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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
874729a7-8d34-4ca9-9d80-487cc6f2f51b | mlbf-net-a-multi-lead-branch-fusion-network | 2008.07263 | null | https://arxiv.org/abs/2008.07263v1 | https://arxiv.org/pdf/2008.07263v1.pdf | MLBF-Net: A Multi-Lead-Branch Fusion Network for Multi-Class Arrhythmia Classification Using 12-Lead ECG | Automatic arrhythmia detection using 12-lead electrocardiogram (ECG) signal plays a critical role in early prevention and diagnosis of cardiovascular diseases. In the previous studies on automatic arrhythmia detection, most methods concatenated 12 leads of ECG into a matrix, and then input the matrix to a variety of feature extractors or deep neural networks for extracting useful information. Under such frameworks, these methods had the ability to extract comprehensive features (known as integrity) of 12-lead ECG since the information of each lead interacts with each other during training. However, the diverse lead-specific features (known as diversity) among 12 leads were neglected, causing inadequate information learning for 12-lead ECG. To maximize the information learning of multi-lead ECG, the information fusion of comprehensive features with integrity and lead-specific features with diversity should be taken into account. In this paper, we propose a novel Multi-Lead-Branch Fusion Network (MLBF-Net) architecture for arrhythmia classification by integrating multi-loss optimization to jointly learning diversity and integrity of multi-lead ECG. MLBF-Net is composed of three components: 1) multiple lead-specific branches for learning the diversity of multi-lead ECG; 2) cross-lead features fusion by concatenating the output feature maps of all branches for learning the integrity of multi-lead ECG; 3) multi-loss co-optimization for all the individual branches and the concatenated network. We demonstrate our MLBF-Net on China Physiological Signal Challenge 2018 which is an open 12-lead ECG dataset. The experimental results show that MLBF-Net obtains an average $F_1$ score of 0.855, reaching the highest arrhythmia classification performance. The proposed method provides a promising solution for multi-lead ECG analysis from an information fusion perspective. | ['Aiping Liu', 'Min Gao', 'Jing Zhang', 'Xu Zhang', 'Xun Chen', 'Xiang Chen', 'Deng Liang'] | 2020-08-17 | null | null | null | null | ['arrhythmia-detection'] | ['medical'] | [ 4.54159640e-02 -4.55007821e-01 1.38686061e-01 -3.89151007e-01
-1.06028724e+00 -4.22122329e-01 -5.09074569e-01 1.23484001e-01
-2.30546549e-01 5.80360949e-01 3.41058634e-02 -2.58635789e-01
-5.49197197e-01 -4.12948221e-01 -2.50345945e-01 -8.16092372e-01
-4.34493333e-01 7.73363784e-02 -5.34760714e-01 -1.58592127e-02
-1.36636227e-01 5.81558108e-01 -9.72624898e-01 3.15709442e-01
9.96000111e-01 1.45476532e+00 -1.06176227e-01 7.19936490e-01
4.52373266e-01 4.24772263e-01 -8.34710479e-01 -4.23767865e-02
2.89715439e-01 -9.46656764e-01 -1.78997412e-01 -1.74135908e-01
-6.95260540e-02 -2.57587224e-01 -7.44158179e-02 7.74400413e-01
1.30743682e+00 -4.09615725e-01 5.29015362e-01 -9.21452820e-01
1.32090524e-01 6.85552418e-01 -5.56353092e-01 5.17176986e-01
-6.12635426e-02 2.80023515e-01 9.94449317e-01 -9.58285391e-01
2.40230411e-01 5.92794597e-01 1.06895423e+00 -9.85417366e-02
-9.90141094e-01 -7.86591887e-01 -1.34780601e-01 2.61535466e-01
-1.71218097e+00 -6.92442432e-02 1.21514165e+00 -4.54576224e-01
6.01042211e-01 5.17980099e-01 1.06768072e+00 6.00123882e-01
5.11647999e-01 7.11274505e-01 8.52181911e-01 -1.26089454e-01
-2.50757039e-01 -1.52193427e-01 2.59695292e-01 6.24453247e-01
2.66515583e-01 1.22609459e-01 -3.69521588e-01 -3.97611380e-01
7.54762471e-01 2.03783587e-01 -5.04951715e-01 3.31322998e-02
-1.42864013e+00 5.53263664e-01 6.09591007e-01 3.77154082e-01
-5.93313932e-01 -2.45939389e-01 5.39891064e-01 5.55914342e-01
6.03392161e-02 6.74122810e-01 -5.93450069e-01 9.80724487e-03
-8.62169504e-01 2.03537583e-01 5.56901336e-01 4.48552072e-01
6.66529596e-01 2.83642173e-01 -6.29326403e-01 8.32253039e-01
2.91396439e-01 4.22390699e-01 5.37143350e-01 -6.91209674e-01
6.33180976e-01 6.58560872e-01 -2.11239234e-01 -1.24214435e+00
-8.66885960e-01 -1.35570240e+00 -1.58550882e+00 -1.14064232e-01
2.40129247e-01 -6.86192036e-01 -4.51020569e-01 1.75119901e+00
1.91614524e-01 1.71524242e-01 -3.85790318e-02 1.06118631e+00
9.39061761e-01 4.19079661e-01 -2.65287310e-01 -7.49788761e-01
1.31074286e+00 -3.93267006e-01 -7.17228353e-01 2.35980585e-01
6.01269245e-01 -5.04856348e-01 6.56304419e-01 6.89233124e-01
-9.24091578e-01 -8.11387241e-01 -1.27076244e+00 3.36368084e-01
3.29670876e-01 5.80500960e-01 3.23711663e-01 5.94182074e-01
-5.92633486e-01 5.42378068e-01 -5.70810735e-01 5.01202703e-01
7.28820264e-01 4.41393584e-01 -1.14449874e-01 1.55564353e-01
-1.43898606e+00 5.73734283e-01 2.92284340e-01 5.44638753e-01
-7.61645675e-01 -8.78344715e-01 -7.38470137e-01 4.48949002e-02
1.84892684e-01 -9.10531402e-01 5.40819526e-01 -7.69704521e-01
-9.94285464e-01 4.71958190e-01 3.75924744e-02 -2.37739995e-01
4.56181198e-01 -2.42035300e-01 -4.27622169e-01 5.61065041e-02
-6.14311732e-02 -3.26690413e-02 7.78807759e-01 -9.71531928e-01
-5.37291825e-01 -7.30564415e-01 -3.35277051e-01 4.03309613e-01
-2.55429775e-01 -2.06092164e-01 -1.16428383e-01 -1.00652552e+00
4.23788905e-01 -5.58425486e-01 -3.23321790e-01 -3.58829081e-01
-5.34022033e-01 9.06875283e-02 4.34673518e-01 -9.52343345e-01
1.89077091e+00 -2.31574464e+00 3.39045882e-01 5.04872084e-01
7.68235564e-01 4.28587973e-01 -1.64904278e-02 1.52925044e-01
-9.85298827e-02 1.30981669e-01 -4.44365948e-01 5.93996374e-03
-4.15593892e-01 -8.45043957e-02 -1.61345024e-02 4.31227893e-01
2.06284732e-01 1.05088639e+00 -5.98062217e-01 -5.09026229e-01
9.72625315e-02 6.20603383e-01 -3.51029187e-01 3.12294513e-01
4.91101444e-01 8.45815539e-01 -5.33661306e-01 7.44305730e-01
6.99881315e-01 -1.61117464e-01 1.25266701e-01 -5.81541896e-01
2.11678967e-01 -1.02316648e-01 -1.25733650e+00 1.86732268e+00
-5.07716201e-02 -3.70513718e-03 -1.49410814e-01 -1.01061130e+00
1.16134644e+00 6.34301484e-01 1.05121768e+00 -3.40320647e-01
2.15447813e-01 3.53907704e-01 5.72272658e-01 -5.50457597e-01
-3.99396658e-01 -1.94249421e-01 -6.90548569e-02 4.97839808e-01
-2.50219498e-02 7.23445788e-02 -8.10413361e-02 -6.60875440e-02
9.46082175e-01 -1.83000043e-01 4.62369829e-01 -8.36351141e-02
6.53907597e-01 -6.25621974e-01 1.25018096e+00 7.72938550e-01
-2.62241244e-01 9.09508228e-01 5.66975117e-01 -9.04552400e-01
-6.20732009e-01 -9.64024961e-01 -4.17778492e-01 2.57242918e-01
-8.27723816e-02 -5.67972839e-01 -4.78800803e-01 -8.13394129e-01
-1.97355505e-02 3.76116708e-02 -2.36973614e-01 -3.64007205e-01
-7.19893217e-01 -1.45972490e+00 9.91278708e-01 5.67295372e-01
3.76856148e-01 -1.02968907e+00 -9.19024944e-01 4.74721968e-01
-5.80089211e-01 -5.37234485e-01 -3.98599833e-01 3.17182690e-01
-1.12592840e+00 -1.13977408e+00 -7.47962058e-01 -4.69712764e-01
4.22739059e-01 -2.99590170e-01 1.05864859e+00 2.44879946e-01
-6.15001380e-01 -1.94481283e-01 -2.28230149e-01 -7.46365845e-01
5.82768805e-02 2.35028341e-02 -7.06758723e-03 6.09089255e-01
5.46651408e-02 -7.59738386e-01 -8.66672337e-01 2.98551410e-01
-4.97775823e-01 -5.23282029e-02 8.01111341e-01 1.22541726e+00
8.44353795e-01 1.47026435e-01 1.13744032e+00 -7.16468513e-01
8.42738450e-01 -3.76062691e-01 -7.87094161e-02 2.94801921e-01
-7.35935211e-01 -4.72213089e-01 6.78363383e-01 -1.48001671e-01
-4.87227112e-01 2.22032517e-01 -4.92196023e-01 -4.27547127e-01
1.43383160e-01 7.58693516e-01 -5.46519935e-01 4.88313250e-02
6.30132794e-01 3.63461733e-01 6.90823980e-03 -2.76673913e-01
-1.77317098e-01 5.94345510e-01 2.99267650e-01 -4.38080788e-01
3.69116753e-01 -9.00995359e-02 2.32593745e-01 -5.04980862e-01
-7.36931920e-01 -4.76743847e-01 -6.80326581e-01 -9.10526067e-02
6.82360649e-01 -1.01828492e+00 -7.01744020e-01 7.67705023e-01
-9.80609119e-01 3.33052576e-01 -3.14018667e-01 6.77228749e-01
-2.42541254e-01 4.64881629e-01 -5.42332947e-01 -8.09439957e-01
-9.06393349e-01 -1.11452520e+00 7.95747459e-01 3.38727459e-02
-2.29943186e-01 -6.42912805e-01 -1.73790045e-02 5.78133203e-02
2.72046506e-01 7.42992103e-01 8.94264519e-01 -7.07762361e-01
-1.91303208e-01 -3.37068498e-01 -7.23351836e-02 7.21448898e-01
3.68595630e-01 -4.31991071e-01 -9.01494801e-01 -3.90382528e-01
4.76775169e-01 -2.07559362e-01 8.00244510e-01 6.83857799e-01
1.25005257e+00 -2.02603921e-01 -6.27199933e-02 1.23047614e+00
1.29749477e+00 4.10250694e-01 5.54390252e-01 -1.50015905e-01
1.01317906e+00 1.31777257e-01 3.54773611e-01 7.02170789e-01
3.98054123e-01 4.10328895e-01 1.04438007e-01 -5.21652877e-01
1.66563451e-01 3.14506263e-01 -4.41698991e-02 1.21669686e+00
-3.91643286e-01 2.09599689e-01 -8.67444992e-01 2.31539577e-01
-1.81986213e+00 -7.53898144e-01 -2.39110261e-01 2.29766059e+00
9.41224873e-01 -2.20704824e-02 2.74097055e-01 6.56827211e-01
4.71594781e-01 -1.02661155e-01 -9.71871495e-01 -5.26602156e-02
-5.00424266e-01 1.18390836e-01 -1.99691385e-01 1.10822454e-01
-1.13930607e+00 -1.04513414e-01 5.27740669e+00 5.71705818e-01
-1.30113947e+00 -1.45125119e-02 1.01439297e+00 -9.93188843e-02
-1.06058583e-01 -2.62760818e-01 -6.22080505e-01 4.37447846e-01
5.13873100e-01 -7.94149563e-02 1.83291271e-01 2.32438594e-01
1.37646824e-01 3.71266067e-01 -8.92766654e-01 1.57922900e+00
7.93081801e-03 -1.04559517e+00 -6.43584132e-03 -1.95746228e-01
4.67187494e-01 -1.73889622e-01 -1.10931106e-01 1.83599144e-01
-6.74779892e-01 -1.03708148e+00 3.89570415e-01 9.04207230e-01
1.10088158e+00 -9.05175567e-01 1.06733906e+00 5.34567356e-01
-1.20787501e+00 -4.02460992e-01 -1.24294475e-01 8.86244252e-02
1.00857101e-01 1.19993305e+00 -5.56221485e-01 1.22249722e+00
6.42290890e-01 1.27859843e+00 -5.17166972e-01 1.22515810e+00
4.11081091e-02 8.14203739e-01 -2.95706570e-01 4.39626575e-01
-2.97435343e-01 -1.46033406e-01 7.78886616e-01 1.14174891e+00
4.61097151e-01 1.98852628e-01 5.31002402e-01 7.35293984e-01
1.22539394e-01 2.97410011e-01 -4.51588959e-01 3.27224910e-01
2.93714434e-01 1.35420835e+00 -3.12549233e-01 -3.27804923e-01
-9.96342152e-02 3.51577312e-01 -2.68686432e-02 3.13883960e-01
-6.60808742e-01 -9.27795947e-01 2.13996410e-01 -6.37368858e-02
-1.29089430e-01 3.21112752e-01 -1.02000487e+00 -1.17802262e+00
3.71192843e-01 -1.35521114e+00 7.35870719e-01 -2.57779688e-01
-1.30430341e+00 1.11013913e+00 -4.05378968e-01 -1.60971391e+00
-2.61226505e-01 -1.76929384e-01 -5.51291406e-01 1.37685907e+00
-1.28632784e+00 -1.03756928e+00 -2.12789714e-01 7.03478038e-01
1.57967716e-01 -4.72180158e-01 1.08059084e+00 6.25395954e-01
-6.97092712e-01 9.03610647e-01 -3.73724461e-01 3.38393807e-01
5.96022844e-01 -1.19398260e+00 -7.36537576e-02 7.96796381e-01
2.53140450e-01 7.20096052e-01 -1.45484684e-02 -4.92928892e-01
-1.20545042e+00 -9.70786810e-01 8.87192965e-01 -3.02300632e-01
-1.58394963e-01 -6.92956969e-02 -9.44944799e-01 2.01239154e-01
-2.99633682e-01 3.07802886e-01 9.95110989e-01 1.94780961e-01
1.79602522e-02 -8.28551292e-01 -8.16745162e-01 1.69812962e-01
7.10215926e-01 -4.72591847e-01 -4.23354894e-01 -5.63940853e-02
1.46450967e-01 -4.03804719e-01 -1.25556660e+00 1.06719899e+00
9.96693254e-01 -9.67496991e-01 1.08610523e+00 -4.62894857e-01
1.80491135e-01 -5.07872164e-01 1.25155106e-01 -1.28844404e+00
-5.30613124e-01 -8.23575318e-01 -2.75221914e-01 8.95166755e-01
3.99155945e-01 -6.66597545e-01 2.23803461e-01 7.00962022e-02
-4.20402229e-01 -1.46449113e+00 -8.32619727e-01 -1.95513517e-01
-2.80270994e-01 -4.42753851e-01 5.42967379e-01 8.45400691e-01
-1.86767593e-01 5.18958628e-01 -7.05552876e-01 -2.03724131e-02
6.19060993e-01 2.49003410e-01 3.86088878e-01 -1.43079436e+00
-5.39206266e-01 -4.64025646e-01 -4.08395529e-01 -5.28978705e-01
-4.26109880e-01 -1.28608119e+00 -2.85423189e-01 -1.32772315e+00
3.13937545e-01 -7.16976762e-01 -1.20297694e+00 4.72373486e-01
-7.40532339e-01 3.29763472e-01 1.74056828e-01 2.58281142e-01
-3.50197881e-01 2.53209203e-01 1.40026379e+00 -1.01086706e-01
-5.72623134e-01 5.24405479e-01 -1.07828236e+00 6.70452416e-01
6.71442151e-01 -4.95745301e-01 -4.90456790e-01 -2.12015703e-01
1.80686146e-01 6.02511823e-01 2.73725748e-01 -1.23854101e+00
5.75982640e-03 3.56637925e-01 9.39335644e-01 -6.54552996e-01
1.30086213e-01 -6.18582666e-01 4.41287339e-01 6.33192360e-01
-1.85423255e-01 1.34756625e-01 7.66653940e-02 3.48828346e-01
-5.35397947e-01 2.28415623e-01 4.67805296e-01 -1.18646868e-01
-6.70358241e-02 6.44808769e-01 -4.71506007e-02 1.86164513e-01
7.97370255e-01 -1.74527034e-01 2.24833176e-01 -6.83244541e-02
-1.07065678e+00 2.76500642e-01 -3.35232437e-01 1.46794751e-01
9.63903248e-01 -1.39085782e+00 -1.29455304e+00 6.29161000e-01
7.23901540e-02 2.11382955e-01 6.32034838e-01 1.35553443e+00
-4.49136198e-01 -1.30836172e-02 -3.76409620e-01 -9.79108155e-01
-1.12739253e+00 1.32821158e-01 5.91071010e-01 -6.83701336e-01
-7.55084515e-01 8.31138909e-01 2.43798465e-01 -1.27580598e-01
2.61446297e-01 -2.01877981e-01 -3.93130839e-01 2.75090069e-01
6.09328210e-01 3.31333160e-01 3.35603476e-01 -4.23753202e-01
-4.97934282e-01 8.97718787e-01 1.41135305e-01 2.54372597e-01
1.25991356e+00 -8.44211057e-02 -2.62920111e-01 7.02251375e-01
1.18682158e+00 -1.93397388e-01 -9.24352884e-01 -1.86350763e-01
-4.60181773e-01 -2.75349677e-01 2.49305125e-02 -1.20623851e+00
-1.40977955e+00 1.28503084e+00 8.52953017e-01 -2.03910261e-01
1.75855219e+00 -6.74038470e-01 9.05728340e-01 3.27561945e-01
3.40887725e-01 -6.86380029e-01 -9.26044285e-02 2.00762406e-01
9.79869962e-01 -9.44660246e-01 1.47704512e-01 -6.89614862e-02
-8.83211613e-01 1.17009020e+00 2.28392482e-01 7.09900027e-03
1.04186499e+00 1.53939098e-01 5.26148021e-01 -1.48963973e-01
-4.04340237e-01 2.32183814e-01 7.17145920e-01 4.35764194e-01
4.02750909e-01 1.48739427e-01 -5.65705001e-01 1.45324707e+00
6.52367324e-02 1.98230028e-01 -4.57151160e-02 6.47657394e-01
-3.82480174e-02 -1.09743190e+00 -2.64644563e-01 8.92362118e-01
-8.46063375e-01 -1.83235690e-01 -3.00014820e-02 1.04198597e-01
5.13125479e-01 8.41493189e-01 -5.08022249e-01 -7.12662220e-01
2.96188980e-01 1.96837991e-01 3.29293340e-01 -4.28977966e-01
-1.11851740e+00 4.93028879e-01 -2.32275024e-01 -3.39120984e-01
-5.90311699e-02 -6.43238068e-01 -1.02943671e+00 4.73323762e-01
-3.23040813e-01 1.24531344e-01 3.66342247e-01 7.57441044e-01
5.73734283e-01 9.81147349e-01 1.00873232e+00 -4.85509157e-01
-7.44258463e-01 -1.02205992e+00 -9.21863258e-01 4.17029440e-01
7.38628864e-01 -1.94616050e-01 -1.72720790e-01 4.73336801e-02] | [14.272746086120605, 3.251847982406616] |
cca66376-a139-4213-bc79-e157afc4481a | unssor-unsupervised-neural-speech-separation | 2305.20054 | null | https://arxiv.org/abs/2305.20054v1 | https://arxiv.org/pdf/2305.20054v1.pdf | UNSSOR: Unsupervised Neural Speech Separation by Leveraging Over-determined Training Mixtures | In reverberant conditions with multiple concurrent speakers, each microphone acquires a mixture signal of multiple speakers at a different location. In over-determined conditions where the microphones out-number speakers, we can narrow down the solutions to speaker images and realize unsupervised speech separation by leveraging each mixture signal as a constraint (i.e., the estimated speaker images at a microphone should add up to the mixture). Equipped with this insight, we propose UNSSOR, an algorithm for $\textbf{u}$nsupervised $\textbf{n}$eural $\textbf{s}$peech $\textbf{s}$eparation by leveraging $\textbf{o}$ver-determined training mixtu$\textbf{r}$es. At each training step, we feed an input mixture to a deep neural network (DNN) to produce an intermediate estimate for each speaker, linearly filter the estimates, and optimize a loss so that, at each microphone, the filtered estimates of all the speakers can add up to the mixture to satisfy the above constraint. We show that this loss can promote unsupervised separation of speakers. The linear filters are computed in each sub-band based on the mixture and DNN estimates through the forward convolutive prediction (FCP) algorithm. To address the frequency permutation problem incurred by using sub-band FCP, a loss term based on minimizing intra-source magnitude scattering is proposed. Although UNSSOR requires over-determined training mixtures, we can train DNNs to achieve under-determined separation (e.g., unsupervised monaural speech separation). Evaluation results on two-speaker separation in reverberant conditions show the effectiveness and potential of UNSSOR. | ['Shinji Watanabe', 'Zhong-Qiu Wang'] | 2023-05-31 | null | null | null | null | ['speech-separation', 'speaker-separation'] | ['speech', 'speech'] | [ 1.45078003e-01 -7.72938654e-02 3.90592307e-01 -2.71932602e-01
-1.18857980e+00 -4.89987344e-01 3.49145383e-02 -4.26626146e-01
-2.72215605e-01 4.47133571e-01 2.32786462e-01 -2.14992031e-01
-2.43834555e-01 -4.90375459e-01 -8.77505779e-01 -1.09908772e+00
-1.46439284e-01 2.55401395e-02 -2.85242766e-01 1.12898536e-02
-3.23911816e-01 3.51664335e-01 -1.69925416e+00 2.13033020e-01
1.08674729e+00 1.20712090e+00 4.55668092e-01 8.63017559e-01
-1.29959941e-01 4.68947440e-01 -1.04919279e+00 -1.58573627e-01
4.73539740e-01 -7.25166678e-01 -9.13058873e-03 1.89873353e-01
5.46639979e-01 -2.73666918e-01 -3.42581242e-01 1.25795841e+00
7.84767687e-01 4.32941467e-01 6.28703356e-01 -8.22828233e-01
-3.56000066e-01 9.12153125e-01 -5.70645988e-01 1.83325052e-01
-9.24734920e-02 -7.10587017e-03 8.11685562e-01 -1.12843788e+00
-2.75870025e-01 1.18487465e+00 5.66573858e-01 4.22013104e-01
-1.18326044e+00 -1.16218221e+00 4.19991225e-01 -1.30017862e-01
-1.43764710e+00 -9.69819129e-01 8.26652586e-01 -2.97171265e-01
6.68304682e-01 6.60080194e-01 4.99803007e-01 6.46655083e-01
-3.50690871e-01 6.04536831e-01 7.93491542e-01 -4.87628102e-01
1.32533208e-01 1.29343048e-01 1.35654896e-01 2.63232231e-01
-1.64158702e-01 1.98846012e-01 -6.96926355e-01 4.19614129e-02
6.62577629e-01 -2.52248436e-01 -7.76435375e-01 2.62109905e-01
-7.03013241e-01 5.09169102e-01 2.76579112e-01 4.10845667e-01
-2.36242115e-01 5.44164442e-02 -1.24206536e-01 2.07109913e-01
5.03554642e-01 3.31671506e-01 -3.29148710e-01 2.11971119e-01
-1.17971992e+00 -1.03215627e-01 7.14794219e-01 8.58665943e-01
8.43151271e-01 6.86920047e-01 -8.63646902e-03 1.37628186e+00
5.11033893e-01 9.42804575e-01 3.03590477e-01 -1.15598619e+00
6.63430154e-01 -3.01010370e-01 3.62720378e-02 -6.67195559e-01
-1.14724301e-02 -1.01956666e+00 -9.89523768e-01 1.53204769e-01
3.38394701e-01 -5.42160094e-01 -9.39842522e-01 2.04057169e+00
2.78568119e-01 5.02622902e-01 1.64352313e-01 1.11791825e+00
5.41701853e-01 1.08146906e+00 -3.98938656e-01 -7.37905681e-01
1.13824058e+00 -8.81965399e-01 -8.16212058e-01 -2.52545029e-01
-1.08732350e-01 -1.02210820e+00 8.29346955e-01 7.42832363e-01
-1.37317085e+00 -7.94114947e-01 -1.03579950e+00 4.70762402e-01
2.83582240e-01 1.06684528e-01 1.02422222e-01 9.91982818e-01
-9.90939260e-01 2.12615028e-01 -7.27318525e-01 3.74542326e-01
-5.33665828e-02 3.12941164e-01 8.92500281e-02 4.65051718e-02
-1.05192316e+00 3.43511581e-01 -2.32045501e-01 5.97892225e-01
-1.22703207e+00 -7.97372460e-01 -8.02129388e-01 3.28930497e-01
1.92956120e-01 -5.28964520e-01 1.20847607e+00 -1.22799993e+00
-1.59597707e+00 2.04034865e-01 -5.12782991e-01 -4.28661495e-01
2.81300038e-01 -2.27089018e-01 -8.70122552e-01 2.10443467e-01
5.11561185e-02 2.95068711e-01 1.18287909e+00 -1.48494530e+00
-6.83049381e-01 -2.27338508e-01 -2.95001537e-01 4.50460196e-01
-5.56474328e-01 8.38306248e-02 -4.18148816e-01 -7.94788837e-01
4.99129772e-01 -6.02899611e-01 -1.45378217e-01 -4.78764117e-01
-5.43313563e-01 2.72389680e-01 4.81714338e-01 -8.78646255e-01
1.20733678e+00 -2.62930155e+00 2.40574731e-03 4.12597448e-01
2.05132514e-01 1.76790789e-01 -7.88689107e-02 6.25546277e-03
-1.68130934e-01 -1.68315455e-01 -3.80584717e-01 -8.25361371e-01
-3.18714716e-02 -3.64461839e-02 -5.07420599e-01 4.40756649e-01
-1.54537171e-01 6.43844381e-02 -4.99627471e-01 -1.42786607e-01
2.63707578e-01 7.11566746e-01 -7.15340316e-01 4.37325031e-01
4.78167385e-02 5.00948071e-01 -9.49249603e-04 3.36802155e-01
1.10134125e+00 2.81142294e-01 5.68156056e-02 -1.67453587e-01
-2.15567693e-01 4.18189198e-01 -1.66180789e+00 1.23533201e+00
-6.90996289e-01 4.97271657e-01 1.11320806e+00 -1.12211275e+00
8.01271379e-01 5.05303264e-01 4.23056573e-01 -3.74086857e-01
-3.67197320e-02 4.85701740e-01 1.65910527e-01 -2.24004984e-01
1.44879192e-01 -5.60112715e-01 2.20746070e-01 2.43964717e-01
3.09102684e-01 -1.07934661e-01 -1.73268747e-02 -6.36177063e-02
6.66918159e-01 -5.11731446e-01 -3.71455520e-01 -2.23633796e-01
5.87241590e-01 -8.11447740e-01 6.05183363e-01 8.89927447e-01
-1.11905687e-01 6.51246846e-01 -1.40233308e-01 4.17129427e-01
-6.07836723e-01 -1.53094554e+00 -9.85762849e-02 1.25411260e+00
1.95584185e-02 -2.11709575e-03 -9.04061794e-01 -8.50668252e-02
-3.67157102e-01 7.37680197e-01 6.51212633e-02 -1.21094966e-02
-7.44430184e-01 -7.06071675e-01 5.54038048e-01 2.35579699e-01
4.73341972e-01 -6.40732467e-01 -5.43895662e-02 3.64091247e-01
-3.81789297e-01 -8.08696032e-01 -9.08735633e-01 5.59670925e-01
-4.65511978e-01 -4.35586482e-01 -9.55118179e-01 -9.08800542e-01
7.31281281e-01 4.38394457e-01 6.92036927e-01 -3.99535269e-01
1.12748981e-01 2.77509809e-01 2.53763888e-02 -4.55027819e-01
-4.32308018e-01 -4.17829841e-01 4.19961005e-01 4.22989458e-01
-2.79979128e-02 -1.04089582e+00 -7.28802741e-01 4.20576215e-01
-7.29901016e-01 -1.43741980e-01 2.71396279e-01 7.37444818e-01
5.15789926e-01 6.27642512e-01 6.90614939e-01 -6.37670532e-02
5.33238649e-01 -1.51234865e-01 -6.85308099e-01 -1.12502038e-01
-1.25842094e-01 -3.08717787e-01 8.75826895e-01 -7.49600947e-01
-1.32191217e+00 -2.32506737e-01 -3.81837636e-01 -7.21442223e-01
-9.73422602e-02 2.75042325e-01 -6.46165490e-01 3.62137765e-01
6.00053608e-01 5.28784633e-01 -1.54959872e-01 -6.26047671e-01
4.10277039e-01 8.51159096e-01 7.36576200e-01 -4.11401600e-01
8.18882763e-01 2.99596697e-01 -5.13991475e-01 -9.99902129e-01
-6.02624774e-01 -4.47796643e-01 7.42163733e-02 -3.13029319e-01
6.36091232e-01 -1.25189912e+00 -7.57964551e-01 6.05887532e-01
-9.48146701e-01 -2.42499709e-01 -3.48241329e-01 1.11037421e+00
-2.63473839e-01 2.89021194e-01 -5.79823673e-01 -1.44033766e+00
-1.51952192e-01 -1.21174788e+00 5.66042244e-01 2.86709249e-01
5.88402450e-02 -5.59382915e-01 -2.70807058e-01 4.44338441e-01
3.43934566e-01 -4.79683250e-01 5.28527200e-01 -4.64483380e-01
-5.79998195e-01 1.24805830e-01 1.86305657e-01 9.53642607e-01
3.80486608e-01 -2.31164098e-01 -1.38295758e+00 -2.43612230e-01
7.82719791e-01 1.66224614e-01 1.01659548e+00 9.49922502e-01
1.02873051e+00 -6.64476156e-01 -1.22964211e-01 7.35494077e-01
1.08194995e+00 5.53008199e-01 3.60908926e-01 -3.68497103e-01
6.54440999e-01 5.78594685e-01 -7.33409915e-03 3.48218083e-01
6.44488633e-02 2.83471704e-01 2.74384737e-01 -1.98909506e-01
-3.85575742e-01 -6.48850352e-02 6.32593989e-01 1.28938782e+00
2.03371316e-01 -4.06521797e-01 -4.18102324e-01 6.48971379e-01
-1.22872722e+00 -9.67907608e-01 5.33655062e-02 2.44655180e+00
8.84425163e-01 9.81001481e-02 2.12471765e-02 2.85583615e-01
1.09018469e+00 1.38265669e-01 -5.53940773e-01 -1.73237234e-01
-2.67349809e-01 4.01427746e-01 2.98076391e-01 9.31358814e-01
-8.25327098e-01 3.82225722e-01 5.13376999e+00 1.10972774e+00
-1.29571247e+00 2.43695304e-01 6.90414011e-01 -5.28610170e-01
-5.67506075e-01 -3.32264423e-01 -7.33126998e-01 6.13502920e-01
9.19545114e-01 1.79712269e-02 8.60785306e-01 5.24937868e-01
4.18847799e-01 -4.08129022e-02 -9.69848812e-01 1.01676416e+00
1.43009797e-01 -7.81188190e-01 -3.18103522e-01 1.65052991e-02
5.60482919e-01 -5.64253815e-02 5.05261540e-01 2.79495984e-01
3.11631203e-01 -9.68095005e-01 1.08776832e+00 1.05003186e-01
6.95516229e-01 -8.19124699e-01 1.16321288e-01 6.17510498e-01
-1.35921872e+00 -3.10489625e-01 -1.64609373e-01 3.24635804e-02
3.29057366e-01 1.21116579e+00 -6.94761157e-01 5.20481467e-01
7.80083418e-01 7.15799481e-02 4.38077182e-01 9.01848495e-01
-2.85529673e-01 8.84299517e-01 -6.14767253e-01 3.87949824e-01
-1.16946116e-01 -3.23660463e-01 9.03263032e-01 1.23308599e+00
7.93940187e-01 2.31253058e-01 5.41300327e-02 9.73673403e-01
-1.59893423e-01 -1.37766346e-01 -9.87602770e-02 2.03180790e-01
6.91176236e-01 9.76971328e-01 -4.03374940e-01 -3.25578272e-01
-1.12634607e-01 8.03901970e-01 -2.07194109e-02 8.47421825e-01
-9.65731680e-01 -4.79275674e-01 9.06051278e-01 1.76246129e-02
4.50013727e-01 -2.16815457e-01 -2.61569589e-01 -1.07724679e+00
2.09722906e-01 -9.10820663e-01 7.88436905e-02 -6.84358180e-01
-1.33024943e+00 6.73947573e-01 -2.89270192e-01 -1.33598852e+00
-6.37947842e-02 -2.97637761e-01 -5.64134955e-01 1.39240837e+00
-1.36421776e+00 -5.72952271e-01 1.45074487e-01 7.34080732e-01
4.59844202e-01 -7.11553395e-02 6.66707575e-01 7.28134096e-01
-5.99142730e-01 7.97878146e-01 3.46259594e-01 3.55235264e-02
6.69849038e-01 -9.67465878e-01 -4.69946638e-02 1.08033323e+00
1.90795258e-01 7.52677202e-01 7.66113937e-01 -2.06066519e-01
-1.07849467e+00 -1.01301014e+00 5.56053877e-01 7.85450935e-02
2.85346091e-01 -7.12622225e-01 -8.85481358e-01 5.00564218e-01
2.14625508e-01 -1.74911886e-01 9.19262409e-01 -5.44532314e-02
-3.44390094e-01 -5.94934285e-01 -9.98114884e-01 5.24906874e-01
8.62018704e-01 -5.96258044e-01 -4.59135950e-01 1.33914754e-01
8.80002916e-01 -4.27713841e-01 -5.90811431e-01 3.15195233e-01
3.77395898e-01 -1.10424340e+00 1.03170133e+00 2.56742872e-02
1.18959419e-01 -4.86023277e-01 -4.01081204e-01 -1.55743611e+00
-8.21120813e-02 -9.68728006e-01 -1.40517280e-01 1.62269127e+00
7.12589741e-01 -8.76760244e-01 4.43689615e-01 5.32885969e-01
-6.26208007e-01 -2.91069657e-01 -1.22336054e+00 -8.35296452e-01
6.83938563e-02 -7.07201779e-01 5.12483537e-01 7.21264064e-01
-5.95902801e-02 1.80114239e-01 -3.64879757e-01 8.37444782e-01
6.84717298e-01 -2.60167615e-03 4.87166077e-01 -7.33387768e-01
-8.66285861e-01 -4.81015295e-01 3.56582016e-01 -1.67056406e+00
5.33827022e-02 -6.29273713e-01 5.00872791e-01 -1.37859714e+00
-2.61771083e-01 -5.90077341e-01 -5.30104816e-01 2.20583810e-04
-3.41616303e-01 -3.46588492e-02 3.77655208e-01 6.12775721e-02
-2.91491002e-02 5.51653922e-01 1.11729407e+00 -3.47039133e-01
-3.55640888e-01 3.69192958e-01 -8.89634907e-01 8.51430058e-01
4.32308972e-01 -4.04460490e-01 -4.66107458e-01 -6.48663044e-01
-3.10150325e-01 4.41212624e-01 1.11243255e-01 -1.21790969e+00
3.17528099e-01 3.71595584e-02 4.53596145e-01 -5.26864231e-01
8.21438909e-01 -9.40717816e-01 3.64241451e-01 2.22454369e-01
-2.61473417e-01 -7.09112346e-01 2.37459615e-01 4.01818126e-01
-4.14103538e-01 -1.08475521e-01 9.14802194e-01 6.28409907e-02
1.61286935e-01 1.02137581e-01 -5.53188503e-01 -1.89214781e-01
5.31675637e-01 -2.23749816e-01 -6.81326911e-02 -7.39901423e-01
-8.23565781e-01 1.76786199e-01 -2.25435123e-01 -5.32323383e-02
4.42332149e-01 -1.11014199e+00 -7.15646505e-01 5.73795199e-01
-4.98180419e-01 3.48185062e-01 7.67530978e-01 7.88797021e-01
7.71569014e-02 1.21228181e-01 4.35168862e-01 -6.27907157e-01
-1.07993615e+00 3.75646740e-01 7.65658557e-01 7.01263780e-03
-1.01340376e-01 1.56050885e+00 8.24942410e-01 -3.69768977e-01
4.34925318e-01 -3.58071953e-01 1.81805789e-02 -2.11834423e-02
6.05687261e-01 4.12828594e-01 2.01107085e-01 -5.89280486e-01
-1.77419588e-01 3.75433952e-01 2.63129979e-01 -7.25975990e-01
1.08091974e+00 -3.94400060e-01 -1.37907997e-01 3.89644265e-01
1.39109278e+00 8.42016399e-01 -1.49524999e+00 -1.19358756e-01
-7.88313031e-01 -4.05484378e-01 5.78617910e-03 -7.51996458e-01
-1.25617218e+00 9.38043535e-01 6.44364476e-01 3.47479820e-01
1.60984266e+00 -7.47175589e-02 7.82870471e-01 -1.80299394e-02
1.04546785e-01 -9.84418750e-01 9.44167227e-02 4.59433705e-01
6.41847134e-01 -6.30810380e-01 -5.20414889e-01 -4.13959861e-01
-2.69316405e-01 8.20621431e-01 4.82001752e-01 8.11789185e-02
8.20805907e-01 4.82956797e-01 2.82435179e-01 2.08541080e-01
-2.81896055e-01 -4.47192788e-03 2.40070879e-01 4.41485018e-01
4.24552500e-01 3.17231089e-01 1.51732996e-01 1.01357210e+00
-5.77014506e-01 -5.21222532e-01 2.60372132e-01 4.09912229e-01
-7.41773546e-01 -8.40616584e-01 -9.94471133e-01 3.71366143e-01
-5.32591522e-01 -4.24040467e-01 5.77537715e-02 -1.71505585e-02
3.54787260e-01 1.54478550e+00 1.81273952e-01 -1.89262733e-01
4.65089470e-01 1.19705454e-01 3.18107426e-01 -5.53228378e-01
-5.69149435e-01 1.10269105e+00 9.05060396e-02 6.42891228e-02
-2.21277952e-01 -4.47305262e-01 -1.28861380e+00 -1.84441894e-01
-6.65125132e-01 5.01636326e-01 6.81100249e-01 6.92061245e-01
7.24210739e-02 9.20871317e-01 9.99319673e-01 -9.39611852e-01
-4.61349159e-01 -1.05171716e+00 -9.86571074e-01 -3.03268107e-03
7.24432766e-01 -2.05871731e-01 -9.91138041e-01 2.10762233e-01] | [15.099431037902832, 5.919931888580322] |
d93df94d-59df-4e19-abfa-32112263ec95 | jump-starting-item-parameters-for-adaptive | null | null | https://aclanthology.org/2021.emnlp-main.67 | https://aclanthology.org/2021.emnlp-main.67.pdf | Jump-Starting Item Parameters for Adaptive Language Tests | A challenge in designing high-stakes language assessments is calibrating the test item difficulties, either a priori or from limited pilot test data. While prior work has addressed ‘cold start’ estimation of item difficulties without piloting, we devise a multi-task generalized linear model with BERT features to jump-start these estimates, rapidly improving their quality with as few as 500 test-takers and a small sample of item exposures (≈6 each) from a large item bank (≈4,000 items). Our joint model provides a principled way to compare test-taker proficiency, item difficulty, and language proficiency frameworks like the Common European Framework of Reference (CEFR). This also enables new item difficulty estimates without piloting them first, which in turn limits item exposure and thus enhances test item security. Finally, using operational data from the Duolingo English Test, a high-stakes English proficiency test, we find that the difficulty estimates derived using this method correlate strongly with lexico-grammatical features that correlate with reading complexity. | ['Burr Settles', 'Manqian Liao', 'Jesse Egbert', 'Geoff T. LaFlair', 'Kevin P. Yancey', 'Arya D. McCarthy'] | null | null | null | null | emnlp-2021-11 | ['skills-assessment', 'language-acquisition'] | ['computer-vision', 'natural-language-processing'] | [-3.65842789e-01 2.59205908e-01 -4.87657726e-01 -2.35840455e-01
-1.52064514e+00 -1.00365055e+00 9.87883359e-02 5.80108821e-01
-1.05942249e+00 8.26502919e-01 5.57060957e-01 -6.67042732e-01
-6.50039196e-01 -7.08136559e-01 -4.90664214e-01 2.34531835e-01
5.21882534e-01 3.68876487e-01 1.11166947e-01 -3.84997926e-03
5.06336868e-01 -9.43829045e-02 -1.31568754e+00 1.83048770e-01
1.48899925e+00 3.21640342e-01 5.59786975e-01 6.45480275e-01
-3.97632122e-02 6.17837846e-01 -6.32171750e-01 -8.97516727e-01
9.29834470e-02 -3.94091755e-01 -8.72055709e-01 -3.02583039e-01
7.34670103e-01 -7.60783732e-01 4.10946965e-01 8.28473151e-01
5.84124565e-01 3.38212550e-01 4.94836628e-01 -4.32343692e-01
-1.34161770e+00 5.35160959e-01 -4.35627811e-02 4.70230162e-01
8.69107306e-01 4.03629929e-01 1.22258961e+00 -8.62413645e-01
3.06933552e-01 7.59106755e-01 8.91997457e-01 3.90884548e-01
-1.03065014e+00 -7.46245027e-01 1.80519432e-01 1.13823473e-01
-1.05890477e+00 -3.78811181e-01 9.35985669e-02 -6.95184588e-01
1.13565397e+00 -5.05828485e-02 8.31165373e-01 1.08851433e+00
9.44518447e-02 2.68489748e-01 1.80393362e+00 -4.37605411e-01
1.62432045e-01 4.41621631e-01 5.40359437e-01 3.34990382e-01
7.16366470e-01 -1.01766162e-01 -6.59755945e-01 -1.94385052e-02
1.02907336e+00 -4.93768752e-01 -2.65314847e-01 1.31745860e-01
-1.09795868e+00 6.50321484e-01 -1.33234993e-01 4.45899785e-01
-2.41645053e-01 -4.79133785e-01 2.90446758e-01 8.85169148e-01
8.34085569e-02 8.61674130e-01 -7.35496938e-01 -7.92583764e-01
-9.41502690e-01 3.43784660e-01 7.14504004e-01 5.62097788e-01
2.43409500e-01 -1.28371492e-02 -2.77172655e-01 1.26108623e+00
3.37931782e-01 5.07583857e-01 8.15322161e-01 -5.32352805e-01
9.07708824e-01 3.85556132e-01 3.85027081e-01 -5.54394901e-01
-5.58300614e-01 -5.65673113e-01 1.99337289e-01 4.89462793e-01
7.90173531e-01 -3.50678146e-01 -4.34586942e-01 2.03299665e+00
-3.17898788e-03 -5.38479090e-01 -4.63164032e-01 7.30402648e-01
4.25418794e-01 6.68475106e-02 6.38898671e-01 -2.26688460e-01
1.45422649e+00 -5.35256743e-01 -4.83729571e-01 -2.82133102e-01
1.00312340e+00 -7.36350656e-01 2.02977490e+00 7.81721234e-01
-1.78252137e+00 -8.98412228e-01 -1.00452566e+00 -2.87149280e-01
-2.72238076e-01 -1.78783610e-01 5.79174042e-01 1.41854513e+00
-1.12940955e+00 2.93719530e-01 -5.87064683e-01 -1.76641971e-01
3.35802585e-02 3.93934876e-01 -4.87806082e-01 -2.39766464e-01
-9.91274655e-01 1.08163202e+00 2.57499874e-01 -3.19982052e-01
-4.56122339e-01 -9.53276575e-01 -8.40989530e-01 2.09684804e-01
2.49455765e-01 -4.92207229e-01 1.36666322e+00 -7.69932389e-01
-1.66678286e+00 6.53745115e-01 1.73174515e-01 4.28809285e-01
2.86723245e-02 -3.17888111e-01 -5.98154366e-01 -1.61307067e-01
3.89410645e-01 1.95240542e-01 2.18693048e-01 -1.34230256e-01
-7.22754478e-01 -2.41926774e-01 3.12368006e-01 6.86665952e-01
-6.99824929e-01 2.72960663e-01 2.10146233e-01 -6.25932157e-01
-4.81391214e-02 -1.19874157e-01 1.96614966e-01 -5.15022874e-01
3.38671118e-01 -3.92112732e-01 -3.36111099e-01 -1.14616549e+00
1.48216105e+00 -1.85233879e+00 -8.78383145e-02 2.77472913e-01
2.00305864e-01 -5.62388599e-02 -3.18122059e-01 3.36040735e-01
-1.30364090e-01 5.39011717e-01 3.32590401e-01 8.73626396e-02
4.55492765e-01 -3.93113256e-01 3.17277968e-01 3.14566344e-01
1.25634931e-02 1.04702890e+00 -9.69863474e-01 -2.28117153e-01
6.49866648e-03 -1.81485102e-01 -1.05516565e+00 -3.69018726e-02
3.21867377e-01 2.53420562e-01 -7.68834054e-02 4.58101660e-01
4.45211619e-01 1.81472693e-02 -2.31843423e-02 5.51550746e-01
-2.94543177e-01 1.02419007e+00 -1.15819907e+00 1.70888615e+00
-5.20140827e-01 1.86808661e-01 -3.76357101e-02 -5.25303483e-01
7.50084877e-01 2.99213886e-01 -6.63371086e-02 -1.22369957e+00
-2.78645068e-01 4.22284901e-01 4.70797062e-01 -6.71510220e-01
5.25442779e-01 -2.76785761e-01 -1.27954513e-01 6.43193066e-01
3.86401564e-01 -2.60681957e-01 3.06177527e-01 -2.02356026e-01
1.19124615e+00 1.37159437e-01 4.80757087e-01 -7.62674093e-01
4.12691087e-01 -2.47686803e-01 4.18726176e-01 1.01669955e+00
-2.04323530e-01 1.70867860e-01 4.09757107e-01 2.18302667e-01
-1.23070228e+00 -1.22333801e+00 -5.35002410e-01 1.36562681e+00
-7.93365896e-01 -4.21738207e-01 -7.40130544e-01 -6.39794230e-01
-1.49205193e-01 9.52038765e-01 -2.83044487e-01 -1.18320540e-01
-1.42481565e-01 -3.27212483e-01 4.06598657e-01 7.16736317e-01
3.26768339e-01 -7.20917225e-01 -1.79499552e-01 4.53112483e-01
-1.57781631e-01 -6.85996771e-01 -5.23841977e-01 -1.50584534e-01
-5.66530347e-01 -8.11951518e-01 -7.10298061e-01 -8.00805151e-01
1.72772288e-01 6.89110160e-02 1.36066234e+00 1.52045429e-01
4.35514987e-01 8.99875879e-01 -3.63238215e-01 -3.67675304e-01
-3.46489735e-02 2.61461884e-01 1.63586363e-01 -9.69200015e-01
9.13477063e-01 -2.78048545e-01 -3.10822964e-01 2.40585253e-01
-6.50548816e-01 -2.59597003e-01 9.40930367e-01 7.55786598e-01
1.17441528e-01 -2.37622261e-01 1.08775890e+00 -4.54900980e-01
1.12335360e+00 -7.16836691e-01 -4.95627850e-01 1.96334749e-01
-8.69895101e-01 -4.76605117e-01 4.71894592e-01 -7.33651459e-01
-1.01625443e+00 -9.39384937e-01 -5.99960625e-01 6.91523731e-01
-3.81095141e-01 9.88107085e-01 6.58444911e-02 -2.86063075e-01
9.07585382e-01 -2.29174003e-01 -7.91114047e-02 -6.27022922e-01
-1.80490062e-01 7.31600463e-01 4.63400632e-01 -1.04012132e+00
6.76456094e-01 -5.83753943e-01 -4.55534607e-01 -5.92046320e-01
-8.79158318e-01 -1.94454268e-01 -4.95576560e-01 -1.09622926e-02
7.21515059e-01 -1.32671106e+00 -8.32559526e-01 1.91582799e-01
-2.92745084e-01 -9.82500911e-01 -3.76908958e-01 1.31337035e+00
-6.32308245e-01 2.73319870e-01 -8.03124428e-01 -5.23875356e-01
1.36926427e-01 -1.07929277e+00 6.37237966e-01 1.64198264e-01
-6.95541441e-01 -1.29601192e+00 3.22510779e-01 4.73016560e-01
4.13520396e-01 -1.52707249e-01 1.17755711e+00 -7.77205706e-01
-1.63428426e-01 -1.30951032e-01 -6.79448470e-02 3.30019742e-01
-6.57960251e-02 -5.88149726e-01 -3.28339487e-01 -4.60167617e-01
4.26490366e-01 -8.48562539e-01 2.55084932e-01 3.72428834e-01
7.01030374e-01 -8.92005414e-02 6.45053923e-01 2.95014739e-01
1.48428142e+00 -2.60359645e-01 5.21942019e-01 6.61660016e-01
4.73703444e-01 4.39572483e-01 6.14950776e-01 1.15159862e-01
8.24897170e-01 6.46870136e-01 -6.49531960e-01 3.85529220e-01
-1.72105759e-01 -4.37024832e-01 7.79530644e-01 1.36565673e+00
-1.89226074e-03 5.34042111e-03 -1.12169337e+00 4.60489452e-01
-1.00891840e+00 -7.88198352e-01 -3.09846133e-01 2.59097314e+00
9.09749687e-01 3.03824216e-01 4.53461707e-01 2.44214892e-01
3.54348660e-01 -4.04397279e-01 -2.89552540e-01 -7.11552441e-01
7.23807560e-03 6.35413289e-01 3.93082470e-01 7.17867553e-01
-3.49831194e-01 7.48300076e-01 7.69060373e+00 7.75178611e-01
-6.30445957e-01 2.35711291e-01 4.70672548e-01 -2.60961294e-01
-6.94441140e-01 -2.06456512e-01 -9.21572208e-01 4.69943255e-01
1.56744361e+00 -4.32628810e-01 6.30993307e-01 4.05841440e-01
2.49781638e-01 -4.52208370e-01 -9.57470238e-01 4.47104901e-01
3.72318290e-02 -4.96023655e-01 -3.11490119e-01 2.91585654e-01
6.03698730e-01 -2.33415723e-01 3.59778553e-01 8.25757861e-01
4.88623142e-01 -1.11732101e+00 9.59763706e-01 5.25051236e-01
1.30079305e+00 -6.39244735e-01 6.22361064e-01 3.44281763e-01
-6.82430327e-01 -4.34638828e-01 -6.71165705e-01 -8.75099063e-01
-1.09923415e-01 3.88470769e-01 -6.12960577e-01 1.54994681e-01
5.45940280e-01 -3.51157002e-02 -9.04968679e-01 1.12408006e+00
-2.47672409e-01 1.03029013e+00 -2.68334866e-01 -6.53104186e-02
2.16485187e-01 -1.19566701e-01 -7.48936087e-03 9.13287878e-01
6.91502869e-01 2.29371279e-01 -1.78754374e-01 5.56753039e-01
-7.28797866e-03 6.86035752e-01 -3.07154536e-01 -3.38316411e-02
6.32268488e-01 1.20837760e+00 -6.25962079e-01 -6.58945963e-02
-1.24398243e+00 4.95763212e-01 7.56275415e-01 1.95055649e-01
-4.83738244e-01 -2.21657008e-01 2.53687680e-01 1.33126944e-01
-3.10839340e-02 -5.38513541e-01 -7.14156032e-01 -1.03604090e+00
1.09321512e-01 -1.25620961e+00 2.53023893e-01 -7.20826030e-01
-1.42849958e+00 -4.25007612e-01 -4.87563163e-02 -7.97514200e-01
-2.20778838e-01 -1.03863072e+00 -3.55673045e-01 1.56874943e+00
-1.04083884e+00 -6.87469184e-01 -2.21318126e-01 5.41591525e-01
5.13056517e-01 6.19572215e-02 7.06464171e-01 3.17436427e-01
-3.84875596e-01 1.07338023e+00 4.68777604e-02 -1.45377666e-01
1.06712294e+00 -1.65754211e+00 3.95556390e-01 6.77168190e-01
-1.56947225e-01 8.29404294e-01 2.26726860e-01 -8.93541753e-01
-1.03954327e+00 -3.53738904e-01 8.50352764e-01 -9.73630428e-01
8.70206773e-01 -4.70777363e-01 -1.11644697e+00 7.15871155e-01
1.33866102e-01 -7.87163079e-01 1.29518127e+00 7.69329548e-01
-8.26676190e-02 3.12369585e-01 -1.15532541e+00 5.94970286e-01
1.22981977e+00 -8.99293661e-01 -9.11372364e-01 3.12701434e-01
4.48938578e-01 -3.60634595e-01 -1.62312829e+00 9.32208002e-02
8.34975183e-01 -9.62890685e-01 6.44971728e-01 -4.97021645e-01
3.64921361e-01 2.66092002e-01 8.89124572e-02 -1.52928567e+00
-1.03005719e+00 -3.13840717e-01 3.19846451e-01 1.30358112e+00
6.28005564e-01 -6.61760747e-01 5.80347896e-01 1.26610613e+00
-4.27650392e-01 -5.14965177e-01 -8.37888241e-01 -7.73793280e-01
8.41376126e-01 -4.48075622e-01 6.23117507e-01 1.02503645e+00
4.82607007e-01 -3.87404440e-03 2.62130708e-01 7.00972825e-02
1.77983180e-01 -1.04272354e+00 5.36995292e-01 -1.32315540e+00
-6.64994001e-01 -6.97717607e-01 -2.13530719e-01 -8.49652708e-01
3.56590152e-02 -8.41873944e-01 -4.36908066e-01 -1.31226492e+00
2.36533418e-01 -3.36103469e-01 -4.02704060e-01 3.24889272e-01
-6.65699840e-01 -1.06886864e-01 1.08982123e-01 -2.69773722e-01
-5.07706702e-01 2.99271159e-02 1.20794868e+00 5.56273878e-01
-2.06101611e-01 -3.36045980e-01 -1.56460452e+00 5.02540946e-01
9.50128734e-01 -2.51438111e-01 -6.33700371e-01 -8.80515099e-01
7.14387298e-01 7.00318217e-02 1.36900261e-01 -1.23441613e+00
1.94395348e-01 -3.17286283e-01 8.79048228e-01 1.49496675e-01
-2.03491986e-01 -2.75854051e-01 -4.13608924e-02 -8.60827938e-02
-4.22131777e-01 6.34821713e-01 3.83122414e-01 -2.70525217e-01
1.15230260e-02 -1.09404707e+00 3.75596255e-01 -3.04794639e-01
-3.69322538e-01 -1.38741806e-01 -5.03222585e-01 3.67933244e-01
6.30993485e-01 -4.26681668e-01 -3.47655028e-01 -4.02464598e-01
-5.20388305e-01 1.62741486e-02 5.87562978e-01 2.87047446e-01
1.87299624e-01 -1.27049160e+00 -8.80759358e-01 1.81463733e-01
1.67123154e-02 -3.96512151e-01 4.61595863e-01 8.14316571e-01
-2.03833044e-01 7.70938218e-01 -3.57550055e-01 5.31771556e-02
-6.11690998e-01 4.65809494e-01 -8.48776624e-02 -4.00212586e-01
-1.96063399e-01 7.27273166e-01 -1.53931521e-03 -2.92517692e-01
-3.52277458e-01 -3.18023890e-01 -4.35739905e-01 9.34682786e-03
8.93446565e-01 6.39659643e-01 2.93214738e-01 -2.83662945e-01
3.78594339e-01 5.43427885e-01 -2.77235359e-01 -7.38117576e-01
1.02856469e+00 -2.82413542e-01 2.11508840e-01 8.93198073e-01
7.00633228e-01 8.07152510e-01 -9.84079599e-01 4.23537865e-02
2.05727369e-01 -5.46786547e-01 1.95125505e-01 -1.08971703e+00
-2.27885004e-02 3.66167098e-01 3.87096554e-01 1.44712090e-01
9.01778877e-01 -4.24344957e-01 1.73480958e-01 5.52585185e-01
6.00351870e-01 -1.45878685e+00 2.90861931e-02 4.69406307e-01
7.54349470e-01 -8.62411797e-01 -9.87504348e-02 8.39777440e-02
-4.04063612e-01 8.80565763e-01 8.94435585e-01 -5.22860214e-02
2.34513402e-01 1.66193947e-01 -2.24889591e-01 5.63413501e-02
-6.22647405e-01 -1.04121380e-01 3.32210124e-01 7.95866072e-01
8.83555412e-01 3.51784229e-02 -1.14098227e+00 1.23416650e+00
-8.49583030e-01 3.26566398e-02 7.38900363e-01 7.18511879e-01
-7.02021539e-01 -1.19335532e+00 -6.62400424e-01 1.06962907e+00
-7.83458054e-01 -4.39657748e-01 -7.56683871e-02 1.09519756e+00
-1.02333672e-01 9.87147689e-01 1.61341563e-01 -1.71415836e-01
5.26619196e-01 4.86816049e-01 7.07077861e-01 -1.00307024e+00
-7.79808640e-01 -2.18414053e-01 1.76815376e-01 -2.65941203e-01
1.36477677e-02 -8.42140317e-01 -4.60376084e-01 -3.93285066e-01
-4.98652488e-01 2.42373109e-01 4.71443892e-01 1.04222918e+00
4.88614440e-02 2.74491459e-01 1.07766762e-01 -3.00084263e-01
-7.78099179e-01 -1.39943254e+00 -7.51102209e-01 2.49957263e-01
-1.03418380e-01 -6.87186062e-01 -2.25897968e-01 -5.27454019e-01] | [10.953479766845703, 9.912497520446777] |
9592d7dd-eb5b-460a-96c5-dc350a2c8bf0 | ms-ranker-accumulating-evidence-from | 2010.04970 | null | https://arxiv.org/abs/2010.04970v1 | https://arxiv.org/pdf/2010.04970v1.pdf | MS-Ranker: Accumulating Evidence from Potentially Correct Candidates for Answer Selection | As conventional answer selection (AS) methods generally match the question with each candidate answer independently, they suffer from the lack of matching information between the question and the candidate. To address this problem, we propose a novel reinforcement learning (RL) based multi-step ranking model, named MS-Ranker, which accumulates information from potentially correct candidate answers as extra evidence for matching the question with a candidate. In specific, we explicitly consider the potential correctness of candidates and update the evidence with a gating mechanism. Moreover, as we use a listwise ranking reward, our model learns to pay more attention to the overall performance. Experiments on two benchmarks, namely WikiQA and SemEval-2016 CQA, show that our model significantly outperforms existing methods that do not rely on external resources. | ['Jie zhou', 'Ping Jian', 'Peng Li', 'Fandong Meng', 'Yingxue Zhang'] | 2020-10-10 | null | null | null | null | ['answer-selection'] | ['natural-language-processing'] | [-8.71429220e-02 -1.99041087e-02 -2.99148202e-01 -5.07082582e-01
-1.41286206e+00 -6.76349998e-01 3.71724397e-01 6.16152883e-01
-7.36581147e-01 7.86700547e-01 3.66746336e-01 -3.01723987e-01
-2.48345390e-01 -9.94085550e-01 -7.57350087e-01 9.67738777e-03
4.13510233e-01 6.70141697e-01 7.61056602e-01 -3.11856717e-01
5.28322518e-01 -2.72045553e-01 -1.62076926e+00 5.27895927e-01
1.21786833e+00 9.95055020e-01 2.05558345e-01 4.74776357e-01
-1.03977904e-01 1.24979675e+00 -5.66385865e-01 -8.17008018e-01
1.05950302e-02 -6.02453232e-01 -1.24546528e+00 -4.60417062e-01
7.04305112e-01 -6.37877285e-01 -4.23116714e-01 9.02541935e-01
4.38933223e-01 3.93132150e-01 9.77145061e-02 -9.18367803e-01
-8.24735641e-01 7.21268833e-01 -1.17866911e-01 3.29928011e-01
8.72825801e-01 2.20942751e-01 1.76827943e+00 -1.12287831e+00
6.91021323e-01 1.12578571e+00 2.60140777e-01 4.79213148e-01
-1.01880920e+00 -4.62353289e-01 4.28202659e-01 7.74116516e-01
-8.44999075e-01 -3.05735141e-01 5.31616569e-01 1.15340874e-01
8.53231192e-01 3.56175244e-01 3.60896140e-01 6.09422505e-01
-3.27629387e-01 1.12443650e+00 1.40211332e+00 -2.42071465e-01
3.28813612e-01 -1.49838656e-01 6.48617446e-01 7.18755484e-01
4.50881273e-02 -1.24857184e-02 -8.37350786e-01 -4.12501186e-01
2.31240503e-02 -2.10694447e-01 -2.10699871e-01 -1.71576172e-01
-9.37231421e-01 9.28438783e-01 6.03707433e-01 -1.11822732e-01
-4.63861257e-01 4.22423631e-01 1.74475893e-01 6.78589642e-01
-3.88408676e-02 9.57921684e-01 -5.96598268e-01 -5.59214689e-03
-6.63249254e-01 5.43570101e-01 7.08541393e-01 5.78649223e-01
1.02032709e+00 -5.63525259e-01 -9.33033526e-01 6.80175483e-01
4.30950403e-01 3.23895395e-01 4.64621872e-01 -1.24772048e+00
6.34962142e-01 9.30606306e-01 3.61382097e-01 -7.54496515e-01
-2.22343102e-01 -4.11706179e-01 -7.23243039e-03 -1.83724195e-01
2.61788189e-01 6.67681694e-02 -5.71993470e-01 1.89306760e+00
4.41597164e-01 1.99920416e-01 -5.82942478e-02 9.93779600e-01
1.31275010e+00 4.09630507e-01 1.87348023e-01 -1.28691286e-01
1.31132829e+00 -1.22791111e+00 -5.95546067e-01 -3.01514626e-01
5.50578773e-01 -4.57057327e-01 1.29135966e+00 2.13756517e-01
-1.19359636e+00 -4.63686794e-01 -8.04310977e-01 -7.92143866e-02
-4.75711748e-03 1.76815018e-02 5.46608150e-01 2.65608191e-01
-1.05353558e+00 7.07368493e-01 -2.61878699e-01 5.61550371e-02
1.49986610e-01 4.10736173e-01 -3.94265018e-02 -3.03377479e-01
-1.64590836e+00 9.42139626e-01 1.27432346e-01 -3.29144560e-02
-8.25607896e-01 -4.48628306e-01 -6.89925730e-01 3.27606380e-01
8.57493520e-01 -9.30324197e-01 1.69015002e+00 -7.87690222e-01
-1.68846154e+00 7.44557023e-01 -5.44794083e-01 -2.93012977e-01
2.90409207e-01 -4.39408183e-01 -2.31822073e-01 3.26648265e-01
1.43398374e-01 6.03856027e-01 5.65486908e-01 -1.04360330e+00
-8.45011771e-01 -8.50842148e-02 6.72393143e-01 2.14570537e-01
-1.97918713e-01 9.92320850e-02 -6.82742357e-01 -2.23006234e-01
7.37484321e-02 -5.24663687e-01 -3.52288693e-01 -2.34615564e-01
-1.61332339e-01 -8.87961745e-01 1.04020178e-01 -4.02999282e-01
1.57709527e+00 -1.72184968e+00 -2.12984271e-02 2.04936191e-01
3.73740703e-01 2.67879426e-01 -6.36485517e-01 4.79034901e-01
3.27237129e-01 2.46564429e-02 -5.89307360e-02 -9.61698443e-02
1.33543253e-01 8.21136385e-02 -4.69258219e-01 4.49711420e-02
3.80316406e-01 1.24705625e+00 -1.30126429e+00 -5.11474192e-01
-3.40692580e-01 -2.80532211e-01 -9.87319469e-01 5.54794431e-01
-6.04965210e-01 2.20882416e-01 -6.78612888e-01 3.80767494e-01
5.50514162e-01 -7.10675061e-01 1.25078931e-01 7.73488805e-02
3.99039477e-01 1.03684044e+00 -1.25381315e+00 1.38709915e+00
-3.86014640e-01 -7.45503157e-02 -4.90221679e-02 -6.83097541e-01
7.34678805e-01 8.43689591e-02 -1.43041154e-02 -1.38210833e+00
-3.28229964e-01 4.96531993e-01 -1.48534793e-02 -5.35020232e-01
7.77270317e-01 1.76041573e-01 2.59389635e-02 6.67328835e-01
1.05251350e-01 9.60812420e-02 4.28346366e-01 6.60163701e-01
1.47303677e+00 2.00628713e-01 8.08088705e-02 1.86073273e-01
8.24495316e-01 -1.07669711e-01 7.53251374e-01 1.27057850e+00
-2.44861066e-01 4.13445473e-01 5.56140363e-01 -1.23710945e-01
-3.43640268e-01 -1.06780422e+00 2.84561306e-01 1.52019215e+00
4.34156239e-01 -6.29111469e-01 -3.92796934e-01 -1.14153385e+00
9.67248157e-02 7.43860841e-01 -5.39925039e-01 -2.93753624e-01
-9.00438607e-01 -3.85109961e-01 5.55118434e-02 4.10456896e-01
2.71953583e-01 -1.37476683e+00 -5.45004070e-01 4.66095358e-01
-4.61746663e-01 -9.26235914e-01 -6.18506849e-01 1.97649464e-01
-8.56266737e-01 -1.35362303e+00 -1.83843642e-01 -6.78028226e-01
5.13824463e-01 2.93148875e-01 1.71385193e+00 7.86650717e-01
3.13623995e-01 5.95703244e-01 -5.91165245e-01 1.41954422e-01
-7.03272671e-02 8.35229307e-02 -3.40515852e-01 -1.39832139e-01
4.15315360e-01 -1.48257598e-01 -9.54446733e-01 4.15673316e-01
-6.50830328e-01 -2.88440257e-01 4.13266778e-01 1.08991218e+00
7.39043891e-01 -4.62083697e-01 1.30645263e+00 -1.21837437e+00
9.29018557e-01 -7.09560513e-01 -5.61351836e-01 5.64541459e-01
-9.18160737e-01 2.57510990e-01 7.21445322e-01 -1.58662647e-01
-1.02763832e+00 -2.56222129e-01 -2.36892939e-01 1.37809226e-02
3.45082402e-01 8.15208316e-01 8.99082571e-02 1.34925470e-01
5.09915054e-01 5.23721315e-02 -5.30456841e-01 -2.67240733e-01
5.29856622e-01 2.16483846e-01 4.57450390e-01 -7.07769752e-01
8.32707822e-01 1.93715557e-01 -4.58749831e-01 2.99401671e-01
-1.42935050e+00 -7.21849263e-01 3.64467651e-02 -2.25921676e-01
3.69788587e-01 -6.80854201e-01 -1.07590854e+00 -1.35866990e-02
-1.13541949e+00 -1.89942658e-01 -3.35458308e-01 4.37091112e-01
-4.71729845e-01 3.21760744e-01 -8.57656181e-01 -7.39713371e-01
-4.46594149e-01 -9.79867101e-01 6.45370126e-01 4.74210382e-01
-2.53309280e-01 -6.13409042e-01 2.22351313e-01 6.49716079e-01
5.15988231e-01 -4.94522482e-01 1.07440829e+00 -9.32637155e-01
-1.05823600e+00 -2.16734290e-01 -2.69486636e-01 4.81798574e-02
-2.47270465e-01 -3.50599796e-01 -9.31595504e-01 -1.53681159e-01
-1.22813284e-01 -1.05357242e+00 1.40085495e+00 -6.37088120e-02
1.30380428e+00 -3.63697946e-01 9.27376300e-02 -2.03219932e-02
1.24244535e+00 -3.95094365e-01 5.50656855e-01 3.91888559e-01
2.47359499e-01 7.36586928e-01 9.48407531e-01 4.16438669e-01
7.47809052e-01 4.99775857e-01 7.39978850e-01 2.84372658e-01
-1.10747673e-01 -5.78794122e-01 4.28117722e-01 1.00723517e+00
4.72044349e-01 -3.63229632e-01 -3.61187041e-01 6.06046557e-01
-2.07803607e+00 -8.96274149e-01 -3.25978726e-01 2.47356892e+00
1.19308221e+00 2.19642058e-01 -5.50555028e-02 -1.98567603e-02
5.53221285e-01 1.05064116e-01 -7.47607052e-01 -1.90996245e-01
-1.81750789e-01 5.91817081e-01 3.74344923e-02 6.35608554e-01
-7.89157212e-01 1.04690731e+00 6.48754120e+00 8.95708025e-01
-5.26027918e-01 8.86053666e-02 4.34172899e-01 -6.27914220e-02
-8.85541677e-01 3.02887917e-01 -7.05090165e-01 2.93146312e-01
8.21678519e-01 -1.96827035e-02 3.91361564e-01 5.99663734e-01
-1.29179865e-01 -3.75638008e-01 -1.16545951e+00 3.88009906e-01
-2.67451294e-02 -1.25862896e+00 1.22592494e-01 -6.02111518e-01
6.56315267e-01 8.23248625e-02 3.32503393e-02 8.27112317e-01
6.44289911e-01 -9.09493089e-01 6.77375853e-01 6.51396334e-01
3.54225397e-01 -7.16441095e-01 7.74345517e-01 3.22633505e-01
-1.04429567e+00 -2.11856648e-01 -3.70903522e-01 -1.64262652e-01
1.52319567e-02 2.95968235e-01 -6.07567906e-01 5.15979409e-01
5.50750196e-01 3.87934566e-01 -9.84557807e-01 1.32500780e+00
-9.44232821e-01 1.08730972e+00 -4.93262522e-02 -4.50038910e-01
1.62703350e-01 1.52508453e-01 2.50701725e-01 7.56164253e-01
-7.90476799e-04 1.17229015e-01 2.35049382e-01 8.55548441e-01
-7.34368443e-01 3.11950773e-01 1.24074571e-01 2.18773469e-01
7.53489375e-01 1.24213719e+00 -3.00913036e-01 -4.81128931e-01
-4.76451933e-01 8.79922986e-01 9.08266187e-01 3.73473734e-01
-6.14924550e-01 -3.35739136e-01 1.62530452e-01 -1.12174191e-01
3.79686534e-01 3.98123354e-01 -9.52744037e-02 -1.05845857e+00
3.98378402e-01 -9.85966921e-01 9.24868405e-01 -8.29929888e-01
-1.52802730e+00 3.60564709e-01 -5.31395137e-01 -9.09649789e-01
-7.07681850e-02 -3.48001540e-01 -8.00898850e-01 8.03871274e-01
-2.36665010e+00 -7.50714064e-01 -1.43215343e-01 4.20756072e-01
3.16679090e-01 2.18242794e-01 7.24765718e-01 3.69309455e-01
-3.54311645e-01 7.57718146e-01 -1.82575166e-01 -1.05609916e-01
1.07565796e+00 -1.40712202e+00 2.16906965e-01 5.39577007e-01
1.86155528e-01 7.72203982e-01 4.90394652e-01 -5.44029355e-01
-1.43481255e+00 -9.16700542e-01 1.31425023e+00 -6.72804654e-01
6.01284325e-01 1.26048028e-01 -1.40199709e+00 2.48830676e-01
3.70366752e-01 1.51566714e-01 7.98684239e-01 3.63088042e-01
-6.73602402e-01 -1.45980835e-01 -8.85525286e-01 3.78566384e-01
7.58868039e-01 -7.05263674e-01 -9.26327646e-01 3.45130384e-01
1.06619883e+00 -3.31702918e-01 -4.28538173e-01 4.90162045e-01
2.70997226e-01 -9.61729050e-01 8.39713931e-01 -8.93646240e-01
5.32555342e-01 -5.10168254e-01 1.36413425e-02 -1.03950930e+00
-4.24088687e-01 -4.96682078e-01 -6.21461153e-01 1.03977859e+00
7.88097322e-01 -3.69248003e-01 8.48322690e-01 6.49172366e-01
-7.38837197e-02 -9.82344210e-01 -1.07022834e+00 -4.60641682e-01
-9.74497013e-03 -1.89683139e-01 7.40888596e-01 6.93450153e-01
8.51117373e-02 5.51081300e-01 -3.02270055e-01 4.29919958e-02
2.72385746e-01 6.14163518e-01 4.79940981e-01 -1.21964061e+00
-5.96103966e-01 -4.67924535e-01 3.76984477e-01 -1.47209299e+00
2.03986064e-01 -1.06826758e+00 2.96453685e-01 -1.83177316e+00
4.47812378e-01 -5.15256464e-01 -8.61939549e-01 3.17507178e-01
-1.20854163e+00 -1.50689976e-02 6.74818754e-02 1.30539000e-01
-1.80212867e+00 8.42881620e-01 1.31441200e+00 7.56916329e-02
1.87594779e-02 8.64289179e-02 -9.84906554e-01 5.21905124e-01
5.60842693e-01 -7.83799589e-01 -1.81551516e-01 -5.13303280e-01
1.06756091e+00 2.56933421e-01 2.93004990e-01 -6.35976017e-01
6.54575586e-01 -2.68615007e-01 1.01122372e-02 -6.01724446e-01
1.98229384e-02 -4.27600235e-01 -7.85814047e-01 4.79473650e-01
-1.01164734e+00 2.83349544e-01 -2.14386329e-01 9.04864132e-01
-3.26609761e-01 -6.38820529e-01 4.99026924e-01 -1.42247993e-02
-7.39685535e-01 3.14408809e-01 -2.44207412e-01 6.43406868e-01
4.64564979e-01 2.46991694e-01 -6.07685924e-01 -5.41453004e-01
-2.24471807e-01 1.06141484e+00 1.56306684e-01 3.67302626e-01
9.32923436e-01 -1.33393788e+00 -9.19913352e-01 -3.70656461e-01
5.17662227e-01 -1.29059210e-01 2.11216941e-01 6.56324148e-01
9.25554559e-02 3.81272852e-01 3.52595896e-01 3.92513163e-02
-1.00636947e+00 5.72453916e-01 2.86199212e-01 -8.41716528e-01
-1.51701257e-01 9.92838025e-01 -4.08901088e-02 -8.21177065e-01
2.97941238e-01 -1.06501184e-01 -6.29262030e-01 5.88718988e-02
5.40946960e-01 2.82391161e-01 8.22977722e-02 -5.60704432e-02
-4.01215225e-01 1.61951050e-01 -2.11282611e-01 -1.10449307e-01
1.12909520e+00 -1.19882248e-01 -2.63066024e-01 2.78526664e-01
7.40180314e-01 3.45999181e-01 -9.36554253e-01 -8.95377994e-01
4.94511366e-01 -4.31643903e-01 -2.46280938e-01 -1.17354453e+00
-8.14360321e-01 6.14134908e-01 1.71288282e-01 -9.24943984e-02
1.02141094e+00 1.54123977e-01 9.46910620e-01 9.03038681e-01
2.86961079e-01 -1.34305990e+00 7.61426926e-01 7.85128534e-01
7.87593126e-01 -1.53583097e+00 -7.04975277e-02 -1.94953740e-01
-5.88155687e-01 9.54238713e-01 1.25979745e+00 -2.26191849e-01
1.49651393e-01 -5.13453245e-01 -1.01194806e-01 -1.79127365e-01
-1.43357718e+00 -3.76970440e-01 3.72968614e-01 9.46689695e-02
5.39242685e-01 -1.83888078e-01 -6.01965070e-01 9.09867108e-01
1.51562721e-01 -7.56260902e-02 4.20501798e-01 9.77339029e-01
-7.27199674e-01 -1.24933088e+00 -1.15146361e-01 7.46437192e-01
-5.15009642e-01 -3.63528252e-01 -4.03909117e-01 7.45592341e-02
-1.65126026e-01 1.33568394e+00 -2.94761032e-01 -4.12898302e-01
4.22435820e-01 7.95579143e-03 2.78358400e-01 -8.19655776e-01
-1.19454014e+00 -3.80637556e-01 2.38451093e-01 -7.01439679e-01
-3.81017417e-01 -3.89161617e-01 -1.49996710e+00 -1.14479259e-01
-6.23491704e-01 6.18160188e-01 1.23887435e-01 1.00588846e+00
5.78303635e-01 5.47879815e-01 7.26531923e-01 3.82930756e-01
-9.89513040e-01 -7.59644330e-01 -1.11870654e-01 5.38268924e-01
3.93329799e-01 -4.58300918e-01 -2.58832306e-01 -6.76548123e-01] | [11.241872787475586, 8.017742156982422] |
2b4d7038-76f4-4971-9ade-7994116e1c86 | high-temporal-resolution-event-based-vehicle | 2212.14289 | null | https://arxiv.org/abs/2212.14289v2 | https://arxiv.org/pdf/2212.14289v2.pdf | High-temporal-resolution event-based vehicle detection and tracking | Event-based vision has been rapidly growing in recent years justified by the unique characteristics it presents such as its high temporal resolutions (~1us), high dynamic range (>120dB), and output latency of only a few microseconds. This work further explores a hybrid, multi-modal, approach for object detection and tracking that leverages state-of-the-art frame-based detectors complemented by hand-crafted event-based methods to improve the overall tracking performance with minimal computational overhead. The methods presented include event-based bounding box (BB) refinement that improves the precision of the resulting BBs, as well as a continuous event-based object detection method, to recover missed detections and generate inter-frame detections that enable a high-temporal-resolution tracking output. The advantages of these methods are quantitatively verified by an ablation study using the higher order tracking accuracy (HOTA) metric. Results show significant performance gains resembled by an improvement in the HOTA from 56.6%, using only frames, to 64.1% and 64.9%, for the event and edge-based mask configurations combined with the two methods proposed, at the baseline framerate of 24Hz. Likewise, incorporating these methods with the same configurations has improved HOTA from 52.5% to 63.1%, and from 51.3% to 60.2% at the high-temporal-resolution tracking rate of 384Hz. Finally, a validation experiment is conducted to analyze the real-world single-object tracking performance using high-speed LiDAR. Empirical evidence shows that our approaches provide significant advantages compared to using frame-based object detectors at the baseline framerate of 24Hz and higher tracking rates of up to 500Hz. | ['Samir Rawashdeh', 'Zaid El-Shair'] | 2022-12-29 | null | null | null | null | ['event-based-vision'] | ['computer-vision'] | [ 1.84070110e-01 -4.49412853e-01 1.51090175e-01 1.83498207e-02
-9.71018016e-01 -4.61052895e-01 6.27158344e-01 7.36134499e-02
-7.48278975e-01 5.93696773e-01 -3.28714699e-01 -2.08453834e-01
-2.50480790e-02 -5.64071000e-01 -6.09651983e-01 -3.73557180e-01
-2.68473655e-01 1.90003216e-01 1.12891757e+00 1.83531061e-01
-1.99517682e-02 9.15248275e-01 -2.09937835e+00 -9.54835862e-02
5.09850562e-01 1.19588900e+00 5.23271449e-02 8.59334707e-01
-3.26483846e-02 4.79383469e-01 -8.66601229e-01 -1.14375539e-01
5.00411749e-01 1.00074224e-01 1.49218300e-02 -1.23527385e-01
8.52042794e-01 -7.82464921e-01 -1.50137573e-01 5.49474895e-01
4.99240518e-01 -1.70286775e-01 1.37498334e-01 -1.34920967e+00
6.05306365e-02 -1.17407918e-01 -9.69139338e-01 4.87228096e-01
3.56033415e-01 6.11136973e-01 5.35344124e-01 -7.47499466e-01
5.37135363e-01 1.14764738e+00 9.72114742e-01 4.67758626e-01
-1.52755857e+00 -1.16185105e+00 -1.66430503e-01 -7.81803653e-02
-1.47008479e+00 -4.22327489e-01 1.31584108e-01 -6.13337755e-01
1.06868863e+00 1.03821270e-01 7.79823661e-01 6.07795119e-01
3.64842117e-01 2.13151231e-01 1.03954029e+00 -3.01363379e-01
8.90393406e-02 -4.62031085e-03 7.72723109e-02 5.14964044e-01
7.92159498e-01 7.12020695e-01 -6.04745269e-01 -7.29909018e-02
9.01220322e-01 -3.07528321e-02 -8.66157636e-02 -1.00553013e-01
-1.16997695e+00 4.94637966e-01 2.71617144e-01 -1.50343021e-02
-3.85559946e-01 4.45687473e-01 3.51707548e-01 -2.66726702e-01
1.59734353e-01 1.37713343e-01 -5.60387708e-02 -2.40388960e-01
-1.36798024e+00 3.78143132e-01 3.95833284e-01 1.13748550e+00
5.76601982e-01 3.98709208e-01 -4.95145142e-01 -2.17245799e-03
5.17544568e-01 9.28897321e-01 -8.24301168e-02 -9.50717866e-01
1.95526823e-01 4.05748099e-01 5.16001403e-01 -5.44713616e-01
-4.42504525e-01 -3.88285130e-01 -2.14294434e-01 9.41621780e-01
5.69622397e-01 -1.51471466e-01 -1.10542929e+00 1.49559891e+00
6.80344403e-01 3.85440141e-01 -1.87583447e-01 9.19872642e-01
4.30660218e-01 4.95918393e-01 4.41110015e-01 -3.55113804e-01
1.59752476e+00 -3.88101846e-01 -5.78319311e-01 -1.72384530e-02
4.87064570e-02 -9.43567455e-01 7.11952150e-01 2.68380612e-01
-9.64988530e-01 -7.90166795e-01 -1.23074520e+00 3.25622857e-01
4.48970646e-02 8.40690136e-02 4.18106526e-01 8.73937249e-01
-1.00420475e+00 2.30955362e-01 -1.12652731e+00 -5.61300993e-01
4.18583006e-01 5.45583665e-01 1.92777142e-01 1.94221109e-01
-5.33864081e-01 8.30032170e-01 2.68121302e-01 -4.14318949e-01
-9.16630030e-01 -1.16851962e+00 -4.03376311e-01 -1.04083121e-02
3.65402222e-01 -6.95755661e-01 1.43839073e+00 -2.49695316e-01
-1.25863039e+00 4.28288549e-01 -2.14140266e-01 -9.08640087e-01
6.76991045e-01 -5.64052105e-01 -4.88283604e-01 3.40684026e-01
1.63638622e-01 8.60017955e-01 7.42915154e-01 -1.05220568e+00
-1.15090537e+00 -1.87826693e-01 -1.62534282e-01 -2.16255724e-01
-1.92971900e-01 2.47113571e-01 -4.81993765e-01 -4.20315832e-01
-1.95117623e-01 -8.42878103e-01 -1.26731724e-01 3.95110309e-01
2.77066119e-02 1.16295837e-01 1.38264394e+00 -2.84220695e-01
1.13614571e+00 -2.09596086e+00 -6.98038042e-01 -1.78875819e-01
1.17203325e-01 5.27497470e-01 2.69780159e-01 2.26927474e-01
3.61207217e-01 -3.08239490e-01 7.99006447e-02 -3.47922325e-01
-3.02804083e-01 -9.79530588e-02 -3.05507451e-01 3.48879427e-01
3.91364604e-01 5.74037015e-01 -7.45197117e-01 -4.00179505e-01
8.16943765e-01 7.77171731e-01 -2.17987716e-01 -4.95045185e-02
-4.16135453e-02 4.38882470e-01 -2.69538999e-01 8.30836236e-01
7.25815654e-01 -1.44371733e-01 -2.61870027e-02 -4.04053330e-01
-8.35285783e-01 -7.03444853e-02 -1.54809153e+00 1.23030555e+00
-1.84917390e-01 7.26300895e-01 2.06400722e-01 -1.69757346e-03
8.76863301e-01 1.75634637e-01 7.37399876e-01 -7.40200937e-01
2.04715073e-01 1.16242222e-01 -6.01958372e-02 -5.54816164e-02
7.17401922e-01 -1.14113376e-01 1.64305791e-01 9.02245715e-02
-1.05413221e-01 2.00503916e-01 3.42804879e-01 2.95112789e-01
1.30421865e+00 2.54058480e-01 1.82330087e-01 -2.49137744e-01
2.28193507e-01 3.49981129e-01 4.82354075e-01 8.44839096e-01
-4.28650409e-01 3.58235180e-01 -3.32527161e-01 -1.61619678e-01
-1.04353034e+00 -1.33599710e+00 -4.49236095e-01 5.39545476e-01
4.16777790e-01 -5.87309301e-01 -5.05661905e-01 -1.00804597e-01
3.66833866e-01 4.05903429e-01 -1.82542846e-01 2.68227104e-02
-5.66795111e-01 -5.29758871e-01 6.67364419e-01 8.57738078e-01
6.71607852e-01 -4.79315579e-01 -1.61038101e+00 4.93749529e-01
3.01315069e-01 -1.50555146e+00 -5.13500087e-02 -4.61680628e-02
-9.56062138e-01 -1.01023114e+00 -3.83070379e-01 -8.67970064e-02
2.87783176e-01 4.67826158e-01 8.44956040e-01 -7.82187805e-02
-9.29262519e-01 5.55426359e-01 -1.50941938e-01 -7.09070444e-01
-3.20595242e-02 -5.92161596e-01 1.61440596e-01 -7.49357194e-02
2.61456102e-01 -2.30505645e-01 -6.78573072e-01 3.04141253e-01
-7.02348530e-01 3.92387472e-02 6.25684142e-01 4.62088943e-01
6.20115221e-01 -2.33513400e-01 3.67700219e-01 -2.06896976e-01
-1.22164130e-01 -2.53653657e-02 -1.31701076e+00 -1.60718948e-01
-8.21233749e-01 -2.65130013e-01 2.88540483e-01 -7.53267109e-01
-1.05727756e+00 4.54882413e-01 2.60868281e-01 -6.56240404e-01
-1.87751994e-01 -2.66331375e-01 3.56325448e-01 -3.59457850e-01
7.79395044e-01 -1.50894165e-01 5.47341108e-02 -2.58265227e-01
3.17668706e-01 4.21076387e-01 7.93675482e-01 -3.57554257e-01
1.00262034e+00 1.01827276e+00 1.65039048e-01 -7.51739860e-01
-4.20508027e-01 -5.81429422e-01 -3.43841374e-01 -6.69987857e-01
8.56454492e-01 -1.23339438e+00 -9.44286346e-01 3.33874851e-01
-9.43895161e-01 -7.77183101e-02 -4.56940860e-01 8.44500303e-01
-2.31627837e-01 1.82868093e-01 -5.15836477e-01 -1.37791407e+00
-4.12285626e-01 -9.07517672e-01 1.27776802e+00 6.37260437e-01
-1.82830319e-02 -1.49447724e-01 -3.03478628e-01 1.46628901e-01
6.01439416e-01 5.48905551e-01 6.89605698e-02 -1.09086506e-01
-1.27555668e+00 -1.54865921e-01 -5.79764128e-01 -2.00836837e-01
-1.80951968e-01 2.96515077e-01 -9.56617296e-01 -4.65280890e-01
-3.22540313e-01 3.17895144e-01 6.72925770e-01 6.37730598e-01
3.95355314e-01 3.83904427e-01 -7.68270552e-01 2.75575846e-01
1.75290763e+00 5.69128633e-01 5.02112389e-01 4.34622586e-01
4.34875667e-01 6.08038567e-02 1.04508770e+00 6.75484598e-01
1.72374114e-01 1.01216233e+00 6.44981146e-01 7.51840696e-02
-6.49488986e-01 3.23215909e-02 3.31849515e-01 2.93257758e-02
-1.35687768e-01 4.74417061e-02 -9.77829754e-01 5.42986929e-01
-1.78263700e+00 -1.11651123e+00 -5.05184054e-01 2.58419371e+00
3.99402708e-01 5.68245530e-01 4.67385769e-01 6.22098446e-02
1.00679827e+00 -2.14018434e-01 -5.70829809e-01 1.06774837e-01
2.75510818e-01 4.09956753e-01 9.84709144e-01 2.71883965e-01
-1.09821594e+00 7.73409784e-01 6.12558889e+00 5.77109516e-01
-1.24380565e+00 1.70983687e-01 -1.03231497e-01 -6.44112766e-01
4.11917239e-01 6.96390048e-02 -1.48623848e+00 5.53100407e-01
1.31725454e+00 -2.34549433e-01 1.05271123e-01 7.37465680e-01
3.85697722e-01 -5.34399748e-01 -7.98424482e-01 1.04063368e+00
-2.03646749e-01 -1.42510176e+00 -2.51674384e-01 2.20394045e-01
3.18876833e-01 8.75456333e-02 -3.12481225e-01 3.02392304e-01
1.18718266e-01 -5.21061897e-01 1.04901707e+00 4.17546779e-01
9.78885472e-01 -6.93173051e-01 4.12914455e-01 2.40815029e-01
-1.84859955e+00 -7.58996904e-02 -9.07622874e-02 -2.24635512e-01
5.94258010e-01 7.09661901e-01 -9.31388319e-01 5.47282159e-01
1.10228503e+00 2.50850558e-01 -5.27111530e-01 1.41024327e+00
1.46327317e-01 5.56338072e-01 -8.63721490e-01 -1.28025478e-02
-6.41989484e-02 3.63949388e-01 8.71782064e-01 1.38797045e+00
3.30058277e-01 6.87751472e-02 3.48399371e-01 7.74726331e-01
3.38772863e-01 -4.01128501e-01 -3.73706222e-01 3.63110691e-01
7.98465550e-01 1.45854497e+00 -8.20402682e-01 -3.36265683e-01
-5.91762364e-01 3.18975478e-01 -1.90172926e-01 4.28324491e-02
-1.40376961e+00 -4.59977418e-01 7.19011903e-01 6.23487711e-01
7.13372171e-01 -3.89621288e-01 -2.20209077e-01 -5.66006124e-01
9.08503383e-02 -3.38226646e-01 2.91222364e-01 -6.74929678e-01
-9.18958247e-01 4.79951382e-01 3.30414891e-01 -1.49408615e+00
-1.84701860e-01 -4.62653726e-01 -3.79910678e-01 8.34365904e-01
-1.46118724e+00 -1.22452402e+00 -6.13566399e-01 4.76880521e-01
4.26698506e-01 1.99334726e-01 4.49236035e-01 8.19106877e-01
-4.28981096e-01 4.24673855e-01 -1.73742279e-01 -6.12976439e-02
6.14417493e-01 -8.88250887e-01 2.95907915e-01 1.26216614e+00
-1.64012492e-01 4.94699776e-01 7.15647936e-01 -8.39467466e-01
-1.59592354e+00 -1.33752275e+00 4.28603083e-01 -4.24880981e-01
5.41856050e-01 -2.13504359e-01 -7.01065481e-01 4.51195598e-01
-1.82070255e-01 3.46464723e-01 1.97944403e-01 -1.51254430e-01
-1.82633698e-01 -3.62083107e-01 -1.30988467e+00 5.39445579e-01
9.76090670e-01 -2.29879275e-01 -3.23673189e-01 -1.34918302e-01
5.99602103e-01 -6.23284340e-01 -7.53614366e-01 7.78671324e-01
8.38854074e-01 -1.05590212e+00 1.11705530e+00 4.31103669e-02
-3.93864900e-01 -9.95660007e-01 -5.81411943e-02 -3.11464071e-01
-3.48690331e-01 -4.88123298e-01 -4.60938722e-01 1.47601247e+00
-2.46285386e-02 -6.57892942e-01 7.28591144e-01 4.86891240e-01
-1.33729428e-01 -4.75466907e-01 -1.20177424e+00 -1.26748776e+00
-7.90862143e-01 -4.60980594e-01 1.81404978e-01 2.45153472e-01
-6.59278631e-01 1.91645920e-01 -1.16866969e-01 4.67082977e-01
8.99397254e-01 1.25278577e-01 9.37375009e-01 -1.37489784e+00
-9.25509483e-02 -1.68981448e-01 -8.51709604e-01 -7.08088696e-01
-6.44856036e-01 -2.74075866e-01 1.24962524e-01 -1.34015071e+00
1.03351183e-01 -3.65291804e-01 -1.62860483e-01 3.06979686e-01
-3.10074612e-02 5.61835825e-01 5.35018623e-01 2.49351725e-01
-4.34642017e-01 7.82681406e-02 5.97708881e-01 2.86739141e-01
-1.73635602e-01 -1.93216190e-01 -2.61262983e-01 7.07745373e-01
5.38022280e-01 -6.01870060e-01 -1.17585704e-01 -2.04933256e-01
-4.98185188e-01 3.54716927e-02 8.58372271e-01 -1.79324687e+00
4.60035503e-01 9.86382738e-02 6.15131795e-01 -9.58095491e-01
6.66802824e-01 -1.02000713e+00 7.09085524e-01 8.79365206e-01
2.84324765e-01 6.39570877e-02 8.57586920e-01 7.47372866e-01
1.30754054e-01 1.76872164e-01 1.02863932e+00 3.03229213e-01
-9.84077454e-01 8.06471780e-02 -3.43508631e-01 -1.77967876e-01
1.61509502e+00 -7.39459455e-01 -5.39832294e-01 -6.19317740e-02
-3.74849975e-01 -3.21448818e-02 4.40369695e-01 2.39956126e-01
3.64082426e-01 -1.19092834e+00 -4.89822119e-01 4.53339405e-02
2.96352338e-02 -2.79928982e-01 4.79736328e-02 9.45956647e-01
-3.30865204e-01 4.10737216e-01 -4.12146062e-01 -1.24645460e+00
-1.64103115e+00 2.59187251e-01 4.75417040e-02 -1.12436384e-01
-8.97148132e-01 6.37523592e-01 -1.98600948e-01 5.85032403e-01
2.93514520e-01 -3.59620750e-01 3.14561307e-01 -1.00033112e-01
7.41416931e-01 6.81868792e-01 7.88490474e-02 -4.66426104e-01
-7.40466475e-01 6.76158726e-01 1.26666740e-01 -2.93838114e-01
8.34349871e-01 9.73427743e-02 6.51066720e-01 1.81714386e-01
4.41146553e-01 1.97358787e-01 -1.70932508e+00 1.27497353e-02
6.57754913e-02 -6.62967324e-01 2.10197255e-01 -8.67732644e-01
-7.86192358e-01 4.87363994e-01 1.11302388e+00 1.07026771e-01
1.21656334e+00 -2.71538012e-02 7.72764564e-01 -1.39308304e-01
7.91808307e-01 -6.45592690e-01 -5.57212085e-02 3.16933453e-01
3.17503572e-01 -9.52735543e-01 3.16227823e-01 -4.53351378e-01
-1.97663307e-01 9.26666856e-01 8.47922266e-01 -1.44117102e-01
2.28163898e-01 9.67782974e-01 -5.77124692e-02 -6.10292815e-02
-6.53372049e-01 -7.11758435e-01 3.74956690e-02 5.84247172e-01
8.52113292e-02 -6.49028048e-02 -3.06284547e-01 2.16503292e-01
1.72205776e-01 1.70497671e-01 2.91575402e-01 1.25844657e+00
-6.63411319e-01 -8.12958121e-01 -7.46677935e-01 4.90382940e-01
-3.37199301e-01 1.91099778e-01 1.71129391e-01 1.20529783e+00
2.13192120e-01 1.19092000e+00 4.15128201e-01 -3.96062076e-01
5.53968847e-01 -8.08184668e-02 5.31620979e-01 -5.18673182e-01
-7.70155728e-01 3.62045854e-01 8.44150111e-02 -9.43750203e-01
-6.24013662e-01 -8.42899501e-01 -1.62916827e+00 -3.59384358e-01
-6.22984529e-01 -2.84459859e-01 8.51912081e-01 6.44340932e-01
7.61362076e-01 7.32239127e-01 1.54072903e-02 -1.03809428e+00
-3.04699481e-01 -5.23127139e-01 -1.61145926e-01 6.42322302e-02
3.94660592e-01 -9.85204458e-01 -1.42328173e-01 1.61983669e-01] | [6.612081527709961, -2.0080997943878174] |
7d62803d-bb77-49fb-a56c-7c8a187dd381 | time-aware-graph-structure-learning-via | 2306.07699 | null | https://arxiv.org/abs/2306.07699v1 | https://arxiv.org/pdf/2306.07699v1.pdf | Time-aware Graph Structure Learning via Sequence Prediction on Temporal Graphs | Temporal Graph Learning, which aims to model the time-evolving nature of graphs, has gained increasing attention and achieved remarkable performance recently. However, in reality, graph structures are often incomplete and noisy, which hinders temporal graph networks (TGNs) from learning informative representations. Graph contrastive learning uses data augmentation to generate plausible variations of existing data and learn robust representations. However, rule-based augmentation approaches may be suboptimal as they lack learnability and fail to leverage rich information from downstream tasks. To address these issues, we propose a Time-aware Graph Structure Learning (TGSL) approach via sequence prediction on temporal graphs, which learns better graph structures for downstream tasks through adding potential temporal edges. In particular, it predicts time-aware context embedding based on previously observed interactions and uses the Gumble-Top-K to select the closest candidate edges to this context embedding. Additionally, several candidate sampling strategies are proposed to ensure both efficiency and diversity. Furthermore, we jointly learn the graph structure and TGNs in an end-to-end manner and perform inference on the refined graph. Extensive experiments on temporal link prediction benchmarks demonstrate that TGSL yields significant gains for the popular TGNs such as TGAT and GraphMixer, and it outperforms other contrastive learning methods on temporal graphs. We will release the code in the future. | ['Jing Bai', 'Xi Xiao', 'Xueting Han', 'Haozhen Zhang'] | 2023-06-13 | null | null | null | null | ['graph-structure-learning', 'link-prediction'] | ['graphs', 'graphs'] | [ 1.97497323e-01 3.04755241e-01 -6.32033050e-01 -2.08543792e-01
-3.59662235e-01 -5.33604145e-01 6.56199872e-01 4.11016017e-01
2.18067080e-01 6.67640924e-01 4.54168111e-01 -3.80668312e-01
-5.72842002e-01 -8.36113393e-01 -6.46968782e-01 -6.69255972e-01
-7.55507648e-01 3.71501803e-01 4.11119848e-01 -1.75582290e-01
-2.30917171e-01 2.90292501e-01 -9.36433375e-01 -1.50135867e-02
9.82521892e-01 6.48942709e-01 1.59949601e-01 5.72787166e-01
-3.75094749e-02 9.91345882e-01 -1.71785522e-02 -4.52579886e-01
2.86541820e-01 -3.51992905e-01 -6.56783044e-01 1.28193051e-01
2.55456001e-01 -3.61797586e-02 -1.09624875e+00 6.80897176e-01
2.04772606e-01 3.14951837e-01 3.42268705e-01 -1.46525121e+00
-6.23960793e-01 8.93711388e-01 -5.99778891e-01 4.85316336e-01
1.63081840e-01 3.80284637e-01 1.42260671e+00 -5.36114991e-01
7.43472099e-01 1.30775392e+00 7.38236189e-01 3.20573032e-01
-1.47322690e+00 -6.74096107e-01 9.80859995e-01 2.92132944e-01
-1.17799532e+00 -7.38300607e-02 1.15546799e+00 -1.93631321e-01
1.11073399e+00 6.98166341e-02 1.02857411e+00 1.33514333e+00
7.78319314e-02 8.31767559e-01 6.68157160e-01 -7.06851333e-02
-1.02696426e-01 -6.28888965e-01 1.27770185e-01 1.04504371e+00
1.68479294e-01 3.12695593e-01 -6.84092760e-01 -2.73060828e-01
8.75977278e-01 2.27904648e-01 -2.16438174e-01 -5.89995861e-01
-1.20880008e+00 7.31505752e-01 8.87669504e-01 2.31344283e-01
-4.57504511e-01 4.78639394e-01 5.18166661e-01 2.53228337e-01
7.13164389e-01 1.92470729e-01 -5.47336876e-01 -3.41664515e-02
-5.09310007e-01 6.08775057e-02 4.69539165e-01 1.05294061e+00
7.48374045e-01 1.20372295e-01 -2.44637489e-01 6.24235451e-01
2.95203924e-01 2.97785513e-02 1.41767815e-01 -5.93875051e-01
8.31069171e-01 8.66676152e-01 -4.48091596e-01 -1.18750370e+00
-4.28794712e-01 -6.76679373e-01 -9.74871993e-01 -5.25337338e-01
3.23366940e-01 7.87983835e-02 -1.19786382e+00 1.94606102e+00
3.62319946e-01 8.92721295e-01 -2.44712368e-01 6.82165265e-01
6.04602575e-01 7.64903605e-01 2.67201334e-01 -3.42715025e-01
8.20799768e-01 -1.04538560e+00 -6.82451844e-01 -5.13548911e-01
8.72177899e-01 -1.32092923e-01 9.32697773e-01 2.68798769e-02
-7.12368727e-01 -3.12340707e-01 -9.63685930e-01 7.36033395e-02
-2.00144380e-01 -3.09695721e-01 1.01542437e+00 1.28380403e-01
-9.94606197e-01 9.90991235e-01 -1.27703810e+00 -3.94644707e-01
3.47545922e-01 2.42935732e-01 -3.24833602e-01 -2.73282290e-01
-1.18414855e+00 3.33578974e-01 5.90591908e-01 2.77795553e-01
-8.76420557e-01 -7.40535557e-01 -1.19105935e+00 1.06024422e-01
8.20668936e-01 -7.60100424e-01 7.98356712e-01 -5.95329046e-01
-1.16788161e+00 2.64252841e-01 -2.02222690e-01 -7.34356284e-01
3.51417691e-01 -6.21775922e-04 -5.15110195e-01 1.40419811e-01
1.32607535e-01 3.18950266e-01 8.15730631e-01 -9.02613282e-01
-2.89696246e-01 -2.09047124e-01 3.90526056e-02 -2.65165378e-04
-3.56457651e-01 -5.05734563e-01 -9.46271658e-01 -9.26526666e-01
2.13603556e-01 -1.10923278e+00 -5.96780896e-01 -1.63191736e-01
-5.62210500e-01 -2.98275590e-01 8.17647576e-01 -6.66319430e-01
1.65363801e+00 -1.91638303e+00 3.51678401e-01 4.89190817e-01
6.20209217e-01 9.09202769e-02 -4.71390635e-01 7.72275209e-01
-2.61965036e-01 6.87583014e-02 -3.41267616e-01 -2.09392503e-01
-1.82143360e-01 3.88460487e-01 -1.28808588e-01 3.10654402e-01
2.12913781e-01 1.23131180e+00 -1.38906729e+00 -4.78799641e-01
1.22257188e-01 3.45911384e-01 -4.75912750e-01 1.18126966e-01
-5.72094977e-01 4.02317375e-01 -6.90622687e-01 5.59167087e-01
2.51617193e-01 -7.97336996e-01 6.96514428e-01 -2.65939713e-01
3.97130013e-01 4.67707515e-01 -8.39756012e-01 1.60986173e+00
-3.80451858e-01 3.75579953e-01 -3.23390961e-01 -1.09029400e+00
8.70392740e-01 1.85663506e-01 6.96420550e-01 -7.33309090e-01
-2.33847409e-01 -2.60186940e-02 1.37295023e-01 -3.25397074e-01
1.30834669e-01 2.60928839e-01 5.37432730e-03 3.85214329e-01
4.84798029e-02 3.14807177e-01 3.82167250e-01 6.97231770e-01
1.66863489e+00 3.63730103e-01 3.81467670e-01 1.61698729e-01
2.18689397e-01 -2.25292116e-01 8.76865983e-01 3.36963058e-01
-1.57548130e-01 2.19329059e-01 7.54178941e-01 -6.01401269e-01
-6.62593901e-01 -1.01124477e+00 7.10411310e-01 9.58877921e-01
4.13057320e-02 -1.06519997e+00 -5.94665557e-02 -1.19200492e+00
1.91880479e-01 4.67362583e-01 -6.67015970e-01 -6.60134017e-01
-8.53261292e-01 -6.61243260e-01 8.34161416e-03 7.35540450e-01
1.44265607e-01 -7.59301662e-01 2.92980403e-01 5.07070124e-01
-2.31641024e-01 -1.17689967e+00 -8.46144259e-01 7.08186254e-02
-1.27699590e+00 -1.34351158e+00 -1.55096859e-01 -7.25976348e-01
7.48110890e-01 5.43039441e-01 1.12736511e+00 3.74073952e-01
1.90401841e-02 3.51383179e-01 -3.92026484e-01 2.04593509e-01
-2.51551598e-01 2.84317195e-01 -2.46598840e-01 1.10629030e-01
3.57382894e-02 -1.05083132e+00 -6.55748188e-01 1.04298070e-01
-6.58333778e-01 1.84741631e-01 5.74372351e-01 9.32340503e-01
6.95775568e-01 1.65788725e-01 6.85468614e-01 -1.17380834e+00
5.44622421e-01 -6.08882129e-01 -5.94743013e-01 2.79061645e-01
-1.07872915e+00 3.21876466e-01 7.43817329e-01 -4.36309874e-01
-7.57551134e-01 5.83027564e-02 3.16172630e-01 -8.62089217e-01
4.75967348e-01 1.14820600e+00 -4.11994457e-02 1.04343846e-01
4.56522435e-01 2.68637627e-01 5.68967499e-02 -2.75663286e-01
4.83145624e-01 -3.09279591e-01 2.84582764e-01 -5.55914760e-01
1.15717196e+00 2.70520419e-01 2.46477023e-01 -5.65569401e-01
-8.47769916e-01 -4.37121361e-01 -4.73295003e-01 -3.09399396e-01
2.44352236e-01 -8.56796980e-01 -5.71060956e-01 1.28175959e-01
-7.21310019e-01 -6.67064548e-01 -2.57297102e-02 4.59291548e-01
-4.67056334e-01 5.66939831e-01 -7.52886295e-01 -7.35329986e-01
-2.80518830e-01 -7.22668350e-01 1.01227427e+00 -2.36081146e-02
-1.48299783e-01 -1.25876975e+00 1.00313183e-02 1.67887405e-01
6.35951608e-02 5.98818481e-01 1.14960730e+00 -5.40725172e-01
-1.10601091e+00 -2.31929064e-01 -1.76990569e-01 -1.63587362e-01
4.64427412e-01 5.68695366e-02 -5.21192193e-01 -4.74696904e-01
-8.03547084e-01 -8.89030099e-02 1.13806736e+00 4.38165337e-01
1.21585202e+00 -5.81235290e-01 -8.36256862e-01 5.16937435e-01
1.25214219e+00 -7.30921775e-02 2.68952549e-01 -2.34688818e-02
1.22060180e+00 6.25544131e-01 8.43400717e-01 3.22455704e-01
6.63570404e-01 5.89763761e-01 6.29446626e-01 1.53420329e-01
-1.92888558e-01 -8.39934289e-01 2.96857387e-01 9.38057721e-01
-1.03604138e-01 -3.87417227e-01 -9.19818997e-01 7.09981859e-01
-2.30346608e+00 -9.66074765e-01 -1.21049263e-01 2.10371256e+00
6.76113904e-01 4.24142778e-01 3.82981807e-01 -3.90862813e-03
7.57918060e-01 6.16790414e-01 -7.96430230e-01 2.24929854e-01
1.09454997e-01 3.40628363e-02 5.02796292e-01 4.91005510e-01
-9.74213660e-01 1.22034454e+00 5.11036777e+00 7.10743487e-01
-8.93702984e-01 -1.02793276e-01 5.62597454e-01 -2.84835417e-02
-5.98047137e-01 4.04059112e-01 -3.61321926e-01 3.62736970e-01
1.05936599e+00 -5.20873129e-01 6.78508699e-01 5.91782570e-01
3.49468201e-01 5.53454936e-01 -1.12865770e+00 7.71518111e-01
-3.82759541e-01 -1.50085366e+00 5.13794981e-02 1.08804159e-01
7.67287195e-01 2.17767909e-01 -1.70852810e-01 6.92443073e-01
7.38898456e-01 -7.75503099e-01 2.75954068e-01 5.49238980e-01
2.89162189e-01 -7.57324636e-01 2.34524831e-01 3.40454489e-01
-1.99956083e+00 -1.98468655e-01 -8.80344957e-02 2.01805115e-01
1.70836717e-01 6.52946472e-01 -1.19119489e+00 9.42830265e-01
4.70165163e-01 1.41258085e+00 -6.92407072e-01 8.94931436e-01
-3.32165480e-01 8.53849292e-01 -2.32820719e-01 -5.88564388e-02
3.84566456e-01 -3.55830312e-01 6.48201406e-01 8.47246885e-01
1.74420193e-01 -9.57915261e-02 6.18244052e-01 7.01902330e-01
-2.66802043e-01 -8.07659328e-02 -9.24136996e-01 -7.06047177e-01
7.61465788e-01 1.21101177e+00 -7.78374135e-01 -1.04312219e-01
-5.37087262e-01 8.01548898e-01 7.99676716e-01 5.60922742e-01
-8.47386420e-01 7.44524226e-03 6.40313983e-01 1.73068911e-01
2.82330036e-01 -5.28034627e-01 3.13124478e-01 -1.01933575e+00
1.68316111e-01 -8.35460007e-01 1.06921983e+00 -3.40492368e-01
-1.59033775e+00 4.95690584e-01 1.08289182e-01 -1.31793225e+00
-4.07741398e-01 -2.71757871e-01 -6.96345508e-01 5.20096362e-01
-1.43783200e+00 -1.48430824e+00 -2.27722198e-01 6.28653824e-01
4.42654431e-01 1.29876211e-01 4.03366953e-01 9.06881168e-02
-6.99273705e-01 5.28407037e-01 -3.54814172e-01 1.51748165e-01
3.85111243e-01 -1.39292228e+00 8.02341402e-01 1.03287160e+00
2.65517652e-01 6.25863850e-01 5.96437454e-01 -1.15894151e+00
-1.71019483e+00 -1.71203983e+00 7.19076276e-01 -1.36090130e-01
1.11409986e+00 -3.65473181e-01 -1.12303102e+00 1.07780194e+00
-3.78741398e-02 5.69195986e-01 3.99546415e-01 4.55924004e-01
-5.18615007e-01 -3.36959839e-01 -5.59404254e-01 9.58945334e-01
1.75001359e+00 -6.79072082e-01 -3.09364777e-03 6.08167827e-01
1.09364426e+00 -2.91580707e-01 -1.06997705e+00 4.86052334e-01
1.85350090e-01 -5.14576674e-01 8.99090111e-01 -7.17250943e-01
3.91540863e-02 -2.73298323e-01 2.43871883e-01 -1.42696249e+00
-6.16885185e-01 -1.12169039e+00 -7.80241787e-01 1.32251477e+00
5.89638770e-01 -7.54522204e-01 1.11737216e+00 3.48953575e-01
-1.97589308e-01 -9.33857143e-01 -7.17785001e-01 -1.09006763e+00
-3.97831559e-01 -3.55263323e-01 6.81433916e-01 1.11572719e+00
1.49019733e-02 6.27136707e-01 -5.61670303e-01 3.95857781e-01
6.71517193e-01 4.43674922e-01 8.02051127e-01 -1.27486563e+00
-5.16723990e-01 -4.56083894e-01 -4.24356610e-01 -9.34908092e-01
3.91219407e-01 -1.35555017e+00 -1.30194128e-01 -1.78983462e+00
2.57343538e-02 -3.33208680e-01 -4.36200351e-01 6.38262749e-01
-5.95373631e-01 -3.83681267e-01 -1.03207231e-01 -2.32237671e-02
-6.82077050e-01 9.74394023e-01 1.35942042e+00 -3.75723481e-01
-3.84459645e-01 -2.35343762e-02 -6.43898726e-01 3.56587678e-01
7.61070907e-01 -3.26739430e-01 -1.07178938e+00 -1.65958092e-01
2.85277218e-01 3.30058396e-01 3.00359070e-01 -5.68274379e-01
1.69188246e-01 -2.18736008e-01 1.26332074e-01 -6.41483784e-01
2.99727440e-01 -7.75938213e-01 2.68489063e-01 5.30803561e-01
-3.04267526e-01 4.78359461e-02 -1.80355404e-02 1.47075284e+00
-9.65459272e-02 5.28493226e-01 3.34231228e-01 -1.49845500e-02
-7.92487323e-01 1.25364459e+00 7.26780295e-02 -1.02914008e-03
9.43981349e-01 -1.15089685e-01 -4.25598949e-01 -4.89918292e-01
-7.84586728e-01 7.40247250e-01 2.16900393e-01 6.37003243e-01
6.58238828e-01 -1.54266214e+00 -5.47157228e-01 -1.65599272e-01
4.49855596e-01 1.07670262e-01 2.70695657e-01 9.92711008e-01
-8.96849111e-02 1.63605332e-01 3.26370806e-01 -5.34562647e-01
-1.11806178e+00 9.16744947e-01 2.33046040e-01 -7.29444921e-01
-1.03624773e+00 7.21343994e-01 1.61997095e-01 -3.25774521e-01
1.34749904e-01 -3.89819860e-01 -1.98876575e-01 -1.48612797e-01
6.36227578e-02 2.11254001e-01 -3.77108343e-02 -2.06565514e-01
-3.37966204e-01 2.18731388e-01 -2.96938717e-01 3.34951133e-01
1.44792628e+00 -9.20716226e-02 5.84238023e-02 3.63531470e-01
1.12987936e+00 -7.93079436e-02 -1.54119694e+00 -7.39981532e-01
4.95518386e-01 -5.00546157e-01 1.01981536e-02 -3.51536453e-01
-1.35405660e+00 4.22331899e-01 4.44739647e-02 4.54064548e-01
1.19009984e+00 1.48765311e-01 9.69032586e-01 3.56622070e-01
3.06921005e-01 -7.30501711e-01 4.32061404e-01 3.44826669e-01
9.14502025e-01 -1.18196797e+00 5.80422059e-02 -8.92575860e-01
-4.56788719e-01 9.87392783e-01 8.47400069e-01 -1.56827033e-01
7.22444534e-01 -1.78435355e-01 -5.11610210e-01 -4.94104594e-01
-1.46420038e+00 -4.79445666e-01 4.76507545e-01 6.26534998e-01
3.56489927e-01 1.60194740e-01 -1.17572948e-01 3.59616488e-01
1.49304807e-01 -2.90419608e-01 1.62595555e-01 8.24142218e-01
3.19530852e-02 -1.27650535e+00 3.37274313e-01 7.22149432e-01
-9.89425089e-03 -6.71189278e-02 -5.46329498e-01 9.49785233e-01
-3.87287974e-01 8.94533217e-01 -3.24373871e-01 -6.23386562e-01
1.77291185e-01 -2.25795746e-01 4.79327112e-01 -7.03518808e-01
-2.46105134e-01 1.64147809e-01 5.47056496e-01 -9.64544773e-01
-2.11343497e-01 -6.83836639e-01 -1.29723692e+00 -3.37514549e-01
-2.44737431e-01 1.86537892e-01 -5.99181019e-02 8.82718980e-01
6.62681997e-01 7.80760407e-01 7.60639191e-01 -6.57599270e-01
-2.72065341e-01 -8.97510171e-01 -4.31884021e-01 3.59589428e-01
2.90832758e-01 -8.12714338e-01 -2.34580278e-01 -1.95625097e-01] | [7.256490230560303, 6.014803409576416] |
3f456cf4-15f3-4c33-b424-e977ca8245e0 | cspn-learning-context-and-resource-aware | 1911.05377 | null | https://arxiv.org/abs/1911.05377v2 | https://arxiv.org/pdf/1911.05377v2.pdf | CSPN++: Learning Context and Resource Aware Convolutional Spatial Propagation Networks for Depth Completion | Depth Completion deals with the problem of converting a sparse depth map to a dense one, given the corresponding color image. Convolutional spatial propagation network (CSPN) is one of the state-of-the-art (SoTA) methods of depth completion, which recovers structural details of the scene. In this paper, we propose CSPN++, which further improves its effectiveness and efficiency by learning adaptive convolutional kernel sizes and the number of iterations for the propagation, thus the context and computational resources needed at each pixel could be dynamically assigned upon requests. Specifically, we formulate the learning of the two hyper-parameters as an architecture selection problem where various configurations of kernel sizes and numbers of iterations are first defined, and then a set of soft weighting parameters are trained to either properly assemble or select from the pre-defined configurations at each pixel. In our experiments, we find weighted assembling can lead to significant accuracy improvements, which we referred to as "context-aware CSPN", while weighted selection, "resource-aware CSPN" can reduce the computational resource significantly with similar or better accuracy. Besides, the resource needed for CSPN++ can be adjusted w.r.t. the computational budget automatically. Finally, to avoid the side effects of noise or inaccurate sparse depths, we embed a gated network inside CSPN++, which further improves the performance. We demonstrate the effectiveness of CSPN++on the KITTI depth completion benchmark, where it significantly improves over CSPN and other SoTA methods. | ['Ruigang Yang', 'Xinjing Cheng', 'Peng Wang', 'Chenye Guan'] | 2019-11-13 | null | null | null | null | ['stereo-lidar-fusion'] | ['computer-vision'] | [ 2.24305972e-01 -1.82507306e-01 6.82268813e-02 -2.76422203e-01
-5.30897915e-01 -2.05264062e-01 3.91527414e-01 -6.26005381e-02
-6.70011222e-01 3.90682459e-01 1.54094189e-01 -9.84223336e-02
-1.12808738e-02 -1.05107737e+00 -6.21641576e-01 -7.97620416e-01
6.60056099e-02 3.81505728e-01 5.70072353e-01 8.28252211e-02
4.76146609e-01 4.02622998e-01 -1.45550871e+00 1.82411224e-01
1.03726590e+00 1.22068477e+00 6.21793509e-01 5.53788602e-01
-3.81546855e-01 8.16772163e-01 -3.19481105e-01 -9.42888856e-02
3.31365108e-01 -1.09366760e-01 -8.20387304e-01 4.17094007e-02
3.22381616e-01 -4.81142819e-01 -4.64297414e-01 1.09162259e+00
5.30518889e-01 1.62242800e-01 3.12047660e-01 -9.95127976e-01
-3.20488065e-01 5.51307499e-01 -9.44138050e-01 4.34152433e-04
-1.43737048e-02 3.56100559e-01 8.00560415e-01 -1.06404817e+00
3.94327283e-01 1.14105082e+00 3.98162007e-01 4.72553432e-01
-1.15574014e+00 -7.68362641e-01 4.82517511e-01 1.82947010e-01
-1.41340601e+00 -3.60838771e-01 7.72873759e-01 -3.30298841e-01
8.42538059e-01 9.24781151e-03 8.27208698e-01 7.04796135e-01
-9.78377610e-02 7.24406242e-01 8.00083160e-01 -4.47731853e-01
5.51628053e-01 -2.81408906e-01 4.34243679e-02 8.40338409e-01
2.00580791e-01 -6.02196418e-02 -6.10704482e-01 9.27269086e-02
1.16861796e+00 1.18184008e-01 -4.49063361e-01 -4.66653377e-01
-1.18480027e+00 6.25781476e-01 7.21457183e-01 -6.27146959e-02
-2.96681523e-01 4.01100367e-01 3.01783234e-01 -8.57715383e-02
4.63404149e-01 3.37445676e-01 -4.28434581e-01 -1.03413807e-02
-8.94111395e-01 7.29322657e-02 3.71620178e-01 9.64729488e-01
1.32103479e+00 -1.56962499e-01 -2.93364882e-01 9.20650303e-01
3.59423347e-02 3.14289480e-01 3.22326541e-01 -1.02250445e+00
7.29438603e-01 8.71181726e-01 -4.93976846e-03 -7.87164867e-01
-4.11848068e-01 -3.34161997e-01 -1.00184107e+00 2.97513574e-01
3.35788101e-01 -2.31761888e-01 -1.25888526e+00 1.72188115e+00
3.34643453e-01 4.88498598e-01 -2.18973937e-03 1.14969969e+00
7.20477164e-01 8.63412499e-01 -7.52902180e-02 2.20057890e-02
1.21831703e+00 -1.28280365e+00 -4.44799721e-01 -6.71598792e-01
5.12888968e-01 -6.20559871e-01 1.23279798e+00 4.65434462e-01
-9.94370997e-01 -4.06879783e-01 -9.96677458e-01 -2.84520596e-01
-2.47203689e-02 3.53962451e-01 9.95131850e-01 2.79679030e-01
-1.13181698e+00 5.64676523e-01 -1.06846571e+00 -6.88174665e-02
5.51560819e-01 5.02776086e-01 -2.40087494e-01 -5.53549588e-01
-7.42127180e-01 3.52061689e-01 4.04772222e-01 3.62411410e-01
-1.09669888e+00 -8.76826882e-01 -8.18836808e-01 1.70712367e-01
5.08007348e-01 -7.66988218e-01 1.05852079e+00 -8.83174300e-01
-1.41485393e+00 6.64456964e-01 -1.93339199e-01 -1.31191522e-01
2.70619214e-01 -1.87451586e-01 1.49626592e-02 1.04684606e-01
2.48748064e-02 9.82012153e-01 7.75925159e-01 -1.13255179e+00
-7.72646904e-01 -4.42489028e-01 4.12707269e-01 3.44654262e-01
-4.96768475e-01 -2.53045321e-01 -1.27363861e+00 -5.02137363e-01
4.58200485e-01 -8.66744995e-01 -5.58312237e-01 2.30272830e-01
-3.08010519e-01 1.26134697e-02 5.22601783e-01 -4.89944100e-01
1.26667225e+00 -2.29895210e+00 4.86232072e-01 3.01452696e-01
4.98663604e-01 1.03851669e-01 -3.37136060e-01 1.02266409e-01
2.69722432e-01 -8.88976455e-02 -4.37622309e-01 -6.07916474e-01
-2.57192075e-01 3.83580089e-01 8.00733492e-02 2.74967968e-01
1.49617353e-02 5.62053859e-01 -8.16592336e-01 -3.59911114e-01
2.07103804e-01 5.57574034e-01 -9.28320289e-01 3.45852494e-01
-2.99232870e-01 3.80554110e-01 -4.60859716e-01 4.86238509e-01
8.85108471e-01 -4.76424187e-01 5.01411594e-02 -3.92859399e-01
-1.42762452e-01 2.47690737e-01 -1.23156619e+00 2.15234518e+00
-8.50711584e-01 3.67975235e-01 1.48549229e-01 -6.61730826e-01
9.11294580e-01 -1.99488446e-01 1.68905914e-01 -7.36401200e-01
3.95942666e-02 1.32910654e-01 -1.45217910e-01 -2.90866286e-01
6.20475233e-01 2.97858953e-01 9.59237814e-02 2.67894417e-01
-2.73129284e-01 -9.31812748e-02 1.55859411e-01 3.17793459e-01
1.33489907e+00 7.96799883e-02 -3.92098092e-02 -1.59023657e-01
5.47838986e-01 -6.81776404e-02 8.86044264e-01 4.87398624e-01
9.18231681e-02 7.59381413e-01 5.90244949e-01 -4.54488516e-01
-9.30322766e-01 -6.83599353e-01 1.01729810e-01 9.54139531e-01
4.68935698e-01 -3.13737541e-01 -8.84817302e-01 -5.52931964e-01
-1.52391776e-01 2.86109954e-01 -7.11564720e-01 -4.96248230e-02
-6.68217301e-01 -7.34315157e-01 1.17211133e-01 7.14918375e-01
7.86842108e-01 -1.17881179e+00 -6.31078720e-01 8.56222734e-02
-5.89306355e-02 -1.18601084e+00 -6.61558568e-01 2.87145525e-01
-9.20627952e-01 -9.66292322e-01 -7.82069981e-01 -8.13004494e-01
1.10765493e+00 4.97758090e-01 1.01225388e+00 3.05615604e-01
-1.48686707e-01 2.05683932e-02 -4.66493487e-01 5.18794023e-02
2.10316971e-01 3.92062962e-01 -3.91655505e-01 6.47266731e-02
5.53789437e-02 -7.41162479e-01 -1.03238499e+00 1.16183162e-01
-1.15689540e+00 5.34224749e-01 6.58059776e-01 6.80257797e-01
7.68997431e-01 8.29088017e-02 1.43361345e-01 -1.13953054e+00
3.06964219e-01 -2.00585932e-01 -8.75167251e-01 1.15257964e-01
-6.02246821e-01 2.91839123e-01 6.37233436e-01 -3.21813256e-01
-9.88425374e-01 2.59726793e-01 -1.67364255e-01 -6.22989953e-01
1.17780283e-01 4.89015579e-01 -4.62748796e-01 -3.01692396e-01
4.41198498e-01 1.41314045e-01 -2.91333407e-01 -4.89388257e-01
2.33199626e-01 2.32875556e-01 4.85417783e-01 -7.16833413e-01
7.08004594e-01 5.45252264e-01 -1.25812948e-01 -3.21104974e-01
-8.62076283e-01 -4.17961657e-01 -5.08078456e-01 1.07186712e-01
7.10979402e-01 -1.02985334e+00 -4.72511500e-01 7.94153929e-01
-1.13332009e+00 -8.57835352e-01 -7.72232339e-02 2.12699398e-01
-2.92776406e-01 2.82714307e-01 -7.43520021e-01 -4.00348544e-01
-4.71191317e-01 -1.33641946e+00 9.88674164e-01 3.94996345e-01
2.46735647e-01 -8.52220535e-01 -1.31718993e-01 2.02885166e-01
5.07527173e-01 1.40747093e-02 9.38665450e-01 4.97232601e-02
-9.93047774e-01 -1.94124468e-02 -6.68425381e-01 2.05692127e-01
-5.39359748e-02 -1.48448095e-01 -9.09904838e-01 -4.41749424e-01
-3.80722940e-01 -1.49411291e-01 1.04825771e+00 4.05744851e-01
1.62606323e+00 -3.09522331e-01 -2.51274347e-01 1.19424939e+00
1.77108109e+00 5.92027307e-02 1.03782010e+00 4.64184374e-01
1.00604236e+00 3.66590828e-01 5.59427261e-01 7.05488026e-01
4.03858840e-01 6.16360128e-01 8.69803309e-01 -2.81555444e-01
-1.46422863e-01 -2.78628796e-01 3.02010141e-02 6.25985444e-01
-3.04644592e-02 -3.33026759e-02 -8.17789435e-01 5.04399002e-01
-1.82829416e+00 -3.86595428e-01 1.38177708e-01 2.31080699e+00
8.75883818e-01 2.19550431e-01 -1.78242818e-01 1.30080819e-01
6.01301730e-01 3.73706341e-01 -6.99882150e-01 1.18071191e-01
8.20656419e-02 4.73503649e-01 6.49978459e-01 5.78582942e-01
-8.49223197e-01 1.15967417e+00 4.97333622e+00 9.71045077e-01
-1.24988663e+00 -4.27462049e-02 6.92020774e-01 -2.85754234e-01
-5.23925066e-01 2.99115062e-01 -8.61169040e-01 4.59661245e-01
4.24990088e-01 2.26695389e-01 7.67053246e-01 8.05739999e-01
2.00499117e-01 -3.36116880e-01 -1.05957091e+00 1.16653144e+00
-7.32629821e-02 -1.47059476e+00 2.49929633e-02 -7.66268447e-02
8.07503879e-01 5.69523573e-02 -1.30041689e-01 1.96406737e-01
3.21326375e-01 -8.19571078e-01 8.09763312e-01 3.01623166e-01
9.08955216e-01 -9.21323478e-01 6.57539129e-01 2.77984053e-01
-1.33027637e+00 -1.22116253e-01 -6.42271996e-01 4.88340855e-02
8.00129026e-02 1.04701829e+00 -2.15212211e-01 2.90361017e-01
8.09815943e-01 6.87987566e-01 -5.70144117e-01 1.19715846e+00
-5.87260246e-01 4.79315162e-01 -2.84498513e-01 5.09049594e-02
2.14315221e-01 -1.64232820e-01 -3.65716580e-04 1.07394147e+00
3.76433611e-01 2.13986218e-01 2.92777479e-01 7.09287047e-01
-1.25475109e-01 -8.41647293e-03 4.46863100e-03 4.01224792e-01
7.76727974e-01 1.38089657e+00 -7.02757359e-01 -2.46286258e-01
-3.08835715e-01 1.07579052e+00 7.19800711e-01 4.57513392e-01
-7.87191033e-01 -4.93881434e-01 6.74203992e-01 3.29025716e-01
3.90601069e-01 -1.98933899e-01 -4.34410304e-01 -1.13011348e+00
5.02205938e-02 -5.40495634e-01 2.40067676e-01 -8.71638477e-01
-9.54344034e-01 6.69920564e-01 -2.61939257e-01 -1.05115879e+00
3.43033791e-01 -4.92106974e-01 -7.32794285e-01 8.96292984e-01
-1.82241440e+00 -9.16093409e-01 -9.07255471e-01 6.51823163e-01
4.68532801e-01 2.26777270e-01 4.28320348e-01 5.20602167e-01
-1.00560760e+00 6.40958428e-01 -2.93534368e-01 1.43511087e-01
5.27957976e-01 -1.06766319e+00 4.03134704e-01 9.98598456e-01
-2.94959664e-01 3.72244447e-01 3.41705859e-01 -4.25839335e-01
-1.37412989e+00 -1.31441653e+00 5.71604967e-01 2.42431372e-01
3.97541106e-01 -3.63480479e-01 -9.19573426e-01 5.05239248e-01
-1.73442572e-01 3.53741199e-01 2.70753264e-01 6.55988557e-03
-3.80862504e-01 -5.23279130e-01 -9.41902518e-01 7.00286567e-01
1.22997749e+00 -3.31062853e-01 1.52654290e-01 2.55348533e-01
1.01594961e+00 -7.58963645e-01 -6.07817948e-01 3.71013790e-01
3.03836524e-01 -1.15417457e+00 9.17101681e-01 -2.10836399e-02
7.62511134e-01 -5.17722428e-01 -1.77749813e-01 -1.18835986e+00
-4.91702944e-01 -2.84960866e-01 -3.38884280e-03 1.09400570e+00
4.56183940e-01 -5.09959817e-01 1.09658992e+00 7.58912683e-01
-3.90650243e-01 -9.65403318e-01 -7.46304274e-01 -3.14891696e-01
-2.49305725e-01 -5.24461806e-01 8.68392229e-01 7.78836906e-01
-3.49207312e-01 1.99548170e-01 -3.34093571e-01 4.29607928e-01
5.49221694e-01 9.61113349e-02 9.24836814e-01 -9.78133738e-01
-5.20237267e-01 -3.77658844e-01 -2.34472781e-01 -1.54056430e+00
-1.13704160e-01 -5.94146729e-01 1.53669640e-01 -1.70569825e+00
2.29573324e-01 -1.02559471e+00 -2.08458975e-01 8.34428310e-01
-4.15792316e-01 1.58628091e-01 4.22807559e-02 2.59588063e-01
-5.91645658e-01 6.34407222e-01 1.37263668e+00 -1.18154950e-01
-4.16923881e-01 -9.02696401e-02 -6.68289781e-01 5.86973727e-01
6.46477640e-01 -2.84965605e-01 -6.50524497e-01 -9.59697247e-01
4.13022012e-01 2.17808366e-01 2.91834354e-01 -1.10246050e+00
3.47920775e-01 -3.44154805e-01 2.85819173e-01 -5.34842193e-01
3.79632354e-01 -8.08895767e-01 6.37793392e-02 2.82198191e-01
-1.70440838e-01 1.19624920e-02 2.24352390e-01 5.20544827e-01
-1.48964494e-01 -2.62566805e-01 9.04816687e-01 -2.39835694e-01
-9.57212865e-01 7.93449640e-01 5.21747246e-02 -1.04989022e-01
8.43409300e-01 -2.66717076e-01 -3.64768326e-01 -1.32509664e-01
-4.01306838e-01 4.34565336e-01 6.55756652e-01 2.96690203e-02
7.99513102e-01 -1.23154664e+00 -4.74169523e-01 2.43119463e-01
1.94371521e-01 8.06883752e-01 5.13129413e-01 7.88660407e-01
-9.47484851e-01 9.53180268e-02 -4.05769423e-02 -4.71109331e-01
-9.07734692e-01 5.28972268e-01 2.32011110e-01 -3.98902237e-01
-7.50334799e-01 1.18343914e+00 5.55878997e-01 -1.94179609e-01
4.90750134e-01 -4.44163501e-01 -8.91609937e-02 -3.16331506e-01
5.56749105e-01 2.28153124e-01 1.05071723e-01 -1.01552442e-01
-2.54507005e-01 7.08546221e-01 -3.65532905e-01 2.05172058e-02
1.32790923e+00 -1.04548089e-01 -2.44755462e-01 -1.09861739e-01
9.49151218e-01 -3.99971344e-02 -1.81268430e+00 -4.57005829e-01
-3.91227335e-01 -6.00361168e-01 4.04882759e-01 -5.98075390e-01
-1.75738084e+00 9.28862751e-01 4.80149627e-01 -3.80603403e-01
1.59060049e+00 -2.34772772e-01 9.74893212e-01 6.85078651e-02
5.65008104e-01 -8.38925421e-01 1.58537254e-01 4.38677132e-01
5.56976557e-01 -1.11062753e+00 5.17270640e-02 -6.42998815e-01
-4.88271028e-01 8.94892275e-01 9.86884236e-01 -1.50359780e-01
5.61236501e-01 4.26490575e-01 -5.81730232e-02 -1.45456359e-01
-6.23054266e-01 -9.28160623e-02 -5.46391606e-02 4.77286637e-01
1.64776847e-01 3.78847271e-02 -1.04635738e-01 4.94796991e-01
-5.90771511e-02 3.01437620e-02 4.11823511e-01 7.66032755e-01
-5.88159025e-01 -1.02178657e+00 -2.29046270e-01 3.66251200e-01
-1.32927829e-02 -3.59174132e-01 6.61817417e-02 3.35975826e-01
2.06662849e-01 6.41276717e-01 1.18639044e-01 -5.38983583e-01
2.91046381e-01 -4.58914310e-01 3.80608439e-01 -8.01279545e-01
-5.12554705e-01 1.10420853e-01 -4.78425100e-02 -6.33028626e-01
-2.88528532e-01 -2.68987924e-01 -1.53014600e+00 -4.92485881e-01
-3.35598409e-01 -3.24275438e-03 5.37505627e-01 8.99185896e-01
3.60115498e-01 6.46367371e-01 6.74938619e-01 -1.13850129e+00
-9.97211784e-02 -7.96714783e-01 -5.19889772e-01 1.28947213e-01
1.11525215e-01 -4.97637093e-01 -3.95843059e-01 -1.70697451e-01] | [8.776022911071777, -2.2460436820983887] |
d985734d-24a9-43f9-8192-f5111f3b078a | chatgpt-to-replace-crowdsourcing-of | 2305.12947 | null | https://arxiv.org/abs/2305.12947v1 | https://arxiv.org/pdf/2305.12947v1.pdf | ChatGPT to Replace Crowdsourcing of Paraphrases for Intent Classification: Higher Diversity and Comparable Model Robustness | The emergence of generative large language models (LLMs) raises the question: what will be its impact on crowdsourcing. Traditionally, crowdsourcing has been used for acquiring solutions to a wide variety of human-intelligence tasks, including ones involving text generation, manipulation or evaluation. For some of these tasks, models like ChatGPT can potentially substitute human workers. In this study, we investigate, whether this is the case for the task of paraphrase generation for intent classification. We quasi-replicated the data collection methodology of an existing crowdsourcing study (similar scale, prompts and seed data) using ChatGPT. We show that ChatGPT-created paraphrases are more diverse and lead to more robust models. | ['Peter Brusilovsky', 'Jakub Simko', 'Jan Cegin'] | 2023-05-22 | null | null | null | null | ['paraphrase-generation', 'paraphrase-generation', 'intent-classification'] | ['computer-code', 'natural-language-processing', 'natural-language-processing'] | [ 1.35484844e-01 3.37971687e-01 9.32390541e-02 -1.88052371e-01
-8.84898067e-01 -8.26012909e-01 9.77744222e-01 1.71789899e-01
-5.92363417e-01 8.82604599e-01 6.31388187e-01 -2.79107928e-01
2.53595382e-01 -5.80527723e-01 -5.16540110e-01 -4.13148165e-01
5.02398312e-01 8.16857517e-01 3.44139874e-01 -5.53410769e-01
4.07599747e-01 1.50807604e-01 -1.35260737e+00 7.41225600e-01
8.35131228e-01 2.88569957e-01 2.77916223e-01 9.58736658e-01
-2.67038465e-01 1.29708958e+00 -1.15305614e+00 -1.00944555e+00
2.15907142e-01 -5.00505209e-01 -1.11894333e+00 1.63051903e-01
2.26019040e-01 -2.08667174e-01 -1.85014367e-01 5.78358471e-01
7.75743842e-01 2.74373382e-01 4.78785694e-01 -1.47457743e+00
-1.12736368e+00 6.18746936e-01 -2.04531968e-01 1.79851249e-01
1.00600159e+00 3.51090044e-01 7.44166434e-01 -8.99143457e-01
8.78367305e-01 1.54997957e+00 6.83708310e-01 6.98498487e-01
-1.15629923e+00 -3.33648235e-01 -3.70784670e-01 -1.74583822e-01
-1.23910141e+00 -4.10763741e-01 2.10927412e-01 -6.76675797e-01
9.20240402e-01 3.15718442e-01 4.78870034e-01 1.44955444e+00
1.20498642e-01 7.27037430e-01 1.30301023e+00 -5.68393707e-01
5.01099467e-01 6.03230000e-01 -4.59137298e-02 1.44828081e-01
1.79400787e-01 -3.08548152e-01 -8.78319800e-01 -7.44090497e-01
5.59937358e-01 -2.50495106e-01 -3.28629501e-02 3.68230104e-01
-1.04667008e+00 1.17583263e+00 6.48340508e-02 4.25462484e-01
-3.71996611e-01 1.54261268e-03 1.68604806e-01 4.26100612e-01
7.39009440e-01 9.36482966e-01 4.93737645e-02 -5.23949742e-01
-8.00913870e-01 8.80301893e-01 1.31668127e+00 1.26106715e+00
5.79863966e-01 -1.11834377e-01 -8.07794511e-01 8.61528635e-01
1.60291031e-01 4.08114344e-01 8.94464135e-01 -9.25554931e-01
7.07596362e-01 7.57120132e-01 5.85851908e-01 -7.42602766e-01
-1.17029876e-01 4.41029102e-01 -1.90059200e-01 -1.39929712e-01
5.43968737e-01 -6.27241731e-01 -4.47064579e-01 1.16384721e+00
3.41431975e-01 -4.04691428e-01 -1.82900488e-01 8.91369045e-01
8.20796788e-01 3.34321856e-01 2.48534545e-01 1.45605385e-01
1.27840817e+00 -8.32146823e-01 -4.02089238e-01 -5.66565573e-01
7.03674614e-01 -1.09941602e+00 1.25546849e+00 8.74653980e-02
-9.75588202e-01 -4.89674032e-01 -3.88292134e-01 -3.87040555e-01
-2.47542873e-01 -2.30116472e-01 4.40887660e-01 8.42122436e-01
-1.26123250e+00 4.60440457e-01 -1.76677212e-01 -7.14208663e-01
1.13385484e-01 5.53353950e-02 -1.69542670e-01 -2.83900127e-02
-1.36500895e+00 1.08426249e+00 1.05798341e-01 -9.75378901e-02
-6.28659785e-01 -1.75275505e-01 -6.34053290e-01 -4.98242795e-01
1.88939080e-01 -6.16224349e-01 1.70443952e+00 -1.02446365e+00
-1.37492418e+00 1.08546734e+00 -4.81394500e-01 -4.83968824e-01
8.99756849e-01 -1.56032860e-01 5.95609099e-02 -1.68436691e-01
4.91948247e-01 6.06743574e-01 7.56979346e-01 -1.17130196e+00
-4.27310139e-01 -2.29036972e-01 3.97016555e-01 2.71649063e-01
-2.73093227e-02 6.42902315e-01 1.27242252e-01 -5.94721854e-01
-5.59363484e-01 -1.25146627e+00 -3.78158659e-01 -5.40951967e-01
-3.43462586e-01 -5.02080500e-01 9.69567001e-02 -7.68993556e-01
1.00572193e+00 -1.80023849e+00 -3.56075205e-02 -2.24192321e-01
4.71945494e-01 3.34124923e-01 -1.80242941e-01 1.18880844e+00
4.39955443e-01 6.34597480e-01 1.12782806e-01 -6.01574838e-01
1.10181391e-01 1.34282067e-01 -4.36172932e-01 1.01012126e-01
2.94788688e-01 1.38153219e+00 -9.25180376e-01 -6.84132159e-01
-3.09575647e-01 5.04953088e-04 -1.56683102e-01 2.74237007e-01
-3.47863197e-01 5.17821252e-01 -4.08141494e-01 2.67247021e-01
2.81153530e-01 -2.97851652e-01 -6.06478006e-02 9.53954160e-01
-1.18104756e-01 1.52114406e-01 -7.48430908e-01 1.30119395e+00
-5.17023623e-01 8.77290666e-01 -5.98434322e-02 4.99413982e-02
9.21749651e-01 4.65387166e-01 -5.47839664e-02 -5.79524219e-01
6.44218698e-02 3.20625812e-01 2.77373474e-02 -8.44739199e-01
1.16431653e+00 -3.54828477e-01 -8.11436027e-02 9.56900120e-01
-1.63206890e-01 -5.76646984e-01 4.17343587e-01 3.75108331e-01
9.56665635e-01 -1.39481843e-01 3.33017558e-01 -1.80380374e-01
9.38426182e-02 5.57560325e-01 -8.20093006e-02 1.35129273e+00
-1.73597127e-01 8.63438189e-01 5.17394125e-01 -2.22595498e-01
-1.19347942e+00 -6.47143424e-01 2.33972251e-01 1.54990768e+00
-1.20387368e-01 -3.99564266e-01 -1.00519073e+00 -3.53776872e-01
1.39770983e-02 6.90588593e-01 -4.54237163e-01 4.86849481e-03
-5.70529282e-01 -6.37673020e-01 9.78113294e-01 4.64361459e-01
1.04054011e-01 -1.37410033e+00 -4.53460276e-01 3.75778466e-01
-4.37721968e-01 -1.26461160e+00 -4.61379498e-01 -4.98566985e-01
-5.81502914e-01 -8.94060135e-01 -9.55770254e-01 -6.29160762e-01
3.88784021e-01 4.19373184e-01 1.42358422e+00 3.17328691e-01
-1.65769197e-02 4.99041438e-01 -8.57097149e-01 -1.03200614e+00
-1.07458603e+00 4.14504647e-01 -1.33781135e-01 -1.67205960e-01
7.07551956e-01 -2.22634837e-01 -2.29244083e-01 3.59153956e-01
-1.08724189e+00 -5.27724065e-02 3.69102120e-01 4.40976113e-01
-2.88355947e-01 -8.03116024e-01 9.38946426e-01 -9.13451135e-01
1.81467652e+00 -6.75155103e-01 -8.65154937e-02 4.09129679e-01
-2.58457601e-01 -2.79883564e-01 4.70816970e-01 -7.09047854e-01
-8.76343191e-01 -3.51219386e-01 5.98490126e-02 1.35474250e-01
-2.90342867e-01 1.67910770e-01 4.39876705e-01 -6.42385408e-02
1.25569439e+00 1.87405080e-01 3.58757414e-02 -2.42115930e-01
4.10225153e-01 1.17392302e+00 -1.46353483e-01 -6.49716735e-01
8.71963501e-01 2.37906218e-01 -5.77481866e-01 -1.04488623e+00
-4.59082842e-01 -4.60660756e-01 -5.34246504e-01 -1.97321758e-01
9.31390405e-01 -1.06376064e+00 -4.68647301e-01 4.71778780e-01
-1.58397198e+00 -6.93325937e-01 -4.12007242e-01 1.41439378e-01
-3.13729227e-01 3.25894564e-01 -8.48487973e-01 -1.14687991e+00
-3.69998157e-01 -1.01174235e+00 1.46960151e+00 3.81144583e-01
-9.91780519e-01 -1.01940036e+00 3.77656668e-01 7.35480487e-01
6.50737643e-01 4.54931334e-03 5.82117617e-01 -1.23721480e+00
-4.14357364e-01 -5.72508454e-01 4.30990010e-02 3.38628478e-02
-1.47801310e-01 -8.96719769e-02 -1.08085680e+00 6.68457299e-02
2.22048819e-01 -8.08939099e-01 1.80212423e-01 -2.62717903e-01
3.42109203e-01 -5.27318954e-01 5.29525131e-02 -2.99903274e-01
8.38742197e-01 9.53019708e-02 6.64423943e-01 3.05079788e-01
6.35824621e-01 1.00250697e+00 4.51070338e-01 5.91984570e-01
5.63366234e-01 6.34458065e-01 -3.39492470e-01 1.95285946e-01
1.87083468e-01 -3.89354199e-01 3.79656762e-01 5.12366056e-01
-2.22831860e-01 -5.77324331e-01 -1.07352078e+00 6.00734651e-01
-1.97774172e+00 -8.76143813e-01 -5.24282038e-01 1.88974237e+00
1.04611552e+00 -2.72189647e-01 3.97884458e-01 -2.39536241e-01
7.15927541e-01 2.49331519e-01 9.07736942e-02 -5.83533525e-01
-5.65514006e-02 1.47475138e-01 1.15877584e-01 5.61188340e-01
-4.26779121e-01 7.43133843e-01 6.65825796e+00 6.49382651e-01
-5.88000059e-01 4.14103180e-01 5.04757345e-01 -1.60376683e-01
-3.83716077e-01 6.62488192e-02 -7.29463220e-01 6.59336030e-01
1.04568255e+00 -5.53619683e-01 5.69969296e-01 6.49430454e-01
5.44693708e-01 -5.29655993e-01 -1.26833963e+00 5.60442507e-01
4.40353394e-01 -1.04829538e+00 6.30595237e-02 8.02668259e-02
9.51305628e-01 -1.03122428e-01 -4.98855501e-01 4.97849643e-01
6.72518134e-01 -1.11389363e+00 1.00696182e+00 3.14452350e-01
4.25440878e-01 -6.42465577e-02 8.13211620e-01 7.85541594e-01
-6.91846848e-01 -6.43650219e-02 -3.97513032e-01 -5.20720243e-01
6.47856832e-01 3.14376473e-01 -1.46025896e+00 1.78998217e-01
1.99779153e-01 -1.81047216e-01 -1.22146153e+00 9.68428969e-01
-3.94982010e-01 5.68276048e-01 1.79592539e-02 -8.19004953e-01
8.52271914e-02 3.97392064e-02 5.82958460e-01 1.29684472e+00
-7.17417374e-02 -2.24881649e-01 2.60074109e-01 8.63346994e-01
-1.64840683e-01 1.98966071e-01 -9.09564435e-01 -2.70103514e-01
6.83309615e-01 1.07519674e+00 -6.11170232e-01 -3.67446810e-01
-1.90532461e-01 9.95495200e-01 3.18499804e-01 4.08275247e-01
-6.30901694e-01 -2.78285891e-01 1.94165483e-01 5.21908581e-01
-4.24774975e-01 -2.59409398e-01 -3.66992354e-01 -9.59411621e-01
2.60711730e-01 -1.20119667e+00 -9.00526047e-02 -9.59566653e-01
-1.51453042e+00 7.01389849e-01 2.00677678e-01 -7.13576436e-01
-8.41856956e-01 -3.50782216e-01 -7.52078950e-01 1.18851948e+00
-7.42530644e-01 -1.14364123e+00 -5.85840702e-01 4.24418777e-01
8.76050532e-01 -3.12850140e-02 5.52690864e-01 3.54772573e-03
-1.76346615e-01 1.89340949e-01 -2.27374628e-01 2.78144985e-01
7.84283400e-01 -1.28384566e+00 1.13361251e+00 7.05987036e-01
2.64035642e-01 9.17575181e-01 8.13603640e-01 -1.09950507e+00
-1.06028748e+00 -9.50808942e-01 1.28113484e+00 -1.36829865e+00
8.49613905e-01 -7.96315432e-01 -7.97708213e-01 6.84411347e-01
3.36995989e-01 -5.75099468e-01 6.30311012e-01 -2.41995618e-01
7.71076977e-02 6.52523518e-01 -1.01095426e+00 7.22342908e-01
1.08035100e+00 -9.87806320e-01 -6.53917313e-01 8.24492514e-01
8.83802474e-01 -4.05865639e-01 -5.67786336e-01 -4.95584816e-01
3.65637779e-01 -8.33235800e-01 5.32281876e-01 -9.77773845e-01
7.22326159e-01 2.11001739e-01 2.03138024e-01 -1.25202596e+00
-9.59143937e-02 -1.02749670e+00 2.55581081e-01 1.59207535e+00
5.79493165e-01 -7.81811357e-01 5.74862421e-01 1.39660668e+00
1.75030932e-01 -3.83650631e-01 -8.05236995e-01 -7.48105168e-01
2.99862176e-01 -1.76057085e-01 6.33275986e-01 6.61761880e-01
9.46846008e-02 5.10387599e-01 -3.99432063e-01 -4.57080632e-01
6.97366446e-02 -2.20525637e-01 1.22849929e+00 -1.23180866e+00
-3.22081417e-01 -1.38482586e-01 -1.70476183e-01 -6.43340707e-01
1.13576569e-01 -7.69069970e-01 3.72926533e-01 -1.76733720e+00
1.13044076e-01 -1.00483902e-01 9.90431428e-01 1.80600420e-01
-5.49156010e-01 7.32220933e-02 5.83759844e-01 4.56781000e-01
-7.39697993e-01 2.16040090e-01 1.09783232e+00 2.52155393e-01
-4.55887347e-01 1.96545497e-01 -9.81018782e-01 4.75117385e-01
8.53230298e-01 -7.80617058e-01 -4.13876861e-01 -5.10757744e-01
6.13213599e-01 -1.36921331e-01 5.93853354e-01 -6.61337793e-01
3.58743191e-01 -3.07954907e-01 -9.75578874e-02 2.86159813e-01
1.15730360e-01 -1.95058346e-01 1.30474821e-01 5.88125875e-03
-5.44518709e-01 6.12527966e-01 -3.08750477e-02 5.14343798e-01
-3.84512171e-02 -9.65149105e-01 1.02639474e-01 -6.05310857e-01
-2.46560410e-01 -1.62550434e-01 -8.60875666e-01 6.82977617e-01
9.70655084e-01 -3.78916234e-01 -7.04508245e-01 -6.83818579e-01
-2.46644393e-01 1.11132950e-01 8.03609014e-01 6.69438779e-01
2.50945091e-01 -1.06304896e+00 -9.87611890e-01 -3.04087073e-01
1.78376719e-01 3.83481868e-02 -1.01957992e-01 7.23919153e-01
-5.09879649e-01 2.79332846e-01 2.39206880e-01 -1.96259543e-01
-1.02952540e+00 2.38024756e-01 1.68816984e-01 -3.46302986e-01
-4.56088036e-02 7.96545208e-01 -5.50894029e-02 -3.21909606e-01
-2.12263688e-01 -1.36886194e-01 -6.08284362e-02 3.07410598e-01
7.44707942e-01 7.03375161e-01 7.27556692e-03 -6.84379458e-01
-3.04683996e-03 -6.86217919e-02 5.18837534e-02 -5.98752141e-01
9.17186081e-01 -1.45253405e-01 -3.02735955e-01 6.98481202e-01
7.63183773e-01 3.17030191e-01 -8.65656137e-01 3.28321829e-02
2.89346069e-01 -3.96173924e-01 -1.04231393e+00 -5.40142715e-01
-1.84489116e-01 5.96241653e-01 -6.28140718e-02 8.51011992e-01
2.66550571e-01 2.34358802e-01 6.68108881e-01 5.26174843e-01
9.00604844e-01 -9.92275894e-01 2.34172970e-01 6.65574133e-01
1.28740847e+00 -1.37898970e+00 -1.82889655e-01 -1.48451388e-01
-1.31855953e+00 7.68889070e-01 6.00944102e-01 5.96980527e-02
2.15459093e-02 9.23549458e-02 4.72855754e-02 -2.50177860e-01
-7.52795279e-01 -8.44469443e-02 8.30251053e-02 8.55075479e-01
6.43090427e-01 -3.14888805e-02 -4.50956941e-01 7.32477367e-01
-6.41452312e-01 3.73400122e-01 1.01459634e+00 1.09477413e+00
-4.05153781e-01 -9.90961313e-01 -6.17186189e-01 5.14355659e-01
-4.11039203e-01 -2.48411536e-01 -1.31301761e+00 5.79180181e-01
-7.35908300e-02 1.67585778e+00 -2.89668262e-01 -3.52878839e-01
4.23945725e-01 3.84600133e-01 7.07704574e-02 -1.22342992e+00
-1.34339762e+00 -5.91518402e-01 5.63150644e-01 -1.80540919e-01
-3.49135041e-01 -5.59054971e-01 -8.73626828e-01 -3.59693766e-01
-3.64157677e-01 3.06029946e-01 4.27551687e-01 1.09469986e+00
4.31500196e-01 -2.03498229e-01 2.75221735e-01 -9.53081667e-01
-7.45821178e-01 -1.47843945e+00 -3.48825455e-01 8.26063454e-01
-6.06661849e-02 -9.87806022e-02 -3.25570583e-01 1.30988285e-01] | [11.728463172912598, 8.445504188537598] |
4530dd4b-4d74-484e-a6ec-9d650c47a5c2 | reflection-removal-for-in-vehicle-black-box | null | null | http://openaccess.thecvf.com/content_cvpr_2015/html/Simon_Reflection_Removal_for_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Simon_Reflection_Removal_for_2015_CVPR_paper.pdf | Reflection Removal for In-Vehicle Black Box Videos | In-vehicle black box camera becomes an popular equipment in many countries for security monitoring and event capturing. The readability of video contents is the most important capability, which is, however, often degraded due to the reflection on the windscreen. In this paper, we propose a novel method to remove the reflection on the windscreen in the in-vehicle black box videos. The main idea is to exploit spatio-temporal coherence of the reflection which shows that a vehicle moves forward while the reflection layer of internal object remains static. The average image prior is introduced by imposing a heavy-tail distribution with higher peak. The two layered scene is the base of the separation model. In order to remove the reflection, a cost non-convex function is developed based on this property and optimized. Experimental results demonstrate that the proposed approach successfully separates the layers in real black box videos. | ['In Kyu Park', 'Christian Simon'] | 2015-06-01 | null | null | null | cvpr-2015-6 | ['reflection-removal'] | ['computer-vision'] | [ 1.72724620e-01 -1.75654322e-01 1.23761944e-01 -1.64911970e-01
-2.18860209e-01 -2.94024438e-01 3.81002694e-01 -4.08890992e-01
-5.29136717e-01 4.05453652e-01 2.35020500e-02 -1.39823854e-01
-1.33927077e-01 -5.56207895e-01 -8.62590253e-01 -1.05223954e+00
1.15435772e-01 -2.05164537e-01 5.98137021e-01 9.21708867e-02
2.20199287e-01 3.58437568e-01 -1.61850286e+00 1.74755588e-01
7.35443473e-01 8.73458326e-01 6.53723955e-01 5.61363578e-01
7.70884082e-02 6.80974424e-01 -4.52419281e-01 -2.74158597e-01
3.84140998e-01 -2.75681943e-01 1.17967874e-01 4.18198854e-01
2.78195500e-01 -6.65944099e-01 -2.50808150e-01 1.38432682e+00
4.66469899e-02 2.95118988e-01 4.97273475e-01 -1.29596508e+00
1.37145102e-01 6.99768960e-02 -1.16587687e+00 3.55176836e-01
8.40784982e-03 -1.26707610e-02 2.68297106e-01 -8.14073920e-01
3.25855315e-01 1.20030475e+00 6.29220128e-01 3.60134125e-01
-5.13144255e-01 -7.21514523e-01 3.53854716e-01 6.21666968e-01
-1.33350766e+00 -4.18323070e-01 1.03519642e+00 -4.21107739e-01
3.68275464e-01 1.89864203e-01 5.11351764e-01 6.78969800e-01
4.61707294e-01 6.31932855e-01 9.33486581e-01 -1.58413097e-01
-1.48848161e-01 4.73241806e-01 2.25950241e-01 4.86986935e-01
8.05444479e-01 1.51037544e-01 -2.52095520e-01 4.10332456e-02
2.10849240e-01 4.01299208e-01 -4.85206217e-01 -1.18404821e-01
-5.24960101e-01 2.81445235e-01 -1.36295781e-01 -8.25977605e-03
-1.10719196e-01 6.94513926e-03 2.34480470e-01 -2.64196116e-02
5.36734521e-01 -5.71264029e-01 -2.10674092e-01 -9.91752818e-02
-1.00047505e+00 5.22973314e-02 3.75156045e-01 8.21145236e-01
7.10658193e-01 5.75800650e-02 3.09972376e-01 5.33926964e-01
6.97414100e-01 9.16898131e-01 -5.46503104e-02 -5.95050871e-01
7.00162351e-01 3.75148952e-01 3.15239817e-01 -1.17578542e+00
-2.07602546e-01 -4.10771579e-01 -7.13562250e-01 7.26419926e-01
2.87043273e-01 -3.97023290e-01 -7.04891741e-01 1.27446902e+00
6.46243334e-01 6.46954834e-01 5.79263717e-02 1.25653887e+00
7.58182704e-01 1.09138405e+00 -1.80745319e-01 -4.86478508e-01
1.37977898e+00 -7.55491495e-01 -1.10557103e+00 -2.37952605e-01
2.27762476e-01 -7.88818717e-01 5.06272852e-01 6.14641190e-01
-1.00154197e+00 -6.31247044e-01 -1.25572479e+00 3.03214103e-01
7.75848925e-02 1.81649268e-01 -8.23109597e-02 8.53275836e-01
-3.83239657e-01 1.38638720e-01 -8.22568357e-01 -9.88678858e-02
2.14661583e-01 -1.11691214e-01 -4.70783442e-01 -2.26032391e-01
-1.02137077e+00 8.41524124e-01 3.14517021e-01 6.49121344e-01
-8.93813193e-01 -7.63276637e-01 -6.53776765e-01 -4.32650261e-02
7.23410189e-01 -3.42443377e-01 7.93341815e-01 -1.17018628e+00
-1.49305308e+00 5.08426070e-01 -2.88020879e-01 -3.38136554e-01
6.78899527e-01 -5.15722573e-01 -5.54841757e-01 3.67438793e-01
-1.88552156e-01 -3.04468602e-01 1.37327230e+00 -1.70552170e+00
-1.02000523e+00 -3.88722599e-01 -2.38455161e-01 3.11108619e-01
-1.89169988e-01 2.07688436e-01 -6.14051282e-01 -2.63757259e-01
-1.61910191e-01 -8.53979290e-01 7.94610903e-02 -1.82110980e-01
-1.10724807e-01 2.75029182e-01 1.37643886e+00 -1.00271022e+00
1.19886136e+00 -2.56765914e+00 -1.25569090e-01 3.05055916e-01
1.35292798e-01 3.19707453e-01 2.14167774e-01 3.25353444e-01
-3.52083426e-03 -2.72205085e-01 -2.42294565e-01 -1.42638966e-01
-3.95718753e-01 1.46685550e-02 -4.27755266e-01 1.05318117e+00
-6.08426053e-03 1.12068772e-01 -7.01395154e-01 -3.87157261e-01
3.38270634e-01 6.67114615e-01 -4.42683429e-01 1.71692908e-01
1.61807835e-01 4.87086117e-01 -3.70645970e-01 1.31885529e-01
1.54004562e+00 4.20456052e-01 -7.22913444e-03 -2.73036778e-01
-5.67683458e-01 -4.61166471e-01 -1.44766426e+00 1.01362216e+00
-1.92153215e-01 7.64449716e-01 5.95128775e-01 -6.72634125e-01
8.37786853e-01 1.45464048e-01 4.08431917e-01 -6.75161839e-01
2.41850674e-01 1.32243391e-02 -1.03126138e-01 -9.60762203e-01
3.93258691e-01 -1.92719713e-01 2.80862302e-01 -8.27791616e-02
-6.73631191e-01 1.14660375e-01 8.16411898e-02 6.22190870e-02
6.55988753e-01 5.41545600e-02 -8.65386650e-02 -2.16161504e-01
8.67036879e-01 -2.30489060e-01 9.10469413e-01 2.70592898e-01
-6.78661186e-03 5.34142077e-01 2.46496379e-01 -1.37648880e-01
-9.39317524e-01 -7.39843428e-01 4.79372628e-02 4.12003338e-01
8.49640667e-01 -2.08363429e-01 -8.31772983e-01 -3.68353575e-01
-1.01648092e-01 6.72683001e-01 -4.03337359e-01 -2.51357079e-01
-7.84716427e-01 -5.56600988e-01 1.65550947e-01 1.60330638e-01
7.78087258e-01 -4.55350965e-01 -1.03510177e+00 -1.39943615e-01
-2.94737518e-01 -1.42009616e+00 -3.98627311e-01 -3.57116818e-01
-7.00209022e-01 -1.13469708e+00 -6.28666103e-01 -3.92546088e-01
7.63058484e-01 1.04379165e+00 5.03420889e-01 2.25854829e-01
-2.29160085e-01 4.83501881e-01 -2.63026923e-01 -6.06929004e-01
-1.60268560e-01 -6.80479586e-01 -1.50557831e-01 8.55812490e-01
2.04772294e-01 -1.05490983e-02 -7.58778155e-01 4.74442750e-01
-8.65728319e-01 1.71502367e-01 4.50330079e-01 4.42420185e-01
2.90800363e-01 9.65199530e-01 -5.72646223e-02 -5.99738061e-01
-4.71362891e-03 -5.11900187e-01 -1.07052505e+00 -2.24096015e-01
-6.44714236e-02 -4.18432713e-01 5.20580649e-01 -2.57738858e-01
-1.72187150e+00 -1.33561373e-01 2.75824457e-01 -4.97897148e-01
-1.16286993e-01 1.53484479e-01 -5.49683630e-01 1.11896865e-01
-1.28404573e-01 6.86707199e-02 9.74797085e-02 -4.86814290e-01
-9.60182026e-02 4.16486859e-01 3.37855130e-01 -1.99806914e-02
1.25956929e+00 1.19469154e+00 2.79680908e-01 -1.57728577e+00
-5.19541144e-01 -7.70698011e-01 -1.37205705e-01 -7.85892546e-01
9.95474219e-01 -8.68261158e-01 -7.58054316e-01 5.69174349e-01
-1.14419281e+00 -1.89703211e-01 2.12164730e-01 8.43908131e-01
-1.42584413e-01 8.26056302e-01 -2.04383940e-01 -1.32790947e+00
8.09822083e-02 -8.52726161e-01 1.00079918e+00 5.01780510e-01
5.34543931e-01 -8.02866995e-01 6.78893402e-02 4.59707052e-01
-5.00371903e-02 6.46396577e-02 3.75132054e-01 6.37011826e-02
-8.64461899e-01 -2.00821117e-01 -3.24906051e-01 5.88482082e-01
-1.67240292e-01 2.77222574e-01 -9.91113842e-01 -2.48617113e-01
5.76061428e-01 6.23620391e-01 1.02390373e+00 5.50218940e-01
8.95411253e-01 -4.68394123e-02 -5.12350500e-01 8.21739435e-01
1.53452754e+00 4.11922455e-01 9.89979804e-01 4.07100171e-01
6.74832582e-01 9.28196967e-01 1.20850861e+00 5.65652430e-01
-3.14271003e-02 6.49043262e-01 8.79108965e-01 -9.15282741e-02
-8.95901546e-02 1.16670460e-01 8.16930950e-01 5.93971431e-01
-2.81874716e-01 -4.45954114e-01 -5.52612662e-01 3.47192258e-01
-1.80827904e+00 -1.41900671e+00 -8.98431778e-01 2.43000507e+00
7.70109892e-02 2.25183263e-01 -1.43544003e-01 1.91930249e-01
8.26310158e-01 1.15504101e-01 -1.04657058e-02 3.00405193e-02
-1.04943283e-01 -4.54317778e-01 9.20907438e-01 6.78255498e-01
-9.66916323e-01 2.29884282e-01 5.05608177e+00 8.15102041e-01
-1.00279438e+00 1.92608133e-01 7.74466693e-02 1.54212676e-02
-1.75637379e-01 7.01135471e-02 -1.19143295e+00 9.36736047e-01
6.98850095e-01 1.92672759e-03 -9.49060824e-03 4.59958166e-01
8.98164928e-01 -4.66075718e-01 -5.39435804e-01 1.02522516e+00
3.99088949e-01 -7.24120557e-01 -3.44665051e-01 1.46867499e-01
6.02432847e-01 -3.29329282e-01 1.96844205e-01 -3.13543618e-01
-4.79047477e-01 -4.18472677e-01 9.08225298e-01 7.37683058e-01
4.99183685e-01 -8.95144105e-01 7.82286108e-01 5.04496753e-01
-1.41345549e+00 -2.74886012e-01 -2.06087157e-01 1.04774602e-01
6.98834360e-01 7.95352757e-01 -5.04361868e-01 6.70169652e-01
7.86623359e-01 6.45845354e-01 -2.24565849e-01 1.36422503e+00
-1.55863240e-01 6.18644834e-01 -3.06602359e-01 3.48333061e-01
1.69760019e-01 -7.05711305e-01 1.13011050e+00 1.36301196e+00
5.10142565e-01 6.89143389e-02 1.55510291e-01 4.27127749e-01
3.59434694e-01 -1.36724383e-01 -7.32066751e-01 3.88443917e-01
-6.20947443e-02 1.32347035e+00 -6.85819626e-01 -2.32171878e-01
-8.57514799e-01 7.42085397e-01 -6.16002619e-01 6.48251116e-01
-1.34694517e+00 -3.46336663e-01 5.50801158e-01 5.89570761e-01
4.98993486e-01 -3.40050608e-01 -4.95229522e-03 -9.86321509e-01
3.22195709e-01 -4.87807631e-01 1.20945819e-01 -6.62211716e-01
-8.03918898e-01 1.91180721e-01 2.99645007e-01 -1.72425508e+00
3.84949178e-01 -7.15662003e-01 -7.55402982e-01 6.87179625e-01
-1.99300241e+00 -9.52586770e-01 -8.05529535e-01 6.95501864e-01
6.54597819e-01 -3.81532088e-02 -1.67016640e-01 8.86790514e-01
-8.03935826e-01 1.18855186e-01 3.70292217e-01 -1.78583071e-01
6.30033076e-01 -6.16232753e-01 -2.93387383e-01 1.38985395e+00
-4.75003600e-01 3.47980559e-01 1.11400497e+00 -9.01375353e-01
-1.55070674e+00 -1.02266085e+00 4.43967700e-01 -4.80721630e-02
4.93943393e-01 -2.35949695e-01 -9.74884927e-01 4.98781890e-01
3.90550405e-01 -6.30479380e-02 3.89387637e-01 -5.67590296e-01
-5.76888099e-02 -4.85534459e-01 -8.98329139e-01 4.34661090e-01
4.81130332e-01 -1.07855350e-01 -4.95303631e-01 -1.84147879e-02
2.52309650e-01 -1.69659287e-01 -2.07747310e-01 4.77832109e-01
5.37104368e-01 -1.14788246e+00 9.20405149e-01 -1.18590571e-01
-9.81270336e-03 -9.03246105e-01 1.18222430e-01 -9.27487910e-01
-7.45410174e-02 -7.26000905e-01 -4.01938483e-02 1.18616450e+00
-2.86827907e-02 -8.55686009e-01 7.36632943e-01 2.59409785e-01
-3.75031650e-01 3.58481170e-03 -8.17588806e-01 -8.07591856e-01
-6.54206216e-01 -4.45667535e-01 5.73397987e-02 3.99507314e-01
-3.35902900e-01 2.13345215e-02 -7.20880866e-01 8.08020651e-01
1.14503396e+00 -2.06170037e-01 1.15438485e+00 -1.21778381e+00
-3.09059143e-01 -1.22813962e-01 -4.33816791e-01 -1.20408058e+00
7.65653476e-02 -3.07272017e-01 3.47326398e-01 -1.36189258e+00
4.76549596e-01 -2.69104814e-04 -8.20396394e-02 -5.72745562e-01
-2.52955317e-01 6.88031614e-02 2.07715511e-01 -3.29150856e-02
-4.40224379e-01 5.31758368e-01 1.11030722e+00 -4.91446033e-02
-9.42557752e-02 2.12642059e-01 -1.63602188e-01 1.18050432e+00
7.57598281e-01 -4.73545581e-01 -5.57669163e-01 -4.97567356e-01
3.62128466e-01 -1.20594151e-01 5.45224786e-01 -9.71649587e-01
2.30932325e-01 -2.38501072e-01 1.32541716e-01 -1.10010982e+00
5.23779690e-01 -1.62532985e+00 4.58734870e-01 4.59407032e-01
4.30875182e-01 -1.00027367e-01 3.00139308e-01 1.00416231e+00
-2.56660193e-01 -5.49724340e-01 1.02166247e+00 1.33556113e-01
-7.95071006e-01 1.65211767e-01 -5.17099917e-01 -1.42115593e-01
1.37927389e+00 -5.85128248e-01 -4.90391821e-01 -1.99138075e-01
-1.32093340e-01 1.92423761e-01 2.13682368e-01 3.25794518e-01
8.52674484e-01 -1.02730465e+00 -8.38798463e-01 2.96012729e-01
-1.12654209e-01 -2.16585547e-01 8.74134958e-01 1.21142018e+00
-6.81312323e-01 1.31108135e-01 -8.37043524e-02 -6.22949839e-01
-1.79927099e+00 5.75057089e-01 2.25354448e-01 1.35906324e-01
-9.77606535e-01 4.80862349e-01 7.20770180e-01 4.68149781e-01
9.06960368e-02 -1.83885485e-01 -5.12598634e-01 7.70635307e-02
9.72451150e-01 8.48806202e-01 -1.56566009e-01 -1.01035166e+00
-2.37425551e-01 9.74209726e-01 1.11161344e-01 -1.22177795e-01
1.06828415e+00 -6.77758455e-01 -2.60653123e-02 3.81631017e-01
1.14256608e+00 5.82834184e-01 -1.60778511e+00 1.60623878e-01
-3.71473819e-01 -9.10337448e-01 1.62926972e-01 -2.19523180e-02
-1.15492153e+00 8.71132910e-01 4.12718594e-01 2.28275955e-01
1.03208542e+00 -3.90232950e-01 9.09559250e-01 -4.59082983e-02
1.81647956e-01 -1.24569023e+00 -1.34761527e-01 2.64244348e-01
5.63567698e-01 -9.62423325e-01 1.22562058e-01 -7.15409279e-01
-5.91271639e-01 1.14806736e+00 4.69591588e-01 -1.95072711e-01
8.48714113e-01 5.35360813e-01 -1.57075766e-02 -3.06882970e-02
-5.04065096e-01 -2.83850312e-01 1.25300869e-01 5.87025583e-01
-9.34540927e-02 -2.35184580e-01 -1.69705823e-01 3.60837340e-01
4.40003425e-01 -2.56130487e-01 7.26469815e-01 7.58322895e-01
-6.99913800e-01 -4.75525826e-01 -9.51352537e-01 1.27618015e-02
-6.70002639e-01 2.95636445e-01 1.88093230e-01 6.95313990e-01
3.62652302e-01 1.26744080e+00 3.37138288e-02 -8.29982758e-02
5.41941226e-01 -2.88676232e-01 2.91682810e-01 -1.89625397e-01
-1.28074944e-01 3.98626059e-01 6.16270937e-02 -5.48857331e-01
-5.65079749e-01 -6.92431271e-01 -1.34413528e+00 -1.69807151e-01
-4.86484438e-01 2.12221056e-01 7.63487637e-01 7.61640191e-01
-3.85141410e-02 5.31633973e-01 8.11146319e-01 -7.95427680e-01
-1.66889727e-01 -5.92748582e-01 -8.27068150e-01 5.82401216e-01
6.92934215e-01 -9.00370181e-01 -7.68111289e-01 3.52277398e-01] | [9.087042808532715, -1.0335736274719238] |
2ce3b540-5f4d-465c-bdf4-cac76761da55 | sca-net-a-self-correcting-two-layer | 2102.05713 | null | https://arxiv.org/abs/2102.05713v5 | https://arxiv.org/pdf/2102.05713v5.pdf | SCA-Net: A Self-Correcting Two-Layer Autoencoder for Hyper-spectral Unmixing | Hyperspectral unmixing involves separating a pixel as a weighted combination of its constituent endmembers and corresponding fractional abundances, with the current state of the art results achieved by neural models on benchmark datasets. However, these networks are severely over-parameterized and consequently, the invariant endmember spectra extracted as decoder weights have a high variance over multiple runs. These approaches perform substantial post-processing while requiring an exact specification of the number of endmembers and specialized initialization of weights from other algorithms like VCA. We show for the first time that a two-layer autoencoder (SCA), with $2FK$ parameters ($F$ features, $K$ endmembers), achieves error metrics that are scales apart ($10^{-5})$ from previously reported values $(10^{-2})$. SCA converges to this low error solution starting from a random initialization of weights. We also show that SCA, based upon a bi-orthogonal representation, performs a self-correction when the number of endmembers are over-specified. Numerical experiments on Samson, Jasper, and Urban datasets demonstrate that SCA outperforms previously reported error metrics for all the cases while being robust to noise and outliers. | ['Clint Dawson', 'Soumyajit Gupta', 'Gurpreet Singh'] | 2021-02-10 | null | null | null | null | ['hyperspectral-unmixing'] | ['computer-vision'] | [ 4.78543341e-01 -5.97023427e-01 2.88448244e-01 -2.51413286e-01
-5.63578308e-01 -4.57059324e-01 5.61934888e-01 4.08281758e-02
-5.72379529e-01 9.19188142e-01 -5.27574457e-02 -1.22435167e-01
-3.27688754e-01 -7.82603860e-01 -7.81432867e-01 -1.34729564e+00
-3.20694536e-01 1.99036494e-01 -3.72981459e-01 -5.17365001e-02
-8.84220973e-02 3.86438370e-01 -1.97949851e+00 -1.38734460e-01
1.17188692e+00 1.14816344e+00 -1.11614153e-01 5.65407932e-01
8.47777911e-03 4.61556077e-01 -5.17511010e-01 -1.00503214e-01
7.38616407e-01 -4.65371549e-01 -3.14132541e-01 2.46550292e-01
8.84354293e-01 -1.58669069e-01 -7.20264092e-02 1.39689362e+00
4.01559561e-01 3.91259789e-01 9.57048416e-01 -9.24902678e-01
-4.30368930e-01 8.35829794e-01 -5.76874495e-01 -3.29555631e-01
-4.62047070e-01 2.57556736e-01 7.26163685e-01 -7.21088886e-01
1.36539817e-01 7.66156435e-01 1.03302014e+00 6.86777309e-02
-1.69931304e+00 -7.88842678e-01 -5.50618395e-02 -6.69765770e-02
-1.72719479e+00 -3.05550933e-01 6.78745508e-01 -7.93198109e-01
8.95577013e-01 5.16309261e-01 6.44653320e-01 6.96428776e-01
-3.20489436e-01 8.08509663e-02 1.53123116e+00 -6.01603150e-01
4.67606068e-01 8.23283121e-02 4.74233150e-01 4.73001003e-01
7.35468924e-01 2.81185985e-01 -1.01506628e-01 -1.80664897e-01
4.61029619e-01 6.23410754e-03 -6.24536276e-01 -3.29967856e-01
-1.25245869e+00 8.26177359e-01 5.81267476e-01 2.59688973e-01
-7.74939656e-01 6.43883944e-02 7.76967034e-02 1.93848819e-01
6.17901742e-01 5.04677534e-01 -4.22958791e-01 5.54083586e-01
-1.49521780e+00 -7.28915855e-02 6.42395556e-01 5.60083985e-01
1.26632619e+00 6.70971155e-01 3.41379317e-04 1.04310298e+00
2.30504289e-01 9.92096186e-01 5.71587086e-01 -9.17733073e-01
2.36413963e-02 7.71364748e-01 3.01183701e-01 -8.36739480e-01
-4.06758219e-01 -7.55351186e-01 -1.28064978e+00 6.70299888e-01
4.95838672e-01 -4.21560079e-01 -1.20915949e+00 1.72710574e+00
7.99654946e-02 1.39176697e-01 2.80957282e-01 7.64029264e-01
6.36190057e-01 6.67306602e-01 1.89880893e-01 -5.24319470e-01
1.11867046e+00 -6.94865346e-01 -5.09544194e-01 -3.34910482e-01
6.29558787e-02 -6.05928361e-01 7.05307901e-01 4.15634632e-01
-6.64981842e-01 -4.60105866e-01 -1.46105313e+00 5.63483596e-01
-6.72607839e-01 4.65588033e-01 5.48518062e-01 1.03020298e+00
-9.39795971e-01 9.64343548e-01 -5.93072891e-01 -2.75937825e-01
2.20880911e-01 4.11417544e-01 -2.99239367e-01 1.84103027e-01
-9.27695930e-01 7.95081258e-01 7.23300993e-01 3.84230107e-01
-7.42522120e-01 -8.03016841e-01 -7.29293406e-01 -3.02860346e-02
6.08415809e-03 -4.34358329e-01 5.73252976e-01 -1.43710566e+00
-1.46398640e+00 6.11082613e-01 6.58680499e-02 -5.71280897e-01
3.59478295e-01 -2.43969765e-02 -8.18657517e-01 2.00970367e-01
-3.25340360e-01 4.61999953e-01 1.14151132e+00 -1.65547311e+00
-2.44213015e-01 -2.13053599e-01 -2.98623085e-01 -1.08424231e-01
-5.58742523e-01 -2.64817715e-01 3.65818590e-01 -6.82638764e-01
3.96308362e-01 -8.80319893e-01 -2.52165705e-01 -2.47967392e-01
-3.43940049e-01 4.55659002e-01 3.15109968e-01 -8.00623238e-01
1.10348237e+00 -2.05368590e+00 -5.25880568e-02 5.94753504e-01
3.11632045e-02 3.86787683e-01 -8.07638094e-02 2.04192504e-01
-6.06600702e-01 -1.14381649e-02 -1.04462159e+00 -7.47408420e-02
6.06180839e-02 -1.97770577e-02 2.25396249e-02 9.27885115e-01
2.08499506e-02 2.00714588e-01 -7.55814850e-01 5.79583384e-02
4.74373460e-01 6.98992252e-01 -2.62877405e-01 1.71676874e-02
-4.56824601e-02 -5.71032688e-02 4.26063687e-01 5.35776079e-01
1.06244922e+00 -3.61252993e-01 2.32483864e-01 -5.32886386e-01
-4.32409763e-01 -3.67810696e-01 -1.69130659e+00 1.20147884e+00
-1.95226595e-01 4.58416700e-01 4.59589183e-01 -1.05921459e+00
7.90929437e-01 2.04018533e-01 6.27435148e-01 5.52217523e-03
3.40974957e-01 5.30965149e-01 7.40936697e-02 -3.07090804e-02
3.16496879e-01 -5.57383597e-01 4.95291054e-01 4.25504357e-01
3.01630080e-01 -1.07798330e-01 3.67852271e-01 -4.08525616e-01
5.79775631e-01 -9.93657634e-02 3.38415742e-01 -8.10060859e-01
7.30259359e-01 1.27780378e-01 3.18998367e-01 6.92224503e-01
-7.08316937e-02 7.57938325e-01 -1.12552159e-01 -2.88307160e-01
-1.03992915e+00 -9.91939604e-01 -4.51881051e-01 7.64516413e-01
-1.05103374e-01 1.67459130e-01 -1.00820172e+00 -6.95651993e-02
3.84553592e-03 9.45716918e-01 -5.84793627e-01 2.33142618e-02
-2.60901570e-01 -1.63634074e+00 5.29189467e-01 2.40176246e-01
6.38113976e-01 -9.37430799e-01 -5.50273418e-01 1.84245899e-01
-5.52956313e-02 -8.01533461e-01 -9.60423797e-02 6.67836130e-01
-8.88257146e-01 -1.18482792e+00 -8.09905767e-01 -4.63155031e-01
9.84126747e-01 2.54086018e-01 9.66181040e-01 -2.57146537e-01
-1.35422096e-01 8.03096890e-02 -1.95656613e-01 -4.34901237e-01
-4.70930964e-01 -3.35481346e-01 2.48383343e-01 3.79885942e-01
4.54653710e-01 -7.54234254e-01 -6.66840672e-01 1.86480898e-02
-1.21612978e+00 -2.22717315e-01 5.30832648e-01 9.62700307e-01
6.32699907e-01 2.55236268e-01 1.81001142e-01 -7.73532867e-01
3.01652074e-01 -3.80551666e-01 -7.65935898e-01 2.07122371e-01
-9.63153481e-01 1.93101108e-01 7.16086686e-01 -4.05915946e-01
-1.00902832e+00 1.80220619e-01 2.68924236e-01 -5.74449360e-01
-4.30226833e-01 5.53885758e-01 8.48570988e-02 -1.25403121e-01
1.23127413e+00 4.18450117e-01 1.04908004e-01 -4.33665603e-01
3.31456244e-01 6.42574668e-01 6.33517027e-01 -4.21379000e-01
9.97111261e-01 6.44190311e-01 -3.61079872e-02 -1.05168831e+00
-4.14238811e-01 -5.43704510e-01 -6.03929400e-01 -1.15401484e-01
6.24514282e-01 -1.21196938e+00 -3.74700308e-01 8.60912323e-01
-6.63751125e-01 -3.28332126e-01 -4.06769902e-01 6.90505207e-01
-2.94549286e-01 5.10499120e-01 -2.73917824e-01 -9.93930042e-01
-6.73962295e-01 -9.88942206e-01 3.69763315e-01 2.25710914e-01
1.05182223e-01 -8.07943404e-01 -3.41171585e-02 1.76208466e-01
4.88821447e-01 5.59338570e-01 8.49614978e-01 -3.46490920e-01
-8.23877454e-02 -2.51659662e-01 -2.84590542e-01 9.44357932e-01
2.41247654e-01 1.89208224e-01 -1.32758677e+00 -4.30667251e-01
8.33138302e-02 1.70154013e-02 1.22431791e+00 5.98124564e-01
8.76903117e-01 -4.00312722e-01 9.01269987e-02 7.71105766e-01
1.89433610e+00 1.34761762e-02 4.10115272e-01 1.89013034e-01
5.99795341e-01 4.92230684e-01 -2.34163791e-01 3.97496074e-01
-2.38672152e-01 -1.29849002e-01 6.89324141e-01 -1.71824366e-01
-1.46080321e-02 2.65738010e-01 3.91583681e-01 7.34985769e-01
-4.76000965e-01 6.15925808e-03 -6.49542212e-01 6.67027593e-01
-1.49847150e+00 -1.09713733e+00 -4.87507194e-01 2.53879881e+00
9.72562432e-01 -3.04726779e-01 -8.29266105e-03 4.93129253e-01
8.90198469e-01 3.31041694e-01 -5.47441721e-01 1.99813303e-02
-6.01940036e-01 5.69772959e-01 1.07158041e+00 5.94420671e-01
-1.44703448e+00 5.45810819e-01 6.58574772e+00 5.02780616e-01
-1.31207538e+00 -1.26607595e-02 3.89605880e-01 -1.88698843e-01
-1.98228359e-01 -3.61918882e-02 -2.84743816e-01 4.53253418e-01
1.06703460e+00 2.11935446e-01 9.27209377e-01 6.17206097e-01
3.28253880e-02 -1.38791427e-01 -6.07744575e-01 1.07583869e+00
3.02392185e-01 -1.03482115e+00 -1.37777954e-01 -1.68542325e-01
1.22048068e+00 3.23495835e-01 -1.17247939e-01 4.35178494e-03
5.40275156e-01 -1.02288651e+00 7.55646348e-01 5.70921838e-01
7.43377626e-01 -6.44370973e-01 6.30113840e-01 1.42009079e-01
-9.39128220e-01 -3.19329560e-01 -5.23373425e-01 -1.94508165e-01
-1.90667808e-01 1.18189847e+00 -3.76939088e-01 5.66185832e-01
6.55210555e-01 2.97649503e-01 -3.10478479e-01 1.02295613e+00
-4.71407808e-02 7.94188678e-01 -5.97724080e-01 2.26904348e-01
1.83031976e-01 -6.65297568e-01 4.86608922e-01 1.32673991e+00
7.42997110e-01 3.82525846e-02 -4.99360496e-03 1.00341570e+00
-6.24685884e-02 2.32149199e-01 -1.17168322e-01 -2.43420705e-01
3.17388088e-01 1.31323636e+00 -6.43101156e-01 -4.93326455e-01
-9.14443135e-02 6.96254790e-01 -1.45751253e-01 7.48845160e-01
-4.75451201e-01 -3.82885158e-01 1.06447887e+00 -4.79295626e-02
4.47622478e-01 -2.16356128e-01 -3.66544873e-01 -1.15172446e+00
-1.16097577e-01 -8.58109415e-01 3.34170997e-01 -6.21496797e-01
-1.46310318e+00 5.94457626e-01 -8.30141231e-02 -1.34061217e+00
8.02960526e-03 -1.15277779e+00 -3.84490728e-01 1.00856626e+00
-1.50845540e+00 -8.97007585e-01 -5.67414284e-01 4.14103091e-01
-4.67908792e-02 -1.97849154e-01 1.22621286e+00 1.41720474e-01
-8.43577743e-01 2.36692935e-01 7.96785057e-01 1.29180014e-01
6.23132169e-01 -1.27777994e+00 -2.59760082e-01 1.19110739e+00
1.13929480e-01 7.16202080e-01 8.96079242e-01 -3.73520136e-01
-9.69632745e-01 -1.18610907e+00 5.07230520e-01 1.64506376e-01
5.74806571e-01 -9.96515155e-03 -7.72890866e-01 4.25909281e-01
3.96284193e-01 1.60124317e-01 1.04941130e+00 -7.72002116e-02
-5.00031710e-01 -3.85238826e-01 -1.27975321e+00 4.73450661e-01
6.53034151e-01 -4.32118863e-01 -2.35786542e-01 2.29262426e-01
2.25847691e-01 -9.02783647e-02 -9.73177612e-01 5.74319839e-01
6.40163481e-01 -1.32563150e+00 1.11507070e+00 -3.18463326e-01
2.94302523e-01 -7.27056324e-01 -4.08703893e-01 -1.49130523e+00
-3.79472613e-01 -3.58735859e-01 -8.98166746e-02 8.60464156e-01
4.39431906e-01 -6.58641636e-01 4.73600060e-01 3.71300578e-01
2.34071966e-02 -1.06306877e-02 -7.23164558e-01 -7.17123866e-01
1.77888379e-01 -4.95183945e-01 6.33706987e-01 1.13476861e+00
-3.56911123e-01 -2.34568298e-01 -4.47893888e-01 5.01000762e-01
9.86002028e-01 1.84207305e-01 3.36424619e-01 -1.53662431e+00
-2.40084350e-01 -6.97893083e-01 2.91789770e-02 -3.47698271e-01
1.78694069e-01 -8.86462092e-01 2.53737479e-01 -1.16098857e+00
4.07562219e-02 -2.12762773e-01 -7.87596047e-01 5.93854070e-01
-2.04451784e-01 5.70585251e-01 -2.70760898e-02 1.42704144e-01
2.23848194e-01 4.48940128e-01 6.16117775e-01 -5.67948163e-01
-2.79075593e-01 -2.90343255e-01 -7.13239551e-01 8.93666565e-01
8.44200075e-01 -3.04695755e-01 -1.58990413e-01 -3.18289042e-01
7.95420632e-02 -4.55743730e-01 5.32718360e-01 -1.48416281e+00
-6.86652539e-03 -3.07610542e-01 5.92232764e-01 -4.14127111e-01
3.49748641e-01 -1.02369547e+00 6.67593420e-01 2.53025502e-01
8.51076469e-02 -4.65527534e-01 2.78922498e-01 4.31622595e-01
-5.48219234e-02 -5.47607958e-01 9.29870248e-01 -1.91966042e-01
-5.50893366e-01 8.45893696e-02 -4.63218451e-01 -4.21383798e-01
6.54402852e-01 -4.08688962e-01 -2.19824314e-01 -1.73015282e-01
-5.37979782e-01 -1.92117393e-01 5.15764236e-01 -2.89324820e-01
7.94523358e-02 -1.13731825e+00 -8.90707791e-01 3.54151666e-01
3.64465527e-02 -3.71102095e-02 3.84915322e-01 8.29457164e-01
-7.66323626e-01 4.51052152e-02 -1.92509055e-01 -4.21523809e-01
-8.87106895e-01 3.10666740e-01 6.59068942e-01 -3.64914397e-03
-2.96927989e-01 7.50488997e-01 -2.14517176e-01 -5.86229920e-01
1.98358353e-02 -3.52192551e-01 -1.23633854e-01 3.95686418e-01
4.86775786e-01 7.17121661e-01 1.96157709e-01 -7.50592113e-01
-2.15192195e-02 4.58025128e-01 6.00723684e-01 7.16078952e-02
1.53438544e+00 1.17168024e-01 -5.62564731e-01 4.08122361e-01
9.91816938e-01 -3.15734483e-02 -1.39233863e+00 -1.11507021e-01
-3.32969993e-01 -1.38276756e-01 3.08509141e-01 -9.87861931e-01
-1.18793154e+00 7.38635421e-01 1.08841980e+00 2.07993597e-01
1.23862457e+00 -7.29886115e-01 2.02629790e-01 5.00061631e-01
1.31366979e-02 -1.18765414e+00 -5.63564122e-01 3.89604717e-01
7.65158176e-01 -1.12348998e+00 2.77863085e-01 -2.10426703e-01
-1.64710119e-01 1.14500475e+00 3.12958151e-01 -2.02903926e-01
9.02656198e-01 7.29260221e-02 5.35252035e-01 -4.51052301e-02
1.66475609e-01 -3.91108394e-01 3.13246459e-01 4.73403454e-01
4.49902415e-01 4.82788205e-01 -2.80812740e-01 4.10439074e-01
-2.63193220e-01 -2.07636774e-01 2.92621583e-01 5.13657868e-01
-5.65191746e-01 -8.29051077e-01 -8.46430540e-01 6.55158877e-01
-2.38029689e-01 -4.55605060e-01 4.19854634e-02 5.64978778e-01
3.55538905e-01 1.03195977e+00 3.99405099e-02 -2.86386937e-01
1.00043245e-01 6.32181823e-01 1.53352201e-01 -1.70275018e-01
-6.11418307e-01 1.56552032e-01 -2.08640963e-01 -6.38008267e-02
-9.48444486e-01 -6.50059998e-01 -9.83638287e-01 -2.29715988e-01
-5.16820014e-01 3.36739495e-02 7.16938019e-01 8.21279407e-01
-8.60277470e-03 3.81657243e-01 4.68589395e-01 -1.13396192e+00
-8.11412454e-01 -1.42075396e+00 -1.00293374e+00 6.10785484e-01
3.21907431e-01 -5.53702831e-01 -8.84766638e-01 2.45023057e-01] | [10.09930419921875, -2.052030563354492] |
e7fc37d2-ca1d-40ce-a659-2ae0ed2beb0f | sim-to-real-learning-for-casualty-detection | 1908.03057 | null | https://arxiv.org/abs/1908.03057v2 | https://arxiv.org/pdf/1908.03057v2.pdf | Sim-to-Real Learning for Casualty Detection from Ground Projected Point Cloud Data | This paper addresses the problem of human body detection---particularly a human body lying on the ground (a.k.a. casualty)---using point cloud data. This ability to detect a casualty is one of the most important features of mobile rescue robots, in order for them to be able to operate autonomously. We propose a deep-learning-based casualty detection method using a deep convolutional neural network (CNN). This network is trained to be able to detect a casualty using a point-cloud data input. In the method we propose, the point cloud input is pre-processed to generate a depth image-like ground-projected heightmap. This heightmap is generated based on the projected distance of each point onto the detected ground plane within the point cloud data. The generated heightmap -- in image form -- is then used as an input for the CNN to detect a human body lying on the ground. To train the neural network, we propose a novel sim-to-real approach, in which the network model is trained using synthetic data obtained in simulation and then tested on real sensor data. To make the model transferable to real data implementations, during the training we adopt specific data augmentation strategies with the synthetic training data. The experimental results show that data augmentation introduced during the training process is essential for improving the performance of the trained model on real data. More specifically, the results demonstrate that the data augmentations on raw point-cloud data have contributed to a considerable improvement of the trained model performance. | ['Roni Permana Saputra', 'Petar Kormushev', 'Nemanja Rakicevic'] | 2019-08-08 | null | null | null | null | ['body-detection'] | ['computer-vision'] | [ 3.55035514e-01 3.45777005e-01 5.87190032e-01 -3.49349082e-01
-1.28398210e-01 2.06906945e-01 1.43255487e-01 1.83365971e-01
-6.87751293e-01 2.61790067e-01 -2.60165960e-01 -4.74312007e-02
-4.69978712e-02 -1.20355320e+00 -9.86610115e-01 -4.44699615e-01
-5.64346194e-01 8.48609805e-01 2.27292806e-01 -8.59477639e-01
-6.67552054e-02 8.61112058e-01 -1.86253715e+00 -2.31722239e-02
7.39028454e-01 1.09187973e+00 3.39543581e-01 3.90430123e-01
4.51981604e-01 2.67649621e-01 -5.77953696e-01 4.61542644e-02
8.36718678e-01 -1.45163715e-01 -3.23571444e-01 2.67669082e-01
2.94569016e-01 -5.40115297e-01 -3.76284450e-01 6.88289642e-01
4.64307308e-01 1.80990957e-02 6.51388824e-01 -1.21998942e+00
1.25285178e-01 -1.55494764e-01 -2.40791708e-01 -8.24827999e-02
4.77175385e-01 2.20995530e-01 3.19981247e-01 -8.65331471e-01
4.60970461e-01 1.12108171e+00 9.73947465e-01 5.34276187e-01
-6.33347571e-01 -4.06699002e-01 -3.33180130e-01 9.27776694e-02
-1.48123968e+00 1.25898972e-01 7.70223320e-01 -5.08435011e-01
8.29421341e-01 -2.54608020e-02 1.09145463e+00 6.30463123e-01
3.69210184e-01 5.29738247e-01 6.69711590e-01 -5.73763609e-01
2.05844715e-01 -3.43320549e-01 -2.86850721e-01 1.03970754e+00
5.31321347e-01 2.02830359e-01 -1.72058478e-01 1.80211931e-01
9.75548387e-01 3.30942035e-01 -5.23526728e-01 -4.89048392e-01
-8.70738626e-01 7.23441601e-01 1.26290226e+00 2.26372629e-02
-7.52377391e-01 -4.45622392e-03 1.70153245e-01 1.38979316e-01
1.97459400e-01 9.11748111e-02 -9.56637338e-02 2.79323399e-01
-9.04333413e-01 5.34125924e-01 5.09805381e-01 5.72163522e-01
7.53637433e-01 -2.71292441e-02 2.28478506e-01 3.63072842e-01
3.59572500e-01 9.66885388e-01 3.29426020e-01 -3.40991586e-01
7.79025316e-01 1.15557396e+00 8.68760124e-02 -1.29299629e+00
-9.15926516e-01 -4.57096368e-01 -9.72580969e-01 8.37551832e-01
3.01725984e-01 -2.23224238e-01 -1.34472346e+00 1.24900508e+00
4.97268081e-01 -3.21424842e-01 2.82269344e-02 1.41993105e+00
5.16212106e-01 6.67774439e-01 -2.87042171e-01 2.37209678e-01
1.21073020e+00 -4.35730726e-01 -3.57524782e-01 -6.18467987e-01
6.13299370e-01 -2.45506223e-02 9.80217576e-01 4.14992154e-01
-6.30821228e-01 -7.96163738e-01 -1.39489079e+00 9.94236842e-02
-3.42634141e-01 3.08792412e-01 8.66443887e-02 2.02333733e-01
-8.34917426e-01 6.30995750e-01 -8.84740591e-01 -5.59725821e-01
2.55715191e-01 5.25328696e-01 -7.06923723e-01 -1.31178319e-01
-1.12138176e+00 9.69086170e-01 4.65800464e-01 4.76869494e-01
-8.20431471e-01 -3.55331264e-02 -1.08376467e+00 1.26204919e-02
2.01281486e-03 -9.34179902e-01 8.92821670e-01 -6.17326021e-01
-8.88028383e-01 9.60618198e-01 3.54708880e-01 -7.18527853e-01
8.09303939e-01 -4.73179728e-01 -1.71139970e-01 2.93255270e-01
1.58618525e-01 6.41172051e-01 9.07278836e-01 -1.43635774e+00
-7.38964260e-01 -6.70589447e-01 -7.95606524e-02 4.02776957e-01
-1.03378415e-01 -3.35756779e-01 -3.99466008e-01 -3.92233312e-01
5.86574614e-01 -1.06299984e+00 -2.08304584e-01 1.28209665e-01
-3.95323157e-01 2.31682941e-01 8.71079504e-01 -6.29618466e-01
5.66138566e-01 -1.99960101e+00 1.41029432e-01 4.80482697e-01
-6.54212385e-02 3.08082908e-01 2.23362327e-01 5.23587525e-01
-5.58326393e-02 -5.06755054e-01 -6.60722852e-01 -4.36129451e-01
-5.53929448e-01 3.19465131e-01 1.29518565e-03 6.34537756e-01
2.41895631e-01 5.85150123e-01 -5.83834291e-01 -3.01394910e-01
4.24468607e-01 6.13467574e-01 -8.10438171e-02 4.97906417e-01
-4.18965444e-02 4.16504502e-01 -9.70123857e-02 4.78154629e-01
7.66463220e-01 2.61846244e-01 -2.44275123e-01 -2.13308423e-03
-6.84612244e-02 -2.34741643e-01 -9.79681313e-01 1.42401636e+00
-2.99614131e-01 3.21916193e-01 1.26060113e-01 -7.92613149e-01
1.24217629e+00 1.86350062e-01 4.29154754e-01 -6.97504222e-01
3.57833534e-01 2.99459368e-01 1.64242908e-02 -6.78560793e-01
3.45001698e-01 -3.37755680e-01 -7.93552846e-02 -1.39688760e-01
-2.54202962e-01 -1.21670015e-01 -1.10332660e-01 -1.92610025e-01
9.92544353e-01 -7.75697604e-02 -5.56942746e-02 -5.06059565e-02
4.47666734e-01 5.45288742e-01 3.78626168e-01 4.85916287e-01
2.48864949e-01 8.90914679e-01 -1.64680555e-01 -8.03341568e-01
-1.05724037e+00 -8.59183729e-01 9.62524340e-02 4.77949142e-01
2.95361400e-01 8.33301106e-04 -1.06347525e+00 -4.50696796e-01
1.72611967e-01 1.90681055e-01 -6.41956627e-01 -3.09315652e-01
-8.32528472e-01 -5.66537619e-01 4.63793784e-01 6.57780111e-01
7.76077032e-01 -1.46835840e+00 -1.69028974e+00 2.27991462e-01
-1.71559855e-01 -9.92207587e-01 3.57276797e-01 3.93114388e-01
-1.05911660e+00 -1.18970454e+00 -7.11275458e-01 -9.20214057e-01
8.45597029e-01 1.17547594e-01 5.76635659e-01 5.46557128e-01
-5.89595914e-01 1.64531663e-01 -4.66622502e-01 -7.76371479e-01
-1.34605467e-01 -1.55382976e-01 3.75061989e-01 -3.12021337e-02
1.09994300e-01 -4.62832481e-01 -7.50999033e-01 1.62821516e-01
-8.07433605e-01 6.00683577e-02 7.48171866e-01 5.63731492e-01
3.80156159e-01 2.26609200e-01 7.19211847e-02 -3.28821808e-01
5.56459785e-01 -1.54155448e-01 -4.60215062e-01 -1.39325514e-01
-1.94674551e-01 -1.49827436e-01 4.55653071e-01 -1.69071019e-01
-4.73482460e-01 4.96264964e-01 -4.37544942e-01 -5.63333213e-01
-3.41952622e-01 7.80551612e-01 -1.67503208e-01 -1.07025340e-01
9.36032593e-01 9.59200114e-02 3.00089210e-01 -4.18216169e-01
-1.62026688e-01 6.31364226e-01 1.00892437e+00 -1.26558512e-01
1.02824509e+00 7.92457283e-01 4.19894904e-01 -9.38282728e-01
-5.16698956e-01 -6.74476385e-01 -1.04508364e+00 -5.44721484e-01
9.40203607e-01 -1.00022066e+00 -5.65977812e-01 6.99027658e-01
-1.19946277e+00 -5.60865879e-01 -3.54479812e-02 3.80202264e-01
-6.16377175e-01 1.34849444e-01 -4.51029807e-01 -9.29136038e-01
-7.10438788e-01 -7.05636680e-01 1.18194449e+00 1.84499875e-01
1.32317334e-01 -5.86733758e-01 9.50157046e-02 1.55519277e-01
7.29704052e-02 7.88845181e-01 5.33028185e-01 -2.52670646e-01
-1.80162296e-01 -7.61143863e-01 1.14815980e-01 2.42886156e-01
-7.95348808e-02 -3.00542861e-01 -8.49699974e-01 -3.30022514e-01
2.82308280e-01 -1.32007092e-01 6.13447070e-01 3.61787587e-01
6.45176589e-01 -5.08889630e-02 -5.26369989e-01 6.48042202e-01
1.39538193e+00 -4.89656478e-02 8.45048487e-01 7.83403993e-01
7.71851718e-01 5.79609931e-01 1.11646855e+00 5.21696508e-01
2.57810980e-01 8.26730728e-01 1.12847602e+00 -5.56483984e-01
2.74434052e-02 -3.56680781e-01 2.85055786e-01 3.17343533e-01
-4.20986056e-01 -6.10821992e-02 -1.33407485e+00 4.84097660e-01
-1.87752450e+00 -7.47824013e-01 -5.94274282e-01 2.22263312e+00
3.65233123e-02 2.81671166e-01 1.57286018e-01 7.44915724e-01
5.01351357e-01 -4.10321891e-01 -1.87030554e-01 -1.61607489e-01
1.32458627e-01 1.58485219e-01 4.92491066e-01 3.19554865e-01
-1.20133805e+00 7.77363122e-01 4.81148911e+00 -2.12701093e-02
-1.44366002e+00 -1.75548464e-01 -1.45151854e-01 1.43988594e-01
6.14091635e-01 -4.82422471e-01 -5.93945622e-01 1.90581813e-01
6.18565679e-01 3.21569264e-01 -4.58564237e-02 1.09043241e+00
3.51219654e-01 -2.49302357e-01 -8.68634105e-01 1.08344555e+00
3.35374057e-01 -8.36476684e-01 6.41089752e-02 2.21546650e-01
2.30696023e-01 1.50803521e-01 -2.92515934e-01 2.00329617e-01
-3.72110099e-01 -9.36670840e-01 1.02593529e+00 4.82046306e-01
6.94946826e-01 -8.40149343e-01 1.17326903e+00 9.98798788e-01
-1.09606349e+00 -3.39480311e-01 -5.88975012e-01 -4.85077500e-01
5.39679170e-01 6.05475008e-01 -1.25287306e+00 5.63227236e-01
9.45074618e-01 2.67571658e-01 -6.59698129e-01 1.27405632e+00
-5.40620685e-01 1.22554660e-01 -6.15759671e-01 1.87252209e-01
1.11350745e-01 1.32020321e-02 3.29693466e-01 9.66637433e-01
5.73903263e-01 1.36059865e-01 3.69508535e-01 5.26598454e-01
3.25873435e-01 2.64388006e-02 -9.49597239e-01 6.70824826e-01
4.98958156e-02 1.14243150e+00 -7.59805799e-01 -2.40022779e-01
2.12427646e-01 1.17745304e+00 1.66212782e-01 -9.98990834e-02
-7.06330419e-01 -5.75165927e-01 2.68541276e-01 8.66121948e-01
4.72646914e-02 -4.13958222e-01 -2.30933681e-01 -7.16490626e-01
3.43540519e-01 -6.89868093e-01 2.89900213e-01 -9.35209513e-01
-7.88906097e-01 9.89350736e-01 3.21483128e-02 -1.44561803e+00
-3.57511401e-01 -8.33275914e-01 -6.62851155e-01 9.40748513e-01
-9.52638209e-01 -1.20739245e+00 -1.06476021e+00 7.49962389e-01
3.09795380e-01 4.27077711e-02 8.00238907e-01 1.71735585e-01
-1.02240868e-01 1.86482251e-01 -5.76112807e-01 3.39748383e-01
1.26033545e-01 -1.00881231e+00 6.49767578e-01 8.94871891e-01
-2.30378602e-02 3.93568546e-01 7.82419443e-01 -1.10225356e+00
-1.11321735e+00 -1.17102373e+00 5.74567080e-01 -4.38562423e-01
-3.69982868e-02 -4.69932079e-01 -9.21213269e-01 6.06249750e-01
-4.31826741e-01 -2.50346698e-02 -4.72802445e-02 -3.12718302e-01
1.96197405e-01 -4.02770452e-02 -1.42834008e+00 2.43528396e-01
9.72400725e-01 -1.38961017e-01 -8.77745211e-01 2.62746722e-01
2.71748155e-01 -6.95570946e-01 -4.76615161e-01 9.06685829e-01
4.64382768e-01 -1.08123875e+00 1.04598987e+00 -1.06187180e-01
2.89028764e-01 -5.56423306e-01 7.69471228e-02 -1.32992661e+00
-1.39393762e-01 1.15408681e-01 -2.98938472e-02 3.45324636e-01
1.10888205e-01 -2.64297247e-01 1.09456921e+00 2.10194007e-01
-3.72168839e-01 -7.49184787e-01 -1.05132258e+00 -8.51443350e-01
-2.08264768e-01 -3.75318229e-01 4.16623056e-01 5.01601219e-01
2.07336228e-02 1.57582119e-01 -2.02253327e-01 7.50526130e-01
6.47428274e-01 -1.37088329e-01 1.18674338e+00 -1.57309556e+00
2.01318443e-01 2.35512435e-01 -9.86925781e-01 -7.67015755e-01
-3.72240007e-01 -5.08845210e-01 5.74271321e-01 -2.03188682e+00
-4.24074709e-01 -4.47231114e-01 2.40822911e-01 6.80674195e-01
-6.54497594e-02 4.42312598e-01 3.25120836e-01 2.48860806e-01
6.12709224e-02 5.18264532e-01 9.48676646e-01 5.16855754e-02
-3.51223201e-01 3.44190419e-01 7.92379901e-02 7.85340011e-01
9.12757337e-01 -5.30466259e-01 -1.92702830e-01 -3.99305880e-01
7.11744055e-02 9.09084231e-02 8.11960816e-01 -2.09563875e+00
2.90995836e-01 3.19800049e-01 5.53176522e-01 -7.91524947e-01
7.14784741e-01 -1.14978850e+00 1.08338900e-01 1.13510084e+00
3.76382738e-01 2.49314591e-01 2.32461095e-01 3.86216223e-01
-4.40401621e-02 -2.39078894e-01 5.87213695e-01 -2.32901990e-01
-6.07765675e-01 2.45074019e-01 -1.83168307e-01 -4.02452916e-01
9.39917326e-01 -5.09381592e-01 1.41922146e-01 -4.39204752e-01
-7.04535723e-01 1.01534262e-01 6.88239336e-01 3.77218872e-01
1.10574532e+00 -1.00127316e+00 -6.56866968e-01 6.17927015e-01
3.37116122e-01 4.62657452e-01 1.49970144e-01 6.48064017e-01
-1.20992589e+00 1.34830028e-01 -4.69817519e-01 -8.19046140e-01
-1.30729568e+00 6.29937410e-01 6.68482304e-01 1.58614472e-01
-1.24253285e+00 5.49822211e-01 -1.39084965e-01 -5.91233611e-01
2.81779081e-01 -6.61048591e-01 -1.42330945e-01 -4.11196291e-01
4.17330712e-01 3.18634771e-02 3.70520592e-01 -9.40285981e-01
-3.63736719e-01 7.71302879e-01 5.16443372e-01 -2.22466812e-01
1.42731535e+00 2.74427772e-01 1.09869488e-01 2.00019374e-01
8.01822066e-01 -2.07782447e-01 -1.19911015e+00 2.19938964e-01
-1.43333867e-01 -3.07788163e-01 -1.71462998e-01 -6.86209321e-01
-9.58791733e-01 1.15691245e+00 1.01621413e+00 5.95341995e-02
9.98289049e-01 -2.43784159e-01 6.77724600e-01 6.28773868e-01
9.41717088e-01 -8.86759877e-01 -9.26736835e-03 4.42526102e-01
1.18501234e+00 -1.18140626e+00 5.69348165e-04 -5.31635880e-01
-4.19491202e-01 1.18782878e+00 6.67861879e-01 -3.75587344e-01
4.34677869e-01 9.94390845e-02 4.38353479e-01 -7.84000516e-01
2.92854011e-02 -4.44302410e-01 1.58312201e-01 8.81141901e-01
-3.18350196e-01 8.87020305e-02 -3.20631452e-02 3.19708943e-01
-6.33221865e-01 1.64019302e-01 3.61840010e-01 1.21158206e+00
-8.98342550e-01 -5.88861585e-01 -8.24236989e-01 2.40361035e-01
9.77885202e-02 3.90113354e-01 -4.41191882e-01 1.07612348e+00
6.48440063e-01 8.25546145e-01 1.38670221e-01 -6.79537356e-01
9.02648270e-01 -1.03179567e-01 4.14521277e-01 -7.21074045e-01
-6.29530609e-01 -4.03350204e-01 -1.86806038e-01 -4.10787940e-01
-1.37536228e-01 -2.63344556e-01 -1.72043049e+00 -6.15526456e-03
-1.15641192e-01 3.27022038e-02 9.78839040e-01 9.47699428e-01
-2.47707218e-02 4.32898968e-01 3.99312735e-01 -1.49688160e+00
-2.81997591e-01 -1.08669305e+00 -5.75466752e-01 7.89568424e-01
2.53509998e-01 -9.17059660e-01 -2.67463714e-01 -4.99623157e-02] | [7.757058143615723, -1.8780211210250854] |
7a4d53e3-3fbd-4ac5-a47f-64142d00dbdf | human-activity-recognition-using-deep-1 | 2304.14499 | null | https://arxiv.org/abs/2304.14499v1 | https://arxiv.org/pdf/2304.14499v1.pdf | Human activity recognition using deep learning approaches and single frame cnn and convolutional lstm | Human activity recognition is one of the most important tasks in computer vision and has proved useful in different fields such as healthcare, sports training and security. There are a number of approaches that have been explored to solve this task, some of them involving sensor data, and some involving video data. In this paper, we aim to explore two deep learning-based approaches, namely single frame Convolutional Neural Networks (CNNs) and convolutional Long Short-Term Memory to recognise human actions from videos. Using a convolutional neural networks-based method is advantageous as CNNs can extract features automatically and Long Short-Term Memory networks are great when it comes to working on sequence data such as video. The two models were trained and evaluated on a benchmark action recognition dataset, UCF50, and another dataset that was created for the experimentation. Though both models exhibit good accuracies, the single frame CNN model outperforms the Convolutional LSTM model by having an accuracy of 99.8% with the UCF50 dataset. | ['Manoj Kumar Rajagopal', 'Balamurugan MS', 'Pooja', 'Annapoorani Subramanian', 'Sheryl Mathew'] | 2023-04-18 | null | null | null | null | ['action-recognition-in-videos', 'human-activity-recognition', 'human-activity-recognition'] | ['computer-vision', 'computer-vision', 'time-series'] | [ 4.93662477e-01 -9.08396766e-02 -2.10910723e-01 -2.42017671e-01
-9.38222408e-02 -4.83605340e-02 7.69068182e-01 -2.01037243e-01
-8.36026967e-01 8.36131394e-01 3.07076603e-01 -1.44121721e-02
-1.42818406e-01 -7.54342079e-01 -6.72914147e-01 -6.99941993e-01
-2.76108325e-01 -3.31537500e-02 4.98754293e-01 1.03254162e-01
3.11014026e-01 3.99941266e-01 -1.90778804e+00 6.58897758e-01
2.67232001e-01 1.40136909e+00 1.42465547e-01 7.92659640e-01
-2.60289937e-01 1.44853806e+00 -6.13856792e-01 -4.00063358e-02
1.89668715e-01 -3.80889624e-01 -9.62332666e-01 3.46899824e-03
7.00185150e-02 -1.88947871e-01 -3.80433530e-01 5.59578240e-01
4.01691049e-01 1.73001498e-01 3.61563176e-01 -1.05422962e+00
-1.26472294e-01 1.38689294e-01 -1.66076392e-01 5.18153191e-01
6.28713489e-01 1.35468587e-01 2.91581273e-01 -4.96036410e-01
3.31195027e-01 1.06222975e+00 8.23608935e-01 5.20486057e-01
-5.53546548e-01 -4.87144977e-01 -1.63981721e-01 6.27649784e-01
-1.11644912e+00 -3.44838738e-01 5.24042487e-01 -5.92206538e-01
1.31442642e+00 -4.67991680e-02 7.95258820e-01 1.24319708e+00
2.99019337e-01 1.03893828e+00 1.12756419e+00 -4.80486721e-01
2.06576571e-01 -1.40993878e-01 1.51777029e-01 5.30094206e-01
-3.70080732e-02 3.80986840e-01 -5.93356431e-01 1.68700784e-01
9.07068193e-01 2.41036847e-01 -2.75218487e-01 -1.59413591e-01
-1.36187470e+00 7.85583317e-01 4.12129521e-01 8.21310818e-01
-5.37019610e-01 2.23874673e-01 8.29016626e-01 4.03340787e-01
2.80881584e-01 3.40810746e-01 -6.06247067e-01 -6.11166239e-01
-8.41702998e-01 1.30690471e-03 7.31680512e-01 3.48274350e-01
3.95735294e-01 8.58946219e-02 -1.47783011e-01 8.41412187e-01
2.04570040e-01 1.96000025e-01 1.12430334e+00 -6.60410106e-01
2.74473667e-01 7.28842199e-01 -1.27745703e-01 -1.01670206e+00
-4.80818123e-01 -3.00198328e-02 -8.72862697e-01 2.60141462e-01
6.35949016e-01 -1.13740690e-01 -1.06981170e+00 1.33624756e+00
-3.08889914e-02 5.33762753e-01 3.13415915e-01 7.53197193e-01
1.07729280e+00 7.56026924e-01 2.64290690e-01 -6.48211390e-02
1.18556917e+00 -1.06971681e+00 -6.94001555e-01 -3.15976828e-01
8.42218697e-01 -5.39912999e-01 5.80927134e-01 5.98446190e-01
-7.54617095e-01 -8.43455613e-01 -1.11250293e+00 1.75347567e-01
-7.29407430e-01 2.67057598e-01 5.54375172e-01 6.14456236e-01
-7.96933234e-01 7.68920064e-01 -9.63510394e-01 -7.45942533e-01
5.07522881e-01 4.26698834e-01 -5.83338737e-01 4.37104478e-02
-1.20963132e+00 8.80454004e-01 7.72752404e-01 2.22905904e-01
-9.00975227e-01 1.53369829e-02 -8.00807297e-01 -2.21464876e-02
3.20851028e-01 -2.35967845e-01 1.18359435e+00 -1.43232369e+00
-1.55693781e+00 7.34902263e-01 1.26069352e-01 -1.03038836e+00
5.71873724e-01 -2.96483427e-01 -4.80641097e-01 1.80057526e-01
-2.90934950e-01 7.12152481e-01 7.69460976e-01 -4.10097063e-01
-6.62911177e-01 -2.43150339e-01 1.81102797e-01 -8.59327614e-02
-4.62396383e-01 6.32549822e-02 3.49303894e-02 -4.55074728e-01
-2.13431448e-01 -9.79599297e-01 -3.41015197e-02 -1.25792190e-01
4.81379777e-02 -4.32260334e-01 1.08234727e+00 -6.32910907e-01
9.71010625e-01 -1.99546599e+00 -1.12903751e-01 -9.98695865e-02
-3.70604932e-01 1.01496005e+00 2.18176153e-02 2.85484642e-01
-1.81172535e-01 -2.04707488e-01 -6.53266013e-02 9.82481763e-02
-3.82409483e-01 5.61831176e-01 7.23844320e-02 1.27285466e-01
3.08932960e-01 8.79665017e-01 -7.14968920e-01 -4.10464853e-01
5.36687613e-01 6.66054487e-01 8.33335705e-03 2.55687624e-01
-1.47825226e-01 4.74899143e-01 -3.22595656e-01 4.09107208e-01
8.46986845e-02 -1.17288888e-01 9.24840197e-02 -7.90008605e-02
-1.17716394e-01 -1.38877988e-01 -1.06752336e+00 1.85253716e+00
-3.41634005e-01 8.95353556e-01 -5.08484662e-01 -1.41056347e+00
8.63804340e-01 7.25293338e-01 5.98942578e-01 -8.52598727e-01
3.62179577e-01 1.87664613e-01 5.37834503e-02 -9.76274252e-01
3.08586746e-01 -1.11776881e-01 1.16491973e-01 1.77190140e-01
4.23234969e-01 5.81700325e-01 4.38826740e-01 -3.02653998e-01
1.24296296e+00 4.24272954e-01 3.47984970e-01 -4.65115532e-02
9.32164609e-01 -1.48094162e-01 4.25176203e-01 4.73064303e-01
-3.27941805e-01 5.13274670e-01 2.95749903e-01 -1.12025809e+00
-7.28424907e-01 -2.95172960e-01 2.08873481e-01 7.34800994e-01
-2.68813252e-01 -3.16687316e-01 -7.11861670e-01 -6.99267149e-01
-3.86106789e-01 1.55515000e-01 -7.58280337e-01 -2.38823980e-01
-6.37044370e-01 -4.70953137e-01 5.80693662e-01 9.08446491e-01
1.09880579e+00 -1.64135075e+00 -1.39054275e+00 4.33205962e-01
-1.66127950e-01 -1.24878538e+00 2.49139383e-01 1.89939439e-01
-9.63371217e-01 -1.40698290e+00 -9.55303013e-01 -6.47815287e-01
1.49403632e-01 -1.45360837e-02 8.94153714e-01 3.08146533e-02
-3.50736439e-01 3.53422672e-01 -6.15068138e-01 -5.90506673e-01
-2.28812516e-01 2.05590829e-01 -2.18646884e-01 2.63763338e-01
7.84761369e-01 -4.16004926e-01 -4.16609406e-01 2.91005909e-01
-9.56193566e-01 -1.48285821e-01 9.18418825e-01 7.40043640e-01
3.93297136e-01 -6.64157867e-02 4.50955093e-01 -6.30665064e-01
3.43662292e-01 -3.31248075e-01 -3.32017601e-01 1.99353725e-01
-2.68128008e-01 1.04448758e-01 5.12045145e-01 -4.27941591e-01
-9.13522422e-01 3.51012796e-01 -4.78530854e-01 -4.21740979e-01
-6.33445680e-01 6.62309110e-01 2.93340441e-02 -1.79717645e-01
6.37496650e-01 3.41918647e-01 8.27192664e-02 -5.72471321e-01
-3.53511274e-01 7.91589737e-01 3.58348876e-01 -1.63866565e-01
2.99676880e-02 2.15873376e-01 8.65080282e-02 -1.15301919e+00
-7.84112811e-01 -5.72666764e-01 -8.80804956e-01 -5.18453658e-01
1.32635355e+00 -7.54347682e-01 -5.18089592e-01 1.11852562e+00
-9.77841675e-01 -3.48293871e-01 -2.35334724e-01 7.54541695e-01
-6.31292403e-01 1.61195248e-01 -4.72087979e-01 -7.66573966e-01
-2.10646212e-01 -9.62884784e-01 8.11261237e-01 3.98585647e-01
-1.92237914e-01 -1.09580564e+00 2.16595426e-01 4.02928680e-01
6.06659412e-01 6.97688401e-01 2.95479864e-01 -7.57988811e-01
-2.59551585e-01 -4.00173008e-01 -1.62063688e-02 7.46897817e-01
6.24631234e-02 -2.00732216e-01 -1.05974150e+00 -1.16935924e-01
-3.14268395e-02 -6.95964634e-01 1.07520640e+00 4.42941070e-01
1.20045066e+00 -3.57852429e-02 -2.87935078e-01 2.54636258e-01
1.31944692e+00 6.30959094e-01 1.29368329e+00 5.86689174e-01
4.87193108e-01 2.19321355e-01 5.96958280e-01 2.03986570e-01
-6.41438290e-02 4.82678652e-01 4.34686571e-01 -3.03835068e-02
3.14827589e-03 4.07711267e-02 4.69686717e-01 3.91677618e-01
-6.01742983e-01 -5.36800213e-02 -9.70699430e-01 4.49017197e-01
-2.06493902e+00 -1.31818187e+00 -1.27001315e-01 2.05936480e+00
2.90589333e-01 2.99110055e-01 2.49592558e-01 7.21189737e-01
4.00143087e-01 2.98456937e-01 -2.03467354e-01 -6.14928901e-01
3.67639996e-02 4.99005318e-01 3.36604267e-01 -1.34241402e-01
-1.58038080e+00 6.34030402e-01 6.01778507e+00 5.81260681e-01
-1.46482694e+00 1.85846053e-02 5.43690205e-01 2.37821117e-02
8.18983853e-01 -3.90279979e-01 -4.24110174e-01 5.41043580e-01
1.54005504e+00 4.58226025e-01 -8.44111294e-02 8.39559078e-01
2.13159874e-01 -4.90593970e-01 -8.04127574e-01 1.09668183e+00
2.54653811e-01 -1.19790506e+00 -6.61997348e-02 -3.26746516e-02
5.64044595e-01 1.42502468e-02 -4.36479896e-01 4.52687740e-01
-2.74736911e-01 -1.31744349e+00 3.12444746e-01 7.94446409e-01
4.28211987e-01 -7.84186184e-01 1.27806580e+00 4.60167587e-01
-1.18905473e+00 -2.97780842e-01 -1.91753253e-01 -6.36027753e-01
6.36498928e-02 2.30929285e-01 -5.51951766e-01 4.97254312e-01
1.03705251e+00 1.03497422e+00 -6.20473325e-01 1.28762150e+00
-8.56672153e-02 6.09996438e-01 -1.27161652e-01 -2.49835461e-01
4.92368847e-01 7.33796274e-03 1.94821898e-02 1.21165109e+00
3.31307381e-01 6.83913529e-02 1.66100979e-01 1.83088318e-01
2.71022439e-01 -3.87295559e-02 -7.56409943e-01 -3.78276974e-01
-3.39816391e-01 1.10493219e+00 -8.04273665e-01 -4.72835362e-01
-7.09058821e-01 9.94105458e-01 -7.77247772e-02 4.78077382e-02
-9.64639246e-01 -4.45622712e-01 3.79038721e-01 1.32438406e-01
4.10585701e-01 -2.08386719e-01 1.92849845e-01 -9.64506805e-01
-1.42426595e-01 -8.37156951e-01 4.91982788e-01 -7.88282275e-01
-7.28542686e-01 7.41485000e-01 4.49891463e-02 -1.39302421e+00
-4.95317250e-01 -1.02860332e+00 -5.59256554e-01 4.03458178e-01
-1.33089316e+00 -1.19207668e+00 -5.74510276e-01 8.22741807e-01
7.29053557e-01 -2.67117083e-01 8.22543383e-01 4.59223270e-01
-4.71174419e-01 1.07312553e-01 -2.40725011e-01 6.36355639e-01
3.08060855e-01 -8.36582243e-01 -7.70646259e-02 8.26880932e-01
2.60960579e-01 2.10014552e-01 3.24488640e-01 -3.10023516e-01
-1.03015757e+00 -1.17195904e+00 8.03479135e-01 -7.59504288e-02
2.66852349e-01 2.15045229e-01 -7.59631395e-01 7.71864176e-01
4.14372146e-01 2.03313276e-01 7.47592032e-01 -3.03542912e-01
1.62401527e-01 4.55039069e-02 -1.05231571e+00 -4.30461466e-02
9.16439176e-01 -3.30289781e-01 -7.44681239e-01 1.66512579e-01
-3.77597213e-02 -2.52825588e-01 -9.59502399e-01 5.24297237e-01
8.07934821e-01 -1.17886651e+00 8.37276816e-01 -6.50449097e-01
4.82676297e-01 -2.82608032e-01 2.71138754e-02 -1.20646465e+00
-4.38025780e-02 -3.77608910e-02 -2.55435795e-01 9.14246678e-01
2.73171574e-01 -6.57492995e-01 9.29595232e-01 3.34118158e-01
-2.31725280e-03 -7.46611953e-01 -8.73770416e-01 -9.81562316e-01
-2.72282392e-01 -3.22770238e-01 3.11626524e-01 7.59309113e-01
-2.66881764e-01 3.22317153e-01 -6.13819301e-01 -5.17313659e-01
1.12810232e-01 -8.79059285e-02 7.84995675e-01 -1.30532682e+00
-5.73132671e-02 -1.10941552e-01 -1.11180305e+00 -7.77877510e-01
4.18637842e-02 -3.71781468e-01 -1.71077922e-01 -1.58780706e+00
1.34192957e-02 8.01577419e-02 -4.03323650e-01 6.96557105e-01
3.23239207e-01 3.72133821e-01 1.03256844e-01 -9.02383775e-02
-6.74967587e-01 3.64360452e-01 9.14422572e-01 -2.39029393e-01
-1.16231710e-01 2.31242552e-01 -3.88020836e-02 7.87550032e-01
1.03875875e+00 -2.56476820e-01 -3.16590488e-01 -3.22758794e-01
-1.75521135e-01 -7.35423043e-02 5.46301663e-01 -1.76190388e+00
2.53505707e-01 1.36804864e-01 7.66690969e-01 -4.52491254e-01
4.92147774e-01 -9.19144273e-01 3.15269649e-01 9.37064171e-01
-3.29531193e-01 -4.93592173e-02 6.92375675e-02 4.61592704e-01
-7.48502612e-01 -2.46340632e-01 7.56292224e-01 -5.86921692e-01
-1.42542708e+00 1.95842355e-01 -6.54277444e-01 -2.45561078e-01
1.33366132e+00 -6.31270170e-01 -1.14748357e-02 -3.02970409e-01
-8.99525642e-01 -7.52710029e-02 4.33537923e-02 6.85033798e-01
6.21146977e-01 -1.34020841e+00 -3.21363539e-01 3.14088613e-01
1.57016695e-01 -3.06135744e-01 1.88540339e-01 9.51561511e-01
-7.10013330e-01 8.26559722e-01 -7.81514108e-01 -7.24396050e-01
-1.50356257e+00 5.96811056e-01 4.81101185e-01 -3.49509895e-01
-4.09698755e-01 6.79194808e-01 -4.25322860e-01 -1.91120759e-01
3.66898328e-01 -5.96896946e-01 -8.44339669e-01 5.57968989e-02
7.94418514e-01 4.10007060e-01 1.93522602e-01 -9.29941833e-01
-4.06017840e-01 6.07980907e-01 2.33919978e-01 1.73494890e-01
1.39729297e+00 3.06585401e-01 2.09271133e-01 5.90833187e-01
1.22616386e+00 -8.26490939e-01 -1.13886642e+00 -6.67626336e-02
4.87174571e-01 -5.11879265e-01 2.70616245e-02 -6.90606773e-01
-1.26608789e+00 1.15619218e+00 1.07729030e+00 2.94707239e-01
1.13060117e+00 -4.31923836e-01 9.98292267e-01 4.87880737e-01
5.53353131e-01 -1.32498312e+00 3.32697332e-01 6.58975005e-01
6.76543474e-01 -1.25025606e+00 -2.28529841e-01 1.02977827e-01
-6.03146434e-01 1.49214280e+00 6.18081510e-01 -3.02280664e-01
7.37319887e-01 7.56758377e-02 4.67869602e-02 -1.33183926e-01
-6.38760567e-01 -4.80574489e-01 1.67076454e-01 5.38139701e-01
6.14244223e-01 -1.58567116e-01 -4.51696366e-01 3.35786462e-01
2.69161522e-01 8.70696664e-01 3.43498141e-01 1.43724382e+00
-4.62512225e-01 -1.01633883e+00 -2.40465447e-01 6.48568451e-01
-7.80104220e-01 3.86102259e-01 -5.16725302e-01 8.52014959e-01
3.78688782e-01 9.47709978e-01 3.14013250e-02 -5.73209405e-01
2.92244703e-01 4.45018858e-01 5.17984927e-01 -2.77668774e-01
-6.94184721e-01 -3.90524507e-01 2.91037500e-01 -8.88629317e-01
-1.27182758e+00 -6.12603247e-01 -1.06824958e+00 1.26578589e-05
-2.73808807e-01 7.86519125e-02 6.60716712e-01 1.33432496e+00
2.08481491e-01 5.71034074e-01 1.92944929e-01 -1.02336955e+00
-1.10614970e-01 -1.20871139e+00 -3.87435526e-01 4.64924753e-01
1.79556161e-01 -8.13020110e-01 -1.72742326e-02 2.62144953e-01] | [8.015755653381348, 0.522996723651886] |
e76c94bb-2699-4a9a-ad06-b36c8cdb9623 | vicomtech-at-ehealth-kd-challenge-2020-deep | null | null | http://ceur-ws.org/Vol-2664/eHealth-KD_paper3.pdf | http://ceur-ws.org/Vol-2664/eHealth-KD_paper3.pdf | Vicomtech at eHealth-KD Challenge 2020: Deep End-to-End Model for Entity and Relation Extraction in Medical Text | This paper describes the participation of the Vicomtech NLP team in the eHealth-KD 2020 shared task about
detecting and classifying entities and relations in health-related texts written in Spanish. The proposed system
consists of a single end-to-end deep neural network with pre-trained BERT models as the core for the semantic
representation of the input texts. We have experimented with two models: BERT-Base Multilingual Cased and
BETO, a BERT model pre-trained on Spanish text. Our system models all the output variables—entities and
relations—at the same time, modelling the whole problem jointly. Some of the outputs are fed back to latter
layers of the model, connecting the outcomes of the different subtasks in a pipeline fashion. Our system shows
robust results in all the scenarios of the task. It has achieved the first position in the main scenario of the
competition and top-3 in the rest of the scenarios. | ['Montse Cuadros and Elena Zotova', 'Naiara Perez', 'Aitor García-Pablos'] | 2020-09-20 | null | null | null | null | ['medical-procedure', 'multi-label-classification-of-biomedical'] | ['medical', 'medical'] | [-3.31122041e-01 1.03655887e+00 2.92199496e-02 -5.36442578e-01
-6.30631268e-01 -2.22266361e-01 7.92336702e-01 4.46393132e-01
-7.91870058e-01 7.90466785e-01 4.23609406e-01 -4.55991417e-01
-1.92442492e-01 -6.17439687e-01 -7.03608990e-01 -1.32774413e-01
-1.06786393e-01 1.46803057e+00 9.40473601e-02 -4.27699566e-01
-6.16476119e-01 7.16566667e-02 -7.38174915e-01 9.72749233e-01
4.93131757e-01 8.94485176e-01 2.90327668e-02 8.75151932e-01
3.31862830e-02 1.39105988e+00 -2.57503152e-01 -8.06942761e-01
-6.62515238e-02 -1.79168627e-01 -1.42630577e+00 -7.07114697e-01
-4.70289476e-02 1.19273551e-01 -4.49591398e-01 7.55600691e-01
7.41400301e-01 -1.55555680e-01 5.82157433e-01 -1.13579655e+00
-1.26458570e-01 1.18451881e+00 3.88541594e-02 5.67056872e-02
4.11114603e-01 -2.13995382e-01 1.10069561e+00 -8.44120681e-01
1.07497752e+00 1.37455606e+00 8.51589024e-01 5.15735090e-01
-8.58702600e-01 -3.04758638e-01 -8.16486105e-02 3.42945218e-01
-1.18073535e+00 -5.74294567e-01 1.69615269e-01 -5.24761438e-01
1.55581188e+00 1.22246630e-01 3.71664912e-01 1.09399092e+00
1.90052941e-01 8.07147503e-01 6.66115284e-01 -4.06197667e-01
-7.26557672e-02 5.14873981e-01 2.65043288e-01 6.64199769e-01
1.57028794e-01 -1.59275159e-01 -2.33685687e-01 -1.00724757e-01
6.78145066e-02 -3.49820375e-01 -1.54457942e-01 -9.35144052e-02
-1.13046360e+00 7.42429793e-01 7.52368510e-01 6.60859764e-01
-8.19247663e-01 2.98995320e-02 7.22071350e-01 3.30488086e-01
5.72137415e-01 3.92544270e-01 -1.06611693e+00 2.24739194e-01
-8.96156609e-01 4.89984781e-01 1.40525877e+00 1.02079713e+00
5.55300117e-02 -9.38271999e-01 -4.64747101e-01 7.32522666e-01
4.36449915e-01 -8.69930014e-02 3.91133636e-01 -4.18135971e-01
9.00261700e-01 5.46457767e-01 2.66255021e-01 -7.36712754e-01
-1.21588945e+00 -4.63790983e-01 -6.05563700e-01 -4.03850704e-01
4.97904688e-01 -8.21029186e-01 -8.89320850e-01 1.75533009e+00
4.58214670e-01 -1.29417181e-01 4.61267322e-01 5.83045959e-01
1.55604708e+00 4.63538408e-01 6.69276893e-01 1.60500512e-01
1.78804576e+00 -1.36570168e+00 -9.71441209e-01 -1.07030541e-01
9.91085351e-01 -5.48880994e-01 2.15553999e-01 2.57987771e-02
-1.22761869e+00 -3.63754421e-01 -7.96130657e-01 -5.45949876e-01
-1.00378799e+00 3.87124687e-01 2.44651794e-01 3.72002050e-02
-1.34755373e+00 3.90585899e-01 -7.80891120e-01 -6.37950480e-01
1.56803519e-01 3.66636127e-01 -4.55666870e-01 6.52014315e-02
-1.85180855e+00 1.49110162e+00 7.85169601e-01 3.44303459e-01
-8.37831259e-01 -6.51396751e-01 -9.51076329e-01 2.38095537e-01
2.05788493e-01 -1.04296505e+00 1.43246078e+00 -6.96261704e-01
-9.74003494e-01 1.41860640e+00 -9.34448373e-03 -8.92767906e-01
1.11706746e+00 -4.48630720e-01 -6.35917604e-01 3.01160961e-02
3.10232043e-01 9.12551224e-01 1.39132664e-02 -6.35330021e-01
-6.15794361e-01 -2.71877617e-01 -1.43463037e-03 1.75125659e-01
1.71240568e-01 3.56291533e-01 -5.76697409e-01 -2.92913049e-01
-6.19134128e-01 -7.11965263e-01 -5.02393723e-01 -4.71072286e-01
-7.75225222e-01 -5.79043865e-01 2.64281809e-01 -1.02710378e+00
9.39508617e-01 -2.00587773e+00 2.51527607e-01 6.80103479e-03
2.93705434e-01 3.32874686e-01 -2.81082448e-02 8.01356494e-01
-3.87512296e-01 -4.98961918e-02 8.91302004e-02 -5.33247650e-01
5.81798069e-02 3.81347686e-02 9.70529318e-02 2.02415392e-01
3.31244707e-01 1.18734419e+00 -1.05809677e+00 -5.59932947e-01
2.07572412e-02 5.36486208e-01 -2.09430754e-01 2.40248516e-01
-5.09572446e-01 3.51073891e-01 -4.87489849e-01 -9.29862782e-02
4.17848557e-01 -3.21215123e-01 5.71142733e-01 -1.90975189e-01
3.83205302e-02 7.86565006e-01 -8.62389505e-01 1.65875661e+00
-5.35434127e-01 4.26764965e-01 1.67780802e-01 -1.05981004e+00
5.27836263e-01 9.61064160e-01 5.46560943e-01 -6.67640746e-01
3.03044081e-01 4.19378281e-01 8.09915587e-02 -1.08029234e+00
2.77444124e-01 -1.06707312e-01 -2.12739900e-01 -1.99834374e-03
5.94222546e-01 4.52239364e-01 3.05819303e-01 2.70883054e-01
1.05140543e+00 1.65206537e-01 6.76705360e-01 -3.60431999e-01
7.57764518e-01 2.66376436e-01 3.05991262e-01 5.72544575e-01
-7.59584159e-02 3.85775417e-01 9.04149592e-01 -6.78141415e-01
-8.98506343e-01 -5.02099752e-01 -2.67327994e-01 1.00023890e+00
-2.75115609e-01 -5.45517743e-01 -7.83327043e-01 -1.13079536e+00
-1.85310543e-01 8.82431507e-01 -1.04691410e+00 1.89002350e-01
-5.47608554e-01 -6.20407999e-01 8.04751217e-01 4.31720495e-01
4.25583363e-01 -1.38980997e+00 -5.71448326e-01 5.25480092e-01
-3.19631457e-01 -1.38284910e+00 2.70469993e-01 6.03952229e-01
-1.30539715e-01 -1.35729456e+00 -8.84410143e-01 -9.80316341e-01
1.15586199e-01 -7.45642602e-01 1.37629545e+00 -3.42287630e-01
-2.21173689e-01 -1.87086567e-01 -2.61007100e-01 -8.82159710e-01
-6.20294631e-01 5.03425360e-01 -6.50746167e-01 -1.83823407e-01
5.99910557e-01 1.08299755e-01 -5.76663554e-01 -1.05136476e-01
-5.84152281e-01 3.62132341e-01 3.30163747e-01 8.57774317e-01
1.45087838e-01 -2.90254503e-01 6.16421402e-01 -1.38664472e+00
5.20889819e-01 -8.80709946e-01 -2.45646238e-01 4.22109634e-01
-1.79481804e-01 -8.23150203e-03 5.93883932e-01 1.29640922e-01
-9.27311599e-01 7.32691363e-02 -7.47075081e-01 2.42846057e-01
-3.12594593e-01 9.55849588e-01 -3.39020491e-01 7.85998881e-01
6.85137928e-01 -5.30950487e-01 -1.77933976e-01 -7.28758037e-01
6.83671236e-01 9.46220875e-01 4.00046349e-01 -4.83639538e-02
9.70523581e-02 2.30031252e-01 -2.14762568e-01 -5.01208663e-01
-1.06914890e+00 -5.76739669e-01 -6.47091389e-01 1.48672193e-01
1.34453547e+00 -1.34306800e+00 -3.94592315e-01 5.07031024e-01
-1.46314669e+00 -5.50561607e-01 -3.42447132e-01 4.52563405e-01
-2.53225327e-01 -4.14955288e-01 -9.23821926e-01 -4.15677547e-01
-6.78610384e-01 -1.02626956e+00 1.06296277e+00 -1.74856989e-03
-4.33829933e-01 -1.24009907e+00 3.27196389e-01 2.89596409e-01
3.63817066e-01 3.45048755e-01 1.00803268e+00 -1.50752497e+00
3.63715023e-01 -3.90966088e-01 -3.15251261e-01 2.34100521e-02
-4.42030191e-01 -3.34852248e-01 -1.08148336e+00 1.08786024e-01
-1.82107866e-01 -4.76213306e-01 1.00374258e+00 3.26380879e-01
7.21666992e-01 -3.12571049e-01 -6.20938778e-01 2.95184225e-01
1.25152838e+00 -5.60242683e-02 4.52121884e-01 5.08846343e-01
5.52604735e-01 1.03780210e+00 3.96226972e-01 -9.45174042e-03
9.66700852e-01 6.71841919e-01 3.41269493e-01 -5.75115919e-01
-4.82894294e-02 -1.39732778e-01 8.61106515e-02 4.20029074e-01
3.29898417e-01 -4.97234821e-01 -1.30976474e+00 6.77017450e-01
-2.17980647e+00 -6.92330539e-01 -4.99173313e-01 1.50126898e+00
9.35437679e-01 -2.63907462e-02 2.24628687e-01 -3.05182219e-01
4.57263976e-01 -1.72117574e-03 -1.01300210e-01 -6.19575977e-01
3.38411075e-03 3.88780922e-01 2.55468637e-01 6.03852093e-01
-1.51606786e+00 1.07485998e+00 5.81314421e+00 4.12775725e-01
-9.56824303e-01 4.11498338e-01 6.21257067e-01 -1.65612679e-02
1.21491909e-01 -2.28994891e-01 -8.37951958e-01 1.16464935e-01
1.51220882e+00 3.27954106e-02 -1.04492493e-01 6.50031090e-01
1.42383799e-01 1.90796301e-01 -1.31572211e+00 3.72799248e-01
-1.34550229e-01 -1.33474624e+00 -2.22374082e-01 -1.76970139e-01
4.38084602e-01 7.84746885e-01 -4.31365579e-01 6.73197448e-01
5.72528183e-01 -1.28303659e+00 8.98131073e-01 7.14394867e-01
8.39048386e-01 -4.13259596e-01 1.41042864e+00 4.50868994e-01
-8.38350296e-01 -3.36671285e-02 -5.43091893e-02 3.00809532e-01
3.46021086e-01 5.58123469e-01 -1.20900142e+00 1.04055107e+00
6.13154709e-01 7.42866337e-01 -4.66174573e-01 1.04563844e+00
-4.39093590e-01 3.99763197e-01 -2.27475479e-01 -6.57032058e-02
4.89778370e-01 3.78657967e-01 3.39898676e-01 1.79351723e+00
-8.93745050e-02 -1.98843166e-01 -6.96572140e-02 6.92593038e-01
-4.41459507e-01 3.12261969e-01 -4.82468992e-01 -1.79643501e-02
7.65002742e-02 1.28024852e+00 -3.33455920e-01 -7.33177960e-01
-4.27148730e-01 7.45662153e-01 5.95526397e-01 2.10060552e-01
-7.59239435e-01 -3.73968691e-01 2.69966964e-02 -1.37373628e-02
2.60304779e-01 4.92829442e-01 -1.19514190e-01 -9.57411528e-01
-2.15784475e-01 -9.92394269e-01 8.74053657e-01 -7.36611664e-01
-1.10062206e+00 9.14680302e-01 -1.19112730e-01 -4.63794142e-01
-5.99053442e-01 -8.53173673e-01 -2.70324945e-01 1.26535618e+00
-1.39450109e+00 -1.38988435e+00 -3.75039615e-02 5.64775944e-01
2.69834906e-01 3.87150049e-03 1.26629841e+00 6.52806461e-01
-6.33332729e-01 4.20713663e-01 4.95684780e-02 5.32808244e-01
8.41539145e-01 -1.50886786e+00 5.86353540e-01 3.37026209e-01
-7.26647750e-02 4.73152727e-01 6.51048660e-01 -5.79468489e-01
-5.98412931e-01 -1.30929482e+00 2.03652859e+00 -3.95978540e-01
7.58194923e-01 -5.97910404e-01 -5.32094657e-01 1.03839338e+00
6.16054475e-01 -1.89565971e-01 5.14133990e-01 2.37086296e-01
-5.32867201e-02 2.53818661e-01 -1.14429939e+00 1.72819123e-01
7.28896916e-01 -5.36457598e-01 -7.50169933e-01 7.78042793e-01
9.43017423e-01 -6.81273162e-01 -9.92318332e-01 4.63929087e-01
4.63526577e-01 -4.76342231e-01 7.64382780e-01 -1.18737543e+00
6.71090901e-01 3.61285917e-02 2.69483123e-02 -1.33465159e+00
-1.44668698e-01 -2.95201510e-01 1.81757787e-03 1.02599180e+00
1.13521016e+00 -4.79064137e-01 3.17628354e-01 4.16194648e-01
-3.91311981e-02 -7.68702209e-01 -9.57412720e-01 -1.51997000e-01
1.80268303e-01 -3.03658575e-01 4.24536467e-01 1.02846873e+00
3.90308835e-02 9.97206509e-01 -5.01321405e-02 -3.79068893e-03
6.64332584e-02 -3.49683523e-01 5.93690634e-01 -1.24965537e+00
-4.27061230e-01 -4.35094923e-01 -1.23901643e-01 -5.39265931e-01
8.37590024e-02 -1.27959526e+00 -1.03942715e-01 -2.08179164e+00
2.42144465e-01 -2.72163898e-01 -2.61443019e-01 6.49468005e-01
-3.40663761e-01 -1.85393274e-01 9.98589024e-02 2.23106399e-04
-6.85040295e-01 1.98329985e-01 7.55609214e-01 -2.21060798e-01
-3.80100794e-02 -2.40042750e-02 -7.31479645e-01 5.88522077e-01
4.00374770e-01 -5.05201995e-01 -6.17627539e-02 -6.55736148e-01
4.53216076e-01 1.66130409e-01 4.11033392e-01 -5.89801610e-01
8.97426903e-02 4.77805108e-01 3.92445922e-01 -6.69717610e-01
1.16380937e-02 -9.15907085e-01 4.18153375e-01 6.51015341e-01
-8.39501441e-01 -4.83593345e-03 2.15774655e-01 1.71802223e-01
-2.69074976e-01 -2.99023896e-01 5.55439651e-01 -2.16351137e-01
-2.02280343e-01 1.04875781e-01 -1.07894041e-01 2.34072745e-01
9.46098506e-01 6.02132618e-01 -4.94271070e-01 -1.13223337e-01
-1.52245033e+00 8.08012724e-01 -2.56255150e-01 4.18873042e-01
-2.20727101e-01 -7.36784756e-01 -1.13712776e+00 -1.45312011e-01
6.50380254e-02 2.62157172e-01 1.78390220e-01 1.18703735e+00
-8.69533598e-01 9.07377541e-01 3.01727206e-02 -3.45405430e-01
-1.03116047e+00 5.99208295e-01 9.60697889e-01 -1.23047268e+00
-5.20807564e-01 9.63219941e-01 1.09262824e-01 -8.83740902e-01
4.67505306e-01 -2.00418219e-01 -7.96405971e-01 3.50592613e-01
5.42252004e-01 2.11533412e-01 4.38388020e-01 -7.07319260e-01
-5.23924172e-01 -2.20428869e-01 -2.02293083e-01 -3.68388742e-01
1.79876077e+00 1.71316668e-01 -2.10769579e-01 3.39976639e-01
1.43974543e+00 -3.95454168e-01 -5.17688215e-01 -2.33476356e-01
4.64010864e-01 7.25358307e-01 -8.49642158e-02 -1.55108106e+00
-8.73942196e-01 9.09559429e-01 2.80131847e-01 2.52420962e-01
7.29497373e-01 3.16896290e-01 5.95171392e-01 2.60951191e-01
2.03392711e-02 -9.13143396e-01 -6.89190149e-01 6.60437107e-01
1.02223432e+00 -9.84130323e-01 -3.04084301e-01 -2.87598878e-01
-6.50072634e-01 9.38346326e-01 2.97923356e-01 9.06826183e-02
7.79405296e-01 2.33394101e-01 2.61792928e-01 -7.02657998e-01
-1.02850771e+00 -2.36583233e-01 4.96141583e-01 4.26072121e-01
9.43286121e-01 3.47887963e-01 -7.41349757e-01 8.86811554e-01
-2.52819747e-01 2.18983576e-01 -1.54899940e-01 6.09809697e-01
1.77919552e-01 -1.05617714e+00 2.87608445e-01 2.49888346e-01
-9.44304168e-01 -4.03727621e-01 -7.74660826e-01 1.11734354e+00
3.42693329e-01 9.32085454e-01 -1.57277271e-01 -1.56366929e-01
8.77414525e-01 4.75344658e-01 8.97889361e-02 -6.44597650e-01
-1.32702780e+00 3.60361561e-02 9.17242825e-01 -6.74927652e-01
-3.21773797e-01 -6.46735847e-01 -1.43070161e+00 2.03820482e-01
-6.52361847e-03 4.30196643e-01 5.94334543e-01 1.07795036e+00
5.32614529e-01 9.18132722e-01 7.94711635e-02 -3.90308648e-01
-4.16119903e-01 -1.54447925e+00 -2.58845747e-01 5.48077881e-01
2.98155993e-01 -2.49087647e-01 2.26750374e-01 -1.45300388e-01] | [8.796296119689941, 8.794349670410156] |
b71b7123-e520-4b3d-80c3-24d443af6138 | an-iterative-similarity-based-adaptation | null | null | https://aclanthology.org/K15-1006 | https://aclanthology.org/K15-1006.pdf | An Iterative Similarity based Adaptation Technique for Cross-domain Text Classification | null | ['Shourya Roy', 'Himanshu Sharad Bhatt', 'Deepali Semwal'] | 2015-07-01 | null | null | null | conll-2015-7 | ['cross-domain-text-classification'] | ['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] | [-7.278365135192871, 3.7177627086639404] |
3a9021d6-72d7-4416-9f55-5165f00b4a60 | deeptilebars-visualizing-term-distribution | 1811.00606 | null | https://arxiv.org/abs/1811.00606v3 | https://arxiv.org/pdf/1811.00606v3.pdf | DeepTileBars: Visualizing Term Distribution for Neural Information Retrieval | Most neural Information Retrieval (Neu-IR) models derive query-to-document ranking scores based on term-level matching. Inspired by TileBars, a classical term distribution visualization method, in this paper, we propose a novel Neu-IR model that handles query-to-document matching at the subtopic and higher levels. Our system first splits the documents into topical segments, "visualizes" the matchings between the query and the segments, and then feeds an interaction matrix into a Neu-IR model, DeepTileBars, to obtain the final ranking scores. DeepTileBars models the relevance signals occurring at different granularities in a document's topic hierarchy. It better captures the discourse structure of a document and thus the matching patterns. Although its design and implementation are light-weight, DeepTileBars outperforms other state-of-the-art Neu-IR models on benchmark datasets including the Text REtrieval Conference (TREC) 2010-2012 Web Tracks and LETOR 4.0. | ['Zhiwen Tang', 'Grace Hui Yang'] | 2018-11-01 | null | null | null | null | ['ad-hoc-information-retrieval'] | ['natural-language-processing'] | [-1.85933858e-01 -6.46590367e-02 -3.38495344e-01 -2.53605306e-01
-9.75867033e-01 -6.08181059e-01 1.09280741e+00 6.15742505e-01
-1.41795516e-01 2.79183481e-02 8.64626706e-01 -5.65719664e-01
-7.18941629e-01 -6.32937431e-01 -1.57622993e-01 -3.49627763e-01
-3.43358099e-01 6.54515266e-01 3.15842986e-01 -4.27712619e-01
7.55416930e-01 1.04518950e-01 -1.86746907e+00 1.20981288e+00
7.85869420e-01 9.08939362e-01 6.44922480e-02 5.44868529e-01
-7.87943959e-01 5.71275234e-01 -9.64022219e-01 -1.40932873e-01
-3.39117438e-01 -1.35818318e-01 -8.99085999e-01 -9.93282080e-01
6.65266693e-01 8.63448679e-02 -1.41256943e-01 1.03843319e+00
4.63971376e-01 1.51492968e-01 9.98792171e-01 -9.40558195e-01
-1.07344067e+00 9.17990446e-01 -8.28789294e-01 1.15533888e-01
5.78697264e-01 -9.78252947e-01 1.31521666e+00 -1.13465953e+00
9.30137098e-01 1.82969034e+00 1.77686006e-01 2.21341416e-01
-1.16103780e+00 -5.58003843e-01 3.20934057e-01 3.57123882e-01
-1.04352462e+00 3.27661127e-01 7.70697713e-01 -6.75018489e-01
8.36665154e-01 9.10164654e-01 5.95254183e-01 7.72299826e-01
8.53410438e-02 9.88252997e-01 7.37891257e-01 -6.84274197e-01
1.36142850e-01 2.75627542e-02 9.36305165e-01 3.22745234e-01
1.21557750e-02 8.38019922e-02 -7.94145703e-01 -4.95420635e-01
3.51266861e-01 2.05401942e-01 -2.60227472e-01 -1.91848323e-01
-1.18158245e+00 6.81999683e-01 1.05081594e+00 6.53889656e-01
-1.58970103e-01 6.96294382e-03 6.00277901e-01 4.39683944e-01
7.95305908e-01 5.19623518e-01 1.70595616e-01 2.90557295e-01
-1.06912243e+00 5.78739941e-01 4.96786743e-01 6.94599628e-01
4.27611709e-01 -6.36803031e-01 -1.24953377e+00 1.21346915e+00
5.95750690e-01 9.92519874e-03 7.06826031e-01 -6.56283557e-01
4.50147599e-01 1.06902909e+00 4.45596278e-02 -1.23523748e+00
-4.00666654e-01 -6.84964478e-01 -7.04311728e-01 1.48407415e-01
-6.83699250e-02 3.30127746e-01 -9.83828843e-01 1.15634644e+00
8.67736116e-02 -5.82180619e-01 -5.28137013e-02 9.01505291e-01
1.50071084e+00 1.01688552e+00 1.34704500e-01 1.79853886e-01
1.83113372e+00 -8.24906588e-01 -9.37524080e-01 -2.01449671e-04
7.95663476e-01 -6.29767716e-01 1.20722842e+00 3.02361976e-02
-8.77402723e-01 -3.38087559e-01 -1.10629976e+00 -4.94049221e-01
-8.32900524e-01 2.56884187e-01 2.96071172e-01 1.61123738e-01
-1.29544425e+00 2.42965654e-01 -2.68537462e-01 -2.09439144e-01
-6.44047335e-02 -1.88719004e-01 1.93848148e-01 6.50082752e-02
-1.57305682e+00 7.72664666e-01 3.32842708e-01 -9.43899993e-03
-5.54210603e-01 -9.72603321e-01 -5.23007870e-01 6.05827987e-01
7.09603205e-02 -6.01980329e-01 1.27432323e+00 -1.36783674e-01
-9.60287213e-01 6.04014993e-01 -1.97426960e-01 5.52444346e-02
2.30070930e-02 -3.16325694e-01 -4.27713782e-01 -6.35387227e-02
-1.29713714e-01 8.34008634e-01 2.03481719e-01 -1.46085429e+00
-7.13168025e-01 -6.95094824e-01 4.81309891e-02 6.21814787e-01
-6.57199204e-01 4.77275997e-01 -1.07422531e+00 -7.44046986e-01
2.28582501e-01 -4.09129173e-01 1.58404738e-01 -1.88806757e-01
-5.50437450e-01 -1.09168887e+00 1.03344488e+00 -5.42505503e-01
1.87511778e+00 -2.13074327e+00 -1.98822692e-02 4.51287806e-01
4.06805217e-01 4.97849323e-02 -1.93051353e-01 7.89532542e-01
-6.62363395e-02 3.53953987e-01 4.53733623e-01 -2.63026714e-01
3.49684328e-01 -4.76362109e-01 -4.22284037e-01 -2.55590826e-01
-5.49047291e-01 1.10901845e+00 -8.07377994e-01 -7.45594084e-01
-1.29431605e-01 7.49462962e-01 -3.38101596e-01 5.27480245e-02
-3.20966303e-01 -3.77216071e-01 -4.28609937e-01 6.40448689e-01
3.86912674e-01 -3.74515921e-01 2.61083782e-01 -2.62623817e-01
-2.99316853e-01 5.25301456e-01 -6.13747001e-01 1.58569670e+00
-1.65129080e-01 1.18263900e+00 -2.91852564e-01 -5.58118999e-01
1.23958313e+00 2.46602207e-01 1.43033773e-01 -1.45132351e+00
-2.15597451e-01 7.88408890e-02 -4.55406487e-01 -2.54300028e-01
1.08033240e+00 6.09053910e-01 -2.02693939e-01 5.57850718e-01
-2.17194542e-01 4.05445993e-01 5.59326172e-01 6.65664077e-01
8.01998436e-01 -1.05417550e-01 -1.63631514e-01 -5.92599332e-01
1.16502553e-01 -2.57101245e-02 -1.76129699e-01 9.87800896e-01
5.54110348e-01 4.81815875e-01 6.98337495e-01 -5.36404729e-01
-5.65652490e-01 -8.53092015e-01 -1.67863458e-01 1.86738217e+00
1.61393926e-01 -7.87924588e-01 -6.70067668e-01 -4.55032647e-01
8.33083540e-02 7.91734576e-01 -1.07773137e+00 -3.01862527e-02
-3.53192359e-01 -4.75510210e-01 3.51817548e-01 1.75938368e-01
1.60099998e-01 -1.24708450e+00 -5.88880002e-01 -8.65052938e-02
-3.79433155e-01 1.23204567e-01 -4.99040723e-01 1.43918231e-01
-8.43441188e-01 -7.45714366e-01 -1.24736130e+00 -9.48498428e-01
4.00276184e-01 3.02638084e-01 1.53343928e+00 2.68857623e-03
-3.18329722e-01 6.40245527e-02 -4.06024694e-01 -4.38589245e-01
-7.25737726e-03 1.20926499e-01 -7.51919568e-01 -3.69504809e-01
4.88546759e-01 2.50258972e-03 -1.05489397e+00 3.21307153e-01
-9.76312280e-01 9.71636474e-02 3.07568759e-01 5.12962818e-01
3.92490923e-01 -2.34734848e-01 2.22202823e-01 -8.09170902e-01
1.62063873e+00 -4.48413879e-01 -4.88879889e-01 1.05167460e+00
-1.01764786e+00 3.17142636e-01 -1.07621081e-01 -2.38099858e-01
-1.19760942e+00 -7.66573906e-01 4.86162663e-01 -1.47983789e-01
3.29823703e-01 1.18560958e+00 4.23590392e-01 5.00274003e-01
1.01893675e+00 -7.83620998e-02 -4.89103079e-01 -9.67470706e-01
5.55385768e-01 9.21734035e-01 6.33621156e-01 -5.70075750e-01
2.34359369e-01 8.60952660e-02 -3.72566670e-01 -2.91573197e-01
-7.28583753e-01 -8.06568801e-01 -7.09414110e-02 -5.84080517e-01
5.15626073e-01 -5.47787368e-01 -8.75523150e-01 -8.17250237e-02
-1.48078620e+00 2.37725928e-01 -7.85695240e-02 1.44067734e-01
2.26182789e-01 -9.44837276e-03 -5.00053942e-01 -8.54695618e-01
-8.38123143e-01 -7.65680254e-01 1.31930614e+00 3.98051023e-01
-3.86179745e-01 -1.02426136e+00 6.04920030e-01 -3.65052149e-02
6.20620251e-01 -5.91422468e-02 1.55544388e+00 -6.61878526e-01
-1.97739318e-01 -2.45483816e-01 -6.48228347e-01 -5.19237459e-01
-2.13132069e-01 9.79493931e-02 -8.36853504e-01 -1.81810081e-01
-5.46052754e-01 -1.72984570e-01 1.21453321e+00 4.56556052e-01
1.49233079e+00 -5.85551679e-01 -7.30301440e-01 1.03470333e-01
1.08639288e+00 5.85248113e-01 6.87846184e-01 7.34709024e-01
4.27972525e-01 1.13011360e+00 5.36652088e-01 1.13573708e-01
3.86121750e-01 9.51009512e-01 5.02910316e-01 -3.73943180e-01
7.05492275e-04 -3.35850894e-01 -9.96384397e-03 6.15730345e-01
1.88553929e-01 -6.06186867e-01 -9.68195140e-01 3.59667003e-01
-2.13291907e+00 -9.03413534e-01 -3.22193950e-02 2.15096188e+00
8.63744617e-01 -2.06239849e-01 -1.32813066e-01 -9.47877094e-02
7.91267216e-01 3.36613089e-01 -4.20608342e-01 -4.81560469e-01
1.75757423e-01 -2.12619826e-01 -1.62951186e-01 4.30814177e-01
-8.53445172e-01 3.81260544e-01 6.44782400e+00 1.18264759e+00
-8.82707775e-01 -3.08622837e-01 6.89436078e-01 -3.41300815e-01
-8.13153088e-01 -2.23605528e-01 -7.97004998e-01 3.15227538e-01
7.58276224e-01 -4.72871602e-01 -2.88089341e-03 7.18540132e-01
-2.82133728e-01 1.06540680e-01 -1.18148100e+00 9.56740558e-01
1.19029276e-01 -1.64977574e+00 6.24963939e-01 3.44176739e-02
3.30935359e-01 -1.67981938e-01 2.91065603e-01 5.38260877e-01
5.71875274e-01 -1.02857840e+00 7.61143565e-01 7.94303536e-01
7.85293281e-01 -7.31592119e-01 6.75954878e-01 1.35460198e-01
-1.17994213e+00 -3.17720994e-02 -3.40753555e-01 5.35127401e-01
-2.16873214e-01 4.72037256e-01 -6.15407109e-01 7.49166727e-01
1.13431978e+00 6.08106077e-01 -9.34704185e-01 1.20117414e+00
2.83330172e-01 2.75061041e-01 1.89501196e-02 -5.88103056e-01
2.37774387e-01 5.79767600e-02 4.24965143e-01 1.70770144e+00
6.63305342e-01 -3.24933767e-01 -1.01455547e-01 9.27113831e-01
-2.08727822e-01 2.51078635e-01 -3.14426869e-01 -6.65947124e-02
7.69868433e-01 1.32087111e+00 -4.24062461e-01 -4.94505376e-01
6.81219772e-02 4.29828048e-01 3.30657363e-01 6.81562245e-01
-1.76148396e-02 -9.60426390e-01 1.76679626e-01 -6.21047392e-02
-2.63902247e-01 3.11338633e-01 -2.54389733e-01 -5.61683595e-01
-1.16315242e-02 -6.65187299e-01 9.21547592e-01 -1.08502591e+00
-9.80076492e-01 8.82230878e-01 5.10123789e-01 -1.34567976e+00
-5.98514378e-01 -4.44922209e-01 -6.25796139e-01 1.17803872e+00
-1.27418518e+00 -8.74882102e-01 -3.24376136e-01 3.29259038e-01
3.10758829e-01 -1.87767103e-01 9.52058733e-01 9.53783318e-02
-3.40062767e-01 5.54821670e-01 5.65334976e-01 4.90018651e-02
5.62113822e-01 -1.47645664e+00 4.68891174e-01 2.11599082e-01
4.33293521e-01 1.12698209e+00 5.01558065e-01 -5.77875912e-01
-7.29920030e-01 -6.09142959e-01 1.19719708e+00 -2.26379737e-01
4.15496707e-01 -4.78154808e-01 -1.11624038e+00 7.49727488e-02
7.11426377e-01 -6.28796518e-01 7.60845482e-01 6.80787086e-01
-9.07411933e-01 -1.30080983e-01 -5.15376627e-01 8.19956541e-01
7.03061700e-01 -6.96409643e-01 -7.28597581e-01 3.05889308e-01
9.79209125e-01 -2.01074108e-01 -6.59533203e-01 3.96650314e-01
9.63648558e-01 -7.84484267e-01 1.33016145e+00 -5.06378591e-01
3.52896214e-01 -2.05616117e-01 4.12270278e-02 -1.37133420e+00
-6.94329500e-01 -2.00896606e-01 -3.70822072e-01 1.08759058e+00
6.97388589e-01 -1.41729072e-01 5.39128304e-01 3.27214628e-01
-2.24617749e-01 -7.40162671e-01 -8.02169323e-01 -2.93675989e-01
1.12105802e-01 -3.16360994e-04 7.05518842e-01 7.96927512e-01
6.94667161e-01 3.93819511e-01 9.29511804e-03 -4.17547345e-01
4.31691051e-01 3.34291488e-01 7.28954151e-02 -1.76563966e+00
1.71927840e-01 -1.13516998e+00 2.18710035e-01 -1.06758273e+00
-3.91530007e-01 -1.16967404e+00 -6.15763180e-02 -2.18259978e+00
5.46592951e-01 -2.26993635e-01 -9.13206637e-01 4.22598600e-01
-1.47949234e-01 2.30496731e-02 1.52783230e-01 7.69352257e-01
-8.81334662e-01 3.97559434e-01 1.15569735e+00 -5.94619215e-01
-3.24599385e-01 -3.77252549e-01 -8.28403533e-01 2.00410828e-01
4.08887148e-01 -4.19885010e-01 -5.47538996e-01 -7.11612940e-01
7.80759454e-01 1.37662604e-01 2.81263560e-01 -5.55317998e-01
7.71528959e-01 1.69621915e-01 4.40492243e-01 -1.38688934e+00
9.34492573e-02 -4.11580920e-01 -9.65561420e-02 2.68869281e-01
-1.35287142e+00 3.53704482e-01 1.15290374e-01 4.25111204e-01
-5.43215930e-01 -2.66831219e-01 1.86257303e-01 1.78179651e-01
-2.93140143e-01 -1.14887536e-01 -3.71221572e-01 -1.94949582e-01
4.26663384e-02 1.28963053e-01 -1.14958465e+00 -3.55519623e-01
-3.60682011e-01 5.71660519e-01 -9.69731659e-02 8.51953745e-01
6.93601429e-01 -1.61220562e+00 -6.05559587e-01 -1.11287527e-01
6.41933441e-01 -1.00343883e-01 2.76258767e-01 1.99868038e-01
-1.51620135e-01 1.16057599e+00 1.44140586e-01 -6.80575371e-01
-1.60823178e+00 3.05636942e-01 9.03075561e-02 -7.62592971e-01
-4.95476604e-01 7.34311402e-01 6.70970023e-01 -4.77254003e-01
9.34160888e-01 -1.20066434e-01 -1.16165316e+00 5.26426673e-01
1.02058959e+00 4.92834121e-01 3.10137004e-01 -5.65441065e-02
-2.84119070e-01 5.82935989e-01 -6.04023457e-01 -5.77895939e-01
8.64925921e-01 2.84480285e-02 -2.16073468e-01 5.01032531e-01
1.33159089e+00 -4.65233773e-01 -4.94677454e-01 -6.07442617e-01
7.34537780e-01 -2.67773062e-01 3.21154118e-01 -1.38562787e+00
-7.07758605e-01 1.01553869e+00 8.59020293e-01 6.88954949e-01
9.50100005e-01 2.58069366e-01 1.58070922e-01 7.71269560e-01
-1.56048909e-01 -1.23731482e+00 3.51766437e-01 6.27310634e-01
1.49813545e+00 -6.66747510e-01 -1.00607581e-01 6.63478971e-02
-3.04385662e-01 1.09044325e+00 3.97113979e-01 4.25917149e-01
5.42232275e-01 -9.94390622e-03 3.48882675e-01 -8.73477340e-01
-1.01659799e+00 4.60273251e-02 1.23403180e+00 3.10969204e-01
1.09159482e+00 -6.95376620e-02 -4.95789111e-01 2.42174059e-01
-2.34101251e-01 -2.28786469e-01 -3.91667157e-01 8.11042190e-01
-6.12705469e-01 -9.72818255e-01 -3.84773701e-01 5.10039330e-01
-3.25906873e-01 -3.19273025e-01 -8.39035273e-01 6.10407591e-01
-6.23212576e-01 8.71442735e-01 5.38231552e-01 -4.60729420e-01
3.14816415e-01 1.69255018e-01 -2.87135541e-01 -5.13349771e-01
-9.04224634e-01 3.93908262e-01 1.45979360e-01 -5.88321924e-01
-2.20122904e-01 -2.48108655e-01 -1.05574930e+00 -4.68415469e-02
-1.46248221e-01 7.20672548e-01 1.05682683e+00 3.45382601e-01
7.71973193e-01 8.74697804e-01 3.42517287e-01 -7.23126590e-01
-1.59642756e-01 -1.37512362e+00 -3.57362121e-01 2.76450276e-01
2.08488345e-01 -5.23930132e-01 -2.01480001e-01 -5.34836173e-01] | [11.48652458190918, 7.579827308654785] |
e1ea5465-f3d7-4204-8e6f-a4c014d56bda | towards-an-efficient-iris-recognition-system | 2210.13101 | null | https://arxiv.org/abs/2210.13101v1 | https://arxiv.org/pdf/2210.13101v1.pdf | Towards an efficient Iris Recognition System on Embedded Devices | Iris Recognition (IR) is one of the market's most reliable and accurate biometric systems. Today, it is challenging to build NIR-capturing devices under the premise of hardware price reduction. Commercial NIR sensors are protected from modification. The process of building a new device is not trivial because it is required to start from scratch with the process of capturing images with quality, calibrating operational distances, and building lightweight software such as eyes/iris detectors and segmentation sub-systems. In light of such challenges, this work aims to develop and implement iris recognition software in an embedding system and calibrate NIR in a contactless binocular setup. We evaluate and contrast speed versus performance obtained with two embedded computers and infrared cameras. Further, a lightweight segmenter sub-system called "Unet_xxs" is proposed, which can be used for iris semantic segmentation under restricted memory resources. | ['Christoph Busch', 'Enrique Lopez Droguett', 'Leonardo Causa', 'Mauricio Vasquez', 'Juan E. Tapia', 'Daniel P. Benalcazar'] | 2022-10-24 | null | null | null | null | ['iris-recognition'] | ['computer-vision'] | [ 4.68681306e-01 -1.71218380e-01 9.29280967e-02 -2.24794999e-01
-3.64366435e-02 -5.34296632e-01 -1.00164652e-01 -3.93567950e-01
-4.50603038e-01 1.43890545e-01 -2.86007613e-01 -7.35535383e-01
1.03219964e-01 -5.42304337e-01 -3.95276427e-01 -5.26003242e-01
7.34015882e-01 -9.53843147e-02 -2.77552277e-01 2.97502559e-02
5.38434923e-01 6.29301429e-01 -1.85338819e+00 -2.88550407e-01
1.10400999e+00 8.29860330e-01 -2.39475399e-01 7.51658201e-01
1.53163016e-01 1.94841206e-01 -5.51282167e-01 -6.18597806e-01
7.40363836e-01 -6.30252898e-01 -6.50997043e-01 9.47321113e-03
8.06496561e-01 -7.98749030e-01 -4.37560640e-02 9.99838054e-01
8.97465467e-01 -2.87974477e-01 -1.27180189e-01 -4.49102283e-01
-6.13890886e-01 2.04458088e-01 -8.91086221e-01 -7.06925914e-02
5.70071399e-01 5.49165487e-01 1.49504840e-01 -2.95795083e-01
1.63080439e-01 6.70279741e-01 6.17695749e-01 8.23432386e-01
-1.02610064e+00 -5.43349802e-01 -6.17666662e-01 -9.11240578e-02
-1.45534623e+00 -7.73486555e-01 5.08217573e-01 -3.82142067e-01
8.99636209e-01 5.58637381e-01 7.12437749e-01 6.29506350e-01
1.52137235e-01 1.26623437e-01 1.54148412e+00 -8.50261152e-01
-2.46076100e-02 5.38946569e-01 1.50827765e-01 5.64060390e-01
5.90347111e-01 3.54557782e-01 -2.23293453e-01 7.86508713e-03
7.48472571e-01 2.36242935e-02 -2.23575458e-01 1.32516742e-01
-7.53072917e-01 1.77771509e-01 1.65472642e-01 2.52907008e-01
-8.18951800e-02 -2.84863651e-01 -6.21887855e-03 1.53421164e-01
2.88163930e-01 4.81471151e-01 9.54306964e-03 -5.07896245e-01
-7.73731649e-01 -6.76696956e-01 6.00136578e-01 9.56512392e-01
2.90884227e-01 -3.01581502e-01 2.12988287e-01 7.13287115e-01
3.99175435e-01 7.86648512e-01 3.13639611e-01 -4.57764059e-01
-5.16276918e-02 9.44046795e-01 1.08916856e-01 -3.90535861e-01
-4.67290103e-01 -2.48613492e-01 -4.67612714e-01 3.61446142e-01
4.35334653e-01 -2.18783125e-01 -9.49451685e-01 8.02372992e-01
4.37909633e-01 2.72911280e-01 -1.58013329e-01 1.20005953e+00
7.64790535e-01 1.65631119e-02 -2.48918191e-01 -1.55735835e-01
1.51099849e+00 -6.02595568e-01 -5.02093911e-01 3.04080456e-01
5.04948795e-01 -1.35664189e+00 1.10295188e+00 2.88263679e-01
-1.11225617e+00 -7.29119897e-01 -1.08852816e+00 -4.51141000e-01
-3.26309770e-01 5.20701170e-01 5.28952420e-01 1.63546169e+00
-1.14609289e+00 3.11739951e-01 -6.88380659e-01 -5.45899749e-01
1.62899345e-01 9.44177687e-01 -1.16365202e-01 1.62947938e-01
-4.19575006e-01 9.27833021e-01 -8.75385106e-02 3.61398488e-01
1.58934236e-01 -4.88913924e-01 -6.03839040e-01 -3.49556744e-01
-1.49394274e-01 -8.66133273e-01 8.85987222e-01 -1.00429928e+00
-2.35625553e+00 1.37313735e+00 -2.44944230e-01 -1.38784423e-01
2.68775553e-01 -2.25992247e-01 -5.22516668e-01 1.19262390e-01
-6.56885564e-01 1.55649751e-01 9.61326063e-01 -4.92948622e-01
-4.38913465e-01 -6.94116831e-01 2.09263504e-01 1.08525448e-01
-3.43768924e-01 5.98159254e-01 -3.27286690e-01 1.04732923e-01
-1.09087013e-01 -1.03981006e+00 1.86160743e-01 -1.38209730e-01
-5.42476475e-01 1.69395521e-01 6.82790577e-01 -6.24826372e-01
1.03145564e+00 -2.35224962e+00 -6.47371054e-01 3.86885822e-01
2.14790776e-01 9.08993721e-01 1.15111820e-01 -2.63412297e-01
-1.88286215e-01 -1.59200326e-01 2.77715504e-01 -1.23327427e-01
-2.27302253e-01 -2.77600110e-01 9.47551504e-02 8.35338652e-01
-2.50082493e-01 7.50703096e-01 -3.27523112e-01 -3.51501197e-01
7.80174017e-01 7.06052721e-01 -2.05372155e-01 8.53156522e-02
5.31956136e-01 6.31138206e-01 -1.02680735e-01 1.19663334e+00
9.11902249e-01 -1.00680120e-01 -1.27316281e-01 -2.97195822e-01
-6.12281084e-01 7.63890892e-02 -1.23198736e+00 1.43221843e+00
-1.86271623e-01 4.58250165e-01 -3.01198419e-02 -3.92849624e-01
9.16825533e-01 2.60562688e-01 4.21555609e-01 -8.02671075e-01
6.49521828e-01 3.64708811e-01 1.13673776e-03 -7.65846610e-01
5.30974805e-01 1.41787454e-01 6.44627035e-01 5.39076328e-01
-5.21131754e-01 1.67364284e-01 -1.91308543e-01 -5.48377812e-01
6.92003489e-01 3.41120660e-01 1.35385811e-01 -6.44657016e-02
7.71617949e-01 -1.14400618e-01 4.70932759e-02 3.56959403e-01
-3.03191453e-01 4.82187748e-01 -6.51131943e-02 -6.78950846e-01
-9.57068264e-01 -6.50144577e-01 -7.46892571e-01 4.88988101e-01
4.57494766e-01 -1.26457876e-02 -1.06745136e+00 -5.94803244e-02
-1.79551616e-01 8.42674199e-05 -1.24015898e-01 1.20001152e-01
-3.47042382e-01 -5.89482188e-01 5.60438156e-01 6.66865110e-02
6.33669019e-01 -3.37997764e-01 -9.89740729e-01 -3.81724127e-02
2.43243396e-01 -9.50615644e-01 -5.40286481e-01 -6.71092033e-01
-7.29772627e-01 -1.25752628e+00 -6.14624798e-01 -5.90156972e-01
1.06519079e+00 5.07024050e-01 5.97523689e-01 2.77590185e-01
-9.25296962e-01 3.82391751e-01 -8.81631896e-02 -7.18949676e-01
1.56674877e-01 6.49980307e-02 3.15357208e-01 2.38431752e-01
1.08348155e+00 -3.50756384e-02 -1.36020577e+00 4.09778327e-01
-5.52114606e-01 8.09698775e-02 4.91800994e-01 5.26321650e-01
4.98626500e-01 -3.19030374e-01 -2.36808643e-01 -6.67791128e-01
4.12048846e-01 3.18128049e-01 -1.25202072e+00 4.14106011e-01
-9.02159095e-01 -3.79733443e-01 4.75204378e-01 -2.10859731e-01
-8.85081887e-01 2.07251012e-01 4.37390208e-02 -9.78645012e-02
-2.69836187e-01 -4.56727594e-02 1.61681175e-01 -1.00540709e+00
8.80798161e-01 -3.17278318e-02 5.53213716e-01 -5.07008374e-01
7.15974048e-02 1.55673409e+00 5.84420621e-01 -3.13837349e-01
6.32233381e-01 5.24363875e-01 9.33126137e-02 -8.98583412e-01
-3.06619525e-01 -6.81251526e-01 -4.62784022e-01 -2.72641003e-01
5.28955877e-01 -8.34050953e-01 -1.44358599e+00 8.30713809e-01
-8.02580416e-01 1.26132518e-01 -2.22577393e-01 8.12779486e-01
-3.60537507e-02 4.80264395e-01 -5.57172656e-01 -6.58647120e-01
-7.96374142e-01 -1.36475575e+00 9.76426065e-01 1.05100811e+00
7.65824094e-02 -5.22228062e-01 2.00047061e-01 9.16477025e-01
7.59604216e-01 6.80991486e-02 5.74821979e-02 2.24573225e-01
-7.48895109e-01 -4.61742908e-01 -4.38709706e-01 3.62891793e-01
3.27374190e-01 4.95943099e-01 -1.35631990e+00 -5.42304158e-01
1.89787611e-01 1.34238049e-01 1.18355580e-01 6.25788927e-01
1.06616712e+00 -9.63703841e-02 -3.07161242e-01 1.24422979e+00
1.67934847e+00 3.02636623e-01 1.12352312e+00 3.82227123e-01
4.11537617e-01 5.46966672e-01 3.17137599e-01 1.74867615e-01
1.75227448e-01 6.13003552e-01 -4.39387420e-03 -6.37101769e-01
4.30567712e-02 2.03493208e-01 1.37538686e-01 5.08380055e-01
-6.93444192e-01 3.38536143e-01 -8.64810228e-01 9.36000794e-02
-1.09768975e+00 -7.56775439e-01 -9.54458714e-02 2.49704885e+00
8.43617499e-01 -4.76982951e-01 2.46148765e-01 -1.44966930e-01
6.73648477e-01 -6.68239355e-01 -6.36333823e-01 -7.10158885e-01
5.55275306e-02 8.04557800e-01 8.24025750e-01 3.46461564e-01
-9.05533791e-01 6.97062731e-01 6.45451927e+00 7.94145390e-02
-1.76334310e+00 -3.37302946e-02 6.48633420e-01 -3.74964148e-01
1.74887449e-01 4.54736389e-02 -9.59044874e-01 6.11107826e-01
1.17358565e+00 3.51273060e-01 5.85900724e-01 6.46144450e-01
2.97765136e-01 -4.34572875e-01 -7.83804893e-01 1.39596403e+00
1.39711291e-01 -1.01504183e+00 -5.24184942e-01 3.56274307e-01
4.75946158e-01 5.04930736e-03 3.07301402e-01 -5.44790447e-01
-3.90240014e-01 -1.19511700e+00 -2.44501755e-02 7.73240864e-01
1.39720821e+00 -5.55349171e-01 1.02572644e+00 -2.08080947e-01
-8.02841783e-01 2.43638843e-01 -5.62689483e-01 -9.07283872e-02
-2.86514074e-01 2.47648999e-01 -6.29198015e-01 3.23453158e-01
6.26466572e-01 3.45031112e-01 -6.81300223e-01 1.24335396e+00
1.47377908e-01 3.76651287e-01 -4.21945304e-01 1.01464294e-01
-4.47572708e-01 -8.14555585e-01 1.18223883e-01 7.10454464e-01
4.57075387e-01 4.28778648e-01 -5.10795593e-01 7.48856604e-01
3.94003838e-02 1.99444816e-01 -4.33832616e-01 1.00762017e-01
1.48839235e-01 1.31988478e+00 -3.71170431e-01 -2.10431889e-02
-7.83798456e-01 7.73496091e-01 -4.13852006e-01 1.92213878e-01
-6.66573584e-01 -8.32097471e-01 7.74451673e-01 3.43123734e-01
-4.06719029e-01 -1.29508078e-01 -6.79119825e-01 -1.21942675e+00
2.16725394e-01 -9.72364664e-01 -5.16997129e-02 -7.10867822e-01
-6.61933661e-01 4.69744653e-01 -7.20806897e-01 -1.44721973e+00
2.64287680e-01 -7.71699965e-01 -2.88572520e-01 1.42861092e+00
-1.75477982e+00 -1.25362706e+00 -6.97017014e-01 9.08418238e-01
-1.10890388e-01 -1.81293488e-01 9.41860020e-01 3.21342647e-01
-1.04304409e+00 1.16599059e+00 1.21162064e-01 1.37694299e-01
7.64107645e-01 -8.52055550e-01 1.04994662e-01 1.10636604e+00
-2.70911008e-01 1.20566547e+00 2.60976911e-01 -2.79564381e-01
-1.97134280e+00 -3.02871734e-01 7.22144961e-01 -5.00035107e-01
1.58470050e-01 -4.87553626e-02 -4.42311108e-01 2.23884255e-01
3.42070103e-01 4.91046626e-03 1.13854349e+00 4.49852496e-02
-1.70027822e-01 -5.34697950e-01 -1.41613162e+00 6.22234046e-01
7.67270088e-01 -9.35945928e-01 -1.85877711e-01 3.83765787e-01
1.33852214e-01 -8.66726339e-01 -9.78009820e-01 1.31179899e-01
9.76739168e-01 -8.70189309e-01 8.40074301e-01 1.28958344e-01
-3.21386270e-02 -6.31632864e-01 5.14999926e-01 -5.33054590e-01
1.10741168e-01 -1.38056993e+00 2.98158377e-01 1.02219868e+00
1.71861649e-01 -1.29716504e+00 8.91472995e-01 1.29586673e+00
1.72439665e-01 -3.77363026e-01 -7.79705465e-01 -5.49376488e-01
-6.09419882e-01 1.69138294e-02 9.43998873e-01 6.43256247e-01
4.53215599e-01 -1.15972407e-01 -2.08161697e-01 3.85157198e-01
7.75402784e-01 2.65039831e-01 1.05945003e+00 -9.66216028e-01
-5.63338436e-02 -2.09934816e-01 -8.18371773e-01 -8.19969237e-01
-5.23395777e-01 -3.43172729e-01 -4.09852028e-01 -9.34419096e-01
6.81230724e-02 -3.42014045e-01 -1.91191331e-01 4.72971886e-01
5.88982590e-02 5.37837744e-01 -1.60347745e-01 4.27119344e-01
4.51505464e-03 -3.54374021e-01 1.12108481e+00 -7.48438612e-02
-5.43079853e-01 1.27316177e-01 -7.19590724e-01 2.92556494e-01
8.23660076e-01 4.40564603e-02 -2.97736257e-01 -4.29970831e-01
1.84564590e-01 -1.85557455e-01 3.15952748e-01 -1.07200420e+00
6.90794528e-01 7.42902106e-04 3.53744268e-01 -2.58731425e-01
3.40402834e-02 -1.00395846e+00 5.47550440e-01 6.14596725e-01
2.65827507e-01 -2.20218956e-01 1.84199661e-01 -2.55073667e-01
2.06223712e-03 1.03983830e-03 1.02043200e+00 2.45751709e-01
-3.53428364e-01 1.94812775e-01 1.78356364e-01 -5.45607507e-01
1.21036232e+00 -1.15167260e+00 -1.01318741e+00 3.97479743e-01
-3.01626444e-01 -2.69350171e-01 1.06904554e+00 1.61415488e-01
7.19904721e-01 -5.46757102e-01 -3.22480321e-01 9.46388841e-01
2.88851708e-01 -2.81357646e-01 3.34363282e-01 1.18528080e+00
-1.33915436e+00 6.81728482e-01 -3.37157339e-01 -6.73877299e-01
-1.75606775e+00 3.69891763e-01 7.19413400e-01 5.47018886e-01
-6.42511845e-01 7.30534136e-01 -5.18127859e-01 1.40810078e-02
1.42790183e-01 -1.93137065e-01 -1.51413560e-01 -5.04022539e-01
9.32735264e-01 4.08142775e-01 2.79810905e-01 -4.05868620e-01
-1.27004087e-01 1.09595251e+00 -1.70303240e-01 3.42673212e-01
8.83577943e-01 -4.09257799e-01 -6.01249754e-01 -3.60635191e-01
6.59646869e-01 1.37185425e-01 -7.52125680e-01 -1.05568759e-01
-4.50730771e-01 -9.99580920e-01 7.33667165e-02 -7.75064111e-01
-9.06135917e-01 6.39636636e-01 1.35571039e+00 -1.16385959e-01
1.48127353e+00 -5.02261698e-01 8.90380263e-01 -7.75410235e-02
5.38401961e-01 -1.31426537e+00 -7.14351594e-01 -5.96194491e-02
2.90694013e-02 -1.40103936e+00 1.09012321e-01 -3.21915895e-01
-9.07639787e-02 1.27883029e+00 5.08776009e-01 3.08015049e-01
5.77520609e-01 2.40971103e-01 5.08802772e-01 -3.38142425e-01
1.21488936e-01 -3.16529363e-01 5.49570680e-01 6.35805666e-01
6.14599884e-01 4.09545034e-01 -5.79930305e-01 -1.06510684e-01
-3.41560036e-01 4.87906098e-01 6.47749901e-01 8.29505384e-01
-7.82402977e-02 -1.26788342e+00 -5.40812492e-01 5.81126153e-01
-6.67412043e-01 -2.47780848e-02 -2.36359119e-01 2.65593022e-01
3.45871806e-01 1.21387458e+00 -1.02967568e-01 -5.42256117e-01
2.27346763e-01 -2.52609134e-01 4.12589908e-01 -3.52749914e-01
-8.88390779e-01 1.56265065e-01 -3.04360598e-01 -6.84547186e-01
-6.63592577e-01 -4.01449561e-01 -4.80563164e-01 -1.07641459e+00
-5.83197832e-01 -1.85559943e-01 1.16263092e+00 6.99844599e-01
6.55141771e-01 -4.52888831e-02 7.20061898e-01 -9.62143242e-02
-3.59958202e-01 -6.11150146e-01 -7.36120820e-01 -1.76891107e-02
3.13409716e-01 -4.81620096e-02 -1.31426722e-01 1.26474664e-01] | [3.7464237213134766, -3.6280136108398438] |
b4b3588f-a287-45fc-b288-37d3398be2c7 | real-time-multiple-people-tracking-with | 1809.04427 | null | http://arxiv.org/abs/1809.04427v1 | http://arxiv.org/pdf/1809.04427v1.pdf | Real-time Multiple People Tracking with Deeply Learned Candidate Selection and Person Re-Identification | Online multi-object tracking is a fundamental problem in time-critical video
analysis applications. A major challenge in the popular tracking-by-detection
framework is how to associate unreliable detection results with existing
tracks. In this paper, we propose to handle unreliable detection by collecting
candidates from outputs of both detection and tracking. The intuition behind
generating redundant candidates is that detection and tracks can complement
each other in different scenarios. Detection results of high confidence prevent
tracking drifts in the long term, and predictions of tracks can handle noisy
detection caused by occlusion. In order to apply optimal selection from a
considerable amount of candidates in real-time, we present a novel scoring
function based on a fully convolutional neural network, that shares most
computations on the entire image. Moreover, we adopt a deeply learned
appearance representation, which is trained on large-scale person
re-identification datasets, to improve the identification ability of our
tracker. Extensive experiments show that our tracker achieves real-time and
state-of-the-art performance on a widely used people tracking benchmark. | ['Haizhou Ai', 'Chong Shang', 'Zijie Zhuang', 'Long Chen'] | 2018-09-12 | null | null | null | null | ['multiple-people-tracking', 'large-scale-person-re-identification', 'online-multi-object-tracking'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-3.08062524e-01 -6.96472824e-01 -9.76175666e-02 -3.07553858e-01
-7.02021599e-01 -6.19329870e-01 1.62329301e-01 -1.58756793e-01
-6.19947731e-01 6.56242907e-01 -9.37020108e-02 2.19716460e-01
2.87398219e-01 -4.16583598e-01 -8.73214483e-01 -4.26150620e-01
-1.24883093e-01 3.72394174e-01 6.18705094e-01 1.37877613e-01
-2.96380639e-01 4.05251324e-01 -1.65922773e+00 9.45845321e-02
6.70341015e-01 1.11151183e+00 -3.13013382e-02 6.72716498e-01
3.39295626e-01 5.31612754e-01 -8.35632741e-01 -6.67077124e-01
4.13322479e-01 -3.68595906e-02 -2.23204363e-02 1.32748917e-01
9.88643944e-01 -6.34472907e-01 -4.92639601e-01 1.30547094e+00
5.65916896e-01 -4.95016761e-03 2.16565445e-01 -1.45464706e+00
-6.71470881e-01 3.88214260e-01 -6.90051675e-01 4.52722162e-01
4.50525701e-01 3.34362835e-01 6.43210053e-01 -8.53507578e-01
2.72091329e-01 1.37581098e+00 1.07941186e+00 7.09401369e-01
-1.07182395e+00 -1.08663678e+00 6.40764236e-01 1.07691623e-01
-1.57417953e+00 -4.71672952e-01 3.14699024e-01 -5.11434495e-01
2.68721581e-01 2.15594456e-01 7.06235111e-01 1.03437674e+00
6.73682487e-04 1.04865265e+00 5.89726985e-01 -7.47166947e-02
-1.81631610e-01 1.41459137e-01 2.69448459e-01 6.55548513e-01
8.55617166e-01 4.36053157e-01 -5.65483212e-01 -3.48262399e-01
6.53490245e-01 3.26399952e-01 -2.01285720e-01 -3.82239968e-01
-1.14321697e+00 5.24704397e-01 5.02258182e-01 -9.61805955e-02
-1.75048128e-01 4.16722387e-01 3.56403232e-01 2.95502752e-01
2.95566052e-01 -1.87791571e-01 -2.90475219e-01 1.11593984e-01
-9.88029301e-01 4.31008309e-01 2.96302319e-01 1.11948442e+00
3.67905647e-01 3.86167206e-02 -7.94824719e-01 4.36824560e-01
5.07439375e-01 9.75741029e-01 2.46156871e-01 -5.22322834e-01
3.58595848e-01 5.00961602e-01 7.01901913e-01 -7.57403374e-01
-3.75493228e-01 -7.52167463e-01 -6.16533160e-01 3.35372150e-01
6.85399175e-01 -3.15718412e-01 -7.97449470e-01 1.97410870e+00
5.90575814e-01 5.99228442e-01 -4.75401491e-01 1.19285929e+00
6.15772247e-01 1.57723427e-01 1.25328824e-01 -2.19375938e-01
1.51183629e+00 -1.07130361e+00 -7.15928316e-01 -2.71538228e-01
2.80896842e-01 -6.58589184e-01 2.20746830e-01 2.58204043e-01
-8.50091100e-01 -1.08367527e+00 -9.81022358e-01 2.63923913e-01
-1.19778603e-01 6.55029714e-01 3.94206882e-01 7.81393588e-01
-1.03989112e+00 4.20563251e-01 -7.91896641e-01 -3.17313969e-01
3.94376934e-01 5.10332465e-01 -1.76123425e-01 1.62166264e-02
-9.70297277e-01 7.10389197e-01 2.19083577e-01 3.32763731e-01
-8.63057911e-01 -5.03605962e-01 -6.67147577e-01 -1.13648087e-01
4.80125546e-01 -7.03017056e-01 1.35618663e+00 -8.32657814e-01
-9.89689231e-01 7.28481710e-01 -3.61823112e-01 -5.97970605e-01
9.69483614e-01 -7.75983632e-01 -8.42814863e-01 -2.45294988e-01
3.89618814e-01 4.42954659e-01 1.15551805e+00 -1.06461799e+00
-1.08096111e+00 -4.57943618e-01 -2.53423721e-01 -1.35007560e-01
-3.00408840e-01 3.77351373e-01 -1.02245259e+00 -7.70507395e-01
-1.04305729e-01 -1.17983425e+00 -1.61849529e-01 5.70128500e-01
-2.61406183e-01 -3.50633264e-01 6.75452054e-01 -6.50864005e-01
1.07458115e+00 -1.95184863e+00 -9.33338404e-02 9.36093554e-02
3.76272708e-01 4.25273687e-01 -5.16527705e-02 -2.30005607e-01
2.31940493e-01 -4.64703262e-01 5.52130938e-01 -7.38813937e-01
9.66771841e-02 -2.51892924e-01 -2.98089355e-01 8.56553257e-01
8.88267010e-02 9.57597435e-01 -9.93112981e-01 -5.09855270e-01
2.04932764e-01 5.27452171e-01 -2.14922860e-01 2.51079708e-01
-7.14734197e-02 5.08089125e-01 -4.75417793e-01 9.03891981e-01
8.89152348e-01 -4.34735388e-01 -3.81227746e-03 -2.09010616e-01
-3.05296570e-01 -2.46030241e-01 -1.51936877e+00 1.56013525e+00
2.08761439e-01 5.69664657e-01 -1.74333546e-02 -4.14811403e-01
6.12717569e-01 2.20258072e-01 5.77183366e-01 -4.71449047e-01
1.95803270e-01 1.59786306e-02 -7.55565539e-02 3.45989317e-02
7.32201397e-01 6.22481346e-01 -3.37216929e-02 3.65854204e-01
-7.62801319e-02 1.06948662e+00 2.76174575e-01 2.15107381e-01
1.03621864e+00 2.66239733e-01 -2.38273405e-02 2.00082868e-01
5.61008811e-01 -2.06796363e-01 1.01717019e+00 1.19314337e+00
-7.77527153e-01 4.62180316e-01 -2.12712124e-01 -7.77608752e-01
-8.89959931e-01 -1.11873305e+00 2.12478433e-02 1.29001582e+00
5.45347452e-01 -2.55888015e-01 -4.84096020e-01 -7.08868504e-01
4.31825638e-01 -1.11052372e-01 -6.04779243e-01 -1.19986162e-01
-7.01233208e-01 -7.22698569e-01 5.30436337e-01 7.32698321e-01
2.59106040e-01 -7.56270409e-01 -6.26332521e-01 3.89180660e-01
-1.36245295e-01 -1.29133773e+00 -9.24351990e-01 -4.39759523e-01
-4.66653287e-01 -1.23487437e+00 -9.57636416e-01 -5.19102871e-01
6.76750839e-01 8.05791616e-01 9.98242497e-01 5.27927458e-01
-3.50667179e-01 3.67502332e-01 -1.95294380e-01 -4.53204095e-01
-8.74264464e-02 -2.38953739e-01 6.97140396e-01 2.29788810e-01
5.47068715e-01 1.57068059e-01 -6.60599589e-01 6.36059403e-01
-2.06790298e-01 -3.83432359e-01 5.18577874e-01 6.60173893e-01
5.03719211e-01 -5.84131964e-02 4.71240044e-01 -3.15290034e-01
1.93957373e-01 -1.37857065e-01 -1.14603674e+00 5.35338044e-01
-3.86900663e-01 -9.41488594e-02 2.43207216e-01 -8.78294706e-01
-7.55533993e-01 3.71513218e-01 3.21580544e-02 -8.36915672e-01
1.34988829e-01 -2.74999201e-01 -7.89913386e-02 -4.24474478e-01
4.51569706e-01 1.46432385e-01 -1.10119060e-01 -5.23428977e-01
3.23105305e-01 3.75301093e-01 8.43979299e-01 -4.50926065e-01
1.17386556e+00 7.41124749e-01 -2.72096783e-01 -2.80779153e-01
-1.09692776e+00 -8.68036807e-01 -6.13184929e-01 -6.38698220e-01
5.23265064e-01 -1.50953031e+00 -1.16142130e+00 5.36817312e-01
-1.24105275e+00 2.09461510e-01 5.81536517e-02 5.61991572e-01
9.08108521e-03 4.75188911e-01 -4.51779604e-01 -1.23244131e+00
-4.77072895e-01 -1.06088185e+00 1.29425800e+00 6.61434591e-01
2.12157726e-01 -4.29274261e-01 8.94613937e-02 5.60281575e-02
2.61140525e-01 5.81793375e-02 -6.09861791e-01 -3.93356740e-01
-1.05195010e+00 -6.01963699e-01 -4.05512333e-01 -4.26188074e-02
-2.50906944e-02 -1.08567245e-01 -9.88005996e-01 -9.21432555e-01
-5.01547039e-01 -1.49483711e-01 1.18238163e+00 5.35757482e-01
9.46407020e-01 7.03383889e-03 -9.39338923e-01 5.84351182e-01
1.23038030e+00 -2.11001948e-01 2.52304763e-01 3.74995440e-01
8.48183155e-01 5.51852845e-02 8.37216854e-01 5.68556726e-01
5.40513813e-01 1.23051429e+00 2.34267741e-01 9.93004218e-02
-3.06728303e-01 -2.17828810e-01 6.40604973e-01 1.73141479e-01
-4.50727157e-02 -1.07582450e-01 -3.39583695e-01 4.82708007e-01
-2.34978986e+00 -1.21529090e+00 -2.07506537e-01 2.64780998e+00
4.94913340e-01 3.94729912e-01 7.07102895e-01 -3.39253366e-01
1.19861841e+00 -2.84668952e-01 -5.96607029e-01 7.49061942e-01
-1.14118159e-01 -3.65278244e-01 8.58768463e-01 2.67253190e-01
-1.62469244e+00 8.67588818e-01 6.18930197e+00 5.56683302e-01
-7.93577194e-01 2.39553928e-01 6.19546920e-02 -3.42536509e-01
5.74012518e-01 -2.61088461e-01 -1.62956810e+00 9.85230207e-01
7.18175232e-01 -5.96892871e-02 3.64817157e-02 9.39620137e-01
5.52017577e-02 1.37030661e-01 -1.17575490e+00 1.26602709e+00
5.13424128e-02 -1.16722596e+00 -3.83880496e-01 -6.16310723e-02
5.51088393e-01 8.58692974e-02 1.91526040e-01 4.95892107e-01
5.24872541e-01 -5.88705599e-01 1.04000115e+00 7.23494112e-01
5.55154681e-01 -5.41318238e-01 6.65417612e-01 1.62756652e-01
-1.95060337e+00 -2.29073197e-01 -6.28494263e-01 1.39952809e-01
4.26894158e-01 5.49974799e-01 -3.16516459e-01 6.14876449e-01
9.91236448e-01 7.41235733e-01 -8.30312252e-01 1.82461560e+00
-8.03266242e-02 2.20261723e-01 -3.99107128e-01 1.29666761e-01
-3.20799679e-01 2.70457208e-01 6.33004904e-01 1.13838601e+00
3.11843932e-01 -3.14459234e-01 7.95716584e-01 7.49261081e-01
-4.71783914e-02 -4.42565024e-01 -3.03347588e-01 4.58876371e-01
7.70894825e-01 1.32001865e+00 -4.44258034e-01 -4.91378337e-01
-6.17827713e-01 8.62640142e-01 5.16550779e-01 8.56555104e-02
-1.35545838e+00 1.28178984e-01 8.75024557e-01 -1.15604877e-01
6.32392704e-01 -5.85342310e-02 3.29408765e-01 -1.45122373e+00
3.71899515e-01 -6.06404603e-01 6.59009099e-01 -3.61829132e-01
-1.55600774e+00 5.74384153e-01 -3.67046565e-01 -1.68933249e+00
-9.86047555e-03 -5.53309679e-01 -5.20664096e-01 7.23304272e-01
-1.57655752e+00 -1.50739062e+00 -5.95679700e-01 5.04209936e-01
4.05139357e-01 -2.10651562e-01 2.91948527e-01 8.38794529e-01
-8.50442588e-01 1.10316908e+00 4.24761921e-02 5.56065738e-01
1.08652520e+00 -9.43142056e-01 7.34560192e-01 1.27973354e+00
1.98734909e-01 5.49825907e-01 6.91525936e-01 -1.02370608e+00
-1.46876860e+00 -1.51091647e+00 5.12274086e-01 -8.78673196e-01
4.69880551e-01 -3.23276013e-01 -8.73596072e-01 8.16591263e-01
-3.33868504e-01 5.70225656e-01 4.12091941e-01 2.04582438e-01
-3.95799875e-01 -1.53930590e-01 -8.48664641e-01 3.76585573e-01
1.26698315e+00 -2.31545195e-01 -3.97110969e-01 4.00739372e-01
5.24050713e-01 -6.47313654e-01 -5.37019491e-01 2.40755886e-01
1.01293242e+00 -7.02757776e-01 1.37515712e+00 -5.30159235e-01
-7.06496239e-01 -8.95054042e-01 1.86568618e-01 -8.65283489e-01
-6.77534103e-01 -6.09881341e-01 -7.31926680e-01 1.08640981e+00
-3.26355100e-02 -3.63590002e-01 9.05567408e-01 5.81698954e-01
1.89577773e-01 -1.35440305e-01 -9.79238510e-01 -1.24807286e+00
-5.24309218e-01 -2.19044223e-01 6.01086736e-01 4.22795027e-01
-5.64607322e-01 -2.22278778e-02 -1.01828766e+00 8.03626955e-01
1.42304087e+00 2.18887851e-01 1.14057267e+00 -1.66011989e+00
-2.53767371e-01 -1.37948513e-01 -5.89609265e-01 -1.37601328e+00
8.31152424e-02 -3.83477390e-01 2.81981975e-01 -1.01601720e+00
4.84166652e-01 -4.36214954e-01 -6.15700006e-01 2.77536362e-01
-6.52506828e-01 4.33527172e-01 4.80605036e-01 5.12221396e-01
-1.41628718e+00 5.15974939e-01 8.02488446e-01 -3.28116149e-01
1.11205978e-02 3.84300798e-01 -5.40891528e-01 6.61930621e-01
2.83078134e-01 -6.76576912e-01 2.08732501e-01 -5.24282098e-01
-6.22637570e-02 -2.26748735e-01 8.00828159e-01 -1.45219600e+00
7.73563206e-01 1.87105671e-01 1.02063215e+00 -1.09131646e+00
5.65648854e-01 -8.14911902e-01 2.08812028e-01 6.04696751e-01
-1.23082865e-02 2.03379020e-01 2.19275206e-01 9.53060567e-01
1.49588034e-01 1.09526388e-01 7.49403536e-01 6.97595030e-02
-8.78814042e-01 7.55030632e-01 2.27360770e-01 -2.01483190e-01
1.07831359e+00 -1.56324252e-01 -4.09049302e-01 -2.49273807e-01
-5.56259274e-01 7.06921577e-01 5.50205231e-01 7.37941742e-01
4.69816506e-01 -1.69604301e+00 -8.54213536e-01 7.80173168e-02
2.28476703e-01 -4.02662218e-01 1.93809360e-01 8.09302628e-01
1.49523377e-01 2.92617053e-01 -1.60313636e-01 -9.25425529e-01
-1.55880105e+00 7.57266581e-01 3.60487133e-01 -1.63573548e-01
-7.50829995e-01 9.25166905e-01 1.50590017e-01 3.86465229e-02
6.06045663e-01 -1.13210097e-01 -1.28678769e-01 -2.41557896e-01
1.27752566e+00 2.34042808e-01 -2.68001795e-01 -8.65967214e-01
-7.03781068e-01 5.34772158e-01 -3.35200131e-01 1.72758311e-01
7.89157569e-01 -2.54803210e-01 3.28644902e-01 9.30302963e-02
5.55184484e-01 -2.80200224e-03 -1.61012101e+00 -5.14575779e-01
-3.56353000e-02 -7.97668874e-01 -1.70460433e-01 -4.78722900e-01
-1.17745185e+00 2.70316601e-01 1.13643026e+00 6.31541163e-02
6.49047136e-01 -1.10825807e-01 8.13009858e-01 4.61637259e-01
6.37641191e-01 -1.17034543e+00 3.31307948e-02 2.41192698e-01
4.76867229e-01 -1.73578978e+00 2.68141240e-01 -2.49601409e-01
-3.17149609e-01 8.62385690e-01 9.98687625e-01 1.34675121e-02
3.78743500e-01 1.41828269e-01 -1.67798076e-03 8.95782094e-03
-5.09004653e-01 -5.51678777e-01 4.78289664e-01 6.77994907e-01
9.91804153e-02 3.43126394e-02 1.71130762e-01 6.48770809e-01
3.02072376e-01 1.33951992e-01 1.03430934e-01 8.28530073e-01
-5.41667938e-01 -1.05153835e+00 -8.70018065e-01 3.03783417e-01
-4.83495593e-01 3.66062909e-01 -1.56975403e-01 5.08726001e-01
3.20018589e-01 1.04537940e+00 7.58687183e-02 -3.41201276e-01
3.16604912e-01 -3.67500454e-01 5.35209894e-01 -2.80630201e-01
-5.76756895e-01 1.45346835e-01 -1.36059016e-01 -6.45453215e-01
-5.69747686e-01 -8.61749232e-01 -8.30480576e-01 -3.86884570e-01
-7.18971610e-01 3.65012996e-02 3.42696786e-01 9.62374568e-01
4.33679581e-01 6.46262467e-01 4.40732092e-01 -9.34648514e-01
-8.09058607e-01 -9.24520731e-01 -4.45021838e-01 5.76078117e-01
6.75010145e-01 -1.25650227e+00 8.27331021e-02 -7.85994381e-02] | [6.448641777038574, -1.9335602521896362] |
b0d2aa79-0a04-4cd6-ad05-6852e13380bc | instant-image-denoising-plugin-for-imagej | 2006.13801 | null | http://arxiv.org/abs/2006.13801v1 | http://arxiv.org/pdf/2006.13801v1.pdf | Instant Image Denoising Plugin for ImageJ using Convolutional Neural Networks | We present a new convolutional neural network (CNN) based ImageJ plugin for
fluorescence microscopy image denoising with an average improvement of 7.5 dB
in peak signal-to-noise ratio (PSNR) and denoising instantly within 80 msec. | [] | 2020-06-23 | null | null | null | null | ['intensity-image-denoising'] | ['computer-vision'] | [ 1.96899742e-01 -4.76455152e-01 9.53196347e-01 -4.27749395e-01
-8.20968390e-01 -2.73293912e-01 -1.02267839e-01 3.56895745e-01
-1.25577521e+00 8.83885026e-01 -4.10376936e-01 -3.16132784e-01
4.28004980e-01 -2.81881899e-01 -6.91578627e-01 -1.07616329e+00
-1.35363221e-01 -6.59264386e-01 2.54339784e-01 1.07837655e-01
2.08449662e-01 1.02556324e+00 -8.70849609e-01 2.66789466e-01
2.55555391e-01 1.28405309e+00 3.40774022e-02 1.34874225e+00
8.80794749e-02 7.32870281e-01 -1.03701198e+00 -4.00269777e-01
1.24038428e-01 7.97183439e-02 -7.91813195e-01 -3.27033043e-01
1.76409781e-01 -7.36357927e-01 -6.27210259e-01 1.16035295e+00
1.17577672e+00 -2.04180524e-01 1.34046063e-01 -8.32199275e-01
-5.62261224e-01 5.43145649e-02 -6.11926258e-01 1.17635357e+00
8.52657761e-03 8.24118018e-01 -6.85836151e-02 -5.05036116e-01
1.12490201e+00 9.92339253e-01 9.73416150e-01 4.45075840e-01
-1.86198044e+00 -5.81164002e-01 -6.25137568e-01 2.17164472e-01
-1.13674581e+00 -3.95193696e-01 2.77371109e-01 1.01102389e-01
1.18403590e+00 -1.47615463e-01 6.12364829e-01 8.26643169e-01
1.00518191e+00 4.04795289e-01 1.33928192e+00 1.53294548e-01
9.77159888e-02 -7.72838652e-01 -1.38511479e-01 4.94222105e-01
-1.54512629e-01 -1.18712284e-01 -2.24507108e-01 2.51945615e-01
1.15239108e+00 -2.03411639e-01 -1.28556818e-01 6.15818799e-01
-1.24022448e+00 2.37862363e-01 4.85121012e-01 1.55308023e-01
-5.18736362e-01 7.77131557e-01 9.39537823e-01 7.49587536e-01
2.73455501e-01 1.79934099e-01 -4.78331923e-01 -3.33478093e-01
-8.69826078e-01 1.54729649e-01 6.40697181e-01 4.08017695e-01
3.48151743e-01 4.75506902e-01 -1.97690740e-01 3.40134233e-01
-9.91644934e-02 6.50928974e-01 2.24521711e-01 -1.90163088e+00
-3.85283828e-01 -8.33950043e-02 2.46218108e-02 -8.12128007e-01
-4.63796079e-01 -3.19640011e-01 -1.27869844e+00 7.56457508e-01
5.60134947e-01 -4.59557503e-01 -9.27714705e-01 1.48616135e+00
-3.57650071e-02 -5.14889657e-02 -1.50757328e-01 8.06747258e-01
9.62639451e-01 5.62335253e-01 1.36616021e-01 -4.01696563e-01
1.01046884e+00 -4.63274419e-01 -1.12569118e+00 2.58851618e-01
1.58081099e-01 -9.28999126e-01 8.97553205e-01 3.47460330e-01
-1.43582296e+00 -3.47634882e-01 -1.18359709e+00 -3.79152268e-01
-3.32661390e-01 -2.78149158e-01 2.32885763e-01 3.11878830e-01
-1.58329439e+00 8.85868251e-01 -1.03374588e+00 -1.52479112e-01
9.12815690e-01 8.19421411e-01 -9.79189098e-01 2.16292605e-01
-5.96224189e-01 6.89454854e-01 -2.35829145e-01 1.60681531e-01
-1.11259687e+00 -1.18336320e+00 -1.46376222e-01 2.02908337e-01
-4.07330811e-01 -6.38375103e-01 1.28686619e+00 -6.15826428e-01
-1.65969753e+00 1.09571433e+00 1.46964481e-02 -9.74043727e-01
4.87331539e-01 1.35840505e-01 -2.36099929e-01 7.05539465e-01
-6.80707991e-02 8.51338625e-01 3.89245063e-01 -9.15115058e-01
-1.92934290e-01 -2.82822222e-01 -2.42082387e-01 -5.97062230e-01
3.26739401e-02 5.94215453e-01 -9.36070234e-02 -5.05753636e-01
1.09681031e-02 -2.90959239e-01 -4.34751272e-01 8.17701280e-01
-7.84997642e-02 5.03588140e-01 1.01027536e+00 -7.67197132e-01
4.67861205e-01 -1.92709613e+00 -5.12784183e-01 -2.80593187e-01
4.26669389e-01 5.29558778e-01 -6.32887244e-01 9.53113958e-02
-3.08172464e-01 1.08110867e-01 5.26880659e-02 -9.02348310e-02
-5.49955130e-01 1.05963118e-01 2.05535471e-01 9.78890479e-01
4.04462926e-02 1.08885872e+00 -7.74122059e-01 -1.48938134e-01
1.92723542e-01 1.23845816e+00 -1.33033395e-01 1.63352713e-01
5.81830442e-01 8.04661334e-01 1.49412289e-01 8.09337795e-01
1.10333014e+00 5.26858075e-03 -1.04116783e-01 -6.81548715e-01
-1.30423158e-01 -2.58836001e-01 -6.11956596e-01 1.54695499e+00
-1.50972068e-01 1.34950304e+00 1.04080820e+00 -9.40332055e-01
5.55443287e-01 3.36195409e-01 6.36532843e-01 -7.89273024e-01
5.73782444e-01 -6.22834191e-02 -1.11227967e-01 -4.25311893e-01
-3.39865834e-01 -7.57313073e-02 6.55122042e-01 9.39143151e-02
4.98118162e-01 -3.63972336e-02 2.52111465e-01 1.32547125e-01
1.64039063e+00 -5.14763176e-01 -1.31733760e-01 -4.54087406e-01
2.91329086e-01 -4.51431811e-01 4.73410755e-01 5.60992062e-01
-1.00742710e+00 6.87010825e-01 7.96631336e-01 -9.75336134e-01
-1.36960673e+00 -8.92598808e-01 -1.15075953e-01 5.73854387e-01
-3.08618069e-01 2.07458064e-01 -1.16845238e+00 -4.52166833e-02
-3.11871648e-01 -3.25926617e-02 -4.30079728e-01 1.76240176e-01
-3.85378480e-01 -6.49208963e-01 1.01430678e+00 5.24486721e-01
6.74017549e-01 -1.39359009e+00 -7.09184647e-01 4.89245296e-01
-4.64214534e-02 -1.33684933e+00 -2.49182358e-01 8.01747799e-01
-9.69843268e-01 -1.03022730e+00 -7.86649227e-01 -7.86403179e-01
7.34914303e-01 9.08909440e-02 1.08204758e+00 2.26286620e-01
-9.82404649e-01 1.07241392e-01 3.01826984e-01 -3.65398556e-01
-6.40134439e-02 -5.33742309e-01 -1.93366751e-01 -6.24619126e-01
2.82775491e-01 -1.10505462e+00 -1.11075795e+00 -1.50246277e-01
-1.26347983e+00 -6.47399127e-01 3.52538288e-01 1.09456992e+00
9.07499909e-01 1.45972937e-01 4.24059510e-01 -2.50891507e-01
8.45173895e-01 3.35340083e-01 -7.65008390e-01 -4.64794010e-01
-4.57544237e-01 -2.93469071e-01 8.90863776e-01 -3.64892572e-01
-6.87620342e-01 2.47114137e-01 -6.74572349e-01 -4.06226814e-01
-4.18896228e-01 3.21351320e-01 2.42142588e-01 -7.87485600e-01
6.10691309e-01 4.74420227e-02 5.61372697e-01 -8.88030231e-02
7.83615708e-02 2.63726890e-01 1.05819774e+00 -2.07186006e-02
2.30091617e-01 1.00378323e+00 1.76793054e-01 -6.38037205e-01
-4.12572592e-01 -2.80992895e-01 -3.64422798e-01 -2.68801272e-01
9.10265207e-01 -8.93836260e-01 -1.63684905e+00 9.21392918e-01
-1.17283130e+00 -4.60143656e-01 1.10363821e-02 1.71200216e-01
-4.26078349e-01 3.72179985e-01 -1.50586748e+00 -1.78203523e-01
-1.16256368e+00 -1.39369822e+00 7.16412663e-01 3.62620085e-01
1.26327053e-01 -7.50636280e-01 -3.15473586e-01 5.51654957e-02
8.72357011e-01 7.33729005e-01 5.40797472e-01 -5.53140156e-02
-2.24348873e-01 -5.03670514e-01 -7.52718329e-01 7.58000612e-01
-1.44983023e-01 2.30226502e-01 -9.65326846e-01 -4.55136657e-01
2.39131272e-01 -1.82724267e-01 9.16316748e-01 1.23668802e+00
1.52567935e+00 1.04538584e-02 9.96212363e-02 1.08509469e+00
1.78594172e+00 3.60550851e-01 1.12860918e+00 3.83591831e-01
9.30463374e-02 -5.39562246e-03 -2.10263669e-01 4.25558925e-01
-4.88817722e-01 2.45480925e-01 8.26625228e-01 -6.23669565e-01
-2.32160270e-01 6.41004562e-01 -6.97287172e-02 5.58837533e-01
6.65606111e-02 -2.95994192e-01 -5.53463221e-01 6.36259258e-01
-1.33801043e+00 -8.85386944e-01 -2.41166338e-01 1.63730407e+00
8.88501108e-01 7.74944127e-02 -5.63904271e-02 -2.22158320e-02
5.53032219e-01 -2.37642631e-01 -6.76531613e-01 -5.36708057e-01
-3.68646652e-01 6.18963599e-01 9.72241819e-01 3.66372108e-01
-9.60087180e-01 5.26511610e-01 7.78830814e+00 6.83062792e-01
-1.45678675e+00 -1.70712322e-01 1.06533968e+00 -2.83365399e-02
4.54423666e-01 -4.68264043e-01 -7.74238557e-02 5.62967002e-01
1.31504178e+00 -1.72297090e-01 6.00853503e-01 4.84831452e-01
5.50638974e-01 -1.44403502e-01 -4.58381653e-01 1.14258158e+00
-6.99826658e-01 -1.71405363e+00 -3.16142201e-01 -4.27108258e-02
3.35561872e-01 1.76408336e-01 2.83608794e-01 -3.75226289e-01
-3.49229015e-02 -1.10642612e+00 2.01918513e-01 6.98945701e-01
1.08304977e+00 -1.15248656e+00 1.43963897e+00 -2.15288728e-01
-6.16174877e-01 3.07862282e-01 -8.10571611e-01 2.28075251e-01
1.96169242e-01 1.01796722e+00 -3.75270188e-01 -6.01923317e-02
1.32231939e+00 2.57909000e-01 -2.11556524e-01 1.10384381e+00
1.79079518e-01 6.84173107e-01 -3.33313286e-01 3.66444290e-01
3.19621205e-01 2.93579251e-02 5.13590634e-01 1.62486684e+00
2.02528015e-01 2.38160864e-01 -4.40853000e-01 7.45483756e-01
-4.99876261e-01 -3.71623188e-01 -6.81179538e-02 -6.69626817e-02
3.24215919e-01 1.57155871e+00 -1.32092452e+00 -2.11140081e-01
-1.75402984e-01 1.27832150e+00 -2.64761358e-01 5.21320581e-01
-5.55249035e-01 -1.29490387e+00 8.96449029e-01 5.87304346e-02
4.44009095e-01 -1.16263084e-01 -4.25722957e-01 -4.96199846e-01
-1.55859262e-01 -6.32467449e-01 -3.66105109e-01 -9.72197413e-01
-1.31030309e+00 2.77386039e-01 -8.80720675e-01 -6.89638972e-01
5.32218277e-01 -9.65045750e-01 -6.47288084e-01 8.61879408e-01
-1.53400159e+00 -7.64547348e-01 -2.96673387e-01 5.56444764e-01
1.07751496e-01 1.95054442e-01 9.88886595e-01 4.55107987e-01
-4.58795786e-01 3.63583684e-01 4.73900408e-01 3.91160816e-01
7.51670063e-01 -1.57234192e+00 6.19375885e-01 7.87546515e-01
-9.09449518e-01 9.00306761e-01 8.93918097e-01 4.08319198e-02
-1.50113499e+00 -8.03650379e-01 7.56408751e-01 1.59266889e-01
7.85023510e-01 9.94052961e-02 -9.83439922e-01 3.54570746e-01
8.64177346e-01 6.68323576e-01 4.73465532e-01 -7.32041776e-01
-2.76622176e-01 -3.37627530e-01 -1.89734757e+00 7.35503674e-01
2.98081368e-01 -4.52168554e-01 4.03641522e-01 3.05261910e-01
5.93212187e-01 -5.43243766e-01 -1.22958517e+00 8.76109228e-02
6.82080388e-01 -1.08919835e+00 1.25475097e+00 -3.17313999e-01
2.54198253e-01 -4.97173846e-01 2.59432197e-01 -1.11592388e+00
-3.66601974e-01 -1.10036588e+00 2.02991873e-01 7.69315422e-01
2.34300736e-02 -3.85006994e-01 7.73400068e-01 2.38501593e-01
-2.93857068e-01 -7.30821431e-01 -1.37840390e+00 -6.47914827e-01
6.60256818e-02 -2.39143714e-01 -8.46876130e-02 3.26429605e-01
-1.46166384e-01 -8.48289430e-02 -5.33852391e-02 -1.97379932e-01
5.46070993e-01 -6.18066311e-01 2.98529774e-01 -7.15980232e-01
6.38525248e-01 -3.17767471e-01 -1.00250733e+00 -9.11612689e-01
-1.38709405e-02 -2.63267547e-01 1.94333345e-01 -1.49933565e+00
3.76422882e-01 1.05524886e+00 -7.32971013e-01 6.90993488e-01
1.10607989e-01 6.31426036e-01 3.79959345e-02 -5.10742366e-01
-7.08623111e-01 5.12031354e-02 1.14391053e+00 -3.35506588e-01
8.48456547e-02 -3.56139868e-01 -5.04319787e-01 4.54463601e-01
7.27742612e-01 -6.13632560e-01 1.97127610e-01 -6.64145589e-01
-1.78024508e-02 2.14051437e-02 5.33895850e-01 -1.29627717e+00
3.16443682e-01 3.78708303e-01 8.18957686e-01 -4.04868037e-01
1.08521402e-01 -7.70766914e-01 2.96401680e-01 7.49603987e-01
-4.62788463e-01 2.15410277e-01 7.30727077e-01 3.81952137e-01
-1.75035745e-01 1.90442368e-01 1.31276822e+00 -2.72010446e-01
-1.73337370e-01 2.65628666e-01 -1.22125566e+00 -2.92055368e-01
5.49069166e-01 -3.79278719e-01 -5.49646556e-01 -6.70468569e-01
-8.53305876e-01 -2.05419347e-01 4.27821606e-01 -5.66710114e-01
7.30143309e-01 -9.97884691e-01 -5.80872953e-01 6.36246875e-02
-3.57878089e-01 -6.16139233e-01 4.41898584e-01 1.37556636e+00
-1.26675594e+00 2.36074150e-01 -7.88548589e-01 -3.44429046e-01
-1.61443758e+00 4.81504649e-02 5.59314370e-01 -2.21051604e-01
-6.55335367e-01 1.34472656e+00 -4.78414953e-01 -1.65813453e-02
2.69575030e-01 -4.22222316e-01 1.59407362e-01 -5.11675656e-01
9.70875323e-01 5.74718714e-01 3.45961064e-01 -3.38261098e-01
-3.97221833e-01 3.17038208e-01 -1.52007997e-01 -5.15153036e-02
1.74785900e+00 -2.17100814e-01 -6.16926968e-01 -2.20385090e-01
1.69255650e+00 -5.97982585e-01 -1.68447316e+00 3.09366882e-01
-3.19359183e-01 -1.39075473e-01 6.99850678e-01 -1.08704698e+00
-1.56471109e+00 1.00550199e+00 1.36183882e+00 5.60517684e-02
1.52504790e+00 -7.15601802e-01 1.36879790e+00 6.23764277e-01
-1.75446495e-01 -1.09221303e+00 5.15588485e-02 4.88475412e-01
7.11705804e-01 -9.75686729e-01 8.49252641e-02 6.79819752e-03
-3.62465590e-01 1.64153874e+00 1.67470332e-02 -3.02077979e-01
6.99130118e-01 1.06289518e+00 7.03343511e-01 -3.99310708e-01
-6.96877003e-01 3.32622901e-02 -4.99413043e-01 1.09177279e+00
6.19053066e-01 -4.44710582e-01 -1.74081817e-01 3.20904136e-01
3.58350664e-01 5.16411185e-01 4.55331892e-01 1.01911426e+00
-4.99641806e-01 -3.75942528e-01 5.14709279e-02 3.52873772e-01
-1.56113255e+00 -1.13447703e-01 4.19362634e-02 3.31027985e-01
-9.49301347e-02 1.11716890e+00 1.19308472e-01 -2.13318206e-02
4.13408875e-01 -5.38704872e-01 3.73654902e-01 5.34905903e-02
-9.48020637e-01 4.87632871e-01 -4.76363033e-01 -1.20893812e+00
-6.23456776e-01 6.23500235e-02 -1.61627781e+00 -8.86265934e-01
3.32846977e-02 -3.56776923e-01 1.00987887e+00 5.10918856e-01
5.31410038e-01 9.17340100e-01 2.85169631e-01 -9.00907874e-01
2.37391382e-01 -7.67782450e-01 -7.81174898e-01 9.42263007e-02
7.07005084e-01 2.56184876e-01 -6.06276393e-01 6.68339550e-01] | [13.035679817199707, -2.6352291107177734] |
814e75aa-c776-43d0-8300-94a01513ed9e | program-synthesis-for-the-oeis | 2202.11908 | null | https://arxiv.org/abs/2202.11908v3 | https://arxiv.org/pdf/2202.11908v3.pdf | Learning Program Synthesis for Integer Sequences from Scratch | We present a self-learning approach for synthesizing programs from integer sequences. Our method relies on a tree search guided by a learned policy. Our system is tested on the On-Line Encyclopedia of Integer Sequences. There, it discovers, on its own, solutions for 27987 sequences starting from basic operators and without human-written training examples. | ['Josef Urban', 'Thibault Gauthier'] | 2022-02-24 | null | null | null | null | ['self-learning'] | ['natural-language-processing'] | [ 2.56995529e-01 1.61400050e-01 -9.87064362e-01 -2.22385019e-01
-8.62105012e-01 -7.89342523e-01 8.45116153e-02 1.26840934e-01
-3.64505559e-01 1.27100956e+00 -5.12311123e-02 -9.49413180e-01
4.86417785e-02 -8.88264716e-01 -8.40676069e-01 -3.46033096e-01
-5.88998735e-01 5.76506436e-01 3.74084115e-01 -5.94953716e-01
4.89967197e-01 3.41904938e-01 -1.60732996e+00 6.61654830e-01
9.06341136e-01 5.94122529e-01 2.00678200e-01 1.23945475e+00
-2.18498468e-01 1.14124072e+00 -6.71168685e-01 -4.34506804e-01
3.63730520e-01 -4.31690156e-01 -7.50730872e-01 -9.95626301e-03
1.72819734e-01 1.45108830e-02 -4.40295815e-01 9.17511404e-01
1.89205572e-01 -3.32986534e-01 2.73772866e-01 -1.22218323e+00
-3.21202308e-01 1.05700040e+00 -1.42018557e-01 4.72414643e-02
6.80888712e-01 6.62544847e-01 1.30619979e+00 -4.04381841e-01
7.42793441e-01 1.12773216e+00 7.75661170e-01 4.66013670e-01
-1.63886487e+00 -5.55323243e-01 -3.62759680e-01 -5.40139675e-02
-1.15758407e+00 -2.94862330e-01 6.20309949e-01 -2.60390580e-01
1.65054846e+00 2.40223110e-01 6.41144454e-01 7.07483888e-01
5.65273464e-01 9.26183105e-01 9.22824740e-01 -7.77238429e-01
4.31252152e-01 -4.61109616e-02 -1.80350378e-01 1.24390519e+00
2.29255250e-03 6.31452143e-01 -1.95048913e-01 -5.66306472e-01
7.71771371e-01 -7.71608412e-01 8.96658152e-02 -6.48883224e-01
-1.16661227e+00 9.20145094e-01 6.59412593e-02 2.74561673e-01
-2.93492913e-01 4.63858932e-01 9.44996834e-01 8.15422297e-01
-1.01512082e-01 8.21488678e-01 -8.55304599e-01 -5.07664502e-01
-8.14006925e-01 3.74755859e-01 1.43463647e+00 1.04260623e+00
7.27814317e-01 3.79252195e-01 -3.85603383e-02 3.37626010e-01
-6.28822073e-02 4.08787578e-01 6.03382826e-01 -1.11739588e+00
6.72441363e-01 6.70671284e-01 5.02610207e-02 -5.40696025e-01
-3.53867412e-01 -1.48295000e-01 -2.93169528e-01 3.25975597e-01
2.29201451e-01 -8.03368747e-01 -5.19355357e-01 1.42707157e+00
-1.66272774e-01 9.46496949e-02 4.18495595e-01 8.52677003e-02
5.54054320e-01 9.36962843e-01 -3.16189110e-01 -5.57624638e-01
6.29728734e-01 -1.28296411e+00 -4.60899860e-01 -1.87460750e-01
8.95549178e-01 -3.94163221e-01 8.24204445e-01 8.41535389e-01
-1.30742133e+00 -4.80133146e-01 -1.16717398e+00 4.75794196e-01
-1.48647174e-01 1.39785022e-01 1.22523367e+00 7.92250693e-01
-1.26390183e+00 7.02182174e-01 -7.29989231e-01 1.57622974e-02
-9.35835317e-02 6.17714107e-01 -3.25468816e-02 3.57894719e-01
-9.01980639e-01 7.98487306e-01 1.25833499e+00 -2.65527368e-01
-7.11569190e-01 -3.82541597e-01 -1.04171872e+00 -6.40184432e-02
6.79203987e-01 -4.44651514e-01 1.70594811e+00 -1.08585238e+00
-1.95349908e+00 6.73362851e-01 -2.06965432e-01 -1.02786851e+00
3.92221749e-01 2.12266356e-01 -5.97711325e-01 -4.95337993e-02
-2.09819153e-01 4.08678979e-01 9.34466600e-01 -1.17817497e+00
-7.98811913e-01 3.47677290e-01 2.14721248e-01 -2.31226385e-01
1.02896102e-01 2.38972425e-01 -1.48058116e-01 -6.29332900e-01
-3.49157691e-01 -1.03071332e+00 -5.10729969e-01 -4.38421100e-01
-2.40419433e-01 -3.97108555e-01 3.17170024e-01 -4.74647671e-01
1.66337419e+00 -1.92379141e+00 3.94266725e-01 5.16222477e-01
-1.39957413e-01 2.75732398e-01 -2.48018011e-01 7.14024127e-01
-3.08644950e-01 -1.77703112e-01 -3.48512799e-01 4.38056320e-01
2.66188562e-01 4.68821347e-01 -5.27679920e-01 2.06305295e-01
1.08593516e-01 1.03023636e+00 -1.17702997e+00 -8.19031358e-01
-1.29556596e-01 -6.05468571e-01 -1.16045165e+00 3.71962845e-01
-1.02229106e+00 -2.31737241e-01 -1.48537993e-01 7.39403784e-01
1.64220318e-01 -1.68949053e-01 5.79462290e-01 2.56482601e-01
-2.91060925e-01 2.38036022e-01 -1.01315784e+00 1.86461210e+00
-6.23700976e-01 4.90437597e-01 -3.19261909e-01 -1.06831038e+00
1.07486248e+00 2.65530676e-01 3.18753749e-01 -3.96557659e-01
-1.04999334e-01 5.26539743e-01 3.76005232e-01 -5.76918244e-01
2.18420804e-01 1.71833009e-01 -4.30668831e-01 6.43370390e-01
1.94866315e-01 -7.52334058e-01 8.82704973e-01 -9.69431028e-02
1.48930836e+00 2.11445093e-01 8.10395241e-01 -2.73066849e-01
8.28681827e-01 3.95956755e-01 3.41289014e-01 8.52646887e-01
2.39482060e-01 -1.25536993e-01 8.74584794e-01 -8.78339827e-01
-1.14121282e+00 -7.24054515e-01 2.17740029e-01 1.32567751e+00
-4.72575426e-01 -8.40695560e-01 -6.07889354e-01 -6.54064715e-01
1.28200993e-01 8.99067223e-01 -2.75314063e-01 -2.08428308e-01
-1.08869421e+00 -2.06859201e-01 6.85179710e-01 4.30252790e-01
2.97742546e-01 -1.79835033e+00 -8.98980200e-01 5.33572078e-01
3.18606555e-01 -9.11224186e-01 -4.69221771e-01 7.41513431e-01
-9.52047586e-01 -1.02119792e+00 -1.96745440e-01 -1.13046157e+00
4.15964484e-01 -4.39024627e-01 1.50049984e+00 3.70500833e-01
-3.57388407e-01 -9.76075232e-03 1.70744304e-02 -1.44889817e-01
-1.24205184e+00 5.04899085e-01 -1.38396785e-01 -7.51347184e-01
7.33834729e-02 -6.35618865e-01 3.78202677e-01 3.15929651e-02
-6.46478176e-01 -2.65516847e-01 8.28522325e-01 1.10014772e+00
1.91517338e-01 4.39103603e-01 4.04203415e-01 -7.85460830e-01
8.54382753e-01 -2.73998410e-01 -1.31421995e+00 5.55240631e-01
-4.87942010e-01 7.20848203e-01 8.83971989e-01 -3.72637361e-01
-6.51556194e-01 5.82177103e-01 -1.29714727e-01 -1.04341507e-01
-1.29752485e-02 5.10677040e-01 1.37636811e-01 -3.19755346e-01
1.15863657e+00 5.30718327e-01 8.77669156e-02 1.12701029e-01
3.45536679e-01 4.40046221e-01 8.93744886e-01 -1.19191599e+00
9.26042080e-01 -3.11607718e-01 -7.80172572e-02 -5.67527473e-01
-2.81061292e-01 3.26563604e-02 -5.68241417e-01 2.93887824e-01
2.76434094e-01 -6.29673958e-01 -9.44991112e-01 2.57522851e-01
-1.12904930e+00 -8.82115722e-01 -9.24266130e-02 2.00293139e-02
-1.24137199e+00 1.60574332e-01 -5.66036046e-01 -6.37818813e-01
-3.14919382e-01 -1.16601408e+00 8.22755635e-01 -2.13890877e-02
-5.08228183e-01 -8.55002463e-01 4.99681205e-01 -3.08113098e-01
2.01183572e-01 6.24475814e-02 1.36127484e+00 -6.11698151e-01
-6.39365196e-01 -1.78710967e-01 1.62196875e-01 1.86769724e-01
2.30142131e-01 4.06614363e-01 -2.36468136e-01 -2.46826097e-01
-3.43270212e-01 -6.53618634e-01 3.45226765e-01 2.08571609e-02
1.35231435e+00 -6.35524392e-01 -2.49729931e-01 6.24175847e-01
1.55150127e+00 4.63451594e-01 4.64779407e-01 6.28153205e-01
1.39665738e-01 6.98124543e-02 6.19451880e-01 6.10409856e-01
7.14943334e-02 4.26090449e-01 4.17645216e-01 2.57290095e-01
4.75050628e-01 -4.32736099e-01 4.24760312e-01 4.22235191e-01
-6.59774104e-03 2.85060350e-02 -1.32759106e+00 4.90909010e-01
-1.77205312e+00 -9.37972844e-01 4.38814461e-01 1.80041397e+00
1.52607679e+00 7.26816654e-01 4.15127784e-01 2.62210190e-01
3.87948155e-01 1.20856829e-01 -7.55527675e-01 -9.71687317e-01
1.97953463e-01 8.45021129e-01 8.55289996e-01 6.28413737e-01
-1.10132945e+00 1.33475292e+00 8.32787132e+00 6.89093769e-01
-1.03692031e+00 -6.75365567e-01 1.12658814e-01 1.88874230e-01
-3.86920512e-01 3.86548042e-02 -8.05967212e-01 2.10077032e-01
1.19310892e+00 -9.31288004e-01 9.00781155e-01 1.19068742e+00
-2.60295719e-01 -1.19239092e-02 -1.22602820e+00 7.31870532e-01
-6.49773479e-02 -1.82404351e+00 -2.52241731e-01 -3.33388776e-01
9.13060784e-01 -2.56430954e-01 -1.63386613e-01 8.15753937e-01
8.77629042e-01 -1.22991908e+00 5.47215641e-01 1.32002262e-02
6.57883883e-01 -1.18959749e+00 5.20732164e-01 7.28951097e-01
-1.12416899e+00 -5.47640920e-01 6.82313368e-03 -1.76960528e-01
-1.45160735e-01 9.67611372e-02 -1.03997481e+00 3.60118508e-01
9.63018760e-02 7.09448814e-01 -6.10389590e-01 1.11760986e+00
-4.27504063e-01 4.89467055e-01 -1.51731208e-01 -3.56685907e-01
4.31478441e-01 5.12533188e-02 1.98480099e-01 1.37335849e+00
1.45351231e-01 1.78754434e-01 4.56272066e-01 6.88848495e-01
6.96770921e-02 -1.75837744e-02 -7.95041144e-01 -3.58321995e-01
3.15440953e-01 6.70202851e-01 -6.66839123e-01 -5.30241013e-01
-2.66376704e-01 4.91796851e-01 3.00385028e-01 2.25095466e-01
-1.07741976e+00 -7.22129822e-01 2.39096969e-01 -3.07678998e-01
6.54369891e-01 -4.15767521e-01 -2.34558925e-01 -1.00086772e+00
-1.25413746e-01 -1.94701815e+00 3.87075007e-01 -4.28022444e-01
-7.24004984e-01 7.23366201e-01 1.15500323e-01 -8.68281186e-01
-1.06526697e+00 -6.44238114e-01 -6.70692444e-01 6.29393995e-01
-1.05030394e+00 -2.20948815e-01 2.14468345e-01 3.82841617e-01
7.05581963e-01 -7.68109918e-01 9.61030900e-01 -1.38276201e-02
-2.53209203e-01 7.85497963e-01 -1.10440910e-01 -1.51079074e-01
2.46777460e-01 -1.40425909e+00 9.33944404e-01 4.69733357e-01
2.29164720e-01 5.68779409e-01 9.73483860e-01 -5.68174541e-01
-1.90294313e+00 -5.95617414e-01 8.55842710e-01 2.60355938e-02
1.23025131e+00 -1.28136307e-01 -7.41099715e-01 9.12804067e-01
2.98997283e-01 -1.56075537e-01 2.25594312e-01 -1.66293308e-01
-1.97668955e-01 -5.06459922e-03 -1.01801014e+00 6.61924303e-01
1.20765293e+00 -4.18286920e-01 -8.08367670e-01 6.30754530e-01
1.02173209e+00 -1.03419352e+00 -8.00681651e-01 5.56141078e-01
2.24661544e-01 -8.32226932e-01 1.03632486e+00 -9.41228271e-01
8.30845892e-01 -1.48639947e-01 -2.42700111e-02 -1.28633916e+00
-1.08563773e-01 -1.26824975e+00 -3.91474903e-01 4.75040495e-01
7.55639851e-01 -5.87160647e-01 1.10264528e+00 3.20125282e-01
-1.36592552e-01 -8.13157082e-01 -6.30706310e-01 -1.01502144e+00
1.23477494e-02 -2.75150716e-01 8.48983526e-01 4.92649078e-01
3.08432609e-01 2.19345048e-01 -2.41366625e-01 -2.14355752e-01
3.80769074e-01 5.96171618e-01 9.09234226e-01 -7.15622783e-01
-9.79582429e-01 -5.93594015e-01 -1.36634678e-01 -8.76944244e-01
8.29265058e-01 -8.87949586e-01 1.20545596e-01 -7.88552105e-01
-2.93064326e-01 -3.75969768e-01 1.27499416e-01 7.70296037e-01
2.71190554e-01 -4.48773921e-01 -5.31445295e-02 -3.47791404e-01
-5.70853174e-01 2.37481549e-01 8.48235905e-01 -2.07442880e-01
-4.46866632e-01 3.19021106e-01 -3.86481315e-01 8.07226598e-01
1.16630256e+00 -3.73907417e-01 -5.34450352e-01 -4.86515276e-02
5.95635414e-01 6.71470821e-01 -2.15864614e-01 -1.00779355e+00
2.90822685e-01 -5.66782951e-01 -2.78533250e-01 -6.94617927e-01
-2.38412991e-01 -7.30311275e-01 5.58929071e-02 1.10418725e+00
-5.71628213e-01 7.49698818e-01 6.04705513e-01 1.66486531e-01
-1.38537019e-01 -8.20937514e-01 4.65095729e-01 -3.46182734e-01
-9.22064185e-01 -1.77211344e-01 -4.62770820e-01 2.49563888e-01
1.15866876e+00 1.18422069e-01 -2.07459088e-02 -1.57153547e-01
-5.39124846e-01 2.40838185e-01 5.63490808e-01 1.63227618e-01
3.43816489e-01 -1.13317621e+00 -5.42185783e-01 3.59464198e-01
3.57974470e-02 -2.90373296e-01 -7.50634551e-01 7.95588717e-02
-9.41207349e-01 8.53389621e-01 -4.34841663e-01 -4.58951801e-01
-1.32285726e+00 9.24690843e-01 2.63882518e-01 -6.96503520e-01
-4.66233850e-01 5.72128713e-01 -7.08667576e-01 -7.29880393e-01
6.44888878e-01 -7.18119562e-01 2.58765519e-01 -4.93588775e-01
2.88422316e-01 1.00822486e-01 -1.10325120e-01 1.72474876e-01
-2.16965824e-01 3.29915643e-01 1.82496548e-01 -2.11560503e-01
1.26242411e+00 8.54949534e-01 -5.22694051e-01 4.65128511e-01
1.04416597e+00 -2.00458728e-02 -8.38408887e-01 -3.19004923e-01
6.39512420e-01 -3.98302853e-01 -5.44824243e-01 -5.91709614e-01
-7.33123660e-01 1.83088765e-01 5.71831428e-02 1.45394191e-01
1.05857503e+00 -2.81509101e-01 6.56070352e-01 1.32224727e+00
7.79883087e-01 -8.03576112e-01 3.84114027e-01 7.99709439e-01
7.67979980e-01 -1.08280683e+00 1.61255002e-01 -2.59811968e-01
-6.48445964e-01 1.51968646e+00 5.50897896e-01 -5.91238618e-01
2.23081946e-01 8.09154689e-01 -3.59916896e-01 2.10380912e-01
-1.36914849e+00 4.79738452e-02 -7.03467131e-02 5.17781973e-01
3.44454259e-01 -1.11550186e-02 -3.40690315e-01 1.78645819e-01
-6.80433929e-01 4.44103152e-01 5.01153767e-01 1.31980228e+00
-5.36533475e-01 -1.50756240e+00 -3.41834247e-01 2.73065299e-01
-9.01170149e-02 -2.51552224e-01 -1.10621013e-01 9.26185727e-01
-3.76767218e-02 2.72272289e-01 -1.87363133e-01 -1.82355523e-01
1.46701038e-01 2.06909060e-01 8.21961701e-01 -7.42953062e-01
-7.86274910e-01 -3.10852021e-01 4.86142755e-01 -6.48407519e-01
-6.89823031e-02 -6.05452776e-01 -1.37753141e+00 -1.80352241e-01
2.31152818e-01 3.55195343e-01 4.38855469e-01 5.83785415e-01
-1.27064526e-01 5.48137546e-01 1.00895929e+00 -6.33475125e-01
-1.01716232e+00 -3.13613117e-01 -7.16821030e-02 -4.86388445e-01
4.52682137e-01 -8.65534395e-02 -2.65643656e-01 1.56706292e-02] | [8.230254173278809, 7.393875598907471] |
8cc71dfa-f5ba-4135-9674-fd092316784f | open-set-domain-adaptation-by-novel-class | 2203.03329 | null | https://arxiv.org/abs/2203.03329v1 | https://arxiv.org/pdf/2203.03329v1.pdf | Open Set Domain Adaptation By Novel Class Discovery | In Open Set Domain Adaptation (OSDA), large amounts of target samples are drawn from the implicit categories that never appear in the source domain. Due to the lack of their specific belonging, existing methods indiscriminately regard them as a single class unknown. We challenge this broadly-adopted practice that may arouse unexpected detrimental effects because the decision boundaries between the implicit categories have been fully ignored. Instead, we propose Self-supervised Class-Discovering Adapter (SCDA) that attempts to achieve OSDA by gradually discovering those implicit classes, then incorporating them to restructure the classifier and update the domain-adaptive features iteratively. SCDA performs two alternate steps to achieve implicit class discovery and self-supervised OSDA, respectively. By jointly optimizing for two tasks, SCDA achieves the state-of-the-art in OSDA and shows a competitive performance to unearth the implicit target classes. | ['Liang Lin', 'Guanbin Li', 'Pengxu Wei', 'Ziliang Chen', 'Jingyu Zhuang'] | 2022-03-07 | null | null | null | null | ['novel-class-discovery', 'novel-class-discovery'] | ['computer-vision', 'methodology'] | [ 2.22072288e-01 3.89133632e-01 -5.30592084e-01 -5.79046667e-01
-6.21869862e-01 -7.37968564e-01 5.65055430e-01 4.77369828e-03
-3.02090794e-01 1.00681019e+00 2.28160582e-02 1.89095177e-03
-4.41744961e-02 -6.85154974e-01 -6.38687611e-01 -6.57473683e-01
1.97786778e-01 7.65383720e-01 4.00052071e-01 1.38164669e-01
-1.05882414e-01 2.32472748e-01 -1.74897659e+00 2.99014211e-01
1.02495158e+00 8.43880773e-01 -8.69627856e-03 3.00637875e-02
-2.77375162e-01 6.17316961e-01 -5.42837799e-01 -3.81302595e-01
2.09579349e-01 -2.78713077e-01 -8.12545359e-01 2.81527013e-01
4.54732835e-01 -3.79623234e-01 -1.16064169e-01 9.89260495e-01
2.24837407e-01 2.96259880e-01 8.56711686e-01 -1.23609829e+00
-8.47089589e-01 5.48289418e-01 -5.43013871e-01 5.15686423e-02
-1.14713050e-01 5.90632223e-02 1.11375952e+00 -1.03221834e+00
4.57446694e-01 1.07017982e+00 6.87789798e-01 7.28725970e-01
-1.47405601e+00 -9.87970769e-01 6.50745988e-01 1.55931920e-01
-1.56996870e+00 -6.42641485e-01 9.11341190e-01 -4.39521551e-01
6.09086215e-01 1.42068312e-01 2.21531153e-01 1.25862920e+00
-4.72439915e-01 9.59323287e-01 1.05331516e+00 -4.52495724e-01
4.88055319e-01 6.79476142e-01 6.26480699e-01 2.82663345e-01
3.96988928e-01 1.82772785e-01 -4.51161772e-01 -3.88835341e-01
5.00961661e-01 1.46038570e-02 1.05337374e-01 -7.82687545e-01
-9.54313695e-01 7.84166098e-01 1.65735528e-01 8.40925649e-02
-2.33364210e-01 -3.80437434e-01 3.59561145e-01 4.30253714e-01
6.72762096e-01 4.68265921e-01 -9.81622159e-01 5.44523075e-02
-6.00465775e-01 2.69564927e-01 7.29034781e-01 1.07136631e+00
1.19071865e+00 -2.42664993e-01 6.24607131e-03 1.30955648e+00
1.42348751e-01 1.38734266e-01 8.03578913e-01 -6.92822516e-01
1.10850193e-01 8.76728714e-01 2.15239599e-01 -4.77718890e-01
-4.23232093e-02 -7.56524682e-01 -5.12929022e-01 2.09520787e-01
5.13315678e-01 3.42250578e-02 -1.04337764e+00 1.82634282e+00
6.15602076e-01 3.16190332e-01 1.91632092e-01 7.06175566e-01
5.67858040e-01 3.83251637e-01 1.99787885e-01 -2.77305275e-01
1.02136564e+00 -1.03577006e+00 -5.03263295e-01 -5.05491078e-01
5.31776130e-01 -3.14217091e-01 1.30861533e+00 4.88941848e-01
-3.55938405e-01 -6.19379282e-01 -1.17647529e+00 2.06848666e-01
-3.78743380e-01 8.68056193e-02 5.57131886e-01 5.05208194e-01
-4.05714542e-01 4.94577050e-01 -6.32985473e-01 -2.00942345e-02
8.53240728e-01 3.94410014e-01 -3.41742903e-01 -1.50468230e-01
-1.02865613e+00 4.84907031e-01 5.10077417e-01 -4.07199979e-01
-9.10828292e-01 -9.69107926e-01 -5.86860061e-01 -9.42000896e-02
8.05084705e-01 -6.08071864e-01 1.32108819e+00 -1.32456839e+00
-1.45692134e+00 1.09810960e+00 -2.13728532e-01 -3.75663161e-01
4.07014191e-01 -3.97579819e-01 -5.55067718e-01 -3.92323881e-01
3.78337026e-01 5.14026701e-01 1.14113724e+00 -1.41167915e+00
-9.34550941e-01 -4.47205842e-01 -7.83758089e-02 2.55476505e-01
-7.70308197e-01 -5.16873419e-01 -1.69233948e-01 -7.64529765e-01
1.63237095e-01 -7.58433998e-01 -1.06633708e-01 1.51248127e-01
-3.24550152e-01 -5.75714469e-01 9.55500305e-01 -1.27630100e-01
1.39048922e+00 -2.24586916e+00 -1.04223974e-02 4.72827395e-03
5.56046724e-01 4.05833125e-01 -2.22740129e-01 -2.18954176e-01
-1.16345547e-01 -1.45012721e-01 -5.24801552e-01 -3.16292852e-01
-1.14918679e-01 3.78740847e-01 -5.22013962e-01 1.93799600e-01
2.34724686e-01 5.27252376e-01 -1.07019091e+00 -1.61664844e-01
-8.48683864e-02 -1.10186264e-01 -6.56507134e-01 3.79779249e-01
-4.58814770e-01 2.97712028e-01 -4.95836705e-01 6.72832906e-01
8.93763900e-01 -3.50617379e-01 2.70310640e-01 1.62145160e-02
1.26981199e-01 3.64197463e-01 -1.19659221e+00 1.36724377e+00
-1.74263492e-01 2.27941945e-01 -1.99119121e-01 -1.16973388e+00
9.88567948e-01 -1.07217103e-01 4.60045189e-01 -3.10929090e-01
-1.29925847e-01 3.32420439e-01 -5.32603450e-02 -1.92300662e-01
1.64417312e-01 -3.44847620e-01 -1.06076961e-02 4.36372161e-01
2.23524943e-01 2.90243000e-01 -1.57445878e-01 -1.41753748e-01
1.05945468e+00 2.34565854e-01 7.20985293e-01 -3.40631336e-01
5.71446896e-01 2.92653084e-01 8.01899076e-01 8.31495583e-01
-4.23239142e-01 5.33650577e-01 2.88192928e-01 -5.20879447e-01
-1.05917549e+00 -1.47952425e+00 -4.52321708e-01 1.41264999e+00
5.58885485e-02 -3.39739509e-02 -4.98039335e-01 -1.33214808e+00
2.89267451e-01 8.27632487e-01 -7.87179351e-01 -4.73843455e-01
-1.22445442e-01 -6.80346906e-01 3.64959031e-01 3.64061087e-01
3.04547995e-01 -7.53346801e-01 -1.01565406e-01 3.36904734e-01
4.61000465e-02 -8.50136578e-01 -5.42550385e-01 5.18037260e-01
-9.66195941e-01 -1.17914641e+00 -6.64207101e-01 -6.21234477e-01
7.20046341e-01 3.39358807e-01 1.06392670e+00 -2.77899265e-01
1.31216496e-01 2.47272793e-02 -4.64259833e-01 -4.03909743e-01
-4.44349468e-01 2.50592679e-01 3.02424610e-01 3.91265273e-01
9.90000546e-01 -8.70858729e-01 -3.15369368e-01 4.48968917e-01
-6.19696319e-01 -1.87255234e-01 5.58274806e-01 1.03898525e+00
6.26762629e-01 1.29599795e-01 1.20253646e+00 -1.59931433e+00
3.12963486e-01 -9.20596778e-01 -4.15397257e-01 1.01617321e-01
-9.84965980e-01 -2.11410671e-02 7.48134375e-01 -1.06124675e+00
-1.05516791e+00 1.68063059e-01 2.53047675e-01 -6.56869709e-01
-5.22048831e-01 3.13824683e-01 -4.22049254e-01 2.98319727e-01
1.03248000e+00 3.05075616e-01 1.23285949e-01 -7.83201933e-01
3.07774693e-01 9.21685517e-01 5.51461220e-01 -6.55685186e-01
9.53073978e-01 3.72568041e-01 -7.49172628e-01 -6.32777095e-01
-1.48043907e+00 -7.43104041e-01 -9.76856351e-01 2.42706195e-01
2.69520819e-01 -1.22993147e+00 7.26918057e-02 5.31400204e-01
-7.62595177e-01 -3.55532229e-01 -6.95739210e-01 1.63878843e-01
-2.89508492e-01 3.55347425e-01 4.77361642e-02 -5.43265164e-01
-6.48084506e-02 -6.58881903e-01 7.60705948e-01 1.64440498e-01
-4.83284026e-01 -8.86005819e-01 9.61106643e-02 1.80317193e-01
1.59865454e-01 -1.23162009e-01 9.80646968e-01 -1.12756872e+00
-2.38170832e-01 -2.16339678e-01 -6.48678690e-02 5.41622698e-01
6.55704319e-01 -4.64749426e-01 -1.35455573e+00 -3.23488027e-01
-1.72584027e-01 -4.46463495e-01 7.23151088e-01 1.29230008e-01
1.27402735e+00 -4.27423716e-01 -4.53070194e-01 5.15626073e-01
1.01471341e+00 2.08194420e-01 2.04270124e-01 4.05471772e-01
5.57772279e-01 5.50425768e-01 6.86894774e-01 6.35296822e-01
3.94427150e-01 6.03565276e-01 1.81112409e-01 -1.40946796e-02
-1.49297535e-01 -4.17953193e-01 3.12245548e-01 5.75918853e-01
3.79686564e-01 1.80205014e-02 -6.94634378e-01 8.38332295e-01
-1.87789822e+00 -6.14768028e-01 9.73570570e-02 2.47119641e+00
1.38212061e+00 3.22728962e-01 4.16281849e-01 3.82978953e-02
7.95542538e-01 -1.14048414e-01 -1.11703587e+00 6.40931875e-02
-1.09301984e-01 2.39182830e-01 2.50143439e-01 2.19145253e-01
-1.35480571e+00 9.77556705e-01 6.57591200e+00 9.01960492e-01
-9.84852850e-01 3.22695136e-01 5.71164072e-01 -5.71438260e-02
-3.29617709e-01 3.67626362e-02 -9.94702220e-01 4.79149669e-01
7.07043469e-01 -3.94632250e-01 3.91839802e-01 1.25271904e+00
-2.45579720e-01 1.33360684e-01 -1.41271150e+00 7.60648012e-01
-2.18257084e-01 -1.03551507e+00 2.16600865e-01 5.68724796e-02
8.66521418e-01 -5.35118841e-02 6.27645329e-02 7.22760618e-01
5.95130920e-01 -6.07174337e-01 6.91614330e-01 1.98862150e-01
1.00678861e+00 -5.33491373e-01 5.27278066e-01 5.57607234e-01
-8.09635758e-01 -4.94883925e-01 -5.31246245e-01 -1.84964761e-01
-4.44603503e-01 6.87465549e-01 -8.40407491e-01 1.72232717e-01
6.91490054e-01 9.88997757e-01 -5.39687634e-01 7.96437562e-01
-3.27545069e-02 9.80378628e-01 -3.46199900e-01 2.69642621e-01
1.26383677e-01 -7.80251250e-02 6.14148021e-01 7.13856399e-01
-2.80708168e-02 1.59277946e-01 2.39480630e-01 1.03069389e+00
-2.08282411e-01 -9.75878984e-02 -4.94803488e-01 -8.92206654e-02
7.89574742e-01 7.91976929e-01 -2.89474010e-01 -5.74616373e-01
-5.78075588e-01 1.04561496e+00 6.28320158e-01 3.92137825e-01
-5.93300641e-01 -2.29455218e-01 1.10102499e+00 3.30924273e-01
2.99629092e-01 2.38903105e-01 -4.86400992e-01 -1.42563653e+00
-1.10162413e-02 -9.30276752e-01 8.47291291e-01 -3.16588670e-01
-1.81207240e+00 3.98818135e-01 -3.02278716e-02 -1.49544096e+00
-6.41479418e-02 -4.49548572e-01 -3.52700710e-01 7.14452088e-01
-1.55802655e+00 -9.35803473e-01 -1.35285765e-01 5.21926701e-01
8.60511303e-01 -4.59537655e-01 9.16670918e-01 4.00273025e-01
-4.99580264e-01 1.06686556e+00 4.47893471e-01 1.42312124e-02
1.01922739e+00 -1.18353760e+00 1.05135806e-01 6.11408532e-01
5.31941950e-02 6.13511801e-01 4.62908030e-01 -6.37448311e-01
-7.77435482e-01 -1.33553219e+00 5.88414729e-01 -5.95193446e-01
7.09340274e-01 -4.76089358e-01 -1.35132027e+00 8.24293852e-01
-4.75272954e-01 3.04608643e-01 7.87998796e-01 4.85180676e-01
-7.51653254e-01 -3.89038295e-01 -1.14639437e+00 3.11851382e-01
1.15848088e+00 -4.50663358e-01 -9.59490418e-01 8.30506086e-02
8.65256131e-01 -8.34096447e-02 -5.98927140e-01 3.42134148e-01
4.69200343e-01 -6.10504925e-01 1.01305532e+00 -9.09215927e-01
3.65640163e-01 -2.52463281e-01 -1.23677656e-01 -1.38780677e+00
-5.20837188e-01 -3.34608138e-01 -4.85776275e-01 1.33321822e+00
4.03625906e-01 -9.21203017e-01 9.72088873e-01 4.61380541e-01
-1.15521774e-01 -4.83395159e-01 -9.15018857e-01 -1.07119846e+00
2.83965021e-01 -1.67436481e-01 7.70058930e-01 1.44369102e+00
-2.05322519e-01 6.02496445e-01 -2.99750954e-01 3.12742442e-01
8.98830354e-01 2.34851316e-01 8.53468418e-01 -1.62742281e+00
-3.03596884e-01 -2.58864462e-01 -1.26707330e-01 -1.19766974e+00
2.62927324e-01 -9.79599178e-01 3.96646336e-02 -7.22564638e-01
3.10143769e-01 -1.15689623e+00 -7.49813080e-01 8.77878070e-01
-4.84766334e-01 5.53304777e-02 -3.00190210e-01 4.96479005e-01
-7.30370462e-01 6.29307568e-01 9.75867867e-01 -1.62939772e-01
-5.30599236e-01 2.84829021e-01 -1.26179898e+00 6.04185224e-01
4.73546654e-01 -7.54294217e-01 -6.02216780e-01 -2.01030239e-01
-3.59550089e-01 -4.62755114e-01 1.75821230e-01 -8.48149955e-01
-5.14831841e-02 -5.00350654e-01 3.28698188e-01 -4.76237297e-01
1.03049517e-01 -9.68036473e-01 6.19723909e-02 2.74652183e-01
-4.47703123e-01 -7.81552911e-01 8.71304497e-02 8.95290494e-01
-1.70359716e-01 -2.89892018e-01 9.74686146e-01 4.26133722e-02
-9.20930803e-01 3.78912061e-01 -2.70067275e-01 3.48230839e-01
9.87925649e-01 -4.93248373e-01 -1.71728685e-01 5.07882684e-02
-9.42875981e-01 1.98694780e-01 4.21857387e-01 5.39406717e-01
3.49468172e-01 -1.27527189e+00 -6.11694753e-01 4.40938652e-01
5.19447684e-01 3.83854896e-01 1.09347500e-01 3.17161798e-01
1.90702975e-01 1.16555348e-01 5.06992489e-02 -6.03105247e-01
-1.02747512e+00 7.51981437e-01 3.31265867e-01 -2.29557529e-01
-6.38784051e-01 8.65571439e-01 7.27987051e-01 -9.74221826e-01
2.88488209e-01 5.06805815e-02 -2.23309010e-01 6.77748919e-02
4.62433755e-01 1.96614876e-01 -1.19728623e-02 -2.43712276e-01
-2.45300084e-01 6.65909871e-02 -5.42617083e-01 5.19866705e-01
1.57001925e+00 -3.26285779e-01 2.19868049e-01 5.89463413e-01
1.12063265e+00 -5.43491915e-03 -1.65582180e+00 -9.01763439e-01
2.87131131e-01 -5.50450683e-01 -4.87249605e-02 -1.13531291e+00
-6.48968458e-01 5.69062054e-01 7.44515240e-01 -1.81749031e-01
1.11778510e+00 1.63156494e-01 5.73775947e-01 5.37464738e-01
3.29602897e-01 -1.18588483e+00 1.68839350e-01 4.86173779e-01
5.55101752e-01 -1.42175889e+00 9.00975615e-02 -6.67030036e-01
-5.16785920e-01 9.75126922e-01 9.74838495e-01 -1.51588535e-02
6.02539301e-01 9.73378494e-02 1.93432358e-03 1.79291040e-01
-9.14876878e-01 -1.60447821e-01 3.77706200e-01 9.80282545e-01
3.76050659e-02 2.33849436e-01 -2.80225314e-02 1.31977427e+00
1.17467403e-01 5.41240200e-02 3.38409066e-01 7.68343568e-01
-5.27582526e-01 -1.19364250e+00 -3.10188413e-01 7.99548924e-01
-1.17172286e-01 -6.78355694e-02 -4.73099679e-01 6.70035601e-01
2.06671491e-01 5.56440771e-01 1.45374417e-01 -3.42897356e-01
4.09540325e-01 3.01840186e-01 2.43047491e-01 -9.79298234e-01
-1.76747322e-01 -9.84501913e-02 6.73015788e-02 -1.41450703e-01
-8.03268924e-02 -7.68594801e-01 -1.01262116e+00 8.97169672e-03
-4.57960635e-01 1.51033267e-01 5.35399243e-02 9.89343941e-01
4.84657526e-01 2.67607957e-01 8.53493035e-01 -2.72103906e-01
-1.04205799e+00 -9.35659409e-01 -5.26270092e-01 5.45234501e-01
5.26746273e-01 -1.09608126e+00 -5.33034801e-01 2.21401885e-01] | [10.294285774230957, 3.151414155960083] |
0fdaa35d-08d1-4030-8f4c-521a5c0d639c | visual-place-recognition-with-low-resolution | 2305.05776 | null | https://arxiv.org/abs/2305.05776v1 | https://arxiv.org/pdf/2305.05776v1.pdf | Visual Place Recognition with Low-Resolution Images | Images incorporate a wealth of information from a robot's surroundings. With the widespread availability of compact cameras, visual information has become increasingly popular for addressing the localisation problem, which is then termed as Visual Place Recognition (VPR). While many applications use high-resolution cameras and high-end systems to achieve optimal place-matching performance, low-end commercial systems face limitations due to resource constraints and relatively low-resolution and low-quality cameras. In this paper, we analyse the effects of image resolution on the accuracy and robustness of well-established handcrafted VPR pipelines. Handcrafted designs have low computational demands and can adapt to flexible image resolutions, making them a suitable approach to scale to any image source and to operate under resource limitations. This paper aims to help academic researchers and companies in the hardware and software industry co-design VPR solutions and expand the use of VPR algorithms in commercial products. | ['Shoaib Ehsan', 'Klaus McDonald-Maier', 'Michael Milford', 'Bruno Ferrarini', 'Mihnea-Alexandru Tomita'] | 2023-05-09 | null | null | null | null | ['visual-place-recognition'] | ['computer-vision'] | [ 9.26571712e-02 -3.63813490e-01 -1.67715460e-01 -2.19441056e-01
-3.89156640e-01 -6.76204860e-01 6.70993447e-01 -1.06051914e-01
-6.45762682e-01 2.68418640e-01 -8.32038298e-02 4.32605632e-02
-3.12915146e-02 -5.24827898e-01 -5.75269699e-01 -2.33060405e-01
6.45142943e-02 1.89086109e-01 6.37668312e-01 -2.50831366e-01
6.24090374e-01 9.35050368e-01 -1.86547375e+00 -2.17894107e-01
3.10217589e-01 8.26697409e-01 7.17540801e-01 6.59153521e-01
1.97909281e-01 4.97488528e-01 -3.45652670e-01 -2.18315169e-01
5.03486395e-01 -8.57447833e-02 -4.59597439e-01 2.40642101e-01
4.48352039e-01 -2.47268960e-01 -4.38225418e-01 9.53160286e-01
5.28367341e-01 1.27544224e-01 2.22748339e-01 -1.36123753e+00
-5.93291521e-01 -2.57821679e-01 -6.73368514e-01 1.91375077e-01
7.77338922e-01 3.95988971e-01 7.53094494e-01 -7.49629676e-01
8.35008681e-01 8.93642545e-01 8.63042176e-01 3.02767932e-01
-1.22144783e+00 -3.75057340e-01 -2.64257491e-01 2.88838118e-01
-1.55947840e+00 -9.31736112e-01 6.24672890e-01 -4.38021749e-01
1.25231469e+00 -9.32751149e-02 3.95767629e-01 8.33243906e-01
5.70681803e-02 2.21400574e-01 8.12165618e-01 -3.01771134e-01
2.92856783e-01 5.08433320e-02 -6.88490987e-01 5.13242185e-01
4.80068624e-01 9.44616646e-02 -7.60634720e-01 2.21170694e-01
1.20769227e+00 5.12693264e-02 -2.10532919e-01 -9.33712542e-01
-1.40597498e+00 5.68563700e-01 6.62802041e-01 2.65895516e-01
-4.76231575e-01 1.98985279e-01 8.89049470e-02 6.06424436e-02
-1.51033148e-01 5.85814416e-01 -2.07150087e-01 -4.96958584e-01
-9.52487230e-01 2.83997338e-02 4.58429903e-01 1.21976697e+00
9.19473648e-01 -9.10417214e-02 6.90799832e-01 9.24533725e-01
5.61849117e-01 4.22010124e-01 5.47407389e-01 -1.35315633e+00
4.25607979e-01 5.81902444e-01 2.86161602e-01 -1.24947202e+00
-4.85390335e-01 -3.18465792e-02 -4.04318660e-01 5.83606601e-01
2.95142353e-01 1.89286098e-01 -8.33253205e-01 1.06775200e+00
1.54893175e-01 -3.57896119e-01 4.42344286e-02 1.21088409e+00
5.23502409e-01 4.76103991e-01 -1.93981186e-01 3.17240983e-01
1.19691503e+00 -8.03071856e-01 -4.02244657e-01 -1.01458836e+00
3.29555601e-01 -8.75590026e-01 7.88498402e-01 2.91058421e-01
-8.27933431e-01 -3.18293422e-01 -1.42875850e+00 -2.50630170e-01
-3.22786242e-01 5.66457659e-02 3.80779892e-01 5.25549829e-01
-1.22168064e+00 3.68931979e-01 -8.25722992e-01 -9.18780804e-01
2.84999728e-01 5.64014494e-01 -9.03920054e-01 -3.21672827e-01
-6.50654972e-01 1.39702892e+00 3.16018701e-01 3.05495374e-02
-5.78652024e-01 -3.69850665e-01 -1.02979529e+00 -2.25334212e-01
7.57127479e-02 -4.75100279e-01 1.17715490e+00 -1.01844013e+00
-1.60759842e+00 1.06511247e+00 1.24729447e-01 -4.16701615e-01
4.33442652e-01 5.51769398e-02 -3.20549041e-01 2.98151582e-01
1.18010879e-01 8.44747245e-01 7.01033175e-01 -9.68498826e-01
-9.02264237e-01 -3.69250625e-01 -7.12148473e-02 4.85986114e-01
9.98995006e-02 5.73332831e-02 -7.23872185e-01 -1.21125896e-02
3.64543051e-01 -9.87116814e-01 -3.34988475e-01 3.51113200e-01
2.60661036e-01 4.59085524e-01 9.43116307e-01 -4.22853172e-01
5.31964362e-01 -2.49953628e+00 -1.69226006e-01 2.08373256e-02
-1.54985741e-01 3.82249892e-01 -2.72002239e-02 4.73616928e-01
2.93371946e-01 -2.59748846e-01 1.03009388e-01 -1.15486912e-01
-1.32954195e-01 3.85653287e-01 1.11060031e-01 9.61939335e-01
-3.86465788e-02 5.03686249e-01 -8.44705224e-01 -4.34821665e-01
9.20229495e-01 5.91375887e-01 -4.08709586e-01 2.09036432e-02
3.61064188e-02 2.72036761e-01 -1.21973984e-01 6.45872653e-01
5.01492798e-01 -3.63737270e-02 1.36112457e-03 -3.92306224e-02
-4.82701451e-01 3.83392200e-02 -1.17616153e+00 1.84570324e+00
-4.69495386e-01 1.20059955e+00 4.12859350e-01 -4.58475590e-01
1.20900893e+00 -4.04873304e-02 2.69511372e-01 -1.11461806e+00
7.99746737e-02 3.90652478e-01 -1.89021692e-01 -3.01012069e-01
1.10131049e+00 -1.96827911e-02 6.03524186e-02 -3.13750744e-01
5.27134053e-02 -1.36051074e-01 8.23263600e-02 -1.46401882e-01
1.12108803e+00 3.18730921e-01 6.13730311e-01 -2.05070078e-02
2.15053409e-01 3.28401536e-01 3.27523679e-01 4.19358730e-01
-4.24718350e-01 1.00330722e+00 -3.61372642e-02 -2.87127197e-01
-1.41066635e+00 -8.60598147e-01 -9.25198644e-02 8.00455391e-01
6.55377626e-01 -2.33927861e-01 -3.39409441e-01 1.24048293e-01
3.48165557e-02 3.31795186e-01 -6.53288215e-02 6.82209954e-02
-4.48439062e-01 -1.48103371e-01 4.73843634e-01 5.35522640e-01
7.64198482e-01 -7.52240360e-01 -1.50598669e+00 2.32375279e-01
-1.24015763e-01 -1.45995963e+00 -1.96706071e-01 1.00754760e-01
-6.22732997e-01 -9.26530659e-01 -8.22243989e-01 -8.13398719e-01
5.38085520e-01 7.54220605e-01 6.70922518e-01 -2.80696779e-01
-4.88302827e-01 7.60232568e-01 -2.70672023e-01 -5.75568713e-02
1.87396985e-02 6.85204566e-02 1.38759717e-01 -1.92736641e-01
2.64119148e-01 -5.03919542e-01 -6.85724556e-01 6.04690433e-01
-7.04228818e-01 -1.94550335e-01 8.19260478e-01 4.39051807e-01
5.70488393e-01 -8.85167718e-02 -2.46139755e-03 -1.26677036e-01
2.23795578e-01 -2.37040237e-01 -1.02648139e+00 -7.52074923e-03
-6.54912472e-01 9.25077423e-02 4.35615301e-01 -2.88388371e-01
-9.59891438e-01 4.92939323e-01 8.08951855e-02 -5.00560939e-01
-2.73759335e-01 1.77677765e-01 -1.84240609e-01 -6.54213011e-01
8.37588072e-01 1.39672056e-01 1.06280051e-01 -1.42797559e-01
4.98804450e-01 6.97613120e-01 9.22337711e-01 -4.48171087e-02
7.43251026e-01 6.27157271e-01 -4.51559983e-02 -1.22224212e+00
-7.74951652e-02 -8.15008163e-01 -6.14402711e-01 -3.55602384e-01
7.06401706e-01 -1.05168962e+00 -4.38770384e-01 3.37153167e-01
-8.88832033e-01 -1.60021156e-01 2.28689797e-02 4.53248471e-01
-8.05294812e-01 3.66874069e-01 -2.51419425e-01 -5.47354043e-01
4.74384464e-02 -1.23965418e+00 9.97809529e-01 6.41296089e-01
-1.87067449e-01 -7.08940804e-01 1.41727045e-01 3.53638917e-01
4.76834774e-01 1.61610678e-01 3.30963964e-03 -3.47696133e-02
-9.93451297e-01 -2.95069963e-01 -3.23753953e-01 -2.07419824e-02
-6.61343187e-02 -3.52448784e-02 -8.97190571e-01 -2.23077133e-01
-3.25458348e-01 5.96579537e-02 3.24577123e-01 1.82239190e-01
3.15381706e-01 -9.30881947e-02 -4.65711117e-01 7.69742846e-01
1.74103558e+00 2.01527461e-01 7.75260746e-01 1.14695358e+00
5.61951518e-01 4.64702725e-01 8.87125552e-01 2.89215952e-01
3.72233003e-01 1.01287889e+00 5.94421506e-01 1.40759259e-01
7.34783942e-03 -4.00906265e-01 3.64733905e-01 3.36870164e-01
-6.30727559e-02 7.83780739e-02 -1.24418688e+00 7.19193757e-01
-1.94013512e+00 -8.88724804e-01 1.40950195e-02 2.36085033e+00
3.34378183e-01 9.17423591e-02 -1.22970998e-01 -1.52495295e-01
7.70537674e-01 1.71675012e-01 -2.87980229e-01 -3.69414628e-01
-1.12825267e-01 -3.85052681e-01 1.23636007e+00 1.50896206e-01
-9.55306292e-01 7.32258201e-01 6.69384050e+00 3.86901855e-01
-1.15966821e+00 -1.17537878e-01 -3.06471288e-02 -5.51480800e-02
2.29889169e-01 1.76153988e-01 -7.36933053e-01 3.17803293e-01
9.59049344e-01 -1.15817674e-01 5.98486125e-01 1.32322645e+00
3.09800077e-02 -4.64343011e-01 -1.05569398e+00 1.57981277e+00
2.54540592e-01 -1.30450809e+00 -6.48683488e-01 3.34728539e-01
4.68563616e-01 5.15255451e-01 1.89352036e-01 -1.85170874e-01
1.28067568e-01 -9.09916282e-01 9.50680971e-01 8.98693874e-02
8.06826293e-01 -7.28720307e-01 6.14259958e-01 2.95119286e-01
-1.15746510e+00 -2.96580583e-01 -4.83068317e-01 -3.00766945e-01
1.98145092e-01 -3.94642130e-02 -9.59279716e-01 1.71914667e-01
9.02712524e-01 6.83568895e-01 -6.79127574e-01 1.40383446e+00
-9.43270624e-02 -3.24886620e-01 -5.60492516e-01 -3.18354145e-02
3.14674586e-01 -1.03900582e-01 3.54847848e-01 9.99481201e-01
3.43817204e-01 -3.91487330e-02 -1.74658164e-01 4.00721759e-01
1.31092772e-01 -2.25254864e-01 -1.06151474e+00 1.82140350e-01
6.96105480e-01 1.28586245e+00 -8.83388042e-01 2.58720934e-01
-5.75314164e-01 1.16010141e+00 3.03656459e-01 5.33730499e-02
-5.82862437e-01 -4.30311710e-01 9.06236529e-01 3.34451258e-01
5.15370607e-01 -7.59566247e-01 -2.01058045e-01 -8.54532719e-01
6.14918321e-02 -8.14301372e-01 7.74313733e-02 -1.17964733e+00
-7.48459756e-01 2.54157126e-01 -1.88476026e-01 -1.52074194e+00
-3.41521949e-01 -7.61685431e-01 -4.05942351e-02 5.81090748e-01
-1.50320506e+00 -1.20249689e+00 -4.26581830e-01 5.47528803e-01
5.08694410e-01 -6.84064254e-02 6.78581595e-01 3.22495848e-01
-1.44779578e-01 2.81559318e-01 3.87372494e-01 1.28410876e-01
7.99913466e-01 -8.69372547e-01 3.01191241e-01 1.12905979e+00
2.02378511e-01 5.64530432e-01 7.56880403e-01 -3.53817999e-01
-1.84072399e+00 -8.01749110e-01 5.83546758e-01 -5.99852622e-01
4.98030007e-01 -2.46287435e-01 -3.89570326e-01 7.54040599e-01
2.09603488e-01 2.42755845e-01 3.00271064e-01 -6.29226863e-02
-2.96897024e-01 -2.65970290e-01 -1.37997007e+00 8.68947089e-01
8.68524015e-01 -8.04690123e-01 -6.54628396e-01 -2.13546112e-01
1.88685060e-01 -7.04695761e-01 -7.68129230e-01 3.64459008e-02
5.19887269e-01 -1.13487971e+00 1.10819364e+00 4.63864446e-01
1.04165860e-01 -7.10667431e-01 -2.57804364e-01 -1.00868356e+00
-6.02468371e-01 -4.59146023e-01 4.00024861e-01 1.15949392e+00
2.14854524e-01 -7.35977232e-01 6.81088626e-01 9.53933656e-01
8.30184743e-02 1.31833013e-02 -1.06371593e+00 -7.05740988e-01
-7.64720142e-01 -2.71316886e-01 2.59926200e-01 7.54964292e-01
3.48071486e-01 1.96602702e-01 -3.22010577e-01 4.68294621e-01
6.92710221e-01 -1.26019597e-01 7.76068747e-01 -9.39522862e-01
2.19776435e-03 -3.27606231e-01 -1.30859470e+00 -8.38195145e-01
-3.45959455e-01 -2.64094144e-01 3.34684134e-01 -1.49857354e+00
-2.43924871e-01 -2.52565295e-01 1.68269694e-01 2.94454455e-01
5.07258356e-01 5.64058840e-01 3.78181249e-01 4.70007211e-01
-8.75702083e-01 1.91306323e-01 5.66668570e-01 8.10570121e-02
-1.70041546e-01 -2.96014696e-01 -3.21417361e-01 8.01406801e-01
8.29419553e-01 -2.01126441e-01 -3.40812653e-01 -6.07155919e-01
2.59097040e-01 -7.48979971e-02 3.96974832e-01 -1.42799687e+00
8.07722867e-01 -3.52012925e-02 6.35549307e-01 -4.78677273e-01
5.22204936e-01 -1.22967088e+00 5.89237094e-01 1.94580838e-01
1.49869006e-02 3.71504426e-01 1.53946534e-01 4.86802965e-01
-1.96864098e-01 -2.77493447e-01 1.04843140e+00 -4.86554950e-01
-1.53005028e+00 -5.36196120e-02 -5.91808259e-01 -1.75145388e-01
1.40002012e+00 -8.04777741e-01 -4.01991397e-01 -2.07527608e-01
-1.59658119e-01 5.92471771e-02 1.46270049e+00 6.97707653e-01
8.15780997e-01 -1.17354214e+00 -7.75627345e-02 2.96587348e-01
4.58498269e-01 4.38715480e-02 -4.17935988e-03 5.98130286e-01
-1.28094697e+00 5.07381558e-01 -6.41211092e-01 -7.56388843e-01
-1.23613250e+00 5.34905016e-01 3.65508288e-01 4.18238610e-01
-8.34118903e-01 4.97919798e-01 -2.56689787e-01 -2.06814960e-01
9.55921784e-02 7.32818395e-02 9.82843116e-02 -1.66928425e-01
5.83120465e-01 4.05879289e-01 5.33215888e-02 -9.94567692e-01
-7.89704680e-01 6.64511383e-01 1.86459333e-01 -3.24801445e-01
1.12489462e+00 -5.38162708e-01 1.14614651e-01 2.02293932e-01
1.02490544e+00 -2.87129432e-01 -1.60460508e+00 -5.09998249e-03
2.16270447e-01 -8.02531719e-01 2.40452647e-01 -3.73331904e-01
-8.11098099e-01 6.02959275e-01 7.76637733e-01 -1.47974923e-01
9.91537273e-01 1.13294385e-02 2.24359661e-01 3.46594036e-01
9.99676287e-01 -1.35908294e+00 -1.42481923e-01 4.66076404e-01
5.20176113e-01 -1.12724340e+00 1.95702314e-01 -7.16225132e-02
-7.54293382e-01 1.05125833e+00 4.43094432e-01 -1.04378954e-01
7.10612759e-02 4.02843356e-01 1.17522329e-01 2.85581704e-02
-2.68760711e-01 -3.03925246e-01 -5.69731891e-02 1.18736601e+00
5.67713454e-02 -2.47259736e-01 5.83759174e-02 -1.18850000e-01
-1.05470166e-01 2.89856992e-03 5.83007276e-01 1.16510236e+00
-6.24987066e-01 -8.94769251e-01 -6.61260962e-01 -3.63008305e-02
-1.98801219e-01 1.94851667e-01 -2.66263604e-01 7.50578463e-01
5.83710819e-02 1.05215538e+00 -4.76998985e-02 -3.89499456e-01
4.74297911e-01 -2.76476264e-01 5.44010103e-01 -3.28485519e-01
-3.06524396e-01 -2.84577221e-01 1.60720825e-01 -9.57437515e-01
-4.36566472e-01 -7.62292683e-01 -9.84456956e-01 -3.86881918e-01
-1.33812219e-01 -2.93930948e-01 1.19804490e+00 5.88507593e-01
7.41100013e-01 1.28846183e-01 2.78345793e-01 -1.34881771e+00
-1.89432234e-01 -4.29706275e-01 -5.55404723e-01 9.17171389e-02
4.14173692e-01 -6.49006605e-01 -9.63356048e-02 1.34083867e-01] | [7.531856536865234, -1.9235386848449707] |
8b92b888-0a54-413e-b17a-a7e8c25f0778 | abaw-valence-arousal-estimation-expression | 2202.10659 | null | https://arxiv.org/abs/2202.10659v2 | https://arxiv.org/pdf/2202.10659v2.pdf | ABAW: Valence-Arousal Estimation, Expression Recognition, Action Unit Detection & Multi-Task Learning Challenges | This paper describes the third Affective Behavior Analysis in-the-wild (ABAW) Competition, held in conjunction with IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), 2022. The 3rd ABAW Competition is a continuation of the Competitions held at ICCV 2021, IEEE FG 2020 and IEEE CVPR 2017 Conferences, and aims at automatically analyzing affect. This year the Competition encompasses four Challenges: i) uni-task Valence-Arousal Estimation, ii) uni-task Expression Classification, iii) uni-task Action Unit Detection, and iv) Multi-Task-Learning. All the Challenges are based on a common benchmark database, Aff-Wild2, which is a large scale in-the-wild database and the first one to be annotated in terms of valence-arousal, expressions and action units. In this paper, we present the four Challenges, with the utilized Competition corpora, we outline the evaluation metrics and present the baseline systems along with their obtained results. | ['Dimitrios Kollias'] | 2022-02-22 | null | null | null | null | ['action-unit-detection'] | ['computer-vision'] | [ 2.89953709e-01 -1.09497100e-01 -2.26357188e-02 -7.52839029e-01
-9.37880576e-01 -4.31917846e-01 5.13965428e-01 2.03243196e-01
-5.83572567e-01 8.13627481e-01 2.96777397e-01 8.40658188e-01
4.84821469e-01 1.05098329e-01 -6.15658723e-02 -6.38728738e-01
-3.85499477e-01 1.83167025e-01 -3.46201003e-01 -2.81568736e-01
-1.34377033e-01 1.62510723e-01 -1.77544284e+00 7.48795927e-01
2.02660963e-01 1.71343315e+00 -6.68130577e-01 7.12025046e-01
2.27998450e-01 9.91517484e-01 -6.72946155e-01 -7.12074935e-01
-2.15567261e-01 -5.24718344e-01 -8.70933473e-01 -1.68973133e-01
3.40376675e-01 2.06412897e-01 3.80852968e-01 8.95010531e-01
7.50837743e-01 2.24391073e-01 8.40986669e-01 -1.84767878e+00
-1.63050279e-01 1.68765545e-01 -7.92584717e-01 1.57641366e-01
5.84629953e-01 -3.09191674e-01 1.15194356e+00 -9.24968600e-01
5.26165903e-01 1.29125619e+00 6.43628061e-01 8.96495223e-01
-8.92308772e-01 -5.46284139e-01 2.70345151e-01 5.74023843e-01
-1.15980268e+00 -3.57086837e-01 9.16405678e-01 -6.05196059e-01
1.36461830e+00 5.19445121e-01 9.44677830e-01 1.83481276e+00
1.12783544e-01 1.36441982e+00 1.51323950e+00 -3.64653111e-01
3.57481450e-01 1.13695815e-01 5.65862119e-01 2.63009489e-01
-6.08608186e-01 -3.44189972e-01 -7.87643373e-01 -3.47243726e-01
-1.05652548e-01 -8.16141725e-01 1.18872546e-01 -5.01827300e-02
-8.11058283e-01 8.20554852e-01 3.97227816e-02 3.99147183e-01
-6.89426541e-01 1.47586688e-01 1.35353386e+00 3.13577116e-01
8.48661065e-01 3.12401593e-01 -5.19530475e-01 -6.86156154e-01
-5.67118645e-01 1.87727302e-01 7.51329005e-01 6.40242159e-01
4.08799082e-01 1.16028219e-01 -6.59011602e-01 1.42154598e+00
2.02899292e-01 2.15797514e-01 6.56005025e-01 -1.00577295e+00
-8.11880305e-02 5.13101339e-01 2.77680214e-02 -7.12708116e-01
-6.97319984e-01 3.83310229e-01 -8.55334878e-01 2.49438006e-02
-2.21974775e-01 -4.08846647e-01 -6.89542234e-01 1.99048758e+00
2.48355806e-01 -8.58103111e-02 1.00066140e-01 7.47949719e-01
1.02800059e+00 6.65057778e-01 5.34306526e-01 -8.85654926e-01
1.52351487e+00 -1.11982679e+00 -1.38389301e+00 -3.75088990e-01
5.80998361e-01 -6.40277386e-01 8.75428915e-01 7.64799118e-01
-1.06621110e+00 -4.83562648e-01 -1.17241263e+00 -3.82937156e-02
-6.34358823e-01 2.86766171e-01 9.22197700e-01 7.75624216e-01
-1.08385110e+00 6.99007139e-02 -5.26214063e-01 -4.39910978e-01
5.50551057e-01 2.98725009e-01 -3.81268770e-01 6.48770988e-01
-1.52771211e+00 1.14559531e+00 -1.86139364e-02 6.69568703e-02
-7.42022693e-01 -4.96197850e-01 -9.02548611e-01 -4.09426540e-01
-1.40199572e-01 -6.93939477e-02 1.33116198e+00 -1.86652732e+00
-1.62704563e+00 1.66037989e+00 -1.03272766e-01 -2.53472954e-01
8.08733553e-02 -3.00414234e-01 -7.54876018e-01 -1.89389333e-01
-1.91039547e-01 8.23246181e-01 4.67658699e-01 -9.43165243e-01
-4.86341804e-01 -7.47568846e-01 -1.80964366e-01 1.90912083e-01
-3.81309807e-01 4.96091515e-01 -4.40362185e-01 -3.40943754e-01
-7.42263317e-01 -1.05463338e+00 -8.65125563e-03 -3.95053685e-01
-1.82770789e-01 -7.49601066e-01 8.13215673e-01 -3.09305996e-01
1.34632230e+00 -2.33955073e+00 1.34218842e-01 3.60744894e-02
2.57413145e-02 2.91185886e-01 -9.82719660e-02 2.43268669e-01
-4.47593272e-01 -2.41060823e-01 6.55799813e-04 -7.26099610e-01
2.06863701e-01 8.99008512e-02 1.13387024e-02 4.97355998e-01
5.28646894e-02 8.99684966e-01 -6.07722282e-01 -5.30343056e-01
1.24396585e-01 3.92325372e-01 1.07997684e-02 3.33644897e-01
1.57091781e-01 5.14534861e-02 -3.91598225e-01 8.47888172e-01
4.95787352e-01 2.97862887e-01 2.54507601e-01 -5.24831474e-01
-1.01281978e-01 -2.56129563e-01 -7.91619480e-01 1.65536714e+00
-5.01443863e-01 8.99575889e-01 2.79534429e-01 -1.16204703e+00
1.17339003e+00 5.79096437e-01 9.06212330e-01 -8.42076719e-01
4.90524322e-01 1.08745024e-01 -3.18250149e-01 -6.00600600e-01
3.42110932e-01 -1.91568851e-01 -7.01861024e-01 2.70390529e-02
5.01557767e-01 -5.83569556e-02 4.22958225e-01 5.43417558e-02
9.93504465e-01 2.36617610e-01 8.30961168e-01 -3.48622382e-01
8.24023187e-01 -4.36055630e-01 8.85732949e-01 1.02946952e-01
-1.04465556e+00 2.89872617e-01 7.91073918e-01 -5.87293506e-01
-5.54307580e-01 -8.15950513e-01 -5.50167859e-02 1.34647155e+00
-1.39151588e-01 -7.23323345e-01 -8.32812309e-01 -5.76831341e-01
-3.36741388e-01 5.49955666e-01 -1.41270185e+00 -2.66043931e-01
-2.36782506e-02 -9.67720091e-01 8.18216622e-01 7.35211730e-01
6.30714536e-01 -1.53680480e+00 -8.44808757e-01 -9.01163593e-02
-5.91694117e-01 -1.24391818e+00 -2.67700672e-01 4.57981080e-01
-1.53731942e-01 -9.13095593e-01 -3.78702223e-01 -7.30390012e-01
4.64789048e-02 -6.16090536e-01 1.38994694e+00 -5.05968034e-01
-7.67436504e-01 7.57545114e-01 -5.54524541e-01 -1.02693331e+00
6.73758052e-03 -4.20836389e-01 2.07666075e-03 5.11164844e-01
1.12480021e+00 -4.33014631e-01 -4.16683942e-01 2.94191726e-02
-3.64292830e-01 -2.82818437e-01 4.93530661e-01 5.92908382e-01
8.95600200e-01 -8.30805242e-01 7.81233549e-01 -8.09243381e-01
7.71192431e-01 -1.60878479e-01 -1.19139694e-01 4.15173888e-01
-3.31632406e-01 -6.19431615e-01 3.16481858e-01 -2.45152503e-01
-1.28625941e+00 5.11373699e-01 -4.51398253e-01 -3.93799663e-01
-2.18689457e-01 7.63932243e-02 -2.96190977e-01 3.19671392e-01
6.26531303e-01 -1.50654450e-01 -3.81261855e-01 5.21814078e-02
1.51147574e-01 9.50947225e-01 4.86320406e-01 -3.60381544e-01
-4.18651611e-01 3.50380808e-01 -6.83659539e-02 -8.13776851e-01
-1.23216355e+00 -8.77046645e-01 -7.51791000e-01 -8.67100418e-01
1.51505721e+00 -1.14850712e+00 -8.77243340e-01 8.26537907e-01
-1.14011323e+00 -2.73152202e-01 -2.35300466e-01 2.69877166e-01
-8.37643027e-01 -4.71145026e-02 -6.83186173e-01 -1.24848700e+00
-8.86829019e-01 -8.58076632e-01 1.37478077e+00 1.61011800e-01
-1.01412368e+00 -8.68714273e-01 7.56055951e-01 4.63553846e-01
1.21075906e-01 7.61417985e-01 5.73410451e-01 -6.92338824e-01
9.92076933e-01 -1.65503740e-01 9.83468965e-02 8.12341273e-01
-1.83024555e-01 1.23198792e-01 -1.55632305e+00 1.15420647e-01
-5.05398810e-02 -1.44871116e+00 8.08789909e-01 4.17151600e-01
1.27171040e+00 2.61012048e-01 -1.80083573e-01 2.71052510e-01
1.12970448e+00 4.89414543e-01 9.03639078e-01 -1.53613305e-02
2.50635952e-01 6.72632277e-01 1.04310668e+00 7.09019184e-01
3.39861661e-01 7.80911088e-01 3.92858207e-01 -1.64132521e-01
3.02325189e-01 4.36517328e-01 5.27024031e-01 8.97863507e-01
-4.50902045e-01 -1.46079361e-01 -6.63863838e-01 5.49366236e-01
-2.06848407e+00 -9.86126721e-01 -3.86283636e-01 1.75066257e+00
7.99664140e-01 -3.83527011e-01 4.76930320e-01 -5.99742346e-02
4.25839871e-01 3.06388319e-01 -4.10442352e-01 -1.44269526e+00
-4.01020885e-01 3.95427287e-01 -2.08351150e-01 3.34146321e-01
-1.64371896e+00 7.14570582e-01 6.21812296e+00 8.10618758e-01
-1.08219492e+00 4.33391452e-01 1.09891713e+00 -4.71588522e-01
2.34598503e-01 -6.77600920e-01 -5.03686249e-01 2.54779279e-01
1.15421808e+00 -9.42059308e-02 5.02533950e-02 1.11418843e+00
3.55990529e-02 -2.98685461e-01 -1.11867285e+00 1.48431158e+00
7.62825549e-01 -5.09485304e-01 -4.94313300e-01 -3.30538750e-01
6.25877380e-01 9.73348618e-02 -1.45661652e-01 9.59587693e-01
-1.61516204e-01 -9.85837579e-01 4.84608501e-01 3.34912717e-01
8.08319092e-01 -9.82178092e-01 8.62279594e-01 -3.41945112e-01
-1.02467024e+00 -4.49587554e-02 -1.63147271e-01 -3.81056704e-02
8.24137684e-03 5.18965840e-01 2.36116275e-01 1.82812259e-01
1.17622948e+00 8.93078923e-01 -4.02002126e-01 2.85155356e-01
-1.87447876e-01 4.57992256e-01 7.62050897e-02 -5.70160806e-01
1.66645631e-01 -1.70780525e-01 1.23400763e-01 1.74390554e+00
-1.67252421e-01 2.28359491e-01 4.49806266e-02 3.32547069e-01
-1.71348423e-01 3.93229961e-01 -3.77868801e-01 -4.69847769e-02
-1.60130505e-02 2.13535905e+00 -3.61315370e-01 -4.40127313e-01
-4.04867232e-01 1.41587675e+00 3.76017272e-01 1.95692796e-02
-1.29110396e+00 -3.19739133e-01 8.86202455e-01 -5.95545948e-01
-1.37513235e-01 4.79491293e-01 -1.28112854e-02 -9.04398382e-01
-8.18998069e-02 -7.77868450e-01 6.58145845e-01 -9.06323731e-01
-1.53656626e+00 9.49020565e-01 -3.63332219e-02 -9.68080580e-01
-1.88218996e-01 -8.20253253e-01 -6.11032367e-01 4.60250050e-01
-1.00282228e+00 -1.42133343e+00 -5.35898745e-01 7.46547461e-01
7.06789851e-01 -1.01167355e-02 1.39281452e+00 5.47773600e-01
-7.19354749e-01 6.81740344e-01 -3.23057532e-01 -5.79671301e-02
1.16664672e+00 -1.38053274e+00 -2.63220459e-01 1.64468750e-01
-9.37250927e-02 -1.31072477e-01 5.78449428e-01 -7.92447627e-02
-1.03548884e+00 -1.13036358e+00 8.40173304e-01 -4.39992309e-01
5.68823695e-01 -9.34243858e-01 -4.08095479e-01 8.41403306e-01
7.01510668e-01 5.35976663e-02 1.27201319e+00 4.42701012e-01
-2.43802905e-01 -1.92610189e-01 -1.04802310e+00 2.45090142e-01
6.56274557e-01 -3.31541449e-01 -2.50563890e-01 3.91453207e-01
6.43745139e-02 -1.85456157e-01 -9.99641895e-01 6.78299129e-01
8.86355281e-01 -1.07570779e+00 5.46221316e-01 -7.11450577e-01
6.12917542e-01 1.72711357e-01 -5.23508251e-01 -1.49992812e+00
-2.20135868e-01 -4.45643157e-01 -5.18807322e-02 1.47540915e+00
3.91561151e-01 6.51334785e-03 4.26913410e-01 5.10988474e-01
-1.98429793e-01 -1.11325967e+00 -1.20184433e+00 -4.21910167e-01
-9.96582657e-02 -5.97650707e-01 -1.72297254e-01 9.66114640e-01
4.56089675e-01 1.01841748e+00 -6.27333403e-01 -6.39663935e-01
3.19419533e-01 2.74315160e-02 4.98991191e-01 -1.04577529e+00
-1.55863076e-01 -4.68294472e-01 -7.90229797e-01 -1.72300249e-01
4.29726928e-01 -4.73966420e-01 2.89550811e-01 -1.17042065e+00
5.52877009e-01 5.87201834e-01 -4.89782959e-01 7.21948802e-01
-1.12398140e-01 5.66590250e-01 5.65172508e-02 -3.48100275e-01
-1.22036529e+00 8.07731450e-01 6.01643682e-01 -2.31817663e-01
9.00756940e-02 -2.81072795e-01 -4.77614760e-01 9.09960151e-01
7.38876283e-01 9.26149786e-02 -4.56882477e-01 2.28332162e-01
2.12591395e-01 -2.83882260e-01 8.41410644e-03 -8.01905930e-01
-2.40332454e-01 -8.93226489e-02 5.84445596e-01 -5.80260694e-01
1.06860840e+00 -4.88863498e-01 -3.70643228e-01 4.35118377e-02
-4.18343276e-01 7.41672376e-03 5.89445412e-01 1.01660654e-01
-5.31105936e-01 -5.63981831e-02 1.21924067e+00 2.24038698e-02
-1.16028166e+00 1.43761158e-01 -8.23601007e-01 2.35924780e-01
1.82802820e+00 1.06350727e-01 -2.50014454e-01 -4.06124234e-01
-1.10508597e+00 4.60229993e-01 -2.44900122e-01 6.39698625e-01
4.39462006e-01 -1.62502706e+00 -7.18135118e-01 -2.26527333e-01
6.37368381e-01 -9.62174654e-01 7.04682171e-01 1.25388086e+00
1.44459754e-01 2.00938240e-01 -5.99549115e-01 -4.83948469e-01
-2.01782942e+00 4.74558532e-01 4.26001370e-01 -5.00614822e-01
2.94544250e-01 1.11025608e+00 2.87638485e-01 -2.69906759e-01
1.33583263e-01 2.66447961e-01 -6.52987361e-01 7.36176610e-01
6.80993855e-01 2.67365187e-01 1.33426830e-01 -7.98190296e-01
-8.29114616e-01 2.65725046e-01 1.50212422e-01 -1.02342263e-01
1.22173047e+00 -1.36662886e-01 -4.60322887e-01 1.29452407e+00
1.51201904e+00 -5.14494777e-01 -6.51915193e-01 2.83650875e-01
1.25917450e-01 7.26512773e-03 7.79448636e-03 -1.07398796e+00
-9.65014458e-01 1.03168273e+00 8.55729640e-01 9.99582708e-02
1.54878020e+00 1.31789502e-02 5.46894848e-01 1.75146386e-01
7.97709450e-02 -2.07517910e+00 3.73554885e-01 5.35316765e-01
1.21374798e+00 -1.14867520e+00 -1.12570509e-01 -3.14103007e-01
-1.51152265e+00 9.29840565e-01 1.13176656e+00 -5.18244393e-02
7.29119003e-01 4.44816440e-01 4.37537909e-01 -4.01422292e-01
-1.27765214e+00 -3.52232903e-01 2.76193142e-01 5.95676243e-01
1.04715800e+00 1.57971919e-01 -5.19204795e-01 1.05413067e+00
9.85313356e-02 -1.31977359e-02 2.78229147e-01 6.10249639e-01
-6.76211268e-02 -8.27468395e-01 6.92201480e-02 5.61671734e-01
-7.06035733e-01 3.54357839e-01 -1.32307339e+00 6.62690997e-01
3.25241983e-01 9.88776863e-01 2.17796400e-01 -5.82523227e-01
6.91745996e-01 4.69981909e-01 6.47449136e-01 -3.55612427e-01
-8.84460270e-01 1.05007812e-01 7.72471488e-01 -1.06107509e+00
-9.22923803e-01 -8.53383899e-01 -1.17670679e+00 1.80862993e-01
1.51396468e-01 3.09512198e-01 7.66551495e-01 8.07683051e-01
1.71830431e-01 6.65584147e-01 6.80125475e-01 -7.05063224e-01
2.94041820e-03 -1.25623000e+00 -9.76886809e-01 7.92642415e-01
-9.50032249e-02 -7.49350011e-01 -1.45951286e-01 1.23024985e-01] | [13.570419311523438, 2.220968246459961] |
518acb38-bd27-4f8f-a432-b3cebe43bca1 | a-new-humanlike-facial-attractiveness | 1511.02465 | null | http://arxiv.org/abs/1511.02465v1 | http://arxiv.org/pdf/1511.02465v1.pdf | A new humanlike facial attractiveness predictor with cascaded fine-tuning deep learning model | This paper proposes a deep leaning method to address the challenging facial
attractiveness prediction problem. The method constructs a convolutional neural
network of facial beauty prediction using a new deep cascaded fine-turning
scheme with various face inputting channels, such as the original RGB face
image, the detail layer image, and the lighting layer image. With a carefully
designed CNN model of deep structure, large input size and small convolutional
kernels, we have achieved a high prediction correlation of 0.88. This result
convinces us that the problem of facial attractiveness prediction can be solved
by deep learning approach, and it also shows the important roles of the facial
smoothness, lightness, and color information that were involved in facial
beauty perception, which is consistent with the result of recent psychology
studies. Furthermore, we analyze the high-level features learnt by CNN through
visualization of its hidden layers, and some interesting phenomena were
observed. It is found that the contours and appearance of facial features,
especially eyes and moth, are the most significant facial attributes for facial
attractiveness prediction, which is also consistent with the visual perception
intuition of human. | ['Lingyu Liang', 'Lianwen Jin', 'Ziyong Feng', 'Jie Xu', 'Duorui Xie'] | 2015-11-08 | null | null | null | null | ['facial-beauty-prediction'] | ['computer-vision'] | [-3.46680582e-01 3.18923593e-01 -6.20324686e-02 -8.81357133e-01
5.46373785e-01 2.21146047e-01 1.93678662e-01 -3.52045149e-01
-1.52154461e-01 2.15758830e-01 2.03022659e-01 1.04590869e-02
-1.15975380e-01 -8.03065419e-01 -4.13992226e-01 -6.92980945e-01
-6.59655333e-02 -2.21216112e-01 -4.77082044e-01 -7.01687932e-01
3.73668611e-01 6.42596066e-01 -1.91630697e+00 2.86891371e-01
5.21173000e-01 1.72259200e+00 -4.70983386e-01 9.30841789e-02
-2.32585788e-01 5.59007823e-01 -1.33087486e-01 -1.04607046e+00
2.77415931e-01 -2.20031604e-01 -4.90223557e-01 1.19269872e-02
3.96320999e-01 -5.11692762e-01 -7.67779499e-02 1.04693806e+00
4.02213067e-01 -1.57209918e-01 4.75527585e-01 -1.53867185e+00
-1.17343652e+00 9.32333320e-02 -1.00713909e+00 -2.93919057e-01
1.95287347e-01 1.60405859e-02 9.41937327e-01 -9.82852578e-01
2.61727273e-01 1.62787580e+00 6.39342427e-01 6.86567962e-01
-6.58536911e-01 -1.27187395e+00 5.74864894e-02 1.75730750e-01
-1.31448257e+00 -3.73946249e-01 8.10258627e-01 -3.27601135e-01
2.46409312e-01 3.10168743e-01 1.15399802e+00 8.07932138e-01
2.83737212e-01 5.12225091e-01 1.23113179e+00 -2.09083661e-01
-2.32494354e-01 3.77816677e-01 -3.23756099e-01 1.09314084e+00
6.19536005e-02 3.51958573e-01 -3.77700835e-01 4.09368984e-02
9.63850200e-01 5.01197949e-02 1.49536043e-01 8.94594193e-02
-2.66296506e-01 8.67455721e-01 8.65812838e-01 3.24297473e-02
-2.28213578e-01 9.52424556e-02 2.32094303e-01 1.98727220e-01
5.06781220e-01 2.47839496e-01 -3.82662326e-01 2.59972394e-01
-3.70653093e-01 -1.66019917e-01 2.86060631e-01 5.43789327e-01
7.44626164e-01 2.14338869e-01 -7.75793195e-02 7.82543898e-01
6.30534530e-01 5.88466585e-01 3.39174569e-01 -9.83207285e-01
-2.60620952e-01 7.45070159e-01 -2.41048619e-01 -1.53597605e+00
-5.07538497e-01 -1.05753131e-01 -8.93989980e-01 5.97854197e-01
1.06940106e-01 8.04035291e-02 -7.22608864e-01 1.73149431e+00
4.37177509e-01 -3.19816530e-01 -3.10072929e-01 1.14259231e+00
1.33785760e+00 5.28891742e-01 4.57446605e-01 -1.01762727e-01
1.44726729e+00 -8.30175400e-01 -8.38203430e-01 1.27096251e-01
1.28134519e-01 -6.33674264e-01 1.34556484e+00 4.20128077e-01
-9.89970386e-01 -8.98845375e-01 -6.62666500e-01 -2.87114501e-01
-4.65545595e-01 5.21939158e-01 1.29732919e+00 8.27016890e-01
-1.09262633e+00 6.26837373e-01 5.86124277e-03 -2.78979391e-01
7.70074248e-01 5.15615225e-01 -6.49955809e-01 2.25646734e-01
-1.07703364e+00 9.33636487e-01 -7.33746216e-02 6.36971474e-01
-5.36123812e-01 -3.38164538e-01 -7.15021193e-01 3.73887986e-01
-2.84571975e-01 -2.74559408e-01 8.09903145e-01 -1.84444559e+00
-1.62204504e+00 9.68298495e-01 -9.91499946e-02 5.03533304e-01
-6.44083321e-02 -2.91927569e-02 -6.91013753e-01 -1.27226058e-02
-3.49436790e-01 9.62682366e-01 1.03481328e+00 -1.14366412e+00
-2.09643945e-01 -7.69082129e-01 -8.38470981e-02 -4.53128293e-02
-8.34301829e-01 1.67573109e-01 -1.65999457e-01 -2.98871219e-01
8.89455229e-02 -6.14994466e-01 -1.60213798e-01 5.89551568e-01
-1.26402751e-01 -3.31633359e-01 6.46472812e-01 -7.65474319e-01
1.08474386e+00 -2.36232257e+00 -2.96689898e-01 5.55828929e-01
4.45033997e-01 9.47337523e-02 -3.46054554e-01 -6.41731098e-02
-5.57531595e-01 2.33648926e-01 4.46712494e-01 -4.14655842e-02
8.82873014e-02 -7.84528404e-02 -4.42924201e-02 2.95478851e-01
4.65285510e-01 1.03228283e+00 -5.73887467e-01 -5.72497249e-01
1.51776493e-01 5.76963067e-01 -5.11327446e-01 2.88664848e-01
1.88751489e-01 3.28271329e-01 -4.01521713e-01 8.16245139e-01
1.15403867e+00 8.78992826e-02 -3.18975002e-02 -4.70667303e-01
-8.15448239e-02 -3.68347704e-01 -6.89351737e-01 1.04852092e+00
-2.75258690e-01 7.22319305e-01 4.79141995e-02 -4.85244751e-01
1.53056216e+00 3.12044863e-02 3.28992814e-01 -1.36089075e+00
5.75463772e-01 -4.08580378e-02 -3.03504877e-02 -8.36223423e-01
3.93244892e-01 -5.57410777e-01 3.33290964e-01 2.22794697e-01
-6.27095765e-03 1.53640032e-01 -4.75465447e-01 -1.44972280e-01
1.52032554e-01 6.77550361e-02 8.00589621e-02 -3.25567633e-01
4.12144095e-01 -7.45394349e-01 4.08970416e-01 -2.36147493e-01
-3.27489465e-01 3.33934069e-01 9.29996550e-01 -9.77069318e-01
-8.90454650e-01 -6.87399507e-01 -1.64621994e-01 1.43337238e+00
2.21469700e-01 -2.99118876e-01 -7.87918150e-01 -2.91279018e-01
3.48395482e-02 4.87265103e-02 -1.29635155e+00 -5.23223758e-01
4.47814837e-02 -6.09046161e-01 2.98323214e-01 5.15681505e-01
5.91824651e-01 -1.25509727e+00 -3.23377699e-01 -5.20366251e-01
2.25904241e-01 -8.07960331e-01 -1.72693457e-04 -3.76886070e-01
-7.19241858e-01 -1.06111801e+00 -5.61937809e-01 -7.87388027e-01
9.33646381e-01 6.85288291e-03 8.87441337e-01 5.64526021e-01
-4.82345819e-01 -1.10508293e-01 -8.54058340e-02 -6.08729720e-01
1.35793179e-01 -5.23386180e-01 6.79866523e-02 3.93301815e-01
5.11023521e-01 -1.61124483e-01 -8.04603934e-01 2.07632601e-01
-5.86977899e-01 3.72083336e-02 7.74129808e-01 5.96503913e-01
1.86953455e-01 -7.36730844e-02 2.93171763e-01 -4.52906877e-01
8.64088893e-01 -1.40411839e-01 -2.96735466e-01 2.26873279e-01
-6.31489635e-01 -2.47159287e-01 3.20220947e-01 -1.80516332e-01
-1.23091722e+00 -2.37241443e-02 -4.56568271e-01 -3.55647743e-01
-2.43492797e-01 -1.88825149e-02 -2.62288332e-01 -5.67526579e-01
3.36260051e-01 -1.82704687e-01 4.66696590e-01 -1.55814916e-01
2.99014181e-01 6.26335740e-01 -1.82838403e-02 -3.28431815e-01
4.22033221e-01 4.94123310e-01 3.27681094e-01 -8.09432447e-01
-7.86597013e-01 1.19155556e-01 -5.23678064e-01 -7.21223235e-01
8.01152229e-01 -7.53305674e-01 -1.71921384e+00 4.40837592e-01
-1.04636621e+00 8.77181739e-02 -4.62330841e-02 2.83224583e-01
-1.41786382e-01 1.83322161e-01 -6.47623420e-01 -9.88032222e-01
-4.06907380e-01 -9.79300559e-01 1.03048623e+00 7.50758111e-01
-1.07420720e-01 -7.86807775e-01 -2.84453720e-01 1.92362741e-01
5.94603240e-01 3.47152442e-01 1.07678866e+00 1.11246489e-01
4.19765450e-02 -1.01504259e-01 -7.73021996e-01 2.99873292e-01
2.75418814e-02 6.26598179e-01 -1.37249732e+00 9.16924998e-02
-1.82720631e-01 -5.56630492e-01 6.29615843e-01 5.43082237e-01
1.88466907e+00 -2.80375093e-01 3.62163596e-02 7.15802789e-01
1.22586334e+00 3.28580067e-02 1.02703941e+00 1.10142246e-01
4.67283309e-01 1.12720013e+00 5.70961297e-01 6.61493957e-01
4.55031037e-01 2.87089348e-01 8.81314814e-01 -1.05809808e+00
2.08683461e-01 -3.61468822e-01 8.18829387e-02 3.67275327e-01
-6.38477206e-01 4.14813995e-01 -4.08348054e-01 -1.04486555e-01
-1.31220019e+00 -9.15253937e-01 -1.68184921e-01 1.73004973e+00
4.39447552e-01 -2.98365563e-01 2.36263901e-01 -5.82859740e-02
6.16695583e-01 -9.57933962e-02 -3.59219521e-01 -1.08742499e+00
-4.63634580e-02 5.35004973e-01 6.55546263e-02 1.61851108e-01
-9.13663805e-01 1.18607557e+00 7.31513596e+00 5.48741341e-01
-1.45822203e+00 -3.76149297e-01 1.34423530e+00 -2.62443610e-02
-4.19068336e-01 -3.15218806e-01 -4.21239495e-01 4.02659386e-01
5.95707536e-01 1.33132920e-01 2.95018375e-01 1.00244701e+00
2.52118647e-01 1.40504325e-02 -7.09853053e-01 1.18231344e+00
6.47270977e-02 -7.48919725e-01 3.05041105e-01 2.11792424e-01
3.50911796e-01 -8.11529458e-01 7.74491668e-01 2.53976852e-01
-7.51209781e-02 -1.68114662e+00 3.19509208e-01 6.59143448e-01
9.70871091e-01 -6.86066389e-01 7.88204432e-01 -2.73410708e-01
-1.05366683e+00 -4.47183907e-01 -6.83862567e-01 -5.07201672e-01
-3.06342870e-01 3.52316916e-01 -4.11609650e-01 4.94121984e-02
1.22795665e+00 5.82967401e-01 -6.89077854e-01 6.98088527e-01
-1.51707411e-01 1.24940149e-01 3.02006193e-02 -2.75808930e-01
1.59661442e-01 -4.37248051e-01 -5.10761559e-01 9.16731715e-01
3.38240266e-01 4.34808344e-01 -4.60905135e-01 9.21878934e-01
-8.12067464e-02 5.71707189e-01 -6.17351830e-01 -2.76996428e-03
-1.18608423e-01 1.74024057e+00 -3.39383155e-01 1.10872842e-01
-3.42563093e-01 6.65167153e-01 3.01219046e-01 2.16266945e-01
-9.21388030e-01 -2.67026424e-01 8.57046306e-01 1.39122814e-01
-4.82745729e-02 2.32651681e-01 -4.57137525e-01 -6.33992255e-01
-1.31379664e-01 -5.55708170e-01 7.12922290e-02 -1.09046340e+00
-1.08702767e+00 6.07993603e-01 -5.05966783e-01 -8.15701485e-01
6.42847002e-01 -1.02071679e+00 -9.21208560e-01 7.65005171e-01
-1.70353901e+00 -1.32148230e+00 -7.57892013e-01 7.11498201e-01
-1.53846502e-01 -1.83208376e-01 9.78610039e-01 2.66542614e-01
-6.08312547e-01 7.18451023e-01 -4.67684984e-01 3.18379313e-01
5.78072131e-01 -7.47685194e-01 -1.72647554e-02 7.56397918e-02
-3.71535659e-01 5.48437178e-01 2.48851687e-01 -1.62671909e-01
-1.09206176e+00 -6.18962049e-01 8.48173797e-01 -3.82118821e-02
3.73166502e-01 -2.45682120e-01 -5.60731649e-01 1.71987131e-01
3.07611406e-01 -5.93597628e-02 1.01306999e+00 2.93730170e-01
-4.06516671e-01 -5.50951421e-01 -1.16627574e+00 5.16253293e-01
8.87105167e-01 -1.69573650e-01 -5.36810160e-02 7.11104646e-02
3.52001429e-01 -1.69239081e-02 -8.58547568e-01 4.51453209e-01
1.18082845e+00 -1.38531244e+00 7.80856490e-01 -9.66021001e-01
1.09848344e+00 3.52697909e-01 9.70112085e-02 -1.03490710e+00
-7.15772331e-01 -2.01041717e-02 3.56001556e-01 9.67578113e-01
3.82676691e-01 -5.01076519e-01 7.98160970e-01 8.51569831e-01
3.24705034e-01 -1.10590732e+00 -6.06396317e-01 -1.44010663e-01
-7.99061060e-02 1.09399371e-02 9.64673162e-01 9.03932571e-01
-3.94027047e-02 1.74222797e-01 -5.58581769e-01 -1.75487518e-01
3.34849507e-01 1.38071552e-01 7.16573775e-01 -1.60934842e+00
3.95520389e-01 -6.11234248e-01 -4.89298344e-01 -5.15718877e-01
3.55447799e-01 -4.57272500e-01 -4.57336932e-01 -1.16042078e+00
3.44392031e-01 -4.23195183e-01 -4.04224455e-01 7.02495933e-01
-5.65551855e-02 4.33599442e-01 1.66833729e-01 -2.29784906e-01
-2.32900217e-01 9.25181627e-01 1.93838000e+00 -1.04972035e-01
1.11621752e-01 -1.02910310e-01 -1.16288459e+00 8.61613810e-01
7.18152583e-01 1.94633380e-01 -2.37008855e-01 -4.67571974e-01
3.75022590e-01 -2.36326814e-01 3.86247605e-01 -6.08295321e-01
-1.26870424e-01 -4.10848677e-01 1.16147423e+00 1.26094138e-02
7.47198701e-01 -9.65344310e-01 -1.75880164e-01 5.26928067e-01
-3.37882727e-01 2.25736778e-02 3.46614301e-01 3.52865830e-02
-1.74793899e-01 2.51211882e-01 9.70327973e-01 -1.75626755e-01
-1.03856146e+00 6.98568940e-01 1.68334365e-01 -6.09970570e-01
1.06813788e+00 -4.19491798e-01 -1.87845483e-01 -6.77289188e-01
-7.89038301e-01 -1.11257331e-02 1.99906185e-01 6.34480774e-01
1.01447928e+00 -1.60561967e+00 -5.49134493e-01 7.38442779e-01
1.43844709e-01 -5.64119339e-01 2.32594192e-01 7.03313470e-01
-4.54090089e-01 -5.41949384e-02 -1.07451117e+00 -3.38310093e-01
-1.37778592e+00 3.96102905e-01 4.60746348e-01 5.43696582e-01
-1.36733562e-01 1.13168681e+00 5.56573629e-01 -3.53688776e-01
1.47218212e-01 -6.90333843e-02 -7.60239840e-01 2.43201163e-02
5.60736835e-01 2.55282551e-01 -1.44935369e-01 -8.30579877e-01
-2.25535810e-01 1.02062166e+00 1.94935754e-01 4.00013775e-01
1.33778667e+00 3.84374335e-02 -5.16703963e-01 -8.11788216e-02
1.17423034e+00 -1.58951119e-01 -9.39979494e-01 3.21035415e-01
-4.65571165e-01 -8.28718841e-01 4.66687307e-02 -6.62396133e-01
-1.69673073e+00 1.31366694e+00 9.75311399e-01 -5.87806385e-03
1.43827534e+00 -9.05786455e-02 5.78492284e-01 1.05546236e-01
1.31899759e-01 -1.30315018e+00 4.64337796e-01 2.41232172e-01
1.04647100e+00 -1.51966083e+00 -1.31157665e-02 -4.65355814e-01
-9.21253920e-01 1.39307487e+00 1.31583786e+00 8.65096301e-02
9.71812904e-01 -2.40756590e-02 1.74613968e-01 -4.95252937e-01
-5.11012256e-01 -1.07793860e-01 3.90267611e-01 4.82380301e-01
7.28431106e-01 2.63439357e-01 -1.04750462e-01 8.42692614e-01
-4.32310581e-01 4.47363034e-02 -1.30271120e-02 1.39532074e-01
-5.76520681e-01 -6.67360187e-01 -2.53521383e-01 3.27400386e-01
-2.88367599e-01 7.45104253e-02 -7.23257244e-01 8.33325982e-01
5.36349475e-01 5.61786115e-01 3.06545675e-01 -7.77719975e-01
2.79792190e-01 -3.11094403e-01 5.16484618e-01 -1.68851033e-01
-4.47934806e-01 -2.60763884e-01 -3.85482848e-01 -9.27216649e-01
-3.37004423e-01 6.96158782e-02 -1.01488054e+00 -7.19266415e-01
-2.69920439e-01 2.98154671e-02 7.96695352e-01 6.38346851e-01
1.92003369e-01 1.16556190e-01 9.96824026e-01 -7.87482858e-01
-7.67168850e-02 -8.91700387e-01 -8.48732591e-01 6.40264809e-01
1.06787898e-01 -8.42925310e-01 -2.55986780e-01 -1.93645790e-01] | [13.402387619018555, 1.1830917596817017] |
e19bb30d-7407-43c3-9479-0dc11b535333 | perimeter-control-autonomous-vehicle-and | 2307.02156 | null | https://arxiv.org/abs/2307.02156v1 | https://arxiv.org/pdf/2307.02156v1.pdf | Perimeter control, autonomous vehicle, and urban spatial structure | This paper examines the effects of hypercongestion mitigation by perimeter control and the introduction of autonomous vehicles on the spatial structures of cities. By incorporating a bathtub model, we develop a land use model where hypercongestion occurs in the downtown area and interacts with land use. We show that hypercongestion mitigation by perimeter control decreases the commuting cost and results in a less dense urban spatial structure. Furthermore, we reveal that the impact of autonomous vehicles depends on the perimeter control implementation. Introduction of autonomous vehicles may increase the commuting cost in the presence of hypercongestion and cause an increase in downtown population in the long-run. This result contradicts that of the standard bottleneck model. When perimeter control is implemented, the introduction of autonomous vehicles decreases the commuting cost and results in a less dense urban spatial structure. These results show that hypercongestion is a key factor that can change urban spatial structures. | ['Yuki Takayama', 'Takao Dantsuji'] | 2023-07-05 | null | null | null | null | ['autonomous-vehicles'] | ['computer-vision'] | [-7.07685351e-01 3.89840066e-01 -3.53453428e-01 3.38905066e-01
3.55251104e-01 -5.23854792e-01 4.74587321e-01 1.17239378e-01
-7.35264957e-01 1.05836892e+00 1.53973460e-01 -9.84936774e-01
-3.01627070e-01 -1.60253823e+00 -6.01498485e-01 -7.03216791e-01
-1.86289832e-01 6.00036867e-02 5.37554145e-01 -6.23137414e-01
-9.92067158e-02 6.10064864e-01 -1.30636799e+00 -7.66746581e-01
9.08489287e-01 -3.09931636e-02 1.90837204e-01 4.82194811e-01
-2.38837764e-01 -3.07587278e-03 -2.77271450e-01 9.65265483e-02
5.81158936e-01 6.00425415e-02 -3.48460793e-01 1.07354932e-01
-1.92624614e-01 -4.38545316e-01 -6.07574821e-01 5.86882949e-01
3.44721645e-01 1.87148109e-01 7.55708158e-01 -1.74484742e+00
1.99954733e-01 7.06324756e-01 -5.07532001e-01 2.16170028e-01
-2.93217957e-01 4.18619245e-01 5.65949798e-01 -1.09275602e-01
4.55103636e-01 1.30677712e+00 2.38943532e-01 -1.77537411e-01
-1.45323658e+00 -6.14544749e-01 2.19173804e-01 -1.92453012e-01
-1.65176082e+00 -4.58976626e-01 2.19134733e-01 -1.89002559e-01
8.19581449e-01 2.87913859e-01 1.04633117e+00 1.45285293e-01
3.52149785e-01 2.83274055e-01 3.49301368e-01 -4.30763900e-01
4.09411520e-01 3.66977192e-02 6.70460090e-02 3.29002202e-01
1.36604536e+00 2.90616423e-01 4.80033129e-01 1.14451893e-01
7.71667719e-01 -9.81361642e-02 -4.50068563e-02 -5.07886410e-01
-6.66229486e-01 7.69694328e-01 4.91571039e-01 4.98256266e-01
-6.64603710e-01 6.36997819e-01 1.53631061e-01 6.04890287e-01
3.31894785e-01 4.20477271e-01 -2.68149763e-01 7.52582867e-03
-6.96613133e-01 4.43928063e-01 5.64351559e-01 8.73963892e-01
9.45900559e-01 1.99941415e-02 1.92849457e-01 1.93359971e-01
4.82329011e-01 1.32782865e+00 -2.96571553e-01 -1.39665043e+00
7.54494071e-01 7.03863978e-01 4.70774531e-01 -8.00535977e-01
-7.03156173e-01 -3.25168729e-01 -6.80179238e-01 4.42066401e-01
5.78603625e-01 -7.14620113e-01 -5.57960987e-01 1.60725927e+00
2.89080054e-01 -1.10252514e-01 1.20649837e-01 5.35281718e-01
5.10784914e-04 6.33192182e-01 3.72784525e-01 -3.84098738e-01
7.31996655e-01 -4.90069985e-01 -8.39201868e-01 1.24554195e-01
1.19022274e+00 -5.39975286e-01 7.98251569e-01 -5.36238909e-01
-1.03886247e+00 -9.03971568e-02 -7.34909773e-01 5.79858422e-01
-6.96899235e-01 -5.24073720e-01 1.02145284e-01 1.06092632e+00
-1.13281465e+00 2.49356553e-01 -9.37994361e-01 -8.30553472e-01
1.98887646e-01 3.05300742e-01 6.59909891e-03 -1.87079608e-01
-1.00951564e+00 7.95682371e-01 7.52425846e-03 -4.25572127e-01
-1.09701842e-01 -6.74015462e-01 -8.41049016e-01 3.91394258e-01
2.84557849e-01 -7.16643929e-01 6.73732579e-01 -7.58374929e-01
-8.76254976e-01 -5.60773537e-02 1.02149591e-01 -2.65107721e-01
8.10036421e-01 1.73320889e-01 -3.06689769e-01 -1.13741569e-01
7.09595859e-01 6.56005323e-01 4.11143936e-02 -1.35040903e+00
-9.21107948e-01 -4.31307077e-01 1.36209860e-01 3.51595551e-01
-1.37528870e-02 -5.98340869e-01 -5.98700205e-03 1.71183180e-02
-2.42709830e-01 -1.06202865e+00 -6.78663433e-01 -4.14943576e-01
-5.77882707e-01 -1.04805999e-01 8.17872405e-01 1.20533541e-01
1.44264174e+00 -2.25795054e+00 -5.30419409e-01 6.80114985e-01
2.34195381e-01 1.70779049e-01 -2.41814479e-01 6.79718673e-01
3.49143237e-01 6.55279577e-01 -1.17055729e-01 -6.49559451e-03
8.17558095e-02 5.59587240e-01 4.81836945e-01 4.36014235e-01
-1.32635385e-01 5.22222400e-01 -8.84276688e-01 -3.43006819e-01
6.70229435e-01 1.20118551e-01 -4.75728035e-01 -4.39814329e-01
2.12633178e-01 -5.33095524e-02 -6.94517374e-01 1.07375279e-01
1.09691954e+00 2.39904225e-01 2.69597590e-01 8.53992939e-01
-1.18347037e+00 1.79560736e-01 -8.34314346e-01 4.06897753e-01
-4.41685408e-01 9.28986728e-01 1.87618703e-01 -7.00931251e-01
4.19718683e-01 2.27240354e-01 5.32036006e-01 -1.09599733e+00
1.68996736e-01 2.61251211e-01 3.13884884e-01 -7.57891238e-01
4.65256631e-01 1.04341798e-01 1.58731677e-02 5.40003061e-01
-8.89511168e-01 -6.97092637e-02 4.90120888e-01 3.54763806e-01
1.38217866e+00 -6.19324565e-01 1.07358962e-01 -7.34722912e-01
-2.76480950e-02 4.67525899e-01 4.38092768e-01 7.06752479e-01
-5.79898834e-01 -3.93402308e-01 8.01133335e-01 -2.16542855e-01
-1.36629999e+00 -1.31109321e+00 -1.19246393e-01 7.36486614e-01
4.45350647e-01 1.47866039e-02 -6.05226994e-01 -3.23445946e-01
6.46674514e-01 1.15749669e+00 -3.97515446e-01 6.17919415e-02
-6.17007136e-01 -6.82130873e-01 4.29873556e-01 -5.22967987e-02
9.86439943e-01 -4.11290973e-01 -7.61682510e-01 3.11338007e-01
-3.00035834e-01 -7.85912216e-01 -2.10031494e-01 -2.93150693e-01
-9.06307399e-01 -1.09383380e+00 -6.32237732e-01 -2.74736226e-01
7.65420079e-01 1.10895324e+00 6.58020377e-01 3.53321373e-01
2.19483197e-01 5.00894666e-01 -2.96675980e-01 -4.31770831e-01
-3.93317103e-01 7.25423694e-01 -1.20885380e-01 -4.32283133e-01
2.42715061e-01 -6.95367873e-01 -7.73838341e-01 3.72180432e-01
-5.45880079e-01 -1.61506087e-01 3.28776538e-01 -1.64313182e-01
4.69379574e-02 5.53642273e-01 8.86659682e-01 -4.17550355e-01
3.99650335e-01 -9.97962296e-01 -8.23504686e-01 -2.35984966e-01
-1.01249576e+00 3.16931680e-02 2.36299098e-01 7.10497797e-02
-8.09080720e-01 -1.11656487e-01 2.53208339e-01 3.55932415e-01
-3.45745385e-01 3.25811714e-01 -3.92351985e-01 2.57579386e-01
2.98541605e-01 -2.82751381e-01 1.89395905e-01 -1.53378472e-01
6.01375699e-01 3.91556621e-01 -3.82351935e-01 6.26990125e-02
9.98845756e-01 7.06145108e-01 4.74134713e-01 -1.53279495e+00
5.24678528e-01 -4.78524089e-01 -7.79485703e-01 -5.06506860e-01
7.43516505e-01 -8.71659815e-01 -7.31486797e-01 2.81975251e-02
-8.72216642e-01 -9.41490889e-01 -2.80230165e-01 5.97625017e-01
-5.21349370e-01 -5.45792729e-02 8.41780975e-02 -9.19316530e-01
4.50556368e-01 -7.80836523e-01 1.25487939e-01 1.89936772e-01
-5.72148710e-02 -1.13953459e+00 1.57917142e-01 -1.75946146e-01
7.03799009e-01 2.73332238e-01 1.08131766e+00 1.64522305e-01
-1.06321979e+00 1.31509379e-01 -4.49354559e-01 -4.10486728e-01
2.87967980e-01 2.21189290e-01 -1.36548743e-01 -3.05699948e-02
-8.28853667e-01 7.96558499e-01 7.03946114e-01 9.64107990e-01
-2.99397595e-02 -5.04482627e-01 -8.38225603e-01 9.08516124e-02
1.65513623e+00 2.88125306e-01 9.18946445e-01 9.06104445e-01
1.97552666e-01 9.07758236e-01 4.64248478e-01 5.48339307e-01
7.89865315e-01 4.28489923e-01 6.79236472e-01 -4.71865594e-01
-9.64733735e-02 -6.81929514e-02 2.84101725e-01 8.40983838e-02
-4.15448904e-01 -5.54504693e-01 -1.02817070e+00 8.95422876e-01
-2.05046272e+00 -1.13267577e+00 -9.27180648e-01 2.24207592e+00
6.59035845e-03 1.43462166e-01 6.55653834e-01 8.76685083e-02
7.46135771e-01 -7.71893710e-02 -1.53512701e-01 -3.66453260e-01
-5.08536305e-03 -4.68603164e-01 1.61278629e+00 1.08086097e+00
-5.39740920e-01 1.08376193e+00 6.75017071e+00 2.42896259e-01
-7.64036417e-01 -1.22804279e-02 3.73697519e-01 -6.92329407e-02
-5.88061035e-01 3.72813232e-02 -4.32098955e-01 3.79878581e-01
1.05975831e+00 -6.59111679e-01 3.81404050e-02 2.86329299e-01
1.61632001e+00 -7.00130999e-01 -6.78493828e-02 -1.42774843e-02
-8.17845583e-01 -1.16112924e+00 -2.25852385e-01 8.02042484e-01
6.69399619e-01 1.92128226e-01 -1.80493876e-01 9.28077698e-02
9.56749558e-01 -4.88511443e-01 3.23743552e-01 2.64713168e-01
6.78599596e-01 -1.10313189e+00 7.21557558e-01 3.26061368e-01
-1.45913565e+00 -1.93585843e-01 -1.12050883e-01 -4.74933088e-01
5.24407923e-01 4.37660307e-01 -6.31731689e-01 2.02736765e-01
3.35054964e-01 1.33042499e-01 -2.94496298e-01 1.38365078e+00
-2.70758301e-01 7.85964131e-01 -5.30274153e-01 2.77709197e-02
6.00970507e-01 -4.01217997e-01 6.91494465e-01 7.93626904e-01
2.53236711e-01 3.50449055e-01 -2.82771718e-02 8.07643831e-01
3.26811671e-01 4.92400452e-02 -1.26782703e+00 3.05148631e-01
8.18563640e-01 5.38876951e-01 -8.42555165e-01 -1.53003156e-01
-4.32739913e-01 3.77357066e-01 -4.25024003e-01 1.01166570e+00
-6.25926673e-01 -6.95608079e-01 1.13074720e+00 1.00656378e+00
2.19404280e-01 -5.79027474e-01 -6.45683706e-01 -5.02051711e-01
-5.64508975e-01 1.77088708e-01 -2.82440811e-01 -2.63422310e-01
-2.20334828e-01 -2.25117385e-01 2.54284114e-01 -5.45126081e-01
1.44097403e-01 2.13234037e-01 -1.04824531e+00 5.52743137e-01
-1.68950498e+00 -5.81808746e-01 -1.86157338e-02 2.60189354e-01
2.70519286e-01 1.72465608e-01 2.12701082e-01 4.44433242e-01
-9.09550548e-01 1.15261957e-01 7.52374768e-01 -8.30768347e-02
1.80342317e-01 -8.05264235e-01 2.24565193e-01 8.04705322e-01
-9.51012254e-01 1.85553893e-01 7.91880190e-01 -9.89662588e-01
-8.75422001e-01 -1.43066704e+00 1.23696125e+00 1.65015999e-02
4.37295318e-01 -1.74631059e-01 -3.65068674e-01 4.97119725e-01
3.38958174e-01 -6.66744053e-01 2.61205703e-01 6.47917390e-02
3.65100622e-01 -1.32005528e-01 -1.24809098e+00 1.12649643e+00
8.98544133e-01 2.25754872e-01 1.75347522e-01 -2.06393935e-02
9.60105479e-01 9.83047426e-01 -3.59213471e-01 -1.36089861e-01
5.39424598e-01 -6.34444058e-01 6.37450159e-01 -2.42382631e-01
-3.53394561e-02 2.16393899e-02 -1.96767956e-01 -1.44547641e+00
-8.42558801e-01 -4.53923792e-01 9.74273324e-01 9.66666460e-01
8.99797857e-01 -9.99728620e-01 8.61209035e-01 8.48756015e-01
-5.09221666e-02 9.30755341e-05 -1.06024241e+00 -1.16777420e+00
6.30888581e-01 -1.77328706e-01 7.41830826e-01 6.48287356e-01
1.87247887e-01 2.29178399e-01 4.35921788e-01 3.76123011e-01
7.35438108e-01 -6.75390363e-01 1.01412487e+00 -1.24069142e+00
9.43128049e-01 -7.17423975e-01 -2.61450082e-01 -7.20372915e-01
1.12550326e-01 -3.61516565e-01 -1.55619606e-01 -1.92536056e+00
-2.80391574e-01 -8.53115618e-01 3.18704784e-01 1.55097887e-01
4.18905318e-01 -2.87082255e-01 2.87836969e-01 1.54767483e-01
-2.53280580e-01 6.35709226e-01 1.06855690e+00 -7.74277821e-02
-1.09616363e+00 3.64673793e-01 -3.25666457e-01 5.24656296e-01
1.25906122e+00 -6.94008768e-01 -6.38476610e-01 -1.27977431e-01
1.49448574e-01 2.25275204e-01 1.54081956e-01 -8.05057168e-01
-6.70434311e-02 -5.67569196e-01 -3.66299838e-01 -5.78901231e-01
-8.09975713e-02 -1.20694804e+00 1.07125804e-01 1.20870113e+00
1.87918529e-01 3.49667907e-01 3.50052774e-01 5.96440852e-01
4.19920683e-01 2.35059574e-01 8.26260984e-01 1.43163234e-01
-8.52666572e-02 1.02211982e-01 -1.74249864e+00 -3.40657920e-01
1.43598986e+00 -4.69736814e-01 -3.21011305e-01 -5.16564012e-01
-6.39764190e-01 8.42790842e-01 6.65945292e-01 -6.52696490e-02
1.59008190e-01 -1.16827059e+00 -6.74193084e-01 4.50286157e-02
-3.43837321e-01 -5.36513388e-01 -8.15516263e-02 8.87000442e-01
-6.85085654e-01 8.76195192e-01 -3.65541071e-01 -5.05023710e-02
-1.25263536e+00 2.92005569e-01 5.56174636e-01 8.45123082e-03
-3.78001481e-01 1.29819605e-02 2.45847970e-01 -1.97319710e-03
-9.96926129e-02 -1.86521351e-01 -2.45489836e-01 -1.80194527e-02
-1.88060939e-01 1.27473414e+00 -6.74793959e-01 -7.03170776e-01
-1.42693400e-01 3.20104539e-01 4.39597011e-01 -5.81390977e-01
1.03288531e+00 -1.26043010e+00 1.02177309e-02 2.89944977e-01
6.35547817e-01 -1.03571832e-01 -9.91397798e-01 2.06668794e-01
-1.12990290e-01 -4.80266929e-01 3.11699659e-01 -3.95218462e-01
-1.03165662e+00 4.05858636e-01 5.07167578e-01 7.47848153e-01
7.18463063e-01 -2.25953832e-01 5.25208950e-01 3.79943430e-01
3.41831028e-01 -1.33470964e+00 -2.58062184e-01 4.73397613e-01
3.71003926e-01 -1.00699365e+00 -1.15103200e-01 -8.12876523e-01
-1.86944485e-01 6.80791378e-01 7.46409237e-01 -1.21920288e-01
1.28754008e+00 -1.01838831e-03 -2.69269913e-01 -2.72673547e-01
-4.84856933e-01 -1.07155049e+00 -8.54229808e-01 7.90471375e-01
2.59378161e-02 8.46590281e-01 -9.48543608e-01 1.46258259e-02
-1.22668175e-02 -2.69802988e-01 1.13965893e+00 6.62014604e-01
-1.10949934e+00 -1.18064761e+00 -5.13327658e-01 4.10269946e-01
7.29472265e-02 4.47115339e-02 -2.41977602e-01 1.68849230e+00
5.00983179e-01 1.40596354e+00 6.33025706e-01 4.73948717e-02
4.40420508e-01 1.97200719e-02 5.44875823e-02 -2.99596637e-01
-5.03988378e-02 1.09650455e-01 3.91724288e-01 -1.79776594e-01
-1.39782727e-01 -8.82143080e-01 -1.31958282e+00 -1.38596058e+00
-4.31922168e-01 1.62856728e-01 5.40665030e-01 9.40413058e-01
5.99712849e-01 3.97785425e-01 7.92040765e-01 -2.22569883e-01
3.29390556e-01 -6.68872893e-01 -1.00634205e+00 -2.49378204e-01
3.89253855e-01 -6.30508661e-01 -3.75617415e-01 -3.72320980e-01] | [5.639978885650635, 1.569669485092163] |
519eec58-4862-4f4c-ab80-51da13d90700 | becoming-self-instruct-introducing-early | 2307.03692 | null | https://arxiv.org/abs/2307.03692v1 | https://arxiv.org/pdf/2307.03692v1.pdf | Becoming self-instruct: introducing early stopping criteria for minimal instruct tuning | In this paper, we introduce the Instruction Following Score (IFS), a metric that detects language models' ability to follow instructions. The metric has a dual purpose. First, IFS can be used to distinguish between base and instruct models. We benchmark publicly available base and instruct models, and show that the ratio of well formatted responses to partial and full sentences can be an effective measure between those two model classes. Secondly, the metric can be used as an early stopping criteria for instruct tuning. We compute IFS for Supervised Fine-Tuning (SFT) of 7B and 13B LLaMA models, showing that models learn to follow instructions relatively early in the training process, and the further finetuning can result in changes in the underlying base model semantics. As an example of semantics change we show the objectivity of model predictions, as defined by an auxiliary metric ObjecQA. We show that in this particular case, semantic changes are the steepest when the IFS tends to plateau. We hope that decomposing instruct tuning into IFS and semantic factors starts a new trend in better controllable instruct tuning and opens possibilities for designing minimal instruct interfaces querying foundation models. | ['Melisa Russak', 'Parikshith Kulkarni', 'Kiran Kamble', 'Brock Imel', 'Kirk Goddard', 'Manhal Daaboul', 'Waseem AlShikh'] | 2023-07-05 | null | null | null | null | ['instruction-following'] | ['natural-language-processing'] | [ 1.25188604e-01 1.94691122e-01 -3.69050562e-01 -8.95552516e-01
-7.03549385e-01 -8.76542568e-01 6.67919397e-01 2.71507770e-01
-4.79445666e-01 5.62209368e-01 4.64055061e-01 -6.78779542e-01
-2.26553395e-01 -7.09939718e-01 -8.34726989e-01 -2.59549528e-01
3.26488227e-01 6.16219521e-01 4.95164037e-01 -7.24193335e-01
5.58612585e-01 3.61157477e-01 -1.68603683e+00 8.70558083e-01
9.90591943e-01 6.10193253e-01 3.52883607e-01 7.44907081e-01
-4.60144311e-01 8.86533380e-01 -6.54381573e-01 -3.49702299e-01
-1.14369929e-01 -3.81472915e-01 -1.05745745e+00 -5.07667124e-01
5.03622234e-01 -1.15943655e-01 1.40864357e-01 7.23762453e-01
1.09758057e-01 4.99227941e-02 8.19040477e-01 -1.10420763e+00
-4.67272580e-01 1.10167551e+00 4.67581712e-02 2.64600396e-01
6.24192953e-01 2.51744747e-01 1.11752772e+00 -5.61477602e-01
4.48491722e-01 1.52507973e+00 6.05340123e-01 7.85850644e-01
-1.40556514e+00 -4.58704829e-01 1.97250247e-01 1.39019676e-02
-8.08870137e-01 -3.18576425e-01 2.91922688e-01 -6.20999038e-01
9.72706616e-01 6.15825236e-01 4.06069428e-01 9.31405246e-01
2.60945857e-01 9.76678789e-01 1.56294203e+00 -5.14842272e-01
1.19046830e-01 5.62721968e-01 6.97868705e-01 6.85142219e-01
5.73410094e-02 3.06657821e-01 -7.88046658e-01 5.90794683e-02
3.71559173e-01 -2.81214058e-01 -2.44263008e-01 -9.79958400e-02
-1.01501322e+00 1.00482166e+00 3.35195959e-01 4.06556249e-01
9.66522247e-02 1.00710407e-01 2.00945303e-01 8.27443540e-01
4.20594551e-02 1.09095359e+00 -8.50230694e-01 -5.23859620e-01
-7.13825405e-01 1.86802432e-01 8.17075789e-01 6.55858815e-01
8.84441614e-01 -2.94306487e-01 -3.50131452e-01 7.94689655e-01
1.30958363e-01 2.94638038e-01 8.90963376e-01 -9.45988655e-01
3.41659784e-01 7.97022700e-01 5.86989745e-02 -4.94615227e-01
-5.38044691e-01 -2.86368042e-01 -1.02943242e-01 9.33914259e-03
7.48972654e-01 1.57275066e-01 -7.27747679e-01 2.03840971e+00
-4.32372727e-02 -4.61719453e-01 -7.48501122e-02 5.74902654e-01
7.80658066e-01 6.21822238e-01 2.08371893e-01 -1.10409394e-01
1.18492699e+00 -6.91126883e-01 -4.84850943e-01 -3.12956631e-01
1.31123841e+00 -7.92345941e-01 1.96542728e+00 4.40663397e-01
-1.08900762e+00 -7.06858337e-01 -1.05950058e+00 -1.36929946e-02
-6.04154825e-01 -2.96544552e-01 6.43103004e-01 7.42442012e-01
-1.22582102e+00 6.55787945e-01 -6.43939137e-01 -4.42611128e-01
-1.54499695e-01 4.49486524e-01 -8.26838762e-02 1.10907771e-01
-1.13968801e+00 1.08645582e+00 2.58009851e-01 -7.13003933e-01
-7.82448113e-01 -8.81132305e-01 -5.70364416e-01 2.44936258e-01
2.61265248e-01 -4.99261498e-01 1.62916994e+00 -1.07840657e+00
-1.56617618e+00 1.01920938e+00 -1.41218916e-01 -3.61347735e-01
3.11541200e-01 -4.09447402e-02 -2.19882265e-01 -3.70940655e-01
-7.91191384e-02 7.17869043e-01 5.68656206e-01 -1.18080473e+00
-7.69919872e-01 -3.04192007e-01 3.90364528e-01 5.48233353e-02
-4.54721510e-01 2.61955690e-02 -8.29128027e-02 -3.86637092e-01
-2.77259145e-02 -8.24912965e-01 9.13338438e-02 -5.08325279e-01
-3.39572728e-01 -5.00740170e-01 2.67311633e-01 -1.97606534e-01
1.69684315e+00 -1.99010277e+00 -3.86206247e-02 3.04478735e-01
1.56433523e-01 1.17354929e-01 -4.71908718e-01 2.33521134e-01
-2.88242493e-02 4.47313666e-01 1.11432485e-01 1.62028268e-01
2.61173338e-01 2.33441591e-01 -3.74186575e-01 -2.56441802e-01
1.65145531e-01 8.72577250e-01 -6.80642903e-01 -2.68515140e-01
-1.36002094e-01 -4.48854305e-02 -9.12479341e-01 3.80311102e-01
-3.82874668e-01 2.78199047e-01 -4.88587439e-01 8.36730599e-02
1.11842781e-01 -2.03702450e-01 -1.15783148e-01 1.71388134e-01
-2.23893285e-01 9.48266745e-01 -8.33904743e-01 1.20080113e+00
-6.42348707e-01 5.07133126e-01 -2.29984403e-01 -8.43052089e-01
9.70973134e-01 -9.57245827e-02 -4.72108722e-02 -1.06889367e+00
-7.03245550e-02 4.29809004e-01 3.31104964e-01 -4.20326203e-01
5.13849139e-01 -2.11386144e-01 -3.80271375e-01 6.86611116e-01
-1.89378336e-02 -5.96740723e-01 3.10773045e-01 1.53636187e-01
1.07489872e+00 4.31709029e-02 3.12161297e-01 -7.05670536e-01
5.34489036e-01 1.41704485e-01 9.87370387e-02 1.02983093e+00
-5.69078177e-02 2.25293264e-01 6.38469219e-01 -5.75026929e-01
-6.84659421e-01 -9.99416590e-01 -1.72641650e-01 1.96834004e+00
-1.09874196e-01 -5.84834337e-01 -1.03337204e+00 -8.74705195e-01
1.11878380e-01 1.23183560e+00 -7.22823381e-01 -5.42897999e-01
-7.96396792e-01 -4.67724174e-01 3.84880573e-01 5.67746520e-01
1.22058824e-01 -9.61172223e-01 -5.41223466e-01 -8.47211033e-02
-9.13265944e-02 -6.55667365e-01 -6.20502353e-01 7.75552571e-01
-1.09578478e+00 -8.42012227e-01 -1.84697181e-01 -6.83191001e-01
4.73481923e-01 8.82147029e-02 1.65257573e+00 5.67414701e-01
3.49784881e-01 5.10964513e-01 -1.90521613e-01 -4.03536171e-01
-1.04225540e+00 4.45649207e-01 1.11359820e-01 -4.47150648e-01
5.79948902e-01 -2.95738816e-01 -2.37083837e-01 6.62484050e-01
-9.02353048e-01 2.62684017e-01 4.03139770e-01 7.75469959e-01
3.08992088e-01 -3.03791821e-01 5.07943690e-01 -1.23075259e+00
8.88061166e-01 -2.57273316e-01 -6.69690549e-01 4.50216681e-01
-1.06684816e+00 7.51525104e-01 7.73720324e-01 -3.25837702e-01
-9.40080047e-01 -2.88571447e-01 -2.39006579e-01 3.28514695e-01
-2.71007329e-01 5.70655763e-01 7.24680051e-02 6.96084127e-02
1.05923402e+00 7.18789622e-02 -1.34254426e-01 -5.78271210e-01
2.43697360e-01 6.80402696e-01 1.35213330e-01 -9.25131619e-01
7.81474710e-01 -1.63664043e-01 -3.68557900e-01 -5.88393033e-01
-1.10778391e+00 -2.23873466e-01 -4.15379226e-01 -1.25919223e-01
6.27637267e-01 -3.86227787e-01 -9.51128066e-01 -1.93652157e-02
-8.78803730e-01 -9.71420705e-01 -4.24587786e-01 2.27852061e-01
-7.98675656e-01 -1.21135145e-01 -6.07178569e-01 -5.31607151e-01
1.16727419e-01 -1.56884754e+00 1.01806653e+00 2.58494020e-01
-7.98588276e-01 -1.17751038e+00 4.91595417e-02 3.96426618e-01
4.56476241e-01 -3.85652959e-01 1.59627032e+00 -1.09309649e+00
-3.78473043e-01 7.11736977e-02 7.92499632e-02 3.40879530e-01
-1.32196844e-01 -6.35312647e-02 -9.66244817e-01 -1.46178007e-01
1.67544037e-01 -5.77672601e-01 8.30106914e-01 2.03475073e-01
1.06227171e+00 -4.21624005e-01 -5.53205051e-02 6.63880885e-01
1.19595003e+00 3.74956816e-01 4.26793277e-01 5.44929087e-01
4.87205625e-01 6.02100611e-01 7.65828431e-01 4.12574932e-02
2.55826712e-01 7.16546774e-01 2.84968875e-02 3.09253454e-01
-1.36191070e-01 -5.24202347e-01 8.08769703e-01 1.13885248e+00
1.79476246e-01 -5.85342720e-02 -1.03163707e+00 1.06115684e-01
-1.43525136e+00 -6.69465601e-01 -5.85750155e-02 2.43559766e+00
1.31630254e+00 5.44166923e-01 1.15281276e-01 4.50597443e-02
2.28608817e-01 -2.43580699e-01 -3.45268756e-01 -9.80246484e-01
-4.77599213e-03 3.42912406e-01 2.02747762e-01 1.02681363e+00
-4.88346010e-01 1.05755687e+00 7.10353327e+00 8.56025040e-01
-1.15566456e+00 -1.10504389e-01 8.70456159e-01 1.57095477e-01
-7.40252912e-01 -1.33884892e-01 -1.07119095e+00 3.25846821e-01
1.43477607e+00 -1.14154994e-01 4.92397994e-01 8.08676422e-01
1.61807269e-01 -9.18297470e-02 -1.52842033e+00 5.49723327e-01
-2.66452521e-01 -1.10781491e+00 2.36968562e-01 -2.77251124e-01
5.94078541e-01 -1.32422090e-01 2.39623532e-01 9.03662026e-01
4.68578815e-01 -1.08070910e+00 8.65252376e-01 5.41642010e-01
6.03328645e-01 -6.41407192e-01 5.08142471e-01 6.15779161e-01
-6.77041769e-01 -2.09201351e-01 -1.94788352e-01 -2.09875852e-01
-5.23040533e-01 -2.98194345e-02 -1.17740798e+00 -2.52974927e-01
5.67860186e-01 1.96736455e-01 -1.06158733e+00 6.91512108e-01
-1.70846194e-01 9.18402016e-01 -2.21762821e-01 -5.29772341e-01
2.43801981e-01 -9.78218019e-02 3.79650295e-01 1.24953854e+00
1.39347240e-01 1.28272399e-01 5.80583513e-02 8.18633676e-01
9.69811901e-02 1.71264589e-01 -4.02980357e-01 -1.19736977e-01
2.84326404e-01 7.84572661e-01 -5.75562418e-01 -4.57948983e-01
-1.86034158e-01 6.20715380e-01 4.58287776e-01 2.56543249e-01
-6.55144989e-01 -1.86282173e-01 6.92652047e-01 4.85773772e-01
-1.52555928e-01 1.00825198e-01 -5.84422588e-01 -1.06174266e+00
-3.56917322e-01 -1.29312861e+00 3.23442638e-01 -8.65872443e-01
-9.94484782e-01 4.43198383e-01 6.89006299e-02 -7.02629805e-01
-4.04799134e-01 -9.71756518e-01 -4.68034595e-01 6.69308007e-01
-1.19332159e+00 -5.55707633e-01 -3.10165554e-01 3.52577955e-01
6.91013157e-01 8.09472986e-03 9.12627101e-01 -9.90927815e-02
-4.13739711e-01 9.17313814e-01 3.76443379e-02 -2.39314198e-01
8.39220822e-01 -1.68711948e+00 2.83331990e-01 4.35554922e-01
2.78703749e-01 1.04398239e+00 1.04345214e+00 -3.42376560e-01
-1.23017335e+00 -7.48002052e-01 9.17346716e-01 -9.52300906e-01
5.91762185e-01 -3.06625813e-01 -9.41729784e-01 7.87092865e-01
-5.49121294e-03 -7.57890999e-01 6.67125046e-01 4.19304669e-01
-5.31669021e-01 -1.65309146e-01 -7.35553026e-01 7.94709980e-01
8.36493731e-01 -4.85346913e-01 -7.25468159e-01 3.55373174e-01
1.06751180e+00 -3.07902724e-01 -8.22169542e-01 2.97256082e-01
4.89271373e-01 -1.45960283e+00 6.95068538e-01 -9.53624845e-01
2.86787838e-01 9.06451195e-02 -2.36743063e-01 -1.42004824e+00
-4.03029710e-01 -7.86876678e-01 1.92413896e-01 1.03426993e+00
8.96753967e-01 -6.28350556e-01 6.71416461e-01 6.16060615e-01
-3.52730215e-01 -9.47977424e-01 -4.58431154e-01 -6.67392373e-01
4.56819922e-01 -6.88057899e-01 9.37978625e-01 8.12756777e-01
9.24460515e-02 7.31808126e-01 3.28273743e-01 -4.16109473e-01
-5.31555153e-02 -9.03215557e-02 7.96898425e-01 -1.32359123e+00
-4.46359068e-01 -8.34331751e-01 -7.45919719e-02 -1.37444746e+00
-4.21264442e-03 -1.10732687e+00 1.58385098e-01 -1.06504047e+00
2.31587633e-01 -6.47733688e-01 -2.86629111e-01 5.51852405e-01
-3.12053204e-01 -1.57163650e-01 1.53263107e-01 -2.39561591e-02
-5.31542003e-01 1.09693393e-01 1.27888680e+00 -5.24230376e-02
-3.34600389e-01 2.13887975e-01 -8.40095699e-01 8.94656301e-01
8.25450778e-01 -4.48962122e-01 -5.44483304e-01 -3.87663722e-01
5.01162052e-01 -1.18056767e-01 -1.41832167e-02 -9.13652539e-01
-2.05882397e-02 -4.32685226e-01 1.60135180e-01 -1.91810548e-01
-4.17513028e-02 -5.96737087e-01 -3.52585018e-01 5.78957617e-01
-9.60103333e-01 3.78754705e-01 3.91549647e-01 8.73201191e-02
-1.11320227e-01 -5.41397512e-01 8.90060365e-01 -5.00217266e-02
-6.88805640e-01 -2.19247475e-01 -3.97621572e-01 4.02161241e-01
5.18957913e-01 -3.06324482e-01 -3.28754604e-01 -3.28490019e-01
-5.87849140e-01 9.72022712e-02 4.61701065e-01 6.00556076e-01
3.74312937e-01 -1.15073228e+00 -4.82798278e-01 4.41914052e-01
3.36310923e-01 -5.26459634e-01 -1.76679984e-01 8.73414695e-01
-4.45717007e-01 7.46765375e-01 9.32259485e-02 -8.09025645e-01
-1.26251125e+00 4.17463899e-01 3.80501896e-01 -5.09299636e-01
5.12818098e-02 9.59114254e-01 4.78747576e-01 -7.55493522e-01
3.56926352e-01 -9.67256129e-01 -1.80688038e-01 6.41821697e-02
4.83880371e-01 9.00012031e-02 2.41954371e-01 -3.38512361e-01
-6.42190874e-02 4.71893996e-01 -2.30217829e-01 -4.03712802e-02
1.13327444e+00 -2.18746066e-02 -9.33886468e-02 8.62504244e-01
1.15130186e+00 2.41042161e-03 -8.13105226e-01 -5.56446277e-02
5.24315000e-01 -4.79049683e-02 -3.17634046e-01 -1.18785739e+00
-5.14961064e-01 9.69099104e-01 5.45538843e-01 6.24864042e-01
1.04573941e+00 9.17298794e-02 4.77220386e-01 7.98884273e-01
4.19891626e-01 -1.06930745e+00 2.07032442e-01 8.67191732e-01
8.67118537e-01 -1.24975204e+00 -3.48422468e-01 -1.30377188e-01
-4.97206002e-01 1.18770552e+00 8.32795918e-01 1.97287828e-01
3.40963662e-01 3.54138643e-01 1.82666659e-01 -5.56101911e-02
-1.03913224e+00 3.97484228e-02 5.30262470e-01 3.57343823e-01
8.97474766e-01 2.41979420e-01 -2.02157229e-01 7.67883956e-01
-9.44746256e-01 -1.50823250e-01 4.25470799e-01 6.49444580e-01
-9.95213449e-01 -1.19203687e+00 -3.28924686e-01 6.75702631e-01
-2.61774421e-01 -2.48961270e-01 -5.24736285e-01 1.01520109e+00
-2.27573309e-02 8.22104692e-01 -1.34233668e-01 -7.31676400e-01
5.43749332e-01 4.78510350e-01 5.18107891e-01 -8.76672089e-01
-1.04906964e+00 -3.35671276e-01 1.21458381e-01 -7.82045186e-01
1.22809909e-01 -3.41784209e-01 -1.24611056e+00 -3.47116947e-01
-2.66027957e-01 3.84917140e-01 4.92246896e-01 1.03202140e+00
6.89376146e-02 4.58761513e-01 5.06737888e-01 -2.88548499e-01
-1.04562366e+00 -1.10301960e+00 -2.31044382e-01 5.64474583e-01
3.39194745e-01 -4.40394908e-01 -5.64001381e-01 -7.38223642e-02] | [10.654903411865234, 8.221073150634766] |
79e6c126-ddce-4dad-96b1-18d435d07306 | guiding-neural-entity-alignment-with | 2211.15833 | null | https://arxiv.org/abs/2211.15833v1 | https://arxiv.org/pdf/2211.15833v1.pdf | Guiding Neural Entity Alignment with Compatibility | Entity Alignment (EA) aims to find equivalent entities between two Knowledge Graphs (KGs). While numerous neural EA models have been devised, they are mainly learned using labelled data only. In this work, we argue that different entities within one KG should have compatible counterparts in the other KG due to the potential dependencies among the entities. Making compatible predictions thus should be one of the goals of training an EA model along with fitting the labelled data: this aspect however is neglected in current methods. To power neural EA models with compatibility, we devise a training framework by addressing three problems: (1) how to measure the compatibility of an EA model; (2) how to inject the property of being compatible into an EA model; (3) how to optimise parameters of the compatibility model. Extensive experiments on widely-used datasets demonstrate the advantages of integrating compatibility within EA models. In fact, state-of-the-art neural EA models trained within our framework using just 5\% of the labelled data can achieve comparable effectiveness with supervised training using 20\% of the labelled data. | ['Xia Zhang', 'Genghong Zhao', 'Guido Zuccon', 'Wen Hua', 'Harrisen Scells', 'Bing Liu'] | 2022-11-29 | null | null | null | null | ['entity-alignment', 'entity-alignment'] | ['knowledge-base', 'natural-language-processing'] | [ 1.67259112e-01 6.75945103e-01 -5.19133583e-02 -4.89088655e-01
6.73095062e-02 -3.76444429e-01 2.80716538e-01 3.36671650e-01
-6.80082619e-01 6.71935260e-01 -7.68684298e-02 -3.00933868e-01
-3.32150310e-01 -1.01252270e+00 -9.82100546e-01 -3.70113015e-01
-1.75514325e-01 7.79800057e-01 3.17346931e-01 -3.45659137e-01
-1.55236527e-01 2.99447358e-01 -1.46661735e+00 1.37991821e-02
1.14640832e+00 1.01940143e+00 5.70681086e-03 4.03362960e-01
-2.34627083e-01 1.00771892e+00 -3.62523526e-01 -9.68007386e-01
2.13165462e-01 -3.31040710e-01 -1.13783073e+00 -4.10512358e-01
5.25981784e-01 3.19687761e-02 -4.26013358e-02 1.18363845e+00
3.56876880e-01 -3.97255411e-03 7.69977033e-01 -1.49153984e+00
-6.11039519e-01 1.04645550e+00 -1.41631901e-01 1.78822383e-01
-3.31242047e-02 -4.47216690e-01 1.34607124e+00 -7.27472663e-01
8.98781955e-01 6.51946187e-01 9.41339612e-01 5.64094901e-01
-1.10992956e+00 -3.42088461e-01 3.26733500e-01 4.33137178e-01
-1.37739658e+00 -4.87068832e-01 7.66743124e-01 -2.13700801e-01
1.34189939e+00 2.07408816e-01 6.69308722e-01 7.57625997e-01
-1.82418272e-01 7.20365226e-01 7.60750473e-01 -6.87115788e-01
1.63463607e-01 3.47969621e-01 2.84477532e-01 6.34936452e-01
3.56871247e-01 -1.82711631e-01 -3.72969061e-01 2.67219190e-02
4.56984013e-01 -8.26685607e-01 -2.56193012e-01 -7.37633288e-01
-9.23759401e-01 7.15572834e-01 7.68737912e-01 5.45295477e-01
-3.21654409e-01 1.38315141e-01 2.96702534e-01 3.45919102e-01
2.60408193e-01 8.79723132e-01 -7.07733154e-01 2.73879230e-01
-6.80710495e-01 2.64425129e-01 1.04035521e+00 1.14145541e+00
6.72182739e-01 -1.12664543e-01 1.66872874e-01 8.16885531e-01
3.46805096e-01 -5.10037988e-02 3.13738436e-01 -5.02235115e-01
6.60227180e-01 9.69635844e-01 -9.61650535e-02 -1.05158591e+00
-6.32759273e-01 -6.66743755e-01 -7.97806621e-01 -1.58122793e-01
5.12550831e-01 -2.43514314e-01 -6.75019860e-01 2.13699651e+00
3.40659320e-01 3.73136818e-01 3.32432985e-01 4.92990911e-01
1.07489312e+00 2.12564752e-01 3.87572467e-01 -2.51499936e-02
1.01987708e+00 -9.51954603e-01 -6.18251324e-01 -5.69051981e-01
1.01194477e+00 -2.99779296e-01 4.95014876e-01 3.54712422e-04
-1.15177834e+00 -5.56123197e-01 -1.16944778e+00 6.05045352e-03
-9.32127118e-01 1.89056739e-01 6.59198701e-01 6.14081383e-01
-1.27862656e+00 7.41880298e-01 -4.49208587e-01 -4.21301842e-01
1.05843656e-01 7.93065190e-01 -5.18991828e-01 2.29886547e-01
-1.80653775e+00 1.45358324e+00 9.99144197e-01 3.64272714e-01
1.70829222e-02 -3.95544529e-01 -9.02793646e-01 3.02019328e-01
5.26848912e-01 -1.04106474e+00 9.29272056e-01 -1.11492789e+00
-9.01844621e-01 9.53412473e-01 2.24528447e-01 -4.69297171e-01
4.67568398e-01 -2.25524291e-01 -7.47764289e-01 -4.28968281e-01
-1.00842148e-01 9.16320920e-01 3.79997045e-01 -1.21119809e+00
-8.13507438e-01 -2.70590663e-01 1.55830517e-01 4.87082630e-01
-3.66051316e-01 -5.02755158e-02 -6.41686022e-01 -3.65949243e-01
-3.52835003e-03 -1.03897512e+00 -1.91888779e-01 -3.88999403e-01
-5.07654071e-01 -2.73062468e-01 3.65436256e-01 -6.88829303e-01
1.39033818e+00 -1.75134110e+00 4.12140220e-01 4.07618344e-01
1.82072505e-01 4.50470805e-01 -6.15929849e-02 3.23224932e-01
-2.00290352e-01 8.63168538e-02 -2.06166953e-02 -2.55944490e-01
2.76071906e-01 1.94942772e-01 -6.28616139e-02 -2.67878789e-02
3.10802013e-01 1.12083662e+00 -7.76176989e-01 -4.47261572e-01
-3.03192381e-02 2.64278919e-01 -5.61675072e-01 1.42273799e-01
-1.81010067e-01 5.92052676e-02 -2.01536164e-01 2.78701603e-01
5.33025086e-01 -2.84151435e-01 6.83403730e-01 -5.16641498e-01
2.35571682e-01 3.89509648e-01 -1.43189168e+00 1.26435268e+00
-1.51147544e-01 3.12426984e-01 -2.70542115e-01 -1.05086052e+00
9.07439053e-01 2.36414656e-01 1.95242196e-01 -6.40559494e-01
2.28151619e-01 4.86295462e-01 3.61842513e-01 -3.19486737e-01
8.76824915e-01 1.38913304e-01 3.68245803e-02 1.45353973e-01
3.67794544e-01 3.51882190e-01 3.07285845e-01 5.00435382e-02
9.62607384e-01 2.71985233e-01 5.55577099e-01 -2.67673254e-01
5.30325234e-01 -1.22327819e-01 6.54136062e-01 7.83479869e-01
-1.57274947e-01 7.34456256e-02 3.36159348e-01 -5.73680878e-01
-1.01738250e+00 -6.76035047e-01 -7.61824846e-02 1.00077546e+00
2.11218148e-01 -5.35929024e-01 -9.12683189e-01 -1.02424574e+00
-4.28350791e-02 7.99910784e-01 -7.50493944e-01 -2.45563179e-01
-6.51455760e-01 -8.60008836e-01 7.56651998e-01 8.30860794e-01
6.12359583e-01 -1.16803253e+00 -6.63332164e-01 3.98949057e-01
-1.98158443e-01 -1.22955596e+00 1.27961040e-01 6.67758048e-01
-6.24953568e-01 -1.05334914e+00 -2.40407094e-01 -9.74421322e-01
8.56472373e-01 -3.85392040e-01 1.48477006e+00 3.42333972e-01
2.60894716e-01 3.05435304e-02 -1.64947748e-01 -4.06934500e-01
-4.55843985e-01 4.92848217e-01 1.14251941e-01 -3.20278853e-01
6.61002219e-01 -5.23024619e-01 -1.10182442e-01 4.60333407e-01
-7.13344395e-01 3.55834156e-01 5.72861910e-01 8.47601354e-01
3.88645619e-01 3.09481472e-01 7.78197289e-01 -1.34689331e+00
6.60675347e-01 -3.61136943e-01 -3.85397613e-01 8.88390660e-01
-8.74778867e-01 2.60608643e-01 6.01980984e-01 -2.61571854e-01
-8.82282853e-01 2.81771533e-02 -2.65017092e-01 -1.30551383e-01
-2.37325758e-01 9.40962434e-01 -4.46317703e-01 -2.36225501e-02
6.74090564e-01 -9.80770364e-02 -5.13708413e-01 -1.23263478e-01
5.20960927e-01 4.88405198e-01 6.55783832e-01 -6.78056121e-01
5.82228005e-01 -1.22824475e-01 1.16217099e-02 -2.35807449e-01
-6.60588384e-01 -3.25365603e-01 -1.07144392e+00 -2.22059548e-01
6.15165174e-01 -6.96070433e-01 -3.42018694e-01 4.53803629e-01
-1.00327432e+00 -3.55032295e-01 -1.57292768e-01 2.71683633e-01
-3.56972814e-01 2.86190152e-01 -3.84235710e-01 -5.70184469e-01
-3.68989199e-01 -9.73457396e-01 5.06414950e-01 2.73040295e-01
-4.80764717e-01 -1.31211567e+00 7.02467412e-02 2.03069359e-01
4.94720876e-01 5.19078632e-04 1.26613224e+00 -1.18378127e+00
-1.69647440e-01 -2.08721489e-01 -1.43960208e-01 1.61463946e-01
4.31310646e-02 6.62526265e-02 -1.00702918e+00 -5.18197566e-02
-3.27874333e-01 -2.86377043e-01 6.69914067e-01 1.88771218e-01
7.19412446e-01 -2.24882603e-01 -6.32776558e-01 5.11781216e-01
1.26242530e+00 1.76420450e-01 7.55704820e-01 7.52129555e-01
8.41367424e-01 8.46993864e-01 4.73569751e-01 -2.06478447e-01
6.67408824e-01 8.87584507e-01 3.98860961e-01 6.66993065e-03
1.95980340e-01 -3.27691108e-01 -1.68378726e-02 8.56112003e-01
-1.32190377e-01 -4.79037017e-01 -1.14167929e+00 6.31862402e-01
-2.13132787e+00 -7.32409358e-01 -2.21757963e-01 1.96317184e+00
8.31451416e-01 3.76210600e-01 -1.66572519e-02 9.56722423e-02
6.63198531e-01 -1.62173122e-01 -4.64186430e-01 -3.46629202e-01
-2.30194777e-01 -1.22446299e-01 4.29579407e-01 1.85518861e-01
-1.37020552e+00 9.07387555e-01 5.77983189e+00 5.68474829e-01
-9.66040671e-01 -3.10274482e-01 5.02765477e-01 4.45843250e-01
-3.22580904e-01 9.54365060e-02 -1.02092946e+00 3.67871910e-01
9.35173273e-01 5.43275774e-02 2.23888785e-01 8.93340528e-01
-6.15516722e-01 6.29796684e-02 -1.53292930e+00 5.80267370e-01
9.88627449e-02 -1.04698348e+00 -1.21440798e-01 1.73656037e-03
6.57219410e-01 -3.34524526e-03 -2.00303227e-01 7.20375359e-01
5.01368999e-01 -1.11581945e+00 6.63612604e-01 4.93388265e-01
3.44227731e-01 -7.36747980e-01 1.10458803e+00 4.03777599e-01
-1.19568920e+00 2.55301118e-01 -4.18700933e-01 2.70978928e-01
9.72187296e-02 3.44208241e-01 -1.05582213e+00 1.06065154e+00
6.99039876e-01 5.93140721e-01 -9.42047536e-01 9.99511302e-01
-4.52087641e-01 1.97916329e-01 -3.71186584e-01 1.42553747e-01
7.96417221e-02 -1.53971121e-01 2.58276790e-01 1.17674577e+00
2.26758942e-01 -2.79762954e-01 -3.63938898e-01 9.14499640e-01
-2.68945158e-01 2.46781245e-01 -5.73383152e-01 -1.39198631e-01
5.36294401e-01 1.16885817e+00 -6.45703852e-01 -2.02825606e-01
-5.08970678e-01 7.80573606e-01 1.13121140e+00 9.73567516e-02
-8.30852509e-01 -2.33947054e-01 1.40269220e-01 -2.58975923e-01
4.04463261e-01 2.58288831e-01 -2.61666954e-01 -9.80140626e-01
1.24148596e-02 -7.18019962e-01 7.83919454e-01 -7.98247099e-01
-1.45505059e+00 9.29220378e-01 8.74113292e-04 -9.69364822e-01
-4.00596499e-01 -7.76824534e-01 -4.54109788e-01 8.08895648e-01
-1.56715286e+00 -1.24786735e+00 -3.13519895e-01 3.24081838e-01
-2.31758967e-01 -1.46492794e-01 1.06688559e+00 4.04920250e-01
-6.64304137e-01 7.29014695e-01 -4.41907868e-02 3.52904856e-01
6.43285036e-01 -1.55538011e+00 7.74483502e-01 8.68201613e-01
4.13816959e-01 8.69525731e-01 7.35959947e-01 -7.34874129e-01
-8.44888687e-01 -9.57756341e-01 1.44712305e+00 -3.89818281e-01
6.28103733e-01 -7.13100731e-02 -1.16587710e+00 8.67622852e-01
1.40572131e-01 4.82964739e-02 8.40625644e-01 5.13581216e-01
-4.13617402e-01 1.15905404e-01 -9.21183288e-01 7.35845089e-01
1.24257648e+00 -3.98450613e-01 -7.27325439e-01 -2.15500697e-01
2.44367629e-01 -6.03618026e-01 -1.15011895e+00 9.26412344e-01
5.75598598e-01 -8.33445907e-01 8.35973620e-01 -9.13305342e-01
2.86365271e-01 -3.46990854e-01 -8.12094361e-02 -1.63627529e+00
-3.51562142e-01 1.10189743e-01 -4.60090280e-01 1.41303635e+00
7.24843442e-01 -4.83140886e-01 7.64850557e-01 1.09192979e+00
1.09096775e-02 -7.93664813e-01 -7.42503643e-01 -7.74836838e-01
8.17952156e-02 -2.23172978e-01 7.30899215e-01 1.39909911e+00
1.80068895e-01 5.42692304e-01 -3.32409441e-01 3.67134243e-01
1.83815911e-01 -5.61126806e-02 6.22662008e-01 -1.47847080e+00
-4.22469467e-01 -4.80264604e-01 -4.98454452e-01 -6.52302027e-01
4.05079484e-01 -9.81050253e-01 -8.89844969e-02 -1.73116338e+00
1.34072885e-01 -8.41401100e-01 -4.12350416e-01 7.85988748e-01
-4.54058468e-01 -1.09135449e-01 1.15046926e-01 1.23575643e-01
-5.75497746e-01 3.39361995e-01 6.02579534e-01 -4.18437757e-02
1.84755350e-04 -2.28447959e-01 -6.93183124e-01 6.89005256e-01
6.88393176e-01 -4.04226303e-01 -5.75844109e-01 -5.09473503e-01
9.28380251e-01 -2.91542709e-01 2.08920777e-01 -1.00303829e+00
6.75957799e-01 4.81965542e-02 3.98860753e-01 -2.80978650e-01
1.71809331e-01 -1.11762023e+00 7.27870703e-01 8.65788907e-02
-4.13275391e-01 1.86186343e-01 3.05919588e-01 4.82669145e-01
-2.74207890e-01 -5.87669611e-01 4.13005531e-01 7.91508406e-02
-1.10736585e+00 -5.49431555e-02 9.14030448e-02 9.32541043e-02
1.00239110e+00 -3.41599762e-01 -5.75873792e-01 -1.10442340e-01
-8.02322745e-01 3.72980267e-01 4.75872368e-01 5.03719807e-01
2.18112409e-01 -1.29959035e+00 -4.37564254e-01 -5.83565608e-03
2.93819189e-01 2.04462305e-01 8.62917677e-02 7.48523235e-01
-2.59013355e-01 3.96707833e-01 -3.59208554e-01 -2.53162980e-01
-1.41416693e+00 4.99701709e-01 6.54319048e-01 -7.44451940e-01
-1.88545555e-01 9.43815112e-01 -5.49875274e-02 -8.91975284e-01
2.86890626e-01 -1.59162059e-01 -6.44444168e-01 1.69559926e-01
-7.82090332e-03 2.61781886e-02 5.10412514e-01 -8.18808079e-01
-5.03171623e-01 2.43472517e-01 -4.54979748e-01 3.64275783e-01
1.37053025e+00 1.01972938e-01 1.60688497e-02 1.84272528e-01
7.76462078e-01 -1.30379766e-01 -7.44067907e-01 -3.55906785e-01
5.30548453e-01 5.69699705e-03 -7.33357295e-02 -9.72008646e-01
-1.07363141e+00 4.84362036e-01 2.51367480e-01 1.98346958e-01
1.02876186e+00 -1.58245787e-01 3.15172225e-01 6.09544337e-01
4.59900022e-01 -1.27889907e+00 -4.83258814e-01 5.93488872e-01
4.64968234e-01 -1.08657014e+00 -4.42715213e-02 -6.78631723e-01
-8.51954639e-01 1.05687022e+00 9.45172608e-01 2.22912550e-01
3.74365389e-01 1.32140964e-01 9.04579181e-03 -5.18876910e-01
-7.39852846e-01 -3.37259501e-01 6.82478666e-01 5.62408030e-01
4.57373559e-01 -1.47497445e-01 -3.15974623e-01 8.23380053e-01
-2.88335979e-01 -1.66281145e-02 2.61874259e-01 8.58113766e-01
-1.11852884e-01 -1.27709103e+00 2.83242583e-01 4.80998456e-01
-3.60374629e-01 -2.04252928e-01 -6.99588299e-01 9.28429604e-01
3.08466405e-01 6.09867513e-01 -1.57559261e-01 -4.83472943e-01
4.95236397e-01 4.11288738e-01 4.03139353e-01 -3.21264565e-01
-9.02856290e-01 -4.73602891e-01 7.08914340e-01 -1.55386120e-01
-6.48060441e-01 -4.82654512e-01 -1.05434775e+00 -1.58084080e-01
-7.80673623e-01 2.21582040e-01 4.87343222e-01 1.09778023e+00
3.77554685e-01 5.32437265e-01 4.68381755e-02 -3.48641247e-01
-4.26075965e-01 -8.60090494e-01 -5.36006570e-01 5.44946074e-01
-2.80296087e-01 -7.31173396e-01 -1.20611668e-01 -1.24178506e-01] | [9.069852828979492, 8.260058403015137] |
2d0373f3-b5f0-4e3a-b358-57c33c40cca5 | the-sensitivity-of-topic-coherence-evaluation | null | null | https://aclanthology.org/N16-1057 | https://aclanthology.org/N16-1057.pdf | The Sensitivity of Topic Coherence Evaluation to Topic Cardinality | null | ['Timothy Baldwin', 'Jey Han Lau'] | 2016-06-01 | null | null | null | naacl-2016-6 | ['coherence-evaluation'] | ['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] | [-7.339041233062744, 3.7041850090026855] |
377ca261-be3f-457f-b2f7-33636867f8a6 | developing-cooperative-policies-for-multi-1 | 2205.05230 | null | https://arxiv.org/abs/2205.05230v1 | https://arxiv.org/pdf/2205.05230v1.pdf | Developing cooperative policies for multi-stage reinforcement learning tasks | Many hierarchical reinforcement learning algorithms utilise a series of independent skills as a basis to solve tasks at a higher level of reasoning. These algorithms don't consider the value of using skills that are cooperative instead of independent. This paper proposes the Cooperative Consecutive Policies (CCP) method of enabling consecutive agents to cooperatively solve long time horizon multi-stage tasks. This method is achieved by modifying the policy of each agent to maximise both the current and next agent's critic. Cooperatively maximising critics allows each agent to take actions that are beneficial for its task as well as subsequent tasks. Using this method in a multi-room maze domain and a peg in hole manipulation domain, the cooperative policies were able to outperform a set of naive policies, a single agent trained across the entire domain, as well as another sequential HRL algorithm. | ['Chris Lehnert', 'Jordan Erskine'] | 2022-05-11 | null | null | null | null | ['hierarchical-reinforcement-learning'] | ['methodology'] | [ 2.06819579e-01 5.92159748e-01 1.51133627e-01 -4.74622175e-02
-5.22386968e-01 -7.39322901e-01 7.05566227e-01 3.16789091e-01
-9.17377949e-01 1.53607213e+00 1.60284787e-01 -2.22220480e-01
-6.42515302e-01 -5.77358127e-01 -4.20241892e-01 -9.91161823e-01
-2.59382278e-01 9.55074430e-01 6.73405886e-01 -5.83893239e-01
5.61211824e-01 3.92725140e-01 -1.52449965e+00 7.21389949e-02
7.41528213e-01 4.42210346e-01 5.72046220e-01 7.83165693e-01
3.22408885e-01 1.56641328e+00 -7.16612220e-01 6.34665191e-02
4.25503939e-01 -5.59351683e-01 -1.02868450e+00 5.50425127e-02
-4.17637378e-01 -4.30808574e-01 1.78891554e-01 5.53168356e-01
5.30651391e-01 6.89148545e-01 5.79025090e-01 -1.21176553e+00
2.86241528e-02 7.35929430e-01 -4.44934756e-01 3.18837613e-01
5.37878096e-01 3.95039558e-01 8.69717777e-01 3.08998615e-01
5.08208573e-01 1.41518998e+00 2.14874178e-01 6.87572777e-01
-1.19938648e+00 -4.31885123e-01 4.75830346e-01 8.18261802e-02
-4.61034805e-01 -1.23364143e-01 5.37113309e-01 -1.91909611e-01
1.43262184e+00 -1.58910856e-01 1.03692937e+00 9.15871322e-01
5.39894223e-01 8.31738472e-01 1.70345044e+00 -5.87425888e-01
6.12838447e-01 -1.56128824e-01 -4.28617358e-01 6.29503608e-01
-4.68670353e-02 6.13458514e-01 -4.96347040e-01 -2.83546656e-01
9.51188743e-01 -3.95613283e-01 3.05283904e-01 -7.44423211e-01
-1.20341957e+00 9.69513118e-01 1.98229626e-01 4.27575141e-01
-8.59009504e-01 3.06810796e-01 4.21045631e-01 9.39523518e-01
-6.89459443e-02 1.22897315e+00 -5.87721586e-01 -1.81281552e-01
-4.05261457e-01 5.26889324e-01 9.06107366e-01 5.10078073e-01
4.94138569e-01 1.68229386e-01 -2.18832582e-01 4.46090847e-01
3.78709048e-01 -5.14361709e-02 4.10950333e-01 -1.63793135e+00
4.17176992e-01 4.44878131e-01 6.10293567e-01 -3.22952241e-01
-6.70958638e-01 -2.11521417e-01 -1.23373322e-01 1.05842793e+00
4.21659708e-01 -5.92894554e-01 -7.28590369e-01 1.67987454e+00
5.06320834e-01 -4.67135251e-01 5.31946778e-01 5.85221708e-01
1.05080254e-01 5.74907005e-01 4.35195655e-01 -4.48097765e-01
9.32333350e-01 -1.33486664e+00 -3.92177075e-01 -4.57880348e-01
7.07235336e-01 -3.69511157e-01 5.15338540e-01 7.43742585e-01
-1.51460576e+00 -6.06516838e-01 -8.96281064e-01 2.80647516e-01
-2.31870472e-01 -5.52214801e-01 7.47833490e-01 2.83104658e-01
-1.35110617e+00 8.99257362e-01 -7.41286159e-01 -3.05181175e-01
7.16698617e-02 5.18243670e-01 -2.15963006e-01 2.99978465e-01
-1.12335432e+00 1.56079376e+00 8.72549772e-01 -2.79872656e-01
-1.52629220e+00 2.18792353e-02 -6.66140139e-01 1.27046242e-01
7.28792787e-01 -6.08546913e-01 1.77467942e+00 -1.07901931e+00
-1.85342824e+00 4.70954001e-01 4.07364160e-01 -5.06253362e-01
5.47655284e-01 -1.02522649e-01 3.48237187e-01 1.72616735e-01
3.48395139e-01 9.08499479e-01 8.62631381e-01 -1.31775427e+00
-1.20019042e+00 -2.05785006e-01 6.10065401e-01 9.80557621e-01
1.72552019e-01 -9.78011042e-02 4.84166950e-01 -8.45082104e-02
-3.29949379e-01 -9.13534164e-01 -7.82791317e-01 -7.88960516e-01
3.20362955e-01 -6.60075486e-01 2.48684719e-01 -3.30156475e-01
5.69651544e-01 -1.83040667e+00 6.65308475e-01 -1.15890726e-02
5.08465432e-02 2.52817525e-03 -2.68892705e-01 7.65393794e-01
-2.40354706e-02 -3.82266998e-01 1.19352959e-01 -9.32992026e-02
1.62938759e-01 6.69741988e-01 9.64014605e-02 1.74925834e-01
-1.17576957e-01 8.03245902e-01 -1.24018502e+00 -3.19573760e-01
5.41721918e-02 -3.82528961e-01 -4.61956441e-01 3.23649228e-01
-4.98116344e-01 7.68667936e-01 -7.94319451e-01 2.19264883e-03
-5.90880960e-02 1.07721463e-01 5.70757270e-01 1.09799838e+00
-3.88528347e-01 2.42063344e-01 -1.08924615e+00 1.43339014e+00
-2.46525377e-01 1.03450239e-01 2.19533354e-01 -1.00294268e+00
8.69805455e-01 7.47446716e-01 5.89832246e-01 -9.35844719e-01
5.15231974e-02 2.31456265e-01 5.91791213e-01 -2.05230415e-01
2.34909445e-01 -3.31924796e-01 -1.09902829e-01 7.33848095e-01
-2.11714972e-02 -5.51248491e-01 4.04563367e-01 7.06050098e-02
1.51797354e+00 5.49047291e-01 4.82525110e-01 -2.98769027e-01
4.85483944e-01 3.85862768e-01 6.17093682e-01 1.22085619e+00
-5.34087062e-01 -2.84842849e-01 8.88227224e-01 -4.48142678e-01
-1.21272016e+00 -7.88132250e-01 5.05343974e-01 1.53232205e+00
5.41559746e-03 1.62885159e-01 -4.46887195e-01 -7.59167373e-01
7.61893839e-02 1.06733203e+00 -9.16303396e-01 -4.78876233e-02
-6.36495233e-01 -4.05074984e-01 1.11847080e-01 3.62541318e-01
5.97300351e-01 -1.81967092e+00 -1.60947657e+00 6.28279209e-01
1.65536672e-01 -5.12466729e-01 1.33200020e-01 9.35018718e-01
-8.68065059e-01 -9.97297645e-01 -6.80141687e-01 -7.44532466e-01
4.44937587e-01 6.78888755e-03 9.81327534e-01 1.58443630e-01
2.55503297e-01 6.97575688e-01 -6.18853390e-01 -6.57308519e-01
-5.24578869e-01 8.68687779e-02 -3.73634510e-02 -7.85888135e-01
-8.00025016e-02 -5.98077953e-01 -3.92308593e-01 2.07416058e-01
-6.07525110e-01 1.34191230e-01 5.82917988e-01 9.21322286e-01
-9.06220637e-03 4.43442434e-01 7.18752623e-01 -5.37595510e-01
1.27002621e+00 -2.94083506e-01 -6.16629362e-01 2.84805030e-01
-6.02466345e-01 2.85375178e-01 6.46840155e-01 -4.58880246e-01
-1.47182727e+00 7.21306279e-02 1.81568637e-01 2.33671263e-01
-3.06888849e-01 3.03212821e-01 2.91537523e-01 7.56596122e-03
6.97672725e-01 1.67964578e-01 1.23311684e-01 1.73985455e-02
6.20244369e-02 1.37753468e-02 -3.29472311e-02 -8.24969471e-01
4.23443884e-01 -1.91253036e-01 2.08311379e-01 -2.46307775e-01
-6.84914470e-01 -4.65844512e-01 -6.05745971e-01 -5.07924557e-01
8.50916982e-01 -7.13895380e-01 -1.02806222e+00 4.79299396e-01
-9.57682014e-01 -1.12617147e+00 -4.64850336e-01 4.13125753e-01
-1.21721947e+00 1.39779568e-01 -4.94788855e-01 -9.85920787e-01
8.77452493e-02 -1.03762066e+00 6.53967083e-01 6.67489290e-01
-2.22705662e-01 -1.06306148e+00 4.06327903e-01 2.89512515e-01
3.31965208e-01 1.87797919e-01 8.46128404e-01 -7.93101311e-01
-2.25592002e-01 2.50448316e-01 3.64986300e-01 -6.57029003e-02
-1.02763131e-01 -5.23928583e-01 -4.46921796e-01 -6.84440136e-01
3.01648617e-01 -1.12101805e+00 4.82964993e-01 3.13404888e-01
2.15631649e-01 -1.09793425e-01 -1.56905249e-01 -1.90780014e-01
1.21156526e+00 1.01907134e+00 7.61973858e-01 1.13659751e+00
-2.76496615e-02 8.59438360e-01 1.03828299e+00 3.22534293e-01
1.98895305e-01 4.22962129e-01 6.47233844e-01 2.60430932e-01
4.89005983e-01 1.21139083e-03 4.84380841e-01 9.06962454e-02
-5.42789817e-01 -1.42092854e-01 -8.42970729e-01 5.99156439e-01
-2.28320789e+00 -1.26323617e+00 1.67336404e-01 1.72549820e+00
7.92302668e-01 4.08009082e-01 4.83684868e-01 3.41658806e-03
3.53129834e-01 6.13381565e-02 -7.89016187e-01 -9.20394599e-01
2.42280692e-01 2.36193538e-01 3.53788704e-01 5.71713626e-01
-8.73623848e-01 1.24977756e+00 6.66176987e+00 4.77287292e-01
-2.82925874e-01 1.73893720e-01 1.95760697e-01 -8.15451294e-02
8.50427374e-02 2.47186363e-01 -6.09875143e-01 1.38135910e-01
8.97911787e-01 1.70787086e-03 8.75901163e-01 7.36045182e-01
1.39669344e-01 -1.07761228e+00 -9.40399706e-01 -6.35831207e-02
-4.52332973e-01 -7.46911883e-01 -6.14580333e-01 1.23421185e-01
9.00885701e-01 -1.62206396e-01 -2.11994484e-01 7.63433099e-01
1.41346323e+00 -9.61609066e-01 7.51437485e-01 4.94966388e-01
-1.30529761e-01 -8.95038724e-01 6.03166878e-01 1.09300935e+00
-7.03750610e-01 -6.11838102e-01 -2.19780952e-01 -8.69191945e-01
8.16112310e-02 -6.11960351e-01 -1.02990854e+00 5.29560924e-01
5.76733232e-01 1.47763282e-01 -2.92165369e-01 9.27368879e-01
-7.40536034e-01 1.11370713e-01 -1.03254013e-01 -4.77246016e-01
9.73756969e-01 -1.58660665e-01 3.24461073e-01 4.78381068e-01
-1.38336450e-01 5.02498031e-01 5.12683034e-01 2.19196543e-01
8.77626300e-01 -1.95080653e-01 -5.96677721e-01 4.02288623e-02
4.02123183e-01 1.03574300e+00 -1.04729187e+00 -2.97181427e-01
-5.88088259e-02 8.49570930e-01 7.62103260e-01 3.01994920e-01
-6.30181730e-01 -1.14172153e-01 1.36654750e-01 -2.85575420e-01
3.81576866e-01 -3.92052203e-01 -1.19965971e-02 -4.63322103e-01
-4.17765349e-01 -1.11920857e+00 5.30172288e-01 -9.64938819e-01
-8.36537898e-01 2.97169536e-01 2.53695250e-01 -7.91451395e-01
-6.32988274e-01 -3.70512336e-01 -6.37509048e-01 9.07700241e-01
-1.57093537e+00 -7.51337528e-01 7.79177248e-02 6.32526278e-01
6.24831855e-01 -5.60992718e-01 8.86892319e-01 -6.55227482e-01
-2.99664050e-01 -2.57707834e-01 7.83492476e-02 -2.99720705e-01
3.75612110e-01 -1.49588287e+00 3.66195329e-02 3.65471393e-01
-4.65388805e-01 3.84890884e-01 8.25158715e-01 -7.86170065e-01
-8.37658942e-01 -3.32723856e-01 3.98993105e-01 -2.75513619e-01
7.14006841e-01 2.69430935e-01 -4.67014998e-01 8.18459928e-01
7.94279754e-01 -8.26962650e-01 2.70635307e-01 1.02587968e-01
3.88406336e-01 2.94440866e-01 -1.22279465e+00 6.67652965e-01
8.74203026e-01 8.75675753e-02 -1.20860422e+00 3.38266760e-01
4.36638027e-01 -5.51225066e-01 -6.87553525e-01 1.84954032e-01
2.75271356e-01 -1.28007936e+00 8.09544206e-01 -8.78838241e-01
5.61305940e-01 -2.61303652e-02 3.05006266e-01 -1.82279837e+00
-8.55640650e-01 -6.20903015e-01 7.07004070e-02 6.05244219e-01
2.75133282e-01 -9.19705451e-01 7.59786248e-01 4.81148183e-01
-8.25218111e-02 -6.42574728e-01 -9.40051794e-01 -8.06676686e-01
3.16516668e-01 3.19212884e-01 1.93945900e-01 6.53478026e-01
3.05799276e-01 5.61791420e-01 -3.65642071e-01 -3.84925269e-02
7.24069178e-01 -1.81838155e-01 6.71729684e-01 -1.27138793e+00
-4.08282638e-01 -4.73577559e-01 1.43569142e-01 -6.58214033e-01
2.38015875e-01 -4.64180052e-01 3.04992080e-01 -1.90007830e+00
-5.45389354e-02 -7.03569889e-01 -2.02717304e-01 7.38004625e-01
-3.87623459e-02 -5.74158370e-01 4.36751872e-01 1.97385266e-01
-9.63892877e-01 3.64069998e-01 1.81918371e+00 1.45114183e-01
-4.60312486e-01 1.21709080e-02 -7.24734485e-01 8.61117661e-01
1.31608331e+00 -5.29518664e-01 -7.80738115e-01 -2.31614679e-01
4.13305253e-01 6.29490852e-01 2.06523448e-01 -1.04405999e+00
5.26437759e-01 -6.65551245e-01 5.73226333e-01 -1.47218928e-01
3.61570328e-01 -8.03258181e-01 1.07803106e-01 1.04689348e+00
-6.28075123e-01 3.79263431e-01 2.48077706e-01 5.50746500e-01
1.13150559e-01 -8.75180602e-01 8.00008357e-01 -9.21946824e-01
-9.58563924e-01 -4.37579691e-01 -1.02810907e+00 -1.14670791e-01
1.76667917e+00 -3.35296720e-01 -2.29750827e-01 -5.35541654e-01
-9.33412075e-01 9.77691770e-01 3.30329329e-01 3.57916504e-02
3.74997944e-01 -7.19195962e-01 -6.25265718e-01 -2.50778675e-01
-4.68582779e-01 4.57773209e-02 1.20397210e-01 8.23688924e-01
-4.01096761e-01 6.05835259e-01 -9.97131348e-01 -6.75954148e-02
-1.17947662e+00 6.23847604e-01 5.31561255e-01 -1.02102101e+00
-5.81471562e-01 8.53615642e-01 -9.71230790e-02 -4.89138126e-01
3.38411063e-01 1.99873492e-01 -8.97877574e-01 3.16963106e-01
1.90602958e-01 3.08158696e-01 -3.25406969e-01 -4.63787792e-03
-7.05210259e-04 1.81843206e-01 -2.31910169e-01 -6.47480667e-01
1.51503921e+00 -2.06324607e-01 -5.99181019e-02 4.09034640e-01
1.11243442e-01 -4.64206189e-01 -1.73140192e+00 -7.20386654e-02
2.07749963e-01 -2.74254888e-01 -8.32476020e-02 -1.19418848e+00
-4.38570440e-01 3.60137254e-01 1.49211707e-03 7.21453071e-01
9.19220090e-01 -2.53669977e-01 1.05452560e-01 6.11058474e-01
1.07381129e+00 -1.60563278e+00 6.00830615e-01 9.62583542e-01
8.81773889e-01 -1.01728618e+00 2.08710149e-01 2.54859984e-01
-1.27356529e+00 1.11527061e+00 7.78897643e-01 -2.94085801e-01
-8.55156481e-02 -3.18939127e-02 -1.75383359e-01 -3.35276395e-01
-1.34715557e+00 -5.32342553e-01 -4.33874905e-01 7.97551215e-01
-2.25835647e-02 -5.04098944e-02 -3.83003861e-01 -1.38844371e-01
-7.09792748e-02 -3.68991382e-02 6.43467903e-01 1.70989668e+00
-9.61489499e-01 -1.20609629e+00 -5.65867901e-01 3.35084379e-01
-6.67143315e-02 4.68012899e-01 -4.75028723e-01 8.68129492e-01
1.66824505e-01 1.08751667e+00 -9.84070525e-02 1.97132602e-01
2.61184871e-01 1.26444548e-01 9.70720291e-01 -7.66581297e-01
-9.41052258e-01 1.16561152e-01 2.17709124e-01 -4.58061665e-01
-6.48636043e-01 -1.09568512e+00 -1.26941061e+00 -1.43072754e-01
2.00432315e-01 4.87478107e-01 2.45223656e-01 1.04773474e+00
-2.71178782e-01 9.04099345e-01 6.26053870e-01 -9.96844471e-01
-9.55130816e-01 -8.93948078e-01 -7.84547150e-01 3.92905027e-02
2.79949188e-01 -9.93939996e-01 -1.17799781e-01 -4.43134904e-01] | [3.8774595260620117, 1.5840948820114136] |
1f20bfab-d704-4a9e-995a-9d9de7b3b73a | differentiable-time-frequency-scattering-in | 2204.08269 | null | https://arxiv.org/abs/2204.08269v4 | https://arxiv.org/pdf/2204.08269v4.pdf | Differentiable Time-Frequency Scattering on GPU | Joint time-frequency scattering (JTFS) is a convolutional operator in the time-frequency domain which extracts spectrotemporal modulations at various rates and scales. It offers an idealized model of spectrotemporal receptive fields (STRF) in the primary auditory cortex, and thus may serve as a biological plausible surrogate for human perceptual judgments at the scale of isolated audio events. Yet, prior implementations of JTFS and STRF have remained outside of the standard toolkit of perceptual similarity measures and evaluation methods for audio generation. We trace this issue down to three limitations: differentiability, speed, and flexibility. In this paper, we present an implementation of time-frequency scattering in Python. Unlike prior implementations, ours accommodates NumPy, PyTorch, and TensorFlow as backends and is thus portable on both CPU and GPU. We demonstrate the usefulness of JTFS via three applications: unsupervised manifold learning of spectrotemporal modulations, supervised classification of musical instruments, and texture resynthesis of bioacoustic sounds. | ['George Fazekas', 'Mathieu Lagrange', 'Vincent Lostanlen', 'Han Han', 'Changhong Wang', 'Cyrus Vahidi', 'John Muradeli'] | 2022-04-18 | null | null | null | null | ['audio-generation'] | ['audio'] | [ 2.06977576e-01 -4.63153213e-01 5.10228038e-01 -6.24750704e-02
-9.74404871e-01 -7.32223809e-01 5.07547796e-01 1.01924829e-01
-4.89711553e-01 4.07375783e-01 3.78140837e-01 -2.99301893e-01
-2.21827060e-01 -4.92000520e-01 -4.28465098e-01 -6.48173988e-01
-6.61971569e-01 -3.79824519e-01 3.23493540e-01 -1.47445545e-01
4.33389097e-01 3.91543686e-01 -1.97660410e+00 6.92004442e-01
1.63216516e-01 1.22170866e+00 2.38260120e-01 1.25943816e+00
2.07159787e-01 6.42331839e-01 -7.27061629e-01 9.12088342e-03
1.70228079e-01 -2.76554227e-01 -7.45731652e-01 -3.99761915e-01
3.86220366e-01 -2.25213915e-01 -2.03608200e-01 8.81889760e-01
9.19312716e-01 3.70941460e-01 5.51734269e-01 -1.08839929e+00
-4.68896210e-01 7.77594507e-01 -2.57490158e-01 7.47745216e-01
3.95865589e-01 8.76895860e-02 9.63540375e-01 -1.03731298e+00
1.67568266e-01 1.32573676e+00 9.69299018e-01 3.14746559e-01
-1.49924910e+00 -5.77056646e-01 -5.50717413e-01 4.47488576e-02
-1.19439650e+00 -8.50926936e-01 4.19684917e-01 -7.33933091e-01
1.29159009e+00 4.53637540e-01 6.67719126e-01 9.22892392e-01
2.70268649e-01 4.98781204e-01 1.16651011e+00 -4.30000573e-01
4.20328766e-01 -4.82389688e-01 -2.85128325e-01 2.92040050e-01
-6.17085159e-01 3.92839104e-01 -1.07194710e+00 -5.79962790e-01
1.04345810e+00 -7.59366751e-01 -2.74904191e-01 4.93338674e-01
-1.58375227e+00 5.06345749e-01 1.05752289e-01 1.69452265e-01
-1.26681328e-01 4.96351779e-01 5.77709198e-01 3.74606550e-01
5.98272622e-01 4.97928143e-01 -3.48412693e-01 -3.99370342e-01
-1.02790010e+00 2.89423257e-01 6.61665440e-01 3.88862878e-01
2.90504843e-01 5.58384061e-01 1.08099110e-01 9.36078429e-01
2.21188813e-01 2.41404817e-01 7.69875646e-01 -1.62776840e+00
-1.74268186e-01 -3.94059807e-01 -1.61042679e-02 -7.47726142e-01
-4.35229719e-01 -2.75220603e-01 -3.75818431e-01 4.86676037e-01
5.15905857e-01 -1.96041819e-02 -5.66000581e-01 1.69314408e+00
2.83800721e-01 5.64220607e-01 -2.96859324e-01 1.13410711e+00
8.85151207e-01 5.86224437e-01 1.54244021e-01 -4.97556143e-02
1.51109624e+00 -3.19148421e-01 -5.35763979e-01 1.92799538e-01
2.63242960e-01 -1.48056686e+00 1.05479467e+00 6.00494146e-01
-1.29857850e+00 -7.32423842e-01 -1.10752153e+00 -2.60356188e-01
-2.81742841e-01 -1.05989747e-01 9.38768685e-01 6.88985467e-01
-1.40853715e+00 1.14878178e+00 -9.74099398e-01 3.07960268e-02
2.23038986e-01 2.34991685e-01 -1.14653558e-01 8.06837618e-01
-1.09700727e+00 4.91748720e-01 5.80635369e-02 -2.69895643e-01
-7.28563607e-01 -1.14422905e+00 -5.01915216e-01 -3.82708502e-03
-5.55341363e-01 -6.74443185e-01 1.75081468e+00 -7.69973695e-01
-1.73899055e+00 7.63500750e-01 1.60273612e-01 -6.08629227e-01
1.36104245e-02 1.07269241e-02 -6.33181870e-01 2.85145938e-01
7.36986026e-02 8.13026369e-01 1.16913056e+00 -4.16602373e-01
-4.32655990e-01 -2.20329631e-02 -9.40090790e-02 6.18827306e-02
-2.23634630e-01 3.40086192e-01 4.13222045e-01 -1.15107405e+00
1.64161429e-01 -6.54421449e-01 1.73561722e-02 1.51243657e-01
-9.62679461e-02 -7.32809678e-02 2.09965989e-01 -6.25006557e-01
9.44596648e-01 -2.73881817e+00 -9.91287008e-02 -1.74583435e-01
2.39469588e-01 -4.96597067e-02 -2.20006257e-01 3.75733167e-01
-4.19569999e-01 -7.40016326e-02 -2.18733028e-01 -1.09654896e-01
1.83191687e-01 -4.19176698e-01 -6.91837013e-01 5.66259265e-01
1.48015827e-01 5.34774840e-01 -9.88582313e-01 -3.22699219e-01
2.12106425e-02 7.73922920e-01 -6.63381696e-01 -2.27517545e-01
-1.80902287e-01 5.40633142e-01 1.07324973e-01 4.94041145e-01
6.55013680e-01 8.03486779e-02 -2.05580339e-01 -2.64691561e-01
-5.64621150e-01 9.19909418e-01 -1.13810956e+00 2.00637865e+00
-4.28316027e-01 1.03832722e+00 1.69915304e-01 -5.39428055e-01
6.64122641e-01 5.82102239e-01 6.68298662e-01 -4.66486305e-01
6.78897798e-02 4.89180684e-01 2.02085450e-01 -3.36926877e-01
4.92261440e-01 -2.51574725e-01 1.14388451e-01 7.81329691e-01
4.68082786e-01 -3.69787961e-01 -3.74436751e-02 -9.04237553e-02
1.08003783e+00 3.54887903e-01 1.28872305e-01 -5.17441452e-01
6.27497137e-02 -3.55806887e-01 8.35958719e-02 3.53981733e-01
-1.60669982e-01 6.69186890e-01 8.25983882e-02 -3.60312760e-01
-1.17262459e+00 -1.63550019e+00 -5.81215918e-01 1.63247120e+00
-4.77789193e-01 -5.70922852e-01 -8.01410735e-01 4.65831816e-01
-1.31410822e-01 4.93142575e-01 -2.39291936e-01 6.34995988e-03
-2.51679182e-01 -6.90892637e-01 1.15208936e+00 4.41015184e-01
1.82800308e-01 -1.29597592e+00 -9.82057571e-01 4.65923190e-01
-3.82697344e-01 -9.83359694e-01 -4.73819762e-01 2.48158813e-01
-8.30864608e-01 -6.02815568e-01 -4.65232193e-01 -7.31416941e-01
-3.34359974e-01 2.07848310e-01 1.00578272e+00 -5.36207497e-01
-9.06709254e-01 5.19045711e-01 -4.45047878e-02 -7.08934128e-01
-3.82843941e-01 -3.79392684e-01 2.09476113e-01 -1.08265102e-01
1.66459844e-01 -1.33168602e+00 -9.57927167e-01 8.70640725e-02
-1.01946926e+00 -1.76345065e-01 5.57477400e-02 5.15664756e-01
6.33236408e-01 -7.87655190e-02 6.32984877e-01 -1.88024506e-01
9.04947281e-01 -5.30149341e-01 -3.64114344e-01 -4.16894346e-01
8.98556113e-02 -1.97307840e-01 6.40329897e-01 -6.74968958e-01
-8.53683829e-01 -4.82543744e-02 -2.87598878e-01 -3.01654875e-01
-3.60151976e-01 3.76069903e-01 7.91669190e-01 -1.45129114e-01
1.30544090e+00 2.36803204e-01 -6.33922815e-02 -4.79075879e-01
2.70217866e-01 6.88851714e-01 7.92484879e-01 -5.36161602e-01
2.57098615e-01 6.81602418e-01 -1.13201320e-01 -1.16327679e+00
-5.99889338e-01 -3.83579314e-01 -3.82444948e-01 -2.60643065e-01
7.47699201e-01 -1.01006448e+00 -9.09640491e-01 3.84720683e-01
-1.12133789e+00 -4.63518351e-01 -3.71909440e-01 9.14415061e-01
-9.88561988e-01 2.67450303e-01 -9.03230250e-01 -9.23350632e-01
-4.54612553e-01 -7.70113766e-01 1.25012541e+00 -1.81431830e-01
-6.28736913e-01 -7.90881276e-01 1.57691836e-01 -1.84576392e-01
4.98916805e-01 3.94148499e-01 7.88940728e-01 -4.95177135e-02
-2.73030251e-01 1.99496850e-01 1.91828102e-01 2.13677675e-01
-6.33991957e-02 1.52893916e-01 -1.75779366e+00 -1.98413149e-01
3.68611813e-01 -5.12025356e-01 8.80738616e-01 6.75086379e-01
1.29501581e+00 -7.06584826e-02 3.41604322e-01 9.14039016e-01
1.04209685e+00 1.58422753e-01 5.00572026e-01 1.62041485e-01
1.26116918e-02 5.60478330e-01 2.41261005e-01 8.23605537e-01
-1.05222343e-02 2.84106672e-01 8.09201226e-02 1.26365170e-01
-3.78363639e-01 9.18274149e-02 5.72049797e-01 1.19092560e+00
-4.07394499e-01 4.38348204e-01 -6.03945911e-01 5.79823792e-01
-1.29512954e+00 -1.12134302e+00 -8.64550844e-02 2.31587553e+00
9.02824283e-01 -1.27698854e-01 3.18857312e-01 6.63346529e-01
6.74232900e-01 2.89984159e-02 -4.02090579e-01 -6.14148498e-01
-6.83662817e-02 9.28064167e-01 1.13610886e-01 3.11838090e-01
-1.17650890e+00 5.20252347e-01 7.26943302e+00 8.79487157e-01
-1.31649840e+00 2.50628471e-01 4.05140430e-01 -4.29401219e-01
-5.53893335e-02 -2.81443924e-01 -1.41531184e-01 1.55173987e-01
1.37417495e+00 -3.23869959e-02 1.17165363e+00 4.62809712e-01
6.36941865e-02 1.48847103e-01 -1.13163567e+00 1.11503410e+00
-5.48522770e-01 -1.23527730e+00 -4.57155138e-01 -1.24656700e-01
1.86193213e-01 3.54975641e-01 6.21911705e-01 -9.73145440e-02
6.60665780e-02 -1.09640574e+00 1.04945195e+00 3.35443765e-01
1.10639989e+00 -6.25807941e-01 -7.33497515e-02 1.52480854e-02
-1.40669203e+00 1.92598801e-03 -5.95515847e-01 -2.51350373e-01
-1.89322874e-01 4.95028406e-01 -1.02691448e+00 -2.85030995e-02
1.03794324e+00 3.64028394e-01 -1.95289180e-01 1.20400620e+00
3.66267204e-01 7.59978771e-01 -4.13663238e-01 5.04590273e-02
2.83649247e-02 4.00005728e-01 7.74077594e-01 1.52872968e+00
6.38242960e-01 3.76160699e-03 -1.93030819e-01 8.80273700e-01
3.64257127e-01 1.87812880e-01 -3.30074251e-01 -2.72167474e-01
6.67292953e-01 1.36996388e+00 -9.13386047e-01 3.74522321e-02
-1.86890811e-01 4.57128197e-01 6.99062422e-02 2.78923273e-01
-7.11662233e-01 -6.27333105e-01 9.17463005e-01 1.08983845e-01
2.25885957e-01 -4.95665133e-01 -2.77124017e-01 -7.92860627e-01
-1.95125982e-01 -7.40603983e-01 2.67789245e-01 -1.04654157e+00
-1.33576989e+00 6.24039233e-01 -2.11744249e-01 -1.60860741e+00
-2.21279159e-01 -7.99539745e-01 -3.80030811e-01 1.04441988e+00
-1.24184954e+00 -5.39946973e-01 6.54896572e-02 8.83616269e-01
4.30734843e-01 -6.97485209e-02 1.35679173e+00 7.45096654e-02
3.47478449e-01 3.29193830e-01 5.79371564e-02 -1.69402808e-01
6.78302765e-01 -1.40150774e+00 9.14418697e-01 3.16762745e-01
3.34805876e-01 8.44004154e-01 7.81674981e-01 -1.96683407e-01
-1.20210683e+00 -8.94797385e-01 5.93969107e-01 -1.16284296e-01
1.14250898e+00 -2.98800498e-01 -7.25290775e-01 1.79886892e-01
2.69059092e-01 3.64604667e-02 1.24548078e+00 -1.19604938e-01
-7.23587692e-01 2.53187381e-02 -8.56875062e-01 6.98073208e-01
9.45032239e-01 -1.14008939e+00 -4.15505260e-01 4.74362701e-01
7.23291039e-01 -3.88575315e-01 -1.39697027e+00 7.73669332e-02
8.10493708e-01 -1.35657072e+00 1.22143567e+00 -1.72836021e-01
3.58408242e-01 -3.85936201e-01 -4.40476388e-01 -1.36106253e+00
-5.19155502e-01 -1.16664541e+00 4.73515131e-02 7.59821177e-01
1.78918317e-01 -5.38594007e-01 1.61695451e-01 -1.69440687e-01
-4.52688336e-01 -2.63007611e-01 -1.29016101e+00 -6.35773361e-01
5.42416833e-02 -1.05096817e+00 5.84894419e-01 8.60920370e-01
3.01522583e-01 1.89456940e-01 1.35501862e-01 1.19316444e-01
6.92609012e-01 1.60883725e-01 2.84556776e-01 -1.20259309e+00
-8.35195184e-01 -6.25262558e-01 -5.72335601e-01 -6.30733132e-01
-1.29683405e-01 -1.03864574e+00 1.80612311e-01 -8.25891018e-01
-3.05388063e-01 -1.82644844e-01 -5.56207240e-01 1.67499527e-01
3.23965043e-01 6.08120918e-01 -1.49701219e-02 2.38932326e-01
-3.11338659e-02 8.81375149e-02 1.36915708e+00 1.09907523e-01
-3.03020813e-02 -3.45767662e-02 -3.70157361e-01 9.13002968e-01
9.36156154e-01 -2.50776827e-01 -3.54066730e-01 -4.90043372e-01
9.75301564e-02 3.46280903e-01 6.87870026e-01 -1.24082100e+00
1.15973756e-01 9.71892402e-02 5.15048265e-01 -1.89023733e-01
7.64338791e-01 -9.58943516e-02 1.68590426e-01 1.95893154e-01
-6.46602035e-01 1.11092180e-02 4.61295962e-01 4.20263350e-01
-1.53589681e-01 6.31601214e-02 9.34278190e-01 -3.61038238e-01
-5.20278573e-01 -1.43868979e-02 -9.57443297e-01 5.85638173e-02
3.03108096e-01 -1.77561760e-01 -3.94046128e-01 -3.40317696e-01
-9.55951750e-01 -8.71094346e-01 -5.80748357e-02 2.07539275e-01
8.12079370e-01 -1.39865434e+00 -7.70253181e-01 1.85671493e-01
-8.52471441e-02 -6.08255744e-01 3.53974819e-01 8.16928089e-01
-5.91712832e-01 2.41204441e-01 -4.75731909e-01 -7.14177966e-01
-1.01936030e+00 3.66415948e-01 3.51668775e-01 6.66896522e-01
-5.13818681e-01 1.24758172e+00 3.46097410e-01 -1.58964530e-01
1.98307663e-01 -4.87947136e-01 -1.87647134e-01 -6.66597113e-02
8.74679804e-01 4.81121093e-01 2.31497601e-01 -5.83108783e-01
-1.09463044e-01 2.47267008e-01 5.43985546e-01 -7.40763485e-01
1.15805602e+00 8.55495185e-02 -3.67410421e-01 1.08376706e+00
1.19245398e+00 5.67493550e-02 -1.09346545e+00 1.53032895e-02
-3.31415497e-02 -2.87567973e-01 8.36545452e-02 -5.78345060e-01
-4.45718765e-01 1.15009475e+00 7.29252577e-01 7.41045773e-01
1.35479128e+00 -2.40485653e-01 8.86556983e-01 2.33479396e-01
2.90939063e-01 -1.00101590e+00 2.60998368e-01 4.13903087e-01
1.02721238e+00 -2.83625007e-01 -2.86737174e-01 -3.55109304e-01
-2.88440049e-01 1.50398338e+00 1.26271397e-01 -3.06646913e-01
9.39536870e-01 6.98891401e-01 2.68241107e-01 7.26814047e-02
-9.39954937e-01 -3.52687202e-02 3.99869233e-01 7.39593744e-01
1.16384208e+00 2.60061353e-01 7.23549798e-02 3.70215029e-01
-1.11701095e+00 -2.65083402e-01 3.96472842e-01 8.61998022e-01
-5.85756540e-01 -6.46086454e-01 -6.54651225e-01 2.89886713e-01
-1.02989697e+00 -5.80131173e-01 -6.15396686e-02 1.89084709e-01
1.56986505e-01 1.03614795e+00 1.99589819e-01 -6.29896164e-01
-8.39412585e-02 1.78200141e-01 7.84367502e-01 -5.93034744e-01
-6.97692096e-01 8.62239718e-01 4.48156754e-03 -5.41153729e-01
-6.29786789e-01 -6.10604942e-01 -1.31446922e+00 -1.05562426e-01
-1.14675567e-01 -2.30989028e-02 8.99757445e-01 5.80331981e-01
3.73233736e-01 8.37831080e-01 5.26779234e-01 -1.22097230e+00
-6.16195142e-01 -1.13606954e+00 -8.31968844e-01 1.50970131e-01
4.66999799e-01 -4.46695209e-01 -3.92963767e-01 4.31169301e-01] | [15.631645202636719, 5.4766106605529785] |
50ac815d-d839-487d-a9cb-a3b5d71d1b59 | superinfection-and-the-hypnozoite-reservoir | 2306.14329 | null | https://arxiv.org/abs/2306.14329v1 | https://arxiv.org/pdf/2306.14329v1.pdf | Superinfection and the hypnozoite reservoir for Plasmodium vivax: a general framework | Malaria is a parasitic disease, transmitted by mosquito vectors. Plasmodium vivax presents particular challenges for disease control, in light of an undetectable reservoir of latent parasites (hypnozoites) within the host liver. Superinfection, which is driven by temporally proximate mosquito inoculation and/or hypnozoite activation events, is an important feature of P. vivax. Here, we present a model of hypnozoite accrual and superinfection for P. vivax. To couple host and vector dynamics, we construct a density-dependent Markov population process with countably many types, for which disease extinction is shown to occur almost surely. We also establish a functional law of large numbers, taking the form of an infinite-dimensional system of ordinary differential equations that can also be recovered under the hybrid approximation or a standard compartment modelling approach. Recognising that the subset of these equations that models the infection status of human hosts has precisely the same form as the Kolmogorov forward equations for a Markovian network of infinite server queues with an inhomogeneous batch arrival process, we use physical insight into the evolution of the latter to write down a time-dependent multivariate generating function for the solution. We use this characterisation to collapse the infinite-compartment model into a single integrodifferential equation (IDE) governing the intensity of mosquito-to-human transmission. Through a steady state analysis, we recover a threshold phenomenon for this IDE in terms of a bifurcation parameter $R_0$, with the disease-free equilibrium shown to be uniformly asymptotically stable if $R_0<1$ and an endemic equilibrium solution emerging if $R_0>1$. Our work provides a theoretical basis to explore the epidemiology of P. vivax, and introduces a general strategy for constructing tractable population-level models of malarial superinfection. | ['Peter G. Taylor', 'James M. McCaw', 'Somya Mehra'] | 2023-06-25 | null | null | null | null | ['epidemiology'] | ['medical'] | [-9.63017419e-02 -2.05729291e-01 3.13561380e-01 2.27784634e-01
1.84443861e-01 -7.03331590e-01 6.50914550e-01 -8.58319700e-02
-4.62711036e-01 8.62266481e-01 -3.64725053e-01 -5.25742650e-01
-2.65118390e-01 -6.68850183e-01 -3.07631165e-01 -1.46681488e+00
-9.95491028e-01 6.94991589e-01 -1.65102154e-01 -2.82279164e-01
-2.27959692e-01 7.54634380e-01 -1.15385187e+00 -4.00168329e-01
7.74152219e-01 4.75032181e-01 -2.54221074e-02 1.00478554e+00
-6.66845515e-02 3.29975724e-01 -6.82986796e-01 -1.29816532e-01
2.37741426e-01 -4.65747029e-01 -3.55769426e-01 -1.23989262e-01
-6.14644706e-01 -4.83208716e-01 -8.31916742e-03 6.39336109e-01
4.41163003e-01 -4.10822660e-01 1.23297048e+00 -1.53338456e+00
-2.70864129e-01 -8.90045390e-02 -7.42130518e-01 7.50081241e-01
1.67731300e-01 3.52358311e-01 2.83569217e-01 -5.98439336e-01
3.52788448e-01 1.22951162e+00 5.16277075e-01 6.85307264e-01
-1.51646805e+00 -2.21393079e-01 -3.64081740e-01 -2.22943485e-01
-1.79909539e+00 -3.49753529e-01 2.58004338e-01 -7.66152203e-01
9.29449201e-01 3.92043233e-01 1.00859272e+00 7.95969725e-01
9.94433045e-01 3.33202839e-01 1.01930439e+00 -2.33637542e-01
4.36470747e-01 -5.59528023e-02 1.45012259e-01 3.92448336e-01
6.85193598e-01 1.23065569e-01 4.48057465e-02 -6.19164586e-01
1.05777800e+00 2.87238628e-01 -2.85511494e-01 -1.80737332e-01
-5.34573853e-01 1.03805947e+00 -1.46780252e-01 3.12700391e-01
-6.29933059e-01 -1.07887350e-01 -6.99661672e-02 3.09046894e-01
5.44423938e-01 -1.42709628e-01 -3.59199017e-01 9.78434607e-02
-7.51858115e-01 2.00658649e-01 1.34606981e+00 4.97857243e-01
6.38537288e-01 -6.59236610e-02 5.40355779e-02 1.63547605e-01
4.45924371e-01 1.34534466e+00 -2.83058554e-01 -5.83236337e-01
-1.87861636e-01 9.36605781e-02 5.19883573e-01 -4.41297263e-01
-3.16927075e-01 -3.18171233e-01 -1.23435056e+00 -1.32407352e-01
5.61309457e-01 -4.12997246e-01 -3.50382119e-01 1.88618016e+00
2.86249220e-01 1.67525541e-02 1.56467542e-01 3.34885865e-01
-3.86775471e-02 1.26631629e+00 2.01453030e-01 -1.07279706e+00
1.38684332e+00 1.67334136e-02 -6.03419125e-01 5.45140088e-01
3.64275724e-01 -3.70444238e-01 3.37794483e-01 7.97358751e-02
-1.17742598e+00 4.05108571e-01 -4.72897917e-01 8.62501562e-01
-2.54135728e-01 -7.69083023e-01 4.45475906e-01 6.44398689e-01
-1.58787251e+00 3.79863352e-01 -1.15495300e+00 -7.51061499e-01
3.75049077e-02 4.27098960e-01 2.68956996e-03 1.91729683e-02
-1.24121177e+00 6.99743807e-01 -4.71732885e-01 4.83481020e-01
-1.17840505e+00 -6.26844823e-01 -4.83259678e-01 7.10840672e-02
-2.13329732e-01 -9.27404702e-01 9.00656760e-01 -5.82315862e-01
-1.15026808e+00 6.44552529e-01 -5.52590311e-01 -2.39041597e-01
4.84861314e-01 3.28853279e-01 -1.43468499e-01 5.82587957e-01
2.42705628e-01 -1.48358438e-02 5.67803800e-01 -1.03964162e+00
-3.52408916e-01 -5.70670485e-01 -1.25395998e-01 1.32633582e-01
-2.12951928e-01 2.93330669e-01 1.07234292e-01 -2.13852614e-01
-5.58001697e-01 -6.74695909e-01 -4.04309362e-01 -1.30814075e-01
-1.04126312e-01 -4.56572950e-01 4.52013046e-01 -4.03724968e-01
1.09908819e+00 -1.72906053e+00 1.41292512e-01 8.31485018e-02
3.15663189e-01 1.51608691e-01 -1.02735441e-02 1.08856297e+00
2.31078759e-01 3.40131730e-01 -5.00716507e-01 -1.11016616e-01
-2.75479257e-01 3.74287337e-01 -7.79014975e-02 9.85781014e-01
5.44601142e-01 5.07845581e-01 -1.04902732e+00 -4.28767413e-01
-1.58438116e-01 1.12160432e+00 -4.34851915e-01 2.35565737e-01
2.30527967e-01 5.63051999e-01 -5.34425855e-01 4.19719368e-01
9.86175060e-01 -2.95305371e-01 -5.06192297e-02 5.31054258e-01
-6.09432101e-01 -4.12600935e-01 -8.04507852e-01 5.14234245e-01
-2.23352164e-01 2.13659272e-01 6.58306360e-01 -1.08730054e+00
3.32739264e-01 6.48755133e-01 8.48326981e-01 -3.15308198e-02
1.49087176e-01 7.96907097e-02 9.69022587e-02 -4.12560672e-01
-7.24029094e-02 -5.91018558e-01 2.45739430e-01 6.62935674e-01
-7.50559717e-02 3.26685131e-01 3.73575151e-01 3.54703128e-01
1.29441679e+00 -4.88882869e-01 6.41594408e-03 -9.97347474e-01
5.70886135e-01 -3.17375781e-03 3.65823925e-01 5.90254486e-01
-3.75281274e-01 -9.89016220e-02 8.72845769e-01 -3.30198407e-01
-1.08941281e+00 -1.49262428e+00 -6.80285931e-01 4.41377044e-01
4.83402461e-02 -1.62074670e-01 -1.11781728e+00 3.79066467e-01
-1.66831508e-01 1.72786087e-01 -7.02621043e-01 -1.02881044e-01
-5.47580957e-01 -1.59333634e+00 4.60024774e-01 -2.96340913e-01
2.91707754e-01 -9.72499967e-01 -8.06846261e-01 2.68570513e-01
-6.29113689e-02 -4.86440063e-01 -8.69177505e-02 1.42153159e-01
-8.54224563e-01 -1.20784700e+00 -1.12056208e+00 -4.57965612e-01
7.89779902e-01 4.35345024e-02 1.01761353e+00 2.42432818e-01
-6.14798605e-01 7.87617624e-01 2.89526343e-01 -8.64065439e-03
-6.48390114e-01 -4.87803042e-01 4.87166196e-01 9.75813568e-02
5.10610580e-01 -4.42107558e-01 -1.11134636e+00 3.46085399e-01
-1.22662568e+00 -3.63737434e-01 1.26583830e-01 4.52934682e-01
2.39778966e-01 2.75088370e-01 4.73588675e-01 -5.54108135e-02
6.56857669e-01 -9.91761446e-01 -7.64006078e-01 2.90939808e-01
-2.57400513e-01 -2.25988589e-02 7.02918649e-01 -4.54787612e-01
-9.26767945e-01 -1.46276072e-01 2.21390188e-01 1.80965573e-01
-1.08332727e-02 3.59309852e-01 3.46236974e-02 1.70890778e-01
1.13250785e-01 6.28778160e-01 3.25862855e-01 -4.64991719e-01
9.74230394e-02 7.38279223e-01 2.37170130e-01 -4.50287879e-01
7.53991246e-01 9.22105253e-01 4.81395841e-01 -1.61614466e+00
1.05703562e-01 -4.21715081e-01 -4.82630134e-01 -3.05623323e-01
8.23086679e-01 -7.31623709e-01 -1.44112384e+00 8.34450424e-01
-1.31415260e+00 -5.20564675e-01 -3.27875018e-01 2.64782041e-01
-5.56186140e-01 2.14546248e-01 -1.17546630e+00 -1.57343066e+00
-1.51819170e-01 -8.37711036e-01 7.51483142e-01 1.64430052e-01
2.29991511e-01 -1.51808715e+00 7.32490122e-01 -3.67038280e-01
9.20788467e-01 3.46549809e-01 8.54875326e-01 -3.47258687e-01
-5.69545090e-01 -1.91469342e-01 -2.50188857e-02 -1.70982912e-01
1.59429342e-01 5.02911031e-01 -3.67793113e-01 -6.01310551e-01
6.47744060e-01 4.11295414e-01 4.31957632e-01 9.13758516e-01
-2.30962813e-01 -6.88732922e-01 -4.74692851e-01 4.12569314e-01
1.61602044e+00 2.79425949e-01 2.13262871e-01 -4.17949222e-02
9.65339765e-02 1.00384533e+00 -1.81026295e-01 8.18117082e-01
4.10044730e-01 2.34653220e-01 1.78199545e-01 -8.92232172e-03
7.39478290e-01 1.96429655e-01 6.56338930e-01 6.55515254e-01
-3.02685022e-01 -2.89358020e-01 -1.08057392e+00 4.15527523e-01
-1.29258418e+00 -1.06549990e+00 -3.53024989e-01 2.72010946e+00
8.40044141e-01 -3.35047156e-01 6.14762366e-01 -4.03502315e-01
8.80728185e-01 -5.66901803e-01 -3.19113106e-01 -2.36442149e-01
-2.81749755e-01 3.54012787e-01 6.78604007e-01 1.03292334e+00
-5.25234640e-01 4.81961817e-01 7.04582310e+00 2.51302391e-01
-1.10821402e+00 2.28611767e-01 5.72934568e-01 1.46322682e-01
-1.82811022e-01 1.25650153e-01 -8.80425155e-01 5.84174216e-01
1.49982500e+00 -7.08898664e-01 4.01890934e-01 1.04760580e-01
8.08703601e-01 -4.09604669e-01 -9.14943159e-01 5.34266591e-01
-1.73256293e-01 -8.04853082e-01 -6.22611195e-02 8.00380588e-01
3.93010199e-01 -9.37213898e-02 -5.77004403e-02 -1.34114459e-01
4.41351593e-01 -9.05543149e-01 -1.26292989e-01 4.26570684e-01
7.01730013e-01 -7.34909236e-01 6.13203764e-01 6.09801352e-01
-1.33553755e+00 2.95341551e-01 -3.74354482e-01 -4.16974157e-01
6.76622093e-01 4.67845529e-01 -3.06755871e-01 1.39746934e-01
4.59325939e-01 4.46877629e-01 -2.44424090e-01 1.09411335e+00
2.87769079e-01 4.04174477e-01 -9.82969701e-01 -1.60593856e-02
-1.01201154e-01 -6.33487642e-01 7.35076845e-01 1.09117210e+00
2.72284865e-01 4.33584362e-01 -3.84770662e-01 1.09058785e+00
6.76132858e-01 -1.30717278e-01 -7.48841345e-01 1.14977015e-02
1.30589142e-01 1.10011971e+00 -9.81931567e-01 -3.77793193e-01
-6.48010671e-02 1.03994966e+00 -2.78731376e-01 7.00684071e-01
-4.74687338e-01 -2.84331203e-01 8.28589797e-01 3.00588846e-01
1.94696888e-01 -2.82044351e-01 4.90083724e-01 -1.47027421e+00
-2.39121497e-01 -6.61609694e-02 1.96604788e-01 -1.90836102e-01
-1.33889747e+00 8.94269824e-01 4.69110012e-01 -7.37421453e-01
-6.83528900e-01 -4.88849431e-01 -9.48239982e-01 1.19569910e+00
-1.38176751e+00 -7.10833549e-01 3.67150664e-01 8.93404603e-01
-7.50977993e-02 3.34749609e-01 8.88433158e-01 1.31079286e-01
-8.50013673e-01 -4.03116755e-02 5.88082850e-01 -2.54815876e-01
-1.20716408e-01 -1.08162880e+00 1.36192158e-01 8.73966277e-01
-7.77918041e-01 1.01380730e+00 1.12937260e+00 -8.12001348e-01
-1.75069451e+00 -7.25577235e-01 1.28184569e+00 -5.68015993e-01
1.00237119e+00 -5.07267773e-01 -7.39840508e-01 2.85856396e-01
4.12646860e-01 -1.46768406e-01 7.17724800e-01 -6.76187873e-01
3.61637563e-01 2.09230155e-01 -1.30007887e+00 4.87144262e-01
3.85700643e-01 -6.49640933e-02 -1.57609597e-01 6.59641027e-01
2.10603967e-01 5.92789710e-01 -7.81773388e-01 8.04878026e-02
3.80065203e-01 -7.82961011e-01 9.34448421e-01 -5.68723798e-01
-2.61350013e-02 -2.26056233e-01 1.50026023e-01 -9.86076593e-01
-1.12278067e-01 -1.16763568e+00 1.22163743e-01 1.06248093e+00
2.41171084e-02 -9.60222363e-01 2.85285711e-01 4.63130921e-01
6.64048612e-01 -6.53464198e-01 -1.37409317e+00 -1.14428496e+00
6.99272513e-01 2.33396307e-01 1.05198890e-01 5.22612691e-01
1.00112811e-01 2.68897921e-01 -2.11579040e-01 2.33497620e-01
1.27981651e+00 -4.78127539e-01 2.28363127e-01 -1.29117429e+00
-1.88937172e-01 -2.92688698e-01 -8.06551278e-02 -7.53119349e-01
-4.17683646e-02 -3.23695302e-01 2.77002044e-02 -1.30261815e+00
4.36629921e-01 -2.81884015e-01 -5.95615916e-02 -2.54550129e-01
2.19844922e-01 1.70831665e-01 -1.59893870e-01 6.05780602e-01
-2.40689248e-01 3.97723764e-01 6.91462159e-01 3.32502127e-01
-4.93672043e-01 2.77316749e-01 -1.08341046e-01 4.41453367e-01
7.26725459e-01 -4.71677601e-01 -2.07598746e-01 4.68065180e-02
3.09065312e-01 5.15035510e-01 4.68822867e-01 -4.76404637e-01
9.18739438e-02 -2.39548594e-01 -5.41457720e-02 -4.74164993e-01
3.34584415e-01 -7.40384936e-01 5.34398854e-01 1.34656870e+00
2.91728586e-01 4.73297119e-01 8.98596719e-02 4.78175491e-01
3.20548445e-01 -1.38763592e-01 7.11252213e-01 -1.40817061e-01
4.08790767e-01 5.41386902e-01 -1.46843708e+00 1.16051912e-01
1.13207388e+00 -1.50178865e-01 -2.95080721e-01 -3.50237340e-01
-7.62578130e-01 4.13415223e-01 6.26049817e-01 -3.58479679e-01
3.81917179e-01 -8.46399009e-01 -9.18032050e-01 3.17329615e-01
-5.28119624e-01 -3.40509534e-01 2.87027419e-01 1.07578421e+00
-8.68555665e-01 5.56731939e-01 -1.97218955e-01 -8.18691730e-01
-8.59734952e-01 6.82332873e-01 5.47992587e-01 -3.15428138e-01
-1.25720099e-01 4.07126456e-01 4.32798326e-01 2.32017979e-01
-1.09352082e-01 1.56964362e-01 -1.71905071e-01 4.25956473e-02
8.87183249e-01 3.48936021e-01 -5.27480364e-01 -8.36396754e-01
-5.42772651e-01 5.30680060e-01 1.87770337e-01 -3.74311209e-01
1.36645949e+00 -4.55905467e-01 -8.83818150e-01 5.73869348e-01
9.60806191e-01 -3.37450117e-01 -1.39745486e+00 2.62390435e-01
-1.98282257e-01 1.75025817e-02 -5.32533407e-01 -2.53731102e-01
-7.78094113e-01 9.80642855e-01 4.92334813e-01 9.21862841e-01
8.17937076e-01 3.32154781e-01 2.99147964e-01 -4.86427359e-02
2.66422987e-01 -2.53811151e-01 -3.09014738e-01 2.71793753e-01
6.34397805e-01 -8.37086439e-01 -4.18435097e-01 -2.92440057e-01
1.47374988e-01 9.24029827e-01 7.27866963e-02 -2.92969972e-01
1.18317127e+00 6.65653825e-01 -2.01914057e-01 -2.24701896e-01
-9.08948302e-01 1.16376415e-01 -4.50865239e-01 9.24398720e-01
3.43939632e-01 1.95111543e-01 -6.80053830e-01 2.19615728e-01
1.76408708e-01 1.13703519e-01 8.04840565e-01 9.34873164e-01
-3.53135079e-01 -9.80563164e-01 -4.64320034e-01 1.53621510e-01
-7.11490691e-01 -3.90549034e-01 -6.00035526e-02 7.17665017e-01
-3.95540684e-01 1.02037537e+00 5.98830998e-01 5.53367794e-01
-2.14849696e-01 3.40086408e-02 4.25676912e-01 -2.93609738e-01
-3.07625979e-01 3.53579342e-01 -8.23154688e-01 -1.75711680e-02
-4.85930711e-01 -6.99705243e-01 -8.16699445e-01 -1.00742650e+00
-2.96232611e-01 7.79553950e-01 5.01640975e-01 8.17628801e-01
5.02122007e-02 -6.56888336e-02 9.21930313e-01 -8.13882828e-01
-5.33556402e-01 -7.54704058e-01 -1.09876013e+00 -5.01144230e-02
6.58275783e-01 -2.36424848e-01 -1.12014389e+00 8.16317052e-02] | [5.931520462036133, 4.33725643157959] |
fdf83b2c-4cea-4c5f-9fd7-a1d037f5f6d7 | a-novel-feature-selection-and-extraction | 1412.7934 | null | http://arxiv.org/abs/1412.7934v1 | http://arxiv.org/pdf/1412.7934v1.pdf | A Novel Feature Selection and Extraction Technique for Classification | This paper presents a versatile technique for the purpose of feature
selection and extraction - Class Dependent Features (CDFs). We use CDFs to
improve the accuracy of classification and at the same time control
computational expense by tackling the curse of dimensionality. In order to
demonstrate the generality of this technique, it is applied to handwritten
digit recognition and text categorization. | ['Raunaq Vohra', 'Kratarth Goel', 'Ainesh Bakshi'] | 2014-12-26 | null | null | null | null | ['handwritten-digit-recognition'] | ['computer-vision'] | [ 1.71339840e-01 -7.84980059e-01 -2.62261748e-01 -7.35875487e-01
-2.22041830e-01 -5.61189711e-01 5.92696249e-01 2.27497771e-01
-3.55254710e-01 7.48967886e-01 -3.61594319e-01 -3.70577633e-01
-6.56127095e-01 -8.21435273e-01 2.20254585e-01 -8.55440795e-01
-2.01277733e-01 8.91986489e-02 7.25133345e-02 9.91155356e-02
9.00647223e-01 1.19595897e+00 -2.03693438e+00 2.10667238e-01
5.25176406e-01 1.24245036e+00 -7.33616054e-02 8.08127642e-01
-1.48724213e-01 6.50210500e-01 -8.57129455e-01 1.01980597e-01
1.79450661e-01 -1.13696784e-01 -5.83269596e-01 5.45583606e-01
-6.93094358e-02 -2.32342154e-01 -2.10614413e-01 6.30737841e-01
4.43400055e-01 2.85653651e-01 1.26974368e+00 -1.27668989e+00
-4.08010542e-01 4.58875336e-02 -3.13028544e-01 3.56924415e-01
1.62451491e-01 -4.73769993e-01 6.69373274e-01 -8.98568630e-01
3.54303271e-01 1.03076041e+00 3.78421247e-01 3.09721500e-01
-9.10376489e-01 -3.10921907e-01 -5.63959740e-02 4.64801677e-02
-1.43346453e+00 -4.28413600e-01 9.82664466e-01 -5.42215109e-01
7.67838120e-01 5.47787130e-01 4.37427387e-02 6.52674139e-01
5.07715404e-01 8.09876919e-01 9.85599697e-01 -7.22049117e-01
3.76643777e-01 1.94232509e-01 8.35331798e-01 4.92362618e-01
4.30530578e-01 6.63802549e-02 -4.88419235e-01 -2.37177253e-01
5.54525912e-01 -5.88490143e-02 -6.34491518e-02 -3.84402454e-01
-7.58636892e-01 1.11470914e+00 -9.03088972e-02 6.90540910e-01
-1.87333390e-01 -3.29801857e-01 3.59923750e-01 5.14870822e-01
3.33791167e-01 4.63439643e-01 -4.54536945e-01 -6.57657683e-02
-6.64771080e-01 2.05851048e-01 1.00596118e+00 7.96102345e-01
3.06956738e-01 5.12084886e-02 -1.27237707e-01 7.19281316e-01
8.16134661e-02 6.04270339e-01 6.61100864e-01 -3.00398588e-01
2.76173055e-01 9.18254852e-01 1.40077367e-01 -7.76131511e-01
-5.00779450e-01 -1.75827935e-01 -7.17117488e-01 6.62158370e-01
2.37395465e-01 2.29091972e-01 -1.10930896e+00 8.44171345e-01
2.23006874e-01 -7.86811709e-01 6.31977012e-03 3.83064628e-01
3.88772607e-01 4.32509392e-01 1.73310973e-02 -1.57667398e-01
1.05280459e+00 -3.61956745e-01 -6.59488082e-01 2.76606619e-01
3.75709325e-01 -8.35263014e-01 8.06985617e-01 6.96930647e-01
-2.94888139e-01 -6.19041145e-01 -1.08193207e+00 2.15962455e-01
-7.40476429e-01 5.68967044e-01 8.45757425e-01 1.01867342e+00
-2.93829769e-01 5.37669957e-01 -8.22797298e-01 3.21693271e-02
2.94857115e-01 7.51677632e-01 -4.96743172e-01 1.24639347e-01
-7.40328014e-01 7.68567145e-01 3.81529361e-01 -1.84017152e-01
-2.35352367e-01 -2.72912621e-01 -6.77156627e-01 4.16861661e-02
-1.56395733e-01 1.16465323e-01 9.72212493e-01 -7.68424213e-01
-1.37524474e+00 5.93046129e-01 2.07332727e-02 -8.96804780e-02
3.43727201e-01 -5.62299900e-02 -6.16728008e-01 7.63134435e-02
-1.70592695e-01 -6.00630268e-02 1.28149438e+00 -5.00997126e-01
-8.50865901e-01 -5.19993901e-01 -5.97398400e-01 -1.29789665e-01
-6.39989674e-01 6.79930374e-02 1.68325961e-01 -8.59730005e-01
2.02414617e-01 -3.93569529e-01 1.23508647e-01 -8.74934569e-02
-1.19953014e-01 -4.57172960e-01 1.18754637e+00 -2.88238108e-01
1.14127541e+00 -2.44476271e+00 1.67221069e-01 6.99089170e-01
-3.55577134e-02 4.01018828e-01 1.75667673e-01 5.08550525e-01
-5.22816554e-02 -2.12845102e-01 -1.58853680e-01 8.68768170e-02
-1.34144634e-01 3.16265166e-01 -3.24640989e-01 5.51041782e-01
6.17227793e-01 3.75087947e-01 -3.65661442e-01 -3.96824539e-01
6.48451626e-01 3.86247545e-01 -8.90089665e-03 -3.40507482e-03
2.07077280e-01 -1.10210769e-01 -8.26431274e-01 7.93675065e-01
6.23306096e-01 1.54378146e-01 6.09097118e-03 -1.35870054e-01
-3.32073152e-01 9.08667222e-02 -1.50972843e+00 6.78032875e-01
-2.49715343e-01 6.71531618e-01 -5.10060191e-01 -1.04914570e+00
1.38857746e+00 -3.23587246e-02 4.43639487e-01 -4.48523760e-01
5.32579601e-01 3.01921695e-01 -1.43460035e-01 -3.98151219e-01
4.18691635e-01 4.90188673e-02 -1.89988807e-01 3.22705597e-01
2.59026229e-01 -9.03547630e-02 2.35937744e-01 -3.04744691e-01
5.87494075e-01 -3.63214582e-01 7.17270076e-01 -6.84204936e-01
9.11513448e-01 -1.17960289e-01 -1.60584718e-01 4.11208451e-01
5.00527173e-02 1.90860465e-01 4.90301490e-01 -7.29165435e-01
-8.09554279e-01 -7.41341889e-01 -4.76142883e-01 8.66663933e-01
-2.20680699e-01 8.82323012e-02 -2.96082020e-01 -7.24514306e-01
5.34389675e-01 7.52283633e-01 -7.65188456e-01 -1.01819053e-01
-5.00290155e-01 -8.87571514e-01 4.27079856e-01 6.28794730e-01
2.97331840e-01 -6.41874850e-01 -7.96526968e-01 2.29852825e-01
5.51130891e-01 -7.34405518e-01 4.89501050e-03 6.61604822e-01
-8.98436666e-01 -1.34757710e+00 -8.32778335e-01 -7.37860501e-01
6.56986535e-01 -3.21354270e-02 6.08699441e-01 -1.33838326e-01
-7.07872987e-01 4.95615780e-01 -5.41244507e-01 -3.49817246e-01
-3.97958249e-01 3.83358270e-01 -5.73493615e-02 1.15684845e-01
5.54075122e-01 -3.48314285e-01 -7.12859407e-02 5.01008987e-01
-9.03711975e-01 -9.72695827e-01 3.23048234e-01 8.76082361e-01
3.50600570e-01 6.03188038e-01 4.37073797e-01 -6.90484345e-01
1.12655199e+00 -8.64443257e-02 -1.03163791e+00 4.12496507e-01
-7.88620949e-01 1.33224532e-01 6.60400510e-01 -5.55607796e-01
-6.95984721e-01 4.38812464e-01 3.51006463e-02 -4.27250564e-02
-1.09290265e-01 3.37829322e-01 4.97386307e-02 -5.86829543e-01
8.10646355e-01 5.02564788e-01 1.58927009e-01 -7.45834649e-01
-1.46700079e-02 1.37384331e+00 3.56569082e-01 -2.72123426e-01
2.94415712e-01 1.57782093e-01 5.79108834e-01 -9.26667929e-01
-3.75908047e-01 -4.84852046e-01 -9.32602108e-01 -2.60257591e-02
1.62644029e-01 -3.58797371e-01 -6.32296920e-01 7.01087654e-01
-5.95188975e-01 1.92469642e-01 -1.48357064e-01 5.52457631e-01
-3.98944855e-01 2.30991781e-01 -1.75562352e-01 -1.05443287e+00
-2.49474108e-01 -8.05681646e-01 7.29609132e-01 1.04268432e-01
-1.09802070e-03 -8.32311749e-01 -3.14888269e-01 -4.71970081e-01
1.83105573e-01 3.00922334e-01 8.80895793e-01 -9.02063131e-01
-7.50223398e-02 -9.49310720e-01 -3.70306857e-02 6.15278304e-01
5.21268964e-01 3.64980936e-01 -9.35195744e-01 -1.89194188e-01
1.55674800e-01 1.48095697e-01 7.20550954e-01 2.62232840e-01
1.20617568e+00 2.13918999e-01 -2.55900055e-01 2.25575417e-01
1.25786221e+00 5.62436640e-01 5.50277293e-01 4.25578296e-01
6.60881624e-02 4.18469012e-01 7.70947456e-01 7.66169906e-01
-2.49693274e-01 5.65324426e-01 -1.61174834e-01 3.80508780e-01
-1.89570293e-01 2.66303539e-01 -3.17288786e-01 3.95660996e-01
1.62798181e-01 -3.87723088e-01 -9.74617839e-01 3.74679685e-01
-1.42093313e+00 -4.90601838e-01 -1.36391222e-01 1.89053059e+00
3.55905473e-01 1.12471543e-01 2.66365141e-01 1.14008522e+00
7.60280073e-01 -3.21855009e-01 -3.08041126e-01 -5.57533145e-01
-1.04235311e-03 2.96032697e-01 6.34527028e-01 5.43371677e-01
-1.49525034e+00 4.17022258e-01 7.39978170e+00 1.02104378e+00
-1.40596521e+00 -5.27522683e-01 4.68895793e-01 4.40380305e-01
4.24302071e-01 -4.84484792e-01 -1.07842290e+00 3.14272702e-01
3.62078011e-01 -4.60309088e-02 1.32385790e-01 9.56862569e-01
-4.27677989e-01 -3.99210036e-01 -9.51424778e-01 1.00253069e+00
1.17271386e-01 -8.73267710e-01 2.26833552e-01 -1.11390539e-01
2.29698896e-01 -4.96991128e-01 2.52232254e-01 -1.31358299e-02
-4.22862828e-01 -7.51019120e-01 3.84269834e-01 4.27942425e-01
7.63015747e-01 -1.09279156e+00 7.32784092e-01 1.54346481e-01
-9.22512889e-01 -4.46650743e-01 -3.09799731e-01 -2.92988092e-01
-2.32368618e-01 7.47391164e-01 -7.03453004e-01 4.31120396e-01
1.03212602e-01 4.35018599e-01 -5.11742592e-01 1.14673805e+00
1.73370689e-01 3.91444340e-02 -3.45808774e-01 -6.82316720e-01
1.92548707e-01 2.83511460e-01 4.14665699e-01 1.29494941e+00
2.78087825e-01 2.36205921e-01 -1.13670528e-01 3.59409660e-01
3.62594426e-01 2.46259838e-01 -4.19100791e-01 -3.40652257e-01
4.71664280e-01 8.61574471e-01 -1.19244802e+00 -1.22746259e-01
9.78877395e-02 7.99691558e-01 2.81314626e-02 -9.67301875e-02
-2.33748958e-01 -1.16619229e+00 2.97908992e-01 -1.37298733e-01
7.33565569e-01 -6.06071770e-01 -6.21278763e-01 -1.00912249e+00
3.36715542e-02 -5.63843310e-01 4.31758314e-01 9.57567431e-03
-1.34473574e+00 6.66806340e-01 4.50550765e-03 -1.34536374e+00
-4.17998135e-01 -1.29056859e+00 -2.02064067e-01 7.88294911e-01
-1.45059860e+00 -7.37972438e-01 -1.43064678e-01 8.20622921e-01
4.94662255e-01 -7.04572022e-01 1.04450691e+00 3.79189700e-01
-3.68131727e-01 6.41675353e-01 5.02547801e-01 9.53739434e-02
4.53707933e-01 -9.93166447e-01 3.28097165e-01 4.66282368e-01
-5.83949722e-02 3.65293145e-01 6.79844975e-01 -4.23027068e-01
-1.30879498e+00 -6.64819062e-01 9.94495273e-01 -2.00622484e-01
4.39437509e-01 -3.22448283e-01 -5.48616350e-01 1.57836005e-02
-7.52772987e-01 -4.92828600e-02 6.18141890e-01 8.11748859e-03
-3.85499477e-01 -3.93446326e-01 -1.45656919e+00 1.50972800e-02
3.23417962e-01 -2.61810213e-01 -5.51358104e-01 2.05545112e-01
-6.42675608e-02 -2.32086748e-01 -9.80845988e-01 3.65750134e-01
6.81861579e-01 -7.77312815e-01 8.19093347e-01 -4.05333549e-01
-9.67367291e-02 -8.81759822e-02 -3.17183673e-01 -7.54792571e-01
-2.49079376e-01 -1.90565780e-01 3.80244330e-02 1.21915543e+00
2.39794239e-01 -7.84382045e-01 4.59171921e-01 6.01398945e-01
3.57797861e-01 -6.76905036e-01 -1.15097392e+00 -1.10504186e+00
-1.90940425e-01 -2.61171818e-01 6.66691124e-01 6.85575247e-01
1.14385467e-02 4.60648052e-02 -2.23513469e-01 -1.51358977e-01
6.18060172e-01 3.38695109e-01 3.13077003e-01 -1.42705476e+00
-3.24177071e-02 -4.31597680e-01 -1.02713692e+00 -4.85262126e-01
-9.43056047e-02 -3.94200712e-01 -1.36729002e-01 -6.78363979e-01
-3.46634775e-01 -5.88223696e-01 -4.36310530e-01 3.34065825e-01
8.98813456e-02 1.10451363e-01 6.78195208e-02 1.76131681e-01
-1.06216714e-01 1.52132660e-01 7.12102294e-01 7.25074559e-02
-1.62228584e-01 5.56501031e-01 -4.50023293e-01 5.11733890e-01
9.06995714e-01 -6.98692918e-01 -3.84188116e-01 -2.88474262e-02
-3.31031531e-01 -1.22910254e-01 2.27082223e-01 -9.14769650e-01
4.98828199e-03 8.06406792e-03 9.64220941e-01 -5.84168077e-01
1.15432635e-01 -1.00069904e+00 -2.21557721e-01 8.08075368e-01
-3.31068695e-01 -9.45747197e-02 9.53321680e-02 5.34572899e-01
-3.97305936e-01 -6.98060691e-01 8.39799881e-01 2.31515035e-01
-6.95441246e-01 -2.27163389e-01 -6.40588105e-01 -6.33045018e-01
1.13725650e+00 -2.32626170e-01 -4.52825204e-02 7.12384656e-02
-5.65277219e-01 -2.59813607e-01 3.35599780e-02 5.33626676e-01
5.96082449e-01 -8.65762651e-01 -2.81416595e-01 9.17808354e-01
7.72441179e-02 -7.50388741e-01 -2.72050768e-01 2.79031128e-01
-5.93813300e-01 6.28470957e-01 -6.03829861e-01 -4.55481857e-01
-1.52696574e+00 6.21872067e-01 2.05206066e-01 -1.02223572e-03
-4.65077370e-01 8.04933369e-01 -6.47716403e-01 3.17263342e-02
3.71481329e-01 -2.48188540e-01 -5.50182998e-01 2.13532820e-01
7.26110220e-01 5.94866753e-01 4.91258770e-01 -2.20800400e-01
-6.44857168e-01 5.38535893e-01 -6.79998174e-02 1.59194231e-01
1.29181671e+00 1.38731394e-02 -5.54080643e-02 5.04965603e-01
1.18852317e+00 -1.19285002e-01 -8.82346928e-01 2.28807300e-01
3.90996307e-01 -7.86922574e-01 8.82260874e-02 -7.71216333e-01
-7.07333386e-01 7.64347196e-01 8.00205529e-01 5.25933802e-01
1.25464511e+00 -4.58712906e-01 8.07920471e-02 8.33476007e-01
2.71843880e-01 -1.00354111e+00 -4.33352500e-01 3.54138643e-01
5.61285734e-01 -7.67347157e-01 3.72846901e-01 -5.69886088e-01
-5.58195591e-01 1.98872662e+00 -1.23473052e-02 -5.17739773e-01
1.04334235e+00 6.76063716e-01 -2.45465428e-01 1.23818643e-01
-4.10453081e-01 -1.24973118e-01 5.24334371e-01 6.96570218e-01
3.83069903e-01 1.08821746e-02 -8.66050839e-01 4.52780753e-01
1.40220508e-01 1.60990492e-01 3.17149818e-01 1.29632628e+00
-6.47490323e-01 -1.40608537e+00 -3.31181675e-01 6.93263471e-01
-4.34901148e-01 2.87257880e-01 -5.13635576e-01 9.25009727e-01
-1.05591081e-01 8.50993991e-01 -9.99269634e-02 -3.21085542e-01
3.65911186e-01 7.14047253e-02 6.50623322e-01 -1.58940956e-01
-3.69502366e-01 -1.17652133e-01 1.06430411e-01 1.05577707e-03
-2.92517006e-01 -7.24619627e-01 -7.84957349e-01 -1.03535421e-01
-8.64789009e-01 2.80905187e-01 1.11375737e+00 7.86451161e-01
1.49779141e-01 3.18704844e-01 1.08415329e+00 -5.09072423e-01
-9.35833693e-01 -8.49081993e-01 -1.08331215e+00 8.74396116e-02
4.70254242e-01 -9.85564590e-01 -3.39653224e-01 -2.71761813e-03] | [8.232643127441406, 4.205712795257568] |
a0d6d95a-9ba2-49e8-99f7-87bda66f45d6 | semi-supervised-convolutive-nmf-for-automatic | 2202.04989 | null | https://arxiv.org/abs/2202.04989v2 | https://arxiv.org/pdf/2202.04989v2.pdf | Semi-Supervised Convolutive NMF for Automatic Piano Transcription | Automatic Music Transcription, which consists in transforming an audio recording of a musical performance into symbolic format, remains a difficult Music Information Retrieval task. In this work, which focuses on piano transcription, we propose a semi-supervised approach using low-rank matrix factorization techniques, in particular Convolutive Nonnegative Matrix Factorization. In the semi-supervised setting, only a single recording of each individual notes is required. We show on the MAPS dataset that the proposed semi-supervised CNMF method performs better than state-of-the-art low-rank factorization techniques and a little worse than supervised deep learning state-of-the-art methods, while however suffering from generalization issues. | ['Jérémy E. Cohen', 'Axel Marmoret', 'Haoran Wu'] | 2022-02-10 | null | null | null | null | ['music-transcription', 'music-information-retrieval'] | ['music', 'music'] | [ 2.61406362e-01 -2.74640381e-01 -7.11214244e-02 4.94394712e-02
-1.03106534e+00 -9.94225383e-01 2.39936709e-01 -1.47966251e-01
-3.61964703e-01 5.28833926e-01 2.71126598e-01 -1.17198616e-01
-5.94891608e-01 -3.95153821e-01 -5.80147505e-01 -7.04411149e-01
-1.53476149e-02 6.06590450e-01 -3.96828026e-01 -1.90210044e-01
1.00775398e-01 2.54018158e-01 -1.57860720e+00 7.03394353e-01
4.70750064e-01 9.15473759e-01 -5.60638756e-02 8.16362739e-01
1.27200425e-01 9.14073229e-01 -5.58554471e-01 -4.49390948e-01
2.65351146e-01 -6.07525945e-01 -7.47556508e-01 8.77565797e-03
6.47120178e-01 7.72550479e-02 -3.69390130e-01 9.71180797e-01
4.87143785e-01 4.20933098e-01 5.94917476e-01 -1.01381624e+00
-5.46559513e-01 1.08151245e+00 -2.92826951e-01 -2.83223223e-02
4.04742181e-01 -5.45480072e-01 1.55536211e+00 -9.04806614e-01
5.06269574e-01 8.43316138e-01 8.44786286e-01 2.28822246e-01
-1.61647666e+00 -9.18199062e-01 -1.42008036e-01 3.49963129e-01
-1.52378452e+00 -4.45332468e-01 8.92047882e-01 -7.41303742e-01
8.29754293e-01 3.63495976e-01 6.84173048e-01 8.53009939e-01
-9.39729065e-02 8.86821032e-01 1.03848374e+00 -5.16033232e-01
1.12552553e-01 -4.95222807e-01 -4.25545163e-02 3.33920389e-01
-2.74790376e-01 -1.20720811e-01 -1.16040361e+00 -3.98357302e-01
8.93515587e-01 5.44882752e-02 -2.06362158e-01 -3.64728332e-01
-1.64164770e+00 6.31614566e-01 1.02686949e-01 6.17414415e-01
-5.85965931e-01 3.47734034e-01 6.67023778e-01 6.06252134e-01
1.56781644e-01 5.32814324e-01 -3.88544947e-01 -6.47774816e-01
-1.53147781e+00 3.55398059e-01 6.95952356e-01 5.54261088e-01
5.88051260e-01 1.57147974e-01 -1.65429488e-01 8.17130506e-01
-1.21491710e-02 3.29029232e-01 5.79424977e-01 -1.04274356e+00
5.10314524e-01 2.97028065e-01 5.07472083e-02 -8.61466348e-01
-4.13239092e-01 -8.79168272e-01 -1.17937422e+00 -2.96561103e-02
4.15520668e-01 2.41435654e-02 -5.30976832e-01 1.69639921e+00
-2.21111611e-01 5.52826166e-01 1.22537062e-01 9.14091527e-01
6.46294534e-01 5.75562119e-01 -4.36876565e-01 -7.32102573e-01
1.03455639e+00 -1.05797410e+00 -9.26750302e-01 3.32470536e-01
1.93244994e-01 -1.18468702e+00 1.10176039e+00 1.27545547e+00
-1.07965636e+00 -7.48458803e-01 -1.01054871e+00 -4.89509515e-02
2.15789542e-01 7.91285992e-01 8.62539828e-01 5.20623088e-01
-7.66599953e-01 9.79670763e-01 -6.95579886e-01 9.71980691e-02
4.46356833e-03 4.16746855e-01 -6.40736461e-01 1.25936463e-01
-9.58242595e-01 2.26650517e-02 1.55969575e-01 1.92342445e-01
-9.09375250e-01 -6.81932330e-01 -4.70213860e-01 2.41464436e-01
3.71386558e-01 -6.91457868e-01 1.31255579e+00 -9.77730215e-01
-1.82307088e+00 6.10048234e-01 -2.56353706e-01 -4.63078022e-01
2.85589486e-01 -7.00814009e-01 -2.82229990e-01 2.14151964e-01
-3.62296067e-02 1.47226140e-01 1.28699994e+00 -9.60891366e-01
-2.48038828e-01 -4.36221212e-01 7.79748112e-02 1.25172213e-01
-3.38959396e-01 -2.36639846e-02 -4.72065955e-01 -1.19855905e+00
3.79824996e-01 -1.45790136e+00 -2.17618674e-01 -4.96390492e-01
-4.25488681e-01 -1.18169986e-01 4.36327547e-01 -6.00836396e-01
1.56499457e+00 -2.33805656e+00 7.81825364e-01 3.15171093e-01
6.83193654e-02 1.11580551e-01 -7.12856650e-02 5.68212688e-01
-5.17771363e-01 -4.70579445e-01 -9.31529179e-02 -6.31911397e-01
2.30226181e-02 7.59355128e-02 -7.43713081e-01 3.97680461e-01
-2.73736507e-01 6.58915401e-01 -9.13146853e-01 -1.11631453e-01
-4.16177846e-02 6.27177596e-01 -7.37704039e-01 7.21538439e-02
4.68882993e-02 6.90031588e-01 9.76369008e-02 4.80157405e-01
2.01304853e-01 3.57008539e-02 2.12474912e-01 -4.63793874e-01
-1.27653614e-01 4.89281595e-01 -1.60936785e+00 2.61800385e+00
-5.71606338e-01 6.19307935e-01 8.94498900e-02 -9.72550213e-01
7.70318329e-01 6.72277510e-01 6.99449599e-01 -8.97232965e-02
-1.75445918e-02 3.25222671e-01 6.93354979e-02 -9.24349800e-02
6.88170493e-01 -4.18327719e-01 -1.26297787e-01 8.80567849e-01
3.99124563e-01 5.39019629e-02 5.32744527e-01 2.30662301e-01
1.18399692e+00 1.80708855e-01 1.66338637e-01 -6.16936870e-02
4.74610955e-01 -3.69057208e-01 5.12406528e-01 6.20129228e-01
3.60833347e-01 9.24990475e-01 3.67880642e-01 -2.94169575e-01
-6.70495689e-01 -1.08882761e+00 1.32663563e-01 1.36780572e+00
-5.12548804e-01 -1.16521966e+00 -6.42770827e-01 -3.40624809e-01
-4.62022498e-02 2.57636189e-01 -5.32601058e-01 1.11540563e-01
-5.05722165e-01 -4.85813409e-01 7.85908163e-01 5.93417108e-01
8.61000195e-02 -9.37276423e-01 -2.03301474e-01 4.01898801e-01
-3.71154517e-01 -8.65069628e-01 -5.14739513e-01 5.11841714e-01
-9.89864349e-01 -7.02040493e-01 -8.35142255e-01 -6.56007230e-01
2.28219852e-01 2.27393985e-01 1.06763220e+00 -2.40209728e-01
-3.31859314e-03 2.58547127e-01 -5.95415592e-01 -1.63738251e-01
-1.81669623e-01 3.62725973e-01 5.59253395e-01 5.17606974e-01
7.29253562e-03 -1.23164558e+00 -3.67818236e-01 3.08943003e-01
-9.84338820e-01 -8.39405134e-02 6.89965963e-01 8.80975425e-01
9.78542626e-01 3.47866356e-01 6.06810272e-01 -9.47774172e-01
6.64230287e-01 1.33992150e-01 -1.63483620e-01 3.24790068e-02
-3.20695311e-01 4.13838699e-02 8.56988847e-01 -7.02884257e-01
-5.38217485e-01 6.61318660e-01 4.09974009e-02 -9.72485483e-01
7.95784444e-02 7.43228436e-01 -1.08277209e-01 1.09449163e-01
8.93151939e-01 2.42818758e-01 -3.12734634e-01 -9.48500693e-01
5.88429451e-01 6.97041631e-01 9.43315387e-01 -5.08367598e-01
9.97773826e-01 4.70843196e-01 8.56948271e-02 -8.90441656e-01
-1.03211427e+00 -8.54624987e-01 -1.21762264e+00 -3.67799908e-01
6.26558125e-01 -1.12402332e+00 -6.60341918e-01 2.94935822e-01
-9.85479534e-01 -1.72754467e-01 -5.52447379e-01 7.37767220e-01
-9.73206580e-01 4.22483325e-01 -7.52902985e-01 -7.10425854e-01
-3.29345018e-01 -7.14637458e-01 1.19982278e+00 -3.63316506e-01
-5.99182844e-01 -3.86533648e-01 5.33057392e-01 6.16196334e-01
-9.94823221e-03 -2.24671453e-01 7.71177232e-01 -3.50964695e-01
-3.10520470e-01 -2.40965724e-01 1.46017462e-01 5.90546668e-01
1.80099592e-01 -2.04276964e-01 -1.21714842e+00 -3.80179524e-01
-5.21811731e-02 -3.47999334e-01 1.18892229e+00 1.17059648e-01
1.03574491e+00 -1.13150842e-01 2.13930011e-01 5.94987988e-01
9.54798579e-01 -7.34358355e-02 5.83721220e-01 2.80995127e-02
9.78134513e-01 1.26092389e-01 7.14287996e-01 5.74843585e-01
-8.95871744e-02 8.71892512e-01 -2.76570749e-02 1.45336792e-01
-2.10264474e-02 -3.30708534e-01 4.97055173e-01 1.31669509e+00
-5.65993607e-01 2.06476033e-01 -6.42675579e-01 4.08070475e-01
-2.34554791e+00 -1.11809826e+00 -5.31394929e-02 2.18788910e+00
8.29308271e-01 -8.84638578e-02 4.72924471e-01 1.12259245e+00
3.87196511e-01 7.25609809e-02 -3.07139844e-01 -1.93009302e-01
-2.24014834e-01 5.89880526e-01 1.30469993e-01 2.56628931e-01
-1.41049349e+00 8.08578014e-01 6.44858742e+00 8.98046017e-01
-1.16637659e+00 1.72988668e-01 -1.54268041e-01 -4.08064187e-01
-1.16717003e-01 -8.24879259e-02 -1.37539074e-01 -9.82999578e-02
9.31790233e-01 1.28114656e-01 9.32213426e-01 7.40906000e-01
1.46259323e-01 3.79339755e-01 -1.28939927e+00 1.70798123e+00
3.32599908e-01 -1.30863178e+00 2.44923145e-01 -7.78419012e-03
8.52730453e-01 -1.77253380e-01 1.50332436e-01 2.55221665e-01
-1.99475467e-01 -9.71532285e-01 9.61550355e-01 6.18433118e-01
7.36759305e-01 -9.94722843e-01 5.16546428e-01 3.34850460e-01
-1.44842577e+00 -2.84668118e-01 -1.88782319e-01 -5.11355877e-01
2.22618505e-01 9.67579961e-01 -1.53800949e-01 6.66804373e-01
7.27934301e-01 1.12278354e+00 -4.23742354e-01 8.92296910e-01
-3.47964019e-01 9.87121344e-01 -1.77813828e-01 6.82809889e-01
1.35306716e-02 -3.68977487e-01 6.07184112e-01 1.19832408e+00
4.28554982e-01 1.73926875e-02 2.27290064e-01 6.49422884e-01
-2.55576789e-01 4.57715809e-01 -4.74423438e-01 -4.36485857e-01
-2.24174216e-01 1.35131001e+00 -7.24912882e-01 -2.50431180e-01
-9.87915397e-02 1.21418297e+00 2.09886804e-01 2.50263423e-01
-4.04776931e-01 -2.97168255e-01 6.19702876e-01 1.20437451e-01
5.66160500e-01 -5.38447857e-01 -2.89521635e-01 -1.49360871e+00
5.32261841e-02 -9.67953801e-01 3.38928401e-01 -7.79056787e-01
-1.18996119e+00 5.57333410e-01 -4.64929491e-01 -1.78751075e+00
-6.34497166e-01 -3.70004714e-01 -3.40360165e-01 5.07830679e-01
-9.44594324e-01 -1.06238008e+00 1.57090187e-01 8.94724190e-01
3.36658210e-01 -4.83593524e-01 1.29298437e+00 8.26720476e-01
-1.86729461e-01 5.33750534e-01 3.15788180e-01 2.17993543e-01
9.90219474e-01 -1.21081531e+00 1.51797742e-01 5.72906554e-01
1.27981472e+00 9.41071749e-01 5.93540013e-01 -3.29693615e-01
-1.57398367e+00 -8.37856770e-01 9.09233630e-01 -1.96483478e-01
8.27171266e-01 -4.75023955e-01 -6.58583164e-01 6.28184259e-01
4.75630388e-02 -1.48446634e-01 1.26076651e+00 5.60208559e-01
-5.10721385e-01 -2.12804914e-01 -4.69476163e-01 3.10830086e-01
1.10936964e+00 -1.27313471e+00 -6.53592229e-01 3.45239133e-01
4.46172416e-01 -2.65550315e-01 -8.88290346e-01 2.68163055e-01
8.42767775e-01 -8.15682173e-01 9.52073395e-01 -6.49694026e-01
3.60705853e-01 -6.76657319e-01 -3.17311943e-01 -1.31640291e+00
-5.75912237e-01 -1.25093102e+00 -2.77213782e-01 1.08613312e+00
4.74019527e-01 1.70551747e-01 9.10908341e-01 -2.07062840e-01
-4.16398922e-04 -3.26576352e-01 -1.11425281e+00 -8.87899160e-01
-2.74967432e-01 -9.37506437e-01 8.72211009e-02 1.13617659e+00
2.84820795e-01 7.96637058e-01 -8.20679724e-01 -5.33420853e-02
4.86562103e-01 6.80043995e-01 1.08485103e+00 -1.51997423e+00
-9.37359154e-01 -1.15636483e-01 -6.04694724e-01 -9.82790351e-01
4.33946222e-01 -1.07082796e+00 -1.10412046e-01 -1.13598740e+00
9.44212377e-02 -3.83350067e-02 -6.94260895e-01 5.34680724e-01
2.83104956e-01 9.51965809e-01 4.57331777e-01 4.62778538e-01
-7.76176155e-01 4.38997388e-01 8.88531268e-01 -2.44421959e-01
-5.31061232e-01 2.48303398e-01 -5.77445149e-01 7.75630951e-01
4.37350243e-01 -6.30093813e-01 -4.39856976e-01 -1.71245098e-01
7.64715731e-01 2.28467986e-01 5.59005029e-02 -1.30698609e+00
3.06233764e-01 2.27221712e-01 1.56412810e-01 -7.50214577e-01
5.75267315e-01 -5.75448275e-01 4.95185971e-01 1.89554930e-01
-5.27246892e-01 9.12986323e-02 1.52339116e-01 5.59048831e-01
-6.55181348e-01 -8.25037882e-02 2.19818428e-01 -1.19795129e-02
-1.41316786e-01 -1.55697437e-02 -5.57969391e-01 -1.59042984e-01
2.70170927e-01 2.41716683e-01 3.99733484e-01 -5.51572323e-01
-1.41482830e+00 -6.15254343e-01 1.98531836e-01 4.75261897e-01
4.22465920e-01 -1.56300175e+00 -6.36140049e-01 1.44844189e-01
4.48018834e-02 -2.93561518e-01 4.80476558e-01 8.91141653e-01
-1.23738751e-01 5.11966050e-01 -1.69093207e-01 -5.58046579e-01
-1.47148943e+00 5.37846863e-01 -1.66666910e-01 -4.52690929e-01
-5.96052945e-01 6.24729395e-01 1.08805172e-01 -4.14350092e-01
3.50449115e-01 -4.71504033e-01 -1.35050178e-01 5.24198413e-01
7.33020723e-01 3.72426510e-01 3.68149310e-01 -9.28449392e-01
-1.51220843e-01 8.14639926e-01 2.10131213e-01 -5.49574673e-01
1.42802382e+00 2.11403385e-01 -3.10834676e-01 1.12698555e+00
8.88097584e-01 3.75023842e-01 -6.86994135e-01 -3.70418161e-01
-2.11959079e-01 -4.61959690e-01 3.04788798e-01 -6.77601933e-01
-1.11324751e+00 8.86822641e-01 4.17028904e-01 -4.58296314e-02
1.25513983e+00 -2.70384938e-01 8.49714577e-01 1.02604127e+00
5.32844007e-01 -1.00911367e+00 -1.71496607e-02 5.17063856e-01
1.05062270e+00 -7.57414043e-01 1.83213443e-01 -2.94422448e-01
-6.84343934e-01 1.18277979e+00 -1.57851428e-01 -4.23688799e-01
8.57942164e-01 1.41628325e-01 7.37467483e-02 1.65933184e-02
-7.44642317e-01 -1.65473416e-01 7.12803602e-01 2.92198002e-01
6.88830793e-01 3.00903201e-01 -3.28843333e-02 1.15228844e+00
-8.05453420e-01 3.04963112e-01 4.07551974e-01 7.81361341e-01
1.01509236e-01 -1.40332437e+00 -4.91271287e-01 3.25918078e-01
-6.68194711e-01 -2.64310718e-01 -8.53448510e-01 1.63327202e-01
1.76867113e-01 7.86454260e-01 -2.37607092e-01 -6.43624842e-01
3.95703703e-01 2.45158747e-01 6.59896612e-01 -7.29026079e-01
-9.91477907e-01 6.39119327e-01 -1.32020906e-01 -5.87535083e-01
-7.84032881e-01 -8.60374749e-01 -1.15741670e+00 2.44662426e-02
-2.75682747e-01 4.42225486e-01 6.02858245e-01 7.64340103e-01
1.54950574e-01 5.63355029e-01 4.05506492e-01 -1.27642167e+00
-2.92988420e-01 -1.06410038e+00 -1.07610357e+00 4.95668769e-01
3.58635306e-01 -5.83934784e-01 -3.32726389e-01 3.16574901e-01] | [15.765318870544434, 5.346141338348389] |
a0264697-afef-4dbb-b499-9d52ace8680a | lipschitz-recurrent-neural-networks | 2006.12070 | null | https://arxiv.org/abs/2006.12070v3 | https://arxiv.org/pdf/2006.12070v3.pdf | Lipschitz Recurrent Neural Networks | Viewing recurrent neural networks (RNNs) as continuous-time dynamical systems, we propose a recurrent unit that describes the hidden state's evolution with two parts: a well-understood linear component plus a Lipschitz nonlinearity. This particular functional form facilitates stability analysis of the long-term behavior of the recurrent unit using tools from nonlinear systems theory. In turn, this enables architectural design decisions before experimentation. Sufficient conditions for global stability of the recurrent unit are obtained, motivating a novel scheme for constructing hidden-to-hidden matrices. Our experiments demonstrate that the Lipschitz RNN can outperform existing recurrent units on a range of benchmark tasks, including computer vision, language modeling and speech prediction tasks. Finally, through Hessian-based analysis we demonstrate that our Lipschitz recurrent unit is more robust with respect to input and parameter perturbations as compared to other continuous-time RNNs. | ['Liam Hodgkinson', 'Omri Azencot', 'N. Benjamin Erichson', 'Michael W. Mahoney', 'Alejandro Queiruga'] | 2020-06-22 | null | https://openreview.net/forum?id=-N7PBXqOUJZ | https://openreview.net/pdf?id=-N7PBXqOUJZ | iclr-2021-1 | ['sequential-image-classification'] | ['computer-vision'] | [-1.87805109e-02 2.69954860e-01 -1.09923579e-01 -2.04319283e-02
-2.36074761e-01 -5.47219992e-01 5.78637242e-01 -5.99145234e-01
2.25612801e-02 3.58957231e-01 3.79049152e-01 -6.28432155e-01
8.45996290e-02 -2.42560089e-01 -8.35912466e-01 -1.03860235e+00
-2.13096172e-01 -2.16258973e-01 -1.28462940e-01 -5.39828241e-01
-1.65718317e-01 5.06911755e-01 -1.10204566e+00 -2.43074119e-01
2.63366520e-01 8.75552714e-01 5.99858016e-02 8.54946375e-01
5.73010743e-01 1.23139226e+00 -2.43851960e-01 2.41097614e-01
4.68434453e-01 -7.06260860e-01 -5.03464282e-01 -9.86727537e-04
-9.17702690e-02 -1.77918330e-01 -6.39279783e-01 1.11009634e+00
4.63712811e-01 2.82993674e-01 6.97763860e-01 -8.51756096e-01
-9.81493771e-01 7.73051739e-01 -1.64364036e-02 1.90822825e-01
-1.45429954e-01 1.80773601e-01 1.16506922e+00 -9.64371443e-01
6.74807549e-01 1.30433631e+00 1.04195774e+00 6.32127702e-01
-1.36167908e+00 -5.40364623e-01 4.56648290e-01 -2.41273209e-01
-1.18488407e+00 -8.21293473e-01 7.35719919e-01 -5.51217854e-01
1.25744867e+00 7.35925809e-02 4.00409758e-01 1.18122900e+00
5.78270793e-01 8.29494536e-01 6.03995025e-01 -3.06336522e-01
2.06243262e-01 1.93151049e-02 3.47316116e-01 8.07552576e-01
-1.02859274e-01 1.63434610e-01 -7.81559423e-02 -5.60891442e-02
9.58277345e-01 1.54758185e-01 -4.43670034e-01 -2.87800252e-01
-9.76107180e-01 8.29725027e-01 5.71305513e-01 2.88956821e-01
-4.27430511e-01 6.03114963e-01 3.32368553e-01 8.15599322e-01
4.08742040e-01 3.23272675e-01 -2.88512498e-01 -4.56179380e-02
-3.43435913e-01 -1.91691875e-01 9.19362605e-01 1.04757094e+00
5.84190249e-01 6.09708786e-01 1.57467306e-01 6.98286772e-01
4.15410161e-01 5.96220672e-01 8.40522885e-01 -7.38341987e-01
3.45169842e-01 3.86438698e-01 -1.54152408e-01 -9.49474037e-01
-6.01972401e-01 -7.86035061e-01 -1.34623086e+00 5.64337894e-02
-4.42230292e-02 -4.43359554e-01 -8.07759345e-01 1.97291839e+00
-8.31435919e-02 2.39406675e-01 3.22194189e-01 6.28738105e-01
2.83275008e-01 1.03164804e+00 -6.46270931e-01 -5.89180827e-01
6.44954503e-01 -8.08706462e-01 -9.53639030e-01 -3.18088941e-02
7.80389786e-01 -2.01977924e-01 9.27240193e-01 4.26582620e-02
-1.28499115e+00 -3.08987290e-01 -1.07054579e+00 -8.78064930e-02
-2.65469313e-01 3.63386214e-01 -3.50524895e-02 2.54772782e-01
-1.41689050e+00 5.19114077e-01 -1.29755628e+00 -5.41222580e-02
-3.68752062e-01 3.72280657e-01 -6.84662908e-02 7.56848872e-01
-1.33799565e+00 9.63449121e-01 1.45805990e-02 8.50679457e-01
-9.14141953e-01 -6.65962100e-01 -8.57583821e-01 8.21498483e-02
3.00595537e-02 -5.65790892e-01 1.40867364e+00 -7.79766202e-01
-1.93767917e+00 1.34832382e-01 -3.85836512e-01 -8.83446813e-01
5.63725352e-01 -1.33949235e-01 -3.15609604e-01 -7.24805295e-02
-3.21581513e-01 4.44179177e-02 1.14551735e+00 -7.88108170e-01
-1.57751650e-01 3.67042832e-02 -2.88232207e-01 8.70616138e-02
-4.76127088e-01 -3.58427614e-01 -1.61376670e-01 -7.25770056e-01
2.87356079e-01 -1.26230991e+00 -4.85953242e-01 -2.80365527e-01
-4.71569121e-01 -5.87150753e-02 8.99778545e-01 -4.42979515e-01
1.46809459e+00 -2.18908501e+00 5.82324386e-01 3.41426164e-01
3.18340451e-01 1.43714190e-01 -3.52335200e-02 5.56977153e-01
-3.56376827e-01 1.59539506e-01 -1.82707570e-02 -6.17956877e-01
1.06133156e-01 1.23508543e-01 -9.23317075e-01 7.19310343e-01
1.63850829e-01 1.20152950e+00 -5.54539561e-01 3.74542117e-01
-4.59897965e-02 7.74306357e-01 -3.44719708e-01 6.51212558e-02
-6.14851788e-02 7.52448663e-02 -1.88167274e-01 2.21463770e-01
-1.59664065e-01 -3.74140114e-01 1.40772555e-02 1.40994966e-01
-3.08882326e-01 5.18658996e-01 -9.23202515e-01 9.41251516e-01
-6.43291175e-01 1.05592000e+00 1.78633258e-01 -8.93453181e-01
7.98293173e-01 6.12359703e-01 2.04827964e-01 -3.84000182e-01
3.53920400e-01 1.16736323e-01 -2.14172117e-02 -1.86882764e-01
2.22295731e-01 -1.08929828e-01 1.84583873e-01 8.14245582e-01
-1.35677412e-01 3.71402681e-01 -5.22102453e-02 2.15386733e-01
1.18307769e+00 -5.45197606e-01 1.17799111e-01 -4.49203938e-01
4.14555639e-01 -5.98513961e-01 5.41853964e-01 7.34747052e-01
-9.05867219e-02 4.32787359e-01 7.76944339e-01 -2.48204261e-01
-1.17836666e+00 -7.22412944e-01 2.85316184e-02 8.61093342e-01
-3.33891153e-01 -2.25594237e-01 -5.82054257e-01 8.15480500e-02
-1.16332337e-01 2.29521051e-01 -9.87243116e-01 -5.46382070e-01
-6.89404666e-01 -4.16119218e-01 7.36924648e-01 6.68269813e-01
1.30331576e-01 -9.65972006e-01 -6.15657866e-01 3.04756165e-01
1.65893704e-01 -9.04886425e-01 -9.35546041e-01 6.18480980e-01
-9.41079795e-01 -5.17427862e-01 -9.59917128e-01 -9.31239605e-01
7.12739527e-01 1.60279542e-01 4.67764556e-01 -7.39447922e-02
7.24428296e-02 4.49928373e-01 -1.81152001e-02 -5.33069000e-02
-7.45401800e-01 2.41776660e-01 6.37527883e-01 1.97992548e-01
-4.28901136e-01 -6.25565112e-01 -3.73266339e-01 4.14678544e-01
-7.91594386e-01 1.25921845e-01 5.67683876e-01 1.03805828e+00
4.73853081e-01 1.92392897e-02 6.95098758e-01 -5.64330697e-01
1.04300833e+00 -4.27938968e-01 -8.17831933e-01 3.69635224e-01
-6.64407790e-01 4.23677653e-01 9.28784013e-01 -8.75348806e-01
-6.22838557e-01 1.72927693e-01 9.68327895e-02 -8.14955235e-01
6.41600668e-01 7.52606809e-01 1.93713114e-01 7.92689696e-02
6.22310162e-01 4.24992949e-01 4.52478737e-01 -2.75718361e-01
3.84102762e-01 3.53479981e-01 5.19004643e-01 -6.78784847e-02
9.36508834e-01 3.86988342e-01 -1.36212394e-01 -1.33715534e+00
-6.17740870e-01 -4.59231526e-01 -6.46753013e-01 -1.66274279e-01
4.70306039e-01 -9.47860360e-01 -1.03555906e+00 5.27930200e-01
-1.01399720e+00 -7.52781689e-01 -3.67041826e-01 3.91730219e-01
-6.39834583e-01 -1.63929328e-01 -1.23229337e+00 -1.28715885e+00
-3.76622111e-01 -1.04545069e+00 6.60383761e-01 -4.51215245e-02
-4.18304540e-02 -1.43596590e+00 3.30903083e-01 -4.86305863e-01
5.93626678e-01 -1.06806248e-01 7.21484005e-01 -1.67635262e-01
-4.62450176e-01 -2.45263219e-01 1.99887618e-01 6.71779335e-01
1.73783779e-01 3.38732451e-01 -7.90161490e-01 -5.08186579e-01
4.27388668e-01 4.99437191e-03 1.08835089e+00 6.37042165e-01
3.14463437e-01 -7.63147771e-01 -6.69837818e-02 7.62526274e-01
1.06414318e+00 2.36120716e-01 3.51291895e-01 6.98079839e-02
8.39591861e-01 3.62805277e-01 -2.33764157e-01 1.91635653e-01
1.45744503e-01 1.12937704e-01 1.66647926e-01 -1.16531774e-01
2.85615355e-01 -2.14518934e-01 1.11850727e+00 1.54911864e+00
1.25340670e-02 3.15214172e-02 -8.29919636e-01 3.70315641e-01
-2.15194011e+00 -1.13049865e+00 1.37925163e-01 1.90401721e+00
6.96187377e-01 2.76537746e-01 3.04671109e-01 1.51513129e-01
5.98389328e-01 4.27860580e-02 -1.03172100e+00 -4.22338516e-01
-4.01306957e-01 -3.75422388e-01 7.15816021e-01 7.87034690e-01
-9.16868925e-01 8.55217516e-01 7.82204437e+00 5.97679690e-02
-1.45190918e+00 -1.97718769e-01 4.84396279e-01 -1.11294322e-01
-2.83052862e-01 -2.91352183e-01 -1.05950570e+00 4.47980948e-02
1.40393877e+00 -4.65057194e-01 6.07890487e-01 7.62385845e-01
5.63870847e-01 6.49184287e-01 -1.08526886e+00 7.41482615e-01
-1.35878399e-01 -1.55980802e+00 -1.79428861e-01 2.12437347e-01
9.10106182e-01 4.17965978e-01 5.87454557e-01 2.80877560e-01
5.32567561e-01 -9.84910131e-01 8.48738194e-01 6.29712045e-01
5.35420299e-01 -6.40400589e-01 4.07266855e-01 4.23502028e-01
-1.42328334e+00 -3.62469941e-01 -3.93647134e-01 -2.95990407e-01
1.46931186e-01 4.39908117e-01 -8.21713448e-01 -5.31891473e-02
3.51435632e-01 1.24907887e+00 -3.02543849e-01 4.80753899e-01
-5.67784607e-02 8.34401011e-01 -3.34992439e-01 -2.08589718e-01
4.51063007e-01 -3.89009804e-01 6.33186340e-01 1.25308883e+00
1.45210102e-01 1.36519402e-01 -1.65011182e-01 9.35758710e-01
-8.06870684e-02 -2.01116398e-01 -8.67133439e-01 -3.93275350e-01
1.20307498e-01 9.44037557e-01 -6.05411351e-01 -1.24505594e-01
-1.21640824e-01 7.53202796e-01 4.03932810e-01 9.81805503e-01
-6.67947054e-01 -3.45214903e-01 9.90466297e-01 -1.57313883e-01
4.41363513e-01 -5.84841669e-01 -1.41375259e-01 -1.32968271e+00
2.51723737e-01 -6.73206151e-01 8.85723904e-02 -6.64585352e-01
-9.53173101e-01 8.83169711e-01 -3.48365754e-01 -1.23423195e+00
-8.33632767e-01 -6.00341797e-01 -6.20822847e-01 8.15959156e-01
-1.25087285e+00 -7.27943718e-01 4.12807882e-01 5.97809315e-01
3.73349547e-01 -2.74846286e-01 6.31810367e-01 -1.47484960e-02
-1.22834039e+00 6.59639657e-01 4.95147884e-01 3.21985602e-01
1.58586338e-01 -1.19151986e+00 7.93949962e-01 1.16107416e+00
3.91002838e-03 1.26780331e+00 9.74514246e-01 -4.62706536e-01
-1.79705846e+00 -1.05247784e+00 4.93378460e-01 -2.16314927e-01
1.35161769e+00 -6.24832690e-01 -1.05820227e+00 1.20981395e+00
-2.02048526e-04 -6.28775656e-02 9.94834080e-02 3.85804512e-02
-3.66890371e-01 1.23144664e-01 -3.30063045e-01 1.07263398e+00
8.93474996e-01 -1.06227994e+00 -2.96745688e-01 9.22315568e-02
1.00776470e+00 -4.13851738e-01 -5.73726535e-01 1.99965596e-01
7.50578046e-01 -4.77120847e-01 7.62610674e-01 -6.85245275e-01
2.84239024e-01 -1.85164824e-01 -7.17077926e-02 -1.38271201e+00
-5.09195864e-01 -1.46739459e+00 -6.11190498e-01 6.07540011e-01
7.03619421e-01 -1.01743758e+00 4.70998436e-01 5.19974172e-01
-2.11539790e-01 -8.96224320e-01 -6.95004880e-01 -1.07986069e+00
2.60400355e-01 -2.80351043e-01 -1.48245290e-01 7.58939862e-01
7.50185251e-02 5.36668777e-01 -5.23759127e-01 2.82235235e-01
1.17217869e-01 -2.62655169e-01 3.72720957e-01 -9.73579526e-01
-1.79408267e-01 -8.91002893e-01 -4.20648247e-01 -1.49874580e+00
4.68611449e-01 -7.70858228e-01 2.60871351e-01 -1.19588304e+00
-3.81054103e-01 -1.17027283e-01 -5.15729547e-01 3.87935579e-01
2.56754130e-01 -8.60195532e-02 1.16609305e-01 4.38619733e-01
-2.20781669e-01 8.19419205e-01 8.52949262e-01 -2.05007195e-02
-6.52560890e-01 3.38287175e-01 -4.75320399e-01 6.50686324e-01
6.94274247e-01 -2.04447985e-01 -6.81461036e-01 -1.92975476e-01
4.36199635e-01 4.31710593e-02 1.65155366e-01 -7.92424619e-01
6.72188282e-01 4.20929901e-02 -1.59455597e-01 -5.94271302e-01
4.61768299e-01 -6.76336229e-01 1.24825791e-01 7.76226997e-01
-8.46550643e-01 4.82642591e-01 -3.95719521e-02 6.30668223e-01
-4.55095135e-02 1.60437599e-01 7.16882527e-01 3.32823634e-01
-9.38697830e-02 2.44132370e-01 -8.37734997e-01 -1.40403494e-01
6.43751502e-01 -1.22175217e-01 -2.34722629e-01 -9.48545933e-01
-8.38806689e-01 3.33960444e-01 -3.02129629e-04 4.10315484e-01
5.90583563e-01 -1.31590450e+00 -5.05779326e-01 5.39825320e-01
-3.50835860e-01 -3.19466054e-01 -5.96332438e-02 1.15724635e+00
-2.55008847e-01 6.96666062e-01 1.41676158e-01 -6.68280125e-01
-1.11142576e+00 4.95220900e-01 7.50131190e-01 -4.62204255e-02
-1.02629781e+00 7.81819344e-01 2.98275322e-01 -3.37265998e-01
7.43536890e-01 -1.04672885e+00 8.45344514e-02 6.85421079e-02
6.59719646e-01 2.27126494e-01 -1.38625258e-03 -6.32114649e-01
-7.91439265e-02 4.84714359e-01 -5.48781455e-02 -4.52052087e-01
1.34068704e+00 -3.93595964e-01 -8.49086344e-02 1.48604763e+00
1.61444485e+00 -4.15714949e-01 -1.43910885e+00 -4.82812047e-01
1.07188269e-01 5.73561966e-01 3.97202112e-02 -9.76923108e-02
-1.09575891e+00 8.25820446e-01 2.96424121e-01 7.12386250e-01
9.76247370e-01 -3.21316749e-01 6.21984780e-01 1.12149835e+00
-9.92326885e-02 -1.08521473e+00 -7.41028935e-02 1.25397265e+00
1.10655177e+00 -8.59686792e-01 -2.97893137e-01 1.96297035e-01
-4.89336878e-01 1.18760037e+00 -6.92583844e-02 -4.42001164e-01
1.25133753e+00 5.16744375e-01 3.05566043e-01 1.46293402e-01
-1.57594895e+00 7.71339936e-03 3.90942037e-01 1.02541037e-02
4.00086522e-01 -1.38462901e-01 3.24974746e-01 2.22098738e-01
-4.09062117e-01 -5.00019133e-01 8.04262757e-01 7.68754601e-01
-4.37383175e-01 -4.54655290e-01 -2.33948436e-02 1.09204784e-01
-2.23404184e-01 -8.00727978e-02 -4.08727169e-01 6.31243527e-01
-1.02715981e+00 6.54681325e-01 7.89336637e-02 -4.93802696e-01
2.00322360e-01 5.69419116e-02 -8.25868398e-02 -5.30355990e-01
-6.71154797e-01 4.31934595e-01 -3.78227860e-01 -5.20108223e-01
-3.25992368e-02 -7.22765148e-01 -1.38806391e+00 -3.50753158e-01
-6.06422603e-01 1.31448805e-02 7.40977228e-01 9.61989820e-01
3.90511632e-01 7.08568692e-01 8.69566619e-01 -7.73981333e-01
-9.96949255e-01 -7.90520251e-01 -4.72970933e-01 -7.49237090e-02
1.24889576e+00 -2.86039233e-01 -8.37785959e-01 2.07387760e-01] | [7.600526332855225, 3.375199794769287] |
a9b16487-ac22-4ece-867b-83414ccf4eb5 | self-supervised-motion-retargeting-with | 2103.06447 | null | https://arxiv.org/abs/2103.06447v1 | https://arxiv.org/pdf/2103.06447v1.pdf | Self-Supervised Motion Retargeting with Safety Guarantee | In this paper, we present self-supervised shared latent embedding (S3LE), a data-driven motion retargeting method that enables the generation of natural motions in humanoid robots from motion capture data or RGB videos. While it requires paired data consisting of human poses and their corresponding robot configurations, it significantly alleviates the necessity of time-consuming data-collection via novel paired data generating processes. Our self-supervised learning procedure consists of two steps: automatically generating paired data to bootstrap the motion retargeting, and learning a projection-invariant mapping to handle the different expressivity of humans and humanoid robots. Furthermore, our method guarantees that the generated robot pose is collision-free and satisfies position limits by utilizing nonparametric regression in the shared latent space. We demonstrate that our method can generate expressive robotic motions from both the CMU motion capture database and YouTube videos. | ['Joohyung Kim', 'Hyemin Ahn', 'Min Jae Song', 'Sungjoon Choi'] | 2021-03-11 | null | null | null | null | ['motion-retargeting'] | ['computer-vision'] | [-6.71763644e-02 3.43613684e-01 -3.55053693e-01 8.33484437e-03
-7.20876217e-01 -5.72782636e-01 6.94581389e-01 -6.77250922e-01
-4.63605762e-01 7.34716356e-01 4.34917480e-01 2.45496824e-01
-2.25521773e-02 -4.86990631e-01 -7.59182751e-01 -8.01189721e-01
-2.08257928e-01 6.84567988e-01 5.76823503e-02 -6.79283440e-02
-1.15075313e-01 5.89245617e-01 -1.59872282e+00 -1.96112484e-01
4.53686565e-01 4.28009063e-01 4.82123822e-01 9.83014464e-01
4.34140384e-01 8.42852473e-01 -7.92767480e-02 2.43689328e-01
5.24203658e-01 -4.07946706e-01 -6.80667400e-01 2.61885405e-01
3.09084207e-02 -5.76935351e-01 -6.61131024e-01 6.89342499e-01
6.09312773e-01 5.36787748e-01 8.32279027e-01 -1.85573220e+00
-4.18518752e-01 3.94737899e-01 -3.36153179e-01 -4.62257087e-01
7.14507997e-01 3.78234446e-01 7.16705620e-01 -9.93930101e-01
1.07525551e+00 1.29495192e+00 5.73633730e-01 9.06832159e-01
-1.35305285e+00 -6.32174075e-01 -4.50655401e-01 7.75861144e-02
-1.39263499e+00 -4.82558340e-01 8.87472749e-01 -7.88385928e-01
8.53872180e-01 -5.66877089e-02 7.60566473e-01 1.61791575e+00
1.73148096e-01 8.16352129e-01 3.48616362e-01 -3.19757551e-01
3.40628773e-01 -2.88472444e-01 -4.98374760e-01 7.19332099e-01
1.30980864e-01 7.36579373e-02 -5.88204443e-01 -3.54508430e-01
1.45184290e+00 -4.88364100e-02 -3.10677826e-01 -1.43402469e+00
-1.84707296e+00 8.14553976e-01 2.21518263e-01 3.53462510e-02
-4.55553651e-01 5.84415674e-01 3.03892672e-01 4.42018323e-02
-8.69838670e-02 4.14602041e-01 -5.17112374e-01 -4.51380849e-01
-4.18748230e-01 4.58911568e-01 6.93450451e-01 1.49078262e+00
8.27194035e-01 1.73209652e-01 2.86250323e-01 4.35736954e-01
2.79042810e-01 5.10282040e-01 9.46770430e-01 -1.57781637e+00
4.10170466e-01 4.88052547e-01 4.77391005e-01 -1.17131341e+00
-5.22523105e-01 6.87482417e-01 -5.20663738e-01 3.25230569e-01
2.45155588e-01 -3.37382048e-01 -5.46975672e-01 1.94618368e+00
3.66139829e-01 -1.52379006e-01 4.28091317e-01 1.00860786e+00
4.44228530e-01 5.71767569e-01 -8.38282402e-04 4.04967293e-02
7.52637446e-01 -1.15063906e+00 -6.45928442e-01 -1.34078953e-02
9.66658950e-01 -3.68432701e-01 1.27890134e+00 6.60966560e-02
-9.21876013e-01 -6.92385197e-01 -1.03282475e+00 -3.17836165e-01
1.36382341e-01 2.92052984e-01 5.02180934e-01 2.53122300e-02
-8.71046782e-01 6.19046271e-01 -1.24144018e+00 -6.00688159e-01
-1.52907446e-01 5.86589336e-01 -9.95396316e-01 1.96506605e-01
-7.62867749e-01 7.25386739e-01 5.39408565e-01 -4.52283770e-01
-9.74639893e-01 -3.91765177e-01 -1.46742308e+00 -4.51999456e-01
5.16568303e-01 -8.65400910e-01 1.23195803e+00 -6.62365437e-01
-2.05023456e+00 6.45794690e-01 2.58803785e-01 -2.21407011e-01
6.90616846e-01 -6.65405154e-01 9.64616239e-02 2.64421165e-01
3.16418529e-01 1.35680103e+00 1.06646764e+00 -1.19997036e+00
-2.43806958e-01 -1.36029243e-01 -3.81931365e-01 3.57717127e-01
-1.11819759e-01 -4.87716764e-01 -6.26940131e-01 -8.17191422e-01
3.12324047e-01 -1.57953334e+00 -5.15888751e-01 3.05441499e-01
-2.72648484e-01 9.35199205e-03 9.83487606e-01 -2.92397767e-01
3.93941760e-01 -2.29719186e+00 9.20963526e-01 -1.05734833e-01
6.53477712e-03 -4.10694361e-01 -2.70270616e-01 3.89022738e-01
-4.81936820e-02 -3.24440539e-01 -9.73837897e-02 -4.34029579e-01
1.66712701e-01 5.04101574e-01 -4.39756393e-01 7.06559718e-01
3.50124240e-02 1.02560771e+00 -1.17773426e+00 -6.92076921e-01
5.30061305e-01 3.85445982e-01 -8.10456455e-01 4.90008861e-01
-1.85131326e-01 1.01286089e+00 -2.99975455e-01 3.52263123e-01
-2.97168307e-02 -2.39704270e-02 1.72522545e-01 -2.83124208e-01
-6.17460869e-02 -1.67974934e-01 -1.23428535e+00 2.36488748e+00
-2.71355271e-01 3.91347110e-01 -7.21715540e-02 -4.08547521e-01
9.55012023e-01 4.15764511e-01 1.07798958e+00 -6.16012840e-03
2.53856033e-01 1.39825031e-01 -4.96987343e-01 -6.78719878e-01
7.45335460e-01 1.03396833e-01 -5.44159353e-01 7.27332354e-01
2.85129994e-01 -5.15525520e-01 -3.54506016e-01 3.25253382e-02
1.02441061e+00 9.43179607e-01 4.51738089e-01 -3.63017470e-02
1.91558078e-01 2.76787251e-01 8.19567144e-01 1.34684846e-01
-3.06431472e-01 6.22492135e-01 1.57495320e-01 -2.82886624e-01
-1.52856958e+00 -1.39178455e+00 4.81627911e-01 1.02718139e+00
2.70743668e-01 -3.00588697e-01 -4.69672590e-01 -2.76132435e-01
1.79785676e-02 4.11625892e-01 -5.59195638e-01 -4.09312010e-01
-7.83286393e-01 -2.94583905e-02 5.97483814e-01 7.90613532e-01
3.27419013e-01 -1.17704713e+00 -1.23369753e+00 2.05640532e-02
-3.76714855e-01 -1.30865312e+00 -7.43883133e-01 1.07386366e-01
-8.58483672e-01 -1.11640370e+00 -8.21512699e-01 -1.13300455e+00
8.38469684e-01 2.16934010e-01 3.64307195e-01 -7.54664779e-01
-4.48308140e-01 8.73116851e-01 -5.77780187e-01 3.87482584e-01
-4.74334031e-01 -2.55226552e-01 7.56244361e-01 -2.97442973e-01
-4.29844223e-02 -6.27129495e-01 -2.27530658e-01 4.17850226e-01
-7.09478498e-01 4.41299349e-01 4.96173114e-01 8.79463673e-01
6.58370554e-01 -3.76041770e-01 2.15471014e-01 -1.04432404e-01
2.40782142e-01 -3.67620140e-01 -4.07117605e-01 -1.93974987e-01
-3.59998606e-02 3.34787607e-01 3.62014264e-01 -8.78190815e-01
-8.30282331e-01 9.74575400e-01 5.69612741e-01 -8.78477395e-01
-1.54065520e-01 5.60706668e-02 -3.54583859e-01 2.12309837e-01
7.92870283e-01 4.22381349e-02 4.31624413e-01 -2.84455240e-01
9.79819953e-01 5.02264738e-01 1.28270042e+00 -5.78735471e-01
9.21044469e-01 5.72383285e-01 -8.50560963e-02 -7.42997169e-01
1.63946167e-01 -4.81412649e-01 -1.22531044e+00 -1.57859623e-01
1.18499863e+00 -1.06044960e+00 -6.55653119e-01 3.47092450e-01
-1.13468146e+00 -6.65026724e-01 -5.53630650e-01 9.15853620e-01
-1.63570523e+00 5.30261636e-01 -6.82414711e-01 -5.24683356e-01
-1.48361940e-02 -1.28477812e+00 1.18314147e+00 -3.93784612e-01
-7.84868062e-01 -5.94394326e-01 5.56929588e-01 -3.18066888e-02
-3.22467685e-01 7.99072981e-01 6.44455791e-01 -2.80186892e-01
-3.86047244e-01 -1.57579973e-01 3.63303930e-01 -1.86994895e-02
6.16330862e-01 -5.81977963e-02 -4.93976146e-01 -5.17018259e-01
-2.46241257e-01 -8.03021371e-01 2.24651620e-01 1.76014617e-01
5.90398610e-01 -4.94804114e-01 -3.66670519e-01 6.54731989e-01
1.02074039e+00 -1.17132746e-01 3.69559258e-01 4.32106584e-01
9.09346581e-01 7.97001362e-01 7.94574618e-01 7.38150477e-01
4.37261730e-01 8.39665294e-01 3.90955001e-01 3.47488880e-01
1.48976535e-01 -7.34449506e-01 6.98386192e-01 8.57151210e-01
1.03077233e-01 1.29061401e-01 -9.11189556e-01 7.46687531e-01
-2.35761428e+00 -7.54822791e-01 2.55794823e-01 2.03629494e+00
7.85619855e-01 -4.57764864e-01 2.45806828e-01 1.71135306e-01
6.86738133e-01 -1.93394378e-01 -7.90283024e-01 5.00509841e-03
7.26616681e-02 -2.10884526e-01 5.05030513e-01 5.51816344e-01
-1.14867783e+00 1.07788408e+00 6.40748501e+00 1.95161372e-01
-7.81465352e-01 -1.26284584e-01 -2.17002302e-01 -2.02398658e-01
-2.13349357e-01 -1.93080753e-01 -2.36583605e-01 1.59079000e-01
5.98780036e-01 -2.55838364e-01 2.54278302e-01 1.25371146e+00
1.43618479e-01 4.63836640e-02 -1.41932631e+00 1.37825978e+00
3.19350511e-02 -1.07435620e+00 1.19754210e-01 2.74883360e-02
7.88511217e-01 -1.04366161e-01 -2.40788460e-02 1.25994876e-01
6.72996223e-01 -7.03952253e-01 8.03119183e-01 4.82968360e-01
9.46139574e-01 -6.49766207e-01 2.23875612e-01 5.87588489e-01
-1.03235352e+00 -2.03981817e-01 -1.41148373e-01 -9.50992182e-02
3.53657752e-01 -2.67102644e-02 -8.81739020e-01 2.42377520e-01
7.19681680e-01 8.08399498e-01 -1.32330015e-01 4.34454352e-01
-2.10989252e-01 -1.71160430e-01 -2.96029150e-01 2.43236721e-01
5.93263991e-02 -1.25116661e-01 9.59434032e-01 7.86238015e-01
4.08566505e-01 -6.79334626e-02 3.66811842e-01 8.04722667e-01
2.57741153e-01 -8.35015923e-02 -1.24313009e+00 -7.57200345e-02
4.81620401e-01 1.06074202e+00 -5.02862811e-01 -4.55506220e-02
-2.30972059e-02 1.55699837e+00 2.59428024e-01 2.98929989e-01
-7.84481049e-01 -1.75993875e-01 6.85432971e-01 -3.00220281e-01
3.14188808e-01 -1.13568568e+00 3.20239477e-02 -1.19434416e+00
-5.38123176e-02 -5.52067578e-01 7.32094720e-02 -9.07073796e-01
-8.91949475e-01 3.75022769e-01 2.36433163e-01 -1.83270800e+00
-1.23672402e+00 -5.23748875e-01 -1.36307448e-01 2.40224078e-01
-5.17525613e-01 -1.12581122e+00 -5.22191525e-01 9.26852822e-01
5.16447783e-01 -2.86923975e-01 1.02426422e+00 -8.44268501e-02
-2.82821804e-01 4.37185347e-01 -1.02092445e-01 9.82716903e-02
9.25562382e-01 -1.00357878e+00 4.15544927e-01 3.99551123e-01
3.79629768e-02 4.66741562e-01 8.68671119e-01 -6.42887890e-01
-1.63341367e+00 -1.09388685e+00 2.90409952e-01 -6.40387237e-01
6.44866347e-01 -2.91763991e-01 -3.91818196e-01 1.08477199e+00
-3.81513357e-01 9.91535932e-02 6.67770386e-01 -5.66470563e-01
-1.35258257e-01 4.70883399e-01 -1.00621390e+00 1.05494535e+00
1.16394114e+00 -4.20429438e-01 -8.38826835e-01 5.08647859e-02
1.06050634e+00 -4.81679231e-01 -8.56982589e-01 7.54560053e-01
7.13560462e-01 -4.23346579e-01 9.95935559e-01 -6.04391813e-01
3.22972476e-01 -4.13496643e-01 -5.48237801e-01 -1.28069091e+00
-4.19674665e-01 -9.33967710e-01 -2.41415337e-01 7.76374757e-01
-1.02091432e-02 -1.94815382e-01 1.03523827e+00 7.33480692e-01
-7.53892288e-02 -2.97591686e-01 -9.97477829e-01 -1.08227515e+00
-1.06950730e-01 -4.11799550e-01 4.53730017e-01 1.01225793e+00
3.90407920e-01 5.71406297e-02 -7.13286936e-01 5.71537241e-02
5.10201156e-01 2.05398761e-02 1.59602070e+00 -6.94194257e-01
-3.44899803e-01 6.44261166e-02 -8.28371286e-01 -1.16692793e+00
4.21603620e-01 -6.06055915e-01 6.04175627e-01 -1.13212597e+00
6.28234968e-02 -2.57719994e-01 2.98177630e-01 6.70432150e-01
2.10156709e-01 1.33412957e-01 1.52092114e-01 7.41932154e-01
-5.14113247e-01 1.26324189e+00 1.14629936e+00 9.68603566e-02
-6.52499795e-01 -4.67833430e-01 -5.98741397e-02 9.98603523e-01
6.80292189e-01 -2.95425594e-01 -5.83199263e-01 -3.91744405e-01
-2.43877694e-02 2.86234140e-01 6.98672771e-01 -1.14659035e+00
2.20085964e-01 -3.48544717e-01 1.79999098e-01 -6.27102971e-01
5.63130379e-01 -9.16192949e-01 4.14961487e-01 5.98496318e-01
-2.82905936e-01 2.63110608e-01 1.90624669e-01 7.72138238e-01
8.70178193e-02 1.89203426e-01 4.13760424e-01 7.51337362e-03
-9.75197256e-01 3.41612041e-01 -5.87069988e-01 -2.42245063e-01
1.41015685e+00 -4.76254910e-01 2.75484055e-01 -6.06978238e-01
-8.52033973e-01 1.90347791e-01 8.97939563e-01 7.34660506e-01
6.52362645e-01 -1.97897804e+00 -3.80641520e-01 2.64891595e-01
4.87959057e-01 3.60938221e-01 -3.30717675e-03 4.98829246e-01
-6.59155667e-01 3.23474616e-01 -6.32759035e-01 -8.58874738e-01
-9.93514061e-01 6.04406774e-01 -1.42345503e-01 3.40177298e-01
-1.04479778e+00 4.83607352e-01 1.85433432e-01 -1.00664520e+00
1.47951841e-01 -2.52918810e-01 4.93783914e-02 -4.99991864e-01
1.14147551e-01 7.06407368e-01 -6.96063459e-01 -1.06296051e+00
-2.00337991e-01 5.35221934e-01 5.06839573e-01 -6.84873521e-01
1.19003367e+00 -2.14111760e-01 7.44248405e-02 8.63742113e-01
1.35966480e+00 -1.56173944e-01 -1.83424401e+00 -1.04252964e-01
-9.93913636e-02 -3.97103697e-01 -5.45858204e-01 9.37999133e-03
-4.83243346e-01 3.05952519e-01 4.51241225e-01 -7.43169606e-01
6.99435353e-01 1.84315249e-01 9.12246108e-01 8.90244484e-01
9.11723256e-01 -1.29295754e+00 6.57220542e-01 4.54220533e-01
1.21614981e+00 -1.14809704e+00 -3.67935859e-02 -2.54631639e-01
-7.81750083e-01 9.07190263e-01 7.48140752e-01 -2.48549074e-01
4.15416449e-01 1.74152523e-01 1.36814028e-01 9.02926698e-02
-5.59232533e-01 1.69670090e-01 1.28003836e-01 9.38455641e-01
-5.87098487e-02 2.57936418e-01 4.07363996e-02 6.54513717e-01
-6.28639638e-01 -1.00654438e-02 6.06620610e-01 1.25767231e+00
-4.59068507e-01 -9.85742152e-01 -4.24422532e-01 -2.38679752e-01
4.13650095e-01 5.17218828e-01 -5.09502351e-01 1.08736670e+00
-5.43941893e-02 6.83259666e-01 1.40792519e-01 -8.47983062e-01
7.65882283e-02 -5.69655597e-02 7.87857831e-01 -8.15906048e-01
3.64295155e-01 -5.05821221e-02 -2.21379980e-01 -9.93785083e-01
-4.03623104e-01 -8.59578252e-01 -1.76466012e+00 -3.91609482e-02
-5.43049201e-02 -9.56670791e-02 3.82505804e-01 6.48405731e-01
4.71668690e-01 7.88867027e-02 5.99829912e-01 -1.65700364e+00
-5.50918460e-01 -9.10023689e-01 -4.33888078e-01 7.08164513e-01
4.50435102e-01 -1.05663490e+00 -2.40392044e-01 4.79657710e-01] | [7.179248332977295, -0.5015988349914551] |
bbe06b70-f772-41a7-82d4-6f47cb54ff8f | repairing-bugs-in-python-assignments-using | 2209.14876 | null | https://arxiv.org/abs/2209.14876v1 | https://arxiv.org/pdf/2209.14876v1.pdf | Repairing Bugs in Python Assignments Using Large Language Models | Students often make mistakes on their introductory programming assignments as part of their learning process. Unfortunately, providing custom repairs for these mistakes can require a substantial amount of time and effort from class instructors. Automated program repair (APR) techniques can be used to synthesize such fixes. Prior work has explored the use of symbolic and neural techniques for APR in the education domain. Both types of approaches require either substantial engineering efforts or large amounts of data and training. We propose to use a large language model trained on code, such as Codex, to build an APR system -- MMAPR -- for introductory Python programming assignments. Our system can fix both syntactic and semantic mistakes by combining multi-modal prompts, iterative querying, test-case-based selection of few-shots, and program chunking. We evaluate MMAPR on 286 real student programs and compare to a baseline built by combining a state-of-the-art Python syntax repair engine, BIFI, and state-of-the-art Python semantic repair engine for student assignments, Refactory. We find that MMAPR can fix more programs and produce smaller patches on average. | ['Gust Verbruggen', 'Gustavo Soares', 'Ruzica Piskac', 'Vu Le', 'Sumit Gulwani', 'José Cambronero', 'Jialu Zhang'] | 2022-09-29 | null | null | null | null | ['program-repair', 'program-repair'] | ['computer-code', 'reasoning'] | [-1.61639582e-02 1.51189193e-01 2.23835576e-02 -5.72045386e-01
-9.79363620e-01 -8.03304017e-01 -2.85239905e-01 8.11492682e-01
-2.42272884e-01 1.86918348e-01 -3.62949632e-02 -8.34320128e-01
1.82306603e-01 -9.27363038e-01 -1.06205618e+00 2.20100507e-01
4.49153334e-01 1.34333655e-01 6.93674743e-01 -4.62736815e-01
6.27739429e-01 7.31188431e-02 -1.84859300e+00 6.58632576e-01
1.28736496e+00 7.81214982e-02 4.13149983e-01 9.48341191e-01
-8.12970817e-01 1.32460642e+00 -8.56237471e-01 -7.01596856e-01
-1.97472647e-01 -1.79528624e-01 -1.09596896e+00 -7.18360186e-01
7.86111474e-01 -1.83355749e-01 -8.92736763e-03 1.20309877e+00
5.25863886e-01 4.43579435e-01 6.00840785e-02 -9.53847229e-01
-7.93371975e-01 1.28995061e+00 -1.20235316e-01 1.56934708e-01
8.58024418e-01 4.05757219e-01 8.52357209e-01 -8.31385255e-01
5.02922595e-01 8.34795594e-01 1.30671728e+00 6.29157364e-01
-1.32990277e+00 -4.94703829e-01 -2.86296964e-01 2.19738364e-01
-1.15581083e+00 -1.29619434e-01 4.66065258e-01 -7.72331178e-01
1.42292404e+00 1.17348932e-01 6.00659192e-01 6.92237854e-01
1.01154018e-02 9.69036400e-01 5.08267224e-01 -8.64383817e-01
4.34502363e-02 3.10952455e-01 7.59762943e-01 1.29426122e+00
-1.92491919e-01 -3.10855687e-01 -3.26390862e-01 -1.35028198e-01
2.11093798e-01 1.17357709e-01 -2.33072236e-01 -3.02002747e-02
-6.97195292e-01 7.11245656e-01 1.32142037e-01 4.19721827e-02
2.50246674e-01 2.65053332e-01 6.53406799e-01 6.06140196e-01
-2.86364079e-01 9.39395547e-01 -5.93532026e-01 -8.77359688e-01
-1.07880235e+00 5.86372435e-01 1.13668168e+00 1.34123480e+00
7.05714345e-01 -2.66470034e-02 -1.47417635e-01 1.06319463e+00
1.64609224e-01 -1.25341371e-01 7.61160612e-01 -1.06024337e+00
8.52889240e-01 1.01373029e+00 -3.14448029e-01 -6.41967952e-01
-1.59786791e-01 8.02087337e-02 2.61636078e-01 3.74545813e-01
4.98687685e-01 -8.34495872e-02 -7.07349360e-01 1.56453311e+00
1.53490022e-01 2.56510347e-01 -2.78386101e-02 3.86981249e-01
1.20326209e+00 5.74437022e-01 4.37572896e-01 5.68855882e-01
1.06172955e+00 -1.33473921e+00 -2.67541319e-01 -1.64565414e-01
1.34846354e+00 -1.07712507e+00 1.65961468e+00 5.47390401e-01
-1.63987935e+00 -6.08195841e-01 -7.60281384e-01 -5.16664445e-01
-3.90572131e-01 5.36302105e-02 2.93685377e-01 6.93056583e-01
-9.62498009e-01 1.01984560e+00 -8.04164350e-01 -4.04090941e-01
2.10607052e-01 3.86046022e-02 1.19096972e-02 -4.20239329e-01
-5.44616818e-01 7.65696406e-01 4.08298820e-01 -5.57585478e-01
-5.91851652e-01 -1.71929371e+00 -1.05154407e+00 7.16601193e-01
2.38396585e-01 -4.51520503e-01 2.12962866e+00 -6.06381357e-01
-1.59904087e+00 8.05539846e-01 1.89965144e-01 1.23286888e-01
1.59834951e-01 -4.51627374e-01 2.97005504e-01 -8.93959552e-02
3.06480862e-02 5.27620733e-01 1.64570943e-01 -6.72728479e-01
-4.69152242e-01 1.40996441e-01 3.09431225e-01 -2.21583107e-03
-1.46238536e-01 3.74581039e-01 -2.22415581e-01 -6.97555006e-01
-4.11448069e-02 -8.52186024e-01 -2.09666669e-01 -8.94764960e-02
-1.05062850e-01 -5.37372828e-01 4.31091845e-01 -8.19559276e-01
1.58929336e+00 -2.11261320e+00 1.76461205e-01 9.26490799e-02
-4.92513999e-02 4.77824748e-01 -2.57245690e-01 3.42046261e-01
-4.07864988e-01 2.46663630e-01 -9.42888558e-02 -3.10010701e-01
2.39566386e-01 7.17535764e-02 -4.95939493e-01 -1.37290210e-01
-7.13082105e-02 6.45517170e-01 -1.03676760e+00 -5.80284953e-01
1.94159567e-01 -8.68628640e-03 -1.41284955e+00 6.13762319e-01
-6.47736013e-01 -2.60057330e-01 -1.34223253e-01 8.88297379e-01
2.74458289e-01 -2.54560020e-02 -1.02418803e-01 5.20168722e-01
-1.23095445e-01 8.18301201e-01 -1.24023128e+00 2.12309480e+00
-7.44593740e-01 5.16473413e-01 -8.94042104e-02 -9.32654977e-01
9.06278014e-01 3.07197571e-01 -4.22940478e-02 -4.00827020e-01
-2.41816729e-01 2.45662227e-01 -2.09610566e-01 -8.70966434e-01
7.83281088e-01 3.29723030e-01 -3.63317907e-01 7.17038095e-01
5.38218498e-01 -4.36114162e-01 3.36825758e-01 4.43921506e-01
1.71864748e+00 4.57361937e-01 1.39808461e-01 -7.44011179e-02
4.02407050e-01 3.64138484e-01 2.78427005e-01 1.07329345e+00
-1.85859650e-01 5.30158579e-01 4.61888611e-01 -4.60307032e-01
-9.89836276e-01 -8.27932596e-01 2.96264648e-01 1.87176836e+00
-4.56152797e-01 -8.37161124e-01 -1.14634979e+00 -7.11234868e-01
5.16414493e-02 1.07452238e+00 9.27408189e-02 -3.61018360e-01
-9.20418620e-01 -8.11465383e-02 9.58001971e-01 7.39998519e-01
1.08561799e-01 -1.38513899e+00 -7.02410877e-01 4.62455213e-01
-1.30507588e-01 -8.03619862e-01 -4.34482753e-01 3.11915010e-01
-6.89227641e-01 -9.47515726e-01 -1.36254663e-02 -1.17124474e+00
7.60585248e-01 3.03176463e-01 1.68457150e+00 1.03509831e+00
-6.51680768e-01 6.63261890e-01 -3.30815017e-01 -3.37994039e-01
-6.99773431e-01 1.02510616e-01 -3.46380800e-01 -1.25911486e+00
6.49603367e-01 -7.84816444e-01 -2.28994619e-03 -1.18890770e-01
-6.70603812e-01 2.10056603e-02 3.30338985e-01 5.99152803e-01
1.79998115e-01 -1.05723076e-01 8.97692442e-02 -1.17828608e+00
7.10418701e-01 -4.77790564e-01 -8.75132084e-01 5.24140537e-01
-3.28551471e-01 1.37762800e-01 8.30193162e-01 -4.60365862e-01
-1.07217383e+00 1.56373978e-01 -5.28157651e-01 -4.56023276e-01
-2.86726385e-01 9.34280872e-01 2.83788562e-01 -5.58148861e-01
1.24593580e+00 1.35177583e-01 -4.60367411e-01 -7.02691495e-01
3.36272359e-01 4.99898136e-01 7.73417175e-01 -1.42623472e+00
5.15055358e-01 -8.21119666e-01 -5.23073018e-01 -5.30040264e-01
-1.00375831e+00 -5.24122417e-01 -4.49227035e-01 -7.48300776e-02
3.93290639e-01 -1.02815306e+00 -8.11474502e-01 3.21978480e-01
-1.36529374e+00 -8.85745168e-01 -3.32531869e-01 1.91123169e-02
-3.89152437e-01 3.44387054e-01 -1.11683190e+00 -2.16943696e-01
-2.57555395e-01 -1.55678153e+00 4.30803180e-01 4.16686893e-01
-6.41584873e-01 -8.54204774e-01 3.21168780e-01 8.40038240e-01
5.27937472e-01 -2.40091279e-01 1.21920907e+00 -8.08986127e-01
-8.10581863e-01 -9.04899687e-02 6.69522583e-02 1.25249118e-01
-5.47491014e-01 4.25666541e-01 -8.06580305e-01 2.05611289e-01
-3.27500761e-01 -8.21962059e-01 3.98303509e-01 -2.57373247e-02
1.62032163e+00 -5.59379935e-01 -1.31856397e-01 6.86886072e-01
1.39188635e+00 -2.13409916e-01 4.35760170e-01 5.01927018e-01
9.21819687e-01 4.20262486e-01 4.06358212e-01 2.97952771e-01
5.63847661e-01 6.02420926e-01 7.62442350e-02 6.90375090e-01
-1.13757908e-01 -5.28064311e-01 4.12770391e-01 1.07148945e+00
3.43886942e-01 2.42249072e-01 -1.50059485e+00 9.73593593e-01
-1.74311817e+00 -1.00931776e+00 -2.99981236e-01 1.92414963e+00
1.57294309e+00 6.49298653e-02 -7.83001110e-02 -2.06993874e-02
2.95843512e-01 -4.50058579e-01 8.56385157e-02 -1.04858124e+00
8.13811064e-01 7.85861492e-01 1.63210034e-01 7.25429773e-01
-6.72883093e-01 1.10252774e+00 6.10596466e+00 6.72975183e-01
-1.05097854e+00 4.43280973e-02 -1.21968955e-01 -1.05092837e-03
-3.12399596e-01 1.52204689e-02 -1.19357026e+00 4.60837781e-01
1.42555046e+00 -4.02184948e-02 7.12690532e-01 1.59654808e+00
-3.69331747e-01 -1.04016095e-01 -1.28036165e+00 5.84661484e-01
3.52804624e-02 -1.55675447e+00 -3.64623487e-01 -7.40604162e-01
7.97096431e-01 -1.89943150e-01 -3.47270280e-01 1.16127288e+00
1.06379223e+00 -1.02688372e+00 8.89259160e-01 3.95493239e-01
2.67323881e-01 -6.34429395e-01 5.23700178e-01 5.64952672e-01
-1.00302970e+00 -2.44056821e-01 -3.29542786e-01 -2.96635777e-01
-4.35463279e-01 2.54947454e-01 -1.07338798e+00 -5.55029996e-02
1.13783920e+00 4.62166905e-01 -1.00098109e+00 1.29845369e+00
-6.72774732e-01 9.35757339e-01 -2.49702737e-01 -3.02007049e-01
-4.77472730e-02 2.72914588e-01 3.25743973e-01 1.60093927e+00
4.17245477e-01 4.38877016e-01 7.07342684e-01 1.34025812e+00
-1.46764874e-01 -5.22041060e-02 -3.47414553e-01 -1.97956488e-01
8.75161588e-01 1.16800392e+00 -2.59181708e-01 -4.10551518e-01
-5.53080738e-01 6.80350840e-01 9.03253198e-01 9.58781168e-02
-7.66734779e-01 -9.21037674e-01 6.87977552e-01 5.33283763e-02
3.96973580e-01 -1.84143797e-01 -7.35736728e-01 -1.11206734e+00
6.83183149e-02 -1.05867577e+00 4.21166182e-01 -1.03640723e+00
-8.13209772e-01 -9.19340104e-02 -6.04137927e-02 -7.42378056e-01
-1.01198494e-01 -4.26988274e-01 -1.30467498e+00 9.39989626e-01
-1.29018795e+00 -6.49697483e-01 -5.05191743e-01 2.88405985e-01
6.28859937e-01 -1.09650224e-01 8.90332580e-01 5.12266576e-01
-7.62156308e-01 1.00993836e+00 -4.64498281e-01 1.86159149e-01
6.74493968e-01 -1.66352224e+00 4.03018743e-01 1.06756020e+00
-1.89747423e-01 1.10301566e+00 8.63921702e-01 -5.77233732e-01
-1.33111072e+00 -1.19182551e+00 1.24427915e+00 -7.42078006e-01
6.92095339e-01 -3.84194292e-02 -1.57278395e+00 1.06521738e+00
1.03636891e-01 -1.75259858e-02 9.86442506e-01 2.12778196e-01
-5.37797034e-01 3.82171422e-01 -1.09165084e+00 6.29911482e-01
7.10848093e-01 -6.17321789e-01 -1.33305454e+00 4.12105858e-01
1.00863385e+00 -1.14004517e+00 -1.14568996e+00 -2.11157143e-01
1.16468854e-01 -6.73832774e-01 1.08048403e+00 -6.06977940e-01
1.16934562e+00 -2.12187439e-01 1.32879704e-01 -1.25158477e+00
-1.78922221e-01 -5.79869568e-01 -2.59708446e-02 1.47237742e+00
2.83500105e-01 1.34514630e-01 7.39301741e-01 1.21982682e+00
-9.91653562e-01 -6.38063371e-01 -3.66129637e-01 -4.34430391e-01
2.32167348e-01 -5.80579817e-01 5.00555098e-01 1.17958748e+00
7.46005237e-01 1.05721235e-01 5.51136434e-01 2.49925956e-01
3.86379242e-01 4.85266075e-02 1.02588022e+00 -1.09812510e+00
-5.09890854e-01 -6.17615581e-01 -9.44229737e-02 -7.24501908e-01
4.45668340e-01 -1.12780023e+00 4.28424597e-01 -1.13825154e+00
7.06674829e-02 -7.57862210e-01 1.61035687e-01 1.17438948e+00
-4.09829110e-01 -3.00793648e-01 1.23541594e-01 -1.91808552e-01
-7.28876472e-01 1.07383244e-01 4.68161672e-01 6.93612248e-02
-5.27491927e-01 -3.23602825e-01 -6.19955301e-01 9.61136937e-01
5.23535967e-01 -7.31734157e-01 -1.64916396e-01 -5.64032078e-01
5.83973408e-01 2.85558432e-01 2.40740061e-01 -1.48618162e+00
7.18712568e-01 -4.52906132e-01 -8.94236267e-02 -1.68878749e-01
-2.93628901e-01 -5.60433567e-01 -2.58316070e-01 3.44735652e-01
-7.03521550e-01 2.62906820e-01 7.58788943e-01 -1.82625894e-02
-2.25505486e-01 -1.33590651e+00 8.77365887e-01 -6.56902730e-01
-8.54013920e-01 -4.42146301e-01 -4.16244447e-01 3.11572939e-01
1.01484954e+00 -1.69657528e-01 -7.09784150e-01 2.71981239e-01
-5.02121031e-01 2.53573328e-01 4.36202347e-01 2.22786203e-01
5.08595288e-01 -9.47887659e-01 -2.55648583e-01 1.91947266e-01
2.00585902e-01 2.11168766e-01 1.99015379e-01 4.99953628e-01
-1.21081400e+00 2.18629226e-01 -3.04648072e-01 -5.79956055e-01
-1.72627616e+00 5.77447116e-01 2.05774993e-01 -2.98022330e-01
-8.36667240e-01 1.25848877e+00 -5.76170325e-01 -1.44412911e+00
4.81928736e-01 -7.85802245e-01 8.53037313e-02 -4.25996602e-01
8.35578442e-01 1.74108461e-01 4.12557781e-01 3.42499584e-01
3.54012288e-02 3.45979452e-01 -1.33939475e-01 3.63192141e-01
1.49827611e+00 4.09602016e-01 -4.59624194e-02 1.45213887e-01
8.57228577e-01 1.92096159e-01 -6.65699065e-01 -2.61779547e-01
2.41680831e-01 -4.06457126e-01 -1.54755667e-01 -1.07936442e+00
-6.40727818e-01 1.00466132e+00 1.55510589e-01 -1.50809363e-01
6.30958378e-01 -2.22228482e-01 7.73641586e-01 9.49058473e-01
2.78066516e-01 -9.10691381e-01 3.45185310e-01 8.24699581e-01
4.98223871e-01 -1.21305490e+00 -5.80255575e-02 -4.21728760e-01
-2.26218462e-01 1.43698037e+00 1.37750506e+00 -1.16832085e-01
3.63790661e-01 3.29568148e-01 -2.49796018e-01 -6.23395406e-02
-9.11326289e-01 5.03774822e-01 8.54908302e-02 1.24375813e-01
9.18051958e-01 -7.72274509e-02 1.66916296e-01 9.38868940e-01
-4.99641061e-01 2.98073083e-01 1.16717947e+00 1.71831214e+00
-7.68550396e-01 -1.00471759e+00 -5.52406728e-01 1.33874923e-01
-5.22367597e-01 -5.15502095e-01 -4.83789518e-02 4.93760496e-01
5.39672337e-02 4.59771782e-01 -1.35839880e-01 -2.25674495e-01
6.01462960e-01 7.95734823e-01 5.71163476e-01 -1.44990039e+00
-1.44315791e+00 -8.86980057e-01 -4.66772355e-03 -7.66285241e-01
2.35369965e-01 -6.73009932e-01 -1.48850906e+00 -7.14341164e-01
-1.06870234e-01 -1.17090363e-02 5.53709328e-01 9.04158533e-01
4.01266158e-01 7.91182935e-01 -3.41914773e-01 -5.55056095e-01
-9.41793442e-01 -7.44357407e-01 1.88606203e-01 3.68688375e-01
1.15994647e-01 -2.75588155e-01 -8.25353488e-02 3.93519998e-01] | [9.304601669311523, 7.43756103515625] |
4d366a83-bb7a-4d60-bf29-c5bbb3902bb1 | compete-to-win-enhancing-pseudo-labels-for | 2304.07519 | null | https://arxiv.org/abs/2304.07519v1 | https://arxiv.org/pdf/2304.07519v1.pdf | Compete to Win: Enhancing Pseudo Labels for Barely-supervised Medical Image Segmentation | This study investigates barely-supervised medical image segmentation where only few labeled data, i.e., single-digit cases are available. We observe the key limitation of the existing state-of-the-art semi-supervised solution cross pseudo supervision is the unsatisfactory precision of foreground classes, leading to a degenerated result under barely-supervised learning. In this paper, we propose a novel Compete-to-Win method (ComWin) to enhance the pseudo label quality. In contrast to directly using one model's predictions as pseudo labels, our key idea is that high-quality pseudo labels should be generated by comparing multiple confidence maps produced by different networks to select the most confident one (a compete-to-win strategy). To further refine pseudo labels at near-boundary areas, an enhanced version of ComWin, namely, ComWin+, is proposed by integrating a boundary-aware enhancement module. Experiments show that our method can achieve the best performance on three public medical image datasets for cardiac structure segmentation, pancreas segmentation and colon tumor segmentation, respectively. The source code is now available at https://github.com/Huiimin5/comwin. | ['Kwang-Ting Cheng', 'Yiqun Lin', 'Xiaomeng Li', 'Huimin Wu'] | 2023-04-15 | null | null | null | null | ['tumor-segmentation', 'pancreas-segmentation', 'pseudo-label'] | ['computer-vision', 'medical', 'miscellaneous'] | [ 4.52696383e-01 5.86542487e-01 -5.40903807e-01 -6.73132896e-01
-1.05960321e+00 -3.11268359e-01 2.90021062e-01 1.86179146e-01
-4.81801391e-01 8.69302511e-01 -1.03955142e-01 -3.47025812e-01
-4.57795672e-02 -4.67279673e-01 -5.97158015e-01 -9.19256210e-01
4.16480422e-01 5.44870615e-01 6.64817393e-01 2.75199294e-01
1.07274085e-01 -7.49038160e-03 -9.38195229e-01 4.92168546e-01
1.26921046e+00 8.86282682e-01 2.94372290e-01 4.05266374e-01
-1.28765836e-01 7.87376702e-01 -2.98686892e-01 -4.62776393e-01
3.36988747e-01 -7.01666176e-01 -8.27194810e-01 1.34432793e-01
9.94087458e-02 -1.03138387e-01 1.68185204e-01 1.43000031e+00
3.35939258e-01 -3.68155986e-01 6.80259466e-01 -1.18069935e+00
-3.59400094e-01 1.06538785e+00 -7.47238576e-01 7.30119124e-02
-1.93196058e-01 1.96279436e-02 8.29470277e-01 -6.60248220e-01
6.16843104e-01 7.52196014e-01 6.62925541e-01 5.75681567e-01
-1.07469141e+00 -7.62781739e-01 1.19020998e-01 3.20370905e-02
-1.46669841e+00 -1.12542614e-01 6.73457563e-01 -3.16540271e-01
1.70357391e-01 1.72328457e-01 5.02865434e-01 8.15899730e-01
2.01025665e-01 9.83259499e-01 1.35270929e+00 -5.03083766e-01
1.86588585e-01 2.90089190e-01 1.29098207e-01 9.66625094e-01
3.89465302e-01 1.15658708e-01 -2.11886212e-01 -1.32988408e-01
7.49212503e-01 -2.09086370e-02 -4.42921549e-01 -5.26038349e-01
-1.52587557e+00 6.55749738e-01 4.83990729e-01 4.02479619e-01
-2.22794771e-01 -2.77624846e-01 3.06379318e-01 -6.84515834e-02
3.99427533e-01 3.08301985e-01 -5.50347209e-01 7.44637474e-02
-1.16869545e+00 -1.52902812e-01 6.19655013e-01 9.17006850e-01
5.74341953e-01 -4.69029784e-01 -2.72156298e-01 7.19953597e-01
5.24950087e-01 8.24015141e-02 6.38584673e-01 -8.80989134e-01
2.42345408e-01 6.86013043e-01 -1.29665166e-01 -4.68019098e-01
-5.09180427e-01 -5.81182420e-01 -9.60797668e-01 1.75089180e-01
7.43078530e-01 -2.79619277e-01 -1.19102788e+00 1.48282766e+00
5.15287280e-01 4.53583956e-01 5.04646935e-02 1.13716269e+00
9.51398730e-01 2.48349115e-01 9.67012793e-02 -2.23258659e-01
1.44160390e+00 -1.42965758e+00 -6.70547903e-01 -9.18034315e-02
7.21638143e-01 -6.99026644e-01 7.67841280e-01 4.93968040e-01
-7.28271067e-01 -3.74585807e-01 -1.08739221e+00 3.07864726e-01
-1.21667519e-01 3.48521024e-01 5.11237025e-01 6.47583187e-01
-7.64790535e-01 6.93253696e-01 -1.07345927e+00 -1.49098799e-01
6.50569916e-01 1.85212344e-01 -2.64540315e-01 -6.38079420e-02
-1.02847672e+00 7.88153589e-01 7.91303575e-01 9.67006832e-02
-6.49108112e-01 -6.19284630e-01 -7.56048262e-01 -3.85448068e-01
7.77136028e-01 -3.96339267e-01 1.45732427e+00 -1.14978099e+00
-1.46288550e+00 9.36863959e-01 1.84890494e-01 -4.99264508e-01
8.64025414e-01 1.31848708e-01 -1.67918220e-01 4.42137122e-01
1.94719300e-01 8.99922669e-01 5.92114329e-01 -1.48483062e+00
-7.70127952e-01 -1.51484922e-01 -3.33170682e-01 1.09127559e-01
5.20114824e-02 -1.84999958e-01 -6.99359596e-01 -7.30400681e-01
3.99492115e-01 -9.94884968e-01 -5.96457481e-01 1.21664993e-01
-7.57473171e-01 -5.03138192e-02 3.81623536e-01 -6.62459016e-01
1.12606442e+00 -2.10239768e+00 -1.44713163e-01 2.64149040e-01
3.87152106e-01 3.20217997e-01 2.83763051e-01 -2.12207049e-01
-3.50020491e-02 1.54613331e-01 -6.82382107e-01 -3.83100212e-01
-1.67287260e-01 3.16900223e-01 3.36482018e-01 4.55846161e-01
2.00759217e-01 8.64941955e-01 -1.01579785e+00 -1.18658280e+00
1.97349608e-01 2.44272307e-01 -2.38506243e-01 3.41656543e-02
-1.13516085e-01 6.22518122e-01 -5.20634413e-01 8.86276543e-01
8.28253090e-01 -6.00222647e-01 3.01597625e-01 -2.49365717e-01
6.36778548e-02 6.82845572e-03 -1.23200893e+00 1.71188557e+00
-3.83634605e-02 1.63464602e-02 5.00030592e-02 -9.28388894e-01
8.45924556e-01 3.48824531e-01 5.68889678e-01 -3.99315596e-01
2.85961330e-01 4.29016709e-01 1.01144671e-01 -3.12443107e-01
1.41930670e-01 -2.22029522e-01 1.29985973e-01 2.74682641e-01
5.80719188e-02 -4.58557741e-04 3.01787108e-01 2.71219254e-01
8.69830132e-01 3.30894351e-01 4.72573519e-01 -3.90244812e-01
5.23152053e-01 2.87862092e-01 1.14700103e+00 7.57942080e-01
-6.51999414e-01 9.68782306e-01 6.94465518e-01 -1.25969902e-01
-7.34338701e-01 -8.79626989e-01 -5.01408160e-01 5.17071366e-01
4.14236635e-01 -2.43971884e-01 -9.99385774e-01 -1.31496274e+00
-1.87935770e-01 5.96744239e-01 -6.22198164e-01 -8.44266266e-03
-4.55048382e-01 -1.13356435e+00 5.97042680e-01 5.55980265e-01
6.24454498e-01 -1.08703554e+00 -7.07522452e-01 1.42607436e-01
-3.00083101e-01 -1.07711303e+00 -4.94700164e-01 5.17220974e-01
-1.10327458e+00 -1.29929829e+00 -1.10362136e+00 -8.90454471e-01
9.89857972e-01 -1.48166358e-01 1.03020895e+00 1.30125090e-01
-2.17145681e-01 -2.36753047e-01 -5.48908651e-01 -1.46006569e-01
-6.93450272e-01 1.52979359e-01 -3.68347883e-01 -4.94071618e-02
1.48190215e-01 -7.05027059e-02 -7.62325466e-01 5.00823438e-01
-6.94806457e-01 4.98462498e-01 9.52377558e-01 1.07559669e+00
1.04750073e+00 -7.21518397e-02 6.01063788e-01 -1.38263226e+00
8.97317380e-02 -2.80954897e-01 -5.40529490e-01 3.95931989e-01
-9.79633749e-01 1.58090387e-02 5.79734147e-01 -4.87161547e-01
-1.20659935e+00 4.53974754e-01 -2.72613704e-01 -2.09716216e-01
-3.82256716e-01 4.20736164e-01 8.06158036e-02 3.54441628e-02
4.37448472e-01 -3.15905781e-03 6.05612844e-02 -4.53127861e-01
1.66764230e-01 6.49285614e-01 5.55083394e-01 -3.08947951e-01
4.72125590e-01 3.38238329e-01 -1.59707427e-01 -1.52816683e-01
-1.02423704e+00 -5.55449545e-01 -6.95234120e-01 -1.53632522e-01
9.63435233e-01 -7.81653643e-01 -1.71661153e-01 5.11648476e-01
-7.59266198e-01 -5.32451689e-01 -2.96502769e-01 7.40020812e-01
-3.55961084e-01 5.30386806e-01 -8.42740417e-01 -4.53488529e-01
-3.87875289e-01 -1.30731976e+00 9.92497683e-01 5.18614948e-01
1.34126246e-01 -8.13465714e-01 -6.66967481e-02 4.77543235e-01
1.50866419e-01 3.63464475e-01 6.45627379e-01 -9.46895599e-01
-4.05876577e-01 -6.26261309e-02 -4.38885033e-01 4.45535511e-01
1.69993415e-01 -2.95289883e-05 -7.89325356e-01 -1.15846038e-01
-8.50971490e-02 -2.86316812e-01 1.01188290e+00 5.87696970e-01
1.17289472e+00 1.38249993e-01 -5.78566790e-01 5.68309367e-01
1.31493402e+00 2.54371405e-01 3.77781570e-01 2.68233299e-01
5.98233581e-01 3.59241992e-01 1.01063800e+00 2.63245851e-01
3.77352804e-01 5.08839726e-01 3.92017066e-01 -6.18388176e-01
-2.14064687e-01 -1.80943161e-01 -3.11155198e-03 8.46840322e-01
1.65686950e-01 -7.63881728e-02 -9.22492266e-01 3.87449503e-01
-1.89412522e+00 -2.92250127e-01 -2.55142719e-01 2.04567623e+00
1.14471090e+00 5.55463254e-01 5.85747324e-02 8.72419029e-02
9.93102372e-01 -1.22477703e-01 -5.06814420e-01 2.11642295e-01
9.86440480e-02 5.32419682e-02 6.63816035e-01 3.94664437e-01
-1.30338943e+00 7.22344577e-01 5.12031937e+00 1.02104008e+00
-9.08511877e-01 3.90907139e-01 1.07652366e+00 3.54598373e-01
5.06087914e-02 7.62129799e-02 -7.13234723e-01 6.16455257e-01
5.71286976e-01 1.31745726e-01 -1.88004598e-01 9.18769479e-01
9.79247093e-02 -4.10454988e-01 -8.72468591e-01 7.65043914e-01
1.02590106e-01 -1.21916425e+00 -2.43202254e-01 -2.16282070e-01
7.25450695e-01 -5.48046008e-02 -3.28591764e-01 9.06310156e-02
2.45358676e-01 -5.90990543e-01 6.04330599e-01 4.37533677e-01
8.01732898e-01 -5.10484576e-01 1.19353747e+00 5.02912879e-01
-9.71611977e-01 2.25522250e-01 -1.63564295e-01 5.46295166e-01
3.05517405e-01 9.01923597e-01 -9.52324867e-01 8.25188160e-01
6.51015043e-01 6.27055764e-01 -6.79033875e-01 1.36172271e+00
-4.83503908e-01 8.87826085e-01 -2.66428322e-01 5.72372414e-02
1.58283427e-01 -1.16070673e-01 2.13301718e-01 1.27982843e+00
7.51505196e-02 1.70908183e-01 4.14974093e-01 7.74533570e-01
4.25751656e-02 1.46574527e-01 3.44465710e-02 1.63366899e-01
2.23342344e-01 1.57510078e+00 -1.47542608e+00 -5.95693827e-01
-2.93266714e-01 8.78303885e-01 -2.87289848e-03 -3.00003006e-03
-1.00528049e+00 -3.46899122e-01 -3.00645381e-01 -1.43649066e-02
1.52512848e-01 3.02789152e-01 -4.21716422e-01 -1.07074559e+00
-1.46312922e-01 -7.82107711e-01 6.45252705e-01 -4.76199478e-01
-1.28620505e+00 7.64480650e-01 9.89497975e-02 -1.33669603e+00
8.30419064e-02 -5.69318175e-01 -5.15418947e-01 5.04216433e-01
-1.61680317e+00 -1.08292353e+00 -4.31446016e-01 3.45021814e-01
4.72238958e-01 1.05152287e-01 5.45640945e-01 4.77578253e-01
-6.50850296e-01 6.55758262e-01 -7.58025125e-02 2.67848253e-01
9.91177380e-01 -1.50116158e+00 -2.06440175e-03 7.96197474e-01
1.32088661e-01 8.65847468e-02 4.66169000e-01 -7.19101548e-01
-5.37255824e-01 -1.08105671e+00 6.85386360e-01 -6.02057800e-02
2.97742844e-01 -1.52530000e-01 -9.17876184e-01 5.19669354e-01
1.09983183e-01 2.65574843e-01 6.28700554e-01 -3.61445606e-01
1.42678827e-01 -3.34206335e-02 -1.37482584e+00 3.96173835e-01
7.50789762e-01 1.63365096e-01 -5.75946808e-01 2.99420267e-01
7.51255810e-01 -7.29510665e-01 -8.49165857e-01 8.14639032e-01
3.28892857e-01 -9.60313737e-01 5.97881258e-01 -1.61916073e-02
4.72335190e-01 -6.35542631e-01 2.09695637e-01 -1.02563536e+00
-5.29475547e-02 -3.15030992e-01 1.30991161e-01 1.25304222e+00
7.90147364e-01 -6.52623773e-01 9.99409437e-01 4.13381159e-01
-3.78940374e-01 -1.10006058e+00 -8.74913454e-01 -5.96335828e-01
-3.10142022e-02 -2.02281743e-01 3.88751119e-01 1.01758850e+00
-8.88467729e-02 -8.98091123e-02 -1.89643756e-01 2.13433951e-01
7.12572575e-01 2.37215549e-01 2.49746874e-01 -1.08091509e+00
-5.14723063e-01 -3.26905102e-01 -1.49562165e-01 -9.47886229e-01
-1.71223670e-01 -1.02456427e+00 4.09277588e-01 -1.52309990e+00
4.87757266e-01 -8.33418190e-01 -6.71358526e-01 7.37514853e-01
-3.53420913e-01 5.43048918e-01 1.25992104e-01 1.29161626e-01
-8.46728861e-01 3.08063269e-01 1.57888722e+00 4.17396277e-02
-8.15094188e-02 2.78071761e-01 -7.49766350e-01 9.10837233e-01
9.74404216e-01 -8.25411677e-01 -1.58070162e-01 -5.03276996e-02
-3.16682875e-01 1.63141400e-01 2.60762423e-01 -9.55850065e-01
1.95856392e-01 -6.59274831e-02 4.07120705e-01 -6.98325455e-01
-1.22602090e-01 -6.61904573e-01 -2.25890931e-02 7.71115422e-01
-4.21856076e-01 -9.94178727e-02 -8.06373209e-02 4.20255750e-01
-3.21330816e-01 -6.81183755e-01 1.04118145e+00 -2.80145824e-01
-5.13085365e-01 1.99422538e-01 -7.06828460e-02 1.43637210e-01
1.15462959e+00 -9.00800675e-02 -2.30844826e-01 4.49508354e-02
-8.60528111e-01 4.69076425e-01 2.96233654e-01 8.03999156e-02
4.28736538e-01 -1.08904517e+00 -7.19209552e-01 -3.75597342e-03
1.30502969e-01 3.81329983e-01 1.89221740e-01 1.17754209e+00
-5.61109722e-01 1.69394046e-01 6.96504191e-02 -8.42148900e-01
-1.26189160e+00 4.22896504e-01 4.33550030e-01 -6.34625673e-01
-7.54744887e-01 8.75812590e-01 1.62256539e-01 -5.94173908e-01
2.55611628e-01 -6.01846159e-01 -1.59376606e-01 -1.39973685e-01
2.95899600e-01 9.80186760e-02 9.57922712e-02 -4.00661349e-01
-4.67125356e-01 3.80627692e-01 -2.65707076e-01 9.24137235e-02
9.60048378e-01 8.81688949e-03 2.40345392e-02 2.75105596e-01
7.38247097e-01 -1.89009562e-01 -1.45294297e+00 -2.87083954e-01
2.07255170e-01 -3.23886752e-01 7.34515861e-02 -1.20831323e+00
-1.39808893e+00 5.97735643e-01 8.65922511e-01 -1.65119261e-01
1.17393398e+00 1.61019847e-01 6.74488246e-01 -7.38775264e-03
2.95554370e-01 -1.08260429e+00 -5.02122119e-02 3.26090772e-03
2.80929834e-01 -1.61397970e+00 5.94541430e-02 -7.81294286e-01
-1.05496514e+00 9.12376523e-01 7.63256669e-01 9.52645466e-02
6.75791085e-01 5.39018631e-01 4.48278010e-01 -2.21173428e-02
-5.04692316e-01 -1.02380440e-01 2.72590220e-01 2.44031638e-01
3.90392661e-01 2.50017852e-01 -5.30693352e-01 7.37884462e-01
1.04432724e-01 1.38646394e-01 3.50214869e-01 9.57806349e-01
-2.84575403e-01 -1.25898850e+00 -3.83002639e-01 7.27171302e-01
-7.13039994e-01 -6.64976835e-02 -6.63796142e-02 7.58611739e-01
4.79802996e-01 7.79672801e-01 -2.00152740e-01 -1.68662637e-01
4.06315587e-02 -3.52896005e-02 2.56203175e-01 -6.08902514e-01
-5.87982297e-01 3.51255208e-01 -5.52626662e-02 -4.35854614e-01
-6.72419131e-01 -6.77116096e-01 -1.59357536e+00 3.46064270e-01
-6.92505002e-01 2.33529732e-01 4.19978261e-01 8.59830439e-01
-4.87013342e-04 7.37252712e-01 3.65926504e-01 -4.64214742e-01
-5.27066767e-01 -9.52451527e-01 -4.26253825e-01 3.07726860e-01
1.29480079e-01 -7.06186473e-01 -2.22749233e-01 9.71143022e-02] | [14.68511962890625, -2.124584674835205] |
bfffe3f9-b7aa-4ce0-bbd1-6b512dd363c2 | indoor-localization-and-multi-person-tracking | 2305.05062 | null | https://arxiv.org/abs/2305.05062v1 | https://arxiv.org/pdf/2305.05062v1.pdf | Indoor Localization and Multi-person Tracking Using Privacy Preserving Distributed Camera Network with Edge Computing | Localization of individuals in a built environment is a growing research topic. Estimating the positions, face orientation (or gaze direction) and trajectories of people through space has many uses, such as in crowd management, security, and healthcare. In this work, we present an open-source, low-cost, scalable and privacy-preserving edge computing framework for multi-person localization, i.e. estimating the positions, orientations, and trajectories of multiple people in an indoor space. Our computing framework consists of 38 Tensor Processing Unit (TPU)-enabled edge computing camera systems placed in the ceiling of the indoor therapeutic space. The edge compute systems are connected to an on-premise fog server through a secure and private network. A multi-person detection algorithm and a pose estimation model run on the edge TPU in real-time to collect features which are used, instead of raw images, for downstream computations. This ensures the privacy of individuals in the space, reduces data transmission/storage and improves scalability. We implemented a Kalman filter-based multi-person tracking method and a state-of-the-art body orientation estimation method to determine the positions and facing orientations of multiple people simultaneously in the indoor space. For our study site with size of 18,000 square feet, our system demonstrated an average localization error of 1.41 meters, a multiple-object tracking accuracy score of 62%, and a mean absolute body orientation error of 29{\deg}, which is sufficient for understanding group activity behaviors in indoor environments. Additionally, our study provides practical guidance for deploying the proposed system by analyzing various elements of the camera installation with respect to tracking accuracy. | ['Gari D. Clifford', 'Craig M. Zimring', 'Leandro Miletto Tonetto', 'Robert Tweedy', 'ArjunSinh Nakum', 'Ratan Singh', 'Venkata Siva Krishna Madala', 'Yashar Kiarashi', 'Chaitra Hedge', 'Hyeokhyen Kwon'] | 2023-05-08 | null | null | null | null | ['multiple-object-tracking', 'indoor-localization', 'human-detection', 'edge-computing'] | ['computer-vision', 'computer-vision', 'computer-vision', 'time-series'] | [-3.46244216e-01 -4.53486621e-01 5.82275510e-01 -1.36702359e-01
-1.53936550e-01 -6.15407228e-01 -2.58749157e-01 1.09094962e-01
-4.95315820e-01 5.87090909e-01 -4.68911678e-02 -1.34443969e-01
1.46219864e-01 -7.98138559e-01 -5.25373876e-01 -8.19418013e-01
-1.48898453e-01 1.84536740e-01 2.03067750e-01 2.72577971e-01
-2.55127877e-01 4.39031571e-01 -1.44449794e+00 -2.52743006e-01
5.10485172e-01 1.24089026e+00 -3.66269320e-01 9.65177894e-01
8.02031577e-01 2.10836709e-01 -5.46018481e-01 -5.16059339e-01
2.82755882e-01 2.94444293e-01 9.05867219e-02 9.73713323e-02
5.25370657e-01 -6.30147576e-01 -4.44836259e-01 6.02737069e-01
9.14308846e-01 1.66288927e-01 1.23871543e-01 -1.48309553e+00
-2.75771648e-01 -4.52955037e-01 -4.71147060e-01 1.58866554e-01
8.11105371e-01 5.47978431e-02 2.55864173e-01 -4.98786002e-01
1.86410367e-01 5.88679612e-01 1.19596684e+00 4.45454597e-01
-5.22317827e-01 -7.43888199e-01 -3.13721411e-03 9.77343842e-02
-2.05060267e+00 -4.11920846e-01 3.29533756e-01 -3.80344987e-01
5.21937847e-01 7.25608289e-01 1.12945139e+00 8.22729707e-01
4.27898735e-01 1.21918760e-01 8.77301157e-01 -1.44143417e-01
4.58417326e-01 3.91597688e-01 7.83402249e-02 8.40059757e-01
8.21436107e-01 -1.14337280e-01 -7.44576514e-01 -4.08032984e-01
6.58426046e-01 6.10702038e-01 -2.55031019e-01 -4.30705249e-02
-9.87753451e-01 2.56503999e-01 4.67309982e-01 -3.68215959e-04
-3.78039658e-01 2.89103270e-01 -1.69900358e-02 -2.84878880e-01
5.73266745e-01 -3.62939775e-01 -2.21142828e-01 -2.67840832e-01
-6.88818455e-01 2.66032338e-01 7.13931561e-01 1.38737154e+00
3.74386013e-01 -3.94327462e-01 -2.90756404e-01 -8.66802335e-02
7.13294089e-01 1.07968760e+00 8.17876011e-02 -4.92384106e-01
5.34320891e-01 6.39196098e-01 4.80576754e-01 -1.46447694e+00
-6.63617790e-01 -4.22007948e-01 -6.65451467e-01 -2.56195873e-01
4.15504277e-01 -7.91759610e-01 -1.90245345e-01 1.41683853e+00
1.03708708e+00 4.57568198e-01 -3.42745870e-01 1.14183307e+00
6.71528161e-01 2.20594868e-01 -1.13443442e-01 -1.72661856e-01
1.80947614e+00 -6.43612862e-01 -5.09927869e-01 -2.36270398e-01
3.63447756e-01 -5.89709282e-01 4.37098682e-01 1.79193486e-02
-7.80168831e-01 -2.89241880e-01 -8.04737031e-01 1.97396159e-01
-3.60524058e-01 4.83111978e-01 5.66795647e-01 1.37537658e+00
-1.07389152e+00 1.17076308e-01 -1.12919843e+00 -7.02212870e-01
2.20743060e-01 7.78985798e-01 -2.59988546e-01 -9.99161042e-03
-6.90736115e-01 2.73385286e-01 -3.95305455e-01 3.41156274e-01
-3.38122427e-01 -4.67960775e-01 -7.24880695e-01 -1.10842451e-01
1.06733255e-01 -1.11958146e+00 7.79607832e-01 -1.99195445e-01
-1.09260333e+00 5.18635869e-01 -4.59058613e-01 -1.61260977e-01
6.87080443e-01 -3.58104140e-01 -6.86211705e-01 2.64399853e-02
2.50834525e-01 -9.03171301e-02 7.00492501e-01 -6.78015411e-01
-7.44943917e-01 -1.10525262e+00 -2.01298818e-01 3.70409280e-01
-5.45150638e-01 1.18443981e-01 -4.97420400e-01 -1.62493601e-01
2.61522919e-01 -1.30302095e+00 -1.19980812e-01 2.15481699e-01
-3.68054301e-01 3.00760359e-01 8.98639977e-01 -8.90848458e-01
1.21104014e+00 -2.20956230e+00 -5.01658916e-01 3.85622412e-01
4.51858342e-01 -1.26508877e-01 8.39220703e-01 1.59206003e-01
6.08709216e-01 -2.72655994e-01 4.10770476e-01 -8.86288822e-01
-1.84394613e-01 -3.40527475e-01 2.42521867e-01 1.07001555e+00
-9.07712817e-01 5.46236873e-01 -7.34417081e-01 -4.28796560e-01
4.25314933e-01 7.15903938e-01 -5.43757319e-01 3.87457339e-03
6.95727289e-01 7.19207466e-01 -6.31413221e-01 1.01635003e+00
8.40380192e-01 -1.57707840e-01 9.13016573e-02 -4.33606468e-03
-2.43482426e-01 -3.48156899e-01 -1.74733317e+00 1.30296254e+00
-2.75667340e-01 3.76197636e-01 4.08116043e-01 1.23499259e-02
6.55168116e-01 5.34189522e-01 6.92113340e-01 5.16802892e-02
5.52577913e-01 -2.90599838e-02 -6.00675642e-01 -4.93443787e-01
8.00841987e-01 2.88199991e-01 -2.71453887e-01 4.01329726e-01
-4.72922176e-01 6.81604981e-01 -1.95765302e-01 -6.12687878e-02
1.23903942e+00 -1.88377842e-01 1.12308204e-01 -3.42034362e-02
4.22017574e-01 -4.03418303e-01 5.88953793e-01 5.74361145e-01
-6.21885657e-01 2.66849488e-01 -7.14293540e-01 -4.85219002e-01
-6.05585396e-01 -1.07498395e+00 1.39642535e-02 8.50486875e-01
6.34112000e-01 -8.18324268e-01 -8.61555159e-01 -3.21311831e-01
2.83087641e-01 1.57175791e-02 -1.63916558e-01 9.96847525e-02
-4.06776696e-01 -7.65593112e-01 4.74234611e-01 6.27726912e-01
8.88742030e-01 -2.01565504e-01 -9.45673406e-01 -8.82755872e-03
-3.90091121e-01 -1.25502074e+00 -8.36445630e-01 -7.55695760e-01
-4.95188892e-01 -1.05126655e+00 -5.10698795e-01 -4.68212277e-01
1.11995840e+00 5.85214317e-01 4.41315264e-01 1.33288234e-01
-3.16544741e-01 1.00531912e+00 2.00997945e-02 -5.54224610e-01
7.22604096e-01 -2.85237700e-01 7.90789306e-01 5.13419986e-01
6.21609509e-01 -5.45559108e-01 -1.02106380e+00 6.61139905e-01
1.04567170e-01 -2.01597899e-01 -7.67120346e-02 7.48690441e-02
4.42564994e-01 3.19006830e-01 -1.21637583e-01 -1.64986670e-01
3.62943292e-01 -4.57269907e-01 -8.01206291e-01 1.83947742e-01
-3.38815540e-01 -8.43542874e-01 2.98084199e-01 -2.00276941e-01
-7.25841045e-01 3.51003438e-01 1.56773910e-01 -1.32450134e-01
-1.99681178e-01 -3.36172581e-01 -1.98413894e-01 -6.20060682e-01
2.22312748e-01 1.43570498e-01 -4.06580359e-01 -2.06459120e-01
7.89159313e-02 1.27258062e+00 3.46375495e-01 -1.16229042e-01
7.50536382e-01 8.87174666e-01 1.38662368e-01 -1.18877256e+00
-2.79904664e-01 -8.73239934e-01 -5.13774395e-01 -6.70283437e-01
8.97806585e-01 -1.45684242e+00 -1.76568294e+00 7.07589030e-01
-1.02387547e+00 3.05934310e-01 2.34279171e-01 6.20829403e-01
-2.84286235e-02 2.02349007e-01 -4.35232699e-01 -1.47123003e+00
-6.46663547e-01 -7.46922493e-01 1.37786889e+00 7.08987057e-01
-1.65327847e-01 -8.72296512e-01 -1.80612564e-01 9.19187605e-01
5.36255598e-01 3.58076930e-01 -4.70411330e-01 -4.69257794e-02
-8.62139225e-01 -6.91566467e-01 1.20214500e-01 -5.08437753e-01
1.39112268e-02 -4.46873665e-01 -1.02762532e+00 -7.14952707e-01
9.91812348e-02 4.61468756e-01 -4.63579409e-02 7.28690088e-01
1.03858054e+00 -4.83374715e-01 -9.61816609e-01 8.51198316e-01
1.25984430e+00 -1.65757276e-02 2.71507025e-01 -3.62058356e-02
9.02212203e-01 7.03741238e-02 4.57844466e-01 9.10038948e-01
9.10166979e-01 7.68402815e-01 2.51571149e-01 1.93433389e-01
3.06846499e-01 -9.95076522e-02 3.27473134e-01 4.59986955e-01
-5.20368218e-01 -3.70851874e-01 -6.10714555e-01 2.05911800e-01
-1.78574598e+00 -8.57707739e-01 -3.87681901e-01 2.56985688e+00
4.91625443e-03 -2.67172545e-01 4.27107215e-01 -1.01804607e-01
1.11992776e+00 -2.37250045e-01 -3.45006585e-01 1.76631823e-01
4.71201837e-01 -3.66468787e-01 1.14994299e+00 2.88552642e-01
-1.27778447e+00 4.13990945e-01 5.48276138e+00 7.07954764e-02
-9.08764482e-01 3.60449731e-01 3.23645949e-01 -4.49762642e-01
4.36425179e-01 -3.83455455e-01 -1.29121959e+00 8.07737708e-01
9.13546443e-01 1.39124408e-01 4.39224511e-01 1.17959738e+00
2.54898787e-01 -2.62362152e-01 -6.54944181e-01 1.46254647e+00
1.95997655e-01 -1.04578054e+00 -1.04500639e+00 5.89047909e-01
5.32056630e-01 -3.29462923e-02 -8.43484923e-02 -2.06884474e-01
-1.38304859e-01 -3.93874764e-01 6.61407948e-01 4.90606695e-01
6.52564526e-01 -7.09400594e-01 7.87257671e-01 4.94948775e-01
-1.86796665e+00 -2.42856815e-01 -1.75264791e-01 -5.24024665e-01
4.90299553e-01 6.24565423e-01 -7.41340101e-01 3.99924785e-01
1.09191430e+00 2.96226084e-01 -5.49181998e-01 1.22215927e+00
-9.78695825e-02 3.26603085e-01 -8.98934960e-01 -4.18217033e-01
-4.74998742e-01 -3.58282536e-01 6.76267982e-01 7.46100485e-01
7.46016562e-01 3.46050322e-01 3.28791410e-01 4.15809125e-01
5.34636788e-02 -7.09067658e-02 -7.07621932e-01 5.60351551e-01
8.48331392e-01 1.55874979e+00 -7.27912128e-01 -2.53451411e-02
-3.47692698e-01 1.10854113e+00 -2.28677586e-01 2.13120058e-01
-1.06246543e+00 -3.17228585e-01 8.91802669e-01 5.27464807e-01
2.60380447e-01 -5.55892110e-01 -2.84331769e-01 -1.20943367e+00
4.60659444e-01 -1.88905507e-01 4.10222620e-01 -5.11417508e-01
-7.96252608e-01 3.50860208e-01 -3.40609491e-01 -1.38801873e+00
2.29660869e-02 -4.25686747e-01 -5.96327364e-01 6.79557502e-01
-7.09689498e-01 -1.43614697e+00 -1.00456905e+00 8.69319677e-01
-1.93429425e-01 1.44261483e-03 8.43589246e-01 6.70516312e-01
-9.84732091e-01 8.28971207e-01 2.14041442e-01 2.84889966e-01
3.17735136e-01 -8.88689876e-01 2.45354414e-01 1.23634076e+00
-1.14692546e-01 8.59375358e-01 5.41653633e-01 -1.04172862e+00
-1.81028509e+00 -1.23448443e+00 9.85115349e-01 -7.84404874e-01
8.50232393e-02 -7.98892856e-01 8.28890223e-03 9.71761882e-01
-3.51640731e-01 4.80189502e-01 1.06595528e+00 6.46762177e-02
4.83163685e-01 -3.67729962e-01 -1.69527352e+00 5.38908422e-01
1.13541329e+00 -3.05551171e-01 1.33808941e-01 5.38151681e-01
2.65060604e-01 -8.64759684e-01 -8.42405558e-01 -1.57626987e-01
8.61555815e-01 -8.17588925e-01 1.11580896e+00 1.43085018e-01
-7.31262147e-01 -6.91344976e-01 -2.18294576e-01 -7.29400516e-01
-3.54094982e-01 -6.75035417e-01 -3.61648947e-01 1.38368177e+00
-2.06757188e-01 -9.32611883e-01 1.14530432e+00 1.44535160e+00
1.52687848e-01 -4.54628229e-01 -1.23360944e+00 -4.84071106e-01
-1.14573109e+00 -4.81042802e-01 7.39729822e-01 2.28801042e-01
-8.43592584e-02 1.02575190e-01 -4.52739865e-01 9.65482950e-01
1.03712964e+00 -1.16911426e-01 1.02666891e+00 -1.00248051e+00
-1.54386103e-01 5.21974862e-01 -9.52271640e-01 -8.07109535e-01
-4.29407686e-01 -3.73588741e-01 -2.62956381e-01 -1.29917824e+00
-7.57880360e-02 -4.90927994e-01 1.51498199e-01 2.53302783e-01
4.31980267e-02 4.43339795e-01 -5.83197922e-02 -2.95770206e-02
-8.45362723e-01 2.77357876e-01 5.62487662e-01 9.49808285e-02
-1.88388497e-01 7.43480742e-01 -5.02990782e-01 8.45360816e-01
6.39554858e-01 -3.41471583e-01 -1.13731019e-01 -4.43503380e-01
1.58966463e-02 -2.06325665e-01 7.38847315e-01 -1.66498721e+00
7.28945673e-01 1.21494375e-01 9.33543026e-01 -6.33728087e-01
8.21596146e-01 -1.31765950e+00 4.82454181e-01 6.55130863e-01
6.96948469e-01 2.76002824e-01 4.54796925e-02 6.08007610e-01
5.45554638e-01 2.63262063e-01 2.29783788e-01 -1.28704697e-01
-4.37339872e-01 5.27262390e-01 -2.06357881e-01 -5.17090797e-01
1.60513163e+00 -4.55712110e-01 -3.84912521e-01 -4.00848955e-01
-5.08764267e-01 3.17229718e-01 6.53626502e-01 4.62500602e-02
5.52430212e-01 -1.33144391e+00 -2.56872207e-01 5.71096301e-01
2.37480979e-02 -1.36697888e-01 5.34690499e-01 9.41015780e-01
-5.92829525e-01 4.29371059e-01 5.19584045e-02 -6.38446689e-01
-1.85664678e+00 1.74767211e-01 3.19313407e-01 2.44005576e-01
-5.40955126e-01 9.16859448e-01 -4.58309770e-01 -2.58313686e-01
2.41815150e-01 -2.29002476e-01 2.15385571e-01 -3.23876172e-01
9.69575465e-01 1.03684783e+00 1.04130447e-01 -9.13559437e-01
-9.15873170e-01 8.63538563e-01 4.89844441e-01 -7.03602955e-02
9.82969701e-01 -7.41544247e-01 1.02166004e-01 -6.22394271e-02
6.35420382e-01 4.57926035e-01 -1.08885896e+00 1.53931975e-01
-5.95599890e-01 -7.97805429e-01 3.99344191e-02 -2.97175527e-01
-9.64314640e-01 3.01294565e-01 1.18076336e+00 4.80175093e-02
1.05083525e+00 -1.06999524e-01 8.92722070e-01 9.48896855e-02
1.09324718e+00 -1.04973209e+00 -5.25990605e-01 5.55627840e-03
1.40912071e-01 -1.11858582e+00 2.63385624e-01 -5.24744868e-01
-3.11272949e-01 6.76338077e-01 7.37495601e-01 2.03168064e-01
7.84515202e-01 2.22645357e-01 -9.51404199e-02 -2.66100228e-01
-5.92942052e-02 4.43001976e-03 1.30439430e-01 8.40352058e-01
3.01311873e-02 4.35764760e-01 5.70192747e-02 7.78866887e-01
-5.05532563e-01 7.31001794e-02 1.37387678e-01 1.14742696e+00
-2.66476989e-01 -4.93477464e-01 -9.72535372e-01 3.61809343e-01
-2.97344863e-01 3.10962498e-01 6.68251887e-02 2.80803353e-01
5.61078072e-01 1.36791611e+00 1.65786326e-01 -6.09915674e-01
3.51535589e-01 -2.28565559e-01 2.97112197e-01 -3.01401794e-01
-7.10129440e-01 -1.89503416e-01 3.85810696e-02 -6.65643036e-01
-7.23187625e-02 -8.81967545e-01 -9.31102693e-01 -7.79640675e-01
-3.19134355e-01 1.23352604e-02 9.86050725e-01 9.57771003e-01
7.35466301e-01 3.84019315e-01 6.28521442e-01 -8.16321790e-01
5.54236509e-02 -6.69975162e-01 -8.84861946e-01 4.51280326e-02
1.77926645e-01 -5.78418612e-01 -1.14184946e-01 -1.82418391e-01] | [6.96907901763916, 0.30729493498802185] |
40a8cbfd-ff02-45f3-8cec-0c53675a6270 | generating-and-weighting-semantically | 2212.04097 | null | https://arxiv.org/abs/2212.04097v1 | https://arxiv.org/pdf/2212.04097v1.pdf | Generating and Weighting Semantically Consistent Sample Pairs for Ultrasound Contrastive Learning | Well-annotated medical datasets enable deep neural networks (DNNs) to gain strong power in extracting lesion-related features. Building such large and well-designed medical datasets is costly due to the need for high-level expertise. Model pre-training based on ImageNet is a common practice to gain better generalization when the data amount is limited. However, it suffers from the domain gap between natural and medical images. In this work, we pre-train DNNs on ultrasound (US) domains instead of ImageNet to reduce the domain gap in medical US applications. To learn US image representations based on unlabeled US videos, we propose a novel meta-learning-based contrastive learning method, namely Meta Ultrasound Contrastive Learning (Meta-USCL). To tackle the key challenge of obtaining semantically consistent sample pairs for contrastive learning, we present a positive pair generation module along with an automatic sample weighting module based on meta-learning. Experimental results on multiple computer-aided diagnosis (CAD) problems, including pneumonia detection, breast cancer classification, and breast tumor segmentation, show that the proposed self-supervised method reaches state-of-the-art (SOTA). The codes are available at https://github.com/Schuture/Meta-USCL. | ['Li Liu', 'Chris H. Q. Ding', 'Chunhui Zhang', 'Yixiong Chen'] | 2022-12-08 | null | null | null | null | ['tumor-segmentation', 'pneumonia-detection'] | ['computer-vision', 'medical'] | [ 5.65864265e-01 1.88182831e-01 -5.29812813e-01 -4.18021858e-01
-1.25568855e+00 -4.88350168e-02 2.24038497e-01 1.50513306e-01
-3.98134619e-01 5.07114410e-01 1.45924017e-01 -4.32256967e-01
1.14483421e-03 -9.39281285e-01 -8.38315964e-01 -7.25827813e-01
6.07764982e-02 4.21529800e-01 2.67355025e-01 -1.14921033e-01
-5.53747416e-02 2.16200680e-01 -1.38161826e+00 6.60343885e-01
1.21584189e+00 1.14686084e+00 6.79837883e-01 5.92237473e-01
-2.80596167e-01 9.92558181e-01 -1.27799928e-01 -2.09348485e-01
1.23772167e-01 -6.83342218e-01 -9.53737020e-01 6.10361546e-02
4.47270602e-01 -6.28295898e-01 -3.08775932e-01 1.16752195e+00
8.00257385e-01 -4.44671959e-02 7.74673104e-01 -1.09061313e+00
-5.50090134e-01 6.04836702e-01 -6.16344869e-01 4.57445145e-01
-2.49519527e-01 1.94837973e-01 6.67136133e-01 -6.56877995e-01
7.27285206e-01 8.94299805e-01 5.27263761e-01 1.04080606e+00
-6.71140075e-01 -7.15096772e-01 -1.87036455e-01 4.31422204e-01
-7.72657335e-01 -1.95507362e-01 7.79791057e-01 -3.56143624e-01
3.23977321e-01 3.45493078e-01 6.22824132e-01 1.14240694e+00
6.02151901e-02 1.23881018e+00 8.99228096e-01 -4.15332109e-01
2.25304484e-01 6.00618683e-02 1.37312725e-01 1.06212926e+00
1.79500848e-01 6.00477867e-02 -4.93592024e-02 -1.40783906e-01
9.48197603e-01 2.43393749e-01 -4.57462341e-01 -5.84208310e-01
-1.14536166e+00 7.07803845e-01 7.61649191e-01 6.76381767e-01
-5.16063631e-01 -7.77424313e-03 6.59312606e-01 -2.48515774e-02
5.88561952e-01 3.60454947e-01 -3.90608370e-01 1.75740317e-01
-7.00107336e-01 -5.08630238e-02 5.32514513e-01 7.12727129e-01
4.40160245e-01 -1.78498030e-01 -2.88411766e-01 1.12875402e+00
-4.99983020e-02 4.36012566e-01 1.07213151e+00 -7.17142403e-01
4.00783390e-01 6.42871857e-01 -2.06270292e-01 -7.22987175e-01
-4.44020480e-01 -7.13415444e-01 -1.30440712e+00 -1.33698300e-01
2.17248157e-01 -8.50775987e-02 -1.31077898e+00 1.57026803e+00
4.26126778e-01 6.55493796e-01 1.92405388e-01 9.83500063e-01
1.37531614e+00 4.00631458e-01 1.49260029e-01 -2.70616084e-01
1.37402272e+00 -1.12274742e+00 -4.79763925e-01 -1.54593587e-01
8.95671725e-01 -3.36548954e-01 9.87213910e-01 1.55134454e-01
-1.08914959e+00 -5.77896833e-01 -1.03173947e+00 9.71009955e-02
-5.75948283e-02 1.54193610e-01 4.89419192e-01 4.01240379e-01
-7.71194696e-01 5.93035698e-01 -1.10420191e+00 -1.62272751e-01
9.71792400e-01 2.30922014e-01 -1.97302714e-01 -5.00081122e-01
-1.13875425e+00 7.62663484e-01 4.83839840e-01 -1.20398328e-01
-9.26103592e-01 -1.16408849e+00 -9.49300885e-01 -1.62428632e-01
4.49061900e-01 -8.31878543e-01 1.51419497e+00 -1.19361770e+00
-1.18674099e+00 1.05469108e+00 2.86159050e-02 -5.21024466e-01
6.09465122e-01 1.46755993e-01 -2.88980931e-01 6.56918585e-01
1.22610047e-01 8.61739278e-01 7.38315344e-01 -1.39218986e+00
-8.62504005e-01 -2.66032428e-01 -1.30832076e-01 5.84675744e-02
-5.14601767e-01 -4.77904171e-01 -3.61374795e-01 -6.56607985e-01
1.08727559e-01 -7.90082574e-01 -4.93866533e-01 1.14830479e-01
-1.86965808e-01 -2.09504887e-01 6.12148106e-01 -8.39900434e-01
9.84944224e-01 -2.02194500e+00 -1.42201455e-02 6.13198988e-02
3.85078251e-01 6.02417588e-01 -4.52418059e-01 -3.73845041e-01
-2.61428982e-01 -2.58018691e-02 -5.97349882e-01 4.45768684e-02
-4.73741919e-01 7.90637657e-02 4.10465628e-01 3.11406642e-01
3.69901091e-01 1.00387216e+00 -1.36249352e+00 -8.42487395e-01
4.41226631e-01 1.90287381e-01 -5.73771119e-01 4.21197623e-01
-2.48182699e-01 6.12510800e-01 -4.44710076e-01 8.89283538e-01
8.61314476e-01 -6.31055295e-01 1.92483544e-01 -6.97664082e-01
3.22284311e-01 1.97675917e-02 -8.23921859e-01 1.88342619e+00
-7.92438507e-01 2.73996502e-01 1.48358615e-02 -1.41606796e+00
4.33898211e-01 4.39610958e-01 9.59179878e-01 -9.49653089e-01
4.82734084e-01 6.39512658e-01 1.79499373e-01 -9.60571051e-01
1.42964507e-02 -2.98127979e-01 4.77193028e-01 3.85661870e-01
2.68067300e-01 -1.10835105e-01 3.42669666e-01 3.61110978e-02
1.13186848e+00 -1.99064299e-01 3.61157805e-01 -2.51752436e-01
5.41675866e-01 2.21231654e-01 5.11201978e-01 5.12812972e-01
-4.77648139e-01 5.93514383e-01 9.87255052e-02 -3.84586364e-01
-9.46364582e-01 -8.65650237e-01 -3.40751469e-01 7.30139017e-01
3.46673816e-01 1.11839585e-01 -8.55123758e-01 -1.02245545e+00
-2.15571299e-01 4.79138345e-01 -7.37497628e-01 -3.47774774e-01
-7.55641282e-01 -9.84144449e-01 1.64970458e-01 7.20757246e-01
7.54853845e-01 -1.05544674e+00 -6.29507124e-01 1.72325790e-01
-6.22874916e-01 -1.06043172e+00 -3.34167957e-01 2.90752668e-02
-1.13608491e+00 -1.32622111e+00 -1.32567561e+00 -1.23409295e+00
9.56138432e-01 3.53332669e-01 1.07792294e+00 1.62448585e-01
-7.33508706e-01 4.69133228e-01 -5.03486156e-01 -4.41906124e-01
-8.48826289e-01 1.02671660e-01 -3.91793966e-01 -3.09482843e-01
3.14846188e-01 -2.83924669e-01 -9.78895247e-01 -7.62426713e-03
-1.15540242e+00 3.70931327e-01 9.57821012e-01 1.30766356e+00
7.45126903e-01 -6.26659468e-02 6.45187855e-01 -1.00579906e+00
2.88518280e-01 -5.62431753e-01 -1.27206609e-01 4.26295966e-01
-2.49514967e-01 -1.06208712e-01 2.72867292e-01 -4.26778793e-01
-1.12325156e+00 -2.01477967e-02 -2.80388117e-01 -4.92413700e-01
-2.14446425e-01 4.93477136e-01 3.04714411e-01 -2.04549804e-02
8.94184530e-01 3.59939516e-01 2.52008021e-01 -1.85875356e-01
9.00406316e-02 8.72927487e-01 5.21439850e-01 -3.25157613e-01
3.24436396e-01 5.59305370e-01 3.81964259e-02 -6.07229054e-01
-1.06154621e+00 -5.55007279e-01 -3.79159242e-01 -1.57853588e-01
8.48756194e-01 -9.46207643e-01 -2.37069488e-01 4.90848184e-01
-8.37521911e-01 -5.57106674e-01 -2.58371174e-01 5.66845953e-01
-4.87287402e-01 3.72184098e-01 -8.14747334e-01 -3.59117508e-01
-7.49023914e-01 -1.41839027e+00 8.57198060e-01 2.18729973e-01
2.12336957e-01 -9.67550576e-01 -1.28822461e-01 4.47488427e-01
3.32393795e-01 2.67962068e-01 9.46745634e-01 -5.53851902e-01
-4.36419576e-01 6.02806509e-02 -5.06459475e-01 6.84331357e-01
5.76777279e-01 -4.51450676e-01 -8.29269886e-01 -2.92718142e-01
6.81840479e-02 -4.85474288e-01 1.06021881e+00 8.76835465e-01
1.70446873e+00 -5.30591048e-02 -4.56480026e-01 5.90165973e-01
1.42861986e+00 1.37799487e-01 3.10331166e-01 2.41296098e-01
7.21239805e-01 5.38149893e-01 7.90472090e-01 3.48808467e-01
3.61686826e-01 3.45145687e-02 5.43027759e-01 -4.59641904e-01
-3.84103864e-01 6.51582405e-02 -3.34215850e-01 7.57447183e-01
6.99710846e-02 -1.32600114e-01 -1.03516662e+00 7.49523103e-01
-1.74886107e+00 -7.06038952e-01 1.57186046e-01 1.75394070e+00
1.13234317e+00 -3.43970768e-02 -5.11125214e-02 4.73655686e-02
7.30091512e-01 -2.03857094e-01 -7.77456343e-01 2.92439908e-01
2.97599375e-01 3.73989940e-01 4.96730655e-01 1.63121238e-01
-1.40941465e+00 5.36988080e-01 4.81975508e+00 1.07356465e+00
-1.47247529e+00 3.76009315e-01 1.12332642e+00 2.12046534e-01
-2.33259290e-01 -7.68677294e-01 -3.48992169e-01 3.63147944e-01
5.93102455e-01 2.34598950e-01 -8.79529342e-02 1.00055289e+00
-4.29308191e-02 -8.63284320e-02 -1.15073740e+00 1.25142300e+00
2.06676364e-01 -1.72334301e+00 1.73223570e-01 -3.35187256e-01
8.67631614e-01 1.75615951e-01 9.07279253e-02 1.95259720e-01
5.22997677e-02 -6.35198116e-01 1.65413797e-01 2.14298218e-01
1.08576775e+00 -5.66184044e-01 9.76405203e-01 2.54371822e-01
-9.94080067e-01 -7.40607083e-02 -2.80286700e-01 7.29127288e-01
-1.22282811e-01 6.48125172e-01 -1.10044444e+00 7.06505716e-01
7.71712363e-01 9.81248319e-01 -3.76173079e-01 1.08544230e+00
2.76852041e-01 4.80264902e-01 3.91878523e-02 7.63725042e-02
1.78084895e-01 1.90007567e-01 2.78322071e-01 1.25003076e+00
4.15359616e-01 3.64885151e-01 1.91531867e-01 7.27024972e-01
-2.13211954e-01 1.87721014e-01 -2.66908944e-01 1.08829075e-02
2.77728587e-01 1.43373334e+00 -8.03778708e-01 -6.85494542e-01
-3.31415921e-01 8.69092584e-01 -2.67346054e-02 6.84094280e-02
-6.34367883e-01 -1.56886965e-01 3.51912975e-02 6.62531853e-02
3.05510163e-02 3.92591327e-01 -2.35061750e-01 -1.20305526e+00
-1.76947743e-01 -9.47707117e-01 5.52755415e-01 -4.95588988e-01
-1.46538198e+00 6.09998941e-01 -8.69864747e-02 -1.69296789e+00
-2.61329085e-01 -6.78743720e-01 -3.34130943e-01 4.67539161e-01
-1.94242167e+00 -1.15874803e+00 -7.84314811e-01 4.93804872e-01
7.92990446e-01 -2.05841139e-01 5.72222948e-01 5.38975894e-01
-3.13646555e-01 6.84563160e-01 -6.74494952e-02 2.80515283e-01
7.67348886e-01 -1.25622463e+00 -5.59898913e-02 4.81347084e-01
-4.42479998e-01 2.11328030e-01 3.99373442e-01 -5.31042159e-01
-1.17084980e+00 -1.40453720e+00 1.85076490e-01 -4.35418030e-03
2.60018915e-01 2.51595914e-01 -8.74990523e-01 2.95994073e-01
9.49440151e-02 4.38196063e-01 8.77574742e-01 -4.96085465e-01
8.97635221e-02 -1.23670802e-01 -1.58359504e+00 4.58133340e-01
9.61305022e-01 -9.87577215e-02 -4.22882259e-01 5.22401392e-01
7.44180024e-01 -8.10125887e-01 -9.59073424e-01 9.25131142e-01
3.10817301e-01 -8.05488646e-01 9.69880939e-01 -4.42764282e-01
9.47115958e-01 1.17267832e-01 -6.27366155e-02 -1.37037730e+00
-1.37545452e-01 1.43157691e-01 -1.04294755e-01 6.47549272e-01
2.06112072e-01 -4.91448879e-01 9.72791612e-01 2.74132162e-01
-4.35002625e-01 -1.12031448e+00 -7.25915134e-01 -4.74090517e-01
8.64716992e-02 -1.93566874e-01 4.99696732e-01 1.11508203e+00
-7.06201568e-02 5.01124747e-02 1.25245839e-01 1.10357963e-01
7.49684274e-01 4.97800335e-02 2.71598518e-01 -9.86373305e-01
-4.45739418e-01 -5.22360444e-01 -1.34985462e-01 -7.76049972e-01
1.40077192e-02 -1.13864982e+00 8.75011533e-02 -1.79837859e+00
6.57512367e-01 -6.70807123e-01 -5.65182865e-01 3.31613570e-01
-3.88054937e-01 2.85849392e-01 -1.50087878e-01 3.80733772e-03
-4.89279330e-01 4.09758270e-01 1.92920327e+00 -5.28726220e-01
-1.60322800e-01 8.02919120e-02 -5.93962491e-01 6.86050236e-01
8.66003513e-01 -2.49341264e-01 -4.82428282e-01 -3.34724337e-01
-2.76820600e-01 3.68376911e-01 4.16767567e-01 -1.05643713e+00
-5.25243245e-02 -1.59565687e-01 4.51312572e-01 -7.51997054e-01
5.38806431e-02 -7.04540253e-01 -4.36764270e-01 1.12150848e+00
-4.58427072e-01 -5.27629018e-01 2.05062121e-01 3.74479622e-01
-3.90316695e-01 -5.06879389e-01 1.05776811e+00 -6.98701382e-01
-7.72162497e-01 7.21357465e-01 -3.22458670e-02 1.02297023e-01
1.00729394e+00 4.50066477e-02 -1.98518634e-01 -3.44505683e-02
-5.90975702e-01 2.64660448e-01 -6.55953260e-03 3.28246206e-01
8.22016060e-01 -1.18473542e+00 -7.75070667e-01 8.34096745e-02
2.76536644e-01 5.29932678e-01 7.91108847e-01 1.01064563e+00
-6.58727586e-01 1.66469306e-01 -2.12901801e-01 -9.32294905e-01
-1.21636438e+00 3.86475921e-01 6.14835560e-01 -3.55742246e-01
-5.26131094e-01 1.03499615e+00 3.31480324e-01 -6.12118125e-01
2.74581641e-01 -7.64870405e-01 -2.68300444e-01 -8.01748559e-02
5.92858613e-01 2.48573482e-01 2.06962720e-01 -2.13277593e-01
-1.78594515e-01 3.54873210e-01 -3.13305199e-01 2.77773023e-01
1.39726031e+00 2.17884824e-01 -4.37738635e-02 2.64637936e-02
1.39242530e+00 -7.26865888e-01 -1.00760341e+00 -4.53440756e-01
-2.25545555e-01 -2.93993711e-01 2.67053843e-01 -7.10528374e-01
-1.41927361e+00 1.02628183e+00 1.24688148e+00 -1.73516825e-01
1.27652192e+00 9.24727097e-02 1.34040534e+00 3.40287030e-01
8.02856013e-02 -9.46563721e-01 6.02981925e-01 -3.45517434e-02
6.86978102e-01 -1.85311437e+00 -1.41641408e-01 -4.58687603e-01
-6.78732753e-01 1.12993610e+00 9.45483565e-01 -1.37116164e-01
9.14700627e-01 8.56419876e-02 1.07979983e-01 -5.10573201e-02
-3.90023470e-01 -2.38174573e-01 2.97108144e-01 6.11193180e-01
4.46845829e-01 1.86182871e-01 -2.10299820e-01 7.00023234e-01
3.32170695e-01 4.78617102e-01 3.89120340e-01 1.20040071e+00
-5.90811014e-01 -1.09658885e+00 -9.33561251e-02 1.04047143e+00
-4.95379478e-01 -1.08553164e-01 3.05973679e-01 5.78111291e-01
2.01635927e-01 6.08542681e-01 6.77902699e-02 -2.76562721e-01
2.00977534e-01 -3.46141547e-01 7.67649472e-01 -6.80653751e-01
-2.86112934e-01 6.68631643e-02 -2.11193740e-01 -3.17649603e-01
-7.36299098e-01 -4.16599602e-01 -1.33249104e+00 1.19882941e-01
-3.33300382e-01 -1.79509342e-01 4.75288659e-01 9.36735153e-01
1.72836751e-01 9.32936549e-01 7.20865309e-01 -7.54845440e-01
-6.18230760e-01 -1.08078885e+00 -3.61425370e-01 6.03837252e-01
5.65267086e-01 -5.33865750e-01 -1.02599733e-01 1.22029543e-01] | [14.790709495544434, -2.242672920227051] |
68222789-4ab1-4e26-b3ec-64ac9d693054 | multi-view-inverse-rendering-for-large-scale | 2211.10206 | null | https://arxiv.org/abs/2211.10206v4 | https://arxiv.org/pdf/2211.10206v4.pdf | Multi-view Inverse Rendering for Large-scale Real-world Indoor Scenes | We present a efficient multi-view inverse rendering method for large-scale real-world indoor scenes that reconstructs global illumination and physically-reasonable SVBRDFs. Unlike previous representations, where the global illumination of large scenes is simplified as multiple environment maps, we propose a compact representation called Texture-based Lighting (TBL). It consists of 3D mesh and HDR textures, and efficiently models direct and infinite-bounce indirect lighting of the entire large scene. Based on TBL, we further propose a hybrid lighting representation with precomputed irradiance, which significantly improves the efficiency and alleviates the rendering noise in the material optimization. To physically disentangle the ambiguity between materials, we propose a three-stage material optimization strategy based on the priors of semantic segmentation and room segmentation. Extensive experiments show that the proposed method outperforms the state-of-the-art quantitatively and qualitatively, and enables physically-reasonable mixed-reality applications such as material editing, editable novel view synthesis and relighting. The project page is at https://lzleejean.github.io/TexIR. | ['Jiaqi Yang', 'Cihui Pan', 'Mofang Cheng', 'Lingli Wang', 'Zhen Li'] | 2022-11-18 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Li_Multi-View_Inverse_Rendering_for_Large-Scale_Real-World_Indoor_Scenes_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Li_Multi-View_Inverse_Rendering_for_Large-Scale_Real-World_Indoor_Scenes_CVPR_2023_paper.pdf | cvpr-2023-1 | ['mixed-reality'] | ['computer-vision'] | [ 2.93130755e-01 -2.58389235e-01 7.00934649e-01 -3.84618640e-01
-4.88041192e-01 -4.46742266e-01 5.30738771e-01 -4.03424501e-01
2.47174576e-01 7.53989220e-01 3.36230636e-01 -2.02717617e-01
8.66816565e-02 -1.06356549e+00 -8.77040505e-01 -6.57373250e-01
6.62986755e-01 3.74075294e-01 3.69304121e-02 -2.53348142e-01
2.19643861e-02 5.11924982e-01 -1.70901155e+00 1.33717313e-01
1.17917645e+00 8.35781813e-01 5.27128339e-01 7.99995482e-01
-1.76794052e-01 6.67701960e-01 -3.29866707e-01 -2.82100588e-01
3.63824338e-01 -4.54486847e-01 -5.82821250e-01 1.88111678e-01
7.28836656e-01 -6.37456775e-01 -2.58762240e-01 9.25070643e-01
5.53022265e-01 3.99606735e-01 3.21066082e-01 -8.73686254e-01
-7.97967017e-01 -3.71748149e-01 -6.54322445e-01 -4.68396932e-01
8.07159245e-01 -6.11652359e-02 5.64495206e-01 -1.06282890e+00
6.09351635e-01 1.44415760e+00 4.93753880e-01 2.46624634e-01
-1.22759128e+00 -4.31888640e-01 3.79158139e-01 8.90022423e-03
-1.32855844e+00 -4.37507272e-01 1.06029415e+00 -2.74278820e-01
5.51255822e-01 8.76248002e-01 9.22408462e-01 9.21668887e-01
1.70892149e-01 3.61634344e-01 1.71325231e+00 -4.23594415e-01
1.26411110e-01 2.52608843e-02 -3.09729546e-01 9.23882604e-01
-7.16231316e-02 1.58621505e-01 -4.81230766e-01 -2.74935126e-01
1.10020638e+00 1.49475336e-01 -6.48590803e-01 -3.96007061e-01
-1.35093582e+00 2.92399824e-01 3.02910030e-01 -3.17934692e-01
-1.24488428e-01 7.38019794e-02 -2.76724603e-02 -1.14798777e-01
9.43832576e-01 -2.17438489e-02 -1.90820396e-01 1.35567620e-01
-6.55686200e-01 2.56997764e-01 6.04256809e-01 1.10250771e+00
8.99787366e-01 1.69075020e-02 3.36755924e-02 1.08256817e+00
5.60246110e-01 1.02229500e+00 -4.65414971e-01 -1.43819690e+00
2.74437696e-01 1.84371367e-01 4.12949890e-01 -8.78770411e-01
-1.59379654e-02 -1.26170412e-01 -9.38421488e-01 3.77833992e-01
9.66142640e-02 1.88045681e-01 -9.62773383e-01 1.33472621e+00
8.52675855e-01 3.85136157e-01 -2.53497690e-01 9.59652185e-01
1.12866545e+00 8.79093528e-01 -4.81861770e-01 -2.19608068e-01
1.18435538e+00 -1.17061710e+00 -9.57046270e-01 -6.17284402e-02
-7.81347379e-02 -1.06254876e+00 1.39075923e+00 5.21640956e-01
-1.28891909e+00 -4.52857107e-01 -8.26478362e-01 -6.14055812e-01
-1.97970383e-02 1.54439574e-02 6.47369325e-01 4.38548923e-01
-9.93466079e-01 3.71293724e-01 -6.58930838e-01 -1.23262219e-01
1.70806557e-01 -1.53852761e-01 -1.61801632e-02 -5.68942308e-01
-8.34251404e-01 6.09989643e-01 -2.67375201e-01 4.50178772e-01
-7.35266030e-01 -9.12166715e-01 -8.17766964e-01 -2.38654584e-01
4.07000661e-01 -1.20945954e+00 8.75925422e-01 -4.63550568e-01
-2.02450275e+00 7.28510678e-01 -5.07208943e-01 6.44085824e-01
7.99529135e-01 -3.58973175e-01 -2.92923778e-01 3.49859297e-02
-8.93058628e-02 2.02871084e-01 7.69891620e-01 -2.06408548e+00
-9.06299129e-02 -2.53586262e-01 2.57282257e-01 6.44926012e-01
3.07181805e-01 -1.73134357e-01 -5.59171617e-01 -5.22088230e-01
2.84032553e-01 -6.80761039e-01 -2.69597620e-01 4.52414602e-01
-6.69710815e-01 6.39507711e-01 6.41784549e-01 -1.03786719e+00
7.84482718e-01 -2.08621502e+00 2.57456958e-01 1.63488567e-01
2.16858983e-01 -3.44667882e-01 9.45556462e-02 2.84407556e-01
2.72863686e-01 -1.88388795e-01 -3.02282572e-01 -9.29840982e-01
5.40869571e-02 3.39707822e-01 -4.42229748e-01 5.67512929e-01
-5.99167645e-01 6.71957970e-01 -8.66418898e-01 -3.95638496e-01
7.95334935e-01 1.09933782e+00 -5.01786470e-01 4.18665230e-01
-1.74338713e-01 1.15048873e+00 -3.02517474e-01 6.09970331e-01
1.28380346e+00 -3.83531563e-02 -1.01651259e-01 -4.36785460e-01
-3.80627602e-01 2.04669997e-01 -1.45982659e+00 2.09751534e+00
-9.55112338e-01 2.61345923e-01 5.50529957e-01 -1.48852304e-01
7.67446578e-01 3.05041131e-02 3.39093089e-01 -8.53053689e-01
-1.30867675e-01 1.16838247e-01 -9.47107434e-01 -1.90658808e-01
6.25929594e-01 -1.56188324e-01 3.35997164e-01 2.19500691e-01
-5.18200934e-01 -7.56848574e-01 -3.77396286e-01 9.56693813e-02
5.75587630e-01 7.70070672e-01 3.54704037e-02 -3.00219625e-01
4.65355963e-01 -4.71932590e-01 5.68607152e-01 3.34645957e-01
3.66776645e-01 1.03980780e+00 -1.57919079e-01 -3.57250839e-01
-1.02083015e+00 -1.46870935e+00 -2.46855646e-01 8.90357435e-01
6.10862195e-01 -3.04920077e-01 -8.10797691e-01 2.85458174e-02
-1.92510322e-01 8.46670210e-01 -4.76224929e-01 4.45461988e-01
-5.86693048e-01 -6.50863409e-01 -2.30477244e-01 9.21006948e-02
7.94277906e-01 -6.62042320e-01 -3.82412493e-01 -1.88665226e-01
-6.55499220e-01 -1.14045453e+00 -5.68271995e-01 -3.64039779e-01
-6.60351694e-01 -8.15382183e-01 -6.42191112e-01 -4.10877824e-01
9.32041228e-01 6.13154471e-01 1.31695533e+00 7.17243999e-02
-4.84689802e-01 8.00535500e-01 -1.10068224e-01 -5.42375632e-02
-2.44381845e-01 -7.09832251e-01 -2.82662302e-01 1.45952642e-01
-6.30291700e-01 -7.71170139e-01 -9.75805104e-01 5.19671738e-01
-7.11164653e-01 1.00120759e+00 -7.79873356e-02 5.19390404e-01
1.09622765e+00 -1.99109465e-01 -1.96264848e-01 -8.95002186e-01
1.52649745e-01 1.74200889e-02 -7.15738237e-01 4.16981310e-01
-3.38011265e-01 -3.31185222e-01 7.09250510e-01 -1.38234898e-01
-1.83406878e+00 -2.72759616e-01 -2.53133267e-01 -4.97680426e-01
-2.22114235e-01 -3.31088930e-01 -6.58562243e-01 -3.75633866e-01
1.19027272e-01 2.31009588e-01 -6.31168544e-01 -7.17160404e-01
6.10160708e-01 2.26218343e-01 4.86902803e-01 -9.82537508e-01
9.24111307e-01 1.08086360e+00 1.59467950e-01 -9.74632740e-01
-7.99915135e-01 -8.33154619e-02 -5.20648539e-01 -4.62206513e-01
9.47855830e-01 -9.95481133e-01 -7.39111423e-01 6.54150248e-01
-1.24101579e+00 -6.95334911e-01 -5.52105844e-01 2.71758914e-01
-7.19438732e-01 4.49316800e-01 -6.25048637e-01 -7.38494158e-01
-2.06163283e-02 -1.22132480e+00 1.45782042e+00 5.81722185e-02
1.91487893e-01 -1.05930471e+00 1.38940200e-01 6.77304924e-01
3.53899449e-01 4.95129824e-01 7.04585671e-01 8.43076587e-01
-1.17120314e+00 5.25892437e-01 -3.72047305e-01 2.62730539e-01
3.00071925e-01 2.20709503e-01 -1.41998553e+00 -1.90260932e-01
1.08943470e-01 3.37761082e-02 5.36045194e-01 4.90601987e-01
1.46596301e+00 -1.27357140e-01 -6.94659576e-02 1.31111169e+00
1.67990124e+00 -4.52290429e-03 8.50243211e-01 7.86970258e-02
1.40272605e+00 6.35550678e-01 4.90327239e-01 7.12296844e-01
8.71176720e-01 6.53843939e-01 5.66595495e-01 -5.44145286e-01
-5.44165015e-01 -3.58395368e-01 4.10676748e-02 1.30959034e+00
-6.44843400e-01 -2.94025719e-01 -4.35198396e-01 9.71798673e-02
-1.53062630e+00 -6.96068943e-01 -3.98676693e-01 2.41547966e+00
5.56727052e-01 -4.16301072e-01 -5.39127588e-01 -2.80849397e-01
4.95765209e-01 5.72307229e-01 -5.43893158e-01 -4.94169444e-01
-1.67457417e-01 1.17939815e-01 4.36568230e-01 1.10630059e+00
-5.92651308e-01 5.92001677e-01 5.78847694e+00 7.64765084e-01
-6.96434200e-01 4.54427540e-01 7.25580335e-01 -7.32277855e-02
-1.14680958e+00 -2.52440181e-02 -4.85572606e-01 3.21281612e-01
4.11857903e-01 1.67497769e-01 1.17468882e+00 4.58828837e-01
5.96582055e-01 -2.97892988e-01 -7.84829438e-01 1.17504048e+00
1.85399815e-01 -1.15842736e+00 8.28605518e-03 1.31876752e-01
1.01162457e+00 -1.70259729e-01 1.06374621e-01 -2.69100547e-01
3.89353186e-01 -7.25337625e-01 9.87938941e-01 8.95606995e-01
1.13073182e+00 -5.20400345e-01 -7.53570497e-02 1.90847263e-01
-1.55564702e+00 3.05866152e-01 -3.07967693e-01 1.49110824e-01
6.12434268e-01 9.16974247e-01 8.24001580e-02 9.24118936e-01
7.23942935e-01 6.70343220e-01 -2.68273950e-01 7.32124090e-01
-3.15327674e-01 5.45006692e-02 -3.33604097e-01 6.15661740e-01
-4.99805838e-01 -8.98647547e-01 4.39195663e-01 8.42827022e-01
4.54864293e-01 3.64802599e-01 3.54659706e-01 1.16348064e+00
-7.10902885e-02 6.42628819e-02 -4.22706723e-01 6.87872291e-01
2.96886772e-01 1.29759026e+00 -7.11541891e-01 -3.05669129e-01
-4.19020325e-01 1.44447458e+00 1.57278195e-01 8.35516572e-01
-1.03478158e+00 -1.01197854e-01 5.78121960e-01 2.68279791e-01
-3.20327729e-02 -3.45057636e-01 -3.86622012e-01 -1.47766507e+00
2.36670613e-01 -4.22922760e-01 -3.79915535e-01 -1.44529307e+00
-1.09012234e+00 3.82749587e-01 5.96911348e-02 -1.18724835e+00
4.57374960e-01 -3.64342749e-01 -5.86308539e-01 1.05859268e+00
-1.78880024e+00 -1.50743556e+00 -8.52045536e-01 6.43991053e-01
5.30402005e-01 7.12117374e-01 9.24533367e-01 4.52959895e-01
-4.30498421e-01 6.23445725e-03 4.69427496e-01 -5.11852682e-01
7.20823050e-01 -1.15345085e+00 4.13884401e-01 7.69478261e-01
-3.43119025e-01 4.90492046e-01 7.07493603e-01 -6.60123944e-01
-1.66407871e+00 -1.08066559e+00 2.70419031e-01 -3.35173368e-01
9.96382385e-02 -6.85004950e-01 -7.45214939e-01 6.34955704e-01
2.61485130e-01 2.24473193e-01 4.87327605e-01 -1.73085093e-01
-3.17185491e-01 -1.23833150e-01 -1.38598740e+00 6.91767752e-01
1.58174539e+00 -6.66357398e-01 -4.14255746e-02 5.17656684e-01
8.84770155e-01 -9.69016552e-01 -9.25887704e-01 3.47029775e-01
6.55280888e-01 -1.39082158e+00 1.39733458e+00 3.54716659e-01
4.80970651e-01 -4.71676409e-01 -5.36494792e-01 -1.39626217e+00
-3.36066395e-01 -7.39496708e-01 -1.46585777e-01 1.19341505e+00
-1.60395905e-01 -9.28048670e-01 3.04307163e-01 7.50189483e-01
-5.24391711e-01 -6.35066509e-01 -7.10675478e-01 -5.04571855e-01
-4.06339347e-01 -3.08988839e-01 8.69503915e-01 9.64586616e-01
-9.15154517e-01 -4.75967377e-02 -4.93220150e-01 4.21592832e-01
1.18122768e+00 5.81882656e-01 8.59513938e-01 -1.10342777e+00
-5.19676149e-01 3.69247720e-02 3.47155243e-01 -1.33863306e+00
5.01344353e-02 -4.62207705e-01 1.58109725e-01 -2.04492688e+00
2.66134888e-01 -6.76700413e-01 1.84089124e-01 -2.88674328e-02
-1.12257019e-01 2.33237088e-01 3.81678417e-02 2.70522814e-02
-5.08135974e-01 8.85114074e-01 2.01386738e+00 1.30125374e-01
-1.49477109e-01 -2.16210037e-01 -4.16910827e-01 9.16472197e-01
4.83433694e-01 -7.83287082e-03 -6.10851765e-01 -7.84333587e-01
3.67818773e-01 1.67694483e-02 5.80899358e-01 -6.43334508e-01
-3.04735661e-01 -4.96853590e-01 4.85904902e-01 -7.16458976e-01
9.71092224e-01 -8.87337923e-01 7.69764304e-01 -1.50780395e-01
3.29717807e-02 -1.52574226e-01 9.74947140e-02 5.79008758e-01
2.61773050e-01 3.26343805e-01 7.60537267e-01 -3.43942016e-01
-3.31810296e-01 4.46267724e-01 1.46427378e-01 6.57085478e-02
6.36400878e-01 -3.08093160e-01 -5.26161373e-01 -3.13695550e-01
-5.80573261e-01 -1.01552539e-01 1.16341496e+00 4.46491502e-02
8.56259942e-01 -1.46554112e+00 -4.98919964e-01 4.00602043e-01
-1.38211131e-01 4.60670412e-01 8.88538718e-01 6.39827251e-01
-9.78187323e-01 -1.39238909e-01 1.00489758e-01 -5.09416461e-01
-1.25715232e+00 2.52923787e-01 3.39562595e-01 -1.21711763e-02
-1.01117754e+00 8.34591806e-01 8.88338566e-01 -8.59631300e-01
-1.43134758e-01 -4.90298152e-01 3.48493934e-01 -5.10949790e-01
4.06018913e-01 7.95406520e-01 -2.33253371e-03 -7.62051702e-01
-8.58645663e-02 1.08181345e+00 4.77858305e-01 -1.29239455e-01
1.41000116e+00 -7.13561356e-01 -4.89989638e-01 6.70833230e-01
1.03970480e+00 5.75887382e-01 -1.57153940e+00 -5.87088093e-02
-1.16670048e+00 -1.20269620e+00 2.50059962e-01 -7.54325747e-01
-1.00378990e+00 8.81660640e-01 4.26830381e-01 -1.58998668e-01
1.32413495e+00 -2.78004169e-01 9.56392407e-01 4.20313179e-02
7.94501483e-01 -9.12706733e-01 -2.08795965e-01 3.78673553e-01
1.16244721e+00 -9.81100798e-01 4.56027925e-01 -9.80705678e-01
-3.25301468e-01 7.98501790e-01 5.16179621e-01 5.00219502e-02
7.31260359e-01 3.34981889e-01 -1.04004763e-01 -1.79783732e-01
-2.60113955e-01 1.83402687e-01 2.96737134e-01 5.38665950e-01
1.77699178e-01 3.48982841e-01 2.52510697e-01 -2.05880571e-02
-1.36797205e-01 -2.04737976e-01 5.36513567e-01 7.36185312e-01
-1.81760371e-01 -9.37226951e-01 -7.28985548e-01 1.96679249e-01
1.25664398e-01 -2.57524282e-01 1.16278768e-01 2.85337538e-01
3.01577270e-01 7.94458628e-01 -1.02841124e-01 -7.23482296e-02
4.49049830e-01 -3.53685170e-01 8.43161345e-01 -5.04829943e-01
-1.65226102e-01 3.89385015e-01 1.06391624e-01 -9.82568324e-01
-5.46884298e-01 -4.72948670e-01 -1.04006732e+00 -6.50354743e-01
-3.35912675e-01 -3.36299330e-01 8.46196949e-01 5.48334002e-01
3.63677233e-01 8.13611507e-01 8.09651375e-01 -1.47577548e+00
3.25954050e-01 -4.34981674e-01 -8.26316714e-01 3.00953150e-01
4.64166880e-01 -8.05123687e-01 -4.77424145e-01 1.85363352e-01] | [9.638041496276855, -3.087989330291748] |
993cbd40-efbe-4bc9-8c8e-e85dafd0bb6e | cvlnet-cross-view-semantic-correspondence | 2208.03660 | null | https://arxiv.org/abs/2208.03660v1 | https://arxiv.org/pdf/2208.03660v1.pdf | CVLNet: Cross-View Semantic Correspondence Learning for Video-based Camera Localization | This paper tackles the problem of Cross-view Video-based camera Localization (CVL). The task is to localize a query camera by leveraging information from its past observations, i.e., a continuous sequence of images observed at previous time stamps, and matching them to a large overhead-view satellite image. The critical challenge of this task is to learn a powerful global feature descriptor for the sequential ground-view images while considering its domain alignment with reference satellite images. For this purpose, we introduce CVLNet, which first projects the sequential ground-view images into an overhead view by exploring the ground-and-overhead geometric correspondences and then leverages the photo consistency among the projected images to form a global representation. In this way, the cross-view domain differences are bridged. Since the reference satellite images are usually pre-cropped and regularly sampled, there is always a misalignment between the query camera location and its matching satellite image center. Motivated by this, we propose estimating the query camera's relative displacement to a satellite image before similarity matching. In this displacement estimation process, we also consider the uncertainty of the camera location. For example, a camera is unlikely to be on top of trees. To evaluate the performance of the proposed method, we collect satellite images from Google Map for the KITTI dataset and construct a new cross-view video-based localization benchmark dataset, KITTI-CVL. Extensive experiments have demonstrated the effectiveness of video-based localization over single image-based localization and the superiority of each proposed module over other alternatives. | ['Hongdong Li', 'Shan Wang', 'Xin Yu', 'Yujiao Shi'] | 2022-08-07 | null | null | null | null | ['image-based-localization', 'camera-localization'] | ['computer-vision', 'computer-vision'] | [ 1.46261901e-02 -5.84186435e-01 -1.98032647e-01 -2.58469880e-01
-1.07290888e+00 -1.00389624e+00 5.61749220e-01 -8.24413151e-02
-3.50197434e-01 3.03600788e-01 -1.87741175e-01 1.76469699e-01
-6.44900203e-02 -6.24996066e-01 -9.14548576e-01 -8.54796112e-01
-9.73404720e-02 2.48086780e-01 3.30239266e-01 8.31422359e-02
3.55861664e-01 6.08867586e-01 -1.19930387e+00 -2.92308718e-01
5.12877643e-01 1.05466509e+00 6.86826229e-01 5.84349930e-01
3.45511824e-01 4.70801622e-01 -3.84671092e-01 -3.03153712e-02
5.03566563e-01 5.23408055e-02 -5.70774198e-01 7.32153893e-01
8.61406267e-01 -5.18138647e-01 -4.68201071e-01 1.15792775e+00
1.55105382e-01 1.88558340e-01 1.49061069e-01 -1.38101125e+00
-2.82990038e-01 -4.03758734e-02 -9.75219905e-01 4.01908845e-01
7.74067283e-01 -1.58067234e-03 1.00177097e+00 -9.68225896e-01
8.45822871e-01 8.80911648e-01 8.67556393e-01 -2.14689821e-01
-1.09030664e+00 -4.79439437e-01 4.43556011e-01 6.88960105e-02
-1.89120805e+00 -3.39769393e-01 6.63626194e-01 -4.72871453e-01
5.24614871e-01 2.76672915e-02 5.89475036e-01 6.56714618e-01
5.86137734e-02 3.75320613e-01 8.54232669e-01 -2.51243800e-01
1.54589593e-01 2.28388626e-02 -1.81376144e-01 5.85834682e-01
1.39875829e-01 4.01602872e-02 -6.10599637e-01 -2.87899762e-01
6.87854767e-01 4.49274391e-01 -6.88018024e-01 -7.94745147e-01
-1.49243414e+00 5.72350562e-01 5.74979961e-01 1.80629387e-01
-3.53516906e-01 8.86016861e-02 4.17456143e-02 8.48505497e-02
3.88084054e-01 1.28444200e-02 -1.87035948e-01 2.18409836e-01
-1.19416106e+00 1.92377996e-02 4.70377684e-01 1.20665061e+00
1.23522758e+00 -1.90099031e-01 4.57064092e-01 4.15625632e-01
2.21433192e-01 9.48651910e-01 1.23588175e-01 -8.67583156e-01
8.86979461e-01 5.05291402e-01 3.58712077e-01 -1.53820646e+00
-5.23306429e-02 -2.47891426e-01 -6.40217960e-01 -2.28452995e-01
3.65798712e-01 1.60927281e-01 -5.12598336e-01 1.49864542e+00
5.35566866e-01 4.97169942e-01 7.64077231e-02 1.12345088e+00
2.13192865e-01 7.61040032e-01 -4.96510118e-01 -1.04837939e-01
1.12577534e+00 -7.74281740e-01 -2.05877244e-01 -6.27781630e-01
5.04989445e-01 -7.54486084e-01 5.13860345e-01 2.87175160e-02
-4.54176456e-01 -5.49219370e-01 -1.00344419e+00 3.16457421e-01
-1.76943138e-01 5.09374917e-01 2.01923475e-01 1.54900268e-01
-9.95826483e-01 2.82157451e-01 -9.18173254e-01 -7.07757413e-01
-1.12441055e-01 1.69841364e-01 -8.59835565e-01 -5.57682931e-01
-9.18026328e-01 4.75261122e-01 3.41799855e-01 2.11321086e-01
-1.08789849e+00 -4.50495631e-01 -8.83726299e-01 -4.43603918e-02
6.15544915e-01 -4.47169870e-01 8.99898887e-01 -9.45978403e-01
-7.98621476e-01 7.83151209e-01 -3.24072599e-01 -3.20637673e-01
3.39309633e-01 -1.17385380e-01 -4.28642392e-01 2.63340712e-01
6.73584521e-01 3.08035195e-01 9.26017106e-01 -1.27619374e+00
-1.03028917e+00 -7.29153216e-01 9.90457237e-02 4.25575525e-01
2.70329807e-02 -2.79706299e-01 -9.22233105e-01 -4.46125746e-01
7.50369370e-01 -1.34162378e+00 -3.11550144e-02 6.79797977e-02
-2.07310438e-01 2.60239035e-01 1.01129413e+00 -5.88405311e-01
8.29345822e-01 -2.38306594e+00 7.59877488e-02 3.26968223e-01
-1.12466760e-01 -2.41951898e-01 -8.25126842e-02 7.32111156e-01
6.67049550e-03 -2.21449405e-01 -7.28597939e-02 -2.41388559e-01
-4.50147659e-01 2.67358422e-01 -5.66095114e-01 1.03747284e+00
-3.10673028e-01 4.47615564e-01 -1.05399430e+00 -3.78992528e-01
2.18488440e-01 2.76268393e-01 -2.05286711e-01 2.33897448e-01
1.00093722e-01 6.24019563e-01 -4.74886924e-01 1.01402175e+00
1.01917684e+00 -2.81176507e-01 8.83269832e-02 -4.21438307e-01
-2.56365895e-01 -1.76559493e-01 -1.32641673e+00 2.09189034e+00
-4.96664762e-01 6.10202849e-01 1.62463784e-01 -7.58692563e-01
9.13135052e-01 8.02834183e-02 5.65736234e-01 -3.91209036e-01
-4.10307020e-01 2.00737879e-01 -7.14976192e-01 -2.43696675e-01
6.35504246e-01 2.11034715e-01 -1.94175005e-01 2.98089772e-01
-1.07330851e-01 -2.38064304e-01 -7.70412982e-02 2.61106074e-01
9.10203040e-01 1.68425530e-01 4.84933257e-01 -2.04981095e-03
7.75806427e-01 1.38676301e-01 6.49855196e-01 6.78118229e-01
-4.77489345e-02 7.20844746e-01 3.62709835e-02 -5.61738014e-01
-7.92059243e-01 -9.83973920e-01 1.48356363e-01 6.58824682e-01
7.29316413e-01 -4.27180350e-01 -5.43388307e-01 -7.58816183e-01
1.51250251e-02 1.12389252e-01 -4.26255375e-01 3.08215708e-01
-4.65531111e-01 -2.63916582e-01 2.14255825e-01 3.93447697e-01
6.99225187e-01 -1.04955249e-01 -8.11499000e-01 5.88169023e-02
-5.50039172e-01 -1.46187782e+00 -9.14222121e-01 -4.34518866e-02
-9.32279170e-01 -1.20583975e+00 -5.73880672e-01 -6.89984560e-01
8.70237708e-01 1.27002287e+00 8.87145221e-01 -4.12157178e-02
2.10790634e-01 6.80279136e-01 -3.74489903e-01 4.81079191e-01
1.56869590e-01 5.31588718e-02 3.61450166e-01 2.46532977e-01
8.40375870e-02 -5.40330648e-01 -7.17529714e-01 7.54702747e-01
-8.95207763e-01 -2.54774332e-01 5.06846070e-01 8.83666635e-01
9.86795604e-01 2.78750986e-01 -3.80009711e-01 -3.00643921e-01
-2.26550847e-01 -5.69696784e-01 -1.14211643e+00 4.47230399e-01
-2.53457427e-01 -3.20610851e-01 4.35625136e-01 -2.44843945e-01
-5.73584735e-01 7.86899805e-01 5.02901793e-01 -8.14532638e-01
-8.29494819e-02 6.58326328e-01 -2.23878920e-01 -5.10250807e-01
1.18923090e-01 7.45713353e-01 -8.08551759e-02 -2.79279590e-01
2.93779403e-01 6.68624818e-01 8.97877514e-01 -3.72168839e-01
1.18431568e+00 9.54381704e-01 -1.35519296e-01 -9.17659938e-01
-8.68528128e-01 -9.26342309e-01 -1.00107992e+00 -1.77611455e-01
5.97897530e-01 -1.49823713e+00 -3.87856752e-01 3.48082155e-01
-1.16853333e+00 2.69223183e-01 1.68472931e-01 6.00425243e-01
-4.55452025e-01 7.39778221e-01 -4.54843417e-02 -5.84681928e-01
-2.91696154e-02 -1.24019742e+00 1.52049148e+00 2.35937074e-01
2.35172719e-01 -1.03647840e+00 2.51802474e-01 2.65163481e-01
-3.46576087e-02 2.19684616e-01 4.57719266e-02 -3.39208961e-01
-1.03700793e+00 -5.64007878e-01 -1.95396438e-01 8.23977590e-02
2.47224003e-01 -5.61101362e-02 -7.88008749e-01 -8.00615847e-01
1.72525108e-01 1.23565316e-01 5.09342730e-01 2.97917426e-01
5.68299294e-01 -7.45688751e-02 -6.61550522e-01 1.02470326e+00
1.81019449e+00 1.59298375e-01 2.14015588e-01 7.38598585e-01
6.52910411e-01 3.71695668e-01 1.12348807e+00 5.06500602e-01
6.51683450e-01 1.06050646e+00 6.84304595e-01 9.29738581e-02
4.73163396e-01 -6.05378449e-01 4.49301541e-01 6.38986945e-01
1.64424330e-01 -1.59330890e-01 -8.57068300e-01 6.74996853e-01
-1.75360036e+00 -9.01093602e-01 9.62374732e-02 2.47312403e+00
-1.06909454e-01 -3.14469546e-01 -2.37195522e-01 -3.61188143e-01
1.00070393e+00 5.42477190e-01 -5.63364446e-01 6.06320560e-01
-1.54634044e-01 -7.24529445e-01 1.16336775e+00 3.77037078e-01
-1.41336286e+00 8.79982471e-01 5.37595034e+00 6.22434378e-01
-1.33331597e+00 1.48543254e-01 2.10569322e-01 1.77950263e-01
1.04213074e-01 5.72481871e-01 -1.05344963e+00 5.12145162e-01
4.39906836e-01 -6.15794323e-02 4.86817718e-01 1.05816305e+00
1.01625666e-01 -5.52293539e-01 -1.00152576e+00 1.20671201e+00
4.07256275e-01 -1.23704612e+00 -1.52002677e-01 3.19485545e-01
8.10536683e-01 3.84331822e-01 -4.49293256e-02 -1.70101598e-01
-2.86428025e-03 -4.25317138e-01 9.20541883e-01 2.94765532e-01
8.01376462e-01 -5.44064224e-01 7.31967568e-01 4.72838908e-01
-1.72803116e+00 -1.71691358e-01 -4.95929331e-01 5.64232580e-02
3.92239660e-01 4.58261907e-01 -7.43188977e-01 9.79066074e-01
8.83467257e-01 1.16862750e+00 -6.36043131e-01 9.78884220e-01
-1.46658257e-01 1.00841366e-01 -5.28606832e-01 6.13894165e-01
4.93352175e-01 -5.29143989e-01 6.27480030e-01 6.69880569e-01
9.09769356e-01 9.11722630e-02 5.58187902e-01 6.29944205e-01
1.21199451e-01 -1.95634902e-01 -1.08449650e+00 2.23418415e-01
8.56573761e-01 1.18695807e+00 -5.64622819e-01 -8.89430940e-02
-6.66791260e-01 1.17515612e+00 3.61363217e-02 3.42817664e-01
-9.47726190e-01 -8.16160664e-02 6.93646252e-01 1.25934929e-01
6.26789689e-01 -4.95865494e-01 4.46256518e-01 -1.47746682e+00
3.15410703e-01 -6.63130641e-01 3.28972548e-01 -9.39517915e-01
-9.04357135e-01 7.53645837e-01 -1.25831440e-02 -1.88199627e+00
-1.33715600e-01 -8.16660002e-02 -5.20466328e-01 7.62343585e-01
-1.37898469e+00 -1.43307781e+00 -7.28978872e-01 7.47041523e-01
4.51609224e-01 -5.17895520e-02 3.67038548e-01 1.48173138e-01
-3.61893564e-01 2.33004749e-01 5.18049181e-01 1.46272808e-01
7.17189729e-01 -8.10921729e-01 4.19475466e-01 1.23146892e+00
5.32086194e-01 5.14645815e-01 3.49202335e-01 -7.06662118e-01
-1.69619644e+00 -1.30050731e+00 6.91737235e-01 -5.67445695e-01
9.13197815e-01 -2.77904898e-01 -7.17121959e-01 9.68562841e-01
-1.53518289e-01 4.26988810e-01 1.73578694e-01 -4.25154299e-01
-3.26691836e-01 -5.39934874e-01 -8.82045984e-01 7.06129968e-02
8.18568110e-01 -9.16446328e-01 -3.66049677e-01 4.75924194e-01
4.14462745e-01 -6.47999585e-01 -9.60400701e-01 6.28849790e-02
4.04998392e-01 -1.00107419e+00 1.23868501e+00 1.69897109e-01
-7.75756836e-02 -8.81704450e-01 -7.93243527e-01 -1.39713252e+00
6.10687137e-02 -3.56965333e-01 4.32020307e-01 1.42929816e+00
-2.59398550e-01 -5.55439532e-01 7.74307609e-01 2.94619292e-01
2.16023877e-01 -1.89694643e-01 -1.25104892e+00 -1.06925619e+00
-5.07500410e-01 -1.78918377e-01 5.76854050e-01 8.79531264e-01
-3.99223298e-01 7.93934539e-02 -6.88318968e-01 1.08205974e+00
8.14127624e-01 4.43306416e-01 1.13370049e+00 -8.63844991e-01
-1.99453130e-01 3.46393883e-01 -7.95170724e-01 -1.34750104e+00
1.71833158e-01 -6.25244737e-01 1.62476972e-01 -1.08778930e+00
2.13331446e-01 -3.35183740e-01 5.91280796e-02 -5.80600537e-02
5.50441444e-02 1.32049724e-01 2.83536464e-01 7.77495682e-01
-7.89825320e-01 3.94520164e-01 6.69188678e-01 -1.68030351e-01
-1.12728104e-02 8.83354098e-02 -1.27832010e-01 6.93026364e-01
3.30634296e-01 -6.81294799e-01 -3.07696849e-01 -7.93726921e-01
7.35142548e-03 6.32940769e-01 6.18434370e-01 -1.20995176e+00
7.24106431e-01 -9.66307819e-02 1.16532266e-01 -9.29158986e-01
3.66671830e-01 -1.25681913e+00 6.30561113e-01 3.19026828e-01
1.78981453e-01 5.38372934e-01 -2.10674271e-01 1.13899851e+00
-6.24823093e-01 -1.39141262e-01 5.60762882e-01 -2.58899927e-01
-1.21100628e+00 4.90183979e-01 -7.31372088e-02 -1.10044606e-01
1.16667974e+00 -4.45123315e-01 -2.71930873e-01 -4.64143157e-01
-3.19702297e-01 2.95297712e-01 1.17196226e+00 2.29452759e-01
6.27056658e-01 -1.26381516e+00 -3.48743618e-01 3.72962266e-01
6.80407703e-01 1.95776895e-01 2.83226818e-01 1.15517247e+00
-7.38673091e-01 5.17429650e-01 3.08926613e-03 -1.21526098e+00
-1.35590732e+00 6.89613700e-01 3.09766620e-01 1.17366835e-02
-6.19152546e-01 5.47998667e-01 3.03549320e-01 -4.72121269e-01
1.44042045e-01 -3.31430703e-01 1.77225024e-01 -1.08167483e-02
5.68027258e-01 1.48561090e-01 -6.67333305e-02 -1.23698080e+00
-7.01004505e-01 1.28936923e+00 8.55182409e-02 -3.48962620e-02
1.25542498e+00 -8.16155672e-01 -1.59892976e-01 2.50371695e-01
1.35525692e+00 1.59071326e-01 -1.48953700e+00 -5.99938214e-01
1.88146651e-01 -1.15893805e+00 9.65133831e-02 -1.28284544e-01
-1.32720888e+00 6.90620840e-01 8.18584442e-01 -1.98616490e-01
1.08623517e+00 -7.50192022e-03 6.98178530e-01 5.11301041e-01
8.79230082e-01 -7.92871892e-01 -1.08252287e-01 2.88300753e-01
4.75394458e-01 -1.46821010e+00 2.22309440e-01 -2.74845779e-01
-6.50148869e-01 1.16979158e+00 4.30589467e-01 -8.07591081e-02
3.50330740e-01 -1.41899019e-01 4.48396429e-02 -1.65908709e-01
-3.68424952e-01 -1.13855906e-01 -2.22586952e-02 6.71321750e-01
-2.63731658e-01 -4.74957749e-02 4.86073107e-01 -3.97100076e-02
2.25719467e-01 -1.33077815e-01 5.70893526e-01 9.29999113e-01
-2.64656872e-01 -7.20308006e-01 -7.23367572e-01 -1.99853793e-01
-6.81520626e-02 3.81622240e-02 -2.63852417e-01 9.08712566e-01
2.66403537e-02 8.65400732e-01 8.18025097e-02 -6.26323164e-01
3.07086438e-01 -4.85139579e-01 1.03618175e-01 -3.97382468e-01
-4.63209972e-02 1.94057122e-01 -3.39324176e-01 -7.83347011e-01
-6.86870337e-01 -8.84275138e-01 -7.52753437e-01 -2.87056655e-01
-4.84371245e-01 3.11284453e-01 7.40235627e-01 8.05229783e-01
3.49619180e-01 -2.32508153e-01 1.16814625e+00 -1.22756970e+00
-6.45379603e-01 -3.98772955e-01 -8.41895163e-01 2.43564144e-01
5.86510956e-01 -7.00163841e-01 -6.77056611e-01 5.69175109e-02] | [7.670236587524414, -2.145390510559082] |
5e0278a8-5e9e-41fd-9efc-adeb0c9f50a4 | anchor-free-person-search | 2103.11617 | null | https://arxiv.org/abs/2103.11617v2 | https://arxiv.org/pdf/2103.11617v2.pdf | Anchor-Free Person Search | Person search aims to simultaneously localize and identify a query person from realistic, uncropped images, which can be regarded as the unified task of pedestrian detection and person re-identification (re-id). Most existing works employ two-stage detectors like Faster-RCNN, yielding encouraging accuracy but with high computational overhead. In this work, we present the Feature-Aligned Person Search Network (AlignPS), the first anchor-free framework to efficiently tackle this challenging task. AlignPS explicitly addresses the major challenges, which we summarize as the misalignment issues in different levels (i.e., scale, region, and task), when accommodating an anchor-free detector for this task. More specifically, we propose an aligned feature aggregation module to generate more discriminative and robust feature embeddings by following a "re-id first" principle. Such a simple design directly improves the baseline anchor-free model on CUHK-SYSU by more than 20% in mAP. Moreover, AlignPS outperforms state-of-the-art two-stage methods, with a higher speed. Code is available at https://github.com/daodaofr/AlignPS | ['Jinpeng Li', 'Ling Shao', 'Fan Zhu', 'Li Liu', 'Shengcai Liao', 'Song Bai', 'Jie Qin', 'Yichao Yan'] | 2021-03-22 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Yan_Anchor-Free_Person_Search_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Yan_Anchor-Free_Person_Search_CVPR_2021_paper.pdf | cvpr-2021-1 | ['person-search'] | ['computer-vision'] | [-4.86575067e-01 -4.77968842e-01 -1.77556314e-02 -2.13247493e-01
-9.46780801e-01 -4.39199388e-01 6.61369860e-01 -1.22461691e-01
-9.73776639e-01 4.52039897e-01 4.91504014e-01 1.63487464e-01
3.58293205e-01 -4.70891207e-01 -3.96895528e-01 -6.28358960e-01
1.94021419e-01 3.28397423e-01 3.23577106e-01 -1.94937102e-02
-1.23155229e-01 4.51637954e-01 -1.69915843e+00 -1.65404752e-01
7.34937131e-01 7.05729842e-01 1.93479747e-01 6.35512650e-01
4.90121245e-01 1.67906508e-01 -4.26691115e-01 -7.61646509e-01
2.93208688e-01 4.90179472e-02 -6.29108012e-01 -9.56219584e-02
6.86417222e-01 -5.88357091e-01 -7.47160614e-01 1.04963660e+00
1.11069822e+00 2.44020179e-01 4.73857313e-01 -1.32342601e+00
-9.52657938e-01 1.62948430e-01 -6.50937021e-01 4.63478714e-01
4.40231055e-01 4.10588950e-01 9.98680234e-01 -1.43095243e+00
1.62883788e-01 1.33063531e+00 7.28637934e-01 7.71833539e-01
-1.10558331e+00 -7.81460345e-01 3.62820745e-01 5.44466138e-01
-2.03589153e+00 -5.80998003e-01 3.44625413e-01 -4.71905202e-01
5.77648163e-01 4.05026436e-01 6.53279126e-01 1.29128969e+00
-4.64308172e-01 1.10705233e+00 7.06279755e-01 -2.62635231e-01
-7.37149790e-02 4.09848094e-01 3.04368287e-01 5.38198590e-01
4.71070468e-01 1.39506444e-01 -4.94542062e-01 -1.77314252e-01
6.01452529e-01 1.53279319e-01 -3.64417344e-01 -3.34659189e-01
-1.17819452e+00 7.73084521e-01 8.05698514e-01 9.72078145e-02
-2.22179800e-01 3.72096561e-02 3.52977663e-01 -2.37148836e-01
4.03068244e-01 1.35546729e-01 -1.51696175e-01 6.12059496e-02
-7.67348230e-01 5.34336746e-01 4.74674731e-01 1.13056803e+00
6.39833570e-01 -2.64561117e-01 -6.47953808e-01 8.19155514e-01
3.71904939e-01 7.04250336e-01 4.07425553e-01 -4.14911360e-01
4.54334766e-01 4.66769338e-01 5.22491276e-01 -8.54573190e-01
-4.30423856e-01 -7.46582389e-01 -7.74952590e-01 -1.81929410e-01
3.21019322e-01 -1.55152306e-01 -7.59985268e-01 1.78997970e+00
6.34107947e-01 3.94627541e-01 -6.65329844e-02 1.38732302e+00
1.07279837e+00 4.38993156e-01 3.78918350e-01 3.97688985e-01
1.94363296e+00 -1.51568532e+00 -4.17205751e-01 -5.15757740e-01
4.70418096e-01 -5.86630583e-01 8.84222925e-01 -2.57281095e-01
-7.19666064e-01 -8.80580127e-01 -9.60404754e-01 -2.09117964e-01
-5.21862864e-01 8.33825469e-01 2.72906154e-01 7.39359319e-01
-1.25680745e+00 4.87406626e-02 -5.12962282e-01 -6.77708209e-01
3.76237541e-01 3.47816080e-01 -5.61812699e-01 -8.20628628e-02
-1.11180723e+00 8.69728863e-01 1.19445391e-01 3.17244470e-01
-6.53932333e-01 -6.65669322e-01 -8.76826704e-01 5.08127585e-02
4.04356033e-01 -1.07220733e+00 1.18023956e+00 -2.86077589e-01
-9.74665582e-01 1.03984082e+00 -6.79312944e-01 -4.63889599e-01
7.35742450e-01 -5.90740144e-01 -4.64444101e-01 -1.26982287e-01
5.17190933e-01 7.61301816e-01 5.53612471e-01 -1.06915677e+00
-1.00635469e+00 -4.95845169e-01 -6.41830564e-02 3.69562596e-01
-4.53387618e-01 4.13542479e-01 -1.05582130e+00 -6.36378109e-01
-3.44149977e-01 -1.02160001e+00 -2.62687624e-01 1.24219485e-01
-4.39071447e-01 -6.18535101e-01 3.67301375e-01 -7.96047568e-01
1.33152950e+00 -2.09028077e+00 9.65504274e-02 -1.56089395e-01
4.79761511e-01 5.81315935e-01 -1.44893229e-01 2.20479280e-01
4.91818897e-02 -7.39695057e-02 1.97091982e-01 -1.04052341e+00
2.51321584e-01 -3.11983943e-01 1.39914930e-01 6.70860529e-01
1.38615891e-01 1.12809730e+00 -8.20397437e-01 -4.36173558e-01
2.70575523e-01 5.64677060e-01 -3.46654981e-01 3.00496936e-01
6.12246335e-01 4.31199610e-01 -3.96816522e-01 8.51225793e-01
8.12827170e-01 -3.07921916e-01 -4.05768514e-01 -3.58935863e-01
-5.13703287e-01 -1.53127030e-01 -1.23333955e+00 1.41999114e+00
-7.35849962e-02 5.84195316e-01 -1.11221075e-01 -5.88460326e-01
7.96160996e-01 6.45523220e-02 1.85711607e-01 -6.33709192e-01
1.01507060e-01 1.96541503e-01 -3.59966218e-01 -2.70316601e-01
7.52580523e-01 5.95454097e-01 -1.97801858e-01 7.88527727e-02
7.15894392e-03 9.66061652e-01 9.73574296e-02 7.26323500e-02
8.45044076e-01 -1.73316002e-01 3.79077345e-01 -2.77522296e-01
9.34403419e-01 -2.83406436e-01 5.78930080e-01 1.08510995e+00
-8.35128903e-01 7.38934636e-01 -2.07519323e-01 -6.02799177e-01
-1.08120596e+00 -1.01546204e+00 -1.07364975e-01 1.27676010e+00
4.48546469e-01 -4.92515653e-01 -8.55837226e-01 -6.42364383e-01
2.19492912e-01 2.60519147e-01 -6.44079745e-01 3.27242911e-02
-7.07055926e-01 -9.58006918e-01 6.84898734e-01 6.02703929e-01
7.77496338e-01 -6.46433055e-01 -3.15868199e-01 -2.53018755e-02
-3.44988912e-01 -1.31695545e+00 -1.00229979e+00 -4.23139900e-01
1.45028919e-01 -9.67617691e-01 -1.42412937e+00 -1.03458667e+00
7.16343462e-01 8.38491857e-01 7.43183911e-01 1.17026873e-01
-4.78528470e-01 4.02898341e-01 -3.24191988e-01 -2.33705461e-01
3.68901104e-01 1.91977307e-01 4.76848274e-01 3.27176601e-01
8.44694018e-01 -6.74237013e-02 -1.22174990e+00 6.76681876e-01
-2.00218454e-01 -1.49249472e-02 5.63596725e-01 7.50416160e-01
5.56270540e-01 -3.45788330e-01 4.36824262e-01 -1.04523964e-01
5.22793710e-01 -3.00776660e-01 -4.17614430e-01 4.38036054e-01
-3.91126901e-01 -2.76992023e-01 4.11086142e-01 -4.65794832e-01
-7.08648384e-01 2.31636599e-01 -3.99544448e-01 -2.68184841e-01
-2.48495042e-01 -2.50868142e-01 -3.57674450e-01 -3.64183366e-01
6.59821093e-01 5.44661462e-01 -3.40615451e-01 -6.43286824e-01
3.67393166e-01 9.80354309e-01 8.18150342e-01 -2.71314949e-01
1.09728789e+00 5.17277956e-01 -4.30182964e-01 -5.26827931e-01
-8.99848044e-01 -1.08436334e+00 -7.96831667e-01 -2.83473581e-01
9.14770126e-01 -1.60154617e+00 -7.62325466e-01 7.13341653e-01
-1.17632043e+00 4.88690771e-02 3.48549597e-02 4.51413661e-01
-8.36650953e-02 4.83578175e-01 -3.57127458e-01 -8.35225165e-01
-6.07085168e-01 -1.14238811e+00 1.24818778e+00 6.73636079e-01
8.97371471e-02 -5.56228757e-01 -2.20078472e-02 4.36199129e-01
4.26518708e-01 3.89136869e-04 -6.49147853e-02 -7.36843944e-01
-6.22022688e-01 -4.64305997e-01 -7.51493156e-01 9.29416195e-02
1.54873114e-02 -6.05337143e-01 -1.10356605e+00 -6.57440603e-01
-4.54662502e-01 5.34689985e-02 1.06123400e+00 1.73142627e-01
8.90063703e-01 -4.05701250e-01 -6.77446067e-01 6.86023474e-01
1.13191855e+00 -3.08981985e-01 3.03956926e-01 5.63392878e-01
8.60963881e-01 3.84757370e-01 5.92816174e-01 4.61956412e-01
1.03337240e+00 1.23316193e+00 -8.35597329e-03 -3.46103609e-01
-5.10655344e-01 -4.18393672e-01 2.00851783e-01 3.55686188e-01
-1.52380526e-01 -2.11473271e-01 -8.85923624e-01 7.20671356e-01
-2.27127600e+00 -8.69013250e-01 -3.50830816e-02 2.17092037e+00
4.92949963e-01 -1.96025699e-01 6.86350644e-01 -2.89905727e-01
1.14732361e+00 1.03155091e-01 -4.70201820e-01 3.91516358e-01
-8.83370861e-02 -2.87629515e-01 6.86454594e-01 4.66893584e-01
-1.54313815e+00 1.11883581e+00 5.15207863e+00 8.95722210e-01
-5.97884595e-01 3.96175891e-01 3.92423064e-01 -5.95304556e-02
2.74256825e-01 -2.14415208e-01 -1.60443473e+00 8.55825961e-01
5.33574283e-01 -2.51449287e-01 2.40715608e-01 1.08745313e+00
2.70963341e-01 1.18419446e-01 -1.10832703e+00 1.50760627e+00
3.49215060e-01 -1.11418092e+00 -6.22590221e-02 5.96975312e-02
3.59211415e-01 1.11069679e-02 9.92724746e-02 4.06269550e-01
9.34477150e-03 -7.70869672e-01 7.54989803e-01 5.41285455e-01
6.56167746e-01 -6.86070681e-01 9.20329630e-01 2.30812967e-01
-1.76061451e+00 -2.79812157e-01 -4.57706630e-01 2.46130064e-01
5.26723742e-01 2.39471495e-01 -3.46230179e-01 5.74657261e-01
1.11703241e+00 6.32728159e-01 -1.11253476e+00 1.62059426e+00
-1.57072008e-01 1.84426643e-02 -3.00338000e-01 -6.22930415e-02
7.46495053e-02 1.35728896e-01 6.43161118e-01 1.50911069e+00
3.71682316e-01 -7.23219961e-02 3.55959743e-01 7.75039971e-01
1.13012008e-01 4.97461706e-02 -2.05268443e-01 6.21361494e-01
7.96479404e-01 1.29393351e+00 -2.88548321e-01 -3.24825227e-01
-5.94523728e-01 1.31128705e+00 5.77146888e-01 3.63818288e-01
-1.07371008e+00 -2.06282333e-01 9.13377762e-01 1.94115639e-01
4.86437410e-01 -1.93497747e-01 2.25779727e-01 -1.35452533e+00
2.56317258e-01 -5.76450527e-01 4.32217985e-01 -4.88726169e-01
-1.44313920e+00 7.04415679e-01 -1.20176628e-01 -1.23859274e+00
4.95294370e-02 -5.99668384e-01 -5.50006986e-01 1.13099849e+00
-1.75648570e+00 -1.65804696e+00 -7.41705477e-01 6.90273523e-01
5.20588815e-01 -2.69939810e-01 7.06869125e-01 8.54518652e-01
-1.17476857e+00 1.32494283e+00 -1.56309381e-01 4.69778150e-01
8.81643057e-01 -1.08360159e+00 8.68309796e-01 1.21446919e+00
-1.20573409e-01 7.38822103e-01 4.89483953e-01 -5.17949700e-01
-1.03253973e+00 -1.36341751e+00 1.08204389e+00 -6.44229710e-01
4.91212964e-01 -7.07071483e-01 -6.12224877e-01 5.41850269e-01
-1.65120915e-01 2.38123521e-01 6.44921541e-01 9.32070091e-02
-3.86870027e-01 -1.38443947e-01 -9.44271743e-01 8.09210360e-01
1.40217578e+00 -4.67675716e-01 -2.95461148e-01 3.52176696e-01
6.70227408e-01 -3.24973851e-01 -5.91673672e-01 2.52844661e-01
5.92618585e-01 -8.68375421e-01 1.51096869e+00 -2.32365698e-01
-4.32084173e-01 -6.78988695e-01 -7.23538473e-02 -1.00733840e+00
-9.16019797e-01 -5.41537583e-01 -2.79594243e-01 1.39045799e+00
1.58042207e-01 -8.53593826e-01 6.79926813e-01 6.97759569e-01
1.37103081e-01 -6.07438624e-01 -1.12501872e+00 -8.87327731e-01
-3.19586635e-01 5.14886305e-02 7.10588694e-01 4.13031965e-01
-3.68647963e-01 2.70931274e-01 -7.68939197e-01 5.85568666e-01
9.17161286e-01 -2.77015507e-01 1.15253437e+00 -1.14768624e+00
-1.13159321e-01 -3.55633050e-01 -6.79359138e-01 -1.56592512e+00
-3.92073281e-02 -6.23500824e-01 4.61452566e-02 -1.46022308e+00
6.17644250e-01 -4.68297392e-01 -2.72422135e-01 3.11124176e-01
-6.67060971e-01 3.12475771e-01 3.74824196e-01 6.27051711e-01
-8.84597898e-01 6.77484870e-01 6.80768609e-01 -6.40773550e-02
-2.02920362e-02 2.29270399e-01 -9.70700443e-01 4.77785110e-01
7.85869718e-01 -2.57102430e-01 1.63388088e-01 -5.75462580e-01
-2.62186855e-01 -7.37027347e-01 9.93860841e-01 -1.26386428e+00
7.80907035e-01 3.59482825e-01 5.78101993e-01 -7.35253036e-01
5.26425660e-01 -4.58759189e-01 -1.78635549e-02 3.68288994e-01
-9.15643051e-02 2.47770786e-01 4.37296890e-02 6.37069166e-01
-6.86634704e-02 -1.26678646e-01 6.72810733e-01 -6.69967979e-02
-1.09779406e+00 6.52026236e-01 1.37005076e-01 -2.61938304e-01
1.15964484e+00 -1.82202473e-01 -5.52472889e-01 -1.47393659e-01
-5.29428899e-01 5.94449043e-01 3.87517005e-01 6.37533069e-01
5.63812017e-01 -1.59198821e+00 -9.43815351e-01 8.29155222e-02
4.31498647e-01 -1.65915579e-01 5.48783898e-01 8.37540686e-01
-1.62796021e-01 6.30411685e-01 9.94142070e-02 -6.13738179e-01
-1.41197371e+00 5.60477793e-01 4.09094751e-01 -2.04406813e-01
-6.96501732e-01 1.02082908e+00 3.87690604e-01 -3.98043126e-01
4.58590060e-01 4.20433939e-01 -4.70775545e-01 -8.07138067e-03
9.58494067e-01 4.05208319e-01 -3.50801796e-01 -1.06879318e+00
-7.57916391e-01 7.56538093e-01 -3.36481780e-01 1.40381947e-01
8.98142874e-01 -4.84281212e-01 3.14112306e-01 -2.09441051e-01
1.15527070e+00 -1.07865147e-01 -1.35409021e+00 -5.53454220e-01
-1.87866375e-01 -6.45420194e-01 -2.60290027e-01 -4.07406807e-01
-6.68581665e-01 5.56640267e-01 1.02793372e+00 -9.44399908e-02
7.33051240e-01 3.20413917e-01 9.16339576e-01 1.57999292e-01
4.86726344e-01 -1.12012517e+00 2.88579445e-02 1.69968590e-01
8.93183112e-01 -1.46689427e+00 2.81194653e-02 -4.50752378e-01
-5.92769682e-01 6.77937090e-01 9.08944547e-01 -2.76162736e-02
4.71300930e-01 -2.86845177e-01 -1.45304069e-01 1.15026265e-01
-8.31353962e-02 -7.98966825e-01 5.32263100e-01 7.16490328e-01
-1.12758592e-01 2.29673341e-01 -9.07384306e-02 9.35599625e-01
-1.99797645e-01 -2.25455284e-01 -5.57928160e-02 5.70694327e-01
-4.15754586e-01 -8.96751523e-01 -4.16627556e-01 8.11719149e-02
-2.04903126e-01 -1.18178710e-01 -1.45328328e-01 7.88142562e-01
2.95377761e-01 9.45511460e-01 -4.54529114e-02 -5.87610781e-01
5.82642138e-01 -3.36302876e-01 7.91401416e-02 -3.75854999e-01
-3.82268608e-01 -1.82829291e-01 5.12563922e-02 -5.77003419e-01
-3.07451636e-01 -6.97277606e-01 -5.80595016e-01 -4.43476170e-01
-3.98951054e-01 -1.28180778e-04 3.73552203e-01 6.91595972e-01
7.16581702e-01 1.78130046e-01 5.94308317e-01 -1.04566014e+00
-5.86920083e-01 -8.65914226e-01 -1.22322477e-01 3.44106615e-01
4.30197835e-01 -8.39923084e-01 -3.03361565e-01 -2.85300225e-01] | [14.807944297790527, 0.833098292350769] |
d6a8a405-894b-4e33-a70f-83ad4fafa8c5 | learning-clause-representation-from | null | null | https://aclanthology.org/2021.textgraphs-1.6 | https://aclanthology.org/2021.textgraphs-1.6.pdf | Learning Clause Representation from Dependency-Anchor Graph for Connective Prediction | Semantic representation that supports the choice of an appropriate connective between pairs of clauses inherently addresses discourse coherence, which is important for tasks such as narrative understanding, argumentation, and discourse parsing. We propose a novel clause embedding method that applies graph learning to a data structure we refer to as a dependency-anchor graph. The dependency anchor graph incorporates two kinds of syntactic information, constituency structure, and dependency relations, to highlight the subject and verb phrase relation. This enhances coherence-related aspects of representation. We design a neural model to learn a semantic representation for clauses from graph convolution over latent representations of the subject and verb phrase. We evaluate our method on two new datasets: a subset of a large corpus where the source texts are published novels, and a new dataset collected from students’ essays. The results demonstrate a significant improvement over tree-based models, confirming the importance of emphasizing the subject and verb phrase. The performance gap between the two datasets illustrates the challenges of analyzing student’s written text, plus a potential evaluation task for coherence modeling and an application for suggesting revisions to students. | ['Rebecca J. Passonneau', 'Ting-Hao Huang', 'Yanjun Gao'] | null | null | null | null | naacl-textgraphs-2021-6 | ['discourse-parsing'] | ['natural-language-processing'] | [ 2.30417684e-01 8.30260396e-01 -6.21250749e-01 -5.38777769e-01
-5.05670786e-01 -5.97048879e-01 9.12713885e-01 9.04311359e-01
-2.85905421e-01 5.82141638e-01 1.18831122e+00 -5.92767119e-01
-1.83277428e-01 -1.07121718e+00 -5.37884116e-01 -3.26064914e-01
1.19787380e-01 2.15415329e-01 7.98594728e-02 -5.65390110e-01
5.30233502e-01 -3.53699662e-02 -1.27078593e+00 7.17724919e-01
8.98190260e-01 3.97236764e-01 4.47183661e-02 4.21819031e-01
-6.78824127e-01 1.26945782e+00 -8.75795424e-01 -5.24364233e-01
-5.74691653e-01 -5.29806614e-01 -1.28085279e+00 -4.67535444e-02
4.71542180e-01 7.31867328e-02 -1.79104269e-01 9.42195654e-01
9.60704014e-02 2.01421186e-01 5.94673336e-01 -9.16970670e-01
-8.93644273e-01 1.35856771e+00 -2.76614845e-01 3.45022798e-01
5.09384096e-01 -3.39972913e-01 1.85391283e+00 -4.05359238e-01
1.08327103e+00 1.46005487e+00 6.02798522e-01 5.16457796e-01
-1.09490681e+00 -2.83230454e-01 5.45555353e-01 5.26226044e-01
-5.44408858e-01 -1.38186291e-01 1.20071709e+00 -6.37500942e-01
1.12798607e+00 2.23664567e-01 9.50260520e-01 1.30325782e+00
8.53051916e-02 8.90272856e-01 8.24300647e-01 -7.25855529e-01
2.26055771e-01 -1.15159072e-01 1.01941526e+00 6.73524499e-01
1.47549242e-01 -2.10711256e-01 -8.23941886e-01 -5.23700416e-02
1.59377769e-01 -4.46395248e-01 -5.47545910e-01 -3.12787324e-01
-9.61335719e-01 1.48707163e+00 5.54347277e-01 4.36538428e-01
6.26064241e-02 2.14775711e-01 5.38229048e-01 2.94237375e-01
5.22186816e-01 5.98262668e-01 -2.29272932e-01 -1.97522596e-01
-4.43335116e-01 3.77629578e-01 1.00680697e+00 7.57331789e-01
2.19400540e-01 -2.06367165e-01 -3.33714634e-01 6.61532700e-01
5.62444687e-01 -2.56137520e-01 4.05730575e-01 -8.91488075e-01
8.45870912e-01 1.09941697e+00 -3.93561959e-01 -1.08113039e+00
-5.06660521e-01 -3.77873182e-01 -2.53202528e-01 -1.73232377e-01
3.21909398e-01 1.41060233e-01 -1.74479961e-01 1.96235394e+00
3.33853155e-01 -2.13345155e-01 4.65832770e-01 5.46541154e-01
1.61309874e+00 5.74519515e-01 1.24527700e-01 -1.09885991e-01
1.68816149e+00 -1.05197239e+00 -1.05378079e+00 -2.57027596e-01
1.10235631e+00 -5.37882149e-01 1.13132727e+00 -1.72945887e-01
-1.09976363e+00 -1.76375598e-01 -1.28903437e+00 -5.74497283e-01
-3.05664897e-01 -1.39308110e-01 7.76761949e-01 1.83030337e-01
-7.10019588e-01 6.27856135e-01 -6.68365777e-01 -1.59025788e-01
5.02373457e-01 -1.11225665e-01 -3.52202117e-01 8.42080861e-02
-1.31672490e+00 8.66227508e-01 4.20164526e-01 -1.59317121e-01
-1.53513660e-03 -8.09940457e-01 -1.41449428e+00 6.71294391e-01
4.07157153e-01 -6.38356149e-01 1.21669531e+00 -5.26936710e-01
-1.53107893e+00 1.09165597e+00 -1.11299129e-02 -4.77684259e-01
1.89343899e-01 -9.62494537e-02 1.42795388e-02 1.51417911e-01
3.14685017e-01 3.90436888e-01 4.18087095e-01 -7.20583916e-01
-3.81057799e-01 -2.67993778e-01 4.36735719e-01 1.48726255e-01
-4.04818624e-01 -8.25909972e-02 2.80000046e-02 -6.96228325e-01
2.43164062e-01 -6.69768155e-01 1.56379759e-01 -3.36017936e-01
-6.47166252e-01 -9.85023260e-01 7.77727544e-01 -5.76246798e-01
1.37608171e+00 -2.02234650e+00 3.55135083e-01 -1.35684699e-01
5.50042272e-01 -2.55867988e-01 -1.57769501e-01 6.58579171e-01
-2.59488344e-01 3.55872840e-01 -2.26672411e-01 -1.07960463e-01
1.49348035e-01 9.58967283e-02 -4.24289763e-01 2.05752239e-01
4.01025802e-01 1.11537027e+00 -9.71649706e-01 -4.29717302e-01
-1.57451212e-01 3.20772558e-01 -7.00742960e-01 1.46409392e-01
-6.05053782e-01 1.07140273e-01 -5.08265138e-01 9.00547877e-02
9.48711187e-02 -6.58848166e-01 6.69657469e-01 -3.96114528e-01
-4.50525954e-02 1.38580585e+00 -8.65205526e-01 1.71995044e+00
-1.63158357e-01 1.15952659e+00 -1.46368489e-01 -1.01160479e+00
8.55485141e-01 2.88417608e-01 1.94886029e-02 -6.01523399e-01
9.34644192e-02 -2.27535069e-01 2.83778518e-01 -5.67142963e-01
6.12420440e-01 1.55100420e-01 -3.44454795e-01 8.58070731e-01
1.57569110e-01 -4.18467730e-01 5.91657043e-01 8.24800849e-01
1.13060522e+00 1.43340990e-01 5.35892546e-01 -5.46108246e-01
3.22676867e-01 1.55793112e-02 3.53535563e-01 3.39733481e-01
1.34554669e-01 2.34469652e-01 1.36709380e+00 -4.92224306e-01
-7.02953696e-01 -7.38064528e-01 -1.83771223e-01 1.08541632e+00
-8.48097876e-02 -1.05672133e+00 -4.00569975e-01 -8.54224920e-01
2.40807850e-02 1.06331289e+00 -9.27933931e-01 -6.75858855e-02
-9.92783070e-01 -4.03742433e-01 1.64417431e-01 7.28356838e-01
2.34758839e-01 -9.09444451e-01 -7.96020389e-01 2.05021918e-01
-3.92826229e-01 -1.20689046e+00 -2.83804357e-01 4.12058830e-01
-6.52373374e-01 -1.45592022e+00 5.22267744e-02 -9.48493779e-01
4.80088830e-01 -1.21280001e-02 1.54936898e+00 4.79949176e-01
2.63178974e-01 3.23831975e-01 -3.30219805e-01 -3.73061717e-01
-5.94403386e-01 2.24476263e-01 -6.80167556e-01 -7.24352479e-01
3.08316976e-01 -5.14273763e-01 -2.05205083e-01 -5.70913017e-01
-6.07898831e-01 4.37644035e-01 -1.36493519e-01 1.17227662e+00
8.31851512e-02 -3.88572454e-01 4.75450128e-01 -1.41462278e+00
1.24614990e+00 -4.22398031e-01 -5.02411962e-01 3.74847680e-01
-5.18812597e-01 5.17466724e-01 2.57984519e-01 -1.08471043e-01
-8.98722708e-01 -7.55514443e-01 5.22094928e-02 3.28106076e-01
4.00773346e-01 1.02332366e+00 -6.62530661e-02 5.55785596e-01
5.67534268e-01 -4.80071783e-01 -1.33102508e-02 -9.52862874e-02
6.92173243e-01 2.19627842e-01 4.98274922e-01 -8.39313507e-01
3.50937694e-01 1.89823657e-02 1.19395643e-01 -6.18646860e-01
-1.23041272e+00 -1.43216744e-01 -6.37911737e-01 1.29053578e-01
9.36153412e-01 -8.18588138e-01 -8.23142648e-01 -3.39910209e-01
-1.68619800e+00 -2.32978523e-01 -6.37842059e-01 4.42768782e-01
-3.10160488e-01 3.28425109e-01 -9.76692379e-01 -2.38856375e-01
-3.22256684e-01 -1.00399089e+00 8.38623166e-01 2.65846312e-01
-7.73821831e-01 -1.52133334e+00 2.53263205e-01 6.56934321e-01
6.22405224e-02 5.88952541e-01 1.74846363e+00 -8.87914002e-01
-4.91210967e-01 1.97090045e-01 -1.74662009e-01 -1.80197373e-01
5.53780794e-02 1.95662558e-01 -8.66492748e-01 -9.59422812e-02
-7.56296441e-02 -5.68571389e-01 1.05518591e+00 2.03909248e-01
8.19926620e-01 -5.69267154e-01 -2.47826084e-01 2.71050215e-01
1.08779705e+00 -1.41120613e-01 2.39413351e-01 4.83507842e-01
7.28817046e-01 8.51363420e-01 1.69086054e-01 7.09294900e-02
8.50651562e-01 5.14517486e-01 2.90240228e-01 3.14476669e-01
-3.61199230e-01 -2.02051610e-01 1.85334548e-01 1.30672443e+00
2.52113938e-01 -3.36584210e-01 -1.01568079e+00 6.14934385e-01
-1.87621450e+00 -1.02906358e+00 -2.75374532e-01 1.43747890e+00
1.31204212e+00 3.22708309e-01 -2.19750926e-01 9.02172774e-02
4.37575042e-01 6.68110013e-01 -7.76406005e-02 -6.75465524e-01
-1.59349769e-01 6.02846146e-01 -2.71867096e-01 9.12335217e-01
-9.48275864e-01 1.04050481e+00 5.80484962e+00 4.51553494e-01
-9.97607648e-01 -3.32140476e-02 3.30531269e-01 1.31421357e-01
-1.05746090e+00 2.58596539e-01 -7.02992678e-01 1.44793212e-01
7.96586812e-01 -4.93986487e-01 -7.88432732e-02 6.64161444e-01
-1.77496493e-01 -1.90774072e-02 -1.59483659e+00 4.04498130e-01
1.70103148e-01 -2.18230772e+00 1.23242855e-01 -2.44621739e-01
6.72931015e-01 -4.52198267e-01 -1.72391266e-01 4.52004999e-01
4.39700007e-01 -1.20726848e+00 7.11025059e-01 -2.34680213e-02
4.73755389e-01 -5.04985392e-01 6.68415427e-01 1.16290517e-01
-1.10757792e+00 1.70451999e-01 1.82454884e-02 -5.04801571e-01
-9.93608907e-02 2.40581319e-01 -8.94493461e-01 4.00675237e-01
3.80560219e-01 1.07214367e+00 -6.41590714e-01 1.41531795e-01
-1.13824010e+00 8.05881321e-01 1.37315616e-01 -4.25225943e-01
2.91446418e-01 -1.32575765e-01 5.09281754e-01 1.24429250e+00
-1.88124731e-01 2.97874212e-01 1.76435053e-01 1.34539449e+00
-3.24154556e-01 9.95385498e-02 -5.81690311e-01 -3.60895127e-01
8.10898185e-01 1.18575132e+00 -6.84992075e-01 -1.27026647e-01
-4.92926121e-01 3.01730603e-01 8.70542407e-01 2.29451120e-01
-4.78625774e-01 -2.61846155e-01 3.88547063e-01 -9.50372741e-02
3.13989520e-01 -2.86023080e-01 -4.64860618e-01 -1.04281044e+00
2.48889271e-02 -8.33375633e-01 5.44795632e-01 -5.19655228e-01
-1.11082232e+00 3.95502955e-01 2.18673319e-01 -6.30172372e-01
-4.55425680e-01 -5.82482755e-01 -1.25690091e+00 6.89274013e-01
-1.59734118e+00 -1.17166185e+00 -1.80752128e-01 6.44362122e-02
5.32768667e-01 -1.06279746e-01 9.77584779e-01 -3.46357673e-01
-5.66071749e-01 4.69545066e-01 -3.23097944e-01 4.58841354e-01
2.81469822e-01 -1.54980266e+00 5.44146061e-01 6.44944549e-01
5.19166529e-01 8.51852536e-01 4.87399787e-01 -4.82956439e-01
-9.53879476e-01 -6.07518613e-01 1.29525840e+00 -5.11537194e-01
7.97827065e-01 -3.64096254e-01 -1.19014311e+00 8.41057658e-01
7.54335880e-01 -4.06575322e-01 1.21695411e+00 7.02563524e-01
-5.22500753e-01 4.96433347e-01 -6.29720449e-01 7.15433240e-01
8.81695032e-01 -6.33213460e-01 -1.44277239e+00 3.93392056e-01
1.24928701e+00 -6.59932733e-01 -8.32066417e-01 2.42122665e-01
1.78843141e-01 -6.95347250e-01 6.57645643e-01 -1.07756877e+00
1.17730081e+00 2.23178402e-01 -6.34074286e-02 -1.28858840e+00
-3.50254923e-01 -4.94525611e-01 -4.70403105e-01 1.30472147e+00
3.70371282e-01 -2.45781526e-01 7.33372569e-01 5.88438034e-01
-3.96869719e-01 -1.02034247e+00 -8.86733294e-01 -1.69162154e-01
4.65386093e-01 -1.05256639e-01 5.01063347e-01 1.44981790e+00
7.06398606e-01 1.15217328e+00 4.81049895e-01 -3.10967714e-01
4.30694431e-01 7.22457170e-01 5.61480105e-01 -1.51317394e+00
-8.61622840e-02 -8.41646254e-01 -1.68741882e-01 -8.28685641e-01
8.07200372e-01 -1.44205153e+00 -5.04689097e-01 -1.97370708e+00
2.83196002e-01 -8.69652480e-02 1.54231057e-01 5.04380763e-01
-3.94246072e-01 -6.30967498e-01 1.94500595e-01 -4.95559014e-02
-4.30043787e-01 7.70837247e-01 1.29315042e+00 -5.45135796e-01
-1.69896603e-01 -5.32327354e-01 -9.34606194e-01 8.82722557e-01
5.72625458e-01 -3.86744231e-01 -4.86605495e-01 -4.88838434e-01
5.78218758e-01 3.94536927e-02 3.19790512e-01 -3.95343542e-01
3.12117159e-01 -4.07994837e-01 7.17876032e-02 -5.20866454e-01
1.06770426e-01 -2.37332106e-01 -5.62524974e-01 3.80947471e-01
-9.98146772e-01 1.61513761e-01 2.33752489e-01 4.39845234e-01
-4.09469396e-01 -2.88900733e-01 4.55444604e-01 -1.48180917e-01
-5.28364003e-01 -3.80252659e-01 -1.94461912e-01 5.75537920e-01
7.35062540e-01 2.94054579e-02 -1.03900838e+00 -3.73240322e-01
-7.06033468e-01 3.78441721e-01 7.47482628e-02 5.99929333e-01
7.45853543e-01 -1.42323327e+00 -9.12616491e-01 5.90416603e-02
1.19443797e-01 2.37616867e-01 -1.67820588e-01 4.53825086e-01
-5.60811698e-01 3.85098517e-01 -9.60343555e-02 -5.78519940e-01
-1.50425112e+00 1.51702583e-01 -3.55313085e-02 -7.18949974e-01
-8.38854969e-01 1.09279907e+00 3.17293704e-02 -3.48272711e-01
1.10302672e-01 -8.14920604e-01 -8.04205239e-01 2.77182609e-01
1.17343657e-01 7.70801827e-02 -7.31967464e-02 -4.71341103e-01
-2.41093591e-01 2.84298390e-01 -2.45737776e-01 -3.81078050e-02
1.60589910e+00 7.87037686e-02 -3.74117613e-01 6.81187093e-01
1.08062851e+00 3.75943512e-01 -8.80154490e-01 -3.34848434e-01
5.49476147e-01 -6.66515082e-02 -4.62794863e-02 -4.07008201e-01
-6.46488726e-01 1.01273704e+00 -1.39921501e-01 3.35220188e-01
3.12139004e-01 3.15294981e-01 4.98918623e-01 5.32123327e-01
-1.98630422e-01 -1.04653311e+00 6.38428032e-01 9.44975972e-01
1.12345493e+00 -1.12139928e+00 3.19870949e-01 -6.42274082e-01
-4.57167536e-01 1.44273758e+00 6.72571301e-01 -4.47018407e-02
4.12249565e-01 2.69264877e-01 -6.39298186e-02 -7.26057291e-01
-1.14359248e+00 6.59978911e-02 5.20376205e-01 2.85790682e-01
1.11744881e+00 1.09829649e-01 -6.24607623e-01 9.28236783e-01
-7.64507532e-01 -6.44189060e-01 8.81679416e-01 1.09124899e+00
-2.24656209e-01 -1.21610379e+00 -1.64389703e-02 2.07479715e-01
-1.91231921e-01 -3.49477649e-01 -6.81719899e-01 9.16033745e-01
-2.23955140e-01 9.24455345e-01 3.22925210e-01 -1.95587948e-02
7.12918416e-02 3.12595516e-01 6.61179245e-01 -1.07740664e+00
-8.36915135e-01 -3.03538799e-01 5.22991955e-01 -3.57756972e-01
-6.30327165e-01 -4.56209809e-01 -1.66050971e+00 -3.21718216e-01
-4.83481318e-01 5.05520225e-01 5.39726555e-01 1.10323417e+00
2.44865432e-01 1.07467914e+00 1.36688069e-01 -1.76819921e-01
-5.38017452e-01 -9.41295862e-01 -1.46234930e-01 6.77284479e-01
2.94972301e-01 -6.40053689e-01 -1.77913383e-01 -4.06654291e-02] | [10.890220642089844, 9.28274917602539] |
dc58718e-a52b-42bc-9d30-8853f3b66c1e | dl-drl-a-double-layer-deep-reinforcement | 2208.02447 | null | https://arxiv.org/abs/2208.02447v3 | https://arxiv.org/pdf/2208.02447v3.pdf | DL-DRL: A double-level deep reinforcement learning approach for large-scale task scheduling of multi-UAV | Exploiting unmanned aerial vehicles (UAVs) to execute tasks is gaining growing popularity recently. To solve the underlying task scheduling problem, the deep reinforcement learning (DRL) based methods demonstrate notable advantage over the conventional heuristics as they rely less on hand-engineered rules. However, their decision space will become prohibitively huge as the problem scales up, thus deteriorating the computation efficiency. To alleviate this issue, we propose a double-level deep reinforcement learning (DL-DRL) approach based on a divide and conquer framework (DCF), where we decompose the task scheduling of multi-UAV into task allocation and route planning. Particularly, we design an encoder-decoder structured policy network in our upper-level DRL model to allocate the tasks to different UAVs, and we exploit another attention based policy network in our lower-level DRL model to construct the route for each UAV, with the objective to maximize the number of executed tasks given the maximum flight distance of the UAV. To effectively train the two models, we design an interactive training strategy (ITS), which includes pre-training, intensive training and alternate training. Experimental results show that our DL-DRL performs favorably against the learning-based and conventional baselines including the OR-Tools, in terms of solution quality and computation efficiency. We also verify the generalization performance of our approach by applying it to larger sizes of up to 1000 tasks. Moreover, we also show via an ablation study that our ITS can help achieve a balance between the performance and training efficiency. | ['Witold Pedrycz', 'Guohua Wu', 'Zhiguang Cao', 'Mingfeng Fan', 'Xiao Mao'] | 2022-08-04 | null | null | null | null | ['self-learning'] | ['natural-language-processing'] | [ 0.08418955 0.04024578 -0.20249002 -0.08527556 -0.60231835 -0.7560742
0.3043226 -0.0471041 -0.46512237 0.69785535 -0.04505197 -0.6482539
-0.3862844 -0.7116381 -0.88295496 -0.60634667 -0.2331431 0.21021211
0.02072825 -0.23858498 -0.07500894 0.25356305 -1.3763376 -0.08353584
1.1063597 1.1403662 0.7351068 0.49337274 0.46580103 0.8448377
-0.5916327 -0.07818973 0.7048122 -0.07770354 -0.8740153 0.26537913
-0.04481434 -0.61147726 -0.39357486 0.81971186 0.43244454 0.268903
0.29545876 -1.521231 -0.44357476 0.371737 -0.6249233 -0.05672008
0.12842928 0.42966592 1.0185252 -0.27174196 0.22939919 1.0396115
0.4430332 0.41093907 -1.0529357 -0.47756928 0.69130343 0.04127377
-1.0956568 -0.09215537 0.49045795 -0.3734521 1.0512112 -0.14963509
0.5321099 0.8580756 0.25123242 0.9132678 0.8609838 -0.23643802
0.25218844 -0.39259872 -0.37227568 0.97571284 0.14462805 0.40261725
-0.08589327 -0.03934501 0.7918521 0.06408162 -0.30428594 -0.4686452
-1.179892 0.8256493 0.6704297 -0.10694739 -0.64842737 0.37582153
0.565017 0.5287767 0.27858374 0.94056195 -0.7933143 0.09590498
-0.78149456 0.52167326 0.57665575 1.2627742 0.7919531 0.13658287
-0.4036657 0.5491721 0.08808658 0.31189695 0.27252504 -1.1088301
0.7197443 0.56731343 0.4329932 -0.7904431 -0.59021014 -0.5082356
-0.67430407 0.23530768 0.0994259 -0.90230924 -0.8966371 1.8978405
0.32588255 0.06844723 0.19440675 1.1578144 0.14571773 0.7685093
0.09614152 -0.12823632 1.2196678 -1.3875887 -0.46003062 -0.5139662
0.84827536 -0.47022828 1.0338767 0.2937966 -0.9949961 -0.61746246
-1.0766146 0.20172545 -0.17211702 0.61506414 0.82440424 0.26113063
-1.0150869 0.47796562 -0.84248894 -0.00761073 0.36456493 0.54585654
-0.15622562 -0.16236289 -1.1180018 0.72887766 0.43491217 0.22260243
-1.5830588 -0.54792243 -1.0668471 0.2978201 1.1017056 -0.8369139
1.5884346 -0.84179705 -1.5325243 0.2858127 0.15813962 -0.59142756
0.18076085 -0.32136264 0.14460689 -0.06819662 0.24763812 0.8211085
0.9823965 -1.171489 -1.0433244 -0.07206061 0.9789458 0.35207784
-0.2146318 -0.2977365 -0.35477224 -0.62691134 -0.44016343 -1.2140933
-0.5466225 -0.4087661 -0.30568677 -0.29200107 0.6581345 -0.55253804
1.3210437 -2.11511 0.46459246 -0.12411667 0.27439174 0.36023802
-0.5876072 0.44922203 0.15441683 -0.12030018 -0.29933766 -0.13237399
0.16501497 0.30109236 -0.33819592 0.18214573 0.36552685 0.8780117
-1.0646853 -0.07068148 -0.01557174 -0.12981305 -0.8305252 0.5649523
-0.61775976 0.35470003 -0.6291168 0.66803795 0.34964857 -0.22423345
0.5443224 0.02120678 -0.17747535 0.3147218 -0.7750792 1.6858085
-0.8855269 0.41202748 0.3267811 -1.1769112 0.6290467 0.253078
0.5025619 -0.66311765 0.04867209 -0.0348452 0.1565903 -0.5335151
0.4229228 0.17595649 -0.31901342 0.38361892 -0.15334786 -0.09810732
0.39597294 -0.04946493 1.2479749 0.38443664 0.22933789 -0.13022974
0.37266445 0.13417716 0.7168243 0.62013996 -0.15516855 -0.19192745
0.882966 -0.70425284 -0.7630811 -0.40009615 0.40253687 1.3107975
0.20100267 -0.44212562 -0.94421446 -1.0768758 0.15284039 0.6441602
-0.5752756 -0.25387526 -0.54132825 -0.6761233 0.32942745 0.47924176
0.6662352 -1.1009215 -1.2935433 0.26634446 -0.11954025 -1.4475771
-0.54647285 0.49243212 -0.5382998 -1.1315507 -0.6916486 -0.7842034
0.7790457 0.4662518 0.94935846 0.05000476 -0.0526321 0.33956614
-0.48056707 -0.32189298 0.04663948 0.49244297 0.11654392 -0.19187522
-0.10933364 -0.39814636 -0.5031791 0.3638521 -0.83458 0.17010786
0.97201085 0.9111928 0.42822722 0.41137233 0.6074085 -0.69265926
0.9169029 -0.40394557 -1.0979608 0.35822004 -0.6238355 0.26806867
0.9855559 -0.21739076 -0.5925077 0.1953402 0.30181673 -0.7157464
-0.04222629 0.7466496 -0.14284152 -0.08502597 0.15588695 0.18630446
-0.06910979 -0.01255123 0.1793675 0.44462448 0.21720695 -0.70508504
0.77540594 -0.10636871 -0.02017831 -0.43207416 -1.0114484 0.01325707
-0.40357846 -0.07542868 0.7702753 -1.0481622 -1.0936686 0.11217736
-1.1486919 -1.0629158 -0.05133022 0.3288147 -0.82189375 0.12413315
-0.3247539 -0.76225156 -0.10896785 -1.6547444 1.316947 0.0837545
0.13093871 -0.80914545 -0.17417692 0.12661913 0.33489951 0.28104448
1.1890389 -0.13689815 -0.67577195 0.02904866 -0.35957763 0.2863316
0.11297786 -0.47279647 -0.49156627 -0.62850994 -0.32818624 -0.6237356
0.6623013 0.47539195 1.406371 -0.3924122 -0.3340105 0.6761789
1.3532466 0.28739747 0.3148184 0.50757265 0.57060903 0.6884813
1.0499853 0.6773411 0.579145 0.77861565 1.046484 -0.11959693
0.3206324 -0.17175809 0.585091 0.29868007 -0.2064201 -0.39307523
-0.91734123 0.40684155 -2.1252093 -0.5289425 0.5315072 2.1845777
0.5222747 -0.00841136 0.23192683 -0.10143203 0.3407293 0.3253694
-0.7919312 -0.3525161 0.522793 -0.07633811 0.6801569 0.42866856
-1.3916254 1.3316464 5.855057 0.57767636 -1.0367677 -0.19085753
0.35887805 -0.25288808 0.14459614 -0.15414147 -0.7082475 0.41201305
0.8113118 -0.0504627 1.0085468 0.91045827 0.3864931 -0.10499281
-1.1192284 0.7964768 -0.32727376 -0.9626702 -0.01823139 0.19280645
0.60986525 -0.07776572 0.05029744 0.8380262 0.6131244 -0.9451859
0.6855541 0.11991274 0.7818225 -0.9913752 0.76169586 0.6332575
-1.2891345 -0.6315977 -0.6115576 -0.28974497 0.01903519 0.26861805
-0.7073549 0.92739105 0.5165326 0.6464205 -0.3646618 0.75902486
-0.42930442 0.33369273 0.04556407 0.01288758 0.8809608 -0.2103727
0.30846924 0.7428569 0.23180941 0.14933753 0.842618 0.65065116
-0.06852333 -0.3381418 -0.75720954 -0.3150717 0.48926008 1.3583198
-0.5665897 -0.07868902 -0.39009875 0.90818214 0.59473425 0.46114656
-1.1565161 -0.4877176 0.9386999 -0.16776842 0.5411468 -0.4877868
0.19457893 -0.79547167 0.05933179 -1.1959617 0.16221553 -0.70352966
-0.9009902 0.84484637 -0.05226935 -1.2764758 -0.29910415 -0.66227466
-0.26397339 0.77327096 -1.9914696 -0.934944 -0.3383319 0.46439245
0.8181293 -0.29848668 0.6110155 0.20755275 -0.93294865 0.502586
-0.31184575 -0.0869427 0.43299812 -1.1867363 0.2893925 0.7429054
-0.24999553 0.20722431 0.3786956 -0.46586314 -1.6960728 -1.4180247
0.3737347 -0.20733598 0.5802994 -0.30153844 -0.28592968 0.8395114
0.25503865 -0.12149092 0.45157853 0.03986903 0.10894267 -0.06233883
-0.7283434 0.6089971 0.9767127 -0.18481088 -0.24299528 0.57686377
1.1103007 -0.57989156 -0.65800697 0.48356816 0.28425378 -0.7039633
0.6487158 -0.8137517 0.7853796 -0.31228387 -0.24160713 -1.6245631
-0.53570753 -0.86722606 0.00777211 0.7433881 0.46770886 -0.4842789
0.5111493 0.45035726 -0.53863907 -1.3088601 -0.4594834 -0.91328
-0.24466139 -0.22166169 0.7744335 0.7913853 -0.18796557 0.50515187
-0.57123816 0.5617331 0.17836806 0.42472944 0.8357583 -0.9316907
-0.4873434 -0.27575466 0.21253256 -1.4055719 0.46261284 -0.78695285
0.3732439 -1.5826176 -0.34422532 -0.6274454 -0.24471329 0.86942637
-0.17631525 -0.48240495 0.36001188 0.084847 -0.94071364 0.79299766
1.478148 -0.00958331 -0.31294593 0.10174982 -0.7812393 0.61965173
0.7600439 -0.35417107 -0.6696168 -0.91270185 0.25765824 0.4624251
0.17868698 -1.039781 0.12027215 -0.30750513 0.04697999 -0.13570948
0.02926696 -1.0980376 -0.39304563 0.6598482 -0.46167424 0.641761
0.31645727 0.6543421 -0.11108489 -0.12149184 0.47500655 -0.02239804
-0.7003827 0.61229146 -0.51392907 -0.07506497 1.352245 0.2985731
-0.12656802 -0.31706318 -0.43648076 0.9816278 0.24122207 0.38476285
0.42212954 -1.0268103 -0.48104972 0.24237242 -0.01260479 0.26241854
0.09601984 0.90333015 -0.49683765 0.77739996 -0.27452156 -0.21133399
-0.7460877 0.9254662 0.230246 -0.82005864 -0.39568016 0.6088667
0.4612367 -0.60792416 0.39272675 -0.36606345 -0.11577951 -0.1385744
0.1830515 0.32277468 -0.06996393 0.07632873 -0.26769966 0.2763219
-0.10996028 0.20184274 1.2117559 0.06401257 0.04676376 -0.18784183
0.5567699 -0.34246317 -1.703841 -0.02761966 0.00906169 -0.19513448
0.28103173 -0.81548166 -1.1868175 0.7292399 0.19558635 0.47203785
1.3946955 -0.50114995 0.91792125 0.65521866 0.59914464 -0.9543084
0.04291601 0.7179416 0.69699085 -1.2087187 -0.18672293 -0.2761324
-0.8632416 0.9439189 0.92179495 -0.3142912 0.25983226 0.31947196
-0.14709827 -0.1607676 -1.1698695 -0.40840992 -0.04360139 0.51264554
0.22294754 0.11175951 -0.06478449 0.6412 -0.14168727 -0.0091854
0.3405938 1.0284526 -0.4053931 -1.11983 -0.0076529 0.3711118
-0.15397942 -0.06294897 -0.19524114 0.99221694 0.18483996 0.80684936
-0.0620131 -0.59904003 0.49051782 -0.15358731 0.4834855 -0.7410735
-0.6281647 -0.08376721 0.08970217 -0.79251206 -0.28754476 -0.22900216
-0.97657204 -0.16200118 -0.18974456 0.18673235 0.33992222 1.0670584
0.6488488 1.0832981 0.93904746 -1.0867931 -0.87429327 -0.6538509
-0.43663198 -0.29945165 0.4593035 -0.91998196 0.05244771 -0.18508299] | [4.8123884201049805, 2.548185110092163] |
581f3a7f-922e-4bd5-8360-caa092f75de0 | understanding-advertisements-with-bert | null | null | https://aclanthology.org/2020.acl-main.674 | https://aclanthology.org/2020.acl-main.674.pdf | Understanding Advertisements with BERT | We consider a task based on CVPR 2018 challenge dataset on advertisement (Ad) understanding. The task involves detecting the viewer{'}s interpretation of an Ad image captured as text. Recent results have shown that the embedded scene-text in the image holds a vital cue for this task. Motivated by this, we fine-tune the base BERT model for a sentence-pair classification task. Despite utilizing the scene-text as the only source of visual information, we could achieve a hit-or-miss accuracy of 84.95{\%} on the challenge test data. To enable BERT to process other visual information, we append image captions to the scene-text. This achieves an accuracy of 89.69{\%}, which is an improvement of 4.7{\%}. This is the best reported result for this task. | ['Kar', 'Silpa Vadakkeeveetil Sreelatha', 'Shirish e', 'Bhargav Kurma', 'Manasi Patwardhan', 'Kanika Kalra'] | 2020-07-01 | null | null | null | acl-2020-6 | ['sentence-pair-classification'] | ['natural-language-processing'] | [ 7.10071504e-01 3.53994220e-01 -8.00344869e-02 -7.74337590e-01
-1.18372297e+00 -6.27477229e-01 6.58416986e-01 1.86093077e-01
-5.25328517e-01 1.39414608e-01 1.50701165e-01 -4.85100538e-01
5.33680797e-01 -2.85566062e-01 -1.15046787e+00 -2.48731941e-01
9.43367407e-02 3.41356844e-01 2.04464763e-01 4.94350679e-02
8.56369585e-02 -6.04333542e-02 -1.38962853e+00 1.09127772e+00
5.20623684e-01 1.46437442e+00 2.12260455e-01 7.45318353e-01
6.20331801e-02 7.09841609e-01 -8.56616318e-01 -1.03851616e+00
2.89246649e-01 -1.04677022e-01 -5.58735430e-01 2.30884701e-01
1.15743101e+00 -4.85242307e-01 -2.54359066e-01 8.06413174e-01
3.15310001e-01 -3.20870161e-01 6.12163663e-01 -1.09826136e+00
-7.79380023e-01 4.92977232e-01 -8.55743766e-01 6.01052880e-01
4.56712842e-01 1.78386629e-01 1.21346486e+00 -9.28385615e-01
5.69536686e-01 1.16406858e+00 2.75824547e-01 3.78747493e-01
-1.31354141e+00 -8.32012832e-01 2.40865424e-01 1.90897256e-01
-1.30026400e+00 -5.44767976e-01 7.76853085e-01 -3.27155858e-01
6.35857880e-01 5.27218759e-01 6.28769815e-01 1.60213423e+00
9.78496820e-02 1.11392963e+00 1.34798408e+00 -2.83529311e-01
3.92589979e-02 5.24461508e-01 2.05216154e-01 4.55455601e-01
4.81829904e-02 -1.76006258e-01 -4.70588505e-01 1.57391518e-01
3.26287657e-01 -3.62786233e-01 -5.70994392e-02 7.15822428e-02
-7.97823668e-01 9.32632029e-01 7.53927886e-01 -8.06038305e-02
3.33807394e-02 2.20134556e-01 1.71354383e-01 -2.25586891e-02
5.66749156e-01 2.04521969e-01 -1.80885702e-01 -1.48454517e-01
-9.91488934e-01 2.44066134e-01 6.73318088e-01 8.45261335e-01
2.40659624e-01 -1.02327526e-01 -4.71502364e-01 7.83763409e-01
3.46084386e-01 7.62726545e-01 -1.59103245e-01 -8.68093550e-01
7.79274166e-01 5.54047406e-01 2.06352189e-01 -9.38884497e-01
-1.15203768e-01 -7.68095791e-01 -3.58032793e-01 -1.96001619e-01
5.21161318e-01 1.05745010e-01 -1.33380270e+00 1.67193532e+00
1.11114129e-01 -3.97430807e-02 -3.53392571e-01 9.52639163e-01
1.09973991e+00 6.15925014e-01 4.25374478e-01 -9.66882184e-02
1.84427094e+00 -8.18141699e-01 -5.57724953e-01 -7.71964073e-01
4.23694789e-01 -8.34956467e-01 1.37031960e+00 4.87896085e-01
-8.74561906e-01 -6.71076953e-01 -1.18751025e+00 -3.41564007e-02
-1.74366757e-01 4.53990966e-01 3.51929724e-01 7.79503286e-01
-1.07126379e+00 -1.91969901e-01 -2.51480967e-01 -3.53149205e-01
7.69996643e-01 3.19590904e-02 -1.23563901e-01 -2.24394351e-01
-8.81353259e-01 5.40452838e-01 -1.85419679e-01 1.40579239e-01
-1.12895095e+00 -7.29076147e-01 -7.06586480e-01 -4.04265989e-03
3.55153412e-01 -5.46169341e-01 1.39981496e+00 -1.35664618e+00
-7.51248240e-01 1.37522173e+00 -3.66868287e-01 -8.20963204e-01
7.51114964e-01 -2.37227380e-01 -7.67502248e-01 4.95070189e-01
3.64715934e-01 1.04561675e+00 1.11902189e+00 -1.52750862e+00
-5.28156519e-01 -5.04792571e-01 1.30243510e-01 -7.14334613e-03
-1.79264396e-01 1.62403375e-01 -6.48883760e-01 -6.90411568e-01
4.20279913e-02 -1.03696334e+00 2.22251505e-01 7.71460384e-02
-5.42864144e-01 -3.67180035e-02 7.20912814e-01 -1.01891673e+00
8.97698224e-01 -2.32471061e+00 -6.52154267e-01 1.73053425e-02
5.06213367e-01 2.43047073e-01 -2.35136047e-01 1.68922156e-01
4.93693203e-02 4.82475281e-01 -8.94395709e-02 -5.79377651e-01
-7.25103775e-03 -4.87140715e-02 -6.28772616e-01 2.76874930e-01
4.33788687e-01 1.19618678e+00 -3.69772822e-01 -4.94969338e-01
5.20493016e-02 4.94747937e-01 -3.81152272e-01 2.73411851e-02
-4.52779382e-01 4.78446960e-01 -4.26288396e-01 5.09681761e-01
8.86669457e-01 -5.12413561e-01 -3.54754776e-02 -4.01013076e-01
2.32772723e-01 2.08833098e-01 -5.97727299e-01 1.36337352e+00
-2.99513102e-01 1.07385552e+00 1.33684650e-02 -8.38610351e-01
8.09191287e-01 -1.59282207e-01 2.97256798e-01 -1.28725004e+00
7.70789832e-02 -1.62547782e-01 1.98010616e-02 -5.08966148e-01
3.05105209e-01 2.11260527e-01 -2.38500699e-01 5.91771305e-02
-3.11942190e-01 7.46791810e-02 -1.08183190e-01 6.94436729e-01
1.11364150e+00 -1.98423401e-01 -3.14662069e-01 -7.32266158e-02
1.88168094e-01 -6.30063936e-02 2.29948893e-01 9.60098803e-01
-4.39067453e-01 8.38143885e-01 4.77039456e-01 -4.14538860e-01
-9.85844672e-01 -1.36415136e+00 -1.72984481e-01 1.11583889e+00
3.69067401e-01 -6.92568481e-01 -6.90625966e-01 -8.16439927e-01
-2.59692401e-01 1.08755779e+00 -9.47168112e-01 -4.09917347e-02
-3.52792144e-01 -4.37616974e-01 5.68543971e-01 4.76507783e-01
7.14832544e-01 -6.68312132e-01 -4.52311724e-01 -1.17185377e-01
-7.11663365e-01 -1.85109401e+00 -4.29355234e-01 -3.12096268e-01
-2.30222985e-01 -8.06660712e-01 -5.21507800e-01 -6.91825628e-01
3.97623360e-01 4.16745335e-01 1.35908842e+00 -1.37247041e-01
-4.95062858e-01 4.99357015e-01 -2.19112396e-01 -5.87155640e-01
-3.21962208e-01 -3.21571499e-01 -4.65155512e-01 1.30973831e-01
4.40899849e-01 -1.88564628e-01 -1.12788832e+00 5.38704813e-01
-6.19126499e-01 2.53235340e-01 6.64160192e-01 4.72516328e-01
6.06740236e-01 -2.50202745e-01 4.09086734e-01 -6.87607110e-01
1.18180119e-01 -2.74479479e-01 -2.40215898e-01 1.61185816e-01
-4.30807501e-01 -4.20607388e-01 2.32937828e-01 -4.10314530e-01
-9.68499005e-01 5.80103658e-02 -2.23925531e-01 -2.07377821e-01
-3.48563671e-01 -3.34799550e-02 -1.99678659e-01 2.97595322e-01
4.73446071e-01 2.32183561e-01 -3.98330957e-01 -3.18951607e-01
3.27159762e-01 7.27428079e-01 4.28565443e-01 -2.41738260e-01
3.53961498e-01 7.12679803e-01 -3.37313026e-01 -9.57730770e-01
-1.47786319e+00 -4.84295011e-01 -9.24394429e-02 -3.83807659e-01
1.31161451e+00 -1.23220050e+00 -8.94772351e-01 -2.26008311e-01
-9.83608723e-01 -3.02341163e-01 2.04415265e-02 1.87673017e-01
-4.38007265e-01 3.53676856e-01 -3.56723696e-01 -9.16449428e-01
-2.57288069e-01 -1.12330151e+00 1.63798594e+00 -8.28846768e-02
-2.34444603e-01 -5.24536371e-01 -5.09196401e-01 1.32520354e+00
2.45534703e-01 2.18940124e-01 5.90774953e-01 -8.38295341e-01
-4.81137097e-01 -3.06853741e-01 -7.72701561e-01 1.97791532e-01
-5.56250632e-01 -1.71342388e-01 -1.54371130e+00 -2.06942335e-01
1.32953659e-01 -2.68007129e-01 9.72757459e-01 3.46572429e-01
1.41329181e+00 -4.63985801e-01 -4.52539563e-01 2.28098005e-01
1.13067746e+00 1.20171653e-02 7.73877025e-01 2.24748626e-01
5.44954836e-01 5.11061549e-01 7.91331470e-01 3.61132711e-01
6.47147954e-01 1.11097825e+00 6.64701521e-01 -2.56781936e-01
-3.65076929e-01 -5.20512879e-01 5.31811953e-01 -5.94127961e-02
3.87602270e-01 -7.13463426e-01 -9.66627121e-01 3.44876707e-01
-1.58984339e+00 -8.19180727e-01 -5.53024888e-01 1.95779729e+00
3.29366446e-01 4.89948392e-01 2.41656408e-01 -1.63996577e-01
6.53149068e-01 3.23250204e-01 -3.65453035e-01 -5.72384000e-01
-1.89478993e-01 -7.43720739e-04 5.44493020e-01 1.72577098e-01
-1.31251109e+00 1.01065671e+00 5.92853260e+00 8.37105513e-01
-1.08617628e+00 1.44503623e-01 1.21493804e+00 -1.22441851e-01
-2.94175237e-01 -2.14618161e-01 -9.35964465e-01 7.19424427e-01
1.17827916e+00 3.28325152e-01 9.37766805e-02 7.45289087e-01
3.92440081e-01 -3.51378769e-01 -1.07863891e+00 1.23518550e+00
5.66615760e-01 -9.72723484e-01 2.45938361e-01 2.33051851e-01
5.17322600e-01 -2.21924573e-01 5.97374678e-01 4.02106196e-01
1.63041744e-02 -1.28060138e+00 8.59029293e-01 -3.94880511e-02
9.52438891e-01 -3.52372676e-01 5.32454312e-01 1.39597580e-01
-1.02950931e+00 7.17383996e-02 -9.71102789e-02 2.22853467e-01
3.51465732e-01 6.20236278e-01 -1.19802403e+00 1.28783718e-01
1.11537504e+00 5.52023590e-01 -1.22007298e+00 8.22404683e-01
-1.48126468e-01 1.01121938e+00 -1.84265152e-01 1.45394847e-01
1.05738871e-01 2.47487962e-01 6.85342610e-01 1.25704539e+00
7.54832998e-02 -4.01818901e-02 1.41192377e-01 8.63122165e-01
-3.76522809e-01 4.98305121e-03 -5.77044129e-01 -2.23656580e-01
3.37766744e-02 1.25780845e+00 -6.67496443e-01 -1.45787925e-01
-3.49264383e-01 1.43068147e+00 1.62895277e-01 4.22087222e-01
-1.51285565e+00 6.66726232e-02 2.83770740e-01 3.85099798e-01
7.19911814e-01 -1.64845377e-01 -2.72478640e-01 -7.65661120e-01
4.35137868e-01 -6.40910029e-01 3.38602781e-01 -1.23614597e+00
-1.15546060e+00 6.28223658e-01 -1.69377372e-01 -1.04081583e+00
1.36894688e-01 -6.95273101e-01 -2.45309770e-01 3.14771533e-01
-1.33251011e+00 -1.65562260e+00 -5.77693164e-01 4.39612627e-01
8.36111844e-01 1.32054016e-01 4.82413620e-01 3.12918842e-01
-4.18098003e-01 9.70934451e-01 -3.56034547e-01 1.28435507e-01
9.10865009e-01 -1.05094898e+00 3.77664596e-01 7.50954092e-01
4.20256644e-01 2.25796714e-01 8.86657178e-01 -6.13301098e-01
-1.34377468e+00 -1.22030866e+00 7.42060542e-01 -9.11036670e-01
6.70273006e-01 -9.99951005e-01 -5.13364673e-01 8.02658260e-01
2.32399121e-01 -1.82484329e-01 7.15608537e-01 -4.26930673e-02
-6.36723638e-01 -2.50619650e-01 -1.03236616e+00 7.48656154e-01
1.12170362e+00 -4.60230947e-01 -3.33629072e-01 4.95902240e-01
8.88900757e-01 -2.52253622e-01 -7.08897412e-01 4.45542514e-01
6.91485703e-01 -8.97363484e-01 1.24709904e+00 -4.32037145e-01
7.10857868e-01 -3.74104939e-02 -4.09725845e-01 -7.53427029e-01
1.44942477e-02 -2.03803331e-01 9.65436772e-02 1.44419563e+00
6.03324831e-01 -1.72491252e-01 7.21650302e-01 7.54764855e-01
-6.79784790e-02 -4.20936018e-01 -9.59378898e-01 -5.19965410e-01
-3.16146582e-01 -8.57411981e-01 -7.31167346e-02 6.11913085e-01
-6.10178292e-01 7.00708449e-01 -3.95687371e-01 2.39985496e-01
6.29556060e-01 -6.68369457e-02 8.21508110e-01 -9.67208505e-01
-2.37630412e-01 -1.19720086e-01 -4.30598915e-01 -1.25927353e+00
5.79162687e-02 -6.76777303e-01 -7.16405883e-02 -1.53748715e+00
5.80338061e-01 -2.01871037e-01 -6.99580461e-02 1.60413668e-01
-3.98554988e-02 7.96217978e-01 3.84149939e-01 1.11873016e-01
-1.07148516e+00 3.99623394e-01 1.18065584e+00 -5.06053627e-01
1.76779419e-01 4.23204154e-02 -1.09229994e+00 5.59185147e-01
7.77151704e-01 -1.83919057e-01 -3.20404649e-01 -5.90308070e-01
1.08692050e-01 -2.12621510e-01 8.05933118e-01 -6.77194715e-01
-2.82231510e-01 2.08061218e-01 7.84515560e-01 -6.89616144e-01
9.52653706e-01 -7.98343897e-01 -8.90923068e-02 2.86660701e-01
-6.37632072e-01 3.14974636e-02 3.67428958e-01 1.03674030e+00
9.15558729e-03 2.52362996e-01 6.54122591e-01 2.59362487e-03
-6.72851741e-01 5.80312014e-02 -2.27408484e-01 1.89130038e-01
1.05635285e+00 -3.22352588e-01 -7.95206368e-01 -7.13009775e-01
-8.60637963e-01 1.41710982e-01 1.96545124e-01 6.99945450e-01
7.28335679e-01 -8.83550167e-01 -7.78080404e-01 5.84380366e-02
4.38336432e-01 -4.92397875e-01 4.29595500e-01 8.89462888e-01
-2.17923179e-01 4.01433557e-01 9.35044363e-02 -9.24486160e-01
-1.53265345e+00 6.48623645e-01 7.76444376e-02 -1.44807007e-02
-5.16661465e-01 8.68485570e-01 7.13393450e-01 2.97411710e-01
1.59552306e-01 -1.63045123e-01 -2.13896766e-01 -1.00905515e-01
6.74783051e-01 1.01366878e-01 5.75002544e-02 -7.86416531e-01
-5.04332542e-01 3.99429679e-01 -3.65315229e-01 -4.42775548e-01
1.03555930e+00 -4.92998570e-01 4.11381245e-01 4.05018240e-01
1.44254184e+00 8.19891393e-02 -1.20187378e+00 -1.32696179e-03
-3.29107136e-01 -6.61183953e-01 2.63890363e-02 -1.21701133e+00
-8.03229690e-01 1.06900620e+00 1.11070502e+00 2.61081249e-01
7.93916047e-01 4.95909065e-01 6.05252385e-01 2.31205462e-03
-4.58660759e-02 -1.00882804e+00 4.99302208e-01 6.28202483e-02
1.10907197e+00 -1.80102849e+00 -1.04209475e-01 -7.37996876e-01
-1.01279271e+00 6.51412070e-01 7.86025822e-01 4.89374921e-02
3.68839800e-01 -1.12176895e-01 -3.01030204e-02 -2.98017442e-01
-7.50022352e-01 -3.45947087e-01 6.27136171e-01 5.23501933e-01
3.40490252e-01 9.60023403e-02 -9.09067541e-02 8.30115497e-01
-3.61167699e-01 -3.94106865e-01 2.66657203e-01 5.11997461e-01
-3.73424798e-01 -4.80080277e-01 1.08451145e-02 5.43049753e-01
-6.56060696e-01 -1.55439943e-01 -8.24935377e-01 6.26954079e-01
-6.63163289e-02 1.47044516e+00 3.86333376e-01 -5.29463232e-01
2.00749144e-01 -2.14664847e-01 4.68123615e-01 -4.22292948e-01
-3.99372488e-01 1.69188187e-01 5.43906569e-01 -6.83722377e-01
-1.53024405e-01 -4.03587639e-01 -8.72004569e-01 1.26410089e-02
-2.69583482e-02 -1.84371516e-01 8.26595426e-01 8.14408779e-01
4.61395890e-01 4.82570350e-01 5.37957311e-01 -3.09157312e-01
-1.35443047e-01 -7.79404223e-01 -2.47713313e-01 7.48036206e-01
2.53541172e-01 -5.44371784e-01 -4.07878041e-01 1.74115434e-01] | [10.846444129943848, 1.3417291641235352] |
f3ef1a8b-a560-4a28-a070-6197ffea44f8 | weakly-supervised-class-agnostic-motion | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Li_Weakly_Supervised_Class-Agnostic_Motion_Prediction_for_Autonomous_Driving_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Li_Weakly_Supervised_Class-Agnostic_Motion_Prediction_for_Autonomous_Driving_CVPR_2023_paper.pdf | Weakly Supervised Class-Agnostic Motion Prediction for Autonomous Driving | Understanding the motion behavior of dynamic environments is vital for autonomous driving, leading to increasing attention in class-agnostic motion prediction in LiDAR point clouds. Outdoor scenes can often be decomposed into mobile foregrounds and static backgrounds, which enables us to associate motion understanding with scene parsing. Based on this observation, we study a novel weakly supervised motion prediction paradigm, where fully or partially (1%, 0.1%) annotated foreground/background binary masks rather than expensive motion annotations are used for supervision. To this end, we propose a two-stage weakly supervised approach, where the segmentation model trained with the incomplete binary masks in Stage1 will facilitate the self-supervised learning of the motion prediction network in Stage2 by estimating possible moving foregrounds in advance. Furthermore, for robust self-supervised motion learning, we design a Consistency-aware Chamfer Distance loss by exploiting multi-frame information and explicitly suppressing potential outliers. Comprehensive experiments show that, with fully or partially binary masks as supervision, our weakly supervised models surpass the self-supervised models by a large margin and perform on par with some supervised ones. This further demonstrates that our approach achieves a good compromise between annotation effort and performance. | ['Guosheng Lin', 'Zhe Wang', 'Ziang Fu', 'Hanyu Shi', 'Ruibo Li'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['motion-prediction', 'scene-parsing'] | ['computer-vision', 'computer-vision'] | [ 3.84132415e-01 1.87755212e-01 -8.14396620e-01 -6.28277659e-01
-5.62493920e-01 -6.61063910e-01 6.11092091e-01 -8.09378177e-02
-6.23229444e-01 5.41840911e-01 -1.19409949e-01 -2.47243956e-01
1.46670491e-01 -5.34862220e-01 -9.90651309e-01 -8.14499795e-01
3.29992287e-02 4.85826939e-01 9.97795939e-01 1.56931967e-01
3.45498025e-02 4.07399386e-01 -1.72303200e+00 2.06897929e-01
9.87711608e-01 7.14395761e-01 4.52992797e-01 8.50083768e-01
-1.28314972e-01 1.17302895e+00 -1.31850436e-01 -3.22109550e-01
2.53600836e-01 -1.47135869e-01 -8.33188295e-01 4.00864810e-01
5.45242310e-01 -4.22536343e-01 -3.19688112e-01 1.05692852e+00
-1.48256540e-01 1.77344993e-01 3.11277568e-01 -1.36071098e+00
2.94464380e-01 5.14710069e-01 -7.78812945e-01 3.66567045e-01
-8.14812854e-02 3.84850174e-01 9.60164368e-01 -7.61731088e-01
6.07908010e-01 9.97956038e-01 5.63631237e-01 5.55827618e-01
-1.09043610e+00 -4.63216007e-01 7.17906058e-01 4.28214729e-01
-1.31891096e+00 -5.73006749e-01 9.95165110e-01 -4.71033931e-01
4.97773558e-01 3.03146362e-01 5.28240919e-01 1.01664042e+00
-7.67471641e-02 1.22400701e+00 6.62557364e-01 -3.56077217e-02
3.27357501e-01 2.08263531e-01 4.21586633e-01 8.61076832e-01
3.28300625e-01 -7.50925243e-02 -5.36238015e-01 4.48395647e-02
4.34001565e-01 -8.63247141e-02 -1.79084107e-01 -8.60837579e-01
-1.17767048e+00 4.79779929e-01 2.68691927e-01 7.03948140e-02
-1.33557543e-01 4.39134985e-01 1.53754130e-01 -2.72157609e-01
6.20527804e-01 -2.12023526e-01 -4.99416113e-01 -1.42811805e-01
-1.29618812e+00 3.27576473e-02 5.41731536e-01 1.04311526e+00
1.19679523e+00 3.66932601e-02 5.09446375e-02 4.19531405e-01
4.95737344e-01 4.53519374e-01 2.49518260e-01 -1.18627882e+00
5.78086734e-01 5.51576436e-01 2.78711379e-01 -7.09372461e-01
-3.46945584e-01 -3.82910579e-01 -6.83652461e-01 1.08803675e-01
6.28694952e-01 2.07972620e-02 -1.04316950e+00 1.64477050e+00
6.94224298e-01 7.33233631e-01 8.69066417e-02 8.79684925e-01
4.08480227e-01 5.35853803e-01 1.85020909e-01 -2.67311245e-01
1.01454759e+00 -1.25865388e+00 -5.91082752e-01 -6.52021766e-01
8.77143443e-01 -5.03673017e-01 9.97828662e-01 1.67294607e-01
-9.32607591e-01 -7.37359941e-01 -1.01031590e+00 -6.81646094e-02
3.01684123e-02 3.05435807e-01 6.60814881e-01 7.05398023e-01
-7.00098276e-01 7.36644864e-01 -1.47833598e+00 -4.97551523e-02
7.07930684e-01 3.56110841e-01 -1.98930576e-01 -2.70499736e-01
-5.73820591e-01 5.25030136e-01 4.59303200e-01 3.15425545e-01
-9.21196818e-01 -5.66405118e-01 -1.05860436e+00 -3.23870182e-01
6.06909215e-01 -6.61109686e-01 1.13837028e+00 -1.00904024e+00
-1.30457366e+00 8.43392015e-01 -7.30429173e-01 -7.65653908e-01
9.37129915e-01 -5.17114401e-01 -1.71706490e-02 2.87725717e-01
2.84979552e-01 9.10041749e-01 9.55160022e-01 -1.51702142e+00
-1.06681597e+00 -3.27308178e-01 3.05515043e-02 1.35763630e-01
-5.00966124e-02 -3.39964688e-01 -8.30316067e-01 -5.35294354e-01
2.96615332e-01 -1.15115857e+00 -5.79888344e-01 8.37922022e-02
-3.20260674e-01 -1.64099690e-02 1.23919106e+00 -4.61749583e-01
9.55499709e-01 -2.08262610e+00 2.87506692e-02 2.06151262e-01
2.37800375e-01 1.79701731e-01 -2.87441816e-03 -3.34340423e-01
2.83616990e-01 -1.73270419e-01 -7.20363677e-01 -7.31236041e-01
-2.40733281e-01 7.16703236e-01 -2.76656002e-01 5.92292905e-01
5.58084369e-01 9.41739976e-01 -1.17941034e+00 -6.65279329e-01
4.72889185e-01 1.87340453e-01 -5.62932253e-01 1.02381647e-01
-4.63743120e-01 9.08756971e-01 -5.05463123e-01 6.59678876e-01
8.18318248e-01 -2.90290207e-01 9.25048068e-02 -2.38555185e-02
-1.70825813e-02 4.89949957e-02 -1.31417418e+00 1.66662467e+00
-2.19247356e-01 7.27512002e-01 1.91826999e-01 -9.86180961e-01
5.04978418e-01 -6.59900159e-02 6.53506160e-01 -2.46361524e-01
-1.51294217e-01 1.09124251e-01 -9.74632278e-02 -5.38365662e-01
5.30573368e-01 4.46243063e-02 2.05998257e-01 4.71661575e-02
-1.97915718e-01 -2.77397186e-02 2.01828808e-01 1.31531686e-01
1.03221726e+00 6.76634967e-01 -7.31054842e-02 -1.24799088e-01
7.81821787e-01 3.23980808e-01 1.03195262e+00 7.41375208e-01
-5.44878721e-01 7.00273991e-01 2.06814468e-01 -3.26449722e-01
-7.78483331e-01 -7.96091437e-01 -4.24475446e-02 9.16819751e-01
7.27016568e-01 -3.15826803e-01 -7.88905025e-01 -9.45042610e-01
-1.93925887e-01 4.99282002e-01 -3.94236922e-01 -1.46185696e-01
-9.61925030e-01 -7.98931003e-01 3.29964757e-01 9.18969512e-01
6.10169888e-01 -6.26129866e-01 -7.63455212e-01 4.99076135e-02
-3.21101397e-01 -1.69093037e+00 -3.40523303e-01 3.88797611e-01
-1.14008486e+00 -1.11139953e+00 -5.90311587e-01 -6.94053710e-01
6.85044408e-01 6.18294120e-01 9.87824380e-01 1.34823263e-01
6.05454668e-02 3.07654232e-01 -2.71058440e-01 -4.02674936e-02
-3.44050854e-01 1.16424628e-01 4.13030665e-03 3.46995175e-01
2.56729543e-01 -4.74911839e-01 -6.63577735e-01 4.24720109e-01
-7.99670279e-01 2.50526786e-01 6.59268856e-01 4.97452289e-01
7.58290589e-01 4.12953757e-02 2.95232739e-02 -9.74542141e-01
-5.44343591e-01 -3.94406617e-01 -7.07942665e-01 2.08463948e-02
-2.63449460e-01 1.83061779e-01 4.31330949e-01 -4.78225648e-01
-1.16265881e+00 7.09855378e-01 -1.88042209e-01 -6.60352468e-01
-4.53216493e-01 -6.73474222e-02 -5.63702285e-01 -2.12398246e-01
2.92832375e-01 1.08686768e-01 -3.79449248e-01 -2.78386533e-01
5.24162471e-01 2.51418173e-01 8.09448600e-01 -4.08991754e-01
1.21973121e+00 1.04636204e+00 -3.80750895e-02 -9.43400502e-01
-8.75900805e-01 -9.18044150e-01 -1.14354002e+00 -2.92995781e-01
1.11344898e+00 -1.07423186e+00 -2.94643819e-01 3.50366205e-01
-9.67877924e-01 -6.83007956e-01 -3.95593822e-01 5.65681279e-01
-5.40112436e-01 6.03897929e-01 -4.67117757e-01 -1.03585851e+00
3.40714455e-01 -1.14970803e+00 1.32502627e+00 5.10014892e-02
-1.57831833e-01 -9.36292589e-01 -2.39337936e-01 7.60644794e-01
-1.77314535e-01 2.23118111e-01 4.50852573e-01 -4.76308346e-01
-1.10580075e+00 -9.78533328e-02 -1.01451516e-01 2.75084496e-01
-2.97553092e-02 2.05702130e-02 -1.29270327e+00 4.89291828e-03
-8.84998590e-03 -5.80031388e-02 1.28483891e+00 4.77466732e-01
1.11727941e+00 -4.23122905e-02 -4.17601675e-01 8.35947394e-01
1.00176191e+00 -7.82944262e-02 3.45727533e-01 3.75116348e-01
1.30971777e+00 8.88414621e-01 1.02323174e+00 3.25891286e-01
6.24926388e-01 4.50198531e-01 6.28663719e-01 -4.28583547e-02
-5.50731309e-02 -3.08192074e-01 5.05287766e-01 4.77068871e-01
5.18321544e-02 -9.76979211e-02 -9.84980166e-01 6.26503587e-01
-2.18993878e+00 -8.17471802e-01 -6.13968968e-01 2.19570303e+00
5.49966574e-01 5.79699278e-01 1.47626460e-01 3.04580301e-01
6.14521384e-01 5.51266372e-01 -7.57115543e-01 3.67031246e-01
-1.79524392e-01 -2.00807154e-01 8.13272476e-01 7.38101840e-01
-1.57200754e+00 1.35180402e+00 4.99186563e+00 7.37059057e-01
-8.92541945e-01 1.28303573e-01 7.94165134e-01 -5.47366999e-02
-3.21811855e-01 2.81808615e-01 -9.55693603e-01 4.82939839e-01
7.47762680e-01 2.66581953e-01 -6.91180378e-02 9.43358123e-01
5.39933264e-01 -2.86961854e-01 -1.17772830e+00 7.41780341e-01
-2.20013112e-01 -1.22645903e+00 -3.23796608e-02 3.72894332e-02
8.68310153e-01 1.18587814e-01 -1.72329232e-01 1.42564997e-01
2.79859602e-01 -6.13238573e-01 9.21314597e-01 2.79110432e-01
1.73540592e-01 -6.66237354e-01 6.20003462e-01 8.40965748e-01
-1.60692656e+00 -4.98830341e-02 -1.96102798e-01 -4.71157357e-02
4.29909974e-01 7.54258573e-01 -4.71701831e-01 5.93569040e-01
6.76226616e-01 9.97205794e-01 -5.64005911e-01 9.17055309e-01
-2.45784357e-01 8.50508094e-01 -4.58429486e-01 4.26766515e-01
3.98087651e-01 -2.79440075e-01 5.96015573e-01 1.18313408e+00
-7.16809332e-02 1.22469649e-01 4.73843664e-01 6.68532848e-01
3.74604166e-02 -1.71277583e-01 -4.44743156e-01 2.98209667e-01
1.46599859e-01 1.08806884e+00 -1.12957025e+00 -4.83900756e-01
-3.91564101e-01 1.14289117e+00 9.28555280e-02 4.61826563e-01
-1.00225663e+00 3.27681363e-01 6.80033445e-01 1.80437356e-01
3.95293623e-01 -5.13241291e-01 -3.91422719e-01 -1.28267324e+00
3.34402859e-01 -2.74993360e-01 2.28219941e-01 -4.42368031e-01
-7.54758179e-01 2.02120304e-01 -2.27165222e-03 -1.37316906e+00
-1.22404180e-01 -4.93350923e-01 -6.98796809e-01 2.00204819e-01
-2.00321078e+00 -1.21107769e+00 -5.73616505e-01 4.16462392e-01
8.69619489e-01 5.01337588e-01 2.98105069e-02 3.42745781e-01
-8.60663950e-01 2.39938766e-01 -2.12773710e-01 9.62118730e-02
5.20167291e-01 -1.27191710e+00 4.30098683e-01 1.36010945e+00
3.64848465e-01 1.64634153e-01 6.23983860e-01 -7.42079079e-01
-1.21859705e+00 -1.59734154e+00 6.05794132e-01 -6.90915823e-01
6.14969254e-01 -2.79157490e-01 -1.19290125e+00 6.54328167e-01
-2.71355093e-01 5.23645997e-01 3.68706167e-01 -3.80911589e-01
8.22754204e-02 -9.93021876e-02 -8.57021689e-01 6.76559210e-01
1.36277103e+00 -3.01127315e-01 -3.49790603e-01 4.01520520e-01
8.56878281e-01 -4.79917258e-01 -3.69326055e-01 8.30666602e-01
2.67569780e-01 -8.06147993e-01 1.08075333e+00 -4.13528442e-01
2.19437659e-01 -6.08701408e-01 -8.30348432e-02 -5.26091814e-01
1.37838751e-01 -6.57965541e-01 -5.56715310e-01 1.17983067e+00
2.67186880e-01 -9.64317247e-02 1.55383718e+00 7.19974577e-01
-2.13354796e-01 -4.48801786e-01 -8.92466843e-01 -7.55440116e-01
-1.28358364e-01 -9.67253327e-01 8.07552692e-03 8.68423641e-01
-6.04637742e-01 4.32140045e-02 -4.18368757e-01 7.02830374e-01
7.69114494e-01 1.13088146e-01 1.04540765e+00 -1.16425276e+00
-3.20762157e-01 -2.68651307e-01 -4.85708773e-01 -1.73474073e+00
4.59507495e-01 -6.81758642e-01 5.48618078e-01 -1.24376619e+00
9.88468230e-02 -5.00817239e-01 -5.39588779e-02 3.45184237e-01
-6.05346680e-01 3.94636393e-01 -2.07357574e-02 4.68050539e-01
-8.85968328e-01 6.45541906e-01 9.48727012e-01 -2.06807464e-01
-4.50351238e-01 4.62522477e-01 -1.82979465e-01 1.35492897e+00
7.11569607e-01 -5.24806976e-01 -5.47327816e-01 -4.18874413e-01
-1.93679526e-01 -1.81054279e-01 6.15887225e-01 -1.13366473e+00
3.89062732e-01 -3.33060175e-01 2.33801603e-01 -8.81128490e-01
3.36996287e-01 -9.03924882e-01 -2.57292986e-01 4.15120840e-01
-1.68194130e-01 -2.17632174e-01 1.42730162e-01 1.04714775e+00
-2.26385355e-01 -2.49516621e-01 7.16211379e-01 5.81215918e-02
-1.25642931e+00 4.83099282e-01 -4.39688385e-01 5.43081351e-02
1.10392928e+00 -5.77407360e-01 6.10474423e-02 -2.11437538e-01
-8.07030380e-01 5.60991526e-01 5.03249645e-01 3.85030687e-01
5.29823124e-01 -9.96070147e-01 -4.25377548e-01 1.72280163e-01
2.40792945e-01 6.52017057e-01 1.98894322e-01 1.12586272e+00
-5.48394561e-01 1.50551513e-01 3.20803732e-01 -1.31903076e+00
-1.34311688e+00 3.48008901e-01 1.28218815e-01 -1.71744034e-01
-7.66398072e-01 9.38285887e-01 4.70162630e-01 -2.40324765e-01
4.03287679e-01 -7.33942032e-01 -9.11400393e-02 -2.64500380e-01
3.73701125e-01 3.83150846e-01 -1.39239773e-01 -8.44666660e-01
-4.87996876e-01 7.15996683e-01 9.33329239e-02 -1.20974347e-01
9.93750334e-01 -3.51363420e-01 3.21592301e-01 6.43629134e-01
1.00433075e+00 1.86864540e-01 -1.95991409e+00 -2.76252329e-01
5.56389570e-01 -4.68986571e-01 -1.65366773e-02 -2.26182453e-02
-1.08976710e+00 9.32353139e-01 4.30755943e-01 -2.43808433e-01
9.19359922e-01 1.59177050e-01 8.42134118e-01 5.23930490e-01
3.40402484e-01 -1.02402449e+00 1.52837232e-01 4.60021853e-01
9.02998745e-02 -1.57841241e+00 -1.85073495e-01 -7.87649274e-01
-7.82364368e-01 8.02945673e-01 7.27320075e-01 -4.81627174e-02
4.49264884e-01 3.92977804e-01 1.28878072e-01 1.54205367e-01
-5.62521100e-01 -6.35863721e-01 3.70973766e-01 5.32995105e-01
5.92591912e-02 -2.57974565e-01 1.60324365e-01 3.86426061e-01
-1.98066290e-02 -2.54338443e-01 2.63251722e-01 1.14300013e+00
-7.37391055e-01 -1.02224660e+00 -3.44078958e-01 1.00043483e-01
-2.07680508e-01 1.43917039e-01 -4.05392945e-01 7.18056858e-01
3.46654147e-01 1.01496768e+00 6.89419955e-02 -3.17181289e-01
1.65942132e-01 6.77883625e-03 2.26061657e-01 -5.52314043e-01
-1.09324463e-01 1.92324951e-01 2.49182750e-02 -9.02152419e-01
-8.28519821e-01 -7.73630977e-01 -1.49850464e+00 3.88667062e-02
-3.43150795e-01 -1.68148882e-03 3.45029563e-01 1.22066116e+00
8.98064151e-02 5.40787876e-01 5.69674134e-01 -1.23388898e+00
-2.51292378e-01 -4.83749747e-01 -2.29891002e-01 3.43850255e-01
7.20043123e-01 -5.47292888e-01 -4.62237716e-01 4.77148652e-01] | [9.101534843444824, -0.14933444559574127] |
08effa78-015f-4cf8-a4db-7111b6fe41d1 | improving-attacks-on-round-reduced-speck32-64 | null | null | https://eprint.iacr.org/2019/037.pdf | https://eprint.iacr.org/2019/037.pdf | Improving Attacks on Round-Reduced Speck32/64 Using Deep Learning | This paper has four main contributions.1 First, we calculate the predicted difference distribution of Speck32/64 with one specific input difference under the Markov assumption completely for up to eight rounds and verify that this yields a globally fairly good model of the difference distribution of Speck32/64. Secondly, we show that contrary to conventional wisdom, machine learning can produce very powerful cryptographic distinguishers: for instance, in a simple low-data, chosen plaintext attack on nine rounds of Speck, we present distinguishers based on deep residual neural networks that achieve a mean key rank roughly five times lower than an analogous classical distinguisher using the full difference distribution table. Thirdly, we develop a highly selective key
search policy based on a variant of Bayesian optimization which, together with our neural distinguishers, can be used to reduce the remaining security of 11-round Speck32/64 to roughly 38 bits. This is a significant improvement over previous literature. Lastly, we show that our neural distinguishers successfully use features of the ciphertext pair distribution that are invisible to all purely differential distinguishers even given unlimited data.
While our attack is based on a known input difference taken from the literature, we also show that neural networks can be used to rapidly (within a matter of minutes on our machine) find good input differences without using prior human cryptanalysis. | ['Aron Gohr'] | 2019-08-01 | null | null | null | conference-2019-8 | ['cryptanalysis'] | ['miscellaneous'] | [ 3.89768720e-01 -3.45370509e-02 3.79840173e-02 -8.55112746e-02
-1.49826229e+00 -1.22984231e+00 6.26932800e-01 6.64734468e-02
-6.11064017e-01 6.95377469e-01 -1.38326153e-01 -1.26445818e+00
-1.01489268e-01 -6.97367072e-01 -8.97313714e-01 -9.78865981e-01
-4.17316139e-01 5.19079208e-01 9.77749676e-02 -5.25990546e-01
6.82585418e-01 5.73525310e-01 -9.53441381e-01 4.87195492e-01
3.98326278e-01 1.10453093e+00 -6.11483634e-01 1.37215590e+00
4.81363237e-01 5.60702324e-01 -6.86674237e-01 -6.03711426e-01
1.07585096e+00 -1.59660384e-01 -1.22734928e+00 -1.04717720e+00
5.18306971e-01 -9.77971077e-01 -7.16522336e-01 1.25404191e+00
5.63213587e-01 -3.83981466e-01 4.56182241e-01 -1.05978036e+00
-3.42085282e-03 1.31123066e+00 1.22973874e-01 1.72719330e-01
1.79449379e-01 4.64842796e-01 1.27022457e+00 -9.86973047e-02
2.09539026e-01 1.09665275e+00 8.36120963e-01 2.68694371e-01
-1.36445212e+00 -1.11987662e+00 -5.26255310e-01 2.35051259e-01
-1.25928569e+00 -3.44691187e-01 4.75390375e-01 1.00021742e-01
1.00891638e+00 5.65493524e-01 4.31608081e-01 9.29092467e-01
7.37394035e-01 6.01815939e-01 1.39758706e+00 -3.99722189e-01
-5.94160147e-02 -3.98092270e-02 3.76722932e-01 2.78376609e-01
5.17826319e-01 7.76796222e-01 -6.16231710e-02 -8.00775409e-01
1.80828258e-01 -8.97822902e-02 -4.29321468e-01 -1.02112353e-01
-1.35967064e+00 9.57670867e-01 7.56284148e-02 1.30043820e-01
9.84159261e-02 7.36312568e-01 5.71378350e-01 8.62349272e-01
-1.44055560e-01 7.67887414e-01 -9.32605863e-01 -3.01468581e-01
-9.73691821e-01 3.36157233e-01 1.56391490e+00 4.78236198e-01
8.78837526e-01 2.55018305e-02 3.52846682e-01 -4.39635724e-01
1.12350248e-01 8.19300473e-01 1.67102367e-01 -9.00457025e-01
5.00710726e-01 -2.93788284e-01 -2.62193810e-02 -8.31791520e-01
-3.73141855e-01 -2.12662756e-01 -7.70404816e-01 3.80946308e-01
8.75550807e-01 -6.64555430e-01 -4.96082872e-01 1.51164234e+00
-1.05903216e-01 -9.96964797e-02 5.38104296e-01 2.83969849e-01
2.56775647e-01 6.60295904e-01 -7.38424659e-01 2.40274593e-01
1.25037074e+00 -2.47052565e-01 -6.41088784e-02 1.51235983e-01
1.11126101e+00 -9.11623001e-01 1.66186959e-01 8.76332998e-01
-9.74674821e-01 -1.41519696e-01 -1.49575973e+00 4.95846905e-02
-2.87963539e-01 -5.06032467e-01 9.31862772e-01 1.38458550e+00
-1.37287319e+00 9.38282371e-01 -6.01209700e-01 6.16273582e-01
1.14448935e-01 1.07136536e+00 -2.69102097e-01 2.52599627e-01
-1.60332501e+00 6.45508468e-01 8.34172785e-01 9.36071351e-02
-1.05873430e+00 -6.65665209e-01 -6.39090657e-01 8.53123888e-02
6.20330215e-01 -2.81371593e-01 1.52877223e+00 -8.33244383e-01
-1.83988118e+00 4.51239258e-01 2.64353693e-01 -1.04910946e+00
3.21039528e-01 -3.13306414e-02 -1.37786016e-01 9.21679139e-02
-6.57789290e-01 3.28286111e-01 6.10679269e-01 -9.75849569e-01
-6.94888175e-01 -9.11629796e-02 5.39503992e-01 -2.73890853e-01
2.23495334e-01 -4.11319435e-02 3.21201921e-01 -1.46794692e-01
-3.87363462e-03 -1.26610327e+00 -6.09834671e-01 -6.72122002e-01
-7.64145613e-01 6.54718056e-02 3.62851351e-01 -6.37762487e-01
6.88365340e-01 -1.83311498e+00 -3.41036350e-01 9.90021408e-01
3.14055324e-01 4.98538792e-01 -3.52657773e-02 6.38054550e-01
-2.27462322e-01 1.73292965e-01 -1.24528311e-01 2.50136882e-01
2.59237289e-01 -3.09626937e-01 -7.92159617e-01 9.61175203e-01
-2.41933495e-01 7.53602862e-01 -7.33860731e-01 2.10947916e-01
-3.41778547e-01 2.98742145e-01 -7.21344769e-01 -8.62150341e-02
-2.42594242e-01 7.04775006e-02 -1.94873512e-01 1.90203533e-01
1.18303490e+00 -1.35953605e-01 7.05511272e-01 -1.37637734e-01
-2.49067135e-02 9.26969290e-01 -1.52126050e+00 1.35631442e+00
-4.14194703e-01 7.98901975e-01 -2.67842878e-03 -8.92253637e-01
5.29909730e-01 1.45318300e-01 1.71937063e-01 -3.77314448e-01
6.05550349e-01 5.56514025e-01 5.79057693e-01 3.99262607e-01
9.22209024e-01 -3.89098153e-02 -3.81190389e-01 1.08430624e+00
-3.06015283e-01 -3.46720427e-01 -3.52890909e-01 4.06581640e-01
1.17371058e+00 -3.65341216e-01 1.34536222e-01 -4.31277782e-01
6.48327649e-01 -3.05062592e-01 3.94904882e-01 1.38209999e+00
2.30233651e-02 1.87478364e-01 1.07640624e+00 -5.05270720e-01
-8.07911694e-01 -9.28048849e-01 -8.01582187e-02 8.48429859e-01
2.27458313e-01 -6.20844603e-01 -5.45642376e-01 -8.00199926e-01
1.33006237e-02 5.56652427e-01 -3.39792103e-01 -1.32812813e-01
-7.46586502e-01 -8.66070628e-01 1.46367383e+00 3.02673846e-01
5.50794303e-01 -5.31126261e-01 -2.44694173e-01 7.43778050e-02
2.96432257e-01 -9.22651231e-01 -3.86007607e-01 9.00490105e-01
-5.86930096e-01 -1.41611540e+00 -2.68026114e-01 -6.66116178e-01
2.30004817e-01 3.65508571e-02 9.40779626e-01 1.80902138e-01
-1.62986126e-02 3.42060417e-01 8.51255804e-02 -1.86745048e-01
-1.21255398e+00 1.60612032e-01 3.32446955e-02 -4.20088291e-01
3.14983457e-01 -4.26922590e-01 -6.97456241e-01 -1.86065659e-02
-1.13073123e+00 -4.35814261e-01 7.50563622e-01 9.51671183e-01
3.53563391e-02 3.46452564e-01 -1.82302341e-01 -9.80004728e-01
4.33257371e-01 -2.83443421e-01 -1.29859746e+00 -5.80589399e-02
-7.91405678e-01 7.72830904e-01 1.04235005e+00 -5.19062459e-01
-6.20782554e-01 -1.58954412e-01 -3.60752106e-01 3.87093991e-01
1.83679745e-01 -7.49573335e-02 -6.17379546e-02 -7.26071239e-01
7.64223516e-01 3.35886091e-01 1.88989028e-01 -1.55336797e-01
4.81554180e-01 8.48290145e-01 5.64809382e-01 -9.01974678e-01
1.14810789e+00 4.84127730e-01 3.96192998e-01 -1.86710507e-01
-2.73161411e-01 -5.26384637e-02 -3.15044522e-01 6.81906104e-01
1.66189894e-01 -8.12093198e-01 -1.79936147e+00 8.56896460e-01
-1.02047968e+00 -6.75528169e-01 7.90934339e-02 6.26753271e-01
-3.92396688e-01 1.11555886e+00 -1.05083227e+00 -7.07425952e-01
-6.12920582e-01 -1.49620152e+00 7.25411475e-01 -4.71151844e-02
1.07670881e-01 -1.18025815e+00 -2.55821995e-03 -3.92882377e-02
3.48140210e-01 1.34087410e-02 9.92334306e-01 -1.37143254e+00
-1.10116661e+00 -3.34171057e-01 -1.69207722e-01 7.42161512e-01
-9.04383287e-02 -1.62076086e-01 -8.68860006e-01 -6.57932878e-01
1.22921795e-01 -4.77516472e-01 1.07710683e+00 1.26838069e-02
1.09678066e+00 -6.31359458e-01 -5.75781241e-02 1.05815279e+00
1.38178742e+00 2.61892289e-01 8.25977743e-01 5.07506847e-01
4.70666200e-01 1.38760969e-01 7.57901669e-02 4.71831977e-01
2.74673194e-01 2.01679707e-01 4.06599730e-01 2.34735996e-01
4.54341888e-01 1.32110044e-01 7.21626341e-01 4.82758313e-01
1.26881063e-01 -6.36606589e-02 -6.96813405e-01 3.66499871e-02
-9.65545535e-01 -8.63447309e-01 -1.17137842e-01 2.66062570e+00
1.22668183e+00 5.80543518e-01 -1.20881461e-01 3.65275383e-01
2.85667121e-01 3.09164852e-01 -1.93731666e-01 -9.15964782e-01
-2.16960415e-01 9.35066283e-01 1.65026760e+00 8.61734033e-01
-1.07038486e+00 7.96937048e-01 7.05208302e+00 1.12287235e+00
-1.22501898e+00 -3.94807637e-01 7.09381223e-01 2.49031156e-01
-7.48910248e-01 6.19489908e-01 -1.00291967e+00 4.57935393e-01
1.60208881e+00 -4.66602929e-02 8.05145860e-01 6.08534455e-01
-5.68745732e-01 -1.49104625e-01 -1.49999583e+00 1.00922489e+00
-7.62442276e-02 -1.46245420e+00 -1.86432749e-01 4.96851534e-01
7.33160794e-01 -6.57077804e-02 4.70638901e-01 3.98974687e-01
9.52882648e-01 -1.15698671e+00 3.00899953e-01 7.90896863e-02
8.34168613e-01 -1.03549075e+00 9.38352108e-01 1.05811901e-01
-6.20753646e-01 -8.35106596e-02 -3.04864824e-01 8.56708661e-02
-2.93787003e-01 2.87708879e-01 -1.13384461e+00 5.66556275e-01
2.13147506e-01 -1.45465359e-01 -1.93075910e-01 7.70507216e-01
-6.36657774e-02 7.21701860e-01 -6.64865732e-01 -2.72206604e-01
6.55744493e-01 1.45064577e-01 5.41139901e-01 1.22852027e+00
1.72810629e-01 -1.26705989e-01 -1.22982778e-01 3.53716254e-01
-9.80252549e-02 -3.06626260e-01 -3.65490496e-01 6.33736625e-02
3.59916329e-01 8.34659994e-01 -3.78121525e-01 -3.80486637e-01
5.64907417e-02 8.27166498e-01 -2.43891403e-01 1.07581057e-01
-5.06268322e-01 -7.99559951e-01 7.37546086e-01 -4.36500579e-01
6.04739785e-01 -4.10435855e-01 -1.81137383e-01 -1.02957511e+00
-2.85246938e-01 -1.78552198e+00 4.21021640e-01 2.00788528e-01
-9.29392755e-01 3.74686420e-01 -4.95912209e-02 -8.90991807e-01
-7.29225874e-01 -1.06973851e+00 -4.58643705e-01 1.04897201e+00
-1.60152936e+00 -5.29317856e-01 5.48124671e-01 5.02935588e-01
-2.68535465e-01 -3.37196529e-01 8.81656826e-01 -1.61955968e-01
6.68630600e-02 1.28535521e+00 7.84574866e-01 3.35931420e-01
6.83795035e-01 -1.46908975e+00 1.07048678e+00 7.51155019e-01
1.88115492e-01 8.83049071e-01 7.36031234e-01 -4.90604997e-01
-2.12239528e+00 -2.49114141e-01 3.59379411e-01 -3.41547281e-01
9.39500093e-01 -2.83284366e-01 -5.04014254e-01 7.34418213e-01
1.32111624e-01 -2.02171624e-01 6.68911517e-01 -7.02011809e-02
-9.38472867e-01 -2.28391767e-01 -1.14986086e+00 5.12450874e-01
4.11631376e-01 -8.25069666e-01 -4.46448833e-01 2.89163142e-01
4.67523217e-01 -8.16477060e-01 -7.01404572e-01 2.03461736e-01
8.35128009e-01 -1.18345106e+00 1.00735736e+00 -7.94781983e-01
-2.24732175e-01 -5.57263754e-02 -3.76531452e-01 -9.61222053e-01
9.27694067e-02 -1.80281043e+00 -2.14496478e-01 5.14527738e-01
3.71197075e-01 -1.12430608e+00 9.02053773e-01 5.31632543e-01
2.75274903e-01 -5.54571569e-01 -1.03319883e+00 -7.76953161e-01
8.52212667e-01 -7.29022086e-01 8.52733672e-01 3.85850251e-01
-3.43356788e-01 -2.59476781e-01 -5.02819479e-01 6.06647372e-01
6.64192915e-01 2.25286588e-01 1.07941771e+00 -1.01261854e+00
-8.94474983e-01 -5.32324255e-01 -5.48352778e-01 -1.57879865e+00
3.19893926e-01 -1.07033837e+00 5.08180857e-02 -4.60286647e-01
1.66891247e-01 -7.17701852e-01 -3.64378870e-01 2.15353847e-01
1.11445643e-01 2.67368525e-01 -5.59618557e-03 -1.11697599e-01
-2.98099399e-01 -4.11967002e-02 7.38542914e-01 -3.14497769e-01
1.55235715e-02 3.82913738e-01 -1.14949346e+00 5.69929957e-01
8.34454119e-01 -5.71304619e-01 1.95162430e-01 4.76631746e-02
6.06045842e-01 9.51396823e-02 2.68132627e-01 -8.43568325e-01
1.83185339e-01 7.32790902e-02 2.33733878e-01 -5.03870606e-01
2.07732186e-01 -5.21803737e-01 -1.30007803e-01 1.04201925e+00
-2.98502386e-01 -4.18982431e-02 1.84754625e-01 4.01378930e-01
3.15140218e-01 -4.48413968e-01 6.24523699e-01 1.49961174e-01
-3.30824614e-01 2.33883634e-01 -5.25834203e-01 3.48789245e-01
4.79024440e-01 8.11696798e-02 -5.30600309e-01 -7.71848798e-01
9.14798770e-03 -1.27694175e-01 3.75415057e-01 -3.14496636e-01
2.27159962e-01 -8.79103303e-01 -6.90561354e-01 2.60668993e-01
-5.13441265e-01 -1.80209592e-01 1.15130469e-01 6.33142412e-01
-1.01028645e+00 8.23494375e-01 1.31835088e-01 -3.20887864e-01
-1.38034022e+00 4.04306263e-01 4.74401027e-01 -6.96248055e-01
-5.45268714e-01 8.42496455e-01 -4.18819860e-02 -2.37089738e-01
1.87516198e-01 -6.03762388e-01 5.80620646e-01 -3.41081172e-01
8.56752694e-01 2.70761400e-01 1.06351860e-01 -3.47863913e-01
-3.30328286e-01 3.31114918e-01 -4.29252148e-01 -1.73832268e-01
8.93334389e-01 4.06298608e-01 -1.37433812e-01 -1.50933638e-01
1.74416268e+00 5.63530326e-01 -1.16888857e+00 -2.64510691e-01
-3.86318229e-02 -2.90168792e-01 -1.68360174e-01 -4.71996248e-01
-8.09102237e-01 7.46418893e-01 5.34769714e-01 4.01636422e-01
7.64790595e-01 -3.65052968e-01 1.43672216e+00 1.42145491e+00
5.11272788e-01 -7.41049051e-01 -3.49361300e-01 9.11291838e-01
-1.05187241e-02 -1.10775471e+00 1.22834936e-01 3.24805602e-02
-1.24553412e-01 1.58849490e+00 -3.38555664e-01 -3.06553990e-01
6.88988090e-01 6.94207847e-01 9.51229781e-02 2.65432864e-01
-6.38934255e-01 2.33308926e-01 2.35798433e-01 4.15330827e-01
-1.16808675e-01 -3.63203436e-02 -3.76568027e-02 4.32718992e-01
-7.40563393e-01 -5.19712985e-01 9.75063145e-01 7.40829408e-01
-1.79813668e-01 -1.62695873e+00 -6.15309596e-01 2.71448582e-01
-1.31466281e+00 -7.33722687e-01 -1.47605360e-01 7.50056624e-01
-3.56033385e-01 8.38652432e-01 -2.05455706e-01 -7.00364947e-01
-6.05199218e-01 -1.17916912e-01 5.80806971e-01 -4.85976413e-02
-8.90037656e-01 -5.50743043e-01 9.71069038e-02 -6.11836553e-01
2.00489715e-01 -3.80190879e-01 -9.81670499e-01 -9.72166181e-01
-4.97599781e-01 3.67730439e-01 8.09267819e-01 1.00487638e+00
5.97876012e-02 -1.16851099e-01 1.12642837e+00 -9.08159316e-01
-1.52596748e+00 -4.46506590e-01 -6.66771650e-01 -2.17381865e-01
8.08180451e-01 1.98896363e-01 -9.83543754e-01 -4.77596343e-01] | [5.935407638549805, 7.341752529144287] |
bf0f0c28-d7f4-413a-9dc1-01777d881d5a | decoding-dynamic-brain-patterns-from-evoked | 1606.02840 | null | http://arxiv.org/abs/1606.02840v2 | http://arxiv.org/pdf/1606.02840v2.pdf | Decoding dynamic brain patterns from evoked responses: A tutorial on multivariate pattern analysis applied to time-series neuroimaging data | Multivariate pattern analysis (MVPA) or brain decoding methods have become
standard practice in analysing fMRI data. Although decoding methods have been
extensively applied in Brain Computing Interfaces (BCI), these methods have
only recently been applied to time-series neuroimaging data such as MEG and EEG
to address experimental questions in Cognitive Neuroscience. In a
tutorial-style review, we describe a broad set of options to inform future
time-series decoding studies from a Cognitive Neuroscience perspective. Using
example MEG data, we illustrate the effects that different options in the
decoding analysis pipeline can have on experimental results where the aim is to
'decode' different perceptual stimuli or cognitive states over time from
dynamic brain activation patterns. We show that decisions made at both
preprocessing (e.g., dimensionality reduction, subsampling, trial averaging)
and decoding (e.g., classifier selection, cross-validation design) stages of
the analysis can significantly affect the results. In addition to standard
decoding, we describe extensions to MVPA for time-varying neuroimaging data
including representational similarity analysis, temporal generalisation, and
the interpretation of classifier weight maps. Finally, we outline important
caveats in the design and interpretation of time-series decoding experiments. | [] | 2016-09-30 | null | null | null | null | ['brain-decoding', 'brain-decoding'] | ['medical', 'miscellaneous'] | [ 9.00085568e-01 -5.46696484e-01 6.71059549e-01 -7.92214215e-01
-3.27005118e-01 -4.85778302e-01 8.30483615e-01 1.49517104e-01
-7.90542245e-01 5.67306638e-01 2.56055355e-01 -6.78144217e-01
-5.58087707e-01 -2.71504045e-01 -2.57246554e-01 -6.21961474e-01
-6.49933279e-01 2.78912950e-02 1.45967007e-01 6.03701919e-02
7.70785689e-01 6.40769303e-01 -1.82116449e+00 4.83973831e-01
6.91027582e-01 1.01761591e+00 2.80056775e-01 4.54489201e-01
2.64163166e-01 1.52112037e-01 -6.64036572e-01 -9.14256871e-02
-9.56964195e-02 -5.91856360e-01 -4.45138097e-01 -2.70101994e-01
-1.59583688e-02 1.03342295e-01 -1.85027465e-01 1.00410652e+00
9.11543012e-01 5.15908301e-01 7.13244379e-01 -1.03456008e+00
-1.68029368e-01 3.99734110e-01 -3.37843329e-01 1.10250068e+00
2.81712651e-01 2.68926114e-01 2.76280791e-01 -7.67263114e-01
4.04932320e-01 1.17574584e+00 4.99616235e-01 2.32827947e-01
-1.52645183e+00 -8.31425607e-01 2.79656768e-01 6.83210075e-01
-1.22967744e+00 -1.00966585e+00 5.45344949e-01 -7.20855772e-01
1.29716814e+00 4.52405602e-01 8.63479078e-01 1.33418477e+00
8.59252572e-01 2.66703904e-01 1.66913950e+00 -3.06539446e-01
5.09434581e-01 -4.22219187e-01 4.90163982e-01 -1.45566791e-01
8.86815116e-02 3.51611286e-01 -5.34972787e-01 -2.21053541e-01
5.02471626e-01 -5.48393488e-01 -2.67453581e-01 4.60528843e-02
-1.48067713e+00 7.15051234e-01 1.39028296e-01 6.32293880e-01
-7.20415235e-01 -6.07678704e-02 6.31287158e-01 3.12916666e-01
4.81710106e-01 3.98253560e-01 -4.15387571e-01 -4.38443273e-01
-1.08290374e+00 2.78974354e-01 2.81117409e-01 3.79326642e-01
1.36732772e-01 4.49849218e-02 -5.30949354e-01 9.78528440e-01
1.66169390e-01 2.06953615e-01 8.48111451e-01 -6.90724969e-01
2.43231714e-01 -2.72925764e-01 -1.70076460e-01 -9.67928469e-01
-9.16665852e-01 -4.07416858e-02 -4.49931085e-01 3.57238278e-02
4.12759513e-01 -2.25832984e-01 -8.58173132e-01 1.76251650e+00
-2.24536642e-01 1.06161989e-01 -3.63465041e-01 9.23311591e-01
2.54017919e-01 1.13445424e-01 5.50591350e-01 -7.35247910e-01
1.61833072e+00 -3.14467661e-02 -9.70566988e-01 -4.78510231e-01
5.62117994e-01 -4.22216237e-01 5.88654697e-01 5.36769032e-01
-1.15835822e+00 -3.77896190e-01 -9.69714105e-01 1.21550098e-01
-4.21509296e-01 -2.46195570e-01 9.01678622e-01 6.87610984e-01
-1.19658279e+00 7.44004190e-01 -1.28257561e+00 -5.52605689e-01
6.04369581e-01 5.35717845e-01 -4.25948918e-01 2.53924340e-01
-1.18105447e+00 1.41418862e+00 3.24799925e-01 3.61298054e-01
-6.02364838e-01 -4.69342917e-01 -5.60207486e-01 -1.45353705e-01
-2.98442245e-01 -4.77126569e-01 1.11120045e+00 -9.88653064e-01
-1.21836579e+00 9.18545067e-01 -5.64538717e-01 -2.26208925e-01
-1.16590053e-01 4.20583606e-01 -8.53221118e-01 -1.05263226e-01
5.98293953e-02 6.28970146e-01 6.87158227e-01 -2.36943439e-01
-3.23503941e-01 -8.85589898e-01 -4.97590661e-01 2.38628700e-01
2.07613677e-01 5.86580396e-01 6.03732504e-02 -8.15898657e-01
3.73044878e-01 -6.92330539e-01 -2.03844514e-02 -3.23657334e-01
3.04321647e-01 -2.12614954e-01 2.98272312e-01 -8.33681643e-01
1.14666796e+00 -2.37067103e+00 2.87985057e-01 3.56700450e-01
-9.82669834e-03 -2.59226769e-01 -2.34896854e-01 1.41669065e-01
-8.14875901e-01 5.02433665e-02 -5.14862716e-01 7.04993829e-02
-8.42716396e-02 -1.49228886e-01 1.64440349e-02 8.57116342e-01
2.21916258e-01 9.53903317e-01 -7.33291388e-01 1.52820289e-01
8.29776227e-02 4.59527075e-01 -4.57437843e-01 -2.20400244e-01
3.45252603e-01 8.23911667e-01 2.13169411e-01 5.04537284e-01
5.49203753e-01 1.80512771e-01 2.88092494e-01 -3.97546887e-01
-5.13080895e-01 8.20778012e-01 -6.19731486e-01 1.76778281e+00
-1.13641419e-01 1.29489100e+00 -1.49733081e-01 -1.46863294e+00
4.87361103e-01 4.14350808e-01 2.86614925e-01 -1.28339636e+00
4.91038263e-01 1.77309051e-01 1.11566377e+00 -5.54734826e-01
-5.87159991e-02 -3.03566605e-01 1.92578986e-01 6.67585135e-01
2.02496022e-01 2.15222821e-01 2.43495375e-01 -4.00409192e-01
1.20318115e+00 -5.28280027e-02 2.16579944e-01 -6.96600795e-01
1.86814472e-01 -2.56217331e-01 3.37477863e-01 5.28959215e-01
-4.82075304e-01 3.97532135e-01 5.73797882e-01 -8.22501108e-02
-7.01750994e-01 -9.19717193e-01 -6.34716928e-01 1.05166066e+00
-3.97052795e-01 -3.04264933e-01 -5.55164099e-01 1.07610688e-01
-3.72999579e-01 1.06196225e+00 -9.02272582e-01 -5.58579147e-01
-5.96253216e-01 -1.36871707e+00 3.02780807e-01 5.66499591e-01
2.27661561e-02 -1.20427692e+00 -1.28171968e+00 4.66724962e-01
-7.42006153e-02 -8.81951928e-01 -1.91250141e-03 5.88964820e-01
-1.27688491e+00 -8.88799906e-01 -3.75126392e-01 -6.27660036e-01
5.23241282e-01 1.55613959e-01 5.58821976e-01 -4.06013876e-01
-4.15846258e-01 5.15379727e-01 -1.62819490e-01 -5.14108896e-01
1.67610034e-01 -4.03976232e-01 1.43142834e-01 -5.28545231e-02
7.00694263e-01 -9.12223041e-01 -7.77663171e-01 1.11967832e-01
-8.54304135e-01 3.45066376e-02 4.27969456e-01 5.96922576e-01
3.34603488e-01 1.15004271e-01 6.64190590e-01 -5.31805813e-01
1.33036983e+00 -6.47006869e-01 -4.10035163e-01 1.68038681e-01
-4.52457696e-01 -1.43901613e-02 1.40756831e-01 -7.92740822e-01
-9.02059019e-01 -5.62030673e-01 3.40422541e-02 -7.91323259e-02
-2.86175519e-01 9.07333910e-01 6.27229810e-02 -1.39965028e-01
8.04849625e-01 2.68011749e-01 2.21204951e-01 -2.27206513e-01
-1.24722019e-01 5.03750026e-01 2.91475624e-01 -4.23108488e-01
-1.20863311e-01 2.18503684e-01 -1.51268139e-01 -6.64830148e-01
1.97063506e-01 3.77801666e-03 -7.76311934e-01 -3.97290081e-01
9.03253675e-01 -7.99298584e-01 -4.91148502e-01 4.42496032e-01
-1.06655014e+00 -4.74018157e-01 1.39420912e-01 8.67132425e-01
-5.96559823e-01 1.03681562e-02 -3.04703414e-01 -6.98960900e-01
-6.91401884e-02 -1.37909663e+00 7.44497299e-01 3.01456414e-02
-7.11632669e-01 -9.75127280e-01 -1.15352288e-01 -2.16451392e-01
7.40242541e-01 2.52822727e-01 1.11919200e+00 -7.76291072e-01
3.64198208e-01 -2.09180098e-02 5.80050386e-02 -2.46552691e-01
7.09164143e-02 -3.14598858e-01 -1.16883671e+00 -1.10280000e-01
6.63788676e-01 2.16012001e-01 5.50446212e-01 7.72084951e-01
1.34439981e+00 4.89291660e-02 -4.92707789e-01 2.86126405e-01
8.22052121e-01 7.82894671e-01 8.42137516e-01 8.36166516e-02
1.76281005e-01 1.14076054e+00 2.93687314e-01 1.97057724e-01
1.24668069e-02 5.41969538e-01 -9.23824459e-02 7.17068970e-01
2.98601955e-01 4.10252899e-01 5.14851332e-01 7.21078217e-01
-8.73037055e-02 2.07969412e-01 -1.07097101e+00 4.20789599e-01
-1.63757432e+00 -1.06083655e+00 -2.51276314e-01 2.60357642e+00
5.40839374e-01 5.59715107e-02 -1.54462699e-02 2.72277147e-01
6.96184158e-01 -4.58598472e-02 -6.76714540e-01 -5.89265764e-01
-1.24859028e-01 4.41131085e-01 2.40459159e-01 2.36869082e-01
-6.49407744e-01 5.37249148e-01 7.01138878e+00 5.14646471e-01
-1.27349854e+00 4.00054008e-01 4.20353025e-01 -5.14641523e-01
-1.53720230e-01 -5.17003685e-02 -5.73178679e-02 5.89517832e-01
1.74085450e+00 -3.92898738e-01 1.09205675e+00 -7.83029720e-02
7.29420364e-01 -6.19363487e-01 -1.25678170e+00 1.35833085e+00
-1.55706396e-02 -9.15814579e-01 -5.75104654e-01 -2.56159585e-02
-5.13873026e-02 2.31258169e-01 -9.75095388e-03 2.93343157e-01
-5.13646066e-01 -1.08316779e+00 8.19188297e-01 6.43480957e-01
7.88323581e-01 -3.19092900e-01 2.76045322e-01 2.77096301e-01
-8.33787262e-01 -9.80449915e-02 -3.29083562e-01 -4.43478465e-01
2.99114734e-01 5.21004975e-01 -4.51747961e-02 1.84772862e-03
7.51574695e-01 7.72159696e-01 -6.59366786e-01 1.18711638e+00
1.06945992e-01 5.50110638e-01 -2.65866458e-01 2.48750225e-02
-2.00755790e-01 -1.55086130e-01 4.49132770e-01 1.15621436e+00
4.53445256e-01 6.26116753e-01 -7.81816602e-01 9.22238111e-01
7.06355035e-01 -1.63172245e-01 -4.94486928e-01 -3.63308728e-01
5.13573229e-01 1.09164202e+00 -1.24374235e+00 -5.67875765e-02
-2.98600465e-01 8.16251099e-01 2.14332521e-01 5.59585810e-01
-5.92191517e-01 -4.17400837e-01 6.21588528e-01 -4.02094014e-02
1.00278743e-01 -4.39234167e-01 -5.66664398e-01 -1.12901568e+00
1.43116653e-01 -8.68321598e-01 2.91321307e-01 -1.06278813e+00
-1.15018713e+00 5.29032469e-01 7.23993897e-01 -7.99152374e-01
-3.80173236e-01 -8.01127374e-01 -6.78004563e-01 1.34583056e+00
-8.88068140e-01 -4.43819910e-02 5.97877167e-02 6.24737561e-01
3.65420282e-01 5.99786490e-02 9.88409579e-01 3.92240971e-01
-5.77261865e-01 -3.44558023e-02 -6.28323257e-02 -3.47432733e-01
6.07175827e-01 -6.35730982e-01 3.51938337e-01 8.80277216e-01
-1.52446017e-01 1.01546776e+00 6.03282928e-01 -5.47654867e-01
-1.36421072e+00 -3.91830534e-01 6.99660480e-01 -2.35663980e-01
6.67977333e-01 -7.18286455e-01 -9.74749684e-01 5.36283910e-01
1.38475984e-01 -3.19558471e-01 1.04961634e+00 1.19533993e-01
1.06456503e-01 2.19974041e-01 -1.18126535e+00 9.32319045e-01
1.33645022e+00 -6.42209053e-01 -7.27384210e-01 2.37195492e-01
-2.51592195e-04 -1.51773527e-01 -1.04032457e+00 4.33522433e-01
9.08704937e-01 -8.49646032e-01 7.45612860e-01 -7.65083969e-01
-2.45558247e-01 -4.35548946e-02 -4.71257092e-03 -1.60102952e+00
-9.05740023e-01 -4.05270100e-01 3.37174162e-02 7.02491105e-01
1.75913841e-01 -9.79382813e-01 -2.05776058e-02 9.95016873e-01
-1.84547290e-01 -4.07756299e-01 -1.18753409e+00 -4.92039412e-01
9.92370918e-02 -1.06835163e+00 3.70473534e-01 6.82626069e-01
4.71604317e-01 2.45085746e-01 3.16578299e-01 -2.45748758e-02
2.33565077e-01 -3.54290843e-01 -2.26803809e-01 -1.23180985e+00
1.37784675e-01 -9.25490558e-01 -5.58277786e-01 -3.97730440e-01
6.42408356e-02 -1.13990867e+00 9.25645530e-02 -1.40010691e+00
-1.17247023e-01 -7.72097111e-02 -6.08312309e-01 4.22800750e-01
-1.11513279e-01 2.20201891e-02 -5.86712807e-02 1.91019669e-01
-4.92833145e-02 2.42008954e-01 6.63299322e-01 1.04695775e-01
-1.44468889e-01 -3.03485364e-01 -7.61418939e-01 1.42637327e-01
1.04549515e+00 -5.60021818e-01 -6.66771293e-01 -5.78115880e-01
1.36576742e-01 2.21240073e-01 4.66422647e-01 -1.09262681e+00
2.60273486e-01 1.21205725e-01 7.89169431e-01 -2.01455310e-01
2.20284939e-01 -6.16447747e-01 4.92737025e-01 4.07662779e-01
-2.70474792e-01 5.73139966e-01 7.65235543e-01 3.77302289e-01
9.35254991e-02 -1.92766890e-01 6.85545206e-01 3.63785140e-02
-6.72541022e-01 -1.68613456e-02 -1.17716849e+00 -1.13610268e-01
8.26102078e-01 -6.53872192e-01 -2.08446726e-01 2.16074623e-02
-1.09888649e+00 2.83152238e-02 -4.97751236e-02 5.82313240e-01
5.11814654e-01 -1.30355346e+00 -5.70841968e-01 6.66715682e-01
-3.01930513e-02 -7.22450256e-01 3.93208236e-01 1.71257186e+00
-1.40930591e-02 5.92240274e-01 -7.22245455e-01 -5.51172018e-01
-9.08697128e-01 3.70348662e-01 3.55813861e-01 3.35995734e-01
-5.44061363e-01 6.43666267e-01 1.14614114e-01 1.74565703e-01
1.04069993e-01 -3.63185942e-01 -5.10762811e-01 6.55638456e-01
8.41181815e-01 3.00592989e-01 6.41935229e-01 -6.86271906e-01
-7.55506575e-01 1.61509253e-02 -7.42363483e-02 -7.97730505e-01
1.47861433e+00 -1.23362415e-01 -3.20447832e-01 1.07308805e+00
1.03776979e+00 -5.97887397e-01 -1.05329871e+00 2.58180916e-01
1.59019127e-01 -3.21433932e-01 3.85316104e-01 -8.99354577e-01
-8.10407281e-01 1.17145228e+00 1.05179548e+00 1.08487673e-01
1.35242212e+00 -3.66971582e-01 -2.11618114e-02 -5.66284219e-03
5.22471607e-01 -1.10300589e+00 -5.64915836e-01 4.19719696e-01
1.12392557e+00 -8.70045543e-01 4.34134044e-02 2.18364879e-01
-5.54276526e-01 1.16317999e+00 1.63720936e-01 -1.51961237e-01
1.06347799e+00 1.19740821e-01 -4.83135045e-01 -3.35823059e-01
-9.47892010e-01 1.92212816e-02 4.88233358e-01 7.02768624e-01
9.23740625e-01 1.18381910e-01 -1.09453130e+00 9.47483957e-01
-2.77871460e-01 5.95378950e-02 2.16792077e-01 1.16480637e+00
-9.90749970e-02 -8.09869647e-01 -6.01055264e-01 1.22909617e+00
-5.23973405e-01 -2.61525244e-01 4.72476892e-02 5.83816409e-01
-1.53411478e-01 1.01898718e+00 5.88338196e-01 -3.40437770e-01
3.87565583e-01 5.80507815e-01 8.46093833e-01 -6.88103080e-01
-6.73939466e-01 1.75489366e-01 1.36935264e-01 -7.14273214e-01
-4.90583479e-01 -1.41622472e+00 -1.11967731e+00 -8.34019333e-02
-1.51325971e-01 -3.54945250e-02 9.64306056e-01 1.21866810e+00
5.16939521e-01 7.63038099e-01 -1.39365286e-01 -1.19906366e+00
7.24569187e-02 -1.32741892e+00 -4.73890722e-01 -4.99511976e-03
-4.72985543e-02 -1.10868430e+00 -4.00857359e-01 1.49361463e-02] | [12.972424507141113, 3.4114291667938232] |
e457b9d9-efb5-4bf4-97bd-24ee1f9bc6a5 | arthroscopic-multi-spectral-scene | 2103.02465 | null | https://arxiv.org/abs/2103.02465v1 | https://arxiv.org/pdf/2103.02465v1.pdf | Arthroscopic Multi-Spectral Scene Segmentation Using Deep Learning | Knee arthroscopy is a minimally invasive surgical (MIS) procedure which is performed to treat knee-joint ailment. Lack of visual information of the surgical site obtained from miniaturized cameras make this surgical procedure more complex. Knee cavity is a very confined space; therefore, surgical scenes are captured at close proximity. Insignificant context of knee atlas often makes them unrecognizable as a consequence unintentional tissue damage often occurred and shows a long learning curve to train new surgeons. Automatic context awareness through labeling of the surgical site can be an alternative to mitigate these drawbacks. However, from the previous studies, it is confirmed that the surgical site exhibits several limitations, among others, lack of discriminative contextual information such as texture and features which drastically limits this vision task. Additionally, poor imaging conditions and lack of accurate ground-truth labels are also limiting the accuracy. To mitigate these limitations of knee arthroscopy, in this work we proposed a scene segmentation method that successfully segments multi structures. | ['Dr. Ajay K. Pandey', 'Cameron Brown', 'Ross Crawford', 'Jonathan Roberts', 'Yu Takeda', 'Dr. Yaqub Jonmohamadi', 'Shahnewaz Ali'] | 2021-03-03 | null | null | null | null | ['scene-segmentation'] | ['computer-vision'] | [ 1.06473848e-01 -7.04589784e-02 -2.56904095e-01 1.91346914e-01
-5.68970680e-01 -5.26853383e-01 1.12410501e-01 1.32514358e-01
-4.99576449e-01 8.32313299e-01 2.22600088e-01 -2.08765164e-01
5.83221540e-02 -1.09275937e-01 -4.69358772e-01 -6.21376336e-01
1.36240408e-01 6.30859435e-02 5.36000252e-01 2.17600584e-01
5.36074340e-01 5.74072361e-01 -1.28383458e+00 2.58826222e-02
6.69974744e-01 6.51275933e-01 9.43588972e-01 3.04354429e-01
5.50099574e-02 5.18188477e-01 -3.57056051e-01 7.48473629e-02
6.40953302e-01 -3.19410533e-01 -5.80426514e-01 5.33387601e-01
4.58755970e-01 -4.72870618e-01 -1.54608577e-01 1.15622616e+00
2.75677681e-01 5.50656095e-02 5.15639901e-01 -5.78342736e-01
4.07473929e-02 9.96618718e-03 -8.51660550e-01 2.64680535e-01
2.95240670e-01 2.71735966e-01 2.55597830e-01 -9.86753702e-01
8.76730621e-01 5.31599164e-01 4.17857617e-01 3.66729587e-01
-9.99221206e-01 -5.82676649e-01 -1.43364653e-01 5.48598804e-02
-1.32805645e+00 2.17523053e-03 8.22032094e-01 -6.33795619e-01
5.23476005e-01 1.75866440e-01 1.23536015e+00 7.61980772e-01
1.07925749e+00 4.01871413e-01 1.70893490e+00 -4.85753387e-01
2.80151278e-01 3.81589867e-02 -2.74003774e-01 7.82978714e-01
5.53492367e-01 2.92782307e-01 -1.81470931e-01 1.29057288e-01
1.29446971e+00 6.17038727e-01 -3.66026491e-01 -1.07339406e+00
-1.23062873e+00 3.73351634e-01 5.92552781e-01 4.09265608e-01
-3.73811126e-01 -2.24630511e-03 3.17811072e-01 -2.79884309e-01
-4.60064203e-01 4.15486217e-01 9.26484019e-02 -2.47253224e-01
-8.45957696e-01 -5.90197802e-01 3.55858594e-01 7.80485511e-01
4.63394850e-01 4.81388345e-02 3.97116065e-01 6.27272248e-01
2.30952144e-01 1.02442302e-01 6.53334498e-01 -8.66317570e-01
9.27309692e-03 7.58977890e-01 -2.31939659e-01 -9.27158952e-01
-4.33749199e-01 -6.16741359e-01 -7.74759471e-01 5.75701475e-01
4.71433282e-01 9.37583596e-02 -1.49333048e+00 1.07356954e+00
4.12245274e-01 6.03947155e-02 -4.01966780e-01 1.52736998e+00
4.47007298e-01 -3.76441441e-02 3.45759541e-01 -4.93869752e-01
1.58761168e+00 -9.95658338e-01 -7.32348204e-01 -7.57498384e-01
8.46069813e-01 -1.16651058e+00 1.31470215e+00 5.52437961e-01
-6.25090301e-01 -6.80125579e-02 -1.13061357e+00 1.54788852e-01
-1.10406347e-01 3.05159360e-01 1.06405222e+00 4.82963741e-01
-7.22123206e-01 2.21473739e-01 -1.07321227e+00 -7.21132934e-01
1.51328787e-01 4.13914382e-01 -8.76284540e-01 -1.77415803e-01
-5.04849255e-01 1.07858026e+00 2.50805557e-01 3.88494819e-01
-8.12577128e-01 -3.76678437e-01 -8.35181415e-01 -4.14601028e-01
8.00630331e-01 -4.88849252e-01 8.42735350e-01 -8.05726230e-01
-1.48067713e+00 1.32918167e+00 1.69251546e-01 -7.67125264e-02
3.00364286e-01 -3.21499914e-01 -2.77274460e-01 2.60097653e-01
1.69576317e-01 2.39544466e-01 5.23013711e-01 -1.26305103e+00
-3.50424886e-01 -6.56931579e-01 1.53705431e-02 4.06338781e-01
2.35666662e-01 -2.50224650e-01 -3.33867282e-01 -7.63702691e-01
5.46792448e-01 -1.25538194e+00 -7.81901836e-01 3.45079213e-01
-3.99009734e-01 4.74582314e-01 6.44412518e-01 -6.70410395e-01
1.18394184e+00 -2.12679768e+00 -2.94212520e-01 5.15980758e-02
1.37895405e-01 1.71338573e-01 5.68647385e-01 2.49773622e-01
-1.54877171e-01 -2.70549357e-02 -2.90956888e-02 3.05537105e-01
-7.68553972e-01 3.29766005e-01 4.37244833e-01 7.81894624e-01
-5.26513219e-01 3.65410000e-01 -7.29428291e-01 -1.13886845e+00
6.66524231e-01 6.37827888e-02 -5.45781791e-01 1.68034300e-01
2.62582093e-01 7.97479928e-01 -4.39752907e-01 9.57450092e-01
5.56902885e-01 -4.20346484e-02 1.97835967e-01 -4.97924268e-01
-2.07437258e-02 -2.17261761e-01 -1.23959315e+00 2.35859871e+00
-4.95756894e-01 2.79417783e-01 2.33690783e-01 -6.66673958e-01
6.11373723e-01 3.98771793e-01 5.33349991e-01 -5.87235749e-01
2.65913814e-01 4.98342663e-01 7.24494234e-02 -7.88886011e-01
2.74273187e-01 -3.84001225e-01 -6.34170473e-02 -8.92948285e-02
-1.30461261e-01 -2.97372848e-01 -1.99129969e-01 8.13569129e-02
7.71099687e-01 2.83292145e-01 7.84092903e-01 -3.60653341e-01
3.72229040e-01 4.97811705e-01 7.69080162e-01 5.84707975e-01
-4.90505248e-01 1.05028248e+00 1.03668503e-01 -4.04247522e-01
-8.03445816e-01 -1.37108624e+00 -8.65207985e-02 3.90164480e-02
5.61005771e-01 -2.85022616e-01 -5.74798405e-01 -6.97052717e-01
-3.62861365e-01 2.64084935e-01 -4.43976939e-01 -3.04405987e-01
-5.50771236e-01 -1.23511285e-01 -3.74818325e-01 4.72142756e-01
1.88116863e-01 -3.81248564e-01 -8.78006279e-01 1.04838073e-01
-9.93036628e-02 -1.09610939e+00 -5.23351133e-01 2.12915227e-01
-1.34530604e+00 -1.32673371e+00 -7.56708026e-01 -9.94416893e-01
9.73572373e-01 6.59995079e-01 6.74158752e-01 1.13408733e-02
-1.04019678e+00 3.20699483e-01 4.24651895e-03 -1.77626967e-01
-4.08792198e-02 -1.55618340e-01 -9.62712616e-02 -3.41261059e-01
2.72679567e-01 -3.75626445e-01 -1.34053314e+00 2.23085776e-01
-6.31097436e-01 2.61265218e-01 1.20103860e+00 7.75302947e-01
1.01407981e+00 -3.05354416e-01 -5.30477017e-02 -8.41408372e-01
1.57833427e-01 -2.92520374e-01 -3.52780342e-01 1.11599967e-01
-4.38714802e-01 -4.20374870e-01 3.47143143e-01 -2.95737982e-01
-1.12988579e+00 4.05141413e-01 5.69429517e-01 -3.97080541e-01
-6.60511494e-01 4.74101841e-01 1.78718761e-01 -1.82963625e-01
5.29794693e-01 -1.66656807e-01 3.60367745e-01 -3.76404047e-01
-3.00475001e-01 7.02276230e-01 6.65969551e-01 -1.60877824e-01
3.42049479e-01 5.76329470e-01 3.52659374e-01 -7.62339890e-01
-8.01009357e-01 -7.77781010e-01 -7.58737445e-01 -6.41314089e-01
9.92130339e-01 -7.56018281e-01 -2.63180077e-01 -1.93506330e-01
-6.05250359e-01 1.09913833e-01 -1.95659716e-02 1.33149421e+00
-4.44757015e-01 7.12572813e-01 -6.02429748e-01 -4.86345232e-01
-2.05117941e-01 -1.48909247e+00 7.79332101e-01 4.28382248e-01
-2.68385530e-01 -8.61882985e-01 -1.15899809e-01 7.27608263e-01
1.54553920e-01 5.10350227e-01 7.51060665e-01 -1.99248344e-01
-9.08757687e-01 -5.27237296e-01 1.15734525e-01 1.88971207e-01
3.83219779e-01 -1.02502450e-01 -5.17549396e-01 -2.16400787e-01
3.53201866e-01 3.92423477e-03 2.01182276e-01 5.85748494e-01
1.10379589e+00 3.00238073e-01 -5.07018328e-01 4.80949223e-01
1.77934575e+00 3.13856900e-01 5.19993782e-01 6.03343248e-01
6.37830734e-01 4.82207894e-01 1.34563279e+00 1.65475205e-01
-7.39347562e-02 7.68900096e-01 5.73737919e-01 -2.83566833e-01
-4.72789228e-01 -1.61793679e-01 8.54937434e-02 9.99915063e-01
-2.95913190e-01 3.79879951e-01 -1.10106850e+00 4.58225995e-01
-1.41837013e+00 -4.02111024e-01 -4.12542701e-01 2.62176132e+00
7.34868288e-01 -4.59896736e-02 -4.67934757e-01 -2.75734186e-01
6.30457401e-01 -1.85424209e-01 -2.19979644e-01 -1.85871124e-01
3.15258324e-01 -5.48601672e-02 7.85817146e-01 1.70103356e-01
-1.00081611e+00 8.20670068e-01 6.22109318e+00 7.97675073e-01
-1.46615970e+00 6.24550395e-02 1.16403788e-01 -2.47935653e-01
7.65995905e-02 3.94220591e-01 -3.21746141e-01 3.90628338e-01
1.92635134e-01 1.74008846e-01 -1.73953831e-01 9.11744833e-01
3.40377331e-01 -1.04146934e+00 -8.45250964e-01 1.10872269e+00
3.19070220e-01 -1.12849331e+00 -1.39775977e-01 3.20302278e-01
5.70382714e-01 -4.00646985e-01 -8.05155784e-02 -1.66474238e-01
-1.83745101e-01 -8.05478573e-01 4.77173552e-02 6.00452662e-01
7.49605954e-01 -4.82333183e-01 9.16644752e-01 1.00620329e-01
-8.07293952e-01 1.22744165e-01 -1.95767567e-01 -4.84459698e-02
1.87192395e-01 3.77972692e-01 -1.12081230e+00 1.48281485e-01
6.15929484e-01 3.13118011e-01 -6.41856849e-01 1.63459766e+00
-1.08475290e-01 2.93587476e-01 -2.64798164e-01 2.89989382e-01
2.39925206e-01 -2.60874957e-01 7.66598344e-01 6.14816368e-01
4.25947905e-01 3.08998954e-02 4.49043006e-01 2.90478319e-01
4.43328559e-01 3.74705702e-01 -8.91478539e-01 2.61628807e-01
7.42240846e-02 1.31314313e+00 -1.19237185e+00 2.20646977e-01
-7.12526560e-01 9.96260464e-01 -2.20059287e-02 3.27779740e-01
-3.64860147e-01 1.42542005e-01 9.64224264e-02 6.13095582e-01
-2.52117425e-01 -4.28675503e-01 -3.07479352e-01 -1.22241700e+00
1.00757524e-01 -4.74324167e-01 4.99419540e-01 -9.89065170e-01
-4.94920105e-01 1.39936492e-01 -1.02215722e-01 -1.94516349e+00
-1.44408628e-01 -5.85039258e-01 -3.58576447e-01 3.10142338e-01
-9.33710814e-01 -1.21843886e+00 -5.91140091e-01 3.88118178e-01
7.72939801e-01 8.85291696e-02 7.07339525e-01 -7.63549581e-02
-4.83959705e-01 2.35244315e-02 -4.75973561e-02 -1.28736749e-01
1.06390142e+00 -1.24315178e+00 -1.01216459e+00 9.45417225e-01
-3.31093758e-01 8.03273737e-01 8.98786604e-01 -9.59217131e-01
-1.46005464e+00 -4.69180495e-01 1.30011708e-01 -1.11123517e-01
5.62677145e-01 -6.75690100e-02 -6.66812301e-01 6.62793577e-01
1.77469969e-01 -4.67753448e-02 1.01222575e+00 -2.76708394e-01
2.80009896e-01 -1.40253544e-01 -1.04600585e+00 1.16778934e+00
7.93267667e-01 -2.40711302e-01 -8.56568456e-01 2.77398586e-01
-1.18239289e-02 -7.81956136e-01 -1.15520692e+00 5.45941770e-01
5.72726846e-01 -9.99634564e-01 8.77124429e-01 1.18648810e-02
4.50457424e-01 -3.04237843e-01 1.30668461e-01 -9.44360197e-01
1.92921415e-01 -2.35831484e-01 3.09978247e-01 8.04660082e-01
4.96205091e-02 -5.65354884e-01 1.13829529e+00 8.50444436e-01
-4.46252108e-01 -7.21880138e-01 -1.02323794e+00 -7.11337030e-01
-2.17868492e-01 -5.60225062e-02 -2.64642656e-01 8.24644983e-01
2.22536877e-01 -5.81822731e-02 -1.70311481e-01 2.07244143e-01
8.45393777e-01 2.27173969e-01 7.09204495e-01 -1.03979731e+00
-1.22329578e-01 -1.50428101e-01 -7.92793334e-01 -5.76628029e-01
-4.34950471e-01 -8.12548995e-01 -1.48076326e-01 -1.62647676e+00
4.38802153e-01 -3.46329510e-01 -1.72635496e-01 -9.75522324e-02
1.06717713e-01 2.83323795e-01 -8.44862014e-02 5.22195697e-01
-2.45497078e-01 9.35789645e-02 1.83736598e+00 4.20582116e-01
-5.52218370e-02 -2.85612326e-02 -3.43530893e-01 1.02318013e+00
7.98055232e-01 -5.49366474e-01 -5.43853462e-01 -1.92794204e-01
-1.32158011e-01 4.55767661e-01 4.37973499e-01 -1.03317928e+00
6.67889535e-01 -3.26202512e-01 5.36771178e-01 -7.28027523e-01
4.65334505e-01 -1.38210285e+00 4.14188892e-01 8.06769073e-01
1.94083825e-01 -2.44723961e-01 5.96433990e-02 6.79327607e-01
-4.87743795e-01 -4.71365482e-01 7.90431499e-01 -7.83128917e-01
-8.73398602e-01 1.69364363e-02 -7.46250987e-01 -5.16504720e-02
1.47984934e+00 -1.12538540e+00 1.90146282e-01 -6.85637668e-02
-1.08009803e+00 -1.67362332e-01 9.51868713e-01 1.88918531e-01
8.15407395e-01 -1.00898838e+00 -7.42497966e-02 2.23561138e-01
2.04293534e-01 1.59971386e-01 8.20818961e-01 1.60153997e+00
-1.18419206e+00 6.05249166e-01 -4.13435191e-01 -8.43975663e-01
-1.43743861e+00 5.55421054e-01 5.54160595e-01 -5.85068800e-02
-9.41475689e-01 1.61403999e-01 5.54085732e-01 -1.07553944e-01
2.36970633e-01 -1.65184960e-01 -8.59555304e-02 -4.46660042e-01
1.19108772e-02 5.35048172e-02 -8.14263299e-02 -4.74756658e-01
-3.87934536e-01 9.26587403e-01 -3.87952983e-01 1.78982779e-01
8.40588510e-01 -5.30012250e-01 8.30888450e-02 4.43623632e-01
8.34932804e-01 1.64405942e-01 -1.09456384e+00 -9.99913961e-02
-1.60445765e-01 -9.98839855e-01 3.94834280e-01 -9.09803271e-01
-1.01279092e+00 9.32444513e-01 9.01496112e-01 -5.42233527e-01
1.02879453e+00 -6.41299365e-03 5.57336569e-01 -6.02305308e-02
9.57637906e-01 -1.38390911e+00 1.55709975e-03 -6.63556159e-02
8.22343469e-01 -1.39072502e+00 1.88830957e-01 -9.16483879e-01
-7.89202154e-01 1.25569904e+00 1.05555499e+00 -8.41834918e-02
6.89323068e-01 3.60155970e-01 2.69102991e-01 -4.02901530e-01
-1.28188342e-01 -1.25093535e-01 1.83076188e-01 6.67105675e-01
5.54698467e-01 1.29929185e-01 -8.50477457e-01 2.17673197e-01
2.09680498e-02 1.87972412e-01 6.96163476e-01 1.34077930e+00
-4.06483501e-01 -6.45721495e-01 -4.99629557e-01 5.10203421e-01
-8.59822929e-01 1.46862745e-01 -2.46909596e-02 1.17710447e+00
-1.46945909e-01 7.89972484e-01 -1.05769806e-01 3.51317860e-02
2.95020521e-01 -4.59670722e-01 4.25675899e-01 -8.12345386e-01
-4.92155910e-01 5.32924950e-01 -2.01500189e-02 -8.15201759e-01
-1.76231429e-01 -4.24567878e-01 -1.32634974e+00 1.76755503e-01
-4.36527818e-01 -6.79144636e-02 8.38946462e-01 7.78114140e-01
1.00310102e-01 3.37874025e-01 2.12793753e-01 -5.01670539e-01
-8.22192281e-02 -5.95373273e-01 -8.49455237e-01 3.97427171e-01
1.82242803e-02 -9.77104723e-01 -3.55018705e-01 3.52918804e-02] | [13.809867858886719, -3.1339075565338135] |
877db9f1-07ed-4e45-8cdb-a99238525b67 | reinforce-security-a-model-free-approach | 2106.00343 | null | https://arxiv.org/abs/2106.00343v1 | https://arxiv.org/pdf/2106.00343v1.pdf | Reinforce Security: A Model-Free Approach Towards Secure Wiretap Coding | The use of deep learning-based techniques for approximating secure encoding functions has attracted considerable interest in wireless communications due to impressive results obtained for general coding and decoding tasks for wireless communication systems. Of particular importance is the development of model-free techniques that work without knowledge about the underlying channel. Such techniques utilize for example generative adversarial networks to estimate and model the conditional channel distribution, mutual information estimation as a reward function, or reinforcement learning. In this paper, the approach of reinforcement learning is studied and, in particular, the policy gradient method for a model-free approach of neural network-based secure encoding is investigated. Previously developed techniques for enforcing a certain co-set structure on the encoding process can be combined with recent reinforcement learning approaches. This new approach is evaluated by extensive simulations, and it is demonstrated that the resulting decoding performance of an eavesdropper is capped at a certain error level. | ['Gerhard Wunder', 'Rafael F. Schaefer', 'Rick Fritschek'] | 2021-06-01 | null | null | null | null | ['mutual-information-estimation'] | ['methodology'] | [ 4.98011798e-01 5.47869682e-01 1.69590935e-02 -1.53047949e-01
-9.11110044e-01 -4.05034602e-01 6.06522381e-01 4.90323417e-02
-4.67137426e-01 9.31127906e-01 -1.14676408e-01 -6.71554387e-01
-2.51949906e-01 -8.33286166e-01 -7.83649266e-01 -1.03577852e+00
-6.58049822e-01 1.23450875e-01 -5.14132798e-01 -4.10535514e-01
2.09236488e-01 5.12948871e-01 -7.31271505e-01 -3.63625586e-02
4.99803156e-01 1.06161690e+00 -5.73041290e-02 1.14414871e+00
2.05758214e-01 7.14881241e-01 -7.57050157e-01 -6.86490297e-01
3.31917495e-01 -6.67253256e-01 -5.08534133e-01 -9.61503088e-02
-4.68007803e-01 -4.42839980e-01 -7.87971854e-01 1.29428959e+00
5.49127460e-01 -3.23740631e-01 5.80777168e-01 -1.01762736e+00
-3.12621891e-01 9.01091039e-01 1.25433594e-01 -2.33252406e-01
2.12454900e-01 -1.04581505e-01 7.76821554e-01 -1.25218853e-01
1.49867848e-01 9.42718148e-01 6.45922482e-01 7.44916320e-01
-1.27632511e+00 -8.98795605e-01 -2.84675479e-01 7.65106007e-02
-1.57134640e+00 -3.78863633e-01 8.60870302e-01 -2.92692916e-03
7.58201122e-01 1.43445417e-01 7.08130360e-01 1.15579927e+00
5.24513364e-01 8.16734552e-01 1.34393847e+00 -7.96191990e-01
6.04314983e-01 1.48510247e-01 -6.43594086e-01 7.61582911e-01
-4.81029898e-02 8.58748138e-01 -4.46839258e-02 -3.75930697e-01
6.48801625e-01 -3.34236145e-01 -3.54346901e-01 -5.24784565e-01
-5.64054251e-01 1.08920765e+00 3.79221588e-01 3.54570091e-01
-3.08961987e-01 8.09659183e-01 2.56382316e-01 8.67222428e-01
3.55298638e-01 4.68573123e-01 -2.90592760e-01 -2.46480450e-01
-9.83749747e-01 3.58428419e-01 1.45678341e+00 9.26844656e-01
5.08327544e-01 5.35065651e-01 -1.34729799e-02 3.62082124e-01
4.70674545e-01 5.53235292e-01 1.36519685e-01 -8.68681669e-01
3.86034548e-01 -3.30987900e-01 1.12542167e-01 -7.69186854e-01
-2.83255517e-01 -5.59401453e-01 -1.06960261e+00 3.92158180e-01
2.88999051e-01 -7.77494311e-01 -8.06914449e-01 1.87473524e+00
-1.92689300e-01 3.84077758e-01 6.16469860e-01 3.09217840e-01
-2.79967859e-02 7.67762065e-01 -1.13021113e-01 -4.35932904e-01
4.80353922e-01 -2.10791573e-01 -6.97614074e-01 2.59549767e-01
4.77901697e-01 -5.46739638e-01 8.06893855e-02 5.12445331e-01
-1.31581986e+00 -1.29511878e-01 -1.26990855e+00 8.55984092e-01
-2.06578970e-01 -3.26758802e-01 4.85869944e-01 1.45641160e+00
-1.34771287e+00 8.72253835e-01 -7.97637165e-01 2.23691523e-01
4.72744256e-01 1.01766002e+00 5.29321916e-02 -4.82301740e-03
-1.54569244e+00 7.32459962e-01 6.99252784e-01 1.97957069e-01
-1.33505177e+00 -2.71735609e-01 -8.58483732e-01 3.69805217e-01
6.19265959e-02 -3.58896405e-01 1.26867437e+00 -1.19209564e+00
-2.13587713e+00 3.17509711e-01 3.70063454e-01 -1.07360470e+00
6.05277479e-01 2.33790070e-01 -3.61046880e-01 1.75072163e-01
-7.18191683e-01 2.10431457e-01 1.32406950e+00 -1.36749399e+00
-3.89061481e-01 3.13958339e-02 2.05553427e-01 -8.00770521e-02
-4.33644414e-01 -1.77376896e-01 1.16728075e-01 -7.05632567e-01
-2.28680640e-01 -9.66686130e-01 -5.77836514e-01 -5.06409816e-02
-3.38529646e-01 2.48201355e-01 6.51412010e-01 -6.75058246e-01
1.05186081e+00 -1.94195926e+00 3.22736263e-01 9.29956615e-01
-1.78555518e-01 5.20498693e-01 -1.67027816e-01 7.53414929e-01
-6.64559519e-03 2.84441501e-01 -4.20749724e-01 -1.96464807e-01
-3.03643532e-02 4.91291940e-01 -2.62944758e-01 6.16283774e-01
3.72656733e-02 6.77314878e-01 -8.60653341e-01 -1.15298303e-02
6.40246943e-02 7.40504801e-01 -7.22164094e-01 4.99696076e-01
-1.26513675e-01 4.46037322e-01 -4.02532518e-01 1.90985784e-01
6.30645394e-01 1.98854432e-01 4.10718203e-01 3.50302845e-01
2.91558951e-01 3.43901548e-03 -9.16428387e-01 1.53195643e+00
-8.10053170e-01 5.89821637e-01 5.01226366e-01 -1.48536468e+00
8.90173137e-01 8.23996186e-01 4.17102456e-01 -4.98554885e-01
4.59980339e-01 2.94198573e-01 3.44276726e-01 -2.57319659e-01
7.49758184e-02 -3.95820290e-01 -1.30979225e-01 5.20751655e-01
1.89071238e-01 -1.90398455e-01 -2.92975157e-01 7.69162476e-02
1.19317281e+00 -9.60805342e-02 2.92360514e-01 -1.89548552e-01
6.00223839e-01 -6.42836690e-01 6.52112886e-02 1.15644145e+00
5.42387478e-02 7.76003525e-02 6.08160257e-01 1.15545746e-02
-1.34358799e+00 -8.62476408e-01 -9.95548815e-02 4.40501690e-01
-3.94831505e-03 -1.93208784e-01 -1.03454018e+00 -4.94407624e-01
-2.15268299e-01 6.12673283e-01 -5.94198942e-01 -6.07483566e-01
-3.34211528e-01 -7.42990315e-01 1.00111377e+00 8.59405771e-02
5.29707849e-01 -9.20855463e-01 -1.59733132e-01 5.73218703e-01
3.40664446e-01 -8.72555792e-01 2.35422209e-01 7.58471072e-01
-5.70977926e-01 -5.92578471e-01 -9.89406526e-01 -5.74114203e-01
5.36682487e-01 -5.15735924e-01 6.40458107e-01 2.24531755e-01
7.55041763e-02 4.97429639e-01 -4.14441943e-01 -3.59766752e-01
-1.37070632e+00 1.12641610e-01 -8.13861787e-02 8.01124647e-02
-1.90501869e-01 -6.83482111e-01 -2.79581517e-01 -6.86189830e-02
-9.62804794e-01 -3.24558735e-01 6.05341136e-01 1.22566581e+00
-1.00430027e-01 3.02339137e-01 5.64167082e-01 -7.02557445e-01
1.01632822e+00 -4.24248785e-01 -8.42631638e-01 1.74856439e-01
-6.72928214e-01 5.07863939e-01 8.81576777e-01 -3.68761688e-01
-8.48135591e-01 -1.46355378e-02 -7.29891360e-01 -1.44474804e-01
5.69314286e-02 4.15452033e-01 -2.52231538e-01 -7.96998084e-01
5.39620996e-01 5.59133470e-01 3.03824186e-01 1.16906986e-01
5.08519523e-02 8.29536557e-01 1.44650862e-01 -6.85899496e-01
1.17430258e+00 1.62904069e-01 2.84960359e-01 -6.36500418e-01
-4.76195455e-01 1.79766297e-01 -1.96105659e-01 -4.08761874e-02
3.31308395e-01 -7.47191906e-01 -1.01864755e+00 4.71383750e-01
-1.08349681e+00 -5.09823442e-01 -1.96538761e-01 5.35458922e-01
-1.03842580e+00 2.80918360e-01 -5.57066500e-01 -1.34699655e+00
-1.81845799e-01 -1.13168371e+00 5.60560107e-01 4.32077646e-02
3.22865516e-01 -1.23636258e+00 -5.02247959e-02 -3.24261248e-01
6.78639889e-01 1.58852071e-01 9.52111006e-01 -6.53348148e-01
-6.04382157e-01 -5.28149784e-01 4.58939075e-02 7.09028065e-01
-2.32860804e-01 -3.92502517e-01 -1.01264334e+00 -6.00839853e-01
9.88731608e-02 -3.68813694e-01 5.51615655e-01 4.35589373e-01
1.28364289e+00 -5.33484757e-01 -4.73358296e-02 8.24709237e-01
1.61821771e+00 6.23075604e-01 7.68278182e-01 8.48255008e-02
1.62629768e-01 1.33510500e-01 9.84670967e-02 7.54781306e-01
-2.17653915e-01 6.18931234e-01 8.03986549e-01 6.90049976e-02
4.64514285e-01 -1.25067443e-01 3.27840477e-01 3.18770379e-01
-1.06015049e-01 -5.37313879e-01 -5.30308902e-01 8.96345526e-02
-1.55956793e+00 -1.12548065e+00 5.58636427e-01 2.29960394e+00
8.02718937e-01 2.72513449e-01 -1.56144753e-01 5.94649911e-01
5.40917993e-01 6.66382834e-02 -2.65496016e-01 -7.76034474e-01
1.29096225e-01 5.87810636e-01 8.55295599e-01 7.72457421e-01
-1.09807670e+00 6.42896056e-01 6.79055548e+00 1.09817433e+00
-1.03398061e+00 -7.90494308e-03 7.18000472e-01 3.66515368e-01
-2.38649040e-01 -6.89893365e-02 -4.31710482e-01 4.60159302e-01
1.44727457e+00 5.16129211e-02 1.00594389e+00 4.92876083e-01
1.33973852e-01 1.92294881e-01 -9.03381109e-01 8.21569622e-01
-2.54074961e-01 -1.33768702e+00 -2.08260566e-01 4.32451487e-01
5.65075338e-01 -2.96774507e-01 3.00308615e-01 3.84564698e-01
4.51256335e-01 -1.03086984e+00 4.89065140e-01 4.65808332e-01
7.01448262e-01 -1.37100387e+00 7.98721731e-01 3.99708360e-01
-6.63883269e-01 -4.18802321e-01 -3.34924102e-01 -1.00846246e-01
5.53348698e-02 2.07886457e-01 -9.56914604e-01 4.39830929e-01
8.77443254e-02 3.14499557e-01 1.55021295e-01 1.12302971e+00
-2.51866192e-01 9.52034712e-01 -2.72383928e-01 -5.66614747e-01
6.09359562e-01 8.77036452e-02 5.57665408e-01 1.34305942e+00
5.59196353e-01 -1.76962018e-01 1.69349965e-02 5.26605189e-01
-2.36284748e-01 -1.54439628e-01 -7.46232867e-01 -9.36722755e-02
3.34106624e-01 8.79562438e-01 -3.83873969e-01 -3.65139134e-02
-5.13080060e-02 9.97996390e-01 1.59218699e-01 5.34940362e-01
-7.44675457e-01 -4.28836137e-01 4.74936336e-01 -1.14533864e-01
4.57543910e-01 -3.50640863e-01 4.48825173e-02 -8.02861512e-01
-3.21900457e-01 -8.51506889e-01 -2.28267927e-02 -1.56367555e-01
-8.58129025e-01 4.04652923e-01 -1.02286808e-01 -1.08020318e+00
-7.96164811e-01 -5.21582007e-01 -3.84596795e-01 9.40410793e-01
-1.77444983e+00 -7.47315109e-01 4.13693041e-01 8.37471545e-01
5.73289245e-02 -7.89711952e-01 1.32619011e+00 -1.06939273e-02
-1.46268755e-01 9.71341610e-01 8.54936659e-01 3.21557313e-01
-6.71449974e-02 -1.34615159e+00 1.70109898e-01 7.08893657e-01
3.03063035e-01 4.65065032e-01 9.23645854e-01 -2.11045533e-01
-1.54556513e+00 -8.78281713e-01 3.71627420e-01 3.62128675e-01
7.82912016e-01 -5.58457673e-01 -6.29119873e-01 2.72424072e-01
4.62406307e-01 1.63864508e-01 6.28516257e-01 -1.36714637e-01
-1.52612820e-01 -6.17833734e-02 -1.37130415e+00 3.44384730e-01
4.44290400e-01 -5.63546777e-01 1.60159603e-01 3.34967256e-01
2.07680106e-01 -6.08664095e-01 -8.23884666e-01 3.25103104e-03
6.26723588e-01 -7.99942255e-01 9.02452826e-01 -5.50192177e-01
3.12798977e-01 2.31871799e-01 -2.09930539e-01 -1.59024036e+00
5.20132408e-02 -1.42079723e+00 -5.82427323e-01 7.02697814e-01
4.55450237e-01 -3.13992560e-01 1.00956130e+00 3.30027670e-01
1.56257078e-01 -8.85571837e-01 -1.31443429e+00 -8.31127167e-01
4.17902052e-01 -4.69641745e-01 3.11275601e-01 2.97421306e-01
1.08308583e-01 -1.35713279e-01 -8.49300563e-01 3.50937426e-01
7.23495722e-01 -3.99530917e-01 3.99341404e-01 -7.11510241e-01
-9.46878672e-01 -2.17725843e-01 -6.74784303e-01 -1.06988883e+00
4.36674327e-01 -8.78438950e-01 1.05228677e-01 -8.84255528e-01
-2.91257650e-01 -7.63957143e-01 -5.01698554e-01 -5.30947633e-02
3.26802343e-01 2.33302768e-02 1.54146507e-01 -2.67414212e-01
-2.35320240e-01 6.02612376e-01 9.51956391e-01 -2.53237396e-01
2.80104019e-03 6.80396974e-01 -6.61067069e-01 3.14734638e-01
1.05523336e+00 -6.08343542e-01 -4.10372913e-01 9.85346362e-03
3.19342583e-01 6.65673554e-01 3.85886431e-01 -1.14513278e+00
1.04251482e-01 2.56200582e-01 3.06247473e-01 2.22961932e-01
4.89626646e-01 -1.19468939e+00 -1.64233312e-01 1.10075998e+00
-7.18832552e-01 -3.61838162e-01 -2.23933980e-02 9.35168922e-01
-1.85335353e-01 -6.16848469e-01 8.71215820e-01 -1.03391912e-02
-2.61733890e-01 2.41214767e-01 -7.83531964e-01 -1.89755514e-01
9.87920582e-01 5.81330396e-02 4.94998962e-01 -1.13531375e+00
-9.49763358e-01 -2.38621026e-01 -1.64256170e-01 -1.39603898e-01
7.59531200e-01 -1.06164467e+00 -7.82864571e-01 2.99348325e-01
-2.43782386e-01 -6.51622057e-01 -2.14076132e-01 2.77589828e-01
-4.05319214e-01 3.30572069e-01 -1.03272200e-01 -2.98689634e-01
-1.06169343e+00 3.91813666e-01 7.19541669e-01 -6.35436535e-01
-3.64809155e-01 7.01036215e-01 -4.58072811e-01 2.33430341e-02
4.61489379e-01 1.05469443e-01 9.11527798e-02 -6.55713856e-01
4.69523281e-01 -2.03002587e-01 2.36972626e-02 -2.06128433e-01
1.77749336e-01 3.50015685e-02 -2.38780409e-01 -3.51712763e-01
1.19472706e+00 -4.21391912e-02 2.88920134e-01 -1.46709993e-01
1.46440911e+00 -3.35066110e-01 -1.41549802e+00 -2.38349304e-01
-1.29246145e-01 -2.71203190e-01 3.06580186e-01 -7.28511155e-01
-1.07811856e+00 1.05277371e+00 9.07067239e-01 7.05509603e-01
1.14090371e+00 -4.65809196e-01 5.62905312e-01 7.84861207e-01
7.69798636e-01 -8.46422911e-01 -1.89905792e-01 4.77181315e-01
5.07889628e-01 -1.02579999e+00 -3.09662372e-01 3.37071382e-02
-7.40367770e-02 1.36196542e+00 -2.62114465e-01 -4.75370586e-01
9.97684717e-01 9.00732696e-01 -2.77335405e-01 3.48457724e-01
-3.44411373e-01 6.88515082e-02 -1.62390739e-01 1.00262702e+00
4.30395216e-01 9.79518443e-02 -3.20131749e-01 -9.22049209e-03
-3.64035629e-02 -9.06356722e-02 6.02222800e-01 1.05465698e+00
-3.93574119e-01 -1.60778761e+00 -3.44591647e-01 2.95323789e-01
-9.46691930e-01 -3.21583867e-01 -2.93712094e-02 7.29367435e-01
-3.03620636e-01 8.26859772e-01 -3.97626966e-01 -4.39334095e-01
-2.50057161e-01 -2.50161171e-01 6.79293096e-01 -1.50081664e-01
-6.56733871e-01 3.37143689e-02 8.77029896e-02 -3.04898083e-01
-5.03431559e-01 -6.82818055e-01 -9.29316401e-01 -2.06144467e-01
-4.01426792e-01 5.65733194e-01 9.77236867e-01 1.13359642e+00
-5.75498939e-02 5.53872824e-01 1.12681079e+00 -8.34480762e-01
-1.02473629e+00 -5.97678483e-01 -7.35285580e-01 -2.04914346e-01
6.05495334e-01 -2.24795133e-01 -1.24587685e-01 -2.33892754e-01] | [6.433241844177246, 1.5596212148666382] |
3b24f56a-a1ca-4778-a470-2ddebeda81b2 | generative-knowledge-selection-for-knowledge | 2304.04836 | null | https://arxiv.org/abs/2304.04836v1 | https://arxiv.org/pdf/2304.04836v1.pdf | Generative Knowledge Selection for Knowledge-Grounded Dialogues | Knowledge selection is the key in knowledge-grounded dialogues (KGD), which aims to select an appropriate knowledge snippet to be used in the utterance based on dialogue history. Previous studies mainly employ the classification approach to classify each candidate snippet as "relevant" or "irrelevant" independently. However, such approaches neglect the interactions between snippets, leading to difficulties in inferring the meaning of snippets. Moreover, they lack modeling of the discourse structure of dialogue-knowledge interactions. We propose a simple yet effective generative approach for knowledge selection, called GenKS. GenKS learns to select snippets by generating their identifiers with a sequence-to-sequence model. GenKS therefore captures intra-knowledge interaction inherently through attention mechanisms. Meanwhile, we devise a hyperlink mechanism to model the dialogue-knowledge interactions explicitly. We conduct experiments on three benchmark datasets, and verify GenKS achieves the best results on both knowledge selection and response generation. | ['Zhaochun Ren', 'Pengjie Ren', 'Weiwei Sun'] | 2023-04-10 | null | null | null | null | ['response-generation'] | ['natural-language-processing'] | [ 1.80266246e-01 5.36302328e-01 -4.78928715e-01 -4.24875230e-01
-6.87628746e-01 -5.37673831e-01 8.82418394e-01 2.75566373e-02
-2.50612050e-01 1.04875839e+00 7.28084624e-01 -1.40062526e-01
-1.24126546e-01 -8.96378636e-01 -4.78235781e-01 -5.73451281e-01
2.56547630e-01 7.76818514e-01 2.56007165e-01 -6.06018722e-01
2.99068719e-01 -2.68889189e-01 -1.24508011e+00 4.99600202e-01
1.11712098e+00 6.63474381e-01 2.63089210e-01 4.92341667e-01
-6.06662512e-01 9.93655503e-01 -9.36523259e-01 -5.59082210e-01
-3.27505320e-01 -9.99780893e-01 -1.70208955e+00 6.36435673e-02
-3.81236762e-01 -1.78090438e-01 -2.57334471e-01 7.98457682e-01
6.11702979e-01 3.59294116e-01 6.06073439e-01 -1.05802286e+00
-5.75544655e-01 1.41612947e+00 -8.37429310e-04 -6.76051155e-02
7.88619816e-01 2.23834515e-01 1.07272589e+00 -6.16741478e-01
6.54426754e-01 1.58019829e+00 1.23944767e-01 8.81132483e-01
-9.37146127e-01 -3.39152455e-01 2.52462596e-01 4.41165805e-01
-1.23384798e+00 -5.85296631e-01 8.30749989e-01 -2.54675984e-01
9.31717455e-01 3.93963307e-01 7.96920002e-01 1.25032830e+00
-3.47812504e-01 1.23280239e+00 8.11354518e-01 -5.72413087e-01
1.73724383e-01 1.64761350e-01 4.43447202e-01 5.25418282e-01
-1.92886949e-01 -4.09057081e-01 -1.05472624e+00 -4.31064248e-01
6.48000836e-01 -6.68929756e-01 -5.27435243e-01 8.82204995e-02
-1.21001303e+00 9.67150211e-01 -2.62632854e-02 1.17639974e-01
-4.51559007e-01 -2.29709029e-01 5.27301192e-01 3.32333207e-01
1.83961213e-01 5.53474307e-01 -4.11246538e-01 -3.91794264e-01
-9.88836512e-02 4.23382789e-01 1.17001450e+00 1.25360548e+00
8.44067514e-01 -2.56514430e-01 -9.22580898e-01 1.01397836e+00
3.09841275e-01 1.96083546e-01 5.61506510e-01 -7.64309824e-01
6.37620330e-01 9.39704955e-01 3.26433569e-01 -8.59814644e-01
-2.63913721e-01 -3.75262052e-02 -6.14829361e-01 -6.95406020e-01
1.66404173e-01 -5.13534606e-01 -5.08229911e-01 1.70968688e+00
5.63746274e-01 -3.87315489e-02 5.63832700e-01 8.80175889e-01
1.31504834e+00 7.00301945e-01 1.41305730e-01 -4.85105097e-01
1.47174764e+00 -9.38061893e-01 -9.37114120e-01 -1.02590166e-01
6.59631133e-01 -5.15185058e-01 1.21367872e+00 1.58965021e-01
-9.51214135e-01 -1.76990137e-01 -5.73953032e-01 1.30650610e-01
7.45980889e-02 -6.90624863e-02 6.03259921e-01 2.94623822e-01
-6.24034345e-01 2.42928490e-01 -3.12751025e-01 -1.34073943e-01
-3.51308584e-02 1.58934310e-01 1.58642575e-01 3.36708724e-01
-1.91251314e+00 7.96216190e-01 9.44408774e-01 2.47541875e-01
-8.43669772e-01 -3.12995702e-01 -8.37140918e-01 -2.10905131e-02
8.84403646e-01 -7.70647168e-01 1.57712090e+00 -9.87223208e-01
-2.07989097e+00 5.03451347e-01 -3.27060461e-01 -4.00788456e-01
2.49705881e-01 -1.64910048e-01 -2.73972243e-01 8.30870047e-02
-1.12824686e-01 3.63036275e-01 6.30269051e-01 -1.27150548e+00
-6.98512435e-01 1.12805672e-01 4.33632523e-01 8.54800522e-01
-2.20283389e-01 -1.72677096e-02 -7.51075685e-01 -3.43277305e-01
-2.29989827e-01 -8.51502717e-01 -5.47210947e-02 -1.04234838e+00
-1.02546144e+00 -8.25791955e-01 4.43864048e-01 -4.69455987e-01
1.71448195e+00 -1.68080389e+00 4.36692029e-01 1.64350465e-01
7.31983855e-02 4.58086938e-01 6.09317459e-02 7.79249847e-01
5.42392194e-01 1.05988622e-01 2.17680596e-02 1.13841467e-01
1.31256819e-01 3.47194046e-01 -3.07542413e-01 -4.21278149e-01
1.62428126e-01 1.04282510e+00 -1.21510768e+00 -5.92218041e-01
-8.13660100e-02 1.14476949e-01 -4.34221089e-01 6.41854644e-01
-7.52298355e-01 4.77948219e-01 -1.03660250e+00 2.36617655e-01
1.15004994e-01 -4.60701436e-01 5.47778606e-01 -2.54008740e-01
2.47059733e-01 7.47187674e-01 -8.21177840e-01 1.48262525e+00
-4.42664862e-01 1.14086024e-01 -2.21983135e-01 -6.33215129e-01
8.89826953e-01 4.49186742e-01 -1.12537652e-01 -4.17958677e-01
4.65990752e-02 -5.87901659e-02 1.62376165e-01 -7.57455409e-01
7.78856754e-01 -4.14643548e-02 -3.65307122e-01 5.45052290e-01
1.32003799e-01 1.05247065e-01 1.58164203e-01 6.11755311e-01
9.83742058e-01 9.30693001e-02 4.14753944e-01 -1.31862555e-02
6.42066956e-01 1.61992222e-01 6.28013134e-01 8.53645623e-01
1.41844213e-01 9.60359275e-02 7.53417671e-01 -1.76940002e-02
-4.30337489e-01 -5.94355643e-01 3.11376691e-01 1.21407866e+00
3.60220611e-01 -6.35738671e-01 -9.38798547e-01 -1.00806308e+00
-3.05798829e-01 1.05217934e+00 -4.97326285e-01 -5.89197397e-01
-5.60391366e-01 -6.94018483e-01 6.79956436e-01 1.55094787e-01
7.28283882e-01 -1.54347765e+00 -3.02484632e-01 3.93185288e-01
-9.20082808e-01 -9.50261474e-01 -5.99474013e-01 -3.61446775e-02
-2.61300355e-01 -1.15669894e+00 -3.20327282e-01 -6.26951337e-01
5.83257854e-01 1.45537611e-02 1.22558045e+00 1.11836061e-01
4.07621711e-01 4.81823623e-01 -9.41444337e-01 -2.61249185e-01
-6.99267149e-01 5.37716568e-01 -2.03588545e-01 1.98193267e-01
5.58192551e-01 -5.81940971e-02 -5.22502780e-01 4.55171257e-01
-6.27399623e-01 4.15559232e-01 4.19509560e-01 1.17213798e+00
3.34388018e-01 9.62816998e-02 1.02118790e+00 -1.12947941e+00
1.25767362e+00 -6.06902480e-01 -2.02980340e-02 5.62802911e-01
-4.38980103e-01 2.34012827e-01 7.13715553e-01 -2.52061754e-01
-1.55200160e+00 -2.20460787e-01 -1.46603972e-01 4.93472740e-02
-1.42348930e-01 8.82212698e-01 -4.55851436e-01 3.69065434e-01
4.33284014e-01 4.90489513e-01 -1.27421826e-01 -2.57989109e-01
2.92357206e-01 7.44591296e-01 2.76131541e-01 -1.03058696e+00
2.99852818e-01 -9.41067114e-02 -6.63390458e-01 -8.92154813e-01
-1.07402277e+00 -3.79912555e-01 -3.12173009e-01 -2.25469172e-01
6.70778096e-01 -6.38761520e-01 -9.91370440e-01 4.05348718e-01
-1.27513301e+00 -4.11719054e-01 -2.03975439e-01 2.66856253e-01
-5.83444357e-01 3.58186424e-01 -5.24889588e-01 -1.05102277e+00
-4.84828502e-01 -9.50265944e-01 7.45483935e-01 5.14862835e-01
-5.09317398e-01 -9.01451826e-01 -2.09190533e-01 5.47266603e-01
8.92625228e-02 -1.79831535e-01 1.20722628e+00 -1.20590782e+00
-5.30363560e-01 8.69816467e-02 1.50104403e-01 -3.43229342e-03
1.94451839e-01 -7.33342245e-02 -7.59439230e-01 9.76233110e-02
-3.01935345e-01 -7.05885589e-01 6.69768512e-01 -5.88225108e-03
1.02374876e+00 -8.71626794e-01 -2.90467143e-01 -3.84167186e-03
7.47316241e-01 4.72355843e-01 7.03770995e-01 5.61716929e-02
6.65987313e-01 9.08832192e-01 8.88900518e-01 6.04146302e-01
8.86803985e-01 9.60012794e-01 6.96881115e-02 3.31576645e-01
1.91444039e-01 -5.46220899e-01 3.54562283e-01 1.04084039e+00
-2.75208242e-02 -6.37619138e-01 -7.87665844e-01 6.01046622e-01
-2.25809717e+00 -1.02526951e+00 3.59850414e-02 2.07131577e+00
1.66996670e+00 4.87573370e-02 1.05541825e-01 -2.19721928e-01
9.13956940e-01 1.61208902e-02 -4.43991363e-01 -2.92314321e-01
-2.29935914e-01 -1.01357609e-01 -7.20091239e-02 7.13550985e-01
-7.26354718e-01 1.32105422e+00 5.49921989e+00 1.04127765e+00
-5.81346393e-01 -1.70626104e-01 5.36348581e-01 3.07295322e-01
-5.23592114e-01 1.17793940e-01 -1.29279089e+00 6.30370855e-01
7.07530856e-01 -5.37304282e-01 2.48424575e-01 4.84851569e-01
2.32363939e-01 -1.77856758e-01 -9.65013981e-01 5.40362775e-01
-1.39905006e-01 -1.36667907e+00 5.24870276e-01 -4.03506845e-01
5.39863527e-01 -6.23277009e-01 -5.03998280e-01 5.42524576e-01
7.04195142e-01 -8.95772099e-01 3.55037600e-01 7.33494103e-01
3.49045545e-01 -8.38176429e-01 7.27035463e-01 5.08013368e-01
-9.83180881e-01 1.60034612e-01 -1.02998711e-01 2.76321262e-01
2.74991393e-01 1.55724362e-01 -1.37379277e+00 8.64392638e-01
1.73959360e-01 2.71616697e-01 -1.68714195e-01 6.61521196e-01
-7.03238904e-01 9.28131759e-01 6.37782142e-02 -5.16839802e-01
2.95468956e-01 -1.12510063e-02 5.24174511e-01 1.26264286e+00
-5.23394570e-02 4.70532835e-01 5.14917135e-01 7.35638022e-01
-2.24260874e-02 2.98432797e-01 -2.15921387e-01 -2.54596978e-01
1.05163670e+00 1.02100563e+00 -4.06508088e-01 -4.69505221e-01
-1.46754354e-01 9.58594620e-01 2.41260722e-01 4.69731897e-01
-5.19457221e-01 -6.06624782e-01 4.78604347e-01 -2.21848011e-01
1.86405256e-01 1.08900666e-01 1.99519917e-01 -1.02821493e+00
-7.26640299e-02 -1.02918434e+00 5.44766486e-01 -4.80080843e-01
-1.31310976e+00 5.60753644e-01 3.36650014e-01 -1.01548946e+00
-6.27408862e-01 1.07679211e-01 -6.70958996e-01 9.16448951e-01
-1.34344769e+00 -1.05997622e+00 -2.26621926e-01 4.95164156e-01
7.19158530e-01 -1.04958013e-01 9.14844453e-01 -2.11412057e-01
-5.72073162e-01 6.64810658e-01 -3.12160164e-01 3.96830767e-01
5.34007728e-01 -1.17021072e+00 2.17910290e-01 2.59234428e-01
-2.36879244e-01 8.67451608e-01 5.75342178e-01 -8.05220127e-01
-1.40326655e+00 -9.45930183e-01 1.18873012e+00 -3.29564005e-01
3.15023154e-01 -1.07590683e-01 -1.22057986e+00 4.81687009e-01
5.18347263e-01 -7.72889614e-01 9.69366491e-01 4.10624683e-01
-8.77687894e-03 3.92736226e-01 -7.05385745e-01 7.17744112e-01
1.06407273e+00 -4.01359081e-01 -8.75570178e-01 2.62885690e-01
1.10006189e+00 -6.66490376e-01 -7.59732187e-01 2.74180293e-01
2.65362769e-01 -6.70723021e-01 7.45512724e-01 -7.58457601e-01
4.64184403e-01 -2.73798347e-01 1.88132808e-01 -1.56207681e+00
-1.02753967e-01 -1.10356784e+00 -5.12234449e-01 1.68529272e+00
5.96371472e-01 -3.62128556e-01 6.88937843e-01 7.34456301e-01
-1.81221515e-01 -8.37220490e-01 -6.41335368e-01 -5.03203392e-01
-3.25557590e-01 -4.77454253e-02 9.29813504e-01 1.17316020e+00
5.43146253e-01 9.75487828e-01 -7.39431322e-01 -1.34997025e-01
2.24886760e-01 2.04012766e-01 8.29672992e-01 -9.44745898e-01
-3.86755556e-01 -4.59545732e-01 1.05424501e-01 -1.38133383e+00
3.65502983e-01 -7.16557622e-01 2.43858606e-01 -1.60891938e+00
1.00523293e-01 -5.25699675e-01 1.73724648e-02 6.73630297e-01
-7.09714890e-01 -4.95823026e-01 -1.55067608e-01 1.08604901e-01
-9.11111355e-01 1.06469858e+00 1.40800118e+00 -3.77804227e-02
-7.36956596e-01 1.15417756e-01 -9.21739340e-01 5.51964343e-01
8.24671328e-01 -8.64449441e-02 -8.43728840e-01 -1.75151750e-01
3.13656211e-01 4.48477954e-01 1.06803648e-01 -3.12607259e-01
3.48161995e-01 -7.43386209e-01 -1.76626220e-01 -4.29415792e-01
2.98477829e-01 -2.21386954e-01 3.54445400e-03 2.69108973e-02
-8.89428914e-01 -4.70851898e-01 -9.34700519e-02 5.63535869e-01
-3.56959075e-01 -4.36991215e-01 2.61320770e-01 -2.54708648e-01
-9.68892038e-01 1.55056134e-01 -5.37344992e-01 5.08254945e-01
7.30438709e-01 -1.57023836e-02 -2.97843695e-01 -6.25041425e-01
-5.78246057e-01 7.01052010e-01 2.24574078e-02 5.53250551e-01
6.93094492e-01 -1.35019839e+00 -7.12845147e-01 -8.11522752e-02
3.76150221e-01 3.61010253e-01 3.43504637e-01 4.75398183e-01
3.61234546e-02 4.39252645e-01 2.25591615e-01 -2.21701235e-01
-1.32627130e+00 8.89726356e-02 2.37300441e-01 -2.88445354e-01
-4.12655830e-01 1.05954564e+00 1.22639246e-01 -4.90682781e-01
2.85814345e-01 1.26796678e-01 -6.94283724e-01 2.58657247e-01
4.97888774e-01 1.42218739e-01 -1.72808483e-01 -4.93267953e-01
-1.22919984e-01 2.39398107e-02 -3.25155348e-01 -9.93616581e-02
6.58115149e-01 -3.92743587e-01 -5.44478931e-02 5.31317890e-01
7.17683911e-01 -1.86148018e-01 -8.95948887e-01 -7.44937241e-01
1.55775741e-01 -3.06310028e-01 -2.86908150e-01 -9.79467154e-01
-5.04581869e-01 2.98874378e-01 -3.64683330e-01 3.62402141e-01
9.03393447e-01 1.45051837e-01 8.95388126e-01 7.42625356e-01
4.98645753e-01 -1.30576754e+00 1.60636052e-01 9.98910248e-01
9.91545558e-01 -9.34034586e-01 -2.34556019e-01 -6.49275482e-01
-1.35852444e+00 8.57807577e-01 1.09337389e+00 6.04990304e-01
2.68865526e-01 -2.55670398e-01 1.33618549e-01 -3.50013494e-01
-1.20998132e+00 -3.27138066e-01 2.36970469e-01 4.58508283e-01
5.52686989e-01 1.26867309e-01 -6.15451038e-01 8.49246085e-01
-2.14984819e-01 -1.46201700e-01 4.18869346e-01 9.43093359e-01
-4.19535160e-01 -1.24676991e+00 4.17806990e-02 5.30491471e-01
-3.25695798e-02 -1.84037477e-01 -8.68654072e-01 3.78539681e-01
-2.72622734e-01 1.11795962e+00 -3.91285539e-01 -6.87665343e-01
4.10081536e-01 4.13832307e-01 1.49390414e-01 -9.25539255e-01
-1.10241616e+00 2.80137267e-02 7.53954232e-01 -9.40222219e-02
-4.65862542e-01 -4.13117975e-01 -1.53461885e+00 -2.82847166e-01
-7.10350811e-01 9.78802562e-01 6.49708090e-03 1.05367923e+00
5.60826540e-01 5.14906585e-01 7.98921108e-01 -1.59883276e-01
-6.36259675e-01 -9.70752776e-01 -3.39019805e-01 4.14509028e-01
-1.02256499e-01 -5.52056372e-01 8.40879232e-02 -7.75639266e-02] | [12.498544692993164, 8.074708938598633] |
6723f275-08f8-435e-ae51-b6f8e272ad4e | end-to-end-pseudo-lidar-for-image-based-3d | 2004.03080 | null | https://arxiv.org/abs/2004.03080v2 | https://arxiv.org/pdf/2004.03080v2.pdf | End-to-End Pseudo-LiDAR for Image-Based 3D Object Detection | Reliable and accurate 3D object detection is a necessity for safe autonomous driving. Although LiDAR sensors can provide accurate 3D point cloud estimates of the environment, they are also prohibitively expensive for many settings. Recently, the introduction of pseudo-LiDAR (PL) has led to a drastic reduction in the accuracy gap between methods based on LiDAR sensors and those based on cheap stereo cameras. PL combines state-of-the-art deep neural networks for 3D depth estimation with those for 3D object detection by converting 2D depth map outputs to 3D point cloud inputs. However, so far these two networks have to be trained separately. In this paper, we introduce a new framework based on differentiable Change of Representation (CoR) modules that allow the entire PL pipeline to be trained end-to-end. The resulting framework is compatible with most state-of-the-art networks for both tasks and in combination with PointRCNN improves over PL consistently across all benchmarks -- yielding the highest entry on the KITTI image-based 3D object detection leaderboard at the time of submission. Our code will be made available at https://github.com/mileyan/pseudo-LiDAR_e2e. | ['Wei-Lun Chao', 'Yan Wang', 'Divyansh Garg', 'Serge Belongie', 'Rui Qian', 'Kilian Q. Weinberger', 'Yurong You', 'Mark Campbell', 'Bharath Hariharan'] | 2020-04-07 | end-to-end-pseudo-lidar-for-image-based-3d-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Qian_End-to-End_Pseudo-LiDAR_for_Image-Based_3D_Object_Detection_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Qian_End-to-End_Pseudo-LiDAR_for_Image-Based_3D_Object_Detection_CVPR_2020_paper.pdf | cvpr-2020-6 | ['3d-depth-estimation'] | ['computer-vision'] | [ 5.83423488e-03 -1.42266795e-01 -5.04603386e-02 -4.95840073e-01
-7.86754727e-01 -5.48388541e-01 6.01750851e-01 -1.88335422e-02
-7.10946977e-01 1.71922773e-01 -3.83312166e-01 -5.14526486e-01
3.15277338e-01 -8.00814688e-01 -1.14173555e+00 -1.76698357e-01
8.11283439e-02 7.58595347e-01 5.16621828e-01 -1.89920411e-01
3.90225619e-01 1.01609313e+00 -1.83181465e+00 -3.54455374e-02
3.77351046e-01 1.24267018e+00 3.06851357e-01 6.97985351e-01
-3.18072289e-01 3.13880861e-01 -4.86665778e-02 -2.33912960e-01
7.60436535e-01 4.07299817e-01 -2.32578129e-01 -1.74609527e-01
1.10900426e+00 -6.97899759e-01 -4.36965436e-01 8.50286663e-01
5.73788047e-01 -1.43352687e-01 4.82389897e-01 -1.32936203e+00
-3.75562400e-01 8.87667164e-02 -7.32260585e-01 2.08806261e-01
1.48917079e-01 4.25751954e-01 7.70960331e-01 -1.33463240e+00
5.70578456e-01 1.49864292e+00 1.00346398e+00 5.14137447e-01
-1.22047222e+00 -8.64986062e-01 9.73202512e-02 2.63514459e-01
-1.31571341e+00 -5.59794724e-01 7.17161000e-01 -4.41987067e-01
1.57853580e+00 -1.24743566e-01 7.48550773e-01 9.03732121e-01
1.30393803e-01 8.59537780e-01 9.79244649e-01 -1.57203883e-01
3.36915329e-02 8.76862481e-02 -2.88071595e-02 7.21100986e-01
5.18341303e-01 5.01986861e-01 -6.04472995e-01 2.77838051e-01
6.41924679e-01 1.44821942e-01 1.68418735e-01 -1.03981221e+00
-1.08315074e+00 7.84203410e-01 8.04410696e-01 -2.17374057e-01
-1.24668628e-01 5.47238231e-01 3.30771714e-01 1.86561450e-01
5.93694270e-01 1.80162266e-02 -3.49865496e-01 -2.89341602e-02
-8.28473628e-01 6.74442291e-01 4.57004637e-01 1.13850760e+00
1.01795208e+00 -1.59966692e-01 2.99693167e-01 3.38192821e-01
5.05828261e-01 6.92990959e-01 -1.04642294e-01 -1.20330656e+00
5.49641788e-01 7.66132534e-01 2.22192451e-01 -4.49016213e-01
-4.25001800e-01 -3.59525055e-01 -2.67050624e-01 9.33025360e-01
1.87712863e-01 2.83705503e-01 -1.09828043e+00 1.18740952e+00
4.94438350e-01 2.65367627e-01 -7.36499280e-02 8.94046664e-01
1.01792204e+00 4.23455149e-01 -3.37486088e-01 6.25473499e-01
1.06671309e+00 -7.74531007e-01 -4.39561643e-02 -7.27896035e-01
4.81372476e-01 -5.48579514e-01 8.46302748e-01 3.02204549e-01
-1.02631712e+00 -6.58821583e-01 -1.23985350e+00 -7.08751857e-01
-5.62224090e-01 -2.77074445e-02 4.84388471e-01 5.30432165e-01
-1.32340336e+00 4.28114623e-01 -1.15214765e+00 -4.59165126e-01
8.43988240e-01 3.75966251e-01 -4.56271619e-01 -3.27477962e-01
-5.94895661e-01 1.27397859e+00 2.81097978e-01 1.40013695e-01
-1.01088893e+00 -9.06119704e-01 -9.23690975e-01 -3.38284463e-01
2.59195536e-01 -8.86790216e-01 1.61621583e+00 -1.94860116e-01
-1.18951488e+00 1.36875939e+00 -2.17414826e-01 -8.32729220e-01
9.66972470e-01 -6.12617373e-01 2.70344824e-01 6.66602980e-04
3.52727801e-01 1.38541734e+00 6.94430768e-01 -1.18216157e+00
-9.68280971e-01 -5.52730560e-01 1.80046745e-02 2.21226111e-01
2.67400920e-01 -1.70981959e-01 -4.82570082e-01 3.05600315e-01
2.43732989e-01 -8.17776620e-01 -1.72769696e-01 6.57696903e-01
-2.14998648e-01 -4.05853182e-01 1.06094480e+00 -1.38387429e-02
7.73866326e-02 -2.09200311e+00 -2.44619310e-01 -3.93061787e-01
3.30554217e-01 1.69140607e-01 -6.08721040e-02 2.59339064e-01
1.17294364e-01 -1.13044851e-01 -3.40051651e-01 -9.64825153e-01
1.37493536e-01 1.79411426e-01 -4.65667874e-01 7.81434834e-01
5.59148192e-01 1.00944412e+00 -8.21770310e-01 -2.57027715e-01
8.13380063e-01 7.74291635e-01 -4.73833352e-01 7.51782432e-02
-2.71437049e-01 3.08528453e-01 -1.78568840e-01 8.55390549e-01
1.10426652e+00 1.98044658e-01 -5.71621060e-01 7.97679648e-02
-6.10861897e-01 6.62399530e-01 -9.39454854e-01 1.95144379e+00
-5.16937852e-01 8.73782277e-01 1.92419827e-01 -7.24405468e-01
1.07728100e+00 -2.17383370e-01 3.66710246e-01 -4.83752698e-01
2.01773793e-01 4.06656206e-01 -2.86381930e-01 -8.93987343e-02
6.39051199e-01 -8.64451304e-02 1.03654593e-01 3.22690457e-02
-2.28207503e-02 -7.49428511e-01 7.01329038e-02 5.45320846e-02
1.12340963e+00 5.60993910e-01 1.12703979e-01 1.13488836e-02
2.81121403e-01 3.70384097e-01 2.38501579e-01 7.50452816e-01
-3.22957218e-01 7.30780184e-01 2.36880124e-01 -6.47180021e-01
-1.17255032e+00 -1.26034141e+00 -4.53466058e-01 4.97743547e-01
3.36153626e-01 -1.37171552e-01 -2.46613100e-01 -4.29493576e-01
7.44428277e-01 7.58908629e-01 -4.86396521e-01 -9.92227904e-03
-4.89948213e-01 -1.19802721e-01 4.84967500e-01 7.78491735e-01
5.23305655e-01 -6.93885446e-01 -1.14603925e+00 2.07041621e-01
3.46199334e-01 -1.41236329e+00 1.50677994e-01 5.06665170e-01
-1.00954294e+00 -9.13195968e-01 -3.01513493e-01 -3.96226943e-01
3.08317810e-01 7.63132632e-01 1.24937069e+00 -1.83450446e-01
-3.92076194e-01 4.09773707e-01 -1.61819085e-01 -1.06632137e+00
-1.59308031e-01 1.45280644e-01 1.29568502e-01 -4.93374974e-01
8.99436831e-01 -5.26682556e-01 -6.50677800e-01 1.12984672e-01
-5.66083908e-01 -6.57582134e-02 5.70632100e-01 2.49941289e-01
8.09358835e-01 -5.39080739e-01 -9.26393867e-02 -4.79726523e-01
1.16946317e-01 -1.89859271e-01 -1.21577597e+00 -6.43314183e-01
-4.70900863e-01 -1.36911079e-01 1.20213896e-01 4.68975417e-02
-5.31931043e-01 5.08059978e-01 -4.39312279e-01 -9.63121593e-01
-4.49853957e-01 -6.75372453e-03 1.36422455e-01 -3.42857003e-01
6.95971906e-01 6.71046004e-02 2.59127188e-02 -6.73629403e-01
4.27735448e-01 6.50151610e-01 7.22669303e-01 -1.36680916e-01
1.01819599e+00 9.48482752e-01 3.39656651e-01 -7.80906022e-01
-8.68413627e-01 -6.85949624e-01 -9.86526489e-01 -9.55136120e-02
7.47516274e-01 -1.48482466e+00 -6.55828655e-01 5.17693937e-01
-1.40590298e+00 -3.79258066e-01 -3.33233714e-01 3.40696841e-01
-6.36747003e-01 2.20874667e-01 -3.18295568e-01 -9.29430544e-01
-1.74536183e-01 -1.19703996e+00 1.57694733e+00 -3.10019981e-02
1.57780245e-01 -4.04462516e-01 4.02034856e-02 2.45929524e-01
2.43145883e-01 3.86640370e-01 2.48283267e-01 -1.03675857e-01
-1.13485765e+00 -5.83139479e-01 -5.52480638e-01 3.30375910e-01
-2.62912929e-01 3.49022187e-02 -1.33197641e+00 -1.81269273e-01
-2.30076969e-01 -3.16481829e-01 1.43220544e+00 3.41246933e-01
9.19171154e-01 4.20359731e-01 -5.31534135e-01 8.42464268e-01
1.49236262e+00 -3.64265323e-01 3.84454042e-01 5.03780067e-01
7.47719169e-01 3.17476034e-01 6.10404849e-01 1.88622102e-01
8.22337687e-01 5.98613679e-01 1.11258304e+00 7.89776891e-02
-2.91922033e-01 -2.91454732e-01 4.30225492e-01 1.82617739e-01
6.22239336e-02 2.05536764e-02 -1.24219275e+00 7.26487875e-01
-1.59568787e+00 -6.49148166e-01 -4.30804878e-01 2.16473079e+00
3.05628002e-01 6.12933636e-01 1.32990718e-01 5.92029840e-02
4.68415260e-01 8.97729322e-02 -8.99839461e-01 -3.44884753e-01
1.60830468e-01 1.59618095e-01 9.81605411e-01 5.82383037e-01
-1.16972303e+00 1.04795051e+00 5.28953218e+00 3.12055588e-01
-1.07536793e+00 2.55269557e-01 9.78867561e-02 -4.64986920e-01
2.84741558e-02 -1.85841322e-02 -1.44983268e+00 1.89952508e-01
8.69649231e-01 7.70623162e-02 1.23636223e-01 1.21546221e+00
1.25399873e-01 -1.84806973e-01 -1.41115499e+00 1.24597859e+00
-1.48223191e-01 -1.31259227e+00 -2.55857795e-01 1.03714392e-01
3.91171277e-01 1.23945522e+00 -1.93046760e-02 3.95922244e-01
3.03532779e-01 -8.39292884e-01 1.20567334e+00 2.36756027e-01
9.68191981e-01 -6.76239550e-01 6.05901122e-01 6.20814800e-01
-1.25629318e+00 -2.05607880e-02 -8.28183949e-01 -3.15511405e-01
1.18676409e-01 6.87038183e-01 -9.84516263e-01 2.10526928e-01
1.09528196e+00 9.20116603e-01 -5.91726661e-01 1.25523019e+00
-4.29831296e-01 5.07440045e-02 -7.75818706e-01 1.09650018e-02
4.41703111e-01 1.25304749e-02 8.28314543e-01 1.12938511e+00
4.63515490e-01 -3.38806093e-01 4.32742275e-02 1.26211274e+00
-2.50022560e-01 -4.37090188e-01 -1.23559439e+00 4.02913868e-01
5.51399529e-01 1.29376662e+00 -5.68453491e-01 -1.48194566e-01
-5.62731564e-01 6.16551936e-01 5.11023402e-01 -2.47391120e-01
-6.16783738e-01 -3.14468294e-01 1.05981445e+00 4.60496128e-01
6.42850041e-01 -7.17538059e-01 -4.05873626e-01 -8.71264040e-01
1.64428905e-01 -1.31396860e-01 -1.47973403e-01 -9.71465409e-01
-1.18793535e+00 3.72431964e-01 3.70768048e-02 -1.31262946e+00
-1.86653674e-01 -1.06079066e+00 -3.82138968e-01 9.45145965e-01
-2.23628736e+00 -1.03421843e+00 -6.29269779e-01 2.86126971e-01
5.84757030e-01 3.01436812e-01 4.50268775e-01 2.78392226e-01
-1.79290965e-01 2.07280695e-01 -2.21517652e-01 -5.07943556e-02
5.37240326e-01 -1.30995476e+00 1.31700134e+00 8.66814852e-01
-1.39312241e-02 1.92168534e-01 5.07316232e-01 -5.33833802e-01
-1.73412299e+00 -1.26484525e+00 8.06908727e-01 -1.16421127e+00
4.71343815e-01 -8.17129254e-01 -6.52084589e-01 7.64911950e-01
-1.73707843e-01 3.08138102e-01 -9.91605371e-02 -2.15303674e-01
-4.06735063e-01 -2.86499530e-01 -1.04739428e+00 2.93078333e-01
1.39426494e+00 -4.68667805e-01 -4.67906684e-01 3.24227750e-01
8.22469592e-01 -9.21742737e-01 -3.43338996e-01 5.96689880e-01
4.11686480e-01 -1.30504155e+00 1.20237732e+00 -4.49740589e-02
3.48493487e-01 -6.11295402e-01 -2.93450296e-01 -8.45233560e-01
-9.94817987e-02 -2.93443114e-01 -1.87409192e-01 6.57856345e-01
2.22169265e-01 -7.08174765e-01 1.02675867e+00 2.79920965e-01
-6.15310431e-01 -5.24622858e-01 -1.36245036e+00 -9.92051899e-01
1.17906369e-01 -9.08403218e-01 5.45529008e-01 2.95400977e-01
-6.66056573e-01 3.21272194e-01 1.09221488e-01 3.46683294e-01
9.67428803e-01 2.17468053e-01 1.23152184e+00 -1.44959700e+00
1.44452885e-01 -7.22606719e-01 -8.10033381e-01 -1.38974500e+00
3.27730253e-02 -1.06283927e+00 3.08929086e-01 -1.78481174e+00
-2.46280268e-01 -3.55362594e-01 3.13149057e-02 5.32969177e-01
2.45941252e-01 5.71886957e-01 1.89276889e-01 1.25423834e-01
-3.93000335e-01 5.30705214e-01 7.96624303e-01 -1.15352504e-01
-8.85197371e-02 8.73007774e-02 -3.27661872e-01 7.22532332e-01
9.71207380e-01 -6.61713541e-01 -1.29907364e-02 -9.10852969e-01
4.05037791e-01 -3.16507548e-01 8.83843184e-01 -1.33965802e+00
9.73443538e-02 3.22004706e-01 4.49804902e-01 -1.37104964e+00
7.64299572e-01 -8.67794633e-01 -2.88334072e-01 5.44279516e-01
1.23100072e-01 7.71977827e-02 5.23488879e-01 5.63636720e-01
1.07458867e-01 -9.96670350e-02 9.12817240e-01 -3.75825107e-01
-1.01323938e+00 5.15382171e-01 -1.22593097e-01 -2.83175737e-01
8.38515520e-01 -6.27158701e-01 -2.26224497e-01 -6.80919513e-02
-1.42201737e-01 2.95929164e-01 7.44076371e-01 6.35779321e-01
8.40240538e-01 -1.18967772e+00 -8.91375005e-01 3.42508197e-01
3.19766641e-01 7.21087277e-01 -1.23175174e-01 7.31657863e-01
-7.95748055e-01 5.72177529e-01 -1.66334942e-01 -1.21680105e+00
-9.43437874e-01 4.29956555e-01 4.21760648e-01 2.15804130e-01
-9.39019084e-01 1.06293976e+00 6.61629215e-02 -8.72141480e-01
3.19230765e-01 -7.72599936e-01 3.52512479e-01 -2.41448984e-01
3.51475507e-01 3.27345818e-01 3.74218762e-01 -5.33863187e-01
-6.64974570e-01 7.83881783e-01 -6.40636757e-02 8.78373906e-02
1.55347073e+00 -6.24957606e-02 7.32940286e-02 5.49820781e-01
1.08238447e+00 -2.56047368e-01 -1.68283427e+00 -1.67246267e-01
-1.73972204e-01 -6.86192036e-01 4.80878055e-01 -5.52485704e-01
-7.82502532e-01 1.32418048e+00 9.23279107e-01 -6.18594550e-02
5.21531641e-01 1.24382772e-01 6.38250649e-01 7.27911115e-01
5.23604751e-01 -6.59335911e-01 -2.13232845e-01 7.81489968e-01
7.74438679e-01 -1.56707358e+00 2.15867236e-01 -3.25410813e-01
-8.30886438e-02 9.80675220e-01 6.35252416e-01 -5.37614226e-01
4.33439970e-01 3.90791774e-01 1.04086600e-01 -3.24099809e-01
-5.72313011e-01 -5.47185957e-01 9.42091644e-02 7.64616668e-01
9.04139876e-02 -5.19102402e-02 3.88615578e-01 -4.41092588e-02
-2.60024667e-01 1.90722495e-01 4.50119644e-01 1.06318390e+00
-6.78220630e-01 -9.19007838e-01 -3.31714511e-01 2.90417850e-01
6.91196546e-02 -3.83540355e-02 -4.77148324e-01 1.00315380e+00
2.82738149e-01 7.41725028e-01 3.52747887e-01 -2.27453053e-01
6.27994299e-01 -3.40876542e-02 6.73824012e-01 -7.57987380e-01
-1.54789552e-01 -2.45776296e-01 -8.28484893e-02 -9.25054729e-01
-2.57641375e-01 -9.07392263e-01 -1.07749081e+00 -4.78006065e-01
-2.76278704e-01 -5.99479973e-01 1.13913083e+00 5.62066436e-01
5.97414434e-01 9.14088711e-02 2.74766624e-01 -1.72927845e+00
-5.18206418e-01 -8.20888698e-01 -3.90021652e-01 -1.75431103e-01
6.70036435e-01 -8.89424026e-01 -2.71373630e-01 -4.07241076e-01] | [7.805871486663818, -2.5940866470336914] |
aeacc466-1eb4-4e6a-aa32-c5bc058d6588 | towards-end-to-end-optimisation-of-functional | 1610.04079 | null | http://arxiv.org/abs/1610.04079v1 | http://arxiv.org/pdf/1610.04079v1.pdf | Towards end-to-end optimisation of functional image analysis pipelines | The study of neurocognitive tasks requiring accurate localisation of activity
often rely on functional Magnetic Resonance Imaging, a widely adopted technique
that makes use of a pipeline of data processing modules, each involving a
variety of parameters. These parameters are frequently set according to the
local goal of each specific module, not accounting for the rest of the
pipeline. Given recent success of neural network research in many different
domains, we propose to convert the whole data pipeline into a deep neural
network, where the parameters involved are jointly optimised by the network to
best serve a common global goal. As a proof of concept, we develop a module
able to adaptively apply the most suitable spatial smoothing to every brain
volume for each specific neuroimaging task, and we validate its results in a
standard brain decoding experiment. | ['Kristoffer Hougaard Madsen', 'Lars Kai Hansen', 'Albert Vilamala'] | 2016-10-13 | null | null | null | null | ['brain-decoding', 'brain-decoding'] | ['medical', 'miscellaneous'] | [ 2.34396666e-01 1.61051288e-01 4.13742900e-01 -5.55359542e-01
-3.48056823e-01 -4.66548145e-01 7.99476087e-01 1.93845704e-01
-9.08205807e-01 4.00999159e-01 3.27096373e-01 -2.76058614e-01
-3.75720203e-01 -3.91136318e-01 -5.75811386e-01 -5.14385521e-01
-2.03002706e-01 6.76563859e-01 5.26988387e-01 -1.20289363e-01
4.75792080e-01 7.23857224e-01 -1.23391020e+00 4.55244929e-01
8.37882042e-01 8.02004516e-01 6.35497928e-01 2.23147124e-01
1.54348254e-01 2.98592955e-01 -5.14343739e-01 -1.68189302e-01
8.96515325e-02 -3.65905821e-01 -7.23161101e-01 -2.95138031e-01
9.04047564e-02 -4.93746176e-02 -1.91753358e-03 1.00234437e+00
7.44536459e-01 2.72960573e-01 8.15852225e-01 -4.97171849e-01
-1.13086849e-01 7.54410386e-01 -1.68185949e-01 8.31507266e-01
3.16779241e-02 2.35999793e-01 6.89410627e-01 -7.34509766e-01
5.11142015e-01 9.63587463e-01 4.99286830e-01 2.28965685e-01
-1.43748200e+00 -3.53024453e-01 2.04735056e-01 2.38658533e-01
-1.18671846e+00 -7.40033805e-01 4.70404923e-01 -5.96679747e-01
1.17150915e+00 -3.40007871e-01 9.57791686e-01 9.71820891e-01
4.39098686e-01 1.77126661e-01 1.29038167e+00 -2.62315601e-01
5.11594415e-01 -1.56350225e-01 2.05900446e-01 3.97234291e-01
1.88390747e-01 -1.63140967e-01 -2.60395437e-01 -1.14322994e-02
8.52332294e-01 -3.19283783e-01 -1.52139321e-01 -4.84327078e-01
-1.20915806e+00 6.24542594e-01 4.28720802e-01 8.48672032e-01
-7.66208827e-01 8.82780924e-02 5.13686538e-01 1.60853744e-01
3.40220720e-01 3.42322737e-01 -3.96129847e-01 2.17123762e-01
-1.32012355e+00 2.42935672e-01 4.33943301e-01 8.61003846e-02
5.92008948e-01 -2.07833797e-01 -3.48660886e-01 1.00130832e+00
5.14035583e-01 -2.89475739e-01 9.79847014e-01 -6.95107520e-01
1.12832166e-01 5.16421318e-01 -9.95237231e-02 -7.36767828e-01
-1.03939044e+00 -6.65069878e-01 -6.01803482e-01 3.56874079e-01
5.64337730e-01 -1.63557813e-01 -7.64875233e-01 1.77918124e+00
2.81333715e-01 -5.30938543e-02 -4.90686864e-01 1.02742541e+00
2.54769295e-01 -1.73601717e-01 4.62640315e-01 5.26415892e-02
1.41010332e+00 -7.89240122e-01 -3.12720299e-01 -6.09212041e-01
6.61036789e-01 -3.47700536e-01 8.71383369e-01 4.47034955e-01
-1.19739473e+00 -6.03446424e-01 -9.60969031e-01 -7.17983991e-02
-5.30862927e-01 3.26961987e-02 4.75927919e-01 5.80877185e-01
-1.64505303e+00 5.97893894e-01 -8.57297897e-01 -3.65087509e-01
6.28174126e-01 5.96051693e-01 -5.48337877e-01 1.79835394e-01
-1.15909481e+00 1.40735590e+00 6.68097496e-01 3.72796804e-01
-9.30411458e-01 -7.60023713e-01 -5.16938865e-01 1.46401122e-01
1.42890528e-01 -9.86302078e-01 9.73105848e-01 -1.01434481e+00
-1.58784914e+00 1.22845554e+00 -7.53630325e-02 -4.69325155e-01
4.55547959e-01 -7.35640079e-02 -7.73323104e-02 2.37422600e-01
-2.34395658e-04 6.80162489e-01 9.97724771e-01 -7.31159866e-01
-3.05383384e-01 -7.32534170e-01 -1.75970271e-01 2.48777509e-01
-1.07180640e-01 3.57423156e-01 -4.01476681e-01 -5.19971788e-01
1.10758610e-01 -4.73551542e-01 -3.13048661e-01 -3.34724784e-01
-1.31143369e-02 -3.15570265e-01 1.07818432e-01 -8.91463161e-01
8.79651010e-01 -2.02043128e+00 5.77853024e-01 2.72836357e-01
3.04902256e-01 7.74801895e-02 -1.11363009e-01 1.50847649e-02
-5.33044159e-01 -1.21750824e-01 -3.77948046e-01 -3.47178429e-01
5.51529415e-02 -1.62193641e-01 2.06603333e-01 7.60763764e-01
2.08630756e-01 1.09407604e+00 -6.75553262e-01 -1.63867772e-01
2.13088036e-01 6.08414769e-01 -4.51921910e-01 1.17846444e-01
-5.31999543e-02 6.88608766e-01 -3.05836260e-01 3.25297453e-02
4.29690689e-01 -1.14410728e-01 4.85327423e-01 -4.19594347e-02
-2.44614303e-01 5.06888926e-01 -1.04808390e+00 2.07091355e+00
-2.29842886e-01 2.93255657e-01 5.12242496e-01 -1.43192351e+00
7.44906664e-01 3.27688158e-01 5.00635862e-01 -7.81125009e-01
3.87318969e-01 2.46514365e-01 5.21981537e-01 -5.75657666e-01
-1.36139274e-01 -1.69010833e-01 2.83347249e-01 6.04716480e-01
4.14063960e-01 2.03525126e-01 7.75879622e-02 -1.52084038e-01
1.36194122e+00 3.86729389e-01 3.57073069e-01 -5.89405417e-01
5.98469555e-01 -2.23152757e-01 8.61654431e-02 7.83858359e-01
-4.61438447e-01 6.99583232e-01 7.19626009e-01 -4.57285851e-01
-1.09616709e+00 -6.73462808e-01 -2.79646993e-01 1.33573222e+00
-6.20849371e-01 3.80162224e-02 -1.22272742e+00 -6.25945568e-01
-2.38427311e-01 3.93302709e-01 -8.59976053e-01 -1.19952895e-01
-6.49202406e-01 -1.15068889e+00 4.56982493e-01 1.92523763e-01
2.36613572e-01 -1.44677067e+00 -1.03630161e+00 4.67863441e-01
3.66610646e-01 -8.68802965e-01 -7.63808638e-02 5.06114900e-01
-9.38182175e-01 -1.09142947e+00 -9.06762600e-01 -4.55920249e-01
6.21004224e-01 2.13789120e-02 1.22196484e+00 1.59313574e-01
-1.49614558e-01 1.72799751e-01 -1.47892877e-01 -3.46294604e-02
-1.74230024e-01 5.61947584e-01 -1.26676500e-01 1.00985833e-01
1.27761945e-01 -7.94392705e-01 -8.80898058e-01 1.16150454e-02
-1.03135169e+00 -7.01455027e-02 7.31226325e-01 5.74590087e-01
2.48184353e-01 -2.55603522e-01 5.78660548e-01 -6.99442267e-01
9.92384851e-01 -6.87408984e-01 -5.56717455e-01 1.76871836e-01
-3.23223889e-01 9.33281034e-02 6.23948216e-01 -2.53594935e-01
-7.30781376e-01 1.86740950e-01 -3.98673654e-01 -5.60371988e-02
-6.45241380e-01 7.10470557e-01 -3.48389059e-01 -2.00001717e-01
5.78068197e-01 1.69464231e-01 1.31117404e-01 -6.22144401e-01
3.17691982e-01 1.30814597e-01 4.38996613e-01 -3.86766881e-01
2.91089982e-01 2.06307188e-01 -1.08724982e-01 -5.07148683e-01
-5.24424255e-01 -1.40689552e-01 -1.25072873e+00 -4.53898996e-01
9.24843192e-01 -5.52855849e-01 -6.25132918e-01 4.84462500e-01
-1.14770806e+00 -7.66794026e-01 8.68355036e-02 4.13575828e-01
-3.61677498e-01 3.18656474e-01 -1.24589808e-01 -3.94900292e-01
-2.62407213e-01 -1.43630946e+00 9.69612181e-01 7.13786706e-02
-3.12848121e-01 -1.20412588e+00 2.78654873e-01 1.33207053e-01
6.30604088e-01 -6.67650700e-02 9.84033704e-01 -7.98166096e-01
-3.26146148e-02 -1.35451913e-01 -2.74997681e-01 1.55422136e-01
-3.01142573e-01 -3.42274100e-01 -1.11063981e+00 -1.37613088e-01
3.95036697e-01 -4.04849723e-02 1.14419162e+00 6.80932820e-01
1.22982681e+00 2.92809874e-01 -1.53669119e-01 6.60102308e-01
1.18427420e+00 -9.47929174e-02 7.67660379e-01 5.23886442e-01
3.86994660e-01 9.42148447e-01 -1.93173334e-01 3.15369517e-02
3.64133388e-01 7.77341068e-01 5.16906857e-01 8.27319548e-02
-4.71753068e-03 4.63160992e-01 3.40198398e-01 2.93310732e-01
-1.33721277e-01 2.03913763e-01 -1.09176755e+00 4.69635993e-01
-1.75511801e+00 -8.12209308e-01 2.29897499e-02 2.34966540e+00
6.63634121e-01 1.85281619e-01 4.37216610e-01 -1.40457125e-02
4.15919930e-01 2.33009681e-01 -3.55155349e-01 -3.11672986e-01
6.53223470e-02 3.67757499e-01 2.20680416e-01 4.27586526e-01
-8.51086259e-01 8.10316205e-01 7.57507229e+00 5.65513432e-01
-1.26370418e+00 4.25523043e-01 6.02672100e-01 -1.31009832e-01
-9.41007808e-02 -1.94194645e-01 -4.01402742e-01 6.16303504e-01
1.19774103e+00 2.19487950e-01 8.95610154e-01 3.68800133e-01
6.96357787e-01 -4.89073098e-01 -9.94571745e-01 4.14053917e-01
-3.28869708e-02 -1.20318937e+00 -3.06964636e-01 1.70557410e-03
5.49004935e-02 4.26633805e-01 -1.10564373e-01 1.15008108e-01
2.84452997e-02 -1.22101104e+00 7.75614619e-01 8.98516595e-01
3.39984983e-01 -4.75724876e-01 4.02756572e-01 4.77239758e-01
-6.57808006e-01 -9.73434299e-02 -2.64151186e-01 -4.64494005e-02
1.61545843e-01 7.20082879e-01 -5.81242681e-01 1.83493093e-01
5.72047770e-01 3.18771750e-01 -7.40992069e-01 1.25269699e+00
-1.62118375e-01 3.47732097e-01 -3.38995993e-01 2.99352616e-01
1.66171044e-01 -2.70452887e-01 3.75403374e-01 1.43678772e+00
3.07829618e-01 -1.52309418e-01 -2.21403167e-01 1.07748413e+00
1.56132862e-01 3.11810851e-01 -4.51961130e-01 2.33248845e-01
1.78716123e-01 1.62807703e+00 -1.04269373e+00 -1.67050973e-01
-3.09831470e-01 9.65793788e-01 8.58157873e-01 1.90805346e-01
-6.22701883e-01 -1.58844627e-02 3.97459090e-01 2.12629437e-01
3.58936340e-01 -4.69354957e-01 -4.96096283e-01 -9.56189215e-01
4.57152016e-02 -8.13061774e-01 2.13969782e-01 -7.90409684e-01
-8.65058064e-01 6.92985773e-01 3.50028388e-02 -4.52366650e-01
-2.90449351e-01 -7.73431599e-01 -6.69488788e-01 1.45834994e+00
-1.34340906e+00 -8.00786316e-01 -2.05703974e-01 5.64532518e-01
6.82849735e-02 -1.02506086e-01 8.68375421e-01 4.62282866e-01
-8.98177326e-01 5.05956523e-02 -1.94022760e-01 -2.73177028e-01
6.51649594e-01 -1.12083340e+00 3.28158200e-01 8.57476652e-01
-1.07657984e-01 8.53471518e-01 6.45529270e-01 -5.87600589e-01
-1.00764859e+00 -7.26025462e-01 8.17216933e-01 -3.61225516e-01
8.34935427e-01 -4.32138234e-01 -1.12904167e+00 5.03368139e-01
2.76580632e-01 2.95602325e-02 4.53253239e-01 5.37643172e-02
-5.04151210e-02 7.58408457e-02 -1.11607087e+00 2.82177359e-01
1.00366247e+00 -5.35443604e-01 -6.62516654e-01 2.50600070e-01
8.20860714e-02 -2.74148017e-01 -9.41705644e-01 1.06607810e-01
4.21376437e-01 -1.18998849e+00 9.44643378e-01 -5.86941719e-01
1.61389202e-01 -1.98740795e-01 3.67049485e-01 -1.66545141e+00
-7.08960474e-01 -2.11186260e-01 -8.58471096e-02 8.92754138e-01
3.66120338e-01 -5.78061998e-01 4.62214261e-01 5.95755756e-01
-2.93375880e-01 -4.93249983e-01 -1.06312633e+00 -1.57304242e-01
1.04520664e-01 -5.20473003e-01 4.41934735e-01 7.69875526e-01
-9.03481767e-02 3.89812469e-01 3.10098175e-02 6.29148586e-03
4.76454884e-01 -5.11853576e-01 2.41805628e-01 -1.41820991e+00
-3.21409822e-01 -1.03774059e+00 -2.34889388e-01 -5.63997149e-01
3.50158453e-01 -1.18365276e+00 3.49782174e-03 -1.64504790e+00
1.72532037e-01 -1.32994801e-01 -4.90778267e-01 6.17824674e-01
-2.99696445e-01 4.34021801e-01 -1.44360036e-01 1.52892515e-01
-4.01191264e-01 2.34927714e-01 9.71450210e-01 1.92994311e-01
-2.70437032e-01 -2.98147708e-01 -8.65409315e-01 5.68451703e-01
7.65962541e-01 -3.49002868e-01 -2.64679849e-01 -6.42014384e-01
2.76528388e-01 -3.81769270e-01 5.88276803e-01 -1.16861546e+00
8.18652362e-02 2.00302646e-01 7.10876882e-01 9.56589803e-02
9.04048383e-02 -8.49089384e-01 2.28436738e-01 3.73690099e-01
-3.13910067e-01 5.98788410e-02 1.69599622e-01 9.05460343e-02
5.14126122e-02 -4.14168298e-01 9.80651140e-01 -2.01040760e-01
-6.68824971e-01 1.31777465e-01 -4.83563036e-01 -2.55865157e-01
8.60765636e-01 -1.45385012e-01 -9.89073440e-02 6.53729066e-02
-1.16606450e+00 -3.13758850e-02 2.35647187e-01 1.69310287e-01
2.61208385e-01 -9.40612376e-01 -6.39779985e-01 3.68221164e-01
-3.86031002e-01 -3.52301091e-01 3.96384925e-01 1.31600034e+00
-2.82487959e-01 5.28175414e-01 -6.56689107e-01 -4.13531452e-01
-6.88178897e-01 4.53244865e-01 8.87265265e-01 -3.60864043e-01
-8.00453842e-01 5.25534451e-01 1.24149762e-01 -2.20970109e-01
-1.94745660e-02 -2.93900430e-01 -7.53201962e-01 4.09332424e-01
6.25784278e-01 1.55213615e-02 6.48105025e-01 -6.77583516e-01
-2.87129223e-01 1.39807358e-01 9.49271694e-02 -3.66672993e-01
1.81255960e+00 -9.30177122e-02 -4.34742212e-01 2.03535751e-01
7.67357409e-01 -3.15571517e-01 -1.46669436e+00 -1.01476997e-01
2.40361065e-01 8.95934328e-02 4.61446971e-01 -8.37018609e-01
-1.16722691e+00 8.22881877e-01 6.32599771e-01 3.33896995e-01
1.13644528e+00 -1.81075677e-01 -5.12531307e-03 -6.73124492e-02
3.01288188e-01 -1.12126434e+00 -4.08406615e-01 7.19760478e-01
1.00454664e+00 -7.87203908e-01 -1.10697456e-01 1.00211553e-01
-3.78298044e-01 1.17827785e+00 2.54648626e-01 -4.50511307e-01
7.24035501e-01 1.68649077e-01 -1.63495034e-01 -5.71098685e-01
-5.16988635e-01 -1.80409387e-01 5.23704350e-01 6.17033839e-01
7.10567892e-01 -2.70837665e-01 -6.63778663e-01 8.98471892e-01
3.24195921e-02 2.37611309e-01 -4.53922413e-02 5.41462541e-01
-5.73931456e-01 -1.18331790e+00 -4.23796922e-01 6.25918090e-01
-5.40419459e-01 -7.60450438e-02 -2.81547725e-01 5.64281881e-01
1.46671012e-01 3.73734772e-01 1.02063157e-01 1.66189931e-02
4.03689444e-01 6.85729146e-01 6.25443578e-01 -7.03545690e-01
-1.08459902e+00 1.66296497e-01 -1.14860050e-01 -6.98177338e-01
-3.73676807e-01 -9.25443888e-01 -1.18511534e+00 3.60309444e-02
3.36018175e-01 -4.10466224e-01 7.26867378e-01 1.44768012e+00
3.60583454e-01 8.66236567e-01 9.23888981e-02 -1.36280322e+00
-2.12891698e-01 -1.26838934e+00 -5.69948673e-01 3.17897618e-01
-4.31605205e-02 -8.90235245e-01 -1.95826199e-02 -7.78356194e-02] | [14.159839630126953, -2.119623899459839] |
43d385c0-a794-4f10-b27b-1c2ee0eb6ede | nudging-the-envelope-of-direct-transfer | null | null | https://aclanthology.org/W12-1908 | https://aclanthology.org/W12-1908.pdf | Nudging the Envelope of Direct Transfer Methods for Multilingual Named Entity Recognition | null | ['Oscar T{\\"a}ckstr{\\"o}m'] | 2012-06-01 | null | null | null | ws-2012-6 | ['multilingual-named-entity-recognition'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.425469875335693, 3.7895476818084717] |
b59d29fe-b1ed-4e6e-87ea-928fceaa4c34 | nkululeko-a-tool-for-rapid-speaker | null | null | https://aclanthology.org/2022.lrec-1.205 | https://aclanthology.org/2022.lrec-1.205.pdf | Nkululeko: A Tool For Rapid Speaker Characteristics Detection | We present advancements with a software tool called Nkululeko, that lets users perform (semi-) supervised machine learning experiments in the speaker characteristics domain. It is based on audformat, a format for speech database metadata description. Due to an interface based on configurable templates, it supports best practise and very fast setup of experiments without the need to be proficient in the underlying language: Python. The paper explains the handling of Nkululeko and presents two typical experiments: comparing the expert acoustic features with artificial neural net embeddings for emotion classification and speaker age regression. | ['Björn Schuller', 'Florian Eyben', 'Hagen Wierstorf', 'Johannes Wagner', 'Felix Burkhardt'] | null | null | null | null | lrec-2022-6 | ['emotion-classification', 'emotion-classification'] | ['computer-vision', 'natural-language-processing'] | [-4.79268968e-01 1.00735135e-01 2.67579615e-01 -6.77468002e-01
-4.84225452e-01 -4.71928239e-01 4.61265564e-01 2.27609351e-01
-6.59201384e-01 4.07050878e-01 3.05111051e-01 -5.61964393e-01
-4.33713943e-02 -2.27734163e-01 -1.34825319e-01 -5.13559937e-01
-3.31598103e-01 6.15988433e-01 -4.16874811e-02 -2.41804987e-01
9.46159065e-02 4.99540716e-01 -1.98963535e+00 -6.77089989e-02
3.77484024e-01 1.04578364e+00 1.42052710e-01 1.06311166e+00
-2.59523451e-01 5.71437776e-01 -6.68075860e-01 -3.32565993e-01
-2.63863523e-02 1.26354679e-01 -7.64858305e-01 -5.87210715e-01
1.86632231e-01 -7.01064393e-02 -2.82554746e-01 9.00097787e-01
1.13785040e+00 1.13012426e-01 7.93233633e-01 -1.57653129e+00
-4.96479720e-01 9.05498087e-01 5.67536414e-01 3.98689926e-01
7.10756600e-01 -1.24812104e-01 5.83546042e-01 -8.21214616e-01
2.27205753e-01 1.29374158e+00 8.91440332e-01 7.12194443e-01
-8.39881837e-01 -6.96525514e-01 -4.10376996e-01 4.25632119e-01
-1.32101214e+00 -9.14661825e-01 7.57431686e-01 -4.77185309e-01
1.29106474e+00 5.56199014e-01 4.25861031e-01 1.64404929e+00
-3.90547723e-01 6.08970284e-01 9.87000883e-01 -8.27764630e-01
5.22857249e-01 1.04200721e+00 5.19796252e-01 4.26954448e-01
-4.91516620e-01 8.04715827e-02 -8.05298626e-01 -4.33137834e-01
2.27156401e-01 -6.76146328e-01 -9.37840268e-02 -1.20214716e-01
-8.40213656e-01 7.21801400e-01 -3.01289558e-01 5.72937608e-01
-3.00401002e-01 -2.41116300e-01 7.80445337e-01 6.77407861e-01
4.10823315e-01 2.69505680e-01 -9.62059021e-01 -1.03019404e+00
-7.97816515e-01 1.18017465e-01 1.23540580e+00 9.53853190e-01
2.10977331e-01 4.09561038e-01 4.00265694e-01 1.37321770e+00
5.58859527e-01 2.73797363e-01 1.26661706e+00 -6.41802967e-01
-1.60165474e-01 2.78686315e-01 -1.62979126e-01 -2.46360719e-01
-3.58552933e-01 3.39765884e-02 -1.66210935e-01 2.75854290e-01
2.89307207e-01 -3.23923022e-01 -4.29419369e-01 1.35847807e+00
2.59547770e-01 -1.68189302e-03 3.21990669e-01 3.14274400e-01
1.31738245e+00 6.80069745e-01 2.28188217e-01 -1.57910332e-01
1.42991138e+00 -6.28132999e-01 -1.01572824e+00 6.26982078e-02
7.15940833e-01 -8.97456229e-01 1.44301736e+00 5.55590212e-01
-9.95806992e-01 -4.57457960e-01 -1.10731113e+00 -5.54201379e-02
-1.26293826e+00 1.21125139e-01 4.76573557e-01 1.53809488e+00
-1.41852844e+00 5.08074880e-01 -5.56747437e-01 -7.03245401e-01
-2.31197551e-01 5.71699977e-01 -5.23224831e-01 6.51472092e-01
-1.44852781e+00 9.29307699e-01 6.13045990e-01 -3.21638435e-01
-3.95217001e-01 -8.18412364e-01 -1.03230524e+00 2.36280173e-01
-6.85959905e-02 -5.55235893e-02 1.54119456e+00 -7.29457855e-01
-2.14063931e+00 9.56319928e-01 1.64416701e-01 -4.95416135e-01
3.32883239e-01 -1.98058188e-01 -1.01866794e+00 -1.74811035e-01
-6.10700369e-01 5.00583351e-01 7.82151163e-01 -5.97955525e-01
-2.67146230e-01 -4.44183230e-01 -1.94718167e-01 -1.23594385e-02
-9.16695178e-01 5.25280416e-01 -2.46834591e-01 -6.04760230e-01
-5.19948900e-01 -5.15322924e-01 2.90959537e-01 -3.67069900e-01
1.14454823e-02 -5.90132773e-01 1.10277557e+00 -8.80388200e-01
1.51240075e+00 -2.49679565e+00 -2.94547886e-01 -3.69057171e-02
-2.53463387e-01 4.43608195e-01 2.29248136e-01 6.06967270e-01
-4.73604828e-01 4.17025201e-02 7.27983341e-02 -5.08313537e-01
7.30625391e-01 3.34056616e-02 6.23907661e-03 2.81698912e-01
-2.67275125e-01 1.80940151e-01 -5.44337332e-01 -5.04639804e-01
6.23328507e-01 5.81848323e-01 -3.27253163e-01 5.51261663e-01
2.49030158e-01 -1.27912700e-01 1.87011119e-02 4.95725185e-01
6.10092938e-01 7.25551486e-01 -8.37371796e-02 5.57486108e-03
-4.19821858e-01 5.17003119e-01 -1.49071324e+00 1.51156199e+00
-8.17939818e-01 9.50944006e-01 2.38791078e-01 -8.84542704e-01
1.16398382e+00 7.91253269e-01 1.20731875e-01 -1.37293832e-02
3.32028210e-01 3.86825025e-01 -3.13738286e-01 -9.39418018e-01
3.35432619e-01 8.83017257e-02 -5.31247184e-02 2.32866809e-01
7.68705249e-01 -1.91294372e-01 -1.57056272e-01 7.74273723e-02
6.96610093e-01 -2.48924997e-02 6.12132967e-01 -3.85458499e-01
8.46324146e-01 -5.36395907e-01 1.27054200e-01 4.27707672e-01
-6.30781412e-01 3.81594487e-02 3.55367988e-01 -1.85092777e-01
-1.14511037e+00 -9.69295442e-01 -5.14427245e-01 1.57431912e+00
-8.87980402e-01 -5.05914748e-01 -1.21625316e+00 -2.94809461e-01
-2.98154056e-01 9.61197853e-01 -6.84365869e-01 -4.93505411e-03
7.98446164e-02 -2.99920559e-01 8.92995656e-01 4.08533484e-01
1.52119160e-01 -1.43975747e+00 -2.92494684e-01 1.55941052e-02
3.55550706e-01 -1.10432947e+00 -2.28985086e-01 4.59878266e-01
-3.52899253e-01 -4.70477164e-01 -4.91811246e-01 -9.81914163e-01
-9.37004760e-02 -7.29381382e-01 9.65566218e-01 -3.95705163e-01
-4.09021646e-01 1.03176498e+00 -5.24502218e-01 -9.80723381e-01
-6.76203668e-01 1.64730594e-01 4.54331726e-01 -2.14405492e-01
1.06973398e+00 -8.62460315e-01 2.67580077e-02 -1.61339924e-01
-6.76705778e-01 -5.95125973e-01 9.46663916e-02 5.37560225e-01
-4.24368680e-01 -4.10233028e-02 7.77586877e-01 -4.88197327e-01
9.14138615e-01 -4.10755724e-01 -6.28215432e-01 3.92182209e-02
-5.90280950e-01 -6.98162243e-02 3.97597045e-01 -3.16299260e-01
-7.22391367e-01 2.56067246e-01 -8.67434561e-01 -1.11608423e-01
-9.14402902e-01 1.96299136e-01 -5.40956199e-01 -2.98222471e-02
5.04418254e-01 2.31559902e-01 2.37192720e-01 -8.08982134e-01
6.52908385e-02 1.82338881e+00 5.82091987e-01 -2.81756282e-01
3.00019652e-01 -3.43621254e-01 -1.05479932e+00 -1.45390999e+00
8.11711513e-03 -6.85661435e-01 -8.12679529e-01 -5.02532005e-01
7.71985352e-01 -7.78272390e-01 -1.00701666e+00 7.68426001e-01
-8.92131686e-01 -3.89058650e-01 -8.76789019e-02 7.34982431e-01
-5.24280965e-01 2.23839417e-01 -2.88014829e-01 -1.22823155e+00
-4.54965591e-01 -9.61262763e-01 6.68223679e-01 2.61733443e-01
-4.63629752e-01 -1.24309015e+00 2.21364737e-01 1.31099969e-01
4.67732012e-01 -1.87088773e-01 7.33245254e-01 -1.38842773e+00
5.85827291e-01 -4.80492473e-01 4.06740308e-01 8.49707365e-01
-1.62407547e-01 4.70780939e-01 -1.76058519e+00 8.15338418e-02
-2.22643558e-02 -4.95339185e-01 1.49558112e-01 1.94902331e-01
1.38811696e+00 -1.09247714e-01 1.19585976e-01 2.60729909e-01
9.04789805e-01 3.01812202e-01 3.98940474e-01 5.46193480e-01
1.75890267e-01 8.65776360e-01 2.19183654e-01 7.50396371e-01
3.46434265e-01 7.13026166e-01 9.03923810e-02 3.51690203e-01
2.73777902e-01 1.65211514e-01 6.74573839e-01 1.20496154e+00
3.38622153e-01 6.53074607e-02 -9.64249015e-01 4.91151512e-01
-1.35425150e+00 -8.07420611e-01 -2.03728423e-01 2.26955366e+00
8.20397019e-01 -7.76497871e-02 4.99242127e-01 6.26385510e-01
5.58452606e-01 2.08413843e-02 -7.48871872e-03 -1.25100148e+00
1.69865757e-01 3.67345124e-01 1.58588141e-01 7.04457998e-01
-1.12942147e+00 7.02208459e-01 7.08114195e+00 8.16902697e-01
-1.17307913e+00 2.88275808e-01 8.31253380e-02 -8.46502036e-02
2.02499807e-01 -6.23891652e-01 -6.71204507e-01 4.70697463e-01
1.92242932e+00 -3.44926894e-01 5.46532035e-01 1.24590862e+00
1.79294392e-01 8.90014023e-02 -1.17221177e+00 1.06489444e+00
1.33546263e-01 -1.00305486e+00 -4.25075859e-01 -1.83093056e-01
-3.43396991e-01 2.22593248e-01 1.82033852e-01 9.00649846e-01
-1.55033916e-02 -9.62063551e-01 7.92053878e-01 1.63492367e-01
6.41563535e-01 -8.79789352e-01 8.03867459e-01 2.15236485e-01
-7.56826282e-01 -2.51009077e-01 -3.31479192e-01 6.14444092e-02
-2.79192161e-02 3.09081882e-01 -1.16171789e+00 3.34484875e-01
1.17472076e+00 1.09250955e-01 -6.35978401e-01 9.95338142e-01
4.47091788e-01 8.85157526e-01 -5.73688805e-01 -3.84611517e-01
1.32816881e-01 1.67824969e-01 5.47967017e-01 1.74425352e+00
3.64718705e-01 -1.25523165e-01 -3.99698585e-01 5.81909008e-02
3.86957735e-01 6.98485494e-01 -5.61922312e-01 -1.60808533e-01
7.76069045e-01 1.34083247e+00 -1.95765361e-01 -3.12366039e-01
-3.16495597e-01 7.99118042e-01 6.61360472e-02 -6.83198869e-03
-6.02391958e-01 -8.57585728e-01 6.81245744e-01 -5.63399587e-03
2.73373932e-01 -1.97521716e-01 -3.98188606e-02 -6.32765770e-01
-2.48245627e-01 -1.07671964e+00 3.42974037e-01 -8.54795337e-01
-1.08940625e+00 9.40800250e-01 1.52046621e-01 -7.69036233e-01
-4.12227243e-01 -1.10449326e+00 -7.66398251e-01 8.25313687e-01
-7.80733168e-01 -8.23973060e-01 -1.26789585e-01 6.54902756e-01
6.22898757e-01 -7.36427188e-01 1.53194666e+00 5.76816440e-01
-7.00745761e-01 8.08005571e-01 2.95842677e-01 8.17504618e-03
8.84216666e-01 -1.79844916e+00 2.22670138e-01 5.29291853e-02
7.32222423e-02 6.01426423e-01 1.13593495e+00 3.70345749e-02
-9.47076678e-01 -6.15878940e-01 1.16627955e+00 -4.41080183e-01
8.22899938e-01 -8.00822079e-01 -9.13607001e-01 5.13665795e-01
7.32043505e-01 -2.82418072e-01 1.31393099e+00 4.43893313e-01
-1.88834131e-01 -1.91052094e-01 -1.17357647e+00 2.56970465e-01
2.33636945e-01 -8.50455523e-01 -7.98776507e-01 2.86977589e-01
5.61308265e-01 -2.53317446e-01 -1.19148481e+00 1.19082771e-01
7.50994503e-01 -9.73423004e-01 8.28983068e-01 -4.36440110e-01
-2.02869907e-01 1.88631684e-01 -2.11473778e-01 -1.25021064e+00
2.52012789e-01 -7.97528863e-01 -1.30128950e-01 1.69052505e+00
6.53883278e-01 -6.80562854e-01 3.40774387e-01 6.47835195e-01
-1.82466283e-01 -3.71979713e-01 -1.27231622e+00 -7.65461087e-01
2.69514658e-02 -9.43726301e-01 8.42191577e-01 1.06767404e+00
3.34478557e-01 3.13819617e-01 -1.73807397e-01 2.84784615e-01
8.14902037e-02 -9.02874112e-01 7.36283004e-01 -1.41938829e+00
-2.29956150e-01 -5.35143554e-01 -8.17281842e-01 -2.70048976e-01
3.50332469e-01 -4.65236038e-01 -1.29643857e-01 -9.31131124e-01
-5.37029684e-01 -1.71118036e-01 -2.89998084e-01 3.83540034e-01
2.93599099e-01 2.20129211e-02 -2.65307605e-01 -5.20303130e-01
-1.23205647e-01 6.04244769e-01 -7.38518536e-02 1.46797448e-01
-3.09063017e-01 7.53558725e-02 -3.10207933e-01 7.69971907e-01
8.89773428e-01 -4.04279262e-01 -4.35257018e-01 5.87655753e-02
-2.03251362e-01 -3.70411754e-01 1.80285603e-01 -1.37250125e+00
1.47808075e-01 4.50603366e-01 2.14419395e-01 -5.26538610e-01
5.73938906e-01 -8.85243654e-01 -1.40424371e-02 1.23828173e-01
-5.12216687e-01 2.36041412e-01 4.77890402e-01 7.84862339e-02
-2.02714726e-01 -8.82056832e-01 6.92314148e-01 1.78781331e-01
-8.44398737e-01 1.61625855e-02 -9.79056358e-01 -3.47285450e-01
1.09973943e+00 -2.53282577e-01 6.99723139e-02 -6.50272191e-01
-1.16260302e+00 -1.14127181e-01 2.65340209e-01 5.92705548e-01
9.36311930e-02 -1.26230872e+00 -3.99785578e-01 5.03624141e-01
2.74883240e-01 -9.10480738e-01 3.26924324e-01 5.22581935e-01
-5.98857999e-01 5.06939352e-01 -1.97950661e-01 -2.00903326e-01
-1.76864159e+00 5.55641711e-01 3.29078674e-01 3.13135773e-01
-3.34429532e-01 9.43888068e-01 -5.05596399e-01 -1.12017870e+00
7.85865903e-01 2.44043153e-02 -7.58991182e-01 3.05406570e-01
8.42246473e-01 4.81403649e-01 5.61323285e-01 -5.31233132e-01
-5.46941698e-01 -2.17195004e-01 6.53646002e-03 -6.17039979e-01
1.55266619e+00 -2.71573454e-01 1.00669891e-01 1.08407915e+00
1.44850743e+00 2.48575583e-01 -5.30299246e-01 9.17428359e-02
1.35887429e-01 1.80754270e-02 3.60813081e-01 -9.05732095e-01
-4.76253390e-01 9.95568693e-01 1.35305130e+00 6.43678367e-01
8.80000174e-01 -1.04594931e-01 3.90220314e-01 6.03430808e-01
-6.20685406e-02 -1.52146471e+00 -3.66941929e-01 5.37560523e-01
8.88033450e-01 -8.97105753e-01 -3.15254599e-01 -9.74945128e-02
-7.18695998e-01 1.45659065e+00 5.60198724e-01 2.76871741e-01
1.29200768e+00 4.68296379e-01 7.51123965e-01 -2.73244772e-02
-7.08098650e-01 -6.63235337e-02 5.40316589e-02 1.03576458e+00
8.48872900e-01 1.40964836e-01 -1.02046341e-01 9.32274282e-01
-8.39119077e-01 -2.61405557e-01 5.04837275e-01 5.77438593e-01
-3.41155946e-01 -1.37342393e+00 -7.10221112e-01 2.17133969e-01
-6.68808997e-01 -2.10241377e-01 -4.94667888e-01 7.48548150e-01
5.30992113e-02 8.20237517e-01 6.08514026e-02 -3.91818345e-01
3.14202875e-01 7.92307436e-01 1.98840275e-01 -5.18521070e-01
-8.63964200e-01 -3.53031158e-02 4.68411714e-01 -9.59548876e-02
-2.24681243e-01 -9.27667081e-01 -8.48434329e-01 -3.39833528e-01
-2.70281523e-01 5.19603193e-01 1.29205477e+00 7.21039772e-01
1.89833537e-01 4.80616838e-01 6.93683982e-01 -9.22552645e-01
-5.21306872e-01 -1.65823042e+00 -8.70644510e-01 -1.44331142e-01
2.23121867e-01 -5.52083075e-01 -5.90697646e-01 1.52316093e-01] | [13.919981002807617, 6.038158416748047] |
ccb80b81-f43c-4df0-9396-037cd8323641 | deformable-video-transformer | 2203.16795 | null | https://arxiv.org/abs/2203.16795v1 | https://arxiv.org/pdf/2203.16795v1.pdf | Deformable Video Transformer | Video transformers have recently emerged as an effective alternative to convolutional networks for action classification. However, most prior video transformers adopt either global space-time attention or hand-defined strategies to compare patches within and across frames. These fixed attention schemes not only have high computational cost but, by comparing patches at predetermined locations, they neglect the motion dynamics in the video. In this paper, we introduce the Deformable Video Transformer (DVT), which dynamically predicts a small subset of video patches to attend for each query location based on motion information, thus allowing the model to decide where to look in the video based on correspondences across frames. Crucially, these motion-based correspondences are obtained at zero-cost from information stored in the compressed format of the video. Our deformable attention mechanism is optimised directly with respect to classification performance, thus eliminating the need for suboptimal hand-design of attention strategies. Experiments on four large-scale video benchmarks (Kinetics-400, Something-Something-V2, EPIC-KITCHENS and Diving-48) demonstrate that, compared to existing video transformers, our model achieves higher accuracy at the same or lower computational cost, and it attains state-of-the-art results on these four datasets. | ['Lorenzo Torresani', 'Jue Wang'] | 2022-03-31 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Wang_Deformable_Video_Transformer_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Wang_Deformable_Video_Transformer_CVPR_2022_paper.pdf | cvpr-2022-1 | ['action-classification'] | ['computer-vision'] | [ 9.97159258e-02 -3.79451245e-01 -4.25701171e-01 -5.32168634e-02
-5.03503203e-01 -4.70992208e-01 3.63986224e-01 8.70494023e-02
-5.77499330e-01 3.41954738e-01 1.93827331e-01 -4.35201116e-02
-8.77102464e-02 -6.82029486e-01 -7.71733880e-01 -6.36226714e-01
-1.18669324e-01 1.81646988e-01 6.18363142e-01 -3.12613249e-02
3.10027659e-01 3.05371642e-01 -1.65478384e+00 5.06898522e-01
4.27507281e-01 1.32007170e+00 3.32518309e-01 7.33400583e-01
1.37942359e-01 1.29319680e+00 -2.55820006e-01 -3.36461246e-01
2.67315030e-01 -3.97341639e-01 -1.08195138e+00 2.09246367e-01
7.30029404e-01 -5.71628571e-01 -7.60825515e-01 8.09004188e-01
1.89931884e-01 4.38989162e-01 2.62259513e-01 -1.07082820e+00
-7.08747566e-01 1.75878584e-01 -4.32949543e-01 8.06903839e-01
4.48493153e-01 3.20517540e-01 1.05808043e+00 -7.85001040e-01
7.42706180e-01 1.08738673e+00 6.35603666e-01 3.70872825e-01
-1.08108926e+00 -2.68710703e-01 5.68239987e-01 7.84572840e-01
-1.39992976e+00 -5.97920597e-01 5.68069398e-01 -5.14624178e-01
1.13591015e+00 3.72961104e-01 9.05177593e-01 9.98312354e-01
4.07881409e-01 8.26744854e-01 4.05777603e-01 -1.61280379e-01
2.28433549e-01 -5.69287300e-01 -3.97925414e-02 8.08900297e-01
-3.25381666e-01 -2.21034214e-01 -7.41532028e-01 1.44809157e-01
1.02328789e+00 1.60276771e-01 -7.09117830e-01 -5.09175897e-01
-1.46781683e+00 6.56465352e-01 3.80258739e-01 3.54396582e-01
-5.73498547e-01 3.60894740e-01 6.18898332e-01 3.02092344e-01
4.92146999e-01 4.58685756e-02 -4.24493045e-01 -4.80405271e-01
-7.43964374e-01 2.68789381e-01 3.73174965e-01 8.27695906e-01
5.63152909e-01 -4.68408614e-02 -5.65948963e-01 3.98724735e-01
8.72299299e-02 1.52716190e-01 6.19978547e-01 -1.20137525e+00
7.31733084e-01 5.32678723e-01 1.63743168e-01 -1.37587178e+00
-2.17830032e-01 -4.29925099e-02 -7.13211834e-01 -6.83402494e-02
4.69563931e-01 2.10770071e-01 -9.65870678e-01 1.57772660e+00
2.89866596e-01 6.62160456e-01 -3.01340133e-01 1.12784159e+00
5.64261556e-01 5.77839851e-01 7.16419742e-02 -1.46725759e-01
1.10847890e+00 -1.06166685e+00 -5.62775791e-01 -1.14010736e-01
5.41573405e-01 -3.36882144e-01 1.08881104e+00 3.07433575e-01
-1.21571445e+00 -7.95538485e-01 -8.49111915e-01 -2.19981998e-01
-9.25352201e-02 -1.30391017e-01 3.62758070e-01 2.76177526e-01
-1.19205332e+00 8.41424108e-01 -1.13300204e+00 -3.07390720e-01
5.10767579e-01 4.24860984e-01 -3.99561673e-01 6.79695532e-02
-9.72335398e-01 5.00837803e-01 1.27321780e-01 1.81606397e-01
-1.04506242e+00 -6.90608144e-01 -8.13988626e-01 2.36816570e-01
5.47098875e-01 -5.15263677e-01 1.29598606e+00 -1.42085743e+00
-1.54809821e+00 6.49402678e-01 -2.24550545e-01 -7.02488899e-01
5.99965692e-01 -3.35725844e-01 -1.54506028e-01 6.23957634e-01
1.29682228e-01 8.89190018e-01 9.10539746e-01 -6.12853050e-01
-8.74251306e-01 -1.90100521e-01 4.40718561e-01 3.24299604e-01
-3.16694766e-01 7.53587708e-02 -9.90582108e-01 -8.32824111e-01
-7.60393441e-02 -9.42495942e-01 -6.77191019e-02 3.38978171e-01
-3.57947536e-02 -2.46951208e-01 9.63684976e-01 -5.53320467e-01
1.17940354e+00 -2.14912629e+00 5.41154981e-01 -2.08546534e-01
2.57370561e-01 3.57169777e-01 -2.08508447e-01 1.58495098e-01
-3.74244228e-02 -2.14976463e-02 -3.70797478e-02 -1.60801932e-01
-3.10193539e-01 2.33060449e-01 -1.33081153e-01 6.45127237e-01
2.41586775e-01 1.05768704e+00 -1.02509713e+00 -4.46299821e-01
5.51802278e-01 5.05770862e-01 -9.72288787e-01 1.29660949e-01
-1.86070815e-01 5.46398640e-01 -4.54476357e-01 5.04326582e-01
1.46449789e-01 -4.29207861e-01 2.59611636e-01 -2.75779992e-01
2.70255804e-02 2.26783961e-01 -9.72775340e-01 1.82449222e+00
-1.41749769e-01 9.56556559e-01 -1.70739263e-01 -1.32859600e+00
4.09272313e-01 3.75507206e-01 8.85157704e-01 -8.34857166e-01
1.57552451e-01 -2.36422122e-01 -1.22050047e-01 -6.33254111e-01
5.25227010e-01 4.33065206e-01 1.08096890e-01 1.40773818e-01
2.75099147e-02 5.96826136e-01 3.48170221e-01 -1.73359271e-02
1.21219671e+00 3.03485394e-01 5.70146404e-02 -2.80058146e-01
5.80514669e-01 -1.30446523e-01 6.91167116e-01 5.78733802e-01
-3.83445710e-01 6.74773991e-01 4.94819283e-01 -8.09669971e-01
-7.73844123e-01 -7.42743611e-01 1.86671004e-01 1.17193019e+00
4.67540652e-01 -5.89517355e-01 -8.96439612e-01 -7.27041841e-01
-2.12385297e-01 1.16559461e-01 -9.32034433e-01 -2.05333829e-01
-8.78908813e-01 -1.31519303e-01 1.68768510e-01 7.94212997e-01
6.35931909e-01 -1.21481752e+00 -1.17919099e+00 3.27767670e-01
-4.70491081e-01 -1.17658639e+00 -9.41865504e-01 -1.35119185e-01
-7.75655568e-01 -1.29709101e+00 -7.56268144e-01 -7.71275699e-01
5.09506643e-01 5.51801443e-01 1.10368407e+00 1.65846616e-01
-1.17927410e-01 5.37846386e-01 -6.78209841e-01 3.85900706e-01
1.11852042e-01 -2.58001033e-02 -1.34221777e-01 4.56748486e-01
2.24379703e-01 -3.54754150e-01 -8.46936226e-01 5.85889220e-01
-7.52239406e-01 3.44146825e-02 1.31939590e-01 9.01028574e-01
7.39582539e-01 -6.19331449e-02 1.36437178e-01 -4.51083452e-01
1.53662255e-02 -3.85070145e-01 -4.98867452e-01 2.01037034e-01
-9.02108923e-02 -1.33202299e-01 6.41709149e-01 -6.79292142e-01
-5.16267180e-01 1.44296840e-01 1.76998556e-01 -1.12473524e+00
5.10908067e-02 4.72434700e-01 -4.40552719e-02 -1.00800864e-01
2.22525969e-01 3.77716780e-01 -6.52135611e-02 -4.15923655e-01
1.34425581e-01 2.99359769e-01 6.52445436e-01 -3.39654148e-01
2.74369210e-01 5.73955834e-01 -3.13229769e-01 -7.63516843e-01
-6.14243209e-01 -3.88701618e-01 -8.91817510e-01 -4.36770886e-01
1.22142005e+00 -7.89154053e-01 -9.47649181e-01 6.13255441e-01
-9.03030097e-01 -6.59133792e-01 -2.66979814e-01 3.77013296e-01
-7.68263996e-01 4.73255455e-01 -6.34263873e-01 -3.04531783e-01
-1.57691881e-01 -1.36489403e+00 1.07091594e+00 -1.93237513e-02
-1.55449152e-01 -8.70166719e-01 -1.81080490e-01 3.45340401e-01
2.68391937e-01 1.60411224e-01 5.80803454e-01 -1.33081526e-01
-8.30094695e-01 6.17747344e-02 -1.37576193e-01 1.75102726e-01
1.64987981e-01 -1.62005480e-02 -6.71967387e-01 -4.88303006e-01
-1.85816884e-01 -1.56953231e-01 8.29088330e-01 5.11777401e-01
1.65647376e+00 -4.10279870e-01 -2.16322318e-01 7.80841112e-01
1.22450399e+00 4.43367779e-01 7.75797307e-01 4.31360155e-01
8.46887350e-01 2.87432373e-01 6.31887674e-01 5.01524210e-01
3.94682050e-01 1.06178045e+00 6.72914505e-01 7.95600787e-02
-1.00989282e-01 -1.59504265e-01 4.48301077e-01 5.45143604e-01
-4.01392758e-01 -5.00904739e-01 -7.03259945e-01 6.67114377e-01
-2.14491439e+00 -1.26172435e+00 9.50661302e-02 2.31194544e+00
4.48548943e-01 4.08465452e-02 3.34529251e-01 1.96288526e-01
7.08131611e-01 4.23631638e-01 -6.60232484e-01 2.53262110e-02
1.05458274e-01 9.45011675e-02 4.62744802e-01 3.39720905e-01
-1.35700691e+00 7.87813604e-01 6.35321999e+00 6.38209343e-01
-1.35948420e+00 1.28090784e-01 7.85258949e-01 -4.38495129e-01
1.77588835e-01 -1.74483150e-01 -4.54688996e-01 7.53652334e-01
8.89998853e-01 4.43270132e-02 4.29650754e-01 5.66262126e-01
3.73574018e-01 3.22472714e-02 -1.22804010e+00 1.04583645e+00
1.46720745e-02 -1.68025827e+00 1.32454529e-01 7.26518258e-02
3.72770637e-01 3.83650474e-02 2.53269654e-02 2.16856211e-01
-1.17041014e-01 -7.98853576e-01 1.23806071e+00 5.17831683e-01
6.50745869e-01 -6.89262152e-01 4.91651624e-01 -1.26938894e-02
-1.62763250e+00 -3.59464824e-01 -3.32868248e-01 -1.21595852e-01
1.56658426e-01 -5.88103831e-02 -1.07112199e-01 3.86060297e-01
1.11776125e+00 1.17807651e+00 -4.68925565e-01 8.20420861e-01
1.29715949e-01 5.14862001e-01 1.15249045e-02 1.86759457e-01
5.47255993e-01 -5.65313129e-03 4.48522002e-01 9.53443766e-01
2.84022391e-01 2.34017909e-01 3.83757114e-01 3.66044343e-01
6.31934479e-02 -5.15409783e-02 -3.63701761e-01 7.11652488e-02
1.69038028e-01 9.01966035e-01 -7.43349850e-01 -4.43430871e-01
-6.59848630e-01 1.13373804e+00 4.99003649e-01 2.81925619e-01
-1.15290976e+00 -8.24046209e-02 9.00611341e-01 4.42580193e-01
8.53715599e-01 -2.32747644e-01 3.58597517e-01 -1.16659105e+00
1.54844627e-01 -8.92680526e-01 5.19218862e-01 -7.50389934e-01
-7.90267169e-01 5.93671083e-01 -5.31010553e-02 -1.51016116e+00
-1.42056003e-01 -5.41085601e-01 -4.05949950e-01 3.41976106e-01
-1.21751440e+00 -7.80501544e-01 -4.56625134e-01 9.17992830e-01
1.02627480e+00 7.62469992e-02 4.70501035e-01 5.22210419e-01
-7.48628139e-01 6.09136701e-01 -6.62492588e-02 3.18757266e-01
4.78849471e-01 -9.65628982e-01 4.37690467e-01 7.24048316e-01
3.13247204e-01 2.52893716e-01 3.31300914e-01 -3.83036166e-01
-1.48966622e+00 -1.22203481e+00 5.73674381e-01 -4.21121180e-01
5.90058208e-01 -2.01345459e-01 -1.00750661e+00 7.97054172e-01
2.67769486e-01 4.52881902e-01 3.17379981e-01 -2.89904237e-01
-1.65045306e-01 -1.50384009e-01 -7.95952022e-01 4.71503615e-01
1.21783435e+00 -4.89593446e-01 -2.57183701e-01 2.62255818e-01
7.13211656e-01 -5.83989501e-01 -9.28692341e-01 2.39114970e-01
6.24358714e-01 -1.15997720e+00 9.30973411e-01 -7.43774712e-01
5.89797914e-01 -3.46913457e-01 -2.15474203e-01 -9.38535511e-01
-6.18800223e-01 -7.18386471e-01 -2.17218310e-01 9.19386268e-01
3.72555293e-02 -4.42804724e-01 8.63869667e-01 4.64704394e-01
-1.61606699e-01 -9.38864112e-01 -1.13135719e+00 -6.48526847e-01
-3.80145073e-01 -2.77713448e-01 4.33951080e-01 9.60678756e-01
-1.30191728e-01 3.35230604e-02 -4.32926089e-01 6.12400612e-03
2.05936760e-01 6.51991665e-02 5.73493361e-01 -8.25498044e-01
-4.53142405e-01 -4.66330290e-01 -9.19376493e-01 -1.47672987e+00
1.25712022e-01 -5.61178148e-01 -9.86671820e-03 -1.18440890e+00
1.60272971e-01 -1.84604585e-01 -4.91097301e-01 7.12347448e-01
-1.98072791e-01 5.10967612e-01 3.42305511e-01 3.97247404e-01
-9.90008891e-01 6.08894944e-01 1.28289354e+00 -2.69610792e-01
-1.36589453e-01 -2.94273049e-01 -3.57590020e-01 6.33568108e-01
5.55829465e-01 -1.59422413e-01 -5.45693398e-01 -7.86395669e-01
-2.31649831e-01 2.92679459e-01 6.19308054e-01 -1.18419921e+00
2.22907871e-01 -3.65503281e-01 3.11821580e-01 -4.57284480e-01
3.67467165e-01 -7.49626815e-01 1.89680666e-01 4.52807903e-01
-4.58639175e-01 3.79711866e-01 1.38862967e-01 7.82244325e-01
-2.52758741e-01 1.40685976e-01 8.31236959e-01 -6.39644712e-02
-1.07484269e+00 6.87414646e-01 -5.14230490e-01 1.19278401e-01
1.29039359e+00 -4.73215938e-01 -1.30543604e-01 -4.22161639e-01
-6.58051014e-01 2.10335448e-01 6.01472497e-01 6.75654352e-01
4.93731976e-01 -1.37349594e+00 -4.06913400e-01 1.40870139e-01
-1.93671640e-02 2.81700008e-02 5.04585624e-01 9.94580686e-01
-6.54615104e-01 5.25098801e-01 -3.62364620e-01 -8.88983130e-01
-1.26846504e+00 7.70529449e-01 5.28032660e-01 -1.90233305e-01
-9.68795896e-01 7.52696037e-01 3.63646895e-01 2.79200166e-01
3.04672539e-01 -5.72417498e-01 -2.10762978e-01 2.18883231e-02
5.97899675e-01 2.72229433e-01 9.32764858e-02 -9.02477980e-01
-4.02960777e-01 7.75398195e-01 -5.58642261e-02 1.98268905e-01
1.22581685e+00 -1.29265442e-01 2.35272229e-01 2.59748966e-01
1.15642285e+00 -4.37852830e-01 -1.83145595e+00 -1.37166962e-01
-1.97505593e-01 -9.32540596e-01 1.62635311e-01 -2.08592728e-01
-1.57073569e+00 6.86330020e-01 5.73678851e-01 3.03026557e-01
1.16266382e+00 -1.45333558e-01 8.07958961e-01 2.12918624e-01
3.53441536e-01 -9.45400596e-01 3.49441081e-01 3.81647378e-01
7.85507381e-01 -9.30207014e-01 -1.87730819e-01 -1.71773329e-01
-6.85452402e-01 9.14080739e-01 7.33024120e-01 -2.10280508e-01
3.80727917e-01 7.41674453e-02 -1.50373951e-01 -1.34305924e-01
-9.73295867e-01 -1.46711944e-02 2.72794425e-01 5.17452836e-01
3.92702878e-01 -3.01437706e-01 2.62222122e-02 8.53950679e-02
2.37165645e-01 1.20659530e-01 2.84348369e-01 1.04313910e+00
-3.60482752e-01 -8.14493179e-01 -5.73746040e-02 4.39874232e-01
-4.91913617e-01 1.05256498e-01 -1.60788909e-01 7.20420897e-01
-2.55886260e-02 8.28683317e-01 5.44520259e-01 -3.91197503e-01
2.67005146e-01 -2.21917942e-01 5.50751328e-01 -3.22144032e-01
-6.03393078e-01 -5.50372340e-03 -1.00931235e-01 -1.12341189e+00
-7.29687512e-01 -7.36415029e-01 -1.02309108e+00 -2.81042069e-01
-7.61105046e-02 -1.38833262e-02 -4.65439968e-02 9.72241700e-01
7.37389326e-01 6.36767268e-01 4.83074158e-01 -1.24360502e+00
-3.10710430e-01 -6.47399902e-01 -2.92121112e-01 4.85704035e-01
4.42411572e-01 -8.74001980e-01 1.22146970e-02 3.87219101e-01] | [8.857187271118164, 0.4478188753128052] |
37f266c2-16cb-4b3a-bfb1-c11b752f5a9b | occlusion-robust-face-recognition-based-on | 1908.06290 | null | https://arxiv.org/abs/1908.06290v1 | https://arxiv.org/pdf/1908.06290v1.pdf | Occlusion Robust Face Recognition Based on Mask Learning with PairwiseDifferential Siamese Network | Deep Convolutional Neural Networks (CNNs) have been pushing the frontier of the face recognition research in the past years. However, existing general CNN face models generalize poorly to the scenario of occlusions on variable facial areas. Inspired by the fact that a human visual system explicitly ignores occlusions and only focuses on non-occluded facial areas, we propose a mask learning strategy to find and discard the corrupted feature elements for face recognition. A mask dictionary is firstly established by exploiting the differences between the top convoluted features of occluded and occlusion-free face pairs using an innovatively designed Pairwise Differential Siamese Network (PDSN). Each item of this dictionary captures the correspondence between occluded facial areas and corrupted feature elements, which is named Feature Discarding Mask (FDM). When dealing with a face image with random partial occlusions, we generate its FDM by combining relevant dictionary items and then multiply it with the original features to eliminate those corrupted feature elements. Comprehensive experiments on both synthesized and realistic occluded face datasets show that the proposed approach significantly outperforms the state-of-the-arts. | ['Lingxue Song', 'Wei Liu', 'Zhifeng Li', 'Dihong Gong', 'Changsong Liu'] | 2019-08-17 | null | null | null | null | ['robust-face-recognition'] | ['computer-vision'] | [ 1.79938868e-01 -2.31898762e-02 1.33127317e-01 -5.43094456e-01
-5.51924482e-02 -6.22915551e-02 4.89130586e-01 -4.39188451e-01
-2.67939597e-01 4.42102998e-01 1.40194073e-01 2.92870939e-01
-2.30748266e-01 -6.49348676e-01 -6.80553794e-01 -9.00854111e-01
1.16995558e-01 2.46851027e-01 -1.86625928e-01 -1.17633611e-01
9.52459779e-03 8.89046788e-01 -1.76084733e+00 2.70848513e-01
5.90048969e-01 1.18446624e+00 -1.57555014e-01 -3.23255360e-01
-3.30284476e-01 2.85898834e-01 -4.61959362e-01 -4.19889867e-01
8.59276056e-01 -4.04561371e-01 -3.39632660e-01 3.88017714e-01
8.90493989e-01 -2.38463014e-01 -5.71096480e-01 1.25574231e+00
3.95309061e-01 1.08270243e-01 5.56942761e-01 -1.32028759e+00
-5.97420633e-01 1.50617167e-01 -9.16142166e-01 1.89430654e-01
1.34026796e-01 -1.54610379e-02 4.79438990e-01 -1.38305771e+00
7.88517118e-01 1.54786468e+00 4.88806039e-01 5.47341168e-01
-1.18106532e+00 -7.37347722e-01 2.95795411e-01 2.30275035e-01
-1.79051876e+00 -7.98547685e-01 1.08783281e+00 -3.37316453e-01
4.43686366e-01 1.61638066e-01 4.99338120e-01 7.08518207e-01
1.06785916e-01 5.74553311e-01 7.84651041e-01 -3.24032009e-01
8.86686798e-03 -7.44121745e-02 -2.19911054e-01 9.52795446e-01
2.73662686e-01 3.34399529e-02 -6.16982579e-01 -2.20081046e-01
7.56441593e-01 3.45038384e-01 -3.40062261e-01 -6.83465660e-01
-8.92632425e-01 6.82843506e-01 3.59673172e-01 3.91665995e-01
-5.32675207e-01 -2.22580120e-01 6.63179755e-02 3.50320160e-01
3.77343625e-01 -1.96192011e-01 -3.56170446e-01 6.71627700e-01
-1.20423508e+00 1.57486796e-01 6.82345033e-01 6.98906779e-01
1.12668025e+00 3.01923424e-01 -2.52519161e-01 9.82564926e-01
2.44263977e-01 3.17393929e-01 2.17879266e-01 -8.13199639e-01
2.87009776e-01 7.17980921e-01 -1.76162019e-01 -1.42402983e+00
-1.65572032e-01 -6.43302083e-01 -1.06999350e+00 3.41383547e-01
3.96870255e-01 -1.08494714e-01 -9.98587668e-01 1.76296580e+00
6.58851564e-01 3.62300992e-01 -2.32426852e-01 8.82498503e-01
8.80149305e-01 3.01425993e-01 -8.83658528e-02 -4.77775097e-01
1.17926741e+00 -7.88090289e-01 -7.67611325e-01 2.46371012e-02
1.38023883e-01 -7.97555327e-01 4.40894514e-01 2.49983504e-01
-1.01414287e+00 -8.38073015e-01 -9.90875959e-01 6.68135434e-02
-2.22912923e-01 3.21703106e-01 2.55661786e-01 5.44473827e-01
-9.44119930e-01 5.47619343e-01 -6.56261146e-01 -2.67929044e-02
1.01051164e+00 6.42342091e-01 -8.62971365e-01 -3.03041011e-01
-7.96225846e-01 6.40700042e-01 -1.04216427e-01 6.28351212e-01
-1.02472746e+00 -7.50151277e-01 -8.46834779e-01 2.49446169e-01
4.55963135e-01 -3.65889519e-01 4.03034091e-01 -1.47004521e+00
-1.18884730e+00 7.29650378e-01 -3.71955663e-01 -7.60915950e-02
5.36296248e-01 1.24736419e-02 -4.78445739e-01 -3.23361680e-02
-2.23941915e-02 7.79854298e-01 1.41875076e+00 -1.12805176e+00
-2.73603976e-01 -5.54345131e-01 -2.91640013e-01 1.08935364e-01
-3.97870481e-01 4.31734473e-02 -4.56389844e-01 -6.71240628e-01
6.33109391e-01 -7.02553749e-01 -7.33594820e-02 5.41287959e-01
-1.84986398e-01 -2.37769708e-01 1.22933960e+00 -6.33067966e-01
7.85920560e-01 -2.38123512e+00 2.49853715e-01 3.97686124e-01
1.36638150e-01 4.71119910e-01 -5.26720464e-01 1.94440782e-01
-4.36039209e-01 -3.87566179e-01 -4.36750233e-01 -6.00972235e-01
-1.46549776e-01 3.25144142e-01 -1.58597305e-01 7.62048483e-01
4.75386113e-01 5.04241109e-01 -6.09858096e-01 -4.85517532e-01
2.49235943e-01 8.53359520e-01 -7.15174139e-01 -6.62352517e-02
-9.66181085e-02 6.83199883e-01 -3.06686193e-01 7.83321440e-01
1.38883781e+00 2.73384124e-01 1.82431534e-01 -3.86523336e-01
-1.44470155e-01 -2.79603720e-01 -1.60458767e+00 1.62808335e+00
-5.65439537e-02 5.43736756e-01 3.47910553e-01 -1.10490155e+00
1.09318852e+00 3.13808024e-01 7.28301883e-01 -5.27620673e-01
4.08762932e-01 2.44463235e-01 8.99368152e-02 -4.28448886e-01
-1.86825275e-01 5.68789281e-02 6.31137133e-01 9.39056873e-02
4.05817956e-01 2.50280678e-01 8.51067007e-02 -1.47950053e-01
7.60059059e-01 8.87382925e-02 1.10779315e-01 -4.85813171e-01
1.16131818e+00 -6.31456614e-01 8.45903575e-01 2.72138178e-01
-3.96014690e-01 7.23988652e-01 4.82749671e-01 -9.41507339e-01
-6.02714777e-01 -8.89926195e-01 -2.78601795e-01 6.90378308e-01
1.19658217e-01 -6.16088323e-02 -9.05698359e-01 -8.14815402e-01
2.10350499e-01 -4.16869409e-02 -7.31459022e-01 -2.24230886e-02
-7.02553511e-01 -4.91019279e-01 1.92452312e-01 6.17725141e-02
8.06962430e-01 -1.26956081e+00 -8.69205818e-02 4.37497608e-02
5.74145280e-02 -9.86250699e-01 -4.46048617e-01 -1.91740826e-01
-4.71783280e-01 -1.08224666e+00 -6.68342054e-01 -1.12975776e+00
1.01126754e+00 4.20006663e-01 6.38745010e-01 3.23394358e-01
-4.53007758e-01 -6.17216416e-02 1.06302366e-01 -8.83876309e-02
5.76605238e-02 -3.79216611e-01 1.76567152e-01 8.86994898e-01
3.96539450e-01 -7.64366925e-01 -6.26926482e-01 3.75691533e-01
-9.10094380e-01 -2.81819224e-01 5.09342849e-01 8.64773929e-01
6.96251273e-01 2.07406417e-01 4.19070870e-01 -6.06040478e-01
1.66260868e-01 -4.39464569e-01 -6.17156684e-01 2.20998392e-01
-4.56776656e-02 3.86664160e-02 5.04871070e-01 -5.59780657e-01
-9.92488027e-01 4.25431222e-01 -1.04884982e-01 -7.77346671e-01
-1.62783191e-01 7.05251247e-02 -6.68194115e-01 -5.85041344e-01
2.14749366e-01 1.33391097e-01 1.25713244e-01 -6.11653745e-01
1.50677562e-01 2.16860682e-01 5.19520164e-01 -3.48544627e-01
1.00366235e+00 9.24045920e-01 3.56798857e-01 -8.38240802e-01
-5.52803338e-01 -1.37820959e-01 -6.76473081e-01 -3.21097970e-01
6.38913631e-01 -7.19580948e-01 -7.52153993e-01 4.47495639e-01
-1.13535058e+00 3.33754718e-01 -3.00295830e-01 3.33430648e-01
-1.28695101e-01 4.04895753e-01 -1.40439272e-01 -6.36450887e-01
-8.09656382e-02 -1.32555676e+00 9.01248634e-01 3.33595663e-01
1.25987023e-01 -4.66768593e-01 -1.43845782e-01 5.74477948e-02
4.03977484e-01 2.56806672e-01 7.21864820e-01 -4.78616625e-01
-6.16703391e-01 -2.83732265e-01 -1.74421430e-01 5.74908912e-01
2.92553395e-01 4.06422615e-02 -1.01156223e+00 -2.94348121e-01
3.56370777e-01 8.47914442e-02 9.53761935e-01 3.25622320e-01
1.20823240e+00 -2.85723478e-01 -3.25431198e-01 8.62030149e-01
1.29155242e+00 1.22996114e-01 7.49708474e-01 -1.60836205e-01
7.01029241e-01 7.79313684e-01 1.71419680e-01 3.87201130e-01
-1.21265002e-01 6.29867435e-01 5.84515572e-01 -1.47554681e-01
-3.29529971e-01 -1.22150518e-01 1.30975246e-01 4.30787951e-01
7.44379461e-02 2.88350731e-02 -3.82016510e-01 5.19884229e-01
-1.50944197e+00 -8.35326433e-01 2.23236054e-01 2.12323856e+00
4.56225604e-01 -1.56713963e-01 -3.77683192e-01 2.24128962e-01
8.47460866e-01 4.00618583e-01 -4.61705327e-01 -5.37442118e-02
-4.26317751e-01 6.76639318e-01 2.56942343e-02 4.75930423e-01
-1.10371721e+00 8.00160825e-01 5.08080626e+00 9.77765679e-01
-1.04894197e+00 1.48096025e-01 5.83029389e-01 -1.10267147e-01
-1.40274078e-01 1.42643541e-01 -7.90172398e-01 4.24692214e-01
2.61674762e-01 2.53771067e-01 4.15121526e-01 7.42254019e-01
9.11400244e-02 1.51520818e-02 -9.57844734e-01 1.00749886e+00
4.06626850e-01 -1.06450427e+00 3.99221301e-01 4.47912998e-02
1.01280928e+00 -3.47362190e-01 3.81816238e-01 3.72283533e-02
-2.37925291e-01 -1.08336532e+00 5.99683642e-01 7.60304451e-01
5.19694388e-01 -7.81969547e-01 6.71784401e-01 7.86876380e-02
-1.21158850e+00 -1.63584262e-01 -5.92949986e-01 1.99262500e-02
-2.14044660e-01 6.16196930e-01 -2.65174925e-01 4.50609326e-01
5.84837556e-01 5.93975365e-01 -3.41578573e-01 9.47551250e-01
7.99242333e-02 2.77042240e-02 -4.50787663e-01 5.28009355e-01
1.56117275e-01 -5.30470729e-01 5.56146860e-01 6.85460687e-01
3.63713175e-01 5.55602647e-02 -1.08327232e-01 9.45598245e-01
-4.24560010e-01 2.82077760e-01 -5.12458146e-01 3.82396668e-01
3.98442268e-01 1.41474962e+00 -6.18201315e-01 -2.15434924e-01
-5.97829401e-01 9.30791616e-01 3.05101663e-01 3.91809583e-01
-6.10449135e-01 -3.03161517e-02 8.79919887e-01 1.90098044e-02
6.64368331e-01 -1.70666605e-01 -1.12510450e-01 -9.44927514e-01
2.93613404e-01 -9.27973032e-01 1.02369249e-01 -3.00039619e-01
-1.12791276e+00 8.25054705e-01 -1.87118530e-01 -1.26402068e+00
3.59283566e-01 -4.53550160e-01 -5.93196213e-01 8.91873002e-01
-1.53030968e+00 -1.02614224e+00 -4.53313977e-01 8.57143044e-01
5.94286025e-01 -3.51822674e-01 6.14880204e-01 6.17900491e-01
-7.05643177e-01 7.26226151e-01 -2.43768040e-02 2.01224700e-01
5.68192244e-01 -2.40406096e-01 5.45462370e-02 8.11176240e-01
2.64563441e-01 7.21708894e-01 4.76220638e-01 -5.25663912e-01
-1.23129487e+00 -1.13704169e+00 1.00507700e+00 9.56384689e-02
9.54007730e-02 -2.67415076e-01 -1.09759331e+00 4.75348502e-01
1.75869912e-01 6.46033466e-01 2.73720175e-01 -4.82503146e-01
-5.14169037e-01 -5.45980513e-01 -1.39669824e+00 5.77872992e-01
1.19265831e+00 -3.14024895e-01 -3.22815627e-01 3.81185114e-01
2.46148765e-01 -1.68053761e-01 -3.74616712e-01 7.34120905e-01
6.68816328e-01 -1.14468682e+00 9.31370258e-01 -5.40233552e-01
-2.71442439e-02 -5.19890666e-01 -1.75588772e-01 -9.20576155e-01
-2.68520832e-01 -6.39526069e-01 -7.60525614e-02 1.07478845e+00
-6.48589283e-02 -5.66610277e-01 8.56803954e-01 3.64312887e-01
-1.13671660e-01 -7.28307724e-01 -1.29371274e+00 -6.63439035e-01
-1.85230255e-01 1.47092998e-01 8.67311537e-01 9.14024413e-01
-5.79825044e-01 -2.28130877e-01 -4.08580452e-01 2.74541259e-01
7.03821421e-01 1.97394267e-02 5.95094144e-01 -1.44315636e+00
2.05483168e-01 -3.59984189e-01 -6.25419974e-01 -8.38468432e-01
5.52923441e-01 -7.10011601e-01 -1.54507011e-01 -8.97254467e-01
4.54569571e-02 -2.83516079e-01 -3.24613035e-01 4.59135294e-01
9.93306115e-02 7.41889238e-01 1.26896277e-01 1.92719042e-01
-1.87058777e-01 7.47823536e-01 1.35414958e+00 -1.63225159e-01
-1.46631479e-01 -8.44784975e-02 -4.21562403e-01 8.90029550e-01
4.48724687e-01 -5.53996325e-01 -2.35893250e-01 -4.90721643e-01
-4.07539219e-01 -2.01343149e-01 4.10108238e-01 -1.42417979e+00
3.50576133e-01 -2.39932127e-02 7.79183507e-01 -5.01062453e-01
5.09850442e-01 -1.16272116e+00 4.58068073e-01 3.88465941e-01
-5.77478968e-02 -1.88038707e-01 1.51365966e-01 4.61152136e-01
-3.99372101e-01 -1.49774745e-01 9.73771095e-01 -4.07237530e-04
-4.42122728e-01 7.52093136e-01 -4.08787131e-02 -3.58739585e-01
9.58738983e-01 -3.49242479e-01 8.83777365e-02 -1.30039632e-01
-8.47713530e-01 -2.60876149e-01 1.62209675e-01 4.86222118e-01
7.06190407e-01 -1.38323593e+00 -8.41731846e-01 1.18008173e+00
-2.45848119e-01 -1.27589449e-01 6.32201493e-01 8.11613619e-01
-3.82770658e-01 2.53605753e-01 -4.82982188e-01 -4.47638482e-01
-1.32868910e+00 6.09568417e-01 5.10696113e-01 9.00275558e-02
-4.13091421e-01 9.18561518e-01 4.78830755e-01 1.69134140e-03
3.95718694e-01 -1.06843196e-01 -3.69385213e-01 1.24100499e-01
4.89600927e-01 1.67660281e-01 3.12048823e-01 -1.18914568e+00
-4.20019090e-01 8.42751682e-01 -7.45751336e-02 1.54312238e-01
1.37808728e+00 6.01244979e-02 -4.03862774e-01 -2.55928785e-01
1.48066199e+00 7.14677721e-02 -1.34461999e+00 -5.51132798e-01
-2.70282716e-01 -8.01472068e-01 -5.11891991e-02 -2.08631501e-01
-1.84780085e+00 7.13081360e-01 7.18105078e-01 -2.54032314e-01
1.29946876e+00 -3.58289808e-01 6.07227266e-01 1.86648354e-01
1.98971689e-01 -7.10547626e-01 1.19763963e-01 4.03251708e-01
1.10960376e+00 -1.02444601e+00 -1.64575819e-02 -6.44415557e-01
-1.81409463e-01 1.04545319e+00 6.31262839e-01 -5.09533823e-01
1.16514158e+00 1.93494856e-02 -1.02480084e-01 -2.31312424e-01
-3.54103893e-01 -1.97942570e-01 3.59657437e-01 5.59043169e-01
2.69350205e-02 -2.75441170e-01 -4.57266629e-01 3.34701478e-01
1.32658541e-01 -9.52893589e-03 -4.82222624e-02 7.66739309e-01
-2.85503149e-01 -1.10432315e+00 -4.35638607e-01 3.44160557e-01
-3.47652495e-01 -6.01668330e-03 -1.87227681e-01 7.60343790e-01
9.02700186e-01 7.27254331e-01 2.80927181e-01 -1.61726743e-01
2.99675852e-01 1.70681834e-01 5.52416325e-01 -4.45742071e-01
-5.01249671e-01 1.64371073e-01 -4.91289049e-01 -5.85638046e-01
-4.43894625e-01 -7.58844078e-01 -9.35339153e-01 -2.53702521e-01
-3.09481561e-01 6.66810721e-02 5.85412025e-01 8.90586197e-01
4.59633350e-01 3.00942689e-01 8.06257427e-01 -9.99865353e-01
-3.45347553e-01 -9.00646806e-01 -6.18400931e-01 4.55862284e-01
5.01716852e-01 -1.11303473e+00 -5.59033811e-01 6.86124116e-02] | [13.207659721374512, 0.4264540672302246] |
6be82154-7f47-4d01-a5f2-3d6965ef23c3 | norquad-norwegian-question-answering-dataset | 2305.01957 | null | https://arxiv.org/abs/2305.01957v1 | https://arxiv.org/pdf/2305.01957v1.pdf | NorQuAD: Norwegian Question Answering Dataset | In this paper we present NorQuAD: the first Norwegian question answering dataset for machine reading comprehension. The dataset consists of 4,752 manually created question-answer pairs. We here detail the data collection procedure and present statistics of the dataset. We also benchmark several multilingual and Norwegian monolingual language models on the dataset and compare them against human performance. The dataset will be made freely available. | ['Lilja Øvrelid', 'Sondre Wold', 'Matias Jentoft', 'Fredrik Aas Andreassen', 'Sardana Ivanova'] | 2023-05-03 | null | null | null | null | ['reading-comprehension', 'machine-reading-comprehension'] | ['natural-language-processing', 'natural-language-processing'] | [-1.47963852e-01 3.50757867e-01 1.77287415e-01 -3.41986090e-01
-1.38100147e+00 -9.59892929e-01 4.61341858e-01 6.27017200e-01
-1.00492895e+00 9.86197650e-01 4.51439470e-01 -8.65145802e-01
6.64828494e-02 -4.41032976e-01 -5.28819442e-01 1.49369806e-01
4.43091303e-01 9.69584644e-01 4.94119972e-01 -7.83880115e-01
1.91090405e-01 -3.52216482e-01 -1.31596673e+00 4.44957078e-01
1.16659582e+00 6.22602582e-01 2.60321140e-01 1.31808650e+00
-6.82466701e-02 1.34975338e+00 -9.51920390e-01 -7.94776082e-01
-1.32532150e-01 -6.85839295e-01 -1.78642046e+00 -7.92568147e-01
9.99645233e-01 -2.92986244e-01 -1.74770311e-01 6.86370015e-01
7.52250552e-01 3.42457861e-01 4.13128793e-01 -7.60046840e-01
-9.94705379e-01 6.83079004e-01 2.95488983e-01 9.32360053e-01
1.40509307e+00 -7.39510059e-02 1.39441431e+00 -9.62742805e-01
9.26891446e-01 1.20909894e+00 2.63899624e-01 7.50116825e-01
-8.06293964e-01 -1.26048177e-01 -1.73158377e-01 7.11098075e-01
-1.12733841e+00 -5.37668824e-01 4.42394555e-01 -1.22515842e-01
1.34199524e+00 6.40148520e-01 1.89583272e-01 9.57132041e-01
4.67529744e-02 1.05909479e+00 1.21843410e+00 -1.01110673e+00
-2.43887872e-01 -1.10812187e-01 9.83229160e-01 7.02265739e-01
-1.72206923e-01 -3.31484407e-01 -5.36466479e-01 3.08740530e-02
1.21669278e-01 -9.93826687e-01 -5.76507211e-01 2.61847764e-01
-1.25833476e+00 7.58988678e-01 2.10851997e-01 3.04545075e-01
-3.90681550e-02 -2.89701998e-01 5.00558853e-01 1.11596906e+00
2.52442062e-01 9.19026732e-01 -1.01513076e+00 -4.00823921e-01
-3.80783468e-01 6.92846954e-01 1.39897156e+00 9.92500484e-01
3.71266633e-01 -4.69766289e-01 -3.31199050e-01 1.03942394e+00
2.31423825e-02 7.18976676e-01 5.16016006e-01 -9.65951741e-01
9.64446545e-01 3.66562694e-01 8.02257378e-03 -7.96349525e-01
-6.87337041e-01 9.45245177e-02 -4.27214056e-01 -7.94425845e-01
7.98249066e-01 -2.42762968e-01 -4.90235239e-01 1.40343964e+00
1.82782292e-01 -3.19939494e-01 5.90978265e-01 6.08496487e-01
2.11088443e+00 7.24512458e-01 2.29069084e-01 -7.87221342e-02
1.81547046e+00 -1.37960398e+00 -1.04682910e+00 -1.82280019e-01
8.76127422e-01 -1.00100553e+00 1.41076827e+00 1.75100297e-01
-1.19498158e+00 -6.54337525e-01 -5.95958710e-01 -9.15956795e-01
-5.24439633e-01 6.09692261e-02 6.93797618e-02 5.33464313e-01
-1.20486045e+00 -1.73392653e-01 -3.65933478e-01 -4.49744612e-01
-1.95202038e-01 -1.89746954e-02 -3.46207917e-01 -3.99264663e-01
-1.57963729e+00 1.09700012e+00 4.94930178e-01 1.20719813e-01
-6.54881597e-01 -4.73626792e-01 -1.14332092e+00 -3.81796122e-01
3.15699995e-01 -6.61657989e-01 2.18627667e+00 -5.92721939e-01
-1.38168001e+00 1.36666286e+00 -4.32279974e-01 -4.74285990e-01
1.22561119e-01 -4.96616274e-01 -5.13269246e-01 2.81095177e-01
2.41698176e-01 7.45733440e-01 2.08317041e-01 -8.04291129e-01
-4.44900930e-01 -1.08539119e-01 4.74698544e-01 3.37396801e-01
4.53339398e-01 6.16417348e-01 -2.88801849e-01 -6.17989838e-01
-1.78994220e-02 -6.95034981e-01 2.10251231e-02 -8.40982795e-01
-3.79676223e-01 -8.99984837e-01 3.05200338e-01 -1.37307250e+00
1.30916798e+00 -1.58730698e+00 4.72633503e-02 -5.24394810e-01
1.72921374e-01 2.87371933e-01 -5.48359275e-01 7.66095638e-01
-1.50663853e-02 -1.54576022e-02 -2.74347693e-01 -1.51262060e-01
-1.08107522e-01 2.05175281e-01 -2.12248176e-01 1.95553713e-02
2.05146387e-01 1.54908645e+00 -1.27808642e+00 -3.59123141e-01
-2.11585090e-01 -2.96783209e-01 -4.04998004e-01 6.27111852e-01
-7.09242344e-01 6.28112495e-01 -4.45317268e-01 6.78504169e-01
3.70084763e-01 1.74788043e-01 -2.49424383e-01 4.84931082e-01
1.92090541e-01 9.38724637e-01 -4.00953472e-01 1.80065608e+00
-4.02423948e-01 9.85166550e-01 8.21855590e-02 -3.83145392e-01
8.33371341e-01 6.15019321e-01 -3.84107918e-01 -1.24212611e+00
2.29829356e-01 4.60338116e-01 9.46604088e-02 -1.04603052e+00
1.00628603e+00 1.74438320e-02 -4.76320982e-01 3.45163763e-01
2.03819752e-01 -4.97590929e-01 5.90840995e-01 1.90676808e-01
1.30372405e+00 -1.41274542e-01 7.05738187e-01 -5.72560310e-01
1.25288415e+00 4.04814482e-01 -6.23565093e-02 8.92582119e-01
-5.20331800e-01 5.24419546e-01 5.71410596e-01 -4.23811197e-01
-5.09275317e-01 -1.27868879e+00 -1.72088772e-01 1.52301490e+00
-1.54391050e-01 -5.56205928e-01 -8.26912284e-01 -7.78437257e-01
-3.98628026e-01 9.28156257e-01 -6.00269437e-01 3.15356672e-01
-9.15498912e-01 -5.32122739e-02 9.35624421e-01 3.70855302e-01
5.66901088e-01 -1.61919248e+00 -2.20584229e-01 4.41387370e-02
-7.95656264e-01 -1.22388291e+00 -3.26838464e-01 -9.74284783e-02
-6.44458294e-01 -1.53171551e+00 -5.62306941e-01 -1.58167219e+00
2.01625258e-01 -1.01010680e-01 2.28815985e+00 5.39292336e-01
2.91876376e-01 7.53616035e-01 -9.39204991e-01 -7.02450991e-01
-7.03154743e-01 8.53479862e-01 -4.14974928e-01 -8.99766743e-01
9.11969960e-01 1.53038114e-01 -1.33832946e-01 2.44959109e-02
-6.60572529e-01 -1.59160703e-01 1.35415420e-02 8.05300713e-01
4.30156231e-01 -7.75465429e-01 9.30616021e-01 -1.14713359e+00
1.17745340e+00 -5.65692246e-01 -5.51968575e-01 5.54758430e-01
1.09726392e-01 6.84319586e-02 6.11045718e-01 2.63474524e-01
-1.00849843e+00 -2.85356045e-01 -8.87148261e-01 5.08402050e-01
-4.52484697e-01 6.76955223e-01 -1.01081192e-01 3.04561913e-01
7.87062883e-01 -1.12511493e-01 -3.38472307e-01 -8.09657454e-01
7.70190120e-01 8.44200253e-01 9.54138756e-01 -6.29792213e-01
4.41519201e-01 -4.09566551e-01 -6.77199781e-01 -1.18477523e+00
-1.24190664e+00 -8.79202545e-01 -8.32645237e-01 -5.20105660e-02
1.15435743e+00 -8.51750016e-01 -6.46250486e-01 6.17521286e-01
-1.51881909e+00 -3.86072218e-01 -3.05232912e-01 1.05440401e-01
-4.37626034e-01 8.23599286e-03 -9.45535839e-01 -4.29630995e-01
-4.71469015e-01 -9.89526391e-01 1.03741455e+00 1.84804633e-01
-4.65369552e-01 -1.39902878e+00 7.14229047e-01 1.04446149e+00
2.53962487e-01 -1.44304127e-01 8.62178206e-01 -1.02027774e+00
-2.70638108e-01 2.86445767e-02 -1.32506341e-01 3.28667641e-01
-2.60774106e-01 -5.07089972e-01 -8.98026645e-01 -1.61079720e-01
5.99602275e-02 -9.92553234e-01 8.22303951e-01 -2.97733545e-02
7.65949428e-01 -2.28606656e-01 2.89097875e-01 -1.50623530e-01
1.07660317e+00 -2.51133233e-01 5.92986465e-01 3.93894047e-01
8.26728821e-01 1.16708446e+00 3.86128873e-01 -4.39295590e-01
1.39601803e+00 3.08614403e-01 1.12284347e-01 2.73427904e-01
-2.69613802e-01 -4.14075077e-01 3.00512254e-01 1.91098118e+00
3.21826458e-01 -5.12017548e-01 -1.41595006e+00 8.36126029e-01
-1.50389993e+00 -5.42683721e-01 -7.46241927e-01 1.59364343e+00
1.25440621e+00 -4.19431210e-01 2.42484324e-02 -1.23555161e-01
2.47448400e-01 2.37559229e-01 -2.38665286e-02 -8.62624645e-01
-4.61025238e-01 8.08051884e-01 2.81188488e-02 1.06698036e+00
-1.15427113e+00 1.35817766e+00 7.42443943e+00 5.50515115e-01
-1.65954173e-01 3.31443399e-01 3.68425578e-01 4.58521932e-01
-4.26114172e-01 -1.38018414e-01 -5.96143126e-01 -1.71189219e-01
1.52972245e+00 -1.14162847e-01 1.00702241e-01 3.16201985e-01
-1.54981494e-01 -5.10617554e-01 -1.09569669e+00 6.87764525e-01
4.53547180e-01 -9.20011699e-01 1.50487110e-01 -8.69504571e-01
6.16351604e-01 3.79250705e-01 -4.31997120e-01 7.53183484e-01
2.43705317e-01 -1.26710975e+00 5.32408059e-01 6.94254100e-01
6.06244624e-01 -5.96479416e-01 1.11134315e+00 4.62177366e-01
-8.54781449e-01 1.71069831e-01 -2.64778852e-01 -6.18798852e-01
2.71531731e-01 -4.58814688e-02 -5.14545143e-01 6.31921649e-01
7.05088973e-01 5.42966247e-01 -1.30328393e+00 8.81886840e-01
-8.45361054e-01 1.01065278e+00 -3.43455449e-02 -4.59884554e-01
1.45932317e-01 -1.40904829e-01 4.76945400e-01 1.13182807e+00
-5.35964668e-01 2.91756898e-01 1.34838969e-01 3.67933095e-01
-3.97517443e-01 7.52970099e-01 -4.57091480e-01 1.73066422e-01
2.99351454e-01 7.05909848e-01 -2.49334961e-01 -3.85639012e-01
-7.10794806e-01 9.38545048e-01 7.93572247e-01 2.92835146e-01
-1.75791770e-01 -6.08487546e-01 2.30786413e-01 -4.64877039e-02
-2.09907606e-01 -2.87733495e-01 3.21497619e-02 -1.29313242e+00
3.07271749e-01 -1.52343488e+00 6.93799615e-01 -8.84770215e-01
-1.62696874e+00 8.57328117e-01 -3.97958010e-02 -4.88034487e-01
-4.52336222e-01 -8.28624547e-01 -2.82423347e-01 8.89368653e-01
-1.76384950e+00 -7.63215661e-01 -3.25121701e-01 4.14259821e-01
7.92411208e-01 -2.20380425e-01 1.16585374e+00 3.50969791e-01
-4.10420239e-01 6.94123566e-01 -7.17770308e-03 5.43399930e-01
8.16672623e-01 -1.53089178e+00 9.11197960e-01 6.31510258e-01
4.95338440e-01 5.02116084e-01 4.93019879e-01 -4.04116482e-01
-9.87074852e-01 -6.99711084e-01 1.84304261e+00 -1.39001179e+00
9.93539989e-01 -3.56586099e-01 -1.31530631e+00 8.66117120e-01
1.12888837e+00 -5.09513259e-01 7.43590653e-01 1.91565797e-01
-2.10647270e-01 3.44894886e-01 -9.39966321e-01 4.77982283e-01
8.43230426e-01 -8.85756791e-01 -1.46191835e+00 6.31366014e-01
1.29047680e+00 -8.20049584e-01 -8.43317211e-01 2.72477359e-01
9.49160382e-02 -8.21692824e-01 5.44294655e-01 -9.38315570e-01
3.79261196e-01 -1.44173995e-01 -4.32585850e-02 -1.20288432e+00
5.45744821e-02 -6.19781852e-01 8.99666697e-02 9.98921037e-01
9.51475918e-01 -7.62612820e-01 2.90541798e-01 2.68080324e-01
-2.07911417e-01 -6.77924037e-01 -1.38479114e+00 -4.42815870e-01
8.76144230e-01 -5.60224712e-01 3.90397608e-01 7.44905293e-01
1.21046260e-01 1.09833646e+00 3.19533557e-01 -4.02350258e-03
-2.14718394e-02 -1.85296521e-01 6.62073791e-01 -9.66926754e-01
-2.04931051e-01 -3.20532501e-01 -1.24571778e-01 -1.41661334e+00
5.88848472e-01 -1.15954244e+00 2.07838267e-01 -1.68853498e+00
-8.62112641e-02 -1.29607216e-01 1.92223161e-01 3.90132636e-01
-7.74233401e-01 1.69188529e-01 1.11585341e-01 4.33540577e-03
-1.08400559e+00 6.76811218e-01 1.39976299e+00 -3.96327563e-02
-3.24516073e-02 -1.40568540e-01 -5.14415503e-01 4.32106882e-01
1.05560088e+00 -2.73483664e-01 -3.53112876e-01 -1.10189474e+00
3.71876299e-01 5.90306036e-02 1.44085720e-01 -8.32265258e-01
8.96166041e-02 3.68630558e-01 -1.99953392e-01 -7.11766064e-01
-5.50695658e-02 -3.01835924e-01 -9.41580415e-01 2.59494960e-01
-5.47820926e-01 9.70357299e-01 3.81858051e-01 2.26365328e-01
-6.89080238e-01 -6.65268183e-01 1.84820250e-01 -1.02091506e-01
-6.97797894e-01 -2.78517067e-01 -5.88934183e-01 1.13010287e+00
7.06577122e-01 4.44751918e-01 -6.13422275e-01 -7.40861058e-01
-4.76449102e-01 9.48345363e-01 1.46984324e-01 9.93578792e-01
3.91546428e-01 -1.01792693e+00 -1.39183462e+00 -3.32162380e-02
6.02605224e-01 -7.10645169e-02 5.64749464e-02 4.21445847e-01
-1.12586236e+00 9.73600388e-01 2.53577769e-01 -2.23214909e-01
-1.15658462e+00 3.26628268e-01 5.09443283e-01 -3.91092628e-01
-1.45296380e-01 1.11662734e+00 -4.98708159e-01 -1.30285478e+00
1.63308270e-02 -5.17093360e-01 -8.79102528e-01 -1.59164574e-02
7.31168449e-01 4.31825578e-01 1.52328372e-01 -9.18334365e-01
-3.40353221e-01 2.74740517e-01 -1.19295277e-01 -2.73199528e-01
6.10472441e-01 -2.69641310e-01 -7.29671478e-01 8.13557625e-01
1.12316704e+00 1.54765487e-01 -1.66213781e-01 -2.99608141e-01
4.62451547e-01 1.22972868e-01 -4.20217037e-01 -9.87545371e-01
-4.51633990e-01 5.19313216e-01 1.78864330e-01 2.65671939e-01
1.01038575e+00 4.47133362e-01 1.25621068e+00 1.00497031e+00
1.76275909e-01 -9.74817097e-01 -1.78218693e-01 1.58285451e+00
1.07651210e+00 -1.51794720e+00 -3.57244790e-01 -4.03140366e-01
-3.14274848e-01 8.10764134e-01 8.10697556e-01 -5.80696538e-02
5.95483124e-01 -3.20831358e-01 8.53844464e-01 -4.27777946e-01
-7.58626819e-01 -4.34117317e-01 7.08004475e-01 6.90267324e-01
1.05926013e+00 1.41964838e-01 -7.18268692e-01 8.49149883e-01
-1.24032295e+00 -3.99867207e-01 6.74977481e-01 1.05320466e+00
-2.66905308e-01 -1.33941722e+00 -3.34319651e-01 3.92497838e-01
-5.21287382e-01 -5.29749632e-01 -8.21248531e-01 8.41183007e-01
-2.10239127e-01 1.66144145e+00 3.21533009e-02 6.70101196e-02
8.42197239e-01 3.33850443e-01 6.14747524e-01 -9.40280318e-01
-9.10320640e-01 -8.59773457e-01 7.20782161e-01 -3.50533277e-01
-4.12221044e-01 -6.54552937e-01 -1.04485953e+00 -8.71365294e-02
-2.16630340e-01 4.78773415e-01 4.69264528e-03 1.08740580e+00
7.42482170e-02 2.71074384e-01 -1.56834237e-02 3.17304134e-02
-3.25592816e-01 -1.30985773e+00 1.62677560e-02 2.37149760e-01
5.71568787e-01 -1.04751885e-01 -3.64020854e-01 -6.48818165e-02] | [11.356464385986328, 8.1957426071167] |
11b80a44-7ac0-42f9-bf98-b8cdefd80045 | glavnet-global-local-audio-visual-cues-for | null | null | http://openaccess.thecvf.com//content/CVPR2021/html/Shi_GLAVNet_Global-Local_Audio-Visual_Cues_for_Fine-Grained_Material_Recognition_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Shi_GLAVNet_Global-Local_Audio-Visual_Cues_for_Fine-Grained_Material_Recognition_CVPR_2021_paper.pdf | GLAVNet: Global-Local Audio-Visual Cues for Fine-Grained Material Recognition | In this paper, we aim to recognize materials with combined use of auditory and visual perception. To this end, we construct a new dataset named GLAudio that consists of both the geometry of the object being struck and the sound captured from either modal sound synthesis (for virtual objects) or real measurements (for real objects). Besides global geometries, our dataset also takes local geometries around different hitpoints into consideration. This local information is less explored in existing datasets. We demonstrate that local geometry has a greater impact on the sound than the global geometry and offers more cues in material recognition. To extract features from different modalities and perform proper fusion, we propose a new deep neural network GLAVNet that comprises multiple branches and a well-designed fusion module. Once trained on GLAudio, our GLAVNet provides state-of-the-art performance on material identification and supports fine-grained material categorization. | ['Yanwen Guo', 'Xiying Wang', 'Shan Yang', 'Haonan Zhang', 'Jie Guo', 'Fengmin Shi'] | 2021-06-19 | null | null | null | cvpr-2021-1 | ['material-recognition'] | ['computer-vision'] | [-2.38750782e-02 -6.05884731e-01 3.61057937e-01 -6.72894418e-02
-7.96871662e-01 -7.61147380e-01 6.47129118e-01 1.95542619e-01
-1.65668130e-01 5.76747917e-02 9.83603969e-02 4.07936946e-02
-1.55831218e-01 -1.14636779e+00 -7.26018488e-01 -8.33803773e-01
1.56383753e-01 2.68121332e-01 4.78600413e-01 -5.75818084e-02
3.55073005e-01 8.77496481e-01 -2.11069393e+00 5.56957662e-01
2.60914952e-01 1.78023601e+00 4.48748082e-01 7.66185582e-01
-2.76706040e-01 4.33149517e-01 -6.15248621e-01 -5.15454123e-03
3.43545765e-01 1.44676194e-01 -6.92708075e-01 3.60014141e-02
1.07615638e+00 -2.07393274e-01 -3.80026847e-01 7.63698339e-01
7.66040325e-01 2.73265317e-02 9.02361393e-01 -8.81474435e-01
-6.20562375e-01 8.11680675e-01 -1.99896127e-01 1.86701134e-01
5.42137384e-01 2.18031242e-01 1.15760350e+00 -1.05610847e+00
2.11389333e-01 1.17544222e+00 6.87180698e-01 1.45103782e-01
-9.66998160e-01 -4.11073506e-01 1.58538818e-01 3.16098511e-01
-1.43961549e+00 -4.76789355e-01 1.21641243e+00 -5.76166928e-01
8.06479931e-01 4.09202695e-01 5.31287074e-01 1.21993351e+00
6.02301583e-02 6.18532360e-01 1.03918803e+00 -5.43289542e-01
3.28110129e-01 -1.11183161e-02 -6.47573248e-02 5.82283616e-01
2.02173945e-02 7.77969286e-02 -6.79038346e-01 3.84714305e-02
7.22398520e-01 -1.90417558e-01 -2.20691592e-01 -3.25425863e-01
-1.25129175e+00 3.50075305e-01 5.68399906e-01 3.15071583e-01
-3.04189593e-01 3.41218382e-01 1.38702646e-01 5.23843654e-02
1.22751288e-01 6.45180583e-01 -9.56489593e-02 2.39138361e-02
-6.06216490e-01 1.70048043e-01 6.95185721e-01 6.10381722e-01
6.53188050e-01 2.11944297e-01 -1.67285472e-01 1.13239479e+00
3.81970108e-01 8.51625681e-01 2.74049222e-01 -6.93417847e-01
4.17629480e-01 4.24234629e-01 -1.86324984e-01 -1.22099316e+00
-5.23675144e-01 -7.16660857e-01 -5.64973474e-01 2.35949144e-01
6.48319125e-01 5.23984134e-01 -8.96317840e-01 1.44945300e+00
2.54692495e-01 8.41099471e-02 -2.37343684e-01 1.05321503e+00
1.43113208e+00 2.83987820e-01 -2.62056112e-01 5.17996252e-01
1.45296979e+00 -8.95376444e-01 -1.82739168e-01 -1.12914503e-01
1.29390076e-01 -9.86033380e-01 1.08201039e+00 8.59636545e-01
-1.04278564e+00 -9.35581207e-01 -1.19343853e+00 5.72371706e-02
-7.96346068e-01 3.19965333e-01 5.21299303e-01 5.41768968e-01
-8.57126474e-01 3.10827345e-01 -5.25492489e-01 -1.42464355e-01
2.45609775e-01 2.06428945e-01 -1.57479882e-01 -1.12587743e-01
-8.20327401e-01 3.43390912e-01 3.49971689e-02 1.17075466e-01
-8.48609984e-01 -7.70114660e-01 -7.82632470e-01 3.29967365e-02
5.19756317e-01 -6.39270604e-01 1.11216176e+00 -2.82247931e-01
-1.59874523e+00 8.03473651e-01 3.59072298e-01 -4.50722538e-02
3.87574375e-01 -1.12620249e-01 -5.48181474e-01 4.39852148e-01
-9.17183012e-02 3.53404820e-01 1.10831141e+00 -1.80441213e+00
-3.30598146e-01 -3.24082136e-01 2.25559279e-01 -1.89220309e-01
-2.06556946e-01 -8.10400695e-02 -3.86929691e-01 -7.27611184e-01
4.67787534e-01 -6.37050390e-01 3.70565504e-01 7.87957944e-03
-7.65346289e-01 -1.38839841e-01 6.29829466e-01 -5.13725638e-01
8.33546758e-01 -2.22009778e+00 -3.97337824e-02 3.01727206e-01
3.58688235e-01 2.09831238e-01 -3.48974049e-01 4.97643679e-01
2.13727325e-01 -1.00018211e-01 1.93908975e-01 -3.70428175e-01
2.72789299e-01 -3.10221553e-01 -3.92000198e-01 2.78950721e-01
1.59776196e-01 8.38128567e-01 -5.24896681e-01 -3.02185982e-01
5.07312298e-01 4.87965286e-01 -5.19370556e-01 9.94673297e-02
-1.67332470e-01 2.62493014e-01 -2.78604597e-01 1.04947233e+00
1.15218151e+00 -6.73019662e-02 -3.94441485e-01 -6.99224055e-01
-1.65086284e-01 3.41173172e-01 -1.48062825e+00 1.66534126e+00
-7.38797247e-01 5.17579794e-01 2.79111832e-01 -6.12121761e-01
1.10236061e+00 -2.91820280e-02 5.22880495e-01 -8.71812761e-01
3.20898503e-01 2.15262160e-01 -2.08278671e-01 -5.88726938e-01
5.54229081e-01 1.63396060e-01 -7.22027477e-03 3.51885408e-01
2.43985057e-01 -2.91933715e-01 -2.67508119e-01 -1.48610741e-01
9.73233581e-01 -4.67369184e-02 7.82409534e-02 -5.82875274e-02
5.12673438e-01 -4.64303672e-01 7.05107525e-02 8.71335745e-01
1.33763179e-01 9.86735821e-01 -1.28795102e-01 -2.64885604e-01
-7.05533743e-01 -1.43832171e+00 -3.90175849e-01 1.18135083e+00
4.70226198e-01 -3.74117285e-01 -4.87576097e-01 -3.87460321e-01
3.45415533e-01 3.64472568e-01 -6.13561332e-01 -2.40331694e-01
-4.63506907e-01 -3.95863146e-01 5.77954233e-01 8.00251007e-01
6.02355301e-01 -1.05280256e+00 -5.67942262e-01 1.31197602e-01
-1.07098930e-01 -1.34632385e+00 -2.34382913e-01 1.17306791e-01
-1.70088202e-01 -9.33751643e-01 -3.24232012e-01 -5.52255690e-01
1.78821772e-01 4.32687461e-01 1.28701341e+00 -1.42384276e-01
-5.49202740e-01 7.39858270e-01 -5.53231001e-01 -6.12115502e-01
-2.47814924e-01 -4.72154561e-03 -8.39773752e-03 3.86245281e-01
1.07522324e-01 -7.28899658e-01 -6.32407248e-01 4.78754789e-01
-7.45820761e-01 -5.12144864e-02 6.19031668e-01 4.70615774e-01
6.01215243e-01 1.06439732e-01 3.75802249e-01 -1.38551176e-01
3.24385107e-01 -2.81654626e-01 -2.49933824e-01 1.68750241e-01
9.31532830e-02 -2.11827978e-01 7.67476261e-01 -5.72701097e-01
-8.77064168e-01 -1.26785904e-01 -3.21917117e-01 -4.48444784e-01
-7.38671482e-01 1.92566201e-01 -4.99499947e-01 -4.33244288e-01
5.66845655e-01 1.91491067e-01 -4.19144422e-01 -9.80753303e-01
2.47164160e-01 9.37418222e-01 7.88149297e-01 -7.95403302e-01
6.42966628e-01 6.32067382e-01 1.06416017e-01 -1.13347161e+00
-6.47892773e-01 -4.24461186e-01 -7.23080456e-01 -4.85068083e-01
6.89074039e-01 -6.81258500e-01 -1.01149356e+00 9.74174440e-01
-9.01244640e-01 -2.26044029e-01 -3.45360547e-01 4.65409666e-01
-4.69161600e-01 1.75186992e-01 -4.63942289e-01 -9.32072520e-01
-9.21614096e-02 -1.19470477e+00 1.53161538e+00 -1.05931632e-01
1.65926844e-01 -8.01701248e-01 -1.00260489e-01 4.38678414e-01
3.79939854e-01 1.19397402e-01 7.87669659e-01 -3.98052722e-01
-7.82462180e-01 -1.07055925e-01 -4.22196835e-01 3.01540554e-01
3.60947400e-01 1.19297475e-01 -1.56835175e+00 -1.18052319e-01
-2.36181200e-01 -3.62709612e-01 1.28330791e+00 2.68975496e-01
1.32981420e+00 1.28829375e-01 -4.15111408e-02 8.09092879e-01
1.49040556e+00 1.50541037e-01 3.24714065e-01 3.81180555e-01
1.13980031e+00 7.33593583e-01 2.91712731e-01 4.88502085e-01
3.61675888e-01 9.86295879e-01 7.29076147e-01 2.02894770e-02
-7.11507320e-01 -1.84819534e-01 1.59112021e-01 9.48948562e-01
-4.88516092e-02 -4.26723391e-01 -8.94213498e-01 3.51514190e-01
-1.33531451e+00 -7.33321667e-01 1.92842949e-02 2.01565838e+00
3.19299281e-01 -1.46412393e-02 9.24672782e-02 5.95678985e-01
4.01807755e-01 1.68835774e-01 -5.03118455e-01 -2.14139938e-01
-4.98068362e-01 2.00253859e-01 1.71062693e-01 1.90206885e-01
-1.21466601e+00 6.44672990e-01 6.65923166e+00 9.96951699e-01
-1.45367396e+00 -2.53334582e-01 1.81758821e-01 2.02319510e-02
-2.11804271e-01 -6.48910463e-01 -6.38864458e-01 3.59110057e-01
4.42310989e-01 6.21426463e-01 5.30401528e-01 6.87023163e-01
-1.65945396e-01 -6.92318752e-02 -1.13199067e+00 1.33975005e+00
3.33915740e-01 -1.21213734e+00 1.76970974e-01 1.50354505e-01
2.88813204e-01 -3.63594741e-02 3.66378218e-01 -1.31277710e-01
4.15594690e-02 -9.73037064e-01 1.42867506e+00 7.07479537e-01
7.35987604e-01 -3.72937739e-01 5.14993429e-01 4.05267021e-03
-1.72229934e+00 -1.90822959e-01 -2.42999569e-01 4.06432822e-02
-2.30787788e-02 6.75337732e-01 -8.67250860e-01 8.90160739e-01
7.87590086e-01 4.90367889e-01 -8.36920679e-01 1.19381320e+00
9.86126736e-02 6.63634181e-01 -4.71100599e-01 5.18607534e-02
-8.33490491e-03 1.80334374e-01 6.50976002e-01 1.29090536e+00
4.48348284e-01 -4.85676467e-01 4.48523730e-01 9.92435277e-01
9.09053236e-02 2.30955020e-01 -4.71802086e-01 8.99916142e-03
4.19824183e-01 1.33726299e+00 -8.29666734e-01 -1.10859320e-01
-3.15399200e-01 4.06413704e-01 9.05348435e-02 2.52092928e-01
-4.91819352e-01 -2.46454969e-01 7.19628215e-01 1.41012564e-01
5.70657670e-01 -3.65642637e-01 -4.23845530e-01 -1.01074529e+00
1.54745385e-01 -5.67493379e-01 1.71606496e-01 -1.11033571e+00
-1.63443708e+00 7.64009714e-01 -2.71051496e-01 -1.38163507e+00
3.19011994e-02 -1.09469581e+00 -5.83015859e-01 7.10351169e-01
-1.32386768e+00 -1.42413104e+00 -7.18786299e-01 4.42090571e-01
3.11394185e-01 -2.45287776e-01 6.91855609e-01 3.00714403e-01
-2.88022012e-01 7.12129116e-01 -2.07259670e-01 1.52464509e-01
6.48845613e-01 -1.29573560e+00 1.71760127e-01 4.86454844e-01
4.21272367e-01 1.82788149e-01 7.28422940e-01 -3.01739573e-01
-1.60089040e+00 -9.36105490e-01 1.46949336e-01 -4.02608037e-01
6.41120732e-01 -5.83249927e-01 -7.74996579e-01 1.05843998e-01
-2.06034750e-01 3.31901456e-03 6.74579322e-01 -1.32624032e-02
-9.00193930e-01 -3.21203470e-01 -1.09054875e+00 2.22681463e-01
1.20884502e+00 -8.54785502e-01 -3.81948143e-01 -2.06502806e-02
7.13379741e-01 -2.90128380e-01 -9.02825058e-01 4.13116366e-01
6.95542216e-01 -1.08754551e+00 1.40913475e+00 -6.67119622e-02
3.59871000e-01 -2.70646989e-01 -7.16040730e-01 -1.25106394e+00
-4.36206490e-01 1.48321781e-02 -1.26155093e-01 1.42570245e+00
6.38499856e-02 -6.08701229e-01 5.23653269e-01 -1.31656349e-01
-5.15961766e-01 -6.96946502e-01 -8.30382705e-01 -8.83236349e-01
6.62764609e-02 -1.08351254e+00 9.92543578e-01 5.95438480e-01
-4.14961785e-01 5.35130613e-02 -1.55443549e-01 3.88919413e-01
6.51297629e-01 6.12362385e-01 6.58693254e-01 -1.52224660e+00
-4.96853203e-01 -6.90320253e-01 -7.91318119e-01 -1.10473192e+00
-2.92765684e-02 -9.70856547e-01 1.70175567e-01 -1.49020016e+00
9.43814144e-02 -4.73625422e-01 -4.22652543e-01 3.25957924e-01
2.16425225e-01 6.82282627e-01 2.48416096e-01 -8.15289468e-02
-6.35102987e-01 6.52127981e-01 1.39269793e+00 -4.29240435e-01
-1.28621206e-01 -1.60023600e-01 -6.37087047e-01 7.03561127e-01
6.66858375e-01 1.22199297e-01 8.14426239e-05 -5.63774824e-01
3.42872888e-01 -1.94643766e-01 8.11639190e-01 -1.42955053e+00
1.42907292e-01 -2.04123538e-02 3.48226875e-01 -8.61071527e-01
7.96845317e-01 -6.85485244e-01 -3.37940566e-02 -1.70018405e-01
-2.16760382e-01 -3.81318331e-01 1.89101681e-01 3.77869219e-01
-2.70393074e-01 -3.45083587e-02 6.21747613e-01 3.39423157e-02
-6.19802773e-01 2.61926651e-01 -3.63678306e-01 -1.70490205e-01
4.45681572e-01 -3.22198480e-01 -5.48434615e-01 -1.42825589e-01
-7.65124798e-01 -3.54947954e-01 4.45156127e-01 4.75805491e-01
6.79770410e-01 -1.62125564e+00 -4.65685725e-01 4.35338527e-01
3.29217255e-01 5.93642145e-02 5.31229079e-01 6.41107619e-01
-3.81943673e-01 4.65654343e-01 -3.29444259e-01 -7.78601944e-01
-1.24714005e+00 5.29486537e-01 2.60019451e-01 2.44557753e-01
-5.28143108e-01 9.54658508e-01 4.27839577e-01 -5.19274175e-01
3.86286527e-01 -6.76943958e-01 -3.87980372e-01 2.05050573e-01
5.50735116e-01 4.73897517e-01 3.93435061e-01 -9.29752827e-01
-3.77730936e-01 9.86287296e-01 3.78272384e-01 -7.71795288e-02
1.38006473e+00 -1.79305390e-01 -8.76738951e-02 8.33416760e-01
1.27270937e+00 5.29448807e-01 -1.15124202e+00 -5.67168236e-01
-5.06521821e-01 -5.13104677e-01 2.80230135e-01 -9.52926636e-01
-1.26255167e+00 1.17053998e+00 5.71290791e-01 6.38574481e-01
1.12967145e+00 3.91783684e-01 7.03339934e-01 2.84522176e-01
6.13252759e-01 -9.75254059e-01 5.96520722e-01 4.54551578e-01
1.11757338e+00 -1.13527179e+00 -3.55914325e-01 -8.68601441e-01
-6.03127778e-02 1.20598125e+00 5.14550745e-01 -5.87575920e-02
8.35409105e-01 3.96602541e-01 1.90094739e-01 -2.65642792e-01
-4.85451579e-01 -3.83252352e-01 8.92338932e-01 6.46119475e-01
9.08186100e-03 2.98609525e-01 6.26187861e-01 9.80911732e-01
-7.46227026e-01 -5.92491329e-01 1.15079291e-01 7.45543003e-01
-5.95648050e-01 -8.04125428e-01 -7.74057627e-01 3.30934584e-01
-2.01176360e-01 8.39446634e-02 -5.83239079e-01 2.80520767e-01
5.76049984e-01 1.04219854e+00 2.93707609e-01 -9.80702102e-01
3.83766174e-01 -2.25206837e-01 7.06790984e-01 -3.83565038e-01
-4.26884890e-01 1.81859136e-01 8.61826763e-02 -7.24168003e-01
-3.04590940e-01 -5.76224864e-01 -9.28433716e-01 -1.67200014e-01
-2.90931523e-01 -2.81122088e-01 9.45491731e-01 7.65477717e-01
1.71482146e-01 9.92458165e-01 6.95683658e-01 -1.40632331e+00
-2.45903268e-01 -9.44532871e-01 -9.32379901e-01 3.93288314e-01
4.85625237e-01 -1.05930674e+00 -5.22608280e-01 -1.85197711e-01] | [10.214286804199219, -0.16316315531730652] |
209b811e-26a8-44aa-85a5-537126edaf86 | machine-reading-fast-and-slow-when-do-models-1 | null | null | https://aclanthology.org/2022.coling-1.8 | https://aclanthology.org/2022.coling-1.8.pdf | Machine Reading, Fast and Slow: When Do Models “Understand” Language? | Two of the most fundamental issues in Natural Language Understanding (NLU) at present are: (a) how it can established whether deep learning-based models score highly on NLU benchmarks for the ”right” reasons; and (b) what those reasons would even be. We investigate the behavior of reading comprehension models with respect to two linguistic ”skills”: coreference resolution and comparison. We propose a definition for the reasoning steps expected from a system that would be ”reading slowly”, and compare that with the behavior of five models of the BERT family of various sizes, observed through saliency scores and counterfactual explanations. We find that for comparison (but not coreference) the systems based on larger encoders are more likely to rely on the ”right” information, but even they struggle with generalization, suggesting that they still learn specific lexical patterns rather than the general principles of comparison. | ['Isabelle Augenstein', 'Anna Rogers', 'Sagnik Ray Choudhury'] | null | null | null | null | coling-2022-10 | ['coreference-resolution'] | ['natural-language-processing'] | [ 3.02387029e-01 9.54855025e-01 -1.92560583e-01 -3.53211671e-01
-4.95809525e-01 -5.11913121e-01 8.82511377e-01 7.09118545e-01
-4.76138294e-01 6.07453942e-01 6.93963110e-01 -7.02977002e-01
-4.71731722e-01 -9.00042892e-01 -1.11996508e+00 -1.19858131e-01
1.92937002e-01 9.03446376e-01 3.86945844e-01 -5.76180339e-01
5.50234318e-01 1.37330800e-01 -1.76951528e+00 3.52205664e-01
1.08753216e+00 5.69173098e-01 3.29691052e-01 5.47024965e-01
-2.83409625e-01 1.20676184e+00 -3.21636856e-01 -4.97555912e-01
-1.13527127e-01 -7.55546629e-01 -1.58785093e+00 -5.29313743e-01
7.69014299e-01 -3.49854499e-01 -1.35934070e-01 1.10310841e+00
1.18233129e-01 1.93462014e-01 8.36639404e-01 -8.02186251e-01
-8.25387239e-01 1.21737564e+00 -1.84174851e-01 6.54361486e-01
6.76998496e-01 2.47945949e-01 1.50473297e+00 -3.82283956e-01
8.10931206e-01 1.73132694e+00 3.78368258e-01 6.89153731e-01
-1.36561060e+00 -2.99153060e-01 3.36751729e-01 4.78639066e-01
-6.77013397e-01 -2.89270997e-01 2.76307672e-01 -4.67932582e-01
1.15168333e+00 2.52243757e-01 4.22224015e-01 1.13417649e+00
3.29316735e-01 7.16760695e-01 1.25055480e+00 -5.06346047e-01
1.44927666e-01 4.96986806e-02 6.77852273e-01 4.32064980e-01
5.72517157e-01 4.07406360e-01 -7.50590622e-01 1.59833595e-01
6.02720678e-01 -4.35748935e-01 -6.71863616e-01 -3.88011754e-01
-1.39771628e+00 9.83457327e-01 5.57831645e-01 5.53830445e-01
-3.14853907e-01 -8.29319283e-02 2.29748189e-01 6.58368051e-01
-2.17787430e-01 1.23804402e+00 -8.22673202e-01 -1.34847984e-02
-7.58724570e-01 5.21916091e-01 7.95238316e-01 8.45731795e-01
6.57850027e-01 -3.76583278e-01 -2.26289794e-01 3.37486267e-01
1.52394533e-01 2.56081641e-01 6.36132658e-01 -1.15758204e+00
5.44316888e-01 5.38744330e-01 3.19296598e-01 -7.31961906e-01
-7.50237882e-01 -4.50943798e-01 -2.08506092e-01 1.60630763e-01
1.07403195e+00 -6.46902770e-02 -4.11940336e-01 2.35501099e+00
-2.69637853e-01 -5.35254061e-01 3.14666182e-01 8.42394173e-01
5.70047855e-01 4.51716185e-01 2.99547821e-01 -2.35629067e-01
1.33746588e+00 -6.87671185e-01 -4.78549689e-01 -6.64701402e-01
8.39969873e-01 -3.78134757e-01 1.52695918e+00 8.94417018e-02
-1.50288284e+00 -7.69097924e-01 -1.14848161e+00 -4.68382359e-01
-1.07850067e-01 -4.99211669e-01 6.53404653e-01 1.77546814e-01
-1.16363859e+00 9.69674826e-01 -4.05036300e-01 -8.51879299e-01
1.22443594e-01 -2.00117435e-02 1.64068751e-02 2.16381028e-01
-1.45796442e+00 1.70123160e+00 6.26603007e-01 -4.43389565e-01
-8.86528850e-01 -7.62259603e-01 -6.88998222e-01 7.20941782e-01
4.08980399e-01 -7.82801986e-01 1.49877763e+00 -1.20611417e+00
-1.03331327e+00 1.01074243e+00 -1.78591922e-01 -7.12142706e-01
4.71492857e-01 -3.87164772e-01 -4.80063371e-02 1.35775700e-01
1.74221694e-01 7.46874392e-01 3.86111826e-01 -1.17030001e+00
-6.70047224e-01 -3.59967500e-01 4.47127193e-01 4.88110811e-01
3.22944701e-01 -3.27125452e-02 6.50710523e-01 -2.95240581e-01
2.23845169e-01 -5.92083395e-01 2.54230171e-01 -1.69711202e-01
-2.06300318e-01 -5.16247153e-01 1.24959864e-01 -6.83271885e-01
8.47340584e-01 -1.81563318e+00 2.79368132e-01 -3.74146514e-02
1.93651810e-01 -7.26806000e-02 -2.05991253e-01 4.24311578e-01
-5.48360765e-01 2.31420413e-01 5.74870892e-02 5.24810195e-01
2.12305069e-01 -1.40149728e-03 -6.47603750e-01 3.47936936e-02
2.05449790e-01 9.24223661e-01 -1.17059374e+00 -9.66169760e-02
-9.12974253e-02 -3.57064009e-01 -7.08037794e-01 1.37745649e-01
-5.44552088e-01 2.88209826e-01 -2.37202495e-01 -9.75541119e-03
4.82225955e-01 -4.07563329e-01 4.08649445e-01 -1.47855118e-01
-2.39555955e-01 1.00775945e+00 -8.51172030e-01 1.41469407e+00
-1.41161710e-01 8.48280251e-01 -2.26641044e-01 -9.83074963e-01
6.36144996e-01 2.18486443e-01 -5.51660061e-01 -9.02429342e-01
2.03138143e-01 3.34715307e-01 1.08167708e+00 -4.82589632e-01
2.10883632e-01 -3.10459912e-01 2.34331325e-01 8.09710503e-01
2.37164900e-01 2.23975498e-02 1.75785899e-01 4.61050510e-01
8.88890922e-01 1.26072943e-01 5.68008780e-01 -9.00534153e-01
4.82553810e-01 1.85190454e-01 1.71107352e-01 1.31848228e+00
-2.94521868e-01 2.56867588e-01 9.00494576e-01 -4.70105499e-01
-1.14550543e+00 -1.27189231e+00 -2.25623727e-01 1.18455064e+00
2.78821766e-01 -1.55549452e-01 -9.40113664e-01 -4.30664152e-01
-1.70410246e-01 1.77645159e+00 -8.42427909e-01 -4.46081161e-01
-4.13588464e-01 -3.04416567e-01 1.97618231e-01 6.25567555e-01
3.25091511e-01 -1.29354405e+00 -1.22388339e+00 1.02411166e-01
-3.11514884e-01 -6.50612235e-01 -1.24394715e-01 2.70515531e-01
-9.58537877e-01 -1.33390355e+00 -2.74299830e-01 -4.96762693e-01
4.34030652e-01 1.32359818e-01 1.53310907e+00 5.62249899e-01
4.06479985e-01 3.42355132e-01 -3.53566349e-01 -7.10432291e-01
-6.90418720e-01 5.54075241e-02 5.71384653e-02 -7.08120644e-01
7.79416800e-01 -4.27849412e-01 -4.88698840e-01 1.80188864e-01
-5.69653511e-01 2.29010984e-01 7.21693695e-01 8.47111523e-01
-1.06549235e-02 -4.27610576e-01 5.85157394e-01 -9.73841786e-01
8.65675271e-01 -4.36822981e-01 -2.91030765e-01 6.35402203e-01
-6.44318104e-01 4.81833488e-01 4.93776709e-01 -2.60999054e-01
-1.38674295e+00 -7.72160113e-01 1.18545689e-01 2.82723606e-01
-3.91390830e-01 3.29198509e-01 -2.35502511e-01 4.03189093e-01
8.76678705e-01 1.54549032e-01 3.19045782e-02 -3.54329556e-01
3.70897591e-01 6.08452894e-02 4.49955493e-01 -1.17006135e+00
6.82817638e-01 1.54453442e-01 -2.93448597e-01 -6.54500246e-01
-1.55574608e+00 3.67310047e-02 -5.08839428e-01 2.48098165e-01
1.08367646e+00 -6.47828519e-01 -7.72919059e-01 7.98221678e-02
-1.38283587e+00 -4.15241063e-01 -5.35997868e-01 5.46619356e-01
-9.73619699e-01 3.09529185e-01 -6.13743603e-01 -4.48102057e-01
3.80154438e-02 -1.12195110e+00 4.72481459e-01 4.63672251e-01
-8.91083717e-01 -9.93000567e-01 -9.43857729e-02 2.86998063e-01
5.00474691e-01 -2.06753165e-01 1.73163521e+00 -1.25556421e+00
-5.22803426e-01 4.70480889e-01 -4.51719671e-01 3.80261336e-03
-3.41587245e-01 -2.85112649e-01 -9.00237203e-01 -1.03641655e-02
5.16251683e-01 -5.26998997e-01 9.07251120e-01 4.05716628e-01
8.81055772e-01 -3.91546875e-01 -2.00380459e-01 2.64700353e-01
1.27764821e+00 1.41920239e-01 7.42675126e-01 4.00693208e-01
1.81468278e-01 1.14741802e+00 5.06775260e-01 -2.15927541e-01
5.61115205e-01 3.24124306e-01 2.90740192e-01 3.17967743e-01
-6.31678179e-02 -6.42832637e-01 2.54545629e-01 5.30025303e-01
-7.70108309e-03 -1.05629233e-03 -1.00999367e+00 7.73357630e-01
-1.95065594e+00 -1.24556983e+00 -2.28343949e-01 2.20623875e+00
9.33607578e-01 4.23435032e-01 -1.77409440e-01 -2.22583726e-01
6.33378625e-01 6.61401898e-02 -7.20586896e-01 -8.02693367e-01
-2.23473907e-01 2.98697621e-01 1.99242905e-01 7.29739547e-01
-5.05254269e-01 9.73250687e-01 7.10888481e+00 4.79626715e-01
-5.87900102e-01 2.96237208e-02 6.94516003e-01 2.84269661e-01
-6.41809046e-01 3.94219279e-01 -5.69217920e-01 3.85026596e-02
1.12169755e+00 -3.67320538e-01 3.23073149e-01 4.89949197e-01
2.95041222e-02 -5.04638433e-01 -1.87879837e+00 2.80477524e-01
-4.17491607e-02 -1.40450609e+00 5.08709967e-01 -2.52487212e-01
5.83791494e-01 -3.10372356e-02 -2.50377536e-01 3.42809469e-01
5.37983179e-01 -1.19397390e+00 1.11098838e+00 4.97093230e-01
1.79264218e-01 -2.92593628e-01 7.51640737e-01 7.08249569e-01
-3.05051267e-01 -1.29588917e-01 -7.40272880e-01 -6.41069472e-01
-1.70565099e-02 -1.17348917e-01 -2.78860837e-01 1.32142812e-01
4.33812201e-01 1.46938592e-01 -6.98847353e-01 6.66597128e-01
-8.77725363e-01 5.32443523e-01 -2.00412162e-02 -2.29661435e-01
1.86422899e-01 9.13819969e-02 4.82528985e-01 7.15058327e-01
2.32552335e-01 1.95739284e-01 -3.74052763e-01 1.19613767e+00
7.26862177e-02 3.00950520e-02 -3.70959610e-01 2.43917853e-01
6.50774837e-01 5.01752615e-01 -5.65888524e-01 -4.72466111e-01
-4.87864852e-01 6.28661096e-01 5.86494505e-01 1.39775068e-01
-6.82581306e-01 8.60189199e-02 4.11943972e-01 1.51883155e-01
3.45388800e-02 1.08788714e-01 -3.52341473e-01 -1.37052488e+00
-2.31310695e-01 -1.03607178e+00 5.27664840e-01 -1.20111847e+00
-1.08023429e+00 1.78854957e-01 1.16476789e-01 -4.12111819e-01
-2.48362377e-01 -9.34315681e-01 -8.39532912e-01 8.86790276e-01
-1.55702209e+00 -6.31328642e-01 5.39427847e-02 3.27643543e-01
6.10756993e-01 2.11454093e-01 8.70177150e-01 -6.26897514e-01
-1.29487917e-01 2.78329134e-01 -1.11132495e-01 -2.38746796e-02
4.70186830e-01 -1.48121202e+00 5.05624652e-01 7.16204762e-01
3.35309714e-01 1.00376904e+00 1.20074677e+00 -3.56197685e-01
-8.47986519e-01 -2.20848218e-01 1.30611312e+00 -8.62662435e-01
7.40236580e-01 -5.12054227e-02 -1.12680328e+00 1.06228495e+00
6.56615436e-01 -8.23195636e-01 3.91164571e-01 4.21343923e-01
-5.94988704e-01 3.22890222e-01 -1.08278489e+00 6.79770470e-01
1.16125679e+00 -4.27776515e-01 -1.72657883e+00 4.48544830e-01
7.71316886e-01 -1.57198370e-01 -4.36951965e-01 6.20666966e-02
3.39925885e-01 -1.49656200e+00 7.41225660e-01 -1.23223484e+00
7.62301564e-01 -1.54205531e-01 -7.82565624e-02 -1.54162848e+00
-5.23425221e-01 -2.96849698e-01 1.49956957e-01 9.38663721e-01
8.38121951e-01 -7.04884171e-01 1.56589806e-01 6.82628930e-01
1.29486233e-01 -4.26655531e-01 -8.14768970e-01 -5.22421360e-01
6.76159680e-01 -1.46105051e-01 6.27227426e-01 9.74726379e-01
3.23003560e-01 9.42143142e-01 1.68163985e-01 -2.27747597e-02
5.48363209e-01 2.29849026e-01 3.85117590e-01 -1.62176359e+00
-2.48153955e-01 -6.60627484e-01 5.02187535e-02 -9.19836760e-01
3.05909365e-01 -8.04498792e-01 -1.79790854e-01 -1.49861312e+00
5.04063725e-01 4.54263277e-02 -1.58931240e-01 2.20704094e-01
-2.47960001e-01 -7.39457250e-01 3.45607668e-01 2.82616436e-01
-4.65905845e-01 1.73383296e-01 1.09206021e+00 4.01087180e-02
2.74135858e-01 -3.91245693e-01 -1.32181060e+00 1.15589428e+00
7.00733721e-01 -7.25696012e-02 -5.22072375e-01 -8.32783818e-01
8.04908156e-01 1.63708150e-01 2.93746352e-01 -9.67824280e-01
6.05120659e-02 -3.10934752e-01 3.96471947e-01 -1.77706361e-01
-3.65905255e-01 -5.35807192e-01 -3.59065175e-01 8.01072717e-01
-1.08199000e+00 3.16900045e-01 1.75971389e-01 2.51636922e-01
1.00423293e-02 -6.43634617e-01 9.36035573e-01 -5.45169532e-01
-8.48782182e-01 -3.38421494e-01 -3.66502345e-01 5.62843084e-01
7.03539133e-01 -5.45975827e-02 -8.94000947e-01 -6.63801134e-01
-8.46237600e-01 3.33802491e-01 7.00679958e-01 4.12728280e-01
2.01414868e-01 -7.77467251e-01 -7.89324045e-01 -1.66105226e-01
1.54372752e-01 -3.16830903e-01 8.04148763e-02 8.39979053e-01
-4.94820058e-01 8.06474209e-01 -2.86102265e-01 -1.03525057e-01
-4.07874882e-01 8.41212034e-01 6.84638023e-01 -2.80131489e-01
-3.48000497e-01 1.03975713e+00 7.74379969e-01 -3.90021861e-01
-1.00699559e-01 -4.38475728e-01 -4.20662642e-01 2.15166435e-01
6.66991413e-01 5.12436070e-02 -2.43492678e-01 -5.32079458e-01
-2.75899112e-01 3.84316474e-01 -3.19435000e-01 1.70423523e-01
1.09755981e+00 -2.47474134e-01 -3.72257866e-02 5.58380485e-01
5.18190503e-01 -9.37819034e-02 -1.10610950e+00 -1.19712532e-01
5.00282049e-01 -2.28448734e-01 -3.80560726e-01 -1.18604124e+00
-3.26373100e-01 1.22940981e+00 1.55467793e-01 2.61981696e-01
5.24125099e-01 3.98151487e-01 3.82611513e-01 5.83581984e-01
3.69597286e-01 -1.04306901e+00 -2.79796869e-01 7.17974842e-01
1.03870285e+00 -9.74025488e-01 -9.45573673e-02 -1.11293688e-01
-4.68138367e-01 1.12606382e+00 9.09982800e-01 -2.55130053e-01
9.95273516e-02 -2.42565140e-01 1.40391970e-02 -3.99851352e-01
-1.15861595e+00 -3.55061710e-01 -4.72678617e-02 6.79913163e-01
7.60822356e-01 -5.87249584e-02 -7.43399560e-01 4.86748874e-01
-6.74146056e-01 -3.68105263e-01 8.66815686e-01 3.09816539e-01
-7.43313015e-01 -5.75819910e-01 -1.96116060e-01 6.01626277e-01
-2.91518539e-01 -2.74961501e-01 -6.42644763e-01 1.07693338e+00
-6.83886185e-02 7.98489392e-01 3.64935011e-01 1.36164159e-01
2.01921806e-01 3.05859149e-01 8.04395556e-01 -7.69086182e-01
-4.67577577e-01 -4.41748798e-01 1.76702634e-01 -6.30277395e-01
-3.53351861e-01 -9.81453180e-01 -1.27326918e+00 -2.21298859e-01
-3.78421009e-01 3.42094243e-01 5.77278959e-04 1.36545491e+00
8.97889435e-02 4.12473679e-01 -9.34529603e-02 -2.78125584e-01
-1.10851729e+00 -1.03147721e+00 -3.63821715e-01 8.22599471e-01
3.33193243e-01 -5.62107444e-01 -6.38601661e-01 -4.16512340e-01] | [9.954545021057129, 7.748218059539795] |
44f79923-3624-4eaf-8e80-7c30184b5144 | learning-sentence-embeddings-using-recursive | 1805.08353 | null | http://arxiv.org/abs/1805.08353v1 | http://arxiv.org/pdf/1805.08353v1.pdf | Learning sentence embeddings using Recursive Networks | Learning sentence vectors that generalise well is a challenging task. In this
paper we compare three methods of learning phrase embeddings: 1) Using LSTMs,
2) using recursive nets, 3) A variant of the method 2 using the POS information
of the phrase. We train our models on dictionary definitions of words to obtain
a reverse dictionary application similar to Felix et al. [1]. To see if our
embeddings can be transferred to a new task we also train and test on the
rotten tomatoes dataset [2]. We train keeping the sentence embeddings fixed as
well as with fine tuning. | ['Anson Bastos'] | 2018-05-22 | null | null | null | null | ['reverse-dictionary'] | ['natural-language-processing'] | [ 2.69116908e-01 1.82852000e-02 -8.26801583e-02 -2.21626386e-01
-2.26639047e-01 -8.29486728e-01 5.93933582e-01 4.70356941e-01
-8.94960105e-01 6.96554184e-01 3.17385942e-01 -4.45613146e-01
4.38940860e-02 -8.95687342e-01 -8.32624137e-01 -5.33778310e-01
-4.75671515e-02 4.81224209e-01 3.16764385e-01 -4.37127590e-01
4.84386086e-01 3.97118628e-01 -9.95797634e-01 1.80970475e-01
3.05922717e-01 4.91212279e-01 3.62564057e-01 9.85051572e-01
-2.73236930e-01 2.78950930e-01 -6.60014331e-01 -4.18492824e-01
4.25584614e-01 -1.62567616e-01 -9.30955887e-01 -1.80601627e-01
5.59249640e-01 -1.36243314e-01 -1.71984881e-01 9.00612295e-01
2.76929498e-01 3.81341457e-01 5.69777429e-01 -6.63564682e-01
-1.07309997e+00 8.09192717e-01 -2.71071613e-01 3.74350309e-01
3.24491322e-01 -5.52038550e-02 1.24495029e+00 -8.68553579e-01
6.51397705e-01 1.15335906e+00 8.34932387e-01 4.84908611e-01
-1.45754933e+00 -3.97531539e-01 7.26505890e-02 2.35693797e-01
-1.19201756e+00 -1.88008189e-01 3.74907196e-01 -2.37427130e-01
1.58992732e+00 -9.79521498e-02 6.02438450e-01 1.42239809e+00
3.17872792e-01 4.49467421e-01 1.08428085e+00 -7.93580532e-01
-9.83986780e-02 1.92222178e-01 4.73710090e-01 5.20083427e-01
1.32080972e-01 1.25460744e-01 -3.69964868e-01 -1.15096845e-01
6.23805702e-01 -6.10313565e-02 -2.09220260e-01 -2.42765218e-01
-1.48570251e+00 1.17450500e+00 5.01691401e-01 7.46297717e-01
-3.80679429e-01 4.61108088e-01 6.65558517e-01 5.84121346e-01
3.81768316e-01 7.97676921e-01 -8.50560367e-01 -2.39047664e-03
-7.54693687e-01 2.94706404e-01 1.06147254e+00 9.89431977e-01
6.66151226e-01 -6.39226809e-02 -1.01336040e-01 9.01350379e-01
1.62855357e-01 2.70466417e-01 6.58140957e-01 -5.36454976e-01
2.68807560e-01 -1.22712359e-01 -1.00814000e-01 -7.56606698e-01
-3.84967268e-01 -2.41870895e-01 -4.74685729e-01 -3.86924297e-01
2.65896320e-01 -2.00492918e-01 -1.03876805e+00 1.66622138e+00
-2.46663079e-01 1.84977770e-01 -2.28616851e-03 4.39887345e-01
8.11510861e-01 8.76785994e-01 2.73651123e-01 -1.51035106e-02
1.12370861e+00 -1.05065596e+00 -7.33582675e-01 -2.47650325e-01
7.49832749e-01 -8.18425894e-01 9.84671533e-01 6.05881453e-01
-8.52389932e-01 -7.57077277e-01 -1.10711324e+00 -2.57010996e-01
-8.91220450e-01 -1.22005232e-01 4.43734288e-01 5.64242065e-01
-1.35271227e+00 1.01317704e+00 -8.62766325e-01 -9.48412001e-01
-3.79281938e-02 5.77113986e-01 -3.72723967e-01 -2.30051130e-02
-1.40024412e+00 1.57187736e+00 7.61946082e-01 -2.26079091e-01
-6.96351349e-01 -3.74870479e-01 -1.06241167e+00 1.57405362e-01
3.77035104e-02 -7.10466385e-01 1.17696488e+00 -7.86612332e-01
-1.49072850e+00 9.66285348e-01 -1.64845139e-01 -6.64452136e-01
-3.40492576e-01 -2.37817436e-01 -3.65334637e-02 -9.66612548e-02
3.48975770e-02 9.39664781e-01 6.92363918e-01 -9.12790716e-01
-5.61476767e-01 -1.25468090e-01 4.71464753e-01 -4.61444929e-02
-5.21385550e-01 2.54315827e-02 8.96673575e-02 -8.92078280e-01
-3.45527679e-01 -9.44005311e-01 -2.12357342e-01 -1.17342055e-01
-3.32133532e-01 -5.29655099e-01 4.65150744e-01 -6.80888712e-01
1.15194726e+00 -1.95178735e+00 5.54865658e-01 4.31631915e-02
2.90628430e-02 4.67363745e-01 -7.65069544e-01 8.15029860e-01
-1.81339994e-01 3.63355368e-01 -2.41024017e-01 -4.15978968e-01
3.21143478e-01 8.43676567e-01 -5.43735743e-01 2.30230391e-01
5.55516422e-01 8.75309527e-01 -9.93161082e-01 -2.09858626e-01
1.18207268e-01 5.15631974e-01 -5.95695436e-01 8.78603682e-02
-2.46559739e-01 6.85186908e-02 3.13011999e-03 2.51068950e-01
3.10884058e-01 1.56851396e-01 4.99055460e-02 -2.07543448e-01
-1.45370796e-01 6.64893568e-01 -8.47450078e-01 1.92007124e+00
-7.91937947e-01 7.58138955e-01 -3.57553273e-01 -1.22803795e+00
1.03895676e+00 4.02700901e-01 1.40211098e-02 -2.83835858e-01
1.60264373e-01 1.63023442e-01 7.01835379e-02 -5.12247980e-01
7.34723508e-01 -5.57376027e-01 -8.24905634e-02 4.06988829e-01
8.09781551e-01 -1.96687996e-01 4.87047940e-01 -7.29930848e-02
1.24301004e+00 3.36575359e-01 4.85131055e-01 -5.35316527e-01
3.91317487e-01 -4.39471193e-02 1.74448997e-01 7.78543770e-01
-6.57636076e-02 5.14614344e-01 4.22593534e-01 -3.75971645e-01
-1.09312260e+00 -1.22025049e+00 -2.61654109e-01 1.37945354e+00
-3.52994412e-01 -7.66209245e-01 -4.14251685e-01 -7.56934047e-01
-2.02646218e-02 1.05748630e+00 -8.09107959e-01 -1.04008570e-01
-6.86013579e-01 -4.45433408e-01 6.41825259e-01 8.22250962e-01
1.24532543e-01 -1.21964800e+00 -5.43996036e-01 1.80613473e-01
9.06940997e-02 -9.02209163e-01 -4.69569683e-01 8.72340083e-01
-7.48007059e-01 -6.24094248e-01 -5.98575234e-01 -1.23507893e+00
3.20318937e-01 -1.03723139e-01 1.48935783e+00 1.95040286e-01
-9.66068450e-03 5.21354556e-01 -7.41358042e-01 -4.94689375e-01
-3.77817273e-01 4.30458963e-01 1.24826834e-01 -5.69704711e-01
8.02918494e-01 -5.59343994e-01 3.01108286e-02 -1.89263865e-01
-1.21216297e+00 -5.51675916e-01 4.73668069e-01 1.12551582e+00
4.99563366e-01 -1.62248507e-01 4.91161555e-01 -1.04254007e+00
8.46571147e-01 -4.99428362e-01 -4.26735729e-01 8.41697603e-02
-3.22342902e-01 2.54603684e-01 6.68107033e-01 -6.26164138e-01
-3.67150038e-01 2.26515636e-01 -6.32636189e-01 -2.72305071e-01
-3.03777754e-01 6.03151500e-01 2.88963139e-01 -4.61260565e-02
6.25789762e-01 4.16967422e-02 -3.49379539e-01 -4.06575471e-01
6.96795821e-01 6.61552250e-01 1.42029330e-01 -4.73482013e-01
8.96100044e-01 6.03004284e-02 -2.21799582e-01 -9.01158810e-01
-9.81720686e-01 -4.74021018e-01 -9.92505133e-01 3.42685342e-01
1.16078174e+00 -4.36886847e-01 -2.57132292e-01 -3.36659737e-02
-1.68907106e+00 -3.75666678e-01 -5.63906908e-01 4.02388245e-01
-4.82971489e-01 3.46150249e-01 -7.76288807e-01 -2.39163965e-01
-1.85374752e-01 -8.71450901e-01 9.67466116e-01 1.02383606e-01
-5.04721522e-01 -1.60038567e+00 7.19083965e-01 -3.32473636e-01
5.88619888e-01 -1.90187863e-03 9.10880744e-01 -1.15798879e+00
3.54970731e-02 -1.90903116e-02 -3.82528007e-02 8.22775602e-01
1.07118636e-01 -8.11798126e-02 -8.64891708e-01 -3.60566884e-01
2.82585889e-01 -4.72469121e-01 1.04971170e+00 2.11932883e-01
8.29118788e-01 -2.23218754e-01 -2.22250164e-01 5.03730953e-01
1.65763628e+00 7.23375380e-02 6.62984908e-01 3.98604035e-01
5.66614449e-01 5.25436223e-01 3.03644627e-01 -4.49441420e-03
2.10701913e-01 3.46926570e-01 3.29605490e-02 1.24778979e-01
1.60257638e-01 -3.90370160e-01 5.10081589e-01 1.24064267e+00
2.39952922e-01 -4.50718939e-01 -8.42544794e-01 1.05433595e+00
-1.70732784e+00 -8.93903017e-01 -2.56588329e-02 1.84506297e+00
8.62215757e-01 3.21072340e-01 4.78065293e-03 6.00606613e-02
4.07908797e-01 2.56828308e-01 7.17421323e-02 -1.32191193e+00
1.63753286e-01 1.21359575e+00 6.35572553e-01 5.35618722e-01
-1.23607516e+00 1.13498151e+00 7.16443920e+00 5.47240496e-01
-1.12294531e+00 2.90915489e-01 -2.47750413e-02 1.90234885e-01
-2.05513552e-01 1.73562706e-01 -8.83406878e-01 2.64937311e-01
1.20224440e+00 -1.84475502e-03 4.92156684e-01 5.16250372e-01
-4.09857094e-01 8.15469250e-02 -1.43402815e+00 7.38193154e-01
5.08935869e-01 -1.04084527e+00 1.58869952e-01 -2.66742945e-01
4.94926721e-01 2.50579417e-01 -9.98416841e-02 7.75828362e-01
3.20891738e-01 -1.27423203e+00 2.74041772e-01 5.53625524e-01
3.18937659e-01 -5.30414104e-01 8.05054843e-01 3.58501524e-01
-1.02483308e+00 1.93313628e-01 -7.88653255e-01 -3.51370662e-01
2.26330027e-01 2.58992672e-01 -8.06845903e-01 3.50984663e-01
5.42822361e-01 1.02505875e+00 -7.85120010e-01 8.64543378e-01
-5.98960400e-01 7.05639184e-01 -4.06370521e-01 -2.72232503e-01
5.04999220e-01 -9.83972251e-02 4.93788004e-01 1.70620942e+00
2.22747400e-01 -1.49329737e-01 1.04663059e-01 1.02395391e+00
4.48385179e-02 1.32198464e-02 -1.06952024e+00 -3.00224304e-01
1.98449329e-01 1.25487339e+00 -6.39912426e-01 -2.57109940e-01
-4.96812761e-01 1.06352246e+00 3.36020052e-01 2.42359251e-01
-6.22020960e-01 -8.81196082e-01 3.80372941e-01 -2.03329504e-01
8.90471458e-01 -5.87623537e-01 -4.45355624e-02 -1.05940986e+00
-3.17434669e-01 -5.19387126e-01 1.16273507e-01 -7.40363240e-01
-1.81395447e+00 6.79254949e-01 1.72289073e-01 -8.00627410e-01
-1.94637731e-01 -1.07668996e+00 -7.94036984e-01 9.70139861e-01
-1.47308648e+00 -1.07734966e+00 2.09314525e-01 2.48922735e-01
5.98325014e-01 8.65987614e-02 1.25772023e+00 3.16396624e-01
-3.31972539e-01 4.99801815e-01 -3.03198434e-02 2.18010664e-01
8.31050336e-01 -1.55277157e+00 6.25102580e-01 5.21727264e-01
5.63745320e-01 1.01706207e+00 7.06586003e-01 -2.50229001e-01
-1.14545226e+00 -1.00620890e+00 1.51669347e+00 -8.00839841e-01
1.02960205e+00 -6.42947495e-01 -1.07712853e+00 1.22242546e+00
9.99811769e-01 -2.87826389e-01 8.60713124e-01 3.35184306e-01
-3.41784269e-01 1.61645293e-01 -9.95865285e-01 4.10047024e-01
8.99052083e-01 -6.22371435e-01 -1.30137646e+00 6.22628570e-01
1.04934216e+00 -2.27612868e-01 -1.05198312e+00 3.16246033e-01
3.63671333e-01 -5.98566055e-01 7.51145899e-01 -8.47543299e-01
4.64529842e-01 -7.95920417e-02 -3.09939802e-01 -1.76658094e+00
-5.39347708e-01 -1.89051285e-01 2.16290131e-01 1.24115884e+00
5.25051057e-01 -7.83105969e-01 1.13300510e-01 -5.33103570e-02
-1.58454832e-02 -8.51163089e-01 -7.06878006e-01 -9.07951951e-01
6.91494465e-01 -2.19076112e-01 2.35833049e-01 1.00457227e+00
2.45035335e-01 8.04326534e-01 -1.42172994e-02 -9.83449146e-02
1.82972580e-01 -2.82471240e-01 4.24123436e-01 -9.51064348e-01
-5.25851130e-01 -4.71576840e-01 -4.94859993e-01 -1.22297204e+00
5.33913493e-01 -1.22506225e+00 1.73926935e-01 -1.55333900e+00
3.22065800e-02 -1.23323150e-01 -4.46261376e-01 5.83705008e-01
-1.41372969e-02 4.35284525e-01 3.27019423e-01 -2.70093828e-01
-4.87941593e-01 4.77156341e-01 8.89625847e-01 -3.18176508e-01
2.00439364e-01 -4.95180517e-01 -5.10200500e-01 5.14031351e-01
1.01735914e+00 -7.31456578e-01 -1.80990204e-01 -8.39347541e-01
4.40631121e-01 -3.85722339e-01 3.24174672e-01 -7.64536440e-01
4.65553924e-02 1.14909224e-01 2.37115040e-01 -5.11574209e-01
1.37343958e-01 -5.85412443e-01 -3.59992981e-01 5.47224164e-01
-5.52541494e-01 5.13112009e-01 5.87898791e-01 4.94960815e-01
-2.32473627e-01 -1.00064325e+00 5.49705565e-01 -2.74034709e-01
-5.26498139e-01 -7.50696212e-02 -5.76431513e-01 -1.06236309e-01
7.50362337e-01 9.20344070e-02 -1.42535061e-01 -3.32284868e-02
-1.05212414e+00 1.63440078e-01 4.98616368e-01 5.61922073e-01
4.39312309e-01 -1.29736257e+00 -6.26083732e-01 2.30096966e-01
-3.40552442e-02 -2.87398756e-01 -5.18065751e-01 5.89362621e-01
-7.26372242e-01 6.34156168e-01 -3.14277291e-01 -5.17434418e-01
-9.09252882e-01 9.19927716e-01 1.33618042e-01 -3.52748245e-01
-2.48391286e-01 8.99576306e-01 1.96705870e-02 -7.71215796e-01
2.93430626e-01 -6.29383922e-01 -5.14958918e-01 5.70630580e-02
2.84381390e-01 -3.86868156e-02 -8.54337886e-02 -4.21498001e-01
-2.90552378e-01 6.74746394e-01 -1.55935466e-01 -8.86928216e-02
1.47501457e+00 1.21530466e-01 -3.72397453e-01 8.70510817e-01
1.45458543e+00 -1.44872770e-01 -4.64940518e-01 -3.23095113e-01
3.32740724e-01 -1.71029717e-01 -1.17686048e-01 -5.02125144e-01
-7.99137950e-01 1.10929596e+00 4.76047546e-01 4.15944993e-01
9.96226609e-01 -2.87449390e-01 1.05696273e+00 6.51427507e-01
3.25318694e-01 -9.64444637e-01 -2.41521709e-02 1.03252518e+00
8.53456795e-01 -8.97741139e-01 -6.90252483e-02 2.02191651e-01
-3.30434382e-01 1.47936690e+00 3.07952136e-01 -9.45154965e-01
9.28881347e-01 1.97227627e-01 -7.83411935e-02 -1.24389358e-01
-9.32587206e-01 -3.68574798e-01 7.16388896e-02 7.28512466e-01
7.85274446e-01 -2.31824100e-01 -7.74054408e-01 4.08045650e-01
-2.00365305e-01 8.86468664e-02 6.30918026e-01 1.01813173e+00
-3.85157347e-01 -1.49238992e+00 -7.70476088e-02 5.26432514e-01
-3.12842011e-01 -5.27311921e-01 -6.33401990e-01 7.34273136e-01
3.89400840e-01 6.89512551e-01 1.13361493e-01 -5.33642769e-01
4.97857243e-01 3.43093991e-01 7.24650443e-01 -1.39611101e+00
-9.98551667e-01 -2.72763371e-01 1.69812545e-01 -1.61754385e-01
-5.86328030e-01 -4.34373438e-01 -1.09731829e+00 -1.57642722e-01
-5.38182497e-01 -4.14558686e-02 6.76060379e-01 1.01165330e+00
1.79603975e-02 5.64638317e-01 2.81721294e-01 -1.02850699e+00
-8.81993055e-01 -1.32435846e+00 -5.05211115e-01 3.28771830e-01
2.76538581e-01 -4.89312232e-01 -4.13786113e-01 1.69803873e-01] | [10.67034912109375, 8.715271949768066] |
a73b4ae8-8fa9-4cc7-8f69-5ac33ec8f68f | measuring-massive-multitask-language | 2009.03300 | null | https://arxiv.org/abs/2009.03300v3 | https://arxiv.org/pdf/2009.03300v3.pdf | Measuring Massive Multitask Language Understanding | We propose a new test to measure a text model's multitask accuracy. The test covers 57 tasks including elementary mathematics, US history, computer science, law, and more. To attain high accuracy on this test, models must possess extensive world knowledge and problem solving ability. We find that while most recent models have near random-chance accuracy, the very largest GPT-3 model improves over random chance by almost 20 percentage points on average. However, on every one of the 57 tasks, the best models still need substantial improvements before they can reach expert-level accuracy. Models also have lopsided performance and frequently do not know when they are wrong. Worse, they still have near-random accuracy on some socially important subjects such as morality and law. By comprehensively evaluating the breadth and depth of a model's academic and professional understanding, our test can be used to analyze models across many tasks and to identify important shortcomings. | ['Andy Zou', 'Dan Hendrycks', 'Mantas Mazeika', 'Collin Burns', 'Steven Basart', 'Jacob Steinhardt', 'Dawn Song'] | 2020-09-07 | null | null | null | null | ['multi-task-language-understanding', 'elementary-mathematics'] | ['methodology', 'reasoning'] | [-2.62778848e-01 3.62349570e-01 -5.96900940e-01 -1.99164540e-01
-1.02410614e+00 -7.42681384e-01 4.23004389e-01 2.98379749e-01
-4.01255786e-01 1.04800975e+00 -4.26276959e-02 -1.03276491e+00
-5.92621624e-01 -6.80464149e-01 -6.38676107e-01 -1.58598423e-01
5.11431754e-01 5.72007835e-01 -5.26159480e-02 -1.72908664e-01
8.37211668e-01 -9.25527960e-02 -1.11737955e+00 1.78962767e-01
1.89943850e+00 4.24050272e-01 -7.91555271e-02 5.56610882e-01
1.20754447e-02 1.11976933e+00 -6.43611610e-01 -1.04315734e+00
-1.44855663e-01 -6.71419352e-02 -1.23004353e+00 -6.53373003e-01
7.81136453e-01 -2.57510096e-01 -5.56247123e-02 1.07487535e+00
3.45873713e-01 1.95452601e-01 1.05032670e+00 -1.20910323e+00
-9.73768473e-01 8.00472617e-01 -6.11699462e-01 3.57349545e-01
7.17192948e-01 2.03597292e-01 1.33310938e+00 -4.70276147e-01
4.16489214e-01 1.22174287e+00 8.54483604e-01 3.30701798e-01
-1.20575202e+00 -1.12442768e+00 2.17195526e-01 3.04033905e-01
-1.04261434e+00 -3.20709139e-01 3.29907030e-01 -8.78894866e-01
8.52797091e-01 8.67465138e-02 5.62129319e-01 1.08826923e+00
5.95543623e-01 5.15766442e-01 1.52597630e+00 -1.82171598e-01
-6.55557960e-02 2.65375972e-01 5.97963452e-01 6.24183595e-01
7.83708274e-01 -1.96031108e-01 -5.95633209e-01 -2.35093206e-01
5.12188315e-01 -4.11816806e-01 -5.82163893e-02 2.09958091e-01
-1.21752083e+00 6.67556942e-01 3.05012856e-02 5.96978068e-01
-1.36328131e-01 2.01507241e-01 -1.42677084e-01 5.35968482e-01
5.66872060e-01 1.04878867e+00 -7.54117489e-01 -6.99172020e-01
-8.50098550e-01 6.60935104e-01 1.04957151e+00 6.08727336e-01
4.48731929e-01 -2.02243909e-01 -2.51541913e-01 7.59981811e-01
-1.50126619e-02 4.45442528e-01 4.87197965e-01 -1.34316409e+00
6.79591477e-01 6.72805607e-01 3.36840659e-01 -1.20478046e+00
-4.01905090e-01 -8.58588994e-01 -5.12195766e-01 1.58558935e-01
9.51433778e-01 -2.41439700e-01 -5.82052648e-01 1.58478677e+00
-1.30560160e-01 3.70914899e-02 -1.90481052e-01 4.13404137e-01
9.06589389e-01 3.76387537e-01 5.92342675e-01 -3.35280634e-02
1.27600646e+00 -5.90410113e-01 -5.53059638e-01 -6.78383648e-01
1.01905835e+00 -8.08301985e-01 1.26009607e+00 7.75529563e-01
-1.42550051e+00 -6.18644297e-01 -8.21814418e-01 -3.06035906e-01
-3.64573419e-01 -2.16066763e-01 7.42953062e-01 9.68231857e-01
-9.28214967e-01 9.44024205e-01 -2.23747641e-01 -6.92003518e-02
5.03025651e-01 6.10977486e-02 -1.91308439e-01 -2.11411566e-01
-1.22422040e+00 1.51080370e+00 2.41457388e-01 -7.28487372e-01
-6.87127769e-01 -1.12759519e+00 -5.80074728e-01 3.99568260e-01
2.64971733e-01 -8.06424022e-01 1.38555360e+00 -4.84440863e-01
-1.07566428e+00 9.76351798e-01 -2.88471878e-01 -1.11918792e-01
6.13818049e-01 -2.04768732e-01 -1.32664710e-01 -2.04964966e-01
4.07974154e-01 1.58209726e-01 2.39698678e-01 -9.39899623e-01
-7.00666189e-01 -2.51188219e-01 3.01562458e-01 3.23836505e-01
-6.12957656e-01 4.77738380e-02 2.69212753e-01 -5.71111560e-01
-8.60731527e-02 -6.12189293e-01 7.80790821e-02 -4.85336214e-01
-1.19132929e-01 -8.18133891e-01 2.89620340e-01 -6.85432613e-01
1.49321389e+00 -1.67957544e+00 -2.74255220e-03 2.11819038e-02
7.28857100e-01 3.81035581e-02 -5.13614621e-03 1.07290827e-01
-2.27143485e-02 8.49343121e-01 2.42604584e-01 -2.20940169e-02
2.09597692e-01 -2.74191648e-01 3.27069797e-02 2.27252766e-01
-1.74779490e-01 1.19680703e+00 -9.39569235e-01 -7.14824021e-01
-1.24757499e-01 -1.60142407e-01 -5.18140435e-01 -2.74415880e-01
-7.74702206e-02 2.02076197e-01 -5.77043295e-01 5.31912982e-01
2.96199113e-01 -8.41656804e-01 -3.90052535e-02 5.87329328e-01
1.11141674e-01 5.33064008e-01 -6.61378324e-01 1.32983184e+00
-5.64536572e-01 8.00880373e-01 -1.36468813e-01 -9.88137364e-01
6.65035844e-01 3.33911061e-01 1.84852555e-01 -6.86714709e-01
-2.03201384e-03 3.52205247e-01 6.44463897e-01 -3.99032682e-01
4.79919583e-01 -1.54206574e-01 8.21956620e-02 6.68743908e-01
-3.43786746e-01 -5.03663480e-01 1.12207875e-01 2.84294963e-01
1.11019671e+00 -1.20297566e-01 4.19657618e-01 -7.00753808e-01
2.65692413e-01 2.38730013e-01 5.69007039e-01 1.13878047e+00
-1.59193203e-01 -1.02853797e-01 6.55097067e-01 -3.71323586e-01
-7.94407189e-01 -7.43885100e-01 -2.49421284e-01 1.73175848e+00
-2.37742178e-02 -3.17034513e-01 -6.42778754e-01 -8.33196521e-01
2.72486985e-01 1.42236102e+00 -6.20902359e-01 -2.22833127e-01
-4.42123711e-02 -7.92713046e-01 4.46869165e-01 3.30719382e-01
6.01819515e-01 -5.54003716e-01 -2.02015549e-01 6.60594031e-02
-5.94864786e-01 -1.16666758e+00 -1.05151631e-01 -2.33407810e-01
-9.42609131e-01 -1.26520574e+00 -5.07295251e-01 -5.41829705e-01
3.30100507e-01 1.57459453e-01 1.48789799e+00 5.38233519e-01
2.65636533e-01 3.64241749e-01 -1.66027576e-01 -8.57327759e-01
-4.57595021e-01 4.00546908e-01 -2.13996284e-02 -8.91446233e-01
7.05531061e-01 -5.71511924e-01 -2.70932019e-01 4.23525840e-01
-1.30546585e-01 1.90656092e-02 6.52319312e-01 6.00532353e-01
-2.51579255e-01 1.23112388e-01 8.72915983e-01 -9.93880808e-01
1.09220183e+00 -6.32219315e-01 -1.62652329e-01 5.94561815e-01
-1.19822478e+00 -3.58756125e-01 5.23772895e-01 -5.53088725e-01
-9.91013348e-01 -7.46766746e-01 1.83394387e-01 2.95999289e-01
-8.39491710e-02 6.85110927e-01 3.36996466e-01 -3.24888080e-01
1.17201769e+00 -8.54480341e-02 -3.85084212e-01 -2.51288891e-01
-1.52916133e-01 5.64724147e-01 1.89812467e-01 -1.23526585e+00
9.92413163e-01 -2.06150532e-01 -1.15830027e-01 -6.20744407e-01
-1.54502857e+00 -1.40547708e-01 -2.95651525e-01 -2.80041754e-01
7.74121106e-01 -8.30065787e-01 -1.16055346e+00 2.90723026e-01
-1.05141521e+00 -6.25050783e-01 2.88816124e-01 6.91246271e-01
-3.38912159e-01 3.22997898e-01 -5.83798885e-01 -7.79451847e-01
-1.69847608e-01 -1.04961324e+00 6.91968560e-01 3.63318354e-01
-5.95821202e-01 -1.38211322e+00 -1.66160002e-01 9.87536371e-01
3.85694832e-01 -1.02615684e-01 1.47241127e+00 -9.71002221e-01
-1.49925455e-01 -1.63008049e-01 -2.57202327e-01 1.32809922e-01
-2.12354884e-01 -3.96823920e-02 -8.30843151e-01 -6.22827932e-02
-2.51475632e-01 -7.29824424e-01 8.03770542e-01 6.11554146e-01
1.32750082e+00 -1.68994114e-01 -5.01466155e-01 2.95766741e-01
9.02244508e-01 3.58285517e-01 5.27370095e-01 4.63197082e-01
3.56678128e-01 4.84336406e-01 5.19233763e-01 -2.89706001e-03
8.33267987e-01 2.55455345e-01 -1.69769213e-01 2.52653956e-01
2.66721100e-01 -2.35041410e-01 2.91806042e-01 6.25065029e-01
-6.63490593e-01 -2.16784209e-01 -1.58221209e+00 4.43229795e-01
-1.69406724e+00 -1.15854311e+00 -2.59054571e-01 1.99206674e+00
9.00737703e-01 6.60609424e-01 2.97452677e-02 1.45927772e-01
4.24710900e-01 -9.34651718e-02 -5.08844614e-01 -5.27728975e-01
1.14044167e-01 2.47639149e-01 2.96737075e-01 6.41258657e-01
-8.35860074e-01 1.20172334e+00 8.28232861e+00 1.20243728e+00
-4.46309060e-01 1.55555561e-01 1.11286271e+00 2.42055655e-01
-5.75011075e-01 -5.93181513e-02 -7.30788529e-01 3.73996705e-01
8.53125095e-01 -7.17884362e-01 2.45673686e-01 9.67874765e-01
-3.19243222e-03 -3.52072150e-01 -1.13764000e+00 7.81743884e-01
-1.10804625e-01 -1.00497174e+00 -1.39763981e-01 3.23732674e-01
1.06114388e+00 -4.96361732e-01 3.24206233e-01 8.02026212e-01
9.74763274e-01 -1.66600215e+00 4.53221947e-01 5.26326001e-01
7.35354960e-01 -5.08960903e-01 5.79902172e-01 8.41169953e-01
-4.91072565e-01 -2.31111303e-01 -3.38945985e-01 -9.22388315e-01
-3.95903349e-01 5.50152719e-01 -5.18610716e-01 3.17662865e-01
5.25393188e-01 7.03318000e-01 -7.99744844e-01 8.75618219e-01
-3.96672785e-01 9.00530577e-01 -4.41235192e-02 -3.36288005e-01
7.00160116e-02 3.70765179e-02 1.88560650e-01 8.75682294e-01
3.28523189e-01 5.75007498e-01 9.09825861e-02 8.19289029e-01
-1.73961595e-01 1.60944894e-01 -6.77491665e-01 -2.16625050e-01
6.70596778e-01 1.00197637e+00 -4.93786454e-01 -5.55664361e-01
-3.42362344e-01 4.35092330e-01 3.75878185e-01 4.37350214e-01
-7.84362733e-01 -2.76448309e-01 6.57515228e-01 7.58642033e-02
-5.71321189e-01 -2.47963756e-01 -1.01020920e+00 -1.12067580e+00
-2.61008680e-01 -1.04452181e+00 2.79430151e-01 -8.53230715e-01
-1.45410633e+00 -1.43261075e-01 -3.95118184e-02 -4.97533709e-01
-1.84723020e-01 -9.13121760e-01 -8.55529308e-01 7.55105615e-01
-1.25848663e+00 -9.30192947e-01 -3.21263939e-01 3.58870089e-01
2.73103029e-01 -3.82785439e-01 6.10380411e-01 1.03326879e-01
-5.42317867e-01 7.00496972e-01 -1.54585227e-01 2.63838172e-01
9.87814724e-01 -1.45548224e+00 4.33500439e-01 5.09701967e-01
-2.62052447e-01 6.26962423e-01 7.15314209e-01 -8.80915940e-01
-7.63926625e-01 -4.26662982e-01 1.11782193e+00 -1.11701858e+00
7.91243553e-01 1.14383459e-01 -7.66525745e-01 1.03420460e+00
1.25506729e-01 -8.10403764e-01 8.71913850e-01 8.69133055e-01
-4.75933790e-01 1.85354888e-01 -1.02793288e+00 5.21387458e-01
1.36590111e+00 -3.52546036e-01 -1.10199976e+00 7.54165173e-01
4.25080240e-01 -4.54024225e-01 -1.24787807e+00 2.74752557e-01
8.61678898e-01 -9.21269953e-01 9.08999324e-01 -1.16901696e+00
1.07159364e+00 6.13104224e-01 2.58972764e-01 -1.45128644e+00
-7.86281526e-01 -4.33134377e-01 1.62425414e-01 9.11918223e-01
8.11916292e-01 -8.24804127e-01 7.01422036e-01 1.21614015e+00
-3.52650397e-02 -7.59358943e-01 -7.26044118e-01 -8.93776000e-01
1.06546164e+00 -6.32411897e-01 4.88015592e-01 1.55971837e+00
3.02056968e-01 4.01810169e-01 -9.86058079e-03 -1.25971124e-01
8.09318066e-01 9.30892453e-02 7.34466314e-01 -1.87973535e+00
-6.30030483e-02 -1.02219236e+00 -5.87637350e-02 -9.48385298e-01
5.84752202e-01 -9.16366100e-01 -4.66341972e-01 -1.96368909e+00
4.86863911e-01 -5.13448596e-01 -1.08874843e-01 5.99921048e-01
-6.56012595e-01 -1.41207995e-02 3.88550572e-02 3.82381375e-03
-7.03302264e-01 -7.00987428e-02 1.52883911e+00 -9.06288158e-03
6.80833906e-02 -1.91456839e-01 -1.39282072e+00 1.09208882e+00
9.96754706e-01 -4.42957103e-01 -2.69741088e-01 -6.27186239e-01
6.46618187e-01 1.45414229e-02 3.87657404e-01 -1.06929886e+00
4.03051108e-01 -9.17204022e-01 7.29931951e-01 6.24356158e-02
2.03748479e-01 -3.69787365e-01 -1.89728633e-01 4.77954149e-01
-3.93752366e-01 1.59286797e-01 6.65130392e-02 9.83737782e-02
1.53650820e-01 -3.07402879e-01 7.08052516e-01 -3.57956290e-01
-2.67649353e-01 -3.87826469e-03 -3.99799883e-01 5.83654523e-01
1.23807883e+00 -2.23714560e-01 -9.50538635e-01 -6.41439557e-01
-6.34193599e-01 6.02240264e-01 4.19107705e-01 2.72653729e-01
3.21154714e-01 -9.26603794e-01 -9.22646105e-01 -4.25254375e-01
-1.52684540e-01 -3.80890459e-01 2.23415568e-01 8.80925894e-01
-2.94725507e-01 7.10657954e-01 -1.02973627e-02 -9.10615847e-02
-1.00993538e+00 1.77503884e-01 3.59826207e-01 -5.24693072e-01
-4.16361652e-02 8.76752377e-01 1.28411487e-01 -2.59756684e-01
4.80176769e-02 -2.13435441e-01 -3.52327138e-01 2.40891203e-02
4.24809158e-01 8.41302514e-01 -3.42252076e-01 -1.79418057e-01
-6.84482753e-02 8.81281734e-01 -4.86166291e-02 -1.92927057e-03
1.15940452e+00 2.60437787e-01 -1.25264034e-01 4.58362311e-01
4.79255319e-01 1.73692420e-01 -3.79988581e-01 1.32230490e-01
1.79924771e-01 -6.47460282e-01 6.94220141e-02 -1.27865660e+00
-4.78300750e-01 8.40972126e-01 -2.36273870e-01 3.16453546e-01
5.37888169e-01 -2.05566451e-01 4.14820015e-01 7.82973766e-01
4.64615434e-01 -1.36753297e+00 1.34749770e-01 6.09582245e-01
5.18175423e-01 -1.45103633e+00 3.58188421e-01 -3.38943869e-01
-5.89546740e-01 8.81943405e-01 1.07968450e+00 2.60446936e-01
6.99963450e-01 -9.35031921e-02 -2.24006653e-01 -2.41858378e-01
-1.09341645e+00 4.20435145e-02 6.09880567e-01 4.16231096e-01
7.80584395e-01 1.02671765e-01 -6.71423137e-01 8.04085732e-01
-8.21134210e-01 9.93011594e-02 5.82699239e-01 3.19178581e-01
-8.56926382e-01 -7.42049873e-01 -4.20856744e-01 9.26112115e-01
-8.01338911e-01 -1.98967174e-01 -6.16119385e-01 8.95754337e-01
5.54892756e-02 1.28986430e+00 -1.77011892e-01 -5.43029130e-01
5.12613431e-02 5.16625226e-01 4.93282288e-01 -5.87931454e-01
-6.58701479e-01 -5.51008105e-01 2.92826295e-01 -4.12840217e-01
-2.85519689e-01 -4.47404891e-01 -6.75048113e-01 -1.11902940e+00
-2.11872265e-01 3.74707103e-01 2.83217460e-01 1.13560903e+00
4.51441407e-02 2.93024838e-01 4.17461433e-02 -2.14562267e-01
-6.36404514e-01 -1.30087590e+00 -4.46584284e-01 2.25935280e-01
-3.17614563e-02 -8.98420632e-01 -5.69053829e-01 -3.53948653e-01] | [9.820452690124512, 7.507683753967285] |
9b08a1ea-cad9-4b6d-980b-28309078a461 | hicu-leveraging-hierarchy-for-curriculum | 2208.02301 | null | https://arxiv.org/abs/2208.02301v1 | https://arxiv.org/pdf/2208.02301v1.pdf | HiCu: Leveraging Hierarchy for Curriculum Learning in Automated ICD Coding | There are several opportunities for automation in healthcare that can improve clinician throughput. One such example is assistive tools to document diagnosis codes when clinicians write notes. We study the automation of medical code prediction using curriculum learning, which is a training strategy for machine learning models that gradually increases the hardness of the learning tasks from easy to difficult. One of the challenges in curriculum learning is the design of curricula -- i.e., in the sequential design of tasks that gradually increase in difficulty. We propose Hierarchical Curriculum Learning (HiCu), an algorithm that uses graph structure in the space of outputs to design curricula for multi-label classification. We create curricula for multi-label classification models that predict ICD diagnosis and procedure codes from natural language descriptions of patients. By leveraging the hierarchy of ICD codes, which groups diagnosis codes based on various organ systems in the human body, we find that our proposed curricula improve the generalization of neural network-based predictive models across recurrent, convolutional, and transformer-based architectures. Our code is available at https://github.com/wren93/HiCu-ICD. | ['Rahul G. Krishnan', 'Tianshu Zhu', 'Tongzi Wu', 'Ruijing Zeng', 'Weiming Ren'] | 2022-08-03 | null | null | null | null | ['medical-code-prediction'] | ['medical'] | [ 2.36696571e-01 3.23991925e-01 -1.17407300e-01 -3.97051752e-01
-7.16210008e-01 -6.08800948e-01 -2.21544012e-01 6.41663015e-01
5.60436323e-02 2.37739876e-01 3.82374048e-01 -1.02132750e+00
-4.88613755e-01 -6.10827744e-01 -4.81655747e-01 -1.41227722e-01
-9.93138626e-02 6.53865516e-01 -2.50106901e-01 -2.06905268e-02
-1.23853192e-01 1.94555670e-01 -1.09727216e+00 1.04779780e+00
8.69334936e-01 5.87186456e-01 5.80707453e-02 1.09192133e+00
2.86610946e-02 1.62248516e+00 -2.97408193e-01 -3.96416605e-01
1.81106920e-03 -5.77778578e-01 -9.47314262e-01 -3.51064295e-01
4.79551703e-01 -1.99496523e-01 -5.08431077e-01 7.53962636e-01
5.52141845e-01 -3.42814356e-01 8.50245655e-01 -9.65897977e-01
-8.52815568e-01 7.57549345e-01 1.17397092e-01 9.07205343e-02
2.84014374e-01 -3.21619004e-01 9.54034269e-01 -5.24971187e-01
5.61502934e-01 9.36603248e-01 1.26323223e+00 9.78941500e-01
-1.12960935e+00 -6.66919827e-01 -1.27565712e-01 1.45571381e-01
-1.23590863e+00 -5.11293374e-02 5.42297661e-01 -1.08208740e+00
9.06237602e-01 3.19926292e-01 8.23237002e-01 1.17630386e+00
6.67530417e-01 7.58452654e-01 8.50408137e-01 -3.65396351e-01
8.86369050e-02 1.17644511e-01 4.03421313e-01 1.43495166e+00
4.11235899e-01 -2.27339119e-02 -4.24252182e-01 -4.50149924e-01
7.64046192e-01 5.95766068e-01 -2.65594810e-01 -4.49107528e-01
-1.08234251e+00 9.38460410e-01 1.15537778e-01 1.35139853e-01
-3.69487330e-02 2.63461739e-01 7.15940535e-01 3.32827300e-01
1.07848393e-02 7.58884490e-01 -7.15074182e-01 -1.66738816e-02
-8.54269743e-01 1.34555444e-01 1.02928579e+00 1.23709309e+00
2.01885775e-01 -1.51594371e-01 -1.88074574e-01 6.82354271e-01
1.33342057e-01 1.13214441e-01 5.77986538e-01 -7.21462905e-01
2.74424702e-01 7.36512303e-01 -4.00295198e-01 -7.53603458e-01
-8.67517531e-01 -6.70260787e-01 -1.12379897e+00 -3.05551320e-01
1.30086839e-01 -3.79370958e-01 -9.33269322e-01 1.44937086e+00
-4.46537621e-02 3.32422256e-01 1.48981616e-01 4.78398144e-01
1.20756376e+00 2.37810329e-01 3.12652975e-01 -2.12096050e-02
1.51992702e+00 -9.43151176e-01 -7.79287100e-01 1.78825706e-01
1.41139102e+00 -4.17832822e-01 7.73252308e-01 4.51209396e-01
-9.03267801e-01 -5.78833461e-01 -7.27555931e-01 -1.10517338e-01
-2.28420138e-01 1.80208102e-01 7.32364833e-01 4.94881183e-01
-1.15886188e+00 7.12296665e-01 -8.67763400e-01 -1.27640635e-01
4.72528756e-01 4.37872440e-01 -1.28547242e-02 -1.32005811e-01
-1.14601266e+00 9.48252022e-01 1.72657833e-01 -4.83109355e-01
-9.37130213e-01 -1.22560620e+00 -1.00383210e+00 3.81251574e-01
-7.78853223e-02 -9.59498823e-01 1.58473635e+00 -6.31604254e-01
-1.07647347e+00 7.82187164e-01 2.15256274e-01 -4.53056633e-01
3.83885145e-01 6.67040646e-02 -3.14559817e-01 2.19883680e-01
-1.59650221e-01 6.26671791e-01 4.26452935e-01 -8.55813801e-01
-5.07395625e-01 -1.02322735e-01 -5.58760874e-02 7.76237622e-02
-6.47906899e-01 -3.81859332e-01 1.20494992e-01 -6.60849333e-01
-2.47894675e-01 -1.16457021e+00 -5.48946619e-01 -1.91974550e-01
-3.10271055e-01 -3.79416198e-01 4.16290402e-01 -6.11053586e-01
1.36240041e+00 -2.09682894e+00 9.94788036e-02 1.95765570e-01
8.73770237e-01 -2.43572169e-03 -3.83464769e-02 4.14427698e-01
-3.43937099e-01 1.80265397e-01 1.27954960e-01 -1.90900579e-01
-2.62687624e-01 2.09006071e-01 -1.62381098e-01 1.26691714e-01
8.39511752e-02 9.88759696e-01 -1.04642832e+00 -6.41923726e-01
5.44384010e-02 4.68731672e-01 -1.04957724e+00 2.76589543e-01
-1.03392087e-01 5.78646839e-01 -5.01534224e-01 5.34343004e-01
8.75946134e-03 -1.14610004e+00 4.68332857e-01 1.08195893e-01
1.38707533e-01 5.31735301e-01 -7.44160891e-01 1.92815399e+00
-5.62594473e-01 4.45677578e-01 -5.01826942e-01 -1.07322848e+00
6.28333330e-01 8.30082238e-01 9.10799265e-01 -8.83188564e-03
5.28552532e-02 1.72698274e-01 2.58326977e-01 -8.00219715e-01
5.99234598e-03 6.84196278e-02 -1.93536818e-01 3.71993810e-01
3.51738751e-01 -1.88410714e-01 -5.09774983e-02 1.63660213e-01
1.67002773e+00 -2.17267305e-01 4.83602256e-01 -4.44406360e-01
3.63105088e-02 2.19939351e-01 6.17452204e-01 9.69157457e-01
-1.11030065e-01 4.82708931e-01 4.41406727e-01 -1.01246548e+00
-1.14350390e+00 -9.00188744e-01 -2.39898801e-01 9.74110723e-01
-5.26952744e-01 -8.07389557e-01 -6.91583633e-01 -6.95771754e-01
1.31531790e-01 1.96880162e-01 -6.48343861e-01 -3.14741969e-01
-5.09350121e-01 -3.50007236e-01 6.95689201e-01 6.88930511e-01
-1.40422687e-01 -9.62562144e-01 -9.54956472e-01 3.37506413e-01
-1.02038115e-01 -9.25530910e-01 -5.18784761e-01 6.00940943e-01
-1.03673708e+00 -1.29675984e+00 -5.72429657e-01 -1.30753005e+00
8.36708188e-01 -1.88901678e-01 1.20660162e+00 3.90923321e-01
-7.09351897e-01 6.52156532e-01 -2.85996497e-01 -6.76124811e-01
-6.57992721e-01 3.76495600e-01 8.02759007e-02 -5.07690668e-01
3.22976321e-01 -4.63893950e-01 -6.36713624e-01 -2.57608682e-01
-8.08443069e-01 6.23620272e-01 5.57787120e-01 1.01888597e+00
5.01257956e-01 -3.28478962e-01 5.47622502e-01 -1.38888824e+00
8.87508631e-01 -5.49097717e-01 -3.05189997e-01 4.81531531e-01
-7.42135644e-01 2.06106633e-01 8.21380079e-01 -6.06053770e-01
-3.16327393e-01 4.66275305e-01 -1.30054936e-01 -6.20269120e-01
-1.48586258e-01 7.61751413e-01 7.97175825e-01 -7.55392164e-02
8.90244484e-01 1.40415192e-01 -2.55832195e-01 -1.72169954e-01
8.55809376e-02 7.08785594e-01 2.05815807e-01 -4.40055847e-01
2.61589348e-01 -1.19785834e-02 1.57705456e-01 -3.59021127e-01
-1.07519376e+00 -5.03230512e-01 -4.37881529e-01 -1.96371734e-01
1.10655677e+00 -9.67044175e-01 -1.09202158e+00 -1.67365462e-01
-1.11745715e+00 -4.70999002e-01 -1.36732414e-01 5.31176090e-01
-6.51753247e-01 4.80606183e-02 -1.20668006e+00 -4.21388984e-01
-6.60919964e-01 -1.24823380e+00 8.69761229e-01 -2.35558636e-02
-8.13474059e-01 -1.23338914e+00 1.01600364e-01 1.91450223e-01
1.84664369e-01 2.66776353e-01 1.52321982e+00 -8.58937621e-01
-4.59980965e-01 5.15776984e-02 2.60671318e-01 1.30639628e-01
1.87356740e-01 -3.88976544e-01 -7.87357092e-01 -4.70652819e-01
-1.10096708e-01 -4.91749883e-01 5.40041208e-01 4.92214113e-01
1.87492645e+00 -5.34917414e-01 -6.78161085e-01 7.40958273e-01
1.29588115e+00 5.00685930e-01 1.79243028e-01 -9.19395834e-02
9.78410840e-01 4.70759153e-01 -9.01292711e-02 4.17248577e-01
6.66120708e-01 3.12645853e-01 -2.08831765e-02 -2.72773981e-01
-5.82638904e-02 -2.02798769e-01 -9.04088765e-02 1.68406749e+00
1.02606371e-01 1.61770254e-01 -1.55834806e+00 6.05578482e-01
-1.89849877e+00 -6.63816750e-01 -6.56590983e-02 1.70010078e+00
1.22968209e+00 -8.13024044e-02 -1.01212732e-01 -9.35935453e-02
4.50534731e-01 -5.60924828e-01 -5.35562932e-01 -7.14519799e-01
6.29064500e-01 5.08070230e-01 4.28914398e-01 3.99857700e-01
-1.13406265e+00 4.78419930e-01 6.53541517e+00 3.03382099e-01
-8.91967654e-01 2.62922436e-01 7.19647467e-01 -1.81424525e-02
-2.05534860e-01 -4.98480171e-01 -7.23349571e-01 3.43233854e-01
1.49766123e+00 -5.05592003e-02 2.70434886e-01 1.10583127e+00
-5.60612120e-02 8.12526882e-01 -1.68704820e+00 1.03818583e+00
6.48222268e-02 -1.79731798e+00 1.42221987e-01 -6.30267411e-02
9.72735524e-01 -9.95419398e-02 1.33758187e-01 4.60343152e-01
8.45041275e-01 -1.12751043e+00 2.29359016e-01 5.29917955e-01
1.05492353e+00 -5.21398962e-01 5.25981128e-01 1.99606687e-01
-1.20506990e+00 -4.62976277e-01 -1.41595364e-01 1.18035421e-01
-3.44198495e-01 3.77318293e-01 -1.20516789e+00 2.75609285e-01
6.85642242e-01 8.18202794e-01 -4.26244915e-01 1.13655591e+00
-5.78494966e-02 6.13324940e-01 3.44340831e-01 -1.25024959e-01
4.97470275e-02 2.88990140e-01 -8.48949030e-02 1.67463756e+00
4.50030982e-01 3.02482247e-01 5.54570615e-01 5.63300788e-01
-1.91528156e-01 2.11183727e-01 -1.04052877e+00 5.51793613e-02
4.21468467e-01 1.14749873e+00 -5.22301555e-01 -4.49198455e-01
-5.74145198e-01 7.33791947e-01 5.01331568e-01 6.38921335e-02
-7.63441384e-01 -3.86526257e-01 5.25399923e-01 -9.45125590e-04
8.50455761e-02 2.25712471e-02 -6.21338964e-01 -1.17320228e+00
-2.98525959e-01 -1.23845267e+00 6.69160724e-01 -6.54082119e-01
-9.50180590e-01 5.64567626e-01 -4.09242690e-01 -1.46230137e+00
-4.66266602e-01 -7.52658606e-01 -1.06731713e-01 4.78233099e-01
-1.31018925e+00 -8.96572888e-01 -2.24696532e-01 6.54816091e-01
5.72305262e-01 -3.29623699e-01 1.55363262e+00 5.84618032e-01
-2.07939684e-01 8.66254508e-01 1.92473054e-01 5.34578085e-01
4.84507650e-01 -1.33756506e+00 1.83696970e-01 1.89376071e-01
9.90252793e-02 6.99895322e-01 2.73772925e-01 -7.17046678e-01
-1.15497959e+00 -1.57092965e+00 1.14585495e+00 -6.74624324e-01
4.88156438e-01 -5.34782290e-01 -7.14646578e-01 8.97877812e-01
-7.64339045e-02 -1.05607912e-01 1.40462315e+00 1.86098546e-01
-4.39043671e-01 1.22456677e-01 -8.41943204e-01 6.68839753e-01
1.16393995e+00 -7.35645533e-01 -4.80636925e-01 8.41151893e-01
1.00942159e+00 -7.68409967e-01 -1.36620724e+00 4.17515874e-01
6.36675239e-01 -2.30005696e-01 8.32914710e-01 -1.09134698e+00
1.08444035e+00 2.24999934e-01 3.53485286e-01 -1.45614624e+00
-7.37801909e-01 -4.05860215e-01 -4.33026135e-01 2.52279311e-01
5.74305654e-01 -3.40266466e-01 7.18775749e-01 3.98068160e-01
-5.24947107e-01 -1.24958444e+00 -5.57237029e-01 -3.56847733e-01
1.96000621e-01 -2.36582272e-02 3.62155288e-01 1.53931808e+00
5.65751076e-01 3.17456156e-01 -3.44117403e-01 1.09001875e-01
4.04552281e-01 1.94175959e-01 7.10797980e-02 -1.62270570e+00
-6.81557715e-01 -2.60357291e-01 -3.96398008e-01 -6.90560579e-01
-5.49215674e-02 -1.49914289e+00 -3.14570725e-01 -1.72728479e+00
3.78642440e-01 -6.00382030e-01 -5.69002271e-01 9.13231969e-01
-1.29486127e-02 -2.09818840e-01 8.74047279e-02 4.52108681e-01
-7.27047443e-01 -1.73459455e-01 1.50321805e+00 -2.52462000e-01
-1.55585408e-01 -8.49946365e-02 -8.01389337e-01 5.40798485e-01
9.99538779e-01 -8.49726200e-01 -5.05645156e-01 -6.26953483e-01
3.53800952e-01 5.49981594e-01 -6.05855882e-02 -1.11813200e+00
4.47520971e-01 1.57539204e-01 4.51631963e-01 -1.87630177e-01
-4.74878848e-02 -8.39947343e-01 1.17983237e-01 1.18770123e+00
-1.02657437e+00 5.94056487e-01 3.28339547e-01 3.97276223e-01
1.56776994e-01 -1.33359432e-01 5.53006172e-01 -4.40192372e-01
-1.01380393e-01 3.74292225e-01 -8.17716002e-01 4.41394784e-02
7.71300495e-01 2.20866501e-01 -3.97391915e-01 -2.44158730e-01
-1.18000984e+00 2.60859281e-01 -2.87212461e-01 5.25050879e-01
7.76836932e-01 -1.28419864e+00 -6.75628543e-01 1.25880003e-01
2.39099771e-01 -3.51053998e-02 1.37802482e-01 7.31995821e-01
-7.73241460e-01 7.93887854e-01 -2.07300171e-01 -5.60740292e-01
-1.45312583e+00 5.40600717e-01 7.39378691e-01 -7.26945639e-01
-7.56557941e-01 7.32555151e-01 2.31211305e-01 -7.38722384e-01
5.81626058e-01 -7.28833735e-01 -3.98192734e-01 -3.05143028e-01
3.65635574e-01 -6.86523840e-02 1.29478350e-01 2.87521541e-01
-9.48913544e-02 2.76690781e-01 -4.83152568e-01 5.27853370e-01
1.41675699e+00 4.49077278e-01 3.47671136e-02 4.24560487e-01
1.13175786e+00 -6.66758597e-01 -6.43465281e-01 -1.35306790e-01
2.41208434e-01 2.03182444e-01 -7.44264647e-02 -9.86531079e-01
-7.84702599e-01 9.33281839e-01 8.48498702e-01 1.92054585e-01
8.86748731e-01 -7.59111941e-02 6.81029081e-01 7.88377523e-01
3.54794472e-01 -9.99723136e-01 3.23496401e-01 5.94794393e-01
5.06884873e-01 -9.25671697e-01 -2.91797668e-01 -3.56921971e-01
-5.50909519e-01 1.39238906e+00 6.77226365e-01 4.73013781e-02
8.84788990e-01 5.78990817e-01 2.19248027e-01 -5.89496195e-01
-9.66510415e-01 2.46909425e-01 2.55880445e-01 4.65424806e-01
9.56347346e-01 4.08089995e-01 -1.11135282e-01 6.40701652e-01
-2.78975517e-01 2.88035899e-01 6.53199255e-01 1.05780709e+00
-1.13557450e-01 -1.11944616e+00 -1.68716423e-02 1.07329285e+00
-5.92484653e-01 -5.16645074e-01 -1.28392488e-01 2.33021542e-01
2.60304004e-01 7.17354596e-01 -5.80090992e-02 -7.26247609e-01
3.53442192e-01 4.62557524e-01 4.70776021e-01 -1.16234291e+00
-8.40717793e-01 -3.19609523e-01 -4.96344008e-02 -3.41125816e-01
-3.44164148e-02 -6.13809764e-01 -1.29888678e+00 -9.05033872e-02
9.35175046e-02 2.04794422e-01 3.38304639e-01 6.82682872e-01
6.09407067e-01 1.36361790e+00 1.68095633e-01 -1.52147919e-01
-6.47559524e-01 -7.21086800e-01 -3.08228612e-01 3.37755799e-01
6.30479276e-01 -3.59841913e-01 8.59594569e-02 3.90708745e-01] | [7.990718841552734, 6.810083866119385] |
fa0551db-0333-41ca-b170-7ded7270aa56 | three-dimensional-deep-learning-approach-for | 1806.05824 | null | http://arxiv.org/abs/1806.05824v1 | http://arxiv.org/pdf/1806.05824v1.pdf | Three dimensional Deep Learning approach for remote sensing image classification | Recently, a variety of approaches has been enriching the field of Remote
Sensing (RS) image processing and analysis. Unfortunately, existing methods
remain limited faced to the rich spatio-spectral content of today's large
datasets. It would seem intriguing to resort to Deep Learning (DL) based
approaches at this stage with regards to their ability to offer accurate
semantic interpretation of the data. However, the specificity introduced by the
coexistence of spectral and spatial content in the RS datasets widens the scope
of the challenges presented to adapt DL methods to these contexts. Therefore,
the aim of this paper is firstly to explore the performance of DL architectures
for the RS hyperspectral dataset classification and secondly to introduce a new
three-dimensional DL approach that enables a joint spectral and spatial
information process. A set of three-dimensional schemes is proposed and
evaluated. Experimental results based on well knownhyperspectral datasets
demonstrate that the proposed method is able to achieve a better classification
rate than state of the art methods with lower computational costs. | ['A. Benoit', 'Amina Ben Hamida', 'Chokri Ben Amar', 'Patrick Lambert'] | 2018-06-15 | null | null | null | null | ['remote-sensing-image-classification'] | ['miscellaneous'] | [ 4.94201809e-01 -2.83759266e-01 6.04027323e-02 -2.93003321e-01
-6.26475751e-01 -4.93888140e-01 8.23200941e-01 1.03281379e-01
-3.90889525e-01 7.79638708e-01 -2.43122011e-01 -3.41183454e-01
-8.31289649e-01 -9.61135626e-01 -1.10671364e-01 -1.04393065e+00
-2.05540106e-01 1.10351086e-01 -9.01099816e-02 -2.71064103e-01
3.56892906e-02 1.17421353e+00 -1.94449568e+00 2.57260352e-01
1.03492320e+00 1.22910881e+00 6.23842418e-01 2.66570866e-01
-2.85739571e-01 4.03827876e-01 -1.65125474e-01 6.80913553e-02
3.39330018e-01 -2.55682081e-01 -7.67285168e-01 4.13062513e-01
3.77103955e-01 -1.75700665e-01 -1.00920117e-02 1.21662676e+00
6.44042432e-01 1.70189023e-01 5.60916424e-01 -6.32569253e-01
-3.94157916e-01 2.06904247e-01 -4.60220039e-01 3.13080788e-01
-3.19248289e-02 -7.23338947e-02 9.39608634e-01 -9.42923486e-01
3.64869922e-01 7.13806987e-01 7.06630707e-01 -1.36695042e-01
-1.23583651e+00 -1.73601702e-01 6.15952089e-02 3.56195837e-01
-1.55929554e+00 -2.94953048e-01 9.94418502e-01 -5.84152281e-01
7.87363350e-01 2.88469911e-01 5.90721369e-01 6.80449307e-01
-1.97227702e-01 7.06191421e-01 1.61519778e+00 -5.15780210e-01
2.77213156e-01 -1.81386005e-02 4.66007888e-02 2.89104372e-01
1.84006691e-01 1.65805191e-01 -2.28538439e-01 1.93542808e-01
4.51340616e-01 -7.08316639e-02 -2.46033341e-01 -5.02967775e-01
-8.03373039e-01 9.69729483e-01 6.83072805e-01 7.10266948e-01
-7.07509935e-01 -4.85195667e-01 2.17727974e-01 4.52497192e-02
8.03754926e-01 4.25387174e-01 -2.33051315e-01 4.67956960e-01
-1.16597807e+00 1.63553804e-01 3.63261014e-01 2.22261667e-01
8.41292977e-01 2.79857934e-01 2.39230171e-01 1.06735146e+00
2.78880924e-01 6.22065008e-01 2.73538381e-01 -4.75082427e-01
1.21748865e-01 6.80261374e-01 3.90089080e-02 -1.08479202e+00
-7.13588476e-01 -7.42841125e-01 -1.05433810e+00 4.12952662e-01
1.24578200e-01 4.36311141e-02 -7.85345852e-01 1.08944476e+00
1.67862698e-01 -1.33574262e-01 1.91207960e-01 1.07988369e+00
6.89262688e-01 5.50758183e-01 1.26948729e-01 -5.01779430e-02
9.20588136e-01 -3.84749293e-01 -5.27227521e-01 -2.21214980e-01
3.81236255e-01 -6.21443033e-01 9.94097352e-01 3.89428288e-01
-4.92235035e-01 -5.75937510e-01 -1.10389769e+00 2.76787043e-01
-7.15588450e-01 5.97876668e-01 8.18657100e-01 7.01899409e-01
-8.22413445e-01 5.64711154e-01 -7.44021475e-01 -7.05088854e-01
5.97766042e-01 2.12833226e-01 -2.80602485e-01 -7.52382502e-02
-1.24170208e+00 7.47879565e-01 8.15190613e-01 5.23483336e-01
-3.50950688e-01 -5.32469332e-01 -5.62416434e-01 -1.02269962e-01
2.67220318e-01 -1.59899846e-01 7.43785620e-01 -1.16466069e+00
-1.24249268e+00 1.03297329e+00 2.73938715e-01 -4.61759597e-01
5.30909181e-01 -5.13985604e-02 -6.44320071e-01 3.84352863e-01
2.30746847e-02 3.27781260e-01 6.44341707e-01 -1.36502039e+00
-6.52090549e-01 -6.10650361e-01 1.41579658e-01 2.60717750e-01
-4.79475051e-01 -2.32025534e-01 7.21760243e-02 -4.76995677e-01
4.37117308e-01 -7.94006050e-01 -1.79480821e-01 5.35183847e-02
-2.38836080e-01 1.30951479e-01 7.76399493e-01 -4.42119658e-01
9.89705622e-01 -2.31081438e+00 1.06709331e-01 1.76115796e-01
-1.39139667e-01 8.31252694e-01 -1.58175811e-01 6.59575045e-01
-2.38828808e-01 3.10506467e-02 -6.35978162e-01 2.41803401e-03
4.63974550e-02 1.80060089e-01 -3.39012921e-01 7.43301928e-01
3.03657293e-01 6.39185011e-01 -6.71183527e-01 -1.85447752e-01
7.67728031e-01 6.07128263e-01 -1.31511977e-02 1.96374148e-01
-2.49651074e-01 5.30066311e-01 -6.11124098e-01 6.94113195e-01
9.78450596e-01 -3.93504202e-02 1.99537843e-01 -3.51827651e-01
-6.22804403e-01 1.14937760e-02 -1.25825989e+00 1.49136817e+00
-3.19837421e-01 5.52782297e-01 1.83485344e-01 -1.38356054e+00
1.00801456e+00 2.07619399e-01 6.85746253e-01 -8.98679972e-01
-8.62051025e-02 5.45428395e-01 -7.78517872e-02 -6.98471010e-01
3.48393112e-01 -4.41463798e-01 4.46950316e-01 2.04714715e-01
-2.31835261e-01 -2.63712794e-01 5.72088622e-02 -5.32283068e-01
3.41353863e-01 3.25359285e-01 4.11358714e-01 -5.50626338e-01
1.02445912e+00 1.96427211e-01 1.03907278e-02 5.02364099e-01
-1.57303691e-01 3.48446041e-01 1.79661252e-02 -6.72180176e-01
-9.73375320e-01 -8.17173302e-01 -7.10149407e-01 7.95712948e-01
-3.91520932e-03 4.17031258e-01 -3.92692387e-01 -2.87262261e-01
1.79379676e-02 4.73369449e-01 -4.10634905e-01 2.14460716e-01
-1.55461743e-01 -1.35989225e+00 4.76202875e-01 1.45537600e-01
9.49833632e-01 -9.62994576e-01 -7.68287241e-01 2.85800576e-01
6.18049235e-04 -1.13416445e+00 7.01103032e-01 2.07139403e-01
-1.08830869e+00 -9.85732496e-01 -6.09562218e-01 -3.80272806e-01
1.54215500e-01 4.52919722e-01 7.19065309e-01 -3.63105237e-01
-3.21649134e-01 1.41567513e-01 -6.11554980e-01 -3.02183002e-01
-2.51866221e-01 3.64777833e-01 -3.00510466e-01 3.19084942e-01
4.83309269e-01 -5.84602773e-01 -5.14217675e-01 2.41339058e-02
-1.37221396e+00 -1.63432866e-01 8.60105574e-01 6.81370437e-01
5.65599322e-01 4.37540025e-01 6.18000448e-01 -7.35029519e-01
3.87890071e-01 -4.83631164e-01 -7.45809019e-01 2.17536077e-01
-7.22275615e-01 -1.01108208e-01 5.92138648e-01 1.36869058e-01
-1.14891648e+00 2.08005220e-01 -3.69284809e-01 1.46318465e-01
-7.63953745e-01 1.02107525e+00 -1.84435397e-01 -3.52808893e-01
7.51510680e-01 3.90471637e-01 -4.07820866e-02 -7.42022157e-01
1.92440316e-01 9.15390015e-01 2.03055680e-01 -2.50653327e-01
6.97722018e-01 7.30939388e-01 2.60763913e-01 -1.38550556e+00
-8.90175998e-01 -5.57484984e-01 -1.07909203e+00 -2.59274215e-01
8.87022376e-01 -9.01880920e-01 -2.05732092e-01 7.33473063e-01
-7.21439064e-01 -2.03876808e-01 -1.92995504e-01 4.95420188e-01
-4.66668069e-01 6.30958140e-01 -9.61788893e-02 -1.04649508e+00
-2.08514303e-01 -1.09798169e+00 1.00462782e+00 -1.01105116e-01
4.66583401e-01 -1.21998024e+00 -1.42125404e-02 3.50212753e-01
4.82029706e-01 4.26454544e-01 1.03848493e+00 -3.86298209e-01
-5.58891654e-01 -2.36911952e-01 -6.32197559e-01 5.10508239e-01
2.73148030e-01 -2.26526171e-01 -1.37423265e+00 -2.13308766e-01
1.19206859e-02 -2.91600227e-01 9.72974598e-01 4.20842469e-01
1.20559859e+00 1.74580291e-01 -1.09501190e-01 7.69335389e-01
2.05341291e+00 6.61463886e-02 5.93540847e-01 7.62162089e-01
5.31654298e-01 8.23313057e-01 7.08484232e-01 5.75558782e-01
9.04015172e-03 6.03157818e-01 8.65700722e-01 -5.75223088e-01
8.55842885e-03 2.14964673e-01 -8.24998468e-02 2.10460514e-01
-3.60103041e-01 -2.77181834e-01 -1.11852670e+00 3.20958346e-01
-1.67608333e+00 -1.05107462e+00 -3.61242324e-01 2.11606693e+00
2.33937025e-01 -3.57288599e-01 -1.19343981e-01 5.74997783e-01
4.99770433e-01 6.48088336e-01 -4.92205888e-01 1.20300770e-01
-6.30461216e-01 1.91179201e-01 5.64481199e-01 2.80627787e-01
-1.59976363e+00 7.83501863e-01 6.05740547e+00 6.37439072e-01
-1.54974520e+00 -1.77757949e-01 3.61303389e-01 4.80229050e-01
-2.07893640e-01 -4.98166710e-01 -5.02686977e-01 1.20708488e-01
8.60642731e-01 2.03231767e-01 3.88218701e-01 5.36150694e-01
4.34036702e-01 -2.93156922e-01 -5.48405886e-01 8.75386477e-01
8.86147395e-02 -1.14987648e+00 1.24659918e-01 8.87903497e-02
6.62159979e-01 3.48595232e-01 2.54353195e-01 -9.42880958e-02
-2.79461890e-01 -1.03739846e+00 5.73569119e-01 5.97994268e-01
6.82189941e-01 -7.67031431e-01 8.86963606e-01 4.53562200e-01
-1.07726860e+00 -3.25098813e-01 -4.59651202e-01 -1.44572526e-01
-1.74426779e-01 7.51529098e-01 -7.78830886e-01 1.05088210e+00
6.26179695e-01 8.28831494e-01 -5.43653429e-01 1.07084775e+00
-1.06225960e-01 2.60042429e-01 -2.12152869e-01 2.55798846e-01
6.34412229e-01 -5.90692699e-01 4.93928611e-01 1.21347666e+00
5.08581936e-01 9.96986851e-02 2.34426737e-01 8.34196746e-01
4.41840470e-01 3.45167160e-01 -8.13657463e-01 -3.79384756e-01
6.57062978e-02 1.26698482e+00 -7.38093555e-01 3.11787836e-02
-3.92512858e-01 6.89971685e-01 1.85959283e-02 4.21869248e-01
-4.48282570e-01 -5.18349633e-02 5.50260782e-01 1.34196475e-01
3.41719806e-01 -5.64298093e-01 -3.57382298e-01 -9.32766378e-01
-1.61027029e-01 -7.55666196e-01 4.57324684e-01 -8.37278426e-01
-1.22752845e+00 8.29934180e-01 4.20583822e-02 -1.32358289e+00
2.33015418e-02 -8.54908466e-01 -5.20981587e-02 1.12612629e+00
-2.35674214e+00 -1.40338349e+00 -6.18862271e-01 5.08542955e-01
4.09900904e-01 -3.01955193e-01 1.02606678e+00 3.32967758e-01
-3.45584273e-01 -2.46951386e-01 4.87434119e-01 -2.56520331e-01
2.67753422e-01 -1.12957549e+00 -2.02042282e-01 1.02302480e+00
-3.67121100e-02 9.38693136e-02 5.94963372e-01 -1.81609318e-01
-1.13194501e+00 -1.16472638e+00 6.79516494e-01 1.34107798e-01
7.80491114e-01 9.75489318e-02 -9.25264835e-01 3.05503577e-01
-3.42319943e-02 6.54940158e-02 8.68708491e-01 -1.92180619e-01
-1.62354171e-01 -4.20873225e-01 -1.12227201e+00 2.65610784e-01
6.82587922e-01 -7.99713135e-01 -4.67656821e-01 3.74995023e-01
8.94916337e-03 1.00163266e-01 -7.75500953e-01 7.76442349e-01
2.80630141e-01 -1.40562212e+00 1.08986688e+00 -1.88374311e-01
3.44931215e-01 -4.07872260e-01 -4.62026626e-01 -1.17124283e+00
-3.27326238e-01 1.97285116e-01 4.58161622e-01 8.78047168e-01
1.31568357e-01 -6.35617375e-01 5.84005773e-01 4.77738418e-02
-2.90991068e-01 -4.28818285e-01 -8.36222589e-01 -7.33078837e-01
1.62203923e-01 -5.47707021e-01 6.41766965e-01 1.07479966e+00
-5.30044854e-01 -2.30353568e-02 -1.08427554e-01 5.89319885e-01
6.56307280e-01 6.20236337e-01 2.61797696e-01 -1.74850929e+00
1.37958778e-02 -5.90101838e-01 -4.53277200e-01 -5.91899335e-01
2.36100242e-01 -9.54273462e-01 -1.14181712e-01 -1.67758322e+00
-1.49753734e-01 -6.50315940e-01 -4.46819365e-01 3.04187626e-01
1.48145542e-01 5.36759794e-01 -1.23120239e-02 3.28077227e-01
-4.79239933e-02 6.41322434e-01 1.05915594e+00 -2.60624170e-01
-2.31985122e-01 6.04502372e-02 -2.44172946e-01 6.61800742e-01
9.98735189e-01 7.61071444e-02 -4.54085201e-01 -4.80626643e-01
1.24636799e-01 -1.18866503e-01 6.55858934e-01 -1.20295227e+00
-4.85729799e-02 -1.23644248e-01 3.41576725e-01 -7.20119894e-01
2.65564740e-01 -9.74422574e-01 3.03123713e-01 2.49571741e-01
-4.83156964e-02 -4.85442191e-01 2.78829366e-01 3.52539748e-01
-5.72605848e-01 -1.45732537e-01 1.08605194e+00 -2.05438316e-01
-1.23029268e+00 4.07472849e-01 -3.76716167e-01 -4.62347060e-01
9.66989636e-01 -4.49524492e-01 -4.79632728e-02 1.38617262e-01
-8.55774939e-01 -3.05025041e-01 2.48018965e-01 9.46641639e-02
3.79366934e-01 -9.62100983e-01 -5.84349155e-01 3.41046423e-01
3.71738613e-01 -1.66694790e-01 6.38410032e-01 9.56428885e-01
-7.31037974e-01 7.67533064e-01 -5.43609262e-01 -6.60786986e-01
-1.09511483e+00 4.16985065e-01 7.93591857e-01 -5.69797121e-02
-5.83840668e-01 5.59381187e-01 -1.08570307e-02 -5.50692201e-01
-1.20075554e-01 -7.39369020e-02 -7.30561972e-01 5.38043380e-01
2.81689584e-01 4.59824771e-01 4.03349340e-01 -9.00405049e-01
-3.78946662e-01 6.04951322e-01 4.62057173e-01 1.10479318e-01
1.81568766e+00 -1.97133437e-01 -2.52083391e-01 5.16540825e-01
9.57664013e-01 -2.81218410e-01 -1.31200254e+00 -2.89271742e-01
3.17573249e-01 -5.94984353e-01 5.82063675e-01 -9.50093508e-01
-1.03053343e+00 1.10981584e+00 1.04879510e+00 5.69655597e-01
1.47876501e+00 -3.41074646e-01 2.65467733e-01 4.37881023e-01
3.65337282e-01 -1.12166631e+00 -3.57487977e-01 4.18592632e-01
8.24563324e-01 -1.49803829e+00 1.76490322e-01 -4.60176140e-01
-3.41082841e-01 1.34080255e+00 -7.82514084e-03 1.12230748e-01
7.19841838e-01 -9.74274427e-02 2.40505740e-01 -3.12579781e-01
5.45155741e-02 -8.98918331e-01 4.16764140e-01 7.46018589e-01
6.14014387e-01 1.83198214e-01 -4.33870792e-01 -1.97692700e-02
2.02575147e-01 2.75454018e-02 2.42206022e-01 7.40676045e-01
-6.99376762e-01 -1.09856749e+00 -3.96713793e-01 2.35720962e-01
-3.95226151e-01 -5.46917506e-02 -3.51493329e-01 8.94663990e-01
3.37474048e-01 9.20402169e-01 -1.93780154e-01 -4.34075259e-02
2.60697514e-01 5.81739619e-02 2.40094900e-01 -4.17434037e-01
-2.05329508e-01 1.69538215e-01 -1.68680735e-02 -3.25275570e-01
-1.13376403e+00 -7.41960406e-01 -9.30990875e-01 1.31792566e-02
-1.01481684e-01 8.23203067e-04 9.37532127e-01 1.16587150e+00
1.19649619e-01 3.41771483e-01 6.77222192e-01 -9.71957564e-01
-6.03170037e-01 -8.46512735e-01 -1.13110256e+00 1.91537201e-01
4.44718748e-01 -6.72808766e-01 -2.38212124e-01 -1.23133110e-02] | [9.731595993041992, -1.626183032989502] |
5321d606-1c61-4eb7-8e09-86e82d5da94b | on-label-efficient-computer-vision-building | null | null | https://open.library.ubc.ca/soa/cIRcle/collections/ubctheses/24/items/1.0402554 | https://open.library.ubc.ca/media/download/pdf/24/1.0402554/4 | On Label-Efficient Computer Vision: Building Fast and Effective Few-Shot Image Classifiers | Modern deep learning requires large-scale extensively labelled datasets for training. Few-shot learning aims to alleviate this issue by learning effectively from few labelled examples. In previously proposed few-shot visual classifiers, it is assumed that the feature manifold arriving at the classifier has uncorrelated feature dimensions and uniform feature variance. In this work, we focus on addressing the limitations arising from this assumption by proposing a variance-sensitive class of models that operates in a low-label regime. The first method, Simple CNAPS, employs a hierarchically regularized Mahalanobis-distance based classifier combined with a state of the art neural adaptive feature extractor to achieve strong performance on Meta-Dataset, mini-ImageNet and tiered-ImageNet benchmarks. We further extend this approach to a transductive learning setting, proposing Transductive CNAPS. This transductive method combines a soft k-means parameter refinement procedure with a two-step task encoder to achieve improved test-time classification accuracy using unlabelled data. Transductive CNAPS achieves state of the art performance across all major few-shot learning benchmarks. Finally, we explore the use of our methods (Simple and Transductive) for “out of the box” continual and active learning. Extensive experiments on large scale benchmarks illustrate robustness and versatility of this, relatively speaking, simple class of models. All trained model checkpoints and corresponding source codes are made publicly available at github. com/plai-group/simple-cnaps. | ['Peyman Bateni'] | 2021-10-19 | null | null | null | university-of-british-columbia-theses-and | ['cross-domain-few-shot'] | ['computer-vision'] | [ 3.98183942e-01 1.61541730e-01 -3.84511292e-01 -4.24678534e-01
-1.06729531e+00 -2.73149580e-01 8.93484831e-01 5.48461415e-02
-5.24435222e-01 6.56236112e-01 -4.48381854e-03 1.32437482e-01
-3.29479814e-01 -6.14411712e-01 -6.52548492e-01 -9.88173902e-01
5.35332151e-02 4.51942503e-01 3.90860856e-01 -2.25144535e-01
1.77987278e-01 2.07419783e-01 -1.92973828e+00 2.78726041e-01
4.70757604e-01 1.12034941e+00 -2.24619033e-03 7.64529467e-01
-1.34075791e-01 1.15195096e+00 -3.51641774e-01 -3.95753860e-01
4.16622072e-01 -4.84134346e-01 -7.99958885e-01 2.63311237e-01
3.10500115e-01 -8.80135149e-02 -1.00417454e-02 9.24174666e-01
6.87631071e-01 4.54607368e-01 8.71275246e-01 -1.48117948e+00
-6.20884120e-01 2.87499666e-01 -3.69333267e-01 3.18224072e-01
-8.87353420e-02 5.09615242e-01 1.02261114e+00 -1.15736008e+00
7.67496288e-01 9.30674732e-01 7.42586195e-01 7.84166574e-01
-1.34135377e+00 -3.50232333e-01 -1.26918614e-01 4.86427516e-01
-1.34731340e+00 -6.56944513e-01 8.27953041e-01 -5.19109845e-01
1.21479464e+00 2.04929233e-01 6.27617538e-01 1.26604366e+00
3.56983542e-02 8.77188742e-01 1.23707902e+00 -7.45551109e-01
8.15733731e-01 2.43941903e-01 3.85539174e-01 7.50264764e-01
-4.15672772e-02 2.40647927e-01 -5.42369425e-01 -1.54901654e-01
3.18191946e-01 1.81722075e-01 -6.00561760e-02 -9.32341635e-01
-8.85377586e-01 1.16116726e+00 4.55257446e-01 3.39843720e-01
-2.16118991e-01 1.57483965e-01 6.68637156e-01 4.41146344e-01
8.22550058e-01 3.88206691e-01 -3.24994981e-01 -2.56666183e-01
-8.99121284e-01 -1.58901155e-01 7.94818342e-01 9.98506486e-01
1.02153921e+00 1.24251910e-01 -4.82908607e-01 1.07304144e+00
9.73049030e-02 9.12754461e-02 6.83197021e-01 -8.74197125e-01
-3.42466012e-02 4.72178400e-01 -3.52988631e-01 -4.37024832e-01
-2.60563731e-01 -3.92785847e-01 -7.14876950e-01 4.22772735e-01
6.29587248e-02 9.34293400e-03 -1.45707154e+00 1.42424166e+00
4.19114172e-01 4.78229135e-01 2.06835512e-02 7.52185583e-01
8.99918318e-01 5.79876781e-01 2.25243360e-01 -4.15823698e-01
1.03192043e+00 -1.20672858e+00 -3.62336636e-01 -1.20190643e-01
9.27896261e-01 -3.26253176e-01 1.16148543e+00 1.35861188e-01
-6.89267635e-01 -5.11021376e-01 -1.13656461e+00 -4.77797613e-02
-7.58702397e-01 -3.42715621e-01 5.32423973e-01 6.40673399e-01
-1.07135403e+00 6.81611001e-01 -7.06268489e-01 -6.11826420e-01
8.79242063e-01 2.09480375e-01 -4.12080765e-01 -2.61320591e-01
-1.04426301e+00 6.99389160e-01 5.26455581e-01 -2.87149400e-01
-1.22122955e+00 -7.77334630e-01 -9.83137906e-01 4.15184535e-02
5.84064543e-01 -6.48723423e-01 1.46124434e+00 -1.12833762e+00
-1.61143708e+00 1.02128708e+00 1.55405730e-01 -4.41835046e-01
4.99148279e-01 -9.18632969e-02 -2.06858814e-02 2.43627250e-01
1.27488254e-02 8.19570422e-01 1.04974473e+00 -1.22593081e+00
-4.16786999e-01 -2.14302808e-01 -1.88053131e-01 1.72062725e-01
-5.29004455e-01 -5.17830476e-02 -3.35096627e-01 -3.41487676e-01
-3.24742854e-01 -8.78308177e-01 -2.71949410e-01 7.24865347e-02
-8.16941708e-02 -3.76538873e-01 8.59889925e-01 1.07935131e-01
8.73059273e-01 -2.11535883e+00 3.38682458e-02 -9.04775783e-02
2.55260676e-01 6.23531699e-01 -2.22161621e-01 4.84173089e-01
-1.47645921e-01 -1.96265504e-01 -3.85508806e-01 -5.41760862e-01
4.28313054e-02 3.58475715e-01 -2.22081408e-01 6.69833362e-01
4.81552064e-01 1.32771146e+00 -1.06394053e+00 -5.76021850e-01
4.83832955e-01 4.91083235e-01 -2.12086722e-01 2.35535696e-01
-2.43457220e-02 2.41146144e-03 -1.11687407e-02 7.74166763e-01
3.04622233e-01 -4.03930575e-01 -3.38460058e-01 1.20476432e-01
-1.60216875e-02 -2.98220962e-01 -8.85831475e-01 1.81481194e+00
-2.82961488e-01 5.55694401e-01 -3.92231792e-01 -1.41092312e+00
8.84827077e-01 2.21572518e-01 4.40938950e-01 -7.77484715e-01
3.60328406e-01 6.65820986e-02 -2.53315330e-01 -4.37499017e-01
2.25727618e-01 -4.06220645e-01 -3.64991953e-03 2.49555990e-01
7.43212402e-01 -5.63370623e-02 3.32271487e-01 2.36758009e-01
1.35517406e+00 1.02265157e-01 6.96087599e-01 -2.57763356e-01
1.77902028e-01 1.01440445e-01 4.91389632e-01 1.00660014e+00
-6.50544703e-01 7.81920850e-01 3.44179094e-01 -3.47800940e-01
-1.16665030e+00 -1.02603149e+00 -1.49126247e-01 1.52580333e+00
-1.12981223e-01 -3.67885679e-01 -6.37142241e-01 -9.01526928e-01
-1.20090358e-01 8.62657785e-01 -1.17943704e+00 -6.26603842e-01
-6.19705580e-02 -7.30401337e-01 4.21202064e-01 5.74844480e-01
3.03545564e-01 -1.12690771e+00 -7.97462642e-01 8.62497464e-02
3.98654342e-01 -6.78064942e-01 2.25672405e-02 7.63325512e-01
-6.67073607e-01 -9.79876041e-01 -9.50592220e-01 -7.65063703e-01
4.85043526e-01 3.25718015e-01 1.01006389e+00 -1.93142325e-01
-7.78729558e-01 6.77854896e-01 -7.24264562e-01 -4.63179559e-01
-2.81022727e-01 2.60064527e-02 -1.45704746e-01 1.38990879e-01
6.06284916e-01 -3.41766775e-01 -4.88571674e-01 1.89762592e-01
-9.28695798e-01 -2.89725978e-02 6.44741297e-01 1.17561305e+00
5.09226322e-01 -2.88464427e-01 6.20445430e-01 -1.00173664e+00
3.23807061e-01 -7.68483400e-01 -1.98581576e-01 2.73831427e-01
-8.11508000e-01 -2.48044170e-02 6.13347948e-01 -6.41074181e-01
-1.00109148e+00 2.53457129e-01 -2.94767600e-03 -8.39413106e-01
-4.19152260e-01 2.12314457e-01 1.31249115e-01 -1.28146067e-01
1.06044996e+00 3.44966412e-01 -3.87410745e-02 -2.69480973e-01
5.80477417e-01 7.82782257e-01 2.33203664e-01 -1.90914840e-01
7.60166883e-01 6.06906474e-01 -6.46338165e-02 -9.45364237e-01
-9.54982877e-01 -9.36266422e-01 -7.81349003e-01 -4.18778926e-01
6.62719786e-01 -8.15783381e-01 -2.56105453e-01 6.10553920e-01
-7.00390756e-01 -6.49888635e-01 -8.47049832e-01 3.92969877e-01
-9.90295410e-01 1.23990946e-01 -4.81500715e-01 -8.50188315e-01
-4.87449795e-01 -8.85349154e-01 1.04053795e+00 2.18247205e-01
-1.44424617e-01 -1.09512806e+00 5.08820474e-01 2.19163015e-01
4.81891960e-01 9.47748199e-02 5.29899895e-01 -1.07091498e+00
-1.02154903e-01 -3.15959573e-01 -1.65710732e-01 4.17306215e-01
-5.10161892e-02 -1.37134045e-01 -1.40287566e+00 -3.81775916e-01
-1.05594255e-01 -1.02183199e+00 1.29411876e+00 2.92866766e-01
8.26141357e-01 1.22974031e-02 -1.42990395e-01 5.65704226e-01
1.59093046e+00 -3.50868478e-02 6.26859486e-01 2.83275574e-01
7.03121722e-01 4.81579989e-01 8.07947636e-01 4.53381538e-01
1.28508747e-01 5.41251361e-01 3.82663548e-01 -6.40341714e-02
-1.67413056e-01 5.97931780e-02 2.45614067e-01 8.16001058e-01
-6.12416752e-02 -1.18819036e-01 -1.04722524e+00 6.37150347e-01
-2.03049517e+00 -1.20737171e+00 2.61239320e-01 2.18257165e+00
8.19230795e-01 3.19097281e-01 1.58872381e-01 3.35254848e-01
6.72849298e-01 1.85804635e-01 -6.40094757e-01 -4.77350324e-01
8.10296685e-02 2.35490993e-01 4.49161202e-01 2.28289694e-01
-1.33020329e+00 8.91271591e-01 5.29086971e+00 1.19491839e+00
-1.13219535e+00 4.54287648e-01 6.72581434e-01 -3.64711732e-01
1.14257939e-01 6.15277588e-02 -7.78458536e-01 3.54734242e-01
1.09091508e+00 -2.02009857e-01 1.61029100e-01 1.23253226e+00
-2.10948423e-01 -9.26257446e-02 -1.09601116e+00 9.45333362e-01
3.65261227e-01 -1.44324279e+00 -2.71761894e-01 -2.05587894e-02
8.48635912e-01 3.64140421e-01 4.12947200e-02 8.93270314e-01
2.49599889e-01 -8.11257243e-01 6.14246309e-01 5.95323443e-01
9.11759019e-01 -7.09395826e-01 5.94377697e-01 3.17532748e-01
-1.08527660e+00 -3.92206192e-01 -4.92802858e-01 -8.78824815e-02
-1.20496377e-01 5.15193164e-01 -8.53757501e-01 3.53429049e-01
6.65157139e-01 8.22646797e-01 -8.27318966e-01 1.32121980e+00
1.42962635e-01 7.51870871e-01 -7.78457299e-02 -2.11944535e-01
5.17789006e-01 3.16758692e-01 4.83083904e-01 1.31601870e+00
-3.72217558e-02 2.15232838e-02 1.48111030e-01 6.02624178e-01
5.03114164e-02 2.02628657e-01 -8.45636904e-01 2.74191657e-03
2.91044384e-01 1.51333952e+00 -1.10235214e+00 -6.46535695e-01
-4.86090422e-01 1.03938329e+00 6.90512717e-01 1.95310727e-01
-6.32590711e-01 -4.90747035e-01 1.89332947e-01 -6.74024001e-02
5.00044942e-01 1.22793302e-01 -3.94808128e-02 -1.10766077e+00
-3.17036986e-01 -5.19903123e-01 4.69045043e-01 -5.25213122e-01
-1.50940204e+00 4.04759377e-01 1.95958897e-01 -1.44503510e+00
-3.68509382e-01 -5.59781432e-01 -7.05601811e-01 3.38745266e-01
-1.43485212e+00 -1.16367435e+00 -4.33171332e-01 7.06191361e-01
9.55455899e-01 -2.95844406e-01 8.60527933e-01 8.31297264e-02
-5.90304971e-01 6.64449513e-01 3.74043047e-01 6.89716451e-03
7.41719067e-01 -1.36036015e+00 4.10628647e-01 5.95424235e-01
2.79693305e-01 2.89511889e-01 6.06325448e-01 -4.32716936e-01
-1.25420523e+00 -1.31143844e+00 4.45782602e-01 -3.82828414e-01
7.47145295e-01 -5.61169624e-01 -9.51348841e-01 4.91789341e-01
2.07041353e-01 6.18111730e-01 8.84391546e-01 -8.43312219e-02
-4.48492169e-01 -3.03647444e-02 -1.04380298e+00 2.70062298e-01
9.78430629e-01 -4.58426982e-01 -7.02123046e-01 3.51344675e-01
6.87930465e-01 9.52957645e-02 -6.41537547e-01 3.93600792e-01
3.63455176e-01 -1.04073918e+00 8.03499043e-01 -6.32300794e-01
4.61068988e-01 1.60354331e-01 -1.33488744e-01 -1.49300492e+00
-4.62699831e-01 -5.11473894e-01 -4.16613728e-01 1.12647510e+00
2.07395241e-01 -5.36671042e-01 8.07740569e-01 4.11614060e-01
-3.06110144e-01 -1.00149357e+00 -1.06069112e+00 -1.10135138e+00
7.94426426e-02 -3.65248442e-01 -8.78151506e-02 1.12481606e+00
8.98227319e-02 6.16677999e-01 -4.17476445e-01 -5.71048856e-01
9.49550152e-01 -2.83534318e-01 6.81751847e-01 -1.32464933e+00
-2.95324594e-01 -3.16806704e-01 -7.98136711e-01 -3.10437411e-01
1.99337333e-01 -9.60385025e-01 2.70461768e-01 -1.35040629e+00
4.13220555e-01 -2.05273613e-01 -4.86021936e-01 7.19856441e-01
-1.54904872e-01 6.68425977e-01 1.96557909e-01 2.66921759e-01
-1.12414992e+00 7.43797064e-01 7.67292917e-01 -1.38819903e-01
-1.74329370e-01 -2.03347832e-01 -3.31359386e-01 5.21342933e-01
7.66261697e-01 -5.45659900e-01 -6.42211795e-01 1.06494203e-01
-2.27264874e-02 -4.43575144e-01 5.46177745e-01 -1.20334256e+00
3.57422382e-01 -6.00157157e-02 3.81593257e-01 -1.50615424e-01
5.60675025e-01 -6.98823571e-01 -2.24678174e-01 4.71525371e-01
-5.60810864e-01 -6.04483664e-01 -6.67883456e-02 7.49443829e-01
-3.07137892e-02 -3.75197351e-01 1.19334829e+00 -2.21454591e-01
-1.20776856e+00 3.56250077e-01 -2.60144532e-01 3.30199897e-01
1.55642736e+00 -4.32994038e-01 -5.24104416e-01 -1.31293461e-01
-6.95516527e-01 -2.94542629e-02 4.22922164e-01 4.24252480e-01
6.31135881e-01 -1.33872950e+00 -6.04615986e-01 1.91240519e-01
7.60699630e-01 -3.73255342e-01 3.36867660e-01 1.06304133e+00
-2.04269662e-01 1.87624335e-01 -2.65404671e-01 -7.42999613e-01
-1.25487936e+00 9.03090298e-01 3.02987903e-01 4.49322676e-03
-7.38113940e-01 9.83958840e-01 -1.33007369e-03 -3.62697780e-01
1.80988580e-01 2.40423620e-01 -1.35350183e-01 4.13962066e-01
7.04288304e-01 5.62957048e-01 2.65971512e-01 -6.84797585e-01
-3.17307651e-01 3.52047980e-01 -3.04468930e-01 -8.50081444e-03
1.39997995e+00 -6.28257319e-02 3.96722406e-01 1.00702524e+00
1.69433188e+00 -7.37992465e-01 -1.54738545e+00 -5.57838619e-01
-6.26246864e-03 -3.33867997e-01 2.82425433e-01 -5.54495156e-01
-8.20859849e-01 1.05857062e+00 9.28721070e-01 3.46451461e-01
8.44057441e-01 1.63814962e-01 3.66189957e-01 4.75094378e-01
2.38571525e-01 -1.38824689e+00 3.06186050e-01 4.34532225e-01
5.12082934e-01 -1.65230882e+00 -4.55087610e-02 4.20879945e-03
-9.33926761e-01 9.76179302e-01 6.00471258e-01 -1.91870078e-01
7.18141437e-01 4.69015576e-02 1.43972903e-01 -4.06624377e-01
-1.14633036e+00 -6.30660832e-01 2.78502971e-01 6.03915155e-01
1.33810639e-01 -1.71510220e-01 -3.69941853e-02 2.54650235e-01
2.53031373e-01 1.98966056e-01 3.26286584e-01 1.32665777e+00
-7.62350559e-01 -4.89164829e-01 -1.89344510e-02 5.78720689e-01
-6.25441223e-02 -1.59225911e-01 -3.69634777e-01 6.38266087e-01
-4.76968177e-02 8.95896137e-01 1.25556812e-01 -4.23633099e-01
1.56005964e-01 4.69735473e-01 4.25813556e-01 -9.05318618e-01
-4.36282992e-01 -1.02679133e-01 -1.43103659e-01 -6.15981102e-01
-4.37043011e-01 -4.39588904e-01 -9.36905801e-01 -3.91190946e-02
-4.36593115e-01 -1.85608137e-02 5.38292825e-01 1.10324752e+00
3.00277203e-01 3.55473161e-01 8.47484827e-01 -1.18994653e+00
-8.52739871e-01 -9.97037947e-01 -6.41510904e-01 4.56513166e-01
2.74069160e-01 -8.97979558e-01 -6.65452778e-01 -3.67133468e-02] | [9.946184158325195, 2.8480257987976074] |
de550431-7813-4c06-ae1b-01a14e41b5e4 | learning-single-multi-attribute-of-object | 2110.04603 | null | https://arxiv.org/abs/2110.04603v1 | https://arxiv.org/pdf/2110.04603v1.pdf | Learning Single/Multi-Attribute of Object with Symmetry and Group | Attributes and objects can compose diverse compositions. To model the compositional nature of these concepts, it is a good choice to learn them as transformations, e.g., coupling and decoupling. However, complex transformations need to satisfy specific principles to guarantee rationality. Here, we first propose a previously ignored principle of attribute-object transformation: Symmetry. For example, coupling peeled-apple with attribute peeled should result in peeled-apple, and decoupling peeled from apple should still output apple. Incorporating the symmetry, we propose a transformation framework inspired by group theory, i.e., SymNet. It consists of two modules: Coupling Network and Decoupling Network. We adopt deep neural networks to implement SymNet and train it in an end-to-end paradigm with the group axioms and symmetry as objectives. Then, we propose a Relative Moving Distance (RMD) based method to utilize the attribute change instead of the attribute pattern itself to classify attributes. Besides the compositions of single-attribute and object, our RMD is also suitable for complex compositions of multiple attributes and objects when incorporating attribute correlations. SymNet can be utilized for attribute learning, compositional zero-shot learning and outperforms the state-of-the-art on four widely-used benchmarks. Code is at https://github.com/DirtyHarryLYL/SymNet. | ['Cewu Lu', 'Xiaohan Mao', 'Xinyu Xu', 'Yue Xu', 'Yong-Lu Li'] | 2021-10-09 | null | null | null | null | ['compositional-zero-shot-learning'] | ['computer-vision'] | [ 1.33775249e-01 -4.82413210e-02 -9.94384363e-02 -6.52100265e-01
-1.36441916e-01 -4.40775424e-01 6.62572443e-01 -1.77261472e-01
-1.89739704e-01 2.88723260e-01 2.10534975e-01 1.27053693e-01
-2.48089150e-01 -1.10675025e+00 -6.78837717e-01 -8.04409862e-01
3.70373130e-01 6.14377379e-01 2.30122045e-01 -3.24152857e-01
-6.32011071e-02 2.53201962e-01 -1.71712923e+00 3.21598738e-01
9.56745565e-01 1.11742604e+00 1.30991295e-01 9.25738290e-02
-3.09966564e-01 6.70650661e-01 -2.32759088e-01 -6.84853315e-01
4.61155176e-01 -5.61644197e-01 -5.98040164e-01 1.22288518e-01
2.06161618e-01 -1.69529706e-01 -1.63026795e-01 1.10476351e+00
5.01697123e-01 3.29657018e-01 8.04956436e-01 -1.77576888e+00
-1.30746841e+00 8.86820376e-01 -2.48894349e-01 -2.99898207e-01
1.04406729e-01 2.89842427e-01 1.53058445e+00 -9.98969853e-01
3.82126212e-01 1.42541301e+00 3.72755110e-01 5.84785879e-01
-1.29956174e+00 -8.52864861e-01 3.47736210e-01 5.71948588e-01
-1.42802310e+00 -2.74228454e-01 9.27772820e-01 -3.79480124e-01
5.86067796e-01 3.63028854e-01 6.92331672e-01 1.09878361e+00
-1.26636282e-01 9.89614904e-01 7.65633225e-01 -1.42325461e-01
2.47901499e-01 3.19522880e-02 -2.33727694e-02 3.60236615e-01
1.35842443e-01 7.28030689e-03 -2.50929058e-01 6.78398535e-02
3.52603167e-01 4.07491237e-01 -1.00377940e-01 -7.51967788e-01
-1.52253091e+00 7.64010549e-01 5.69300711e-01 9.84757245e-02
-2.62905866e-01 1.40370965e-01 3.66238832e-01 3.68563890e-01
2.14760721e-01 3.38690102e-01 -4.53696430e-01 2.47047752e-01
-2.62243867e-01 2.93267190e-01 5.91020107e-01 1.38803697e+00
8.94658685e-01 -2.74785068e-02 -2.42790952e-01 9.88465488e-01
4.11369771e-01 3.76890957e-01 4.08750743e-01 -9.16180789e-01
8.98825079e-02 8.50950122e-01 -2.22108588e-01 -7.97872782e-01
-2.86148489e-01 -3.37914139e-01 -1.12510157e+00 7.04817697e-02
-5.04382327e-03 1.96148276e-01 -8.62820983e-01 2.00096560e+00
5.24898112e-01 2.86060959e-01 -5.06791994e-02 9.39977944e-01
8.87043357e-01 6.08814001e-01 1.74280554e-01 -2.83924509e-02
1.39205337e+00 -9.63976562e-01 -6.83878303e-01 7.57892504e-02
5.42775750e-01 -6.52272522e-01 1.48610818e+00 2.63494670e-01
-9.03269887e-01 -5.70126534e-01 -1.09271216e+00 -1.04114845e-01
-5.15000701e-01 -1.41228199e-01 6.64779782e-01 2.91613966e-01
-5.33833206e-01 6.51233196e-01 -6.17052019e-01 -4.32174861e-01
5.20288348e-01 3.14288467e-01 -2.56136447e-01 -4.86305356e-03
-1.37134171e+00 6.58368230e-01 6.61723256e-01 -1.62241906e-01
-8.08504343e-01 -9.91697669e-01 -9.10451412e-01 2.81230122e-01
6.44736767e-01 -9.38508511e-01 1.25982857e+00 -1.04538572e+00
-1.55402303e+00 7.15757072e-01 2.43086264e-01 -5.15541099e-02
5.23147941e-01 -1.66997984e-01 -6.08845115e-01 -3.58555228e-01
1.87999755e-01 6.10101581e-01 6.01059496e-01 -1.11694956e+00
-8.20683956e-01 -2.50890017e-01 -4.57955189e-02 2.22166687e-01
-5.84999681e-01 -2.07308196e-02 -2.83728719e-01 -6.52723134e-01
1.50273904e-01 -9.23123777e-01 1.31958514e-01 1.47150978e-01
-5.67150474e-01 -5.66957712e-01 7.38007843e-01 -7.53770918e-02
1.03132248e+00 -2.30672169e+00 1.81577414e-01 1.42954499e-01
2.82850146e-01 7.51700327e-02 -2.32929766e-01 4.50520098e-01
-2.32654989e-01 2.92445440e-02 -4.58345801e-01 -9.47797671e-02
4.60594714e-01 1.94450632e-01 -1.02640010e-01 3.19251806e-01
3.60710979e-01 1.09982276e+00 -1.00728035e+00 -4.16891366e-01
2.18574211e-01 3.75174850e-01 -6.05079174e-01 1.72507659e-01
-3.41056734e-01 1.52800128e-01 -3.78280967e-01 9.16828990e-01
7.43660390e-01 -1.70757547e-01 -1.10018160e-02 -5.80720186e-01
3.35511044e-02 4.15026508e-02 -1.43880224e+00 1.56538737e+00
-3.35452437e-01 4.32414822e-02 -2.96603203e-01 -9.11819816e-01
1.18426359e+00 1.61670044e-01 6.38453245e-01 -5.97348273e-01
2.69437850e-01 2.59993047e-01 3.11236799e-01 -3.64717096e-01
2.46488582e-02 -2.90938616e-01 -1.35129511e-01 4.50429708e-01
1.51866436e-01 8.84094462e-02 2.10140660e-01 9.06109586e-02
8.15564752e-01 3.20591390e-01 5.05564094e-01 -3.41234624e-01
6.48009956e-01 -2.92764992e-01 1.00288916e+00 2.94654191e-01
-6.14693109e-03 5.55472970e-01 5.30350745e-01 -6.21181607e-01
-1.21523726e+00 -1.28724349e+00 3.20919007e-02 1.52570605e+00
4.24596637e-01 -4.28939611e-01 -4.00114626e-01 -5.78968346e-01
1.92955598e-01 1.01358056e+00 -6.79837525e-01 -6.19875729e-01
-4.17621970e-01 -6.67268872e-01 3.28850836e-01 6.50437534e-01
6.46554410e-01 -1.22786713e+00 -3.54045838e-01 2.94830710e-01
-1.75843462e-01 -8.30198884e-01 -8.34150195e-01 1.66317627e-01
-3.91586572e-01 -9.43508983e-01 -2.87714601e-01 -8.19456220e-01
5.02865016e-01 1.47537723e-01 1.24675405e+00 5.86372018e-02
-1.85896158e-01 8.13679919e-02 -5.02869129e-01 -5.31671584e-01
-1.16714537e-01 1.55593321e-01 1.59033954e-01 5.51659703e-01
7.02598155e-01 -9.93472219e-01 -6.44193470e-01 6.51654124e-01
-1.12115884e+00 1.28377691e-01 8.46461833e-01 8.56727540e-01
7.00737417e-01 -1.96559414e-01 5.17455041e-01 -9.32545245e-01
6.09514952e-01 -4.79400814e-01 -3.38514537e-01 4.33412135e-01
-7.38359869e-01 1.44202590e-01 9.10980642e-01 -8.19202840e-01
-8.75190437e-01 2.35259503e-01 9.57075581e-02 -6.10047281e-01
-1.97143868e-01 1.35116801e-01 -1.02113664e+00 3.41998845e-01
4.53531772e-01 2.33114272e-01 -1.60785392e-02 -4.26445544e-01
6.81426227e-01 5.41030467e-01 5.66325188e-01 -8.36507261e-01
8.16999555e-01 4.92169112e-01 7.81124458e-03 -1.89514190e-01
-7.26072669e-01 -2.73825318e-01 -6.97725236e-01 -4.20980342e-02
9.07001138e-01 -6.35207236e-01 -9.90624666e-01 4.90378618e-01
-9.81784105e-01 1.46340802e-02 -4.65814173e-01 3.61461282e-01
-6.80447459e-01 5.56326173e-02 -2.70578533e-01 -1.81224048e-01
-3.98760259e-01 -9.66118813e-01 8.01238179e-01 6.83767870e-02
-2.04263985e-01 -6.56334877e-01 -1.69937134e-01 -1.11762270e-01
3.62225533e-01 2.69074976e-01 1.23992097e+00 -1.06596076e+00
-6.06832027e-01 2.06047986e-02 -1.36409149e-01 4.35502172e-01
4.36376214e-01 3.25133026e-01 -7.78476954e-01 -1.62033811e-02
-8.69332999e-02 -1.94829792e-01 6.96412385e-01 1.50216110e-02
1.16448665e+00 -3.98706317e-01 -1.44457191e-01 9.08263981e-01
1.32316709e+00 3.91848773e-01 6.34475529e-01 5.16042531e-01
9.60160971e-01 7.12947130e-01 6.77695394e-01 6.38393998e-01
5.78036427e-01 6.80060089e-01 4.99360353e-01 1.57242179e-01
-1.02274135e-01 -2.91637450e-01 1.71877593e-01 9.09866333e-01
-1.43990973e-02 -5.41262440e-02 -8.56103241e-01 3.37483823e-01
-2.06804204e+00 -1.04310107e+00 -1.96085665e-02 2.13303328e+00
8.85759294e-01 5.70510812e-02 8.90336335e-02 -1.42496191e-02
9.22562778e-01 -4.24647145e-02 -7.43994355e-01 -2.27474779e-01
-1.38343200e-01 -5.99486493e-02 1.55626148e-01 1.66125551e-01
-1.06747997e+00 9.85417962e-01 4.42505217e+00 7.24799871e-01
-9.35345054e-01 8.69258642e-02 4.18538511e-01 -7.45479017e-02
-6.54370427e-01 6.64337277e-02 -7.46186316e-01 6.76919758e-01
2.93186814e-01 -5.12420118e-01 5.86122751e-01 9.45472240e-01
-1.36876389e-01 5.93477011e-01 -1.55991924e+00 8.95343423e-01
6.24773763e-02 -9.88378286e-01 3.26123118e-01 -1.98801830e-01
5.75302780e-01 -4.53956753e-01 7.36043826e-02 6.03439212e-01
5.20836174e-01 -9.51854825e-01 7.49206901e-01 4.83478129e-01
6.61193669e-01 -8.41671705e-01 4.91392970e-01 2.01236904e-01
-1.49559414e+00 -1.33914083e-01 -6.20249629e-01 1.50069356e-01
-6.03922457e-02 4.40993160e-01 -4.45330322e-01 7.25581050e-01
6.74900711e-01 9.13273513e-01 -4.05717760e-01 8.35069120e-01
-3.36823612e-01 1.15646362e-01 -1.51698321e-01 -8.52173567e-02
1.32123455e-01 -4.33001876e-01 3.14854980e-01 8.71723771e-01
4.47005391e-01 1.82479605e-01 2.33927369e-01 1.14312458e+00
-2.57387042e-01 2.31749430e-01 -4.62768525e-01 2.35637575e-01
9.34837162e-01 1.41859305e+00 -4.24725562e-01 -4.59159821e-01
-5.45360923e-01 8.22194576e-01 1.91155240e-01 1.64198607e-01
-9.85192060e-01 -5.51939666e-01 9.42801595e-01 -5.89906462e-02
4.15789127e-01 2.04127505e-01 -2.99228817e-01 -1.12308979e+00
1.00577369e-01 -1.11982608e+00 5.25461674e-01 -6.71519697e-01
-1.68498683e+00 4.75608706e-01 9.76456329e-02 -1.62754345e+00
2.32295960e-01 -3.76875430e-01 -8.90514851e-01 7.04836369e-01
-1.26455164e+00 -1.39841795e+00 -7.54377186e-01 8.10988009e-01
5.29401481e-01 -3.07614893e-01 6.14176929e-01 3.92849177e-01
-4.61356103e-01 6.23359799e-01 -9.60920453e-02 2.15471387e-02
9.99997437e-01 -1.27727914e+00 4.54108685e-01 4.50820178e-01
6.39594942e-02 6.77889109e-01 5.81839085e-01 -5.39016008e-01
-1.35885060e+00 -1.36562729e+00 7.15180516e-01 -3.35941494e-01
7.34046400e-01 -5.69267333e-01 -1.10470665e+00 8.11331034e-01
1.53964639e-01 1.14480712e-01 7.56571949e-01 4.03488688e-02
-7.88878322e-01 -6.66766226e-01 -8.53313088e-01 7.72637367e-01
1.31438422e+00 -1.98776767e-01 -7.11710334e-01 3.52335721e-01
1.10366106e+00 9.27439556e-02 -9.35633659e-01 6.38726294e-01
5.70143998e-01 -8.06207538e-01 1.00362706e+00 -6.48432970e-01
4.53347385e-01 -4.78325903e-01 -3.96005720e-01 -1.50861263e+00
-9.03515875e-01 -4.53516155e-01 1.74371817e-03 1.59112346e+00
3.99576455e-01 -7.65821576e-01 5.69888294e-01 6.39053226e-01
-3.04965526e-01 -7.48469234e-01 -6.31686091e-01 -1.14881372e+00
6.52337074e-02 -1.18965134e-01 1.33051968e+00 1.13164413e+00
-3.31204653e-01 5.42171299e-01 -4.33392644e-01 -5.99251017e-02
6.36946857e-01 1.90147832e-01 7.07989395e-01 -1.50449359e+00
-2.28542775e-01 -7.31955051e-01 -5.19709587e-01 -6.75926626e-01
2.12116137e-01 -1.22660637e+00 -4.91287261e-02 -1.35932088e+00
2.49551460e-01 -5.03933668e-01 -6.42455399e-01 6.29538298e-01
-2.66547471e-01 -1.91898540e-01 3.47860664e-01 4.55802679e-01
-4.24776316e-01 1.05774319e+00 1.34544897e+00 -3.24689865e-01
-1.46751180e-01 -6.25223145e-02 -9.47790980e-01 5.37562549e-01
1.04782009e+00 -5.88911474e-01 -6.48496270e-01 -2.72434592e-01
2.60081708e-01 -5.73050499e-01 2.72126377e-01 -9.52144980e-01
3.01245242e-01 -4.65192884e-01 2.54355609e-01 -4.01680321e-01
2.40053833e-01 -1.06886196e+00 3.75125110e-01 4.07304913e-01
-5.14868975e-01 2.14853674e-01 -3.63728225e-01 4.47042227e-01
-1.12199828e-01 -4.88368310e-02 8.07059646e-01 -1.60857081e-01
-8.55015874e-01 6.30986989e-01 2.16953933e-01 2.02270612e-01
1.38751709e+00 -6.53499365e-02 -4.19318497e-01 -2.48620100e-02
-5.67376494e-01 4.29957092e-01 5.08249104e-01 7.10947514e-01
6.04585469e-01 -2.05192733e+00 -8.96823168e-01 2.22650990e-01
4.94115353e-01 6.68278039e-02 1.02321163e-01 8.47459674e-01
-2.98082918e-01 -3.47663872e-02 -5.60070038e-01 -5.34272671e-01
-1.11939442e+00 9.77210820e-01 1.48474857e-01 2.12571651e-01
-6.36571944e-01 7.64392078e-01 5.76389015e-01 -9.44089770e-01
2.86148489e-01 -1.88878223e-01 -2.76672021e-02 1.84931368e-01
3.98048609e-01 3.03460538e-01 -5.37549444e-02 -3.07582825e-01
-4.36202109e-01 5.86721897e-01 -1.31341457e-01 2.27098495e-01
1.26578689e+00 1.16387479e-01 -3.74830663e-01 6.64256036e-01
1.10179794e+00 -3.73977125e-01 -1.11609364e+00 -5.69056511e-01
7.36704171e-02 -5.12153327e-01 -4.30697203e-01 -6.09037519e-01
-1.04801714e+00 8.70651007e-01 4.44088429e-01 1.13875128e-01
1.22011817e+00 1.48564920e-01 6.19230151e-01 5.79544663e-01
1.08198933e-01 -1.10944605e+00 2.67049491e-01 3.39780390e-01
8.76513422e-01 -1.11456370e+00 -1.23246782e-01 -4.97599125e-01
-1.01898825e+00 9.66864288e-01 1.00955832e+00 -6.53815418e-02
6.84901118e-01 -4.59880233e-02 -6.69487566e-02 -2.71078479e-02
-9.91802573e-01 -2.28945091e-01 4.00861740e-01 4.67329472e-01
3.70440811e-01 3.74021351e-01 -2.45132446e-01 7.84915745e-01
-2.02724352e-01 -1.66098684e-01 1.86079785e-01 7.38568842e-01
-4.65131342e-01 -1.22406650e+00 -1.20799541e-01 4.66089964e-01
-3.48520116e-03 -1.70579404e-01 -4.73976970e-01 8.05244446e-01
5.63707471e-01 6.38868093e-01 2.02259094e-01 -6.61315620e-01
5.76900721e-01 1.38748482e-01 1.29659623e-01 -7.04452813e-01
-4.56674725e-01 -1.53480947e-01 -2.03189760e-01 -5.46774805e-01
-2.94927955e-01 -6.47415578e-01 -1.22336519e+00 -5.47964931e-01
-5.55260405e-02 -9.66644287e-02 3.59029442e-01 6.30505562e-01
3.71887356e-01 6.76110804e-01 8.13275635e-01 -2.60462582e-01
-8.35229039e-01 -9.05761182e-01 -5.80741584e-01 7.85047650e-01
-8.65929425e-02 -7.72744179e-01 -2.80029118e-01 5.29621504e-02] | [10.118444442749023, 2.3518729209899902] |
7cf294b8-14fe-40ab-b1e7-39f7b6786391 | wise-whitebox-image-stylization-by-example | 2207.14606 | null | https://arxiv.org/abs/2207.14606v1 | https://arxiv.org/pdf/2207.14606v1.pdf | WISE: Whitebox Image Stylization by Example-based Learning | Image-based artistic rendering can synthesize a variety of expressive styles using algorithmic image filtering. In contrast to deep learning-based methods, these heuristics-based filtering techniques can operate on high-resolution images, are interpretable, and can be parameterized according to various design aspects. However, adapting or extending these techniques to produce new styles is often a tedious and error-prone task that requires expert knowledge. We propose a new paradigm to alleviate this problem: implementing algorithmic image filtering techniques as differentiable operations that can learn parametrizations aligned to certain reference styles. To this end, we present WISE, an example-based image-processing system that can handle a multitude of stylization techniques, such as watercolor, oil or cartoon stylization, within a common framework. By training parameter prediction networks for global and local filter parameterizations, we can simultaneously adapt effects to reference styles and image content, e.g., to enhance facial features. Our method can be optimized in a style-transfer framework or learned in a generative-adversarial setting for image-to-image translation. We demonstrate that jointly training an XDoG filter and a CNN for postprocessing can achieve comparable results to a state-of-the-art GAN-based method. | ['Matthias Trapp', 'Jürgen Döllner', 'Amir Semmo', 'Martin Büssemeyer', 'Max Reimann', 'Winfried Lötzsch'] | 2022-07-29 | null | null | null | null | ['image-stylization', 'parameter-prediction'] | ['computer-vision', 'miscellaneous'] | [ 6.58888459e-01 -1.56517997e-02 1.04108177e-01 -4.87167209e-01
-5.45267642e-01 -8.72816801e-01 7.14737177e-01 -4.64558929e-01
-1.71824411e-01 5.83567441e-01 -4.78834771e-02 -1.55495510e-01
2.11959139e-01 -9.80655134e-01 -9.49268222e-01 -5.25917590e-01
5.06460428e-01 3.52097660e-01 -5.07660173e-02 -5.22032559e-01
2.01273635e-01 1.00305104e+00 -1.51312351e+00 4.16329712e-01
8.81524861e-01 9.36366796e-01 3.07069123e-02 8.83023202e-01
-3.21476400e-01 4.36030865e-01 -7.69598901e-01 -6.80690706e-01
4.81147796e-01 -5.45676589e-01 -4.29153115e-01 3.03809404e-01
7.26930022e-01 -3.43567938e-01 -1.30092567e-02 1.00516951e+00
4.03331548e-01 1.09083988e-02 6.13164842e-01 -1.12488651e+00
-9.81562316e-01 1.92716584e-01 -4.96749043e-01 -3.89998525e-01
3.11373383e-01 4.63942558e-01 6.60788298e-01 -7.54824698e-01
7.38048434e-01 1.63922513e+00 4.90632385e-01 9.23408091e-01
-1.80086064e+00 -8.44707906e-01 9.76935476e-02 -2.52498955e-01
-9.18247819e-01 -4.06486392e-01 1.00007308e+00 -3.49339932e-01
2.94603974e-01 4.55378205e-01 9.37214315e-01 1.25002694e+00
2.26455078e-01 5.99552691e-01 1.49351287e+00 -4.14414167e-01
2.70314306e-01 -3.94044295e-02 -7.47331381e-01 8.47995222e-01
-1.53042480e-01 1.73470542e-01 -5.23076594e-01 1.31674949e-02
1.42890155e+00 -9.09642726e-02 -2.13866889e-01 -3.09879541e-01
-1.13000822e+00 8.85034442e-01 4.67988700e-01 -9.88618731e-02
-1.60067320e-01 5.19642532e-01 2.06610173e-01 5.43007672e-01
4.93483096e-01 1.06368971e+00 -5.68719745e-01 1.13816303e-03
-9.87979352e-01 5.47239244e-01 7.10616350e-01 9.77874875e-01
1.05142999e+00 4.71039355e-01 -2.86298454e-01 8.61094534e-01
-2.29822262e-03 5.15664577e-01 2.26195768e-01 -1.22599876e+00
2.90487912e-02 3.63047779e-01 7.64750615e-02 -7.60302484e-01
-1.18765816e-01 -2.94784665e-01 -8.88818264e-01 8.94237101e-01
3.87119502e-01 -2.25902408e-01 -1.34116530e+00 1.71865654e+00
2.08446503e-01 -5.12658134e-02 -2.61327416e-01 8.08151841e-01
5.78263760e-01 5.41399240e-01 2.25436315e-01 7.82646611e-02
1.25749123e+00 -9.44500864e-01 -6.61217034e-01 -1.34752333e-01
1.88595597e-02 -1.07727921e+00 1.64452803e+00 5.32883406e-01
-1.37195051e+00 -6.36014044e-01 -9.11377370e-01 -2.57641375e-01
-3.77334297e-01 9.08193588e-02 6.92011416e-01 7.61325836e-01
-1.24918044e+00 8.54307294e-01 -7.12019682e-01 -2.32619956e-01
5.41505158e-01 3.77092689e-01 -1.39547870e-01 3.56224179e-01
-8.58778834e-01 6.98382735e-01 1.03953116e-01 -1.74168125e-01
-7.82482803e-01 -1.13424027e+00 -6.42242014e-01 2.21899673e-02
2.05542237e-01 -1.23023093e+00 1.09253395e+00 -1.78986418e+00
-2.25721574e+00 9.60145831e-01 -5.30931242e-02 -2.55064815e-01
6.33603215e-01 -2.23562032e-01 -3.35624337e-01 2.97880098e-02
-1.45791590e-01 1.03439569e+00 1.54450119e+00 -1.36425805e+00
-3.97861362e-01 4.58383784e-02 3.45444471e-01 2.00038210e-01
-2.08912313e-01 1.36997074e-01 -5.20062804e-01 -1.31059003e+00
-2.38267586e-01 -8.89460206e-01 -4.08616960e-01 6.25729382e-01
-3.61490071e-01 4.59282756e-01 9.50477123e-01 -4.03528422e-01
7.80564308e-01 -1.91741896e+00 5.55608213e-01 2.95083851e-01
8.93240348e-02 2.27018818e-01 -3.91956896e-01 2.30865300e-01
-4.32112142e-02 4.08659369e-01 -1.75412491e-01 -3.94187152e-01
-1.13044932e-01 2.34631166e-01 -4.30448502e-01 -2.79733492e-03
5.73333800e-01 1.03597605e+00 -8.01101685e-01 -3.27121258e-01
4.46815223e-01 7.04745829e-01 -9.04195428e-01 3.76537383e-01
-6.02336466e-01 1.08056974e+00 -4.56978887e-01 6.67756915e-01
5.37264526e-01 6.79957718e-02 -2.38468163e-02 -4.70377386e-01
8.15173760e-02 -2.22366571e-01 -9.66672599e-01 1.91278946e+00
-1.03940117e+00 6.91919982e-01 2.30624974e-01 -5.73736370e-01
1.02131128e+00 -2.10858118e-02 2.19775066e-01 -3.55542421e-01
2.27360561e-01 1.87912598e-01 -2.19974473e-01 -1.40300378e-01
5.41659772e-01 -2.89570779e-01 5.12807742e-02 3.04606080e-01
1.58185542e-01 -8.48984778e-01 6.85085207e-02 -2.47425333e-01
8.45457017e-01 4.85494405e-01 1.58609841e-02 -2.40104854e-01
4.14299846e-01 -2.99656034e-01 3.77977252e-01 5.80471218e-01
4.68356937e-01 8.65932047e-01 5.57625592e-01 -6.37115002e-01
-1.33626401e+00 -9.87383544e-01 -2.72257030e-02 1.16856992e+00
-1.43219814e-01 -2.77501374e-01 -1.08162284e+00 -4.06460285e-01
-4.40177917e-02 6.99725568e-01 -7.68783569e-01 -1.41436562e-01
-7.23468125e-01 -3.37643266e-01 4.07545716e-01 5.20052731e-01
6.23427749e-01 -1.20637989e+00 -5.42521536e-01 2.12884665e-01
4.31137919e-01 -1.01990592e+00 -8.17574501e-01 1.29953828e-02
-9.01651204e-01 -7.29545772e-01 -8.63193095e-01 -6.35859847e-01
8.47478569e-01 -2.96149284e-01 1.45471728e+00 1.84567511e-01
-2.79985577e-01 2.93525398e-01 -1.20425664e-01 -4.34848875e-01
-8.27172279e-01 6.16073348e-02 -2.42237180e-01 2.64745951e-01
-2.85731584e-01 -8.68066072e-01 -9.19388890e-01 3.26248229e-01
-1.30753481e+00 4.93018925e-01 4.44819987e-01 1.00892341e+00
7.67116010e-01 -3.03634673e-01 3.18980634e-01 -1.28265595e+00
8.67796779e-01 2.10588664e-01 -7.76248991e-01 2.30631903e-01
-3.82872760e-01 2.84764767e-01 9.45142984e-01 -7.74593294e-01
-1.15266228e+00 2.67258167e-01 -1.81900471e-01 -7.44292438e-01
-2.57770985e-01 1.29777715e-01 -4.60030526e-01 -5.30352414e-01
7.54222274e-01 1.60680502e-03 1.35207560e-03 -3.60035747e-01
1.00862181e+00 1.47703305e-01 6.66284263e-01 -9.12037671e-01
1.30091631e+00 4.63887662e-01 1.97567850e-01 -5.08514285e-01
-6.76820040e-01 4.46859449e-01 -6.91280365e-01 -2.01733768e-01
8.80167305e-01 -7.23568857e-01 -5.74776173e-01 6.12174392e-01
-1.18747878e+00 -8.66186023e-01 -6.01805806e-01 -2.84675360e-01
-8.19884241e-01 -1.05598509e-01 -4.90769833e-01 -2.55808264e-01
-3.11523795e-01 -1.44155836e+00 1.34590745e+00 3.74999553e-01
-3.75299484e-01 -1.09205997e+00 -3.25305790e-01 -2.75872853e-02
7.98649311e-01 7.10392475e-01 9.14950728e-01 7.34485984e-02
-6.91230536e-01 -8.65749922e-03 -1.31801307e-01 2.86086977e-01
4.15331006e-01 4.02860075e-01 -8.04557145e-01 -2.24130020e-01
-2.79461741e-01 -3.52493167e-01 6.56574428e-01 2.28807092e-01
1.70680630e+00 -5.25413752e-01 9.67645943e-02 1.31655276e+00
1.30635798e+00 1.32681578e-01 7.67235518e-01 3.36902022e-01
9.86997724e-01 3.12325835e-01 2.87850261e-01 2.97711253e-01
-2.05842122e-01 9.13715303e-01 2.64948159e-01 -5.45716345e-01
-4.13605362e-01 -4.47014570e-01 2.75344737e-02 4.57565226e-02
-2.91575044e-01 -5.47605716e-02 -3.63244444e-01 -6.54734001e-02
-1.48841989e+00 -8.09611082e-01 4.03995425e-01 1.93138754e+00
1.24283791e+00 3.84526663e-02 1.84795052e-01 -1.73889890e-01
5.22198975e-01 1.70971468e-01 -7.56505013e-01 -8.06562960e-01
-1.07718892e-01 8.47339451e-01 5.81607997e-01 3.50785941e-01
-8.99746835e-01 1.44482780e+00 6.40979958e+00 1.01151848e+00
-1.57320440e+00 -5.11714518e-02 9.36356187e-01 -7.92998448e-02
-7.47919977e-01 -1.80044457e-01 -3.98872077e-01 2.79570132e-01
4.89321619e-01 -1.60550687e-03 1.04467630e+00 7.48897314e-01
2.71289974e-01 3.35154206e-01 -1.17364323e+00 1.09040809e+00
-1.02314325e-02 -1.67491913e+00 4.68023807e-01 -1.74563602e-01
8.75587463e-01 -7.40502596e-01 5.99702537e-01 5.65096326e-02
4.58992660e-01 -1.29602742e+00 9.86629248e-01 5.62348723e-01
1.34108043e+00 -8.11349630e-01 -2.32006423e-02 -2.02304631e-01
-8.99685144e-01 1.75132364e-01 -9.74369794e-02 2.45749116e-01
1.77924067e-01 2.85804182e-01 -2.79365331e-01 7.24370405e-02
6.59434259e-01 5.26027262e-01 -5.68711698e-01 6.02374196e-01
-5.45088232e-01 4.16066259e-01 -2.19509639e-02 8.21132436e-02
3.29760797e-02 -4.49770182e-01 4.89720732e-01 1.03729606e+00
3.64106327e-01 6.70687184e-02 -5.95412068e-02 1.37314057e+00
-3.92638713e-01 1.11950394e-02 -5.01561284e-01 -4.75615636e-02
2.46392474e-01 1.50982726e+00 -7.26466954e-01 -2.17606008e-01
-1.47200212e-01 1.36604393e+00 2.02728793e-01 6.34116530e-01
-8.97896647e-01 -3.59626144e-01 8.51471663e-01 3.14985424e-01
2.76022881e-01 -2.80065686e-01 -5.93713403e-01 -1.21012807e+00
-2.81232476e-01 -1.32390761e+00 -2.79697984e-01 -1.04024720e+00
-1.16273963e+00 8.45644653e-01 -1.79712594e-01 -1.17009747e+00
-1.97580904e-01 -7.42217422e-01 -8.57039809e-01 9.20575678e-01
-1.34774446e+00 -1.48566937e+00 -3.89818609e-01 6.82924569e-01
6.15308583e-01 -3.66805822e-01 8.24408889e-01 2.91011706e-02
-2.55051941e-01 8.11508298e-01 -1.62427768e-01 1.26620661e-02
8.98483157e-01 -1.32485640e+00 5.96340775e-01 6.54582202e-01
8.35425109e-02 5.05216300e-01 8.75791252e-01 -3.56407791e-01
-1.56067848e+00 -1.33457232e+00 8.33019018e-02 -2.62826949e-01
5.46152890e-01 -5.40141761e-01 -7.46766686e-01 5.86237550e-01
4.13768947e-01 6.74068630e-02 4.46045756e-01 -3.91330756e-02
-4.35540617e-01 -2.70426869e-01 -1.24442589e+00 1.25095212e+00
1.27725160e+00 -4.64858711e-01 -6.35469034e-02 3.13706219e-01
6.84710383e-01 -8.26855302e-01 -9.28065717e-01 2.18430504e-01
5.82560360e-01 -9.08196270e-01 1.13501918e+00 -6.80904746e-01
7.28371084e-01 -2.95004457e-01 1.24706700e-01 -1.72163808e+00
-3.75230342e-01 -1.22990072e+00 2.73308992e-01 1.24285769e+00
2.24300802e-01 -5.46651542e-01 6.94660127e-01 7.48065650e-01
-1.28544003e-01 -6.68166399e-01 -4.97777164e-01 -4.76768076e-01
2.48642847e-01 -1.58093810e-01 1.12923026e+00 6.20992482e-01
-6.65792763e-01 9.52563211e-02 -5.37552714e-01 -8.98040533e-02
4.30562347e-01 1.90559506e-01 1.21828306e+00 -9.22468066e-01
-4.67611849e-01 -7.53080845e-01 -1.70084462e-01 -9.09095824e-01
3.99679661e-01 -6.37108684e-01 -9.79622453e-02 -1.10404527e+00
-3.67094040e-01 -5.63516140e-01 1.03765897e-01 4.98486102e-01
-1.53831780e-01 5.15127420e-01 4.15538222e-01 7.24476203e-02
-1.09645547e-02 5.44048011e-01 1.91325665e+00 -2.59394556e-01
-3.60866904e-01 -1.05063029e-01 -8.36157680e-01 7.95286298e-01
7.97933042e-01 -6.96056038e-02 -3.35390300e-01 -5.23833156e-01
2.72660077e-01 5.91161959e-02 4.68125969e-01 -9.23357487e-01
-2.24755004e-01 -5.13076484e-01 6.68471456e-01 1.17428958e-01
4.01071638e-01 -6.62442744e-01 4.52490032e-01 1.62741765e-01
-6.21316433e-01 -1.38797536e-02 3.20475370e-01 3.72316718e-01
-6.63881823e-02 6.63147122e-02 1.01762676e+00 -2.33697563e-01
-5.29599428e-01 5.98136008e-01 -1.61727697e-01 3.16547230e-02
6.85906708e-01 -2.70301580e-01 -1.31465912e-01 -6.64245129e-01
-6.31678879e-01 -2.33080968e-01 9.61591482e-01 4.82156903e-01
5.96791625e-01 -1.50685728e+00 -6.01021051e-01 4.11156744e-01
-1.02156021e-01 -5.55762090e-02 -7.50633106e-02 2.70722538e-01
-7.48514593e-01 -2.87536025e-01 -6.26424491e-01 -4.60932761e-01
-1.05424690e+00 5.39253473e-01 5.03543139e-01 -1.41234174e-01
-4.34416711e-01 6.57493711e-01 5.30453265e-01 -3.28111142e-01
-1.41535893e-01 -3.83354604e-01 1.29502237e-01 -2.50744164e-01
3.26320767e-01 -7.24410638e-02 -5.01720496e-02 -3.27063054e-01
1.75738096e-01 9.32915926e-01 1.55115962e-01 -2.58243918e-01
1.21831942e+00 8.82962123e-02 -2.91133039e-02 -3.48222349e-03
1.00549233e+00 1.06851883e-01 -1.72656250e+00 -8.99568293e-03
-5.10522902e-01 -7.72605181e-01 7.13624880e-02 -8.36427331e-01
-1.63880098e+00 7.39433169e-01 5.42524040e-01 -5.02700657e-02
1.54809475e+00 -2.97154576e-01 6.99706972e-01 1.21456407e-01
3.84280235e-01 -9.87289548e-01 3.64879191e-01 1.40348405e-01
1.40056038e+00 -1.01862228e+00 -2.19068944e-01 -5.13855636e-01
-6.13076091e-01 1.41862702e+00 4.91909921e-01 -4.74247307e-01
5.82143724e-01 4.97660309e-01 2.84492761e-01 8.28761458e-02
-4.63146985e-01 9.32434425e-02 6.86487257e-01 7.34033465e-01
5.18219650e-01 6.81596920e-02 -9.09303203e-02 9.90457609e-02
-4.87345487e-01 1.97452344e-02 3.55727851e-01 6.12914622e-01
-5.05253933e-02 -1.52486002e+00 -4.03789341e-01 4.80831444e-01
-3.47528309e-01 -1.79885045e-01 -2.26767883e-01 7.15176344e-01
2.00583488e-01 5.43085515e-01 1.05934337e-01 -2.89128125e-01
4.28599536e-01 -1.59383386e-01 7.79326797e-01 -5.86401701e-01
-8.28836262e-01 2.91732550e-01 -1.03492029e-01 -7.49810457e-01
-4.37595427e-01 -5.60087919e-01 -7.49398947e-01 -3.60088527e-01
7.32357204e-02 -5.63684881e-01 4.80202228e-01 7.21609354e-01
4.05555606e-01 7.73140728e-01 5.79086602e-01 -1.23312354e+00
-2.15857342e-01 -5.72269261e-01 -4.59813863e-01 7.51053572e-01
4.26943824e-02 -6.31119251e-01 -1.14553288e-01 4.63287830e-01] | [11.662017822265625, -0.4645424485206604] |
4d9770ca-cb35-4c4f-b8b1-d7c00e26a742 | v-nas-neural-architecture-search-for | 1906.02817 | null | https://arxiv.org/abs/1906.02817v2 | https://arxiv.org/pdf/1906.02817v2.pdf | V-NAS: Neural Architecture Search for Volumetric Medical Image Segmentation | Deep learning algorithms, in particular 2D and 3D fully convolutional neural networks (FCNs), have rapidly become the mainstream methodology for volumetric medical image segmentation. However, 2D convolutions cannot fully leverage the rich spatial information along the third axis, while 3D convolutions suffer from the demanding computation and high GPU memory consumption. In this paper, we propose to automatically search the network architecture tailoring to volumetric medical image segmentation problem. Concretely, we formulate the structure learning as differentiable neural architecture search, and let the network itself choose between 2D, 3D or Pseudo-3D (P3D) convolutions at each layer. We evaluate our method on 3 public datasets, i.e., the NIH Pancreas dataset, the Lung and Pancreas dataset from the Medical Segmentation Decathlon (MSD) Challenge. Our method, named V-NAS, consistently outperforms other state-of-the-arts on the segmentation task of both normal organ (NIH Pancreas) and abnormal organs (MSD Lung tumors and MSD Pancreas tumors), which shows the power of chosen architecture. Moreover, the searched architecture on one dataset can be well generalized to other datasets, which demonstrates the robustness and practical use of our proposed method. | ['Dong Yang', 'Zhuotun Zhu', 'Daguang Xu', 'Alan Yuille', 'Chenxi Liu'] | 2019-06-06 | null | null | null | null | ['volumetric-medical-image-segmentation'] | ['medical'] | [-4.84508984e-02 2.35079765e-01 -3.69461536e-01 -3.41447890e-01
-5.81972182e-01 -6.38510644e-01 2.76556402e-01 -4.75305915e-02
-4.11006808e-01 3.45298141e-01 -1.42356288e-02 -7.72781193e-01
1.80091634e-02 -6.60376430e-01 -6.75110698e-01 -8.31763625e-01
-2.42235333e-01 6.48338437e-01 1.04081370e-01 2.60841638e-01
-1.86592743e-01 8.77421916e-01 -6.46632969e-01 -1.37408888e-02
9.32558000e-01 1.40978718e+00 9.05933231e-02 6.13331556e-01
-4.95421201e-01 5.45828938e-01 -1.55723989e-01 -1.38122037e-01
4.49063778e-01 -2.79450178e-01 -9.15906489e-01 1.17229156e-01
4.39755172e-01 -3.28121573e-01 -4.36192721e-01 1.04610157e+00
5.64286828e-01 -3.29396904e-01 7.88100541e-01 -1.03547800e+00
-4.92379993e-01 6.43904805e-01 -5.99982262e-01 4.51260418e-01
-4.20994401e-01 4.28540081e-01 6.63361847e-01 -7.70341992e-01
4.77981627e-01 1.06932437e+00 9.79289353e-01 6.79461122e-01
-1.02806163e+00 -4.49906617e-01 9.12148803e-02 -3.42504054e-01
-1.26563263e+00 5.81971966e-02 6.00241065e-01 -7.08687663e-01
7.25265324e-01 1.84231460e-01 9.32480037e-01 9.51815665e-01
1.48862019e-01 1.10028839e+00 8.52433860e-01 1.10143319e-01
1.15419909e-01 -2.72164077e-01 2.62125969e-01 1.03094161e+00
4.37651634e-01 1.25921905e-01 1.23203889e-01 -7.90804699e-02
1.23722041e+00 6.14067130e-02 -3.79741341e-01 -5.29191494e-01
-1.47755337e+00 8.19416344e-01 8.10346842e-01 2.02567995e-01
-6.43807232e-01 3.49338114e-01 5.57018757e-01 -2.59149307e-03
3.35692942e-01 2.69067556e-01 -6.46605432e-01 1.90459147e-01
-8.06788385e-01 6.53705001e-02 8.26321244e-01 9.26728606e-01
3.50997001e-01 -2.83166561e-02 -5.66597223e-01 5.90392351e-01
5.13773263e-01 2.24772036e-01 5.31194031e-01 -6.70943439e-01
3.93445700e-01 8.41177344e-01 -3.31827193e-01 -3.85431856e-01
-1.01495266e+00 -6.19676292e-01 -1.29202163e+00 2.19393820e-02
7.08428562e-01 -3.02912444e-01 -1.46726251e+00 1.42060483e+00
5.41831017e-01 1.90037891e-01 7.18652904e-02 1.20722592e+00
1.56120670e+00 3.31294686e-01 2.26432338e-01 1.35073930e-01
1.53579247e+00 -1.13785267e+00 -2.28828251e-01 2.63149329e-02
7.80869842e-01 -3.30075860e-01 8.22105467e-01 9.95294079e-02
-1.11323011e+00 -2.35526949e-01 -8.42599809e-01 -6.24946281e-02
-1.48009658e-01 5.13199866e-02 1.07243454e+00 6.39682353e-01
-1.13200653e+00 5.42495191e-01 -1.15869296e+00 -2.40292042e-01
1.10389698e+00 6.03542209e-01 -5.97988032e-02 7.18727149e-03
-9.17043209e-01 5.61848700e-01 5.12058854e-01 4.31134813e-02
-9.99990046e-01 -1.28890955e+00 -6.90972745e-01 1.04237184e-01
1.60019919e-01 -1.08344114e+00 1.27098095e+00 -8.54136050e-01
-1.44460189e+00 1.28903627e+00 4.21950489e-01 -6.55220628e-01
8.19130898e-01 2.33413547e-01 1.04338221e-01 2.70954043e-01
-2.45850220e-01 9.87700701e-01 4.20772076e-01 -9.57687080e-01
-4.82834756e-01 -4.13583159e-01 -1.51251301e-01 8.90005007e-02
3.17658745e-02 -3.67809683e-01 -7.41523564e-01 -6.81429446e-01
4.49189454e-01 -8.22283566e-01 -6.23020113e-01 4.63006258e-01
-8.23173165e-01 -1.40229493e-01 5.67846477e-01 -5.40451467e-01
7.15793192e-01 -2.04520273e+00 2.36507803e-01 4.19783503e-01
6.01358652e-01 1.38880163e-01 1.88370496e-02 -5.78574121e-01
3.53827514e-02 1.93196312e-01 -3.74623418e-01 -3.11264843e-01
1.77118871e-02 3.49787384e-01 2.00972185e-01 6.31371439e-01
1.48549080e-01 1.46411765e+00 -7.40091026e-01 -8.39634717e-01
9.99895632e-02 6.30962968e-01 -5.45965850e-01 8.06260929e-02
-3.06756169e-01 7.47390211e-01 -7.81520247e-01 9.76279974e-01
6.84272885e-01 -9.57626402e-01 -1.02519058e-01 -3.49509805e-01
2.97855325e-02 2.18862984e-02 -4.70021367e-01 1.81046450e+00
-8.95187929e-02 3.37835670e-01 1.96405903e-01 -8.78340006e-01
7.81012237e-01 2.51119852e-01 1.04141831e+00 -7.41372705e-01
3.38547468e-01 4.16738659e-01 2.48533878e-02 -5.48280120e-01
-7.27688894e-02 -3.16781178e-02 7.82825947e-02 1.04111150e-01
1.15415908e-01 -1.83363333e-02 9.58136693e-02 -1.47615179e-01
9.77334499e-01 -8.87427106e-02 1.60854638e-01 -5.75053036e-01
2.65835315e-01 3.17700237e-01 3.54771376e-01 7.61486351e-01
-4.21559393e-01 5.74410081e-01 8.28064322e-01 -6.79752409e-01
-9.42780137e-01 -1.03977573e+00 -4.78794128e-01 6.83045268e-01
3.14136058e-01 2.53047049e-01 -8.33683908e-01 -1.15500367e+00
2.47082964e-01 1.89465225e-01 -7.18625605e-01 1.38160706e-01
-9.05245483e-01 -1.10351765e+00 8.66569221e-01 5.78430474e-01
6.24122739e-01 -9.15940940e-01 -7.50547409e-01 2.42147088e-01
1.33554727e-01 -1.15187764e+00 -6.23128235e-01 4.43053722e-01
-1.32406795e+00 -1.21510899e+00 -1.19418907e+00 -1.06548798e+00
7.81616688e-01 -9.17733461e-02 1.40174222e+00 7.74297789e-02
-3.53465497e-01 3.00558388e-01 -2.35541463e-02 -1.71923682e-01
-4.64550108e-01 4.42599386e-01 -5.29554486e-01 -3.62557948e-01
1.45031750e-01 -4.78636712e-01 -9.71645236e-01 2.92644888e-01
-8.00411165e-01 3.12181115e-01 8.25306475e-01 8.68655920e-01
1.05002093e+00 -3.78436834e-01 3.34082901e-01 -9.61763144e-01
3.95575345e-01 -6.15314841e-01 -7.40084887e-01 2.17304960e-01
-5.27299106e-01 4.41570347e-03 3.26284200e-01 -4.16409612e-01
-5.16120315e-01 3.67959738e-01 -3.37373912e-01 -7.26433218e-01
-2.61157393e-01 6.30025506e-01 1.42898589e-01 -3.23640108e-01
5.30760944e-01 2.13275477e-01 2.44654119e-01 -6.00091875e-01
1.77976698e-01 4.15234528e-02 5.97373247e-01 -4.40725535e-01
5.52141011e-01 4.12555963e-01 3.58092844e-01 -3.61800164e-01
-7.10067749e-01 -2.81751513e-01 -6.29530370e-01 -5.89868501e-02
1.23008156e+00 -8.31790090e-01 -7.26655662e-01 3.98199171e-01
-1.06324625e+00 -7.25971401e-01 -3.31330895e-01 4.07492101e-01
-3.92185718e-01 2.34889835e-01 -9.24647570e-01 -1.64653420e-01
-8.99826884e-01 -1.68108106e+00 1.12214053e+00 2.99169660e-01
5.81977591e-02 -1.22515059e+00 -2.27135479e-01 -1.92792844e-02
6.60997033e-01 7.68967986e-01 1.34956646e+00 -8.84268880e-01
-8.14713776e-01 -7.66018331e-02 -6.24045908e-01 1.22426048e-01
-8.49459544e-02 -2.19722360e-01 -5.88241100e-01 -2.15642437e-01
-2.94393331e-01 -2.65438646e-01 1.00692749e+00 9.64492440e-01
1.69056809e+00 -2.48903975e-01 -4.94390965e-01 1.29005098e+00
1.28717947e+00 1.15985878e-01 2.95094192e-01 1.93498150e-01
8.65715206e-01 5.07933013e-02 5.93976788e-02 3.24672729e-01
3.80648434e-01 3.31482738e-01 8.80370736e-01 -8.44633281e-01
-2.85991132e-01 9.74558890e-02 -2.54740655e-01 6.12859368e-01
9.59670842e-02 -1.88886687e-01 -1.10265863e+00 3.65617573e-01
-1.45471370e+00 -1.92135841e-01 -1.43876255e-01 1.69588637e+00
7.52583742e-01 2.79885717e-02 1.48982480e-01 -3.34214628e-01
5.47273934e-01 -3.60136367e-02 -9.06461537e-01 -9.13117267e-03
-3.10054682e-02 2.50738472e-01 8.88502121e-01 -1.43802324e-02
-1.46595860e+00 5.48015893e-01 6.17955399e+00 7.01896071e-01
-1.40379965e+00 6.43639490e-02 1.17818666e+00 -5.63049363e-03
-2.53088623e-01 -5.32042563e-01 -7.10669518e-01 3.47373575e-01
4.31156069e-01 2.97114104e-01 2.00673088e-01 8.99208844e-01
-8.84089544e-02 3.63432854e-01 -1.30889201e+00 1.02666390e+00
-4.20244157e-01 -1.63582182e+00 7.60981739e-02 2.71111518e-01
7.73549080e-01 5.43740571e-01 2.42392331e-01 4.35046166e-01
1.92244545e-01 -1.41107297e+00 4.67390746e-01 1.72330379e-01
1.00354266e+00 -5.08955598e-01 6.76301420e-01 2.00560734e-01
-9.79627848e-01 2.17697456e-01 -1.17442906e-01 8.15181494e-01
6.03797054e-03 5.60956299e-01 -9.86115515e-01 2.73649782e-01
6.74594402e-01 5.12187719e-01 -4.42852527e-01 1.38418114e+00
4.39184792e-02 6.49948716e-01 -5.14080465e-01 -1.60162330e-01
7.86634982e-01 -1.39870003e-01 4.95090455e-01 1.36111104e+00
4.29244250e-01 1.04521938e-01 1.61376715e-01 1.19297743e+00
-5.24278879e-01 9.83022526e-02 -1.50208429e-01 -3.51797268e-02
-2.32160906e-03 1.28268278e+00 -1.03934526e+00 -2.88793921e-01
-2.77237386e-01 6.90359354e-01 2.55385898e-02 3.43551844e-01
-1.05230188e+00 2.84425747e-02 5.93546808e-01 -1.38619140e-01
4.65331852e-01 -1.84946752e-04 -5.95399559e-01 -9.34589684e-01
-2.64679074e-01 -5.84012806e-01 6.55556798e-01 -2.33905137e-01
-1.46955061e+00 6.12874925e-01 -2.56761938e-01 -1.05925548e+00
2.37929180e-01 -9.37334776e-01 -5.99156857e-01 7.12112844e-01
-1.75983250e+00 -1.07538211e+00 -5.74462712e-01 5.86081743e-01
3.22469324e-01 -4.63687740e-02 4.51415956e-01 5.17565489e-01
-5.14218152e-01 7.42228329e-01 4.21118457e-03 5.39294899e-01
2.02185959e-01 -1.26285279e+00 3.35830569e-01 2.74821401e-01
-2.88759649e-01 1.38424680e-01 7.46580660e-02 -4.22686577e-01
-1.72672653e+00 -1.47696936e+00 3.13079834e-01 8.06919113e-03
3.46657783e-01 -1.18275911e-01 -8.17549109e-01 6.55596972e-01
-2.99492776e-02 5.08799255e-01 5.38640082e-01 -2.30850846e-01
-4.67093587e-02 2.51574427e-01 -1.54112673e+00 5.87096989e-01
1.07474697e+00 2.46493861e-01 -1.77744299e-01 6.66714489e-01
1.13847268e+00 -1.15472686e+00 -1.24562442e+00 7.87536144e-01
3.26896846e-01 -6.44723773e-01 1.31165338e+00 -6.45036817e-01
6.06701612e-01 9.01016407e-03 -1.18025951e-02 -9.55037653e-01
-2.22847149e-01 -3.72304767e-01 -3.75251502e-01 5.93040943e-01
6.01979792e-01 -6.41923785e-01 1.13011098e+00 5.18767178e-01
-5.84132373e-01 -1.40053082e+00 -1.21754003e+00 -4.55239058e-01
5.14671504e-01 -2.09650815e-01 7.50197589e-01 9.47128117e-01
-5.62511027e-01 -5.38424514e-02 2.06699595e-01 2.04911992e-01
7.68193841e-01 2.52567440e-01 4.84248936e-01 -1.12448108e+00
-3.41443658e-01 -1.18829882e+00 -2.83648789e-01 -1.37863922e+00
-1.08824179e-01 -1.40553153e+00 -1.83850914e-01 -1.59696651e+00
2.79720962e-01 -7.34841466e-01 -4.40767318e-01 5.80136418e-01
-4.17122357e-02 1.35050058e-01 -3.76440845e-02 2.11063102e-01
-4.49972719e-01 3.26636434e-01 1.99522603e+00 -3.72000039e-01
-3.25890601e-01 2.13608444e-01 -6.57259881e-01 6.47146583e-01
7.58499384e-01 -2.87597358e-01 -7.57203326e-02 -5.21523356e-01
-1.41074181e-01 3.84005904e-01 5.93715906e-01 -7.28062868e-01
1.77866712e-01 -9.32645425e-02 7.23158598e-01 -8.93562019e-01
1.83365382e-02 -8.20790708e-01 4.28452808e-03 8.14178765e-01
-3.63795638e-01 -1.67611301e-01 4.14159298e-01 2.47795194e-01
-2.18389072e-02 -1.47948429e-01 1.13002813e+00 -4.22199696e-01
-4.51781571e-01 1.11991739e+00 -1.47907138e-01 2.99169451e-01
8.31731796e-01 -2.69542843e-01 -1.72355503e-01 2.01365516e-01
-6.94411695e-01 5.24619102e-01 1.95853859e-01 6.08422458e-02
6.43163502e-01 -1.32539463e+00 -7.17024326e-01 1.65345192e-01
-1.08585022e-01 7.81177700e-01 2.89294422e-01 1.45659542e+00
-9.96998250e-01 4.89260346e-01 5.52866347e-02 -1.27825344e+00
-8.41448784e-01 4.25605178e-01 8.68442535e-01 -5.63783348e-01
-1.14307940e+00 1.03920197e+00 4.90151167e-01 -5.72784662e-01
5.62925994e-01 -8.79259944e-01 -4.77522723e-02 -3.49474877e-01
9.23027769e-02 -1.59615539e-02 5.49959019e-02 -2.52164215e-01
-3.56976241e-01 4.14990723e-01 4.11502011e-02 5.37052512e-01
1.33540058e+00 2.42705986e-01 -7.40008950e-02 -1.89779446e-01
1.41753721e+00 -7.08788335e-01 -1.47355294e+00 -3.47659260e-01
6.95926994e-02 -1.24420166e-01 3.15332562e-01 -8.88676822e-01
-1.85117936e+00 7.88132310e-01 7.62263536e-01 7.53040239e-02
1.08779705e+00 3.33639055e-01 1.04947054e+00 2.65006185e-01
4.51136008e-03 -5.05024731e-01 -6.46407604e-02 5.38741887e-01
6.47581160e-01 -1.23402393e+00 -1.61679924e-01 -3.87442529e-01
-4.94399667e-01 1.29997873e+00 6.84045494e-01 -1.17434561e-01
7.38712788e-01 4.19874012e-01 8.58675539e-02 -5.30950367e-01
-3.71533513e-01 -3.22224014e-02 4.44741607e-01 3.29013318e-01
3.70477021e-01 1.42121896e-01 1.53959587e-01 6.74913287e-01
6.84206262e-02 -4.33099195e-02 4.35798280e-02 7.74715543e-01
-2.58084744e-01 -5.40808618e-01 -2.56102443e-01 7.29313076e-01
-5.41107059e-01 -1.49132386e-01 -2.10855380e-01 1.09460771e+00
4.79207486e-02 8.97005722e-02 8.32003728e-02 1.78044707e-01
3.34197253e-01 -2.74970997e-02 5.86250424e-01 -1.99504927e-01
-8.92006040e-01 2.79121786e-01 -2.86805242e-01 -5.66704154e-01
-1.52007073e-01 -4.64289784e-01 -1.57693732e+00 -2.34524682e-01
-3.45746502e-02 -2.69431204e-01 7.99074888e-01 6.94364309e-01
2.55412698e-01 8.32330167e-01 3.56127441e-01 -9.25494790e-01
-7.84199953e-01 -6.99782908e-01 -4.47243243e-01 2.17904598e-01
4.07977432e-01 -4.73576874e-01 -1.49516284e-01 -2.31590524e-01] | [14.579432487487793, -2.6307849884033203] |
aa7dadf2-9e97-4d99-88f1-85193ef31f17 | universal-dependency-treebank-for-odia | 2205.11976 | null | https://arxiv.org/abs/2205.11976v1 | https://arxiv.org/pdf/2205.11976v1.pdf | Universal Dependency Treebank for Odia Language | This paper presents the first publicly available treebank of Odia, a morphologically rich low resource Indian language. The treebank contains approx. 1082 tokens (100 sentences) in Odia selected from "Samantar", the largest available parallel corpora collection for Indic languages. All the selected sentences are manually annotated following the ``Universal Dependency (UD)" guidelines. The morphological analysis of the Odia treebank was performed using machine learning techniques. The Odia annotated treebank will enrich the Odia language resource and will help in building language technology tools for cross-lingual learning and typological research. We also build a preliminary Odia parser using a machine learning approach. The accuracy of the parser is 86.6% Tokenization, 64.1% UPOS, 63.78% XPOS, 42.04% UAS and 21.34% LAS. Finally, the paper briefly discusses the linguistic analysis of the Odia UD treebank. | ['Bijayalaxmi Dash', 'Satya Ranjan Dash', 'Saraswati Sahoo', 'Atul Kr. Ojha', 'Kalyanamalini Sahoo', 'Shantipriya Parida'] | 2022-05-24 | null | https://aclanthology.org/2022.wildre-1.15 | https://aclanthology.org/2022.wildre-1.15.pdf | wildre-lrec-2022-6 | ['morphological-analysis'] | ['natural-language-processing'] | [-6.68536186e-01 1.87306553e-01 -3.01579177e-01 -1.46889463e-01
-8.74083519e-01 -6.85532808e-01 1.84318379e-01 5.73576987e-01
-6.44158483e-01 1.07243919e+00 2.66382158e-01 -4.68878746e-01
3.74169886e-01 -8.25469255e-01 -2.23735482e-01 -5.50232708e-01
-5.03834616e-03 7.62648523e-01 9.50898230e-02 -2.94264615e-01
4.47127581e-01 6.31738782e-01 -9.66951013e-01 1.85808912e-01
9.57614839e-01 1.44320950e-01 7.70960867e-01 6.39205933e-01
-4.21895444e-01 8.52391362e-01 -9.04367685e-01 -5.28159618e-01
7.75494501e-02 -1.81682974e-01 -1.49544883e+00 -3.11867625e-01
9.28342491e-02 -1.41340062e-01 1.47629857e-01 1.19821572e+00
3.94142419e-01 -1.08332448e-01 5.14992237e-01 -4.25945133e-01
-7.70769596e-01 1.46648824e+00 -6.10916078e-01 5.04185796e-01
7.63291121e-02 -4.56734955e-01 1.32299078e+00 -9.01839554e-01
1.06320500e+00 1.57897258e+00 3.26263040e-01 3.27622414e-01
-6.45011425e-01 -8.03105593e-01 -2.79793561e-01 9.24844444e-02
-1.19579184e+00 -2.75227785e-01 5.58226526e-01 -4.46444005e-01
1.36236632e+00 -1.85563996e-01 4.02119786e-01 3.56338143e-01
3.63449633e-01 6.15869105e-01 1.67007601e+00 -1.07162821e+00
-2.41539683e-02 2.49848083e-01 4.89124566e-01 7.00767100e-01
6.55643046e-01 -6.49223745e-01 -2.30842113e-01 6.90862909e-02
5.88952184e-01 -9.38543260e-01 5.56827903e-01 6.31340981e-01
-6.01944745e-01 9.97250199e-01 -3.88009310e-01 9.97677982e-01
-2.13808790e-01 -3.62713069e-01 1.05077493e+00 3.03195328e-01
5.18356204e-01 3.25206161e-01 -8.65336418e-01 -3.65324348e-01
-2.75460750e-01 2.63339076e-02 8.37693989e-01 1.01317084e+00
6.53435051e-01 4.37432110e-01 7.22101510e-01 1.56541193e+00
3.44179630e-01 7.08141029e-01 5.66955984e-01 -6.64251268e-01
6.32159531e-01 5.99149406e-01 -2.44428694e-01 -4.88460898e-01
-2.74639040e-01 1.96288556e-01 -2.03640580e-01 -1.52182013e-01
7.11121142e-01 -5.75408876e-01 -8.01202536e-01 1.31471729e+00
1.45289332e-01 -8.91611576e-01 5.61220288e-01 1.88652575e-01
9.66992497e-01 8.46588671e-01 7.76071906e-01 -2.96725631e-01
1.65488708e+00 -5.55468559e-01 -8.14161956e-01 9.81971920e-02
9.77810740e-01 -1.33749354e+00 1.05627704e+00 4.39470977e-01
-1.36509252e+00 -3.64373833e-01 -8.06948245e-01 -1.63329855e-01
-5.15985012e-01 5.20911098e-01 6.49181962e-01 1.06826663e+00
-6.22086942e-01 1.90984413e-01 -7.99102545e-01 -5.69794357e-01
9.36178640e-02 3.27851236e-01 -6.54176056e-01 7.92742968e-02
-1.13222349e+00 1.23797071e+00 9.30621147e-01 -3.21929336e-01
-2.30051205e-01 -2.60389656e-01 -1.04391921e+00 -4.29979920e-01
-1.24671206e-01 5.19391537e-01 1.15199089e+00 -8.39130819e-01
-1.51501548e+00 1.58802009e+00 -9.64059532e-02 -4.01720017e-01
-8.05196464e-02 -1.92558125e-01 -7.39110708e-01 4.20865566e-01
5.70784569e-01 1.68111309e-01 -1.78732917e-01 -7.97887444e-01
-1.03454101e+00 -5.57417214e-01 -1.90820798e-01 -6.68762252e-02
9.32255611e-02 1.06354678e+00 -1.12833232e-01 -8.75398636e-01
6.16241470e-02 -8.00205529e-01 -8.14971104e-02 -1.18735933e+00
-1.59118474e-01 -7.01732993e-01 5.96697450e-01 -1.39773107e+00
1.24243701e+00 -1.68943572e+00 -5.74032903e-01 1.85486358e-02
-4.75979924e-01 5.33607721e-01 1.70790777e-01 7.56594479e-01
-1.00445129e-01 2.62076646e-01 1.55140422e-02 1.02378003e-01
-1.45513669e-01 7.43935108e-01 1.30344793e-01 4.63090092e-01
2.83624709e-01 4.09717530e-01 -8.48419726e-01 -8.69399965e-01
1.56285137e-01 -1.00072622e-02 -2.79371083e-01 -9.78150442e-02
2.96134800e-01 3.65165174e-01 -2.15860188e-01 9.79222000e-01
8.10198367e-01 9.28558290e-01 6.11626923e-01 1.39748186e-01
-9.66356874e-01 8.99886310e-01 -1.03990483e+00 1.60308182e+00
-8.02183568e-01 5.17274618e-01 -6.36188686e-02 -9.91396546e-01
1.44585633e+00 2.71144897e-01 2.13907436e-01 -5.13286412e-01
3.68982136e-01 7.51629829e-01 5.28686464e-01 -7.75519609e-01
9.23377216e-01 -1.71479851e-01 -7.18135238e-01 3.00578892e-01
5.67861736e-01 -8.33176374e-02 7.49972165e-01 1.65104896e-01
5.11885107e-01 1.27131999e-01 9.76695359e-01 -1.00530362e+00
7.95010567e-01 3.95552099e-01 9.64066327e-01 1.51536405e-01
-2.90703297e-01 -3.16861719e-02 5.70190966e-01 -1.82025343e-01
-1.25148189e+00 -7.22292066e-01 -8.05234551e-01 1.11088896e+00
-5.24777114e-01 -2.78451890e-01 -7.48113155e-01 -5.23217678e-01
-4.45597768e-01 6.99199617e-01 -2.23252639e-01 6.19216859e-01
-1.09705007e+00 -7.59041786e-01 9.79052961e-01 5.43600798e-01
5.75378239e-01 -1.50062549e+00 -2.14672804e-01 5.88154614e-01
-1.02336507e-03 -1.18473506e+00 2.26787403e-01 2.48180896e-01
-6.19120598e-01 -9.13646579e-01 -1.36467174e-01 -1.49943900e+00
2.17340365e-01 -4.01444376e-01 9.82704401e-01 -2.75781721e-01
-2.26923779e-01 -3.15128922e-01 -5.73300362e-01 -7.42955446e-01
-1.01593232e+00 2.14372769e-01 1.14386506e-01 -9.49118674e-01
8.51413369e-01 -3.22949529e-01 3.16044003e-01 -4.18247104e-01
-4.00438458e-01 -3.93093228e-01 3.81958365e-01 7.64828920e-01
4.46293116e-01 1.10237494e-01 1.13118577e+00 -1.37392557e+00
2.47293338e-01 -5.14535129e-01 -5.89675486e-01 -6.96952194e-02
-2.07631633e-01 -1.30229190e-01 7.43322790e-01 8.55845138e-02
-1.53568423e+00 -1.32579640e-01 -1.07380760e+00 7.07876444e-01
-5.59704065e-01 6.62978828e-01 -5.79191327e-01 1.56424120e-01
2.62857139e-01 -2.01189369e-01 -4.87823814e-01 -8.27229738e-01
1.06015943e-01 1.05701792e+00 7.81444132e-01 -1.18866360e+00
2.93643534e-01 9.92108211e-02 -2.92539299e-01 -1.24706113e+00
-4.28171337e-01 -7.35384107e-01 -1.38055670e+00 7.62956310e-03
8.19162130e-01 -8.92392993e-01 -4.19279128e-01 6.33717358e-01
-1.05700481e+00 -3.60414773e-01 -2.37297297e-01 5.89838207e-01
-2.51929581e-01 4.49961513e-01 -1.23778832e+00 -8.83349776e-01
-7.12856948e-01 -7.72160172e-01 3.61132443e-01 3.41426015e-01
-3.58902335e-01 -1.31201231e+00 3.39412808e-01 1.01306319e-01
-3.91370684e-01 2.14702412e-01 1.21724856e+00 -9.12706852e-01
5.50229549e-01 1.03842005e-01 -1.79995775e-01 5.23348391e-01
3.74480456e-01 5.44826865e-01 -7.29189634e-01 1.46281511e-01
-2.87314087e-01 -3.56131643e-01 2.48400047e-01 2.00352430e-01
2.23082155e-01 -2.66620308e-01 2.78885216e-01 -3.18122730e-02
1.72332013e+00 7.51229823e-01 4.51131761e-01 8.62957358e-01
4.09017891e-01 8.43046784e-01 9.80825067e-01 4.67659354e-01
6.63657606e-01 1.91769451e-02 -2.79716581e-01 4.00583327e-01
-3.20477962e-01 -2.13246450e-01 6.87388599e-01 1.65026641e+00
-5.71465977e-02 4.61108774e-01 -1.51290011e+00 1.28232658e+00
-1.21034384e+00 -6.63843691e-01 -6.42065048e-01 1.79347336e+00
1.26837635e+00 -4.86662574e-02 3.63841057e-01 4.38724428e-01
7.44118929e-01 -2.35686585e-01 3.22416574e-01 -1.53158474e+00
-3.12187523e-01 1.09984863e+00 7.17970848e-01 6.07891619e-01
-1.03442740e+00 1.70525587e+00 5.70854712e+00 6.11976385e-01
-9.64307904e-01 2.54688025e-01 2.35800352e-02 6.96050107e-01
2.15894833e-01 1.04095049e-01 -1.29686487e+00 2.61706680e-01
1.49041462e+00 -3.50762993e-01 -2.30232671e-01 8.50613415e-01
3.59191626e-01 -5.03965974e-01 -1.62745014e-01 4.81940776e-01
-1.99951336e-01 -1.28834796e+00 -9.90238413e-02 -5.28174527e-02
5.82660258e-01 4.53445762e-01 -3.72382969e-01 4.46610272e-01
9.60899830e-01 -5.23736238e-01 9.31760252e-01 -4.08036590e-01
9.80362236e-01 -1.28589714e+00 9.25277710e-01 2.09391519e-01
-1.34721756e+00 1.42458647e-01 -9.39605057e-01 -2.78816700e-01
1.04123950e-01 2.07330346e-01 -9.25987303e-01 7.16487527e-01
8.42421830e-01 5.70222020e-01 -5.34856677e-01 6.57801211e-01
-5.95677972e-01 1.26390553e+00 -2.34478593e-01 2.21939292e-02
5.49444318e-01 -5.16629338e-01 4.94905204e-01 1.98786938e+00
-5.46622872e-02 1.14841990e-01 1.43297866e-01 -2.56646099e-03
3.70326564e-02 1.09741950e+00 -4.69425410e-01 1.99075602e-02
7.71930337e-01 1.07195258e+00 -8.37624490e-01 -4.75747377e-01
-5.22411406e-01 4.56330538e-01 4.89112943e-01 -3.19698155e-01
-1.64754063e-01 -8.59655261e-01 4.89121079e-01 -2.88753003e-01
4.10562605e-02 -4.62156087e-01 -5.89992225e-01 -1.00201452e+00
-3.38269591e-01 -7.85661519e-01 8.89684916e-01 -2.39666060e-01
-1.18256485e+00 7.20964730e-01 2.06667632e-01 -6.65980995e-01
-3.92611951e-01 -8.32254291e-01 -5.58857143e-01 1.19163251e+00
-1.31350076e+00 -1.46896362e+00 5.40187418e-01 2.73538381e-01
8.64050090e-01 -3.94775540e-01 1.22210097e+00 3.37560594e-01
-9.09945905e-01 5.04639924e-01 5.93344793e-02 7.09998906e-01
6.15075052e-01 -1.51884007e+00 3.77542168e-01 1.06615758e+00
-3.10263604e-01 5.77067018e-01 4.97141480e-01 -8.14649761e-01
-9.69134331e-01 -1.01694071e+00 1.69921148e+00 5.11191562e-02
1.33951437e+00 -1.09896965e-01 -9.35880363e-01 8.82192850e-01
6.81840897e-01 -3.65004063e-01 9.93744195e-01 -5.60227141e-04
-2.60507733e-01 2.12490872e-01 -1.49329555e+00 4.78953838e-01
3.50091130e-01 -4.74065214e-01 -9.98054206e-01 1.33979917e-01
2.72949755e-01 -1.98458880e-01 -1.66856670e+00 -2.98961103e-01
5.11593461e-01 -3.20850402e-01 3.62994790e-01 -5.24034202e-01
2.13204682e-01 -2.83418316e-02 -1.33071363e-01 -1.17283988e+00
-2.11840212e-01 -4.99924481e-01 8.25029194e-01 2.01816225e+00
4.56396818e-01 -7.25007534e-01 4.10964072e-01 2.92237282e-01
-4.73028660e-01 2.18164432e-03 -8.88592601e-01 -8.13395381e-01
9.67615664e-01 -6.49873137e-01 3.30925465e-01 1.23667574e+00
5.97866595e-01 3.85868430e-01 -3.63684036e-02 1.05086481e-02
5.20053267e-01 -1.61380604e-01 3.26101035e-01 -1.18709826e+00
1.74135771e-02 -1.04129612e-01 -2.46450841e-01 -2.56561399e-01
5.13640583e-01 -9.88991201e-01 -2.85160661e-01 -1.41757238e+00
-9.99009013e-02 -9.01699901e-01 2.80237615e-01 7.03190446e-01
1.11771829e-01 1.53859228e-01 9.30338576e-02 1.29751191e-01
3.35288793e-01 -1.58292264e-01 9.58640635e-01 2.69885480e-01
-3.94541860e-01 -4.30493712e-01 -5.37316263e-01 1.07668340e+00
1.47480822e+00 -7.07291603e-01 1.99247986e-01 -4.21405911e-01
-1.70682341e-01 -2.05414414e-01 -6.88454211e-01 -4.03665096e-01
-3.04880649e-01 -4.72620964e-01 -1.29018232e-01 -8.04176152e-01
-2.15756088e-01 -2.96293199e-01 -2.63124704e-01 7.03587949e-01
1.46471396e-01 4.78472650e-01 5.89052200e-01 -4.64266509e-01
-2.78685868e-01 -8.97633553e-01 1.23892641e+00 -4.02203500e-01
-1.16156435e+00 -4.00351807e-02 -1.01531792e+00 2.32357994e-01
1.00577891e+00 -4.08234671e-02 -2.39873588e-01 4.93963569e-01
-5.85097253e-01 -2.46547744e-01 3.66343260e-01 1.44230247e-01
4.93868850e-02 -9.69361782e-01 -8.72407258e-01 -4.45628986e-02
3.20197418e-02 -3.54059041e-01 4.99362014e-02 3.53225380e-01
-1.24186647e+00 5.31613946e-01 -7.98977256e-01 1.51517689e-01
-1.32301712e+00 1.33579493e-01 -2.22313866e-01 -4.08191413e-01
-7.73149371e-01 4.37057644e-01 -4.06426191e-01 -3.57730120e-01
-3.41285348e-01 -2.40903482e-01 -9.73836064e-01 1.78615138e-01
3.65436673e-01 5.93961716e-01 9.41356793e-02 -1.45709980e+00
-4.59243029e-01 2.50883549e-01 -1.45381868e-01 -4.81344163e-01
1.32397461e+00 -2.60393113e-01 -7.89118648e-01 7.13349164e-01
9.26879704e-01 8.56468678e-01 -1.91650927e-01 5.73253743e-02
5.77260852e-01 -2.16888309e-01 -3.04398566e-01 -6.34683490e-01
-4.46687907e-01 7.09228516e-01 -5.38234739e-03 3.28921489e-02
8.59852433e-01 -5.71032055e-02 8.25610280e-01 1.60812333e-01
3.55128020e-01 -1.56240153e+00 -8.24216545e-01 1.20970249e+00
5.76779187e-01 -7.30147660e-01 -1.26036733e-01 -5.39405525e-01
-8.26739967e-01 1.26403439e+00 4.88992959e-01 -5.43586195e-01
6.95636690e-01 6.92210376e-01 4.00193036e-01 5.55632152e-02
-5.21486402e-01 -4.12946016e-01 -3.29719037e-01 9.00518358e-01
1.06883633e+00 4.49024171e-01 -1.49305272e+00 8.33316147e-01
-9.19428170e-01 -5.87283432e-01 1.03497314e+00 9.42851603e-01
-6.17977798e-01 -1.69163251e+00 -4.35714334e-01 8.02156776e-02
-1.45044684e+00 -3.41671497e-01 -2.18963698e-01 1.51252460e+00
1.86936021e-01 1.00928879e+00 2.27072954e-01 1.19251408e-01
1.12714760e-01 1.35547310e-01 5.01590848e-01 -1.02735126e+00
-9.24076498e-01 3.76683503e-01 8.53831470e-01 4.59827870e-01
-5.27452111e-01 -1.02755094e+00 -1.81011510e+00 -5.43651283e-01
-1.60681739e-01 7.63339341e-01 7.77995169e-01 9.26432371e-01
-5.75987637e-01 -8.92421082e-02 3.16763133e-01 -3.16976249e-01
4.95589264e-02 -1.22157180e+00 -9.47715521e-01 -5.56642115e-02
-5.71819425e-01 -3.58172506e-01 2.59516835e-01 3.03172469e-01] | [10.364304542541504, 10.113279342651367] |
51e96b02-a4c3-4ac3-b6af-66a08ac1503d | kafk-at-semeval-2020-task-8-extracting | null | null | https://aclanthology.org/2020.semeval-1.152 | https://aclanthology.org/2020.semeval-1.152.pdf | KAFK at SemEval-2020 Task 8: Extracting Features from Pre-trained Neural Networks to Classify Internet Memes | This paper presents two approaches for the internet meme classification challenge of SemEval-2020 Task 8 by Team KAFK (cosec). The first approach uses both text and image features, while the second approach uses only the images. Error analysis of the two approaches shows that using only the images is more robust to the noise in the text on the memes. We utilize pre-trained DistilBERT and EfficientNet to extract features from the text and image of the memes respectively. Our classification systems obtained macro f1 score of 0.3286 for Task A and 0.5005 for Task B. | ['Kuntal Dey', 'Ferdous Ahmed Barbhuiya', 'Arup Baruah', 'Kaushik Amar Das'] | 2020-12-01 | null | null | null | semeval-2020 | ['meme-classification'] | ['natural-language-processing'] | [-2.93903649e-01 -2.95469344e-01 2.63149470e-01 -7.12551475e-02
-4.95451808e-01 -6.83606923e-01 1.07165909e+00 1.12893984e-01
-9.73106325e-01 8.42712045e-01 -3.67819145e-03 -9.77820605e-02
3.00635099e-01 -6.61135256e-01 -7.71268308e-01 -4.01754618e-01
1.04861744e-01 2.43124679e-01 4.66308624e-01 -1.49501711e-01
6.98676527e-01 2.39180066e-02 -1.34592938e+00 9.90285993e-01
3.23330015e-01 1.46723640e+00 3.09724331e-01 1.12305093e+00
-4.83279020e-01 1.22118080e+00 -8.25676262e-01 -3.40066075e-01
4.17158335e-01 -3.11346382e-01 -1.09595037e+00 2.23829448e-02
1.03241158e+00 -4.36487585e-01 -7.40411460e-01 9.67347503e-01
4.70898986e-01 2.62205780e-01 8.87724400e-01 -9.56340253e-01
-4.11261678e-01 3.45472217e-01 -6.31984413e-01 8.38137388e-01
2.71360338e-01 -6.87563568e-02 7.57143140e-01 -1.31365657e+00
9.18943048e-01 1.09351611e+00 6.99937820e-01 4.30857658e-01
-1.00987780e+00 -7.73425579e-01 -4.05738294e-01 3.03714395e-01
-1.32313001e+00 -6.04602456e-01 1.35566086e-01 -7.11186826e-01
1.13533592e+00 -3.86386812e-02 3.04769695e-01 1.34059680e+00
7.08937705e-01 8.91968608e-01 1.68606472e+00 -3.87122482e-01
1.56864122e-01 7.27181852e-01 1.82314023e-01 8.93609762e-01
-1.18714057e-01 -1.79970324e-01 -7.98308730e-01 -1.11877009e-01
4.59064394e-01 -1.65476978e-01 1.33334130e-01 4.70874041e-01
-1.19192743e+00 7.37711132e-01 4.15680349e-01 5.37403822e-01
-4.20557946e-01 1.92375228e-01 7.21851110e-01 7.53331065e-01
1.01931357e+00 3.73988330e-01 -1.47916123e-01 -8.69670138e-02
-1.47299695e+00 7.79048279e-02 8.60442877e-01 7.45985031e-01
6.26797795e-01 -1.86443701e-02 -4.79493476e-02 9.69322026e-01
7.60934129e-02 5.44206619e-01 7.79099524e-01 -5.53615749e-01
7.74374247e-01 3.76985580e-01 1.27354711e-01 -1.31093907e+00
-3.81396025e-01 -2.43593439e-01 -6.32691264e-01 -1.93467699e-02
2.63068795e-01 -1.62879482e-01 -1.29258788e+00 1.13351202e+00
-1.60915956e-01 -3.96378897e-02 -8.14520195e-02 5.94981551e-01
1.24932086e+00 1.02813292e+00 -4.53451537e-02 2.45592356e-01
1.09720182e+00 -1.38467741e+00 -6.45211995e-01 -2.87708700e-01
6.24834895e-01 -9.80270147e-01 7.02641785e-01 2.17483565e-01
-1.12955427e+00 -4.89471167e-01 -1.35420358e+00 1.59924462e-01
-9.57831800e-01 2.89944351e-01 1.57696426e-01 4.74042088e-01
-1.57507586e+00 5.47398865e-01 -4.02915061e-01 -8.18248689e-01
3.77359778e-01 3.57207179e-01 -6.01233304e-01 -1.48300603e-02
-1.03834403e+00 9.59510624e-01 4.91869539e-01 -3.55036378e-01
-1.10467696e+00 -1.11492686e-01 -3.88938069e-01 -2.08672464e-01
-1.32969782e-01 -1.97038636e-01 8.85569394e-01 -1.22197890e+00
-1.15600896e+00 1.37800443e+00 1.44580230e-01 -7.87609696e-01
7.60528386e-01 -1.38297811e-01 -4.52728450e-01 6.16101146e-01
1.94184497e-01 9.86555338e-01 1.01479435e+00 -1.02202809e+00
-1.06126297e+00 -1.45071939e-01 -7.88437426e-02 2.51033574e-01
-8.53278637e-01 7.38211870e-02 -4.37936693e-01 -6.62415743e-01
-7.13706911e-02 -8.53939652e-01 3.49930227e-01 -1.29854217e-01
-3.28295857e-01 -1.14387788e-01 1.06943357e+00 -1.10908329e+00
9.29565310e-01 -2.04549193e+00 -2.39557236e-01 1.32453769e-01
5.14451146e-01 2.41431966e-01 -5.89757025e-01 4.45542634e-01
1.14304982e-01 2.87447155e-01 3.03627223e-01 -4.31186259e-01
-3.94330293e-01 -2.66349822e-01 -3.54927406e-02 5.59249878e-01
-2.13594362e-01 6.65099978e-01 -6.17571771e-01 -6.12467945e-01
-1.01040922e-01 1.70886874e-01 -1.19808510e-01 2.77955055e-01
-1.49636656e-01 2.11211681e-01 -3.14919531e-01 2.69196272e-01
6.72775626e-01 -3.40186596e-01 -5.70047796e-02 -9.84288752e-02
-1.14896558e-01 4.31240112e-01 -7.06862867e-01 1.35132098e+00
-4.04456407e-01 1.25201619e+00 2.03380138e-01 -9.64587331e-01
7.96184421e-01 5.82359791e-01 5.13243914e-01 -1.09095323e+00
1.78285345e-01 3.22101384e-01 -3.10452402e-01 -5.29781699e-01
5.21231234e-01 1.31649569e-01 2.72202283e-01 7.85489440e-01
4.14622337e-01 1.00882716e-01 3.31693262e-01 6.20677888e-01
1.37094486e+00 -1.90038994e-01 -1.28416261e-02 -6.50766730e-01
4.73451018e-01 1.69757437e-02 -2.72354074e-02 1.10580087e+00
-3.97572637e-01 5.06904483e-01 1.82636663e-01 -6.74735904e-01
-1.31654513e+00 -8.21984708e-01 4.43384908e-02 1.20982707e+00
-4.07777689e-02 -3.15107793e-01 -8.72095704e-01 -1.23653114e+00
-1.18737407e-01 3.37466657e-01 -9.48681235e-01 1.19593807e-01
-4.96100724e-01 -4.94084120e-01 6.87567234e-01 2.11328492e-01
1.16026926e+00 -1.01462972e+00 -9.88796279e-02 5.85136488e-02
-5.06399035e-01 -1.24135315e+00 -6.05743051e-01 9.29843751e-04
-8.88772607e-01 -7.87053108e-01 -9.15803492e-01 -1.03679109e+00
5.30303597e-01 3.03671986e-01 1.22502840e+00 3.98337960e-01
-1.08240955e-01 6.70681417e-01 -3.76577437e-01 -1.14419721e-02
-5.11870742e-01 4.17537659e-01 1.59896538e-02 -4.01687026e-02
3.63568515e-01 -2.61654109e-01 -5.80441117e-01 5.15210986e-01
-7.51199722e-01 5.30470312e-02 4.52978641e-01 8.88617098e-01
2.49667034e-01 7.11280778e-02 6.21843159e-01 -7.44159162e-01
5.28775454e-01 -9.27666128e-01 -1.83208019e-01 7.56287798e-02
-5.13229489e-01 -3.57331067e-01 4.36007351e-01 -4.77124125e-01
-8.06209147e-01 -1.16056345e-01 -5.31773120e-02 -1.77894086e-01
7.77235627e-02 4.05070931e-01 6.00121796e-01 -2.32146323e-01
8.06499064e-01 4.11978841e-01 -6.17499575e-02 -5.55291235e-01
-2.78300434e-01 1.11037266e+00 2.84191400e-01 -3.22831690e-01
3.72521311e-01 5.43250859e-01 -2.54936039e-01 -1.18868613e+00
-9.11473155e-01 -6.69130147e-01 -4.27793264e-01 -4.70223486e-01
9.69132960e-01 -9.05944824e-01 -2.02130720e-01 6.48300529e-01
-1.20511532e+00 -2.47858033e-01 2.99416006e-01 2.83481717e-01
-4.57794487e-01 7.96397626e-02 -1.01567304e+00 -5.98809302e-01
-6.55771613e-01 -9.08032835e-01 6.33989573e-01 4.00861911e-03
-2.37390753e-02 -1.20654178e+00 3.76685783e-02 5.65119445e-01
6.73108160e-01 -6.05156049e-02 5.55612266e-01 -1.22408116e+00
-4.28763062e-01 -5.06077826e-01 -4.20342863e-01 5.93128979e-01
-2.15478808e-01 -3.61304820e-01 -1.08888400e+00 -6.40442371e-01
1.73060179e-01 -5.95959961e-01 1.29040813e+00 7.67554119e-02
8.54535997e-01 -4.55980569e-01 -3.13713312e-01 1.61716223e-01
1.52714014e+00 1.07229322e-01 7.19397485e-01 6.17190719e-01
6.15373194e-01 6.25356615e-01 3.65296543e-01 1.68175578e-01
1.43538877e-01 5.16121805e-01 1.01283640e-02 1.68620840e-01
-4.38199818e-01 -2.09817141e-01 6.54738903e-01 9.47872758e-01
1.73631176e-01 -5.62613070e-01 -1.09607112e+00 8.07865798e-01
-1.68335497e+00 -9.24216449e-01 -6.67758584e-02 1.79214478e+00
5.47593057e-01 1.09054081e-01 2.09150389e-01 -1.93329856e-01
1.00478315e+00 2.91625559e-01 -1.34608254e-01 -2.53872991e-01
-2.54180998e-01 -4.20511849e-02 6.33045435e-01 2.43519723e-01
-1.25936019e+00 9.41553950e-01 7.80594110e+00 1.15094483e+00
-9.14029956e-01 6.15903497e-01 8.79498601e-01 -1.89717114e-01
3.13641101e-01 -4.65044826e-01 -7.78119087e-01 7.58504570e-01
1.39557290e+00 3.00783515e-01 4.74001318e-01 5.13089955e-01
-1.85989827e-01 -5.21358311e-01 -9.46739197e-01 8.88957560e-01
3.82799894e-01 -1.54654634e+00 5.65371737e-02 5.11836819e-02
1.11424649e+00 5.65697193e-01 3.12532455e-01 3.50829512e-01
-8.56513083e-02 -1.05140316e+00 8.80647182e-01 6.81285322e-01
5.97358763e-01 -5.41088223e-01 1.01493466e+00 3.59230965e-01
-8.94831002e-01 -2.38025691e-02 -4.33098525e-01 8.20923150e-02
-5.40629625e-01 4.41998988e-01 -8.36618721e-01 -1.08503319e-01
1.05110490e+00 7.11161256e-01 -9.18882430e-01 8.78279924e-01
2.77899921e-01 8.54265690e-01 -3.21741253e-01 -2.95780241e-01
4.67648119e-01 1.88442364e-01 8.11361969e-01 1.61148846e+00
1.79595739e-01 -4.21509415e-01 1.58856302e-01 6.16483450e-01
-5.94958425e-01 3.18419456e-01 -6.59704864e-01 -5.02799571e-01
3.00579399e-01 1.31316829e+00 -1.03058124e+00 -6.65965199e-01
-4.18738097e-01 1.33294523e+00 3.79257739e-01 3.07241052e-01
-4.78078485e-01 -5.51292956e-01 -9.66746882e-02 2.52997935e-01
3.11777949e-01 -2.88560361e-01 -2.82099247e-01 -1.03468060e+00
-1.12108141e-01 -8.64142239e-01 4.53263015e-01 -6.76319540e-01
-1.51982701e+00 8.24086368e-01 -3.49086732e-01 -8.07485640e-01
1.64791942e-01 -8.17911088e-01 -6.29193783e-01 8.04504991e-01
-9.66090620e-01 -1.09345043e+00 -1.41859576e-01 5.22562683e-01
7.22340524e-01 -7.60091364e-01 6.13248587e-01 5.40307879e-01
-3.15110087e-01 6.63093150e-01 5.79718769e-01 2.94391841e-01
8.32422316e-01 -1.10837531e+00 3.08348000e-01 4.55729216e-01
-4.05914970e-02 3.52938086e-01 5.31221867e-01 -8.39764535e-01
-1.03019941e+00 -8.96053851e-01 9.75509882e-01 -6.84562266e-01
8.09276879e-01 -4.37992126e-01 -5.51801980e-01 5.77531338e-01
7.21516669e-01 -1.77332506e-01 2.61177778e-01 -2.25985557e-01
-3.06479394e-01 1.58009499e-01 -1.40511370e+00 2.98392594e-01
3.99715751e-01 -7.75697708e-01 -3.78255695e-01 6.79552436e-01
2.47420415e-01 -6.69558644e-02 -9.13602173e-01 -1.10155739e-01
4.76660401e-01 -8.82722020e-01 6.29491031e-01 -5.45628369e-01
9.07907307e-01 3.78379524e-02 -3.40543270e-01 -1.13897789e+00
9.29504558e-02 -1.96704566e-01 -8.19046572e-02 1.12596714e+00
6.63680792e-01 -5.71887016e-01 7.54899859e-01 -1.16961591e-01
3.64354670e-01 -4.62182164e-01 -9.19490933e-01 -8.16843927e-01
3.04800779e-01 -1.74761713e-02 -2.13068575e-01 1.02297354e+00
-1.49596959e-01 5.64316034e-01 -5.55165827e-01 -3.45682561e-01
3.29861134e-01 -4.06248212e-01 5.21875978e-01 -9.28702652e-01
-7.15783760e-02 -3.09938014e-01 -6.04973733e-01 -8.66004288e-01
1.22895546e-01 -9.30275261e-01 2.30860747e-02 -1.60762882e+00
7.30143964e-01 -1.30920216e-01 -4.33941513e-01 2.05588758e-01
1.34285241e-01 8.18889737e-01 3.59612852e-02 5.76649368e-01
-9.38698709e-01 1.40587538e-01 8.97753119e-01 -5.04836440e-01
2.16009691e-01 -2.03444541e-01 -6.66604117e-02 5.99765182e-01
9.15185094e-01 -7.87591517e-01 -1.69336647e-02 -3.76399994e-01
3.81703198e-01 -1.86039597e-01 2.01868728e-01 -1.21973538e+00
2.72259682e-01 2.14852974e-01 6.41428173e-01 -8.72207224e-01
3.74617398e-01 -5.98780453e-01 -3.93557876e-01 2.57363826e-01
-4.59290743e-01 2.54400760e-01 3.53205293e-01 5.54724872e-01
-1.36958271e-01 -5.38294196e-01 1.09778321e+00 -5.52703738e-01
-4.85995352e-01 1.66435882e-01 -9.30066824e-01 4.97538447e-02
7.57676721e-01 -1.54699445e-01 -6.12498105e-01 -6.27767026e-01
-7.89162457e-01 1.00168832e-01 3.38090539e-01 3.64871502e-01
5.39064944e-01 -1.27478540e+00 -8.27330709e-01 -1.89594299e-01
-1.40524074e-01 -9.72402513e-01 1.85710549e-01 9.93992984e-01
-8.03244412e-01 3.03347498e-01 -4.79446471e-01 -3.48703206e-01
-1.28033376e+00 2.64839888e-01 6.46122634e-01 -4.40000504e-01
-4.58242476e-01 8.30362439e-01 -5.36536351e-02 -4.05973345e-01
3.56974095e-01 6.51734233e-01 -3.19037735e-01 2.56159991e-01
8.27487409e-01 7.39582598e-01 2.05858842e-01 -7.96665251e-01
-2.45991215e-01 1.98412791e-01 -6.14323139e-01 -5.43972194e-01
1.39354086e+00 -3.54264528e-01 -3.53701144e-01 6.36235535e-01
1.59957290e+00 -2.74660349e-01 -6.89679801e-01 -3.56584162e-01
2.35166848e-01 -2.83976436e-01 4.97485131e-01 -9.94765162e-01
-1.15390813e+00 1.02611268e+00 1.04158211e+00 4.84200656e-01
6.55485988e-01 -3.30182314e-01 9.46344197e-01 7.10103273e-01
2.22432956e-01 -1.70732033e+00 5.77492118e-01 8.72753441e-01
9.57222342e-01 -1.26480329e+00 8.36552605e-02 1.98519766e-01
-4.66938645e-01 1.19554889e+00 6.94522917e-01 -3.17184389e-01
7.57618368e-01 7.05027115e-03 2.44048871e-02 -4.02622312e-01
-7.10530221e-01 2.47571930e-01 4.84095961e-01 2.86384642e-01
4.83781636e-01 -2.36316741e-01 -4.10655528e-01 1.96087986e-01
4.56090868e-02 -3.03443223e-01 5.13975084e-01 1.27196538e+00
-7.36751735e-01 -4.77424204e-01 -3.57210040e-01 8.20933461e-01
-9.54486668e-01 -3.05315703e-01 -7.40928352e-01 6.34285033e-01
-3.10440157e-02 1.13856232e+00 3.15534443e-01 -8.23833048e-01
-1.13263287e-01 2.93528110e-01 3.49210620e-01 -5.66652000e-01
-8.80563200e-01 -2.42690966e-02 3.24481100e-01 -3.80912840e-01
-4.10550088e-01 -1.96367666e-01 -8.65964293e-01 -7.90280521e-01
-4.12191331e-01 1.26479551e-01 1.08249426e+00 8.71360004e-01
3.15777510e-01 1.95220441e-01 7.69086540e-01 -9.60494280e-01
-3.18309247e-01 -1.36375344e+00 -9.12892282e-01 2.99536794e-01
2.09445581e-01 -4.64564592e-01 -5.34247696e-01 1.79114819e-01] | [8.419997215270996, 10.696388244628906] |
f937fb6c-2084-4457-a0d2-20cc6eae0b3b | dynamic-graph-based-label-propagation-for | null | null | https://www.sciencedirect.com/science/article/abs/pii/S0957417418304998 | https://www.sciencedirect.com/science/article/abs/pii/S0957417418304998 | Dynamic Graph-Based Label Propagation for Density Peaks Clustering | Clustering is a major approach in data mining and machine learning and has been successful in many real-world applications. Density peaks clustering (DPC) is a recently published method that uses an intuitive to cluster data objects efficiently and effectively. However, DPC and most of its improvements suffer from some shortcomings to be addressed. For instance, this method only considers the global structure of data which leading to missing many clusters. The cut-off distance affects the local density values and is calculated in different ways depending on the size of the datasets, which can influence the quality of clustering. Then, the original label assignment can cause a “chain reaction”, whereby if a wrong label is assigned to a data point, and then there may be many more wrong labels subsequently assigned to the other points. In this paper, a density peaks clustering method called DPC-DLP is proposed. The proposed method employs the idea of k-nearest neighbors to compute the global cut-off parameter and the local density of each point. Moreover, the proposed method uses a graph-based label propagation to assign labels to remaining points and form final clusters. The proposed label propagation can effectively assign true labels to those of data instances which located in border and overlapped regions. The proposed method can be applied to some applications. To make the method practical for image clustering, the local structure is used to achieve low-dimensional space. In addition, proposed method considers label space correlation, to be effective in the gene expression problems. Several experiments are performed to evaluate the performance of the proposed method on both synthetic and real-world datasets. The results demonstrate that in most cases, the proposed method outperformed some state-of-the-art methods. | ['Abdulrahman Lotfi', 'Nooruldeen Nasih Qader', 'Seyed Amjad Seyedi', 'Parham Moradi'] | 2019-01-01 | null | null | null | null | ['image-clustering'] | ['computer-vision'] | [-5.44806533e-02 -3.06166321e-01 -1.46459863e-01 -1.08436719e-01
-1.76130667e-01 -2.94389188e-01 2.40899265e-01 6.04116619e-01
-3.15546125e-01 5.47532260e-01 -7.47349411e-02 3.95282805e-02
-5.69617450e-01 -9.55298007e-01 -1.88560128e-01 -1.12692797e+00
-1.29201338e-01 6.95139885e-01 6.17891371e-01 1.99624643e-01
4.78554934e-01 4.51728851e-01 -1.65066612e+00 2.86935508e-01
8.60786676e-01 6.67859375e-01 3.34675848e-01 2.87830420e-02
-6.24162018e-01 3.58979642e-01 -6.25670075e-01 2.03914970e-01
1.48244292e-01 -4.55451310e-01 -5.31134963e-01 1.85700625e-01
-3.90077561e-01 2.43432969e-01 4.59091634e-01 1.13315797e+00
3.27795684e-01 1.61502644e-01 1.08470809e+00 -1.27729714e+00
-2.83074707e-01 3.12794805e-01 -1.24541295e+00 -4.08781022e-02
-5.08203246e-02 -1.86540753e-01 5.02701223e-01 -6.31872177e-01
3.96268070e-01 1.25459170e+00 5.29542148e-01 8.48122165e-02
-1.00677705e+00 -7.88541138e-01 -1.80468038e-02 2.69176841e-01
-1.72412896e+00 1.14354990e-01 8.32707286e-01 -5.74568808e-01
2.83754468e-01 5.12839742e-02 4.96494412e-01 5.44580780e-02
1.27494574e-01 2.55419046e-01 1.30157506e+00 -4.02754486e-01
5.75184345e-01 2.34341845e-01 3.04553241e-01 5.43952286e-01
3.66040379e-01 -3.29216540e-01 2.49239691e-02 -2.52857715e-01
2.36853257e-01 2.76014596e-01 -2.32304975e-01 -4.06621665e-01
-9.31599855e-01 8.11398983e-01 5.55652261e-01 6.72653913e-01
-3.05959553e-01 -2.68732160e-01 4.16309863e-01 -3.30956101e-01
1.09538093e-01 -1.44210393e-02 -7.52389058e-02 -4.39914241e-02
-8.70907247e-01 -7.00902864e-02 4.46909845e-01 6.60843134e-01
9.41249371e-01 -5.20401180e-01 1.45022646e-01 9.15476859e-01
5.12824416e-01 1.40581414e-01 6.30855560e-01 -5.81219792e-01
3.49030077e-01 1.06880462e+00 2.54756883e-02 -1.60105491e+00
-6.85282588e-01 -2.96388090e-01 -1.00859714e+00 2.21062973e-01
3.90658289e-01 1.02534825e-02 -9.28955793e-01 1.41393209e+00
6.14752173e-01 2.82992274e-01 2.89597344e-02 1.00531518e+00
5.83733618e-01 1.08502352e+00 2.21497431e-01 -6.35257900e-01
1.29851103e+00 -3.82890314e-01 -7.99558163e-01 3.27862293e-01
5.62645435e-01 -1.01771557e+00 7.70660043e-01 5.75291514e-01
-5.26301801e-01 -6.32481813e-01 -9.27029967e-01 4.82514232e-01
-3.87755424e-01 1.04733884e-01 3.24626356e-01 4.72054183e-01
-6.64529920e-01 3.21944088e-01 -7.17565358e-01 -7.18649805e-01
2.76610970e-01 4.48563993e-01 -7.85785019e-02 -2.26565763e-01
-9.41921055e-01 4.41993833e-01 7.59579480e-01 5.92114478e-02
-2.26920187e-01 -3.27352792e-01 -3.00466001e-01 8.94326866e-02
3.02589297e-01 -2.09164619e-01 4.37648714e-01 -7.09946275e-01
-9.69958544e-01 4.62477177e-01 -1.90254718e-01 2.10484420e-03
1.22228697e-01 4.81538415e-01 -3.43769640e-01 1.49135545e-01
1.31705657e-01 4.76938218e-01 1.93904012e-01 -1.36578000e+00
-8.77428532e-01 -6.92169130e-01 -4.92802531e-01 4.52796102e-01
-2.50540763e-01 -2.46462107e-01 -6.63012743e-01 -3.26122433e-01
5.23848116e-01 -8.77981961e-01 -1.43998593e-01 -3.78690928e-01
-2.63902456e-01 -5.26744664e-01 1.08267510e+00 -1.52818725e-01
1.35233510e+00 -2.35940361e+00 -1.40342489e-02 7.63898432e-01
7.03421682e-02 2.86534131e-01 3.24695021e-01 5.78181565e-01
1.61421642e-01 2.45788708e-01 -3.45376045e-01 1.86795160e-01
-4.47278649e-01 1.28994629e-01 4.04993862e-01 5.47615528e-01
-1.73778251e-01 -2.78521646e-02 -7.87749171e-01 -8.79182160e-01
3.37845534e-01 4.90864575e-01 -2.75615215e-01 -1.10304490e-01
-3.14710625e-02 4.08464760e-01 -5.19435465e-01 3.46833408e-01
1.16014826e+00 -9.59439278e-02 2.97813624e-01 -4.02700454e-01
-2.38180622e-01 -5.49220681e-01 -1.63566542e+00 1.06130445e+00
-3.49764451e-02 9.37197283e-02 -8.45234990e-02 -1.18628073e+00
1.35133219e+00 2.35923707e-01 8.36043000e-01 -3.43010306e-01
3.01920533e-01 2.40926102e-01 1.00311480e-01 -5.56367099e-01
1.17608167e-01 -2.34578535e-01 2.39203826e-01 2.95884609e-01
-4.10491586e-01 2.93139011e-01 3.08877558e-01 1.03007309e-01
8.02071035e-01 -3.15070361e-01 3.02841187e-01 -3.45607758e-01
8.29717457e-01 2.41093323e-01 8.98406446e-01 2.36781746e-01
-1.61868006e-01 3.84802669e-01 5.55653453e-01 -6.48104474e-02
-7.73665249e-01 -7.66747832e-01 -2.34878719e-01 4.26533580e-01
6.01497769e-01 -2.27198794e-01 -8.31535101e-01 -5.67469478e-01
-5.32397553e-02 4.94706720e-01 -3.75126362e-01 -9.83032063e-02
-1.22479245e-01 -1.04531908e+00 7.31289759e-02 -2.22136285e-02
7.40764081e-01 -9.61022794e-01 -3.00702035e-01 3.03438723e-01
-8.86796191e-02 -6.36908710e-01 -1.76079854e-01 -1.27498776e-01
-9.42309797e-01 -1.19007456e+00 -5.35651982e-01 -9.37985361e-01
9.45287704e-01 4.10689592e-01 4.18635577e-01 3.28963965e-01
-1.84702650e-01 -1.61806062e-01 -6.90905273e-01 -2.24439055e-01
-2.98624396e-01 1.98046371e-01 -1.26100585e-01 2.78293133e-01
7.54079163e-01 -4.45459366e-01 -6.58153236e-01 6.51290178e-01
-9.60608244e-01 -1.06645852e-01 4.88460958e-01 5.60722172e-01
7.51478851e-01 9.34546590e-01 8.68323445e-01 -1.14603353e+00
7.41666555e-01 -7.27859080e-01 -7.30063796e-01 1.58298403e-01
-7.94969320e-01 -6.30874708e-02 8.98610890e-01 -1.19666025e-01
-1.00373292e+00 1.82408139e-01 3.92286479e-02 -2.48964593e-01
-3.14207256e-01 5.84156334e-01 -3.00080538e-01 2.83344209e-01
3.82911980e-01 1.00903541e-01 2.47396275e-01 -5.17971933e-01
3.75626869e-02 9.37815368e-01 1.92293629e-01 -3.18591565e-01
4.31284130e-01 4.20483410e-01 4.12329972e-01 -6.31018698e-01
-2.40452304e-01 -8.18638384e-01 -6.11160278e-01 -3.14109415e-01
1.04605246e+00 -5.07199228e-01 -8.66824150e-01 4.76670742e-01
-7.91712463e-01 3.59085202e-01 3.45325261e-01 6.13247156e-01
-1.27739385e-01 4.44024324e-01 -2.90250510e-01 -9.12513673e-01
-1.37310684e-01 -1.23513627e+00 5.82821548e-01 5.10535121e-01
1.34482980e-01 -1.01486170e+00 -8.61601532e-02 3.42039540e-02
6.46026805e-02 3.53212476e-01 1.29247546e+00 -5.56960642e-01
-3.61411244e-01 -8.79413188e-02 -2.69133538e-01 -3.49393189e-02
4.34584588e-01 2.61337727e-01 -4.58204299e-01 -2.46068224e-01
-8.98503736e-02 1.90227792e-01 5.39154470e-01 3.44797283e-01
1.02690566e+00 -1.00456052e-01 -7.92536259e-01 2.77712941e-01
1.78550816e+00 7.67804444e-01 6.84177876e-01 2.34508574e-01
5.89455426e-01 7.58041561e-01 9.09876943e-01 4.69002157e-01
1.07452929e-01 5.88386059e-01 3.57290894e-01 -3.64561230e-01
1.16478682e-01 -1.29960403e-01 -1.05491817e-01 9.76641834e-01
6.61160573e-02 -3.02612394e-01 -9.73251104e-01 5.02038002e-01
-1.93538344e+00 -9.63342607e-01 -6.20373964e-01 2.29461360e+00
6.60327435e-01 5.91268428e-02 1.48801520e-01 5.77403545e-01
1.22915399e+00 -2.99985558e-01 -4.05464262e-01 -2.97682554e-01
1.72416121e-01 -2.01940518e-02 4.32650417e-01 3.71381193e-01
-7.64931917e-01 5.71383297e-01 5.01008892e+00 9.74862516e-01
-1.03062463e+00 2.86551807e-02 7.74326503e-01 2.14953348e-01
-2.54413765e-02 1.10028863e-01 -7.69384384e-01 9.37054574e-01
4.38384026e-01 -4.66619022e-02 2.10891917e-01 6.09714448e-01
5.98942816e-01 -6.54969990e-01 -4.71144617e-01 9.86962676e-01
-2.02874586e-01 -8.72561634e-01 -2.46007312e-02 2.61578023e-01
7.14449108e-01 -4.62894738e-01 -1.83152437e-01 -1.34718627e-01
1.13037467e-01 -7.31462777e-01 -6.08733017e-03 5.19774973e-01
4.82671380e-01 -1.27869523e+00 1.11302102e+00 5.93325496e-01
-1.16942573e+00 -6.43803328e-02 -6.61785543e-01 4.28964868e-02
-2.52043623e-02 1.11738420e+00 -1.08329189e+00 5.89814365e-01
6.37813210e-01 4.91924614e-01 -5.29481113e-01 1.16426170e+00
1.85583457e-01 6.09298885e-01 -5.86037576e-01 -2.70568818e-01
3.14505249e-01 -6.20934308e-01 1.59456670e-01 9.20337439e-01
5.59119880e-01 2.41601646e-01 2.52077460e-01 6.60272956e-01
2.19716072e-01 7.43281662e-01 -5.32521904e-01 3.37540746e-01
8.54974389e-01 1.29328966e+00 -1.29001975e+00 -2.79792905e-01
-2.78626144e-01 5.29811859e-01 1.39724940e-01 2.10427582e-01
-7.89699256e-01 -6.29862010e-01 -4.28374931e-02 4.87718850e-01
8.76578763e-02 -3.70663218e-02 -1.45320952e-01 -4.55591291e-01
-2.97823966e-01 -5.04332244e-01 5.43456972e-01 -5.79465926e-01
-1.23319197e+00 2.52078801e-01 1.25178218e-01 -1.23056614e+00
1.22840948e-01 -1.71955973e-01 -6.44370437e-01 7.29718149e-01
-1.05163455e+00 -6.30457520e-01 -4.99674290e-01 5.23327351e-01
2.76024073e-01 -1.29772380e-01 5.05817711e-01 5.40706277e-01
-5.58976471e-01 2.77999580e-01 5.72574437e-01 -6.97487146e-02
7.20179081e-01 -8.93185616e-01 -6.14395082e-01 6.51693046e-01
-2.44791672e-01 6.17088079e-01 4.86110121e-01 -8.36874068e-01
-7.07465351e-01 -1.01730275e+00 4.99645799e-01 3.81607711e-01
6.31471574e-02 -8.72766152e-02 -9.07651305e-01 1.78343624e-01
9.08305272e-02 -1.70877337e-01 6.23789370e-01 8.00510347e-02
3.47730547e-01 -5.16652346e-01 -1.48511469e+00 3.11952591e-01
4.50771749e-01 2.23266602e-01 -2.06309944e-01 2.26645619e-01
1.79505587e-01 1.03236571e-01 -9.14624155e-01 4.05853540e-01
2.82816201e-01 -1.34059405e+00 4.97137457e-01 8.14691409e-02
7.67782703e-02 -1.11368024e+00 2.27716014e-01 -1.27243078e+00
-5.49127400e-01 7.81084225e-02 5.38393617e-01 1.78705156e+00
2.82847941e-01 -5.60917556e-01 7.86352515e-01 3.20853442e-01
1.29591182e-01 -8.40855658e-01 -7.51828849e-01 -4.64941829e-01
-1.32257029e-01 1.10360645e-01 7.81090975e-01 9.92834449e-01
1.54526427e-01 3.59310120e-01 -4.52889837e-02 8.43184963e-02
6.82053864e-01 -7.33297467e-02 5.94564080e-01 -1.49825299e+00
5.17057776e-02 -2.24336505e-01 -6.62895083e-01 -5.66742063e-01
-2.60486633e-01 -8.92826378e-01 -6.47765771e-02 -1.76132214e+00
2.34848738e-01 -1.12613344e+00 -2.86455125e-01 2.12427020e-01
-2.85070449e-01 1.95514306e-01 1.25192165e-01 6.75772130e-01
-4.07565832e-01 1.54316097e-01 1.12469733e+00 2.92811729e-02
-5.03626764e-01 1.16304876e-02 -3.77228349e-01 5.78121781e-01
1.07389700e+00 -6.08140349e-01 -7.07042754e-01 -2.03273986e-02
-1.01823267e-02 -8.26965347e-02 -2.23247156e-01 -1.23895216e+00
4.51161176e-01 -2.39972055e-01 4.84051764e-01 -8.14651251e-01
-5.24598435e-02 -1.18923163e+00 5.96550226e-01 3.93965065e-01
2.90249940e-02 5.06925695e-02 -6.89338669e-02 5.76345146e-01
-2.24838838e-01 -6.62720501e-01 1.10309803e+00 -2.28713274e-01
-5.48332155e-01 -1.98974125e-02 -4.30029541e-01 -2.11970389e-01
1.60011089e+00 -3.56782436e-01 -1.79900050e-01 -1.47479504e-01
-5.48540056e-01 3.61909240e-01 6.52259052e-01 -4.14374620e-02
3.73604357e-01 -1.46726406e+00 -4.48158443e-01 8.13055784e-02
1.68631166e-01 2.74404168e-01 2.18647555e-01 9.07559395e-01
-7.99264550e-01 2.11331412e-01 -4.39636149e-02 -8.87992442e-01
-1.28587139e+00 7.71979690e-01 3.58707532e-02 -8.96496773e-02
-3.27201545e-01 1.48281649e-01 9.00058225e-02 -1.88108340e-01
4.02276888e-02 -1.70777693e-01 -5.82483590e-01 1.13934420e-01
1.73813313e-01 5.20328641e-01 -4.11676019e-02 -6.95720077e-01
-3.88405144e-01 1.07879138e+00 4.95882854e-02 -5.26756532e-02
9.40586209e-01 -1.69756800e-01 -4.18087512e-01 5.42377174e-01
1.23413479e+00 6.97531253e-02 -6.96115196e-01 2.03110259e-02
1.02068439e-01 -4.88606274e-01 -6.99160108e-03 -6.89262748e-01
-1.15312791e+00 7.35204935e-01 8.66439044e-01 2.58898616e-01
1.29516232e+00 -1.15660369e-01 6.22255981e-01 -9.68433768e-02
3.92946661e-01 -1.34593105e+00 -2.06475064e-01 5.62126888e-03
2.00509951e-01 -9.63816702e-01 1.79425955e-01 -6.10395133e-01
-6.23995185e-01 1.08721817e+00 7.33509362e-01 8.40770230e-02
8.47572744e-01 1.12777129e-01 7.79189616e-02 -2.63748556e-01
-3.09841365e-01 -5.34235649e-02 -1.43737063e-01 6.72765374e-01
4.09726053e-01 1.22076292e-02 -9.56357300e-01 3.94409746e-01
1.40903205e-01 1.31269142e-01 3.91200632e-01 6.69860244e-01
-8.08943033e-01 -1.11796415e+00 -8.73210430e-01 5.33466280e-01
-3.19837570e-01 4.00173694e-01 -2.16370076e-01 8.51064861e-01
8.19913983e-01 1.18503499e+00 3.12539846e-01 -4.32298452e-01
8.99714828e-02 2.07051933e-02 9.73638222e-02 -3.77390087e-01
-2.28313640e-01 3.66419464e-01 -4.26649839e-01 -3.47478651e-02
-7.75636196e-01 -5.70760608e-01 -2.01948905e+00 -3.56439650e-01
-5.60984671e-01 6.93199039e-01 6.50024414e-01 6.05448365e-01
4.51310813e-01 3.48148137e-01 8.63660514e-01 -1.38356328e-01
1.18390001e-01 -8.86034131e-01 -8.46160531e-01 5.62892795e-01
-2.68044859e-01 -9.35465157e-01 -4.35364574e-01 8.91790912e-02] | [7.643976211547852, 4.569610595703125] |
146c0e05-3b5c-462b-9e1d-213fb9b4e2be | level-generation-and-style-enhancement-deep | 2107.07397 | null | https://arxiv.org/abs/2107.07397v1 | https://arxiv.org/pdf/2107.07397v1.pdf | Level generation and style enhancement -- deep learning for game development overview | We present practical approaches of using deep learning to create and enhance level maps and textures for video games -- desktop, mobile, and web. We aim to present new possibilities for game developers and level artists. The task of designing levels and filling them with details is challenging. It is both time-consuming and takes effort to make levels rich, complex, and with a feeling of being natural. Fortunately, recent progress in deep learning provides new tools to accompany level designers and visual artists. Moreover, they offer a way to generate infinite worlds for game replayability and adjust educational games to players' needs. We present seven approaches to create level maps, each using statistical methods, machine learning, or deep learning. In particular, we include: - Generative Adversarial Networks for creating new images from existing examples (e.g. ProGAN). - Super-resolution techniques for upscaling images while preserving crisp detail (e.g. ESRGAN). - Neural style transfer for changing visual themes. - Image translation - turning semantic maps into images (e.g. GauGAN). - Semantic segmentation for turning images into semantic masks (e.g. U-Net). - Unsupervised semantic segmentation for extracting semantic features (e.g. Tile2Vec). - Texture synthesis - creating large patterns based on a smaller sample (e.g. InGAN). | ['Błażej Podgórski', 'Bartłomiej Olechno', 'Piotr Migdał'] | 2021-07-15 | null | null | null | null | ['texture-synthesis', 'unsupervised-semantic-segmentation'] | ['computer-vision', 'computer-vision'] | [ 3.64171565e-01 3.52706671e-01 3.65880460e-01 -8.35156515e-02
-7.15952575e-01 -5.91533482e-01 5.07889271e-01 -5.40325522e-01
1.83376651e-02 6.71483338e-01 3.00548553e-01 -2.21118212e-01
2.47654870e-01 -1.30320489e+00 -9.43514884e-01 -4.85836864e-01
1.00363247e-01 1.95896626e-01 3.11919630e-01 -7.14502275e-01
4.05595213e-01 4.99151200e-01 -1.71595323e+00 9.04386163e-01
6.86508775e-01 7.85448909e-01 2.19667807e-01 9.58381176e-01
-6.12627447e-01 9.49884593e-01 -8.79777491e-01 -6.01615906e-01
5.54413259e-01 -7.83433080e-01 -5.99853337e-01 3.99431556e-01
3.79049659e-01 -2.72591144e-01 -2.68290520e-01 1.19404674e+00
4.65780884e-01 3.62229645e-02 5.08786023e-01 -1.33586514e+00
-9.73833919e-01 4.93128449e-01 -5.95474601e-01 -2.12580517e-01
4.10830945e-01 3.71744514e-01 4.25085515e-01 -5.31804442e-01
7.19358563e-01 1.36692941e+00 8.63282323e-01 8.13195050e-01
-1.34625137e+00 -7.08065689e-01 -2.90458649e-01 -6.86827162e-03
-1.14289677e+00 -1.60160348e-01 7.74583340e-01 -5.84198177e-01
5.79725266e-01 5.06275415e-01 1.10335267e+00 1.19367003e+00
2.54613489e-01 5.85609198e-01 1.59466875e+00 -6.45569205e-01
3.36833596e-01 1.95458382e-01 -6.48077548e-01 6.68303907e-01
-2.12640405e-01 1.72625124e-01 -4.24608052e-01 3.93717706e-01
1.62356126e+00 -2.65813828e-01 1.87981278e-01 -2.29689628e-01
-9.90269363e-01 7.33158767e-01 3.96737158e-01 2.86285162e-01
-1.69216648e-01 3.92912716e-01 2.33779907e-01 5.69013357e-01
4.68006551e-01 8.70075524e-01 -2.25284338e-01 -2.45629966e-01
-1.01861668e+00 4.70812500e-01 5.08533418e-01 1.01598859e+00
8.64356577e-01 6.36188745e-01 -2.02760085e-01 8.38986516e-01
-2.27386340e-01 2.72598743e-01 5.06757438e-01 -1.20652544e+00
-1.26842987e-02 3.23205709e-01 -2.84401309e-02 -1.00786126e+00
-1.57733843e-01 -3.71955112e-02 -8.20209086e-01 1.18502581e+00
4.53841507e-01 -3.30909371e-01 -1.38277757e+00 1.55886245e+00
-1.67997705e-03 4.15141314e-01 -7.52109438e-02 6.21531963e-01
1.04845393e+00 9.03811753e-01 -9.24427658e-02 4.81364518e-01
1.51316822e+00 -9.46119964e-01 -6.53672576e-01 -1.91529348e-01
2.06093386e-01 -8.12064707e-01 1.63034368e+00 5.90730190e-01
-1.60675764e+00 -7.77489603e-01 -1.15103197e+00 -7.42019266e-02
-7.97908604e-01 -3.05479616e-01 5.78193665e-01 9.54091251e-01
-1.51241326e+00 8.08382750e-01 -6.00391984e-01 -2.18328819e-01
5.66652358e-01 3.32327664e-01 -2.73745030e-01 2.32896909e-01
-1.07093084e+00 6.69339120e-01 2.41217405e-01 -4.08432811e-01
-5.70285022e-01 -8.62015307e-01 -1.01050651e+00 -9.65147540e-02
1.07072741e-01 -7.06454933e-01 1.13187957e+00 -1.52800477e+00
-1.90337682e+00 1.08208358e+00 2.85375178e-01 -3.25921118e-01
6.57283247e-01 1.34702027e-01 -3.28395158e-01 -1.57162305e-02
2.89841712e-01 1.24732137e+00 1.02340269e+00 -1.52865744e+00
-8.27305913e-01 -1.55180395e-02 3.30448955e-01 2.90031195e-01
-6.18274994e-02 -1.64863110e-01 -5.17366707e-01 -1.23763657e+00
-6.09072633e-02 -5.57984054e-01 -4.48060542e-01 -9.79446992e-02
-3.28246027e-01 4.00333017e-01 9.14440095e-01 -1.05890441e+00
9.46087718e-01 -2.04350257e+00 1.63244948e-01 3.45401347e-01
4.50506032e-01 1.64207801e-01 -1.62272841e-01 2.17364114e-02
-1.71669632e-01 3.38478327e-01 5.15275728e-03 -5.76757714e-02
-4.93774265e-02 1.13643102e-01 1.42046317e-01 -1.94618180e-01
1.11708835e-01 1.16492224e+00 -6.07628882e-01 -2.20357254e-01
5.32081664e-01 4.95020270e-01 -1.00644720e+00 1.01754725e-01
-3.75529975e-01 5.71543574e-01 3.84486951e-02 3.38849038e-01
5.42567372e-01 -1.17165809e-02 -1.85114518e-01 -1.31199155e-02
-1.67999461e-01 6.54609799e-02 -1.39616048e+00 1.68363547e+00
-8.29617441e-01 8.71450186e-01 9.43018496e-02 -7.00004935e-01
1.06533277e+00 -1.40731275e-01 2.51265347e-01 -9.79211330e-01
8.97085220e-02 1.92777235e-02 -3.43193352e-01 -3.15676033e-01
7.91002512e-01 -3.60689312e-01 -4.27897543e-01 1.56742379e-01
-2.74927747e-02 -6.84754372e-01 8.46703723e-02 -9.95541289e-02
9.73123670e-01 2.45946243e-01 -6.76697046e-02 -1.72233507e-01
1.66701123e-01 3.37724313e-02 2.79245049e-01 7.01865375e-01
3.31889004e-01 9.04433310e-01 6.94646955e-01 -7.64595926e-01
-1.70426095e+00 -1.12031984e+00 4.26116139e-01 1.09057164e+00
1.61236629e-03 -1.28549233e-01 -1.32872188e+00 -1.31417528e-01
-3.74588877e-01 7.37872243e-01 -6.54825747e-01 -2.11827811e-02
-4.17231381e-01 -1.16722651e-01 5.38140297e-01 5.07851362e-01
6.88255847e-01 -1.50180197e+00 -4.45815951e-01 4.03038055e-01
7.71551281e-02 -9.39065397e-01 -5.09336293e-01 9.87324119e-02
-5.30422568e-01 -4.66158926e-01 -7.92705059e-01 -1.29341948e+00
5.35373509e-01 -4.49193977e-02 1.37899494e+00 -1.09947369e-01
-4.31492209e-01 1.34690434e-01 -3.50444496e-01 -4.00190085e-01
-7.19774187e-01 -1.17334552e-01 -2.59587198e-01 -2.49859601e-01
3.61994021e-02 -8.26154232e-01 -6.22729599e-01 1.97901174e-01
-1.23109734e+00 6.93733394e-01 4.76097882e-01 5.70432186e-01
6.79726064e-01 3.31748605e-01 1.47258803e-01 -1.26573145e+00
8.87436509e-01 4.84976359e-02 -5.47902286e-01 -5.31788431e-02
-1.52386889e-01 -4.60576359e-03 7.08517790e-01 -4.88847017e-01
-1.05132473e+00 -2.89438337e-01 -4.67775166e-01 -4.15063381e-01
-2.46909603e-01 -1.00046042e-02 -2.72381335e-01 -1.27179101e-01
9.25902426e-01 1.21880010e-01 9.10357311e-02 -7.94591233e-02
8.97859395e-01 4.99145031e-01 9.41273093e-01 -5.28130770e-01
9.97626603e-01 4.49586779e-01 -2.20286936e-01 -8.07674885e-01
-3.11427802e-01 2.09725246e-01 -4.92546231e-01 -3.61291468e-01
1.25331306e+00 -8.51223588e-01 -1.50529534e-01 7.12513983e-01
-9.02344227e-01 -9.40814912e-01 -8.83207083e-01 -2.48989061e-01
-7.76270211e-01 -6.27959222e-02 -8.85044456e-01 -1.14297576e-01
-4.54851892e-03 -1.22827053e+00 9.08401191e-01 6.21504545e-01
-2.88137794e-01 -1.08899105e+00 1.11822551e-03 4.79576796e-01
5.44541240e-01 7.86881804e-01 6.95972621e-01 4.21852395e-02
-5.79472542e-01 7.64200464e-02 -3.55454296e-01 3.60950142e-01
2.91500539e-01 -8.28081928e-03 -8.02547932e-01 6.89408602e-03
-4.42470491e-01 -2.37914473e-01 3.77449185e-01 5.82801819e-01
1.39154506e+00 -7.31933355e-01 8.60886499e-02 8.98198247e-01
1.36073542e+00 6.70945585e-01 1.34180427e+00 5.85295916e-01
9.74573314e-01 4.14184421e-01 2.43883699e-01 3.61031532e-01
2.32459128e-01 7.13229775e-01 -1.93100907e-02 -7.63193190e-01
-5.40957153e-01 -5.28838336e-01 1.21409163e-01 5.63048363e-01
-1.54643312e-01 1.31548166e-01 -5.86881220e-01 2.04290882e-01
-1.61316836e+00 -1.12434685e+00 3.96395586e-02 1.79153979e+00
8.93632233e-01 3.99521321e-01 2.53167212e-01 3.70427892e-02
7.95909107e-01 6.47098497e-02 -2.71072447e-01 -9.73817348e-01
-1.30677208e-01 1.03556871e+00 4.74191368e-01 6.29954875e-01
-9.49134231e-01 1.50770283e+00 5.74335623e+00 1.19184828e+00
-1.04830933e+00 1.83685079e-01 1.10617244e+00 2.39857193e-02
-6.61272883e-01 -2.53993720e-01 -2.88665295e-01 5.46578646e-01
4.00759339e-01 -5.75190037e-02 8.20593119e-01 8.70064199e-01
1.93680972e-01 2.32433066e-01 -5.30344784e-01 1.19784224e+00
-4.74442467e-02 -1.93689394e+00 2.01067477e-01 -5.88873066e-02
1.15701973e+00 -6.17002964e-01 3.30867171e-01 4.70121115e-01
7.36143172e-01 -1.30892491e+00 9.83226597e-01 5.10408282e-01
1.51772642e+00 -9.15302217e-01 4.21101063e-01 -1.35034353e-01
-1.19379544e+00 3.36799979e-01 -3.00098568e-01 -2.64919519e-01
-1.72241911e-01 2.71033734e-01 -3.74757320e-01 -4.63032685e-02
8.16206753e-01 2.05259815e-01 -5.39030313e-01 5.83400607e-01
-1.79450989e-01 6.77900314e-02 -9.98742506e-02 -7.49050602e-02
3.63631874e-01 -3.70959938e-01 1.88275769e-01 1.32169557e+00
4.44654852e-01 1.57478303e-01 -2.47491035e-03 1.09527576e+00
-1.64522767e-01 4.12767231e-02 -7.34826982e-01 7.78824463e-02
2.86484867e-01 1.19355273e+00 -1.08716512e+00 -3.94397259e-01
-2.96754241e-01 1.48030949e+00 -3.45771834e-02 2.64415592e-01
-7.60443211e-01 -7.81828940e-01 8.22711945e-01 7.24288046e-01
1.61361396e-01 -1.13636129e-01 -8.05070281e-01 -9.58149731e-01
-2.74135292e-01 -1.37863600e+00 -8.40188861e-02 -1.19891787e+00
-9.58991528e-01 6.52847886e-01 -1.94833860e-01 -9.95484710e-01
-3.85321677e-01 -5.13063073e-01 -8.76810193e-01 8.33846331e-01
-8.61639798e-01 -1.17172015e+00 -5.05684495e-01 7.47931421e-01
9.84597862e-01 -4.86556202e-01 4.25238192e-01 2.30170414e-01
1.03123441e-01 6.23483181e-01 7.31869368e-03 2.29160592e-01
2.69600689e-01 -1.42634737e+00 9.87706721e-01 8.51914048e-01
2.00443685e-01 -6.33075535e-02 6.41572356e-01 -5.39510965e-01
-1.00832224e+00 -9.74659026e-01 3.30983400e-01 -2.84294009e-01
5.03305972e-01 -5.95366776e-01 -7.27066875e-01 3.66014749e-01
2.62975484e-01 -5.18660247e-01 4.48253572e-01 -2.45883524e-01
-1.95523053e-01 9.84433964e-02 -1.32034457e+00 1.34253740e+00
1.03783011e+00 -5.09354949e-01 -2.34949455e-01 5.95218763e-02
6.95451379e-01 -8.62796307e-01 -5.67528248e-01 -1.74388997e-02
4.72199023e-01 -1.41354585e+00 1.12072217e+00 -3.98270309e-01
8.83220136e-01 -3.15191180e-01 1.41527176e-01 -1.42524755e+00
-7.67203450e-01 -9.50870633e-01 6.78055584e-01 1.09960485e+00
3.33457530e-01 -1.53018996e-01 1.26493514e+00 5.72842836e-01
-1.27589062e-01 -4.44627494e-01 -5.04609704e-01 -3.51149887e-01
1.88493863e-01 -6.03576958e-01 8.41150820e-01 1.01385975e+00
-2.72018492e-01 1.75105389e-02 -4.28905487e-01 -4.85522076e-02
3.93436462e-01 -2.39886105e-01 1.00878775e+00 -6.53960228e-01
-5.65816998e-01 -7.99685657e-01 -6.16343319e-01 -7.11896658e-01
-3.17304134e-02 -7.30792224e-01 -9.18171182e-02 -1.69108999e+00
-1.38485864e-01 -4.37472969e-01 2.05992758e-01 4.18366551e-01
-3.44253108e-02 9.16460037e-01 3.06019545e-01 -1.45080030e-01
-1.41333506e-01 5.70219979e-02 1.73055243e+00 -1.95077598e-01
-5.16159773e-01 -1.04627386e-01 -1.13438404e+00 8.06921124e-01
1.07371831e+00 -2.00980112e-01 -3.76179338e-01 -1.34431869e-01
3.96963656e-01 2.89994985e-01 4.48907644e-01 -1.25479198e+00
-2.31936142e-01 -1.16971284e-01 7.00339854e-01 -1.23129874e-01
1.07383147e-01 -3.93520117e-01 4.30659145e-01 3.74739200e-01
-1.85535684e-01 1.01493485e-01 4.83933419e-01 1.28191262e-01
-1.28151700e-01 -1.81316331e-01 1.12919438e+00 -6.83727860e-01
-1.13626504e+00 3.73430029e-02 -6.44668758e-01 4.68120091e-02
8.67258012e-01 -7.71607101e-01 4.22262624e-02 -8.71209741e-01
-1.12294197e+00 -2.04868108e-01 7.48769820e-01 5.96670330e-01
6.13306224e-01 -1.65118599e+00 -5.83548009e-01 4.44852978e-01
-3.48683476e-01 5.36283292e-02 2.48646691e-01 -2.02163190e-01
-1.24048913e+00 -2.41852075e-01 -8.83906662e-01 -2.78386712e-01
-1.19097638e+00 4.07783657e-01 2.20599890e-01 -3.64179909e-01
-6.42786741e-01 1.01805639e+00 5.27557969e-01 -4.66915965e-01
-1.47007219e-02 -2.65053153e-01 -5.40699735e-02 -6.39413595e-02
4.92590427e-01 2.87069470e-01 -2.41075736e-02 -1.68086171e-01
1.90779448e-01 5.52240014e-01 1.43900573e-01 -4.23589975e-01
1.30205822e+00 1.49127424e-01 -6.35774508e-02 1.63073212e-01
1.05161774e+00 8.59972313e-02 -1.48490107e+00 1.55689001e-01
-4.81515050e-01 -5.19113123e-01 -8.60531181e-02 -7.38765061e-01
-1.14602017e+00 7.32358336e-01 7.65102923e-01 3.19076002e-01
1.29937160e+00 -1.45726725e-01 7.23325551e-01 -1.66270196e-01
5.37295759e-01 -1.25070786e+00 6.55526578e-01 3.73342276e-01
8.50682259e-01 -8.14404905e-01 -3.56790304e-01 -2.59791285e-01
-8.27094078e-01 1.02093685e+00 6.88142836e-01 -1.97467342e-01
3.93077642e-01 7.33839154e-01 2.98384994e-01 -6.60539120e-02
-2.82076865e-01 -2.60030389e-01 2.00782806e-01 1.00463974e+00
1.40662864e-01 4.44478154e-01 8.24579671e-02 4.17204559e-01
-8.45432043e-01 2.82463408e-03 7.72233248e-01 6.85343325e-01
-5.04098415e-01 -1.26958430e+00 -7.09444940e-01 5.11875033e-01
-4.67506051e-01 -2.21299648e-01 -3.07373822e-01 7.66597450e-01
5.32594562e-01 7.17004001e-01 2.36897066e-01 -6.77456796e-01
3.60880852e-01 -1.26006395e-01 6.34595037e-01 -7.84044683e-01
-6.06365502e-01 2.05426261e-01 -5.63911237e-02 -5.80994427e-01
-9.64404643e-02 -3.64205271e-01 -8.83452415e-01 -7.59088695e-01
3.91550064e-01 -2.21205667e-01 5.73794127e-01 4.62342560e-01
2.16549397e-01 9.05641556e-01 4.27828789e-01 -1.15406060e+00
2.62282342e-01 -6.32592559e-01 -7.13450611e-01 6.05038047e-01
-2.66767025e-01 -2.99290180e-01 -8.09908360e-02 4.46389526e-01] | [11.641351699829102, -0.4539114534854889] |
58548c42-3746-4703-be6b-26c322ebca6b | fisheye8k-a-benchmark-and-dataset-for-fisheye | 2305.17449 | null | https://arxiv.org/abs/2305.17449v2 | https://arxiv.org/pdf/2305.17449v2.pdf | FishEye8K: A Benchmark and Dataset for Fisheye Camera Object Detection | With the advance of AI, road object detection has been a prominent topic in computer vision, mostly using perspective cameras. Fisheye lens provides omnidirectional wide coverage for using fewer cameras to monitor road intersections, however with view distortions. To our knowledge, there is no existing open dataset prepared for traffic surveillance on fisheye cameras. This paper introduces an open FishEye8K benchmark dataset for road object detection tasks, which comprises 157K bounding boxes across five classes (Pedestrian, Bike, Car, Bus, and Truck). In addition, we present benchmark results of State-of-The-Art (SoTA) models, including variations of YOLOv5, YOLOR, YOLO7, and YOLOv8. The dataset comprises 8,000 images recorded in 22 videos using 18 fisheye cameras for traffic monitoring in Hsinchu, Taiwan, at resolutions of 1080$\times$1080 and 1280$\times$1280. The data annotation and validation process were arduous and time-consuming, due to the ultra-wide panoramic and hemispherical fisheye camera images with large distortion and numerous road participants, particularly people riding scooters. To avoid bias, frames from a particular camera were assigned to either the training or test sets, maintaining a ratio of about 70:30 for both the number of images and bounding boxes in each class. Experimental results show that YOLOv8 and YOLOR outperform on input sizes 640$\times$640 and 1280$\times$1280, respectively. The dataset will be available on GitHub with PASCAL VOC, MS COCO, and YOLO annotation formats. The FishEye8K benchmark will provide significant contributions to the fisheye video analytics and smart city applications. | ['Fang-Pang Lin', 'Mohammed Abduljabbar', 'Fady Alnajjar', 'Ganzorig Batnasan', 'Hamad Al Jassmi', 'Byambaa Dorj', 'Ping-Yang Chen', 'Ming-Ching Chang', 'Jun-Wei Hsieh', 'Erkhembayar Ganbold', 'Munkh-Erdene Otgonbold', 'Munkhjargal Gochoo'] | 2023-05-27 | null | null | null | null | ['2d-object-detection'] | ['computer-vision'] | [-1.48174375e-01 -3.42239201e-01 -1.77329361e-01 -2.01191440e-01
-6.58245385e-01 -5.42524397e-01 3.50187182e-01 -6.80030704e-01
-5.97248197e-01 6.15748286e-01 -2.75029093e-01 -4.45467383e-01
2.74650007e-01 -8.75437975e-01 -8.31484258e-01 -8.82544339e-01
1.09352924e-01 1.11930802e-01 8.78539443e-01 -1.74747631e-01
-1.67087875e-02 5.01301765e-01 -1.87398696e+00 1.24476150e-01
4.90577132e-01 1.08330166e+00 1.04753293e-01 1.07757306e+00
5.16804516e-01 6.50501013e-01 -3.78089726e-01 -7.76970029e-01
6.12992525e-01 1.77023247e-01 -2.12426722e-01 1.18816629e-01
1.54002476e+00 -7.62804091e-01 -8.87667418e-01 1.05440843e+00
3.22980225e-01 2.82235503e-01 4.75146174e-01 -1.66717613e+00
-4.93390560e-01 -1.05014063e-01 -7.63062477e-01 7.13018775e-01
1.12870976e-01 5.24958193e-01 7.82706678e-01 -1.03885293e+00
4.65903372e-01 1.26775420e+00 4.74670738e-01 3.90646785e-01
-5.54636121e-01 -1.13528311e+00 6.60684183e-02 7.56844997e-01
-1.52721047e+00 -6.52217269e-01 2.11494759e-01 -2.57986724e-01
9.96810734e-01 2.38637403e-01 6.86643362e-01 9.84007657e-01
1.71198919e-01 6.96816802e-01 8.20663512e-01 -5.28771393e-02
-2.32678980e-01 2.30167404e-01 2.90554613e-01 8.32539797e-01
5.33131659e-01 4.32628155e-01 -1.90934852e-01 2.35279873e-01
6.21562362e-01 7.02332631e-02 -2.88598329e-01 -2.78923541e-01
-9.59343016e-01 8.08276653e-01 3.97268891e-01 -5.16017973e-01
1.73315853e-01 8.98346305e-02 4.30541784e-01 3.04694474e-01
2.80586839e-01 -2.20743686e-01 -3.74143302e-01 -2.87953675e-01
-3.65447730e-01 2.74209082e-01 4.53940421e-01 1.60666656e+00
7.45941818e-01 8.82496238e-02 4.60636258e-01 7.18717754e-01
7.89504647e-02 1.19257867e+00 -1.85204133e-01 -1.16438591e+00
1.00225103e+00 3.28462988e-01 9.41641703e-02 -1.08684754e+00
-1.48665935e-01 4.51936871e-01 -5.50753355e-01 4.32277650e-01
7.15769649e-01 -2.09571987e-01 -8.54380190e-01 9.42632556e-01
5.36830246e-01 1.53115600e-01 1.68190878e-02 9.56028759e-01
1.24682498e+00 8.91918361e-01 -8.48732144e-02 1.38040483e-01
1.64515901e+00 -1.28662550e+00 -4.32415575e-01 -1.42916098e-01
6.62258208e-01 -8.01724732e-01 1.01432395e+00 3.61903369e-01
-8.10327649e-01 -6.71538830e-01 -7.68240333e-01 -1.92247674e-01
-6.30990624e-01 2.02734619e-01 9.51326266e-02 1.03822386e+00
-7.93729961e-01 -3.45767975e-01 -3.16518426e-01 -3.64191234e-01
5.53503096e-01 5.92744863e-03 -6.10499799e-01 -5.82425058e-01
-1.15977585e+00 1.04927170e+00 1.02210753e-01 -2.70656552e-02
-9.16755021e-01 -5.55613995e-01 -1.11755085e+00 -2.47825533e-02
7.96869278e-01 1.44806318e-02 9.16864812e-01 -5.63351512e-01
-1.29504991e+00 9.03352201e-01 2.56546307e-02 -5.46752274e-01
4.80324000e-01 -3.82119149e-01 -8.38892937e-01 4.48632270e-01
1.59869507e-01 1.00278091e+00 8.14316332e-01 -7.74017870e-01
-1.42542052e+00 -2.48098344e-01 3.00014347e-01 2.27788150e-01
-1.15013130e-01 3.05540383e-01 -7.60421991e-01 -1.92739621e-01
-6.19268060e-01 -1.20941079e+00 1.24718994e-01 1.66401342e-01
-2.81350017e-01 -3.89488935e-01 1.59025121e+00 -4.80354309e-01
1.00654674e+00 -2.14884591e+00 -5.28561711e-01 -7.15165660e-02
3.31652373e-01 6.77389503e-01 1.51645122e-02 -8.45575109e-02
9.91934463e-02 -1.08854242e-01 1.84095368e-01 1.95765316e-01
-4.10583496e-01 1.64884135e-01 -2.82854438e-01 7.07204521e-01
-2.26584449e-01 6.86906517e-01 -8.45613122e-01 -6.19125366e-01
7.57396877e-01 3.35509092e-01 -3.51571232e-01 6.76926672e-02
4.50767040e-01 -3.37883011e-02 -1.81601152e-01 1.06368315e+00
1.04709899e+00 2.33342469e-01 -6.00453496e-01 -1.85357213e-01
-4.62570876e-01 -2.69784063e-01 -1.13938558e+00 6.49661243e-01
-3.77573282e-01 1.47799993e+00 1.26311511e-01 -6.99016094e-01
8.23343873e-01 1.13576494e-01 2.35732004e-01 -7.37847507e-01
3.92186679e-02 3.54079492e-02 -2.04152301e-01 -6.62830412e-01
7.90509284e-01 5.27738452e-01 -1.91806927e-01 -2.58954391e-02
-1.22792184e-01 -2.38886058e-01 6.98305368e-01 5.45029677e-02
7.81737626e-01 -1.48638189e-01 1.23020418e-01 1.41015127e-01
6.21095121e-01 3.19442928e-01 4.50871557e-01 8.12906682e-01
-6.11836851e-01 6.42223835e-01 2.83638269e-01 -8.98621380e-01
-1.28898406e+00 -1.07409120e+00 -4.83524650e-01 1.02324378e+00
4.00979996e-01 -2.00159803e-01 -8.04343462e-01 -6.50137722e-01
-2.11536914e-01 5.14367819e-01 -2.51804471e-01 1.50934264e-01
-9.86059964e-01 -3.70917648e-01 9.10892427e-01 6.57348990e-01
1.13923812e+00 -8.51083696e-01 -1.02362347e+00 -2.52198905e-01
-2.13026807e-01 -1.68786693e+00 -8.78826141e-01 -4.45031077e-01
-3.72036994e-01 -1.63690269e+00 -8.54905903e-01 -6.60443127e-01
4.72468376e-01 1.21185184e+00 9.43053365e-01 -3.09139878e-01
-4.55553442e-01 4.25267726e-01 -3.97897512e-01 -5.40175974e-01
9.31040570e-02 -2.05022573e-01 5.15737757e-02 -2.00859874e-01
7.68253505e-01 2.74723619e-01 -6.94090366e-01 1.18227386e+00
-4.44187343e-01 -1.49433419e-01 3.41797978e-01 5.69882691e-01
2.06961945e-01 1.81860596e-01 2.60280930e-02 -5.45166433e-01
-2.21984401e-01 -3.75465125e-01 -1.20885468e+00 -5.86655624e-02
-4.44804370e-01 -9.91819739e-01 6.28058076e-01 -1.57387778e-01
-1.12618220e+00 -3.61022130e-02 7.50042871e-02 -6.58142507e-01
-5.99458814e-01 -3.46843362e-01 -2.51191705e-01 -1.40618861e-01
3.78315002e-01 4.26531620e-02 -5.45621254e-02 -2.45136674e-02
5.85316777e-01 9.89219904e-01 7.19812930e-01 -5.30719049e-02
7.21902192e-01 7.92048216e-01 -2.15407863e-01 -1.49384737e+00
-5.26706457e-01 -7.92692721e-01 -4.34095800e-01 -5.99699199e-01
1.08704245e+00 -1.20606756e+00 -7.86251664e-01 9.06122565e-01
-8.29057634e-01 -1.25287428e-01 -3.79895023e-03 6.32347643e-01
-2.71739274e-01 3.18216264e-01 -3.95496935e-01 -5.26335716e-01
-1.41025171e-01 -1.32236910e+00 1.10406089e+00 4.96387064e-01
3.61069411e-01 -6.07935667e-01 -2.68641055e-01 9.08250093e-01
1.73954248e-01 -8.75010118e-02 3.03290308e-01 -1.10140517e-01
-8.14580739e-01 -5.46242714e-01 -6.91295445e-01 5.25063217e-01
-2.09375858e-01 3.29636931e-01 -8.75440478e-01 -9.83621627e-02
-5.58065593e-01 -2.48311937e-01 9.66273785e-01 5.19398391e-01
1.09512210e+00 -1.43109232e-01 -4.24680144e-01 7.38858104e-01
9.72674489e-01 4.80773836e-01 1.02520680e+00 4.42478716e-01
9.39095438e-01 6.60693824e-01 9.70404923e-01 -1.72830988e-02
7.14590371e-01 8.38252425e-01 5.52833319e-01 -6.34358078e-02
-1.74631953e-01 9.40691978e-02 6.69681907e-01 7.74522573e-02
-4.66968358e-01 -5.01709342e-01 -1.00421035e+00 4.94399518e-01
-1.31883132e+00 -1.44635499e+00 -5.77686012e-01 2.26378345e+00
1.79896474e-01 1.83074042e-01 4.05200303e-01 6.32275641e-02
9.21060562e-01 2.26137593e-01 -3.13629448e-01 -3.36391538e-01
-1.18563548e-01 -3.74919474e-01 1.24512041e+00 3.28188241e-01
-1.56123662e+00 9.48928237e-01 5.57760572e+00 1.00254512e+00
-1.07244134e+00 -9.54767391e-02 6.54880285e-01 -1.67174190e-01
4.83669251e-01 -3.14614624e-01 -1.66459548e+00 6.25006080e-01
1.19705200e+00 2.98154235e-01 5.72665110e-02 1.22053516e+00
2.41869405e-01 -5.14554322e-01 -7.56702542e-01 1.14960599e+00
3.74520659e-01 -1.32993782e+00 -1.93271190e-01 2.78647751e-01
6.97994888e-01 4.16005313e-01 8.62595439e-02 4.15523857e-01
2.68137276e-01 -9.15291190e-01 6.49917662e-01 4.58842032e-02
1.04594886e+00 -8.18731964e-01 7.33949482e-01 1.99037522e-01
-1.51886499e+00 -2.77995288e-01 -6.74017489e-01 2.88818032e-01
3.16119283e-01 6.78473618e-03 -5.32494843e-01 1.52530834e-01
1.37469363e+00 9.26881909e-01 -5.47015250e-01 8.71824741e-01
-4.23693582e-02 4.82049197e-01 -4.33730870e-01 3.84948738e-02
4.86897975e-01 -2.90907502e-01 6.84809268e-01 1.30255067e+00
2.90873080e-01 2.13840723e-01 1.14583232e-01 1.07753731e-01
1.30847931e-01 -3.18506986e-01 -1.12398815e+00 6.33174360e-01
6.07021332e-01 1.22906983e+00 -5.79720378e-01 -4.23824638e-01
-1.01416373e+00 3.88918787e-01 -2.23040536e-01 3.08447123e-01
-1.38017333e+00 -6.01736546e-01 7.99588323e-01 3.76907974e-01
7.16149032e-01 -2.58833528e-01 1.74515173e-01 -1.13575816e+00
-1.31417066e-01 -6.65350437e-01 5.10315716e-01 -9.37585771e-01
-7.38615513e-01 5.16623735e-01 6.20806873e-01 -1.55279553e+00
1.76388204e-01 -1.05434239e+00 -5.26589334e-01 3.65536153e-01
-1.61331177e+00 -1.08636367e+00 -8.05844843e-01 7.04994440e-01
1.06469238e+00 -3.49319279e-01 1.87456816e-01 7.38709152e-01
-9.25660551e-01 5.45005143e-01 2.22917333e-01 3.78076136e-01
7.94264376e-01 -9.16355073e-01 3.80584627e-01 8.02199125e-01
-1.75587490e-01 8.16022083e-02 3.42527092e-01 -2.08889216e-01
-1.20875573e+00 -1.41058826e+00 6.72035694e-01 -8.51952553e-01
3.97895098e-01 -1.95832446e-01 -4.95176822e-01 7.44794309e-01
1.11309573e-01 5.90266407e-01 3.10345709e-01 -5.53643584e-01
-2.32175112e-01 -5.48354864e-01 -1.00189519e+00 5.58236241e-01
9.55573678e-01 -1.44032404e-01 -1.79346249e-01 1.42894641e-01
4.63965297e-01 -8.14524531e-01 -5.97930670e-01 9.20665786e-02
8.80206823e-01 -9.32601690e-01 1.24745667e+00 -2.07410082e-01
2.80052871e-01 -3.62767905e-01 -3.29655349e-01 -8.53839815e-01
1.44295275e-01 -1.87450051e-01 2.01444700e-02 9.95789587e-01
2.03830168e-01 -7.32811809e-01 7.55190611e-01 5.66511452e-01
-4.27326769e-01 -4.99120474e-01 -1.25032890e+00 -7.95689404e-01
-7.79080838e-02 -5.36146164e-01 3.62050474e-01 3.76320451e-01
-5.19127071e-01 1.49985313e-01 -7.28015065e-01 1.84812218e-01
7.64669657e-01 -2.35423848e-01 1.46943355e+00 -8.79477203e-01
4.43361878e-01 -2.57293284e-01 -7.27338493e-01 -1.35566342e+00
-6.29797354e-02 -1.81837156e-01 -1.31152913e-01 -8.47054780e-01
1.19496919e-02 -5.00873923e-01 2.88604110e-01 1.78253084e-01
-9.06761922e-03 6.92947030e-01 2.25168616e-01 3.49518031e-01
-6.75630569e-01 2.95576990e-01 1.19976187e+00 -3.62030417e-01
1.17730841e-01 2.07252637e-01 -2.25198686e-01 1.10681534e+00
6.06046796e-01 -2.33490184e-01 -3.22634280e-01 -4.43120778e-01
-2.21208796e-01 -9.02086496e-02 8.11876714e-01 -1.03579211e+00
2.85380542e-01 -2.68937975e-01 2.41440281e-01 -1.15611053e+00
5.75035155e-01 -1.06807566e+00 -6.06163107e-02 3.82313997e-01
1.30110934e-01 2.52759457e-01 1.14675067e-01 7.31012464e-01
-2.27762938e-01 1.36362001e-01 1.06327689e+00 -1.13696009e-01
-1.39533150e+00 4.70111281e-01 -6.00345433e-01 1.45363539e-01
1.66883624e+00 -8.34886789e-01 -8.49802136e-01 -3.15723896e-01
-1.93253800e-01 6.35139585e-01 2.38235086e-01 6.91711068e-01
9.02853370e-01 -1.19957793e+00 -7.23099530e-01 3.83531362e-01
2.96530038e-01 2.06918344e-01 3.96351188e-01 8.62098753e-01
-9.21354651e-01 9.59818006e-01 -3.81800234e-01 -8.80635440e-01
-1.68412840e+00 3.58577669e-01 3.01934898e-01 2.39929631e-01
-9.55248713e-01 5.45954943e-01 4.09024447e-01 -1.82936028e-01
2.48181731e-01 -3.05571496e-01 -4.70746696e-01 1.03256419e-01
7.79812694e-01 1.23095036e+00 -7.17005059e-02 -1.19874477e+00
-4.32203829e-01 9.00115609e-01 -1.49623454e-01 2.65772909e-01
9.05921638e-01 -6.03959680e-01 4.86230612e-01 -1.29428655e-01
1.09063804e+00 -2.53214747e-01 -1.55862069e+00 -8.76702368e-02
-6.66649818e-01 -9.49639320e-01 -1.70127023e-02 -1.02794006e-01
-1.20746648e+00 8.75739574e-01 8.34282041e-01 -3.63266468e-02
7.24331141e-01 7.04567879e-02 1.14297295e+00 6.19103611e-01
4.09803480e-01 -1.20600557e+00 -1.56600729e-01 6.53823078e-01
6.11033916e-01 -1.64478254e+00 -3.46961319e-02 -5.56031048e-01
-8.07404160e-01 1.33694530e+00 1.16214514e+00 -1.06379226e-01
5.12633562e-01 8.03147331e-02 1.63118437e-01 -2.40606084e-01
-5.62060118e-01 -2.19911486e-01 8.83255526e-02 8.68154705e-01
-1.94666862e-01 8.21984485e-02 3.14379692e-01 -2.34771565e-01
-6.00837097e-02 -3.16828787e-01 7.52217054e-01 6.32755756e-01
-6.58175349e-01 -3.97413403e-01 -7.81481266e-01 4.22259033e-01
-2.80258417e-01 7.10930908e-04 1.31391898e-01 1.30384541e+00
3.39309007e-01 1.21018064e+00 3.90260845e-01 -1.84219226e-01
5.70153654e-01 -5.56704879e-01 1.75042734e-01 -1.47602618e-01
-2.22695805e-03 -2.09892079e-01 4.02870268e-01 -5.63561022e-01
-3.22578609e-01 -7.76885748e-01 -7.65731037e-01 -7.96345532e-01
-6.34771943e-01 -1.73780218e-01 3.37774783e-01 6.30173624e-01
3.87652293e-02 2.61802878e-02 7.02335298e-01 -1.05512750e+00
-1.32116079e-01 -7.36824393e-01 -4.72774029e-01 6.02421202e-02
4.26242888e-01 -1.02118278e+00 -4.09394622e-01 3.55734140e-01] | [8.08397102355957, -1.0126968622207642] |
a1950272-71c9-469b-acbe-f7f919ba637a | does-character-level-information-always | 2306.02302 | null | https://arxiv.org/abs/2306.02302v1 | https://arxiv.org/pdf/2306.02302v1.pdf | Does Character-level Information Always Improve DRS-based Semantic Parsing? | Even in the era of massive language models, it has been suggested that character-level representations improve the performance of neural models. The state-of-the-art neural semantic parser for Discourse Representation Structures uses character-level representations, improving performance in the four languages (i.e., English, German, Dutch, and Italian) in the Parallel Meaning Bank dataset. However, how and why character-level information improves the parser's performance remains unclear. This study provides an in-depth analysis of performance changes by order of character sequences. In the experiments, we compare F1-scores by shuffling the order and randomizing character sequences after testing the performance of character-level information. Our results indicate that incorporating character-level information does not improve the performance in English and German. In addition, we find that the parser is not sensitive to correct character order in Dutch. Nevertheless, performance improvements are observed when using character-level information. | ['Hitomi Yanaka', 'Tomoya Kurosawa'] | 2023-06-04 | null | null | null | null | ['semantic-parsing'] | ['natural-language-processing'] | [ 2.19958887e-01 1.59350738e-01 -2.28158608e-01 -2.13747233e-01
-5.78579426e-01 -7.88980067e-01 5.74278593e-01 6.22745216e-01
-9.64721501e-01 4.57853377e-01 8.13170135e-01 -5.88782132e-01
1.66636407e-01 -9.08303320e-01 -5.61168313e-01 -3.06119442e-01
2.03080535e-01 3.80569071e-01 4.13141549e-01 -3.34555268e-01
5.99250019e-01 3.73367459e-01 -1.46558416e+00 9.37228560e-01
6.66517913e-01 1.25455752e-01 3.45939428e-01 6.46413743e-01
-3.92020613e-01 9.05300260e-01 -9.86268759e-01 -3.45343679e-01
-2.25933492e-01 -6.18737280e-01 -1.00649667e+00 -4.73661095e-01
3.42432469e-01 -1.89314842e-01 -3.39215577e-01 9.57429349e-01
1.99312925e-01 6.64689392e-02 5.15562713e-01 -4.83443826e-01
-6.45238757e-01 9.67826307e-01 -2.67419904e-01 5.87993503e-01
4.94073957e-01 9.79677662e-02 1.33147144e+00 -5.51091433e-01
1.02396274e+00 1.69024670e+00 5.03375709e-01 6.39070272e-01
-1.25167298e+00 -2.38305017e-01 3.74356776e-01 2.07833260e-01
-9.82031703e-01 -2.99600333e-01 5.94106376e-01 -3.87813598e-01
1.64365232e+00 4.60876748e-02 4.90552425e-01 9.79223311e-01
1.27313167e-01 7.56125271e-01 1.11222446e+00 -8.26991558e-01
1.99182004e-01 -1.82668194e-01 6.45322025e-01 5.43048263e-01
6.74619675e-01 1.09747373e-01 -6.89069629e-01 -1.28366247e-01
5.82416475e-01 -8.26301873e-01 4.48843688e-02 1.63733676e-01
-1.20361519e+00 1.16470885e+00 3.07616264e-01 8.97993743e-01
-3.52135301e-01 1.45899862e-01 8.71148527e-01 2.62345701e-01
4.09618199e-01 7.85459697e-01 -4.18663621e-01 -4.51086193e-01
-7.80960560e-01 2.54727483e-01 8.68364930e-01 4.20228064e-01
3.02577049e-01 2.75198340e-01 -1.78446740e-01 9.13608432e-01
8.76904428e-02 -1.20236434e-01 8.00584614e-01 -1.02882159e+00
8.74417305e-01 4.39954907e-01 -1.29808351e-01 -1.01283634e+00
-6.50049567e-01 -1.73282996e-01 -2.31817409e-01 -1.85451388e-01
9.25081611e-01 -2.12851658e-01 -6.95102572e-01 2.10224628e+00
-2.99943089e-01 -4.90973979e-01 3.92022967e-01 8.09305489e-01
7.68814087e-01 8.10260057e-01 6.28272474e-01 -2.50624493e-02
1.68310440e+00 -7.44608760e-01 -6.31088793e-01 -7.71086872e-01
1.09185302e+00 -6.33983433e-01 1.18984473e+00 7.51521206e-03
-1.11753142e+00 -6.14916444e-01 -1.09410286e+00 -2.53810525e-01
-2.63010234e-01 6.06314875e-02 7.67568231e-01 9.47242200e-01
-9.19339895e-01 5.67050338e-01 -9.51081455e-01 -4.95275885e-01
6.56126961e-02 -4.57729287e-02 -4.25011128e-01 1.44752655e-02
-1.48713863e+00 1.34624112e+00 9.30302143e-01 -4.14015859e-01
-2.25633353e-01 -3.23298991e-01 -1.04035676e+00 1.99388713e-01
4.20797318e-02 -4.12527353e-01 1.17993855e+00 -1.04388189e+00
-1.37102270e+00 1.03735435e+00 -3.33346665e-01 -6.00606620e-01
7.49695525e-02 -1.76831648e-01 -9.83308107e-02 3.16605657e-01
4.59529124e-02 8.39725196e-01 1.13404438e-01 -9.12780583e-01
-5.27687252e-01 -3.49961758e-01 1.90279141e-01 3.03044289e-01
-1.68931559e-01 4.05102402e-01 1.05697073e-01 -5.85923433e-01
1.06873535e-01 -8.15553844e-01 1.61991443e-03 -4.81228948e-01
4.02229130e-02 -5.86950898e-01 4.15323824e-02 -9.52155173e-01
1.30681813e+00 -2.31446314e+00 -5.53449355e-02 -1.73477739e-01
-2.00706676e-01 2.36265793e-01 -3.79097998e-01 4.06296253e-01
-1.54521316e-01 6.01801038e-01 -2.17814654e-01 7.25783557e-02
-8.73800963e-02 3.84618104e-01 1.06348239e-01 2.95082301e-01
5.70173383e-01 9.33517516e-01 -7.64100015e-01 -3.15662950e-01
4.80250232e-02 2.61758029e-01 -8.79497170e-01 -2.85533667e-01
-2.04620793e-01 8.40024836e-03 -2.11415991e-01 1.14309743e-01
1.76273510e-01 -1.95757560e-02 5.70578158e-01 1.91837013e-01
-2.21083477e-01 1.23308671e+00 -7.91248441e-01 1.64358532e+00
-4.22358841e-01 8.99989426e-01 -1.74838036e-01 -9.97561932e-01
8.67971599e-01 4.47633117e-01 -2.01266348e-01 -1.06546867e+00
2.94817751e-03 3.90877157e-01 1.04261744e+00 -2.87601084e-01
7.91188836e-01 -3.90561730e-01 -4.87979680e-01 3.46895635e-01
-1.77557662e-01 5.82391247e-02 5.42925000e-01 8.69207233e-02
1.14291227e+00 3.59741435e-03 3.11096072e-01 -4.17615652e-01
2.33498052e-01 3.12739402e-01 5.94751656e-01 7.10684180e-01
-1.72829553e-02 6.01210356e-01 9.35002744e-01 -2.96757191e-01
-1.24346042e+00 -7.56243467e-01 -1.66470528e-01 1.30694759e+00
-3.33568186e-01 -6.13333881e-01 -1.00053906e+00 -3.39274794e-01
-1.24264903e-01 1.37909496e+00 -5.78250349e-01 -1.84711382e-01
-1.10080326e+00 -9.18354928e-01 1.05347824e+00 7.03576207e-01
2.97758371e-01 -1.34228230e+00 -1.06006229e+00 3.69264811e-01
-3.54664952e-01 -1.16468585e+00 1.84190616e-01 2.95193225e-01
-1.01156509e+00 -8.28765213e-01 -4.89653111e-01 -7.15842426e-01
3.56548250e-01 -1.25381947e-01 1.25614250e+00 1.69127554e-01
1.43115625e-01 -1.21128783e-01 -5.04117787e-01 -2.09364533e-01
-7.65133381e-01 5.30749142e-01 -4.75898713e-01 -8.08885336e-01
4.62923586e-01 -4.27267030e-02 -4.18076180e-02 -1.74016356e-01
-7.40721166e-01 -8.55919570e-02 4.90192682e-01 8.29895973e-01
1.83970898e-01 -3.40436697e-01 6.56603694e-01 -1.26110518e+00
9.89922464e-01 -3.76746118e-01 -1.83187053e-01 -1.38224419e-02
-1.79024860e-01 3.25533897e-01 4.53579873e-01 -2.62458920e-01
-1.10768211e+00 -3.41393471e-01 -4.57901388e-01 4.01442319e-01
-2.73288310e-01 8.75504971e-01 1.25006065e-01 5.86708784e-01
8.97287488e-01 -1.20714314e-01 -7.96092153e-02 -4.47176099e-01
2.15233415e-01 3.83562446e-01 3.02747250e-01 -8.57844353e-01
8.61743242e-02 -1.68915257e-01 -3.62749428e-01 -1.06734443e+00
-7.08941579e-01 -2.37497855e-02 -5.38692772e-01 2.93353349e-01
1.16437709e+00 -8.07408690e-01 -2.60321945e-01 2.52910823e-01
-1.56407094e+00 -4.49064761e-01 -3.01665336e-01 5.06709218e-01
-4.30172354e-01 4.42957491e-01 -9.87508655e-01 -4.20926571e-01
-3.48807275e-02 -1.15081143e+00 4.99787748e-01 1.28676444e-01
-8.42879713e-01 -1.02873516e+00 -1.23493493e-01 1.82280421e-01
3.23842794e-01 9.67755616e-02 1.59050393e+00 -9.69940126e-01
-5.32638691e-02 -7.10456911e-03 -2.14083329e-01 2.01977521e-01
-1.71756044e-01 -2.33207300e-01 -8.46977115e-01 9.48811322e-03
-6.64696842e-02 -2.55738735e-01 9.94434118e-01 2.76723921e-01
8.53296757e-01 -1.25604019e-01 2.80605257e-02 1.25732541e-01
1.17091382e+00 2.10522503e-01 8.62535059e-01 8.23824883e-01
3.13183993e-01 9.32276130e-01 2.95913339e-01 -9.98587087e-02
4.51594979e-01 3.61765087e-01 4.50466312e-02 2.87936419e-01
-3.49786699e-01 -2.33068064e-01 3.84597868e-01 6.69896960e-01
8.93442333e-02 -4.94731724e-01 -1.05099964e+00 7.50832617e-01
-1.54098105e+00 -8.54223430e-01 -2.05996603e-01 2.00420618e+00
9.27010715e-01 5.34980118e-01 -1.58841833e-02 2.41411164e-01
6.80643678e-01 4.97308850e-01 -1.46838397e-01 -1.15690958e+00
-2.70209879e-01 3.23991776e-01 4.53511477e-01 5.23416579e-01
-6.60097361e-01 1.38348162e+00 6.81605148e+00 4.34453398e-01
-9.71160889e-01 7.16488585e-02 5.74067414e-01 1.19996838e-01
-5.00645876e-01 7.71520436e-02 -6.47623658e-01 3.86833280e-01
1.53566146e+00 -1.12922959e-01 2.04605028e-01 6.81325197e-01
-1.01666257e-01 -4.64175403e-01 -1.01358891e+00 4.28642869e-01
1.43154368e-01 -1.41000569e+00 1.06143773e-01 -2.07401618e-01
3.16912860e-01 8.82017612e-02 -5.10318160e-01 3.47720802e-01
2.21758679e-01 -9.92989600e-01 8.76569808e-01 8.53307471e-02
4.75911498e-01 -7.09512770e-01 8.81171882e-01 4.38554704e-01
-6.38658762e-01 1.82676867e-01 -4.99985486e-01 -4.52497184e-01
1.68566316e-01 -1.72435809e-02 -6.16279900e-01 2.00255588e-01
3.27067375e-01 2.62319207e-01 -8.21458757e-01 7.06565320e-01
-5.47644556e-01 1.12791502e+00 -2.78803766e-01 -2.83417374e-01
3.93910587e-01 2.09659204e-01 3.57069641e-01 1.64812720e+00
-1.37117103e-01 1.18067294e-01 -1.07717067e-01 5.18710077e-01
3.39332111e-02 2.01979309e-01 -4.73177969e-01 -4.13166016e-01
5.70408344e-01 5.93432605e-01 -1.01341379e+00 -3.50097507e-01
-5.45437217e-01 8.05410504e-01 5.24710119e-01 1.10117964e-01
-3.67749035e-01 -4.12454963e-01 5.35746872e-01 2.24370971e-01
2.02877045e-01 -5.01596212e-01 -6.74837947e-01 -7.94713855e-01
-8.74623209e-02 -9.03212786e-01 4.36908066e-01 -7.13240921e-01
-1.00975311e+00 5.91830969e-01 2.32775167e-01 -2.53425837e-01
-6.09877169e-01 -9.61847126e-01 -4.34604347e-01 1.01802981e+00
-1.15199423e+00 -5.24484158e-01 1.96150333e-01 -3.19186524e-02
6.76466227e-01 -1.37331381e-01 1.08484805e+00 9.74861626e-03
-2.88249403e-01 5.77974439e-01 -1.29794613e-01 6.59319997e-01
4.06851470e-01 -1.07738066e+00 9.95057762e-01 7.22967684e-01
2.57568955e-01 7.91369200e-01 7.37447917e-01 -7.13115931e-01
-9.36754823e-01 -3.41196120e-01 1.36465824e+00 -1.64809465e-01
6.76585555e-01 -1.25363275e-01 -1.19474435e+00 7.27157652e-01
4.47581589e-01 -3.55888128e-01 6.58252418e-01 3.38222414e-01
-5.72641611e-01 6.17994487e-01 -9.65063155e-01 5.35537660e-01
8.73338938e-01 -6.67468309e-01 -1.26934826e+00 -1.91979879e-03
7.06416428e-01 -3.39735895e-01 -6.52663887e-01 1.67372242e-01
4.79581833e-01 -8.00216436e-01 8.19716036e-01 -7.59967566e-01
6.75716758e-01 -1.69584248e-02 -2.87836015e-01 -1.38017809e+00
-5.43673217e-01 1.07121073e-01 4.31868374e-01 1.02472663e+00
8.08265626e-01 -7.09885597e-01 7.24581003e-01 4.64638114e-01
-3.46581191e-01 -3.45637590e-01 -1.06576538e+00 -6.45999670e-01
7.59910047e-01 -4.15205270e-01 3.04097056e-01 7.55337536e-01
1.50293604e-01 6.42414093e-01 3.47109407e-01 -3.05452496e-01
9.12136286e-02 -1.43947065e-01 9.56005454e-02 -1.34423959e+00
-3.09629023e-01 -6.71599567e-01 -3.57961833e-01 -6.55328572e-01
5.92489541e-01 -1.07301807e+00 3.05241067e-02 -2.01902628e+00
-4.10708319e-03 -2.34245434e-01 2.90749781e-02 5.01595199e-01
-3.13769281e-01 -1.08439304e-01 6.56580329e-01 3.81277688e-02
-1.18554696e-01 1.59262136e-01 8.93518269e-01 1.80160761e-01
-2.05871314e-02 -4.77368265e-01 -7.73288727e-01 8.54997456e-01
1.23724604e+00 -6.27405405e-01 -1.73723623e-01 -1.00430894e+00
3.09240639e-01 -1.56893600e-02 1.06061064e-01 -8.87306035e-01
-2.42492668e-02 -4.96002212e-02 3.90539765e-01 -3.39667916e-01
2.90507346e-01 -1.55726567e-01 -3.74972194e-01 6.88618124e-01
-7.22895265e-01 6.78134918e-01 8.18084538e-01 1.01993680e-01
-2.24838823e-01 -8.50165725e-01 8.17524374e-01 -5.84102273e-01
-9.25837874e-01 -8.14100981e-01 -1.04301977e+00 4.90647554e-01
4.62889880e-01 -3.67463291e-01 -5.90359807e-01 -7.92422425e-03
-4.45470393e-01 -1.78584307e-01 4.91743088e-01 6.09371006e-01
3.30139309e-01 -9.54358459e-01 -9.28972006e-01 1.35592714e-01
9.79450271e-02 -3.43309224e-01 -1.13705941e-01 2.60999590e-01
-8.97086740e-01 6.06489837e-01 -4.01375324e-01 -2.12716103e-01
-1.22980177e+00 1.45477071e-01 2.21763849e-01 -4.20131385e-01
-4.13829297e-01 8.19646895e-01 -5.10781258e-02 -3.91425610e-01
-2.09246818e-02 -4.93471444e-01 -4.93012309e-01 4.07893330e-01
4.11541402e-01 2.33042076e-01 6.89394400e-02 -7.75457919e-01
-4.99926269e-01 2.57354587e-01 -2.51611352e-01 -4.98897254e-01
1.22791851e+00 2.82143205e-01 -1.32025257e-01 7.58387923e-01
9.39500093e-01 -9.06218812e-02 -5.86800456e-01 1.54588401e-01
5.46662569e-01 -8.67344737e-02 -2.15368513e-02 -7.59223998e-01
-6.12895131e-01 9.73747671e-01 2.36036807e-01 8.08996484e-02
5.46751201e-01 -8.06640834e-03 5.60719132e-01 4.26202834e-01
2.12584704e-01 -1.20610511e+00 -8.40801224e-02 1.18482196e+00
6.66900098e-01 -6.43218338e-01 -1.40316322e-01 -3.29804599e-01
-6.71725512e-01 9.81449544e-01 6.13717914e-01 -3.67816955e-01
1.12255745e-01 1.86790586e-01 -1.26120478e-01 -1.12611562e-01
-8.01908493e-01 -1.02207437e-01 -1.26767531e-01 3.85464251e-01
1.18338728e+00 2.35903770e-01 -9.75406826e-01 5.10964930e-01
-6.36678100e-01 -4.32344377e-01 6.14358664e-01 1.04026389e+00
-7.02274323e-01 -1.21573663e+00 -4.51647043e-01 4.38237131e-01
-6.86346889e-01 -4.91407752e-01 -6.16167843e-01 1.05153370e+00
-1.15785211e-01 9.71417487e-01 6.01567030e-01 -3.74082569e-03
4.26882356e-01 7.00953245e-01 4.30840224e-01 -1.02306533e+00
-9.81758952e-01 -1.81852430e-01 7.91620195e-01 -2.35509589e-01
-1.81269422e-01 -9.07486975e-01 -1.72708821e+00 -2.80878961e-01
-8.95866901e-02 2.97661692e-01 6.76900029e-01 9.22167718e-01
1.98432922e-01 7.17901111e-01 -2.57136494e-01 -4.02399838e-01
-4.34256196e-01 -1.23899591e+00 -7.37994239e-02 3.67879421e-01
-6.95942268e-02 -4.12597746e-01 -2.37078965e-01 -1.93845436e-01] | [10.66201114654541, 9.35583782196045] |
826b1a55-5243-43d1-96ae-a98e6ffd220f | byzantine-robust-loopless-stochastic-variance | 2303.04560 | null | https://arxiv.org/abs/2303.04560v1 | https://arxiv.org/pdf/2303.04560v1.pdf | Byzantine-Robust Loopless Stochastic Variance-Reduced Gradient | Distributed optimization with open collaboration is a popular field since it provides an opportunity for small groups/companies/universities, and individuals to jointly solve huge-scale problems. However, standard optimization algorithms are fragile in such settings due to the possible presence of so-called Byzantine workers -- participants that can send (intentionally or not) incorrect information instead of the one prescribed by the protocol (e.g., send anti-gradient instead of stochastic gradients). Thus, the problem of designing distributed methods with provable robustness to Byzantine workers has been receiving a lot of attention recently. In particular, several works consider a very promising way to achieve Byzantine tolerance via exploiting variance reduction and robust aggregation. The existing approaches use SAGA- and SARAH-type variance-reduced estimators, while another popular estimator -- SVRG -- is not studied in the context of Byzantine-robustness. In this work, we close this gap in the literature and propose a new method -- Byzantine-Robust Loopless Stochastic Variance Reduced Gradient (BR-LSVRG). We derive non-asymptotic convergence guarantees for the new method in the strongly convex case and compare its performance with existing approaches in numerical experiments. | ['Eduard Gorbunov', 'Nikita Fedin'] | 2023-03-08 | null | null | null | null | ['distributed-optimization'] | ['methodology'] | [-3.33605140e-01 -5.35001792e-02 1.42623544e-01 -1.56615451e-01
-1.21012735e+00 -6.48749530e-01 1.97775319e-01 2.49634370e-01
-4.54303116e-01 1.15930057e+00 2.54677702e-02 -6.31654114e-02
-4.75674540e-01 -5.60093939e-01 -8.08503389e-01 -1.05856502e+00
-2.03562587e-01 3.94487321e-01 -1.96546931e-02 -1.43935144e-01
2.22313672e-01 1.69410914e-01 -9.09461915e-01 -3.48986924e-01
1.25103736e+00 7.61143625e-01 -1.73423901e-01 6.07792258e-01
2.71539271e-01 4.31635559e-01 -6.48472667e-01 -4.08199072e-01
7.81459987e-01 -5.86974561e-01 -4.09205645e-01 5.51346838e-02
3.42535079e-01 -1.71698898e-01 2.61844099e-01 1.40583003e+00
7.97717154e-01 3.57580066e-01 1.71127394e-01 -1.36276639e+00
-3.59219998e-01 8.52517366e-01 -1.00140131e+00 -9.37469006e-02
1.63788170e-01 1.08873323e-01 8.42221141e-01 -6.96229458e-01
5.98570049e-01 1.22958672e+00 7.69036412e-01 3.91162217e-01
-1.16173828e+00 -5.07562220e-01 1.14755400e-01 -2.25073919e-01
-1.41521251e+00 -2.27401659e-01 6.16797626e-01 -3.11920434e-01
1.12804607e-01 8.31768632e-01 1.37497991e-01 7.82535911e-01
3.66169177e-02 5.29341877e-01 1.43796003e+00 -9.21647176e-02
4.62118089e-01 2.93135494e-01 1.53458342e-01 4.94579464e-01
7.57955670e-01 -1.55671701e-01 -5.23683131e-01 -8.02383840e-01
5.06364644e-01 1.37562886e-01 -5.10503352e-01 -3.71547133e-01
-1.23945045e+00 8.62251580e-01 4.08952534e-01 1.64116651e-01
-5.96887529e-01 2.29333773e-01 2.88539469e-01 4.87372309e-01
1.04552066e+00 2.71799654e-01 -5.25417924e-01 -2.09941790e-01
-9.24090624e-01 4.55360651e-01 1.01034164e+00 6.17708027e-01
7.73310542e-01 -8.32205936e-02 -3.34043205e-01 5.15787959e-01
3.54420960e-01 6.01029396e-01 1.53070256e-01 -9.33979630e-01
8.06965590e-01 2.87865847e-01 6.34375155e-01 -1.20618498e+00
-2.18660593e-01 -5.93650103e-01 -1.28703105e+00 7.93682784e-02
7.96594441e-01 -6.43628240e-01 1.31040007e-01 1.71329546e+00
1.01698983e+00 2.20377266e-01 -2.38276348e-01 1.29229009e+00
-1.08158335e-01 5.63243151e-01 -6.19949520e-01 -6.60394847e-01
8.85713816e-01 -1.06346905e+00 -7.56610870e-01 2.26951942e-01
5.81140220e-01 -4.80922222e-01 7.59914815e-01 5.64636588e-01
-1.09785116e+00 4.22461294e-02 -6.86779261e-01 4.02481914e-01
-2.74777506e-02 -2.66834855e-01 2.11453706e-01 1.14541161e+00
-1.04510963e+00 7.38021314e-01 -6.17651165e-01 -7.86625817e-02
4.67485249e-01 3.29846174e-01 -2.41905004e-01 -1.50128752e-02
-7.44894445e-01 4.23869371e-01 -1.70936018e-01 3.99773300e-01
-7.53419876e-01 -9.01978016e-01 -2.90059030e-01 -1.81022827e-02
8.76202464e-01 -7.43387818e-01 1.03301799e+00 -9.38694358e-01
-1.63640344e+00 3.36910158e-01 1.18196562e-01 -4.34937984e-01
1.42376757e+00 -2.98793584e-01 3.31726432e-01 -4.64503318e-02
7.00427443e-02 -4.02195632e-01 9.90445495e-01 -1.05979490e+00
-4.85010207e-01 -8.62327814e-01 -3.56838480e-02 4.44919467e-02
-6.50024951e-01 2.83454627e-01 2.97256410e-01 -5.65595329e-01
-2.06315920e-01 -1.04211783e+00 -6.61062360e-01 -1.97282489e-02
-7.08505213e-01 -2.25066125e-01 6.87710702e-01 -5.57231247e-01
1.15471315e+00 -2.04690266e+00 6.89272359e-02 3.48049194e-01
5.06934941e-01 2.28588417e-01 6.13777675e-02 6.23805702e-01
3.76579195e-01 4.17307556e-01 -2.99858719e-01 -5.70345640e-01
1.93443432e-01 -2.09624067e-01 -1.95638999e-01 1.25490236e+00
-5.16914606e-01 4.62193012e-01 -1.12977004e+00 -2.22586349e-01
-1.05893001e-01 3.21621209e-01 -4.71581340e-01 1.41043067e-01
-7.44424984e-02 7.42655516e-01 -7.99243271e-01 2.54143864e-01
9.93949771e-01 -2.51043171e-01 2.28016734e-01 3.89350206e-01
-3.71999681e-01 -3.18572283e-01 -1.77715433e+00 1.20068443e+00
-5.60907662e-01 2.82445431e-01 9.14453745e-01 -9.26461935e-01
7.53230631e-01 3.36104155e-01 6.29338086e-01 1.23375922e-01
8.33622143e-02 7.45423317e-01 -2.47626781e-01 -1.64939314e-01
2.93841422e-01 3.94713655e-02 -3.59383896e-02 7.68885314e-01
-6.79455638e-01 -1.39644435e-02 2.01816246e-01 2.35989720e-01
1.10200596e+00 -2.65149385e-01 1.74284682e-01 -4.83946800e-01
7.17034042e-01 -5.60228765e-01 7.56159484e-01 1.24087596e+00
-3.14307898e-01 5.18417776e-01 4.86302912e-01 -1.00890778e-01
-7.62426853e-01 -6.45953774e-01 1.70055225e-01 1.15229523e+00
6.15607351e-02 -4.23689365e-01 -8.86896372e-01 -7.86262989e-01
2.74863213e-01 1.63873792e-01 -4.70001340e-01 2.64474481e-01
-2.89040536e-01 -8.08438182e-01 4.09931153e-01 -3.96085158e-02
6.24714315e-01 -4.14283395e-01 -6.10843658e-01 2.50481695e-01
-1.85211569e-01 -8.09077084e-01 -8.83775294e-01 -3.27054739e-01
-8.60980988e-01 -9.90415394e-01 -1.17450130e+00 -7.85585791e-02
8.14780474e-01 4.14883286e-01 6.64861560e-01 5.92207946e-02
-4.76304367e-02 5.00653386e-01 -1.93905458e-01 -4.87673372e-01
-1.47170991e-01 -1.25195503e-01 2.02382028e-01 8.43776166e-01
-5.10318995e-01 -4.36251849e-01 -7.40525544e-01 5.57309866e-01
-7.50621080e-01 -4.65506107e-01 4.04108204e-02 8.29181910e-01
3.76432270e-01 -1.01747662e-01 7.56148875e-01 -1.14156520e+00
1.02187026e+00 -6.03408575e-01 -1.13580418e+00 1.71522081e-01
-6.72845244e-01 -5.30052632e-02 9.38751698e-01 -3.96369725e-01
-1.07992387e+00 -1.95576966e-01 4.25807595e-01 -2.92795956e-01
4.10674572e-01 4.71820265e-01 -3.77280638e-02 -4.47626889e-01
6.34910226e-01 -5.13655059e-02 1.88850492e-01 -4.93512064e-01
5.01335502e-01 8.01351070e-01 -4.36131284e-03 -6.82163060e-01
8.94858897e-01 7.86558867e-01 2.04001337e-01 -6.61944509e-01
-7.37086475e-01 -5.20672023e-01 2.03067616e-01 -1.88586727e-01
2.38715008e-01 -6.37048185e-01 -9.41311061e-01 7.31550276e-01
-1.12840307e+00 -2.94885963e-01 -4.44237262e-01 4.46612686e-01
-3.51839393e-01 6.58208072e-01 -5.41482270e-01 -1.29574037e+00
-5.07564008e-01 -1.19200015e+00 7.36777067e-01 3.83480459e-01
2.13724285e-01 -1.07854259e+00 2.65987337e-01 4.20408756e-01
9.32490110e-01 5.20380676e-01 -9.20791104e-02 -8.14294636e-01
-6.34057462e-01 -4.24088925e-01 -1.05404556e-01 3.76880050e-01
-7.09928721e-02 -1.87972292e-01 -6.10419095e-01 -6.26642764e-01
3.83877605e-01 -2.29386106e-01 4.90139514e-01 4.49699193e-01
1.08484066e+00 -8.81512105e-01 -1.93700507e-01 4.73694772e-01
1.42623210e+00 -7.27082193e-01 2.07518354e-01 1.40019298e-01
4.34114665e-01 5.57116151e-01 6.80641890e-01 1.22906399e+00
2.65422761e-01 6.21800125e-01 4.61182952e-01 2.46341020e-01
4.05147702e-01 2.12729834e-02 3.75771612e-01 6.69391930e-01
-2.34246343e-01 -3.29629779e-01 -4.74469781e-01 4.26471889e-01
-2.32565665e+00 -7.65067697e-01 -6.68810904e-01 2.62376761e+00
9.62305784e-01 -4.17653590e-01 3.33679676e-01 3.52247898e-03
1.13933575e+00 1.41160831e-01 -2.43836418e-01 -2.43887350e-01
-2.07771122e-01 -2.72779495e-01 7.91855514e-01 4.93001044e-01
-6.75239801e-01 2.31081516e-01 4.71862936e+00 1.04057693e+00
-7.83695281e-01 6.88264132e-01 6.62473440e-01 -9.95437428e-02
-1.66348651e-01 2.35767066e-01 -5.99096537e-01 8.14377248e-01
8.33909810e-01 -5.38818359e-01 6.48207068e-01 8.11533153e-01
6.37397885e-01 -2.48665363e-01 -7.48114944e-01 1.07681799e+00
-2.14069765e-02 -1.12312675e+00 -7.20381320e-01 3.42701972e-01
1.36229432e+00 -8.56444761e-02 1.41937081e-02 -1.75088212e-01
3.32958817e-01 -4.61232483e-01 3.72811168e-01 6.13878965e-01
3.07473451e-01 -7.43600726e-01 7.25170434e-01 6.30564690e-01
-8.30783606e-01 -2.84708619e-01 -3.60903770e-01 -1.50342643e-01
3.07183266e-01 1.31406665e+00 -3.12091738e-01 8.05452704e-01
5.68045020e-01 3.39559674e-01 -2.35932663e-01 1.43835902e+00
-2.14887694e-01 7.38545060e-01 -7.51996160e-01 -3.52228343e-01
1.18978962e-01 -6.33178830e-01 1.09961414e+00 8.35114360e-01
4.82213169e-01 -1.67576432e-01 2.78023750e-01 6.91692591e-01
-3.02775174e-01 4.39909399e-01 -4.88850683e-01 2.56324440e-01
3.80558133e-01 1.32718515e+00 -5.99670529e-01 -2.79089481e-01
-1.93948358e-01 8.39740932e-01 1.78060681e-01 4.02634978e-01
-6.78885102e-01 -5.35441875e-01 4.78146613e-01 1.80278998e-02
1.09176755e-01 -2.36532494e-01 -3.39247584e-01 -1.18162882e+00
4.28645134e-01 -1.00058734e+00 4.22472268e-01 -1.50654331e-01
-1.60708797e+00 9.68700349e-02 -2.88121969e-01 -7.94804335e-01
-3.97155136e-02 -1.45901144e-02 -6.02231443e-01 7.55156934e-01
-1.22406983e+00 -8.08062673e-01 -1.52506083e-01 7.29674399e-01
2.10743785e-01 6.20115735e-02 3.03168923e-01 3.19992781e-01
-6.33894265e-01 6.34613454e-01 8.74092817e-01 -3.40857744e-01
8.68030310e-01 -1.27380610e+00 -2.76402831e-01 9.64664698e-01
-1.56383201e-01 7.08196342e-01 1.03255343e+00 -6.09795511e-01
-1.83062792e+00 -1.00893569e+00 7.78407097e-01 -3.96867305e-01
8.39014709e-01 -4.29990441e-01 -6.61809981e-01 4.36062843e-01
1.53189197e-01 4.95761216e-01 2.64364064e-01 7.21812248e-02
-1.20519616e-01 -5.89927554e-01 -1.39656603e+00 4.80944037e-01
7.60673821e-01 -1.73941910e-01 8.09304193e-02 8.85564327e-01
4.52871054e-01 -2.96896279e-01 -6.03643417e-01 -4.05838899e-02
1.36403993e-01 -1.07221949e+00 6.41827703e-01 -2.65237004e-01
-1.38955086e-01 -2.70282924e-01 1.47987038e-01 -1.42260540e+00
3.24369192e-01 -1.65726876e+00 -1.64197743e-01 1.35486448e+00
4.88520153e-02 -1.29939198e+00 5.65451920e-01 7.16679990e-01
2.21815363e-01 -5.67269921e-01 -1.30412245e+00 -1.17578495e+00
-5.96991181e-03 -1.57957003e-01 4.38919395e-01 8.51596415e-01
-1.44296149e-02 -1.62485391e-01 -8.11421037e-01 1.88778579e-01
1.09051657e+00 7.71665275e-02 1.17895675e+00 -9.18040335e-01
-4.44640696e-01 -3.61796021e-01 -1.18411459e-01 -9.15239036e-01
-3.16716079e-03 -6.04602098e-01 2.98441261e-01 -1.12746453e+00
1.28443331e-01 -5.53819835e-01 -1.55682132e-01 1.23145692e-01
-2.77898073e-01 1.33961454e-01 4.37393695e-01 1.52057096e-01
-1.01498497e+00 5.48713505e-01 1.09342504e+00 1.00934230e-01
-3.95268559e-01 4.64811832e-01 -7.21333683e-01 3.89295191e-01
6.11167192e-01 -8.64265382e-01 -2.17615187e-01 -1.56341240e-01
6.56713724e-01 3.45022738e-01 4.85781848e-01 -7.10907340e-01
4.08498883e-01 -3.25967401e-01 -5.85233092e-01 -1.72029659e-01
-1.85252145e-01 -7.40541220e-01 2.38826275e-01 4.30786252e-01
-3.74456584e-01 -1.91945955e-02 -5.40475726e-01 1.06029904e+00
1.20702706e-01 -2.55806893e-01 5.41139662e-01 -5.69799878e-02
3.53427619e-01 3.17475736e-01 -2.52233177e-01 4.54160720e-01
1.08544087e+00 2.24378109e-01 -4.45873529e-01 -6.78436458e-01
-2.59325892e-01 5.08640051e-01 3.18053544e-01 -1.96452960e-02
4.00258482e-01 -9.12862778e-01 -1.13296974e+00 -2.77416915e-01
-3.85271698e-01 -5.13693616e-02 4.16395962e-01 1.51838946e+00
-2.92857796e-01 -3.39555554e-02 4.26551491e-01 -3.24635178e-01
-1.05466223e+00 4.26790237e-01 2.14426711e-01 -4.82725263e-01
-4.38279957e-01 8.96785438e-01 -5.83922975e-02 -5.94307661e-01
1.74178019e-01 7.79902413e-02 5.96879840e-01 -3.18253711e-02
5.48654556e-01 1.04841626e+00 4.75544669e-02 -1.85986638e-01
-3.36785585e-01 2.20217526e-01 3.02604377e-01 -2.68934160e-01
1.38984096e+00 -5.33541799e-01 -5.51284492e-01 4.14060265e-01
1.18009698e+00 5.82675517e-01 -1.20724714e+00 -1.68124601e-01
-2.98203249e-02 -8.12575817e-01 -1.08351909e-01 -4.93114054e-01
-1.17004919e+00 6.31535470e-01 4.20555472e-01 5.47998548e-01
7.80057549e-01 -3.97999883e-01 7.70001531e-01 3.91769767e-01
6.93515241e-01 -1.39269483e+00 4.91189435e-02 2.65748143e-01
7.90686488e-01 -1.14607215e+00 1.64330661e-01 -3.46592486e-01
-3.57080996e-01 8.33076835e-01 2.47514948e-01 -4.78775620e-01
6.81763589e-01 -5.34748957e-02 -1.29634485e-01 1.65468365e-01
-4.87369895e-01 9.39651355e-02 -1.32313132e-01 3.09203923e-01
1.86295509e-02 1.96702287e-01 -1.01384568e+00 6.78986907e-01
2.65948653e-01 2.85640340e-02 8.58691335e-01 8.12018454e-01
-9.63060558e-02 -9.59702313e-01 -8.12981069e-01 4.18261737e-01
-7.52531648e-01 2.26948291e-01 -2.75129467e-01 3.35024446e-01
-2.31878087e-01 1.40169227e+00 -6.41726494e-01 1.71643972e-01
-3.33496998e-03 -4.19971228e-01 8.34785551e-02 -2.88990557e-01
-1.16266572e+00 1.41582474e-01 -1.19961224e-01 -6.50919795e-01
-4.46714848e-01 -5.90219259e-01 -7.23114967e-01 -4.68004376e-01
-6.79523945e-01 5.94064593e-01 9.06011164e-01 8.85330260e-01
5.32561123e-01 1.34779629e-03 1.25604105e+00 -7.25606561e-01
-1.49371672e+00 -7.38513947e-01 -9.95177209e-01 3.46778125e-01
2.93261349e-01 -2.23070383e-01 -8.80822182e-01 -5.06524086e-01] | [6.293889045715332, 4.888726711273193] |
0352dd1f-b338-4277-9aa0-968a2f7f61e5 | joint-communication-and-computation-design-in | 2210.15399 | null | https://arxiv.org/abs/2210.15399v1 | https://arxiv.org/pdf/2210.15399v1.pdf | Joint Communication and Computation Design in Transmissive RMS Transceiver Enabled Multi-Tier Computing Networks | In this paper, a novel transmissive reconfigurable meta-surface (RMS) transceiver enabled multi-tier computing network architecture is proposed for improving computing capability, decreasing computing delay and reducing base station (BS) deployment cost, in which transmissive RMS equipped with a feed antenna can be regarded as a new type of multi-antenna system. We formulate a total energy consumption minimization problem by a joint optimization of subcarrier allocation, task input bits, time slot allocation, transmit power allocation and RMS transmissive coefficient while taking into account the constraints of communication resources and computing resources. This formulated problem is a non-convex optimization problem due to the high coupling of optimization variables, which is NP-hard to obtain its optimal solution. To address the above challenging problems, block coordinate descent (BCD) technique is employed to decouple the optimization variables to solve the problem. Specifically, the joint optimization problem of subcarrier allocation, task input bits, time slot allocation, transmit power allocation and RMS transmissive coefficient is divided into three subproblems to solve by applying BCD. Then, the decoupled three subproblems are optimized alternately by using successive convex approximation (SCA) and difference-convex (DC) programming until the convergence is achieved. Numerical results verify that our proposed algorithm is superior in reducing total energy consumption compared to other benchmarks. | ['Jianmin Lu', 'Hongying Tang', 'Ziwei Liu', 'Wen Chen', 'Zhendong Li'] | 2022-10-27 | null | null | null | null | ['total-energy'] | ['miscellaneous'] | [ 3.64072561e-01 -2.00259343e-01 -1.14683837e-01 1.25220031e-01
-4.75978911e-01 -5.24001002e-01 -3.24383788e-02 -4.35500741e-01
-3.78271848e-01 1.02322185e+00 -2.75213957e-01 -4.58539903e-01
-6.20411396e-01 -7.18351960e-01 -3.29891950e-01 -1.29359984e+00
-6.12405241e-02 -1.01258412e-01 -1.95356756e-01 -6.84107393e-02
8.02603960e-02 4.78250891e-01 -1.08783960e+00 -5.00036359e-01
1.01716864e+00 1.22491086e+00 3.93338680e-01 2.78408766e-01
3.58687371e-01 1.80822060e-01 -5.41473806e-01 2.37806235e-02
3.16503078e-01 -3.54681939e-01 -3.45711291e-01 7.02837780e-02
-5.32771051e-01 -1.52319252e-01 -4.30807322e-02 9.43321168e-01
7.04003930e-01 3.82312030e-01 6.32239282e-01 -1.37704551e+00
-1.39656842e-01 3.18160981e-01 -1.09102607e+00 2.46447116e-01
-2.32986525e-01 -4.95187432e-01 3.63356620e-01 -5.47191083e-01
1.24112047e-01 8.17504346e-01 2.21065730e-01 1.41113043e-01
-1.05650592e+00 -6.64979696e-01 1.39204144e-01 2.64182270e-01
-1.74361181e+00 -4.49949831e-01 8.60694945e-01 -8.68796743e-03
7.35450149e-01 9.29747999e-01 6.71981335e-01 7.87457600e-02
4.45602953e-01 1.41370639e-01 1.00518680e+00 -6.45905435e-01
3.11387539e-01 7.71820843e-02 -8.80372003e-02 6.50170505e-01
4.94155467e-01 -1.20613150e-01 -3.41387317e-02 -1.84526458e-01
6.59728765e-01 -1.63227424e-01 -2.70972341e-01 9.15734470e-03
-1.21212065e+00 3.03431869e-01 5.10971427e-01 3.24024230e-01
-5.28521538e-01 2.88528681e-01 4.48803380e-02 1.29336178e-01
4.97302294e-01 2.71002978e-01 -3.65123540e-01 9.46696326e-02
-8.05034935e-01 -1.70511499e-01 5.65476716e-01 1.15736568e+00
3.13869178e-01 3.41337651e-01 -5.54606915e-01 7.24592686e-01
3.81215066e-01 8.39672267e-01 -7.73689151e-02 -7.06319928e-01
4.91354734e-01 3.34618688e-01 2.06425995e-01 -1.00058639e+00
-6.52326226e-01 -1.14774787e+00 -1.21413028e+00 -2.75441408e-01
-1.62044674e-01 -8.23661566e-01 -3.24508786e-01 1.49971163e+00
2.81969726e-01 2.11934239e-01 -1.79399792e-02 1.28438258e+00
3.41863096e-01 1.22422862e+00 -1.03467405e-02 -1.16515720e+00
1.61528504e+00 -8.90802920e-01 -7.36246884e-01 7.40232915e-02
3.44610006e-01 -8.92117202e-01 1.95731610e-01 2.05102071e-01
-1.23407662e+00 -6.21180497e-02 -1.28047872e+00 3.91092956e-01
-7.69488886e-02 7.52302706e-01 5.66518247e-01 1.03439224e+00
-7.54388273e-01 -1.79674655e-01 -5.18175662e-01 8.33919272e-02
1.43046781e-01 8.74071777e-01 4.53109622e-01 1.19241238e-01
-7.61908352e-01 4.86241817e-01 2.60732055e-01 4.99798030e-01
-4.48682785e-01 -5.89010835e-01 -1.83406711e-01 2.85959482e-01
7.27416396e-01 -9.05994117e-01 7.04002678e-01 -6.17047727e-01
-1.88240671e+00 2.31295988e-01 -8.12806264e-02 4.32258695e-02
9.24901143e-02 4.82115209e-01 -4.81741220e-01 -3.42410088e-01
-3.00603837e-01 -1.70246243e-01 6.48444831e-01 -1.06666863e+00
-7.42497087e-01 -3.96432668e-01 2.39390343e-01 5.77485979e-01
-5.26249349e-01 1.48526272e-02 -5.60054779e-01 -5.55767059e-01
2.61787977e-03 -1.02782059e+00 -3.82378340e-01 -4.60680008e-01
-5.36230326e-01 -1.51059479e-01 7.73555636e-01 -3.77351552e-01
1.50128078e+00 -1.96257329e+00 5.74282110e-01 5.68816662e-01
2.81659998e-02 -6.31776825e-02 -7.13629276e-02 3.43806930e-02
3.54357034e-01 -6.39449209e-02 -4.02357951e-02 1.13591373e-01
-1.45781726e-01 -1.24959476e-01 2.38532528e-01 6.52258694e-01
-3.76298457e-01 5.16975164e-01 -5.90087295e-01 -4.04617667e-01
8.54464546e-02 4.18922216e-01 -5.42255461e-01 1.50188774e-01
7.81973451e-02 4.50899988e-01 -9.87519503e-01 7.90956318e-01
1.17660415e+00 -1.26713812e-01 4.70267117e-01 -6.68458283e-01
-5.85390687e-01 -3.01768541e-01 -1.23505664e+00 1.34420133e+00
-7.92498350e-01 3.98484290e-01 8.02940011e-01 -1.48181498e+00
7.68101037e-01 1.58289239e-01 8.84681284e-01 -1.00796688e+00
5.08135080e-01 2.23906890e-01 -3.00999153e-02 -3.69091332e-01
4.76656765e-01 -1.21249199e-01 -1.99905574e-01 3.68137926e-01
-5.22370338e-01 3.67346913e-01 6.75911456e-02 -2.64368176e-01
8.84369731e-01 -1.14963815e-01 3.36852185e-02 -7.35320270e-01
9.45866346e-01 -1.83187146e-02 8.55002105e-01 2.21718147e-01
3.50881457e-01 -2.65981555e-01 2.36519784e-01 2.43744571e-02
-8.03500295e-01 -7.34531641e-01 -2.81050175e-01 1.03967261e+00
7.99197853e-01 2.32065041e-02 -5.55032969e-01 -1.58048764e-01
-2.38958836e-01 4.95287478e-01 3.38656083e-02 -1.40335336e-01
-6.99017465e-01 -1.37125337e+00 -9.84690115e-02 -9.71382856e-02
6.99886978e-01 -1.52552605e-01 -9.38639998e-01 3.73472422e-01
-1.53597623e-01 -9.97221053e-01 -5.32979012e-01 3.39200050e-01
-6.30125701e-01 -5.67645609e-01 -8.96026254e-01 -6.88421309e-01
1.12914443e+00 6.31005108e-01 8.13555241e-01 4.51118238e-02
-4.19994891e-01 2.43526697e-01 -8.86137560e-02 -3.93760264e-01
4.20506179e-01 1.91170618e-01 1.08524472e-01 1.31049573e-01
-4.63777661e-01 -4.90844667e-01 -7.78734267e-01 6.57680750e-01
-6.34383202e-01 2.42682427e-01 8.36945415e-01 5.08842409e-01
9.12757874e-01 5.54895043e-01 5.81109881e-01 -5.45410275e-01
6.21714413e-01 -8.20674419e-01 -9.83939230e-01 5.70844352e-01
-3.08015466e-01 -6.05210587e-02 8.80919337e-01 -1.70325428e-01
-1.09420669e+00 -1.32701620e-02 4.02031660e-01 -7.94417113e-02
7.14212298e-01 5.31858325e-01 -4.47952300e-01 -5.57851732e-01
-2.06087809e-02 5.91711938e-01 -2.77099729e-01 -1.46216035e-01
-1.27802700e-01 8.58520925e-01 1.99724272e-01 -6.70907438e-01
8.76694322e-01 2.66938299e-01 7.96870649e-01 -7.57110000e-01
-4.61521924e-01 -2.25814171e-02 -2.29328796e-02 -6.12713158e-01
5.51807344e-01 -9.91964340e-01 -1.21896088e+00 -1.37898931e-02
-9.12483394e-01 8.25488716e-02 4.19444025e-01 6.86211646e-01
-3.10918391e-01 2.14235902e-01 2.17240572e-01 -1.03564131e+00
-6.93294585e-01 -9.63686287e-01 8.20486307e-01 4.44716066e-01
4.84939754e-01 -5.69659352e-01 -4.09414917e-01 3.32257748e-01
7.97998607e-01 7.01893210e-01 1.04672408e+00 3.56916755e-01
-8.48332703e-01 1.97504424e-02 -3.11744153e-01 -1.46387026e-01
1.46213859e-01 -3.52251261e-01 -1.32256791e-01 -7.31871486e-01
2.58358661e-02 3.65407825e-01 1.73592374e-01 5.91196001e-01
1.53635406e+00 -5.62031746e-01 -7.18394995e-01 9.54475820e-01
1.72263932e+00 3.82582635e-01 2.73632258e-01 2.49307588e-01
2.64917046e-01 7.82709848e-03 8.45002651e-01 8.67396712e-01
4.41173464e-01 1.04701018e+00 6.57028019e-01 -7.81722292e-02
4.39742506e-01 5.54760456e-01 1.67542607e-01 9.57197726e-01
-4.63202268e-01 -6.58870816e-01 -4.09590274e-01 1.04900479e-01
-1.74889410e+00 -5.31715274e-01 -1.97182834e-01 2.30739760e+00
3.76257509e-01 -3.77062172e-01 8.20102692e-02 1.65450096e-01
8.10275078e-01 -1.28250524e-01 -8.28441441e-01 -3.49871755e-01
1.23618990e-01 -1.55332625e-01 9.25500631e-01 -6.85956404e-02
-6.88986540e-01 7.06804171e-02 4.79422760e+00 1.02206612e+00
-1.31698000e+00 3.55083197e-01 5.65544724e-01 -5.81293941e-01
-4.33261901e-01 -1.91386223e-01 -4.24965262e-01 9.84037220e-01
8.20623875e-01 -6.48949504e-01 1.05119073e+00 3.72931838e-01
4.93354112e-01 -1.92349017e-01 -6.56302512e-01 1.29079390e+00
-1.88482150e-01 -1.18181920e+00 -3.36659968e-01 2.92038143e-01
6.61889374e-01 -4.04200196e-01 6.02525473e-02 1.56666115e-01
-3.34265828e-01 -6.71428561e-01 6.20463550e-01 5.42684436e-01
9.29694176e-01 -1.21978056e+00 4.55591261e-01 3.73217285e-01
-1.53555870e+00 -3.78217101e-01 -4.39692497e-01 -1.32330060e-02
1.03174835e-01 8.54896605e-01 1.07929204e-02 1.17864168e+00
4.13162529e-01 2.10349113e-01 2.28163242e-01 9.75704432e-01
4.84077871e-01 8.73287320e-02 -5.58432698e-01 -5.18885374e-01
6.18638769e-02 -6.83751464e-01 7.73969948e-01 6.97762132e-01
7.79650033e-01 9.79588747e-01 2.20178127e-01 7.35152364e-01
-1.79141238e-01 1.71852574e-01 2.13165358e-02 3.40490282e-01
8.07619154e-01 1.74279058e+00 -6.76045656e-01 1.22153535e-01
-3.18528652e-01 6.17044568e-01 -2.60285269e-02 4.62952882e-01
-1.18404996e+00 -6.73587024e-01 5.38242459e-01 -1.25573963e-01
-4.39411812e-02 -5.89236379e-01 -4.16055351e-01 -8.45665276e-01
3.10613573e-01 -1.81366950e-01 1.65789425e-01 -4.58907753e-01
-4.21156853e-01 4.80122179e-01 -2.22982749e-01 -1.39933884e+00
4.31087762e-01 -2.12592795e-01 -6.64305151e-01 9.85241950e-01
-1.91665626e+00 -8.09596062e-01 -5.99218607e-01 6.16335928e-01
2.81402737e-01 -2.18696401e-01 4.37046140e-01 6.67051613e-01
-1.15361464e+00 7.59306729e-01 6.89023674e-01 -5.77727914e-01
-3.48484926e-02 -6.19209230e-01 -9.13152337e-01 7.34345198e-01
-7.97241211e-01 2.74961591e-01 4.98699963e-01 -1.51260778e-01
-2.47740889e+00 -1.18189001e+00 2.54121780e-01 6.20872796e-01
5.83837442e-02 -3.11226785e-01 5.13292551e-02 -2.71606073e-03
1.48231804e-01 -3.32666226e-02 8.88259292e-01 -3.52028370e-01
6.66366935e-01 -4.88652945e-01 -1.29228020e+00 5.52887082e-01
9.66557384e-01 2.83298999e-01 4.53997135e-01 6.36033297e-01
4.06367302e-01 -5.44602811e-01 -1.07942247e+00 4.49010909e-01
4.43066895e-01 -1.52698904e-01 9.30839181e-01 5.20142503e-02
1.26544595e-01 -5.33423305e-01 -1.87578559e-01 -1.32212400e+00
-6.42463624e-01 -9.49952602e-01 -1.76588684e-01 1.18774056e+00
3.39480907e-01 -6.17953479e-01 6.91485226e-01 3.81222963e-01
-3.44099551e-01 -1.15531063e+00 -1.33990359e+00 -8.33907902e-01
-4.46826220e-01 1.86936766e-01 7.64627576e-01 6.20844126e-01
2.20425963e-01 4.29798067e-01 -4.47511673e-01 4.66974050e-01
8.15160334e-01 3.48415822e-01 4.71169591e-01 -1.00818312e+00
-2.55823106e-01 -2.38083079e-01 -2.46120766e-02 -1.27255034e+00
-5.63414942e-04 -9.95048940e-01 -1.02326006e-01 -1.50089788e+00
1.73685431e-01 -1.00707161e+00 -3.14601243e-01 9.65511575e-02
1.10859506e-01 3.48015688e-02 1.07640065e-01 1.95737079e-01
-6.93765998e-01 8.47294390e-01 1.41508138e+00 -2.35278144e-01
-3.73224258e-01 1.40769035e-01 -7.00710833e-01 6.28608912e-02
6.81151748e-01 -2.74102181e-01 -4.70756263e-01 -8.77294719e-01
3.22976381e-01 7.71471441e-01 -8.03250670e-02 -1.10021400e+00
4.43789750e-01 -4.83120292e-01 2.30877236e-01 -4.79665607e-01
4.92239803e-01 -1.52454174e+00 4.28354591e-01 6.66453123e-01
1.49738684e-01 -6.56008124e-02 2.47088894e-01 5.47991812e-01
2.32182071e-01 1.33070692e-01 6.24166250e-01 3.84058982e-01
-3.30991715e-01 3.80377203e-01 -6.52683973e-01 -5.02750218e-01
1.68485713e+00 -4.13526356e-01 -3.83249164e-01 9.72924531e-02
-4.55545664e-01 7.45825052e-01 -1.33581474e-01 2.05089778e-01
2.83331662e-01 -1.54468679e+00 -7.16943860e-01 -1.93152040e-01
-3.64103913e-01 -2.51429230e-01 6.11782074e-01 1.25381184e+00
-3.04789037e-01 7.59280384e-01 -2.35409573e-01 -4.34871167e-01
-1.29364264e+00 7.49609247e-02 4.95161742e-01 -2.77924836e-01
1.18912920e-01 6.31956160e-01 -3.20636660e-01 5.77084757e-02
-6.58866540e-02 -2.72640362e-02 -1.20668538e-01 -9.49982777e-02
-9.68354866e-02 1.03582621e+00 1.52264744e-01 -4.10539180e-01
-6.66121185e-01 8.11444521e-01 3.25075090e-01 3.43387395e-01
1.47171998e+00 -5.84400177e-01 -5.50086617e-01 -5.11192262e-01
1.25435042e+00 -5.72042838e-02 -6.99098051e-01 -2.14300632e-01
-6.26048923e-01 -4.72367793e-01 7.69127488e-01 -6.72159553e-01
-1.35986722e+00 1.96527079e-01 7.24399149e-01 2.24195823e-01
1.88627255e+00 -5.16740382e-01 7.10376561e-01 2.44988471e-01
7.55552113e-01 -1.30951440e+00 -3.08963269e-01 2.78355002e-01
6.23886108e-01 -4.64449733e-01 3.78781289e-01 -8.15127909e-01
1.55616596e-01 1.30203032e+00 7.22216785e-01 1.18599080e-01
7.44799078e-01 3.17422926e-01 -7.14683294e-01 -3.37341912e-02
-6.02295160e-01 2.08640248e-01 1.65828153e-01 8.69568810e-02
3.68069887e-01 5.07285178e-01 -1.01292968e+00 9.07753050e-01
-6.72954842e-02 9.22509655e-03 4.08257276e-01 1.00293481e+00
-3.28437746e-01 -9.46145296e-01 -5.21619380e-01 7.52606809e-01
-3.88275772e-01 1.45574763e-01 2.44622529e-01 1.87745243e-01
2.98589796e-01 1.13119745e+00 1.95846274e-01 -3.77799243e-01
3.80622238e-01 -6.82837009e-01 5.75349927e-01 -2.88043201e-01
-3.26389581e-01 1.91564158e-01 1.63963094e-01 -2.04714999e-01
-3.60843927e-01 -3.08036059e-01 -1.11849499e+00 -3.86246830e-01
-7.69286096e-01 4.71209317e-01 9.54257965e-01 9.64361489e-01
6.79565966e-01 1.25543845e+00 1.36559451e+00 -7.97035992e-01
-3.12912226e-01 -4.38632280e-01 -7.07262218e-01 -8.65324080e-01
5.46667688e-02 -5.54301739e-01 -3.12386062e-02 -4.84008729e-01] | [6.015433311462402, 1.520037055015564] |
a20e6cf4-3ff3-4641-a96a-7679328814de | pose-agnostic-cross-spectral-hallucination | 1909.04365 | null | https://arxiv.org/abs/1909.04365v2 | https://arxiv.org/pdf/1909.04365v2.pdf | Cross-Spectral Face Hallucination via Disentangling Independent Factors | The cross-sensor gap is one of the challenges that have aroused much research interests in Heterogeneous Face Recognition (HFR). Although recent methods have attempted to fill the gap with deep generative networks, most of them suffer from the inevitable misalignment between different face modalities. Instead of imaging sensors, the misalignment primarily results from facial geometric variations that are independent of the spectrum. Rather than building a monolithic but complex structure, this paper proposes a Pose Aligned Cross-spectral Hallucination (PACH) approach to disentangle the independent factors and deal with them in individual stages. In the first stage, an Unsupervised Face Alignment (UFA) module is designed to align the facial shapes of the near-infrared (NIR) images with those of the visible (VIS) images in a generative way, where UV maps are effectively utilized as the shape guidance. Thus the task of the second stage becomes spectrum translation with aligned paired data. We develop a Texture Prior Synthesis (TPS) module to achieve complexion control and consequently generate more realistic VIS images than existing methods. Experiments on three challenging NIR-VIS datasets verify the effectiveness of our approach in producing visually appealing images and achieving state-of-the-art performance in HFR. | ['Boyan Duan', 'Yi Li', 'Xingguang Song', 'Ran He', 'Chaoyou Fu'] | 2019-09-10 | cross-spectral-face-hallucination-via | http://openaccess.thecvf.com/content_CVPR_2020/html/Duan_Cross-Spectral_Face_Hallucination_via_Disentangling_Independent_Factors_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Duan_Cross-Spectral_Face_Hallucination_via_Disentangling_Independent_Factors_CVPR_2020_paper.pdf | cvpr-2020-6 | ['heterogeneous-face-recognition', 'face-hallucination'] | ['computer-vision', 'computer-vision'] | [ 6.58705592e-01 -3.49362604e-02 3.91753703e-01 -4.95973200e-01
-4.37505960e-01 -2.15810239e-01 6.79496169e-01 -7.41545379e-01
1.13429032e-01 5.12618840e-01 2.50199527e-01 7.13529289e-02
-7.87870660e-02 -8.21962774e-01 -7.69469082e-01 -1.08722198e+00
6.70554936e-01 6.98617622e-02 -3.42687041e-01 -4.02913749e-01
-1.01499371e-01 5.64975321e-01 -1.85383391e+00 1.66395858e-01
1.07517111e+00 1.09784901e+00 1.03603877e-01 2.15249524e-01
-9.21967402e-02 3.12643230e-01 -3.33145440e-01 -6.28482699e-01
5.30341864e-01 -6.43846631e-01 -1.12952121e-01 5.62627673e-01
5.75234413e-01 -2.47110024e-01 -3.87761831e-01 1.27763736e+00
7.28652179e-01 -9.81606320e-02 6.58377826e-01 -1.17728281e+00
-9.84053910e-01 2.37839192e-01 -9.67878938e-01 -5.92359185e-01
2.65567392e-01 1.43551007e-01 5.47239959e-01 -1.08936095e+00
4.34242040e-01 1.47912157e+00 4.95018125e-01 6.61783814e-01
-1.34810221e+00 -7.08337069e-01 -1.33352682e-01 4.34693210e-02
-1.38393724e+00 -8.44198525e-01 1.17574799e+00 -4.16809708e-01
3.13363522e-01 3.37811053e-01 6.13422692e-01 1.33164692e+00
3.01623400e-02 1.47652045e-01 1.28836644e+00 -4.00306731e-01
-3.75459269e-02 1.81641310e-01 -5.30739844e-01 5.28487444e-01
1.37820452e-01 1.73892647e-01 -6.19540036e-01 -1.52726397e-02
8.80764365e-01 1.34355620e-01 -5.68219662e-01 -2.77642548e-01
-9.41737652e-01 4.74589705e-01 3.71268690e-01 1.22543104e-01
-4.78571892e-01 -2.57938981e-01 -1.84759840e-01 2.38882359e-02
5.23883998e-01 2.82321960e-01 2.76721772e-02 4.91801471e-01
-7.79282033e-01 1.09945893e-01 3.12152952e-01 7.45499194e-01
7.89290965e-01 3.33277106e-01 -2.63984621e-01 1.25052130e+00
6.14055276e-01 9.18486476e-01 2.13414624e-01 -6.67579055e-01
3.41599226e-01 4.87706989e-01 2.92238668e-02 -1.06619942e+00
-8.86285901e-02 -5.72150886e-01 -1.01034272e+00 2.77562380e-01
2.99193203e-01 -2.41288424e-01 -1.14447367e+00 1.84977770e+00
4.69645530e-01 1.03657410e-01 9.62794498e-02 1.10680151e+00
7.96263695e-01 6.17493927e-01 -1.12176038e-01 -4.85149473e-01
1.34107602e+00 -6.77539229e-01 -8.66445124e-01 -2.11139157e-01
-3.13698947e-01 -1.17631328e+00 8.16495597e-01 2.26751640e-01
-1.05921030e+00 -7.08297789e-01 -1.08091414e+00 2.49398891e-02
7.94237945e-03 2.89791375e-01 4.03682500e-01 6.06397271e-01
-1.04785264e+00 3.93614739e-01 -4.44210827e-01 -1.67339444e-01
3.91676098e-01 6.08397909e-02 -4.42987919e-01 -2.35128388e-01
-9.76508319e-01 6.26762986e-01 -1.25249019e-02 5.24227977e-01
-6.21685445e-01 -7.12197006e-01 -7.76307881e-01 -5.94348982e-02
4.44731981e-01 -6.64796889e-01 6.56238854e-01 -1.18171108e+00
-1.86962759e+00 7.50008941e-01 -1.32851794e-01 3.44585538e-01
5.31875193e-01 -4.35378440e-02 -5.77393711e-01 2.04923525e-02
-2.61788249e-01 5.02827406e-01 1.25918770e+00 -1.68053329e+00
4.35533561e-02 -6.79629982e-01 -3.39508295e-01 3.72967333e-01
-4.47387934e-01 8.17702115e-02 -4.62323517e-01 -6.93359315e-01
3.11963558e-01 -8.79646897e-01 1.44666359e-01 1.51078820e-01
-5.24512410e-01 2.24630252e-01 8.39001477e-01 -7.18663514e-01
7.65117645e-01 -2.19810367e+00 3.03196669e-01 2.27146074e-01
2.20066264e-01 3.67885649e-01 -3.37285846e-01 3.32118094e-01
-4.23876643e-01 -3.07308167e-01 -1.97191909e-01 -5.00890374e-01
-9.18240026e-02 -3.20991711e-03 -3.50697815e-01 4.70175624e-01
3.99908841e-01 8.33039343e-01 -4.72304612e-01 -2.79592723e-01
7.29743540e-02 1.02796650e+00 -4.00677174e-01 5.07906377e-01
-1.91928610e-01 7.77749121e-01 -2.88069606e-01 8.14177930e-01
1.15570617e+00 -1.59097277e-02 2.35435709e-01 -7.30825961e-01
-1.14307433e-01 -3.00215989e-01 -1.04350102e+00 1.61838424e+00
-2.64332384e-01 2.10191891e-01 2.95797765e-01 -7.82499254e-01
1.26195049e+00 4.11485374e-01 7.19368994e-01 -9.32017207e-01
2.79129416e-01 2.64242500e-01 4.43091914e-02 -5.48040867e-01
1.43610448e-01 -5.11949658e-01 4.74265128e-01 8.39326829e-02
7.76915327e-02 -2.01777309e-01 -3.39957118e-01 -3.34361851e-01
3.72247666e-01 4.11685079e-01 -1.00901291e-01 -6.09765612e-02
6.33470058e-01 -5.90028822e-01 6.12972438e-01 1.24431379e-01
6.36894703e-02 1.05313122e+00 1.80965438e-01 -2.18891129e-01
-1.14549804e+00 -1.16357124e+00 -2.12489814e-01 5.27374625e-01
1.76061258e-01 -2.18041047e-01 -9.63540256e-01 -7.50481859e-02
-1.42263129e-01 3.35721523e-01 -5.44609427e-01 -3.82297933e-02
-3.29021126e-01 -8.99864435e-01 2.46585190e-01 3.01335782e-01
8.10904920e-01 -8.04645836e-01 -2.59678483e-01 -4.89710309e-02
-2.14664325e-01 -1.21921062e+00 -4.69779760e-01 -3.66875350e-01
-4.18608069e-01 -8.77088904e-01 -8.42296064e-01 -5.02504408e-01
8.28381121e-01 5.01749098e-01 6.75818920e-01 -1.13008574e-01
-5.53973734e-01 2.51323074e-01 -1.63714260e-01 -4.15664166e-01
-2.92344302e-01 -6.15061045e-01 4.21445481e-02 8.85336339e-01
-3.00625935e-02 -8.04025233e-01 -8.04907143e-01 4.05410886e-01
-1.04406059e+00 4.82276440e-01 6.32214010e-01 8.31155241e-01
6.04700863e-01 -1.32740140e-01 4.31568176e-01 -7.11349547e-01
3.74970704e-01 -2.58820057e-01 -5.83105564e-01 3.60701084e-01
-6.19086504e-01 -2.08218932e-01 6.15989566e-01 -5.32571077e-01
-1.49479282e+00 1.33641288e-01 -1.66179940e-01 -7.45636761e-01
-9.21011642e-02 1.45065084e-01 -7.39853323e-01 -2.74255902e-01
5.94447732e-01 3.03638309e-01 3.69770139e-01 -4.07218993e-01
2.41095841e-01 7.57372379e-01 5.53644121e-01 -6.20960951e-01
1.18902588e+00 4.70347643e-01 1.63818389e-01 -1.01950741e+00
-6.72558010e-01 2.22076587e-02 -4.07820225e-01 -5.76465964e-01
9.46961462e-01 -1.04884684e+00 -7.11036384e-01 8.76296461e-01
-1.00423121e+00 6.83930516e-03 6.10393845e-03 3.28364104e-01
-3.55053425e-01 2.57016301e-01 -3.67760777e-01 -8.84605348e-01
-2.06940711e-01 -1.27085483e+00 1.25352788e+00 6.60645962e-01
3.13168019e-01 -5.33540070e-01 -9.91213247e-02 7.00144410e-01
6.65869117e-01 4.11428899e-01 7.58974373e-01 1.08156957e-01
-6.73170865e-01 8.78865421e-02 -4.00857419e-01 4.00710434e-01
5.04591227e-01 1.36811420e-01 -1.42896307e+00 -2.75254101e-01
2.43615061e-01 -1.46424606e-01 5.51817238e-01 2.51964450e-01
1.13569176e+00 -3.16848755e-01 -2.62188856e-02 8.94585252e-01
1.50082839e+00 2.62417734e-01 9.62110758e-01 -1.22280523e-01
9.98395801e-01 9.13279474e-01 3.80899310e-01 4.08851773e-01
2.08526298e-01 9.16426659e-01 4.55266923e-01 -4.48341370e-01
-3.51452649e-01 -3.43901515e-01 4.22712892e-01 4.55843538e-01
-3.21689695e-01 -1.41685426e-01 -5.42797267e-01 5.27697951e-02
-1.73830843e+00 -9.42334056e-01 5.63700236e-02 2.18789148e+00
8.08140039e-01 -5.94616473e-01 -1.59976602e-01 -1.40413931e-02
8.22647214e-01 1.34825364e-01 -6.29293859e-01 1.16766863e-01
-4.45235521e-01 2.62460113e-01 1.17895724e-02 3.45698148e-01
-7.26724029e-01 7.22571611e-01 5.44136190e+00 6.60520196e-01
-1.45477712e+00 -9.93458331e-02 7.73508191e-01 7.15268552e-02
-6.08993411e-01 -1.75859973e-01 -5.37568271e-01 3.35892975e-01
4.35924023e-01 1.70990571e-01 7.25025415e-01 4.94720906e-01
2.16363892e-01 1.27179593e-01 -8.62272441e-01 1.40968955e+00
2.63036788e-01 -9.68550563e-01 1.37126908e-01 1.88630998e-01
8.37491810e-01 -3.77808809e-01 4.74630564e-01 -2.71912485e-01
-2.90358335e-01 -1.22473037e+00 9.25706983e-01 9.00087774e-01
1.13392222e+00 -6.49327695e-01 3.47964168e-01 -1.42361913e-02
-1.12472725e+00 -3.00666373e-02 -3.75008881e-01 3.18715572e-01
5.00870608e-02 6.99191868e-01 -4.99119371e-01 8.12693119e-01
5.95365584e-01 5.24245143e-01 -4.16515648e-01 6.89827859e-01
-1.64054558e-01 1.08211152e-01 -6.90558702e-02 4.69739199e-01
-3.81245673e-01 -6.67790771e-01 5.96022904e-01 6.56068683e-01
5.22647142e-01 1.37331769e-01 -9.92669910e-03 1.39136577e+00
-1.89959764e-01 1.43164173e-01 -4.80457991e-01 -2.68616289e-01
3.29036474e-01 1.47556317e+00 -3.17788929e-01 9.68154073e-02
-4.25534695e-01 9.00348902e-01 -4.62270901e-02 6.01949811e-01
-9.31560338e-01 1.20443217e-02 7.78738320e-01 2.56158501e-01
7.75807276e-02 -1.41535178e-01 -2.33140558e-01 -1.27594090e+00
3.45432341e-01 -9.97079074e-01 -1.03766471e-01 -1.07732570e+00
-1.40539610e+00 6.40260637e-01 -1.58226162e-01 -1.21052063e+00
-8.50946549e-03 -5.27583718e-01 -5.29202580e-01 1.32375121e+00
-1.75751293e+00 -1.65871775e+00 -6.35894775e-01 7.80290067e-01
1.83245540e-01 -1.27604038e-01 7.28305995e-01 4.14535165e-01
-6.93608224e-01 6.05870128e-01 -1.64251402e-01 -1.00160733e-01
8.10254574e-01 -7.01854110e-01 -3.98199186e-02 9.16283667e-01
-1.14961356e-01 6.59510493e-01 6.66260719e-01 -6.08845353e-01
-1.74679255e+00 -1.09830081e+00 2.18353048e-01 -5.95766505e-05
1.93565026e-01 -4.44101244e-01 -8.18139434e-01 2.83186555e-01
2.35210255e-01 1.10905744e-01 6.52602732e-01 -2.83075064e-01
-6.91251576e-01 -4.52362984e-01 -1.07864940e+00 6.10556901e-01
1.15731144e+00 -5.97904980e-01 -1.39132172e-01 8.64095986e-02
4.76349056e-01 -4.14801925e-01 -8.43291461e-01 6.56177759e-01
6.86320484e-01 -1.20300937e+00 9.67653811e-01 -1.20967373e-01
5.41312754e-01 -5.41173041e-01 -2.52799183e-01 -1.36281180e+00
-7.17163831e-02 -8.79412770e-01 2.09499776e-01 1.60646081e+00
1.75949350e-01 -8.95733178e-01 4.39272195e-01 5.70521653e-01
-8.91339257e-02 -6.33747280e-01 -4.94406730e-01 -3.99692357e-01
-3.27343345e-01 -2.20995005e-02 8.54112327e-01 1.00122952e+00
-3.91513884e-01 3.45046788e-01 -4.49188530e-01 2.84362048e-01
8.91734004e-01 2.83900857e-01 8.04183006e-01 -1.25520849e+00
-3.92545819e-01 -2.78333187e-01 -1.15826733e-01 -6.03004098e-01
2.12884936e-02 -6.29767776e-01 1.95210233e-01 -1.19960666e+00
3.96727622e-01 -3.00594270e-01 1.10265642e-01 5.31394601e-01
-1.36185706e-01 3.32087964e-01 1.91965178e-01 2.52254039e-01
1.27325714e-01 1.02399302e+00 1.68180120e+00 2.13559531e-02
-3.71964127e-02 -2.71372795e-01 -9.60864007e-01 4.59825784e-01
4.46300745e-01 1.18434474e-01 -6.58024728e-01 -4.31481779e-01
8.23192298e-02 2.07780182e-01 4.70072538e-01 -9.80929494e-01
5.24321944e-02 -3.30526561e-01 6.86207652e-01 -2.33913153e-01
6.12736046e-01 -9.41267312e-01 6.28245711e-01 -2.66762823e-02
-1.85723379e-02 -2.22487450e-01 2.88565662e-02 4.71213490e-01
-1.81114450e-01 3.74570429e-01 1.18665683e+00 1.37932271e-01
-5.30340746e-02 5.35322189e-01 3.33767146e-01 -3.26458424e-01
9.40970004e-01 -2.80885816e-01 -5.58490455e-01 -3.04299295e-01
-3.92454624e-01 -1.59838915e-01 5.94200075e-01 5.41912973e-01
7.35615075e-01 -1.41378570e+00 -7.67726421e-01 7.24684715e-01
6.16374537e-02 1.01496324e-01 6.78377271e-01 8.52377772e-01
-2.81747997e-01 1.80717260e-02 -3.97041857e-01 -6.46005690e-01
-1.21958208e+00 3.30317378e-01 6.42960906e-01 2.83705652e-01
-5.20568728e-01 7.62352526e-01 5.46012819e-01 -3.17448467e-01
7.69648552e-02 3.43035042e-01 -3.41884673e-01 2.57894080e-02
5.65760314e-01 6.90913126e-02 1.51856273e-01 -9.49939787e-01
-3.02378833e-02 9.09507334e-01 1.66973457e-01 -7.62322545e-02
1.65497375e+00 -1.47554070e-01 -3.38524252e-01 9.49100628e-02
1.08694124e+00 1.48278207e-01 -1.49702656e+00 -2.31439054e-01
-6.69125080e-01 -7.58299768e-01 -5.18227331e-02 -7.36329973e-01
-1.44501483e+00 8.79948199e-01 7.17445195e-01 -1.02614798e-01
1.43820763e+00 -3.03909361e-01 7.02835619e-01 -2.18290001e-01
1.65824383e-01 -9.72443998e-01 2.89601266e-01 6.10581301e-02
1.11320913e+00 -9.53136683e-01 -1.84632793e-01 -8.64477754e-01
-5.16031623e-01 1.22564471e+00 7.37378597e-01 2.53292024e-01
5.22008598e-01 2.34081835e-01 1.11119486e-01 -1.74938992e-01
-3.63557220e-01 -1.55337095e-01 4.88830417e-01 6.74618483e-01
3.91341865e-01 -5.35261519e-02 3.13080940e-03 4.48822647e-01
-1.25595614e-01 -1.97584823e-01 1.82652593e-01 6.14969492e-01
-1.39263526e-01 -1.17676163e+00 -6.85462415e-01 -3.44140232e-02
-1.66329727e-01 1.18090309e-01 -2.52319992e-01 4.26355958e-01
1.88526154e-01 9.45806623e-01 -1.69279084e-01 -6.24419153e-01
3.88302445e-01 1.91479027e-01 7.71899521e-01 -4.17142808e-01
-9.77066606e-02 5.17157733e-01 -2.40319744e-01 -6.05513334e-01
-4.21328038e-01 -5.50858617e-01 -8.26386988e-01 -2.61373401e-01
-2.28831038e-01 -2.67447889e-01 8.77585351e-01 8.08879733e-01
4.62555230e-01 5.42470634e-01 1.00841570e+00 -1.00379002e+00
-4.95796084e-01 -8.54070246e-01 -7.10061073e-01 6.00038230e-01
2.36897349e-01 -8.31856310e-01 -2.14823976e-01 5.27433082e-02] | [12.884637832641602, -0.0009344199788756669] |
a3e9c69f-56fd-4e9b-9430-d9ab5c7926f9 | multi-agent-path-finding-with-capacity | 1907.12648 | null | https://arxiv.org/abs/1907.12648v1 | https://arxiv.org/pdf/1907.12648v1.pdf | Multi-Agent Path Finding with Capacity Constraints | In multi-agent path finding (MAPF) the task is to navigate agents from their starting positions to given individual goals. The problem takes place in an undirected graph whose vertices represent positions and edges define the topology. Agents can move to neighbor vertices across edges. In the standard MAPF, space occupation by agents is modeled by a capacity constraint that permits at most one agent per vertex. We suggest an extension of MAPF in this paper that permits more than one agent per vertex. Propositional satisfiability (SAT) models for these extensions of MAPF are studied. We focus on modeling capacity constraints in SAT-based formulations of MAPF and evaluation of performance of these models. We extend two existing SAT-based formulations with vertex capacity constraints: MDD-SAT and SMT-CBS where the former is an approach that builds the model in an eager way while the latter relies on lazy construction of the model. | ['Sven Koenig', 'Pavel Surynek', 'T. K. Satish Kumar'] | 2019-07-21 | null | null | null | null | ['multi-agent-path-finding'] | ['playing-games'] | [ 1.38331681e-01 5.99333584e-01 -2.10576519e-01 -1.07366689e-01
-1.32777199e-01 -1.01294923e+00 7.05788314e-01 4.66450870e-01
-3.70413452e-01 1.23303246e+00 -2.63484925e-01 -5.13543248e-01
-9.49061036e-01 -1.45519459e+00 -5.32671809e-01 -4.25083578e-01
-6.23525441e-01 1.41883636e+00 8.48647237e-01 -4.99620646e-01
1.49874315e-01 6.29136920e-01 -8.89854431e-01 6.78123683e-02
4.76600885e-01 1.68499708e-01 2.94277221e-01 9.48997021e-01
-3.29815894e-01 9.24251735e-01 -4.53904867e-01 -1.29564255e-01
4.89946127e-01 -3.38814497e-01 -1.33740306e+00 2.32502818e-01
-2.97289282e-01 -1.55221373e-01 -4.19425398e-01 9.03141797e-01
-3.21654350e-01 1.37831673e-01 4.94882673e-01 -2.33123469e+00
-1.08641721e-02 8.93308997e-01 -4.98187304e-01 1.28506199e-01
7.41263211e-01 -2.94011552e-02 9.89300013e-01 2.31605321e-01
8.49891603e-01 1.21814501e+00 1.59562811e-01 4.69200373e-01
-1.12194562e+00 2.16579661e-01 6.83492064e-01 5.10453343e-01
-1.24338555e+00 -1.62462965e-01 1.23196311e-01 -2.60213375e-01
1.70232356e+00 7.01662421e-01 6.27352297e-01 2.60234177e-01
2.55063564e-01 2.99987346e-01 1.02064443e+00 -6.43665910e-01
4.87572283e-01 -1.16968546e-02 3.38694066e-01 6.53939068e-01
6.57008171e-01 2.02510003e-02 -1.31129697e-01 -3.27921957e-01
8.65505040e-01 -2.68672556e-01 -1.67180300e-02 -5.96313238e-01
-1.13389039e+00 1.01056755e+00 1.38739005e-01 1.97472617e-01
-4.06761825e-01 6.31220341e-01 2.29256839e-01 4.40798134e-01
-2.52290040e-01 5.40365815e-01 -3.56275320e-01 -6.12791479e-02
-4.58037168e-01 6.51037931e-01 1.19640386e+00 1.35656071e+00
5.89784980e-01 -4.36924636e-01 9.93190035e-02 2.97076087e-02
2.75784165e-01 5.29878974e-01 -4.78998423e-01 -1.24545443e+00
7.24401474e-01 6.66500151e-01 5.70468187e-01 -8.51377964e-01
-7.31928051e-01 -4.46454436e-02 -1.26361132e-01 5.66173375e-01
4.48315084e-01 -3.27464379e-02 -6.39001191e-01 1.89538312e+00
4.80162621e-01 6.17707483e-02 2.41140947e-01 5.04089534e-01
3.42739969e-01 9.34211254e-01 -3.25104058e-01 -5.93739867e-01
1.06764126e+00 -1.49189222e+00 -5.90056062e-01 -2.17679724e-01
7.44349003e-01 -1.34667248e-01 7.05182552e-02 4.62467015e-01
-1.58239496e+00 4.36848342e-01 -9.20895100e-01 1.94118083e-01
-5.83792150e-01 -7.47410417e-01 8.03947031e-01 5.80092311e-01
-1.62959361e+00 1.95461765e-01 -9.67208207e-01 -4.71147299e-01
-9.57888290e-02 8.44039440e-01 -5.05115569e-01 -3.07168067e-01
-8.66310477e-01 9.57508445e-01 5.09693265e-01 -6.85940757e-02
-9.55822349e-01 8.85609537e-02 -8.88919771e-01 9.23505425e-02
1.08966899e+00 -8.61161947e-01 1.44404173e+00 -5.37392795e-01
-1.39194775e+00 4.30676907e-01 -2.03857064e-01 -5.94512701e-01
5.28398514e-01 7.00682759e-01 -1.94793791e-01 1.56303421e-01
3.26377660e-01 4.88601893e-01 5.82760088e-02 -1.27547061e+00
-9.63614404e-01 -2.66723216e-01 1.08674788e+00 2.08256662e-01
3.90427798e-01 -9.90364701e-03 -3.93587440e-01 4.41817462e-01
1.30698830e-01 -1.13753629e+00 -5.87404251e-01 -4.36835170e-01
-5.75614274e-01 -3.59750211e-01 3.90774935e-01 1.66639492e-01
1.31731951e+00 -1.39885116e+00 5.62161744e-01 6.39896393e-01
3.68442774e-01 -2.89907008e-01 -4.86042947e-01 1.34529591e+00
1.98223367e-01 5.88743016e-02 1.02787450e-01 -9.30354372e-02
4.38688070e-01 7.37648010e-01 1.91569820e-01 6.04191065e-01
-1.85908854e-01 8.23536932e-01 -9.83432949e-01 -4.50092405e-01
1.56092152e-01 -1.63014799e-01 -7.39853263e-01 -3.26186150e-01
-7.03111589e-01 -2.40379453e-01 -6.02035522e-01 2.91662335e-01
8.29415500e-01 -2.57681966e-01 6.88677728e-01 7.96158314e-01
-6.00753963e-01 3.93529654e-01 -1.62650633e+00 1.41960561e+00
-2.06016123e-01 2.35180318e-01 3.32982510e-01 -6.82401597e-01
4.37780410e-01 1.10609911e-01 3.89220446e-01 -6.37477458e-01
6.91966936e-02 1.30891368e-01 2.32689127e-01 -3.24702978e-01
5.18424392e-01 1.15280874e-01 -3.23175967e-01 7.26763010e-01
-6.49713576e-01 1.31947368e-01 9.25116360e-01 6.48880064e-01
1.69670105e+00 -2.77945697e-01 3.41641665e-01 -3.00955057e-01
7.45465994e-01 6.97811604e-01 4.70636547e-01 9.68226910e-01
-7.98275694e-02 -2.77906537e-01 9.18394148e-01 -4.80040848e-01
-8.70454192e-01 -9.64785516e-01 3.43286395e-01 7.85068929e-01
5.28347433e-01 -8.34987283e-01 -8.23774576e-01 -5.43550253e-01
-1.08654931e-01 7.81493247e-01 -6.40868068e-01 2.68175095e-01
-7.37087667e-01 -3.57023001e-01 1.01229563e-01 2.54189074e-01
2.44322494e-01 -9.04552042e-01 -9.70448554e-01 4.58597273e-01
-1.38636574e-01 -1.02307713e+00 -1.30528137e-01 2.08748564e-01
-4.17909294e-01 -1.31585062e+00 -1.30807962e-02 -7.63605714e-01
1.05555058e+00 5.65501809e-01 8.81077468e-01 3.10305417e-01
1.34563550e-01 4.72623587e-01 -5.30115306e-01 -3.46916348e-01
-2.09558427e-01 2.17787862e-01 -9.51535404e-02 -6.26220405e-01
1.94198802e-01 -3.42600405e-01 -3.79213244e-02 3.58577073e-01
-6.74634278e-01 2.23867610e-01 5.34082912e-02 4.36293155e-01
6.18530631e-01 9.06417370e-01 3.24423969e-01 -1.03029454e+00
5.67706287e-01 -6.17151737e-01 -1.00361490e+00 4.56628382e-01
-3.60294253e-01 -1.23554595e-01 5.50947368e-01 3.47422883e-02
-3.80472630e-01 -3.66628468e-02 4.91583645e-01 2.30501845e-01
2.30099689e-02 7.80996382e-01 -2.70576984e-01 -1.94822267e-01
9.22535807e-02 -6.65457696e-02 3.37625272e-03 1.88474029e-01
1.80555016e-01 -2.34181676e-02 2.82847583e-01 -7.58013129e-01
6.48083985e-01 2.83486664e-01 7.68465698e-01 -2.29448527e-01
9.66736823e-02 -8.97313133e-02 -3.14325035e-01 -1.22016981e-01
5.38615584e-01 -1.93455487e-01 -1.28068674e+00 1.43645346e-01
-1.17045140e+00 -8.02455246e-01 -1.71819791e-01 5.00058606e-02
-8.62917125e-01 1.38646096e-01 -4.64322388e-01 -1.18819809e+00
3.08474749e-01 -1.25597978e+00 6.22016668e-01 1.56999826e-02
-1.67842105e-01 -1.20752788e+00 4.79380786e-01 2.14027286e-01
3.66364270e-01 4.18767840e-01 1.08343089e+00 -7.57451117e-01
-1.04464841e+00 -7.18092844e-02 -4.96582985e-02 -8.75003755e-01
-2.87359115e-02 -1.64478138e-01 6.45807609e-02 -4.97017652e-01
-5.21399677e-01 3.07104170e-01 -3.14226560e-03 6.03650689e-01
3.98117483e-01 -5.66962600e-01 -9.51063514e-01 -7.48371780e-02
1.88397944e+00 7.75880456e-01 6.30147040e-01 1.02566993e+00
-2.53672246e-02 6.69450283e-01 4.86994445e-01 2.17708096e-01
8.39548171e-01 8.07780206e-01 8.48735809e-01 3.10103029e-01
4.29291338e-01 1.47047311e-01 5.89863285e-02 7.61636272e-02
-3.43705535e-01 -1.14451814e+00 -1.03399146e+00 6.34015560e-01
-2.31640768e+00 -1.02696073e+00 -7.18858480e-01 2.07842422e+00
2.60329157e-01 3.23751479e-01 4.91773605e-01 3.19992900e-01
6.96720600e-01 -1.95793688e-01 -1.63986057e-01 -8.93603146e-01
1.32905051e-01 -5.33104353e-02 8.61655593e-01 1.37183499e+00
-6.99192882e-01 8.34670007e-01 6.66435051e+00 1.33945316e-01
-3.02125901e-01 1.67197615e-01 -1.04971327e-01 -3.30851078e-01
-3.00670654e-01 3.06093782e-01 -6.24519706e-01 1.90183371e-01
1.06905377e+00 -3.08397055e-01 1.04955137e+00 5.39187610e-01
3.00918758e-01 -4.23351049e-01 -1.20636797e+00 3.97925407e-01
-2.89672315e-01 -1.34946275e+00 -2.33672395e-01 6.01148546e-01
7.59818852e-01 -2.72725880e-01 -2.65160918e-01 1.62595231e-02
8.06167126e-01 -1.10891354e+00 7.36535370e-01 1.85651973e-01
3.08487684e-01 -1.04125750e+00 7.50704587e-01 5.21085143e-01
-1.30778956e+00 -2.77798891e-01 -1.28300518e-01 -5.82554340e-01
7.35345840e-01 -1.92843273e-01 -9.50096130e-01 8.21378112e-01
2.34177068e-01 -3.42871249e-02 9.92893726e-02 1.29778683e+00
-2.09903926e-01 -1.89345375e-01 -5.21635056e-01 -1.48611605e-01
7.62691677e-01 -4.21053171e-01 7.30054915e-01 8.35409939e-01
2.49194831e-01 3.69351268e-01 5.64942598e-01 6.09088063e-01
6.19029641e-01 -2.51308739e-01 -5.38363934e-01 1.61888182e-01
7.77200341e-01 9.17515278e-01 -1.16361499e+00 -9.51996818e-02
-5.11805475e-01 6.57580554e-01 2.87748069e-01 3.21239084e-01
-9.19940591e-01 -2.49109268e-01 6.41968012e-01 4.48341697e-01
1.96323588e-01 -4.49789494e-01 -4.87002283e-02 -3.45680892e-01
-1.91628292e-01 -6.26633823e-01 4.50147420e-01 -7.52854884e-01
-7.30573297e-01 7.22886980e-01 5.40940404e-01 -6.00002527e-01
-3.59485388e-01 -3.87436420e-01 -6.91107571e-01 8.30091536e-01
-1.54363072e+00 -1.03623700e+00 -1.85410455e-01 7.81801224e-01
2.98356295e-01 -5.88037297e-02 8.61615360e-01 -8.88767987e-02
-6.16508842e-01 1.25570402e-01 -2.47360349e-01 -5.06117940e-01
-2.69099653e-01 -1.32866812e+00 3.41818362e-01 9.93609607e-01
-2.17908129e-01 6.26151562e-01 9.46995616e-01 -7.38814652e-01
-1.79923511e+00 -9.20836985e-01 1.11950123e+00 -2.47093678e-01
6.17819607e-01 -1.11922614e-01 -3.04302096e-01 1.38717282e+00
4.31449920e-01 -2.94998378e-01 2.91916341e-01 -1.19628802e-01
1.18739530e-01 2.91488439e-01 -1.37400985e+00 4.77569401e-01
1.12509000e+00 1.74868539e-01 -2.68756956e-01 6.43560767e-01
5.74879825e-01 -5.56283057e-01 -2.94452667e-01 -1.56300306e-01
3.56555497e-03 -7.15783298e-01 7.49211490e-01 -7.83799648e-01
5.48780337e-02 -8.81972790e-01 -2.97783613e-01 -1.30218983e+00
-8.17240000e-01 -8.99034739e-01 1.21052869e-01 7.62004673e-01
7.99980938e-01 -1.07073033e+00 9.19844270e-01 8.78206432e-01
-2.88019121e-01 -6.10678971e-01 -1.27440715e+00 -1.05538571e+00
-2.34237760e-02 -2.48085499e-01 1.01229227e+00 6.88067615e-01
6.70105278e-01 1.74251333e-01 -2.99895275e-02 7.60013938e-01
5.57950139e-01 1.82581339e-02 6.61663890e-01 -1.23680615e+00
-5.20792782e-01 -5.08265436e-01 -3.82429540e-01 -7.06309140e-01
1.39231533e-01 -8.67852747e-01 -1.46346241e-01 -2.48597813e+00
2.26262376e-01 -6.63405240e-01 2.36547366e-01 7.86100030e-01
8.20962548e-01 -2.87230641e-01 2.33331770e-01 -2.10521907e-01
-9.60580409e-01 -2.30518445e-01 1.27463245e+00 -1.51447728e-01
-3.55780631e-01 6.73420504e-02 -4.24508095e-01 3.44182551e-01
8.14144492e-01 -5.22536159e-01 -9.91554201e-01 -4.75307703e-01
8.21860373e-01 6.62989736e-01 2.88440883e-01 -6.35159612e-01
6.82926178e-01 -1.08048928e+00 -5.65821469e-01 -3.56177717e-01
4.62400943e-01 -1.18502724e+00 8.93642604e-01 7.11639225e-01
-1.82047099e-01 5.87328076e-01 1.06871419e-01 3.67718071e-01
1.13519318e-01 -7.73425400e-01 2.67547846e-01 -4.43710536e-01
-6.57937825e-01 1.91247836e-01 -6.35824025e-01 -3.29038233e-01
1.87625456e+00 -4.28490430e-01 -1.00175571e+00 -5.50445378e-01
-8.24693680e-01 8.24110448e-01 6.08328283e-01 -1.73080564e-01
4.44619805e-01 -1.03402615e+00 -5.13412833e-01 -1.65029943e-01
-1.20688148e-01 1.38714775e-01 1.56543016e-01 8.71599674e-01
-9.80863392e-01 8.39727998e-01 -2.84077734e-01 1.14069097e-01
-1.33602452e+00 1.13011217e+00 4.23550993e-01 -4.33416456e-01
-5.12198091e-01 7.48106003e-01 2.37813070e-02 -2.17450969e-03
4.28971127e-02 -2.52676189e-01 -1.96671739e-01 -4.68267620e-01
6.93513513e-01 7.13693559e-01 -2.25921735e-01 -5.69689810e-01
-7.49362111e-01 3.04296792e-01 -2.99943477e-01 -4.51113999e-01
1.48699558e+00 -3.36780876e-01 -6.32238209e-01 -1.77493140e-01
3.71168882e-01 3.79307158e-02 -6.46783650e-01 1.18827321e-01
-1.39202237e-01 -5.65455496e-01 -4.01578248e-02 -9.52896118e-01
-8.20800841e-01 1.04801551e-01 -3.04820836e-01 9.54819083e-01
8.69938374e-01 1.91337705e-01 4.53599185e-01 3.32996249e-01
1.22645605e+00 -9.48826194e-01 -3.57067287e-01 6.95677042e-01
6.46781445e-01 -4.50690806e-01 2.08572317e-02 -1.02603614e+00
-3.11493248e-01 1.05222487e+00 6.94719374e-01 -4.13138568e-02
1.06288835e-01 6.86494946e-01 -5.24282038e-01 -4.78205383e-01
-1.02644122e+00 -3.02606553e-01 -6.40086651e-01 9.52140570e-01
-5.06169140e-01 2.59012312e-01 -4.99798536e-01 3.88503611e-01
-3.42985183e-01 -8.38085189e-02 1.23284698e+00 1.43767118e+00
-7.16594696e-01 -1.65444481e+00 -5.35328090e-01 2.86826473e-02
7.55445659e-02 1.50221094e-01 -6.27145350e-01 1.12342906e+00
8.15701112e-02 1.46452940e+00 2.33140573e-01 -1.14243671e-01
4.05715317e-01 -4.02394801e-01 9.22875285e-01 -7.01902330e-01
-3.60909641e-01 -2.12297067e-01 7.30772197e-01 -4.28553194e-01
-4.98172343e-01 -8.95112157e-01 -1.65824032e+00 -9.26111996e-01
-2.97671646e-01 5.00529349e-01 4.02087629e-01 8.75420868e-01
-5.07429801e-02 7.15005040e-01 4.18354541e-01 -3.89985323e-01
-9.46101919e-02 -3.57142746e-01 -7.60541677e-01 -4.51590002e-01
8.05772021e-02 -8.06971490e-01 -2.08604276e-01 -5.46811163e-01] | [4.988661289215088, 1.8399194478988647] |
502eb50f-a3b3-4ab5-9870-ce3b7f391228 | prediction-intervals-for-economic-fixed-event | 2210.13562 | null | https://arxiv.org/abs/2210.13562v1 | https://arxiv.org/pdf/2210.13562v1.pdf | Prediction intervals for economic fixed-event forecasts | The fixed-event forecasting setup is common in economic policy. It involves a sequence of forecasts of the same ('fixed') predictand, so that the difficulty of the forecasting problem decreases over time. Fixed-event point forecasts are typically published without a quantitative measure of uncertainty. To construct such a measure, we consider forecast postprocessing techniques tailored to the fixed-event case. We propose heteroscedastic and quantile regression methods that impose parametric constraints motivated by the problem at hand, and use these methods to construct prediction intervals for gross domestic product (GDP) growth in Germany and the US. | ['Hendrik Plett', 'Fabian Krüger'] | 2022-10-24 | null | null | null | null | ['prediction-intervals'] | ['miscellaneous'] | [-3.14127237e-01 5.26872911e-02 -3.18375915e-01 -6.55785978e-01
-9.04138029e-01 -5.92682123e-01 9.77351844e-01 2.69408047e-01
-5.73788695e-02 1.07098413e+00 5.02901435e-01 -9.07547534e-01
-4.01695758e-01 -9.00369585e-01 -4.74990249e-01 -6.93770468e-01
-7.92677477e-02 3.91121805e-01 -2.57736087e-01 9.48552936e-02
5.17149091e-01 3.89905155e-01 -1.34769189e+00 -1.22241415e-01
1.08273268e+00 1.19608176e+00 3.79553959e-02 5.19476950e-01
-1.49800286e-01 7.21223652e-01 -6.27885580e-01 -4.07366693e-01
2.97389209e-01 -1.41294301e-01 -5.86369097e-01 2.01152995e-01
-2.67434508e-01 -1.03637286e-01 1.52359143e-01 9.12175775e-01
1.47755057e-01 3.99786115e-01 1.14952385e+00 -1.26211441e+00
-5.51849544e-01 7.72932231e-01 -4.65329081e-01 3.89093250e-01
2.56590486e-01 -1.62052497e-01 7.19006538e-01 -9.94340956e-01
2.59706914e-01 1.16383612e+00 7.37978756e-01 -1.69439942e-01
-1.43585396e+00 -4.68505293e-01 2.76088834e-01 -3.72998297e-01
-1.13201857e+00 -6.36710763e-01 4.28378046e-01 -1.01810634e+00
7.85371244e-01 5.78915663e-02 3.45055312e-01 9.72180784e-01
7.96818793e-01 -4.29507419e-02 1.21255600e+00 -4.55770612e-01
5.26822746e-01 5.54113928e-03 9.98009071e-02 -3.36753964e-01
2.68933505e-01 4.86325473e-01 -1.08693279e-01 -2.60174751e-01
5.75403273e-01 2.54246771e-01 -1.41248554e-01 1.71328709e-01
-9.39237475e-01 1.17934978e+00 -4.09095258e-01 3.13506484e-01
-8.83191884e-01 -1.32301629e-01 4.85500395e-01 6.03137314e-01
1.31637275e+00 -8.94379243e-02 -8.53515625e-01 -1.20866187e-01
-1.32678998e+00 4.05013353e-01 9.89581168e-01 6.26781821e-01
4.60662335e-01 2.41682723e-01 -3.08690697e-01 4.02763933e-01
1.72774583e-01 8.02685380e-01 2.58164048e-01 -7.89845347e-01
4.57317621e-01 -1.57053530e-01 7.36734927e-01 -1.16284502e+00
-4.63376999e-01 -3.43061000e-01 -7.82052994e-01 2.53798515e-01
7.10610569e-01 -5.40218234e-01 -7.22389102e-01 1.54995799e+00
1.78380147e-01 1.83551252e-01 1.77160829e-01 3.61217678e-01
8.31841007e-02 1.02669621e+00 1.99552462e-01 -7.74538875e-01
9.56662774e-01 -3.97630215e-01 -6.96860254e-01 3.31904590e-01
6.28752112e-01 -7.91118741e-01 5.62937737e-01 4.61959839e-01
-9.93062437e-01 -4.36275750e-01 -2.86316067e-01 4.08843905e-01
-3.77797574e-01 -3.91773395e-02 2.79634088e-01 6.57444894e-01
-9.13783908e-01 8.69536519e-01 -9.37073290e-01 1.71224385e-01
-2.83238739e-01 -1.45904804e-02 9.82290581e-02 3.90791327e-01
-1.07987750e+00 8.80581260e-01 4.04532492e-01 1.17426805e-01
-4.74563211e-01 -9.46442187e-01 -7.32565820e-01 3.53637427e-01
9.24389139e-02 -4.96458083e-01 1.53157890e+00 -8.86825681e-01
-1.61783910e+00 3.21829855e-01 -1.64229572e-01 -4.24149901e-01
4.78773743e-01 1.36058912e-01 -7.87395656e-01 -3.72338086e-01
1.72580853e-01 -2.27298424e-01 8.80183518e-01 -8.61189663e-01
-8.02928150e-01 -2.79893219e-01 -4.26579773e-01 -2.27298036e-01
3.40885639e-01 4.74963188e-01 6.37456000e-01 -1.02230966e+00
-7.47321695e-02 -6.29333794e-01 -4.39696580e-01 -9.24079657e-01
-3.37663323e-01 -4.47747350e-01 2.99016088e-01 -1.03388536e+00
1.32378745e+00 -2.16582227e+00 -2.83914387e-01 6.06567800e-01
-3.01828027e-01 -4.71738428e-01 2.90053427e-01 8.26024055e-01
-4.86402988e-01 1.72872454e-01 -1.49966151e-01 -4.75799233e-01
2.18753234e-01 1.75545767e-01 -8.84650826e-01 6.81915879e-01
1.32600829e-01 4.24513251e-01 -5.77050447e-01 -4.60694991e-02
2.66260475e-01 1.12028480e-01 -2.39566773e-01 1.01491354e-01
-2.27755122e-02 6.44058466e-01 -3.66783410e-01 3.03659946e-01
7.53593087e-01 -2.86785513e-01 4.94019426e-02 3.54088783e-01
-9.80711699e-01 4.92481232e-01 -1.20929301e+00 1.03995728e+00
-4.95251060e-01 3.08876216e-01 -2.38310754e-01 -1.47268891e+00
1.11190593e+00 5.49281955e-01 5.70105314e-01 -3.70544344e-01
3.80302854e-02 2.41112769e-01 -3.23502839e-01 -2.25057349e-01
4.45728421e-01 -5.25911570e-01 -1.89096481e-01 1.89666808e-01
-2.69575287e-02 -1.05309449e-01 2.04062387e-01 -2.69947380e-01
6.11074328e-01 -1.57238498e-01 6.59651935e-01 -8.54024649e-01
1.08222194e-01 -2.81346608e-02 7.78467596e-01 6.94761813e-01
2.75551826e-01 2.32084423e-01 7.59993613e-01 -5.30851722e-01
-1.22296929e+00 -9.51175690e-01 -4.90853995e-01 1.05573189e+00
-8.45628977e-01 -4.16839682e-02 -2.79004782e-01 -2.19300836e-01
4.12663549e-01 1.18428147e+00 -6.92158818e-01 4.28797871e-01
-9.98890474e-02 -8.11170578e-01 -2.76461214e-01 5.02206087e-01
-4.26406115e-02 -6.63920403e-01 -5.40827096e-01 5.61455607e-01
1.60259947e-01 -6.45298660e-01 -2.29562730e-01 2.81785280e-01
-7.61721194e-01 -7.01793551e-01 -9.45245802e-01 -2.19243333e-01
2.89065301e-01 -2.92278856e-01 1.48736727e+00 -6.57735169e-01
5.92858851e-01 2.69447953e-01 -1.98083460e-01 -7.31463730e-01
-2.53549844e-01 -2.27297768e-01 5.49859889e-02 1.66390650e-02
1.11614339e-01 -6.14399076e-01 -4.98189658e-01 6.90075606e-02
-6.69271231e-01 -3.99873525e-01 3.20104882e-02 9.03740287e-01
4.39633936e-01 3.07987422e-01 1.10321915e+00 -6.95893466e-01
6.81854129e-01 -9.74960089e-01 -1.32675660e+00 3.18446338e-01
-8.56524289e-01 -1.88083798e-01 6.61039114e-01 -4.47400540e-01
-1.38218558e+00 -1.65426955e-01 1.87174737e-01 -1.88026458e-01
-1.16768591e-01 1.30998373e+00 1.97332487e-01 3.01959306e-01
2.78773010e-01 -1.34149328e-01 -3.12874407e-01 -6.10794842e-01
1.51143715e-01 5.06810844e-01 7.31121004e-01 -5.42497635e-01
5.72336018e-01 1.97304562e-01 1.28249437e-01 -3.79752845e-01
-7.64252841e-01 -5.07822633e-01 -4.42437232e-01 -1.43909514e-01
6.41898334e-01 -1.02627432e+00 -7.72951663e-01 2.64492363e-01
-1.12639165e+00 -3.75828683e-01 -6.04172409e-01 8.97167087e-01
-7.59585917e-01 -1.82884365e-01 -3.34479988e-01 -1.49205577e+00
-7.39006996e-02 -8.30112457e-01 8.18065345e-01 2.50271112e-01
-1.20560549e-01 -1.53023148e+00 5.03622234e-01 -5.85282683e-01
4.17433739e-01 5.45067728e-01 7.54171014e-01 -8.77545953e-01
5.23948260e-02 -2.01888174e-01 -2.30792195e-01 2.91015059e-01
1.70688592e-02 3.46030533e-01 -7.02543497e-01 -1.19550221e-01
3.35903734e-01 1.53043553e-01 6.29229188e-01 1.24238992e+00
1.15873742e+00 -6.23261511e-01 -1.98693231e-01 2.55618215e-01
1.46549392e+00 4.64325219e-01 5.87811887e-01 1.68811411e-01
-1.53004512e-01 8.41372192e-01 4.95134979e-01 1.13545144e+00
4.46237773e-01 3.73411447e-01 -1.35811076e-01 1.29358377e-02
8.52526784e-01 -1.93627745e-01 1.36713713e-01 6.92487597e-01
-2.46695906e-01 -3.26996356e-01 -1.26027048e+00 7.47596860e-01
-1.78672540e+00 -1.28493392e+00 -8.96759182e-02 2.36193848e+00
6.60718918e-01 9.45142955e-02 4.49998707e-01 9.25886333e-02
7.15258002e-01 8.27300847e-02 -1.08937941e-01 -7.19313443e-01
-8.70300084e-02 1.96547404e-01 7.81106830e-01 5.58371663e-01
-9.01792169e-01 2.23950833e-01 7.74402142e+00 7.41997659e-01
-9.65346634e-01 -2.35254969e-02 1.21185160e+00 4.64999765e-01
-5.16389072e-01 2.26501331e-01 -8.71025443e-01 7.89683282e-01
1.72757196e+00 -7.47517884e-01 6.41983673e-02 9.56130087e-01
8.00386250e-01 -2.98132330e-01 -8.59616697e-01 5.84125578e-01
-6.10868275e-01 -1.17397535e+00 -4.93345320e-01 3.73114258e-01
9.76239383e-01 -2.19761565e-01 4.18673396e-01 4.40555036e-01
7.66932428e-01 -1.02511632e+00 6.67807400e-01 1.12024486e+00
7.23731995e-01 -1.15177023e+00 8.29277635e-01 4.63530719e-01
-9.95686829e-01 -3.16235572e-01 -4.77711678e-01 -5.06488085e-01
5.26019454e-01 1.43910909e+00 -5.15842140e-01 6.93739712e-01
5.04486203e-01 5.65655768e-01 1.53239250e-01 1.03334451e+00
2.30697855e-01 9.76624787e-01 -3.02582771e-01 4.20749664e-01
2.82618791e-01 -6.80754364e-01 1.87642336e-01 1.01018608e+00
1.04027832e+00 4.12401974e-01 4.23513800e-01 7.10275590e-01
1.95193961e-01 2.44359836e-01 -6.63021922e-01 6.58316091e-02
5.04943490e-01 6.11448944e-01 -6.26036048e-01 -4.76043642e-01
-8.20886970e-01 2.59744346e-01 -7.31037483e-02 5.74037552e-01
-5.51579714e-01 -2.06548661e-01 3.61967713e-01 3.70618366e-02
2.58853972e-01 -4.24163699e-01 -5.42638958e-01 -1.07459521e+00
-1.31001532e-01 -5.01022637e-01 4.52711225e-01 -4.48535234e-01
-1.57546592e+00 1.05195157e-01 3.86133075e-01 -1.00599933e+00
-9.87256348e-01 -4.93124068e-01 -9.89285409e-01 1.36577392e+00
-1.42607045e+00 -5.84662020e-01 3.89469236e-01 4.39130843e-01
3.68021846e-01 3.36158387e-02 7.85507441e-01 -1.84028104e-01
-5.58863699e-01 -5.09100817e-02 9.86689150e-01 -2.06818119e-01
4.20302987e-01 -1.34212720e+00 5.48672676e-01 8.49350512e-01
-3.41060132e-01 3.76425415e-01 9.87372339e-01 -9.18282330e-01
-5.77534497e-01 -8.86549950e-01 1.39632475e+00 -3.06582242e-01
1.11880445e+00 2.96788965e-03 -8.53770912e-01 8.73897731e-01
1.11418769e-01 3.30013931e-02 5.41795075e-01 4.82273549e-01
-1.52785763e-01 7.65030757e-02 -1.10551322e+00 -3.33133265e-02
2.09157467e-01 -2.88541913e-01 -5.38942695e-01 4.55666810e-01
5.89606345e-01 -1.05486088e-01 -1.46193492e+00 2.98934549e-01
4.70062256e-01 -6.86029077e-01 6.88632607e-01 -7.84755290e-01
3.23609680e-01 2.10972622e-01 -1.90252393e-01 -1.77468038e+00
-4.19439375e-01 -8.45056772e-01 1.22652784e-01 1.42845595e+00
4.96235520e-01 -1.11866236e+00 2.11693987e-01 8.35499823e-01
-2.00953946e-01 -3.05896491e-01 -1.16651487e+00 -1.20345616e+00
6.81106925e-01 -4.27951813e-01 9.96923804e-01 1.01825070e+00
1.24787427e-01 -2.61998445e-01 -4.31624442e-01 3.32753301e-01
5.57861328e-01 3.94683510e-01 4.99721825e-01 -1.70974839e+00
-5.41756898e-02 -3.45430583e-01 -8.31457675e-02 -7.52786934e-01
4.01813179e-01 -2.88800925e-01 2.09762938e-02 -9.13394272e-01
2.25073006e-03 -3.25883180e-01 -2.30541795e-01 -3.29605155e-02
2.43015047e-02 -4.41094667e-01 -1.39581963e-01 8.80805030e-02
1.40565008e-01 6.46611273e-01 6.20782912e-01 3.25035423e-01
-3.16337794e-01 3.92564803e-01 -3.78434449e-01 7.10531771e-01
9.33238089e-01 -5.95258296e-01 -2.51812249e-01 -3.42930779e-02
3.75577658e-01 8.03457260e-01 4.61508483e-01 -5.64982235e-01
1.60349518e-01 -8.88613939e-01 5.61159134e-01 -9.98283088e-01
-2.81134486e-01 -6.07262254e-01 6.13268673e-01 1.64798990e-01
-2.51424730e-01 6.10120595e-01 -3.03624570e-02 6.63570285e-01
-3.95702332e-01 -2.59903848e-01 6.02387726e-01 1.45335510e-01
3.88132073e-02 1.61221594e-01 -6.13124669e-01 7.01841936e-02
7.91927695e-01 6.79503605e-02 -2.70773824e-02 -7.01173842e-01
-1.08097863e+00 2.34197661e-01 2.34283119e-01 -3.05535138e-01
7.06612244e-02 -1.19006193e+00 -1.11674464e+00 -1.31054088e-01
-2.32555047e-01 -2.58382529e-01 1.38161644e-01 1.11152518e+00
2.11709440e-01 7.57750690e-01 2.15884056e-02 -2.94934660e-01
-3.44316185e-01 7.54251122e-01 3.17501128e-01 -4.68787014e-01
-5.50424576e-01 3.50040376e-01 2.09421128e-01 -1.06374301e-01
-2.74539083e-01 -6.64609909e-01 -2.97089309e-01 4.86143500e-01
6.06054604e-01 6.37154222e-01 1.13650754e-01 -4.87760484e-01
4.90159579e-02 1.22453973e-01 2.65717566e-01 -5.41460693e-01
1.41323864e+00 -6.39701724e-01 1.41503528e-01 1.05694294e+00
8.07331979e-01 2.27682781e-03 -1.34841347e+00 -1.06762880e-02
5.53634703e-01 -4.70223725e-01 1.72532603e-01 -5.95752120e-01
-8.31498265e-01 5.50448418e-01 1.51070654e-01 8.27072501e-01
1.07006729e+00 -4.17554766e-01 3.73271614e-01 -6.01248331e-02
1.75379485e-01 -1.06591082e+00 -8.39970767e-01 5.51803231e-01
1.11594570e+00 -1.15306997e+00 -1.44266799e-01 -7.34552294e-02
-3.54812056e-01 1.22908580e+00 -4.54202056e-01 -2.55886674e-01
1.31614363e+00 3.17767262e-01 -1.99261338e-01 1.07229061e-01
-8.96237910e-01 -3.87440771e-02 5.18374324e-01 5.24063051e-01
5.17175019e-01 4.97836381e-01 -7.57115483e-01 7.76974976e-01
-3.66094410e-01 6.00403808e-02 4.57810283e-01 6.13254964e-01
-3.34561467e-01 -8.10419619e-01 -6.13049686e-01 6.65026963e-01
-7.33434737e-01 -1.86800450e-01 3.33206177e-01 7.45730102e-01
-3.86001259e-01 1.21126509e+00 3.93202782e-01 2.17436939e-01
2.18256995e-01 4.10432756e-01 -2.91965604e-01 -3.73005688e-01
-3.20854723e-01 3.35309476e-01 2.98947901e-01 -2.99861580e-01
-4.22623754e-01 -1.28663635e+00 -7.75381625e-01 -8.19184840e-01
-3.41080487e-01 3.13778996e-01 7.96408832e-01 1.19767225e+00
8.05999339e-02 2.07595304e-01 1.22774625e+00 -9.39062238e-01
-1.00858510e+00 -1.12256539e+00 -1.05651820e+00 1.96660697e-01
3.97310942e-01 -8.63362908e-01 -7.81450152e-01 1.95834637e-01] | [6.002164840698242, 3.9471030235290527] |
82723c2f-1956-41ec-a9f8-2bb0af7eb821 | limit-distribution-theory-for-the-smooth-1 | 2107.13494 | null | https://arxiv.org/abs/2107.13494v5 | https://arxiv.org/pdf/2107.13494v5.pdf | Limit Distribution Theory for the Smooth 1-Wasserstein Distance with Applications | The smooth 1-Wasserstein distance (SWD) $W_1^\sigma$ was recently proposed as a means to mitigate the curse of dimensionality in empirical approximation while preserving the Wasserstein structure. Indeed, SWD exhibits parametric convergence rates and inherits the metric and topological structure of the classic Wasserstein distance. Motivated by the above, this work conducts a thorough statistical study of the SWD, including a high-dimensional limit distribution result for empirical $W_1^\sigma$, bootstrap consistency, concentration inequalities, and Berry-Esseen type bounds. The derived nondegenerate limit stands in sharp contrast with the classic empirical $W_1$, for which a similar result is known only in the one-dimensional case. We also explore asymptotics and characterize the limit distribution when the smoothing parameter $\sigma$ is scaled with $n$, converging to $0$ at a sufficiently slow rate. The dimensionality of the sampled distribution enters empirical SWD convergence bounds only through the prefactor (i.e., the constant). We provide a sharp characterization of this prefactor's dependence on the smoothing parameter and the intrinsic dimension. This result is then used to derive new empirical convergence rates for classic $W_1$ in terms of the intrinsic dimension. As applications of the limit distribution theory, we study two-sample testing and minimum distance estimation (MDE) under $W_1^\sigma$. We establish asymptotic validity of SWD testing, while for MDE, we prove measurability, almost sure convergence, and limit distributions for optimal estimators and their corresponding $W_1^\sigma$ error. Our results suggest that the SWD is well suited for high-dimensional statistical learning and inference. | ['Kengo Kato', 'Ziv Goldfeld', 'Ritwik Sadhu'] | 2021-07-28 | null | null | null | null | ['hypothesis-testing', 'hypothesis-testing'] | ['methodology', 'miscellaneous'] | [-1.93858035e-02 1.05479792e-01 -7.10166618e-02 -2.70547092e-01
-8.35319221e-01 -3.26199293e-01 1.33767068e-01 1.69029206e-01
-7.64360666e-01 9.81839120e-01 -3.18822861e-01 -3.30207229e-01
-8.75005007e-01 -8.18698108e-01 -6.90023601e-01 -1.18170202e+00
-5.04000247e-01 1.55559301e-01 1.17589533e-01 -4.93356772e-02
2.96920329e-01 3.56711030e-01 -1.44474781e+00 -8.46134126e-01
1.14896417e+00 1.17234612e+00 -7.76841417e-02 7.35645056e-01
7.84005523e-02 -5.93419522e-02 -2.70992398e-01 -2.92115122e-01
3.82838249e-01 -6.82112217e-01 -6.01877332e-01 -3.56376261e-01
4.50265765e-01 -1.01656251e-01 1.02742732e-01 1.54253948e+00
5.10482848e-01 4.41264838e-01 9.36473846e-01 -1.09776199e+00
-6.86730921e-01 4.04724509e-01 -6.68727398e-01 2.72041500e-01
7.55374786e-03 -1.66055143e-01 9.79203045e-01 -8.07470918e-01
4.17079747e-01 9.82797265e-01 8.95894825e-01 5.50872803e-01
-1.49990094e+00 -5.54042041e-01 -3.62016588e-01 -3.09209704e-01
-1.56030405e+00 -3.94242778e-02 5.20396888e-01 -6.68427467e-01
2.23564908e-01 -1.35013647e-02 3.67262125e-01 7.46547937e-01
3.69083077e-01 3.38007867e-01 1.11968541e+00 -7.30309486e-01
5.30523479e-01 1.79918423e-01 5.04305303e-01 1.00765240e+00
8.44166696e-01 1.51681483e-01 -1.79397330e-01 -8.74637738e-02
9.21654284e-01 -2.94363856e-01 -2.77967483e-01 -6.96904123e-01
-8.46424878e-01 9.18775141e-01 4.88444492e-02 4.54738736e-01
1.30580980e-02 2.38811642e-01 1.83465883e-01 3.53061229e-01
6.48537397e-01 3.11799109e-01 -2.84473181e-01 -2.98284560e-01
-7.45502830e-01 2.47900382e-01 6.55686557e-01 9.24775362e-01
6.60319388e-01 -6.76131248e-02 -4.13507074e-02 6.79846168e-01
2.77621746e-01 1.01833916e+00 7.03793392e-02 -1.04139578e+00
3.89373302e-01 1.12723343e-01 4.67410803e-01 -7.73197711e-01
-4.78916287e-01 -5.08697271e-01 -1.05177796e+00 1.37975052e-01
1.04293692e+00 -3.97134721e-02 -2.94110566e-01 2.23468757e+00
3.64773631e-01 -8.17309879e-03 -1.51899502e-01 6.59851670e-01
-8.47303346e-02 2.16764733e-01 -6.57490268e-02 -5.66244900e-01
9.87115085e-01 -1.41425768e-03 -5.74042618e-01 3.30286831e-01
8.24087024e-01 -3.50455135e-01 1.43801117e+00 2.39409670e-01
-1.14868402e+00 -3.55855942e-01 -1.05476272e+00 2.17950106e-01
-1.85539335e-01 -2.95390159e-01 1.30579665e-01 9.66975808e-01
-8.35648358e-01 1.06298709e+00 -8.40064883e-01 -2.89113164e-01
2.97039390e-01 1.11805901e-01 -1.90359980e-01 -5.88962398e-02
-1.08555019e+00 7.11663544e-01 3.92932678e-03 2.03501865e-01
-3.83623958e-01 -9.94548023e-01 -7.41964519e-01 -2.04692900e-01
8.68895184e-03 -3.07109654e-01 6.57800198e-01 -5.34732454e-02
-1.22288895e+00 6.48590922e-01 -1.23076349e-01 -4.06493038e-01
7.72653937e-01 -7.33375549e-02 -3.17450345e-01 1.33778349e-01
2.85949290e-01 -1.42791137e-01 6.66892290e-01 -6.66409850e-01
-2.60927886e-01 -7.29195356e-01 -3.53221774e-01 -3.40902716e-01
-3.56027961e-01 -4.87891853e-01 3.98590952e-01 -4.08917397e-01
1.81640536e-01 -6.74184024e-01 -7.88629651e-02 2.86934972e-02
-1.56903535e-01 -2.10871652e-01 2.82316536e-01 -1.70486316e-01
1.27752495e+00 -2.33645511e+00 2.24972777e-02 5.24059236e-01
2.33314350e-01 -8.98638144e-02 -1.02383107e-01 2.91548878e-01
9.01288912e-02 2.49033034e-01 -5.81293225e-01 -6.26624823e-02
3.94324273e-01 -9.43021104e-02 -7.84768164e-03 1.19428408e+00
2.90217381e-02 4.70869958e-01 -1.12834370e+00 -3.92021000e-01
5.22632077e-02 3.91810924e-01 -6.37328148e-01 -1.41688868e-01
1.61236569e-01 2.13219836e-01 -4.67405409e-01 1.72393456e-01
1.09251237e+00 -2.18506768e-01 -5.97558245e-02 -8.53779837e-02
-2.40463898e-01 -2.92513877e-01 -1.35271406e+00 1.24000001e+00
-2.16758043e-01 4.39386040e-01 3.43620986e-01 -1.44207060e+00
1.01192880e+00 -1.36753991e-01 5.12700260e-01 -5.69461048e-01
2.26460576e-01 5.79864502e-01 -1.35023147e-01 -4.28891480e-01
3.20023566e-01 -9.44101334e-01 -9.75809395e-02 3.31743956e-01
-1.08279735e-01 -2.34333836e-02 2.58095145e-01 -1.12257272e-01
1.03033471e+00 2.83721816e-02 1.90854788e-01 -1.16905832e+00
4.47916806e-01 -3.90473485e-01 4.07607853e-01 9.04208064e-01
-4.81856197e-01 2.41342008e-01 9.12744761e-01 2.65236162e-02
-1.16141164e+00 -1.51638305e+00 -1.09269512e+00 6.88719153e-01
3.98075640e-01 -2.09226757e-02 -9.16610479e-01 -3.82157981e-01
3.59579474e-01 7.35431671e-01 -1.06283283e+00 -2.65127420e-01
-3.09717357e-01 -1.00117660e+00 5.18021345e-01 4.11581278e-01
5.82015574e-01 -3.31452221e-01 -4.81735140e-01 -6.51262626e-02
2.65834033e-01 -7.51217306e-01 -5.80957711e-01 1.37831882e-01
-9.62604344e-01 -1.30493212e+00 -8.36249650e-01 -3.40410084e-01
5.22203743e-01 -1.79607242e-01 6.92964375e-01 -4.26075131e-01
-2.54717767e-01 5.33349216e-01 -6.06231168e-02 -9.62534249e-02
-3.46492708e-01 -1.60083726e-01 3.86574417e-01 -9.24146548e-02
5.24274945e-01 -5.48883557e-01 -8.04530621e-01 6.17532313e-01
-9.20742452e-01 -7.65046537e-01 2.43807703e-01 8.97030473e-01
6.29694104e-01 7.99838677e-02 7.32648969e-01 -5.19648731e-01
7.31318831e-01 -4.41622078e-01 -1.03005493e+00 2.48569399e-02
-9.26823974e-01 6.52831614e-01 7.65487313e-01 -4.53191996e-01
-4.62454945e-01 -5.88694274e-01 3.48558165e-02 -1.84293330e-01
2.03654245e-01 2.80848920e-01 7.02118948e-02 -4.98469584e-02
6.85567677e-01 9.93892923e-02 2.95681417e-01 -6.23989105e-01
1.45651340e-01 5.90697527e-01 4.74014193e-01 -9.49306309e-01
6.81860745e-01 5.68225026e-01 5.73309958e-01 -1.09942007e+00
-1.06773126e+00 -1.95726305e-01 -6.10772491e-01 2.95859743e-02
7.39943445e-01 -1.88975245e-01 -1.12196624e+00 1.15135148e-01
-5.44031620e-01 -3.37855965e-01 -7.89751768e-01 8.58999074e-01
-8.86888742e-01 4.77597296e-01 -4.19135541e-01 -1.34523571e+00
7.82499015e-02 -7.91038454e-01 7.41216242e-01 -2.00269684e-01
1.08283181e-02 -1.30516696e+00 3.65184665e-01 -1.89884543e-01
4.41065431e-01 5.90157986e-01 1.11060846e+00 -3.75864416e-01
3.90947908e-02 -3.75739843e-01 -3.29946458e-01 6.19759321e-01
-6.81938604e-02 -2.08627984e-01 -4.25002724e-01 -4.74961519e-01
3.16107929e-01 1.34037867e-01 6.90949440e-01 7.92332411e-01
8.70608866e-01 -2.14616105e-01 -5.20266853e-02 4.71854866e-01
1.62008393e+00 -7.45477900e-02 3.82944196e-01 1.34065419e-01
7.91309476e-02 5.83813787e-01 4.45523679e-01 5.77345371e-01
-9.28948671e-02 2.14201301e-01 1.33210659e-01 3.49960446e-01
2.81699121e-01 -2.25787967e-01 3.26597035e-01 8.49232256e-01
-5.47489412e-02 3.96917075e-01 -6.42219543e-01 4.66628015e-01
-1.42912877e+00 -9.91545081e-01 -2.30205044e-01 2.99705315e+00
8.39971125e-01 2.40788117e-01 4.64826614e-01 2.25693405e-01
8.92545581e-01 -7.92333484e-02 -7.01392829e-01 -2.81444579e-01
-9.45678577e-02 6.38956189e-01 8.84047806e-01 8.81660998e-01
-7.01537371e-01 1.94084555e-01 6.66643238e+00 9.81548786e-01
-6.05916739e-01 2.25204080e-01 3.08276266e-01 -7.10627660e-02
-3.89948159e-01 -1.55274391e-01 -8.88841510e-01 7.83470631e-01
1.04689801e+00 -3.11047941e-01 1.20156504e-01 9.24113929e-01
2.82689542e-01 -3.13728511e-01 -1.02442670e+00 8.14354062e-01
-3.16215962e-01 -1.05214286e+00 -4.81260598e-01 6.67464554e-01
6.73992336e-01 -1.83984488e-01 1.29906207e-01 8.81614387e-02
1.11483835e-01 -9.99415576e-01 4.02008682e-01 5.27120531e-01
1.18383002e+00 -1.01349103e+00 8.60656559e-01 3.69415879e-01
-1.26074231e+00 -3.06848735e-02 -6.24348283e-01 4.29864638e-02
-2.93023791e-02 1.16513383e+00 -1.28326952e-01 2.83571154e-01
5.03463864e-01 5.48368216e-01 -1.28255457e-01 8.23181748e-01
3.30922544e-01 3.89938980e-01 -5.75341225e-01 -3.91599089e-01
1.87495351e-01 -8.06611657e-01 5.62918782e-01 1.08024502e+00
7.53928065e-01 8.89958516e-02 -3.51836801e-01 9.65560734e-01
4.68820222e-02 1.58530235e-01 -5.66489160e-01 -1.32235065e-01
6.03563547e-01 6.06872499e-01 -5.01335382e-01 -2.47081406e-02
-1.29803315e-01 4.94801879e-01 1.25082478e-01 3.69339406e-01
-7.38230109e-01 -1.08160853e+00 8.68878424e-01 4.77724373e-01
4.71261144e-01 -5.66826522e-01 -2.64058918e-01 -1.10011387e+00
3.62046361e-01 -1.73354432e-01 3.51438016e-01 -5.21523878e-02
-1.52002776e+00 1.98554233e-01 2.91820675e-01 -1.12975073e+00
-7.49502257e-02 -9.10758674e-01 -2.78585315e-01 9.33136344e-01
-1.19458532e+00 -1.70771033e-01 2.11382136e-01 3.56016636e-01
-2.77244836e-01 1.64888993e-01 6.44696653e-01 9.55964625e-02
-6.44155562e-01 8.20340574e-01 9.52235639e-01 1.38499662e-01
3.67562115e-01 -1.41686106e+00 -1.12775527e-01 7.12861955e-01
-5.03721893e-01 7.94011712e-01 1.07951605e+00 -5.19654810e-01
-1.42539287e+00 -6.75949097e-01 6.35307431e-01 -3.39585364e-01
1.14108193e+00 -4.14770365e-01 -9.11786973e-01 3.25261354e-01
-4.70211357e-01 2.70464957e-01 7.34058201e-01 1.96942165e-01
-2.28846028e-01 -3.52246284e-01 -1.67982054e+00 4.64678079e-01
1.11499310e+00 -4.15079117e-01 -3.88583481e-01 4.03449833e-02
5.45431137e-01 1.97742999e-01 -1.39946091e+00 3.59335870e-01
6.72483146e-01 -1.10067511e+00 8.25329483e-01 -5.90958893e-01
1.56641960e-01 -6.79757027e-03 -4.96436179e-01 -9.33291912e-01
-8.13216642e-02 -8.96065235e-01 4.93180268e-02 8.97574663e-01
3.24213922e-01 -1.06062889e+00 5.89542568e-01 5.82905889e-01
2.29479313e-01 -8.94994974e-01 -1.50800312e+00 -1.39974904e+00
8.08699787e-01 -6.43787026e-01 1.67415306e-01 6.93823814e-01
2.65669316e-01 -1.36256039e-01 -4.58253287e-02 3.99473403e-03
1.44238532e+00 -1.68196380e-01 3.22048008e-01 -1.39661598e+00
-3.46395940e-01 -5.63779294e-01 -4.99039471e-01 -1.21104538e+00
6.51749969e-02 -7.74914980e-01 1.09385014e-01 -8.70222211e-01
-1.22822113e-02 -5.15408814e-01 -4.06952411e-01 -3.76088530e-01
1.67678103e-01 -3.95544693e-02 -2.64137238e-01 -3.41408364e-02
-3.65199596e-01 6.42110467e-01 1.24642599e+00 3.82960320e-01
-2.10707933e-01 3.99319045e-02 -5.67904174e-01 7.36772656e-01
5.33417642e-01 -4.13972437e-01 -3.37232232e-01 7.54147246e-02
3.89254600e-01 2.62849107e-02 2.49131501e-01 -9.20627177e-01
7.36842025e-03 -2.58472353e-01 1.29666179e-01 -2.67808110e-01
1.05008155e-01 -3.75371575e-01 -3.76289934e-01 4.52026784e-01
-6.16965294e-01 -2.57309675e-01 -1.77876100e-01 9.13495123e-01
1.99560374e-01 -5.42336047e-01 1.33905876e+00 2.45068207e-01
7.60308579e-02 2.86435217e-01 -2.03881457e-01 7.13823140e-01
7.95058370e-01 -2.39850223e-01 -2.39927754e-01 -4.83185872e-02
-7.26287007e-01 1.52435405e-02 3.89790714e-01 -3.40480536e-01
4.79423851e-01 -1.42704129e+00 -6.39425159e-01 3.99094552e-01
-6.21506236e-02 -1.79275975e-01 3.04251760e-01 1.26305199e+00
-3.30938846e-01 1.42393783e-01 1.72115043e-01 -6.17509067e-01
-5.31504750e-01 3.94831061e-01 4.09738809e-01 5.53428084e-02
-5.30202091e-01 6.24511480e-01 -1.38726279e-01 -9.65307653e-02
2.46508330e-01 -6.13093853e-01 5.59140027e-01 -1.24501728e-01
6.20116830e-01 8.12059820e-01 -2.48812407e-01 -1.37959570e-01
-3.26535344e-01 9.78643179e-01 3.20849746e-01 -3.09739441e-01
1.28356135e+00 -3.05988967e-01 -1.51847139e-01 8.38336945e-01
1.60498714e+00 7.18180016e-02 -1.45459235e+00 -2.50920713e-01
1.87597796e-01 -5.72944641e-01 -2.59168923e-01 -8.42842683e-02
-5.24033189e-01 9.27574098e-01 6.93892241e-01 6.86620474e-01
5.96654296e-01 2.68712461e-01 5.31230092e-01 2.10593238e-01
3.65993142e-01 -1.35843694e+00 -6.81810752e-02 4.44061786e-01
5.74833989e-01 -9.73048866e-01 -2.01098010e-01 -3.29548605e-02
3.86963412e-02 1.04875147e+00 2.44217053e-01 -4.00679946e-01
1.07198191e+00 2.72912234e-01 -5.94004214e-01 4.93449718e-03
-1.46244839e-01 -2.01920494e-02 1.49450570e-01 4.87185091e-01
4.78333592e-01 5.23155518e-02 -6.86033309e-01 5.51646829e-01
-2.70860076e-01 -9.16250870e-02 3.68682176e-01 5.71874380e-01
-8.38154733e-01 -7.55705655e-01 -1.22824453e-01 4.83738899e-01
-2.72905946e-01 2.11650915e-02 2.00416133e-01 1.10665143e+00
-9.11115929e-02 6.69965804e-01 2.36578003e-01 -1.09933697e-01
2.58399099e-01 2.34305754e-01 7.60405719e-01 -4.10511568e-02
4.01337236e-01 -1.34059295e-01 -4.60486799e-01 -4.80165094e-01
-4.01822180e-01 -6.54772639e-01 -1.14044142e+00 -7.24738121e-01
-2.76235908e-01 6.61711454e-01 5.79557836e-01 1.02265978e+00
3.44428793e-02 -1.77151844e-01 7.04156220e-01 -3.01464856e-01
-1.20838749e+00 -9.68608081e-01 -1.38354218e+00 2.15779603e-01
5.68391025e-01 -7.53242850e-01 -1.10811746e+00 -5.18638074e-01] | [7.157359600067139, 4.1748433113098145] |
c9eed412-58bc-462a-ad64-1ed897db8b27 | evaluating-gaussian-grasp-maps-for-generative | 2206.00432 | null | https://arxiv.org/abs/2206.00432v1 | https://arxiv.org/pdf/2206.00432v1.pdf | Evaluating Gaussian Grasp Maps for Generative Grasping Models | Generalising robotic grasping to previously unseen objects is a key task in general robotic manipulation. The current method for training many antipodal generative grasping models rely on a binary ground truth grasp map generated from the centre thirds of correctly labelled grasp rectangles. However, these binary maps do not accurately reflect the positions in which a robotic arm can correctly grasp a given object. We propose a continuous Gaussian representation of annotated grasps to generate ground truth training data which achieves a higher success rate on a simulated robotic grasping benchmark. Three modern generative grasping networks are trained with either binary or Gaussian grasp maps, along with recent advancements from the robotic grasping literature, such as discretisation of grasp angles into bins and an attentional loss function. Despite negligible difference according to the standard rectangle metric, Gaussian maps better reproduce the training data and therefore improve success rates when tested on the same simulated robot arm by avoiding collisions with the object: achieving 87.94\% accuracy. Furthermore, the best performing model is shown to operate with a high success rate when transferred to a real robotic arm, at high inference speeds, without the need for transfer learning. The system is then shown to be capable of performing grasps on an antagonistic physical object dataset benchmark. | ['Ulrik Beierholm', 'Magnus Bordewich', 'Toby P. Breckon', 'William Prew'] | 2022-06-01 | null | null | null | null | ['robotic-grasping'] | ['robots'] | [ 0.3776438 0.40174103 0.1992215 -0.29342386 -0.5596069 -0.76174814
0.40082398 -0.27362096 -0.16188543 0.60506135 -0.4277556 0.06327792
-0.5036291 -0.8862594 -1.3637611 -1.0589532 -0.25414765 1.1064075
0.10781762 -0.04136741 0.36289203 0.78832376 -1.679607 0.38867986
0.4802865 1.0507485 0.7452062 0.64338905 0.41336417 0.12247776
-0.7375488 -0.28010628 0.49904546 0.23422658 -0.7918258 -0.17294171
0.24119301 -0.66634697 -0.25008214 0.7426759 0.35939473 -0.15837222
1.0303103 -1.493768 -0.9737909 0.9190355 -0.14836262 -0.5301747
0.16411315 0.11688442 0.75046533 -0.7153308 0.7983064 1.4480162
0.5648089 0.6300894 -1.2025639 -0.4729316 -0.27540913 -0.38197762
-0.906904 0.06835578 0.56924105 -0.34777242 0.889825 0.0307109
0.304305 1.5523709 0.56610286 0.6999807 0.85831434 -0.28877005
0.30236268 -0.35859197 -0.26790503 0.37945196 0.46104923 0.13954324
-0.04799023 -0.12506628 1.1915933 0.04725721 -0.10636193 -1.1357394
-1.4568366 0.7373758 0.95277536 0.27301183 -0.74984294 0.55181664
0.22916114 -0.06453195 -0.04160717 0.78607535 -0.33854678 0.04154851
-0.18789573 0.73382854 1.0677098 1.5736227 0.3094033 -0.01044182
-0.13490754 0.5212957 0.2082595 0.8140138 0.12824409 -1.2064431
0.4658852 0.45741826 0.45537645 -0.8559949 -0.37440166 -0.15812972
-0.47874868 0.6628835 0.6424104 0.02779873 -1.2857279 1.6176172
0.03459741 -0.7289496 0.24206662 0.9685243 0.39266384 0.444358
0.15222543 0.5040782 0.9483841 -0.4365348 -0.19702622 -0.16678298
0.08670887 -0.67878836 0.8700074 0.7877692 -1.0610175 -0.5403886
-1.0454092 0.10999838 -0.47457123 0.52491254 1.0068256 0.37586346
-0.89350504 1.0584757 -1.1252506 -0.2969869 0.5806031 0.660968
-0.35711834 -0.31933972 -0.45795086 1.2264508 0.83542687 0.2475767
-1.1743722 -0.5790995 -0.55262697 0.08821225 0.14608727 -0.4539316
1.1536516 -0.33124882 -1.5123098 0.61868113 0.49863654 -0.37102392
0.60701746 -0.45103967 0.5396618 0.15380532 -0.00905619 1.2863433
1.0065559 -1.8974177 -0.1991897 -0.487023 0.17986582 -0.03487112
0.06718569 -0.60455626 0.18721944 -0.5318636 0.5542668 -1.3146716
0.11075001 0.09476341 -0.30864975 -0.5492115 1.0730112 -0.6239858
-0.03607065 -1.8218188 0.69296545 0.2041005 -0.12365349 0.15157166
-0.1431008 0.56660116 0.07106832 -0.28236407 -0.23880985 0.14388712
0.39604792 0.51402766 -0.783368 0.16659667 0.46909648 1.1282353
-1.0610836 0.04046296 0.259859 0.5442303 -0.5282854 0.2625413
-0.41678116 0.28197786 -0.41025463 0.73859245 0.48927316 0.14456254
0.16308878 -0.33990434 0.3337144 -0.3148928 -0.68251127 1.8109871
-0.39818704 0.20422612 0.20143522 -0.99480146 1.4628593 0.18604477
0.50905097 -0.04959343 0.29203498 0.46166733 0.28305018 -0.3389871
0.32662243 0.14455058 -0.16761842 0.27776238 0.39010644 -0.93041736
0.04117541 -0.15356547 0.9593801 0.9581187 -0.48622957 -0.4528019
-0.34301072 0.12247352 -0.2512105 0.8797455 0.29110235 0.83214897
0.3568211 -0.31402394 -1.4833537 -1.6047182 -0.314773 0.96094334
0.31058633 0.2965271 -0.80580455 -0.31864995 0.54891723 0.913494
-0.7919546 -0.40810752 -0.5977487 -0.78967124 0.38151062 0.9020893
0.11372158 -1.6832258 -1.2986214 0.35174772 0.01491422 -0.9978109
0.48795027 0.4693136 -0.9070577 -1.1947063 -1.0927647 -0.95916563
0.9107383 0.08673969 0.8175555 -0.20813793 -0.71624845 0.5855076
-0.5901516 -0.6243276 -0.7287936 0.13788009 0.07366448 -0.7965229
0.1377996 -0.46900526 -0.38560393 0.4214003 -0.86295164 -0.17652044
1.0415661 0.9287512 0.23588596 -0.27195737 0.7495986 0.07778684
0.667918 -0.31661075 -0.39877212 0.37255234 -0.23840639 0.16644883
0.32558855 -0.7869936 -0.6762864 0.10566948 0.3526208 -0.78152794
-0.11769817 -0.04029332 0.14240895 -0.12601848 0.7583133 -0.03302584
0.3296662 -0.29406 0.40868637 0.5574202 0.8079837 -1.1410269
0.4900002 0.21620367 0.2826587 -0.50191885 -0.22233832 0.08437448
-0.845599 -0.07109202 0.76902336 -0.19454718 -1.2709413 0.61530113
-1.3045474 -0.57781065 -0.24232462 0.45595995 -1.2036184 0.09737201
-0.50482535 -0.86596715 -0.5335948 -1.3612344 1.6073222 0.0102194
-0.1224979 -0.26958206 -0.61095905 -0.09644347 0.4621554 0.70491326
1.0332434 -0.43725517 -0.6003845 -0.4989215 -0.30651313 0.46083525
0.300126 0.09257317 -0.7126658 -0.62771887 -0.22866254 -0.7145522
0.5179623 0.21159707 1.2062907 -0.03499021 -0.57770944 -0.08450942
1.162996 0.402445 0.7105308 -0.03103882 0.4465167 0.75525063
0.6435508 0.10452691 -0.2361247 0.58503634 1.1205812 0.5218784
-0.08974556 -0.17743275 -0.01171712 -0.04291813 -0.4534982 -0.2889676
-1.0742856 0.48022628 -1.7453514 -0.69227517 0.18445168 2.2043502
0.51855123 0.3364458 -0.12400411 0.17315035 0.6088523 -0.642727
-0.73750263 -0.17005102 0.17396279 0.43465668 0.52353126 0.1064172
-1.0356543 0.8463677 6.152952 0.35958794 -0.9606249 -0.40084702
0.00698023 0.18228465 0.27348557 -0.35737768 -0.5458881 0.43490973
0.39663917 0.21537448 0.709143 1.3073574 -0.6470211 -0.2001634
-1.5541295 0.71302474 0.02520309 -0.89017755 0.29100278 -0.09567245
0.37273526 -0.07333004 0.18878618 0.45519745 0.46519613 -1.2859626
0.92581064 0.68294203 0.63625467 -0.4980139 0.66746074 0.3554894
-0.44522873 -0.2723272 -0.7093382 0.33077225 -0.06992649 -0.16755132
-1.3561037 0.5961869 1.0369036 0.08742602 -0.11567718 0.87934107
-0.16384465 0.02591701 -0.49961916 -0.3534841 0.3018604 0.05238755
0.55579793 0.8171279 0.53032756 -0.02407664 -0.0936431 1.2895074
0.25630233 -0.6229317 -0.72625756 -0.091589 0.34941575 0.98769236
-0.82343596 0.01329204 0.5840742 0.7801774 0.17258586 0.20733663
-0.71015304 -0.6233614 0.16039869 -0.18608028 0.5601324 -0.4598396
-0.25516123 -0.3679837 0.2693709 -0.46114033 -0.45311436 -1.4134996
-1.3119959 0.6045134 0.55411696 -0.85572267 -0.51609826 -1.4501144
-0.2036642 0.8597831 -0.59760875 -1.3672011 -0.6882042 -0.03786621
0.35310623 -0.0455121 1.1383399 -0.29845488 0.26307112 0.21363096
-0.09457093 -0.03981033 0.4225604 -1.2817115 0.35018033 0.05045209
-0.28124374 0.7230061 0.9355445 -0.6858385 -1.7093871 -0.88605416
-0.17302716 -0.8482103 0.36693728 -0.46387836 -0.9450665 0.87873244
-0.20162295 -0.21858715 -0.3316576 -0.22315994 -0.18038666 0.12891695
-1.5160778 0.38621444 1.202439 0.09906235 -0.9491027 0.64751667
0.45052823 -0.83184355 -1.1885691 0.98466665 0.93781114 -0.49601218
1.2413073 -0.7220356 0.7688918 0.04035385 -0.36767238 -1.2272457
-0.17270957 -0.25589132 -0.08226769 0.7051418 0.0089861 -0.5425601
0.785681 0.59304476 -0.3223637 -0.87986755 -0.97284424 -1.1219893
0.49519253 0.1927942 0.5162717 0.3297801 -0.01006515 -0.20924424
0.21281482 0.16080645 0.84858567 0.27236727 0.8263254 -1.3390774
-0.01195312 -0.5679072 -0.5777076 -0.76661325 0.14610124 -0.95894057
0.61141694 -1.6912146 -0.05293604 -0.9912376 0.16169325 0.7701389
0.287327 0.18667606 0.15402259 0.17103659 0.09006791 0.590459
1.4190183 -0.18724413 0.07347946 -0.01571779 -0.04485054 0.44320762
0.94659835 -0.3136011 -0.30545482 -0.5569182 -0.21280444 0.02368266
0.95808285 -1.1180018 -0.16477722 0.09094483 0.6698709 -0.4791102
0.7025811 -0.9855746 0.3975471 0.7369892 -0.23654889 -0.10788111
0.30700305 0.55834264 0.48032007 -0.51829726 0.5087791 -0.10606974
-0.3025721 -0.11403174 0.12007134 -0.5626056 1.3478853 -0.23577936
-0.48041278 0.00849741 -0.9292471 -0.0791382 0.5084668 0.9706205
0.525786 -1.1010952 -0.6287116 0.1133702 -0.09314395 0.29014173
-0.15769829 0.21828091 -0.6696287 0.52206606 -0.75095403 -1.0490667
-0.74466026 0.67540896 0.13038619 0.3998479 -0.68451077 0.75450855
-0.04570075 -0.6045451 0.38125837 -0.65552753 0.47373873 -0.50482184
-0.19458131 0.39236388 0.1324564 -0.18906377 -0.10081016 0.50673425
0.20296004 0.07730684 1.5042876 0.7766322 -0.03647699 0.170911
0.7985772 -0.68242204 -1.7511591 0.33661285 -0.18179731 -0.29573917
-0.5691813 -1.1739848 -0.73728806 0.7590946 0.43640128 0.30922666
0.37127507 0.24114949 0.37376615 0.9374655 1.1746768 -0.72682524
0.3454851 0.29273105 1.8503155 -1.0537143 -0.13468027 -0.49279973
-0.46987605 1.4056158 0.7334059 -0.72919667 0.24195866 0.44436675
-0.29981714 -0.14058657 -0.12348826 0.47015172 0.09412156 0.93167543
-0.13468868 0.21743955 -0.01549167 0.4256593 -0.5051114 0.02046281
0.24897657 1.2420287 -0.43745375 -0.75815725 -0.23519444 0.47242978
-0.06888884 0.4654966 -0.2267138 1.1396672 -0.21486908 0.6195053
0.13258958 -0.26167417 0.38612294 0.05101798 1.3316636 -0.5072544
-0.3287552 -0.5019559 -0.51432055 -0.431285 -0.20137475 -0.6008168
-1.346537 0.11326844 -0.32112065 -0.04261942 1.1718849 0.5371277
0.33063188 0.5203722 0.11897092 -1.9260557 -1.2604914 -1.3259902
-0.19971865 0.3881744 -0.11866711 -1.2308804 -0.2010764 -0.03861377] | [5.738926410675049, -0.8216999173164368] |
d6c7dbaf-6982-4559-b56f-e19623e1a36c | automated-segmentation-of-microvessels-in | 2210.00166 | null | https://arxiv.org/abs/2210.00166v2 | https://arxiv.org/pdf/2210.00166v2.pdf | Automated segmentation of microvessels in intravascular OCT images using deep learning | To analyze this characteristic of vulnerability, we developed an automated deep learning method for detecting microvessels in intravascular optical coherence tomography (IVOCT) images. A total of 8,403 IVOCT image frames from 85 lesions and 37 normal segments were analyzed. Manual annotation was done using a dedicated software (OCTOPUS) previously developed by our group. Data augmentation in the polar (r,{\theta}) domain was applied to raw IVOCT images to ensure that microvessels appear at all possible angles. Pre-processing methods included guidewire/shadow detection, lumen segmentation, pixel shifting, and noise reduction. DeepLab v3+ was used to segment microvessel candidates. A bounding box on each candidate was classified as either microvessel or non-microvessel using a shallow convolutional neural network. For better classification, we used data augmentation (i.e., angle rotation) on bounding boxes with a microvessel during network training. Data augmentation and pre-processing steps improved microvessel segmentation performance significantly, yielding a method with Dice of 0.71+/-0.10 and pixel-wise sensitivity/specificity of 87.7+/-6.6%/99.8+/-0.1%. The network for classifying microvessels from candidates performed exceptionally well, with sensitivity of 99.5+/-0.3%, specificity of 98.8+/-1.0%, and accuracy of 99.1+/-0.5%. The classification step eliminated the majority of residual false positives, and the Dice coefficient increased from 0.71 to 0.73. In addition, our method produced 698 image frames with microvessels present, compared to 730 from manual analysis, representing a 4.4% difference. When compared to the manual method, the automated method improved microvessel continuity, implying improved segmentation performance. The method will be useful for research purposes as well as potential future treatment planning. | ['David L. Wilson', 'Hiram G. Bezerra', 'Giulio Guagliumi', 'Sadeer Al-Kindi', 'Ammar Hoori', 'Luis A. P. Dallan', 'Vladislav N. Zimin', 'Ga-briel T. R. Pereira', 'Issam Motairek', 'Yazan Gharaibeh', 'Lia Gomez-Perez', 'Justin N. Kim', 'Juhwan Lee'] | 2022-10-01 | null | null | null | null | ['shadow-detection'] | ['computer-vision'] | [ 9.18796882e-02 1.03727661e-01 -9.41211060e-02 -3.18046182e-01
-7.22904086e-01 -5.34378707e-01 -2.23569125e-02 3.48047078e-01
-5.83240509e-01 7.28404403e-01 -1.45143857e-02 -6.46535456e-01
4.03597683e-01 -6.87047064e-01 -2.65002459e-01 -6.80801272e-01
-3.67153168e-01 2.69376665e-01 3.44013005e-01 4.19790655e-01
2.47275770e-01 8.52319539e-01 -7.31613100e-01 4.86036330e-01
7.42902577e-01 1.24718606e+00 -3.27923745e-02 9.53123629e-01
-7.09496960e-02 5.97059906e-01 -5.45317292e-01 -1.39193594e-01
4.91565675e-01 -1.98200032e-01 -5.48813760e-01 4.64680195e-02
6.81036115e-01 -5.75395346e-01 -1.89034566e-01 7.48278320e-01
7.85744131e-01 -2.91562080e-01 6.99324012e-01 -4.62308884e-01
-2.06726119e-01 1.42377123e-01 -9.00961459e-01 8.00614536e-01
-3.90467525e-01 6.01398349e-01 6.35014832e-01 -5.58914542e-01
9.09622788e-01 5.55821419e-01 6.55215979e-01 3.23034525e-01
-1.58583832e+00 -5.23859918e-01 -3.92667800e-01 -1.94842275e-02
-1.29758787e+00 -3.30208600e-01 4.05344307e-01 -9.91835833e-01
9.23660934e-01 2.24043518e-01 1.18637943e+00 2.88619071e-01
5.25261104e-01 2.82155633e-01 1.08083630e+00 -2.25187376e-01
1.04465105e-01 3.08458984e-01 3.29679102e-01 6.12593591e-01
4.79381979e-01 -1.05611242e-01 4.59137946e-01 -1.35868400e-01
1.31191671e+00 -2.68772095e-01 -2.73855120e-01 -2.11686328e-01
-9.30048108e-01 9.00062501e-01 5.08890688e-01 2.66861051e-01
-4.12157267e-01 -4.73758727e-02 5.50659180e-01 -5.38805239e-02
3.39228243e-01 5.26507497e-01 -1.49790898e-01 5.02088815e-02
-8.23669136e-01 -2.08123043e-01 4.16170597e-01 2.49882385e-01
4.21512008e-01 -7.73109612e-04 -1.47027433e-01 9.09571052e-01
7.63513427e-03 1.23819835e-01 4.88648534e-01 -1.25136614e+00
-3.06915194e-02 6.68132246e-01 4.20778133e-02 -8.03323746e-01
-8.87655556e-01 -6.28974497e-01 -1.00318205e+00 6.44772947e-01
8.83975863e-01 -2.88461477e-01 -9.04993951e-01 1.17737925e+00
1.18541390e-01 -2.69944847e-01 -2.16758028e-01 1.08463919e+00
7.71299005e-01 2.76283324e-01 9.84692127e-02 -3.58262450e-01
1.43559980e+00 -4.90122139e-01 -4.18146819e-01 6.95049167e-02
1.01521945e+00 -9.09376383e-01 1.08074760e+00 2.22782716e-01
-1.14568210e+00 -4.29487646e-01 -9.42727625e-01 1.09244116e-01
1.46679521e-01 3.35155487e-01 5.78060269e-01 6.88290417e-01
-1.23759437e+00 3.64995271e-01 -9.12121892e-01 -1.65588826e-01
1.02507246e+00 5.32690585e-01 -3.10333282e-01 3.07915747e-01
-6.36806905e-01 7.76692808e-01 -1.48668751e-01 -1.18332118e-01
-1.96135551e-01 -1.05835950e+00 -5.25261760e-01 -1.26528502e-01
-2.39078149e-01 -8.20464849e-01 1.10988915e+00 -7.99784124e-01
-1.20117819e+00 1.11829650e+00 -3.81106883e-01 -7.60256410e-01
5.70340633e-01 1.79086864e-01 -1.88828006e-01 8.77335250e-01
2.53819585e-01 5.81867635e-01 3.84654790e-01 -9.13497984e-01
-6.78851306e-01 -4.34773535e-01 -3.37747216e-01 -2.17578664e-01
2.30917052e-04 -5.56083806e-02 -1.45465747e-01 -3.73487800e-01
3.16623122e-01 -7.30421662e-01 -5.06019175e-01 5.28032780e-01
-2.43144408e-01 1.27221048e-01 3.93351018e-01 -7.96707034e-01
1.08513820e+00 -1.84413874e+00 -8.30269516e-01 5.74230552e-01
1.06057000e+00 4.73495990e-01 6.57297075e-02 -4.29881006e-01
-4.44921196e-01 4.81711477e-01 -1.53805405e-01 4.42366488e-02
-7.39091992e-01 -4.28723633e-01 2.96518803e-01 7.41908967e-01
4.38304432e-02 8.50416005e-01 -5.76554477e-01 -6.41530097e-01
6.93123639e-01 6.19809747e-01 -4.59341913e-01 -1.92482010e-01
4.23433840e-01 5.26863456e-01 -1.70990437e-01 7.18385220e-01
7.92321444e-01 -5.70934534e-01 9.64985415e-02 -4.07546759e-01
-3.43366623e-01 -3.83631214e-02 -9.11419392e-01 1.07668781e+00
-4.25253808e-01 1.10972464e+00 -2.96372618e-03 -7.00142622e-01
9.14258659e-01 3.20575237e-01 1.07228339e+00 -7.01531947e-01
4.57872421e-01 2.68613517e-01 6.32105529e-01 -7.90085256e-01
-1.19628161e-01 -2.69536853e-01 7.73812890e-01 2.99230069e-01
-4.28952157e-01 8.11624900e-02 2.70996064e-01 9.56510082e-02
9.69775915e-01 -5.29593349e-01 2.77715445e-01 -2.98283160e-01
3.45927179e-01 3.88483882e-01 4.67750579e-01 6.76749110e-01
-6.45919919e-01 8.36164594e-01 8.36530864e-01 -9.21464443e-01
-1.08868229e+00 -9.51597273e-01 -8.93408954e-01 -1.52276322e-01
-2.09165234e-02 -1.88127220e-01 -7.09966600e-01 -2.59062886e-01
-1.99749634e-01 2.55003482e-01 -3.41088206e-01 1.54013455e-01
-6.00559413e-01 -1.01187932e+00 2.32186154e-01 4.88560766e-01
3.96217436e-01 -7.57359087e-01 -8.39228332e-01 5.91431558e-01
-2.72994041e-01 -1.17027295e+00 -1.07647777e-01 -1.83029845e-01
-1.10122430e+00 -1.32649124e+00 -7.28810012e-01 -6.68930471e-01
7.02732563e-01 2.85293669e-01 8.78992975e-01 2.34980136e-02
-1.01503623e+00 -2.13133574e-01 -6.85857758e-02 -1.19939961e-01
-2.40620211e-01 -3.62743169e-01 -2.55600303e-01 -4.50894907e-02
4.48866010e-01 -4.25802827e-01 -1.23269665e+00 1.73233494e-01
-3.14003915e-01 -3.18700932e-02 5.64633548e-01 8.29113245e-01
9.71858203e-01 -2.63761580e-01 4.29150254e-01 -6.73711121e-01
2.92127758e-01 6.99450076e-02 -6.92174613e-01 -2.79621869e-01
-4.87770796e-01 -6.03919148e-01 5.00837803e-01 -3.16636771e-01
-6.29403234e-01 1.43934384e-01 2.70046797e-02 -3.24342340e-01
-2.04285488e-01 2.36752078e-01 5.72202981e-01 -3.54939699e-01
1.03191447e+00 -1.86187133e-01 5.92799067e-01 1.79849826e-02
-1.36926904e-01 7.58362889e-01 4.59595114e-01 -2.57196873e-02
1.16326854e-01 9.67026651e-01 1.63170591e-01 -9.67026651e-01
-2.85560548e-01 -5.31390011e-01 -6.61451995e-01 -2.91155964e-01
1.13403308e+00 -5.60604990e-01 -7.64658511e-01 1.39888123e-01
-1.04683375e+00 -3.97853076e-01 -3.62813652e-01 9.03069615e-01
-2.86194324e-01 5.33986092e-01 -7.72729933e-01 -4.34220135e-01
-7.53997564e-01 -1.57150722e+00 3.36069673e-01 2.86558598e-01
-7.47646630e-01 -1.02209365e+00 -2.48113676e-04 4.05847758e-01
4.80112404e-01 5.79949915e-01 9.29953218e-01 -4.70195174e-01
-4.31162238e-01 -5.28021932e-01 -5.89822590e-01 4.67500985e-01
2.84902006e-01 4.06788617e-01 -7.55885661e-01 1.60978004e-01
-1.44568309e-01 2.45307252e-01 8.07994843e-01 1.30966270e+00
1.35337365e+00 6.30317405e-02 -3.81909579e-01 7.61605263e-01
1.27238297e+00 6.59910142e-01 7.19663084e-01 4.37798679e-01
4.29052204e-01 5.50620973e-01 1.35163933e-01 3.43773007e-01
-1.10044070e-01 3.90955836e-01 3.69208395e-01 -7.73939908e-01
-3.42611909e-01 6.39076114e-01 -4.13369268e-01 -1.12315498e-01
-2.54425704e-01 2.40428016e-01 -1.17208624e+00 5.39594352e-01
-1.26460016e+00 -7.75356948e-01 -8.40566516e-01 2.23565769e+00
6.04671836e-01 3.70984137e-01 2.00421676e-01 -3.11626554e-01
8.44655097e-01 -3.20063889e-01 -6.04878783e-01 -3.11177760e-01
-1.28200889e-01 6.31524682e-01 6.22539878e-01 6.26740158e-01
-1.20901668e+00 7.01519430e-01 5.72460079e+00 2.72936076e-01
-1.50748789e+00 -3.47175360e-01 1.15987313e+00 -2.74857998e-01
1.94381297e-01 -3.97636853e-02 -3.96186531e-01 3.87793094e-01
5.52791178e-01 4.96589355e-02 6.06326340e-03 6.14751220e-01
6.60646379e-01 -4.93801177e-01 -5.25890768e-01 1.06642354e+00
-2.78147608e-01 -1.84644568e+00 -2.81234831e-01 3.45978409e-01
5.14903665e-01 1.66675851e-01 -2.67887302e-02 -2.25411922e-01
-2.02302054e-01 -1.04372334e+00 -7.75320549e-03 4.11041915e-01
1.10824573e+00 -3.98305029e-01 1.05654323e+00 -3.10003668e-01
-9.54084456e-01 2.53838241e-01 -2.23717809e-01 8.72922987e-02
3.92603278e-01 1.10717165e+00 -1.20174038e+00 -8.25177357e-02
7.42770016e-01 6.76266372e-01 -3.98936033e-01 1.34536719e+00
2.04982296e-01 4.70830441e-01 -4.46958542e-01 1.94368705e-01
-2.53395010e-02 -3.72978598e-01 6.68458283e-01 1.05295944e+00
2.49255006e-03 2.59496003e-01 -2.13750795e-01 8.18711996e-01
2.32284620e-01 3.99540454e-01 -2.03668520e-01 3.02595228e-01
4.31254119e-01 1.60836864e+00 -1.05358565e+00 -6.21850133e-01
-5.16731679e-01 3.48597288e-01 -9.99003425e-02 4.31224823e-01
-6.59324467e-01 -6.73492789e-01 6.21303976e-01 8.32298815e-01
-1.44643649e-01 5.10373525e-02 -1.06591296e+00 -8.29453111e-01
1.24296904e-01 -3.52026761e-01 3.88084173e-01 -5.08992851e-01
-1.16018152e+00 7.15398908e-01 -4.85652030e-01 -1.35280788e+00
3.38756479e-03 -7.11746693e-01 -8.76369715e-01 1.05472660e+00
-1.48594475e+00 -5.85514128e-01 -6.08778596e-01 2.17144176e-01
5.86873293e-02 -1.41669601e-01 6.51377618e-01 1.88631877e-01
-5.36205709e-01 3.89278710e-01 -3.34766001e-01 2.54556775e-01
9.22493160e-01 -1.22953367e+00 -1.80708673e-02 6.34445250e-01
-6.78074837e-01 8.82874608e-01 1.46537796e-01 -4.91367698e-01
-6.02385759e-01 -1.08138108e+00 8.06157827e-01 -1.56897306e-01
7.48637199e-01 3.11104476e-01 -8.28478754e-01 5.85128307e-01
-6.30382895e-02 3.19993049e-01 1.05271959e+00 -3.35859179e-01
2.24234127e-02 -7.87807480e-02 -1.29643083e+00 7.70493448e-01
2.55172342e-01 1.32191703e-02 -6.53189942e-02 4.19951767e-01
1.14038393e-01 -6.59618378e-01 -1.15391994e+00 5.91509521e-01
8.31322849e-01 -1.14058280e+00 1.03834403e+00 -2.73864031e-01
3.55287701e-01 -1.54837102e-01 1.77313551e-01 -6.48069322e-01
-4.97581542e-01 -5.06012559e-01 3.89741063e-01 3.86654437e-01
4.80768025e-01 -9.39129829e-01 1.04605734e+00 6.81927443e-01
-5.76576233e-01 -9.31025088e-01 -1.00653303e+00 -9.63867903e-02
3.28744680e-01 -3.75395924e-01 -7.85663575e-02 6.76343977e-01
7.93824643e-02 -7.40643917e-03 3.08300495e-01 -7.88940042e-02
5.21660626e-01 3.15107964e-02 6.31793618e-01 -1.24185288e+00
1.08243097e-02 -9.03362989e-01 -5.77424407e-01 -7.75383294e-01
-2.02030122e-01 -1.02758610e+00 -7.18813300e-01 -1.85908401e+00
7.53747672e-02 -4.77578789e-01 -9.67843756e-02 5.22990882e-01
4.01366763e-02 5.02045035e-01 -2.99492210e-01 3.63147885e-01
3.89153361e-01 -2.72393167e-01 1.63122594e+00 -3.87193374e-02
-7.50480354e-01 1.01883531e-01 -8.09131682e-01 7.60472178e-01
1.57577586e+00 -1.27896875e-01 -1.29346311e-01 1.81846088e-03
-3.14694792e-01 1.16207972e-01 6.60098910e-01 -9.24072564e-01
-1.30963534e-01 2.05856666e-01 8.50921035e-01 -4.99575317e-01
2.23899558e-01 -5.24559319e-01 -1.27479240e-01 8.29810143e-01
-4.73983027e-02 -3.28577876e-01 3.29740137e-01 -1.73428223e-01
1.74278184e-03 -9.89868306e-03 1.33542717e+00 -3.47651124e-01
-4.01197165e-01 1.77035183e-01 -9.21772897e-01 3.32390033e-02
1.18638325e+00 -7.25180089e-01 -5.33989608e-01 -1.51837558e-01
-1.26438761e+00 3.15443799e-02 3.57673347e-01 -4.22337472e-01
6.20066404e-01 -7.17188954e-01 -7.98216224e-01 3.12097430e-01
1.50466964e-01 -9.15454477e-02 6.69259667e-01 1.53689766e+00
-1.14514601e+00 2.92581081e-01 -3.05136442e-01 -1.06229568e+00
-1.32672405e+00 -1.03282690e-01 9.59479988e-01 -9.14153829e-02
-1.01437998e+00 8.15191984e-01 1.91781431e-01 2.26653993e-01
-1.36524796e-01 -4.53540236e-01 -1.91661790e-01 -1.77183449e-01
6.36366248e-01 3.47436517e-01 1.53148666e-01 -4.24480706e-01
-2.35826418e-01 8.63488615e-01 -3.08788687e-01 3.81881706e-02
9.06736135e-01 7.36617856e-03 -2.43340015e-01 -2.71295421e-02
1.30997598e+00 -6.41060248e-02 -1.10067952e+00 6.05880618e-02
-6.34181142e-01 -4.06921655e-01 3.36925775e-01 -8.66598725e-01
-1.25168610e+00 9.88014102e-01 1.10797763e+00 1.03777722e-01
9.54908192e-01 -8.97036269e-02 7.84038007e-01 -2.80808628e-01
-1.15440473e-01 -7.66480565e-01 -1.49950489e-01 1.08825915e-01
5.26664793e-01 -1.30698633e+00 5.38286427e-03 -6.68265820e-01
-6.55966878e-01 1.47111273e+00 6.74815595e-01 -1.80161327e-01
7.17556000e-01 3.55860978e-01 4.59555268e-01 -2.92860270e-01
-1.77476585e-01 6.00380562e-02 1.52910218e-01 6.60293639e-01
7.83419728e-01 1.59716815e-01 -4.95495260e-01 3.83424461e-01
-2.24058792e-01 1.01455957e-01 7.95039296e-01 7.42198050e-01
-6.41541421e-01 -6.63348675e-01 -1.50850117e-01 9.49962318e-01
-5.51874816e-01 -2.25610524e-01 7.43956491e-02 9.88893270e-01
-1.17165816e-03 9.71333683e-01 4.66192871e-01 -1.32870153e-01
3.19961131e-01 -2.75838763e-01 2.17835814e-01 -5.27190149e-01
-5.60711801e-01 7.54146874e-01 -4.13399711e-02 -6.42253041e-01
-1.26947969e-01 -5.50809979e-01 -1.68023503e+00 -1.52977780e-01
5.22090830e-02 -1.13175109e-01 6.29580379e-01 3.72502059e-01
3.20876271e-01 6.92417085e-01 4.16901022e-01 -4.64745164e-01
-1.40797608e-02 -8.53249013e-01 -6.02691650e-01 3.10365260e-02
5.26228011e-01 -4.10474032e-01 -6.30191505e-01 3.73273015e-01] | [14.269331932067871, -2.444283962249756] |
e00c0d0c-2ca2-406b-ac4c-82f7367e736c | datnet-dual-adversarial-transfer-for-low | null | null | https://openreview.net/forum?id=HkGzUjR5tQ | https://openreview.net/pdf?id=HkGzUjR5tQ | DATNet: Dual Adversarial Transfer for Low-resource Named Entity Recognition | We propose a new architecture termed Dual Adversarial Transfer Network (DATNet) for addressing low-resource Named Entity Recognition (NER). Specifically, two variants of DATNet, i.e., DATNet-F and DATNet-P, are proposed to explore effective feature fusion between high and low resource. To address the noisy and imbalanced training data, we propose a novel Generalized Resource-Adversarial Discriminator (GRAD). Additionally, adversarial training is adopted to boost model generalization. We examine the effects of different components in DATNet across domains and languages and show that significant improvement can be obtained especially for low-resource data. Without augmenting any additional hand-crafted features, we achieve new state-of-the-art performances on CoNLL and Twitter NER---88.16% F1 for Spanish, 53.43% F1 for WNUT-2016, and 42.83% F1 for WNUT-2017. | ['Kenneth Kwok', 'Hao Zhang', 'Joey Tianyi Zhou', 'Hongyuan Zhu', 'Rick Siow Mong Goh', 'Di Jin'] | 2019-05-01 | null | null | null | iclr-2019-5 | ['low-resource-named-entity-recognition'] | ['natural-language-processing'] | [-3.00415128e-01 -2.11242139e-01 -2.62474045e-02 -4.53462929e-01
-1.16802800e+00 -7.95897424e-01 6.80700719e-01 -5.15360758e-02
-9.97678757e-01 1.07535410e+00 1.19570114e-01 -2.02781230e-01
2.75191456e-01 -8.30745339e-01 -6.39268100e-01 -3.46371949e-01
1.88681185e-01 2.68992931e-01 -1.38631195e-01 -4.95720059e-01
-4.68706012e-01 5.54927289e-01 -9.72788751e-01 2.85081387e-01
1.13366532e+00 8.93023014e-01 -2.29511514e-01 5.59474111e-01
-2.78171629e-01 5.97217202e-01 -9.88528013e-01 -8.45257282e-01
2.61139125e-01 -1.50321841e-01 -7.71242499e-01 -6.12522781e-01
1.94549933e-01 -1.28076673e-01 -5.42623281e-01 8.78118992e-01
1.04645193e+00 2.00756222e-01 6.93912387e-01 -1.09012318e+00
-9.23070908e-01 6.74649835e-01 -3.99619520e-01 2.69934356e-01
1.89845845e-01 2.04199687e-01 7.80078411e-01 -1.03750384e+00
5.03058970e-01 1.14190304e+00 1.06051564e+00 8.72195780e-01
-8.79621923e-01 -1.08675945e+00 1.97536405e-02 -3.84942666e-02
-1.53812432e+00 -4.87341255e-01 3.78912985e-01 -1.17193041e-02
8.82792830e-01 2.67441452e-01 -1.07608363e-01 1.35586095e+00
-4.58028764e-02 8.54406774e-01 1.19347978e+00 -1.13757044e-01
-5.65949082e-02 7.23691434e-02 -3.00163031e-02 4.09497797e-01
3.89495730e-01 -2.61754990e-02 -2.59980768e-01 -1.40339300e-01
3.75226647e-01 4.50366437e-02 -1.23912446e-01 6.49121344e-01
-8.99119020e-01 6.57324910e-01 5.98589361e-01 4.42231238e-01
-4.44212407e-01 -1.90601528e-01 6.08834147e-01 2.36446440e-01
7.92087972e-01 3.80861551e-01 -9.23128843e-01 -1.39306784e-01
-7.27866232e-01 -4.10816493e-03 8.09970856e-01 1.09780896e+00
5.37961125e-01 5.03206909e-01 -5.06729126e-01 1.11952591e+00
4.80898172e-02 8.81943047e-01 6.20727479e-01 -2.69657433e-01
8.24140906e-01 4.58223909e-01 4.90804985e-02 -4.74166512e-01
-5.02447546e-01 -5.06038308e-01 -1.29363799e+00 -4.28139001e-01
1.97455496e-01 -8.04539263e-01 -1.25977361e+00 1.82573032e+00
2.09290981e-01 3.85021061e-01 5.20578325e-01 5.80951452e-01
1.31706643e+00 6.42378628e-01 7.38712132e-01 2.66674399e-01
1.41947019e+00 -9.87572312e-01 -6.59872472e-01 -2.55100608e-01
5.76870203e-01 -6.01363599e-01 1.00634539e+00 -1.06579274e-01
-8.92368853e-01 -6.41935885e-01 -8.08301270e-01 -1.15212493e-01
-8.61317098e-01 4.37660515e-01 4.39176232e-01 8.57299805e-01
-7.26143718e-01 6.04534686e-01 -7.24382222e-01 -1.71407983e-01
4.79141504e-01 4.38961118e-01 -5.89809477e-01 -2.74566174e-01
-1.64824462e+00 7.65432417e-01 3.66709441e-01 9.65298563e-02
-8.98845851e-01 -9.10322070e-01 -8.32079709e-01 1.16659336e-01
-4.39850651e-02 -5.30306637e-01 1.13791668e+00 -7.42922723e-01
-1.38773358e+00 8.10663998e-01 1.76725239e-01 -3.97464424e-01
6.23799264e-01 -4.87672836e-01 -1.12990606e+00 -2.25374624e-01
1.19849943e-01 5.84768951e-01 4.06954408e-01 -9.82083082e-01
-4.43288893e-01 -1.78611130e-01 9.22327675e-03 6.08028285e-02
-6.78018034e-01 2.33697414e-01 -3.90279382e-01 -9.10233021e-01
-6.24744177e-01 -9.03121173e-01 -2.42210045e-01 -4.56069708e-01
-5.98109543e-01 -2.93842018e-01 7.26992309e-01 -1.00240970e+00
1.20497394e+00 -2.18765903e+00 -3.91217142e-01 -1.43931285e-01
2.66514551e-02 9.95203137e-01 -5.09796858e-01 1.93704695e-01
-1.32396430e-01 7.00782716e-01 -1.51258379e-01 -4.10069346e-01
4.30349372e-02 2.08372518e-01 -8.43015537e-02 2.94978827e-01
6.94468677e-01 1.06965399e+00 -8.62442851e-01 -1.14901662e-01
-3.91970873e-02 8.15831065e-01 -4.90416825e-01 4.64393258e-01
8.70889351e-02 5.22446632e-01 -7.04246104e-01 6.90656483e-01
9.89361167e-01 -6.76704273e-02 8.14691037e-02 -6.69335276e-02
3.74045945e-03 2.03861296e-01 -1.12997782e+00 1.42371535e+00
-7.42762744e-01 2.26061493e-01 -7.67071848e-04 -5.28094709e-01
1.11940527e+00 4.66322690e-01 1.93844408e-01 -9.27299380e-01
1.99909031e-01 2.75564253e-01 -1.44682497e-01 -1.51001483e-01
6.65207624e-01 -1.05881393e-01 -5.85742533e-01 2.01405361e-02
4.39132690e-01 3.21971595e-01 4.23241593e-02 -1.48475975e-01
1.19917381e+00 -9.30692181e-02 2.65195966e-01 -1.85404912e-01
6.03993118e-01 -2.60180980e-01 8.85805011e-01 6.54858768e-01
-3.89435470e-01 5.28270245e-01 1.71295598e-01 -3.26344371e-01
-8.96513939e-01 -8.54285717e-01 -2.51562715e-01 1.42547596e+00
-1.21543504e-01 -1.42572939e-01 -8.16418886e-01 -1.08212852e+00
-3.12043875e-02 5.54338157e-01 -5.63461125e-01 -1.01880841e-01
-5.94353616e-01 -1.15719748e+00 1.47048533e+00 7.53834784e-01
1.06850839e+00 -1.05602765e+00 1.04316160e-01 4.03430313e-02
-1.93874896e-01 -1.29821408e+00 -6.28782511e-01 2.52077579e-01
-4.83498394e-01 -7.73653984e-01 -1.02681446e+00 -5.77268124e-01
3.48891556e-01 -2.20322385e-01 1.21075511e+00 -1.27673134e-01
3.80904116e-02 1.40043825e-01 -5.88249683e-01 -5.51582098e-01
-2.73044169e-01 6.05605781e-01 1.58989459e-01 -1.08766235e-01
3.05559337e-01 -2.08715037e-01 -4.70107734e-01 4.60713416e-01
-8.14903855e-01 -4.42538679e-01 6.88993216e-01 1.14614034e+00
5.73556185e-01 -1.54152066e-01 8.71397197e-01 -1.08501804e+00
4.80151832e-01 -8.95060122e-01 -1.77531078e-01 4.03896958e-01
-3.39059323e-01 3.29717733e-02 1.21357703e+00 -5.81921041e-01
-1.24814916e+00 -3.23389053e-01 -5.29916227e-01 -3.00341696e-01
-3.36647928e-01 3.35724175e-01 -6.19526207e-01 2.81736832e-02
7.51162171e-01 -2.83586048e-02 -7.50673652e-01 -8.26420844e-01
3.75857443e-01 1.03609955e+00 5.69195926e-01 -5.92920899e-01
7.91695058e-01 1.39624178e-01 -4.45040464e-01 -5.86739838e-01
-9.26496387e-01 -2.99913228e-01 -4.34663951e-01 4.09677416e-01
9.34061289e-01 -1.41624951e+00 -5.33392906e-01 6.97247326e-01
-1.11506486e+00 -3.12551975e-01 -2.39683881e-01 4.33638871e-01
1.09783776e-01 -1.00125618e-01 -8.04882824e-01 -5.63860059e-01
-8.78993332e-01 -9.35380816e-01 9.26933944e-01 5.55492282e-01
2.70920664e-01 -1.09705305e+00 -1.25858253e-02 2.49873325e-01
7.33493090e-01 3.63643348e-01 4.50738966e-01 -1.55031538e+00
7.12082088e-02 -3.45808625e-01 -5.25921345e-01 6.66456819e-01
-1.72217097e-02 -3.16184193e-01 -1.24707675e+00 -4.90644664e-01
-4.29744661e-01 -4.63663787e-01 5.94980359e-01 -1.91840500e-01
1.15899813e+00 -3.63141239e-01 -2.16330260e-01 8.92370880e-01
1.46835554e+00 6.30267709e-02 6.45241082e-01 2.37146094e-01
8.41678143e-01 8.94301385e-02 4.91881520e-01 4.88580734e-01
5.10774970e-01 4.81490403e-01 4.38607745e-02 -4.37399030e-01
-3.16086560e-01 -4.06900436e-01 3.40674043e-01 8.07061970e-01
-1.53725326e-01 -6.40479982e-01 -1.13611031e+00 5.27542293e-01
-1.31140924e+00 -6.53083622e-01 4.63201739e-02 1.79104984e+00
1.02237868e+00 -9.73975360e-02 -4.56482880e-02 -2.23066568e-01
9.08745468e-01 2.14851379e-01 -7.75835991e-01 -3.26954335e-01
-3.31347793e-01 7.68740654e-01 9.30337846e-01 1.29928552e-02
-1.46136022e+00 1.25395632e+00 5.90236664e+00 1.23137093e+00
-1.22106683e+00 4.73628938e-01 8.21915209e-01 3.04662019e-01
-2.47480288e-01 -4.34541970e-01 -1.20058584e+00 5.84373653e-01
1.36297679e+00 -1.50641829e-01 1.34932756e-01 8.84474874e-01
-3.35579425e-01 5.15418470e-01 -5.38918793e-01 7.88486838e-01
-1.30715519e-02 -8.82006407e-01 9.20297857e-03 -1.18358098e-01
8.46585512e-01 5.05278587e-01 -2.05366276e-02 9.97219026e-01
7.55332351e-01 -1.20438182e+00 3.55892032e-01 3.77892107e-01
1.39030278e+00 -1.00567591e+00 1.21713209e+00 1.58913106e-01
-1.32609367e+00 6.16110004e-02 -4.61409003e-01 3.39252383e-01
1.36402836e-02 4.71503854e-01 -5.87291360e-01 1.07228446e+00
7.13480234e-01 5.71454346e-01 -5.61380982e-01 8.22351456e-01
-3.30906928e-01 7.33076096e-01 -4.16137964e-01 2.91459225e-02
1.87056050e-01 3.43028814e-01 2.89436758e-01 1.52394247e+00
3.35980773e-01 2.15800037e-03 2.60516256e-01 4.14663136e-01
-8.79427195e-01 2.55110353e-01 -3.97602230e-01 -3.31837032e-03
7.15550005e-01 1.44289267e+00 -2.84386069e-01 -2.64091671e-01
-4.44754660e-01 1.19054294e+00 5.37142456e-01 3.62127692e-01
-1.15441632e+00 -7.69506395e-01 8.00035954e-01 -2.92866319e-01
4.09206092e-01 -9.25809294e-02 -1.46816954e-01 -1.41778064e+00
-3.08609247e-01 -9.33558583e-01 6.39718056e-01 -4.43216920e-01
-1.67492783e+00 1.08946574e+00 -3.05802971e-01 -1.10570216e+00
-5.05816890e-03 -5.25416970e-01 -5.80055296e-01 1.06434536e+00
-1.91475904e+00 -1.52724218e+00 -3.03751796e-01 7.51797736e-01
2.32202575e-01 -5.08388519e-01 1.12436366e+00 9.45629895e-01
-8.27159584e-01 1.40230739e+00 4.48038101e-01 7.36235201e-01
1.03511417e+00 -1.25012243e+00 6.61594391e-01 8.68638396e-01
2.68661301e-03 2.75258690e-01 1.69613779e-01 -4.99028265e-01
-1.19425476e+00 -1.75520241e+00 9.62986648e-01 -4.44933265e-01
4.54605520e-01 -4.55383539e-01 -7.98937023e-01 7.31060386e-01
1.89542949e-01 4.82584149e-01 1.01426554e+00 9.06159282e-02
-6.85676098e-01 -1.40559286e-01 -1.71533859e+00 4.63704109e-01
9.60298121e-01 -5.33365846e-01 -4.17115808e-01 1.39464751e-01
9.95970130e-01 -6.36861324e-01 -1.41338527e+00 6.09816074e-01
2.08853781e-01 -3.44210178e-01 1.03302109e+00 -1.05846858e+00
2.37014726e-01 2.40141749e-02 -2.83361822e-01 -1.53137577e+00
-2.20828369e-01 -4.71609324e-01 1.19969733e-01 1.82676601e+00
5.56267619e-01 -7.70309806e-01 4.87059772e-01 5.55326641e-01
-2.18674988e-01 -5.01827359e-01 -9.25502777e-01 -9.03711438e-01
4.74746853e-01 -2.47939661e-01 9.51367855e-01 1.42795837e+00
-7.10758269e-01 4.10841972e-01 -4.81624693e-01 2.63761848e-01
2.17495605e-01 -4.73747611e-01 6.87327206e-01 -9.65418935e-01
9.16571170e-02 2.74159219e-02 -8.00604969e-02 -6.98235512e-01
4.59903806e-01 -1.08575094e+00 -1.14934310e-01 -1.18717849e+00
-8.52335468e-02 -7.57120192e-01 -7.81678319e-01 1.03940642e+00
-4.71076459e-01 5.02932549e-01 4.31080699e-01 1.02700375e-01
-6.32795334e-01 7.25319803e-01 1.09474397e+00 -8.86928663e-02
-6.61160648e-02 -1.37498036e-01 -9.06022906e-01 4.65079010e-01
8.90060663e-01 -7.47779787e-01 1.43756807e-01 -6.59829259e-01
-2.31402710e-01 -2.39236727e-01 3.55772264e-02 -1.00701272e+00
9.87628996e-02 4.26578894e-02 6.14147425e-01 -3.47785771e-01
3.55301537e-02 -6.00257397e-01 5.26736416e-02 2.24192470e-01
-1.55826226e-01 3.38336855e-01 6.24921501e-01 4.17770535e-01
-3.49374413e-01 -5.81324436e-02 7.57587433e-01 1.70554921e-01
-7.15559006e-01 6.54009283e-01 9.65027362e-02 6.25296295e-01
8.72773170e-01 3.71501952e-01 -7.30158567e-01 -8.99158511e-03
-5.88886619e-01 3.73876959e-01 3.31882015e-02 4.72611606e-01
1.65755913e-01 -1.50267863e+00 -1.09197533e+00 1.82322681e-01
1.11173622e-01 -8.21945351e-03 6.04802072e-01 3.25413436e-01
-3.65378290e-01 2.94258147e-01 -2.80783623e-01 2.73947548e-02
-9.27590907e-01 2.48448446e-01 3.54945600e-01 -9.59602177e-01
-1.01182267e-01 8.16738665e-01 -1.40419409e-01 -1.06548524e+00
-1.15232922e-01 4.55943011e-02 -2.78637767e-01 1.28011167e-01
5.18815160e-01 5.36446571e-01 2.81194538e-01 -7.93294668e-01
-6.04334295e-01 2.02378288e-01 -1.61059380e-01 2.70226687e-01
1.49849641e+00 2.26741314e-01 2.53648460e-01 -7.37046972e-02
1.27066958e+00 2.69316703e-01 -9.53172326e-01 -3.05465728e-01
-1.21591993e-01 1.43047683e-02 -7.32350647e-02 -1.28735697e+00
-1.22698200e+00 8.80053222e-01 7.81708777e-01 -2.47106850e-02
1.16998589e+00 -3.98582578e-01 1.15001857e+00 3.65833819e-01
1.43458098e-01 -8.82230341e-01 -2.43604422e-01 1.04506457e+00
5.76770484e-01 -1.34687710e+00 -2.71244079e-01 -2.16308367e-02
-7.38187969e-01 6.98615193e-01 7.48757839e-01 -1.30345762e-01
6.78525805e-01 3.15382957e-01 1.98711485e-01 2.67607182e-01
-4.42802280e-01 -4.58851844e-01 4.42931443e-01 6.65337205e-01
5.69816351e-01 2.62938768e-01 -2.81301379e-01 1.27272618e+00
1.80794690e-02 -1.42078683e-01 1.06553316e-01 6.82799697e-01
1.60782188e-01 -1.26081705e+00 -1.59991100e-01 4.33947921e-01
-1.14666116e+00 -4.48871315e-01 -1.59786880e-01 9.71645117e-01
2.41592616e-01 7.68691480e-01 4.83108573e-02 -6.26070142e-01
7.54130721e-01 2.34218270e-01 -3.65459323e-02 -6.17251873e-01
-1.19983840e+00 -3.60924423e-01 3.86534750e-01 -2.63763785e-01
-2.81013519e-01 -2.55106300e-01 -1.15033519e+00 -6.13976598e-01
-3.82986426e-01 2.70512611e-01 7.47754395e-01 7.73006976e-01
8.59651566e-01 7.88596511e-01 7.09021688e-01 -4.89838034e-01
-6.18546128e-01 -1.21978295e+00 -4.56576586e-01 4.83234316e-01
1.01875544e-01 -4.21642154e-01 -1.74524605e-01 -2.62619376e-01] | [9.793269157409668, 9.572829246520996] |
cf0a8189-d683-4389-84f9-304c4af82ede | benchmarking-single-image-reflection-removal | null | null | http://openaccess.thecvf.com/content_iccv_2017/html/Wan_Benchmarking_Single-Image_Reflection_ICCV_2017_paper.html | http://openaccess.thecvf.com/content_ICCV_2017/papers/Wan_Benchmarking_Single-Image_Reflection_ICCV_2017_paper.pdf | Benchmarking Single-Image Reflection Removal Algorithms | Removing undesired reflections from a photo taken in front of a glass is of great importance for enhancing the efficiency of visual computing systems. Various approaches have been proposed and shown to be visually plausible on small datasets collected by their authors. A quantitative comparison of existing approaches using the same dataset has never been conducted due to the lack of suitable benchmark data with ground truth. This paper presents the first captured Single-image Reflection Removal dataset 'SIR2' with 40 controlled and 100 wild scenes, ground truth of background and reflection. For each controlled scene, we further provide ten sets of images under varying aperture settings and glass thicknesses. We perform quantitative and visual quality comparisons for four state-of-the-art singleimage reflection removal algorithms using four error metrics. Open problems for improving reflection removal algorithms are discussed at the end.
| ['Ah-Hwee Tan', 'Ling-Yu Duan', 'Alex C. Kot', 'Renjie Wan', 'Boxin Shi'] | 2017-10-01 | null | null | null | iccv-2017-10 | ['reflection-removal'] | ['computer-vision'] | [ 9.13348734e-01 -7.96856955e-02 9.54436302e-01 -1.77951068e-01
-9.64775860e-01 -3.44888479e-01 5.06049216e-01 -1.85224429e-01
-6.17238097e-02 4.56819981e-01 2.77471572e-01 -1.07590474e-01
4.40113395e-02 -4.60424304e-01 -5.85227251e-01 -8.80061805e-01
1.28236637e-01 1.70023013e-02 6.18912697e-01 -1.59832567e-01
3.83272737e-01 5.97029448e-01 -1.75063014e+00 6.25274539e-01
5.90041101e-01 7.55903959e-01 2.83273041e-01 7.60604203e-01
4.46630001e-01 5.48777461e-01 -6.79604173e-01 -4.00876552e-01
8.51660490e-01 -5.28950453e-01 -2.72626668e-01 5.03564835e-01
1.21608782e+00 -2.80556113e-01 -2.45742872e-01 9.42119360e-01
7.80019283e-01 7.52217248e-02 4.87639576e-01 -7.98566878e-01
-4.53342557e-01 -3.22388671e-02 -8.42069268e-01 1.01379395e-01
5.94642043e-01 4.52257037e-01 5.56615889e-01 -7.17674553e-01
4.96221870e-01 1.08141625e+00 7.65501082e-01 2.41805956e-01
-1.23186350e+00 -4.77751315e-01 -2.27991328e-01 -1.58708930e-01
-1.10766506e+00 -7.09148765e-01 7.41778672e-01 -2.05979094e-01
1.00790572e+00 6.63126171e-01 5.61465204e-01 7.80429602e-01
9.34378654e-02 2.56180942e-01 1.55163074e+00 -7.92449832e-01
2.89822798e-02 3.35114598e-01 1.49815395e-01 6.05726540e-01
6.29745126e-01 3.12607974e-01 -3.30564559e-01 2.22468749e-02
4.72808123e-01 -2.01250225e-01 -6.48047447e-01 -1.42195955e-01
-7.60646045e-01 9.79798660e-02 3.73538673e-01 7.61009604e-02
-1.48094706e-02 -1.03948288e-01 -1.43556714e-01 3.20365638e-01
6.28043413e-01 3.96379530e-01 -3.15729678e-02 5.36764145e-01
-7.51826167e-01 1.69345528e-01 6.55155599e-01 9.50289011e-01
5.65676928e-01 1.74812838e-01 -1.88946977e-01 8.62147689e-01
4.35020506e-01 9.06506240e-01 -2.11121172e-01 -8.51545453e-01
4.56735194e-01 4.84103620e-01 5.94186366e-01 -1.04369164e+00
-2.53840506e-01 -2.75890946e-01 -7.33794391e-01 8.06834519e-01
3.31370205e-01 1.03550628e-01 -1.11048079e+00 7.37763107e-01
1.65413916e-01 1.02816619e-01 2.03386247e-01 1.11897588e+00
1.15180063e+00 5.84872186e-01 -5.64188778e-01 -3.13967228e-01
1.15650487e+00 -9.84448552e-01 -5.12711525e-01 -3.62076372e-01
-1.13124400e-01 -1.45513022e+00 1.04090881e+00 9.47256625e-01
-1.19334948e+00 -3.15508634e-01 -1.02884328e+00 -7.93781057e-02
1.33796096e-01 4.88847047e-01 3.87189239e-01 1.08528531e+00
-9.32699919e-01 3.41959029e-01 -3.61921579e-01 -1.97682202e-01
2.86676407e-01 -1.08662099e-01 -2.00898528e-01 -5.32632530e-01
-5.43520272e-01 7.71768212e-01 -5.42713463e-01 4.22788471e-01
-1.00337565e+00 -7.08353221e-01 -5.72874010e-01 -4.21085000e-01
3.55621934e-01 -5.42143583e-01 9.28094208e-01 -7.96455085e-01
-1.44345987e+00 1.42080057e+00 -3.75709198e-02 -3.44501227e-01
6.96564913e-01 -4.77409422e-01 -6.91887617e-01 1.48263156e-01
-3.79374892e-01 -4.05984432e-01 1.08603263e+00 -2.00905371e+00
-2.10001186e-01 -2.77906507e-01 3.14308964e-02 1.84539765e-01
3.83621544e-01 3.61825854e-01 -6.41362607e-01 -2.88701564e-01
4.87148426e-02 -8.17491949e-01 -3.41175675e-01 -1.47098899e-01
-6.80032551e-01 7.60993659e-01 5.62438190e-01 -5.58925211e-01
8.87538791e-01 -2.21990323e+00 -3.53582919e-01 1.29514616e-02
1.04940355e-01 6.60322249e-01 -1.94427073e-01 5.49424529e-01
-8.49019364e-02 -1.98319048e-01 -2.79089838e-01 -6.91078603e-01
-4.04072493e-01 -3.52170050e-01 -3.64068389e-01 8.93668175e-01
-1.13836251e-01 3.21713626e-01 -4.84520286e-01 -1.25126943e-01
5.48158169e-01 7.66437948e-01 -3.07886004e-01 2.82010645e-01
2.99536996e-02 3.22498113e-01 -1.10353105e-01 6.33331001e-01
1.17124760e+00 1.53846353e-01 7.25432411e-02 -6.54426336e-01
-4.17218357e-01 -1.37738600e-01 -1.29321826e+00 1.01039088e+00
-6.11809969e-01 9.18762565e-01 1.97237521e-01 -2.24100992e-01
1.06867695e+00 -6.51838556e-02 1.81882069e-01 -1.08877134e+00
8.28408897e-02 4.91499864e-02 -2.58591622e-01 -6.47222281e-01
5.20515323e-01 -2.42104515e-01 3.03981006e-01 1.74502969e-01
-7.38372803e-01 -6.70151889e-01 -1.77184977e-02 4.18833792e-02
1.21128988e+00 9.81091894e-03 -1.38438111e-02 -2.79371701e-02
5.30099213e-01 3.53197046e-02 1.89960510e-01 8.34444940e-01
1.30274102e-01 1.60679996e+00 -1.92404628e-01 -3.71970266e-01
-1.01280403e+00 -1.29813576e+00 -5.25260121e-02 5.14233291e-01
4.25089002e-01 -2.41423339e-01 -6.71270907e-01 2.72408009e-01
-3.07194263e-01 5.37346661e-01 -5.13190687e-01 1.81120753e-01
-4.47945237e-01 -1.04608452e+00 2.17694893e-01 3.84227484e-02
6.78664744e-01 -1.01445580e+00 -1.03468144e+00 -2.93011129e-01
-1.33694857e-01 -1.51443875e+00 2.48817895e-02 -2.57244855e-01
-6.36994600e-01 -1.57149315e+00 -5.13453722e-01 -4.96921241e-01
8.17856371e-01 1.00469613e+00 1.61908352e+00 4.37549621e-01
-1.10392642e+00 6.33748651e-01 -3.02131265e-01 -4.99078751e-01
-3.58338207e-01 -8.44806671e-01 -3.55844706e-01 1.33184418e-01
1.39241340e-02 -2.44370520e-01 -8.97870600e-01 5.23564696e-01
-9.61950362e-01 1.93862274e-01 5.12303531e-01 3.00930560e-01
4.81414974e-01 2.24138781e-01 -5.46062887e-01 -1.02151418e+00
4.08356190e-01 2.09109306e-01 -9.86282706e-01 2.01532066e-01
-1.73076987e-01 -3.76327038e-01 4.30982053e-01 4.43716869e-02
-1.65722024e+00 -7.41313398e-02 3.74284208e-01 -1.69665471e-01
-3.32200527e-01 -5.02358258e-01 -1.71221182e-01 -3.01567733e-01
8.91880333e-01 -6.31010905e-03 -2.55234867e-01 -5.83409905e-01
1.69122800e-01 4.77627724e-01 5.66653252e-01 -1.78192183e-01
8.18259895e-01 1.05431080e+00 1.66722938e-01 -1.51000035e+00
-1.13315415e+00 -8.02769184e-01 -3.67037892e-01 -4.51382220e-01
6.80531144e-01 -8.89721453e-01 -6.69432402e-01 7.97146857e-01
-8.37351680e-01 -5.79871774e-01 -5.68233319e-02 2.08251089e-01
-3.81993711e-01 5.58029830e-01 -3.66836131e-01 -1.13461709e+00
-3.38726670e-01 -1.09403229e+00 1.33839703e+00 2.53081381e-01
3.21094722e-01 -6.02522314e-01 2.40611404e-01 7.96253860e-01
5.10220647e-01 2.37269074e-01 3.37474138e-01 6.22902691e-01
-8.61848235e-01 -1.26240477e-01 -5.30096114e-01 6.21716082e-01
1.50420532e-01 2.46016145e-01 -1.47394741e+00 -3.66385937e-01
2.35094860e-01 3.51428054e-02 1.11247039e+00 4.82992828e-01
6.91672683e-01 -2.22127140e-02 -3.01056474e-01 6.77124143e-01
1.95805347e+00 -5.65170757e-02 1.45260584e+00 2.80342102e-01
6.51529908e-01 6.36550426e-01 7.79185951e-01 1.82292446e-01
-2.51564920e-01 6.57590389e-01 6.59194767e-01 -5.36731303e-01
-7.26911604e-01 4.20489490e-01 3.72213364e-01 3.10026348e-01
-4.58636701e-01 -7.94056177e-01 -6.34372294e-01 3.95211518e-01
-1.20631862e+00 -1.02177250e+00 -8.94044459e-01 2.43013000e+00
1.46008998e-01 -1.58299357e-01 -8.93102735e-02 4.09571171e-01
4.49269354e-01 7.24541545e-02 4.21243459e-02 -1.33051634e-01
-5.13481736e-01 2.35917732e-01 6.65987074e-01 8.85558546e-01
-8.25773656e-01 6.18173361e-01 6.71540165e+00 1.59673259e-01
-9.39929545e-01 -1.15683466e-01 5.56628823e-01 -2.28907317e-01
-3.02572280e-01 -3.24616358e-02 -8.32324445e-01 8.95785093e-02
5.60525596e-01 4.48800117e-01 2.93371230e-01 2.06492156e-01
5.81711173e-01 -7.46426105e-01 -6.40188277e-01 1.21795499e+00
5.51625311e-01 -8.46676290e-01 -1.94489971e-01 -2.15221480e-01
9.31716561e-01 1.52755648e-01 1.81984648e-01 -6.68436766e-01
2.43506193e-01 -1.02474797e+00 3.23803872e-01 9.77601886e-01
6.28011584e-01 -4.05174822e-01 3.65136117e-01 -2.84536958e-01
-9.24642742e-01 2.02932879e-01 -4.81613070e-01 2.17120633e-01
1.92275286e-01 7.42640674e-01 -6.00028038e-01 6.95929050e-01
1.10147774e+00 4.52692986e-01 -6.77481890e-01 1.30520117e+00
-2.38389984e-01 4.59510624e-01 -4.23065186e-01 2.78817713e-01
-1.24024890e-01 -6.60569191e-01 7.18367994e-01 1.30898058e+00
1.28028929e-01 1.63812086e-01 -2.05233872e-01 6.09082818e-01
4.94107425e-01 -2.10420147e-01 -6.51773036e-01 4.45491493e-01
-1.19095504e-01 1.30684221e+00 -7.07278550e-01 4.67208810e-02
-5.78644574e-01 1.08861780e+00 -2.20440730e-01 5.93725979e-01
-6.17389917e-01 -1.93433970e-01 6.28908038e-01 7.42402911e-01
5.12985848e-02 -2.49252036e-01 -2.02765554e-01 -9.69752192e-01
2.49261379e-01 -7.90606558e-01 1.43135816e-01 -1.41893768e+00
-1.16604066e+00 6.77645564e-01 3.37898470e-02 -1.41640866e+00
5.30868411e-01 -9.76092756e-01 -5.10054052e-01 8.36442232e-01
-1.62244415e+00 -1.03343189e+00 -1.09121144e+00 6.27002180e-01
4.09244269e-01 7.32647115e-03 5.75331092e-01 4.24685240e-01
-3.70038003e-01 1.42672313e-02 4.09675270e-01 -3.09588850e-01
7.74164259e-01 -8.59470367e-01 2.95042366e-01 1.13766766e+00
-2.40955204e-02 4.42972571e-01 1.26207089e+00 -3.46314371e-01
-1.57008731e+00 -8.96460414e-01 3.65003675e-01 -5.26997268e-01
1.40731320e-01 -3.79757762e-01 -8.37394655e-01 5.00698090e-01
4.17192787e-01 -1.60700437e-02 4.33477104e-01 -3.46764863e-01
-1.97775036e-01 -3.40465844e-01 -1.14237893e+00 7.70826876e-01
1.11078990e+00 -1.27554238e-01 -2.15045035e-01 3.66816968e-01
2.04861745e-01 -4.28496480e-01 -3.23516965e-01 3.12247097e-01
6.06992781e-01 -1.80030930e+00 1.47751319e+00 5.33397645e-02
3.01701605e-01 -5.66917241e-01 -2.53817499e-01 -1.14202571e+00
8.31193998e-02 -7.13797927e-01 5.44770896e-01 1.09759200e+00
2.71467656e-01 -6.05442703e-01 6.06074512e-01 4.27264035e-01
-1.86801046e-01 -5.56419641e-02 -1.84516415e-01 -6.16452336e-01
-5.70644379e-01 -5.74766457e-01 1.50761306e-01 5.22972226e-01
-8.16208899e-01 2.35619053e-01 -7.27123141e-01 5.41706443e-01
1.12008142e+00 6.18825138e-01 1.38083899e+00 -9.86858428e-01
-2.74638087e-01 -1.08379543e-01 -3.91091019e-01 -6.38509572e-01
-4.62410629e-01 -2.62920350e-01 2.43330762e-01 -1.81053603e+00
2.85157323e-01 -3.54045182e-01 1.91414967e-01 6.09781221e-02
-1.35576446e-02 8.97394657e-01 1.18850656e-02 -5.84043264e-02
-6.12330437e-01 3.12968105e-01 1.25081146e+00 1.29384836e-02
-1.30266368e-01 1.48121826e-02 -6.18314564e-01 8.60886574e-01
7.33865678e-01 -3.20818156e-01 -5.02967536e-01 -5.67303896e-01
4.41837430e-01 -1.55615449e-01 7.28604674e-01 -1.17091191e+00
-1.48014471e-01 1.43661033e-02 3.22349787e-01 -4.97419029e-01
7.98397303e-01 -1.03658795e+00 4.57274079e-01 1.93521529e-01
1.27446234e-01 -2.92910814e-01 3.48015666e-01 7.00043678e-01
7.32879853e-03 -1.39755696e-01 1.14149797e+00 -3.04668963e-01
-5.91686010e-01 -1.88366994e-01 -1.70757487e-01 1.49747446e-01
8.70427251e-01 -4.53567117e-01 -6.35411859e-01 -2.74415106e-01
-5.32897413e-01 -3.21727037e-01 8.63610208e-01 1.63191855e-01
9.17055845e-01 -5.80159068e-01 -8.18137884e-01 1.87238067e-01
2.22585365e-01 -2.19173551e-01 6.26907885e-01 6.60310984e-01
-9.78895962e-01 -1.36628589e-02 -1.35311320e-01 -5.25633574e-01
-2.15550351e+00 3.97010297e-01 6.22376323e-01 7.44257541e-03
-1.02077556e+00 5.29657364e-01 4.72727865e-01 9.89019945e-02
4.93007191e-02 -4.57346350e-01 7.11585805e-02 -4.15802926e-01
7.66933441e-01 6.22261405e-01 4.49074924e-01 -6.07918084e-01
-2.61717260e-01 9.72577035e-01 1.95592657e-01 2.85569988e-02
1.51430166e+00 -3.88091117e-01 8.97726864e-02 2.15034530e-01
8.73431265e-01 5.77466607e-01 -1.03494585e+00 -1.54567510e-01
-6.06178820e-01 -1.14014435e+00 9.96556655e-02 -7.79204845e-01
-1.26742315e+00 7.80756831e-01 7.73938596e-01 1.13370195e-01
1.51939690e+00 -3.33151639e-01 2.29167983e-01 1.64714172e-01
2.40618140e-01 -7.40629435e-01 3.07945728e-01 7.65888020e-02
1.15767062e+00 -1.42947292e+00 4.78182077e-01 -1.08920658e+00
-7.61153519e-01 9.32201803e-01 3.47268075e-01 -2.44886383e-01
5.39394379e-01 6.49872839e-01 5.28617799e-01 -5.58017075e-01
-3.94169390e-01 -3.37120116e-01 4.23888117e-01 8.80113482e-01
5.23860097e-01 -2.91589469e-01 1.13394624e-02 -8.73518884e-02
-9.27252248e-02 -2.55042434e-01 7.60756016e-01 7.88025200e-01
-3.82816613e-01 -6.54754162e-01 -6.97474539e-01 1.89928904e-01
-5.52700818e-01 -1.50380537e-01 -5.62750220e-01 7.35332489e-01
-3.22550774e-01 1.31403399e+00 -2.20239118e-01 1.97232082e-01
5.95189035e-01 -7.34821200e-01 9.23332334e-01 -5.93864620e-01
-6.98269784e-01 1.24615766e-01 3.84904206e-01 -6.79070771e-01
-8.18772018e-01 -5.41686654e-01 -7.27159441e-01 -9.77689028e-02
-2.88341552e-01 -2.76546627e-01 5.50142348e-01 3.30886871e-01
4.67165485e-02 5.87567747e-01 4.06000137e-01 -9.71313357e-01
7.29562417e-02 -7.49397457e-01 -8.42216611e-01 7.67403245e-01
3.84600371e-01 -4.64461625e-01 -7.40454853e-01 2.06561595e-01] | [10.50070858001709, -2.8453896045684814] |
54c07edc-7d93-4b00-838d-d2cd205b3370 | translation-transformers-rediscover-inherent | 2109.07864 | null | https://arxiv.org/abs/2109.07864v1 | https://arxiv.org/pdf/2109.07864v1.pdf | Translation Transformers Rediscover Inherent Data Domains | Many works proposed methods to improve the performance of Neural Machine Translation (NMT) models in a domain/multi-domain adaptation scenario. However, an understanding of how NMT baselines represent text domain information internally is still lacking. Here we analyze the sentence representations learned by NMT Transformers and show that these explicitly include the information on text domains, even after only seeing the input sentences without domains labels. Furthermore, we show that this internal information is enough to cluster sentences by their underlying domains without supervision. We show that NMT models produce clusters better aligned to the actual domains compared to pre-trained language models (LMs). Notably, when computed on document-level, NMT cluster-to-domain correspondence nears 100%. We use these findings together with an approach to NMT domain adaptation using automatically extracted domains. Whereas previous work relied on external LMs for text clustering, we propose re-using the NMT model as a source of unsupervised clusters. We perform an extensive experimental study comparing two approaches across two data scenarios, three language pairs, and both sentence-level and document-level clustering, showing equal or significantly superior performance compared to LMs. | ['Mark Fishel', 'Elizaveta Korotkova', 'Maksym Del'] | 2021-09-16 | null | https://aclanthology.org/2021.wmt-1.65 | https://aclanthology.org/2021.wmt-1.65.pdf | wmt-emnlp-2021-11 | ['text-clustering'] | ['natural-language-processing'] | [ 4.81826156e-01 1.09490275e-01 -4.26133156e-01 -5.94546497e-01
-1.19457138e+00 -9.94618535e-01 9.55516160e-01 3.40056904e-02
-3.89475346e-01 9.25288618e-01 4.32140827e-01 -3.45259994e-01
1.12701073e-01 -2.97874063e-01 -8.30265880e-01 -5.49187243e-01
4.32808369e-01 1.31051755e+00 -4.91433442e-02 -2.68548578e-01
6.33033738e-02 -2.94644199e-02 -1.07196498e+00 8.34510505e-01
1.22932279e+00 2.24500179e-01 2.83489585e-01 5.06265283e-01
-4.80051130e-01 5.23423672e-01 -8.76611233e-01 -2.75044411e-01
3.91833216e-01 -8.01844060e-01 -9.86828923e-01 2.48031959e-01
6.30030274e-01 1.47888213e-01 -1.29811317e-01 9.46412623e-01
4.04363424e-01 -3.62348370e-02 1.15251923e+00 -9.38808441e-01
-9.91376519e-01 1.02386975e+00 -5.48451543e-01 4.36812863e-02
1.74666598e-01 -9.53471437e-02 9.02155399e-01 -8.30452025e-01
1.25828958e+00 1.24622941e+00 5.13471305e-01 8.78574491e-01
-1.61099887e+00 -5.98942637e-01 7.16427565e-02 1.07598461e-01
-9.76176977e-01 -5.74666440e-01 7.13035345e-01 -4.55260962e-01
1.12924576e+00 -1.32560432e-01 -9.26863328e-02 1.50246656e+00
-1.52485922e-01 7.05747962e-01 1.41713762e+00 -9.86165762e-01
2.33540729e-01 6.70173764e-01 2.16568127e-01 -1.28449410e-01
1.92322269e-01 -2.98357904e-01 -4.62543935e-01 7.83569296e-04
4.44092005e-01 -4.95192498e-01 1.57135800e-01 -5.91863930e-01
-1.53671801e+00 9.28098619e-01 -4.63714413e-02 7.78116763e-01
-7.92571604e-02 -4.29708481e-01 6.93701625e-01 7.21959591e-01
8.70660067e-01 6.69980109e-01 -7.24422216e-01 3.37726958e-02
-1.25669372e+00 -6.16818145e-02 7.18606412e-01 1.31085885e+00
7.50555575e-01 -4.19793352e-02 -1.38364211e-01 1.14765120e+00
-1.56478122e-01 5.21591067e-01 9.12834883e-01 -8.47827315e-01
1.06607485e+00 6.70397520e-01 -1.06281862e-01 -5.76041639e-01
-1.86060742e-01 -4.03285146e-01 -8.64005029e-01 -2.15588421e-01
4.97051179e-01 -3.72135282e-01 -8.02498877e-01 1.89357567e+00
-9.24707428e-02 -1.63631290e-01 5.76876938e-01 5.42002976e-01
4.77455556e-01 5.96128941e-01 -3.82272080e-02 -3.80194038e-01
1.01054394e+00 -1.02467203e+00 -8.22641134e-01 -3.19778532e-01
1.13846469e+00 -8.19119811e-01 1.27714622e+00 1.59392744e-01
-1.08420110e+00 -6.97875857e-01 -8.97560000e-01 3.01911030e-02
-6.40999734e-01 4.19731885e-01 2.30460968e-02 4.32922840e-01
-1.41712713e+00 5.82906902e-01 -5.35638452e-01 -9.83219147e-01
2.45239779e-01 5.19463480e-01 -3.55514795e-01 1.07742567e-02
-1.31047344e+00 1.24212074e+00 5.40372074e-01 -8.09808016e-01
-6.19863510e-01 -3.78575236e-01 -6.20296121e-01 -2.83481717e-01
-8.22232515e-02 -6.74370229e-01 1.33238602e+00 -1.49844587e+00
-1.43971896e+00 1.19868755e+00 -6.63044572e-01 -7.61635303e-01
3.34281236e-01 -1.87156033e-02 -3.68363291e-01 3.05559576e-01
4.70639050e-01 8.38013172e-01 6.79998577e-01 -1.44742441e+00
-4.64594990e-01 -3.16846251e-01 -3.58813167e-01 4.60564613e-01
-9.49785769e-01 1.93593830e-01 -5.88700734e-02 -5.64390361e-01
-4.00179885e-02 -9.77232456e-01 -1.78205863e-01 -6.18271828e-01
-3.27502906e-01 -4.53523457e-01 7.05054283e-01 -7.93128192e-01
1.09895682e+00 -1.96284020e+00 3.90566379e-01 1.30244838e-02
-1.83375515e-02 1.70299888e-01 -3.99365515e-01 6.27743363e-01
-2.61436075e-01 3.16785723e-01 -4.20552462e-01 -7.15046585e-01
2.12728113e-01 3.45727414e-01 -4.38729286e-01 1.90167353e-01
3.93165171e-01 8.60517919e-01 -6.70078397e-01 -7.63428867e-01
-1.54420501e-02 -3.93556245e-02 -4.70337480e-01 4.18910161e-02
-3.79334837e-01 6.49520099e-01 1.40396627e-02 1.44748211e-01
7.14578450e-01 -3.10287267e-01 7.40650237e-01 3.37798119e-01
1.05716065e-01 6.22531474e-01 -7.55174339e-01 1.94078565e+00
-3.35276008e-01 9.06555831e-01 9.23914239e-02 -1.46023881e+00
1.10658514e+00 5.02029419e-01 4.24139857e-01 -6.69010282e-01
-1.80995136e-01 4.15607154e-01 1.02510870e-01 -2.72729546e-01
4.40706939e-01 -2.66239047e-01 -2.08689839e-01 7.18656003e-01
4.84525621e-01 2.51646247e-02 4.18437779e-01 4.05284703e-01
9.81292725e-01 2.18469560e-01 1.34759471e-01 -4.69789565e-01
3.41906220e-01 7.66443729e-01 3.40384066e-01 5.64308047e-01
-2.57466845e-02 7.67352939e-01 4.58290398e-01 3.16575430e-02
-1.43601203e+00 -1.04515040e+00 -1.53629914e-01 1.37291408e+00
-2.16280252e-01 -2.88727254e-01 -1.10514414e+00 -9.50646996e-01
-2.28950292e-01 8.47668171e-01 -6.30078495e-01 -1.12873323e-01
-8.28585148e-01 -6.71700656e-01 6.49202168e-01 3.60769391e-01
2.71079421e-01 -8.82666111e-01 1.55451342e-01 2.49657393e-01
-6.08552575e-01 -1.30697012e+00 -4.13530976e-01 5.74740946e-01
-1.23167598e+00 -5.31992078e-01 -8.12309086e-01 -1.28619277e+00
7.49263823e-01 2.22533464e-01 1.50028050e+00 -4.18633193e-01
4.21373516e-01 1.21602021e-01 -5.98944187e-01 -2.17385426e-01
-1.11782742e+00 6.42652631e-01 4.00617629e-01 -3.41386974e-01
9.70355511e-01 -6.85317755e-01 -5.23657240e-02 3.25280905e-01
-8.01039338e-01 4.18211855e-02 7.39170074e-01 7.66875803e-01
2.94542402e-01 -2.01992571e-01 8.98649156e-01 -1.09318697e+00
8.31147552e-01 -6.00983441e-01 -6.73648715e-02 3.53621572e-01
-4.80528265e-01 2.64951110e-01 1.02891755e+00 -6.71390235e-01
-1.06687868e+00 1.14331692e-01 1.93647325e-01 -4.92293358e-01
-6.31436467e-01 2.24574506e-01 -1.55826882e-01 5.65558374e-01
1.11615515e+00 4.65421557e-01 3.99319977e-02 -5.35067797e-01
4.61980790e-01 1.08959126e+00 3.71403009e-01 -8.77928436e-01
9.59511280e-01 2.85061508e-01 -5.71788609e-01 -6.44926727e-01
-6.11283839e-01 -5.89711607e-01 -1.47187340e+00 2.92174608e-01
8.57388854e-01 -1.09442699e+00 3.56319785e-01 -2.60169178e-01
-1.40003228e+00 -5.20531476e-01 -2.63173312e-01 4.55289781e-01
-6.76536381e-01 4.50018793e-01 -5.86916029e-01 -5.03111780e-01
-1.98983774e-01 -9.08385158e-01 1.09129560e+00 -2.69778430e-01
-7.99200952e-01 -1.54181886e+00 2.41412252e-01 3.94439787e-01
2.54391998e-01 9.26251486e-02 1.33926332e+00 -1.26657581e+00
-6.53412119e-02 3.01748067e-01 -1.52451443e-02 4.84730870e-01
2.17188135e-01 -2.10300490e-01 -1.02663088e+00 -3.94896448e-01
6.93300646e-03 -1.35246813e-01 6.72008216e-01 3.44925851e-01
5.22416592e-01 -2.49003381e-01 -4.55806106e-01 3.06947142e-01
1.27631640e+00 4.10700627e-02 2.42629796e-01 5.87961316e-01
5.09114206e-01 1.00259376e+00 5.12542844e-01 9.78079140e-02
4.19882596e-01 7.58142829e-01 -8.30407068e-02 -3.66977155e-01
-3.98344368e-01 -2.66218394e-01 8.77577960e-01 1.44053006e+00
4.11436468e-01 -9.75781232e-02 -1.06663656e+00 9.36383188e-01
-1.82289088e+00 -9.61296678e-01 -3.41080964e-01 1.93987715e+00
1.22716844e+00 1.62388295e-01 3.59952033e-01 -1.70558870e-01
9.35128927e-01 -4.15224761e-01 -6.35713160e-01 -7.77490735e-01
-5.90862215e-01 2.17690557e-01 3.65960449e-01 2.32537925e-01
-1.11825860e+00 1.22034109e+00 6.94524431e+00 9.14279878e-01
-7.30225682e-01 3.30711126e-01 2.78139055e-01 -4.38036993e-02
-3.32936555e-01 -4.01215926e-02 -7.18242586e-01 5.01086175e-01
1.42800653e+00 -2.65949488e-01 4.01810020e-01 7.88338065e-01
3.11844796e-01 3.46410364e-01 -1.44720745e+00 6.50551617e-01
1.67938411e-01 -1.16168785e+00 3.26611161e-01 2.16313511e-01
1.28658235e+00 1.99977666e-01 4.38880026e-02 5.63069701e-01
7.67577231e-01 -8.77770603e-01 5.72668612e-01 2.85245869e-02
8.91039133e-01 -5.28302670e-01 6.52719021e-01 7.42131650e-01
-6.95129395e-01 8.44290480e-02 -6.35788918e-01 -9.20857489e-02
-2.84817189e-01 4.26984191e-01 -1.48848736e+00 7.74320662e-01
5.20342171e-01 9.65420723e-01 -8.64129722e-01 2.70602942e-01
-1.23757925e-02 8.72817814e-01 -6.46846667e-02 5.66645041e-02
2.95717090e-01 -1.97694078e-01 3.46892208e-01 1.68031585e+00
2.34213531e-01 -5.89320302e-01 8.11682120e-02 9.05459404e-01
-2.16848195e-01 3.11508477e-01 -1.02793777e+00 -8.30525607e-02
6.00861013e-01 9.32866454e-01 -6.50708675e-01 -8.14552009e-01
-4.49586481e-01 1.20662081e+00 4.14096236e-01 4.90880162e-01
-6.09853923e-01 -2.12240353e-01 7.67148137e-01 1.25374105e-02
1.05567552e-01 -3.58395249e-01 -9.13083434e-01 -1.24251699e+00
1.18004806e-01 -1.19680429e+00 2.67750800e-01 -5.94256043e-01
-1.68619990e+00 6.92533612e-01 1.55915707e-01 -1.55593729e+00
-6.86719418e-01 -5.86030900e-01 -4.48253721e-01 9.44222629e-01
-1.39504027e+00 -1.11628747e+00 3.31598282e-01 6.54577792e-01
8.99760425e-01 -5.73317349e-01 1.01796925e+00 1.22804902e-01
-3.60395104e-01 7.99864531e-01 1.01957309e+00 4.47842747e-01
1.44330955e+00 -1.43808174e+00 5.84088385e-01 8.15042675e-01
2.57746518e-01 9.86603975e-01 6.14756584e-01 -6.76805556e-01
-8.86416316e-01 -1.33159065e+00 1.16424298e+00 -1.19568837e+00
4.73839611e-01 -6.83790565e-01 -9.87260520e-01 7.52692103e-01
9.91030216e-01 -7.28067279e-01 9.80533481e-01 2.43476242e-01
-2.59181887e-01 2.44443238e-01 -9.99630451e-01 5.40536046e-01
1.05300570e+00 -5.48415959e-01 -1.14091456e+00 7.04834938e-01
9.46505904e-01 3.91532518e-02 -9.27771926e-01 1.40059218e-01
5.10870554e-02 -6.89794242e-01 7.86673248e-01 -7.32351124e-01
6.81182086e-01 -2.18860120e-01 -2.99703717e-01 -1.61738241e+00
-3.45304817e-01 -3.01352859e-01 2.64288306e-01 1.46409619e+00
8.26953471e-01 -4.33071673e-01 7.04850614e-01 9.93044749e-02
-3.33440632e-01 -4.58705649e-02 -8.93209577e-01 -1.11800086e+00
1.09169352e+00 -6.65724929e-03 4.83398497e-01 1.60231829e+00
2.82295078e-01 7.88911402e-01 -8.01990107e-02 -1.01141416e-01
4.91394252e-01 1.18694574e-01 9.54368889e-01 -1.34994197e+00
-1.50401905e-01 -4.92291600e-01 1.72863707e-01 -9.75000739e-01
5.78403056e-01 -1.50394034e+00 -6.59637749e-02 -1.71741116e+00
3.25103730e-01 -8.79580975e-02 -1.26151636e-01 3.36821079e-01
1.06981158e-01 5.19782677e-02 2.33584672e-01 6.63102567e-01
-7.72473514e-01 2.04646930e-01 1.12116551e+00 -2.48230115e-01
-2.10585266e-01 -3.21944803e-01 -8.68581116e-01 6.15891695e-01
1.18238902e+00 -7.16507494e-01 -3.71448547e-01 -9.29838300e-01
-2.06958547e-01 -2.63263762e-01 -1.51732475e-01 -1.11711431e+00
2.27212369e-01 -1.78246588e-01 5.09219825e-01 -5.40870905e-01
8.42732340e-02 -6.37590289e-01 -3.20224166e-01 3.89023088e-02
-8.43528152e-01 1.35869533e-01 2.89620787e-01 3.17463815e-01
-3.45888168e-01 -1.78604394e-01 7.51973152e-01 -3.80254388e-01
-5.34169853e-01 -3.20393622e-01 -5.81604242e-01 2.71863669e-01
8.26773822e-01 -4.61533844e-01 -3.04788768e-01 -4.52761769e-01
-6.88464582e-01 1.84736386e-01 8.66357088e-01 4.85310227e-01
1.94619611e-01 -1.41453516e+00 -1.03286183e+00 1.25253901e-01
1.77764624e-01 -1.56953350e-01 -2.11387470e-01 7.96269715e-01
7.84905776e-02 6.45227134e-01 -1.65146753e-01 -9.77375805e-01
-1.20403659e+00 5.15836477e-01 1.83245689e-02 -6.10004604e-01
-1.81978151e-01 4.29765403e-01 1.53354108e-01 -1.13098776e+00
6.31355029e-03 -2.53484100e-01 -1.59321234e-01 2.30363980e-01
5.75268939e-02 1.71215776e-02 1.67429611e-01 -6.47694826e-01
-3.29867393e-01 7.48927355e-01 -2.89379597e-01 -5.07734597e-01
1.21830869e+00 -3.39749902e-01 -1.61181957e-01 5.59485674e-01
1.26059103e+00 -1.86875999e-01 -9.50165689e-01 -6.11732185e-01
5.72994828e-01 4.46743742e-02 -7.46299803e-01 -1.02837253e+00
-1.44675255e-01 1.32869506e+00 4.45698857e-01 1.64020613e-01
1.00847983e+00 1.31606817e-01 5.79505622e-01 4.89488333e-01
2.42818639e-01 -1.64463651e+00 3.40708852e-01 9.07230139e-01
5.81750691e-01 -1.29538786e+00 -3.38814199e-01 -5.34432754e-03
-1.01809430e+00 1.02798617e+00 7.19874322e-01 -1.47561207e-01
-1.24795754e-02 1.01426750e-01 4.06183004e-01 4.16228056e-01
-9.34320450e-01 -3.17424595e-01 9.32207555e-02 1.16050482e+00
7.65284956e-01 1.01062953e-02 -1.65347517e-01 5.65558374e-01
-4.63732809e-01 -3.96148950e-01 5.09552956e-01 6.53955579e-01
-5.28427124e-01 -1.58251321e+00 -5.59663177e-01 3.29280555e-01
-4.13242906e-01 -2.08815604e-01 -1.12947690e+00 8.82827163e-01
8.36788416e-02 1.26312554e+00 1.39751405e-01 -5.46160102e-01
1.46433905e-01 7.13013113e-01 2.07501069e-01 -1.16277397e+00
-5.98325908e-01 1.62294582e-01 2.09532678e-02 1.30223736e-01
-6.46237433e-01 -9.72372532e-01 -1.18210149e+00 -1.15221031e-02
-1.07552156e-01 4.41353917e-01 6.06285393e-01 8.89913917e-01
6.32996976e-01 4.10439670e-01 6.26231551e-01 -6.98212981e-01
-6.01556361e-01 -1.48603129e+00 -1.26855806e-01 7.53444135e-01
1.49236947e-01 -3.59205276e-01 -2.94916570e-01 5.63403130e-01] | [11.520003318786621, 10.182589530944824] |
af876539-c342-4736-9870-e2d10bc22ee1 | evaluating-and-predicting-the-efficiency | 2201.07718 | null | https://arxiv.org/abs/2201.07718v1 | https://arxiv.org/pdf/2201.07718v1.pdf | Evaluating and predicting the Efficiency Index for Stereotactic Radiosurgery Plans using RapidMiner GO(JAVA) Based Artificial Intelligence Algorithms | Evaluation the prediction of Efficiency index by DVH parameter for SRS treatment plans using Supervised Machine learning and the performance of predictive model algorithms of RapidMiner GO in the parameter prediction are investigated. Dose volume histogram (DVH) based Efficiency index was calculated for 100 clinical SRS plans generated by Leksell Gamma plan, and the results were compared to predicted values produced by machine learning toolbox of RapidMiner Go, algorithms are namely, Generalized linear model (GLR), Decision Tree Model, Support Vector Machine (SVM), Gradient Boosted Trees (GBT), Random Forest (RF) and Deep learning Model (DL). Root mean square error (RMSE), Average absolute error, Absolute relative error, squared correlation and model building time were determined to evaluate the performance of each algorithm. The GLR algorithm model had square correlation of 0.974 with the smallest RMSE of 0.01, relatively high prediction speed, and fast model building time with 2.812 s, according to the results. The RMSE values for all models were between 0.01 upto 0.021, all algorithms performed well. The RMSE of the Gradient Boosted Tree, Random Forest, and Decision Tree regression algorithms was found to be greater than 0.01, suggesting that they are not appropriate for predicting EI in this analysis. RapidMiner GO machine learning models can be used to predict DVH parameters like EI in SRS treatment planning QA. To effectively evaluate the parameter, it is necessary to choose a suitable machine learning algorithm. | ['Alexis Dimitriadis', 'Sheikh Othman', 'Hossam Donya'] | 2022-01-19 | null | null | null | null | ['parameter-prediction'] | ['miscellaneous'] | [ 5.86010069e-02 1.04219645e-01 -6.20373905e-01 -3.11725587e-01
-7.59832561e-01 -4.41098846e-02 2.37560764e-01 4.77146596e-01
-3.17083806e-01 1.11038303e+00 2.60820359e-01 -7.56976306e-01
-8.46539736e-01 -8.34543824e-01 8.20700377e-02 -1.05742037e+00
8.22879747e-02 8.14852118e-01 2.59254277e-01 -1.56304732e-01
2.98712373e-01 9.18697417e-01 -7.05972373e-01 3.97974908e-01
8.64712894e-01 1.19232678e+00 4.80925113e-01 6.91313207e-01
2.38778800e-01 1.44823897e+00 -2.13666543e-01 3.95277739e-01
-3.44771653e-01 -3.23483288e-01 -7.82881141e-01 -9.07792389e-01
-5.32647669e-01 -1.67834625e-01 -4.93810892e-01 3.04906100e-01
8.73615265e-01 -2.29476005e-01 1.15578735e+00 -1.02084303e+00
2.21969679e-01 3.61195564e-01 -3.89099926e-01 4.81560320e-01
3.43010932e-01 1.83943644e-01 2.89645463e-01 -5.23400307e-01
5.55274665e-01 3.87613505e-01 1.19007647e+00 4.20189470e-01
-8.92719865e-01 -6.68772936e-01 -7.60405719e-01 8.36415961e-02
-1.39559841e+00 1.36281893e-01 1.67830706e-01 -9.06848848e-01
1.37393904e+00 7.30165362e-01 7.63679802e-01 5.14555812e-01
1.40208793e+00 1.40573308e-01 1.55053198e+00 -4.70986873e-01
4.82970089e-01 2.32619047e-01 1.27743155e-01 8.35120797e-01
-1.05041169e-01 7.22375393e-01 -9.01680440e-03 -2.80916989e-01
6.71415865e-01 -9.73697677e-02 -2.72647560e-01 2.57239398e-02
-9.00840342e-01 1.25631320e+00 1.00225592e+00 4.38925385e-01
-4.89910275e-01 -2.33684853e-01 5.36394060e-01 3.47976200e-02
3.31729621e-01 1.98212743e-01 -6.02787375e-01 1.98411103e-02
-7.53020346e-01 9.75276455e-02 5.74682474e-01 6.43911600e-01
3.14131565e-02 -6.30252585e-02 -7.17737198e-01 8.72027576e-01
4.78424162e-01 2.91408688e-01 8.64831686e-01 -1.54135317e-01
1.15167752e-01 4.39181954e-01 -1.65457591e-01 -4.43016142e-01
-1.22302115e+00 -6.27641380e-01 -1.12008226e+00 2.40475312e-01
1.68876037e-01 1.79988295e-02 -1.11620939e+00 8.36745858e-01
2.52166688e-01 -6.54691219e-01 9.66104046e-02 6.35180295e-01
1.38487899e+00 4.93006319e-01 5.13176918e-01 -6.41409039e-01
9.65980351e-01 -9.21903253e-01 -7.67535508e-01 8.53953883e-02
1.31562197e+00 -7.64640272e-01 7.96024859e-01 2.35646918e-01
-8.50448132e-01 -2.31054604e-01 -8.95505846e-01 2.80097544e-01
8.78871083e-02 3.11388195e-01 9.24223781e-01 8.02619636e-01
-5.63316643e-01 1.05235803e+00 -9.39328253e-01 -2.34661832e-01
6.51787281e-01 7.59200752e-01 -1.67584717e-01 -1.35027021e-01
-8.60808134e-01 1.52041352e+00 4.02153820e-01 -1.30264461e-01
-8.07978034e-01 -9.51915622e-01 -4.92511272e-01 -1.54977351e-01
-3.70923936e-01 -7.87973881e-01 1.42439842e+00 -3.64593118e-01
-1.50974464e+00 4.71795291e-01 8.04947987e-02 -3.05423111e-01
5.41710377e-01 4.02276784e-01 -4.00439471e-01 -2.03984171e-01
3.79916877e-02 6.85003102e-02 4.24885750e-02 -6.68080449e-01
-3.95031869e-01 -6.21438861e-01 -5.74771941e-01 1.73771143e-01
3.63702595e-01 1.69671014e-01 5.06722689e-01 -1.12047873e-01
5.38236320e-01 -7.61274576e-01 -5.47840357e-01 -1.95638105e-01
-2.89733529e-01 -2.65543997e-01 4.93729144e-01 -9.58381176e-01
1.57747054e+00 -1.54807639e+00 -5.63134730e-01 5.76467514e-01
-2.43096636e-03 -1.12515576e-01 5.53810775e-01 4.19780582e-01
-4.85844612e-01 -1.29365027e-01 -2.53122926e-01 8.18634212e-01
-6.26148760e-01 -2.46226136e-02 4.72560614e-01 6.72086418e-01
-5.23540676e-01 1.13493645e+00 -6.35435045e-01 -5.21478653e-01
4.75177675e-01 2.07327560e-01 -1.37777507e-01 1.80504322e-01
1.85264736e-01 5.79492450e-01 -4.34847891e-01 7.39216030e-01
5.89797497e-01 -2.58373946e-01 -1.77575991e-01 -5.39571643e-01
-2.03533381e-01 1.53792188e-01 -5.34253061e-01 1.19166100e+00
-5.63141882e-01 4.61113006e-01 -5.60143530e-01 -6.41049981e-01
1.19494784e+00 2.97317863e-01 7.51003623e-01 -9.45099175e-01
4.06775683e-01 2.52941072e-01 3.52529436e-01 -6.42702103e-01
-1.77971229e-01 -9.13923442e-01 3.57656568e-01 4.00460064e-02
-2.08985522e-01 -3.42814147e-01 -7.57008791e-01 -1.16511479e-01
1.29581356e+00 -3.97133619e-01 8.95553052e-01 -4.70556349e-01
4.09872442e-01 4.67757583e-01 3.13288085e-02 4.54638898e-01
-7.74453357e-02 4.24848437e-01 2.37322927e-01 -5.73849082e-01
-8.72631729e-01 -8.07888687e-01 -9.09385145e-01 4.05431032e-01
-1.55248523e-01 -2.14601800e-01 -4.13010657e-01 -5.37164390e-01
-1.32064283e-01 1.11326313e+00 -4.28350687e-01 -1.34668157e-01
-2.17243910e-01 -1.22092247e+00 1.17838785e-01 7.70972729e-01
2.69133687e-01 -9.75749135e-01 -4.07391727e-01 2.88716167e-01
1.47948526e-02 -7.65113711e-01 2.25460440e-01 4.60632414e-01
-1.39671087e+00 -1.18467116e+00 -3.93055916e-01 -2.97219753e-01
5.72445214e-01 -2.83340544e-01 9.44589734e-01 1.79848999e-01
-5.09210527e-01 1.03623651e-01 -2.02157989e-01 -4.08751428e-01
-7.83813655e-01 -1.26427248e-01 -1.72503546e-01 -1.05931735e+00
6.10652007e-02 -2.71931648e-01 -6.98420703e-01 7.47815430e-01
-5.18189907e-01 3.54156196e-01 6.85554624e-01 1.10667384e+00
9.95961130e-01 3.32973838e-01 3.64846438e-01 -9.48585629e-01
3.71394038e-01 -5.43283045e-01 -3.37027788e-01 2.10023463e-01
-1.22958803e+00 -3.16563755e-01 4.72184181e-01 -9.25135314e-02
-8.65652263e-01 9.29530114e-02 -5.29599488e-01 1.51134610e-01
3.42063643e-02 7.07059979e-01 9.21175331e-02 -5.29468060e-01
1.16707325e+00 -5.75438775e-02 -5.78098185e-02 -3.55283618e-02
-3.38955522e-01 1.13933659e+00 -1.14698164e-01 6.41516817e-05
3.07320774e-01 1.34430632e-01 3.59745055e-01 -5.93256414e-01
-6.33970678e-01 -2.92865068e-01 -5.07360041e-01 -4.76883680e-01
9.54538345e-01 -7.63654828e-01 -7.31290758e-01 4.60596949e-01
-6.58265769e-01 -5.93150556e-01 -1.95153415e-01 9.48363185e-01
-8.09025645e-01 -2.34855469e-02 -6.45275533e-01 -6.95460141e-01
-1.06315732e+00 -9.95982826e-01 4.09607977e-01 3.91824663e-01
-5.87586820e-01 -1.06818521e+00 9.35858041e-02 5.49937367e-01
7.68914461e-01 5.97175896e-01 1.18778265e+00 -8.11073244e-01
-5.17334193e-02 -5.79548419e-01 -1.66479006e-01 -2.87871831e-03
1.13507211e-01 3.89328264e-02 -8.39774966e-01 -3.58043909e-01
2.13000461e-01 -1.37843803e-01 3.00138980e-01 1.09047592e+00
1.40608203e+00 -1.14961781e-01 -8.84705782e-01 6.66322827e-01
1.73602414e+00 9.49834466e-01 9.37200367e-01 5.29514849e-01
4.00011927e-01 1.76659618e-02 8.09870005e-01 5.20312786e-01
1.02289446e-01 6.63126171e-01 3.59619290e-01 -2.11154655e-01
8.96608680e-02 -3.25257599e-01 -2.66772538e-01 5.16394079e-01
-3.67104530e-01 6.00687191e-02 -1.39307213e+00 -2.32332408e-01
-1.40321505e+00 -7.06850290e-01 -7.72929072e-01 2.12003565e+00
2.96112716e-01 3.22314888e-01 -1.57474577e-01 3.63197565e-01
3.37373316e-01 -3.97433639e-01 -5.23769796e-01 -5.60448885e-01
1.59242317e-01 5.50449312e-01 8.49283695e-01 3.88524801e-01
-7.71265209e-01 5.48980236e-01 6.67385530e+00 1.19774282e+00
-1.21943712e+00 1.58776641e-01 8.20296407e-01 1.09145910e-01
-4.82098348e-02 9.62296724e-02 -7.07945108e-01 4.12835091e-01
1.27747452e+00 -1.91186845e-01 6.00948073e-02 8.21787477e-01
4.71971482e-01 -6.83785260e-01 -8.13841045e-01 1.01081729e+00
-4.91218686e-01 -1.42526460e+00 -5.03999889e-01 1.85890198e-01
4.40165669e-01 2.55974203e-01 -4.37204331e-01 4.65060860e-01
1.41990781e-01 -1.60913229e+00 1.82675698e-03 6.88189089e-01
8.30522418e-01 -6.28984332e-01 1.24652278e+00 6.98448598e-01
-7.27561235e-01 -2.63062507e-01 -4.85398769e-01 -1.52776456e-02
-8.51365086e-03 7.53997922e-01 -1.34796286e+00 7.58862555e-01
6.41496420e-01 3.84074479e-01 -3.47490758e-01 1.26471496e+00
5.69368415e-02 6.25746429e-01 -4.14905101e-01 -1.77327782e-01
-6.83065727e-02 1.80020422e-01 -4.61361147e-02 7.65634418e-01
4.40558016e-01 7.10222661e-01 -3.41272026e-01 2.59532660e-01
7.72326052e-01 4.61042583e-01 -2.40992621e-01 3.18108976e-01
3.61117840e-01 1.10365295e+00 -8.49071860e-01 -1.12909684e-03
-1.72245249e-01 4.15989518e-01 -1.13448240e-01 -2.64667660e-01
-1.17201996e+00 1.42106518e-01 -2.08493143e-01 6.25558853e-01
-2.30695024e-01 2.92700142e-01 -7.68041134e-01 -3.03076774e-01
-5.44729650e-01 -4.09689188e-01 8.46717358e-01 -1.26718557e+00
-1.22423434e+00 9.26782191e-01 7.75487348e-02 -1.41310060e+00
1.40358496e-03 -6.75349891e-01 -8.40511143e-01 1.03362095e+00
-1.08720696e+00 -8.57154250e-01 -6.53853476e-01 5.31370163e-01
2.17847124e-01 -3.76586378e-01 1.12346101e+00 -1.08201288e-01
-4.06215668e-01 4.29710954e-01 3.33624929e-01 -2.61841178e-01
1.15742832e-01 -8.77002060e-01 -4.51025844e-01 -3.19789976e-01
-5.95448375e-01 -1.76695362e-01 5.29081762e-01 -7.37327933e-01
-1.08205187e+00 -7.98510075e-01 6.10061169e-01 -1.69495568e-01
3.11966866e-01 3.96553189e-01 -6.09237492e-01 3.99761885e-01
-5.10497354e-02 1.07522756e-01 1.06055558e+00 -2.35396042e-01
5.08486092e-01 -1.36943161e-01 -1.62156534e+00 -1.20668307e-01
4.38124508e-01 -1.04572296e-01 8.74140952e-03 9.53693509e-01
1.98453546e-01 -9.49135244e-01 -1.72257435e+00 1.10721183e+00
5.73820472e-01 -9.78834927e-01 1.01107919e+00 -6.34380698e-01
5.56029499e-01 -7.59690851e-02 -7.62449726e-02 -1.07881880e+00
-5.92952967e-01 2.05355167e-01 1.53219298e-01 6.07445955e-01
8.06824565e-01 -7.13907361e-01 9.58785534e-01 9.86413717e-01
-2.85980523e-01 -1.59849417e+00 -1.34815824e+00 -7.05896974e-01
3.65902871e-01 -4.73273367e-01 3.46990168e-01 8.12171698e-01
6.76534548e-02 2.85130739e-01 1.23067088e-01 -5.91070950e-02
1.80676132e-01 -2.28347227e-01 2.71999270e-01 -1.05785418e+00
-2.14832246e-01 -2.28351206e-01 -5.99945962e-01 -3.85880381e-01
-4.44950461e-01 -1.08151186e+00 -3.63682687e-01 -2.10293794e+00
3.91683340e-01 -9.73299801e-01 -3.50283504e-01 5.20158947e-01
2.14460686e-01 -3.44979465e-01 -1.78922877e-01 4.61827368e-01
3.20940733e-01 5.90857267e-01 1.38053966e+00 -1.50735676e-01
-4.05759722e-01 7.14308381e-01 -2.93176770e-01 6.18795514e-01
9.98218834e-01 -8.40088665e-01 -2.82564521e-01 1.97716206e-01
-2.92512849e-02 7.99387813e-01 1.27053782e-01 -1.12332988e+00
1.60826474e-01 -2.52477080e-01 9.73322451e-01 -8.70766521e-01
-8.65591317e-02 -9.71953332e-01 7.87853360e-01 9.25455093e-01
2.14337129e-02 -1.31887645e-01 3.49416971e-01 2.18900554e-02
-2.99756303e-02 -5.42514086e-01 1.20301557e+00 8.29877630e-02
-2.67157495e-01 5.24350405e-01 -2.43674189e-01 -5.83113968e-01
1.39174056e+00 -6.29056573e-01 -1.89901710e-01 -1.62594557e-01
-1.02710199e+00 -1.04678959e-01 4.07599956e-02 -3.04130018e-01
6.18849277e-01 -1.38909996e+00 -4.93432164e-01 -6.64099529e-02
1.60756543e-01 1.56821869e-02 7.81707942e-01 1.42950141e+00
-1.04243922e+00 4.24554795e-01 -2.17869610e-01 -7.02699721e-01
-1.37696362e+00 3.39261025e-01 8.35844874e-01 -7.46451437e-01
-6.15375876e-01 7.09868014e-01 -2.21831560e-01 -4.67328489e-01
-2.75037229e-01 -3.60223323e-01 -5.28992891e-01 -3.41730863e-01
2.86162287e-01 6.28938854e-01 5.48183501e-01 -4.84433860e-01
-5.48778713e-01 7.32729316e-01 1.77700594e-02 3.72269392e-01
1.47608769e+00 3.26984107e-01 -9.85986670e-04 1.00811301e-02
1.18590057e+00 -4.19268638e-01 -5.58602631e-01 2.01405883e-01
-3.46440151e-02 -2.57141262e-01 5.78323066e-01 -1.45628917e+00
-8.41827691e-01 5.04792273e-01 1.29758680e+00 -1.80478096e-01
1.43199515e+00 1.33793905e-01 3.94996554e-01 -2.17914239e-01
3.04794580e-01 -8.50333571e-01 -3.96807641e-01 2.89422095e-01
1.01428103e+00 -1.35320234e+00 6.37249351e-01 -6.67566597e-01
-8.95633578e-01 1.38879550e+00 6.54197395e-01 1.95146203e-01
1.11450398e+00 5.60839593e-01 -6.36302680e-03 -2.44400069e-01
-6.43697321e-01 3.33722383e-01 1.34255797e-01 4.72369254e-01
6.31745696e-01 2.30914161e-01 -7.30255663e-01 9.24786985e-01
-5.76486170e-01 5.17127693e-01 1.80181473e-01 8.72830749e-01
-5.95804930e-01 -6.97467744e-01 -1.86578989e-01 9.09502447e-01
-1.06377468e-01 4.31695804e-02 1.45353392e-01 1.17873752e+00
-2.35232919e-01 6.87457383e-01 -2.25862980e-01 -5.04647970e-01
6.61262929e-01 -1.79918677e-01 5.83296359e-01 -1.88750029e-01
-7.28687048e-01 -2.37956643e-01 6.48460761e-02 -3.93147469e-01
-1.19420663e-01 -2.30014265e-01 -1.37807822e+00 -2.54743159e-01
-7.17184484e-01 7.00245574e-02 9.49244201e-01 1.04067969e+00
-3.85029942e-01 5.90834677e-01 8.52740169e-01 -3.03232551e-01
-2.87554324e-01 -1.33359075e+00 -6.67953670e-01 -3.02902132e-01
-3.01637024e-01 -6.90199852e-01 -3.82050574e-01 -6.61831319e-01] | [15.157181739807129, -2.4612159729003906] |
54093628-6275-42b4-9e26-25942aa3ad46 | constant-approximation-for-normalized | 2212.14334 | null | https://arxiv.org/abs/2212.14334v1 | https://arxiv.org/pdf/2212.14334v1.pdf | Constant Approximation for Normalized Modularity and Associations Clustering | We study the problem of graph clustering under a broad class of objectives in which the quality of a cluster is defined based on the ratio between the number of edges in the cluster, and the total weight of vertices in the cluster. We show that our definition is closely related to popular clustering measures, namely normalized associations, which is a dual of the normalized cut objective, and normalized modularity. We give a linear time constant-approximate algorithm for our objective, which implies the first constant-factor approximation algorithms for normalized modularity and normalized associations. | ['Christian Sohler', 'Vahab Mirrokni', 'Jakub Łącki'] | 2022-12-29 | null | null | null | null | ['graph-clustering'] | ['graphs'] | [-1.35285869e-01 3.39213997e-01 -3.31597030e-01 -2.66040377e-02
3.63721848e-02 -7.10379422e-01 4.98702936e-02 6.74533784e-01
-3.79715890e-01 2.27588177e-01 -1.07990243e-02 -1.05086520e-01
-8.02248061e-01 -9.96433735e-01 -3.99398655e-01 -6.91562951e-01
-7.26366818e-01 6.76881731e-01 3.82982850e-01 5.13493493e-02
3.16838562e-01 4.58081067e-01 -1.06792545e+00 -2.35554129e-01
7.72628665e-01 5.29893279e-01 -1.32742867e-01 8.40894461e-01
1.46075770e-01 1.81360260e-01 -4.34247643e-01 -3.47858757e-01
1.78297058e-01 -6.02994263e-01 -1.19259322e+00 5.02080679e-01
-3.05670172e-01 5.75660825e-01 -3.24867666e-01 1.29241002e+00
5.89902811e-02 9.96113569e-02 7.53783166e-01 -1.80652714e+00
-3.18305314e-01 8.59959126e-01 -1.00432193e+00 2.08440721e-01
3.96758378e-01 -3.50622803e-01 1.44809127e+00 -2.93092787e-01
6.20133758e-01 1.22934961e+00 4.91758555e-01 2.80307513e-02
-1.68409169e+00 -1.85497314e-01 9.49482843e-02 -2.07158029e-01
-1.47169054e+00 -9.18783918e-02 2.29257166e-01 -5.59042335e-01
6.77386105e-01 3.33049744e-01 5.60826004e-01 -2.23209292e-01
-1.02590419e-01 2.79705435e-01 6.94508195e-01 -4.61627364e-01
2.15121135e-01 -1.47821099e-01 4.42949176e-01 7.35781670e-01
7.75923669e-01 -2.44834721e-01 -1.87494263e-01 -4.18730974e-01
6.49891078e-01 -1.95284277e-01 -3.06038111e-01 -8.72667611e-01
-1.31386268e+00 9.39403355e-01 5.67734957e-01 3.99267226e-01
-9.80838435e-04 3.42448205e-01 1.18078001e-01 3.31305742e-01
2.15244114e-01 4.11720067e-01 -3.18578452e-01 2.45652705e-01
-8.57454121e-01 -1.57772034e-01 1.07054698e+00 8.92302155e-01
9.83002722e-01 -5.01468241e-01 3.91556114e-01 5.06056547e-01
4.13918674e-01 2.62291223e-01 -1.78207368e-01 -1.30768836e+00
1.34022877e-01 1.04830480e+00 -2.55400211e-01 -1.25439656e+00
-5.14761209e-01 -4.79735762e-01 -9.20881391e-01 -5.65528348e-02
3.11223060e-01 8.28870982e-02 -3.88182491e-01 2.02143312e+00
2.76601523e-01 -1.55662790e-01 -3.24797720e-01 6.13387764e-01
2.78079271e-01 3.77123624e-01 -3.54969531e-01 -6.73301756e-01
8.16353500e-01 -9.00183618e-01 -4.98185992e-01 1.50483638e-01
6.26284778e-01 -7.45528102e-01 5.49605191e-01 -7.04653910e-04
-1.23757076e+00 -5.14484644e-02 -9.44661915e-01 5.20538807e-01
-1.89040780e-01 -5.98798513e-01 5.65727353e-01 6.60248935e-01
-1.43286335e+00 6.68770790e-01 -6.05411470e-01 -6.50830448e-01
-1.33248270e-01 5.32930553e-01 -4.20243055e-01 2.60125808e-02
-3.56811702e-01 2.92574227e-01 4.87464696e-01 -4.58648384e-01
-3.77464324e-01 -3.87504071e-01 -6.04349017e-01 4.47640181e-01
3.77551198e-01 -6.73125386e-01 6.57095671e-01 -1.08093429e+00
-6.43954813e-01 1.00503457e+00 8.43161792e-02 -1.35103032e-01
1.18885763e-01 4.75719512e-01 -2.37842798e-01 2.48786569e-01
4.69693273e-01 4.61844057e-01 2.63896316e-01 -1.18865180e+00
-6.16614223e-01 -3.73713136e-01 3.38749111e-01 4.43651304e-02
-3.71683687e-01 4.28695418e-02 -6.29311740e-01 -3.78581315e-01
4.00484055e-01 -9.08717811e-01 -6.25649571e-01 3.49583589e-02
-4.99242306e-01 -3.06327879e-01 3.91722113e-01 -1.42504811e-01
1.52303481e+00 -2.26275253e+00 4.77836967e-01 9.03228045e-01
7.83205092e-01 -4.18580174e-01 -2.24903032e-01 5.14104128e-01
-2.94113815e-01 4.65482801e-01 -4.68800545e-01 3.85264486e-01
1.04454041e-01 2.20352843e-01 4.19978797e-01 6.92330778e-01
5.81346042e-02 3.58462036e-01 -1.14992893e+00 -5.66625416e-01
-1.79771811e-01 -1.34602442e-01 -6.68141365e-01 -1.91769004e-02
2.10821345e-01 -2.75343567e-01 -9.19094533e-02 7.70308748e-02
6.82104349e-01 -5.70202291e-01 7.66860843e-01 1.15515038e-01
-8.16483200e-02 -1.50881254e-03 -1.57161367e+00 1.18492103e+00
5.69188111e-02 3.97271782e-01 5.92913926e-01 -1.04467702e+00
4.79484975e-01 1.83350861e-01 1.03080332e+00 -1.00062065e-01
1.37348607e-01 -7.55376965e-02 1.17722243e-01 -1.28498971e-01
2.32002720e-01 -7.16774836e-02 -2.35296056e-01 9.24006164e-01
-1.15263179e-01 3.31489414e-01 8.63508999e-01 1.03383410e+00
1.59857309e+00 -5.40795445e-01 4.85260576e-01 -8.29769313e-01
1.94068998e-01 -5.42980283e-02 5.09103179e-01 4.99366164e-01
-2.94444323e-01 1.28497824e-01 1.18805957e+00 6.36667088e-02
-1.11846972e+00 -1.39208925e+00 1.09325230e-01 1.05967844e+00
4.71660078e-01 -8.64335299e-01 -9.53942060e-01 -4.85765129e-01
4.77212816e-02 -1.99786201e-01 -5.79970121e-01 -1.20766088e-01
-1.13161765e-01 -7.63314784e-01 2.03066450e-02 3.00290257e-01
-1.04833938e-01 -4.30220485e-01 -2.67190546e-01 1.74395517e-01
-2.88721263e-01 -7.69796491e-01 -9.80330348e-01 6.31648302e-02
-1.05740082e+00 -1.71596503e+00 -2.11901620e-01 -1.01652730e+00
1.00836110e+00 6.61006510e-01 1.18049014e+00 5.52637815e-01
-2.98821896e-01 4.56011742e-01 -4.34955597e-01 3.48654777e-01
-2.15783134e-01 6.56830221e-02 9.41697583e-02 -4.82824445e-02
-1.59258377e-02 -7.44473636e-01 -3.30657989e-01 4.40160096e-01
-1.05760717e+00 -2.33388886e-01 2.79893756e-01 4.34055805e-01
5.22880673e-01 5.22932470e-01 3.94618243e-01 -7.72821248e-01
6.84895873e-01 -6.35307372e-01 -4.69212085e-01 2.39445567e-01
-7.18794227e-01 3.39973241e-01 2.48609841e-01 -2.33209088e-01
-9.80561748e-02 2.55349636e-01 3.44870955e-01 1.04886763e-01
4.05598164e-01 4.57846016e-01 -2.53209502e-01 -1.31919324e-01
2.43865848e-01 -3.72751266e-01 -1.71302736e-01 -2.68838614e-01
6.47777975e-01 2.47048631e-01 4.60788131e-01 -4.47944582e-01
8.81500244e-01 5.96570969e-01 3.67427289e-01 -6.22583449e-01
-4.90388393e-01 -8.78949046e-01 -8.39255154e-01 -3.16664994e-01
7.19020724e-01 -4.59202349e-01 -1.06681335e+00 -2.65387446e-01
-7.96578228e-01 9.57506374e-02 -7.13119805e-02 2.67286211e-01
-4.97128874e-01 7.00150967e-01 -5.73051751e-01 -7.42472351e-01
2.96239182e-02 -7.14160979e-01 3.82144541e-01 -5.88672720e-02
-6.77161515e-02 -9.58263814e-01 4.07775670e-01 -7.27924109e-02
4.12046053e-02 6.34388745e-01 1.33258772e+00 -3.46751183e-01
-4.77413177e-01 -1.53990507e-01 -4.67963576e-01 5.30685335e-02
7.02353939e-02 2.91077912e-01 -5.58617041e-02 -5.80309868e-01
-3.70195329e-01 2.10861817e-01 8.48732650e-01 3.96338463e-01
6.09696269e-01 -5.82569897e-01 -6.56681955e-01 4.19836462e-01
1.85319424e+00 -1.01837665e-01 2.82662600e-01 -2.33109873e-02
5.47521710e-01 4.97862637e-01 2.65555024e-01 5.01900375e-01
2.32160628e-01 4.71947223e-01 3.91282529e-01 -8.29994120e-03
2.41913483e-01 2.58680403e-01 4.62322682e-02 1.33029854e+00
-1.35121524e-01 -2.95842916e-01 -7.86378920e-01 8.18134844e-01
-2.11953354e+00 -8.64623427e-01 -5.10443449e-01 2.38318777e+00
3.60571951e-01 2.81723261e-01 8.10042024e-01 5.04304707e-01
1.30397594e+00 -1.44286528e-01 -1.10830851e-01 -5.56941211e-01
-7.95611814e-02 2.46844009e-01 6.68712854e-01 7.59098232e-01
-9.09032106e-01 4.85106856e-01 8.05864048e+00 6.16476834e-01
-2.21546933e-01 1.55088529e-01 3.35888594e-01 -1.67395443e-01
-1.64229199e-01 2.24293008e-01 1.28944829e-01 3.23330611e-01
8.52826357e-01 -7.52335012e-01 7.25611627e-01 7.27054060e-01
1.67013574e-02 -1.20462533e-02 -1.14831460e+00 4.73035932e-01
-2.57841825e-01 -8.45024884e-01 -2.64786005e-01 6.46754444e-01
9.56622660e-01 -1.78000838e-01 -2.14046657e-01 -4.49891925e-01
6.87672853e-01 -6.83888972e-01 4.62742954e-01 1.98612735e-01
8.13872278e-01 -1.27762556e+00 5.08957207e-01 3.81812900e-02
-1.70662272e+00 -6.33569434e-02 -3.40561211e-01 2.77212681e-03
9.54422206e-02 9.80223238e-01 -3.80985796e-01 6.06490374e-01
7.54406393e-01 4.50999111e-01 -2.66151786e-01 1.47781169e+00
9.73599777e-02 2.98982888e-01 -4.30411041e-01 1.54063880e-01
5.86916842e-02 -5.19629419e-01 5.56132197e-01 1.25361514e+00
7.08049256e-03 1.97122380e-01 4.58705008e-01 6.11163259e-01
-4.02388871e-01 8.97889882e-02 -4.07950938e-01 -8.38753060e-02
6.06812537e-01 1.43132389e+00 -1.40798926e+00 -1.10713884e-01
-2.52995580e-01 7.43328393e-01 4.90629137e-01 -6.80371076e-02
-6.06241524e-01 -7.62275398e-01 8.15328836e-01 2.54526705e-01
3.96249652e-01 -4.78271514e-01 -1.80354998e-01 -6.64587259e-01
-1.07443638e-01 -4.89488095e-01 7.50884950e-01 -3.15790266e-01
-1.32949829e+00 2.77774572e-01 -2.77327001e-01 -7.84031153e-01
3.69304180e-01 -4.43446934e-01 -6.53770089e-01 2.88679302e-01
-7.99606562e-01 -2.96318829e-01 3.22038829e-02 5.53134084e-01
-2.13579878e-01 4.63807940e-01 6.22637868e-01 3.18863094e-01
-5.99904358e-01 2.99991816e-01 4.07094926e-01 1.38675019e-01
4.24681932e-01 -1.62694132e+00 -4.38762195e-02 1.10271847e+00
-1.49092808e-01 5.15954196e-01 6.49908006e-01 -4.86814469e-01
-1.26090252e+00 -9.46889579e-01 1.01583278e+00 -5.20718135e-02
8.30259860e-01 -3.91264483e-02 -3.91126335e-01 5.58222294e-01
1.94629282e-01 -2.76853532e-01 7.78395593e-01 3.11279863e-01
-4.35255289e-01 -6.59080520e-02 -1.19438219e+00 3.34962994e-01
1.33175170e+00 -1.82563305e-01 -1.64982900e-01 4.44537580e-01
1.02138257e+00 4.30095434e-01 -1.17249274e+00 2.20624179e-01
3.40000689e-01 -7.80820906e-01 8.15654516e-01 -5.71004808e-01
-7.58270472e-02 -4.02228981e-01 -2.06777483e-01 -1.21402168e+00
-1.16662097e+00 -4.78787631e-01 7.59197026e-02 9.79415238e-01
5.53665042e-01 -3.84706944e-01 8.41239631e-01 5.94000444e-02
2.49269903e-01 -7.38608360e-01 -9.10967231e-01 -9.22300518e-01
-2.19282776e-01 1.23906218e-01 5.79335093e-01 1.18731356e+00
7.66297400e-01 7.29377091e-01 1.09738700e-01 2.01722682e-01
1.02057040e+00 1.03500918e-01 2.91318297e-01 -1.67951632e+00
-3.29780519e-01 -8.56560767e-01 -7.79547811e-01 -5.63710213e-01
-3.02570686e-02 -1.16922641e+00 -1.76731169e-01 -1.88865423e+00
8.71722937e-01 -4.04268533e-01 -3.07237536e-01 1.91584364e-01
-1.25649527e-01 2.66944677e-01 1.74546719e-01 1.65790990e-01
-1.15005231e+00 5.34760393e-03 8.19281518e-01 -1.84778497e-02
-1.99603423e-01 -1.92197412e-02 -8.07740688e-01 5.97387135e-01
6.96121514e-01 -7.73225307e-01 -9.24078077e-02 -1.64807215e-01
5.89854658e-01 9.25402567e-02 -1.27680421e-01 -9.05155241e-01
3.61118793e-01 -3.78879786e-01 -8.97047594e-02 -5.02674401e-01
-7.75356293e-02 -8.64957452e-01 3.72035503e-01 8.53255570e-01
-3.56768996e-01 6.54332101e-01 -4.27499354e-01 7.23902464e-01
2.55009755e-02 -1.61902592e-01 9.79747474e-01 -1.23456590e-01
-1.08891152e-01 3.35744262e-01 -4.78318244e-01 2.53578365e-01
1.18328738e+00 -3.58458698e-01 -3.68649125e-01 -3.48390818e-01
-8.13138306e-01 5.41608930e-01 7.88985610e-01 -4.05061804e-02
4.80419129e-01 -1.37950778e+00 -7.24639893e-01 -4.12892073e-01
4.11764570e-02 -4.27528024e-01 -2.49306977e-01 9.66916442e-01
-5.54766119e-01 3.02619368e-01 -1.10721312e-01 -2.79689729e-01
-1.50182247e+00 1.18094599e+00 2.81347215e-01 -2.92824000e-01
-7.57066086e-02 5.35431266e-01 3.71098459e-01 -2.28940591e-01
2.05213964e-01 2.53333718e-01 2.56633729e-01 -1.85074314e-01
2.42220998e-01 9.62532520e-01 -2.69959003e-01 -6.93171144e-01
-5.72298825e-01 5.28647780e-01 3.28275919e-01 -2.21586689e-01
1.27173150e+00 -2.12226972e-01 -1.00042892e+00 1.98906958e-01
1.02480888e+00 -1.15512475e-01 -4.90814418e-01 -2.02596545e-01
3.77567798e-01 -5.31414330e-01 -2.39700049e-01 -3.85014892e-01
-1.24699450e+00 3.44293088e-01 3.59693795e-01 7.90918529e-01
1.38436520e+00 4.81423140e-01 3.42234910e-01 3.31550628e-01
3.64468664e-01 -1.14854014e+00 2.45064422e-01 2.38605961e-01
2.29985669e-01 -6.47295713e-01 1.37004092e-01 -9.03421700e-01
-1.17961064e-01 8.27437639e-01 4.04377848e-01 -2.69728005e-01
9.73784804e-01 3.03980261e-01 -5.51929533e-01 -3.22424859e-01
-6.66932464e-01 -6.04128957e-01 -1.53250396e-02 6.02991104e-01
5.52143633e-01 6.04827046e-01 -7.10173726e-01 3.38084817e-01
-1.26451105e-01 -6.04889512e-01 5.87698162e-01 6.94647074e-01
-1.13043511e+00 -1.10625696e+00 -2.92949200e-01 6.40137136e-01
-2.88513988e-01 5.67312278e-02 -9.05458987e-01 4.18683916e-01
1.80372760e-01 1.41322625e+00 2.29049295e-01 -4.27971840e-01
2.24426910e-01 -3.18339467e-01 6.34225428e-01 -6.32618070e-01
-3.95652980e-01 -1.83123425e-02 -7.59996623e-02 -5.30837297e-01
-6.56007588e-01 -2.95464396e-01 -1.22308743e+00 -1.06098640e+00
-8.79543722e-01 5.67754865e-01 3.07691187e-01 6.60018206e-01
1.60055533e-01 3.88815910e-01 9.90658939e-01 -3.19542855e-01
8.63319710e-02 -6.78429723e-01 -1.05742824e+00 3.37447464e-01
4.52820696e-02 -3.35737497e-01 -6.03411674e-01 -9.25796106e-02] | [6.962836265563965, 5.25604248046875] |
11f51e6c-2f77-407a-81a5-9a41755757f4 | nonblind-image-deconvolution-via-leveraging | null | null | https://link.springer.com/article/10.1007/s11263-022-01621-9 | https://link.springer.com/article/10.1007/s11263-022-01621-9 | Nonblind image deconvolution via leveraging model uncertainty in an untrained deep neural network | Nonblind image deconvolution (NID) is about restoring the latent image with sharp details from
a noisy blurred one using a known blur kernel. This paper presents a dataset-free deep learning
approach for NID using untrained deep neural networks (DNNs), which does not require any external
training data with ground-truth images. Based on a spatially-adaptive dropout scheme, the proposed
approach learns a DNN with model uncertainty from the input blurred image, and the deconvolution result is obtained by aggregating the multiple predictions from the learned dropout DNN.
It is shown that the solution approximates a minimum-mean-squared-error estimator in Bayesian
inference. In addition, a self-supervised loss function for training is presented to efficiently handle
the noise in blurred images. Extensive experiments show that the proposed approach not only performs noticeably better than existing non-learning-based methods and unsupervised learning-based
methods, but also performs competitively against recent supervised learning-based methods. | ['Mingqin Chen; Yuhui Quan; Tongyao Pang; Hui Ji'] | 2022-05-18 | null | null | null | international-journal-of-computer-vision-2022 | ['image-deconvolution'] | ['computer-vision'] | [ 7.00189620e-02 9.28552896e-02 1.18042901e-01 -7.43419647e-01
-9.50505495e-01 -1.99495882e-01 3.76281261e-01 -8.72599781e-01
-4.24952507e-01 1.16912019e+00 4.33769107e-01 6.26167804e-02
-3.36889058e-01 -1.86603814e-01 -1.05947673e+00 -1.06281102e+00
3.58581245e-01 3.00922960e-01 -1.05216034e-01 5.52024126e-01
2.07590446e-01 2.64864922e-01 -1.12694335e+00 7.64013827e-02
1.25728643e+00 1.05320728e+00 5.75995862e-01 6.10346854e-01
2.19368249e-01 1.48315084e+00 -4.86209452e-01 -2.85130382e-01
2.88612485e-01 -5.32762349e-01 -7.88560867e-01 4.17005211e-01
8.48136961e-01 -1.04065454e+00 -1.03878057e+00 1.67109025e+00
3.50203544e-01 1.18656352e-01 5.94564676e-01 -6.90218449e-01
-1.35860693e+00 4.00624841e-01 -4.45715278e-01 1.08530596e-01
-2.91873962e-01 3.44776481e-01 5.16081512e-01 -1.21401918e+00
4.04206485e-01 1.23525393e+00 6.96921706e-01 6.61384940e-01
-1.39195633e+00 -4.05981839e-01 -1.93989277e-01 2.27102935e-01
-1.19208276e+00 -5.72868526e-01 6.42622232e-01 -4.40723717e-01
3.49323243e-01 -2.99680941e-02 8.04799721e-02 1.17056882e+00
7.16700032e-02 7.24075139e-01 1.53834569e+00 -3.13280314e-01
4.76804554e-01 1.75822765e-01 1.26671299e-01 7.36706436e-01
4.30805266e-01 3.73799384e-01 -5.47370970e-01 -1.57939926e-01
1.18896043e+00 2.99911737e-01 -8.90217185e-01 -4.24402148e-01
-9.45974290e-01 5.44596612e-01 8.18936288e-01 -4.53355629e-03
-5.75045228e-01 3.55322510e-01 -1.81322217e-01 1.17982544e-01
7.43692160e-01 2.22877026e-01 -4.31270689e-01 -2.26410513e-04
-1.35617638e+00 -3.59091064e-04 7.10027933e-01 8.36108506e-01
9.74218011e-01 4.68310088e-01 -3.60277593e-01 8.38544071e-01
4.16790485e-01 6.08523667e-01 3.84283453e-01 -1.46661651e+00
1.63510397e-01 1.87691189e-02 5.86052597e-01 -3.41476351e-01
9.37089548e-02 -4.02873456e-01 -1.07666755e+00 7.31300294e-01
5.07213354e-01 -4.08206612e-01 -1.30383301e+00 1.78841591e+00
-1.94502085e-01 6.75721705e-01 2.99539834e-01 1.47698081e+00
6.79861724e-01 3.93877923e-01 -3.68736893e-01 -2.53706336e-01
8.33490670e-01 -1.08523917e+00 -1.01499593e+00 -3.87833655e-01
-3.59613329e-01 -7.42751062e-01 8.69866788e-01 5.81378222e-01
-1.16325533e+00 -6.64605141e-01 -1.03212357e+00 -3.34158331e-01
2.20789909e-01 3.50511491e-01 3.20456356e-01 5.07852256e-01
-1.25101221e+00 8.34727407e-01 -8.24324250e-01 -5.94179630e-02
8.29428673e-01 3.08627605e-01 -2.65802205e-01 -5.05886137e-01
-9.23020363e-01 9.69331205e-01 1.92551002e-01 4.90181327e-01
-1.54897749e+00 -8.22135031e-01 -8.04239273e-01 7.37382397e-02
1.37269154e-01 -8.25679481e-01 1.42409956e+00 -1.27893424e+00
-1.75413084e+00 6.08452916e-01 -3.20212781e-01 -7.20297098e-01
6.96053267e-01 -8.03107560e-01 9.83425975e-03 4.15585130e-01
-1.37017310e-01 5.54735661e-01 1.45384860e+00 -1.67383778e+00
-1.89044192e-01 -2.85683870e-01 -4.94269617e-02 1.21168226e-01
-1.17016152e-01 -1.44516870e-01 -1.78456798e-01 -6.99957311e-01
1.01956993e-01 -5.64115644e-01 -1.33146793e-01 2.69759148e-01
-4.24498200e-01 3.96505237e-01 9.14094746e-01 -9.61912036e-01
5.14355540e-01 -2.19791532e+00 1.36616185e-01 -3.70546937e-01
5.15598059e-01 3.05016935e-01 8.96581635e-03 -7.59053603e-02
-8.89652781e-03 -4.64787841e-01 -7.37498224e-01 -7.29478538e-01
1.27656125e-02 4.56462651e-01 -6.28062665e-01 6.80306733e-01
2.08918735e-01 7.92576909e-01 -1.02723908e+00 -9.24545899e-02
3.44668865e-01 8.04018259e-01 -2.67162323e-01 6.96397007e-01
-9.91675705e-02 5.95250487e-01 7.99715966e-02 4.39253598e-01
1.15671313e+00 -3.24858844e-01 -8.55927765e-02 -4.33012247e-01
-2.56000757e-02 -6.23762757e-02 -8.54012609e-01 1.84507430e+00
-3.40046167e-01 9.20178652e-01 4.25782651e-01 -9.55287635e-01
8.11099291e-01 3.89976650e-01 -2.26095557e-01 -1.42744347e-01
3.54212709e-02 3.36869061e-01 -2.39764065e-01 -7.87552595e-01
1.30991220e-01 -3.76085401e-01 5.33267736e-01 1.91549510e-01
5.65183818e-01 -7.72908852e-02 -4.36632186e-01 6.60728589e-02
1.12050617e+00 3.56206000e-01 -2.66048461e-01 -3.27535599e-01
2.76660472e-01 -3.07870656e-01 7.50255525e-01 1.09044349e+00
-3.31936747e-01 1.23979616e+00 1.40477106e-01 -3.92969459e-01
-1.01725662e+00 -1.42819321e+00 -1.19897462e-01 3.53475809e-01
2.69333065e-01 5.69349945e-01 -1.16296458e+00 -5.82420766e-01
-5.13116978e-02 6.85499489e-01 -5.93386948e-01 -2.09571689e-01
-1.89476833e-01 -9.22694623e-01 3.41210306e-01 4.80069399e-01
1.00032163e+00 -8.44729722e-01 -1.01649892e-02 -5.81306405e-02
-1.35108292e-01 -1.16123879e+00 -6.13698959e-01 3.13702315e-01
-1.09209979e+00 -1.10722446e+00 -1.25596976e+00 -1.03824067e+00
1.09162736e+00 3.69715661e-01 7.31012166e-01 -3.44244897e-01
-7.54804015e-02 3.20776343e-01 2.41153404e-01 4.85820770e-02
-3.19059879e-01 -5.55776894e-01 3.52159441e-01 4.58355039e-01
3.74105304e-01 -6.51673317e-01 -7.38275886e-01 -6.23092055e-02
-9.28483129e-01 -6.67403489e-02 9.81509268e-01 1.21224105e+00
4.15177554e-01 1.68290988e-01 5.50170004e-01 -7.55511880e-01
6.04812622e-01 -2.21507907e-01 -8.21923196e-01 1.35288656e-01
-6.72105968e-01 4.01759177e-01 5.55491984e-01 -5.95780134e-01
-1.81176865e+00 6.39575049e-02 3.40294570e-01 -1.16310620e+00
-2.73730755e-01 2.62538165e-01 -1.39865369e-01 -2.06369236e-01
7.07976222e-01 3.45273256e-01 1.06778413e-01 -9.51011002e-01
4.79717821e-01 7.92525709e-01 1.21232903e+00 -3.73013020e-01
1.04096425e+00 5.86117029e-01 -3.34323257e-01 -4.59825128e-01
-1.26237154e+00 -1.85921580e-01 -6.17496729e-01 -5.77109046e-02
8.74506354e-01 -1.33543265e+00 -5.78281701e-01 1.03545618e+00
-1.43039048e+00 -5.17584264e-01 -2.01879978e-01 9.47132707e-01
-5.89897096e-01 4.33799326e-01 -1.07553172e+00 -8.97656739e-01
-1.50466233e-01 -1.06335771e+00 5.54392636e-01 6.31211936e-01
2.84282118e-01 -1.05887055e+00 -3.37183438e-02 5.94311357e-01
5.46351135e-01 -1.13370322e-01 5.36010265e-01 -2.23525226e-01
-1.12076795e+00 4.11819555e-02 -7.27744401e-01 1.31203949e+00
3.84172499e-01 -4.53660458e-01 -1.54062450e+00 -3.15839589e-01
7.23215342e-01 -4.88798529e-01 1.23677468e+00 8.24853063e-01
1.20922112e+00 -6.67179346e-01 2.75605142e-01 8.81065071e-01
1.71349204e+00 -2.54788280e-01 7.58995771e-01 1.44141689e-01
9.26879287e-01 2.28399813e-01 7.98965469e-02 1.09385245e-01
1.34801567e-01 9.07405317e-02 7.57188261e-01 -2.63976961e-01
-5.12986064e-01 -1.88806169e-02 5.51760614e-01 5.22926807e-01
2.30085030e-02 1.86853558e-01 -3.82967412e-01 6.55868351e-01
-1.99455893e+00 -9.46956217e-01 -1.31300181e-01 2.25852132e+00
1.27060986e+00 6.52290434e-02 -5.60755134e-01 -3.64345789e-01
8.89615357e-01 9.60417688e-02 -9.15516794e-01 4.06207666e-02
-3.14807028e-01 2.47099400e-01 6.64296210e-01 8.86437118e-01
-1.22921348e+00 7.64000356e-01 6.40409613e+00 5.87952495e-01
-9.51626241e-01 3.30546647e-01 7.38481522e-01 8.11652690e-02
5.85316727e-03 -1.02249328e-02 -2.81283468e-01 5.61593592e-01
7.84089923e-01 2.24011634e-02 9.51696694e-01 8.79910052e-01
5.59078455e-01 -2.20948696e-01 -1.06160200e+00 1.18637073e+00
1.72043383e-01 -1.28830338e+00 -1.89423531e-01 -1.66861385e-01
1.13744867e+00 2.72314429e-01 3.08427095e-01 -9.37211290e-02
4.90558475e-01 -1.00484073e+00 7.26932108e-01 1.18448031e+00
6.55875981e-01 -4.00586903e-01 9.73810673e-01 2.64068782e-01
-2.14650825e-01 -1.18733801e-01 -6.82942033e-01 -2.24467695e-01
2.62448080e-02 1.00611269e+00 -3.53313684e-01 3.93282205e-01
9.33401525e-01 7.77718246e-01 -2.56669253e-01 1.47690046e+00
-6.57612622e-01 8.96438837e-01 2.73386031e-01 5.74091077e-01
1.01362139e-01 -4.39788133e-01 5.54717481e-01 1.06141317e+00
3.26323599e-01 -1.02084592e-01 -2.42383525e-01 1.46363270e+00
-4.96528268e-01 -8.87543201e-01 -4.13017094e-01 2.82591432e-01
2.15626717e-01 1.19893003e+00 -8.27679038e-02 -4.72699076e-01
-4.01949197e-01 1.49966288e+00 2.94668317e-01 1.01095855e+00
-6.39387667e-01 -4.44941729e-01 7.42779791e-01 -3.03073943e-01
5.55089235e-01 -4.22724187e-02 -4.22476143e-01 -1.36675215e+00
1.84977464e-02 -5.62086940e-01 -1.69359177e-01 -1.52011836e+00
-1.82522643e+00 6.07564151e-01 -1.80455089e-01 -1.16555560e+00
2.30089929e-02 -5.99189878e-01 -6.21739745e-01 1.25045252e+00
-1.83480906e+00 -8.50980818e-01 -5.06686747e-01 4.49853569e-01
4.47401911e-01 -1.70521110e-01 5.17071128e-01 2.01740906e-01
-4.86153543e-01 1.69688404e-01 7.36521482e-01 3.12802225e-01
1.01736307e+00 -1.63476753e+00 -9.20990482e-02 1.28291249e+00
-1.06639631e-01 6.61991537e-01 9.25154269e-01 -4.03458983e-01
-1.12496614e+00 -1.33577085e+00 5.52895784e-01 -3.87111068e-01
6.06473327e-01 -1.87353417e-01 -1.24166453e+00 7.09602535e-01
6.79641187e-01 5.11335909e-01 1.13163665e-01 -4.01848346e-01
-4.34301704e-01 -4.44779605e-01 -1.33831418e+00 2.79394805e-01
6.55316055e-01 -8.58296335e-01 -1.00557911e+00 2.36052066e-01
8.28797340e-01 -6.08242273e-01 -3.75934154e-01 1.97331786e-01
7.12184384e-02 -1.03267801e+00 9.30020690e-01 -4.26216781e-01
5.16789794e-01 -5.36253154e-01 -1.57324262e-02 -1.47997880e+00
-3.81359726e-01 -9.10572410e-01 -5.07358551e-01 1.09809208e+00
9.17365123e-03 -5.02927005e-01 7.49890864e-01 6.47482991e-01
-3.24190587e-01 -1.93468973e-01 -8.22375238e-01 -9.08366621e-01
-1.31228670e-01 5.24666868e-02 4.30196822e-02 7.18026042e-01
-6.23638391e-01 8.74343216e-02 -8.17230225e-01 7.68419445e-01
1.45057333e+00 -3.94461639e-02 2.50914484e-01 -9.56642091e-01
-3.20405483e-01 2.36733537e-02 -1.32095039e-01 -1.36915779e+00
1.66337952e-01 -4.84123975e-01 5.65379620e-01 -1.56252742e+00
5.51237166e-01 1.43707320e-01 -4.66957569e-01 3.71804714e-01
-2.50352681e-01 3.82694155e-01 -1.77835733e-01 3.85805160e-01
-5.10107040e-01 7.30216861e-01 1.23750436e+00 -2.93093503e-01
1.07130624e-01 9.72880274e-02 -6.85660124e-01 9.34811354e-01
5.23437202e-01 -7.24727571e-01 -5.10544658e-01 -8.18623126e-01
-5.20922661e-01 -1.46021426e-01 9.38170612e-01 -9.80808377e-01
4.29945707e-01 -1.01424791e-01 6.08375371e-01 -2.55247056e-01
3.70541334e-01 -6.64679825e-01 -1.81247927e-02 3.00354660e-01
-3.52772176e-01 -6.82327032e-01 -1.32279500e-01 8.55172932e-01
-2.69091725e-01 -4.93666291e-01 1.09024632e+00 -1.74626425e-01
-4.91748989e-01 1.70477808e-01 -2.86437631e-01 1.87758096e-02
3.21986049e-01 2.90854797e-02 -5.44079125e-01 -5.89411378e-01
-8.20041180e-01 -2.44562775e-01 3.89304698e-01 1.36616036e-01
8.49788129e-01 -1.28018904e+00 -6.16809726e-01 -3.76677886e-03
-3.16320479e-01 2.59847194e-01 2.18815431e-01 7.38871336e-01
-5.13132930e-01 7.97646865e-02 -6.92544952e-02 -4.48912024e-01
-7.19481826e-01 4.44381952e-01 6.27889037e-01 2.28028893e-01
-5.84895849e-01 1.02944076e+00 4.58067596e-01 -3.10431153e-01
5.85591733e-01 -3.95800620e-01 5.04009187e-01 -6.75614893e-01
7.47601330e-01 3.11209619e-01 -1.13562644e-01 -2.86253929e-01
-1.30301043e-02 1.83374584e-01 -1.12669468e-01 -2.35452026e-01
1.59748924e+00 -3.40373933e-01 -4.08446372e-01 3.63512605e-01
1.17152941e+00 -2.25125462e-01 -2.26655316e+00 -6.52182341e-01
4.61098813e-02 -7.19112039e-01 5.43243945e-01 -1.09131539e+00
-1.29406238e+00 8.15881252e-01 8.57066751e-01 -2.46765599e-01
1.16405523e+00 -1.09925158e-01 6.64941072e-01 4.07574117e-01
6.31990358e-02 -8.83252442e-01 2.89492279e-01 4.04092938e-01
9.31046784e-01 -1.46335840e+00 -7.43053854e-02 1.33441806e-01
-4.66779679e-01 1.15921390e+00 6.42874300e-01 -4.63480711e-01
6.57260776e-01 1.30255952e-01 3.04652900e-01 6.38392940e-02
-5.53270340e-01 -8.79756138e-02 2.32703298e-01 7.41699994e-01
-7.91199654e-02 -3.52921844e-01 2.63473660e-01 6.59583449e-01
4.38418150e-01 4.44535941e-01 6.77973270e-01 5.55881798e-01
-5.22398233e-01 -7.14203656e-01 -5.99769533e-01 2.12269038e-01
-4.26578701e-01 -3.97736758e-01 -2.37684660e-02 3.39182287e-01
1.27758965e-01 8.18115532e-01 -7.67418817e-02 5.36988825e-02
-8.03406723e-03 -1.00900479e-01 6.44181430e-01 -5.04810333e-01
-2.33101994e-02 1.88117653e-01 -3.85796666e-01 -3.02533805e-01
-5.53612649e-01 -4.08283859e-01 -9.43925023e-01 -1.88144609e-01
-2.26210535e-01 -5.03851622e-02 3.61574441e-01 1.04065633e+00
2.49270603e-01 3.04281563e-01 7.17751622e-01 -1.10895932e+00
-9.15297627e-01 -1.13799775e+00 -7.90459275e-01 2.92846888e-01
9.49138522e-01 -5.39642274e-01 -9.60976303e-01 5.57971001e-01] | [11.62326431274414, -2.7044129371643066] |
b8d66374-3078-4559-b954-afd0d0766d8d | mesh-guided-one-shot-face-reenactment-using | 2008.07783 | null | https://arxiv.org/abs/2008.07783v2 | https://arxiv.org/pdf/2008.07783v2.pdf | Mesh Guided One-shot Face Reenactment using Graph Convolutional Networks | Face reenactment aims to animate a source face image to a different pose and expression provided by a driving image. Existing approaches are either designed for a specific identity, or suffer from the identity preservation problem in the one-shot or few-shot scenarios. In this paper, we introduce a method for one-shot face reenactment, which uses the reconstructed 3D meshes (i.e., the source mesh and driving mesh) as guidance to learn the optical flow needed for the reenacted face synthesis. Technically, we explicitly exclude the driving face's identity information in the reconstructed driving mesh. In this way, our network can focus on the motion estimation for the source face without the interference of driving face shape. We propose a motion net to learn the face motion, which is an asymmetric autoencoder. The encoder is a graph convolutional network (GCN) that learns a latent motion vector from the meshes, and the decoder serves to produce an optical flow image from the latent vector with CNNs. Compared to previous methods using sparse keypoints to guide the optical flow learning, our motion net learns the optical flow directly from 3D dense meshes, which provide the detailed shape and pose information for the optical flow, so it can achieve more accurate expression and pose on the reenacted face. Extensive experiments show that our method can generate high-quality results and outperforms state-of-the-art methods in both qualitative and quantitative comparisons. | ['Tianjia Shao', 'Yi Yuan', 'Kun Zhou', 'Guangming Yao'] | 2020-08-18 | null | null | null | null | ['face-reenactment'] | ['computer-vision'] | [-1.60703734e-01 1.46195456e-01 -5.81274144e-02 -2.27611437e-01
-1.19676977e-01 -3.06733817e-01 4.15881813e-01 -9.84459519e-01
2.40429550e-01 4.29402560e-01 3.69225055e-01 4.36588705e-01
3.20757061e-01 -9.80613112e-01 -7.48399675e-01 -8.18432570e-01
3.56630236e-01 3.68823528e-01 -3.69152606e-01 -4.02093261e-01
7.22900182e-02 7.53649652e-01 -1.63364434e+00 5.44771254e-02
4.79990512e-01 1.01524782e+00 -4.52381819e-02 5.34092665e-01
-2.30703607e-01 1.04984367e+00 -4.21674788e-01 -5.25480747e-01
3.22024375e-01 -7.83423066e-01 -8.04669142e-01 4.28162128e-01
8.59890997e-01 -8.46737683e-01 -7.68454432e-01 1.08623385e+00
4.73556459e-01 7.06030577e-02 4.18908209e-01 -1.71395874e+00
-8.88395965e-01 8.79918039e-02 -6.57609701e-01 -3.80941123e-01
6.51673794e-01 3.19394857e-01 6.45787537e-01 -1.09630704e+00
1.21902096e+00 1.67339969e+00 5.66886365e-01 1.00694764e+00
-1.12312734e+00 -1.02540851e+00 8.67835730e-02 1.07840396e-01
-1.42980397e+00 -8.67295980e-01 1.32932818e+00 -4.79605287e-01
1.35602131e-01 -1.60251856e-01 1.07136989e+00 1.31102180e+00
1.52122006e-01 6.00151122e-01 6.16231024e-01 5.70032597e-02
3.02719418e-03 -1.45422220e-01 -6.64518356e-01 1.21206975e+00
-1.14154302e-01 3.63256991e-01 -5.64069569e-01 8.25819001e-03
1.25461924e+00 7.51718804e-02 -5.99618316e-01 -6.25113368e-01
-1.04250991e+00 9.38388526e-01 5.16767979e-01 1.70533493e-01
-3.90019804e-01 2.61649370e-01 1.99714944e-01 2.43140683e-01
5.02193213e-01 4.80324142e-02 -1.93777103e-02 3.28376889e-02
-8.96735728e-01 2.70475537e-01 9.89987254e-01 9.77034092e-01
1.19527411e+00 4.92285252e-01 -1.14671394e-01 5.36140144e-01
4.86584127e-01 5.46318889e-01 1.74498543e-01 -1.57201862e+00
8.92686993e-02 5.67333221e-01 3.91417593e-02 -1.48728681e+00
-2.36011297e-02 -8.09254497e-02 -1.01692593e+00 3.68141383e-01
2.16942400e-01 -3.09150547e-01 -8.35765958e-01 1.98804641e+00
5.93188822e-01 8.08887482e-01 3.09276767e-02 1.07268322e+00
1.19754994e+00 6.42279983e-01 -4.13665116e-01 -3.77224058e-01
9.23230231e-01 -1.03445935e+00 -1.13937497e+00 -4.33296990e-03
4.63851303e-01 -7.09189713e-01 6.39004290e-01 -3.25282407e-03
-1.14965653e+00 -6.73275709e-01 -9.62789059e-01 -2.65055627e-01
1.26848340e-01 -8.93983096e-02 3.08309346e-01 2.41087392e-01
-1.26426220e+00 6.66309237e-01 -6.79105759e-01 -2.53121823e-01
5.62020123e-01 3.50644439e-01 -7.24259436e-01 -1.51363939e-01
-1.10769200e+00 6.36927903e-01 -1.42750308e-01 3.11162382e-01
-1.13093770e+00 -9.35100436e-01 -1.26763391e+00 -5.68426736e-02
1.15952156e-01 -9.38911557e-01 1.08443570e+00 -1.37855637e+00
-2.00708294e+00 9.85808730e-01 -2.30757177e-01 3.17898154e-01
5.95330119e-01 8.49769861e-02 -2.48917758e-01 3.04845512e-01
2.60677040e-01 9.71902192e-01 1.18240047e+00 -1.45744264e+00
-2.94093311e-01 -2.26948977e-01 7.15342090e-02 9.58088636e-02
-2.41789326e-01 -2.47165054e-01 -6.10677004e-01 -7.26583660e-01
4.21165824e-02 -8.81673038e-01 1.08596906e-01 6.68616176e-01
-1.92176089e-01 9.05432329e-02 1.41984272e+00 -6.40411258e-01
8.16524446e-01 -2.07529521e+00 5.78754425e-01 -1.62817180e-01
2.60206938e-01 1.00457110e-01 -3.71679753e-01 2.47521117e-01
-1.78361818e-01 -1.42050371e-01 -2.07118914e-01 -6.15967095e-01
-3.46382439e-01 3.43769461e-01 -2.67383486e-01 7.06173539e-01
3.08260262e-01 1.06730735e+00 -1.24446189e+00 -6.02373183e-01
2.87570238e-01 9.75527585e-01 -7.61237144e-01 6.13333702e-01
-2.36602798e-02 9.01529551e-01 -2.77908593e-01 5.15423000e-01
8.85695815e-01 -1.33225054e-01 -5.20013236e-02 -5.93517900e-01
-2.59349067e-02 -2.60564178e-01 -1.24520993e+00 2.17076206e+00
-4.72829103e-01 4.71867591e-01 5.94247222e-01 -6.26760602e-01
9.89797413e-01 4.37482774e-01 8.80416989e-01 -5.51164389e-01
4.60634530e-01 9.39433798e-02 -3.94001186e-01 -5.61882198e-01
1.99429579e-02 -4.29567665e-01 2.23908678e-01 4.87761617e-01
3.48830104e-01 -3.71924400e-01 -1.81203231e-01 1.18398689e-01
5.70868313e-01 3.35686654e-01 -1.64302200e-01 -1.58082008e-01
7.80065894e-01 -4.40014958e-01 9.49151933e-01 -7.54168779e-02
-1.39871657e-01 7.89843976e-01 4.15676117e-01 -7.86400974e-01
-9.76325572e-01 -9.19269741e-01 1.29282907e-01 4.87207264e-01
4.37414110e-01 -3.42404276e-01 -9.37331319e-01 -5.95406890e-01
-2.99634580e-02 2.03949451e-01 -8.64033520e-01 -4.39581484e-01
-9.07770097e-01 -5.41615747e-02 3.05471510e-01 2.71388233e-01
7.24524140e-01 -1.17428708e+00 -2.13532701e-01 2.38005146e-01
-3.61956745e-01 -1.12104595e+00 -9.39363658e-01 -7.93432295e-01
-5.89047432e-01 -1.20072329e+00 -7.89653182e-01 -1.11487973e+00
8.12415898e-01 1.78657874e-01 8.15195262e-01 3.88875306e-01
-1.29415646e-01 2.97120184e-01 -4.51515242e-02 6.96990937e-02
-5.24612725e-01 -3.07430774e-01 9.81005058e-02 7.00216174e-01
-1.56507879e-01 -8.25451732e-01 -8.11527431e-01 3.23482662e-01
-8.00202012e-01 2.91137129e-01 5.57073839e-02 9.27508771e-01
3.36539418e-01 -3.37628841e-01 3.92902404e-01 -7.18316138e-01
1.65626854e-01 -3.77271473e-01 -3.68641973e-01 -8.27550218e-02
-2.33937800e-01 -1.93010606e-02 7.57447302e-01 -4.09064084e-01
-1.04864085e+00 2.76735723e-01 -2.63990849e-01 -1.16324401e+00
1.45894930e-01 -1.68099076e-01 -4.88192022e-01 -5.93747139e-01
2.65475005e-01 9.07019824e-02 5.82085311e-01 -3.14580142e-01
6.37920439e-01 3.59919727e-01 6.72569871e-01 -3.97606730e-01
1.06983578e+00 9.92429137e-01 1.32546917e-01 -6.33205593e-01
-7.80152619e-01 3.98014672e-02 -8.20905328e-01 -6.03044510e-01
8.40660870e-01 -8.51774037e-01 -1.00041056e+00 8.57520938e-01
-1.50585115e+00 -2.13985607e-01 -2.62734175e-01 2.36084089e-01
-8.39827538e-01 2.08174735e-01 -8.54577661e-01 -2.42221162e-01
-4.03067350e-01 -1.29830599e+00 1.35245478e+00 3.63513112e-01
5.02343439e-02 -1.08012843e+00 1.46673188e-01 1.06136359e-01
2.79281199e-01 6.77613437e-01 5.62301815e-01 3.98128003e-01
-6.93406224e-01 5.73805608e-02 -8.30947794e-03 2.92936414e-01
5.40199637e-01 3.57571840e-01 -9.43850160e-01 -6.03212357e-01
2.30182081e-01 -2.53132492e-01 4.78410661e-01 1.88755482e-01
9.55114841e-01 -5.82361400e-01 -2.53814995e-01 1.25446093e+00
1.20200372e+00 1.07400343e-01 6.97806001e-01 -2.22144619e-01
1.25238764e+00 8.73569369e-01 3.28205585e-01 4.17556942e-01
3.20440620e-01 6.98334455e-01 6.51014566e-01 -2.22471029e-01
-4.75754708e-01 -6.34583831e-01 4.84422207e-01 9.84599888e-01
-8.87042785e-04 2.90680323e-02 -4.47521776e-01 4.23073113e-01
-1.76017189e+00 -9.43732381e-01 3.29876721e-01 1.85939682e+00
6.64058805e-01 -4.43588167e-01 -1.86361253e-01 -9.98233780e-02
8.88614058e-01 4.70209211e-01 -5.84428847e-01 -2.41223797e-01
6.40053079e-02 9.28975195e-02 -1.44084647e-01 8.57695401e-01
-6.75155222e-01 1.16663241e+00 5.78294230e+00 4.95501041e-01
-1.51615369e+00 7.80386180e-02 3.94972354e-01 -9.74280089e-02
-3.09574217e-01 -4.02642973e-02 -5.54994404e-01 5.00049591e-01
4.66246426e-01 -2.50736594e-01 5.34161627e-01 6.77910268e-01
2.54980952e-01 5.25753796e-01 -1.14940846e+00 1.31789994e+00
4.01692092e-01 -1.66070330e+00 4.84430969e-01 4.11549620e-02
9.47015524e-01 -5.58058560e-01 2.17629094e-02 -1.88669860e-02
1.02494575e-01 -1.06126130e+00 7.52704144e-01 7.68459320e-01
1.15335202e+00 -9.05676782e-01 4.29162711e-01 1.22004084e-01
-1.33073545e+00 9.68516171e-02 -1.89951897e-01 -4.08061920e-03
3.36039484e-01 1.27941713e-01 -3.32952797e-01 5.31592607e-01
4.96850193e-01 1.30021322e+00 -3.63943316e-02 5.59345424e-01
-2.82861680e-01 5.88199012e-02 7.31646270e-02 3.88217151e-01
1.19504414e-01 -4.69000340e-01 7.12127686e-01 5.69927812e-01
3.64683449e-01 1.53879896e-01 1.72255695e-01 1.10509288e+00
-4.85666335e-01 1.14237234e-01 -8.84424448e-01 2.69884586e-01
3.60345989e-01 1.44131589e+00 -3.64335597e-01 -2.31650591e-01
-4.13692117e-01 1.19157803e+00 2.63497472e-01 4.36863720e-01
-7.18763053e-01 -3.39654297e-01 1.16392016e+00 2.45747462e-01
1.66964665e-01 7.50852525e-02 3.72900218e-01 -1.27710795e+00
-1.56379342e-01 -7.10729122e-01 -2.93842837e-04 -1.03183663e+00
-1.21358573e+00 6.96729481e-01 -3.06050152e-01 -1.29359972e+00
-4.06054676e-01 -4.49676424e-01 -8.41328263e-01 7.19424367e-01
-1.40656340e+00 -1.33331621e+00 -8.37562442e-01 8.69265199e-01
3.56249362e-01 -1.33399785e-01 7.33185649e-01 3.86488676e-01
-5.16660452e-01 5.78625321e-01 -3.51075768e-01 5.95129907e-01
7.02570200e-01 -5.85044682e-01 2.71303475e-01 5.48896849e-01
1.92774627e-02 3.82703245e-01 3.88292551e-01 -5.43910563e-01
-1.75331843e+00 -1.34050393e+00 5.63373983e-01 -2.46699765e-01
3.40342373e-01 -3.72253060e-01 -8.79552484e-01 7.57017195e-01
1.41747639e-01 5.17997324e-01 2.06991345e-01 -5.82851231e-01
-2.13888526e-01 -3.65145296e-01 -1.25456202e+00 6.95523143e-01
1.41107559e+00 -6.85302436e-01 -2.80054033e-01 9.28692222e-02
7.99610317e-01 -5.89250386e-01 -8.44116330e-01 3.59459668e-01
5.95728874e-01 -8.82489562e-01 8.07061911e-01 -5.66447794e-01
6.75665140e-01 -4.30272430e-01 1.96732536e-01 -1.38409281e+00
-3.45938563e-01 -9.63655651e-01 -2.26181239e-01 1.27116895e+00
-3.02643448e-01 -3.51017922e-01 9.31462348e-01 3.03964674e-01
-7.91234821e-02 -7.94590950e-01 -9.02012050e-01 -4.22944158e-01
7.62080252e-02 5.58940461e-03 9.83400822e-01 1.25683653e+00
-5.41340947e-01 3.60903174e-01 -6.69720411e-01 -2.62692664e-02
6.84455276e-01 3.06327701e-01 8.81348908e-01 -1.15651405e+00
2.22573712e-01 -2.61738747e-01 -5.66041172e-01 -1.09269512e+00
8.00255537e-01 -9.81087983e-01 -5.69415875e-02 -1.30906713e+00
-3.76210324e-02 -4.21811789e-02 2.00384885e-01 3.86034161e-01
1.60977840e-01 4.54868108e-01 2.21595511e-01 1.14853449e-01
-4.56313901e-02 1.03659558e+00 2.06864238e+00 -1.70664698e-01
-1.55349195e-01 -3.95936191e-01 -5.44030845e-01 8.07117462e-01
3.61686200e-01 -3.47130597e-01 -4.29253757e-01 -5.44736505e-01
-4.83800992e-02 4.52412009e-01 5.35110533e-01 -7.76982248e-01
3.54468107e-01 -2.85294056e-01 4.45400387e-01 -2.74942249e-01
5.19514084e-01 -7.67033577e-01 5.16681910e-01 4.89688307e-01
8.61852430e-03 -5.95317930e-02 -1.05685867e-01 4.55027401e-01
-2.46552140e-01 1.10823371e-01 1.05769312e+00 -1.65775344e-01
-5.44400990e-01 1.17549169e+00 1.10570572e-01 9.89394635e-02
1.00463903e+00 -3.27068418e-01 -7.64907375e-02 -6.66472554e-01
-6.21618032e-01 2.40308225e-01 8.56876016e-01 8.50808144e-01
9.03371632e-01 -1.94281507e+00 -8.55569601e-01 6.42918169e-01
-1.05498768e-01 2.42506847e-01 1.96195930e-01 5.48554242e-01
-7.54760623e-01 -1.48184851e-01 -6.54740393e-01 -6.18978918e-01
-1.01843226e+00 5.55250466e-01 7.94440269e-01 3.79161924e-01
-8.23267937e-01 8.26224804e-01 5.38988352e-01 -4.00327593e-01
4.82810065e-02 2.97810227e-01 -1.17511675e-01 1.23796694e-01
5.80371320e-01 3.20193201e-01 -2.76069045e-01 -1.32801712e+00
-2.88020700e-01 1.17687404e+00 2.09925905e-01 -6.77869618e-02
1.28020108e+00 -9.21810269e-02 -4.58073080e-01 2.08122581e-01
1.85247684e+00 -3.22212018e-02 -1.65195668e+00 -9.01618525e-02
-6.81880176e-01 -8.89869094e-01 6.99848905e-02 2.05341186e-02
-1.99862754e+00 9.17669356e-01 3.19729507e-01 -4.72059906e-01
1.05419934e+00 -9.32479948e-02 1.14029729e+00 -9.00140703e-02
5.11987627e-01 -7.52422035e-01 5.01166761e-01 3.50837797e-01
1.22227967e+00 -9.98981059e-01 -2.72921741e-01 -5.18344164e-01
-3.50991577e-01 1.30788743e+00 9.62312877e-01 -2.94395357e-01
9.14115489e-01 1.96975574e-01 1.98485866e-01 -3.98837507e-01
-5.65455079e-01 1.97188675e-01 7.94721097e-02 6.53311133e-01
6.16927408e-02 -2.60493040e-01 3.08925837e-01 2.05168694e-01
-3.16564500e-01 9.05376747e-02 5.33109069e-01 6.65243447e-01
-5.99964485e-02 -9.85151649e-01 -1.92738220e-01 3.10999118e-02
-7.05969557e-02 3.24978113e-01 -4.38979179e-01 6.42507613e-01
4.73583221e-01 7.73768783e-01 3.11186880e-01 -4.92847234e-01
4.00092661e-01 -2.23854274e-01 6.66890383e-01 -6.24233961e-01
-2.82974243e-01 -9.05035362e-02 -4.26410973e-01 -9.89623725e-01
-5.00297010e-01 -3.90510529e-01 -1.37345767e+00 -8.73792470e-01
-6.35353178e-02 4.18923274e-02 2.76404530e-01 6.09601319e-01
4.37846094e-01 2.76844651e-01 1.18373787e+00 -1.22410297e+00
6.57440862e-03 -5.62538445e-01 -6.76579773e-01 8.47279429e-01
7.31960654e-01 -9.47609782e-01 -6.12546861e-01 2.18885809e-01] | [12.796710968017578, -0.2652316093444824] |
b6d96eb8-24ad-4180-a0eb-956c5e37ec79 | metaregnet-metamorphic-image-registration | 2303.09088 | null | https://arxiv.org/abs/2303.09088v1 | https://arxiv.org/pdf/2303.09088v1.pdf | MetaRegNet: Metamorphic Image Registration Using Flow-Driven Residual Networks | Deep learning based methods provide efficient solutions to medical image registration, including the challenging problem of diffeomorphic image registration. However, most methods register normal image pairs, facing difficulty handling those with missing correspondences, e.g., in the presence of pathology like tumors. We desire an efficient solution to jointly account for spatial deformations and appearance changes in the pathological regions where the correspondences are missing, i.e., finding a solution to metamorphic image registration. Some approaches are proposed to tackle this problem, but they cannot properly handle large pathological regions and deformations around pathologies. In this paper, we propose a deep metamorphic image registration network (MetaRegNet), which adopts time-varying flows to drive spatial diffeomorphic deformations and generate intensity variations. We evaluate MetaRegNet on two datasets, i.e., BraTS 2021 with brain tumors and 3D-IRCADb-01 with liver tumors, showing promising results in registering a healthy and tumor image pair. The source code is available online. | ['Yi Hong', 'Ankita Joshi'] | 2023-03-16 | null | null | null | null | ['medical-image-registration'] | ['medical'] | [-1.94041416e-01 2.56687663e-02 -6.35840073e-02 -3.13594729e-01
-8.45088601e-01 -3.15472364e-01 3.49328548e-01 -8.93562809e-02
-4.56560999e-01 4.25658494e-01 3.09731603e-01 7.56234303e-02
-5.27080372e-02 -5.66724718e-01 -3.82345021e-01 -8.18183124e-01
-6.82294294e-02 6.01892412e-01 1.09280415e-01 -3.18572700e-01
-9.99712199e-02 3.97552639e-01 -5.20866156e-01 6.79780915e-02
9.68927741e-01 5.72770894e-01 2.73488432e-01 3.08003187e-01
1.32493712e-02 3.94722670e-01 -9.95231345e-02 -1.23247243e-01
3.74298483e-01 -6.00344479e-01 -1.01090384e+00 2.70683616e-02
4.07155126e-01 -3.51515085e-01 -5.51603258e-01 1.18404484e+00
5.39904892e-01 -9.30202305e-02 5.82391202e-01 -1.33928740e+00
-4.79593277e-01 3.72785717e-01 -6.52013004e-01 6.40801847e-01
-4.12678346e-02 3.16170931e-01 4.59258348e-01 -6.66246176e-01
7.33763218e-01 8.01591456e-01 8.91248465e-01 6.04236066e-01
-1.08529723e+00 -4.60018128e-01 -2.26388335e-01 6.41308576e-02
-1.21405518e+00 -4.45371807e-01 7.44972169e-01 -6.60869837e-01
6.52597189e-01 1.09555602e-01 7.35912561e-01 9.84049618e-01
6.40412569e-01 3.75381857e-01 1.19837320e+00 3.10345497e-02
-7.40820542e-02 -5.46387017e-01 8.19051340e-02 6.94684148e-01
1.88913152e-01 1.26791477e-01 -1.06895126e-01 -1.77176610e-01
1.04506671e+00 2.54050344e-01 -5.89248121e-01 -3.31835687e-01
-1.63144577e+00 7.00788260e-01 6.49530292e-01 7.56950438e-01
-3.88998657e-01 1.27524724e-02 3.23148042e-01 4.13858265e-01
6.25273407e-01 2.76193947e-01 -1.76103711e-01 1.70943573e-01
-8.58175993e-01 3.58852930e-02 4.17801291e-01 5.61563075e-01
4.62913275e-01 -5.11059090e-02 -3.56079862e-02 6.30316913e-01
3.64639521e-01 1.87318802e-01 1.05419874e+00 -6.96714640e-01
3.33328843e-01 6.18841648e-01 -2.97055036e-01 -9.21133161e-01
-7.94268072e-01 -2.24232882e-01 -1.25311565e+00 1.66488692e-01
7.45304585e-01 -7.68338218e-02 -8.55131388e-01 1.79001272e+00
4.41117167e-01 4.70504045e-01 1.15594864e-01 1.17030466e+00
1.00935268e+00 1.70662999e-01 5.43613844e-02 -1.54485196e-01
1.27936125e+00 -8.15096259e-01 -6.74881577e-01 -1.71204403e-01
8.90457630e-01 -6.41384721e-01 9.66441810e-01 -2.98618704e-01
-1.24140370e+00 -2.85913367e-02 -6.55413449e-01 -8.77274871e-02
-8.64903033e-02 -3.06030601e-01 4.04960036e-01 1.53765485e-01
-1.28164029e+00 6.38952792e-01 -1.38799298e+00 -4.33925390e-01
4.71578777e-01 5.36826909e-01 -6.68418109e-01 4.51200195e-02
-1.07132065e+00 9.86454904e-01 4.31593172e-02 1.72749206e-01
-8.22893202e-01 -1.15830529e+00 -8.43213260e-01 -2.28476465e-01
-3.17756534e-02 -7.28673279e-01 9.05933142e-01 -9.58616793e-01
-1.09578037e+00 1.30182981e+00 8.60823877e-03 -3.22540194e-01
9.46560204e-01 3.12947661e-01 -2.73914218e-01 1.14910558e-01
6.76689595e-02 6.57976031e-01 4.68751431e-01 -9.43281889e-01
6.93968609e-02 -6.14543259e-01 -3.47263753e-01 3.75457644e-01
-2.09782958e-01 1.68576032e-01 -1.03392623e-01 -7.48821318e-01
3.21028620e-01 -1.05851817e+00 -4.40035582e-01 4.75159407e-01
-4.16859686e-01 2.32089177e-01 9.98519123e-01 -1.03648949e+00
4.47509587e-01 -1.97357464e+00 3.09645563e-01 1.71918854e-01
6.09556973e-01 1.69440106e-01 -3.48871559e-01 -2.63117403e-01
-3.01448584e-01 -5.99928945e-03 -6.19016469e-01 -2.61922181e-01
-4.12933946e-01 2.67925560e-01 1.36783719e-01 8.80293131e-01
-7.19085187e-02 1.33993423e+00 -8.34855676e-01 -5.24220765e-01
5.50621971e-02 7.08456576e-01 -4.63313013e-01 1.04411893e-01
1.63157478e-01 1.05409765e+00 -4.99255866e-01 4.72145408e-01
7.15116024e-01 -2.96548128e-01 7.04799891e-02 -5.24150431e-01
-9.60590914e-02 -5.92246428e-02 -5.91616392e-01 1.93404448e+00
-4.53571767e-01 5.37535191e-01 4.27770346e-01 -1.17753005e+00
7.22616971e-01 5.46708643e-01 1.35832644e+00 -6.86154366e-01
4.04041380e-01 3.33842546e-01 3.14822257e-01 -5.66576242e-01
-1.64753683e-02 -3.42177629e-01 2.20272750e-01 5.80167294e-01
-7.61875808e-02 -1.92156330e-01 -8.17409605e-02 -5.73453158e-02
1.26720083e+00 -2.73250431e-01 1.56020731e-01 -5.38572788e-01
5.24541140e-01 -2.54758205e-02 6.29579604e-01 9.46110338e-02
-6.79342687e-01 8.77142429e-01 4.36077744e-01 -7.89602220e-01
-1.05286086e+00 -1.14700437e+00 -1.55703127e-01 2.76048984e-02
3.57422918e-01 -7.52891824e-02 -9.01329815e-01 -6.99231148e-01
-1.72318235e-01 -1.36214793e-01 -6.03770554e-01 -2.99506545e-01
-8.88893902e-01 -1.24805546e+00 4.83808100e-01 2.66949147e-01
7.23142385e-01 -1.03030491e+00 -4.88406092e-01 2.86078930e-01
-5.11691570e-01 -1.16012180e+00 -9.63327825e-01 -2.79710293e-01
-1.15578318e+00 -1.29964316e+00 -1.04756284e+00 -9.84270215e-01
1.12987554e+00 8.49286467e-02 9.87220526e-01 3.91702890e-01
-7.42740154e-01 2.45530665e-01 1.55420566e-03 2.01502070e-01
-7.25044727e-01 7.18806610e-02 -4.21758369e-02 -1.71343371e-01
-9.07895640e-02 -7.59329379e-01 -7.03949511e-01 7.02702522e-01
-9.47197437e-01 1.41053528e-01 4.39333200e-01 9.52981472e-01
7.15519607e-01 -2.34710321e-01 2.92773545e-01 -5.78888953e-01
5.07521033e-01 -4.74861592e-01 -3.65362257e-01 2.59013534e-01
-4.02600527e-01 -7.51781687e-02 3.53740096e-01 -4.95387107e-01
-7.53502071e-01 2.12271243e-01 -3.02912444e-01 -4.68356401e-01
-4.45729345e-02 3.98063630e-01 -1.45473694e-02 -5.54544210e-01
5.42449892e-01 2.08880484e-01 6.33844316e-01 -1.91326559e-01
-3.71823199e-02 1.79403082e-01 7.13385880e-01 -4.28466111e-01
9.18253958e-01 7.21607208e-01 9.37666968e-02 -5.04603267e-01
-5.14841795e-01 -3.34728539e-01 -9.23281491e-01 -2.19319776e-01
1.08640254e+00 -6.80722177e-01 -3.24867338e-01 8.14013302e-01
-1.03655815e+00 -7.23725736e-01 -5.43323755e-01 5.97189963e-01
-7.90768623e-01 3.13803285e-01 -8.61202002e-01 4.16037768e-01
-7.01298833e-01 -1.57267570e+00 8.04304361e-01 1.98028624e-01
-1.37747973e-01 -1.42852306e+00 4.50202495e-01 2.41096437e-01
8.62028837e-01 7.56594121e-01 1.01030183e+00 -4.41379994e-01
-3.81850123e-01 2.85269357e-02 -1.57350630e-01 -9.72172990e-02
6.07565761e-01 -4.15165633e-01 -4.85690057e-01 -4.30899888e-01
8.41237828e-02 -6.35828227e-02 6.40997767e-01 6.28191471e-01
1.01543903e+00 -2.36777797e-01 -3.38989288e-01 1.11706257e+00
1.08316135e+00 4.08487134e-02 7.65701115e-01 1.43629238e-01
8.65547001e-01 5.68284094e-01 6.00431673e-02 2.43225433e-02
5.35259008e-01 9.42513049e-01 4.15463239e-01 -7.51592100e-01
-5.52281976e-01 2.22007871e-01 8.35712105e-02 1.09792924e+00
-1.68527558e-01 1.39057025e-01 -1.07967532e+00 5.08560240e-01
-1.59663391e+00 -7.45264471e-01 -2.85837263e-01 1.87371469e+00
1.01753998e+00 -3.60171676e-01 -8.41019973e-02 -2.56754190e-01
7.01512337e-01 2.25340202e-02 -6.71396375e-01 1.79965973e-01
9.10813641e-03 1.75442528e-02 2.97159523e-01 5.04141927e-01
-1.06095314e+00 7.28664339e-01 6.11383533e+00 2.49148756e-01
-1.29694259e+00 6.89989567e-01 8.45333457e-01 9.73502696e-02
-2.28923231e-01 -1.03018738e-01 -1.20136000e-01 4.45384532e-01
5.83678246e-01 -3.04476649e-01 3.37469786e-01 2.07892403e-01
3.65504891e-01 2.15055838e-01 -9.11503196e-01 8.43418658e-01
3.03422641e-02 -1.28039515e+00 -2.80396193e-01 2.75937617e-01
6.81651771e-01 2.28768289e-01 8.79223272e-02 -2.05182597e-01
2.76417196e-01 -9.86593366e-01 1.82706833e-01 6.08048916e-01
9.22080100e-01 -3.49155962e-01 6.07372642e-01 4.76136692e-02
-1.04783440e+00 5.39067864e-01 -1.60240695e-01 6.01589501e-01
2.76203394e-01 6.35202765e-01 -5.35714090e-01 4.26416248e-01
6.57329500e-01 9.86642718e-01 -4.99540508e-01 1.21307588e+00
-7.97970500e-03 1.55545652e-01 -1.47957876e-01 7.70088136e-01
-4.21574973e-02 -4.08529967e-01 6.17183566e-01 7.96264052e-01
7.28728712e-01 1.46721646e-01 1.36768579e-01 1.10345054e+00
-1.44203722e-01 1.02754109e-01 -5.94011188e-01 4.03741986e-01
-9.57610384e-02 1.51767945e+00 -9.54399168e-01 -2.53681149e-02
-4.34246093e-01 1.01993406e+00 1.81598678e-01 1.62274048e-01
-7.22995520e-01 1.43914953e-01 8.47017646e-01 4.13143992e-01
-3.94320071e-01 -3.84531856e-01 -8.68137106e-02 -1.52260530e+00
-1.90635368e-01 -7.03473270e-01 4.49845433e-01 -5.60257018e-01
-1.31789172e+00 6.43667936e-01 -1.94097012e-02 -1.25078523e+00
6.74205180e-03 -3.76411885e-01 -1.19644451e+00 6.73897982e-01
-1.55686545e+00 -1.10924375e+00 -7.62008667e-01 9.81928706e-01
1.58568293e-01 -9.51827876e-03 6.39620662e-01 5.63591361e-01
-4.66934383e-01 6.14940703e-01 -1.63196415e-01 3.70042413e-01
9.72304046e-01 -1.01353085e+00 2.71168172e-01 7.76095510e-01
-9.96270478e-02 3.16820145e-01 2.72974193e-01 -6.55190587e-01
-1.37943375e+00 -1.23771954e+00 7.45441616e-01 -1.20547161e-01
7.30847538e-01 -1.48387104e-01 -1.08305955e+00 7.97842562e-01
1.67106435e-01 8.64923954e-01 4.10186350e-01 -7.59273589e-01
-5.45970462e-02 -1.34600773e-02 -1.63980746e+00 5.44443667e-01
1.07233918e+00 -2.07447127e-01 -3.20251018e-01 7.03805387e-01
4.43038106e-01 -9.86483574e-01 -1.30004013e+00 3.26659411e-01
1.90275937e-01 -6.26151860e-01 1.14704061e+00 -3.36624563e-01
2.91614562e-01 -2.09105134e-01 3.46231312e-01 -1.73009801e+00
2.32356810e-03 -6.53391242e-01 3.58114362e-01 8.54738414e-01
1.20543852e-01 -8.69741678e-01 6.15929484e-01 6.73032701e-01
-4.67247248e-01 -4.66064215e-01 -1.34356415e+00 -7.13294387e-01
6.02278709e-01 1.21761806e-01 4.68796194e-01 1.34398425e+00
-5.23200892e-02 -4.58628327e-01 -7.07570538e-02 -9.03745741e-02
6.30764961e-01 -2.06300586e-01 3.55097681e-01 -1.13569152e+00
1.15610883e-01 -5.47201335e-01 -5.75708330e-01 -2.65607208e-01
5.14038026e-01 -1.45165360e+00 4.44523953e-02 -1.34825611e+00
3.37012947e-01 -7.07436979e-01 -1.48201823e-01 7.58374393e-01
6.83044046e-02 4.86809224e-01 -1.05566017e-01 5.30751467e-01
-1.10688917e-01 5.08096516e-01 1.78068185e+00 -3.95136654e-01
-1.32266983e-01 -1.82995930e-01 -5.02581060e-01 6.96708024e-01
1.06070745e+00 -4.83840555e-01 -1.31355241e-01 -3.89007032e-01
-2.95008600e-01 2.31736466e-01 7.71647513e-01 -9.25708830e-01
3.38009715e-01 -4.14713286e-02 2.90158987e-01 -1.03744501e-02
-4.36497629e-02 -7.96782374e-01 3.77181411e-01 8.07838738e-01
-2.74897546e-01 5.43799698e-01 1.45340964e-01 4.25460003e-02
-3.22746128e-01 -2.82153301e-03 1.14556801e+00 -2.29431212e-01
-3.09980065e-01 9.68930423e-01 -4.05027479e-01 4.08937931e-01
1.10155272e+00 1.77887037e-01 -4.95869011e-01 -1.76932275e-01
-9.10257101e-01 1.41592667e-01 5.60526371e-01 3.43129963e-01
7.09516764e-01 -1.59360039e+00 -9.04168248e-01 2.79483885e-01
-8.13661218e-02 1.07547738e-01 2.82201171e-01 1.71107066e+00
-6.59422278e-01 -8.93557444e-03 -5.65493762e-01 -6.32285118e-01
-1.23221123e+00 1.34571522e-01 1.09980297e+00 -6.53232813e-01
-1.15417159e+00 3.90690476e-01 4.18288589e-01 -5.34249067e-01
-2.25688800e-01 -5.58818460e-01 -1.07655451e-01 -2.65459120e-01
3.35910201e-01 -5.68830855e-02 3.24683428e-01 -1.01146102e+00
-4.37908590e-01 9.19983804e-01 -7.39751682e-02 1.92966953e-01
1.31973505e+00 -7.01032877e-02 -4.32262391e-01 -2.48946935e-01
1.42852318e+00 -5.06836712e-01 -1.30548573e+00 -5.72257817e-01
-1.67373475e-02 -2.98651993e-01 2.69300073e-01 -5.56972563e-01
-2.09016418e+00 6.67548239e-01 1.00024891e+00 -1.90963149e-01
1.01339650e+00 1.43521100e-01 1.17336738e+00 -7.03861117e-02
2.38309652e-01 -4.77045029e-01 1.20372675e-01 4.27618951e-01
9.30370629e-01 -1.33246160e+00 -1.92081600e-01 -2.83471793e-01
-3.91532809e-01 1.26302350e+00 6.52220905e-01 -1.65509760e-01
7.61712313e-01 7.87469149e-01 2.73885489e-01 -4.39834893e-01
-2.97680736e-01 2.02216282e-01 3.41888309e-01 6.67613804e-01
4.53922123e-01 5.52651137e-02 -3.13726693e-01 2.70385653e-01
-1.26370713e-01 -1.33023754e-01 5.64428389e-01 7.63060868e-01
8.88610631e-02 -1.13104856e+00 -3.73138875e-01 3.18347543e-01
-4.83697265e-01 5.05042449e-02 -2.30237171e-01 9.61929560e-01
-5.18798828e-02 3.78702044e-01 3.42326730e-01 2.03310885e-02
2.83901900e-01 -2.85983264e-01 6.06590509e-01 -2.77888834e-01
-7.78032541e-01 3.77583764e-02 -4.87727433e-01 -7.09460437e-01
-5.11994600e-01 -9.02236044e-01 -1.46154022e+00 -1.33752003e-01
1.50728881e-01 -2.30631784e-01 5.35262465e-01 9.06720400e-01
1.46867841e-01 5.90815067e-01 5.29401362e-01 -8.43084574e-01
-2.82376885e-01 -5.96074820e-01 -4.79282290e-01 6.98196352e-01
3.71037513e-01 -5.90061963e-01 -2.96584755e-01 1.38480708e-01] | [13.983650207519531, -2.5773894786834717] |
59ee690d-657b-40f5-a231-ca46fae2cb86 | onion-peel-networks-for-deep-video-completion | 1908.08718 | null | https://arxiv.org/abs/1908.08718v1 | https://arxiv.org/pdf/1908.08718v1.pdf | Onion-Peel Networks for Deep Video Completion | We propose the onion-peel networks for video completion. Given a set of reference images and a target image with holes, our network fills the hole by referring the contents in the reference images. Our onion-peel network progressively fills the hole from the hole boundary enabling it to exploit richer contextual information for the missing regions every step. Given a sufficient number of recurrences, even a large hole can be inpainted successfully. To attend to the missing information visible in the reference images, we propose an asymmetric attention block that computes similarities between the hole boundary pixels in the target and the non-hole pixels in the references in a non-local manner. With our attention block, our network can have an unlimited spatial-temporal window size and fill the holes with globally coherent contents. In addition, our framework is applicable to the image completion guided by the reference images without any modification, which is difficult to do with the previous methods. We validate that our method produces visually pleasing image and video inpainting results in realistic test cases. | ['Joon-Young Lee', 'Seoung Wug Oh', 'Seon Joo Kim', 'Sungho Lee'] | 2019-08-23 | onion-peel-networks-for-deep-video-completion-1 | http://openaccess.thecvf.com/content_ICCV_2019/html/Oh_Onion-Peel_Networks_for_Deep_Video_Completion_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Oh_Onion-Peel_Networks_for_Deep_Video_Completion_ICCV_2019_paper.pdf | iccv-2019-10 | ['video-inpainting'] | ['computer-vision'] | [ 4.88966286e-01 3.12843651e-01 -6.47380427e-02 2.02914521e-01
-5.43078959e-01 -3.25054318e-01 -5.12833856e-02 -1.54144794e-01
-1.66405827e-01 8.59333694e-01 1.96395516e-01 5.25039174e-02
6.25749454e-02 -7.61984706e-01 -9.11345601e-01 -4.24085498e-01
1.44836083e-01 -5.21747656e-02 6.27767026e-01 -1.13574982e-01
2.48650178e-01 4.43459481e-01 -1.35631704e+00 4.02150571e-01
8.50407124e-01 9.69158888e-01 7.40980744e-01 5.91967702e-01
7.13087479e-03 8.57795060e-01 -3.62082630e-01 -1.32467330e-01
6.53681397e-01 -4.51751381e-01 -7.37824321e-01 5.17294765e-01
6.98348284e-01 -7.03177989e-01 -6.57703459e-01 1.01270044e+00
1.26831844e-01 2.75567323e-01 -1.45198759e-02 -1.02563202e+00
-6.19856119e-01 4.02261972e-01 -1.19980812e+00 1.20810285e-01
4.20655459e-01 1.41846851e-01 8.02616179e-01 -1.10930955e+00
1.07776201e+00 1.20485592e+00 6.41020298e-01 4.43092704e-01
-1.08672953e+00 -4.84821022e-01 4.10855949e-01 2.59152532e-01
-1.36633146e+00 -4.80469316e-01 9.73970830e-01 -3.93079706e-02
6.34103954e-01 3.44457775e-01 6.86620295e-01 4.68921363e-01
1.04919828e-01 5.14976859e-01 6.34430408e-01 -4.95199919e-01
8.60464722e-02 -3.35222065e-01 -3.41673017e-01 7.13157713e-01
4.80111577e-02 9.71299484e-02 -6.49489880e-01 2.96831697e-01
1.34819663e+00 4.14444774e-01 -7.90421486e-01 -3.89899284e-01
-1.18689013e+00 5.81640303e-01 6.23331189e-01 2.89842874e-01
-6.37273252e-01 3.52814287e-01 7.81082138e-02 1.73590645e-01
5.53468823e-01 2.79563010e-01 -3.29773158e-01 2.25856990e-01
-1.35322130e+00 -3.92796844e-02 4.57970738e-01 1.03322482e+00
1.05984163e+00 7.30096325e-02 -1.49892792e-01 6.93403363e-01
-4.62941639e-03 8.74280483e-02 -1.02986716e-01 -1.49602926e+00
7.53331363e-01 3.81869465e-01 6.07457697e-01 -1.32196736e+00
8.88780951e-02 -3.56285423e-01 -1.11650634e+00 4.28048968e-01
4.13729757e-01 -8.21694955e-02 -9.19383526e-01 1.74469316e+00
4.07160878e-01 3.57883334e-01 -1.14953913e-01 9.87429202e-01
6.01336062e-01 8.88479829e-01 -3.55550081e-01 -6.07222021e-01
1.11548781e+00 -1.40019631e+00 -8.12113881e-01 -5.02174735e-01
2.49695539e-01 -8.53379786e-01 9.50545430e-01 2.65958220e-01
-1.69764590e+00 -6.95079863e-01 -9.75559890e-01 -4.55632985e-01
3.72519523e-01 6.06911592e-02 1.90480560e-01 -2.18825359e-02
-1.29956603e+00 6.83105111e-01 -7.01453686e-01 -2.46819988e-01
4.02445704e-01 1.82012931e-01 -4.07011539e-01 -4.31629598e-01
-9.28979933e-01 6.42914593e-01 3.84507746e-01 3.39124799e-01
-9.21343327e-01 -7.87964106e-01 -9.12308812e-01 2.71249473e-01
6.36048853e-01 -6.71891868e-01 8.97953749e-01 -1.36967468e+00
-1.08187854e+00 5.40776610e-01 -5.22415519e-01 -4.06180263e-01
7.63449788e-01 -2.38618046e-01 5.26560955e-02 5.33797562e-01
3.61818880e-01 1.09414840e+00 1.10164464e+00 -1.42938983e+00
-7.04758823e-01 -7.02766180e-02 2.18168512e-01 4.65017676e-01
-7.99967945e-02 -2.17757240e-01 -9.31985021e-01 -9.49705899e-01
4.23519790e-01 -5.31849980e-01 -2.84450024e-01 6.10173643e-01
-2.92232096e-01 3.55931312e-01 1.15521491e+00 -1.05980635e+00
1.28979266e+00 -2.32379222e+00 2.95628488e-01 -3.16714942e-02
5.45977414e-01 1.09994277e-01 -3.32785636e-01 4.53030616e-01
-1.19661584e-01 9.72097963e-02 -4.53625649e-01 -5.34151196e-01
-5.77075958e-01 3.02532166e-01 -4.76275772e-01 3.56250018e-01
7.75076672e-02 7.80084312e-01 -9.78279352e-01 -4.59550798e-01
2.09479019e-01 5.18823445e-01 -8.37137818e-01 2.59701014e-01
-1.59680203e-01 6.36124074e-01 -2.78072059e-01 4.91744339e-01
9.54263806e-01 -2.67946273e-01 -2.68290397e-02 -3.78145546e-01
-2.86387563e-01 -7.61848018e-02 -1.19493771e+00 1.95061409e+00
-4.87295508e-01 8.88313591e-01 5.30814469e-01 -5.84189713e-01
7.81644344e-01 9.70511436e-02 4.38593626e-01 -5.59566796e-01
-1.90723762e-01 1.65622428e-01 -1.63830936e-01 -3.96545768e-01
7.75948286e-01 6.04929626e-02 4.53743607e-01 4.47879851e-01
-3.00503075e-01 2.25331658e-03 2.36780703e-01 2.19452798e-01
9.80387688e-01 2.30710115e-02 1.10617459e-01 9.86837000e-02
5.23301303e-01 -3.45423162e-01 7.13377953e-01 7.69487977e-01
-2.30162162e-02 1.12626445e+00 3.39916080e-01 -6.63578093e-01
-1.32547510e+00 -1.01806414e+00 3.02856117e-01 7.91407883e-01
6.14830256e-01 -3.81991625e-01 -6.31004751e-01 -3.01309735e-01
-4.82726038e-01 2.10700631e-01 -6.29281938e-01 1.33074775e-01
-7.24301875e-01 1.14051498e-01 -1.28987432e-01 4.41666812e-01
9.48526978e-01 -1.25451863e+00 -7.80322373e-01 4.03179228e-01
-5.22445679e-01 -1.03307712e+00 -1.07776546e+00 -1.20755814e-01
-1.07733190e+00 -1.17379892e+00 -9.95650232e-01 -1.15173197e+00
1.06500816e+00 6.91192150e-01 9.09412861e-01 5.41411817e-01
-1.24744460e-01 2.98536886e-02 -2.25616485e-01 4.35124427e-01
-2.93193907e-01 -2.25704163e-01 -4.88656253e-01 1.15058161e-01
-4.33648914e-01 -7.57922053e-01 -1.01040685e+00 4.93909419e-01
-1.16995883e+00 5.53579986e-01 3.18063706e-01 1.01641965e+00
7.06571460e-01 2.40119249e-01 2.64907122e-01 -4.88436788e-01
2.74175256e-01 -2.68456668e-01 -4.78607476e-01 1.18986748e-01
-6.63023144e-02 -1.61884278e-01 5.18933117e-01 -4.53624934e-01
-1.00064647e+00 1.58645641e-02 9.54865664e-02 -9.73264456e-01
2.40036190e-01 2.21165285e-01 -1.62812933e-01 -4.46822010e-02
3.80383879e-01 1.71004921e-01 -8.43499526e-02 -4.02922839e-01
4.68054861e-01 1.96553916e-01 7.23694086e-01 -3.25250804e-01
6.74416959e-01 6.95472300e-01 -8.31904337e-02 -5.32739162e-01
-5.73162913e-01 -2.42176205e-01 -5.87851763e-01 -1.43651500e-01
7.41884887e-01 -8.41454685e-01 -4.83647823e-01 3.66790414e-01
-1.63028276e+00 -4.08744752e-01 -5.82486808e-01 1.07759997e-01
-4.24701124e-01 8.56116533e-01 -8.54755878e-01 -5.06934464e-01
-3.22360635e-01 -1.21970069e+00 9.69964683e-01 2.22138196e-01
-1.36851907e-01 -8.19676399e-01 -2.13420540e-01 7.54283145e-02
1.94175392e-01 9.37546268e-02 7.65847683e-01 1.60636440e-01
-1.18084228e+00 1.43667728e-01 -5.46513200e-01 2.68705904e-01
2.08891466e-01 -8.97303224e-02 -5.71685731e-01 -4.84896749e-01
3.92345786e-02 1.27804955e-03 1.02245164e+00 4.53456730e-01
1.10297441e+00 -6.54728711e-01 -2.65346706e-01 6.55524254e-01
1.54279864e+00 3.52103382e-01 1.03229213e+00 2.15942591e-01
7.61063397e-01 5.02704620e-01 5.12010098e-01 3.97226572e-01
2.35209689e-02 5.98180652e-01 7.13520169e-01 -4.43304092e-01
-3.58176261e-01 -5.04237652e-01 1.42271116e-01 5.54904759e-01
-4.92777787e-02 -4.20517087e-01 -5.26380360e-01 9.04861748e-01
-1.87522197e+00 -1.01880443e+00 -4.97357212e-02 2.31661129e+00
8.23470771e-01 -1.51491493e-01 -3.81983489e-01 4.24841307e-02
1.07151604e+00 4.96239394e-01 -6.30136073e-01 -2.39681885e-01
-8.36727768e-02 9.60574746e-02 2.84832358e-01 9.44243252e-01
-7.28254497e-01 9.83418167e-01 6.07196379e+00 9.55482900e-01
-8.79537880e-01 2.04077050e-01 8.15748990e-01 -1.45075858e-01
-5.11954248e-01 2.26062268e-01 -4.34368670e-01 4.12709564e-01
1.47334307e-01 1.64286107e-01 6.55643404e-01 4.97458190e-01
4.24963862e-01 -7.09915876e-01 -1.08965182e+00 1.13245356e+00
1.60900086e-01 -1.72236097e+00 1.09273873e-01 -1.86256409e-01
1.16409206e+00 -3.16777289e-01 4.16156016e-02 -2.29735866e-01
-1.10741414e-01 -9.02976036e-01 7.81944275e-01 3.22311729e-01
9.88956511e-01 -7.67191172e-01 4.70099241e-01 3.06157261e-01
-1.22691464e+00 -1.23398513e-01 -4.21188504e-01 -2.68405974e-01
4.51238483e-01 5.65938771e-01 -4.23417866e-01 4.86162364e-01
6.67574763e-01 8.28382313e-01 -2.80344129e-01 1.16911256e+00
-2.23294452e-01 7.53505668e-03 -2.33154610e-01 7.05082953e-01
2.07375839e-01 -4.40306872e-01 5.97766161e-01 6.07284307e-01
6.32381022e-01 1.44316614e-01 7.81855509e-02 1.10442865e+00
-2.91875452e-01 -3.96920852e-02 -6.17709339e-01 4.04006541e-01
4.91194963e-01 1.02879047e+00 -8.32833350e-01 -4.05160874e-01
-4.51187104e-01 1.40453005e+00 2.93051988e-01 6.95246994e-01
-7.89910018e-01 -3.55908722e-01 4.05399114e-01 3.65595758e-01
6.04999244e-01 -5.34959324e-02 -3.04649144e-01 -1.10561430e+00
3.19266200e-01 -6.94341063e-01 1.24855042e-01 -1.32253563e+00
-9.40916657e-01 7.02580273e-01 -1.64796501e-01 -1.38647878e+00
4.74871993e-02 1.70843676e-02 -7.80340075e-01 9.04776394e-01
-1.53428733e+00 -9.18663979e-01 -6.01199806e-01 6.63071990e-01
8.67021978e-01 2.63472766e-01 3.11212540e-01 3.06172997e-01
-2.41208360e-01 2.66102344e-01 -1.06762037e-01 7.16920942e-02
5.10273218e-01 -5.12666404e-01 1.69523790e-01 1.17976081e+00
-1.48256421e-01 3.38639587e-01 5.82959175e-01 -8.81725371e-01
-8.63907874e-01 -1.11184132e+00 8.13414574e-01 1.70817316e-01
3.80293220e-01 -8.61793607e-02 -1.00215709e+00 7.51516581e-01
5.77547073e-01 1.07192419e-01 -1.75311252e-01 -5.39518416e-01
-2.10110575e-01 -8.85810703e-02 -1.06446445e+00 8.68355393e-01
1.17022622e+00 -3.62819254e-01 -4.28872883e-01 2.09812626e-01
8.65185022e-01 -6.94270611e-01 -4.06168014e-01 2.58727312e-01
2.97347784e-01 -1.13600147e+00 1.02167177e+00 -2.61189640e-01
9.58311975e-01 -7.06255198e-01 5.05909659e-02 -1.06977355e+00
-1.70830205e-01 -1.02155280e+00 4.27835397e-02 1.03070152e+00
9.13186520e-02 -3.07509989e-01 8.86056662e-01 5.37757814e-01
-2.71019667e-01 -6.99759483e-01 -1.02894199e+00 -3.43590707e-01
-3.04959893e-01 -3.65560293e-01 3.76217037e-01 5.83738923e-01
-2.98463285e-01 -9.76698175e-02 -8.55431080e-01 2.51318991e-01
5.40911376e-01 3.34762573e-01 6.51117742e-01 -7.78992951e-01
-3.31397086e-01 -1.15992904e-01 -4.76020277e-02 -1.66597939e+00
-6.04751967e-02 -5.25164127e-01 9.56150517e-02 -1.57047749e+00
3.65175128e-01 -3.43054324e-01 1.39532059e-01 6.15018547e-01
-1.74804047e-01 6.28869057e-01 5.32742679e-01 2.93680757e-01
-5.27193844e-01 5.44493973e-01 1.80456114e+00 -1.31464705e-01
-4.76182312e-01 -3.14817816e-01 -5.32474935e-01 7.77345598e-01
6.76035762e-01 -4.20968682e-01 -5.74525177e-01 -6.10254169e-01
1.95965528e-01 6.39711797e-01 5.38060486e-01 -9.23797607e-01
3.08732629e-01 -1.29862025e-01 5.79016685e-01 -8.54486585e-01
4.86773074e-01 -8.76998246e-01 5.61583340e-01 2.95281976e-01
-2.35547140e-01 1.67584017e-01 2.16492832e-01 6.29835606e-01
-3.17416817e-01 -4.11506295e-01 8.41376543e-01 -3.66644382e-01
-5.78487813e-01 4.81953233e-01 -2.01316148e-01 1.34395063e-01
1.02804792e+00 -7.14493632e-01 -1.91796809e-01 -6.36305451e-01
-9.53851998e-01 2.60182083e-01 9.41929102e-01 2.23077670e-01
1.04775202e+00 -1.30881345e+00 -3.99099588e-01 5.20899117e-01
-3.39483202e-01 3.93459439e-01 7.51319408e-01 7.83770442e-01
-8.77448440e-01 1.14572883e-01 -4.25813913e-01 -5.62929034e-01
-1.21971548e+00 8.44131172e-01 3.84892523e-01 -3.02441329e-01
-8.36242080e-01 6.17756307e-01 8.78308117e-01 3.71025890e-01
3.29836905e-01 -4.19632077e-01 9.02535766e-02 -1.55341670e-01
6.94670618e-01 2.72563249e-01 -2.97389448e-01 -4.73272294e-01
9.29006189e-02 7.79596508e-01 -1.99550554e-01 -2.73591012e-01
1.31607938e+00 -5.31749070e-01 -3.02630991e-01 2.18768101e-02
1.09897578e+00 2.95269459e-01 -1.90771914e+00 -1.56132534e-01
-5.31317890e-01 -9.54636931e-01 -2.93543655e-02 -4.06704456e-01
-1.60314238e+00 7.74368942e-01 2.02718243e-01 -9.05401111e-02
1.51715553e+00 -2.40005478e-01 1.03523171e+00 -1.58411711e-01
4.00154680e-01 -8.12585235e-01 3.80018979e-01 2.82902271e-01
1.20462430e+00 -8.47225130e-01 -6.55760244e-02 -7.58545160e-01
-3.63361895e-01 1.15773857e+00 6.51409864e-01 -1.72504410e-01
2.45498687e-01 1.45487577e-01 -2.73978472e-01 7.80045614e-02
-6.84684455e-01 2.19917111e-02 1.74769629e-02 6.45615220e-01
5.49704619e-02 -5.00984251e-01 -3.66806090e-01 5.79667166e-02
3.84720892e-01 1.39447570e-01 8.58494103e-01 6.73683941e-01
-4.28108424e-01 -1.06802571e+00 -6.10391736e-01 4.90461402e-02
-2.24529162e-01 -3.43838304e-01 -3.89833376e-02 8.11211646e-01
2.36570328e-01 8.76324952e-01 2.91702062e-01 -2.11439859e-02
1.21617101e-01 -3.12554926e-01 4.45832014e-01 -5.95937312e-01
-2.42779702e-01 4.33569759e-01 -1.42568320e-01 -8.04460585e-01
-4.24053371e-01 -2.81798393e-01 -1.31140769e+00 -3.46304417e-01
-5.44157661e-02 -5.99244162e-02 -1.44027382e-01 4.54734087e-01
5.56409836e-01 5.22516489e-01 5.45487463e-01 -1.13989496e+00
1.36078164e-01 -7.21747339e-01 -5.85258543e-01 4.17398781e-01
5.33818364e-01 -4.19631094e-01 -4.14460033e-01 2.28468820e-01] | [10.803101539611816, -1.3335800170898438] |
aa91eda6-5ddb-4613-91de-8f5d783b49a0 | covid-widenet-a-capsule-network-for-covid-19 | null | null | https://www.sciencedirect.com/science/article/pii/S1568494622002046 | https://www.sciencedirect.com/science/article/pii/S1568494622002046 | COVID-WideNet—A capsule network for COVID-19 detection | Ever since the outbreak of COVID-19, the entire world is grappling with panic over its rapid spread. Consequently, it is of utmost importance to detect its presence. Timely diagnostic testing leads to the quick identification, treatment and isolation of infected people. A number of deep learning classifiers have been proved to provide encouraging results with higher accuracy as compared to the conventional method of RT-PCR testing. Chest radiography, particularly for using X-ray images, is a prime imaging modality for detecting the suspected COVID-19 patients. However, the performance of these approaches still needs to be improved. In this paper, we propose a capsule network called COVID-WideNet for diagnosing COVID-19 cases using Chest X-ray (CXR) images. Experimental results have demonstrated that a discriminative trained, multi-layer capsule network achieves state-of-the-art performance on the COVIDx dataset. In particular, COVID-WideNet performs better than any other CNN based approaches for the diagnosis of COVID-19 infected patients. Further, the proposed COVID-WideNet has the number of trainable parameters that is 20 times less than that of other CNN based models. This results in a fast and efficient diagnosing COVID-19 symptoms, and with achieving the 0.95 of Area Under Curve (AUC), 91% of accuracy, sensitivity and specificity, respectively. This may also assist radiologists to detect COVID and its variant like delta. | ['Harsh Panwar'] | 2022-03-22 | null | null | null | applied-soft-computing-2022-3 | ['covid-19-detection'] | ['medical'] | [-2.35756695e-01 -5.02688766e-01 -4.80523258e-02 1.08861074e-01
-3.61563981e-01 -4.85318869e-01 2.70489514e-01 2.11802408e-01
-5.10919750e-01 7.48016775e-01 -1.95364296e-01 -5.26994169e-01
-2.13617697e-01 -6.20119154e-01 -2.16617107e-01 -8.94219041e-01
-4.27079529e-01 8.99110317e-01 1.18154883e-01 3.15354526e-01
-3.53792757e-01 6.98194683e-01 -8.13524365e-01 1.77740589e-01
7.61948586e-01 9.69911814e-01 5.31289399e-01 8.81220222e-01
2.67292142e-01 6.07824504e-01 -5.71787417e-01 -2.03746110e-01
-4.48766910e-02 -1.62905216e-01 -5.96879005e-01 -2.67810732e-01
-1.61818445e-01 -7.57626653e-01 -1.21691957e-01 5.19061983e-01
4.68009561e-01 -2.98242986e-01 8.67809653e-01 -9.44660783e-01
-4.51997668e-01 1.05047613e-01 -5.17684579e-01 6.86208427e-01
8.61388631e-03 1.84016809e-01 3.93955082e-01 -7.11382926e-01
5.31651199e-01 8.54583442e-01 9.75848258e-01 4.97027785e-01
-5.48873127e-01 -8.37602019e-01 -3.20365995e-01 1.74282223e-01
-1.44052696e+00 6.27718329e-01 -3.54010873e-02 -7.00827539e-01
1.13019395e+00 1.06958233e-01 1.14842892e+00 1.32543159e+00
7.35984921e-01 5.03334701e-01 9.33121800e-01 1.11825667e-01
-7.94029459e-02 -2.33541075e-02 1.24505669e-01 8.46712589e-01
8.35290372e-01 -2.67071482e-02 1.95973828e-01 -3.38938475e-01
9.70917284e-01 7.69132257e-01 -3.55899423e-01 1.31030753e-01
-1.20028961e+00 1.03825569e+00 5.63275635e-01 6.02018714e-01
-4.82237637e-01 7.34517649e-02 6.94296241e-01 -1.75665170e-01
2.49009341e-01 3.73074651e-01 -4.73318934e-01 1.19608499e-01
-6.14334702e-01 2.22844318e-01 4.03149486e-01 3.11952949e-01
-1.61437094e-01 5.08738048e-02 -2.48494372e-01 6.38028860e-01
3.11988860e-01 9.00298238e-01 5.48852623e-01 -2.49552071e-01
1.64011791e-01 6.50082946e-01 7.11336583e-02 -1.06885588e+00
-7.76751518e-01 -8.35018992e-01 -1.36056232e+00 -3.49356323e-01
1.28403902e-01 -3.79659027e-01 -1.24283886e+00 1.39416301e+00
1.17953591e-01 3.99892271e-01 -8.11605155e-02 9.56783772e-01
7.99597681e-01 7.09440589e-01 8.83737728e-02 -1.39538839e-01
1.88121819e+00 -6.68964326e-01 -6.67525470e-01 -8.78242217e-03
3.98646891e-01 -7.06207156e-01 4.02207881e-01 5.37203670e-01
-4.16484237e-01 -3.51084083e-01 -1.09958720e+00 5.15162170e-01
-3.95539314e-01 3.28095585e-01 8.27697873e-01 6.66641414e-01
-8.43621731e-01 2.14437693e-01 -1.02127957e+00 -6.80938303e-01
4.37417865e-01 5.26481569e-01 -4.06571627e-01 -3.21994394e-01
-1.18351042e+00 9.80201840e-01 4.32778478e-01 3.34920198e-01
-1.19856095e+00 -5.56051254e-01 -5.21817625e-01 -3.56298946e-02
2.09120378e-01 -9.76225436e-01 9.76950526e-01 -6.52402490e-02
-6.15314364e-01 7.41870761e-01 6.62801862e-02 -3.99544328e-01
4.14139479e-01 -2.28679895e-01 -4.54097122e-01 3.72124106e-01
9.56371725e-02 3.26757431e-01 6.40264094e-01 -9.40308630e-01
-6.59584165e-01 -4.32420850e-01 -1.30456761e-01 -3.69911909e-01
-4.86271568e-02 3.48272562e-01 -4.84034717e-01 -5.37810028e-01
-2.13920131e-01 -1.00559735e+00 -2.99103558e-01 -1.54117808e-01
-3.98451865e-01 -4.91720438e-01 1.14518571e+00 -8.00605714e-01
8.40334594e-01 -1.89660370e+00 -5.41934073e-01 2.29444697e-01
8.20277095e-01 1.03784573e+00 2.69202769e-01 2.66324729e-01
3.47663648e-02 2.85340130e-01 -1.16297089e-01 6.45069480e-02
-4.58086759e-01 1.62861109e-01 7.41065964e-02 7.11961806e-01
5.86207032e-01 8.98272574e-01 -8.21325064e-01 -6.04723275e-01
2.97581047e-01 1.03584802e+00 -2.59915173e-01 4.99707967e-01
6.66059703e-02 6.05150461e-01 -7.78052986e-01 8.35263312e-01
7.48931706e-01 -1.06198585e+00 3.28863382e-01 7.58034140e-02
3.60286266e-01 -3.55762690e-01 -7.24010170e-01 8.70543420e-01
-1.07877813e-01 6.89376235e-01 -1.40567152e-02 -7.94060469e-01
6.75204337e-01 8.05805326e-01 4.13687885e-01 -4.31108028e-01
6.11491740e-01 2.36890316e-01 1.96588904e-01 -8.74349713e-01
-2.72784214e-02 -1.66891009e-01 2.51458883e-01 2.90327162e-01
-3.35343748e-01 3.88411790e-01 1.58671327e-02 6.40115142e-02
1.10323834e+00 -4.18026567e-01 4.93305057e-01 -5.41063920e-02
5.35030842e-01 1.69376880e-01 4.52790201e-01 7.17424035e-01
-3.18778694e-01 5.77088237e-01 2.10465729e-01 -6.19915485e-01
-8.93259525e-01 -1.11707163e+00 -5.64140320e-01 1.79488048e-01
-1.76241606e-01 7.70605803e-02 -6.19893134e-01 -3.46498579e-01
6.19161949e-02 -1.33996038e-02 -8.14245403e-01 1.21296100e-01
-7.83089459e-01 -9.37734008e-01 8.56213391e-01 8.84782195e-01
6.52852595e-01 -8.94246578e-01 -8.94101501e-01 3.01833808e-01
-2.05174327e-01 -1.13562298e+00 -1.06426449e-02 2.65488356e-01
-8.50533068e-01 -1.49162316e+00 -1.09059215e+00 -8.94016266e-01
5.27162910e-01 2.25961775e-01 7.86985874e-01 4.49356914e-01
-8.08209360e-01 -1.06345331e-02 -3.25899243e-01 -5.22012174e-01
-1.59867406e-01 -1.97231732e-02 4.13098410e-02 -4.01513875e-01
5.08107960e-01 -2.42771637e-02 -9.44804251e-01 5.81621714e-02
-7.37669051e-01 -1.52528867e-01 6.98736370e-01 7.86418974e-01
5.61318994e-01 -1.39040649e-01 6.11335039e-01 -7.04374671e-01
6.58264279e-01 -7.90108740e-01 -2.18125477e-01 1.27714932e-01
-5.55115521e-01 -6.31500363e-01 7.78716207e-01 -3.93198818e-01
-7.95026124e-01 -3.61128271e-01 -1.19281992e-01 -5.98849237e-01
-1.02402024e-01 5.16680598e-01 7.44360328e-01 2.48281598e-01
3.29975694e-01 1.33268818e-01 7.80859739e-02 -3.21887970e-01
-3.49495977e-01 7.90642023e-01 5.42067170e-01 -1.03954419e-01
5.42181075e-01 6.09898448e-01 2.39719916e-02 -8.04782748e-01
-6.45302176e-01 -9.09180939e-01 -2.67195582e-01 -2.46640429e-01
1.58365870e+00 -1.06096292e+00 -1.03027558e+00 5.94988823e-01
-1.22441626e+00 2.47305706e-01 5.58837593e-01 8.56172442e-01
3.72090749e-02 2.73119837e-01 -9.13257897e-01 -6.97601199e-01
-8.56849074e-01 -1.43862689e+00 9.18634355e-01 2.78742790e-01
-1.86087966e-01 -1.15887201e+00 2.08921328e-01 2.85503030e-01
5.99072874e-01 6.73422575e-01 1.02932131e+00 -1.02425575e+00
-6.45635247e-01 -5.80147266e-01 -6.59444392e-01 3.06388021e-01
1.72282979e-01 5.68086915e-02 -8.85693371e-01 -5.56275666e-01
7.05987066e-02 -3.39850992e-01 8.22459817e-01 6.39082551e-01
1.07320189e+00 9.59776565e-02 -7.94745207e-01 8.66829813e-01
1.68036580e+00 8.71807158e-01 3.70509118e-01 2.07898602e-01
6.01938188e-01 -5.35560735e-02 4.29508954e-01 5.61329067e-01
1.18363321e-01 2.06827253e-01 5.80848396e-01 -4.81420994e-01
2.16141447e-01 2.81261593e-01 -2.73118138e-01 7.57909954e-01
-4.08417791e-01 -5.18985808e-01 -1.20389581e+00 6.56609774e-01
-1.40952992e+00 -7.50592947e-01 -5.84156275e-01 1.56424916e+00
3.96345675e-01 -2.40958884e-01 -4.64255214e-02 -3.89796570e-02
8.08028340e-01 -2.60536283e-01 -4.12982345e-01 -4.31501687e-01
1.16583407e-01 4.68138367e-01 2.75112242e-01 -6.37555122e-02
-1.27074051e+00 2.58886606e-01 6.75773239e+00 4.11952764e-01
-1.52238250e+00 3.45578521e-01 6.39155388e-01 2.57438272e-01
2.86869019e-01 -6.90165102e-01 -7.19137669e-01 4.18956846e-01
6.16268098e-01 2.01715142e-01 -1.05920866e-01 8.83532703e-01
1.67462215e-01 4.42270897e-02 -7.17959225e-01 8.76323938e-01
1.15633480e-01 -1.44541895e+00 -2.33627275e-01 1.25681430e-01
5.59334993e-01 3.23751152e-01 -3.80748250e-02 2.77113587e-01
1.60290405e-01 -1.23490882e+00 -3.63783650e-02 1.22630619e-01
1.00116348e+00 -7.99631178e-01 1.58078897e+00 1.39951751e-01
-1.36398733e+00 2.61834636e-02 -3.55564952e-01 2.73153216e-01
4.65805903e-02 5.24659872e-01 -1.56790280e+00 3.63947839e-01
9.38494146e-01 3.63479346e-01 -3.12724322e-01 1.18282008e+00
-7.68118575e-02 6.38219237e-01 -2.32236207e-01 -3.33807260e-01
5.40386856e-01 5.67952432e-02 1.59281895e-01 1.61434591e+00
3.50360781e-01 2.50523150e-01 2.43978694e-01 6.80245638e-01
2.10806280e-01 1.17722422e-01 -7.19689012e-01 -1.34901419e-01
1.91552550e-01 1.12162936e+00 -1.01763237e+00 -5.97271442e-01
-3.20156693e-01 5.78790605e-01 -3.58583741e-02 1.72476128e-01
-1.17906737e+00 -1.64976835e-01 5.48663020e-01 -2.51251347e-02
6.24518335e-01 7.20831752e-02 -1.53662730e-02 -8.58558059e-01
-2.56753087e-01 -7.77642429e-01 6.08156264e-01 -6.70263350e-01
-1.24470556e+00 8.49365056e-01 -3.42706107e-02 -1.08767843e+00
-3.32320869e-01 -9.83016193e-01 -6.61022127e-01 8.77739966e-01
-1.50094724e+00 -1.04312396e+00 -6.19038939e-01 5.87290525e-01
3.56727481e-01 -1.61388770e-01 1.03898549e+00 1.32325456e-01
-6.00665092e-01 3.64676118e-01 3.53133231e-01 4.37608927e-01
4.54516500e-01 -9.54355955e-01 -7.61822506e-04 7.17152476e-01
-8.89632523e-01 1.05982745e+00 3.24234992e-01 -8.85898411e-01
-1.16941011e+00 -1.31580997e+00 8.23337138e-01 -1.27172858e-01
4.05113131e-01 -4.97631133e-02 -7.91111410e-01 6.94598198e-01
3.05599153e-01 -1.12058863e-01 8.07441652e-01 -2.55015761e-01
-2.77885705e-01 1.38616621e-01 -1.35326183e+00 2.13211536e-01
4.55471814e-01 -3.45387220e-01 -5.99648595e-01 4.69372571e-01
7.29180574e-01 -2.65167058e-01 -1.02235246e+00 6.22685611e-01
6.48396969e-01 -8.97196829e-01 1.13496232e+00 -5.41348636e-01
3.54965121e-01 -1.36250660e-01 6.49775267e-02 -8.12521756e-01
-4.66975421e-01 7.76217431e-02 1.02354921e-01 4.49154884e-01
6.83469847e-02 -7.88166642e-01 6.28818035e-01 -1.38266400e-01
1.52162211e-02 -1.19700956e+00 -6.13539040e-01 -5.95819890e-01
-2.60004371e-01 -2.81323969e-01 4.05083776e-01 1.15800869e+00
-4.03158933e-01 6.69446364e-02 -3.04891914e-01 1.98217392e-01
5.08763731e-01 -2.32048128e-02 3.95915002e-01 -1.28972828e+00
-4.12671119e-01 -2.39395455e-01 -4.84079123e-01 -3.93511951e-01
-5.91253221e-01 -7.42612958e-01 -3.08807671e-01 -1.95407403e+00
6.71726286e-01 -3.19669783e-01 -4.61303294e-01 3.84790301e-01
-3.39133859e-01 4.57453698e-01 6.22279271e-02 3.28936607e-01
-2.99675435e-01 -1.46619037e-01 1.57576716e+00 -1.06290765e-01
1.14862911e-01 -7.84373749e-03 -3.09862405e-01 5.97309172e-01
9.31252897e-01 -5.05612552e-01 -3.34051788e-01 -3.53639752e-01
9.08402726e-02 2.67838717e-01 3.03126067e-01 -8.85212719e-01
-8.84202048e-02 -8.01327601e-02 8.20234537e-01 -1.18767595e+00
3.20896298e-01 -6.61709726e-01 3.11599106e-01 1.35590339e+00
3.72195035e-01 3.70052189e-01 2.28160411e-01 4.58692193e-01
-1.75432131e-01 -3.95213105e-02 8.19329560e-01 -3.21931154e-01
-4.93886232e-01 4.15467530e-01 -9.24519658e-01 -4.49258238e-02
1.21451068e+00 -1.84048250e-01 -4.88212883e-01 1.21198095e-01
-3.95009071e-01 1.05121613e-01 1.95195545e-02 2.58145392e-01
9.26686227e-01 -9.27910030e-01 -7.40965366e-01 2.79094487e-01
1.42380774e-01 2.73920357e-01 3.91649812e-01 1.21016061e+00
-1.32653689e+00 1.02835321e+00 -2.13126317e-01 -1.07281899e+00
-1.35832047e+00 6.89477205e-01 4.76126343e-01 -6.72316730e-01
-7.35627353e-01 8.96651030e-01 3.87957126e-01 -9.67545658e-02
3.27197462e-01 -4.95589525e-01 -3.96700680e-01 -1.10731080e-01
6.95621908e-01 1.58633992e-01 1.33234441e-01 -2.50191987e-01
-7.48443842e-01 5.33670545e-01 -3.28463644e-01 6.26852572e-01
1.36632228e+00 4.73667204e-01 -2.62037724e-01 4.55661528e-02
1.11550558e+00 -4.03495401e-01 -5.33091187e-01 7.30423480e-02
-4.39140916e-01 -2.96227545e-01 -1.59374177e-01 -1.01725030e+00
-1.06157506e+00 8.78463089e-01 9.83208299e-01 8.64289626e-02
9.83540177e-01 2.46126369e-01 8.61851633e-01 3.10456604e-01
2.47589573e-01 -4.61964905e-01 1.90716773e-01 3.88552189e-01
4.65265810e-01 -1.30535376e+00 -8.63516778e-02 -2.83853650e-01
-3.08573186e-01 1.00951731e+00 5.91378927e-01 -2.58031398e-01
6.54424667e-01 4.28007424e-01 3.09231311e-01 -6.45628572e-01
-6.83345556e-01 7.02294782e-02 4.32801805e-02 9.15615797e-01
5.84248960e-01 5.48695564e-01 -2.63762325e-01 4.77622747e-01
1.46651387e-01 5.89352159e-04 1.55569330e-01 7.70716488e-01
-4.51990992e-01 -6.80103481e-01 -6.97637856e-01 8.24480355e-01
-9.36655223e-01 -4.56241220e-02 -2.13505313e-01 1.30816054e+00
3.40187371e-01 8.23624134e-01 1.34209216e-01 -3.12586784e-01
-5.18394969e-02 -1.13377266e-01 2.51211345e-01 -3.74292493e-01
-7.34079123e-01 1.52428135e-01 -4.70522605e-02 -2.49133125e-01
-3.72673035e-01 -3.97222996e-01 -1.47251570e+00 -4.94397998e-01
-4.45679188e-01 -1.85987223e-02 7.22882152e-01 9.36899900e-01
2.32893545e-02 7.80704200e-01 3.53965640e-01 -2.03096062e-01
-4.13281649e-01 -9.76043880e-01 -5.54983497e-01 2.27141634e-01
4.70148683e-01 -6.14814162e-01 -9.90041047e-02 -2.82943517e-01] | [15.564651489257812, -1.70636785030365] |
87f4d953-e61e-48a7-a609-4a3c8cbcdcf6 | efficient-cavity-searching-for-gene-network | 2211.02935 | null | https://arxiv.org/abs/2211.02935v1 | https://arxiv.org/pdf/2211.02935v1.pdf | Efficient Cavity Searching for Gene Network of Influenza A Virus | High order structures (cavities and cliques) of the gene network of influenza A virus reveal tight associations among viruses during evolution and are key signals that indicate viral cross-species infection and cause pandemics. As indicators for sensing the dynamic changes of viral genes, these higher order structures have been the focus of attention in the field of virology. However, the size of the viral gene network is usually huge, and searching these structures in the networks introduces unacceptable delay. To mitigate this issue, in this paper, we propose a simple-yet-effective model named HyperSearch based on deep learning to search cavities in a computable complex network for influenza virus genetics. Extensive experiments conducted on a public influenza virus dataset demonstrate the effectiveness of HyperSearch over other advanced deep-learning methods without any elaborated model crafting. Moreover, HyperSearch can finish the search works in minutes while 0-1 programming takes days. Since the proposed method is simple and easy to be transferred to other complex networks, HyperSearch has the potential to facilitate the monitoring of dynamic changes in viral genes and help humans keep up with the pace of virus mutations. | ['Zheng Kou', 'Xinyue Fan', 'Yaohua Liu', 'Jiahao Shen', 'Yanqing Su', 'Jietong Zhao', 'Junjie Li'] | 2022-11-05 | null | null | null | null | ['virology'] | ['miscellaneous'] | [ 1.78959835e-02 -2.28471488e-01 -1.95853531e-01 -1.16927788e-01
3.48822951e-01 -6.46710813e-01 1.06381148e-01 5.59884571e-02
-2.04075441e-01 7.02651083e-01 -3.40337574e-01 -5.69528341e-01
-2.62903184e-01 -9.76396680e-01 -7.56348729e-01 -8.67602170e-01
-6.71525776e-01 8.78992379e-01 1.89912133e-02 -5.70745647e-01
1.55714199e-01 6.30218625e-01 -1.13602686e+00 2.46816695e-01
9.64490175e-01 5.37883997e-01 2.13295147e-01 5.77027321e-01
-2.20081180e-01 2.56271452e-01 -5.00989377e-01 -2.13754065e-02
1.13694631e-01 -2.39294812e-01 -6.10274851e-01 -1.84978142e-01
-1.55241996e-01 1.35970533e-01 -1.40936017e-01 9.46929395e-01
6.20455086e-01 -2.72240847e-01 3.81638974e-01 -1.29935455e+00
-2.73620844e-01 2.14313835e-01 -5.87972045e-01 4.21035796e-01
1.23024084e-01 2.20418140e-01 1.03014779e+00 -6.86362803e-01
8.67491603e-01 1.26669347e+00 1.04314649e+00 2.89492130e-01
-1.16198993e+00 -7.37414181e-01 -6.55308962e-02 1.70997620e-01
-1.48225987e+00 -1.43484086e-01 6.32151008e-01 -6.70163691e-01
1.18606806e+00 2.76568860e-01 1.04862547e+00 7.72988856e-01
2.73357064e-01 4.07513052e-01 7.97620237e-01 1.64575994e-01
3.82027365e-02 -3.13756108e-01 -1.43201083e-01 1.39622903e+00
2.84486473e-01 7.15271831e-02 -1.70550421e-01 -7.41174698e-01
6.88233495e-01 5.40724874e-01 -3.97443116e-01 -2.39165843e-01
-1.00506043e+00 9.94605660e-01 7.30806649e-01 3.62362653e-01
-4.09335345e-01 -8.11058357e-02 4.41661030e-01 4.18448746e-01
3.79739791e-01 5.12816727e-01 -6.95765734e-01 2.98701245e-02
-6.37310743e-01 -4.75469679e-02 9.97108340e-01 3.66815120e-01
9.25563872e-01 -1.59354076e-01 3.19770515e-01 7.40126967e-01
2.94368118e-01 3.10695440e-01 -1.21010639e-01 -3.79095376e-01
1.12735286e-01 1.12409365e+00 -2.31631204e-01 -1.49238670e+00
-1.02116370e+00 -5.54871202e-01 -1.52486515e+00 -3.09792548e-01
2.13128909e-01 -3.42447549e-01 -7.48863757e-01 1.60130513e+00
7.03811467e-01 3.43061566e-01 -3.93828660e-01 6.66663766e-01
5.73709488e-01 9.89465237e-01 -2.94502318e-01 -4.70804334e-01
1.30990803e+00 -6.95156872e-01 -4.32838678e-01 2.51664042e-01
6.33817315e-01 -3.06991369e-01 8.22618246e-01 2.26409718e-01
-5.95113158e-01 -2.06724331e-01 -7.82254457e-01 4.22817767e-01
-3.87798101e-01 -4.21814650e-01 8.31673324e-01 4.96338576e-01
-1.28421915e+00 6.36346936e-01 -7.62962162e-01 -4.69601572e-01
5.76376379e-01 7.73840547e-01 -3.30236107e-02 8.70131180e-02
-1.43925905e+00 4.73636091e-01 3.40933383e-01 5.25034189e-01
-7.30487347e-01 -6.70140386e-01 -3.85182291e-01 1.42478064e-01
4.71778005e-01 -6.05428696e-01 4.26705092e-01 -6.67659581e-01
-1.21220613e+00 7.50191152e-01 -1.74707040e-01 -1.20621860e-01
1.71143368e-01 3.59151185e-01 -2.57392645e-01 -1.42334566e-01
-4.93224204e-01 4.03847247e-01 6.79832399e-01 -9.69823360e-01
-2.02693835e-01 -4.78185117e-01 -6.26026979e-03 -5.09760417e-02
-1.64316386e-01 6.65715635e-02 -3.85350585e-01 -2.14404061e-01
-2.59915382e-01 -1.29276454e+00 -5.67537725e-01 -7.45115802e-02
-4.19411629e-01 -4.96692866e-01 6.63724124e-01 -4.76830512e-01
1.13569677e+00 -1.77535617e+00 2.71243513e-01 7.12710798e-01
6.14201665e-01 5.44173121e-01 -3.50522220e-01 6.55327737e-01
1.22637324e-01 3.73642147e-01 -4.05673862e-01 5.09987056e-01
-4.27665532e-01 3.52254689e-01 2.10926682e-01 5.61929345e-01
2.04895556e-01 1.01851213e+00 -8.59163642e-01 -2.73944557e-01
-3.57192487e-01 7.26532757e-01 -8.16904068e-01 1.49499044e-01
-6.63113594e-01 5.86944044e-01 -5.42863786e-01 6.97865844e-01
7.92127311e-01 -8.88137341e-01 8.31028521e-01 1.79000832e-02
6.60195202e-02 -3.80133302e-03 -6.63937092e-01 1.10961545e+00
1.13995411e-01 4.44617093e-01 3.54641825e-01 -1.26398933e+00
9.11552370e-01 2.02700973e-01 6.35605872e-01 -4.23074156e-01
6.43289313e-02 -1.86151583e-02 3.79554302e-01 -5.53438008e-01
-1.45298764e-01 1.11696593e-01 3.09812427e-01 5.70676625e-01
-5.27835846e-01 2.55050659e-01 2.24457011e-01 -1.15154214e-01
1.09078896e+00 -5.62155724e-01 3.21861714e-01 -3.58148426e-01
5.99052310e-01 1.15975656e-01 7.80740261e-01 5.13972044e-01
-8.91346633e-02 -2.54981607e-01 8.84865701e-01 -1.06797266e+00
-9.32391405e-01 -5.26577711e-01 4.86118086e-02 1.23397887e+00
-3.92737798e-02 -2.29347810e-01 -7.48627782e-01 -5.15470445e-01
1.68279279e-02 -3.48041415e-01 -4.36422914e-01 1.21228866e-01
-8.67858768e-01 -1.21795070e+00 6.62573576e-01 1.52698876e-02
2.59435654e-01 -1.23631799e+00 -2.66768873e-01 2.03743249e-01
-3.46693933e-01 -6.36739552e-01 -5.74867427e-01 -8.27890933e-02
-1.03531384e+00 -1.57001996e+00 -3.58498067e-01 -1.21608651e+00
9.31618869e-01 2.60406047e-01 1.03057110e+00 9.27600324e-01
-7.31694937e-01 -2.35253662e-01 -7.33676180e-02 -1.68064192e-01
-5.18795848e-01 3.61232579e-01 1.74524084e-01 -1.04431391e-01
3.65580648e-01 -5.78026474e-01 -7.59924293e-01 4.78831649e-01
-8.89718831e-01 -2.12374747e-01 3.79910409e-01 1.18095529e+00
4.32506680e-01 4.40387011e-01 4.96644229e-01 -1.02169406e+00
7.39524961e-01 -7.52859890e-01 -1.14812338e+00 4.60197449e-01
-7.18177497e-01 -9.53664444e-03 7.15352058e-01 -2.17519075e-01
-6.09862864e-01 6.27555475e-02 -1.64227784e-02 -2.20802875e-04
2.39060342e-01 9.18485224e-01 2.27311894e-01 -2.36260593e-01
4.98055488e-01 3.73890191e-01 4.46835399e-01 -2.55355388e-01
8.08517486e-02 4.99907225e-01 -1.39955595e-01 -3.56316157e-02
6.63472950e-01 5.43630958e-01 4.15429056e-01 -1.17475164e+00
-4.42976475e-01 -5.81650734e-01 -5.88064611e-01 -6.89718798e-02
6.90725803e-01 -7.43574560e-01 -1.46099794e+00 5.68181634e-01
-1.21301103e+00 -2.67652184e-01 5.68197250e-01 -3.35652344e-02
-6.07693195e-02 5.76935768e-01 -9.24551368e-01 -4.56006736e-01
-6.15314245e-01 -1.06491125e+00 6.92463577e-01 -3.96782048e-02
1.58722140e-02 -1.25238717e+00 5.78666806e-01 4.04465050e-01
5.16729951e-01 3.20888788e-01 1.43494880e+00 -5.41519701e-01
-9.05544460e-01 6.66976944e-02 -2.40634382e-01 -1.45534739e-01
2.78720319e-01 2.25006267e-01 -4.57441986e-01 -7.51802325e-01
-1.58103555e-01 -2.38182575e-01 8.34828198e-01 3.18280905e-01
1.14883769e+00 -5.54239750e-01 -4.90997046e-01 8.67189705e-01
1.26532280e+00 3.86046767e-01 4.63901073e-01 6.59675971e-02
6.97209656e-01 5.90032458e-01 2.92731702e-01 6.21012032e-01
1.88113272e-01 4.90247995e-01 5.71284413e-01 -5.03352463e-01
7.21807480e-01 6.14675432e-02 2.17420325e-01 1.23329175e+00
-2.96360999e-01 -3.11449230e-01 -1.10380900e+00 1.57490030e-01
-1.66820359e+00 -1.05146348e+00 -1.13829449e-01 1.77097321e+00
1.07910740e+00 -1.78122103e-01 1.54582068e-01 -3.15316021e-01
8.12536001e-01 2.93551415e-01 -8.32225740e-01 -4.14819688e-01
-1.51578844e-01 -4.87137772e-02 1.77711755e-01 4.83757347e-01
-8.70292306e-01 8.33271921e-01 6.71876955e+00 5.52167475e-01
-1.24884474e+00 -1.11265048e-01 6.74555957e-01 1.31924391e-01
-2.46542349e-01 -2.12836310e-01 -5.00605524e-01 2.53174722e-01
8.15274775e-01 2.82121688e-01 9.13996935e-01 5.27973652e-01
2.06112042e-01 4.09265995e-01 -8.02447140e-01 8.73125613e-01
-2.65554577e-01 -1.83000803e+00 -1.10069409e-01 4.03325438e-01
8.06652129e-01 4.97166753e-01 -1.65403932e-01 -1.20280653e-01
3.48421961e-01 -1.28267539e+00 -5.84125757e-01 3.64091605e-01
7.57212460e-01 -9.20004249e-01 7.15762317e-01 4.32965666e-01
-1.22123754e+00 -6.99901534e-03 -6.40545309e-01 -9.91952606e-03
8.04851577e-02 6.73201203e-01 -1.38374567e+00 3.96837480e-02
5.76203287e-01 6.71999216e-01 -9.56429467e-02 9.10950541e-01
1.49118528e-01 5.85869908e-01 -2.78431714e-01 -4.63986993e-01
2.22372666e-01 -6.10122681e-01 5.10492206e-01 1.22078657e+00
4.02079672e-02 1.56172648e-01 4.55294698e-01 6.85826898e-01
-2.96834618e-01 4.97489907e-02 -6.58543706e-01 -5.05522788e-01
3.24307621e-01 1.13901258e+00 -8.75087321e-01 -1.31881908e-01
-2.34318748e-01 6.86591506e-01 4.46667939e-01 3.56997132e-01
-7.13459074e-01 -2.56012827e-01 9.14150894e-01 9.20997784e-02
4.22726631e-01 -2.60437489e-01 2.66842037e-01 -8.64565432e-01
-3.29956770e-01 -1.33543277e+00 4.28273737e-01 -1.94859490e-01
-1.27476108e+00 6.75338149e-01 -5.79167962e-01 -6.66622698e-01
-7.07153755e-04 -7.17070580e-01 -5.11925936e-01 5.00021577e-01
-1.38249874e+00 -5.01233459e-01 -2.56628096e-01 7.85247147e-01
2.98082381e-01 -2.09549442e-01 8.71172190e-01 1.95902973e-01
-8.06635380e-01 3.54148358e-01 5.42823017e-01 5.78768589e-02
1.39346551e-02 -7.20109940e-01 7.20146477e-01 2.59904891e-01
-3.52075137e-02 8.27802360e-01 3.95437479e-01 -9.64167476e-01
-1.80361259e+00 -9.09696817e-01 8.37319076e-01 -9.38951299e-02
7.36226559e-01 -6.02972269e-01 -1.03252053e+00 2.60501176e-01
1.83230728e-01 -2.01293975e-01 7.98101604e-01 2.48019367e-01
-3.51648927e-01 -1.06131591e-01 -9.42019045e-01 5.28055966e-01
1.26789975e+00 -6.34398043e-01 -2.07563862e-02 8.95741820e-01
8.69175255e-01 -5.32974489e-02 -6.90285981e-01 5.71581662e-01
5.26373327e-01 -7.81327963e-01 9.82575536e-01 -9.96033072e-01
-1.87453270e-01 -3.21997136e-01 3.38092417e-01 -1.27440298e+00
-4.96690631e-01 -9.88755703e-01 -5.34691103e-02 4.05807912e-01
4.79137629e-01 -8.96664679e-01 9.15943205e-01 -1.16119057e-01
3.70976120e-01 -9.83747542e-01 -9.88514304e-01 -4.77308601e-01
-4.57890391e-01 2.26751417e-01 7.40589201e-01 1.29856324e+00
-3.47990803e-02 3.78521502e-01 -5.89774787e-01 1.81564987e-01
5.09540617e-01 2.76552439e-01 6.67145193e-01 -1.70108807e+00
-3.05224031e-01 -4.22415495e-01 -1.60332963e-01 -7.88699448e-01
1.01228869e-02 -1.00223637e+00 -7.79816089e-03 -1.21475208e+00
4.36236888e-01 -7.72995472e-01 -3.99405986e-01 4.49941635e-01
-3.25527415e-02 -1.28414229e-01 -2.82755941e-01 4.19165969e-01
-4.99651045e-01 4.67876762e-01 1.35345268e+00 -2.75683790e-01
-3.34863126e-01 2.34918222e-01 -3.55930150e-01 5.86916387e-01
9.32013929e-01 -5.33295989e-01 -4.16264951e-01 -1.75830424e-01
7.87855506e-01 7.52583519e-02 -1.83027033e-02 -2.76580542e-01
2.92981267e-01 -3.55972230e-01 8.69341195e-02 -6.51628852e-01
1.69854447e-01 -8.15297782e-01 4.12058234e-01 1.23419988e+00
4.12803888e-02 6.20610952e-01 3.11690494e-02 7.87189484e-01
-2.21226178e-02 1.97825924e-01 6.09223723e-01 -2.26956427e-01
-2.49182224e-01 7.33658373e-01 -5.81546366e-01 1.43931344e-01
7.83547282e-01 -7.09718019e-02 -4.84529376e-01 -2.11984426e-01
-4.54255521e-01 4.83396649e-01 4.19040412e-01 3.36175770e-01
7.58493066e-01 -6.75726414e-01 -5.68032861e-01 3.45142275e-01
-2.83557832e-01 -4.85713258e-02 2.40355402e-01 9.02290940e-01
-1.08380902e+00 6.84108436e-01 -1.72066748e-01 -8.98130536e-01
-1.48079336e+00 8.54699254e-01 4.79984939e-01 -5.30353069e-01
-3.80719095e-01 9.00921822e-01 1.85224384e-01 -8.92966628e-01
3.62677813e-01 6.00934355e-03 -3.22876006e-01 1.90325812e-01
4.15578067e-01 3.44932884e-01 -2.40799069e-01 -3.59981894e-01
-6.29401088e-01 6.11913860e-01 -1.39815748e-01 8.08238506e-01
1.68958318e+00 1.17871597e-01 -1.03272784e+00 7.20961764e-02
1.07822096e+00 -2.93922335e-01 -9.75042105e-01 -3.21451843e-01
1.73142105e-01 -1.11569196e-01 -4.58848119e-01 -4.05434906e-01
-1.22396266e+00 7.15512931e-01 5.88370144e-01 4.73057300e-01
1.11035192e+00 -9.50472429e-02 8.65805268e-01 9.76840973e-01
3.93663079e-01 -8.37352812e-01 1.00750834e-01 6.83453202e-01
7.02342808e-01 -1.19530535e+00 -6.19704239e-02 -4.47567046e-01
-3.61237437e-01 9.93643045e-01 3.71924251e-01 2.25120962e-01
1.00371492e+00 3.82239968e-02 -1.31977089e-02 -7.90788114e-01
-9.49342072e-01 5.46063110e-02 8.53710398e-02 5.63757122e-01
2.87679046e-01 2.76792403e-02 -1.92143455e-01 -3.50544959e-01
2.02215239e-01 -4.22937423e-02 1.92656100e-01 6.35102153e-01
-6.06355429e-01 -1.12039757e+00 -5.40844277e-02 4.87574220e-01
-3.02189857e-01 -3.51203740e-01 -6.35157406e-01 4.91257459e-01
-2.22096760e-02 8.59701037e-01 3.40020917e-02 -3.68587166e-01
-6.49501458e-02 -3.49728353e-02 1.03808954e-01 -3.42932522e-01
-6.39262259e-01 -1.39006019e-01 -5.80105139e-03 -5.80326617e-01
-4.61958677e-01 -2.66359895e-01 -1.25785935e+00 -7.02666819e-01
-3.75176340e-01 2.35700235e-01 6.01051748e-01 7.86097229e-01
8.99046302e-01 2.02481091e-01 1.00681317e+00 -4.87092048e-01
-4.01118398e-01 -6.58856392e-01 -3.59282970e-01 6.52475581e-02
4.92867112e-01 -3.72424781e-01 -1.85141623e-01 -4.37335372e-01] | [4.997255325317383, 5.560227870941162] |
b30db9c7-a355-490f-bb33-2381f5b7057e | reinforced-path-reasoning-for-counterfactual | 2207.06674 | null | https://arxiv.org/abs/2207.06674v1 | https://arxiv.org/pdf/2207.06674v1.pdf | Reinforced Path Reasoning for Counterfactual Explainable Recommendation | Counterfactual explanations interpret the recommendation mechanism via exploring how minimal alterations on items or users affect the recommendation decisions. Existing counterfactual explainable approaches face huge search space and their explanations are either action-based (e.g., user click) or aspect-based (i.e., item description). We believe item attribute-based explanations are more intuitive and persuadable for users since they explain by fine-grained item demographic features (e.g., brand). Moreover, counterfactual explanation could enhance recommendations by filtering out negative items. In this work, we propose a novel Counterfactual Explainable Recommendation (CERec) to generate item attribute-based counterfactual explanations meanwhile to boost recommendation performance. Our CERec optimizes an explanation policy upon uniformly searching candidate counterfactuals within a reinforcement learning environment. We reduce the huge search space with an adaptive path sampler by using rich context information of a given knowledge graph. We also deploy the explanation policy to a recommendation model to enhance the recommendation. Extensive explainability and recommendation evaluations demonstrate CERec's ability to provide explanations consistent with user preferences and maintain improved recommendations. We release our code at https://github.com/Chrystalii/CERec. | ['Guandong Xu', 'Dianer Yu', 'Qian Li', 'Xiangmeng Wang'] | 2022-07-14 | null | null | null | null | ['counterfactual-explanation'] | ['miscellaneous'] | [ 2.20864505e-01 5.82434893e-01 -9.20149624e-01 -4.64756966e-01
-2.81336635e-01 -5.68118155e-01 4.70189542e-01 -6.03360124e-02
-3.26863155e-02 1.14262724e+00 8.56565058e-01 -8.71179581e-01
-6.64894283e-01 -9.90267813e-01 -9.29338336e-01 -6.42049164e-02
1.20852981e-02 4.25867081e-01 -2.71344066e-01 -3.00563723e-01
6.82851851e-01 -6.02897592e-02 -1.70250356e+00 4.57682580e-01
1.39089429e+00 3.18062514e-01 3.08871746e-01 5.39487422e-01
-1.88270226e-01 1.81071803e-01 -1.20568588e-01 -7.06637442e-01
1.98439255e-01 -4.39774513e-01 -8.19798410e-01 4.06482071e-02
1.39302567e-01 -4.35754627e-01 1.21404259e-02 7.83004999e-01
1.60708860e-01 5.56527078e-01 5.03077149e-01 -1.52662134e+00
-1.47620559e+00 1.34921014e+00 -3.95257145e-01 -1.73808709e-02
4.67178702e-01 2.39363909e-01 1.49750662e+00 -8.01254928e-01
4.43958938e-01 1.36446142e+00 2.46232569e-01 7.72300184e-01
-1.36664414e+00 -7.64929354e-01 1.06861377e+00 4.77559537e-01
-6.71334982e-01 -4.25007101e-03 7.43516326e-01 2.70723160e-02
7.66947806e-01 9.26548898e-01 8.12398553e-01 1.19131982e+00
2.62199901e-03 6.91895425e-01 1.04370499e+00 -3.01115245e-01
4.79292929e-01 1.00064985e-01 2.66555786e-01 6.61568165e-01
1.06423318e+00 4.21774775e-01 -5.30313373e-01 -3.67343664e-01
6.83933794e-01 5.85769236e-01 -4.34143633e-01 -5.10654092e-01
-1.08704233e+00 1.30929565e+00 4.95286584e-01 -2.35454455e-01
-8.21931481e-01 4.30783719e-01 -8.76324400e-02 2.03539774e-01
3.22668523e-01 1.18348145e+00 -8.64394486e-01 1.38401896e-01
-1.14588626e-01 3.94743919e-01 7.39782810e-01 7.47034252e-01
8.45109940e-01 7.16352165e-02 -4.15298939e-01 6.36436582e-01
6.79192066e-01 5.97103596e-01 6.28733099e-01 -9.38139558e-01
3.19634885e-01 4.94505018e-01 3.73287290e-01 -9.16331470e-01
-2.49187142e-01 -7.18182266e-01 -4.87270892e-01 5.38747460e-02
1.66139200e-01 -2.84375906e-01 -7.69177377e-01 1.77096534e+00
4.92683113e-01 2.13546097e-01 -1.01556726e-01 1.24329531e+00
4.75127012e-01 2.40851611e-01 3.83594930e-01 -2.76124775e-01
1.37058914e+00 -1.07764375e+00 -7.82112300e-01 -2.82778263e-01
5.78743041e-01 -4.88314211e-01 1.57161319e+00 2.74550617e-01
-6.90271139e-01 -4.05634671e-01 -6.98030233e-01 5.17502785e-01
-3.27030927e-01 -1.46718159e-01 1.39616001e+00 7.58428574e-01
-6.58553302e-01 7.39890754e-01 -4.38912183e-01 -1.91887766e-01
2.81385988e-01 5.41105390e-01 2.14492202e-01 -6.25885352e-02
-1.42094314e+00 3.38819832e-01 2.23168164e-01 -3.16708446e-01
-5.09626448e-01 -7.66990244e-01 -6.73797607e-01 4.13210273e-01
1.02458239e+00 -1.31618941e+00 1.27292395e+00 -9.04857457e-01
-1.49216926e+00 -2.90251791e-01 -1.14019409e-01 -5.02590120e-01
1.91057876e-01 -2.80705690e-01 -5.63949883e-01 -4.07985240e-01
2.36627057e-01 6.49690688e-01 7.75104463e-01 -1.36423433e+00
-7.73737907e-01 -2.71542847e-01 5.19053340e-01 4.43606406e-01
-1.62097886e-01 -7.03920722e-01 -9.41260234e-02 -7.84455180e-01
-2.29522295e-04 -1.11657739e+00 -7.14374006e-01 -3.60534519e-01
-6.36964440e-01 -3.28882784e-01 3.70642096e-01 -3.24820250e-01
1.29659307e+00 -1.59478438e+00 -4.03242528e-01 4.01146561e-01
1.66351378e-01 -1.87393427e-01 -3.54896992e-01 3.34034741e-01
-9.14914981e-02 7.39947081e-01 2.90933460e-01 1.57753140e-01
3.47684294e-01 2.36760870e-01 -4.15405750e-01 2.59038024e-02
-3.77058446e-01 1.08611989e+00 -9.93303716e-01 1.05761543e-01
6.39066473e-02 1.80866629e-01 -1.30936420e+00 -7.11246431e-02
-6.41332269e-01 1.49631575e-01 -8.43472838e-01 3.31202000e-01
4.53844398e-01 -6.03327394e-01 6.07587874e-01 -1.34669757e-02
5.77744432e-02 8.21494102e-01 -1.26680660e+00 1.29037893e+00
-7.44422972e-01 -1.56047335e-02 -7.39057124e-01 -4.69926059e-01
7.28878736e-01 8.77782479e-02 9.24979709e-03 -6.20466530e-01
-2.30745390e-01 1.84867620e-01 1.45411789e-01 -2.37127632e-01
7.49783874e-01 -6.43551871e-02 1.43098503e-01 1.10200715e+00
-5.32295406e-01 4.05943573e-01 -1.41357422e-01 4.13864017e-01
7.39823043e-01 1.57639593e-01 7.44940519e-01 -2.18164757e-01
2.57696033e-01 -7.55744204e-02 4.88621563e-01 1.27175879e+00
2.77044743e-01 2.10359111e-01 1.08607285e-01 -4.71614361e-01
-5.90967834e-01 -7.27830887e-01 4.87692982e-01 1.27794933e+00
3.62472564e-01 -4.10714597e-01 -3.24191034e-01 -1.42662168e+00
3.56336862e-01 1.56078744e+00 -7.76207626e-01 -5.12535930e-01
-8.05459172e-02 -4.36037004e-01 -2.84221977e-01 6.10822499e-01
3.04897249e-01 -1.06944191e+00 -2.72439480e-01 1.05047248e-01
-3.25100183e-01 -1.23215556e-01 -1.01715982e+00 -1.35188192e-01
-1.02513146e+00 -1.05450225e+00 -2.05235526e-01 8.92586540e-03
7.71857440e-01 8.09029222e-01 1.14000285e+00 4.89141852e-01
3.16329628e-01 5.67207217e-01 -5.58771431e-01 -4.09400791e-01
1.11726768e-01 -1.42323270e-01 3.05181712e-01 -1.49228916e-01
2.51587898e-01 -6.04685605e-01 -1.31626523e+00 5.37139773e-01
-6.31830752e-01 2.98804879e-01 5.93742549e-01 8.09061468e-01
6.99491084e-01 -1.56236500e-01 8.78767192e-01 -1.46450937e+00
9.92967546e-01 -8.93683076e-01 -3.22734177e-01 2.95463383e-01
-1.63874090e+00 5.37739277e-01 7.56564915e-01 -7.18172312e-01
-1.52557039e+00 -2.09505931e-01 5.40576130e-02 2.11681530e-01
-1.58404127e-01 4.36248153e-01 -1.16480716e-01 3.83934796e-01
8.27194750e-01 -7.19630420e-02 -3.63823593e-01 -5.18259346e-01
9.60247695e-01 3.25800210e-01 3.20696359e-04 -4.73695278e-01
8.44804704e-01 4.35008377e-01 -3.76612216e-01 1.30683303e-01
-1.03642607e+00 -2.60784030e-01 2.21458346e-01 -1.20864296e-02
4.55375046e-01 -4.04184133e-01 -1.16223586e+00 -7.81941354e-01
-9.05740857e-01 -2.93114811e-01 -4.94587123e-01 9.19974029e-01
-5.84430516e-01 7.61548197e-03 -5.23857884e-02 -8.65670919e-01
-3.64100724e-01 -8.01671326e-01 6.44630849e-01 3.44731092e-01
-6.37105405e-01 -1.07212722e+00 1.57413967e-02 5.36832452e-01
3.72715890e-01 -8.28645602e-02 9.12934721e-01 -9.42642570e-01
-9.41869438e-01 2.48211056e-01 -7.66349956e-02 -5.15485466e-01
2.91453987e-01 -3.10606450e-01 -5.25594890e-01 -5.27285561e-02
-5.62909484e-01 2.19659477e-01 7.52250791e-01 6.14455640e-01
1.39451909e+00 -1.22444093e+00 -5.12506187e-01 2.31444731e-01
1.29155874e+00 2.68595487e-01 3.50163728e-01 5.19308984e-01
4.73018676e-01 4.52630371e-01 1.07425559e+00 7.66009986e-01
5.03228724e-01 6.31664693e-01 7.34676242e-01 3.35340351e-02
-1.36767611e-01 -8.92664313e-01 1.38784319e-01 2.52106916e-02
-3.99840415e-01 -3.45442504e-01 -2.54776210e-01 4.51048195e-01
-2.20229888e+00 -1.13458681e+00 -3.77430677e-01 2.04053831e+00
4.72540617e-01 2.11167019e-02 9.19757113e-02 -2.04342797e-01
6.73494220e-01 -2.21917972e-01 -9.98587668e-01 -5.36126137e-01
2.40932181e-01 -2.89880186e-01 5.14707625e-01 7.23420024e-01
-4.18907583e-01 8.25512886e-01 5.15519285e+00 4.67474312e-01
-5.76752722e-01 -6.52202815e-02 6.16701245e-01 -2.24476710e-01
-1.54248142e+00 2.36713558e-01 -6.46582842e-01 3.66889507e-01
7.07526565e-01 -6.00856423e-01 7.61385322e-01 1.10065866e+00
6.41565084e-01 2.37338647e-01 -1.00162899e+00 4.10071641e-01
-3.74201566e-01 -1.81206775e+00 5.16982198e-01 3.63378853e-01
9.57254410e-01 -4.45799887e-01 3.17870587e-01 5.07728994e-01
1.10310328e+00 -7.57577598e-01 6.73334360e-01 4.75669652e-01
5.01224399e-01 -7.69527256e-01 5.00078917e-01 2.73744285e-01
-8.90738368e-01 -2.58851290e-01 -4.88056540e-01 -3.40583026e-01
3.45268622e-02 3.59463125e-01 -1.19866848e+00 5.73327065e-01
4.49020684e-01 5.99958718e-01 -3.62539619e-01 8.36658180e-01
-8.25418532e-01 9.32525754e-01 1.92089066e-01 -4.59008664e-01
-7.58212060e-03 -1.34771302e-01 4.61873561e-01 6.84392869e-01
5.22807479e-01 5.01963973e-01 -4.96281423e-02 9.37329352e-01
-7.86349401e-02 2.73253083e-01 -5.11077285e-01 2.78432876e-01
7.41083145e-01 1.12006581e+00 -6.51135385e-01 -4.05448705e-01
-2.91968197e-01 9.10197079e-01 1.34192944e-01 6.74441993e-01
-8.97009373e-01 1.31451666e-01 9.74368215e-01 1.85227260e-01
4.74488914e-01 5.55131376e-01 -6.23698652e-01 -1.19240177e+00
-4.22441155e-01 -1.02608335e+00 7.53458440e-01 -9.28003073e-01
-1.24206924e+00 3.17518234e-01 -1.56716838e-01 -9.86184895e-01
-2.76298016e-01 -1.75081640e-01 -8.01714778e-01 6.07036769e-01
-1.50289702e+00 -7.28722394e-01 -6.07556477e-02 5.28657019e-01
8.96894991e-01 7.16560706e-02 6.39895320e-01 -2.30255008e-01
-2.76592195e-01 5.62825561e-01 -6.84857965e-02 -8.40504766e-01
5.30805647e-01 -1.46437156e+00 5.13324082e-01 5.35867989e-01
5.12618482e-01 1.24375618e+00 1.02486539e+00 -9.49642420e-01
-1.38135779e+00 -1.07291925e+00 6.95058405e-01 -3.32169592e-01
4.09464091e-01 1.44645259e-01 -7.56581485e-01 7.55612969e-01
1.56234682e-01 -5.48096061e-01 1.03994977e+00 8.73727143e-01
-4.08979058e-01 8.70046243e-02 -1.03528023e+00 1.16419721e+00
1.44030321e+00 1.09996907e-02 -5.31618834e-01 3.86861950e-01
1.23491788e+00 1.43056318e-01 -3.68096858e-01 -1.74249383e-03
6.67235374e-01 -8.79513383e-01 1.00776148e+00 -1.13636279e+00
5.03154516e-01 -3.80138785e-01 3.81858423e-02 -1.84606409e+00
-8.72878850e-01 -8.19802642e-01 -2.94970512e-01 9.16619778e-01
9.67039466e-01 -9.04471576e-01 8.90245318e-01 9.33537424e-01
-6.83073178e-02 -8.57860982e-01 -1.33935288e-01 -6.74117446e-01
-5.63996553e-01 -4.68000919e-01 1.40920734e+00 9.04781818e-01
3.22956234e-01 5.54686189e-01 -6.08234286e-01 3.53049636e-01
5.54259419e-01 8.11047137e-01 7.06984758e-01 -1.08004069e+00
-8.39361548e-01 -2.39396542e-01 6.12561107e-01 -1.11766124e+00
-7.51788542e-02 -8.64813328e-01 -2.54112810e-01 -1.80932188e+00
3.55096549e-01 -4.50414330e-01 -4.65007067e-01 5.84869981e-01
-6.80371344e-01 3.53239998e-02 1.00846410e-01 1.48499340e-01
-8.28045189e-01 5.18350542e-01 1.52695954e+00 5.51758036e-02
-3.29972208e-01 4.50760990e-01 -1.64588356e+00 6.77491248e-01
1.28791642e+00 -6.21932864e-01 -9.21658337e-01 -7.62384310e-02
5.78554034e-01 1.71408832e-01 2.18869224e-01 2.45999079e-02
-1.72660589e-01 -7.35692143e-01 2.87801176e-01 -2.42494985e-01
-1.71680272e-01 -6.71716928e-01 3.59476447e-01 7.78226376e-01
-8.65125597e-01 -3.91842518e-03 -5.69449030e-02 1.24433064e+00
3.99308711e-01 -9.42809358e-02 -6.70274859e-03 -1.14489727e-01
-6.48230493e-01 3.72811168e-01 -3.65900636e-01 -3.89881760e-01
6.94846451e-01 -2.53555685e-01 -5.79269469e-01 -8.78395379e-01
-6.99249268e-01 2.73335010e-01 1.42152995e-01 5.81387877e-01
7.91387320e-01 -1.45622897e+00 -4.45756853e-01 -1.55211210e-01
2.51414508e-01 -7.76385605e-01 3.73378992e-01 4.00419295e-01
4.21522379e-01 7.11642802e-01 -6.93664998e-02 2.13744313e-01
-1.28450799e+00 8.01276028e-01 -3.45631205e-02 -4.62008178e-01
-2.30241343e-01 6.63062453e-01 4.82105702e-01 -6.59330785e-01
-1.89876989e-01 -2.06639767e-01 -5.66653132e-01 -4.49095547e-01
4.55578446e-01 5.30038297e-01 -5.49113691e-01 4.21796799e-01
-2.58983105e-01 -5.15195169e-02 -1.43692002e-01 -4.60134782e-02
1.34602582e+00 -4.70757276e-01 4.75904882e-01 -1.09188937e-01
4.82407182e-01 4.55836743e-01 -1.37954152e+00 -1.05462056e-02
3.95354405e-02 -9.15077806e-01 -1.34331077e-01 -1.40800977e+00
-8.81605685e-01 2.55094051e-01 2.06995398e-01 5.08265316e-01
9.60851669e-01 9.21021327e-02 4.88417149e-01 4.41035509e-01
2.82952875e-01 -7.48882592e-01 8.91249329e-02 -1.30507708e-01
1.02590072e+00 -1.27004588e+00 2.25778863e-01 -6.46474838e-01
-1.00493288e+00 7.74813533e-01 8.13485861e-01 2.46451586e-01
3.96000177e-01 -5.32388270e-01 -2.08031103e-01 -2.41469175e-01
-1.26494122e+00 -2.59625405e-01 6.42288625e-01 5.94199717e-01
6.73170507e-01 5.41158319e-01 -9.71154928e-01 1.16274965e+00
-3.30172390e-01 -9.00488868e-02 5.35576224e-01 2.14985833e-02
-5.74932873e-01 -1.15954471e+00 -2.55411714e-01 9.06905591e-01
-1.84661999e-01 -3.28843951e-01 -4.11571831e-01 7.41962314e-01
-7.59856850e-02 1.25433218e+00 -2.74465293e-01 -4.04356748e-01
4.73458350e-01 -2.50189692e-01 2.69396398e-02 -6.96238458e-01
-4.12464499e-01 -6.10179901e-02 3.55596423e-01 -7.66158819e-01
-1.51170462e-01 -6.21893942e-01 -1.42882884e+00 -5.69987416e-01
-5.92590749e-01 6.59980059e-01 4.83312398e-01 8.45061660e-01
8.04800332e-01 6.31959438e-01 6.52911484e-01 -1.27141610e-01
-7.59069264e-01 -6.54272556e-01 -4.22332406e-01 5.17445147e-01
3.37609649e-02 -8.15133393e-01 -1.99200839e-01 -1.13692537e-01] | [9.448583602905273, 5.687960624694824] |
db29d918-6244-4689-955f-4c1b200ab916 | link-and-code-fast-indexing-with-graphs-and | 1804.09996 | null | http://arxiv.org/abs/1804.09996v2 | http://arxiv.org/pdf/1804.09996v2.pdf | Link and code: Fast indexing with graphs and compact regression codes | Similarity search approaches based on graph walks have recently attained
outstanding speed-accuracy trade-offs, taking aside the memory requirements. In
this paper, we revisit these approaches by considering, additionally, the
memory constraint required to index billions of images on a single server. This
leads us to propose a method based both on graph traversal and compact
representations. We encode the indexed vectors using quantization and exploit
the graph structure to refine the similarity estimation.
In essence, our method takes the best of these two worlds: the search
strategy is based on nested graphs, thereby providing high precision with a
relatively small set of comparisons. At the same time it offers a significant
memory compression. As a result, our approach outperforms the state of the art
on operating points considering 64-128 bytes per vector, as demonstrated by our
results on two billion-scale public benchmarks. | ['Hervé Jégou', 'Matthijs Douze', 'Alexandre Sablayrolles'] | 2018-04-26 | link-and-code-fast-indexing-with-graphs-and-1 | http://openaccess.thecvf.com/content_cvpr_2018/html/Douze_Link_and_Code_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Douze_Link_and_Code_CVPR_2018_paper.pdf | cvpr-2018-6 | ['image-similarity-search'] | ['computer-vision'] | [ 1.96089178e-01 -3.00593704e-01 -3.79114032e-01 -6.77746162e-02
-8.15611899e-01 -5.38663030e-01 6.40441179e-01 8.39031398e-01
-4.84117389e-01 3.33568007e-01 8.10549185e-02 -3.89748961e-01
-2.69741625e-01 -1.14863575e+00 -4.44536537e-01 -4.54096645e-01
-1.77396744e-01 4.14598793e-01 7.52321005e-01 -3.15389007e-01
7.51944125e-01 5.53467035e-01 -1.80563712e+00 -8.61293972e-02
7.29788423e-01 1.36078429e+00 1.58857182e-01 6.97515011e-01
-2.49474958e-01 6.29566908e-01 -2.19191700e-01 -8.29372764e-01
4.02765900e-01 -1.29115626e-01 -7.23482192e-01 -1.27448499e-01
5.95870554e-01 -5.67466497e-01 -6.28039956e-01 1.07743025e+00
4.27819669e-01 1.75093472e-01 2.20790625e-01 -9.69124258e-01
-3.54018152e-01 5.17514706e-01 -6.35272026e-01 3.02582443e-01
6.89503610e-01 7.95906335e-02 1.45268631e+00 -5.37438393e-01
6.80581152e-01 8.47648561e-01 6.64503336e-01 -1.06590942e-01
-1.30582917e+00 -2.78427511e-01 -9.99430418e-02 6.43306673e-01
-1.73962033e+00 -5.41312039e-01 4.20137733e-01 5.09407409e-02
1.45385599e+00 5.51310480e-01 7.56925702e-01 3.72316092e-01
2.46062785e-01 3.78227502e-01 8.26423883e-01 -3.67005855e-01
4.40160394e-01 -4.03616160e-01 3.47320855e-01 8.62616837e-01
6.61251962e-01 -1.91033125e-01 -5.88957846e-01 -6.34963036e-01
4.51734930e-01 -1.53876975e-01 -3.31242055e-01 -6.73068225e-01
-1.13300979e+00 9.11929369e-01 3.58578116e-01 2.44152397e-01
-4.35199410e-01 4.89603579e-01 7.66566157e-01 3.92788619e-01
1.79223105e-01 2.42055699e-01 1.64100051e-01 -2.30195299e-01
-1.19650698e+00 2.16551751e-01 1.00699306e+00 9.25034344e-01
8.91472697e-01 -3.33799422e-01 -1.25282302e-01 6.21870279e-01
2.04892293e-01 4.28586006e-01 4.19087917e-01 -8.14882398e-01
5.42688906e-01 4.86816287e-01 -3.28373700e-01 -1.54408145e+00
-1.38409734e-01 -3.14009070e-01 -7.38763332e-01 -2.72040945e-02
1.32810012e-01 7.55571783e-01 -6.60247386e-01 1.29977751e+00
3.60307574e-01 6.76279590e-02 -2.04418376e-01 5.77553034e-01
4.02378321e-01 5.23917556e-01 -2.36944452e-01 -2.98886389e-01
1.53049541e+00 -1.05760491e+00 -6.01957560e-01 5.60473613e-02
7.28645921e-01 -7.16137052e-01 9.51377988e-01 3.99322242e-01
-1.30358112e+00 -4.45496589e-01 -1.36356282e+00 -7.69454986e-03
-3.06185424e-01 -4.34348047e-01 7.69120336e-01 7.66659915e-01
-1.55276859e+00 9.87834036e-01 -9.43839788e-01 -5.55705488e-01
1.01029925e-01 4.81321633e-01 -2.68410921e-01 -7.32078552e-02
-8.72551620e-01 5.23252249e-01 3.95170510e-01 -4.55989510e-01
-1.00703672e-01 -3.43925625e-01 -6.64642394e-01 4.66073155e-01
5.20006299e-01 -6.34829402e-01 9.51411903e-01 -2.30538577e-01
-1.26829600e+00 7.47951388e-01 -1.81038409e-01 -8.56075883e-01
2.33137518e-01 -3.80386487e-02 -1.77934542e-01 6.16969347e-01
2.62446352e-03 2.67574281e-01 5.46514928e-01 -6.00543618e-01
-5.31276882e-01 -5.41635811e-01 8.80478881e-03 1.10569313e-01
-7.29574203e-01 4.57857642e-03 -9.10007954e-01 -6.13911450e-01
2.19554633e-01 -8.76339495e-01 -4.16634262e-01 9.39251631e-02
-1.20474584e-01 -2.78825551e-01 3.15197408e-01 -3.05265337e-01
1.77848792e+00 -2.11851120e+00 1.80749372e-02 5.44556916e-01
6.27697170e-01 2.57412553e-01 -5.79344258e-02 8.20103526e-01
3.07984412e-01 7.51933381e-02 -1.14007205e-01 -1.25797242e-01
1.34556800e-01 1.74669579e-01 -2.95082837e-01 6.30187452e-01
-2.62334406e-01 1.05078721e+00 -8.31101775e-01 -8.08005154e-01
9.55675021e-02 2.70339400e-01 -6.73226893e-01 1.30801603e-01
1.30080923e-01 -4.01098341e-01 -5.66162527e-01 4.30194080e-01
7.11495817e-01 -7.12859511e-01 4.75568533e-01 -3.99171531e-01
-8.61342251e-02 3.71849507e-01 -1.24308872e+00 1.75746965e+00
-1.29274875e-01 2.36483306e-01 -1.77771181e-01 -1.00230169e+00
8.27958822e-01 -7.73202181e-02 6.10393345e-01 -9.86387193e-01
-1.58134326e-01 5.11427343e-01 -4.45639014e-01 -5.76234283e-03
7.36942172e-01 4.59679663e-01 -1.36253629e-02 6.13755524e-01
-2.40113333e-01 -2.01047793e-01 6.06398284e-01 3.76460612e-01
1.51987123e+00 -2.56883085e-01 6.28948569e-01 -3.42647344e-01
6.01066828e-01 2.44781431e-02 -4.37610783e-02 8.48448098e-01
5.29132318e-03 2.83664227e-01 4.09413338e-01 -3.81292999e-01
-9.60872352e-01 -1.07349956e+00 7.84974024e-02 8.77827168e-01
3.34425926e-01 -1.10270810e+00 -7.68782437e-01 -3.93511593e-01
-1.93370879e-02 4.23443347e-01 -2.88498133e-01 -1.08245358e-01
-7.76515841e-01 -3.39787453e-01 3.73325646e-01 3.32669854e-01
3.58154774e-01 -5.76439559e-01 -1.07691050e+00 3.50852996e-01
1.60921752e-01 -1.10429192e+00 -6.09873652e-01 -5.36334468e-04
-1.03678942e+00 -1.04824150e+00 -6.26634717e-01 -6.02289200e-01
1.81015328e-01 6.42815232e-01 1.38746977e+00 3.04596782e-01
-2.42552042e-01 4.49271500e-01 -4.67775494e-01 3.57665151e-01
-1.78238347e-01 2.90804446e-01 -2.16078281e-01 -2.12003022e-01
3.14168930e-01 -9.38875318e-01 -7.54855156e-01 9.53182876e-02
-1.04539824e+00 -1.78731441e-01 6.15033448e-01 7.71688700e-01
9.57814634e-01 -2.22056992e-02 4.49053459e-02 -4.97096986e-01
7.16745794e-01 -4.47224945e-01 -9.98542488e-01 3.39223593e-01
-1.25193226e+00 3.88968796e-01 6.05795860e-01 -1.61816731e-01
-3.58380020e-01 1.54150978e-01 -1.88175619e-01 -3.63377124e-01
2.85222501e-01 4.51386303e-01 3.58737826e-01 -4.63727266e-01
4.05178934e-01 6.30900383e-01 2.62166888e-01 -3.45212489e-01
5.38641572e-01 5.17917097e-01 5.39087474e-01 -3.77840012e-01
5.72757542e-01 5.01510084e-01 3.72672886e-01 -6.39865816e-01
-1.14090033e-01 -7.63988316e-01 -4.39665377e-01 1.21111877e-01
4.66893613e-01 -5.19962847e-01 -8.27432215e-01 5.66966906e-02
-9.75406408e-01 1.14081381e-02 -2.23883182e-01 3.85785103e-01
-7.25610733e-01 1.18747640e+00 -6.31982803e-01 -5.12587190e-01
-6.95062280e-01 -1.24230015e+00 1.15524578e+00 -2.41897210e-01
-2.26977393e-01 -6.43703580e-01 2.47278675e-01 5.38752787e-03
5.79020083e-01 1.33788228e-01 7.39133716e-01 -7.64287293e-01
-9.43871737e-01 -3.52329016e-01 -4.80352879e-01 -1.79147378e-01
-1.27260670e-01 -2.26435840e-01 -4.48906779e-01 -4.82228428e-01
-1.57757297e-01 -1.56350851e-01 7.76942194e-01 -7.19542280e-02
1.20418084e+00 -1.94896713e-01 -3.88922095e-01 8.31629276e-01
1.78120804e+00 -8.32326710e-02 7.59772599e-01 4.50350940e-01
4.89704072e-01 3.53725493e-01 6.61064923e-01 7.29859173e-01
3.48148763e-01 1.10061240e+00 6.55223310e-01 2.84869105e-01
-2.40793288e-01 -1.41107738e-01 -8.29233304e-02 1.17961776e+00
-3.45997848e-02 -4.33560818e-01 -8.71885955e-01 5.42457163e-01
-1.82184231e+00 -8.46719146e-01 -1.60974741e-01 2.61619997e+00
6.10247612e-01 1.41792729e-01 1.90782070e-01 4.02010024e-01
4.88604128e-01 5.83511353e-01 -2.89272219e-01 -5.72187662e-01
5.51695786e-02 5.92945635e-01 9.06548381e-01 3.50650460e-01
-7.07012057e-01 6.40725732e-01 6.81702995e+00 1.12821865e+00
-9.89545941e-01 -5.01314402e-02 4.09726113e-01 -7.65649825e-02
-4.01436150e-01 -3.58851478e-02 -7.27999032e-01 3.81305546e-01
1.13163137e+00 -5.02227247e-01 6.65307641e-01 8.27656686e-01
-3.89685303e-01 -4.07869034e-02 -9.45380926e-01 1.36114430e+00
8.79162550e-02 -1.56415260e+00 3.72263670e-01 3.73009920e-01
3.19753945e-01 2.06955552e-01 -1.18576162e-01 -2.08736703e-01
-4.01927717e-03 -8.81574392e-01 5.28189421e-01 2.59864241e-01
8.98248434e-01 -7.43574023e-01 6.07358634e-01 -3.22370641e-02
-1.73840547e+00 1.11200340e-01 -3.51085365e-01 3.48761156e-02
2.90035903e-01 6.45423412e-01 -4.27533776e-01 7.00672984e-01
6.28750563e-01 3.44697267e-01 -7.13713884e-01 1.12382853e+00
3.55661631e-01 3.80378097e-01 -6.04794860e-01 -3.75542521e-01
3.47827792e-01 -2.34990746e-01 4.14870828e-01 1.26297474e+00
4.99790519e-01 3.91138624e-03 1.36658236e-01 4.14124250e-01
7.81882647e-03 5.86157441e-01 -6.37698293e-01 -5.11908233e-02
6.37597859e-01 1.09171319e+00 -9.83959615e-01 -4.76952255e-01
-5.78164816e-01 1.18739629e+00 5.97132921e-01 -8.30264837e-02
-8.42746377e-01 -7.59111166e-01 4.90967691e-01 2.98980683e-01
7.10783243e-01 -4.82956648e-01 -8.12101439e-02 -9.99540448e-01
3.47495079e-01 -8.90887141e-01 4.82394934e-01 -2.06788093e-01
-1.00859046e+00 8.65395904e-01 8.83636475e-02 -1.04433012e+00
-4.09763902e-01 -4.90099043e-01 -1.02690712e-01 6.09338522e-01
-1.55478525e+00 -6.16534472e-01 -3.91383916e-01 5.39848983e-01
2.57181376e-01 7.99970999e-02 8.86274993e-01 4.58149821e-01
-5.71227707e-02 8.64000380e-01 1.73463762e-01 -3.37717950e-01
3.56008381e-01 -1.00381911e+00 1.06849301e+00 6.47450686e-01
4.25849468e-01 6.52374089e-01 4.86484379e-01 -6.00796342e-01
-2.03643131e+00 -5.75332582e-01 1.02562618e+00 8.49488750e-02
9.91143763e-01 -1.46343186e-01 -9.87263620e-01 1.30226791e-01
1.81393549e-01 3.73392612e-01 5.31796992e-01 -1.28691271e-01
-6.32633150e-01 -1.69391468e-01 -9.77205515e-01 6.05707467e-01
1.28960431e+00 -6.00581050e-01 -2.82884777e-01 4.09338444e-01
6.16788447e-01 -3.33257467e-01 -1.15111470e+00 2.75050968e-01
6.84660256e-01 -1.32352257e+00 1.25701582e+00 -1.42473027e-01
3.19604203e-02 -7.05409646e-02 -5.89603722e-01 -6.52395070e-01
-3.81054044e-01 -8.43678892e-01 -4.86820787e-01 8.23499441e-01
2.80962456e-02 -8.86948884e-01 8.18951368e-01 2.34403372e-01
3.06252986e-01 -1.20077229e+00 -9.93928611e-01 -1.09694493e+00
-4.40563887e-01 -3.39601278e-01 8.80633533e-01 5.87338209e-01
1.39844000e-01 2.82839000e-01 -7.80451968e-02 -1.83533102e-01
8.52188885e-01 4.31619912e-01 7.30858922e-01 -1.10632527e+00
-5.51102459e-01 -6.47723854e-01 -8.69868398e-01 -1.40106356e+00
-6.50832653e-02 -8.23098421e-01 -3.43851984e-01 -1.32925689e+00
2.94957221e-01 -4.86211449e-01 -2.75775790e-01 1.83866605e-01
-1.78763181e-01 5.98596871e-01 2.48471692e-01 4.29786235e-01
-1.01321101e+00 3.14728796e-01 7.58628607e-01 2.40756925e-02
1.10608600e-01 -3.65208954e-01 -4.56617028e-01 3.75835806e-01
8.47308218e-01 -2.36382395e-01 -4.75089490e-01 -4.29158896e-01
3.75493288e-01 6.20393790e-02 1.81686208e-01 -1.16137516e+00
5.55192888e-01 1.05979539e-01 -3.61871272e-01 -6.57671094e-01
4.30049568e-01 -7.49994040e-01 3.27725649e-01 8.13045144e-01
-2.51413554e-01 5.46431005e-01 2.84148119e-02 8.03376853e-01
-2.57139027e-01 -2.52564996e-01 6.43060267e-01 -4.88981046e-02
-7.58259892e-01 3.54141027e-01 3.62380617e-03 1.67618822e-02
9.50675309e-01 -5.49516380e-01 -2.01810107e-01 -3.74868214e-01
-1.74975902e-01 -5.87481558e-02 8.07051837e-01 -3.29688564e-02
5.54903865e-01 -1.43301880e+00 -6.00899994e-01 1.29510269e-01
2.50648975e-01 -5.41819274e-01 -5.79951517e-03 7.44114220e-01
-9.63404596e-01 6.47281229e-01 -1.34604678e-01 -6.27612710e-01
-1.60917008e+00 1.00503659e+00 -1.48653120e-01 -7.28915095e-01
-7.31808066e-01 5.57863832e-01 -1.15700670e-01 2.43033767e-01
1.62822410e-01 -8.95293802e-02 4.77220081e-02 -2.29315422e-02
6.49256945e-01 6.74123287e-01 3.24322611e-01 -5.13391972e-01
-4.38672513e-01 8.82167161e-01 2.75243931e-02 3.69988754e-02
1.08695769e+00 -2.26145118e-01 -4.04672593e-01 1.50892278e-02
1.52722526e+00 3.26203763e-01 -6.96685791e-01 -2.77731925e-01
2.33636066e-01 -7.41473615e-01 1.59590140e-01 -3.82564329e-02
-1.00862467e+00 4.79652196e-01 6.29439592e-01 6.20186388e-01
1.28690922e+00 -3.02715637e-02 1.32931876e+00 6.27422571e-01
8.59951437e-01 -8.74377310e-01 -3.73393819e-02 1.66267470e-01
4.68431741e-01 -9.38609362e-01 4.63813215e-01 -7.06995904e-01
-6.24824874e-02 1.21569669e+00 -7.90148601e-02 -3.51128638e-01
4.33105469e-01 2.63033062e-01 -4.10032213e-01 -3.15085441e-01
-7.14374661e-01 -3.31910074e-01 3.76703620e-01 3.89975876e-01
2.48483777e-01 -2.21964736e-02 -8.63698781e-01 -7.74220228e-02
-1.34897918e-01 -1.13414630e-01 1.11097256e-02 1.01588523e+00
-6.13427103e-01 -1.30899715e+00 -2.46746883e-01 6.37096405e-01
-4.97811168e-01 -4.80732888e-01 -1.47291794e-01 6.91798627e-01
-5.86541414e-01 8.39209855e-01 7.97255039e-02 -4.79886591e-01
2.27707908e-01 -3.13140661e-01 5.00549555e-01 -2.12614670e-01
-6.49854362e-01 -1.74161077e-01 1.08803064e-01 -1.30039752e+00
-1.40107289e-01 -4.22999650e-01 -9.74274755e-01 -7.01592803e-01
-2.93092877e-01 3.26150239e-01 6.14852965e-01 3.26460034e-01
6.95695281e-01 2.65791476e-01 6.28440976e-01 -7.43019044e-01
-1.15396011e+00 -3.07721913e-01 -6.02926493e-01 3.00095618e-01
-3.51814218e-02 -3.74223262e-01 -3.05579305e-01 -3.90745372e-01] | [8.495582580566406, 3.5244452953338623] |
cfff2d1a-5c6e-41d9-80f3-8bf5bc3b5cd4 | hierarchical-multi-label-classification | null | null | https://icml.cc/Conferences/2018/Schedule?showEvent=2306 | http://proceedings.mlr.press/v80/wehrmann18a/wehrmann18a.pdf | Hierarchical Multi-Label Classification Networks |
One of the most challenging machine learning problems is a particular case of data classification in which classes are hierarchically structured and objects can be assigned to multiple paths of the class hierarchy at the same time. This task is known as hierarchical multi-label classification (HMC), with applications in text classification, image annotation, and in bioinformatics problems such as protein function prediction. In this paper, we propose novel neural network architectures for HMC called HMCN, capable of simultaneously optimizing local and global loss functions for discovering local hierarchical class-relationships and global information from the entire class hierarchy while penalizing hierarchical violations. We evaluate its performance in 21 datasets from four distinct domains, and we compare it against the current HMC state-of-the-art approaches. Results show that HMCN substantially outperforms all baselines with statistical significance, arising as the novel state-of-the-art for HMC.
| ['Rodrigo Barros', 'Jonatas Wehrmann', 'Ricardo Cerri'] | 2018-07-01 | null | null | null | icml-2018-7 | ['protein-function-prediction'] | ['medical'] | [ 4.90614504e-01 8.62754360e-02 -3.51915747e-01 -6.69120729e-01
-9.42999125e-01 -4.09690320e-01 2.77825624e-01 7.42412984e-01
-4.22353208e-01 7.92659760e-01 3.68685126e-02 -2.12327287e-01
-4.59147364e-01 -3.81201714e-01 -5.93952894e-01 -6.96260691e-01
-1.71708003e-01 1.00813639e+00 2.62799799e-01 2.80736834e-01
1.76361501e-01 4.07990515e-01 -1.60042202e+00 1.03353059e+00
4.82941717e-01 1.15858245e+00 -2.64994800e-01 5.29502034e-01
1.25359073e-01 8.65490973e-01 -3.39171350e-01 -3.89239073e-01
-1.94755718e-01 -1.16312400e-01 -1.38845289e+00 -4.71681822e-03
8.02103937e-01 2.39671320e-01 2.81872809e-01 1.01125526e+00
4.18032557e-01 5.78494109e-02 9.76532340e-01 -1.30076468e+00
-3.86883825e-01 5.04350603e-01 -7.31469274e-01 -3.31274122e-01
-9.20865387e-02 -3.70723456e-01 1.56786382e+00 -6.58944666e-01
7.17141688e-01 1.59684932e+00 7.69902527e-01 5.09129822e-01
-1.71601772e+00 -5.47803819e-01 3.74901354e-01 3.77141148e-01
-1.43188918e+00 -7.17206597e-02 4.24826682e-01 -8.06837201e-01
1.04251957e+00 2.73132026e-01 -1.74635932e-01 9.52976346e-01
2.51753300e-01 9.41223204e-01 1.12513912e+00 -1.64753452e-01
2.91015744e-01 -2.32636049e-01 6.65490389e-01 8.26510549e-01
-1.16472222e-01 -3.13703239e-01 -4.57791924e-01 -5.48405111e-01
-5.05534001e-02 -2.71416396e-01 -4.99279937e-03 -4.24586862e-01
-1.04231775e+00 9.95272219e-01 3.73417437e-01 2.01932162e-01
-7.10280016e-02 1.32404491e-01 5.90492129e-01 1.02576390e-01
5.59605300e-01 6.15225673e-01 -9.90531623e-01 3.97477180e-01
-5.84800899e-01 3.43907356e-01 8.29678893e-01 7.90855467e-01
6.60895586e-01 -5.22207677e-01 -2.88274348e-01 1.21767032e+00
2.42744952e-01 -3.08775365e-01 3.95539969e-01 -1.11402869e+00
3.25317174e-01 9.33350563e-01 -2.30835155e-01 -7.91651428e-01
-1.07064748e+00 -5.18224239e-01 -1.00644016e+00 1.11886121e-01
2.54849106e-01 2.38559321e-01 -6.81882799e-01 1.82793343e+00
4.52901185e-01 -6.10334426e-03 -1.27707079e-01 4.19260085e-01
8.61978114e-01 5.85908890e-01 3.45159739e-01 -1.95445567e-01
1.34919751e+00 -1.12452388e+00 -4.34221923e-01 -7.56398290e-02
7.80227721e-01 -3.41556102e-01 8.46394598e-01 6.35417104e-01
-7.30914295e-01 -4.50646073e-01 -8.77161443e-01 -5.24364889e-01
-6.50042713e-01 -9.02382433e-02 4.56625074e-01 7.65400529e-02
-8.74071181e-01 8.04158270e-01 -6.11446798e-01 -2.15741768e-01
4.06128854e-01 5.16667545e-01 -4.23678279e-01 -6.65138811e-02
-1.19235420e+00 7.08537579e-01 7.64229536e-01 -7.45211467e-02
-9.36205626e-01 -7.28544116e-01 -6.64208412e-01 2.88058132e-01
2.56420374e-01 -2.09764808e-01 1.21710300e+00 -6.65278077e-01
-9.39485610e-01 1.25964785e+00 -3.28849927e-02 -4.04012978e-01
3.62221479e-01 -1.15948871e-01 -2.89505631e-01 -1.63615093e-01
4.16785657e-01 9.93141830e-01 2.58882344e-01 -1.39948940e+00
-1.03910816e+00 -6.03891611e-01 -3.15111220e-01 5.79335429e-02
-2.77052909e-01 2.54601836e-01 -4.12620723e-01 -3.01373869e-01
6.02482520e-02 -9.04179275e-01 -3.11407000e-01 -8.30889568e-02
-7.66271770e-01 -9.91493165e-01 8.48948359e-01 -3.26294124e-01
1.22772217e+00 -2.02232337e+00 5.19986391e-01 5.73448800e-02
4.37723488e-01 -7.74351731e-02 -1.10796742e-01 2.01668188e-01
-1.51673958e-01 2.26163581e-01 -4.22532052e-01 -4.89530116e-01
1.99450999e-01 2.04979911e-01 5.41205443e-02 6.30680382e-01
1.02713697e-01 7.00000167e-01 -5.55634797e-01 -6.11046374e-01
-2.70036310e-01 2.04381242e-01 -4.01612848e-01 3.61978374e-02
-6.57618582e-01 4.78931755e-01 -5.42041175e-02 5.63101351e-01
3.84484738e-01 -1.03468990e+00 6.91112518e-01 -1.60778165e-01
2.21402664e-02 2.54860282e-01 -7.90468752e-01 1.29225707e+00
-6.36227503e-02 5.09228289e-01 -9.08060148e-02 -1.19728422e+00
5.82400560e-01 3.28698277e-01 6.85733438e-01 -5.74600220e-01
-3.91460136e-02 1.38874188e-01 3.44067551e-02 -1.77097172e-01
7.36186579e-02 6.08569421e-02 -2.55404502e-01 2.00149179e-01
8.23890939e-02 3.70409667e-01 1.69778660e-01 -3.84732243e-03
9.85222220e-01 -2.21278623e-01 7.94995070e-01 -5.81840456e-01
5.32169521e-01 -1.72099948e-01 1.04487193e+00 7.40289569e-01
-2.37797543e-01 3.62568766e-01 8.96141827e-01 -6.27536178e-01
-9.99434650e-01 -6.03425145e-01 -5.16298473e-01 1.67974210e+00
-8.72976482e-02 -4.73796368e-01 -5.04245818e-01 -9.67252374e-01
3.08557123e-01 3.44102859e-01 -7.44270682e-01 1.11358762e-02
-4.74602133e-01 -8.09621394e-01 5.77204108e-01 4.30208862e-01
3.41935903e-01 -1.00171781e+00 -2.25072995e-01 4.18862194e-01
-1.79741204e-01 -1.28050673e+00 -2.28240460e-01 7.03060448e-01
-5.06716371e-01 -1.23290646e+00 -3.48190308e-01 -9.05066371e-01
2.88871497e-01 -1.66706458e-01 1.40187907e+00 -2.67106318e-03
-4.09635037e-01 -1.99836925e-01 -3.13855708e-01 -1.44753858e-01
-4.11157012e-01 6.09400272e-01 -1.05581418e-01 7.32089579e-02
2.08832130e-01 -2.79730350e-01 -3.07141036e-01 6.10077381e-01
-7.77621806e-01 2.51817852e-01 2.04921558e-01 9.59874868e-01
8.47421706e-01 3.68299663e-01 6.59022689e-01 -1.55637527e+00
2.44726405e-01 -5.60378075e-01 -8.17228854e-01 6.06696308e-01
-7.29799151e-01 8.74061510e-02 8.14562976e-01 -1.82699084e-01
-6.73080504e-01 2.73494869e-01 -1.31072283e-01 -2.50518396e-02
-4.46572989e-01 5.90925336e-01 -4.16767687e-01 1.01602292e-02
5.11316359e-01 -1.54311672e-01 -5.93351007e-01 -7.15543807e-01
1.72904193e-01 7.15630293e-01 3.58767360e-01 -7.78388917e-01
-1.14696607e-01 2.81006694e-01 4.59253639e-01 -6.02030873e-01
-1.67691255e+00 -8.49048913e-01 -9.36921775e-01 1.06570469e-02
1.27171242e+00 -6.39357448e-01 -1.14136767e+00 5.13853669e-01
-1.20981920e+00 -5.04958868e-01 4.40664142e-01 -1.17109738e-01
-5.33103883e-01 2.18231753e-01 -1.07745504e+00 -4.13691491e-01
-1.45895973e-01 -1.30860937e+00 1.47402310e+00 -1.49050489e-01
-1.08908258e-01 -9.74268556e-01 3.48224416e-02 6.28563404e-01
-1.20516069e-01 4.11268085e-01 1.81261909e+00 -9.24738944e-01
-3.47587675e-01 5.83913401e-02 -3.90983105e-01 1.68063492e-01
-7.30979145e-02 -1.39447108e-01 -9.86197889e-01 -6.54165506e-01
-6.57282650e-01 -8.82062197e-01 1.07697320e+00 1.85278133e-01
1.89056766e+00 -4.26178575e-01 -4.36406970e-01 6.65908575e-01
1.45373654e+00 1.94969699e-01 1.92596391e-01 4.13680941e-01
8.37423742e-01 9.94842887e-01 3.46035123e-01 3.12201530e-01
4.96831954e-01 9.88452077e-01 5.84876239e-01 -1.67189896e-01
1.71577826e-01 1.44484118e-01 -2.62376934e-01 5.53102791e-01
4.19940591e-01 -5.81319928e-01 -1.22913074e+00 1.76240116e-01
-2.23746610e+00 -4.33635890e-01 -4.80351120e-01 1.90727019e+00
1.07842517e+00 8.24641287e-02 1.73394587e-02 -1.41381808e-02
8.76058102e-01 -7.95102268e-02 -8.56443942e-01 -2.75286973e-01
-2.77266920e-01 -5.05521894e-02 4.04161751e-01 4.42727178e-01
-1.85267746e+00 9.11894262e-01 6.72756720e+00 8.79243314e-01
-6.80703223e-01 1.07611597e-01 1.06087005e+00 1.45097822e-01
3.28586251e-01 -2.16159150e-01 -1.28237998e+00 2.99650818e-01
1.04401970e+00 4.09004331e-01 1.23285845e-01 6.37590468e-01
-2.15507418e-01 2.47227788e-01 -1.40731025e+00 7.77996063e-01
-1.04061887e-01 -1.18877816e+00 7.82640427e-02 2.60103047e-01
9.13826346e-01 1.29373863e-01 8.23254958e-02 3.57932836e-01
6.60442352e-01 -1.02163708e+00 3.86853755e-01 1.65821090e-01
8.32420409e-01 -6.64293230e-01 5.58870375e-01 3.42710078e-01
-1.09484768e+00 -2.66274452e-01 -1.43568233e-01 2.86454767e-01
-1.83532804e-01 5.28295279e-01 -6.50985062e-01 4.94789809e-01
7.27447927e-01 8.59869540e-01 -8.76905560e-01 9.61599350e-01
-7.09695518e-02 6.80167615e-01 6.12315759e-02 2.25988522e-01
4.49037492e-01 2.45875403e-01 3.12275589e-02 1.26521254e+00
-2.90007651e-01 3.16392118e-03 8.25886250e-01 4.60230798e-01
-5.02695084e-01 1.95192620e-01 -2.42807776e-01 1.77305698e-01
2.97422051e-01 1.22707939e+00 -8.11957955e-01 -3.30831796e-01
-5.55759072e-01 8.73944044e-01 9.32352960e-01 9.51828733e-02
-7.79909074e-01 -3.01199019e-01 3.61435771e-01 -3.52328986e-01
-2.18086615e-02 5.55716977e-02 -2.44674683e-01 -8.20906639e-01
-4.41203117e-01 -9.37825084e-01 1.11773157e+00 -3.15932900e-01
-1.69810069e+00 4.91490871e-01 -2.16187388e-01 -9.14805949e-01
9.22845230e-02 -1.11888051e+00 5.81121556e-02 3.53377551e-01
-1.42512500e+00 -1.32587659e+00 -3.20815784e-03 4.02913898e-01
5.85527182e-01 -2.25956231e-01 9.80345488e-01 5.45983315e-01
-7.16060400e-01 7.10295856e-01 6.38403058e-01 -6.61145523e-02
9.01940525e-01 -1.60787511e+00 1.62431419e-01 3.29385102e-01
-6.00508749e-02 2.87284434e-01 3.34621757e-01 -5.60143530e-01
-4.41652954e-01 -1.51164460e+00 1.13317692e+00 -2.23307580e-01
6.68297231e-01 -6.73859954e-01 -1.20338225e+00 6.28653169e-01
-4.20557894e-02 6.88239485e-02 1.14836013e+00 4.93925303e-01
-7.72637665e-01 2.11505741e-02 -9.01762426e-01 1.64969876e-01
8.65126610e-01 -5.04979789e-01 2.19375081e-02 7.97066867e-01
9.90766406e-01 -6.68442622e-02 -1.14627731e+00 6.93140328e-01
7.16844499e-01 -7.97204137e-01 7.83441186e-01 -1.30103910e+00
5.97405136e-01 -2.14021251e-01 -5.76016903e-01 -9.26578820e-01
-8.21683168e-01 -1.49464518e-01 -3.51144299e-02 1.04108107e+00
8.42877150e-01 -3.50226402e-01 7.88057446e-01 3.83420229e-01
-2.81411141e-01 -9.14645016e-01 -8.97660971e-01 -6.41033709e-01
4.53338861e-01 2.29292605e-02 2.64878035e-01 1.31926298e+00
-2.98553169e-01 8.52778316e-01 -5.36228597e-01 1.96927994e-01
9.71828222e-01 3.53667885e-01 2.43983954e-01 -1.94284153e+00
-4.51790065e-01 -5.99916399e-01 -2.71020561e-01 -6.71138346e-01
8.48531842e-01 -1.19080472e+00 2.51322210e-01 -1.47548497e+00
6.93337440e-01 -3.10907781e-01 -5.78842402e-01 9.59696651e-01
-2.69988030e-01 1.01542234e-01 -1.72610618e-02 3.96884978e-01
-1.22528613e+00 3.30275863e-01 7.93585837e-01 -4.50813293e-01
1.25127092e-01 -3.81279923e-02 -5.94981313e-01 6.10833466e-01
6.17985189e-01 -5.89829326e-01 -2.65133474e-02 -3.45990032e-01
3.02596301e-01 -7.58819166e-04 1.99045375e-01 -8.16562533e-01
3.38558316e-01 -1.86504662e-01 3.46388876e-01 -7.58232355e-01
1.75568148e-01 -6.96631432e-01 -3.65907103e-02 3.55538636e-01
-1.08713055e+00 5.43425120e-02 -2.05380004e-02 6.28966033e-01
-1.36129349e-01 -1.09457523e-01 1.35348558e+00 -9.02165398e-02
-6.65201485e-01 2.68699318e-01 -3.13126922e-01 8.64766687e-02
9.06022370e-01 5.02076209e-01 -6.88402653e-01 8.84845480e-02
-9.20533061e-01 6.95979834e-01 8.03183168e-02 2.92695194e-01
3.71068150e-01 -1.22755897e+00 -7.55344629e-01 -6.23477586e-02
4.88553435e-01 2.58782525e-02 5.54376878e-02 6.39888465e-01
-2.40443051e-01 8.21581781e-01 6.94074258e-02 -7.23407447e-01
-1.62307322e+00 6.66072786e-01 3.85760397e-01 -7.87985623e-01
-3.11637580e-01 8.01321387e-01 6.50539756e-01 -9.17472184e-01
6.25608981e-01 5.76502867e-02 -3.21454346e-01 2.71075815e-02
2.05932423e-01 3.70801479e-01 2.95757920e-01 -5.69499791e-01
-4.14263844e-01 4.66116458e-01 -4.83195812e-01 3.52446735e-01
1.17957628e+00 9.19042379e-02 -6.58830822e-01 6.81526303e-01
1.72179961e+00 -8.76849771e-01 -8.95756662e-01 -4.30218697e-01
7.38980174e-01 1.28799289e-01 -7.43792057e-02 -1.17182732e+00
-6.36362135e-01 8.20104599e-01 5.49817264e-01 1.67442277e-01
9.06953514e-01 2.09689990e-01 3.03859234e-01 6.17123067e-01
2.31472254e-01 -1.09167182e+00 7.96552002e-02 8.35313320e-01
5.52753329e-01 -1.35676432e+00 -1.17222570e-01 -4.02750850e-01
-5.45100272e-01 9.10060167e-01 8.15113127e-01 3.86819422e-01
6.74213529e-01 -2.71006059e-02 -2.95985699e-01 -4.39390242e-01
-1.25391304e+00 -2.09854752e-01 5.39772809e-01 7.69399330e-02
8.54530990e-01 1.36788324e-01 -2.56462216e-01 3.69251877e-01
2.95845926e-01 -5.08167207e-01 -5.91934994e-02 7.62415409e-01
-5.00575602e-01 -1.06327760e+00 6.09758124e-02 6.44728184e-01
-8.66709232e-01 -5.43460846e-02 -7.52757728e-01 5.34502625e-01
3.76399517e-01 9.02603984e-01 -8.35008696e-02 -3.35401177e-01
2.34037414e-01 2.46338606e-01 1.49632782e-01 -6.76401734e-01
-5.47869802e-01 3.22100937e-01 3.02409142e-01 -5.46006680e-01
-4.26302165e-01 -6.42318964e-01 -1.33978796e+00 7.84007460e-02
-2.16540888e-01 3.08256090e-01 3.60424101e-01 7.63914645e-01
3.79838794e-01 6.08651280e-01 6.30350530e-01 -5.57661295e-01
-4.58548158e-01 -8.56461525e-01 -6.77343607e-01 6.61871552e-01
5.20632684e-01 -7.34594703e-01 -1.03075989e-01 -3.27671441e-04] | [9.580009460449219, 4.347387790679932] |
c85e8079-2e54-4f12-b9ba-75c38fa6a9b9 | a-language-guided-benchmark-for-weakly | 2302.14163 | null | https://arxiv.org/abs/2302.14163v1 | https://arxiv.org/pdf/2302.14163v1.pdf | A Language-Guided Benchmark for Weakly Supervised Open Vocabulary Semantic Segmentation | Increasing attention is being diverted to data-efficient problem settings like Open Vocabulary Semantic Segmentation (OVSS) which deals with segmenting an arbitrary object that may or may not be seen during training. The closest standard problems related to OVSS are Zero-Shot and Few-Shot Segmentation (ZSS, FSS) and their Cross-dataset variants where zero to few annotations are needed to segment novel classes. The existing FSS and ZSS methods utilize fully supervised pixel-labelled seen classes to segment unseen classes. Pixel-level labels are hard to obtain, and using weak supervision in the form of inexpensive image-level labels is often more practical. To this end, we propose a novel unified weakly supervised OVSS pipeline that can perform ZSS, FSS and Cross-dataset segmentation on novel classes without using pixel-level labels for either the base (seen) or the novel (unseen) classes in an inductive setting. We propose Weakly-Supervised Language-Guided Segmentation Network (WLSegNet), a novel language-guided segmentation pipeline that i) learns generalizable context vectors with batch aggregates (mean) to map class prompts to image features using frozen CLIP (a vision-language model) and ii) decouples weak ZSS/FSS into weak semantic segmentation and Zero-Shot segmentation. The learned context vectors avoid overfitting on seen classes during training and transfer better to novel classes during testing. WLSegNet avoids fine-tuning and the use of external datasets during training. The proposed pipeline beats existing methods for weak generalized Zero-Shot and weak Few-Shot semantic segmentation by 39 and 3 mIOU points respectively on PASCAL VOC and weak Few-Shot semantic segmentation by 5 mIOU points on MS COCO. On a harder setting of 2-way 1-shot weak FSS, WLSegNet beats the baselines by 13 and 22 mIOU points on PASCAL VOC and MS COCO, respectively. | ['Brejesh lall', 'Monish Natarajan', 'Mustafa Chasmai', 'Prashant Pandey'] | 2023-02-27 | null | null | null | null | ['zero-shot-segmentation'] | ['computer-vision'] | [ 6.12911105e-01 2.91645259e-01 -3.27561170e-01 -6.19524002e-01
-1.27742851e+00 -8.02150726e-01 4.76610512e-01 1.17116734e-01
-6.85544550e-01 4.94873196e-01 -4.67489451e-01 -2.20580980e-01
3.14838231e-01 -7.90883541e-01 -1.11778605e+00 -6.24642313e-01
4.30768788e-01 6.54702365e-01 1.03385282e+00 -2.30251655e-01
-6.61087176e-03 1.16613895e-01 -1.87129807e+00 2.17763990e-01
7.44414330e-01 1.13688791e+00 4.10892516e-01 9.67429876e-01
-5.07136226e-01 2.87162691e-01 -5.67262769e-01 -1.61949292e-01
5.45011282e-01 -3.60312700e-01 -1.07304764e+00 2.50979543e-01
9.32074368e-01 -1.30871266e-01 1.55797988e-01 1.09105682e+00
4.26047742e-01 5.06030262e-01 5.86354911e-01 -1.23213661e+00
-5.44465482e-01 3.90347213e-01 -5.70558190e-01 -4.56405059e-02
7.17563108e-02 6.23214781e-01 9.66651917e-01 -8.77141893e-01
8.50953877e-01 1.20102537e+00 6.28375053e-01 9.00287092e-01
-1.24573314e+00 -3.77878368e-01 3.43602210e-01 -9.87669080e-03
-1.32659078e+00 -2.91064233e-01 3.49007487e-01 -4.89109397e-01
8.44618797e-01 2.41332337e-01 4.86433178e-01 1.12736285e+00
-5.50008595e-01 1.31636715e+00 1.04778862e+00 -3.64275813e-01
5.42063832e-01 3.16000313e-01 6.40963912e-01 6.40253365e-01
-9.12361518e-02 -9.18177515e-02 -2.38664433e-01 2.82757878e-01
5.74535668e-01 4.38360460e-02 -1.38450891e-01 -3.23449165e-01
-1.11074245e+00 8.39747012e-01 4.39904690e-01 1.74738243e-01
1.44312620e-01 2.22490683e-01 5.31492829e-01 1.67171642e-01
4.64498520e-01 2.05910176e-01 -8.14102471e-01 -4.66893725e-02
-1.37798119e+00 7.89398625e-02 6.78509176e-01 1.37739336e+00
1.24500751e+00 1.17071114e-01 -5.89577854e-01 9.57262039e-01
-1.54343685e-02 6.06476009e-01 4.10718709e-01 -1.03433359e+00
1.97503582e-01 5.69888771e-01 -1.15386315e-01 -9.91373062e-02
-1.38617799e-01 -3.13718736e-01 -3.43283147e-01 1.46648705e-01
5.09929955e-01 -4.53893952e-02 -1.93960214e+00 1.64973569e+00
5.80564976e-01 4.34290797e-01 1.76164344e-01 8.63216937e-01
1.26036286e+00 8.54765296e-01 3.05385947e-01 9.23207328e-02
1.32556617e+00 -1.34580350e+00 -2.61422366e-01 -5.98842978e-01
6.98510349e-01 -4.77355421e-01 1.78660011e+00 7.06987903e-02
-8.72889996e-01 -7.64110744e-01 -9.35718596e-01 -4.77424145e-01
-8.43123555e-01 -2.86233813e-01 4.59790230e-01 5.50536692e-01
-1.13619649e+00 4.42688584e-01 -7.72899687e-01 -4.22899485e-01
7.98803627e-01 3.12350482e-01 -9.99169797e-02 -2.37094209e-01
-1.08813405e+00 3.55132520e-01 6.67642176e-01 -2.50303656e-01
-1.28548861e+00 -7.91502655e-01 -1.19156671e+00 -5.95144294e-02
9.58157003e-01 -5.85443795e-01 1.31175721e+00 -1.28701198e+00
-1.36672902e+00 1.26744199e+00 -1.63684607e-01 -4.79519874e-01
4.12400007e-01 -6.66837618e-02 -8.94647613e-02 2.31989563e-01
6.01389349e-01 1.46536720e+00 9.56666291e-01 -1.28491640e+00
-7.51289189e-01 -2.52335429e-01 1.35015741e-01 1.37880668e-01
2.29273294e-03 -3.69570971e-01 -9.56467211e-01 -5.65973639e-01
-7.98735116e-03 -8.40383351e-01 -3.37911934e-01 3.50567792e-03
-6.41036808e-01 -4.84485656e-01 1.02532816e+00 -2.50318199e-01
7.32265592e-01 -2.19886541e+00 -5.16708568e-03 -1.64087474e-01
-1.02344818e-01 5.10677457e-01 -4.50884074e-01 -1.65685102e-01
2.23899111e-01 -3.46541479e-02 -8.34065080e-01 -3.87075305e-01
4.81727682e-02 6.37071133e-01 -1.55659005e-01 4.83257473e-02
3.01296681e-01 1.28746521e+00 -1.03201509e+00 -6.89546347e-01
3.63583952e-01 1.65485144e-01 -4.55573767e-01 2.67659783e-01
-7.37222791e-01 2.82005608e-01 -6.93501458e-02 9.81489420e-01
5.51095188e-01 -2.28618205e-01 -5.94940841e-01 2.82926932e-02
4.55534421e-02 -2.71096587e-01 -1.17020178e+00 2.17864656e+00
-3.31980377e-01 3.42762649e-01 -5.52406497e-02 -1.07789028e+00
6.98211849e-01 1.46422982e-02 7.48534277e-02 -4.86844718e-01
1.05306573e-01 2.53695101e-01 -5.08715808e-01 -5.13474226e-01
4.13068920e-01 -2.28286624e-01 -4.85606164e-01 1.56117633e-01
6.24264777e-01 -6.76762700e-01 5.23459375e-01 3.83945107e-01
7.82347202e-01 4.01745170e-01 6.73204660e-02 -2.27044508e-01
3.91731709e-01 3.10356826e-01 7.32864797e-01 1.08517015e+00
-3.75538796e-01 1.02366567e+00 3.46726745e-01 -9.94908959e-02
-7.67645419e-01 -1.41103256e+00 -2.10241288e-01 1.60767388e+00
5.43298960e-01 -4.04380895e-02 -1.16443944e+00 -9.67164636e-01
-1.17921852e-01 8.18725288e-01 -4.99622256e-01 4.20666523e-02
-9.15921926e-02 -4.03415620e-01 6.02046728e-01 7.52027750e-01
4.96460676e-01 -1.22498441e+00 -5.58196485e-01 8.73004943e-02
5.25523871e-02 -1.26365650e+00 -6.39491916e-01 5.25764465e-01
-6.61190927e-01 -1.04135275e+00 -1.13085115e+00 -1.10040069e+00
6.20329976e-01 2.36872181e-01 1.04350114e+00 -2.51955271e-01
-6.53521419e-01 5.95669448e-01 -3.96534115e-01 -3.63558024e-01
-8.72653052e-02 1.42030925e-01 -2.52635509e-01 5.46707660e-02
5.35155296e-01 -1.14223503e-01 -5.63621879e-01 2.94928610e-01
-1.13758647e+00 1.67172983e-01 3.65580648e-01 8.43446076e-01
1.11019123e+00 -5.10982394e-01 5.28666675e-01 -1.33169830e+00
-1.09432325e-01 -4.53135341e-01 -5.77185571e-01 3.68134260e-01
-3.14107269e-01 -1.28584012e-01 4.71343458e-01 -5.09736896e-01
-1.00806153e+00 3.32716554e-01 -2.09628180e-01 -6.45231962e-01
-6.41150951e-01 5.54736853e-02 -1.40517250e-01 -4.68306169e-02
8.25051129e-01 2.10266739e-01 -3.55991006e-01 -3.52542043e-01
8.24843228e-01 7.99915612e-01 8.47460806e-01 -6.38470054e-01
6.73456907e-01 5.80908358e-01 -5.52309513e-01 -1.08279800e+00
-1.25502396e+00 -1.01062500e+00 -7.37723470e-01 1.22447563e-02
1.34449244e+00 -1.01919270e+00 -2.58817337e-02 5.47275245e-01
-8.75196934e-01 -9.55188751e-01 -1.04679799e+00 -7.76980370e-02
-7.72273123e-01 2.45260671e-01 -5.92782676e-01 -5.63042939e-01
-4.20097768e-01 -1.30458188e+00 1.71994436e+00 4.94153500e-01
-1.28072932e-01 -8.60182285e-01 -3.65670264e-01 7.11727798e-01
2.27453895e-02 3.41356933e-01 4.71814692e-01 -7.38261163e-01
-6.20335937e-01 -1.11348629e-01 -4.19315964e-01 7.08384871e-01
-1.93141401e-01 -1.49596393e-01 -1.24880517e+00 -2.69606799e-01
-3.64447832e-01 -8.85615706e-01 1.20991147e+00 5.00132203e-01
1.19452035e+00 2.16335252e-01 -1.73180416e-01 8.61553967e-01
1.48366523e+00 1.03274547e-01 4.21729624e-01 8.63785371e-02
1.02432680e+00 4.96151149e-01 7.97127068e-01 7.04897940e-02
2.45627373e-01 3.89223903e-01 1.84201673e-01 -4.83538955e-01
-4.73125905e-01 -1.97597980e-01 2.56768078e-01 4.12139803e-01
2.27618530e-01 -1.25958681e-01 -8.30947816e-01 9.99765575e-01
-1.80339277e+00 -5.70369303e-01 -1.65979579e-01 2.05405235e+00
9.04615760e-01 4.07198280e-01 1.33156031e-02 -1.09744251e-01
8.20191681e-01 1.71362549e-01 -9.16030943e-01 -4.39776659e-01
-2.37514660e-01 5.77555597e-01 6.42408311e-01 5.43891191e-01
-1.19649363e+00 1.58457768e+00 4.78089523e+00 1.29623902e+00
-9.75889325e-01 5.03964305e-01 8.48535061e-01 -2.05296069e-01
-2.02815965e-01 7.18126297e-02 -1.09024107e+00 2.26438388e-01
6.94263577e-01 2.34852061e-01 1.68422893e-01 1.07425940e+00
-2.02326402e-01 -3.78776252e-01 -1.19767404e+00 1.08695412e+00
1.27587482e-01 -1.28009737e+00 2.05977540e-03 -4.64353889e-01
1.20869112e+00 4.53185499e-01 -6.87564984e-02 6.91719592e-01
3.53594065e-01 -7.48670220e-01 8.48481536e-01 2.18512878e-01
1.26210141e+00 -3.09458584e-01 5.69387555e-01 4.09126878e-01
-1.16261685e+00 1.26693889e-01 -2.91161269e-01 2.04088911e-01
2.79165328e-01 3.43033671e-01 -6.58029616e-01 2.64546961e-01
8.15843344e-01 6.11920655e-01 -5.50937116e-01 7.02573180e-01
-4.73195285e-01 6.72783256e-01 -5.94981492e-01 1.80904910e-01
8.66255045e-01 3.15199457e-02 4.37388062e-01 1.36813235e+00
-1.90412644e-02 1.31873608e-01 6.10091507e-01 9.18444574e-01
-6.28360733e-02 -6.59032017e-02 -3.00658584e-01 5.87864332e-02
1.17959298e-01 1.22951221e+00 -1.28064060e+00 -8.32665980e-01
-3.83146614e-01 1.28655255e+00 -2.76635382e-02 6.27172768e-01
-7.56211460e-01 -6.59810126e-01 4.83037949e-01 1.03039525e-01
5.50626636e-01 1.24167651e-01 -3.67255211e-01 -1.20382893e+00
-2.37685546e-01 -4.28228468e-01 6.15002334e-01 -6.83649421e-01
-1.11440575e+00 3.93969387e-01 1.13997474e-01 -8.97092819e-01
-3.36122662e-02 -5.31050861e-01 -6.52752697e-01 5.65423012e-01
-1.65055799e+00 -1.40542388e+00 -4.12505895e-01 8.04821014e-01
1.32197750e+00 1.09020501e-01 5.61057806e-01 9.85424444e-02
-4.78835315e-01 5.37575603e-01 -1.00445814e-01 9.53605995e-02
6.42268300e-01 -1.51229489e+00 5.26314378e-01 8.59703958e-01
2.44728565e-01 1.54719830e-01 4.94444788e-01 -5.43824017e-01
-7.74328411e-01 -1.55621588e+00 5.70990622e-01 -4.67631847e-01
4.36213940e-01 -6.37583911e-01 -1.04667270e+00 6.39622092e-01
-1.04129978e-01 6.40352130e-01 4.49366510e-01 -1.65405080e-01
-3.29103380e-01 7.81089906e-03 -1.21582913e+00 4.84658450e-01
1.21835709e+00 -5.63108027e-01 -6.70818865e-01 5.53374946e-01
1.20957386e+00 -4.46944445e-01 -4.98668134e-01 4.73495692e-01
7.28600100e-02 -7.01405823e-01 8.74743760e-01 -4.17017013e-01
2.17382863e-01 -4.92315769e-01 -2.55631357e-01 -9.55614865e-01
1.65432498e-01 -4.84502792e-01 2.07076266e-01 1.13865006e+00
4.85150814e-01 -3.27107221e-01 9.12775218e-01 3.86809945e-01
-5.43423772e-01 -6.31876469e-01 -9.01191354e-01 -9.54575717e-01
-1.23593286e-02 -7.65447915e-01 2.77192265e-01 8.63121450e-01
-5.72108388e-01 3.23030442e-01 1.04856081e-01 2.66626012e-02
6.68319166e-01 2.09166348e-01 9.08802330e-01 -1.03161323e+00
-3.91653419e-01 -2.54010141e-01 -5.26951075e-01 -1.07772613e+00
1.94867462e-01 -1.11775434e+00 5.04162252e-01 -1.61981213e+00
7.85210729e-02 -4.36265767e-01 -3.04991364e-01 8.11028123e-01
-1.92222163e-01 6.52191937e-01 1.93337098e-01 -1.50256291e-01
-1.14292061e+00 3.85699183e-01 1.22748733e+00 -4.66455311e-01
-4.30988818e-01 -1.04351856e-01 -4.87526089e-01 8.46520960e-01
3.78623635e-01 -4.16868716e-01 -6.28981709e-01 -2.83788383e-01
-2.36943439e-01 -9.78696570e-02 4.57103491e-01 -9.76376414e-01
2.66314030e-01 -1.61309451e-01 6.88787177e-02 -5.68566263e-01
2.55486935e-01 -3.82775545e-01 -2.98287213e-01 2.63456166e-01
-2.86031783e-01 -9.08464074e-01 1.93836108e-01 7.51034379e-01
-1.68631747e-01 -5.01179457e-01 1.08535671e+00 -4.49715078e-01
-1.43867266e+00 5.57195187e-01 -6.54981937e-03 6.08774245e-01
1.47457302e+00 -7.35594153e-01 1.82991549e-02 9.50874761e-02
-1.15261662e+00 5.69052935e-01 4.07215416e-01 5.51422298e-01
5.27834058e-01 -8.63782704e-01 -3.53984565e-01 2.56459445e-01
5.49985349e-01 8.30772102e-01 4.78784919e-01 5.74466348e-01
-4.98348296e-01 6.52211532e-02 1.87426761e-01 -1.04630804e+00
-1.15167356e+00 4.83264446e-01 1.52141780e-01 1.63693890e-01
-5.63178658e-01 1.49419880e+00 6.40569389e-01 -7.15340495e-01
3.37833464e-01 -3.98790777e-01 9.72411633e-02 2.08398208e-01
3.69730175e-01 2.42900655e-01 -1.15944587e-01 -5.76658010e-01
-1.31373629e-01 7.16988623e-01 -1.54666185e-01 2.48065758e-02
1.17428601e+00 -1.21146135e-01 1.75571173e-01 8.83364499e-01
1.35051882e+00 -6.04296744e-01 -1.66615689e+00 -3.11825335e-01
9.33241397e-02 -5.91654070e-02 9.11493152e-02 -8.35499883e-01
-1.03525114e+00 1.05707860e+00 7.60799229e-01 -8.24106634e-02
8.93407345e-01 4.20635790e-01 1.29400969e+00 2.09994867e-01
3.63442034e-01 -1.52841508e+00 -4.83712591e-02 5.63720703e-01
3.15649152e-01 -1.66424823e+00 -5.66829264e-01 -5.69217801e-01
-6.92894340e-01 8.76761317e-01 7.70810187e-01 -6.69338331e-02
4.40079451e-01 8.35220739e-02 2.04451695e-01 -1.75262377e-01
-2.84357607e-01 -8.02726030e-01 3.65113020e-01 6.69315577e-01
-3.71536352e-02 1.77330345e-01 4.40408289e-03 6.58918917e-01
6.18068613e-02 1.02520194e-02 3.33244592e-01 9.21904325e-01
-6.73207581e-01 -8.03134024e-01 -2.37377524e-01 5.22333980e-01
-1.45302221e-01 -2.40308508e-01 -1.66406512e-01 6.19976699e-01
6.54316008e-01 7.88224161e-01 2.09716856e-01 2.54118405e-02
3.42716753e-01 2.89267927e-01 2.40802199e-01 -1.29698551e+00
-5.13761282e-01 1.94383562e-01 -1.63465038e-01 -6.33230329e-01
-4.38026160e-01 -4.41877246e-01 -1.61148560e+00 2.86475569e-01
-4.01928127e-01 -7.17248917e-02 5.59281945e-01 1.13224423e+00
1.00696847e-01 5.59051096e-01 1.75759703e-01 -9.19882834e-01
-3.28692943e-01 -6.32583320e-01 -6.25560105e-01 7.03576863e-01
2.77130723e-01 -4.94106352e-01 -4.26710278e-01 3.33949625e-01] | [9.672394752502441, 0.8729109168052673] |
9c46b83b-8402-466b-87cf-14679e094db1 | cross-modal-cross-domain-moment-alignment | null | null | http://openaccess.thecvf.com/content_CVPR_2020/html/Jing_Cross-Modal_Cross-Domain_Moment_Alignment_Network_for_Person_Search_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Jing_Cross-Modal_Cross-Domain_Moment_Alignment_Network_for_Person_Search_CVPR_2020_paper.pdf | Cross-Modal Cross-Domain Moment Alignment Network for Person Search | Text-based person search has drawn increasing attention due to its wide applications in video surveillance. However, most of the existing models depend heavily on paired image-text data, which is very expensive to acquire. Moreover, they always face huge performance drop when directly exploiting them to new domains. To overcome this problem, we make the first attempt to adapt the model to new target domains in the absence of pairwise labels, which combines the challenges from both cross-modal (text-based) person search and cross-domain person search. Specially, we propose a moment alignment network (MAN) to solve the cross-modal cross-domain person search task in this paper. The idea is to learn three effective moment alignments including domain alignment (DA), cross-modal alignment (CA) and exemplar alignment (EA), which together can learn domain-invariant and semantic aligned cross-modal representations to improve model generalization. Extensive experiments are conducted on CUHK Person Description dataset (CUHK-PEDES) and Richly Annotated Pedestrian dataset (RAP). Experimental results show that our proposed model achieves the state-of-the-art performances on five transfer tasks.
| [' Tieniu Tan', ' Liang Wang', ' Wei Wang', 'Ya Jing'] | 2020-06-01 | null | null | null | cvpr-2020-6 | ['person-search'] | ['computer-vision'] | [ 3.91687714e-02 -5.49372017e-01 -1.41695306e-01 -4.46695715e-01
-9.47335422e-01 -3.04423571e-01 8.37106228e-01 -3.17225873e-01
-7.25401700e-01 6.86018288e-01 3.34313780e-01 2.75424868e-01
-1.20331064e-01 -4.15576011e-01 -5.45180798e-01 -6.71641111e-01
2.98768073e-01 7.65261948e-01 2.86988705e-01 -2.12130502e-01
-2.16722101e-01 1.26127169e-01 -1.41015840e+00 2.20249191e-01
9.76118147e-01 9.41761374e-01 1.78589985e-01 1.21940106e-01
9.29737836e-02 4.09709334e-01 -4.73035634e-01 -8.61503482e-01
2.70140082e-01 -2.17130244e-01 -9.87962067e-01 3.16968739e-01
5.06233811e-01 -3.04416806e-01 -6.36676729e-01 1.07837605e+00
8.17492068e-01 3.84397328e-01 5.79269648e-01 -1.52129686e+00
-9.39281166e-01 -1.93052739e-03 -8.62803578e-01 1.50810048e-01
6.24241292e-01 2.49182284e-01 6.83638155e-01 -9.14883852e-01
5.23162663e-01 1.59544706e+00 7.91845679e-01 8.47751260e-01
-9.51716483e-01 -8.14980567e-01 2.55414128e-01 5.73774278e-01
-1.50565028e+00 -3.23881030e-01 8.26076567e-01 -4.31987852e-01
6.61008000e-01 -1.87931061e-02 6.19667292e-01 1.63398659e+00
-3.68765980e-01 1.22644389e+00 1.05994654e+00 -2.60537565e-01
-2.67641246e-01 2.35118330e-01 -3.86881083e-02 5.11714160e-01
1.34693295e-01 1.51235759e-01 -5.01225233e-01 -3.53085510e-02
6.46563232e-01 1.04015730e-01 -3.65251869e-01 -7.55587757e-01
-1.33523786e+00 6.12505913e-01 5.16627252e-01 2.19232872e-01
-1.98462754e-01 -2.74055511e-01 7.86479414e-01 1.02267563e-01
3.05341393e-01 7.45184422e-02 -4.41286474e-01 1.29241049e-01
-5.98701894e-01 4.89826769e-01 3.45115036e-01 1.15455985e+00
4.52075124e-01 -3.87439489e-01 -2.66495496e-01 1.34540164e+00
1.77613646e-01 8.23997557e-01 6.19569123e-01 -3.89855325e-01
8.24824631e-01 5.31977892e-01 2.65132189e-01 -1.08030248e+00
-3.61512214e-01 -4.48697150e-01 -1.08359933e+00 -5.31236112e-01
6.07854664e-01 -1.81506500e-02 -8.76013517e-01 1.92662752e+00
4.65089470e-01 1.56756520e-01 1.85776547e-01 1.19015121e+00
1.08609796e+00 4.78143454e-01 4.06528682e-01 4.07266319e-02
1.59343481e+00 -1.41851199e+00 -4.64713246e-01 -5.35553038e-01
3.16007197e-01 -6.12703800e-01 9.38181341e-01 -2.64099874e-02
-7.86960304e-01 -9.16841567e-01 -7.13956892e-01 -5.21724857e-02
-5.84602654e-01 3.11611682e-01 2.63311356e-01 3.75609159e-01
-7.50639677e-01 -9.50159654e-02 -4.32988524e-01 -8.19209039e-01
4.55664903e-01 3.29436868e-01 -7.52092361e-01 -4.48321193e-01
-1.48796916e+00 9.47481334e-01 5.17536342e-01 2.45249152e-01
-7.23802924e-01 -2.90397406e-01 -9.27638710e-01 -2.20746353e-01
5.13687611e-01 -1.06517792e+00 8.85952234e-01 -9.78398144e-01
-1.04006994e+00 1.29766607e+00 -1.49125174e-01 -2.71914721e-01
7.79194713e-01 -2.87843227e-01 -6.34452999e-01 1.78045318e-01
5.63672721e-01 8.36081624e-01 7.67682195e-01 -1.20089185e+00
-8.45704496e-01 -6.81043744e-01 -4.68250215e-02 3.92080009e-01
-5.25659740e-01 2.27653459e-01 -9.79212940e-01 -6.69138968e-01
-2.19993785e-01 -1.00273550e+00 -3.81696969e-02 -9.84054059e-02
-3.43138456e-01 -5.80276489e-01 7.00138628e-01 -8.97415042e-01
9.07113254e-01 -1.80523264e+00 5.17701030e-01 -1.99034810e-01
-1.57439947e-01 4.86254185e-01 -1.03372969e-01 2.36234933e-01
-3.42077352e-02 -4.52334493e-01 -1.07666075e-01 -4.79442060e-01
1.00621708e-01 8.67056325e-02 -3.84405516e-02 3.94764006e-01
7.94580951e-02 1.01508880e+00 -9.00192261e-01 -9.21420634e-01
2.36116171e-01 4.06639457e-01 -2.58529752e-01 1.85050786e-01
9.89689827e-02 7.84965932e-01 -5.62029123e-01 8.07045400e-01
8.74325812e-01 -3.73178512e-01 -2.34660115e-02 -4.13835138e-01
9.00956839e-02 -3.53618413e-01 -9.01127398e-01 1.90611517e+00
-1.07323736e-01 2.56833494e-01 -2.08864853e-01 -1.28953528e+00
7.11003780e-01 1.84282750e-01 4.61347908e-01 -8.63462269e-01
2.08495352e-02 3.45086195e-02 -4.62101817e-01 -6.26063108e-01
2.56928712e-01 3.86003256e-02 -4.32193071e-01 -1.96579471e-02
1.98081419e-01 5.41527152e-01 2.70592898e-01 -1.37510672e-01
4.63792443e-01 2.02551156e-01 3.07018638e-01 -1.96546361e-01
1.13978767e+00 1.00833997e-01 6.54605865e-01 5.91915131e-01
-5.48182905e-01 4.32852805e-01 1.02005333e-01 -7.90505230e-01
-9.60651457e-01 -1.03424788e+00 -6.89829439e-02 1.08246481e+00
6.92403316e-01 -2.01564223e-01 -7.60121763e-01 -9.28554118e-01
-1.00170761e-01 7.59679601e-02 -5.45409322e-01 -1.73818871e-01
-6.43980205e-01 -7.97481239e-01 6.98868275e-01 6.49839103e-01
1.28082561e+00 -1.00637901e+00 -3.55097264e-01 -6.74234629e-02
-8.51141632e-01 -1.50084746e+00 -8.51871669e-01 -4.69752580e-01
-3.51381898e-01 -1.20348907e+00 -1.44520748e+00 -1.20971131e+00
6.32927299e-01 4.10491616e-01 9.18490648e-01 -2.59557635e-01
-3.47749531e-01 7.22388983e-01 -3.04325461e-01 -1.67597428e-01
2.72809297e-01 1.02456056e-01 3.29676300e-01 1.76098213e-01
8.81703734e-01 2.69992687e-02 -6.94375634e-01 7.80439258e-01
-6.06334627e-01 1.29036726e-02 6.82147741e-01 1.26960981e+00
6.16431355e-01 2.24567298e-02 4.79944587e-01 -2.67906994e-01
4.32420045e-01 -2.93024689e-01 -3.91478866e-01 6.87386513e-01
-1.07301980e-01 -2.30832249e-01 3.79210055e-01 -5.93113661e-01
-1.24710226e+00 9.11315680e-02 -7.44214794e-03 -6.50213420e-01
-3.45903337e-01 3.08493376e-01 -5.82965136e-01 -6.11435249e-02
3.52545083e-01 6.01151764e-01 -2.34641463e-01 -4.45536762e-01
7.43129253e-02 6.09564364e-01 8.62251103e-01 -7.52929270e-01
8.67443860e-01 5.25532544e-01 -2.54975885e-01 -7.06022024e-01
-1.01343584e+00 -7.49130547e-01 -9.39427018e-01 -3.35872859e-01
1.21662974e+00 -1.34506738e+00 -5.55454135e-01 9.20887232e-01
-1.11527956e+00 7.24982247e-02 2.61763483e-01 3.68347526e-01
-3.97520006e-01 6.94189429e-01 -2.45769382e-01 -6.43757164e-01
-3.69186938e-01 -1.19987416e+00 1.31710982e+00 4.45184082e-01
2.22378120e-01 -9.91107285e-01 -7.03302994e-02 7.62947202e-01
7.88680092e-02 3.14539969e-01 4.61895049e-01 -6.76138163e-01
-4.10400063e-01 -3.07737023e-01 -5.57459891e-01 1.79113105e-01
4.30036448e-02 -8.09984565e-01 -6.77494586e-01 -5.60142815e-01
-3.67616177e-01 -5.40068686e-01 8.69389355e-01 2.46240735e-01
1.02752471e+00 -9.34300870e-02 -7.42429495e-01 6.14401102e-01
1.11203778e+00 -9.75473449e-02 4.30725873e-01 6.72956645e-01
7.44003236e-01 6.75160050e-01 9.08221960e-01 3.32675397e-01
8.83015394e-01 1.31089342e+00 -2.37126406e-02 -8.97042304e-02
-6.73858896e-02 -4.25244898e-01 2.09467217e-01 2.34817445e-01
-9.56870839e-02 -1.36367843e-01 -9.97772455e-01 7.89164186e-01
-2.19678879e+00 -1.26355267e+00 1.84979260e-01 2.01137042e+00
6.70645654e-01 -1.37778565e-01 5.21569371e-01 -3.90262753e-01
1.02217877e+00 9.20743681e-03 -5.32994092e-01 4.64121729e-01
-2.43214995e-01 -3.59335899e-01 2.72572666e-01 -3.72827835e-02
-1.77120280e+00 1.02861428e+00 4.92049837e+00 9.45219219e-01
-6.78744316e-01 2.71185964e-01 4.43942159e-01 2.17707425e-01
3.82214516e-01 -4.36849445e-01 -1.07186627e+00 7.55797029e-01
1.97589591e-01 -6.06771260e-02 1.09992683e-01 1.05558789e+00
-2.41161287e-01 8.91636088e-02 -1.10340846e+00 1.68597937e+00
4.00370598e-01 -7.35288441e-01 2.35543683e-01 -6.81772381e-02
6.66401505e-01 -4.04059410e-01 2.36734867e-01 6.27682269e-01
6.91479892e-02 -8.22310686e-01 5.52077770e-01 5.39973080e-01
7.92234719e-01 -7.58858860e-01 9.56896007e-01 4.58422482e-01
-1.59343767e+00 -2.74453849e-01 -5.41756809e-01 4.24050629e-01
3.62401664e-01 6.00554282e-03 -4.57740426e-01 7.69040585e-01
1.19593835e+00 8.75543952e-01 -6.34714723e-01 1.08730829e+00
1.61975414e-01 -1.85322702e-01 -2.75247306e-01 2.76508182e-01
2.83246726e-01 -6.88233525e-02 5.36326647e-01 1.27182674e+00
2.73710608e-01 2.55709868e-02 6.36775613e-01 6.59785986e-01
1.33733600e-01 -1.19514003e-01 -4.83564168e-01 3.06573302e-01
4.00713742e-01 8.84415984e-01 -3.59181166e-01 -3.70686620e-01
-6.29415512e-01 1.44262993e+00 3.24835002e-01 3.12387288e-01
-1.01957500e+00 -1.07578635e-01 7.74592578e-01 -1.55481726e-01
2.91867882e-01 -8.00190866e-02 2.89627820e-01 -1.42045116e+00
2.90214568e-01 -9.70461488e-01 9.00357246e-01 -6.01911068e-01
-1.82590568e+00 5.10831475e-01 3.64323884e-01 -1.42623687e+00
-2.33182654e-01 -5.14257848e-01 -1.57353610e-01 9.45608616e-01
-1.71969771e+00 -1.83125126e+00 -6.11254334e-01 1.20878422e+00
7.21087098e-01 -5.21250546e-01 5.35655558e-01 8.12224150e-01
-6.46328032e-01 1.05825508e+00 -1.31001487e-01 6.39268100e-01
1.12005317e+00 -8.25574994e-01 2.57348865e-01 7.24334538e-01
-2.76502877e-01 7.55013525e-01 5.07346034e-01 -6.37993276e-01
-1.24267364e+00 -1.28238273e+00 8.37335289e-01 -6.93645120e-01
4.04236883e-01 -1.21753454e-01 -6.66242599e-01 7.23081231e-01
3.13628614e-02 5.64783588e-02 5.96499979e-01 -6.54091174e-03
-4.52968866e-01 -1.13786444e-01 -1.00303721e+00 5.61995983e-01
1.44048798e+00 -4.55485910e-01 -6.52451754e-01 5.35543203e-01
4.47407961e-01 -4.88306522e-01 -6.45781815e-01 5.17644882e-01
6.20596886e-01 -9.66055155e-01 1.65690410e+00 -7.28363037e-01
9.14231986e-02 -3.84196997e-01 -1.99221671e-01 -1.14192295e+00
-3.45099717e-01 -5.16325496e-02 1.70492738e-01 1.35516036e+00
-9.74860489e-02 -5.93865633e-01 5.30435920e-01 5.05207002e-01
8.17512199e-02 -3.91955882e-01 -1.08941376e+00 -1.05917156e+00
6.90701604e-02 6.18205518e-02 5.77188253e-01 1.00613892e+00
-1.63194373e-01 4.19237763e-01 -9.90516126e-01 3.80893975e-01
1.03481925e+00 1.60005987e-01 1.04803121e+00 -1.27883661e+00
-2.00221479e-01 -3.55118692e-01 -5.25820434e-01 -1.40924370e+00
3.61981302e-01 -7.18720555e-01 8.69619008e-03 -1.43082833e+00
6.88968182e-01 -3.37904185e-01 -2.24702939e-01 2.30252862e-01
-4.52741116e-01 1.34392887e-01 2.92539984e-01 4.70557392e-01
-1.18207836e+00 8.27364206e-01 1.14219511e+00 -3.68138611e-01
2.22112946e-02 4.03166339e-02 -5.26903629e-01 6.90714002e-01
4.39565688e-01 -3.54035012e-02 -2.33975947e-01 -7.81732261e-01
-3.19903076e-01 -4.07034419e-02 8.45257521e-01 -1.27610052e+00
5.28481007e-01 1.64023996e-03 8.09293032e-01 -7.66411543e-01
6.12132311e-01 -9.50461566e-01 7.76708648e-02 2.99543172e-01
-2.45105132e-01 -1.41390935e-02 9.33336318e-02 8.79843891e-01
-4.16204304e-01 1.60094053e-01 8.86365652e-01 -3.04120034e-01
-1.24380302e+00 6.70117438e-01 3.02980483e-01 1.98648810e-01
1.13130999e+00 -2.17672616e-01 -3.36800247e-01 -2.64980137e-01
-6.81219637e-01 7.80465961e-01 3.84003699e-01 7.63267398e-01
5.85247517e-01 -1.62487805e+00 -7.43896723e-01 8.41079578e-02
5.70231438e-01 -7.07899779e-02 7.37509012e-01 7.87457466e-01
-7.70377293e-02 7.21126437e-01 -3.73861283e-01 -8.36681187e-01
-1.37933636e+00 8.26096296e-01 4.88783807e-01 -4.69263107e-01
-5.25876701e-01 1.05615437e+00 5.97787559e-01 -6.76425397e-01
2.88658112e-01 5.76744378e-01 -3.40096444e-01 7.80612184e-03
6.07137144e-01 2.06369132e-01 -4.42073613e-01 -1.17743015e+00
-6.42128587e-01 1.08536828e+00 -1.84239030e-01 1.73162967e-01
8.81503999e-01 -3.83349508e-01 9.16012973e-02 -6.96608573e-02
1.14127159e+00 -5.73285878e-01 -1.14515781e+00 -7.66703606e-01
-2.73150648e-03 -6.31007791e-01 -4.95415360e-01 -7.89080679e-01
-8.21024776e-01 8.53872836e-01 8.13970625e-01 -3.44619095e-01
1.12958789e+00 2.96392560e-01 1.30680513e+00 4.23893154e-01
6.39624178e-01 -1.30823851e+00 3.39163512e-01 4.34001207e-01
8.15268278e-01 -1.62365997e+00 -1.45928472e-01 -3.12065840e-01
-9.02569592e-01 7.49029219e-01 1.03960705e+00 2.77775109e-01
4.41775590e-01 -5.88022947e-01 -1.24938451e-01 -5.94040453e-02
-3.07829112e-01 -5.89586139e-01 4.98618722e-01 1.05299759e+00
-1.05506599e-01 -7.89709538e-02 -2.47446280e-02 7.95008063e-01
2.20082607e-02 -8.46514255e-02 -2.77355134e-01 6.39941573e-01
-2.56267399e-01 -1.10128796e+00 -5.75667202e-01 9.73415598e-02
-1.87278315e-01 1.47320405e-01 -4.32245642e-01 1.01301563e+00
3.43336433e-01 9.57777679e-01 -2.01782823e-01 -3.10210973e-01
5.98965287e-01 1.79758623e-01 4.81882840e-01 -9.25767794e-02
-8.91203135e-02 -8.54458213e-02 -6.91308007e-02 -3.17967683e-01
-8.31111491e-01 -8.84427071e-01 -6.54141784e-01 -2.30328709e-01
-1.18689597e-01 6.66552708e-02 2.00193182e-01 1.04591954e+00
2.62557894e-01 2.35788047e-01 1.44903734e-01 -8.10026228e-01
-4.78764683e-01 -8.77669334e-01 -2.27429971e-01 8.25049162e-01
3.06837350e-01 -1.20908916e+00 6.42963797e-02 1.91528305e-01] | [14.682085990905762, 0.8722730278968811] |
adca26b0-66a1-4574-90f0-1547d73114bd | active-implicit-object-reconstruction-using | 2303.16739 | null | https://arxiv.org/abs/2303.16739v2 | https://arxiv.org/pdf/2303.16739v2.pdf | Active Implicit Object Reconstruction using Uncertainty-guided Next-Best-View Optimziation | Actively planning sensor views during object reconstruction is crucial for autonomous mobile robots. An effective method should be able to strike a balance between accuracy and efficiency. In this paper, we propose a seamless integration of the emerging implicit representation with the active reconstruction task. We build an implicit occupancy field as our geometry proxy. While training, the prior object bounding box is utilized as auxiliary information to generate clean and detailed reconstructions. To evaluate view uncertainty, we employ a sampling-based approach that directly extracts entropy from the reconstructed occupancy probability field as our measure of view information gain. This eliminates the need for additional uncertainty maps or learning. Unlike previous methods that compare view uncertainty within a finite set of candidates, we aim to find the next-best-view (NBV) on a continuous manifold. Leveraging the differentiability of the implicit representation, the NBV can be optimized directly by maximizing the view uncertainty using gradient descent. It significantly enhances the method's adaptability to different scenarios. Simulation and real-world experiments demonstrate that our approach effectively improves reconstruction accuracy and efficiency of view planning in active reconstruction tasks. The proposed system will open source at https://github.com/HITSZ-NRSL/ActiveImplicitRecon.git. | ['Mengmeng Fu', 'Haoyao Chen', 'Fengyu Quan', 'Jianheng Liu', 'Dongyu Yan'] | 2023-03-29 | null | null | null | null | ['object-reconstruction'] | ['computer-vision'] | [ 8.46752375e-02 4.26944822e-01 -2.30546787e-01 -4.86907542e-01
-1.07409644e+00 -5.32059312e-01 5.99590182e-01 -1.99001543e-02
-4.16364312e-01 7.25331426e-01 3.05310309e-01 6.64762512e-04
-2.14237198e-01 -1.06095099e+00 -9.47658122e-01 -7.88183749e-01
2.61064380e-01 5.83026707e-01 2.22612932e-01 -6.52912036e-02
4.60496962e-01 4.31878388e-01 -1.59545231e+00 -2.44721815e-01
1.24924636e+00 1.05832613e+00 7.81864464e-01 3.26826125e-01
3.34645845e-02 3.48349482e-01 -4.19619791e-02 1.53624695e-02
5.42637706e-01 -1.84375281e-03 -4.54074472e-01 7.26557598e-02
2.64752191e-02 -4.60910350e-01 -6.82977587e-02 1.07477701e+00
4.60744679e-01 2.17515528e-01 6.85667574e-01 -8.81237447e-01
2.60742530e-02 1.57362923e-01 -3.64299655e-01 -1.33350477e-01
4.54235792e-01 8.57397765e-02 7.32647002e-01 -1.29204178e+00
7.24433482e-01 1.01239622e+00 5.71714520e-01 3.37949932e-01
-1.00178885e+00 -3.77535820e-01 2.82763720e-01 5.13998717e-02
-1.55193233e+00 -7.15688825e-01 8.77447426e-01 -2.83953726e-01
5.82720339e-01 1.79510921e-01 8.02757680e-01 8.83919597e-01
4.29732054e-01 6.87675536e-01 8.49795640e-01 -2.77317017e-01
5.85543692e-01 2.52315462e-01 -3.20507765e-01 9.05455947e-01
3.40554267e-01 2.09190130e-01 -6.59343481e-01 -1.72442660e-01
8.58908832e-01 6.46371320e-02 -4.26418751e-01 -1.14555025e+00
-1.26191032e+00 8.21991920e-01 4.88348573e-01 -2.19546720e-01
-4.12244827e-01 2.19437897e-01 -1.15332767e-01 -9.13807973e-02
3.92377079e-01 3.01306218e-01 -1.90246612e-01 -6.47307262e-02
-5.26969969e-01 1.80092573e-01 6.32715762e-01 1.07541597e+00
1.00815439e+00 -1.92237094e-01 2.95098752e-01 6.47029161e-01
8.15222383e-01 6.52092755e-01 -1.40164137e-01 -1.34529042e+00
5.10661781e-01 5.36812127e-01 4.68949944e-01 -7.70033717e-01
-4.35887948e-02 -3.81922841e-01 -4.92912710e-01 5.70654154e-01
2.29342312e-01 1.02863483e-01 -8.00655425e-01 1.43873882e+00
7.73508489e-01 -9.64478925e-02 5.94826303e-02 8.66283059e-01
3.63489509e-01 5.65399230e-01 -5.43214917e-01 -3.90961587e-01
9.20027733e-01 -7.68856943e-01 -6.06140018e-01 -2.56582290e-01
2.38110557e-01 -4.65164185e-01 8.54706168e-01 3.31439197e-01
-1.05802739e+00 -1.18781663e-01 -1.39035201e+00 3.23002599e-02
-1.11338787e-01 -2.98109613e-02 3.34292680e-01 3.81084949e-01
-8.96360219e-01 4.57398325e-01 -1.06061566e+00 -2.04972610e-01
4.44676280e-01 3.86830568e-01 -2.59297192e-01 -1.83538973e-01
-8.24771643e-01 9.44561422e-01 4.75451559e-01 -6.06974727e-03
-1.05605721e+00 -5.32097101e-01 -9.34458733e-01 -4.39910591e-01
7.53534317e-01 -9.52153444e-01 1.09157860e+00 -4.55001652e-01
-1.73524296e+00 4.46521550e-01 -2.39983067e-01 -3.32500309e-01
6.87486470e-01 -2.74436653e-01 2.24495023e-01 1.31227076e-01
2.38586023e-01 7.18284309e-01 8.03827584e-01 -1.88509130e+00
-5.51713407e-01 -5.34731746e-01 2.95149177e-01 8.68837416e-01
2.27707148e-01 -8.91473472e-01 -6.14857793e-01 -1.28191039e-01
7.58711338e-01 -9.40180004e-01 -4.26934838e-01 3.40913832e-01
-1.16283521e-01 2.34626725e-01 6.81846559e-01 -4.06125307e-01
8.19202662e-01 -1.88708019e+00 2.73728281e-01 3.25328588e-01
1.46428719e-01 -4.51744258e-01 3.72710884e-01 2.42071718e-01
7.57616103e-01 -7.29205683e-02 -5.56209862e-01 -5.25955796e-01
-1.69437006e-01 3.40864331e-01 -1.71481505e-01 7.47701824e-01
-6.21390678e-02 7.46510506e-01 -8.91453147e-01 -4.76610541e-01
5.25354147e-01 5.01164794e-01 -6.40214741e-01 1.46054968e-01
-2.32991844e-01 7.96218336e-01 -6.05458140e-01 7.01691568e-01
7.85020769e-01 -2.17021748e-01 1.03402741e-01 -1.21397249e-01
-3.83216769e-01 2.90316790e-01 -1.26553631e+00 2.11977077e+00
-7.23830938e-01 2.27455139e-01 3.28185111e-01 -6.93291247e-01
8.75024319e-01 -4.35688943e-02 7.04588890e-01 -5.24246275e-01
1.26516685e-01 3.15088689e-01 -4.91933972e-01 -2.28548720e-01
6.41504526e-01 -9.19454545e-03 -1.88267976e-01 1.72045290e-01
-1.81897759e-01 -4.88890022e-01 -3.73688698e-01 5.70291169e-02
8.37633252e-01 5.86113095e-01 7.94618189e-01 -4.52297181e-01
4.01212573e-01 -6.78086057e-02 6.90520406e-01 4.58469391e-01
-3.09930835e-02 6.79395497e-01 -1.55745089e-01 -1.03010699e-01
-1.04291999e+00 -1.43722534e+00 -4.12370324e-01 4.14751410e-01
8.58182967e-01 -2.14351013e-01 -5.90647697e-01 -6.05913460e-01
-4.53356281e-02 7.68050075e-01 -5.37241399e-01 5.31350635e-02
-5.68736315e-01 -3.41908336e-01 -1.75010383e-01 2.39030406e-01
5.36398113e-01 -5.69029927e-01 -9.84625638e-01 1.07420154e-01
-4.40799415e-01 -8.12907815e-01 -2.80250549e-01 2.65614063e-01
-1.05623484e+00 -8.49444330e-01 -3.83817375e-01 -2.68603265e-01
9.27918196e-01 3.89575154e-01 9.24402654e-01 -2.16215178e-01
1.86470225e-02 7.28779495e-01 -3.69806945e-01 -3.65526319e-01
-2.09492430e-01 -2.16064677e-02 1.94978237e-01 -2.74768889e-01
-2.67163813e-01 -8.61026287e-01 -8.14555824e-01 4.12489027e-01
-5.16753852e-01 2.65059263e-01 4.86884207e-01 6.86408281e-01
1.09282446e+00 -2.73289591e-01 2.65819371e-01 -4.76704419e-01
1.56721845e-02 -6.55783534e-01 -9.03552592e-01 8.67427886e-02
-7.32315242e-01 3.88140976e-01 1.69215530e-01 -7.78287202e-02
-1.25142431e+00 4.61601794e-01 -6.54121265e-02 -4.28364456e-01
9.15529877e-02 4.04943526e-01 -4.65693206e-01 -1.35799810e-01
4.44305688e-01 1.61989093e-01 5.38847186e-02 -2.20384896e-01
4.61083680e-01 4.55583304e-01 2.35548645e-01 -5.43084919e-01
7.26507783e-01 1.00908136e+00 -1.56111578e-02 -7.98888564e-01
-6.52615368e-01 -5.41672707e-01 -7.39963651e-01 -5.75506985e-01
8.48363042e-01 -9.04134035e-01 -5.53103328e-01 5.41576790e-03
-1.01554847e+00 -1.14285216e-01 -5.42721450e-01 6.42651260e-01
-9.69294071e-01 3.76720041e-01 2.29913980e-01 -1.21118724e+00
-3.19378227e-01 -1.26855493e+00 1.26491964e+00 2.86351610e-02
1.06902234e-01 -8.77668321e-01 1.96713313e-01 3.97303522e-01
1.29779994e-01 4.43318725e-01 2.59619117e-01 -2.32779030e-02
-1.32203102e+00 4.56962995e-02 1.27051085e-01 2.92509561e-03
6.94133639e-02 -2.35241354e-01 -9.50313926e-01 -3.27048153e-01
4.09019947e-01 -2.35325232e-01 8.12205136e-01 4.69559401e-01
8.38107347e-01 -2.73959696e-01 -5.02651453e-01 7.80610740e-01
1.70988321e+00 3.20274413e-01 7.20069706e-01 3.07940602e-01
5.93638957e-01 5.07960141e-01 1.05749166e+00 6.77863300e-01
7.15753913e-01 7.46514261e-01 1.01210010e+00 4.83234674e-01
1.73501983e-01 -3.78563851e-01 4.30836469e-01 9.37083662e-01
-1.22667529e-01 -3.47878337e-01 -9.91782129e-01 3.47554833e-01
-1.85602736e+00 -7.86388874e-01 2.21917450e-01 2.60025430e+00
4.26645577e-01 2.39522308e-01 -4.60607320e-01 -1.69639915e-01
4.27630812e-01 1.13477215e-01 -7.60502458e-01 2.19084516e-01
3.16969007e-01 -2.65449286e-01 6.96515739e-01 1.02473569e+00
-9.39407110e-01 6.07625961e-01 5.14442301e+00 6.57592535e-01
-8.07901919e-01 2.13673905e-01 2.73406327e-01 -1.72174469e-01
-7.22289026e-01 1.87059119e-01 -9.67921138e-01 3.24768066e-01
5.63574135e-01 3.51403542e-02 4.59814757e-01 8.91122103e-01
2.29768127e-01 -6.36595726e-01 -9.10848200e-01 9.79084909e-01
2.64689326e-02 -1.41938913e+00 -1.58217445e-01 4.18065846e-01
6.68037713e-01 1.86660156e-01 -1.27764419e-01 2.50841752e-02
1.05558276e-01 -6.63488865e-01 1.21143007e+00 8.79927874e-01
7.17439711e-01 -7.84948945e-01 4.53842223e-01 7.82512963e-01
-1.39930141e+00 -8.99595469e-02 -2.88831085e-01 1.45587489e-01
5.10151982e-01 5.88776052e-01 -1.10173118e+00 8.34982872e-01
5.29838502e-01 5.19476354e-01 -1.37855455e-01 9.20863509e-01
7.24338517e-02 -1.17183708e-01 -6.80053353e-01 1.21447742e-01
-9.80531797e-02 -5.42095065e-01 1.05757689e+00 5.12273729e-01
5.55914938e-01 1.95412815e-01 2.61726975e-01 9.14218605e-01
1.64687321e-01 -8.89763981e-02 -8.79377902e-01 3.49263549e-01
7.88969338e-01 1.10889220e+00 -9.43437338e-01 -1.09067872e-01
-1.71319813e-01 8.51314902e-01 4.15868968e-01 1.53128719e-02
-9.43815053e-01 1.91241845e-01 3.62927973e-01 2.78970212e-01
3.60421151e-01 -4.46529120e-01 -5.36575437e-01 -1.17525613e+00
1.98098749e-01 -3.49112660e-01 -9.85490382e-02 -6.40167952e-01
-8.27488244e-01 3.81180972e-01 3.88403624e-01 -1.71552050e+00
-3.15958977e-01 -2.09265381e-01 -1.95232481e-01 6.38564706e-01
-1.18235815e+00 -9.02706027e-01 -4.70777571e-01 3.14007998e-01
8.12166035e-01 1.43014356e-01 5.57916462e-01 7.46284947e-02
-1.47046968e-02 1.02404252e-01 2.13859066e-01 -5.16048133e-01
3.05675328e-01 -1.11726785e+00 1.66321740e-01 7.38465548e-01
-4.79169264e-02 5.13175905e-01 7.80536115e-01 -8.92212510e-01
-1.56284285e+00 -9.43263531e-01 -1.03364410e-02 -5.97728014e-01
4.91345637e-02 -3.92890066e-01 -5.56054056e-01 4.42244262e-01
-1.11581273e-02 -6.75134808e-02 2.33828023e-01 -2.74041593e-01
6.26443252e-02 -7.41724819e-02 -1.35827148e+00 5.71686625e-01
1.25336921e+00 -2.95395851e-01 -4.28471178e-01 -2.51238812e-02
7.51240730e-01 -6.75663948e-01 -8.40808570e-01 9.12303329e-01
7.90186882e-01 -1.08032882e+00 1.03240597e+00 3.48898798e-01
1.05721943e-01 -4.86247063e-01 -6.92236841e-01 -1.14287961e+00
-1.85317621e-02 -3.88537288e-01 -3.66637498e-01 8.98437977e-01
3.56776863e-01 -7.70712912e-01 7.97946274e-01 2.78812677e-01
-4.60131556e-01 -1.19548070e+00 -1.01774228e+00 -6.38996840e-01
-3.43289793e-01 -4.21836853e-01 3.73133719e-01 5.66413939e-01
-3.87358397e-01 1.99271426e-01 -1.01410396e-01 5.94339252e-01
1.02725756e+00 3.57212611e-02 8.39706659e-01 -1.19682837e+00
-2.73863345e-01 9.72517207e-02 -1.27777696e-01 -1.37977982e+00
-1.91219598e-01 -6.50481224e-01 5.47080219e-01 -1.90212476e+00
3.72103713e-02 -7.87124693e-01 1.42281041e-01 -8.91047493e-02
1.75298199e-01 -5.17104976e-02 8.27626884e-02 4.28784072e-01
-5.55833936e-01 1.07957077e+00 1.30451810e+00 1.35267794e-01
-3.72506648e-01 1.00984968e-01 -3.49902630e-01 1.01513493e+00
9.44297910e-01 -4.06190008e-01 -7.38900125e-01 -3.25094253e-01
3.66037101e-01 4.10527080e-01 2.88319498e-01 -1.08110309e+00
1.37800246e-01 -3.95109266e-01 2.35487774e-01 -1.10702109e+00
1.01751995e+00 -9.81417477e-01 4.79604125e-01 2.79945046e-01
3.10127437e-02 -7.51740709e-02 -1.69599727e-01 9.98196304e-01
1.65100321e-01 -4.75244105e-01 7.69137084e-01 -4.49815542e-01
-6.24758065e-01 4.10661101e-01 -1.87276453e-01 -2.28386968e-01
1.13118207e+00 -6.26629353e-01 -1.08129896e-01 -3.76234800e-01
-5.82210600e-01 3.09198290e-01 1.05075693e+00 1.90388739e-01
1.00718081e+00 -1.30597711e+00 -3.02200615e-01 1.16039924e-01
1.69518247e-01 5.85571945e-01 2.04972669e-01 8.65683854e-01
-6.18735254e-01 1.77797928e-01 3.83971073e-02 -9.98476923e-01
-9.02031839e-01 2.57283688e-01 4.04593438e-01 -8.57599825e-02
-7.01107800e-01 6.68620408e-01 3.05731446e-01 -7.03898311e-01
1.30545169e-01 -3.55641186e-01 -8.20725635e-02 -2.46862019e-03
1.23172484e-01 5.53160548e-01 -7.66384974e-02 -6.28403902e-01
-3.07767272e-01 6.21646225e-01 2.27500513e-01 -5.61014175e-01
1.31682682e+00 -7.76118517e-01 1.28899395e-01 6.97669387e-01
9.98413563e-01 3.81186187e-01 -1.77499318e+00 7.62326270e-02
-1.73923865e-01 -6.32349968e-01 2.07181513e-01 -4.41714317e-01
-5.64121723e-01 4.55553412e-01 7.96966791e-01 1.40669961e-02
7.56624043e-01 1.14262141e-01 3.43489826e-01 4.50623900e-01
1.17947781e+00 -8.88223529e-01 2.02550024e-01 4.48036104e-01
1.18205249e+00 -1.39967525e+00 4.59231198e-01 -6.91981196e-01
-5.20348430e-01 8.36882710e-01 5.44704497e-01 -1.32666379e-01
7.39488423e-01 3.17854255e-01 -1.49636835e-01 -3.24190855e-01
-5.57498753e-01 -8.37705210e-02 1.19968183e-01 4.20238882e-01
-1.65381834e-01 8.93191434e-03 -1.10120513e-01 6.19978048e-02
-2.87047237e-01 -3.23057622e-01 4.65758055e-01 1.22635508e+00
-8.21574986e-01 -8.94500911e-01 -4.50442255e-01 4.00460392e-01
5.13248816e-02 2.13231415e-01 1.00462608e-01 6.69915318e-01
2.00866207e-01 6.67524040e-01 2.97773201e-02 -3.57051194e-01
7.85124302e-02 -1.49810642e-01 7.83545792e-01 -6.77018762e-01
2.15843111e-01 1.37594379e-02 9.01299268e-02 -7.90997624e-01
-4.14940417e-01 -8.23828459e-01 -1.42497754e+00 3.86349298e-02
-6.14230812e-01 9.00002122e-02 8.74292135e-01 8.30564559e-01
3.37156057e-01 2.65735298e-01 7.58260727e-01 -1.18374205e+00
-5.10696769e-01 -6.77736998e-01 -2.61887610e-01 -2.40791813e-01
4.12250429e-01 -1.19642293e+00 -4.07912850e-01 -1.75583243e-01] | [8.464186668395996, -2.658402681350708] |
d613cf96-bb89-4e4b-b00d-0e1feb1b7d9b | 3d-face-reconstruction-with-dense-landmarks | 2204.02776 | null | https://arxiv.org/abs/2204.02776v2 | https://arxiv.org/pdf/2204.02776v2.pdf | 3D face reconstruction with dense landmarks | Landmarks often play a key role in face analysis, but many aspects of identity or expression cannot be represented by sparse landmarks alone. Thus, in order to reconstruct faces more accurately, landmarks are often combined with additional signals like depth images or techniques like differentiable rendering. Can we keep things simple by just using more landmarks? In answer, we present the first method that accurately predicts 10x as many landmarks as usual, covering the whole head, including the eyes and teeth. This is accomplished using synthetic training data, which guarantees perfect landmark annotations. By fitting a morphable model to these dense landmarks, we achieve state-of-the-art results for monocular 3D face reconstruction in the wild. We show that dense landmarks are an ideal signal for integrating face shape information across frames by demonstrating accurate and expressive facial performance capture in both monocular and multi-view scenarios. This approach is also highly efficient: we can predict dense landmarks and fit our 3D face model at over 150FPS on a single CPU thread. Please see our website: https://microsoft.github.io/DenseLandmarks/. | ['Julien Valentin', 'Tom Cashman', 'Ivan Stojiljkovic', 'Toby Sharp', 'Jamie Shotton', 'Chirag Raman', 'Stephan Garbin', 'Daniel Wilde', 'Nikola Milosavljevic', 'Jingjing Shen', 'Matthew Johnson', 'Charlie Hewitt', 'Tadas Baltrusaitis', 'Erroll Wood'] | 2022-04-06 | null | null | null | null | ['3d-face-reconstruction', 'face-model', 'face-reconstruction'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [-2.45728105e-01 2.34295934e-01 2.10402682e-02 -6.23990297e-01
-8.69812965e-01 -4.82746422e-01 5.20325005e-01 -4.95182097e-01
-8.45386237e-02 4.96746540e-01 3.70962650e-01 2.62509823e-01
5.09645522e-01 -5.09609759e-01 -7.95333803e-01 -4.11038429e-01
6.90108836e-02 5.84251344e-01 -3.30558628e-01 8.31159763e-03
-9.43067744e-02 9.38758433e-01 -1.70717096e+00 -1.56883430e-02
1.53334767e-01 1.00902438e+00 -2.18956769e-01 5.26638389e-01
-1.77449629e-01 3.61361384e-01 -2.34968066e-01 -6.77448273e-01
5.37225127e-01 -2.16389015e-01 -5.48033297e-01 2.16163605e-01
8.58080149e-01 -8.30861151e-01 -2.70002782e-01 8.78798962e-01
7.17212856e-01 -2.22601563e-01 2.34667242e-01 -1.06392968e+00
-3.00131679e-01 5.26828580e-02 -9.98892367e-01 -2.63576955e-01
8.05112541e-01 -2.13674139e-02 6.66323125e-01 -1.30136132e+00
8.17034185e-01 1.41230726e+00 8.18310797e-01 9.83530164e-01
-1.32911646e+00 -8.71194124e-01 -4.38236110e-02 -2.02922508e-01
-1.64941812e+00 -1.33934176e+00 7.87981808e-01 -2.51657784e-01
4.78584468e-01 3.35570157e-01 7.34973788e-01 8.97753179e-01
-1.80679724e-01 5.95031321e-01 8.98722112e-01 -2.83912688e-01
-1.01005971e-01 -5.48466854e-02 -4.87796545e-01 1.13888574e+00
-1.48089200e-01 7.73759633e-02 -8.54024053e-01 -2.35033244e-01
1.01409745e+00 1.17648944e-01 -3.91641617e-01 -2.34858289e-01
-9.97148156e-01 5.14100492e-01 2.82874435e-01 -3.88882682e-02
-2.08294526e-01 4.64047194e-01 1.13656431e-01 1.37268826e-01
8.22160304e-01 -4.13908949e-03 -3.80443215e-01 -2.18647763e-01
-1.23738277e+00 1.17873073e-01 5.85399687e-01 9.85831618e-01
1.03142679e+00 1.20721117e-01 3.14225107e-01 8.24224174e-01
4.45004225e-01 7.79563189e-01 9.21759158e-02 -1.65582812e+00
-1.33925676e-01 1.73092097e-01 7.19838068e-02 -9.14347827e-01
-4.32003528e-01 1.63461398e-02 -6.42152369e-01 4.47729379e-01
4.97774065e-01 -2.13665187e-01 -8.06628942e-01 1.78965676e+00
8.74557674e-01 4.89446163e-01 -5.66384494e-01 9.19081509e-01
1.06651151e+00 3.61222118e-01 -3.54598016e-01 -1.28323913e-01
1.39236903e+00 -5.77731788e-01 -5.73710382e-01 -1.62864953e-01
2.93363214e-01 -9.13377762e-01 9.39752579e-01 2.98914343e-01
-1.38430130e+00 -3.20517868e-01 -6.28751755e-01 -4.50018346e-01
1.41324401e-01 2.71679405e-02 7.46509612e-01 7.26457179e-01
-1.47622478e+00 4.89047438e-01 -9.76374149e-01 -1.51859030e-01
6.80712581e-01 5.82518160e-01 -9.54727530e-01 -2.27739975e-01
-4.86527741e-01 6.13331854e-01 -6.22301817e-01 1.19081907e-01
-7.88774014e-01 -1.00256598e+00 -1.04806650e+00 -5.15807271e-02
4.53494564e-02 -7.45988190e-01 1.21902561e+00 -8.98817420e-01
-1.64510262e+00 1.41973531e+00 -7.97107399e-01 5.44706173e-02
4.40091133e-01 -1.19754389e-01 1.16404615e-01 4.23878729e-01
-7.50628933e-02 9.84721780e-01 9.69487727e-01 -1.29154992e+00
-1.12106465e-01 -8.90533328e-01 1.46477418e-02 5.08477986e-02
-3.02849244e-02 3.51110220e-01 -6.54752910e-01 -3.47074509e-01
2.92580932e-01 -8.87415051e-01 8.43979940e-02 9.19833839e-01
-2.54432470e-01 2.35669672e-01 6.07220054e-01 -8.67406607e-01
4.41565365e-01 -2.29855180e+00 -1.32481176e-02 2.01683879e-01
5.15833855e-01 -4.97335829e-02 3.45980539e-03 -1.23426959e-01
4.09956761e-02 3.22096609e-02 2.06819531e-02 -1.18595910e+00
1.47365615e-01 -8.22420232e-03 -9.17142257e-02 8.67118418e-01
-1.38009220e-01 1.04793251e+00 -6.01811469e-01 -5.41234076e-01
1.85058698e-01 1.10143316e+00 -6.84253156e-01 9.30600166e-02
-4.29123789e-02 8.10848296e-01 -6.20829500e-02 1.04861116e+00
1.01939321e+00 -1.70657590e-01 9.94173437e-02 -8.22345912e-02
1.95779607e-01 2.75458246e-02 -1.08384991e+00 2.00467348e+00
-5.26441216e-01 6.35415018e-01 7.72693515e-01 -4.59209234e-01
9.14297640e-01 4.58684713e-01 8.10516477e-01 -6.29071772e-01
1.82325076e-02 2.07794562e-01 -5.43968856e-01 -1.04326066e-02
7.07798079e-02 -3.86930346e-01 4.13118124e-01 5.54742873e-01
1.30271003e-01 -4.58856314e-01 -4.38590229e-01 -3.68327200e-02
7.66068935e-01 3.60732585e-01 5.83436005e-02 -2.45617386e-02
1.91834375e-01 -4.74029809e-01 6.11519575e-01 -8.97634402e-02
-8.70698616e-02 9.64105010e-01 4.40727830e-01 -5.70177853e-01
-9.96886134e-01 -1.01085806e+00 -3.43315035e-01 8.48479033e-01
-1.61472574e-01 -5.24024069e-01 -9.37899470e-01 -2.25042671e-01
5.67999259e-02 1.52449280e-01 -6.35657489e-01 2.87454963e-01
-5.05396664e-01 -4.02827114e-01 5.99005222e-01 2.64958173e-01
2.82682955e-01 -6.51072204e-01 -4.42910552e-01 -2.20831230e-01
-3.75395007e-02 -1.21748948e+00 -6.08225405e-01 -2.57809222e-01
-6.83540642e-01 -9.99249041e-01 -9.98833060e-01 -4.96576130e-01
9.57296968e-01 2.45292887e-01 1.18166482e+00 3.22645545e-01
-4.30389971e-01 5.48929095e-01 1.73297703e-01 -2.04919621e-01
-1.06259044e-02 -5.08992851e-01 1.73221141e-01 1.83525056e-01
2.38071531e-01 -9.09254849e-01 -7.99535811e-01 2.93050677e-01
-3.15314889e-01 9.99489650e-02 2.98440922e-02 5.52442551e-01
8.51038039e-01 -6.57596111e-01 1.23738542e-01 -8.00162375e-01
-4.79662307e-02 -3.31878811e-01 -7.08469748e-01 -1.25565335e-01
-5.09042740e-02 -2.12588713e-01 4.08276528e-01 -1.18148200e-01
-8.54182601e-01 4.13988233e-01 -5.46965957e-01 -7.78858066e-01
-2.52045363e-01 -7.65247941e-02 -2.20894009e-01 -5.16299009e-01
4.66477036e-01 1.52079225e-01 4.92395967e-01 -6.26026869e-01
4.01701152e-01 4.45700586e-01 4.65458125e-01 -6.93629622e-01
7.48225331e-01 9.84248340e-01 3.41359973e-01 -1.04510653e+00
-6.76277518e-01 -2.35015497e-01 -8.68635178e-01 -3.33084643e-01
5.00210106e-01 -1.09952354e+00 -1.10002482e+00 3.39702845e-01
-1.28355730e+00 -5.03315270e-01 -3.59315991e-01 2.34639004e-01
-7.44956672e-01 3.15541238e-01 -5.44533074e-01 -6.77156866e-01
-3.34528714e-01 -1.08903861e+00 1.67992043e+00 1.70830607e-01
-1.57063320e-01 -8.29541624e-01 2.97263805e-02 3.41573685e-01
3.01127911e-01 5.17788410e-01 3.24866623e-01 4.35803831e-03
-5.34462094e-01 -1.17074572e-01 -2.04683661e-01 1.12726435e-01
1.54346943e-01 1.18133925e-01 -1.40921640e+00 -2.62326241e-01
-1.89295281e-02 -3.89484435e-01 3.94386798e-01 4.54597533e-01
1.16549623e+00 -3.34906369e-01 -2.31712073e-01 1.38576663e+00
1.12271011e+00 -2.37662598e-01 6.94253147e-01 -2.63170928e-01
8.56874228e-01 6.23753369e-01 2.23880008e-01 5.59194982e-01
5.54484785e-01 9.67251658e-01 3.50917578e-01 -1.93522975e-01
-4.61235821e-01 -2.94009298e-01 2.30990365e-01 5.82804203e-01
-2.34002233e-01 4.03861910e-01 -9.26798522e-01 2.79379696e-01
-1.18388069e+00 -9.32832718e-01 1.27271727e-01 2.23408389e+00
9.67751801e-01 -5.06494880e-01 1.55574277e-01 -7.06194267e-02
3.79172713e-01 1.14080638e-01 -5.26135564e-01 -1.56552151e-01
-5.44226877e-02 5.09644926e-01 1.20505579e-01 8.37500095e-01
-8.29263628e-01 9.41856623e-01 6.82806110e+00 5.75842798e-01
-1.46950424e+00 2.92231023e-01 9.68120277e-01 -5.67081273e-01
-4.61632133e-01 -2.15229318e-01 -8.54411364e-01 4.24499243e-01
9.02637839e-01 -5.06141558e-02 6.55370891e-01 7.87859559e-01
1.79003134e-01 -3.57694738e-02 -1.07662630e+00 1.53034437e+00
3.51360738e-01 -1.37532103e+00 -2.74346620e-01 3.57837915e-01
6.58869147e-01 -2.47692112e-02 1.72810122e-01 -3.07407707e-01
3.00635453e-02 -1.37545562e+00 6.81459606e-01 6.97519839e-01
1.42959201e+00 -6.25772774e-01 2.27164745e-01 2.31797233e-01
-9.97240961e-01 4.94243860e-01 -5.97510636e-01 1.63438708e-01
1.87796757e-01 6.42727435e-01 -6.33055687e-01 2.01266795e-01
5.57919204e-01 6.66497171e-01 -3.50249529e-01 9.07020152e-01
-8.54343325e-02 2.39713922e-01 -8.07782710e-01 4.64802682e-01
-3.14945579e-01 -1.72834411e-01 3.01596045e-01 9.10957515e-01
6.05599105e-01 4.07551467e-01 -2.22900629e-01 7.66962707e-01
-4.78721112e-01 1.11540154e-01 -8.27442944e-01 4.21845883e-01
4.26574916e-01 1.36032379e+00 -5.14478147e-01 1.09249586e-03
-5.32200634e-01 9.91683781e-01 2.92839646e-01 1.22890055e-01
-7.06064165e-01 1.21011235e-01 1.05837142e+00 5.35372257e-01
3.68986651e-02 -3.64998043e-01 -1.84640020e-01 -1.24003804e+00
9.05959159e-02 -7.60630608e-01 -1.67210266e-01 -8.90517175e-01
-8.09885621e-01 5.94580472e-01 -3.15657139e-01 -9.65838850e-01
-3.13451797e-01 -4.75017130e-01 -3.46002847e-01 8.01253557e-01
-1.61715472e+00 -1.42995822e+00 -6.47391379e-01 8.70703638e-01
2.07529604e-01 4.83562201e-02 1.09106851e+00 5.55677891e-01
-2.31359839e-01 8.51449311e-01 -1.79803997e-01 2.05146447e-01
7.94411302e-01 -9.64608729e-01 5.62106490e-01 5.43908119e-01
5.93836963e-01 5.74326694e-01 4.33785051e-01 -3.14612031e-01
-1.76547337e+00 -7.43189096e-01 7.17124343e-01 -5.55623829e-01
1.67889327e-01 -6.01877630e-01 -5.62263310e-01 9.74808216e-01
-1.39524058e-01 7.24075973e-01 7.71112561e-01 1.49092944e-02
-4.41980362e-01 -2.07325354e-01 -1.41261196e+00 4.65817094e-01
1.37581575e+00 -6.33801281e-01 6.23027496e-02 4.08818126e-01
3.27077657e-01 -7.81649828e-01 -8.84717941e-01 7.97210559e-02
7.96211362e-01 -1.23189116e+00 1.17614257e+00 -1.69524252e-01
2.65826285e-01 -1.78686768e-01 -3.88395399e-01 -1.10655391e+00
8.60345289e-02 -1.01475441e+00 -3.28633517e-01 1.17478991e+00
1.11886986e-01 -5.95746398e-01 1.11324346e+00 8.58210981e-01
5.72951548e-02 -8.78780365e-01 -1.27229011e+00 -4.13475275e-01
-1.27818614e-01 -3.69055867e-01 1.00111222e+00 7.91698515e-01
-6.33080751e-02 -1.60307467e-01 -4.62813258e-01 -1.30221248e-02
8.22598219e-01 6.98887855e-02 1.11909091e+00 -1.17019618e+00
3.62294982e-03 -2.51403242e-01 -7.28750646e-01 -1.30845284e+00
6.03528082e-01 -9.25731122e-01 -2.61108607e-01 -1.24953783e+00
1.12859659e-01 -6.30957365e-01 4.35135633e-01 5.73258579e-01
2.36633092e-01 1.00869882e+00 2.13424414e-01 1.29985549e-02
-2.40383804e-01 5.37220359e-01 1.36432803e+00 2.10167438e-01
2.42932126e-01 -1.46008030e-01 -8.07831109e-01 9.55596268e-01
5.48795640e-01 -1.54753059e-01 -1.37461662e-01 -7.61829197e-01
9.55113471e-02 3.36604893e-01 5.14846921e-01 -7.87242949e-01
3.08980018e-01 -2.40152925e-02 7.65243113e-01 -1.69647187e-01
1.13750148e+00 -8.63493621e-01 5.01588583e-01 2.84472276e-02
1.38339609e-01 -4.76268195e-02 2.79092401e-01 8.81460905e-02
-1.12300046e-01 1.11150518e-01 1.01749301e+00 -2.80345082e-01
-3.61706883e-01 7.48945475e-01 4.28525954e-01 4.69264947e-02
9.08760846e-01 -2.88856834e-01 -1.83708936e-01 -6.94149256e-01
-9.92097020e-01 -1.50572449e-01 9.54972386e-01 -3.45821753e-02
8.33502531e-01 -1.47846401e+00 -9.60653782e-01 6.71579897e-01
-3.36755216e-01 6.96673021e-02 1.88652858e-01 8.54153156e-01
-8.55010033e-01 2.18278319e-01 -1.99876472e-01 -7.07557321e-01
-1.46169639e+00 1.97646350e-01 6.53756499e-01 4.44496989e-01
-8.06787372e-01 1.11677003e+00 3.44779730e-01 -4.90201384e-01
1.40389845e-01 1.16525590e-01 2.12491438e-01 -2.97383056e-04
9.27536309e-01 1.34800643e-01 1.29337227e-02 -1.05114591e+00
-4.68392164e-01 1.13269758e+00 4.09883529e-01 -2.75217891e-01
1.48443544e+00 -2.78051674e-01 -2.53407538e-01 1.77169055e-01
1.39125669e+00 6.14843309e-01 -1.52568257e+00 -1.82132255e-02
-5.23107409e-01 -1.05308068e+00 -5.48302121e-02 -4.15631562e-01
-1.51288915e+00 1.08512580e+00 3.87128264e-01 -3.03899229e-01
1.07470620e+00 2.22866982e-01 7.70613909e-01 -2.01823935e-01
8.74026537e-01 -4.99480128e-01 -1.12949125e-01 1.16886094e-01
1.01824927e+00 -1.15542877e+00 5.11293523e-02 -6.01629734e-01
-3.61597419e-01 1.03446901e+00 3.17525834e-01 -1.76919624e-01
9.71737742e-01 6.53522134e-01 1.35697857e-01 -2.77592450e-01
-5.91417730e-01 -3.77528965e-02 1.87591314e-01 7.20507622e-01
5.17552257e-01 7.98884965e-03 4.28386867e-01 2.13453978e-01
-5.97354889e-01 -1.22574106e-01 3.21318984e-01 4.43017036e-01
-1.56620950e-01 -1.01464880e+00 -5.12789011e-01 3.76369327e-01
-6.83319271e-01 1.77374613e-02 -2.07800940e-01 5.40997267e-01
-7.63829574e-02 6.32078588e-01 2.27889732e-01 -1.32054612e-01
1.46137252e-01 1.34639949e-01 9.11078095e-01 -6.24812603e-01
-1.23444699e-01 2.39293337e-01 -7.68614113e-02 -1.09449697e+00
-2.73128331e-01 -8.05526376e-01 -1.23343873e+00 -9.87060130e-01
7.42919371e-02 -1.86761111e-01 8.47212136e-01 5.48261344e-01
6.31305039e-01 -9.98194739e-02 6.59163237e-01 -1.47083116e+00
-1.06819704e-01 -5.91605961e-01 -6.63404226e-01 1.89021990e-01
5.07473290e-01 -6.52442157e-01 -4.75460112e-01 1.70329466e-01] | [13.17725944519043, -0.04004315659403801] |
17eb255d-b459-48d0-960c-775aa2ba2500 | shadow-removal-via-shadow-image-decomposition | 1908.08628 | null | https://arxiv.org/abs/1908.08628v1 | https://arxiv.org/pdf/1908.08628v1.pdf | Shadow Removal via Shadow Image Decomposition | We propose a novel deep learning method for shadow removal. Inspired by physical models of shadow formation, we use a linear illumination transformation to model the shadow effects in the image that allows the shadow image to be expressed as a combination of the shadow-free image, the shadow parameters, and a matte layer. We use two deep networks, namely SP-Net and M-Net, to predict the shadow parameters and the shadow matte respectively. This system allows us to remove the shadow effects on the images. We train and test our framework on the most challenging shadow removal dataset (ISTD). Compared to the state-of-the-art method, our model achieves a 40% error reduction in terms of root mean square error (RMSE) for the shadow area, reducing RMSE from 13.3 to 7.9. Moreover, we create an augmented ISTD dataset based on an image decomposition system by modifying the shadow parameters to generate new synthetic shadow images. Training our model on this new augmented ISTD dataset further lowers the RMSE on the shadow area to 7.4. | ['Hieu Le', 'Dimitris Samaras'] | 2019-08-23 | shadow-removal-via-shadow-image-decomposition-1 | http://openaccess.thecvf.com/content_ICCV_2019/html/Le_Shadow_Removal_via_Shadow_Image_Decomposition_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Le_Shadow_Removal_via_Shadow_Image_Decomposition_ICCV_2019_paper.pdf | iccv-2019-10 | ['shadow-removal'] | ['computer-vision'] | [ 5.57853460e-01 1.80880532e-01 7.05530047e-01 -3.27406585e-01
-2.83443570e-01 -2.23745674e-01 4.13071096e-01 -6.85718775e-01
-2.65132755e-01 6.53971732e-01 2.65143476e-02 -4.56009507e-01
5.13120115e-01 -7.20836997e-01 -8.25588107e-01 -1.04023266e+00
3.61345172e-01 9.59579349e-02 3.99780005e-01 -1.59090877e-01
5.62487207e-02 6.17646396e-01 -1.27862060e+00 1.99372210e-02
1.22577393e+00 1.10155344e+00 6.13979757e-01 5.91204762e-01
1.12696387e-01 5.22965610e-01 -9.24098849e-01 3.95021215e-02
4.75634187e-01 -2.56582052e-01 -2.80734617e-02 1.07480153e-01
4.97754991e-01 -8.96289647e-01 -6.28679216e-01 5.17025709e-01
6.10168874e-01 1.51737183e-01 5.04347146e-01 -1.09821153e+00
-6.41100526e-01 -2.03550592e-01 -7.43935227e-01 -3.69415104e-01
1.31380647e-01 6.72736242e-02 3.64682376e-01 -1.10821331e+00
5.16893506e-01 1.27645314e+00 9.04431224e-01 1.21407494e-01
-1.22617996e+00 -6.29009485e-01 -5.16011473e-03 1.95918486e-01
-1.18066716e+00 -2.62469500e-01 7.01204658e-01 3.91360708e-02
6.07937694e-01 3.90905410e-01 7.25643873e-01 1.22385144e+00
6.60386562e-01 8.14077556e-01 1.60025871e+00 -4.71625924e-01
9.33396891e-02 -6.53546602e-02 -1.99435562e-01 8.75898540e-01
1.84771016e-01 1.67490453e-01 -5.13795853e-01 4.88714650e-02
5.40403783e-01 1.01772062e-01 -5.16365886e-01 -3.37716877e-01
-8.46514463e-01 5.13692915e-01 7.13530958e-01 -3.59867007e-01
-3.64696473e-01 1.69869334e-01 -2.55724877e-01 -1.16950110e-01
5.23334920e-01 2.47167662e-01 -3.57228786e-01 5.57856798e-01
-8.96379590e-01 2.61676848e-01 8.94094706e-01 5.81904709e-01
8.77322316e-01 2.29259118e-01 -3.45778435e-01 7.98615754e-01
3.84873837e-01 1.44006515e+00 -1.69755682e-01 -1.11018896e+00
8.39897543e-02 4.20075953e-01 3.72464240e-01 -1.03960872e+00
-5.34030914e-01 -6.33318365e-01 -9.32691157e-01 5.55775046e-01
1.67791829e-01 -2.48703599e-01 -1.45668912e+00 1.49631202e+00
2.99253881e-01 2.93840200e-01 8.99679661e-02 1.00417745e+00
9.64973629e-01 8.63577127e-01 -6.14955485e-01 -1.05171114e-01
9.45294797e-01 -1.25097179e+00 -8.37441623e-01 -5.31847119e-01
2.01050326e-01 -9.85978723e-01 1.30853128e+00 2.38058373e-01
-6.58296645e-01 -2.88112223e-01 -1.28543472e+00 -4.40659411e-02
-3.48804951e-01 4.39449400e-01 4.79601085e-01 3.31330895e-01
-9.96901810e-01 3.80351841e-01 -6.52373075e-01 -1.59306616e-01
3.78834784e-01 4.54961136e-02 1.16022386e-01 -2.03135550e-01
-9.65849638e-01 9.09878433e-01 -9.33678523e-02 4.86673743e-01
-1.20474887e+00 -8.72577488e-01 -7.86180854e-01 2.02374056e-01
7.20132053e-01 -5.64616501e-01 9.58932281e-01 -8.62436831e-01
-1.64143753e+00 4.52881187e-01 -2.70220578e-01 -3.11133355e-01
5.73779523e-01 -5.94634533e-01 -1.74247965e-01 -6.06221147e-02
-6.27491251e-02 3.51997763e-01 1.14427412e+00 -2.07763696e+00
-2.02004924e-01 6.15695119e-03 2.09808856e-01 2.99100578e-01
2.23816391e-02 -4.53955024e-01 -6.95388734e-01 -6.85848832e-01
1.18906766e-01 -1.42743921e+00 -2.42728135e-03 2.82711089e-01
-8.20007443e-01 7.13666379e-01 1.28773713e+00 -1.00941634e+00
7.86725879e-01 -1.93984818e+00 -2.36140728e-01 2.08018452e-01
3.42242837e-01 3.48490477e-01 -3.06226641e-01 1.92109883e-01
2.51252025e-01 -4.04273182e-01 -7.81489193e-01 -7.97933400e-01
2.14874167e-02 4.78029102e-01 -6.25396609e-01 3.90742064e-01
-1.85911611e-01 9.31117773e-01 -4.49987590e-01 -2.23654166e-01
4.03481126e-01 8.76219988e-01 -7.18903467e-02 5.16919613e-01
-1.19255021e-01 1.48412094e-01 -1.26057878e-01 6.16135776e-01
1.40015244e+00 -6.96136579e-02 2.64099926e-01 -2.18134403e-01
-1.38329744e-01 7.27182925e-02 -8.12952995e-01 1.32608104e+00
-7.86699712e-01 9.96927559e-01 3.18496168e-01 -2.00463682e-01
9.24398959e-01 -2.39423305e-01 3.71450514e-01 -1.04220748e+00
1.44673631e-01 -1.23731596e-02 -1.27982378e-01 -1.11610003e-01
3.79136950e-01 -3.49699408e-02 2.42824718e-01 6.95717096e-01
-4.91503716e-01 -2.31160268e-01 -3.95948201e-01 2.19763935e-01
1.07218587e+00 3.56492013e-01 -3.57143342e-01 -1.85054958e-01
4.05669808e-01 -3.05466145e-01 5.76409280e-01 7.82273471e-01
2.17417002e-01 9.53491569e-01 4.06785309e-01 -4.92663264e-01
-7.71070480e-01 -1.28529930e+00 7.80426338e-02 9.28719223e-01
4.86345947e-01 -9.24756154e-02 -8.84985924e-01 -6.14910066e-01
2.18103930e-01 9.97607946e-01 -8.48625898e-01 -2.45828882e-01
-5.69067061e-01 -1.08583915e+00 3.50143224e-01 2.05934301e-01
9.54161167e-01 -1.05399823e+00 -8.25618207e-01 -2.80137658e-01
-3.34579706e-01 -1.26956248e+00 -3.28676432e-01 2.14384481e-01
-2.80892015e-01 -1.04399240e+00 -6.94054723e-01 -2.35921383e-01
7.01296031e-01 8.21205378e-01 9.86641586e-01 3.63457054e-01
-3.06325406e-01 1.79153476e-02 -2.26561800e-01 -8.10101092e-01
-3.94951820e-01 -2.42041409e-01 -1.38636932e-01 2.41236463e-01
-5.05600393e-01 -5.80823600e-01 -9.25575316e-01 3.49058568e-01
-1.06205213e+00 6.07858896e-01 8.97003889e-01 6.93645060e-01
4.34417218e-01 -1.51142344e-01 -9.18192640e-02 -9.55991566e-01
3.49215835e-01 -1.63165972e-01 -7.55713940e-01 2.46009991e-01
-9.04061198e-01 -5.17708659e-02 5.25167704e-01 -1.50140479e-01
-1.86111975e+00 1.68794513e-01 1.83682621e-01 -5.42280912e-01
1.81417510e-01 5.37166419e-03 -2.49510840e-01 -4.06227648e-01
6.09570205e-01 3.61366391e-01 -2.33846322e-01 -4.89564806e-01
2.98127770e-01 2.14504570e-01 5.54068744e-01 -1.32959142e-01
1.41560698e+00 1.00913703e+00 2.10046232e-01 -8.99984956e-01
-1.33166659e+00 2.52778959e-02 -7.10607827e-01 -2.56880999e-01
7.07244635e-01 -7.61690199e-01 -5.24386406e-01 9.38782215e-01
-1.21391439e+00 -1.20596814e+00 4.69066985e-02 -4.19189185e-02
-2.53270060e-01 4.84603256e-01 -1.97424382e-01 -6.10335529e-01
-5.43687165e-01 -9.01118398e-01 1.32909203e+00 1.98324606e-01
3.61893892e-01 -7.90175617e-01 -3.60292345e-02 2.94958651e-01
5.44715047e-01 4.73837674e-01 7.84074187e-01 2.89505422e-01
-8.48499715e-01 6.18048608e-02 -7.06979096e-01 6.50081098e-01
2.17138067e-01 -1.79370120e-01 -1.31108296e+00 -2.23297998e-01
1.46897182e-01 -1.02036260e-02 1.27561021e+00 5.10546148e-01
1.19436347e+00 -2.01507375e-01 -3.42517197e-01 1.09152353e+00
1.45037794e+00 2.49705091e-02 9.37651157e-01 3.84117931e-01
9.88126278e-01 2.09066331e-01 7.15210676e-01 5.15954196e-01
3.56723428e-01 6.65726006e-01 6.09592438e-01 -7.97053695e-01
-6.84694231e-01 2.80405600e-02 1.87499791e-01 3.61405134e-01
-1.11772142e-01 -6.64785504e-01 -8.55029881e-01 6.42301217e-02
-1.75683260e+00 -3.99750531e-01 -3.34870994e-01 1.95894241e+00
3.72598261e-01 1.50678784e-01 -6.49773121e-01 -2.09900811e-01
1.95020646e-01 6.83465838e-01 -7.64917970e-01 -1.40651897e-01
-4.01268661e-01 1.02407023e-01 8.11442733e-01 8.31490874e-01
-8.77920210e-01 1.20077026e+00 6.23360634e+00 3.50283355e-01
-1.36266065e+00 -5.31804562e-02 3.37853312e-01 7.96085298e-02
-4.54179406e-01 1.54560030e-01 -3.43870312e-01 4.11325097e-01
5.78879654e-01 3.80018532e-01 7.17700601e-01 4.53612566e-01
3.33854496e-01 -7.10345149e-01 -3.97400826e-01 7.98459053e-01
4.86550391e-01 -1.04990315e+00 -1.35902241e-01 1.55684069e-01
8.88276279e-01 7.85045698e-02 7.30578452e-02 1.67550176e-01
1.80878744e-01 -9.14366424e-01 6.44171059e-01 9.76662338e-01
8.16065311e-01 -4.12463009e-01 8.99327278e-01 2.35227510e-01
-8.99761975e-01 -1.80511940e-02 -3.22115302e-01 3.09177414e-02
9.20647830e-02 9.13581789e-01 -1.17537868e+00 4.04516697e-01
8.74936461e-01 2.70132720e-01 -6.80140078e-01 6.68512285e-01
-6.01394296e-01 4.26873356e-01 -3.29814821e-01 4.28344309e-01
-3.14638810e-03 -5.23526073e-01 4.19679463e-01 1.16631818e+00
2.12909296e-01 1.20614864e-01 -8.36132541e-02 9.94685173e-01
-2.40561485e-01 -4.65395331e-01 -4.28298771e-01 4.66127962e-01
2.49092609e-01 1.47618997e+00 -7.12853789e-01 -3.22986901e-01
-1.02501191e-01 1.50780904e+00 -2.95131970e-02 7.64946043e-01
-1.20256686e+00 -3.59183669e-01 5.02879500e-01 8.43390971e-02
2.85171062e-01 -2.41271451e-01 -4.30374116e-01 -9.58491564e-01
1.33075044e-01 -7.96858609e-01 -4.89279121e-01 -1.34323835e+00
-6.38287246e-01 4.19609874e-01 -1.78614005e-01 -7.94081986e-01
4.07709569e-01 -4.71410543e-01 -8.12492311e-01 8.81497145e-01
-1.82153654e+00 -1.30206132e+00 -1.18750381e+00 3.66806090e-01
3.26890916e-01 1.17607854e-01 7.88644254e-01 2.12180521e-02
-5.59511960e-01 2.89437950e-01 3.59055609e-01 -1.92870289e-01
1.02512944e+00 -1.22662771e+00 6.79747224e-01 9.41590428e-01
-3.86438549e-01 1.73658520e-01 8.46221685e-01 -6.94057047e-01
-1.57515645e+00 -1.13722765e+00 5.27561307e-01 -3.74894708e-01
2.67217427e-01 -7.21210897e-01 -9.87787902e-01 4.83801097e-01
9.84247550e-02 -3.62483114e-02 1.72186509e-01 -2.13652194e-01
-3.85034829e-01 -5.73566556e-01 -9.62276995e-01 7.64306545e-01
1.08823228e+00 -4.56315577e-01 -9.95970145e-02 5.25019586e-01
8.97053361e-01 -7.77688920e-01 -2.53721267e-01 5.86662591e-01
8.23389173e-01 -1.40733910e+00 1.01034117e+00 2.51730293e-01
4.23385739e-01 -3.50182801e-01 -2.91716576e-01 -1.46607292e+00
-1.90547153e-01 -3.94028693e-01 -3.73187214e-01 8.84158373e-01
3.12046587e-01 -7.10702062e-01 7.33823836e-01 3.39055121e-01
-6.35950327e-01 -1.00814378e+00 -5.21263599e-01 -5.41111469e-01
-4.86846358e-01 -2.87449002e-01 5.36520302e-01 4.13377464e-01
-1.14878237e+00 2.89389938e-01 -5.70675731e-01 5.73500812e-01
8.62073481e-01 6.30343974e-01 1.17047179e+00 -1.08530259e+00
-1.51320145e-01 1.57559097e-01 5.29235482e-01 -1.07772815e+00
3.44270140e-01 -3.44525039e-01 6.10667408e-01 -1.89918709e+00
3.80982161e-01 -5.16748548e-01 -1.35921344e-01 7.14249134e-01
-2.20069483e-01 4.53659892e-01 3.87463063e-01 1.66304767e-01
-3.38699728e-01 9.67045009e-01 1.54828560e+00 -1.18936896e-01
-3.47411484e-01 1.18157268e-01 -4.79178518e-01 8.88845682e-01
7.78620481e-01 -4.45804566e-01 -2.76058555e-01 -6.70316160e-01
-1.38731617e-02 -1.73802108e-01 6.14362657e-01 -1.02026296e+00
-2.65800923e-01 -3.10256362e-01 8.00949752e-01 -8.79067540e-01
9.90474641e-01 -8.25927019e-01 -2.44515128e-02 5.57925105e-01
1.57798484e-01 -6.21919632e-01 3.07814777e-01 5.37856221e-01
3.76391441e-01 3.92586142e-01 7.86939561e-01 1.67357594e-01
-3.38474959e-01 1.12769037e-01 -3.59546989e-01 -4.69029158e-01
7.87281632e-01 -6.47573173e-02 -6.08920336e-01 -6.78712606e-01
-1.82812989e-01 1.13592803e-01 6.66249096e-01 2.29425386e-01
9.04195011e-01 -1.02557898e+00 -5.10654032e-01 3.57313871e-01
-2.56289989e-01 9.35226530e-02 3.03150088e-01 9.05113161e-01
-7.50462830e-01 1.76333874e-01 -1.36529818e-01 -3.76023531e-01
-1.39401889e+00 1.62016556e-01 5.17912567e-01 -1.39333501e-01
-7.65822053e-01 5.93372047e-01 9.14648533e-01 -7.36198783e-01
2.20332667e-01 -3.34566683e-01 3.54293555e-01 -3.87519568e-01
1.67506903e-01 5.22753358e-01 8.88520256e-02 -4.52203631e-01
-3.37900788e-01 4.50644255e-01 3.79216850e-01 1.45906359e-02
1.48563194e+00 -2.19561145e-01 -2.02235088e-01 1.86529070e-01
1.06030321e+00 2.84574509e-01 -1.89458466e+00 -7.68082291e-02
-6.20182037e-01 -5.31738400e-01 2.60308504e-01 -1.27896333e+00
-1.15552652e+00 7.90707648e-01 9.61495936e-01 -1.35476798e-01
1.34273517e+00 -2.52204865e-01 1.26398849e+00 6.98826730e-01
-8.41214061e-02 -1.09370291e+00 2.77932972e-01 9.04312611e-01
1.34023046e+00 -1.26747465e+00 4.90222007e-01 -5.61958790e-01
-4.75147158e-01 8.84417355e-01 5.24268031e-01 -2.05538109e-01
6.56681895e-01 5.05362570e-01 5.29903412e-01 -1.68008938e-01
-3.55482787e-01 3.54934717e-03 4.30088997e-01 5.93695343e-01
-2.11434990e-01 1.54940024e-01 1.68870255e-01 2.43299499e-01
-2.43758351e-01 -3.11515689e-01 6.12899482e-01 6.83060884e-01
-5.34198880e-01 -6.95613921e-01 -7.33854115e-01 3.44318658e-01
1.93622455e-01 -2.28557780e-01 -7.21194923e-01 7.49338150e-01
3.25657606e-01 8.85930240e-01 -1.78959608e-01 -5.30461967e-01
3.64912003e-01 -2.13834837e-01 3.34918171e-01 -4.23633844e-01
-3.86429727e-02 7.53254071e-02 -4.68712300e-02 -7.88873494e-01
-8.73213485e-02 -3.03735524e-01 -1.38104200e+00 -3.77622247e-01
-2.92883903e-01 -4.62328374e-01 9.28578436e-01 1.03153729e+00
2.57166088e-01 9.08745408e-01 8.01701427e-01 -1.28971386e+00
4.58121151e-02 -9.31224704e-01 -6.38701677e-01 4.00232114e-02
6.19444907e-01 -8.10734808e-01 -4.73955065e-01 -1.47253543e-01] | [10.846614837646484, -4.107207298278809] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.