 -2.20882678e+00 4.23340887e-01 -1.01476066e-01
1.81952298e-01 2.86386963e-02 -2.41261214e-01 3.73526543e-01
-9.89306197e-02 -2.05299616e-01 -5.72757304e-01 -9.77945209e-01
-4.84576195e-01 4.48804677e-01 -6.54461324e-01 6.92323029e-01
2.73486555e-01 8.51106405e-01 -1.19056845e+00 -5.02335072e-01
8.20741415e-01 8.16378951e-01 -8.31289589e-01 6.02537155e-01
-2.87643820e-01 9.95572031e-01 8.93553253e-03 1.28487036e-01
1.10509837e+00 1.41433552e-01 -1.09394334e-01 -2.00716943e-01
-3.20078731e-01 1.89914092e-01 -1.19436944e+00 2.38635445e+00
-5.81116259e-01 8.65207613e-01 2.21465498e-01 -6.16234601e-01
5.51920950e-01 2.80051768e-01 6.47675514e-01 -5.33799231e-01
-1.64309025e-01 -1.09048277e-01 -4.79635835e-01 -7.27219045e-01
5.70597172e-01 -5.28894886e-02 3.70664388e-01 3.25105846e-01
1.83653012e-01 -4.80214238e-01 4.11863387e-01 3.25698286e-01
9.03055251e-01 6.22241139e-01 1.08110532e-01 -3.20275009e-01
5.76375723e-01 -1.67910933e-01 3.11897457e-01 6.66322589e-01
1.47715092e-01 1.28664768e+00 3.88789982e-01 -6.27642035e-01
-1.44212103e+00 -1.21557546e+00 1.00859582e-01 4.22140628e-01
1.70081466e-01 -5.22450745e-01 -1.05113506e+00 -3.36101800e-01
-2.71466255e-01 8.21296036e-01 -8.17337215e-01 5.59574924e-02
-7.08274484e-01 1.01412646e-01 1.28969267e-01 1.75794527e-01
6.09031618e-01 -8.94117117e-01 -8.30714464e-01 3.63365829e-01
-5.91770947e-01 -1.54184508e+00 -9.13503468e-01 -5.62999368e-01
-1.01493859e+00 -9.96982157e-01 -8.13642919e-01 -6.06119812e-01
8.10765982e-01 4.07363653e-01 1.32559550e+00 5.60546443e-02
-3.98598194e-01 2.56360501e-01 -8.75553414e-02 2.53366917e-01
-8.23184848e-01 -6.03615642e-01 -3.63775641e-01 4.55084801e-01
-6.20793223e-01 -6.73813641e-01 -7.78252959e-01 1.48363918e-01
-1.30839992e+00 6.56094313e-01 -9.06311721e-02 5.73094666e-01
4.10368323e-01 -4.67303768e-02 -2.29586989e-01 -8.55301261e-01
1.46682054e-01 -2.56849349e-01 -5.75069308e-01 1.02724969e-01
6.57666326e-02 7.43484497e-02 6.53181493e-01 -1.36044025e-01
-1.38388360e+00 4.20265108e-01 1.60302311e-01 -1.01814866e+00
-2.78998494e-01 -1.11293159e-01 -7.17330575e-02 -1.43078700e-01
6.46233857e-01 2.19642311e-01 -6.93952888e-02 -5.04510283e-01
7.81190097e-01 1.41192347e-01 1.07395089e+00 -5.57328939e-01
8.00797939e-01 9.99330997e-01 2.04725578e-01 -7.68411875e-01
-7.69882202e-01 -3.47370774e-01 -1.03838325e+00 -3.90440375e-01
9.51487958e-01 -9.21866834e-01 -1.24357857e-01 3.89050514e-01
-1.72736967e+00 -4.71888870e-01 -8.77602637e-01 3.22800875e-01
-1.02659357e+00 6.10674798e-01 -7.42396116e-01 -6.75275147e-01
1.76320538e-01 -1.17399216e+00 1.17748570e+00 -4.68456559e-02
-1.85662165e-01 -1.07230628e+00 3.02254289e-01 1.62056208e-01
3.81141126e-01 4.28373158e-01 5.26622415e-01 5.78589618e-01
-9.52953398e-01 4.00335521e-01 -2.92046994e-01 3.20105046e-01
3.56378376e-01 3.13341282e-02 -1.30324078e+00 -7.16814026e-02
3.19517225e-01 3.59060615e-01 8.69330108e-01 7.59486794e-01
1.03533018e+00 -4.50027615e-01 -4.54660691e-02 9.54878628e-01
1.60011840e+00 1.71912953e-01 1.23593771e+00 2.49792580e-02
8.95848751e-01 7.88698912e-01 2.98452944e-01 5.69935262e-01
1.81616604e-01 8.64585578e-01 4.08241689e-01 -1.50410205e-01
-8.49571109e-01 -4.41686094e-01 3.89631540e-01 2.52349555e-01
-1.71239734e-01 -2.45330572e-01 -4.90349710e-01 8.13091516e-01
-1.93243480e+00 -1.19481027e+00 -3.43889296e-01 2.27515411e+00
7.81743348e-01 -2.69017577e-01 -4.27596644e-03 5.29435351e-02
7.41756678e-01 2.67671466e-01 -1.43105596e-01 -3.15892786e-01
-1.03953764e-01 2.91977972e-01 4.03317392e-01 1.26698720e+00
-8.64796579e-01 7.96137333e-01 6.83722591e+00 5.67354798e-01
-8.41423154e-01 2.90219754e-01 8.75299990e-01 -8.95464495e-02
-4.50585544e-01 3.02044630e-01 -2.57961035e-01 4.75977004e-01
7.65745640e-01 -1.17997125e-01 8.02145123e-01 3.00989777e-01
7.19556510e-01 -2.70158499e-01 -1.28211594e+00 1.08700991e+00
1.89277321e-01 -2.01981091e+00 1.43693522e-01 -1.48798234e-03
1.23001158e+00 -3.87875885e-01 -2.36254781e-01 -4.94586378e-01
3.16173793e-03 -8.42612267e-01 1.31062186e+00 8.47388089e-01
1.18589175e+00 -4.58349019e-01 1.58211976e-01 2.46209845e-01
-1.06276143e+00 2.62916446e-01 -1.91485539e-01 -1.90605298e-01
7.53634930e-01 8.23665261e-01 -4.13582593e-01 7.05326855e-01
6.31077290e-01 1.13160908e+00 -2.53993720e-01 1.08979762e+00
-1.58732802e-01 1.84762597e-01 6.38515800e-02 1.18024552e+00
-7.54786357e-02 -5.04470825e-01 6.23282373e-01 1.36968434e+00
4.22649056e-01 1.27107859e-01 3.51025946e-02 1.33394444e+00
-4.34599221e-02 -3.56531113e-01 -9.07475114e-01 5.75582802e-01
1.05191939e-01 9.86966193e-01 -7.23293066e-01 -5.80449700e-01
-3.11456710e-01 1.53822601e+00 -1.56014189e-01 6.23323917e-01
-7.09615231e-01 -6.88396171e-02 8.00139070e-01 5.26312649e-01
1.57916099e-01 -2.86978573e-01 -3.60299915e-01 -1.33987033e+00
8.43735710e-02 -4.59276080e-01 1.11268535e-01 -1.22679651e+00
-9.51385856e-01 5.24656117e-01 1.96641758e-01 -1.39712393e+00
-3.98296863e-01 -1.50871769e-01 -5.76940835e-01 1.12214553e+00
-1.57621729e+00 -8.73544216e-01 -4.38796729e-01 7.97351778e-01
8.98975492e-01 3.11753124e-01 4.21586394e-01 3.54373395e-01
7.59103894e-02 -9.94048640e-02 -8.97679180e-02 -3.15671898e-02
7.50722051e-01 -8.56849551e-01 6.98457658e-01 1.36859512e+00
1.41250014e-01 2.48957649e-01 8.08065534e-01 -6.30024493e-01
-1.22012830e+00 -1.29813945e+00 9.98583674e-01 -4.81383562e-01
1.98771700e-01 -3.45037550e-01 -8.91450465e-01 7.65483558e-01
5.92191219e-01 4.21840847e-01 -8.53967518e-02 -9.08947408e-01
-2.86270708e-01 -1.43630886e-02 -1.32557487e+00 4.06654418e-01
1.27222157e+00 -7.35033751e-01 -2.99258709e-01 2.80404299e-01
7.51267374e-01 -7.81594455e-01 -5.15695095e-01 -1.34737343e-01
4.54292119e-01 -1.43479586e+00 1.22052157e+00 -4.19782013e-01
9.32421863e-01 -7.14860082e-01 -2.02438105e-02 -1.43592644e+00
-4.03666943e-01 -1.10493433e+00 -3.28495532e-01 1.04579556e+00
-2.17233121e-01 -1.23220466e-01 7.35923409e-01 6.45544827e-01
-2.89166898e-01 -1.43825486e-01 -9.02002573e-01 -5.87131381e-01
-3.23679000e-01 -4.48762119e-01 2.71524549e-01 9.32635009e-01
-3.37561429e-01 -1.83719873e-01 -7.33386636e-01 -5.11662737e-02
8.38446736e-01 -3.28039601e-02 7.05456555e-01 -7.55836666e-01
-4.68258470e-01 -2.91432619e-01 -3.63436699e-01 -1.17086601e+00
1.13702409e-01 -5.04807472e-01 7.89529905e-02 -1.57454610e+00
-2.15465352e-02 -9.97014716e-02 3.89484882e-01 7.39140213e-02
-1.28431525e-02 4.84926015e-01 4.65358406e-01 2.56920993e-01
-2.84948677e-01 3.62792999e-01 1.59975183e+00 3.76843289e-02
-2.63072968e-01 -4.24725860e-01 -3.67357254e-01 5.61541080e-01
2.91735977e-01 -3.79135698e-01 -5.99837482e-01 -8.16108346e-01
-4.75593247e-02 6.16133034e-01 4.95751560e-01 -9.74814117e-01
7.09832385e-02 -3.34723055e-01 5.37017524e-01 -2.10344344e-01
4.89306509e-01 -8.28087687e-01 6.76066816e-01 3.96645874e-01
-3.67167681e-01 1.16431303e-01 2.21589193e-01 4.39867139e-01
-2.61282206e-01 -1.36777431e-01 9.02750850e-01 -2.43323550e-01
-6.35508955e-01 2.20819667e-01 -1.36736885e-01 1.78630695e-01
8.59794974e-01 -4.46078837e-01 -9.10441726e-02 -7.10263193e-01
-8.13031018e-01 -2.27667227e-01 9.10953164e-01 3.77732188e-01
6.77660465e-01 -1.27308047e+00 -9.13508654e-01 6.24919116e-01
-3.42862487e-01 1.90476120e-01 4.76417482e-01 6.12885535e-01
-1.14333653e+00 2.30837628e-01 -5.11595547e-01 -6.87054396e-01
-9.24938262e-01 5.74727774e-01 5.45576274e-01 -3.44134718e-02
-7.78689265e-01 6.19722247e-01 7.91919291e-01 2.39021540e-01
-2.77588535e-02 -3.72308522e-01 3.81031305e-01 -4.04607058e-01
7.48601079e-01 4.25748795e-01 1.11219786e-01 -8.07912946e-01
-1.93232432e-01 7.55520046e-01 3.64576876e-01 -5.61405361e-01
1.20885551e+00 -4.19879407e-01 -2.12323412e-01 1.69050202e-01
1.18704069e+00 1.90175489e-01 -2.04172754e+00 3.08538042e-02
-4.64326352e-01 -1.30651689e+00 2.44897082e-02 -3.82141471e-01
-1.27315140e+00 8.12294304e-01 1.45740747e-01 3.69801074e-02
1.32674229e+00 -1.44800618e-01 6.73356116e-01 -5.81472218e-01
2.87920386e-01 -6.21901453e-01 -1.19860142e-01 3.67720991e-01
1.02933395e+00 -5.35044372e-01 1.45226792e-02 -8.33053112e-01
-3.70354176e-01 1.14240575e+00 1.58567250e-01 -4.73585099e-01
5.78197241e-01 3.64118129e-01 -1.65702850e-01 1.57827303e-01
-7.18535662e-01 1.51139647e-01 4.16794628e-01 8.68934989e-01
3.97461414e-01 -3.59856814e-01 2.78509334e-02 -3.67652386e-01
7.97886997e-02 3.96905631e-01 8.77495170e-01 7.59761810e-01
-7.69890174e-02 -1.12835789e+00 -6.53187811e-01 -1.40963346e-01
-1.06812291e-01 -4.74374779e-02 -3.57252397e-02 4.59963471e-01
3.87531728e-01 9.92676556e-01 3.51715654e-01 1.81762204e-02
3.63381386e-01 -1.14490792e-01 9.77113903e-01 -4.77746844e-01
-3.59879732e-01 3.49840790e-01 -5.83588742e-02 -1.16750014e+00
-6.82519734e-01 -7.16710508e-01 -9.74228680e-01 -3.61763567e-01
2.45526731e-01 -9.12265331e-02 3.20739716e-01 8.25826168e-01
6.66067898e-01 5.03621399e-01 7.23395944e-01 -1.37699628e+00
3.16692561e-01 -5.73744774e-01 -4.89575326e-01 8.74356866e-01
6.53904855e-01 -2.87776887e-01 -2.22691283e-01 1.02566600e+00] | [10.739469528198242, -1.3715078830718994] |
582f5d42-e9b0-49b4-9e3c-c775391d9fa8 | pca-event-based-otical-flow-for-visual | 2105.03760 | null | https://arxiv.org/abs/2105.03760v2 | https://arxiv.org/pdf/2105.03760v2.pdf | PCA Event-Based Optical Flow for Visual Odometry | With the advent of neuromorphic vision sensors such as event-based cameras, a paradigm shift is required for most computer vision algorithms. Among these algorithms, optical flow estimation is a prime candidate for this process considering that it is linked to a neuromorphic vision approach. Usage of optical flow is widespread in robotics applications due to its richness and accuracy. We present a Principal Component Analysis (PCA) approach to the problem of event-based optical flow estimation. In this approach, we examine different regularization methods which efficiently enhance the estimation of the optical flow. We show that the best variant of our proposed method, dedicated to the real-time context of visual odometry, is about two times faster compared to state-of-the-art implementations while significantly improves optical flow accuracy. | ['Samia Bouchafa', 'David Roussel', 'Fabien Bonardi', 'Mahmoud Z. Khairallah'] | 2021-05-08 | null | null | null | null | ['event-based-optical-flow'] | ['computer-vision'] | [ 1.54337183e-01 -3.62264693e-01 2.08206370e-01 5.62888496e-02
2.48657942e-01 -3.11298519e-01 8.14016402e-01 -7.09459707e-02
-9.67714548e-01 8.56275380e-01 1.10987142e-01 2.61906922e-01
-1.79160953e-01 -5.59068024e-01 -5.71665883e-01 -5.80748260e-01
1.10582314e-01 4.76256832e-02 4.96437132e-01 3.19307186e-02
7.80788600e-01 7.62009323e-01 -1.88600957e+00 -3.00641030e-01
4.51612502e-01 1.04971194e+00 3.15574139e-01 7.34933436e-01
-1.80930004e-01 7.05108643e-01 -1.51075542e-01 -2.23828718e-01
3.11252773e-01 -3.94969046e-01 -5.32028675e-01 8.62107649e-02
7.72853434e-01 -1.20653413e-01 -7.72701919e-01 1.21500969e+00
3.90762866e-01 3.17183256e-01 5.30770600e-01 -1.22032726e+00
-3.90610814e-01 -1.47146881e-01 -3.06490898e-01 7.20907092e-01
4.82648283e-01 1.27845660e-01 6.75771773e-01 -8.15288484e-01
1.04058957e+00 1.02115738e+00 2.70089835e-01 5.95851064e-01
-1.21373308e+00 -1.08841605e-01 -2.88165659e-01 8.30621362e-01
-8.92511785e-01 -5.16012609e-01 8.51806045e-01 -6.12033725e-01
1.03405070e+00 -2.81783432e-01 1.23205137e+00 1.03287232e+00
6.67450070e-01 5.60989261e-01 1.15629828e+00 -2.70681381e-01
5.73974490e-01 -2.16768399e-01 2.88085312e-01 7.12795615e-01
4.42886740e-01 2.17188016e-01 -9.54412043e-01 1.67471409e-01
8.20672750e-01 9.93577316e-02 -4.10123765e-01 -5.16340494e-01
-1.26150227e+00 5.29933929e-01 3.16257745e-01 8.01239237e-02
-6.00091994e-01 5.04619300e-01 3.06292534e-01 -7.61315525e-02
-1.19658366e-01 3.60627711e-01 1.10168114e-01 -6.27927601e-01
-8.12764049e-01 3.27391446e-01 9.45612431e-01 5.62743843e-01
8.88063490e-01 2.50556111e-01 1.21167168e-01 3.77257198e-01
5.53768039e-01 3.20704848e-01 8.15611720e-01 -1.24570441e+00
-1.50521591e-01 4.92853582e-01 3.47704887e-02 -9.82812285e-01
-3.18074226e-01 -8.44760612e-02 -5.75457513e-01 5.11534452e-01
5.76044858e-01 1.01682603e-01 -5.49921155e-01 1.52503204e+00
2.30881333e-01 7.95807004e-01 5.09237424e-02 1.03097713e+00
2.50544578e-01 4.94385988e-01 -1.87239230e-01 -4.15562004e-01
1.26364565e+00 -5.96285939e-01 -7.78471649e-01 -2.39887655e-01
-2.16957480e-01 -6.43560708e-01 4.73969996e-01 6.05181813e-01
-8.11792612e-01 -3.66369784e-01 -1.30754793e+00 -1.35306537e-01
-3.20152223e-01 -2.06297085e-01 8.72688234e-01 6.64961278e-01
-1.13573098e+00 7.15015769e-01 -1.05877852e+00 -8.08368742e-01
3.39065939e-01 3.52275193e-01 -5.79972684e-01 1.99664176e-01
-5.52617133e-01 1.06764698e+00 1.50598913e-01 7.30514526e-02
-8.17704320e-01 -4.37733322e-01 -6.44700766e-01 -7.68350214e-02
-1.55228660e-01 -8.14253807e-01 9.02786672e-01 -7.77685225e-01
-1.96070576e+00 8.03138316e-01 -3.19862902e-01 -8.33320975e-01
4.08142686e-01 -2.63058901e-01 -5.81816696e-02 5.07970870e-01
-2.91658312e-01 8.05000186e-01 1.15473557e+00 -5.40347874e-01
-4.58720505e-01 -5.95177710e-01 -1.05464971e-02 -3.36916745e-02
-4.67931330e-01 -1.58878401e-01 -4.66664098e-02 -6.35531619e-02
1.18749991e-01 -1.08523476e+00 -1.83298290e-01 5.39849579e-01
2.34845117e-01 1.80316642e-02 9.91424084e-01 -1.57376945e-01
7.21084297e-01 -1.85025597e+00 3.29260439e-01 -3.85324299e-01
1.71101779e-01 3.76832873e-01 2.94322789e-01 1.75260440e-01
2.30278969e-01 -6.56436801e-01 -3.48611116e-01 -3.10626864e-01
-2.60845006e-01 3.34458321e-01 -3.15231085e-01 8.58013153e-01
2.96963453e-01 7.55993187e-01 -9.36283231e-01 -2.92168170e-01
7.95083761e-01 8.51673424e-01 -6.73747301e-01 9.00197327e-02
1.73157994e-02 7.36821532e-01 -2.17584670e-01 2.11418152e-01
5.21761298e-01 1.42071739e-01 -7.40438253e-02 -2.43561253e-01
-6.08414769e-01 -1.01569975e-02 -1.33286119e+00 2.07661891e+00
-3.16955775e-01 1.23877299e+00 -9.12338346e-02 -1.14642930e+00
9.94561076e-01 7.64544085e-02 7.68600643e-01 -6.02794886e-01
4.54144031e-01 2.10719943e-01 1.09943658e-01 -6.51366353e-01
4.75740790e-01 1.22328186e-02 5.85647643e-01 2.13276088e-01
5.56279361e-01 -3.44182216e-02 3.82814527e-01 -7.42330551e-02
1.27830195e+00 5.93946993e-01 5.98016083e-01 -3.92844230e-01
6.98968530e-01 -3.11980128e-01 5.52639723e-01 2.74153173e-01
-8.87995005e-01 4.88876253e-01 2.79846758e-01 -6.31142616e-01
-9.59057689e-01 -9.91250336e-01 -3.17639738e-01 1.75276086e-01
1.87875316e-01 -1.44963831e-01 -7.01964319e-01 2.16661468e-02
-8.12467411e-02 -9.52311754e-02 -3.84159774e-01 -4.76051196e-02
-5.80362201e-01 -7.25338161e-01 3.53058994e-01 1.25165045e-01
7.08748579e-01 -1.14024210e+00 -1.54646587e+00 5.63843429e-01
2.45374367e-01 -1.52014863e+00 8.98475274e-02 5.99459857e-02
-1.21973395e+00 -1.09959340e+00 -6.72779560e-01 -4.13137615e-01
4.39441025e-01 1.94641113e-01 5.75428188e-01 -5.26041031e-01
-6.99821889e-01 6.74582601e-01 -1.61435828e-01 -2.93361843e-01
4.26092446e-02 -3.66476536e-01 1.55954883e-01 5.48953176e-01
4.83075798e-01 -1.17861068e+00 -7.82958865e-01 -3.68955404e-01
-8.81386757e-01 -3.84713143e-01 5.27803779e-01 5.35799861e-01
4.76517946e-01 -6.68155789e-01 1.04450755e-01 -6.27059519e-01
5.51477730e-01 -4.22219664e-01 -8.53534222e-01 -2.37931177e-01
-6.36265934e-01 4.26567614e-01 7.02346563e-01 -4.09614742e-01
-8.13173294e-01 6.11946166e-01 1.35738671e-01 -5.63335299e-01
-2.43309960e-01 1.84197798e-01 3.07502121e-01 -7.08811581e-01
7.19199479e-01 3.55314642e-01 1.27110049e-01 -1.66730896e-01
1.97338521e-01 2.49730095e-01 8.02694499e-01 -2.71278143e-01
2.92585731e-01 9.31282282e-01 5.94869614e-01 -1.22597277e+00
-1.31983221e-01 -6.99877620e-01 -6.59677744e-01 -6.70162201e-01
7.90579498e-01 -4.31964517e-01 -1.14175999e+00 7.81074584e-01
-1.50805140e+00 2.02037320e-01 -1.57557070e-01 9.32156742e-01
-8.99672568e-01 5.03616512e-01 -4.51957345e-01 -9.84752715e-01
-2.11995691e-01 -1.31019461e+00 6.48894846e-01 7.18720734e-01
2.17017367e-01 -8.98241103e-01 6.50382102e-01 -1.24765061e-01
3.94939572e-01 2.32172817e-01 3.23502123e-01 7.41446996e-03
-7.38284528e-01 -5.23185730e-02 -2.79198110e-01 2.06294104e-01
-2.09562629e-01 2.19152987e-01 -1.31460643e+00 6.66365772e-02
2.77054429e-01 1.53739110e-01 8.59503567e-01 5.45784771e-01
7.34739661e-01 2.24244833e-01 -6.68239370e-02 6.30827308e-01
1.78633618e+00 2.42531851e-01 9.04268920e-01 2.82506257e-01
5.90189278e-01 5.23304045e-01 -1.49322329e-02 6.62127912e-01
3.11092377e-01 7.98678517e-01 7.14320600e-01 7.94903934e-01
-3.69266361e-01 -1.07336126e-03 4.47936118e-01 7.66849160e-01
-4.11844909e-01 2.30035618e-01 -7.32078910e-01 6.15441799e-01
-1.92958832e+00 -9.69157875e-01 -2.45319471e-01 2.37450099e+00
2.63667017e-01 8.37403908e-02 -2.73710769e-02 4.17125076e-01
6.03271961e-01 1.66909784e-01 -4.15768862e-01 -5.30269384e-01
-3.78568545e-02 4.90020365e-01 3.60856533e-01 4.53431666e-01
-8.62567365e-01 7.86253333e-01 6.52391100e+00 4.81494144e-03
-1.38922012e+00 1.46510243e-01 -3.00230414e-01 1.52249262e-01
1.83772177e-01 1.83716491e-01 -7.49891400e-01 6.10114217e-01
1.05457997e+00 -1.27676636e-01 6.97313309e-01 6.21028721e-01
2.20802158e-01 -4.65254664e-01 -1.00968039e+00 1.48495114e+00
3.43841702e-01 -1.34798622e+00 -1.66571718e-02 2.25930348e-01
5.39294422e-01 1.24601744e-01 -1.36389866e-01 -3.03275973e-01
-3.94542485e-01 -5.46213448e-01 5.24943769e-01 8.84954631e-01
1.05786696e-01 -4.40604031e-01 3.38365316e-01 7.99922422e-02
-1.01752436e+00 -1.53313830e-01 -5.74975252e-01 -6.44776642e-01
3.43624830e-01 8.57942820e-01 -3.41097325e-01 1.13342926e-01
5.61151445e-01 1.26005626e+00 -4.25122559e-01 1.60640776e+00
-8.55487585e-02 1.83207765e-01 -3.56745034e-01 -3.37558389e-01
1.24814928e-01 -4.92154658e-01 9.06727672e-01 1.01141357e+00
3.76710892e-01 -2.82570183e-01 -3.15654337e-01 8.48112643e-01
5.57756498e-02 -9.17025656e-02 -9.37811255e-01 3.77692394e-02
2.62350649e-01 1.46699333e+00 -7.11974502e-01 -9.23456401e-02
-5.50982714e-01 1.06040120e+00 4.07363951e-01 1.74886584e-01
-4.26462173e-01 -4.79087085e-01 8.32854927e-01 -1.83101431e-01
1.61833972e-01 -7.82906413e-01 -6.67569339e-02 -1.35504007e+00
2.82008722e-02 -1.66573316e-01 -1.45073280e-01 -8.04791868e-01
-8.50714386e-01 4.87164497e-01 -4.25977647e-01 -1.30545807e+00
-3.61762345e-01 -1.24708974e+00 -3.89161110e-01 4.36477602e-01
-1.79260218e+00 -6.46380961e-01 -6.00235045e-01 8.62610757e-01
4.15859848e-01 -1.80430070e-01 7.98419595e-01 4.16350812e-01
-3.79828572e-01 -1.55941606e-01 2.71629710e-02 -1.14137709e-01
6.03835046e-01 -1.17194653e+00 1.44997522e-01 1.48684120e+00
6.14534616e-01 4.73884821e-01 8.64807725e-01 -1.90071866e-01
-1.93514681e+00 -5.46629608e-01 7.22225487e-01 -2.57725954e-01
8.82296145e-01 -4.22695838e-02 -5.51956296e-01 4.55579013e-01
1.98777333e-01 5.26909232e-01 3.10324460e-01 -5.30498862e-01
-2.38555908e-01 -4.02421117e-01 -1.04527760e+00 5.20111501e-01
9.26475406e-01 -7.67598808e-01 -6.17514312e-01 -1.51397318e-01
-1.90446340e-02 -1.18149236e-01 -7.29347765e-01 -4.48698588e-02
7.98878431e-01 -1.36346519e+00 8.63122940e-01 -2.39443287e-01
2.63334632e-01 -4.41787660e-01 -7.70458877e-02 -1.20393395e+00
-8.09293315e-02 -8.98703277e-01 -6.35477543e-01 7.21206248e-01
-3.46901417e-01 -7.75882244e-01 8.81014526e-01 3.70178998e-01
-6.57407194e-02 -2.47071028e-01 -1.26499081e+00 -5.75647473e-01
-5.50056219e-01 -5.04283965e-01 -2.88340151e-01 5.29386580e-01
1.86081916e-01 1.90777317e-01 -2.91446418e-01 -1.10218845e-01
9.35714006e-01 -1.13267198e-01 3.96706194e-01 -1.51442277e+00
-3.35242987e-01 -4.98369873e-01 -1.55759251e+00 -1.01555908e+00
2.58112162e-01 -7.67423868e-01 9.05475840e-02 -1.17028725e+00
-3.05648986e-02 2.89443940e-01 -3.24580431e-01 -2.26281613e-01
1.58047184e-01 4.93028313e-01 3.30764771e-01 1.39109969e-01
-3.32477927e-01 4.48419780e-01 1.07551599e+00 1.57429487e-01
-5.32443635e-02 -1.62278205e-01 2.17646100e-02 6.25061274e-01
4.95631248e-01 -3.70485604e-01 -2.81756163e-01 -4.97070372e-01
1.45128161e-01 -8.46399665e-02 6.23921096e-01 -1.69922578e+00
7.64192641e-01 1.65077314e-01 2.27407992e-01 7.42950961e-02
5.76996028e-01 -7.30180800e-01 -3.73423100e-02 7.82810092e-01
-3.52053493e-02 2.09996209e-01 -1.40543012e-02 7.40032256e-01
-3.67121726e-01 -4.45658326e-01 9.00471151e-01 -2.65336901e-01
-1.12497330e+00 4.03733045e-01 -7.64365435e-01 -5.48889525e-02
8.76364827e-01 -3.96077693e-01 -4.72978443e-01 5.98055171e-03
-3.86295438e-01 -5.14119506e-01 2.15444550e-01 2.74393111e-01
7.62400508e-01 -1.04115939e+00 -2.78250247e-01 2.09832907e-01
8.95231068e-02 -4.12117004e-01 -7.83697069e-02 9.90807295e-01
-9.14761961e-01 4.96953994e-01 -1.08966839e+00 -7.00671315e-01
-8.77298474e-01 4.58789647e-01 3.13573420e-01 1.56660512e-01
-7.28003442e-01 4.96382177e-01 -3.13728094e-01 2.57911146e-01
1.59800947e-01 -1.91199839e-01 -4.51902866e-01 -4.29764464e-02
6.08785331e-01 8.67479861e-01 4.29059044e-02 -6.67568684e-01
-5.67420185e-01 9.17188287e-01 4.92841750e-01 -5.29894352e-01
1.18738341e+00 -1.57224685e-01 -1.98019534e-01 5.99802315e-01
1.01597428e+00 -3.36802691e-01 -1.59261966e+00 1.26798421e-01
2.38701478e-01 -4.27510500e-01 3.43651772e-01 -1.41107082e-01
-9.74084675e-01 1.36729169e+00 8.64353001e-01 1.91513747e-01
1.01233125e+00 -6.87374294e-01 6.90544307e-01 5.58318973e-01
6.87073648e-01 -9.35353458e-01 6.28951937e-02 5.39404213e-01
4.67580080e-01 -1.07604837e+00 -6.94666523e-03 -1.70488194e-01
-1.26341403e-01 1.53938746e+00 3.98582101e-01 -9.40747261e-01
7.53779054e-01 2.49726132e-01 -2.12382063e-01 1.80905581e-01
-6.00559533e-01 -4.06394333e-01 -8.03254098e-02 8.15539420e-01
2.10450768e-01 -2.24601641e-01 -6.69927776e-01 -1.81620553e-01
9.41263139e-02 4.22552794e-01 8.99512827e-01 8.20036292e-01
-4.54289585e-01 -9.84977067e-01 -1.65890023e-01 1.42779008e-01
-4.09997404e-01 1.33635357e-01 -1.05377249e-01 3.73126775e-01
-1.05339386e-01 7.12858260e-01 3.68132778e-02 -2.28300974e-01
2.26543427e-01 1.22074224e-01 1.13201594e+00 -1.36568248e-01
-3.35267603e-01 -1.79899275e-01 -3.97237271e-01 -9.48642850e-01
-1.04571080e+00 -6.63843751e-01 -1.33261764e+00 -1.79792315e-01
1.72590122e-01 -2.69360840e-01 1.23052323e+00 1.05233765e+00
4.54646736e-01 3.02834779e-01 3.61406326e-01 -1.05521691e+00
4.17914316e-02 -6.36292219e-01 -4.67351764e-01 4.45772439e-01
3.12614620e-01 -9.22287107e-01 -2.28864655e-01 2.66104668e-01] | [8.650249481201172, -1.3145411014556885] |
bebb5f95-90a4-4bed-b94c-dfcd63651275 | on-residual-cnn-in-text-dependent-speaker | 1705.10134 | null | http://arxiv.org/abs/1705.10134v2 | http://arxiv.org/pdf/1705.10134v2.pdf | On Residual CNN in text-dependent speaker verification task | Deep learning approaches are still not very common in the speaker
verification field. We investigate the possibility of using deep residual
convolutional neural network with spectrograms as an input features in the
text-dependent speaker verification task. Despite the fact that we were not
able to surpass the baseline system in quality, we achieved a quite good
results for such a new approach getting an 5.23% ERR on the RSR2015 evaluation
part. Fusion of the baseline and proposed systems outperformed the best
individual system by 18% relatively. | ['Egor Malykh', 'Oleg Kudashev', 'Sergey Novoselov'] | 2017-05-29 | null | null | null | null | ['text-dependent-speaker-verification'] | ['speech'] | [-2.28996705e-02 1.00603141e-01 5.31642318e-01 -5.96709430e-01
-1.12552083e+00 -3.53753835e-01 8.90412390e-01 -2.89340347e-01
-7.46958315e-01 6.93018734e-01 3.93515080e-01 -3.64856511e-01
1.47157282e-01 -2.25308850e-01 -6.31259024e-01 -8.99392903e-01
6.97671697e-02 1.59213141e-01 -8.01150948e-02 -5.25738537e-01
1.48264587e-01 3.92920941e-01 -1.68258321e+00 2.74961084e-01
4.96197551e-01 9.65185404e-01 -2.88024962e-01 7.89877892e-01
1.60422549e-01 3.40590417e-01 -1.13489842e+00 -3.86898547e-01
1.77209362e-01 -2.75013834e-01 -8.55251014e-01 -3.18460435e-01
6.76061749e-01 -1.96464434e-01 -3.21459085e-01 9.28776681e-01
1.16777003e+00 6.99985623e-02 4.39945042e-01 -9.05189455e-01
-5.06998777e-01 1.13863802e+00 -1.59855574e-01 4.86246318e-01
7.12069988e-01 -3.20884734e-02 7.38454461e-01 -9.84997332e-01
3.41325611e-01 1.13650298e+00 6.55489445e-01 7.50502646e-01
-1.03513825e+00 -8.28295290e-01 -2.25918666e-01 4.14553821e-01
-1.58931494e+00 -1.28803205e+00 6.09513760e-01 -6.08436875e-02
9.76391137e-01 2.32022569e-01 -4.81285974e-02 1.46753252e+00
-2.75222838e-01 6.14849269e-01 1.21655858e+00 -5.39678693e-01
-7.61908963e-02 4.98973876e-01 3.85636896e-01 3.22250515e-01
-1.62933409e-01 6.05786562e-01 -7.79501975e-01 7.39914775e-02
-2.48008501e-02 -7.64896810e-01 -5.14020622e-01 5.75754881e-01
-1.26269948e+00 7.86178231e-01 4.28144842e-01 1.02775347e+00
-2.94847578e-01 -2.67379712e-02 6.20926201e-01 6.96627498e-01
4.18425560e-01 2.43443280e-01 -4.93747026e-01 -2.55745798e-01
-1.36688626e+00 1.57092974e-01 6.97735131e-01 3.56831938e-01
2.73030818e-01 7.28542566e-01 -1.24639310e-01 8.25842261e-01
1.88110620e-01 4.22802627e-01 8.44598055e-01 -2.89831758e-01
2.87171870e-01 5.04669286e-02 6.81355149e-02 -4.95005786e-01
-3.99049401e-01 -6.66087866e-01 -8.13868880e-01 1.42005056e-01
4.91127372e-01 -4.88568187e-01 -1.11286449e+00 1.62443876e+00
-1.09881228e-02 2.74114132e-01 3.39902669e-01 9.03323710e-01
1.31220996e+00 7.16052830e-01 -1.63240090e-01 -8.91158879e-02
1.06741154e+00 -6.70112371e-01 -1.02704585e+00 2.80675083e-01
3.96137387e-01 -1.05702078e+00 7.57296264e-01 6.65730298e-01
-9.62456763e-01 -8.34076703e-01 -1.42817533e+00 2.60272533e-01
-5.36968768e-01 1.39810175e-01 3.56306642e-01 1.44792068e+00
-1.73914897e+00 6.23241305e-01 -3.26947540e-01 -2.62097627e-01
1.49928734e-01 4.90324050e-01 -6.54145718e-01 3.20464939e-01
-1.56102467e+00 1.25821507e+00 2.25643024e-01 3.35707575e-01
-1.03068984e+00 -5.55689275e-01 -7.25056112e-01 7.57931471e-02
-1.31305531e-01 -6.22139778e-03 1.29766977e+00 -6.28361166e-01
-2.02926207e+00 9.79267538e-01 -2.82466948e-01 -8.82327497e-01
6.71402276e-01 -1.23228960e-01 -1.09838891e+00 -3.64856511e-01
-6.70015514e-01 4.01523709e-01 9.28290367e-01 -9.01413918e-01
-3.39821309e-01 -2.41365746e-01 -2.18103856e-01 -1.69885546e-01
-1.08016185e-01 3.37380171e-01 3.14615041e-01 -2.43137971e-01
5.49596064e-02 -8.08465838e-01 4.59655225e-02 -6.59667373e-01
-4.58638340e-01 -4.54549491e-01 8.02855074e-01 -9.77039278e-01
7.99715579e-01 -2.14865732e+00 -2.17713621e-02 -8.87180120e-02
-3.29943419e-01 6.81474805e-01 -1.20877840e-01 2.48376653e-01
-4.50462282e-01 1.39796779e-01 -1.58139735e-01 -6.98657691e-01
2.50157267e-01 -3.89476627e-01 -2.61923939e-01 6.97040737e-01
2.12444305e-01 5.25139928e-01 -3.08441877e-01 -4.29711193e-02
3.92745167e-01 1.01554000e+00 -6.60249144e-02 1.18105136e-01
3.60861719e-01 4.50463802e-01 3.50684822e-01 4.93251324e-01
9.57468092e-01 5.53373635e-01 -3.17100912e-01 3.05858254e-02
-2.31244698e-01 9.61677015e-01 -1.37604821e+00 1.65514386e+00
-5.31698823e-01 1.19145620e+00 2.18497992e-01 -1.09499514e+00
1.21711659e+00 1.07243538e+00 1.29003286e-01 -5.73075235e-01
2.15394899e-01 3.88998121e-01 4.41918612e-01 -2.09155858e-01
6.12601459e-01 -6.28082037e-01 -1.94581971e-02 3.14245313e-01
4.35115278e-01 -5.30503597e-03 -2.98657417e-01 -1.50850132e-01
8.40978503e-01 -2.54152596e-01 -1.91633077e-03 -3.54401857e-01
1.14752245e+00 -6.23328149e-01 6.50064275e-02 4.99230146e-01
-6.47240758e-01 8.47052753e-01 8.93267319e-02 -2.75916398e-01
-6.48673356e-01 -8.25958252e-01 -4.69521880e-01 9.83555794e-01
-4.50275362e-01 -5.20620607e-02 -8.47059965e-01 -6.50757253e-01
-2.76572913e-01 7.35058725e-01 -6.43239439e-01 9.74597558e-02
-4.32545871e-01 -6.78293288e-01 1.27351308e+00 1.90891936e-01
4.88627166e-01 -1.08910787e+00 1.43785030e-01 3.37861568e-01
-1.82507545e-01 -1.21415186e+00 -2.61711955e-01 4.02112752e-01
-5.35299599e-01 -4.29377645e-01 -1.05623341e+00 -7.59297132e-01
-1.03079841e-01 -1.66585505e-01 8.12138319e-01 1.32582914e-02
5.89502864e-02 -9.18747708e-02 -3.63971055e-01 -5.43852866e-01
-8.98619890e-01 3.16373557e-01 3.18878531e-01 1.35456294e-01
6.39723718e-01 -4.24047351e-01 -3.79368544e-01 1.98047206e-01
-4.82231855e-01 -7.88366020e-01 2.04590499e-01 7.67476022e-01
-4.01168555e-01 -2.10282028e-01 1.04733801e+00 -2.85570174e-01
5.58945954e-01 -8.22763741e-02 -5.30026913e-01 -1.39007151e-01
-4.62002188e-01 2.13133365e-01 3.22649926e-01 -2.33643249e-01
-9.67577755e-01 7.34874010e-02 -9.47065651e-01 2.13932041e-02
-6.78238332e-01 2.62881279e-01 -1.27933323e-01 -1.56858712e-01
7.99840152e-01 2.95900434e-01 -6.32366464e-02 -6.38576925e-01
8.58138651e-02 1.24345589e+00 3.88224185e-01 1.10669933e-01
8.17612410e-01 -7.31408447e-02 -5.30099332e-01 -1.01074469e+00
-2.68187672e-01 -5.39849997e-01 -5.19071281e-01 -1.47089854e-01
6.98415339e-01 -1.07809210e+00 -9.81863379e-01 7.70642102e-01
-1.37265301e+00 1.03396386e-01 -6.38222173e-02 5.16574621e-01
-1.71996772e-01 3.25032413e-01 -4.99861211e-01 -1.10339308e+00
-5.36735475e-01 -1.36431265e+00 9.90516603e-01 1.83984414e-02
-1.12746648e-01 -8.31603408e-01 1.68831870e-01 5.11110544e-01
1.00533271e+00 -1.06306873e-01 1.85533091e-01 -1.14417458e+00
9.83160827e-03 -5.48083484e-01 7.71549121e-02 6.98097587e-01
1.66308850e-01 4.96667959e-02 -1.99651170e+00 -3.49787802e-01
1.75936833e-01 -1.60367772e-01 1.05664575e+00 2.47136325e-01
7.53763855e-01 8.88215005e-02 1.94483578e-01 2.58176237e-01
1.15385091e+00 6.75803348e-02 8.39402497e-01 9.62360948e-02
1.33725375e-01 5.65018594e-01 3.03824507e-02 1.06143907e-01
6.85496032e-02 1.03126884e+00 1.38444826e-01 6.95992634e-03
-4.44611549e-01 1.51549503e-01 8.49059045e-01 8.74850452e-01
-1.53487682e-01 -2.34587044e-01 -1.03457355e+00 6.40284657e-01
-1.17536473e+00 -1.11589801e+00 -3.01145375e-01 2.29995298e+00
6.14783287e-01 1.97534069e-01 2.18627498e-01 8.31786990e-01
8.82040739e-01 3.51660013e-01 -1.62146855e-02 -8.59900713e-01
-3.10684800e-01 4.79261935e-01 2.67839909e-01 7.85387814e-01
-1.21888578e+00 8.06469142e-01 7.30288887e+00 5.53810716e-01
-1.56473827e+00 3.45072240e-01 2.46201292e-01 6.19693622e-02
1.48931026e-01 -5.07409155e-01 -8.89469087e-01 3.98715436e-01
1.82528567e+00 -1.18760444e-01 3.17348599e-01 5.57781279e-01
1.06197037e-01 9.48262513e-02 -9.54969049e-01 1.11016691e+00
2.76758969e-01 -1.05851901e+00 -2.60828584e-01 1.75920144e-01
5.71362436e-01 3.33884299e-01 1.27270639e-01 5.60190916e-01
7.89900795e-02 -1.31755340e+00 5.96025348e-01 2.15759709e-01
6.00626409e-01 -7.85342455e-01 1.32166481e+00 2.63513345e-02
-8.30481827e-01 1.57285914e-01 -2.37873569e-01 -9.97450203e-04
1.54690936e-01 5.10045469e-01 -1.29485595e+00 6.31720722e-01
6.39961243e-01 2.47058257e-01 -3.67203802e-01 1.07869744e+00
4.72004488e-02 1.00148988e+00 -3.72309417e-01 -1.36505261e-01
2.36186966e-01 3.95184696e-01 6.63767517e-01 1.42649603e+00
3.45732749e-01 -5.18169582e-01 -6.78656101e-01 5.12740195e-01
-1.43214494e-01 1.27787665e-01 -6.51016653e-01 8.32926258e-02
1.06605329e-01 1.03720641e+00 -1.62719220e-01 -3.29214215e-01
-1.18911363e-01 9.21532869e-01 2.72280425e-02 2.66151369e-01
-8.69189918e-01 -4.60590303e-01 6.27277613e-01 -1.04650542e-01
4.37637329e-01 -1.18105978e-01 -1.92727327e-01 -1.05179524e+00
1.01539232e-01 -1.08279181e+00 5.97166121e-02 -3.73324811e-01
-1.02961719e+00 1.25053155e+00 -4.14339542e-01 -1.04833746e+00
-5.44040799e-01 -6.67228401e-01 -5.99642158e-01 1.25734961e+00
-1.54401648e+00 -8.88135910e-01 5.54320067e-02 7.19822466e-01
5.01735330e-01 -5.90764165e-01 1.25494051e+00 6.66787624e-01
-4.13133770e-01 1.15844107e+00 5.74786291e-02 9.98759493e-02
9.23583865e-01 -1.20704508e+00 7.30074763e-01 9.86082733e-01
4.56071913e-01 5.37395060e-01 1.06038272e+00 1.38989225e-01
-9.84063387e-01 -5.50061345e-01 1.25650930e+00 -3.62132311e-01
4.38096851e-01 -5.12804270e-01 -9.61108685e-01 4.03432369e-01
8.08113158e-01 -1.49770901e-01 6.38268292e-01 4.17279571e-01
-5.14897346e-01 -1.79590285e-01 -1.35639238e+00 2.12490916e-01
5.93373120e-01 -9.53138947e-01 -7.72359192e-01 1.13791369e-01
6.23739898e-01 -1.96099460e-01 -8.10363233e-01 4.82302248e-01
4.08819497e-01 -1.10554683e+00 8.49491417e-01 -4.04330522e-01
-5.85591123e-02 -1.21927746e-01 -2.90363580e-01 -1.51807249e+00
1.50037080e-01 -6.99760497e-01 1.47780716e-01 1.59626639e+00
8.93364668e-01 -7.61906385e-01 6.80763602e-01 2.52689332e-01
-1.95284739e-01 -6.36950284e-02 -1.61285317e+00 -8.89540732e-01
4.19067532e-01 -7.14364350e-01 6.99293137e-01 9.19776976e-01
8.20359215e-02 3.88300180e-01 -5.32366753e-01 2.52075791e-01
3.09487492e-01 -3.98662299e-01 6.98692977e-01 -1.13313615e+00
-3.61915752e-02 -6.52472079e-01 -8.02862048e-01 -5.18304944e-01
4.21738505e-01 -8.66361141e-01 3.25291663e-01 -1.23049605e+00
-1.39881864e-01 -7.74687603e-02 -7.52977908e-01 2.28708327e-01
-1.57443911e-01 4.18775588e-01 1.74717426e-01 -3.67108583e-01
-1.02242731e-01 5.49637079e-01 5.49591064e-01 -2.72087544e-01
-6.34518564e-02 4.03327346e-01 -5.58302283e-01 1.65429816e-01
9.04741824e-01 -3.30078721e-01 8.96140635e-02 -1.56605989e-01
-4.68757480e-01 1.50943831e-01 1.25349998e-01 -1.34222877e+00
2.55246520e-01 7.24067390e-01 3.05802077e-01 -7.12039649e-01
5.13304293e-01 -7.23663330e-01 3.59651893e-02 4.66083199e-01
-4.44421738e-01 -3.21170688e-01 5.09294808e-01 3.76637280e-02
-5.87030053e-01 -3.02549392e-01 7.85687566e-01 4.14434582e-01
-4.15241510e-01 -4.15482782e-02 -6.34323001e-01 -5.39271116e-01
3.62080812e-01 -1.86701462e-01 -1.99478254e-01 -6.04287565e-01
-8.81127179e-01 -3.70965064e-01 -5.99043183e-02 7.87623644e-01
3.48636448e-01 -1.06708407e+00 -1.25336182e+00 2.40614355e-01
-5.43842129e-02 -7.54700363e-01 3.46174985e-01 6.91426158e-01
-2.11068287e-01 8.14761937e-01 -1.75733477e-01 -6.67462647e-01
-1.54546940e+00 3.53326708e-01 8.57587814e-01 1.47476653e-02
-2.99145043e-01 9.52158809e-01 -4.25089657e-01 -8.41459751e-01
6.85981810e-01 -2.69829094e-01 -3.12620252e-01 1.67470872e-01
7.90941000e-01 2.73747355e-01 8.82471859e-01 -1.03012443e+00
-7.14456618e-01 1.41263083e-01 -6.97021857e-02 -6.14408195e-01
1.47473085e+00 -5.64177185e-02 3.28785241e-01 6.06562853e-01
1.38986731e+00 1.94334045e-01 -6.46317959e-01 -1.13622591e-01
1.46902338e-01 -6.31547719e-02 3.87866795e-01 -1.10825694e+00
-1.13646376e+00 1.22047913e+00 1.17613339e+00 4.72225666e-01
7.34891176e-01 -4.10168856e-01 6.15141094e-01 4.31752205e-01
2.21265227e-01 -8.02173018e-01 -5.82755923e-01 6.62312150e-01
1.19128716e+00 -1.63504815e+00 -1.89596489e-01 9.10432190e-02
-4.82654452e-01 1.16094267e+00 1.41628295e-01 -9.75183919e-02
8.36862206e-01 3.34112614e-01 6.06743813e-01 1.23916298e-01
-5.80136061e-01 -2.71013618e-01 4.01421309e-01 6.87169969e-01
1.06997490e+00 1.06293023e-01 -3.47968191e-01 3.13919127e-01
-6.32189929e-01 -1.41314805e-01 5.34605861e-01 3.94884825e-01
-1.90999091e-01 -1.13410115e+00 -5.64683199e-01 -1.66256934e-01
-9.03180063e-01 -2.87603424e-03 -5.84530652e-01 7.91457117e-01
-1.10544369e-01 1.44641626e+00 -2.64530838e-01 -7.67698407e-01
4.26983416e-01 6.05236411e-01 1.83673531e-01 -2.66496718e-01
-1.20777512e+00 -7.52611980e-02 2.72310197e-01 -3.54467273e-01
-6.36302829e-01 -7.02244639e-01 -8.42692494e-01 -4.53611314e-01
-7.14713335e-01 2.79920995e-01 1.43087685e+00 1.03370309e+00
3.18109430e-02 5.85243285e-01 9.64191675e-01 -8.45130980e-01
-6.25234723e-01 -1.70872056e+00 -4.90170509e-01 1.31768584e-01
1.03284585e+00 -4.83608216e-01 -8.40905309e-01 -9.23956931e-02] | [14.32711410522461, 6.077618598937988] |
7cb2ab81-88d2-4b48-9d22-96656af60f7e | monogrnet-a-geometric-reasoning-network-for | 1811.10247 | null | https://arxiv.org/abs/1811.10247v2 | https://arxiv.org/pdf/1811.10247v2.pdf | MonoGRNet: A Geometric Reasoning Network for Monocular 3D Object Localization | Detecting and localizing objects in the real 3D space, which plays a crucial role in scene understanding, is particularly challenging given only a single RGB image due to the geometric information loss during imagery projection. We propose MonoGRNet for the amodal 3D object detection from a monocular RGB image via geometric reasoning in both the observed 2D projection and the unobserved depth dimension. MonoGRNet is a single, unified network composed of four task-specific subnetworks, responsible for 2D object detection, instance depth estimation (IDE), 3D localization and local corner regression. Unlike the pixel-level depth estimation that needs per-pixel annotations, we propose a novel IDE method that directly predicts the depth of the targeting 3D bounding box's center using sparse supervision. The 3D localization is further achieved by estimating the position in the horizontal and vertical dimensions. Finally, MonoGRNet is jointly learned by optimizing the locations and poses of the 3D bounding boxes in the global context. We demonstrate that MonoGRNet achieves state-of-the-art performance on challenging datasets. | ['Yan Lu', 'Zengyi Qin', 'Jinglu Wang'] | 2018-11-26 | null | null | null | null | ['monocular-3d-object-localization'] | ['computer-vision'] | [ 1.66315690e-01 2.00539261e-01 -1.07207581e-01 -3.55261266e-01
-6.41961455e-01 -4.45685893e-01 4.53096032e-01 -2.35818848e-01
-3.83392483e-01 1.20224312e-01 7.92365372e-02 -2.90826082e-01
1.52187526e-01 -6.01701081e-01 -9.85823512e-01 -6.96851254e-01
2.56528050e-01 4.79102284e-01 3.77819180e-01 2.08396420e-01
3.22050333e-01 8.68403912e-01 -1.38569105e+00 1.24226883e-01
3.82562578e-01 1.53822339e+00 4.25626069e-01 6.06598556e-01
3.61111127e-02 6.27111018e-01 -7.36906230e-02 1.70601860e-01
5.31094491e-01 1.92344040e-02 -3.44911456e-01 5.07311881e-01
9.09054101e-01 -8.66109371e-01 -3.94540340e-01 1.02256393e+00
2.86086559e-01 -9.55690965e-02 5.61713815e-01 -1.19793200e+00
-1.92329213e-01 -5.44026010e-02 -9.88146484e-01 -1.25157103e-01
5.85914671e-01 2.79661804e-01 8.48324120e-01 -1.00323570e+00
6.49607956e-01 1.45096195e+00 5.33605874e-01 3.74245733e-01
-1.13467395e+00 -6.16333544e-01 7.94270933e-01 -1.39314458e-01
-1.37451303e+00 -1.50159836e-01 1.00404251e+00 -5.63044846e-01
9.95119274e-01 -1.54786363e-01 5.82534730e-01 9.04116273e-01
-1.48806959e-01 8.84210765e-01 1.03352332e+00 -1.65010571e-01
7.55030811e-02 -1.77817881e-01 1.11132354e-01 8.76512706e-01
2.69014716e-01 9.30467248e-02 -6.73788726e-01 2.08344072e-01
1.24631608e+00 3.76420945e-01 -3.17705214e-01 -8.68326843e-01
-1.12983930e+00 5.63603342e-01 7.87922978e-01 -3.33247393e-01
-2.94054210e-01 4.40902501e-01 -4.33164909e-02 -3.13438147e-01
4.34844196e-01 1.15031391e-01 -7.18364954e-01 2.28528500e-01
-5.55005491e-01 2.33846486e-01 4.08793718e-01 1.10722601e+00
9.93898392e-01 -1.85436144e-01 1.35598049e-01 2.26183668e-01
5.63505352e-01 7.31388927e-01 -2.77282279e-02 -1.05730546e+00
7.67277241e-01 1.12266684e+00 2.90107697e-01 -9.96015728e-01
-7.15219915e-01 -3.35529059e-01 -7.41679490e-01 3.28052342e-01
5.15343964e-01 8.13451409e-03 -1.16256642e+00 1.47279429e+00
6.90671563e-01 -8.41042679e-03 -2.40579739e-01 1.28403080e+00
9.85490143e-01 5.72133124e-01 -3.33251774e-01 3.44338119e-01
1.17006099e+00 -9.14559722e-01 -5.00438251e-02 -8.47705662e-01
4.21631604e-01 -5.65380037e-01 8.73534203e-01 2.79888332e-01
-9.41372633e-01 -3.40547979e-01 -8.01315308e-01 -6.63456500e-01
-2.57176459e-01 5.83173156e-01 7.12705970e-01 1.31866172e-01
-7.27343500e-01 2.70168986e-02 -8.11022997e-01 -2.21561834e-01
5.60565829e-01 2.51541257e-01 -5.43312550e-01 -3.02417010e-01
-5.36102653e-01 6.38129473e-01 2.63054132e-01 4.24800307e-01
-1.14944160e+00 -5.81648409e-01 -1.18856597e+00 -3.03667188e-02
5.82490325e-01 -8.64922643e-01 9.69162941e-01 -2.42889747e-01
-1.23147023e+00 1.21449327e+00 -2.27318749e-01 -2.50062197e-01
6.88715518e-01 -4.23017949e-01 4.54448909e-01 2.42065042e-01
2.24502057e-01 9.13176537e-01 8.95395815e-01 -1.35495806e+00
-7.31068969e-01 -9.89592671e-01 1.85488820e-01 6.48444772e-01
2.49975890e-01 -5.39524794e-01 -8.14319551e-01 -1.49010748e-01
1.09517789e+00 -7.42773652e-01 -3.51471901e-01 5.80859125e-01
-8.73480260e-01 -1.98634565e-02 7.93530166e-01 -4.81635362e-01
4.62288469e-01 -2.16079760e+00 2.97805130e-01 1.05090253e-02
2.82973200e-01 -3.23296160e-01 2.24285617e-01 -7.54771978e-02
1.39580682e-01 -2.11902902e-01 -9.70144719e-02 -7.18661606e-01
-4.72622514e-02 -8.78631547e-02 -4.74778086e-01 7.77986884e-01
1.72913760e-01 9.52737808e-01 -7.75266588e-01 -2.13094279e-01
5.68661988e-01 4.73864377e-01 -5.44848442e-01 3.78750771e-01
-4.25795972e-01 4.91326392e-01 -5.68609834e-01 9.83188868e-01
1.04958844e+00 -3.24108511e-01 -1.25473753e-01 -4.34014529e-01
-2.50983804e-01 3.57222587e-01 -1.35829318e+00 2.01109982e+00
-4.55176651e-01 4.49354291e-01 1.45165831e-01 -4.52996671e-01
8.66290212e-01 -2.60029525e-01 2.37763360e-01 -5.75578392e-01
7.66754672e-02 1.38414264e-01 -5.69754660e-01 -3.70188922e-01
3.19117218e-01 2.29390234e-01 -2.22981408e-01 4.18795526e-01
-7.93765858e-02 -5.99959075e-01 -2.70996064e-01 9.72248912e-02
8.03271770e-01 3.70574266e-01 2.58196414e-01 2.75499701e-01
4.03141767e-01 1.03877494e-02 5.10909140e-01 7.03225076e-01
-2.55709868e-02 8.44628990e-01 6.45841062e-01 -5.19384265e-01
-9.23300385e-01 -1.03559983e+00 6.39459409e-04 5.41251838e-01
6.15092754e-01 -4.69828621e-02 -2.52377212e-01 -6.29082918e-01
3.40009153e-01 3.28693718e-01 -7.67072201e-01 8.49525034e-02
-5.12021124e-01 -3.74784797e-01 -6.99277082e-03 5.71131229e-01
7.25496054e-01 -7.17185915e-01 -1.03863347e+00 -3.03343862e-01
-1.05197065e-01 -1.53640556e+00 -4.81369704e-01 6.50346935e-01
-9.90806162e-01 -1.30265057e+00 -4.93312031e-01 -5.97135425e-01
8.80473912e-01 7.88999736e-01 8.58582675e-01 -3.00990462e-01
-4.24950659e-01 2.27473751e-01 -4.83485758e-02 -2.42798373e-01
5.01853883e-01 1.35238424e-01 -1.90307036e-01 -3.20712314e-03
3.11881542e-01 -4.15503830e-01 -9.01805937e-01 4.22284573e-01
-3.92979205e-01 4.50984746e-01 5.92881858e-01 6.19673193e-01
8.19871187e-01 -1.79006502e-01 -3.53728890e-01 -6.09390557e-01
-2.94231206e-01 -1.70886904e-01 -9.83316958e-01 -1.66588530e-01
-1.13569222e-01 -1.83154687e-01 1.46424353e-01 -1.59412310e-01
-9.13167000e-01 7.13041961e-01 2.20890820e-01 -6.53134942e-01
-4.53107148e-01 -5.15764356e-02 -3.47132474e-01 -2.06724703e-01
4.33526039e-01 2.14700580e-01 -3.02380145e-01 -5.77555060e-01
3.61171842e-01 3.29895377e-01 3.89742225e-01 -1.60667360e-01
8.86441648e-01 1.03075683e+00 3.45379055e-01 -7.22073078e-01
-1.44492948e+00 -7.42188215e-01 -1.02338040e+00 -1.79044530e-01
1.09378421e+00 -1.48404765e+00 -1.02516985e+00 6.93460882e-01
-1.46831846e+00 -4.82671976e-01 -6.01792485e-02 4.70433086e-01
-5.67265928e-01 5.47509193e-02 -3.33251953e-01 -7.83134401e-01
-1.75666586e-01 -1.08437133e+00 1.90268576e+00 7.09282830e-02
2.82706678e-01 -6.20160878e-01 -2.39429444e-01 3.44186336e-01
-2.53421515e-01 3.30440760e-01 8.23474348e-01 2.38216430e-01
-1.40971673e+00 -2.50345170e-01 -7.07462788e-01 9.58294421e-02
-3.21745694e-01 -3.95549387e-01 -1.25333786e+00 5.88312186e-03
5.08134067e-02 -3.86945307e-01 1.12048316e+00 6.49493575e-01
1.33832085e+00 9.63521227e-02 -3.91553730e-01 1.33676982e+00
1.46966660e+00 -6.59698099e-02 2.17000484e-01 3.60957623e-01
1.19081056e+00 4.94430155e-01 7.75474072e-01 4.92391497e-01
5.09112000e-01 6.61889255e-01 1.11257124e+00 -1.96285412e-01
-1.10144332e-01 -6.01613402e-01 2.57350951e-01 -1.00167215e-01
1.77515209e-01 -3.23195420e-02 -8.96534920e-01 2.83844322e-01
-1.86505723e+00 -5.12155175e-01 -1.17627375e-01 2.09077334e+00
1.85121626e-01 2.33085692e-01 -1.65359899e-02 -9.08344015e-02
5.19647419e-01 3.19505244e-01 -1.11925244e+00 2.60691166e-01
-2.91259408e-01 -3.92140269e-01 9.62895393e-01 6.21023059e-01
-1.12925220e+00 1.05052292e+00 5.74666262e+00 2.76276827e-01
-1.06849539e+00 -8.26563910e-02 6.12609982e-01 -2.70160705e-01
-1.40210360e-01 6.45014271e-02 -1.27939022e+00 4.13334034e-02
-2.22742453e-01 6.18423998e-01 2.69293368e-01 1.02288103e+00
2.19461635e-01 -5.36296666e-01 -1.13519466e+00 1.38891280e+00
1.69530258e-01 -1.14092970e+00 5.28245885e-03 1.70489252e-01
8.37020457e-01 2.86779344e-01 9.21087191e-02 -1.63773745e-01
9.70015749e-02 -6.91762686e-01 1.04207623e+00 3.61186862e-01
8.43956351e-01 -5.80703378e-01 4.62095499e-01 6.34054899e-01
-1.20025945e+00 -4.33389097e-01 -5.43917656e-01 -1.83267996e-01
3.48004818e-01 7.21598506e-01 -9.22615349e-01 1.33184537e-01
8.25455010e-01 1.09417069e+00 -5.67556798e-01 9.43993807e-01
-8.13589633e-01 -1.20381363e-01 -5.45386732e-01 3.28016847e-01
4.10177708e-01 -3.29107314e-01 5.23341238e-01 7.05654860e-01
3.06304008e-01 1.15290359e-01 2.37270534e-01 1.07856357e+00
-5.80654405e-02 -4.21143830e-01 -6.71381831e-01 4.41219866e-01
3.21176946e-01 1.23816478e+00 -7.62330353e-01 -4.42392565e-03
-2.06594065e-01 1.11665905e+00 3.02684933e-01 4.39071178e-01
-6.96152270e-01 -1.55791461e-01 7.08080828e-01 2.90016741e-01
6.15022719e-01 -4.86551404e-01 -5.83154440e-01 -1.36335111e+00
3.57154399e-01 -2.45774239e-01 1.40107721e-01 -1.25744605e+00
-9.43908334e-01 3.38694096e-01 -7.79592693e-02 -1.19235170e+00
3.80425006e-02 -1.06949174e+00 -3.88083667e-01 9.12176013e-01
-1.72218192e+00 -1.25181770e+00 -1.01208818e+00 6.79598451e-01
3.97289932e-01 3.67926866e-01 3.13561469e-01 3.83022125e-03
-5.24675965e-01 1.00836009e-01 -4.35462207e-01 1.43416822e-01
5.20016313e-01 -1.13558292e+00 6.13731027e-01 7.78214335e-01
-1.37804583e-01 2.49968573e-01 2.61884928e-01 -6.03716671e-01
-1.68156445e+00 -1.16632426e+00 7.21995056e-01 -6.49132669e-01
3.94805521e-01 -9.80576396e-01 -3.64948720e-01 6.78656042e-01
-6.81280851e-01 2.10582241e-01 -1.21906392e-01 -4.77337539e-02
-5.63386738e-01 -1.83598503e-01 -9.89543796e-01 4.27478969e-01
1.54188466e+00 -7.67664790e-01 -1.92063600e-01 5.76601446e-01
8.65060508e-01 -1.00669539e+00 -2.10917264e-01 2.51734287e-01
4.71235991e-01 -1.08571804e+00 1.32167161e+00 -2.38552764e-01
5.56120932e-01 -6.10900164e-01 -4.14863557e-01 -6.73141479e-01
4.64653187e-02 -1.83419108e-01 -3.15108895e-01 5.18129110e-01
1.83333486e-01 -3.26674938e-01 1.24338841e+00 4.55635011e-01
-1.77720800e-01 -7.75164068e-01 -9.15833473e-01 -4.43453401e-01
-4.57152098e-01 -7.00893104e-01 3.76168132e-01 3.60482663e-01
-5.69048703e-01 4.69849825e-01 -2.91718721e-01 5.91788352e-01
9.07784343e-01 6.41844213e-01 1.21454871e+00 -1.15811110e+00
-2.12146938e-01 -2.84352988e-01 -7.21748114e-01 -1.96610558e+00
-1.08558405e-02 -6.70758843e-01 2.38263816e-01 -1.51075613e+00
2.36559093e-01 -2.68118322e-01 1.72045350e-01 2.69868255e-01
1.47141337e-01 3.68460268e-01 2.46518567e-01 1.14114530e-01
-6.50391757e-01 6.01605475e-01 1.41654742e+00 6.27055466e-02
-2.91425079e-01 -8.93555358e-02 -4.90509689e-01 9.65815663e-01
2.83962905e-01 -3.94902736e-01 -1.91814497e-01 -8.01017642e-01
3.72112542e-01 8.83512646e-02 8.62521529e-01 -9.01885509e-01
3.13738704e-01 -2.20810845e-01 8.15368950e-01 -1.42422330e+00
8.49985540e-01 -9.42814827e-01 -2.55506337e-01 3.16458970e-01
-2.14325890e-01 -1.30042523e-01 2.18335334e-02 7.87574649e-01
3.81094292e-02 7.60352165e-02 7.42248237e-01 -4.74087507e-01
-9.18944359e-01 7.47670949e-01 2.64513463e-01 -1.63432062e-01
1.08288562e+00 -3.71617436e-01 -2.74137825e-01 -2.11694077e-01
-7.13186622e-01 3.16186816e-01 7.16596663e-01 1.79734498e-01
1.04548812e+00 -9.95102465e-01 -2.52357453e-01 4.54177022e-01
2.35108614e-01 9.92245436e-01 4.37501162e-01 8.78137052e-01
-6.41127050e-01 5.82410693e-01 3.89545076e-02 -1.17772663e+00
-1.04612219e+00 4.86983269e-01 5.95110118e-01 4.42957915e-02
-7.31650233e-01 1.10988247e+00 9.45039213e-01 -5.65362215e-01
6.08371019e-01 -6.74618483e-01 2.29987890e-01 -1.55503362e-01
3.99034053e-01 2.37501651e-01 -3.81082259e-02 -4.90831822e-01
-4.74618047e-01 1.13812327e+00 -2.52386846e-04 -1.94393285e-02
1.42985606e+00 -4.24639374e-01 -1.22405760e-01 4.75686103e-01
1.19890976e+00 -5.28295338e-01 -2.02881646e+00 -2.85278589e-01
-5.26883423e-01 -8.95328045e-01 3.34002495e-01 -5.75897932e-01
-1.00393188e+00 1.23295188e+00 5.05136967e-01 -5.70663512e-01
9.77166951e-01 2.26740286e-01 3.56097639e-01 6.15284204e-01
4.89855528e-01 -8.59438479e-01 3.17384213e-01 7.55331814e-01
7.98512995e-01 -1.62434721e+00 2.70052999e-01 -5.75212419e-01
-3.62827539e-01 1.15403378e+00 9.00403202e-01 -1.91471040e-01
6.44718766e-01 -2.11024016e-01 -1.36348665e-01 -4.37366277e-01
-4.69937444e-01 -2.57513225e-01 3.57577205e-01 4.99746948e-01
-8.67330208e-02 -6.55607060e-02 6.20805383e-01 2.47370407e-01
6.75304681e-02 -4.08161134e-01 2.92047977e-01 8.00133407e-01
-4.31687325e-01 -4.09994304e-01 -4.93979871e-01 1.23731293e-01
1.39299095e-01 4.36018407e-02 -5.94651103e-01 8.78441095e-01
3.84352446e-01 4.97359455e-01 2.56721318e-01 -1.85142756e-01
2.89138019e-01 -3.06937993e-01 6.09652519e-01 -8.95080149e-01
-2.34263092e-02 1.53591573e-01 -2.71492183e-01 -9.44486201e-01
-2.32285321e-01 -7.25376189e-01 -1.03897631e+00 5.94270565e-02
-2.07136318e-01 -6.33196056e-01 1.00840271e+00 8.77247095e-01
2.11652145e-01 9.12620127e-02 6.34785354e-01 -1.38815176e+00
-2.69784272e-01 -6.61340058e-01 -6.38833225e-01 1.28294125e-01
6.97430134e-01 -7.87762523e-01 -5.04257917e-01 -2.55656630e-01] | [7.794857978820801, -2.632065773010254] |
01e94139-6cd2-4ae2-a26c-44e4d4ea0383 | spatial-attentive-single-image-deraining-with | 1904.01538 | null | https://arxiv.org/abs/1904.01538v2 | https://arxiv.org/pdf/1904.01538v2.pdf | Spatial Attentive Single-Image Deraining with a High Quality Real Rain Dataset | Removing rain streaks from a single image has been drawing considerable attention as rain streaks can severely degrade the image quality and affect the performance of existing outdoor vision tasks. While recent CNN-based derainers have reported promising performances, deraining remains an open problem for two reasons. First, existing synthesized rain datasets have only limited realism, in terms of modeling real rain characteristics such as rain shape, direction and intensity. Second, there are no public benchmarks for quantitative comparisons on real rain images, which makes the current evaluation less objective. The core challenge is that real world rain/clean image pairs cannot be captured at the same time. In this paper, we address the single image rain removal problem in two ways. First, we propose a semi-automatic method that incorporates temporal priors and human supervision to generate a high-quality clean image from each input sequence of real rain images. Using this method, we construct a large-scale dataset of $\sim$$29.5K$ rain/rain-free image pairs that covers a wide range of natural rain scenes. Second, to better cover the stochastic distribution of real rain streaks, we propose a novel SPatial Attentive Network (SPANet) to remove rain streaks in a local-to-global manner. Extensive experiments demonstrate that our network performs favorably against the state-of-the-art deraining methods. | ['Xin Yang', 'Tianyu Wang', 'Rynson Lau', 'Qiang Zhang', 'Ke Xu', 'Shaozhe Chen'] | 2019-04-02 | spatial-attentive-single-image-deraining-with-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Wang_Spatial_Attentive_Single-Image_Deraining_With_a_High_Quality_Real_Rain_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Wang_Spatial_Attentive_Single-Image_Deraining_With_a_High_Quality_Real_Rain_CVPR_2019_paper.pdf | cvpr-2019-6 | ['single-image-deraining'] | ['computer-vision'] | [ 1.55533701e-01 -5.94939768e-01 4.29775983e-01 -7.22586513e-01
-6.07142389e-01 -3.53788167e-01 1.48354068e-01 -5.61444700e-01
-3.45598400e-01 1.08044040e+00 -1.14854842e-01 -1.55145049e-01
2.35232621e-01 -8.57747853e-01 -6.95015430e-01 -1.07937717e+00
-2.02969611e-02 -1.31690189e-01 3.23399782e-01 -5.73311865e-01
-1.75790280e-01 5.62418342e-01 -1.61393535e+00 -8.03570971e-02
1.56131065e+00 5.66480875e-01 6.11752450e-01 6.27822340e-01
1.40760452e-01 8.39310229e-01 -7.77632296e-01 -7.02780336e-02
4.27008420e-01 -8.19691479e-01 -1.39964938e-01 -2.92879418e-02
1.02197683e+00 -5.69120646e-01 -4.33157951e-01 1.18885553e+00
5.52820086e-01 8.82347152e-02 2.24774063e-01 -6.41053021e-01
-5.55095553e-01 2.04364389e-01 -7.67310560e-01 5.18286169e-01
-9.49939713e-02 4.78010863e-01 6.57067955e-01 -9.04254317e-01
4.58486706e-01 1.08788276e+00 5.93034029e-01 3.27597290e-01
-9.40308928e-01 -8.48449349e-01 3.18071991e-01 9.50942859e-02
-1.04834282e+00 -3.00882250e-01 4.90715951e-01 -6.52579963e-02
3.81449491e-01 4.72211987e-01 8.92125785e-01 9.17074323e-01
2.00168684e-01 5.14873564e-01 1.66937816e+00 -1.57962218e-01
1.33968249e-01 -2.73684025e-01 1.70271814e-01 5.40860295e-01
6.58584356e-01 2.88051128e-01 -2.71860957e-01 3.82596731e-01
6.64747417e-01 2.04072058e-01 -8.66700292e-01 5.25237583e-02
-7.97813714e-01 7.69112825e-01 9.62789178e-01 1.06495798e-01
-3.65013003e-01 5.92291802e-02 -2.90452927e-01 3.80332798e-01
7.28920043e-01 2.94992507e-01 -2.54947186e-01 3.83496493e-01
-1.27516580e+00 7.13333189e-01 6.32531941e-01 4.18780059e-01
8.12059522e-01 5.59141219e-01 -1.19604923e-01 8.55612874e-01
2.09285572e-01 1.55251479e+00 -2.30457082e-01 -7.39506960e-01
5.39485812e-01 -1.04316026e-01 4.94761407e-01 -1.00816965e+00
-3.09349000e-01 -4.39888865e-01 -1.34194911e+00 7.21270442e-01
4.65396971e-01 -1.56935155e-01 -1.39444041e+00 1.44058716e+00
1.69774368e-01 3.75104874e-01 1.92975745e-01 1.57098496e+00
9.09710586e-01 9.14098203e-01 -1.57348663e-01 -5.36806583e-01
1.04199684e+00 -9.29813504e-01 -9.18628812e-01 -6.81270659e-01
-2.75012672e-01 -7.98013866e-01 9.48850453e-01 4.33112979e-01
-9.19115365e-01 -5.71103871e-01 -1.09086466e+00 2.71949559e-01
-1.41006097e-01 -9.41471159e-02 5.97942591e-01 3.72976333e-01
-8.48355591e-01 4.69942212e-01 -6.61767483e-01 -1.92041472e-01
4.12364721e-01 -2.40826994e-01 -5.76938502e-02 -8.38364184e-01
-1.41614795e+00 8.77472341e-01 -3.40149202e-03 9.81391549e-01
-1.18808866e+00 -6.65008068e-01 -8.05097580e-01 -1.80490941e-01
1.07923701e-01 -6.00525916e-01 8.50529611e-01 -1.17373550e+00
-1.24715436e+00 4.94620442e-01 -3.22967082e-01 -4.50907081e-01
4.35763001e-01 -5.56485116e-01 -5.72681069e-01 1.50044262e-01
-2.11765268e-03 3.48634511e-01 1.35508776e+00 -1.94550741e+00
-7.13580608e-01 -2.09767982e-01 1.26666337e-01 4.20833886e-01
4.15262580e-01 -2.20375896e-01 -5.27150095e-01 -8.88185859e-01
-4.04682197e-02 -9.48089480e-01 -4.67555523e-01 -2.71310247e-02
-2.09422082e-01 6.11544907e-01 8.52877200e-01 -7.97259212e-01
9.68011677e-01 -1.95261121e+00 -8.92558247e-02 -1.32401392e-01
4.19093668e-01 7.87707329e-01 -3.31344694e-01 2.51640994e-02
-2.47638281e-02 -1.11347467e-01 -8.00384104e-01 -3.61982018e-01
-5.89756250e-01 5.26556551e-01 -7.20496178e-01 6.73516154e-01
2.79551089e-01 6.97237909e-01 -9.51610446e-01 -2.44970173e-01
3.26361567e-01 6.35852516e-01 -1.31200150e-01 7.06843555e-01
-3.53784025e-01 5.45815468e-01 -2.37084806e-01 7.61500537e-01
1.45409083e+00 2.43625030e-01 -1.24530219e-01 -1.29231796e-01
-1.56662807e-01 -2.04282716e-01 -1.10597622e+00 1.11517489e+00
-5.76627910e-01 7.37643123e-01 3.62410277e-01 -4.94895548e-01
9.59902525e-01 1.20010860e-02 -1.22766577e-01 -9.52242613e-01
-1.12408489e-01 3.15419108e-01 9.87265073e-03 -7.37857103e-01
4.23470467e-01 -3.81979734e-01 4.33585346e-01 1.04385093e-01
-4.02692646e-01 -5.54028273e-01 5.02051301e-02 1.01911597e-01
9.16086495e-01 -1.10200159e-02 -1.11306928e-01 8.07126798e-03
8.67163017e-02 -9.20183435e-02 9.13361847e-01 1.06076801e+00
-2.79869586e-01 1.46542192e+00 -1.18046798e-01 -4.50125605e-01
-8.33186567e-01 -1.38184321e+00 -9.36736912e-02 6.44946873e-01
5.70544899e-01 3.41456890e-01 -5.23547947e-01 -5.51054299e-01
-1.21094175e-01 4.30527031e-01 -7.57527173e-01 2.42504865e-01
-8.32787931e-01 -1.38999379e+00 3.29133362e-01 1.91525981e-01
9.49062884e-01 -1.38158131e+00 -5.45182109e-01 6.02456974e-03
-4.50968474e-01 -1.26271367e+00 -1.98753819e-01 1.63626090e-01
-8.34246218e-01 -1.03905904e+00 -8.29861939e-01 -6.06083035e-01
4.82531548e-01 1.00077140e+00 1.57272184e+00 2.23996297e-01
-5.38172483e-01 -3.10217500e-01 -7.32163012e-01 -5.89895427e-01
1.41109630e-01 -4.51466739e-01 -2.23177552e-01 -4.63735498e-02
-1.33739889e-01 -6.82769775e-01 -1.02637553e+00 1.96864620e-01
-1.22486925e+00 -6.68302327e-02 7.01410115e-01 8.88248146e-01
5.29947519e-01 2.87115276e-02 4.13327962e-01 -9.62503552e-01
4.06923324e-01 -5.05471885e-01 -7.07551658e-01 -2.24260297e-02
-3.61413032e-01 -3.04330260e-01 6.02653205e-01 -8.83213803e-02
-1.42545509e+00 -4.46426757e-02 2.09546252e-03 -4.39915597e-01
-1.64034307e-01 4.34016705e-01 -1.75722197e-01 -1.15529738e-01
6.73861384e-01 3.70126128e-01 -9.89721566e-02 -3.76048863e-01
4.02646631e-01 2.93971509e-01 6.82885587e-01 -7.92528614e-02
1.42615402e+00 9.28881764e-01 -2.29271129e-01 -1.02750361e+00
-1.30270696e+00 -4.05294836e-01 -2.17593163e-01 -2.25018755e-01
7.68042564e-01 -1.33480883e+00 -2.87884623e-01 9.32867706e-01
-1.02631938e+00 -7.22547233e-01 -1.04872473e-01 3.66171122e-01
-2.77342759e-02 3.96115214e-01 -5.06922543e-01 -9.38309491e-01
-6.92059398e-01 -9.39151049e-01 8.58387470e-01 7.15167165e-01
4.89902645e-01 -5.92312932e-01 2.17300713e-01 3.78822386e-01
7.74375021e-01 3.07223529e-01 2.86483973e-01 5.46062469e-01
-9.99429405e-01 1.41478777e-01 -5.88274121e-01 6.08094692e-01
3.01981151e-01 5.66438138e-02 -9.69231904e-01 -4.78538007e-01
8.36802572e-02 -3.08833927e-01 1.42228425e+00 6.31922245e-01
8.31456363e-01 -9.40851048e-02 7.58172944e-02 9.60487664e-01
1.72709596e+00 -3.53542087e-03 1.10346162e+00 3.17416549e-01
8.43782485e-01 3.60600740e-01 1.02347565e+00 2.89399505e-01
1.67514771e-01 4.57660377e-01 9.26697850e-01 -6.81376696e-01
-5.17809570e-01 3.17317069e-01 3.37978959e-01 5.12493670e-01
-4.20054376e-01 -6.46272421e-01 -4.86261219e-01 8.66428673e-01
-1.69738126e+00 -1.18539357e+00 -2.33753756e-01 2.10111737e+00
8.56945872e-01 -1.26756385e-01 -3.97238463e-01 -2.44934410e-01
3.97521645e-01 7.87910819e-01 -4.94300663e-01 1.46318391e-01
-6.37667000e-01 6.24390304e-01 6.82220042e-01 6.88474238e-01
-1.10230947e+00 1.11514211e+00 5.40425539e+00 5.42652607e-01
-1.37441254e+00 -4.28211577e-02 4.75800991e-01 -1.07217334e-01
-1.69923827e-01 -3.18444222e-02 -7.00682104e-01 5.72969258e-01
4.44389284e-01 4.59105939e-01 4.82931018e-01 3.79712492e-01
7.79487073e-01 -5.15488863e-01 -2.70452768e-01 1.00077415e+00
2.43565291e-01 -9.97424006e-01 -1.32681787e-01 -3.77904326e-01
9.82263029e-01 4.01712418e-01 5.52031845e-02 1.65832460e-01
6.17470741e-01 -1.30308914e+00 4.68145400e-01 9.19207513e-01
7.45864630e-01 -5.68807781e-01 9.38710034e-01 -6.73307292e-03
-9.73523498e-01 2.70763546e-01 -4.09518987e-01 -3.40951420e-02
2.75123328e-01 1.36636353e+00 -3.85291725e-01 8.13612461e-01
1.27951837e+00 7.77200580e-01 -5.30921221e-01 1.27593374e+00
-7.45246589e-01 9.37171578e-01 -3.88590753e-01 5.51445186e-01
1.82207242e-01 -5.67623973e-01 5.76825023e-01 1.31364954e+00
1.91149622e-01 4.83217955e-01 -2.57759094e-02 6.23377860e-01
-4.50970270e-02 -3.73651922e-01 -6.54093027e-01 4.48829383e-01
2.61773497e-01 1.27835500e+00 -4.17545080e-01 -1.27205193e-01
-2.26683334e-01 1.15932155e+00 2.90830433e-02 8.08387756e-01
-9.62815106e-01 -5.77491939e-01 1.18430245e+00 -8.45228322e-03
5.34082532e-01 -2.98809797e-01 -2.75556028e-01 -1.36523128e+00
2.57328540e-01 -1.13304305e+00 -1.65715471e-01 -9.37402308e-01
-1.31760430e+00 8.99106920e-01 -2.96650469e-01 -1.24624014e+00
2.61675656e-01 -1.78525254e-01 -9.03460145e-01 1.05622995e+00
-2.26905966e+00 -1.04138994e+00 -1.35872555e+00 5.07787585e-01
6.33686543e-01 3.14065009e-01 3.40282738e-01 3.80656898e-01
-6.58005118e-01 1.55141994e-01 3.57774377e-01 3.13674621e-02
1.06620574e+00 -1.21119297e+00 4.12800759e-01 1.52740026e+00
-5.60104325e-02 2.15393394e-01 1.20197952e+00 -5.50673366e-01
-1.25239706e+00 -1.58215821e+00 5.33087075e-01 -1.72975257e-01
2.27411911e-01 -2.60790288e-01 -1.36802351e+00 2.47967780e-01
1.99751988e-01 6.15531325e-01 -1.90515980e-01 -1.83069199e-01
-4.38701242e-01 -6.57152116e-01 -9.52961743e-01 6.43981516e-01
8.99971724e-01 -1.19025402e-01 -3.93766671e-01 3.33789885e-01
6.53003693e-01 -5.01172602e-01 -3.20878565e-01 7.67990232e-01
3.29574615e-01 -1.52054226e+00 9.89409864e-01 6.43296987e-02
6.60252213e-01 -7.46046662e-01 -2.03477323e-01 -1.70126295e+00
-1.04400247e-01 -2.81020373e-01 -1.51621357e-01 1.11166728e+00
9.03255120e-02 -6.05702996e-01 5.17913282e-01 -1.04475051e-01
-2.75762994e-02 -4.73090529e-01 -5.86548746e-01 -6.75408959e-01
-1.38777032e-01 -8.95738527e-02 2.73862988e-01 8.44933569e-01
-8.97607625e-01 2.19042599e-01 -8.79924297e-01 6.51260793e-01
9.85587358e-01 7.36823559e-01 8.57801914e-01 -9.31757629e-01
-1.14949942e-01 -7.16919228e-02 4.96454956e-03 -1.00447154e+00
-1.19508758e-01 -1.82168350e-01 8.91302049e-01 -1.66721594e+00
1.09504469e-01 -6.93556070e-01 -1.56658143e-01 2.57049143e-01
-7.29683697e-01 5.82267463e-01 3.13586056e-01 4.25722927e-01
-4.63529408e-01 9.51052010e-01 1.48879826e+00 -1.22394368e-01
-1.49987340e-01 -2.74170022e-02 -3.63266110e-01 7.10267544e-01
9.18823242e-01 -5.69571435e-01 -1.91608176e-01 -8.39237571e-01
-3.81942242e-02 8.56412575e-02 4.42759275e-01 -1.21615231e+00
-2.56715566e-01 -5.79001963e-01 3.49324405e-01 -4.84783679e-01
4.06043977e-01 -5.36594450e-01 -7.41678551e-02 2.50623703e-01
2.78048575e-01 -2.76379049e-01 2.28532404e-02 7.90539205e-01
-6.10443950e-01 1.82208300e-01 1.38102531e+00 -6.55012354e-02
-6.61348104e-01 4.74309683e-01 -2.43751198e-01 1.39593765e-01
6.38684630e-01 6.60630465e-02 -6.64457142e-01 -5.55834115e-01
-3.06853086e-01 3.18224788e-01 4.76635218e-01 3.07543755e-01
8.00837338e-01 -5.41002870e-01 -1.11231697e+00 7.14520290e-02
4.50223982e-02 4.34382886e-01 6.11994326e-01 7.09649861e-01
-8.84552538e-01 -1.70317829e-01 -1.05957799e-01 -3.10815960e-01
-1.34207976e+00 1.97029203e-01 5.41530252e-01 -2.85067379e-01
-1.00324368e+00 8.58833373e-01 5.28001726e-01 -2.03561693e-01
4.60902415e-02 -4.27977264e-01 -1.44145325e-01 -1.09683819e-01
7.36808598e-01 -8.27306882e-02 1.87708274e-01 -3.12342942e-01
-3.03608701e-02 4.65242743e-01 -1.22710183e-01 2.03610703e-01
1.56846583e+00 -3.16523314e-01 -8.17585588e-02 3.11605483e-01
6.60842419e-01 4.61264886e-02 -1.53323162e+00 -3.23959678e-01
-7.02533662e-01 -8.99147034e-01 1.18706174e-01 -9.38275456e-01
-1.81330228e+00 8.05449903e-01 9.28911030e-01 -1.84391942e-02
1.49416065e+00 -4.04203475e-01 1.01929116e+00 3.61840099e-01
1.26720369e-01 -6.71171904e-01 3.62758003e-02 8.65215302e-01
9.27386403e-01 -1.63107049e+00 1.41635433e-01 -5.38381696e-01
-6.90938473e-01 8.80003452e-01 6.76351786e-01 -3.69737864e-01
5.04467010e-01 3.49516809e-01 7.78807402e-01 -2.49373287e-01
-3.46015006e-01 -5.94591022e-01 -2.57840574e-01 6.64944947e-01
1.44894525e-01 5.10907695e-02 -2.27344662e-01 1.43674314e-01
8.76884013e-02 1.59397572e-01 8.45549524e-01 8.81551921e-01
-6.86570644e-01 -6.12392843e-01 -7.70570457e-01 3.43958795e-01
-2.28565767e-01 -4.89784569e-01 1.74830839e-01 5.90503156e-01
2.36755699e-01 1.13224828e+00 -1.45153105e-01 -1.69689152e-02
3.38724762e-01 -6.54816806e-01 4.00114864e-01 -3.87920648e-01
-3.24690521e-01 -2.06351653e-01 -1.41820073e-01 -5.12431383e-01
-6.34433448e-01 -4.37308043e-01 -7.90160120e-01 -2.67111421e-01
-1.09336346e-01 5.06725572e-02 4.35818940e-01 9.26336944e-01
1.07267871e-01 5.62541068e-01 8.13458502e-01 -1.32672465e+00
-6.49063438e-02 -9.79034424e-01 -1.02592611e+00 3.38183165e-01
8.85892630e-01 -5.76367795e-01 -6.26280069e-01 1.28288761e-01] | [10.921175003051758, -3.251901865005493] |
b43c23f8-8d66-43f6-af0b-22b35d2a589b | cross-lingual-question-answering-over | 2302.13241 | null | https://arxiv.org/abs/2302.13241v1 | https://arxiv.org/pdf/2302.13241v1.pdf | Cross-Lingual Question Answering over Knowledge Base as Reading Comprehension | Although many large-scale knowledge bases (KBs) claim to contain multilingual information, their support for many non-English languages is often incomplete. This incompleteness gives birth to the task of cross-lingual question answering over knowledge base (xKBQA), which aims to answer questions in languages different from that of the provided KB. One of the major challenges facing xKBQA is the high cost of data annotation, leading to limited resources available for further exploration. Another challenge is mapping KB schemas and natural language expressions in the questions under cross-lingual settings. In this paper, we propose a novel approach for xKBQA in a reading comprehension paradigm. We convert KB subgraphs into passages to narrow the gap between KB schemas and questions, which enables our model to benefit from recent advances in multilingual pre-trained language models (MPLMs) and cross-lingual machine reading comprehension (xMRC). Specifically, we use MPLMs, with considerable knowledge of cross-lingual mappings, for cross-lingual reading comprehension. Existing high-quality xMRC datasets can be further utilized to finetune our model, greatly alleviating the data scarcity issue in xKBQA. Extensive experiments on two xKBQA datasets in 12 languages show that our approach outperforms various baselines and achieves strong few-shot and zero-shot performance. Our dataset and code are released for further research. | ['Dongyan Zhao', 'Haowei Du', 'Xingyu Shen', 'Yansong Feng', 'Yuxuan Lai', 'Chen Zhang'] | 2023-02-26 | null | null | null | null | ['cross-lingual-question-answering', 'reading-comprehension', 'machine-reading-comprehension'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [-1.60817906e-01 9.06257108e-02 -3.79912019e-01 -3.36498111e-01
-1.50524473e+00 -7.83543646e-01 3.62020224e-01 3.28056812e-01
-4.47180271e-01 1.04688919e+00 5.46832979e-01 -6.96010530e-01
-1.10299341e-01 -9.39254224e-01 -9.62657034e-01 -9.35077369e-02
6.16024673e-01 6.33807302e-01 3.78135473e-01 -9.92340326e-01
-9.89096090e-02 -3.01597178e-01 -1.14621425e+00 5.35235107e-01
1.49756682e+00 7.76347876e-01 3.67000967e-01 4.87681270e-01
-6.79501235e-01 1.03411603e+00 -3.04117352e-01 -9.32904661e-01
-2.05674723e-01 -3.06038767e-01 -1.53946853e+00 -5.22835374e-01
6.74397111e-01 -2.64924973e-01 -2.37895638e-01 8.96098793e-01
3.79120737e-01 8.37889686e-02 4.52003747e-01 -1.02966368e+00
-1.18859303e+00 8.30878615e-01 -2.90136844e-01 3.39201063e-01
6.88677251e-01 -2.15608001e-01 1.30708957e+00 -6.94649518e-01
5.43633044e-01 1.28749919e+00 5.37626386e-01 6.74148858e-01
-9.69510674e-01 -5.09273708e-01 9.70159024e-02 7.22175539e-01
-1.41690421e+00 -5.46455085e-01 4.91551459e-01 -1.57532081e-01
1.22783852e+00 2.95150608e-01 2.18427047e-01 1.27333736e+00
-1.95910096e-01 9.83297646e-01 1.14162028e+00 -7.03865707e-01
-6.91530034e-02 2.76485980e-01 3.93381149e-01 6.67366326e-01
5.98920956e-02 -6.41717732e-01 -6.56363606e-01 2.07044147e-02
3.96985799e-01 -6.42267764e-01 -6.44972682e-01 -1.85554385e-01
-1.23415899e+00 8.23081434e-01 1.91665962e-01 1.40645713e-01
1.67514241e-04 -2.52129614e-01 4.81859654e-01 6.00235820e-01
3.01671952e-01 6.98408961e-01 -9.59576249e-01 -2.04299018e-01
-4.00183141e-01 2.18829080e-01 1.00375509e+00 1.37468898e+00
9.22522962e-01 -5.23302555e-01 -1.02987036e-01 1.27782464e+00
6.28101081e-02 7.61493504e-01 5.82528293e-01 -7.69772828e-01
1.04254627e+00 7.40225852e-01 7.94102848e-02 -8.52810204e-01
-1.08867630e-01 -1.12237111e-01 -6.21495247e-01 -9.07297075e-01
5.27540207e-01 -7.37868808e-03 -3.91847849e-01 1.84118390e+00
3.76526892e-01 -7.44502023e-02 6.13135517e-01 7.08318233e-01
1.03893375e+00 9.11730349e-01 8.57333615e-02 6.10953122e-02
1.76577914e+00 -1.13643324e+00 -7.60652006e-01 -3.84990573e-01
8.95594656e-01 -5.42753577e-01 1.71259379e+00 1.72859490e-01
-8.86331260e-01 -4.42007899e-01 -7.38630056e-01 -7.82313824e-01
-5.86793303e-01 5.20127714e-02 4.29927677e-01 5.70633590e-01
-9.79648411e-01 -4.25377548e-01 -4.73031044e-01 -5.55019379e-01
5.77752367e-02 -2.24832952e-01 -2.62831092e-01 -6.75499678e-01
-1.82529628e+00 1.16275966e+00 5.52317739e-01 3.75276133e-02
-8.14064324e-01 -9.06647503e-01 -1.04967904e+00 -9.68272015e-02
9.26982582e-01 -5.93841493e-01 1.44874525e+00 -5.35172582e-01
-1.30293775e+00 7.73881257e-01 -3.40158522e-01 -3.09305757e-01
6.52098507e-02 -5.32044947e-01 -5.58988333e-01 2.50014830e-02
5.22405803e-01 6.28310025e-01 4.24768358e-01 -9.94306326e-01
-7.80912936e-01 -3.50242466e-01 5.96426129e-01 4.86977160e-01
-3.62655789e-01 -1.01154052e-01 -8.67656291e-01 -3.36784989e-01
-2.00685561e-01 -5.33669055e-01 8.27209428e-02 -7.15985119e-01
-3.40292335e-01 -6.56756103e-01 4.21434909e-01 -1.03813291e+00
1.33359993e+00 -1.72302723e+00 2.88938135e-01 -3.45977843e-01
-2.36040622e-01 1.91229701e-01 -3.90903324e-01 6.58937931e-01
3.99319082e-01 5.65321818e-02 -9.98134762e-02 -1.00226335e-01
7.53646791e-02 6.09676421e-01 -9.06892478e-01 -2.01036543e-01
3.01880240e-01 1.37023032e+00 -1.17903650e+00 -6.19725525e-01
-2.43663326e-01 4.55952771e-02 -3.47618967e-01 3.49183440e-01
-6.08998179e-01 3.19200099e-01 -5.47423065e-01 8.56224000e-01
4.38432276e-01 -2.35549927e-01 5.60842231e-02 -2.74457484e-01
1.26358837e-01 7.50894368e-01 -6.61059737e-01 2.17445755e+00
-8.25844467e-01 3.15031409e-01 -1.94344640e-01 -1.02385879e+00
8.36582899e-01 3.76853049e-01 -9.25050005e-02 -1.14255667e+00
-2.48846725e-01 3.92594039e-01 -4.09958005e-01 -8.57523203e-01
8.07070792e-01 -2.52391338e-01 -4.96338218e-01 4.29942757e-01
2.19926417e-01 -3.40812117e-01 3.40416282e-01 5.56542814e-01
8.34955990e-01 1.94626555e-01 4.22573000e-01 -3.02713960e-01
7.64891446e-01 4.26882952e-01 2.26414397e-01 7.53584445e-01
-3.21718585e-03 -3.46697844e-03 3.77226651e-01 -2.32948717e-02
-7.46732891e-01 -1.01440740e+00 -1.38560370e-01 1.44172359e+00
2.61734039e-01 -5.01773775e-01 -7.13481426e-01 -7.90024996e-01
-7.24661350e-02 1.05644262e+00 -3.53556633e-01 -1.35780707e-01
-6.93996251e-01 -4.51409876e-01 9.36663747e-01 4.41996157e-01
4.72715795e-01 -9.53297138e-01 -3.15832704e-01 3.45021963e-01
-9.54778969e-01 -1.48320210e+00 -3.04080963e-01 -9.23725069e-02
-4.43517298e-01 -1.07330871e+00 -5.22401273e-01 -9.51589465e-01
3.74063909e-01 2.83394128e-01 1.58796585e+00 -1.97158143e-01
-1.29734026e-02 8.02020013e-01 -7.81067550e-01 -4.50297982e-01
-4.86293316e-01 5.22254586e-01 -1.22853614e-01 -3.52902353e-01
7.47169733e-01 -1.36746228e-01 -2.95860291e-01 1.94083139e-01
-8.69312108e-01 1.49689212e-01 4.24288303e-01 7.22383320e-01
6.25463068e-01 -1.53093338e-01 1.16274881e+00 -9.73002911e-01
1.03417385e+00 -7.45400965e-01 -5.78593969e-01 1.06312931e+00
-3.11830938e-01 1.57816246e-01 6.86323345e-01 -1.71945274e-01
-1.43588090e+00 -6.34491265e-01 -1.92474440e-01 1.10843122e-01
-2.37798821e-02 1.07377994e+00 -3.49707156e-01 2.92407095e-01
5.49688339e-01 4.04288322e-01 -5.19162893e-01 -6.06834471e-01
8.89581442e-01 8.04242730e-01 7.07315981e-01 -1.11000025e+00
3.25896144e-01 1.15732864e-01 -7.65828609e-01 -8.01268220e-01
-1.49776101e+00 -7.25822747e-01 -6.31240249e-01 1.44991502e-02
8.40497553e-01 -1.20381665e+00 -4.17451560e-01 3.00971836e-01
-1.18130982e+00 -3.46670866e-01 -1.66446343e-01 8.97490978e-02
-5.04123628e-01 3.65178555e-01 -5.63616276e-01 -5.40881932e-01
-3.91468287e-01 -9.57664490e-01 1.06287992e+00 1.37726933e-01
-1.31965667e-01 -1.24194574e+00 3.51196498e-01 1.13725221e+00
2.91188508e-01 -4.97028947e-01 1.56708300e+00 -5.69123983e-01
-6.27329409e-01 2.27084041e-01 -2.81061381e-01 4.67546701e-01
2.21283838e-01 -6.62988603e-01 -6.62615895e-01 -2.40154639e-01
-4.56866547e-02 -1.30799794e+00 6.24618411e-01 -9.30722728e-02
1.00059378e+00 -2.61202037e-01 2.89755333e-02 2.83984214e-01
1.40479112e+00 -2.38671854e-01 4.93881494e-01 4.91483182e-01
8.03898752e-01 8.67520452e-01 8.14636350e-01 -7.59186521e-02
1.44033515e+00 6.37813270e-01 2.58727651e-02 1.48697838e-01
-1.40729859e-01 -5.71773112e-01 4.10799056e-01 1.68960595e+00
3.52772027e-01 -2.84604281e-01 -1.30273032e+00 1.19852471e+00
-1.81031990e+00 -4.15404826e-01 -1.28263131e-01 1.75004578e+00
1.57812178e+00 -4.52375591e-01 -4.70915556e-01 -5.38771808e-01
1.95096508e-01 -4.15879041e-02 -6.25972450e-01 -3.97106886e-01
-3.83801788e-01 3.00242871e-01 1.90914631e-01 5.37041247e-01
-8.29702973e-01 1.37263787e+00 5.47914791e+00 1.09360647e+00
-6.88898444e-01 2.93474555e-01 3.18078041e-01 2.74052083e-01
-5.23993492e-01 9.33835506e-02 -1.19091117e+00 2.19955102e-01
8.72049034e-01 -3.85389924e-01 5.25516868e-01 5.74428797e-01
-3.41353029e-01 -2.07037956e-01 -1.05648005e+00 8.38373184e-01
4.08857614e-01 -1.15440774e+00 4.36779588e-01 -5.39580584e-01
7.66768992e-01 2.63256341e-01 -2.08077654e-01 1.00790191e+00
5.92172623e-01 -1.06704700e+00 4.50906187e-01 4.22176272e-01
8.92688215e-01 -7.63773322e-01 7.19576359e-01 7.36854434e-01
-1.08259237e+00 9.54111665e-02 -7.91927159e-01 1.65940896e-01
3.15229326e-01 5.25497980e-02 -8.37493002e-01 9.39902306e-01
8.59575927e-01 4.66982573e-01 -6.58788502e-01 2.97817975e-01
-5.50397575e-01 7.19281793e-01 -2.00016834e-02 -1.38147563e-01
3.51858556e-01 -2.49865979e-01 1.62625358e-01 1.14622986e+00
2.46311724e-01 3.30027729e-01 8.16439390e-02 7.14372754e-01
-3.74965727e-01 5.80039203e-01 -4.05050784e-01 -9.67800319e-02
5.66123307e-01 1.03880036e+00 1.11735994e-02 -4.74842548e-01
-9.26646292e-01 9.47248459e-01 1.08744240e+00 4.05764252e-01
-5.49178362e-01 -2.38438353e-01 3.92255425e-01 -1.81323320e-01
1.78202480e-01 -1.19661532e-01 1.03650220e-01 -1.69218612e+00
1.23442151e-01 -1.49877155e+00 7.85332024e-01 -9.22404766e-01
-1.65169907e+00 4.40540045e-01 1.22353084e-01 -5.40520668e-01
-2.45545432e-01 -5.07787645e-01 4.75479551e-02 1.03965998e+00
-2.05246210e+00 -1.52011120e+00 -2.68147111e-01 8.63564074e-01
7.79127836e-01 -4.25218679e-02 1.02978504e+00 2.69341737e-01
-3.87199014e-01 6.13438547e-01 1.35818183e-01 1.69752255e-01
1.00748169e+00 -1.28521132e+00 3.25714529e-01 7.09451675e-01
3.87450814e-01 8.08295667e-01 2.56259918e-01 -6.35614872e-01
-1.73080230e+00 -1.18333650e+00 1.34212101e+00 -9.47577059e-01
1.10103440e+00 -4.75881010e-01 -1.42771196e+00 9.32632565e-01
4.81769949e-01 -4.23743725e-01 9.74211574e-01 5.66069901e-01
-5.96323013e-01 -1.48333311e-01 -5.65529823e-01 6.99898064e-01
7.95623660e-01 -1.06465209e+00 -1.22704220e+00 2.10344434e-01
1.09828234e+00 -4.49145585e-01 -1.08470368e+00 4.39513654e-01
2.29586318e-01 -5.11407912e-01 8.39789033e-01 -9.37977552e-01
3.25581372e-01 -2.21885070e-01 -3.73667330e-01 -1.50346398e+00
2.33160689e-01 -1.71006754e-01 -1.91002488e-02 1.33795333e+00
6.52845919e-01 -4.47816193e-01 2.03946963e-01 5.64993203e-01
-1.31847247e-01 -7.80287743e-01 -1.17880857e+00 -6.17389381e-01
5.39360166e-01 -6.50741696e-01 7.62500823e-01 1.17795897e+00
3.26254994e-01 8.13850105e-01 -3.12698215e-01 2.11865619e-01
2.16149688e-01 3.54721487e-01 8.50237370e-01 -6.41629815e-01
-2.67054498e-01 2.77993251e-02 2.00569093e-01 -1.54603803e+00
4.84824747e-01 -1.05755627e+00 9.42284614e-02 -1.58714020e+00
3.74354690e-01 -6.33203685e-01 -1.34087712e-01 6.46498263e-01
-5.44752479e-01 -5.59141103e-04 -2.90313601e-01 8.20050165e-02
-9.56841469e-01 7.74712741e-01 1.34185410e+00 -1.88549712e-01
5.82568608e-02 -5.53311288e-01 -9.16999519e-01 4.97111410e-01
6.03361666e-01 -2.11805239e-01 -8.34632218e-01 -1.11827898e+00
6.46516502e-01 2.11970463e-01 1.33088529e-01 -3.40235084e-01
4.82544899e-01 -6.30567148e-02 -3.02333802e-01 -5.45710504e-01
1.38733506e-01 -4.44643736e-01 -4.56779093e-01 -1.44027784e-01
-4.38015789e-01 9.87502113e-02 3.73814523e-01 4.56761867e-01
-6.14663899e-01 -3.83676887e-01 2.47675404e-01 -2.09980741e-01
-1.08394337e+00 1.11249335e-01 -1.36420652e-01 8.89872789e-01
6.23674929e-01 2.56366849e-01 -7.47203946e-01 -5.31692564e-01
-2.74821430e-01 7.45558858e-01 3.17880452e-01 8.04300666e-01
4.77274090e-01 -1.20537162e+00 -8.66276145e-01 2.68527819e-03
8.47110212e-01 4.03356493e-01 4.24082160e-01 7.33469129e-01
-2.86438942e-01 1.07825160e+00 -1.02564581e-02 -3.44561964e-01
-9.33104277e-01 4.27646786e-01 1.69825301e-01 -4.32851136e-01
-2.39564821e-01 9.40112352e-01 1.81373075e-01 -1.00480163e+00
7.84889385e-02 -4.09651160e-01 -3.60385895e-01 9.59889591e-02
5.27171373e-01 2.22591370e-01 2.29684263e-01 -6.55364871e-01
-2.13288784e-01 4.31239069e-01 -3.62391174e-01 6.56596869e-02
9.88028944e-01 -5.56429088e-01 -4.82587039e-01 5.95726430e-01
9.65630114e-01 1.69410139e-01 -6.40921891e-01 -6.97584391e-01
3.34883362e-01 -2.01917551e-02 -1.57527745e-01 -1.13167536e+00
-3.61874968e-01 9.38775957e-01 -1.12420738e-01 -6.19822294e-02
1.13936126e+00 5.34750700e-01 1.16462839e+00 8.65406930e-01
7.08071768e-01 -1.12852931e+00 -4.54686058e-04 9.64795589e-01
9.86092448e-01 -1.40295649e+00 -4.18119192e-01 -2.93651044e-01
-7.64873028e-01 7.47795105e-01 1.00007546e+00 5.83063483e-01
1.95501134e-01 -1.77838311e-01 5.41645765e-01 -3.03012103e-01
-9.23416674e-01 -2.82805800e-01 5.11930108e-01 4.86471266e-01
3.39641154e-01 1.14050895e-01 -1.39633954e-01 9.73899841e-01
-5.84677815e-01 -1.99970409e-01 2.50454515e-01 7.81292856e-01
-2.92280018e-01 -1.24444199e+00 -7.79490769e-02 1.60558134e-01
-3.14344198e-01 -5.67798972e-01 -4.10908848e-01 8.45318675e-01
-1.60763822e-02 1.10343874e+00 -1.46471530e-01 6.50721267e-02
4.93871063e-01 2.24620432e-01 5.65235436e-01 -6.57931805e-01
-2.06412166e-01 -4.49198395e-01 3.70668650e-01 -4.58512843e-01
-3.48929822e-01 -4.17750657e-01 -9.23429012e-01 -3.17283869e-02
-5.30798674e-01 4.18886483e-01 4.08706605e-01 1.05344176e+00
2.54961103e-01 2.10080951e-01 8.46103281e-02 3.99600118e-01
-6.16804659e-01 -1.05142510e+00 -2.85404354e-01 3.75507623e-01
1.50455534e-01 -2.52597958e-01 5.31332456e-02 1.55796289e-01] | [10.878190994262695, 8.357419967651367] |
6430de06-d563-432b-aa63-ed1c7dd78cc0 | learning-occupancy-function-from-point-clouds | 2010.11378 | null | https://arxiv.org/abs/2010.11378v1 | https://arxiv.org/pdf/2010.11378v1.pdf | Learning Occupancy Function from Point Clouds for Surface Reconstruction | Implicit function based surface reconstruction has been studied for a long time to recover 3D shapes from point clouds sampled from surfaces. Recently, Signed Distance Functions (SDFs) and Occupany Functions are adopted in learning-based shape reconstruction methods as implicit 3D shape representation. This paper proposes a novel method for learning occupancy functions from sparse point clouds and achieves better performance on challenging surface reconstruction tasks. Unlike the previous methods, which predict point occupancy with fully-connected multi-layer networks, we adapt the point cloud deep learning architecture, Point Convolution Neural Network (PCNN), to build our learning model. Specifically, we create a sampling operator and insert it into PCNN to continuously sample the feature space at the points where occupancy states need to be predicted. This method natively obtains point cloud data's geometric nature, and it's invariant to point permutation. Our occupancy function learning can be easily fit into procedures of point cloud up-sampling and surface reconstruction. Our experiments show state-of-the-art performance for reconstructing With ShapeNet dataset and demonstrate this method's well-generalization by testing it with McGill 3D dataset \cite{siddiqi2008retrieving}. Moreover, we find the learned occupancy function is relatively more rotation invariant than previous shape learning methods. | ['Matthew Kyan', 'Meng Jia'] | 2020-10-22 | null | null | null | null | ['3d-shape-representation'] | ['computer-vision'] | [-8.15840662e-02 -1.99850965e-02 -4.58328985e-02 -4.59245533e-01
-7.16964185e-01 -2.77706116e-01 5.42524695e-01 -2.21807718e-01
-5.75138479e-02 4.58191216e-01 -1.47325769e-01 -2.50264019e-01
-1.32474318e-01 -1.24407721e+00 -1.27659571e+00 -5.71962953e-01
1.12728961e-02 1.09184432e+00 2.61597097e-01 2.11194460e-03
4.23013538e-01 1.25650859e+00 -1.67110431e+00 2.22561136e-01
6.18963242e-01 1.10862899e+00 2.70349294e-01 1.46145776e-01
-7.06036448e-01 1.30316898e-01 -1.33855060e-01 1.80561766e-01
3.63708079e-01 4.13574189e-01 -6.61847591e-01 -1.39520183e-01
5.26153684e-01 -2.49406070e-01 -3.45367610e-01 6.43886745e-01
4.55344170e-01 1.29263718e-02 1.09382498e+00 -1.03206754e+00
-9.24801052e-01 -3.51577178e-02 -4.71421272e-01 -2.90512681e-01
-1.48983732e-01 4.05502319e-02 6.74510360e-01 -1.46666169e+00
3.99146080e-01 1.37845004e+00 1.01867127e+00 4.37750965e-01
-1.12563372e+00 -9.36638594e-01 1.40447728e-02 -2.28805408e-01
-1.53905463e+00 -1.62701890e-01 1.09336519e+00 -4.12275463e-01
1.19016337e+00 3.96787286e-01 8.33711922e-01 7.61622846e-01
5.80747984e-03 8.66090238e-01 5.64694107e-01 -6.94360808e-02
3.65390539e-01 -3.07547837e-01 -2.84326732e-01 6.02834642e-01
1.80299819e-01 1.62523270e-01 -2.88026899e-01 -5.09060979e-01
1.71106958e+00 4.91345882e-01 -1.33797243e-01 -7.94455051e-01
-1.14048481e+00 7.27877676e-01 7.48466015e-01 5.15567921e-02
-3.02259237e-01 5.29779315e-01 -1.47419302e-02 7.16360360e-02
8.16691101e-01 2.32491136e-01 -7.17661798e-01 1.22404337e-01
-7.30012596e-01 4.77773875e-01 5.45144379e-01 1.17805243e+00
1.16710901e+00 1.88077003e-01 6.47542924e-02 8.63998890e-01
6.23627663e-01 8.48179281e-01 6.40955344e-02 -9.67991650e-01
1.66873276e-01 9.59514439e-01 1.48652211e-01 -8.60369027e-01
-2.24680796e-01 -2.86259234e-01 -1.13830698e+00 4.42511737e-01
-1.46251515e-01 5.08026659e-01 -1.17386067e+00 1.17786419e+00
4.94785130e-01 7.39556313e-01 -3.47298354e-01 8.36967885e-01
1.00187969e+00 7.65250027e-01 -2.88372338e-01 1.48753732e-01
8.42908859e-01 -3.23315084e-01 -1.21923715e-01 2.23747358e-01
3.34857076e-01 -3.71427000e-01 1.05931425e+00 3.17029625e-01
-1.05091381e+00 -6.08046710e-01 -9.94693696e-01 -4.09794152e-01
-1.56407908e-01 -1.94172218e-01 8.40424895e-01 1.64947525e-01
-1.07456326e+00 8.97088528e-01 -1.04695225e+00 8.61825943e-02
1.02908087e+00 5.96596777e-01 -2.43747279e-01 -1.25998463e-02
-6.57445312e-01 5.35141826e-01 -7.71858310e-03 6.96476921e-03
-8.40980411e-01 -9.75769341e-01 -9.41712141e-01 -1.30980555e-02
-4.79010157e-02 -9.44691837e-01 1.03806674e+00 -5.20785391e-01
-1.47189021e+00 9.66374755e-01 -3.54410410e-01 -1.08708888e-01
1.49003103e-01 3.61839943e-02 1.48777412e-02 -1.92082718e-01
1.07547864e-01 8.65773141e-01 1.01959467e+00 -1.53465092e+00
-1.55567676e-01 -6.24194920e-01 -3.04711163e-01 4.64095399e-02
1.81354195e-01 -6.57155991e-01 -3.89582455e-01 -4.66866791e-01
8.15172315e-01 -6.88771248e-01 -2.21608713e-01 5.83578348e-01
-4.20072600e-02 -7.23835588e-01 9.78784621e-01 -2.05507159e-01
6.14363611e-01 -2.16442394e+00 -2.63359081e-02 4.52743292e-01
2.40828902e-01 1.34550914e-01 -1.64394751e-01 2.13293374e-01
-8.12309980e-02 2.06106305e-01 -6.69919312e-01 -6.14790618e-01
1.52098909e-01 5.84780157e-01 -5.52478433e-01 6.89474285e-01
4.13279384e-01 1.30354261e+00 -5.64723074e-01 -3.19392085e-02
4.81972009e-01 8.96996915e-01 -7.21298158e-01 1.04130790e-01
-5.86383939e-01 3.93033326e-01 -5.05626380e-01 1.00380611e+00
1.08667326e+00 -3.28979671e-01 -2.76261359e-01 -2.57579654e-01
-2.15741172e-01 2.19899669e-01 -1.00130630e+00 1.90371573e+00
-3.06985170e-01 1.18797317e-01 -2.64997687e-03 -9.65505481e-01
1.44711769e+00 1.47597015e-01 7.38858104e-01 -5.88544250e-01
-1.25998005e-01 3.73452604e-01 -4.95178401e-01 -1.07429788e-01
3.02687824e-01 -3.71657550e-01 5.52304648e-02 1.36831969e-01
-2.80745119e-01 -7.30900347e-01 -8.57109666e-01 -4.10671204e-01
8.32131922e-01 4.68076736e-01 1.91985946e-02 -4.07934606e-01
3.53857249e-01 -1.09619670e-01 4.70439941e-01 4.28360611e-01
2.10449949e-01 9.34467554e-01 1.05437383e-01 -9.97122705e-01
-1.31191468e+00 -1.39360809e+00 -5.49727261e-01 6.12136185e-01
2.23797947e-01 -2.99211144e-02 -2.31909245e-01 -3.52207780e-01
6.62411332e-01 5.55952787e-01 -6.16479933e-01 3.31358761e-02
-8.52236807e-01 -1.67898998e-01 4.17868495e-01 6.36676252e-01
3.33772987e-01 -1.33023632e+00 -2.42084175e-01 1.02365918e-01
3.24959308e-01 -5.07366538e-01 -3.38507891e-01 9.13242325e-02
-1.32388961e+00 -1.06261373e+00 -4.92394745e-01 -9.38786030e-01
8.07351649e-01 3.32489729e-01 1.11155570e+00 1.93234533e-01
-1.00491337e-01 3.37770909e-01 -5.22013828e-02 -6.69779956e-01
5.50773591e-02 6.72908947e-02 1.55351475e-01 -2.15290323e-01
5.99960685e-01 -1.13666499e+00 -6.49702430e-01 2.37655640e-01
-9.94880736e-01 -4.50492799e-02 5.63174069e-01 6.33908689e-01
1.42225122e+00 -1.98517993e-01 4.85006571e-01 -7.11744308e-01
2.10903004e-01 -5.51477432e-01 -6.75799727e-01 -1.51553452e-01
-2.48407885e-01 1.60725772e-01 4.10932869e-01 -1.43251911e-01
-5.36277652e-01 2.92028368e-01 -4.76175934e-01 -1.15418088e+00
-2.82861561e-01 1.44116551e-01 -1.20489120e-01 -2.79895127e-01
5.92747569e-01 5.28300464e-01 2.68038750e-01 -8.13143909e-01
2.12450936e-01 4.43068653e-01 3.44403565e-01 -8.91001940e-01
8.92735064e-01 1.02406847e+00 1.71835914e-01 -9.42986965e-01
-4.35999602e-01 -4.71790731e-01 -8.59823048e-01 1.30386651e-01
6.00728571e-01 -1.00131011e+00 -9.31300342e-01 3.93994093e-01
-1.54492795e+00 -3.04538488e-01 -4.92945760e-01 7.39871264e-02
-7.67704844e-01 2.20571548e-01 -2.02267557e-01 -8.61274600e-01
-5.41497111e-01 -1.01477623e+00 1.74002278e+00 -2.75084853e-01
1.90813109e-01 -8.95313978e-01 1.55045405e-01 -2.29475781e-01
3.79305363e-01 5.04305840e-01 9.89043117e-01 -1.58377409e-01
-1.12080359e+00 3.11732609e-02 -2.58046091e-01 2.06401557e-01
1.37468427e-01 -2.20513821e-01 -8.97457898e-01 -3.71289134e-01
1.63923591e-01 1.52610745e-02 8.43960583e-01 5.81426144e-01
1.94930136e+00 -2.58105159e-01 -5.82348764e-01 1.18195069e+00
1.46695125e+00 -1.72937766e-01 9.26264584e-01 8.70896876e-02
9.19679821e-01 1.40590817e-01 2.49496371e-01 3.98811787e-01
5.12376904e-01 4.54650015e-01 8.65071237e-01 -8.56327116e-02
1.39694354e-02 -5.57124019e-01 -6.90111443e-02 9.53516245e-01
-2.45463699e-01 3.55436027e-01 -9.96881127e-01 3.36937726e-01
-1.81664610e+00 -6.67218685e-01 -1.66904822e-01 2.08688545e+00
6.62413418e-01 -7.18980376e-03 -1.70461789e-01 -8.99166092e-02
4.10365790e-01 8.66902992e-02 -9.91799533e-01 -3.14568281e-02
-1.49602089e-02 7.43384600e-01 6.05686367e-01 4.45931375e-01
-9.86846685e-01 9.45636988e-01 5.96472788e+00 8.26786816e-01
-1.26024950e+00 4.40983735e-02 4.54784334e-01 -7.95120671e-02
-8.48330200e-01 -1.98342308e-01 -6.27376974e-01 3.26070160e-01
4.17707592e-01 2.42135197e-01 4.83838975e-01 1.02627170e+00
-2.40393933e-02 5.00876367e-01 -1.15908504e+00 1.48400760e+00
-1.50023565e-01 -1.77773082e+00 3.81749570e-01 2.71380663e-01
5.99342644e-01 5.34814715e-01 1.62202097e-03 4.11402196e-01
1.59437135e-01 -1.57765663e+00 6.43352628e-01 8.68304670e-01
1.31920040e+00 -7.28371143e-01 3.03962260e-01 6.72470987e-01
-1.37750936e+00 3.53456616e-01 -1.03390813e+00 -1.86881617e-01
7.42314830e-02 6.85178578e-01 -7.84706533e-01 4.10359234e-01
6.91228986e-01 9.91379082e-01 3.80207375e-02 1.00597298e+00
8.12113509e-02 2.91079432e-01 -6.55671418e-01 -6.51912950e-03
4.87742350e-02 -4.05773580e-01 5.41873932e-01 6.75906897e-01
5.17786026e-01 3.23020756e-01 7.43047893e-02 1.42437327e+00
-1.75210610e-01 -6.04029559e-02 -1.06265044e+00 2.69417584e-01
6.49183273e-01 9.55145121e-01 -4.70370412e-01 2.12805923e-02
-1.09554872e-01 6.08006299e-01 6.77715123e-01 2.15799794e-01
-5.61271489e-01 -4.16721776e-02 8.89514744e-01 6.85187578e-01
6.04474843e-01 -7.60693192e-01 -6.07844234e-01 -8.10050189e-01
-9.23035759e-03 -2.97833204e-01 -2.81117409e-01 -8.77241313e-01
-1.56450713e+00 2.73692638e-01 -2.79623605e-02 -1.37506413e+00
3.67309809e-01 -8.32992613e-01 -5.21563709e-01 1.08042049e+00
-1.70471728e+00 -1.20716810e+00 -2.16635674e-01 5.01840651e-01
4.95169848e-01 -1.32981956e-01 9.01107192e-01 2.36163288e-01
1.38385981e-01 2.82581419e-01 -1.02256477e-01 2.76633874e-02
7.25616794e-03 -1.09727323e+00 9.19142187e-01 -1.04125252e-03
1.95756048e-01 7.24374712e-01 1.97531339e-02 -8.97563815e-01
-1.94743347e+00 -1.27977586e+00 2.21931621e-01 -7.31298447e-01
3.89696471e-02 -5.48984587e-01 -1.25970173e+00 6.48506284e-01
-5.45690298e-01 4.19050545e-01 3.15844834e-01 -2.82052085e-02
-4.08397913e-01 -6.43638801e-03 -1.34724653e+00 1.73429847e-01
1.53979230e+00 -5.46860933e-01 -6.76417530e-01 3.44188362e-01
1.01512742e+00 -6.90136671e-01 -8.97747874e-01 9.53177989e-01
3.80117744e-01 -7.21717000e-01 1.27879763e+00 -5.47551394e-01
1.61733374e-01 -4.19720471e-01 -3.91685218e-01 -9.80544329e-01
-7.59920537e-01 -1.72872260e-01 -3.68978709e-01 5.64064443e-01
8.99871439e-02 -6.47600651e-01 1.13417089e+00 3.18993300e-01
-6.15232170e-01 -1.26753652e+00 -1.34315097e+00 -5.60979247e-01
4.32241291e-01 -6.60319269e-01 1.15174580e+00 1.02903807e+00
-6.47906899e-01 1.73910469e-01 -1.51530700e-02 4.07004505e-01
6.95535898e-01 2.66192436e-01 9.32509601e-01 -1.74594617e+00
9.90527943e-02 -4.18149382e-01 -3.05570453e-01 -1.53124714e+00
3.12549531e-01 -1.33616483e+00 -1.60175741e-01 -1.80109179e+00
-2.31603429e-01 -1.04794645e+00 -5.39182536e-02 6.66179538e-01
3.58501345e-01 1.83770567e-01 -2.66315103e-01 5.59095860e-01
-2.76553154e-01 1.13148987e+00 1.51837683e+00 -1.98957890e-01
-1.45290688e-01 8.04309770e-02 -4.77306217e-01 7.04929352e-01
6.38522327e-01 -2.64868557e-01 -3.44913572e-01 -7.20047474e-01
2.77264774e-01 -4.66785580e-02 5.15677750e-01 -8.22341442e-01
1.68349102e-01 -2.52659827e-01 6.60864949e-01 -1.35261142e+00
6.75680578e-01 -1.11018014e+00 4.80203331e-01 1.96079642e-01
3.24977249e-01 -5.94525747e-02 3.73065412e-01 6.37328088e-01
1.64023519e-01 9.39958617e-02 6.26939118e-01 -5.25509298e-01
-3.82148176e-01 1.16064334e+00 3.91883373e-01 -3.13595712e-01
7.70539641e-01 -5.10464072e-01 2.32734196e-02 1.79596469e-01
-4.33074594e-01 3.83570194e-02 7.49719739e-01 3.78457427e-01
1.33104491e+00 -1.78846085e+00 -6.77499771e-01 7.89312422e-01
-9.31979716e-02 1.18970454e+00 1.05191998e-01 4.23944652e-01
-8.59322011e-01 3.87045354e-01 1.87536553e-02 -1.25636256e+00
-5.25730252e-01 3.61685932e-01 5.28278470e-01 3.00821751e-01
-1.13656366e+00 8.20871055e-01 3.11674207e-01 -1.13111937e+00
9.54113603e-02 -8.28159094e-01 2.01098353e-01 -5.06452262e-01
1.24039836e-01 9.11939219e-02 1.32750824e-01 -4.93532956e-01
-3.11258614e-01 8.95792902e-01 2.12466314e-01 3.58922899e-01
1.80915868e+00 5.12379050e-01 -3.96106541e-01 6.21813834e-01
1.34065378e+00 -1.70638248e-01 -1.43778181e+00 -2.36200601e-01
-3.07279706e-01 -6.19489014e-01 -3.47289443e-02 -2.96474874e-01
-9.18299854e-01 8.53843153e-01 4.50400382e-01 1.36760339e-01
4.24493939e-01 2.94467866e-01 7.74254978e-01 4.37792987e-01
8.12557876e-01 -6.39200389e-01 -3.01036518e-02 8.78780782e-01
1.35405791e+00 -1.12464130e+00 6.79132491e-02 -4.36610430e-01
-5.27787860e-03 1.04639328e+00 5.62324822e-01 -6.55139089e-01
1.09854209e+00 2.03678817e-01 -3.68234396e-01 -7.23816514e-01
-4.89637583e-01 1.70372710e-01 3.85165662e-01 7.35373318e-01
1.13842860e-01 3.58820707e-01 2.92554826e-01 4.01705325e-01
-3.80231202e-01 8.21554661e-03 -1.83073401e-01 7.36484170e-01
-7.60326743e-01 -9.26271498e-01 -4.29932058e-01 7.30849028e-01
4.11316544e-01 -7.15559423e-02 -5.72663918e-03 6.67199314e-01
1.08986795e-01 -1.37986064e-01 5.85954130e-01 -2.56279767e-01
5.52014351e-01 6.06560856e-02 6.25636458e-01 -8.60620677e-01
-1.84167586e-02 -1.19105168e-01 -5.81430972e-01 -5.62278748e-01
-3.05221558e-01 -6.10871911e-01 -1.67386246e+00 -3.09934437e-01
-5.28513379e-02 -1.14333808e-01 7.90619493e-01 7.93983102e-01
7.45205224e-01 4.26489532e-01 5.97097397e-01 -1.48205090e+00
-4.38813597e-01 -8.33901048e-01 -6.83758974e-01 3.45444828e-01
2.79557437e-01 -9.87884402e-01 -2.36806110e-01 -5.11431694e-01] | [8.311874389648438, -3.6449122428894043] |
51f62dfd-b750-4ce0-b72d-d68d8a4f25ae | search-time-efficient-device-constraints | 2307.04443 | null | https://arxiv.org/abs/2307.04443v1 | https://arxiv.org/pdf/2307.04443v1.pdf | Search-time Efficient Device Constraints-Aware Neural Architecture Search | Edge computing aims to enable edge devices, such as IoT devices, to process data locally instead of relying on the cloud. However, deep learning techniques like computer vision and natural language processing can be computationally expensive and memory-intensive. Creating manual architectures specialized for each device is infeasible due to their varying memory and computational constraints. To address these concerns, we automate the construction of task-specific deep learning architectures optimized for device constraints through Neural Architecture Search (NAS). We present DCA-NAS, a principled method of fast neural network architecture search that incorporates edge-device constraints such as model size and floating-point operations. It incorporates weight sharing and channel bottleneck techniques to speed up the search time. Based on our experiments, we see that DCA-NAS outperforms manual architectures for similar sized models and is comparable to popular mobile architectures on various image classification datasets like CIFAR-10, CIFAR-100, and Imagenet-1k. Experiments with search spaces -- DARTS and NAS-Bench-201 show the generalization capabilities of DCA-NAS. On further evaluating our approach on Hardware-NAS-Bench, device-specific architectures with low inference latency and state-of-the-art performance were discovered. | ['Sumeet Agarwal', 'Tanu Kanvar', 'Oshin Dutta'] | 2023-07-10 | null | null | null | null | ['architecture-search', 'edge-computing'] | ['methodology', 'time-series'] | [-3.3621567e-01 -4.9310288e-01 -2.7194011e-01 -2.5797215e-01
-5.1505083e-01 -3.9958113e-01 2.4086151e-01 -1.8698423e-01
-7.0611203e-01 3.0511463e-01 -3.1825933e-01 -8.8480639e-01
-5.9800144e-02 -7.6232725e-01 -8.2983422e-01 -2.1978311e-01
-2.6538840e-01 4.5426238e-01 1.3593054e-01 2.1655242e-01
2.9519282e-02 5.3641391e-01 -1.5602869e+00 4.5782784e-01
2.3584527e-01 1.8098564e+00 2.7708250e-01 8.9194554e-01
-7.2241038e-02 6.4084333e-01 -3.3562300e-01 -3.6555436e-01
6.7194641e-01 2.6547816e-01 -5.7036716e-01 -5.8263552e-01
6.5767771e-01 -3.2972911e-01 -3.8052481e-01 1.0390991e+00
8.2107633e-01 -5.7817034e-02 3.1504142e-01 -1.4509171e+00
-5.6826174e-01 5.5386549e-01 -1.3636835e-01 6.3099670e-01
-3.2403207e-01 1.7802572e-01 9.5903814e-01 -1.0592252e+00
3.7051609e-01 8.3951789e-01 1.0561265e+00 6.7513388e-01
-9.0374267e-01 -7.5499326e-01 3.2839647e-01 5.5936015e-01
-1.7991643e+00 -8.1659621e-01 4.8343983e-01 -6.1801702e-02
1.9156125e+00 3.0857140e-01 8.4016156e-01 1.0592432e+00
2.8390485e-01 8.2319599e-01 4.4985300e-01 -3.4351486e-01
9.1531688e-01 -1.2505916e-01 1.1325967e-01 9.2860085e-01
4.2425311e-01 8.4348945e-03 -7.6903468e-01 -7.2755225e-02
6.9348133e-01 -4.2302508e-02 6.1154366e-02 -1.7024826e-01
-9.5496750e-01 4.5579126e-01 6.5079957e-01 1.5642400e-01
-6.1496383e-01 7.4349076e-01 7.8001654e-01 3.4720275e-01
3.0343911e-01 2.8349656e-01 -9.7998029e-01 -3.6814350e-01
-1.0490974e+00 -4.8792850e-02 9.1697502e-01 1.2803757e+00
4.3233240e-01 4.0391847e-01 7.5739250e-03 3.7168053e-01
3.6314052e-01 2.8816551e-01 6.1948210e-01 -8.6646509e-01
4.2698765e-01 3.9108926e-01 -1.7214468e-01 -6.0203743e-01
-7.1672940e-01 -6.5010011e-01 -8.1529647e-01 3.5983378e-01
2.0273196e-02 -3.5825279e-01 -9.9902195e-01 1.4082847e+00
6.4981081e-02 5.0235599e-01 2.0614689e-02 9.4688338e-01
8.9900815e-01 3.7077263e-01 2.0872383e-01 3.1308368e-01
1.6469220e+00 -1.3330684e+00 -1.8165132e-01 -6.2070012e-01
9.4710577e-01 -3.7709036e-01 1.0207258e+00 3.6195171e-01
-1.2366190e+00 -5.5780768e-01 -1.2593985e+00 -2.2008891e-01
-5.5959392e-01 3.7658432e-01 1.1210680e+00 9.6438718e-01
-1.6091925e+00 3.6968675e-01 -1.1226841e+00 -1.4524548e-01
8.3789122e-01 8.8878834e-01 2.0816885e-02 4.0252724e-01
-8.1950057e-01 5.6143141e-01 5.1555687e-01 -8.9014340e-03
-6.9830143e-01 -9.0024161e-01 -5.3403956e-01 5.4125780e-01
-3.4598466e-02 -1.0878302e+00 1.2288646e+00 -8.8634276e-01
-1.6901017e+00 8.0393088e-01 -1.5640944e-01 -9.9981987e-01
1.7626487e-01 -1.3525987e-01 -7.3837191e-01 -7.5796068e-02
-3.1084254e-01 1.1266069e+00 8.7665564e-01 -4.1179776e-01
-9.1413784e-01 -3.5431519e-01 5.9057746e-02 6.7234121e-02
-8.1275147e-01 -1.5518355e-01 -9.0202379e-01 -4.0472701e-01
-8.0945678e-02 -1.2907646e+00 -3.0872706e-01 2.0670584e-01
-1.1430525e-01 -1.9753657e-01 8.1196523e-01 -1.8736838e-01
1.3252246e+00 -2.1953394e+00 -3.7765005e-01 2.9458049e-01
2.8657928e-01 3.0081609e-01 -1.6469494e-01 -4.6604583e-01
1.0377379e-01 3.7711382e-02 4.8865134e-01 -4.9448007e-01
1.0628691e-01 7.7730991e-02 2.0696807e-03 1.1792048e-01
-2.7185956e-01 1.2042135e+00 -4.9476451e-01 -3.3866256e-01
1.4930063e-01 5.2747113e-01 -9.4149983e-01 -5.2584219e-01
-2.5919104e-01 -2.9933235e-01 -3.8386998e-01 9.9081838e-01
3.7417969e-01 -6.8876022e-01 6.5914713e-02 -5.5398357e-01
1.1997522e-01 4.0485406e-01 -9.2560726e-01 2.0394583e+00
-9.5288664e-01 1.0237464e+00 1.6159138e-01 -7.4421018e-01
4.7466078e-01 2.6406446e-01 3.0979812e-01 -1.0337538e+00
4.7545570e-01 3.7659186e-01 -1.2657815e-01 -2.4071251e-01
4.5914999e-01 7.4647182e-01 5.2999008e-02 1.7742939e-01
2.9881448e-02 4.7610098e-01 -1.3563538e-01 -2.0981234e-01
1.2842501e+00 -3.6280537e-01 -1.4579017e-01 -6.7297626e-01
2.0286447e-01 4.4733796e-02 2.4676397e-01 9.6531636e-01
-2.9697496e-01 1.3727659e-01 -2.8811949e-01 -1.0309389e+00
-8.5426110e-01 -1.0161610e+00 -2.1328397e-01 1.3723000e+00
8.4428087e-02 -6.7600775e-01 -9.7885787e-01 -5.2308488e-01
-2.4946187e-01 6.7655319e-01 -2.6233211e-01 -9.3164131e-02
-3.9379367e-01 -8.4632105e-01 6.9911593e-01 9.6139711e-01
8.8967651e-01 -7.4900520e-01 -1.3059714e+00 1.9505754e-01
4.0148908e-01 -1.4083277e+00 -4.8670724e-01 5.2826965e-01
-1.0386814e+00 -6.0750353e-01 -4.1667107e-01 -9.8695928e-01
5.6321496e-01 1.9945063e-01 1.4048160e+00 1.4232442e-01
-4.3695122e-01 2.0289852e-01 1.4158830e-01 -4.6753430e-01
2.3191741e-01 5.5760854e-01 2.5465089e-01 -3.4623182e-01
7.6666701e-01 -6.5370041e-01 -9.7779089e-01 7.4441038e-02
-4.2009890e-01 7.5670227e-02 6.9240576e-01 7.1133089e-01
7.5954258e-01 2.3251520e-01 2.7226436e-01 -4.9458277e-01
6.5209138e-01 -3.2436115e-01 -8.6101973e-01 2.6708186e-01
-1.0801967e+00 2.1521275e-01 4.8133138e-01 -6.3970464e-01
-5.2340817e-01 5.1809567e-01 -2.7881545e-01 -8.4854227e-01
3.4760695e-03 4.5590329e-01 -8.6845398e-02 -4.8086071e-01
9.3846941e-01 6.4250588e-02 -5.1901174e-01 -2.4892269e-01
1.5478919e-01 7.5286418e-01 7.0756578e-01 -4.2334676e-01
9.0765581e-02 4.8012608e-01 5.6677468e-02 -6.3115287e-01
-3.5637644e-01 -2.3387730e-01 -5.2090116e-02 5.8755312e-02
9.0159279e-01 -1.4280381e+00 -9.5701867e-01 3.3684081e-01
-1.0450563e+00 -6.2908220e-01 -6.0536355e-02 5.7800108e-01
-2.1539289e-01 -3.1355342e-01 -7.2309065e-01 -4.7132298e-01
-1.0600902e+00 -1.4319695e+00 1.1776539e+00 3.6893812e-01
-2.7893797e-01 -1.0778505e+00 -6.1726075e-01 3.4263071e-02
9.8797506e-01 -4.4449961e-01 8.0683774e-01 -6.4969951e-01
-9.0757942e-01 -2.3625854e-01 -3.4036204e-01 -1.1932332e-01
-2.9135501e-01 -5.0716877e-01 -1.1174655e+00 -3.0438238e-01
-1.3673308e-01 -4.9045235e-02 6.4852387e-01 6.4143747e-01
1.6701634e+00 -2.6069024e-01 -5.4267716e-01 1.0848694e+00
1.4446770e+00 2.9934612e-01 4.8699871e-01 5.7915628e-01
6.2057555e-01 -3.9180201e-01 -1.4866920e-01 3.9234078e-01
4.4023663e-01 7.3644447e-01 5.0234103e-01 -1.8339491e-02
-1.3114494e-01 6.2630929e-02 1.7263034e-01 6.6954434e-01
3.3785064e-02 -1.5707891e-01 -8.8652253e-01 5.4334456e-01
-1.7377589e+00 -5.7586181e-01 8.7472729e-02 1.9055263e+00
3.9691898e-01 7.2148854e-01 -4.9979415e-02 -6.0704038e-02
2.8301445e-01 -2.1635936e-01 -9.3947971e-01 -6.2848318e-01
1.5792495e-01 3.5780779e-01 1.1193131e+00 2.1268293e-01
-1.0914820e+00 9.3171614e-01 6.9535255e+00 9.6790022e-01
-1.5397896e+00 4.0771642e-01 8.7597716e-01 -5.5055708e-01
-1.0782785e-01 -3.4954226e-01 -1.0302284e+00 6.1901551e-01
1.3170477e+00 1.8230399e-01 8.4731996e-01 1.6330080e+00
-3.3887163e-01 3.0912545e-01 -1.4368509e+00 1.8680962e+00
-1.7393669e-01 -1.9687648e+00 -2.1116737e-01 6.4493567e-02
6.9499844e-01 8.8932252e-01 2.5958097e-01 5.0386053e-01
2.2417410e-01 -9.7136611e-01 9.6835375e-01 8.2136706e-02
9.4315821e-01 -5.1669437e-01 4.6231011e-01 1.5820105e-01
-1.4501256e+00 -2.9083863e-01 -3.8575071e-01 -2.6607698e-01
-1.2240587e-01 5.4497635e-01 -7.0421535e-01 -1.6730697e-01
1.2645661e+00 2.1942529e-01 -5.4894960e-01 9.8768538e-01
2.8929812e-01 3.4685138e-01 -7.4513292e-01 -2.6751962e-01
2.2805004e-01 3.7253609e-01 1.0285259e-01 1.1779530e+00
6.2866670e-01 -2.2746244e-01 -1.1958288e-01 7.8981358e-01
-4.7464252e-01 -3.6059076e-01 -3.4898761e-01 1.3817006e-01
9.5839792e-01 1.1939877e+00 -9.8973811e-01 -4.6911263e-01
-4.9878958e-01 1.1305832e+00 1.6181086e-01 2.5081900e-01
-1.1270982e+00 -3.7516352e-02 9.7618979e-01 -1.4213637e-01
3.4360301e-01 -3.1878927e-01 -7.4902993e-01 -9.6471959e-01
-1.1047519e-02 -7.8292179e-01 3.0321041e-01 -7.1748966e-01
-8.4796834e-01 1.0067453e+00 -2.3320563e-01 -8.9868295e-01
-1.6832580e-01 -9.7731578e-01 -6.4502299e-01 5.8627939e-01
-1.3971624e+00 -9.6812296e-01 -3.1603974e-01 6.5660542e-01
8.7036550e-01 -5.9907800e-01 1.0053452e+00 8.6803138e-01
-6.8113148e-01 1.2217283e+00 -6.6342093e-02 -9.1715446e-03
5.6502510e-02 -8.4983718e-01 1.2497808e+00 6.6149288e-01
5.5385828e-01 4.9906263e-01 3.4062269e-01 -2.7975494e-01
-1.9486485e+00 -1.0521685e+00 5.0446600e-01 -1.7041454e-01
6.5356457e-01 -4.8054609e-01 -4.0156960e-01 7.3796076e-01
8.3292343e-02 6.4361477e-01 4.7644985e-01 1.6171178e-01
-2.3382869e-01 -2.6864630e-01 -9.1414404e-01 8.2580978e-01
1.5410700e+00 -8.0836898e-01 2.6059374e-01 4.7410551e-01
5.7338041e-01 -7.0678043e-01 -6.4045250e-01 2.8261855e-01
6.5587306e-01 -6.4894652e-01 1.2224219e+00 -4.6504563e-01
-2.5089195e-01 -7.5389616e-02 -3.2273635e-01 -1.0151985e+00
-3.4821746e-01 -5.8702886e-01 -6.6096646e-01 5.3639334e-01
7.3838216e-01 -6.6562015e-01 1.3421514e+00 9.7383624e-01
-3.6442980e-01 -9.5073491e-01 -1.1610280e+00 -9.6922988e-01
-3.2779545e-01 -9.1568124e-01 8.4891129e-01 6.1817533e-01
-4.0704373e-01 3.6566126e-01 1.3783561e-01 3.8880304e-01
3.6713687e-01 -1.3569626e-01 3.6254981e-01 -1.1325055e+00
-5.5974871e-01 -7.6475042e-01 -5.5152071e-01 -1.2031784e+00
9.0481006e-02 -7.6591992e-01 -2.3147914e-01 -1.1239377e+00
-4.7198817e-01 -8.3942109e-01 -4.9673182e-01 6.0288489e-01
3.3209810e-01 3.6006063e-01 1.2862328e-01 3.5041010e-01
-9.0727425e-01 4.8839867e-02 4.7672001e-01 -3.5648516e-01
-2.3716860e-01 -1.8174833e-01 -5.2285707e-01 8.8596469e-01
9.6416581e-01 -2.9475224e-01 -6.5845424e-01 -1.2236068e+00
6.1131793e-01 -3.8713896e-01 2.7055031e-01 -1.6499462e+00
8.2097292e-01 4.1236699e-01 4.7715983e-01 -2.8434518e-01
4.4913378e-01 -1.2157642e+00 2.4036473e-01 5.8681351e-01
-1.2584841e-01 4.9885061e-01 5.9327805e-01 3.2455727e-01
2.1223941e-01 -1.0208042e-01 3.2283247e-01 1.8684980e-02
-1.4164935e+00 4.4875965e-01 -2.8655109e-01 -1.5952206e-01
1.0534686e+00 -3.3659792e-01 -3.8510737e-01 6.2179647e-02
-5.4095709e-01 1.5500899e-01 1.5901561e-01 4.5875773e-01
5.7361466e-01 -1.2878045e+00 -2.1350549e-01 4.6643662e-01
6.2031844e-03 -1.5381782e-02 1.8950328e-01 6.8690079e-01
-1.0042796e+00 7.2025746e-01 -2.6266301e-01 -6.7514670e-01
-1.1238569e+00 6.1876428e-01 6.2191874e-01 -9.3434252e-02
-6.2696397e-01 1.2537854e+00 -7.0315078e-02 1.7355729e-02
6.9503164e-01 -6.5983093e-01 3.5038048e-01 -5.0233012e-01
6.0154331e-01 2.9107651e-01 6.7619652e-01 -1.0206971e-01
-6.2914217e-01 2.4453093e-01 4.3657672e-02 9.8818772e-02
9.4050926e-01 -4.9415316e-02 3.0639666e-01 -2.3470506e-01
1.2741501e+00 -4.1700310e-01 -1.2281522e+00 -2.0655468e-02
1.6623102e-01 1.9996773e-02 1.0055230e+00 -7.7397889e-01
-1.5036587e+00 6.9910777e-01 1.3779973e+00 -3.9698295e-02
1.4424397e+00 -2.5705233e-01 1.1818255e+00 9.6250463e-01
7.0478088e-01 -1.3397689e+00 -3.2910311e-01 4.9485829e-01
2.4887332e-01 -1.1796322e+00 -3.0311054e-01 -1.3140699e-01
-1.1315090e-01 9.0300310e-01 7.0171750e-01 5.5382624e-02
1.0709931e+00 8.2660294e-01 -9.1092162e-02 -3.3818763e-01
-8.8008678e-01 -6.4778864e-02 2.5658593e-01 4.1299644e-01
4.3306235e-02 1.2522130e-01 2.3303483e-01 6.7860538e-01
-2.0315091e-01 3.7490311e-01 -3.0560693e-01 9.4214839e-01
-2.6782580e-02 -7.8178304e-01 -4.2087480e-02 4.9760097e-01
-5.3668910e-01 -5.8248925e-01 2.7782332e-02 4.2856565e-01
1.4365099e-01 6.4059460e-01 5.4659867e-01 -6.9656527e-01
1.6751483e-03 1.9031939e-01 2.8907672e-01 -1.1923497e-01
-7.2701991e-01 -2.8981277e-01 1.1969976e-01 -7.6101249e-01
9.9821888e-02 -7.1324721e-02 -1.4347889e+00 -4.8727176e-01
-3.3408895e-01 -2.6239744e-01 1.2974639e+00 8.0173010e-01
1.1451349e+00 7.6300591e-01 -8.0908075e-02 -1.0588816e+00
-4.6259817e-01 -3.5460624e-01 -1.2580061e-01 -1.6475666e-01
6.1811581e-02 -3.3254907e-01 -3.5280056e-02 -1.0579394e-01] | [8.516139030456543, 2.918426513671875] |
fcd455eb-8328-4db9-a7f8-e518ea2c439c | nuwa-xl-diffusion-over-diffusion-for | 2303.12346 | null | https://arxiv.org/abs/2303.12346v1 | https://arxiv.org/pdf/2303.12346v1.pdf | NUWA-XL: Diffusion over Diffusion for eXtremely Long Video Generation | In this paper, we propose NUWA-XL, a novel Diffusion over Diffusion architecture for eXtremely Long video generation. Most current work generates long videos segment by segment sequentially, which normally leads to the gap between training on short videos and inferring long videos, and the sequential generation is inefficient. Instead, our approach adopts a ``coarse-to-fine'' process, in which the video can be generated in parallel at the same granularity. A global diffusion model is applied to generate the keyframes across the entire time range, and then local diffusion models recursively fill in the content between nearby frames. This simple yet effective strategy allows us to directly train on long videos (3376 frames) to reduce the training-inference gap, and makes it possible to generate all segments in parallel. To evaluate our model, we build FlintstonesHD dataset, a new benchmark for long video generation. Experiments show that our model not only generates high-quality long videos with both global and local coherence, but also decreases the average inference time from 7.55min to 26s (by 94.26\%) at the same hardware setting when generating 1024 frames. The homepage link is \url{https://msra-nuwa.azurewebsites.net/} | ['Nan Duan', 'Houqiang Li', 'Zicheng Liu', 'Lijuan Wang', 'Gong Ming', 'Jianlong Fu', 'Fan Yang', 'Shuguang Liu', 'Linjie Li', 'Zhengyuan Yang', 'Minheng Ni', 'Xiaodong Wang', 'JianFeng Wang', 'Huan Yang', 'Chenfei Wu', 'Shengming Yin'] | 2023-03-22 | null | null | null | null | ['video-generation'] | ['computer-vision'] | [-3.92236449e-02 -3.44988704e-02 -2.21100092e-01 6.07790388e-02
-1.00384450e+00 -5.34549296e-01 4.87861246e-01 -4.24724162e-01
-3.09199274e-01 9.59085822e-01 2.27406591e-01 -4.08480406e-01
2.83936799e-01 -9.57357526e-01 -8.62416744e-01 -7.10403562e-01
-2.48369891e-02 2.60350943e-01 5.77952325e-01 1.06411651e-01
2.00858247e-02 1.43536061e-01 -1.32566178e+00 3.83992225e-01
7.96638787e-01 8.51614833e-01 5.31180859e-01 1.12678647e+00
2.99889725e-02 1.16896558e+00 -8.83670211e-01 -4.30123031e-01
2.85904467e-01 -8.02713275e-01 -7.52214372e-01 1.09402619e-01
4.55534309e-01 -9.75337565e-01 -6.45383596e-01 8.59414041e-01
7.89863050e-01 6.14778176e-02 2.25259647e-01 -1.12689888e+00
-4.59350795e-01 9.52530026e-01 -7.68383026e-01 3.43644023e-01
4.27933931e-01 6.88177943e-01 7.75808156e-01 -7.52287030e-01
8.27129602e-01 1.19203305e+00 4.92330849e-01 7.21691608e-01
-8.60293269e-01 -8.93100083e-01 6.46662116e-02 1.53392300e-01
-1.52686858e+00 -5.71716726e-01 3.32610965e-01 -2.56057322e-01
7.09610164e-01 2.35289112e-01 8.82158935e-01 1.31387031e+00
7.35215098e-02 9.22218442e-01 6.41675115e-01 -1.14713736e-01
7.92334303e-02 -6.34474695e-01 -2.70894051e-01 7.38208890e-01
1.74893349e-01 9.69138071e-02 -5.45674801e-01 -2.02261619e-02
1.19230342e+00 -1.01558067e-01 -4.35214877e-01 4.67283368e-01
-1.64476836e+00 6.95584655e-01 7.44837970e-02 1.52799919e-01
-3.98903817e-01 5.33718348e-01 4.46331888e-01 3.02860051e-01
3.92295659e-01 3.41199905e-01 -1.22444697e-01 -6.17753088e-01
-1.29217565e+00 5.53154171e-01 8.18755031e-01 1.19377291e+00
7.45640695e-01 2.34020337e-01 -5.77558339e-01 6.14869356e-01
5.13577946e-02 6.58229411e-01 4.58932459e-01 -1.49148977e+00
6.44002199e-01 -4.94893342e-02 1.18294396e-01 -9.25888360e-01
9.95754823e-02 -1.89063519e-01 -1.10032153e+00 -1.75632507e-01
4.16364640e-01 -8.04588914e-01 -7.52293169e-01 1.54822922e+00
3.34261984e-01 6.96480751e-01 -8.27079266e-02 8.67499530e-01
6.42033160e-01 1.28263283e+00 -2.34924823e-01 -4.69933540e-01
1.19972539e+00 -1.41036654e+00 -6.94515049e-01 1.25179276e-01
4.88903999e-01 -9.23961878e-01 9.82430339e-01 4.86920089e-01
-1.60622656e+00 -8.11995327e-01 -9.78489339e-01 -1.11054823e-01
2.58408338e-01 4.13906574e-02 3.83604318e-01 1.53197810e-01
-1.27758169e+00 4.82385814e-01 -9.28636014e-01 -9.51144670e-04
3.52088124e-01 1.25447989e-01 1.46645546e-01 -1.76594675e-01
-1.25226784e+00 2.02856407e-01 5.00788033e-01 8.83387178e-02
-1.09542191e+00 -6.41700506e-01 -6.09556258e-01 8.23004767e-02
6.32471085e-01 -8.45897138e-01 1.30156922e+00 -9.27440882e-01
-1.68022740e+00 1.10663794e-01 -2.91685402e-01 -6.83558166e-01
8.04554462e-01 -4.10719246e-01 -2.73065716e-01 4.35361564e-01
3.53847891e-02 1.14628458e+00 9.30680096e-01 -8.84806216e-01
-7.39025474e-01 2.72398889e-01 3.32481652e-01 9.52233300e-02
-1.81523740e-01 -7.64438510e-02 -1.14379656e+00 -9.83916759e-01
-4.17178750e-01 -1.11933029e+00 -2.42859915e-01 -1.91612229e-01
-4.59154010e-01 -1.88941851e-01 7.50179291e-01 -6.59042478e-01
1.64742160e+00 -2.11128664e+00 8.69591907e-02 -2.41845357e-03
2.83078700e-01 3.09995562e-01 -2.46794477e-01 4.21957165e-01
3.42733026e-01 1.28370196e-01 -1.26147926e-01 -3.57991040e-01
-7.03866407e-02 1.67518169e-01 -4.95678991e-01 1.45714562e-02
-1.54514149e-01 1.00330126e+00 -1.05313182e+00 -6.99038804e-01
6.38238387e-03 5.67890108e-01 -8.11209619e-01 3.59395027e-01
-5.29473424e-01 3.19580317e-01 -3.16173583e-01 2.94894189e-01
6.58683360e-01 -4.41734403e-01 2.59899467e-01 -1.70342758e-01
-5.95951676e-02 1.16272338e-01 -1.21002340e+00 1.93487144e+00
-5.06251335e-01 7.77649939e-01 -2.88097918e-01 -3.92670572e-01
5.60966790e-01 4.32251394e-01 4.32803363e-01 -4.78358299e-01
-1.22771241e-01 1.12365454e-01 -2.51634806e-01 -4.84291881e-01
5.58079541e-01 3.96007538e-01 7.67692477e-02 6.72684491e-01
-1.25951678e-01 -5.49166650e-03 8.09834003e-01 4.12312955e-01
1.03220868e+00 3.85153294e-01 -5.22858836e-02 1.21325240e-01
2.99545944e-01 -1.22660972e-01 5.89960635e-01 7.89730132e-01
3.14289331e-01 7.60229290e-01 6.17823422e-01 -4.36736345e-01
-1.19832015e+00 -1.03980696e+00 3.85378927e-01 8.06578398e-01
2.78949976e-01 -9.52902913e-01 -1.05916810e+00 -6.32171929e-01
-4.59348738e-01 6.08085334e-01 -2.36865416e-01 5.17769568e-02
-8.63072097e-01 -5.83469927e-01 8.74428988e-01 4.86486346e-01
7.83502162e-01 -1.17841232e+00 -6.84407711e-01 3.02904665e-01
-5.50511003e-01 -1.11464763e+00 -8.90946031e-01 -7.21945584e-01
-8.33428919e-01 -8.88430655e-01 -9.41651344e-01 -5.74351788e-01
5.34505427e-01 3.30968827e-01 1.09402847e+00 3.30179185e-01
-1.08809806e-02 -2.12317631e-01 -4.04596746e-01 2.20121846e-01
-6.16913080e-01 2.30336398e-01 -2.54832923e-01 -2.30813578e-01
1.72840226e-02 -4.33954209e-01 -9.05978322e-01 4.03492808e-01
-1.00335181e+00 5.23314655e-01 6.37667060e-01 8.52064192e-01
7.09494650e-01 1.47700474e-01 5.46596110e-01 -7.68600583e-01
4.22924966e-01 -5.50468206e-01 -6.89803660e-01 1.25242129e-01
-3.41809928e-01 -1.42985940e-01 7.76301205e-01 -5.09407640e-01
-1.08597291e+00 -2.41600826e-01 -4.48969364e-01 -6.48729861e-01
-1.09985089e-02 3.66242230e-01 3.48898955e-02 5.23962617e-01
3.86625707e-01 4.40523744e-01 1.21293515e-02 -2.16552436e-01
3.17365110e-01 5.04641473e-01 4.87359583e-01 -5.86175740e-01
6.06430709e-01 3.23836833e-01 -2.97979057e-01 -8.01639497e-01
-6.79017782e-01 -8.99908170e-02 -2.53060013e-01 -1.87380672e-01
8.55840981e-01 -1.18242419e+00 -6.76956236e-01 6.55621171e-01
-1.27951705e+00 -9.33407009e-01 -3.01685601e-01 5.26789486e-01
-4.88107860e-01 5.71761727e-01 -1.09285009e+00 -2.82727033e-01
-4.73153263e-01 -1.27893281e+00 1.14445424e+00 1.03458740e-01
-2.95772880e-01 -7.51831651e-01 -6.62890077e-02 1.78843528e-01
3.92062336e-01 1.23076156e-01 2.69658536e-01 8.85741040e-02
-1.04665673e+00 1.36966094e-01 -2.16740891e-01 3.93107414e-01
-6.03179373e-02 3.41735184e-01 -4.75300491e-01 -4.48380470e-01
-2.55406022e-01 -3.07889611e-01 8.33102882e-01 2.69025028e-01
1.51681149e+00 -5.70323706e-01 -1.30585790e-01 7.51811147e-01
1.25051606e+00 1.71187714e-01 1.01838732e+00 -6.35763677e-03
8.19775403e-01 -5.16658165e-02 6.89756274e-01 7.19967723e-01
4.34489369e-01 6.89634740e-01 1.17016278e-01 9.67518426e-03
-5.27471483e-01 -4.14979607e-01 7.40704298e-01 1.13817728e+00
-3.64641994e-01 -7.27102637e-01 -5.76173723e-01 4.21162456e-01
-1.82503057e+00 -1.41796982e+00 -3.90597545e-02 2.04586935e+00
1.11300492e+00 1.86414003e-01 1.93678230e-01 -1.61510393e-01
7.64123261e-01 5.16083717e-01 -3.16797167e-01 -4.44133505e-02
3.94699201e-02 1.85440972e-01 3.68161321e-01 7.25911140e-01
-7.18804896e-01 1.03710377e+00 5.69380140e+00 1.32114565e+00
-1.19683194e+00 1.39694139e-01 7.99319565e-01 -6.45308793e-01
-3.19324702e-01 -5.27625419e-02 -1.00455081e+00 9.56655502e-01
1.18992949e+00 -9.06694382e-02 4.82724786e-01 6.32017970e-01
4.42386061e-01 -8.24451074e-02 -7.81254888e-01 1.05423665e+00
4.69777212e-02 -1.75628948e+00 3.49692494e-01 7.85353854e-02
9.06733930e-01 -2.17726063e-02 -9.35506970e-02 2.44815767e-01
3.53383839e-01 -6.83875501e-01 8.66174281e-01 3.71389925e-01
1.08130622e+00 -8.28591824e-01 5.88369071e-01 4.00877386e-01
-1.30205846e+00 1.52568534e-01 -5.07269740e-01 1.40564293e-01
6.23934329e-01 8.36103976e-01 -6.47141218e-01 4.05797482e-01
6.07142568e-01 6.90651357e-01 -3.77986550e-01 8.56637836e-01
-3.22183907e-01 8.11058342e-01 -3.83792877e-01 1.22259147e-01
3.03821772e-01 -2.06580311e-01 3.84137005e-01 1.37578833e+00
8.95122945e-01 1.28410071e-01 4.03610826e-01 5.84589779e-01
-2.26571769e-01 -3.30197692e-01 -3.34844321e-01 1.58603236e-01
6.92616940e-01 1.08722413e+00 -6.49998009e-01 -8.37433636e-01
-2.69136995e-01 1.26912022e+00 5.80147617e-02 4.09265250e-01
-1.56249237e+00 -5.08338451e-01 3.74809593e-01 1.84879526e-01
5.80550253e-01 -4.60410893e-01 3.20175052e-01 -1.38256228e+00
7.51324138e-03 -1.05024564e+00 3.26073259e-01 -8.13258469e-01
-7.52083540e-01 7.47145951e-01 2.60164682e-02 -1.26585388e+00
-6.92710280e-01 1.35468796e-01 -4.73856568e-01 5.84760129e-01
-1.36116278e+00 -9.73501682e-01 -5.03643274e-01 7.36627400e-01
1.07686770e+00 1.61129862e-01 4.23629284e-01 5.39394319e-01
-5.28936565e-01 6.91343307e-01 -1.07560329e-01 7.61352926e-02
7.32487559e-01 -9.87576306e-01 6.75725877e-01 9.83713210e-01
1.24312989e-01 5.09213924e-01 4.93141323e-01 -7.19993353e-01
-1.42588472e+00 -1.31830120e+00 7.66457081e-01 -3.63025926e-02
5.29369891e-01 -2.47823596e-01 -6.85757935e-01 5.75576842e-01
6.43619716e-01 -4.59751897e-02 4.81033444e-01 -5.06235123e-01
-9.93900448e-02 -1.94297850e-01 -5.74967444e-01 9.07557547e-01
1.23344195e+00 -2.77307868e-01 -6.47411123e-02 3.16750079e-01
9.70225394e-01 -6.66860521e-01 -1.04580677e+00 3.75404358e-02
5.53944826e-01 -1.19378459e+00 8.33298504e-01 6.66311681e-02
8.45041633e-01 -4.82686192e-01 1.23261727e-01 -1.08045244e+00
-1.58355653e-01 -1.19999993e+00 -7.28070974e-01 1.21130371e+00
4.49668556e-01 -2.90575862e-01 7.25640118e-01 -4.03006524e-02
-1.08631931e-01 -9.91723001e-01 -5.61353564e-01 -6.31387234e-01
-3.15195143e-01 -4.22064096e-01 8.92526627e-01 5.82523227e-01
-3.63235503e-01 3.23268712e-01 -6.49727464e-01 1.15025640e-01
4.41728860e-01 3.24577391e-01 1.06919336e+00 -5.13578176e-01
-7.95585990e-01 -2.98392594e-01 4.39375453e-02 -1.88701141e+00
-7.36381635e-02 -4.07839656e-01 4.99810427e-02 -1.42215264e+00
6.53833374e-02 -3.57552737e-01 1.44428477e-01 2.23694518e-01
-2.65187860e-01 4.38249528e-01 4.03751343e-01 3.45840514e-01
-8.53492439e-01 4.75538999e-01 1.63359571e+00 -1.32696116e-02
-8.22870433e-02 -1.30884647e-01 -4.19373661e-01 6.41808689e-01
7.30618000e-01 -1.71560168e-01 -7.39211559e-01 -8.30959678e-01
2.22432494e-01 4.31875408e-01 3.01895559e-01 -1.26958859e+00
1.63997650e-01 -1.53330624e-01 3.13703865e-01 -6.43763304e-01
3.77801597e-01 -4.06985819e-01 4.96517509e-01 5.71488798e-01
-2.19601288e-01 3.59449357e-01 -6.45140186e-02 4.43951845e-01
-3.18921000e-01 -3.35239917e-02 5.91411471e-01 -2.24346995e-01
-7.82874525e-01 7.27896571e-01 -3.03248435e-01 1.66817918e-01
1.10309172e+00 -1.18769124e-01 -3.96811634e-01 -6.22731924e-01
-5.39034367e-01 2.67089665e-01 5.49535990e-01 3.21678281e-01
6.94497705e-01 -1.36881685e+00 -8.12573075e-01 1.06990442e-01
-3.81084710e-01 3.75273943e-01 5.37424564e-01 8.21167588e-01
-9.48710084e-01 1.38894156e-01 7.15494305e-02 -5.97982407e-01
-1.08428299e+00 4.97503221e-01 -4.45804819e-02 -3.74909461e-01
-7.29038537e-01 8.65807235e-01 1.50778621e-01 3.66042018e-01
8.95026550e-02 -2.47248203e-01 9.29803327e-02 4.67663212e-03
8.65199506e-01 5.13864160e-01 -4.19216782e-01 -4.03316349e-01
2.43996903e-01 4.75286961e-01 -1.33776054e-01 -2.54453301e-01
1.09591901e+00 -2.32025146e-01 -2.78668627e-02 1.53517798e-01
1.20322359e+00 2.74239570e-01 -1.67937922e+00 9.37723145e-02
-6.38309479e-01 -6.68663800e-01 -1.98522091e-01 -3.64509016e-01
-1.41712296e+00 7.36541569e-01 1.12324253e-01 2.73807019e-01
1.15098214e+00 -1.60433069e-01 1.65331471e+00 1.83975860e-01
3.96873802e-01 -9.54572320e-01 2.30605632e-01 4.94094729e-01
7.17610419e-01 -7.36284673e-01 -4.73299548e-02 -3.09961021e-01
-7.57399976e-01 1.05244172e+00 6.73252642e-01 -1.55679017e-01
2.70526141e-01 4.03081954e-01 -1.95943505e-01 9.68345702e-02
-1.03036904e+00 2.27840409e-01 -1.82664171e-01 2.16497123e-01
3.95065933e-01 -1.26378089e-01 -3.05582255e-01 2.17929363e-01
-2.36569196e-01 3.37025285e-01 7.91866124e-01 6.86026514e-01
-3.09737206e-01 -1.18017566e+00 -2.20836952e-01 3.63291860e-01
-5.24808407e-01 -2.15892062e-01 2.84956872e-01 6.25414848e-01
2.89854258e-01 7.90079057e-01 2.69297779e-01 -2.69655854e-01
-6.12188168e-02 -3.49735022e-01 3.89810830e-01 -2.94370800e-01
-3.40166062e-01 4.29151684e-01 2.56963611e-01 -7.83480346e-01
-4.32753235e-01 -5.02623558e-01 -1.27187908e+00 -8.81238878e-01
-2.77209789e-01 3.04549515e-01 1.31735012e-01 5.45205772e-01
7.23960638e-01 6.27588093e-01 6.03656352e-01 -9.45142388e-01
-3.78206640e-01 -8.38637650e-01 -2.54058599e-01 2.65209705e-01
1.90660253e-01 -2.43236482e-01 -1.62107021e-01 3.75448585e-01] | [10.94238567352295, -0.6056921482086182] |
f4d6a5b6-c7f7-409d-8967-84c51ac9ce24 | deep-speech-scaling-up-end-to-end-speech | 1412.5567 | null | http://arxiv.org/abs/1412.5567v2 | http://arxiv.org/pdf/1412.5567v2.pdf | Deep Speech: Scaling up end-to-end speech recognition | We present a state-of-the-art speech recognition system developed using
end-to-end deep learning. Our architecture is significantly simpler than
traditional speech systems, which rely on laboriously engineered processing
pipelines; these traditional systems also tend to perform poorly when used in
noisy environments. In contrast, our system does not need hand-designed
components to model background noise, reverberation, or speaker variation, but
instead directly learns a function that is robust to such effects. We do not
need a phoneme dictionary, nor even the concept of a "phoneme." Key to our
approach is a well-optimized RNN training system that uses multiple GPUs, as
well as a set of novel data synthesis techniques that allow us to efficiently
obtain a large amount of varied data for training. Our system, called Deep
Speech, outperforms previously published results on the widely studied
Switchboard Hub5'00, achieving 16.0% error on the full test set. Deep Speech
also handles challenging noisy environments better than widely used,
state-of-the-art commercial speech systems. | ['Sanjeev Satheesh', 'Jared Casper', 'Greg Diamos', 'Awni Hannun', 'Andrew Y. Ng', 'Adam Coates', 'Erich Elsen', 'Carl Case', 'Shubho Sengupta', 'Ryan Prenger', 'Bryan Catanzaro'] | 2014-12-17 | null | null | null | null | ['accented-speech-recognition'] | ['speech'] | [-1.48924351e-01 -3.48420322e-01 4.81565177e-01 -4.61473018e-01
-1.05664301e+00 -5.17505944e-01 5.74315786e-01 -1.73046008e-01
-5.18927693e-01 4.00856078e-01 2.86556065e-01 -7.51350403e-01
4.18993473e-01 -4.45772618e-01 -6.57282293e-01 -6.95152998e-01
-3.19540575e-02 5.23452401e-01 2.77600199e-01 -7.21154809e-01
-6.78333566e-02 3.65760118e-01 -1.62665105e+00 1.84974432e-01
4.60213900e-01 8.28885555e-01 5.46104550e-01 1.13468874e+00
8.56734172e-04 8.01885188e-01 -1.25717831e+00 4.61133048e-02
1.19827829e-01 -3.19074512e-01 -6.20752931e-01 -1.94325373e-01
4.68264937e-01 -4.03346002e-01 -6.25135601e-01 6.75946295e-01
1.18882537e+00 5.03014386e-01 -1.00606024e-01 -4.81996357e-01
-3.97601634e-01 5.49014509e-01 1.54010460e-01 3.83425176e-01
1.25324741e-01 4.51778054e-01 7.11583674e-01 -8.61489594e-01
2.10554600e-01 1.17389071e+00 7.35151231e-01 6.68974459e-01
-1.17030668e+00 -6.13374174e-01 4.48111743e-02 5.36945499e-02
-1.12627196e+00 -1.28789890e+00 3.22143137e-01 -1.14279710e-01
1.69488299e+00 3.48859936e-01 5.14349937e-01 1.54760289e+00
-2.00370923e-01 5.51618576e-01 9.29718018e-01 -5.39715946e-01
4.42832500e-01 8.20566118e-02 3.77820395e-02 6.35494828e-01
-4.45170164e-01 3.70432615e-01 -7.36025631e-01 5.95133640e-02
4.48137432e-01 -4.48319823e-01 -5.10811508e-01 2.66165525e-01
-1.27955496e+00 4.19738322e-01 1.29205436e-01 3.14704955e-01
-2.60001481e-01 2.70900965e-01 6.90154791e-01 3.99256080e-01
4.92919773e-01 3.46922636e-01 -8.04646969e-01 -8.07299852e-01
-1.08026898e+00 -1.34828031e-01 1.19756532e+00 9.20244992e-01
3.81212085e-01 7.54648685e-01 -2.76091397e-02 1.26874685e+00
1.94602057e-01 5.80184281e-01 8.09439898e-01 -7.29192734e-01
4.25893158e-01 -2.27387637e-01 -2.67109245e-01 -3.39344352e-01
-4.37735498e-01 -9.27887321e-01 -7.11072922e-01 2.04524323e-01
2.62956768e-01 -3.39633256e-01 -1.08228791e+00 1.59738696e+00
4.55511659e-02 4.21836913e-01 3.03419620e-01 7.70080447e-01
8.90281737e-01 8.76842976e-01 -4.82301623e-01 1.59782171e-01
1.20634329e+00 -1.41147316e+00 -6.18752241e-01 -4.93946195e-01
3.96397859e-01 -1.07257295e+00 1.24884820e+00 7.48047411e-01
-1.17055392e+00 -5.57700396e-01 -1.23010576e+00 -2.86817640e-01
-3.69357288e-01 2.21857712e-01 6.34083688e-01 1.16775286e+00
-1.60468316e+00 4.04660702e-01 -9.62207913e-01 -4.55123633e-01
-1.83321140e-03 5.62999368e-01 -2.45556653e-01 1.99257731e-01
-9.41101611e-01 9.96255219e-01 -1.83699355e-02 8.47386867e-02
-1.27387917e+00 -6.69658542e-01 -7.16902971e-01 2.16854796e-01
2.55832404e-01 -5.46688378e-01 1.92687905e+00 -9.14882421e-01
-2.50727892e+00 5.56805968e-01 -3.41001600e-01 -6.11203372e-01
1.91189155e-01 -2.13028014e-01 -6.68733776e-01 -1.13436356e-01
-5.70715666e-01 2.73091048e-01 8.62132907e-01 -8.76938164e-01
-2.66547114e-01 1.51151896e-01 -2.49891192e-01 -9.08417553e-02
-4.43738431e-01 3.41258347e-01 -6.44944966e-01 -7.89633393e-01
-1.78668946e-01 -7.96457946e-01 -2.13455543e-01 -3.54729354e-01
-2.89789110e-01 6.82287887e-02 7.84583151e-01 -8.04795802e-01
9.59132850e-01 -2.23438883e+00 -2.89862633e-01 5.98255917e-02
-1.91858839e-02 8.28698695e-01 -2.45751590e-01 3.70026588e-01
4.84661236e-02 -1.28416359e-01 -6.70058504e-02 -9.02067721e-01
1.04817942e-01 2.20465317e-01 -4.49175417e-01 3.16788405e-01
5.72676361e-02 3.39337885e-01 -9.50130761e-01 2.88644880e-01
4.07054365e-01 7.02104628e-01 -7.98743725e-01 3.94341975e-01
-1.65019304e-01 2.93216765e-01 2.51904130e-01 3.81933361e-01
4.31252152e-01 1.23120703e-01 9.64751318e-02 1.90750614e-01
-2.87481874e-01 1.25277472e+00 -8.83385658e-01 1.90135324e+00
-9.93002534e-01 1.05533910e+00 5.78778505e-01 -9.62134957e-01
1.06192303e+00 5.43059707e-01 -2.54255887e-02 -6.11078084e-01
2.62324959e-01 5.10218740e-01 9.78062600e-02 -3.38098645e-01
5.13996005e-01 9.68503803e-02 3.42256457e-01 1.83803156e-01
2.91293472e-01 -4.25725609e-01 -1.23267792e-01 5.96156381e-02
1.38690662e+00 -2.05395341e-01 -5.66068478e-02 -2.91782111e-01
3.42724234e-01 -4.28462714e-01 5.40251434e-01 9.25134718e-01
-1.26910046e-01 8.65918875e-01 1.52917072e-01 -1.81502074e-01
-9.90586042e-01 -9.47162151e-01 1.49583267e-02 1.31009173e+00
-5.07666230e-01 -5.42088628e-01 -8.87111783e-01 -1.03463337e-01
-4.92552251e-01 8.16512525e-01 -6.30687252e-02 -5.67155555e-02
-5.63675165e-01 -7.03568757e-01 1.01693356e+00 4.73580658e-01
5.25779605e-01 -9.32269871e-01 -1.42672122e-01 5.13863802e-01
-4.04912047e-02 -1.22183371e+00 -6.50824845e-01 5.91994822e-01
-4.66557831e-01 -4.16151881e-01 -4.72332031e-01 -9.45587993e-01
1.52546898e-01 4.79563236e-01 1.38422775e+00 1.74276367e-01
-7.05907047e-02 2.37103000e-01 -3.55654538e-01 -3.49004567e-01
-8.77979517e-01 2.13361472e-01 3.67486924e-01 -2.16570929e-01
2.59679914e-01 -6.54587507e-01 -3.97660762e-01 2.57734895e-01
-5.59737802e-01 -1.27878815e-01 5.52061498e-01 1.17382884e+00
2.16473758e-01 -1.95508040e-02 6.59164190e-01 -4.05335397e-01
5.06049335e-01 -3.84439737e-01 -7.17489243e-01 1.58236444e-01
-2.31723443e-01 1.64109282e-02 9.77344275e-01 -3.81596774e-01
-1.01052380e+00 -6.48675635e-02 -8.69840980e-01 -1.12429410e-01
-2.85585314e-01 1.69071540e-01 -1.81094915e-01 -7.30657158e-03
7.72654235e-01 4.08241540e-01 8.88779312e-02 -6.87596619e-01
3.51134866e-01 1.19518137e+00 7.63922155e-01 -5.04608452e-01
3.96704853e-01 2.29497015e-01 -5.02477705e-01 -1.45957136e+00
-2.51086563e-01 -5.64765394e-01 -1.37481615e-01 1.37886554e-01
4.21769738e-01 -1.14060044e+00 -8.93715739e-01 7.57827461e-01
-1.32449830e+00 -7.57158399e-01 -5.76505661e-02 6.05817020e-01
-4.59560782e-01 1.32920340e-01 -8.92951012e-01 -8.57690632e-01
-6.70157313e-01 -1.37528813e+00 1.17553437e+00 -1.63373157e-01
-1.46919578e-01 -6.95840120e-01 8.85920748e-02 5.35422802e-01
9.33962286e-01 -5.64923465e-01 4.73816544e-01 -7.38888621e-01
-5.63821256e-01 -9.70398709e-02 1.25854984e-01 9.47349608e-01
7.49078542e-02 1.06123358e-01 -1.53552806e+00 -4.49427903e-01
1.29970282e-01 -3.48001957e-01 9.27694321e-01 1.50146052e-01
1.20569074e+00 -2.35037714e-01 6.65917322e-02 8.99169564e-01
7.78118789e-01 2.79138029e-01 6.02209389e-01 1.70401856e-01
4.02221709e-01 3.52498561e-01 -1.61325395e-01 3.49490881e-01
2.59359628e-01 8.65447760e-01 1.08730316e-01 -2.19983995e-01
-4.26135421e-01 8.89275130e-03 9.29074466e-01 1.45588100e+00
3.10771108e-01 -4.86340791e-01 -9.37202394e-01 4.49679077e-01
-1.40440738e+00 -1.03597462e+00 -5.77099249e-02 2.12611222e+00
9.14943099e-01 1.55646503e-01 1.03306659e-01 2.87574321e-01
3.53996247e-01 1.64259121e-01 -4.03482497e-01 -4.94709343e-01
-2.73689061e-01 8.35002005e-01 4.10158426e-01 8.10633957e-01
-9.46609080e-01 1.22615123e+00 7.29434919e+00 8.76243949e-01
-1.49020362e+00 2.44931981e-01 4.06228572e-01 -5.59119046e-01
-2.17464179e-01 -2.22657815e-01 -7.08060563e-01 3.33823323e-01
1.65177870e+00 3.32735032e-01 9.33133185e-01 8.64596248e-01
5.10250330e-01 1.55131608e-01 -9.57599998e-01 1.03040266e+00
2.06305265e-01 -1.33049357e+00 -5.44414878e-01 -2.16234490e-01
3.51447254e-01 7.17779338e-01 9.70779881e-02 5.81743300e-01
6.79886937e-01 -1.01150298e+00 9.21595454e-01 1.50539935e-01
6.10449851e-01 -5.49124122e-01 4.99257863e-01 2.48483539e-01
-8.45117331e-01 1.39156669e-01 -2.83075958e-01 -1.88646719e-01
4.32852134e-02 7.75896847e-01 -1.17087722e+00 5.52321225e-02
7.20474243e-01 4.02209908e-01 -3.90023828e-01 9.35158610e-01
-2.78296143e-01 1.16059601e+00 -6.35901332e-01 -2.41382137e-01
2.11280346e-01 2.62665272e-01 4.45873380e-01 1.57583463e+00
4.30816680e-01 -4.54968177e-02 6.45268941e-04 3.22811872e-01
-2.02366039e-01 -1.13648757e-01 -2.98456818e-01 -2.31322106e-02
7.12089598e-01 1.01958776e+00 -2.33993322e-01 -4.64173555e-01
-4.20482785e-01 1.08495796e+00 3.12683225e-01 4.51008230e-01
-6.41057789e-01 -4.98383969e-01 1.16140354e+00 -2.52966613e-01
4.87174213e-01 -6.77819014e-01 -1.79954916e-01 -1.28837419e+00
5.46777584e-02 -1.53794634e+00 -2.79569924e-01 -6.67552292e-01
-1.04554284e+00 1.12545204e+00 -8.44146669e-01 -8.10896575e-01
-4.22758460e-01 -7.71343410e-01 -6.90758109e-01 1.04478455e+00
-1.40262592e+00 -7.66479254e-01 -1.59182817e-01 5.00015795e-01
8.54514599e-01 -6.22029841e-01 1.24153709e+00 6.10144794e-01
-7.93553650e-01 8.08886826e-01 4.82780397e-01 -1.83616430e-02
8.11543226e-01 -1.19127858e+00 1.22015309e+00 1.01317811e+00
3.06385070e-01 8.31210613e-01 7.49998152e-01 -1.05407871e-01
-1.60886097e+00 -7.82644689e-01 6.63622558e-01 -3.97267073e-01
7.74325192e-01 -9.98232901e-01 -7.93848276e-01 4.98805434e-01
3.98425311e-01 7.92034939e-02 8.16027462e-01 4.31311011e-01
-5.99836230e-01 -4.04515952e-01 -7.94413149e-01 6.40442312e-01
1.02970183e+00 -8.23964119e-01 -3.68561149e-01 3.94095033e-01
1.02532327e+00 -6.20042443e-01 -3.96187574e-01 -9.16473344e-02
4.11296040e-01 -1.00343657e+00 9.87507701e-01 -4.92944241e-01
-6.86003640e-02 -2.39126533e-01 -3.93698722e-01 -1.76548982e+00
-1.83490515e-01 -1.27223396e+00 -1.14509374e-01 1.17378032e+00
6.40136719e-01 -8.27120841e-01 5.97509146e-01 1.64500564e-01
-9.05530751e-01 -4.64674085e-01 -9.77097631e-01 -1.00202799e+00
-6.82226121e-02 -8.10681939e-01 7.32008517e-01 6.93275213e-01
-4.73909639e-03 3.78121227e-01 -3.56168479e-01 3.86472225e-01
2.01392785e-01 -5.15014648e-01 9.74984705e-01 -6.78447962e-01
-9.08167541e-01 -6.68535173e-01 -2.11807325e-01 -1.39311564e+00
2.03990877e-01 -7.34104812e-01 5.62438905e-01 -1.22247791e+00
-6.33241415e-01 -6.39691412e-01 -1.07613176e-01 4.75611955e-01
5.08710858e-04 -5.29311039e-02 -1.76554429e-04 -1.68546900e-01
-2.56965190e-01 7.40889370e-01 5.58591962e-01 -9.71945822e-02
-2.29449257e-01 -3.95340985e-03 -5.47072172e-01 5.41376233e-01
8.68213594e-01 -1.35667175e-01 -3.96618754e-01 -9.70375001e-01
-2.49906600e-01 -2.09897310e-01 2.24216565e-01 -1.25854862e+00
4.45345670e-01 3.43769014e-01 -3.86530422e-02 -2.97881216e-01
7.43208468e-01 -4.91344124e-01 -1.79077774e-01 1.99207038e-01
-1.23687647e-01 3.10014319e-02 5.12991905e-01 1.68905795e-01
-2.47224376e-01 1.04007516e-02 8.03220153e-01 2.33075209e-02
-5.97746193e-01 -6.31572679e-02 -8.51805747e-01 -4.51363996e-02
3.03307563e-01 3.11126292e-01 -6.23869479e-01 -5.15707195e-01
-4.64996606e-01 -1.20936640e-01 1.23474330e-01 5.80829501e-01
3.35620433e-01 -8.93598497e-01 -7.19188452e-01 6.57045186e-01
-2.39336550e-01 -1.00123145e-01 1.32377073e-01 5.03555894e-01
-8.36053908e-01 4.34928864e-01 3.44878823e-01 -4.53930646e-01
-1.04496181e+00 1.62803262e-01 6.30465448e-01 8.39748830e-02
-6.05275273e-01 1.20204496e+00 -2.94556797e-01 -7.32532024e-01
6.49553239e-01 -6.70613587e-01 5.12749732e-01 -2.50054747e-01
8.42655122e-01 3.04503709e-01 8.73764575e-01 -3.46342742e-01
-4.33673531e-01 -1.20550752e-01 -5.37311211e-02 -5.07095277e-01
1.33853006e+00 1.59624010e-01 1.19167775e-01 3.50975573e-01
1.18408692e+00 1.46387383e-01 -1.10036600e+00 -2.95510739e-01
-2.28598878e-01 -2.13921532e-01 5.59840322e-01 -1.00105965e+00
-1.02766228e+00 1.01566005e+00 5.46194017e-01 1.43006042e-01
1.17239928e+00 -4.28010404e-01 1.28617811e+00 8.21600735e-01
4.25602198e-01 -1.17122781e+00 1.00564156e-02 1.20099819e+00
9.24753785e-01 -1.16101384e+00 -5.58460593e-01 -1.16201460e-01
-2.82204449e-01 1.02914751e+00 3.68863016e-01 2.53576338e-01
6.94325984e-01 9.45530474e-01 4.64691758e-01 2.47467980e-01
-1.13040221e+00 -2.62047380e-01 -6.11427911e-02 7.21914828e-01
6.13535464e-01 9.77696478e-02 3.41679245e-01 5.22972465e-01
-8.87928724e-01 -3.58666927e-01 4.04485941e-01 8.31859410e-01
-4.71206874e-01 -1.18552196e+00 -4.56246346e-01 6.82196766e-02
-3.21153492e-01 -6.37326717e-01 -2.11091906e-01 2.66578645e-01
-1.17593028e-01 1.39816594e+00 1.57744244e-01 -6.68274224e-01
4.43745911e-01 1.04964808e-01 2.87943929e-01 -7.03493595e-01
-1.11174548e+00 7.47139901e-02 4.30844575e-01 -4.47029501e-01
1.11053802e-01 -5.74997187e-01 -1.14437437e+00 -6.62479758e-01
-3.14365208e-01 -6.94927275e-02 1.23098671e+00 9.10555482e-01
5.14741063e-01 8.61014068e-01 6.03748143e-01 -9.63015079e-01
-6.26817822e-01 -1.16107142e+00 -2.77753472e-01 -5.57460859e-02
8.12339008e-01 -1.99014246e-01 -4.48547006e-01 -1.22375153e-01] | [14.466208457946777, 6.442663192749023] |
06a9eb27-15ca-4a9b-a4dd-c532e9be9921 | enabling-automatic-repair-of-source-code | 2202.03055 | null | https://arxiv.org/abs/2202.03055v1 | https://arxiv.org/pdf/2202.03055v1.pdf | Enabling Automatic Repair of Source Code Vulnerabilities Using Data-Driven Methods | Users around the world rely on software-intensive systems in their day-to-day activities. These systems regularly contain bugs and security vulnerabilities. To facilitate bug fixing, data-driven models of automatic program repair use pairs of buggy and fixed code to learn transformations that fix errors in code. However, automatic repair of security vulnerabilities remains under-explored. In this work, we propose ways to improve code representations for vulnerability repair from three perspectives: input data type, data-driven models, and downstream tasks. The expected results of this work are improved code representations for automatic program repair and, specifically, fixing security vulnerabilities. | ['Anastasiia Grishina'] | 2022-02-07 | null | null | null | null | ['program-repair', 'program-repair'] | ['computer-code', 'reasoning'] | [-1.65274948e-01 2.97882140e-01 -3.17821383e-01 -3.48344058e-01
-7.27917254e-01 -8.02243412e-01 -4.93820496e-02 6.05481088e-01
4.63119209e-01 1.13659211e-01 1.57179490e-01 -8.76338720e-01
1.68649212e-01 -1.03682005e+00 -9.03126121e-01 1.86867341e-01
-1.48025334e-01 -5.58339596e-01 2.06971884e-01 -4.92754638e-01
6.57585800e-01 2.54633324e-03 -1.41231179e+00 4.89191592e-01
1.12669694e+00 1.30017981e-01 -1.54639766e-01 9.31191266e-01
-4.68348324e-01 7.48198926e-01 -6.10775590e-01 -4.98066097e-01
3.91159728e-02 3.10386135e-03 -1.02730155e+00 -3.89160633e-01
4.23657537e-01 2.18352620e-02 1.64395392e-01 1.44652271e+00
-1.01093039e-01 -2.49137908e-01 -3.05152987e-03 -1.33219731e+00
-1.28343248e+00 7.50947773e-01 -4.06176656e-01 1.91918835e-01
7.00367332e-01 2.73175299e-01 1.00676215e+00 -6.67732239e-01
3.19942921e-01 1.12541926e+00 9.39136624e-01 7.90050924e-01
-1.21322334e+00 -3.83737445e-01 3.28446269e-01 -8.22739452e-02
-1.04321897e+00 -1.62087277e-01 6.45159662e-01 -8.60566914e-01
1.88003206e+00 3.39086175e-01 2.75010258e-01 8.87278140e-01
8.73537958e-01 1.35023594e-01 2.98916221e-01 -6.87003434e-01
-1.63263492e-02 9.95120108e-02 9.50400293e-01 1.17003584e+00
6.77112520e-01 1.92557260e-01 -7.53000826e-02 -7.00820684e-01
1.35451987e-01 2.49647439e-01 -2.01804847e-01 1.39430712e-03
-6.97147310e-01 8.09039831e-01 4.91993994e-01 5.31964242e-01
6.49638772e-02 6.53391957e-01 6.38085425e-01 5.72248518e-01
2.31802985e-01 1.05471075e+00 -6.93980277e-01 -3.38290900e-01
-5.32764554e-01 2.99212396e-01 6.91746593e-01 1.02988935e+00
9.29740608e-01 2.88432956e-01 4.08439711e-02 4.65281963e-01
5.55707276e-01 1.20272674e-01 3.30352336e-01 -3.21290851e-01
8.86928201e-01 1.29851508e+00 -3.27201225e-02 -1.16522205e+00
-3.60409796e-01 -2.44192272e-01 -2.59102046e-01 2.81303495e-01
-1.94665313e-01 1.96522385e-01 -6.94649041e-01 1.59577882e+00
-1.84912086e-02 -4.39996161e-02 4.91740108e-02 2.72137403e-01
4.93755430e-01 6.03113949e-01 1.23157129e-01 2.23771125e-01
1.11099362e+00 -8.81567478e-01 -5.26788235e-01 -4.36240673e-01
1.32042944e+00 -5.55563986e-01 1.45859003e+00 3.54609489e-01
-1.07346535e+00 -5.78596532e-01 -1.08903444e+00 -9.35225114e-02
-4.83954310e-01 -8.79278556e-02 5.73689818e-01 1.15949976e+00
-1.08257306e+00 5.27259588e-01 -9.54083145e-01 -1.45228669e-01
1.21943876e-01 1.04770750e-01 -3.87862295e-01 1.82237566e-01
-8.99786949e-01 8.60454321e-01 3.38598132e-01 -2.46758446e-01
-1.02270389e+00 -1.07931340e+00 -1.31814265e+00 4.35523659e-01
3.80346864e-01 -5.09782910e-01 1.38258088e+00 -7.27763653e-01
-6.59287333e-01 5.07386208e-01 2.63425820e-02 -3.09118956e-01
-4.52534407e-01 -4.72876310e-01 -5.06368458e-01 -6.83256686e-01
-2.57254019e-02 -1.31443292e-01 7.05673575e-01 -1.07527483e+00
-3.94731641e-01 -1.92873597e-01 6.07453227e-01 -7.39848137e-01
-5.55231214e-01 3.63995135e-01 1.57289594e-01 -7.15707421e-01
-3.85146588e-01 -7.64176011e-01 -3.66401553e-01 -3.52933526e-01
-2.73185611e-01 1.02154026e-02 4.59700763e-01 -1.17846119e+00
1.98300016e+00 -2.26743531e+00 3.60831618e-01 1.61550969e-01
4.03669387e-01 3.53236884e-01 -4.00219798e-01 5.07537544e-01
-5.69342792e-01 9.39098895e-01 -2.97529757e-01 3.80981006e-02
7.53148645e-02 -1.16565667e-01 -7.44754374e-01 3.75841223e-02
5.96290469e-01 1.00864804e+00 -9.26679850e-01 1.21398084e-01
-1.97363809e-01 1.67269051e-01 -1.10265553e+00 3.85790706e-01
-5.95505595e-01 -1.19338877e-01 -4.28411007e-01 9.94442046e-01
5.46542823e-01 9.61255431e-02 -9.25447792e-02 1.97630674e-01
-2.34523401e-01 5.85014880e-01 -6.94709599e-01 1.55208230e+00
-8.79526138e-01 4.29563284e-01 -2.29414970e-01 -4.83382702e-01
1.00529385e+00 2.62163192e-01 -2.13277742e-01 -5.29886544e-01
-2.37222448e-01 5.05008325e-02 3.47367562e-02 -8.92466724e-01
6.85854614e-01 4.84353423e-01 -6.39541090e-01 6.65741563e-01
3.67221460e-02 4.53245528e-02 5.48603274e-02 1.77863687e-01
1.56973946e+00 2.14064956e-01 6.19199932e-01 -9.58507508e-02
5.19802272e-01 5.40133975e-02 5.62198102e-01 5.06206274e-01
1.54540434e-01 3.41884702e-01 9.57268476e-01 -7.74823785e-01
-9.93491828e-01 -7.86183834e-01 2.75163323e-01 1.08148313e+00
-4.13516909e-01 -1.24205613e+00 -1.21925008e+00 -1.15176070e+00
-1.74934059e-01 1.00964439e+00 -6.98511183e-01 -1.01537895e+00
-7.50363648e-01 -3.51749808e-01 8.43119144e-01 8.31866562e-01
-5.15941009e-02 -1.03441334e+00 -5.73986888e-01 1.39650851e-01
-4.36354503e-02 -2.17697263e-01 -6.45233512e-01 2.27335561e-02
-7.18587041e-01 -1.38999581e+00 2.03209773e-01 -4.90673006e-01
9.45781589e-01 2.74494231e-01 1.26814663e+00 9.18893337e-01
-6.34012163e-01 4.30229038e-01 -5.70151269e-01 -1.44218177e-01
-9.20864403e-01 8.94500911e-02 -1.64057538e-01 -6.75272465e-01
1.33909091e-01 -3.13199103e-01 1.53212532e-01 3.78687158e-02
-9.58935618e-01 -5.46720088e-01 2.77381182e-01 6.93957388e-01
7.23035187e-02 6.30644113e-02 2.53370672e-01 -1.01246667e+00
9.72811937e-01 -8.55014980e-01 -8.19000781e-01 7.70453513e-01
-7.33809829e-01 3.51364434e-01 6.16743684e-01 -1.76291913e-01
-9.69078958e-01 -5.43277226e-02 -4.85188782e-01 -2.94740289e-01
-1.69234537e-02 9.38998640e-01 -1.84297696e-01 -2.88468271e-01
1.34458661e+00 -9.88367274e-02 -2.65090704e-01 -4.30039823e-01
2.51420796e-01 3.94763559e-01 3.28308284e-01 -9.65197265e-01
9.71992493e-01 -5.64074814e-01 -5.32740951e-01 -2.99965143e-01
-4.26320732e-01 -8.92516151e-02 -3.69849801e-01 1.45367458e-01
5.46113431e-01 -5.44435859e-01 -4.00891632e-01 1.75548345e-01
-1.50369215e+00 -2.52600878e-01 -8.47572163e-02 -4.70560312e-01
-1.40569329e-01 4.87790346e-01 -2.32789591e-01 -7.06724942e-01
-5.71102262e-01 -1.46673167e+00 9.62421954e-01 1.00389294e-01
-7.40770876e-01 -1.07251620e+00 6.08585715e-01 -1.60425678e-01
6.67421460e-01 2.16994032e-01 1.62638807e+00 -2.32483521e-01
-8.20104122e-01 -2.65292048e-01 6.88906834e-02 3.95775765e-01
4.27766532e-01 4.66981083e-01 -7.43643165e-01 -3.42461854e-01
-8.38059932e-02 5.10070799e-03 3.67606014e-01 -1.93985924e-02
1.33674717e+00 -6.67148232e-01 -4.17182803e-01 5.53063393e-01
1.34927130e+00 1.51733264e-01 8.37735951e-01 3.11484367e-01
7.21829176e-01 6.75264657e-01 6.55348897e-01 2.95298338e-01
4.44592267e-01 6.02620125e-01 8.36588323e-01 4.32970285e-01
1.57462806e-01 -3.29976410e-01 7.73240983e-01 3.37697864e-01
3.31548721e-01 -4.59569916e-02 -1.56216502e+00 8.92362893e-01
-1.77683842e+00 -8.14669073e-01 -4.53744262e-01 2.24083495e+00
7.47211754e-01 1.31884009e-01 -4.62447852e-02 1.97474808e-01
5.02773821e-01 -2.95182675e-01 -1.97499588e-01 -9.73707139e-01
8.10584247e-01 2.83944577e-01 6.50356859e-02 6.13374293e-01
-8.70476127e-01 7.74621606e-01 6.55021524e+00 1.16717547e-01
-9.27685201e-01 2.51289666e-01 9.05863717e-02 1.91476718e-01
-8.80295157e-01 4.22905326e-01 -9.76424456e-01 3.61927748e-01
1.22291064e+00 -3.51743072e-01 5.43059707e-01 1.49146116e+00
-1.82926178e-01 2.94375867e-01 -1.26057959e+00 4.47706848e-01
5.73939295e-04 -1.41363859e+00 -1.31191894e-01 -8.83880630e-02
4.54309255e-01 -4.92608547e-01 2.08636746e-01 7.21116662e-01
3.58166397e-01 -1.01686084e+00 6.78688347e-01 7.40135014e-01
4.79480505e-01 -7.58150637e-01 6.48709238e-01 2.67177343e-01
-1.24540174e+00 -4.86226052e-01 -1.92733645e-01 -1.35944948e-01
-1.72901794e-01 5.80717504e-01 -7.52663076e-01 2.75376916e-01
8.80992293e-01 6.43413067e-01 -1.35877776e+00 8.77177656e-01
-3.53348076e-01 5.91662228e-01 3.60927910e-01 1.55036807e-01
-2.21221536e-01 4.66000855e-01 2.78621942e-01 1.30921602e+00
6.57884717e-01 -3.58938366e-01 1.30292848e-01 1.39195430e+00
2.79081792e-01 -3.90856147e-01 -9.56143618e-01 -4.38109338e-01
3.85679096e-01 1.09602821e+00 -2.24745482e-01 1.60511479e-01
-4.71955121e-01 3.35476696e-01 6.82163537e-01 5.55376988e-03
-1.03733242e+00 -6.63774252e-01 1.28844166e+00 2.93589115e-01
-1.52623355e-01 -3.78105313e-01 -6.29141152e-01 -1.23450458e+00
4.68817860e-01 -1.15219188e+00 3.51971895e-01 -6.21945024e-01
-9.81562138e-01 6.16142571e-01 7.49638379e-02 -1.06171310e+00
-3.44780356e-01 -4.76773560e-01 -1.16719484e+00 9.15171444e-01
-1.01651216e+00 -1.17447066e+00 -1.48106143e-01 1.27549469e-01
5.04736125e-01 -3.27243567e-01 1.02545500e+00 2.47188106e-01
-6.72548056e-01 1.00028431e+00 -6.31412446e-01 2.00948492e-02
2.92755008e-01 -1.26007116e+00 1.60664833e+00 1.49189365e+00
-2.99249236e-02 1.55853581e+00 4.89214599e-01 -1.08239555e+00
-1.59339976e+00 -1.44326055e+00 9.67916012e-01 -8.51284564e-01
8.47831428e-01 -5.32927454e-01 -1.57981145e+00 1.02017879e+00
7.54103512e-02 -1.32497519e-01 6.83955729e-01 3.58394653e-01
-1.04780090e+00 2.40531504e-01 -1.25758910e+00 5.59250832e-01
8.83781433e-01 -7.70132303e-01 -6.35619998e-01 1.01713605e-01
1.20592153e+00 -5.10726869e-01 -7.60378778e-01 1.38546433e-02
1.00845948e-01 -7.27048814e-01 8.98304462e-01 -8.05735946e-01
6.99897945e-01 -4.82991368e-01 -2.87587848e-02 -1.45149660e+00
-4.96305168e-01 -5.89190781e-01 -3.56219858e-01 1.49652421e+00
5.78481436e-01 -6.21097922e-01 3.51122051e-01 8.42748165e-01
-5.55445075e-01 -4.08122033e-01 -4.19925511e-01 -5.66637516e-01
6.80017471e-02 -6.71362042e-01 9.13271308e-01 9.51024771e-01
4.87999290e-01 -1.42691478e-01 -1.46321207e-01 5.91978669e-01
6.29860535e-02 -1.46061271e-01 7.27198362e-01 -1.12624931e+00
-6.17725790e-01 -4.27869260e-01 -3.67094368e-01 -1.71668634e-01
4.31852758e-01 -1.02198017e+00 4.57544550e-02 -1.17302740e+00
1.34070758e-02 -3.17807287e-01 4.95904172e-03 1.16083717e+00
-4.67960268e-01 -4.38407958e-01 -1.16736121e-01 -2.66929030e-01
-2.68570811e-01 1.20905273e-01 2.70978421e-01 -4.82412279e-01
-1.44716546e-01 -5.71958013e-02 -1.14336610e+00 5.09086549e-01
9.56419826e-01 -8.60119939e-01 -3.98375422e-01 -6.92128956e-01
9.68658924e-01 9.55633521e-02 5.25549829e-01 -7.84148335e-01
-7.18853548e-02 -4.32440728e-01 -4.46716547e-01 5.92434555e-02
-3.20371836e-01 -6.02047026e-01 2.94665843e-01 7.92690158e-01
-3.64908695e-01 6.43865943e-01 6.43676877e-01 2.89027393e-01
-1.50885344e-01 -9.42820013e-01 5.51244020e-01 -1.69116691e-01
-6.85857773e-01 4.06699674e-03 -5.16099930e-01 -2.11346745e-01
1.16995347e+00 6.06799088e-02 -8.29264522e-01 3.23581755e-01
-5.28088510e-01 -3.05045638e-02 6.58389509e-01 1.03112650e+00
7.99740613e-01 -1.20268857e+00 -3.23018730e-01 6.20046496e-01
5.39167225e-01 -3.19592893e-01 1.65774867e-01 3.46637636e-01
-5.28854966e-01 3.12565178e-01 -4.02955264e-01 -2.48929113e-01
-1.50765896e+00 9.44761276e-01 3.83325219e-01 -2.91055173e-01
-3.53471339e-01 7.11024106e-01 4.77961749e-02 -9.44190800e-01
4.99631837e-02 -8.07120800e-01 1.80708840e-02 -5.76254368e-01
8.45836043e-01 2.89480835e-01 5.42838991e-01 3.64416167e-02
-3.84713084e-01 3.90690625e-01 -4.61319685e-01 6.43425822e-01
1.28922522e+00 4.41598862e-01 -6.34702981e-01 1.33139893e-01
9.62361455e-01 -3.00616268e-02 -6.54948831e-01 3.75308216e-01
4.62590516e-01 -8.56510937e-01 -1.65694177e-01 -8.32713485e-01
-8.65591407e-01 1.02794957e+00 4.35931414e-01 3.99968952e-01
9.09060657e-01 -8.88104960e-02 5.20271242e-01 7.01673567e-01
7.13491440e-01 -3.27876866e-01 1.76446378e-01 6.32767916e-01
1.05859447e+00 -9.42554414e-01 -3.49484473e-01 -5.12878478e-01
-1.04922883e-01 1.25064814e+00 1.11901510e+00 -1.80931225e-01
4.19780999e-01 6.96725607e-01 -4.52439040e-01 -2.23744988e-01
-9.82935071e-01 3.19884777e-01 2.26080909e-01 1.01661861e+00
8.52680802e-01 -1.59839541e-01 -5.26728965e-02 9.76789296e-01
-8.74201879e-02 -9.24070925e-02 8.75149190e-01 1.26446438e+00
-4.15257812e-01 -1.57551277e+00 -6.44738674e-01 2.26285368e-01
-2.55372941e-01 -3.80086720e-01 -6.27677917e-01 4.72169191e-01
1.45273786e-02 9.15246725e-01 -3.74000251e-01 -7.56517649e-01
6.16413236e-01 1.59100011e-01 4.34105188e-01 -1.29228866e+00
-1.31285763e+00 -7.93501437e-01 8.16416815e-02 -7.41314828e-01
5.38092554e-01 -4.46993828e-01 -1.15271723e+00 -3.22013795e-01
-3.17197204e-01 -3.96523479e-04 6.04980886e-01 6.03797317e-01
8.49193931e-01 8.89978051e-01 4.27241325e-01 -4.89554524e-01
-5.72569311e-01 -5.61575532e-01 2.00489879e-01 1.14390098e-01
5.29541612e-01 -5.52500010e-01 -1.00935534e-01 2.28832915e-01] | [7.520407676696777, 7.7369065284729] |
a5c6b02c-1663-46ad-99a6-e67917067d8e | optic-net-a-novel-convolutional-neural | 1910.05672 | null | https://arxiv.org/abs/1910.05672v1 | https://arxiv.org/pdf/1910.05672v1.pdf | Optic-Net: A Novel Convolutional Neural Network for Diagnosis of Retinal Diseases from Optical Tomography Images | Diagnosing different retinal diseases from Spectral Domain Optical Coherence Tomography (SD-OCT) images is a challenging task. Different automated approaches such as image processing, machine learning and deep learning algorithms have been used for early detection and diagnosis of retinal diseases. Unfortunately, these are prone to error and computational inefficiency, which requires further intervention from human experts. In this paper, we propose a novel convolution neural network architecture to successfully distinguish between different degeneration of retinal layers and their underlying causes. The proposed novel architecture outperforms other classification models while addressing the issue of gradient explosion. Our approach reaches near perfect accuracy of 99.8% and 100% for two separately available Retinal SD-OCT data-set respectively. Additionally, our architecture predicts retinal diseases in real time while outperforming human diagnosticians. | ['Ali Shihab Sabbir', 'Sharif Amit Kamran', 'Alireza Tavakkoli', 'Sourajit Saha'] | 2019-10-13 | null | null | null | null | ['retinal-oct-disease-classification'] | ['computer-vision'] | [ 1.25759661e-01 -1.77007571e-01 3.64835054e-01 -2.71602273e-01
-3.87580663e-01 -2.95693308e-01 -5.77366240e-02 7.63750216e-03
-4.63786572e-01 8.97935212e-01 -9.74946246e-02 -6.43756449e-01
-3.88653517e-01 -5.36249518e-01 -1.00074276e-01 -6.00647688e-01
-2.98276190e-02 3.24595064e-01 2.70559937e-01 3.91829699e-01
5.23343265e-01 7.94569254e-01 -1.63259041e+00 4.40416306e-01
1.30557930e+00 1.22838593e+00 1.39983565e-01 1.18760538e+00
1.35203257e-01 9.28645194e-01 -9.31596309e-02 -1.44102834e-02
6.24351144e-01 -3.71682465e-01 -9.45231259e-01 2.81653225e-01
9.85832155e-01 -7.11313188e-01 -1.07132956e-01 1.03997517e+00
9.45472360e-01 -4.59218472e-01 5.78081071e-01 -4.97514099e-01
-5.29639781e-01 -2.97905326e-01 -6.29899919e-01 7.13254273e-01
-3.18148375e-01 5.79003990e-01 2.67631471e-01 -5.17141581e-01
2.68693238e-01 7.52011418e-01 7.29296446e-01 3.89483631e-01
-9.26477432e-01 -3.89737248e-01 -5.70858538e-01 3.71177614e-01
-9.15037692e-01 -4.51159179e-01 1.22733086e-01 -9.97678578e-01
1.00257349e+00 7.30928630e-02 8.86042237e-01 2.85230070e-01
1.57417864e-01 2.03268886e-01 1.69638932e+00 -3.75936300e-01
-1.34942338e-01 4.01519053e-02 3.44069362e-01 9.71144140e-01
5.68650663e-01 4.15581286e-01 1.55474737e-01 -2.03661118e-02
1.01701844e+00 -1.92195848e-01 -3.74523491e-01 1.85494184e-01
-8.93534720e-01 4.63042855e-01 4.94950712e-01 -5.64487949e-02
-7.04791129e-01 -8.57368931e-02 3.57899070e-01 2.43053138e-01
3.48942161e-01 7.12528050e-01 -5.11152804e-01 4.52915989e-02
-8.49225819e-01 -4.86477204e-02 2.45340288e-01 2.93361753e-01
2.34465033e-01 -2.91169658e-02 -2.68103391e-01 9.00314629e-01
2.51452655e-01 2.25911349e-01 5.64048886e-01 -8.73970568e-01
-7.81853944e-02 8.26684058e-01 4.02143747e-01 -4.95412529e-01
-6.84107304e-01 -7.50919819e-01 -9.96269584e-01 7.74788141e-01
4.59920317e-01 -5.43161094e-01 -1.44730020e+00 6.08132839e-01
1.05113223e-01 4.55992579e-01 -1.05076551e-01 1.26213121e+00
8.53426814e-01 1.46067925e-02 7.52458395e-03 -2.80773878e-01
1.47880244e+00 -7.56678224e-01 -2.23331079e-01 -1.94445789e-01
4.13549751e-01 -1.12200451e+00 5.09582222e-01 6.58983767e-01
-1.25722063e+00 -3.06659639e-01 -8.28730702e-01 -2.83879250e-01
2.53086928e-02 1.13473177e+00 8.47678483e-01 5.03897905e-01
-1.36680770e+00 4.97060180e-01 -8.96556139e-01 -6.30815148e-01
1.09317005e+00 7.44352758e-01 -2.45188087e-01 -5.96481562e-02
-2.81209975e-01 9.27554131e-01 2.03217566e-01 1.62500918e-01
-4.49759096e-01 -6.72201514e-01 -2.16690704e-01 -2.85061896e-01
-1.50431067e-01 -1.33310544e+00 1.16646767e+00 -7.84869373e-01
-1.51597881e+00 1.37101197e+00 -3.90831918e-01 -9.08348918e-01
4.66661423e-01 -2.57612646e-01 -3.80829901e-01 4.04696465e-01
-3.47489327e-01 4.86169040e-01 8.80958796e-01 -5.91314018e-01
-1.09232020e+00 -6.25139952e-01 -1.24628372e-01 -1.02904283e-01
1.66818365e-01 2.79647350e-01 7.26844892e-02 -1.90325622e-02
2.93383867e-01 -8.30976725e-01 -3.42727333e-01 4.49214578e-01
-6.66901648e-01 -8.62215534e-02 3.81533831e-01 -6.69103384e-01
8.77002716e-01 -1.96402550e+00 -4.93767619e-01 -1.75422892e-01
8.31657946e-01 1.12220812e+00 -3.17045115e-02 -1.89370632e-01
-2.28649661e-01 2.18960401e-02 2.00622097e-01 1.97202235e-01
-7.85561085e-01 -1.89854220e-01 5.30475751e-02 5.67151487e-01
6.35006726e-01 6.82398558e-01 -5.86002290e-01 -2.36849979e-01
3.89785379e-01 5.26748240e-01 -4.78074402e-01 1.07536443e-01
4.21610624e-02 5.12547851e-01 -1.95748135e-01 8.56758833e-01
7.54685938e-01 -7.81909287e-01 -5.71206920e-02 -4.17451233e-01
-3.34176928e-01 2.87634164e-01 -8.41259956e-01 8.73084664e-01
-1.77045077e-01 1.14858854e+00 -4.33646113e-01 -9.08812463e-01
7.25725591e-01 2.06812993e-01 1.52508035e-01 -5.56237221e-01
3.46964687e-01 5.29352486e-01 7.81529605e-01 -1.06820226e+00
2.18254211e-03 -9.62572917e-02 1.24843264e+00 2.08472461e-01
-1.73471048e-01 5.89832127e-01 2.57283803e-02 -4.99366820e-01
1.14949048e+00 -2.79285163e-01 6.06415212e-01 6.55597597e-02
5.03490329e-01 1.21888191e-01 3.48770678e-01 5.10293484e-01
-5.63178778e-01 6.47859871e-01 5.63764095e-01 -9.31769967e-01
-1.15028548e+00 -8.28891695e-01 -4.76458549e-01 -7.42213130e-02
-1.20420910e-01 2.27101535e-01 -4.37674850e-01 -3.02201390e-01
5.59282303e-03 -1.91278294e-01 -5.71438193e-01 2.30590060e-01
-3.57821196e-01 -1.08863699e+00 2.66329974e-01 3.19171339e-01
8.91199708e-01 -5.13851345e-01 -9.47223485e-01 2.70263702e-01
2.51828611e-01 -1.12665558e+00 3.12013894e-01 -4.58708584e-01
-1.27871490e+00 -1.63409638e+00 -7.22770274e-01 -9.40093219e-01
7.71063924e-01 3.23408455e-01 9.89607692e-01 9.80634168e-02
-1.29280961e+00 -1.29314721e-01 1.24540068e-01 -6.16167605e-01
-1.28924564e-01 -3.82356793e-01 -1.51856914e-01 3.14720035e-01
6.70796156e-01 -7.03975379e-01 -1.43225229e+00 -1.57499224e-01
-8.91527608e-02 3.50545458e-02 1.23222280e+00 8.18814814e-01
6.45068884e-01 1.03584938e-02 3.05865973e-01 -8.29402685e-01
6.57803059e-01 -2.86687344e-01 -9.39453542e-01 -1.87195931e-02
-9.22246158e-01 -1.99914306e-01 8.68007019e-02 -1.97510689e-01
-6.67836666e-01 2.10593473e-02 1.96248919e-01 -3.46207023e-01
-6.60881758e-01 2.84396410e-01 7.25607097e-01 -5.75214624e-01
9.47900414e-01 1.22512914e-01 2.85635531e-01 -6.65727019e-01
-2.13441372e-01 8.72468710e-01 4.37900513e-01 2.89566606e-01
2.56296068e-01 6.43703341e-01 4.05905902e-01 -7.59746194e-01
-9.17166114e-01 -6.07493043e-01 -4.35023338e-01 -7.40033910e-02
8.57025921e-01 -9.72220361e-01 -7.70109653e-01 1.00106931e+00
-1.29668581e+00 1.62398130e-01 1.39353007e-01 8.82483304e-01
-1.24091662e-01 2.88607210e-01 -5.55913985e-01 -7.46493697e-01
-5.84966063e-01 -1.00462604e+00 7.35925734e-01 5.36607563e-01
2.05037072e-01 -6.98818803e-01 3.72674689e-03 3.59302342e-01
5.58784544e-01 3.11199456e-01 1.26156831e+00 -1.39922276e-01
-8.54500473e-01 -2.25694627e-01 -1.13496447e+00 6.90216482e-01
3.02196711e-01 4.63712275e-01 -1.00886703e+00 -8.90107602e-02
-3.67304564e-01 -1.59768641e-01 9.97126818e-01 1.07957768e+00
1.04445875e+00 -2.53027380e-01 -3.51372600e-01 7.38275290e-01
1.86955690e+00 3.43734801e-01 1.03329349e+00 2.80612648e-01
2.42767781e-01 4.41975057e-01 1.78582117e-01 4.81993288e-01
1.38919353e-01 8.73986483e-02 4.18401897e-01 -4.56457138e-01
-5.29648721e-01 5.34072459e-01 -4.11330283e-01 2.07536638e-01
-7.14402735e-01 3.94086763e-02 -1.20418441e+00 7.98843145e-01
-1.55918860e+00 -8.27912569e-01 -6.31598890e-01 2.13625360e+00
6.03624523e-01 -7.24311918e-02 1.43561035e-01 -2.52470877e-02
6.35102928e-01 -7.41809070e-01 -7.48720169e-01 -2.06834555e-01
-1.91437125e-01 7.03190088e-01 6.56801403e-01 1.53069869e-01
-1.14709163e+00 6.72026515e-01 6.60038662e+00 -8.35143402e-03
-1.61497140e+00 -2.30403654e-02 6.62091553e-01 -3.15943658e-01
6.03276730e-01 -1.93524480e-01 -4.68870699e-01 4.20897454e-01
6.14962161e-01 5.99458627e-02 1.60264716e-01 1.84598386e-01
5.59239030e-01 -4.29454029e-01 -6.66949868e-01 1.35159481e+00
-2.31331900e-01 -1.64791083e+00 4.15957421e-02 1.93037495e-01
6.26962543e-01 5.43300331e-01 2.56699800e-01 -4.81502175e-01
1.45432547e-01 -1.39239049e+00 -4.87643719e-01 9.60570037e-01
9.71923053e-01 -4.63802993e-01 9.76881444e-01 -2.29507193e-01
-4.05974537e-01 -3.28420430e-01 -3.37143034e-01 -4.34985757e-01
-3.75145912e-01 7.40700722e-01 -1.19670272e+00 7.75431395e-02
7.80586123e-01 9.23497140e-01 -8.03762078e-01 2.05662179e+00
7.95242004e-03 6.93625152e-01 -8.82244762e-03 2.25549638e-01
2.82073826e-01 -1.80521086e-01 6.13916516e-01 6.11205459e-01
4.79494065e-01 2.74936736e-01 -3.26760173e-01 7.88786352e-01
3.26944858e-01 1.00902565e-01 -4.84903932e-01 -1.88608065e-01
2.11129293e-01 1.21727133e+00 -3.03860366e-01 -1.29929125e-01
-4.35257941e-01 5.21695077e-01 2.28407338e-01 4.77512836e-01
-2.77473658e-01 -4.17783588e-01 9.98108506e-01 5.35377443e-01
7.51024410e-02 -1.07842721e-01 -5.12163043e-01 -7.92709470e-01
4.80860099e-02 -5.77906013e-01 1.32930443e-01 -1.08544707e+00
-1.19108248e+00 8.95186126e-01 -9.44140911e-01 -1.59546983e+00
3.97704691e-02 -1.12687886e+00 -5.85753620e-01 1.18596780e+00
-2.16314983e+00 -1.00905323e+00 -6.20223880e-01 6.92656279e-01
8.36886540e-02 -6.78354859e-01 5.65331697e-01 4.30537462e-01
-6.99239075e-01 2.17916816e-01 -1.88033842e-02 2.04157427e-01
8.07052910e-01 -1.21355009e+00 -2.51030698e-02 9.22514975e-01
-5.43646812e-01 4.76551175e-01 3.52344573e-01 -3.24723274e-01
-8.33970010e-01 -1.25513661e+00 9.03662503e-01 4.20692898e-02
4.85049546e-01 7.76417136e-01 -6.11523211e-01 3.95024419e-01
3.20880532e-01 3.62256169e-01 7.01729417e-01 3.92052578e-03
-1.08704329e-01 -2.14484453e-01 -1.13819647e+00 4.84711677e-01
7.48611510e-01 -3.72774422e-01 -3.16604465e-01 7.36109793e-01
-9.46334377e-03 -3.63847286e-01 -6.83483183e-01 5.16807497e-01
5.12937903e-01 -1.45584130e+00 7.86772966e-01 -1.10457289e+00
6.66793108e-01 -3.09931010e-01 4.28358167e-01 -9.78900075e-01
-3.16363811e-01 -6.76265776e-01 1.35059521e-01 4.66983318e-01
2.59602934e-01 -1.16529191e+00 7.29547799e-01 3.26183587e-01
1.69971902e-02 -9.71345067e-01 -5.88623226e-01 -3.81590754e-01
-2.18444794e-01 1.29104676e-02 1.15188822e-01 7.93006361e-01
-7.17601359e-01 1.36052608e-01 -1.61855608e-01 6.85647607e-01
7.08239079e-01 2.67182648e-01 4.30411547e-01 -1.56669247e+00
-4.72631771e-04 -5.91640472e-01 -1.21289933e+00 -5.54487586e-01
-4.61649835e-01 -6.08830571e-01 -6.51692033e-01 -1.79568028e+00
1.49720818e-01 -4.23289448e-01 -2.90481269e-01 4.16681200e-01
9.77105182e-03 5.97785532e-01 -4.30927902e-01 2.24830881e-01
9.34033915e-02 -1.85435385e-01 1.47536552e+00 1.56059731e-02
-2.13065118e-01 1.34429321e-01 -6.67184949e-01 9.59073365e-01
1.12854505e+00 -1.63118586e-01 -4.08193141e-01 -7.16633558e-01
2.84362584e-01 -2.35177577e-02 1.04904747e+00 -1.22662926e+00
4.26052749e-01 3.66218723e-02 3.84339839e-01 -4.30422664e-01
1.43338945e-02 -2.86053658e-01 -2.23122820e-01 6.54763281e-01
7.43612200e-02 -4.77306992e-01 3.55113775e-01 3.46305490e-01
-4.47506666e-01 2.00966910e-01 1.24640834e+00 -2.78693765e-01
-7.57244647e-01 5.73852599e-01 -3.87526035e-01 -2.30804719e-02
9.66773272e-01 -5.07821798e-01 -7.81122983e-01 1.74145639e-01
-9.55256999e-01 -2.22992264e-02 1.70105338e-01 -1.28227592e-01
9.32610691e-01 -8.05119574e-01 -1.12440479e+00 2.86352307e-01
9.76076722e-02 -1.82840854e-01 4.91680205e-01 1.54140306e+00
-1.13991368e+00 6.25664890e-01 -6.55583918e-01 -7.53542125e-01
-1.64140797e+00 -7.77517585e-03 1.03282213e+00 1.31116837e-01
-6.91492379e-01 1.01515663e+00 -1.50491133e-01 4.04865324e-01
-4.87248003e-02 -6.19349897e-01 -3.77428114e-01 -1.88222885e-01
6.30241632e-01 5.93267083e-01 2.14638934e-01 7.58344382e-02
-6.63704053e-02 9.20089602e-01 -5.13046443e-01 5.23204982e-01
1.20865941e+00 -1.81387380e-01 -5.61928034e-01 -8.19792077e-02
8.05235624e-01 -3.49347860e-01 -8.59664023e-01 -2.87369132e-01
-8.35794657e-02 -7.37259209e-01 6.15049124e-01 -1.39271021e+00
-1.04323971e+00 1.20151293e+00 1.43458140e+00 4.50665429e-02
1.53993809e+00 -3.54258835e-01 8.22448969e-01 3.11203212e-01
1.14469886e-01 -6.05410993e-01 -3.82043064e-01 1.29098266e-01
5.22334814e-01 -1.66653407e+00 -3.23420465e-02 -6.07916951e-01
-2.36855090e-01 1.31725383e+00 6.92244768e-01 -4.26212728e-01
5.89996934e-01 -1.10805921e-01 4.39761311e-01 -3.40502143e-01
-8.47155988e-01 -7.12323904e-01 5.95796824e-01 8.05643678e-01
5.15284777e-01 -9.09007248e-03 -3.45148563e-01 1.86113462e-01
1.25724614e-01 6.97717190e-01 8.68195474e-01 5.07258534e-01
-5.76964021e-01 -8.02265346e-01 1.10698454e-01 1.13747728e+00
-6.76726043e-01 -4.64652956e-01 -4.14052248e-01 6.09561145e-01
4.47956800e-01 7.33051300e-01 2.78215379e-01 -5.22289565e-03
2.53772568e-02 -2.30104610e-01 7.47070491e-01 -5.96258581e-01
-3.06080371e-01 8.06960538e-02 2.42669582e-01 -5.53074718e-01
-5.74282229e-01 -3.85237247e-01 -8.09346437e-01 -8.23648348e-02
-1.82254240e-03 -5.47302961e-01 3.37670594e-01 7.60197818e-01
9.90630329e-01 4.55232471e-01 3.92073303e-01 -1.93943143e-01
-3.56600940e-01 -1.06183827e+00 -8.07484567e-01 -1.34613633e-01
9.62767184e-01 -5.20256698e-01 -3.43910940e-02 2.51654088e-01] | [15.82298755645752, -3.994412422180176] |
b0053006-f6f5-483e-b0d8-3eebeaba30d9 | from-time-series-to-networks-in-r-with-the | 2208.09660 | null | https://arxiv.org/abs/2208.09660v1 | https://arxiv.org/pdf/2208.09660v1.pdf | From Time Series to Networks in R with the ts2net Package | Network science established itself as a prominent tool for modeling time series and complex systems. This modeling process consists of transforming a set or a single time series into a network. Nodes may represent complete time series, segments, or single values, while links define associations or similarities between the represented parts. R is one of the main programming languages used in data science, statistics, and machine learning, with many packages available. However, no single package provides the necessary methods to transform time series into networks. This paper presents ts2net, an R package for modeling one or multiple time series into networks. The package provides the time series distance functions that can be easily computed in parallel and in supercomputers to process larger data sets and methods to transform distance matrices into networks. Ts2net also provides methods to transform a single time series into a network, such as recurrence networks, visibility graphs, and transition networks. Together with other packages, ts2net permits using network science and graph mining tools to extract information from time series. | ['Leonardo N. Ferreira'] | 2022-08-20 | null | null | null | null | ['graph-mining'] | ['graphs'] | [-7.36498907e-02 -2.54701644e-01 -2.17805281e-01 -1.50788084e-01
2.92696685e-01 -8.94470215e-01 5.33305824e-01 6.15130246e-01
7.70522431e-02 2.48370647e-01 -2.02885509e-01 -8.46286595e-01
-7.95095742e-01 -1.35951817e+00 8.91438946e-02 -3.44375044e-01
-9.93991613e-01 3.96851331e-01 3.08843613e-01 -3.73563230e-01
7.98268020e-02 1.01908529e+00 -1.30254507e+00 -2.66236424e-01
5.63342035e-01 8.00502777e-01 -2.28315294e-01 7.29195416e-01
-3.44002545e-01 6.90403342e-01 -5.93325317e-01 6.10831231e-02
1.92925319e-01 -4.90912616e-01 -3.57880861e-01 -4.59045291e-01
-5.53556502e-01 2.41486937e-01 -5.31722426e-01 7.60903537e-01
2.26602450e-01 1.49065673e-01 5.25459409e-01 -1.90449381e+00
-2.27931395e-01 8.15496206e-01 -3.78139824e-01 4.77507085e-01
6.69320703e-01 -2.27036431e-01 6.07947350e-01 -4.47895169e-01
8.11378300e-01 1.16482365e+00 8.80969882e-01 -3.93310755e-01
-1.31327498e+00 -4.41118568e-01 -3.36035937e-01 1.68110803e-01
-1.44280946e+00 1.39744163e-01 9.68017161e-01 -5.30619800e-01
1.12558508e+00 8.15967083e-01 1.21254361e+00 2.92607427e-01
3.39885384e-01 2.78491192e-02 1.05358052e+00 -2.57399589e-01
1.26097649e-01 -2.67048329e-01 5.44310749e-01 4.22503203e-01
1.22303858e-01 -1.73560698e-02 -8.84148404e-02 -5.09301364e-01
1.07757068e+00 5.16480207e-01 1.54160455e-01 -2.64115457e-04
-1.35957837e+00 7.47488022e-01 3.69613379e-01 8.30460727e-01
-3.70092332e-01 2.28836481e-02 6.72993004e-01 1.02681887e+00
6.61559463e-01 4.58831221e-01 -1.95977464e-01 -3.30071986e-01
-7.59462059e-01 1.15436371e-02 1.03035545e+00 7.47087300e-01
7.96102226e-01 1.94533125e-01 3.71628255e-01 5.69881320e-01
1.62661672e-02 3.03567499e-01 2.78972685e-01 -8.17558050e-01
1.34050459e-01 9.56351697e-01 -5.55096984e-01 -1.72180367e+00
-8.83433104e-01 -3.44946474e-01 -1.07521319e+00 -7.32496008e-02
4.33773220e-01 -1.83736891e-01 -4.52265441e-01 1.34137011e+00
3.63292873e-01 6.64213076e-02 -3.18304628e-01 3.82340133e-01
7.44542062e-01 9.36404765e-01 -3.57765824e-01 -6.99024498e-01
1.09878540e+00 -3.41832429e-01 -9.07986820e-01 5.19927919e-01
8.83660018e-01 -7.28996634e-01 3.88175786e-01 3.27547014e-01
-1.01113522e+00 -3.32924038e-01 -6.99121416e-01 1.89288780e-01
-9.26987529e-01 -5.32687366e-01 8.01923037e-01 2.95372009e-01
-1.40363109e+00 1.16101062e+00 -1.00451040e+00 -6.65979624e-01
-1.34351179e-01 2.43207991e-01 -4.06751335e-01 1.43424958e-01
-1.34500611e+00 9.57650781e-01 2.54679471e-01 5.67117929e-02
-1.49494261e-01 -7.43553042e-01 -1.00474691e+00 -1.00614481e-01
9.59874094e-02 -4.14128959e-01 5.04394710e-01 -5.54937661e-01
-1.02702022e+00 5.72530150e-01 -9.03025568e-02 -4.48800325e-01
3.88136268e-01 6.08638346e-01 -1.02802992e+00 2.35088822e-02
2.29329709e-02 -1.69457301e-01 4.97981369e-01 -4.21202809e-01
4.20221500e-02 -4.07135218e-01 -3.09287816e-01 -2.31220379e-01
-2.99787313e-01 5.46191156e-01 -1.74811691e-01 -7.53225625e-01
4.24298793e-01 -6.62269771e-01 -4.43286061e-01 -7.27192312e-02
-4.75112021e-01 -6.38243139e-01 9.17297363e-01 -6.82277203e-01
1.68870819e+00 -2.05035496e+00 2.61930615e-01 1.08234501e+00
5.39572418e-01 -3.68118078e-01 -4.50587049e-02 1.25024855e+00
-6.54803216e-01 2.41261065e-01 -1.01904653e-01 2.09003225e-01
-7.09006712e-02 1.91863164e-01 4.53547612e-02 6.07509017e-01
-2.15644613e-01 6.00775361e-01 -9.85950947e-01 -2.31062248e-01
2.91774631e-01 3.54933828e-01 1.31107941e-01 -1.51365712e-01
1.92579061e-01 2.29704902e-01 -1.75996885e-01 4.04611349e-01
4.10409808e-01 -1.88791409e-01 4.49345648e-01 -4.71356846e-02
-6.30013525e-01 2.49623895e-01 -1.18758214e+00 1.21240652e+00
-7.82912299e-02 9.69751894e-01 -2.33038589e-01 -1.18263805e+00
1.57338858e+00 4.01830286e-01 1.01396739e+00 -4.33421880e-01
1.47997379e-01 1.15015343e-01 1.91371933e-01 -5.13613045e-01
2.15303153e-01 4.11277294e-01 -1.05364472e-02 8.42142105e-01
-2.86960244e-01 -4.21112627e-01 7.96012521e-01 2.23001748e-01
1.57200718e+00 -5.46547413e-01 3.80002499e-01 -2.31575832e-01
2.41519168e-01 1.84823871e-01 3.91223073e-01 1.63433507e-01
2.76615530e-01 3.40471119e-02 1.09300220e+00 -7.03062952e-01
-1.37038422e+00 -1.12465739e+00 1.62240118e-02 8.24644804e-01
-3.57968360e-01 -9.40396369e-01 -1.13986433e-01 1.01422846e-01
1.83593154e-01 4.01335627e-01 -5.47909260e-01 9.85439047e-02
-4.20110613e-01 -2.89765567e-01 3.99892807e-01 2.65542895e-01
-7.66418204e-02 -1.02491391e+00 5.59997149e-02 3.33751470e-01
1.86220586e-01 -6.10451102e-01 -4.14170474e-01 2.67018527e-02
-1.31008720e+00 -1.20727074e+00 -2.56676883e-01 -8.14971209e-01
8.92255545e-01 4.03407544e-01 9.29065228e-01 2.22514957e-01
-4.94874954e-01 2.50426024e-01 -3.81316334e-01 -2.30992720e-01
-3.64318222e-01 -1.41194269e-01 1.95059210e-01 -2.00761750e-01
1.55953735e-01 -1.23731148e+00 -1.90039486e-01 4.77312565e-01
-1.12309945e+00 -2.16518074e-01 -1.52517542e-01 1.66893557e-01
4.54108179e-01 6.81708336e-01 5.70223868e-01 -5.34300447e-01
1.10621679e+00 -9.44376707e-01 -6.78520560e-01 1.54458076e-01
-6.66350782e-01 -4.77461845e-01 7.23159432e-01 -4.37292099e-01
5.95011078e-02 -1.45954743e-01 3.83425951e-01 -4.56632286e-01
2.91606277e-01 1.37954485e+00 3.96488279e-01 3.66868638e-02
5.11383951e-01 2.69334406e-01 5.51120043e-01 -3.53272378e-01
2.85914272e-01 3.90258431e-01 2.59299725e-01 -1.82935283e-01
8.74371350e-01 3.27812761e-01 3.25632542e-01 -1.11261034e+00
1.36572629e-01 -5.75322747e-01 -9.14309263e-01 -7.02652037e-01
4.50176477e-01 -4.55242544e-01 -6.67587221e-01 3.82478356e-01
-9.37450409e-01 -8.26151222e-02 -2.99073637e-01 4.88918483e-01
-1.87810123e-01 1.65244877e-01 -6.72363043e-01 -6.72669053e-01
-1.92861453e-01 -3.32994342e-01 9.99081135e-02 1.19860187e-01
-6.76527500e-01 -1.53061807e+00 3.94998759e-01 -3.23651344e-01
5.65721095e-01 9.80066240e-01 8.85127544e-01 -8.36098850e-01
-3.42805199e-02 -8.08274806e-01 3.06298248e-02 -1.00724541e-01
4.14345682e-01 9.79863763e-01 8.98706634e-03 -3.11708510e-01
-1.89127445e-01 5.91873705e-01 1.30892828e-01 4.46744144e-01
1.16393244e+00 -3.87520045e-01 -4.89798784e-01 5.38984656e-01
1.11357248e+00 7.12162614e-01 7.77448237e-01 1.64504707e-01
5.54249585e-01 9.81797218e-01 2.85427809e-01 4.86511528e-01
4.52615142e-01 3.10589761e-01 1.08117647e-01 -2.07085252e-01
4.59761083e-01 -8.28300044e-02 2.32971892e-01 1.42654848e+00
-4.21532452e-01 2.66663522e-01 -1.15326023e+00 4.87085938e-01
-1.84908223e+00 -1.37348664e+00 -8.71784508e-01 2.07954812e+00
4.70265448e-01 1.51859894e-01 7.14030802e-01 5.52262723e-01
1.04243815e+00 8.87824744e-02 -4.79526103e-01 -4.69886929e-01
-1.19638100e-01 1.73711568e-01 3.25089157e-01 2.54646748e-01
-5.73947370e-01 4.31534708e-01 7.21918106e+00 4.41623360e-01
-1.20204675e+00 -2.23081335e-01 2.48892114e-01 1.20566338e-01
-5.01163721e-01 1.89238057e-01 1.50478274e-01 2.73988187e-01
1.19083953e+00 -1.14710498e+00 5.96168339e-01 6.24517918e-01
8.17545652e-01 1.89774141e-01 -9.44802463e-01 9.67584670e-01
-6.01849079e-01 -1.40095985e+00 -3.38936269e-01 1.07433371e-01
3.95336330e-01 -7.72163330e-04 -3.58285069e-01 -1.02183208e-01
3.51268828e-01 -1.04760528e+00 2.22890452e-01 7.35920250e-01
7.48891532e-01 -7.67002583e-01 4.26477015e-01 -3.17906365e-02
-1.83476663e+00 1.77137986e-01 -1.20705552e-01 -5.72326541e-01
2.90576398e-01 9.94485199e-01 -7.24402428e-01 9.55381274e-01
6.35853291e-01 1.38981533e+00 -5.02418339e-01 1.32890952e+00
1.85454279e-01 4.68771517e-01 -6.27827048e-01 -2.58208752e-01
-2.96551287e-01 -8.25608552e-01 6.36974514e-01 8.17976654e-01
6.68978333e-01 -1.78228486e-02 1.12865441e-01 7.10329831e-01
5.85227251e-01 1.26469240e-01 -1.11186349e+00 -5.18903255e-01
1.00187159e+00 1.43669856e+00 -1.18526518e+00 -3.32942992e-01
-2.77113169e-01 2.53249347e-01 -1.58119902e-01 3.94253850e-01
-6.47746921e-01 -1.20254147e+00 6.74038529e-01 4.47637796e-01
-2.94306338e-01 -9.14229333e-01 -2.17581853e-01 -7.97447264e-01
-1.84068963e-01 -7.07561135e-01 6.85064971e-01 -9.72137570e-01
-1.39257729e+00 7.42102206e-01 2.49862775e-01 -1.48683631e+00
-4.61689323e-01 -3.26273113e-01 -1.00023019e+00 8.05638731e-01
-5.55699408e-01 -4.55688059e-01 -3.19213420e-01 1.06576049e+00
-1.76455423e-01 1.04934387e-01 7.32240260e-01 3.35749954e-01
-6.90823376e-01 -1.18219368e-01 1.42606184e-01 2.76107311e-01
2.14492619e-01 -1.11991155e+00 5.55698454e-01 5.28588235e-01
-2.38606229e-01 7.46475875e-01 7.82954395e-01 -8.47120702e-01
-1.51822102e+00 -7.98460126e-01 1.04106128e+00 7.71150738e-02
1.44864535e+00 -3.59049320e-01 -1.00765741e+00 6.90386295e-01
7.51747563e-02 -9.40561760e-03 7.36474514e-01 1.00547910e-01
-2.56268829e-01 -3.56555074e-01 -9.63555157e-01 6.86115861e-01
1.03848362e+00 -7.16984987e-01 -3.28973085e-01 5.63877583e-01
7.65107095e-01 -4.35051695e-02 -1.74618530e+00 9.56042558e-02
3.22450936e-01 -7.47308791e-01 9.09781873e-01 -3.30048501e-01
2.59658583e-02 -4.08863425e-01 3.52730572e-01 -1.42941880e+00
-3.72382462e-01 -1.20370471e+00 2.94538937e-03 1.07490683e+00
4.48990017e-01 -1.25464213e+00 3.07864219e-01 3.67100686e-01
4.64429446e-02 -3.07770044e-01 -7.79329300e-01 -1.10093224e+00
-4.05392259e-01 -6.45827830e-01 9.51138675e-01 1.61151969e+00
7.27459133e-01 4.65539545e-02 1.19185865e-01 -2.17960134e-01
6.32554054e-01 7.35826939e-02 6.68759525e-01 -1.79729939e+00
2.04468161e-01 -6.67832673e-01 -7.99622893e-01 -1.17126323e-01
-2.17442244e-01 -1.03926694e+00 -6.67345583e-01 -1.75137877e+00
-5.22205412e-01 -7.48449862e-01 5.83669404e-03 5.07055104e-01
5.75062156e-01 -1.47026688e-01 -1.87530518e-02 5.51600575e-01
1.85582153e-02 1.14283435e-01 1.13423145e+00 2.41837978e-01
-6.08679295e-01 2.42951468e-01 -1.57656968e-01 5.00421345e-01
1.00135279e+00 -5.81155717e-01 -6.12272084e-01 2.37812847e-01
6.23204529e-01 6.85735583e-01 2.15849161e-01 -7.16697574e-01
4.73147571e-01 -1.51222527e-01 6.99470118e-02 -8.63962591e-01
-9.60640684e-02 -9.61359918e-01 1.19055545e+00 7.19847620e-01
-1.26424909e-01 1.10687757e+00 2.31769279e-01 3.45780969e-01
-3.13215673e-01 2.26345271e-01 1.76994771e-01 1.90822989e-01
-3.43305796e-01 4.63384926e-01 -5.72772443e-01 -4.18569028e-01
1.18189383e+00 -3.84081960e-01 -5.82367778e-01 -6.80484235e-01
-1.31927431e+00 4.39092845e-01 3.13797891e-01 5.24263382e-01
6.66705012e-01 -1.58780754e+00 -4.98588085e-01 1.32573202e-01
-2.78840035e-01 -3.21780652e-01 -6.72358926e-03 1.20867586e+00
-9.92656887e-01 1.99993521e-01 -5.25520027e-01 -5.79526365e-01
-1.45192027e+00 5.43011487e-01 2.50916183e-01 -2.51599163e-01
-6.02954268e-01 6.33759722e-02 -6.76043451e-01 -6.16495550e-01
-9.05508250e-02 -4.89909172e-01 -4.53976810e-01 4.30412471e-01
5.08037388e-01 8.22015285e-01 -1.84350625e-01 -2.72246450e-01
-5.22051275e-01 6.95501864e-01 4.51455832e-01 -5.63213155e-02
1.87120366e+00 -5.00912741e-02 -1.21308780e+00 9.87164140e-01
1.37192774e+00 -3.12046111e-01 -4.86855626e-01 -4.34159786e-02
2.22745985e-01 -1.58161893e-01 -3.75935555e-01 -1.66229621e-01
-1.06311965e+00 1.60642907e-01 -2.27407087e-02 1.58865178e+00
1.33841169e+00 -4.77482267e-02 2.51600266e-01 9.00741220e-02
3.30915540e-01 -6.63623512e-01 -2.26265490e-01 5.83162427e-01
1.00918889e+00 -4.04252052e-01 1.57772526e-02 -6.27990723e-01
-3.21451500e-02 1.59841251e+00 5.53520303e-03 -4.32184130e-01
1.28322303e+00 3.89116317e-01 -9.68854800e-02 -6.37033463e-01
-8.30997765e-01 -4.49458733e-02 2.96479076e-01 6.82624459e-01
6.35408819e-01 2.14951530e-01 -4.45751578e-01 -6.41703024e-04
-6.62930250e-01 -1.52040228e-01 6.44518435e-01 8.18899930e-01
-2.08586380e-01 -1.07979333e+00 -5.59641421e-01 7.56375253e-01
-5.40464818e-02 6.05244190e-02 -4.77575064e-01 1.01123106e+00
-4.47267920e-01 1.14949477e+00 6.08064294e-01 -7.50332296e-01
4.97922957e-01 -2.00324416e-01 1.33100823e-01 -3.23403925e-01
-8.51088881e-01 -7.09745809e-02 2.29788169e-01 -6.43524826e-01
-3.59605879e-01 -7.91343570e-01 -1.32265949e+00 -1.13750398e+00
-1.72761157e-01 3.13701808e-01 8.60677898e-01 5.00904799e-01
4.10721272e-01 7.41170824e-01 9.58952665e-01 -6.54987156e-01
3.61808151e-01 -8.58745694e-01 -1.01764286e+00 -8.25179508e-04
1.41547754e-01 -1.85219243e-01 -6.06199741e-01 -2.34337375e-01] | [7.267972946166992, 3.436136245727539] |
03fb0353-04b9-4b5e-8a06-22eb0aeb6286 | combining-particle-and-tensor-network-methods | 2305.17884 | null | https://arxiv.org/abs/2305.17884v2 | https://arxiv.org/pdf/2305.17884v2.pdf | Combining Particle and Tensor-network Methods for Partial Differential Equations via Sketching | In this paper, we propose a general framework for solving high-dimensional partial differential equations with tensor networks. Our approach offers a comprehensive solution methodology, wherein we employ a combination of particle simulations to update the solution and re-estimations of the new solution as a tensor-network using a recently proposed tensor train sketching technique. Our method can also be interpreted as an alternative approach for performing particle number control by assuming the particles originate from an underlying tensor network. We demonstrate the versatility and flexibility of our approach by applying it to two specific scenarios: simulating the Fokker-Planck equation through Langevin dynamics and quantum imaginary time evolution via auxiliary-field quantum Monte Carlo. | ['Yuehaw Khoo', 'Yian Chen'] | 2023-05-29 | null | null | null | null | ['tensor-networks'] | ['methodology'] | [ 9.26619917e-02 7.54374862e-02 3.54866832e-01 1.72413155e-01
-1.59118250e-01 -6.21062458e-01 9.96071815e-01 -4.46441084e-01
-6.43774092e-01 9.76553440e-01 -8.36058185e-02 -4.35966164e-01
-4.80865359e-01 -8.32460523e-01 -4.21968371e-01 -9.73455727e-01
-2.67619669e-01 7.37858474e-01 1.50718600e-01 -7.03544021e-01
5.03177226e-01 8.64400566e-01 -1.08379602e+00 -1.39457047e-01
7.42479563e-01 7.66847193e-01 -4.62081850e-01 1.02129006e+00
6.30658716e-02 6.42627895e-01 -3.51968169e-01 -5.28077662e-01
4.05059338e-01 -3.47870767e-01 -1.12821722e+00 -9.34428722e-02
2.67526418e-01 -1.50794908e-01 -6.33219063e-01 1.05148637e+00
1.90161809e-01 5.48791409e-01 6.78594351e-01 -9.30762351e-01
-5.09311616e-01 4.39072788e-01 2.14754492e-01 2.80588835e-01
2.53815204e-01 3.24646056e-01 9.63651776e-01 -6.20607018e-01
7.17702687e-01 1.02922726e+00 1.03872526e+00 6.83659315e-01
-1.94707572e+00 -5.54753654e-02 -5.37890494e-01 7.63122141e-02
-1.36464441e+00 -1.83788016e-01 9.29308772e-01 -4.86507446e-01
8.89798224e-01 3.68355155e-01 8.80469203e-01 8.84927392e-01
3.83919388e-01 2.84655988e-01 1.28709733e+00 -5.84230542e-01
5.43957889e-01 -4.06331271e-02 3.15836430e-01 9.20915425e-01
2.49409765e-01 5.49542367e-01 -4.81472433e-01 -7.70830631e-01
1.01296270e+00 -2.05125540e-01 -7.69286826e-02 -5.91794372e-01
-1.76559043e+00 8.75651836e-01 1.57321364e-01 4.74021405e-01
-5.08054256e-01 6.78486407e-01 4.22732145e-01 1.67751864e-01
5.52097321e-01 6.56323731e-01 -1.99137926e-01 -1.80624366e-01
-8.46858978e-01 9.08342123e-01 1.36323595e+00 3.87650341e-01
9.41269159e-01 2.35247649e-02 -2.17936650e-01 -5.59434257e-02
2.04710498e-01 5.10226071e-01 -2.85037942e-02 -1.36336780e+00
-2.63061002e-03 -4.59325053e-02 6.41372144e-01 -8.20587814e-01
-3.90112579e-01 -2.25104526e-01 -9.08113480e-01 2.21330374e-01
5.69950342e-01 -8.28304961e-02 -4.79599625e-01 1.40659320e+00
5.01407504e-01 4.21940386e-01 7.86666423e-02 6.81366265e-01
1.33603334e-01 4.93950844e-01 -3.22902262e-01 -9.75471660e-02
1.03369689e+00 -6.88088894e-01 -7.70799398e-01 7.82079577e-01
7.03545332e-01 -5.76309264e-01 3.72880071e-01 5.16630709e-01
-1.29623353e+00 -1.93360135e-01 -8.71844947e-01 -7.24741444e-02
-3.46545130e-01 5.28079132e-03 1.19684601e+00 7.81854451e-01
-1.23644853e+00 1.73379445e+00 -1.03276014e+00 -1.62446886e-01
-8.35795142e-03 4.97692883e-01 -1.38665468e-01 4.12912905e-01
-1.10423505e+00 9.47558284e-01 2.72881448e-01 4.08847153e-01
-5.77560067e-01 -7.95222759e-01 -3.05918097e-01 -2.71394432e-01
5.14260959e-03 -1.06730723e+00 1.29393804e+00 -1.80881903e-01
-2.21251655e+00 2.92995840e-01 2.39165332e-02 -5.43411255e-01
3.80076379e-01 3.39569002e-01 -1.80971682e-01 2.05341876e-01
-2.08130345e-01 4.57307510e-02 7.02677667e-01 -9.26543176e-01
1.49021879e-01 -7.52862468e-02 3.43131453e-01 -3.02640945e-01
1.39196560e-01 -3.30127627e-01 3.79421890e-01 -4.09523755e-01
3.68806899e-01 -1.28349388e+00 -4.85441357e-01 -6.23306409e-02
-3.86142701e-01 -1.75057054e-01 4.83272970e-01 -1.95478559e-01
1.01697779e+00 -1.55632937e+00 7.64735103e-01 5.68225384e-01
6.23486757e-01 3.10870260e-01 1.23718180e-01 1.02536893e+00
2.81966627e-02 1.74496040e-01 -3.90257448e-01 -5.32379687e-01
4.69688684e-01 4.76814717e-01 -4.90040094e-01 7.07787097e-01
2.24562079e-01 9.16613519e-01 -1.03835750e+00 -1.35192141e-01
2.64187127e-01 7.84368813e-01 -7.61634350e-01 -3.95416796e-01
-2.11782306e-01 8.82703125e-01 -5.54790914e-01 1.24172352e-01
8.16698492e-01 -2.65907764e-01 5.47388755e-02 -6.35920703e-01
-5.86250961e-01 2.93540001e-01 -1.38793302e+00 1.58536172e+00
-1.12208277e-01 2.76825756e-01 1.57098189e-01 -1.06644320e+00
4.91352797e-01 2.52911270e-01 5.23630798e-01 -2.66143441e-01
2.34399751e-01 4.37398404e-01 1.77551463e-01 -4.92561668e-01
8.42322350e-01 -5.97847164e-01 8.24410617e-02 8.40763927e-01
1.52681693e-01 -6.08107448e-01 4.39957321e-01 4.65227127e-01
1.18016720e+00 3.93469870e-01 -3.58935408e-02 -5.58334351e-01
7.11654127e-01 8.99058059e-02 -8.42040628e-02 1.00755072e+00
-4.42691296e-02 3.42786983e-02 6.03206754e-01 -7.63246894e-01
-1.65903997e+00 -1.04128420e+00 -3.85265738e-01 4.61254597e-01
-1.91167682e-01 -5.88641703e-01 -7.28084147e-01 -1.34588167e-01
2.11333528e-01 5.29895842e-01 -8.57589781e-01 6.35444745e-02
-7.03799307e-01 -8.87549758e-01 7.16971219e-01 -1.05822362e-01
4.02560383e-01 -6.25648916e-01 -1.10072970e-01 2.45512277e-01
2.05622450e-01 -8.71683180e-01 -6.36544600e-02 -6.31284714e-02
-1.01401687e+00 -8.48627806e-01 -6.42091990e-01 8.65445137e-02
4.54707175e-01 -1.46963298e-01 7.92386770e-01 1.57138228e-01
-2.64488369e-01 5.65911472e-01 5.72723821e-02 3.56921792e-01
-8.66096377e-01 9.48840752e-03 4.13294345e-01 1.90973431e-01
-2.63259001e-02 -7.25315034e-01 -5.33074617e-01 3.45656350e-02
-1.00842905e+00 -5.51756173e-02 7.82277659e-02 7.73620963e-01
1.89778313e-01 6.14671372e-02 -2.28079274e-01 -7.16865838e-01
9.65151727e-01 -1.88186929e-01 -8.63997340e-01 -2.83006448e-02
-3.35965276e-01 6.73209727e-01 6.11446440e-01 -4.57986981e-01
-8.18826377e-01 -1.06033519e-01 -1.23648450e-01 -1.16459899e-01
3.58891577e-01 2.69542396e-01 6.66465461e-01 -1.20810533e+00
6.07219875e-01 2.61140317e-01 1.95983022e-01 -5.44076025e-01
5.32062888e-01 9.03233290e-02 4.34485316e-01 -1.06415558e+00
1.21055818e+00 9.15669858e-01 9.30520773e-01 -6.55000806e-01
-5.21739960e-01 -1.08285323e-01 -9.21405733e-01 -5.20279445e-02
6.64597332e-01 -1.24932803e-01 -1.68346381e+00 6.35466874e-01
-1.55252838e+00 -2.11778060e-01 -5.15977979e-01 7.16082633e-01
-5.64176202e-01 8.07265520e-01 -8.03363919e-01 -1.02158129e+00
-9.14570391e-02 -1.10447752e+00 9.58274543e-01 -1.25320986e-01
4.34216261e-02 -1.55490661e+00 6.96906388e-01 -1.69166297e-01
8.69537830e-01 2.27824792e-01 6.10243499e-01 -2.07839295e-01
-8.72071326e-01 -3.13995689e-01 -1.17228575e-01 2.77093351e-01
-5.06017387e-01 2.85576791e-01 -6.97418094e-01 -2.18975723e-01
6.57002330e-02 9.04710293e-02 7.33982682e-01 -6.23479895e-02
6.82582855e-01 -1.75413430e-01 -2.46279731e-01 5.49391448e-01
1.38811624e+00 -5.13445258e-01 5.63484907e-01 -5.96317798e-02
8.20808887e-01 3.73587132e-01 -1.60561129e-01 4.12749916e-01
3.34953427e-01 7.95140445e-01 9.37581509e-02 4.35101032e-01
1.71499148e-01 4.70035784e-02 1.24995194e-01 1.15862608e+00
-7.93245971e-01 2.40132481e-01 -7.75097132e-01 7.11577535e-02
-1.74096811e+00 -1.31523025e+00 -6.23496711e-01 2.18211317e+00
6.38706446e-01 -1.04551286e-01 1.25325069e-01 5.82894534e-02
3.85509908e-01 -5.85892536e-02 -2.78173089e-01 -4.01953965e-01
6.91684932e-02 6.74560905e-01 7.65285730e-01 8.36342394e-01
-8.15631211e-01 6.69826150e-01 8.03941250e+00 5.16333699e-01
-9.34123337e-01 3.37207764e-01 -3.70824635e-01 2.80351341e-01
-4.81747895e-01 4.30913031e-01 -6.22608185e-01 3.16694409e-01
1.06470251e+00 -1.92581400e-01 8.91958594e-01 1.48867577e-01
1.44432098e-01 -2.91948766e-01 -9.12182868e-01 6.82131648e-01
-2.06387252e-01 -1.83720005e+00 -1.02590332e-02 3.58353853e-01
7.49112725e-01 1.20078109e-01 -1.69328496e-01 1.06491812e-01
1.95396677e-01 -4.11043078e-01 4.32232141e-01 1.08332384e+00
3.53600949e-01 -4.01012540e-01 2.61907667e-01 1.94534436e-01
-9.30822670e-01 2.86938936e-01 -2.95302212e-01 -5.83106458e-01
3.55460793e-01 8.05297136e-01 -4.44327116e-01 4.52678591e-01
1.65186487e-02 4.36234981e-01 -1.66542187e-01 1.11847055e+00
2.24618644e-01 4.78030384e-01 -5.92128396e-01 -3.26598734e-01
3.74111861e-01 -8.42621982e-01 8.91269743e-01 8.35210145e-01
2.33446836e-01 9.34136286e-02 -9.16745067e-02 1.32813776e+00
3.29829127e-01 -1.24800012e-01 -5.35825372e-01 -4.07436162e-01
-9.55447331e-02 1.40548813e+00 -7.96038508e-01 -5.61829567e-01
6.64486140e-02 9.29584980e-01 1.79166436e-01 5.14353991e-01
-5.82826495e-01 -4.85222697e-01 5.86662948e-01 -2.36346513e-01
4.94774938e-01 -8.87973785e-01 1.38876885e-01 -1.70702362e+00
-1.53969690e-01 -1.97371364e-01 -3.85649681e-01 -6.16337657e-01
-1.23801494e+00 4.00710225e-01 2.29753345e-01 -9.32424486e-01
-1.76511124e-01 -1.03721428e+00 -5.73015690e-01 1.04510069e+00
-1.23400867e+00 -7.73389697e-01 1.97395295e-01 4.36433017e-01
-6.39226913e-01 9.92637351e-02 1.02687240e+00 1.85688704e-01
-3.99446517e-01 -6.11585118e-02 4.99069661e-01 -2.69531250e-01
-9.91973430e-02 -1.35733664e+00 7.29151130e-01 6.08581126e-01
-1.11394964e-01 1.02914727e+00 1.11979461e+00 -5.64030766e-01
-1.84803009e+00 -5.66067934e-01 8.36704910e-01 -5.45435131e-01
1.42195225e+00 -6.09434843e-01 -9.62816477e-01 6.14775479e-01
1.29698604e-01 2.69665182e-01 4.50805902e-01 -2.84644477e-02
-1.29315943e-01 2.24459276e-01 -1.21813989e+00 4.74409997e-01
9.13274109e-01 -8.66329372e-01 -3.60083699e-01 8.42641592e-01
5.11694908e-01 -5.07514954e-01 -1.24937057e+00 5.30006066e-02
5.01879990e-01 -8.41329932e-01 8.71552527e-01 -6.26029670e-01
-8.07886496e-02 -3.55230987e-01 2.43121497e-02 -1.22456121e+00
-3.25662971e-01 -1.52470648e+00 -1.70022488e-01 6.16699338e-01
-2.18773745e-02 -1.07760954e+00 6.06034636e-01 5.70246279e-01
3.38719897e-02 -4.49488431e-01 -1.16727865e+00 -6.43972099e-01
3.25233787e-01 -5.73552787e-01 5.93918324e-01 6.80618286e-01
9.73529965e-02 8.68049338e-02 -4.69336033e-01 1.01091303e-01
9.53805208e-01 -3.86349231e-01 6.07938528e-01 -1.32339013e+00
-5.90798497e-01 -3.80497783e-01 -4.92426544e-01 -9.68583405e-01
3.88345063e-01 -1.14022851e+00 -4.15429085e-01 -1.06029654e+00
-7.53292069e-03 -3.74285519e-01 7.42758960e-02 -1.07516438e-01
5.51360965e-01 5.70412755e-01 1.40659586e-01 4.03420001e-01
-5.64000487e-01 4.99464184e-01 1.43581176e+00 3.28369349e-01
8.47326592e-02 1.24854177e-01 5.08203395e-02 5.56089163e-01
5.43832123e-01 -4.56933826e-01 1.44089572e-02 -4.92008068e-02
8.76490176e-01 1.79541707e-01 1.03464627e+00 -1.06998217e+00
3.99516135e-01 9.29407626e-02 -1.50872290e-01 -5.18824577e-01
6.03580952e-01 -3.01485151e-01 4.39219028e-01 6.54596806e-01
-1.74025118e-01 1.53137818e-01 7.42995813e-02 5.59669733e-01
3.27563256e-01 -5.67979515e-01 5.59980989e-01 -2.97575772e-01
-2.66595215e-01 2.36672983e-01 -6.22353435e-01 -2.27668390e-01
5.16565323e-01 9.51299816e-02 -4.00392950e-01 -4.54271957e-02
-1.18781757e+00 -2.89375812e-01 5.88605583e-01 -4.52485442e-01
2.56317765e-01 -1.53246367e+00 -4.62345809e-01 3.52691352e-01
-1.61582187e-01 -7.04192340e-01 3.95703852e-01 1.29105806e+00
-9.48103964e-01 6.02619171e-01 -8.77456516e-02 -4.11619902e-01
-6.08955383e-01 3.57964247e-01 8.87144148e-01 -4.15913194e-01
-5.97631276e-01 4.63690579e-01 -4.70025480e-01 -7.76004374e-01
-3.73582333e-01 -5.63565969e-01 3.27013552e-01 -4.18947458e-01
5.09435892e-01 7.06143737e-01 -4.53989618e-02 -9.04792547e-01
-7.66730160e-02 5.71541488e-01 2.24295378e-01 -4.99710411e-01
1.04127812e+00 -2.36275885e-02 -7.58528531e-01 5.82969248e-01
1.32346261e+00 1.02255128e-01 -7.74957836e-01 -2.90365160e-01
-2.70681202e-01 -1.16166271e-01 2.42678046e-01 -3.64178479e-01
-4.99145567e-01 7.62401462e-01 1.83484510e-01 7.97990084e-01
1.83367059e-01 -2.08525490e-02 7.64448583e-01 8.76196742e-01
4.66145188e-01 -9.45536911e-01 -1.62594974e-01 7.24169374e-01
6.32959664e-01 -8.13351989e-01 1.75102517e-01 -2.74489820e-01
1.93196014e-02 1.56734622e+00 -3.25407714e-01 -7.13940322e-01
8.79610121e-01 -8.56897328e-03 -5.91592073e-01 -4.73225743e-01
-6.30987227e-01 -1.75827727e-01 3.21750939e-01 4.25911754e-01
3.31051856e-01 2.25511849e-01 -4.84442294e-01 -2.95435995e-01
-2.71162659e-01 2.39869833e-01 1.00277281e+00 7.02171028e-01
-1.09905303e-01 -1.47009325e+00 -4.69913006e-01 2.29770184e-01
-1.73898995e-01 -1.33787900e-01 -5.24993278e-02 6.79114640e-01
-8.44602883e-02 3.88072759e-01 -7.37965479e-02 -2.39178404e-01
1.28805459e-01 3.30904782e-01 1.12015235e+00 -4.02870446e-01
-4.37687188e-01 -2.69806445e-01 7.93933049e-02 -5.61715305e-01
-8.54628563e-01 -8.08561087e-01 -1.16660202e+00 -8.97050560e-01
-2.79691219e-01 4.21956271e-01 1.07590902e+00 1.11793387e+00
3.54026407e-01 3.53061646e-01 4.19203877e-01 -1.54715514e+00
-1.00112534e+00 -8.60192060e-01 -7.60676920e-01 1.97827145e-01
5.26319563e-01 -9.36240911e-01 -6.97959840e-01 -5.16228378e-01] | [5.66885232925415, 4.945534706115723] |
a7433bf4-1983-4f9e-8f4a-e613c45a6c8e | open-set-object-detection-using | 2211.11530 | null | https://arxiv.org/abs/2211.11530v1 | https://arxiv.org/pdf/2211.11530v1.pdf | Open-Set Object Detection Using Classification-free Object Proposal and Instance-level Contrastive Learning with Appendix | Detecting both known and unknown objects is a fundamental skill for robot manipulation in unstructured environments. Open-set object detection (OSOD) is a promising direction to handle the problem consisting of two subtasks: objects and background separation, and open-set object classification. In this paper, we present Openset RCNN to address the challenging OSOD. To disambiguate unknown objects and background in the first subtask, we propose to use classification-free region proposal network (CF-RPN) which estimates the objectness score of each region purely using cues from object's location and shape preventing overfitting to the training categories. To identify unknown objects in the second subtask, we propose to represent them using the complementary region of known categories in a latent space which is accomplished by a prototype learning network (PLN). PLN performs instance-level contrastive learning to encode proposals to a latent space and builds a compact region centering with a prototype for each known category. Further, we note that the detection performance of unknown objects can not be unbiasedly evaluated on the situation that commonly used object detection datasets are not fully annotated. Thus, a new benchmark is introduced by reorganizing GraspNet-1billion, a robotic grasp pose detection dataset with complete annotation. Extensive experiments demonstrate the merits of our method. We finally show that our Openset RCNN can endow the robot with an open-set perception ability to support robotic rearrangement tasks in cluttered environments. More details can be found in https://sites.google.com/view/openest-rcnn/ | ['Rong Xiong', 'Yue Wang', 'Yifei Yang', 'Zhongxiang Zhou'] | 2022-11-21 | null | null | null | null | ['robot-manipulation'] | ['robots'] | [ 1.62156105e-01 1.92325726e-01 -1.20522141e-01 -2.71767706e-01
-5.37273109e-01 -7.18274236e-01 3.23531151e-01 -6.59655407e-02
-2.59309530e-01 3.76146287e-01 -2.81838685e-01 1.38852715e-01
-8.20335373e-02 -5.14006913e-01 -1.03117406e+00 -7.30997801e-01
-1.59941539e-01 6.08147860e-01 6.63053095e-01 -1.22336864e-01
2.52581269e-01 7.48831689e-01 -1.79658818e+00 4.75199163e-01
8.17447484e-01 1.31525552e+00 9.14520919e-01 2.93569684e-01
-1.02161087e-01 2.86515236e-01 -5.54167867e-01 1.41846105e-01
6.76733255e-01 2.63617933e-01 -6.79929614e-01 3.12211830e-02
5.08027494e-01 -4.62409437e-01 -1.93872616e-01 1.16434515e+00
3.71844828e-01 1.90935910e-01 8.04074764e-01 -1.48807514e+00
-5.78845561e-01 5.70544362e-01 -5.57175517e-01 2.72602551e-02
2.55872309e-01 3.38085264e-01 8.04273367e-01 -1.27347529e+00
5.22168756e-01 1.54609466e+00 4.43113893e-01 4.92690623e-01
-1.21359718e+00 -6.53898001e-01 7.04547763e-01 8.41704458e-02
-1.35918152e+00 -1.64731383e-01 7.76740193e-01 -6.11048400e-01
5.36506653e-01 1.96118876e-01 5.62520683e-01 1.29748189e+00
8.48809034e-02 1.13995552e+00 9.14658964e-01 -1.25485674e-01
2.74990559e-01 7.32922256e-02 3.31965208e-01 5.97314537e-01
4.61020052e-01 1.12306863e-01 -1.68542519e-01 3.14977989e-02
1.00162446e+00 3.36574316e-01 -3.50893736e-01 -1.17555416e+00
-1.55350578e+00 4.64208901e-01 1.03158522e+00 3.67271192e-02
-3.77885044e-01 1.59650132e-01 2.75065154e-01 -1.24245929e-02
1.04503885e-01 5.75317144e-01 -5.88778615e-01 3.99654955e-01
-1.87427863e-01 1.63603023e-01 7.31835425e-01 1.50417387e+00
7.38972723e-01 -3.47401887e-01 -4.16667968e-01 8.69784534e-01
3.12907487e-01 5.15108645e-01 2.38321960e-01 -8.92748594e-01
4.31310862e-01 6.88392341e-01 2.23802000e-01 -8.12803090e-01
-4.44336593e-01 -4.35491443e-01 -5.23746610e-01 3.75680029e-01
5.34051299e-01 2.59297639e-01 -1.18504012e+00 1.59947944e+00
5.13652861e-01 -1.89440370e-01 5.96054830e-02 1.15914226e+00
8.80354285e-01 4.05592591e-01 -1.55848160e-01 2.71258831e-01
1.46998489e+00 -1.18545210e+00 -3.82535130e-01 -4.31709170e-01
2.80731678e-01 -5.27519226e-01 1.11130059e+00 4.97817218e-01
-7.08775640e-01 -6.14791334e-01 -9.39288557e-01 -1.59997249e-03
-5.34027696e-01 5.92443883e-01 6.27750218e-01 8.16276763e-03
-5.73122799e-01 3.63461018e-01 -9.64156568e-01 -3.47312242e-01
7.84109592e-01 3.62771451e-01 -5.12777269e-01 -3.24675471e-01
-5.34526169e-01 7.64549255e-01 8.70438337e-01 4.07375127e-01
-1.30492210e+00 -2.57161945e-01 -9.34622288e-01 -9.55635607e-02
1.01434267e+00 -3.80683362e-01 9.79665697e-01 -6.73228741e-01
-1.21343100e+00 7.28284776e-01 2.24301636e-01 -1.03613302e-01
5.30459225e-01 -4.47020829e-01 1.90968007e-01 3.00093256e-02
3.45944792e-01 9.48614419e-01 9.49306786e-01 -1.82979000e+00
-6.14638448e-01 -5.72526872e-01 2.19543129e-01 1.16866417e-01
4.86893132e-02 -2.25808024e-01 -5.02252340e-01 -7.10116982e-01
9.26166892e-01 -1.05838847e+00 -1.48851737e-01 4.20008153e-01
-5.31458676e-01 -5.21344006e-01 8.97380590e-01 -4.06380683e-01
4.03435916e-01 -2.16125178e+00 3.02925318e-01 6.66044559e-03
2.67253160e-01 6.93897009e-02 -2.92595088e-01 -3.11604291e-02
-2.98592858e-02 -3.62997204e-01 -1.48250371e-01 -5.85307134e-04
2.00775787e-01 3.37318391e-01 -4.63288218e-01 5.07720768e-01
4.04954523e-01 9.42166209e-01 -1.06736076e+00 -2.73849666e-01
1.14383802e-01 7.71976188e-02 -4.85488713e-01 3.57175857e-01
-5.20497501e-01 4.31344450e-01 -4.07998413e-01 1.10490429e+00
9.13718700e-01 -2.00307053e-02 -6.47620708e-02 -3.70137125e-01
-7.32776523e-02 -1.10959940e-01 -1.47968233e+00 1.78906536e+00
-8.87524113e-02 2.29938656e-01 3.68344069e-01 -9.35249329e-01
1.08064985e+00 -2.04715207e-01 2.88051128e-01 -1.51506603e-01
7.14594796e-02 3.80520344e-01 9.38030258e-02 -6.18556678e-01
4.98738915e-01 3.28899503e-01 -1.77647337e-01 -1.41960746e-02
3.83127242e-01 -1.74193591e-01 4.74327840e-02 -2.34262217e-02
1.04061103e+00 5.36461353e-01 1.42337963e-01 -3.09417486e-01
1.48176014e-01 -1.96986664e-02 6.80938125e-01 1.06510746e+00
-4.78548616e-01 6.61737204e-01 2.89215863e-01 -4.34055060e-01
-6.28841341e-01 -1.41903007e+00 -2.01889798e-01 1.25669956e+00
7.42201209e-01 1.46510556e-01 -4.18150693e-01 -9.39841330e-01
3.43398452e-01 3.42911392e-01 -5.38239777e-01 -2.19957426e-01
-6.15958869e-01 -3.89323562e-01 1.42733619e-01 7.44449854e-01
1.72657043e-01 -1.35209084e+00 -8.23932052e-01 2.33012047e-02
-2.54470170e-01 -1.07274199e+00 -3.32493216e-01 7.14678228e-01
-6.47062480e-01 -1.15580559e+00 -7.93860435e-01 -1.13952470e+00
8.70579064e-01 6.20039344e-01 7.24908650e-01 -1.98832422e-01
-4.96191859e-01 4.25347209e-01 -5.77326357e-01 -6.54066443e-01
-1.94949493e-01 -2.65064780e-02 4.34140444e-01 -1.81196183e-01
2.00861678e-01 -2.17018723e-01 -6.29398227e-01 7.07939684e-01
-6.69396102e-01 -7.03468174e-02 9.02114332e-01 7.60878623e-01
6.67497516e-01 -2.17210099e-01 5.49248934e-01 -2.54603714e-01
1.23699054e-01 -5.26731014e-01 -7.70369947e-01 3.54133010e-01
1.71651319e-02 6.60963356e-02 1.33512914e-01 -8.56795549e-01
-7.50991881e-01 4.15638834e-01 4.30433691e-01 -8.10742199e-01
-5.13075888e-01 1.33448794e-01 -4.59352791e-01 -4.39890884e-02
5.83480775e-01 5.29203191e-02 -5.08525176e-04 -5.44133127e-01
4.02706623e-01 6.80737317e-01 8.56635392e-01 -8.65768194e-01
6.90310180e-01 4.46553379e-01 -3.25655371e-01 -4.33420867e-01
-9.29025114e-01 -8.92380118e-01 -9.74573195e-01 -2.46388212e-01
8.06662798e-01 -1.14206421e+00 -6.84780300e-01 4.58192647e-01
-1.36689270e+00 -3.54269743e-01 -3.76942933e-01 4.87341464e-01
-7.68644273e-01 3.53435397e-01 -2.71590114e-01 -9.34963763e-01
-3.92574258e-02 -1.25330067e+00 1.40423608e+00 1.64535344e-02
2.52616584e-01 -1.25415891e-01 -4.80664968e-01 2.29730204e-01
1.26702748e-02 2.25699797e-01 5.81556082e-01 -9.53339696e-01
-8.66343617e-01 -1.84427440e-01 -4.60608214e-01 3.24883550e-01
2.31768623e-01 -3.59094799e-01 -1.03035700e+00 -5.83274126e-01
-3.87433991e-02 -5.35079181e-01 1.14400542e+00 2.44015872e-01
1.42766035e+00 -1.50136515e-01 -6.66872084e-01 4.42951977e-01
1.14908671e+00 9.58765671e-02 3.15782875e-01 2.04556689e-01
7.58105159e-01 8.23759556e-01 1.12362373e+00 4.02130187e-01
1.78005368e-01 6.54600680e-01 8.84258330e-01 1.54263347e-01
-1.20110229e-01 -9.84465107e-02 3.94337833e-01 2.35601351e-01
1.17439358e-02 -2.81792909e-01 -8.56805027e-01 5.68723738e-01
-2.01768255e+00 -7.41479993e-01 8.19170326e-02 2.05020070e+00
5.07495642e-01 2.74707079e-01 1.72351487e-02 -2.02810347e-01
9.60633457e-01 -2.59505928e-01 -8.21546197e-01 3.97428364e-01
-1.39850870e-01 -2.69856632e-01 3.68758768e-01 -5.34811355e-02
-1.38971555e+00 9.73625541e-01 4.72891283e+00 8.14806402e-01
-8.61063719e-01 -1.13784606e-02 1.07915320e-01 1.53042853e-01
1.98912770e-01 -1.05231136e-01 -9.04332101e-01 3.41924697e-01
-5.86708337e-02 5.06954074e-01 3.30383837e-01 1.30606973e+00
-1.81541249e-01 -1.44482583e-01 -1.42902744e+00 9.37050641e-01
9.80768129e-02 -8.87085736e-01 2.20798142e-02 -3.76226902e-02
4.04842556e-01 2.13221416e-01 -6.16976097e-02 6.40871704e-01
1.96079582e-01 -5.95395625e-01 1.22806096e+00 4.56359953e-01
6.44340336e-01 -1.76894113e-01 6.54698133e-01 5.46247602e-01
-1.21443772e+00 -6.78818405e-01 -6.90979123e-01 1.41207993e-01
-1.57673940e-01 1.55095294e-01 -1.00006604e+00 4.00212586e-01
1.03194284e+00 5.87477326e-01 -4.43067223e-01 1.34827197e+00
-1.79459333e-01 1.21953394e-02 -4.26158190e-01 9.08338279e-02
6.66464716e-02 -4.61373441e-02 8.82147253e-01 8.94478440e-01
2.43984446e-01 9.98814031e-02 8.01442802e-01 1.29411888e+00
1.16001353e-01 -3.25354248e-01 -5.67347407e-01 2.13539392e-01
5.02701223e-01 1.28569317e+00 -1.05204642e+00 -2.46517032e-01
-4.17034850e-02 9.68606114e-01 5.55020690e-01 2.22580597e-01
-7.00570524e-01 -4.45152879e-01 4.34656411e-01 -7.20790029e-02
5.96646845e-01 -2.92244583e-01 5.96995056e-02 -1.21666145e+00
4.13053840e-01 -6.56801999e-01 3.03801924e-01 -9.61420596e-01
-1.41786802e+00 5.46817303e-01 3.11182827e-01 -1.44013989e+00
3.45197678e-01 -1.20689547e+00 -3.06848258e-01 5.05350888e-01
-1.39956462e+00 -1.29131818e+00 -7.25816011e-01 2.37909153e-01
8.92805457e-01 -5.63840941e-02 5.66899121e-01 -1.17698289e-01
-5.22156119e-01 2.73059666e-01 -1.16115250e-02 1.57019839e-01
7.50594437e-01 -1.34786272e+00 6.63727596e-02 6.11220419e-01
-4.36790921e-02 6.47943497e-01 5.67662060e-01 -8.62934768e-01
-1.41737199e+00 -1.20831525e+00 -2.40295261e-01 -6.79635286e-01
5.72520018e-01 -8.60362470e-01 -1.02198792e+00 6.54510796e-01
-4.61941719e-01 6.25129461e-01 1.60168037e-01 -1.40472248e-01
-4.45716918e-01 5.97092323e-02 -1.11746144e+00 3.12475711e-01
1.41787887e+00 -1.13555260e-01 -7.51127541e-01 5.13934433e-01
1.10574281e+00 -6.81777716e-01 -7.61132419e-01 8.39360237e-01
6.14504993e-01 -5.36914766e-01 1.10402286e+00 -4.88205850e-01
2.25845665e-01 -5.52661777e-01 -4.94568288e-01 -1.08710897e+00
-3.15568089e-01 -2.56116241e-01 -2.63576567e-01 9.90949869e-01
1.93963256e-02 -5.70924759e-01 7.33173490e-01 2.14982986e-01
-6.22060239e-01 -7.13251412e-01 -8.30234349e-01 -1.18471575e+00
-1.09254546e-01 -3.32905918e-01 5.00101686e-01 5.84844828e-01
-2.15328127e-01 4.56005447e-02 1.43657908e-01 6.56874537e-01
6.18915319e-01 4.03868288e-01 8.97018194e-01 -1.56853402e+00
-8.91657323e-02 -2.03221172e-01 -5.53199887e-01 -1.37000489e+00
2.29448020e-01 -9.01246250e-01 7.62747526e-01 -1.50299680e+00
3.78963977e-01 -9.37740207e-01 -4.21664447e-01 6.67849600e-01
-3.09966947e-03 1.47244290e-01 3.62127215e-01 4.22657013e-01
-9.73070562e-01 8.22703183e-01 1.32564664e+00 -4.68479067e-01
-2.80374169e-01 1.35567039e-01 -3.98097843e-01 8.26807678e-01
6.23655498e-01 -3.78604949e-01 -6.57433569e-02 -3.47655445e-01
-2.74142027e-01 -2.44487658e-01 8.76862407e-01 -1.13499546e+00
5.06191440e-02 -2.70728260e-01 4.45284843e-01 -9.62217510e-01
5.51522315e-01 -1.02186799e+00 -2.34158605e-01 6.07470810e-01
-1.72162831e-01 -3.50742102e-01 1.98476329e-01 9.85363066e-01
1.87562928e-01 -3.01035851e-01 5.88781476e-01 -3.49893421e-01
-1.08342683e+00 2.68396050e-01 -7.98168406e-02 -2.91148335e-01
1.22096956e+00 -4.32016492e-01 -4.59775597e-01 2.94098228e-01
-8.42852294e-01 4.91677850e-01 4.93006647e-01 8.33955884e-01
8.65062356e-01 -1.09764850e+00 -4.18898433e-01 4.07152981e-01
6.46576047e-01 6.70026422e-01 1.26331255e-01 7.66795993e-01
-2.69576252e-01 2.55564392e-01 -2.24767074e-01 -1.12603402e+00
-9.92291987e-01 9.31614220e-01 2.14690834e-01 2.90119767e-01
-6.47703707e-01 1.10086524e+00 7.40441144e-01 -9.83115673e-01
6.83709204e-01 -7.07740545e-01 -2.29568958e-01 -1.71389267e-01
1.86261192e-01 3.29344690e-01 -7.33922645e-02 -6.04640305e-01
-3.88627708e-01 4.24780190e-01 -8.86543915e-02 5.57369113e-01
1.32648408e+00 -5.12783453e-02 -3.20445180e-01 5.00687659e-01
8.91605020e-01 -3.81749272e-01 -1.64403689e+00 -3.68583620e-01
1.33858202e-03 -4.21766549e-01 -3.36733729e-01 -8.03516805e-01
-7.22251296e-01 5.86405218e-01 9.38898861e-01 -1.01033330e-01
6.75876677e-01 4.45703387e-01 3.00907850e-01 8.86869133e-01
8.02076459e-01 -1.02123964e+00 5.88264763e-01 6.38318717e-01
1.37135684e+00 -1.63404810e+00 -1.68383658e-01 -7.24992692e-01
-3.50879729e-01 1.13905668e+00 1.25237966e+00 -1.26368776e-01
4.70499605e-01 1.16011307e-01 -1.58475637e-01 -4.05235648e-01
-3.68983686e-01 -2.10608691e-01 3.96399915e-01 7.08407998e-01
-2.48441264e-01 2.00841635e-01 2.95496196e-01 7.82308221e-01
7.93004707e-02 -3.95120949e-01 4.59652692e-01 1.28607380e+00
-7.89627016e-01 -4.85368490e-01 -6.45535529e-01 4.36519235e-01
9.43745449e-02 3.18426847e-01 -2.63463289e-01 5.90634227e-01
4.35464323e-01 6.81822479e-01 1.72225043e-01 -3.02755237e-01
3.61996502e-01 -3.10603499e-01 5.80497503e-01 -1.02091396e+00
-5.95952459e-02 -3.45539264e-02 -4.35431927e-01 -7.14439631e-01
-2.97542334e-01 -5.79953671e-01 -1.20576262e+00 6.03966713e-01
-9.32401359e-01 -1.41774431e-01 6.48211718e-01 6.50482535e-01
2.76432484e-01 6.30240619e-01 3.21341395e-01 -1.57262039e+00
-9.17957842e-01 -9.45555091e-01 -5.14541388e-01 1.29904464e-01
3.55883121e-01 -1.33139181e+00 -2.33716026e-01 -2.26098761e-01] | [6.010989665985107, -1.0403622388839722] |
2d1af57d-c310-4db7-b8b0-eb2dcde3750d | joint-routing-and-resource-allocation-for | 1809.07470 | null | http://arxiv.org/abs/1809.07470v1 | http://arxiv.org/pdf/1809.07470v1.pdf | Joint Routing and Resource Allocation for Millimeter Wave Picocellular Backhaul | Picocellular architectures are essential for providing the spatial reuse
required to satisfy the ever-increasing demand for mobile data. A key
deployment challenge is to provide backhaul connections with sufficiently high
data rate. Providing wired support (e.g., using optical fiber) to pico base
stations deployed opportunistically on lampposts and rooftops is impractical,
hence wireless backhaul becomes an attractive approach. A multihop mesh network
comprised of directional millimeter wave links is considered here for this
purpose. The backhaul design problem is formulated as one of joint routing and
resource allocation, accounting for mutual interference across simultaneously
active links. A computationally tractable formulation is developed by
leveraging the localized nature of interference and the provable existence of a
sparse optimal allocation. Numerical results are provided for millimeter (mm)
wave mesh networks, which are well suited for scaling backhaul data rates due
to abundance of spectrum, and the ability to form highly directional,
electronically steerable beams. | [] | 2018-09-20 | null | null | null | null | ['pico'] | ['natural-language-processing'] | [ 3.63221206e-03 5.02001882e-01 -3.45309854e-01 1.56037793e-01
-5.76793179e-02 -7.91192889e-01 1.23723652e-02 -4.64423627e-01
-2.51039267e-01 1.50746441e+00 -8.97087529e-02 -7.07262337e-01
-7.50314236e-01 -1.06668413e+00 7.88395330e-02 -1.03150320e+00
-5.04039586e-01 2.43872583e-01 -3.91410828e-01 3.02283224e-02
7.10597113e-02 7.96162069e-01 -8.45705152e-01 -7.40048468e-01
7.71136224e-01 1.16601241e+00 3.23128730e-01 5.72070599e-01
-1.69028684e-01 1.47127956e-01 -4.93559927e-01 -1.92344919e-01
1.24528095e-01 -2.23426312e-01 -2.98939049e-01 2.18742430e-01
1.12426251e-01 -4.47937638e-01 -4.08990234e-01 6.04447186e-01
6.92197859e-01 -1.74464598e-01 5.45112908e-01 -1.23103166e+00
-4.59148996e-02 1.24118239e-01 -3.83583665e-01 1.34348229e-01
1.54520661e-01 -3.79727297e-02 1.07418501e+00 -4.00260001e-01
6.45880759e-01 5.33731818e-01 3.51144403e-01 3.27699810e-01
-1.08735049e+00 -6.32974923e-01 -2.26803705e-01 -3.36587816e-01
-1.24536741e+00 -6.91044271e-01 4.37040597e-01 -1.53805792e-01
5.39802849e-01 4.71041650e-01 1.00896132e+00 5.02066851e-01
3.79396200e-01 5.96531667e-02 7.25657701e-01 -3.48521650e-01
3.71522009e-01 2.02696502e-01 -3.87409449e-01 9.11276996e-01
9.72984254e-01 -1.63974583e-01 -5.24612665e-01 -4.30777222e-02
1.08937895e+00 -2.92132556e-01 -6.96675181e-01 -3.80073249e-01
-8.97330284e-01 3.62277687e-01 7.34490156e-01 4.88350779e-01
-6.63958311e-01 5.99567056e-01 -8.08826029e-01 4.29532230e-01
2.51111835e-01 7.13285923e-01 9.92356567e-04 1.15645476e-01
-9.26752388e-01 -2.31620967e-01 9.76754487e-01 1.53248537e+00
4.45124865e-01 -6.19363040e-03 2.01975524e-01 4.83881772e-01
7.11099803e-01 9.11100268e-01 -3.75741929e-01 -1.07249117e+00
5.59509218e-01 -1.01406291e-01 6.60424829e-01 -6.20857716e-01
-8.60759020e-01 -1.60311973e+00 -8.48448098e-01 1.44226089e-01
2.86271393e-01 -9.38579559e-01 -6.26741529e-01 1.50590432e+00
1.52285516e-01 -5.27396142e-01 3.20718944e-01 5.04289985e-01
2.80185610e-01 7.11087465e-01 -2.27995217e-01 -9.20828998e-01
1.08353341e+00 -5.18099129e-01 -8.71803701e-01 -3.67706180e-01
1.64804593e-01 -8.00432146e-01 1.98813658e-02 2.36851826e-01
-1.27539527e+00 3.73950869e-01 -1.14038086e+00 5.06918311e-01
2.61545271e-01 -1.37118667e-01 7.51771092e-01 8.41501713e-01
-9.65478420e-01 -8.19757804e-02 -3.77553582e-01 -5.30284047e-01
7.45861888e-01 4.77892071e-01 2.10635141e-01 -5.24537027e-01
-4.64689851e-01 3.32619309e-01 -1.99564874e-01 3.24139223e-02
-6.18087500e-02 -6.35521531e-01 -2.59808630e-01 1.79758251e-01
1.01862326e-01 -1.14027286e+00 9.26491618e-01 -5.34202754e-02
-1.23207378e+00 7.26279020e-02 -2.69540638e-01 -3.63072067e-01
4.58492577e-01 4.60276455e-01 -2.56159604e-01 4.96596038e-01
8.98885056e-02 2.80719757e-01 5.02971351e-01 -1.33961606e+00
-8.78856122e-01 -3.48759592e-01 2.76146322e-01 1.31183580e-01
-4.55184460e-01 -3.58549356e-01 -1.08010314e-01 -2.42892474e-01
6.43258989e-01 -1.01388943e+00 -4.42757636e-01 7.02617690e-02
-7.54590690e-01 9.80508402e-02 8.12635064e-01 -6.50773421e-02
1.09820163e+00 -1.95807683e+00 2.50162154e-01 5.46145082e-01
3.20761293e-01 -2.52997547e-01 1.84075877e-01 7.46653914e-01
7.11912632e-01 2.06818640e-01 -9.43683647e-03 -4.30442065e-01
-1.85120836e-01 -1.08674496e-01 6.75907061e-02 4.52739984e-01
-5.59748650e-01 4.09019619e-01 -8.35037768e-01 -7.66828805e-02
-7.72405881e-03 2.47083560e-01 -4.57451791e-01 6.75790478e-03
-9.69214961e-02 6.45963311e-01 -7.92314053e-01 1.00381124e+00
6.74825072e-01 -4.34621453e-01 3.95321131e-01 -2.39968225e-01
-7.60438323e-01 -8.33410695e-02 -8.72684479e-01 1.28216028e+00
-8.81791592e-01 5.61417580e-01 6.82008862e-01 -3.87820661e-01
6.59617841e-01 5.34449339e-01 1.00513101e+00 -4.95628834e-01
-1.04998119e-01 6.68341279e-01 -8.14943686e-02 -5.78828037e-01
-8.77747014e-02 -5.89384556e-01 1.04049958e-01 1.67315885e-01
-1.97411329e-02 -2.36177251e-01 1.59344986e-01 -6.63206633e-03
1.27149236e+00 -4.10758704e-01 2.87255317e-01 -4.95219886e-01
-1.16567370e-02 -1.21030480e-01 5.74135005e-01 5.61922193e-01
3.55202332e-02 -5.23784421e-02 8.03928971e-02 -9.09993425e-02
-5.96355379e-01 -1.37115693e+00 -3.13146502e-01 7.86933824e-02
5.35071909e-01 7.19480589e-02 -1.39464006e-01 3.59689556e-02
1.37113810e-01 4.00170624e-01 3.74677181e-01 5.33047318e-01
-9.77804512e-02 -7.86452472e-01 1.54275820e-02 -2.14982018e-01
8.75250340e-01 3.70636843e-02 -6.05694592e-01 8.56017470e-01
-1.15526520e-01 -1.36263192e+00 -2.92477608e-02 1.33923050e-02
-6.16364837e-01 -8.53319883e-01 -8.66723061e-01 -7.05277324e-01
6.64777696e-01 7.01592565e-01 7.46150553e-01 -8.27425346e-02
-3.39231193e-01 9.70453084e-01 4.99945767e-02 -3.25213969e-01
4.56681073e-01 1.16722845e-01 3.04295003e-01 1.03106059e-01
-4.08628047e-01 -1.22997141e+00 -9.60025549e-01 4.95928437e-01
-1.91483602e-01 1.98460788e-01 7.59799361e-01 1.21444300e-01
3.68177772e-01 4.36736465e-01 1.18968296e+00 -2.95530081e-01
2.64834166e-01 -8.35186541e-01 -9.68474925e-01 -8.34486913e-04
-2.22953796e-01 -5.21743834e-01 3.47961962e-01 4.34201002e-01
-9.05140519e-01 -1.56641543e-01 8.59719366e-02 6.54346764e-01
3.21112841e-01 5.90849459e-01 -3.10454279e-01 -8.52383733e-01
1.07022166e-01 -3.31536889e-01 -2.30315924e-01 -6.68083504e-02
2.57357687e-01 9.86409783e-01 -3.06681842e-02 -1.27785981e-01
1.05325484e+00 8.69244874e-01 7.33254731e-01 -1.69092417e+00
-6.63108349e-01 -2.95501590e-01 6.45166785e-02 -6.90845549e-01
7.88567901e-01 -9.41477895e-01 -6.25550807e-01 -1.17922142e-01
-1.07988107e+00 -2.44831085e-01 2.86908541e-02 8.55288506e-01
-4.67379391e-01 -1.42411575e-01 -2.13605568e-01 -1.31599820e+00
-6.43235892e-02 -5.91666400e-01 3.10366690e-01 5.60685158e-01
-7.52684996e-02 -7.86697090e-01 -4.17827070e-01 5.12470245e-01
1.06879842e+00 4.93870676e-01 9.59303200e-01 7.50236809e-01
-1.53553891e+00 -5.55578843e-02 -3.58639866e-01 -4.94774491e-01
5.29273033e-01 -4.92920578e-01 -5.48994541e-01 -4.93645906e-01
-2.67095834e-01 6.26929551e-02 5.63982427e-01 9.36565936e-01
6.89382434e-01 -4.55741882e-01 -7.88881183e-01 7.09838867e-01
1.77313495e+00 2.10557938e-01 5.47713101e-01 -7.59444907e-02
1.02057971e-01 4.50465232e-01 1.91189826e-01 5.31243801e-01
3.63428116e-01 4.74711657e-01 9.37233925e-01 1.19868696e-01
-2.54397333e-01 3.69934708e-01 -3.14566463e-01 4.69842434e-01
-1.51184052e-01 -1.12783468e+00 -2.71680027e-01 4.86595541e-01
-1.36563516e+00 -8.43101263e-01 -8.32231790e-02 2.20304036e+00
3.57083529e-02 1.63852692e-01 -7.13968463e-03 -1.27013966e-01
3.61059934e-01 -3.68041880e-02 -3.88526618e-01 4.54921633e-01
-1.63422987e-01 7.98684731e-02 1.04660428e+00 7.56692827e-01
-6.12935424e-01 1.85944930e-01 5.30627060e+00 4.01603162e-01
-8.37706447e-01 -3.60049047e-02 2.20069334e-01 -3.70787650e-01
-8.00017536e-01 -3.09287131e-01 -7.58054674e-01 3.55099112e-01
3.38453561e-01 -2.26534829e-01 4.11878228e-01 2.60322187e-02
5.97300887e-01 -4.49591100e-01 -4.81123984e-01 1.05930853e+00
-4.84431624e-01 -1.78786552e+00 -3.04586619e-01 7.56497383e-01
9.25649703e-01 -5.70542365e-02 9.41935480e-02 -4.87444729e-01
6.47869110e-02 -6.41837239e-01 2.91556954e-01 6.83347106e-01
8.60811174e-01 -6.21843398e-01 2.33755127e-01 1.97897509e-01
-1.22510922e+00 -5.10393083e-01 -6.33615628e-02 -1.07084170e-01
8.34527493e-01 1.39926898e+00 -3.83798391e-01 7.93895781e-01
2.02712819e-01 3.64680499e-01 6.30896270e-01 1.77949083e+00
-1.11109316e-01 3.61161888e-01 -7.77581692e-01 -4.49121267e-01
7.73541182e-02 -5.00663936e-01 1.04005516e+00 5.43031633e-01
1.02563596e+00 2.49015689e-01 -1.38704225e-01 6.05969429e-01
-4.35620725e-01 -7.62457475e-02 -8.30461979e-01 -3.39416750e-02
1.09031284e+00 1.25355625e+00 -7.20753133e-01 5.36303997e-01
-6.08679175e-01 5.27770936e-01 -1.11659206e-01 8.32487464e-01
-2.11618468e-01 -6.94875360e-01 1.04281318e+00 3.05984914e-01
1.82483748e-01 -1.15067124e+00 -5.60206890e-01 -5.92544317e-01
7.63735995e-02 1.41498253e-01 -5.78025207e-02 -4.16481256e-01
-9.78819966e-01 2.30339035e-01 -5.29546440e-01 -1.27985549e+00
2.72667289e-01 -3.08493853e-01 -4.74740952e-01 8.50165725e-01
-1.93660021e+00 -9.12664115e-01 -3.68651837e-01 3.41541469e-02
1.24240905e-01 -9.31873620e-02 7.54240692e-01 5.12878895e-01
-3.08269083e-01 4.03461307e-02 3.53109747e-01 -4.51147348e-01
1.09458297e-01 -8.48718584e-01 -6.58130229e-01 7.69962549e-01
-2.67368287e-01 6.01741850e-01 7.84907818e-01 -4.85986859e-01
-1.89446068e+00 -9.99433994e-01 7.44324148e-01 3.32063586e-01
4.72954720e-01 -5.06723858e-02 3.06943148e-01 3.68361235e-01
3.13566715e-01 -1.97914749e-01 1.08938241e+00 -2.31738195e-01
4.71413761e-01 -3.75203341e-01 -1.39872634e+00 7.30391204e-01
1.40888095e+00 1.34527892e-01 5.64592302e-01 7.40148723e-01
3.42745185e-01 -1.87554061e-01 -6.80621088e-01 1.85513750e-01
6.68281138e-01 -6.49061322e-01 9.19271410e-01 -9.78843272e-02
3.47623564e-02 -3.28618288e-01 -6.56333625e-01 -1.41095817e+00
-4.94380832e-01 -1.32199407e+00 -1.74724340e-01 1.06361198e+00
7.07746804e-01 -9.29353893e-01 9.33857739e-01 2.16210157e-01
9.15127769e-02 -8.94162059e-01 -1.52402759e+00 -1.08275962e+00
-4.81478781e-01 -3.80166732e-02 3.09377193e-01 5.57103634e-01
1.10255338e-01 7.56007135e-01 -3.69949460e-01 8.59217107e-01
1.20625246e+00 -1.79809369e-02 5.62521875e-01 -1.32181227e+00
-2.50661463e-01 -4.65917945e-01 -2.55519331e-01 -1.67534494e+00
-3.52676779e-01 -7.43722856e-01 -2.78905541e-01 -2.26778793e+00
-6.56908751e-01 -1.30819559e+00 2.57572174e-01 -2.39155218e-01
6.99569762e-01 6.07413054e-01 1.59493178e-01 4.61671688e-02
-5.05380630e-02 4.97907758e-01 1.39046168e+00 -1.13111027e-01
-1.73117310e-01 7.70294785e-01 -7.84416676e-01 4.83593911e-01
9.94094372e-01 1.33506507e-01 -7.32136786e-01 -6.12178564e-01
5.29709101e-01 4.59804326e-01 1.32357836e-01 -1.57808828e+00
2.69803077e-01 -4.37861502e-01 1.16492949e-01 -1.52819589e-01
7.85734475e-01 -1.31681728e+00 2.53472090e-01 5.47477305e-01
2.95076787e-01 -6.46498978e-01 -3.12277764e-01 9.99747753e-01
3.86163980e-01 -6.69765472e-02 9.05362725e-01 -3.26862074e-02
-5.33803776e-02 5.15304148e-01 -8.40027034e-01 -3.62139642e-01
1.33151925e+00 -1.36441737e-01 -5.33896863e-01 -8.61329377e-01
-7.63619721e-01 5.72611034e-01 -3.16134579e-02 -3.50272119e-01
2.48474717e-01 -1.16264451e+00 -2.10294485e-01 -1.48554847e-01
-2.01473236e-01 -4.39217746e-01 6.64711595e-02 8.11136901e-01
-4.41930890e-01 7.70617247e-01 -4.87759113e-02 -1.20985694e-01
-1.02191412e+00 -4.16606992e-01 7.02337861e-01 2.59298921e-01
-4.32177216e-01 1.12560439e+00 -4.81279641e-01 4.44014579e-01
1.20982617e-01 7.55345356e-03 1.64718702e-01 1.16714239e-02
2.62988120e-01 3.93592298e-01 5.37638851e-02 7.57566541e-02
-1.85550839e-01 5.74914575e-01 5.79026103e-01 -2.98505872e-01
1.39835668e+00 -7.31402278e-01 -3.89302403e-01 -1.10352106e-01
6.16498351e-01 6.32427931e-01 -9.75634992e-01 -2.17432603e-02
-4.20893312e-01 -6.80306792e-01 4.27928209e-01 -6.24805570e-01
-8.88589263e-01 2.79836386e-01 2.63257176e-01 8.02725136e-01
8.39326203e-01 9.75532904e-02 7.49816120e-01 6.71541333e-01
1.35509479e+00 -9.00359809e-01 3.66179571e-02 2.07112581e-02
5.50087690e-01 -5.24445653e-01 3.85172367e-02 -1.09729350e+00
7.22089887e-01 1.18089902e+00 3.44239026e-01 4.30433422e-01
8.06033254e-01 1.62902921e-01 -3.79928291e-01 -3.36430341e-01
-2.51861721e-01 -4.30880725e-01 -4.80876803e-01 5.64028680e-01
4.31013376e-01 2.64777929e-01 -5.03079534e-01 -1.35777825e-02
-2.30737299e-01 -2.27536857e-01 7.64970899e-01 1.15674031e+00
-1.15109968e+00 -1.27351272e+00 -4.53085229e-02 9.76423323e-01
-2.60365933e-01 1.71380490e-01 2.89116859e-01 5.88696480e-01
4.89577353e-02 1.29488337e+00 -1.39740437e-01 7.03858808e-02
1.51585281e-01 -4.00251359e-01 8.16946507e-01 -5.74324369e-01
7.98057556e-01 -6.74029365e-02 6.74006402e-01 -5.16884681e-03
-3.94938797e-01 -3.56062323e-01 -1.12877512e+00 -3.34151946e-02
-1.69380575e-01 3.47368240e-01 8.18015873e-01 1.06669044e+00
2.90845513e-01 5.76842427e-01 9.29716945e-01 -5.00121355e-01
6.47276193e-02 -2.18966037e-01 -1.00348878e+00 -8.01474690e-01
7.35307038e-01 -7.21117914e-01 -2.95958489e-01 -5.86961091e-01] | [6.20794153213501, 1.3018405437469482] |
9d911968-04ba-40ca-8409-b16a9548fa86 | technical-report-disentangled-action-parsing | 2111.03225 | null | https://arxiv.org/abs/2111.03225v1 | https://arxiv.org/pdf/2111.03225v1.pdf | Technical Report: Disentangled Action Parsing Networks for Accurate Part-level Action Parsing | Part-level Action Parsing aims at part state parsing for boosting action recognition in videos. Despite of dramatic progresses in the area of video classification research, a severe problem faced by the community is that the detailed understanding of human actions is ignored. Our motivation is that parsing human actions needs to build models that focus on the specific problem. We present a simple yet effective approach, named disentangled action parsing (DAP). Specifically, we divided the part-level action parsing into three stages: 1) person detection, where a person detector is adopted to detect all persons from videos as well as performs instance-level action recognition; 2) Part parsing, where a part-parsing model is proposed to recognize human parts from detected person images; and 3) Action parsing, where a multi-modal action parsing network is used to parse action category conditioning on all detection results that are obtained from previous stages. With these three major models applied, our approach of DAP records a global mean of $0.605$ score in 2021 Kinetics-TPS Challenge. | ['Jingkuan Song', 'Lechao Chen', 'Lianli Gao', 'Xiaojia Chen', 'Xuanhan Wang'] | 2021-11-05 | null | null | null | null | ['action-parsing'] | ['natural-language-processing'] | [ 7.39233792e-01 4.12778527e-01 -4.43499744e-01 -3.11442584e-01
-1.07594085e+00 -4.01065618e-01 6.64273083e-01 -2.51234502e-01
-4.60701317e-01 4.66590494e-01 6.66409016e-01 1.03324495e-01
4.29773808e-01 -5.81309557e-01 -6.91316605e-01 -7.12547839e-01
-2.00452536e-01 3.13370228e-01 3.85848343e-01 1.73770398e-01
-1.84934214e-02 3.72922450e-01 -1.53237796e+00 7.96348691e-01
2.10976735e-01 8.07402670e-01 -2.58377135e-01 1.22457159e+00
1.09002896e-01 1.46321023e+00 -5.81933558e-01 -4.56951588e-01
3.57751340e-01 -8.67511988e-01 -9.74171638e-01 7.02176571e-01
3.81594121e-01 -8.24993610e-01 -3.80512327e-01 9.66375113e-01
2.88185537e-01 -1.02227993e-01 3.09922785e-01 -1.28666604e+00
-3.78855481e-03 6.56388700e-01 -6.36865258e-01 3.40574294e-01
6.95516288e-01 3.93313617e-01 1.07113361e+00 -6.45587087e-01
5.31187475e-01 1.48429835e+00 4.19568866e-01 9.39938426e-01
-1.07664657e+00 -5.96426725e-01 4.81971562e-01 2.39431083e-01
-6.63640440e-01 -4.40271378e-01 5.65859735e-01 -5.11397362e-01
1.12376797e+00 1.71720132e-01 8.07097793e-01 1.26718867e+00
7.83871636e-02 1.42616868e+00 8.75633657e-01 -4.07788932e-01
2.71423578e-01 -1.62397191e-01 4.67336565e-01 8.30360174e-01
2.44024903e-01 -9.62136984e-02 -6.94578052e-01 -7.52637163e-02
6.00254357e-01 -2.48151589e-02 5.79971299e-02 -4.79465365e-01
-1.22818053e+00 6.54340744e-01 1.33702001e-02 4.62960973e-02
-7.25322723e-01 3.14982057e-01 6.36949003e-01 -1.79604605e-01
3.19553614e-01 1.93779171e-02 -3.17980647e-01 -4.05936897e-01
-8.89348686e-01 2.82374680e-01 6.83290720e-01 6.27574742e-01
4.15066123e-01 -1.63894877e-01 -4.11675215e-01 3.78285944e-01
1.35147616e-01 4.60566312e-01 9.42257866e-02 -1.26466715e+00
6.79076850e-01 8.41184735e-01 1.48686588e-01 -5.64348459e-01
-2.87734032e-01 1.27811819e-01 -4.57705975e-01 3.82563621e-01
5.59057713e-01 -3.72554064e-01 -9.27615345e-01 1.72143126e+00
3.81005645e-01 1.36332378e-01 1.47987589e-01 7.57961333e-01
6.69564128e-01 3.75649303e-01 6.58261716e-01 2.90384144e-02
1.67807817e+00 -1.22468066e+00 -4.36809629e-01 -6.79588258e-01
7.12331176e-01 -2.33290762e-01 3.81976783e-01 4.01463211e-01
-1.37082994e+00 -6.19808018e-01 -9.26419795e-01 7.32110590e-02
-1.59549356e-01 3.63905996e-01 7.08006859e-01 9.09110308e-01
-8.33731890e-01 3.03426236e-01 -1.15492225e+00 -5.76967478e-01
8.08180511e-01 3.49112779e-01 -6.16937280e-01 -1.87171072e-01
-8.99169922e-01 7.47238636e-01 3.66727561e-01 -2.08190791e-02
-1.17133331e+00 -1.44300565e-01 -9.37603712e-01 1.24017432e-01
6.91537142e-01 -7.52949357e-01 1.32513177e+00 -1.29218948e+00
-1.17108536e+00 9.97915328e-01 -1.83009177e-01 -7.22124159e-01
5.93824148e-01 -5.41094065e-01 -1.12315319e-01 6.88770175e-01
1.69193596e-01 9.66992736e-01 9.21956480e-01 -9.26061392e-01
-1.12405860e+00 -5.72880626e-01 3.23587507e-01 4.66591924e-01
2.39687622e-01 5.10984004e-01 -6.16790056e-01 -3.01820189e-01
1.62561554e-02 -8.88106823e-01 -3.23951185e-01 -2.15696439e-01
-3.49995434e-01 -3.18863064e-01 6.52769685e-01 -1.22349656e+00
8.17056596e-01 -1.98205972e+00 2.59661138e-01 -4.23159748e-01
2.83628643e-01 2.31493622e-01 -2.49160498e-01 3.01242471e-01
-3.04484874e-01 -3.00892517e-02 7.28529785e-03 -5.34514248e-01
-1.62264749e-01 7.04019442e-02 -2.07213998e-01 4.02333230e-01
5.90331793e-01 9.27168846e-01 -9.46270823e-01 -5.58762074e-01
3.53255868e-01 2.18769774e-01 -8.07250679e-01 2.43582010e-01
-9.05940309e-02 5.49921453e-01 -7.31373489e-01 9.87382472e-01
3.76735657e-01 -1.95453763e-01 4.01483685e-01 4.19616140e-02
2.57768221e-02 3.17705572e-01 -1.14166057e+00 1.68843651e+00
3.93641472e-01 3.32262963e-01 1.71184555e-01 -1.22845745e+00
1.48656815e-01 3.80763769e-01 9.87279654e-01 -4.63750094e-01
1.41675055e-01 -4.00400460e-01 -1.00205027e-01 -6.08659327e-01
2.31027424e-01 2.02286802e-02 -3.15203488e-01 4.32209820e-01
1.67538390e-01 4.28373903e-01 4.48920697e-01 5.10059059e-01
1.68063903e+00 7.24289000e-01 5.34083009e-01 2.66953588e-01
7.53966033e-01 3.12099695e-01 7.70267725e-01 8.12217236e-01
-8.80936384e-01 4.23220724e-01 9.47113931e-01 -5.58393955e-01
-7.70977914e-01 -9.55532968e-01 5.09366095e-01 1.35389292e+00
4.15784791e-02 -3.12583953e-01 -1.19912565e+00 -1.15558541e+00
-2.29187876e-01 4.17622149e-01 -6.83368146e-01 4.33103107e-02
-1.00876939e+00 -7.03881383e-01 6.40839636e-01 9.66170788e-01
8.69952738e-01 -1.35314691e+00 -1.08197570e+00 7.71642849e-02
-3.84160131e-01 -1.31385863e+00 -1.74409658e-01 5.11199534e-02
-8.11307549e-01 -1.49337244e+00 -7.21442640e-01 -5.04821658e-01
5.10072470e-01 3.32321495e-01 1.22644532e+00 -1.18607089e-01
-5.34522772e-01 9.39229071e-01 -5.63068271e-01 -2.72429466e-01
-4.73284453e-01 -4.67681080e-01 -4.24210392e-02 1.75172925e-01
7.30655789e-01 -1.61045287e-02 -7.68775702e-01 1.65852368e-01
-4.35628414e-01 2.83197790e-01 1.10900569e+00 4.19546664e-01
3.62657338e-01 -4.56625223e-02 1.45234957e-01 -7.67789125e-01
-1.23202950e-01 -8.87047350e-02 -1.72588363e-01 4.34716612e-01
-1.01327877e-02 -2.32786894e-01 7.37072751e-02 -2.72129476e-01
-1.30400717e+00 8.46201837e-01 -1.09494992e-01 -7.66654685e-02
-6.85571611e-01 -1.91686496e-01 -5.67656159e-01 3.84375066e-01
4.87334490e-01 4.91132975e-01 3.55919488e-02 -2.53159821e-01
3.80614787e-01 4.69644636e-01 5.64275146e-01 -3.59691054e-01
4.25556719e-01 6.60889089e-01 -1.21994533e-01 -8.67967069e-01
-1.06333506e+00 -6.51237190e-01 -9.06877995e-01 -5.82441092e-01
1.78750587e+00 -1.30895615e+00 -7.06061184e-01 6.59358859e-01
-1.13177633e+00 -3.90458465e-01 -3.37061912e-01 4.34979409e-01
-7.02319086e-01 6.21124983e-01 -7.52428114e-01 -1.05648816e+00
-1.11669175e-01 -1.01730406e+00 1.37558544e+00 1.45277098e-01
-2.34696269e-01 -4.97936606e-01 -3.23063582e-02 1.13291216e+00
-2.44686142e-01 3.07876348e-01 4.58693653e-01 -7.48152614e-01
-8.82135212e-01 -5.56249797e-01 -1.52937755e-01 6.50556207e-01
-1.73864916e-01 -3.49552780e-01 -9.85832691e-01 -9.49401334e-02
1.51443435e-02 -4.75230575e-01 8.62586558e-01 5.35157382e-01
8.90430927e-01 5.57735227e-02 -4.52852666e-01 2.56877188e-02
8.45529795e-01 2.22042948e-01 7.24868119e-01 8.92713293e-02
5.83495319e-01 6.67029083e-01 6.24396443e-01 2.18235761e-01
2.07992703e-01 4.96433347e-01 3.59505951e-01 -7.20767900e-02
-3.97369444e-01 -3.59222233e-01 8.04061115e-01 -2.54705967e-03
-4.36351687e-01 -1.48007080e-01 -6.74328983e-01 1.64171204e-01
-1.97815895e+00 -1.48788035e+00 -9.06868875e-02 1.99308562e+00
1.34702563e-01 4.22789067e-01 7.05633938e-01 1.82376549e-01
7.77542233e-01 3.74070287e-01 -4.64300007e-01 2.82850601e-02
1.27880320e-01 -1.83856279e-01 5.13547540e-01 4.98090349e-02
-1.61395586e+00 8.76877427e-01 6.75663900e+00 2.98602313e-01
-3.30513000e-01 1.52178062e-02 5.90207040e-01 -4.21717688e-02
6.25675559e-01 1.15169995e-01 -9.62748587e-01 2.92346895e-01
7.06213653e-01 2.75481969e-01 6.13441095e-02 1.06367934e+00
2.10501254e-01 -5.36426008e-01 -1.32740593e+00 8.48918974e-01
4.03911799e-01 -9.71777260e-01 1.75926104e-01 1.22608490e-01
2.18179777e-01 -2.19502851e-01 -4.29121107e-01 7.04907179e-01
3.10196251e-01 -5.95449865e-01 7.07660139e-01 4.23527926e-01
3.10166061e-01 -5.28588235e-01 5.21314442e-01 7.04761863e-01
-1.33826959e+00 -3.59113336e-01 7.97628239e-02 -3.14266294e-01
4.21228260e-01 5.10455929e-02 -4.88794982e-01 4.25559640e-01
6.18709385e-01 8.35523248e-01 -4.78146315e-01 6.48286819e-01
-5.39510369e-01 7.84976542e-01 2.19007730e-01 2.61882901e-01
1.06138043e-01 -6.59468845e-02 4.72613305e-01 1.27520001e+00
-2.22560883e-01 5.10737360e-01 5.28929591e-01 3.84538054e-01
1.54359519e-01 -3.21733803e-01 -3.02119255e-01 -2.43588269e-01
-1.39800012e-01 1.22445142e+00 -1.08701146e+00 -7.41378725e-01
-8.54807198e-01 1.29764736e+00 8.99992585e-02 1.64658964e-01
-1.01535249e+00 1.97099149e-01 7.02999771e-01 3.67794596e-02
4.70524937e-01 -1.07301407e-01 -4.49407510e-02 -1.25319195e+00
-1.75429061e-01 -1.02979255e+00 8.94844532e-01 -8.13214779e-01
-8.41718614e-01 2.18665719e-01 1.86203271e-01 -1.07253337e+00
-3.12367380e-01 -7.52446473e-01 -4.49739069e-01 4.52456713e-01
-9.52723801e-01 -1.23414016e+00 -2.50318080e-01 5.10269105e-01
1.14941370e+00 -1.57814417e-02 7.04771817e-01 2.23263413e-01
-9.25901711e-01 1.23852640e-01 -8.18197548e-01 7.39292860e-01
1.97636858e-01 -1.14478981e+00 4.62692559e-01 1.38223600e+00
2.61692464e-01 3.03381383e-01 6.45774961e-01 -9.33935583e-01
-1.39070988e+00 -8.31396520e-01 9.16719615e-01 -9.35261607e-01
4.94569391e-01 -3.39838117e-01 -3.61891657e-01 9.74691212e-01
8.92762244e-02 -7.74283782e-02 6.53308749e-01 -6.18451089e-02
-2.78297782e-01 1.98192656e-01 -9.94843483e-01 4.74797785e-01
1.17681289e+00 -2.77919501e-01 -8.74868214e-01 3.91824961e-01
3.54457915e-01 -3.72441143e-01 -3.30704302e-01 2.20360816e-01
6.77702427e-01 -1.18091643e+00 1.16407645e+00 -1.08862889e+00
5.02544284e-01 -8.43575820e-02 -1.71539292e-01 -4.48584795e-01
-4.22193348e-01 -2.28497550e-01 -3.91620040e-01 9.53906298e-01
7.74574131e-02 -2.33177468e-01 1.22773254e+00 1.00156105e+00
4.35563885e-02 -4.19001073e-01 -7.68372595e-01 -6.01478875e-01
-3.82517874e-01 -4.76049781e-01 -7.33798221e-02 4.02982086e-01
-2.58205738e-02 4.47884798e-01 -5.78319192e-01 2.76483178e-01
7.84857452e-01 -8.18101615e-02 1.02412665e+00 -9.14416850e-01
-5.82745731e-01 -2.31205583e-01 -5.96625686e-01 -1.25070214e+00
4.16274220e-02 -3.72238308e-01 1.96077958e-01 -1.71153569e+00
8.97222102e-01 4.77632284e-01 -2.50892282e-01 6.34157896e-01
-2.20245034e-01 1.47985399e-01 3.07589710e-01 1.44923851e-01
-1.04607439e+00 9.06674564e-02 9.18994188e-01 -4.98557054e-02
8.65589157e-02 3.58113080e-01 -6.22437596e-01 1.13422608e+00
4.23912227e-01 -5.42852581e-01 -1.98368996e-01 -9.27366838e-02
-2.13215366e-01 4.43756968e-01 8.10854793e-01 -1.29648161e+00
5.22152856e-02 -6.38843188e-03 6.13748968e-01 -8.27119529e-01
5.57048202e-01 -6.51332617e-01 -9.90693569e-02 9.65097845e-01
-3.18712413e-01 -1.61873192e-01 -1.63248897e-01 7.81637490e-01
-3.47448215e-02 -2.44756695e-02 6.48966432e-01 -5.18489897e-01
-1.16599119e+00 1.11557499e-01 -6.88044846e-01 -4.48502265e-02
1.34576237e+00 -4.12580341e-01 -2.26947948e-01 -3.17863286e-01
-1.13690722e+00 1.98324680e-01 1.11381024e-01 3.84443671e-01
2.60284156e-01 -9.51125026e-01 -7.47446477e-01 -4.23970930e-02
4.03326117e-02 -3.97155464e-01 5.45104504e-01 8.65116954e-01
-2.85158217e-01 4.87611681e-01 -1.75061449e-01 -6.10359967e-01
-1.77450204e+00 6.70827448e-01 2.61804938e-01 -4.50566888e-01
-9.30931032e-01 9.79280472e-01 2.78431296e-01 6.88419268e-02
4.82020020e-01 -2.45448142e-01 -3.54949445e-01 6.54103979e-02
9.01566684e-01 6.23103619e-01 -4.40799296e-01 -9.47154522e-01
-7.02012539e-01 2.52688169e-01 1.11601226e-01 -2.18660176e-01
1.20526266e+00 1.24580838e-01 2.68239439e-01 8.81211609e-02
6.85196698e-01 -3.03795397e-01 -1.78022540e+00 7.00155124e-02
-4.57365485e-03 -2.54449487e-01 -2.73097873e-01 -8.35830688e-01
-8.42978001e-01 7.65124440e-01 6.65441930e-01 2.20241994e-01
1.17092693e+00 2.22892284e-01 6.45180345e-01 1.97511464e-01
4.94165033e-01 -1.22798324e+00 3.87229770e-01 2.21835718e-01
5.46120882e-01 -1.28152668e+00 2.50201464e-01 -5.38653553e-01
-9.61042821e-01 7.98879921e-01 6.24379396e-01 -1.56565070e-01
2.26342037e-01 8.02844241e-02 -9.08548981e-02 -2.65765101e-01
-5.68834126e-01 -5.04750907e-01 1.90201923e-01 4.49445397e-01
2.23674715e-01 3.90330628e-02 -2.37022832e-01 6.50595307e-01
4.63858724e-01 1.37720183e-01 3.19888204e-01 1.13871467e+00
-7.15111613e-01 -1.06883061e+00 -3.17854762e-01 3.46453190e-01
-6.53831720e-01 3.06145012e-01 -7.39612281e-01 9.00069237e-01
3.28766435e-01 1.06553996e+00 -9.38412249e-02 -2.88285613e-01
4.14920181e-01 4.96804625e-01 5.81870615e-01 -7.81323373e-01
-4.44513232e-01 -6.92738146e-02 3.64358008e-01 -1.16954303e+00
-8.71553838e-01 -1.20833421e+00 -1.28077328e+00 2.18025059e-01
1.48393407e-01 -2.10979298e-01 4.41266000e-01 1.17591906e+00
9.69824046e-02 4.22974348e-01 2.83921629e-01 -1.18040717e+00
-6.09787703e-01 -7.16836512e-01 -4.94500369e-01 5.04702628e-01
1.40055522e-01 -5.75507343e-01 -3.35058808e-01 3.93776119e-01] | [8.33236312866211, 0.47201403975486755] |
2314a065-bc25-416f-aa78-a628c54e8de1 | document-level-relation-extraction-with-2 | 2201.04826 | null | https://arxiv.org/abs/2201.04826v1 | https://arxiv.org/pdf/2201.04826v1.pdf | Document-level Relation Extraction with Context Guided Mention Integration and Inter-pair Reasoning | Document-level Relation Extraction (DRE) aims to recognize the relations between two entities. The entity may correspond to multiple mentions that span beyond sentence boundary. Few previous studies have investigated the mention integration, which may be problematic because coreferential mentions do not equally contribute to a specific relation. Moreover, prior efforts mainly focus on reasoning at entity-level rather than capturing the global interactions between entity pairs. In this paper, we propose two novel techniques, Context Guided Mention Integration and Inter-pair Reasoning (CGM2IR), to improve the DRE. Instead of simply applying average pooling, the contexts are utilized to guide the integration of coreferential mentions in a weighted sum manner. Additionally, inter-pair reasoning executes an iterative algorithm on the entity pair graph, so as to model the interdependency of relations. We evaluate our CGM2IR model on three widely used benchmark datasets, namely DocRED, CDR, and GDA. Experimental results show that our model outperforms previous state-of-the-art models. | ['Jianhua Dai', 'Lu Xu', 'Daojian Zeng', 'Chao Zhao'] | 2022-01-13 | null | null | null | null | ['document-level-relation-extraction'] | ['natural-language-processing'] | [-1.60333980e-02 5.19819796e-01 -3.64306450e-01 -3.73796433e-01
-8.84410501e-01 -6.41304135e-01 5.47190964e-01 7.63654411e-01
-3.18306714e-01 7.63847768e-01 5.19981921e-01 -4.04977173e-01
-1.97996929e-01 -8.92350852e-01 -3.65471244e-01 -3.06288987e-01
4.67498899e-02 3.69118154e-01 5.31412959e-01 -4.37181920e-01
1.77206025e-01 1.94634855e-01 -8.83665681e-01 6.33954883e-01
1.16686320e+00 5.86124897e-01 -1.36090457e-01 7.24883601e-02
-5.13206840e-01 1.05063891e+00 -6.88650191e-01 -1.09999681e+00
-1.42638996e-01 -3.41976106e-01 -1.25041795e+00 -3.83704334e-01
7.15228990e-02 1.80014640e-01 -1.18791662e-01 1.10017896e+00
2.45255858e-01 1.69945449e-01 7.79438317e-01 -8.99565756e-01
-7.68028378e-01 1.15397763e+00 -7.62901545e-01 3.07866842e-01
4.92417842e-01 -6.24501824e-01 1.69746447e+00 -9.69783783e-01
8.85342479e-01 1.13411272e+00 5.94596744e-01 1.60778940e-01
-8.70743990e-01 -4.17859674e-01 4.94919002e-01 4.06219423e-01
-1.44355559e+00 -2.31139317e-01 6.97093904e-01 -3.73338521e-01
1.48010302e+00 3.54716361e-01 4.89658974e-02 4.22041148e-01
-3.85662317e-02 6.30064964e-01 6.84163451e-01 -6.55757189e-01
-1.32898003e-01 -9.69155133e-02 7.97235489e-01 2.83572167e-01
5.34003854e-01 -5.43256760e-01 -3.62987161e-01 -1.32695973e-01
2.31359959e-01 -3.52216184e-01 -3.99977356e-01 1.82324871e-01
-1.03257596e+00 6.38400018e-01 6.40499830e-01 6.38547599e-01
-4.98604327e-01 -3.93145561e-01 4.40601975e-01 1.49832696e-01
6.25181615e-01 5.89173198e-01 -7.14751184e-01 1.20000765e-01
-5.61426699e-01 2.32357889e-01 1.04835904e+00 1.15713549e+00
7.55091012e-01 -9.65301752e-01 -3.36985081e-01 9.50506747e-01
4.10704762e-01 -1.10216126e-01 1.78565189e-01 -5.47489583e-01
1.06340158e+00 1.21255124e+00 2.76547354e-02 -1.21209145e+00
-4.14155006e-01 -2.64953315e-01 -7.41346121e-01 -5.62367558e-01
2.33145237e-01 -2.58237869e-01 -4.61825430e-01 1.52173305e+00
5.46362519e-01 1.37462288e-01 4.54328537e-01 7.33201385e-01
1.43505740e+00 4.66551185e-01 3.21674407e-01 -2.78993309e-01
1.64852715e+00 -1.10097432e+00 -1.08801830e+00 -3.62108916e-01
1.07587075e+00 -8.79422903e-01 4.38188255e-01 -2.79877722e-01
-9.97288108e-01 -8.74657258e-02 -8.79544914e-01 -3.18788528e-01
-5.71432173e-01 1.60187095e-01 6.68660760e-01 1.54726639e-01
-5.23181856e-01 4.76186156e-01 -6.36867583e-01 -2.95586526e-01
-8.91734287e-02 1.14632115e-01 -4.59156632e-01 1.16993792e-01
-1.81439328e+00 1.29741108e+00 6.41949415e-01 4.24664289e-01
1.87478602e-01 -6.47903025e-01 -1.05428314e+00 2.45059267e-01
6.34450555e-01 -5.60135484e-01 1.14180446e+00 -1.82795897e-01
-8.12597573e-01 9.30047631e-01 -5.70959628e-01 -3.95663083e-01
-1.02949366e-01 -3.62269461e-01 -6.99925303e-01 -1.17186412e-01
2.41030440e-01 8.68664086e-02 -2.67291337e-01 -1.31709790e+00
-7.34423339e-01 -2.56916940e-01 4.74057347e-01 4.79164928e-01
-5.55652454e-02 5.04350424e-01 -7.17441678e-01 -5.19233048e-01
4.21956897e-01 -5.90462685e-01 -1.16051897e-01 -8.44642937e-01
-8.98828566e-01 -7.53224015e-01 5.86422563e-01 -7.63487160e-01
1.98086607e+00 -1.90787387e+00 1.41063035e-01 8.94110873e-02
1.97818696e-01 3.60550851e-01 1.22352153e-01 8.85721564e-01
-2.35221133e-01 4.45895553e-01 -2.02557549e-01 -2.16711268e-01
-3.74509953e-02 1.82250276e-01 -2.76049078e-01 -1.21386342e-01
3.53930682e-01 1.01941717e+00 -9.82269168e-01 -7.44436860e-01
-3.84556144e-01 3.14903319e-01 -2.46456474e-01 1.69405475e-01
-1.51849225e-01 4.22062539e-02 -5.76094687e-01 5.52739859e-01
7.07420051e-01 -2.85338461e-01 7.97147751e-01 -6.28213286e-01
1.02033569e-02 1.05905461e+00 -1.03305411e+00 1.46244109e+00
-3.73330653e-01 1.63351938e-01 -3.03959727e-01 -8.18794668e-01
1.03719950e+00 3.87032539e-01 3.22580785e-01 -4.47310776e-01
-4.04253192e-02 2.17301607e-01 2.59624440e-02 -6.08289659e-01
7.52137125e-01 1.28984243e-01 -2.79987663e-01 1.47815153e-01
-1.01767424e-02 3.53989005e-01 6.22202158e-01 5.80924094e-01
1.21490240e+00 2.36962065e-01 7.88884699e-01 2.50192583e-02
6.73981428e-01 1.65834569e-03 1.05636823e+00 2.99156547e-01
1.08066864e-01 4.23950404e-01 9.09338117e-01 2.09218897e-02
-3.98918033e-01 -9.43631411e-01 -2.53422391e-02 7.47240365e-01
5.16821802e-01 -8.42428923e-01 -5.91860831e-01 -1.23399019e+00
-2.52590448e-01 8.58633339e-01 -5.81671357e-01 3.77041735e-02
-9.73035097e-01 -9.18528914e-01 4.35533464e-01 6.72494173e-01
6.08199179e-01 -1.05850303e+00 -5.24604581e-02 4.49337155e-01
-8.28181326e-01 -1.44329441e+00 -5.17582655e-01 9.14691389e-02
-5.30882299e-01 -1.28657269e+00 -9.82491225e-02 -9.34486508e-01
5.90071321e-01 1.62404746e-01 1.43887353e+00 4.19742376e-01
2.04865679e-01 -1.98214591e-01 -6.58795893e-01 7.18422979e-02
1.31784547e-02 4.47054058e-01 -4.34871018e-01 -2.59918004e-01
9.87666428e-01 -5.77330828e-01 -3.73115182e-01 2.47934178e-01
-5.51601112e-01 -5.22077493e-02 5.90833068e-01 7.43747711e-01
4.81494099e-01 1.92731209e-02 8.46674085e-01 -1.53264832e+00
7.87498772e-01 -6.61487997e-01 -1.12787172e-01 9.29797173e-01
-5.63336730e-01 6.24910034e-02 2.39847496e-01 -1.15163088e-01
-1.50929880e+00 -3.69407326e-01 -1.85754076e-01 3.06488663e-01
9.52302665e-02 1.26835704e+00 -5.05710244e-01 5.25644243e-01
4.64494288e-01 -3.46672505e-01 -8.93573046e-01 -4.48112190e-01
5.19294977e-01 6.90646112e-01 4.74776268e-01 -7.67065287e-01
5.93203485e-01 -1.63970999e-02 -1.57581076e-01 -2.83975333e-01
-1.24203694e+00 -8.44248831e-01 -1.08759367e+00 9.27062705e-02
9.40430999e-01 -8.79192233e-01 -5.57986319e-01 9.62897204e-03
-1.63983238e+00 1.38253063e-01 -3.96323903e-03 3.85187805e-01
1.80144086e-01 3.88523757e-01 -9.73865807e-01 -6.79669917e-01
-5.10967195e-01 -7.50841796e-01 8.65222275e-01 3.88944775e-01
-4.85613018e-01 -1.12350786e+00 1.73851639e-01 4.82157052e-01
-1.06678881e-01 2.58300573e-01 1.05009294e+00 -1.08261943e+00
-6.00511491e-01 -2.10514694e-01 -5.35984278e-01 -1.07341066e-01
4.50422883e-01 4.11464982e-02 -5.53929508e-01 2.60006368e-01
-5.15966058e-01 -1.51640717e-02 7.68713593e-01 -1.20234258e-01
4.36725289e-01 -4.59501207e-01 -6.59655571e-01 1.14172511e-01
1.27232277e+00 3.19269449e-01 8.10664535e-01 4.78877902e-01
7.69586205e-01 8.36750329e-01 1.06476521e+00 -8.81627295e-03
1.01301491e+00 7.78581023e-01 1.06152862e-01 1.23450756e-01
-1.78792894e-01 -2.91710466e-01 2.91402079e-02 1.20604610e+00
-2.94783115e-01 -3.02770674e-01 -8.42983365e-01 7.05983996e-01
-2.18261433e+00 -9.20559347e-01 -6.62063181e-01 1.76716232e+00
1.24214089e+00 1.75863355e-01 -2.97448784e-01 -4.65373173e-02
8.25723350e-01 6.68009222e-02 -2.24520601e-02 -3.55106771e-01
-2.09738180e-01 4.58658114e-02 2.09874734e-01 6.41293585e-01
-1.14660597e+00 1.18923819e+00 5.00354815e+00 6.42377734e-01
-5.01448870e-01 1.03289604e-01 2.25712270e-01 5.38052022e-01
-4.90277439e-01 4.68653560e-01 -1.31956148e+00 1.58450037e-01
5.91672480e-01 -2.08577633e-01 -1.66464299e-01 5.10960698e-01
-1.85042530e-01 -1.51422009e-01 -8.91189992e-01 2.69557029e-01
-2.50828490e-02 -1.17404211e+00 -2.51679420e-01 -2.56174773e-01
6.85050130e-01 -3.13580155e-01 -5.18498540e-01 5.04304171e-01
4.92277741e-01 -5.49938202e-01 2.54807472e-01 4.85362738e-01
3.56145561e-01 -7.81264067e-01 1.24116898e+00 1.63292080e-01
-1.67093480e+00 4.11715776e-01 -2.00038761e-01 1.25954345e-01
4.47037071e-01 8.77585471e-01 -7.07991540e-01 1.40813506e+00
5.92406392e-01 7.61674583e-01 -4.59377348e-01 7.59118617e-01
-9.30382133e-01 5.05374014e-01 -5.09596802e-02 -5.01141138e-02
-2.38975491e-02 -1.98496789e-01 3.52244794e-01 1.59310150e+00
1.04672261e-01 5.53653002e-01 1.62937064e-02 6.30264163e-01
-3.06594551e-01 4.26083714e-01 -2.34344229e-01 1.64084986e-01
8.77308786e-01 1.48153305e+00 -5.66839635e-01 -3.50558877e-01
-7.00912058e-01 7.50301361e-01 8.02836299e-01 3.84204417e-01
-7.05987632e-01 -7.06578195e-01 6.02828920e-01 -2.38234058e-01
4.05848324e-01 -9.61669534e-02 -3.55521709e-01 -1.41471076e+00
3.55899841e-01 -5.46493948e-01 8.07081521e-01 -5.98067880e-01
-1.48604667e+00 7.48368561e-01 3.05278078e-02 -9.44030225e-01
-1.15426049e-01 -9.81886387e-02 -7.84110129e-01 1.02068567e+00
-1.77546418e+00 -1.32715356e+00 -6.25296161e-02 1.90901577e-01
1.63939372e-01 3.94532204e-01 9.29919302e-01 6.26095474e-01
-8.76418054e-01 7.06843913e-01 -5.34668624e-01 6.59528792e-01
7.38508284e-01 -1.33374703e+00 4.78268176e-01 1.13662910e+00
1.72090769e-01 1.24424088e+00 4.47074741e-01 -9.74090040e-01
-6.32576585e-01 -1.13423431e+00 1.92898059e+00 -3.51462543e-01
7.69302964e-01 -2.02546775e-01 -1.32340002e+00 1.04474640e+00
5.53203285e-01 1.32183265e-02 8.67702544e-01 8.12888384e-01
-6.35393679e-01 1.67072434e-02 -9.14494455e-01 6.05652571e-01
1.16287780e+00 -4.85710829e-01 -1.14645422e+00 1.73432529e-01
9.67877030e-01 -4.80417997e-01 -1.22517490e+00 7.44172931e-01
1.19543128e-01 -6.27852201e-01 8.84348869e-01 -6.49489939e-01
6.77581906e-01 -6.17520571e-01 -1.42658964e-01 -1.32475352e+00
-2.90051877e-01 -3.49598706e-01 -5.54442048e-01 2.10962439e+00
1.04375339e+00 -5.51325679e-01 3.67557883e-01 7.82731712e-01
-2.30538622e-01 -9.04762149e-01 -6.25684619e-01 -6.03376031e-01
-8.79969746e-02 -6.46644309e-02 8.05902600e-01 1.28936887e+00
8.04289937e-01 8.09610605e-01 8.38185847e-03 5.61121821e-01
3.48399222e-01 4.78145659e-01 3.07933688e-01 -1.08276474e+00
-2.77519822e-01 -3.50644797e-01 -1.89077348e-01 -1.03861880e+00
2.72324771e-01 -1.02482903e+00 9.94718075e-02 -2.02465653e+00
2.80015737e-01 -6.72515988e-01 -3.07332933e-01 7.51900136e-01
-8.22800517e-01 -1.07398584e-01 -5.50308824e-03 2.18305260e-01
-6.75991356e-01 3.05856377e-01 1.03263199e+00 -1.17480107e-01
-2.25003317e-01 -1.62822902e-01 -1.03094864e+00 7.31433213e-01
6.05795681e-01 -4.80637908e-01 -2.93742687e-01 -4.20793742e-01
4.98313785e-01 6.75577223e-02 -1.61550552e-01 -3.45443577e-01
5.85120201e-01 -2.35794932e-01 -1.42750323e-01 -7.48678267e-01
9.92322713e-02 -6.34876966e-01 8.04260746e-03 -6.11677095e-02
-4.11802918e-01 -3.45951207e-02 -1.49280131e-01 1.88501135e-01
-6.69636190e-01 -5.09965658e-01 2.06110299e-01 -8.24976526e-03
-6.74351037e-01 2.06134487e-02 1.26378208e-01 3.54729861e-01
9.89778399e-01 3.98193210e-01 -8.53942752e-01 4.53634076e-02
-8.52616191e-01 4.15259480e-01 5.24052382e-02 4.46704119e-01
4.11166847e-01 -1.36287045e+00 -7.66714454e-01 -5.17999709e-01
2.50514448e-01 3.39582384e-01 2.41327256e-01 8.73405874e-01
-1.10434018e-01 4.16597813e-01 5.48494339e-01 4.49437723e-02
-1.54073393e+00 6.58736348e-01 2.52450407e-01 -1.00148249e+00
-5.20331323e-01 9.29874480e-01 8.02738816e-02 -6.02409303e-01
-5.34890257e-02 -2.84393638e-01 -8.85945618e-01 3.49686474e-01
5.49057662e-01 -8.56924057e-03 3.18422675e-01 -6.86860740e-01
-7.51559913e-01 5.57920039e-01 -4.40501064e-01 1.82582304e-01
1.23248756e+00 -2.48534411e-01 -6.50868773e-01 2.89260328e-01
9.41310167e-01 2.74272382e-01 -5.40221512e-01 -5.58931410e-01
7.40889609e-01 -2.10295677e-01 -3.87234479e-01 -8.64274561e-01
-1.03964281e+00 5.02297044e-01 -5.03638566e-01 4.65464294e-01
1.06380510e+00 3.55239421e-01 8.17188859e-01 3.04792821e-01
3.02381873e-01 -8.44504952e-01 -4.27010536e-01 7.67544091e-01
7.62213171e-01 -9.79191422e-01 2.31515616e-01 -1.32357037e+00
-7.66153276e-01 9.56964552e-01 9.74120200e-01 1.74851239e-01
6.69220686e-01 4.95282769e-01 6.31484091e-02 -2.52677411e-01
-8.48553717e-01 -4.91188049e-01 5.32831192e-01 2.67081708e-01
1.02021813e+00 4.51766700e-02 -1.01923537e+00 9.03487086e-01
-2.06562094e-02 -2.56469429e-01 3.28482330e-01 1.00272024e+00
-2.26749673e-01 -1.51584268e+00 7.34493434e-02 2.65760303e-01
-6.69130325e-01 -5.39155841e-01 -5.16693532e-01 6.40533566e-01
9.57536548e-02 1.26962340e+00 -6.99105710e-02 -4.57962990e-01
6.15375102e-01 1.04811817e-01 2.29874641e-01 -8.82202208e-01
-8.53332400e-01 -2.20266446e-01 8.03263307e-01 -1.63791806e-01
-7.78236151e-01 -8.02003264e-01 -1.73757386e+00 -1.14318699e-01
-7.61615038e-01 4.53518361e-01 2.37783417e-01 1.25402749e+00
4.08473611e-01 8.58334780e-01 4.09173042e-01 9.22059342e-02
-3.73324305e-02 -1.19607913e+00 -3.18966031e-01 4.75829780e-01
-1.45799235e-01 -6.15914881e-01 -2.18418598e-01 -1.28096402e-01] | [9.281805038452148, 8.653569221496582] |
2babdd23-1f57-4884-b48b-d878ba7651d5 | robust-and-accurate-object-detection-via-self | 2111.07239 | null | https://arxiv.org/abs/2111.07239v1 | https://arxiv.org/pdf/2111.07239v1.pdf | Robust and Accurate Object Detection via Self-Knowledge Distillation | Object detection has achieved promising performance on clean datasets, but how to achieve better tradeoff between the adversarial robustness and clean precision is still under-explored. Adversarial training is the mainstream method to improve robustness, but most of the works will sacrifice clean precision to gain robustness than standard training. In this paper, we propose Unified Decoupled Feature Alignment (UDFA), a novel fine-tuning paradigm which achieves better performance than existing methods, by fully exploring the combination between self-knowledge distillation and adversarial training for object detection. We first use decoupled fore/back-ground features to construct self-knowledge distillation branch between clean feature representation from pretrained detector (served as teacher) and adversarial feature representation from student detector. Then we explore the self-knowledge distillation from a new angle by decoupling original branch into a self-supervised learning branch and a new self-knowledge distillation branch. With extensive experiments on the PASCAL-VOC and MS-COCO benchmarks, the evaluation results show that UDFA can surpass the standard training and state-of-the-art adversarial training methods for object detection. For example, compared with teacher detector, our approach on GFLV2 with ResNet-50 improves clean precision by 2.2 AP on PASCAL-VOC; compared with SOTA adversarial training methods, our approach improves clean precision by 1.6 AP, while improving adversarial robustness by 0.5 AP. Our code will be available at https://github.com/grispeut/udfa. | ['Hongcheng Huang', 'Xiongziyan Xiao', 'Renhao Xie', 'Pengzhi Chu', 'Weipeng Xu'] | 2021-11-14 | null | null | null | null | ['self-knowledge-distillation'] | ['computer-vision'] | [-1.49117127e-01 -1.60362750e-01 1.16309769e-01 -1.21793061e-01
-1.11950910e+00 -8.64039183e-01 6.54992402e-01 -3.08191389e-01
-4.30128098e-01 5.40791750e-01 -3.21037434e-02 -1.70116857e-01
3.36370170e-01 -8.74931395e-01 -9.85738337e-01 -8.89031351e-01
2.52341270e-01 -8.77061412e-02 5.99216163e-01 -4.01361853e-01
-2.09436387e-01 4.64279503e-01 -1.19818187e+00 3.36546034e-01
1.02322388e+00 9.01360929e-01 -2.31516227e-01 9.40587640e-01
8.65065306e-02 7.66954184e-01 -9.21287477e-01 -7.98615336e-01
6.86324000e-01 -1.08636342e-01 -5.12187958e-01 -5.71371496e-01
9.58716512e-01 -3.88297319e-01 -7.51096547e-01 1.48297322e+00
8.64771128e-01 -1.52357802e-01 5.06053746e-01 -1.17798579e+00
-9.92927432e-01 6.55379415e-01 -6.20607913e-01 3.18421334e-01
-6.21788800e-02 6.06586754e-01 5.98412156e-01 -8.18468869e-01
2.31821075e-01 1.26872349e+00 7.74162352e-01 8.38343799e-01
-1.03301680e+00 -1.16340041e+00 2.09955245e-01 3.68588656e-01
-1.41222584e+00 -2.79812038e-01 5.62295914e-01 -4.56858605e-01
7.08305418e-01 3.89799535e-01 2.33375147e-01 1.40731227e+00
1.84308633e-01 9.14860189e-01 1.13512170e+00 -2.70224452e-01
-1.06054053e-01 3.19590211e-01 1.09272890e-01 6.19008422e-01
4.48550344e-01 6.46353483e-01 -7.42666125e-02 1.07492551e-01
5.83768845e-01 -2.86833290e-02 -2.24176973e-01 -2.41980985e-01
-8.71275127e-01 6.86113656e-01 1.02827597e+00 7.18332455e-02
1.12308346e-01 3.05808723e-01 5.82262516e-01 4.26499665e-01
2.25866541e-01 5.74998200e-01 -4.17403013e-01 2.78722256e-01
-6.77576005e-01 3.18439901e-01 6.57728851e-01 1.01748443e+00
5.77389538e-01 4.94847417e-01 -6.39553547e-01 7.35171795e-01
2.21357673e-01 1.04816711e+00 4.21726644e-01 -6.66828871e-01
3.56373459e-01 4.39109623e-01 -1.83824122e-01 -7.07279980e-01
-4.06594388e-03 -8.54577959e-01 -9.63690042e-01 6.32227182e-01
3.83903891e-01 -3.09491545e-01 -1.18732476e+00 1.62414861e+00
4.92568284e-01 3.00794065e-01 2.95325518e-01 8.82469535e-01
1.07477343e+00 4.88992512e-01 7.38754049e-02 2.77944446e-01
1.40458691e+00 -1.34648645e+00 -4.93365198e-01 -2.30872631e-01
2.97604918e-01 -9.82433975e-01 8.37528467e-01 5.61165214e-01
-7.87042022e-01 -1.00629139e+00 -1.26801968e+00 -1.05929874e-01
-5.31768322e-01 3.86528596e-02 4.37671006e-01 9.31487739e-01
-5.01929939e-01 5.56649685e-01 -7.27197587e-01 4.42408621e-02
7.52369404e-01 1.69246897e-01 -4.12332445e-01 -9.80186611e-02
-1.30241477e+00 9.13475573e-01 4.49936390e-01 -1.00486338e-01
-1.30663967e+00 -9.25352275e-01 -8.10233474e-01 -5.32858521e-02
6.27192080e-01 -8.35004508e-01 1.29685771e+00 -9.18470263e-01
-1.43538833e+00 6.58090651e-01 4.12761241e-01 -7.42873549e-01
7.44930446e-01 -8.70486259e-01 -5.72708726e-01 -1.17303751e-01
-1.44284777e-02 5.73718309e-01 1.21383560e+00 -1.41108406e+00
-5.98179102e-01 -1.77104175e-01 2.51830190e-01 -3.95426974e-02
-3.51095200e-01 -5.69984969e-03 -5.88781297e-01 -1.01683366e+00
-4.27903146e-01 -8.66746545e-01 -1.93525106e-01 -4.75659296e-02
-5.46562135e-01 -1.95843354e-02 1.02529144e+00 -5.07497668e-01
1.16088247e+00 -2.27788568e+00 -1.13372490e-01 -6.56752586e-02
2.66541719e-01 1.07186544e+00 -2.95254678e-01 -4.55474481e-02
-2.91148633e-01 1.15223721e-01 -2.65357792e-01 -2.53020376e-01
5.42346202e-03 1.01089358e-01 -6.08777344e-01 6.36566460e-01
3.72541785e-01 1.11275709e+00 -1.10497928e+00 -3.49869132e-01
3.45676154e-01 5.04935265e-01 -6.08780324e-01 4.14551556e-01
9.56171900e-02 3.71436328e-01 -3.01221341e-01 7.30548024e-01
1.05156410e+00 3.23037922e-01 -3.64417285e-01 -3.50271553e-01
2.41974536e-02 -8.45313743e-02 -1.43971777e+00 1.40475750e+00
-3.11005682e-01 5.25386453e-01 1.38743371e-01 -6.38159513e-01
8.10116768e-01 -3.90045345e-02 -7.11529888e-03 -6.81466818e-01
2.75777608e-01 6.69417381e-02 1.07444346e-01 -2.52697915e-01
4.27231997e-01 6.26829937e-02 -2.16728255e-01 -2.27107689e-01
3.85539442e-01 -3.02475363e-01 2.23863851e-02 3.06863457e-01
1.16015661e+00 1.47586271e-01 1.94360405e-01 -1.96866468e-01
6.45398736e-01 -1.63228646e-01 7.20361173e-01 9.70791698e-01
-4.68937606e-01 7.25845039e-01 1.81895614e-01 -3.14709067e-01
-6.77411854e-01 -1.43444741e+00 -2.71998316e-01 1.03517759e+00
2.47314647e-01 -3.91757041e-01 -1.01778829e+00 -1.19431996e+00
2.65407175e-01 5.54261625e-01 -7.06986248e-01 -4.68954504e-01
-6.93966389e-01 -6.03540123e-01 1.03947568e+00 7.37073958e-01
8.72935772e-01 -8.98064196e-01 -1.09118044e-01 -1.45189688e-01
1.58623055e-01 -1.10269165e+00 -5.40077627e-01 1.96951136e-01
-3.38880897e-01 -1.14463878e+00 -6.19998813e-01 -5.00676215e-01
5.49750209e-01 4.11564171e-01 9.89795029e-01 1.33109409e-02
-5.00925720e-01 2.40537301e-01 -4.44526255e-01 -8.71318340e-01
-5.46776652e-01 6.18422888e-02 2.74641722e-01 -1.99566707e-01
1.78084970e-01 -3.37539494e-01 -6.14381969e-01 3.38072449e-01
-1.07045591e+00 -3.13750446e-01 6.28203750e-01 9.50839996e-01
6.22362912e-01 -1.77991852e-01 4.51016694e-01 -9.36903059e-01
1.88512221e-01 -4.29193944e-01 -7.85262108e-01 1.10876128e-01
-4.97467995e-01 1.32062078e-01 9.39708889e-01 -4.74993616e-01
-1.08274961e+00 5.20625822e-02 -5.02806008e-01 -8.76707256e-01
-2.65826344e-01 -2.59685218e-01 -3.86242956e-01 -3.06678295e-01
1.31309605e+00 -2.32683402e-03 -3.62934589e-01 -4.02065456e-01
7.61559486e-01 5.82953095e-01 1.12033045e+00 -6.75532341e-01
1.59389377e+00 3.95709485e-01 -3.13790470e-01 -3.81992280e-01
-9.78048861e-01 -4.88239318e-01 -4.68474150e-01 2.07265079e-01
7.49734640e-01 -1.23937631e+00 -4.66439873e-01 7.49827027e-01
-9.97996569e-01 -2.90582865e-01 -6.50591910e-01 2.20255196e-01
-2.60875136e-01 4.63701755e-01 -5.17943084e-01 -4.16940898e-01
-6.81662858e-01 -1.17720449e+00 9.60812628e-01 3.43331784e-01
3.80592287e-01 -5.60697198e-01 2.26789415e-01 3.45915854e-01
5.36463559e-01 4.92597908e-01 1.41642347e-01 -7.89744914e-01
-5.18960714e-01 -4.02073979e-01 -3.85560364e-01 1.01309693e+00
3.29680741e-02 -3.54897603e-02 -1.30805039e+00 -5.71852922e-01
6.84367865e-02 -3.86478931e-01 1.27011240e+00 -3.19285020e-02
1.23100674e+00 -3.73615146e-01 -2.33226255e-01 1.09334564e+00
1.42440307e+00 -2.21633390e-01 8.63091528e-01 4.25103247e-01
9.89159048e-01 1.75865609e-02 5.77444017e-01 -1.75964157e-03
6.16528802e-02 5.83815277e-01 7.31355429e-01 -2.26745844e-01
-7.10555315e-01 -3.19747239e-01 6.62903488e-01 4.47211415e-01
-4.91546951e-02 -4.68179546e-02 -7.05119848e-01 3.79152924e-01
-1.70289278e+00 -1.04376817e+00 -1.12712383e-01 2.16126728e+00
1.01140761e+00 3.46136332e-01 9.95852053e-02 1.77649841e-01
6.72582924e-01 1.39531389e-01 -6.08025193e-01 -4.00564224e-01
-9.70985144e-02 3.70192736e-01 8.50437403e-01 4.93089646e-01
-1.52120376e+00 1.16384244e+00 5.26554537e+00 1.33649325e+00
-1.16355085e+00 3.55965793e-01 3.07828158e-01 -1.50001608e-02
1.23604134e-01 -1.86402932e-01 -1.17434371e+00 3.94565523e-01
6.79497600e-01 2.04450544e-03 2.21172094e-01 1.23786509e+00
-4.48674470e-01 2.38454118e-01 -1.02810252e+00 8.56074691e-01
-3.60106525e-04 -1.07406318e+00 1.30631700e-02 -2.59323239e-01
1.05883920e+00 2.86813080e-01 2.12553054e-01 9.26573098e-01
7.73031652e-01 -9.91540670e-01 7.39840984e-01 3.32247406e-01
9.38117623e-01 -6.79684162e-01 8.54790449e-01 2.89908201e-01
-1.26828849e+00 -1.94062471e-01 -4.16471362e-01 2.78790683e-01
-1.58373818e-01 6.59882843e-01 -6.57580674e-01 8.58148336e-01
9.19552982e-01 3.56945157e-01 -8.95790219e-01 1.08718681e+00
-7.87592649e-01 6.99678183e-01 -1.14620663e-01 3.84516865e-01
5.44911288e-02 3.78378719e-01 8.24897826e-01 1.65113270e+00
-1.27232388e-01 -4.84034233e-02 1.62875667e-01 7.42838502e-01
-3.42497855e-01 -1.52136803e-01 -5.23977935e-01 4.10415947e-01
4.86535519e-01 1.45338011e+00 -1.45500809e-01 -4.27505791e-01
-2.21230134e-01 1.06316185e+00 3.73837173e-01 1.55190766e-01
-1.43313181e+00 -7.52907276e-01 9.47937369e-01 -3.04662958e-02
4.62887317e-01 -8.57753977e-02 -2.99184084e-01 -1.34244871e+00
1.43480012e-02 -1.19750357e+00 3.22981894e-01 -2.51737446e-01
-1.36544979e+00 6.78065479e-01 -1.18753232e-01 -1.33869660e+00
1.41974300e-01 -7.37266183e-01 -8.72491419e-01 8.30151737e-01
-1.77742779e+00 -1.35972703e+00 -5.08649886e-01 7.28106737e-01
6.25209808e-01 -2.24003583e-01 7.62232006e-01 4.54979122e-01
-8.01858485e-01 1.37356639e+00 2.60200799e-01 5.29399753e-01
1.12969339e+00 -1.46081042e+00 7.75599957e-01 1.32700551e+00
2.38850340e-01 4.43342060e-01 5.24863839e-01 -5.03004730e-01
-1.39665830e+00 -1.53948629e+00 5.04562445e-02 -8.56374562e-01
6.07305586e-01 -3.81694674e-01 -1.02236652e+00 5.28422058e-01
2.51712441e-01 6.35200620e-01 3.77758831e-01 -1.54557064e-01
-1.03001857e+00 -4.47603047e-01 -1.32880402e+00 5.17867982e-01
1.07799089e+00 -3.59039128e-01 -7.43145049e-01 2.49244809e-01
1.11421084e+00 -8.23519826e-01 -8.16749036e-01 7.20129311e-01
3.13667297e-01 -9.37730312e-01 1.33092546e+00 -4.61781144e-01
2.45979905e-01 -6.67090237e-01 -2.24122897e-01 -1.43426013e+00
-5.88221788e-01 -7.66349673e-01 -3.96803170e-01 1.46849132e+00
1.34418070e-01 -6.68167830e-01 4.51114208e-01 5.68859726e-02
-4.97329801e-01 -7.03303814e-01 -7.78891921e-01 -1.22882164e+00
5.15504301e-01 -4.50577825e-01 6.00348830e-01 6.72405779e-01
-5.96077502e-01 1.77166328e-01 -3.44251692e-01 5.04241109e-01
7.72235990e-01 -2.03453645e-01 1.31318283e+00 -7.04699814e-01
-5.56101859e-01 -4.84730899e-01 -4.37610745e-01 -9.54975247e-01
-1.35566279e-01 -9.68435407e-01 1.04081534e-01 -1.17699945e+00
9.32737216e-02 -3.33941579e-01 -5.20907223e-01 7.39421844e-01
-6.73766911e-01 4.86730427e-01 5.58712542e-01 -3.50540280e-02
-5.74988902e-01 5.13899982e-01 1.16668952e+00 -4.08204913e-01
-2.55282689e-02 -1.71119064e-01 -9.22073066e-01 7.61730194e-01
7.72085905e-01 -6.63552940e-01 -4.68669198e-02 -5.53827703e-01
-3.16879541e-01 -5.79833031e-01 4.94762480e-01 -1.38423157e+00
2.67710239e-01 5.01991832e-04 6.38730288e-01 -2.19172537e-01
9.72689539e-02 -6.47131920e-01 -1.32502720e-01 7.13655531e-01
1.00149795e-01 -3.91891062e-01 6.70100152e-01 5.79635561e-01
-1.72188088e-01 -1.05189435e-01 1.24842560e+00 6.56377599e-02
-7.33282626e-01 3.20576221e-01 2.95901597e-01 2.82423526e-01
1.20165169e+00 1.36354836e-02 -8.07216525e-01 1.54155016e-01
-4.99346137e-01 3.50136667e-01 2.98955530e-01 6.95678353e-01
5.07178068e-01 -1.43693280e+00 -9.62902725e-01 3.07311684e-01
1.01951053e-02 1.76446229e-01 3.28176349e-01 4.59533632e-01
-3.02253366e-01 6.29717335e-02 -2.21743792e-01 -5.24898291e-01
-1.26350033e+00 8.83844435e-01 4.21395987e-01 -4.37227875e-01
-4.41284031e-01 1.18470216e+00 5.53241253e-01 -4.75993991e-01
4.13473159e-01 -1.49315432e-01 2.52610631e-02 -3.26760262e-01
7.31602192e-01 6.02055609e-01 1.29287079e-01 -5.40611863e-01
-3.08397502e-01 5.89111209e-01 -2.65286475e-01 4.97891158e-01
9.66038167e-01 3.65688831e-01 2.31465131e-01 -2.58518040e-01
1.19964015e+00 4.15780395e-01 -1.37411070e+00 -3.77095699e-01
-5.39396584e-01 -5.87982357e-01 -6.73781559e-02 -9.60016191e-01
-1.26589024e+00 8.26278985e-01 9.55610335e-01 2.83566248e-02
1.18938494e+00 -1.63167834e-01 8.26959670e-01 3.31552505e-01
1.65500984e-01 -6.82457507e-01 3.05779427e-01 6.09608412e-01
1.13204420e+00 -1.43361568e+00 1.04040159e-02 -5.80729067e-01
-5.97662210e-01 8.47678721e-01 1.16783786e+00 -5.37158608e-01
4.96344239e-01 5.07714033e-01 1.79459289e-01 8.82076398e-02
-2.61330724e-01 -3.49120051e-01 5.69545329e-01 6.50398672e-01
3.28712240e-02 1.59729064e-01 -8.27277824e-03 9.07140076e-01
-3.56520355e-01 -2.91199833e-01 1.46225378e-01 6.68784857e-01
-4.98309165e-01 -1.16996694e+00 -7.25581706e-01 2.13942036e-01
-7.50818491e-01 -1.56530663e-01 -1.73904464e-01 7.47549832e-01
6.09085381e-01 8.80374551e-01 -3.28889996e-01 -7.07375348e-01
7.91461766e-01 -2.19366014e-01 4.98696387e-01 -6.96031332e-01
-9.94671583e-01 -2.17622057e-01 -2.33889595e-01 -6.45057619e-01
1.01560377e-01 -2.76703566e-01 -1.10759556e+00 -3.09507310e-01
-6.87370896e-01 6.21119142e-03 4.78953034e-01 5.99346042e-01
3.31978232e-01 8.46109807e-01 7.78915346e-01 -7.24453747e-01
-1.00382662e+00 -8.80268514e-01 -5.54148220e-02 3.30155879e-01
4.79118288e-01 -5.50736725e-01 -3.78436059e-01 -3.39669809e-02] | [5.5436835289001465, 7.932278156280518] |
e267d5d9-a27e-403e-ae78-a509027d6d68 | classification-of-lung-pathologies-in | 2302.07157 | null | https://arxiv.org/abs/2302.07157v2 | https://arxiv.org/pdf/2302.07157v2.pdf | Classification of Lung Pathologies in Neonates using Dual Tree Complex Wavelet Transform | Annually 8500 neonatal deaths are reported in the US due to respiratory failure. Recently, Lung Ultrasound (LUS), due to its radiation free nature, portability, and being cheaper is gaining wide acceptability as a diagnostic tool for lung conditions. However, lack of highly trained medical professionals has limited its use especially in remote areas. To address this, an automated screening system that captures characteristics of the LUS patterns can be of significant assistance to clinicians who are not experts in lung ultrasound (LUS) images. In this paper, we propose a feature extraction method designed to quantify the spatially-localized line patterns and texture patterns found in LUS images. Using the dual-tree complex wavelet transform (DTCWT) and four types of common image features we propose a method to classify the LUS images into 6 common neonatal lung conditions. These conditions are normal lung, pneumothorax (PTX), transient tachypnea of the newborn (TTN), respiratory distress syndrome (RDS), chronic lung disease (CLD) and consolidation (CON) that could be pneumonia or atelectasis. The proposed method using DTCWT decomposition extracted global statistical, grey-level co-occurrence matrix (GLCM), grey-level run length matrix (GLRLM) and linear binary pattern (LBP) features to be fed to a linear discriminative analysis (LDA) based classifier. Using 15 best DTCWT features along with 3 clinical features the proposed approach achieved a per-image classification accuracy of 92.78% with a balanced dataset containing 720 images from 24 patients and 74.39% with the larger unbalanced dataset containing 1550 images from 42 patients. Likewise, the proposed method achieved a maximum per-subject classification accuracy of 81.53% with 43 DTCWT features and 3 clinical features using the balanced dataset and 64.97% with 13 DTCWT features and 3 clinical features using the unbalanced dataset. | ['Karthikeyan Umapathy', 'Naimul Khan', 'Lei Gao', 'Randy Tan', 'Ryan Tan', 'Adel Mohamed', 'Sagarjit Aujla'] | 2023-02-14 | null | null | null | null | ['respiratory-failure'] | ['medical'] | [ 3.22391599e-01 -3.52046460e-01 1.06341958e-01 -1.35283740e-02
-5.14109135e-01 -4.57481831e-01 6.53407350e-02 1.96195468e-01
-1.40538618e-01 4.76695120e-01 -9.48348548e-03 -5.00002861e-01
-5.53827882e-01 -6.69872701e-01 -3.67954105e-01 -1.07368946e+00
-2.82564074e-01 3.42037350e-01 7.18882859e-01 6.31247461e-01
1.32696986e-01 9.19878125e-01 -1.51291490e+00 6.61018074e-01
6.28407598e-01 1.16538799e+00 4.43059713e-01 1.22579730e+00
-2.01229155e-01 1.08059347e+00 -1.23346239e-01 3.70340957e-03
1.89636186e-01 -7.09843338e-01 -6.28411829e-01 -3.15139741e-01
4.42578286e-01 -4.32480991e-01 6.55742968e-03 4.74669546e-01
1.04389739e+00 -8.38481262e-02 1.09026122e+00 -8.61064494e-01
8.20354670e-02 -8.11692476e-02 -8.94362688e-01 8.10463488e-01
2.98468441e-01 -1.87108785e-01 5.22083402e-01 -1.01933742e+00
2.75380015e-01 9.29138660e-01 1.01850128e+00 3.26704741e-01
-7.23983943e-01 -7.05555081e-01 -1.08937597e+00 6.03609741e-01
-1.00001073e+00 1.72183514e-02 3.52745026e-01 -7.99898148e-01
8.57649744e-01 6.88719928e-01 8.71891975e-01 3.55658650e-01
4.91035223e-01 6.93488866e-02 1.55722988e+00 -6.49908185e-01
-6.51474521e-02 -1.08417600e-01 -3.14746290e-01 1.22906733e+00
6.91366792e-01 -2.06233725e-01 -1.17133126e-01 -3.82970840e-01
6.30953252e-01 3.42151433e-01 -2.82069236e-01 -2.24648327e-01
-8.98474097e-01 7.04390347e-01 -2.58786649e-01 8.51365805e-01
-6.41750157e-01 -1.21334039e-01 3.79898220e-01 -5.09030111e-02
6.34085685e-02 3.50199752e-02 -5.69987774e-01 -3.74281794e-01
-9.25372720e-01 -2.83237755e-01 7.66132116e-01 6.11418664e-01
5.45573711e-01 -1.09319165e-01 -5.04875362e-01 1.04965246e+00
1.55404359e-01 9.60610747e-01 8.02860141e-01 -7.81383395e-01
3.93977135e-01 4.00587708e-01 -4.61772501e-01 -1.06602776e+00
-6.81487381e-01 5.83598167e-02 -8.10415804e-01 -2.04678327e-01
-8.85034204e-02 -2.88501084e-01 -7.63098598e-01 1.16291142e+00
3.96571189e-01 2.90913731e-01 -8.40906873e-02 4.49759513e-01
1.17551458e+00 5.83983481e-01 -5.38695082e-02 -6.88012898e-01
1.42027533e+00 -5.40819407e-01 -5.92937231e-01 5.31116784e-01
6.19722307e-01 -9.59075451e-01 3.91221285e-01 2.21765965e-01
-9.13946748e-01 -3.07112277e-01 -5.36304712e-01 6.85447991e-01
-2.62521598e-02 4.52849150e-01 -3.53110246e-02 8.73420894e-01
-1.15654778e+00 3.95209670e-01 -9.30718124e-01 -8.97133529e-01
4.53840673e-01 4.57892865e-01 -5.13165951e-01 -1.88091293e-01
-5.74805677e-01 9.15820539e-01 8.54904577e-02 -3.56904685e-01
-2.45947316e-01 -5.58820784e-01 -4.60272253e-01 -3.99523899e-02
9.32774041e-03 -2.50951648e-01 6.83649123e-01 -1.62866578e-01
-9.72083449e-01 8.01802397e-01 -1.89726204e-01 1.88487187e-01
1.95275471e-01 1.06830716e-01 -3.15676033e-01 1.03089392e+00
1.74757540e-01 -2.43791984e-03 7.10960031e-01 -8.29722404e-01
-7.55406320e-01 -2.23664910e-01 -9.96448755e-01 1.03827650e-02
-3.89750063e-01 5.00606298e-01 -1.07108064e-01 -6.08292580e-01
5.48540235e-01 -7.65987217e-01 2.52318174e-01 -1.19619928e-01
1.58430412e-01 -1.22010894e-01 1.29128754e+00 -1.12517548e+00
1.30118680e+00 -1.82856953e+00 -5.37137866e-01 3.59099656e-01
1.74502462e-01 6.14904463e-01 1.69869810e-01 4.56183791e-01
-2.40489081e-01 1.82529036e-02 -1.85664997e-01 4.68419790e-01
-6.01220191e-01 3.78127843e-01 5.25081635e-01 5.80307543e-01
3.03405404e-01 7.50312388e-01 -7.46152878e-01 -1.10854483e+00
3.94349337e-01 4.20978397e-01 -4.50297862e-01 6.79008603e-01
7.78832793e-01 3.59004170e-01 -6.99362636e-01 7.62014329e-01
8.52107048e-01 1.13140672e-01 5.48292063e-02 -3.44111741e-01
-2.95801431e-01 -3.32893223e-01 -8.66477609e-01 7.71247804e-01
-2.45556653e-01 4.80131447e-01 -5.74478246e-02 -1.16966569e+00
8.26818883e-01 9.79978859e-01 9.81326222e-01 -4.18464035e-01
1.71057999e-01 4.13445443e-01 2.34705985e-01 -1.62240267e+00
-6.48241401e-01 -3.40617687e-01 7.51909673e-01 4.95059520e-01
-8.19922611e-02 -1.49764434e-01 8.45348090e-02 -2.93133467e-01
1.67307794e+00 -1.94861442e-01 6.00423574e-01 -4.38875556e-01
6.40618086e-01 -3.64090174e-01 3.58890146e-01 6.93888187e-01
-2.11739957e-01 1.05044377e+00 4.36060548e-01 -2.58937120e-01
-8.09136033e-01 -7.84952879e-01 -4.42183167e-01 7.48936176e-01
-5.95399201e-01 3.16838235e-01 -5.51697195e-01 -6.11893117e-01
-2.86940206e-02 9.26924273e-02 -4.60490733e-01 2.23015696e-01
-8.76747489e-01 -8.47020090e-01 6.40167475e-01 5.40124238e-01
4.45433855e-01 -1.20310616e+00 -1.41067708e+00 1.44275442e-01
5.36486953e-02 -8.10837567e-01 -1.09510370e-01 2.57240862e-01
-9.54617620e-01 -1.17772746e+00 -1.11118722e+00 -9.33542490e-01
6.61218226e-01 2.04193473e-01 6.43050492e-01 2.67765492e-01
-1.09682024e+00 5.27056217e-01 -5.50573945e-01 -1.72244519e-01
-5.08475959e-01 -4.92993891e-01 2.62411544e-03 -2.40326270e-01
1.05054438e-01 -5.54698229e-01 -7.73632586e-01 6.66696355e-02
-8.89641225e-01 -2.21463233e-01 9.15250361e-01 8.31069708e-01
3.68609279e-01 1.11981280e-01 5.06955028e-01 -8.51107001e-01
2.71851391e-01 -8.73735249e-01 -7.66005181e-03 3.71451646e-01
-4.79900867e-01 -4.68858331e-01 5.70424914e-01 -2.72140026e-01
-9.21828270e-01 -3.79382551e-01 -7.88726881e-02 -5.92944145e-01
-3.20839167e-01 3.68826628e-01 4.54822809e-01 -2.88286150e-01
5.06385505e-01 2.38263309e-01 2.99721062e-02 -5.81660986e-01
-5.90325058e-01 9.51179385e-01 3.49269181e-01 -5.29055893e-01
6.68352723e-01 1.96547866e-01 6.55793667e-01 -9.59697306e-01
-3.93787682e-01 -8.85159612e-01 -8.19972336e-01 -3.37556958e-01
1.23034763e+00 -5.04382312e-01 -3.22687775e-01 4.40195143e-01
-8.16277623e-01 1.40279800e-01 -2.64484547e-02 9.88509238e-01
-3.57690215e-01 7.80325592e-01 -3.91461343e-01 -9.98914063e-01
-5.87302744e-01 -7.85568774e-01 5.32885134e-01 1.79862410e-01
-1.12814292e-01 -8.59209657e-01 1.84378162e-01 4.99913633e-01
7.33316064e-01 7.72602201e-01 1.38617539e+00 -8.04700255e-01
6.52722269e-02 -4.78960633e-01 -4.90269929e-01 6.29859746e-01
6.99710608e-01 2.21400037e-01 -7.28460968e-01 -3.06442529e-01
4.72716093e-01 -2.41057158e-01 4.05287027e-01 4.41443354e-01
1.23007524e+00 -4.52674359e-01 -2.19620287e-01 3.88914406e-01
1.58756709e+00 9.21209514e-01 1.18490577e-01 -1.98120952e-01
7.12463677e-01 5.94809592e-01 4.97556001e-01 6.01564467e-01
-4.78059649e-02 1.18284516e-01 1.57817945e-01 -2.48371080e-01
-3.61518115e-01 9.73687991e-02 -1.83395907e-01 1.19046628e+00
-5.96044779e-01 -8.14986974e-02 -1.34175920e+00 5.26224911e-01
-1.22173142e+00 -1.07526791e+00 -3.78628075e-01 1.91834295e+00
7.44682372e-01 -4.42017674e-01 -4.26905565e-02 4.89301801e-01
8.07528019e-01 -2.86135882e-01 -1.23143373e-02 -7.78539836e-01
1.32740676e-01 9.54733551e-01 6.08352244e-01 -2.58784354e-01
-9.81189966e-01 -2.18734564e-03 5.45979357e+00 8.49280059e-01
-1.30393124e+00 1.31548077e-01 5.97759604e-01 4.15904939e-01
1.21254139e-01 -3.56241584e-01 -1.37567922e-01 5.95249772e-01
6.05700493e-01 4.17918831e-01 -1.93639174e-02 5.11070907e-01
2.36502528e-01 -5.24492800e-01 -6.44874454e-01 9.00302649e-01
2.08399311e-01 -8.51908684e-01 -2.10674465e-01 -2.44090140e-01
8.50560069e-01 6.18778951e-02 -1.68792561e-01 -3.99791777e-01
-7.00290918e-01 -1.22817421e+00 8.57512876e-02 5.24740696e-01
1.52526903e+00 -3.54624718e-01 1.06277442e+00 4.13032025e-01
-1.30317855e+00 -5.39104454e-02 -2.61076301e-01 2.15546414e-01
-6.75343633e-01 2.31298581e-01 -1.38352013e+00 4.30028200e-01
1.11412597e+00 3.10333818e-01 -4.33781415e-01 1.33695829e+00
4.10013407e-01 1.19574177e+00 -6.46416962e-01 -3.45439106e-01
1.59209631e-02 -1.85688242e-01 3.90235305e-01 1.72546840e+00
8.34620595e-01 6.91074073e-01 -4.98410940e-01 2.52165824e-01
4.33543265e-01 5.69047451e-01 -5.17317295e-01 2.10851058e-01
4.08316612e-01 1.30261445e+00 -1.10113370e+00 -1.64545313e-01
-7.65320778e-01 3.21620971e-01 -2.53171235e-01 8.95969197e-02
-5.64787328e-01 -4.57585037e-01 -4.28385794e-01 4.20985609e-01
5.91083705e-01 4.15346146e-01 -1.94573522e-01 -5.75907290e-01
-1.66496515e-01 -7.77973890e-01 4.87888843e-01 -5.88153005e-01
-9.51521158e-01 4.22517776e-01 3.12314928e-01 -1.17612696e+00
3.20282727e-02 -5.14911771e-01 -1.02839577e+00 9.56899405e-01
-1.22578156e+00 -8.24362099e-01 -5.40698767e-01 2.74322957e-01
2.36731917e-01 -4.37310994e-01 7.94712901e-01 3.24129254e-01
-3.99520397e-01 3.65392715e-01 3.19719464e-01 -3.80718671e-02
2.94056177e-01 -1.17080307e+00 -6.94347978e-01 5.91543853e-01
-6.05197608e-01 3.01802278e-01 4.25958753e-01 -6.03235304e-01
-1.23674786e+00 -9.84251976e-01 8.90803933e-01 -7.73490081e-03
2.32097402e-01 4.37825024e-01 -8.99895132e-01 5.72903231e-02
-1.56648770e-01 4.49537754e-01 8.47524405e-01 -1.05875325e+00
2.89237171e-01 -2.14318335e-02 -1.67928290e+00 -1.11039229e-01
5.65909386e-01 -3.39750834e-02 -5.34790993e-01 2.12606505e-01
-1.95068866e-01 -1.92077383e-01 -1.15675628e+00 1.07746279e+00
7.81889379e-01 -1.28766215e+00 7.04364240e-01 -1.77512810e-01
4.48318034e-01 -1.95550203e-01 -1.60587996e-01 -3.65600228e-01
-4.03489649e-01 -1.82359889e-01 6.37134314e-02 9.89347279e-01
5.03653772e-02 -8.06822479e-01 6.89211845e-01 3.47403526e-01
6.54389849e-03 -1.37210584e+00 -1.21301699e+00 -3.00768822e-01
-4.49090973e-02 -1.04883261e-01 1.06659755e-01 6.67212903e-01
-2.67780751e-01 -3.36159587e-01 5.41846380e-02 -1.20685130e-01
3.75968784e-01 4.36825231e-02 5.32955602e-02 -9.64445770e-01
-9.44179296e-02 -1.53250247e-01 -6.19307518e-01 -1.01734377e-01
-4.90963429e-01 -6.76891565e-01 -1.10828616e-01 -1.86198163e+00
6.80074811e-01 -7.72031784e-01 -4.72878665e-01 5.14253080e-01
-1.82697281e-01 3.75358373e-01 -1.51419058e-01 2.75259018e-01
-8.70995671e-02 -1.46607414e-01 1.30423224e+00 5.19223750e-01
3.37991446e-01 3.25747699e-01 -4.08014618e-02 6.61209702e-01
7.57497013e-01 -6.43640101e-01 -3.03536654e-01 4.55728434e-02
-2.51540989e-01 6.18447244e-01 1.90214142e-02 -1.23502016e+00
-4.74886857e-02 -2.43727997e-01 7.14624703e-01 -9.82559741e-01
3.52462381e-02 -1.05688143e+00 3.69790375e-01 1.12023282e+00
2.57745117e-01 5.67063354e-02 3.55392806e-02 1.08769223e-01
-1.18297204e-01 -6.20929360e-01 1.04662633e+00 1.09941624e-02
-1.95733935e-01 2.89726347e-01 -7.74551928e-01 1.97998434e-02
1.29972124e+00 -6.38140917e-01 -3.88399437e-02 -2.98791174e-02
-3.30922097e-01 -1.39182389e-01 2.08554849e-01 -1.51864603e-01
7.71884859e-01 -1.09659696e+00 -8.42292130e-01 4.38075751e-01
-1.98634967e-01 -4.58183847e-02 3.35387558e-01 1.48807144e+00
-1.44905543e+00 5.95039308e-01 -1.58724755e-01 -7.74103463e-01
-1.85085344e+00 3.07558440e-02 1.39237672e-01 -3.29940468e-02
-5.23615658e-01 1.00267661e+00 3.10315136e-02 9.02557746e-02
2.53960839e-03 -4.62882549e-01 -7.76843309e-01 1.92038834e-01
-8.69118944e-02 1.00299585e+00 2.19173610e-01 -8.05906773e-01
-5.55408001e-01 1.37445104e+00 4.57333177e-01 3.61966431e-01
1.19067252e+00 1.18357033e-01 -5.58637679e-01 2.15123162e-01
1.69793153e+00 1.91116437e-01 -5.12317061e-01 3.48068401e-02
4.44400758e-02 -4.45779771e-01 -3.65641415e-01 -6.19211257e-01
-1.04947340e+00 9.85310793e-01 9.94070172e-01 4.84613240e-01
1.44599640e+00 1.43899575e-01 6.50728822e-01 1.72134399e-01
-4.07304652e-02 -7.03542829e-01 1.52986407e-01 9.43237394e-02
8.37640226e-01 -1.01750708e+00 1.07386932e-01 -3.77324224e-01
-6.04621947e-01 1.47707856e+00 3.72056544e-01 -2.05398072e-02
9.19655383e-01 3.75321239e-01 1.48965552e-01 -1.70521379e-01
-8.33516836e-01 4.46894281e-02 4.26701039e-01 6.49195194e-01
6.61009550e-01 7.70380720e-02 -6.31454349e-01 4.41525608e-01
8.61445516e-02 -6.86769262e-02 2.46434912e-01 1.37075138e+00
-9.53419566e-01 -4.44321811e-01 -6.89427495e-01 1.35620630e+00
-1.33646250e+00 -3.90666304e-03 1.88973352e-01 4.88191068e-01
8.29430163e-01 7.93547332e-01 -1.68192700e-01 -1.37310460e-01
1.49642467e-01 1.23502292e-01 5.38530350e-01 -7.47592807e-01
-3.77330601e-01 2.70465791e-01 -7.05583766e-02 -1.36168659e-01
-6.40681624e-01 -9.38324153e-01 -1.36575973e+00 1.88260123e-01
-4.07063425e-01 2.27241993e-01 5.86273134e-01 7.38845289e-01
-2.46829882e-01 3.56966704e-01 9.38282251e-01 -2.26399779e-01
-2.01369897e-01 -1.11070943e+00 -6.55112624e-01 1.49485409e-01
3.85802984e-01 -6.35528982e-01 -4.92670000e-01 2.16522753e-01] | [15.383807182312012, -1.970115065574646] |
7615def0-5b46-4447-8fc7-eba9a7f58c77 | choosing-to-rank | 1809.05139 | null | http://arxiv.org/abs/1809.05139v2 | http://arxiv.org/pdf/1809.05139v2.pdf | Choosing to Rank | Ranking data arises in a wide variety of application areas but remains
difficult to model, learn from, and predict. Datasets often exhibit
multimodality, intransitivity, or incomplete rankings---particularly when
generated by humans---yet popular probabilistic models are often too rigid to
capture such complexities. In this work we leverage recent progress on similar
challenges in discrete choice modeling to form flexible and tractable
choice-based models for ranking data. We study choice representations, maps
from rankings (complete or top-$k$) to collections of choices, as a way of
forming ranking models from choice models. We focus on the repeated selection
(RS) choice representation, first used to form the Plackett-Luce ranking model
from the conditional multinomial logit choice model. We fully characterize, for
a prime number of alternatives, the choice representations that admit ranking
distributions with unit normalization, a desirably property that greatly
simplifies maximum likelihood estimation. We further show that only specific
minor variations on repeated selection exhibit this property. Our choice-based
ranking models provide higher out-of-sample likelihood when compared to
Plackett-Luce and Mallows models on a broad collection of ranking tasks
including food preferences, ranked-choice elections, car racing, and search
engine relevance tasks. | ['Stephen Ragain', 'Johan Ugander'] | 2018-09-13 | null | null | null | null | ['carracing-v0'] | ['playing-games'] | [ 4.72751632e-02 -1.28351256e-01 -6.57302082e-01 -7.57359087e-01
-1.23829961e+00 -8.10609221e-01 1.08604741e+00 1.93222195e-01
-5.15174627e-01 8.77981067e-01 4.43688214e-01 -4.21167076e-01
-9.59213436e-01 -5.58557153e-01 -5.67861497e-01 -3.88185441e-01
-1.90088987e-01 1.01232946e+00 -1.83408245e-01 -5.00111163e-01
2.99792051e-01 1.23162210e-01 -1.79594183e+00 1.92658827e-01
7.71589041e-01 7.64596105e-01 5.32294856e-03 4.14065808e-01
3.37823898e-01 7.06687987e-01 1.18819490e-01 -5.20131350e-01
2.28896692e-01 1.60383224e-01 -8.04411709e-01 -3.24424297e-01
3.91040593e-01 -2.61152685e-01 -2.77776361e-01 6.82958841e-01
3.28674972e-01 3.89748693e-01 1.44755626e+00 -1.43560672e+00
-5.91171563e-01 1.06000876e+00 -3.94838154e-02 -2.96727180e-01
6.36136532e-01 -2.31978118e-01 1.72661662e+00 -9.03840423e-01
7.29379475e-01 1.63476110e+00 3.73629659e-01 3.75207037e-01
-1.84684825e+00 -5.19587517e-01 2.75072724e-01 -1.20860189e-01
-1.21448433e+00 -4.68093932e-01 6.43315494e-01 -4.84078586e-01
4.60689813e-01 6.68277502e-01 2.65618086e-01 1.42976570e+00
2.75921673e-01 9.83609080e-01 1.70944095e+00 -4.41385716e-01
1.02441266e-01 4.81500514e-02 4.64423001e-01 2.65327990e-01
3.00888240e-01 4.00942296e-01 -5.47517300e-01 -7.15599537e-01
6.40844584e-01 2.43045628e-01 1.34569377e-01 -8.24907303e-01
-1.18612230e+00 1.19912767e+00 1.83485344e-01 -3.41425598e-01
-1.28879100e-01 4.34299320e-01 -1.38234660e-01 5.09095430e-01
2.67154574e-01 6.74914479e-01 -7.18770206e-01 1.31502569e-01
-6.61574006e-01 8.10396075e-01 9.57010627e-01 9.52390075e-01
8.44437242e-01 -5.76882064e-01 -4.30900395e-01 1.17687762e+00
6.92730010e-01 5.10612249e-01 3.04103255e-01 -1.02439737e+00
2.48222411e-01 1.79614842e-01 5.59143841e-01 -7.62577057e-01
-5.23071408e-01 -3.28306794e-01 -4.42005694e-01 1.69074893e-01
5.42482555e-01 2.31433660e-01 -7.94033051e-01 2.08385110e+00
-8.94344896e-02 -7.28085577e-01 -3.55853736e-01 5.78696847e-01
4.17176545e-01 3.78162891e-01 4.09315169e-01 1.54227111e-02
1.30225360e+00 -2.55963683e-01 -2.08503366e-01 -2.81248927e-01
5.26190996e-01 -6.56486630e-01 1.28531241e+00 4.88759339e-01
-8.25232565e-01 -1.52338564e-01 -7.18640685e-01 -1.23930156e-01
-3.58724952e-01 -6.98722675e-02 1.14359045e+00 6.00132942e-01
-1.04419780e+00 6.76219702e-01 -3.10775042e-01 -1.45290732e-01
5.57368919e-02 6.69909537e-01 -2.77710259e-01 -3.91633421e-01
-1.36915505e+00 1.18820620e+00 -1.11346185e-01 -3.39347541e-01
-9.40499365e-01 -2.67327189e-01 -9.60723042e-01 -8.47053528e-03
5.75586319e-01 -7.11578488e-01 1.54682183e+00 -4.26945537e-01
-1.22593045e+00 6.54344440e-01 5.11758178e-02 1.06002212e-01
3.18393767e-01 -2.48231404e-02 -3.96296442e-01 -6.62661910e-01
4.20566767e-01 7.72978067e-01 5.10155678e-01 -1.23583949e+00
-5.76114178e-01 -2.50255525e-01 3.78859252e-01 5.82868159e-01
1.25551313e-01 1.86565369e-01 -2.92876214e-02 -4.73736763e-01
1.46644011e-01 -1.23103845e+00 -5.36132574e-01 -3.49378645e-01
-4.00810003e-01 -7.00605512e-01 -3.27968568e-01 1.67952746e-01
1.41658354e+00 -1.87263036e+00 5.29423282e-02 4.91083235e-01
2.28714719e-01 -6.54956400e-01 -3.06742907e-01 5.22551775e-01
-1.97447285e-01 5.23425102e-01 -3.55663076e-02 -3.45712364e-01
9.61272001e-01 1.67540312e-01 -2.19250306e-01 4.65194583e-01
8.44671428e-02 9.18142557e-01 -7.92643487e-01 -4.63253200e-01
1.39185786e-01 -3.46746668e-02 -7.36406207e-01 -1.06726848e-01
-4.52057242e-01 -2.53048301e-01 -5.45703650e-01 7.18519568e-01
2.26820365e-01 -2.40887523e-01 4.75224942e-01 2.34533995e-01
-1.24771520e-01 8.51961613e-01 -1.48954082e+00 1.44765532e+00
-3.08709115e-01 1.99939638e-01 -2.83737957e-01 -5.09020805e-01
5.91008425e-01 -2.03748673e-01 3.95881951e-01 -5.72895408e-01
4.29829881e-02 3.84999931e-01 -1.09416589e-01 -1.12532023e-02
7.90877342e-01 -2.84835905e-01 -1.10547066e+00 4.02473837e-01
-9.01123583e-02 -2.89711803e-01 4.26681161e-01 8.83901492e-02
8.88784885e-01 1.69139206e-01 4.11283612e-01 -5.77068329e-01
-1.84498832e-01 1.13283552e-01 4.94277984e-01 1.28575540e+00
1.61763594e-01 5.92887402e-01 4.86103743e-01 -2.89139092e-01
-6.93027794e-01 -1.37350464e+00 -5.53251028e-01 1.74091756e+00
1.14336222e-01 -3.82890373e-01 1.36579219e-02 -3.15299958e-01
2.29086459e-01 7.67029762e-01 -4.76478159e-01 7.57799484e-03
-2.68170983e-01 -9.61014986e-01 4.06385124e-01 2.82004714e-01
-1.80580601e-01 -7.87831426e-01 -5.58925830e-02 2.87769079e-01
-2.40928978e-01 -3.38375121e-01 -5.70347607e-01 6.88740253e-01
-5.88699222e-01 -9.83047605e-01 -3.34957480e-01 -5.20586491e-01
1.26178011e-01 -2.16730032e-02 1.62184870e+00 -3.31879854e-01
2.13135290e-03 3.25539470e-01 -6.27286434e-02 -1.91855997e-01
-9.94700417e-02 -1.06022291e-01 4.04065430e-01 -3.98517489e-01
5.75902641e-01 -3.06843013e-01 -4.82582599e-01 4.91148889e-01
-9.64185476e-01 -4.18582499e-01 7.71599054e-01 9.72393811e-01
6.49142265e-01 -1.39001340e-01 3.19068730e-01 -9.80660796e-01
9.44919705e-01 -8.53675842e-01 -6.26150370e-01 3.04275960e-01
-1.03012133e+00 6.27752542e-01 2.49268711e-01 -6.27336025e-01
-7.08884418e-01 1.27561808e-01 3.15108538e-01 1.16750941e-01
-3.94986272e-02 7.73823738e-01 -2.25650325e-01 2.15395570e-01
8.43372703e-01 -2.90329784e-01 -1.47427261e-01 -4.85655755e-01
7.98119664e-01 6.27421260e-01 3.18737179e-01 -1.15331888e+00
6.88903332e-01 7.91724399e-02 1.60365969e-01 -2.48305529e-01
-5.08984506e-01 -3.40105116e-01 -2.44636104e-01 1.58991963e-01
4.28332865e-01 -9.33499873e-01 -7.33929753e-01 1.06012657e-01
-5.99531591e-01 -2.70880580e-01 -1.67590141e-01 6.77246809e-01
-7.20527470e-01 1.39695138e-01 -5.83457768e-01 -1.00183225e+00
4.15887773e-01 -1.23172259e+00 9.52329874e-01 -3.97472382e-02
-6.79154813e-01 -9.35664415e-01 2.99416065e-01 1.09763131e-01
3.67912471e-01 1.71132505e-01 1.33934641e+00 -7.30043411e-01
-7.04806566e-01 -2.63338625e-01 1.06845863e-01 -2.34595358e-01
-2.33155027e-01 -5.65968640e-02 -5.12387574e-01 -5.03025502e-02
-3.66520822e-01 -5.65175653e-01 1.12430978e+00 5.39960504e-01
9.23663676e-01 -1.90209463e-01 -3.30304205e-01 1.15702227e-01
1.33761227e+00 -6.85706139e-02 4.69178915e-01 2.31885716e-01
2.00339869e-01 7.82820344e-01 7.07061231e-01 4.74546611e-01
9.07003403e-01 8.62105966e-01 2.89585441e-01 1.80552810e-01
3.70658606e-01 -5.50504088e-01 4.02673274e-01 3.74298990e-01
2.29228675e-01 -3.86929214e-01 -6.57956600e-01 3.80170166e-01
-1.92761266e+00 -9.67430890e-01 -1.09574318e-01 2.66760111e+00
1.01652622e+00 2.66559780e-01 3.96894276e-01 -6.38643727e-02
6.11800909e-01 9.80369896e-02 -3.19569468e-01 -4.53388423e-01
-1.74961627e-01 3.59402061e-01 6.63300872e-01 8.35169315e-01
-1.18277001e+00 7.42435098e-01 7.57967949e+00 1.00367236e+00
-3.73626888e-01 -2.59024978e-01 7.35447526e-01 -2.02699140e-01
-1.11610138e+00 3.38988423e-01 -8.78918290e-01 3.82221788e-01
8.46835494e-01 2.87969094e-02 7.75836945e-01 9.19936061e-01
-2.74320245e-01 -8.85685384e-02 -1.69009829e+00 9.58391190e-01
-2.22806424e-01 -7.25187182e-01 -1.73566584e-03 2.36218840e-01
4.86415267e-01 -4.09121215e-02 5.10046601e-01 6.93666518e-01
1.48227692e+00 -1.51604784e+00 9.58267331e-01 5.15756607e-01
1.20270085e+00 -6.41529560e-01 2.90276438e-01 3.34514141e-01
-9.43950832e-01 -3.73179406e-01 -4.21644181e-01 -3.03842336e-01
-4.95065823e-02 6.32238150e-01 -4.19930637e-01 3.07847321e-01
2.56786525e-01 4.38414991e-01 -6.01956010e-01 8.06709886e-01
1.32017359e-01 2.46635139e-01 -6.42952025e-01 -3.42752039e-01
8.75589848e-02 -2.58092463e-01 2.90181220e-01 9.77132082e-01
3.05143654e-01 1.09677143e-01 3.00765693e-01 8.60624075e-01
-9.97233838e-02 -1.46709487e-01 -6.76498532e-01 -9.61007923e-03
8.40213180e-01 1.00144565e+00 -2.25884706e-01 9.80349705e-02
-2.99539208e-01 2.46452898e-01 4.53305304e-01 3.58599514e-01
-5.70217609e-01 1.65130831e-02 8.30885172e-01 1.65385351e-01
-7.95704722e-02 -1.55701995e-01 -1.11213923e-01 -1.28516960e+00
-1.29129648e-01 -9.96056974e-01 4.81119037e-01 -5.06314695e-01
-1.76352000e+00 2.62173593e-01 8.72540593e-01 -1.03871441e+00
-7.46554613e-01 -8.81982386e-01 -1.63440809e-01 8.43960583e-01
-1.36362362e+00 -9.45661306e-01 3.73639703e-01 5.42824864e-01
2.55726993e-01 1.62613243e-01 8.44766617e-01 1.28354788e-01
-2.74115026e-01 5.22228718e-01 3.89115214e-01 -3.65213782e-01
8.31374049e-01 -1.63787889e+00 1.41631812e-01 2.08457038e-01
1.38600349e-01 1.04644418e+00 6.21967077e-01 -4.94818389e-01
-1.45933890e+00 -5.61047018e-01 1.23810542e+00 -1.10223579e+00
4.82676208e-01 -5.13327479e-01 -1.24517649e-01 6.28616571e-01
-2.73325980e-01 -4.40245211e-01 8.62169504e-01 8.32636535e-01
-6.55004978e-01 -3.36105675e-02 -9.68574524e-01 9.90837216e-01
1.04453111e+00 -7.51946688e-01 -7.50865221e-01 2.56743371e-01
4.61013198e-01 -1.08546196e-02 -6.95807099e-01 3.93297255e-01
1.02245188e+00 -8.30429316e-01 1.23983312e+00 -8.46161962e-01
5.02411723e-01 -1.95799083e-01 -8.28040838e-01 -1.32857859e+00
-8.52823734e-01 -7.17408895e-01 3.96615028e-01 9.22290921e-01
8.63675594e-01 -4.24898088e-01 3.46429884e-01 1.34722674e+00
7.83369541e-02 -7.10129261e-01 -1.01626611e+00 -5.36937475e-01
4.79041576e-01 -5.19665956e-01 8.04110706e-01 7.97599077e-01
2.98241898e-02 3.90392691e-01 -5.01321197e-01 -2.75037467e-01
8.27185929e-01 6.65529668e-02 6.62001491e-01 -1.68537164e+00
-5.12895584e-01 -7.27915466e-01 -5.38641028e-02 -1.39011228e+00
2.31839776e-01 -1.04917932e+00 1.80207223e-01 -1.39122128e+00
4.56289411e-01 -9.32560742e-01 -5.21421075e-01 3.33458543e-01
-1.84175879e-01 -4.16571870e-02 7.12989643e-02 3.42130631e-01
-8.88888419e-01 3.48893136e-01 8.45848978e-01 -5.67204580e-02
-3.20114270e-02 2.51906157e-01 -1.37374020e+00 5.79248548e-01
5.49398780e-01 -7.92248189e-01 -5.18911481e-01 -2.54476249e-01
7.15962887e-01 4.23125267e-01 3.46939534e-01 -3.63092989e-01
-2.71704979e-02 -7.52228081e-01 3.58456552e-01 -3.64498347e-01
5.61988950e-01 -6.22359037e-01 5.03752947e-01 -3.42230126e-02
-9.98009920e-01 3.51342738e-01 -4.46122527e-01 7.85667300e-01
1.96841881e-01 -2.29858249e-01 2.47734994e-01 -1.95566460e-01
-5.20455062e-01 3.10492605e-01 -6.16757512e-01 6.69409186e-02
2.97568321e-01 1.85240388e-01 -3.38890791e-01 -6.29855990e-01
-8.07289600e-01 4.24674034e-01 6.55066490e-01 3.63638103e-01
3.13906223e-01 -1.68997824e+00 -7.75789976e-01 -9.52446014e-02
3.52088749e-01 -1.57691196e-01 -4.05547500e-01 3.64619553e-01
7.60398358e-02 5.40139675e-01 2.73316707e-02 -2.15546370e-01
-7.29011774e-01 3.87770891e-01 8.34160950e-03 -6.97785318e-01
1.49740368e-01 7.27952123e-01 2.38292009e-01 -1.00531209e+00
6.63648695e-02 -4.49105054e-01 -2.71041989e-01 -1.92104671e-02
-6.79340810e-02 4.04705226e-01 -3.99052113e-01 -4.92093801e-01
-4.34261739e-01 3.93155426e-01 -1.51078045e-01 -6.52047992e-01
9.83152390e-01 -2.82234550e-01 -3.86582278e-02 6.01329207e-01
9.59686160e-01 3.72029506e-02 -9.90086436e-01 -2.64397621e-01
2.35054776e-01 -4.28617120e-01 -2.71299720e-01 -9.76961017e-01
-2.18278691e-02 3.09870869e-01 3.09059322e-01 1.71135545e-01
5.57640553e-01 1.45702183e-01 8.98467377e-02 6.37843132e-01
7.18107343e-01 -1.07725835e+00 -2.78407693e-01 4.51321751e-01
8.23493481e-01 -1.44461095e+00 7.96703994e-02 -2.01241523e-01
-6.69802070e-01 6.83901131e-01 1.54253870e-01 -2.30550319e-01
8.54639053e-01 -2.11227670e-01 -3.11664999e-01 -1.94246233e-01
-1.11610401e+00 -2.76766002e-01 5.82652569e-01 5.27630568e-01
5.17424285e-01 7.42489636e-01 -4.84392136e-01 1.07961535e+00
-3.70539188e-01 -3.21739376e-01 4.68332440e-01 6.93210125e-01
-3.57657939e-01 -1.63550997e+00 -2.47612938e-01 1.05595207e+00
-3.38457465e-01 -3.75759959e-01 -5.23517430e-01 8.28959405e-01
-1.75815016e-01 1.42305541e+00 -1.57170266e-01 -5.61102271e-01
2.10232824e-01 1.03342995e-01 4.09500957e-01 -7.68461227e-01
-4.00579870e-01 1.70561910e-01 4.81608421e-01 -5.54674387e-01
-2.61755705e-01 -9.74637687e-01 -5.31017184e-01 -4.51569974e-01
-5.18138945e-01 2.55093455e-01 5.81526101e-01 8.08872759e-01
1.04894660e-01 -8.72337669e-02 7.48397768e-01 -6.86931491e-01
-1.40990567e+00 -9.05313253e-01 -9.73621249e-01 6.11413717e-01
2.67014027e-01 -1.03635550e+00 -7.21635699e-01 -6.74430132e-01] | [9.295289039611816, 5.470141410827637] |
7a18a03d-a12f-4ae1-b3bc-84108d6d8f1f | geovln-learning-geometry-enhanced-visual-1 | 2305.17102 | null | https://arxiv.org/abs/2305.17102v1 | https://arxiv.org/pdf/2305.17102v1.pdf | GeoVLN: Learning Geometry-Enhanced Visual Representation with Slot Attention for Vision-and-Language Navigation | Most existing works solving Room-to-Room VLN problem only utilize RGB images and do not consider local context around candidate views, which lack sufficient visual cues about surrounding environment. Moreover, natural language contains complex semantic information thus its correlations with visual inputs are hard to model merely with cross attention. In this paper, we propose GeoVLN, which learns Geometry-enhanced visual representation based on slot attention for robust Visual-and-Language Navigation. The RGB images are compensated with the corresponding depth maps and normal maps predicted by Omnidata as visual inputs. Technically, we introduce a two-stage module that combine local slot attention and CLIP model to produce geometry-enhanced representation from such input. We employ V&L BERT to learn a cross-modal representation that incorporate both language and vision informations. Additionally, a novel multiway attention module is designed, encouraging different phrases of input instruction to exploit the most related features from visual input. Extensive experiments demonstrate the effectiveness of our newly designed modules and show the compelling performance of the proposed method. | ['Yanwei Fu', 'Haitao Lin', 'Boyan Jiang', 'Qiang Sun', 'Jingyang Huo'] | 2023-05-26 | geovln-learning-geometry-enhanced-visual | http://openaccess.thecvf.com//content/CVPR2023/html/Huo_GeoVLN_Learning_Geometry-Enhanced_Visual_Representation_With_Slot_Attention_for_Vision-and-Language_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Huo_GeoVLN_Learning_Geometry-Enhanced_Visual_Representation_With_Slot_Attention_for_Vision-and-Language_CVPR_2023_paper.pdf | cvpr-2023-1 | ['vision-and-language-navigation'] | ['robots'] | [-6.93474859e-02 1.93643454e-03 -1.04517587e-01 -5.59371889e-01
-5.74674428e-01 -4.57563400e-01 4.32425946e-01 -3.70466746e-02
-2.40339860e-01 3.62268120e-01 5.59293330e-01 -4.10945296e-01
1.51446849e-01 -8.39687645e-01 -1.05346048e+00 -4.02183384e-01
6.57870948e-01 -7.14224502e-02 2.97728688e-01 -6.23056710e-01
4.71738666e-01 4.16216254e-01 -1.64656317e+00 3.91662776e-01
8.79287958e-01 1.05587566e+00 9.54004288e-01 6.69384301e-01
-5.73753119e-01 9.81911302e-01 -2.46637076e-01 2.57113129e-01
2.11130008e-01 -1.98520750e-01 -5.11973500e-01 1.44173289e-02
6.31306708e-01 -3.02745163e-01 -5.15892863e-01 9.01051581e-01
4.99103904e-01 6.46580338e-01 4.89271164e-01 -1.00912273e+00
-9.74057853e-01 1.89377680e-01 -6.76425099e-01 6.83395118e-02
7.09120691e-01 3.20738345e-01 6.43501699e-01 -1.20463204e+00
4.83838826e-01 1.40009427e+00 2.33890563e-01 4.85444099e-01
-6.94032073e-01 -4.91930753e-01 7.95232415e-01 3.26377809e-01
-1.24731827e+00 -1.70665130e-01 1.18564034e+00 -8.12857896e-02
1.01669991e+00 1.50328472e-01 6.87955022e-01 9.19479907e-01
1.37303725e-01 9.58754778e-01 1.14110649e+00 -4.74310607e-01
3.41817513e-02 4.39655304e-01 -1.32845089e-01 9.85946238e-01
-1.60429984e-01 -1.16173655e-01 -6.27168298e-01 5.30610383e-01
1.04061317e+00 5.74308932e-01 -5.98174334e-01 -1.01405764e+00
-1.18143964e+00 5.41008830e-01 1.10669923e+00 2.09716022e-01
-1.75019652e-01 1.91170111e-01 7.26762936e-02 -1.48469120e-01
-1.22348387e-02 2.07131013e-01 -4.73130286e-01 2.90985137e-01
-3.72268736e-01 -1.09176032e-01 -2.33025365e-02 1.44765258e+00
9.89414811e-01 1.21806055e-01 -1.52123377e-01 5.70719302e-01
7.80438662e-01 6.66110396e-01 6.13914430e-01 -8.43982697e-01
8.27945590e-01 7.76475608e-01 -1.09618828e-01 -1.17655861e+00
-5.36039174e-01 -2.51834840e-01 -6.66499972e-01 4.20004159e-01
6.21625781e-02 4.45545048e-01 -1.03153121e+00 1.65797102e+00
3.87969494e-01 -3.63809541e-02 2.99653232e-01 1.12541485e+00
1.29522657e+00 6.85625196e-01 1.12818532e-01 1.58200741e-01
1.16107988e+00 -1.46497750e+00 -8.75439227e-01 -6.14284933e-01
4.10985827e-01 -7.33193755e-01 1.54825532e+00 1.98853254e-01
-9.21894789e-01 -1.08710039e+00 -1.12080646e+00 -6.68732882e-01
-7.94797897e-01 1.78916529e-01 6.47570372e-01 1.57652199e-01
-9.85397339e-01 9.67055112e-02 -5.76509058e-01 -3.67240846e-01
1.86759144e-01 2.31312826e-01 -6.11597598e-01 -4.37201947e-01
-8.39545190e-01 6.82665229e-01 2.99323022e-01 6.27345741e-01
-9.27198052e-01 -3.17084312e-01 -1.42110801e+00 -1.54109299e-01
4.22914326e-01 -7.09284127e-01 9.46524680e-01 -7.08580017e-01
-1.36042345e+00 5.48121393e-01 -4.16799039e-01 4.07701850e-01
1.71637475e-01 -2.91414410e-01 -6.37687966e-02 4.78280149e-02
2.14734480e-01 7.41991937e-01 3.55240673e-01 -1.81995916e+00
-4.14725810e-01 -6.47106826e-01 3.66068363e-01 7.92790174e-01
-4.52972949e-03 -6.79099917e-01 -7.18686104e-01 -2.56632149e-01
7.26403534e-01 -3.72433662e-01 -5.17613292e-01 4.04740013e-02
-3.77410918e-01 -1.47971362e-01 6.53445244e-01 -5.82966506e-01
9.64223981e-01 -1.99190223e+00 1.58268303e-01 9.39906910e-02
1.84225306e-01 -9.75706652e-02 -1.63055196e-01 2.84834236e-01
-7.56971212e-03 -2.16101259e-01 1.87972710e-01 -3.60148698e-01
-1.88103199e-01 2.87682563e-01 -3.37055355e-01 3.08408439e-01
-4.21002917e-02 1.08506751e+00 -1.05155766e+00 -4.94717389e-01
7.58541405e-01 5.59962332e-01 -6.75553858e-01 5.89867055e-01
-2.58134484e-01 5.17485321e-01 -7.22494245e-01 7.02218354e-01
7.64649272e-01 -2.49438658e-02 -5.92352785e-02 -4.11935061e-01
-2.25321874e-01 8.28139484e-02 -9.37481761e-01 2.36037326e+00
-8.75274062e-01 4.90834266e-01 -3.16862240e-02 -7.80570388e-01
1.05854285e+00 -1.40627757e-01 -1.58472210e-01 -1.13559985e+00
2.20671877e-01 4.67241945e-04 -6.12717628e-01 -8.53057623e-01
6.10827982e-01 1.65654883e-01 -6.60526752e-02 4.50535081e-02
-1.78734194e-02 -3.78708571e-01 -6.94966733e-01 1.35575861e-01
6.82686925e-01 6.96024656e-01 3.67134720e-01 7.11409934e-03
7.63717353e-01 -9.39735249e-02 2.26832956e-01 5.99283457e-01
-2.63835579e-01 7.58637011e-01 3.75521332e-02 -4.08213943e-01
-7.47892678e-01 -1.12519670e+00 2.99076617e-01 1.16789794e+00
7.44603693e-01 -2.72753060e-01 -4.94065434e-01 -7.70304918e-01
-3.66747230e-01 7.09210515e-01 -8.00313115e-01 -9.83877331e-02
-5.55342734e-01 -5.00095896e-02 -2.16989785e-01 1.04996097e+00
5.31556427e-01 -1.26125741e+00 -6.93864942e-01 -2.44923353e-01
-2.35227108e-01 -1.10323584e+00 -5.15229106e-01 3.87877613e-01
-7.84139633e-01 -1.09934819e+00 -5.21241367e-01 -1.16064668e+00
9.90452170e-01 8.48627388e-01 1.04352844e+00 7.68782524e-03
-2.50398427e-01 6.87321723e-01 -3.21374387e-01 -1.54985264e-01
3.69449824e-01 -2.56081104e-01 -1.34260967e-01 -2.80103147e-01
3.60328972e-01 -3.19368839e-01 -8.49408507e-01 4.19307686e-02
-5.38319468e-01 2.52899826e-01 7.83872008e-01 7.87202418e-01
8.01019609e-01 -4.81914848e-01 7.03630969e-02 -5.55698156e-01
2.50153869e-01 -3.54660362e-01 -4.44924235e-01 3.74302745e-01
-2.74406165e-01 2.15085000e-01 6.81371272e-01 -5.51347136e-02
-1.29738057e+00 3.73543143e-01 -1.55266181e-01 -7.14713037e-01
-4.26742613e-01 1.45615995e-01 -7.62056351e-01 -3.71990129e-02
2.14214906e-01 4.58030760e-01 -2.26519048e-01 -2.96092480e-01
7.65917718e-01 3.51141572e-01 6.39999986e-01 -5.24492383e-01
6.29815459e-01 4.13854092e-01 -2.10246876e-01 -5.68991542e-01
-9.79969144e-01 -6.09929919e-01 -7.84589887e-01 -2.61474252e-01
1.26810920e+00 -1.15898860e+00 -1.01545954e+00 1.03045940e-01
-1.21663737e+00 -2.93926656e-01 -1.31287351e-01 6.60374999e-01
-7.26772428e-01 1.73166126e-01 -2.47631207e-01 -8.70509982e-01
1.51623622e-01 -1.31219268e+00 1.31059325e+00 6.01908147e-01
3.86933744e-01 -9.66165602e-01 -1.09909847e-01 5.57663918e-01
2.82772899e-01 -2.51809750e-02 7.71193743e-01 -1.58029348e-01
-1.14327252e+00 2.39543527e-01 -5.56298494e-01 -6.44646771e-03
7.20356852e-02 -3.05386484e-01 -1.20910585e+00 8.24396440e-04
-6.23195991e-02 -5.78563452e-01 9.00331795e-01 4.16555583e-01
1.57343721e+00 1.20208129e-01 -3.41527581e-01 9.51014459e-01
1.82364333e+00 3.00400317e-01 4.96834934e-01 4.74776775e-01
1.31486046e+00 7.81138003e-01 7.70180166e-01 2.04452738e-01
8.27162981e-01 5.70427239e-01 9.10661161e-01 -4.30605620e-01
3.59883383e-02 -6.42115355e-01 2.59056300e-01 1.13272464e+00
-1.73480660e-01 -1.14438742e-01 -7.36599088e-01 3.96681637e-01
-1.81599510e+00 -6.70970798e-01 1.16629694e-02 1.77156532e+00
3.47447991e-01 2.56484165e-03 -5.63345671e-01 -2.64694840e-01
1.69421285e-01 2.95339376e-01 -5.59269547e-01 -4.57399875e-01
-3.49205546e-02 1.09258577e-01 4.19788003e-01 6.44683540e-01
-8.82842302e-01 1.09642768e+00 5.45266390e+00 5.46140015e-01
-8.63032699e-01 -4.99588102e-02 5.15527844e-01 9.68748629e-02
-8.98578882e-01 -2.35290423e-01 -7.83214867e-01 -1.83117315e-01
2.30547190e-01 2.78340727e-01 2.76130050e-01 1.21144629e+00
1.23185344e-01 -2.15831667e-01 -8.86200249e-01 1.21578968e+00
4.36824203e-01 -1.12065959e+00 5.76692633e-02 -2.03880057e-01
7.93914437e-01 -3.05103481e-01 3.15181941e-01 6.25063479e-01
3.02549183e-01 -1.20548618e+00 7.52849579e-01 8.25725973e-01
6.24206960e-01 -8.14440787e-01 6.91313326e-01 1.71055242e-01
-1.67698026e+00 -2.53867745e-01 -5.92431247e-01 -8.15029293e-02
-1.01393852e-02 -1.75293699e-01 -6.56468749e-01 7.16616333e-01
9.12449419e-01 8.28514159e-01 -7.67492533e-01 6.42835379e-01
-3.82693946e-01 -6.43625706e-02 3.64367217e-02 -1.76432207e-01
4.83983725e-01 -1.56579122e-01 -6.24177456e-02 9.03673887e-01
3.87405992e-01 2.56810546e-01 3.41299951e-01 9.47287202e-01
3.02587211e-01 2.78038830e-01 -1.07666779e+00 5.11328161e-01
1.88897237e-01 1.14361441e+00 -6.96190774e-01 -2.12975532e-01
-7.54662156e-01 1.06920576e+00 5.78312099e-01 7.97065139e-01
-8.68829131e-01 -2.18792662e-01 5.88971078e-01 -2.50139926e-02
4.13676322e-01 -3.32717717e-01 -2.85466075e-01 -1.15370536e+00
7.32838660e-02 -4.95571256e-01 1.91276241e-02 -1.47543740e+00
-9.40955281e-01 7.61099875e-01 -2.22625658e-01 -1.30455267e+00
3.76704559e-02 -9.07759011e-01 -6.07501805e-01 9.95749533e-01
-1.81404853e+00 -1.32720554e+00 -9.73480523e-01 9.03194308e-01
9.04028058e-01 -2.36562770e-02 6.86471939e-01 -1.05310604e-01
-4.41572547e-01 4.44828719e-01 -1.24004766e-01 -9.48521867e-03
6.58738196e-01 -1.38228083e+00 4.82419953e-02 6.78329349e-01
-2.71867421e-02 8.19099844e-01 4.39582467e-01 -3.44601661e-01
-1.53396428e+00 -9.80882585e-01 5.22258162e-01 -6.56163633e-01
1.62835300e-01 -5.29420435e-01 -6.29144549e-01 8.79492104e-01
3.05529386e-01 3.85356694e-01 5.11724412e-01 -5.16600013e-02
-4.29414690e-01 -1.18645482e-01 -7.62881815e-01 6.93464220e-01
1.28099763e+00 -7.79103935e-01 -6.88442945e-01 1.20851077e-01
1.18376803e+00 -6.64021432e-01 -2.76883334e-01 3.05627078e-01
4.55210090e-01 -1.29437959e+00 1.27822959e+00 -4.84995067e-01
6.36536181e-01 -6.25200629e-01 -7.42183685e-01 -1.03823841e+00
-2.07207903e-01 1.44370869e-01 1.78657323e-01 1.15799999e+00
2.26125076e-01 -1.92842901e-01 8.23143423e-01 3.55988681e-01
-4.91769433e-01 -9.60961580e-01 -5.44546008e-01 -1.80833712e-01
-2.79660285e-01 -6.51032567e-01 5.33184528e-01 8.27166319e-01
-8.16898718e-02 4.32370245e-01 -2.19352141e-01 2.74280548e-01
4.08874035e-01 3.30922067e-01 9.34112549e-01 -7.05698311e-01
-1.08591795e-01 -2.87715882e-01 -3.71195644e-01 -1.67369533e+00
2.28970975e-01 -7.27939188e-01 4.60027963e-01 -1.87060630e+00
2.05132276e-01 -3.40188801e-01 -4.87232000e-01 3.23085427e-01
-3.86561751e-01 1.53501257e-01 3.04883689e-01 -2.87278384e-01
-8.75593245e-01 1.06247020e+00 1.72561133e+00 -1.58471987e-01
-2.21043661e-01 -4.10369903e-01 -7.26678491e-01 8.09469283e-01
5.40928125e-01 2.23525822e-01 -9.19098318e-01 -5.94645381e-01
9.46200266e-02 1.97542354e-01 4.80194151e-01 -1.12953413e+00
2.14885488e-01 -4.89209950e-01 9.21514630e-01 -1.05754232e+00
5.62208295e-01 -1.07738864e+00 -4.42211837e-01 7.45221898e-02
-3.52263659e-01 4.20340389e-01 2.96919346e-01 8.34771812e-01
-3.45248163e-01 6.98562562e-02 2.67959118e-01 -5.00774801e-01
-1.34089363e+00 2.60486245e-01 -7.00297803e-02 -1.75530374e-01
1.00598311e+00 -2.90185899e-01 -2.05594689e-01 -3.49594086e-01
-7.18658745e-01 4.50685352e-01 4.60475236e-01 6.80263281e-01
1.24537337e+00 -1.53233886e+00 6.99693710e-02 6.89472497e-01
4.89656240e-01 6.51229262e-01 7.73881495e-01 4.62892175e-01
-6.69654787e-01 6.51195526e-01 -3.91715825e-01 -6.75158381e-01
-1.08512521e+00 1.09149981e+00 2.70616710e-01 -5.24853077e-03
-6.20197892e-01 1.12047362e+00 8.82629812e-01 -9.11112130e-01
3.72802883e-01 -8.15705538e-01 -6.31132722e-01 -3.23232144e-01
4.94122535e-01 -1.64294705e-01 -3.27636927e-01 -7.45624661e-01
-4.25041407e-01 1.09521341e+00 2.34619543e-01 1.21483468e-02
1.07066107e+00 -6.25283659e-01 9.68456790e-02 7.52371371e-01
1.18336749e+00 2.62555897e-01 -1.40630424e+00 -2.21531361e-01
-5.80084980e-01 -4.50340480e-01 -7.65576400e-03 -5.61779618e-01
-1.11521900e+00 1.44754350e+00 7.55218446e-01 -5.12765288e-01
1.17039275e+00 -2.48283539e-02 4.67969239e-01 3.37715507e-01
4.47455615e-01 -8.88174057e-01 5.44582605e-01 4.57667887e-01
9.82359827e-01 -1.62073076e+00 -2.10556746e-01 -4.61895198e-01
-6.06876969e-01 1.12594533e+00 1.53785968e+00 1.23333536e-01
5.68469942e-01 -1.00977726e-01 3.94732952e-01 -2.37150252e-01
-6.17963374e-01 -4.52002168e-01 3.59439909e-01 9.30661559e-01
4.69435215e-01 -4.00982872e-02 4.41217422e-01 7.09179938e-01
-1.53988913e-01 -3.91911566e-01 3.41494709e-01 8.57489824e-01
-6.13179326e-01 -6.17615163e-01 -4.26096767e-01 -1.72532961e-01
2.89259434e-01 -2.44236529e-01 -5.51698618e-02 7.70712018e-01
3.48579347e-01 7.95057714e-01 1.66186497e-01 -5.23707509e-01
5.49360037e-01 -2.00485930e-01 5.28752208e-01 -6.70441628e-01
-2.64522135e-01 5.24432659e-02 -5.20458460e-01 -1.08089864e+00
-3.80642384e-01 -2.06511900e-01 -1.51965511e+00 -4.12776601e-03
-2.22280435e-02 -1.28334433e-01 6.61104798e-01 8.99876237e-01
9.52915251e-02 1.11379802e+00 4.68091011e-01 -1.20930278e+00
-2.24092267e-02 -6.69998467e-01 -3.98305833e-01 3.06867093e-01
5.85209548e-01 -7.90931404e-01 -1.45548761e-01 -1.85862944e-01] | [4.463918685913086, 0.4149438440799713] |
9b5c3e67-6d22-4042-8793-73cdb0fc6414 | joint-speech-recognition-and-audio-captioning | 2202.01405 | null | https://arxiv.org/abs/2202.01405v1 | https://arxiv.org/pdf/2202.01405v1.pdf | Joint Speech Recognition and Audio Captioning | Speech samples recorded in both indoor and outdoor environments are often contaminated with secondary audio sources. Most end-to-end monaural speech recognition systems either remove these background sounds using speech enhancement or train noise-robust models. For better model interpretability and holistic understanding, we aim to bring together the growing field of automated audio captioning (AAC) and the thoroughly studied automatic speech recognition (ASR). The goal of AAC is to generate natural language descriptions of contents in audio samples. We propose several approaches for end-to-end joint modeling of ASR and AAC tasks and demonstrate their advantages over traditional approaches, which model these tasks independently. A major hurdle in evaluating our proposed approach is the lack of labeled audio datasets with both speech transcriptions and audio captions. Therefore we also create a multi-task dataset by mixing the clean speech Wall Street Journal corpus with multiple levels of background noises chosen from the AudioCaps dataset. We also perform extensive experimental evaluation and show improvements of our proposed methods as compared to existing state-of-the-art ASR and AAC methods. | ['Shinji Watanabe', 'Michael Hentschel', 'Yosuke Kashiwagi', 'Xuankai Chang', 'Emiru Tsunoo', 'Chaitanya Narisetty'] | 2022-02-03 | null | null | null | null | ['audio-captioning'] | ['audio'] | [ 3.70233715e-01 -3.49737316e-01 6.24224603e-01 -4.45521861e-01
-1.80343771e+00 -5.45787454e-01 5.97266018e-01 3.36655267e-02
-1.41069770e-01 7.12750196e-01 8.52894306e-01 -2.44590297e-01
2.75786728e-01 -2.04007495e-02 -7.10734725e-01 -4.24162775e-01
2.48252109e-01 2.10514933e-01 9.37910676e-02 -1.26308039e-01
-1.41467139e-01 1.86848119e-01 -1.88780522e+00 7.34895527e-01
7.20446765e-01 1.04703474e+00 5.73289514e-01 1.38435245e+00
-7.49496222e-02 9.56101000e-01 -9.39020038e-01 -2.55472809e-01
-4.74739335e-02 -4.86099303e-01 -5.71375132e-01 2.11294323e-01
4.49326545e-01 6.62656575e-02 -4.35888946e-01 8.89032125e-01
1.15511477e+00 4.51253444e-01 5.54447532e-01 -1.03439939e+00
-4.90007967e-01 7.01201439e-01 3.50695178e-02 1.83724225e-01
6.71295166e-01 7.29500800e-02 9.38003242e-01 -1.03886569e+00
-7.00353831e-02 1.25527012e+00 5.30795932e-01 7.24167168e-01
-1.21064341e+00 -7.24573731e-01 7.73945749e-02 3.92281413e-01
-1.40028644e+00 -1.35541463e+00 9.79896128e-01 -3.18687826e-01
9.43093479e-01 7.30348647e-01 2.98940629e-01 1.56066442e+00
-5.71878970e-01 9.69471753e-01 1.12209594e+00 -6.27345264e-01
4.48272556e-01 1.21054374e-01 9.83714825e-04 -2.68266678e-01
-5.25600970e-01 1.07151000e-02 -7.90890634e-01 -1.74617350e-01
2.17090368e-01 -6.04368985e-01 -5.99608004e-01 3.01871628e-01
-1.17078292e+00 1.68218955e-01 -5.89586012e-02 1.93232790e-01
-5.61880469e-01 3.13828588e-01 4.04747784e-01 1.58496886e-01
4.04334426e-01 1.66535825e-01 -5.27348578e-01 -5.72237968e-01
-1.25733030e+00 2.09483653e-01 6.01400912e-01 1.02697217e+00
1.19834885e-01 5.98255515e-01 -3.45431507e-01 1.59056258e+00
6.39167428e-01 5.02162576e-01 5.63916385e-01 -6.65122032e-01
5.93049109e-01 -4.64467287e-01 4.24441636e-01 -3.41227233e-01
1.45411953e-01 -7.60349810e-01 -6.21143162e-01 -8.92856941e-02
3.68167534e-02 -1.11099415e-01 -1.13582683e+00 1.69533002e+00
1.90980006e-02 6.10928714e-01 4.69571054e-01 8.23775113e-01
8.81211638e-01 1.15328074e+00 1.62612364e-01 -2.93992609e-01
1.33166313e+00 -1.24717784e+00 -1.31587493e+00 -3.57310951e-01
1.88751563e-01 -1.42512071e+00 1.26139915e+00 6.33808196e-01
-1.23021603e+00 -9.11309659e-01 -8.92562270e-01 -5.51782455e-03
-1.36703253e-01 4.65175569e-01 -1.55027499e-02 8.15828919e-01
-1.03323102e+00 3.49069387e-02 -6.26128852e-01 4.48011123e-02
1.91182271e-01 -6.56962469e-02 -1.74514592e-01 -2.37885192e-02
-1.02426028e+00 6.78497732e-01 -6.14795797e-02 1.44559160e-01
-1.51035106e+00 -6.45706058e-01 -8.78404558e-01 1.04003303e-01
1.15865223e-01 -3.45370620e-01 1.97093892e+00 -9.44117188e-01
-1.82458925e+00 3.90225053e-01 -5.85336745e-01 -6.44604743e-01
3.39208901e-01 -6.41732991e-01 -9.27156746e-01 -8.29751566e-02
-1.84781894e-01 4.76572901e-01 9.53283012e-01 -1.57690430e+00
-5.42872488e-01 -2.44130287e-02 -4.97693241e-01 3.20194930e-01
-1.30174845e-01 4.41292048e-01 -2.47603565e-01 -9.61078703e-01
-1.57163486e-01 -7.15041220e-01 -5.62439412e-02 -4.39400077e-01
-4.25698936e-01 1.23252831e-01 7.59851515e-01 -1.13140535e+00
1.18393254e+00 -2.30068421e+00 -1.50415033e-01 -3.10975015e-01
-3.91027749e-01 5.61010599e-01 -3.39615911e-01 4.33216453e-01
-3.40542883e-01 -7.62929544e-02 -1.53995827e-01 -1.05665767e+00
9.78975743e-02 -2.01563183e-02 -8.38056922e-01 1.31677449e-01
1.82046592e-01 2.66238570e-01 -7.52122223e-01 -3.22006524e-01
4.14014846e-01 1.06117153e+00 -3.70310903e-01 6.35656297e-01
-2.16262326e-01 5.86691678e-01 -2.14754958e-02 5.26934206e-01
4.29465353e-01 4.44533110e-01 -4.86025482e-01 -8.00370723e-02
-7.51984268e-02 9.93859768e-01 -1.38774335e+00 1.78676021e+00
-8.63980174e-01 8.85340452e-01 4.08640325e-01 -5.89125395e-01
1.05041683e+00 1.13191843e+00 -6.35284185e-02 -4.16886687e-01
6.86656334e-04 5.85974276e-01 -4.89976332e-02 -3.84278029e-01
3.53898346e-01 -2.62274683e-01 2.48360962e-01 -6.04119785e-02
1.80328920e-01 -5.02872586e-01 -2.21834213e-01 -1.39128834e-01
9.05304432e-01 -5.42586595e-02 3.59069332e-02 1.18011408e-01
8.66935670e-01 -4.25614119e-01 2.48554468e-01 7.40692675e-01
-3.48848283e-01 1.13033223e+00 -2.63149232e-01 1.70104519e-01
-1.04818296e+00 -1.21160972e+00 5.05590672e-03 1.08650911e+00
-4.80303049e-01 -6.10985816e-01 -8.23767185e-01 -1.16770431e-01
-7.36369491e-01 1.20638251e+00 -9.21747759e-02 2.05739245e-01
-3.56166244e-01 -3.72954756e-01 9.38190281e-01 5.00057638e-01
2.65674174e-01 -9.38649178e-01 1.70565218e-01 3.60628307e-01
-6.79182291e-01 -1.50338149e+00 -5.42430639e-01 1.02501087e-01
-3.13121110e-01 -3.68147314e-01 -1.00788021e+00 -9.69158530e-01
1.21253483e-01 2.05322132e-01 9.66328859e-01 -4.66662377e-01
-4.72849831e-02 6.20153606e-01 -5.77920556e-01 -8.14509571e-01
-1.03831732e+00 -2.79199600e-01 2.46462911e-01 4.37779278e-01
5.90815321e-02 -8.34021807e-01 -3.37601691e-01 4.06505167e-01
-7.01705754e-01 1.93346888e-01 4.46910173e-01 6.03532732e-01
7.48848379e-01 -3.25936414e-02 9.57502425e-01 -3.01673934e-02
9.23444629e-01 -1.69692606e-01 -3.83738071e-01 9.09454469e-03
5.86639792e-02 -3.66552919e-01 8.11982393e-01 -5.94492197e-01
-1.28048372e+00 1.55514732e-01 -7.76328921e-01 -4.99659449e-01
-8.16031098e-01 2.43783981e-01 -5.42707264e-01 4.18529242e-01
6.95265472e-01 6.53766453e-01 -3.84935617e-01 -8.91141415e-01
2.95031786e-01 1.59825420e+00 8.17861676e-01 -5.43448269e-01
5.17456710e-01 -2.63317600e-02 -4.31868821e-01 -1.37455773e+00
-6.54977441e-01 -6.64561868e-01 -7.81261995e-02 -2.37424120e-01
6.31775856e-01 -1.40461707e+00 -2.61858523e-01 5.41706026e-01
-1.41704464e+00 -2.97609776e-01 -1.95666432e-01 8.08016419e-01
-8.84032667e-01 2.89522141e-01 -5.44700563e-01 -1.57711995e+00
-3.82838994e-01 -1.42747056e+00 1.26801145e+00 -1.93913266e-01
-2.97872931e-01 -4.61588144e-01 1.65368691e-01 7.11781979e-01
6.39169872e-01 -3.41064245e-01 4.83850449e-01 -9.77376878e-01
-2.13819370e-01 -2.97528744e-01 2.00310037e-01 9.71355081e-01
3.24958652e-01 -1.71268031e-01 -1.74776399e+00 -6.60344064e-02
1.50011405e-01 -3.67532074e-01 3.96544009e-01 3.67084503e-01
1.26120770e+00 -3.64956975e-01 1.71692014e-01 7.32358620e-02
9.86502886e-01 5.09146392e-01 7.80380309e-01 -7.35807195e-02
4.55008417e-01 5.03137887e-01 5.85338593e-01 2.87880272e-01
6.63311109e-02 9.16094422e-01 2.63798058e-01 2.48578284e-02
-7.64089227e-01 -3.56650263e-01 5.66701412e-01 1.53177798e+00
2.76445836e-01 -5.83562732e-01 -9.50988948e-01 9.68521833e-01
-1.48670387e+00 -7.74150789e-01 -1.54482886e-01 2.26459455e+00
1.05933857e+00 -4.43405658e-02 3.60089988e-02 7.54065335e-01
8.58165443e-01 2.82383002e-02 5.63248023e-02 -4.24161851e-01
-1.15544006e-01 2.79193014e-01 -1.65336743e-01 8.53954017e-01
-1.05290425e+00 6.40420258e-01 6.33642244e+00 1.17372441e+00
-8.70973229e-01 2.54394054e-01 5.58757603e-01 -2.39297763e-01
-6.72038272e-02 -3.68383437e-01 -5.38188636e-01 3.42320502e-01
1.54145789e+00 1.28482640e-01 7.05856502e-01 7.88698137e-01
6.62008941e-01 3.88431370e-01 -1.28070545e+00 1.31240523e+00
6.49066735e-03 -9.50065494e-01 9.67773125e-02 -3.56543720e-01
6.58191621e-01 1.87151253e-01 1.23564981e-01 3.22530299e-01
2.57835202e-02 -1.10660875e+00 1.11134768e+00 3.79004031e-01
6.55882120e-01 -5.29497206e-01 4.98570532e-01 2.60787040e-01
-1.26526785e+00 5.59927262e-02 4.48216796e-02 -8.64112750e-02
3.52125823e-01 2.71520585e-01 -1.05760312e+00 4.09892887e-01
6.27115726e-01 1.90293163e-01 -2.23183483e-01 1.51211488e+00
-2.24715695e-01 1.19412136e+00 -3.23605657e-01 -2.83925664e-02
8.70959461e-02 3.03290248e-01 1.04517674e+00 1.65851247e+00
4.97188002e-01 -1.63460702e-01 -2.08720818e-01 6.33549511e-01
-8.07298943e-02 4.14602846e-01 -3.78549397e-01 -3.11160237e-01
6.47440135e-01 7.73477793e-01 -1.29911244e-01 -3.76221180e-01
-3.62659246e-01 8.24983358e-01 -3.50389510e-01 5.61753213e-01
-8.11398983e-01 -4.53674227e-01 9.47608054e-01 -7.13357478e-02
3.70109111e-01 -3.23185861e-01 -1.07880995e-01 -1.00626802e+00
9.24431160e-02 -1.24250102e+00 -2.07042545e-01 -1.43497241e+00
-1.21239352e+00 9.85663950e-01 -1.99593350e-01 -1.38451970e+00
-3.35705221e-01 -4.38383013e-01 -5.09674788e-01 1.11415339e+00
-1.58167362e+00 -1.14858961e+00 -2.37589508e-01 5.67438722e-01
1.40256906e+00 -2.52519667e-01 1.09738231e+00 7.27396548e-01
-2.48313904e-01 4.78306115e-01 2.35102460e-01 -1.62632787e-03
7.32162595e-01 -1.14046693e+00 6.55589581e-01 1.08130372e+00
4.90079492e-01 2.99320072e-01 1.28643954e+00 -3.08409601e-01
-1.06958675e+00 -1.30354333e+00 9.06508446e-01 -4.34962958e-01
6.25369310e-01 -7.93427050e-01 -9.57897127e-01 6.18886471e-01
3.45596999e-01 3.30345854e-02 9.43874419e-01 -2.23348796e-01
-1.62662089e-01 -2.52063215e-01 -7.43781269e-01 6.83029652e-01
6.70992494e-01 -8.85817170e-01 -8.73509407e-01 2.64843762e-01
1.02130938e+00 -2.42681414e-01 -4.46111619e-01 2.69677639e-01
2.97671705e-01 -6.50579154e-01 9.56524193e-01 -5.67312419e-01
1.55428350e-01 -6.40245080e-01 -7.93552458e-01 -1.63026035e+00
2.59018064e-01 -1.14808071e+00 -4.21990529e-02 1.76337445e+00
5.04053414e-01 -2.82162637e-01 2.88311571e-01 2.50715613e-01
-6.92376912e-01 -1.66043952e-01 -1.08958256e+00 -8.20741773e-01
-2.70232707e-01 -1.31195736e+00 5.73044121e-01 5.29677331e-01
-3.26544613e-01 5.03942251e-01 -7.53563941e-01 5.00759006e-01
5.18287301e-01 -4.52586323e-01 8.81312490e-01 -8.33209097e-01
-4.40287292e-01 -1.19479544e-01 -3.81782293e-01 -1.06086957e+00
-1.12921365e-01 -4.33719695e-01 6.21059895e-01 -1.66994989e+00
-4.83608335e-01 -7.77242258e-02 -2.85404176e-01 2.26437658e-01
1.04417644e-01 2.22936288e-01 1.53224126e-01 -9.97344628e-02
-4.38680142e-01 9.91812885e-01 9.35896158e-01 -1.65409401e-01
-3.08876246e-01 1.78130001e-01 -4.47522819e-01 6.57122254e-01
6.36508882e-01 -5.42060733e-01 -5.53597331e-01 -4.36351478e-01
-2.68137783e-01 3.58488530e-01 3.23783666e-01 -1.41268933e+00
2.06348285e-01 2.13903829e-01 9.65289655e-04 -6.39393210e-01
1.07743490e+00 -8.54490399e-01 2.42382273e-01 -1.51070789e-01
-7.51467109e-01 -3.82767558e-01 5.40193081e-01 6.03367567e-01
-6.36225462e-01 -1.07908577e-01 7.80571282e-01 1.06215321e-01
-1.60646647e-01 -1.79821461e-01 -8.25740635e-01 -5.78662865e-02
4.57484990e-01 8.48184079e-02 -1.02858044e-01 -8.91245127e-01
-1.05723226e+00 -1.85045034e-01 -1.27526537e-01 7.29659855e-01
7.59759188e-01 -1.28098917e+00 -1.07447851e+00 1.40204519e-01
4.47685756e-02 -1.04760669e-01 3.47591758e-01 5.62728107e-01
-1.47706047e-01 5.17814457e-01 1.03405811e-01 -4.90795344e-01
-1.53208244e+00 3.06626141e-01 4.49890703e-01 3.34678292e-01
-2.63288766e-01 8.35256279e-01 2.27098778e-01 -1.48726463e-01
7.65774965e-01 -3.45026165e-01 -2.65703857e-01 -3.14786106e-01
8.72439027e-01 3.90573919e-01 3.90842944e-01 -9.39673126e-01
-2.14499280e-01 4.26411815e-02 8.87553841e-02 -6.70706153e-01
1.31154752e+00 -3.42366457e-01 2.99807906e-01 8.06583226e-01
1.08998728e+00 3.08292955e-01 -8.60141814e-01 -1.67358845e-01
-3.77561420e-01 -4.91674215e-01 3.51211607e-01 -1.06734478e+00
-3.48064482e-01 1.27978539e+00 9.00942147e-01 3.34544986e-01
1.21862197e+00 -1.49758011e-01 8.42099667e-01 4.06864941e-01
7.39060864e-02 -1.01253796e+00 1.62482888e-01 3.79949600e-01
1.35622394e+00 -9.77231741e-01 -6.88578248e-01 -3.75562221e-01
-6.72925353e-01 8.41607034e-01 3.88104171e-01 2.38783687e-01
3.59200209e-01 3.31053883e-01 4.04615134e-01 4.60139632e-01
-8.58243048e-01 -3.52351427e-01 3.96157086e-01 8.65343750e-01
5.21663368e-01 -9.63903442e-02 2.70866245e-01 1.05249882e+00
-5.45286357e-01 -2.02955455e-01 5.19996703e-01 6.91027284e-01
-3.65497112e-01 -1.06147242e+00 -8.70556891e-01 -3.82326245e-02
-8.39065969e-01 -3.19299668e-01 -3.82543802e-01 7.31112361e-02
-3.72037083e-01 1.57192647e+00 -2.22207263e-01 -3.47200900e-01
4.59259629e-01 5.90350330e-01 8.43818784e-02 -5.12934208e-01
-4.37691033e-01 6.59082353e-01 4.76923436e-01 -1.66739710e-02
-2.99687266e-01 -5.87022662e-01 -8.48310709e-01 5.21473527e-01
-4.29786593e-01 4.22281086e-01 1.19737089e+00 9.10424173e-01
3.56444538e-01 1.12285471e+00 5.46152353e-01 -1.03163350e+00
-5.01342177e-01 -1.42033815e+00 -3.57867718e-01 2.94891983e-01
8.24106753e-01 -2.91219145e-01 -5.71748674e-01 4.38023716e-01] | [14.871525764465332, 6.014614582061768] |
ab8dd45b-31e2-4240-81c3-9d2323bf7441 | emc2-net-joint-equalization-and-modulation | 2303.10934 | null | https://arxiv.org/abs/2303.10934v1 | https://arxiv.org/pdf/2303.10934v1.pdf | EMC2-Net: Joint Equalization and Modulation Classification based on Constellation Network | Modulation classification (MC) is the first step performed at the receiver side unless the modulation type is explicitly indicated by the transmitter. Machine learning techniques have been widely used for MC recently. In this paper, we propose a novel MC technique dubbed as Joint Equalization and Modulation Classification based on Constellation Network (EMC2-Net). Unlike prior works that considered the constellation points as an image, the proposed EMC2-Net directly uses a set of 2D constellation points to perform MC. In order to obtain clear and concrete constellation despite multipath fading channels, the proposed EMC2-Net consists of equalizer and classifier having separate and explainable roles via novel three-phase training and noise-curriculum pretraining. Numerical results with linear modulation types under different channel models show that the proposed EMC2-Net achieves the performance of state-of-the-art MC techniques with significantly less complexity. | ['Junil Choi', 'Hyun Ryu'] | 2023-03-20 | null | null | null | null | ['intelligent-communication'] | ['time-series'] | [ 6.63712680e-01 -8.21878165e-02 -3.12378019e-01 -2.86396652e-01
-6.20206952e-01 -8.66122171e-02 6.49980664e-01 -2.87858814e-01
-4.42451268e-01 9.45085645e-01 -3.58359963e-01 -9.77200031e-01
-1.96270451e-01 -4.25407231e-01 -6.01931751e-01 -9.82357502e-01
-6.48877561e-01 -2.18067273e-01 -2.24288732e-01 -2.15483069e-01
4.14579123e-01 4.29265290e-01 -1.25139749e+00 2.92947441e-01
9.55153048e-01 1.27450597e+00 1.88741311e-01 1.09830356e+00
2.56317139e-01 5.73125482e-01 -1.09427106e+00 -2.85859317e-01
3.81141275e-01 -6.97316349e-01 -4.90273759e-02 2.96184421e-01
1.28945559e-01 5.65016493e-02 -5.66527128e-01 1.14336073e+00
6.00246668e-01 -2.78518409e-01 9.63122189e-01 -1.30214429e+00
-2.52431482e-01 4.31188315e-01 -5.51511049e-01 1.63303494e-01
3.40424702e-02 -1.71480641e-01 4.59050536e-01 -4.99635637e-01
-8.78876634e-03 1.25800884e+00 6.40642881e-01 1.44045442e-01
-9.57377195e-01 -1.23258650e+00 -1.96987733e-01 8.07340264e-01
-1.73451483e+00 -3.67056459e-01 5.69666564e-01 -3.21381949e-02
6.89529479e-01 4.12558347e-01 2.79043734e-01 4.16458815e-01
6.57685995e-01 5.40451646e-01 1.53444433e+00 -9.72830534e-01
2.57407397e-01 1.24027245e-01 -1.73608869e-01 5.38740993e-01
1.50344968e-01 3.52600485e-01 -2.30942294e-02 -1.73785284e-01
5.14317453e-01 -6.05526865e-01 -5.30441880e-01 9.48996916e-02
-7.92069674e-01 7.66443968e-01 2.94332117e-01 7.47466162e-02
-1.40702769e-01 3.25791299e-01 -1.43645123e-01 8.12991798e-01
-1.44572249e-02 6.19773567e-02 -1.89973891e-01 1.11864299e-01
-9.64496315e-01 -2.52249688e-01 9.05453682e-01 1.27769315e+00
5.12657642e-01 5.05130649e-01 7.26846829e-02 4.41862166e-01
6.88174486e-01 5.69878340e-01 3.58055204e-01 -6.51244044e-01
5.85844696e-01 8.93841833e-02 -2.20056489e-01 -1.01431644e+00
-2.33737275e-01 -1.27701712e+00 -1.25612235e+00 5.11038661e-01
-5.27117066e-02 -4.44098979e-01 -1.07094264e+00 1.35779333e+00
-7.31160641e-02 7.11335719e-01 5.83512008e-01 5.80877781e-01
5.01189888e-01 7.64877260e-01 -9.02778283e-02 -3.02876860e-01
1.07753551e+00 -7.26814508e-01 -9.20622945e-01 2.50718202e-02
6.25946701e-01 -1.01341307e+00 -2.08457693e-01 6.91890359e-01
-7.27181673e-01 -6.67986631e-01 -1.82385993e+00 7.08002806e-01
-2.20155403e-01 3.01282257e-01 2.63032645e-01 1.60331428e+00
-8.76192033e-01 6.97204828e-01 -2.31840819e-01 3.00969839e-01
5.43745458e-01 9.95929599e-01 4.05945443e-03 -2.07113270e-02
-1.36061871e+00 1.10401917e+00 5.82494140e-01 2.26183608e-01
-7.06951678e-01 8.46801978e-03 -8.61529827e-01 -1.01442020e-02
5.02852015e-02 -2.12408677e-01 1.23226130e+00 -1.20354438e+00
-1.68766463e+00 7.92693868e-02 -1.86006591e-01 -8.91036212e-01
2.88361818e-01 1.38962016e-01 -1.17105353e+00 2.51851022e-01
-4.90145713e-01 5.23348272e-01 1.33393824e+00 -1.36469817e+00
-1.08783460e+00 2.50833452e-01 -2.70637006e-01 1.90701932e-01
1.70346454e-01 -1.54435217e-01 -3.83405536e-01 -7.84066021e-01
2.69439518e-01 -6.82677209e-01 -2.87052900e-01 -1.91545174e-01
-2.47805521e-01 2.71254033e-01 1.08700407e+00 -8.23937118e-01
1.58954930e+00 -2.06724024e+00 -2.17040762e-01 7.31239498e-01
-1.60995990e-01 6.28318548e-01 -3.32971886e-02 1.73973218e-01
-2.38376677e-01 -7.91090950e-02 -3.92547369e-01 -4.22977269e-01
4.64381650e-02 2.64664918e-01 2.90363599e-02 9.77454543e-01
6.13403544e-02 6.08439207e-01 -5.58134973e-01 -4.01176780e-01
4.22743589e-01 4.64007646e-01 -3.99681181e-01 -2.46441394e-01
9.86827984e-02 4.86674100e-01 4.68000732e-02 7.06088901e-01
1.29474699e+00 -1.23427860e-01 5.32412827e-01 -1.25941545e-01
5.93400151e-02 1.73941348e-02 -1.36154532e+00 9.58718538e-01
-4.15444642e-01 1.14140308e+00 2.58367181e-01 -1.43066812e+00
6.85125887e-01 7.59183943e-01 -1.23606622e-01 -8.02965939e-01
6.61893547e-01 5.30741632e-01 5.82429767e-01 -6.10163808e-02
4.48109120e-01 -2.13389307e-01 9.33639631e-02 -1.32114425e-01
1.56084985e-01 1.82236746e-01 -6.23375103e-02 -2.32526377e-01
5.91111064e-01 -1.77852318e-01 8.15697134e-01 -2.90248245e-02
9.52822924e-01 -3.63596082e-01 7.60464847e-01 9.43769157e-01
-1.00155629e-01 1.66170835e-01 2.43260842e-02 5.59012234e-01
-9.12123859e-01 -6.88631773e-01 -5.88820636e-01 1.53388247e-01
5.74267209e-01 -1.32738248e-01 -5.38819253e-01 -3.67817014e-01
-6.20202273e-02 7.79857159e-01 1.85981169e-02 -1.12057000e-03
-4.42986488e-01 -9.75840986e-01 8.41141403e-01 3.46058346e-02
1.05854976e+00 -2.88791895e-01 2.85173059e-02 3.96876663e-01
9.70425308e-02 -1.38612068e+00 -1.14833433e-02 4.25260425e-01
-6.11167431e-01 -8.33493054e-01 -7.86096334e-01 -9.96098459e-01
7.44461060e-01 2.00188577e-01 4.15816337e-01 8.66605416e-02
4.60310839e-02 4.57857996e-02 -5.87190092e-01 -4.51698482e-01
-8.19513202e-01 -3.27994406e-01 3.49901468e-02 4.25003558e-01
-1.54024549e-02 -7.34935284e-01 -4.62334007e-01 5.31924009e-01
-5.33833683e-01 2.10205726e-02 1.15044761e+00 7.72774339e-01
1.49924010e-01 8.82419229e-01 7.01162457e-01 -5.17944813e-01
2.38180175e-01 -6.31253481e-01 -8.38407636e-01 1.39387816e-01
-5.87168574e-01 -1.94665883e-02 7.18050599e-01 -3.77819508e-01
-7.69442558e-01 -2.03541741e-02 -1.81499317e-01 -1.21233381e-01
-2.15425536e-01 3.87104958e-01 -4.87092614e-01 -9.99251902e-01
3.96653652e-01 4.68244940e-01 -1.13056954e-02 -3.83261740e-01
8.48223791e-02 1.56653666e+00 7.09116578e-01 -3.59817557e-02
1.19010103e+00 1.48900300e-01 6.17659628e-01 -1.18766654e+00
-3.78407478e-01 -3.68175626e-01 -5.18583179e-01 -1.26840755e-01
5.25355339e-01 -1.22419155e+00 -7.02095807e-01 5.41525364e-01
-1.00692582e+00 7.83365965e-02 6.03723705e-01 9.01959836e-01
-3.91197532e-01 7.99873769e-01 -1.31703824e-01 -1.08212364e+00
5.19728009e-03 -1.12316608e+00 5.17919421e-01 1.02119498e-01
2.34023884e-01 -1.00676000e+00 -6.17939711e-01 1.48198888e-01
3.90932858e-01 2.01434404e-01 9.99425173e-01 -3.85419160e-01
-9.13085461e-01 -2.70032823e-01 -2.09185049e-01 6.34120166e-01
2.78893560e-01 -6.55833840e-01 -9.13268268e-01 -4.49989945e-01
1.11449704e-01 3.15001160e-01 6.45870268e-01 4.19277787e-01
1.27833641e+00 -3.74525547e-01 -2.75683910e-01 1.03591549e+00
1.63357818e+00 8.87192369e-01 1.08082032e+00 3.51761878e-01
1.94218438e-02 -1.37954652e-01 4.77547914e-01 3.45499933e-01
1.80975214e-01 6.80236459e-01 4.62096781e-01 -2.03961879e-01
-1.60748199e-01 2.73794651e-01 2.94936895e-01 7.41823792e-01
2.57385552e-01 -6.73288226e-01 -3.15120399e-01 -9.57515240e-02
-1.52478349e+00 -9.89772320e-01 -3.38188142e-01 2.18492651e+00
7.31498957e-01 2.87254632e-01 -4.60339218e-01 8.75910163e-01
1.01297700e+00 -8.80347788e-02 -3.89960557e-02 -4.79594469e-01
-3.73890460e-01 6.07609153e-01 1.06464851e+00 6.76420510e-01
-1.53649247e+00 2.29277521e-01 6.34449863e+00 1.63707352e+00
-1.22785008e+00 -3.81007493e-02 4.15585518e-01 5.73210001e-01
3.25363904e-01 2.47847885e-01 -9.40715671e-01 6.61592782e-01
1.06926882e+00 2.26562202e-01 2.09824622e-01 2.60383427e-01
9.28121209e-02 -3.84312719e-01 -8.58514369e-01 1.28175998e+00
1.85268864e-01 -1.10296929e+00 -4.95783798e-02 1.65771171e-01
8.63091707e-01 -5.82461238e-01 3.35794896e-01 4.18281227e-01
-2.62395144e-01 -1.18218410e+00 5.52469671e-01 3.79342258e-01
9.34803545e-01 -7.85996974e-01 8.66661966e-01 5.26913464e-01
-1.06433618e+00 -4.37867463e-01 -1.91005081e-01 -8.58952552e-02
2.10408345e-01 3.92977208e-01 -8.15541208e-01 1.28923118e+00
-8.50405023e-02 4.00215656e-01 -4.96075898e-01 1.73616576e+00
-2.83895850e-01 9.85234618e-01 -4.37217534e-01 -1.08254202e-01
3.20340931e-01 3.27933058e-02 8.10597539e-01 1.41418016e+00
7.15872288e-01 9.71220881e-02 1.50924241e-02 4.22628187e-02
2.05578580e-02 -1.78714141e-01 3.04076035e-04 2.28748798e-01
6.32769346e-01 9.28826153e-01 -3.68196845e-01 -3.84684175e-01
-4.14304167e-01 8.74750674e-01 -5.55032611e-01 4.03472155e-01
-8.12433183e-01 -9.03219223e-01 1.97417289e-01 -3.18870425e-01
8.24252605e-01 -4.06058669e-01 -4.08595413e-01 -6.81521475e-01
-2.64603108e-01 -1.00707257e+00 -1.29227955e-02 -3.89095634e-01
-7.63853908e-01 2.34482273e-01 4.72241692e-04 -2.06111956e+00
-3.45068723e-02 -8.01262319e-01 -5.63455701e-01 1.03084421e+00
-2.26031256e+00 -7.12760985e-01 -1.09523714e-01 3.02380294e-01
2.58963078e-01 -7.45215893e-01 6.94370031e-01 6.45082891e-01
-2.23505706e-01 1.05307209e+00 8.05419981e-01 -5.27666211e-02
3.88092965e-01 -1.14065802e+00 2.52156723e-02 8.10359597e-01
-1.07488632e-01 3.12754333e-01 7.40734816e-01 -4.60714817e-01
-1.08951795e+00 -1.23658597e+00 6.25909507e-01 4.24099505e-01
1.57934561e-01 -2.36426339e-01 -3.36206764e-01 2.10886881e-01
6.75493419e-01 -3.98119867e-01 8.20013702e-01 -5.44657767e-01
1.71119407e-01 -1.21632118e-04 -1.07904303e+00 5.07073224e-01
4.88940299e-01 -1.66498899e-01 -4.65790421e-01 1.06307670e-01
9.24759731e-02 -5.98458827e-01 -6.59421086e-01 6.72626376e-01
4.09224123e-01 -6.89293146e-01 7.45210409e-01 1.26771659e-01
-1.62969589e-01 -7.25912333e-01 -6.56611323e-01 -1.14742458e+00
-4.60524000e-02 -1.06429195e+00 -3.30551386e-01 5.92016637e-01
5.60738385e-01 -7.32544422e-01 6.95059359e-01 -6.39077008e-01
-3.28557730e-01 -8.05250466e-01 -1.15622520e+00 -1.10131991e+00
-7.87989795e-02 -5.84009588e-01 5.62705457e-01 5.82354605e-01
-2.20870391e-01 7.59632811e-02 -7.09467113e-01 8.72528970e-01
1.13161385e+00 -1.23722307e-01 7.35217273e-01 -9.53678668e-01
-7.40007818e-01 -2.49484092e-01 -9.52389240e-01 -1.63619053e+00
2.01041177e-02 -1.08562422e+00 -1.54012784e-01 -1.07041335e+00
-4.21187907e-01 -5.01396239e-01 -2.11943656e-01 -4.52008098e-02
1.80068985e-02 3.69187653e-01 6.04148842e-02 3.43575850e-02
-4.09197539e-01 6.42891049e-01 1.15116930e+00 -8.16277340e-02
8.54042396e-02 5.29637098e-01 -6.48343027e-01 5.56942403e-01
9.14630234e-01 -2.63429254e-01 -2.55038649e-01 1.20267242e-01
-4.37306762e-01 1.84583187e-01 4.28959697e-01 -1.63531792e+00
3.78954113e-01 2.66239285e-01 7.98209965e-01 -9.51420188e-01
6.49936020e-01 -1.01905656e+00 2.29976401e-01 6.59276009e-01
1.11367531e-01 -5.42521954e-01 8.25155079e-02 7.66215086e-01
-3.46140742e-01 -4.94458884e-01 9.88815486e-01 4.56035584e-01
-9.24525261e-01 -1.06006086e-01 -8.95554006e-01 -6.40203536e-01
1.12636101e+00 -7.80677378e-01 -2.59852618e-01 -7.99879968e-01
-9.45630491e-01 4.81366247e-01 -1.58452302e-01 -9.60264429e-02
7.14554012e-01 -1.40201581e+00 -5.59222519e-01 2.80352533e-01
-3.07079196e-01 -5.32482088e-01 4.36371453e-02 8.20735574e-01
-4.71235812e-01 9.04754400e-01 5.50329387e-02 -7.34590113e-01
-1.64714384e+00 -9.48315188e-02 6.15657449e-01 -2.12597102e-02
-2.60837466e-01 4.71941769e-01 -3.34039509e-01 -2.18874589e-01
5.17960966e-01 3.83870825e-02 -3.44775498e-01 -5.00473797e-01
2.05927774e-01 2.35939905e-01 1.55504316e-01 -7.51484394e-01
-1.95880067e-02 5.90330660e-01 9.35303494e-02 -4.00968939e-01
7.41074383e-01 -2.33925536e-01 1.09351128e-01 -1.83298677e-01
1.41399562e+00 -2.52781808e-01 -8.56138051e-01 -3.23783636e-01
-2.37041414e-02 -6.14463329e-01 1.71506435e-01 -7.42303431e-01
-6.44791543e-01 8.59247983e-01 1.22774410e+00 -2.22013369e-01
1.43170166e+00 -8.52775633e-01 5.09214818e-01 5.73251486e-01
5.14259875e-01 -1.06480002e+00 -3.89230072e-01 5.46939969e-01
3.76902997e-01 -8.71761441e-01 3.25946510e-01 -6.57173932e-01
-2.01499745e-01 1.11400199e+00 1.31550387e-01 6.04811460e-02
1.11955929e+00 1.42246082e-01 4.12453525e-02 3.65583450e-01
-5.65353155e-01 -1.58607975e-01 4.65036660e-01 7.86848485e-01
1.14347674e-02 1.50330234e-02 -5.14365971e-01 3.35247159e-01
-3.78982782e-01 -3.01072836e-01 6.16625130e-01 1.33903265e+00
-8.17791700e-01 -1.53751600e+00 -8.46993148e-01 5.73265493e-01
-6.23577297e-01 -1.57039925e-01 1.68243423e-01 8.19084406e-01
4.64682937e-01 1.35731494e+00 -1.24117799e-01 -6.56903863e-01
-1.55535936e-01 -1.58768743e-01 6.81916058e-01 -3.38544190e-01
-9.99323353e-02 2.83877820e-01 2.67939121e-01 5.85454591e-02
-5.93946576e-01 -4.30244863e-01 -1.29047132e+00 -2.36878395e-02
-8.56499553e-01 4.96125400e-01 1.01940763e+00 1.21611738e+00
2.28823319e-01 7.36449003e-01 1.23263669e+00 -7.79092371e-01
-6.81909025e-01 -9.65224981e-01 -7.46993244e-01 -3.28422427e-01
7.73961663e-01 -5.63920140e-01 -4.68555957e-01 -5.43757007e-02] | [6.426156997680664, 1.459825873374939] |
272b5366-1ac1-43a6-8d3d-4b82b0d3e8ee | weakly-supervised-temporal-action-5 | 2203.02925 | null | https://arxiv.org/abs/2203.02925v5 | https://arxiv.org/pdf/2203.02925v5.pdf | Weakly Supervised Temporal Action Localization via Representative Snippet Knowledge Propagation | Weakly supervised temporal action localization aims to localize temporal boundaries of actions and simultaneously identify their categories with only video-level category labels. Many existing methods seek to generate pseudo labels for bridging the discrepancy between classification and localization, but usually only make use of limited contextual information for pseudo label generation. To alleviate this problem, we propose a representative snippet summarization and propagation framework. Our method seeks to mine the representative snippets in each video for propagating information between video snippets to generate better pseudo labels. For each video, its own representative snippets and the representative snippets from a memory bank are propagated to update the input features in an intra- and inter-video manner. The pseudo labels are generated from the temporal class activation maps of the updated features to rectify the predictions of the main branch. Our method obtains superior performance in comparison to the existing methods on two benchmarks, THUMOS14 and ActivityNet1.3, achieving gains as high as 1.2% in terms of average mAP on THUMOS14. | ['Hongsheng Li', 'Liang Wang', 'Linjiang Huang'] | 2022-03-06 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Huang_Weakly_Supervised_Temporal_Action_Localization_via_Representative_Snippet_Knowledge_Propagation_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Huang_Weakly_Supervised_Temporal_Action_Localization_via_Representative_Snippet_Knowledge_Propagation_CVPR_2022_paper.pdf | cvpr-2022-1 | ['weakly-supervised-temporal-action', 'action-localization'] | ['computer-vision', 'computer-vision'] | [ 6.28245294e-01 -7.28565082e-02 -8.82935584e-01 -4.32784498e-01
-9.02128339e-01 -4.31042880e-01 5.10423362e-01 2.31717259e-01
-4.62514013e-01 8.75054359e-01 4.70957041e-01 3.64968032e-01
1.59464836e-01 -2.82882899e-01 -7.04743564e-01 -8.03874016e-01
-2.46916100e-01 3.20583321e-02 7.44197667e-01 2.48959526e-01
4.40418720e-01 -9.51808095e-02 -1.43188739e+00 1.04843628e+00
5.85198641e-01 1.40303230e+00 1.53413191e-01 4.27907497e-01
-1.88738346e-01 1.47761881e+00 -5.59008837e-01 6.16989695e-02
1.49898097e-01 -6.68019950e-01 -9.02733743e-01 3.48713309e-01
4.40228850e-01 -4.82500881e-01 -5.04364252e-01 8.66900563e-01
1.02378644e-01 2.86870480e-01 6.30595148e-01 -1.43473792e+00
-7.15335533e-02 9.42476988e-01 -7.86610663e-01 5.71542025e-01
4.22665119e-01 2.66095847e-01 1.07877266e+00 -9.27952707e-01
7.80476153e-01 9.46769655e-01 5.23944557e-01 6.93654597e-01
-9.86177802e-01 -6.83908165e-01 6.96423888e-01 6.49986207e-01
-1.45101702e+00 -4.26832229e-01 5.42474747e-01 -4.12472814e-01
8.98940980e-01 -1.53641291e-02 5.64254344e-01 1.18139517e+00
-4.36370708e-02 1.19780242e+00 8.47163498e-01 -7.42050335e-02
3.10971528e-01 -1.78115010e-01 2.23833457e-01 7.48545408e-01
-3.50929052e-01 -3.68233353e-01 -1.16310835e+00 -2.80284453e-02
6.50098085e-01 1.58891559e-01 -1.97932199e-01 1.16562983e-02
-1.59145117e+00 5.09801626e-01 5.58928967e-01 3.70451570e-01
-6.28227472e-01 5.88074863e-01 6.40842676e-01 -1.32450551e-01
4.43981230e-01 2.55935460e-01 -4.93946224e-01 -3.78412992e-01
-1.01925135e+00 -2.86369976e-02 2.65173405e-01 1.02108026e+00
9.61856186e-01 -1.21885926e-01 -7.39351034e-01 7.16737449e-01
-3.87452208e-02 1.48541089e-02 5.13618529e-01 -1.33922613e+00
6.85558021e-01 7.87443340e-01 2.51929939e-01 -8.57939959e-01
-2.45163143e-01 -2.30850473e-01 -4.74510491e-01 -2.92397469e-01
3.11659545e-01 8.76062550e-03 -1.08993137e+00 1.61710453e+00
2.25556687e-01 8.90761077e-01 -1.05986558e-01 8.17070246e-01
6.87627971e-01 8.75160515e-01 4.38876629e-01 -3.69015455e-01
9.49301362e-01 -1.37941551e+00 -6.19633913e-01 -4.04583961e-01
9.33363616e-01 -3.56484652e-01 8.75635922e-01 2.26203263e-01
-9.29058194e-01 -6.42543077e-01 -7.69441009e-01 3.26781362e-01
1.36086205e-02 3.98817986e-01 3.19581896e-01 -1.55257642e-01
-7.97417641e-01 7.04639316e-01 -9.70503688e-01 -3.07844698e-01
8.77669156e-01 2.17309386e-01 -2.34855801e-01 6.57408014e-02
-9.95077133e-01 3.41384053e-01 7.23520875e-01 9.00655985e-03
-1.23513436e+00 -6.57503963e-01 -7.52132297e-01 -4.47703823e-02
7.07108438e-01 -1.75484374e-01 1.32341731e+00 -1.51296604e+00
-1.05459476e+00 5.00482202e-01 -4.43120211e-01 -7.66499221e-01
3.08712900e-01 -1.62653849e-01 -1.30844638e-01 4.27228302e-01
4.80693907e-01 1.29118717e+00 8.20537627e-01 -8.26302648e-01
-1.49749374e+00 -9.17939618e-02 -1.65527947e-02 4.79584605e-01
-5.11569560e-01 -1.25993028e-01 -7.71162570e-01 -7.80669808e-01
1.34818286e-01 -9.25715029e-01 -1.68351457e-01 -2.17518449e-01
-4.22619939e-01 -4.28953022e-01 9.03920949e-01 -5.36349177e-01
1.56696427e+00 -2.17568326e+00 2.01226011e-01 -1.47861555e-01
1.55626148e-01 1.86845064e-01 -1.12563021e-01 -3.42073590e-02
1.78856459e-02 -3.72754596e-02 -2.15818644e-01 -1.11256637e-01
-3.16585332e-01 4.99448366e-02 -2.40711272e-01 3.70191008e-01
1.78876981e-01 8.29245329e-01 -1.11634231e+00 -6.81887925e-01
4.57936227e-02 9.39601213e-02 -4.44566548e-01 1.12867830e-02
-4.27833438e-01 6.24992430e-01 -5.76056242e-01 6.49069130e-01
6.29042387e-02 -3.45830590e-01 3.74059007e-02 -2.81271368e-01
4.11088243e-02 4.07170802e-01 -1.02529311e+00 1.82745910e+00
-5.56993894e-02 6.69159591e-01 -4.42230880e-01 -1.12350988e+00
5.56591034e-01 1.84832647e-01 9.39173460e-01 -5.50550580e-01
-1.19922221e-01 3.43764313e-02 -3.57148737e-01 -4.13305551e-01
3.60082567e-01 2.49952987e-01 -2.74646699e-01 4.24781322e-01
7.18724951e-02 6.47782743e-01 5.73269486e-01 3.65907669e-01
1.49210095e+00 4.86485988e-01 9.87616107e-02 8.07792172e-02
7.02308238e-01 2.64170438e-01 9.01566923e-01 6.82822704e-01
-4.46693957e-01 3.94834459e-01 6.40158176e-01 -5.20270407e-01
-6.73202276e-01 -9.08627272e-01 2.75517851e-01 1.30876958e+00
3.49141836e-01 -7.38043487e-01 -8.33343565e-01 -1.37093151e+00
-2.80762166e-01 5.51364005e-01 -6.73587739e-01 -4.04834986e-01
-7.70856440e-01 -4.53643471e-01 4.89373147e-01 8.59598458e-01
8.36459637e-01 -1.31801915e+00 -7.48305261e-01 4.46584761e-01
-6.72733247e-01 -1.21762872e+00 -8.01232636e-01 2.34375685e-01
-9.86068606e-01 -1.04650867e+00 -7.07575858e-01 -8.10902476e-01
8.20325077e-01 2.39909381e-01 7.69229770e-01 -2.44936198e-01
9.48569214e-04 7.00481683e-02 -5.55418789e-01 2.93392658e-01
-3.01673681e-01 1.13264218e-01 4.20873649e-02 3.82018089e-01
4.90577877e-01 -3.38685095e-01 -6.97882414e-01 6.56551838e-01
-5.91941059e-01 3.04207921e-01 5.39073467e-01 7.41498172e-01
9.18366551e-01 1.11035936e-01 7.21086979e-01 -7.93385446e-01
2.61352118e-02 -5.90383112e-01 -2.65989453e-01 1.18044198e-01
-4.18358535e-01 -4.38486114e-02 5.36884189e-01 -6.83281064e-01
-1.06320000e+00 5.41088223e-01 2.68716604e-01 -4.38928246e-01
-2.89465398e-01 3.02600414e-01 1.39330685e-01 3.58338386e-01
6.35286748e-01 4.98265356e-01 -5.20332158e-01 -2.58050561e-01
3.61897141e-01 4.99712110e-01 6.36549652e-01 -1.48487538e-01
2.41195187e-01 4.88620877e-01 -3.83138418e-01 -5.26610851e-01
-1.08114111e+00 -6.89126551e-01 -7.94396400e-01 -6.44184470e-01
1.05269325e+00 -9.48335230e-01 -3.49483192e-01 4.63967890e-01
-9.69208539e-01 -5.02521813e-01 -4.24051642e-01 4.14332479e-01
-6.07675135e-01 2.38924772e-01 -6.97645843e-01 -4.62421209e-01
-7.43572041e-02 -1.09892261e+00 1.10447776e+00 2.70410776e-01
-3.58087271e-01 -4.93127614e-01 -3.02243024e-01 3.79962176e-01
1.26602352e-02 3.33581716e-02 7.01662183e-01 -6.09155595e-01
-6.82070196e-01 -1.12939648e-01 -1.96793020e-01 4.26887393e-01
3.05023819e-01 -2.71010101e-01 -7.37524152e-01 -1.04353577e-01
-4.67203379e-01 -3.92046988e-01 1.22819340e+00 5.33793449e-01
1.31783319e+00 -4.21059310e-01 -6.38578892e-01 2.75196970e-01
9.95762408e-01 3.94944102e-01 3.77240539e-01 1.23478696e-01
8.35333228e-01 4.13133383e-01 1.04985392e+00 6.37589276e-01
2.06605047e-01 8.48215103e-01 3.62117529e-01 2.89220661e-01
-3.21673870e-01 -3.91146690e-01 8.27947378e-01 5.78613937e-01
6.35460243e-02 -1.00978628e-01 -6.99896991e-01 5.63958168e-01
-2.10779262e+00 -1.12149823e+00 1.65193025e-02 2.10804176e+00
8.83353710e-01 5.09280086e-01 2.99684674e-01 -1.82454772e-02
1.00505972e+00 4.18596953e-01 -7.45551825e-01 2.27234200e-01
7.01420307e-02 -2.91738003e-01 4.87692803e-01 1.34439886e-01
-1.44982636e+00 1.24675953e+00 5.32412052e+00 9.38682735e-01
-8.70286286e-01 3.10754210e-01 9.44941223e-01 -3.82533491e-01
3.32354963e-01 1.19483769e-01 -1.04611266e+00 6.91769481e-01
8.56097162e-01 9.76202190e-02 1.11437000e-01 8.74071896e-01
4.46603030e-01 -6.12713218e-01 -1.35583091e+00 9.75784898e-01
1.51295692e-01 -1.50407898e+00 6.80033863e-02 -2.74251372e-01
1.09291220e+00 -9.76792630e-03 -1.39218107e-01 4.13345039e-01
-3.35762911e-02 -5.06588519e-01 8.97154748e-01 5.35129130e-01
6.20592892e-01 -6.75065875e-01 5.97354949e-01 3.14668655e-01
-1.51968265e+00 -3.74179214e-01 -8.99713337e-02 -2.81641670e-02
1.93174973e-01 2.50236303e-01 -7.90306330e-01 1.14911556e-01
7.08415687e-01 1.28770959e+00 -6.31884992e-01 1.07804692e+00
-3.28741074e-01 9.51661944e-01 -1.68953419e-01 1.31757662e-01
6.21964157e-01 1.80665433e-01 3.30568641e-01 1.24810648e+00
1.40689611e-01 -2.38408390e-02 5.34904182e-01 4.48471725e-01
-1.97675765e-01 1.14552476e-01 -2.61656076e-01 -5.23792990e-02
5.17873883e-01 1.07744944e+00 -1.09478748e+00 -7.73741543e-01
-3.84419143e-01 1.23740077e+00 9.40636918e-02 3.02299798e-01
-1.19323194e+00 -1.14493020e-01 3.43853116e-01 2.13649407e-01
3.24319482e-01 7.09117576e-02 -1.39400214e-01 -1.02589476e+00
4.78798263e-02 -6.22274697e-01 7.97057390e-01 -6.85661554e-01
-7.60779023e-01 4.70781744e-01 1.88164413e-01 -1.48895824e+00
-2.86936581e-01 6.16811365e-02 -5.24493217e-01 2.86773592e-01
-1.06972313e+00 -7.82339931e-01 -4.99290109e-01 4.31272537e-01
1.28297985e+00 -1.15283430e-01 4.14853781e-01 3.80332559e-01
-7.23765612e-01 4.33284819e-01 -2.84523040e-01 1.41149402e-01
7.08383620e-01 -1.02871990e+00 1.86430782e-01 9.74910438e-01
3.70778799e-01 -5.07507958e-02 3.59455258e-01 -8.49940300e-01
-7.25759149e-01 -1.53746605e+00 8.24242949e-01 -2.56731331e-01
5.38665116e-01 -1.79153323e-01 -7.51432538e-01 7.00757504e-01
-9.51342732e-02 1.90965772e-01 4.06659752e-01 -4.87304658e-01
-2.37704366e-01 -2.68294454e-01 -7.86566198e-01 6.05433702e-01
1.39125156e+00 -2.02271357e-01 -4.31656241e-01 6.19339466e-01
5.69900036e-01 -1.03017002e-01 -5.20553529e-01 3.46455783e-01
3.18665415e-01 -7.60430634e-01 5.91938317e-01 -4.88853484e-01
6.03811920e-01 -5.62452793e-01 -1.02168851e-01 -9.04403448e-01
-4.06724900e-01 -6.80789649e-01 -1.78518862e-01 1.31270564e+00
6.21051371e-01 3.41194384e-02 1.11404586e+00 2.75021225e-01
-1.71874493e-01 -7.96739101e-01 -9.13996041e-01 -5.64373195e-01
-7.65743256e-01 -5.37585020e-01 7.03713000e-02 6.19907737e-01
1.36371583e-01 3.73004496e-01 -6.04185760e-01 -8.55384246e-02
5.13154209e-01 -9.68124792e-02 4.94730949e-01 -7.23323584e-01
-9.84469578e-02 -3.91793340e-01 -4.83280659e-01 -1.28927457e+00
3.62599909e-01 -9.85466421e-01 4.57279146e-01 -1.55027318e+00
4.47685778e-01 -4.90974456e-01 -6.85118496e-01 9.40813243e-01
-7.12596998e-02 6.79789901e-01 1.20187752e-01 3.70435536e-01
-1.35880268e+00 3.40659380e-01 9.66896534e-01 -3.17591578e-01
-4.83394086e-01 6.73277155e-02 -4.24387246e-01 9.99442577e-01
7.06399322e-01 -7.59036720e-01 -5.08093655e-01 -2.59258032e-01
-1.95545465e-01 1.14619412e-01 2.21651256e-01 -1.32962310e+00
3.40923071e-01 -3.26216936e-01 4.09985065e-01 -8.52861464e-01
2.84177333e-01 -5.52185535e-01 -9.21591148e-02 4.52350050e-01
-8.93844664e-01 -2.82446653e-01 -1.18259221e-01 8.69429886e-01
-2.95653224e-01 -1.31968275e-01 8.25213969e-01 -8.80413577e-02
-1.43407655e+00 4.44444060e-01 -5.73111176e-01 3.33728381e-02
1.42582607e+00 -3.27373087e-01 -2.31046781e-01 -3.76626313e-01
-9.39050734e-01 4.13015574e-01 1.86117962e-01 3.92144501e-01
6.64706349e-01 -1.48150003e+00 -5.90369225e-01 -8.25606380e-03
1.74369246e-01 -2.32334837e-01 4.95792836e-01 1.01922619e+00
-1.49276927e-01 4.55199718e-01 -2.65539527e-01 -7.18947709e-01
-1.26255476e+00 4.32130218e-01 2.02938318e-01 -1.95272401e-01
-6.66540802e-01 1.12133396e+00 5.36493897e-01 5.11139512e-01
5.82220733e-01 -2.93463081e-01 -3.38318259e-01 3.38647932e-01
6.86063766e-01 5.20503104e-01 -2.65949875e-01 -9.46760297e-01
-5.72103620e-01 3.58292639e-01 -1.75383478e-01 -1.86335687e-02
1.16996419e+00 -1.65222079e-01 1.69037655e-01 4.84556615e-01
1.19230056e+00 -5.16215444e-01 -1.88431287e+00 -3.89235437e-01
2.17097536e-01 -3.72210652e-01 -5.10521159e-02 -7.49618471e-01
-1.25730455e+00 5.07188857e-01 5.33452392e-01 -3.24762285e-01
1.31659210e+00 3.89762700e-01 9.41678405e-01 2.09882349e-01
3.63133579e-01 -1.36243379e+00 6.56853139e-01 2.87192076e-01
5.82256377e-01 -1.00568843e+00 -5.62616512e-02 -3.82422328e-01
-9.92403567e-01 8.47117901e-01 9.93057251e-01 -3.28712314e-02
3.15652817e-01 6.99738413e-02 -1.67340845e-01 6.83859587e-02
-9.58689153e-01 -1.68532953e-01 2.44133949e-01 3.90990704e-01
1.40144438e-01 -1.07523970e-01 -4.72595155e-01 5.97718179e-01
7.13330746e-01 1.55266896e-01 4.59873736e-01 1.08698046e+00
-7.49281347e-01 -9.25385118e-01 -5.88851310e-02 7.63005435e-01
-4.21522170e-01 5.90525568e-02 -3.22772443e-01 2.76184231e-01
3.41143161e-01 8.01030695e-01 1.62631631e-01 -6.73379481e-01
1.42389044e-01 2.23073766e-01 1.71952844e-01 -8.32150996e-01
-3.57421726e-01 4.51230079e-01 1.51367694e-01 -9.50334430e-01
-7.23262131e-01 -8.50333273e-01 -1.64310718e+00 2.25010663e-01
-1.35057688e-01 1.15921289e-01 1.36295795e-01 9.70951259e-01
3.65031749e-01 6.39168024e-01 5.80297947e-01 -9.32277441e-01
-2.01313809e-01 -1.02928221e+00 -4.93483663e-01 3.72110814e-01
3.76488380e-02 -7.84334779e-01 -2.21804157e-01 6.35918796e-01] | [8.493468284606934, 0.6134655475616455] |
47d5a06a-6b5c-4bdf-9f68-bcacb9b9d479 | self-point-flow-self-supervised-scene-flow | 2105.08248 | null | https://arxiv.org/abs/2105.08248v1 | https://arxiv.org/pdf/2105.08248v1.pdf | Self-Point-Flow: Self-Supervised Scene Flow Estimation from Point Clouds with Optimal Transport and Random Walk | Due to the scarcity of annotated scene flow data, self-supervised scene flow learning in point clouds has attracted increasing attention. In the self-supervised manner, establishing correspondences between two point clouds to approximate scene flow is an effective approach. Previous methods often obtain correspondences by applying point-wise matching that only takes the distance on 3D point coordinates into account, introducing two critical issues: (1) it overlooks other discriminative measures, such as color and surface normal, which often bring fruitful clues for accurate matching; and (2) it often generates sub-par performance, as the matching is operated in an unconstrained situation, where multiple points can be ended up with the same corresponding point. To address the issues, we formulate this matching task as an optimal transport problem. The output optimal assignment matrix can be utilized to guide the generation of pseudo ground truth. In this optimal transport, we design the transport cost by considering multiple descriptors and encourage one-to-one matching by mass equality constraints. Also, constructing a graph on the points, a random walk module is introduced to encourage the local consistency of the pseudo labels. Comprehensive experiments on FlyingThings3D and KITTI show that our method achieves state-of-the-art performance among self-supervised learning methods. Our self-supervised method even performs on par with some supervised learning approaches, although we do not need any ground truth flow for training. | ['Lihua Xie', 'Guosheng Lin', 'Ruibo Li'] | 2021-05-18 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Li_Self-Point-Flow_Self-Supervised_Scene_Flow_Estimation_From_Point_Clouds_With_Optimal_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Li_Self-Point-Flow_Self-Supervised_Scene_Flow_Estimation_From_Point_Clouds_With_Optimal_CVPR_2021_paper.pdf | cvpr-2021-1 | ['scene-flow-estimation'] | ['computer-vision'] | [-1.86499149e-01 -2.88966328e-01 -4.72692847e-01 -2.76019514e-01
-4.87358540e-01 -5.13099611e-01 5.30580103e-01 1.80533469e-01
-2.66831458e-01 5.96753061e-01 -9.40370336e-02 -8.69792923e-02
-1.80818602e-01 -9.29863572e-01 -6.14006042e-01 -6.08989596e-01
2.64493264e-02 5.77632606e-01 5.83511770e-01 -1.55588880e-01
4.64974225e-01 6.46490157e-01 -1.51891971e+00 -1.47703573e-01
1.16718006e+00 9.61176217e-01 2.54315585e-01 2.46475622e-01
-6.81194305e-01 6.76651061e-01 -3.36466610e-01 -3.26132387e-01
5.57890177e-01 -5.35696507e-01 -9.76938248e-01 3.14712048e-01
7.05700576e-01 -2.25498244e-01 -4.25128132e-01 1.06818926e+00
1.32685572e-01 2.04650164e-01 7.18581259e-01 -1.57221925e+00
-3.54931921e-01 1.57075934e-03 -7.91673005e-01 9.12662968e-02
4.47161824e-01 3.24057549e-01 1.12957489e+00 -8.26521277e-01
5.26129842e-01 1.02029777e+00 4.94248688e-01 2.16603592e-01
-1.12832606e+00 -5.90813577e-01 1.77862912e-01 2.19827458e-01
-1.51402855e+00 -1.86436176e-01 1.11268377e+00 -5.35437226e-01
4.10987675e-01 4.14582379e-02 9.24163043e-01 4.35084492e-01
-1.40933335e-01 7.85558343e-01 7.83489645e-01 -2.38808826e-01
2.74962425e-01 9.96028930e-02 -9.99026075e-02 1.00125420e+00
1.79359049e-01 4.59920540e-02 -5.50311983e-01 1.31811462e-02
8.39895546e-01 2.06650615e-01 -3.12854290e-01 -8.67318034e-01
-1.41859913e+00 8.02720845e-01 9.77542102e-01 2.49150753e-01
-1.03491195e-01 -3.42186019e-02 2.03591704e-01 7.06163868e-02
4.48257923e-01 4.12546009e-01 -1.73933849e-01 -7.24562928e-02
-8.90602529e-01 1.39650777e-01 6.49676323e-01 1.25016510e+00
1.52113235e+00 -3.34175006e-02 1.38954058e-01 5.51364124e-01
2.72499591e-01 4.61317718e-01 1.89634815e-01 -1.03474891e+00
7.12337315e-01 1.08482492e+00 1.15103073e-01 -1.44783127e+00
-1.81837961e-01 -2.27156073e-01 -9.59801733e-01 2.20851526e-01
6.50480509e-01 1.17046855e-01 -5.65501928e-01 1.34424543e+00
4.67899859e-01 5.34620702e-01 -1.71153069e-01 1.11278033e+00
7.79988706e-01 7.24606872e-01 -1.66102499e-01 -2.33609617e-01
8.11339021e-01 -1.15624940e+00 -4.57094878e-01 -2.19606429e-01
7.64056683e-01 -7.96582937e-01 1.08898473e+00 -7.98245519e-02
-8.31167519e-01 -6.68655992e-01 -9.45574641e-01 6.18274324e-02
-2.63865411e-01 4.04727645e-02 7.16946661e-01 2.85069317e-01
-7.53772616e-01 7.34833241e-01 -6.83619499e-01 -2.57535458e-01
4.88604397e-01 1.53470993e-01 -3.90949279e-01 -7.56183788e-02
-8.26202333e-01 7.70951211e-01 2.13060290e-01 1.88510120e-01
-4.34566230e-01 -8.00561607e-01 -8.89388204e-01 -2.15272263e-01
3.28731388e-01 -7.70430505e-01 6.30467534e-01 -7.25599885e-01
-1.36246181e+00 9.17215466e-01 -3.26152653e-01 -1.14073856e-02
7.49610543e-01 -1.15220994e-02 -1.38321757e-01 3.10219586e-01
4.68500853e-01 7.84268320e-01 6.46842360e-01 -1.41259348e+00
-8.39670122e-01 -1.60524607e-01 1.04594156e-01 3.56391013e-01
-1.87280804e-01 -4.32139695e-01 -5.15469253e-01 -3.50175738e-01
5.95601797e-01 -8.65079641e-01 -2.79097646e-01 4.68758136e-01
-4.86213863e-01 -4.33278739e-01 9.69443738e-01 -2.82421689e-02
1.06119394e+00 -2.20597816e+00 9.41241905e-02 2.36896008e-01
3.30251932e-01 1.74837023e-01 6.71233013e-02 2.44402081e-01
1.60554096e-01 -1.10539012e-02 -5.27543366e-01 -4.22214806e-01
-3.44763637e-01 3.27402085e-01 -2.51762956e-01 6.68267548e-01
3.14643532e-01 7.79261708e-01 -1.41047084e+00 -9.89248693e-01
6.30984902e-01 9.49378237e-02 -5.29444933e-01 3.96555543e-01
-7.43486211e-02 6.25478089e-01 -6.93788230e-01 5.72497487e-01
7.76323736e-01 -4.25894678e-01 -4.45724100e-01 -4.41152066e-01
-3.31901997e-01 9.32904109e-02 -1.47518420e+00 2.10729289e+00
-3.26641321e-01 5.39979219e-01 -4.36248630e-01 -9.02671874e-01
1.10668314e+00 -1.53395355e-01 8.42889845e-01 -3.44509482e-01
-3.66665088e-02 2.03648910e-01 -2.02932298e-01 -4.48261946e-01
3.88883740e-01 -4.76897173e-02 2.01014936e-01 2.84575492e-01
-6.34560809e-02 -5.11396408e-01 1.89140543e-01 2.72479683e-01
7.95918882e-01 3.58630955e-01 1.06365316e-01 -3.24218541e-01
7.22762764e-01 5.52991271e-01 6.99053049e-01 3.81785393e-01
-2.04946503e-01 8.50578129e-01 3.10102254e-01 -6.25348866e-01
-9.11609590e-01 -8.80354941e-01 -1.61582306e-01 3.77295762e-01
9.66479123e-01 -4.42337513e-01 -4.52842087e-01 -7.76690960e-01
4.44051586e-02 2.29350969e-01 -3.42220962e-01 -2.32299343e-02
-6.16517842e-01 -4.41208333e-01 1.71860948e-01 3.97332311e-01
7.92965949e-01 -7.80714273e-01 -5.94720900e-01 1.08167455e-01
-3.08329672e-01 -1.22694612e+00 -6.56866431e-01 -1.15003102e-01
-1.01878166e+00 -1.30964077e+00 -6.82910919e-01 -7.90506840e-01
1.11295474e+00 8.35030317e-01 9.75204527e-01 4.23680753e-01
-1.54946342e-01 3.97194587e-02 -3.19890290e-01 1.13255240e-01
-6.22118264e-02 2.12247401e-01 1.10686040e-02 1.57982588e-01
2.34775752e-01 -5.56132495e-01 -7.42487371e-01 7.01641023e-01
-6.62290871e-01 1.75661817e-01 2.52096713e-01 7.00993419e-01
6.69057548e-01 9.34059545e-02 1.74888939e-01 -6.40843213e-01
5.83381057e-02 -2.15786844e-01 -6.36531293e-01 1.81317866e-01
-4.51360524e-01 8.41173306e-02 6.84176385e-01 -2.86938727e-01
-7.72903621e-01 3.77894908e-01 1.61577880e-01 -7.46959567e-01
-8.40801597e-02 1.67985201e-01 -1.51032284e-01 -4.99717474e-01
6.26540661e-01 7.42620006e-02 1.12381414e-01 -1.61180705e-01
5.43196619e-01 5.21485567e-01 4.02390569e-01 -5.50444484e-01
1.27651918e+00 7.25903630e-01 3.88967156e-01 -7.51697361e-01
-9.61241603e-01 -8.15817535e-01 -1.14777052e+00 -4.92831200e-01
8.51773739e-01 -7.40061343e-01 -7.29842246e-01 3.28446656e-01
-1.21572912e+00 -1.13753699e-01 -3.05895269e-01 6.16254866e-01
-5.26135087e-01 5.11983514e-01 -3.50919515e-01 -6.46957040e-01
7.71060139e-02 -1.25484180e+00 1.03271747e+00 2.95427859e-01
-1.00323059e-01 -1.16511929e+00 1.84272993e-02 2.81330526e-01
1.62359431e-01 3.00509363e-01 6.90404713e-01 -1.91898778e-01
-1.05888820e+00 -5.17825969e-02 -4.26394254e-01 1.93232968e-01
4.81467903e-01 1.78371400e-01 -7.83652544e-01 -9.83688608e-02
-2.73907572e-01 -1.23026893e-01 5.19145191e-01 1.09261036e-01
1.04560542e+00 6.79973699e-03 -3.71449798e-01 8.76552999e-01
1.41569555e+00 2.40186998e-03 4.34600770e-01 3.50471795e-01
1.18401515e+00 8.20380449e-01 6.86914206e-01 2.20405176e-01
5.68374991e-01 5.92775762e-01 5.63610673e-01 -2.65975565e-01
-1.62778586e-01 -6.96536183e-01 -1.15642250e-01 9.35104430e-01
-7.29737729e-02 -1.73325688e-02 -9.54420865e-01 4.41140771e-01
-1.99189663e+00 -7.95611024e-01 -4.95729268e-01 2.37217999e+00
5.68125248e-01 1.45440385e-01 1.43910348e-01 2.30993629e-01
6.41703486e-01 3.57692540e-01 -4.65124696e-01 3.02664250e-01
-1.05360769e-01 -2.50406027e-01 5.04258096e-01 6.08312190e-01
-9.73618031e-01 1.08591497e+00 5.70608187e+00 6.30220115e-01
-1.25284016e+00 -1.21980861e-01 4.45378661e-01 1.98552221e-01
-3.53635579e-01 4.68237489e-01 -7.74865746e-01 5.65585554e-01
-3.47921513e-02 -1.61329225e-01 2.65468001e-01 7.58196950e-01
1.68724149e-01 -2.10692763e-01 -1.11123252e+00 1.27563727e+00
5.72890267e-02 -1.42307091e+00 9.84655842e-02 1.46003263e-02
7.13859797e-01 -1.17630124e-01 -3.94648671e-01 -1.50852308e-01
1.00376055e-01 -5.62130094e-01 5.26792824e-01 4.93441433e-01
6.61632597e-01 -6.02808058e-01 4.91012096e-01 5.26271820e-01
-1.48957598e+00 3.17774385e-01 -4.84336555e-01 -4.43964526e-02
3.65338296e-01 6.23421609e-01 -4.21846688e-01 7.90535927e-01
6.22485101e-01 1.16500139e+00 -5.84543943e-01 1.39680433e+00
-2.53692508e-01 2.34278068e-01 -3.98276985e-01 2.73042414e-02
3.11952502e-01 -7.71736324e-01 6.20751023e-01 8.51939142e-01
2.42302462e-01 -2.33222488e-02 5.06518662e-01 9.47536469e-01
1.26827598e-01 1.93270728e-01 -7.80734658e-01 5.46772778e-01
4.33205903e-01 1.21952868e+00 -9.82071280e-01 -2.53049433e-01
-4.09486175e-01 8.91742289e-01 3.61068875e-01 3.42453897e-01
-5.84354520e-01 -3.34417820e-01 5.24827063e-01 2.60706902e-01
-2.26488352e-01 -3.70043963e-01 -3.34781855e-01 -1.29488909e+00
1.72236204e-01 -2.12705314e-01 2.19033033e-01 -8.51955891e-01
-1.34775317e+00 5.57817221e-01 1.01562269e-01 -1.95216370e+00
-5.23570441e-02 -3.28713834e-01 -8.08666408e-01 7.12159395e-01
-1.72087085e+00 -8.83257568e-01 -8.02280962e-01 6.71884835e-01
3.69293720e-01 -4.48839180e-02 3.45943600e-01 4.66453046e-01
-6.18293643e-01 2.71142960e-01 -3.33085418e-01 1.94206581e-01
7.07534611e-01 -1.26933515e+00 1.94166049e-01 9.66005385e-01
3.15058559e-01 4.35668290e-01 4.67855215e-01 -5.36458433e-01
-1.31344724e+00 -9.94367838e-01 8.25597227e-01 -4.31003362e-01
6.46625578e-01 -2.16973230e-01 -1.13405299e+00 3.11985850e-01
-1.90529197e-01 5.21507919e-01 2.52958775e-01 -2.16867492e-01
-8.40240344e-02 -4.00361955e-01 -9.50265467e-01 5.91893852e-01
1.52514088e+00 -4.52648818e-01 -3.97738665e-01 5.38754165e-01
7.15210617e-01 -6.31068766e-01 -6.47417605e-01 4.02749240e-01
4.36551832e-02 -1.00851035e+00 9.16994333e-01 -3.82516325e-01
5.04514456e-01 -8.68904531e-01 7.72939697e-02 -1.36512685e+00
-2.16143429e-01 -6.64670050e-01 5.90698831e-02 1.35401547e+00
1.78609639e-01 -5.24998367e-01 1.11153996e+00 5.62344193e-01
-1.73814714e-01 -6.78647935e-01 -8.65313709e-01 -8.47413182e-01
-2.10048780e-01 -2.77774960e-01 7.65767276e-01 1.19006002e+00
-1.44322753e-01 3.48034710e-01 -2.93558657e-01 1.57234505e-01
7.98138916e-01 4.28280592e-01 1.11591554e+00 -1.31779134e+00
1.50446087e-01 -5.35860479e-01 -6.65987790e-01 -1.59485483e+00
3.45107824e-01 -9.08668518e-01 7.72304684e-02 -1.54015589e+00
-1.80524603e-01 -1.00785077e+00 1.37509122e-01 3.07737857e-01
-2.17390209e-01 9.85845029e-02 1.87195525e-01 6.40092373e-01
-5.96215606e-01 7.15987504e-01 1.59977353e+00 -1.04957566e-01
-3.91850054e-01 1.69048980e-01 -4.85526204e-01 7.28508115e-01
5.90313554e-01 -4.33345526e-01 -5.10158360e-01 -5.23451626e-01
6.74254149e-02 1.09733947e-01 4.33508635e-01 -1.08090985e+00
6.70771480e-01 -4.86657679e-01 1.06362015e-01 -6.52166486e-01
1.62327811e-01 -1.05627787e+00 -6.65027276e-02 3.13446760e-01
-5.14357835e-02 2.77449042e-01 -1.81163862e-01 5.71700513e-01
-4.99623567e-01 -3.02891910e-01 6.91286325e-01 -1.46610826e-01
-8.09379995e-01 7.86099911e-01 2.47256234e-01 1.60937622e-01
1.05114782e+00 -5.17720640e-01 -2.30096564e-01 -2.13553503e-01
-1.46895930e-01 5.44956744e-01 6.26789272e-01 3.32747728e-01
6.25716567e-01 -1.48407233e+00 -5.16209245e-01 2.64041811e-01
4.54642802e-01 7.87255883e-01 -6.95764124e-02 6.96671426e-01
-8.13220203e-01 1.55148372e-01 -3.03012803e-02 -1.11693072e+00
-7.60704398e-01 3.20018977e-01 3.73890340e-01 1.47762135e-01
-8.28152478e-01 6.07976973e-01 1.69490725e-01 -4.16042447e-01
1.86156690e-01 -2.50430495e-01 -2.92514507e-02 3.72959822e-02
2.10938156e-01 4.73052114e-01 -5.24565130e-02 -7.64613152e-01
-4.05819267e-01 1.09153485e+00 2.98376173e-01 2.05900013e-01
1.03452444e+00 -1.53646052e-01 -1.05236702e-01 4.87375498e-01
1.34492707e+00 -5.33039682e-02 -1.48774338e+00 -4.13051993e-01
-1.37272105e-01 -1.01156259e+00 9.07441303e-02 -4.75909933e-03
-1.36860132e+00 9.87357676e-01 2.66945899e-01 2.49280885e-01
8.52461278e-01 -8.25198293e-02 7.74364233e-01 2.41066381e-01
6.14789307e-01 -6.93231285e-01 1.77705869e-01 3.23069721e-01
5.08516610e-01 -1.50096154e+00 1.93023950e-01 -7.49382973e-01
-5.60150027e-01 1.12590337e+00 9.59780455e-01 -3.02369922e-01
6.86947465e-01 -3.87530737e-02 1.12423643e-01 -1.43779844e-01
-2.54155427e-01 -3.70917618e-01 4.04650569e-01 5.35227418e-01
8.70899335e-02 -2.15498120e-01 5.65287052e-03 -3.63944113e-01
-3.51264924e-01 -9.67965946e-02 2.22772151e-01 7.60013998e-01
-3.84387106e-01 -1.07650852e+00 -1.81123525e-01 3.53724629e-01
1.84726045e-01 2.21349373e-01 -1.42436013e-01 7.87110686e-01
2.80649774e-02 8.40433478e-01 3.79950851e-01 -3.94437462e-01
5.90092123e-01 -4.36086088e-01 2.51657516e-01 -4.77211148e-01
-3.64500761e-01 -1.36853069e-01 -3.44831377e-01 -6.21642768e-01
-8.18254471e-01 -5.39405942e-01 -1.32114911e+00 -4.34421629e-01
-4.39635634e-01 3.11287493e-01 4.04774457e-01 1.02493167e+00
2.50092536e-01 1.67729363e-01 1.09837937e+00 -8.53202760e-01
-1.56808615e-01 -4.65299875e-01 -4.12149072e-01 7.05608010e-01
4.27143902e-01 -8.41501534e-01 -7.29632795e-01 3.32276896e-02] | [8.487618446350098, -2.1079494953155518] |
8bf1b646-7f1f-4942-973c-3a2a550a1544 | reflect-summarizing-robot-experiences-for | 2306.15724 | null | https://arxiv.org/abs/2306.15724v1 | https://arxiv.org/pdf/2306.15724v1.pdf | REFLECT: Summarizing Robot Experiences for Failure Explanation and Correction | The ability to detect and analyze failed executions automatically is crucial for an explainable and robust robotic system. Recently, Large Language Models (LLMs) have demonstrated strong common sense reasoning skills on textual inputs. To leverage the power of LLM for robot failure explanation, we propose a framework REFLECT, which converts multi-sensory data into a hierarchical summary of robot past experiences and queries LLM with a progressive failure explanation algorithm. Conditioned on the explanation, a failure correction planner generates an executable plan for the robot to correct the failure and complete the task. To systematically evaluate the framework, we create the RoboFail dataset and show that our LLM-based framework is able to generate informative failure explanations that assist successful correction planning. Project website: https://roboreflect.github.io/ | ['Shuran Song', 'Arpit Bahety', 'Zeyi Liu'] | 2023-06-27 | null | null | null | null | ['common-sense-reasoning'] | ['reasoning'] | [ 1.95509955e-01 7.92062044e-01 -3.46026331e-01 -4.21078533e-01
-6.91997945e-01 -5.06513834e-01 7.08783567e-01 2.56971061e-01
2.15148628e-01 5.67445040e-01 5.01564205e-01 -5.89414179e-01
-2.03597292e-01 -4.62592572e-01 -9.89714622e-01 3.59087676e-01
-7.17908442e-02 6.43280208e-01 7.91686475e-02 -1.28040552e-01
4.90703464e-01 2.24445343e-01 -1.79820704e+00 3.72840077e-01
5.17638206e-01 4.40563262e-01 6.71024084e-01 9.00353014e-01
1.91312551e-01 1.54392099e+00 -2.36905813e-01 2.04773188e-01
-2.89089105e-04 -1.84229627e-01 -1.40149260e+00 -3.31512801e-02
-3.99878383e-01 -4.01402205e-01 -4.12690669e-01 7.82353699e-01
-3.53427976e-01 2.19645590e-01 5.88965535e-01 -2.07622266e+00
-2.73081452e-01 1.15631020e+00 1.89416796e-01 -2.08047792e-01
1.00300419e+00 6.00256979e-01 1.01680911e+00 -8.22646081e-01
8.86165023e-01 1.43761659e+00 3.43674868e-01 8.07229280e-01
-1.06001472e+00 -1.27698958e-01 1.37947768e-01 4.20536518e-01
-1.09471476e+00 -4.46640104e-01 3.58536333e-01 -4.49354291e-01
1.82670486e+00 1.69118330e-01 4.55684602e-01 1.12520623e+00
8.79400134e-01 7.96149671e-01 5.93001485e-01 -3.58427048e-01
2.89951175e-01 -3.24361145e-01 2.99973100e-01 9.11441922e-01
2.70477653e-01 3.23679745e-01 -8.00383389e-01 -4.28731479e-02
3.81082177e-01 3.10766727e-01 1.19813859e-01 -3.62979829e-01
-1.30585599e+00 4.10828680e-01 2.41420865e-01 1.01599358e-01
-4.60543960e-01 9.32894826e-01 3.34174067e-01 2.55730063e-01
-2.69383371e-01 1.04779911e+00 -6.68833613e-01 -4.39325243e-01
-2.73936629e-01 5.62976301e-01 1.10141838e+00 1.21333373e+00
7.32791483e-01 -1.97704151e-01 3.63527089e-02 1.69983268e-01
4.35267359e-01 4.08577949e-01 4.15850878e-01 -1.64264774e+00
2.75374055e-01 9.56443369e-01 3.21340501e-01 -7.53730953e-01
-8.19088399e-01 1.69696242e-01 -1.57747418e-01 3.94085735e-01
-2.77489275e-01 3.47580433e-01 -8.08560908e-01 1.79410815e+00
-2.03468837e-02 -3.53576392e-02 6.58795476e-01 8.03975463e-01
3.59620780e-01 4.82821822e-01 8.42131972e-02 -1.43528981e-02
8.88046741e-01 -1.03086042e+00 -4.61695522e-01 -9.87097979e-01
1.03276777e+00 -2.64147371e-01 1.23199832e+00 6.14102483e-01
-1.07994235e+00 -2.63632655e-01 -1.28591454e+00 -1.18010558e-01
1.03847571e-01 5.76026887e-02 8.64944458e-01 -6.00738645e-01
-9.09527063e-01 6.97743893e-01 -1.28930688e+00 -4.78217572e-01
2.08318718e-02 2.71916330e-01 -5.03739417e-01 -4.11890686e-01
-5.80743074e-01 1.25793910e+00 8.23196888e-01 -1.14528798e-01
-1.71475351e+00 -2.75572836e-01 -1.34439707e+00 -8.57519954e-02
8.12029898e-01 -9.21227574e-01 2.00166368e+00 -3.16036493e-01
-1.31818151e+00 3.64534438e-01 -2.91763544e-01 -6.91323876e-01
2.29861423e-01 -6.26286447e-01 1.12623610e-02 -8.96158740e-02
5.29598594e-01 7.80874550e-01 4.38613713e-01 -1.11014616e+00
-6.03267431e-01 -1.82400540e-01 4.03338760e-01 -8.64515454e-03
5.07750511e-01 -2.65643954e-01 -3.24381471e-01 1.29062817e-01
4.45400357e-01 -1.15880322e+00 -5.13064444e-01 -5.25364995e-01
-8.56554627e-01 -4.30608749e-01 2.25552887e-01 -5.12224078e-01
9.60654259e-01 -1.99425256e+00 6.52792454e-01 5.75513653e-02
3.01433355e-01 -7.50425935e-01 -3.57742310e-01 8.85752201e-01
2.24012062e-02 2.23193675e-01 -1.65757567e-01 -5.07325709e-01
4.57063168e-01 5.54835141e-01 -6.92372441e-01 1.42947197e-01
3.38403106e-01 9.22253370e-01 -1.04387510e+00 -3.67286921e-01
3.99526268e-01 -1.73655525e-01 -7.78878868e-01 4.77425933e-01
-1.03525352e+00 3.55497539e-01 -5.75999320e-01 6.95497453e-01
-1.55877143e-01 -4.26944017e-01 4.68659669e-01 3.38280350e-01
-4.18155417e-02 6.51255429e-01 -7.70441890e-01 2.05404949e+00
-6.03087425e-01 4.92444098e-01 -3.57551724e-01 -2.62485534e-01
6.39768422e-01 2.19684765e-01 5.36193624e-02 -3.96735549e-01
5.46528175e-02 4.14395958e-01 -3.22243005e-01 -8.31305206e-01
6.95299625e-01 9.14057717e-02 -7.37732947e-01 6.19754791e-01
1.88878439e-02 -7.00827777e-01 1.55860782e-01 6.15099132e-01
1.70254087e+00 4.50989574e-01 6.85530186e-01 2.32200101e-01
1.81799039e-01 7.89696634e-01 5.25297225e-01 1.08567619e+00
5.36764413e-02 3.11253965e-01 4.66416061e-01 -5.85681498e-01
-1.07724929e+00 -7.97578752e-01 6.20793700e-01 8.15840662e-01
5.74888825e-01 -9.19427156e-01 -5.19602656e-01 -5.86841226e-01
-2.20306125e-02 1.43206418e+00 -4.31604296e-01 -5.57389796e-01
-3.44904751e-01 2.38887221e-02 5.85449517e-01 6.84185565e-01
1.33401081e-01 -1.46434581e+00 -1.23949671e+00 1.72239229e-01
-5.22669971e-01 -9.87937510e-01 9.48064849e-02 5.59241056e-01
-6.45485461e-01 -1.31275225e+00 7.11250484e-01 -6.23897254e-01
6.72116935e-01 1.41497508e-01 1.21644568e+00 6.10018671e-01
-2.03788355e-01 5.99554658e-01 -4.97066975e-01 -2.67305672e-01
-1.04164481e+00 -6.04996011e-02 3.47114414e-01 -1.15768623e+00
1.02278560e-01 -4.46988165e-01 -3.61749679e-02 1.46500573e-01
-6.82170093e-01 6.09688401e-01 6.45578623e-01 6.21507406e-01
6.87980533e-01 1.11313485e-01 2.09370285e-01 -5.02390027e-01
6.78573608e-01 -6.44603610e-01 -5.10376513e-01 2.12766349e-01
-8.12742829e-01 5.69217026e-01 4.78988081e-01 -6.94323853e-02
-7.35494733e-01 3.27917278e-01 3.14979792e-01 -2.83720374e-01
-3.36513221e-01 9.85777676e-01 1.82764865e-02 5.92247427e-01
6.99372470e-01 1.62310094e-01 -1.44337878e-01 -1.68029234e-01
4.58830565e-01 2.96186358e-01 1.07788706e+00 -8.58731866e-01
6.01960838e-01 1.27228573e-01 7.85913244e-02 -7.14179277e-02
-6.97288394e-01 -1.12657480e-01 -2.79272944e-01 -1.37626573e-01
6.01005793e-01 -5.95795870e-01 -1.16247356e+00 -3.89535986e-02
-1.47109795e+00 -9.26640809e-01 -3.28095883e-01 3.65829885e-01
-1.43025303e+00 5.58440797e-02 -5.97463429e-01 -1.16948354e+00
6.90862015e-02 -1.34046996e+00 1.09947658e+00 1.36311240e-02
-1.00616086e+00 -3.23970079e-01 1.48306966e-01 4.23297286e-02
-7.90471509e-02 1.33793861e-01 1.15980899e+00 -7.35018492e-01
-8.79705608e-01 -1.88716814e-01 2.04159230e-01 -2.00011551e-01
-1.93676844e-01 -2.22039446e-01 -4.88402814e-01 8.78341198e-02
-1.80925012e-01 -7.09594905e-01 3.44044089e-01 -6.30110735e-03
9.32975352e-01 -5.25982141e-01 -6.56441271e-01 1.55021073e-02
1.22029650e+00 -1.21447839e-01 5.63271761e-01 6.82420790e-01
2.05790743e-01 5.74827015e-01 1.20051908e+00 5.76329350e-01
6.66742265e-01 5.16550422e-01 9.93671656e-01 8.17265809e-01
-2.87865866e-02 -7.57683754e-01 7.22667277e-01 3.65540922e-01
2.71455735e-01 -3.06862593e-01 -1.37114084e+00 5.99225938e-01
-2.29399800e+00 -8.08424890e-01 -3.13071609e-02 1.63694715e+00
6.19316876e-01 2.76904196e-01 -4.85949457e-01 8.90625343e-02
7.76424259e-02 -3.67209971e-01 -8.12510014e-01 -4.01718795e-01
2.82311380e-01 -4.16361809e-01 2.06580326e-01 1.03356242e+00
-5.85585356e-01 1.21549010e+00 6.30427122e+00 2.86628872e-01
-5.03790438e-01 -4.77903187e-02 -1.31345198e-01 -6.92170337e-02
-4.13376182e-01 3.70401978e-01 -4.80362207e-01 -2.11913079e-01
1.00643814e+00 -4.16697174e-01 8.16775084e-01 1.28550732e+00
1.52610064e-01 -4.14474100e-01 -1.69082248e+00 6.28945172e-01
-5.64470589e-02 -1.28130698e+00 -1.65699959e-01 -2.77404279e-01
3.10736358e-01 3.31531048e-01 -5.26856370e-02 4.99464393e-01
8.99847984e-01 -1.29919553e+00 1.33344817e+00 7.47416914e-01
3.96134406e-01 -5.76216638e-01 6.49205029e-01 9.32172716e-01
-8.20352376e-01 -5.50724387e-01 -2.50704139e-01 -6.36158288e-01
4.68798131e-01 1.92450225e-01 -1.65736806e+00 5.58762491e-01
4.47296053e-01 9.14212167e-01 -5.06530106e-01 4.88618106e-01
-9.16973293e-01 4.43173461e-02 -2.29559675e-01 1.61374677e-02
-2.07998097e-01 4.57965583e-01 8.53066742e-01 5.47528267e-01
4.73576128e-01 6.30512685e-02 4.02529150e-01 1.17639196e+00
4.05151635e-01 -7.87856698e-01 -8.98569942e-01 -3.63149673e-01
7.23876595e-01 8.20598781e-01 -4.76789892e-01 -3.73607069e-01
1.15348093e-01 1.11590338e+00 4.66760755e-01 2.99939990e-01
-7.48007834e-01 1.38574958e-01 6.90538764e-01 -2.95327842e-01
-2.53646284e-01 -5.94227910e-01 -6.36044204e-01 -8.84801924e-01
1.24328449e-01 -1.01189590e+00 3.20605665e-01 -1.61573350e+00
-7.56007493e-01 7.28323877e-01 2.13564843e-01 -9.54275429e-01
-9.11777616e-01 -5.29012084e-01 -3.26285988e-01 4.35357451e-01
-1.13193202e+00 -1.12383056e+00 -4.59584326e-01 2.43110493e-01
9.13284421e-01 7.18191564e-02 1.07626164e+00 -6.26858771e-01
-4.80165273e-01 -2.75024652e-01 -9.83057976e-01 -6.06068730e-01
2.63586491e-01 -1.15510237e+00 5.56832969e-01 9.52013612e-01
-1.23431444e-01 1.10418046e+00 1.42589438e+00 -1.22058427e+00
-1.99673855e+00 -1.04680288e+00 9.53391254e-01 -1.11564565e+00
6.39835954e-01 4.95047197e-02 -5.16592920e-01 1.47828305e+00
-5.94246984e-02 -5.35480738e-01 1.38974011e-01 -6.69974312e-02
-4.62862909e-01 5.67332983e-01 -9.08033848e-01 9.76481795e-01
1.37560141e+00 -5.17485619e-01 -1.20811784e+00 5.38996577e-01
1.21439719e+00 -5.67493677e-01 -4.98508185e-01 5.07628858e-01
3.99800777e-01 -8.03633451e-01 6.42887115e-01 -7.17849731e-01
1.02413225e+00 -7.67632306e-01 -6.48792744e-01 -1.14053416e+00
-1.01435751e-01 -5.90931654e-01 -3.45908642e-01 6.43120050e-01
7.41290152e-01 -4.93269175e-01 3.41356933e-01 1.01513445e+00
-7.61990905e-01 -5.17557859e-01 -6.37766302e-01 -6.71380520e-01
-6.00403368e-01 -1.24668109e+00 6.69730842e-01 1.36016920e-01
9.26697135e-01 2.05519974e-01 -2.57362574e-01 7.15520501e-01
2.48348936e-01 3.12887281e-01 9.96942818e-01 -1.05214179e+00
-5.49485624e-01 -8.84118453e-02 -1.69661075e-01 -9.31077480e-01
7.42887795e-01 -1.04041731e+00 9.77194667e-01 -1.85392845e+00
4.58991945e-01 -3.11121166e-01 1.69220656e-01 1.23431075e+00
2.90400684e-01 -1.86810285e-01 2.17521057e-01 4.60406870e-01
-1.12843060e+00 3.31886232e-01 7.51440108e-01 -1.71127364e-01
-2.80574769e-01 -4.59024340e-01 -6.25375926e-01 9.61886346e-01
8.84412706e-01 -6.67114556e-01 -4.23212379e-01 -6.65259123e-01
9.16183650e-01 5.73589563e-01 7.27392614e-01 -1.01598132e+00
4.63844895e-01 -6.70937359e-01 -2.19246849e-01 -3.84175569e-01
3.74823153e-01 -7.38906324e-01 5.23760796e-01 8.90674174e-01
-7.88500905e-01 3.79545331e-01 3.06709319e-01 6.37085378e-01
-4.60518077e-02 -3.33925724e-01 -3.72013189e-02 -1.08341336e-01
-1.18640697e+00 -4.94757108e-02 -7.86730230e-01 -4.29503679e-01
1.02614379e+00 1.33980647e-01 -5.26557803e-01 -4.22806442e-01
-7.14559853e-01 6.29028201e-01 9.57045078e-01 7.04150796e-01
8.91123772e-01 -9.84030128e-01 -1.74940735e-01 -4.51354794e-02
5.94713271e-01 1.55239971e-02 2.69407663e-03 7.46603966e-01
-3.63091826e-01 5.22548854e-01 -1.42343163e-01 -3.99338037e-01
-8.47964227e-01 6.47113562e-01 1.47706509e-01 -9.55792069e-02
-8.90605927e-01 5.91468930e-01 5.44063710e-02 -7.28218019e-01
3.36049050e-02 -8.30281436e-01 2.25331690e-02 -1.04212546e+00
5.44508457e-01 7.93574899e-02 -1.90460533e-01 -1.62990600e-01
-7.83346236e-01 4.15166095e-02 3.21751326e-01 -5.93635380e-01
1.38575029e+00 -2.97171950e-01 -1.77766904e-01 6.68380022e-01
3.72209281e-01 -2.63302028e-01 -1.21478343e+00 2.40478277e-01
4.67739254e-01 -1.28734708e-01 -3.96428347e-01 -1.09546971e+00
-9.84948725e-02 4.39719915e-01 -4.23325241e-01 1.73797850e-02
6.70203865e-01 5.24436414e-01 4.57632065e-01 9.08758223e-01
1.09938717e+00 -7.93145299e-01 5.42062581e-01 9.26046729e-01
1.42314839e+00 -1.01624084e+00 -2.72204310e-01 -3.89273852e-01
-7.88046837e-01 1.13200486e+00 8.85450304e-01 2.28604432e-02
-2.63564140e-01 5.28792977e-01 -2.28849471e-01 -4.83036995e-01
-1.40875828e+00 2.25799810e-03 -2.48323634e-01 6.07234240e-01
-7.65053481e-02 1.92669377e-01 1.66032836e-01 1.12352502e+00
-5.28562069e-01 3.27499449e-01 9.94864225e-01 1.25654745e+00
-6.14862144e-01 -8.83716226e-01 -1.02111362e-01 1.76701933e-01
1.12781264e-01 1.35165513e-01 -6.83779299e-01 6.21476352e-01
-2.89003164e-01 1.33474183e+00 -2.56805599e-01 -8.36298287e-01
3.38861495e-01 2.31920630e-01 5.68770587e-01 -1.13196361e+00
-1.16863623e-01 -5.27117074e-01 4.96187180e-01 -1.48038054e+00
-5.17993532e-02 -6.30671382e-01 -2.22389054e+00 -1.91172153e-01
3.51384878e-02 -1.24913054e-02 6.10620022e-01 1.25489831e+00
4.96968746e-01 7.40045190e-01 9.94206518e-02 -9.35594082e-01
-5.73429108e-01 -8.35357785e-01 -3.82646956e-02 2.22588569e-01
3.18680942e-01 -6.76601231e-01 -4.17680144e-01 1.38091296e-01] | [4.418815612792969, 0.9061102271080017] |
7d747612-c5f9-4d08-af5d-9ed28964fa99 | from-pretraining-data-to-language-models-to | 2305.08283 | null | https://arxiv.org/abs/2305.08283v3 | https://arxiv.org/pdf/2305.08283v3.pdf | From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Models | Language models (LMs) are pretrained on diverse data sources, including news, discussion forums, books, and online encyclopedias. A significant portion of this data includes opinions and perspectives which, on one hand, celebrate democracy and diversity of ideas, and on the other hand are inherently socially biased. Our work develops new methods to (1) measure political biases in LMs trained on such corpora, along social and economic axes, and (2) measure the fairness of downstream NLP models trained on top of politically biased LMs. We focus on hate speech and misinformation detection, aiming to empirically quantify the effects of political (social, economic) biases in pretraining data on the fairness of high-stakes social-oriented tasks. Our findings reveal that pretrained LMs do have political leanings that reinforce the polarization present in pretraining corpora, propagating social biases into hate speech predictions and misinformation detectors. We discuss the implications of our findings for NLP research and propose future directions to mitigate unfairness. | ['Yulia Tsvetkov', 'YuHan Liu', 'Chan Young Park', 'Shangbin Feng'] | 2023-05-15 | null | null | null | null | ['misinformation'] | ['miscellaneous'] | [-2.54139483e-01 7.79691100e-01 -5.21964610e-01 -4.54551965e-01
-3.51190150e-01 -7.62058914e-01 1.25504601e+00 4.05112714e-01
-6.50883377e-01 7.09844410e-01 1.15611899e+00 -6.10764325e-01
2.87306339e-01 -6.27828360e-01 -5.16970277e-01 -2.87840545e-01
4.85388756e-01 2.06769690e-01 -3.86027128e-01 -2.42558554e-01
6.09887064e-01 2.45163128e-01 -8.82287562e-01 5.33681810e-01
1.10635710e+00 5.53753912e-01 -7.77906060e-01 3.57553720e-01
-2.56449401e-01 1.58804333e+00 -7.33000994e-01 -1.35202456e+00
3.79154310e-02 -2.71112770e-01 -7.10782886e-01 -5.76239944e-01
1.01999116e+00 -2.02200130e-01 -3.69722784e-01 1.45286739e+00
7.06471562e-01 -1.36135891e-01 1.05152106e+00 -1.05072951e+00
-1.30766606e+00 1.34348965e+00 -4.71642375e-01 4.68479216e-01
-2.82849461e-01 4.87335980e-01 1.19390130e+00 -8.19812000e-01
1.00237286e+00 1.75506926e+00 5.98659635e-01 6.74974442e-01
-1.16652513e+00 -1.13601494e+00 1.98469102e-01 -2.98760206e-01
-6.96719646e-01 -8.15031350e-01 6.09333873e-01 -1.01003468e+00
4.22413915e-01 1.45231053e-01 3.23258787e-01 1.94586468e+00
4.38385010e-01 6.43287182e-01 1.55201709e+00 -7.75520280e-02
4.55637984e-02 5.29790401e-01 3.78187269e-01 6.79429352e-01
5.14370859e-01 3.44436705e-01 -7.84871519e-01 -4.28337693e-01
-4.29779291e-02 -3.55025530e-01 4.11057547e-02 2.15339258e-01
-1.05281425e+00 1.38400960e+00 7.00788856e-01 3.68624151e-01
-2.23995045e-01 -2.02992544e-01 5.11251628e-01 4.38983321e-01
1.25298619e+00 1.04000676e+00 -1.56166255e-01 5.77819049e-02
-9.23317969e-01 3.05182904e-01 1.07052791e+00 4.03938293e-01
4.01910961e-01 -6.85532913e-02 -4.74366844e-01 8.00611436e-01
2.91193813e-01 8.91381025e-01 5.57761252e-01 -9.83975768e-01
7.99698532e-01 2.83090413e-01 5.58905229e-02 -1.73602164e+00
-3.51363331e-01 -6.80033147e-01 -7.51846552e-01 -1.55586526e-02
5.39240837e-01 -6.38858318e-01 -2.86423653e-01 1.84646010e+00
-1.09079354e-01 -5.43777823e-01 -1.81276053e-01 7.54935145e-01
7.19257057e-01 5.27853191e-01 7.18599081e-01 -1.48105085e-01
9.26239967e-01 -7.81670630e-01 -5.50409317e-01 -4.91786450e-01
9.91750717e-01 -9.28252637e-01 1.05241632e+00 -1.00112349e-01
-9.37564850e-01 1.53052472e-02 -5.80175459e-01 -3.80900294e-01
-5.01070142e-01 -1.79524094e-01 1.45972282e-01 7.47934103e-01
-8.27582300e-01 7.37912416e-01 -9.67870001e-03 -1.03917360e-01
8.91498923e-01 -4.97796685e-01 3.15397121e-02 3.02835315e-01
-1.53273082e+00 1.53198433e+00 -2.44091868e-01 -1.94388941e-01
-7.89226770e-01 -8.54213834e-01 -6.53460801e-01 9.09480155e-02
2.40870878e-01 -3.01115870e-01 1.14319801e+00 -1.48327196e+00
-1.25528145e+00 1.29859805e+00 2.15111196e-01 -3.97686094e-01
9.04071927e-01 -3.83723706e-01 -3.84225905e-01 -3.62769246e-01
3.88212860e-01 4.73156005e-01 8.31880927e-01 -1.02742016e+00
-2.37397775e-01 -1.98601201e-01 -1.16904750e-01 1.28259733e-01
-7.48379648e-01 3.99220109e-01 1.13663280e+00 -7.42365062e-01
-6.42010689e-01 -7.81343102e-01 1.08227015e-01 -1.41665444e-01
-7.99448907e-01 -1.72212780e-01 3.90472412e-01 -8.75338972e-01
1.37902987e+00 -2.04724216e+00 -3.72854993e-02 2.22688124e-01
7.15881944e-01 6.35644674e-01 -1.06807763e-03 3.68789643e-01
5.39242588e-02 1.00446188e+00 1.73298135e-01 -2.39202917e-01
3.67334068e-01 -2.76587278e-01 -7.90606916e-01 6.44510090e-01
3.51483487e-02 9.07379150e-01 -9.77058232e-01 -5.77055633e-01
-5.86283095e-02 1.22719288e-01 -7.80757487e-01 -1.56938002e-01
-2.09017441e-01 2.44530559e-01 -2.16807455e-01 1.31385952e-01
4.00703967e-01 -9.96160954e-02 7.10041523e-02 1.54234648e-01
-5.54800749e-01 1.07897687e+00 -5.04580215e-02 6.63130462e-01
-2.63737798e-01 1.29729533e+00 1.69500619e-01 -4.62207407e-01
7.24292576e-01 -7.22756386e-02 -2.42157310e-01 -6.87446535e-01
6.69429779e-01 5.68924770e-02 1.57919079e-01 -5.04108310e-01
6.05004787e-01 -5.88934779e-01 -2.32941985e-01 7.43392885e-01
-1.75110653e-01 7.28471130e-02 -2.37319186e-01 5.11494637e-01
6.77061260e-01 -5.95027208e-01 2.16730297e-01 -6.78491473e-01
4.36169565e-01 -2.35806927e-02 5.53107023e-01 8.02150726e-01
-6.69894636e-01 -9.54732224e-02 1.08033371e+00 -4.68787342e-01
-1.24217522e+00 -7.71162331e-01 -2.99610160e-02 1.51365781e+00
-3.46188098e-01 -2.41889700e-01 -6.24858975e-01 -1.04155087e+00
3.48381042e-01 1.35229957e+00 -6.70267045e-01 -2.72093296e-01
-3.88284355e-01 -5.83869159e-01 9.85297799e-01 -1.13021836e-01
2.92532623e-01 -9.63566840e-01 -1.86975494e-01 -3.10276270e-01
-3.11358631e-01 -7.94025064e-01 -4.44895416e-01 -3.39321733e-01
-3.84786278e-01 -1.05412948e+00 -4.77534026e-01 -3.22243810e-01
5.66801131e-01 4.60498296e-02 1.11844480e+00 1.25695571e-01
3.21400136e-01 4.77595208e-03 -3.21157463e-02 -8.45932961e-01
-1.25143385e+00 1.98429897e-01 9.91846398e-02 -1.19999044e-01
4.23977166e-01 -2.80848742e-01 -1.94095701e-01 1.08384967e-01
-6.34942353e-01 1.18832678e-01 4.59070593e-01 6.95158720e-01
-5.15996337e-01 -5.04890859e-01 4.76097226e-01 -1.55115092e+00
1.07197642e+00 -8.72162282e-01 -4.69568282e-01 1.54218450e-01
-8.04776549e-01 1.25990495e-01 6.16751611e-01 -2.91287661e-01
-1.45854807e+00 -9.78676975e-01 2.05847204e-01 1.08540058e-01
1.26568422e-01 5.73110700e-01 5.43680429e-01 -1.38882458e-01
1.21228528e+00 -5.75765312e-01 9.94265378e-02 -4.42097723e-01
6.61424458e-01 9.40684915e-01 1.03945263e-01 -5.44318795e-01
9.33651745e-01 2.77209282e-01 -3.61668229e-01 -9.16855812e-01
-1.44409418e+00 -6.12908490e-02 -3.57548773e-01 -4.62106407e-01
7.39226580e-01 -8.53362620e-01 -6.69752717e-01 3.71792436e-01
-1.19248724e+00 -3.22268754e-01 3.90303656e-02 3.82333368e-01
-2.14166343e-02 2.08806202e-01 -1.04888451e+00 -1.02919221e+00
-3.58110934e-01 -5.69547176e-01 4.25726891e-01 -5.67345358e-02
-7.43588984e-01 -1.04627335e+00 2.64361054e-01 6.83384359e-01
6.46192610e-01 3.00398469e-01 1.22354484e+00 -1.00810456e+00
3.85482609e-02 9.18097273e-02 -3.98450434e-01 5.22769570e-01
-3.98694038e-01 3.84244740e-01 -1.03159547e+00 7.46643096e-02
-6.98951036e-02 -7.66161382e-01 1.03247058e+00 7.87144378e-02
4.91335601e-01 -1.06917703e+00 -3.07085644e-02 2.77284142e-02
8.32755864e-01 -5.99796355e-01 1.68775380e-01 2.18108073e-01
5.23275495e-01 1.12155151e+00 1.17900573e-01 2.97281653e-01
4.26020563e-01 6.38001189e-02 1.05516970e-01 1.91608608e-01
-2.09306758e-02 -8.76121163e-01 7.95492649e-01 9.44797039e-01
2.86965489e-01 -1.74980670e-01 -1.13265562e+00 3.33922416e-01
-1.55007362e+00 -1.34005404e+00 -1.12709612e-01 1.78700221e+00
8.54461074e-01 9.30440649e-02 -2.93073617e-02 -4.47780699e-01
9.57219303e-01 7.46617317e-01 -3.01771492e-01 -8.57647955e-01
-4.55065101e-01 -3.20873022e-01 8.61374140e-01 7.90323853e-01
-1.04392314e+00 1.23424196e+00 6.37606764e+00 4.48377550e-01
-1.25029111e+00 3.31752867e-01 1.07793486e+00 -3.94558132e-01
-7.00395644e-01 -3.80513132e-01 -5.60324192e-01 6.53882921e-01
1.01220274e+00 -7.00387061e-01 2.35810593e-01 9.55094993e-01
3.91007423e-01 1.36714116e-01 -8.67733300e-01 3.62819761e-01
1.85148373e-01 -1.30390108e+00 -2.23413445e-02 3.46575618e-01
1.07909846e+00 3.37655783e-01 5.84918022e-01 5.47999322e-01
1.08484232e+00 -1.07675087e+00 1.18348145e+00 2.08425298e-01
3.69830847e-01 -3.92576933e-01 7.65264988e-01 7.35188007e-01
3.46473306e-01 -3.18844050e-01 -5.64086139e-01 -5.34039617e-01
-1.87018719e-02 8.22169840e-01 -6.92980409e-01 -2.88208395e-01
3.58720392e-01 7.30918050e-01 -6.21093631e-01 3.07399482e-01
-6.79673493e-01 8.83920908e-01 1.52363405e-01 -4.69310522e-01
5.59084892e-01 7.39384219e-02 6.23992026e-01 1.39117277e+00
-1.05093852e-01 2.66291201e-02 -4.64444876e-01 1.14814925e+00
-7.63580739e-01 3.23045999e-01 -9.84647810e-01 -8.12384963e-01
5.00302792e-01 1.05730224e+00 -1.75425991e-01 -3.19322199e-01
-4.29974884e-01 3.68119150e-01 9.15175736e-01 2.31630608e-01
-6.22124195e-01 2.82515526e-01 9.38816488e-01 9.32288095e-02
-4.81830567e-01 -3.77337378e-03 -5.19035876e-01 -1.48067367e+00
-4.75374073e-01 -1.22202849e+00 3.57652068e-01 -6.10182703e-01
-1.87418342e+00 -3.54756601e-02 -4.66884494e-01 -2.82524645e-01
-2.92433513e-04 -6.94039464e-01 -4.12072957e-01 8.16840410e-01
-1.41454518e+00 -7.59560466e-01 2.13073552e-01 9.70725901e-03
1.85670182e-01 -2.40738064e-01 5.62199175e-01 1.27156422e-01
-7.49443293e-01 3.48379701e-01 7.39835724e-02 5.73382676e-01
1.24098837e+00 -8.16682875e-01 4.83729571e-01 7.12628365e-01
-3.11349239e-02 6.63789213e-01 7.89071321e-01 -8.42052579e-01
-4.72358197e-01 -9.55127776e-01 1.71258104e+00 -9.45048213e-01
1.18821347e+00 -2.50583142e-01 -5.93433738e-01 9.27932978e-01
3.46751481e-01 -6.47526145e-01 1.05458379e+00 5.81345737e-01
-1.07666361e+00 4.97621685e-01 -1.15290606e+00 8.06158602e-01
1.20955241e+00 -9.01871264e-01 -9.54225600e-01 5.24742484e-01
5.37295759e-01 -2.09576547e-01 -3.57219994e-01 -1.67099237e-01
5.77673435e-01 -1.14683163e+00 5.14843524e-01 -1.39782262e+00
1.43778372e+00 5.89155436e-01 1.78853676e-01 -1.77161181e+00
-6.91477478e-01 -5.47468364e-01 3.65501583e-01 1.24170041e+00
7.34575212e-01 -7.11456358e-01 5.26684582e-01 1.07317197e+00
2.72509873e-01 -9.81080383e-02 -6.00242734e-01 -2.39559814e-01
8.27746153e-01 -4.17233780e-02 3.01216580e-02 1.64829838e+00
2.52115101e-01 5.92693150e-01 -4.65892524e-01 -1.21304378e-01
6.97373807e-01 -1.28051952e-01 4.52498227e-01 -1.31844175e+00
3.69161695e-01 -8.91266227e-01 1.94392949e-02 -4.49740410e-01
8.08437169e-01 -1.19889569e+00 -2.79623598e-01 -8.81705403e-01
2.73072034e-01 -2.58159995e-01 5.05997688e-02 1.24372184e-01
-2.11116493e-01 -7.00822026e-02 7.17787623e-01 3.56176555e-01
-4.35264468e-01 3.77390653e-01 1.13705003e+00 -2.09776193e-01
1.94201529e-01 -3.45268279e-01 -1.33944130e+00 1.15942228e+00
6.93162203e-01 -7.59091437e-01 1.14820816e-01 -7.21462488e-01
9.00309145e-01 -5.44073880e-01 4.53809321e-01 -3.71270686e-01
1.09005712e-01 -4.63121146e-01 2.04587430e-01 1.71595708e-01
-1.40062064e-01 -4.76005167e-01 -4.63315099e-01 6.05043709e-01
-1.36369181e+00 -9.98716056e-02 -1.11334242e-01 5.24486780e-01
2.95476094e-02 -1.02913558e-01 9.68237936e-01 -2.57315099e-01
-2.13071536e-02 -8.33820030e-02 -6.70181811e-01 7.49613464e-01
5.38507998e-01 6.20242059e-01 -1.17166090e+00 -7.45746493e-01
-4.73100245e-01 1.65579885e-01 4.96896625e-01 5.34234166e-01
-1.15406118e-01 -1.10941339e+00 -1.08478999e+00 -1.95665911e-01
-1.51136620e-02 -8.99418175e-01 -9.87551138e-02 7.59218812e-01
-3.44404668e-01 5.53055346e-01 -2.58744687e-01 3.17772388e-01
-1.11475921e+00 4.04069841e-01 2.53139406e-01 -7.11277649e-02
-7.73836225e-02 1.12073624e+00 4.50492561e-01 -4.88936126e-01
2.14231033e-02 1.46106645e-01 -1.93089664e-01 5.88032484e-01
4.24089968e-01 7.17069805e-01 -5.17498970e-01 -1.01848173e+00
-7.09789917e-02 -1.21361621e-01 -5.17564528e-02 -1.82006031e-01
1.26373875e+00 -1.19981006e-01 -4.75581050e-01 5.51160157e-01
1.11092758e+00 6.73127949e-01 -5.96998215e-01 -4.86733556e-01
3.40061784e-01 -6.87505126e-01 1.29356071e-01 -1.08261549e+00
-6.42220080e-01 1.04556584e+00 -3.32359493e-01 3.55029345e-01
-1.12445559e-02 -1.95190832e-01 4.17532116e-01 2.78835744e-01
-8.95014219e-03 -1.45972061e+00 8.16665068e-02 6.25924706e-01
9.90746021e-01 -1.13463354e+00 -5.80835529e-02 -3.21074933e-01
-1.01179659e+00 9.41622019e-01 5.51109791e-01 2.28672624e-02
6.07801914e-01 -4.36267667e-02 4.78783429e-01 -2.85077661e-01
-9.52839553e-01 2.30397493e-01 1.41633928e-01 1.59790859e-01
8.48759234e-01 3.59484792e-01 -7.61174083e-01 6.35900617e-01
-6.31030738e-01 -3.66259187e-01 5.89795828e-01 1.86594933e-01
-6.52607322e-01 -5.24593830e-01 -2.37803951e-01 5.45404196e-01
-8.86910915e-01 -6.14688635e-01 -1.37039852e+00 4.61619824e-01
1.15721002e-01 8.95983636e-01 6.66364431e-02 -3.67409766e-01
-4.06689718e-02 4.09796476e-01 -3.27602565e-01 -5.61101854e-01
-1.09794593e+00 -6.77166343e-01 9.63657558e-01 -4.84068334e-01
-1.53481603e-01 -6.95107043e-01 -1.91684559e-01 -1.22041118e+00
-7.61334300e-02 9.26907286e-02 5.81430256e-01 9.02316570e-01
5.49457431e-01 2.01493707e-02 4.18221951e-01 -3.96032006e-01
-1.03000712e+00 -9.45924282e-01 -2.42935687e-01 7.14942873e-01
3.33554357e-01 -3.49732757e-01 -8.85274351e-01 -3.57142389e-01] | [8.89455509185791, 10.280733108520508] |
64f25a0f-204e-4782-a9fa-9bb5f95e8ebb | zero-shot-query-contextualization-for | 2204.10613 | null | https://arxiv.org/abs/2204.10613v1 | https://arxiv.org/pdf/2204.10613v1.pdf | Zero-shot Query Contextualization for Conversational Search | Current conversational passage retrieval systems cast conversational search into ad-hoc search by using an intermediate query resolution step that places the user's question in context of the conversation. While the proposed methods have proven effective, they still assume the availability of large-scale question resolution and conversational search datasets. To waive the dependency on the availability of such data, we adapt a pre-trained token-level dense retriever on ad-hoc search data to perform conversational search with no additional fine-tuning. The proposed method allows to contextualize the user question within the conversation history, but restrict the matching only between question and potential answer. Our experiments demonstrate the effectiveness of the proposed approach. We also perform an analysis that provides insights of how contextualization works in the latent space, in essence introducing a bias towards salient terms from the conversation. | ['Evangelos Kanoulas', 'Andrew Yates', 'Antonios Minas Krasakis'] | 2022-04-22 | null | null | null | null | ['passage-retrieval', 'conversational-search'] | ['natural-language-processing', 'natural-language-processing'] | [ 1.58273488e-01 1.25272870e-01 -2.04244271e-01 -3.55753243e-01
-1.37461925e+00 -6.87115192e-01 1.06309974e+00 3.77661526e-01
-6.68036044e-01 5.77477694e-01 8.31438661e-01 -3.14476222e-01
-4.05904531e-01 -7.41669357e-01 -1.89381316e-01 -4.25200343e-01
1.43596977e-01 6.79697752e-01 5.08820951e-01 -5.58880508e-01
4.69702452e-01 2.54200608e-01 -1.56581819e+00 7.05403566e-01
8.90235901e-01 5.42583883e-01 4.69962537e-01 9.01158035e-01
-4.20932174e-01 7.49858379e-01 -7.60581434e-01 -2.42368221e-01
-3.30280066e-02 -7.19687879e-01 -1.29621959e+00 -1.01527460e-01
3.92407775e-01 -3.52177620e-01 -4.19557720e-01 6.10030830e-01
5.58254182e-01 8.60853672e-01 2.14196354e-01 -8.13858926e-01
-4.20263320e-01 5.48247874e-01 2.18086690e-01 5.74554205e-01
8.75498712e-01 4.20781877e-03 1.44748664e+00 -9.44075882e-01
8.72846723e-01 1.20451450e+00 3.34997058e-01 6.06813848e-01
-9.28489387e-01 -3.52323025e-01 3.46721530e-01 5.37402451e-01
-1.20021367e+00 -5.82707405e-01 8.84726465e-01 -2.04905123e-01
1.20116162e+00 7.00176656e-01 4.10448670e-01 9.92768526e-01
-3.09887141e-01 7.43734241e-01 7.57196128e-01 -7.16639936e-01
2.80021876e-01 2.37865582e-01 4.99730200e-01 3.62274766e-01
-4.38693672e-01 -3.54720980e-01 -7.78488278e-01 -7.41125405e-01
1.79931700e-01 -8.26633647e-02 -5.05840063e-01 -2.57876486e-01
-7.85368145e-01 8.41833770e-01 2.50537068e-01 5.92626870e-01
-5.40000260e-01 -3.08413148e-01 3.56808692e-01 5.56728601e-01
4.73850667e-01 7.04636395e-01 -2.21148685e-01 -4.48548079e-01
-1.12678587e+00 4.37598825e-01 1.18877745e+00 9.08950806e-01
8.34366918e-01 -8.41615736e-01 -6.98567390e-01 1.19108510e+00
-7.03645172e-03 5.82727864e-02 4.92085993e-01 -9.65871453e-01
5.78191221e-01 5.06959379e-01 4.25969571e-01 -9.22647476e-01
-1.77666515e-01 -2.75533825e-01 -5.23523390e-02 -6.39535248e-01
3.64205539e-01 1.19882626e-02 -3.66327733e-01 1.69340515e+00
4.61001635e-01 -2.36291327e-02 -3.20040202e-03 9.11274910e-01
6.50092244e-01 6.43146992e-01 8.46373737e-02 -3.51398051e-01
1.60090363e+00 -1.18269622e+00 -7.96409726e-01 -1.80760235e-01
5.33654571e-01 -9.42430913e-01 1.39231944e+00 -1.82632700e-01
-1.10031152e+00 -2.65835226e-01 -7.71574736e-01 -4.64664906e-01
-1.99976623e-01 -2.20514134e-01 3.96479785e-01 3.42778206e-01
-9.99736488e-01 2.61392206e-01 -4.03354883e-01 -8.92966390e-01
-3.61961424e-01 -1.12032262e-03 2.43229140e-02 -2.43004873e-01
-1.54663193e+00 9.76284087e-01 -1.44316569e-01 8.27253386e-02
-6.93712354e-01 -4.93366569e-01 -4.35547382e-01 5.01458406e-01
6.43001437e-01 -7.87233949e-01 1.57678938e+00 -5.94955325e-01
-1.54459691e+00 7.39114583e-01 -7.77395070e-01 -3.34969699e-01
2.46934980e-01 -3.13855052e-01 -2.91310579e-01 6.50331259e-01
1.07379802e-01 2.97830522e-01 6.60752594e-01 -1.00586104e+00
-7.95569897e-01 -1.04087479e-01 7.30175376e-01 5.90985179e-01
-2.73024738e-01 1.60334975e-01 -8.06862891e-01 -3.36831331e-01
1.05796799e-01 -8.82766604e-01 8.06151256e-02 -4.62423980e-01
-1.22792512e-01 -8.39015305e-01 7.70989478e-01 -6.44169629e-01
1.59256947e+00 -1.91038024e+00 -2.24738270e-02 9.41589028e-02
3.68870632e-03 7.94907063e-02 -2.74003834e-01 1.30031359e+00
3.97518009e-01 -3.58539261e-02 2.86966208e-02 -5.44610560e-01
2.12932810e-01 9.65029299e-02 -5.22338092e-01 2.62657434e-01
-1.63696870e-01 8.72033358e-01 -1.06090581e+00 -7.42356241e-01
-8.20762366e-02 2.63030201e-01 -8.24692786e-01 5.63026428e-01
-3.84168029e-01 2.60305703e-01 -7.06525743e-01 2.64455229e-01
2.80587256e-01 -2.46239677e-01 2.97688603e-01 1.02121338e-01
-1.47253588e-01 1.16057146e+00 -5.75914681e-01 1.88934720e+00
-6.88439846e-01 6.85361624e-01 3.82546097e-01 -7.44310200e-01
6.12909138e-01 5.02327800e-01 3.24380755e-01 -8.90041947e-01
-3.36640745e-01 -6.91847205e-02 -1.06368951e-01 -7.14434445e-01
9.87019598e-01 -1.48822963e-01 -1.85458943e-01 8.33538353e-01
-6.88540190e-02 6.20030798e-02 1.34355366e-01 5.45588970e-01
1.21312082e+00 -1.28492057e-01 3.95412520e-02 -2.19955564e-01
7.50262320e-01 2.15890840e-01 1.57801464e-01 1.33890009e+00
-2.08230168e-01 2.59537727e-01 3.00298870e-01 4.69133630e-02
-7.40248501e-01 -7.79575109e-01 1.99673343e-02 1.74360204e+00
2.21756384e-01 -4.78692800e-01 -6.55744553e-01 -6.05852544e-01
-2.03431159e-01 8.39938343e-01 -4.49593574e-01 -8.53479654e-02
-9.30672705e-01 -3.38763297e-01 4.39094901e-01 -1.45552695e-01
4.39060450e-01 -1.28001821e+00 -3.97039264e-01 2.28984743e-01
-8.69375944e-01 -1.01275790e+00 -6.42176747e-01 -1.78473443e-01
-5.30905068e-01 -1.01443815e+00 -5.08422077e-01 -8.51846516e-01
2.84544855e-01 4.91586804e-01 1.27029836e+00 2.91549832e-01
-3.28522809e-02 9.19155002e-01 -5.97164989e-01 2.47949570e-01
-2.73621023e-01 6.50355101e-01 -3.73924613e-01 -1.07816391e-01
7.01339841e-01 -6.74671710e-01 -1.04249299e+00 3.73255551e-01
-7.48854637e-01 -3.18832695e-01 1.33120507e-01 8.83834660e-01
2.48900075e-02 -3.47101897e-01 9.66661632e-01 -7.75537312e-01
1.40235305e+00 -6.50136590e-01 -2.03467250e-01 4.35953528e-01
-7.39322424e-01 6.03440329e-02 2.87404180e-01 -2.59487778e-01
-1.25796652e+00 -5.33122003e-01 -2.20838711e-01 4.26541418e-02
-2.09116116e-02 5.99408925e-01 -1.47163728e-03 3.14706624e-01
5.73942542e-01 3.10638666e-01 -3.99098217e-01 -7.41441071e-01
5.78993559e-01 7.62826562e-01 4.70521063e-01 -7.06049740e-01
4.61673498e-01 3.30211610e-01 -6.44950867e-01 -8.43385160e-01
-8.49897385e-01 -1.23307061e+00 -3.93549651e-01 -2.39747390e-01
6.69545829e-01 -6.04213238e-01 -9.48410511e-01 -2.04231963e-01
-1.07162130e+00 -2.14103520e-01 -1.39113337e-01 3.13506395e-01
-3.49362701e-01 6.48688078e-01 -7.83575356e-01 -9.26789582e-01
-4.01392400e-01 -6.85472131e-01 8.90555680e-01 1.05518430e-01
-5.18607974e-01 -8.81388128e-01 4.04678434e-01 7.31970847e-01
6.24725997e-01 -4.79615867e-01 1.10430527e+00 -1.14227915e+00
-8.30316007e-01 -3.45636785e-01 -1.00008249e-01 -1.73285678e-01
1.98912948e-01 -4.82501477e-01 -1.12922108e+00 -1.92902550e-01
1.21419862e-01 -1.86419949e-01 7.89339840e-01 -1.40519917e-01
5.35419285e-01 -4.20703560e-01 -3.20370764e-01 -5.76811805e-02
9.76102948e-01 1.13266639e-01 3.34916860e-01 4.58753169e-01
1.75490767e-01 9.53767240e-01 5.66236138e-01 2.98143297e-01
6.36839271e-01 8.48541737e-01 1.22787198e-02 2.74716586e-01
-8.63344222e-02 -4.24466938e-01 5.04704379e-02 9.20776904e-01
3.79897267e-01 -4.11099970e-01 -7.61617422e-01 8.13658118e-01
-1.99136353e+00 -1.24087882e+00 3.90096426e-01 2.17230225e+00
1.05393171e+00 -3.90035287e-02 6.55334536e-03 -2.82563120e-01
7.39483893e-01 3.37840080e-01 -3.30359459e-01 -4.70802546e-01
1.28938556e-01 1.56345233e-01 -1.89011022e-01 1.19377899e+00
-7.10828960e-01 8.87061417e-01 6.59092522e+00 5.77244282e-01
-8.65153968e-01 1.74534589e-01 -3.79771404e-02 -3.80740970e-01
-5.76315105e-01 2.75026113e-01 -7.10926354e-01 1.60302520e-01
9.03220594e-01 -3.87215108e-01 5.68371773e-01 6.40827298e-01
3.63411635e-01 -2.02307627e-01 -1.34753585e+00 6.68898106e-01
1.90549985e-01 -1.24905980e+00 5.28499931e-02 -1.95271611e-01
2.80989021e-01 -1.83433339e-01 -3.19105864e-01 8.60528767e-01
9.47554968e-03 -6.15323484e-01 1.15307979e-01 6.33339643e-01
-9.66847781e-03 -5.87026417e-01 5.46644211e-01 7.53361881e-01
-9.31124330e-01 -8.59897509e-02 -2.24769190e-01 -2.15476409e-01
1.76650122e-01 1.59325033e-01 -1.34847438e+00 3.01734507e-01
3.95279825e-01 -3.65245193e-02 -4.31533188e-01 8.64886642e-01
-1.10754013e-01 6.30442500e-01 -2.88255394e-01 -2.61723518e-01
4.05926555e-01 -1.39280125e-01 8.08884203e-01 1.49269581e+00
2.71697044e-02 2.80595571e-01 3.24983209e-01 5.75857282e-01
-1.58041805e-01 3.58934641e-01 -2.78136075e-01 7.22257495e-02
8.67173493e-01 1.04967320e+00 -3.22133064e-01 -5.29166818e-01
-3.81637424e-01 1.06152165e+00 3.89523268e-01 7.19726682e-01
-2.78710753e-01 -4.03487891e-01 6.77628458e-01 2.04006303e-02
3.58719468e-01 -7.23412260e-02 2.50760555e-01 -1.17653787e+00
3.78265291e-01 -9.36311901e-01 6.58098936e-01 -7.06607938e-01
-1.15726185e+00 5.53167522e-01 1.66705519e-01 -7.10995436e-01
-8.62474799e-01 1.83207631e-01 -6.30912900e-01 1.13642216e+00
-1.59020388e+00 -8.66502285e-01 -1.35790423e-01 6.23735607e-01
7.99207628e-01 2.54694402e-01 1.01830876e+00 5.00055850e-01
-2.54147977e-01 5.81496239e-01 -1.25461677e-02 1.85409840e-02
9.56717253e-01 -1.01708221e+00 1.94067836e-01 6.34528458e-01
-3.48756425e-02 1.25095630e+00 9.46773231e-01 -4.34724867e-01
-1.35927606e+00 -3.49159807e-01 1.60632288e+00 -5.95031202e-01
8.36956143e-01 -6.32300675e-01 -1.20209146e+00 4.77862954e-01
6.31406784e-01 -5.96791685e-01 7.74361372e-01 8.25550556e-01
-2.68326610e-01 2.90187970e-02 -8.75316381e-01 7.72002101e-01
8.66223395e-01 -1.35470784e+00 -1.26958346e+00 3.67194951e-01
9.25886512e-01 -1.90566570e-01 -5.97571969e-01 9.83945727e-02
5.37050843e-01 -8.46542716e-01 9.81614232e-01 -6.24663591e-01
4.56675030e-02 2.75288802e-02 -4.01845351e-02 -9.66435075e-01
-1.59965307e-01 -9.10134733e-01 -4.47520167e-02 1.36461985e+00
5.26021063e-01 -4.89702076e-01 6.25520885e-01 1.07799351e+00
-1.83031991e-01 -6.11776829e-01 -1.19747615e+00 -2.99926668e-01
2.09352225e-02 -2.51861066e-01 5.91450334e-01 9.46060061e-01
7.32837617e-01 6.45133495e-01 -2.25831538e-01 1.69463679e-01
2.89414823e-02 6.13835454e-01 7.58091748e-01 -9.53148544e-01
-3.13625634e-01 -4.27408099e-01 2.94048250e-01 -1.59273005e+00
2.85989553e-01 -7.94969916e-01 2.22090840e-01 -1.66906202e+00
1.57046288e-01 -3.25623900e-01 -4.98213738e-01 1.03447877e-01
-4.46051657e-01 -3.16596121e-01 7.31854290e-02 4.91078049e-01
-9.51354146e-01 5.71291625e-01 1.10443997e+00 -3.70542072e-02
-4.25767034e-01 1.09489262e-01 -6.52462482e-01 3.18163961e-01
5.21047413e-01 -5.14705956e-01 -7.27996290e-01 -2.64060080e-01
2.57779211e-01 3.50548744e-01 1.92615673e-01 -4.64158267e-01
8.06389511e-01 -3.66214812e-02 -3.50069702e-01 -5.81407428e-01
6.09915674e-01 -6.00602031e-01 -4.53519434e-01 -1.55391749e-02
-1.06503260e+00 1.21620735e-02 -4.68945354e-02 6.28373563e-01
-4.17940348e-01 -3.41435730e-01 2.74459749e-01 -1.02529809e-01
-5.49742281e-01 -7.89297000e-02 -6.69587374e-01 2.98467755e-01
2.68794864e-01 6.79334328e-02 -3.79588842e-01 -8.95899892e-01
-8.12815130e-01 4.77235049e-01 3.03267300e-01 6.21781230e-01
2.88681120e-01 -9.57027197e-01 -5.33074319e-01 -2.06374362e-01
3.88472736e-01 -3.50140929e-01 3.51319313e-01 4.63723660e-01
3.90362442e-02 9.68385220e-01 3.19264412e-01 -4.05458242e-01
-1.22537541e+00 5.32942474e-01 2.28853390e-01 -5.24961114e-01
-7.82463610e-01 7.31352508e-01 1.74537584e-01 -3.69658798e-01
4.45179254e-01 -2.89293200e-01 -4.06862944e-01 3.57545286e-01
5.62878489e-01 1.32893920e-01 1.79149792e-01 -2.43103191e-01
-4.10995752e-01 9.83967185e-02 -2.73476720e-01 -5.16754866e-01
8.47381592e-01 -7.75494754e-01 -2.60318786e-01 5.44872165e-01
1.21882033e+00 3.14559996e-01 -8.10488701e-01 -6.66908979e-01
4.85500097e-01 -3.62334609e-01 -1.28192782e-01 -7.73529947e-01
-1.23278126e-01 5.95002413e-01 1.89242885e-01 5.28294563e-01
9.54492748e-01 2.03204527e-01 9.75422859e-01 9.88676906e-01
2.18511939e-01 -1.15073419e+00 7.81442970e-03 6.55288994e-01
8.34119916e-01 -9.19831753e-01 -1.12740621e-01 -2.67482430e-01
-2.98970491e-01 7.97626376e-01 3.32242459e-01 6.99314773e-02
4.65335697e-01 -3.48118424e-01 2.30557829e-01 -3.50531042e-01
-1.23059535e+00 -2.74639815e-01 2.94876397e-01 6.93287477e-02
4.90326732e-01 -2.54833996e-01 -6.47642314e-01 4.21822369e-01
-2.65394151e-01 -1.99259996e-01 1.85976818e-01 1.05086827e+00
-5.39343238e-01 -1.13054943e+00 -1.00423701e-01 5.77283092e-02
-4.50614125e-01 -5.03511667e-01 -5.82899630e-01 5.83151639e-01
-5.04373789e-01 1.41239405e+00 5.90262413e-02 -3.15988660e-02
4.50578749e-01 7.84150898e-01 5.94104938e-02 -7.19854176e-01
-1.03633344e+00 4.00806665e-02 5.65900445e-01 -4.56946284e-01
-3.87049675e-01 -6.22157514e-01 -8.49377275e-01 -4.80031446e-02
-4.19238538e-01 1.09955716e+00 4.64752913e-01 1.19503450e+00
6.49469554e-01 2.19606742e-01 6.61031842e-01 -3.21138114e-01
-5.55785477e-01 -1.17070937e+00 -7.68937916e-02 4.87052113e-01
5.79276800e-01 -3.67313713e-01 -5.90604722e-01 -2.56635696e-01] | [12.013164520263672, 7.810952186584473] |
2c62eaa0-43ca-4f1f-ba96-4a501f1ef4ad | scaling-vision-transformers-to-22-billion | 2302.05442 | null | https://arxiv.org/abs/2302.05442v1 | https://arxiv.org/pdf/2302.05442v1.pdf | Scaling Vision Transformers to 22 Billion Parameters | The scaling of Transformers has driven breakthrough capabilities for language models. At present, the largest large language models (LLMs) contain upwards of 100B parameters. Vision Transformers (ViT) have introduced the same architecture to image and video modelling, but these have not yet been successfully scaled to nearly the same degree; the largest dense ViT contains 4B parameters (Chen et al., 2022). We present a recipe for highly efficient and stable training of a 22B-parameter ViT (ViT-22B) and perform a wide variety of experiments on the resulting model. When evaluated on downstream tasks (often with a lightweight linear model on frozen features), ViT-22B demonstrates increasing performance with scale. We further observe other interesting benefits of scale, including an improved tradeoff between fairness and performance, state-of-the-art alignment to human visual perception in terms of shape/texture bias, and improved robustness. ViT-22B demonstrates the potential for "LLM-like" scaling in vision, and provides key steps towards getting there. | ['Neil Houlsby', 'Jeremiah Harmsen', 'Daniel Keysers', 'Xiaohua Zhai', 'Mario Lučić', 'Thomas Kipf', 'Dustin Tran', 'Filip Pavetić', 'Alexander Kolesnikov', 'Thomas Mensink', 'Yi Tay', 'Cristina Vasconcelos', 'Vighnesh Birodkar', 'Alexey Gritsenko', 'Mark Patrick Collier', 'Jasmijn Bastings', 'Fantine Huot', 'Avital Oliver', 'Fisher Yu', 'Aravindh Mahendran', 'Gamaleldin F. Elsayed', 'Sjoerd van Steenkiste', 'Manoj Kumar', 'Utku Evci', 'Joan Puigcerver', 'Matthias Minderer', 'Carlos Riquelme', 'Xiao Wang', 'Anurag Arnab', 'Michael Tschannen', 'Lucas Beyer', 'Rodolphe Jenatton', 'Ibrahim Alabdulmohsin', 'Robert Geirhos', 'Mathilde Caron', 'Andreas Steiner', 'Justin Gilmer', 'Jonathan Heek', 'Piotr Padlewski', 'Basil Mustafa', 'Josip Djolonga', 'Mostafa Dehghani'] | 2023-02-10 | null | null | null | null | ['zero-shot-transfer-image-classification', 'action-classification'] | ['computer-vision', 'computer-vision'] | [-2.48390622e-03 1.26773044e-01 6.96615726e-02 -4.88701344e-01
-7.74124384e-01 -5.13424695e-01 1.08913219e+00 -4.12952125e-01
-5.83380461e-01 1.53470069e-01 3.48255396e-01 -4.24844205e-01
3.59452307e-01 -1.17809027e-01 -6.79346800e-01 -4.67740506e-01
8.90553221e-02 4.71540630e-01 3.95008564e-01 -2.78608829e-01
1.81518227e-01 2.88198113e-01 -1.44670117e+00 5.53984761e-01
2.81650513e-01 1.04177380e+00 3.68396342e-01 9.01558220e-01
3.86725552e-02 9.55402851e-01 -2.54416049e-01 -7.35450923e-01
4.64802623e-01 -1.29455552e-01 -7.00999379e-01 1.66296244e-01
1.10233915e+00 -3.96561027e-01 -6.71932578e-01 7.45647430e-01
5.42107999e-01 -1.45223901e-01 4.55378532e-01 -1.24688315e+00
-1.01074636e+00 3.98326099e-01 -7.85677075e-01 3.35719228e-01
-7.56452307e-02 6.88167214e-01 9.34522629e-01 -9.29931581e-01
4.90471900e-01 1.65351892e+00 8.43737543e-01 6.88206732e-01
-1.37928021e+00 -5.14809966e-01 2.00495690e-01 1.76473036e-01
-1.29374218e+00 -9.42840874e-01 -6.91507608e-02 -4.59517598e-01
1.34954393e+00 2.07404658e-01 9.06570077e-01 1.19395673e+00
2.01647624e-01 7.46141791e-01 1.41807795e+00 -4.27524269e-01
-9.90425870e-02 3.17373604e-01 -1.66645825e-01 8.79269838e-01
-4.00237665e-02 1.19722135e-01 -8.81769359e-01 4.53083925e-02
1.01335776e+00 -5.74660957e-01 -1.52655512e-01 -3.53735834e-01
-1.28045630e+00 6.07683837e-01 3.67608935e-01 -4.89168316e-02
2.02905253e-01 6.48129165e-01 4.67889607e-01 3.95829022e-01
4.28294867e-01 3.29769731e-01 -3.66055638e-01 -1.48031369e-01
-1.05356240e+00 2.21516386e-01 4.81144637e-01 1.14107490e+00
4.68464762e-01 1.08750045e-01 -2.13789999e-01 9.19884920e-01
5.09685457e-01 7.32589006e-01 5.87356925e-01 -1.49196792e+00
1.46912217e-01 4.30181101e-02 7.03301206e-02 -7.04279780e-01
-4.83267277e-01 -2.95529515e-01 -6.85059786e-01 5.04076362e-01
6.05570972e-01 3.33191216e-01 -1.23997676e+00 1.91786015e+00
-3.62047181e-02 1.29032940e-01 -2.52487779e-01 8.31313133e-01
7.10569859e-01 6.59227133e-01 2.61112124e-01 2.04966411e-01
1.45228028e+00 -1.12288773e+00 -1.47100791e-01 -4.66619790e-01
3.28563362e-01 -1.08950365e+00 1.38641262e+00 5.02206802e-01
-1.39369547e+00 -6.49297595e-01 -7.42398620e-01 -5.17180026e-01
-6.94280490e-02 5.89112528e-02 8.45051110e-01 8.04298937e-01
-1.91606057e+00 2.07924396e-01 -8.06438327e-01 -8.45164895e-01
5.00921428e-01 3.80437642e-01 -3.68002295e-01 -6.28619865e-02
-7.10329652e-01 1.24558699e+00 -2.65382499e-01 -3.39821950e-02
-1.16743553e+00 -7.82890379e-01 -7.37795532e-01 -8.46485198e-02
2.31457889e-01 -1.09403133e+00 1.39322186e+00 -1.02992713e+00
-1.40343201e+00 1.45943689e+00 -4.07890975e-01 -6.90781832e-01
6.38574541e-01 -6.24544211e-02 -1.48075715e-01 2.05557439e-02
-1.67224944e-01 1.40884066e+00 1.15379107e+00 -1.18863738e+00
-4.72451806e-01 -2.40937486e-01 1.30299300e-01 3.80989045e-01
-4.26657170e-01 4.10770446e-01 -9.50467408e-01 -5.22860289e-01
-2.27286920e-01 -1.01644027e+00 -7.93343782e-02 6.02752626e-01
1.07670546e-01 -7.50810280e-02 3.68867248e-01 -6.66307688e-01
8.21467698e-01 -2.02433968e+00 -3.04402746e-02 -1.29403368e-01
7.49766052e-01 3.95941436e-01 -6.29789829e-01 1.71845645e-01
1.80363119e-01 1.97790086e-01 8.96136388e-02 -8.93717587e-01
8.00381899e-02 1.87566593e-01 -3.78769785e-01 4.84263569e-01
1.10718176e-01 1.26231539e+00 -5.59481502e-01 -5.42645395e-01
3.80339533e-01 5.68816721e-01 -6.29356503e-01 -4.89039719e-02
-1.17807701e-01 1.42296031e-01 8.09657127e-02 4.33477670e-01
5.96129656e-01 -5.44374645e-01 8.58278349e-02 -3.08178097e-01
-1.79818511e-01 2.62507871e-02 -6.49851859e-01 1.60437810e+00
-3.74308854e-01 1.22413909e+00 2.92175621e-01 -7.30885088e-01
5.80841422e-01 9.92107466e-02 1.58559367e-01 -9.17498708e-01
9.09079164e-02 -1.55701507e-02 6.84172288e-02 -4.60054964e-01
6.87869430e-01 -1.85574532e-01 2.95690387e-01 2.10441977e-01
2.16390476e-01 -5.75459003e-01 6.85432181e-02 5.14371634e-01
8.24555039e-01 2.97268122e-01 6.90936521e-02 -4.00828362e-01
2.16715381e-01 -1.71996966e-01 2.15542674e-01 1.00148344e+00
-5.13524055e-01 8.89129519e-01 3.72137398e-01 -4.67470348e-01
-1.65331793e+00 -9.33054149e-01 -3.59737314e-02 1.23819196e+00
-8.74038860e-02 -6.16709411e-01 -9.04703557e-01 1.34183094e-03
8.73760581e-02 3.82030785e-01 -7.28319347e-01 -7.64841959e-02
-2.99505591e-01 -7.66275585e-01 9.67189014e-01 5.59569478e-01
3.99279326e-01 -7.43481934e-01 -5.83502352e-01 -1.96095511e-01
-7.78397769e-02 -1.32706535e+00 -5.39701104e-01 -1.55395404e-01
-5.32525241e-01 -6.33795500e-01 -8.39417875e-01 -7.39848375e-01
3.76307100e-01 6.23012900e-01 1.32187092e+00 2.32960045e-01
-3.51453215e-01 5.95487297e-01 1.31920680e-01 -4.87846911e-01
-5.22018671e-01 -3.61512005e-01 1.74821466e-01 -1.37952507e-01
1.56561583e-01 -1.68750569e-01 -6.59270644e-01 3.60318452e-01
-5.83897114e-01 4.17841554e-01 4.91684437e-01 7.77351677e-01
2.90639073e-01 -7.14960039e-01 1.10749923e-01 -5.25354028e-01
4.66528386e-01 1.59823135e-01 -6.54916048e-01 3.96083593e-01
-7.49406278e-01 -4.63443296e-03 4.62428272e-01 -4.59286720e-01
-1.06379473e+00 -2.76687384e-01 6.66203126e-02 -5.64510047e-01
1.70393139e-01 9.06060264e-02 2.86843717e-01 -5.17259061e-01
6.70848966e-01 2.08040699e-01 2.19689548e-01 -3.32696110e-01
7.93616354e-01 4.92393881e-01 5.70515096e-01 -5.58632851e-01
5.44667006e-01 6.21022344e-01 -1.17982872e-01 -9.68855798e-01
-6.26203537e-01 -2.12619945e-01 -3.09644282e-01 -7.90903121e-02
7.44816720e-01 -1.27662909e+00 -8.63510549e-01 8.39704514e-01
-9.39948142e-01 -8.27403724e-01 -2.41573766e-01 2.49770910e-01
-7.53675342e-01 5.51461041e-01 -7.18534946e-01 -7.69809544e-01
-9.65961367e-02 -1.23511565e+00 1.10767508e+00 1.53400615e-01
-1.33427188e-01 -8.39290738e-01 -5.54028675e-02 7.14537740e-01
7.75247872e-01 -4.50995922e-01 8.98723185e-01 -5.99800237e-02
-5.40154517e-01 3.35149914e-01 -7.83823073e-01 3.83374482e-01
-3.52512270e-01 1.56929657e-01 -1.27759480e+00 -5.51140964e-01
-2.20353618e-01 -7.56763160e-01 1.05479467e+00 6.62917137e-01
9.55419838e-01 2.26747617e-02 -2.31812641e-01 9.79154289e-01
1.21588480e+00 -1.88193187e-01 7.19988465e-01 3.87713909e-01
6.72880590e-01 3.49417090e-01 1.47609115e-01 2.12490708e-01
7.44024694e-01 6.66063488e-01 1.75137773e-01 -2.84342527e-01
-7.17561603e-01 -1.46447480e-01 4.56699431e-01 7.42823124e-01
-1.88199893e-01 -1.93108842e-01 -9.99468446e-01 3.79308701e-01
-1.71289039e+00 -8.33863735e-01 3.10660619e-02 2.18659830e+00
7.55363047e-01 2.76963323e-01 9.76616442e-02 -6.62419736e-01
3.51863027e-01 1.95985451e-01 -7.53083229e-01 -5.07808208e-01
-4.06895101e-01 9.88361314e-02 5.36302567e-01 7.68253148e-01
-8.82728338e-01 1.39322257e+00 8.32456112e+00 8.33521366e-01
-1.22500026e+00 1.28567547e-01 1.11927867e+00 -4.68477845e-01
-3.35047483e-01 1.74520984e-01 -7.87155390e-01 1.93646207e-01
1.06627619e+00 -2.18390413e-02 8.71381640e-01 5.20739555e-01
4.02198911e-01 -2.34384745e-01 -9.49981689e-01 1.28134358e+00
3.72946382e-01 -1.47817147e+00 3.35397452e-01 1.85759634e-01
6.84355676e-01 7.08911002e-01 4.46468592e-01 2.98032254e-01
3.23922873e-01 -1.32798493e+00 1.13731432e+00 6.00094557e-01
1.19862628e+00 -2.89833546e-01 1.53616682e-01 9.88253206e-02
-1.00346446e+00 -1.18880011e-01 -9.03186917e-01 1.51691595e-02
1.39697045e-01 3.19180377e-02 -5.36867678e-01 -7.37408847e-02
9.24697518e-01 6.28347933e-01 -1.09606731e+00 8.49600077e-01
1.98491350e-01 6.23087943e-01 -4.36886489e-01 1.71091363e-01
3.01848590e-01 -6.97767958e-02 2.42430314e-01 1.39827704e+00
-8.90441239e-03 -8.76929387e-02 -7.73059055e-02 5.94783902e-01
-1.23716608e-01 -8.26657936e-02 -4.28916067e-01 -8.44617411e-02
3.98491204e-01 1.33256018e+00 -6.20331824e-01 -6.09917700e-01
-7.18210757e-01 8.99395168e-01 5.47907591e-01 3.30515414e-01
-9.40474749e-01 3.65345508e-01 8.26073527e-01 1.65357932e-01
2.87802458e-01 -4.14125592e-01 -3.73220265e-01 -1.31470716e+00
-2.26556182e-01 -1.29740679e+00 -1.27320308e-02 -1.37908661e+00
-1.28619385e+00 5.93123198e-01 -6.06439784e-02 -7.48215854e-01
7.47544318e-02 -1.01753855e+00 -1.88668802e-01 1.00545943e+00
-1.44352329e+00 -1.66658497e+00 -3.14850986e-01 6.89061821e-01
5.84728658e-01 -2.91135907e-01 8.68826330e-01 6.10268861e-02
-3.22498798e-01 8.89274359e-01 1.18635044e-01 -2.24678412e-01
1.12542951e+00 -1.14390862e+00 1.08959126e+00 7.90013015e-01
3.48908901e-01 8.38806033e-01 7.43180096e-01 -2.10827932e-01
-1.38674998e+00 -6.21212840e-01 7.09251583e-01 -9.91778255e-01
8.24572086e-01 -5.38206875e-01 -4.73512530e-01 1.01973796e+00
5.09495139e-01 5.03789335e-02 2.81232297e-01 1.23332217e-01
-7.64813840e-01 -1.10727824e-01 -8.82631123e-01 8.65883470e-01
1.22179747e+00 -7.49475479e-01 -3.78692687e-01 1.34036675e-01
7.28464782e-01 -3.98616552e-01 -5.76994836e-01 8.09966773e-02
9.65331316e-01 -1.27017879e+00 1.08534610e+00 -6.35842681e-01
2.58723319e-01 -1.65591851e-01 -2.71376818e-01 -1.04376018e+00
-7.01514781e-01 -9.62129951e-01 8.19838885e-03 1.00114644e+00
2.31820196e-01 -5.33587217e-01 6.32425606e-01 8.34028125e-01
-8.48416463e-02 -7.70904958e-01 -8.51115346e-01 -5.81979692e-01
1.99103624e-01 -6.22619748e-01 2.15893760e-01 5.93641639e-01
-2.38747969e-01 2.84794092e-01 -6.27754033e-01 -2.45444551e-01
6.59213185e-01 -3.82390231e-01 9.19810236e-01 -7.55175531e-01
-4.93803799e-01 -7.82131910e-01 -5.71553469e-01 -1.46120071e+00
-2.66147882e-01 -7.28198528e-01 -1.73564553e-01 -1.38190126e+00
6.24779344e-01 -3.98005664e-01 -2.02143326e-01 5.42943656e-01
-6.27530664e-02 7.28981137e-01 6.72813177e-01 3.83751661e-01
-5.88465273e-01 2.43673474e-01 1.27225924e+00 -8.35419893e-02
3.50339323e-01 -5.30486226e-01 -1.05060148e+00 8.08133245e-01
4.66586560e-01 5.58917522e-02 -5.94516039e-01 -9.10647810e-01
2.45102286e-01 -4.33297127e-01 5.74602425e-01 -8.58356476e-01
1.27220392e-01 -1.22576796e-01 5.26314557e-01 -7.70516247e-02
7.42293537e-01 -3.39747250e-01 1.17682072e-03 2.27869332e-01
-4.47753400e-01 2.20436737e-01 4.20158565e-01 3.33816528e-01
2.58603841e-01 1.28555670e-01 1.04267180e+00 -3.04609776e-01
-1.06905341e+00 2.54258335e-01 -3.97313476e-01 2.04530835e-01
6.62574291e-01 -2.90578693e-01 -5.98744154e-01 -5.83310664e-01
-6.20291293e-01 2.03202724e-01 7.66328037e-01 5.10442197e-01
5.01162767e-01 -9.28660810e-01 -8.10057461e-01 2.02070311e-01
1.63002193e-01 -3.84519637e-01 1.17140181e-01 9.77783799e-01
-5.03292799e-01 4.25405592e-01 -2.34153539e-01 -7.91558325e-01
-1.36151290e+00 4.98086423e-01 5.05220234e-01 -5.42270616e-02
-6.60920024e-01 1.42429090e+00 6.55278146e-01 -9.82266441e-02
1.68718547e-01 -4.06730771e-01 2.57824600e-01 -8.45922008e-02
6.82107985e-01 1.45463824e-01 -1.73103571e-01 -6.68041945e-01
-3.77855152e-01 7.16401994e-01 -2.30526999e-01 -4.85807210e-01
1.24768639e+00 -5.29209971e-01 -1.16227344e-01 3.92075628e-01
7.42214441e-01 -1.87141314e-01 -1.84663713e+00 -2.24330097e-01
-4.39343274e-01 -4.98842686e-01 7.35367909e-02 -9.61455166e-01
-9.61367726e-01 9.83917475e-01 7.25774348e-01 2.95985285e-02
8.52868438e-01 1.03032306e-01 5.82566142e-01 2.72842973e-01
5.34253836e-01 -9.92121577e-01 9.98152643e-02 5.41191638e-01
9.04812753e-01 -1.28568149e+00 1.40136909e-02 5.11013390e-03
-9.60467994e-01 8.69387031e-01 5.93920350e-01 3.02710570e-02
4.16393667e-01 4.75296855e-01 4.48950827e-01 -8.03168714e-02
-1.09714115e+00 -1.00766145e-01 2.27027893e-01 8.17916512e-01
5.39845884e-01 -6.43915683e-02 8.60875174e-02 -1.13261782e-01
-3.87356102e-01 -2.13811975e-02 4.51309979e-01 4.49954361e-01
-5.75885773e-01 -8.67412865e-01 -3.38435709e-01 3.82615596e-01
-8.10546279e-02 -5.56382239e-01 -2.64768869e-01 7.17523217e-01
2.72121504e-02 9.53070104e-01 6.60373494e-02 -3.01562101e-01
1.08897485e-01 3.13594230e-02 1.06694365e+00 -3.92419517e-01
-5.36563694e-01 3.22974622e-01 3.74056287e-02 -9.14449513e-01
-3.51896554e-01 -6.14258647e-01 -8.89664650e-01 -8.12914789e-01
-6.04858808e-02 -4.85614479e-01 6.59315348e-01 7.97432959e-01
4.72731262e-01 1.54389501e-01 1.90397561e-01 -1.12733006e+00
-7.29701221e-01 -9.23087835e-01 -4.31823552e-01 3.42836797e-01
3.62851232e-01 -4.14074749e-01 -1.43395692e-01 3.90694201e-01] | [9.978906631469727, 1.4301166534423828] |
9f3e63fb-cc75-43d2-99ac-99b575be06c2 | self-supervised-audio-teacher-student | 2306.04186 | null | https://arxiv.org/abs/2306.04186v1 | https://arxiv.org/pdf/2306.04186v1.pdf | Self-supervised Audio Teacher-Student Transformer for Both Clip-level and Frame-level Tasks | In recent years, self-supervised learning (SSL) has emerged as a popular approach for learning audio representations. The ultimate goal of audio self-supervised pre-training is to transfer knowledge to downstream audio tasks, generally including clip-level and frame-level tasks. Clip-level tasks classify the scene or sound of an entire audio clip, e.g. audio tagging, instrument recognition, etc. While frame-level tasks detect event-level timestamps from an audio clip, e.g. sound event detection, speaker diarization, etc. Prior studies primarily evaluate on clip-level downstream tasks. Frame-level tasks are important for fine-grained acoustic scene/event understanding, and are generally more challenging than clip-level tasks. In order to tackle both clip-level and frame-level tasks, this paper proposes two self-supervised audio representation learning methods: ATST-Clip and ATST-Frame, responsible for learning clip-level and frame-level representations, respectively. ATST stands for Audio Teacher-Student Transformer, which means both methods use a transformer encoder and a teacher-student training scheme.Experimental results show that our ATST-Frame model obtains state-of-the-art (SOTA) performance on most of the clip-level and frame-level downstream tasks. Especially, it outperforms other models by a large margin on the frame-level sound event detection task. In addition, the performance can be further improved by combining the two models through knowledge distillation. | ['Xiaofei Li', 'Nian Shao', 'Xian Li'] | 2023-06-07 | null | null | null | null | ['audio-tagging', 'sound-event-detection', 'instrument-recognition', 'speaker-diarization'] | ['audio', 'audio', 'audio', 'speech'] | [ 4.35707927e-01 -3.33720148e-01 -2.04258099e-01 -4.55258042e-01
-1.66340411e+00 -4.16852325e-01 3.87360930e-01 5.62621653e-01
-1.84079319e-01 4.78268683e-01 3.25793296e-01 5.59474342e-02
-1.08198933e-02 -5.11566281e-01 -7.58253932e-01 -6.70502007e-01
-2.31661424e-01 -6.69643059e-02 5.09255588e-01 1.34550408e-01
-4.91098501e-02 1.43542727e-02 -1.95664787e+00 7.52633274e-01
4.78606880e-01 1.43455350e+00 3.21349949e-01 9.66324329e-01
-1.97112903e-01 1.17749619e+00 -6.82019472e-01 1.20567560e-01
-2.12104753e-01 -6.54098511e-01 -6.74158633e-01 -1.09159805e-01
4.43762422e-01 -7.63300285e-02 -1.53483152e-01 8.18581998e-01
6.91039979e-01 2.75095373e-01 4.16689962e-01 -1.35536194e+00
1.04612313e-01 1.02927852e+00 -3.39563221e-01 4.78782445e-01
5.38347602e-01 -1.75577119e-01 1.12993741e+00 -1.12390041e+00
-1.15182415e-01 1.28395545e+00 7.80543685e-01 7.69252554e-02
-8.81140292e-01 -1.18125272e+00 2.35471040e-01 6.37975454e-01
-1.41678119e+00 -6.28038406e-01 9.94250536e-01 -5.07551432e-01
6.85483873e-01 5.83275259e-02 4.69298631e-01 8.46836269e-01
5.16354805e-03 1.13991237e+00 9.92595017e-01 -5.88783383e-01
3.64971846e-01 -3.00058663e-01 -7.38204494e-02 2.52271324e-01
-5.89845657e-01 8.27956796e-02 -1.21924603e+00 -3.96738239e-02
7.12589145e-01 -1.37118191e-01 -1.91002235e-01 1.91212952e-01
-1.18166041e+00 5.37276745e-01 1.27534047e-01 3.43832165e-01
-2.46138051e-01 1.72112003e-01 8.78159046e-01 3.63256216e-01
7.19738603e-01 -6.52573407e-02 -5.25531471e-01 -5.02362132e-01
-1.31252956e+00 6.32046089e-02 4.94766176e-01 9.12538111e-01
6.13798976e-01 6.22587502e-01 -3.41522008e-01 1.01586258e+00
1.28841877e-01 3.06938887e-01 6.81864917e-01 -7.48924494e-01
4.94969487e-01 3.57041396e-02 -2.56170273e-01 -3.86835486e-01
-2.18766227e-01 -5.46271920e-01 -8.17347288e-01 -1.69084921e-01
6.33183494e-02 -1.97157040e-01 -7.58772254e-01 1.76041710e+00
3.40151012e-01 1.05303359e+00 -1.12186768e-03 6.77820921e-01
1.08802903e+00 1.20843637e+00 3.37487370e-01 -5.75758874e-01
1.64059114e+00 -9.47750509e-01 -9.18258786e-01 -6.13246001e-02
3.08888018e-01 -9.64463592e-01 8.18001807e-01 5.55590749e-01
-1.20843697e+00 -1.24335134e+00 -8.39523017e-01 8.52875710e-02
-6.98888972e-02 2.97606736e-01 2.76252747e-01 3.53023201e-01
-7.09360123e-01 2.83546984e-01 -8.55985284e-01 4.00485732e-02
2.79989660e-01 -2.73576695e-02 -3.59002538e-02 2.64439851e-01
-1.42703211e+00 1.26924232e-01 4.01231915e-01 -2.42048770e-01
-1.48331332e+00 -1.23011565e+00 -1.03658116e+00 3.66883248e-01
3.80725890e-01 -1.39187604e-01 1.94910264e+00 -7.35392630e-01
-1.64257109e+00 6.05689764e-01 -3.96532416e-01 -8.40555012e-01
2.17021964e-02 -5.07209778e-01 -8.35859001e-01 3.41631502e-01
2.27354407e-01 4.22246903e-01 1.35800815e+00 -9.19092298e-01
-1.31690347e+00 1.58436783e-02 -1.19990006e-01 1.72750384e-01
-4.59637761e-01 2.99234986e-01 -4.61884528e-01 -1.06443453e+00
-7.61869177e-02 -5.94608068e-01 2.95776755e-01 -2.58290559e-01
-2.15976238e-01 -5.46914279e-01 1.20402157e+00 -5.30850172e-01
1.48387742e+00 -2.69561362e+00 -2.69748956e-01 -3.50139588e-01
-1.32496551e-01 3.45446259e-01 -1.50815576e-01 4.05349076e-01
-4.73066837e-01 -3.34089726e-01 3.00240424e-02 -4.62793797e-01
-5.24647981e-02 -1.36619300e-01 -7.81641960e-01 -6.87961057e-02
2.69620299e-01 3.75618041e-01 -1.10633898e+00 -7.52581954e-01
2.61094570e-01 5.17148495e-01 -4.38953370e-01 5.25313675e-01
-1.95168644e-01 5.56621194e-01 -2.45624065e-01 5.57241559e-01
1.88799262e-01 1.99525744e-01 -1.62286952e-01 -3.85890543e-01
-3.32939059e-01 7.81359136e-01 -1.35692585e+00 1.85076845e+00
-8.62246871e-01 8.34704161e-01 1.62459433e-01 -1.10650837e+00
8.68913233e-01 9.27529037e-01 5.96411884e-01 -5.19321322e-01
-1.57670319e-01 7.65722692e-02 -2.05791324e-01 -4.64019328e-01
3.45821470e-01 -3.75243604e-01 -3.30961168e-01 2.23455191e-01
4.17140692e-01 -2.39567757e-01 1.52525902e-01 1.89666227e-02
1.03322899e+00 1.22660495e-01 3.27343494e-01 8.52240548e-02
8.19032311e-01 -3.95581752e-01 7.47115076e-01 5.51723182e-01
-1.20585896e-01 6.30180359e-01 9.82713178e-02 2.04710979e-02
-3.95299166e-01 -1.20921314e+00 -4.85421233e-02 1.91323519e+00
-1.52302384e-01 -8.55806231e-01 -7.92061448e-01 -3.86990786e-01
-1.37719914e-01 6.56895101e-01 -1.06109247e-01 -3.41284722e-01
-7.18488514e-01 -9.15655717e-02 8.12526405e-01 7.67117321e-01
4.49224323e-01 -1.16025078e+00 -4.43382621e-01 6.42922044e-01
-3.47990602e-01 -1.28962994e+00 -6.53452158e-01 5.43333173e-01
-8.03929567e-01 -7.63169050e-01 -6.25297189e-01 -1.06272697e+00
-2.28357725e-02 3.34427834e-01 9.94959712e-01 -5.23763835e-01
-2.21631199e-01 4.47124958e-01 -6.67490661e-01 -7.57346332e-01
-2.77384222e-01 -1.68454424e-02 2.35563233e-01 4.86886710e-01
7.27324486e-02 -8.03873122e-01 -4.83763248e-01 4.12291259e-01
-9.10359919e-01 -2.56232202e-01 3.25320750e-01 6.42897546e-01
9.03681159e-01 4.96536314e-01 9.81131971e-01 -5.48904121e-01
4.51423407e-01 -2.96462029e-01 -2.92704999e-01 2.75736529e-04
6.03607222e-02 -4.59355861e-01 8.87768388e-01 -6.99721992e-01
-1.17782867e+00 8.97697657e-02 -3.15987825e-01 -7.79354095e-01
-4.23347026e-01 5.32097518e-01 -2.89446175e-01 4.12228376e-01
4.31903273e-01 3.65456372e-01 -4.73843306e-01 -7.06195712e-01
1.46231964e-01 9.08390701e-01 9.25892830e-01 -7.23478556e-01
7.75199234e-01 2.26700827e-01 -4.33691144e-01 -1.07211626e+00
-1.45765901e+00 -6.52602851e-01 -4.06549484e-01 -4.08126891e-01
8.03880453e-01 -1.45526028e+00 -4.38971937e-01 5.47711849e-01
-8.58972788e-01 -3.19103956e-01 -6.90652013e-01 7.15799868e-01
-6.59086645e-01 -4.54073921e-02 -6.04178488e-01 -1.06848669e+00
-1.92751020e-01 -9.18072343e-01 1.49155951e+00 2.37018064e-01
-2.06490040e-01 -7.91797221e-01 2.89110988e-02 2.66790003e-01
2.37365648e-01 -7.11109936e-02 7.46205926e-01 -6.00373089e-01
-2.36979216e-01 -7.23043978e-02 1.86996341e-01 5.31242669e-01
3.07892293e-01 -2.53952712e-01 -1.47828257e+00 -3.00382286e-01
1.11554675e-01 -5.93239069e-01 8.56230915e-01 5.25725842e-01
1.71077037e+00 -6.25366494e-02 -8.46955478e-02 4.70818907e-01
8.47841740e-01 5.23548722e-01 4.24210191e-01 -4.74569947e-02
6.01925492e-01 2.97674239e-01 1.03051150e+00 5.70437491e-01
3.42654347e-01 9.22743261e-01 9.68771353e-02 -1.27062812e-01
-6.12814188e-01 -6.09053791e-01 7.14358747e-01 1.27372408e+00
2.01075554e-01 -1.69715937e-02 -6.99494421e-01 5.87647855e-01
-1.75393379e+00 -1.05816746e+00 4.26827036e-02 2.19033742e+00
1.17866242e+00 2.65468061e-01 3.73470515e-01 8.63685429e-01
8.91084731e-01 3.08827877e-01 -5.06271064e-01 -5.70866689e-02
1.14227206e-01 5.17047644e-01 -1.77930072e-01 1.95017472e-01
-1.45085871e+00 9.28874075e-01 5.39562988e+00 1.31259000e+00
-1.14696431e+00 4.17265326e-01 2.19847336e-01 -1.51617348e-01
2.95676917e-01 -8.98222998e-02 -1.01540780e+00 6.13927722e-01
1.30817747e+00 -1.64670482e-01 -6.00570217e-02 7.30548739e-01
5.14550507e-01 3.85294631e-02 -1.45373929e+00 1.26270795e+00
-1.84987264e-03 -1.13657975e+00 -6.93807378e-02 -5.21172822e-01
5.33365786e-01 -3.29466164e-01 2.06102245e-02 8.63944471e-01
-1.68388933e-01 -5.75146735e-01 1.01695383e+00 2.77997106e-01
1.04232073e+00 -9.09800410e-01 4.45949584e-01 1.99943051e-01
-2.13193321e+00 -1.39109150e-01 4.77008102e-03 -4.19945754e-02
4.44364846e-01 8.17146659e-01 -6.50321066e-01 5.62209606e-01
1.03368866e+00 9.04717147e-01 -1.70982882e-01 1.26154816e+00
-3.63999754e-01 1.50633752e+00 -1.32072538e-01 4.63645577e-01
6.16864860e-02 3.58863920e-01 5.00862181e-01 1.60169101e+00
3.46970588e-01 1.03099398e-01 5.62981367e-01 2.60413527e-01
-1.06216781e-01 -2.97110435e-02 -1.71154425e-01 6.44032583e-02
7.42330492e-01 1.09699070e+00 -4.48294461e-01 -3.91126245e-01
-2.09119409e-01 6.03558302e-01 -1.35482594e-01 2.50833869e-01
-1.00262356e+00 -6.47648811e-01 8.39927852e-01 9.26089287e-02
6.12448037e-01 4.03181203e-02 2.54272968e-01 -8.99508238e-01
-8.78526494e-02 -8.08787942e-01 6.51383758e-01 -8.38075817e-01
-1.22088647e+00 4.21641469e-01 2.43405756e-02 -1.76601195e+00
-5.41457295e-01 -1.63559586e-01 -8.96691203e-01 3.37764502e-01
-1.80140734e+00 -9.93711412e-01 -1.11714728e-01 9.42531586e-01
1.19229329e+00 -2.08892494e-01 8.18833768e-01 5.86819053e-01
-3.95080149e-01 6.93965256e-01 -7.58092031e-02 3.02649349e-01
9.08266664e-01 -1.16954339e+00 -6.20093234e-02 6.16554320e-01
6.08095527e-01 8.18760917e-02 4.34221953e-01 -3.51213008e-01
-1.24399447e+00 -1.47535276e+00 8.82318318e-01 -2.13272367e-02
7.74625063e-01 -5.00784457e-01 -1.00963700e+00 4.27682072e-01
-2.01612078e-02 2.26490691e-01 9.12697792e-01 1.17702782e-01
-4.36164856e-01 -6.45569265e-01 -6.13676190e-01 -1.85713619e-02
7.49962747e-01 -1.20310605e+00 -8.37295830e-01 3.39490563e-01
9.04330909e-01 -4.95539933e-01 -9.62085009e-01 5.27727842e-01
3.06826144e-01 -8.71150613e-01 1.04402006e+00 -3.33766073e-01
3.23581934e-01 -5.77409029e-01 -8.92311707e-02 -1.22371435e+00
-1.69481158e-01 -7.93311357e-01 -5.14644623e-01 1.89708996e+00
5.01307398e-02 -1.36677939e-02 5.09715438e-01 -4.78019774e-01
-4.59045857e-01 -3.78436536e-01 -1.10700977e+00 -1.05134046e+00
-3.17080170e-01 -1.07985413e+00 4.99975652e-01 8.01017225e-01
-3.83809512e-03 5.67074180e-01 -4.48849261e-01 4.15564656e-01
5.99529862e-01 2.39515841e-01 5.92304468e-01 -1.36495447e+00
-3.33426386e-01 -2.82275587e-01 -4.05932814e-01 -1.10344839e+00
3.21870357e-01 -7.56462634e-01 4.20748085e-01 -1.11438990e+00
-1.53465360e-01 -1.20797776e-01 -6.92033231e-01 5.26414275e-01
-9.91227180e-02 3.52046341e-01 2.00114101e-01 1.65336534e-01
-8.83721352e-01 5.90927482e-01 8.43353271e-01 -2.46108487e-01
-3.92884105e-01 4.11091745e-01 -2.77801871e-01 9.67532396e-01
7.22450733e-01 -5.27475774e-01 -6.35865629e-01 -2.04881161e-01
-2.36987188e-01 4.78712231e-01 3.37060928e-01 -1.47139490e+00
4.86696333e-01 1.47586823e-01 1.73764080e-01 -9.97930884e-01
5.28998375e-01 -7.06884563e-01 -2.13084981e-01 1.95571274e-01
-7.29930460e-01 -3.53394240e-01 3.74088317e-01 7.27031112e-01
-9.43213165e-01 -8.54625553e-02 8.38844061e-01 1.37315482e-01
-7.26138175e-01 1.58521563e-01 -6.26336396e-01 3.18974465e-01
8.40367913e-01 -1.71130467e-02 1.36115104e-01 -6.16201341e-01
-8.33279490e-01 5.79669923e-02 -5.58866203e-01 6.51892960e-01
6.67700887e-01 -1.49547815e+00 -7.75715232e-01 2.67026305e-01
3.02161455e-01 2.36861616e-01 3.83853644e-01 6.88588440e-01
2.61460274e-01 3.79904717e-01 1.25806600e-01 -9.15260553e-01
-1.34059596e+00 1.06659949e-01 -3.12055945e-02 -1.22339875e-01
-5.33091843e-01 1.04973888e+00 4.99173641e-01 3.86864580e-02
8.58666122e-01 -6.11723244e-01 -2.65603304e-01 3.57862562e-01
9.13212180e-01 3.66045088e-01 1.68427110e-01 -6.03704691e-01
-2.38972083e-01 5.48438549e-01 5.46986647e-02 -1.19272754e-01
1.24522567e+00 3.37635502e-02 2.10052669e-01 1.05503559e+00
1.14343715e+00 -1.24486350e-01 -1.34724796e+00 -5.00628710e-01
-9.63868424e-02 -8.47888887e-02 4.13042456e-01 -5.50892591e-01
-9.49489415e-01 1.24919868e+00 7.07707584e-01 3.94770294e-01
1.44455636e+00 8.34675580e-02 1.02059913e+00 7.89092034e-02
4.66266155e-01 -1.11621320e+00 3.53640318e-01 7.32453167e-01
7.29179859e-01 -8.50241840e-01 -2.25294054e-01 -4.06387448e-01
-3.26608181e-01 1.04880702e+00 5.98368943e-01 1.37924850e-01
8.07411253e-01 5.61608493e-01 3.30579393e-02 2.17332035e-01
-7.99341381e-01 -3.72175664e-01 4.61017579e-01 6.27671957e-01
7.01034248e-01 -5.08516654e-02 3.08049262e-01 8.57751012e-01
-3.71757776e-01 3.25304419e-02 3.49475332e-02 9.94416356e-01
-6.77241385e-01 -1.08433127e+00 -6.12478197e-01 2.36219272e-01
-4.48819369e-01 -9.48773995e-02 -2.19177902e-02 2.88329691e-01
2.08627388e-01 1.25273466e+00 2.29574963e-01 -4.26228642e-01
4.27287340e-01 3.71273428e-01 1.93339273e-01 -9.19647515e-01
-7.68343747e-01 6.76033080e-01 -4.97387582e-03 -4.19923365e-01
-6.47566438e-01 -6.86254859e-01 -1.40345860e+00 4.35558736e-01
-3.96837533e-01 5.49297869e-01 3.94431859e-01 9.40144777e-01
2.01558262e-01 1.09541047e+00 9.17229950e-01 -1.10220289e+00
-2.16517031e-01 -9.89318430e-01 -7.14278400e-01 1.64943591e-01
5.62587738e-01 -5.78805506e-01 -3.31554800e-01 5.79200089e-01] | [15.213295936584473, 5.1764607429504395] |
4252bf79-5eb6-42ac-9a00-c64939968b63 | gym-gazebo2-a-toolkit-for-reinforcement | 1903.06278 | null | http://arxiv.org/abs/1903.06278v2 | http://arxiv.org/pdf/1903.06278v2.pdf | gym-gazebo2, a toolkit for reinforcement learning using ROS 2 and Gazebo | This paper presents an upgraded, real world application oriented version of
gym-gazebo, the Robot Operating System (ROS) and Gazebo based Reinforcement
Learning (RL) toolkit, which complies with OpenAI Gym. The content discusses
the new ROS 2 based software architecture and summarizes the results obtained
using Proximal Policy Optimization (PPO). Ultimately, the output of this work
presents a benchmarking system for robotics that allows different techniques
and algorithms to be compared using the same virtual conditions. We have
evaluated environments with different levels of complexity of the Modular
Articulated Robotic Arm (MARA), reaching accuracies in the millimeter scale.
The converged results show the feasibility and usefulness of the gym-gazebo 2
toolkit, its potential and applicability in industrial use cases, using modular
robots. | ['Víctor Mayoral Vilches', 'Lander Usategui San Juan', 'Elias Barba Moral', 'Nestor Gonzalez Lopez', 'Alejandro Solano Rueda', 'Yue Leire Erro Nuin', 'Risto Kojcev'] | 2019-03-14 | null | null | null | null | ['transfer-reinforcement-learning'] | ['methodology'] | [-6.09777093e-01 4.58678305e-01 -9.79375020e-02 9.48432162e-02
-1.28399491e-01 -5.44044554e-01 3.42166334e-01 -3.44454914e-01
-4.15853679e-01 9.61165905e-01 -2.01170132e-01 -2.35744402e-01
-8.33601892e-01 -2.92859375e-01 -5.66413939e-01 -8.54475260e-01
-6.45803571e-01 3.97752881e-01 3.46572250e-01 -1.13545024e+00
4.12256926e-01 7.45820880e-01 -2.01714182e+00 -4.63981777e-01
4.01469409e-01 5.36236346e-01 9.37210083e-01 1.16617119e+00
7.01484799e-01 9.34327543e-01 -7.34158933e-01 6.92805946e-01
7.06669450e-01 1.50842428e-01 -7.83378720e-01 -2.25259200e-01
2.45048646e-02 4.75159436e-02 -2.78825879e-01 4.51366603e-01
8.32889497e-01 5.34690797e-01 3.22977044e-02 -1.71129310e+00
6.89722002e-02 6.04652703e-01 -1.61547344e-02 4.40029874e-02
6.43325686e-01 5.59787214e-01 4.66328859e-01 -1.64974540e-01
5.53916812e-01 1.56048727e+00 5.35771728e-01 5.19919038e-01
-6.91568494e-01 -2.99044877e-01 -7.98352137e-02 1.27592787e-01
-1.01657557e+00 -4.63100374e-02 3.87984544e-01 -5.55243939e-02
1.26460636e+00 1.05249934e-01 7.79192150e-01 1.15670049e+00
8.83672476e-01 3.93314481e-01 1.33207893e+00 -5.38170099e-01
6.23688936e-01 -2.34022036e-01 -1.13531888e-01 6.83814883e-01
1.88282758e-01 4.63787705e-01 -3.11021477e-01 2.66613197e-02
7.14262009e-01 -4.00392562e-01 3.07916433e-01 -8.03138375e-01
-1.42407930e+00 5.29127717e-01 6.10536933e-01 -9.19091143e-03
-7.17393219e-01 8.28342378e-01 2.69565493e-01 5.26395738e-01
-6.54461205e-01 7.51470625e-01 -8.90668988e-01 -4.53421921e-01
2.54299551e-01 8.00135136e-01 1.12164843e+00 1.30565381e+00
3.20977271e-01 3.53396297e-01 2.66435236e-01 6.94542170e-01
9.13312256e-01 2.89998710e-01 6.65699005e-01 -1.92557526e+00
1.10803492e-01 2.47880995e-01 6.21089339e-01 -3.58402312e-01
-1.09380615e+00 1.22451976e-01 3.02796185e-01 1.35986066e+00
3.96877289e-01 -6.35507405e-01 -2.23175958e-01 1.13554585e+00
6.87091351e-01 -4.80823457e-01 6.78550065e-01 9.87375557e-01
4.59605157e-01 4.74096864e-01 -3.31346989e-02 2.54068583e-01
1.35081339e+00 -1.08570850e+00 -6.80874407e-01 -1.47955894e-01
7.09515572e-01 -6.74979210e-01 1.30037975e+00 9.03237224e-01
-9.98009980e-01 -4.89831150e-01 -1.62642670e+00 2.83954412e-01
-5.34034491e-01 1.06941976e-01 6.45302117e-01 7.13119209e-01
-1.16422045e+00 8.35532010e-01 -1.08408439e+00 -9.30281699e-01
-4.20459956e-01 8.45177948e-01 -9.03242305e-02 3.08855772e-01
-5.91125906e-01 1.39129639e+00 6.47402585e-01 -1.68932211e-02
-1.08791447e+00 -2.68802971e-01 -6.59016907e-01 -3.05027962e-01
4.33105022e-01 -5.25772095e-01 1.74695134e+00 -1.29397482e-01
-2.89124393e+00 2.70686299e-01 8.17653716e-01 -5.35317779e-01
3.53158385e-01 -6.31556988e-01 7.15484396e-02 5.16902916e-02
-2.90946960e-02 6.48290813e-01 3.32980931e-01 -1.12181473e+00
-8.82002890e-01 -2.45607972e-01 5.20431638e-01 4.44246590e-01
5.09766996e-01 -5.42892218e-02 5.25487781e-01 -1.07790671e-01
-1.14257172e-01 -1.50401556e+00 -4.61711675e-01 -3.50544184e-01
2.00236514e-01 -5.77123702e-01 8.84345949e-01 -9.96353254e-02
1.54093906e-01 -1.88099802e+00 2.58552432e-01 -8.56797099e-02
-5.63376069e-01 -2.37821862e-01 1.56799778e-01 1.01012266e+00
-3.53044644e-02 -4.13385212e-01 4.56896544e-01 3.87182653e-01
2.00197399e-01 6.22341216e-01 -1.16861209e-01 4.41650480e-01
-2.29763746e-01 4.40598309e-01 -1.34170747e+00 -1.28282577e-01
6.92907631e-01 2.18881741e-02 -3.04004073e-01 5.70905022e-02
-2.06106946e-01 7.16476321e-01 -6.17742419e-01 6.56666458e-01
1.18563855e-02 5.05857289e-01 3.06827754e-01 3.40347648e-01
-7.78246403e-01 9.59348083e-02 -1.46845269e+00 2.26828885e+00
-6.58539236e-01 4.98731613e-01 3.21364313e-01 -7.46165395e-01
1.12412369e+00 2.25137398e-01 9.46250200e-01 -4.03807014e-01
4.57272649e-01 2.82040119e-01 -2.50701718e-02 -9.69308078e-01
6.99037790e-01 5.58121979e-01 -1.06985189e-01 -1.59199566e-01
3.68952930e-01 -6.65721953e-01 3.41746807e-01 -1.79339200e-01
1.38599050e+00 1.35834908e+00 3.98017377e-01 -7.96880424e-01
5.55741549e-01 6.81415200e-01 4.45102826e-02 5.98312259e-01
-7.89541423e-01 -1.63664836e-02 2.65548766e-01 -2.33707234e-01
-9.40208077e-01 -1.12999785e+00 1.24131650e-01 1.32902181e+00
5.92589915e-01 -2.98744589e-01 -5.89245141e-01 -2.43839756e-01
2.92921774e-02 7.42610097e-01 -2.37851843e-01 1.32875502e-01
-7.88720012e-01 -3.62034410e-01 4.22025621e-01 5.85023202e-02
2.59277552e-01 -1.32558978e+00 -1.50446582e+00 4.69606608e-01
3.27538818e-01 -1.17292011e+00 6.66237116e-01 2.61255413e-01
-1.06164670e+00 -1.48138011e+00 -2.65276968e-01 -8.49470973e-01
-1.09371163e-01 3.12734932e-01 6.46677017e-01 -2.97137648e-01
-4.97296721e-01 9.97705579e-01 -7.57897139e-01 -7.77852893e-01
-4.00041312e-01 4.13143113e-02 3.87825936e-01 -1.10760808e+00
-3.40091348e-01 -6.52190149e-01 -5.94919324e-01 5.94930410e-01
-3.57388169e-01 -4.23404723e-01 5.96085608e-01 5.05859196e-01
7.63493255e-02 -3.78961116e-01 6.37562335e-01 1.71693832e-01
6.92865014e-01 -3.38261962e-01 -1.10987735e+00 -2.12057367e-01
-5.78394532e-01 7.68324733e-02 2.10902289e-01 -3.88054937e-01
-8.79206896e-01 5.98737001e-01 -1.60403922e-02 -2.44138129e-02
-2.60251880e-01 6.65627867e-02 3.75973076e-01 -3.66784543e-01
1.22807026e+00 -3.92736137e-01 5.08159935e-01 -1.78626448e-01
4.70011175e-01 7.94456840e-01 4.32638347e-01 -7.50770152e-01
2.38766968e-01 3.21619570e-01 4.17112440e-01 -8.47724259e-01
5.32042584e-04 -4.74764526e-01 -3.02514642e-01 -7.81601071e-01
6.68810546e-01 -9.01491106e-01 -1.56222546e+00 1.73367888e-01
-6.87876165e-01 -1.06664038e+00 -5.22124767e-01 8.37835789e-01
-1.47298062e+00 1.03474082e-02 -4.74298000e-01 -9.99566376e-01
-9.21896994e-02 -1.42750394e+00 9.63241875e-01 6.54656708e-01
-4.15670462e-02 -3.86829019e-01 4.24039096e-01 1.71251655e-01
2.60573298e-01 6.04885221e-01 1.48870751e-01 -2.04667956e-01
-3.88353616e-01 6.82975957e-03 6.20473444e-01 -6.01418018e-02
-1.08259015e-01 5.01968078e-02 -6.69957340e-01 -4.57485616e-01
-3.34787548e-01 -6.68958902e-01 -5.11853099e-01 2.84486771e-01
3.63948554e-01 2.75139194e-02 -1.55388996e-01 -2.04732016e-01
1.42183685e+00 2.29079023e-01 6.69131339e-01 1.32955599e+00
1.72741711e-01 7.06037521e-01 1.56525409e+00 5.26607215e-01
2.78187871e-01 7.53868937e-01 1.31930971e+00 8.24673235e-01
8.54619667e-02 1.46751508e-01 9.22380209e-01 5.59799314e-01
-4.22826707e-01 2.97157019e-01 -7.18442261e-01 3.41292083e-01
-2.26636600e+00 -5.31640232e-01 -4.67527628e-01 2.08710217e+00
2.67163038e-01 -2.13536337e-01 4.80852783e-01 6.01024739e-02
4.73634094e-01 -4.74408507e-01 -3.92958671e-01 -8.14522803e-01
4.87288386e-01 1.92947522e-01 9.63038266e-01 4.91604030e-01
-7.80364096e-01 1.07529819e+00 7.15725946e+00 2.05186784e-01
-1.00732672e+00 1.62908942e-01 -6.70618832e-01 -7.24327937e-02
1.10296261e+00 1.44336984e-01 -6.92597151e-01 3.21156561e-01
1.35891759e+00 -5.11228815e-02 9.54998672e-01 1.65643311e+00
6.85693860e-01 -4.35696661e-01 -7.65044272e-01 5.64067304e-01
-4.42753851e-01 -9.18319285e-01 -8.97036135e-01 2.72482336e-01
4.79202062e-01 8.43145847e-01 -3.96242231e-01 8.53936851e-01
1.07737660e+00 -5.32910526e-01 1.17591619e+00 3.24525774e-01
-4.72094752e-02 -7.29753792e-01 6.52043819e-01 4.48675156e-01
-8.70275974e-01 -8.70155573e-01 -5.74346125e-01 -4.71252739e-01
-1.60901338e-01 -4.06053305e-01 -1.04099512e+00 9.22790825e-01
1.17869675e+00 3.37384611e-01 -3.47494364e-01 1.25408065e+00
-3.82353365e-01 5.97701408e-02 -5.19903183e-01 -6.75800323e-01
3.27852964e-01 -2.74121791e-01 8.37042749e-01 5.85418701e-01
2.84209698e-01 -2.11831957e-01 3.25425029e-01 2.65475988e-01
7.55433679e-01 9.97616202e-02 -7.87711561e-01 5.62065482e-01
1.76620185e-01 1.59589946e+00 -5.24981022e-01 2.58029383e-02
1.48830727e-01 4.65696961e-01 1.73012957e-01 1.02878548e-01
-7.50952065e-01 -7.05091655e-01 8.08918595e-01 -4.20565307e-01
7.94291571e-02 -8.78210247e-01 -1.61446005e-01 -3.70007932e-01
-2.39037022e-01 -7.34532535e-01 3.21430117e-01 -1.49287760e+00
-5.59196055e-01 5.05812615e-02 3.90236109e-01 -1.62142348e+00
-5.24793684e-01 -1.23343647e+00 -2.09001869e-01 3.36169392e-01
-1.32015181e+00 -7.83261895e-01 -6.16981268e-01 2.48851970e-01
9.93920326e-01 -1.37191638e-01 8.66658568e-01 -3.07442695e-01
-2.81726003e-01 -1.51017398e-01 3.62874925e-01 -7.41794288e-01
6.20601356e-01 -1.42847621e+00 -3.58664721e-01 1.38874024e-01
-6.73223674e-01 3.24346483e-01 1.17984045e+00 -2.94304818e-01
-2.02397513e+00 -6.28681481e-01 -2.94766963e-01 -5.53375721e-01
8.43777001e-01 1.61063135e-01 -1.16012521e-01 7.95743465e-01
6.31778538e-01 -2.61974722e-01 -8.05791989e-02 -2.84759045e-01
5.34821033e-01 -2.70622194e-01 -1.37778258e+00 9.40696061e-01
7.59559512e-01 5.01359820e-01 -8.16938281e-01 4.87544596e-01
7.74741709e-01 -8.84826183e-01 -1.24962759e+00 5.05189478e-01
8.80408883e-01 -7.87966609e-01 9.52460110e-01 -3.59102249e-01
9.53188539e-02 -6.60096228e-01 -7.01998174e-02 -1.50883436e+00
-4.86380868e-02 -1.19167018e+00 -2.85966605e-01 6.06521130e-01
-5.41360900e-02 -8.35731626e-01 5.39835930e-01 3.47520635e-02
-7.14273930e-01 -3.75879854e-01 -1.07496631e+00 -1.09091270e+00
-7.96264336e-02 -1.36426955e-01 3.97643059e-01 6.34131968e-01
6.64817810e-01 2.43399858e-01 9.56564471e-02 7.12828994e-01
1.71918333e-01 -6.25449002e-01 1.35662973e+00 -1.11603844e+00
-3.06735724e-01 -1.72216088e-01 -8.53358030e-01 -4.77352113e-01
-1.33553952e-01 -3.70176464e-01 5.45330048e-01 -1.57209861e+00
-9.20106351e-01 -5.85514784e-01 -9.69379991e-02 5.22610188e-01
4.68104601e-01 -1.55440226e-01 3.61174405e-01 2.91656375e-01
-4.21382785e-01 4.38730061e-01 9.70413208e-01 1.66573599e-01
-6.60392880e-01 -2.34355703e-02 -8.31838027e-02 7.91350305e-01
1.43585408e+00 -4.73974615e-01 -3.99125665e-01 -5.31161278e-02
2.60411710e-01 -2.57258285e-02 4.88852113e-01 -1.68603277e+00
1.14762709e-01 -3.60620707e-01 -2.31708214e-01 -3.05296630e-01
3.91304225e-01 -1.10228562e+00 -9.99876931e-02 1.20692742e+00
-1.28061026e-01 4.28232193e-01 3.35645616e-01 6.05272114e-01
6.18969023e-01 -5.17937720e-01 6.94788873e-01 -1.92204461e-01
-8.94685984e-01 -6.34285986e-01 -9.90814507e-01 -6.25195026e-01
1.74255168e+00 -2.09207147e-01 -3.37272108e-01 4.12819386e-02
-7.56922841e-01 4.25696284e-01 4.66087580e-01 7.17546046e-01
2.36362927e-02 -8.80943060e-01 -2.81607509e-01 -4.59925503e-01
2.50413001e-01 -1.17508650e-01 -2.67709970e-01 7.61212945e-01
-1.35419226e+00 2.26145118e-01 -9.24181521e-01 -8.16495657e-01
-1.15409434e+00 3.99562091e-01 4.34125394e-01 -6.91006426e-03
-9.02736664e-01 -8.68649855e-02 -8.10279846e-01 -1.10127413e+00
2.94413656e-01 -5.46308398e-01 -5.61268389e-01 -9.96618330e-01
1.25169873e-01 8.76835465e-01 9.77181867e-02 -1.32554635e-01
-1.12150773e-01 2.33283013e-01 6.06294930e-01 -5.08439660e-01
1.36621952e+00 -1.90663666e-01 2.20980048e-01 7.79103637e-01
9.82373953e-03 -1.86047092e-01 -1.45257366e+00 6.51629329e-01
3.54107708e-01 -9.97131243e-02 -2.03499749e-01 -7.29114175e-01
-3.27751994e-01 2.37151906e-01 1.32096779e+00 2.39747837e-01
3.84409398e-01 -2.68245041e-01 -7.06524923e-02 1.22767961e+00
1.05299973e+00 -1.42288005e+00 2.07070678e-01 6.80398583e-01
1.15514541e+00 -8.87103260e-01 3.06933612e-01 -3.77393484e-01
-5.52940130e-01 1.36269283e+00 7.64603555e-01 -9.11314130e-01
3.22586000e-01 1.26835868e-01 3.34071249e-01 -2.18365327e-01
-4.63159144e-01 -2.08450705e-01 -7.05544889e-01 1.19953001e+00
5.18439934e-02 -2.98602097e-02 -6.56602740e-01 3.04964799e-02
-5.79484284e-01 2.49960404e-02 9.89705145e-01 1.97215533e+00
-6.98330998e-01 -1.01082563e+00 -8.48731458e-01 -5.24877071e-01
-3.53853583e-01 8.60389173e-01 -1.21763952e-01 1.41232741e+00
-3.18361014e-01 1.20428097e+00 -4.63882059e-01 -1.69239715e-01
6.42905056e-01 -1.48111179e-01 8.01066220e-01 -3.11688393e-01
-8.84154081e-01 -4.11676645e-01 3.08706969e-01 -1.09558201e+00
-4.17844027e-01 -4.92446959e-01 -1.64538491e+00 -1.61489248e-02
-2.16033965e-01 2.89427452e-02 1.74842453e+00 6.47329986e-01
5.08482337e-01 9.55251396e-01 7.27437675e-01 -1.54198802e+00
-8.50524485e-01 -1.21674252e+00 -5.64203739e-01 -3.66698593e-01
2.64843944e-02 -9.60115969e-01 -2.87524406e-02 -9.67168882e-02] | [4.3522748947143555, 1.2371909618377686] |
cb5e8403-bcce-4212-a846-763c62d29b82 | 190910094 | 1909.10094 | null | https://arxiv.org/abs/1909.10094v2 | https://arxiv.org/pdf/1909.10094v2.pdf | Deep Structured Neural Network for Event Temporal Relation Extraction | We propose a novel deep structured learning framework for event temporal relation extraction. The model consists of 1) a recurrent neural network (RNN) to learn scoring functions for pair-wise relations, and 2) a structured support vector machine (SSVM) to make joint predictions. The neural network automatically learns representations that account for long-term contexts to provide robust features for the structured model, while the SSVM incorporates domain knowledge such as transitive closure of temporal relations as constraints to make better globally consistent decisions. By jointly training the two components, our model combines the benefits of both data-driven learning and knowledge exploitation. Experimental results on three high-quality event temporal relation datasets (TCR, MATRES, and TB-Dense) demonstrate that incorporated with pre-trained contextualized embeddings, the proposed model achieves significantly better performances than the state-of-the-art methods on all three datasets. We also provide thorough ablation studies to investigate our model. | ['Ralph Weischedel', 'I-Hung Hsu', 'Nanyun Peng', 'Aram Galstyan', 'Rujun Han', 'Mu Yang'] | 2019-09-22 | deep-structured-neural-network-for-event | https://aclanthology.org/K19-1062 | https://aclanthology.org/K19-1062.pdf | conll-2019-11 | ['temporal-relation-extraction'] | ['natural-language-processing'] | [ 2.71839261e-01 4.29287553e-01 -5.56609631e-01 -7.20296621e-01
-9.35225129e-01 -2.03788225e-02 8.14939320e-01 4.90975618e-01
-3.92445177e-01 7.19739914e-01 5.73314786e-01 -1.86218500e-01
-3.86518538e-01 -8.07007790e-01 -5.66037059e-01 -3.39400113e-01
-5.24217725e-01 4.79814559e-01 4.57443774e-01 -2.62849152e-01
-1.79198250e-01 3.81513387e-01 -1.15539634e+00 5.13144851e-01
6.49086654e-01 1.40137374e+00 -6.05851635e-02 3.14217448e-01
2.53627867e-01 1.73471749e+00 -1.75539210e-01 -3.93612444e-01
-1.22611679e-01 -1.76345259e-01 -1.01380265e+00 -3.24020982e-01
-5.54539919e-01 -8.80588964e-02 -7.75654018e-01 4.43760574e-01
3.15274447e-01 5.09366095e-01 6.31028473e-01 -9.92278636e-01
-6.73613608e-01 8.13632846e-01 -3.15549701e-01 6.22939289e-01
2.88371444e-01 -2.53497243e-01 1.49915099e+00 -8.99586082e-01
7.80417681e-01 1.03866708e+00 5.85999072e-01 5.68879023e-02
-1.20490181e+00 -5.66837013e-01 3.53602529e-01 7.18159616e-01
-1.38882864e+00 -5.15257001e-01 7.88423717e-01 -1.51437193e-01
1.89705670e+00 9.81802121e-02 5.33389866e-01 1.32040274e+00
3.21214199e-01 6.23923123e-01 5.63630521e-01 -3.62385988e-01
3.09685767e-01 -2.06180930e-01 2.71287709e-01 4.70796496e-01
-6.32081255e-02 3.73014957e-01 -7.83050537e-01 -3.04093987e-01
5.51629126e-01 -1.75978504e-02 5.21085644e-03 -1.94223568e-01
-1.09070420e+00 6.11335456e-01 6.50740743e-01 3.86337519e-01
-7.12696731e-01 1.55360132e-01 6.44258440e-01 1.20941445e-01
7.06929445e-01 3.63749802e-01 -7.57413149e-01 -1.46802455e-01
-6.77474082e-01 1.30223319e-01 6.28229678e-01 7.17077076e-01
5.33118904e-01 -2.20278114e-01 -4.26955462e-01 9.04498219e-01
3.73260409e-01 -9.19581652e-02 3.82817954e-01 -6.63308799e-01
8.31947267e-01 9.73173916e-01 -1.27706051e-01 -1.08137286e+00
-6.30847275e-01 -3.78929108e-01 -7.00222194e-01 -4.65366960e-01
-1.00347340e-01 1.48262735e-02 -6.78700149e-01 1.87833190e+00
2.78369576e-01 6.36183858e-01 1.93032697e-01 7.82318294e-01
8.42697442e-01 7.33449996e-01 2.42284507e-01 -4.28291976e-01
1.24244547e+00 -1.02606475e+00 -1.04293704e+00 -4.27448392e-01
7.54696429e-01 -3.30807388e-01 4.91934597e-01 6.13813894e-03
-1.04820979e+00 -3.15194905e-01 -1.27356970e+00 -3.35727483e-01
-3.46358359e-01 6.45910054e-02 8.39312494e-01 -1.24352768e-01
-6.95110857e-01 7.27435768e-01 -1.29586375e+00 -3.43395889e-01
5.60569227e-01 3.26207012e-01 -3.51294339e-01 2.43386075e-01
-1.81410527e+00 1.17817533e+00 5.40463030e-01 3.44120681e-01
-9.08442020e-01 -5.15702188e-01 -1.12988925e+00 3.21273148e-01
5.34162343e-01 -6.05230570e-01 1.15275562e+00 -2.59044319e-01
-1.33423555e+00 8.19628596e-01 -5.87930918e-01 -8.33948374e-01
6.80608749e-02 -4.81510431e-01 -6.92291260e-01 2.83360302e-01
5.07200286e-02 2.59726457e-02 4.94248718e-01 -8.76756549e-01
-5.12958109e-01 -1.48350880e-01 -1.86978474e-01 1.00185513e-01
-3.29500109e-01 4.52453703e-01 -4.99412209e-01 -7.25540698e-01
4.86860313e-02 -6.56763375e-01 -4.72641438e-01 -4.37363148e-01
-6.52815998e-01 -6.13544762e-01 5.70173383e-01 -6.37067437e-01
1.72401702e+00 -2.06643820e+00 3.40092957e-01 2.10796624e-01
1.33689836e-01 2.00292051e-01 -5.90731576e-02 6.72047794e-01
-6.26837075e-01 -1.26828298e-01 -1.47143200e-01 -6.04826331e-01
-2.55781233e-01 4.40871000e-01 -5.47150016e-01 2.58021176e-01
7.83058703e-01 1.26178074e+00 -1.12931466e+00 -5.13213575e-01
7.54224556e-03 4.87065375e-01 -1.86572507e-01 5.04103601e-01
-2.21904770e-01 1.82012260e-01 -6.84580564e-01 3.84097606e-01
8.69528726e-02 -6.36434317e-01 7.45829642e-01 -1.93484142e-01
2.51871526e-01 1.07781553e+00 -7.88063288e-01 1.63911986e+00
-3.80006462e-01 3.46383601e-01 -5.62282264e-01 -1.38069475e+00
1.05475402e+00 6.42362595e-01 4.86499906e-01 -8.92501533e-01
1.26061603e-01 -5.77504523e-02 -2.25861624e-01 -5.77257454e-01
3.18391860e-01 -5.60546890e-02 -2.41425142e-01 5.23747802e-01
2.84234285e-01 3.69160384e-01 -1.87742282e-02 4.45770651e-01
1.48728287e+00 2.92997032e-01 6.55593812e-01 9.78169665e-02
3.69438171e-01 -3.31783831e-01 1.14932263e+00 4.35180902e-01
-1.10197328e-01 4.17350501e-01 7.21680641e-01 -7.87117898e-01
-5.70467710e-01 -1.13573325e+00 3.66264656e-02 9.77876127e-01
7.04473779e-02 -8.72066081e-01 -2.01750104e-03 -1.05422282e+00
-2.02487186e-01 7.47622371e-01 -9.05044079e-01 -5.06078541e-01
-8.77071798e-01 -7.74214983e-01 4.26187217e-01 9.80346978e-01
3.21883082e-01 -1.34034431e+00 -4.35766369e-01 4.05937046e-01
-3.79711151e-01 -1.33554018e+00 -2.01332256e-01 6.04242086e-01
-8.38917375e-01 -1.08811247e+00 1.97536990e-01 -6.94008112e-01
2.76178718e-01 -2.24544048e-01 1.25602484e+00 -1.71812758e-01
-1.26526535e-01 -2.11683080e-01 -6.50689662e-01 -1.50722891e-01
-4.68573160e-02 6.91402256e-02 6.30887151e-02 1.21626943e-01
4.93871391e-01 -9.62993741e-01 -4.03576285e-01 2.51034915e-01
-6.15023971e-01 -4.64542843e-02 4.20977056e-01 1.02544069e+00
7.47152328e-01 1.52785599e-01 8.40554953e-01 -9.77978528e-01
5.21591783e-01 -7.20662832e-01 -2.21914649e-01 4.24382746e-01
-7.07288802e-01 1.88083202e-01 3.58010292e-01 -3.81405473e-01
-1.30145407e+00 -1.22486018e-01 4.40141931e-02 -3.12881380e-01
5.41991480e-02 9.96573925e-01 3.61295510e-03 6.23439610e-01
7.38404691e-01 5.97206019e-02 -4.90159690e-01 -4.22227144e-01
4.45228815e-01 3.38139623e-01 6.03433251e-01 -5.92648625e-01
6.61304533e-01 4.85620648e-01 -6.44548312e-02 -3.31969470e-01
-1.42864859e+00 -2.78408200e-01 -8.16636562e-01 1.63465276e-01
9.83245850e-01 -1.09800720e+00 -3.76035035e-01 1.49517104e-01
-1.28395629e+00 -4.85840082e-01 -5.12201786e-01 6.29676282e-01
-5.18059552e-01 1.29011618e-02 -9.68417823e-01 -8.98598909e-01
-3.66662443e-01 -6.71151817e-01 9.80747283e-01 1.33252472e-01
-6.22827470e-01 -1.07948995e+00 2.89479673e-01 3.07702929e-01
8.57693031e-02 5.26118755e-01 1.08565307e+00 -1.11149442e+00
-5.61131060e-01 -1.47614285e-01 -3.46687973e-01 1.47235459e-02
2.45327696e-01 -2.93862760e-01 -9.89517868e-01 1.03407510e-01
-1.79050326e-01 -6.44772470e-01 1.10681534e+00 1.94348425e-01
1.01508403e+00 -1.67065039e-01 -6.61666572e-01 6.12936616e-01
9.23656404e-01 3.07375401e-01 7.83489347e-01 2.79865801e-01
5.67221463e-01 5.05886734e-01 9.30238903e-01 5.28855324e-01
8.91930819e-01 7.80210614e-01 4.01266962e-01 3.86611074e-02
4.05530035e-02 -3.14802766e-01 2.72751898e-01 8.97489667e-01
-3.09789270e-01 -7.69477896e-03 -8.27739835e-01 7.93661654e-01
-2.40355158e+00 -1.08157945e+00 1.35946035e-01 1.73855364e+00
1.15339184e+00 4.16206926e-01 -7.35476166e-02 3.40498835e-01
3.27980310e-01 5.20415962e-01 -4.85455394e-01 -4.03149426e-01
-1.06545702e-01 4.70293224e-01 -2.04652280e-01 2.16601983e-01
-1.27632380e+00 1.07163990e+00 6.09492207e+00 7.08810031e-01
-7.64252901e-01 2.59577125e-01 5.05508006e-01 -2.27143422e-01
-1.80508241e-01 1.90441579e-01 -5.60419381e-01 1.35360271e-01
1.07508111e+00 4.36472632e-02 2.53805310e-01 4.96774465e-01
1.64736018e-01 1.75801381e-01 -1.37482297e+00 6.96557701e-01
-2.67242402e-01 -1.57344711e+00 -2.65037566e-01 -1.34756029e-01
5.98243952e-01 1.74916178e-01 -1.75036013e-01 5.92583835e-01
5.39006352e-01 -1.14069057e+00 5.12549520e-01 6.92515671e-01
5.51021934e-01 -7.20016003e-01 8.35931778e-01 1.27996847e-01
-1.58096778e+00 -9.35513526e-02 -6.92856312e-02 -2.78103590e-01
4.17171955e-01 8.77039075e-01 -7.60459840e-01 1.06686759e+00
6.08024836e-01 1.37216568e+00 -3.12443405e-01 5.81602335e-01
-8.22946668e-01 8.04572761e-01 -1.99155182e-01 1.94375977e-01
-2.41965503e-02 1.47786155e-01 4.39569324e-01 1.11947727e+00
-1.62643537e-01 3.49755615e-01 1.00145422e-01 9.05005455e-01
-7.03807846e-02 -1.40159994e-01 -4.02519077e-01 -2.05521360e-01
7.53255427e-01 1.18725371e+00 -4.21264797e-01 -1.64150238e-01
-4.71972734e-01 8.05098832e-01 1.05142152e+00 5.36665201e-01
-8.90422523e-01 -3.22029203e-01 7.17720091e-01 -2.42084384e-01
5.00943780e-01 -1.93460137e-01 -3.79086584e-01 -1.31899381e+00
6.44699708e-02 -4.02872533e-01 1.13755310e+00 -5.83407104e-01
-1.46295774e+00 8.93280983e-01 3.98177356e-02 -9.35226738e-01
-5.22577047e-01 -9.99213904e-02 -7.76574850e-01 9.10746038e-01
-1.60753047e+00 -1.29785109e+00 2.36993894e-01 6.06780529e-01
1.56195119e-01 -1.46120369e-01 1.04882467e+00 2.78867006e-01
-1.02939427e+00 5.32797694e-01 -5.20896137e-01 3.66635203e-01
6.04916036e-01 -1.07614541e+00 4.95355874e-01 7.75232673e-01
3.43054354e-01 7.42548585e-01 2.52529293e-01 -7.83686757e-01
-1.01198936e+00 -1.35714567e+00 1.53773141e+00 -5.07108569e-01
8.63728642e-01 -3.79272491e-01 -9.55100596e-01 1.15081179e+00
8.44536200e-02 4.71926451e-01 8.92641962e-01 7.34153092e-01
-6.64362490e-01 -1.81017801e-01 -8.17090809e-01 3.90508324e-01
1.18472803e+00 -8.39874089e-01 -1.07395148e+00 2.61774272e-01
1.05132556e+00 -3.89427632e-01 -1.18401444e+00 6.65334225e-01
2.86104053e-01 -7.10456550e-01 1.13290310e+00 -9.88430679e-01
6.19580388e-01 -1.09714411e-01 -2.06486851e-01 -1.10045445e+00
-3.91407788e-01 -6.65237486e-01 -9.74241853e-01 1.27114320e+00
7.73441672e-01 -6.56980395e-01 4.04687107e-01 6.11928761e-01
5.88918142e-02 -1.34881854e+00 -1.03108895e+00 -6.30976200e-01
-3.43707025e-01 -7.83399880e-01 5.12018859e-01 1.16225934e+00
3.78684103e-01 7.14976966e-01 -4.83516723e-01 4.61355150e-01
3.04693043e-01 4.07504916e-01 1.61819741e-01 -1.14772439e+00
-3.59808743e-01 -7.46894330e-02 -3.40862811e-01 -8.38155568e-01
4.49869990e-01 -1.05439985e+00 -2.26739496e-01 -1.54537928e+00
2.99994111e-01 -4.23624068e-01 -9.29669023e-01 9.06566262e-01
-4.45327312e-01 -1.55195460e-01 -2.38186762e-01 6.21134741e-03
-9.49579000e-01 9.07778263e-01 8.42208147e-01 8.80112201e-02
-4.03796911e-01 -8.88159499e-02 -7.03943193e-01 5.51702082e-01
5.33080935e-01 -6.99224830e-01 -5.08552313e-01 -2.83192635e-01
5.38592756e-01 2.92185903e-01 2.54154056e-01 -5.14185727e-01
3.64983290e-01 -1.71438664e-01 2.86247611e-01 -5.48972011e-01
4.84638572e-01 -5.57983577e-01 -1.71255127e-01 -2.59214249e-02
-7.19743609e-01 -2.82614261e-01 1.18534490e-01 7.82211959e-01
-3.80930603e-01 3.35497618e-01 4.04038757e-01 3.52448553e-01
-5.96205294e-01 3.66423905e-01 -8.02024007e-02 3.11566293e-01
9.52921033e-01 4.40078497e-01 -2.74618238e-01 -5.51025629e-01
-9.84012544e-01 5.79333007e-01 -2.87250191e-01 6.06895387e-01
8.25335383e-01 -1.63386643e+00 -6.01770103e-01 3.73526700e-02
1.67764887e-01 2.36471638e-01 8.80746767e-02 8.72797430e-01
1.39518633e-01 3.68375748e-01 3.43228281e-01 -2.39734247e-01
-9.04657364e-01 6.98623002e-01 2.04908147e-01 -1.13696301e+00
-7.33797908e-01 1.09788573e+00 -5.09029105e-02 -4.08887833e-01
2.04762906e-01 -6.32015824e-01 -2.70027012e-01 2.02554524e-01
5.95008314e-01 5.78473434e-02 1.35180101e-01 -4.54164773e-01
-8.26889038e-01 1.48902506e-01 -2.85266310e-01 -1.42367527e-01
1.80154324e+00 3.19826342e-02 -8.34772214e-02 5.67067206e-01
1.16451621e+00 -3.22973222e-01 -1.10413444e+00 -8.20601285e-01
4.69899476e-01 -3.53914574e-02 1.35331661e-01 -9.02825713e-01
-8.35605800e-01 6.57778382e-01 -1.71160251e-01 1.39274061e-01
1.26506650e+00 2.88187891e-01 9.19814348e-01 3.72561246e-01
2.99482852e-01 -9.31505382e-01 2.46793687e-01 7.44243264e-01
9.02412891e-01 -1.05067623e+00 1.11062318e-01 -5.61377466e-01
-8.30603838e-01 9.86725271e-01 4.52278286e-01 -2.37437800e-01
9.13583696e-01 4.30544555e-01 -1.40436992e-01 -3.68382573e-01
-1.44603896e+00 -3.40788990e-01 5.56217968e-01 4.36157197e-01
3.29273999e-01 -8.67384076e-02 -1.16874672e-01 1.32974112e+00
9.66814309e-02 9.85982195e-02 -1.37725160e-01 1.01959217e+00
1.27613112e-01 -1.24178481e+00 2.28716403e-01 4.29024011e-01
-3.70780736e-01 -1.40725091e-01 -3.09937239e-01 5.21043062e-01
-7.48632029e-02 1.18512094e+00 4.02425565e-02 -7.47194529e-01
5.22884786e-01 1.72183350e-01 2.84395874e-01 -7.67377734e-01
-6.99536264e-01 -1.28958479e-01 6.78040504e-01 -1.04086912e+00
-4.99371171e-01 -7.34996498e-01 -1.42997921e+00 2.01850653e-01
-2.68121332e-01 -9.41691548e-03 -9.01945084e-02 1.32838595e+00
5.07351160e-01 1.01076961e+00 7.16675043e-01 -6.22821271e-01
-2.51934588e-01 -1.02145696e+00 -4.75129098e-01 4.87490207e-01
2.21298054e-01 -8.60853314e-01 -1.29917324e-01 -1.20894030e-01] | [9.081171035766602, 9.084399223327637] |
f3b3491e-5540-4552-bfb8-95ab3fc526fe | variational-em-based-deep-learning-for-noise | null | null | http://openaccess.thecvf.com/content_CVPR_2020/html/Nan_Variational-EM-Based_Deep_Learning_for_Noise-Blind_Image_Deblurring_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Nan_Variational-EM-Based_Deep_Learning_for_Noise-Blind_Image_Deblurring_CVPR_2020_paper.pdf | Variational-EM-Based Deep Learning for Noise-Blind Image Deblurring | Non-blind deblurring is an important problem encountered in many image restoration tasks. The focus of non-blind deblurring is on how to suppress noise magnification during deblurring. In practice, it often happens that the noise level of input image is unknown and varies among different images. This paper aims at developing a deep learning framework for deblurring images with unknown noise level. Based on the framework of variational expectation maximization (EM), an iterative noise-blind deblurring scheme is proposed which integrates the estimation of noise level and the quantification of image prior uncertainty. Then, the proposed scheme is unrolled to a neural network (NN) where image prior is modeled by NN with uncertainty quantification. Extensive experiments showed that the proposed method not only outperformed existing noise-blind deblurring methods by a large margin, but also outperformed those state-of-the-art image deblurring methods designed/trained with known noise level.
| [' Hui Ji', ' Yuhui Quan', 'Yuesong Nan'] | 2020-06-01 | null | null | null | cvpr-2020-6 | ['blind-image-deblurring'] | ['computer-vision'] | [ 4.44920182e-01 -5.97331107e-01 2.51438022e-01 1.46564394e-01
-6.28458261e-01 -2.56516665e-01 5.65532684e-01 -6.54126883e-01
-3.26432347e-01 7.95170248e-01 6.23625696e-01 -2.94374138e-01
-1.01013280e-01 -3.48512650e-01 -5.70767939e-01 -1.18227804e+00
2.38061085e-01 8.31703003e-03 -1.80898264e-01 8.31163302e-02
4.26381767e-01 2.22111315e-01 -1.22580302e+00 -3.24618161e-01
9.12971258e-01 8.90953302e-01 5.14305890e-01 9.72558141e-01
8.74471143e-02 9.85214472e-01 -6.22061670e-01 1.50933012e-01
3.28842252e-01 -5.88191092e-01 -6.89381719e-01 3.78947735e-01
5.72643042e-01 -6.86759770e-01 -7.41701186e-01 1.68659937e+00
7.42222250e-01 4.06971306e-01 7.70861447e-01 -7.22779095e-01
-1.38654995e+00 4.75491643e-01 -8.92962098e-01 6.71870589e-01
-1.35098740e-01 -3.19712758e-02 3.33031297e-01 -8.70386243e-01
2.59015143e-01 1.06741118e+00 8.04148376e-01 5.84105134e-01
-9.80677664e-01 -5.95166206e-01 -2.12958544e-01 4.74338442e-01
-1.53302062e+00 -6.63572729e-01 7.92796135e-01 -5.68268299e-01
2.90516853e-01 1.00938521e-01 2.04500452e-01 8.83214176e-01
2.62751251e-01 7.66342103e-01 1.33595753e+00 -5.17466545e-01
2.72799075e-01 -4.43749219e-01 1.03283435e-01 2.12119699e-01
3.06610793e-01 4.09621090e-01 -3.30199242e-01 -5.42570464e-02
1.07962453e+00 -1.47775382e-01 -9.44117069e-01 -2.12703124e-01
-1.19518030e+00 4.83154058e-01 4.56620097e-01 4.91132289e-01
-5.16589046e-01 3.61087173e-01 1.87331736e-01 2.21365005e-01
6.72610164e-01 2.26190746e-01 -1.27490222e-01 3.47828865e-02
-1.52623391e+00 -1.11061573e-01 4.48184341e-01 4.22758788e-01
5.98062277e-01 3.60028595e-01 -4.31780607e-01 1.07154632e+00
4.46392775e-01 8.04745793e-01 6.85577035e-01 -1.05570495e+00
-3.72403078e-02 -2.20939979e-01 5.13555288e-01 -1.19492149e+00
-7.84437135e-02 -5.90964615e-01 -1.63044393e+00 4.00776565e-01
1.10808879e-01 -3.05516034e-01 -1.25220680e+00 1.53503692e+00
8.01707208e-02 7.66825736e-01 7.52211735e-02 1.45987988e+00
6.80631816e-01 7.10398793e-01 -2.93970436e-01 -5.68045974e-01
1.00177538e+00 -1.10357141e+00 -1.29323411e+00 -3.44663709e-01
-1.70917600e-01 -9.69336808e-01 5.15649796e-01 5.30627668e-01
-9.20169294e-01 -6.44044638e-01 -1.27180529e+00 3.08820675e-03
9.92274135e-02 1.67602867e-01 2.07744926e-01 8.15726519e-01
-1.30204558e+00 5.79116523e-01 -6.51013255e-01 -1.70929596e-01
5.04511595e-01 1.02554029e-02 -1.88950494e-01 -4.35369015e-01
-1.29821348e+00 1.18352103e+00 2.09377766e-01 5.99916458e-01
-1.38002360e+00 -3.62798810e-01 -7.56937802e-01 -1.41667590e-01
2.55587280e-01 -7.55704939e-01 1.18364501e+00 -1.11339617e+00
-1.71749640e+00 5.81228197e-01 -3.13751668e-01 -2.66778708e-01
7.01161087e-01 -5.88256657e-01 -4.00150418e-01 -1.75046846e-02
2.25821212e-02 2.82772094e-01 1.62145960e+00 -1.65264606e+00
-1.44326895e-01 -1.86971426e-01 -5.54920077e-01 4.04865623e-01
-4.19979207e-02 -9.47632417e-02 -5.30855775e-01 -9.93652284e-01
3.34561855e-01 -4.99209076e-01 -2.40749374e-01 -9.95428208e-03
-3.37233365e-01 2.87847191e-01 9.71878886e-01 -1.28700554e+00
1.25265539e+00 -1.94889843e+00 4.41898465e-01 -1.53517634e-01
3.88570338e-01 4.90855426e-01 -2.17294231e-01 -1.69202778e-02
-3.11923593e-01 -2.16674116e-02 -6.69132829e-01 -3.73792320e-01
-3.62828314e-01 -2.81225201e-02 -2.67401993e-01 9.91568029e-01
-3.70426625e-01 7.85672545e-01 -9.36605513e-01 -2.31419027e-01
5.37302732e-01 6.72453761e-01 -7.35632479e-02 4.03218806e-01
-1.02169672e-02 7.56959081e-01 1.19093237e-02 4.80787963e-01
1.24142468e+00 -3.41741145e-01 -2.18608111e-01 -4.29625809e-01
-3.39714205e-03 -5.74298084e-01 -1.38893950e+00 1.70325196e+00
-3.49255979e-01 1.08920419e+00 5.68494797e-01 -9.27821040e-01
6.05606019e-01 3.81952047e-01 2.08831072e-01 -2.20169753e-01
3.97790492e-01 1.95880398e-01 -3.60881031e-01 -7.68682361e-01
7.99461424e-01 -2.13114187e-01 4.56672281e-01 5.21974027e-01
-1.75011188e-01 -2.09994808e-01 -2.55128592e-01 -4.26310822e-02
8.21499705e-01 -4.40087765e-02 3.35938722e-01 -3.47684324e-01
8.18247914e-01 -2.77115971e-01 3.31159174e-01 1.11898422e+00
-5.00135899e-01 1.11925876e+00 -2.89982378e-01 -1.10240765e-01
-1.00987947e+00 -7.88897157e-01 -1.91389173e-01 7.11746514e-01
6.68890417e-01 3.17592710e-01 -9.80232358e-01 -2.08979294e-01
-5.87460637e-01 5.65032899e-01 -5.09605169e-01 4.81163785e-02
-3.71485919e-01 -1.16395450e+00 4.15648997e-01 2.26462573e-01
1.07638240e+00 -6.53839767e-01 1.14395387e-01 9.23854113e-02
-6.46918893e-01 -7.83041000e-01 -8.98321509e-01 -1.88562617e-01
-7.33899176e-01 -9.22908664e-01 -1.51710939e+00 -8.62255514e-01
7.93620825e-01 7.84719050e-01 8.23033750e-01 7.74597749e-02
-9.75975022e-02 2.58276522e-01 -1.76223323e-01 1.03086136e-01
-5.78202188e-01 -4.07115608e-01 9.65061933e-02 1.29533321e-01
1.92562476e-01 -6.02748632e-01 -8.42807829e-01 5.10877192e-01
-1.25992477e+00 -1.42050251e-01 6.00421071e-01 1.10112715e+00
3.27300102e-01 6.37557149e-01 4.39473331e-01 -2.12180898e-01
9.65764761e-01 -5.28708220e-01 -6.31919265e-01 2.53380924e-01
-6.91254020e-01 1.58006307e-02 2.27110356e-01 -8.35422516e-01
-1.23970282e+00 -2.46607348e-01 1.39059693e-01 -5.85430682e-01
-1.37140736e-01 4.70596075e-01 -2.57282287e-01 -2.94614524e-01
8.69484067e-01 5.87301552e-01 9.99230593e-02 -6.04521930e-01
5.92769861e-01 1.15732718e+00 9.38787460e-01 -2.51078606e-01
7.40822732e-01 4.04484540e-01 -2.17732713e-01 -9.08699274e-01
-5.44247627e-01 -4.74097282e-01 -3.11019570e-01 -2.59605497e-01
8.29916596e-01 -1.12896025e+00 -2.61922985e-01 1.44188797e+00
-1.25383818e+00 -1.88915744e-01 1.95664093e-01 5.17441273e-01
-3.42579097e-01 1.11394203e+00 -7.55107641e-01 -7.46904314e-01
-4.66470391e-01 -1.35487509e+00 6.56429172e-01 3.53464991e-01
1.51587054e-01 -1.02263153e+00 1.74423352e-01 6.78969324e-01
1.00959826e+00 -3.67105126e-01 4.32066321e-01 -2.59095374e-02
-4.86298680e-01 -1.37756199e-01 -6.78529799e-01 8.24975610e-01
5.25832236e-01 -6.66476071e-01 -1.00309348e+00 -3.76761228e-01
5.19712687e-01 1.60408560e-02 1.12923276e+00 1.11955404e+00
1.01912344e+00 -2.49225333e-01 -1.02630921e-01 5.95776975e-01
1.45007503e+00 -1.88776046e-01 9.93506193e-01 4.11679268e-01
8.66227329e-01 3.31415758e-02 -2.89860964e-02 1.52881399e-01
4.42303233e-02 4.41163957e-01 5.07614434e-01 -9.47186276e-02
-5.99216163e-01 1.42330229e-01 1.27233416e-01 6.98140144e-01
1.72022581e-01 -6.06225073e-01 -6.25991702e-01 8.28382373e-01
-1.91940510e+00 -8.16287518e-01 -7.04449117e-02 2.06307578e+00
1.11656356e+00 -4.24041569e-01 -6.67943120e-01 -2.48605665e-02
1.32626975e+00 5.11272907e-01 -7.85009563e-01 3.02362531e-01
-3.91870081e-01 -3.52081746e-01 7.41117895e-01 1.01317656e+00
-1.35917795e+00 8.07243168e-01 6.66989136e+00 9.58884537e-01
-9.98146951e-01 2.34527737e-01 6.40513122e-01 4.47251886e-01
7.94143230e-02 -1.82501227e-01 -1.95920646e-01 4.89365011e-01
4.81581241e-01 4.67314459e-02 8.88996124e-01 4.63480592e-01
4.54433620e-01 -3.95597339e-01 -5.92672706e-01 1.33104014e+00
4.29675728e-01 -1.20800924e+00 -1.07630678e-01 -1.75294414e-01
1.31010008e+00 1.25062913e-01 2.91157346e-02 -2.60397643e-01
5.68241656e-01 -1.30329180e+00 3.35898101e-01 8.89298856e-01
7.17001021e-01 -2.92986065e-01 1.05495310e+00 3.88356209e-01
-5.49392045e-01 1.25659123e-01 -3.43508929e-01 -9.68727923e-04
3.55233878e-01 1.08500707e+00 -1.90420926e-01 4.45731163e-01
7.25308120e-01 9.41608965e-01 -1.65350679e-02 1.30153871e+00
-4.88530844e-01 6.92338467e-01 4.03627567e-02 5.25634110e-01
-1.17314033e-01 -2.15534121e-01 9.12618458e-01 1.03195250e+00
5.88286579e-01 1.90832302e-01 -3.29634637e-01 8.68979514e-01
-2.90161252e-01 -3.20618927e-01 -1.34345427e-01 1.53981283e-01
3.89295429e-01 8.71266365e-01 -4.17018443e-01 -3.45108718e-01
-9.40879136e-02 1.47512257e+00 -3.88562024e-01 7.38813698e-01
-4.44700360e-01 -2.22555667e-01 5.44345200e-01 -2.21945763e-01
4.38561887e-01 -2.27303267e-01 -3.45861793e-01 -1.37335169e+00
-2.95331597e-01 -9.13041770e-01 -9.53220129e-02 -1.33742750e+00
-1.74477541e+00 4.82081175e-01 -1.48742840e-01 -1.18764603e+00
1.22026578e-01 -3.76872510e-01 -5.77222705e-01 1.39551854e+00
-1.74855304e+00 -8.82253826e-01 -5.40974140e-01 4.00046319e-01
6.44203901e-01 -2.05380589e-01 4.58058834e-01 3.09760392e-01
-5.42377591e-01 1.34622246e-01 6.56800091e-01 9.06947702e-02
8.37622821e-01 -9.59564388e-01 1.98558673e-01 1.54994738e+00
-4.06464845e-01 5.78766525e-01 1.11651957e+00 -8.88519347e-01
-1.11232984e+00 -9.55975413e-01 3.93825471e-01 -5.32060722e-03
5.51516235e-01 4.07165080e-01 -1.10003126e+00 3.45767498e-01
6.46293700e-01 3.20269406e-01 1.95266142e-01 -5.81785321e-01
-1.30103648e-01 1.64322719e-01 -1.30658448e+00 4.41015333e-01
5.51574647e-01 -5.53680003e-01 -7.34668612e-01 7.59578794e-02
4.61123556e-01 -4.93604988e-01 -4.68904555e-01 2.54126012e-01
1.82131007e-01 -7.09102035e-01 1.09174120e+00 -2.75666490e-02
5.49466372e-01 -7.94800937e-01 -2.73298949e-01 -1.73870850e+00
-6.14498019e-01 -7.93806851e-01 -4.42822278e-01 1.09048951e+00
-2.35772550e-01 -3.67651522e-01 4.89500582e-01 2.66423762e-01
9.07832533e-02 1.36467114e-01 -8.77908051e-01 -4.20620441e-01
1.36954695e-01 -3.10309291e-01 3.25852007e-01 9.97025192e-01
-6.18542850e-01 4.34584655e-02 -1.11672020e+00 6.98531270e-01
1.26145303e+00 -2.90219724e-01 2.74456561e-01 -9.60770190e-01
-2.51754194e-01 -4.56198007e-01 -1.85750768e-01 -1.64871728e+00
1.15671530e-01 -3.93934876e-01 5.96002698e-01 -1.79624617e+00
3.86439890e-01 7.74717182e-02 -3.76002312e-01 1.98267698e-01
-6.15784585e-01 3.48547876e-01 -3.52347195e-01 6.81908309e-01
-2.39080474e-01 6.75043821e-01 1.44121087e+00 -4.46298182e-01
-3.91696244e-02 -9.72866192e-02 -7.98799098e-01 5.71069658e-01
6.48180604e-01 -4.03366804e-01 -3.25348467e-01 -8.55343044e-01
7.24430801e-03 1.97605669e-01 5.08065343e-01 -8.60514641e-01
5.97415626e-01 -1.30962789e-01 3.70412678e-01 -6.62050009e-01
-5.21497317e-02 -8.00120950e-01 1.73201799e-01 7.04830512e-02
-2.01316491e-01 -5.46382189e-01 1.02917820e-01 7.98269570e-01
-2.79355198e-01 -4.68348265e-01 9.27011669e-01 -1.92207471e-01
-7.61355698e-01 -6.38410123e-03 -6.58418596e-01 -2.05880135e-01
4.96379256e-01 -2.19968975e-01 -5.11340261e-01 -6.50011122e-01
-5.62684000e-01 -7.12875128e-02 4.70544994e-01 3.05889577e-01
7.23695040e-01 -1.14819288e+00 -9.07041848e-01 1.99390754e-01
-3.02454650e-01 2.30848063e-02 6.38927877e-01 7.46120930e-01
-5.02055705e-01 1.70188874e-03 -8.69968310e-02 -4.74283636e-01
-1.12569141e+00 4.30243164e-01 7.62590826e-01 6.56403005e-02
-3.84857088e-01 1.04990852e+00 1.12488428e-02 -2.71458209e-01
3.12533021e-01 1.34585172e-01 -2.51444966e-01 -2.77785152e-01
1.06105411e+00 7.18773961e-01 2.35117599e-03 -9.01795864e-01
-1.48666371e-02 8.13452184e-01 -6.98188785e-03 -1.96993664e-01
1.15478909e+00 -9.16921258e-01 -5.60269117e-01 -1.65452715e-02
1.09687495e+00 -1.15942493e-01 -1.54899395e+00 -5.73272645e-01
-3.32027793e-01 -5.83730340e-01 1.12737572e+00 -9.21613693e-01
-1.33373046e+00 6.61141932e-01 1.21427095e+00 -1.55747578e-01
1.36781275e+00 -3.87216419e-01 7.85223305e-01 1.78613886e-01
1.00802174e-02 -7.75058746e-01 6.20925240e-02 4.30300295e-01
9.70768929e-01 -1.57837737e+00 2.62477666e-01 7.96338767e-02
-2.55076706e-01 1.05002666e+00 2.02744916e-01 -3.99073660e-02
8.16357315e-01 1.99947089e-01 4.03155476e-01 4.46277224e-02
4.80792858e-02 1.28715768e-01 3.14142495e-01 6.91458166e-01
7.76608661e-02 -1.61507666e-01 -6.37851059e-02 3.67860645e-01
4.07157451e-01 5.74170411e-01 6.61394894e-01 6.64598882e-01
-6.89918458e-01 -6.17668808e-01 -1.14595628e+00 1.52461678e-01
-5.78958869e-01 -4.98032957e-01 1.45402372e-01 -3.97168472e-02
1.70934096e-01 1.42356491e+00 -2.14464903e-01 -2.37454802e-01
-3.68616730e-01 -4.61029649e-01 2.94290960e-01 -1.22229904e-01
1.84014067e-02 1.69020772e-01 -4.27621871e-01 -6.64135143e-02
-7.99765110e-01 -3.31605822e-01 -7.39565015e-01 -2.90248275e-01
-6.94777429e-01 7.15718642e-02 6.81815505e-01 1.03366482e+00
1.72594517e-01 4.50787932e-01 5.04869401e-01 -1.36450303e+00
-5.75100482e-01 -1.32410574e+00 -6.81088746e-01 2.81965863e-02
1.04995573e+00 -4.25119281e-01 -8.89399350e-01 5.63026249e-01] | [11.623723983764648, -2.7111918926239014] |
060bbed3-82a6-4fa3-9a50-c2bd38605328 | multilogue-net-a-context-aware-rnn-for-multi | 2002.08267 | null | https://arxiv.org/abs/2002.08267v3 | https://arxiv.org/pdf/2002.08267v3.pdf | Multilogue-Net: A Context Aware RNN for Multi-modal Emotion Detection and Sentiment Analysis in Conversation | Sentiment Analysis and Emotion Detection in conversation is key in several real-world applications, with an increase in modalities available aiding a better understanding of the underlying emotions. Multi-modal Emotion Detection and Sentiment Analysis can be particularly useful, as applications will be able to use specific subsets of available modalities, as per the available data. Current systems dealing with Multi-modal functionality fail to leverage and capture - the context of the conversation through all modalities, the dependency between the listener(s) and speaker emotional states, and the relevance and relationship between the available modalities. In this paper, we propose an end to end RNN architecture that attempts to take into account all the mentioned drawbacks. Our proposed model, at the time of writing, out-performs the state of the art on a benchmark dataset on a variety of accuracy and regression metrics. | ['Ashish Sardana', 'Aman Shenoy'] | 2020-02-19 | multilogue-net-a-context-aware-rnn-for-multi-1 | https://128.84.21.199/abs/2002.08267 | https://128.84.21.199/pdf/2002.08267.pdf | arxiv-preprint-2020-2 | ['multimodal-emotion-recognition', 'emotion-recognition-in-conversation', 'multimodal-emotion-recognition'] | ['computer-vision', 'natural-language-processing', 'speech'] | [ 1.67824954e-01 -1.59780204e-01 -1.65259182e-01 -5.05241156e-01
-6.49981260e-01 -5.61007977e-01 6.78050876e-01 1.38563558e-01
-4.46597457e-01 3.68977815e-01 7.09443390e-01 1.89040631e-01
5.94870262e-02 -4.21491176e-01 1.04575483e-02 -3.07793468e-01
1.00408144e-01 1.00981392e-01 -2.32698888e-01 -5.43973744e-01
-8.91362354e-02 3.15307766e-01 -1.73293161e+00 6.72510445e-01
1.51224166e-01 1.31003857e+00 -1.91440042e-02 8.69124115e-01
-2.15863913e-01 1.03120470e+00 -4.13970709e-01 -3.45182806e-01
-2.20872179e-01 -4.97305006e-01 -6.82935834e-01 -2.10420355e-01
-3.03945374e-02 -8.72238725e-02 1.55968014e-02 6.37147784e-01
6.95275068e-01 2.87457526e-01 3.75789672e-01 -1.07017875e+00
-5.00972457e-02 4.30362433e-01 -1.93184942e-01 7.64675364e-02
7.17268467e-01 -1.66872472e-01 1.11564982e+00 -8.06029201e-01
5.63546777e-01 8.47179949e-01 4.07192826e-01 8.06087494e-01
-7.90522456e-01 -4.07851398e-01 1.13181829e-01 1.57687426e-01
-8.89566779e-01 -7.87893295e-01 1.12004471e+00 -2.04905644e-01
1.19624352e+00 3.76016229e-01 4.63271648e-01 1.41452086e+00
-1.49059772e-01 6.48941159e-01 1.32162273e+00 -6.59867525e-01
2.06366941e-01 5.64139247e-01 7.30266422e-02 2.33936220e-01
-6.25617027e-01 -2.30493098e-01 -1.17711544e+00 -3.97184230e-02
-4.44745906e-02 -3.44368406e-02 -2.02279016e-01 -3.62343602e-02
-1.03828943e+00 5.76052785e-01 9.44170505e-02 7.52158403e-01
-7.37612188e-01 -6.22775219e-02 9.33203638e-01 3.95404369e-01
5.41853070e-01 3.16590041e-01 -7.50993073e-01 -9.06656444e-01
-9.93473172e-01 -3.85272980e-01 9.90412414e-01 3.38305116e-01
5.53309202e-01 -1.13925539e-01 9.50934365e-02 8.83509338e-01
4.32723820e-01 2.21090272e-01 5.98606765e-01 -7.12138295e-01
3.42652470e-01 8.22106481e-01 -1.76430847e-02 -8.92667592e-01
-7.29657531e-01 -8.89655873e-02 -8.00602853e-01 1.12929344e-01
1.49009138e-01 -4.84868824e-01 -5.64466655e-01 1.96668589e+00
4.70646411e-01 -3.52359680e-03 5.20851374e-01 8.83951843e-01
1.04392409e+00 5.69449425e-01 1.76744103e-01 -4.07504410e-01
1.63469899e+00 -5.86859405e-01 -8.44743192e-01 -6.18382931e-01
4.29119736e-01 -7.98151851e-01 1.05022645e+00 3.12655598e-01
-8.32203209e-01 -3.33677709e-01 -9.56062615e-01 5.91702461e-02
-6.54684186e-01 1.55760929e-01 6.50204062e-01 7.31935501e-01
-1.00362051e+00 2.95962870e-01 -5.96493542e-01 -5.78949094e-01
-1.35649130e-01 4.43558186e-01 -6.33122921e-01 2.39639714e-01
-1.31385946e+00 1.11768520e+00 1.05981350e-01 5.66448867e-01
-2.67577559e-01 -2.13285238e-01 -7.97500432e-01 5.52653857e-02
2.01983154e-01 -4.32221830e-01 1.15083253e+00 -1.39285362e+00
-1.95945144e+00 6.70316994e-01 -3.27327967e-01 -1.42164484e-01
2.98965991e-01 -2.94852972e-01 -7.02006459e-01 1.49959654e-01
-5.07848084e-01 4.32013452e-01 6.32647514e-01 -9.62437391e-01
-5.24826586e-01 -4.53125179e-01 2.30287671e-01 4.56223994e-01
-7.40836561e-01 3.60528231e-01 -3.84279341e-01 -7.16681853e-02
-2.73503780e-01 -8.96441519e-01 4.09959033e-02 -4.58775252e-01
-2.94595361e-01 -3.92044373e-02 8.82019222e-01 -5.30279279e-01
1.02807331e+00 -2.33422446e+00 3.62412721e-01 2.08156362e-01
-2.05422312e-01 7.08038658e-02 -1.27082601e-01 7.31244922e-01
-1.55287042e-01 -9.28898379e-02 2.62001723e-01 -8.02816689e-01
1.70488194e-01 1.22141108e-01 -1.99568197e-01 3.18039745e-01
7.17661381e-02 5.16909957e-01 -4.82487679e-01 -3.01654965e-01
3.59639049e-01 9.01699722e-01 -1.35523528e-01 4.58549678e-01
-1.29664302e-01 4.34131891e-01 -3.67014289e-01 4.04573083e-01
2.41451919e-01 5.09390756e-02 2.85461426e-01 -2.74075091e-01
-9.24700424e-02 4.31036413e-01 -1.09956157e+00 1.83961892e+00
-1.00101483e+00 9.22558188e-01 3.08732420e-01 -5.88763118e-01
7.79018760e-01 8.45257819e-01 4.50589985e-01 -7.30373800e-01
3.46598893e-01 2.83498224e-02 -2.84536220e-02 -5.95948279e-01
6.23147786e-01 -4.20296252e-01 -3.70271534e-01 7.13490307e-01
1.09976187e-01 2.03078628e-01 6.62534386e-02 6.66052252e-02
8.47242892e-01 2.67420243e-02 4.08222467e-01 4.14800882e-01
6.26011431e-01 -4.79614794e-01 2.81214476e-01 4.24585253e-01
-1.76847801e-01 2.70215005e-01 6.00774109e-01 -1.59170747e-01
-4.38560665e-01 -3.39647114e-01 3.29070538e-01 1.59914398e+00
-2.19751492e-01 -2.81623930e-01 -4.37289655e-01 -4.65587676e-01
-5.94868898e-01 5.19483149e-01 -9.08926308e-01 -7.74739608e-02
-1.48390383e-01 -5.45228601e-01 5.66219211e-01 3.92333090e-01
2.03600243e-01 -1.53492141e+00 -9.83183742e-01 9.89783853e-02
-4.22556102e-01 -1.41139817e+00 5.93348816e-02 4.25004363e-01
-5.55593789e-01 -7.75092244e-01 -2.43562788e-01 -2.84397990e-01
1.21769018e-01 -1.89399272e-01 1.09997690e+00 -1.91837952e-01
8.17720518e-02 7.74522424e-01 -5.52291155e-01 -4.79168981e-01
-5.33897936e-01 2.81730920e-01 -1.53596789e-01 3.72895569e-01
3.51755232e-01 -6.44759238e-01 -4.06417072e-01 3.59696671e-02
-8.90383780e-01 -2.64746398e-02 3.81016135e-01 4.78743017e-01
1.09385617e-01 -2.46529415e-01 7.93146491e-01 -6.84102118e-01
7.90406823e-01 -4.79612797e-01 6.59857020e-02 1.91164792e-01
-2.05465659e-01 -1.47392392e-01 6.09251499e-01 -3.36647749e-01
-1.29569197e+00 2.54742615e-02 -3.46865475e-01 -1.20109230e-01
-5.95730662e-01 9.29460645e-01 -1.81673691e-01 1.49630278e-01
4.06401038e-01 -3.77540290e-02 -1.93882301e-01 -2.02983037e-01
5.32391667e-01 1.05241740e+00 3.07186842e-01 -2.95972228e-01
1.19843118e-01 5.26524246e-01 -4.98820432e-02 -8.06668520e-01
-8.83214235e-01 -7.02572942e-01 -5.46598375e-01 -6.27571285e-01
7.29754984e-01 -8.69513929e-01 -9.59209919e-01 3.69541496e-01
-1.06859481e+00 -1.67470485e-01 -7.08844438e-02 5.81453800e-01
-4.16844696e-01 1.87384084e-01 -5.64161181e-01 -1.24283290e+00
-5.72695255e-01 -9.61362720e-01 7.91655779e-01 3.68504286e-01
-6.75995052e-01 -1.03155422e+00 1.63107827e-01 5.03444552e-01
6.75452113e-01 2.56201804e-01 5.47830701e-01 -9.20852184e-01
1.40603825e-01 -5.64380348e-01 8.85504708e-02 3.96304637e-01
1.31392211e-01 -4.07796800e-02 -1.52569413e+00 6.90530464e-02
9.14374813e-02 -5.71931541e-01 6.27653301e-01 6.81194887e-02
5.68226039e-01 -1.82922110e-02 8.92447010e-02 2.87026968e-02
1.21222436e+00 -6.00516237e-02 3.59198719e-01 2.17507958e-01
4.81742859e-01 9.11849141e-01 4.65189934e-01 6.70600235e-01
5.39284170e-01 6.28953516e-01 7.16755569e-01 -1.22246087e-01
2.72946179e-01 2.32505277e-01 6.12984598e-01 7.75374174e-01
-1.40487580e-02 -3.33213061e-01 -7.04579055e-01 3.19761753e-01
-1.93805337e+00 -1.08996522e+00 6.22949675e-02 1.97023523e+00
7.91913509e-01 9.72632468e-02 1.86607361e-01 7.74370730e-02
3.25404495e-01 4.72854376e-01 -5.15947640e-01 -9.51668501e-01
-7.18113109e-02 3.25848092e-03 -8.54235068e-02 3.26375365e-01
-9.98202085e-01 5.21998823e-01 6.09595442e+00 3.18734795e-01
-1.45761144e+00 8.00898075e-02 4.35768306e-01 -4.25352275e-01
-1.74872056e-01 -2.88964123e-01 -4.43539560e-01 2.27828369e-01
1.41648602e+00 4.57919270e-01 6.54259980e-01 6.01361156e-01
3.96747500e-01 -3.10513526e-01 -1.32032418e+00 9.54209805e-01
4.29003775e-01 -8.62569511e-01 -5.74560761e-01 -2.20213190e-01
2.29284450e-01 2.91518927e-01 2.80691721e-02 3.95082116e-01
-2.50127256e-01 -8.77663016e-01 6.74107194e-01 6.21979713e-01
4.88926798e-01 -8.13377500e-01 1.03014827e+00 4.69496012e-01
-1.00376189e+00 -8.08881149e-02 4.28224742e-01 -3.66723329e-01
2.95475632e-01 4.35605139e-01 -7.41194427e-01 4.83368367e-01
5.50396800e-01 4.67467368e-01 -1.84604660e-01 2.36075044e-01
-2.22950980e-01 4.51825768e-01 -5.31184852e-01 -3.18833828e-01
3.08549255e-02 2.20437702e-02 3.67453903e-01 1.49550688e+00
3.12953234e-01 -4.71939370e-02 -1.69518739e-01 2.71765202e-01
-1.78996146e-01 3.51508856e-01 -4.74968135e-01 -2.08309859e-01
1.26274914e-01 1.63245249e+00 -6.65765703e-01 -2.05314189e-01
-6.93229675e-01 9.79410708e-01 2.55030036e-01 2.36659646e-01
-5.97811639e-01 -1.31562382e-01 7.01413274e-01 -5.86483359e-01
2.06059963e-01 -4.20188718e-02 -3.87732759e-02 -1.05855620e+00
2.01417521e-01 -1.11110437e+00 5.56803226e-01 -8.73627543e-01
-1.01023662e+00 7.81740367e-01 -3.84547144e-01 -9.85352993e-01
-6.41952038e-01 -5.30223966e-01 -4.95048970e-01 8.39556396e-01
-1.54181957e+00 -1.44509363e+00 -1.27623960e-01 4.97973680e-01
3.90306294e-01 4.34005708e-02 1.35433853e+00 2.70989865e-01
-5.88445306e-01 3.04992139e-01 -2.14893386e-01 5.09006977e-02
8.90605032e-01 -9.90969181e-01 -3.58811706e-01 6.37013495e-01
2.23793626e-01 4.78342682e-01 9.60519314e-01 -5.29430360e-02
-1.60868216e+00 -3.46351683e-01 1.04758871e+00 -3.69555414e-01
7.31532037e-01 -3.23438317e-01 -6.04363322e-01 5.10283351e-01
6.75761342e-01 -1.01694472e-01 1.21961856e+00 8.01718056e-01
-4.68226194e-01 -1.61904946e-01 -1.00698125e+00 4.57641065e-01
2.63285011e-01 -1.13571990e+00 -3.26145887e-01 -1.75412223e-01
1.49351239e-01 -4.68850195e-01 -9.35016334e-01 2.95402497e-01
8.81371796e-01 -1.29594195e+00 6.40361965e-01 -3.84884596e-01
5.48516691e-01 -2.41956711e-02 -3.59974086e-01 -1.43367290e+00
3.00518900e-01 -5.21467507e-01 -2.48947278e-01 1.57376969e+00
6.33192897e-01 -4.12577301e-01 5.79981506e-01 9.03267086e-01
1.74674809e-01 -7.94929385e-01 -1.07327890e+00 2.08592147e-01
-5.59544861e-01 -9.73495603e-01 4.06657279e-01 9.69538927e-01
6.47679806e-01 7.43725479e-01 -7.24586785e-01 1.26033381e-01
-2.25881264e-01 2.79219687e-01 7.26077795e-01 -1.08675551e+00
-3.18723083e-01 -5.56424677e-01 -2.58264869e-01 -4.83639389e-01
3.45317125e-01 -4.52461302e-01 -7.54418690e-03 -1.56564987e+00
1.27212092e-01 1.69829819e-02 -5.28028905e-01 6.61324620e-01
1.15845287e-02 3.28575522e-01 3.56606930e-01 -2.21718192e-01
-8.02946448e-01 4.09314215e-01 6.50004923e-01 8.61282498e-02
-4.45857137e-01 -1.00127064e-01 -7.68530190e-01 8.17450583e-01
8.57548773e-01 -1.78133890e-01 -2.31091797e-01 -1.60202250e-01
8.99494290e-01 2.60213286e-01 1.33716181e-01 -7.94306815e-01
3.38848978e-01 -8.03590715e-02 2.62841165e-01 -5.41275084e-01
1.08540666e+00 -1.16828442e+00 4.15906450e-03 -1.97799370e-01
-5.10730386e-01 1.20264636e-02 3.53975475e-01 3.40171188e-01
-4.17936146e-01 -1.48514226e-01 4.47935551e-01 7.88031965e-02
-7.69150257e-01 -2.34321475e-01 -5.55106461e-01 -1.53683096e-01
6.82047069e-01 -1.31636307e-01 -2.56070644e-01 -7.76334226e-01
-8.46350610e-01 -1.95039362e-02 1.66042462e-01 7.09127545e-01
3.39949369e-01 -9.55644250e-01 -2.90969312e-01 -4.28647082e-03
1.15180559e-01 -6.03615761e-01 5.53223908e-01 9.50007856e-01
2.51690358e-01 1.84532285e-01 -9.77530032e-02 -1.39059395e-01
-1.47051871e+00 9.42612290e-02 5.01613915e-01 -3.77140433e-01
-5.74501231e-03 5.46911716e-01 -3.10050339e-01 -4.83681232e-01
2.32386589e-01 -1.73899010e-01 -7.18727052e-01 7.91734815e-01
4.88388836e-01 1.78774431e-01 2.89769322e-01 -9.60240304e-01
-5.27177989e-01 4.03789073e-01 1.76203653e-01 -4.49113160e-01
1.51912630e+00 -3.91158789e-01 -1.30941793e-01 1.06295907e+00
1.17513406e+00 1.50752544e-01 -7.83778787e-01 -2.41252333e-01
-8.32986385e-02 7.11984783e-02 2.75999367e-01 -1.12398505e+00
-9.07693386e-01 9.29294944e-01 6.71762288e-01 5.20834506e-01
1.41616082e+00 5.26684523e-02 6.24195099e-01 3.35267842e-01
4.51024249e-02 -1.37819719e+00 -7.31931329e-02 5.88640869e-01
6.08825564e-01 -1.39105403e+00 -1.21102400e-01 3.22166272e-02
-9.91977155e-01 1.23812914e+00 3.45945865e-01 4.35423523e-01
6.06004238e-01 3.49606872e-01 6.71795905e-01 -2.71628439e-01
-1.07678354e+00 -2.85347670e-01 3.13568711e-01 1.59192115e-01
9.08700407e-01 4.59231585e-02 1.02841727e-01 6.85531139e-01
2.62339078e-02 -6.42222315e-02 3.09797198e-01 7.88054764e-01
-1.01986632e-01 -1.11909437e+00 -3.78532767e-01 2.72892237e-01
-7.30884314e-01 -1.94207821e-02 -5.73687792e-01 4.78366256e-01
1.72931999e-02 1.30746019e+00 -6.23866506e-02 -3.88586402e-01
6.23812377e-01 4.47029978e-01 1.44837499e-01 -3.73732239e-01
-8.97533596e-01 1.36847287e-01 5.13496876e-01 -6.00797117e-01
-1.03879213e+00 -7.21180201e-01 -1.11578882e+00 8.03421885e-02
-3.13313633e-01 -6.15267977e-02 1.11329484e+00 1.36671138e+00
5.15309453e-01 5.72299778e-01 7.19060421e-01 -1.00706124e+00
-1.06677704e-01 -1.21103764e+00 -4.16227192e-01 3.47465068e-01
5.59186578e-01 -2.10863844e-01 -4.56642330e-01 -1.01982467e-01] | [13.332121849060059, 5.436816215515137] |
8c7982cc-5d87-4e58-b64d-f04889da0e8c | epic-graph-augmentation-with-edit-path | 2306.01310 | null | https://arxiv.org/abs/2306.01310v1 | https://arxiv.org/pdf/2306.01310v1.pdf | EPIC: Graph Augmentation with Edit Path Interpolation via Learnable Cost | Graph-based models have become increasingly important in various domains, but the limited size and diversity of existing graph datasets often limit their performance. To address this issue, we propose EPIC (Edit Path Interpolation via learnable Cost), a novel interpolation-based method for augmenting graph datasets. Our approach leverages graph edit distance to generate new graphs that are similar to the original ones but exhibit some variation in their structures. To achieve this, we learn the graph edit distance through a comparison of labeled graphs and utilize this knowledge to create graph edit paths between pairs of original graphs. With randomly sampled graphs from a graph edit path, we enrich the training set to enhance the generalization capability of classification models. We demonstrate the effectiveness of our approach on several benchmark datasets and show that it outperforms existing augmentation methods in graph classification tasks. | ['Dongwoo Kim', 'Sungsoo Ahn', 'Seungbeom Lee', 'Jaeseung Heo'] | 2023-06-02 | null | null | null | null | ['graph-classification'] | ['graphs'] | [ 5.06551147e-01 4.55706298e-01 -3.91819179e-01 -4.01148081e-01
-3.74884874e-01 -8.47663701e-01 5.16092300e-01 6.06877685e-01
-7.96418488e-02 7.74322212e-01 4.15386558e-02 -3.19141805e-01
-1.15171514e-01 -1.15886199e+00 -8.79222691e-01 -2.14908928e-01
-3.16175312e-01 4.19012755e-01 1.49924248e-01 -2.26541981e-01
2.16256917e-01 3.87333572e-01 -1.07796288e+00 1.46349937e-01
1.32600927e+00 3.93202245e-01 -2.57065415e-01 3.50186408e-01
-1.81049332e-01 5.08975625e-01 -3.39451700e-01 -6.26347780e-01
5.03784239e-01 -5.48045397e-01 -8.29931855e-01 2.22018689e-01
5.68750441e-01 -1.74570113e-01 -6.77672684e-01 1.10026991e+00
4.71405871e-02 3.65731984e-01 5.73929012e-01 -1.55132961e+00
-1.15739644e+00 9.80056942e-01 -5.37864983e-01 9.86108743e-03
4.72163141e-01 -1.44075006e-02 1.28263021e+00 -7.46855736e-01
8.79853129e-01 1.11902046e+00 9.62408960e-01 4.53728646e-01
-1.60256684e+00 -4.67572898e-01 2.99730122e-01 8.99204239e-02
-1.27851045e+00 7.31867254e-02 1.15560687e+00 -2.36446708e-01
6.72773063e-01 1.66345134e-01 7.60581851e-01 8.48939836e-01
-2.13831756e-02 3.94067913e-01 9.82396007e-01 -4.39330965e-01
3.65579762e-02 -1.60475329e-01 2.13123932e-01 1.08583868e+00
3.98344636e-01 -5.29363342e-02 -3.03230911e-01 -2.56825417e-01
5.64227462e-01 -9.11935698e-03 -2.90072173e-01 -7.36717105e-01
-1.08189166e+00 8.64096761e-01 9.17633533e-01 4.69204644e-03
-1.28724605e-01 2.09042519e-01 3.82026762e-01 4.01259691e-01
5.74863732e-01 8.25332165e-01 -2.24639937e-01 3.33069265e-01
-3.40423226e-01 2.59992957e-01 7.79081464e-01 1.11417687e+00
1.03009939e+00 -1.30512919e-02 -2.01464161e-01 7.97006667e-01
8.54707658e-02 -3.04134190e-02 1.76292747e-01 -7.75638700e-01
6.12419844e-01 1.23053885e+00 -4.66769248e-01 -1.25601625e+00
-1.69926673e-01 -3.43226582e-01 -8.05086851e-01 -9.59463865e-02
3.48218828e-01 1.79841444e-01 -1.01525545e+00 1.82169747e+00
5.59816957e-01 3.06395501e-01 -2.96528220e-01 4.34324086e-01
7.42362320e-01 3.43706995e-01 7.55122304e-02 1.55991375e-01
8.01271737e-01 -1.16360497e+00 -4.32507306e-01 -2.34549239e-01
1.15063345e+00 -1.86853915e-01 1.40047801e+00 8.55078995e-02
-7.63330996e-01 -3.23313445e-01 -1.06748736e+00 -1.80444047e-01
-4.77272838e-01 -3.93247187e-01 7.83167183e-01 6.65017545e-01
-9.70037043e-01 1.01548207e+00 -6.56367600e-01 -1.73301294e-01
5.26509941e-01 2.69360691e-01 -5.96173227e-01 -5.02948284e-01
-8.97403181e-01 4.53508288e-01 6.62854195e-01 -2.20423996e-01
-5.94624937e-01 -9.48041975e-01 -1.26345253e+00 -9.29125473e-02
5.59233129e-01 -6.79004967e-01 8.05290222e-01 -8.14755857e-01
-1.04999471e+00 7.58208990e-01 1.71499550e-01 -4.71163124e-01
4.36874479e-01 1.75508320e-01 -2.87835121e-01 -2.07980033e-02
-8.54004100e-02 7.06620395e-01 7.95922518e-01 -1.12839389e+00
-2.23803058e-01 -3.91649336e-01 2.75089860e-01 9.01183411e-02
-5.98306894e-01 -4.80259746e-01 -4.65300441e-01 -1.05617929e+00
2.20140684e-02 -1.11223412e+00 -4.36231464e-01 1.94289908e-01
-6.23539805e-01 -7.22954795e-02 8.62302840e-01 -6.41508698e-01
1.27031004e+00 -1.94278693e+00 2.67497152e-01 4.18890595e-01
6.96988344e-01 9.95677039e-02 -4.52929884e-01 5.35361409e-01
-8.24086592e-02 4.21801835e-01 -8.55933309e-01 -3.06377858e-01
-2.19323456e-01 4.61405128e-01 9.15943682e-02 1.96202189e-01
2.73225784e-01 1.10859656e+00 -1.30440712e+00 -4.26754743e-01
-2.62890831e-02 3.07021409e-01 -7.88485467e-01 1.83525845e-01
-4.32618171e-01 6.41690910e-01 -3.61695796e-01 4.11477447e-01
6.59026623e-01 -2.77788997e-01 5.35040200e-01 -2.95047283e-01
6.53355539e-01 1.92230478e-01 -1.03512692e+00 1.86276472e+00
-2.92835593e-01 4.05352890e-01 -4.38668370e-01 -1.10002232e+00
1.03346896e+00 -2.00203896e-01 4.53113645e-01 -3.44119370e-01
-1.26668692e-01 7.39552453e-02 9.80437174e-02 -1.77492544e-01
4.84040141e-01 2.88665712e-01 4.20495458e-02 6.45543754e-01
-5.40609024e-02 -2.91253626e-01 5.04693925e-01 6.03183866e-01
1.36696684e+00 6.36001006e-02 2.30658904e-01 -6.55358806e-02
3.65638137e-01 -1.13508806e-01 4.25860435e-01 5.59624732e-01
8.57995544e-03 5.03848314e-01 5.40717244e-01 -4.71632898e-01
-1.12174070e+00 -1.14093804e+00 2.29718342e-01 7.35165536e-01
-2.18416583e-02 -7.66701937e-01 -8.22344244e-01 -1.53316343e+00
2.89363503e-01 6.64985359e-01 -8.11214149e-01 -6.50418937e-01
-6.57195270e-01 -5.37480772e-01 4.86938566e-01 6.14961147e-01
2.93306708e-01 -6.98440850e-01 3.56893539e-01 1.50656745e-01
-1.67220943e-02 -1.27622151e+00 -9.87184107e-01 -3.30513656e-01
-1.13210392e+00 -1.43725502e+00 -2.60998696e-01 -1.01504028e+00
1.31113529e+00 3.07821095e-01 1.24655557e+00 5.78860581e-01
-3.16004813e-01 3.85750830e-01 -4.92641270e-01 -4.93078865e-02
-8.24235499e-01 3.48776549e-01 -2.29082197e-01 4.56841663e-02
-9.18650776e-02 -8.63161623e-01 -3.61434549e-01 1.87132880e-01
-1.05653179e+00 1.69613838e-01 1.46064475e-01 7.73483038e-01
7.31660545e-01 -1.12728477e-02 6.87178433e-01 -1.46652997e+00
9.05398130e-01 -5.79533994e-01 -6.37872577e-01 2.74916351e-01
-1.00593507e+00 3.58753085e-01 7.95892596e-01 -4.08104092e-01
-6.22367978e-01 2.06592586e-02 2.44171008e-01 -4.45573211e-01
3.37487251e-01 9.10381198e-01 -6.77177683e-02 -5.68707824e-01
7.68237233e-01 -8.99527669e-02 1.43446788e-01 -4.68564004e-01
9.15972352e-01 1.39221400e-01 5.22265255e-01 -5.80975890e-01
1.03469825e+00 2.40961537e-01 3.77782017e-01 -4.95100498e-01
-6.74872160e-01 7.26335123e-02 -8.31671000e-01 -6.82193339e-02
4.14597303e-01 -6.04187667e-01 -6.89617470e-02 2.33897045e-01
-9.10108447e-01 -5.37424088e-01 -4.45077479e-01 2.92126149e-01
-2.05000743e-01 7.59925902e-01 -5.75194538e-01 -3.10232401e-01
-3.25319678e-01 -9.20437515e-01 7.65345156e-01 -9.26793143e-02
-9.92112309e-02 -1.43143725e+00 6.41710460e-02 9.56524983e-02
2.89344843e-02 8.64598811e-01 1.30122149e+00 -6.93136990e-01
-7.15062559e-01 -4.42530245e-01 -2.60503531e-01 3.96227300e-01
6.35655820e-01 -2.42324006e-02 -4.29243684e-01 -4.87792879e-01
-6.67729020e-01 -2.09376588e-01 7.25161672e-01 -7.36269131e-02
1.60349941e+00 -4.95533407e-01 -6.05211616e-01 6.86084867e-01
1.20112753e+00 -1.38775855e-01 4.54883307e-01 -7.71865249e-02
1.25086451e+00 5.45567513e-01 4.74621236e-01 2.03829885e-01
5.34361184e-01 5.99069834e-01 3.45030189e-01 5.08277602e-02
-4.00421321e-01 -7.57701337e-01 1.31297663e-01 9.95756924e-01
4.87640128e-03 -2.88759291e-01 -8.65779221e-01 6.13321185e-01
-1.63918114e+00 -4.55413073e-01 -3.40906739e-01 2.37370348e+00
8.48716617e-01 9.94136184e-02 1.63281947e-01 8.98882970e-02
9.09940600e-01 1.83409184e-01 -5.54128408e-01 -3.10657084e-01
3.06067877e-02 4.06905770e-01 4.47333932e-01 5.93432188e-01
-8.55488300e-01 9.87810194e-01 6.50982380e+00 4.49365020e-01
-7.60666609e-01 -1.06223665e-01 4.73825514e-01 3.11742961e-01
-7.24029720e-01 2.49429613e-01 -2.48666584e-01 3.77713352e-01
6.41109645e-01 -5.01665890e-01 7.59548306e-01 7.15669334e-01
-2.56043404e-01 4.14745420e-01 -1.49451339e+00 6.68495297e-01
8.90671238e-02 -1.47753513e+00 3.30096334e-01 9.33120251e-02
1.00118363e+00 -7.35604465e-02 -9.73004103e-02 3.28175515e-01
6.67440653e-01 -1.01771736e+00 1.34677723e-01 2.52071202e-01
7.95325279e-01 -6.53600395e-01 2.90714413e-01 1.40152825e-02
-1.51986194e+00 1.64445594e-01 -1.51786938e-01 5.64614050e-02
4.62344699e-02 7.31957018e-01 -1.20019495e+00 9.06208217e-01
2.61269838e-01 9.59793270e-01 -1.06384015e+00 9.96297717e-01
-4.93615121e-01 6.35610461e-01 -1.64831236e-01 2.21514612e-01
1.55548647e-03 -6.31247520e-01 6.60683632e-01 8.09562027e-01
1.96499214e-01 -1.95794925e-01 4.81620729e-01 1.05579484e+00
-7.52718031e-01 1.49456561e-01 -1.15397501e+00 -4.11176533e-01
7.64143169e-01 1.23376632e+00 -6.21217132e-01 -1.66943952e-01
-4.51546460e-01 1.05476010e+00 7.79501259e-01 2.26638660e-01
-9.11261976e-01 -6.00499153e-01 4.75008070e-01 5.48642687e-02
-3.91293019e-02 -4.47919279e-01 -1.01788431e-01 -1.06628549e+00
3.33177656e-01 -1.00956488e+00 7.44085193e-01 -5.87378442e-01
-1.42584062e+00 5.78584969e-01 5.18788993e-02 -1.00162482e+00
3.20388563e-02 -2.56126016e-01 -6.11196339e-01 4.90996301e-01
-1.32048321e+00 -1.42424500e+00 -8.16364586e-01 4.38385278e-01
3.08513880e-01 2.81509832e-02 6.14893019e-01 3.52682203e-01
-4.81130689e-01 9.94115233e-01 -7.20107257e-02 2.80369639e-01
6.13515437e-01 -1.38675058e+00 1.17289293e+00 7.59632230e-01
4.37006891e-01 5.05256951e-01 2.95195520e-01 -1.00123024e+00
-1.53554666e+00 -1.68956232e+00 4.65518773e-01 -4.92454886e-01
6.96120381e-01 -4.32690591e-01 -1.31338584e+00 1.08747029e+00
-9.94082540e-02 4.31188107e-01 4.72987533e-01 -6.92017972e-02
-7.26715565e-01 3.68971489e-02 -1.26505303e+00 8.51896167e-01
1.82287490e+00 -5.51878512e-01 -2.92044014e-01 5.24278879e-01
1.14867544e+00 -4.19273466e-01 -1.13824165e+00 5.18801570e-01
2.52265960e-01 -5.15110731e-01 8.42705309e-01 -1.05763841e+00
3.15364599e-01 -2.27319226e-01 5.91850355e-02 -1.79827392e+00
-1.77561164e-01 -7.27222264e-01 -2.31536537e-01 1.27644479e+00
6.14807129e-01 -1.03780603e+00 1.05063570e+00 5.90405643e-01
-2.31167495e-01 -8.31565320e-01 -4.58820015e-01 -1.10694611e+00
1.18212644e-02 2.57094391e-03 9.96053278e-01 1.36261916e+00
1.21799912e-02 2.69652575e-01 -8.13939422e-02 7.17424080e-02
7.94840157e-01 2.49977410e-02 1.07004929e+00 -1.31248403e+00
-3.22426379e-01 -2.49059305e-01 -6.16566241e-01 -6.29604995e-01
5.98057151e-01 -1.67329895e+00 -3.45468998e-01 -1.62464571e+00
7.84777179e-02 -7.71751106e-01 -1.13718383e-01 6.06844723e-01
-5.59388816e-01 2.37374842e-01 -1.27282627e-02 -2.83597335e-02
-2.41685659e-01 7.75280952e-01 1.38307369e+00 -2.80715317e-01
-3.39129776e-01 -2.53747880e-01 -6.62526071e-01 4.42803442e-01
8.04605782e-01 -5.59618831e-01 -8.26959372e-01 -5.50185680e-01
1.67248636e-01 -3.81847888e-01 1.95850328e-01 -8.18103313e-01
-1.50692806e-01 -9.05167907e-02 -5.47554381e-02 2.09105127e-02
-3.76789905e-02 -6.21598661e-01 2.79396743e-01 4.87766623e-01
-5.35361886e-01 3.09431195e-01 8.76312405e-02 1.03021038e+00
-6.19772486e-02 -7.32087940e-02 7.14343429e-01 3.74175198e-02
-4.56692308e-01 8.32676768e-01 3.15494359e-01 2.99640626e-01
1.17782426e+00 -1.54590979e-01 -3.80087465e-01 -2.88118929e-01
-6.39701128e-01 1.78998724e-01 7.51238346e-01 4.70933080e-01
6.01911962e-01 -1.52940261e+00 -6.24874413e-01 1.15116209e-01
4.86245692e-01 2.12909490e-01 -1.87541857e-01 6.73761368e-01
-5.53614616e-01 -2.03365743e-01 -1.51359275e-01 -3.29729170e-01
-1.28795493e+00 8.01084757e-01 3.39491695e-01 -4.53723162e-01
-6.95220470e-01 6.32701874e-01 1.00647889e-01 -9.89522815e-01
-1.05624404e-02 -3.44145089e-01 1.16634324e-01 -3.88368964e-01
6.55778125e-02 4.44050491e-01 1.63713276e-01 -2.03626171e-01
-4.15381528e-02 3.44898790e-01 -2.70525455e-01 2.83235699e-01
1.18978858e+00 1.38757572e-01 -2.62419492e-01 1.50681153e-01
1.41616976e+00 1.79134920e-01 -1.12264681e+00 -3.94429922e-01
2.53816962e-01 -8.07168186e-01 -2.17479363e-01 -4.40678120e-01
-1.26694191e+00 5.05285382e-01 1.45598784e-01 3.25736433e-01
9.59070742e-01 -3.06087472e-02 9.27291155e-01 4.11595464e-01
2.85810053e-01 -6.83058679e-01 3.61715287e-01 1.64479285e-01
9.56451714e-01 -1.30523705e+00 2.13653103e-01 -1.16133320e+00
-4.45191920e-01 9.21615779e-01 7.35402822e-01 -2.80097276e-01
5.39087772e-01 -1.71278924e-01 -4.50535238e-01 -2.74712384e-01
-4.71520394e-01 -3.54917645e-02 4.82218862e-01 1.03123057e+00
2.72176385e-01 1.35048777e-01 -3.84519130e-01 1.57953292e-01
-5.95096163e-02 -1.63167611e-01 6.23729050e-01 9.31202412e-01
9.51687805e-03 -1.57775688e+00 1.56334639e-01 7.52512753e-01
3.33155803e-02 -1.99005499e-01 -9.40368474e-01 8.66244614e-01
-3.21591526e-01 7.64597356e-01 -1.56127557e-01 -5.77089190e-01
2.97090590e-01 -9.13897604e-02 7.27157712e-01 -9.31626678e-01
-4.03093398e-01 -7.69344866e-01 4.05019075e-01 -4.86993223e-01
-1.60811562e-02 -3.86646658e-01 -1.21393478e+00 -3.33654732e-01
-3.76481205e-01 1.42570823e-01 4.58655298e-01 6.39565766e-01
7.81436443e-01 5.07219553e-01 8.42878878e-01 -4.34108913e-01
-5.17800689e-01 -6.00274980e-01 -5.44700384e-01 8.93514335e-01
6.28983378e-02 -6.96292639e-01 -2.70713300e-01 1.42723829e-01] | [7.162132740020752, 6.259637355804443] |
15a21c7d-4439-4506-a2f4-09a3405f06f7 | 360o-surface-regression-with-a-hyper-sphere | 1909.07043 | null | https://arxiv.org/abs/1909.07043v1 | https://arxiv.org/pdf/1909.07043v1.pdf | $360^o$ Surface Regression with a Hyper-Sphere Loss | Omnidirectional vision is becoming increasingly relevant as more efficient $360^o$ image acquisition is now possible. However, the lack of annotated $360^o$ datasets has hindered the application of deep learning techniques on spherical content. This is further exaggerated on tasks where ground truth acquisition is difficult, such as monocular surface estimation. While recent research approaches on the 2D domain overcome this challenge by relying on generating normals from depth cues using RGB-D sensors, this is very difficult to apply on the spherical domain. In this work, we address the unavailability of sufficient $360^o$ ground truth normal data, by leveraging existing 3D datasets and remodelling them via rendering. We present a dataset of $360^o$ images of indoor spaces with their corresponding ground truth surface normal, and train a deep convolutional neural network (CNN) on the task of monocular 360 surface estimation. We achieve this by minimizing a novel angular loss function defined on the hyper-sphere using simple quaternion algebra. We put an effort to appropriately compare with other state of the art methods trained on planar datasets and finally, present the practical applicability of our trained model on a spherical image re-lighting task using completely unseen data by qualitatively showing the promising generalization ability of our dataset and model. The dataset is available at: vcl3d.github.io/HyperSphereSurfaceRegression. | ['Stamatis Samaras', 'Dimitrios Zarpalas', 'Dimitrios Ataloglou', 'Petros Daras', 'Antonis Karakottas', 'Vasileios Gkitsas', 'Nikolaos Zioulis'] | 2019-09-16 | null | null | null | null | ['surface-normals-estimation'] | ['computer-vision'] | [ 2.54483879e-01 2.05456868e-01 4.96341318e-01 -4.39715028e-01
-6.35331750e-01 -5.26451170e-01 5.80263674e-01 -2.67434061e-01
-6.22781038e-01 7.27449656e-01 1.35740649e-03 -3.38628262e-01
5.36132284e-05 -7.59873688e-01 -1.06558645e+00 -5.68128586e-01
5.67605942e-02 2.07694829e-01 -2.15405464e-01 -3.10882092e-01
2.32171923e-01 6.30607784e-01 -1.75985932e+00 -1.25417367e-01
7.24468112e-01 1.27190840e+00 8.12906101e-02 6.72349989e-01
3.55043918e-01 3.03547710e-01 -4.41478997e-01 -3.96586150e-01
6.59146726e-01 -1.13867588e-01 -5.10694444e-01 -3.12630422e-02
9.69350696e-01 -4.23554569e-01 -2.64446557e-01 1.00515068e+00
6.62005067e-01 1.74017362e-02 5.68378150e-01 -9.10114288e-01
-5.81212580e-01 -4.72373128e-01 -2.10160792e-01 -4.58714087e-03
6.89253092e-01 -9.48547870e-02 8.55688035e-01 -1.05222344e+00
7.92067826e-01 7.90954709e-01 6.78414702e-01 4.22288984e-01
-8.98599625e-01 -2.71411926e-01 -6.08138107e-02 1.21617712e-01
-1.50381422e+00 -4.27848995e-01 8.80831599e-01 -5.45676947e-01
1.07486868e+00 3.52508217e-01 8.01726460e-01 9.89948928e-01
6.91552311e-02 7.61431754e-01 1.23409379e+00 -4.25550401e-01
2.75859714e-01 -2.89469734e-02 -6.20650470e-01 7.21402884e-01
1.63497910e-01 1.09026149e-01 -4.24297780e-01 3.49735558e-01
1.09119570e+00 -1.92638963e-01 -6.88187480e-01 -1.06677830e+00
-1.27962816e+00 6.18634701e-01 7.65206337e-01 -4.72017601e-02
-1.92963988e-01 8.41183413e-04 8.39681551e-02 1.35725513e-01
6.54481888e-01 6.21830463e-01 -5.74223995e-01 -1.55399472e-01
-6.67948842e-01 2.75477886e-01 5.99844158e-01 1.07195854e+00
7.48871088e-01 8.30679983e-02 5.53072572e-01 7.26495445e-01
4.19435292e-01 7.62555063e-01 1.95440426e-01 -1.11746955e+00
3.83933306e-01 5.70988715e-01 3.01489919e-01 -9.90951359e-01
-5.69316864e-01 -3.60598743e-01 -8.42746258e-01 6.47960007e-01
4.98445630e-01 -1.20794242e-02 -7.37331569e-01 1.55135322e+00
3.79272342e-01 -2.17414051e-01 2.84929574e-02 1.12288964e+00
8.90955567e-01 2.91249275e-01 -6.81699097e-01 1.50788471e-01
9.53238308e-01 -6.67656958e-01 -3.00222993e-01 -1.78340718e-01
4.87302214e-01 -7.54415691e-01 1.17508543e+00 7.69300580e-01
-8.84548187e-01 -2.74555564e-01 -1.21462500e+00 -3.46362501e-01
-4.85201865e-01 2.15900764e-01 6.34102345e-01 6.52935386e-01
-1.29360223e+00 3.78615469e-01 -7.58918166e-01 -4.07063514e-01
3.62477779e-01 3.81864011e-01 -6.93836391e-01 -2.41739586e-01
-8.73876870e-01 8.62852514e-01 -1.81302458e-01 2.68692374e-01
-5.68739593e-01 -8.04933369e-01 -1.32051301e+00 -5.06014884e-01
1.42883137e-01 -8.55263352e-01 1.19124651e+00 -5.48596323e-01
-1.65588546e+00 1.19619799e+00 -1.04347407e-03 -2.80736089e-01
7.53390312e-01 -4.23023850e-01 3.81118767e-02 3.74317281e-02
1.24751497e-02 5.61709106e-01 6.72037244e-01 -1.37516105e+00
-4.60023791e-01 -7.02025056e-01 4.45239156e-01 6.20141089e-01
7.61927143e-02 -5.30538559e-01 -3.66032422e-01 -1.85578674e-01
4.93137240e-01 -9.06204820e-01 -1.50039688e-01 3.21964532e-01
-3.85659188e-01 -3.61923836e-02 3.05114329e-01 -6.68612301e-01
6.44102156e-01 -1.88712740e+00 1.36395663e-01 -5.71289212e-02
7.37298653e-02 1.31640434e-01 2.11279094e-01 1.25609130e-01
-4.50917371e-02 -1.48106307e-01 -2.89469659e-01 -4.70723301e-01
1.33344978e-01 9.88382148e-04 -1.24196582e-01 9.31126237e-01
-5.35855517e-02 7.70920515e-01 -7.94785678e-01 -6.66347072e-02
7.05820918e-01 8.58639240e-01 -7.16523528e-01 2.48844609e-01
2.52724756e-02 5.85899591e-01 4.69249673e-02 7.87228942e-01
7.63752162e-01 -6.32524118e-02 -2.01473191e-01 -2.79353201e-01
-3.83337080e-01 2.90323764e-01 -1.23676372e+00 1.98924279e+00
-7.76409566e-01 7.38993227e-01 2.15817660e-01 -9.63876843e-01
9.46788073e-01 -6.82764426e-02 5.18915474e-01 -1.04085195e+00
4.40206937e-03 4.83777165e-01 -3.45672339e-01 -2.92121053e-01
4.30571854e-01 -1.01647951e-01 9.89303067e-02 -1.52219340e-01
-1.21181369e-01 -1.02243578e+00 -2.59395838e-01 -2.85266906e-01
7.41333663e-01 6.34730875e-01 3.53556752e-01 -2.56828070e-01
5.28029561e-01 -2.21841425e-01 5.83242849e-02 2.26858363e-01
-1.09795183e-01 1.08805335e+00 3.57157774e-02 -6.71515107e-01
-1.34070361e+00 -1.08239150e+00 -6.00062847e-01 4.27889884e-01
2.03457668e-01 -1.38285697e-01 -6.59789681e-01 -2.98559397e-01
-7.06603378e-02 5.74796200e-01 -7.04831660e-01 2.20259234e-01
-5.14362454e-01 -7.86614537e-01 1.35775119e-01 3.83380979e-01
6.48886085e-01 -6.77586019e-01 -9.82370496e-01 -1.33861884e-01
-1.27441734e-01 -1.30820799e+00 6.58563077e-02 3.11710555e-02
-5.19368947e-01 -1.11163676e+00 -9.90188956e-01 -4.06696588e-01
5.97512245e-01 3.16997498e-01 1.25614262e+00 -3.58855277e-01
-4.35584813e-01 7.30865598e-01 -3.70196477e-02 -6.81634307e-01
2.07467660e-01 -4.18332852e-02 1.41365126e-01 -2.23887950e-01
2.23154172e-01 -6.74971759e-01 -1.12238359e+00 2.47981235e-01
-6.71336591e-01 2.25436300e-01 3.15946370e-01 5.53677022e-01
7.58214355e-01 -3.97818148e-01 -4.51409258e-02 -3.82795423e-01
1.13522030e-01 -2.55626470e-01 -9.39375699e-01 -3.06061536e-01
-4.37520474e-01 -1.45377755e-01 5.66799819e-01 3.20739061e-01
-8.22229028e-01 1.26709893e-01 -5.43494642e-01 -2.94564605e-01
-2.26026401e-01 1.88609943e-01 -1.33623227e-01 -3.54253173e-01
7.83877850e-01 2.68192232e-01 5.39170997e-03 -3.29507232e-01
2.16437981e-01 5.40279090e-01 4.70312923e-01 -3.47944230e-01
7.78288066e-01 1.10489047e+00 3.41696590e-01 -1.25671649e+00
-9.23510730e-01 -4.66251463e-01 -7.80604899e-01 -2.12761298e-01
8.92941475e-01 -1.03709865e+00 -8.93339455e-01 5.62861502e-01
-9.55834866e-01 -5.64306617e-01 -3.35439146e-01 5.90738714e-01
-9.72859681e-01 4.55763698e-01 -3.47983018e-02 -7.47706234e-01
-1.74034625e-01 -1.06759000e+00 1.49079919e+00 1.48467813e-02
4.39664759e-02 -1.03162813e+00 2.41869748e-01 5.15036702e-01
3.48932326e-01 6.06312752e-01 4.66929227e-01 -7.58070052e-02
-7.80433655e-01 -1.81132808e-01 -3.10547084e-01 5.74545503e-01
1.24584131e-01 -1.79344773e-01 -1.19769228e+00 -3.93747956e-01
1.28997609e-01 -5.58048308e-01 5.44469655e-01 4.96882290e-01
1.08025205e+00 4.05631214e-02 1.45725057e-01 1.26986229e+00
1.63375497e+00 -2.67673451e-02 6.73090279e-01 7.97340393e-01
8.15151155e-01 6.34141862e-01 4.81853545e-01 4.19703186e-01
8.53180289e-01 8.92029822e-01 1.03596556e+00 -2.91190416e-01
-9.43192765e-02 -4.68064360e-02 -4.50762622e-02 6.58503234e-01
-5.37097573e-01 3.32890218e-03 -1.08128238e+00 5.04399121e-01
-1.25323522e+00 -4.23710614e-01 1.94542129e-02 2.45021486e+00
6.09552860e-01 -4.16950285e-02 -3.38247627e-01 3.20328712e-01
1.22701237e-02 1.89761460e-01 -5.30654550e-01 -2.45470196e-01
-2.59529859e-01 2.71340728e-01 4.80832458e-01 8.07855964e-01
-1.19076431e+00 6.41128659e-01 5.35120201e+00 3.11743975e-01
-1.48521662e+00 -3.11819941e-01 4.85958487e-01 -1.14122733e-01
-3.87401789e-01 -3.81465882e-01 -7.24834442e-01 1.06479026e-01
5.00249922e-01 1.70228019e-01 4.29100126e-01 1.04665077e+00
7.31028169e-02 -3.09074789e-01 -1.19837010e+00 1.52463937e+00
4.96799618e-01 -1.18490767e+00 -2.27096364e-01 1.10002108e-01
8.16615999e-01 4.82026637e-01 1.64380848e-01 -3.94639559e-02
5.11576328e-03 -1.32134950e+00 6.75149262e-01 5.54669201e-01
1.17593479e+00 -4.01943356e-01 6.03178978e-01 3.18332404e-01
-9.16930497e-01 2.97600657e-01 -3.88410330e-01 -2.48770028e-01
-1.74098741e-02 5.31150639e-01 -1.00800455e+00 7.66147017e-01
1.02132559e+00 6.40236199e-01 -5.68492174e-01 1.02827322e+00
-3.15736800e-01 -1.37947813e-01 -5.66245139e-01 -2.30786037e-02
1.04677394e-01 -4.56078202e-01 4.36320901e-01 8.98776531e-01
5.97814441e-01 4.75085154e-02 -3.81525040e-01 6.74874306e-01
7.08597451e-02 2.36703068e-01 -9.92585719e-01 5.17504156e-01
-6.72707930e-02 1.22260869e+00 -3.90315741e-01 1.06039345e-01
-5.16880095e-01 1.14893222e+00 3.44083756e-01 4.28853661e-01
-5.14960170e-01 -3.21386516e-01 7.21843600e-01 2.23366916e-01
1.59541354e-01 -7.27503657e-01 -4.37615216e-01 -1.53035998e+00
3.98071468e-01 -6.35751188e-01 -1.11725517e-01 -1.17642415e+00
-9.11734581e-01 6.43292129e-01 -1.03601262e-01 -1.17560041e+00
-2.62332141e-01 -1.18799365e+00 1.55546125e-02 9.31007802e-01
-1.93739212e+00 -1.08726013e+00 -8.02627623e-01 6.12208247e-01
3.55799198e-01 1.59010395e-01 9.23037231e-01 1.48333013e-01
-7.23944558e-03 3.52899164e-01 1.87583223e-01 2.19963044e-02
5.47687173e-01 -1.56182218e+00 5.12494445e-01 7.48273551e-01
1.02344647e-01 5.59310377e-01 6.96126461e-01 -1.50712982e-01
-1.60771716e+00 -8.27686608e-01 5.87918937e-01 -8.99694264e-01
3.21596295e-01 -6.72117412e-01 -5.45540392e-01 6.67644858e-01
-7.32668862e-02 2.68060476e-01 5.38085818e-01 -2.24810526e-01
-1.90624818e-01 -1.07302934e-01 -1.10590518e+00 7.25660682e-01
1.31937587e+00 -5.48362255e-01 -2.61267245e-01 3.98019701e-01
5.06639838e-01 -9.57822680e-01 -9.28921044e-01 7.17848957e-01
5.13126075e-01 -1.38823354e+00 1.22581184e+00 -4.24842872e-02
5.52784085e-01 -3.08173478e-01 -6.60290897e-01 -1.38031483e+00
2.31565163e-01 -5.12948215e-01 1.12971604e-01 3.93801838e-01
2.04908505e-01 -6.82725310e-01 1.03182375e+00 4.59630430e-01
-4.37967658e-01 -1.12643397e+00 -1.30468142e+00 -5.31816185e-01
1.64431393e-01 -6.88410044e-01 3.84178728e-01 7.95999765e-01
-3.05603445e-01 1.53758982e-02 -6.69229999e-02 2.01230004e-01
7.36663938e-01 6.66483790e-02 1.18295658e+00 -1.09933174e+00
-2.53589060e-02 -1.49851352e-01 -6.62116230e-01 -1.38444078e+00
1.81496441e-02 -6.67093456e-01 -1.27166444e-02 -1.74127889e+00
-4.21183795e-01 -4.25225765e-01 2.65037626e-01 4.98518012e-02
2.54820853e-01 7.76330531e-01 -7.29275048e-02 4.36866209e-02
-4.30326253e-01 8.10913205e-01 1.58485854e+00 2.27390170e-01
9.81456637e-02 -4.49262187e-02 -4.59009975e-01 1.02048922e+00
6.19488716e-01 4.34157521e-01 -4.99813527e-01 -6.89784646e-01
7.93336272e-01 -2.32596740e-01 5.74239552e-01 -1.07365787e+00
8.05680752e-02 1.00922629e-01 6.13804281e-01 -5.18570483e-01
8.49868476e-01 -7.67546296e-01 -1.25000477e-01 1.28040060e-01
2.59496778e-01 -9.90109295e-02 1.51127532e-01 2.93826550e-01
-9.13448259e-02 1.92430392e-01 7.17070103e-01 -2.61733353e-01
-6.71444952e-01 4.49950635e-01 1.28240436e-01 6.91342801e-02
8.57216001e-01 -5.59538543e-01 -8.52990523e-02 -5.52335918e-01
-5.67125142e-01 -1.80902228e-01 9.65977073e-01 2.74049550e-01
8.44079792e-01 -1.25310981e+00 -4.83335316e-01 5.90689242e-01
3.03999692e-01 6.93774819e-01 3.17293927e-02 7.30696321e-01
-1.08475113e+00 7.33386517e-01 -1.61425889e-01 -8.72930884e-01
-8.89882803e-01 2.77898729e-01 7.61444211e-01 1.68685451e-01
-5.77992857e-01 1.03264332e+00 4.60081190e-01 -1.00109720e+00
9.27640274e-02 -5.84629297e-01 -2.42371067e-01 -2.08058238e-01
8.11081976e-02 2.93653309e-01 5.11086047e-01 -7.58222640e-01
-4.75247473e-01 9.98483419e-01 4.78954554e-01 -1.83719471e-01
1.54475141e+00 -2.47448564e-01 -7.49608353e-02 3.83791745e-01
1.40345800e+00 2.17057958e-01 -1.60718179e+00 -1.86902788e-02
-4.24017757e-01 -6.84281528e-01 -6.31850585e-02 -5.79992175e-01
-6.59529865e-01 9.95645404e-01 5.93801796e-01 1.80925541e-02
9.84837770e-01 -1.33580282e-01 4.02614594e-01 4.56442744e-01
6.30545020e-01 -8.70503187e-01 5.43862693e-02 6.29777491e-01
1.27837002e+00 -1.66377485e+00 1.49806157e-01 -5.02356172e-01
-2.22845569e-01 1.09269011e+00 5.23895264e-01 -3.08237880e-01
7.16762483e-01 1.97328627e-02 2.64670193e-01 -3.23554873e-01
-5.50365560e-02 -2.35373452e-02 3.47530633e-01 8.79370391e-01
5.59467435e-01 -5.30431885e-03 3.79876308e-02 1.45759210e-01
-8.26402187e-01 -8.29902217e-02 6.16893828e-01 8.35162103e-01
-1.06709898e-01 -7.45446503e-01 -4.05518711e-01 7.07866326e-02
-4.39399272e-01 -1.27482042e-01 -1.13160066e-01 8.97288561e-01
1.42176986e-01 5.67563951e-01 4.58296351e-02 -6.43864945e-02
5.68814993e-01 -2.76186556e-01 7.54840612e-01 -5.57940304e-01
2.02134147e-01 -2.62139738e-01 9.84959528e-02 -7.60782242e-01
-5.24817646e-01 -7.63603210e-01 -9.13046539e-01 -9.32141244e-02
8.31026286e-02 -2.56316930e-01 1.29242778e+00 6.70473993e-01
1.79857448e-01 2.43092433e-01 5.09379029e-01 -1.43914402e+00
-3.65599990e-01 -7.16179192e-01 -4.06446517e-01 4.27515328e-01
5.91679573e-01 -6.94187284e-01 -5.69370389e-01 -3.15037034e-02] | [8.492130279541016, -2.6757993698120117] |
5caed062-3452-4d84-a3ff-04aee02f35aa | edge-based-video-analytics-a-survey | 2303.14329 | null | https://arxiv.org/abs/2303.14329v1 | https://arxiv.org/pdf/2303.14329v1.pdf | Edge-Based Video Analytics: A Survey | Edge computing has been getting a momentum with ever-increasing data at the edge of the network. In particular, huge amounts of video data and their real-time processing requirements have been increasingly hindering the traditional cloud computing approach due to high bandwidth consumption and high latency. Edge computing in essence aims to overcome this hindrance by processing most video data making use of edge servers, such as small-scale on-premises server clusters, server-grade computing resources at mobile base stations and even mobile devices like smartphones and tablets; hence, the term edge-based video analytics. However, the actual realization of such analytics requires more than the simple, collective use of edge servers. In this paper, we survey state-of-the-art works on edge-based video analytics with respect to applications, architectures, techniques, resource management, security and privacy. We provide a comprehensive and detailed review on what works, what doesn't work and why. These findings give insights and suggestions for next generation edge-based video analytics. We also identify open issues and research directions. | ['Di wu', 'Yipeng Zhou', 'Young Choon Lee', 'Amirmohammad Pasdar', 'Zhenxiao Luo', 'Miao Hu'] | 2023-03-25 | null | null | null | null | ['edge-computing'] | ['time-series'] | [-4.01879460e-01 -4.03991878e-01 -2.09373876e-01 -6.35749176e-02
2.05724849e-03 -5.25979996e-01 2.72686742e-02 7.61829615e-02
-2.94666082e-01 3.70929629e-01 1.01484852e-02 -5.55583179e-01
1.69412732e-01 -7.57156312e-01 -1.34955168e-01 -4.86570060e-01
-2.87835598e-01 2.46917337e-01 5.56188881e-01 -2.11892705e-02
3.36943157e-02 3.34761113e-01 -1.77202260e+00 4.64508623e-01
5.55002928e-01 1.60012686e+00 -2.25343220e-02 8.30321670e-01
-5.23376614e-02 9.10192788e-01 -2.59654492e-01 -5.68187177e-01
3.24522942e-01 -2.77620200e-02 -1.80216134e-01 6.07986972e-02
-5.52189313e-02 -5.77366889e-01 -5.97387850e-01 1.01801574e+00
4.80398715e-01 -8.47451091e-02 -2.67243888e-02 -1.89002430e+00
-3.60770583e-01 -2.76542362e-02 -5.08628368e-01 5.69809675e-01
4.48429823e-01 -8.49049911e-02 6.57514393e-01 -7.33158112e-01
7.63339043e-01 1.95597559e-01 5.82101882e-01 2.95381427e-01
-4.01501387e-01 -3.81677628e-01 1.19439006e-01 9.98421311e-01
-1.47966540e+00 -8.66894305e-01 5.28379440e-01 -2.77757585e-01
9.91772652e-01 6.27462327e-01 8.18964183e-01 4.23419625e-01
1.45435199e-01 7.61409461e-01 5.81171870e-01 -2.96313405e-01
4.95234847e-01 3.05084050e-01 -7.63769522e-02 1.93696246e-01
3.95592481e-01 -4.77017105e-01 -6.94749415e-01 -2.10243419e-01
4.87338722e-01 3.82694900e-01 -3.52612019e-01 -3.42280358e-01
-7.74346650e-01 2.71325529e-01 -4.11515534e-02 2.82767922e-01
-8.93621266e-01 1.46875635e-01 9.31652248e-01 4.22040761e-01
7.97956049e-01 -4.56110895e-01 -4.77204323e-01 -7.51375616e-01
-1.10335886e+00 -2.09629893e-01 8.24414551e-01 1.53104281e+00
5.63344598e-01 -7.22846612e-02 1.66774347e-01 4.61582802e-02
1.14394575e-01 2.04690531e-01 -2.57734880e-02 -9.03365791e-01
3.11307251e-01 4.14064944e-01 4.26977351e-02 -1.02844989e+00
-1.86577871e-01 -1.24857619e-01 -7.24608958e-01 1.53023601e-01
1.24263994e-01 -3.87896657e-01 -4.01072562e-01 1.12964642e+00
4.72043037e-01 5.66972673e-01 -2.36477390e-01 1.02390969e+00
7.18562782e-01 5.43673754e-01 -1.66094154e-01 -5.76209128e-01
1.60828531e+00 -1.08544016e+00 -8.69679987e-01 1.46768361e-01
4.66914296e-01 -7.89612830e-01 5.76098919e-01 3.71921211e-01
-1.24947834e+00 -2.03962699e-02 -5.27385652e-01 1.68097958e-01
-4.05141503e-01 -2.31879979e-01 7.46911347e-01 9.26605523e-01
-1.40219021e+00 1.97812498e-01 -8.43299448e-01 -4.52390581e-01
5.22378266e-01 3.44355017e-01 -3.61689627e-01 -1.61638975e-01
-7.33801842e-01 2.97276437e-01 1.20201753e-02 -2.54599750e-01
-2.85522491e-01 -6.85476601e-01 -2.52957731e-01 4.77706105e-01
8.72684479e-01 -7.56207705e-01 9.47708368e-01 -8.51888418e-01
-1.43465865e+00 7.12871671e-01 -2.50657678e-01 -1.68083176e-01
4.90074694e-01 -1.51617080e-01 -7.64345109e-01 4.47469503e-01
-1.84308514e-01 -2.30531365e-01 6.03098154e-01 -6.38328791e-01
-9.23889160e-01 -5.85333943e-01 -4.96004187e-02 -4.76990193e-02
-6.45346999e-01 5.95436692e-01 -9.21516478e-01 -1.03627428e-01
-2.08023474e-01 -9.24034357e-01 -2.29070231e-01 -4.98608164e-02
-3.30346301e-02 -2.39219982e-02 1.27909303e+00 -4.16249603e-01
1.69250405e+00 -2.33731341e+00 -5.10699034e-01 1.75322756e-01
8.78332376e-01 5.41564882e-01 3.64713967e-01 6.26365066e-01
1.44414663e-01 1.49072379e-01 6.95124865e-01 -7.58993104e-02
-2.51230806e-01 -2.97404170e-01 7.95605928e-02 3.26103956e-01
-6.82660997e-01 6.92006946e-01 -8.96432698e-01 -3.41683894e-01
3.13479543e-01 6.19569719e-01 -7.55714953e-01 5.19308671e-02
1.39076009e-01 3.80873859e-01 -4.70367372e-01 9.33798671e-01
7.69149125e-01 -5.04431844e-01 2.35274836e-01 -1.14282824e-01
-3.48325849e-01 -5.05606011e-02 -9.96154130e-01 1.21890378e+00
-2.95466214e-01 9.94101167e-01 6.46376014e-01 -8.22535574e-01
4.37228560e-01 8.99490297e-01 8.64970803e-01 -5.09629905e-01
2.60157406e-01 3.81232500e-01 -3.01289976e-01 -7.83909023e-01
6.33202672e-01 3.50157738e-01 3.07614207e-01 4.13388371e-01
-5.23810267e-01 7.97249496e-01 1.84480816e-01 3.48588675e-01
1.36489916e+00 -2.63928980e-01 2.15037525e-01 -2.85676438e-02
4.28669840e-01 -1.40617117e-01 6.88913822e-01 2.40428030e-01
-7.61003315e-01 2.66110808e-01 5.01760066e-01 -6.49902761e-01
-1.00949490e+00 -7.25883901e-01 3.45864117e-01 1.17360401e+00
1.62955090e-01 -1.26181579e+00 -9.06456649e-01 -3.95593286e-01
-2.42322952e-01 2.43328214e-01 -8.51580948e-02 1.11145608e-01
-1.29337370e-01 -4.40482169e-01 -4.23384681e-02 4.55712974e-01
4.96530116e-01 -7.13275433e-01 -9.57176745e-01 1.24532826e-01
-2.33736455e-01 -1.53182161e+00 -4.51746434e-01 -4.05965358e-01
-8.83272648e-01 -9.89323318e-01 -4.27906424e-01 -2.60542542e-01
5.66782773e-01 9.60063875e-01 1.26939356e+00 3.63319278e-01
-2.23247752e-01 5.75240791e-01 -6.76026464e-01 -3.86397362e-01
3.08656126e-01 -1.95829421e-01 3.29847902e-01 2.37857446e-01
9.40660357e-01 -8.87101591e-01 -9.03841853e-01 2.51583666e-01
-7.40048170e-01 -7.48092532e-02 2.13863268e-01 2.56586999e-01
4.37815249e-01 4.54594314e-01 5.37285864e-01 -1.00358617e+00
5.11217296e-01 -9.96055305e-01 -5.26998758e-01 1.59892216e-01
-7.23536193e-01 -8.37470353e-01 7.66542077e-01 1.20904036e-02
-6.23827159e-01 -3.50177169e-01 1.24372996e-01 -6.24800026e-01
2.95464583e-02 4.65412021e-01 -2.96260536e-01 -1.92042124e-02
2.16782108e-01 -8.18278790e-02 -3.50950919e-02 -2.29547396e-01
2.32826769e-02 1.20026028e+00 2.41217718e-01 -6.37248158e-02
2.73316771e-01 7.95131028e-01 -1.45787805e-01 -1.06946516e+00
-1.28129855e-01 -8.07795167e-01 -1.10748149e-01 -6.65131629e-01
7.00167477e-01 -9.86552536e-01 -1.19021225e+00 4.04163301e-01
-1.03334534e+00 -7.62414858e-02 -1.31794661e-01 5.01452565e-01
-4.01667029e-01 3.44525069e-01 -7.29730606e-01 -1.04275084e+00
-4.77180451e-01 -1.06787646e+00 7.07324624e-01 4.02977556e-01
-5.21064848e-02 -8.07638347e-01 -3.28574449e-01 5.68536758e-01
9.86415982e-01 5.94348870e-02 2.70088971e-01 -4.22833920e-01
-9.89555955e-01 -6.81283116e-01 -4.23270732e-01 -1.77012570e-02
-2.49176808e-02 2.14138240e-01 -9.07455504e-01 -5.19969285e-01
8.56936425e-02 3.02767366e-01 1.56911835e-01 4.68480349e-01
1.27554572e+00 -1.55567691e-01 -4.61403459e-01 9.05838907e-01
1.54323733e+00 3.19899559e-01 8.45990896e-01 2.00814888e-01
5.25802672e-01 2.61652440e-01 5.98623574e-01 1.05006742e+00
3.63104343e-01 6.26652062e-01 7.50046432e-01 2.75305539e-01
3.02924991e-01 1.52929619e-01 2.62009919e-01 1.12055206e+00
-8.02189887e-01 -5.12626469e-01 -7.20735133e-01 4.10643220e-01
-2.02160001e+00 -1.20223677e+00 -5.11318982e-01 2.37245369e+00
-1.16228789e-01 8.28426108e-02 3.89637440e-01 1.23998009e-01
8.41002524e-01 -1.26192346e-01 -5.21467328e-01 -3.05178404e-01
3.97938043e-01 -2.36859784e-01 7.12636113e-01 -1.86670333e-01
-7.44956374e-01 8.09036732e-01 6.77562571e+00 6.34003103e-01
-1.32190418e+00 4.05066133e-01 7.19111979e-01 -3.60260308e-01
-7.37342685e-02 1.00213245e-01 -4.26314116e-01 9.22483802e-01
1.23037410e+00 -4.84926671e-01 7.13443220e-01 1.40149641e+00
4.71629292e-01 -2.11863473e-01 -8.74868870e-01 1.59828818e+00
-9.26892310e-02 -1.63619387e+00 -4.96025622e-01 5.40319800e-01
6.12598777e-01 3.62362027e-01 -2.19541028e-01 -1.55888751e-01
-3.29506427e-01 -5.46827495e-01 5.56934297e-01 3.75792503e-01
1.09584630e+00 -7.19038308e-01 9.76314783e-01 1.61476985e-01
-1.45831752e+00 -2.77337916e-02 -3.58405650e-01 -6.14693284e-01
5.82041919e-01 1.00499022e+00 -1.56824127e-01 3.57499242e-01
1.22631896e+00 5.52418411e-01 -6.11832105e-02 1.07176268e+00
2.89527088e-01 3.68290454e-01 -4.02613372e-01 -6.50761127e-02
-2.00163141e-01 -5.39075494e-01 3.90869707e-01 9.16570485e-01
5.75917780e-01 3.26122493e-01 1.81219712e-01 2.08120883e-01
-1.91027775e-01 1.52680248e-01 -7.27664948e-01 -1.81857213e-01
9.01703000e-01 1.59421468e+00 -1.01103520e+00 -6.86730027e-01
-1.03786945e+00 1.04173458e+00 -5.85844554e-02 2.89666176e-01
-7.68744349e-01 -3.61011386e-01 1.31820595e+00 5.49478769e-01
2.19820246e-01 -5.44849217e-01 -1.10880904e-01 -1.36533535e+00
2.80019432e-01 -6.39443636e-01 3.39236945e-01 -7.48283267e-01
-1.02370179e+00 8.02581072e-01 -3.78448814e-01 -1.28799069e+00
-9.44320112e-02 -5.62586725e-01 -8.41423631e-01 3.62751484e-01
-1.24053276e+00 -6.73057497e-01 -7.32871711e-01 8.94983292e-01
2.99446791e-01 -2.04406068e-01 6.24180734e-01 1.01764977e+00
-6.07387662e-01 5.68473220e-01 3.66257429e-01 3.83070707e-02
5.50075293e-01 -6.15430653e-01 2.61759639e-01 8.72074902e-01
-1.05661623e-01 3.13818693e-01 5.69635272e-01 -3.93825144e-01
-2.01266503e+00 -5.75221837e-01 7.94287264e-01 -3.23359549e-01
7.33032465e-01 -4.96395648e-01 -5.17796218e-01 7.50343204e-01
1.40097886e-01 6.05425715e-01 1.12687957e+00 -2.51470152e-02
3.45366709e-02 -3.00128788e-01 -1.07148182e+00 7.33860433e-01
1.21582747e+00 -6.48225367e-01 4.90894258e-01 3.12988222e-01
3.44896793e-01 -1.60390049e-01 -7.29054153e-01 -1.14207841e-01
4.33496237e-01 -1.42946541e+00 5.57447314e-01 -5.96751571e-01
2.68369224e-02 -2.90389508e-01 -2.07366824e-01 -9.08794880e-01
-4.33261991e-01 -1.07132804e+00 -4.47906822e-01 1.03550637e+00
-5.77054061e-02 -7.03399003e-01 1.44297278e+00 1.14420593e+00
-2.75557220e-01 -6.53370440e-01 -8.68363619e-01 -7.00994730e-01
-9.57171202e-01 -7.55019724e-01 5.64618647e-01 9.28714216e-01
6.04034305e-01 1.73960626e-01 -3.66103768e-01 -8.43564197e-02
5.61932802e-01 -7.84474909e-02 9.39914882e-01 -1.03698742e+00
-6.39472008e-02 -1.33288413e-01 -1.14061058e+00 -8.25992405e-01
-5.18855631e-01 -5.86134851e-01 -6.27944708e-01 -1.31384647e+00
1.31068587e-01 -4.86261129e-01 -4.58148830e-02 -6.12521730e-02
1.59921031e-02 5.40640891e-01 5.60242355e-01 3.53057265e-01
-1.30630350e+00 5.80605082e-02 7.92519391e-01 6.09483182e-01
-3.40823494e-02 7.14468583e-02 -6.50228083e-01 7.04590976e-01
8.11190009e-01 -1.29850194e-01 -5.39267838e-01 -5.27466595e-01
7.99124241e-01 2.10034564e-01 -8.48136544e-02 -9.89116609e-01
7.37022638e-01 -1.79271638e-01 -8.98561180e-02 -3.04595768e-01
2.77277201e-01 -1.54312885e+00 5.47008693e-01 2.19824180e-01
6.83149159e-01 4.39440906e-01 -1.67622194e-01 5.86280882e-01
-3.51901501e-01 1.55341446e-01 2.38074780e-01 -1.10927738e-01
-9.44898903e-01 7.58912325e-01 -7.37095892e-01 7.20053120e-03
1.67257380e+00 -5.90128183e-01 -2.65493125e-01 -8.21441948e-01
-7.01870918e-01 2.22126409e-01 9.48633075e-01 1.24395095e-01
4.88358140e-01 -1.19162261e+00 -5.16748190e-01 3.11651468e-01
-2.25046165e-02 -3.90261263e-01 5.56800187e-01 1.15217972e+00
-9.39688027e-01 2.79191583e-01 -2.94894397e-01 -4.69859183e-01
-1.65659225e+00 1.13189197e+00 -2.45052084e-01 2.65153572e-02
-7.27173150e-01 4.62162673e-01 -1.46716878e-01 4.75448817e-01
1.80054437e-02 2.72046298e-01 1.20845653e-01 -1.02729619e-01
1.03209424e+00 9.62194979e-01 1.81478396e-01 -4.39110398e-01
-5.82515657e-01 2.35300466e-01 2.16328859e-01 2.02648804e-01
1.35523033e+00 -5.77234685e-01 -3.73265207e-01 1.08657621e-01
1.02442551e+00 3.10579866e-01 -1.02233267e+00 1.15340287e-02
2.11407691e-02 -9.63568091e-01 2.11349100e-01 -6.69728294e-02
-1.68196607e+00 6.83236361e-01 4.18474853e-01 6.68103397e-01
1.49712420e+00 -3.35351340e-02 1.17969739e+00 -1.07314415e-01
8.13819587e-01 -1.47239029e+00 -2.33306244e-01 2.49998763e-01
-1.57028865e-02 -1.13446796e+00 -6.92113265e-02 -9.18715417e-01
-4.24543262e-01 1.06659651e+00 3.16671073e-01 6.73410222e-02
1.16380143e+00 4.12668914e-01 1.20509614e-03 -3.07606161e-01
-9.69401538e-01 -5.49570331e-03 -4.16560054e-01 7.02791154e-01
6.32014275e-01 2.44095847e-01 -3.50055158e-01 5.75681031e-01
1.48319304e-01 4.49744582e-01 6.83198929e-01 1.19058597e+00
-2.25725234e-01 -1.11885452e+00 -3.35872263e-01 6.43178165e-01
-8.48271132e-01 -2.20187098e-01 -6.15670299e-03 2.17217460e-01
-7.20863417e-02 1.25328696e+00 4.91458297e-01 -6.77372694e-01
-8.12160969e-02 -3.74186318e-04 -1.83715597e-01 -2.35961020e-01
-3.43059033e-01 -1.46062210e-01 1.92732424e-01 -8.32508266e-01
-1.28614873e-01 -5.18977225e-01 -1.01914477e+00 -1.29700065e+00
-4.47675049e-01 2.31015533e-01 1.14913714e+00 4.77800101e-01
1.22768295e+00 4.80750710e-01 5.88106036e-01 -6.53863966e-01
9.48032662e-02 -4.53896940e-01 -7.50145972e-01 2.72321105e-01
-8.69641379e-02 -2.18427047e-01 -2.65259892e-01 7.20635429e-02] | [8.401601791381836, -0.4851190149784088] |
23d4d5b5-fc60-49a7-902b-c616cfdffdff | using-human-guided-causal-knowledge-for-more | 2110.04664 | null | https://arxiv.org/abs/2110.04664v1 | https://arxiv.org/pdf/2110.04664v1.pdf | Using Human-Guided Causal Knowledge for More Generalized Robot Task Planning | A major challenge in research involving artificial intelligence (AI) is the development of algorithms that can find solutions to problems that can generalize to different environments and tasks. Unlike AI, humans are adept at finding solutions that can transfer. We hypothesize this is because their solutions are informed by causal models. We propose to use human-guided causal knowledge to help robots find solutions that can generalize to a new environment. We develop and test the feasibility of a language interface that na\"ive participants can use to communicate these causal models to a planner. We find preliminary evidence that participants are able to use our interface and generate causal models that achieve near-generalization. We outline an experiment aimed at testing far-generalization using our interface and describe our longer terms goals for these causal models. | ['Steven Sloman', 'R. Iris Bahar', 'Emily Sheetz', 'Yanqi Liu', 'Semir Tatlidil'] | 2021-10-09 | null | null | null | null | ['robot-task-planning'] | ['robots'] | [ 1.63058308e-03 7.44989038e-01 5.82562573e-03 -5.53658545e-01
-2.38104492e-01 -4.04609621e-01 6.09720051e-01 2.66441882e-01
-2.09134698e-01 1.01431155e+00 5.64101040e-01 -5.09118676e-01
-5.00700355e-01 -7.11788654e-01 -6.86434448e-01 4.21760418e-02
-3.46994847e-01 8.62201929e-01 1.10756300e-01 -3.02169710e-01
5.97856998e-01 5.57981908e-01 -1.57953644e+00 4.35255080e-01
1.16239727e+00 -7.76786655e-02 6.68487430e-01 7.65941799e-01
-1.09979004e-01 8.31909955e-01 -7.64079928e-01 3.90066713e-01
1.21265538e-01 -6.55880213e-01 -1.43963099e+00 -3.12879592e-01
-4.05017845e-02 -3.77568543e-01 -9.72162560e-02 5.94212711e-01
6.66243657e-02 4.11004603e-01 8.65896702e-01 -1.88669705e+00
-9.30498481e-01 8.80641639e-01 4.55669984e-02 -3.55244130e-02
1.34510195e+00 5.73189974e-01 4.50825065e-01 -5.84962033e-02
8.48958075e-01 1.87716591e+00 5.21031559e-01 8.44636381e-01
-1.28429091e+00 -6.81768000e-01 3.89937997e-01 3.05399209e-01
-1.09383333e+00 -2.16858700e-01 3.86501104e-01 -5.31323373e-01
1.32393944e+00 1.19325943e-01 5.98036706e-01 1.11767101e+00
4.98648971e-01 4.76301074e-01 1.14748049e+00 -6.61157012e-01
3.38393211e-01 1.20664932e-01 -8.65008235e-02 5.90199649e-01
3.27268034e-01 4.95514899e-01 -9.24854875e-01 -3.93850565e-01
1.01846468e+00 -6.51942730e-01 -1.47912383e-01 -2.51030862e-01
-1.29498386e+00 6.32078886e-01 7.21183360e-01 4.71044779e-01
-8.37165177e-01 4.58004564e-01 -2.53864229e-01 3.66298556e-01
-2.36530274e-01 1.38186944e+00 -6.64337933e-01 -1.56567752e-01
1.45368934e-01 8.10309112e-01 1.21546769e+00 1.21840823e+00
5.13480008e-01 -1.44004509e-01 7.41120949e-02 2.58524150e-01
5.31824291e-01 3.59214693e-01 4.09257174e-01 -1.50160086e+00
5.58233559e-02 6.93392873e-01 5.65494061e-01 -9.60582435e-01
-8.22508574e-01 3.90597992e-02 1.35263026e-01 4.35400009e-01
2.48316169e-01 -5.22005081e-01 -6.14076853e-01 1.82977688e+00
1.94944322e-01 -1.28128454e-01 4.49551880e-01 9.50880110e-01
4.47522134e-01 5.65825522e-01 6.08883858e-01 -2.59272665e-01
8.85193348e-01 -3.36459994e-01 -5.61017752e-01 -7.29564369e-01
9.29545760e-01 -3.77361566e-01 1.28877044e+00 3.43750209e-01
-8.84384930e-01 -3.60051543e-01 -1.13526392e+00 7.17055425e-02
-1.13528691e-01 -6.74128354e-01 1.18919396e+00 5.00823259e-01
-1.16200864e+00 7.07945883e-01 -7.90677428e-01 -1.10124493e+00
-3.25682834e-02 1.83528468e-01 -2.99065679e-01 -1.61340237e-01
-9.45116401e-01 1.25199902e+00 9.03292358e-01 -1.88387081e-01
-9.04931545e-01 -5.30476928e-01 -8.68848920e-01 -1.85359389e-01
1.18592441e-01 -1.16371679e+00 1.66396666e+00 -7.86924124e-01
-1.38897872e+00 1.21497914e-01 -2.80166715e-01 -1.16593488e-01
1.42141253e-01 -2.37220570e-01 4.57925797e-02 -7.62983635e-02
5.94186723e-01 1.04800642e+00 -1.68697998e-01 -1.62902510e+00
-7.73038507e-01 -3.46908122e-01 1.59406975e-01 4.40952361e-01
-1.08559012e-01 1.06070735e-01 4.09041286e-01 -2.23207265e-01
1.87890619e-01 -8.60300362e-01 -5.44942617e-01 -3.55253905e-01
-1.21014677e-01 -5.54962993e-01 2.94814378e-01 -2.19304532e-01
8.18846643e-01 -1.51315427e+00 2.94589788e-01 1.56893998e-01
-1.62876993e-01 -4.53814000e-01 -3.74184489e-01 6.03628933e-01
3.20033953e-02 4.51089770e-01 2.47725993e-01 3.99573892e-01
1.69273853e-01 3.57248366e-01 -1.76058829e-01 -8.32384303e-02
1.45929679e-01 7.65837133e-01 -1.30579138e+00 -3.49941581e-01
3.96317020e-02 -3.88838612e-02 -7.48399973e-01 4.76700127e-01
-9.12721872e-01 5.36334932e-01 -7.32566714e-01 1.16071291e-01
4.09393571e-02 1.07164517e-01 4.89288539e-01 5.04284263e-01
-3.68051410e-01 5.85914731e-01 -1.22190511e+00 1.81661963e+00
-5.19788086e-01 3.90474558e-01 -2.87119538e-01 -3.74089450e-01
8.19312990e-01 5.45269191e-01 2.38421652e-02 -3.99926513e-01
7.44037479e-02 2.32930705e-01 2.28672668e-01 -8.97078216e-01
3.02112311e-01 -3.19441110e-01 -2.18354061e-01 8.66015434e-01
-2.31211931e-01 -7.70738482e-01 -3.60339247e-02 2.37664998e-01
1.39051294e+00 3.72675002e-01 2.83084989e-01 -5.60736418e-01
-1.86156690e-01 8.82820964e-01 5.51399112e-01 1.08379650e+00
-1.81141384e-02 4.49820720e-02 5.05299032e-01 -7.63490260e-01
-6.91590607e-01 -7.93292642e-01 4.16703939e-01 1.06522608e+00
2.84417182e-01 -4.29930270e-01 -4.98287976e-01 -4.35862243e-01
-3.04233938e-01 1.93584812e+00 -4.65559334e-01 -2.65977591e-01
-4.23356861e-01 -3.56915206e-01 3.37962985e-01 5.99641442e-01
3.58279139e-01 -1.60928822e+00 -1.37019134e+00 2.52709091e-01
-2.87358731e-01 -6.81670249e-01 -1.35259321e-02 2.17745379e-01
-9.92012739e-01 -1.05547559e+00 1.51995301e-01 -6.68802679e-01
5.92409492e-01 2.48722255e-01 8.33457708e-01 2.48798743e-01
-2.56847918e-01 8.38182628e-01 -4.21924412e-01 -8.75373185e-01
-7.31281281e-01 -1.49269849e-01 1.88776404e-01 -1.17042005e+00
3.53394628e-01 -6.84499443e-01 -1.93899766e-01 4.93853867e-01
-5.49631238e-01 3.76675278e-01 5.66746593e-01 3.64879638e-01
-2.45776877e-01 4.21769470e-01 5.52025795e-01 -3.84439051e-01
1.35818958e+00 -7.28181720e-01 -3.87972742e-01 2.85741687e-01
-7.97279775e-01 4.34566081e-01 2.34125793e-01 -4.26750928e-01
-1.50870419e+00 3.12176287e-01 4.34713632e-01 2.47432441e-01
-6.80703104e-01 7.49578238e-01 -1.68304548e-01 2.28186041e-01
1.41691053e+00 -5.23460329e-01 -5.91367967e-02 -1.13774568e-01
7.57566214e-01 4.48791891e-01 4.44807231e-01 -1.33823061e+00
5.49991071e-01 -1.47407919e-01 -9.82735902e-02 -4.60283190e-01
-3.70500982e-01 7.43177310e-02 -4.31378722e-01 -3.17135334e-01
8.35807264e-01 -4.62485224e-01 -1.17953432e+00 -9.61685255e-02
-1.50809813e+00 -1.00477624e+00 -7.09810928e-02 7.12123990e-01
-1.02133453e+00 -3.96622777e-01 -1.58829227e-01 -9.33239281e-01
3.15910041e-01 -9.18144584e-01 4.26761210e-01 5.51518202e-01
-1.17890942e+00 -8.42554688e-01 1.45524293e-01 6.29006177e-02
3.91146392e-01 8.88730213e-02 1.29642260e+00 -5.49917758e-01
-5.16573608e-01 2.88858980e-01 -1.11017190e-01 -6.17725074e-01
1.98002812e-02 -3.73232007e-01 -4.92587268e-01 2.48521551e-01
-9.72174555e-02 -4.50337768e-01 -1.10108890e-02 2.75241792e-01
4.62878406e-01 -5.18058121e-01 -9.25482571e-01 -1.81884393e-01
9.15307999e-01 7.68904626e-01 5.31733871e-01 5.58881581e-01
2.28308082e-01 1.06252170e+00 8.89127493e-01 1.56144142e-01
7.65657544e-01 6.48806095e-01 9.82704684e-02 4.37306315e-01
6.45560324e-02 -4.97309774e-01 2.07009763e-01 -3.35542679e-01
-1.61980093e-01 -3.17251205e-01 -1.49164426e+00 7.12356269e-01
-2.21720243e+00 -8.84075046e-01 -3.27707648e-01 1.65621126e+00
9.32209909e-01 -6.46879822e-02 -6.49696290e-02 -3.83561909e-01
4.15468961e-01 -8.19076836e-01 -4.74520177e-01 -6.98394954e-01
5.75414658e-01 1.49606928e-01 1.64215803e-01 9.21542466e-01
-4.93248075e-01 1.23630476e+00 7.90895939e+00 -1.40390828e-01
-3.23601961e-01 -2.12333992e-01 3.40740532e-01 1.81170166e-01
-6.95799112e-01 3.62186253e-01 -2.85419941e-01 -3.94903898e-01
9.10611868e-01 -5.26331842e-01 5.63876748e-01 7.70160437e-01
5.06706834e-01 -4.30864692e-01 -1.47892559e+00 1.57353818e-01
-2.10243821e-01 -1.07196295e+00 9.11436602e-02 -2.42977738e-01
5.05552649e-01 -4.92737204e-01 -3.52524877e-01 2.71227151e-01
1.32479203e+00 -1.39520133e+00 7.68175662e-01 4.85266656e-01
1.48312032e-01 -6.65542722e-01 4.21612740e-01 7.05326915e-01
-5.22859097e-01 -3.02533120e-01 -2.33170688e-01 -9.72795248e-01
1.29451886e-01 -2.17965052e-01 -1.63757169e+00 2.84812689e-01
7.15784729e-01 1.02632277e-01 -2.69204319e-01 1.05177355e+00
-7.51706362e-01 8.47087726e-02 -2.74088651e-01 -5.18312633e-01
-1.61729813e-01 2.39838585e-01 6.08224392e-01 7.75529802e-01
5.33297956e-01 8.53424072e-01 3.24028730e-01 1.45009160e+00
7.41994917e-01 -2.54950583e-01 -9.77070093e-01 -1.75028443e-01
7.34159946e-01 4.38636571e-01 -7.16506600e-01 -2.70546451e-02
1.53093621e-01 8.02905738e-01 1.69558972e-01 5.51551223e-01
-3.79856914e-01 -2.80601472e-01 6.57658160e-01 -3.54705304e-02
-5.54000974e-01 -4.63818610e-01 -7.28529632e-01 -5.89209616e-01
-4.97230440e-01 -9.67460394e-01 2.34203458e-01 -1.54231465e+00
-1.09909368e+00 4.55760151e-01 7.35179424e-01 -2.58915424e-01
-6.80392861e-01 -4.26043987e-01 -5.11400700e-01 1.07562673e+00
-6.78414285e-01 -9.64895546e-01 -5.66210926e-01 3.42316061e-01
5.35728097e-01 -8.04040730e-02 1.12939036e+00 -5.91796577e-01
-1.71194851e-01 -8.56036916e-02 -1.07954752e+00 -4.82853055e-01
6.30515635e-01 -9.87505734e-01 3.98561776e-01 7.15347767e-01
-2.64629811e-01 1.20797944e+00 1.34223378e+00 -1.40428627e+00
-1.31894517e+00 -5.02116501e-01 8.42975080e-01 -6.66471601e-01
4.20110881e-01 8.47092941e-02 -7.84429908e-01 1.22464156e+00
5.27640045e-01 -8.33461821e-01 4.94669169e-01 4.72277641e-01
-3.49521786e-01 5.57591617e-01 -1.26865649e+00 1.02007270e+00
1.53437603e+00 4.90973052e-03 -1.21622562e+00 3.26772720e-01
1.00480247e+00 -2.07165256e-01 -5.02264380e-01 3.70682687e-01
3.52536947e-01 -7.72671044e-01 7.37075388e-01 -9.59285021e-01
4.73517329e-01 -4.64409769e-01 -7.78671578e-02 -1.89791536e+00
-5.50785899e-01 -9.29232717e-01 7.58778155e-01 8.99353206e-01
6.70841634e-01 -7.65315413e-01 5.12886643e-01 1.58354616e+00
-1.78184047e-01 3.40410084e-01 -4.24408495e-01 -6.56443954e-01
2.24576488e-01 -8.02581787e-01 7.15047777e-01 9.83180881e-01
9.82209325e-01 5.43574154e-01 3.81890327e-01 5.41678369e-01
5.09777606e-01 -1.68857157e-01 7.76758492e-01 -1.31093442e+00
-2.54007936e-01 -4.35840636e-01 -2.59683486e-02 -6.45459056e-01
3.85498255e-01 -6.30538046e-01 6.47766113e-01 -2.09938884e+00
1.01418950e-01 -9.30866778e-01 4.10638362e-01 9.02398765e-01
-1.24866828e-01 -6.92343354e-01 2.58470297e-01 1.77628443e-01
-2.24726528e-01 2.77457684e-01 9.39214706e-01 1.78954348e-01
-7.21706271e-01 -4.46683913e-01 -1.20868361e+00 9.17976439e-01
1.07640982e+00 -6.92279875e-01 -8.62165868e-01 -5.99996090e-01
5.41606009e-01 1.38479024e-01 5.03433347e-01 -1.13214397e+00
5.70638597e-01 -1.08103240e+00 4.87061828e-01 2.34928846e-01
-1.87249020e-01 -7.67719686e-01 5.65194309e-01 6.88661158e-01
-5.93958616e-01 1.25336885e-01 8.08069885e-01 3.14986408e-01
4.34206009e-01 -4.76720899e-01 1.90063156e-02 -3.08581650e-01
-8.52776289e-01 -5.95290124e-01 -1.06631434e+00 -3.03457260e-01
1.27028048e+00 -6.14449345e-02 -4.91000712e-01 -6.65888727e-01
-4.52062100e-01 6.83588028e-01 5.72345376e-01 6.37787879e-01
7.93592513e-01 -1.01409042e+00 -5.29410779e-01 -1.74823090e-01
1.29578531e-01 -8.02905932e-02 -3.39820445e-01 1.80587426e-01
-6.22536838e-01 5.16681731e-01 -5.32975316e-01 -1.82302415e-01
-7.56151557e-01 6.07323945e-01 4.41948086e-01 5.05934119e-01
-5.54057837e-01 8.71324301e-01 2.42637098e-01 -6.21178806e-01
1.66814372e-01 -2.22046241e-01 -1.94936499e-01 -5.71347177e-01
5.95184684e-01 2.07245514e-01 -5.82546353e-01 1.79734647e-01
-4.18869376e-01 1.71799719e-01 2.96363980e-01 -8.42780054e-01
1.36607528e+00 -9.57374573e-02 -2.57677883e-01 3.69347960e-01
2.57394165e-01 -4.74775851e-01 -1.13074827e+00 2.03611180e-01
2.31231764e-01 -5.51320434e-01 -1.97577566e-01 -1.32209063e+00
-7.81199411e-02 3.09577465e-01 2.31285125e-01 8.83873105e-02
6.87182605e-01 3.03896993e-01 7.09867552e-02 8.06126356e-01
9.36977386e-01 -7.66138792e-01 3.55226338e-01 4.83376384e-01
1.44958556e+00 -8.81653607e-01 1.93280745e-02 -4.49336201e-01
-7.39277124e-01 1.25571501e+00 1.06906724e+00 1.91749990e-01
2.28081733e-01 4.00782526e-01 1.73188984e-01 -4.98113900e-01
-1.15510190e+00 -5.28449230e-02 -1.55320972e-01 1.30032277e+00
4.96481031e-01 2.41660357e-01 -3.66124153e-01 4.63582546e-01
-5.40601671e-01 4.39088345e-01 8.91159534e-01 1.08512640e+00
-5.95303416e-01 -1.07606041e+00 -7.30583310e-01 1.34645645e-02
4.15854663e-01 2.86309719e-01 -9.33119833e-01 7.61685193e-01
-8.27968568e-02 1.71327531e+00 -2.75709741e-02 -4.33391541e-01
4.70572501e-01 1.97156936e-01 7.37250447e-01 -9.38752472e-01
-3.80824447e-01 -4.24929500e-01 4.01918590e-01 -9.60196793e-01
-1.79683656e-01 -9.57176745e-01 -1.90051901e+00 -2.43344516e-01
2.72481501e-01 3.24111611e-01 5.77227533e-01 8.34080458e-01
3.91588956e-01 6.29274309e-01 -9.51769799e-02 -7.18429863e-01
-8.86665359e-02 -9.75374460e-01 1.54498041e-01 1.68020561e-01
-2.59718567e-01 -6.97090745e-01 -1.55654460e-01 2.48362973e-01] | [4.394287586212158, 1.103908896446228] |
5f5e5c35-7d83-4faa-9316-4f6a2262e3e2 | sketching-without-worrying-noise-tolerant | 2203.14817 | null | https://arxiv.org/abs/2203.14817v1 | https://arxiv.org/pdf/2203.14817v1.pdf | Sketching without Worrying: Noise-Tolerant Sketch-Based Image Retrieval | Sketching enables many exciting applications, notably, image retrieval. The fear-to-sketch problem (i.e., "I can't sketch") has however proven to be fatal for its widespread adoption. This paper tackles this "fear" head on, and for the first time, proposes an auxiliary module for existing retrieval models that predominantly lets the users sketch without having to worry. We first conducted a pilot study that revealed the secret lies in the existence of noisy strokes, but not so much of the "I can't sketch". We consequently design a stroke subset selector that {detects noisy strokes, leaving only those} which make a positive contribution towards successful retrieval. Our Reinforcement Learning based formulation quantifies the importance of each stroke present in a given subset, based on the extent to which that stroke contributes to retrieval. When combined with pre-trained retrieval models as a pre-processing module, we achieve a significant gain of 8%-10% over standard baselines and in turn report new state-of-the-art performance. Last but not least, we demonstrate the selector once trained, can also be used in a plug-and-play manner to empower various sketch applications in ways that were not previously possible. | ['Yi-Zhe Song', 'Tao Xiang', 'Pinaki Nath Chowdhury', 'Aneeshan Sain', 'Abdullah Faiz Ur Rahman Khilji', 'Subhadeep Koley', 'Ayan Kumar Bhunia'] | 2022-03-28 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Bhunia_Sketching_Without_Worrying_Noise-Tolerant_Sketch-Based_Image_Retrieval_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Bhunia_Sketching_Without_Worrying_Noise-Tolerant_Sketch-Based_Image_Retrieval_CVPR_2022_paper.pdf | cvpr-2022-1 | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 2.06748605e-01 -4.57343161e-02 -8.28548074e-02 -1.23109315e-02
-1.06057274e+00 -8.47698569e-01 7.74836957e-01 -1.39272377e-01
-1.83139771e-01 2.21529201e-01 3.51371169e-01 -2.33271420e-01
-1.80825174e-01 -7.06613362e-01 -5.46920896e-01 -6.02677345e-01
3.43783289e-01 2.38690168e-01 2.77188029e-02 -4.41198528e-01
7.28317618e-01 7.19112456e-01 -1.47814918e+00 4.17798191e-01
2.46768877e-01 7.32453942e-01 1.63070485e-01 7.00381875e-01
-3.57968360e-01 7.26264656e-01 -8.79203856e-01 -6.54081225e-01
1.24058016e-01 -2.19788477e-02 -4.75170940e-01 -2.55759805e-01
8.81957114e-01 -9.81375754e-01 -4.33131307e-01 6.68334961e-01
6.94921672e-01 -6.82512149e-02 6.94884777e-01 -1.11757469e+00
-6.77086592e-01 3.82500231e-01 -4.77427155e-01 -1.30535662e-01
4.71399486e-01 2.00206608e-01 1.23371708e+00 -1.09083712e+00
9.94474828e-01 1.19126260e+00 4.59308356e-01 7.31564164e-01
-1.14202034e+00 -9.68582869e-01 2.30609864e-01 -3.76408935e-01
-1.24790919e+00 -5.52388489e-01 1.07728302e+00 -1.80045217e-01
8.34856749e-01 4.76778299e-01 5.53278267e-01 1.34927893e+00
-8.81473795e-02 1.44427323e+00 8.60028088e-01 -3.62316310e-01
7.03550875e-02 1.48872197e-01 -1.73983481e-02 5.01272202e-01
-6.81061074e-02 -5.02277873e-02 -6.26215041e-01 -2.15914577e-01
9.12121892e-01 1.19776912e-01 -2.44480576e-02 -4.22801614e-01
-1.04162526e+00 6.51593566e-01 3.55032921e-01 3.49644303e-01
-3.19294661e-01 6.24641240e-01 2.65038162e-01 4.28411901e-01
3.72585982e-01 4.81062382e-01 -2.46007089e-02 -3.04304957e-01
-1.40672028e+00 4.98120099e-01 7.75134027e-01 8.88431609e-01
4.19558287e-01 -4.83767092e-02 -4.65841025e-01 8.78956378e-01
-1.15100360e-02 6.13800049e-01 -4.78336029e-02 -8.37860823e-01
2.16247678e-01 3.68521333e-01 3.84016007e-01 -1.10913014e+00
1.41769916e-01 -3.23649794e-01 -7.56667912e-01 4.92244095e-01
1.98646516e-01 2.02685222e-01 -7.96152174e-01 1.69673789e+00
-1.73663259e-01 -4.62053865e-02 -3.48330468e-01 8.87708187e-01
7.60288060e-01 4.45703536e-01 1.64582297e-01 3.02394241e-01
1.30141044e+00 -7.58796334e-01 -7.06686914e-01 -9.30835307e-03
1.31924957e-01 -1.27927113e+00 1.55085003e+00 7.66341686e-01
-1.09478748e+00 -3.73874187e-01 -1.20875609e+00 -6.40346780e-02
-4.81400102e-01 4.80285943e-01 7.09227562e-01 7.60121763e-01
-1.12997270e+00 8.39112997e-01 -1.76532239e-01 -3.59272838e-01
7.19499469e-01 2.97705472e-01 -3.65676463e-01 -2.13329464e-01
-7.33001828e-01 8.25194418e-01 -4.71017241e-01 -1.19714938e-01
-5.96191764e-01 -7.65579104e-01 -2.01248184e-01 1.79997846e-01
6.51061535e-01 -7.33753622e-01 1.31094003e+00 -8.92715633e-01
-1.43706191e+00 6.10130250e-01 -2.58970618e-01 5.91560863e-02
8.46216023e-01 -6.69986367e-01 -1.36985570e-01 3.65722269e-01
2.74872929e-02 8.87882411e-01 1.32550347e+00 -1.58987343e+00
-1.65814131e-01 -5.94936125e-02 1.26049444e-01 8.69976804e-02
-4.79198039e-01 1.16446398e-01 -8.16831172e-01 -1.10629559e+00
-1.14444911e-01 -7.84852564e-01 2.23018043e-02 4.93760467e-01
-3.59337091e-01 -3.28239799e-01 9.29224908e-01 -4.50906843e-01
1.43702841e+00 -2.07921696e+00 -8.99890661e-02 4.60802317e-01
2.53510892e-01 4.86374438e-01 -4.35918421e-01 9.27892864e-01
2.53030956e-01 4.41085279e-01 3.74762528e-02 -3.87877226e-01
2.20563903e-01 6.58133999e-02 -9.34475005e-01 -6.22301269e-03
6.62916303e-01 1.00385642e+00 -9.57773745e-01 -2.66644686e-01
3.56462747e-01 6.95991993e-01 -4.50550705e-01 1.99281707e-01
-1.96471915e-01 -1.28401190e-01 -4.75481927e-01 7.18428731e-01
6.86735213e-01 -8.12298879e-02 5.74283376e-02 -2.24313781e-01
6.45665526e-02 7.03295246e-02 -1.21264505e+00 1.80000484e+00
-3.65599334e-01 5.93533874e-01 1.32489175e-01 -4.09742653e-01
9.49315190e-01 2.24287763e-01 1.92492053e-01 -9.70703959e-01
-2.00022474e-01 1.54756531e-01 -4.10640836e-01 -3.50334078e-01
9.31262612e-01 -1.74670056e-01 -6.40777200e-02 7.38709092e-01
-2.87398472e-02 -3.30063522e-01 3.68858352e-02 7.49200523e-01
1.25018954e+00 3.52495760e-01 -1.36762902e-01 -4.99340594e-02
1.94296136e-01 -3.54992121e-01 -2.51881063e-01 1.21562874e+00
1.88126904e-03 9.95234072e-01 7.12960303e-01 -2.07886174e-01
-9.94890630e-01 -1.08218765e+00 4.58144307e-01 1.33658218e+00
2.09088162e-01 -5.39640427e-01 -5.12554884e-01 -6.26154900e-01
1.65442035e-01 7.08698452e-01 -6.54157043e-01 -1.04884893e-01
-5.10388970e-01 1.15118651e-02 8.17778468e-01 4.71161276e-01
1.55003190e-01 -1.37655354e+00 -6.43644631e-01 8.54494050e-02
1.97972029e-01 -6.32913053e-01 -6.28092289e-01 -4.92139310e-02
-6.97444201e-01 -8.43390167e-01 -1.03228164e+00 -3.93800408e-01
5.43548763e-01 7.23639131e-01 1.30844235e+00 6.83806539e-01
-4.30936575e-01 7.01774597e-01 -3.76521081e-01 -4.12230074e-01
-7.59375542e-02 2.14331411e-02 -4.43272948e-01 -2.14411035e-01
8.96712020e-02 -5.60749710e-01 -9.49202299e-01 1.31224319e-01
-1.14846885e+00 -9.48736072e-02 9.56745684e-01 8.03185105e-01
1.45127848e-01 -3.46832633e-01 6.47245288e-01 -1.03539848e+00
1.09446204e+00 -1.24625571e-01 -2.45397519e-02 3.37592185e-01
-5.07632554e-01 1.39513895e-01 3.35694492e-01 -7.07435668e-01
-8.26464891e-01 -1.56757936e-01 -2.67119169e-01 -5.78683913e-01
-2.35528182e-02 6.15401641e-02 -3.38510461e-02 7.99448937e-02
4.02923226e-01 1.21095508e-01 2.67269667e-02 -6.26002491e-01
8.75175297e-01 4.02339935e-01 3.19103420e-01 -7.87483633e-01
9.62533057e-01 5.43029010e-01 -1.97693393e-01 -8.97502840e-01
-3.16602319e-01 -6.18753016e-01 -2.30611116e-01 -3.25019807e-01
2.01226830e-01 -8.09648693e-01 -7.63644516e-01 1.82691097e-01
-1.26179576e+00 -3.07344705e-01 -1.77537292e-01 -3.23033899e-01
-2.91041642e-01 4.08727229e-01 -4.34500754e-01 -1.13509119e+00
-6.42390966e-01 -9.73287642e-01 1.36004424e+00 1.86372429e-01
-5.40857852e-01 -4.55708295e-01 -1.90887019e-01 1.24190167e-01
9.11377311e-01 -2.02328354e-01 9.92363334e-01 -3.38515162e-01
-8.23606193e-01 -4.43965018e-01 -6.82412922e-01 2.07449645e-01
-1.78266600e-01 2.82781094e-01 -1.21508014e+00 -2.33755752e-01
-6.48576021e-01 -3.72807056e-01 1.02308464e+00 -4.93998192e-02
1.00835919e+00 -2.48102412e-01 -1.63462058e-01 3.27173364e-03
1.42110193e+00 -8.16429332e-02 9.96484220e-01 1.49611644e-02
4.62393492e-01 3.86058837e-01 3.70002329e-01 3.91530961e-01
1.14124445e-02 7.15364695e-01 2.18309835e-01 -2.28463396e-01
-5.98345280e-01 -6.05720341e-01 1.93364292e-01 4.83010292e-01
1.96630463e-01 -3.31550658e-01 -4.19807315e-01 5.18633842e-01
-1.68105471e+00 -1.05172801e+00 2.71485150e-01 2.11258912e+00
6.45139694e-01 1.98235452e-01 9.03939232e-02 1.03846468e-01
1.78520471e-01 3.93891186e-01 -2.52359778e-01 -4.64719474e-01
-2.99549066e-02 8.60309243e-01 1.68087453e-01 4.82011169e-01
-9.78687048e-01 1.29501128e+00 6.10898972e+00 1.10694659e+00
-1.28122449e+00 -3.28957677e-01 4.02308404e-01 -1.99585140e-01
-6.86661005e-01 1.61786661e-01 -3.33451033e-01 2.50633597e-01
2.23113537e-01 3.41122359e-01 6.45104349e-01 7.33147860e-01
4.43003001e-03 -2.46602491e-01 -9.37442362e-01 9.17668760e-01
3.82036567e-02 -1.26975417e+00 5.50228715e-01 -1.12206414e-01
3.17159384e-01 -4.83527064e-01 3.01705688e-01 4.18323666e-01
1.41269788e-01 -1.09591174e+00 9.42753494e-01 7.63444602e-01
9.13273036e-01 -6.64260805e-01 3.90344411e-01 2.44143471e-01
-9.58873987e-01 2.11088717e-01 -2.76503712e-01 4.60689887e-02
-1.93235159e-01 5.07135153e-01 -7.65342057e-01 3.70077491e-01
3.81687522e-01 3.09103608e-01 -6.43199146e-01 9.19186831e-01
-3.33664477e-01 4.97510016e-01 -3.01337659e-01 -3.41019690e-01
2.10936129e-01 4.05266397e-02 4.77218419e-01 1.47399569e+00
3.29212308e-01 7.63363717e-03 -1.11836545e-01 9.25763249e-01
-3.34144413e-01 1.91834688e-01 -7.98534513e-01 -3.20869923e-01
5.73664665e-01 1.26166749e+00 -7.77333260e-01 -5.28030038e-01
-2.54422277e-01 1.35413635e+00 2.67833680e-01 4.78778899e-01
-3.75436425e-01 -6.77317023e-01 5.15393317e-01 1.87071994e-01
5.55073440e-01 -1.54600255e-02 -4.67644304e-01 -1.00811172e+00
2.34488219e-01 -1.00868034e+00 -1.26572520e-01 -1.04255760e+00
-1.32701850e+00 2.69933552e-01 -3.60460311e-01 -9.09064770e-01
-9.45256427e-02 -3.69828820e-01 -8.94635320e-01 9.06351745e-01
-1.53366375e+00 -1.41094112e+00 -2.49470994e-01 2.20703155e-01
6.31315947e-01 -1.63328405e-02 1.00442493e+00 2.86728680e-01
-2.08141923e-01 7.42713809e-01 -1.01917163e-01 1.54816970e-01
1.09958899e+00 -1.16441905e+00 5.25652528e-01 5.00582576e-01
5.22614479e-01 1.04686224e+00 6.25441015e-01 -6.29431486e-01
-1.79219890e+00 -5.35208702e-01 9.65622187e-01 -6.28981233e-01
4.60114181e-01 -4.06848103e-01 -7.80312181e-01 4.14538346e-02
2.20148668e-01 -2.33770773e-01 3.16276371e-01 8.25327635e-02
-8.63136828e-01 4.95097926e-03 -1.01135015e+00 9.21680391e-01
9.22424376e-01 -9.04170930e-01 -4.54786360e-01 -7.88101330e-02
3.81902248e-01 -9.53262765e-03 -2.76174515e-01 2.50102617e-02
1.28597951e+00 -1.15122950e+00 1.38162136e+00 -7.01753318e-01
7.75352657e-01 -7.54196793e-02 -1.06708094e-01 -8.87563467e-01
-7.93545619e-02 -8.25847864e-01 -3.63715366e-02 1.41026461e+00
8.54295120e-02 6.89330548e-02 1.07390320e+00 6.61885083e-01
2.74575442e-01 -7.54985392e-01 -5.95572054e-01 -4.56322730e-01
1.64908797e-01 -5.65028787e-01 4.71338987e-01 5.96369088e-01
-2.35949442e-01 2.56699145e-01 -5.81269979e-01 -2.64324635e-01
4.60147828e-01 7.38777667e-02 1.15350926e+00 -1.07778382e+00
-1.41147912e-01 -8.73889267e-01 1.49076730e-01 -1.20161867e+00
-2.15451553e-01 -6.08568430e-01 5.39199039e-02 -1.53001511e+00
2.93087095e-01 -5.27939677e-01 -3.37478548e-01 5.14425457e-01
-2.72531152e-01 4.96280193e-01 6.36044443e-01 3.10323507e-01
-6.71329677e-01 2.09851205e-01 1.23964179e+00 -3.16958606e-01
-1.30749092e-01 -1.54525459e-01 -1.03141379e+00 2.82142282e-01
6.04796767e-01 -2.99021959e-01 -2.93790251e-01 -2.62310296e-01
4.47380602e-01 -2.08240561e-02 6.00077689e-01 -7.55249381e-01
1.79421708e-01 1.45037249e-01 4.39268649e-01 -6.12862468e-01
5.19274712e-01 -8.17634284e-01 -3.50999236e-02 2.32157767e-01
-5.35755694e-01 -1.47388309e-01 2.77173489e-01 6.22619748e-01
1.04123071e-01 -2.58034050e-01 2.90417194e-01 -1.79678515e-01
-4.91094381e-01 -7.24509954e-02 -3.89886141e-01 -2.43439928e-01
3.32951277e-01 -3.43628675e-02 -2.16186672e-01 -7.99453497e-01
-3.00932407e-01 -1.49252772e-01 4.01185393e-01 5.68434000e-01
8.20021570e-01 -1.27914381e+00 -4.14879799e-01 1.60598665e-01
1.07816048e-01 -4.97338116e-01 1.72269493e-01 2.65201688e-01
-2.36352637e-01 2.55965829e-01 -1.22547716e-01 -1.34942785e-01
-1.38291597e+00 4.71080452e-01 -3.62563938e-01 -3.36419225e-01
-7.32298493e-01 7.87118018e-01 -1.00036360e-01 -9.27079394e-02
5.94803631e-01 -2.34814808e-01 1.03143886e-01 3.00640345e-01
5.10280132e-01 2.38234475e-01 7.26058409e-02 -1.51061296e-01
-2.77291179e-01 5.80253303e-01 -3.08297843e-01 -3.88040245e-01
1.27312291e+00 2.80391276e-01 1.50775641e-01 1.60668507e-01
1.15947485e+00 3.40269566e-01 -1.08860123e+00 -6.64922968e-02
-6.11830354e-02 -7.18251288e-01 -5.23544811e-02 -1.33451617e+00
-9.08445239e-01 1.18022048e+00 5.30132234e-01 8.36513788e-02
8.02106142e-01 -2.34827220e-01 9.17232096e-01 6.46995366e-01
2.40244046e-01 -1.03886855e+00 5.23755491e-01 1.21946432e-01
1.11113620e+00 -1.14995289e+00 2.75440902e-01 -1.83345661e-01
-6.87349439e-01 1.27316070e+00 9.70139056e-02 -5.73739886e-01
4.15311456e-01 3.85252595e-01 9.81667563e-02 -3.35685372e-01
-7.63170362e-01 -1.45188689e-01 3.58414173e-01 4.11769837e-01
5.04162014e-01 1.18917629e-01 -2.19084606e-01 6.04861259e-01
1.96309850e-01 3.03037763e-01 1.24707513e-01 1.05103540e+00
-4.13959652e-01 -1.50899804e+00 -2.54312456e-01 5.70341170e-01
-1.92712575e-01 -1.81878060e-01 -9.28602636e-01 8.87949407e-01
-1.49228588e-01 6.97227597e-01 -1.87176719e-01 -4.22977716e-01
5.28292239e-01 3.03701758e-01 6.37052774e-01 -1.10838927e-01
-1.05673885e+00 2.95917839e-01 7.10045248e-02 -6.00114346e-01
-1.32625047e-02 -2.43479088e-01 -7.40580738e-01 -3.98331404e-01
-2.57888377e-01 -3.03770036e-01 6.59209132e-01 5.92751563e-01
5.04620910e-01 2.58252949e-01 3.45451087e-01 -1.08894813e+00
-7.18434870e-01 -8.00778210e-01 -2.73272276e-01 5.40871322e-01
2.67214388e-01 -5.06252527e-01 -1.42410740e-01 -2.47128308e-01] | [11.655597686767578, 0.5566099882125854] |
66a94325-98d0-4414-810e-2c72b2f5b68d | towards-imperceptible-document-manipulations | 2305.01860 | null | https://arxiv.org/abs/2305.01860v1 | https://arxiv.org/pdf/2305.01860v1.pdf | Towards Imperceptible Document Manipulations against Neural Ranking Models | Adversarial attacks have gained traction in order to identify potential vulnerabilities in neural ranking models (NRMs), but current attack methods often introduce grammatical errors, nonsensical expressions, or incoherent text fragments, which can be easily detected. Additionally, current methods rely heavily on the use of a well-imitated surrogate NRM to guarantee the attack effect, which makes them difficult to use in practice. To address these issues, we propose a framework called Imperceptible DocumEnt Manipulation (IDEM) to produce adversarial documents that are less noticeable to both algorithms and humans. IDEM instructs a well-established generative language model, such as BART, to generate connection sentences without introducing easy-to-detect errors, and employs a separate position-wise merging strategy to balance relevance and coherence of the perturbed text. Experimental results on the popular MS MARCO benchmark demonstrate that IDEM can outperform strong baselines while preserving fluency and correctness of the target documents as evidenced by automatic and human evaluations. Furthermore, the separation of adversarial text generation from the surrogate NRM makes IDEM more robust and less affected by the quality of the surrogate NRM. | ['Yingfei Sun', 'Le Sun', 'Zheng Ye', 'Ben He', 'Xuanang Chen'] | 2023-05-03 | null | null | null | null | ['adversarial-text'] | ['adversarial'] | [ 5.60813785e-01 1.95542276e-01 3.24753821e-01 -1.71558067e-01
-1.03946817e+00 -1.10090518e+00 9.89492595e-01 1.04260251e-01
-2.91329682e-01 6.15144849e-01 2.25114241e-01 -3.18655103e-01
1.40393987e-01 -6.10103965e-01 -9.04590011e-01 -4.09426957e-01
3.25649641e-02 3.33912641e-01 2.56352481e-02 -5.35163701e-01
1.67556390e-01 1.98359743e-01 -1.21616864e+00 4.06453162e-01
1.31077790e+00 2.63713866e-01 -1.11566998e-01 7.58549333e-01
2.82900959e-01 9.55577075e-01 -1.33526564e+00 -1.04921317e+00
2.78976798e-01 -5.64678669e-01 -5.77556074e-01 -5.20643532e-01
7.24586606e-01 -5.50114930e-01 -3.58527094e-01 1.37782609e+00
7.33731985e-01 6.82979003e-02 6.97371840e-01 -1.14091897e+00
-8.56079638e-01 1.07878411e+00 -3.32560688e-01 1.31926894e-01
2.71196336e-01 3.23602498e-01 9.96356428e-01 -6.89603925e-01
6.50153518e-01 1.47191823e+00 5.13237417e-01 7.97775745e-01
-1.33417308e+00 -8.34697604e-01 3.59782055e-02 -9.96039063e-02
-1.06674314e+00 -7.56544650e-01 7.67851710e-01 -9.87086073e-02
7.66780734e-01 7.58571148e-01 4.72914428e-02 1.80369079e+00
4.36560631e-01 7.20606625e-01 9.48500037e-01 -3.85617465e-01
2.00208411e-01 1.86729461e-01 -8.65075458e-03 4.46225405e-01
4.05173689e-01 2.97593415e-01 -5.51154554e-01 -2.46158257e-01
1.70432940e-01 -4.36415702e-01 -5.91154277e-01 -9.60671753e-02
-9.30521309e-01 7.40476191e-01 3.14478844e-01 2.99218565e-01
-1.77773461e-01 1.25576660e-01 5.20845115e-01 4.56153303e-01
5.12055576e-01 1.01912344e+00 -6.41575130e-03 -2.10863501e-01
-1.06436253e+00 4.79306459e-01 6.73345685e-01 7.04256654e-01
3.23560834e-02 2.28380889e-01 -5.70091128e-01 7.73702681e-01
1.42824337e-01 7.15986073e-01 5.81465542e-01 -6.56620204e-01
8.21335912e-01 3.48214775e-01 -3.85985896e-02 -1.16668952e+00
6.90459609e-02 -7.73126483e-01 -7.61095047e-01 3.78480613e-01
3.75210434e-01 -1.06276415e-01 -8.72151315e-01 2.13507223e+00
2.66885832e-02 -8.53954777e-02 2.60782331e-01 7.94068575e-01
5.26676118e-01 5.80352664e-01 -2.94603575e-02 1.90080449e-01
8.80611360e-01 -8.23827744e-01 -5.71684062e-01 -5.94643652e-01
6.35181427e-01 -7.37401128e-01 1.37253463e+00 4.12574738e-01
-1.22637260e+00 -1.54376432e-01 -1.39580393e+00 4.14056219e-02
-3.39301497e-01 -1.95446268e-01 1.08565725e-01 9.20900524e-01
-8.26900482e-01 7.69869447e-01 -5.25891781e-01 2.88176060e-01
4.49851811e-01 1.07418187e-01 -3.58587652e-01 -1.41459435e-01
-1.70731056e+00 1.09843206e+00 4.01083142e-01 2.21889585e-01
-1.09701622e+00 -7.99460948e-01 -8.19578707e-01 -4.60109748e-02
1.17285453e-01 -6.17759228e-01 1.20805418e+00 -1.00767255e+00
-1.57458091e+00 6.53692424e-01 1.87228635e-01 -4.88165438e-01
9.10929263e-01 -4.62381095e-01 -4.61158574e-01 1.70131594e-01
-1.76803812e-01 3.90230566e-01 1.13494825e+00 -1.24285662e+00
3.97597998e-02 -1.51747152e-01 2.04343945e-01 2.05429792e-01
-5.44057846e-01 2.02956013e-02 -1.57003909e-01 -1.13952100e+00
-4.28997755e-01 -8.45545411e-01 -8.94723311e-02 -5.53164363e-01
-9.26262498e-01 1.83406308e-01 4.91590947e-01 -8.91963601e-01
1.33036149e+00 -2.06186533e+00 2.46738315e-01 3.20442557e-01
4.36984330e-01 7.01989591e-01 -5.75350285e-01 4.10826951e-01
-1.01430014e-01 4.06258851e-01 -2.21602157e-01 -5.39693832e-01
2.87969142e-01 -1.20874628e-01 -8.44354630e-01 2.68872082e-01
2.24935710e-01 1.07031536e+00 -1.04043972e+00 -1.83771014e-01
-2.31984928e-01 4.85016614e-01 -6.90082550e-01 1.90112412e-01
-3.97139192e-01 2.57155508e-01 -7.38815889e-02 3.58564287e-01
4.71385628e-01 1.31006345e-01 -7.65706375e-02 1.79597780e-01
4.17884320e-01 6.29817247e-01 -7.89997995e-01 1.19412029e+00
-3.92585844e-01 6.38660014e-01 -1.57458186e-01 -4.55431461e-01
6.35726213e-01 1.64251864e-01 -4.16190922e-01 -7.35959113e-01
2.15576619e-01 1.65020853e-01 1.74687594e-01 -7.16805682e-02
6.85018837e-01 1.40830830e-01 -2.45794088e-01 6.55062318e-01
-2.61615813e-01 -1.73169792e-01 3.37719411e-01 6.89435065e-01
1.40262115e+00 -2.66822785e-01 -9.92332175e-02 2.54380330e-02
2.53112167e-01 -2.22723052e-01 4.15547132e-01 1.20857739e+00
-2.61137933e-02 6.05804384e-01 4.72790927e-01 7.03243613e-02
-9.46251571e-01 -1.15609491e+00 1.92468971e-01 1.04699719e+00
2.68387813e-02 -5.91388524e-01 -1.16915739e+00 -1.01052952e+00
-2.47058362e-01 1.19056404e+00 -5.84715426e-01 -7.71735191e-01
-5.95116138e-01 -6.19356632e-01 1.24722493e+00 2.73685753e-01
2.50280023e-01 -1.10075200e+00 -4.11400646e-01 1.54190585e-01
-2.76630223e-01 -8.78109157e-01 -6.89411819e-01 -9.89702716e-03
-4.95819479e-01 -8.35711896e-01 -4.88605022e-01 -4.33685273e-01
8.64163458e-01 1.55750895e-02 1.22406089e+00 1.63492456e-01
-1.16960719e-01 2.70030480e-02 -3.95662367e-01 -4.87423241e-01
-1.09228754e+00 1.55764697e-02 -4.53798361e-02 -2.30950445e-01
-4.18336168e-02 -4.99209046e-01 -2.39666432e-01 1.97218120e-01
-1.34402573e+00 4.37906757e-02 7.21273541e-01 9.16791081e-01
2.94928788e-03 1.47329986e-01 5.35735011e-01 -1.11282194e+00
1.18262672e+00 -1.58059165e-01 -3.88508290e-01 3.70398670e-01
-6.86670721e-01 2.23585442e-01 9.22626078e-01 -8.05635154e-01
-7.62740612e-01 -7.04195797e-01 -1.31859615e-01 -3.00994128e-01
1.28470629e-01 5.20757437e-01 -5.04185975e-01 2.95038894e-02
1.06924438e+00 2.84995943e-01 -6.33500144e-02 -1.76881775e-01
5.92618525e-01 6.33150399e-01 7.94318318e-01 -5.03722429e-01
1.37238753e+00 -5.06227911e-02 -4.08160925e-01 -4.79620963e-01
-7.56260872e-01 2.91867852e-01 -8.00769180e-02 -1.18776105e-01
2.44812012e-01 -7.42849708e-01 -2.64887989e-01 7.50757635e-01
-1.33427894e+00 -3.80253583e-01 -1.38378125e-02 -3.81922089e-02
-6.82254881e-02 5.05017579e-01 -7.78514683e-01 -6.44467294e-01
-7.47213423e-01 -1.13612890e+00 8.96835268e-01 -1.00201666e-01
-6.72742665e-01 -8.49715650e-01 5.85068241e-02 4.88648713e-01
5.91154933e-01 3.05682540e-01 1.20442700e+00 -9.58393037e-01
-3.98127854e-01 -5.74411988e-01 1.85154378e-01 6.08854175e-01
-1.90918669e-01 6.62865071e-03 -1.01440263e+00 -5.05909204e-01
-4.56887409e-02 -5.76611161e-01 8.49722505e-01 -2.82802612e-01
7.59888947e-01 -6.68939173e-01 -7.64133036e-02 4.41153407e-01
7.39672959e-01 -4.27364819e-02 8.25406909e-01 3.05127680e-01
7.42215931e-01 6.21678650e-01 2.60073304e-01 -7.63779739e-03
1.09964535e-02 5.80252826e-01 4.11175847e-01 -1.79753825e-02
-1.49691701e-01 -4.91973221e-01 9.10741508e-01 7.22754180e-01
4.77957845e-01 -6.85898900e-01 -7.07123280e-01 1.97313637e-01
-1.38873553e+00 -1.09487188e+00 8.72476399e-02 2.20931077e+00
1.36240172e+00 6.74312532e-01 -3.32113802e-01 3.85955215e-01
6.58413887e-01 3.59136134e-01 -4.53708231e-01 -4.24644798e-01
-2.03105375e-01 2.19245777e-01 2.77696311e-01 5.08554935e-01
-9.62766111e-01 9.72609103e-01 6.16879606e+00 9.16173756e-01
-1.09591579e+00 3.41466516e-02 4.64062721e-01 -4.07205850e-01
-7.88756013e-01 -2.12077692e-01 -6.23167753e-01 6.85746610e-01
1.06659639e+00 -2.85531133e-01 6.06444001e-01 6.37319684e-01
1.01299196e-01 3.47423345e-01 -1.17473197e+00 6.83966339e-01
3.07605326e-01 -1.03941739e+00 3.74952555e-01 -4.83720005e-02
5.69736421e-01 -1.90652966e-01 5.47187924e-01 5.37809014e-01
5.70645392e-01 -1.28512001e+00 1.09251475e+00 3.41231287e-01
6.84443474e-01 -9.46351171e-01 6.35455191e-01 5.53043127e-01
-4.36709285e-01 7.50988871e-02 -2.76957005e-01 1.70586765e-01
-5.77078499e-02 8.10725451e-01 -9.25425649e-01 3.90919745e-01
2.17031643e-01 2.30225146e-01 -9.98475134e-01 5.62654078e-01
-7.27983356e-01 8.14645648e-01 -8.74296352e-02 -1.58193603e-01
4.27882262e-02 1.75982207e-01 1.02320349e+00 1.10601544e+00
-9.82362330e-02 -4.45745081e-01 -1.37023821e-01 9.98983562e-01
-5.31589270e-01 2.55172439e-02 -6.11337543e-01 -5.04035652e-01
6.13685727e-01 1.05289447e+00 -3.12294096e-01 -1.55398235e-01
1.82040423e-01 1.20736516e+00 5.30739903e-01 3.18073869e-01
-8.31839085e-01 -6.91611648e-01 5.75188935e-01 -1.09679349e-01
7.26016518e-03 9.14010257e-02 -3.53532791e-01 -1.10033727e+00
3.29888254e-01 -1.75060046e+00 9.92795154e-02 -5.60579300e-01
-1.31395853e+00 9.54658329e-01 -2.31614560e-01 -1.06497991e+00
-5.12818038e-01 -2.06400737e-01 -7.04448581e-01 9.50803041e-01
-1.23110187e+00 -1.04317415e+00 9.92362723e-02 2.63695925e-01
3.76835585e-01 -2.61893034e-01 8.50269854e-01 2.52162427e-01
-8.26957941e-01 1.34826279e+00 1.83810778e-02 3.00179392e-01
1.02270722e+00 -1.28841412e+00 8.71363759e-01 1.42456472e+00
2.88911968e-01 1.10854673e+00 1.09555244e+00 -8.75451505e-01
-1.13284492e+00 -1.35963368e+00 9.01685774e-01 -8.45947146e-01
5.59854329e-01 -7.92085767e-01 -1.02375543e+00 4.36226100e-01
1.38391614e-01 -3.90974522e-01 5.53135693e-01 1.10520646e-02
-8.86217535e-01 8.42190906e-02 -1.02138114e+00 1.14266896e+00
8.63587618e-01 -7.06138372e-01 -7.08899081e-01 3.88631463e-01
8.63333464e-01 -5.81101358e-01 -3.52750182e-01 3.31295729e-01
4.32321459e-01 -8.11950624e-01 9.43963051e-01 -7.61618674e-01
7.63953745e-01 -1.89352810e-01 7.41668940e-02 -1.65347135e+00
-1.88154474e-01 -9.47190464e-01 -2.11308122e-01 1.45402789e+00
6.21840060e-01 -6.60646141e-01 3.52527529e-01 6.25414431e-01
-1.11647360e-01 -4.35615003e-01 -7.47535348e-01 -1.01508439e+00
2.70685315e-01 -3.08394998e-01 5.61046362e-01 9.20371294e-01
-9.70364884e-02 4.43870723e-01 -5.38717985e-01 2.81816036e-01
5.05042613e-01 -3.00387412e-01 8.15116763e-01 -9.41763937e-01
-5.82632124e-01 -6.18338645e-01 -2.48555809e-01 -6.88592494e-01
3.80437732e-01 -1.16779602e+00 2.94449508e-01 -1.02707982e+00
5.92944734e-02 -1.66872412e-01 -1.41285807e-01 4.50496346e-01
-7.14495838e-01 3.99905801e-01 2.50150889e-01 2.18737602e-01
-5.35413027e-01 5.87100506e-01 8.69702458e-01 -3.37472409e-01
-3.69085521e-02 -1.43826544e-01 -1.06292641e+00 5.92662573e-01
6.91550016e-01 -7.74473786e-01 -6.07410491e-01 -5.75589836e-01
5.23921371e-01 -4.39958334e-01 3.82921040e-01 -7.96066105e-01
4.85591590e-03 1.72812939e-01 3.02058011e-01 -9.55805406e-02
4.08436321e-02 -2.48792425e-01 -6.81460649e-02 3.99208039e-01
-7.84885645e-01 1.71460778e-01 2.10305020e-01 5.28755605e-01
-2.44601313e-02 -4.52861428e-01 7.99262166e-01 1.51163638e-01
-2.55698133e-02 -1.16274431e-01 -3.19261372e-01 4.58573848e-01
7.49793410e-01 2.56382436e-01 -7.75415480e-01 -5.03169656e-01
-1.26962624e-02 5.71971107e-03 6.40491843e-01 5.95113575e-01
5.79067588e-01 -1.10881901e+00 -8.79640222e-01 7.55667463e-02
9.82235074e-02 -2.76850015e-01 1.31817043e-01 4.76100504e-01
-3.24872226e-01 1.26260579e-01 1.82858422e-01 -7.55484328e-02
-1.35949624e+00 4.81588632e-01 3.29331040e-01 -5.98070502e-01
-5.13000250e-01 9.43353832e-01 8.30553696e-02 -4.68891442e-01
4.40466523e-01 1.08981825e-01 3.89710218e-02 -2.13823900e-01
8.38365436e-01 1.36778995e-01 3.35241586e-01 -2.96063542e-01
-1.82845473e-01 -2.44352296e-01 -5.24710834e-01 -3.56129676e-01
9.89291608e-01 2.76351750e-01 -8.11979100e-02 9.39125791e-02
8.84271324e-01 5.94875813e-01 -9.75160837e-01 -4.96271215e-02
-7.11284727e-02 -3.68784517e-01 2.24625766e-02 -1.22707641e+00
-6.73369348e-01 6.84298098e-01 1.49474144e-01 1.50050610e-01
8.96673620e-01 -4.29838508e-01 9.92448688e-01 4.18240130e-01
1.08990751e-01 -7.66118824e-01 3.20018947e-01 6.23186350e-01
1.17756677e+00 -9.06776130e-01 -2.10808247e-01 -2.21170187e-01
-7.01510489e-01 7.33939886e-01 5.70986688e-01 1.18906915e-01
-3.74681130e-02 3.75686586e-01 2.64460653e-01 1.04492582e-01
-9.17624652e-01 4.32270497e-01 5.61436355e-01 6.29652679e-01
2.59504229e-01 -1.69972360e-01 -6.08543232e-02 7.74826646e-01
-7.53387690e-01 -6.34990990e-01 5.24797142e-01 8.84175539e-01
-1.40753657e-01 -1.24174476e+00 -4.87317383e-01 2.76205361e-01
-6.64525867e-01 -5.45197070e-01 -8.18414032e-01 4.96892333e-01
-2.95829654e-01 1.15852702e+00 -3.84647876e-01 -5.31626463e-01
3.57246518e-01 1.26901284e-01 3.84412378e-01 -5.41664541e-01
-8.38574648e-01 -3.02769899e-01 2.17772678e-01 -4.57867652e-01
3.80678356e-01 -4.28956866e-01 -8.65448058e-01 -3.70720029e-01
-3.87259036e-01 1.83032528e-01 5.81558228e-01 1.06741142e+00
4.99673605e-01 6.16599679e-01 5.85697174e-01 -5.08074760e-01
-1.25309837e+00 -1.11168659e+00 -1.57515511e-01 7.76580155e-01
3.30475241e-01 -2.97146201e-01 -5.82303524e-01 -9.58574116e-02] | [6.089151382446289, 8.15313720703125] |
abab3cca-9763-46b7-8bb7-26d1b85bb352 | real-esrgan-training-real-world-blind-super | 2107.10833 | null | https://arxiv.org/abs/2107.10833v2 | https://arxiv.org/pdf/2107.10833v2.pdf | Real-ESRGAN: Training Real-World Blind Super-Resolution with Pure Synthetic Data | Though many attempts have been made in blind super-resolution to restore low-resolution images with unknown and complex degradations, they are still far from addressing general real-world degraded images. In this work, we extend the powerful ESRGAN to a practical restoration application (namely, Real-ESRGAN), which is trained with pure synthetic data. Specifically, a high-order degradation modeling process is introduced to better simulate complex real-world degradations. We also consider the common ringing and overshoot artifacts in the synthesis process. In addition, we employ a U-Net discriminator with spectral normalization to increase discriminator capability and stabilize the training dynamics. Extensive comparisons have shown its superior visual performance than prior works on various real datasets. We also provide efficient implementations to synthesize training pairs on the fly. | ['Ying Shan', 'Chao Dong', 'Liangbin Xie', 'Xintao Wang'] | 2021-07-22 | null | null | null | null | ['video-super-resolution'] | ['computer-vision'] | [ 5.26697636e-01 -4.27336335e-01 6.11402243e-02 3.09211351e-02
-6.36186779e-01 -3.51461381e-01 6.01905286e-01 -1.02514541e+00
8.66273344e-02 1.01405728e+00 5.59886336e-01 -1.43968761e-01
1.17158093e-01 -3.38108212e-01 -5.39600134e-01 -8.66610110e-01
7.54962265e-02 -2.61228263e-01 -1.29344612e-02 -4.83140081e-01
1.10839428e-02 3.68854791e-01 -1.43452728e+00 1.66791514e-01
1.26260078e+00 6.20840073e-01 3.18221390e-01 9.27887499e-01
5.80751896e-01 8.44817758e-01 -7.82672405e-01 -1.30699694e-01
6.82134628e-01 -8.35072637e-01 -3.14490587e-01 4.39258069e-01
4.16471511e-01 -7.35464394e-01 -8.97926450e-01 1.33944976e+00
9.85458016e-01 1.57163024e-01 3.67788345e-01 -6.90894127e-01
-1.10393929e+00 2.01498717e-01 -6.31641209e-01 6.28917038e-01
3.17952722e-01 5.99721432e-01 3.13676357e-01 -5.83568811e-01
5.63829899e-01 1.45563710e+00 5.64029396e-01 8.23473871e-01
-1.19166183e+00 -6.42581999e-01 -9.56831798e-02 4.34426486e-01
-1.09259140e+00 -8.26876223e-01 7.43160009e-01 -2.27570221e-01
5.13914764e-01 1.71123505e-01 4.30981487e-01 1.48381519e+00
8.27192590e-02 4.36241239e-01 1.55341220e+00 -1.33020520e-01
6.26744032e-02 -3.85587484e-01 -5.32972574e-01 2.41876811e-01
4.10636365e-02 6.36594772e-01 -4.02759790e-01 1.55259609e-01
1.38538313e+00 -3.17419887e-01 -9.17422712e-01 4.90289144e-02
-1.20266306e+00 4.07773048e-01 5.87053120e-01 1.53381094e-01
-3.94525588e-01 1.13073535e-01 1.03127666e-01 6.49541259e-01
6.04050159e-01 3.63336384e-01 -1.98454946e-01 1.02813512e-01
-9.04532373e-01 5.28963767e-02 2.31600568e-01 6.75202429e-01
4.23054636e-01 7.69063294e-01 -4.25408214e-01 9.82662082e-01
-4.24792431e-02 3.54568571e-01 6.88600540e-01 -1.42381454e+00
3.41524094e-01 -4.01612818e-02 3.81457865e-01 -6.84550524e-01
-3.62544544e-02 -9.04478490e-01 -1.55140114e+00 5.27841985e-01
1.98011756e-01 -1.22768663e-01 -1.21549404e+00 1.69639719e+00
-2.03055814e-02 6.23000681e-01 3.65842640e-01 1.43449044e+00
4.93326873e-01 5.98761916e-01 -4.00729299e-01 -6.33053362e-01
9.84580338e-01 -1.07024229e+00 -1.13077581e+00 -2.53488123e-01
-3.80322546e-01 -8.63431931e-01 1.26380360e+00 5.07676363e-01
-1.16816223e+00 -7.18105793e-01 -1.18887782e+00 -4.98878919e-02
2.15683028e-01 2.64526993e-01 3.09493124e-01 5.50876737e-01
-1.25108135e+00 9.57057714e-01 -7.44868994e-01 -2.34378055e-01
4.09076750e-01 -3.97200361e-02 -1.02838658e-01 -2.79626876e-01
-1.26836598e+00 8.27960372e-01 -2.60840729e-02 2.97392845e-01
-1.25389946e+00 -5.22014499e-01 -5.24794996e-01 -2.22167209e-01
2.11123034e-01 -1.06585586e+00 1.15052557e+00 -1.20175588e+00
-1.98334002e+00 4.27101642e-01 -1.88764349e-01 -5.00666976e-01
7.79969931e-01 -1.26571789e-01 -8.10574472e-01 7.71591365e-02
-1.71654701e-01 2.20816806e-01 1.37789154e+00 -1.58369803e+00
-1.47523686e-01 -2.10751355e-01 -9.41897184e-02 4.69003290e-01
-1.30294606e-01 1.68686226e-01 -3.60244125e-01 -1.23877430e+00
2.56841153e-01 -6.01329088e-01 -2.26607144e-01 2.95886192e-02
-2.58447528e-01 6.93453074e-01 8.01532507e-01 -1.21658885e+00
1.14685726e+00 -2.06285000e+00 3.11974317e-01 -4.33861256e-01
1.06000856e-01 5.33849239e-01 -3.25626314e-01 1.35990843e-01
-9.13985670e-02 -1.56896468e-02 -5.59813440e-01 -4.71985847e-01
-3.16579938e-01 3.68545502e-01 -5.68870425e-01 7.07073331e-01
8.97375867e-02 6.87510431e-01 -7.83990681e-01 -4.01981734e-02
1.87271774e-01 8.77971709e-01 -3.55700672e-01 2.11021334e-01
-6.67993724e-02 9.40732062e-01 1.05730161e-01 7.61826575e-01
8.54355037e-01 -1.98766381e-01 3.81780304e-02 -3.88333887e-01
6.57802224e-02 3.03670689e-02 -1.23635793e+00 1.59093368e+00
-3.90573204e-01 8.40222597e-01 3.10293078e-01 -6.38815284e-01
7.18848407e-01 2.91226685e-01 -1.55789545e-02 -8.95752490e-01
4.39653993e-02 1.73160136e-01 3.23785432e-02 -4.21779007e-01
5.14400125e-01 -2.18281776e-01 6.12002254e-01 1.00977629e-01
-1.16874710e-01 -1.33936867e-01 -3.25716883e-02 -6.37328401e-02
9.81537402e-01 3.17275465e-01 1.29486009e-01 -2.65529510e-02
4.64993834e-01 -3.78253341e-01 7.39623189e-01 4.94264245e-01
-2.67522097e-01 1.19027162e+00 1.93746403e-01 -1.24675654e-01
-1.32178235e+00 -1.08750629e+00 5.05764000e-02 4.14696008e-01
4.55203861e-01 -2.17031799e-02 -7.13156223e-01 -9.25259739e-02
-4.17815775e-01 3.68814975e-01 -2.32819572e-01 -1.30040109e-01
-5.73989272e-01 -1.09262121e+00 5.33437371e-01 2.88485497e-01
9.96576846e-01 -7.50950456e-01 -3.26844871e-01 4.79788408e-02
-4.72147852e-01 -1.22179949e+00 -6.99951112e-01 -2.09211990e-01
-9.65002000e-01 -9.45680618e-01 -9.59441006e-01 -7.13925302e-01
6.41309202e-01 4.49434906e-01 8.28113496e-01 -2.08445862e-01
-2.90434003e-01 -7.28866011e-02 -2.00790778e-01 1.77557111e-01
-5.40792644e-01 -6.15494788e-01 3.54608715e-01 1.02602780e-01
-5.35710931e-01 -8.56772840e-01 -8.51633549e-01 4.21276987e-01
-9.87114787e-01 1.75702661e-01 5.31650007e-01 1.25578380e+00
6.11603260e-01 4.79143679e-01 4.68743414e-01 -1.92454293e-01
7.89253056e-01 -1.61649182e-01 -4.79836255e-01 3.46995145e-03
-7.41356015e-01 5.22886142e-02 9.45489943e-01 -8.42852771e-01
-1.35852110e+00 -2.44361550e-01 2.78028585e-02 -7.99739063e-01
1.20201014e-01 1.92341059e-02 -2.33124629e-01 -3.42031181e-01
8.20794821e-01 4.65935320e-01 1.42001614e-01 -7.56190062e-01
2.90456325e-01 7.59770036e-01 1.17691123e+00 -2.40942091e-01
1.11564076e+00 4.65625435e-01 -2.05624387e-01 -7.00192750e-01
-6.18981719e-01 -1.64723396e-02 -6.35574386e-02 -1.67114571e-01
4.99890268e-01 -1.37734950e+00 -5.32668948e-01 1.06563020e+00
-9.58887160e-01 -7.25597024e-01 -3.10987055e-01 3.65994722e-01
-6.70478463e-01 5.78709185e-01 -9.98488247e-01 -6.87163830e-01
-2.71704108e-01 -1.08387589e+00 6.50471747e-01 5.52023232e-01
5.66789627e-01 -5.93006372e-01 -6.48127422e-02 3.41426909e-01
7.92890370e-01 2.03517869e-01 3.91635180e-01 4.64830756e-01
-7.64824450e-01 3.92974734e-01 -4.33711022e-01 8.94497693e-01
3.59503806e-01 -1.03801399e-01 -8.89381886e-01 -6.99698746e-01
3.09013039e-01 -3.14377509e-02 1.06427729e+00 4.32345539e-01
9.70257998e-01 -5.42094469e-01 1.46531895e-01 1.16760337e+00
1.48520124e+00 1.35780603e-01 1.19189441e+00 4.40094411e-01
4.70924109e-01 1.16281301e-01 3.95818442e-01 4.03002143e-01
1.32315055e-01 7.05550969e-01 6.12309217e-01 -3.23553443e-01
-9.46468949e-01 -2.83619314e-01 6.80298507e-01 7.32700109e-01
-4.01736975e-01 -4.18128341e-01 -4.16137129e-01 4.86079663e-01
-1.50581431e+00 -1.09241009e+00 8.16286821e-03 1.92077565e+00
1.02468944e+00 -1.69791698e-01 -3.10394794e-01 7.27023855e-02
9.13999796e-01 5.03624260e-01 -8.55463743e-01 1.68629333e-01
-8.52195859e-01 1.57353222e-01 6.80635035e-01 6.38465226e-01
-9.08789158e-01 9.10802782e-01 6.97393656e+00 5.80603778e-01
-1.21342385e+00 1.30572915e-01 6.33972526e-01 -4.36126664e-02
-1.29741043e-01 -2.02878118e-02 -2.27192044e-01 5.47110260e-01
6.36535227e-01 -8.91630873e-02 1.20000374e+00 3.19119155e-01
4.72093850e-01 2.30718777e-02 -3.93826991e-01 1.19035411e+00
7.49988034e-02 -1.04707563e+00 -2.53313202e-02 -1.37755498e-01
1.16895401e+00 -1.21646173e-01 2.60237843e-01 -1.29486635e-01
4.98484671e-01 -1.08189428e+00 4.40274507e-01 8.07032168e-01
1.27224481e+00 -4.68109280e-01 4.12738323e-01 2.07943290e-01
-8.13536167e-01 -3.32232594e-01 -5.33914089e-01 -1.49796396e-01
4.01256472e-01 5.10527134e-01 -3.69499534e-01 7.55147219e-01
7.38205612e-01 9.03050840e-01 -5.14095783e-01 1.07464731e+00
-6.60942852e-01 6.06113672e-01 4.02303301e-02 7.73742139e-01
-3.71779263e-01 -3.11899781e-01 8.49938929e-01 7.37416029e-01
6.33754790e-01 1.81881711e-01 -3.49880099e-01 7.33051181e-01
-6.57868460e-02 -5.25210738e-01 -1.64104789e-01 1.18039742e-01
2.40685493e-01 9.24072802e-01 -5.08962609e-02 -1.62757739e-01
-1.71862289e-01 1.53995860e+00 -2.62841910e-01 8.45160961e-01
-8.28206003e-01 -1.09272286e-01 1.22989094e+00 2.34521516e-02
2.01079845e-01 -4.06310081e-01 -2.21781820e-01 -1.61273122e+00
1.53302429e-02 -1.27140510e+00 2.78483257e-02 -1.33897591e+00
-1.20346653e+00 7.33885884e-01 -4.80949402e-01 -1.42781973e+00
-1.86683580e-01 -3.09172183e-01 -4.33666855e-01 1.10985351e+00
-1.94759881e+00 -9.15885150e-01 -6.37059629e-01 9.21764731e-01
6.33279979e-01 -2.42147684e-01 4.09201443e-01 3.84894252e-01
-7.68505216e-01 3.56288016e-01 4.72524285e-01 -2.35226721e-01
1.11685514e+00 -9.95680928e-01 6.70124352e-01 1.63011849e+00
-3.20390463e-01 3.63375455e-01 1.16003597e+00 -7.10989654e-01
-1.37691164e+00 -1.13175666e+00 2.15880409e-01 8.50009471e-02
5.57253540e-01 -3.46843116e-02 -1.02329659e+00 5.50915658e-01
3.47443938e-01 2.94554710e-01 -1.91782370e-01 -7.65424728e-01
-4.14085478e-01 -2.41308331e-01 -1.23193240e+00 7.76823282e-01
1.46309543e+00 -6.94060028e-01 -3.00576478e-01 1.10047095e-01
9.18746173e-01 -7.16176748e-01 -7.18927383e-01 5.27050495e-01
2.70670056e-01 -1.04260480e+00 1.27083373e+00 -2.41007820e-01
5.65203190e-01 -7.82429755e-01 -2.97221988e-01 -1.72065341e+00
-3.72894764e-01 -1.10014164e+00 -4.90211964e-01 1.09645057e+00
-1.86301202e-01 -7.75348604e-01 4.30441409e-01 -5.02883308e-02
-2.10290596e-01 -2.31517583e-01 -7.86320627e-01 -9.89083767e-01
-2.56415337e-01 1.16109438e-01 5.28948665e-01 8.62485170e-01
-6.28546655e-01 8.24950784e-02 -1.14690685e+00 5.97266614e-01
1.04648304e+00 -8.08124393e-02 6.72036529e-01 -7.87889779e-01
-5.22128284e-01 -1.13370024e-01 -1.95971370e-01 -1.26084149e+00
-1.94697514e-01 -2.44380429e-01 1.11426108e-01 -1.53644347e+00
-1.81514192e-02 1.98679306e-02 -1.52129248e-01 5.26640899e-02
-3.30794394e-01 5.55357456e-01 -2.65646335e-02 5.82369983e-01
-2.57204585e-02 7.57883370e-01 1.80937839e+00 -9.50716585e-02
-2.81493872e-01 -9.35331136e-02 -8.02784085e-01 3.28670681e-01
9.21912491e-01 -7.25224465e-02 -5.27089357e-01 -6.62487745e-01
-1.11958988e-01 4.90477383e-01 6.75692677e-01 -1.18656349e+00
1.40484974e-01 -1.69413790e-01 5.14471233e-01 -9.77573246e-02
3.69218051e-01 -3.78605604e-01 5.89789331e-01 3.92117053e-01
-1.31823555e-01 -2.10632637e-01 6.27935957e-03 5.35383523e-01
-4.70818788e-01 4.25522923e-01 1.29992807e+00 -1.70776263e-01
-6.27798557e-01 4.73657489e-01 -5.33373915e-02 -7.83429965e-02
6.07360303e-01 -2.03993708e-01 -8.87846947e-01 -6.20713592e-01
-7.48532414e-01 -2.87151169e-02 9.28819537e-01 3.28597695e-01
6.64451897e-01 -1.32583165e+00 -9.56405997e-01 2.51176506e-01
-3.62100154e-01 -1.97539583e-01 7.31164575e-01 5.77558577e-01
-3.95855695e-01 -1.68747991e-01 -4.88083065e-01 -2.61871904e-01
-1.14805317e+00 7.18037546e-01 7.14983702e-01 -2.15033460e-02
-8.03987682e-01 5.89999259e-01 -5.99221233e-03 -2.16144007e-02
1.54735565e-01 1.86739247e-02 -9.54861790e-02 -2.15342820e-01
7.69403994e-01 6.01027191e-01 -1.88000232e-01 -5.55358529e-01
-2.16179993e-02 4.74284232e-01 4.23875928e-01 -1.31233171e-01
1.33165431e+00 -7.08894968e-01 -7.30485767e-02 7.31393471e-02
7.88266957e-01 -1.33706883e-01 -1.84227777e+00 -3.87535691e-01
-4.84102309e-01 -7.59691775e-01 3.19167733e-01 -1.02538669e+00
-1.31871903e+00 6.46019757e-01 9.67236578e-01 -9.43303257e-02
1.93430734e+00 -4.23014313e-01 7.32714713e-01 -7.80014917e-02
1.88786805e-01 -8.05911243e-01 3.40031296e-01 3.00961733e-01
1.26744723e+00 -1.03330278e+00 6.36386797e-02 -3.67810577e-01
-5.45331597e-01 9.64741647e-01 5.61925650e-01 -1.79418266e-01
2.31677061e-03 3.47826451e-01 2.49332473e-01 4.87025172e-01
-6.87299907e-01 -1.92383990e-01 -4.55353670e-02 9.43369746e-01
-1.06201075e-01 -1.30389914e-01 -3.20015877e-01 3.58100682e-01
-6.90900460e-02 2.63468087e-01 1.01417160e+00 3.93634796e-01
-2.90584832e-01 -8.96282077e-01 -6.34342909e-01 8.26878101e-02
-5.37721097e-01 -2.23241270e-01 1.28968507e-01 3.23153615e-01
1.68084018e-02 1.04652464e+00 -2.67622769e-01 -4.63434666e-01
3.06974590e-01 -3.45247656e-01 5.66828907e-01 -1.22400604e-01
-3.27851623e-01 1.75457686e-01 -1.38021350e-01 -7.80800343e-01
-5.72410583e-01 -5.01679301e-01 -6.21224761e-01 -4.74217027e-01
-1.25801221e-01 -3.48569751e-01 5.56116045e-01 4.39193726e-01
4.13995057e-01 8.63025606e-01 8.37650359e-01 -1.03124058e+00
-7.90499568e-01 -1.07122540e+00 -6.60035968e-01 2.31175438e-01
1.04783821e+00 -4.81650114e-01 -8.53050351e-01 3.10151041e-01] | [11.411760330200195, -2.1670775413513184] |
59b3c6d2-da17-4854-b073-b905ed82d397 | pixelsteganalysis-destroying-hidden | 1902.11113 | null | http://arxiv.org/abs/1902.11113v2 | http://arxiv.org/pdf/1902.11113v2.pdf | PixelSteganalysis: Destroying Hidden Information with a Low Degree of Visual Degradation | Steganography is the science of unnoticeably concealing a secret message
within a certain image, called a cover image. The cover image with the secret
message is called a stego image. Steganography is commonly used for illegal
purposes such as terrorist activities and pornography. To thwart covert
communications and transactions, attacking algorithms against steganography,
called steganalysis, exist. Currently, there are many studies implementing deep
learning to the steganography algorithm. However, conventional steganalysis is
no longer effective for deep learning based steganography algorithms. Our
framework is the first one to disturb covert communications and transactions
via the recent deep learning-based steganography algorithms. We first extract a
sophisticated pixel distribution of the potential stego image from the
auto-regressive model induced by deep learning. Using the extracted pixel
distributions, we detect whether an image is the stego or not at the pixel
level. Each pixel value is adjusted as required and the adjustment induces an
effective removal of the secret image. Because the decoding method of deep
learning-based steganography algorithms is approximate (lossy), which is
different from the conventional steganography, we propose a new quantitative
metric that is more suitable for measuring the accurate effect. We evaluate our
method using three public benchmarks in comparison with a conventional
steganalysis method and show up to a 20% improvement in terms of decoding rate. | ['Hyun-Soo Choi', 'Dahuin Jung', 'Sungroh Yoon', 'Ho Bae'] | 2019-01-30 | null | null | null | null | ['steganalysis', 'image-steganography'] | ['computer-vision', 'computer-vision'] | [ 8.92667770e-01 1.35036409e-01 3.55125107e-02 1.95318460e-01
-1.92826405e-01 -1.19939655e-01 5.52525878e-01 -4.41464037e-01
-3.17666829e-01 5.63769400e-01 -2.13638961e-01 -6.02681518e-01
4.56442446e-01 -1.18900478e+00 -1.05810559e+00 -1.23900855e+00
-4.44593549e-01 -9.99838114e-02 7.71216974e-02 -6.33175254e-01
5.65320969e-01 1.47621483e-01 -1.26333988e+00 5.73719144e-01
5.60330272e-01 9.88495290e-01 -5.48526160e-02 8.04456770e-01
8.06428716e-02 7.12684572e-01 -7.06606030e-01 -1.54282272e-01
4.55744117e-01 -1.03011537e+00 -5.19387126e-01 1.62380889e-01
-6.63496256e-02 -3.24913532e-01 -5.74694812e-01 1.45593858e+00
2.23727107e-01 -6.07228637e-01 5.34408092e-01 -1.05147362e+00
-3.64787251e-01 6.87351644e-01 -7.03362048e-01 -8.37858580e-03
1.31208643e-01 5.06969869e-01 5.86263120e-01 -3.95621926e-01
3.59712213e-01 1.30223083e+00 6.31933391e-01 4.39303547e-01
-9.56654668e-01 -1.17993295e+00 -4.97034818e-01 3.17136288e-01
-1.09809411e+00 -3.50673318e-01 8.56036961e-01 -1.07395418e-01
5.75928092e-01 3.85423422e-01 8.84082079e-01 8.79478574e-01
1.12108886e+00 7.35810041e-01 1.57624388e+00 -5.65023124e-01
-1.50827676e-01 2.48258989e-02 -5.17042220e-01 8.78288925e-01
4.68956709e-01 8.64836037e-01 -4.16418687e-02 1.27193734e-01
6.45623684e-01 1.72524050e-01 -3.92887473e-01 -2.12706700e-01
-1.31430471e+00 1.11739993e+00 5.43568969e-01 4.38437879e-01
-1.30150706e-01 5.96020758e-01 3.10105324e-01 1.12808347e+00
3.26354623e-01 3.15331370e-01 -6.30197898e-02 3.11016619e-01
-9.59542334e-01 -1.38783693e-01 1.10410500e+00 3.41365218e-01
8.94825220e-01 4.58017796e-01 3.20263654e-01 5.10690138e-02
5.40939867e-01 9.15321231e-01 3.98282230e-01 -5.09342372e-01
2.76005536e-01 3.52906495e-01 -2.74344593e-01 -1.66459215e+00
-1.03090219e-01 -4.38618839e-01 -1.36702800e+00 8.19797099e-01
2.52500474e-01 -1.25453815e-01 -1.04022300e+00 1.15701699e+00
-8.20396766e-02 4.40055490e-01 2.81339020e-01 5.17477632e-01
5.69237530e-01 1.05202413e+00 -3.82994026e-01 -2.51614600e-01
1.08199811e+00 -8.27919364e-01 -7.13618457e-01 -4.22959387e-01
8.12464833e-01 -7.62951791e-01 3.96319866e-01 5.27788937e-01
-7.51623392e-01 -2.73633510e-01 -1.48816848e+00 6.21468008e-01
-5.47877729e-01 -7.65606344e-01 2.63951838e-01 1.14251959e+00
-8.30432594e-01 6.15647554e-01 -3.60760421e-01 1.72071308e-01
5.97129941e-01 6.81084871e-01 -3.10993820e-01 -1.67530075e-01
-1.68296123e+00 8.87756586e-01 8.72902393e-01 -8.98663253e-02
-1.33018863e+00 4.53945361e-02 -1.01372373e+00 1.75082177e-01
2.09301785e-01 -7.08304271e-02 4.90223616e-01 -1.57277751e+00
-1.21627188e+00 1.08234143e+00 3.23395431e-01 -9.04821813e-01
5.77841878e-01 5.61938882e-01 -6.98560119e-01 3.70354578e-02
-5.70353746e-01 3.72011334e-01 1.81138611e+00 -1.39033711e+00
-6.39084101e-01 -8.65420327e-02 -2.10039720e-01 -1.39485270e-01
-2.28960007e-01 -2.40124315e-01 3.24271768e-02 -7.51144290e-01
2.60961890e-01 -1.20819497e+00 -3.29184122e-02 -2.64098823e-01
-3.49512398e-01 5.40138125e-01 1.39491940e+00 -9.44305956e-01
1.06557238e+00 -2.18980622e+00 -1.91113755e-01 5.67747593e-01
3.79177421e-01 7.38766372e-01 -1.49055168e-01 4.27849799e-01
-1.23185024e-01 3.16138119e-01 -5.12504041e-01 2.78262407e-01
-4.34408128e-01 -2.22875760e-03 -1.66683599e-01 9.62458551e-01
-3.44086647e-01 1.10472441e+00 -8.44637215e-01 -4.56368953e-01
4.60185438e-01 5.77690899e-01 -2.02235386e-01 -1.85849562e-01
-8.78932979e-03 5.60473502e-01 -2.64164478e-01 4.70328629e-01
1.03897345e+00 -2.48112395e-01 3.40662956e-01 1.16365559e-01
3.00509691e-01 -2.27044653e-02 -7.34017849e-01 9.80034351e-01
-4.02697921e-01 1.10182214e+00 -8.45548734e-02 -1.40401280e+00
1.09622180e+00 3.88633251e-01 3.20392698e-01 -7.11115718e-01
6.23155177e-01 5.83630204e-01 3.16207260e-01 -6.10137939e-01
1.46484748e-01 -2.23155409e-01 -1.43062532e-01 7.01679051e-01
-3.18951786e-01 -1.14646941e-01 -4.21585351e-01 -2.29066953e-01
1.09557307e+00 -2.78706312e-01 4.82666701e-01 -2.25694124e-02
7.56220758e-01 -2.05931112e-01 1.26351848e-01 8.91401589e-01
-4.97178733e-02 3.23551327e-01 5.42523444e-01 -6.53147459e-01
-1.38323057e+00 -3.24669302e-01 2.40329042e-01 5.13361514e-01
5.05971789e-01 3.04219127e-01 -8.28159750e-01 -8.10396612e-01
-1.90583438e-01 3.63072604e-01 -4.54024047e-01 -6.09508872e-01
-6.89014077e-01 -6.76658630e-01 8.44801724e-01 -4.61829215e-01
1.32465649e+00 -1.54529083e+00 -4.22137201e-01 1.65264755e-01
-1.30666941e-01 -9.24687028e-01 -1.74596235e-01 -1.66911110e-01
-8.60823691e-01 -1.13987088e+00 -7.65678346e-01 -1.02206612e+00
7.03780890e-01 3.46062958e-01 7.43883550e-01 7.32272267e-01
5.74038550e-02 -3.46615374e-01 -5.27576149e-01 -2.87988842e-01
-1.38164580e+00 -3.67310345e-02 -4.13997680e-01 2.68473417e-01
3.63361597e-01 -2.92713851e-01 -6.21248126e-01 2.39464208e-01
-1.12520170e+00 2.03770712e-01 1.16536307e+00 9.32051480e-01
8.79402831e-02 7.70954788e-01 1.61018655e-01 -1.22461069e+00
2.54731148e-01 -5.07070363e-01 -5.69104671e-01 -7.19022155e-02
-9.41731274e-01 9.72552449e-02 4.19283003e-01 -3.71188045e-01
-4.15828586e-01 -1.90146372e-01 -2.24501982e-01 -2.12210119e-01
1.46301433e-01 4.58786815e-01 -1.30018592e-01 -9.15217757e-01
3.81217986e-01 8.55758190e-01 5.88802159e-01 4.22053412e-02
-2.58211166e-01 7.65684366e-01 1.76279142e-01 5.76253772e-01
1.22623217e+00 6.80563748e-01 2.83669949e-01 -8.26540411e-01
-1.63041264e-01 -2.06681751e-02 -6.49945363e-02 -2.30772108e-01
5.27777672e-01 -5.48488796e-01 -8.91253591e-01 1.13447940e+00
-1.17134917e+00 -3.07593364e-02 3.01363736e-01 3.31430048e-01
-3.91783595e-01 7.09660590e-01 -4.79676396e-01 -6.13009572e-01
-3.94142270e-01 -1.24211311e+00 4.73599553e-01 -2.76202619e-01
2.83353001e-01 -1.28685391e+00 -1.65260226e-01 2.07743928e-01
4.37643290e-01 8.18108141e-01 7.71856844e-01 -6.95751250e-01
-8.02543700e-01 -7.06101418e-01 -1.23053476e-01 8.13553572e-01
1.18261315e-01 -5.71835399e-01 -6.52526915e-01 -7.90868640e-01
5.48815727e-01 9.79500189e-02 1.28707004e+00 3.38689625e-01
1.16545212e+00 -8.56921196e-01 -3.63663971e-01 9.19785261e-01
1.75074732e+00 5.22391319e-01 1.51870918e+00 7.30314910e-01
7.54402578e-01 4.52561855e-01 2.93986857e-01 1.60184473e-01
-1.35336503e-01 1.03354312e-01 9.92533207e-01 -2.91166574e-01
-4.23748493e-02 -1.45747095e-01 6.40277505e-01 7.50408769e-01
-3.06508355e-02 -7.94236660e-01 -7.35900462e-01 2.43414506e-01
-1.37429404e+00 -1.17655587e+00 -2.05768406e-01 2.17432070e+00
6.20355368e-01 4.49356556e-01 -4.88719046e-01 7.82048881e-01
1.07923079e+00 7.96553195e-01 -4.20383602e-01 -6.40641034e-01
-2.20261231e-01 1.03929497e-01 1.25530112e+00 6.30637705e-01
-1.10817599e+00 9.77955461e-01 6.08614349e+00 1.09726214e+00
-1.44493866e+00 1.66972801e-01 7.14925826e-01 5.33621371e-01
-3.65745783e-01 7.26418793e-02 -4.57744449e-01 7.79931307e-01
7.52747416e-01 1.32907391e-01 6.57663763e-01 3.06249470e-01
1.78047791e-01 -8.17924179e-03 -5.96738398e-01 1.02880669e+00
4.26306903e-01 -1.47051406e+00 1.10971048e-01 7.04856277e-01
1.04236853e+00 -2.91107595e-01 3.73233318e-01 1.94820706e-02
4.50029485e-02 -1.33590150e+00 2.57191598e-01 2.38667786e-01
9.19180989e-01 -9.16861534e-01 1.14262879e+00 3.59795600e-01
-6.56425476e-01 -1.11178771e-01 -2.02459365e-01 -1.27605021e-01
-1.70608103e-01 5.43219745e-01 -8.70774567e-01 1.56660482e-01
4.18859422e-01 7.67948568e-01 -6.71576560e-02 7.54118443e-01
-4.21398759e-01 9.55102861e-01 2.13771798e-02 -6.66339621e-02
5.91114759e-01 7.72189125e-02 1.04486501e+00 1.02975237e+00
6.67682469e-01 -2.70545989e-01 -1.52084187e-01 5.91763675e-01
-2.00328052e-01 -1.90694839e-01 -1.07887554e+00 -2.12737367e-01
2.80138403e-01 7.32639492e-01 -7.75552988e-01 -3.48633915e-01
-1.56559050e-01 9.81038749e-01 -8.13498974e-01 2.45902672e-01
-7.15049684e-01 -6.55630887e-01 1.59221530e-01 3.88198555e-01
5.83398819e-01 -4.25428636e-02 -4.41189289e-01 -1.00017989e+00
-5.11850536e-01 -1.60645044e+00 -4.36727107e-02 -2.99374580e-01
-4.45723057e-01 3.60976279e-01 -3.50226015e-01 -1.57107866e+00
-2.61801392e-01 -3.59280497e-01 -6.09662831e-01 5.95189691e-01
-1.79454219e+00 -1.18340695e+00 -3.19232613e-01 5.77083051e-01
3.48821670e-01 -7.26011395e-01 5.27235329e-01 -6.75176678e-04
1.97124127e-02 5.11013031e-01 4.23610926e-01 3.14078003e-01
2.25476936e-01 -6.01790607e-01 6.40385211e-01 9.55762804e-01
-8.16824436e-02 -1.27800047e-01 1.08686864e+00 -7.19131827e-01
-1.27840459e+00 -8.18495095e-01 7.41412282e-01 9.42254812e-02
3.48156065e-01 -1.55673236e-01 -8.26686323e-01 7.61557937e-01
3.22300434e-01 -2.03624487e-01 1.31022781e-01 -8.60421896e-01
-1.48366064e-01 -8.90005659e-03 -1.53925049e+00 3.76790941e-01
4.76894289e-01 -2.53085971e-01 -3.79441053e-01 4.80506979e-02
6.58624887e-01 -2.53072917e-01 -2.23938808e-01 3.47136915e-01
6.89734876e-01 -1.09194362e+00 9.39792931e-01 1.05973221e-01
7.08434701e-01 -1.10171229e-01 2.70210914e-02 -1.30374658e+00
-8.02771896e-02 -9.12555397e-01 -2.97755510e-01 4.48291808e-01
1.29306391e-01 -1.02723420e+00 9.52358484e-01 -5.81062317e-01
3.89656156e-01 -4.40547317e-01 -8.32254589e-01 -7.90079951e-01
1.06409833e-01 -1.30820945e-01 8.72009873e-01 1.02845752e+00
-5.20645499e-01 -4.06794548e-01 -1.12766278e+00 1.62351072e-01
1.13783658e+00 -2.30149925e-01 7.50505388e-01 -1.07035160e+00
-8.56571794e-02 -3.41850877e-01 -7.90450037e-01 -7.71794379e-01
2.15863120e-02 -7.65469253e-01 5.61709963e-02 -1.14825928e+00
-1.44774184e-01 -3.96086395e-01 -4.20889735e-01 2.43503973e-01
4.15810019e-01 7.48899460e-01 1.39559656e-02 4.78157759e-01
-1.09818198e-01 9.76758748e-02 1.63082230e+00 -7.11113393e-01
-2.54484185e-04 1.68970466e-01 -5.44698775e-01 8.20552289e-01
9.84722197e-01 -1.00648129e+00 1.42148545e-03 -1.38482198e-01
4.33579385e-01 5.76243959e-02 6.04674518e-01 -1.20616472e+00
-2.80465316e-02 -6.07486442e-02 1.37745127e-01 -2.61752933e-01
4.24041227e-02 -1.07001746e+00 1.55036852e-01 1.57055151e+00
-2.34092206e-01 -6.70157135e-01 -2.42365748e-01 6.12429738e-01
-3.06641608e-01 -3.74838769e-01 1.12814331e+00 -4.38460886e-01
-1.02629709e+00 2.28635713e-01 -5.60435474e-01 -2.70741701e-01
1.16427219e+00 -6.46591663e-01 -5.47250696e-02 -8.41851354e-01
-4.14650530e-01 -2.65906006e-01 5.25239170e-01 1.51389793e-01
1.09859443e+00 -1.12962985e+00 -9.14842010e-01 6.78626597e-01
-2.18819082e-01 -5.54205060e-01 -5.76868951e-02 6.69344485e-01
-1.11577702e+00 3.38058859e-01 -4.06651050e-01 -4.34388041e-01
-1.44251275e+00 5.84108233e-01 4.79547322e-01 -5.38318396e-01
-6.78571045e-01 5.22432208e-01 1.18694767e-01 -1.98763306e-03
-4.91231540e-03 7.16584697e-02 -4.50613320e-01 -3.34743321e-01
5.73987186e-01 3.35408241e-01 -2.74547994e-01 -7.90256202e-01
6.74292892e-02 5.05400002e-01 -2.85073482e-02 8.91572759e-02
1.07628548e+00 -3.29784960e-01 -4.04808968e-01 5.27581200e-03
1.60085034e+00 -2.28276178e-01 -1.00317013e+00 -1.42367333e-01
-3.58516395e-01 -6.97384655e-01 4.10491526e-01 -4.29037541e-01
-1.45229232e+00 8.92975986e-01 9.16697562e-01 7.43170500e-01
1.05229819e+00 -5.57726085e-01 1.43496943e+00 4.97951418e-01
4.82539654e-01 -7.72485852e-01 1.12860195e-01 5.24530530e-01
5.21915257e-01 -1.53890324e+00 2.11599506e-02 -4.69480120e-02
-3.52170765e-01 1.08075333e+00 -4.34798524e-02 -5.94549239e-01
8.74072134e-01 2.23258212e-01 -1.36392161e-01 -4.07113880e-01
-2.12296426e-01 1.27105594e-01 1.49599360e-02 6.67443395e-01
-2.17774302e-01 -1.78457778e-02 -3.64267260e-01 -4.54326868e-01
-1.39418527e-01 7.09757730e-02 6.63260400e-01 8.47313821e-01
-9.10215378e-01 -8.73499334e-01 -8.73851001e-01 3.08202118e-01
-7.83781290e-01 -2.50266016e-01 -3.08600161e-02 7.73400128e-01
2.62507915e-01 9.36593652e-01 -2.22553555e-02 -9.41686690e-01
-4.78830576e-01 -4.22498018e-01 2.59237111e-01 -3.20907950e-01
-3.37028325e-01 -8.02736804e-02 -3.07953984e-01 -2.15008765e-01
-5.30281842e-01 -1.07238479e-02 -1.01638722e+00 -1.05948222e+00
-4.59334522e-01 -8.43576267e-02 9.25983369e-01 1.19630909e+00
-2.90665865e-01 4.63990152e-01 1.29131567e+00 -7.63663352e-01
-1.97824389e-01 -6.05920970e-01 -7.16536999e-01 3.09818983e-01
8.94009829e-01 -1.62406251e-01 -9.84156311e-01 8.97841528e-02] | [4.306778907775879, 8.058269500732422] |
9a84435b-adb7-4059-ac9d-852d7ea746af | whole-body-tumor-segmentation-of-18f-fdg-pet | 2210.08068 | null | https://arxiv.org/abs/2210.08068v1 | https://arxiv.org/pdf/2210.08068v1.pdf | Whole-body tumor segmentation of 18F -FDG PET/CT using a cascaded and ensembled convolutional neural networks | Background: A crucial initial processing step for quantitative PET/CT analysis is the segmentation of tumor lesions enabling accurate feature ex-traction, tumor characterization, oncologic staging, and image-based therapy response assessment. Manual lesion segmentation is however associated with enormous effort and cost and is thus infeasible in clinical routine. Goal: The goal of this study was to report the performance of a deep neural network designed to automatically segment regions suspected of cancer in whole-body 18F-FDG PET/CT images in the context of the AutoPET challenge. Method: A cascaded approach was developed where a stacked ensemble of 3D UNET CNN processed the PET/CT images at a fixed 6mm resolution. A refiner network composed of residual layers enhanced the 6mm segmentation mask to the original resolution. Results: 930 cases were used to train the model. 50% were histologically proven cancer patients and 50% were healthy controls. We obtained a dice=0.68 on 84 stratified test cases. Manual and automatic Metabolic Tumor Volume (MTV) were highly correlated (R2 = 0.969,Slope = 0.947). Inference time was 89.7 seconds on average. Conclusion: The proposed algorithm accurately segmented regions suspicious for cancer in whole-body 18F -FDG PET/CT images. | ['Lei Xiang', 'Xinrui Zhan', 'Ludovic Sibille'] | 2022-10-14 | null | null | null | null | ['tumor-segmentation'] | ['computer-vision'] | [ 3.53120446e-01 3.93376082e-01 -4.63236779e-01 -4.68745977e-01
-9.31020975e-01 -4.64615554e-01 4.80643839e-01 3.93720984e-01
-8.86300147e-01 9.23219323e-01 -3.96615453e-02 -4.90230978e-01
1.18066393e-01 -6.40293241e-01 -3.35841984e-01 -1.00560844e+00
1.06504537e-01 1.02084661e+00 3.54710191e-01 4.32382911e-01
-5.28976507e-02 9.01281953e-01 -4.17348325e-01 3.92974436e-01
6.90419912e-01 1.06503427e+00 6.66929543e-01 9.17071640e-01
1.32288396e-01 7.28053987e-01 -2.38593727e-01 -9.28712413e-02
9.85836238e-02 -6.06764913e-01 -9.20403302e-01 2.21811589e-02
1.89917833e-01 -3.76959324e-01 -2.91759968e-01 9.03870285e-01
4.79796052e-01 -6.39676452e-02 9.76099312e-01 -8.12777579e-01
1.28200948e-01 7.00482309e-01 -5.56753516e-01 7.27271199e-01
-3.77320379e-01 3.01663965e-01 2.34564975e-01 -6.12287223e-01
4.48286355e-01 3.39004189e-01 8.73365879e-01 4.38141078e-01
-1.00984192e+00 -5.90449154e-01 -4.35105234e-01 -8.98863152e-02
-1.36336946e+00 -2.03313231e-01 -9.53882188e-02 -5.39560974e-01
1.02095544e+00 4.71545190e-01 1.04753149e+00 8.09497178e-01
9.20910418e-01 3.09354007e-01 1.07219851e+00 -1.41691446e-01
8.49993601e-02 1.39721528e-01 -5.62719852e-02 8.25967491e-01
3.52701247e-01 1.40904203e-01 2.71128356e-01 1.22419164e-01
1.06830645e+00 -6.73847124e-02 -6.38567507e-02 -1.89702854e-01
-1.20333421e+00 9.32014346e-01 9.24090803e-01 6.63652539e-01
-5.23433685e-01 4.56595838e-01 6.55074775e-01 -2.30143011e-01
5.80529749e-01 4.74663489e-02 1.30365744e-01 3.82807791e-01
-1.16496348e+00 -2.52835155e-01 1.33323446e-01 3.66517276e-01
-8.33433941e-02 -1.49853751e-01 -2.46247545e-01 5.08483171e-01
3.35454345e-01 3.54440778e-01 1.24210000e+00 -4.53220367e-01
5.64314309e-04 3.92307550e-01 -1.34826273e-01 -2.82197267e-01
-1.02019596e+00 -6.96168423e-01 -1.09279883e+00 1.49197131e-02
5.26708305e-01 -8.26392844e-02 -1.44603610e+00 1.14967704e+00
2.66276687e-01 -1.38871297e-01 -3.14467043e-01 1.08041644e+00
7.29550898e-01 1.18534014e-01 5.72866678e-01 -3.42455238e-01
1.35131466e+00 -1.00059950e+00 -4.06122446e-01 5.82531653e-02
1.03250682e+00 -5.00597417e-01 4.54309911e-01 -1.16798356e-01
-1.26710176e+00 -9.10658613e-02 -8.48375261e-01 2.60241479e-01
-1.87456727e-01 3.00909728e-01 7.61870742e-01 1.04074538e+00
-1.17727149e+00 6.11092210e-01 -1.28242350e+00 -4.32336092e-01
8.85752320e-01 9.40682828e-01 -4.04308677e-01 -1.04428250e-02
-8.83289158e-01 1.40641117e+00 8.60618532e-01 6.03416152e-02
-1.25384605e+00 -9.39025998e-01 -5.03868937e-01 -2.60839045e-01
1.29800707e-01 -1.00632763e+00 1.54233813e+00 -1.06257999e+00
-1.26545227e+00 1.11357415e+00 -3.08719818e-02 -9.03272390e-01
9.32113826e-01 6.59226894e-01 -2.00128369e-02 3.70572537e-01
2.35694051e-01 8.27754080e-01 5.30650020e-01 -6.52975023e-01
-7.13489532e-01 -6.98158264e-01 -8.09023976e-01 6.31722987e-01
4.45984215e-01 -1.11856639e-01 -1.76302329e-01 -2.94245929e-01
2.42268026e-01 -9.71441746e-01 -5.38228631e-01 -9.02814674e-04
-1.01417847e-01 1.44410923e-01 6.65357113e-01 -8.95718753e-01
6.51251972e-01 -1.46092439e+00 -2.85236359e-01 4.65740770e-01
5.33305347e-01 3.30755040e-02 3.83453876e-01 -5.72336197e-01
-3.94488573e-01 3.06581020e-01 -5.55712759e-01 3.60663272e-02
-2.87598938e-01 -4.30013612e-02 6.82421088e-01 9.70610440e-01
-1.14507973e-01 1.48217750e+00 -8.85052025e-01 -8.66963804e-01
6.02693260e-01 3.56432229e-01 -1.26396745e-01 -1.03715882e-01
5.99416792e-02 6.98542655e-01 -2.55422384e-01 6.50270641e-01
4.21266526e-01 -1.19783685e-01 2.41625398e-01 -3.23876321e-01
-9.06308144e-02 -1.56184450e-01 -3.52649510e-01 1.61312795e+00
-3.64765435e-01 7.57201970e-01 9.84651595e-02 -7.58923352e-01
2.91665614e-01 4.13596213e-01 1.25130272e+00 -1.04781997e+00
7.66523421e-01 4.44464386e-01 6.10598445e-01 -2.47079849e-01
6.94731772e-02 -7.70385265e-01 1.37633845e-01 3.36181223e-01
-2.58887589e-01 -5.02488494e-01 -5.25711067e-02 -9.30252671e-02
1.20446062e+00 -2.24754125e-01 4.67236578e-01 -3.67552221e-01
4.83084857e-01 5.11871040e-01 1.09567218e-01 5.01511693e-01
-7.10524261e-01 6.11429334e-01 3.01762283e-01 -5.05517900e-01
-1.10209501e+00 -1.07309866e+00 -4.80692327e-01 4.06685144e-01
-3.39005113e-01 3.56400639e-01 -8.43670189e-01 -9.10879672e-01
-3.57031703e-01 7.60041237e-01 -9.87530529e-01 8.32071155e-02
-5.93057513e-01 -1.45477033e+00 4.84770805e-01 5.97362399e-01
4.97602016e-01 -9.24980164e-01 -8.76956284e-01 5.71592450e-01
-3.15723419e-01 -1.05954194e+00 -1.47803679e-01 7.33542979e-01
-1.27992284e+00 -1.19908333e+00 -1.16422236e+00 -5.64289033e-01
9.62939560e-01 -1.17629245e-02 1.05629492e+00 -3.61396521e-02
-6.25168681e-01 -3.92471766e-03 6.17549159e-02 -3.38925391e-01
-7.06406057e-01 8.07532966e-02 -2.07648173e-01 -5.91119051e-01
2.13885074e-03 -1.73427939e-01 -7.95456290e-01 5.89080334e-01
-9.05218601e-01 3.39430928e-01 9.05474603e-01 7.73209155e-01
1.16877699e+00 1.99169144e-01 6.96534961e-02 -6.88137293e-01
2.80865431e-01 -4.07551020e-01 -4.50256854e-01 1.73909813e-02
-3.31769913e-01 -5.32921314e-01 4.12879914e-01 -1.69190943e-01
-6.50055468e-01 5.13245761e-01 -1.69069171e-01 -1.48686454e-01
-7.84412026e-02 2.60462195e-01 3.10725331e-01 -3.82248044e-01
7.39677131e-01 9.92858186e-02 1.35596991e-01 4.20267045e-01
-9.92368236e-02 1.59177020e-01 5.36524475e-01 -2.02986598e-02
2.08803371e-01 5.56817174e-01 5.10512352e-01 -6.38254106e-01
-4.54447001e-01 -3.59060675e-01 -1.00048447e+00 -3.38237137e-01
1.22602713e+00 -6.42452955e-01 -4.63123649e-01 2.57437259e-01
-6.69598043e-01 -9.00557160e-01 -4.25360352e-01 8.40724945e-01
-7.15468585e-01 3.18336487e-02 -5.03285944e-01 -1.85247451e-01
-8.22412431e-01 -1.55583251e+00 8.03550065e-01 2.43972033e-01
-4.19624656e-01 -1.14828956e+00 -6.81567192e-02 2.81061381e-01
6.16008043e-01 6.66516304e-01 6.99220657e-01 -7.69374549e-01
-1.74233019e-02 -4.53262895e-01 -5.71230710e-01 -3.61742415e-02
1.17901638e-01 -9.22486410e-02 -8.78189921e-01 5.40226921e-02
7.28018284e-02 -2.00200319e-01 7.65343726e-01 1.24772167e+00
1.35360086e+00 3.12634945e-01 -6.45775378e-01 6.72547162e-01
1.46381521e+00 5.60129225e-01 3.86497796e-01 1.25260502e-01
7.57534146e-01 5.25695011e-02 3.53076667e-01 7.10324794e-02
-1.05593368e-01 4.41364288e-01 8.11115682e-01 -2.13634655e-01
-2.89783567e-01 3.85637768e-02 -9.50237885e-02 1.87934116e-01
-1.60788774e-01 -4.03082281e-01 -1.26391995e+00 4.64496702e-01
-1.08953714e+00 -6.16160274e-01 -7.03707814e-01 1.92655408e+00
3.35304737e-01 2.89295018e-01 5.80711253e-02 -1.01564573e-02
8.15547526e-01 -2.82939970e-01 -7.23591030e-01 -2.99869388e-01
2.23120883e-01 3.42100203e-01 1.04487038e+00 6.38732970e-01
-1.13927996e+00 5.74565411e-01 6.03628254e+00 8.96175265e-01
-1.21682656e+00 5.13416231e-01 1.25431192e+00 -2.56797642e-01
1.50464848e-01 -5.02285600e-01 -3.79985482e-01 3.70585889e-01
1.32240033e+00 1.12572089e-01 -1.50838941e-01 4.59383518e-01
4.38965708e-01 -8.07494342e-01 -9.99014914e-01 8.11793447e-01
-1.84327960e-01 -1.37604570e+00 -2.76704848e-01 4.06141967e-01
6.89128578e-01 4.83366072e-01 4.00539897e-02 9.86624062e-02
1.15225509e-01 -1.46073568e+00 3.85813475e-01 5.01167595e-01
1.18366635e+00 -6.12517595e-01 1.21296167e+00 4.52886581e-01
-8.44539940e-01 3.42635483e-01 -1.71206325e-01 5.67365944e-01
8.57578516e-02 4.26248163e-01 -1.87653518e+00 4.19130653e-01
2.30074704e-01 2.90425956e-01 -5.61191738e-01 1.02579558e+00
1.98114336e-01 4.19543505e-01 -6.51840150e-01 -3.25365714e-03
5.64731956e-01 1.58113733e-01 -2.92899180e-02 1.24328840e+00
3.37721229e-01 4.34073150e-01 -2.94978470e-01 6.28016531e-01
2.05584317e-02 9.84543562e-02 -1.31347761e-01 -8.32559168e-02
-1.00431189e-01 1.54192126e+00 -1.68871927e+00 -4.87592936e-01
3.83593999e-02 9.56677735e-01 -1.25116542e-01 -2.86560237e-01
-1.35535347e+00 3.85900646e-01 -2.99360901e-01 4.94576514e-01
-1.16680540e-01 2.08711281e-01 -5.91377020e-01 -6.75892293e-01
-6.86991334e-01 -2.98735708e-01 5.77503383e-01 -7.99754679e-01
-8.84917974e-01 6.57798946e-01 -4.06056307e-02 -6.43465221e-01
-1.66156739e-01 -5.55726767e-01 -7.46969879e-01 1.02419221e+00
-1.24738801e+00 -9.25267041e-01 -7.06958294e-01 4.36260313e-01
4.52521205e-01 2.34143332e-01 9.68959808e-01 1.18173890e-01
-4.76659626e-01 4.27741915e-01 -4.50024717e-02 5.91638684e-02
2.88360447e-01 -1.37487280e+00 -8.06687772e-02 4.32024002e-01
-5.12852252e-01 -4.76032123e-02 1.58381894e-01 -8.94686103e-01
-9.11781192e-01 -1.35221970e+00 4.67006743e-01 -3.66888434e-01
6.91839337e-01 3.20624173e-01 -3.99053097e-01 5.23510754e-01
1.34320274e-01 4.96024877e-01 8.13959897e-01 -1.17432284e+00
6.89847946e-01 2.11120516e-01 -1.78483939e+00 2.29142502e-01
3.34966630e-01 -1.04585499e-01 -1.40542537e-01 4.79496688e-01
9.77798104e-02 -8.88419807e-01 -1.17909348e+00 5.39735734e-01
5.32679617e-01 -8.88815939e-01 9.66711640e-01 -1.88786030e-01
2.22102389e-01 -6.69464618e-02 -1.43969774e-01 -1.07439649e+00
-4.26918358e-01 1.49363324e-01 2.96905071e-01 1.29552186e-01
4.80207175e-01 -4.26373243e-01 1.25083911e+00 7.79565632e-01
-5.60618699e-01 -9.32705522e-01 -1.11147916e+00 -2.77370453e-01
3.82784337e-01 -5.59547007e-01 -7.89288711e-03 7.30530500e-01
-5.04851460e-01 -1.61641896e-01 6.18720114e-01 2.37751435e-02
5.30047596e-01 -4.73633051e-01 -4.84793596e-02 -8.69459033e-01
8.12536180e-02 -9.56468344e-01 -4.63573962e-01 -1.65090948e-01
-1.85302213e-01 -1.29754961e+00 -3.22398007e-01 -1.85258412e+00
5.07760882e-01 -3.50523680e-01 -3.51621568e-01 3.27579319e-01
2.05834076e-01 6.57035947e-01 -1.96231067e-01 1.17987618e-01
-1.66775435e-02 1.58255979e-01 1.62038195e+00 -1.40743300e-01
3.63420434e-02 3.41997683e-01 -2.68591374e-01 8.25949490e-01
1.04316056e+00 -6.01444483e-01 -3.70195895e-01 8.88958052e-02
-3.61148924e-01 5.20117939e-01 6.52017057e-01 -1.11007440e+00
-5.69810122e-02 -2.31424898e-01 1.31173849e+00 -1.02659798e+00
1.81789756e-01 -1.15487170e+00 4.94032949e-01 1.17976141e+00
-9.64078531e-02 1.06269130e-02 4.08516169e-01 2.01520119e-02
1.07851572e-01 -4.44660008e-01 1.24024081e+00 -7.31468201e-01
-3.14342320e-01 5.87320566e-01 -6.24611795e-01 -3.32578599e-01
1.46469688e+00 -4.58649129e-01 2.74607614e-02 -2.80167460e-02
-1.11156905e+00 -2.02236772e-01 3.02265197e-01 -3.95799398e-01
4.58717823e-01 -1.23031652e+00 -5.75985730e-01 -2.65531927e-01
-4.14174378e-01 3.83160859e-01 3.38594258e-01 1.73619390e+00
-1.15409386e+00 7.77054012e-01 -3.36975753e-01 -8.17977905e-01
-1.06341505e+00 1.74719781e-01 1.29177344e+00 -6.96097374e-01
-6.40073359e-01 1.10474980e+00 2.85530299e-01 8.33248661e-04
-1.74885556e-01 -7.78505266e-01 9.20499787e-02 1.43163234e-01
2.38910347e-01 9.44299102e-02 3.58511955e-01 -8.59158278e-01
-3.67802173e-01 3.46226752e-01 -1.69285655e-01 -2.69948900e-01
1.10990036e+00 -1.68041270e-02 -7.26278592e-03 3.33336778e-02
1.24907255e+00 -6.26853824e-01 -9.97132480e-01 2.21898258e-01
-3.23112965e-01 -3.85737084e-02 7.53669918e-01 -1.31756270e+00
-1.11240506e+00 6.69418275e-01 1.20975089e+00 -2.98244283e-02
1.15655565e+00 -1.87402442e-01 5.40280282e-01 -1.67048424e-01
-3.31117362e-02 -6.70498669e-01 -2.98561305e-01 2.28935272e-01
5.62702000e-01 -1.37931323e+00 2.32773826e-01 -3.96237940e-01
-6.55565262e-01 1.16150177e+00 6.36303008e-01 -1.07642449e-01
2.66078413e-01 3.32115710e-01 -1.07166938e-01 -3.01992178e-01
-2.93632865e-01 8.37407485e-02 2.95019150e-01 4.67825592e-01
5.75962961e-01 3.87546122e-01 -1.49103895e-01 5.16713083e-01
-2.50621170e-01 1.40775247e-02 3.82067084e-01 7.97632992e-01
-5.18880010e-01 -4.63176221e-01 -3.38405877e-01 7.03634322e-01
-6.42382681e-01 1.99485268e-03 -3.87648523e-01 1.26550603e+00
1.73581630e-01 2.57783353e-01 1.33959353e-01 1.69477671e-01
-1.65827235e-03 9.31681544e-02 7.10092425e-01 -4.32574868e-01
-9.81258869e-01 2.81417429e-01 -9.46070626e-02 -3.27464610e-01
-3.47995520e-01 -6.52189910e-01 -1.62651050e+00 -1.11885250e-01
-2.80045301e-01 2.35494804e-02 1.19783449e+00 8.40005577e-01
-4.41607028e-01 1.01766455e+00 4.67347413e-01 -7.73606122e-01
-7.71558359e-02 -1.05822480e+00 -4.52724516e-01 -4.04135168e-01
1.06312841e-01 -4.73099262e-01 -1.06995337e-01 -1.09968714e-01] | [14.719883918762207, -2.4950523376464844] |
29b45e59-fa00-41cc-b49b-76d8f9713a2d | temporal-cycle-consistency-learning | 1904.07846 | null | http://arxiv.org/abs/1904.07846v1 | http://arxiv.org/pdf/1904.07846v1.pdf | Temporal Cycle-Consistency Learning | We introduce a self-supervised representation learning method based on the
task of temporal alignment between videos. The method trains a network using
temporal cycle consistency (TCC), a differentiable cycle-consistency loss that
can be used to find correspondences across time in multiple videos. The
resulting per-frame embeddings can be used to align videos by simply matching
frames using the nearest-neighbors in the learned embedding space.
To evaluate the power of the embeddings, we densely label the Pouring and
Penn Action video datasets for action phases. We show that (i) the learned
embeddings enable few-shot classification of these action phases, significantly
reducing the supervised training requirements; and (ii) TCC is complementary to
other methods of self-supervised learning in videos, such as Shuffle and Learn
and Time-Contrastive Networks. The embeddings are also used for a number of
applications based on alignment (dense temporal correspondence) between video
pairs, including transfer of metadata of synchronized modalities between videos
(sounds, temporal semantic labels), synchronized playback of multiple videos,
and anomaly detection. Project webpage:
https://sites.google.com/view/temporal-cycle-consistency . | ['Yusuf Aytar', 'Debidatta Dwibedi', 'Jonathan Tompson', 'Andrew Zisserman', 'Pierre Sermanet'] | 2019-04-16 | temporal-cycle-consistency-learning-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Dwibedi_Temporal_Cycle-Consistency_Learning_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Dwibedi_Temporal_Cycle-Consistency_Learning_CVPR_2019_paper.pdf | cvpr-2019-6 | ['video-alignment'] | ['computer-vision'] | [ 7.40203559e-02 -2.71546364e-01 -3.31147939e-01 -4.84190613e-01
-5.79246998e-01 -6.70228124e-01 6.64248049e-01 1.32358328e-01
-2.96086371e-01 2.20826477e-01 4.56793875e-01 2.84702510e-01
-2.35636368e-01 -2.72132933e-01 -8.25723052e-01 -4.38939720e-01
-7.47729778e-01 4.74583134e-02 2.64145613e-01 1.12730205e-01
2.07176730e-01 3.86123449e-01 -1.83217835e+00 4.81457800e-01
2.47241437e-01 1.09216487e+00 -3.30783948e-02 8.28206420e-01
-1.01149743e-02 1.02847838e+00 -1.11006148e-01 -2.36545466e-02
4.33352560e-01 -7.22766697e-01 -9.07688677e-01 -1.42814502e-01
7.49118268e-01 -2.64430135e-01 -9.12808955e-01 9.35543835e-01
9.39764753e-02 4.65938956e-01 4.92253125e-01 -1.79816353e+00
-5.81619501e-01 2.25934237e-01 -2.36711666e-01 6.02951705e-01
7.12814569e-01 1.57009289e-02 1.21903408e+00 -8.08092296e-01
7.49746978e-01 9.55139816e-01 9.93542731e-01 4.98316169e-01
-1.05841053e+00 -4.93960142e-01 -3.58615555e-02 8.59333694e-01
-1.18350124e+00 -6.86584413e-01 6.45266950e-01 -6.85907245e-01
9.60044682e-01 2.01033533e-01 9.67312753e-01 1.25829935e+00
1.35327563e-01 7.21930265e-01 5.85170746e-01 -2.42602363e-01
1.39392063e-01 -2.84887493e-01 -1.81152180e-01 7.20377028e-01
-3.87919158e-01 9.68639776e-02 -9.86345530e-01 5.93161099e-02
6.95506930e-01 4.25588936e-01 -2.74865031e-01 -5.90282202e-01
-1.73669803e+00 6.30362391e-01 1.27232879e-01 5.83088756e-01
-4.14070413e-02 4.98122185e-01 9.21525955e-01 7.14214742e-01
3.72009009e-01 6.91766560e-01 -3.47472370e-01 -6.49833679e-01
-8.85688245e-01 2.43160538e-02 6.68150127e-01 9.94677842e-01
7.86970079e-01 8.91783461e-02 -5.08034714e-02 4.24490303e-01
1.67947650e-01 1.07493185e-01 1.01625276e+00 -1.48854542e+00
1.77443042e-01 2.47691095e-01 -4.74789217e-02 -1.18899608e+00
-2.72505462e-01 4.50527638e-01 -5.36450446e-01 -6.96044117e-02
4.14818466e-01 3.70680511e-01 -4.08054024e-01 1.81405723e+00
1.67112336e-01 9.84260321e-01 -2.13545375e-02 6.73665404e-01
5.03377557e-01 6.52041078e-01 -2.12394595e-01 -3.21045130e-01
9.91918504e-01 -9.93298173e-01 -9.52944517e-01 -3.73640880e-02
8.90798807e-01 -6.45576596e-01 9.88763630e-01 6.24183714e-02
-1.04534423e+00 -8.07137311e-01 -1.08970141e+00 -1.51284235e-02
-4.82490569e-01 -3.33912671e-01 4.00932759e-01 -3.39817852e-02
-1.11390126e+00 1.33783722e+00 -1.20752943e+00 -6.02791250e-01
2.48620555e-01 2.10081890e-01 -9.18779552e-01 8.34864900e-02
-1.10080945e+00 6.89423501e-01 3.55969787e-01 -1.01675093e-01
-8.08016181e-01 -9.22854781e-01 -1.33859110e+00 -2.54732192e-01
2.89055612e-03 -1.68912187e-01 1.14652979e+00 -1.41500485e+00
-1.41845965e+00 1.00602853e+00 -1.11578651e-01 -4.94857103e-01
2.82965869e-01 -5.78963041e-01 -6.63836896e-01 6.05683029e-01
2.79832959e-01 6.89740539e-01 1.01693523e+00 -5.49629807e-01
-6.90871954e-01 -2.04854578e-01 4.19962406e-03 1.71752572e-01
-6.44676447e-01 -2.21619196e-02 -3.28789979e-01 -7.34793186e-01
1.00093991e-01 -9.83606517e-01 2.54465193e-01 4.61358070e-01
-1.42695820e-02 -1.76361933e-01 1.23807216e+00 -7.72901893e-01
1.03813887e+00 -2.55031919e+00 4.24038887e-01 -1.41335681e-01
7.74662197e-03 -5.08884154e-02 -4.38728452e-01 5.65924823e-01
-5.24146616e-01 -6.07581288e-02 -1.88391823e-02 -2.76634216e-01
-1.36973634e-01 4.16759104e-01 -2.36066252e-01 8.99935603e-01
2.72280991e-01 7.79445767e-01 -1.37262833e+00 -4.80683625e-01
3.00295740e-01 2.86270827e-01 -5.91137052e-01 6.67571902e-01
1.86465941e-02 6.47489190e-01 1.04976356e-01 4.11489159e-01
-9.39826015e-03 -8.04148838e-02 2.30416387e-01 -7.72215724e-01
-1.41688928e-01 2.62238413e-01 -1.06742048e+00 2.27737904e+00
-3.32281709e-01 1.24774349e+00 -5.47197223e-01 -1.29563904e+00
5.80221355e-01 4.59626734e-01 1.15421653e+00 -7.06125975e-01
-1.14356250e-01 2.85811014e-02 -2.49333739e-01 -1.00661373e+00
5.07150233e-01 2.71391898e-01 -6.49689883e-02 6.68994308e-01
6.35848999e-01 4.57852185e-02 4.84205782e-01 1.19222924e-01
1.24553204e+00 5.89433670e-01 1.85929969e-01 -2.52005029e-02
2.63965100e-01 -1.82034388e-01 5.83649457e-01 2.64327079e-01
-4.94942307e-01 7.22371578e-01 4.01434839e-01 -7.49172568e-01
-1.30829465e+00 -1.04250181e+00 2.17934102e-01 1.22223175e+00
2.29135767e-01 -7.91185975e-01 -4.27953959e-01 -6.38890505e-01
2.15607099e-02 1.67540848e-01 -7.61549652e-01 -3.33662897e-01
-8.08207273e-01 1.90265298e-01 5.36824644e-01 8.11913908e-01
2.43833154e-01 -1.01949024e+00 -6.72716200e-01 3.57684195e-02
-3.41527492e-01 -1.25260651e+00 -8.09038877e-01 2.02331647e-01
-8.24144483e-01 -1.47338998e+00 -3.62946421e-01 -9.70686555e-01
3.76442075e-01 3.24098915e-01 9.16530848e-01 -8.94864351e-02
-3.15378249e-01 1.18620193e+00 -6.13748372e-01 3.31895739e-01
-3.91743958e-01 -4.86238033e-01 6.77537858e-01 8.34589601e-02
3.42508525e-01 -9.27590191e-01 -5.96352279e-01 5.23473799e-01
-8.74785066e-01 -1.53255761e-01 -7.68122226e-02 8.80370557e-01
7.30765104e-01 -3.32350045e-01 1.46436945e-01 -3.49641204e-01
1.12301826e-01 -5.96550524e-01 -3.30375403e-01 1.56735986e-01
-1.23605698e-01 -5.98596185e-02 5.37241042e-01 -7.59786665e-01
-2.69870520e-01 4.06367369e-02 3.32193255e-01 -1.22806048e+00
-1.24976777e-01 2.19514847e-01 2.41900057e-01 6.51997924e-02
5.15764773e-01 1.09656617e-01 3.69791240e-01 -1.94466561e-01
6.41223490e-01 4.06259805e-01 9.60080028e-01 -2.59238303e-01
8.35005820e-01 6.13460541e-01 -2.21007079e-01 -6.94212317e-01
-7.78292954e-01 -7.96318054e-01 -1.08066761e+00 -4.51191545e-01
1.11669886e+00 -8.71684194e-01 -5.34970403e-01 9.68056843e-02
-9.74291563e-01 -5.51956058e-01 -7.52357483e-01 9.30652499e-01
-1.23741961e+00 6.68147564e-01 -6.36133254e-01 -2.25885987e-01
1.69291511e-01 -7.12690830e-01 7.67056048e-01 1.75450802e-01
-6.90879703e-01 -1.12550664e+00 5.90775311e-01 -1.01511799e-01
1.06990375e-02 3.97946447e-01 5.76163650e-01 -8.03700924e-01
-2.85368025e-01 -1.29445329e-01 1.67209521e-01 5.18328071e-01
4.81700301e-01 2.36518070e-01 -9.24666941e-01 -5.09358883e-01
-1.52576149e-01 -4.26855385e-01 4.48996514e-01 3.06450009e-01
1.34255803e+00 -4.73738432e-01 -2.38899171e-01 7.95589805e-01
1.08082914e+00 1.82015583e-01 6.99429929e-01 5.17300546e-01
8.76076818e-01 6.59093797e-01 6.56009138e-01 4.59484279e-01
1.14062473e-01 8.48387420e-01 3.71616960e-01 3.05746973e-01
-1.13119975e-01 -4.09898549e-01 7.46431947e-01 1.29086673e+00
-5.40114641e-02 1.65879354e-01 -6.19657338e-01 7.21246898e-01
-2.17179322e+00 -1.57584906e+00 4.09518033e-01 2.20517516e+00
6.34340525e-01 -2.11343884e-01 2.79604644e-01 2.70910084e-01
8.11118603e-01 4.32365388e-01 -5.05749762e-01 -3.52114975e-01
1.59251869e-01 -3.19123687e-03 2.11326599e-01 1.78861260e-01
-1.47074258e+00 5.64784288e-01 6.34101200e+00 4.21433479e-01
-8.34300816e-01 3.69386494e-01 1.95649087e-01 -2.30097204e-01
5.90276383e-02 -3.78951356e-02 -1.76699832e-01 6.85491145e-01
1.11915755e+00 -1.92630813e-01 4.32831854e-01 7.96306133e-01
2.21200675e-01 2.79572666e-01 -1.92041051e+00 1.19582498e+00
3.18975538e-01 -1.48709023e+00 -2.53178179e-01 -3.74238968e-01
6.85021877e-01 -9.06483736e-03 -6.29396215e-02 2.17747033e-01
-5.52225374e-02 -8.33403111e-01 7.16902494e-01 5.86182773e-01
9.25478458e-01 -3.94306898e-01 4.33234304e-01 -2.66095787e-01
-1.52618396e+00 -1.77246153e-01 -1.90146565e-01 7.50424480e-03
3.21324587e-01 8.14595371e-02 -3.25401694e-01 3.62618744e-01
1.19736385e+00 1.67582822e+00 -3.64096642e-01 1.02473378e+00
1.11157492e-01 2.71842808e-01 -1.64368674e-01 4.56303954e-01
1.52514160e-01 -2.73816019e-01 6.33319139e-01 1.26758611e+00
4.80196327e-01 -3.15213114e-01 1.27759784e-01 3.68723512e-01
4.33940999e-02 -1.78371862e-01 -1.03572798e+00 -3.71854156e-01
4.88720596e-01 9.81306672e-01 -4.20901418e-01 -2.81833202e-01
-6.07166409e-01 1.34889805e+00 2.56091416e-01 1.62588775e-01
-9.46538985e-01 -3.03332627e-01 1.10238564e+00 -4.45142277e-02
3.94835055e-01 -3.57003570e-01 5.45403481e-01 -1.10582507e+00
1.57098800e-01 -7.86846042e-01 5.90813875e-01 -8.69422019e-01
-1.39947927e+00 3.09253067e-01 1.35518029e-01 -2.05952501e+00
-6.32166266e-01 -5.03591061e-01 -7.78165460e-01 -4.28601392e-02
-1.15183520e+00 -7.51850903e-01 -4.43756968e-01 9.24090624e-01
7.14289308e-01 -3.91408861e-01 9.91185665e-01 4.97969627e-01
-3.49197149e-01 5.07168710e-01 1.58626691e-01 3.42873663e-01
9.69294608e-01 -1.30201602e+00 2.34353557e-01 6.71168506e-01
5.71373403e-01 3.06987822e-01 5.63697040e-01 -2.83814728e-01
-1.46581733e+00 -1.20185161e+00 5.51446795e-01 -4.28089172e-01
1.10031688e+00 -2.19224557e-01 -1.03629279e+00 9.30041194e-01
2.20440432e-01 5.37877142e-01 1.00668752e+00 -1.35016993e-01
-7.39409328e-01 -3.02960366e-01 -8.08237672e-01 5.40335417e-01
1.22474778e+00 -1.06469810e+00 -6.61652803e-01 6.65286899e-01
8.34685385e-01 -3.74797910e-01 -1.15234721e+00 1.22386426e-01
9.05706167e-01 -1.04115188e+00 9.53871608e-01 -9.29095685e-01
5.08752882e-01 -3.11128199e-01 -4.09080893e-01 -1.20203388e+00
-3.61763775e-01 -9.17091131e-01 -4.30263966e-01 1.09601760e+00
-8.94293264e-02 -2.84986109e-01 4.98321176e-01 2.49658212e-01
-2.47119322e-01 -4.07402635e-01 -1.14143384e+00 -1.20979071e+00
-4.40146565e-01 -3.70004058e-01 2.43508860e-01 1.31478357e+00
4.76778299e-01 -1.93645641e-01 -5.04160941e-01 8.63957107e-02
2.24285156e-01 -2.35683508e-02 6.67555451e-01 -9.56486404e-01
-2.77976751e-01 -3.32996637e-01 -1.15106785e+00 -8.92096221e-01
4.61356342e-01 -9.48229671e-01 8.97950605e-02 -9.94753063e-01
-1.86932161e-02 -1.07352912e-01 -5.27535856e-01 5.23644507e-01
3.95283401e-01 4.37086374e-01 1.39983863e-01 4.79582071e-01
-1.04190290e+00 5.68354368e-01 7.16415942e-01 -1.43899709e-01
-8.23092461e-02 -3.42928916e-01 1.22258611e-01 7.18463898e-01
8.25175345e-01 -5.21018326e-01 -3.01069856e-01 -3.18960428e-01
8.81084129e-02 -4.40435037e-02 4.50539678e-01 -1.34646297e+00
2.53587514e-01 -1.14255853e-01 2.14330226e-01 -2.86799371e-01
4.72074181e-01 -9.23397720e-01 2.81113416e-01 4.67247784e-01
-5.09914219e-01 6.29447341e-01 1.85406268e-01 8.55701983e-01
-4.63799655e-01 -1.69362739e-01 6.54332280e-01 -2.73431204e-02
-1.17445087e+00 4.05610025e-01 -3.05415243e-01 2.03762576e-01
1.25319028e+00 -4.72048402e-01 -1.26532897e-01 -4.74595934e-01
-7.99458921e-01 5.61925210e-02 4.90882218e-01 7.55842388e-01
6.79589808e-01 -1.94860661e+00 -2.09400967e-01 2.01051906e-01
4.13649470e-01 -4.17575717e-01 2.55900502e-01 1.08942974e+00
-5.07676065e-01 -2.16204952e-02 -6.01627588e-01 -9.71122861e-01
-1.21020985e+00 4.92379487e-01 3.88325721e-01 1.85357556e-01
-8.69977772e-01 5.78738749e-01 -7.23029524e-02 -1.80010155e-01
3.37734610e-01 -1.97861075e-01 -1.84239060e-01 3.04419279e-01
6.63196921e-01 4.64374751e-01 -2.63855934e-01 -7.17327476e-01
-4.55071568e-01 6.99010789e-01 1.43634900e-01 -3.01981010e-02
1.41926181e+00 -1.40195042e-01 -4.50422838e-02 1.02732837e+00
1.83371627e+00 -4.05893445e-01 -1.48156762e+00 -2.67657280e-01
6.29183501e-02 -6.63614988e-01 -3.57211053e-01 1.42328471e-01
-9.76610422e-01 7.06195831e-01 9.03061271e-01 2.60624915e-01
8.40033352e-01 -5.90343494e-03 7.52318025e-01 4.66824800e-01
-1.54397509e-03 -1.26564741e+00 8.36395919e-01 4.16950941e-01
7.17687726e-01 -1.14554572e+00 -1.41925931e-01 1.81668282e-01
-5.48560083e-01 1.39918017e+00 5.13518691e-01 -3.46481264e-01
6.98013783e-01 2.49782093e-02 -8.28937143e-02 -2.63814747e-01
-6.20317817e-01 -7.22140446e-02 2.75461853e-01 8.69890034e-01
3.35279047e-01 -2.92913795e-01 1.88958302e-01 -4.96982634e-02
8.58297497e-02 -1.53897762e-01 5.46594262e-01 1.05378568e+00
-8.67930949e-02 -8.22161257e-01 1.07938714e-01 4.24345106e-01
-1.52724922e-01 2.06226483e-01 3.26549523e-02 8.24622989e-01
-2.80606076e-02 6.79787695e-01 7.65264332e-01 -6.13558590e-01
2.27489859e-01 1.93498284e-01 5.76161265e-01 -5.90124667e-01
-1.90191314e-01 -2.23394617e-01 -4.69839722e-02 -1.32530665e+00
-1.01008272e+00 -7.12418139e-01 -9.03638959e-01 -2.86643207e-01
3.11129261e-02 6.54951483e-02 3.14738303e-01 8.18827927e-01
4.32052732e-01 2.43843183e-01 8.94464970e-01 -1.17057538e+00
-1.77760780e-01 -8.17261100e-01 -6.01472497e-01 9.39189196e-01
4.57953155e-01 -8.28773379e-01 -6.27967417e-01 5.78042865e-01] | [8.69175910949707, 0.7312559485435486] |
d15988fa-cfd5-436e-ac57-a9e6f8dc2eee | temporal-feature-networks-for-cnn-based | 2103.12213 | null | https://arxiv.org/abs/2103.12213v1 | https://arxiv.org/pdf/2103.12213v1.pdf | Temporal Feature Networks for CNN based Object Detection | For reliable environment perception, the use of temporal information is essential in some situations. Especially for object detection, sometimes a situation can only be understood in the right perspective through temporal information. Since image-based object detectors are currently based almost exclusively on CNN architectures, an extension of their feature extraction with temporal features seems promising. Within this work we investigate different architectural components for a CNN-based temporal information extraction. We present a Temporal Feature Network which is based on the insights gained from our architectural investigations. This network is trained from scratch without any ImageNet information based pre-training as these images are not available with temporal information. The object detector based on this network is evaluated against the non-temporal counterpart as baseline and achieves competitive results in an evaluation on the KITTI object detection dataset. | ['J. Marius Zöllner', 'Tassilo Wald', 'Michael Weber'] | 2021-03-22 | null | null | null | null | ['temporal-information-extraction'] | ['natural-language-processing'] | [ 1.97496489e-01 -2.84337223e-01 1.88353341e-02 -4.66633946e-01
-6.14959635e-02 -3.87877941e-01 9.88608301e-01 2.93452233e-01
-9.87867415e-01 3.84757161e-01 -2.44284391e-01 -1.72451720e-01
-4.11777943e-01 -5.49859464e-01 -4.65965211e-01 -7.44925320e-01
-3.15334529e-01 1.33233830e-01 1.02047360e+00 -4.04511303e-01
1.61807388e-01 8.31421316e-01 -1.72224998e+00 6.09124422e-01
1.69721693e-01 1.30190730e+00 4.45632786e-01 8.75594974e-01
6.11263327e-02 7.22498298e-01 -4.42532629e-01 1.20864980e-01
4.32743192e-01 -1.94620192e-01 -5.75439572e-01 2.84907948e-02
1.04363509e-01 -4.59245384e-01 -3.87265742e-01 5.28546512e-01
2.25438818e-01 1.34289742e-01 4.62618113e-01 -1.02674997e+00
1.23103283e-01 2.78335631e-01 -8.42704326e-02 6.63783848e-01
2.10644960e-01 3.07657719e-01 7.23988831e-01 -7.26451993e-01
8.31376553e-01 9.43193853e-01 5.21046042e-01 2.47072890e-01
-1.25299716e+00 -7.15352222e-02 3.74262631e-01 7.53588676e-01
-1.19026291e+00 -3.52963865e-01 8.16645265e-01 -5.71662843e-01
1.38365054e+00 -1.05183132e-01 9.44305599e-01 1.29721427e+00
1.37417853e-01 6.33325636e-01 1.25493526e+00 -4.66212422e-01
1.86051235e-01 2.75908828e-01 1.75909996e-01 2.11762741e-01
-1.06982276e-01 8.61741006e-01 -6.38366044e-01 5.97696304e-01
4.87735152e-01 -1.72878489e-01 6.16299622e-02 -5.27507782e-01
-1.24605024e+00 5.37060380e-01 8.64784002e-01 9.98183727e-01
-3.81787002e-01 2.25597650e-01 5.91608644e-01 5.82799554e-01
2.77397484e-01 7.33110160e-02 -4.77446824e-01 -1.12655461e-01
-9.76108015e-01 1.68826401e-01 5.30100226e-01 5.50790489e-01
5.43737113e-01 4.32640910e-02 -2.00103238e-01 2.62647420e-01
2.04580858e-01 1.21568859e-01 3.74587595e-01 -5.60539126e-01
1.91773996e-02 5.76267779e-01 3.34761947e-01 -7.57999420e-01
-8.30204785e-01 -5.26979983e-01 -4.98232812e-01 6.30343914e-01
6.98585629e-01 3.04815024e-01 -1.00311148e+00 1.47234964e+00
3.02933931e-01 -2.00135354e-02 -7.46509293e-03 1.01978970e+00
6.36240780e-01 3.72534752e-01 5.90170398e-02 -1.93624273e-01
1.18770242e+00 -5.90092123e-01 -5.69568396e-01 -7.29625002e-02
5.37888467e-01 -7.09496617e-01 3.31989974e-01 6.19891942e-01
-4.47322041e-01 -8.45777571e-01 -1.40748203e+00 1.74892936e-02
-9.00464654e-01 3.57717484e-01 3.86901557e-01 5.79608738e-01
-1.11987388e+00 8.91614556e-01 -1.08795381e+00 -8.32871258e-01
-1.92852411e-02 5.41032851e-01 -5.86703837e-01 1.22799262e-01
-9.22527730e-01 1.31200802e+00 9.26358819e-01 5.68807423e-01
-9.08644378e-01 -2.48842109e-02 -6.89095438e-01 -1.91275939e-01
3.93153965e-01 -3.59912336e-01 1.32310259e+00 -1.02120137e+00
-1.41817605e+00 7.30946362e-01 -1.87962279e-01 -1.12546551e+00
9.20996428e-01 1.03239059e-01 -2.49555841e-01 3.48256677e-01
-1.83371961e-01 8.50965381e-01 9.37663853e-01 -1.08672440e+00
-6.53389990e-01 -2.07938179e-01 1.73519894e-01 -2.43827090e-01
-2.55869776e-01 1.39546946e-01 -2.06926703e-01 -3.43146116e-01
1.79544464e-01 -9.99307692e-01 -2.48150676e-01 4.17142548e-02
-4.96725291e-02 -1.77205488e-01 1.17112482e+00 -5.11929333e-01
7.96738744e-01 -2.05547738e+00 -4.86054011e-02 -3.60053554e-02
2.88611948e-02 3.57235223e-01 -6.62486926e-02 3.98791552e-01
-2.59019941e-01 -2.13759288e-01 -1.14671148e-01 -3.73679787e-01
-2.43762042e-02 7.66100213e-02 -1.09377868e-01 5.70911169e-01
3.38886619e-01 7.76425958e-01 -7.27606714e-01 -4.25214738e-01
8.10109794e-01 4.91763800e-01 -2.05180511e-01 4.73735183e-02
-2.72687346e-01 6.77009523e-01 -3.85749459e-01 4.01984841e-01
4.80104059e-01 2.09172100e-01 7.36529604e-02 -3.31556827e-01
-7.16469407e-01 3.12441587e-01 -7.82141626e-01 1.83699417e+00
-4.95078802e-01 1.08533466e+00 -2.73951292e-01 -1.12278175e+00
9.35355425e-01 6.18617594e-01 5.53152144e-01 -1.02547121e+00
3.74746442e-01 4.00510520e-01 4.20320302e-01 -5.78764796e-01
5.74488401e-01 -3.91993336e-02 3.15757543e-01 2.73615509e-01
3.62869054e-01 2.69141704e-01 5.42067885e-01 -1.24564081e-01
1.26477671e+00 4.95890290e-01 2.87126034e-01 -1.61038443e-01
6.03328466e-01 9.06516984e-02 3.42917442e-01 6.86061025e-01
-4.88963962e-01 5.51384985e-01 2.65210897e-01 -8.76107156e-01
-1.18210757e+00 -7.73074269e-01 -2.87151217e-01 7.77342379e-01
-2.59277374e-02 -4.10880893e-01 -2.43831664e-01 -7.26322234e-01
-2.75777400e-01 4.53166783e-01 -9.38354909e-01 -4.67129983e-02
-6.28369987e-01 -4.26462948e-01 8.25251043e-02 5.92625797e-01
5.32322824e-01 -1.34096515e+00 -1.28200495e+00 5.46150327e-01
1.56517059e-01 -1.34298027e+00 3.58288169e-01 7.01187789e-01
-9.21864033e-01 -9.69202518e-01 -4.41745192e-01 -2.82242447e-01
1.92243129e-01 3.18481058e-01 6.72403574e-01 -9.17305276e-02
-4.56993043e-01 4.69729096e-01 -6.95433199e-01 -7.26205468e-01
-2.59590566e-01 1.10966213e-01 -1.08496711e-01 1.09192140e-01
4.74478900e-01 -8.12563300e-01 -7.09537685e-01 3.28050643e-01
-8.54246616e-01 -2.04898566e-01 6.22518241e-01 6.04460895e-01
4.98138107e-02 2.82322168e-02 3.09096158e-01 -1.13174610e-01
-1.50209442e-01 -4.61159274e-02 -7.97548056e-01 2.90530995e-02
-3.49126607e-01 5.51264323e-02 3.04793060e-01 -4.53425318e-01
-1.24327183e+00 4.17827874e-01 -2.28765652e-01 -5.52658856e-01
-4.93611902e-01 3.65239918e-01 1.77760944e-01 -2.71439075e-01
8.42045069e-01 1.91306144e-01 -2.37778589e-01 -5.59541523e-01
1.65991321e-01 1.96221754e-01 2.67126650e-01 -3.07071954e-01
7.39742815e-01 8.38235855e-01 8.40194821e-02 -1.04028916e+00
-5.46088994e-01 -8.81315172e-01 -1.21454465e+00 -5.50307691e-01
9.27571476e-01 -6.13720000e-01 -7.45385826e-01 4.26289260e-01
-1.52716088e+00 -2.70830840e-01 -4.13067222e-01 7.37621367e-01
-5.84992409e-01 -4.70158570e-02 -2.30461240e-01 -1.10484815e+00
1.66059524e-01 -9.79499876e-01 1.01734054e+00 -2.42854785e-02
1.76374745e-02 -8.01683843e-01 1.70583334e-02 -3.19843292e-02
5.51850259e-01 2.82283902e-01 3.42583388e-01 -6.83369517e-01
-9.36981320e-01 -5.82123697e-01 -2.34838471e-01 3.30138862e-01
-2.21476287e-01 1.38114110e-01 -1.32523131e+00 -1.75378308e-01
1.56920105e-01 -1.40899047e-01 1.37023342e+00 3.77468407e-01
5.72721958e-01 2.49691501e-01 -4.05301332e-01 2.84264505e-01
1.56325698e+00 4.18510288e-01 5.47251999e-01 6.73289597e-01
4.00804460e-01 8.34434450e-01 6.78625882e-01 5.02172589e-01
3.67922559e-02 1.02479041e+00 7.24001765e-01 1.92414939e-01
-2.89692748e-02 2.79336631e-01 3.46155405e-01 2.11173624e-01
-3.51859331e-01 -2.03206718e-01 -8.80388379e-01 7.84097433e-01
-1.99210703e+00 -9.12449718e-01 -1.19742967e-01 2.15334392e+00
9.95681137e-02 6.26431584e-01 2.31735080e-01 4.76779908e-01
3.06699365e-01 1.81229398e-01 -2.20459133e-01 -3.23100716e-01
-2.21755058e-01 1.16009146e-01 3.88575047e-01 6.80551156e-02
-1.27204609e+00 7.57085681e-01 5.78608799e+00 4.94336516e-01
-1.31663501e+00 2.41019085e-01 9.40575898e-02 -1.50491044e-01
5.33943117e-01 1.94773227e-01 -7.51464188e-01 4.31113467e-02
1.07711661e+00 1.33672524e-02 -1.09416090e-01 7.70005465e-01
6.02511346e-01 -5.10632575e-01 -1.28210199e+00 8.43953609e-01
-3.20614755e-01 -1.05389380e+00 -4.18622404e-01 8.92640576e-02
1.75227299e-01 1.94009557e-01 -3.34365070e-02 2.09940165e-01
-2.04118520e-01 -8.24975669e-01 1.12066782e+00 5.38391590e-01
2.36073852e-01 -4.48958278e-01 8.98519695e-01 4.26358014e-01
-1.34136462e+00 -2.91289389e-01 -1.49386957e-01 -2.82851040e-01
2.88148910e-01 5.02273738e-01 -1.07441545e+00 6.44205153e-01
7.75646448e-01 8.29812586e-01 -7.99069941e-01 1.43575585e+00
-1.06675655e-01 3.98129940e-01 -5.97319782e-01 -4.01521623e-02
4.17106479e-01 8.49667490e-02 8.64597797e-01 1.44306064e+00
3.91096175e-01 -7.69135654e-02 5.14727011e-02 6.87603533e-01
5.41637659e-01 -4.09235694e-02 -1.01732898e+00 1.33254258e-02
-1.58617646e-01 1.44425356e+00 -1.23319161e+00 -1.45352021e-01
-3.91958594e-01 7.99064636e-01 1.24176368e-01 2.72635281e-01
-5.78709185e-01 -1.22508325e-01 3.09928566e-01 6.18588440e-02
7.83269584e-01 -5.93405902e-01 3.79241109e-02 -8.59886706e-01
1.48495689e-01 -3.07223022e-01 2.71527052e-01 -8.02148521e-01
-7.87286043e-01 8.64092588e-01 2.98176676e-01 -1.37393332e+00
-5.08698106e-01 -1.09973907e+00 -5.32951474e-01 6.72434986e-01
-1.31043029e+00 -1.32774091e+00 -1.85233399e-01 3.75745237e-01
4.15131092e-01 1.59046471e-01 4.62025017e-01 2.12714851e-01
-2.88901806e-01 5.01985615e-03 -2.94022053e-01 -1.04376702e-02
3.54118943e-01 -1.23125601e+00 1.00522056e-01 1.22708392e+00
2.81701267e-01 4.14710701e-01 1.04453194e+00 -3.15890312e-01
-9.94699061e-01 -8.40422630e-01 9.52998459e-01 -4.75174546e-01
7.33268499e-01 -4.00262177e-01 -9.39470828e-01 5.39492011e-01
4.34864998e-01 1.37652040e-01 3.56928334e-02 -1.87529121e-02
-3.02895814e-01 -2.42830113e-01 -8.05577397e-01 1.93512544e-01
8.86590183e-01 -5.25384009e-01 -5.74533165e-01 1.23280920e-01
6.77281260e-01 -1.22162677e-01 -5.52891552e-01 7.01109111e-01
5.92672825e-01 -1.39369357e+00 9.85623717e-01 -2.48641968e-01
-9.49105807e-03 -5.64329863e-01 -9.65321362e-02 -7.21294701e-01
-1.43059000e-01 -3.82813126e-01 1.31878912e-01 7.73471832e-01
1.62692487e-01 -3.91013384e-01 7.58615494e-01 4.33996208e-02
-4.82100770e-02 -1.23279594e-01 -1.23978388e+00 -1.21836042e+00
-4.20361936e-01 -1.04356980e+00 6.65552393e-02 4.35408175e-01
-3.92922699e-01 2.72895604e-01 -2.33900353e-01 -1.19903810e-01
4.30929631e-01 1.11472771e-01 4.83023196e-01 -1.19457650e+00
-2.40799755e-01 -5.74291289e-01 -9.14832652e-01 -5.52626193e-01
-1.00667235e-02 -5.82068443e-01 2.20483765e-01 -1.21040058e+00
-1.09484486e-01 -2.34160293e-02 -5.05381942e-01 4.24945563e-01
5.40568769e-01 4.11784083e-01 2.19502732e-01 -1.92359284e-01
-5.67146540e-01 5.52727461e-01 8.22117686e-01 -1.48585379e-01
-1.90662369e-01 1.34381503e-01 3.95284265e-01 6.16850615e-01
7.50092745e-01 -5.05151927e-01 -2.91151881e-01 6.19558617e-03
2.51705885e-01 3.25767673e-03 8.12158406e-01 -1.43154407e+00
4.41353381e-01 1.14187323e-01 3.13117713e-01 -9.32671607e-01
6.45280540e-01 -1.02954066e+00 1.45658717e-01 6.40270591e-01
-1.40095770e-01 -2.71438845e-02 4.86581117e-01 6.50411069e-01
-4.15863335e-01 -3.12086701e-01 6.80245876e-01 -2.71983892e-01
-1.22825015e+00 9.41505581e-02 -6.23554766e-01 -8.45262468e-01
9.15380180e-01 -5.23109555e-01 -1.62973553e-02 -5.12010604e-02
-1.25200284e+00 -2.08429620e-01 1.89343080e-01 4.07718331e-01
6.04256272e-01 -1.14430380e+00 -3.82475823e-01 -1.72639623e-01
4.26262230e-01 -4.77004856e-01 3.11298162e-01 1.29868507e+00
-3.77747566e-01 9.44101632e-01 -4.98808444e-01 -9.67831552e-01
-1.41739869e+00 1.08608508e+00 4.54031974e-01 -2.80818999e-01
-7.08186269e-01 5.46448767e-01 3.01408559e-01 6.71861991e-02
1.39999464e-01 -5.78501821e-01 -5.27528703e-01 4.55543071e-01
3.74047637e-01 1.23125255e-01 5.66331685e-01 -7.17581809e-01
-4.43697780e-01 3.55507851e-01 2.39237044e-02 -7.04990685e-01
1.35269451e+00 -1.01132225e-02 1.53050516e-02 7.82831907e-01
1.24064672e+00 -5.84376514e-01 -1.31463015e+00 -3.27737898e-01
4.35466409e-01 -2.88682491e-01 1.22057036e-01 -6.55482531e-01
-7.08463967e-01 1.20542538e+00 9.48422074e-01 3.50876153e-01
1.18774700e+00 -2.71343827e-01 1.56122833e-01 5.73147357e-01
6.03956938e-01 -8.69374454e-01 3.20557743e-01 5.16814232e-01
1.09184122e+00 -1.43749261e+00 -4.87604514e-02 -1.54917404e-01
-1.71759129e-01 1.49896884e+00 4.18756753e-01 -2.63739787e-02
8.28473985e-01 9.30247307e-02 -1.08534247e-01 -1.95445210e-01
-1.03459132e+00 -1.02761292e+00 4.42272365e-01 6.18617594e-01
2.34503478e-01 -1.23641323e-02 -3.49697232e-01 1.82960480e-01
1.20552234e-01 6.59474917e-03 2.64626980e-01 1.17741883e+00
-4.42895204e-01 -1.31720150e+00 -1.87040538e-01 7.04434738e-02
-3.06671739e-01 1.06629223e-01 -4.81150776e-01 1.09248078e+00
4.46549147e-01 9.19842184e-01 -1.50856137e-01 -5.72143137e-01
2.76258737e-01 9.98421665e-03 8.00113976e-01 -4.57662970e-01
-7.34427214e-01 2.58627117e-01 1.78948566e-01 -7.39328623e-01
-8.00306499e-01 -1.00077033e+00 -7.92770505e-01 1.35428444e-01
-2.15717822e-01 -2.65332967e-01 1.15575039e+00 1.23996985e+00
-1.22309908e-01 7.53565967e-01 2.32994005e-01 -1.31002927e+00
8.74444190e-03 -1.05278671e+00 -3.02831203e-01 2.39204448e-02
5.45391321e-01 -8.61353874e-01 -2.20196843e-01 8.31998661e-02] | [8.446159362792969, -0.39331290125846863] |
1ca48970-b24a-47c4-b4bc-a38221699f82 | deep-implicit-templates-for-3d-shape | 2011.14565 | null | https://arxiv.org/abs/2011.14565v2 | https://arxiv.org/pdf/2011.14565v2.pdf | Deep Implicit Templates for 3D Shape Representation | Deep implicit functions (DIFs), as a kind of 3D shape representation, are becoming more and more popular in the 3D vision community due to their compactness and strong representation power. However, unlike polygon mesh-based templates, it remains a challenge to reason dense correspondences or other semantic relationships across shapes represented by DIFs, which limits its applications in texture transfer, shape analysis and so on. To overcome this limitation and also make DIFs more interpretable, we propose Deep Implicit Templates, a new 3D shape representation that supports explicit correspondence reasoning in deep implicit representations. Our key idea is to formulate DIFs as conditional deformations of a template implicit function. To this end, we propose Spatial Warping LSTM, which decomposes the conditional spatial transformation into multiple affine transformations and guarantees generalization capability. Moreover, the training loss is carefully designed in order to achieve high reconstruction accuracy while learning a plausible template with accurate correspondences in an unsupervised manner. Experiments show that our method can not only learn a common implicit template for a collection of shapes, but also establish dense correspondences across all the shapes simultaneously without any supervision. | ['Yebin Liu', 'Qionghai Dai', 'Tao Yu', 'Zerong Zheng'] | 2020-11-30 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Zheng_Deep_Implicit_Templates_for_3D_Shape_Representation_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Zheng_Deep_Implicit_Templates_for_3D_Shape_Representation_CVPR_2021_paper.pdf | cvpr-2021-1 | ['3d-shape-representation'] | ['computer-vision'] | [ 8.21439326e-02 6.18921295e-02 5.11342846e-03 -5.62726378e-01
-4.03732747e-01 -5.12787223e-01 7.51365483e-01 -1.01244509e-01
2.20656261e-01 3.27621430e-01 2.67167866e-01 -1.40227586e-01
-3.01911503e-01 -1.13644278e+00 -7.36664116e-01 -7.34662414e-01
5.70680857e-01 6.30469561e-01 1.61207736e-01 4.72009443e-02
1.06964640e-01 1.09726870e+00 -1.11703205e+00 2.42932826e-01
8.31520796e-01 1.01696038e+00 2.31427491e-01 -3.00498098e-01
-3.07912618e-01 3.74621689e-01 -1.31827921e-01 -3.69802624e-01
2.59956360e-01 1.21302523e-01 -7.11626172e-01 3.68056625e-01
5.29517055e-01 -3.31936747e-01 -5.44639051e-01 9.68848526e-01
3.58122557e-01 8.29176530e-02 9.44685578e-01 -1.07321620e+00
-1.11037028e+00 8.49991012e-03 -5.55779636e-01 -4.68665630e-01
8.50843266e-02 1.24648362e-01 8.92447352e-01 -1.19514513e+00
5.13071299e-01 1.58908248e+00 6.60472095e-01 4.34279233e-01
-1.45984888e+00 -5.77452660e-01 3.27289313e-01 -1.13361530e-01
-1.50652909e+00 -4.28268522e-01 9.51732934e-01 -3.95102918e-01
7.48432875e-01 2.89851695e-01 5.92300534e-01 9.89689946e-01
-5.43087758e-02 8.84491622e-01 9.54115629e-01 -2.20598951e-01
9.72956568e-02 -1.34010389e-01 -2.36516178e-01 6.93357527e-01
2.06905022e-01 -2.03200296e-01 -3.18385363e-01 -1.16738543e-01
1.66456914e+00 3.93255353e-01 -3.86686385e-01 -6.67010784e-01
-1.39003944e+00 7.18527913e-01 7.84652770e-01 2.70493865e-01
-4.72464383e-01 3.09679925e-01 9.66134891e-02 8.30479935e-02
6.17703259e-01 1.16715826e-01 -3.37637454e-01 2.09541574e-01
-4.20556307e-01 1.87335983e-01 3.98482591e-01 1.06647503e+00
8.36416602e-01 1.73930094e-01 -1.25402912e-01 8.83659005e-01
5.68526626e-01 7.35775948e-01 1.12440223e-02 -8.69861126e-01
4.17336911e-01 8.70004833e-01 7.13630021e-02 -1.36846650e+00
2.26082914e-02 -4.69405919e-01 -1.29873621e+00 2.53958881e-01
2.29185984e-01 5.98000288e-01 -9.56828594e-01 1.81415105e+00
3.13030839e-01 3.28224629e-01 -3.31921607e-01 1.00133073e+00
7.19053507e-01 5.90872407e-01 -1.54398456e-01 1.56737730e-01
1.13366604e+00 -4.44883466e-01 -4.40900087e-01 1.04784586e-01
2.61358649e-01 -6.97770059e-01 1.10367930e+00 -6.89267069e-02
-1.04414046e+00 -4.14068133e-01 -7.52007127e-01 -5.01090050e-01
-1.67185187e-01 1.05582386e-01 6.24463677e-01 6.06233701e-02
-8.66666377e-01 4.20246303e-01 -8.92022312e-01 -1.81015059e-01
6.49722338e-01 5.39441228e-01 -4.48232174e-01 -5.06028235e-02
-8.59266460e-01 8.29790950e-01 5.34459539e-02 3.11294556e-01
-4.31169719e-01 -5.66727281e-01 -1.11906660e+00 7.87019879e-02
1.19809806e-01 -9.30929422e-01 8.76524210e-01 -7.85191119e-01
-1.56187832e+00 1.07870376e+00 -2.69974142e-01 -5.74798286e-02
4.28588599e-01 -9.32681337e-02 6.09806702e-02 -2.98102349e-01
1.63591608e-01 6.26054585e-01 1.00913191e+00 -1.36003923e+00
1.77325457e-02 -4.89613384e-01 2.29653075e-01 2.03850001e-01
-1.42202467e-01 -4.41918671e-01 -4.63105887e-01 -1.00101650e+00
6.55923903e-01 -6.37455761e-01 -1.91448092e-01 6.34773433e-01
-4.01645094e-01 -4.46503431e-01 9.20974135e-01 -5.05296707e-01
5.69879532e-01 -2.12673187e+00 4.67689961e-01 4.26718563e-01
3.39528888e-01 1.64828718e-01 -2.87598282e-01 1.88965313e-02
1.35681331e-01 5.86770400e-02 -4.94658202e-01 -3.81367356e-01
2.24362090e-01 6.05293453e-01 -7.32042253e-01 4.21334296e-01
5.42820632e-01 1.15534592e+00 -7.65291929e-01 -8.50478709e-02
2.81759083e-01 7.71385849e-01 -4.43381190e-01 2.37028182e-01
-4.09026086e-01 5.45305073e-01 -9.08851266e-01 5.12548447e-01
9.58424926e-01 -2.82985896e-01 5.71903540e-03 -5.39124191e-01
-2.35021804e-02 4.12708700e-01 -1.06530333e+00 1.92383659e+00
-5.22395611e-01 1.36707038e-01 1.47019168e-02 -1.25291550e+00
1.47442496e+00 3.54005575e-01 4.01572436e-01 -6.46082997e-01
1.88598465e-02 5.29863417e-01 -3.34904343e-01 -1.50080726e-01
4.50270325e-02 -2.58073509e-01 1.15038477e-01 5.06480277e-01
-1.94997817e-01 -4.44061488e-01 -5.62779784e-01 -2.05681399e-01
5.75744748e-01 4.02311146e-01 1.00592047e-01 -3.96738946e-01
6.79765522e-01 -3.82756233e-01 8.20086479e-01 1.55061349e-01
4.65111375e-01 6.89472616e-01 1.49707168e-01 -8.13157022e-01
-1.07675421e+00 -1.32540798e+00 -2.29003891e-01 2.26536915e-01
4.21551704e-01 2.84469631e-02 -4.27401632e-01 -5.41427553e-01
3.20014715e-01 5.33184290e-01 -6.82237685e-01 -9.08701941e-02
-7.85491645e-01 -2.49699473e-01 2.85595447e-01 7.53277063e-01
6.70567811e-01 -9.71552253e-01 -2.15920553e-01 1.29460916e-01
-1.85822826e-02 -9.18342888e-01 -8.09224546e-01 -2.08659112e-01
-1.09355927e+00 -9.63948965e-01 -9.48852539e-01 -9.96839285e-01
1.12424421e+00 5.50744772e-01 9.04956520e-01 1.51640043e-01
-8.35080340e-04 3.99875581e-01 -1.00567073e-01 -1.81696832e-01
-1.78477317e-01 -6.14090823e-02 7.14743137e-02 3.16178948e-01
3.14678490e-01 -9.64034021e-01 -3.84409428e-01 5.04508793e-01
-8.63412857e-01 4.83264536e-01 5.96868813e-01 9.52326357e-01
1.09936059e+00 -2.38060713e-01 5.64864874e-01 -6.11866355e-01
3.44121784e-01 -2.79317111e-01 -5.93567014e-01 3.40644449e-01
-1.14133306e-01 2.41681322e-01 4.61500078e-01 -4.20926362e-01
-1.01633036e+00 2.87980773e-02 -1.62170202e-01 -9.76705194e-01
-1.36500284e-01 4.30700153e-01 -4.97892916e-01 -5.65480813e-02
3.74054372e-01 5.23923874e-01 1.93774477e-01 -8.49960864e-01
4.06407982e-01 2.26320073e-01 4.64449555e-01 -1.00932097e+00
1.20431912e+00 5.68004131e-01 2.36337259e-01 -6.93542004e-01
-8.02679658e-01 -4.03783238e-03 -1.03206563e+00 1.77542523e-01
7.29639292e-01 -7.44876742e-01 -6.89002812e-01 3.92775953e-01
-1.57461417e+00 -2.97169715e-01 -3.48209232e-01 2.99486339e-01
-6.78384840e-01 3.47592235e-01 -4.39884216e-01 -5.92736363e-01
-3.63238871e-01 -1.16901529e+00 1.31675577e+00 3.80026586e-02
-1.39569804e-01 -1.04619086e+00 -2.31869400e-01 -5.28417877e-04
4.16439384e-01 5.34099400e-01 1.14551580e+00 -5.62903956e-02
-1.10498881e+00 3.12655568e-02 -6.05037689e-01 1.45668581e-01
6.59950435e-01 -2.61587203e-01 -7.54889846e-01 -2.77379960e-01
5.51585034e-02 -1.61781043e-01 6.20986879e-01 2.07640126e-01
1.49587035e+00 -4.90942717e-01 -1.86076388e-01 7.75486171e-01
1.17121649e+00 6.94661587e-02 5.20850599e-01 1.81887433e-01
8.69120777e-01 5.24192214e-01 2.85334945e-01 2.20020413e-01
4.43494856e-01 8.21217477e-01 5.33078611e-01 -2.82084167e-01
-1.83103830e-01 -4.73250449e-01 6.61070272e-02 1.08696902e+00
-3.39185029e-01 2.65786380e-01 -7.12659240e-01 3.23839724e-01
-1.90968573e+00 -6.87810302e-01 5.19770477e-03 2.37536478e+00
8.67456079e-01 -9.08408239e-02 -2.90997058e-01 4.16188166e-02
7.04804242e-01 8.83358046e-02 -7.57739484e-01 -1.96829632e-01
-1.89836100e-01 3.70358735e-01 -9.06881243e-02 3.90431076e-01
-8.70252013e-01 8.62018883e-01 4.71122932e+00 8.79604042e-01
-1.39348888e+00 -7.24199861e-02 6.56566858e-01 4.00257945e-01
-9.52028513e-01 -6.74895123e-02 -3.02556813e-01 1.20802775e-01
-1.33368045e-01 -1.74745083e-01 4.98568922e-01 6.23364091e-01
1.38303608e-01 6.10022604e-01 -1.20975864e+00 1.26266491e+00
-2.80822158e-01 -1.58546209e+00 6.20281279e-01 2.52725720e-01
7.51104116e-01 -2.59205461e-01 1.64796293e-01 1.58562139e-02
2.67924629e-02 -1.25833762e+00 1.04803097e+00 6.75205529e-01
1.06342077e+00 -4.56618935e-01 3.24528575e-01 4.57038939e-01
-1.43363130e+00 5.44868112e-01 -8.28045547e-01 5.31059913e-02
-8.02541524e-03 6.05912089e-01 -4.11114603e-01 6.23574436e-01
3.64274085e-01 1.11522806e+00 -4.73609865e-02 8.82650077e-01
-2.95671403e-01 9.43556279e-02 -3.88820857e-01 1.67769730e-01
2.13078618e-01 -4.73735392e-01 4.92705315e-01 7.92966545e-01
5.19114256e-01 4.15310085e-01 2.29604080e-01 1.52018535e+00
-3.37878048e-01 3.41993570e-02 -8.80733311e-01 7.09611624e-02
6.15504503e-01 1.03745878e+00 -4.80677366e-01 -3.76273482e-03
-3.73681277e-01 8.75489593e-01 4.74203348e-01 4.68684763e-01
-5.74262142e-01 -5.85690066e-02 8.46403718e-01 2.26081058e-01
1.54892772e-01 -6.19084358e-01 -6.00266278e-01 -1.32038963e+00
3.05348277e-01 -4.99377668e-01 -1.03807755e-01 -7.68144608e-01
-1.64739835e+00 4.08412218e-01 -2.10878998e-01 -1.53981233e+00
2.48608395e-01 -6.16199195e-01 -7.99997449e-01 1.17264104e+00
-1.50567627e+00 -1.61265385e+00 -4.66799885e-01 6.68097913e-01
4.95183825e-01 -4.03788872e-02 8.47368121e-01 2.80021727e-01
-1.29156500e-01 5.20497084e-01 -1.57993019e-01 2.88900346e-01
3.91413331e-01 -9.84227121e-01 7.05051899e-01 2.56821126e-01
2.51667917e-01 8.84749353e-01 1.08540334e-01 -4.85246181e-01
-1.72661066e+00 -1.26038146e+00 5.56290329e-01 -2.73399621e-01
2.93889850e-01 -3.51676494e-01 -1.28163624e+00 7.44632185e-01
-1.69983804e-01 1.73919022e-01 2.01391548e-01 -5.67319877e-02
-6.54185593e-01 -1.18269302e-01 -1.07088923e+00 6.28725290e-01
1.35933995e+00 -7.30036318e-01 -7.25702822e-01 2.32564613e-01
7.28193641e-01 -6.41108632e-01 -1.01806664e+00 7.13559389e-01
4.09127533e-01 -6.92183018e-01 1.16436827e+00 -4.97178018e-01
2.90535688e-01 -4.08666909e-01 -3.12622964e-01 -1.13800526e+00
-5.47969460e-01 -4.06878799e-01 -1.13095254e-01 1.13604069e+00
1.68454736e-01 -9.87206042e-01 7.50249565e-01 6.11252785e-01
-3.16263646e-01 -1.13371563e+00 -1.06379580e+00 -8.55220556e-01
2.20136195e-01 -2.31557220e-01 9.61288095e-01 1.20045781e+00
-5.73369145e-01 1.90337092e-01 -2.22134575e-01 1.93524018e-01
7.14868665e-01 5.84462047e-01 8.58320177e-01 -1.42624116e+00
-8.37207735e-02 -6.30219817e-01 -5.27745724e-01 -1.64071631e+00
3.65274340e-01 -1.13624084e+00 -7.30743483e-02 -1.66346204e+00
3.04288529e-02 -1.22153330e+00 -1.49992362e-01 8.00863445e-01
2.18384027e-01 2.20115095e-01 -3.63183208e-02 4.44513291e-01
-3.30883488e-02 1.16885412e+00 1.76220274e+00 -3.93522918e-01
-4.12479155e-02 -6.07214458e-02 -5.54537177e-01 8.17553043e-01
5.65432489e-01 -9.35296342e-02 -4.99671966e-01 -1.07660484e+00
-2.57139504e-02 3.98759842e-02 5.55415750e-01 -4.88937527e-01
1.95638202e-02 -5.16073406e-01 4.83284980e-01 -6.18620992e-01
6.11636043e-01 -8.74112606e-01 3.58057618e-01 2.49957487e-01
-7.12098926e-02 -1.50471240e-01 1.64564431e-01 5.82059920e-01
-2.97858059e-01 1.18724227e-01 6.28443003e-01 -1.28216177e-01
-3.44796777e-01 1.05298531e+00 1.43734857e-01 -9.26618874e-02
6.60446048e-01 -3.79228115e-01 -1.00561529e-01 -1.72362581e-01
-6.33581996e-01 -1.30724460e-02 7.17206180e-01 5.80253720e-01
9.94627237e-01 -1.91163516e+00 -6.94464147e-01 4.42116767e-01
4.91364710e-02 7.88746834e-01 6.07220568e-02 6.75742090e-01
-4.46054488e-01 4.01194274e-01 -2.49228984e-01 -8.37869406e-01
-7.60220706e-01 2.97608882e-01 4.54974145e-01 1.39747635e-01
-1.11914837e+00 6.93803549e-01 9.02606487e-01 -7.36798286e-01
1.34956658e-01 -4.96118635e-01 1.20828584e-01 -5.19584715e-01
2.40486860e-01 -1.17090620e-01 -6.98284730e-02 -7.70397365e-01
-2.98216254e-01 1.08748007e+00 2.22159877e-01 2.31334448e-01
1.50893927e+00 1.27380162e-01 -3.50060105e-01 2.98382759e-01
1.14327526e+00 -5.39597720e-02 -1.37404966e+00 -7.79266715e-01
-8.25653747e-02 -8.42523873e-01 -9.99058634e-02 -3.68148088e-01
-1.20674908e+00 1.16830051e+00 1.05595902e-01 6.14193790e-02
9.58172083e-01 2.22558770e-02 9.80136156e-01 4.12034661e-01
6.08707905e-01 -4.48452383e-01 1.72239661e-01 7.12512493e-01
1.51405072e+00 -8.53677571e-01 -1.28630355e-01 -6.96281552e-01
-2.77761787e-01 1.43959439e+00 6.02962673e-01 -2.81974196e-01
7.36698389e-01 -6.93705827e-02 -2.04470471e-01 -2.89549530e-01
-4.53081101e-01 5.70568368e-02 6.74867094e-01 6.31589532e-01
3.64346176e-01 1.67309821e-01 1.11194633e-01 4.81665850e-01
1.07048750e-01 -1.31725192e-01 -9.89948139e-02 5.43183804e-01
-2.69607395e-01 -1.31232548e+00 -2.58790135e-01 3.70248437e-01
2.72487760e-01 1.53837770e-01 -2.49974102e-01 5.18761575e-01
7.51382485e-02 2.28963867e-01 1.16560690e-01 -2.51370549e-01
5.42470157e-01 -2.38225207e-01 7.65699446e-01 -7.04093397e-01
-1.54329082e-02 4.54732813e-02 -2.74835944e-01 -3.96984696e-01
-4.33459789e-01 -5.61490238e-01 -1.36028433e+00 -3.94278228e-01
-1.41471148e-01 -1.66100636e-01 3.79166424e-01 1.08877552e+00
4.90998745e-01 1.92846805e-01 6.81758642e-01 -1.09289289e+00
-5.03027558e-01 -5.17727554e-01 -4.52508003e-01 6.05326056e-01
1.20439723e-01 -1.16672456e+00 4.70708534e-02 -2.51179505e-02] | [8.584914207458496, -3.6599552631378174] |
375b858d-6f62-4d85-b549-c813e1fc1f83 | coarse-to-fine-q-attention-with-tree | 2204.12471 | null | https://arxiv.org/abs/2204.12471v2 | https://arxiv.org/pdf/2204.12471v2.pdf | Coarse-to-fine Q-attention with Tree Expansion | Coarse-to-fine Q-attention enables sample-efficient robot manipulation by discretizing the translation space in a coarse-to-fine manner, where the resolution gradually increases at each layer in the hierarchy. Although effective, Q-attention suffers from "coarse ambiguity" - when voxelization is significantly coarse, it is not feasible to distinguish similar-looking objects without first inspecting at a finer resolution. To combat this, we propose to envision Q-attention as a tree that can be expanded and used to accumulate value estimates across the top-k voxels at each Q-attention depth. When our extension, Q-attention with Tree Expansion (QTE), replaces standard Q-attention in the Attention-driven Robot Manipulation (ARM) system, we are able to accomplish a larger set of tasks; especially on those that suffer from "coarse ambiguity". In addition to evaluating our approach across 12 RLBench tasks, we also show that the improved performance is visible in a real-world task involving small objects. | ['Pieter Abbeel', 'Stephen James'] | 2022-04-26 | null | null | null | null | ['robot-manipulation'] | ['robots'] | [ 2.78989077e-01 3.81197006e-01 1.44106746e-01 1.43952891e-01
-1.02905619e+00 -4.69626695e-01 9.41321999e-02 2.05201566e-01
-2.05742404e-01 7.05880702e-01 1.44725934e-01 -1.30105674e-01
-4.42725748e-01 -7.30081141e-01 -7.46488094e-01 -4.54269677e-01
-9.36753973e-02 9.74647045e-01 3.83973300e-01 -4.17002052e-01
6.08117878e-01 7.69208074e-01 -1.68312132e+00 3.48500937e-01
7.82182813e-01 8.12815070e-01 6.04520738e-01 8.41043949e-01
-4.34301980e-02 6.51370168e-01 -6.19925499e-01 1.36642694e-01
4.07696813e-01 -1.82693481e-01 -1.48413301e+00 8.48327391e-03
6.07670248e-01 -3.84499192e-01 2.98759669e-01 1.08914959e+00
1.51115760e-01 3.70257109e-01 7.26826906e-01 -1.19071949e+00
-7.19528377e-01 7.00551212e-01 -8.23840499e-01 2.84446120e-01
3.52951527e-01 3.01280409e-01 1.22096598e+00 -1.04497898e+00
6.63977325e-01 1.78373539e+00 5.04288733e-01 2.01014072e-01
-1.47950017e+00 -3.21703076e-01 3.76067132e-01 2.03957319e-01
-1.10671484e+00 3.55797336e-02 4.25358862e-01 -4.82793629e-01
1.30017722e+00 1.25194535e-01 7.71110177e-01 6.05266511e-01
5.98493099e-01 7.38774896e-01 9.56183791e-01 -2.14252815e-01
3.96559775e-01 -5.86716533e-01 4.67812680e-02 6.70813859e-01
1.44179940e-01 3.68225537e-02 -3.19650084e-01 3.24913077e-02
1.39570653e+00 -1.25175789e-01 4.52377759e-02 -6.54790640e-01
-1.61482024e+00 7.26332486e-01 1.14278960e+00 3.78365755e-01
-8.96584153e-01 6.93435669e-01 3.99721026e-01 3.03490460e-01
1.21803463e-01 1.24330616e+00 -6.11638665e-01 -8.81250110e-03
-7.12943435e-01 5.54075658e-01 4.25465316e-01 1.41184163e+00
8.70465159e-01 -1.31182611e-01 -4.63657945e-01 3.77117395e-01
-7.15366974e-02 5.04951812e-02 3.25828135e-01 -1.62776339e+00
3.54679853e-01 4.06679064e-01 4.46980566e-01 -6.77764297e-01
-6.51654124e-01 -4.40747738e-01 -6.39279962e-01 8.49735498e-01
5.86646974e-01 1.69618160e-01 -1.20530534e+00 1.59006965e+00
3.94753516e-01 -2.74114728e-01 -2.84394801e-01 1.04337597e+00
1.92599952e-01 4.63132888e-01 2.68433958e-01 -7.08602322e-03
1.54317367e+00 -1.16897893e+00 -6.02575064e-01 -1.13908693e-01
5.10934889e-01 -3.09483290e-01 1.47049308e+00 7.81745017e-01
-1.43407035e+00 -8.49812210e-01 -1.07500052e+00 -6.26742661e-01
-4.55445528e-01 -4.47613955e-01 7.27453351e-01 1.53079972e-01
-1.25551677e+00 9.11648631e-01 -7.01103806e-01 -4.70902950e-01
4.94966328e-01 4.73542988e-01 -2.51325727e-01 2.46900320e-02
-8.31393123e-01 1.31380212e+00 4.15761024e-01 6.52284846e-02
-1.27995646e+00 -6.44473672e-01 -6.98030233e-01 3.20016772e-01
5.51765203e-01 -1.08498836e+00 1.59387684e+00 -7.30471492e-01
-1.41900122e+00 4.28703845e-01 -8.21451948e-04 -2.54497707e-01
4.46197212e-01 -5.30965626e-01 6.87003791e-01 9.83270928e-02
4.05668080e-01 1.17352927e+00 1.05585313e+00 -1.48155808e+00
-7.53084719e-01 -5.61521649e-01 5.38346291e-01 4.73249853e-01
2.40399227e-01 -2.82358170e-01 -1.66041583e-01 -6.32830143e-01
4.10309821e-01 -9.17772055e-01 -5.46622992e-01 1.54177651e-01
-1.82779968e-01 -4.06020969e-01 5.86254060e-01 -4.32890028e-01
6.05339885e-01 -1.83921921e+00 7.62866974e-01 -1.09929971e-01
5.17307043e-01 -3.42035532e-01 -3.63342822e-01 2.12422520e-01
-5.35895824e-02 9.12188292e-02 -3.73217195e-01 -3.43163133e-01
1.12717077e-01 4.94627774e-01 -8.00346360e-02 2.72445053e-01
4.95592684e-01 1.05379486e+00 -1.48735869e+00 -5.77283800e-01
1.54085100e-01 1.00515231e-01 -9.00068760e-01 -1.63726240e-01
-4.59338874e-01 6.43246472e-01 -6.37544513e-01 7.32674181e-01
3.36941391e-01 -1.80947691e-01 -1.84279427e-01 -5.89908846e-02
-2.15386823e-01 2.65326574e-02 -1.02447450e+00 2.10922813e+00
-6.38093174e-01 4.47348714e-01 3.98458213e-01 -6.42538905e-01
5.11377811e-01 -1.03780806e-01 4.15546805e-01 -6.01766765e-01
1.42892554e-01 7.72749931e-02 1.53284833e-01 -2.28013501e-01
1.19106662e+00 -8.32773298e-02 -4.32868421e-01 2.17017561e-01
-4.84601334e-02 -7.26000547e-01 1.90824196e-01 2.23148614e-02
1.18343067e+00 3.90657902e-01 3.13679934e-01 -5.61554730e-01
1.88884690e-01 3.76296967e-01 1.86228380e-01 9.00411546e-01
-2.87636727e-01 5.51581740e-01 5.40660083e-01 -2.94490635e-01
-1.06775439e+00 -1.28563678e+00 6.13728026e-03 1.80207193e+00
3.46270323e-01 -8.67352188e-02 -7.83567607e-01 -6.37180984e-01
1.99344754e-01 6.50563955e-01 -9.73189175e-01 6.37117773e-02
-7.32721508e-01 -3.02866213e-02 1.74667284e-01 8.95102501e-01
1.75747037e-01 -1.54702115e+00 -1.29101896e+00 3.90344024e-01
-1.09342054e-01 -6.06276512e-01 -2.52014875e-01 5.79994559e-01
-1.00078559e+00 -9.18762982e-01 -8.32841754e-01 -7.94739306e-01
7.25835621e-01 1.23557761e-01 1.40360606e+00 -1.93901584e-01
-3.47799778e-01 2.71560282e-01 -4.81329203e-01 -4.06975478e-01
-1.35580570e-01 2.80037135e-01 -1.22340463e-01 -8.10708284e-01
-2.31541038e-01 -2.63288409e-01 -4.56217527e-01 3.05377040e-02
-5.88811398e-01 -1.56457454e-01 8.90288830e-01 8.74785304e-01
7.82519817e-01 -8.63287002e-02 6.23044431e-01 -6.12131238e-01
7.16172338e-01 -2.72309989e-01 -6.01048946e-01 -2.23921657e-01
-5.18653169e-02 3.39774549e-01 5.88814616e-01 -3.50558847e-01
-6.93602145e-01 -6.25945441e-03 -1.33925094e-03 -6.04023457e-01
-6.11200146e-02 4.29039687e-01 1.05049416e-01 -1.09293228e-02
5.17025590e-01 -3.97382706e-01 -3.62760067e-01 -4.64358002e-01
8.75103652e-01 9.43794698e-02 6.36996865e-01 -7.35920787e-01
4.67548758e-01 2.21647248e-01 1.53648168e-01 -5.99198759e-01
-5.38467407e-01 -1.86624259e-01 -8.97538602e-01 -3.00581176e-02
1.14080644e+00 -4.31305081e-01 -1.04587209e+00 8.32333118e-02
-1.36584353e+00 -7.42708743e-01 -8.43032658e-01 1.90894067e-01
-1.08052838e+00 -2.59890445e-02 -6.81514263e-01 -7.26204813e-01
-2.65278351e-02 -1.41475427e+00 1.52783108e+00 -2.09650487e-01
-5.48919201e-01 -4.05081987e-01 -2.05463499e-01 -1.60983372e-02
3.25816780e-01 2.08716661e-01 1.14205968e+00 -1.83687553e-01
-8.28495204e-01 3.63621175e-01 -4.28721935e-01 -3.02497923e-01
1.61299314e-02 -3.62530231e-01 -6.61492884e-01 -3.83354098e-01
-2.32121095e-01 -5.06454110e-01 8.25077653e-01 4.95731890e-01
1.48212624e+00 -1.01183303e-01 -3.49128813e-01 2.47911960e-01
1.16315687e+00 9.79468971e-02 5.17414391e-01 3.23714286e-01
6.93728268e-01 6.73915505e-01 1.09379208e+00 3.36187661e-01
3.65544051e-01 6.99815929e-01 1.03171313e+00 -1.31847873e-01
-2.13710248e-01 -1.30643556e-02 8.20572954e-03 2.45441973e-01
-2.00736567e-01 -4.58423980e-02 -8.87967527e-01 8.20183277e-01
-1.74496222e+00 -6.92064643e-01 3.38505418e-03 2.05841351e+00
6.05622828e-01 3.38681430e-01 1.95881724e-01 1.85376763e-01
4.28285211e-01 -7.52526149e-02 -7.79409230e-01 -7.88400173e-01
7.12215900e-01 4.63929355e-01 5.42811275e-01 7.49639928e-01
-9.02733445e-01 1.23152137e+00 7.13741302e+00 6.36559308e-01
-7.92533100e-01 9.46091861e-02 1.88532472e-01 -9.86303613e-02
-8.70744213e-02 -2.47899592e-01 -4.24694508e-01 -5.00861257e-02
3.74982595e-01 9.80383009e-02 5.24564087e-01 9.31497455e-01
-1.28328606e-01 -4.30304497e-01 -1.54308629e+00 6.13583446e-01
-3.98878872e-01 -1.01937902e+00 -4.23986726e-02 -3.84759195e-02
5.33436775e-01 7.10481927e-02 2.15002105e-01 7.17458367e-01
7.40260541e-01 -1.21078074e+00 9.66654360e-01 1.74053416e-01
8.14627528e-01 -7.49080479e-01 5.19488692e-01 4.65809435e-01
-1.35193050e+00 -3.60680521e-01 -5.79117119e-01 -2.23281801e-01
1.24595284e-01 6.70648366e-02 -1.01234925e+00 4.42080230e-01
9.05089378e-01 2.90884197e-01 -1.89993799e-01 7.69015431e-01
-2.25841671e-01 -3.07529420e-01 -1.88698828e-01 -8.32825676e-02
4.34538513e-01 1.09644517e-01 5.36914647e-01 1.03791487e+00
2.39753276e-01 1.72954023e-01 3.38380903e-01 1.17734396e+00
1.47615150e-01 -2.60363072e-01 -4.82282609e-01 2.99626082e-01
4.27682728e-01 1.00558615e+00 -9.24456298e-01 -5.50409138e-01
4.24303897e-02 1.17160380e+00 6.95247710e-01 2.16608107e-01
-8.63616765e-01 -5.30500591e-01 6.56818151e-01 1.03360079e-01
5.59053004e-01 -6.23825192e-01 -5.27758181e-01 -4.23751503e-01
-3.80930364e-01 -7.13813961e-01 1.44997060e-01 -1.17238963e+00
-9.84063685e-01 6.28403783e-01 2.87370682e-01 -9.42849994e-01
-3.08122903e-01 -6.55697703e-01 1.49309710e-02 1.07602024e+00
-1.36666834e+00 -1.01133394e+00 -4.65647668e-01 4.48518217e-01
9.60012376e-01 6.27315342e-01 8.48497629e-01 -2.57117897e-01
-2.37930156e-02 2.09663406e-01 -2.79192090e-01 -3.01435590e-01
2.58328348e-01 -1.58072889e+00 5.69375873e-01 3.89228523e-01
-9.80607942e-02 6.54486239e-01 8.33860636e-01 -7.30424821e-01
-1.24007869e+00 -9.34565902e-01 4.80127752e-01 -6.46785080e-01
5.36738276e-01 -3.20347011e-01 -8.90498877e-01 9.31102395e-01
1.10361166e-01 -7.67888799e-02 -1.20036967e-01 3.65050524e-01
-1.48281291e-01 2.99525648e-01 -1.37381101e+00 5.64941764e-01
1.29801464e+00 -3.21511596e-01 -1.05756938e+00 1.50980145e-01
9.81370091e-01 -6.30283594e-01 -1.09843695e+00 3.02167237e-01
6.37847841e-01 -8.71420741e-01 9.58759964e-01 -5.09426415e-01
5.03553808e-01 -4.88726228e-01 -8.20975602e-02 -1.51045525e+00
-1.06645811e+00 -3.92257154e-01 -2.07993817e-02 5.90839624e-01
2.59913623e-01 -5.41214049e-01 7.82775402e-01 1.99848592e-01
-4.48551923e-01 -6.45029604e-01 -9.03564155e-01 -7.46600151e-01
4.91546720e-01 -3.04725826e-01 6.73212171e-01 6.16324306e-01
1.64223060e-01 2.85811365e-01 1.56436265e-01 3.89193296e-01
4.81424361e-01 1.97289556e-01 7.37995148e-01 -1.28602123e+00
-2.42709026e-01 -6.39839530e-01 -1.74614668e-01 -1.39333010e+00
-6.09444045e-02 -6.37973726e-01 7.27703989e-01 -1.88889611e+00
1.41314641e-01 -5.26639819e-01 -3.62584144e-01 5.91349006e-01
-1.21012174e-01 2.34965980e-01 5.15276790e-01 1.88782379e-01
-6.83567643e-01 5.54703414e-01 1.95250773e+00 -2.19827354e-01
-2.64147133e-01 -5.66745698e-01 -6.14268482e-01 7.11201549e-01
6.05403244e-01 -3.79646301e-01 -3.13296050e-01 -5.89683712e-01
-6.81490228e-02 3.63736212e-01 3.08816582e-01 -1.18405926e+00
5.85631020e-02 -1.67822078e-01 5.55819631e-01 -8.39572906e-01
5.03779829e-01 -7.26437151e-01 -2.13295728e-01 6.01280391e-01
-5.03584266e-01 3.12281072e-01 3.30760956e-01 4.45583582e-01
9.04397741e-02 -6.08281419e-02 7.69085884e-01 -5.18677235e-01
-9.18750823e-01 2.56144941e-01 -6.06107652e-01 -1.63306817e-01
9.68126774e-01 -4.56469446e-01 -1.76101372e-01 -3.41156796e-02
-1.03529000e+00 4.44053113e-01 6.96512341e-01 3.63144159e-01
5.64254105e-01 -1.19538534e+00 -5.22734165e-01 1.85676992e-01
1.51012525e-01 6.86156452e-01 6.01736605e-02 6.90131068e-01
-5.05170882e-01 3.89746606e-01 -6.86814427e-01 -6.62396967e-01
-9.65829730e-01 9.13846850e-01 9.27712396e-02 -5.23941398e-01
-6.14507198e-01 1.10341489e+00 4.92160350e-01 -1.42333299e-01
3.04219782e-01 -1.20324612e+00 -1.51938364e-01 2.32867636e-02
1.72452629e-01 4.05984700e-01 -1.17077269e-01 -4.49221075e-01
-3.46747905e-01 8.25975478e-01 2.02712752e-02 -2.04165250e-01
1.01005423e+00 -1.34390637e-01 -1.15691304e-01 4.15237606e-01
6.86851740e-01 -2.34828532e-01 -1.65364015e+00 3.60240161e-01
4.44756411e-02 -4.14588809e-01 -2.65968405e-02 -6.69662774e-01
-6.22917116e-01 8.72300088e-01 2.53467441e-01 4.64473128e-01
1.01233196e+00 1.68266565e-01 5.84427238e-01 5.48445523e-01
9.45933759e-01 -9.19993401e-01 4.81344461e-01 9.01270866e-01
1.61041224e+00 -8.56822133e-01 5.02081495e-03 -3.41808945e-01
-4.36840475e-01 8.48169267e-01 8.55384648e-01 -3.42820257e-01
1.03171021e-02 3.12070757e-01 -3.75369638e-01 -3.78509909e-01
-6.26191914e-01 -4.80545223e-01 -2.18510702e-02 8.26385796e-01
2.66655594e-01 2.53437776e-02 2.26872668e-01 1.68969318e-01
-4.22458827e-01 1.33421144e-03 4.31937635e-01 1.10369170e+00
-8.60478878e-01 -6.72120392e-01 -5.48537672e-01 3.67254168e-01
-1.33957133e-01 -4.03912961e-02 -2.79943824e-01 1.02050388e+00
2.36989230e-01 6.69800878e-01 4.39532399e-01 -1.21340193e-01
5.49865723e-01 -2.97018439e-01 1.11252344e+00 -8.79319370e-01
-7.10049033e-01 -6.54937625e-02 -3.90837900e-02 -1.10703218e+00
-1.39074773e-01 -5.98549187e-01 -1.53672123e+00 -1.28062144e-02
-2.04813495e-01 -1.08950593e-01 3.73254299e-01 6.70525074e-01
2.56048918e-01 1.25919569e+00 7.56961405e-02 -1.80682313e+00
-6.06477916e-01 -1.13971066e+00 -5.64536572e-01 2.60925353e-01
6.17513478e-01 -1.14948320e+00 -9.38360579e-03 -2.80953646e-01] | [4.737096309661865, 0.592918336391449] |
89b3cc43-b77c-4a40-828e-c613ac11ffcf | domain-agnostic-few-shot-learning-for | 2111.00007 | null | https://arxiv.org/abs/2111.00007v1 | https://arxiv.org/pdf/2111.00007v1.pdf | Domain Agnostic Few-Shot Learning For Document Intelligence | Few-shot learning aims to generalize to novel classes with only a few samples with class labels. Research in few-shot learning has borrowed techniques from transfer learning, metric learning, meta-learning, and Bayesian methods. These methods also aim to train models from limited training samples, and while encouraging performance has been achieved, they often fail to generalize to novel domains. Many of the existing meta-learning methods rely on training data for which the base classes are sampled from the same domain as the novel classes used for meta-testing. However, in many applications in the industry, such as document classification, collecting large samples of data for meta-learning is infeasible or impossible. While research in the field of the cross-domain few-shot learning exists, it is mostly limited to computer vision. To our knowledge, no work yet exists that examines the use of few-shot learning for classification of semi-structured documents (scans of paper documents) generated as part of a business workflow (forms, letters, bills, etc.). Here the domain shift is significant, going from natural images to the semi-structured documents of interest. In this work, we address the problem of few-shot document image classification under domain shift. We evaluate our work by extensive comparisons with existing methods. Experimental results demonstrate that the proposed method shows consistent improvements on the few-shot classification performance under domain shift. | ['Glenn Fung', 'Eric Bunch', 'Jaya Krishna Mandivarapu'] | 2021-10-29 | null | null | null | null | ['document-image-classification', 'cross-domain-few-shot', 'cross-domain-few-shot-learning'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 5.66850781e-01 -7.84296840e-02 -5.00741243e-01 -6.43591285e-01
-9.08179700e-01 -1.28603667e-01 9.44897652e-01 2.71000355e-01
-1.26074657e-01 8.20371568e-01 -7.31319189e-02 7.55416080e-02
-3.23805392e-01 -8.52903187e-01 -5.35946131e-01 -5.50651848e-01
2.66802102e-01 7.01517463e-01 6.05336368e-01 -2.50270694e-01
6.60527110e-01 2.51646936e-01 -1.95544124e+00 7.00952590e-01
5.89041352e-01 8.51945043e-01 3.38638246e-01 6.76829517e-01
-6.30843878e-01 6.84299290e-01 -7.04261363e-01 -1.29262403e-01
8.20591897e-02 -7.78939247e-01 -9.65905309e-01 5.71724474e-01
4.06759083e-01 -3.30204219e-01 -1.64237425e-01 9.62435484e-01
3.61095697e-01 7.80155182e-01 1.00515771e+00 -1.15328276e+00
-7.40589917e-01 3.78705829e-01 -4.91956234e-01 3.03703099e-01
5.19291341e-01 -2.35711530e-01 5.83396673e-01 -1.11819589e+00
9.22608733e-01 1.11256695e+00 5.68357408e-01 6.88740075e-01
-9.38998878e-01 -4.42835987e-01 -9.44855809e-02 5.74839652e-01
-1.07757914e+00 -4.50087398e-01 7.88283527e-01 -4.73081976e-01
8.35722208e-01 -1.91082686e-01 3.32739204e-01 1.29665029e+00
1.01806834e-01 9.64280784e-01 1.22143507e+00 -9.87428486e-01
8.85474801e-01 6.56577945e-01 4.49004680e-01 3.10560673e-01
8.26470405e-02 -9.37435552e-02 -6.29231632e-01 -8.72738138e-02
1.35848045e-01 4.05158281e-01 6.94326982e-02 -6.42175555e-01
-6.53545916e-01 9.59042013e-01 -1.70711249e-01 7.21630394e-01
-1.78471476e-01 -4.84248668e-01 6.21060014e-01 5.13312340e-01
7.94665098e-01 2.75999516e-01 -2.69502968e-01 -3.40282977e-01
-1.30965960e+00 1.77202031e-01 9.56747234e-01 1.44190192e+00
9.57587898e-01 -5.74808121e-02 -2.55217344e-01 1.12415004e+00
-9.36732218e-02 -2.47618258e-02 1.10438514e+00 -6.67429745e-01
3.37785274e-01 4.77785498e-01 4.38275263e-02 -6.16893470e-01
3.04274056e-02 5.36981486e-02 -4.02933002e-01 8.00359547e-02
1.05715290e-01 -4.26490679e-02 -1.20788229e+00 1.01303542e+00
2.18154326e-01 2.59451956e-01 1.06587112e-01 4.82959419e-01
8.17421854e-01 8.02859843e-01 -1.87994137e-01 -5.01120687e-01
8.83091092e-01 -9.69582260e-01 -7.42229760e-01 -2.67387152e-01
7.85409629e-01 -6.98713005e-01 1.14576423e+00 4.52178150e-01
-9.04051125e-01 -7.79217660e-01 -1.18156111e+00 3.80276829e-01
-8.86034429e-01 -6.32468522e-01 4.22119021e-01 8.89935136e-01
-4.14885670e-01 8.82227778e-01 -4.69843626e-01 -1.03273141e+00
6.15242183e-01 -6.13401532e-02 -2.74089903e-01 -7.48723388e-01
-1.02537441e+00 9.50839341e-01 7.26719201e-01 -7.49655485e-01
-8.01111698e-01 -6.25670433e-01 -9.28651989e-01 -1.33330580e-02
5.74959576e-01 -2.59683341e-01 1.51387000e+00 -8.75999570e-01
-1.54778862e+00 9.22060311e-01 3.78417969e-02 -4.59355384e-01
3.15512925e-01 7.06486125e-03 -6.06539488e-01 8.56331214e-02
4.10533063e-02 2.85130382e-01 1.08332705e+00 -1.18929446e+00
-8.83381546e-01 -4.14206415e-01 -9.30116847e-02 -3.30881290e-02
-7.00529575e-01 -9.33481567e-03 -1.44304201e-01 -5.57660699e-01
-2.32845411e-01 -7.87125170e-01 3.33091873e-03 -3.42234135e-01
5.77310920e-02 -3.25705290e-01 1.32427537e+00 -2.11210579e-01
1.01771235e+00 -2.17999196e+00 -2.50072211e-01 -1.41960144e-01
-3.83906066e-01 5.20561695e-01 5.99003732e-02 6.58139288e-01
-1.22933721e-04 -1.70480177e-01 -2.24859238e-01 3.53034143e-03
-1.21491544e-01 2.63983339e-01 -3.53942633e-01 2.94107229e-01
-9.66744125e-02 6.92219436e-01 -1.08983743e+00 -6.92050040e-01
4.56746668e-01 -6.81934208e-02 -2.51829565e-01 8.88907835e-02
-1.79898262e-01 -8.00931528e-02 -3.21141601e-01 7.73321450e-01
5.56327164e-01 -2.19907835e-01 8.11649710e-02 1.44148588e-01
8.14349502e-02 -2.15364739e-01 -1.21147573e+00 1.99498987e+00
-5.62643349e-01 5.69172263e-01 -5.27616143e-01 -1.63542330e+00
1.16583645e+00 3.25695276e-01 4.67343897e-01 -5.28647840e-01
1.57594338e-01 1.12985611e-01 -9.91919637e-02 -6.57947421e-01
5.48499882e-01 -7.66784072e-01 -3.71394195e-02 6.76911354e-01
6.05754316e-01 -3.83396864e-01 5.11446178e-01 -1.38136046e-02
1.11093891e+00 1.48901805e-01 6.39802277e-01 6.00975975e-02
2.36820653e-01 3.22089285e-01 3.55518609e-01 1.00016308e+00
-4.15803641e-01 7.43755043e-01 -4.96147424e-02 -4.69073206e-01
-1.20496821e+00 -9.42321897e-01 -2.64722288e-01 1.43533039e+00
4.53855321e-02 -3.02273422e-01 -5.64779997e-01 -7.27898657e-01
-2.32471973e-02 1.08251333e+00 -5.91380835e-01 -5.21139324e-01
-1.66832566e-01 -5.86137652e-01 4.57152314e-02 6.11491859e-01
2.95041561e-01 -1.08603036e+00 -7.85606563e-01 4.12009001e-01
3.02873999e-01 -9.00416374e-01 -5.15002385e-02 4.12589520e-01
-1.23485458e+00 -1.14840615e+00 -1.04052436e+00 -1.03833508e+00
4.20788080e-01 5.06343663e-01 7.57043660e-01 -5.20995617e-01
-6.53322816e-01 6.72178447e-01 -9.01080966e-01 -8.08380365e-01
-5.48535049e-01 -6.33068010e-02 1.26906008e-01 -3.75544769e-03
1.04625893e+00 -3.89603376e-01 -4.60186191e-02 2.93291301e-01
-1.01486158e+00 -4.60189015e-01 4.72635418e-01 1.25280702e+00
4.20513034e-01 2.95234472e-01 7.78299510e-01 -1.47687411e+00
7.70236313e-01 -6.05896294e-01 -6.24592714e-02 4.78921771e-01
-8.50374877e-01 -2.54827261e-01 5.80505192e-01 -6.71132088e-01
-1.42689967e+00 -2.53858089e-01 5.81717789e-01 -5.54887295e-01
-5.79385579e-01 4.06764925e-01 1.83290944e-01 1.36080146e-01
1.05094492e+00 2.89646626e-01 6.87268972e-02 -5.59782982e-01
4.68221933e-01 1.28557539e+00 1.97316885e-01 -4.23004240e-01
4.16173458e-01 5.56293786e-01 -2.76085287e-01 -1.19426823e+00
-1.01919675e+00 -9.98379409e-01 -1.00184119e+00 -2.12587550e-01
4.32843983e-01 -3.95896792e-01 2.90183574e-01 1.76255822e-01
-7.28885353e-01 -4.01059203e-02 -8.46335769e-01 6.22990727e-01
-1.09681213e+00 4.28675532e-01 -3.15391868e-01 -8.29396904e-01
-3.09438799e-02 -5.62363744e-01 7.65897810e-01 2.56702632e-01
-3.63192618e-01 -1.02244878e+00 2.90610492e-01 3.13624084e-01
2.58173585e-01 1.58032939e-01 1.05363905e+00 -1.03558195e+00
2.59174611e-02 -6.93511724e-01 2.19964907e-01 5.08274078e-01
5.20708025e-01 -2.32152641e-01 -1.05781877e+00 -4.08710420e-01
2.20655903e-01 -7.37788498e-01 9.46514666e-01 4.29445833e-01
1.10961831e+00 1.80577621e-01 -5.42198718e-01 1.89602897e-01
1.26216185e+00 6.85290575e-01 6.13977134e-01 4.23160940e-01
1.44742668e-01 8.17259789e-01 1.23444891e+00 8.18172693e-01
-2.39599660e-01 3.59635890e-01 -3.19044478e-02 5.65769494e-01
-1.30057007e-01 1.96957402e-02 1.11754619e-01 6.29932702e-01
4.57623191e-02 -1.04508206e-01 -9.61226642e-01 8.02518725e-01
-1.94587755e+00 -1.39710593e+00 3.93849313e-01 2.12260485e+00
6.89899147e-01 3.42356235e-01 -1.60169434e-02 4.31873649e-01
9.94000256e-01 7.99224004e-02 -6.79456115e-01 -5.95320582e-01
2.30464667e-01 3.65888447e-01 -1.45029053e-02 -7.41507933e-02
-1.24602282e+00 8.77705276e-01 6.34544516e+00 9.14692342e-01
-7.18268514e-01 3.23768675e-01 3.28844845e-01 -2.56752610e-01
2.81391323e-01 8.30693766e-02 -8.79576981e-01 2.90673494e-01
1.06233716e+00 -3.23990673e-01 1.15136802e-01 1.25724494e+00
-1.64770633e-01 -1.79917201e-01 -1.41226637e+00 1.10603940e+00
6.54243290e-01 -1.30317259e+00 1.28415540e-01 -1.19242229e-01
1.20807719e+00 -2.42916986e-01 -2.94967107e-02 8.53693962e-01
7.65697658e-02 -7.58846402e-01 2.60942400e-01 5.43780506e-01
8.19739759e-01 -9.20119166e-01 7.45290875e-01 7.17928886e-01
-6.36639059e-01 -3.95574033e-01 -8.75085711e-01 -9.46109220e-02
3.70106027e-02 5.37260354e-01 -1.25974929e+00 3.23993504e-01
7.00888157e-01 8.91669810e-01 -3.84771615e-01 1.07405198e+00
3.39560598e-01 4.33778405e-01 2.95717686e-01 -3.15205753e-01
3.17534208e-01 -4.12714966e-02 4.24352765e-01 1.21925819e+00
4.42435235e-01 3.58626954e-02 3.32746029e-01 5.37843466e-01
8.07286426e-02 1.87116072e-01 -1.15977108e+00 -2.35094398e-01
3.06828648e-01 9.29464936e-01 -7.78371513e-01 -7.32175767e-01
-8.07794154e-01 9.80380476e-01 -3.93468216e-02 1.46585807e-01
-3.46224308e-01 -9.49967384e-01 1.79067403e-01 2.98152119e-01
5.01543641e-01 1.81180984e-01 -1.02698386e-01 -8.46340120e-01
-1.31446287e-01 -6.46349669e-01 6.25199437e-01 -6.40096724e-01
-1.66517687e+00 1.63916051e-01 3.91720057e-01 -1.58941174e+00
-7.68720031e-01 -7.31940687e-01 -6.42015755e-01 4.61594671e-01
-1.31261253e+00 -9.94684577e-01 -3.16793233e-01 6.28186107e-01
1.31011450e+00 -8.70793998e-01 1.08413661e+00 -1.56475902e-02
-2.41227057e-02 3.37913066e-01 7.54131079e-01 -1.31696388e-01
1.12737894e+00 -1.10388756e+00 1.04569942e-01 3.76818180e-01
1.51251569e-01 5.36051393e-01 6.75135672e-01 -7.02416539e-01
-1.14346600e+00 -9.75716054e-01 4.95308697e-01 -3.86327744e-01
5.10627747e-01 -2.33135656e-01 -1.14091623e+00 5.21504104e-01
5.07621244e-02 3.54874760e-01 1.13372278e+00 1.76909670e-01
-3.11129093e-01 -2.26039402e-02 -1.26974261e+00 2.72681952e-01
6.70742333e-01 -5.38358867e-01 -1.16212833e+00 5.41932106e-01
3.10433060e-01 5.31578884e-02 -8.57431710e-01 1.33981913e-01
5.43840110e-01 -9.15107727e-01 7.46742964e-01 -1.13729966e+00
4.63648438e-01 9.35366377e-02 -5.24692237e-01 -1.55480349e+00
-4.03945029e-01 -4.23570015e-02 -5.25825679e-01 1.24469471e+00
-9.44199879e-03 -6.16357513e-02 1.02195156e+00 3.15842867e-01
-1.88669637e-01 -4.93222266e-01 -7.64934838e-01 -1.40526366e+00
1.42922789e-01 -3.68866891e-01 3.86963993e-01 1.14134562e+00
3.08688343e-01 3.48737657e-01 -3.14622909e-01 -5.46490848e-01
7.21573770e-01 4.63914543e-01 7.88113654e-01 -1.63311088e+00
-2.80701667e-01 -1.68577954e-01 -6.73229814e-01 -4.56193328e-01
2.48684689e-01 -9.36298668e-01 1.69877753e-01 -1.66769803e+00
4.85435367e-01 -1.25715393e-03 -3.94382745e-01 2.02009141e-01
2.34556004e-01 -2.39937324e-02 1.53314754e-01 1.37824625e-01
-7.88292825e-01 4.65255082e-01 1.11418784e+00 -5.98115921e-01
-3.20983410e-01 2.00988412e-01 -5.12872756e-01 8.83186877e-01
7.86594808e-01 -5.60550749e-01 -6.82031810e-01 2.51027465e-01
-3.81915718e-01 -1.48082078e-01 -9.51981843e-02 -1.11255658e+00
3.27362448e-01 -4.31037039e-01 5.51563144e-01 -6.61749601e-01
3.46786588e-01 -6.78384781e-01 -3.57479870e-01 3.05347294e-01
-1.98021755e-01 -6.61763310e-01 -1.49229124e-01 9.48475003e-01
-3.78825128e-01 -1.07939017e+00 1.07174945e+00 -7.43873894e-01
-1.37353337e+00 1.23583153e-01 -4.37861919e-01 2.68566430e-01
1.50240350e+00 -7.66043663e-01 -1.43210441e-01 -2.29502544e-01
-8.60882223e-01 -2.29352921e-01 4.27579612e-01 7.17241883e-01
8.61830115e-01 -1.36902845e+00 -5.18392265e-01 9.25687701e-02
8.65945458e-01 -1.10932432e-01 2.18358442e-01 4.84462947e-01
-9.87454653e-02 4.31499958e-01 -3.55408162e-01 -5.94618320e-01
-1.05468464e+00 9.04923916e-01 -2.15949476e-01 1.17286667e-02
-7.79003918e-01 5.77042043e-01 -1.51781410e-01 -3.47737789e-01
2.91614681e-01 5.42573743e-02 -2.44026572e-01 5.11992991e-01
8.61735344e-01 7.74793148e-01 3.10926765e-01 -1.86307669e-01
-1.35045648e-01 3.17390412e-01 -5.63751638e-01 1.68326143e-02
1.60452974e+00 3.04077286e-02 3.81268173e-01 1.23849940e+00
1.18730247e+00 -7.13600397e-01 -1.02594709e+00 -5.99038780e-01
4.28526640e-01 -7.49885798e-01 -6.85900226e-02 -5.39727867e-01
-2.97232002e-01 1.06034505e+00 8.00447524e-01 1.14893220e-01
7.57698953e-01 9.81713459e-02 6.23133540e-01 8.37950885e-01
8.12794268e-01 -2.01912022e+00 6.34703875e-01 6.85430110e-01
4.90512580e-01 -1.58308363e+00 4.57227081e-02 8.16723555e-02
-6.10598505e-01 1.32638359e+00 6.05038345e-01 -1.67464465e-02
9.45752144e-01 4.71880957e-02 -3.99725944e-01 -9.46150199e-02
-6.86050117e-01 -5.68640865e-02 8.24800655e-02 1.15938258e+00
4.31771755e-01 -1.00687318e-01 -8.74841660e-02 6.09039962e-01
1.58776313e-01 3.27047884e-01 7.71555066e-01 1.71251488e+00
-1.03774631e+00 -1.04960597e+00 -4.66209203e-01 9.35359180e-01
-1.21659741e-01 1.99883848e-01 -3.46537054e-01 7.14822590e-01
4.55945060e-02 9.19280112e-01 1.25071302e-01 -1.80967703e-01
4.46730375e-01 7.19610691e-01 6.93102241e-01 -1.49391437e+00
1.16028357e-02 -2.70343930e-01 -1.99065328e-01 -5.19693606e-02
-6.22524738e-01 -7.98194408e-01 -1.02832937e+00 -4.56545036e-03
-4.97149497e-01 4.70396057e-02 7.03168631e-01 1.06872499e+00
-4.55680937e-02 5.32439291e-01 6.87867820e-01 -9.43281889e-01
-8.40386510e-01 -1.29127693e+00 -1.12266946e+00 5.29829979e-01
7.35077485e-02 -8.84296060e-01 -8.74704644e-02 3.51439834e-01] | [10.024373054504395, 3.0845348834991455] |
2e8e0a96-c717-4c9f-9853-f257f0af013d | using-context-information-for-dialog-act | null | null | https://aclanthology.org/D17-1231 | https://aclanthology.org/D17-1231.pdf | Using Context Information for Dialog Act Classification in DNN Framework | Previous work on dialog act (DA) classification has investigated different methods, such as hidden Markov models, maximum entropy, conditional random fields, graphical models, and support vector machines. A few recent studies explored using deep learning neural networks for DA classification, however, it is not clear yet what is the best method for using dialog context or DA sequential information, and how much gain it brings. This paper proposes several ways of using context information for DA classification, all in the deep learning framework. The baseline system classifies each utterance using the convolutional neural networks (CNN). Our proposed methods include using hierarchical models (recurrent neural networks (RNN) or CNN) for DA sequence tagging where the bottom layer takes the sentence CNN representation as input, concatenating predictions from the previous utterances with the CNN vector for classification, and performing sequence decoding based on the predictions from the sentence CNN model. We conduct thorough experiments and comparisons on the Switchboard corpus, demonstrate that incorporating context information significantly improves DA classification, and show that we achieve new state-of-the-art performance for this task. | ['Yang Liu', 'Kun Han', 'Zhao Tan', 'Yun Lei'] | 2017-09-01 | null | null | null | emnlp-2017-9 | ['dialog-act-classification'] | ['natural-language-processing'] | [ 1.29325375e-01 1.43087640e-01 -1.90875277e-01 -7.51324713e-01
-1.36170134e-01 -5.84465027e-01 6.47235930e-01 -3.07964701e-02
-3.38290840e-01 7.96098351e-01 8.05739641e-01 -6.36000633e-01
5.79954624e-01 -5.58385551e-01 1.53826401e-01 -6.47251487e-01
9.06460881e-02 4.60426718e-01 2.98531830e-01 -5.03692210e-01
2.21123785e-01 3.03505003e-01 -9.63812411e-01 7.07774401e-01
3.82353008e-01 1.13883138e+00 1.79164886e-01 9.42251146e-01
-6.80116117e-01 1.75523221e+00 -1.00001585e+00 -4.19678003e-01
-2.21963257e-01 -5.59928656e-01 -1.39500916e+00 -5.94022647e-02
-1.46829695e-01 -8.92680585e-01 -4.80893105e-01 4.44701314e-01
3.03697914e-01 2.21220925e-01 5.87334096e-01 -9.15615261e-01
-6.73340976e-01 9.41103041e-01 2.79131591e-01 1.16319254e-01
5.35531580e-01 1.56705841e-01 1.28910458e+00 -5.82853734e-01
4.68026131e-01 1.51637530e+00 5.57942390e-01 9.12343442e-01
-1.13028097e+00 -3.98104846e-01 2.51796067e-01 4.17554602e-02
-7.46983409e-01 -3.59044969e-01 9.77641404e-01 -4.90200698e-01
1.65056360e+00 4.97922152e-02 3.79794210e-01 1.58323622e+00
4.76832867e-01 1.25553310e+00 8.79352391e-01 -7.13603437e-01
3.66023421e-01 -8.01167171e-03 9.18478608e-01 7.47818530e-01
-7.64252782e-01 -1.26051918e-01 -4.01772112e-01 -5.48977911e-01
3.84149194e-01 1.43999130e-01 1.18527919e-01 1.47831798e-01
-8.83343756e-01 1.48441803e+00 3.50522488e-01 4.58148718e-01
-2.86985636e-01 -1.73298270e-01 7.11619854e-01 4.46519762e-01
3.66848946e-01 2.91469902e-01 -8.43140900e-01 -2.70651013e-01
-5.28122306e-01 3.29785466e-01 1.32375479e+00 8.33046317e-01
4.56606984e-01 7.60056898e-02 -5.02879024e-01 1.05919516e+00
4.84347343e-01 2.50576269e-02 7.03107059e-01 -1.02468562e+00
5.44272065e-01 5.85736573e-01 1.24429222e-02 -4.56042320e-01
-6.04742467e-01 3.67109984e-01 -7.31676519e-01 5.72255987e-04
4.99310821e-01 -8.34289432e-01 -7.24690974e-01 1.77893794e+00
-1.41398221e-01 -3.60267460e-01 3.95515442e-01 5.55619001e-01
9.88525093e-01 8.35345089e-01 5.10562539e-01 -1.70400053e-01
1.22788906e+00 -1.19528687e+00 -1.11302078e+00 -2.81021774e-01
1.03813231e+00 -6.73388898e-01 7.07228243e-01 2.86612064e-01
-9.75790441e-01 -4.75189894e-01 -1.17749166e+00 -3.36948574e-01
-5.55971503e-01 3.59685361e-01 7.08276689e-01 6.74333274e-01
-1.18276393e+00 4.42462951e-01 -8.85975599e-01 -4.44958806e-01
-3.58741395e-02 5.11593819e-01 -8.47654138e-03 4.38028753e-01
-1.51521921e+00 1.26390636e+00 3.51273775e-01 6.85810819e-02
-8.19901705e-01 1.30176485e-01 -1.13819003e+00 3.36102545e-01
5.52411824e-02 -4.17652458e-01 1.93124080e+00 -9.42404032e-01
-2.31155705e+00 3.88380498e-01 -3.09470981e-01 -8.37723315e-01
-1.60396263e-01 -1.21955931e-01 -1.42414778e-01 -2.97084242e-01
-3.94703984e-01 9.30173695e-01 3.78596038e-01 -8.38218570e-01
-4.89003420e-01 -2.21916184e-01 4.20928478e-01 3.40055116e-02
-1.89778849e-01 2.33790264e-01 2.66838491e-01 -3.59943032e-01
1.21273831e-01 -1.18110883e+00 -4.17477518e-01 -5.51811993e-01
-5.82566023e-01 -7.29042411e-01 8.65705132e-01 -1.00631571e+00
1.39852262e+00 -2.08324647e+00 3.03830117e-01 -2.61259705e-01
6.11594394e-02 3.53325814e-01 2.25560933e-01 7.50277042e-01
2.94312716e-01 -1.61839098e-01 -2.52620071e-01 -6.95194781e-01
1.58207178e-01 5.64347982e-01 -5.13303697e-01 -2.03428745e-01
2.31208324e-01 8.23595047e-01 -4.88682151e-01 -1.49736777e-01
3.16505700e-01 2.38748714e-01 -4.21491921e-01 6.83385193e-01
-5.97216249e-01 5.59047818e-01 -2.48072654e-01 2.62063533e-01
1.48591205e-01 -1.08854361e-01 9.15706396e-01 1.64523900e-01
2.23977342e-02 1.02629066e+00 -4.68593627e-01 1.70025957e+00
-6.18279755e-01 9.79456782e-01 8.91516916e-03 -8.76946688e-01
1.19942760e+00 6.89473450e-01 -2.07991060e-02 -2.73939073e-01
3.18855852e-01 -1.84128895e-01 8.47437903e-02 -4.32392865e-01
5.55703878e-01 -5.96365631e-02 -5.09431481e-01 5.00089526e-01
5.16734600e-01 2.11739749e-01 -2.66169488e-01 2.13896543e-01
1.07413256e+00 -2.17298314e-01 5.90155065e-01 2.06344634e-01
7.68332243e-01 -3.34249474e-02 3.86345774e-01 5.63904583e-01
-5.05611360e-01 1.97855681e-01 8.32111061e-01 -7.27004170e-01
-8.19891274e-01 -7.33785868e-01 3.08159232e-01 1.50488758e+00
-3.74465376e-01 -5.15603602e-01 -9.14251208e-01 -1.08882856e+00
-3.37060064e-01 1.00469887e+00 -5.33440769e-01 3.61007378e-02
-7.35602379e-01 -4.01558667e-01 7.05512762e-01 8.76698911e-01
8.84976149e-01 -1.50805354e+00 -2.22387090e-01 3.90382916e-01
-1.79474413e-01 -1.02225101e+00 -1.63766086e-01 5.71112275e-01
-1.01371658e+00 -5.07729530e-01 -3.50391388e-01 -1.06815386e+00
8.99055004e-02 -1.48200214e-01 9.53353465e-01 7.63888583e-02
4.13343459e-01 3.20948437e-02 -3.57285619e-01 -1.08724631e-01
-9.13206458e-01 3.67559314e-01 -1.82502821e-01 -3.50963920e-01
9.17854249e-01 -3.75584811e-01 -1.68312386e-01 9.73102301e-02
-5.22913933e-01 -1.11194447e-01 5.05767405e-01 1.10132349e+00
-3.14118177e-01 -4.81880128e-01 6.79793775e-01 -9.72518384e-01
9.87636030e-01 -4.14368093e-01 -2.60253489e-01 2.64125496e-01
-2.39258587e-01 3.61791134e-01 3.57542813e-01 -1.75109997e-01
-1.48754060e+00 4.93399024e-01 -7.69978404e-01 9.18162465e-02
-7.51011968e-01 4.90843326e-01 -8.01755488e-02 5.70981205e-01
5.54113388e-01 2.09553331e-01 2.47488022e-01 -5.24973094e-01
5.62694192e-01 1.18657196e+00 1.47721812e-01 -1.70213670e-01
-2.47570425e-01 -5.09867370e-02 -5.87616622e-01 -6.77745104e-01
-9.44351494e-01 -3.73506963e-01 -1.21628129e+00 -2.17618346e-02
1.62912548e+00 -6.72349274e-01 -9.59380329e-01 6.00290418e-01
-1.48822165e+00 -6.29778922e-01 4.84372765e-01 3.98694038e-01
-4.76836026e-01 4.01289642e-01 -1.49976635e+00 -9.90501165e-01
-5.09395361e-01 -1.40301514e+00 5.71391702e-01 2.07445715e-02
-6.46097600e-01 -1.31789660e+00 2.88673371e-01 5.85983694e-01
4.62526172e-01 -3.93286735e-01 1.29421854e+00 -1.29845166e+00
-1.48281723e-01 -5.19355871e-02 2.07651079e-01 5.93923032e-01
1.48230314e-01 -2.48808935e-01 -1.34531200e+00 1.09461427e-01
2.81594068e-01 -6.09162509e-01 8.11719418e-01 4.86473709e-01
8.98650587e-01 -5.58816493e-01 -2.90329337e-01 -3.41421008e-01
8.33521962e-01 8.23646247e-01 6.74221158e-01 -1.87874511e-02
3.91572118e-01 6.17038548e-01 3.03450495e-01 2.20833927e-01
7.07314849e-01 7.01703370e-01 2.27847070e-01 3.81258100e-01
4.76777665e-02 6.80774003e-02 7.48963356e-01 1.07262707e+00
3.67415577e-01 -3.93430024e-01 -9.18637693e-01 2.40305290e-01
-2.00269008e+00 -9.75015342e-01 1.57103971e-01 1.68915629e+00
8.95521581e-01 3.04252833e-01 3.46088648e-01 -1.07988462e-01
5.35021424e-01 3.06543291e-01 -1.30587116e-01 -9.95930791e-01
1.45113423e-01 1.13768622e-01 -5.29840440e-02 7.25287259e-01
-1.29583001e+00 1.33096588e+00 6.81396198e+00 2.77768046e-01
-1.10337532e+00 2.07052127e-01 8.20835769e-01 2.93134719e-01
4.42355037e-01 -9.64977406e-03 -1.11053097e+00 3.39612126e-01
1.49743652e+00 5.56034267e-01 2.34315813e-01 1.21716559e+00
-3.27888802e-02 -1.34352878e-01 -1.23328733e+00 5.55570066e-01
-4.41902392e-02 -1.45538843e+00 -1.25982873e-02 -1.23794749e-01
4.49204415e-01 1.12034883e-02 -2.75613248e-01 9.62643564e-01
8.94938290e-01 -8.67608607e-01 2.96310782e-01 3.38086814e-01
-1.41272068e-01 -5.42989492e-01 1.18146372e+00 4.56819355e-01
-7.81950057e-01 -3.02762717e-01 -4.28267211e-01 -5.42805433e-01
1.43339530e-01 8.83428939e-03 -1.54081321e+00 2.06999257e-01
2.68330306e-01 5.50486863e-01 -3.69790852e-01 2.44816527e-01
-3.26275200e-01 1.06308556e+00 2.03200057e-01 -6.38471484e-01
5.63014686e-01 -4.43326384e-02 2.29817241e-01 1.44429088e+00
-2.58112907e-01 2.51621932e-01 4.45452899e-01 5.65006316e-01
2.52582831e-03 3.27638909e-03 -6.48106873e-01 4.20103408e-02
3.41823459e-01 9.65350807e-01 -4.47026730e-01 -5.91181278e-01
-7.02931702e-01 9.88923252e-01 5.73695064e-01 1.17147855e-01
-4.29749072e-01 -2.98053503e-01 6.57665312e-01 -7.39236236e-01
4.79260057e-01 -6.46452308e-01 -5.19232929e-01 -1.00513566e+00
-3.69578540e-01 -5.73505282e-01 6.35827363e-01 -6.80034697e-01
-1.45723236e+00 8.81861985e-01 -8.29597861e-02 -8.20824802e-01
-8.80806386e-01 -9.12747085e-01 -9.09139454e-01 8.42909217e-01
-8.03039849e-01 -1.07323921e+00 1.51643246e-01 4.48755026e-01
1.13809538e+00 -5.00717402e-01 1.27445292e+00 -4.08062115e-02
-4.17097837e-01 3.49617243e-01 -1.04129620e-01 8.79405737e-01
5.22102654e-01 -1.39914978e+00 4.18738186e-01 1.60939768e-01
-5.28561212e-02 5.80220640e-01 3.18530172e-01 -7.70262778e-01
-7.27016985e-01 -6.55317307e-01 1.29215217e+00 -7.25153804e-01
5.87515891e-01 -7.56053746e-01 -8.87896180e-01 1.15277362e+00
8.79976869e-01 -7.35456407e-01 8.97684097e-01 3.31108660e-01
-1.58459827e-01 4.25825655e-01 -9.95785892e-01 5.89189470e-01
4.93106693e-01 -7.43034601e-01 -1.01473546e+00 2.93480039e-01
1.15834343e+00 -3.09899360e-01 -7.41947114e-01 -3.72864790e-02
4.00416106e-01 -9.33030784e-01 6.88956141e-01 -9.34459090e-01
4.68667269e-01 3.11619818e-01 -3.83077204e-01 -1.17459726e+00
-1.64205819e-01 -3.31301868e-01 -1.12973273e-01 1.26836038e+00
7.59460747e-01 -3.96641314e-01 7.11459458e-01 8.80807579e-01
-3.36381435e-01 -7.35207140e-01 -7.96119273e-01 -3.72480035e-01
4.79970068e-01 -3.58468652e-01 4.05315191e-01 1.08998871e+00
5.20735264e-01 1.09626591e+00 -5.78523695e-01 -1.46698415e-01
-2.56980211e-01 3.04999612e-02 5.98454356e-01 -1.00842130e+00
-1.43826172e-01 -3.26380700e-01 -3.53179127e-01 -1.63023245e+00
5.79629898e-01 -7.08370328e-01 2.38744825e-01 -1.39987779e+00
-3.95338237e-02 -3.23893540e-02 4.73151952e-02 6.42441452e-01
-8.97089466e-02 -5.78762650e-01 2.44610041e-01 1.28883883e-01
-5.09060204e-01 7.19391286e-01 6.23468399e-01 -3.29353273e-01
-4.80412185e-01 3.12487990e-01 -4.52053361e-02 5.73711634e-01
1.09928811e+00 -1.85652077e-01 -2.55520821e-01 -2.08132535e-01
-4.08259958e-01 5.86729348e-01 -9.22357216e-02 -9.69205499e-01
2.74707139e-01 3.16046216e-02 3.99214089e-01 -8.40142965e-01
7.38015354e-01 -5.24048567e-01 -5.40888250e-01 5.45115709e-01
-1.13104796e+00 5.10071628e-02 1.42410189e-01 4.17223722e-01
-3.05914611e-01 -6.43411219e-01 4.51432467e-01 -4.27466154e-01
-6.00697398e-01 -2.59015918e-01 -1.33100939e+00 -4.91591632e-01
4.83867466e-01 4.76809978e-01 -3.42411488e-01 -9.42049146e-01
-1.36088192e+00 1.07481509e-01 -2.42885381e-01 3.64787102e-01
3.12413067e-01 -1.14855731e+00 -2.87098229e-01 -1.88200362e-02
-3.16920698e-01 -4.36006039e-01 1.24803029e-01 5.14204621e-01
-2.53614306e-01 9.14529085e-01 -2.60620058e-01 -5.98046303e-01
-1.39385164e+00 4.15980935e-01 2.87117839e-01 -6.67956412e-01
-1.53017625e-01 8.15413773e-01 8.13673213e-02 -9.34716344e-01
4.72429752e-01 -6.12112463e-01 -7.03205764e-01 -1.36463139e-02
4.45866495e-01 -4.34007542e-03 3.16334777e-02 -5.52582562e-01
-1.38010308e-01 -3.52283269e-01 -5.51704824e-01 -4.50555831e-01
1.05945325e+00 -2.35880196e-01 -3.00678480e-02 6.99520826e-01
1.21882868e+00 -4.41693515e-01 -1.03943610e+00 -4.47814941e-01
4.25964773e-01 1.68653920e-01 -6.12499900e-02 -9.31578755e-01
-6.71145022e-01 1.24451017e+00 4.94655073e-01 5.98053277e-01
7.26792216e-01 4.15483601e-02 1.02719903e+00 1.03470182e+00
5.70520684e-02 -9.85275984e-01 3.63145649e-01 1.34137738e+00
8.06028426e-01 -1.27357650e+00 -3.74845088e-01 -2.01621085e-01
-1.27832818e+00 1.55680275e+00 8.56025159e-01 -6.01771660e-02
1.03749263e+00 5.20418771e-02 4.65769529e-01 -2.62335718e-01
-1.22248363e+00 -2.88742501e-03 -1.22018807e-01 3.33911300e-01
7.99299955e-01 6.04774654e-02 -9.98287648e-02 7.96928346e-01
-3.23410809e-01 -2.73720473e-01 5.15849411e-01 1.31791961e+00
-6.08455539e-01 -1.35277057e+00 6.76286221e-02 3.60174805e-01
-5.39183617e-01 -1.79032281e-01 -1.02769518e+00 4.29016918e-01
-3.85427922e-01 1.28076363e+00 1.82934701e-01 -8.56336653e-01
-6.53141225e-03 1.05199993e+00 -2.21560411e-02 -8.27768028e-01
-1.09905815e+00 -1.77654997e-01 6.39908254e-01 -1.64878532e-01
-4.66623068e-01 -6.91704512e-01 -1.31937480e+00 -2.11524129e-01
-4.87776369e-01 5.95701709e-02 6.79737985e-01 1.48812938e+00
1.96734443e-01 1.71247452e-01 7.34553218e-01 -6.99240983e-01
-6.74434960e-01 -1.58083534e+00 -6.19678572e-02 7.13208914e-02
3.53212625e-01 -4.85984415e-01 -1.68443561e-01 1.12389408e-01] | [12.815994262695312, 7.723491668701172] |
9021bcb9-38a4-4805-994d-e3740ed61f50 | densely-constrained-depth-estimator-for | 2207.10047 | null | https://arxiv.org/abs/2207.10047v3 | https://arxiv.org/pdf/2207.10047v3.pdf | Densely Constrained Depth Estimator for Monocular 3D Object Detection | Estimating accurate 3D locations of objects from monocular images is a challenging problem because of lacking depth. Previous work shows that utilizing the object's keypoint projection constraints to estimate multiple depth candidates boosts the detection performance. However, the existing methods can only utilize vertical edges as projection constraints for depth estimation. So these methods only use a small number of projection constraints and produce insufficient depth candidates, leading to inaccurate depth estimation. In this paper, we propose a method that utilizes dense projection constraints from edges of any direction. In this way, we employ much more projection constraints and produce considerable depth candidates. Besides, we present a graph matching weighting module to merge the depth candidates. The proposed method DCD (Densely Constrained Detector) achieves state-of-the-art performance on the KITTI and WOD benchmarks. Code is released at https://github.com/BraveGroup/DCD. | ['Yuntao Chen', 'Zhaoxiang Zhang', 'JiaWei He', 'Yingyan Li'] | 2022-07-20 | null | null | null | null | ['graph-matching'] | ['graphs'] | [-2.47333631e-01 -2.90509135e-01 -4.62921798e-01 -1.92305773e-01
-4.11588460e-01 -4.55725849e-01 3.57881784e-01 -2.29093418e-01
-3.62155855e-01 2.80480534e-01 1.33313552e-01 -1.88845679e-01
3.75979871e-01 -8.05642307e-01 -4.41016793e-01 -4.71139193e-01
4.03116584e-01 2.67430991e-01 1.05709636e+00 2.24008322e-01
5.56929827e-01 7.03503251e-01 -1.34564412e+00 1.53650537e-01
4.47278380e-01 9.62624729e-01 5.35341740e-01 5.16021252e-01
-1.43390223e-01 3.95998895e-01 -1.14877753e-01 -2.35843778e-01
7.13007569e-01 -2.23648861e-01 -2.62850791e-01 2.15024486e-01
8.07179034e-01 -9.41476882e-01 -7.24405050e-01 1.08711302e+00
7.30383813e-01 -2.16895148e-01 4.26478028e-01 -1.17945766e+00
-2.60262907e-01 8.85462984e-02 -1.17559528e+00 3.14626217e-01
5.34195185e-01 1.48950085e-01 8.25697303e-01 -1.38668048e+00
8.31077754e-01 1.14787912e+00 5.21598518e-01 5.61344683e-01
-8.53040874e-01 -7.52606452e-01 2.95935512e-01 2.85386175e-01
-1.52312422e+00 -4.51935917e-01 9.40705776e-01 -2.96976894e-01
1.04356229e+00 6.01548888e-02 7.83066511e-01 6.94171309e-01
5.39692231e-02 9.57023799e-01 9.29814517e-01 -4.59825426e-01
6.21539541e-02 2.20001377e-02 1.64225191e-01 8.57935369e-01
6.32453799e-01 1.88997731e-01 -5.96716285e-01 3.18440609e-03
9.26366389e-01 2.86678642e-01 -4.92531866e-01 -7.60067344e-01
-1.10689545e+00 6.84247494e-01 5.05792081e-01 -5.40441684e-02
-9.45609957e-02 2.30287522e-01 6.79713562e-02 -1.92634970e-01
2.96492100e-01 2.33710110e-02 -1.82464778e-01 1.22608237e-01
-7.37340510e-01 1.59311846e-01 4.44469690e-01 1.27209830e+00
9.31291223e-01 -2.25932717e-01 1.08030893e-01 5.37558913e-01
3.80442202e-01 5.33756912e-01 1.78294957e-01 -9.63202715e-01
7.27490067e-01 8.23817015e-01 6.99955821e-02 -1.10274363e+00
-5.23982584e-01 -1.24640211e-01 -4.30129170e-01 4.52358693e-01
4.57129329e-01 -1.05890550e-01 -1.04094636e+00 1.22033811e+00
6.78184032e-01 1.08702891e-01 -2.15546399e-01 1.21033633e+00
9.76326704e-01 5.21609247e-01 -5.66333354e-01 -4.74345026e-04
1.06844616e+00 -1.11646390e+00 -5.21111310e-01 -6.73871577e-01
5.94963133e-01 -9.65855122e-01 8.46219301e-01 5.53751707e-01
-1.22698092e+00 -4.22923476e-01 -1.20298254e+00 -3.81643683e-01
-1.01211630e-01 3.20218027e-01 8.69427979e-01 5.73838532e-01
-8.36742282e-01 9.53819677e-02 -8.77401650e-01 -3.43738586e-01
2.10150048e-01 2.87739068e-01 -3.33076596e-01 -5.42049527e-01
-6.15502834e-01 8.19990993e-01 3.32811594e-01 -3.57759073e-02
-5.97465754e-01 -3.90879542e-01 -8.97276580e-01 -1.95616558e-01
5.21049798e-01 -7.54506350e-01 1.13878477e+00 -1.67536229e-01
-1.16178322e+00 9.50412154e-01 -2.78316587e-01 -5.05912378e-02
6.84761643e-01 -2.70135164e-01 1.33991763e-01 3.17615271e-01
-6.46714913e-03 8.07893813e-01 4.67764854e-01 -1.15233207e+00
-9.51144159e-01 -7.06357360e-01 1.14877827e-01 5.49354970e-01
-2.97833711e-01 -1.01983517e-01 -1.22925663e+00 -2.98197031e-01
8.65225792e-01 -9.05057669e-01 -3.09924304e-01 4.58858669e-01
-4.83495504e-01 -1.31016195e-01 9.61730599e-01 -3.11250210e-01
1.21780252e+00 -2.11756420e+00 3.67536247e-02 -1.60253118e-03
4.25500751e-01 6.14527101e-03 2.19252184e-01 2.80637652e-01
4.04758692e-01 -2.07835138e-01 5.94890602e-02 -5.63995540e-01
-1.27986863e-01 -1.14693202e-01 6.61429465e-02 6.23044789e-01
-1.49643406e-01 6.61325216e-01 -7.36063421e-01 -6.15289807e-01
6.45168483e-01 3.52046400e-01 -6.75542474e-01 -1.16574440e-04
6.66476339e-02 -1.56759899e-02 -4.82732713e-01 1.01001930e+00
1.25085258e+00 -1.40541047e-01 -1.13736801e-02 -3.68928939e-01
-3.13241303e-01 2.08845764e-01 -1.47920072e+00 1.81860054e+00
-1.44431785e-01 5.83990455e-01 -5.46319596e-02 -3.54723722e-01
9.06366646e-01 -1.69211388e-01 4.52890605e-01 -5.30904174e-01
2.12271944e-01 2.90886402e-01 -6.49423823e-02 -1.30400777e-01
4.64615613e-01 1.74894676e-01 1.99155301e-01 2.73421884e-01
-2.49778762e-01 -4.02686834e-01 8.18467662e-02 3.00264269e-01
1.15080273e+00 1.18642218e-01 4.31648433e-01 1.22250468e-01
3.03553492e-01 5.96911050e-02 8.71241033e-01 4.41098183e-01
-2.83350050e-01 9.73477602e-01 3.24222803e-01 -3.21812928e-01
-9.95696604e-01 -1.07051754e+00 -2.34866783e-01 3.59081119e-01
7.75626242e-01 -4.16108996e-01 -3.06744784e-01 -6.25367343e-01
2.15653017e-01 3.03202748e-01 -3.05341184e-01 1.07097551e-01
-5.82863569e-01 -5.06781936e-01 8.43367577e-02 7.81427324e-01
5.56523860e-01 -6.11819267e-01 -8.46480429e-01 -9.59466919e-02
-3.26576978e-02 -1.30519342e+00 -6.78635299e-01 4.76468019e-02
-9.45908010e-01 -1.13800299e+00 -8.09531689e-01 -7.56458163e-01
8.34399223e-01 9.88959491e-01 8.61952841e-01 1.37562007e-01
-4.84637976e-01 4.94294651e-02 -2.76173264e-01 -2.53362447e-01
3.64338309e-01 -1.70722574e-01 -4.97787148e-02 -4.51481789e-01
7.39299238e-01 -4.11140084e-01 -1.09665203e+00 5.38115263e-01
-4.50426012e-01 3.95186126e-01 6.99801385e-01 4.70036238e-01
7.25983143e-01 -1.31001696e-01 -8.92269835e-02 -6.42458856e-01
-1.55160418e-02 -1.73279464e-01 -1.06903279e+00 -1.26092255e-01
-5.20598769e-01 -2.17028186e-01 1.62365824e-01 -4.53942388e-01
-9.65144336e-01 5.76336920e-01 9.46238786e-02 -8.04214716e-01
6.21663593e-02 5.30610569e-02 -2.07305431e-01 -4.44700986e-01
4.93792474e-01 9.27207544e-02 -3.91493767e-01 -4.78689700e-01
1.86321616e-01 4.56004590e-01 2.38320097e-01 -9.05604810e-02
8.16875815e-01 8.45562220e-01 1.39584571e-01 -5.69474638e-01
-6.38253689e-01 -9.86965060e-01 -8.05234849e-01 -2.17018619e-01
6.22179985e-01 -9.99181986e-01 -4.33639437e-01 4.25109744e-01
-1.15641558e+00 -1.63533404e-01 2.36154869e-01 8.71119380e-01
-2.66293883e-01 6.91814840e-01 -5.98592401e-01 -7.83946276e-01
-2.53379583e-01 -1.14329064e+00 1.11311018e+00 1.79428101e-01
1.46134973e-01 -5.81850469e-01 -3.26082036e-02 1.28989384e-01
-1.54589573e-02 1.97921261e-01 2.14873031e-01 2.20843956e-01
-1.10912943e+00 -3.76432925e-01 -5.74723125e-01 -3.03911809e-02
-3.73288170e-02 -5.87141402e-02 -9.06485736e-01 -1.94031253e-01
-1.34363109e-02 -2.01601926e-02 1.07799506e+00 5.40448427e-01
9.26190436e-01 2.11471707e-01 -6.87636971e-01 8.80846977e-01
1.64898050e+00 2.19289690e-01 7.53753841e-01 2.81337410e-01
9.65051174e-01 4.03886914e-01 8.96922827e-01 5.52052259e-01
4.10519660e-01 9.04440403e-01 4.33527440e-01 -9.62895155e-02
-4.29293543e-01 -3.12791884e-01 2.16898158e-01 6.34292841e-01
4.06382643e-02 -3.52240533e-01 -1.02758169e+00 6.45839512e-01
-1.84627223e+00 -7.07987368e-01 -5.58171451e-01 2.25283241e+00
4.98569399e-01 4.26698893e-01 -9.63046681e-03 7.11796358e-02
7.37014353e-01 2.69729327e-02 -6.26576245e-01 -5.36355935e-02
-3.13506238e-02 -1.34133175e-01 9.45106447e-01 6.28294826e-01
-8.96863937e-01 9.31904972e-01 6.02823257e+00 6.19621098e-01
-8.96557629e-01 -3.03131831e-03 1.88178957e-01 -2.93419093e-01
-1.88871428e-01 1.47038370e-01 -1.43961120e+00 3.75561893e-01
-4.87371199e-02 3.63082588e-02 -4.28910144e-02 1.06453383e+00
1.15606487e-01 -6.74333096e-01 -1.07805288e+00 1.33017826e+00
2.00785264e-01 -1.08329809e+00 -2.56642312e-01 3.08073759e-01
8.70163560e-01 1.40900716e-01 -6.20349161e-02 -1.80297598e-01
6.80843219e-02 -5.05561411e-01 5.49248993e-01 1.12548120e-01
6.86147869e-01 -5.93476474e-01 6.46520853e-01 5.13691545e-01
-1.35220790e+00 5.81289120e-02 -7.84238935e-01 -2.74570674e-01
3.30292314e-01 9.10375953e-01 -8.75776291e-01 3.14131051e-01
7.08628714e-01 7.11006224e-01 -5.71480274e-01 1.50615132e+00
-3.94415379e-01 9.58946571e-02 -6.56853974e-01 -7.82522187e-03
6.65169535e-03 -7.62223601e-02 4.47970450e-01 1.03997707e+00
5.71539700e-01 3.39071304e-01 3.13336521e-01 6.61395609e-01
-5.60418889e-02 -1.83941331e-02 -8.61221552e-01 5.06163299e-01
6.41314864e-01 1.18040490e+00 -9.40716863e-01 -2.27953374e-01
-8.45905840e-01 1.01799691e+00 2.65860021e-01 3.38852555e-02
-7.59424925e-01 -2.47675836e-01 4.23041999e-01 4.04883474e-01
2.97108918e-01 -4.70341980e-01 -5.46541035e-01 -1.30864048e+00
2.26670265e-01 -3.59029949e-01 3.38694751e-01 -8.96403193e-01
-9.92490530e-01 2.28794277e-01 1.74718313e-02 -1.47843528e+00
7.81771764e-02 -6.78209960e-01 -5.19435227e-01 7.53723145e-01
-1.66001153e+00 -9.06369507e-01 -8.46036673e-01 5.29042065e-01
6.95512295e-01 3.45934063e-01 1.51344121e-01 5.52030027e-01
-3.84682417e-01 4.31577772e-01 -3.14994425e-01 2.00766996e-02
7.19400585e-01 -1.00648832e+00 5.62653422e-01 1.09568131e+00
4.17834930e-02 4.13935393e-01 4.25358415e-01 -7.98104048e-01
-1.51502323e+00 -7.53301024e-01 7.40263224e-01 -5.48898637e-01
2.93912560e-01 -4.38445747e-01 -7.44628847e-01 6.39407337e-01
-6.13467135e-02 3.08655947e-01 9.99216884e-02 -2.27098882e-01
-1.81536779e-01 4.80982922e-02 -1.01647770e+00 6.31561577e-01
1.44571400e+00 -2.72923350e-01 -3.54812115e-01 3.25185835e-01
5.38690984e-01 -7.82244444e-01 -2.45524883e-01 5.16407669e-01
6.10289395e-01 -1.28303790e+00 1.11168325e+00 1.88451365e-01
3.15907001e-01 -6.73900068e-01 -1.61299840e-01 -6.84528172e-01
-2.46136948e-01 -1.87773421e-01 -3.33455473e-01 9.09596682e-01
3.02659571e-01 -5.55460989e-01 1.25619650e+00 5.66290855e-01
-2.14214072e-01 -8.27239096e-01 -7.70781636e-01 -7.30037212e-01
-4.60181117e-01 -3.70313197e-01 1.35927558e-01 8.03360224e-01
3.06291264e-02 2.37230375e-01 -3.41405123e-01 2.80087292e-01
7.56447136e-01 2.64490396e-01 1.09791720e+00 -9.54261482e-01
-3.34911078e-01 -3.16657990e-01 -7.04170585e-01 -1.63143563e+00
-3.57183069e-01 -6.28726423e-01 7.28740692e-02 -1.76697969e+00
4.15533781e-01 -4.92542773e-01 -4.39497754e-02 2.22353712e-01
-1.93617448e-01 6.42950535e-01 1.73850045e-01 1.59160927e-01
-4.58419561e-01 2.95912325e-01 1.40803456e+00 2.58645415e-01
-1.32810444e-01 8.53553880e-03 -3.72488141e-01 1.01837909e+00
9.60115433e-01 -4.43885833e-01 -3.79440606e-01 -6.66392028e-01
1.90697163e-01 2.40826476e-02 2.86269099e-01 -1.09504199e+00
5.27672768e-01 -2.23800108e-01 6.68267965e-01 -1.30342126e+00
7.59803712e-01 -6.93914652e-01 -4.74052392e-02 5.94209492e-01
3.35390776e-01 1.10127270e-01 1.95581988e-01 4.72750783e-01
-1.99709386e-02 -3.57846618e-01 7.86295950e-01 -3.03806424e-01
-1.06932771e+00 4.54321295e-01 1.45227179e-01 -2.19649196e-01
1.26348495e+00 -6.44932449e-01 -3.82929206e-01 -2.13072211e-01
-3.01196843e-01 3.68937075e-01 8.59809101e-01 2.69997537e-01
1.12678337e+00 -1.37530267e+00 -5.69269300e-01 1.09965518e-01
9.44482461e-02 2.33487412e-01 5.89846484e-02 8.61759841e-01
-8.94532681e-01 4.79788959e-01 -7.45414346e-02 -7.78093636e-01
-1.57343173e+00 6.32885158e-01 8.96095708e-02 1.03528030e-01
-8.85597527e-01 1.06034780e+00 6.72027230e-01 -1.27143472e-01
2.55689055e-01 -4.18092400e-01 1.73464924e-01 -1.12547144e-01
5.00208259e-01 4.86164391e-01 -5.76131269e-02 -3.34875554e-01
-5.33738613e-01 1.10039234e+00 -1.03705481e-01 -1.18202507e-01
9.91961002e-01 -3.35122168e-01 3.02670598e-01 5.69625199e-02
1.14340401e+00 3.49643230e-01 -1.43140042e+00 -2.79420048e-01
-2.54507959e-01 -1.19729197e+00 4.34280962e-01 -3.36738825e-01
-1.07709789e+00 1.07977378e+00 5.70622265e-01 -3.12164456e-01
1.08384693e+00 5.85208684e-02 8.69010448e-01 1.69340387e-01
6.40142262e-01 -9.54074681e-01 2.80835837e-01 3.39603245e-01
6.59856379e-01 -1.44102979e+00 4.50151831e-01 -9.64919448e-01
-5.04912913e-01 1.10600603e+00 1.06117380e+00 -1.43252030e-01
5.00857353e-01 4.11054701e-01 -9.86634716e-02 -2.36033201e-01
-5.28870344e-01 -4.41498041e-01 2.50358820e-01 5.33118188e-01
1.75922126e-01 -1.18185498e-01 -4.77320492e-01 -3.33666429e-02
8.55728015e-02 -1.49038225e-01 6.52039170e-01 9.94011641e-01
-7.35085011e-01 -1.03426445e+00 -5.16578078e-01 2.94195235e-01
-2.35925749e-01 -1.28301218e-01 -3.97390842e-01 8.65524054e-01
4.49509136e-02 7.61056006e-01 -8.74415189e-02 -3.79405588e-01
3.67108434e-01 -4.02621865e-01 6.61517739e-01 -9.37660336e-01
1.19906902e-01 3.78114313e-01 -2.11144350e-02 -7.77258277e-01
-2.42124036e-01 -6.66353285e-01 -1.29942083e+00 -4.34965968e-01
-7.54297793e-01 -4.43609446e-01 6.78254306e-01 3.53894055e-01
1.84768662e-01 8.99547189e-02 5.44658661e-01 -1.00360608e+00
-1.98711783e-01 -7.14273632e-01 -4.48483527e-01 8.51322338e-03
1.05262384e-01 -9.13730621e-01 -4.72575665e-01 -3.13737333e-01] | [7.936544418334961, -2.490060567855835] |
46bf72af-c734-47d4-982a-813469ad30af | offline-reinforcement-learning-with-implicit | 2110.06169 | null | https://arxiv.org/abs/2110.06169v1 | https://arxiv.org/pdf/2110.06169v1.pdf | Offline Reinforcement Learning with Implicit Q-Learning | Offline reinforcement learning requires reconciling two conflicting aims: learning a policy that improves over the behavior policy that collected the dataset, while at the same time minimizing the deviation from the behavior policy so as to avoid errors due to distributional shift. This trade-off is critical, because most current offline reinforcement learning methods need to query the value of unseen actions during training to improve the policy, and therefore need to either constrain these actions to be in-distribution, or else regularize their values. We propose an offline RL method that never needs to evaluate actions outside of the dataset, but still enables the learned policy to improve substantially over the best behavior in the data through generalization. The main insight in our work is that, instead of evaluating unseen actions from the latest policy, we can approximate the policy improvement step implicitly by treating the state value function as a random variable, with randomness determined by the action (while still integrating over the dynamics to avoid excessive optimism), and then taking a state conditional upper expectile of this random variable to estimate the value of the best actions in that state. This leverages the generalization capacity of the function approximator to estimate the value of the best available action at a given state without ever directly querying a Q-function with this unseen action. Our algorithm alternates between fitting this upper expectile value function and backing it up into a Q-function. Then, we extract the policy via advantage-weighted behavioral cloning. We dub our method implicit Q-learning (IQL). IQL demonstrates the state-of-the-art performance on D4RL, a standard benchmark for offline reinforcement learning. We also demonstrate that IQL achieves strong performance fine-tuning using online interaction after offline initialization. | ['Sergey Levine', 'Ashvin Nair', 'Ilya Kostrikov'] | 2021-10-12 | null | null | null | null | ['d4rl'] | ['robots'] | [-1.02139607e-01 2.37439856e-01 -5.68358004e-01 -2.00510040e-01
-1.00738406e+00 -1.04507267e+00 2.84367710e-01 1.89058006e-01
-9.29067314e-01 1.05252528e+00 6.87862486e-02 -5.00147879e-01
-1.95771307e-01 -8.18367064e-01 -1.00817466e+00 -9.73903656e-01
-1.84682176e-01 7.37255096e-01 -3.20505574e-02 -1.68185562e-01
2.32076243e-01 3.48684877e-01 -1.27472317e+00 -1.19543895e-01
9.84357536e-01 1.04683936e+00 1.25446133e-02 7.66531885e-01
1.96933597e-01 8.67484152e-01 -7.47698545e-01 -6.90976158e-02
4.59620267e-01 -5.59836984e-01 -6.46635473e-01 -3.61529104e-02
-1.35476864e-03 -8.70741487e-01 -2.67982125e-01 1.15115690e+00
4.50133920e-01 5.69888353e-01 1.83486655e-01 -1.03085613e+00
-4.13786739e-01 7.38278091e-01 -2.41955370e-01 -6.81739151e-02
1.95087984e-01 7.63086140e-01 1.02562332e+00 5.81980683e-02
5.87973952e-01 1.21306467e+00 2.31114015e-01 7.75877833e-01
-1.56486905e+00 -6.16889536e-01 5.41341901e-01 -1.89389497e-01
-9.32918847e-01 -3.34996432e-01 5.30971766e-01 -9.16047767e-02
8.99217725e-01 -1.81853771e-01 8.68649542e-01 9.72961962e-01
1.94003209e-01 7.59724200e-01 1.31974030e+00 -9.11441222e-02
9.74775910e-01 3.25486064e-02 -3.41080725e-01 6.67877793e-01
-1.14328168e-01 6.15779996e-01 -2.81367451e-01 -3.74471873e-01
7.98730195e-01 4.48273942e-02 -1.77620247e-01 -5.68270981e-01
-8.53821337e-01 6.69943392e-01 2.69108742e-01 -1.10335708e-01
-6.69226468e-01 6.35777116e-01 3.57590586e-01 6.79621637e-01
-4.26321588e-02 7.97682047e-01 -8.86659920e-01 -8.31283510e-01
-5.24079919e-01 6.26554310e-01 8.66440594e-01 3.78661245e-01
9.62447405e-01 1.33687288e-01 -4.41891909e-01 4.25489187e-01
-3.43244560e-02 6.11734807e-01 5.26746333e-01 -1.55396724e+00
5.36929905e-01 4.47277039e-01 6.74177885e-01 -3.83989751e-01
1.07146623e-02 -4.34847862e-01 -4.36121263e-02 6.57876313e-01
6.72972977e-01 -5.99942625e-01 -9.74541903e-01 2.35339332e+00
5.61278284e-01 5.59476241e-02 1.09752744e-01 8.30957532e-01
-4.08265322e-01 6.46843731e-01 1.30920738e-01 -4.09015864e-01
6.64714634e-01 -6.77559972e-01 -4.29360360e-01 -1.91837415e-01
8.26636612e-01 3.76466736e-02 1.29927790e+00 4.74484175e-01
-1.17652082e+00 -3.66387367e-01 -9.22179580e-01 4.34192061e-01
6.56637326e-02 -1.38057292e-01 3.81748438e-01 3.63986373e-01
-1.00272059e+00 1.22006214e+00 -1.24699390e+00 1.89436272e-01
3.42644513e-01 6.95236981e-01 -8.67072642e-02 4.17358577e-01
-1.13653910e+00 8.78271043e-01 5.78901470e-01 -2.57748902e-01
-1.56474936e+00 -7.11581767e-01 -4.97810274e-01 1.04552306e-01
1.06394064e+00 -4.51816708e-01 1.74903738e+00 -1.33828056e+00
-2.17546368e+00 2.64379412e-01 8.15798268e-02 -9.22671974e-01
5.63674390e-01 -2.78911144e-01 -1.51666790e-01 -4.67300303e-02
-7.08134994e-02 5.84648013e-01 9.93177295e-01 -9.89538312e-01
-7.83941925e-01 -4.26938027e-01 5.04592180e-01 4.68143791e-01
-1.96623489e-01 -7.34260440e-01 -2.98027515e-01 -2.78163910e-01
-2.59003073e-01 -1.11686993e+00 -3.57804090e-01 -7.32754320e-02
-5.75092062e-03 -2.45961368e-01 5.07872522e-01 -4.79352385e-01
1.37717664e+00 -1.91296577e+00 1.08797260e-01 3.68846536e-01
1.32914791e-02 1.79428324e-01 -2.45165929e-01 3.38616461e-01
1.43548220e-01 -1.18511664e-02 -2.29172781e-01 -8.79772678e-02
1.73895031e-01 5.15786767e-01 -7.33651042e-01 5.76732635e-01
-1.52585953e-01 9.92956877e-01 -1.18398893e+00 -5.84154874e-02
1.72006935e-01 -9.49940458e-02 -1.07853913e+00 4.38626260e-01
-9.35609281e-01 8.03569317e-01 -8.65864098e-01 1.50359079e-01
2.93663085e-01 -5.03159054e-02 5.53294063e-01 2.68795699e-01
-1.05983928e-01 3.89817983e-01 -1.36261666e+00 1.60548723e+00
-5.58952212e-01 -2.04983473e-01 6.30998760e-02 -1.19124281e+00
7.50397623e-01 1.72895975e-02 7.77834535e-01 -1.00148356e+00
9.97513980e-02 1.36362329e-01 1.51536971e-01 -3.46534252e-01
9.78169665e-02 -1.06340006e-01 -2.54572723e-02 7.49012351e-01
2.71647703e-02 -1.53193071e-01 -3.91397020e-03 5.74191893e-03
1.05787051e+00 6.76448882e-01 2.39331201e-01 -6.08024746e-02
3.78798336e-01 -2.01167881e-01 7.22033501e-01 1.04104435e+00
-3.44501436e-01 -2.21479312e-01 9.97811437e-01 -3.07036638e-01
-1.03772795e+00 -1.23332191e+00 2.92119831e-01 1.24214530e+00
-4.27487567e-02 -1.49167001e-01 -7.52000451e-01 -1.23169410e+00
4.87553805e-01 9.89593506e-01 -8.11692417e-01 -4.44380224e-01
-6.62596345e-01 -1.67442515e-01 3.10629219e-01 3.96275818e-01
5.27370512e-01 -1.06592166e+00 -8.99246812e-01 5.39590538e-01
2.84333527e-01 -4.23051149e-01 -6.20785594e-01 4.56999063e-01
-9.44645643e-01 -9.03692186e-01 -3.43523324e-01 -1.48949206e-01
5.22546053e-01 -4.71066982e-01 9.18252110e-01 -6.53024763e-02
2.82286435e-01 5.80278814e-01 1.01108335e-01 2.73562409e-02
-4.97987241e-01 -6.45458326e-02 2.42220417e-01 -6.98917210e-02
3.43239456e-02 -6.32958829e-01 -7.87827253e-01 1.57299101e-01
-7.98299849e-01 -4.07820523e-01 2.80743659e-01 1.01601088e+00
8.17806065e-01 1.98972914e-02 8.46891105e-01 -6.78248405e-01
6.77558780e-01 -4.67868388e-01 -1.27370644e+00 2.46438712e-01
-8.52361441e-01 8.30515683e-01 1.11090910e+00 -7.99596488e-01
-9.62446988e-01 9.32770874e-03 -1.80931106e-01 -6.42857313e-01
1.05808444e-01 1.77989244e-01 -7.27358013e-02 3.81808043e-01
6.14906430e-01 3.12197924e-01 4.28629249e-01 -3.83019239e-01
6.31475806e-01 4.22912568e-01 5.14269114e-01 -1.35160434e+00
4.62010711e-01 3.81599903e-01 -8.55638832e-02 -1.11523733e-01
-1.02531624e+00 -2.69002263e-02 -9.60590616e-02 -9.26937237e-02
4.83248800e-01 -6.27539158e-01 -1.49715126e+00 1.87619820e-01
-4.52135056e-01 -9.81319666e-01 -9.18446302e-01 4.07977283e-01
-9.93715703e-01 -3.00673675e-02 -2.78902292e-01 -1.13083637e+00
-1.75420880e-01 -1.27017021e+00 7.86816478e-01 2.58397430e-01
7.23187923e-02 -9.22510445e-01 4.30377036e-01 -7.79198706e-02
2.49620706e-01 1.13248773e-01 9.23914373e-01 -6.22764349e-01
-4.39238101e-01 8.39055926e-02 4.00898039e-01 5.66184342e-01
1.62500873e-01 -3.23496431e-01 -7.32235730e-01 -7.01092243e-01
1.28250137e-01 -7.57861912e-01 6.42009318e-01 4.31410134e-01
1.44443214e+00 -6.78345859e-01 -5.86232655e-02 6.37761831e-01
1.27293599e+00 5.86671829e-01 3.74490976e-01 4.42846864e-01
1.43286094e-01 9.36308652e-02 7.44734824e-01 8.74972641e-01
2.56117374e-01 4.39901292e-01 5.64353704e-01 5.05851209e-01
4.23670739e-01 -8.28353703e-01 7.54863620e-01 1.48942888e-01
2.16442779e-01 3.60253416e-02 -4.65753913e-01 3.20407540e-01
-2.00687218e+00 -1.02992237e+00 9.71830964e-01 2.75364542e+00
1.38360071e+00 3.83439422e-01 3.83148372e-01 -4.10145968e-01
1.46643668e-01 8.03771056e-03 -1.58584595e+00 -5.92282653e-01
4.03530419e-01 4.74976599e-01 7.61142731e-01 7.89491296e-01
-7.95162499e-01 1.05635536e+00 5.80379295e+00 7.44608998e-01
-1.33395755e+00 -1.70175195e-01 8.30691814e-01 -3.51049006e-01
-3.91256541e-01 3.11386913e-01 -7.57571757e-01 5.46236396e-01
1.07693756e+00 -2.61467576e-01 1.31500864e+00 9.95171607e-01
4.83937889e-01 -2.56672531e-01 -1.25156271e+00 4.99378651e-01
-5.82522213e-01 -1.07618201e+00 -2.96948433e-01 7.28254393e-02
8.37482810e-01 -8.29600394e-02 2.05147296e-01 8.76426816e-01
8.45677495e-01 -9.07923043e-01 7.57252693e-01 5.23555994e-01
6.15146756e-01 -9.85138893e-01 1.94630474e-01 6.66945875e-01
-7.93446839e-01 -5.69815338e-01 -1.89341381e-01 -2.59465307e-01
-2.32461572e-01 8.40971395e-02 -7.23689377e-01 -1.27444476e-01
3.20720673e-01 4.19032246e-01 -1.10423736e-01 6.36915386e-01
-4.56565738e-01 7.74525225e-01 -4.48874176e-01 -1.81788906e-01
4.35270399e-01 -5.51242054e-01 4.32557017e-01 3.89451772e-01
9.22056362e-02 -7.99169466e-02 6.52275264e-01 1.07302022e+00
-1.30732104e-01 -1.59378752e-01 -3.59984428e-01 -4.02505070e-01
5.67350984e-01 8.26090097e-01 -2.88622707e-01 -4.17951912e-01
3.62963416e-02 8.26528132e-01 7.16219723e-01 6.17697775e-01
-9.54613924e-01 -6.22987486e-02 1.07805455e+00 -2.75416464e-01
3.63654643e-01 -2.65759468e-01 1.69160273e-02 -9.76503551e-01
1.03016384e-02 -1.17519104e+00 5.00344217e-01 -2.30451524e-01
-9.45172668e-01 5.26791962e-04 -1.18644826e-01 -1.11341739e+00
-7.53777444e-01 -2.38765493e-01 -2.57475048e-01 8.65128160e-01
-1.28197372e+00 -3.57747972e-01 5.20050585e-01 5.49434841e-01
3.30517501e-01 9.82681811e-02 6.23644233e-01 -2.10377544e-01
-4.29512858e-01 6.44558132e-01 4.78404790e-01 -2.53130615e-01
6.63924098e-01 -1.52416575e+00 1.23758011e-01 2.97382563e-01
-1.56712323e-01 5.91032803e-01 5.78171015e-01 -6.84575915e-01
-1.69635391e+00 -7.63597250e-01 4.69668135e-02 -3.77770424e-01
9.09259617e-01 -1.69060871e-01 -8.49227965e-01 7.28490531e-01
-1.55677706e-01 2.70492770e-02 -3.24954763e-02 -1.29679054e-01
-1.91627264e-01 -4.44586605e-01 -1.22989464e+00 8.84648442e-01
7.89547980e-01 -3.91390294e-01 -4.59906965e-01 5.46411574e-02
9.39801753e-01 -6.28962040e-01 -8.26846898e-01 5.47049828e-02
6.15535915e-01 -6.97349668e-01 7.69124508e-01 -1.18809855e+00
1.64686784e-01 -2.83296257e-01 -7.19571114e-02 -1.66457140e+00
9.27341431e-02 -1.11502326e+00 -6.01080000e-01 7.90986836e-01
3.05890471e-01 -1.00721204e+00 9.42095578e-01 8.24197233e-01
2.27124810e-01 -1.18777204e+00 -9.85756516e-01 -8.98006618e-01
4.88103688e-01 -4.07210559e-01 9.11952436e-01 4.95861381e-01
-6.71722442e-02 -5.67559488e-02 -3.75833392e-01 6.78969994e-02
5.09704590e-01 3.15324038e-01 8.93030286e-01 -4.91806030e-01
-9.72695827e-01 -3.57371598e-01 2.39453301e-01 -1.58980465e+00
4.16539729e-01 -6.33626223e-01 2.09065646e-01 -1.04729784e+00
-5.72695807e-02 -5.69628358e-01 -5.45621455e-01 7.14746594e-01
-1.80163294e-01 -6.42796695e-01 2.62758583e-01 1.72547618e-04
-6.60512924e-01 8.65670323e-01 1.64416206e+00 6.42713457e-02
-7.95311093e-01 2.63401568e-01 -6.47131085e-01 4.57513660e-01
9.00367439e-01 -5.43745100e-01 -7.25580513e-01 -1.61335573e-01
2.18994141e-01 5.70432365e-01 1.50174364e-01 -7.28368819e-01
9.54441950e-02 -7.50865996e-01 3.36601287e-01 -1.85134664e-01
1.29770249e-01 -7.11038530e-01 -2.61793524e-01 7.45485842e-01
-8.20597291e-01 1.25141412e-01 9.32174250e-02 7.54089892e-01
2.52543241e-01 -1.34151608e-01 8.83436084e-01 -1.40857279e-01
-4.50256795e-01 4.60850835e-01 -3.42098027e-01 4.77988034e-01
1.01705027e+00 1.36580378e-01 -2.17774466e-01 -6.28072500e-01
-7.80343354e-01 6.15504265e-01 5.98202586e-01 2.21825868e-01
2.62723237e-01 -1.07673359e+00 -1.69537112e-01 2.70770222e-01
-2.56608039e-01 -1.63015261e-01 8.71930271e-02 5.39866745e-01
4.69541289e-02 1.46086276e-01 -1.29798045e-02 -3.29220146e-01
-4.46161926e-01 7.53964126e-01 7.83023179e-01 -6.87741876e-01
-4.70178932e-01 3.37076336e-01 -9.97341946e-02 -6.12787187e-01
4.27079380e-01 -4.45921719e-01 1.51170656e-01 -1.84738860e-01
3.00081909e-01 3.10736388e-01 -1.91204265e-01 1.05645515e-01
-7.10194930e-02 2.28912786e-01 9.25700143e-02 -6.27423465e-01
1.15710235e+00 4.54836562e-02 2.15400428e-01 3.60070914e-01
1.22669864e+00 -9.44968611e-02 -2.07497358e+00 -1.40892655e-01
1.12072751e-02 -3.45905870e-01 5.65460734e-02 -1.23226643e+00
-9.00570393e-01 4.38674659e-01 6.48779094e-01 7.51649439e-02
1.06259298e+00 -3.36560875e-01 7.13067293e-01 7.87006736e-01
5.71792364e-01 -1.66957819e+00 2.19465569e-01 5.57235837e-01
5.18421292e-01 -1.07846487e+00 -1.29966244e-01 7.71202564e-01
-9.44184422e-01 9.69411075e-01 7.48557150e-01 -4.13588941e-01
4.39051330e-01 3.00657272e-01 -7.24056661e-02 2.56968290e-01
-1.09270978e+00 -2.17407838e-01 -4.05771732e-02 3.39121908e-01
-4.47208844e-02 2.07328528e-01 -1.02674454e-01 2.18279243e-01
1.25597576e-02 1.67023882e-01 1.11201137e-01 1.17189050e+00
-6.13597214e-01 -1.30956995e+00 -1.92312911e-01 5.17028570e-01
-3.89337599e-01 2.78193891e-01 -5.65040223e-02 6.37615442e-01
-1.75164178e-01 7.05485344e-01 1.66192293e-01 -1.32366538e-01
3.06000113e-01 1.05328858e-01 5.04072070e-01 -3.96054953e-01
-6.10901773e-01 1.04540318e-01 -9.94431004e-02 -1.13268888e+00
2.82709599e-01 -5.44155300e-01 -1.60492051e+00 -1.31863952e-01
2.72540823e-02 3.38303924e-01 5.78976691e-01 9.98191774e-01
4.32949305e-01 3.78629446e-01 1.05101681e+00 -4.81706411e-01
-1.68301022e+00 -4.62186366e-01 -3.75756562e-01 4.63743031e-01
7.48499930e-01 -6.25506878e-01 -3.96280348e-01 -3.92536968e-01] | [4.106921672821045, 2.2142112255096436] |
66472b8e-8a1f-4d90-a49f-ec35c5a244ba | lastustaln-at-complex-word-identification-cwi | null | null | https://aclanthology.org/W18-0517 | https://aclanthology.org/W18-0517.pdf | LaSTUS/TALN at Complex Word Identification (CWI) 2018 Shared Task | This paper presents the participation of the LaSTUS/TALN team in the Complex Word Identification (CWI) Shared Task 2018 in the English monolingual track . The purpose of the task was to determine if a word in a given sentence can be judged as complex or not by a certain target audience. For the English track, task organizers provided a training and a development datasets of 27,299 and 3,328 words respectively together with the sentence in which each word occurs. The words were judged as complex or not by 20 human evaluators; ten of whom are natives. We submitted two systems: one system modeled each word to evaluate as a numeric vector populated with a set of lexical, semantic and contextual features while the other system relies on a word embedding representation and a distance metric. We trained two separate classifiers to automatically decide if each word is complex or not. We submitted six runs, two for each of the three subsets of the English monolingual CWI track. | ["Ahmed Abura{'}ed", 'Horacio Saggion'] | 2018-06-01 | null | null | null | ws-2018-6 | ['complex-word-identification'] | ['natural-language-processing'] | [ 4.92943600e-02 -1.51631432e-02 1.10795513e-01 -4.70786631e-01
-8.42610478e-01 -9.95800018e-01 7.68524945e-01 7.21076310e-01
-1.23469460e+00 5.81765413e-01 4.53898787e-01 -6.82923138e-01
2.26226106e-01 -3.43705922e-01 -2.94971019e-01 -2.17402637e-01
2.00005785e-01 7.62172103e-01 -9.36635137e-02 -4.05224830e-01
3.31379473e-01 -2.93110441e-02 -1.37519073e+00 2.38119140e-01
8.95975828e-01 9.20307577e-01 4.67498094e-01 8.14888954e-01
-6.24054559e-02 4.37337846e-01 -7.31598616e-01 -5.41645110e-01
1.57978654e-01 -1.76695913e-01 -9.61829543e-01 -3.00966829e-01
6.25051558e-01 3.38274270e-01 1.05978139e-01 9.47594047e-01
6.61240876e-01 9.14308354e-02 9.28189218e-01 -8.72562468e-01
-6.22172177e-01 1.13254285e+00 9.31780785e-02 5.55156291e-01
6.16315842e-01 7.02511370e-02 1.49567270e+00 -1.55078161e+00
9.26187992e-01 1.32529414e+00 4.45939958e-01 3.95331413e-01
-1.28205740e+00 -8.19356382e-01 1.17553219e-01 4.21643645e-01
-1.43955827e+00 -7.31839240e-01 2.52897978e-01 -7.83095479e-01
1.54572153e+00 2.57179774e-02 8.06902170e-01 1.05281115e+00
1.63223725e-02 6.68487251e-01 1.42156637e+00 -8.45358551e-01
1.65350035e-01 4.13803190e-01 8.69589686e-01 3.16650212e-01
1.09894961e-01 5.58484085e-02 -6.27751529e-01 1.30374730e-01
-3.69050324e-01 -7.72574961e-01 -3.26644748e-01 2.40698665e-01
-1.52671289e+00 1.11415493e+00 -1.77381802e-02 5.98895013e-01
-8.24567154e-02 -3.55098367e-01 7.24935412e-01 8.27321947e-01
6.27380192e-01 7.37469673e-01 -6.45261109e-01 -2.25579515e-01
-7.95129299e-01 2.86305070e-01 8.28594208e-01 7.81849325e-01
6.02504194e-01 -2.32639775e-01 -5.00509679e-01 1.03188252e+00
3.54081452e-01 7.24515975e-01 9.58533645e-01 -2.23305851e-01
6.85678065e-01 5.48347771e-01 -4.37121131e-02 -8.10559571e-01
-4.79471385e-01 -2.98208356e-01 -2.80562460e-01 3.68780866e-02
4.15982813e-01 -2.31535465e-01 -7.68138587e-01 1.84007454e+00
4.81341369e-02 -6.48553550e-01 2.47607201e-01 6.27289891e-01
1.12127185e+00 8.46436501e-01 3.03986281e-01 -4.85575236e-02
1.67095578e+00 -9.03489709e-01 -7.58183181e-01 -3.93396765e-01
9.59764659e-01 -1.29993200e+00 1.38682413e+00 3.55269849e-01
-1.07067204e+00 -7.66533852e-01 -1.48160100e+00 -2.15526178e-01
-9.03051555e-01 3.53439212e-01 2.54505053e-02 6.99459970e-01
-1.07609403e+00 1.09091625e-01 -2.16246247e-01 -3.61724734e-01
-2.58292198e-01 -1.59821995e-02 -6.60304129e-01 -1.05421834e-01
-1.70505786e+00 1.56942177e+00 4.90224153e-01 -1.16087317e-01
-7.36916661e-01 -6.41379058e-01 -1.25906003e+00 -1.83834597e-01
-2.64342129e-01 -1.84507206e-01 1.15555382e+00 -9.07595634e-01
-1.05735898e+00 1.83184564e+00 -8.16901922e-02 -3.10490191e-01
3.77504289e-01 -1.92664593e-01 -7.37307608e-01 -3.86058360e-01
6.58663273e-01 4.28982466e-01 7.85428584e-01 -7.39392638e-01
-8.89338434e-01 -5.35325229e-01 3.24108149e-03 3.68877590e-01
-4.34664100e-01 5.01948953e-01 -1.67792618e-01 -7.36826658e-01
-2.38156274e-01 -1.04840279e+00 3.21573079e-01 -5.78229666e-01
-7.22185016e-01 -9.18976605e-01 1.71334788e-01 -9.53645945e-01
1.49117351e+00 -2.05715275e+00 4.11654711e-01 6.88141659e-02
1.85910717e-01 4.05264020e-01 -3.54882419e-01 6.42393649e-01
-2.86278456e-01 3.64132673e-01 4.37190272e-02 -7.19161153e-01
1.37121573e-01 -1.05196103e-01 -1.07040308e-01 4.22924846e-01
2.57067293e-01 5.87769091e-01 -9.69035208e-01 -3.43568146e-01
-1.17013514e-01 2.14524955e-01 -1.71487406e-01 2.39426896e-01
1.57173172e-01 1.31703675e-01 2.13964626e-01 1.94680110e-01
4.44198430e-01 4.43832964e-01 1.42490610e-01 -1.82255313e-01
-4.67231870e-01 7.74224520e-01 -1.08574164e+00 1.29306972e+00
-8.93948555e-01 8.75879645e-01 -6.77024350e-02 -5.93400240e-01
8.55133355e-01 4.36129600e-01 -4.67382669e-01 -7.03302741e-01
1.37529328e-01 5.78626275e-01 4.07553047e-01 -2.88309246e-01
6.59567714e-01 -1.05651736e-01 -5.96239805e-01 5.40342927e-01
3.25573951e-01 -1.32837012e-01 8.00388217e-01 1.86386243e-01
1.11514330e+00 -2.64702976e-01 4.28648263e-01 -9.21671808e-01
8.14274490e-01 3.46524105e-03 4.46413845e-01 5.95813751e-01
-3.05823386e-01 3.76742184e-01 3.91229391e-01 -5.23537815e-01
-1.11323404e+00 -1.07351780e+00 -3.65819514e-01 1.51857984e+00
-1.60769746e-01 -6.71380579e-01 -5.95915496e-01 -4.12900537e-01
-6.39799982e-02 1.02423882e+00 -6.54737055e-01 -2.68940747e-01
-6.59342825e-01 -1.28116861e-01 5.83167732e-01 2.35030472e-01
-1.84323192e-01 -1.40302098e+00 -3.39424789e-01 1.25227317e-01
-4.34118897e-01 -9.25356328e-01 -9.83965576e-01 4.16490972e-01
2.40606934e-01 -9.98809993e-01 -5.98387957e-01 -1.31429183e+00
2.66398579e-01 -3.89000028e-02 1.51980448e+00 5.16862832e-02
-3.94082427e-01 -3.43168005e-02 -6.58385277e-01 -5.32485068e-01
-6.69242442e-01 4.26333964e-01 4.88257736e-01 -2.52858102e-02
6.27549350e-01 1.99085146e-01 -2.92829692e-01 -7.58916512e-02
-6.16356313e-01 -1.27388090e-01 2.17988849e-01 8.18279207e-01
3.45509559e-01 -3.69865507e-01 2.40774080e-01 -9.29395199e-01
1.08429611e+00 -5.08218050e-01 -2.61154801e-01 3.64910364e-01
-5.87658107e-01 2.42969841e-02 4.69920963e-01 -5.72209060e-01
-5.34284532e-01 -8.41051340e-03 -4.06182498e-01 3.38666856e-01
3.19572501e-02 8.39320898e-01 -3.15283805e-01 4.57489312e-01
7.64427841e-01 2.01911926e-02 -3.71800154e-01 -4.77966666e-01
4.89273936e-01 8.96525264e-01 3.69961262e-01 -6.30071223e-01
7.29402781e-01 -5.37823677e-01 -7.81968296e-01 -9.63561237e-01
-1.12127614e+00 -7.67829835e-01 -7.30383098e-01 -1.78785980e-01
1.07960117e+00 -1.03790867e+00 -3.72208416e-01 6.52218401e-01
-1.44966912e+00 -2.06973180e-01 -1.21672209e-02 4.64094937e-01
-1.42592415e-01 1.07061930e-01 -4.80627775e-01 -4.53549892e-01
-6.82268739e-01 -1.23946059e+00 9.18600380e-01 -1.73613966e-01
-9.16540444e-01 -9.70773280e-01 4.41506058e-01 2.62055248e-01
5.19194484e-01 -1.23244174e-01 1.23782730e+00 -9.73500073e-01
5.20381570e-01 -3.32692802e-01 -4.79032584e-02 5.79628050e-01
-9.22206715e-02 5.31944484e-02 -7.49267280e-01 -4.70728844e-01
-1.38800949e-01 -8.22593212e-01 7.37600386e-01 -3.17352340e-02
3.32621843e-01 -2.58354038e-01 3.20166200e-02 2.30520919e-01
1.19785070e+00 -3.35667059e-02 1.46077320e-01 2.82119125e-01
5.76498687e-01 6.41754687e-01 4.78140652e-01 1.02378197e-01
6.98089242e-01 7.66206324e-01 -1.51576981e-01 1.99863493e-01
-2.61534661e-01 -2.13768080e-01 7.59298861e-01 1.39084589e+00
3.05236131e-01 -3.36720437e-01 -1.36869025e+00 8.28990936e-01
-1.19615614e+00 -5.24509251e-01 -3.50584924e-01 2.14217448e+00
1.23681831e+00 2.62364566e-01 4.60290089e-02 5.36404192e-01
8.13793421e-01 9.12497640e-02 -6.27844930e-02 -8.11329782e-01
-5.01615822e-01 5.25878310e-01 2.13385046e-01 9.35923159e-01
-1.15168595e+00 1.38209653e+00 5.68904066e+00 9.34010923e-01
-1.02860689e+00 2.43008107e-01 5.09510458e-01 2.98316061e-01
-3.40287685e-01 -1.53062031e-01 -1.24253917e+00 3.94368708e-01
1.31036580e+00 -4.45111781e-01 2.82244772e-01 5.74627578e-01
-1.49562716e-01 -1.52625218e-01 -1.24226880e+00 8.47371161e-01
3.25613141e-01 -8.27872336e-01 6.71154931e-02 -3.26348871e-01
4.50196862e-01 3.92596632e-01 -4.57211174e-02 7.54802406e-01
3.43084186e-01 -1.26958358e+00 1.27304649e+00 3.92410755e-01
1.11216962e+00 -7.15306759e-01 7.34184384e-01 4.41625983e-01
-1.21468282e+00 3.36494476e-01 -1.01113267e-01 -2.87161142e-01
-1.14248499e-01 2.89681405e-01 -5.73028505e-01 -5.38246185e-02
4.82514977e-01 5.56038976e-01 -1.07984114e+00 6.51696920e-01
-6.17140710e-01 8.39758575e-01 -1.32110730e-01 -4.34617251e-01
2.06865877e-01 1.21518277e-01 7.16994882e-01 1.67764485e+00
7.34237134e-02 -4.83824819e-01 3.95844072e-01 4.32574123e-01
4.97770980e-02 7.73794591e-01 -5.36408544e-01 -2.36250222e-01
8.05284381e-01 1.50216162e+00 -4.71873492e-01 -3.93319637e-01
-2.07818940e-01 1.02104318e+00 8.71904850e-01 6.54955283e-02
-1.01946108e-01 -8.07587683e-01 6.04309738e-01 -1.53312057e-01
1.38535351e-01 -2.90178448e-01 -2.71954328e-01 -1.19666147e+00
1.46488369e-01 -8.23742688e-01 3.80753845e-01 -5.54346025e-01
-1.59797847e+00 1.26848447e+00 -4.11397994e-01 -8.71567726e-01
-3.32108438e-01 -1.08657205e+00 -5.31781554e-01 1.38602662e+00
-1.39143562e+00 -1.06110513e+00 -1.40732024e-02 4.99860644e-01
5.75801194e-01 -5.13158560e-01 1.21134114e+00 2.74729580e-01
-4.36742336e-01 7.84074426e-01 -1.68250248e-01 3.93423021e-01
9.23554242e-01 -1.58753026e+00 6.29164696e-01 5.66796958e-01
3.31679225e-01 7.37257183e-01 8.63140285e-01 -4.46134448e-01
-9.78826940e-01 -9.54074442e-01 2.11714029e+00 -6.32388413e-01
1.05923653e+00 -8.15229535e-01 -5.81973493e-01 4.35377568e-01
5.64238131e-01 -1.48093373e-01 7.92166471e-01 3.42906505e-01
-4.57878381e-01 3.08660597e-01 -8.03478599e-01 4.57963198e-01
7.64909446e-01 -8.45533431e-01 -1.03857648e+00 5.00398099e-01
7.80972242e-01 -1.29663169e-01 -9.12025273e-01 -3.83987091e-02
3.81398708e-01 -4.76678610e-01 5.43974578e-01 -8.15113068e-01
5.00645518e-01 -2.77771860e-01 -4.09170866e-01 -2.01582050e+00
-3.52005154e-01 -2.03838021e-01 3.86263996e-01 1.28359425e+00
9.81804192e-01 -3.36279541e-01 -9.40157995e-02 5.92636056e-02
-2.55268961e-01 -4.75151420e-01 -1.11925912e+00 -8.06802511e-01
5.41064799e-01 -6.31434619e-01 1.50181472e-01 9.37221646e-01
1.94646612e-01 1.19847238e+00 9.65519920e-02 -3.55086923e-01
1.80675715e-01 -3.52826893e-01 2.86743671e-01 -1.35544145e+00
5.81685863e-02 -6.60101354e-01 -6.01404309e-01 -3.92185807e-01
6.21299148e-01 -1.55920982e+00 1.89151123e-01 -1.37475634e+00
8.74708667e-02 -2.03957304e-01 -1.96980253e-01 4.68742549e-01
-4.22755361e-01 3.34457785e-01 3.26635450e-01 3.65591198e-02
-4.01757181e-01 5.03417730e-01 6.42767012e-01 -3.43444824e-01
1.92001343e-01 -1.22975111e-01 -4.83884364e-01 4.65242893e-01
6.71770453e-01 -2.75930405e-01 -2.01337375e-02 -6.84866786e-01
7.24365413e-01 -4.25689399e-01 -8.95710960e-02 -9.24374938e-01
4.11188379e-02 4.02790993e-01 1.97707176e-01 -3.99552703e-01
5.81397824e-02 -4.42280263e-01 -2.80879617e-01 4.43065614e-01
-4.96600866e-01 5.93683541e-01 1.79005310e-01 -1.62575975e-01
-3.83487672e-01 -5.55947185e-01 9.14410770e-01 1.54759372e-02
-6.50467157e-01 1.23505063e-01 -8.12920570e-01 6.78064108e-01
8.66958797e-01 2.05302358e-01 -4.43899781e-02 -9.84401405e-02
-8.08364809e-01 2.25798130e-01 1.81143805e-01 8.55758309e-01
4.86851990e-01 -1.53282964e+00 -1.26013458e+00 2.31770203e-01
7.87711918e-01 -6.98435068e-01 -2.28506178e-01 4.34292257e-01
-5.33323169e-01 4.93561298e-01 -4.38331775e-02 -1.67561173e-01
-1.23892367e+00 3.89050007e-01 8.69456232e-02 -4.11221623e-01
-2.32909679e-01 1.07168543e+00 -2.90477902e-01 -9.60693300e-01
2.49856323e-01 -1.96621805e-01 -6.62272811e-01 6.86276019e-01
8.18394482e-01 2.00439394e-01 3.60373944e-01 -1.38528848e+00
-4.70294237e-01 5.88251173e-01 -3.38251442e-01 -5.90547204e-01
1.00407469e+00 -6.90869391e-02 -3.22110265e-01 1.11797285e+00
1.50743592e+00 1.72464669e-01 -2.29647934e-01 -5.51049769e-01
4.51541871e-01 8.05637538e-02 1.27163202e-01 -1.03881598e+00
-4.98798698e-01 8.44193637e-01 6.10238075e-01 2.56619692e-01
5.04734218e-01 3.27159800e-02 7.09579706e-01 3.73995274e-01
3.01217884e-02 -1.65531242e+00 -3.53333056e-01 1.32847917e+00
1.20759058e+00 -1.40759432e+00 -5.53972483e-01 -1.01258233e-01
-7.66572416e-01 1.04430902e+00 3.53214860e-01 -1.09503828e-01
7.84951448e-01 9.97800380e-03 4.40009713e-01 -1.75280362e-01
-8.31879377e-01 -3.44417989e-01 7.76581824e-01 5.24121225e-01
9.50371265e-01 5.52684009e-01 -9.75804687e-01 8.07063401e-01
-9.02685881e-01 -7.50041068e-01 2.72105008e-01 5.26371896e-01
-4.10942197e-01 -1.06766307e+00 1.56078208e-02 4.56269950e-01
-3.00838649e-01 -6.30922437e-01 -4.30929929e-01 5.52379608e-01
2.28394255e-01 1.26651073e+00 1.40161172e-01 -3.98230702e-01
5.24155617e-01 2.45438546e-01 2.05584243e-01 -1.02835226e+00
-8.91173780e-01 -4.87501770e-01 4.93443251e-01 -2.33405605e-01
1.63317159e-01 -8.88278842e-01 -1.00847149e+00 -4.24349643e-02
3.62844318e-02 2.54748762e-01 9.35198605e-01 1.17719388e+00
-2.95705020e-01 1.42238900e-01 5.90333283e-01 -7.03861475e-01
-5.91222048e-01 -1.62018871e+00 -5.35478055e-01 6.28443003e-01
2.82914251e-01 -4.65733916e-01 -5.52639425e-01 8.31276830e-03] | [10.527769088745117, 10.440348625183105] |
3b82a0e6-2295-4813-b76d-ac591e7780af | explainability-by-parsing-neural-module-tree | 1812.03299 | null | https://arxiv.org/abs/1812.03299v3 | https://arxiv.org/pdf/1812.03299v3.pdf | Learning to Assemble Neural Module Tree Networks for Visual Grounding | Visual grounding, a task to ground (i.e., localize) natural language in images, essentially requires composite visual reasoning. However, existing methods over-simplify the composite nature of language into a monolithic sentence embedding or a coarse composition of subject-predicate-object triplet. In this paper, we propose to ground natural language in an intuitive, explainable, and composite fashion as it should be. In particular, we develop a novel modular network called Neural Module Tree network (NMTree) that regularizes the visual grounding along the dependency parsing tree of the sentence, where each node is a neural module that calculates visual attention according to its linguistic feature, and the grounding score is accumulated in a bottom-up direction where as needed. NMTree disentangles the visual grounding from the composite reasoning, allowing the former to only focus on primitive and easy-to-generalize patterns. To reduce the impact of parsing errors, we train the modules and their assembly end-to-end by using the Gumbel-Softmax approximation and its straight-through gradient estimator, accounting for the discrete nature of module assembly. Overall, the proposed NMTree consistently outperforms the state-of-the-arts on several benchmarks. Qualitative results show explainable grounding score calculation in great detail. | ['Zheng-Jun Zha', 'Feng Wu', 'Daqing Liu', 'Hanwang Zhang'] | 2018-12-08 | learning-to-assemble-neural-module-tree | http://openaccess.thecvf.com/content_ICCV_2019/html/Liu_Learning_to_Assemble_Neural_Module_Tree_Networks_for_Visual_Grounding_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Liu_Learning_to_Assemble_Neural_Module_Tree_Networks_for_Visual_Grounding_ICCV_2019_paper.pdf | iccv-2019-10 | ['natural-language-visual-grounding'] | ['reasoning'] | [-9.62772465e-04 6.92402542e-01 -7.05057979e-02 -5.01583695e-01
-6.22862041e-01 -4.68051523e-01 5.19403517e-01 1.58396974e-01
-1.20380215e-01 3.31195682e-01 3.93873364e-01 -5.13138950e-01
3.01773220e-01 -8.05897534e-01 -1.01229537e+00 -4.77196604e-01
1.44205302e-01 1.32212117e-01 -7.15115666e-02 1.79947615e-02
2.52846591e-02 -7.60280155e-03 -1.16932619e+00 7.40272701e-01
9.09694791e-01 8.60650897e-01 6.20921969e-01 5.22056520e-01
-4.40815598e-01 1.17898715e+00 -4.86561328e-01 -8.09525490e-01
4.46317643e-02 -7.12385476e-01 -7.50645518e-01 2.56162316e-01
8.51995349e-01 -1.24355346e-01 -1.68432668e-01 1.23001099e+00
6.83283880e-02 -1.05054952e-01 4.67989236e-01 -1.17560744e+00
-1.33201098e+00 7.44133770e-01 -5.81234276e-01 -4.52518389e-02
1.56668007e-01 3.08767378e-01 1.48831344e+00 -1.18307102e+00
6.77352190e-01 1.51657867e+00 3.16085756e-01 5.62457144e-01
-1.53516030e+00 -3.10653806e-01 7.00080037e-01 -1.21368067e-02
-1.13915360e+00 -1.10104434e-01 7.57840514e-01 -7.44223893e-01
1.09786499e+00 5.98771386e-02 6.11661494e-01 8.71507525e-01
5.40234983e-01 8.20531368e-01 1.03025925e+00 -3.55034351e-01
3.69569629e-01 2.32272185e-02 2.55185932e-01 1.33103526e+00
3.75407726e-01 -2.77939320e-01 -7.31060326e-01 2.22536922e-01
5.92483282e-01 9.80182886e-02 -2.89637484e-02 -6.93980992e-01
-1.37799144e+00 8.16544175e-01 1.16068578e+00 1.77643858e-02
-5.29240370e-01 3.78411084e-01 1.57621831e-01 3.00027058e-02
4.55047667e-01 2.66249686e-01 -1.20351672e-01 4.85939234e-01
-1.00514519e+00 1.05235048e-01 3.68653804e-01 8.82412136e-01
1.00237525e+00 6.50119921e-03 -5.58027685e-01 4.61287558e-01
5.84983170e-01 2.12148309e-01 2.27110744e-01 -6.30708039e-01
9.00153100e-01 1.08255363e+00 -2.11571097e-01 -1.04128838e+00
-5.66461802e-01 -6.69215500e-01 -8.96574557e-01 5.03102779e-01
2.99651742e-01 -2.60692853e-02 -1.06456006e+00 2.08664012e+00
3.03894371e-01 -1.24192595e-01 1.63014472e-01 1.02710891e+00
1.11066628e+00 6.41183257e-01 5.78281343e-01 1.93087026e-01
1.65100098e+00 -1.28090012e+00 -6.12716675e-01 -8.10218751e-01
5.04598141e-01 -3.02653551e-01 1.57702303e+00 1.23994753e-01
-1.13561082e+00 -8.23876441e-01 -1.08295798e+00 -5.21839976e-01
-3.26130867e-01 4.23717529e-01 6.43064082e-01 2.37079218e-01
-1.26569259e+00 3.28090131e-01 -8.82812083e-01 -1.89952746e-01
6.48084223e-01 -7.65775666e-02 -5.75593352e-01 1.38752058e-01
-8.90602589e-01 8.35391879e-01 4.72382396e-01 2.67231435e-01
-7.85923719e-01 -5.62346041e-01 -1.34223747e+00 3.13883096e-01
3.38197172e-01 -1.40361083e+00 9.28475201e-01 -1.00737560e+00
-1.09020483e+00 1.15601194e+00 -2.39942864e-01 -5.11884272e-01
1.51035786e-01 -1.41515285e-01 9.27942991e-02 2.35752881e-01
3.02327961e-01 1.08551586e+00 7.93684959e-01 -1.45270526e+00
-4.83441740e-01 -2.90035874e-01 5.29906750e-01 2.38402709e-01
-1.86067045e-01 -3.29907149e-01 -5.03222764e-01 -7.08220184e-01
4.08307940e-01 -6.79581523e-01 -1.43434748e-01 1.58366799e-01
-7.59684563e-01 -1.41235664e-01 1.29069343e-01 -9.68940914e-01
1.25453651e+00 -2.29933214e+00 5.86602986e-01 -1.72232032e-01
6.31732404e-01 -3.84863943e-01 -4.22056429e-02 4.09655362e-01
-1.41955107e-01 5.25162332e-02 -3.19186509e-01 -7.11622655e-01
3.30240667e-01 1.27013177e-01 -2.85512805e-01 3.94481927e-01
5.99171877e-01 1.17586112e+00 -1.03172731e+00 -6.33062363e-01
2.69054081e-02 4.81626600e-01 -7.06206560e-01 2.33675003e-01
-4.08881038e-01 1.42534643e-01 -1.01118691e-01 6.18345797e-01
4.98633415e-01 -6.08099222e-01 2.02837378e-01 -4.37982529e-01
-1.49393883e-02 3.16461861e-01 -9.05397832e-01 2.03093147e+00
-6.61891460e-01 7.47157812e-01 -2.90349107e-02 -8.46955776e-01
8.84082615e-01 -1.05246037e-01 -1.40776724e-01 -7.89300740e-01
-1.53870538e-01 -1.75452918e-01 -7.16913044e-02 -3.78586441e-01
3.87278855e-01 -1.81816831e-01 -2.26941586e-01 1.98710635e-01
2.55575985e-01 2.39545628e-01 1.57810092e-01 5.29810429e-01
8.32832158e-01 3.02630931e-01 4.14745271e-01 -2.81007886e-01
2.19864413e-01 -5.26608825e-02 2.60892540e-01 6.21725380e-01
5.50938174e-02 5.22533834e-01 8.07001829e-01 -3.86290401e-01
-7.67625630e-01 -1.37900615e+00 2.78035343e-01 1.34079218e+00
3.16347986e-01 -3.72274280e-01 -1.06209326e+00 -7.80844390e-01
2.24364433e-03 8.66767764e-01 -8.46066654e-01 -2.72197098e-01
-4.82216865e-01 -3.26462716e-01 5.32191470e-02 7.38021612e-01
5.06772578e-01 -1.34711754e+00 -8.53558004e-01 3.67454812e-02
-1.45890579e-01 -1.15204906e+00 -4.67479378e-01 3.27343464e-01
-6.96549177e-01 -8.84322405e-01 -3.67625445e-01 -9.46312428e-01
1.01307106e+00 2.92701870e-01 1.47964370e+00 1.79096386e-01
-1.94447804e-02 1.93992570e-01 -1.16933867e-01 -6.29228801e-02
-2.03898549e-01 5.98003790e-02 -4.83147025e-01 2.61054754e-01
1.09912768e-01 -2.25203916e-01 -6.34382546e-01 -2.62912124e-01
-7.83417463e-01 8.53382647e-01 7.81242609e-01 9.28277075e-01
8.10108602e-01 -3.21980000e-01 1.35937229e-01 -8.46352160e-01
5.42762399e-01 -4.32193935e-01 -4.93557006e-01 4.62736130e-01
-3.87759894e-01 3.79953086e-01 5.17488301e-01 -3.78709808e-02
-9.25503671e-01 1.22588180e-01 4.02014330e-02 -2.47240022e-01
3.55465896e-02 6.49295032e-01 -4.27787662e-01 2.04351708e-01
5.13533592e-01 1.74485549e-01 -3.08886737e-01 -3.26205224e-01
8.71671557e-01 1.66789889e-01 7.78389513e-01 -5.84721923e-01
9.00786817e-01 5.17056584e-01 5.45324795e-02 -5.63245475e-01
-1.16461253e+00 -6.21486530e-02 -5.49733043e-01 -2.54576892e-01
1.39330041e+00 -1.05771601e+00 -6.11299694e-01 -7.26724863e-02
-1.46717167e+00 -5.58009624e-01 -2.36333892e-01 5.05296551e-02
-3.74199718e-01 1.50832325e-01 -5.53233624e-01 -6.53347075e-01
-4.27572757e-01 -1.01201832e+00 1.38789117e+00 1.85969666e-01
-2.17165768e-01 -8.37400019e-01 -1.35723874e-01 2.57309020e-01
8.46866220e-02 5.08320630e-01 1.10698557e+00 -2.39877805e-01
-8.30324948e-01 1.53712258e-01 -6.46403134e-01 1.90045208e-01
-1.82494298e-01 -1.66666687e-01 -1.01458216e+00 -2.07674816e-01
-2.03000262e-01 -1.59551933e-01 1.19756067e+00 3.52317929e-01
1.01038742e+00 -5.08882165e-01 -5.88411391e-02 8.18702161e-01
1.55948973e+00 -1.37290776e-01 4.80597854e-01 4.52434927e-01
1.08884525e+00 7.12240100e-01 2.27300331e-01 5.86704686e-02
8.44268620e-01 5.94089866e-01 8.78956676e-01 -6.93422079e-01
-4.31042224e-01 -6.29925787e-01 5.04355192e-01 4.71033633e-01
2.82976151e-01 -1.08432271e-01 -8.17251086e-01 3.58161718e-01
-1.86290157e+00 -9.53659415e-01 -3.96985933e-02 1.94733369e+00
7.04243064e-01 3.89212936e-01 -7.09739700e-02 -3.35507929e-01
7.47822762e-01 3.15638721e-01 -3.73314679e-01 -5.36728978e-01
-2.17736751e-01 2.85422485e-02 7.98041746e-02 7.41890073e-01
-1.11928153e+00 1.16303408e+00 5.16615391e+00 3.62150699e-01
-1.08792353e+00 2.38531739e-01 7.82854378e-01 -6.08789921e-02
-5.07251740e-01 7.94457123e-02 -7.09729791e-01 3.62461507e-01
4.24578518e-01 1.68891147e-01 4.89544749e-01 9.88069236e-01
5.49983308e-02 6.70734569e-02 -1.29058266e+00 9.19241667e-01
3.89886610e-02 -1.53322148e+00 2.30251148e-01 -4.77037318e-02
5.80181777e-01 -1.51228577e-01 1.09509669e-01 3.74224573e-01
1.57054156e-01 -9.54602480e-01 1.57187784e+00 5.52112818e-01
6.38090670e-01 -3.80984664e-01 3.81954461e-01 2.23489881e-01
-1.49010479e+00 -4.61688377e-02 -2.47491524e-01 -1.78596705e-01
1.14897311e-01 5.40760040e-01 -4.05357867e-01 4.20277536e-01
5.91543436e-01 4.11082804e-01 -9.09833372e-01 6.77422225e-01
-7.90911973e-01 3.48229378e-01 2.60176480e-01 -1.51833072e-01
4.98903275e-01 -7.84568489e-02 3.05892348e-01 1.46434200e+00
9.14580375e-02 -1.93787634e-01 3.60910743e-01 1.31229174e+00
-1.51228160e-01 -2.91508846e-02 -4.14134771e-01 8.03888738e-02
2.22657710e-01 1.45156133e+00 -9.10251319e-01 -4.49485093e-01
-5.80430686e-01 1.15332711e+00 8.74096215e-01 6.28628671e-01
-8.72158587e-01 -2.30762780e-01 4.87345219e-01 1.10636719e-01
4.22143847e-01 -2.57463455e-01 -6.34444714e-01 -1.27436888e+00
4.17255074e-01 -6.45407438e-01 4.18898970e-01 -1.14727807e+00
-1.17077911e+00 8.76989603e-01 -8.82248357e-02 -9.01328504e-01
-6.62639886e-02 -8.48974168e-01 -5.92699111e-01 1.06399786e+00
-1.43473589e+00 -1.49306667e+00 -4.89133656e-01 3.29515636e-01
5.91542900e-01 1.40863493e-01 5.46085358e-01 -3.61023396e-02
-6.51348174e-01 5.66524148e-01 -5.53634286e-01 1.65255859e-01
2.54071027e-01 -1.56532717e+00 7.36730278e-01 1.19483972e+00
5.58751822e-01 9.35774148e-01 7.03205228e-01 -5.63934445e-01
-1.25502467e+00 -1.22887242e+00 1.20025706e+00 -4.40521717e-01
7.08239377e-01 -8.53218138e-01 -9.99344230e-01 6.31293952e-01
5.67200601e-01 -6.77908654e-04 3.62248689e-01 1.81497887e-01
-7.41990268e-01 -1.67040035e-01 -8.25774431e-01 8.00191641e-01
1.00700366e+00 -8.36838841e-01 -7.45303690e-01 2.25825310e-01
1.00949335e+00 -4.68392193e-01 -2.84868240e-01 -1.02006152e-01
3.51801932e-01 -1.07231688e+00 9.42340553e-01 -6.66149914e-01
8.37039888e-01 -5.77053249e-01 -2.12860048e-01 -1.17208576e+00
-6.74571037e-01 -3.83967757e-01 -2.23885015e-01 1.14036453e+00
5.97665131e-01 -3.54126155e-01 7.20758915e-01 3.98739874e-01
-4.36911404e-01 -1.04135478e+00 -9.40425158e-01 -5.49476266e-01
-8.01722258e-02 -2.49369994e-01 5.35616219e-01 7.36428320e-01
1.56970605e-01 6.53572500e-01 -1.37613788e-01 4.36797410e-01
7.76586950e-01 3.99899691e-01 6.70345187e-01 -9.60509479e-01
-5.83657861e-01 -5.97888947e-01 -3.86347234e-01 -1.06113851e+00
2.90327430e-01 -1.13704288e+00 1.35676146e-01 -2.10877156e+00
5.41305900e-01 1.84982672e-01 -3.36550415e-01 7.48129964e-01
-6.04676247e-01 1.96950659e-01 5.00799417e-01 -5.25341481e-02
-8.33819032e-01 4.03293759e-01 1.27201414e+00 -4.24273729e-01
1.22340946e-02 -5.56613922e-01 -1.02070713e+00 7.33214259e-01
5.62217951e-01 -3.88093323e-01 -4.14686620e-01 -8.42606843e-01
5.47058582e-01 -3.41764055e-02 9.37575340e-01 -7.87850857e-01
1.74871191e-01 -5.49514964e-02 3.66120547e-01 -4.74964172e-01
9.38902423e-02 -6.29046440e-01 -6.78864717e-02 4.09360021e-01
-3.95563960e-01 3.35704833e-01 2.13774368e-01 4.39359695e-01
-2.16276661e-01 -2.01761443e-03 6.08555675e-01 -1.77946985e-01
-7.84037173e-01 1.30799681e-01 -4.37730402e-02 1.03523359e-01
7.89319873e-01 -1.60759896e-01 -5.89949369e-01 -7.69104958e-02
-6.57992542e-01 9.87575725e-02 4.24174488e-01 3.11234325e-01
6.03419662e-01 -1.51360285e+00 -6.16298020e-01 8.29372630e-02
2.97161013e-01 -3.80301513e-02 3.65378976e-01 6.71720982e-01
-5.03042102e-01 2.70624578e-01 -1.28832147e-01 -6.93747461e-01
-9.94393408e-01 7.85913408e-01 3.60096782e-01 -3.31024528e-01
-7.46844888e-01 1.06953931e+00 9.69614387e-01 -7.57455379e-02
3.07457656e-01 -7.86035299e-01 -3.24057825e-02 -7.42649660e-03
3.93989325e-01 -2.19140008e-01 -6.31068274e-02 -6.33791506e-01
-4.44524139e-01 5.31589091e-01 2.50538707e-01 -7.04280958e-02
1.04898381e+00 -1.09456226e-01 -2.59989619e-01 4.08130527e-01
9.50011790e-01 -1.44127771e-01 -1.60970140e+00 -1.07513659e-01
-4.21778932e-02 -1.77726194e-01 5.85079417e-02 -7.44051516e-01
-1.07272506e+00 1.20977390e+00 2.64623106e-01 3.49010229e-01
1.06936920e+00 4.55790848e-01 3.56349379e-01 2.24841282e-01
1.73488736e-01 -6.13558650e-01 1.91163510e-01 1.93457469e-01
1.10554671e+00 -1.20024860e+00 -1.23045869e-01 -2.87660092e-01
-7.20187604e-01 9.68687594e-01 6.85338259e-01 -9.74178314e-02
9.85323712e-02 3.01578199e-03 7.13842213e-02 -3.91155630e-01
-7.32219100e-01 -2.81964123e-01 7.11323738e-01 4.40672129e-01
5.30160666e-01 2.39156961e-01 -1.91775504e-02 8.40150297e-01
-1.79119751e-01 -4.35034364e-01 -6.96927682e-02 5.33680916e-01
-5.80462039e-01 -5.84021270e-01 -2.73280472e-01 1.09183550e-01
-8.90487060e-02 -5.22472799e-01 -4.54418153e-01 7.93269873e-01
3.46227795e-01 7.63806999e-01 1.44406602e-01 -2.89427191e-01
4.50325012e-01 -1.07566370e-02 3.87508243e-01 -7.86032617e-01
-5.48303723e-01 -1.16447374e-01 -4.57229242e-02 -8.12514603e-01
-9.48558971e-02 -3.16000581e-01 -1.43253052e+00 -8.25810060e-03
4.31513377e-02 -2.21354917e-01 5.15315533e-01 9.76307869e-01
3.99094164e-01 8.71284306e-01 1.83935031e-01 -8.92268658e-01
-3.18413496e-01 -7.72587895e-01 -2.80543625e-01 4.87818390e-01
4.02841806e-01 -6.29249513e-01 -1.61337644e-01 2.95197368e-01] | [10.598921775817871, 1.6051007509231567] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.