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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2b4fe9f6-0040-4f28-b7ab-aa7c39d413ec | u-det-a-modified-u-net-architecture-with | 2003.09293 | null | https://arxiv.org/abs/2003.09293v1 | https://arxiv.org/pdf/2003.09293v1.pdf | U-Det: A Modified U-Net architecture with bidirectional feature network for lung nodule segmentation | Early diagnosis and analysis of lung cancer involve a precise and efficient lung nodule segmentation in computed tomography (CT) images. However, the anonymous shapes, visual features, and surroundings of the nodule in the CT image pose a challenging problem to the robust segmentation of the lung nodules. This article proposes U-Det, a resource-efficient model architecture, which is an end to end deep learning approach to solve the task at hand. It incorporates a Bi-FPN (bidirectional feature network) between the encoder and decoder. Furthermore, it uses Mish activation function and class weights of masks to enhance segmentation efficiency. The proposed model is extensively trained and evaluated on the publicly available LUNA-16 dataset consisting of 1186 lung nodules. The U-Det architecture outperforms the existing U-Net model with the Dice similarity coefficient (DSC) of 82.82% and achieves results comparable to human experts. | ['Chandra Sekhara Rao Annavarapu', 'Samson Anosh Babu P', 'Nikhil Varma Keetha'] | 2020-03-20 | null | null | null | null | ['lung-nodule-segmentation'] | ['medical'] | [-1.13632372e-02 2.27669358e-01 -4.26320165e-01 -2.61858076e-01
-7.99456358e-01 -3.84188592e-01 3.03973526e-01 -2.00744763e-01
-4.49531347e-01 2.46802583e-01 1.57417312e-01 -5.09279728e-01
2.01865345e-01 -5.46624184e-01 -5.70737302e-01 -6.45750999e-01
-3.53352390e-02 8.31069529e-01 6.43001854e-01 2.65782118e-01
-3.27279001e-01 5.65902710e-01 -6.11591518e-01 5.49470723e-01
6.26463532e-01 1.13797033e+00 5.17544508e-01 9.16770279e-01
7.47402087e-02 9.88547921e-01 7.50406384e-02 -3.36624086e-01
3.63664240e-01 -5.59974134e-01 -1.11813021e+00 1.61073282e-01
4.40270782e-01 -7.35204458e-01 -6.41527712e-01 8.28791499e-01
5.87772191e-01 -1.98871613e-01 9.91922736e-01 -6.33053482e-01
-4.75285709e-01 4.97984558e-01 -5.10102689e-01 6.46022797e-01
-3.90870512e-01 2.49977201e-01 8.88363779e-01 -9.15968657e-01
4.86084670e-01 8.68033886e-01 8.38110447e-01 5.42537868e-01
-8.01822424e-01 -3.93881232e-01 -5.54475248e-01 1.15881981e-02
-1.24379694e+00 8.24814513e-02 8.33328590e-02 -5.54306567e-01
8.30960214e-01 1.98741719e-01 8.40216279e-01 7.66849995e-01
6.80257380e-01 9.11024034e-01 5.25694668e-01 -3.66082974e-02
-1.77802891e-01 -5.44011220e-02 -2.43863896e-01 1.25414824e+00
4.15770888e-01 2.69627839e-01 3.81226808e-01 1.11252099e-01
1.28134143e+00 3.60232353e-01 -2.08985955e-01 -6.17232382e-01
-1.46126711e+00 7.65793800e-01 1.44081938e+00 4.52929527e-01
-4.10444587e-01 5.50789773e-01 3.94841611e-01 -3.53276759e-01
1.41662955e-01 1.99100897e-01 1.29491776e-01 3.85091603e-01
-1.06439614e+00 -3.15106630e-01 6.80331767e-01 8.14382315e-01
1.16144620e-01 -7.49903023e-02 -9.76187110e-01 5.83377659e-01
5.79434633e-01 3.03802848e-01 6.60541356e-01 -6.80672228e-01
2.41032630e-01 5.94046056e-01 -2.47932866e-01 -4.81411844e-01
-5.45954764e-01 -7.49250352e-01 -1.06000078e+00 9.34879035e-02
4.28157449e-01 -1.46254659e-01 -1.47897732e+00 1.06676197e+00
3.62620443e-01 2.24951506e-01 -3.73598486e-01 1.09562552e+00
1.29240406e+00 2.46106967e-01 1.68567255e-01 2.29565918e-01
1.37428188e+00 -1.58772182e+00 -5.68713427e-01 -3.10437024e-01
5.69757998e-01 -6.55677438e-01 6.49109244e-01 -4.04679865e-01
-1.14762592e+00 -4.71268117e-01 -8.93290997e-01 -4.75127459e-01
1.68327570e-01 6.91976249e-01 3.19253504e-01 4.71399993e-01
-1.11786997e+00 5.14552891e-01 -1.21815634e+00 -2.44702682e-01
9.46845651e-01 6.29880548e-01 -1.48180887e-01 -1.35517821e-01
-7.51618564e-01 8.20264518e-01 4.92845774e-01 1.42231673e-01
-1.22069669e+00 -7.71781266e-01 -5.96190810e-01 3.45642805e-01
3.81063789e-01 -1.14979804e+00 1.77874446e+00 -7.93461919e-01
-1.28320849e+00 8.90427172e-01 1.08852066e-01 -5.25238991e-01
9.75609064e-01 1.69962153e-01 2.51233757e-01 4.83873844e-01
8.18656608e-02 9.97341156e-01 6.02823079e-01 -9.93053913e-01
-5.92684031e-01 1.61429401e-02 -5.24636567e-01 3.97349894e-01
5.47146611e-02 -2.86290288e-01 -8.17996323e-01 -5.57289600e-01
1.76509395e-01 -8.97796273e-01 -6.31931663e-01 7.19377041e-01
-4.78691578e-01 2.61740456e-03 8.51072013e-01 -9.41899896e-01
9.53838646e-01 -1.84881246e+00 -2.60295868e-01 3.18792701e-01
6.39443576e-01 2.98725456e-01 9.79237482e-02 -4.31163073e-01
-5.03630079e-02 1.32407382e-01 -3.83592904e-01 1.18087851e-01
-2.49442205e-01 2.98853368e-01 5.72726011e-01 5.39525270e-01
-4.58723828e-02 1.53523874e+00 -7.68029094e-01 -1.05732191e+00
4.45667446e-01 4.49813873e-01 -5.84764242e-01 3.76797169e-01
-1.14224993e-01 3.94780248e-01 -5.62034547e-01 5.96661031e-01
5.00614822e-01 -8.14675093e-01 2.93336153e-01 -4.00476992e-01
1.34207085e-01 2.64009666e-02 -4.86397356e-01 1.47589362e+00
-3.54986578e-01 4.09715503e-01 1.35115325e-01 -4.11417663e-01
3.41129184e-01 7.00734258e-01 9.74041820e-01 -5.66447794e-01
6.06150806e-01 4.28278953e-01 5.09150863e-01 -6.48957849e-01
-2.43140861e-01 -2.02644989e-01 3.91985506e-01 4.93905634e-01
-1.78314392e-02 -2.50110745e-01 2.48789057e-01 3.60609405e-02
1.26409233e+00 -3.97125930e-01 5.86533666e-01 -2.87477463e-01
4.60012764e-01 1.80338416e-02 2.49172956e-01 6.09492719e-01
-5.93526304e-01 1.00683510e+00 3.05429071e-01 -4.13824171e-01
-1.12058282e+00 -1.12103987e+00 -2.34003425e-01 5.57815850e-01
-1.37429265e-02 8.97008926e-02 -6.07346177e-01 -1.24511063e+00
-1.34845050e-02 2.58305162e-01 -7.59603500e-01 2.33356282e-02
-7.20656753e-01 -4.61801022e-01 3.48382682e-01 1.01978564e+00
6.99881256e-01 -9.68266845e-01 -6.63891673e-01 1.92420170e-01
-3.47725809e-01 -1.15105247e+00 -9.22061324e-01 4.52003509e-01
-1.05798507e+00 -1.23378062e+00 -1.00535798e+00 -1.11211753e+00
9.67999995e-01 2.85680175e-01 1.40947843e+00 3.29992115e-01
-7.00870693e-01 9.96085405e-02 1.86833516e-01 3.20526995e-02
-4.32359666e-01 6.61219358e-02 -8.10038328e-01 -5.92995346e-01
4.38597500e-02 8.23146179e-02 -9.86467421e-01 5.33872306e-01
-8.71526897e-01 1.98602647e-01 1.25503802e+00 8.96884322e-01
1.00085199e+00 -4.88121156e-03 -1.56866312e-02 -9.95982170e-01
1.96231484e-01 -6.34852767e-01 -3.27618897e-01 3.71864349e-01
-1.93964690e-01 -1.39918178e-01 4.18291837e-01 -1.40525987e-02
-9.20942247e-01 5.55868924e-01 -2.13479623e-01 -4.75226969e-01
2.58004248e-01 1.93448290e-01 2.55194813e-01 -5.22709370e-01
5.58166444e-01 3.23649459e-02 9.52903479e-02 -1.13514014e-01
1.90595806e-01 4.25685436e-01 5.58858752e-01 1.56206712e-01
7.65280962e-01 4.81845438e-01 1.68161482e-01 -5.23755908e-01
-9.27699447e-01 -6.79974675e-01 -1.01177967e+00 -2.38317803e-01
1.15287137e+00 -8.43211770e-01 -2.23741353e-01 -5.60460985e-02
-8.73254418e-01 -5.05492508e-01 -3.61360699e-01 7.49766588e-01
-5.39858758e-01 3.36743265e-01 -8.38680267e-01 3.61602344e-02
-7.51830399e-01 -1.30358160e+00 9.24779058e-01 3.07258338e-01
-7.95469731e-02 -1.06078827e+00 -6.41334504e-02 4.14489418e-01
5.70470154e-01 1.35629714e-01 9.31510568e-01 -6.51576221e-01
-8.29127789e-01 -3.16522449e-01 -7.60838866e-01 2.84864366e-01
3.02003115e-01 1.43858045e-02 -6.92418098e-01 -3.56785774e-01
8.64476897e-03 -3.03451747e-01 1.11700678e+00 9.08147693e-01
1.60165644e+00 -1.82813525e-01 -7.18234360e-01 8.82316530e-01
1.34552705e+00 5.36165759e-02 3.38390380e-01 -1.93039626e-01
8.91884387e-01 -1.30739287e-01 1.64576560e-01 8.36739242e-02
1.70806542e-01 2.00561985e-01 8.12782526e-01 -5.11309206e-01
-7.28593886e-01 -9.31861699e-02 -5.36722541e-02 5.97260416e-01
1.29359951e-02 -3.48079830e-01 -1.13872528e+00 5.34765661e-01
-1.39000940e+00 -6.34101570e-01 -2.48470068e-01 1.73762214e+00
6.77645743e-01 -2.60047894e-02 -1.29560560e-01 -3.65804225e-01
6.85793936e-01 -1.15549555e-02 -5.72480619e-01 -8.69420916e-02
4.83767211e-01 1.77870527e-01 6.97542608e-01 3.11782509e-01
-1.47416818e+00 4.52309847e-01 6.96186829e+00 9.13155496e-01
-1.19187224e+00 2.74061292e-01 9.68399644e-01 -2.15740036e-02
-1.55641705e-01 -5.89109063e-01 -4.24324125e-01 1.69053450e-01
5.34179866e-01 -6.20833449e-02 -6.95992783e-02 9.02278304e-01
3.21208611e-02 -1.41796842e-01 -1.27274513e+00 6.99224710e-01
-3.62846851e-02 -1.41264415e+00 -1.66741759e-01 9.21825618e-02
9.18557644e-01 6.29719079e-01 2.19039261e-01 2.17185289e-01
2.99788564e-01 -1.45758641e+00 2.17466831e-01 2.03634948e-01
1.00661647e+00 -5.39822698e-01 1.08897579e+00 3.15142721e-01
-1.36150455e+00 -2.89275944e-02 -4.50580776e-01 8.70016277e-01
-6.40003011e-02 5.27959228e-01 -1.73689079e+00 3.22566420e-01
5.86974561e-01 6.30043268e-01 -8.46679270e-01 1.58235073e+00
-4.38632183e-02 6.64968967e-01 -4.43049967e-01 -1.79504789e-02
6.21344864e-01 1.46370620e-01 1.74020037e-01 1.31437445e+00
6.35737479e-01 3.58422548e-02 3.84274662e-01 9.95589435e-01
-5.04035652e-01 1.07513823e-01 -3.24375689e-01 -8.67104307e-02
-8.53298232e-02 1.61857378e+00 -1.15212703e+00 -3.00030470e-01
-4.36667502e-01 9.47650850e-01 1.39624149e-01 3.37858088e-02
-1.09145319e+00 2.59691179e-01 -1.37753710e-01 3.74312848e-01
5.21581352e-01 1.39518976e-01 -3.33621651e-01 -6.15359604e-01
-2.89952338e-01 -5.86198151e-01 5.78531086e-01 -5.42162418e-01
-1.25101376e+00 7.30184972e-01 -4.70539540e-01 -1.28610694e+00
-5.91535792e-02 -7.37474501e-01 -8.85174811e-01 6.25586033e-01
-1.47092855e+00 -1.16719306e+00 -5.90223193e-01 4.06694472e-01
4.30698037e-01 7.08589107e-02 4.94665444e-01 2.58869112e-01
-4.58685666e-01 2.95426965e-01 2.56919056e-01 5.82685232e-01
4.91933048e-01 -1.51832867e+00 2.81296372e-01 6.26926959e-01
-1.05706669e-01 -1.02761827e-01 -4.19537835e-02 -8.38139713e-01
-9.70003426e-01 -1.60195732e+00 7.50451088e-01 -3.59097987e-01
3.92387599e-01 1.09245770e-01 -7.45100796e-01 7.75689662e-01
3.91388446e-01 7.86335289e-01 5.63681424e-01 -8.08936775e-01
2.48635635e-01 2.26887032e-01 -1.30438256e+00 3.56478125e-01
5.80601692e-01 -1.19004160e-01 -2.11980551e-01 5.85680902e-01
3.40889454e-01 -8.51373911e-01 -9.07874167e-01 5.49486935e-01
4.26615626e-01 -9.63673413e-01 1.24710691e+00 -2.45546177e-01
5.69918096e-01 -1.39257029e-01 2.76864052e-01 -9.61560130e-01
-7.92470753e-01 9.23222154e-02 -1.67081822e-02 2.19705671e-01
5.66719770e-01 1.00588605e-01 1.19491565e+00 5.00644207e-01
-3.53808194e-01 -1.17737710e+00 -9.37863648e-01 -2.46418893e-01
2.91836917e-01 8.09318051e-02 -4.22911048e-02 4.23059255e-01
-6.10523522e-01 -1.20879281e-02 1.63524985e-01 -9.95308831e-02
5.61942756e-01 5.31901838e-03 1.97617281e-02 -9.31826115e-01
-1.48597717e-01 -6.69334769e-01 -1.42726764e-01 -1.08331645e+00
-3.28495741e-01 -1.43023121e+00 2.21217036e-01 -2.10444069e+00
6.40570343e-01 -1.36009276e-01 -2.52327651e-01 4.17058647e-01
-2.72155672e-01 3.90984267e-01 -1.84049625e-02 2.51794279e-01
-5.70273519e-01 5.06218016e-01 2.27027535e+00 -3.54237616e-01
3.18223387e-01 4.90487903e-01 -2.97427773e-01 7.37623751e-01
7.27801621e-01 -5.64754605e-01 -2.20030829e-01 -5.29686630e-01
-4.07047033e-01 2.00475827e-01 5.58855176e-01 -1.22437167e+00
4.58823532e-01 1.04067475e-01 1.03336179e+00 -1.04384053e+00
8.81644189e-02 -1.27217472e+00 -1.41607419e-01 1.31052411e+00
-3.08438212e-01 -1.46269873e-01 -8.35756063e-02 6.59289539e-01
-1.82647854e-01 -3.51666272e-01 1.19710243e+00 -6.07115448e-01
-4.66777682e-01 8.47437024e-01 -4.28639203e-01 9.20362324e-02
1.27984369e+00 -1.42246962e-01 5.90832233e-02 -1.22323819e-01
-7.14254618e-01 5.52201867e-01 -1.90532603e-03 -2.04712152e-01
7.12417960e-01 -1.52928972e+00 -8.32841575e-01 2.36111939e-01
-2.39754900e-01 5.44387400e-01 2.42846355e-01 1.31437147e+00
-1.25577271e+00 7.16250896e-01 -1.75436750e-01 -8.87233138e-01
-1.19076467e+00 1.76072851e-01 1.17261100e+00 -7.45347500e-01
-8.94387603e-01 1.19558597e+00 3.95367742e-01 -3.80899698e-01
3.12391669e-01 -6.10928774e-01 -2.17957780e-01 -2.97683716e-01
2.26863530e-02 2.34035283e-01 9.43715423e-02 -4.32751477e-01
-3.02728951e-01 2.92799413e-01 -2.25217730e-01 3.41919363e-01
9.63731647e-01 1.20378971e-01 -9.80227813e-02 -2.52407044e-01
1.29582083e+00 -5.37415802e-01 -1.20166934e+00 -4.95304257e-01
-4.04210016e-02 -1.79504991e-01 3.87821823e-01 -9.86104429e-01
-1.57408917e+00 1.03628123e+00 8.80173862e-01 -1.78116232e-01
7.35034943e-01 2.21859459e-02 1.02105796e+00 5.01559734e-01
-4.64840889e-01 -7.47456908e-01 5.08926988e-01 4.86728936e-01
7.18028426e-01 -1.63824451e+00 9.82044563e-02 -5.93554676e-01
-8.10413182e-01 1.36267006e+00 1.05576980e+00 -1.70245454e-01
8.39028955e-01 4.51263845e-01 2.61778295e-01 -2.73923814e-01
-6.62705779e-01 -2.80391246e-01 8.10713530e-01 3.72232527e-01
8.76417100e-01 8.83105323e-02 -1.32045135e-01 5.13426125e-01
9.77423042e-02 9.02274698e-02 1.58898741e-01 7.29044318e-01
-7.59743631e-01 -5.99067569e-01 -2.74942994e-01 8.83724988e-01
-5.59208870e-01 6.29798025e-02 -4.28450346e-01 9.07659292e-01
8.20243880e-02 3.11053425e-01 6.53296113e-02 7.01453388e-02
-1.60049453e-01 -2.54056185e-01 4.01176840e-01 -8.33889425e-01
-8.70361745e-01 4.16640848e-01 -1.86667666e-01 -4.83319193e-01
-2.00862750e-01 -2.59224743e-01 -1.57356071e+00 -3.26722488e-02
-2.26684615e-01 -8.34619403e-02 3.61260772e-01 7.12265432e-01
-1.83013171e-01 1.12367558e+00 7.12839305e-01 -6.41331196e-01
-7.41273820e-01 -8.01069140e-01 -4.04999495e-01 1.28707618e-01
2.20524922e-01 -3.61096203e-01 -4.76249196e-02 -2.07526654e-01] | [15.384090423583984, -2.0899295806884766] |
87007eda-28db-4b58-8519-8b20654336a1 | online-clustering-by-penalized-weighted-gmm | 1902.02544 | null | http://arxiv.org/abs/1902.02544v1 | http://arxiv.org/pdf/1902.02544v1.pdf | Online Clustering by Penalized Weighted GMM | With the dawn of the Big Data era, data sets are growing rapidly. Data is
streaming from everywhere - from cameras, mobile phones, cars, and other
electronic devices. Clustering streaming data is a very challenging problem.
Unlike the traditional clustering algorithms where the dataset can be stored
and scanned multiple times, clustering streaming data has to satisfy
constraints such as limit memory size, real-time response, unknown data
statistics and an unknown number of clusters. In this paper, we present a novel
online clustering algorithm which can be used to cluster streaming data without
knowing the number of clusters a priori. Results on both synthetic and real
datasets show that the proposed algorithm produces partitions which are close
to what you could get if you clustered the whole data at one time. | ['Shlomo Bugdary', 'Shay Maymon'] | 2019-02-07 | null | null | null | null | ['online-clustering'] | ['computer-vision'] | [-6.83903396e-02 -3.98866832e-01 -1.64155718e-02 -4.61058259e-01
-2.91740865e-01 -6.04489028e-01 5.17425872e-02 7.48809636e-01
-5.74498177e-01 3.61727357e-01 -1.38386637e-01 1.11609548e-01
-4.07173961e-01 -1.02609253e+00 -3.81252378e-01 -6.77852273e-01
-5.90096533e-01 9.63662505e-01 9.78287458e-01 -5.67727163e-02
2.79552102e-01 2.42881119e-01 -1.90788555e+00 5.20884514e-01
6.16866529e-01 1.03597915e+00 3.04031730e-01 8.11004579e-01
-3.51722270e-01 5.20324111e-01 -7.59038508e-01 1.55695006e-01
4.28327948e-01 -2.45047346e-01 -5.94661772e-01 2.89660394e-01
-4.29632872e-01 -9.58753526e-02 -2.62917560e-02 9.08609927e-01
4.12426263e-01 3.46765429e-01 1.49012223e-01 -1.54234886e+00
1.47505596e-01 8.35000694e-01 -8.16595435e-01 3.73630792e-01
3.39559436e-01 -1.56802475e-01 4.76098925e-01 -3.61710161e-01
5.25896490e-01 9.45491850e-01 3.34944040e-01 1.96254328e-01
-7.06074595e-01 -4.74380493e-01 9.08068940e-03 6.29664183e-01
-1.73698342e+00 -1.61657631e-01 3.61484796e-01 -3.01022440e-01
3.97681028e-01 4.07044053e-01 5.71283758e-01 3.13214362e-02
-3.08325261e-01 4.11330581e-01 4.21088725e-01 -1.40072450e-01
8.76401901e-01 8.70920196e-02 1.35401770e-01 -2.62584418e-01
3.67387682e-01 -4.51022655e-01 -2.57358223e-01 -2.55643129e-01
7.64529258e-02 5.00788093e-01 -1.93104863e-01 -2.37325534e-01
-1.37382710e+00 6.52361333e-01 -2.29977161e-01 3.28730762e-01
-5.76508880e-01 -1.43613666e-01 6.54201150e-01 3.58855933e-01
1.06904879e-01 -2.77868062e-01 -3.23507756e-01 -3.11398208e-01
-1.13166606e+00 -1.69451963e-02 1.05758643e+00 1.17818141e+00
6.53615594e-01 -1.66917399e-01 5.33191860e-01 5.32652915e-01
-8.52657184e-02 4.17206436e-01 7.06050813e-01 -9.12033081e-01
5.55322289e-01 7.22188532e-01 2.33021438e-01 -1.11570978e+00
-4.80405688e-01 3.09700519e-01 -1.26647544e+00 -2.25721329e-01
4.49624926e-01 -2.14194745e-01 -5.11730134e-01 1.07649553e+00
8.25024545e-01 4.57002372e-01 1.00927554e-01 9.94883835e-01
5.00887036e-01 1.14288580e+00 -3.90563011e-01 -7.36987352e-01
9.83652532e-01 -3.54166538e-01 -8.08871329e-01 3.20568383e-01
3.41598213e-01 -6.38820410e-01 6.93201959e-01 8.63733292e-01
-1.16623247e+00 -5.18828034e-01 -7.62167335e-01 4.82868969e-01
-3.34197521e-01 -6.23360991e-01 2.13991195e-01 7.69941211e-01
-9.93335485e-01 2.80986071e-01 -1.02209997e+00 -5.59854388e-01
1.10540517e-01 4.78134871e-01 -1.26227900e-01 -3.24822277e-01
-7.94265389e-01 1.65846143e-02 9.50192273e-01 -4.21134792e-02
-6.36693299e-01 -5.38858950e-01 -1.71436280e-01 2.36965790e-01
5.56565642e-01 -1.98008403e-01 8.78072381e-01 -9.07849848e-01
-1.03409266e+00 2.70397484e-01 -1.96228430e-01 -3.39768529e-01
2.92994499e-01 2.95797795e-01 -6.71316683e-01 5.33039629e-01
-2.66083539e-01 1.03002720e-01 5.65106988e-01 -1.01721251e+00
-9.51841354e-01 -4.83269960e-01 -3.56156349e-01 1.32108614e-01
-5.64281881e-01 1.36549816e-01 -4.45002645e-01 3.00598722e-02
6.98212460e-02 -6.96161270e-01 -6.41253293e-01 -4.11793530e-01
6.60059461e-03 -3.04817200e-01 1.27457249e+00 -1.33916989e-01
1.37360656e+00 -2.39198279e+00 -1.88349962e-01 5.29592514e-01
9.92941186e-02 2.63421923e-01 1.12599820e-01 8.33376646e-01
9.27394107e-02 3.64868701e-01 -1.56358823e-01 1.17832080e-01
-3.05264473e-01 3.25916767e-01 -2.81467987e-03 5.19725263e-01
-3.69011790e-01 4.23483476e-02 -1.03901935e+00 -6.39158309e-01
2.72900522e-01 1.47827908e-01 -3.83072913e-01 2.27977172e-01
-9.79705602e-02 1.88176170e-01 -2.00175449e-01 3.02831054e-01
9.26158845e-01 -4.30269986e-01 3.14532965e-01 3.84574324e-01
-1.09293394e-01 -4.25579160e-01 -2.07730842e+00 1.08211040e+00
8.75567645e-02 3.62182289e-01 2.48465627e-01 -1.35658932e+00
8.63684416e-01 6.67959511e-01 1.05159914e+00 -4.70873773e-01
-3.84703558e-03 4.32860292e-02 -1.83717683e-02 -6.70907795e-01
7.18662441e-01 -1.48831174e-01 -3.31321768e-02 7.69837856e-01
-5.41478753e-01 1.92686737e-01 6.56762898e-01 4.78640229e-01
1.27468300e+00 -1.02607250e+00 1.46303043e-01 -9.09051895e-02
4.21911567e-01 3.14891189e-01 7.35367417e-01 4.31104988e-01
4.45068292e-02 8.04197371e-01 3.68156165e-01 -5.10037243e-01
-1.25303411e+00 -8.69686365e-01 2.72378832e-01 7.08137631e-01
4.08857882e-01 -4.79998231e-01 -4.73787695e-01 -2.56777331e-02
-1.68002307e-01 2.54089713e-01 -3.22858155e-01 1.30854562e-01
-5.06760299e-01 -5.60530424e-01 -3.20199728e-02 7.73117095e-02
1.93701968e-01 -8.62532139e-01 -9.66868103e-01 4.14232492e-01
-2.73282170e-01 -1.15288043e+00 -4.13123131e-01 -2.91151404e-01
-9.05164182e-01 -1.07929957e+00 -1.93471730e-01 -7.52670884e-01
7.25748658e-01 7.37441003e-01 8.42768431e-01 5.16980767e-01
-2.95392394e-01 2.51156807e-01 -8.57683301e-01 -5.04532933e-01
-2.65158355e-01 6.06249161e-02 1.04244299e-01 3.25738519e-01
5.54621160e-01 -6.15059316e-01 -6.65037036e-01 5.57905793e-01
-1.51507938e+00 -3.43727946e-01 -6.81567416e-02 9.26714316e-02
5.38772643e-01 1.12714648e+00 8.63464952e-01 -9.77269113e-01
5.63103795e-01 -1.01529872e+00 -6.16621792e-01 4.41944264e-02
-4.32627857e-01 -6.54641092e-01 1.14656639e+00 -6.17663383e-01
-5.53953826e-01 2.97724217e-01 2.28051558e-01 -5.21083593e-01
-4.41925764e-01 3.94678235e-01 -3.47183496e-01 6.36763692e-01
3.75460476e-01 1.46284059e-01 1.28471091e-01 -2.28140771e-01
2.59455115e-01 1.16626310e+00 4.22502130e-01 -1.24922663e-01
5.44014573e-01 8.50501180e-01 -1.90202415e-01 -1.15162253e+00
-7.96010196e-02 -1.22353601e+00 -8.50459874e-01 -4.95134890e-01
4.97666419e-01 -6.50509417e-01 -1.14177895e+00 3.79791766e-01
-7.73995638e-01 -1.50391057e-01 -4.13206428e-01 4.38264966e-01
-1.66180089e-01 5.89772105e-01 -1.75488517e-01 -8.92559409e-01
-9.35748965e-02 -6.83974028e-01 4.57820743e-01 3.61844748e-01
2.60029379e-02 -8.97039175e-01 -5.91790639e-02 -2.89990902e-02
4.08660710e-01 5.85482717e-01 3.69957656e-01 -9.12204921e-01
-6.55539870e-01 -5.65395176e-01 1.73709262e-02 -1.17137350e-01
2.65699923e-01 3.76752734e-01 -3.25504959e-01 -5.78996181e-01
1.16124570e-01 8.12666267e-02 3.18829894e-01 2.83662975e-01
1.36761475e+00 -5.72981477e-01 -5.14131188e-01 2.93220073e-01
1.69555008e+00 7.26931930e-01 6.82541430e-01 -3.37338783e-02
5.07600427e-01 7.42319942e-01 8.68794680e-01 1.30653560e+00
4.42192674e-01 9.92239565e-02 4.80121344e-01 3.42005849e-01
6.34316742e-01 1.98203683e-01 5.77887148e-02 1.14071131e+00
2.77040899e-01 -6.80414438e-01 -1.12928772e+00 1.08766699e+00
-2.00670838e+00 -1.31618249e+00 -6.17765784e-01 2.61595154e+00
5.06116450e-01 1.83305487e-01 6.02460146e-01 8.08144510e-01
1.05601406e+00 -3.20655495e-01 -5.81006467e-01 -4.35959071e-01
1.18250564e-01 -2.03467920e-01 5.53980589e-01 1.70767620e-01
-5.95225155e-01 4.42829996e-01 6.26314831e+00 6.21828914e-01
-1.08838844e+00 2.88461167e-02 4.19588447e-01 -3.85865718e-01
-2.36222580e-01 4.73456047e-02 -5.82598388e-01 9.64460552e-01
1.49414825e+00 -7.20285952e-01 4.89003807e-01 5.82397401e-01
6.69258237e-01 -5.05927563e-01 -1.04637527e+00 1.14973497e+00
-1.13178104e-01 -1.20375717e+00 -2.30570227e-01 2.50533462e-01
6.98290825e-01 -4.82149273e-02 -4.78960633e-01 -2.08284855e-01
4.91236836e-01 -7.18483686e-01 3.02287132e-01 5.05018950e-01
4.13024783e-01 -1.16349292e+00 7.90445387e-01 9.46649194e-01
-1.31880605e+00 -4.09621894e-01 -5.57625353e-01 -2.49649480e-01
5.41868210e-01 9.80615854e-01 -7.69968867e-01 4.97087091e-01
1.12010920e+00 4.18015480e-01 -2.32183248e-01 1.44178879e+00
6.10249698e-01 8.55325341e-01 -1.07111776e+00 -1.91172123e-01
5.36694415e-02 -3.63114566e-01 1.86543420e-01 1.01586044e+00
4.57835585e-01 7.05741227e-01 5.24175406e-01 6.96527511e-02
9.98868421e-02 2.76639462e-01 -4.62383717e-01 5.33235557e-02
1.01376021e+00 1.27881408e+00 -1.28960133e+00 -6.80403829e-01
-3.04481924e-01 6.13865554e-01 -2.37715855e-01 4.26791236e-02
-6.76184714e-01 -6.60840154e-01 2.60833442e-01 6.30470812e-01
4.54538077e-01 -3.32066089e-01 -1.45596683e-01 -6.77434385e-01
-3.92787047e-02 -4.57833767e-01 7.68375516e-01 -4.18180019e-01
-1.25752342e+00 2.13358819e-01 -4.70705479e-02 -1.38414872e+00
-1.21006303e-01 1.67392254e-01 -7.81918466e-01 7.60539845e-02
-9.03556526e-01 -1.47542045e-01 -4.98415738e-01 1.13188577e+00
4.37380433e-01 1.18353125e-02 3.58963966e-01 6.11419916e-01
-4.56019640e-01 2.55539954e-01 8.40807319e-01 -2.55968627e-02
4.92422581e-01 -9.79753494e-01 -8.40905607e-02 8.06986749e-01
-2.29457781e-01 1.24143958e-01 6.39092028e-01 -5.19972384e-01
-1.33280671e+00 -8.69298577e-01 5.31957030e-01 -2.32280158e-02
4.25730079e-01 -4.25814211e-01 -1.14012384e+00 1.91241413e-01
1.52873456e-01 2.42448777e-01 8.38376284e-01 -2.99595445e-01
2.65149087e-01 -5.39643168e-01 -1.36343014e+00 -1.77939832e-01
5.30409336e-01 2.56025553e-01 -1.04786009e-01 3.59600097e-01
6.30587637e-01 2.44830027e-02 -1.07154012e+00 6.50096834e-02
1.75721511e-01 -1.10646784e+00 5.48597336e-01 -4.06898141e-01
-1.47526518e-01 -5.70930600e-01 -5.73148280e-02 -1.03534997e+00
-8.59115571e-02 -7.59673953e-01 9.84266922e-02 1.35516846e+00
4.10841107e-02 -4.40108448e-01 9.56659973e-01 8.01262796e-01
1.69931978e-01 -2.13426083e-01 -1.03182971e+00 -8.44502151e-01
-5.03369629e-01 -3.67784888e-01 9.37436700e-01 1.22090232e+00
4.72218871e-01 1.60740718e-01 -1.98877320e-01 4.92745042e-01
8.55122685e-01 2.12508544e-01 8.95098686e-01 -1.51084065e+00
3.34914513e-02 -1.10117964e-01 -5.36259174e-01 -7.26354659e-01
-5.33675611e-01 -3.51782411e-01 -7.30166212e-02 -1.59557962e+00
1.39343351e-01 -8.05854261e-01 -4.04647365e-02 7.71519095e-02
6.06887676e-02 -1.48389339e-01 2.51139462e-01 4.82656807e-01
-1.16036010e+00 8.19070414e-02 7.74956703e-01 2.54778862e-01
-3.00218672e-01 3.33215535e-01 -2.02406079e-01 4.51826692e-01
1.07817507e+00 -7.67990828e-01 -7.77479470e-01 -2.90996283e-01
3.81202042e-01 5.15708983e-01 -3.87176484e-01 -1.15556800e+00
9.42667842e-01 -4.67172176e-01 6.81287274e-02 -1.10040963e+00
-1.84025262e-02 -1.39640486e+00 6.18843257e-01 4.45354402e-01
7.16901571e-02 2.81410992e-01 -1.56497359e-01 8.09101641e-01
-4.63300616e-01 -1.86381310e-01 7.85945356e-01 -2.49754995e-01
-6.06933773e-01 5.56569993e-01 -7.03898847e-01 2.54634589e-01
1.71665061e+00 -7.02419937e-01 4.05002907e-02 -5.51670611e-01
-7.18666375e-01 9.34027433e-01 6.69022501e-01 3.97069246e-01
7.17995644e-01 -1.11549401e+00 -7.44661748e-01 1.55774532e-02
-1.57925487e-01 6.52605414e-01 4.63204712e-01 5.92986643e-01
-5.70614219e-01 8.32661837e-02 -1.26948059e-01 -8.35809290e-01
-1.41316628e+00 1.31228209e+00 -2.39084929e-01 1.54320985e-01
-5.14773726e-01 2.93491393e-01 -5.33845186e-01 -1.45595074e-01
3.52788329e-01 -4.35462557e-02 3.28422338e-02 2.90578067e-01
9.22909558e-01 8.32118094e-01 7.90787190e-02 -4.11650538e-01
-2.78138101e-01 3.31328541e-01 1.19658612e-01 -3.54323275e-02
1.48797810e+00 -6.57609105e-01 -1.73870802e-01 8.25212836e-01
1.07223475e+00 -2.96278775e-01 -7.14100063e-01 -1.49149582e-01
2.56674588e-01 -8.41437399e-01 -4.70925301e-01 -1.58245906e-01
-1.24429870e+00 7.92400718e-01 6.20014369e-01 1.09492731e+00
1.29160583e+00 4.57835989e-03 1.05462921e+00 1.82717353e-01
5.11003792e-01 -1.53555453e+00 4.78931703e-02 5.61575964e-02
6.06198646e-02 -1.25647354e+00 -2.30549686e-02 -7.32146949e-02
-7.26803064e-01 1.01402628e+00 4.27548677e-01 -9.44937170e-02
1.03685796e+00 5.65154910e-01 -1.58230171e-01 -4.90080342e-02
-1.12322617e+00 6.06916361e-02 -5.34072340e-01 7.16940820e-01
-1.00477077e-01 5.16213104e-02 -3.38917673e-01 4.09279674e-01
-1.95881620e-01 8.73254463e-02 1.14638579e+00 1.06239736e+00
-1.04185510e+00 -9.48330283e-01 -7.41515696e-01 6.03670239e-01
-5.20183325e-01 5.15076518e-01 -8.15336332e-02 3.06685388e-01
1.16646551e-01 1.39045155e+00 5.65697730e-01 -2.98179358e-01
2.09193006e-01 -1.78357288e-01 -3.75098407e-01 -4.75050956e-01
-3.74827981e-01 1.09087275e-02 -4.82055932e-01 -4.07628417e-01
-4.74485725e-01 -7.30516851e-01 -1.78371119e+00 -6.86360359e-01
-2.84016967e-01 6.11321032e-01 8.29253733e-01 4.97173160e-01
4.24376935e-01 2.11560563e-03 1.25754678e+00 -5.47920048e-01
-2.61860460e-01 -6.52947366e-01 -8.10024440e-01 4.25520718e-01
1.36550903e-01 1.64140970e-01 -2.90284842e-01 5.61239243e-01] | [7.197559833526611, 4.523019790649414] |
6ca608ba-6491-45fc-8450-64739196b631 | scene-graph-generation-with-external | 1904.00560 | null | http://arxiv.org/abs/1904.00560v1 | http://arxiv.org/pdf/1904.00560v1.pdf | Scene Graph Generation with External Knowledge and Image Reconstruction | Scene graph generation has received growing attention with the advancements
in image understanding tasks such as object detection, attributes and
relationship prediction,~\etc. However, existing datasets are biased in terms
of object and relationship labels, or often come with noisy and missing
annotations, which makes the development of a reliable scene graph prediction
model very challenging. In this paper, we propose a novel scene graph
generation algorithm with external knowledge and image reconstruction loss to
overcome these dataset issues. In particular, we extract commonsense knowledge
from the external knowledge base to refine object and phrase features for
improving generalizability in scene graph generation. To address the bias of
noisy object annotations, we introduce an auxiliary image reconstruction path
to regularize the scene graph generation network. Extensive experiments show
that our framework can generate better scene graphs, achieving the
state-of-the-art performance on two benchmark datasets: Visual Relationship
Detection and Visual Genome datasets. | ['Handong Zhao', 'Jiuxiang Gu', 'Jianfei Cai', 'Zhe Lin', 'Sheng Li', 'Mingyang Ling'] | 2019-04-01 | scene-graph-generation-with-external-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Gu_Scene_Graph_Generation_With_External_Knowledge_and_Image_Reconstruction_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Gu_Scene_Graph_Generation_With_External_Knowledge_and_Image_Reconstruction_CVPR_2019_paper.pdf | cvpr-2019-6 | ['visual-relationship-detection'] | ['computer-vision'] | [ 5.76029122e-01 2.54818738e-01 -3.07255477e-01 -6.16835237e-01
-8.79150927e-02 -3.25319231e-01 5.03817081e-01 2.19455734e-01
-8.49423036e-02 5.58668613e-01 2.78013259e-01 -1.52413502e-01
4.70197164e-02 -1.03973389e+00 -9.29538786e-01 -3.93147051e-01
4.31692719e-01 3.29699367e-01 3.76080960e-01 -8.63184780e-02
1.63856924e-01 1.14214070e-01 -1.58512402e+00 3.49008292e-01
1.03312600e+00 9.05438006e-01 3.88798922e-01 2.46011391e-01
-3.04862261e-01 1.17920077e+00 -2.39365906e-01 -7.07219601e-01
2.61206757e-02 -3.50818008e-01 -9.87530768e-01 4.01402384e-01
4.38360989e-01 -2.49564275e-01 -5.75538516e-01 1.29904830e+00
3.95861745e-01 4.89187613e-02 4.93117541e-01 -1.58173954e+00
-1.02249360e+00 8.71434748e-01 -6.49552643e-01 -6.57712445e-02
3.71203989e-01 2.12522388e-01 1.24721813e+00 -7.75161088e-01
9.03678954e-01 1.37871623e+00 4.36868668e-01 4.70298469e-01
-1.40794563e+00 -7.78485894e-01 4.38198477e-01 5.42214096e-01
-1.55611742e+00 -3.39364707e-01 9.22183216e-01 -4.08822447e-01
7.97246516e-01 1.34332895e-01 6.72420442e-01 1.03269315e+00
-4.24792171e-01 1.02700591e+00 9.53451753e-01 -2.77289301e-01
-1.70341328e-01 3.18073541e-01 -3.34065333e-02 8.08643997e-01
4.67437774e-01 -2.84909904e-01 -4.92125064e-01 2.40673825e-01
8.37852538e-01 -9.39833522e-02 -3.22124451e-01 -6.19586647e-01
-1.30855477e+00 6.13310635e-01 9.16516185e-01 -1.64172798e-01
-1.99050400e-02 2.17177525e-01 2.84742832e-01 -6.98280856e-02
4.30723339e-01 4.27506983e-01 -2.63475537e-01 3.00605953e-01
-2.30041802e-01 2.12672561e-01 6.10415697e-01 1.48341227e+00
8.81380558e-01 -1.61677197e-01 -3.11441869e-01 9.19258237e-01
2.65463114e-01 4.55509454e-01 1.04399435e-01 -7.41165757e-01
6.84045672e-01 1.12379253e+00 -2.70856559e-01 -1.47788036e+00
-3.70132267e-01 -3.80568326e-01 -1.02301288e+00 -4.21385884e-01
1.73177451e-01 3.52708340e-01 -9.43498969e-01 1.78273797e+00
4.48452622e-01 3.89316797e-01 -7.93923438e-02 9.58826959e-01
1.47777629e+00 4.73098725e-01 2.61562407e-01 -9.72966999e-02
1.27370203e+00 -1.00512481e+00 -7.22605944e-01 -3.44811678e-01
5.78779161e-01 -5.94496429e-01 1.23634565e+00 9.38500538e-02
-7.95909226e-01 -6.87735558e-01 -7.00854897e-01 -3.08874995e-01
-3.26978564e-01 6.11962425e-03 1.01506221e+00 1.93112373e-01
-5.87267995e-01 1.63060144e-01 -4.70060825e-01 -4.60561305e-01
8.51555169e-01 1.91391394e-01 -5.50148547e-01 -4.45011020e-01
-9.55367446e-01 5.62193334e-01 8.76343846e-01 5.90946861e-02
-8.67282271e-01 -5.44776678e-01 -1.03623378e+00 -1.46714792e-01
7.58015811e-01 -1.03188348e+00 8.17110419e-01 -8.81068170e-01
-9.07414794e-01 1.02834153e+00 -5.52554578e-02 -2.36336663e-01
3.55331838e-01 -1.45940512e-01 -3.63175869e-01 -1.35505229e-01
3.75805676e-01 9.70715106e-01 7.10518003e-01 -1.58601403e+00
-6.63496375e-01 -3.26084018e-01 1.87682554e-01 3.42764705e-01
-3.03072274e-01 -1.55348644e-01 -1.00200284e+00 -6.46891713e-01
2.95349956e-01 -8.12340498e-01 -3.87322187e-01 -1.19791187e-01
-9.19089794e-01 -3.76555771e-01 7.96899557e-01 -3.22264075e-01
1.09681749e+00 -2.11677599e+00 1.72387645e-01 1.05233453e-01
3.54215503e-01 2.80263629e-02 -2.84856260e-01 9.37443525e-02
-1.27250567e-01 1.77011564e-01 -1.23119615e-01 -2.75225103e-01
-1.98011994e-01 5.30839980e-01 -4.16936368e-01 8.85594860e-02
4.23946649e-01 1.10594678e+00 -1.15069115e+00 -8.99810970e-01
1.30940259e-01 2.51213253e-01 -6.73770666e-01 3.49106073e-01
-5.13919950e-01 5.66103339e-01 -6.95088506e-01 7.12747037e-01
5.98285139e-01 -6.93946183e-01 4.78390120e-02 -6.30066931e-01
3.18233490e-01 2.00602248e-01 -1.12646353e+00 1.78939402e+00
-9.27049220e-02 4.99156922e-01 -4.46370006e-01 -9.92356360e-01
9.04609144e-01 -1.19887717e-01 2.96957731e-01 -6.46911442e-01
1.47700563e-01 -2.23628700e-01 -7.46279210e-02 -6.77201748e-01
4.72047418e-01 -5.33502270e-03 5.95526099e-02 6.42796159e-02
1.07044829e-02 -4.83173013e-01 2.59779841e-01 6.07600152e-01
9.16836739e-01 2.39488915e-01 2.54238874e-01 5.81170619e-02
4.57624286e-01 1.46912545e-01 7.22566724e-01 6.44802809e-01
-8.38057622e-02 5.98338664e-01 5.78118682e-01 -3.72699916e-01
-7.95487106e-01 -8.91791999e-01 -4.85508703e-02 9.05991673e-01
7.83955276e-01 -7.15838611e-01 -4.69984353e-01 -9.62096512e-01
-9.12204161e-02 6.21427953e-01 -3.92771780e-01 -4.27987635e-01
-3.95918876e-01 -7.74926305e-01 4.07726973e-01 6.10974371e-01
6.71621442e-01 -1.28750813e+00 4.42882441e-02 5.52976206e-02
-5.71189582e-01 -1.74615717e+00 -2.28362441e-01 -9.59358811e-02
-6.13320768e-01 -1.21522772e+00 -2.42867647e-03 -8.84681404e-01
1.07759857e+00 4.75130796e-01 1.49733329e+00 2.54969358e-01
-3.90221357e-01 2.85748005e-01 -4.61134017e-01 -4.56258535e-01
-2.33294889e-01 -8.57762434e-03 -1.79280818e-01 6.24527447e-02
2.98595458e-01 -4.03377265e-01 -6.07378900e-01 3.16448689e-01
-8.07103932e-01 6.61713302e-01 5.26430845e-01 8.50220919e-01
9.03338015e-01 2.43435651e-01 3.91343683e-01 -1.20509565e+00
3.90121639e-01 -3.55176985e-01 -5.41874647e-01 2.28086337e-01
-5.91061831e-01 9.55891237e-02 4.97643650e-01 -2.78447866e-01
-1.23696625e+00 2.50064850e-01 1.65545255e-01 -2.99854517e-01
-2.83754647e-01 3.82873207e-01 -6.62239730e-01 -6.75850511e-02
5.89803874e-01 2.95295388e-01 -3.54944646e-01 -2.27572262e-01
8.35048079e-01 2.57195622e-01 6.87259555e-01 -5.19073009e-01
9.81230319e-01 5.04105389e-01 1.40107185e-01 -5.56391180e-01
-1.46884143e+00 -6.05604053e-01 -6.04280472e-01 -1.15535550e-01
8.75268400e-01 -1.15964675e+00 -4.63626176e-01 2.92882413e-01
-1.18374538e+00 -1.02077134e-01 -3.02065402e-01 2.40174562e-01
-5.32044530e-01 4.06265855e-01 -3.74381542e-01 -4.14853245e-01
-1.47683859e-01 -1.29143882e+00 1.19437146e+00 2.69507527e-01
9.10750777e-02 -7.10691392e-01 -2.86734551e-01 6.54386759e-01
4.63929353e-03 3.10131043e-01 1.08924079e+00 -3.17534089e-01
-1.23442638e+00 6.13579527e-02 -8.77632201e-01 3.18397909e-01
1.52680874e-01 -1.35423183e-01 -8.77963483e-01 3.58164497e-02
-4.65441495e-01 -5.15923440e-01 9.10249233e-01 1.41018182e-01
1.64264798e+00 -2.34588176e-01 -6.05832636e-01 8.48940492e-01
1.43813026e+00 -1.76566020e-01 6.30131781e-01 1.02319486e-01
1.42528510e+00 7.28782356e-01 7.71214962e-01 4.81153369e-01
7.94379950e-01 7.17162251e-01 6.52280033e-01 -1.79788828e-01
-5.27522087e-01 -6.01178646e-01 -5.89147955e-02 7.10584760e-01
-7.40418732e-02 -3.66423070e-01 -8.07943523e-01 7.69526720e-01
-2.04165363e+00 -8.53004813e-01 -5.19596934e-01 1.68457878e+00
9.97442901e-01 1.54217646e-01 -1.86802968e-01 -1.50686786e-01
7.31124222e-01 1.23670712e-01 -5.11170924e-01 2.53790885e-01
-3.41418833e-01 -2.80632347e-01 5.15642047e-01 1.73325151e-01
-1.13486052e+00 1.50128818e+00 5.49842215e+00 8.29556942e-01
-6.30532324e-01 -1.49691358e-01 7.18695521e-01 3.63841951e-01
-5.21203399e-01 3.39887589e-01 -9.50928807e-01 1.50729224e-01
1.07409451e-02 4.90385145e-02 3.40342999e-01 9.95817006e-01
-7.26772398e-02 1.47572920e-01 -1.17046785e+00 1.37014449e+00
2.51067728e-01 -1.29266524e+00 6.49240255e-01 -1.45057380e-01
7.76644051e-01 -1.05416521e-01 -1.68222770e-01 2.47739717e-01
4.32139099e-01 -1.00498950e+00 6.31201148e-01 3.54545206e-01
7.03023195e-01 -6.09418452e-01 5.95977306e-01 3.03624719e-01
-1.41086388e+00 1.97895646e-01 -5.15257120e-01 -2.22580344e-03
1.34890959e-01 6.78277552e-01 -1.06868315e+00 5.97408235e-01
7.92601764e-01 1.04168570e+00 -9.03791904e-01 9.37768519e-01
-5.21247149e-01 5.85864186e-01 -1.31229952e-01 1.10528633e-01
-4.34209891e-02 -1.13484986e-01 4.73555475e-01 9.28554952e-01
-1.44456193e-01 2.27520898e-01 4.93833750e-01 1.09885430e+00
-4.22606736e-01 1.91847071e-01 -8.65803242e-01 -8.72147677e-04
4.23404366e-01 1.39604175e+00 -8.30187023e-01 -2.98992008e-01
-5.29279828e-01 8.52584660e-01 5.96010685e-01 3.63695174e-01
-9.70935643e-01 -1.03034385e-01 5.22649705e-01 1.05800092e-01
2.76249293e-02 1.35765970e-02 -3.11429024e-01 -1.29880691e+00
-3.23864594e-02 -7.97352970e-01 4.25694078e-01 -8.69775534e-01
-1.50767052e+00 3.20187777e-01 1.78128049e-01 -1.02819037e+00
1.95587516e-01 -5.18806696e-01 -2.18742371e-01 2.96657950e-01
-1.70029223e+00 -1.56567788e+00 -8.22270334e-01 8.73901367e-01
4.94501829e-01 2.28448361e-02 4.41715211e-01 2.37413853e-01
-5.84723175e-01 4.76694435e-01 -5.66243351e-01 2.22232416e-01
7.37818837e-01 -1.27023792e+00 4.37230408e-01 8.23361039e-01
3.82585108e-01 4.30162638e-01 5.70480466e-01 -8.22740912e-01
-1.44848967e+00 -1.53821623e+00 5.53924024e-01 -6.12150371e-01
6.21125042e-01 -4.48760986e-01 -7.98963666e-01 8.10503304e-01
-1.26497939e-01 4.09658998e-01 5.60481668e-01 1.03311598e-01
-4.99232978e-01 -1.27084598e-01 -7.49882340e-01 8.96805644e-01
1.88407707e+00 -4.43173826e-01 -3.48869413e-01 5.41974127e-01
1.03838277e+00 -4.04375196e-01 -7.22851753e-01 8.31171811e-01
1.67487025e-01 -7.27503717e-01 1.18665230e+00 -6.76143587e-01
4.52199697e-01 -5.97059071e-01 -2.00977519e-01 -9.85829413e-01
-4.62728173e-01 -3.26516241e-01 -6.26632273e-02 1.53471184e+00
3.27658206e-01 -3.14989805e-01 7.92708755e-01 4.96932477e-01
-1.37345523e-01 -4.89813507e-01 -3.99760753e-01 -4.54072475e-01
-6.26547873e-01 -5.69554985e-01 6.18615329e-01 1.09341228e+00
-1.87576890e-01 9.21557546e-01 -4.84511018e-01 1.56408593e-01
6.83013082e-01 4.50748831e-01 1.32809639e+00 -1.09894693e+00
-2.86542565e-01 -2.15249956e-01 -6.96460783e-01 -1.26910186e+00
3.34796518e-01 -1.08877385e+00 1.44179240e-01 -1.83672893e+00
7.07080543e-01 -7.25461483e-01 -1.75100818e-01 6.28506899e-01
-6.20980322e-01 3.49287838e-01 1.56120300e-01 3.03320680e-02
-1.06753838e+00 7.65177190e-01 1.72908151e+00 -3.69140804e-01
3.91502492e-02 -3.13170642e-01 -9.15519595e-01 9.86849427e-01
4.49947983e-01 -4.17145610e-01 -8.71285439e-01 -4.99223202e-01
5.24526119e-01 -1.99146166e-01 6.27672195e-01 -7.48635709e-01
9.72948670e-02 -3.87365937e-01 4.03792500e-01 -7.08580732e-01
1.36072695e-01 -8.54256153e-01 1.16394654e-01 1.30244300e-01
-2.29802847e-01 -2.51890570e-01 -5.24816550e-02 8.93583357e-01
-2.77306527e-01 6.70313910e-02 5.49040139e-01 -2.24758446e-01
-1.09621489e+00 5.85681260e-01 4.24680650e-01 2.52879649e-01
9.89264965e-01 -2.15404883e-01 -6.15723431e-01 -2.53973186e-01
-5.03570497e-01 3.83471757e-01 5.40693521e-01 6.64547384e-01
8.77760351e-01 -1.41624832e+00 -7.41020322e-01 1.13883764e-01
7.46537328e-01 5.08003354e-01 1.03569798e-01 6.24433756e-01
-3.24768275e-01 5.04310569e-03 7.23934099e-02 -8.49628031e-01
-1.30328965e+00 7.52144992e-01 -5.82046136e-02 -1.74472973e-01
-7.28105426e-01 1.03294122e+00 7.93833435e-01 -3.48331839e-01
1.74324289e-01 -4.16660964e-01 -3.22910756e-01 -2.35960081e-01
1.56587780e-01 2.82297190e-02 -2.52094984e-01 -8.37593019e-01
-2.55919874e-01 5.94314575e-01 -1.00173190e-01 5.47340751e-01
1.07246757e+00 -2.56440043e-01 -3.43894631e-01 2.04578042e-01
8.37169349e-01 -8.82751495e-02 -9.89519477e-01 -4.34771866e-01
-2.98503749e-02 -6.75913393e-01 -1.47013679e-01 -5.84198594e-01
-1.06371951e+00 6.17490888e-01 2.45930061e-01 1.69828027e-01
1.20578051e+00 4.42131162e-01 7.96836078e-01 4.22696739e-01
4.18149501e-01 -8.02405179e-01 3.77867341e-01 3.56029958e-01
8.57534349e-01 -1.62875736e+00 2.05499515e-01 -1.20710504e+00
-7.72825301e-01 5.88341236e-01 1.10597324e+00 6.03460073e-02
5.24829030e-01 -1.07840694e-01 -2.54200190e-01 -4.46702659e-01
-6.33686006e-01 -5.19893944e-01 5.54301381e-01 7.24602878e-01
3.41017962e-01 1.80635720e-01 -1.14233509e-01 4.69961554e-01
-1.49582833e-01 -1.73736274e-01 2.10051879e-01 5.90282440e-01
-2.85205215e-01 -1.03904045e+00 5.19194007e-02 4.96119767e-01
-1.90985143e-01 -3.31182867e-01 -5.18806100e-01 7.35062897e-01
4.09402043e-01 8.87009680e-01 -1.70799851e-01 -3.55708927e-01
4.41098541e-01 -3.89627039e-01 4.89288956e-01 -8.55883420e-01
-4.35517542e-02 -1.10505134e-01 4.24112827e-01 -7.00820744e-01
-6.52367949e-01 -3.24578434e-01 -1.36726594e+00 9.85301682e-04
-7.28534758e-01 -2.33371645e-01 2.28949904e-01 9.40253973e-01
3.76783520e-01 6.63711727e-01 3.78434658e-01 -4.23288792e-01
-2.55216975e-02 -7.60769248e-01 -2.88194150e-01 1.11666310e+00
-6.09506071e-02 -6.96676075e-01 2.26003379e-02 5.05178511e-01] | [10.327539443969727, 1.6361089944839478] |
f3480a0f-2157-4097-b17b-9bcdc6f57000 | mult-an-end-to-end-multitask-learning | 2205.08303 | null | https://arxiv.org/abs/2205.08303v1 | https://arxiv.org/pdf/2205.08303v1.pdf | MulT: An End-to-End Multitask Learning Transformer | We propose an end-to-end Multitask Learning Transformer framework, named MulT, to simultaneously learn multiple high-level vision tasks, including depth estimation, semantic segmentation, reshading, surface normal estimation, 2D keypoint detection, and edge detection. Based on the Swin transformer model, our framework encodes the input image into a shared representation and makes predictions for each vision task using task-specific transformer-based decoder heads. At the heart of our approach is a shared attention mechanism modeling the dependencies across the tasks. We evaluate our model on several multitask benchmarks, showing that our MulT framework outperforms both the state-of-the art multitask convolutional neural network models and all the respective single task transformer models. Our experiments further highlight the benefits of sharing attention across all the tasks, and demonstrate that our MulT model is robust and generalizes well to new domains. Our project website is at https://ivrl.github.io/MulT/. | ['Mathieu Salzmann', 'Sabine Süsstrunk', 'Tong Zhang', 'Deblina Bhattacharjee'] | 2022-05-17 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Bhattacharjee_MulT_An_End-to-End_Multitask_Learning_Transformer_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Bhattacharjee_MulT_An_End-to-End_Multitask_Learning_Transformer_CVPR_2022_paper.pdf | cvpr-2022-1 | ['edge-detection'] | ['computer-vision'] | [ 1.76103771e-01 1.07873611e-01 -7.40359500e-02 -4.23940897e-01
-1.06135809e+00 -3.49811882e-01 5.80061376e-01 -2.35819682e-01
-4.00401264e-01 2.51647353e-01 2.57380366e-01 -1.32400051e-01
2.72635847e-01 -4.55249578e-01 -1.09311318e+00 -4.13911432e-01
2.61665910e-01 5.43380558e-01 6.14534736e-01 1.98452666e-01
2.30691761e-01 -4.48943228e-02 -1.29303575e+00 6.30661488e-01
7.50409126e-01 1.25844717e+00 6.73838019e-01 8.70951056e-01
5.29464567e-03 1.15293372e+00 -1.03041276e-01 -3.76032978e-01
2.61343598e-01 2.47901380e-01 -8.60219300e-01 -2.22175084e-02
8.61349106e-01 -5.56825101e-01 -3.69342864e-01 7.92703509e-01
4.39427733e-01 -1.46084890e-01 6.35654151e-01 -1.20040560e+00
-8.41865063e-01 2.24568531e-01 -8.29920828e-01 2.35341981e-01
-4.17124815e-02 1.82801977e-01 1.17176008e+00 -1.13712788e+00
3.81004363e-01 1.32102430e+00 8.11680734e-01 4.57055390e-01
-9.98547733e-01 -6.27888441e-01 4.97886389e-01 2.31516004e-01
-9.24483061e-01 -4.63941842e-01 6.04839027e-01 -4.89350349e-01
1.41065383e+00 -4.33734447e-01 4.36413288e-01 1.20408785e+00
5.55524111e-01 1.38461590e+00 1.25121748e+00 -1.46750689e-01
-6.13789819e-02 -4.61265534e-01 4.44886610e-02 1.07793128e+00
-7.52955154e-02 -1.59578308e-01 -7.38450408e-01 3.55355531e-01
1.09824896e+00 -8.12136307e-02 -1.72268488e-02 -4.25573677e-01
-1.23816645e+00 6.50396347e-01 5.76091945e-01 -1.21886954e-01
-3.47762674e-01 7.83151031e-01 5.10591030e-01 2.36341450e-02
9.61097836e-01 -1.41483307e-01 -8.36628377e-01 -1.13488762e-02
-6.70505047e-01 1.15208164e-01 4.13790584e-01 8.31025243e-01
8.14984858e-01 -2.81365395e-01 -3.92504305e-01 8.32006216e-01
4.32596207e-01 4.21902239e-01 3.38598251e-01 -9.05024946e-01
4.10188466e-01 3.97757381e-01 -2.87309110e-01 -3.16064209e-01
-4.81223047e-01 -4.64872420e-01 -4.36551511e-01 3.66962254e-01
2.65184760e-01 -1.93765029e-01 -1.45520806e+00 1.69868767e+00
6.66600764e-02 5.67305207e-01 -2.58934572e-02 8.33631217e-01
9.93506491e-01 4.05871034e-01 1.73596933e-01 5.30681610e-01
1.34126878e+00 -1.61799502e+00 -3.34086418e-01 -1.01804805e+00
5.48938632e-01 -8.09143245e-01 9.20628667e-01 4.69864637e-01
-1.09397280e+00 -5.16109288e-01 -7.38185585e-01 -8.00284326e-01
-1.39867261e-01 3.02001864e-01 8.87259126e-01 5.10188416e-02
-1.38658333e+00 1.79478854e-01 -1.18249190e+00 -3.08589846e-01
9.00954604e-01 1.47206649e-01 -2.89785504e-01 -3.35846066e-01
-5.93722820e-01 9.90409374e-01 1.15071595e-01 5.58187664e-02
-1.45281863e+00 -8.34037423e-01 -1.12036228e+00 -8.08197334e-02
2.17223331e-01 -1.13748705e+00 1.68247962e+00 -7.29595840e-01
-1.25601494e+00 1.26845074e+00 -4.92411524e-01 -5.94569802e-01
5.09748161e-01 -5.63484013e-01 3.77237797e-01 -2.03757025e-02
3.79286617e-01 9.86625791e-01 9.61844444e-01 -9.92005885e-01
-8.69375229e-01 -5.36282599e-01 1.45413820e-02 3.23880106e-01
-4.74923961e-02 -8.76234844e-02 -9.29576337e-01 -5.91772676e-01
-1.28581777e-01 -7.55213797e-01 -1.71674430e-01 3.05374891e-01
-5.33033311e-01 -4.72010076e-01 8.59118283e-01 -5.84298790e-01
4.30397213e-01 -2.06965756e+00 4.96959805e-01 -3.98618132e-01
6.52246594e-01 -1.88098475e-01 -2.35089093e-01 -9.10982192e-02
1.78136677e-01 -2.82545000e-01 -1.90572515e-01 -1.16711485e+00
-7.16474727e-02 2.26134822e-01 -7.17512816e-02 2.93866068e-01
1.82419673e-01 1.38520265e+00 -6.43405497e-01 -3.20037186e-01
4.22362328e-01 5.60490847e-01 -3.15366685e-01 1.12278134e-01
-4.93540585e-01 4.27142560e-01 -5.18035531e-01 8.56530309e-01
5.42164385e-01 -5.85708082e-01 -2.58878022e-01 -4.41685498e-01
-7.96091035e-02 4.69849765e-01 -4.60086644e-01 2.33211398e+00
-8.53185833e-01 9.17133033e-01 2.03935787e-01 -9.29063618e-01
6.77374065e-01 5.34548201e-02 2.06709787e-01 -9.33603466e-01
1.18161820e-01 1.78052008e-01 -3.84587705e-01 -4.32689071e-01
1.86487526e-01 7.22208172e-02 -3.15129459e-02 3.92423183e-01
5.24110794e-01 -1.16105244e-01 -2.71051258e-01 1.05591476e-01
1.05587888e+00 4.34824824e-01 1.19470721e-02 -2.08129466e-01
7.12877586e-02 -2.01918721e-01 5.08058488e-01 5.52897036e-01
-1.39012054e-01 6.87124372e-01 4.27408159e-01 -5.40895879e-01
-9.05458033e-01 -1.04600930e+00 -4.73788455e-02 1.42383265e+00
1.18577786e-01 -2.64914513e-01 -5.73463500e-01 -6.48651779e-01
2.97722638e-01 3.97398382e-01 -1.02388060e+00 -2.43523810e-03
-2.49332279e-01 -6.47732198e-01 5.29812872e-01 9.04693902e-01
7.43555903e-01 -8.94682169e-01 -8.24108899e-01 5.38638681e-02
-1.95722133e-01 -1.61634910e+00 -5.85844517e-01 4.36303735e-01
-8.12442899e-01 -1.05550408e+00 -7.52513826e-01 -9.88661706e-01
3.07441473e-01 3.84193152e-01 1.25851047e+00 -2.81367809e-01
-3.73072147e-01 5.50393522e-01 -9.89301577e-02 -5.77850342e-01
1.29228294e-01 2.63718784e-01 -4.75743085e-01 -3.93024366e-03
2.46129200e-01 -5.51218569e-01 -6.99962378e-01 1.64187074e-01
-6.19132638e-01 5.84383428e-01 6.28393352e-01 6.10346079e-01
7.78783798e-01 -5.05677998e-01 3.69769424e-01 -8.75679433e-01
2.88552821e-01 -2.09305495e-01 -5.26274681e-01 3.59531373e-01
-3.52145821e-01 2.90176272e-01 7.18401149e-02 -6.87296316e-02
-1.09556901e+00 8.10749978e-02 -7.90125877e-02 -7.34755635e-01
-9.82550830e-02 2.24912494e-01 -2.93069817e-02 -3.48606855e-01
2.93721706e-01 2.72973120e-01 1.80530790e-02 -5.35636306e-01
3.74027550e-01 2.75169075e-01 6.87322795e-01 -5.76341569e-01
4.77593571e-01 6.33208036e-01 -9.33605954e-02 -5.25379002e-01
-1.27569568e+00 -2.85856873e-01 -6.17168248e-01 -1.83588624e-01
1.28263044e+00 -1.51883280e+00 -5.66711187e-01 1.20246935e+00
-1.33107042e+00 -1.13790381e+00 -9.98088974e-04 2.57306933e-01
-7.31878638e-01 1.39660716e-01 -7.82181263e-01 -4.46575552e-01
-5.39240837e-01 -1.49461186e+00 1.73514664e+00 1.33208692e-01
2.27594748e-01 -1.31673622e+00 -8.32157061e-02 6.72255158e-01
3.12354207e-01 2.30601251e-01 8.36082518e-01 -2.73766041e-01
-9.92985606e-01 2.72333771e-01 -5.79603910e-01 2.74070770e-01
-7.30033815e-02 -2.72675365e-01 -1.33250475e+00 -2.70796239e-01
-2.83292383e-01 -8.22360694e-01 1.73132920e+00 8.92647028e-01
1.41971993e+00 2.54090488e-01 -4.41751599e-01 1.26051247e+00
1.39892375e+00 -2.78294384e-01 5.22680461e-01 4.71046418e-01
1.16932261e+00 2.98217863e-01 3.16908181e-01 1.43513516e-01
1.17442918e+00 5.28521895e-01 9.07618582e-01 -4.01951432e-01
-4.25046384e-01 -1.31501913e-01 4.36441749e-01 2.32963964e-01
1.39479944e-02 -1.59476176e-01 -9.90432084e-01 7.40737021e-01
-1.99811018e+00 -4.75012988e-01 -1.23370804e-01 1.82116532e+00
3.92402321e-01 3.66664767e-01 8.19502696e-02 -4.28820699e-01
3.94223005e-01 2.84544796e-01 -1.06087863e+00 -3.31247061e-01
-9.69599411e-02 3.73707861e-01 8.38462055e-01 5.30691147e-01
-1.33726096e+00 1.36282730e+00 5.96437931e+00 5.19613683e-01
-1.16707921e+00 4.69877988e-01 8.00962627e-01 -2.21714020e-01
-1.85036466e-01 -1.87062427e-01 -7.39442408e-01 1.10324778e-01
4.66258824e-01 -7.37815425e-02 2.67295986e-01 7.35335469e-01
-1.02240987e-01 -1.75115779e-01 -1.14210474e+00 1.03246808e+00
2.50245243e-01 -1.35198700e+00 -1.49697945e-01 6.16992712e-02
7.35843718e-01 1.02500260e+00 3.08788449e-01 2.86530703e-01
5.39967000e-01 -1.06650817e+00 9.49797690e-01 3.57658714e-01
9.40949619e-01 -4.64431852e-01 4.23430830e-01 1.07709132e-01
-1.39884746e+00 -1.13139421e-01 -1.71130866e-01 -5.86957820e-02
1.27875924e-01 5.55291235e-01 -6.11366630e-01 3.08973789e-01
8.57516587e-01 1.26894379e+00 -7.35167086e-01 1.02876210e+00
-4.72718686e-01 2.85380989e-01 -2.47434318e-01 5.15100002e-01
3.59930694e-01 2.30651915e-01 2.65040874e-01 1.17953038e+00
1.54115736e-01 -2.71126747e-01 3.24366182e-01 7.81996787e-01
-3.14655602e-01 -3.27492744e-01 -4.88523424e-01 3.58541071e-01
2.10339755e-01 1.25960159e+00 -7.07653284e-01 -1.84043810e-01
-8.02249253e-01 1.41225207e+00 6.54683113e-01 3.75350952e-01
-8.56339335e-01 -2.74903961e-02 1.02289712e+00 -1.25328064e-01
6.37967825e-01 -2.92260498e-01 -6.39591455e-01 -1.29547775e+00
1.24956958e-01 -4.47954804e-01 3.43353182e-01 -1.11458170e+00
-1.31692898e+00 5.12425542e-01 -1.96390972e-01 -6.63231373e-01
-5.98908514e-02 -9.59927261e-01 -8.67937207e-01 9.25115645e-01
-2.12887335e+00 -1.85138202e+00 -5.11233211e-01 8.99954498e-01
1.01225412e+00 -1.96716078e-02 5.25065482e-01 1.34579420e-01
-6.61246538e-01 5.45918524e-01 -3.31455827e-01 2.27757320e-01
8.76507342e-01 -1.46870327e+00 1.08799815e+00 8.39371383e-01
-2.02256292e-02 3.54454643e-03 2.36877859e-01 -4.67179030e-01
-1.37624681e+00 -1.31582153e+00 6.40950680e-01 -6.04942679e-01
7.34047949e-01 -6.34068310e-01 -7.32813299e-01 1.23998821e+00
2.48815164e-01 3.74872565e-01 2.91657418e-01 2.39676863e-01
-6.20008409e-01 1.76244806e-02 -7.86659002e-01 1.45089880e-01
1.19378924e+00 -7.80133903e-01 -3.53986412e-01 4.93081778e-01
9.01173592e-01 -8.28304946e-01 -6.15596473e-01 3.27120930e-01
6.17352188e-01 -1.01157463e+00 1.06059134e+00 -1.88656360e-01
7.01508760e-01 6.26079813e-02 -1.42145813e-01 -1.22815275e+00
-3.33571583e-01 -1.61335498e-01 -2.36688018e-01 8.24781299e-01
4.77293342e-01 -6.96637809e-01 7.71189690e-01 4.09970522e-01
-5.23448646e-01 -9.86634672e-01 -1.04258299e+00 -4.31739002e-01
2.52609760e-01 -7.53195107e-01 1.20457351e-01 5.53113282e-01
-5.73718369e-01 7.28191316e-01 -4.24623698e-01 2.79785573e-01
1.06997120e+00 2.32844979e-01 5.97185731e-01 -1.35847223e+00
-4.90737230e-01 -4.54786152e-01 -2.75929362e-01 -1.42802560e+00
3.72991443e-01 -9.88609791e-01 1.74355563e-02 -2.06981850e+00
2.76717812e-01 -2.24961162e-01 -2.24589482e-01 1.02162158e+00
-2.02370912e-01 3.77348304e-01 2.96065629e-01 1.23452865e-01
-7.64924169e-01 5.48424780e-01 1.44071841e+00 -1.60712868e-01
9.15876105e-02 -1.36459559e-01 -8.77904654e-01 8.88150394e-01
6.28785253e-01 -2.40885869e-01 -4.34534848e-01 -1.46154630e+00
5.41884638e-02 -1.49848238e-01 8.68281960e-01 -1.01497531e+00
4.38292742e-01 1.55337349e-01 5.25288939e-01 -4.88269091e-01
5.44227004e-01 -5.02687156e-01 -5.37526250e-01 8.92378613e-02
-2.75526434e-01 5.19884452e-02 5.27549326e-01 5.79777300e-01
5.48446588e-02 4.16453421e-01 7.76685417e-01 -1.58708200e-01
-1.19798100e+00 6.87060654e-01 -2.02434268e-02 2.50004739e-01
9.48879421e-01 -1.03306271e-01 -4.83009875e-01 -1.02859892e-01
-7.37656951e-01 6.78877473e-01 4.00054365e-01 5.82007945e-01
7.70141006e-01 -8.74550939e-01 -8.49642456e-01 2.15428412e-01
3.02820534e-01 4.91892606e-01 3.14297885e-01 8.56458127e-01
-2.58356363e-01 3.43360394e-01 -2.95304477e-01 -8.72591078e-01
-1.13643229e+00 1.16571911e-01 6.43228710e-01 -2.46668756e-01
-9.09983933e-01 1.29597187e+00 7.95395255e-01 -3.04222941e-01
3.12970012e-01 -7.13436425e-01 1.54356956e-01 -2.79469997e-01
3.31829846e-01 -4.53199781e-02 1.61253229e-01 -2.98135668e-01
-4.00077939e-01 9.38343406e-01 -4.35121149e-01 -1.14647136e-03
1.55549431e+00 -8.51629749e-02 -6.97402805e-02 4.34912920e-01
1.11103189e+00 -5.90722024e-01 -1.85947335e+00 -3.81426990e-01
-2.44766310e-01 -2.82891840e-01 6.17263556e-01 -9.38121200e-01
-1.24111402e+00 1.07741666e+00 4.23929542e-01 -4.29412395e-01
1.37898552e+00 2.52192885e-01 8.71629000e-01 2.41603836e-01
4.50672597e-01 -7.53206611e-01 2.49795854e-01 7.91023195e-01
7.83692598e-01 -1.49462759e+00 -3.48848343e-01 -3.53822857e-01
-8.28451335e-01 8.68184328e-01 9.54174459e-01 -3.12650561e-01
5.09658456e-01 5.83232164e-01 5.27020358e-03 -3.77212137e-01
-1.18712664e+00 -4.69451994e-01 2.23603547e-01 7.06916869e-01
4.53685015e-01 -1.84193701e-01 4.13242400e-01 3.45400393e-01
1.45279393e-01 2.55144417e-01 2.65835255e-01 6.39065742e-01
-4.19165969e-01 -9.77460563e-01 -1.19005339e-02 4.47304845e-01
-3.57928604e-01 -2.71927297e-01 -3.16773146e-01 5.89176476e-01
1.09455168e-01 5.45740247e-01 3.10425758e-01 -3.60119164e-01
1.89092159e-01 7.98915997e-02 7.05080092e-01 -7.70329714e-01
-3.67021710e-01 3.06315105e-02 -8.53779092e-02 -7.89875865e-01
-3.14395666e-01 -4.90501791e-01 -1.16015172e+00 2.93822885e-02
2.86794901e-01 -4.48842824e-01 6.91647232e-01 1.12066925e+00
4.74341393e-01 8.89904797e-01 1.73390135e-01 -1.13218546e+00
-4.61528264e-02 -1.00509381e+00 -2.20346093e-01 9.66914743e-03
7.16312528e-01 -7.80116796e-01 7.41551444e-02 3.90618257e-02] | [9.689898490905762, 1.2588812112808228] |
2abb756d-dc21-45d5-8a3b-0dac741aa5cd | face-attention-network-an-effective-face | 1711.07246 | null | http://arxiv.org/abs/1711.07246v2 | http://arxiv.org/pdf/1711.07246v2.pdf | Face Attention Network: An Effective Face Detector for the Occluded Faces | The performance of face detection has been largely improved with the
development of convolutional neural network. However, the occlusion issue due
to mask and sunglasses, is still a challenging problem. The improvement on the
recall of these occluded cases usually brings the risk of high false positives.
In this paper, we present a novel face detector called Face Attention Network
(FAN), which can significantly improve the recall of the face detection problem
in the occluded case without compromising the speed. More specifically, we
propose a new anchor-level attention, which will highlight the features from
the face region. Integrated with our anchor assign strategy and data
augmentation techniques, we obtain state-of-art results on public face
detection benchmarks like WiderFace and MAFA. The code will be released for
reproduction. | ['Jianfeng Wang', 'Ye Yuan', 'Gang Yu'] | 2017-11-20 | null | null | null | null | ['occluded-face-detection'] | ['computer-vision'] | [ 6.84486628e-02 -6.81150258e-02 1.54296011e-02 -3.22803795e-01
-4.25530821e-01 -2.84148246e-01 3.25307995e-01 -5.36086440e-01
-2.30219558e-01 4.47363228e-01 -2.59280726e-02 -1.61614474e-02
1.90887243e-01 -6.91159546e-01 -5.50524890e-01 -5.90698957e-01
8.87701735e-02 -1.70377329e-01 1.89007327e-01 -8.60693678e-03
-6.09961115e-02 7.68630564e-01 -1.85802126e+00 4.14899558e-01
6.11959577e-01 1.12199783e+00 -5.85941561e-02 2.96302795e-01
-1.48378178e-01 3.66414070e-01 -6.40784442e-01 -6.64115369e-01
2.24173620e-01 -1.46763483e-02 -3.88825148e-01 3.77919711e-02
8.79165053e-01 -7.27487862e-01 -4.46291327e-01 1.20102406e+00
5.64503491e-01 -9.77149457e-02 3.69028151e-01 -1.40918541e+00
-5.65830648e-01 3.21610630e-01 -1.16303301e+00 3.89511108e-01
2.61761785e-01 6.64490610e-02 5.12184143e-01 -1.40231109e+00
3.86333436e-01 1.73478055e+00 8.53939533e-01 5.87608933e-01
-8.41177046e-01 -1.15152359e+00 2.65438020e-01 2.55436957e-01
-1.68841410e+00 -7.79748917e-01 6.99688137e-01 -2.06293136e-01
6.06369972e-01 1.95505157e-01 3.46065819e-01 1.14572275e+00
-1.50248423e-01 6.72094524e-01 7.52623141e-01 -3.74624461e-01
-3.58771920e-01 2.27594346e-01 1.84927192e-02 8.25422227e-01
5.62432528e-01 1.81132525e-01 -2.24930838e-01 2.55611595e-02
7.54879117e-01 2.24650577e-01 -2.79452175e-01 -2.43934784e-02
-3.06823939e-01 7.54018068e-01 6.31945908e-01 2.45651692e-01
-3.27449590e-02 -1.69679914e-02 2.85019994e-01 -1.10926926e-01
5.95666409e-01 -1.13331743e-01 -2.58084565e-01 3.61672372e-01
-7.74294913e-01 1.14264719e-01 4.31355625e-01 8.02562177e-01
5.15894651e-01 7.99649507e-02 -5.53951561e-01 9.01856065e-01
5.24574518e-01 5.09993970e-01 5.94240390e-02 -6.56932533e-01
2.80451000e-01 8.57356191e-01 4.46670726e-02 -1.13836110e+00
-5.44343472e-01 -6.45338953e-01 -5.41376472e-01 3.44277322e-01
2.99556434e-01 -2.12588653e-01 -1.09681535e+00 1.81839406e+00
6.84448838e-01 3.23244929e-01 -2.70981818e-01 8.50325286e-01
1.14142823e+00 3.99935037e-01 2.34160289e-01 -2.38423064e-01
1.63454676e+00 -1.01627851e+00 -1.07772505e+00 -2.41739824e-01
2.88546860e-01 -9.68091309e-01 8.12071621e-01 1.45365924e-01
-8.34245622e-01 -8.15666318e-01 -1.07126617e+00 -2.14915663e-01
-4.02335405e-01 6.25764787e-01 6.43565536e-01 1.16888070e+00
-9.18887019e-01 2.12487102e-01 -5.61719835e-01 -5.71025610e-01
1.00473499e+00 6.05390608e-01 -4.89222527e-01 -1.43542781e-01
-8.09367716e-01 6.69937015e-01 6.02896921e-02 2.96709538e-01
-8.18004489e-01 -5.20235658e-01 -7.91914642e-01 2.65526414e-01
5.86872041e-01 -1.56690091e-01 1.10965073e+00 -8.90809178e-01
-1.16812921e+00 8.07485938e-01 -1.38065264e-01 -1.57286167e-01
5.80354214e-01 -5.14777422e-01 -6.20703042e-01 -5.90353645e-03
-5.54287024e-02 8.66017997e-01 1.04803026e+00 -1.09511244e+00
-5.39575517e-01 -5.67178965e-01 3.73000167e-02 -2.13770270e-01
-6.51875019e-01 5.90495765e-01 -7.47251570e-01 -7.42965043e-01
-1.72506809e-01 -7.66794205e-01 2.00586498e-01 4.16031927e-01
-2.45275900e-01 -5.19480586e-01 1.49817550e+00 -6.44887686e-01
1.14916205e+00 -2.50324559e+00 -3.54943663e-01 -1.54546708e-01
9.99420509e-02 7.21634030e-01 -2.79531896e-01 -8.74249116e-02
-2.36806095e-01 1.92463592e-01 -2.28061620e-02 -5.27324915e-01
-2.77410954e-01 -1.48512200e-01 -2.27466553e-01 5.09528160e-01
6.43869460e-01 6.81303620e-01 -4.24241781e-01 -5.12789965e-01
1.22666761e-01 8.72472763e-01 -6.28906727e-01 1.80093095e-01
7.90054500e-02 1.91303566e-01 -3.04457992e-01 1.02449608e+00
1.21596026e+00 1.33571737e-02 -1.83745176e-01 -4.04436857e-01
-1.68465208e-02 -1.80752933e-01 -1.29727602e+00 1.32991993e+00
-8.87426287e-02 5.95304847e-01 2.40737319e-01 -4.26433772e-01
8.32174778e-01 2.83146411e-01 1.22617066e-01 -4.78899062e-01
4.30649012e-01 -1.51073381e-01 2.01848179e-01 -5.91121852e-01
2.42588654e-01 1.79985642e-01 5.42647600e-01 5.81687428e-02
7.88066350e-03 6.65929019e-01 -2.10626852e-02 -1.50581561e-02
6.57762587e-01 3.10717762e-01 3.43959741e-02 -2.49277368e-01
6.77444041e-01 -4.75009739e-01 6.51874244e-01 4.64408427e-01
-4.51520771e-01 5.51418602e-01 5.86429596e-01 -4.57526535e-01
-6.82090521e-01 -8.17225993e-01 -4.62808609e-01 1.20258784e+00
-3.45192514e-02 -4.34017748e-01 -1.06956530e+00 -9.41731095e-01
1.99669629e-01 1.68840617e-01 -8.53359103e-01 -2.95208335e-01
-5.98536551e-01 -9.72506344e-01 5.96682012e-01 7.79646933e-01
7.85972655e-01 -9.65638280e-01 -2.60193795e-01 -1.50561303e-01
1.08605437e-01 -1.33013713e+00 -5.66515684e-01 -2.87572861e-01
-4.67166275e-01 -1.25637126e+00 -5.90772271e-01 -8.33451211e-01
7.78751373e-01 5.13427496e-01 6.97235882e-01 4.10373360e-01
-6.89800680e-01 -3.26716416e-02 -2.21516520e-01 -6.65908992e-01
5.85936867e-02 -5.55503704e-02 1.39692560e-01 3.55109036e-01
5.11351049e-01 -3.98387909e-02 -7.90434301e-01 4.43025798e-01
-7.07340479e-01 -3.25477630e-01 5.32899678e-01 6.49676502e-01
2.05921993e-01 -2.54850723e-02 8.35160792e-01 -7.68493950e-01
1.83898732e-01 -2.47760653e-01 -6.29859388e-01 1.23180561e-01
-2.44995505e-01 -2.65441298e-01 2.40155026e-01 -4.97333556e-01
-1.23192370e+00 3.47075671e-01 -2.66125768e-01 -5.64772546e-01
-2.22260952e-01 -2.25631416e-01 -5.20499349e-01 -5.03175676e-01
4.58388537e-01 -3.47170681e-01 5.13315946e-02 -8.51628482e-01
1.85625777e-01 8.43356311e-01 4.00079012e-01 -1.73267066e-01
7.34869301e-01 6.30927801e-01 -1.34499371e-01 -7.69641459e-01
-9.04763401e-01 -2.12288514e-01 -3.89779210e-01 -2.42557064e-01
6.72709346e-01 -9.16224599e-01 -7.85914540e-01 4.92551208e-01
-1.23184574e+00 2.59833366e-01 2.42503509e-01 2.32843265e-01
8.76654312e-02 4.31009948e-01 -5.58388889e-01 -1.15765953e+00
-4.69753534e-01 -1.24617481e+00 1.21515059e+00 5.86056590e-01
2.26824626e-01 -3.02253097e-01 -4.10283148e-01 2.11871907e-01
4.24301982e-01 1.73442081e-01 4.32852924e-01 -4.38759476e-01
-5.57441413e-01 -3.95545453e-01 -6.79481685e-01 3.07110131e-01
2.72044808e-01 3.39429736e-01 -1.50999105e+00 -5.70069373e-01
-1.95252493e-01 -1.35442661e-02 1.11597943e+00 3.47728997e-01
1.41284323e+00 -1.93404108e-02 -6.97443902e-01 7.15331852e-01
1.08931768e+00 2.24592268e-01 7.35062063e-01 3.81657891e-02
6.62919581e-01 7.19231665e-01 5.93636274e-01 2.52472848e-01
-9.41729546e-02 8.75982463e-01 5.26858449e-01 -4.19147849e-01
-5.12253881e-01 -1.11392923e-01 2.15986654e-01 -1.05146497e-01
-1.23618141e-01 -1.47267938e-01 -6.76689267e-01 2.90063143e-01
-1.57614172e+00 -6.69410825e-01 -1.06252596e-01 2.00693560e+00
4.17744935e-01 -3.77319544e-03 2.86972046e-01 1.52801588e-01
1.07935202e+00 -6.16839454e-02 -4.03860807e-01 -1.50411710e-01
-7.29903430e-02 2.84981817e-01 2.81321287e-01 3.35109979e-01
-1.50827110e+00 1.11066782e+00 6.82535744e+00 8.65548015e-01
-9.98177826e-01 3.23874086e-01 7.75666833e-01 -2.27422819e-01
4.40253973e-01 -4.91068393e-01 -1.38963687e+00 5.44950843e-01
5.86610675e-01 3.14968735e-01 1.33313522e-01 9.91487861e-01
1.59280542e-02 9.77835134e-02 -8.68062377e-01 1.08458602e+00
2.82607257e-01 -8.25363576e-01 -6.28522038e-02 7.53364041e-02
4.45215523e-01 -3.54767203e-01 3.82217765e-01 4.68807489e-01
-2.74006307e-01 -1.08983326e+00 3.07720721e-01 1.70954943e-01
9.84185815e-01 -9.96521771e-01 8.92183542e-01 -2.32646152e-01
-1.30336308e+00 -3.81261826e-01 -4.54298943e-01 2.60394216e-02
-1.80691674e-01 2.73592204e-01 -4.97010469e-01 1.44834667e-01
9.30594563e-01 3.74234110e-01 -7.86692083e-01 1.24868953e+00
-1.20233096e-01 3.75622272e-01 -3.17505121e-01 2.21540049e-01
-6.34225234e-02 1.36605859e-01 3.89093101e-01 1.19319260e+00
2.38846168e-01 8.75506841e-04 1.12913229e-01 8.49730968e-01
-4.51350361e-01 1.60566583e-01 -6.21973276e-01 2.58680969e-01
3.62200320e-01 1.63554072e+00 -7.00881600e-01 -1.17935672e-01
-7.30565012e-01 7.65868127e-01 3.71235549e-01 2.49531060e-01
-1.10950994e+00 -3.47395629e-01 7.78209090e-01 3.16740274e-01
5.18255115e-01 1.77381307e-01 -1.34772304e-02 -7.50154853e-01
2.60953665e-01 -7.40262687e-01 4.56861347e-01 -4.68706191e-01
-9.73951578e-01 7.18899310e-01 -2.16570407e-01 -8.16849828e-01
4.10619318e-01 -8.37392211e-01 -6.49475813e-01 6.58029497e-01
-1.57150722e+00 -1.18943524e+00 -6.03876233e-01 4.28785592e-01
5.61353683e-01 -2.39568278e-01 6.40401900e-01 9.19697404e-01
-1.22838545e+00 1.16385961e+00 -1.99307352e-01 4.13909703e-01
8.46128881e-01 -6.29883349e-01 3.35645527e-01 8.92088056e-01
-2.66739004e-03 5.54131806e-01 4.58624214e-01 -5.77506006e-01
-1.17192101e+00 -1.28263903e+00 3.93933952e-01 -3.49475443e-01
1.77598417e-01 -5.26590943e-01 -9.49775755e-01 6.82582557e-01
-1.10625513e-02 3.14757079e-01 4.07557279e-01 2.07794160e-01
-5.51814616e-01 -3.72174650e-01 -1.50661755e+00 4.70099717e-01
1.17756963e+00 -3.36293757e-01 -2.35692918e-01 4.62444574e-01
6.88682199e-01 -2.94851571e-01 -4.46263015e-01 7.77343810e-01
7.54566729e-01 -8.74512017e-01 1.14968765e+00 -6.06600583e-01
8.05576295e-02 -3.00312072e-01 1.85093489e-02 -7.36354709e-01
-4.63048428e-01 -3.00218582e-01 -3.59658509e-01 1.60794616e+00
8.37293565e-02 -5.42027354e-01 8.66596818e-01 3.41271073e-01
2.26596426e-02 -5.75030744e-01 -1.00255263e+00 -7.68286169e-01
-2.14206323e-01 -1.69640977e-03 6.53961182e-01 5.97127020e-01
-4.89343554e-01 1.12160765e-01 -4.68643546e-01 4.77835774e-01
6.10660255e-01 -1.36199385e-01 4.99483913e-01 -1.23163497e+00
1.27774030e-01 -3.44455093e-01 -5.65717638e-01 -6.82957053e-01
4.74208295e-02 -4.38526660e-01 -1.00505456e-01 -1.07949555e+00
3.73632938e-01 -5.73261036e-03 -3.34784716e-01 5.60557187e-01
-4.95941788e-01 7.48009443e-01 1.95386186e-01 -1.53576612e-01
-5.17708540e-01 5.33956230e-01 1.00332332e+00 -4.46809158e-02
-7.14161992e-02 -1.61792368e-01 -7.25186527e-01 9.01061535e-01
9.23793554e-01 -4.81166512e-01 2.91833840e-02 -4.36956793e-01
-3.91637385e-01 -4.61305678e-01 3.74154061e-01 -1.16217327e+00
1.88524187e-01 3.31615865e-01 9.58626807e-01 -7.61139393e-01
5.13491809e-01 -8.67523193e-01 -6.47003949e-02 6.72324181e-01
1.71123132e-01 -4.59167324e-02 6.99905217e-01 5.20852745e-01
3.55581902e-02 -4.60126437e-02 1.16080892e+00 1.50874346e-01
-6.15919054e-01 4.92761225e-01 2.40066722e-02 -3.64271045e-01
1.20656753e+00 -7.26000294e-02 -4.68002468e-01 7.52650574e-02
-5.73740602e-01 2.05087945e-01 1.97905168e-01 7.64921606e-01
5.62049568e-01 -1.37047350e+00 -8.40854347e-01 6.38650358e-01
6.42948970e-02 -3.49655658e-01 3.78575504e-01 6.73306942e-01
-2.76691496e-01 2.88262397e-01 -2.33523145e-01 -5.31276107e-01
-1.65630627e+00 8.59551191e-01 5.39600968e-01 2.26831302e-01
-5.67309082e-01 1.04251385e+00 5.16419888e-01 -1.64297468e-03
5.82304716e-01 -1.05221003e-01 -5.61811149e-01 6.78276345e-02
1.12589490e+00 4.58312213e-01 5.55068217e-02 -6.94494009e-01
-5.82726777e-01 6.72936201e-01 -4.23872918e-01 5.62964141e-01
1.00840116e+00 3.29335965e-02 3.89992297e-02 -2.74992555e-01
1.00405180e+00 2.58510076e-02 -1.35768509e+00 -9.17726085e-02
-1.84458733e-01 -8.11719060e-01 2.31865793e-01 -7.53724396e-01
-1.51075268e+00 8.38944376e-01 1.32417107e+00 1.04180006e-02
1.01089919e+00 9.13852360e-03 5.46707571e-01 7.41018578e-02
7.14372322e-02 -8.14072132e-01 1.72363520e-02 1.82529151e-01
1.07391405e+00 -1.56992364e+00 3.22857662e-03 -9.27043319e-01
-5.12716249e-02 9.63994741e-01 1.09009147e+00 7.54555911e-02
6.38804793e-01 4.52179581e-01 -1.00581080e-01 -1.65170237e-01
-3.33442599e-01 -4.05781031e-01 2.90486634e-01 5.79334676e-01
5.20228744e-01 -1.89112648e-01 -1.37004465e-01 7.30992496e-01
2.03098804e-01 -1.62868723e-01 3.48898135e-02 6.02777719e-01
-6.12781823e-01 -8.78505409e-01 -7.30822504e-01 4.25013244e-01
-8.84378672e-01 -1.02960896e-02 -5.10730684e-01 8.85442138e-01
4.87728477e-01 1.00509918e+00 1.90674692e-01 -1.82925865e-01
4.06375915e-01 6.09263331e-02 5.32474518e-01 -5.70497990e-01
-4.19306606e-01 1.87457815e-01 -5.46999536e-02 -5.50428629e-01
-2.78123796e-01 -4.96382177e-01 -9.24326241e-01 -3.82671714e-01
-6.70344055e-01 -1.87906861e-01 4.39896166e-01 5.84135473e-01
4.77358431e-01 6.22741818e-01 4.82294232e-01 -6.79428041e-01
-4.17025894e-01 -1.33159578e+00 -4.65047985e-01 3.60185653e-01
3.26019973e-01 -1.26122582e+00 -2.37539947e-01 -3.38340908e-01] | [13.283285140991211, 0.6246783137321472] |
4442e1c7-4cf2-4d79-b821-17c3a8722f98 | examining-european-press-coverage-of-the | 2305.00182 | null | https://arxiv.org/abs/2305.00182v1 | https://arxiv.org/pdf/2305.00182v1.pdf | Examining European Press Coverage of the Covid-19 No-Vax Movement: An NLP Framework | This paper examines how the European press dealt with the no-vax reactions against the Covid-19 vaccine and the dis- and misinformation associated with this movement. Using a curated dataset of 1786 articles from 19 European newspapers on the anti-vaccine movement over a period of 22 months in 2020-2021, we used Natural Language Processing techniques including topic modeling, sentiment analysis, semantic relationship with word embeddings, political analysis, named entity recognition, and semantic networks, to understand the specific role of the European traditional press in the disinformation ecosystem. The results of this multi-angle analysis demonstrate that the European well-established press actively opposed a variety of hoaxes mainly spread on social media, and was critical of the anti-vax trend, regardless of the political orientation of the newspaper. This confirms the relevance of studying the role of high-quality press in the disinformation ecosystem. | ['Daniel Gatica-Perez', 'David Alonso del Barrio'] | 2023-04-29 | null | null | null | null | ['misinformation'] | ['miscellaneous'] | [-3.62018883e-01 2.79992789e-01 -7.25298643e-01 3.41356277e-01
-1.47452548e-01 -9.21401560e-01 1.38972354e+00 1.14275205e+00
-8.17467749e-01 4.03486013e-01 1.24623251e+00 -6.71824098e-01
-1.43468186e-01 -1.00155199e+00 -6.83205903e-01 -4.14896816e-01
-2.60646373e-01 5.56288123e-01 2.17503645e-02 -6.51649535e-01
4.79853362e-01 1.42724708e-01 -5.77578008e-01 -4.71359044e-02
5.28769970e-01 7.09506333e-01 -1.28000572e-01 3.94164383e-01
-5.45293748e-01 6.96346879e-01 -8.04500043e-01 -4.72473443e-01
-6.62137717e-02 -4.76692291e-03 -2.93855876e-01 -4.92562115e-01
-4.30376679e-02 -1.55210510e-01 -3.48287046e-01 1.13486195e+00
4.14964557e-01 -5.56844056e-01 5.37192702e-01 -6.76943004e-01
-4.52793628e-01 5.95292151e-01 -3.32670510e-01 7.14738727e-01
2.98403949e-01 -1.90648243e-01 1.09797907e+00 -6.94867611e-01
1.83627498e+00 1.24931049e+00 8.77048612e-01 -1.40631720e-01
-9.00746524e-01 -5.36384284e-01 7.21634775e-02 -1.59465134e-01
-9.81331050e-01 2.63262421e-01 4.79706317e-01 -1.25280023e+00
6.57341778e-01 2.16149613e-01 1.10470712e+00 1.39388382e+00
1.15509033e+00 -9.11672860e-02 1.39231122e+00 8.14305991e-02
1.83908045e-01 5.08235514e-01 5.05236268e-01 2.53800660e-01
7.04406083e-01 7.78290331e-02 -5.05891204e-01 -8.35112393e-01
1.05206586e-01 2.74855644e-02 -1.42488763e-01 2.05356386e-02
-1.18507636e+00 1.39552665e+00 3.56874377e-01 8.88175189e-01
-8.42657328e-01 -5.34172595e-01 6.50422394e-01 3.47778887e-01
1.24814761e+00 7.29444623e-01 -6.35042787e-01 -1.83398351e-01
-6.14672720e-01 2.21641898e-01 1.03240955e+00 2.21005410e-01
2.26753518e-01 -3.65776271e-01 1.81131884e-01 2.10562333e-01
4.97039139e-01 9.14634883e-01 7.06922635e-02 -5.54275624e-02
5.33464253e-01 4.69176531e-01 5.90366572e-02 -2.02487636e+00
-5.39266646e-01 -7.24176943e-01 -3.75661105e-01 -2.24053457e-01
1.29092231e-01 -6.72848165e-01 -6.81151032e-01 1.63410401e+00
4.73511636e-01 -2.66151160e-01 -1.70988336e-01 7.13802516e-01
5.33733010e-01 8.93693149e-01 4.52840030e-01 -3.09997112e-01
1.85866630e+00 -4.19338822e-01 -1.27098691e+00 -2.82566935e-01
2.93157429e-01 -9.71711755e-01 4.55969712e-03 8.09334144e-02
-5.46244919e-01 3.26553136e-02 -1.16775179e+00 3.82161945e-01
-1.23417592e+00 -7.00780511e-01 2.96641469e-01 4.10259545e-01
-7.72758782e-01 3.73174638e-01 -2.43329644e-01 -9.61921334e-01
3.27797234e-01 -3.68553162e-01 -2.63296455e-01 1.77797630e-01
-1.71828663e+00 1.47104001e+00 2.16688603e-01 -2.72727996e-01
-6.84255183e-01 -7.94399440e-01 -6.05078936e-01 -3.64869803e-01
3.12822580e-01 -3.59797865e-01 4.38591361e-01 -8.14299703e-01
-6.80604279e-01 8.10059905e-01 3.72703880e-01 -5.97167909e-01
4.00003135e-01 -4.15725052e-01 -1.07666600e+00 2.87822157e-01
2.26610586e-01 -1.92392450e-02 5.25072098e-01 -7.46670604e-01
-3.23366851e-01 -6.81584895e-01 -1.01523943e-01 -2.77623117e-01
-3.02065194e-01 7.66034603e-01 2.61904597e-01 -1.03816152e+00
-2.68717021e-01 -7.55876422e-01 -2.08015606e-01 -5.62502265e-01
-2.09864184e-01 -1.32054120e-01 7.38433421e-01 -1.20639634e+00
1.32429969e+00 -2.09967756e+00 -1.47071853e-01 2.00531185e-01
6.66403532e-01 6.54720962e-02 7.91854933e-02 1.20181513e+00
1.07274547e-01 6.41106129e-01 4.30216104e-01 4.81582850e-01
-5.33530302e-02 -4.39845361e-02 -5.57108760e-01 8.38622153e-01
2.85909325e-02 5.59560359e-01 -1.18077588e+00 -3.93473171e-02
-7.17846006e-02 5.97292840e-01 -3.59595716e-01 -3.91730726e-01
-5.93164504e-01 2.60845244e-01 -9.06380177e-01 3.81261468e-01
7.87101805e-01 -1.21229447e-01 7.86205888e-01 -3.01418781e-01
-8.00792277e-01 4.70420957e-01 -1.46737233e-01 1.01512730e+00
5.26597425e-02 1.10114098e+00 5.08604884e-01 -2.82914400e-01
5.30125618e-01 3.12614173e-01 5.62989831e-01 -6.33295894e-01
6.56580567e-01 5.90774938e-02 -1.85938049e-02 -5.57442486e-01
6.89241290e-01 -1.75500572e-01 -3.44609141e-01 2.54892945e-01
-1.30650684e-01 1.82385013e-01 -3.02253664e-01 4.88812119e-01
9.35987711e-01 -5.33329606e-01 2.78536171e-01 -7.64761269e-01
-4.87622842e-02 6.24312878e-01 5.14208317e-01 3.64648879e-01
-4.18323040e-01 1.51948230e-02 6.63569629e-01 -3.43884945e-01
-9.73101616e-01 -8.51385653e-01 -4.24932361e-01 8.17574799e-01
-5.24279885e-02 -6.51160717e-01 -6.25842631e-01 -2.95851797e-01
1.66442603e-01 8.88266742e-01 -8.53896081e-01 4.10003692e-01
-3.21466893e-01 -8.55561435e-01 1.27636850e-01 -4.12667841e-01
2.48277083e-01 -4.44758981e-01 -4.10727501e-01 4.55026835e-01
-5.08531071e-02 -9.28817153e-01 -1.42678931e-01 -1.14922144e-01
-4.42810923e-01 -1.40544140e+00 -7.03564465e-01 -4.00260210e-01
4.44733977e-01 -1.93114012e-01 7.56810546e-01 -7.89674997e-01
5.00900894e-02 3.64811838e-01 -4.12083752e-02 -8.91331613e-01
-7.65344739e-01 -6.33357689e-02 1.38537154e-01 -3.52713555e-01
5.08389831e-01 -7.55345598e-02 -5.63847899e-01 -2.00316355e-01
-8.43484581e-01 -3.40956479e-01 2.41283357e-01 8.11319798e-02
1.27159143e-02 1.76032096e-01 5.25856137e-01 -1.07846427e+00
7.48862624e-01 -1.47507989e+00 -4.94437307e-01 -1.09326050e-01
-7.69033015e-01 -5.04135966e-01 -1.27885133e-01 -2.70535946e-01
-1.02958453e+00 -1.06285000e+00 -9.09303650e-02 5.16133904e-01
2.39036471e-01 1.15069366e+00 5.12890935e-01 2.79187500e-01
5.62079906e-01 -3.35140675e-01 3.98551434e-01 -5.47402084e-01
5.71882844e-01 7.71220386e-01 -1.10870466e-01 1.75053239e-01
7.51089573e-01 7.28815675e-01 -7.04604506e-01 -1.07453203e+00
-5.68313658e-01 -3.90036345e-01 2.40232497e-01 -5.46392381e-01
1.49084616e+00 -1.16230357e+00 -3.20613295e-01 6.14702344e-01
-1.35485637e+00 2.57674545e-01 2.09191546e-01 7.06498861e-01
3.15652251e-01 -3.39781158e-02 -8.05868089e-01 -5.33295572e-01
-3.55976731e-01 -5.83173454e-01 -9.58266202e-03 -6.27881512e-02
-5.13610780e-01 -1.36235321e+00 7.91447163e-01 4.90595639e-01
1.11701262e+00 7.20483899e-01 1.08119106e+00 -1.23520851e+00
-2.31148854e-01 -1.93804473e-01 -1.29902422e-01 -3.51461433e-02
3.05371910e-01 -9.42391083e-02 -3.27350438e-01 -2.32262626e-01
5.22217691e-01 1.92598537e-01 5.99273682e-01 2.40310669e-01
-1.32457361e-01 -8.83092165e-01 -4.93211567e-01 -4.50163364e-01
1.81403434e+00 1.94553703e-01 3.16417217e-01 8.37108195e-01
1.09829754e-01 7.65734434e-01 4.09312814e-01 6.24673963e-01
5.30363679e-01 4.18226421e-01 2.17450410e-01 9.71553251e-02
8.82374123e-02 -4.50318456e-01 6.22158468e-01 1.17220116e+00
2.47367799e-01 -8.91015753e-02 -1.10336471e+00 6.40096366e-01
-1.20129383e+00 -7.19601870e-01 -1.54582962e-01 1.73083746e+00
7.41991699e-01 3.70507240e-01 1.58758145e-02 -7.51573861e-01
8.72873664e-01 8.87464106e-01 1.83470964e-01 -5.63100338e-01
-4.99398232e-01 1.82389561e-02 1.15732968e+00 3.99073571e-01
-9.46139455e-01 6.61760628e-01 7.17009687e+00 7.01096833e-01
-1.19228458e+00 5.91757536e-01 2.87952185e-01 3.00991923e-01
-6.99657679e-01 -2.33157963e-01 -5.34484684e-01 6.26348495e-01
1.29713368e+00 -3.98357987e-01 -8.61535072e-02 5.06513715e-01
5.72251201e-01 -2.23572761e-01 1.10217594e-01 1.05835218e-02
4.78224128e-01 -1.87359142e+00 -3.09717178e-01 2.25470215e-01
7.62482762e-01 9.18944478e-01 4.10222560e-02 -1.39358670e-01
1.68525070e-01 -3.07245672e-01 8.65756452e-01 4.89701807e-01
2.47849539e-01 -3.12996298e-01 9.29493189e-01 -2.75560077e-02
-4.38690275e-01 9.53018367e-02 5.60014248e-02 2.06011590e-02
6.01433575e-01 1.08977258e+00 -6.13607645e-01 5.28327405e-01
3.28719854e-01 6.01646662e-01 -4.39826339e-01 5.22658706e-01
-2.09041327e-01 7.53943861e-01 -9.10132155e-02 -1.43486574e-01
6.00745440e-01 -2.39636153e-01 1.39686477e+00 1.07203484e+00
-3.51748243e-02 -1.16243497e-01 -3.03833455e-01 3.26072127e-01
2.30321974e-01 3.20792705e-01 -8.14353347e-01 -1.19475257e+00
1.71940133e-01 8.34704757e-01 -6.26152098e-01 -3.06627005e-01
-3.30494076e-01 2.15126187e-01 -1.69151619e-01 3.12789410e-01
-7.06298828e-01 -1.47198339e-03 8.25808287e-01 5.44320703e-01
3.19859326e-01 -3.29471290e-01 -9.45355818e-02 -7.19264209e-01
-5.55949867e-01 -1.08072925e+00 3.57607424e-01 -4.74091351e-01
-1.14232540e+00 4.68382537e-01 -1.55648723e-01 -5.23057282e-01
5.07706523e-01 -3.01397830e-01 -3.73916328e-01 4.31773007e-01
-1.14183283e+00 -8.06927800e-01 6.55472636e-01 -1.59407675e-01
1.57386690e-01 -1.75202042e-01 6.15531921e-01 1.92314520e-01
-2.05428183e-01 -4.01683509e-01 4.88261282e-01 -8.63799378e-02
4.99548256e-01 -6.23293757e-01 2.82415450e-01 1.31882161e-01
-6.12502337e-01 8.42798352e-01 1.09402812e+00 -1.54386115e+00
-1.50598693e+00 -8.66618574e-01 1.60212934e+00 -6.24767423e-01
1.82309294e+00 -5.05939186e-01 -3.54121685e-01 4.05341327e-01
1.07116640e+00 -1.14027107e+00 7.77389467e-01 2.44236365e-01
-5.10646999e-01 3.53492379e-01 -1.25793195e+00 5.96783340e-01
3.94459754e-01 -5.52067459e-01 -1.16626525e+00 7.89814413e-01
1.05343962e+00 1.98276535e-01 -1.08573282e+00 8.78773034e-02
5.61275363e-01 -1.99426398e-01 9.08584356e-01 -9.05244470e-01
5.93072772e-01 8.86179358e-02 -1.01721369e-01 -1.32803643e+00
-4.15894806e-01 -5.07436693e-01 5.58361173e-01 9.75373864e-01
4.75142449e-01 -1.27017188e+00 1.44920185e-01 6.10336475e-02
4.26325083e-01 -4.14429456e-01 -1.01678741e+00 -2.84277081e-01
8.08374435e-02 -1.57163695e-01 6.81354804e-03 1.28519225e+00
3.02200504e-02 3.02605510e-01 -1.97071463e-01 2.01102957e-01
3.06897670e-01 -2.51362383e-01 3.19614649e-01 -1.09463596e+00
1.80043459e-01 -2.86706060e-01 -4.87677544e-01 -1.88833594e-01
-3.58262807e-01 -6.69330478e-01 -6.03375018e-01 -1.52584291e+00
2.06439257e-01 -2.25624129e-01 -3.71094376e-01 -3.71377289e-01
2.78598368e-01 -3.39824200e-01 2.97284275e-01 3.52112025e-01
-2.11383462e-01 1.84354872e-01 9.28747058e-01 -3.74931693e-01
1.83749408e-01 -5.01906335e-01 -7.77848125e-01 7.75407434e-01
6.45239592e-01 -8.06922019e-01 1.06611609e-01 -2.32215300e-01
1.13024056e+00 -1.85210034e-01 1.24630295e-01 -5.36642015e-01
1.23435497e-01 -8.45009983e-02 -5.06443679e-02 -7.49995530e-01
-6.34726882e-02 -1.04935801e+00 6.05474651e-01 9.51499403e-01
-1.01546943e-01 3.67734164e-01 2.81143755e-01 8.35674524e-01
-3.74255329e-01 1.92198381e-01 2.28264064e-01 1.51188552e-01
-2.35357165e-01 -6.21656813e-02 -1.08842158e+00 4.34790730e-01
8.79936934e-01 3.14911604e-01 -1.00855660e+00 -2.21595377e-01
-7.19566166e-01 1.11209325e-01 5.04622698e-01 6.40159309e-01
-3.29559222e-02 -9.13919985e-01 -8.88404667e-01 -1.84290513e-01
-1.87912777e-01 -1.12403333e+00 3.80612761e-01 9.93627906e-01
-7.67381012e-01 6.66329026e-01 -9.83254313e-02 7.59268776e-02
-8.90365541e-01 6.57333851e-01 -1.09879076e-01 -3.78739655e-01
-5.40720940e-01 3.88428986e-01 6.93134516e-02 -1.76655471e-01
-3.95395532e-02 1.21014766e-01 -6.19947255e-01 1.13463640e+00
5.87714911e-01 4.35000122e-01 -3.38765085e-01 -7.99700856e-01
-6.93693101e-01 2.55610347e-01 -1.19358391e-01 -1.47470891e-01
1.53259444e+00 -1.66511551e-01 -5.32227099e-01 7.70523727e-01
1.44608665e+00 6.14356399e-01 -1.90717623e-01 -1.40240252e-01
2.36355513e-01 -2.49922022e-01 6.32303357e-02 -1.21318638e+00
-8.59797716e-01 1.21519901e-01 5.26792467e-01 7.04953909e-01
2.58123696e-01 2.28306130e-01 7.59644389e-01 -2.83061892e-01
1.49771452e-01 -1.37219000e+00 -3.61246318e-01 5.24609804e-01
1.16758692e+00 -6.82185233e-01 2.25025535e-01 -2.76729822e-01
-6.03064954e-01 6.10287189e-01 -1.66975677e-01 -2.55828798e-02
1.38968158e+00 8.01627934e-02 3.33123356e-01 -9.66176867e-01
-5.02777994e-01 3.44591260e-01 -1.30586505e-01 3.30390722e-01
1.21631622e-01 4.74523187e-01 -1.30629253e+00 5.97409189e-01
1.10751569e-01 -7.11909682e-02 5.51826119e-01 9.42111015e-01
-4.38503534e-01 -8.03444207e-01 -3.42111200e-01 2.86721975e-01
-9.95538056e-01 -2.78288901e-01 -7.40852714e-01 8.50313067e-01
4.20066059e-01 8.34729195e-01 2.84271598e-01 -3.15338701e-01
7.88581744e-02 -3.46477591e-02 -2.35633105e-01 -1.87373161e-01
-1.16404831e+00 2.31027097e-01 7.64284551e-01 -2.94847488e-01
-4.92005378e-01 -7.50894010e-01 -7.63697386e-01 -6.40441239e-01
-3.48556608e-01 3.03750366e-01 1.28747296e+00 9.10305023e-01
6.26802623e-01 3.44130814e-01 6.11124575e-01 1.07931666e-01
-3.86259735e-01 -9.41943705e-01 -4.92641389e-01 1.32685214e-01
2.85398871e-01 -3.93660873e-01 -6.47125185e-01 -6.52431130e-01] | [8.47394847869873, 9.785730361938477] |
eab11ef5-3414-46b2-8ba5-c7ac84b97476 | agilegan-stylizing-portraits-by-inversion | null | null | https://dl.acm.org/doi/10.1145/3450626.3459771 | https://guoxiansong.github.io/homepage/paper/AgileGAN.pdf | AgileGAN: stylizing portraits by inversion-consistent transfer learning | Portraiture as an art form has evolved from realistic depiction into a plethora of creative styles. While substantial progress has been made in automated stylization, generating high quality stylistic portraits is still a challenge, and even the recent popular Toonify suffers from several artifacts when used on real input images. Such StyleGAN-based methods have focused on finding the best latent inversion mapping for reconstructing input images; however, our key insight is that this does not lead to good generalization to different portrait styles. Hence we propose AgileGAN, a framework that can generate high quality stylistic portraits via inversion-consistent transfer learning. We introduce a novel hierarchical variational autoencoder to ensure the inverse mapped distribution conforms to the original latent Gaussian distribution, while augmenting the original space to a multi-resolution latent space so as to better encode different levels of detail. To better capture attribute-dependent stylization of facial features, we also present an attribute-aware generator and adopt an early stopping strategy to avoid overfitting small training datasets. Our approach provides greater agility in creating high quality and high resolution (1024×1024) portrait stylization models, requiring only a limited number of style exemplars (~100) and short training time (~1 hour). We collected several style datasets for evaluation including 3D cartoons, comics, oil paintings and celebrities. We show that we can achieve superior portrait stylization quality to previous state-of-the-art methods, with comparisons done qualitatively, quantitatively and through a perceptual user study. We also demonstrate two applications of our method, image editing and motion retargeting. | ['Tat-Jen Cham', 'Chuanxia Zheng', 'Chun-Pong Lai', 'Wan-Chun Ma', 'Jing Liu', 'Linjie Luo', 'Guoxian Song'] | 2021-07-19 | null | null | null | acm-transactions-on-graphics-2021-7 | ['motion-retargeting'] | ['computer-vision'] | [ 4.20668572e-01 -1.00877382e-01 -4.07690741e-02 -1.50881216e-01
-6.11232638e-01 -7.90067315e-01 7.41547585e-01 -6.10516965e-01
4.54357266e-02 9.63270843e-01 1.11012064e-01 5.54804243e-02
1.45448878e-01 -9.23640728e-01 -6.96768343e-01 -5.70472956e-01
3.83786857e-01 5.21578968e-01 -2.12955251e-01 -3.54731381e-01
-7.19487816e-02 7.55240560e-01 -1.38388264e+00 2.36801147e-01
9.06392157e-01 6.35113955e-01 5.38929813e-02 6.28364623e-01
-8.94244164e-02 3.44030857e-01 -8.11882854e-01 -7.49016762e-01
3.93982023e-01 -6.59776330e-01 -5.44907331e-01 5.17234921e-01
6.92746639e-01 -4.27639604e-01 -2.08206534e-01 9.42217946e-01
4.64104027e-01 -1.43630132e-01 8.95309091e-01 -1.33193076e+00
-1.14707923e+00 2.78369039e-01 -8.89626205e-01 -3.87543619e-01
3.28098893e-01 3.80523503e-01 8.67621303e-01 -6.05928898e-01
1.05302918e+00 1.54733455e+00 6.15665317e-01 7.99163163e-01
-1.88245082e+00 -9.27559912e-01 -1.47996649e-01 -2.92282641e-01
-1.39031732e+00 -3.65800261e-01 1.04989660e+00 -3.01514089e-01
1.89727515e-01 3.32268625e-01 9.85842466e-01 1.68145072e+00
2.11097851e-01 7.70141602e-01 1.45190561e+00 -2.41380394e-01
8.83590057e-02 3.41942877e-01 -8.08381021e-01 6.79004252e-01
1.03774965e-01 -5.96617088e-02 -4.46855038e-01 -6.06847331e-02
1.54445601e+00 -8.68135989e-02 -2.12185130e-01 -3.56277466e-01
-1.28101254e+00 8.62600982e-01 4.25250500e-01 2.25342780e-01
-3.32206219e-01 3.95184427e-01 2.11448409e-02 3.02422673e-01
4.99449104e-01 7.20286489e-01 1.23538777e-01 -9.09951255e-02
-1.22428238e+00 3.62458885e-01 6.22076690e-01 9.31297183e-01
7.60824740e-01 5.70137501e-01 -2.65748233e-01 8.34280193e-01
-4.03179601e-02 8.08636129e-01 2.54134864e-01 -1.16159201e+00
-1.42401010e-02 4.31637496e-01 1.09939642e-01 -1.10956621e+00
1.62281156e-01 -3.96012604e-01 -1.04863214e+00 6.81430638e-01
2.37827212e-01 -1.71706021e-01 -9.10681665e-01 1.79072046e+00
1.59012690e-01 -4.44296785e-02 -1.11253478e-01 7.83496976e-01
4.74165022e-01 7.99279630e-01 -3.80656980e-02 7.71400928e-02
1.29376006e+00 -5.88057637e-01 -8.52991283e-01 7.89154842e-02
-4.54049446e-02 -9.58221793e-01 1.50062907e+00 3.81204814e-01
-1.17765987e+00 -5.74043274e-01 -1.04107535e+00 -5.31676188e-02
-3.32301930e-02 9.71483216e-02 6.94532156e-01 7.95504928e-01
-1.00801802e+00 8.03718686e-01 -4.55899864e-01 -4.09902275e-01
7.74317563e-01 1.75180301e-01 -5.74601173e-01 1.11132517e-01
-9.34123158e-01 6.72349691e-01 -6.23853356e-02 -3.83466303e-01
-8.13837469e-01 -9.14157808e-01 -6.87628090e-01 -1.26727268e-01
9.51052755e-02 -1.02032173e+00 9.22147214e-01 -1.59641826e+00
-1.97608376e+00 8.49684238e-01 -3.11162677e-02 -1.58351734e-01
7.66491055e-01 8.09925795e-02 -4.14005220e-01 8.05005208e-02
2.26252228e-02 1.19472778e+00 1.37139869e+00 -1.71339869e+00
-2.65801787e-01 -8.10216274e-03 -3.04604582e-02 1.85394987e-01
-4.65155214e-01 -3.09773445e-01 -4.95415777e-01 -1.15484190e+00
-2.24070385e-01 -1.11355484e+00 1.92499124e-02 3.66556048e-01
-4.64774370e-01 3.19558263e-01 9.90699470e-01 -4.74842519e-01
9.14655983e-01 -1.91729462e+00 3.70975763e-01 9.44895893e-02
2.98152119e-01 7.26785883e-02 -2.59249568e-01 3.53866041e-01
2.25215964e-02 4.25651908e-01 -3.28801543e-01 -6.64206982e-01
-6.34460943e-04 3.59029025e-01 -4.60756809e-01 2.50396430e-01
3.36231738e-01 1.11626804e+00 -7.88075507e-01 -4.84913409e-01
1.10907666e-01 7.38668740e-01 -5.81962347e-01 7.83869252e-02
-2.35743806e-01 7.45361686e-01 -2.05558524e-01 7.28426516e-01
6.34534299e-01 -1.47734970e-01 -4.09747437e-02 -3.52851152e-01
1.82920322e-01 -4.20299023e-01 -1.11502051e+00 1.70467997e+00
-5.49528778e-01 8.99655521e-01 2.25979788e-03 -2.98041046e-01
1.04073071e+00 2.00834438e-01 4.84791398e-01 -5.51585674e-01
2.03328460e-01 1.15060890e-02 -3.78632218e-01 -1.37727901e-01
6.13957644e-01 -5.75568020e-01 -1.44666284e-01 6.61417842e-01
1.58118717e-02 -6.31002188e-01 1.71361137e-02 1.17205620e-01
5.40782988e-01 4.59670037e-01 3.18182968e-02 -2.55719036e-01
1.37296021e-01 -1.03501037e-01 4.01798695e-01 3.80363911e-01
1.70388684e-01 9.31956053e-01 6.09612703e-01 -5.01512825e-01
-1.51729894e+00 -1.22145677e+00 -3.25461850e-02 6.03776336e-01
-4.89113070e-02 -2.07604393e-01 -1.17848933e+00 -4.82187152e-01
-5.34298224e-03 7.52069712e-01 -9.04027939e-01 -3.16029787e-02
-5.24503887e-01 -5.06913304e-01 7.58788586e-01 3.11265260e-01
7.00364172e-01 -1.12647629e+00 -4.70571637e-01 -3.28331888e-02
-3.41409668e-02 -9.36296523e-01 -7.36524165e-01 -5.08480310e-01
-7.84920752e-01 -6.55032814e-01 -1.09129894e+00 -5.31001031e-01
7.76691735e-01 7.85592571e-02 1.17654145e+00 -2.16050044e-01
-2.14513302e-01 3.69862467e-01 -1.29510313e-01 -3.32611173e-01
-7.14401782e-01 4.49539647e-02 1.86932549e-01 3.65523547e-01
-2.98793882e-01 -8.86548281e-01 -6.11423016e-01 3.31301957e-01
-1.25680566e+00 4.77522373e-01 6.84462368e-01 9.24548864e-01
6.92098379e-01 -1.31141633e-01 4.48509306e-01 -1.04756629e+00
7.62699604e-01 -4.79081757e-02 -4.45250303e-01 5.39676286e-03
-4.96415883e-01 1.86269805e-01 7.50840068e-01 -8.00109327e-01
-1.12098122e+00 -7.03146085e-02 -1.43530875e-01 -8.06475580e-01
-1.15871608e-01 -2.87993159e-03 -1.99175760e-01 -1.18042432e-01
7.53117085e-01 2.60373890e-01 3.34117532e-01 -3.33788961e-01
7.11264849e-01 3.29155058e-01 5.82296669e-01 -6.64008558e-01
1.21658993e+00 8.01611483e-01 9.22030061e-02 -9.68175352e-01
-3.71344149e-01 4.99022543e-01 -7.17670381e-01 -2.69752204e-01
7.52336562e-01 -7.42346346e-01 -7.53431797e-01 3.61758739e-01
-8.96281004e-01 -5.32301486e-01 -7.32171834e-01 -1.83431983e-01
-6.86696887e-01 2.04344630e-01 -3.97392869e-01 -5.37835598e-01
-4.12321210e-01 -1.22868145e+00 1.25293660e+00 1.64519295e-01
-6.73501253e-01 -9.54771519e-01 1.14819378e-01 1.57343760e-01
4.77403969e-01 8.92510653e-01 9.91784692e-01 2.27314964e-01
-7.23455906e-01 -8.15010667e-02 -1.53363392e-01 1.54189155e-01
3.62315089e-01 2.21648514e-01 -8.73106539e-01 -3.60910356e-01
-4.29060906e-01 -1.93038970e-01 5.96196353e-01 2.25950465e-01
1.03450179e+00 -4.37000901e-01 -1.23736463e-01 9.19925630e-01
1.34318841e+00 6.16732612e-03 7.49784470e-01 4.83508296e-02
1.04592752e+00 4.81915206e-01 1.89509109e-01 3.07148606e-01
9.87067968e-02 7.44227588e-01 2.30374143e-01 -4.07383114e-01
-4.43247974e-01 -6.58303618e-01 3.14775586e-01 3.78789067e-01
-2.73394942e-01 -2.83141106e-01 -3.35260987e-01 4.49407011e-01
-1.38698959e+00 -1.01360822e+00 3.01476032e-01 2.12070537e+00
9.58866537e-01 -5.74740916e-02 2.99918890e-01 7.11612552e-02
4.53933954e-01 2.40270570e-01 -6.62557244e-01 -5.27279019e-01
-3.82132053e-01 3.29667419e-01 3.66335511e-01 4.53419358e-01
-7.72504926e-01 1.24620914e+00 6.41244745e+00 1.01099241e+00
-1.41123319e+00 -2.55032536e-02 9.07624006e-01 -1.80692837e-01
-9.02044475e-01 -1.56624600e-01 -3.71824741e-01 3.79733801e-01
4.30646747e-01 -2.58338246e-02 6.50601745e-01 7.09311128e-01
1.06756896e-01 1.64400816e-01 -9.14773405e-01 1.29103625e+00
1.07115149e-01 -1.69704235e+00 6.36417270e-01 1.85488999e-01
1.10950994e+00 -6.86848998e-01 6.07158124e-01 5.62388971e-02
3.44046623e-01 -1.34166050e+00 9.52946007e-01 6.40261710e-01
1.55586600e+00 -9.45466816e-01 1.37796447e-01 -1.26833664e-02
-9.06379163e-01 2.56710768e-01 -2.26244375e-01 2.51917630e-01
2.21988708e-01 2.94348478e-01 -5.79248667e-01 3.02061856e-01
5.08268237e-01 6.62640214e-01 -5.13662279e-01 3.72772604e-01
-2.19002306e-01 3.53328437e-01 -7.66871646e-02 -3.97856608e-02
1.32244229e-01 -3.88781279e-01 7.82498896e-01 9.81805086e-01
4.71578926e-01 -4.24471647e-02 -1.42754048e-01 1.35606742e+00
-3.38577658e-01 2.65025385e-02 -7.96147585e-01 -1.00791551e-01
2.97148347e-01 1.34386683e+00 -7.08235383e-01 -2.55385369e-01
8.58252943e-02 1.55892241e+00 6.46297038e-02 5.21026134e-01
-9.07376528e-01 -1.66752070e-01 8.90456080e-01 4.39841807e-01
1.98644772e-01 -3.38442415e-01 -5.65253079e-01 -1.13690805e+00
-3.47379088e-01 -1.13556302e+00 -2.06084803e-01 -9.63917971e-01
-1.15841186e+00 8.91505182e-01 -9.17561948e-02 -1.23616087e+00
-2.12806746e-01 -3.94141108e-01 -6.70548081e-01 9.36132371e-01
-1.04358172e+00 -1.59905112e+00 -4.50857282e-01 4.81892049e-01
5.56377649e-01 -2.65468150e-01 9.46342349e-01 1.36520505e-01
-3.82494181e-01 8.64443004e-01 3.22953351e-02 -1.07978601e-02
9.86074626e-01 -1.28982484e+00 5.77987015e-01 5.66694021e-01
3.58785391e-01 4.94667053e-01 9.22631681e-01 -6.44749045e-01
-1.51178086e+00 -1.08175790e+00 2.58614272e-01 -6.12881660e-01
2.60049313e-01 -4.66111302e-01 -6.70677304e-01 6.39891386e-01
4.93232191e-01 -3.35648626e-01 5.46445191e-01 -1.94893107e-01
-2.60426253e-01 -1.92892656e-01 -1.28210068e+00 1.19583583e+00
1.07835662e+00 -3.84323180e-01 -6.55949861e-02 -7.05720931e-02
5.08180916e-01 -3.45101178e-01 -9.46408451e-01 1.34958610e-01
8.41165006e-01 -9.31473970e-01 9.79087949e-01 -2.59884089e-01
6.42983496e-01 -3.91321927e-01 7.20282197e-02 -1.53869605e+00
-4.36735541e-01 -1.13965189e+00 6.05008416e-02 1.34964514e+00
2.45044455e-01 -3.70750248e-01 9.39575732e-01 3.86249959e-01
3.53826910e-01 -7.54985154e-01 -6.17276073e-01 -5.61609507e-01
1.52991876e-01 -1.05053477e-01 8.96869898e-01 9.96664882e-01
-6.32095575e-01 3.77136320e-01 -8.52020979e-01 -3.37736338e-01
7.03886330e-01 2.32751042e-01 1.24171937e+00 -1.06871700e+00
-2.75335699e-01 -5.01938164e-01 -1.69512689e-01 -7.85975933e-01
1.06858298e-01 -6.92114055e-01 -4.22516495e-01 -1.30464244e+00
-9.80744977e-03 -4.56365705e-01 2.46208817e-01 3.67957830e-01
-2.72714831e-02 9.00522828e-01 3.70023757e-01 2.75824130e-01
-1.37787312e-02 7.13808239e-01 1.92440653e+00 -1.95216656e-01
-2.68718302e-01 -2.20115826e-01 -9.77340698e-01 6.07687891e-01
6.01742446e-01 -2.83774376e-01 -4.70921993e-01 -4.68019068e-01
2.17791483e-01 -4.57729325e-02 4.43451464e-01 -1.02894998e+00
-2.36557752e-01 -1.67835936e-01 8.97781193e-01 -2.57808626e-01
6.20224416e-01 -6.72813058e-01 6.70871079e-01 2.61254132e-01
-2.34222412e-01 1.15681402e-01 3.45199317e-01 5.37969410e-01
3.00184288e-03 2.29516536e-01 1.15998018e+00 -1.46377951e-01
-4.44390237e-01 5.47734857e-01 -3.06323141e-01 -1.62208527e-01
9.04751241e-01 -5.06021738e-01 1.08667552e-01 -8.27785373e-01
-4.89448965e-01 -2.52109021e-01 1.06387854e+00 4.75229472e-01
5.57867229e-01 -1.74153411e+00 -8.32267284e-01 4.49831247e-01
-5.72583489e-02 -3.15929025e-01 2.54078865e-01 2.69314349e-01
-7.51324713e-01 -1.25982547e-02 -7.46749759e-01 -5.88344097e-01
-1.13730752e+00 4.11511093e-01 1.25530586e-01 -5.19982949e-02
-6.95621252e-01 6.51619673e-01 2.76181847e-01 -2.19249427e-01
-1.91973805e-01 -1.39935479e-01 -2.36179661e-02 9.93537530e-02
3.65424067e-01 1.63230419e-01 -4.48616475e-01 -7.61494279e-01
1.49877459e-01 8.31175327e-01 1.18179925e-01 -4.18050796e-01
1.18922150e+00 -1.20113650e-02 8.29504877e-02 3.51781845e-01
1.07329214e+00 3.11847687e-01 -1.79292715e+00 2.73217652e-02
-5.26412606e-01 -6.58096313e-01 -5.57222553e-02 -5.71126640e-01
-1.07560480e+00 7.59272397e-01 5.60796678e-01 6.25261441e-02
1.17574656e+00 -2.77326018e-01 9.24109399e-01 1.31260259e-02
3.63038570e-01 -8.36800992e-01 3.89245510e-01 1.74699929e-02
1.19158959e+00 -1.02353859e+00 5.74015006e-02 -2.14023516e-01
-1.00652814e+00 9.97874022e-01 3.30322266e-01 -4.16319549e-01
2.98618376e-01 1.98972628e-01 1.94129780e-01 -3.35361138e-02
-3.29525441e-01 1.41363531e-01 4.85518187e-01 6.08186841e-01
2.37524956e-01 1.42916247e-01 2.33058438e-01 1.74375445e-01
-5.34531116e-01 -1.36119083e-01 5.20547748e-01 5.17356575e-01
5.17157279e-03 -1.37304783e+00 -5.68073928e-01 2.44607717e-01
-3.77303183e-01 9.65829045e-02 -4.63717371e-01 1.01143062e+00
1.26648292e-01 4.01564926e-01 1.53550297e-01 -3.57991606e-01
2.54025459e-01 -1.03392206e-01 7.78607070e-01 -3.75014544e-01
-3.85294110e-01 2.83827692e-01 -1.31648630e-01 -4.59210843e-01
-3.25463474e-01 -5.02719045e-01 -6.26162648e-01 -7.58075416e-01
1.27235293e-01 -1.50897697e-01 5.07892370e-01 5.54187119e-01
4.69695330e-01 6.78641677e-01 5.49407065e-01 -1.21169519e+00
-5.49549535e-02 -8.08429658e-01 -6.37184024e-01 6.78343713e-01
1.64410263e-01 -6.21333420e-01 3.60750072e-02 4.20591146e-01] | [11.888160705566406, -0.4252932667732239] |
e67ca740-c443-420f-8ca9-519d3e15151d | evaluating-the-performance-of-ann-prediction | 1612.02666 | null | http://arxiv.org/abs/1612.02666v1 | http://arxiv.org/pdf/1612.02666v1.pdf | Evaluating the Performance of ANN Prediction System at Shanghai Stock Market in the Period 21-Sep-2016 to 11-Oct-2016 | This research evaluates the performance of an Artificial Neural Network based
prediction system that was employed on the Shanghai Stock Exchange for the
period 21-Sep-2016 to 11-Oct-2016. It is a follow-up to a previous paper in
which the prices were predicted and published before September 21. Stock market
price prediction remains an important quest for investors and researchers. This
research used an Artificial Intelligence system, being an Artificial Neural
Network that is feedforward multi-layer perceptron with error backpropagation
for prediction, unlike other methods such as technical, fundamental or time
series analysis. While these alternative methods tend to guide on trends and
not the exact likely prices, neural networks on the other hand have the ability
to predict the real value prices, as was done on this research. Nonetheless,
determination of suitable network parameters remains a challenge in neural
network design, with this research settling on a configuration of 5:21:21:1
with 80% training data or 4-year of training data as a good enough model for
stock prediction, as already determined in a previous research by the author.
The comparative results indicate that neural network can predict typical stock
market prices with mean absolute percentage errors that are as low as 1.95%
over the ten prediction instances that was studied in this research. | ['Barack Wamkaya Wanjawa'] | 2016-12-05 | null | null | null | null | ['stock-prediction'] | ['time-series'] | [-7.69942701e-01 -8.02169219e-02 -2.89374709e-01 -5.99826537e-02
2.41418526e-01 -3.17367494e-01 4.62560385e-01 -6.58016056e-02
-4.66652989e-01 1.08821285e+00 1.46986237e-02 -8.14580977e-01
-2.99599737e-01 -1.03429842e+00 -2.98645258e-01 -4.39570516e-01
-2.10351184e-01 3.35013360e-01 1.30368859e-01 -6.54579103e-01
7.58467078e-01 5.71500957e-01 -1.56806934e+00 -6.26466647e-02
4.54000443e-01 1.40162289e+00 -8.81624594e-03 4.00752842e-01
-2.81845629e-01 1.19884050e+00 -6.15293026e-01 -4.97212201e-01
8.59328628e-01 -1.43952325e-01 -1.61939517e-01 -5.13811827e-01
-3.21600288e-01 -6.49628878e-01 -1.49137706e-01 8.06781173e-01
2.31144086e-01 -1.91907182e-01 5.37371695e-01 -1.06494427e+00
-9.52585459e-01 9.13587928e-01 -2.36156166e-01 5.57079911e-01
-2.46758178e-01 -1.56082259e-02 8.89009833e-01 -9.53701735e-01
7.99295604e-02 4.82879639e-01 8.48679125e-01 -2.31517442e-02
-7.58655608e-01 -1.05298471e+00 -9.39712301e-02 2.10232317e-01
-1.04896677e+00 1.27500594e-01 1.00827861e+00 -7.49391913e-01
1.03941190e+00 2.12992057e-01 1.20934701e+00 2.67699987e-01
5.75590134e-01 3.58660132e-01 1.37158906e+00 -4.48919147e-01
3.25550348e-01 6.48187339e-01 1.53374225e-01 -1.19931839e-01
6.69232726e-01 4.46313083e-01 -1.16670936e-01 1.77163258e-02
8.98196757e-01 7.65979961e-02 -1.82874147e-02 5.26788056e-01
-7.71185100e-01 1.03672183e+00 3.52052122e-01 8.16998243e-01
-1.05058420e+00 -3.97365898e-01 1.38387635e-01 6.75074637e-01
5.51499069e-01 6.70562029e-01 -1.00517964e+00 -2.12870166e-01
-1.37887454e+00 5.04054606e-01 1.13771880e+00 3.67993087e-01
2.12578550e-01 9.25541580e-01 3.02951962e-01 3.42033654e-01
3.57125819e-01 3.26200724e-01 1.06305802e+00 -3.89057010e-01
2.34828144e-01 7.66825259e-01 3.28561962e-01 -1.06851292e+00
-4.24519509e-01 -8.66995573e-01 -7.23071456e-01 7.92084157e-01
5.23932278e-01 -6.45539880e-01 -6.64128602e-01 9.40416038e-01
-3.00814182e-01 1.09467171e-01 3.29123318e-01 6.06099725e-01
3.64808857e-01 1.01981056e+00 -8.74432251e-02 -4.17775065e-01
9.57677662e-01 -7.61349082e-01 -7.93419898e-01 1.30502403e-01
1.72810644e-01 -8.72128963e-01 2.97820121e-01 4.91992831e-01
-1.13985431e+00 -5.19602895e-01 -1.12406862e+00 7.64226377e-01
-9.21839893e-01 9.94456559e-02 4.36293095e-01 7.65016675e-01
-9.67092276e-01 9.93693709e-01 -4.62891042e-01 2.28827596e-01
-2.58188725e-01 3.74508888e-01 2.27803558e-01 1.03843725e+00
-1.60105550e+00 1.56576240e+00 1.00518227e+00 4.36480075e-01
2.24562749e-01 -7.96769261e-01 -1.68941468e-01 9.75523666e-02
-3.86808008e-01 -1.11685522e-01 1.30056834e+00 -1.19166648e+00
-1.61959362e+00 2.24876657e-01 3.95584792e-01 -1.04522598e+00
6.50504649e-01 4.05377448e-02 -9.96310294e-01 -2.90387660e-01
-3.07300895e-01 1.33338451e-01 1.39646783e-01 -7.75631547e-01
-1.01554549e+00 1.89284310e-02 -1.64026812e-01 -8.88396874e-02
-2.10872576e-01 1.55414671e-01 4.71862912e-01 -1.07505631e+00
2.84379482e-01 -7.05331266e-01 -1.00437798e-01 -5.46567917e-01
1.36024624e-01 -1.65898964e-01 5.90927124e-01 -1.13201225e+00
1.47140539e+00 -1.39931154e+00 -8.71806204e-01 5.48479974e-01
-4.18498427e-01 4.94209141e-01 6.15892828e-01 7.16177166e-01
-6.45300210e-01 1.21666662e-01 -8.72258171e-02 2.99695939e-01
1.43174216e-01 -4.84862067e-02 -6.00305378e-01 1.92598179e-01
1.77777901e-01 7.39426434e-01 -2.65155107e-01 1.86185874e-02
2.24907190e-01 1.30704656e-01 6.18602559e-02 -7.53465248e-03
-1.09745015e-03 -1.75266370e-01 -1.78582475e-01 6.89940274e-01
7.77065098e-01 -1.16561875e-01 -2.62159199e-01 4.24093120e-02
-7.73786008e-01 1.20186962e-01 -1.30461228e+00 2.73548692e-01
-6.43235520e-02 8.25066626e-01 -4.53390837e-01 -1.07343793e+00
1.55069208e+00 7.94414222e-01 6.04178667e-01 -8.34064245e-01
1.86908662e-01 8.51700723e-01 2.90933818e-01 -3.26831847e-01
7.33247638e-01 -4.08666342e-01 4.87649798e-01 4.11838800e-01
-4.50081527e-01 3.06874633e-01 3.46335411e-01 -6.59939587e-01
2.80986637e-01 6.36822581e-02 1.86127365e-01 -3.67997944e-01
5.48456848e-01 1.55690566e-01 6.82929218e-01 2.47874245e-01
-1.40827000e-01 2.22142190e-01 3.33661348e-01 -8.60477507e-01
-1.29734719e+00 -5.96488535e-01 -5.77381432e-01 3.78689468e-01
-4.22502011e-01 3.14880431e-01 -4.43500429e-01 1.84649527e-01
1.27746284e-01 1.06307840e+00 -3.84837091e-01 3.44018251e-01
-2.25203320e-01 -9.95455861e-01 3.69153261e-01 3.70959371e-01
6.34201646e-01 -1.35870636e+00 -7.94911683e-01 6.77087963e-01
6.37741804e-01 -5.56533456e-01 4.08701509e-01 3.47190797e-01
-1.10724068e+00 -9.08436000e-01 -1.16282439e+00 -6.98180437e-01
4.03956741e-01 -4.98770177e-01 1.03164256e+00 -1.37746166e-02
4.31103796e-01 -3.15343231e-01 -1.82414293e-01 -1.18210971e+00
-2.17422128e-01 6.04175813e-02 -4.67103645e-02 -1.03138983e-02
7.45386243e-01 -6.84658229e-01 -5.49356878e-01 2.59696860e-02
-5.55771530e-01 -5.37521951e-02 8.21668386e-01 6.21959567e-01
1.46669030e-01 5.50145328e-01 1.10551655e+00 -4.69776362e-01
9.49939907e-01 -7.83092082e-01 -1.27356267e+00 1.39536172e-01
-1.45437038e+00 -1.20619252e-01 7.00441658e-01 -1.41448170e-01
-8.92078817e-01 -2.18684882e-01 -1.30798906e-01 -1.79369375e-01
1.41304061e-01 1.02194941e+00 7.64968634e-01 -4.99249920e-02
2.15017006e-01 6.45715535e-01 2.55675852e-01 -4.54651713e-01
-5.60397208e-01 6.42965674e-01 3.29313576e-01 9.41619426e-02
8.97745728e-01 -1.47581741e-01 -2.13643326e-03 -7.13679790e-01
-1.96643248e-01 -1.06977053e-01 -6.01303816e-01 -3.92632723e-01
4.73187476e-01 -9.27944005e-01 -7.14077592e-01 9.01911199e-01
-8.04395258e-01 -7.51072448e-03 -7.33646899e-02 8.91932428e-01
-9.75716934e-02 -2.67171979e-01 -9.13373470e-01 -1.43592536e+00
-5.83156645e-01 -8.01865339e-01 -1.63216352e-01 4.70729858e-01
-3.17074269e-01 -1.31072712e+00 9.74246487e-02 -1.16579659e-01
9.97922659e-01 3.89239460e-01 5.67680120e-01 -1.06659770e+00
-4.30381805e-01 -5.89344621e-01 -1.82957444e-02 6.38420999e-01
9.41642225e-02 3.47975165e-01 -6.42926455e-01 3.55462730e-02
2.39280149e-01 1.80160999e-01 3.77813756e-01 5.98994255e-01
2.81745493e-01 -6.07227743e-01 3.50053877e-01 1.09759413e-01
1.87792051e+00 1.12393045e+00 6.79853559e-01 1.24358320e+00
-1.45511642e-01 4.85578656e-01 4.51081634e-01 6.42437339e-01
2.99467266e-01 -1.21318186e-02 2.61263102e-01 3.88437435e-02
6.01712942e-01 -7.80278817e-02 2.98659444e-01 9.11671817e-01
-4.38116312e-01 4.51565951e-01 -9.25016105e-01 3.27918470e-01
-1.55542433e+00 -1.26104307e+00 -1.55188709e-01 1.98763740e+00
6.93255544e-01 8.44419956e-01 2.80157685e-01 7.51323044e-01
5.32001674e-01 6.50178343e-02 -3.32014561e-01 -6.15004778e-01
-2.76668936e-01 2.19804034e-01 1.03995478e+00 1.92085743e-01
-9.80070353e-01 2.74845898e-01 6.72356749e+00 1.86945975e-01
-1.70633852e+00 -5.57488263e-01 7.12865055e-01 4.64624502e-02
-1.77663818e-01 -4.09849063e-02 -7.92344570e-01 1.14252651e+00
1.31987059e+00 -6.63655162e-01 1.69256732e-01 1.05388856e+00
5.74127197e-01 -7.77761042e-02 -3.30615491e-01 6.84032381e-01
-4.29006845e-01 -1.62307942e+00 4.31333967e-02 1.28257036e-01
8.69814157e-01 -1.17934942e-01 7.25783855e-02 3.35601389e-01
-1.45360216e-01 -7.63713479e-01 1.06622863e+00 1.11144471e+00
-2.23973230e-01 -9.25431252e-01 1.44017541e+00 4.88438606e-01
-9.63176489e-01 -3.90731752e-01 -3.56061429e-01 -8.75249386e-01
1.11341923e-01 5.94491661e-01 -6.95810854e-01 5.03368080e-01
7.10791051e-01 4.31915641e-01 -2.77064830e-01 1.31642497e+00
2.79654730e-02 8.34018648e-01 -3.88674527e-01 -5.88470638e-01
4.66744930e-01 -5.76631010e-01 1.84890062e-01 7.98451006e-01
7.28437543e-01 2.31025711e-01 -3.88731837e-01 7.66713560e-01
4.49552715e-01 2.48163328e-01 -4.31026936e-01 3.65305245e-02
5.59144855e-01 6.75021648e-01 -5.83995640e-01 -2.73600936e-01
-6.50876760e-01 5.25175519e-02 -3.84611547e-01 3.47485125e-01
-4.87915009e-01 -5.27966440e-01 2.65619606e-01 4.21908885e-01
3.58646303e-01 -5.83816059e-02 -8.70296419e-01 -7.70788848e-01
1.94698334e-01 -4.32352573e-01 3.54447328e-02 -7.67035186e-01
-1.26734531e+00 6.21553183e-01 6.08221479e-02 -1.57895589e+00
-6.82613075e-01 -1.08478165e+00 -9.87332940e-01 1.35329235e+00
-1.63009715e+00 -6.54372990e-01 4.47835028e-01 9.71621126e-02
3.04019958e-01 -9.27785337e-01 6.96450412e-01 2.47052982e-01
-6.10622227e-01 1.57496199e-01 7.10048616e-01 4.35200572e-01
-4.72074114e-02 -1.31207073e+00 2.52357334e-01 9.19383526e-01
-3.51352006e-01 3.26786608e-01 8.49585950e-01 -8.06998670e-01
-7.52165198e-01 -5.46705365e-01 1.28567672e+00 2.52411276e-01
1.11747229e+00 5.22490442e-01 -8.94868255e-01 8.25886071e-01
6.18356764e-01 -5.21948040e-01 4.62543070e-01 -3.75822663e-01
3.92204612e-01 -3.13962251e-01 -1.11101425e+00 5.13612866e-01
-2.34182104e-01 -8.03778991e-02 -1.06198502e+00 -3.34080666e-01
6.46103919e-02 -2.25472301e-01 -1.33544099e+00 2.73628533e-01
8.07357848e-01 -1.30139732e+00 5.91403484e-01 -3.18929940e-01
3.90690982e-01 -3.08060735e-01 2.09765762e-01 -1.22816777e+00
-4.59468573e-01 -3.72926414e-01 -4.04819939e-03 1.14013529e+00
1.13223469e+00 -1.28500569e+00 1.14078712e+00 1.26042056e+00
1.05471969e-01 -1.00099289e+00 -8.70464444e-01 -7.48524964e-01
3.27482015e-01 -3.33770663e-01 9.50560331e-01 1.08853972e+00
-1.44832492e-01 -2.53271759e-01 -2.72607744e-01 -3.76183949e-02
4.84161437e-01 6.06974401e-02 3.15013468e-01 -1.44563711e+00
6.91432059e-02 -1.02008319e+00 -3.94204855e-01 -3.57940733e-01
-3.51920500e-02 -3.35457653e-01 -7.63934433e-01 -1.41927683e+00
-6.39903724e-01 -3.29026252e-01 -6.99042499e-01 3.32536519e-01
2.44213358e-01 -1.44121915e-01 3.70321512e-01 5.57306588e-01
8.04731786e-01 2.28984088e-01 9.58922446e-01 -1.91087984e-02
-3.24040830e-01 6.63163126e-01 -4.99214500e-01 7.13683605e-01
1.10814798e+00 -9.67363045e-02 -3.54330450e-01 1.70907930e-01
4.20279264e-01 1.80925459e-01 -2.36128699e-02 -1.18692660e+00
3.69948268e-01 -3.19473833e-01 8.80023122e-01 -1.14125586e+00
1.42299116e-01 -1.19398320e+00 8.09258401e-01 8.13123345e-01
-8.58527645e-02 6.82776511e-01 3.22062463e-01 -1.52470320e-01
-7.45415151e-01 -5.98535597e-01 3.64181995e-01 -2.74969339e-01
-1.06426692e+00 3.05778421e-02 -3.77239317e-01 -5.68276465e-01
1.42039835e+00 -8.86280835e-01 -2.43740939e-02 -4.40636635e-01
-6.05308592e-01 1.54256150e-01 5.13533726e-02 2.39790142e-01
3.31294239e-01 -1.36975312e+00 -8.11531901e-01 -1.10337511e-02
-4.03526068e-01 -6.29237950e-01 -9.92615893e-02 5.08542478e-01
-1.17732131e+00 8.39740098e-01 -7.66288698e-01 1.04249366e-01
-5.03342330e-01 3.68422478e-01 6.33347988e-01 -2.93260604e-01
-4.38984782e-01 2.80479759e-01 -8.03741932e-01 6.15116656e-02
6.60256892e-02 -5.33620298e-01 -8.37199271e-01 5.13343394e-01
5.21354318e-01 5.31691432e-01 2.12825239e-01 -5.58246017e-01
-6.49304315e-03 5.81841707e-01 1.49579108e-01 -1.52238414e-01
1.80969274e+00 1.86899841e-01 -1.02594405e-01 1.01222098e+00
9.35425282e-01 -2.91553736e-01 -1.01425517e+00 9.45613310e-02
6.09389544e-01 -1.21554382e-01 1.11805938e-01 -8.88652146e-01
-1.27081347e+00 9.40993279e-02 6.24999821e-01 9.46413636e-01
1.23127115e+00 -9.15640473e-01 7.44503140e-01 3.67832094e-01
6.25002161e-02 -1.59697187e+00 -7.64159858e-01 6.33276880e-01
1.05689085e+00 -1.06915390e+00 -5.06611206e-02 2.43719488e-01
-4.18034881e-01 1.60118902e+00 2.65158176e-01 -4.78466123e-01
1.40156543e+00 3.66405219e-01 2.69313484e-01 1.69628695e-01
-8.08450162e-01 1.70681223e-01 6.13178499e-02 6.38239235e-02
5.96680284e-01 1.88510329e-01 -7.92069435e-01 7.95221984e-01
-8.87210250e-01 6.70113385e-01 7.05617487e-01 1.13254941e+00
-4.79314446e-01 -8.71572495e-01 -7.22142041e-01 7.10377038e-01
-9.61696863e-01 -1.40752003e-01 -4.90709841e-02 1.26422334e+00
-1.39774084e-01 6.40908718e-01 5.56617916e-01 -4.35986757e-01
5.16986966e-01 2.76956886e-01 -3.47216606e-01 9.32883173e-02
-9.90450859e-01 -2.90246129e-01 4.36329283e-02 3.43021840e-01
-4.01710689e-01 -7.81609714e-01 -1.09465551e+00 -7.73199916e-01
-1.96323484e-01 5.40173829e-01 8.86664569e-01 9.40210640e-01
-1.98692203e-01 3.45105529e-01 9.93517876e-01 -7.55939782e-01
-9.39844787e-01 -1.26523328e+00 -1.17866373e+00 -8.96403119e-02
3.88580501e-01 -3.79150271e-01 -5.80478728e-01 1.39348498e-02] | [4.494966506958008, 4.243381977081299] |
bbee4978-3e3e-4865-b496-1a838b6507ba | dcf-asn-coarse-to-fine-real-time-visual | 2103.10607 | null | https://arxiv.org/abs/2103.10607v1 | https://arxiv.org/pdf/2103.10607v1.pdf | DCF-ASN: Coarse-to-fine Real-time Visual Tracking via Discriminative Correlation Filter and Attentional Siamese Network | Discriminative correlation filters (DCF) and siamese networks have achieved promising performance on visual tracking tasks thanks to their superior computational efficiency and reliable similarity metric learning, respectively. However, how to effectively take advantages of powerful deep networks, while maintaining the real-time response of DCF, remains a challenging problem. Embedding the cross-correlation operator as a separate layer into siamese networks is a popular choice to enhance the tracking accuracy. Being a key component of such a network, the correlation layer is updated online together with other parts of the network. Yet, when facing serious disturbance, fused trackers may still drift away from the target completely due to accumulated errors. To address these issues, we propose a coarse-to-fine tracking framework, which roughly infers the target state via an online-updating DCF module first and subsequently, finely locates the target through an offline-training asymmetric siamese network (ASN). Benefitting from the guidance of DCF and the learned channel weights obtained through exploiting the given ground-truth template, ASN refines feature representation and implements precise target localization. Systematic experiments on five popular tracking datasets demonstrate that the proposed DCF-ASN achieves the state-of-the-art performance while exhibiting good tracking efficiency. | ['Qiang Shen', 'Xiaoyue Yin', 'Ying Li', 'Xizhe Xue'] | 2021-03-19 | null | null | null | null | ['real-time-visual-tracking'] | ['computer-vision'] | [-3.24042380e-01 -5.56452155e-01 -2.05830008e-01 -1.33387923e-01
-6.09136164e-01 -7.17597663e-01 4.89647210e-01 -2.48030469e-01
-4.70464796e-01 6.29045665e-01 -1.93063542e-01 2.26904079e-01
-2.90908575e-01 -3.59366864e-01 -6.43720269e-01 -1.00402439e+00
-1.91351324e-01 2.62711346e-01 4.69550490e-01 1.26167610e-01
-6.80290684e-02 5.83344698e-01 -1.07573187e+00 -4.78364110e-01
9.36384439e-01 1.46509850e+00 1.61151677e-01 2.01950163e-01
1.69999927e-01 3.45797002e-01 -5.63346088e-01 -2.97947854e-01
5.25850475e-01 -1.08610699e-02 3.17158639e-01 -1.73120663e-01
5.22812068e-01 -2.56797284e-01 -6.95400536e-01 1.38568044e+00
4.32749957e-01 -5.07384315e-02 1.57788500e-01 -1.27851212e+00
-2.91306525e-01 3.49015206e-01 -6.20126605e-01 3.93204957e-01
-1.30748793e-01 4.64409471e-01 6.61152542e-01 -6.38184786e-01
3.44698101e-01 1.07897484e+00 9.26143229e-01 5.33867598e-01
-1.33462942e+00 -1.27863514e+00 3.37923855e-01 -8.58678967e-02
-1.45642018e+00 -4.14409101e-01 6.94309294e-01 -4.05668348e-01
2.33678311e-01 -2.29211420e-01 8.64141166e-01 1.06836915e+00
2.40705550e-01 5.38475156e-01 7.07985163e-01 4.59978193e-01
9.52552333e-02 -3.27367559e-02 -1.37489766e-01 6.81363165e-01
6.59679353e-01 7.73871064e-01 -5.87427497e-01 4.67877544e-04
9.40677941e-01 2.51998842e-01 -4.97341156e-01 -7.93940604e-01
-1.43630850e+00 4.85453337e-01 9.84233439e-01 1.90597028e-01
-4.99484956e-01 3.08372915e-01 2.69865066e-01 1.62861794e-01
7.99641088e-02 7.93133825e-02 -1.47113115e-01 -1.15624413e-01
-1.11461318e+00 1.75645709e-01 3.75593126e-01 8.82157743e-01
4.68079865e-01 3.38030934e-01 -5.40254653e-01 2.54912198e-01
6.09205544e-01 1.02070308e+00 1.56106338e-01 -6.90228403e-01
4.18563008e-01 6.28812373e-01 3.83469969e-01 -1.15476406e+00
-3.47057581e-01 -1.39411640e+00 -9.55162644e-01 3.31927508e-01
5.94821513e-01 -2.67787218e-01 -6.12723470e-01 2.04448509e+00
6.37874901e-01 5.42273462e-01 -1.55292720e-01 1.28756511e+00
1.20776206e-01 1.89096272e-01 2.20034048e-01 -2.00496525e-01
1.17179179e+00 -6.36339247e-01 -7.24354923e-01 -2.08575010e-01
1.73910320e-01 -7.13607669e-01 1.83902949e-01 1.67833209e-01
-6.38179362e-01 -8.23774576e-01 -1.22584248e+00 5.74668527e-01
-5.22061214e-02 6.30300164e-01 6.95122242e-01 5.21256149e-01
-8.06785047e-01 5.11811197e-01 -1.04698181e+00 -1.94961488e-01
6.47814214e-01 5.49876571e-01 -2.27182627e-01 2.05384150e-01
-1.11416137e+00 6.29224658e-01 3.20633113e-01 6.13796830e-01
-1.10623252e+00 -7.99719810e-01 -5.86650431e-01 6.80085644e-02
5.44235110e-01 -7.05436289e-01 8.59722555e-01 -6.15945101e-01
-1.56022310e+00 9.04743224e-02 5.00201732e-02 -7.56091535e-01
6.30280495e-01 -3.64293545e-01 -5.92218459e-01 -3.53045389e-02
7.52173960e-02 4.81108606e-01 1.18432403e+00 -8.78039062e-01
-9.25623715e-01 -6.21863842e-01 -1.74628705e-01 -5.36569543e-02
-3.10801327e-01 -3.40758264e-01 -6.50407553e-01 -7.18970060e-01
2.21026748e-01 -1.00323093e+00 -5.68853095e-02 6.68929040e-01
-1.87544838e-01 -1.72153607e-01 1.18853283e+00 -3.66309464e-01
1.23689950e+00 -2.29031873e+00 3.13026384e-02 2.86750436e-01
4.61154073e-01 6.32965446e-01 -7.20353648e-02 -1.95335731e-01
2.66308963e-01 -7.68945277e-01 2.12989241e-01 -1.46107927e-01
7.50819519e-02 -1.50105804e-01 -2.58472681e-01 8.24540257e-01
1.42387599e-01 9.63176191e-01 -9.83878314e-01 -3.73772651e-01
2.46223271e-01 7.63321817e-01 -2.93358505e-01 3.61633524e-02
-1.65886611e-01 7.77999520e-01 -8.40217888e-01 5.47180414e-01
7.81762481e-01 -3.59478056e-01 -5.05698211e-02 -6.42388165e-01
-4.34846103e-01 -3.04370016e-01 -1.23463523e+00 1.89934742e+00
-1.70067251e-01 4.87118959e-01 3.46472234e-01 -8.64363849e-01
1.00138712e+00 7.56838173e-02 5.80147982e-01 -7.96224236e-01
3.60670000e-01 3.12215477e-01 2.42854625e-01 7.82571435e-02
5.55639341e-02 1.64975464e-01 9.57320258e-02 1.41942762e-02
3.79003398e-02 6.27289295e-01 3.10651418e-02 1.47084400e-01
9.80918169e-01 3.84251356e-01 -7.11069023e-03 -1.90847307e-01
9.04876888e-01 -8.80164132e-02 9.45595741e-01 5.58681965e-01
-6.02305353e-01 -7.04512969e-02 -9.04351249e-02 -2.87594229e-01
-6.36553228e-01 -1.20146823e+00 -3.09313703e-02 6.62558794e-01
4.75287527e-01 -8.92394930e-02 -4.08671558e-01 -6.40451729e-01
3.69285017e-01 -8.35344344e-02 -6.22017920e-01 -3.23038787e-01
-5.27163684e-01 -3.43211889e-01 4.84249949e-01 6.83565617e-01
8.93582761e-01 -3.29914629e-01 -6.67330921e-01 4.58119214e-01
2.34946057e-01 -1.11152399e+00 -7.62562275e-01 4.73019592e-02
-8.12257230e-01 -1.09851205e+00 -7.97482908e-01 -4.06560481e-01
5.47919512e-01 6.44161582e-01 3.16632718e-01 -6.21039979e-03
-4.84237075e-02 1.76490843e-01 1.21962324e-01 -2.62462143e-02
1.54366195e-01 -1.21673249e-01 2.00068071e-01 3.72319251e-01
3.65373045e-01 -4.88469332e-01 -7.72333503e-01 4.62225795e-01
-4.65353429e-01 -3.72225016e-01 8.68399978e-01 8.28887165e-01
6.38186395e-01 1.20253548e-01 3.41458917e-01 -1.72474578e-01
1.44610509e-01 -1.46321699e-01 -1.50081515e+00 1.16890930e-01
-4.29605573e-01 2.59463072e-01 6.79971993e-01 -5.81076205e-01
-8.81926477e-01 3.35521996e-01 2.77004957e-01 -1.01408494e+00
3.71323705e-01 1.78865582e-01 -2.33587980e-01 -6.38336897e-01
3.07453841e-01 3.46193016e-01 2.59984702e-01 -3.22563469e-01
1.58100441e-01 2.47242823e-01 8.57120275e-01 -3.87315810e-01
1.61287284e+00 7.14913011e-01 1.66399300e-01 -3.56468886e-01
-1.02974999e+00 -5.23489833e-01 -4.14016455e-01 -5.34284890e-01
7.04941511e-01 -1.18959534e+00 -1.20435536e+00 4.20239329e-01
-9.55803335e-01 1.14961356e-01 -5.63562326e-02 9.16976392e-01
2.33240630e-02 1.99505702e-01 -2.63227850e-01 -8.25374067e-01
-3.54492784e-01 -9.17264283e-01 1.10189724e+00 7.34515727e-01
4.15033102e-01 -8.46074760e-01 4.28090570e-03 -1.04313187e-01
8.68252575e-01 2.19718754e-01 1.31407648e-01 -2.90664375e-01
-1.06667018e+00 -3.90786439e-01 -4.45419699e-01 1.14582248e-01
3.15820873e-02 -1.09649964e-01 -7.17683196e-01 -6.95985734e-01
-8.73441622e-02 3.76328733e-03 9.41674948e-01 5.34750164e-01
6.99152052e-01 4.83895764e-02 -7.43372262e-01 8.64207387e-01
1.40888715e+00 1.82270870e-01 1.21435761e-01 2.94143260e-01
4.60816264e-01 1.63309984e-02 8.34866643e-01 5.19078434e-01
2.62029380e-01 8.32628131e-01 5.77523768e-01 3.28492045e-01
-2.99673200e-01 -4.02741015e-01 5.10201454e-01 5.76650739e-01
8.20097998e-02 2.77242362e-01 -3.38493526e-01 2.43298545e-01
-1.90728486e+00 -1.05134022e+00 1.79977909e-01 2.57029700e+00
4.05761242e-01 4.50884789e-01 6.66590258e-02 -3.04256409e-01
8.95718992e-01 3.23889017e-01 -9.30025637e-01 8.18928719e-01
-6.22183420e-02 -2.54460484e-01 6.60311341e-01 6.25045672e-02
-1.25338328e+00 8.52984726e-01 4.41217661e+00 9.11601603e-01
-1.44479561e+00 1.23437397e-01 -1.08746484e-01 -1.71115875e-01
2.75873005e-01 -7.44701969e-03 -1.18590569e+00 6.65996492e-01
6.17975414e-01 -1.22343734e-01 5.08207142e-01 8.08188856e-01
1.33616760e-01 8.17194283e-02 -8.10732663e-01 1.04543507e+00
-2.01072752e-01 -1.25226605e+00 -4.06794757e-01 2.44703367e-02
4.82699364e-01 2.39881575e-01 2.97283471e-01 4.85339820e-01
1.20388605e-01 -3.71580929e-01 7.61269271e-01 6.45392179e-01
4.87724006e-01 -4.80385542e-01 5.88879347e-01 2.45458096e-01
-1.91029501e+00 -1.56493217e-01 -5.02764225e-01 2.88257271e-01
-3.64099555e-02 6.57660723e-01 -1.55184716e-01 6.62798882e-01
7.70078182e-01 1.09064519e+00 -5.29506445e-01 1.51896763e+00
-2.37884000e-01 2.72548497e-01 -5.36672413e-01 3.41081843e-02
3.43938470e-01 -1.41334414e-01 8.02918971e-01 7.25278318e-01
3.28562051e-01 -1.69599280e-01 3.97264123e-01 9.23964858e-01
9.26948860e-02 -4.40148741e-01 -2.85480499e-01 -1.64593041e-01
7.50965953e-01 1.52503967e+00 -5.71921647e-01 -1.62498653e-01
-3.40267032e-01 4.73117769e-01 2.68085927e-01 4.11461502e-01
-1.19260085e+00 -3.50162446e-01 9.26444054e-01 -2.25170434e-01
6.28289223e-01 -4.16572481e-01 1.90351039e-01 -1.18322897e+00
2.23545194e-01 -5.54147482e-01 1.91988215e-01 -2.85275549e-01
-1.20969069e+00 6.10253453e-01 -5.86705029e-01 -1.90973067e+00
2.40267664e-01 -3.87502432e-01 -4.30953354e-01 8.20400774e-01
-1.62947702e+00 -1.09659231e+00 -6.08928859e-01 7.51856565e-01
1.31722137e-01 -2.74668425e-01 2.17815429e-01 6.11308336e-01
-8.12150121e-01 6.71671212e-01 2.22749338e-01 3.12883377e-01
7.83984363e-01 -8.18965018e-01 -1.90957244e-02 1.02964950e+00
-5.79534704e-03 7.93352187e-01 4.73903298e-01 -8.39266479e-01
-1.80049765e+00 -1.27368641e+00 4.14588749e-01 -1.08633727e-01
1.03694248e+00 -2.38387451e-01 -7.95677423e-01 3.86080712e-01
-1.30985826e-01 5.76249897e-01 7.10969120e-02 -1.51322797e-01
-3.43337387e-01 -6.72246516e-01 -7.20806062e-01 3.23728740e-01
1.16410756e+00 -4.35832173e-01 -2.97822177e-01 9.47553441e-02
4.15605426e-01 -5.73890924e-01 -8.47749352e-01 3.59081089e-01
7.86993742e-01 -7.29823828e-01 9.43720460e-01 -2.10934177e-01
-5.27117193e-01 -1.01215279e+00 1.73573568e-01 -1.27167785e+00
-4.73095715e-01 -6.71063185e-01 -4.37468320e-01 1.08444107e+00
-4.58639115e-02 -8.53396297e-01 9.70479012e-01 1.57867938e-01
-1.89228412e-02 -3.62136781e-01 -1.15784109e+00 -1.20194650e+00
-5.76205611e-01 -2.22855911e-01 4.66167390e-01 5.63337743e-01
-4.78270262e-01 2.12266177e-01 -3.94488811e-01 6.76600993e-01
1.30425727e+00 3.33749741e-01 9.00156200e-01 -1.42769027e+00
-1.79265052e-01 -3.64413381e-01 -6.49936140e-01 -1.52879167e+00
1.40888482e-01 -7.09158897e-01 2.01935112e-01 -9.57186759e-01
-5.43945655e-03 -5.19973159e-01 -5.26650906e-01 1.74034759e-01
-1.84829041e-01 2.16373116e-01 3.05069208e-01 3.73442888e-01
-9.47167456e-01 9.13523912e-01 1.30147493e+00 -2.71603882e-01
1.31235465e-01 2.43395761e-01 -4.37037140e-01 4.11645412e-01
4.08304125e-01 -5.79129994e-01 -2.17021734e-01 -4.78041053e-01
-2.28030831e-01 1.48835227e-01 6.63307667e-01 -1.59026110e+00
9.65278387e-01 2.22346649e-01 7.58096516e-01 -7.14665473e-01
3.35425168e-01 -1.33291733e+00 4.26487386e-01 8.75418782e-01
1.36366384e-02 -1.98344976e-01 1.96591914e-01 1.04672098e+00
-4.27183270e-01 3.77749503e-01 9.39045191e-01 3.98772806e-01
-7.53314257e-01 7.32051432e-01 2.04209551e-01 -7.27059618e-02
1.07310414e+00 -2.31817946e-01 -4.62523758e-01 -4.99880575e-02
-4.44422901e-01 5.56921601e-01 2.63218164e-01 4.65381831e-01
3.10106784e-01 -1.57075167e+00 -2.66819239e-01 3.36644500e-01
8.21416359e-03 -3.31969619e-01 3.90351623e-01 1.37552822e+00
1.03005640e-01 7.76097953e-01 -2.23729640e-01 -9.94705260e-01
-8.90973330e-01 5.50646603e-01 5.45899153e-01 -1.68066874e-01
-5.73637903e-01 7.54541159e-01 2.27282926e-01 3.85509357e-02
6.05265796e-01 -2.76993215e-01 -4.43597212e-02 -6.06320016e-02
6.13653302e-01 1.39257342e-01 -2.26093933e-01 -5.87274194e-01
-6.04814947e-01 9.07160223e-01 3.21921930e-02 3.13621998e-01
1.02310705e+00 -1.75342619e-01 3.57795209e-01 -1.41158387e-01
8.33378792e-01 -1.92279309e-01 -2.02863836e+00 -6.79686189e-01
1.13363005e-01 -6.15143239e-01 2.75952369e-01 -6.77647829e-01
-1.65132666e+00 7.58054078e-01 1.03786230e+00 -2.03109935e-01
1.09938908e+00 -2.99437344e-01 7.64153898e-01 3.08188558e-01
6.21496439e-01 -9.38936412e-01 1.06005929e-01 3.58312815e-01
4.30947453e-01 -1.30338812e+00 -1.06476754e-01 -1.35893881e-01
-2.70938039e-01 1.03713739e+00 7.69850373e-01 -2.98638284e-01
6.90260530e-01 2.50978231e-01 7.42952898e-03 -1.37142405e-01
-5.18893719e-01 -3.45486820e-01 3.45040888e-01 5.20133018e-01
-7.26208314e-02 -1.12078339e-01 2.23312676e-02 4.66882586e-01
2.04778031e-01 1.55042887e-01 -4.20574605e-01 7.69295216e-01
-4.46419984e-01 -6.86944783e-01 -3.76977503e-01 2.63307571e-01
-1.37138784e-01 3.42415392e-01 2.62499712e-02 9.01685834e-01
1.03311002e-01 5.52180707e-01 -1.40058383e-01 -4.78001982e-01
4.04758573e-01 -4.72656339e-01 4.49140310e-01 -4.37818542e-02
-6.49269462e-01 3.84608567e-01 -3.26251239e-01 -9.88769889e-01
-5.53055763e-01 -6.50691509e-01 -1.10768080e+00 -1.79605752e-01
-5.53978443e-01 3.26687068e-01 6.28384471e-01 9.62670565e-01
4.90891308e-01 6.57491148e-01 6.11309111e-01 -8.29155803e-01
-9.20577228e-01 -7.70608246e-01 -5.68039536e-01 4.58762012e-02
4.51102734e-01 -1.24045873e+00 -2.67347574e-01 -4.18937743e-01] | [6.2835516929626465, -2.1301283836364746] |
5c68091b-ffc0-47e6-b10b-185f6cf7205a | explainable-data-driven-optimization-from | 2301.10074 | null | https://arxiv.org/abs/2301.10074v1 | https://arxiv.org/pdf/2301.10074v1.pdf | Explainable Data-Driven Optimization: From Context to Decision and Back Again | Data-driven optimization uses contextual information and machine learning algorithms to find solutions to decision problems with uncertain parameters. While a vast body of work is dedicated to interpreting machine learning models in the classification setting, explaining decision pipelines involving learning algorithms remains unaddressed. This lack of interpretability can block the adoption of data-driven solutions as practitioners may not understand or trust the recommended decisions. We bridge this gap by introducing a counterfactual explanation methodology tailored to explain solutions to data-driven problems. We introduce two classes of explanations and develop methods to find nearest explanations of random forest and nearest-neighbor predictors. We demonstrate our approach by explaining key problems in operations management such as inventory management and routing. | ['Thibaut Vidal', 'Axel Parmentier', 'Alexandre Forel'] | 2023-01-24 | null | null | null | null | ['counterfactual-explanation'] | ['miscellaneous'] | [ 5.85436344e-01 1.03776503e+00 -1.02804351e+00 -1.10085654e+00
-4.92957413e-01 -4.34254616e-01 4.06630278e-01 4.25309449e-01
1.16494395e-01 7.85164416e-01 6.49804831e-01 -1.38125575e+00
-1.02805519e+00 -5.35883248e-01 -7.25449204e-01 -2.17006460e-01
-6.31795600e-02 7.38997400e-01 -7.00216234e-01 -1.96054708e-02
7.54202843e-01 4.36347425e-01 -1.54805267e+00 5.64585686e-01
8.37742031e-01 7.65787721e-01 -7.98508748e-02 6.81506217e-01
-2.74289995e-01 9.50626493e-01 -3.28704774e-01 -6.62848532e-01
4.36319619e-01 -3.05627882e-01 -9.23794091e-01 2.39926308e-01
-1.63516283e-01 -4.58263338e-01 1.57089218e-01 4.18267339e-01
4.43841405e-02 -1.37862042e-01 7.61508942e-01 -1.85062778e+00
-9.29363668e-01 1.19667578e+00 -3.50771278e-01 9.89406854e-02
1.49192914e-01 3.21919411e-01 1.42703688e+00 -5.77976286e-01
3.77102613e-01 1.29161012e+00 5.22292316e-01 4.65654820e-01
-1.48122370e+00 -3.22510779e-01 6.83442116e-01 2.39367247e-01
-6.71709359e-01 -1.87453166e-01 5.52712977e-01 -5.93161225e-01
8.48840117e-01 4.64346081e-01 5.98331928e-01 8.90911341e-01
5.28452754e-01 8.07291687e-01 1.18322492e+00 -5.98331869e-01
5.68336785e-01 4.62851435e-01 3.45165908e-01 2.72928178e-01
7.59905040e-01 7.07265258e-01 -8.18772197e-01 -1.27374589e-01
4.46719378e-01 4.33227777e-01 1.17543256e-02 -5.95086336e-01
-1.14960110e+00 1.26773012e+00 3.65925580e-01 -1.93820104e-01
-6.55636787e-01 2.29754508e-01 3.01697459e-02 3.11106831e-01
4.03188735e-01 1.13090158e+00 -1.14936113e+00 2.42070127e-02
-6.42992496e-01 4.18178380e-01 1.03225446e+00 1.14852619e+00
6.95723057e-01 -9.20697898e-02 -8.89272764e-02 1.52428657e-01
6.44460618e-01 9.53899026e-02 2.34807525e-02 -1.18418896e+00
5.38410425e-01 5.91263771e-01 3.85308743e-01 -9.51038480e-01
-5.66737592e-01 -6.09991372e-01 -2.78681189e-01 2.25690976e-01
3.32141280e-01 -1.55552462e-01 -8.66845489e-01 1.09503603e+00
-1.95507690e-01 -5.16168416e-01 2.74499655e-01 1.06255925e+00
6.85776174e-02 2.49466538e-01 7.05801547e-02 -3.38556498e-01
8.76432896e-01 -8.73348713e-01 -9.01766777e-01 -7.70392299e-01
7.54236698e-01 -3.17164242e-01 1.00650287e+00 6.08943224e-01
-7.85000145e-01 -1.28440693e-01 -1.02340114e+00 1.91881001e-01
-2.45453700e-01 -4.02305454e-01 1.39427841e+00 4.50155824e-01
-3.69385123e-01 7.32818842e-01 -7.30209589e-01 -2.12118134e-01
7.01372206e-01 4.08768803e-01 1.38342783e-01 -1.57339051e-01
-8.02747905e-01 1.03562772e+00 1.83348641e-01 3.52873176e-01
-7.75008798e-01 -8.99300933e-01 -7.82685518e-01 9.62443277e-02
6.67825520e-01 -8.81324649e-01 1.51700675e+00 -9.10569012e-01
-9.34979022e-01 4.32498813e-01 -1.84655219e-01 -8.67774069e-01
5.43213785e-01 -3.86684477e-01 -4.08172905e-01 -6.51781440e-01
2.78689474e-01 4.40775603e-01 4.71724600e-01 -1.39780998e+00
-9.65943038e-01 -6.36909068e-01 1.32738873e-01 -3.97057869e-02
2.84656972e-01 -3.63029152e-01 4.64913934e-01 -3.98580879e-01
4.90422159e-01 -7.20210731e-01 -1.03561771e+00 -2.21091762e-01
-7.87822962e-01 2.41909981e-01 4.21218783e-01 -2.20091268e-01
1.27284467e+00 -1.46359277e+00 -2.49630928e-01 2.26974472e-01
1.36665270e-01 -8.14140081e-01 2.90242344e-01 5.10788202e-01
-2.09479421e-01 7.10062087e-01 -3.82319927e-01 2.27938034e-02
3.77459556e-01 6.83711112e-01 -8.32468629e-01 1.92454770e-01
4.74128246e-01 8.18718135e-01 -8.12150121e-01 -5.22860512e-02
3.52548063e-01 -2.89423436e-01 -7.85430729e-01 2.76850939e-01
-4.26133633e-01 2.91988969e-01 -5.32059371e-01 1.01372778e+00
4.49095160e-01 -2.67510444e-01 2.86380827e-01 2.54196048e-01
-2.43043333e-01 6.58270180e-01 -1.11551046e+00 1.40910053e+00
-4.84712005e-01 6.10081136e-01 -1.04632497e-01 -1.08563113e+00
8.34538519e-01 -2.01825723e-01 4.63767871e-02 -2.74398923e-01
-6.67014793e-02 3.11144263e-01 1.71428189e-01 -6.29979372e-01
2.77007669e-01 -6.05895638e-01 -1.67705759e-01 7.45606840e-01
-4.41013336e-01 -2.97935754e-01 -4.28436548e-01 -1.13576591e-01
1.05532086e+00 2.64601022e-01 7.86681831e-01 -3.48298877e-01
-2.01569423e-01 9.35957789e-01 7.72749662e-01 9.78845417e-01
1.13732055e-01 6.52189553e-01 7.11133003e-01 -1.12285304e+00
-7.60657549e-01 -7.05531836e-01 -2.57364690e-01 1.13690424e+00
-6.17065001e-03 -1.97758436e-01 -2.20793724e-01 -1.03022540e+00
6.85719907e-01 1.72607744e+00 -9.16295707e-01 -5.96171096e-02
-4.90505015e-03 -8.91368568e-01 -2.58002698e-01 7.18379080e-01
-2.36198530e-01 -5.82990170e-01 -8.44143808e-01 2.52935201e-01
2.84172267e-01 -5.32959342e-01 1.19387984e-01 9.88631904e-01
-1.25407887e+00 -1.38553596e+00 8.39452669e-02 6.53020069e-02
8.24937642e-01 2.43495062e-01 1.41592038e+00 -1.02067478e-01
-1.56321630e-01 1.82254270e-01 -1.39270827e-01 -1.09010029e+00
-5.53801298e-01 -1.77504532e-02 6.27694800e-02 -4.49058115e-01
8.59450996e-01 -2.50599772e-01 -6.04911983e-01 5.05493701e-01
-4.80401546e-01 1.66693762e-01 8.53612959e-01 9.31573689e-01
6.83889449e-01 -3.46237086e-02 4.36364472e-01 -1.43745267e+00
7.82700181e-01 -7.60400951e-01 -6.08725905e-01 2.97229201e-01
-1.68800139e+00 5.91549814e-01 2.60507941e-01 -7.43661076e-02
-1.00302100e+00 2.17877015e-01 5.79707563e-01 -1.02010369e-01
-3.83443713e-01 7.70147383e-01 -1.03345513e-01 6.74913704e-01
7.30152249e-01 -4.67433721e-01 -8.57731923e-02 -3.80870014e-01
6.33180797e-01 5.68491757e-01 3.20591837e-01 -3.66008729e-01
6.10851049e-01 3.71054471e-01 1.23568699e-01 1.12387776e-01
-1.19640267e+00 5.87060377e-02 -7.07116604e-01 1.48780383e-02
4.80034977e-01 -4.80201244e-01 -5.32134712e-01 -7.01775312e-01
-8.74318480e-01 -1.69692114e-01 -7.46343076e-01 7.28245914e-01
-9.91038859e-01 -5.76166272e-01 3.21565956e-01 -1.04363561e+00
2.30628863e-01 -1.12992918e+00 7.58717775e-01 -1.05098030e-02
-9.63458121e-01 -1.03233063e+00 -1.48935392e-01 7.27564335e-01
2.63103932e-01 3.67252171e-01 1.20638824e+00 -1.22303545e+00
-9.23668325e-01 -2.51249850e-01 1.30384058e-01 -2.67830849e-01
2.34203398e-01 2.78873239e-02 -9.66341078e-01 2.68541545e-01
9.17409360e-02 1.02130920e-02 4.66890186e-01 9.20366645e-01
1.40091944e+00 -8.26129675e-01 -5.30886292e-01 4.77749020e-01
1.11633027e+00 3.13937247e-01 8.11353251e-02 8.32700670e-01
1.16439164e-01 1.51089466e+00 1.24313009e+00 7.49598742e-01
6.17922485e-01 4.38672632e-01 8.93857300e-01 -7.48572499e-02
6.38118267e-01 -5.83141387e-01 -1.48920655e-01 -1.75996646e-01
2.32918382e-01 -2.63712406e-01 -1.32882130e+00 7.02238441e-01
-2.24333906e+00 -7.45544136e-01 -5.38653612e-01 2.06153226e+00
2.12861568e-01 3.00111055e-01 -6.90266564e-02 1.93334654e-01
4.65297818e-01 -1.27392411e-01 -1.00635684e+00 -1.01263511e+00
1.22799622e-02 -4.67898458e-01 9.46999252e-01 6.29343390e-01
-7.63451457e-01 4.65311974e-01 7.66551876e+00 -1.13173194e-01
-4.50963259e-01 -2.65666187e-01 1.21103358e+00 -3.23840827e-01
-9.03343976e-01 6.32682681e-01 -6.05915666e-01 3.97142023e-03
1.03897512e+00 -3.90082151e-01 3.60879093e-01 1.20679259e+00
6.49051785e-01 -8.97131935e-02 -1.84170270e+00 5.14571488e-01
-3.58926237e-01 -1.67987740e+00 1.45923898e-01 2.09116280e-01
7.72043824e-01 -4.16483253e-01 2.68663257e-01 1.15822507e-02
9.94012654e-01 -1.52879512e+00 7.80589402e-01 5.35799742e-01
2.22005785e-01 -8.22119474e-01 9.53887522e-01 3.01823109e-01
-4.01861191e-01 -9.32656884e-01 -2.44084120e-01 -8.34025979e-01
8.60256031e-02 6.83964729e-01 -1.25378573e+00 3.67234707e-01
6.19688332e-01 7.77604997e-01 -1.91323295e-01 5.84979355e-01
-4.21013355e-01 6.15049839e-01 1.81902915e-01 2.03171492e-01
2.35280991e-01 -8.61840472e-02 1.39364615e-01 8.37196767e-01
1.35384142e-01 3.02253962e-01 -2.06617624e-01 1.17865539e+00
4.84991342e-01 -2.84528434e-01 -8.70035470e-01 3.07303052e-02
4.21636522e-01 7.28303313e-01 -5.23678184e-01 1.57333612e-01
-1.78406507e-01 5.11225760e-01 -1.94206014e-01 3.55684906e-01
-4.48686123e-01 5.94488019e-03 9.72429872e-01 2.48427913e-01
-1.31028593e-01 -2.07417402e-02 -1.38008642e+00 -6.53257847e-01
-1.48748934e-01 -1.00043309e+00 5.97780704e-01 -7.16132820e-01
-1.27769995e+00 2.83325948e-02 1.98613167e-01 -1.02621114e+00
-6.91034794e-01 -6.33755147e-01 -5.55775642e-01 6.77916527e-01
-1.41834450e+00 -7.99394786e-01 -1.59304023e-01 -2.40117311e-01
7.48343706e-01 -1.20497130e-01 6.14590108e-01 -7.12632298e-01
-5.20488262e-01 9.72716361e-02 7.86208063e-02 -5.82808435e-01
5.68765461e-01 -1.53993285e+00 4.78826404e-01 5.73058426e-01
1.69084072e-02 9.04151618e-01 1.17257130e+00 -6.70899391e-01
-1.60894966e+00 -9.49463189e-01 1.03839302e+00 -8.31287980e-01
4.68078762e-01 -9.52371880e-02 -6.01130486e-01 1.02921712e+00
-4.84339558e-02 -3.02999765e-01 1.13927495e+00 5.60404956e-01
-2.76296921e-02 -3.25065762e-01 -1.31642067e+00 5.29079735e-01
1.03553510e+00 -1.03579260e-01 -7.66053855e-01 5.38723111e-01
7.79939592e-01 -1.25100553e-01 -7.06375837e-01 3.46521258e-01
6.24671280e-01 -1.09810555e+00 6.84000909e-01 -1.38079023e+00
7.83550739e-01 -1.27408102e-01 -4.77862060e-01 -1.36995900e+00
-1.55217722e-01 -8.77066493e-01 8.74836277e-03 1.10476184e+00
1.06990528e+00 -6.61403537e-01 1.01451504e+00 1.49135852e+00
1.00720348e-02 -9.71736491e-01 -6.46412969e-01 -3.63184512e-01
-1.16947293e-01 -9.46225762e-01 1.13899982e+00 9.96801674e-01
6.31127954e-02 4.99315606e-03 -2.39182875e-01 3.92389268e-01
6.37294233e-01 6.00182772e-01 8.74077260e-01 -1.24397898e+00
-7.58531094e-02 -2.84743816e-01 -1.37821585e-01 -8.09343696e-01
1.16593145e-01 -6.19029999e-01 5.54243848e-02 -1.83878958e+00
1.22744955e-01 -7.02372551e-01 -3.32631886e-01 5.78425944e-01
-1.60216428e-02 -5.83769381e-01 3.46206874e-02 1.40164614e-01
-1.55790478e-01 3.12334120e-01 8.34573448e-01 -7.09083080e-02
-3.54352415e-01 3.98348451e-01 -1.68761313e+00 9.75965679e-01
8.08723152e-01 -8.53643119e-01 -6.43528819e-01 -5.68128943e-01
4.52639550e-01 1.19099043e-01 4.03039694e-01 -2.64862210e-01
1.48266912e-01 -9.76841211e-01 3.90873998e-01 -6.91948533e-01
-3.23237270e-01 -1.15894556e+00 4.44122076e-01 7.20994234e-01
-1.10012317e+00 2.91380584e-01 8.81499276e-02 8.53047669e-01
1.75611451e-02 -2.08098143e-01 7.72157088e-02 4.02876321e-05
-5.61511219e-01 -1.24544770e-01 -4.89870131e-01 -3.13881516e-01
1.08777630e+00 -4.57197905e-01 -2.96533465e-01 -5.84020972e-01
-8.51798892e-01 6.78410649e-01 4.54576194e-01 5.79076290e-01
6.68764591e-01 -8.46189201e-01 -6.52710617e-01 1.23203263e-01
1.87825382e-01 1.44063264e-01 -3.26906979e-01 5.11548162e-01
-3.38441432e-01 9.94836986e-01 -1.26638571e-02 -3.92106205e-01
-6.32411480e-01 6.41720474e-01 3.43548417e-01 -1.18096136e-01
-1.20879963e-01 6.74201012e-01 1.48054689e-01 -6.95381284e-01
1.69449717e-01 -6.44054234e-01 1.05895646e-01 9.08687860e-02
2.89748102e-01 6.64973676e-01 -8.70662108e-02 1.56456023e-01
-5.90780556e-01 3.46323214e-02 -1.36578381e-01 3.20987180e-02
1.67807710e+00 -3.67554575e-01 3.06036711e-01 7.12366402e-01
4.08555269e-01 -3.65528911e-01 -1.50599825e+00 7.24008977e-02
6.43864095e-01 -9.01628911e-01 2.53837913e-01 -1.35166824e+00
-7.11593866e-01 9.32881176e-01 3.55136663e-01 2.29431942e-01
9.12941039e-01 8.54317173e-02 -1.06627233e-01 6.25263989e-01
2.30512723e-01 -8.48965466e-01 -3.84960353e-01 -2.07376018e-01
1.10420918e+00 -1.69133699e+00 3.74152690e-01 -2.34223530e-01
-1.00149632e+00 1.17146349e+00 5.42251170e-01 3.66572410e-01
5.25821805e-01 2.46653527e-01 2.82957733e-01 -3.60307366e-01
-1.32610261e+00 4.36352454e-02 7.86707923e-02 7.88253486e-01
4.30208713e-01 2.09759936e-01 -1.45002857e-01 1.05726862e+00
-4.69229937e-01 5.15834950e-02 6.47494435e-01 9.53791618e-01
-3.24901670e-01 -9.84361470e-01 -4.72131014e-01 8.76632512e-01
-4.68136042e-01 -1.60932988e-01 -7.13107884e-01 1.16052282e+00
-1.19010704e-02 1.45856535e+00 9.44830328e-02 -4.35128361e-01
6.09173357e-01 -2.38299975e-03 -2.47019470e-01 -8.12655866e-01
-4.36909407e-01 -2.85405248e-01 2.15559125e-01 -6.99563146e-01
-2.74588943e-01 -1.00229526e+00 -1.13044262e+00 -2.44555309e-01
-3.28877956e-01 2.05907270e-01 9.05579567e-01 1.02508104e+00
5.29609919e-01 7.19658017e-01 7.76366532e-01 -5.09215117e-01
-9.75150824e-01 -3.91095728e-01 -4.77792680e-01 -3.61347049e-02
4.85081047e-01 -6.22685492e-01 -5.07655203e-01 -5.55133969e-02] | [8.732169151306152, 5.7935471534729] |
b786ae7a-0d0d-4138-930c-b1063b3bb018 | multilingual-neural-rst-discourse-parsing | 2012.01704 | null | https://arxiv.org/abs/2012.01704v1 | https://arxiv.org/pdf/2012.01704v1.pdf | Multilingual Neural RST Discourse Parsing | Text discourse parsing plays an important role in understanding information flow and argumentative structure in natural language. Previous research under the Rhetorical Structure Theory (RST) has mostly focused on inducing and evaluating models from the English treebank. However, the parsing tasks for other languages such as German, Dutch, and Portuguese are still challenging due to the shortage of annotated data. In this work, we investigate two approaches to establish a neural, cross-lingual discourse parser via: (1) utilizing multilingual vector representations; and (2) adopting segment-level translation of the source content. Experiment results show that both methods are effective even with limited training data, and achieve state-of-the-art performance on cross-lingual, document-level discourse parsing on all sub-tasks. | ['Nancy F. Chen', 'Ke Shi', 'Zhengyuan Liu'] | 2020-12-03 | null | https://aclanthology.org/2020.coling-main.591 | https://aclanthology.org/2020.coling-main.591.pdf | coling-2020-8 | ['discourse-parsing'] | ['natural-language-processing'] | [ 1.42538041e-01 4.58455235e-01 -5.08791208e-01 -3.67492288e-01
-1.13287818e+00 -8.47927809e-01 8.70773137e-01 5.41453600e-01
-4.68515068e-01 9.40956235e-01 8.27590585e-01 -1.02719176e+00
4.83196199e-01 -5.88340700e-01 -6.92216456e-01 -2.22670808e-01
3.39158699e-02 3.51159453e-01 2.92505145e-01 -4.35552269e-01
3.35602582e-01 -6.86889589e-02 -8.66159201e-01 7.08588421e-01
9.82397616e-01 4.10538167e-01 2.60967731e-01 6.84152544e-01
-8.24718595e-01 1.45639277e+00 -8.81137788e-01 -6.31840646e-01
-3.59808236e-01 -9.11225796e-01 -1.46686149e+00 -2.35507235e-01
5.60020022e-02 -3.48199636e-01 3.39978701e-03 9.82962132e-01
2.33924285e-01 -2.16682479e-01 4.38095093e-01 -6.22349620e-01
-8.48697543e-01 1.13675213e+00 -5.32520294e-01 5.49888790e-01
4.74160552e-01 -3.45951557e-01 1.33417249e+00 -5.48172891e-01
9.28238094e-01 1.55668259e+00 2.91260242e-01 7.17877805e-01
-9.03997481e-01 -2.98055917e-01 4.44889992e-01 2.94373930e-01
-5.14738917e-01 -3.17254543e-01 9.97148931e-01 -4.66266304e-01
1.17583704e+00 -1.97815653e-02 2.56776482e-01 1.29013574e+00
1.42108724e-01 1.18007481e+00 1.14077902e+00 -8.40431869e-01
-6.90760911e-02 1.12612903e-01 5.50450683e-01 5.37870407e-01
-4.57305694e-04 -2.85982907e-01 -4.26274210e-01 8.31264257e-02
4.04879063e-01 -9.31921184e-01 -2.36126035e-01 2.75318772e-01
-1.08718896e+00 1.39183068e+00 2.07733303e-01 7.97840118e-01
-1.76477313e-01 -2.79422611e-01 1.06817675e+00 2.66979933e-01
7.80136764e-01 3.90089780e-01 -4.20571923e-01 -2.84065574e-01
-4.02150393e-01 1.35861725e-01 1.05599940e+00 6.39012396e-01
2.03682140e-01 7.77591392e-02 -1.17008500e-01 9.45121467e-01
3.65402669e-01 2.83482820e-01 6.70019507e-01 -7.62965620e-01
1.26688695e+00 6.01682186e-01 -1.79745451e-01 -7.72831500e-01
-4.35279936e-01 1.01551235e-01 -5.24515688e-01 -2.72073239e-01
5.59283555e-01 -4.78868037e-01 -2.84011513e-01 1.90167117e+00
3.71784836e-01 -4.33789402e-01 7.40532517e-01 7.27216542e-01
1.34494472e+00 1.04652739e+00 4.91808891e-01 -5.32120228e-01
1.68117905e+00 -1.15857923e+00 -1.06965268e+00 -4.27969247e-01
9.79267240e-01 -8.97328556e-01 1.06217933e+00 -2.11153924e-01
-1.28016984e+00 -4.96139139e-01 -1.07867253e+00 -5.09914100e-01
-3.20112586e-01 1.07268214e-01 3.81279707e-01 5.17501235e-01
-7.02877760e-01 1.87152214e-02 -8.80349636e-01 3.10690068e-02
3.07246089e-01 -3.12503994e-01 -1.86364844e-01 -3.24730650e-02
-1.42018890e+00 1.18233979e+00 6.18655920e-01 2.45973025e-03
-3.24146092e-01 -5.74733853e-01 -1.26556051e+00 -1.53231949e-01
2.43399650e-01 -2.12934464e-01 1.56127656e+00 -1.12682879e+00
-1.58750498e+00 1.17562830e+00 -3.69778693e-01 -5.32646477e-01
2.72038847e-01 -5.65230548e-01 -2.93682199e-02 3.88869941e-02
3.99701178e-01 2.87749171e-01 1.29076779e-01 -1.23797333e+00
-6.53698087e-01 -5.09289980e-01 3.36679876e-01 5.01197517e-01
-2.89360225e-01 6.49251699e-01 -2.07826033e-01 -4.15722847e-01
-2.10568100e-01 -5.81509352e-01 -3.63214053e-02 -6.34152830e-01
-4.22297210e-01 -8.85261416e-01 8.85907054e-01 -1.24874401e+00
1.33205426e+00 -1.71816945e+00 1.84605315e-01 -6.50744915e-01
-1.23147145e-01 5.17516553e-01 2.37813126e-02 5.77708364e-01
2.35624343e-01 3.06575775e-01 -2.19509825e-01 -1.96569070e-01
-1.86505690e-01 3.89891028e-01 -3.20713580e-01 7.51385838e-02
5.09025872e-01 1.05581105e+00 -8.70137632e-01 -7.19776690e-01
3.40536647e-02 3.52099895e-01 -2.77829766e-01 4.08400774e-01
-4.39758003e-01 6.97647393e-01 -6.91112995e-01 3.75094503e-01
1.99944019e-01 -5.10921068e-02 8.51164043e-01 9.96834859e-02
-4.17422175e-01 1.16753376e+00 -7.24893332e-01 1.63934731e+00
-5.86687386e-01 9.93726432e-01 2.07701966e-01 -1.33073783e+00
6.28967345e-01 4.78975475e-01 1.90722737e-02 -9.47937310e-01
3.74754399e-01 1.62942469e-01 2.88979024e-01 -7.89329886e-01
5.11461079e-01 -1.89051405e-01 -5.08342981e-01 6.21184647e-01
-1.15009852e-01 1.99914336e-01 4.76732671e-01 1.21589899e-01
7.33668625e-01 3.13185722e-01 7.27349997e-01 -3.35993260e-01
8.08302939e-01 3.03535283e-01 4.63534296e-01 2.50042647e-01
-2.06696376e-01 1.90155774e-01 8.53190720e-01 -2.56119877e-01
-7.69816637e-01 -7.86193252e-01 -2.57134736e-01 1.35659003e+00
-3.77168246e-02 -2.99981624e-01 -1.10351300e+00 -8.64589810e-01
-3.34596157e-01 1.02245343e+00 -4.14475560e-01 5.74197173e-01
-1.48965037e+00 -8.07022274e-01 7.56511807e-01 7.41559446e-01
6.27122819e-01 -1.32424760e+00 -7.28063405e-01 4.63363469e-01
-6.06951177e-01 -1.53654730e+00 -8.31702910e-03 -4.43714857e-02
-6.19552016e-01 -1.37380755e+00 -3.36735010e-01 -1.18607962e+00
3.03013623e-01 9.70721245e-02 1.18182266e+00 9.57793072e-02
2.16474414e-01 2.84201708e-02 -6.00025356e-01 -5.23530841e-01
-9.71481919e-01 1.60651997e-01 -5.65619588e-01 -5.47742486e-01
3.34353834e-01 7.51768872e-02 -2.44480167e-02 -2.43394002e-01
-6.15223289e-01 1.94979325e-01 2.94841856e-01 9.71151054e-01
2.46949479e-01 -5.22489369e-01 8.46249938e-01 -1.43436205e+00
1.25966644e+00 -5.72057962e-01 -3.77089083e-01 2.90119499e-01
-1.64860114e-01 5.90798929e-02 7.00231850e-01 -2.04296187e-01
-1.68133390e+00 -6.48817897e-01 -5.14786303e-01 5.35285771e-01
-3.66332978e-02 9.21329260e-01 -2.52625853e-01 6.90477550e-01
5.32002628e-01 -1.94207668e-01 -6.33924678e-02 -3.57292414e-01
6.42191172e-01 6.51042283e-01 4.50869203e-01 -9.75439906e-01
3.13464463e-01 -8.42748024e-03 -3.76400530e-01 -9.76061285e-01
-1.16700613e+00 -2.57451177e-01 -7.48183787e-01 5.31394640e-03
1.41272438e+00 -9.70605075e-01 -5.06989777e-01 1.89542964e-01
-1.67025483e+00 -3.93997580e-01 6.18737238e-03 4.72744226e-01
-2.30658397e-01 4.44986671e-01 -1.06405699e+00 -7.14149296e-01
-5.44161737e-01 -1.05706167e+00 9.29487765e-01 2.57533014e-01
-3.14113200e-01 -1.46945488e+00 3.62028241e-01 6.24849617e-01
1.35250747e-01 3.83904099e-01 1.40822351e+00 -8.52587998e-01
-1.96151540e-01 2.16443315e-01 -3.85559499e-01 4.11320359e-01
4.96901199e-02 -1.61067694e-01 -8.20433438e-01 -4.44331169e-02
1.39140204e-01 -8.22008371e-01 4.46919560e-01 3.33792388e-01
5.31448960e-01 -4.43191141e-01 -8.09508562e-02 1.63391512e-02
1.12166703e+00 2.13210076e-01 3.47157925e-01 5.25107443e-01
6.86062753e-01 1.09973645e+00 6.21788144e-01 -1.16940111e-01
9.47049499e-01 3.22273791e-01 5.74390553e-02 1.02161907e-03
-3.23881596e-01 -3.35205168e-01 5.08174539e-01 1.39070320e+00
7.81957619e-03 -3.52983654e-01 -1.06752694e+00 7.90825248e-01
-1.97064745e+00 -7.76278317e-01 -3.51886362e-01 1.57029974e+00
1.24252081e+00 2.08227932e-01 -9.87223759e-02 -1.49105355e-01
5.76936126e-01 6.31293178e-01 -2.44279787e-01 -8.06172311e-01
-2.61960745e-01 1.20382234e-01 -3.48099917e-02 7.18201041e-01
-1.19151843e+00 1.33036983e+00 5.73956776e+00 4.72269267e-01
-9.12733614e-01 4.24434900e-01 7.04329967e-01 5.00753880e-01
-3.89288366e-01 -3.86474431e-02 -9.88012493e-01 1.93634897e-01
1.23160243e+00 -2.87071228e-01 -4.40223590e-02 7.81585038e-01
4.54549259e-03 -1.03419259e-01 -9.37128365e-01 4.85556692e-01
2.79321611e-01 -1.56267285e+00 -5.99401630e-02 -2.51883000e-01
6.51909471e-01 5.47213741e-02 -5.49504399e-01 6.35606527e-01
4.82516795e-01 -9.41304922e-01 6.26013577e-01 -3.26820999e-01
5.18939197e-01 -5.87956429e-01 8.94129813e-01 6.26385152e-01
-1.18362606e+00 2.79076070e-01 -2.51703650e-01 -2.54620820e-01
5.57858050e-01 1.03511242e-02 -7.46703744e-01 6.79261386e-01
4.46583807e-01 4.76045638e-01 -3.86917174e-01 7.11621866e-02
-8.97506356e-01 1.00240302e+00 4.55491319e-02 -4.71040130e-01
4.90251184e-01 -1.35330051e-01 3.82127255e-01 1.53894901e+00
-3.25947404e-01 7.42451921e-02 4.00779277e-01 5.98383963e-01
-3.73683155e-01 5.10566056e-01 -6.23016119e-01 -2.12412924e-01
4.19811159e-01 1.01579261e+00 -5.76984286e-01 -2.95102000e-01
-6.98442459e-01 5.14486492e-01 9.04401004e-01 1.97675511e-01
-5.97830594e-01 2.99604191e-03 3.71433556e-01 -3.01111042e-01
8.55754688e-02 -5.07359266e-01 -4.44293410e-01 -9.25336838e-01
3.18934053e-01 -9.89416122e-01 5.53030491e-01 -3.29002440e-01
-1.27418077e+00 8.85644734e-01 2.35160172e-01 -7.23322451e-01
-7.58720398e-01 -8.71662080e-01 -7.94578910e-01 8.63818228e-01
-1.88596165e+00 -1.39246941e+00 2.30051830e-01 2.09548205e-01
1.13274300e+00 -1.64304465e-01 1.05223536e+00 1.03130594e-01
-6.26423776e-01 3.76853317e-01 -2.19002515e-02 6.67314291e-01
4.28934753e-01 -1.20060384e+00 3.43205839e-01 8.65655899e-01
1.66126102e-01 4.82004762e-01 4.36142921e-01 -5.44011533e-01
-1.30921745e+00 -7.32229054e-01 1.38202846e+00 -5.35257578e-01
9.90291178e-01 -2.74676889e-01 -1.20422399e+00 7.79429078e-01
9.46344018e-01 -5.07702112e-01 9.33249354e-01 4.92333710e-01
-2.96845853e-01 4.88762349e-01 -7.41878271e-01 5.49649477e-01
7.29364336e-01 -6.63450301e-01 -1.18573070e+00 3.40377420e-01
1.01726294e+00 -7.32395887e-01 -9.77778971e-01 2.02795848e-01
2.89471984e-01 -6.07156396e-01 7.16355085e-01 -1.12762642e+00
1.04453456e+00 1.59562409e-01 -1.90455154e-01 -1.08176315e+00
7.85195157e-02 -3.72391582e-01 6.53232494e-03 1.68631577e+00
6.41148746e-01 -4.35529351e-01 2.81863332e-01 3.20130348e-01
-2.66471148e-01 -7.39100754e-01 -1.02591920e+00 -4.29668993e-01
7.11177468e-01 -3.55235100e-01 1.67066738e-01 1.10544288e+00
2.78524190e-01 1.22446227e+00 3.15974932e-03 -9.95157436e-02
1.51412487e-01 3.21341008e-01 6.59500360e-01 -1.02499616e+00
-1.28383100e-01 -5.55947721e-01 1.60896063e-01 -1.11966372e+00
8.99504542e-01 -8.99629533e-01 1.49970025e-01 -1.80066729e+00
8.61472934e-02 -1.24381296e-01 2.02342644e-01 1.96251899e-01
-4.74031270e-01 -3.78593981e-01 2.39345506e-01 1.85973138e-01
-4.16468173e-01 5.81324637e-01 1.16196036e+00 -2.26519536e-02
-2.17525005e-01 -2.85536796e-01 -8.96129966e-01 9.13562238e-01
9.42944765e-01 -3.61017823e-01 -3.78989607e-01 -8.66492271e-01
-1.28613055e-01 4.02075887e-01 -8.72490108e-02 -2.18953475e-01
-4.80392911e-02 -2.33899951e-01 2.76931329e-03 -5.36996424e-01
1.14981636e-01 -9.47292671e-02 -7.94323623e-01 3.60108495e-01
-6.55215859e-01 4.16137874e-01 4.01457578e-01 2.92724967e-01
-5.64569414e-01 -5.86791873e-01 5.35765886e-01 -2.66018867e-01
-7.14877486e-01 -4.80888188e-01 -4.52865273e-01 8.16109061e-01
7.75918722e-01 2.31245264e-01 -1.03239560e+00 -1.41979247e-01
-1.22987375e-01 2.94418424e-01 -3.13170962e-02 6.91217721e-01
2.14616746e-01 -1.08065462e+00 -1.08258784e+00 -1.32801667e-01
-1.78381637e-01 1.48258820e-01 -7.98706487e-02 4.83219296e-01
-5.30339420e-01 6.91393316e-01 1.69845056e-02 -4.47425306e-01
-1.37104309e+00 2.31155008e-01 -6.53753430e-02 -9.29606497e-01
-4.89685148e-01 7.44013429e-01 1.81713447e-01 -4.93331134e-01
8.85218158e-02 -8.35920721e-02 -8.26458991e-01 2.71049708e-01
6.44724905e-01 1.63902901e-02 -7.48776421e-02 -1.01545787e+00
-2.83259809e-01 3.49524200e-01 -1.31280899e-01 -2.00241953e-01
1.14025342e+00 -1.51240990e-01 -2.52053171e-01 6.82460904e-01
1.29712141e+00 1.67120710e-01 -7.85000920e-01 -2.83203393e-01
3.90148193e-01 6.15545362e-02 -1.34729356e-01 -4.74866033e-01
-3.71228695e-01 1.17278194e+00 3.29128169e-02 4.46635485e-01
6.58634067e-01 2.96422273e-01 9.24587309e-01 4.14700091e-01
8.66502002e-02 -1.22495472e+00 -1.96033195e-02 1.00438428e+00
9.93432760e-01 -1.36872888e+00 -1.00518458e-01 -4.85473186e-01
-8.86000335e-01 1.19934607e+00 7.41551757e-01 -3.27704996e-02
4.73535717e-01 2.14172125e-01 4.35574889e-01 -2.63099134e-01
-7.57552445e-01 -1.57922894e-01 2.11732164e-01 4.53220308e-01
1.33571088e+00 1.04942791e-01 -8.94270122e-01 7.53283203e-01
-3.74136955e-01 -6.17054582e-01 3.30496043e-01 1.13196647e+00
-4.23997194e-01 -1.32820356e+00 -2.13703886e-01 -4.44357321e-02
-7.76052594e-01 -1.11015573e-01 -5.71748972e-01 1.15641916e+00
-3.55302662e-01 1.18590617e+00 1.23235002e-01 2.07036242e-01
4.05764610e-01 1.99250251e-01 3.37772191e-01 -6.49113715e-01
-6.84183896e-01 1.43829495e-01 9.86145794e-01 -2.73131520e-01
-8.95273507e-01 -7.46251464e-01 -1.57553267e+00 -4.56965193e-02
-1.15841851e-01 3.51582736e-01 6.72946036e-01 1.31520367e+00
-4.06067586e-03 7.86337793e-01 2.97311187e-01 -5.98595619e-01
-4.94739234e-01 -1.27929926e+00 1.14365600e-01 3.96799922e-01
3.49942641e-03 -3.97666872e-01 3.04282401e-02 2.71439373e-01] | [10.775768280029297, 9.364920616149902] |
f578cead-8673-42b4-8369-04438679774c | ucm-2-a-rule-based-approach-to-infer-the | null | null | https://aclanthology.org/S12-1038 | https://aclanthology.org/S12-1038.pdf | UCM-2: a Rule-Based Approach to Infer the Scope of Negation via Dependency Parsing | null | ["Pablo Gerv{\\'a}s", "Alberto D{\\'\\i}az", 'Miguel Ballesteros', 'Laura Plaza', 'Virginia Francisco', 'Jorge Carrillo de Albornoz'] | 2012-07-01 | null | null | null | semeval-2012-7 | ['negation-detection'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.355528831481934, 3.5328574180603027] |
e56a8905-eb04-490b-8f68-6be86b7deffa | using-object-information-for-spotting-text | null | null | http://openaccess.thecvf.com/content_ECCV_2018/html/Shitala_Prasad_Using_Object_Information_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Shitala_Prasad_Using_Object_Information_ECCV_2018_paper.pdf | Using Object Information for Spotting Text | Text spotting, also called text detection, is a challenging computer vision task because of cluttered backgrounds, diverse imaging environments, various text sizes and similarity between some objects and characters, e.g., tyre and âoâ. However, text spotting is a vital step in numerous AI and computer vision systems, such as autonomous robots and systems for visually impaired. Due to its potential applications and commercial values, researchers have proposed various deep architectures and methods for text spotting. These methods and architectures concentrate only on text in images, but neglect other information related to text. There exists a strong relationship between certain objects and the presence of text, such as signboards or the absence of text, such as trees. In this paper, a text spotting algorithm based on text and object dependency is proposed. The proposed algorithm consists of two sub-convolutional neural networks and three training stages. For this study, a new NTU-UTOI dataset containing over 22k non-synthetic images with 277k bounding boxes for text and 42 text-related object classes is established. According to our best knowledge, it is the second largest non-synthetic text image database. Experimental results on three benchmark datasets with clutter backgrounds, COCO-Text, MSRA-TD500 and SVT show that the proposed algorithm provides comparable performance to state-of-the-art text spotting methods. Experiments are also performed on our newly established dataset to investigate the effectiveness of object information for text spotting. The experimental results indicate that the object information contributes significantly on the performance gain. | ['Shitala Prasad', 'Adams Wai Kin Kong'] | 2018-09-01 | null | null | null | eccv-2018-9 | ['text-spotting'] | ['computer-vision'] | [ 6.04720533e-01 -6.48046315e-01 2.31136605e-01 -3.01816314e-01
2.43969820e-02 -1.38248891e-01 7.06682384e-01 -3.06005388e-01
-4.88700747e-01 5.29507518e-01 -1.04986958e-01 -3.42838794e-01
5.11156470e-02 -5.22203863e-01 -6.07210219e-01 -8.02228749e-01
6.40532494e-01 5.43061674e-01 6.46794260e-01 -2.52455086e-01
4.86557633e-01 3.21155280e-01 -1.52827704e+00 5.12543917e-01
8.89709055e-01 9.96684134e-01 6.08993948e-01 5.15519440e-01
-4.20592099e-01 7.62427449e-01 -6.97202146e-01 -3.68248492e-01
3.54519814e-01 -3.48675042e-01 -2.30353490e-01 4.23234254e-01
7.59961307e-01 -6.93547010e-01 -5.89802206e-01 1.02176189e+00
5.61031163e-01 -1.07030176e-01 7.55996704e-01 -1.14732742e+00
-7.23890901e-01 5.11601627e-01 -9.61246729e-01 2.24080399e-01
-1.62287846e-01 1.13283262e-01 6.81036532e-01 -1.06865680e+00
3.84721965e-01 1.13674700e+00 6.41313672e-01 2.81500280e-01
-6.69437587e-01 -8.87885571e-01 1.30645812e-01 3.81356478e-01
-1.25887430e+00 -2.81552523e-01 7.38635838e-01 -5.89789450e-01
5.88212848e-01 3.44440550e-01 3.88146788e-01 1.06913376e+00
3.99786055e-01 1.23907351e+00 9.61036325e-01 -6.86029553e-01
-5.67754768e-02 2.27850154e-01 1.74662888e-01 6.57457113e-01
5.73002517e-01 -2.62117416e-01 -3.98274630e-01 2.80814171e-01
6.33723199e-01 9.84032452e-02 -3.43363017e-01 -2.44503930e-01
-1.48356426e+00 6.44637644e-01 1.51001558e-01 2.30715051e-01
1.18007481e-01 -3.78497243e-02 5.31637788e-01 3.46950465e-03
2.06213042e-01 -1.81849405e-01 -2.18557745e-01 1.52204201e-01
-8.64085734e-01 8.60732570e-02 4.61481005e-01 1.10099316e+00
2.82683223e-01 2.84389585e-01 -2.41499349e-01 1.12849879e+00
3.43873948e-01 1.08648992e+00 8.05519700e-01 1.89129159e-01
9.51328337e-01 7.25097597e-01 1.08644404e-01 -1.11297369e+00
-3.50507647e-01 -9.52963829e-02 -9.81971741e-01 3.81374866e-01
4.38407868e-01 -1.55390888e-01 -1.31738174e+00 8.06903660e-01
1.90740600e-01 -1.33779496e-01 -1.06617026e-01 1.07904863e+00
1.09331715e+00 6.08017623e-01 -4.20710474e-01 1.10364735e-01
1.31818914e+00 -1.18541145e+00 -9.71300304e-01 -5.94639480e-01
4.56168085e-01 -1.27035141e+00 1.17591631e+00 5.28240561e-01
-3.98195714e-01 -4.47414130e-01 -1.11757720e+00 -1.29949123e-01
-5.93178689e-01 9.76579845e-01 3.39745909e-01 8.49822223e-01
-4.34423447e-01 8.02266300e-02 -4.80833203e-01 -8.18312466e-01
3.94379109e-01 1.38400733e-01 5.32442927e-02 -1.05628461e-01
-9.16175306e-01 7.13613927e-01 6.74563289e-01 4.47187632e-01
-5.19786060e-01 5.34049608e-02 -4.15991604e-01 -1.16528116e-01
6.13249660e-01 -1.95792437e-01 9.02825415e-01 -1.16183603e+00
-1.26827657e+00 6.80089056e-01 8.02165568e-02 -3.30518901e-01
1.06156790e+00 -3.59044671e-01 -5.76670647e-01 -9.51905549e-03
1.59276709e-01 3.18178505e-01 1.33214533e+00 -9.48559403e-01
-8.61255467e-01 -4.31977630e-01 -4.72656578e-01 4.12875026e-01
-4.19945419e-01 2.98564762e-01 -6.83753192e-01 -8.26518297e-01
7.78571218e-02 -9.04147267e-01 3.30384016e-01 4.51519102e-01
-7.94910908e-01 -9.20757204e-02 1.54347658e+00 -7.55453467e-01
1.02810776e+00 -2.00241232e+00 -3.26353490e-01 -1.60997257e-01
9.78426486e-02 5.59018254e-01 6.05613813e-02 2.72653550e-01
1.46582216e-01 -2.73174701e-05 -2.09526181e-01 4.25196998e-02
-6.68556169e-02 -6.41586408e-02 -2.91298747e-01 5.40259004e-01
9.15038958e-02 7.44773090e-01 -2.10026652e-01 -7.98941195e-01
7.23782778e-01 2.31136546e-01 3.10132802e-01 -2.40739107e-01
-3.51114810e-01 -1.18136592e-02 -6.43852890e-01 9.52368438e-01
8.72113824e-01 -9.29540470e-02 -1.84639975e-01 -1.74247816e-01
-7.48547465e-02 -4.77583915e-01 -1.33645463e+00 1.04736114e+00
1.56602472e-01 1.34558761e+00 8.10249597e-02 -9.03513491e-01
1.05752575e+00 -8.63638371e-02 1.29770085e-01 -7.94831157e-01
6.05666697e-01 2.00806990e-01 1.78370565e-01 -8.54482889e-01
8.02329838e-01 2.55107105e-01 4.76327032e-01 3.12945932e-01
-5.69282174e-01 -7.93954730e-03 4.37936097e-01 7.77000338e-02
7.32688785e-01 1.27311110e-01 -6.35656416e-02 -1.82928607e-01
4.71664131e-01 1.99191988e-01 1.45892620e-01 7.95664191e-01
-1.65736824e-01 7.59506345e-01 2.66877979e-01 -5.96726835e-01
-1.26794648e+00 -5.34704208e-01 -4.97841865e-01 1.07746685e+00
5.21893501e-01 -5.84064536e-02 -5.45891583e-01 -4.29310143e-01
-3.38366069e-02 5.84346950e-01 -8.16269457e-01 2.32232153e-01
-6.24313951e-01 -9.59064662e-01 6.65087938e-01 5.14058232e-01
1.15551412e+00 -1.22064376e+00 -6.99271858e-01 -1.23621888e-01
-1.65161252e-01 -1.39040542e+00 -4.87574250e-01 1.54581979e-01
-7.35922813e-01 -1.07507658e+00 -9.56245184e-01 -1.07104576e+00
6.22507334e-01 6.23694479e-01 5.03972709e-01 5.72888032e-02
-6.87366068e-01 -1.28501832e-01 -5.30958295e-01 -7.23265707e-01
-1.91610202e-01 -2.61119068e-01 -2.71467060e-01 4.20433789e-01
2.72417247e-01 2.85520762e-01 -4.60061789e-01 7.73231328e-01
-9.98841703e-01 4.84852612e-01 9.60856318e-01 1.01647294e+00
2.93957412e-01 4.33228642e-01 1.02108559e-02 -7.49728620e-01
4.60954696e-01 1.58793107e-02 -6.88768268e-01 4.76225048e-01
-3.35722119e-01 -2.40081236e-01 6.70620739e-01 -6.86571479e-01
-1.31552148e+00 2.24596635e-01 3.64278913e-01 -2.92431504e-01
-2.88452417e-01 2.47000262e-01 -2.14602858e-01 -1.50364995e-01
6.70376122e-01 7.05327868e-01 -6.26578182e-02 -2.50996262e-01
4.07952294e-02 1.14547729e+00 3.24136704e-01 -2.78540462e-01
7.55466044e-01 6.14204049e-01 -1.58368111e-01 -1.40947068e+00
-4.44859773e-01 -5.40377259e-01 -6.49361908e-01 -2.87679970e-01
8.55988920e-01 -5.35553515e-01 -3.40322793e-01 1.15627730e+00
-9.82015848e-01 -3.38991076e-01 4.27566677e-01 4.84210193e-01
-1.72158912e-01 7.80531645e-01 -2.99633920e-01 -8.33797097e-01
-5.27206302e-01 -1.10917044e+00 1.29329622e+00 2.98334092e-01
5.49988389e-01 -6.48390114e-01 -6.29083812e-01 6.47634208e-01
3.73660952e-01 1.84464920e-02 8.43544781e-01 -6.66907012e-01
-6.69684529e-01 -3.18084121e-01 -7.19425499e-01 2.01082602e-01
1.35263592e-01 1.25707790e-01 -8.98231804e-01 -1.94284290e-01
-2.52235144e-01 -2.99279302e-01 9.56179082e-01 3.68940055e-01
8.97475839e-01 5.43748327e-02 -4.81392950e-01 5.06884158e-01
1.32869661e+00 6.45570576e-01 6.48438632e-01 6.93770707e-01
9.75637257e-01 4.23139691e-01 7.90724695e-01 4.99202460e-01
-2.50098079e-01 5.46060860e-01 3.70395094e-01 -3.13946575e-01
-2.41390198e-01 2.34737962e-01 1.68951809e-01 4.25389767e-01
3.40485215e-01 -6.72498226e-01 -1.11116695e+00 3.28440428e-01
-1.90994239e+00 -6.86910987e-01 -7.80495882e-01 1.98124647e+00
3.97268057e-01 2.96088755e-01 -2.09863469e-01 3.73334199e-01
1.15088415e+00 3.88851650e-02 -9.33322012e-01 8.65999609e-02
-5.51145434e-01 -9.70349461e-02 7.46369481e-01 -1.58885971e-01
-1.41378081e+00 1.16789579e+00 5.01739597e+00 1.10072970e+00
-1.39151859e+00 -2.75007755e-01 5.49300730e-01 1.05376735e-01
6.19657993e-01 -4.81902540e-01 -9.34220016e-01 5.59152246e-01
1.40368611e-01 8.32776949e-02 2.13990614e-01 7.11333215e-01
2.02376783e-01 -3.17703068e-01 -7.93366194e-01 1.19379008e+00
5.42838812e-01 -9.73285317e-01 2.28543013e-01 -2.28306070e-01
6.33165061e-01 7.83782750e-02 2.43471954e-02 1.01577729e-01
8.80947337e-03 -9.82285082e-01 9.28429902e-01 1.61084250e-01
8.19299877e-01 -4.35258955e-01 9.33668852e-01 2.68344760e-01
-1.20951259e+00 -8.24674591e-02 -6.17598236e-01 2.13611498e-01
-2.33085588e-01 3.91509682e-01 -9.71676588e-01 3.42436433e-01
9.06992912e-01 8.12630475e-01 -1.04120028e+00 1.25415194e+00
-3.45576294e-02 6.20593429e-01 -4.69560802e-01 -6.30695939e-01
2.84266680e-01 -2.79800296e-01 4.02696788e-01 1.15794456e+00
1.96970627e-01 -1.21193007e-01 8.24786201e-02 8.75070989e-01
1.13534160e-01 5.59046447e-01 -6.32572472e-01 -1.22947328e-01
1.44304991e-01 1.16525590e+00 -1.33733082e+00 -5.51810861e-01
-6.03061795e-01 1.05727935e+00 -2.81827956e-01 4.52507615e-01
-9.39762771e-01 -9.23675120e-01 -6.95385784e-02 6.65604370e-03
5.97923040e-01 -1.90659076e-01 -6.69366717e-01 -1.21141601e+00
3.83030474e-01 -9.81664836e-01 1.21664524e-01 -1.25011337e+00
-8.57927382e-01 5.28619409e-01 -3.18371594e-01 -1.29884756e+00
5.48434317e-01 -1.10515034e+00 -6.21867836e-01 7.50369608e-01
-1.26887476e+00 -1.46975791e+00 -8.84266496e-01 6.54357374e-01
1.17419624e+00 -4.76982832e-01 9.94042754e-02 1.87344298e-01
-1.01144552e+00 4.80569303e-01 8.49488676e-01 6.08473063e-01
8.25469077e-01 -8.36633980e-01 4.83389050e-01 1.00834751e+00
-6.41209958e-03 1.35157466e-01 5.37898958e-01 -9.76123631e-01
-1.41470826e+00 -1.11290431e+00 2.18614534e-01 -2.81451672e-01
7.15891719e-01 -5.42212486e-01 -9.04916167e-01 6.33759141e-01
1.56719193e-01 -1.98881984e-01 -1.33741528e-01 -4.34516191e-01
-8.21567792e-03 -7.02297240e-02 -7.52980649e-01 7.80282438e-01
6.59277260e-01 -4.97313067e-02 -5.04348576e-01 5.57934880e-01
3.89874786e-01 -5.13047040e-01 -7.91739002e-02 2.53205806e-01
7.60614038e-01 -9.57144678e-01 6.87859595e-01 -1.14164710e-01
4.17184830e-01 -4.51506406e-01 -1.83215499e-01 -6.56224430e-01
4.68368270e-02 -1.21634223e-01 4.65662152e-01 1.15320432e+00
2.04199165e-01 -5.45026243e-01 9.92553055e-01 2.86619157e-01
-1.19356103e-01 -3.32767844e-01 -7.10914791e-01 -6.91683650e-01
-1.34898573e-01 -2.36587793e-01 1.82011172e-01 9.35193717e-01
-5.99563956e-01 4.69967782e-01 -6.78824604e-01 2.38859765e-02
5.33512592e-01 1.19930595e-01 1.00687337e+00 -1.34521353e+00
1.10008970e-01 -6.40573680e-01 -5.23860276e-01 -9.41589773e-01
-1.72626361e-01 -5.56472659e-01 1.83555603e-01 -1.59055841e+00
2.12946773e-01 -3.60548913e-01 1.30108550e-01 3.92278999e-01
-3.34233403e-01 1.95845515e-01 1.37090921e-01 3.22679192e-01
-4.98561889e-01 6.15669847e-01 1.41579282e+00 -6.58846617e-01
1.09984465e-02 -1.14808150e-01 -5.32927662e-02 7.37112105e-01
8.61035347e-01 -2.36129403e-01 -2.41550818e-01 -6.48866892e-01
-6.35438263e-02 -2.35745862e-01 2.24941477e-01 -1.06554270e+00
3.31283957e-01 -2.54456788e-01 7.40331292e-01 -1.20321918e+00
2.73429453e-01 -8.90561044e-01 -2.29418844e-01 4.54598993e-01
-2.03076258e-01 -3.05803418e-01 3.54792655e-01 7.25111842e-01
4.20451351e-02 -5.40679276e-01 8.76159072e-01 6.80923611e-02
-9.75382090e-01 -4.77082804e-02 -5.04217923e-01 -3.40345092e-02
1.05209374e+00 -5.87571919e-01 -6.65555894e-01 -7.92632252e-02
-1.21395610e-01 1.69552416e-01 2.97185659e-01 6.54254019e-01
7.99679101e-01 -9.29382205e-01 -9.36371446e-01 3.87843013e-01
3.97209048e-01 8.57576132e-02 3.24398503e-02 8.33039999e-01
-9.57335234e-01 5.97626150e-01 -3.89331013e-01 -8.33337426e-01
-1.69228482e+00 4.09340620e-01 3.12218070e-01 2.29014292e-01
-1.00655878e+00 4.08854663e-01 5.46991348e-01 -1.61663726e-01
5.08049667e-01 -5.96843779e-01 -1.31428242e-01 -1.53389737e-01
5.73880196e-01 3.82380575e-01 2.25910008e-01 -6.12945020e-01
-9.94406939e-02 9.54990566e-01 -4.12626147e-01 1.89872116e-01
7.85085261e-01 -9.17629153e-02 8.31325725e-02 2.93866545e-01
5.59202433e-01 -1.67375028e-01 -9.18570280e-01 -4.95104164e-01
-4.86918725e-03 -6.29129291e-01 5.98827787e-02 -1.10229611e+00
-1.01790178e+00 9.83382940e-01 9.36416388e-01 -9.72569063e-02
9.71189082e-01 -4.11180735e-01 6.13053560e-01 9.01889384e-01
7.13832974e-02 -1.38674510e+00 3.67251009e-01 4.70063865e-01
9.05173242e-01 -1.62329161e+00 1.85209036e-01 -3.76402318e-01
-7.18653917e-01 1.41616952e+00 1.00583339e+00 2.52640486e-01
1.19723670e-01 2.69153506e-01 2.39634335e-01 1.06211901e-02
-3.63559663e-01 -2.64549524e-01 1.27956420e-01 5.77312112e-01
2.91113317e-01 -5.46915550e-03 -2.65364468e-01 1.13212608e-01
-4.82069999e-02 -1.06505066e-01 7.62655616e-01 1.09205747e+00
-7.58261323e-01 -5.46628416e-01 -8.97726178e-01 9.98139679e-01
-4.08779323e-01 -3.16556722e-01 -7.69804895e-01 1.14425218e+00
-3.09404563e-02 9.18711841e-01 -1.75859239e-02 -3.13946962e-01
1.25547156e-01 -7.74210766e-02 1.08365960e-01 -1.57103047e-01
-2.83465534e-01 3.09805930e-01 8.45848843e-02 1.52325854e-01
-1.59226805e-01 -4.49989915e-01 -1.11742330e+00 -2.15437576e-01
-9.88360584e-01 -3.59178752e-01 9.98571396e-01 7.68872142e-01
-4.99712229e-02 6.29965663e-01 2.10304186e-01 -5.94938219e-01
-2.90375441e-01 -1.28155279e+00 -7.85046160e-01 3.87609333e-01
1.13501675e-01 -7.47318447e-01 -2.54485369e-01 4.17154849e-01] | [12.037247657775879, 2.2684683799743652] |
f730e5f6-f302-49ad-a49a-8c8e87d10e0d | automated-optical-multi-layer-design-via-deep | 2006.11940 | null | https://arxiv.org/abs/2006.11940v1 | https://arxiv.org/pdf/2006.11940v1.pdf | Automated Optical Multi-layer Design via Deep Reinforcement Learning | Optical multi-layer thin films are widely used in optical and energy applications requiring photonic designs. Engineers often design such structures based on their physical intuition. However, solely relying on human experts can be time-consuming and may lead to sub-optimal designs, especially when the design space is large. In this work, we frame the multi-layer optical design task as a sequence generation problem. A deep sequence generation network is proposed for efficiently generating optical layer sequences. We train the deep sequence generation network with proximal policy optimization to generate multi-layer structures with desired properties. The proposed method is applied to two energy applications. Our algorithm successfully discovered high-performance designs, outperforming structures designed by human experts in task 1, and a state-of-the-art memetic algorithm in task 2. | ['L. Jay Guo', 'Haozhu Wang', 'Zeyu Zheng', 'Chengang Ji'] | 2020-06-21 | null | null | null | null | ['physical-intuition'] | ['reasoning'] | [ 4.04768944e-01 -8.72428194e-02 -5.92530295e-02 1.87913835e-01
-5.51087737e-01 -4.49149430e-01 1.88564993e-02 -4.68815953e-01
-2.82794714e-01 1.41007578e+00 -9.12772417e-02 -4.97485250e-01
1.52598113e-01 -7.99902201e-01 -9.14926350e-01 -9.85822976e-01
4.20468897e-01 4.42947298e-01 -1.65465400e-01 -6.30044332e-03
7.52908528e-01 5.33189774e-01 -1.53659749e+00 5.63648522e-01
1.05799901e+00 1.04857028e+00 8.02673161e-01 6.96378410e-01
-4.82467473e-01 4.17110860e-01 -5.69354117e-01 -8.63792002e-02
3.49236399e-01 -8.09057057e-01 -8.12085509e-01 8.39840770e-02
1.22375920e-01 -1.18946590e-01 2.40695886e-02 1.02916157e+00
8.27988982e-01 -6.31744042e-02 7.02150822e-01 -5.57964563e-01
-1.00896764e+00 7.25513816e-01 -4.44014370e-01 -1.85562775e-01
2.96854794e-01 5.57009399e-01 1.01401615e+00 -6.12326264e-01
6.15222931e-01 9.71378684e-01 2.45006561e-01 1.27549744e+00
-1.47551441e+00 -8.60632598e-01 -4.74561937e-02 2.50636727e-01
-1.09784102e+00 -6.56170964e-01 8.99228275e-01 -4.79919434e-01
1.39278126e+00 -1.29660711e-01 9.24475491e-01 1.16055357e+00
2.85947889e-01 6.49714708e-01 1.09414232e+00 -5.11709571e-01
6.31980121e-01 -1.51349753e-02 -4.56226408e-01 9.03579712e-01
4.08466429e-01 4.26539332e-01 -7.17333674e-01 -1.83793113e-01
5.88982224e-01 -5.00378311e-01 -4.91980687e-02 5.31129017e-02
-8.71293724e-01 6.41937554e-01 4.57042158e-01 1.27223656e-01
-6.31234825e-01 4.90664810e-01 1.89971421e-02 1.05003938e-01
3.64496857e-01 1.34114861e+00 -3.87124456e-02 1.55568682e-03
-9.72867131e-01 4.07861859e-01 5.94254613e-01 8.71064544e-01
8.28772366e-01 2.69237846e-01 -4.29326236e-01 7.07787871e-01
1.78523570e-01 2.99522877e-01 1.50002256e-01 -8.76417994e-01
1.42872334e-04 4.46542561e-01 4.90414143e-01 -1.55703634e-01
-5.00970840e-01 -3.52693707e-01 -8.64765823e-01 2.50418514e-01
-1.76881284e-01 -6.42582059e-01 -1.00698948e+00 1.39243436e+00
1.91921711e-01 -5.17053762e-03 -1.05587088e-01 1.03324223e+00
4.94601637e-01 1.06984496e+00 -2.89118230e-01 -5.75072646e-01
1.00692630e+00 -1.17692971e+00 -6.33390665e-01 -1.91729888e-01
2.53437608e-01 -8.58640850e-01 9.23476219e-01 4.86094713e-01
-1.48877907e+00 -4.33274359e-01 -1.41911137e+00 1.43121570e-01
4.85087708e-02 2.60925740e-01 5.95695019e-01 6.44470274e-01
-1.17270970e+00 7.69127607e-01 -3.32570881e-01 -1.48124576e-01
5.67490458e-01 5.66911101e-01 4.90851492e-01 4.88707684e-02
-6.73784971e-01 5.47191024e-01 6.67948961e-01 2.09780455e-01
-1.10236192e+00 -9.61725891e-01 -1.44306421e-02 -1.28247231e-01
4.42744017e-01 -1.38025069e+00 1.21559405e+00 -7.61886895e-01
-2.37424421e+00 6.57443047e-01 -2.93960243e-01 -3.06824237e-01
3.82268727e-01 1.25699878e-01 -2.15466499e-01 -7.34566972e-02
-4.53290284e-01 6.75158799e-01 1.23700988e+00 -1.25995851e+00
-5.18099308e-01 3.28671038e-01 -1.85146928e-01 -3.02417576e-01
-3.62150341e-01 -9.45041794e-03 2.02865764e-01 -4.86261547e-01
-1.48015529e-01 -1.10258412e+00 -5.68033695e-01 -8.97004753e-02
-8.25673282e-01 -2.64244407e-01 4.10833269e-01 -2.26061746e-01
1.20044160e+00 -1.38206744e+00 4.97463882e-01 5.79643361e-02
2.07336590e-01 3.30582321e-01 -2.27821067e-01 5.66163123e-01
2.56485194e-01 3.83679360e-01 -9.62436050e-02 -8.89289156e-02
-8.25530514e-02 -2.95484483e-01 1.96023323e-02 1.95353910e-01
1.55221269e-01 1.01411116e+00 -8.40915024e-01 -8.64462703e-02
-1.15914144e-01 -6.55337349e-02 -8.95304859e-01 4.18761432e-01
-1.17595315e+00 5.21312892e-01 -6.56635106e-01 5.84465325e-01
4.81844336e-01 -7.80793548e-01 4.51736838e-01 -3.88809502e-01
-4.95619774e-01 3.51854324e-01 -5.25747716e-01 1.77455258e+00
-4.95500386e-01 6.03930473e-01 3.35146114e-02 -1.05489004e+00
1.03127730e+00 1.86604336e-01 5.86810768e-01 -8.95842195e-01
-6.79681599e-02 3.46099049e-01 2.66157299e-01 -5.51267982e-01
5.60960293e-01 -4.55977440e-01 1.03844777e-01 8.10578942e-01
-4.43526566e-01 -2.98433840e-01 2.05773041e-01 -2.80710131e-01
1.24208069e+00 7.68258199e-02 -1.42175838e-01 -5.99035919e-01
3.71654212e-01 -9.89441648e-02 6.43486381e-01 8.65474224e-01
3.01991135e-01 2.37018973e-01 4.86473083e-01 -6.80109382e-01
-1.89925933e+00 -8.49518597e-01 2.47380733e-01 5.71600497e-01
7.29289204e-02 -3.47782552e-01 -5.28295636e-01 -3.05304259e-01
-2.20450401e-01 6.88801408e-01 -2.27909967e-01 -3.14464867e-01
-6.11968338e-01 -8.24530005e-01 2.71718465e-02 7.24598765e-04
4.05481994e-01 -1.17110646e+00 -5.85695446e-01 6.70642197e-01
2.43727058e-01 -1.21918368e+00 -3.93020064e-01 -1.75638020e-01
-6.27447605e-01 -4.67193782e-01 -8.74567807e-01 -8.39568973e-01
8.93570781e-01 -6.80020228e-02 1.16149187e+00 2.17012092e-01
-9.03632820e-01 -3.09256256e-01 -1.33526787e-01 -3.36246967e-01
-8.69372725e-01 4.61039543e-01 2.11931154e-01 -8.69912058e-02
-2.69039065e-01 -6.71774328e-01 -9.77801681e-01 1.29744142e-01
-4.73495543e-01 9.13040340e-01 9.93140996e-01 8.62365901e-01
5.67349851e-01 1.06973574e-01 8.45841050e-01 -7.21472204e-01
7.80306041e-01 -1.74300015e-01 -1.03549612e+00 5.56095779e-01
-4.80729401e-01 7.33601749e-01 8.84755492e-01 -4.37174469e-01
-1.01830482e+00 1.38251543e-01 -3.95431519e-02 -3.08032840e-01
2.80815333e-01 2.02894688e-01 -9.88604128e-02 -4.34033960e-01
6.23046756e-01 3.05696219e-01 -2.73262262e-01 -3.87535006e-01
7.27804601e-02 3.15424979e-01 -1.75140575e-01 -1.12462926e+00
6.83409393e-01 1.41406981e-02 3.89333338e-01 -1.06622124e+00
-8.51446509e-01 2.99904257e-01 -3.09807867e-01 -3.72009784e-01
8.96108150e-01 -4.62989420e-01 -1.19691563e+00 4.86474812e-01
-1.47013950e+00 -5.24096668e-01 2.56843772e-02 -3.38198282e-02
-4.76491541e-01 -6.95743337e-02 -3.74541968e-01 -9.78708088e-01
-6.63491249e-01 -1.27672195e+00 8.35776329e-01 3.85439575e-01
4.35100216e-03 -6.23489976e-01 -4.19184342e-02 1.93890169e-01
4.92709726e-01 1.99218169e-01 1.50870347e+00 3.84765327e-01
-1.34595573e+00 4.30376291e-01 -3.22826713e-01 1.17925212e-01
3.15367520e-01 6.63132221e-02 -7.69683719e-01 -3.30375761e-01
-4.55221742e-01 -4.60047394e-01 9.30957198e-01 5.27107716e-01
1.54161906e+00 -5.63586533e-01 -4.19280916e-01 5.43746829e-01
1.53274393e+00 3.56124729e-01 5.72915256e-01 1.85386278e-02
7.79691696e-01 3.55014890e-01 -3.22784553e-03 8.35091591e-01
-1.57630414e-01 7.00051606e-01 1.21529557e-01 6.95535764e-02
-3.71960290e-02 -2.49040172e-01 3.08805764e-01 7.08892763e-01
1.18034542e-01 -5.85625768e-01 -7.58281291e-01 4.31460381e-01
-1.76768553e+00 -7.98424006e-01 2.40712836e-01 1.85943675e+00
1.04338527e+00 1.17332421e-01 1.85385749e-01 -4.95744616e-01
8.33143413e-01 -7.05734417e-02 -1.31072652e+00 -5.75191200e-01
-7.92843774e-02 4.82597828e-01 5.67507088e-01 1.17143013e-01
-4.08059210e-01 1.18142092e+00 6.39687920e+00 9.50643539e-01
-1.15970218e+00 -1.76030844e-01 6.44501030e-01 -7.74555862e-01
-8.56684327e-01 9.19784680e-02 -1.00639629e+00 7.69196868e-01
8.53324890e-01 -3.25131983e-01 9.64432359e-01 1.17262624e-01
4.04732317e-01 1.01844862e-01 -1.15496373e+00 1.15553606e+00
-5.28605640e-01 -2.33797479e+00 2.11672544e-01 4.41470921e-01
1.46400940e+00 -4.54968065e-01 1.06306836e-01 -1.66663691e-01
2.62693256e-01 -1.24015164e+00 7.29252636e-01 6.64776862e-01
6.65360570e-01 -9.52111065e-01 -4.73776609e-02 2.09923148e-01
-8.34687114e-01 -1.05695888e-01 -5.60056210e-01 -1.35330232e-02
4.55446631e-01 9.73607719e-01 -8.42965901e-01 2.61434853e-01
3.01734537e-01 3.39284599e-01 1.62600636e-01 1.08291113e+00
2.20908802e-02 3.82147014e-01 -2.00215414e-01 -9.89091516e-01
4.39773738e-01 -3.79943937e-01 5.38005531e-01 6.77749276e-01
7.04400837e-01 2.54085004e-01 2.25465536e-01 1.71152425e+00
-5.73391736e-01 -2.42540315e-01 -4.47543681e-01 -7.24347949e-01
5.21405041e-01 1.12996316e+00 -5.35787702e-01 -8.90839398e-02
5.45168761e-03 1.01184523e+00 3.56684297e-01 3.67354453e-01
-8.10084045e-01 -2.23808438e-01 1.04538727e+00 1.58277303e-01
4.65331078e-01 -4.91416156e-01 -4.64666545e-01 -6.91230297e-01
-1.20182157e-01 -6.32702887e-01 -3.14490378e-01 -7.04364657e-01
-1.41297054e+00 5.01500010e-01 -4.95776832e-01 -7.82317817e-01
-8.36277977e-02 -7.03895152e-01 -5.63210607e-01 8.02160442e-01
-1.60148871e+00 -6.10481620e-01 9.38817114e-03 -1.17013589e-01
4.46832001e-01 -3.92247528e-01 3.66263688e-01 2.48541683e-01
-9.40981865e-01 7.30170012e-01 1.40391514e-01 -5.90481579e-01
2.76724368e-01 -8.16938221e-01 6.64259613e-01 6.04029775e-01
-3.60720873e-01 3.25781822e-01 8.47460926e-01 -8.17606747e-01
-2.04066992e+00 -9.36257005e-01 2.76260078e-01 3.79095703e-01
5.27739942e-01 -4.87796217e-01 -6.44579709e-01 -3.52964550e-01
4.50318515e-01 -5.09670854e-01 5.18617749e-01 -3.40895265e-01
3.55598062e-01 2.02997833e-01 -1.14424574e+00 7.82860219e-01
1.68393826e+00 -3.99934083e-01 4.04870659e-01 7.32017636e-01
9.61236656e-01 -2.96230644e-01 -5.40131509e-01 3.29647392e-01
5.88633657e-01 -8.08187604e-01 9.03632343e-01 -6.98254585e-01
8.01462650e-01 -2.02386782e-01 1.71043634e-01 -1.21551108e+00
-4.14265394e-01 -1.49957585e+00 -1.55418962e-01 9.19146657e-01
8.56245041e-01 -3.73031288e-01 1.27010739e+00 3.83245140e-01
-2.85137087e-01 -8.89618874e-01 -6.58849597e-01 -8.21245372e-01
4.91180234e-02 -3.04827397e-03 7.71322370e-01 2.69729555e-01
-5.82393289e-01 4.74851787e-01 -5.97567737e-01 6.55553490e-02
7.87286460e-01 5.03621101e-01 4.12471771e-01 -9.83360648e-01
-6.58325613e-01 -8.16559434e-01 1.82911873e-01 -1.24373078e+00
6.06865227e-01 -6.98594809e-01 3.33434969e-01 -1.50415993e+00
1.11882865e-01 -7.16854930e-01 1.51789906e-02 -1.64267421e-02
4.14659306e-02 -6.56239241e-02 1.12215087e-01 -1.61106631e-01
-2.51612723e-01 1.07500958e+00 1.56296182e+00 -2.43094295e-01
-3.54710519e-01 4.65896800e-02 -7.45710135e-01 1.71763852e-01
9.78165030e-01 -3.41857433e-01 -2.75232166e-01 -5.57679355e-01
9.91152942e-01 -3.92503291e-03 6.24227151e-02 -1.11367166e+00
2.28980660e-01 -4.91999060e-01 2.57207960e-01 -4.72171605e-01
3.20557833e-01 -3.21329027e-01 2.40677029e-01 5.48182786e-01
-3.41410965e-01 -1.63311526e-01 3.56873363e-01 2.44722798e-01
5.71353555e-01 -5.84675431e-01 1.01640475e+00 -4.25785214e-01
-5.47738671e-01 4.54829752e-01 -6.00140691e-01 -6.21864293e-03
8.62203240e-01 -1.81973368e-01 -4.25817013e-01 -5.41225076e-03
-1.96084753e-01 9.02300104e-02 5.99158645e-01 1.02696851e-01
9.11476254e-01 -1.28993762e+00 -7.56992459e-01 1.95171177e-01
-1.80301532e-01 -1.65829271e-01 3.17385584e-01 1.15728691e-01
-7.14550674e-01 3.86737764e-01 -2.69599766e-01 -4.72257435e-01
-8.59918594e-01 5.32059968e-01 4.60023969e-01 -1.71920598e-01
-3.80491257e-01 1.14167714e+00 2.28321001e-01 5.95915094e-02
-2.13460118e-01 -2.76363432e-01 6.01279914e-01 -2.44470030e-01
3.57382387e-01 3.60101253e-01 -5.15101962e-02 2.38075897e-01
-1.05856597e-01 7.82334447e-01 -2.14092061e-01 -2.09513858e-01
1.41949534e+00 1.44093707e-01 -3.41667026e-01 1.33837020e-04
1.19642580e+00 -4.34383094e-01 -1.18121696e+00 -1.54486224e-01
1.19460776e-01 -1.01103589e-01 4.12112147e-01 -5.95488846e-01
-1.03272736e+00 7.38483548e-01 5.65270185e-01 -4.54229936e-02
1.15202451e+00 -2.30113417e-01 1.38166976e+00 7.82727897e-01
5.80070376e-01 -1.22515905e+00 4.81106192e-01 3.51908952e-01
7.06475854e-01 -9.18663204e-01 2.38797978e-01 -3.08913104e-02
-2.00497299e-01 1.15669274e+00 8.84062588e-01 2.92217620e-02
3.03759187e-01 3.01357269e-01 -4.96004879e-01 -1.00392982e-01
-1.18739736e+00 -1.33666202e-01 4.05535936e-01 3.04189712e-01
3.79090309e-01 1.16557933e-01 -9.76461023e-02 3.49247195e-02
-3.25127780e-01 3.43652368e-02 4.82723564e-01 1.10508657e+00
-6.33757949e-01 -1.53182065e+00 1.00932673e-01 5.04387498e-01
-1.73570275e-01 -3.19629759e-01 -5.56552768e-01 -1.20619558e-01
-6.77378327e-02 6.93327606e-01 3.31190154e-02 -5.95109224e-01
2.70777997e-02 -2.64866203e-01 9.98151183e-01 -5.71875095e-01
-4.44960117e-01 -9.05963033e-02 3.08941364e-01 -4.01529461e-01
-1.59781054e-01 -2.51568884e-01 -1.18225133e+00 -5.67505956e-01
-1.45107493e-01 -1.41334161e-02 8.36281478e-01 8.40024948e-01
7.68107951e-01 8.45957398e-01 9.80333030e-01 -9.68371749e-01
-1.92990854e-01 -4.64830786e-01 -2.61933327e-01 -3.43858063e-01
5.17339349e-01 -4.10039812e-01 3.67601782e-01 -2.85849631e-01] | [5.168126106262207, 5.335023403167725] |
ab22f59a-d40a-484d-9c51-ea9af3d64abe | scoop-self-supervised-correspondence-and | 2211.14020 | null | https://arxiv.org/abs/2211.14020v2 | https://arxiv.org/pdf/2211.14020v2.pdf | SCOOP: Self-Supervised Correspondence and Optimization-Based Scene Flow | Scene flow estimation is a long-standing problem in computer vision, where the goal is to find the 3D motion of a scene from its consecutive observations. Recently, there have been efforts to compute the scene flow from 3D point clouds. A common approach is to train a regression model that consumes source and target point clouds and outputs the per-point translation vector. An alternative is to learn point matches between the point clouds concurrently with regressing a refinement of the initial correspondence flow. In both cases, the learning task is very challenging since the flow regression is done in the free 3D space, and a typical solution is to resort to a large annotated synthetic dataset. We introduce SCOOP, a new method for scene flow estimation that can be learned on a small amount of data without employing ground-truth flow supervision. In contrast to previous work, we train a pure correspondence model focused on learning point feature representation and initialize the flow as the difference between a source point and its softly corresponding target point. Then, in the run-time phase, we directly optimize a flow refinement component with a self-supervised objective, which leads to a coherent and accurate flow field between the point clouds. Experiments on widespread datasets demonstrate the performance gains achieved by our method compared to existing leading techniques while using a fraction of the training data. Our code is publicly available at https://github.com/itailang/SCOOP. | ['Michael Rubinstein', 'Shai Avidan', 'Forrester Cole', 'Dror Aiger', 'Itai Lang'] | 2022-11-25 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Lang_SCOOP_Self-Supervised_Correspondence_and_Optimization-Based_Scene_Flow_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Lang_SCOOP_Self-Supervised_Correspondence_and_Optimization-Based_Scene_Flow_CVPR_2023_paper.pdf | cvpr-2023-1 | ['scene-flow-estimation'] | ['computer-vision'] | [ 1.26698622e-02 -3.52867275e-01 1.21756755e-02 -2.21437722e-01
-5.64474642e-01 -6.08923554e-01 5.95572233e-01 1.08935818e-01
-4.76994514e-01 3.64550918e-01 -1.90749377e-01 -2.48848274e-01
2.72075623e-01 -7.55866349e-01 -8.05666089e-01 -5.60664058e-01
9.29948986e-02 6.61120534e-01 5.29599965e-01 7.03140348e-03
5.43739557e-01 7.95940161e-01 -1.29696572e+00 -2.15599149e-01
9.68841851e-01 9.13277745e-01 2.84421712e-01 9.56847727e-01
-2.67474115e-01 8.00450861e-01 -2.91979581e-01 -2.29744539e-01
6.41659975e-01 -3.58677834e-01 -9.72030878e-01 2.61998534e-01
8.54523242e-01 -3.10312629e-01 -3.92274380e-01 8.77910495e-01
1.66247189e-01 3.50001007e-01 4.90956783e-01 -1.36159301e+00
8.67810771e-02 -2.31594086e-01 -7.00903416e-01 2.52282977e-01
3.95111799e-01 4.69338208e-01 8.05476129e-01 -8.51284921e-01
7.06713378e-01 9.94744420e-01 5.53463936e-01 2.89774835e-01
-1.39292526e+00 -4.18449432e-01 3.50019932e-02 2.16364861e-01
-1.28977776e+00 -3.16931754e-01 1.02128243e+00 -8.50429535e-01
8.32302928e-01 2.35722796e-03 8.46505404e-01 5.21484673e-01
-1.18873119e-01 5.38158715e-01 6.15184009e-01 -3.42177659e-01
2.84490675e-01 -6.78817332e-02 -7.78867304e-02 8.30271602e-01
6.79818764e-02 4.08811942e-02 -3.90047550e-01 -1.47358247e-03
9.30336654e-01 2.23112583e-01 -4.40971881e-01 -7.80562818e-01
-1.34733057e+00 8.38398635e-01 8.32690120e-01 -4.10180166e-02
-4.72305626e-01 2.21081793e-01 1.08078301e-01 1.61807805e-01
6.22718811e-01 2.45088309e-01 -3.96656781e-01 -2.15744898e-01
-1.08169818e+00 3.93197447e-01 9.23122704e-01 8.13101768e-01
1.35166395e+00 -2.53137261e-01 1.48245171e-01 2.64695287e-01
3.65712851e-01 5.39519608e-01 1.50041714e-01 -1.37693465e+00
6.89954460e-01 7.54250348e-01 1.52966067e-01 -1.21751428e+00
-9.58690941e-02 -1.13144688e-01 -6.97526395e-01 5.49064159e-01
8.49621654e-01 -4.82570715e-02 -7.76889265e-01 1.31953669e+00
7.73016691e-01 9.44678009e-01 -1.64434880e-01 1.10096407e+00
3.69237423e-01 8.04238319e-01 -3.62491876e-01 -4.26193923e-02
7.53245413e-01 -1.13518083e+00 -8.23214576e-02 -3.25777888e-01
6.97710156e-01 -7.88696945e-01 8.99490893e-01 1.36951581e-01
-1.01950586e+00 -4.83210981e-01 -7.39524424e-01 -2.40092859e-01
-1.46133006e-01 -2.40568608e-01 4.78597909e-01 1.57522887e-01
-7.81277895e-01 7.82865882e-01 -1.23002374e+00 -2.30038479e-01
5.81860602e-01 1.81159675e-01 -4.60587084e-01 -2.14991465e-01
-4.28928047e-01 7.18449295e-01 1.88130230e-01 2.06367463e-01
-6.42363667e-01 -1.04467773e+00 -9.59259272e-01 -1.71839267e-01
4.64455277e-01 -9.66189086e-01 1.04984224e+00 -7.81958938e-01
-1.57270551e+00 8.48622024e-01 -3.85507256e-01 -2.10575461e-01
8.87141943e-01 -4.07878906e-01 4.32654649e-01 2.52917290e-01
3.40163589e-01 6.70808792e-01 7.05428720e-01 -1.16572762e+00
-7.46465445e-01 -3.12361300e-01 -1.69831589e-02 2.11822525e-01
2.19540447e-01 -2.68760473e-01 -6.80814624e-01 -5.38145639e-02
2.42366627e-01 -1.07532585e+00 -4.77133214e-01 4.01239365e-01
-3.81966919e-01 -9.66908857e-02 8.85159373e-01 -3.77353668e-01
6.83353543e-01 -2.00023174e+00 3.15923214e-01 1.38410702e-01
2.97893405e-01 2.11296052e-01 9.61785689e-02 7.62119815e-02
1.23834834e-01 -1.65837839e-01 -6.48153961e-01 -5.99872291e-01
-3.76112521e-01 1.32247955e-01 -3.58182102e-01 8.66128206e-01
3.45608443e-01 8.30690742e-01 -1.19344032e+00 -5.76024473e-01
8.11530471e-01 6.28224909e-01 -9.00584102e-01 4.98301446e-01
-2.26681501e-01 9.10309792e-01 -4.62158263e-01 3.16617340e-01
7.45221794e-01 -5.20420074e-01 -3.51810396e-01 -8.61352012e-02
-3.26262474e-01 2.65157968e-01 -1.45918155e+00 2.22653055e+00
-4.84626442e-01 6.76705360e-01 -1.84918404e-01 -9.49630737e-01
8.29963982e-01 3.07741538e-02 8.75756979e-01 -2.30440617e-01
1.32078007e-01 2.18286023e-01 -2.94517279e-01 -4.30412114e-01
2.90831387e-01 3.48456614e-02 2.30617940e-01 2.63987631e-01
1.39971660e-03 -6.32477760e-01 2.27211162e-01 1.16206951e-01
1.12736607e+00 5.41626632e-01 2.71054327e-01 1.81107149e-01
7.91626990e-01 4.91201609e-01 6.04222894e-01 4.11600351e-01
-1.06494412e-01 8.71185720e-01 4.39856976e-01 -6.52802825e-01
-1.12861669e+00 -9.35231447e-01 1.02930784e-01 3.36158156e-01
5.56683719e-01 -4.65158582e-01 -5.56716144e-01 -7.46479273e-01
-1.19272657e-02 2.51443118e-01 -3.54448050e-01 2.13515252e-01
-9.85381842e-01 -3.26974779e-01 -2.99525633e-02 2.79142052e-01
5.80904663e-01 -8.47261190e-01 -1.12464261e+00 1.17693156e-01
-3.72048050e-01 -1.37963617e+00 -6.93144679e-01 -8.80693570e-02
-1.06765890e+00 -1.27354097e+00 -5.09579301e-01 -4.92021441e-01
7.94064760e-01 5.04010201e-01 1.11734796e+00 2.52027422e-01
-4.01478469e-01 1.41195595e-01 -1.71352819e-01 -9.60898027e-02
-1.81422129e-01 1.68039322e-01 -2.55047917e-01 1.75126910e-01
1.49144918e-01 -6.56792343e-01 -8.49787056e-01 2.52278179e-01
-6.18823290e-01 1.68000937e-01 2.29585484e-01 5.22516012e-01
7.52422035e-01 -3.36269021e-01 -8.88512060e-02 -7.35918820e-01
-7.51936734e-02 -3.72024328e-01 -9.27080393e-01 -1.53707057e-01
-2.39693165e-01 8.64869207e-02 6.23260438e-01 -2.80650467e-01
-8.16642523e-01 6.89194858e-01 -2.78826226e-02 -1.03808737e+00
-2.11079836e-01 1.15704879e-01 4.51440662e-02 -1.77373707e-01
6.07409477e-01 1.96522381e-02 1.53408840e-01 -2.58916736e-01
5.17690599e-01 1.58957690e-01 6.64418638e-01 -4.33831155e-01
1.11082995e+00 7.98081696e-01 3.01619411e-01 -8.44047308e-01
-8.86910975e-01 -8.74589860e-01 -1.25366390e+00 -3.93402964e-01
8.94368589e-01 -6.95061624e-01 -8.29895914e-01 3.12517434e-01
-1.38869143e+00 -5.83313227e-01 -4.43633407e-01 7.37469077e-01
-8.39021623e-01 4.13928926e-01 -2.46221200e-01 -4.50335085e-01
-2.07461536e-01 -1.23480690e+00 1.17734289e+00 1.82268515e-01
-7.44087398e-02 -1.15543830e+00 5.29073358e-01 2.34601632e-01
9.07033756e-02 6.22630239e-01 2.98976690e-01 -4.79064556e-03
-9.72685814e-01 -2.14660868e-01 -2.15655684e-01 1.36324197e-01
1.94333762e-01 2.03974307e-01 -8.75204384e-01 -1.90235317e-01
5.46595268e-02 -1.77943155e-01 6.70233369e-01 5.12088835e-01
9.66686666e-01 -1.08895287e-01 -3.74356955e-01 1.23346877e+00
1.57576048e+00 -2.04008088e-01 2.85398543e-01 2.95052111e-01
1.05335343e+00 5.40987015e-01 5.95495582e-01 3.06564480e-01
4.99198705e-01 6.73851013e-01 5.95733941e-01 -6.38975203e-02
-1.52481854e-01 -4.00064617e-01 5.85153932e-03 6.49097145e-01
-2.60044903e-01 1.60367582e-02 -1.03494775e+00 4.37988937e-01
-1.99726653e+00 -9.21187520e-01 -3.73400331e-01 2.38798904e+00
5.53331792e-01 6.94656223e-02 7.24299178e-02 8.76780972e-02
4.00835365e-01 1.61683246e-01 -6.83346987e-01 1.68114081e-01
3.42521697e-01 7.87260011e-02 5.05882502e-01 9.64500844e-01
-1.03685904e+00 1.15253580e+00 4.97434044e+00 8.13708231e-02
-1.59186828e+00 -1.98164172e-02 3.49429160e-01 -3.04460645e-01
2.12006733e-01 3.96900922e-01 -6.53680265e-01 4.40620095e-01
6.79673851e-01 -1.58189073e-01 2.84724683e-01 6.84522748e-01
3.45083356e-01 -1.60691291e-01 -1.24546981e+00 1.18627191e+00
-1.75989629e-03 -1.50864410e+00 -2.30118588e-01 1.38638556e-01
6.09764397e-01 3.71767491e-01 -4.82119888e-01 -7.96564296e-02
3.01260144e-01 -7.06154823e-01 5.37222326e-01 5.38027346e-01
6.43406332e-01 -4.74880755e-01 3.41458797e-01 6.40721738e-01
-1.22482359e+00 2.59822786e-01 -3.11450422e-01 -1.91515148e-01
5.12514353e-01 5.19953609e-01 -8.90315294e-01 5.75731039e-01
7.55881548e-01 1.25360906e+00 -4.44905281e-01 1.36954689e+00
-1.85738415e-01 4.04830635e-01 -5.75489163e-01 3.37984294e-01
1.61028892e-01 -5.55952370e-01 6.70186341e-01 1.07008064e+00
2.78064877e-01 -3.96682136e-02 4.75293368e-01 1.05419290e+00
6.87178373e-02 5.99135552e-03 -6.73474133e-01 6.66745722e-01
2.33160019e-01 1.46880043e+00 -9.04030681e-01 -2.07957014e-01
-3.46764863e-01 1.07420015e+00 5.21394372e-01 2.61667460e-01
-6.72478735e-01 -2.26953536e-01 7.72081256e-01 2.79969275e-01
2.11078763e-01 -4.83173996e-01 -3.01597983e-01 -1.48244786e+00
1.96434155e-01 -2.20060050e-01 1.75799355e-01 -6.47989631e-01
-1.00943863e+00 5.83633006e-01 -4.30450588e-02 -1.58564544e+00
-4.14583683e-01 -2.98866659e-01 -8.46815288e-01 1.02270234e+00
-1.88592482e+00 -8.60959888e-01 -8.41508687e-01 7.89303899e-01
5.94228745e-01 3.24165493e-01 3.22284430e-01 1.92375213e-01
-4.85564351e-01 -4.05453779e-02 -3.58047247e-01 2.66286969e-01
5.59705734e-01 -1.29358172e+00 5.39181292e-01 1.06357026e+00
3.06887269e-01 1.21935621e-01 6.55340910e-01 -5.31312227e-01
-1.32626343e+00 -1.14489686e+00 8.26878846e-01 -8.11536252e-01
6.07475698e-01 -3.04794431e-01 -1.09236634e+00 8.30877066e-01
-7.37038553e-02 7.83989608e-01 5.18082343e-02 -4.55125481e-01
-6.83667213e-02 -1.84753716e-01 -8.99705648e-01 3.86720151e-01
1.05405676e+00 -2.29569912e-01 -3.32840115e-01 2.22481936e-01
7.15031862e-01 -7.20108151e-01 -6.57899976e-01 1.83816493e-01
1.36810884e-01 -9.67723429e-01 1.07325494e+00 -3.17619473e-01
6.24667823e-01 -7.03911960e-01 1.53381526e-01 -1.37594986e+00
3.54277641e-02 -7.62356222e-01 -2.40394145e-01 8.12453926e-01
9.15701091e-02 -4.59891528e-01 1.15128088e+00 6.81293786e-01
-7.73001611e-02 -4.93941993e-01 -9.12890553e-01 -4.46692884e-01
-2.56962404e-02 -4.81069446e-01 3.65018249e-01 8.29811215e-01
-4.44948345e-01 5.25700569e-01 -1.19199671e-01 4.05723155e-01
9.46271896e-01 3.41950417e-01 1.37517846e+00 -1.31318426e+00
-9.99484360e-02 -4.21116769e-01 -6.76994443e-01 -1.43497133e+00
3.37954164e-01 -1.03792620e+00 2.31843978e-01 -1.45852745e+00
-3.36088121e-01 -6.73418581e-01 2.89785028e-01 3.04698348e-01
-2.36590385e-01 1.73358396e-01 4.19101089e-01 4.58504677e-01
-3.90452981e-01 4.72443581e-01 1.41969919e+00 -9.30036604e-03
-6.46699607e-01 2.41185024e-01 -1.24444485e-01 7.78209507e-01
6.20563745e-01 -5.43683589e-01 -4.76738602e-01 -5.90232968e-01
-1.99755356e-02 2.13076934e-01 6.96410298e-01 -9.91343498e-01
7.02782512e-01 -4.14188236e-01 2.55495638e-01 -7.39952207e-01
4.49405074e-01 -9.15273011e-01 7.57710040e-02 3.84933174e-01
-1.23616457e-01 1.41513929e-01 -7.06604570e-02 5.00385404e-01
-1.69323623e-01 -1.77717268e-01 7.62521863e-01 -2.64330149e-01
-6.66226566e-01 7.91843832e-01 3.05048317e-01 2.45376527e-01
1.23588526e+00 -3.29228550e-01 -6.24685697e-02 -2.19066367e-02
-4.01361674e-01 2.57714093e-01 6.72283947e-01 3.92459124e-01
7.39372671e-01 -1.02503395e+00 -8.01540792e-01 3.62295300e-01
1.00798942e-02 1.02297652e+00 -6.53693751e-02 8.21701288e-01
-1.00446463e+00 1.89835206e-01 -3.83513421e-02 -1.23977172e+00
-9.16503251e-01 4.10398155e-01 6.47367716e-01 -2.29007080e-02
-1.18267524e+00 6.17537200e-01 1.72729060e-01 -4.12449956e-01
6.63290396e-02 -3.33748400e-01 -9.51651391e-03 -2.76461959e-01
4.14921820e-01 5.34259140e-01 6.88785613e-02 -8.23227048e-01
-4.11542624e-01 1.01030958e+00 1.58109859e-01 -8.94282907e-02
1.40904975e+00 -9.31285247e-02 -1.05482191e-01 5.57430506e-01
1.52404761e+00 -1.40991375e-01 -1.80762148e+00 -2.90574640e-01
-4.41875868e-02 -1.07066441e+00 1.37643546e-01 -2.44177058e-02
-1.45217931e+00 9.26976502e-01 3.28449458e-01 -2.72372495e-02
8.48607361e-01 1.38979554e-02 7.31653869e-01 1.80480570e-01
2.91290641e-01 -5.76766014e-01 6.14677407e-02 4.15268421e-01
5.93803644e-01 -1.44731951e+00 -1.76098216e-02 -6.02970362e-01
-3.95917267e-01 1.15639687e+00 6.90310478e-01 -6.50394320e-01
7.61704266e-01 1.07582420e-01 1.16181783e-01 -9.15468633e-02
-6.97687387e-01 -9.19230133e-02 1.61654338e-01 3.80027264e-01
1.65038988e-01 -2.89740294e-01 3.09946507e-01 -5.80370784e-01
-3.28636765e-01 2.85397649e-01 5.74128449e-01 9.07285929e-01
-1.82952344e-01 -9.66064692e-01 -2.94799507e-01 2.09269926e-01
-1.06541343e-01 2.16586187e-01 -5.73782921e-02 4.91449982e-01
-1.00731969e-01 8.19757521e-01 4.45105910e-01 -1.28409013e-01
5.76268613e-01 -2.69768983e-01 3.97586554e-01 -5.87759912e-01
-4.26327020e-01 -6.84287250e-02 -5.63955605e-01 -9.43009079e-01
-7.28017509e-01 -7.97229707e-01 -1.34499907e+00 -4.04507428e-01
-1.10888384e-01 5.47673739e-02 6.78789377e-01 8.21104109e-01
3.84443223e-01 1.75180122e-01 9.31913912e-01 -1.34313452e+00
-1.74922124e-01 -4.75748181e-01 -7.71358535e-02 6.71903312e-01
8.36194813e-01 -5.69568753e-01 -5.93473077e-01 3.42406869e-01] | [8.526656150817871, -2.0467145442962646] |
9b217d3e-fb59-4162-9c36-672ccaffbd02 | thinking-the-fusion-strategy-of-multi | 2202.10758 | null | https://arxiv.org/abs/2202.10758v1 | https://arxiv.org/pdf/2202.10758v1.pdf | Thinking the Fusion Strategy of Multi-reference Face Reenactment | In recent advances of deep generative models, face reenactment -manipulating and controlling human face, including their head movement-has drawn much attention for its wide range of applicability. Despite its strong expressiveness, it is inevitable that the models fail to reconstruct or accurately generate unseen side of the face of a given single reference image. Most of existing methods alleviate this problem by learning appearances of human faces from large amount of data and generate realistic texture at inference time. Rather than completely relying on what generative models learn, we show that simple extension by using multiple reference images significantly improves generation quality. We show this by 1) conducting the reconstruction task on publicly available dataset, 2) conducting facial motion transfer on our original dataset which consists of multi-person's head movement video sequences, and 3) using a newly proposed evaluation metric to validate that our method achieves better quantitative results. | ['Tamaki Kojima', 'Takuya Narihira', 'Takuya Yashima'] | 2022-02-22 | null | null | null | null | ['face-reenactment'] | ['computer-vision'] | [ 2.09673807e-01 3.15343827e-01 3.37806433e-01 -5.23080230e-01
-5.56836843e-01 -4.99232531e-01 8.00629497e-01 -1.17026997e+00
-8.94573890e-03 8.62043202e-01 4.50011522e-01 3.28463465e-01
2.82373577e-01 -6.07783794e-01 -9.36240017e-01 -6.22021735e-01
1.24116160e-01 5.38138032e-01 -1.19262971e-01 -6.40595108e-02
-5.50811738e-02 5.17486453e-01 -1.79988933e+00 1.88842669e-01
4.03276682e-01 7.86700487e-01 -3.70234437e-02 6.60035312e-01
3.89532924e-01 7.13644505e-01 -6.02916360e-01 -7.77499795e-01
4.59915519e-01 -7.17372298e-01 -6.27629399e-01 5.03108263e-01
7.88243413e-01 -8.29187095e-01 -3.37620497e-01 9.06372070e-01
7.63019264e-01 9.45473984e-02 5.97864032e-01 -1.35558879e+00
-9.13565397e-01 1.97107643e-01 -8.01348507e-01 -2.53903151e-01
6.52989149e-01 2.10735753e-01 4.98234451e-01 -9.01286960e-01
9.69357252e-01 1.66934133e+00 6.32547140e-01 1.22636247e+00
-1.22951066e+00 -9.76182044e-01 -6.82389885e-02 -1.01711705e-01
-1.47423422e+00 -9.98242438e-01 8.75909030e-01 -2.83201784e-01
4.45623130e-01 2.29884967e-01 5.74390709e-01 1.73517287e+00
-5.30564860e-02 5.88061929e-01 1.19048798e+00 -2.75989145e-01
1.43954428e-02 -3.91160250e-02 -4.29774582e-01 6.90601230e-01
7.96780735e-02 3.61036450e-01 -4.14237380e-01 -1.36963859e-01
1.05798328e+00 -1.94867253e-01 -5.41720212e-01 -2.69368619e-01
-9.66227353e-01 6.44733667e-01 6.33949339e-02 -9.16060954e-02
-2.55455583e-01 3.23271304e-01 -6.40341491e-02 1.04180820e-01
3.47440302e-01 -1.73441246e-02 -8.12584609e-02 -4.10263762e-02
-1.12018156e+00 5.14963210e-01 6.79278612e-01 1.11507559e+00
5.01557589e-01 3.53521228e-01 -1.32670030e-01 7.35704422e-01
3.67812216e-01 6.67755544e-01 3.87754798e-01 -1.30757368e+00
-1.25007750e-03 1.61526844e-01 2.82232165e-01 -1.01368403e+00
-6.20913990e-02 -1.04529962e-01 -9.23557937e-01 3.60570580e-01
5.02950490e-01 -3.69000733e-01 -1.08128893e+00 2.21982121e+00
5.33380151e-01 4.70059097e-01 -1.53055057e-01 9.06748712e-01
8.74719679e-01 4.88274574e-01 -7.46457353e-02 -3.54234338e-01
1.11587119e+00 -8.41203690e-01 -7.75636375e-01 1.07995920e-01
-9.44790244e-02 -7.67684996e-01 9.38785434e-01 4.39459026e-01
-1.25439179e+00 -7.62460411e-01 -7.51672387e-01 -9.11432952e-02
1.67349100e-01 -5.61494790e-02 6.03337467e-01 7.49117911e-01
-1.42032146e+00 4.55936432e-01 -6.88177824e-01 -4.16821867e-01
5.97732127e-01 4.47520643e-01 -8.57984066e-01 -1.41282171e-01
-8.84350896e-01 6.17883980e-01 -2.19740003e-01 1.76653147e-01
-1.32161880e+00 -7.05774605e-01 -8.71288657e-01 -9.94934440e-02
5.64796152e-03 -1.06461191e+00 1.29295123e+00 -1.31185770e+00
-1.80804586e+00 9.63591695e-01 -3.67076278e-01 -7.79932663e-02
8.92286003e-01 -3.06045741e-01 -3.76105905e-01 1.53015032e-01
-1.65785238e-01 1.01447380e+00 1.25175321e+00 -1.57900894e+00
-1.94705725e-01 -6.16484880e-01 -1.96517967e-02 -1.03328619e-02
-1.89204648e-01 1.36688516e-01 -4.47061718e-01 -7.21345901e-01
-1.78791642e-01 -1.15777445e+00 8.53266641e-02 2.57782966e-01
-4.68233526e-01 1.02043867e-01 1.07690060e+00 -6.53497696e-01
8.55993032e-01 -1.75559092e+00 7.05893859e-02 -1.25396758e-01
1.85283571e-01 2.78898120e-01 -4.11771536e-01 1.28832877e-01
-3.39331388e-01 8.80978033e-02 -3.10935862e-02 -7.87424505e-01
-8.95393640e-02 2.00564206e-01 -5.27201116e-01 5.43908060e-01
2.34805197e-01 1.02069902e+00 -5.72321951e-01 -4.71271038e-01
6.67889789e-02 1.05116832e+00 -8.74699116e-01 2.35374883e-01
8.56563915e-03 8.54884803e-01 -2.02946097e-01 5.81982374e-01
9.54675198e-01 -1.37035266e-01 1.26059592e-01 -2.77401894e-01
4.03300732e-01 -3.56283426e-01 -1.05650187e+00 1.66678238e+00
-2.89480239e-01 4.46808219e-01 1.44351453e-01 -5.32031119e-01
6.20651484e-01 6.45077467e-01 3.88841808e-01 -4.92543936e-01
1.40216529e-01 -2.65988111e-01 -4.13316302e-02 -5.30070662e-01
1.83052838e-01 -5.13106763e-01 2.69126713e-01 5.04893124e-01
9.77338180e-02 -1.56975344e-01 -1.75121631e-02 -4.52075899e-02
7.54932702e-01 6.47329628e-01 7.80799463e-02 -1.01897143e-01
3.94409984e-01 -5.49736977e-01 5.05048931e-01 3.10430557e-01
-1.43118680e-01 1.27706099e+00 2.82988250e-01 -4.73699629e-01
-1.08511949e+00 -1.17531252e+00 -1.25836834e-01 7.71951497e-01
-2.01114699e-01 -8.86486024e-02 -1.08080781e+00 -5.52495956e-01
-7.61068910e-02 6.70597792e-01 -1.02586699e+00 -3.03305276e-02
-5.56799054e-01 -8.40305030e-01 5.75083315e-01 4.97480541e-01
5.21395564e-01 -1.08845258e+00 -4.93617296e-01 -1.44692212e-01
-2.43539765e-01 -1.17910349e+00 -5.44000566e-01 -6.91633821e-01
-5.47059476e-01 -1.00075793e+00 -9.60986733e-01 -5.29876173e-01
9.54155982e-01 1.53146923e-01 1.20923173e+00 1.96234331e-01
-3.64091337e-01 5.67941427e-01 -1.06074503e-02 -1.82263941e-01
-5.32275558e-01 -4.17126417e-01 1.73880249e-01 2.83965170e-01
-1.17646560e-01 -8.66066575e-01 -7.10222661e-01 3.65918547e-01
-8.46188128e-01 3.28527153e-01 2.02571511e-01 6.82684958e-01
2.37297609e-01 -2.20824212e-01 6.51279509e-01 -9.60741520e-01
4.46923763e-01 -2.74278522e-01 -4.67999160e-01 1.20137222e-01
-3.32177460e-01 -1.27227694e-01 4.58535016e-01 -4.21527386e-01
-1.40233862e+00 1.58170223e-01 -2.75357932e-01 -6.87119365e-01
-2.48853728e-01 -3.39579731e-01 -4.72991556e-01 2.07917430e-02
5.03647447e-01 2.74802774e-01 2.50955701e-01 -3.09303313e-01
3.78981441e-01 3.68026465e-01 7.58647978e-01 -7.08204567e-01
1.10116017e+00 7.82575727e-01 1.31637007e-01 -7.24725783e-01
-6.36213660e-01 2.82749832e-01 -6.94524705e-01 -3.79354507e-01
8.71348858e-01 -9.90991294e-01 -1.01733387e+00 5.42054296e-01
-1.15842366e+00 -1.94086626e-01 2.08789436e-03 1.85079679e-01
-6.39252067e-01 2.96015561e-01 -5.37323236e-01 -9.01719272e-01
-1.97908401e-01 -1.12014031e+00 1.47145748e+00 1.51622757e-01
-3.61013979e-01 -8.87446880e-01 1.87963828e-01 4.55911100e-01
4.99884456e-01 6.05546474e-01 5.65593481e-01 -8.78325105e-02
-6.29426360e-01 -2.76753366e-01 1.08990911e-02 3.77233863e-01
3.59691501e-01 3.47416878e-01 -1.41767991e+00 -4.37380284e-01
2.38825977e-02 -3.77744108e-01 5.74556231e-01 3.84103417e-01
1.21724308e+00 -3.98120999e-01 -3.42357337e-01 6.74731314e-01
1.19317889e+00 1.32381573e-01 1.01039457e+00 -3.26987714e-01
6.98499441e-01 7.31670916e-01 1.20812327e-01 4.77354318e-01
3.41625869e-01 7.70953655e-01 4.73439068e-01 -3.83789800e-02
-4.21371043e-01 -5.30478895e-01 5.17194092e-01 4.10615742e-01
-5.76794207e-01 -2.83493012e-01 -5.52997351e-01 2.94121295e-01
-1.53283024e+00 -1.41538656e+00 2.55830646e-01 2.20278692e+00
7.76461840e-01 -3.50676090e-01 2.52012819e-01 -1.08143128e-01
7.49748409e-01 -6.48685172e-02 -4.51068848e-01 1.12711899e-01
-1.19957246e-01 2.93168277e-01 -3.26851696e-01 5.44509649e-01
-8.02553654e-01 8.79768789e-01 7.13001442e+00 4.53942716e-01
-1.21552646e+00 9.79019105e-02 9.21317339e-01 -3.88742864e-01
-3.72748911e-01 -1.86139539e-01 -8.48968923e-01 5.55454314e-01
6.90380633e-01 -1.31343141e-01 4.97458637e-01 7.60007203e-01
2.46198848e-01 1.84682861e-01 -1.33776891e+00 1.16052723e+00
4.31332797e-01 -1.16005683e+00 4.00754958e-01 1.39466166e-01
8.40884089e-01 -5.53126335e-01 3.93097281e-01 2.19867513e-01
3.35819274e-01 -1.43842638e+00 8.47842157e-01 7.55002975e-01
9.52140391e-01 -6.18650198e-01 2.86070228e-01 2.04817787e-01
-8.29609394e-01 2.33746514e-01 -2.84559280e-01 7.80079663e-02
2.67353356e-01 2.65300572e-01 -7.34521270e-01 3.25993419e-01
5.18721104e-01 4.65824604e-01 -5.78673184e-01 5.81212819e-01
-1.17225476e-01 4.43611652e-01 -2.28597060e-01 5.27430356e-01
-3.29864889e-01 -8.67105573e-02 4.04473513e-01 8.33354294e-01
5.46642303e-01 1.22835763e-01 -2.64260143e-01 1.08757555e+00
-2.81246871e-01 -2.87362635e-01 -8.06689322e-01 1.32810116e-01
2.33410314e-01 1.33235884e+00 -4.13649231e-01 -1.90344885e-01
-3.01463962e-01 1.17491817e+00 2.06644550e-01 4.70938087e-01
-1.16774333e+00 2.29357943e-01 7.39299476e-01 4.63419020e-01
8.66630226e-02 -3.31973024e-02 2.35889912e-01 -1.16588390e+00
-1.44006878e-01 -8.87810946e-01 1.18880514e-02 -1.13687789e+00
-1.23621809e+00 9.20206666e-01 7.99745098e-02 -1.00316417e+00
-8.09431851e-01 -3.20976466e-01 -5.33216953e-01 7.22690940e-01
-1.10467482e+00 -1.58879554e+00 -5.79696536e-01 8.57176900e-01
4.59453702e-01 -8.99337828e-02 8.09207857e-01 4.50167120e-01
-5.99474132e-01 9.62851524e-01 -5.07388592e-01 -9.78777092e-03
8.40066314e-01 -8.22277606e-01 4.59690660e-01 7.09704936e-01
9.30748060e-02 7.01757610e-01 8.78720164e-01 -4.01862323e-01
-1.40207291e+00 -1.07659018e+00 5.17217219e-01 -8.34005475e-01
4.62433584e-02 -5.81424356e-01 -7.07748771e-01 1.02769268e+00
3.00981075e-01 3.10102046e-01 4.66005832e-01 -4.08744454e-01
-2.98312664e-01 -1.12644486e-01 -1.32749856e+00 8.56645226e-01
1.48675740e+00 -2.05694243e-01 -3.05681705e-01 2.01196864e-01
3.62279475e-01 -3.72863799e-01 -6.17385685e-01 4.99085516e-01
8.53054404e-01 -1.19905007e+00 8.97953868e-01 -6.47562265e-01
7.27056086e-01 -3.18634778e-01 -1.95095599e-01 -1.14083743e+00
-1.73040688e-01 -9.90141094e-01 -6.15239143e-02 1.43056631e+00
8.58191028e-03 -4.25329775e-01 7.85230160e-01 7.93709219e-01
1.00364096e-01 -6.17202044e-01 -7.43397057e-01 -4.82855022e-01
1.61137447e-01 -1.08916186e-01 8.37414145e-01 8.89652371e-01
-4.31513399e-01 4.17785734e-01 -9.43640649e-01 3.21107395e-02
7.30975747e-01 1.16608933e-01 1.37085593e+00 -1.16821420e+00
-1.92460462e-01 -1.39719874e-01 -4.45124924e-01 -9.02967930e-01
5.35544515e-01 -7.17916250e-01 -9.42943692e-02 -1.12483180e+00
5.20220220e-01 3.62781845e-02 3.51264983e-01 4.89809543e-01
-2.68453453e-02 6.89888537e-01 2.12455485e-02 1.13905855e-01
-3.23116034e-01 7.90175080e-01 1.72712123e+00 3.51815343e-01
3.06377649e-01 -1.20636076e-01 -8.26634586e-01 8.12097430e-01
2.66466588e-01 -2.69329876e-01 -7.11784899e-01 -4.39978510e-01
-7.38190264e-02 2.61675417e-01 6.90130532e-01 -8.51100683e-01
-1.12979375e-01 -1.07535429e-01 1.01219034e+00 -1.03761092e-01
6.08919084e-01 -7.06728756e-01 8.34180653e-01 2.11543649e-01
-1.39470413e-01 1.79385617e-01 1.78258762e-01 4.53241915e-01
-6.85016513e-02 1.74117520e-01 1.08535194e+00 -8.75341967e-02
-4.03059542e-01 5.77033460e-01 1.75258256e-02 4.46400866e-02
9.86377060e-01 -3.94997746e-01 -1.38239622e-01 -8.55591059e-01
-8.58025014e-01 -3.34373742e-01 8.76044691e-01 6.01001620e-01
5.45308888e-01 -1.57082069e+00 -8.33898723e-01 5.97298443e-01
-2.61487871e-01 -1.68190777e-01 2.86580205e-01 5.24397790e-01
-3.58734041e-01 2.25325059e-02 -4.46555525e-01 -5.15088022e-01
-1.34552991e+00 4.72299814e-01 4.70702857e-01 2.61069894e-01
-6.54292643e-01 7.51498640e-01 8.31196487e-01 -1.32933140e-01
-8.13770816e-02 1.39068216e-01 6.14789054e-02 -3.07738572e-01
6.46550834e-01 1.44333005e-01 -1.83043495e-01 -1.06399155e+00
-2.27829561e-01 7.42377281e-01 -2.07233559e-02 -3.18502307e-01
1.34266043e+00 -9.20997709e-02 6.83520138e-02 -1.35822400e-01
1.10709989e+00 2.27859899e-01 -1.65564787e+00 2.85093874e-01
-5.42287111e-01 -7.12869465e-01 -3.53525966e-01 -6.45833015e-01
-1.55174100e+00 7.42224932e-01 5.91622949e-01 -3.82730305e-01
1.17405140e+00 -1.40142679e-01 7.41950214e-01 -7.89725129e-03
6.43817484e-01 -5.34044564e-01 3.32792640e-01 1.01518162e-01
1.24944746e+00 -1.22737360e+00 -1.89919606e-01 -3.80627036e-01
-6.19706035e-01 9.28607702e-01 7.60660827e-01 -2.10551545e-01
7.70559072e-01 5.20249486e-01 8.68538022e-02 -5.14074489e-02
-9.13236797e-01 1.75347105e-01 2.49492124e-01 8.09778273e-01
5.46825051e-01 -2.66390562e-01 1.68822035e-01 6.51165724e-01
-5.93013108e-01 2.78695494e-01 4.94206071e-01 5.32808602e-01
8.02185088e-02 -1.01187682e+00 -3.57325256e-01 1.49855427e-02
-5.85074663e-01 1.46940753e-01 -2.75681019e-01 1.03310454e+00
7.11687207e-02 7.81755090e-01 1.20327652e-01 -2.22562939e-01
1.69035703e-01 4.52413112e-02 1.09381855e+00 -4.94680345e-01
-8.62784684e-02 6.23022355e-02 -2.40770131e-01 -6.65444016e-01
-4.40952569e-01 -6.43925071e-01 -1.01803398e+00 -5.79549789e-01
-5.61926663e-02 -2.20608100e-01 2.62421489e-01 7.11416543e-01
4.54408973e-01 1.95341125e-01 6.16451681e-01 -1.24826610e+00
-5.99126637e-01 -1.04590523e+00 -6.28833115e-01 8.69292378e-01
3.00105482e-01 -9.16655123e-01 -3.73999387e-01 6.13050938e-01] | [12.822565078735352, -0.24185024201869965] |
6a6dae82-36f7-4f7b-9232-132d296c5f2c | factorizable-graph-convolutional-networks | 2010.05421 | null | https://arxiv.org/abs/2010.05421v1 | https://arxiv.org/pdf/2010.05421v1.pdf | Factorizable Graph Convolutional Networks | Graphs have been widely adopted to denote structural connections between entities. The relations are in many cases heterogeneous, but entangled together and denoted merely as a single edge between a pair of nodes. For example, in a social network graph, users in different latent relationships like friends and colleagues, are usually connected via a bare edge that conceals such intrinsic connections. In this paper, we introduce a novel graph convolutional network (GCN), termed as factorizable graph convolutional network(FactorGCN), that explicitly disentangles such intertwined relations encoded in a graph. FactorGCN takes a simple graph as input, and disentangles it into several factorized graphs, each of which represents a latent and disentangled relation among nodes. The features of the nodes are then aggregated separately in each factorized latent space to produce disentangled features, which further leads to better performances for downstream tasks. We evaluate the proposed FactorGCN both qualitatively and quantitatively on the synthetic and real-world datasets, and demonstrate that it yields truly encouraging results in terms of both disentangling and feature aggregation. Code is publicly available at https://github.com/ihollywhy/FactorGCN.PyTorch. | ['Xinchao Wang', 'Mingli Song', 'Zunlei Feng', 'Yiding Yang'] | 2020-10-12 | null | http://proceedings.neurips.cc/paper/2020/hash/ea3502c3594588f0e9d5142f99c66627-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/ea3502c3594588f0e9d5142f99c66627-Paper.pdf | neurips-2020-12 | ['graph-regression'] | ['graphs'] | [-1.37944311e-01 3.85318726e-01 -2.89919317e-01 -1.73287198e-01
1.37717798e-01 -8.83893788e-01 9.03198719e-01 1.05930679e-02
2.84020036e-01 6.86287344e-01 4.47174370e-01 -3.42734665e-01
-4.71567482e-01 -1.05772531e+00 -4.47922766e-01 -6.08607233e-01
-4.03947622e-01 3.41590047e-01 -2.32728705e-01 -1.15539886e-01
-3.97605151e-01 4.29968297e-01 -7.45389581e-01 2.26973183e-02
8.35011780e-01 3.51905555e-01 -3.48181218e-01 5.10403812e-01
1.68517247e-01 7.42835522e-01 -3.38440895e-01 -8.63109469e-01
2.03775600e-01 -3.17627639e-01 -5.93244553e-01 1.05444632e-01
1.38962001e-01 9.12600569e-03 -1.05726993e+00 9.63938534e-01
-5.39085492e-02 -2.43547142e-01 6.07761323e-01 -1.73803306e+00
-1.09439170e+00 1.03628814e+00 -5.52487135e-01 3.23830661e-03
3.10431182e-01 -2.96994727e-02 1.68917072e+00 -6.72363281e-01
4.67358381e-01 1.34214842e+00 2.60145605e-01 1.71088174e-01
-1.60501587e+00 -9.56676304e-01 1.63015857e-01 -1.78547978e-01
-1.38753295e+00 -3.57299954e-01 9.26842988e-01 -6.66320562e-01
6.53965533e-01 4.11228657e-01 1.00764883e+00 1.11729705e+00
3.37293774e-01 6.36265457e-01 7.95593977e-01 3.86806019e-02
-2.60848761e-01 -2.35283837e-01 3.07044297e-01 9.12834764e-01
8.76897871e-01 6.87172934e-02 -4.91968125e-01 -3.02689672e-01
8.85369122e-01 3.26215714e-01 -4.90476221e-01 -6.62567854e-01
-1.52615952e+00 8.76610875e-01 9.88746047e-01 2.62506962e-01
-1.43648818e-01 2.21533164e-01 1.20132692e-01 3.98204923e-01
5.87495446e-01 5.18042028e-01 -1.46299839e-01 3.20983917e-01
-3.62051070e-01 4.06583399e-02 1.01100302e+00 8.24460626e-01
9.48982835e-01 -2.07421035e-01 7.65157863e-02 4.63042974e-01
5.56900561e-01 2.83746570e-01 -1.17540453e-03 -7.49778509e-01
5.56684077e-01 1.18962157e+00 -1.80004463e-01 -1.73434341e+00
-3.83982480e-01 -5.29387832e-01 -1.40111911e+00 -1.35777861e-01
1.30427152e-01 -1.94245458e-01 -7.24582732e-01 1.91026545e+00
1.20426901e-01 5.14042258e-01 -8.74296278e-02 7.55076766e-01
1.16151834e+00 4.19267803e-01 -2.27825001e-01 7.68255517e-02
1.31199217e+00 -1.07974660e+00 -7.85853386e-01 -2.05122784e-01
4.60389316e-01 -3.73124868e-01 5.69777787e-01 8.76330063e-02
-8.29819143e-01 -1.33654967e-01 -1.16540086e+00 -4.99872379e-02
-3.65323126e-01 -9.52693596e-02 1.19496310e+00 3.12431008e-01
-1.19717765e+00 6.09702766e-01 -9.33174193e-01 -1.46679446e-01
5.17948270e-01 4.91867542e-01 -1.01095593e+00 -9.18883756e-02
-1.48973978e+00 4.01365459e-01 3.01238060e-01 4.83861268e-01
-6.83156252e-01 -4.42129999e-01 -1.06618094e+00 3.92370284e-01
5.59763789e-01 -9.91229057e-01 5.71167827e-01 -5.33875585e-01
-9.34744120e-01 6.97537959e-01 5.74060753e-02 1.40591450e-02
2.32439488e-01 -5.24742827e-02 -8.16952348e-01 -2.08408222e-01
6.31094500e-02 2.25724235e-01 7.06169248e-01 -1.23597562e+00
1.00744203e-01 -3.11089784e-01 5.50822318e-01 6.57436848e-02
-3.21315706e-01 -2.10210621e-01 -4.84257996e-01 -6.27996385e-01
3.09705049e-01 -1.03695762e+00 -7.17409700e-02 -1.05352461e-01
-1.17833674e+00 -1.84385583e-01 6.69553459e-01 -2.84632385e-01
1.49287093e+00 -2.01797390e+00 6.15405679e-01 2.29555696e-01
1.55306542e+00 1.73394606e-01 -3.08902323e-01 8.65171254e-01
-4.37037826e-01 4.58632350e-01 -1.50762592e-02 -2.61478364e-01
2.99868379e-02 3.04668963e-01 4.19451967e-02 6.82558835e-01
2.44588017e-01 1.43151712e+00 -1.18592083e+00 -1.18835330e-01
-1.13427974e-02 5.15829206e-01 -3.61663312e-01 2.09659785e-01
5.32186106e-02 4.97476131e-01 -6.45534992e-01 3.18064243e-01
7.41450071e-01 -8.38442087e-01 5.43330848e-01 -2.72002935e-01
2.57825017e-01 3.30063045e-01 -1.01004326e+00 1.53398669e+00
-1.81873754e-01 7.72112846e-01 6.58416897e-02 -8.46868515e-01
9.01512027e-01 3.21788013e-01 4.34830487e-01 2.58483994e-03
1.40556961e-01 -8.16234052e-02 4.41654056e-01 -2.74195671e-01
3.09363931e-01 2.83072323e-01 -2.01088965e-01 7.51034319e-01
3.17666262e-01 1.95593581e-01 2.53741771e-01 9.42780733e-01
1.39016354e+00 -3.10119927e-01 6.64707065e-01 -3.18596452e-01
2.83778131e-01 -6.45997882e-01 5.35146058e-01 4.01426584e-01
-2.16212366e-02 3.71431142e-01 1.24707568e+00 -4.57281142e-01
-8.05582047e-01 -1.31465650e+00 2.29450569e-01 5.05292416e-01
2.89464235e-01 -9.40340877e-01 -2.57093936e-01 -6.96261764e-01
3.39962482e-01 1.24217704e-01 -8.67577374e-01 -3.02171707e-01
-3.03941011e-01 -6.93048656e-01 4.88774657e-01 1.44851565e-01
2.81892121e-01 -6.93203747e-01 3.89860243e-01 -8.07357579e-02
-5.25682792e-02 -1.05136991e+00 -5.04660904e-01 -9.47406963e-02
-5.28308809e-01 -1.35332155e+00 -2.63434738e-01 -4.11310434e-01
8.43955755e-01 6.02005303e-01 1.19736600e+00 5.03581762e-01
5.21234162e-02 -4.13391478e-02 -2.90965766e-01 2.99799353e-01
-1.20224312e-01 3.47883999e-01 7.65556470e-02 3.99609327e-01
2.36297220e-01 -1.07966530e+00 -6.25470579e-01 2.78429747e-01
-9.27036464e-01 4.73278016e-01 4.04120177e-01 8.57176304e-01
-1.32401241e-02 3.29329446e-02 3.45590532e-01 -1.33710253e+00
9.39899921e-01 -8.94351363e-01 -3.70448083e-01 3.16996574e-01
-4.66375113e-01 1.28634840e-01 4.66518849e-01 -2.23500863e-01
-4.37705487e-01 -4.04375046e-01 4.42694306e-01 -4.37021613e-01
2.01396421e-01 8.11003268e-01 -5.86921155e-01 5.10410704e-02
4.69077259e-01 -1.77142233e-01 -3.21986130e-03 -3.82990241e-01
6.91633165e-01 3.98600668e-01 1.03782713e-01 -3.49837005e-01
1.26997256e+00 3.63043398e-01 2.17574283e-01 -4.37304884e-01
-7.50667632e-01 -1.34338886e-01 -7.87135303e-01 -7.89351836e-02
6.41318500e-01 -1.01761460e+00 -8.73240650e-01 4.21310127e-01
-1.15204668e+00 -8.05227235e-02 4.09574136e-02 4.02625531e-01
4.61464897e-02 2.22500503e-01 -6.72995508e-01 -2.79221147e-01
-9.39904992e-03 -9.72108841e-01 8.20129395e-01 2.62368351e-01
-3.22010696e-01 -1.38437545e+00 2.86375642e-01 2.90564239e-01
2.08183169e-01 5.57676613e-01 1.01447737e+00 -7.07375705e-01
-8.44331741e-01 -4.18784827e-01 -5.78352332e-01 2.68264506e-02
6.08600616e-01 2.22607091e-01 -5.74721575e-01 -4.80303884e-01
-6.60561979e-01 -4.05673459e-02 9.28202868e-01 -9.44775790e-02
6.54303730e-01 -5.34687161e-01 -6.13969505e-01 8.03913951e-01
1.05922902e+00 -2.74559826e-01 4.06104565e-01 -2.64760524e-01
1.42728388e+00 3.82520229e-01 -2.74485290e-01 1.47534892e-01
5.61489880e-01 5.34970105e-01 4.66828406e-01 -2.24789023e-01
-7.31622353e-02 -5.58662355e-01 9.30391997e-02 1.30363011e+00
-4.01766330e-01 -7.61039793e-01 -7.56963670e-01 2.22831026e-01
-2.08329105e+00 -6.20437860e-01 -4.17035490e-01 1.66344082e+00
3.95724326e-01 -2.33309567e-02 -1.21057034e-01 -1.30961880e-01
9.37188506e-01 7.20742226e-01 -6.80531144e-01 -7.07644001e-02
-2.72812068e-01 5.26128262e-02 2.22572684e-01 8.35660338e-01
-8.90108705e-01 9.20385063e-01 5.14799261e+00 4.48259503e-01
-7.28217065e-01 6.38282895e-02 4.91095006e-01 4.18834686e-02
-8.27110469e-01 4.00518864e-01 -1.47429392e-01 3.32216173e-01
4.51229632e-01 -6.72378242e-01 7.24509239e-01 3.52475166e-01
4.94032679e-03 3.52387160e-01 -1.12449896e+00 7.66774297e-01
-2.44068488e-01 -1.32834363e+00 2.60790616e-01 3.35854024e-01
8.13863277e-01 -7.71749169e-02 1.03765517e-01 8.49216208e-02
8.06336164e-01 -1.49778652e+00 2.01594114e-01 5.74221194e-01
6.74929261e-01 -5.96073866e-01 5.62150061e-01 1.55433223e-01
-1.43597543e+00 2.95830816e-01 -1.09003618e-01 -3.63933802e-01
5.39802723e-02 8.37131977e-01 -5.09666204e-01 1.06752777e+00
2.07851976e-01 1.18672252e+00 -6.10132396e-01 6.53103232e-01
-7.83019662e-01 4.62395221e-01 8.85061473e-02 6.44377545e-02
7.12897256e-02 -7.12764204e-01 7.16435373e-01 6.68761313e-01
1.30772948e-01 1.98810905e-01 2.11018268e-02 1.25929201e+00
-5.54580510e-01 -2.55403727e-01 -8.58963549e-01 -7.44516671e-01
5.38263679e-01 1.67450416e+00 -7.24913955e-01 -1.13356151e-01
-5.58155775e-01 9.75241482e-01 8.61382544e-01 7.21266925e-01
-6.82180643e-01 -3.46446693e-01 1.08051324e+00 -3.62882880e-03
-8.74205604e-02 -4.07919347e-01 2.04952303e-02 -1.78656363e+00
-4.02530544e-02 -5.94396055e-01 2.03994125e-01 -7.26031780e-01
-1.50584793e+00 9.18111026e-01 -1.56652793e-01 -1.00869703e+00
7.89090320e-02 -5.24833202e-01 -7.21190393e-01 1.10321438e+00
-9.73808467e-01 -1.31567287e+00 -4.61874127e-01 6.35770559e-01
-3.44621390e-01 -1.48099929e-01 8.84870589e-01 2.10138977e-01
-8.55047524e-01 5.81512928e-01 -6.55083358e-02 4.68361527e-01
3.41194332e-01 -1.26660407e+00 7.89931715e-01 7.74071872e-01
5.15415728e-01 1.10188890e+00 3.73598218e-01 -8.05807710e-01
-1.39555955e+00 -1.13312793e+00 1.01404250e+00 -5.33820212e-01
1.21624684e+00 -9.82593119e-01 -8.60882521e-01 1.06891549e+00
3.75937790e-01 4.06129003e-01 8.09444070e-01 4.92372751e-01
-7.95656502e-01 9.20998827e-02 -7.45510161e-01 8.02652359e-01
1.59513474e+00 -7.52641797e-01 -2.80866385e-01 3.84736240e-01
1.26847339e+00 -2.06405357e-01 -1.03374505e+00 3.97494376e-01
6.92074299e-01 -1.12077343e+00 8.57976437e-01 -9.44737375e-01
6.01470053e-01 -2.34669283e-01 1.16344556e-01 -1.74689150e+00
-8.02123904e-01 -8.55806649e-01 -3.81074578e-01 1.06805134e+00
6.38057888e-01 -1.14630973e+00 8.19600224e-01 5.10488510e-01
2.78963685e-01 -9.09081280e-01 -5.62591612e-01 -5.27309477e-01
-1.92064792e-01 2.18700860e-02 1.05632389e+00 1.45910299e+00
2.30184600e-01 7.90430784e-01 -5.88974953e-01 3.20620745e-01
5.19235969e-01 3.54237914e-01 7.84734786e-01 -1.41179717e+00
-4.28149194e-01 -4.89653587e-01 -8.44939828e-01 -9.46447790e-01
3.05335641e-01 -1.26945233e+00 -6.55953646e-01 -1.54841781e+00
5.46184897e-01 -5.06779134e-01 -2.93799549e-01 4.76374269e-01
-3.75812083e-01 1.98185101e-01 2.16420025e-01 2.41955638e-01
-4.74514693e-01 5.92361093e-01 1.71846914e+00 -2.57305413e-01
-1.31754935e-01 1.93677992e-02 -1.15254295e+00 4.53122646e-01
6.63363636e-01 -4.07976389e-01 -6.13297999e-01 -3.69325727e-01
6.41233087e-01 3.26557219e-01 4.28171933e-01 -4.42384869e-01
1.94884375e-01 -5.90210855e-02 3.77608389e-02 -6.99522570e-02
3.44354063e-01 -6.84652388e-01 8.01764190e-01 2.75300533e-01
-2.60164708e-01 8.04214776e-02 -3.52616280e-01 8.14357340e-01
-3.10352385e-01 3.90769094e-01 1.59176350e-01 2.33642943e-02
-8.04496482e-02 8.03223670e-01 2.02678382e-01 -1.03799306e-01
8.80789161e-01 2.06589431e-01 -8.62356305e-01 -7.24835634e-01
-9.92092371e-01 3.07173222e-01 4.13974017e-01 4.99336660e-01
4.99747843e-01 -1.64385331e+00 -7.66806781e-01 4.57322299e-01
1.45421639e-01 1.13399819e-01 2.63437003e-01 8.84744883e-01
-3.86932880e-01 3.77016693e-01 -1.25292346e-01 -3.35739672e-01
-1.11905587e+00 5.46512663e-01 2.36818939e-01 -4.97441113e-01
-5.13396442e-01 8.68945479e-01 5.77113807e-01 -6.50379598e-01
-1.50970772e-01 -3.03512216e-01 -2.44213209e-01 8.93466547e-02
6.21448867e-02 1.62605166e-01 -3.23890001e-01 -9.20206428e-01
-2.80676246e-01 2.78479904e-01 -1.40397698e-01 2.10928753e-01
1.34214711e+00 -3.59347574e-02 -7.13380814e-01 4.16868299e-01
1.45131433e+00 3.43830973e-01 -9.65761542e-01 -4.77738231e-01
-2.05505475e-01 -6.56006575e-01 -2.32633904e-01 -3.58986706e-01
-1.53187168e+00 5.86357892e-01 -4.08238024e-01 9.05388296e-01
8.25255930e-01 3.70552003e-01 4.75114584e-01 1.91000864e-01
1.83512062e-01 7.98256025e-02 9.45231766e-02 3.45192015e-01
9.89597499e-01 -1.04778516e+00 2.69505203e-01 -9.18927252e-01
-4.50602770e-01 9.51472402e-01 4.54446644e-01 -5.64180136e-01
1.09717584e+00 7.94019550e-02 -3.17382872e-01 -6.53880358e-01
-8.13914776e-01 4.30605263e-02 5.50485253e-01 3.89818549e-01
3.03166270e-01 7.58853972e-01 -1.72036931e-01 7.01294959e-01
-3.94556701e-01 -3.46031517e-01 5.37914872e-01 3.31913561e-01
1.43118396e-01 -1.25276208e+00 2.49444097e-01 6.33161008e-01
-1.12659119e-01 -2.25841299e-01 -7.77093291e-01 8.94842088e-01
5.37021458e-02 8.66878211e-01 1.09195709e-04 -8.74122202e-01
5.98557666e-02 -4.44183290e-01 2.21367821e-01 -8.58851075e-01
-2.33200818e-01 -1.96901679e-01 2.98820347e-01 -6.69202447e-01
-5.01872838e-01 -3.94540966e-01 -8.45196247e-01 -6.66216075e-01
-4.23858523e-01 2.81113207e-01 2.76246704e-02 7.36780941e-01
5.62813044e-01 6.81207955e-01 8.07339907e-01 -6.28775477e-01
1.42340466e-01 -9.12143052e-01 -9.60733593e-01 3.56664151e-01
5.42473972e-01 -8.30019593e-01 -6.66829586e-01 -2.87664413e-01] | [7.244682788848877, 6.239664554595947] |
5e699986-3422-4f1b-9e4d-91ea97e3d28c | service-choreography-sbvr-and-time | 1512.07685 | null | http://arxiv.org/abs/1512.07685v1 | http://arxiv.org/pdf/1512.07685v1.pdf | Service Choreography, SBVR, and Time | We propose the use of structured natural language (English) in specifying
service choreographies, focusing on the what rather than the how of the
required coordination of participant services in realising a business
application scenario. The declarative approach we propose uses the OMG standard
Semantics of Business Vocabulary and Rules (SBVR) as a modelling language. The
service choreography approach has been proposed for describing the global
orderings of the invocations on interfaces of participant services. We
therefore extend SBVR with a notion of time which can capture the coordination
of the participant services, in terms of the observable message exchanges
between them. The extension is done using existing modelling constructs in
SBVR, and hence respects the standard specification. The idea is that users -
domain specialists rather than implementation specialists - can verify the
requested service composition by directly reading the structured English used
by SBVR. At the same time, the SBVR model can be represented in formal logic so
it can be parsed and executed by a machine. | ['Sotiris Moschoyiannis', 'Paul Krause', 'Nurulhuda A. Manaf'] | 2015-12-24 | null | null | null | null | ['service-composition', 'formal-logic'] | ['miscellaneous', 'reasoning'] | [ 2.36748129e-01 5.46402693e-01 1.41260505e-01 -6.64905131e-01
1.69579402e-01 -6.76756203e-01 1.29893577e+00 -5.24009392e-02
-1.67631969e-01 1.63193464e-01 2.05307335e-01 -8.46828580e-01
-3.45557451e-01 -1.15198576e+00 -1.21046960e-01 -2.81757206e-01
-1.63943291e-01 6.96554303e-01 9.66240048e-01 -5.99008024e-01
-1.49934843e-01 6.37495577e-01 -1.61146533e+00 6.62828624e-01
1.31355599e-01 9.41358685e-01 1.47378489e-01 7.13059604e-01
-7.57884264e-01 1.07887065e+00 -2.70586371e-01 -2.77477831e-01
9.00456011e-02 -4.74189460e-01 -1.26064813e+00 4.14810449e-01
-6.36952817e-01 -2.47687414e-01 2.76647151e-01 7.15969741e-01
-4.46114302e-01 -1.60812452e-01 3.36178869e-01 -1.73479676e+00
-9.37469527e-02 7.50188649e-01 2.11913869e-01 -4.73317951e-01
8.26573193e-01 1.45931557e-01 1.05123842e+00 -3.09684664e-01
9.01959419e-01 1.32015026e+00 1.08612351e-01 7.92452335e-01
-1.34327412e+00 8.40432644e-02 1.42423451e-01 1.02687009e-01
-1.17364299e+00 -6.06590211e-01 2.34677523e-01 -5.12699127e-01
1.14942193e+00 1.03602159e+00 4.26931709e-01 4.60106999e-01
8.63543972e-02 2.05064476e-01 8.65904212e-01 -9.52297747e-01
4.67163235e-01 4.81468171e-01 4.38559294e-01 -3.95084321e-02
3.78869951e-01 -1.15436934e-01 4.53121066e-02 -4.66793031e-01
6.72277033e-01 2.68800974e-01 -4.69587184e-02 -7.25541294e-01
-1.16643655e+00 3.25061440e-01 -4.83967543e-01 7.73729324e-01
-1.93823352e-01 2.63242096e-01 6.87478542e-01 7.66429782e-01
-3.07632297e-01 1.07726559e-01 -7.29960859e-01 -2.58698821e-01
-3.32407653e-01 4.10737544e-01 1.80935085e+00 1.17847180e+00
1.48614168e-01 -2.05932155e-01 2.56271631e-01 3.54344174e-02
1.15446305e+00 1.38280913e-01 -6.14346974e-02 -1.00376976e+00
-1.81984603e-01 8.56714964e-01 9.22596931e-01 -5.99860430e-01
1.25345141e-01 3.33504915e-01 -4.60327417e-02 5.19241095e-01
5.32346547e-01 3.76192957e-01 -4.82811391e-01 1.13852024e+00
4.30545509e-01 -3.05595756e-01 3.97389740e-01 6.09669030e-01
4.40557629e-01 6.14796460e-01 2.24809468e-01 -4.43318009e-01
1.71081054e+00 -4.15133446e-01 -8.03878427e-01 2.93980539e-01
6.89808309e-01 -2.36223146e-01 8.47727358e-01 4.19035017e-01
-1.27867270e+00 1.95231602e-01 -8.03661108e-01 2.06064716e-01
-3.31469268e-01 -6.65855229e-01 7.05115557e-01 5.12179077e-01
-1.10605097e+00 1.76314324e-01 -8.91836107e-01 -4.12115812e-01
-6.23490930e-01 5.26927710e-01 -2.89829403e-01 1.70753285e-01
-1.08266640e+00 8.35874498e-01 5.14362514e-01 1.72016695e-01
-9.69428718e-01 5.76026104e-02 -9.28306699e-01 2.39153117e-01
5.99605262e-01 -2.95416564e-01 1.95207787e+00 -1.25940657e+00
-1.68123603e+00 9.20085669e-01 -6.09937012e-02 -4.99487579e-01
6.88142419e-01 5.93534589e-01 -1.11768007e+00 -1.46917924e-01
-1.07552566e-01 -1.98455855e-01 9.03698131e-02 -1.38503909e+00
-1.16346800e+00 -1.59783900e-01 7.29668677e-01 -3.93445432e-01
3.88261467e-01 8.29332888e-01 -4.43924785e-01 3.00873429e-01
4.19667244e-01 -5.92467070e-01 -2.16080442e-01 -3.00678045e-01
3.12083036e-01 -3.91118318e-01 3.41855794e-01 -2.05566332e-01
1.13074112e+00 -2.03785610e+00 -3.72038111e-02 7.00157285e-01
2.25995392e-01 4.28433605e-02 2.57098109e-01 1.29377043e+00
-3.27812438e-03 2.06060633e-01 4.48620394e-02 2.15366051e-01
9.43941176e-01 7.07407892e-01 -3.83925468e-01 3.10420692e-01
-1.57262385e-01 4.10774112e-01 -8.21247756e-01 -6.19426705e-02
2.32908174e-01 2.66549468e-01 -2.75829643e-01 2.84646422e-01
-8.10184121e-01 4.10102725e-01 -5.85889399e-01 3.49623799e-01
5.60412943e-01 -9.64040980e-02 1.26154602e+00 5.49457550e-01
-4.67822194e-01 9.55430865e-01 -1.83828843e+00 1.17617965e+00
-7.35952735e-01 -1.83338717e-01 5.79858363e-01 -4.51241046e-01
7.85225034e-01 1.07586181e+00 2.20235527e-01 -5.88175774e-01
-2.03730479e-01 5.76826453e-01 3.00298750e-01 -4.19472009e-01
-1.64169997e-01 -2.02000946e-01 -2.24349797e-01 9.61156189e-01
-4.23899233e-01 1.35076836e-01 5.71815789e-01 4.38987501e-02
8.15473378e-01 6.01900220e-01 6.25334144e-01 -2.96724021e-01
1.06554151e+00 4.19507995e-02 6.07222974e-01 5.35279095e-01
3.55286956e-01 -4.30424988e-01 1.02384126e+00 -9.64312673e-01
-1.02067006e+00 -7.60703385e-01 1.79081172e-01 1.05657804e+00
1.33279040e-01 -1.01634431e+00 -4.86726463e-01 -4.05333102e-01
-3.12798113e-01 9.07383978e-01 -7.14660734e-02 3.02486748e-01
-6.66307330e-01 1.59253150e-01 2.96492040e-01 2.19362602e-01
2.21309662e-02 -1.19761169e+00 -1.35427141e+00 4.09029007e-01
1.92307517e-01 -1.26275170e+00 8.69542584e-02 -1.25999242e-01
-5.81249654e-01 -1.08040202e+00 3.83243054e-01 -3.72991085e-01
4.25695896e-01 -7.35730827e-02 1.16392171e+00 6.46189153e-01
2.87599504e-01 3.74972314e-01 -6.25475228e-01 -4.39291686e-01
-8.81261766e-01 -5.54102778e-01 -2.73566484e-01 2.38768414e-01
5.63443244e-01 -6.62957430e-01 -1.80347160e-01 5.45424342e-01
-1.40258455e+00 4.24833417e-01 -1.40279174e-01 1.48313716e-01
2.45486274e-01 2.90688068e-01 1.65534779e-01 -1.02010632e+00
4.69276458e-01 -1.80984184e-01 -1.09870315e+00 3.97249192e-01
-7.46376157e-01 1.19188271e-01 6.85733259e-01 -9.71215498e-03
-1.02566111e+00 -4.43517298e-01 -1.30480796e-01 5.84364176e-01
-6.86621606e-01 4.31222737e-01 -9.65071082e-01 1.40346006e-01
-9.29333456e-03 1.67124718e-01 -1.01994328e-01 -5.21222115e-01
6.84274137e-02 7.48548567e-01 5.90089820e-02 -9.26763654e-01
6.84292912e-01 8.11154902e-01 1.59151524e-01 -2.27998078e-01
4.65699852e-01 -4.97599006e-01 -1.56783059e-01 -1.21452399e-01
4.29087579e-01 -1.50717869e-01 -1.52596331e+00 1.05286911e-01
-1.17471802e+00 -4.87394869e-01 -5.34567714e-01 4.91472557e-02
-6.87157035e-01 1.46353468e-01 -5.30089617e-01 -1.37571990e+00
8.78614485e-02 -1.21232390e+00 9.10890281e-01 -4.50047493e-01
-6.23813748e-01 -7.13987887e-01 -1.30472332e-01 -1.77045330e-01
3.35612535e-01 6.08104244e-02 8.98505867e-01 -1.07282579e+00
-5.98257005e-01 -4.69852388e-01 -6.75406829e-02 2.95223631e-02
9.29397866e-02 2.97263861e-01 -4.90934789e-01 1.24704003e-01
1.70588315e-01 6.69952393e-01 -7.07312524e-01 -2.75641769e-01
4.20657009e-01 -4.47575450e-01 -1.86650217e-01 1.00182113e-03
1.70771182e+00 9.33131695e-01 9.91487741e-01 5.31290472e-01
-5.17419493e-03 9.14639413e-01 5.26248813e-01 2.32770294e-01
5.94590545e-01 1.16769528e+00 1.91937253e-01 2.25679383e-01
2.75498152e-01 -1.05073184e-01 3.66183907e-01 3.91557634e-01
-2.90286094e-01 3.15698892e-01 -1.34968579e+00 2.60606438e-01
-2.18825388e+00 -1.03042674e+00 -6.62224472e-01 2.03913093e+00
6.64005756e-01 1.54149495e-02 1.55563533e-01 3.32795799e-01
2.24921718e-01 -1.76361427e-01 4.09112453e-01 -1.21507537e+00
6.27071261e-01 -5.93994930e-02 2.41591558e-01 8.63725901e-01
-2.05712587e-01 5.17828703e-01 5.69712591e+00 8.77745897e-02
-7.95176923e-01 4.97951955e-01 -3.84930253e-01 4.80529487e-01
-7.49271512e-01 8.19659650e-01 -6.95628285e-01 2.71510035e-01
1.43820548e+00 -1.49072662e-01 8.18647206e-01 6.05900586e-01
8.83725405e-01 -1.07103273e-01 -1.55116844e+00 4.02785361e-01
-4.71834868e-01 -1.31800663e+00 -9.11400765e-02 1.24103375e-01
2.00455666e-01 -4.31304514e-01 -1.09018004e+00 -4.92112301e-02
4.24521893e-01 -7.21600413e-01 1.33928418e+00 7.86038637e-01
5.04262507e-01 -4.00112987e-01 8.18299592e-01 5.87270379e-01
-1.15401244e+00 -2.34317154e-01 3.28348130e-01 -3.91808271e-01
4.47245389e-01 7.77845159e-02 -6.16355538e-01 9.55341637e-01
5.12460172e-01 -2.05247983e-01 3.13441008e-01 7.38430321e-01
-2.52916813e-01 2.11816683e-01 -3.53679568e-01 -9.32587311e-02
7.76851252e-02 -6.71566606e-01 4.00641441e-01 1.29003620e+00
-1.49881005e-01 1.13880329e-01 1.22391872e-01 7.07501292e-01
6.63261890e-01 -1.75703093e-02 -3.52942526e-01 1.99795105e-02
1.62807673e-01 8.13655138e-01 -6.79121554e-01 -5.74425280e-01
-7.71456599e-01 3.44234377e-01 -6.14287794e-01 6.08777463e-01
-5.30067563e-01 -3.10726136e-01 5.68295062e-01 5.92879415e-01
2.01821595e-01 -3.77290577e-01 1.04198031e-01 -1.12430084e+00
4.12164032e-01 -1.31247497e+00 2.11601779e-01 -8.07490528e-01
-8.00073624e-01 6.16868317e-01 4.64159280e-01 -7.63541043e-01
-5.20586729e-01 -5.10761678e-01 -5.09431660e-01 9.49879110e-01
-1.14402008e+00 -1.21534407e+00 1.85887814e-01 5.70362926e-01
1.06245475e-02 3.82810533e-01 1.27667677e+00 2.49811158e-01
1.30816013e-01 -5.26739597e-01 -6.07919574e-01 -1.63391903e-01
-1.50081456e-01 -1.13747835e+00 4.23768967e-01 1.02435005e+00
1.34867951e-02 8.95842612e-01 1.13568294e+00 -4.64016706e-01
-1.60859859e+00 -3.78189474e-01 1.83897579e+00 -2.77368158e-01
9.71718431e-01 -5.93967855e-01 -7.31439054e-01 1.16107202e+00
2.47505143e-01 -1.97309151e-01 3.59015912e-01 -2.33310178e-01
-2.48680264e-01 -6.19836032e-01 -1.13129580e+00 7.34188676e-01
9.99631703e-01 -7.50778556e-01 -8.38846624e-01 1.44493669e-01
7.06053615e-01 4.26559001e-02 -1.03731453e+00 1.21114127e-01
6.76205873e-01 -9.90741849e-01 4.67603415e-01 -9.61375952e-01
-4.03072000e-01 -1.08586395e+00 -3.34053397e-01 -2.91741669e-01
1.86406925e-01 -8.81319523e-01 -7.27681369e-02 1.28547335e+00
3.76519144e-01 -1.21064234e+00 1.01706766e-01 1.38831973e+00
1.14868090e-01 -1.72238275e-01 -8.51841092e-01 -6.67497635e-01
-8.22263300e-01 -8.41332078e-01 1.37756848e+00 7.36345470e-01
6.74159586e-01 -2.92596340e-01 1.37025625e-01 2.16693327e-01
1.32574096e-01 2.12611347e-01 6.89737320e-01 -1.39863873e+00
-6.28486454e-01 -1.88297078e-01 -4.43081409e-01 -4.33518231e-01
-1.12484880e-01 -6.57794058e-01 1.99049879e-02 -2.04611349e+00
-4.83729810e-01 -4.16809201e-01 2.43941143e-01 5.68837941e-01
1.17844784e+00 -5.50369978e-01 2.60266870e-01 3.62605274e-01
-5.21302342e-01 -3.65668446e-01 8.85802448e-01 2.96454877e-01
-1.27397969e-01 3.35546374e-01 -3.68511528e-01 5.49606502e-01
4.71742958e-01 -6.36569977e-01 -5.65525055e-01 -2.36043110e-01
8.73535454e-01 4.13536966e-01 5.55095434e-01 -5.10723114e-01
2.97111422e-01 -9.49288845e-01 -7.53007352e-01 2.88951039e-01
-1.28015801e-01 -1.92078650e+00 1.34463882e+00 4.96848077e-01
-5.61660051e-01 2.06375077e-01 -3.56472075e-01 1.69817116e-02
-2.56906033e-01 -6.11961961e-01 1.36803061e-01 -3.33878338e-01
-7.90598333e-01 -7.37499520e-02 -8.53895128e-01 -6.23355508e-01
1.24259138e+00 -2.96321660e-01 6.58133253e-02 -2.01537684e-01
-1.10770333e+00 1.58172213e-02 8.11001122e-01 7.46786967e-02
2.57313818e-01 -8.09566081e-01 -6.37301803e-02 4.21186775e-01
1.79837257e-01 -2.37642944e-01 -4.45919633e-01 9.83584702e-01
-8.38656068e-01 5.21897078e-01 -1.33846268e-01 -2.98710048e-01
-1.11347914e+00 5.75554669e-01 6.81065321e-01 -1.19415261e-01
-6.23530805e-01 -2.39039928e-01 -2.07974866e-01 -4.97039378e-01
7.00388029e-02 -5.44342101e-01 -1.29692659e-01 -3.97877365e-01
1.01866686e+00 1.11374341e-01 2.89648622e-01 -4.77510840e-01
-7.11650312e-01 9.23935696e-02 4.74963307e-01 -8.81193221e-01
1.29976487e+00 -2.51798064e-01 -1.01636279e+00 6.14709854e-01
3.99147034e-01 2.59536088e-01 -6.73478007e-01 -3.94614227e-02
7.77218401e-01 -3.72960627e-01 -5.24618626e-01 -6.39804482e-01
-4.06133920e-01 4.32281584e-01 -3.88467722e-02 1.17932165e+00
8.42975914e-01 1.05817854e-01 2.19250858e-01 1.88157350e-01
1.04158247e+00 -8.59894335e-01 -1.06495571e+00 2.32810095e-01
8.16534758e-01 -2.13218763e-01 -4.43123817e-01 -6.49043202e-01
-5.83576262e-01 1.38132834e+00 -3.11773960e-02 9.11675096e-02
5.27278185e-01 7.74353087e-01 1.69606611e-01 -5.29681742e-01
-1.22769880e+00 -1.43856823e-01 -3.28261405e-01 6.20596111e-01
1.37703463e-01 3.45125526e-01 -1.19009233e+00 6.69688106e-01
4.99933213e-01 6.53017461e-01 7.18493462e-01 1.34724176e+00
-2.05943078e-01 -1.92843461e+00 -2.78824180e-01 -1.12362906e-01
-5.47492146e-01 1.79155231e-01 -2.14630276e-01 1.15574300e+00
3.59744251e-01 1.15682209e+00 2.08626822e-01 8.49474818e-02
7.31532395e-01 2.28839606e-01 2.73966342e-01 -9.05553937e-01
-8.40391219e-01 1.07949264e-01 9.00549293e-01 -6.61599755e-01
-5.69068670e-01 -6.47369385e-01 -1.66867077e+00 -2.68516153e-01
1.09215699e-01 9.13396955e-01 8.10045183e-01 1.25585544e+00
-3.52180526e-02 3.09565514e-01 3.97318929e-01 -2.17958286e-01
-6.24435425e-01 -3.75613034e-01 -7.87228703e-01 5.00479817e-01
8.01324472e-02 -1.10480137e-01 -3.08099002e-01 3.50139469e-01] | [8.634760856628418, 6.893792152404785] |
06934474-f5d0-457e-922f-14a1495d1ed6 | real-time-audio-visual-end-to-end-speech | 2303.07005 | null | https://arxiv.org/abs/2303.07005v1 | https://arxiv.org/pdf/2303.07005v1.pdf | Real-Time Audio-Visual End-to-End Speech Enhancement | Audio-visual speech enhancement (AV-SE) methods utilize auxiliary visual cues to enhance speakers' voices. Therefore, technically they should be able to outperform the audio-only speech enhancement (SE) methods. However, there are few works in the literature on an AV-SE system that can work in real time on a CPU. In this paper, we propose a low-latency real-time audio-visual end-to-end enhancement (AV-E3Net) model based on the recently proposed end-to-end enhancement network (E3Net). Our main contribution includes two aspects: 1) We employ a dense connection module to solve the performance degradation caused by the deep model structure. This module significantly improves the model's performance on the AV-SE task. 2) We propose a multi-stage gating-and-summation (GS) fusion module to merge audio and visual cues. Our results show that the proposed model provides better perceptual quality and intelligibility than the baseline E3net model with a negligible computational cost increase. | ['Huaming Wang', 'Sefik Emre Eskimez', 'ZiYi Yang', 'Min Tang', 'Hemin Yang', 'Zirun Zhu'] | 2023-03-13 | null | null | null | null | ['speech-enhancement'] | ['speech'] | [-6.97906613e-02 -1.81731552e-01 4.44328815e-01 -2.19164059e-01
-9.41423953e-01 2.09357813e-02 3.23099703e-01 -5.89024164e-02
-5.65188289e-01 3.65283281e-01 5.37127495e-01 -3.50089729e-01
2.42211401e-01 -4.53014165e-01 -3.47929209e-01 -5.90217948e-01
1.45979553e-01 -6.81351364e-01 5.33544660e-01 -4.82101232e-01
5.40277176e-02 2.62008309e-01 -1.81575620e+00 4.70995188e-01
9.54477549e-01 1.06538689e+00 6.60201788e-01 1.12486553e+00
7.36944601e-02 6.25104845e-01 -6.65044665e-01 -2.53026187e-01
-1.57693267e-01 -3.92795503e-01 -3.87306869e-01 1.58465222e-01
2.74892479e-01 -5.47297180e-01 -4.67524976e-01 1.05447626e+00
1.40244031e+00 3.86592388e-01 1.87006533e-01 -1.21160948e+00
-5.24303317e-01 2.83391804e-01 -6.45962179e-01 3.61499608e-01
2.18884796e-01 2.57772595e-01 7.58811831e-01 -1.19921124e+00
3.55594277e-01 1.30355060e+00 6.85911059e-01 6.26178682e-01
-9.70851064e-01 -5.39202690e-01 1.91612780e-01 7.25775659e-01
-1.20826304e+00 -9.72273171e-01 8.39519441e-01 1.21401712e-01
1.07145536e+00 3.47334892e-01 7.56661594e-01 9.50491011e-01
7.22440109e-02 9.01030004e-01 1.20183289e+00 -3.55624944e-01
3.14629257e-01 1.56630099e-01 -1.78010866e-01 4.10685211e-01
-4.42816198e-01 5.43539524e-01 -8.40966284e-01 2.67836273e-01
5.37850916e-01 -5.29474437e-01 -3.43304217e-01 2.68207401e-01
-6.54370487e-01 4.30034697e-01 5.08551598e-01 4.68046486e-01
-5.61579645e-01 2.20832035e-01 5.21073580e-01 4.66248274e-01
6.97829068e-01 -1.12915516e-01 -1.36542201e-01 -2.66520590e-01
-1.33328569e+00 7.43727461e-02 3.58032882e-01 5.72341084e-01
1.85993314e-01 5.78934789e-01 -4.28502560e-01 1.08349013e+00
6.48509502e-01 4.20391083e-01 2.78320342e-01 -8.27297211e-01
3.47038120e-01 -3.05594981e-01 -1.40063874e-02 -6.97162986e-01
-4.20323282e-01 -8.18897188e-01 -8.11963379e-01 7.25926220e-01
-3.13864738e-01 -2.87475377e-01 -7.33036816e-01 1.79048371e+00
3.77365619e-01 3.21616828e-01 9.51143727e-02 1.16444576e+00
1.36311960e+00 8.78538311e-01 3.40857893e-01 -3.01425368e-01
1.45562208e+00 -1.28113949e+00 -1.24516356e+00 4.72374968e-02
9.91465710e-03 -1.16668689e+00 1.02451670e+00 4.57839966e-01
-1.65086222e+00 -1.03913438e+00 -1.20718932e+00 -2.72934109e-01
-1.10509336e-01 2.06603974e-01 4.19232428e-01 9.53598261e-01
-1.96029782e+00 3.19704115e-01 -8.01907837e-01 -4.88146469e-02
6.12017699e-02 2.87986040e-01 -1.95300564e-01 4.17387992e-01
-1.31649482e+00 8.11783433e-01 2.45182291e-02 3.65732819e-01
-1.11273670e+00 -4.91809428e-01 -8.62722456e-01 3.76205742e-01
2.85491765e-01 -8.64270508e-01 1.54967403e+00 -1.01034558e+00
-1.87869811e+00 3.61864269e-01 -5.87643027e-01 -1.62556455e-01
3.37916255e-01 -1.75081670e-01 -7.18607605e-01 3.38273764e-01
-4.94013339e-01 9.53532398e-01 9.43918347e-01 -1.35026491e+00
-5.68594933e-01 -1.13924965e-01 -6.40927702e-02 3.55147749e-01
-3.93168718e-01 4.00626570e-01 -5.94877064e-01 -9.37427580e-01
6.29484057e-02 -4.30861026e-01 -1.20308243e-01 2.98733562e-01
-1.80153489e-01 -8.00404977e-03 8.15845191e-01 -1.20700049e+00
1.51028931e+00 -2.49405766e+00 -1.58589050e-01 -1.65531471e-01
3.47285599e-01 8.55794370e-01 -5.05055249e-01 2.14245871e-01
-8.35654065e-02 -1.31947801e-01 1.04149461e-01 -9.23452973e-01
8.17341506e-02 -2.15731740e-01 3.09456065e-02 1.64794892e-01
9.08129886e-02 7.18921959e-01 -6.29418731e-01 -5.94094753e-01
5.55313349e-01 1.20100987e+00 -7.25477874e-01 3.62648547e-01
2.75765657e-01 2.45845973e-01 1.37157872e-01 4.15214777e-01
1.07671678e+00 2.27019742e-01 -2.05609992e-01 -2.13570267e-01
-3.53250891e-01 4.30762172e-01 -1.38472676e+00 1.61939263e+00
-9.06161308e-01 8.75437379e-01 7.89487541e-01 -3.51267010e-01
8.51659358e-01 8.61707449e-01 5.84266484e-02 -8.23963940e-01
2.25677222e-01 2.22104341e-01 -6.32635923e-03 -5.04765570e-01
7.96422422e-01 -1.62952870e-01 6.73720300e-01 1.26867602e-02
1.78696632e-01 1.66796491e-01 -1.33223504e-01 1.65009812e-01
8.70516598e-01 -1.70231476e-01 1.05230622e-01 7.31926337e-02
8.70328069e-01 -7.83608973e-01 4.41411644e-01 5.10324478e-01
-7.15561330e-01 5.45712829e-01 1.74327224e-01 2.69278139e-01
-9.63187873e-01 -1.30621684e+00 1.49244025e-01 1.27478158e+00
1.53322116e-01 -5.11945844e-01 -9.13961172e-01 -8.30363855e-02
-7.33886778e-01 5.16494095e-01 -1.30674392e-01 -3.60636041e-02
-2.09636435e-01 -4.27880049e-01 5.42747200e-01 6.85039759e-01
7.90760040e-01 -1.27051842e+00 -4.15895522e-01 5.06515503e-01
-4.26665008e-01 -1.09377801e+00 -6.58870399e-01 -1.43554330e-01
-7.87518620e-01 -2.83674240e-01 -1.01395643e+00 -8.57658386e-01
2.89107800e-01 4.72454399e-01 6.69992268e-01 -4.95374985e-02
2.77554896e-02 2.20795944e-01 -3.86443436e-01 -2.46294647e-01
-4.32907432e-01 -3.17142010e-01 -1.73554108e-01 -3.95311378e-02
1.78222030e-01 -8.19486797e-01 -8.50246787e-01 1.49620011e-01
-7.88940668e-01 2.28456825e-01 5.59914052e-01 8.04585218e-01
5.25974512e-01 6.16161432e-03 9.10299897e-01 1.43933371e-02
1.03108931e+00 9.40228924e-02 -4.61843640e-01 1.90740693e-02
-4.70840842e-01 -2.51312941e-01 3.41672778e-01 -4.91267622e-01
-1.54053915e+00 -2.21226528e-01 -1.08272290e+00 -3.44031394e-01
-1.24417514e-01 2.88935453e-01 -4.51634020e-01 -1.27410010e-01
2.82829970e-01 1.60777360e-01 -1.28528357e-01 -7.37827063e-01
3.86040032e-01 1.16902733e+00 7.62275696e-01 -4.41766679e-02
3.41605932e-01 2.52597451e-01 -1.38339385e-01 -1.00796700e+00
-2.23587543e-01 -6.33223891e-01 -2.05480475e-02 -5.44080913e-01
9.13021505e-01 -1.29457986e+00 -8.96869898e-01 6.56002998e-01
-1.46671653e+00 -2.00745270e-01 -1.08217692e-03 7.67892540e-01
-3.76464546e-01 6.28311634e-01 -9.69601393e-01 -1.12328672e+00
-7.34666169e-01 -1.38921189e+00 1.04685497e+00 2.74739802e-01
1.89128846e-01 -7.35573292e-01 -1.30634867e-02 2.10801646e-01
9.64244902e-01 -3.38653237e-01 2.48052061e-01 5.35606183e-02
-2.84172863e-01 5.09605467e-01 -4.97355729e-01 7.17462420e-01
-1.87722832e-01 -1.07601441e-01 -1.64091778e+00 -3.94705743e-01
2.17132494e-01 7.52910748e-02 1.01604068e+00 8.27686727e-01
1.13622928e+00 -5.00456207e-02 1.91508368e-01 7.28545547e-01
1.22916424e+00 3.94720823e-01 9.96506035e-01 4.78692092e-02
2.07665324e-01 5.16901076e-01 5.52061915e-01 5.41164577e-01
2.28189066e-01 9.17629004e-01 4.13929880e-01 -8.10061157e-01
-9.10180449e-01 -2.55302638e-01 8.24151039e-01 1.33027422e+00
-1.13772005e-01 -4.60799009e-01 -2.93386281e-01 7.14306653e-01
-1.47682476e+00 -9.28143799e-01 -2.07638502e-01 1.83164680e+00
7.07244992e-01 4.39635590e-02 1.36856273e-01 5.18657029e-01
9.21668887e-01 3.86409104e-01 -1.64420903e-01 -7.84592390e-01
-1.05767481e-01 5.71395755e-01 -1.99040491e-02 9.24688220e-01
-1.06322622e+00 7.39261270e-01 6.02795982e+00 1.15584588e+00
-1.19014108e+00 6.75358891e-01 4.39962417e-01 -1.87689781e-01
-2.18792796e-01 -4.55998629e-01 -5.33148229e-01 2.21595213e-01
1.17825055e+00 7.53229335e-02 3.24529469e-01 6.96128488e-01
8.03188086e-01 1.29257143e-01 -5.08468688e-01 1.28502607e+00
2.70702150e-02 -8.38364959e-01 -3.10054123e-01 -1.18406639e-01
3.48140866e-01 -1.28551409e-01 3.33233237e-01 2.12918475e-01
-2.78815240e-01 -7.33385026e-01 8.01393747e-01 2.02000886e-01
1.05867660e+00 -9.66109812e-01 7.62904763e-01 -4.71347459e-02
-1.61521983e+00 1.30871877e-01 -2.13651255e-01 2.84385383e-01
6.39852703e-01 6.51498973e-01 -4.91518825e-01 4.93519187e-01
7.31112421e-01 2.07017004e-01 -3.34002644e-01 1.31675565e+00
-6.28869712e-01 6.49219275e-01 -7.36251241e-03 7.41260052e-02
2.57727236e-01 2.88500339e-01 9.60400701e-01 1.51667297e+00
4.81841683e-01 -4.95400839e-02 -3.89674217e-01 6.89490855e-01
8.14440136e-04 1.26037806e-01 -1.56559929e-01 2.33933911e-01
2.22870395e-01 1.24069965e+00 -2.78535157e-01 -2.60499567e-01
-4.28655148e-01 1.07685435e+00 -2.39563882e-01 4.29283708e-01
-9.85972703e-01 -8.27168345e-01 6.71029449e-01 3.52278980e-03
4.73998219e-01 -6.70855343e-02 -1.23937711e-01 -7.79274821e-01
1.70692369e-01 -8.93754065e-01 -1.21425852e-01 -1.28120494e+00
-8.25618505e-01 1.09397519e+00 -5.01467884e-01 -1.18346202e+00
-5.98247014e-02 -4.93560165e-01 -4.76261646e-01 1.00488997e+00
-1.90032709e+00 -9.72393394e-01 -4.26975191e-01 8.09782803e-01
5.90628743e-01 9.08147618e-02 5.43357790e-01 9.43723261e-01
-1.76474631e-01 9.11209106e-01 -1.08418696e-01 -2.61832684e-01
7.71202326e-01 -1.04011476e+00 4.10955042e-01 1.14149594e+00
-8.84116739e-02 2.20409602e-01 7.98257649e-01 -3.60168695e-01
-7.65223086e-01 -1.00872779e+00 1.22859800e+00 5.20540893e-01
2.67151684e-01 -3.71323854e-01 -9.06946719e-01 1.51926547e-01
7.14251041e-01 -8.72537792e-02 5.99875748e-01 -7.70854950e-02
-2.03353807e-01 -2.94061244e-01 -1.24278653e+00 6.42711282e-01
9.64391768e-01 -7.83364832e-01 -4.95615393e-01 -3.27180624e-01
1.02596629e+00 -3.05792749e-01 -6.48583770e-01 1.72155172e-01
3.51314396e-01 -1.08511412e+00 1.02765727e+00 -8.73326361e-02
4.05962020e-01 -6.19041800e-01 -9.43368673e-02 -1.50013793e+00
-1.92925557e-01 -1.00947475e+00 -1.98795140e-01 1.39132059e+00
2.90613294e-01 -4.21854377e-01 1.77266970e-01 -3.71851888e-03
-5.23394942e-01 -5.46551585e-01 -1.12117350e+00 -7.14779913e-01
-5.20298183e-01 -7.79554486e-01 5.58395028e-01 3.74836892e-01
1.65780087e-03 2.96325803e-01 -7.17709422e-01 2.22961441e-01
3.93888414e-01 -4.65115249e-01 3.07615399e-01 -7.42766023e-01
-6.20512426e-01 -3.91057789e-01 -3.87323916e-01 -1.35742414e+00
-2.41691425e-01 -4.84744877e-01 3.21078867e-01 -1.71591854e+00
8.09343606e-02 1.13841176e-01 -4.68053401e-01 2.38461301e-01
-3.80806476e-01 1.59848452e-01 3.63502651e-01 -4.82894570e-01
-6.66656196e-01 1.00588524e+00 1.46406293e+00 9.04762596e-02
-4.98825699e-01 -8.44455957e-02 -6.57176256e-01 4.65147734e-01
5.83319128e-01 -2.39011407e-01 -4.74244267e-01 -4.77091163e-01
-7.34432563e-02 4.19073433e-01 5.24243891e-01 -1.07047284e+00
4.83421177e-01 3.48423660e-01 1.14609070e-01 -7.74151981e-01
7.83355832e-01 -6.01637900e-01 -8.91000107e-02 3.85709584e-01
-3.03881377e-01 1.02350317e-01 5.33666372e-01 3.17151278e-01
-7.24673927e-01 -1.05090231e-01 1.11529315e+00 3.80190521e-01
-7.14557171e-01 1.19027644e-01 -7.49412656e-01 -3.12214941e-01
6.77299619e-01 6.01554476e-03 -3.94518703e-01 -7.51801193e-01
-1.05400550e+00 -9.87992883e-02 -1.36582494e-01 2.66751528e-01
1.04555929e+00 -1.31939065e+00 -8.70872498e-01 5.85537814e-02
-3.16018552e-01 -6.94370210e-01 9.20667708e-01 1.01446652e+00
-3.97426784e-01 2.98236430e-01 -2.54847586e-01 -3.97899508e-01
-1.78703094e+00 5.08139253e-01 4.23181415e-01 -2.18081161e-01
-5.22529244e-01 9.73992169e-01 1.74922228e-01 -2.09622514e-02
6.13788009e-01 -2.92523950e-02 -3.33700031e-01 -2.15176523e-01
1.03160632e+00 7.20835268e-01 1.91690117e-01 -6.06330752e-01
-3.15993190e-01 1.90150142e-01 1.10843480e-01 -7.64597714e-01
1.44047403e+00 -4.42756295e-01 1.82840854e-01 -1.00575192e-02
1.00870955e+00 1.65560141e-01 -1.18666923e+00 -8.68316144e-02
-7.12217927e-01 -5.05875587e-01 7.97055304e-01 -9.37718570e-01
-1.34345257e+00 1.59553230e+00 1.33631063e+00 -1.43507972e-01
1.81528056e+00 -3.48164469e-01 9.51348782e-01 -1.98646143e-01
-1.65529221e-01 -1.20838332e+00 2.25736260e-01 2.52567172e-01
1.09204054e+00 -9.91286933e-01 -4.08487797e-01 -6.26223028e-01
-6.69068098e-01 9.41471875e-01 5.02163172e-01 4.29740906e-01
6.91346645e-01 4.78688300e-01 1.93628117e-01 1.53412610e-01
-8.90490234e-01 -6.79938138e-01 4.13672388e-01 8.80927026e-01
5.88738799e-01 -8.13683197e-02 -4.85776395e-01 7.62263894e-01
-4.30749804e-02 4.25789766e-02 2.59735256e-01 5.42185187e-01
-5.33032775e-01 -9.94633853e-01 -4.48842525e-01 -3.48823249e-01
-6.10274673e-01 -4.51086164e-01 -4.34442237e-02 4.58633989e-01
2.58389086e-01 1.74669147e+00 -9.88152176e-02 -6.40841663e-01
4.32210684e-01 -3.04684967e-01 3.51362288e-01 -4.63081747e-02
-1.02755797e+00 7.63807476e-01 1.58384934e-01 -6.05245471e-01
-4.78456527e-01 -2.33429790e-01 -1.04488504e+00 -5.52816391e-01
-3.67202610e-01 -4.37239893e-02 1.03463626e+00 5.33194065e-01
5.63298106e-01 1.24730539e+00 6.74764395e-01 -7.90304601e-01
-2.93863267e-01 -1.08766460e+00 -7.05800951e-01 1.55632719e-01
5.89131474e-01 -4.78778273e-01 -4.99860287e-01 -1.61640227e-01] | [14.808537483215332, 5.828755855560303] |
15256d7f-ca7a-4541-a1ad-8b81c675bcd7 | learning-invariance-from-generated-variance | 2301.00725 | null | https://arxiv.org/abs/2301.00725v1 | https://arxiv.org/pdf/2301.00725v1.pdf | Learning Invariance from Generated Variance for Unsupervised Person Re-identification | This work focuses on unsupervised representation learning in person re-identification (ReID). Recent self-supervised contrastive learning methods learn invariance by maximizing the representation similarity between two augmented views of a same image. However, traditional data augmentation may bring to the fore undesirable distortions on identity features, which is not always favorable in id-sensitive ReID tasks. In this paper, we propose to replace traditional data augmentation with a generative adversarial network (GAN) that is targeted to generate augmented views for contrastive learning. A 3D mesh guided person image generator is proposed to disentangle a person image into id-related and id-unrelated features. Deviating from previous GAN-based ReID methods that only work in id-unrelated space (pose and camera style), we conduct GAN-based augmentation on both id-unrelated and id-related features. We further propose specific contrastive losses to help our network learn invariance from id-unrelated and id-related augmentations. By jointly training the generative and the contrastive modules, our method achieves new state-of-the-art unsupervised person ReID performance on mainstream large-scale benchmarks. | ['Francois Bremond', 'Antitza Dantcheva', 'Benoit Lagadec', 'Yaohui Wang', 'Hao Chen'] | 2023-01-02 | null | null | null | null | ['person-re-identification', 'unsupervised-person-re-identification'] | ['computer-vision', 'computer-vision'] | [ 3.55421305e-01 1.27538383e-01 1.67221546e-01 -6.56173289e-01
-5.98005593e-01 -6.62114799e-01 1.07962692e+00 -4.70507026e-01
-4.12898183e-01 7.08958685e-01 5.60306728e-01 3.53002071e-01
1.27576470e-01 -7.14039207e-01 -7.77062058e-01 -6.85347617e-01
2.41978988e-01 6.81145668e-01 -8.41881156e-01 -2.19965264e-01
-6.21155240e-02 5.41107297e-01 -1.58816004e+00 -9.84893143e-02
8.77796113e-01 4.56402421e-01 -4.20471936e-01 5.44399083e-01
2.64898092e-01 7.15293288e-01 -6.68001592e-01 -9.31642950e-01
8.15152586e-01 -8.43511343e-01 -8.40707541e-01 6.16751686e-02
1.10222065e+00 -4.95847046e-01 -7.71883249e-01 1.03316438e+00
7.72764206e-01 2.95218796e-01 1.00427020e+00 -1.56244421e+00
-1.29511893e+00 2.72935003e-01 -8.91347468e-01 2.81179458e-01
6.55419052e-01 1.38599917e-01 4.55254108e-01 -6.84740305e-01
7.09676445e-01 1.59153593e+00 7.40323901e-01 1.17627084e+00
-1.44908273e+00 -1.00478733e+00 1.61664903e-01 8.46865624e-02
-1.39623499e+00 -4.94559437e-01 1.08919454e+00 -4.16783810e-01
3.79499286e-01 2.62201339e-01 6.07942581e-01 1.71999717e+00
-1.27806649e-01 7.38468170e-01 1.37559605e+00 -2.42352530e-01
-1.40467718e-01 1.90510154e-02 -1.46316424e-01 3.69653702e-01
4.03436333e-01 5.05732477e-01 -4.64384556e-01 3.12639847e-02
1.01950562e+00 2.25960955e-01 6.81946725e-02 -5.47278643e-01
-1.16742456e+00 7.62441874e-01 6.48716390e-01 -2.17849568e-01
-2.44978622e-01 -2.81949434e-02 3.31129402e-01 4.18142915e-01
4.92419541e-01 5.13626456e-01 1.42429441e-01 2.38630697e-01
-6.30390465e-01 6.01652086e-01 3.73582006e-01 1.11173022e+00
7.86228955e-01 1.80558547e-01 -5.43550789e-01 8.09018016e-01
1.89786274e-02 6.39325976e-01 3.81831050e-01 -7.71907806e-01
3.86196852e-01 5.72613418e-01 -7.99358040e-02 -6.68834627e-01
-1.25175640e-01 -7.10323513e-01 -1.35772169e+00 2.93730617e-01
4.72474903e-01 -2.03243017e-01 -1.09961641e+00 2.22215676e+00
2.84403026e-01 1.82460040e-01 1.32610396e-01 9.04065669e-01
1.00569785e+00 1.40696436e-01 1.95175007e-01 2.54008383e-01
1.31725562e+00 -8.93686056e-01 -5.39199352e-01 -2.99273372e-01
1.56369567e-01 -4.94110197e-01 7.69362807e-01 -1.86802968e-01
-1.21348941e+00 -9.37486947e-01 -1.00715029e+00 -2.47983620e-01
-4.72965121e-01 -7.17000589e-02 4.78921294e-01 8.25356126e-01
-1.06692004e+00 2.06140414e-01 -2.95851797e-01 -4.27853972e-01
6.73086047e-01 4.95086581e-01 -8.75542641e-01 -2.02809319e-01
-1.10604036e+00 8.29634607e-01 -1.06683649e-01 -8.39151666e-02
-1.00100112e+00 -8.52553189e-01 -1.17586243e+00 -2.47058064e-01
-9.94059369e-02 -1.40757787e+00 7.12494850e-01 -1.17801726e+00
-1.22724867e+00 1.48624408e+00 -1.43761754e-01 -3.35588574e-01
9.64497566e-01 -1.74394757e-01 -2.88318038e-01 -8.50053951e-02
3.89764994e-01 1.08059180e+00 1.21585941e+00 -1.62609756e+00
-1.72065333e-01 -7.35022604e-01 1.27230570e-01 4.66410249e-01
-3.08266461e-01 -3.35682034e-02 -1.50437415e-01 -1.10575140e+00
-1.89623654e-01 -9.99462843e-01 1.21452613e-02 -1.92200188e-02
-6.22996330e-01 -6.93065971e-02 6.91838741e-01 -8.10973704e-01
1.84092402e-01 -1.89978063e+00 3.16778570e-01 3.60055268e-03
2.51696616e-01 -2.38111950e-02 -4.30989236e-01 3.00318818e-03
-4.53959018e-01 -9.20764208e-02 -6.37629777e-02 -8.95558238e-01
1.65391136e-02 1.62253436e-02 -1.27852678e-01 7.47951448e-01
2.91172862e-01 1.26670957e+00 -9.59332466e-01 -3.53197873e-01
2.67345130e-01 7.39839792e-01 -5.83899200e-01 5.25691807e-01
5.09024441e-01 1.19020343e+00 -1.19965181e-01 4.86925483e-01
1.07047069e+00 1.51783094e-01 -2.74894148e-01 -1.79183960e-01
2.72486418e-01 -6.68907389e-02 -7.38798261e-01 1.87176383e+00
-3.41613531e-01 4.09047186e-01 -2.94188052e-01 -1.05414522e+00
9.78450179e-01 6.69099391e-03 3.85272056e-01 -8.74078274e-01
5.92227280e-02 -3.66260022e-01 -2.03749642e-01 1.76146626e-02
5.17170489e-01 -1.58765316e-01 -1.57626927e-01 5.73774815e-01
3.00700665e-01 2.33454570e-01 -2.19982460e-01 1.50310233e-01
7.16611862e-01 3.46885324e-01 4.37998883e-02 -3.43097985e-01
6.69367969e-01 -4.71897751e-01 5.06066978e-01 8.29835653e-01
-1.83288783e-01 1.06398070e+00 8.40117112e-02 -5.00748158e-01
-1.36866820e+00 -1.45262241e+00 -1.04327686e-01 9.81606662e-01
5.06725609e-01 6.22353666e-02 -8.78756642e-01 -1.02564502e+00
1.33526042e-01 4.69020933e-01 -1.21131718e+00 -4.44731504e-01
-6.56634629e-01 -7.01017201e-01 7.48241544e-01 7.63214529e-01
8.36009502e-01 -7.06351519e-01 2.41524786e-01 -2.35635132e-01
-2.33439773e-01 -9.50912952e-01 -9.59714770e-01 -3.65256280e-01
-3.97333384e-01 -8.95916343e-01 -1.26036942e+00 -8.88219178e-01
1.30815196e+00 4.31648225e-01 1.06576276e+00 -2.55120367e-01
-3.44107538e-01 8.16383958e-01 -1.89039320e-01 -5.18478692e-01
-3.40431929e-01 4.43516951e-03 4.20531660e-01 2.87908465e-01
3.83507341e-01 -7.10936427e-01 -8.66694272e-01 4.97127771e-01
-5.89026392e-01 9.77063775e-02 3.60931128e-01 1.16042113e+00
4.04055625e-01 -3.93604994e-01 8.30446780e-01 -6.99320734e-01
3.45524132e-01 -3.33927304e-01 -1.13978051e-01 1.32394150e-01
-5.03836751e-01 6.61977977e-02 6.17125928e-01 -5.56546867e-01
-1.31945503e+00 -3.30002531e-02 8.96870643e-02 -6.41534328e-01
-3.21950108e-01 -3.85072947e-01 -5.22153795e-01 -2.12990820e-01
7.93468535e-01 4.25278336e-01 2.10229263e-01 -3.55680525e-01
4.76239532e-01 3.89732599e-01 1.09036243e+00 -6.16869748e-01
1.45090628e+00 6.80535853e-01 -2.37159561e-02 -3.72970164e-01
-7.05162168e-01 -3.24227691e-01 -8.71597111e-01 -1.34568438e-01
9.55132663e-01 -1.30400753e+00 -6.64954245e-01 7.04832613e-01
-9.01777089e-01 3.43450904e-02 -6.83249831e-01 2.56767869e-01
-6.92721367e-01 3.45069110e-01 -3.32395107e-01 -5.49551964e-01
-6.17804766e-01 -8.17956328e-01 1.12810576e+00 4.67530102e-01
-2.70559132e-01 -8.69273543e-01 1.70648038e-01 6.79662228e-01
3.93476188e-01 4.89275962e-01 4.27912652e-01 -6.90392256e-01
-2.71041006e-01 -4.46008503e-01 -3.25305372e-01 3.53256404e-01
4.86548126e-01 -8.31628382e-01 -1.24049556e+00 -8.05492759e-01
-2.41452187e-01 -3.94208282e-01 8.63994300e-01 1.03768490e-01
1.18771803e+00 -5.46709418e-01 -2.67639101e-01 1.16265345e+00
1.01690710e+00 -8.98751244e-02 8.07210326e-01 4.13588464e-01
1.14615929e+00 6.66934431e-01 1.84996367e-01 2.43011028e-01
6.00296855e-01 7.61676550e-01 1.70066461e-01 -5.07756650e-01
-4.63914186e-01 -7.32706726e-01 1.80724800e-01 -5.72672300e-02
-5.01069069e-01 -2.16549203e-01 -4.18449968e-01 5.95153511e-01
-1.53501630e+00 -1.17831969e+00 4.30401951e-01 2.30784464e+00
6.29592538e-01 -3.31584632e-01 5.15261412e-01 -2.33842090e-01
1.09153402e+00 5.97585700e-02 -8.81374896e-01 1.51050752e-02
-3.28739613e-01 1.43520787e-01 4.91177440e-01 1.11720204e-01
-1.26042497e+00 6.86792731e-01 5.78151417e+00 3.95726532e-01
-7.11783648e-01 1.71504766e-01 7.25463986e-01 -2.13358119e-01
-4.68455911e-01 -3.55353415e-01 -6.41070783e-01 5.10519266e-01
3.62717241e-01 -4.26118433e-01 4.19693738e-01 8.51779103e-01
-1.89587086e-01 5.04620612e-01 -1.53859019e+00 1.42365980e+00
6.00127578e-01 -9.98085439e-01 4.25201148e-01 3.38229448e-01
1.10930884e+00 -5.85456371e-01 5.87602258e-01 3.01947445e-01
3.96689475e-01 -1.33090353e+00 5.91773748e-01 6.99553907e-01
1.17681479e+00 -1.06514716e+00 8.50908875e-01 -1.46672383e-01
-1.01319444e+00 1.10419504e-01 -2.40330145e-01 1.32929981e-01
3.23751345e-02 -8.60174522e-02 -4.17444915e-01 7.54241645e-01
7.07297862e-01 8.61430168e-01 -6.78757608e-01 5.90523601e-01
-1.84595227e-01 -1.70878749e-02 3.35270017e-02 6.50281310e-01
-2.73338377e-01 -2.18529329e-01 6.54295802e-01 9.34691966e-01
1.28049310e-02 -4.57439050e-02 -1.10617682e-01 1.21374679e+00
-5.71980000e-01 -3.15999240e-01 -8.41354549e-01 4.67827827e-01
3.23679507e-01 9.98060763e-01 -1.39120102e-01 -3.19606394e-01
-3.15318495e-01 1.61144996e+00 3.70694906e-01 4.05139685e-01
-6.92466557e-01 -6.77863583e-02 1.06460786e+00 2.54037291e-01
2.75520999e-02 1.15514241e-01 -1.89026222e-01 -1.38746107e+00
1.52863041e-01 -8.28207135e-01 4.64365989e-01 -5.19922674e-01
-2.09033728e+00 4.37431514e-01 8.91965479e-02 -1.36405849e+00
-6.23145044e-01 -1.09726235e-01 -7.95730591e-01 1.25848222e+00
-1.43032146e+00 -1.85444415e+00 -6.16689563e-01 1.01520753e+00
4.01146114e-01 -5.68791211e-01 8.38225305e-01 2.25807279e-01
-5.41744232e-01 1.56993794e+00 -3.15235406e-02 6.25537097e-01
1.12871540e+00 -1.42113376e+00 1.03389120e+00 9.65536237e-01
-1.29191786e-01 8.89833272e-01 4.93099064e-01 -6.33351386e-01
-1.41126323e+00 -1.30784810e+00 4.30783123e-01 -8.25208545e-01
-3.31214592e-02 -5.97164273e-01 -6.76498473e-01 9.76938486e-01
1.76838979e-01 2.69409388e-01 6.39624178e-01 5.22689475e-03
-7.98366606e-01 -1.89535201e-01 -1.63084579e+00 5.50529182e-01
1.66794395e+00 -7.39931107e-01 -6.01669967e-01 1.35728851e-01
3.00339073e-01 -3.62180203e-01 -8.29389930e-01 4.10128683e-01
7.09764659e-01 -7.44297802e-01 1.51528442e+00 -8.53008449e-01
4.28963751e-01 -1.46507159e-01 2.51120746e-01 -1.37229455e+00
-5.47777057e-01 -5.67256331e-01 3.46073397e-02 1.54423010e+00
-2.11916313e-01 -8.32752228e-01 9.53895867e-01 6.51370943e-01
1.25977844e-01 -2.32042372e-02 -8.13933432e-01 -1.05928946e+00
3.97057950e-01 3.45624983e-01 8.53023589e-01 1.24008620e+00
-3.94799918e-01 2.84103632e-01 -7.27515280e-01 1.77881926e-01
1.35075366e+00 -4.31571305e-02 1.25632620e+00 -1.20699632e+00
-8.21934417e-02 -2.82833487e-01 -8.86913538e-01 -8.55809569e-01
5.15749574e-01 -1.05810142e+00 -2.50483751e-01 -1.24817741e+00
7.81528115e-01 -3.08084965e-01 -2.47800931e-01 3.95139456e-01
-4.67788190e-01 6.42737031e-01 1.75425112e-01 2.56330818e-01
-1.91725373e-01 7.89246798e-01 1.38135326e+00 -6.31871283e-01
1.04673170e-02 -8.37938190e-02 -1.09505391e+00 3.79026651e-01
6.16120398e-01 -2.07746103e-01 -5.19956708e-01 -3.22847724e-01
-2.68758625e-01 -4.28018302e-01 9.40552711e-01 -9.12367523e-01
2.26941511e-01 1.99880898e-01 1.16574621e+00 -3.48465383e-01
4.36117649e-01 -4.40106362e-01 2.47722372e-01 1.28825843e-01
-4.91958946e-01 6.73300698e-02 -1.46956872e-02 5.78812540e-01
-7.09161237e-02 2.13726476e-01 8.83225739e-01 -1.40426338e-01
-4.84806687e-01 7.36663878e-01 1.36377245e-01 2.18236551e-01
8.31867456e-01 -3.59207571e-01 -4.57563967e-01 -6.64525449e-01
-6.14271283e-01 8.95468071e-02 9.40707803e-01 7.36501038e-01
6.17556691e-01 -1.67116570e+00 -1.21051407e+00 5.68424523e-01
4.04627919e-01 -5.73404804e-02 5.83445609e-01 2.65371680e-01
-4.91343588e-02 1.73000935e-02 -7.71233797e-01 -4.68314409e-01
-1.21732366e+00 8.21600974e-01 5.24049938e-01 -6.55717254e-02
-6.81137562e-01 9.10574019e-01 8.82017314e-01 -7.02438772e-01
6.49002672e-04 6.69095755e-01 -2.07126498e-01 -1.18820123e-01
7.06748247e-01 4.65159208e-01 -2.35744193e-01 -1.06868136e+00
-2.95781732e-01 6.97627962e-01 -4.62636530e-01 -1.73768364e-02
1.12027597e+00 -1.84749454e-01 1.40814513e-01 -1.83360234e-01
1.29824126e+00 -1.21763356e-01 -1.41492212e+00 -2.28150979e-01
-7.09688067e-01 -6.45491898e-01 -3.50258082e-01 -7.62545049e-01
-1.07325613e+00 4.49929386e-01 9.14901793e-01 -3.27076077e-01
1.00049472e+00 1.51804671e-01 5.66905975e-01 4.97239977e-02
4.55210000e-01 -8.01122367e-01 2.42121041e-01 7.74973929e-02
1.09521198e+00 -1.51020229e+00 3.35152261e-02 -3.86506647e-01
-5.25012970e-01 5.88297725e-01 9.39208269e-01 -3.25528413e-01
1.75056860e-01 -5.54515049e-02 -5.82502745e-02 3.77455540e-02
4.40374762e-02 -1.95685297e-01 6.39032722e-01 1.24586391e+00
1.63610384e-01 8.77476633e-02 7.89666101e-02 4.54630524e-01
-6.49282992e-01 -3.99325371e-01 2.67700881e-01 6.78341210e-01
5.24364650e-01 -1.30994916e+00 -5.91681242e-01 4.36009288e-01
-1.30044937e-01 9.28164572e-02 -6.74530685e-01 8.82405579e-01
2.52176970e-01 7.31883764e-01 4.46330547e-01 -4.36471820e-01
3.83222491e-01 -1.84041888e-01 9.57001448e-01 -3.72493953e-01
-4.53237355e-01 -6.19756222e-01 -2.37211853e-01 -2.99211323e-01
-5.50180197e-01 -9.45304573e-01 -6.34254098e-01 -5.81846058e-01
1.26077175e-01 -2.03373045e-01 3.70958209e-01 8.17136765e-01
3.63502204e-01 3.06185275e-01 8.71092796e-01 -1.03943789e+00
-3.57235014e-01 -1.01832068e+00 -4.67582256e-01 1.08036447e+00
4.02378708e-01 -8.31798375e-01 -3.53630871e-01 1.37849599e-01] | [14.671342849731445, 0.9199724793434143] |
b88c7638-537a-403f-99ab-9a63c30977aa | a-simple-approach-to-improve-single-model | 2205.00403 | null | https://arxiv.org/abs/2205.00403v2 | https://arxiv.org/pdf/2205.00403v2.pdf | A Simple Approach to Improve Single-Model Deep Uncertainty via Distance-Awareness | Accurate uncertainty quantification is a major challenge in deep learning, as neural networks can make overconfident errors and assign high confidence predictions to out-of-distribution (OOD) inputs. The most popular approaches to estimate predictive uncertainty in deep learning are methods that combine predictions from multiple neural networks, such as Bayesian neural networks (BNNs) and deep ensembles. However their practicality in real-time, industrial-scale applications are limited due to the high memory and computational cost. Furthermore, ensembles and BNNs do not necessarily fix all the issues with the underlying member networks. In this work, we study principled approaches to improve uncertainty property of a single network, based on a single, deterministic representation. By formalizing the uncertainty quantification as a minimax learning problem, we first identify distance awareness, i.e., the model's ability to quantify the distance of a testing example from the training data, as a necessary condition for a DNN to achieve high-quality (i.e., minimax optimal) uncertainty estimation. We then propose Spectral-normalized Neural Gaussian Process (SNGP), a simple method that improves the distance-awareness ability of modern DNNs with two simple changes: (1) applying spectral normalization to hidden weights to enforce bi-Lipschitz smoothness in representations and (2) replacing the last output layer with a Gaussian process layer. On a suite of vision and language understanding benchmarks, SNGP outperforms other single-model approaches in prediction, calibration and out-of-domain detection. Furthermore, SNGP provides complementary benefits to popular techniques such as deep ensembles and data augmentation, making it a simple and scalable building block for probabilistic deep learning. Code is open-sourced at https://github.com/google/uncertainty-baselines | ['Balaji Lakshminarayanan', 'Dustin Tran', 'Jasper Snoek', 'Zack Nado', 'Ghassen Jerfel', 'Yeming Wen', 'Zi Lin', 'Jie Ren', 'Shreyas Padhy', 'Jeremiah Zhe Liu'] | 2022-05-01 | null | null | null | null | ['probabilistic-deep-learning'] | ['computer-vision'] | [-1.44945472e-01 1.34763181e-01 -1.85318999e-02 -6.44452810e-01
-9.86437082e-01 -5.51774204e-01 5.95102489e-01 -3.81614864e-02
-3.17148924e-01 9.26194549e-01 -8.12768191e-02 -3.17026764e-01
-3.42152625e-01 -8.81516039e-01 -1.07228863e+00 -8.78107011e-01
1.84065327e-01 6.07563734e-01 1.56229213e-01 3.56015563e-01
-3.81022505e-02 4.49910581e-01 -1.35915852e+00 -5.66332787e-02
9.99410212e-01 1.54244339e+00 4.50013690e-02 3.09385836e-01
-1.97813511e-01 5.72243333e-01 -5.59220791e-01 -5.36852717e-01
1.28962785e-01 1.35338344e-02 -3.69994938e-01 -5.17624557e-01
3.78567934e-01 -4.55350995e-01 -3.09969068e-01 1.36029124e+00
5.52741289e-01 3.19614261e-01 1.02392936e+00 -1.42283571e+00
-6.08843505e-01 1.04834926e+00 -5.53836048e-01 3.75638865e-02
-4.01118428e-01 -6.35885447e-02 9.54937875e-01 -9.28036094e-01
-5.84897846e-02 1.37617803e+00 8.87308896e-01 5.79005003e-01
-1.38414419e+00 -9.31095660e-01 1.66954905e-01 -1.05968907e-01
-1.45675588e+00 -4.75939959e-01 5.41607082e-01 -5.24894416e-01
6.19843960e-01 -1.77475139e-01 1.29585043e-02 1.44970024e+00
3.99024129e-01 7.74322450e-01 7.97704518e-01 -3.28151323e-02
6.51078701e-01 5.77015430e-02 4.41419594e-02 3.48411620e-01
3.54682893e-01 2.99826354e-01 -3.87556583e-01 -2.39341527e-01
8.63160014e-01 1.08946659e-01 -2.44733214e-01 -4.02909696e-01
-8.72423112e-01 8.10112476e-01 4.51276779e-01 2.98824571e-02
-2.52139539e-01 5.67458749e-01 2.40612298e-01 -1.74879730e-01
5.91901898e-01 2.20879808e-01 -5.84730566e-01 -2.04269022e-01
-9.66447651e-01 1.82755351e-01 8.06397676e-01 7.98490644e-01
6.48061514e-01 1.46108031e-01 -3.20459545e-01 7.91034162e-01
6.24529004e-01 5.49839079e-01 2.65214831e-01 -1.02447188e+00
3.89509797e-01 1.13548607e-01 4.06075977e-02 -8.29097509e-01
-2.02879846e-01 -6.82056069e-01 -1.16610324e+00 4.29684013e-01
4.14368004e-01 -4.24542814e-01 -1.02384710e+00 1.99025643e+00
1.31890148e-01 5.42397857e-01 1.10571116e-01 6.61573827e-01
6.55349076e-01 5.84114969e-01 1.26745269e-01 1.09410606e-01
9.63372707e-01 -6.95291400e-01 -4.18696254e-01 -2.01004386e-01
3.51592034e-01 -3.82854491e-01 6.50511682e-01 4.43953842e-01
-7.23451912e-01 -5.84739566e-01 -1.12669063e+00 2.59842694e-01
-2.79961497e-01 1.73193514e-02 3.96069109e-01 5.99728048e-01
-8.13385367e-01 1.13512015e+00 -1.07792425e+00 3.05992305e-01
7.73745716e-01 2.71705449e-01 -2.75607020e-01 -7.83113912e-02
-1.27587688e+00 8.85086596e-01 8.94132912e-01 1.80443257e-01
-1.10623705e+00 -8.85360181e-01 -8.73616338e-01 1.54700980e-01
3.00855577e-01 -5.93223572e-01 1.29747808e+00 -7.90996432e-01
-1.51645947e+00 2.70036876e-01 4.90124151e-02 -6.92061126e-01
7.46808052e-01 -5.91189802e-01 -2.19837666e-01 -3.22770238e-01
-1.59787998e-01 7.78444350e-01 9.87798750e-01 -1.30363107e+00
-5.69969177e-01 -3.48654985e-01 -2.52477884e-01 -4.84665036e-02
-1.07550278e-01 -2.92987972e-01 -3.21365684e-01 -5.80359638e-01
3.59133065e-01 -7.99123466e-01 -2.64897853e-01 9.44306403e-02
-6.77903414e-01 -4.55889851e-01 6.14388645e-01 -5.46773612e-01
9.43674743e-01 -2.10803032e+00 -3.57351005e-02 4.52951491e-01
3.88967246e-01 1.81253478e-01 2.99039334e-02 -1.73087969e-01
6.17385320e-02 2.97108233e-01 -4.66496289e-01 -6.49140239e-01
4.09367383e-01 4.54890698e-01 -4.82734919e-01 3.59845042e-01
4.06902939e-01 8.35604429e-01 -9.24096465e-01 -2.95889258e-01
1.89782917e-01 7.78819799e-01 -1.55738547e-01 -3.16770226e-02
-4.22517180e-01 2.86464006e-01 -2.87424743e-01 4.81927991e-01
7.98852444e-01 -2.86845058e-01 -2.28162512e-01 -1.57968342e-01
2.85628408e-01 2.19954431e-01 -1.60973644e+00 1.45334482e+00
-4.28433418e-01 5.05460858e-01 -2.57437021e-01 -8.40794981e-01
1.05162847e+00 2.16355309e-01 2.57074177e-01 -2.17701234e-02
3.41142654e-01 3.29554260e-01 -1.55342489e-01 2.67463565e-01
2.33741015e-01 -8.59356746e-02 -2.01915540e-02 3.31620544e-01
4.67461765e-01 -1.38702884e-01 -6.24037236e-02 2.91319024e-02
8.64696622e-01 2.04845250e-01 1.51676685e-01 -2.11446568e-01
1.24281265e-01 -5.76044798e-01 8.25331748e-01 9.82010961e-01
-2.99673468e-01 7.80540943e-01 5.99565566e-01 -1.51663750e-01
-9.42583919e-01 -1.48689222e+00 -4.74798739e-01 8.68885100e-01
-8.47239941e-02 -3.01664453e-02 -6.97749257e-01 -7.02458620e-01
3.06728661e-01 1.01344013e+00 -5.73662639e-01 -3.26949388e-01
7.41340145e-02 -7.20340312e-01 6.93784297e-01 9.57293928e-01
4.54629362e-01 -6.92494929e-01 -9.93873999e-02 2.12756053e-01
8.52026343e-02 -1.05053234e+00 -2.32442230e-01 5.40451586e-01
-8.48391533e-01 -6.49548531e-01 -6.62484527e-01 -2.91416526e-01
4.23433274e-01 -3.47795546e-01 1.09434116e+00 -5.89590013e-01
2.71981150e-01 1.12810493e-01 6.71212822e-02 -7.30748057e-01
-3.59617144e-01 -2.19316155e-01 5.48533738e-01 -2.05950826e-01
3.87322485e-01 -7.34421849e-01 -3.20631295e-01 2.98217803e-01
-7.23872006e-01 -2.46311322e-01 7.21443295e-01 8.55168164e-01
1.02462232e+00 3.66840035e-01 4.58483011e-01 -5.76164961e-01
7.27966487e-01 -5.22499681e-01 -8.88760209e-01 2.65881449e-01
-6.47102773e-01 4.62723553e-01 3.55051696e-01 -5.44292748e-01
-1.02727926e+00 3.75649929e-02 -1.39150426e-01 -8.54500949e-01
-1.81988850e-01 4.82804745e-01 -3.55302930e-01 1.24203019e-01
8.79407465e-01 3.69549841e-02 -6.14951663e-02 -3.67177218e-01
4.01680052e-01 4.48319882e-01 6.93340838e-01 -9.26611781e-01
6.87715828e-01 1.82885006e-01 5.16230986e-02 -2.97882259e-01
-1.10069633e+00 -1.77024186e-01 -6.29751325e-01 -2.80601941e-02
7.02536821e-01 -9.69805539e-01 -7.54588842e-01 5.45690358e-01
-1.23022926e+00 -2.60790676e-01 -2.29935855e-01 7.38318324e-01
-4.55655634e-01 1.83598876e-01 -4.16979998e-01 -1.06128657e+00
-4.22841281e-01 -1.07478094e+00 9.47101235e-01 3.68161082e-01
-2.09105872e-02 -1.00033247e+00 3.58495861e-02 -1.16791259e-02
3.54678512e-01 2.64472067e-01 5.60403883e-01 -1.03475606e+00
-4.24677908e-01 -1.38543457e-01 -4.16087806e-01 1.14050388e+00
-2.13588104e-01 2.93129385e-01 -1.28105783e+00 1.95051908e-01
4.66586873e-02 -2.83203125e-01 1.21691203e+00 7.47898221e-01
1.32866621e+00 -1.07643224e-01 -2.63651937e-01 5.62671185e-01
1.33099127e+00 -7.10870549e-02 5.41348517e-01 5.44966012e-02
5.89810908e-01 3.46746892e-01 1.35175884e-01 4.65742201e-01
1.19126372e-01 1.93785310e-01 5.59683144e-01 4.55435485e-01
2.32779250e-01 -4.19409424e-01 3.41390252e-01 5.06209970e-01
-7.31763840e-02 -2.12138206e-01 -1.12981749e+00 2.26054266e-01
-2.02740073e+00 -8.90401781e-01 1.12593561e-01 2.31378675e+00
1.00820065e+00 4.84035611e-01 -2.52611130e-01 -4.45289910e-02
8.24193418e-01 -1.28590912e-01 -8.10107827e-01 -1.31408684e-02
-8.51258337e-02 3.53124253e-02 5.66991925e-01 3.63793552e-01
-1.44708180e+00 6.13886297e-01 5.45238161e+00 1.24316573e+00
-8.45150828e-01 1.27058804e-01 1.00543046e+00 4.57018008e-03
-3.22836339e-01 -3.56313467e-01 -1.14229238e+00 6.45277441e-01
1.03178263e+00 -3.65412124e-02 1.74240038e-01 1.33910489e+00
-1.33110300e-01 2.93289330e-02 -1.45816016e+00 1.07402372e+00
-2.76593983e-01 -1.33891582e+00 -1.16183586e-01 2.00799424e-02
9.62513506e-01 3.88027936e-01 2.58852541e-01 4.98035133e-01
8.25323701e-01 -1.31003821e+00 8.27192724e-01 8.41256499e-01
5.60983777e-01 -9.81157839e-01 1.16127574e+00 4.97020036e-01
-7.91805506e-01 9.31510478e-02 -6.85147405e-01 3.24023664e-01
1.60345763e-01 1.19726288e+00 -5.47928393e-01 3.44670832e-01
8.02041650e-01 4.66432273e-01 -8.73811096e-02 1.03469598e+00
-3.82647425e-01 5.71121454e-01 -7.78453946e-01 -1.88847436e-04
1.94973662e-01 -1.96005866e-01 4.94195491e-01 1.01567101e+00
5.72254062e-01 -1.90187141e-01 -8.13995302e-02 1.45702994e+00
-3.98568332e-01 -4.57214177e-01 -4.23161924e-01 -2.67022196e-02
6.81099772e-01 9.54690397e-01 -4.45873857e-01 -5.91423064e-02
-1.64034590e-01 6.24455273e-01 3.71000916e-01 3.81037891e-01
-1.03109014e+00 -2.96237499e-01 9.27632391e-01 -3.91875118e-01
2.95581579e-01 -1.24932952e-01 -4.72841054e-01 -8.93979371e-01
3.50351408e-02 -5.07208288e-01 1.73713490e-01 -6.78580463e-01
-1.68449247e+00 6.06328130e-01 -3.49416584e-02 -1.08459711e+00
-2.85484672e-01 -9.23376918e-01 -5.46200752e-01 1.19992697e+00
-1.34273136e+00 -8.87634337e-01 -1.00362159e-01 4.28447157e-01
1.01103894e-01 -1.34965390e-01 7.72999883e-01 7.65138213e-03
-6.22522652e-01 7.01686740e-01 5.71532249e-01 2.50770330e-01
6.72280490e-01 -1.41716051e+00 2.77479649e-01 8.94165039e-01
3.05229992e-01 5.20843506e-01 6.96340561e-01 -5.47724426e-01
-7.45656908e-01 -1.27584505e+00 4.08576041e-01 -6.89199090e-01
9.45713878e-01 -2.42120832e-01 -1.12792838e+00 5.85012317e-01
-3.82878095e-01 2.33201519e-01 5.71022153e-01 2.99907118e-01
-5.33618331e-01 -2.41916493e-01 -1.16571343e+00 3.51687223e-01
6.00009680e-01 -5.25234401e-01 -5.98066807e-01 2.21676409e-01
9.59566355e-01 -5.70799589e-01 -1.05352688e+00 6.62730515e-01
5.67750633e-01 -9.44733083e-01 9.39130545e-01 -4.99606490e-01
6.21938944e-01 -2.41732478e-01 -4.50065255e-01 -1.45616806e+00
-1.89656466e-01 -3.05419058e-01 -6.58657253e-01 1.30287969e+00
5.94898343e-01 -7.62274861e-01 7.47045517e-01 9.89788711e-01
-2.57134706e-01 -9.31826591e-01 -9.94351089e-01 -9.82792675e-01
2.92056918e-01 -1.04600203e+00 6.15624189e-01 5.26364028e-01
-4.80426669e-01 2.48363218e-03 -2.39549622e-01 5.57230771e-01
9.37954903e-01 -4.18067068e-01 3.44655037e-01 -1.60602093e+00
-3.75829160e-01 -7.09132850e-01 -5.38192689e-01 -9.50234771e-01
2.76086450e-01 -5.09507775e-01 4.23609018e-01 -1.45658445e+00
-7.03162805e-04 -5.67956269e-01 -7.16847956e-01 3.46654534e-01
-6.58055842e-02 -1.34167314e-01 -3.36103663e-02 2.08490819e-01
-5.77228665e-01 7.08691716e-01 8.04806530e-01 -1.55272827e-01
-8.36185887e-02 4.87628490e-01 -6.77610099e-01 9.04221892e-01
9.26276147e-01 -5.18878460e-01 -4.72653538e-01 -3.69376659e-01
2.65585005e-01 -3.02874804e-01 5.37103534e-01 -1.24261343e+00
3.90949816e-01 -3.80155668e-02 6.73740149e-01 -6.42133296e-01
3.39444309e-01 -8.01699102e-01 1.23239882e-01 1.38035774e-01
-3.27870548e-01 -3.80534470e-01 3.83266091e-01 1.00425160e+00
-2.44780645e-01 -4.53337848e-01 8.04362237e-01 5.58468699e-02
-6.55257225e-01 6.25507712e-01 6.23045377e-02 7.28719383e-02
8.54074180e-01 1.03585459e-01 -2.92332262e-01 -3.14740926e-01
-6.49021566e-01 2.90423900e-01 1.04364060e-01 2.79033065e-01
5.78161240e-01 -1.33572590e+00 -6.63747728e-01 -2.31362469e-02
-3.71776149e-02 6.33656085e-01 3.33661318e-01 5.98975301e-01
-1.46122351e-01 3.36644828e-01 8.13345537e-02 -9.06426072e-01
-7.28656054e-01 2.76378185e-01 6.20423198e-01 -1.41281173e-01
-2.43609369e-01 1.32870746e+00 2.67419785e-01 -5.83980381e-01
7.31256485e-01 -5.82075834e-01 5.89342266e-02 -1.97315544e-01
6.77919865e-01 3.09437931e-01 1.00027695e-01 -2.87607402e-01
-2.10443482e-01 1.14325300e-01 6.01440892e-02 -1.44710407e-01
1.25010252e+00 7.25083053e-02 4.63066883e-02 7.20904171e-01
9.34309661e-01 -4.63006586e-01 -1.76025915e+00 -5.53069293e-01
-3.04849427e-02 -9.04597640e-02 4.01626974e-01 -7.78637886e-01
-9.52111006e-01 1.14648092e+00 6.43982768e-01 -8.68585184e-02
7.27276564e-01 1.30936727e-01 4.00272220e-01 6.07157707e-01
2.16436163e-01 -1.17459464e+00 -3.12426519e-02 6.18024349e-01
8.40665936e-01 -1.43773329e+00 -1.76348090e-01 1.44716755e-01
-7.29542792e-01 1.18620133e+00 6.47017360e-01 7.30017349e-02
1.20050275e+00 4.15875703e-01 -2.57910043e-01 1.93500012e-01
-6.74607515e-01 1.43682733e-01 6.38253033e-01 6.58757508e-01
1.61651671e-01 2.68759072e-01 5.48716962e-01 1.22396183e+00
-1.17373921e-01 -8.07518810e-02 5.92326634e-02 4.24067169e-01
-5.56571960e-01 -5.31206250e-01 -3.04044515e-01 7.30713487e-01
-3.73461366e-01 -2.59028941e-01 1.53976977e-01 4.85159695e-01
1.44335717e-01 6.80730700e-01 1.80592820e-01 -4.71588165e-01
5.94255030e-02 2.50092983e-01 2.17950821e-01 -5.15729249e-01
1.78273872e-01 -2.10655794e-01 -1.89638630e-01 -4.68063354e-01
-1.10479973e-01 -5.91009736e-01 -1.25195312e+00 -4.72048819e-01
-4.16742772e-01 -1.11189328e-01 8.45556736e-01 1.10360098e+00
4.14999902e-01 5.34467578e-01 1.15613483e-01 -8.87908399e-01
-1.20834064e+00 -1.11812317e+00 -7.85224795e-01 -1.57230475e-03
1.65107846e-01 -9.07546043e-01 -6.34188712e-01 -2.04391524e-01] | [7.463440895080566, 3.756512403488159] |
4213709b-cefd-4a11-8bca-a6406e7291e8 | gene-communities-in-co-expression-networks | 2305.12963 | null | https://arxiv.org/abs/2305.12963v1 | https://arxiv.org/pdf/2305.12963v1.pdf | Gene communities in co-expression networks across different tissues | With the recent availability of tissue-specific gene expression data, e.g., provided by the GTEx Consortium, there is interest in comparing gene co-expression patterns across tissues. One promising approach to this problem is to use a multilayer network analysis framework and perform multilayer community detection. Communities in gene co-expression networks reveal communities of genes similarly expressed across individuals, potentially involved in related biological processes responding to specific environmental stimuli or sharing common regulatory variations. We construct a multilayer network in which each layer is a tissue-specific gene co-expression network. We develop methods for multilayer community detection with correlation matrix input and an appropriate null model. Our correlation matrix input method identifies groups of genes that are similarly co-expressed in multiple tissues (a community that spans multiple layers, which we call a generalist community) and some groups of genes that are co-expressed in just one tissue (a community that lies primarily within just one layer, which we call a specialist community). We further found gene co-expression communities where the genes physically cluster across the genome significantly more than expected by chance. This clustering hints at underlying regulatory elements determining similar expression patterns across individuals and cell types. The results indicate that our multilayer community detection method for correlation matrix input extracts biologically interesting communities of genes. | ['Naoki Masuda', 'Omer Gokcumen', 'Marie Saitou', 'Madison Russell'] | 2023-05-22 | null | null | null | null | ['community-detection'] | ['graphs'] | [ 1.56604171e-01 -2.12898731e-01 2.71147430e-01 -1.49678811e-02
1.18571490e-01 -9.13180828e-01 3.87344629e-01 5.34380436e-01
2.92164832e-02 5.25539994e-01 2.12645024e-01 -3.26274008e-01
-3.91097277e-01 -8.32869053e-01 -3.82829368e-01 -1.10063267e+00
-7.29780674e-01 2.96119422e-01 1.98556155e-01 -5.93378022e-02
-1.14839770e-01 4.47750390e-01 -1.29537845e+00 6.44307435e-01
5.07318914e-01 2.26418927e-01 -2.77972296e-02 8.80806327e-01
-3.02466452e-01 1.37172788e-01 -4.79532510e-01 1.37396052e-01
1.79732665e-01 -7.59064555e-01 -5.72989762e-01 2.76075959e-01
1.35524720e-01 1.01697397e+00 9.66158211e-02 1.17978489e+00
3.03696811e-01 -3.90950948e-01 6.30471647e-01 -1.30032563e+00
-5.92241064e-02 5.98418176e-01 -9.80018258e-01 3.60507779e-02
1.91111758e-01 9.98025090e-02 1.16048622e+00 -7.53164411e-01
1.04749537e+00 1.35327709e+00 7.19091356e-01 3.02789032e-01
-2.13508844e+00 -3.25567275e-01 -4.42173034e-02 -2.68459916e-01
-1.99634314e+00 -2.83656836e-01 2.50227928e-01 -8.62736523e-01
9.51434672e-01 7.63457775e-01 1.13174164e+00 3.54029059e-01
3.42480421e-01 1.23373978e-01 1.24297690e+00 -3.76660049e-01
4.20645833e-01 -2.74235964e-01 -7.92013183e-02 6.28510475e-01
5.80057740e-01 -4.10683453e-01 -3.09158564e-01 -7.93230355e-01
4.58797514e-01 8.71414393e-02 -3.95220488e-01 -3.92577708e-01
-1.48091578e+00 5.14335334e-01 2.54368067e-01 8.43643904e-01
-2.06041172e-01 8.34120065e-02 4.45195317e-01 5.28769135e-01
4.21852350e-01 3.90808821e-01 -4.01362777e-01 4.44926053e-01
-9.32259321e-01 -2.35441655e-01 9.37229991e-01 6.28955483e-01
8.43518496e-01 -6.20973289e-01 3.84282112e-01 9.77059662e-01
5.00952661e-01 -1.74625993e-01 1.73637435e-01 -8.12503517e-01
-2.52518564e-01 9.97860014e-01 -4.26645607e-01 -1.35314929e+00
-5.78448713e-01 -4.25232828e-01 -1.28906310e+00 3.12562525e-01
5.13227642e-01 -3.23358446e-01 -1.83570802e-01 1.82231867e+00
4.39445108e-01 1.75957978e-01 2.22326070e-02 6.33991420e-01
2.58445710e-01 3.85299981e-01 7.87076168e-03 -3.86540890e-01
1.63573694e+00 -2.06570789e-01 -4.78137702e-01 2.22576767e-01
6.59231365e-01 -6.36344969e-01 3.77295911e-01 7.41053075e-02
-6.47858024e-01 2.07894854e-02 -8.88935387e-01 5.09275854e-01
-3.26888919e-01 -3.31547976e-01 7.02147603e-01 4.52032536e-01
-1.37181890e+00 6.03453755e-01 -5.58103919e-01 -1.11690402e+00
2.98188865e-01 3.83695543e-01 -8.13475788e-01 -1.38988361e-01
-8.19219828e-01 3.09397399e-01 1.59408525e-01 1.71947584e-01
-6.63530946e-01 -3.36179316e-01 -2.94186503e-01 2.22171992e-01
6.68484047e-02 -1.01815438e+00 1.80120021e-01 -8.40115786e-01
-7.62823105e-01 1.08295512e+00 -2.84878790e-01 2.49278545e-01
2.04131044e-02 6.75223827e-01 -3.95991564e-01 1.37246981e-01
1.50744885e-01 4.05283093e-01 3.15792821e-02 -1.13641620e+00
-3.12393099e-01 -5.25006950e-01 -3.71415019e-01 -1.08986497e-01
-3.03837448e-01 4.29510295e-01 -2.31970817e-01 -2.15710402e-01
5.57359993e-01 -1.24385166e+00 -6.44849181e-01 2.69492924e-01
-6.99230433e-01 -5.59476539e-02 3.83279383e-01 7.17943311e-02
9.92846131e-01 -2.00021768e+00 5.07543325e-01 8.12039077e-01
6.80914342e-01 -4.27971542e-01 -2.78758258e-01 8.39622736e-01
-6.76072717e-01 8.89492273e-01 -2.34725326e-01 2.23531961e-01
-4.34148014e-01 2.15138122e-01 4.73312318e-01 1.05910385e+00
4.40884709e-01 4.62752044e-01 -1.13416553e+00 -4.22186881e-01
-4.61940616e-01 2.18473777e-01 -4.30010378e-01 -2.40835343e-02
-2.87728142e-02 3.39450538e-01 -2.46885151e-01 7.57544398e-01
6.79219246e-01 -7.11956382e-01 1.00932443e+00 -2.06791148e-01
-3.35950851e-01 -2.68228441e-01 -1.26067829e+00 1.26074696e+00
3.80652100e-01 8.66129994e-01 6.42675996e-01 -1.02084923e+00
8.06686997e-01 5.57238460e-01 7.62828350e-01 4.02507722e-01
-1.06262460e-01 1.52533120e-02 7.35162854e-01 -4.92858022e-01
-1.50081515e-01 -3.01800698e-01 1.36872619e-01 6.22387290e-01
8.54249448e-02 2.70414025e-01 2.14834332e-01 5.44478059e-01
1.92066252e+00 -6.44734979e-01 5.06369889e-01 -8.20043325e-01
4.06373858e-01 8.82225037e-02 9.13152635e-01 4.30784583e-01
-1.48026556e-01 5.02373219e-01 1.22347355e+00 2.94456929e-02
-1.19099605e+00 -7.59506106e-01 -1.70603067e-01 8.88017893e-01
-2.86531329e-01 -7.30025411e-01 -2.55827099e-01 -4.86820936e-02
9.84985009e-03 -4.96992111e-01 -8.25377285e-01 1.46571100e-01
6.87135458e-02 -1.30404043e+00 9.23381448e-01 -1.14305645e-01
1.17234036e-01 -3.85852933e-01 1.55240908e-01 1.64701387e-01
-1.36893466e-01 -5.62615156e-01 -3.39075357e-01 5.26338458e-01
-1.14449596e+00 -1.28570580e+00 -4.66901064e-01 -7.00508118e-01
1.10145748e+00 4.34488744e-01 1.00847173e+00 5.15243530e-01
-7.75298178e-01 2.32889667e-01 -6.99484050e-02 -1.29175842e-01
-4.82634515e-01 -4.27481711e-01 2.33553410e-01 7.24161640e-02
1.58199459e-01 -7.92238593e-01 -3.85032147e-01 4.68364358e-01
-8.34615588e-01 -9.03693289e-02 3.16074669e-01 8.41281533e-01
4.73136961e-01 2.64292955e-01 2.96104014e-01 -7.70605683e-01
7.20365286e-01 -1.18086743e+00 -4.25874889e-01 4.24640596e-01
-2.28118584e-01 -1.45449504e-01 1.76024318e-01 -2.38852948e-01
-3.56436163e-01 2.91815609e-01 2.85554647e-01 2.49758549e-03
-1.91118538e-01 1.09680915e+00 -3.12722951e-01 -3.81176844e-02
7.49604821e-01 1.37893349e-01 3.76690984e-01 -1.58618748e-01
1.02500744e-01 4.75130260e-01 3.03869899e-02 -4.09027606e-01
3.61706913e-01 8.20711732e-01 6.51503563e-01 -1.10339785e+00
3.27258110e-02 -8.72117817e-01 -8.72946620e-01 -4.10799205e-01
7.96056271e-01 -1.01118910e+00 -8.73174787e-01 1.64803594e-01
-9.47544038e-01 -7.97836632e-02 3.68801415e-01 3.21176082e-01
-5.79667352e-02 4.90825146e-01 -9.14808512e-01 -7.39013016e-01
1.05929539e-01 -5.49517512e-01 5.30111372e-01 -2.01360315e-01
-6.07375503e-01 -1.27913046e+00 7.89619207e-01 -4.84009869e-02
8.99471492e-02 6.27599239e-01 1.03237915e+00 -5.19202352e-01
-4.57723022e-01 -2.27163173e-02 3.39446701e-02 -1.26871198e-01
3.60713184e-01 7.62185931e-01 -7.20012307e-01 -4.02232587e-01
-5.93255401e-01 2.46391118e-01 7.27738976e-01 1.68081656e-01
6.57720089e-01 -2.84432173e-01 -9.20185089e-01 3.83124202e-01
1.50357711e+00 -8.31280500e-02 6.39271617e-01 -8.38045925e-02
4.74881589e-01 1.22046041e+00 8.87555107e-02 1.29196167e-01
-2.59737164e-01 4.95983899e-01 1.68177620e-01 -3.96561205e-01
3.07972074e-01 3.76069188e-01 3.25057268e-01 9.63581741e-01
-3.68192136e-01 -2.84029365e-01 -1.08835328e+00 7.11636841e-01
-2.06667566e+00 -1.21755695e+00 -9.21182632e-01 2.01744628e+00
9.15484667e-01 -2.90247649e-01 2.97840834e-01 -2.43692487e-01
1.11566520e+00 -3.52378905e-01 -2.62667805e-01 -2.31429905e-01
-6.85640693e-01 -1.76045135e-01 1.98242188e-01 3.51407945e-01
-6.96047425e-01 3.57214570e-01 7.45045805e+00 3.58118415e-01
-7.88439155e-01 -2.31850401e-01 5.29073894e-01 -1.12590581e-01
-3.86253953e-01 4.08720195e-01 -3.22708964e-01 2.53826857e-01
8.95520866e-01 -6.76135302e-01 7.75471330e-02 4.47938800e-01
5.11030018e-01 -1.29562423e-01 -1.03777993e+00 4.21311319e-01
-3.85917306e-01 -1.44517064e+00 -5.10289073e-01 6.04784846e-01
8.15429807e-01 2.98594773e-01 -5.81702113e-01 -4.47129190e-01
7.99483716e-01 -9.54503119e-01 -1.13909915e-01 5.35316825e-01
5.99815130e-01 -2.72714913e-01 6.00224495e-01 1.93031535e-01
-1.38792038e+00 2.36866325e-01 -6.37213647e-01 -2.18470216e-01
-1.34749711e-01 1.35692954e+00 -1.09820235e+00 3.95580232e-01
7.80419767e-01 8.03137302e-01 -4.40345764e-01 1.28698516e+00
1.16991721e-01 7.87760556e-01 -3.37799311e-01 -1.68303102e-01
-1.98320881e-01 -4.15875196e-01 6.68884158e-01 1.52202082e+00
3.76958311e-01 3.25147480e-01 1.33420154e-01 1.08530593e+00
2.06282035e-01 2.07239360e-01 -7.46720791e-01 -2.82845289e-01
5.15197456e-01 1.86256957e+00 -1.23698306e+00 -3.28624219e-01
-4.59815681e-01 9.61247206e-01 3.48870873e-01 3.64697814e-01
-8.76569189e-03 -2.58636117e-01 1.45963562e+00 1.90378219e-01
-4.12828773e-02 -2.34024197e-01 -3.00675025e-03 -1.11792886e+00
-5.86200655e-01 -5.98415017e-01 3.50781620e-01 -5.52654684e-01
-1.81825650e+00 1.18719511e-01 -2.98184752e-01 -1.17255998e+00
5.18741123e-02 -5.36904931e-01 -8.59795928e-01 1.10902202e+00
-5.86342216e-01 -6.14946604e-01 -2.53909022e-01 4.74953920e-01
-3.65120262e-01 -6.82778358e-02 1.05058122e+00 2.05677003e-01
-7.33893275e-01 -8.28419030e-02 4.13758337e-01 8.64521936e-02
5.93134403e-01 -1.19100761e+00 6.21302165e-02 5.03405154e-01
-1.25914719e-02 1.44005036e+00 7.17114389e-01 -8.57960582e-01
-1.07533276e+00 -1.01676774e+00 1.02785015e+00 3.84822004e-02
9.16382551e-01 -8.07968080e-01 -1.11762071e+00 3.18661273e-01
1.72674805e-01 2.48802349e-01 1.53158271e+00 4.05009001e-01
-3.13998163e-01 2.28252038e-02 -1.02915525e+00 7.63462067e-01
1.07245922e+00 -5.70218265e-01 -1.00109830e-01 6.03389502e-01
2.71131009e-01 5.60201526e-01 -1.24202204e+00 9.40108150e-02
7.14193940e-01 -9.78161573e-01 6.83291316e-01 -7.40453541e-01
1.64696515e-01 -1.11566758e+00 -5.14379069e-02 -1.59674501e+00
-9.80202794e-01 -3.17438364e-01 7.15955973e-01 1.14046574e+00
4.97851610e-01 -6.19439602e-01 6.14699841e-01 2.00662494e-01
-1.45705640e-02 -3.82921457e-01 -9.23524737e-01 -7.41314292e-01
-1.97385661e-02 1.60107017e-01 1.77886024e-01 1.71618462e+00
8.02185774e-01 5.52788854e-01 4.39554788e-02 -5.55076152e-02
6.54370308e-01 1.34122759e-01 5.88325143e-01 -1.47121882e+00
-3.32728565e-01 -5.51074147e-01 -7.82908082e-01 -3.60567629e-01
-7.04367608e-02 -1.16810727e+00 1.52721152e-01 -1.34668207e+00
8.07345688e-01 -3.97839934e-01 -2.47340620e-01 7.03190446e-01
-4.78367694e-02 2.50290990e-01 -2.37521380e-01 3.32778484e-01
-5.10497570e-01 -1.62121400e-01 9.67166126e-01 -5.81135675e-02
-1.23204710e-02 -5.96997082e-01 -9.41383719e-01 5.27937174e-01
6.79569006e-01 -6.24156713e-01 1.24744147e-01 2.42817834e-01
6.77846789e-01 -2.57596131e-02 2.88842231e-01 -8.14782917e-01
6.23393893e-01 -2.19538510e-01 5.08370280e-01 -1.77407712e-01
-1.46800205e-01 -8.22879791e-01 1.12924230e+00 8.47578883e-01
-3.07828814e-01 1.30628318e-01 1.73869699e-01 7.22547710e-01
-2.44089812e-01 3.89698446e-02 5.73978662e-01 -4.30122674e-01
-2.54741848e-01 -1.14711516e-01 -1.32106209e+00 -5.67260385e-01
1.03548717e+00 -2.53151417e-01 -4.51488614e-01 -2.56773114e-01
-1.11348772e+00 1.66839659e-01 6.19150996e-01 2.43694093e-02
3.50154966e-01 -1.38099790e+00 -1.10135031e+00 -8.99037421e-02
4.59914386e-01 -4.88951743e-01 8.58106092e-02 1.09939981e+00
-4.56480473e-01 4.51395474e-02 -2.96771377e-01 -1.02117443e+00
-2.06832123e+00 1.57648310e-01 4.14487392e-01 -5.55249974e-02
-8.33999738e-03 9.20455098e-01 3.29954326e-01 -5.00972748e-01
-6.12381458e-01 1.80643186e-01 -2.73274064e-01 2.58711100e-01
3.85480464e-01 1.85458094e-01 -2.14636758e-01 -6.03745222e-01
-5.96706450e-01 4.83097613e-01 3.06118131e-01 -4.28512022e-02
1.26150608e+00 -2.09018871e-01 -1.48306084e+00 8.47560942e-01
1.25381732e+00 2.27255493e-01 -4.48062599e-01 1.10176049e-01
3.73603553e-01 -4.12645549e-01 -4.32156444e-01 -5.18643498e-01
-9.52348709e-01 4.30652618e-01 2.19805121e-01 3.31595838e-01
8.51410747e-01 3.10005277e-01 -4.78263080e-01 2.70684332e-01
2.44719431e-01 -5.78017116e-01 1.02252513e-01 5.15696347e-01
7.20043063e-01 -6.69199467e-01 1.49563193e-01 -7.81955361e-01
-5.20930216e-02 1.23027039e+00 3.95115316e-01 -2.07953632e-01
8.03980470e-01 4.82452154e-01 -1.74342722e-01 -7.02456832e-01
-1.33935666e+00 -8.48013386e-02 9.47824270e-02 4.57420230e-01
1.15596485e+00 3.26780051e-01 -5.99894762e-01 2.79135019e-01
2.78725326e-01 -2.17839882e-01 6.07540607e-01 7.13976860e-01
-5.59659958e-01 -1.19763386e+00 -5.93498230e-01 4.98476624e-01
-3.94102365e-01 -1.98588580e-01 -1.11665738e+00 4.80350971e-01
2.32044160e-01 1.10089850e+00 4.03860062e-01 -5.63227654e-01
-5.53171299e-02 1.92080379e-01 3.78898740e-01 -9.89253581e-01
-6.85255408e-01 7.37963259e-01 2.53010899e-01 -3.33363831e-01
-5.59022248e-01 -8.65469396e-01 -1.15067053e+00 -4.35844988e-01
-5.38554132e-01 3.79106849e-01 3.83307338e-01 4.45439696e-01
5.71622849e-01 5.45056105e-01 5.03372014e-01 -3.13589334e-01
3.76704425e-01 -1.04810500e+00 -1.07989883e+00 2.98242152e-01
-5.21118045e-02 -1.24722436e-01 -5.38927257e-01 2.94490814e-01] | [6.663071632385254, 5.364526748657227] |
4d295134-6f76-4f2a-a431-628190475b3c | analytical-modelling-of-exoplanet-transit | 2112.11600 | null | https://arxiv.org/abs/2112.11600v1 | https://arxiv.org/pdf/2112.11600v1.pdf | Analytical Modelling of Exoplanet Transit Specroscopy with Dimensional Analysis and Symbolic Regression | The physical characteristics and atmospheric chemical composition of newly discovered exoplanets are often inferred from their transit spectra which are obtained from complex numerical models of radiative transfer. Alternatively, simple analytical expressions provide insightful physical intuition into the relevant atmospheric processes. The deep learning revolution has opened the door for deriving such analytical results directly with a computer algorithm fitting to the data. As a proof of concept, we successfully demonstrate the use of symbolic regression on synthetic data for the transit radii of generic hot Jupiter exoplanets to derive a corresponding analytical formula. As a preprocessing step, we use dimensional analysis to identify the relevant dimensionless combinations of variables and reduce the number of independent inputs, which improves the performance of the symbolic regression. The dimensional analysis also allowed us to mathematically derive and properly parametrize the most general family of degeneracies among the input atmospheric parameters which affect the characterization of an exoplanet atmosphere through transit spectroscopy. | ['Alexander Roman', 'Katia Matcheva', 'Konstantin T. Matchev'] | 2021-12-22 | null | null | null | null | ['physical-intuition'] | ['reasoning'] | [-1.19603068e-01 -2.44387478e-01 2.50922620e-01 -2.15450615e-01
-6.93533421e-02 -8.12482595e-01 9.03970361e-01 9.12340507e-02
-2.99022108e-01 7.14207947e-01 -5.60845315e-01 -7.73155928e-01
-2.52121389e-01 -9.17676508e-01 -4.19418246e-01 -1.02685726e+00
-2.04275116e-01 7.27574408e-01 -1.07984208e-01 -4.84296322e-01
-4.66585420e-02 1.05567944e+00 -1.53180528e+00 -4.92622733e-01
9.02242541e-01 8.95435810e-01 -4.12004411e-01 7.56377339e-01
-4.40035649e-02 4.41052020e-01 -4.57068413e-01 1.00972662e-02
5.85723460e-01 -5.14304817e-01 -5.73022842e-01 -2.16822013e-01
1.71647489e-01 -8.63105431e-02 -4.57225055e-01 7.17613459e-01
1.00771666e-01 5.08991301e-01 1.08156610e+00 -1.10817814e+00
9.02802497e-03 3.70648772e-01 -4.48318452e-01 2.22170085e-01
-3.36203128e-01 2.95687020e-01 9.92713511e-01 -2.82937437e-01
-8.62528980e-02 9.67625916e-01 7.08841383e-01 2.20339019e-02
-1.11987484e+00 -3.74038696e-01 -4.83322710e-01 1.85410697e-02
-1.36310911e+00 -2.06870317e-01 5.64483702e-01 -5.11674464e-01
8.68794322e-01 4.71542269e-01 8.40493321e-01 3.96362394e-01
-1.48995250e-01 3.09265461e-02 1.03123236e+00 -6.74335182e-01
1.59267426e-01 3.25899154e-01 1.09489828e-01 8.50926161e-01
4.48421389e-01 5.68185210e-01 -1.39431417e-01 -4.60022628e-01
6.48710787e-01 -4.50118035e-01 -1.38060808e-01 4.88880724e-02
-8.47546101e-01 8.98176372e-01 3.68722409e-01 5.95764965e-02
-3.10044885e-01 2.52076685e-01 3.02164435e-01 5.01180112e-01
2.75196493e-01 9.89007771e-01 -5.90757370e-01 4.74514402e-02
-7.39619613e-01 4.57827866e-01 1.06716180e+00 2.76847810e-01
1.12790287e+00 2.06557289e-01 5.03364027e-01 2.44069189e-01
1.82383895e-01 1.02869642e+00 2.78176457e-01 -1.08548999e+00
-7.24191144e-02 4.03973848e-01 4.86439407e-01 -6.68148220e-01
-7.79634595e-01 -2.15856612e-01 -5.17156541e-01 2.37287015e-01
6.40781045e-01 -6.58119023e-01 -5.01925051e-01 1.16599953e+00
5.50229311e-01 -2.57587880e-01 3.63685787e-01 8.82530928e-01
5.21833420e-01 8.52817357e-01 -8.65703076e-02 -1.41015276e-01
9.72317934e-01 -5.30726135e-01 -6.33239821e-02 4.32543084e-02
7.06765473e-01 -4.67958808e-01 1.02293026e+00 1.54492766e-01
-4.41367269e-01 -2.91612595e-01 -1.13172507e+00 -1.01541886e-02
-4.72515762e-01 1.24166161e-01 1.23794651e+00 7.35988975e-01
-5.65376341e-01 1.02180576e+00 -9.37034607e-01 3.73899541e-03
-4.03464615e-01 4.37832743e-01 -1.46786449e-02 7.64117002e-01
-1.27928364e+00 7.68675089e-01 3.13119560e-01 2.00949594e-01
-7.44325519e-01 -7.53010929e-01 -4.48806584e-01 3.89819779e-02
1.36247277e-01 -8.33843589e-01 1.41851330e+00 -8.14680874e-01
-1.62611127e+00 7.07097709e-01 2.06012949e-02 -6.40116572e-01
4.21101719e-01 1.46998093e-01 -3.15624058e-01 4.21569079e-01
-3.96670014e-01 2.82714844e-01 6.99519634e-01 -7.33384967e-01
-1.44163802e-01 -9.16858763e-02 1.08280636e-01 1.15575148e-02
-1.33595824e-01 1.43138161e-02 1.86247215e-01 4.29605655e-02
5.48182055e-02 -1.26968741e+00 -2.81572253e-01 8.03755298e-02
-2.32803464e-01 8.06987807e-02 3.88286710e-01 -3.51754636e-01
4.85662639e-01 -2.00484586e+00 2.16275454e-01 3.73942643e-01
1.67731389e-01 9.64024886e-02 2.61790574e-01 5.55228591e-01
-3.39238971e-01 3.79354656e-02 -3.73427778e-01 -5.61072454e-02
3.16595286e-02 -4.28158091e-03 -5.80885768e-01 6.23985350e-01
1.85742840e-01 5.86606860e-01 -6.11883700e-01 -9.38004926e-02
2.17440575e-01 2.54035950e-01 -1.32333994e-01 2.58490771e-01
-6.29324317e-01 4.59140360e-01 -5.74644744e-01 1.38176054e-01
4.81108308e-01 4.61338647e-03 -6.32489622e-02 1.82712991e-02
-5.74594200e-01 6.23067439e-01 -5.70383728e-01 8.82644475e-01
-6.79992080e-01 9.14582312e-01 2.64981687e-02 -8.25867712e-01
1.09943724e+00 3.25939432e-02 6.30962551e-01 -5.02632320e-01
6.15052462e-01 7.25273713e-02 3.51476938e-01 -6.61818802e-01
7.36140668e-01 -8.83465648e-01 5.12411706e-02 7.46529639e-01
-2.86042571e-01 -6.16853714e-01 -1.13885347e-02 1.42307863e-01
4.09413010e-01 1.07923485e-01 7.44047910e-02 -5.69793701e-01
4.09994304e-01 2.67490029e-01 7.84665495e-02 4.45311487e-01
2.40404874e-01 2.90892929e-01 7.43065596e-01 -7.83766210e-01
-1.22015417e+00 -8.02043617e-01 -3.02771896e-01 6.78886592e-01
-5.16480356e-02 -2.95062333e-01 -5.33544958e-01 1.23712435e-01
1.88417047e-01 6.81835651e-01 -6.40940249e-01 -3.61190826e-01
-2.95056015e-01 -1.37666047e+00 8.49377394e-01 7.54849389e-02
2.98582971e-01 -6.04694188e-01 -8.24761927e-01 -2.79641300e-01
3.54671180e-02 -1.23356676e+00 5.27994812e-01 4.54168260e-01
-8.06775510e-01 -1.14158988e+00 -7.72235394e-02 1.26581028e-01
2.86039263e-01 1.29074261e-01 6.52713835e-01 2.14520529e-01
-4.17147398e-01 1.36401623e-01 -3.11242901e-02 -6.73983932e-01
-8.05640101e-01 3.82296555e-02 5.21650851e-01 -2.14695185e-02
1.66093409e-01 -7.22825468e-01 -2.66121358e-01 1.08010873e-01
-6.75375402e-01 -5.52604673e-03 9.33279917e-02 1.40841618e-01
2.20830813e-01 8.15383270e-02 1.51937008e-01 -4.32373077e-01
3.49698842e-01 -4.51114208e-01 -1.41146779e+00 2.56192476e-01
-3.07091653e-01 7.13956296e-01 9.60559666e-01 -3.61591309e-01
-9.48215663e-01 -1.31335929e-01 9.46634337e-02 -3.18857044e-01
4.68147732e-02 4.77657765e-01 3.05759758e-01 -4.35654908e-01
1.04225850e+00 1.56313196e-01 8.44065025e-02 -3.43841344e-01
3.65152001e-01 5.37017941e-01 6.20250762e-01 -1.08675981e+00
1.18627942e+00 6.79450750e-01 7.05646217e-01 -1.48906720e+00
-7.10159898e-01 -2.69752294e-01 -6.47967875e-01 -9.84201729e-02
4.16309386e-01 -7.96031654e-01 -1.11453474e+00 3.40421766e-01
-8.24002504e-01 -4.81617540e-01 -2.10942194e-01 5.97881496e-01
-3.25020939e-01 5.46961010e-01 -3.42445463e-01 -1.15787733e+00
-4.61873174e-01 -8.79218340e-01 7.57559061e-01 3.64287645e-01
-1.82619214e-01 -1.13419878e+00 3.17894280e-01 -3.63254547e-02
2.40241066e-01 5.16962826e-01 1.28067458e+00 -2.83557892e-01
-5.27843297e-01 -1.93388432e-01 -2.18110114e-01 -4.96812537e-03
3.48881185e-02 8.75249743e-01 -1.19355106e+00 -4.82041463e-02
3.02785784e-01 3.38979624e-02 7.24325538e-01 2.33620957e-01
8.35814595e-01 -7.32207745e-02 -7.82311708e-02 1.08051622e+00
1.11286366e+00 -7.82567710e-02 1.53255418e-01 3.77975792e-01
5.49951971e-01 6.55911863e-01 2.66588569e-01 5.16452610e-01
-1.19876429e-01 2.51699597e-01 2.40082100e-01 -3.39450873e-02
3.93257231e-01 -3.87226790e-02 1.60789400e-01 6.75182641e-01
-3.65839243e-01 6.34626672e-02 -9.73133683e-01 1.50599718e-01
-1.19212008e+00 -5.06713271e-01 -4.91924256e-01 2.40575647e+00
6.03110850e-01 -2.14949090e-04 3.50701362e-01 -1.14222862e-01
2.44355679e-01 -1.15957305e-01 -7.02284038e-01 -5.33742249e-01
-8.14286694e-02 3.26545626e-01 9.99468923e-01 6.57423139e-01
-7.83099592e-01 7.03397989e-01 7.08090353e+00 1.01498820e-01
-1.66453302e+00 -4.10301834e-01 2.49601156e-01 -8.26947689e-02
-4.17796731e-01 2.52507329e-01 -7.05673397e-01 -1.36559203e-01
1.52559435e+00 -6.32400334e-01 6.87422395e-01 3.71328861e-01
3.87771189e-01 -2.34255806e-01 -1.02721989e+00 7.26909459e-01
-5.26025355e-01 -1.12771606e+00 -9.40948427e-02 3.09046451e-02
5.63391268e-01 5.42936683e-01 -6.98830606e-03 1.52937314e-02
-3.98563631e-02 -6.65798783e-01 5.08434296e-01 6.72917664e-01
7.85015464e-01 -6.00237370e-01 8.41023326e-02 3.60670596e-01
-7.17027485e-01 -5.57433404e-02 -6.78816617e-01 -5.77714920e-01
-1.80834904e-01 6.02979720e-01 -1.18718314e+00 6.02110028e-01
3.42205554e-01 6.88053444e-02 -4.77025747e-01 9.37155306e-01
-1.70720667e-01 7.32054830e-01 -1.16834271e+00 -2.31178939e-01
3.23705971e-01 -7.26918519e-01 5.33102691e-01 7.74181068e-01
2.54035056e-01 3.58106136e-01 -6.68482721e-01 1.19323230e+00
8.26840028e-02 5.76284230e-02 -4.91387188e-01 -4.88094956e-01
4.09970507e-02 1.54608715e+00 -8.15516055e-01 -1.16422318e-01
-1.20599456e-01 2.42818310e-03 9.17093530e-02 4.01459247e-01
-8.58851671e-01 -3.84132028e-01 8.76126170e-01 1.26343071e-01
8.21128711e-02 -8.06230366e-01 -1.60068229e-01 -1.15600944e+00
-3.17211062e-01 -8.04486036e-01 1.49867069e-02 -7.45776534e-01
-7.71149814e-01 5.00693142e-01 3.17444533e-01 -1.00285852e+00
-3.06191385e-01 -8.82221043e-01 -1.02818763e+00 9.87423539e-01
-1.36403680e+00 -3.68445784e-01 -3.56172800e-01 1.82963997e-01
-3.02988082e-01 -1.00241929e-01 8.92924964e-01 -2.01973125e-01
-7.08459020e-01 1.75468355e-01 5.54146707e-01 -2.47266650e-01
3.66441578e-01 -1.01902854e+00 7.37929642e-01 7.59311199e-01
6.38566492e-03 5.94251156e-01 1.25779343e+00 -3.93604904e-01
-1.55760837e+00 -6.40303433e-01 3.24640334e-01 -2.29832470e-01
1.17630255e+00 -3.37099046e-01 -9.27491724e-01 4.89255100e-01
-1.70705780e-01 -2.09166080e-01 6.65868104e-01 1.20138094e-01
-2.57786781e-01 -1.88112572e-01 -8.59935164e-01 4.89350557e-01
3.19780082e-01 -7.11037040e-01 -4.38692510e-01 5.90683818e-01
8.17916036e-01 -8.95527378e-02 -7.01315284e-01 2.85706460e-01
5.76938570e-01 -9.14862275e-01 8.47944796e-01 -9.19395387e-01
1.68528602e-01 -4.35996711e-01 1.23398095e-01 -1.01964760e+00
1.19700767e-01 -1.04534316e+00 1.83267802e-01 4.91415292e-01
5.39752603e-01 -7.34545946e-01 4.56210971e-01 9.58989024e-01
2.01512530e-01 -3.18079382e-01 -5.26963115e-01 -7.91485012e-01
4.04537320e-01 -4.03686732e-01 6.99633360e-01 9.97178614e-01
-3.23016763e-01 2.84648269e-01 -5.51535711e-02 6.73226297e-01
5.44458270e-01 6.57032788e-01 7.67789960e-01 -1.57028508e+00
-6.13348663e-01 -2.53556132e-01 -1.74968801e-02 -7.66855061e-01
2.95445740e-01 -7.39525080e-01 -1.56062379e-01 -6.58293188e-01
-4.48735029e-01 -7.88757503e-01 6.88507855e-02 2.52386898e-01
3.70684743e-01 -4.64297868e-02 -3.21520753e-02 1.07323155e-01
2.79799759e-01 5.99569559e-01 8.71112049e-01 7.71934837e-02
-2.64124751e-01 2.83796579e-01 -3.20608974e-01 8.54862154e-01
9.47145343e-01 -6.39891684e-01 -1.95582703e-01 -3.34018350e-01
8.25346172e-01 1.60631090e-01 4.31033224e-01 -9.82842743e-01
4.79145907e-02 -3.41690898e-01 1.17676072e-01 -3.26828033e-01
3.62432718e-01 -4.69467938e-01 1.48361251e-01 2.60738552e-01
-2.25782841e-01 -1.82636470e-01 5.82611203e-01 5.91942370e-02
6.97686523e-02 -6.27685070e-01 8.88002276e-01 -1.32565117e-02
-1.70921132e-01 1.98431194e-01 -5.61041892e-01 -1.99135855e-01
6.11999273e-01 2.92696744e-01 -2.36522675e-01 -3.40588868e-01
-4.85000700e-01 1.18360564e-01 5.51609755e-01 -1.31521560e-02
-9.28494632e-02 -5.10722399e-01 -5.01590312e-01 3.68371457e-01
-3.03402454e-01 -1.91197600e-02 -1.19342797e-01 7.89901912e-01
-1.26307690e+00 3.98488492e-01 -2.17837468e-01 -4.07341182e-01
-9.01605725e-01 5.13703406e-01 1.13920188e+00 3.19769233e-01
-3.39774519e-01 5.47320664e-01 1.52021348e-01 -3.85971189e-01
-3.84370863e-01 -5.81127286e-01 2.00390294e-01 -1.82754751e-02
3.22173387e-01 3.77124816e-01 1.31290123e-01 -4.51652110e-01
-5.67045175e-02 5.26902258e-01 3.91926020e-01 -3.50312553e-02
1.26131082e+00 -1.08308114e-01 -3.60179275e-01 6.97249115e-01
1.20486462e+00 1.37466798e-02 -1.10847425e+00 -2.05388248e-01
-3.23979378e-01 -4.80364822e-02 1.89613879e-01 -3.76286179e-01
-8.55482221e-01 1.16917145e+00 5.36463857e-02 6.18179739e-01
8.64277422e-01 -1.10192433e-01 5.65212429e-01 1.24737453e+00
4.01957706e-03 -5.57674289e-01 -7.55273461e-01 5.30722499e-01
5.71917117e-01 -1.10983539e+00 2.88040936e-01 -1.85135543e-01
-1.39805406e-01 1.77877212e+00 2.90552944e-01 7.07347095e-02
5.75502276e-01 2.81282008e-01 2.63880342e-01 -2.50457227e-01
-6.16416395e-01 -2.88540479e-02 1.96480379e-01 1.46716654e-01
1.40513271e-01 2.18403369e-01 3.45615409e-02 9.95019823e-02
-8.01068723e-01 -5.16791821e-01 8.24651301e-01 2.23150864e-01
-4.97934759e-01 -9.30943668e-01 -4.87051487e-01 1.01775378e-01
-2.37007186e-01 -1.12828493e-01 -5.50070584e-01 7.75939345e-01
-2.91251361e-01 6.36386275e-01 9.80507508e-02 -1.38253838e-01
-1.23198368e-01 1.52737856e-01 5.81808269e-01 -1.16625041e-01
-3.47581834e-01 -3.29373926e-01 1.83203414e-01 1.55246228e-01
-2.81423688e-01 -6.36049151e-01 -1.54053175e+00 -5.61303496e-01
-2.44712666e-01 7.43592620e-01 9.89729464e-01 1.23557341e+00
-1.04798444e-01 1.44992977e-01 7.84663856e-01 -9.64703560e-01
-6.62220001e-01 -7.91014552e-01 -6.92750812e-01 -7.57741183e-02
7.21644580e-01 -5.91733158e-01 -7.97935128e-01 -2.57062495e-01] | [6.697889804840088, 3.356170892715454] |
07055ba3-1104-495d-9e10-c6c8b7bfdd46 | deep-mixture-of-linear-inverse-regressions | null | null | http://openaccess.thecvf.com/content_cvpr_2017/html/Lathuiliere_Deep_Mixture_of_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Lathuiliere_Deep_Mixture_of_CVPR_2017_paper.pdf | Deep Mixture of Linear Inverse Regressions Applied to Head-Pose Estimation | Convolutional Neural Networks (ConvNets) have become the state-of-the-art for many classification and regression problems in computer vision. When it comes to regression, approaches such as measuring the Euclidean distance of target and predictions are often employed as output layer. In this paper, we propose the coupling of a Gaussian mixture of linear inverse regressions with a ConvNet, and we describe the methodological foundations and the associated algorithm to jointly train the deep network and the regression function. We test our model on the head-pose estimation problem. In this particular problem, we show that inverse regression outperforms regression models currently used by state-of-the-art computer vision methods. Our method does not require the incorporation of additional data, as it is often proposed in the literature, thus it is able to work well on relatively small training datasets. Finally, it outperforms state-of-the-art methods in head-pose estimation using a widely used head-pose dataset. To the best of our knowledge, we are the first to incorporate inverse regression into deep learning for computer vision applications.
| ['Rafael Munoz-Salinas', 'Pablo Mesejo', 'Remi Juge', 'Radu Horaud', 'Stephane Lathuiliere'] | 2017-07-01 | null | null | null | cvpr-2017-7 | ['head-pose-estimation'] | ['computer-vision'] | [-1.46763295e-01 2.32624739e-01 9.32430550e-02 -5.29308081e-01
-5.20742536e-01 -6.31958544e-02 6.99791372e-01 1.66028142e-02
-1.03506696e+00 6.01133287e-01 -1.33084767e-02 -5.28260879e-02
-1.23627409e-02 -3.82903844e-01 -8.54805410e-01 -6.35093987e-01
9.63663608e-02 7.28947043e-01 7.65952766e-02 -3.26008648e-01
1.01422131e-01 7.05548346e-01 -1.56039941e+00 -3.00812453e-01
4.56605554e-01 9.48863447e-01 -3.42334174e-02 3.83362085e-01
5.23935318e-01 4.80730623e-01 -2.89040804e-01 -2.68067032e-01
1.97040997e-02 -1.26572162e-01 -4.55755651e-01 -1.10060163e-01
6.73225939e-01 -3.18232298e-01 -4.93136585e-01 6.49466574e-01
8.45910966e-01 2.17810243e-01 9.48906600e-01 -1.16769767e+00
-5.35029829e-01 2.65127152e-01 -6.67557597e-01 1.27952337e-01
2.47386634e-01 -2.43664742e-01 7.57414997e-01 -9.52935815e-01
4.32093561e-01 1.03465366e+00 9.39289391e-01 8.06275785e-01
-1.38178241e+00 -7.01323748e-01 1.01165757e-01 4.96328086e-01
-1.44641292e+00 -7.48463869e-01 7.89222181e-01 -4.18925405e-01
9.97356057e-01 -2.32060384e-02 5.33089399e-01 1.17375243e+00
-6.65135309e-02 9.45202887e-01 1.06486118e+00 -4.82546091e-01
2.63074130e-01 1.44080371e-01 1.36819139e-01 6.08062327e-01
1.28054187e-01 1.22412838e-01 -3.96784812e-01 2.45899126e-01
6.38844788e-01 -1.93326220e-01 -2.92376161e-01 -5.35916388e-01
-7.34049499e-01 1.03157222e+00 7.76432157e-01 4.27139074e-01
-3.78020883e-01 2.81542093e-01 2.16879532e-01 -1.45675898e-01
7.70338535e-01 6.84554502e-02 -5.35358965e-01 1.22769915e-01
-1.16632521e+00 4.74478602e-01 9.06603098e-01 8.23201239e-01
2.98364997e-01 -1.08653553e-01 1.04460783e-01 9.94613111e-01
7.11029172e-01 2.45927274e-01 5.08925378e-01 -7.41232574e-01
1.88474655e-01 2.34928414e-01 -2.66723335e-01 -7.00564086e-01
-9.03912306e-01 -6.33726060e-01 -7.53345788e-01 4.07618940e-01
7.54793525e-01 -1.44006386e-01 -8.94681692e-01 1.81069565e+00
2.81150460e-01 1.25491604e-01 -3.21723312e-01 1.02729094e+00
9.86793518e-01 -1.92106869e-02 -1.01739973e-01 -1.34273589e-01
1.21490085e+00 -8.34595859e-01 -6.43229306e-01 -3.88941735e-01
6.07515872e-01 -6.76888943e-01 4.95959252e-01 5.12599766e-01
-1.05442631e+00 -5.07884324e-01 -9.86900449e-01 -2.46154666e-01
-4.20235604e-01 3.60172629e-01 4.83665317e-01 6.89294755e-01
-1.22925460e+00 6.59409285e-01 -9.84560072e-01 -6.04714155e-01
3.97184402e-01 7.83216774e-01 -6.86804354e-01 2.88637299e-02
-8.81232500e-01 1.58198237e+00 -3.99630032e-02 4.48199421e-01
-5.33569157e-01 -4.89373326e-01 -9.45130944e-01 -2.75926352e-01
2.80004352e-01 -6.94917262e-01 1.51863086e+00 -7.43055463e-01
-1.71984625e+00 1.21530712e+00 -1.91473469e-01 -5.42815626e-01
7.75254726e-01 -5.99104404e-01 -5.67742735e-02 -3.32083106e-01
-7.65321255e-02 8.46915185e-01 8.56261909e-01 -9.36626434e-01
-2.38280952e-01 -9.54660177e-01 -2.93754280e-01 -8.03129748e-02
-7.46076405e-02 1.55677218e-02 -5.63142657e-01 -2.67889738e-01
2.75205553e-01 -1.28174412e+00 -3.58419210e-01 1.68595105e-01
-3.75507146e-01 -5.32784402e-01 4.84571666e-01 -6.19963109e-01
1.01209724e+00 -1.85216379e+00 4.70402211e-01 1.12453643e-02
3.31498086e-01 3.39724511e-01 8.05945396e-02 -1.22629039e-01
-3.79941374e-01 -4.25663620e-01 -2.63726950e-01 -9.38497603e-01
1.36073247e-01 3.43628488e-02 6.38105199e-02 1.38574064e+00
-1.50612146e-01 9.05899286e-01 -5.08847237e-01 -3.41203690e-01
3.96252841e-01 9.95274723e-01 -6.90410733e-01 2.43337408e-01
5.81461526e-02 5.89617312e-01 -2.68695295e-01 5.47411181e-02
5.92561305e-01 3.22810769e-01 -1.67639598e-01 -2.36570239e-01
-2.03497961e-01 1.42401576e-01 -9.22753751e-01 1.73540390e+00
-5.39543092e-01 8.86650622e-01 -2.43452936e-02 -1.45414197e+00
1.02836907e+00 2.40550354e-01 3.95033091e-01 -5.80176532e-01
3.99763465e-01 3.17587793e-01 2.77515352e-01 -4.56131965e-01
1.87564492e-01 -1.43740565e-01 6.74272701e-02 1.70261443e-01
3.42755526e-01 -2.31055602e-01 1.56886932e-02 -5.30201256e-01
7.35552847e-01 5.57813168e-01 4.96503145e-01 -1.85748354e-01
8.97852421e-01 -4.51213032e-01 1.56596929e-01 4.60714132e-01
-9.40994546e-02 7.04816163e-01 3.17271233e-01 -5.75256765e-01
-9.44895566e-01 -6.78772390e-01 -3.87938946e-01 1.38343477e+00
-4.95962083e-01 -9.26556066e-02 -1.18153369e+00 -4.03393775e-01
7.18192980e-02 7.59458780e-01 -8.91406238e-01 4.49801423e-02
-8.80805552e-01 -6.39072835e-01 6.63837373e-01 7.18935370e-01
2.20836639e-01 -1.03998053e+00 -4.72943842e-01 1.86296538e-01
2.21108034e-01 -1.31636274e+00 -2.27247268e-01 3.54038388e-01
-7.73664236e-01 -1.00779176e+00 -9.34573889e-01 -6.24004126e-01
5.28979778e-01 -1.94672361e-01 1.03186643e+00 -6.33721724e-02
-3.74825805e-01 4.88130510e-01 3.59029956e-02 -5.62878966e-01
-4.55061011e-02 3.66926402e-01 1.17613375e-01 -1.43194310e-02
7.06262589e-01 -7.37145185e-01 -6.28501177e-01 1.86048761e-01
-5.18197238e-01 -3.01210254e-01 5.85988104e-01 6.21534109e-01
3.28805506e-01 -6.01515114e-01 1.65351585e-01 -6.87773645e-01
6.03128791e-01 -3.37303013e-01 -7.38257229e-01 -2.45067015e-01
-5.58885455e-01 3.04456890e-01 5.95340669e-01 -4.44685608e-01
-6.06623352e-01 2.99611419e-01 -4.88290220e-01 -3.93479198e-01
-7.22417593e-01 3.80462617e-01 1.10157669e-01 -2.29737371e-01
7.62983441e-01 4.03450169e-02 1.33101732e-01 -6.19150996e-01
4.44950730e-01 5.94673395e-01 5.43188095e-01 -1.65776223e-01
5.47428429e-01 5.09502709e-01 5.40600359e-01 -9.03014600e-01
-8.34574103e-01 -5.42080700e-01 -1.13536751e+00 -2.28134125e-01
1.13642776e+00 -7.05704451e-01 -1.01274109e+00 4.70823228e-01
-1.33532131e+00 -3.05950165e-01 2.28120297e-01 6.91149592e-01
-9.58033979e-01 5.45860305e-02 -3.50877553e-01 -7.78484046e-01
-1.69638410e-01 -1.32674944e+00 1.21791434e+00 4.33154777e-02
-3.45143706e-01 -1.10207009e+00 1.29785106e-01 2.28018150e-01
4.55873996e-01 1.83459148e-01 2.34858513e-01 -8.31902623e-01
-1.78323239e-01 -4.09676731e-01 -8.98585469e-02 1.22781985e-01
-3.02840441e-01 -1.29573807e-01 -1.23280692e+00 -2.93174863e-01
1.10302448e-01 -2.60436773e-01 1.10770595e+00 8.73133123e-01
1.32619107e+00 1.56829059e-01 -3.07367146e-01 7.61777639e-01
1.01355731e+00 -2.70770103e-01 6.68918490e-01 6.03291810e-01
5.61031580e-01 8.46071184e-01 3.83678585e-01 3.32009256e-01
6.47308826e-01 1.12309361e+00 5.77313900e-01 -2.05127448e-01
1.66252647e-02 1.77591539e-03 1.54200137e-01 4.63601112e-01
-4.52893376e-01 1.03973106e-01 -8.95932615e-01 2.46413648e-01
-1.99048889e+00 -7.73818910e-01 -3.19134891e-01 2.22021461e+00
3.76453906e-01 2.63882130e-02 2.84564167e-01 3.12269568e-01
5.04270375e-01 -7.17618782e-03 -4.52774972e-01 -2.27503702e-01
4.10585761e-01 4.73270535e-01 5.89712799e-01 3.92118275e-01
-1.43029010e+00 8.96745324e-01 5.73367119e+00 3.63708913e-01
-1.22756135e+00 2.30675921e-01 3.42730075e-01 -1.51555806e-01
4.91694033e-01 -4.90463525e-01 -1.12245762e+00 1.69281021e-01
9.44991529e-01 4.28485125e-01 4.90459979e-01 1.07651913e+00
2.86339790e-01 -1.62610754e-01 -1.67447186e+00 1.19768894e+00
5.17533422e-01 -7.38824606e-01 -7.27094769e-01 1.79219216e-01
2.42615297e-01 2.80897021e-01 2.07944930e-01 5.31565189e-01
-9.64707285e-02 -1.49498296e+00 7.21668243e-01 4.62431967e-01
4.92931932e-01 -4.99750733e-01 7.53600419e-01 5.61929762e-01
-8.49464297e-01 -5.82575239e-02 -2.97887176e-01 9.24051460e-03
1.51302218e-01 3.02539110e-01 -6.53015196e-01 -9.93513316e-03
8.02582502e-01 7.20247805e-01 -5.47149241e-01 1.09520316e+00
-4.90439355e-01 3.90784502e-01 -5.29253364e-01 -2.67313987e-01
1.67007953e-01 -1.93590581e-01 3.73089105e-01 1.37220454e+00
1.46714985e-01 -2.50477850e-01 -3.94822098e-02 8.10024738e-01
2.15933565e-02 1.99954346e-01 -6.13291562e-01 5.52654922e-01
-4.82044369e-02 1.31267035e+00 -4.35204476e-01 1.37701049e-01
-4.91782635e-01 6.19474411e-01 7.82369852e-01 2.45719850e-01
-7.90914297e-01 -9.37518254e-02 5.76125622e-01 1.75061777e-01
4.47058529e-01 -3.61969143e-01 -3.92246097e-01 -8.26307535e-01
3.43091935e-02 -4.99444604e-01 8.09305012e-02 -6.89953685e-01
-1.16984153e+00 6.60595953e-01 1.19947478e-01 -9.55402732e-01
-6.66314781e-01 -1.04247189e+00 -3.58476639e-01 8.52453291e-01
-1.57530761e+00 -1.16756189e+00 -3.50666642e-01 6.74411237e-01
5.06186426e-01 -2.20323935e-01 9.75079596e-01 2.79028803e-01
-5.12618542e-01 6.12167239e-01 -1.71047792e-01 2.66708016e-01
7.72973537e-01 -1.20089293e+00 3.25249642e-01 4.05719310e-01
-4.89017256e-02 6.24759793e-01 1.18699169e+00 -1.00011453e-02
-1.33683336e+00 -7.40503788e-01 7.74977982e-01 -3.28827262e-01
6.95458591e-01 -4.94841188e-01 -7.11281359e-01 7.85848498e-01
7.77554279e-03 5.67674756e-01 6.91758871e-01 3.37204099e-01
-2.85982400e-01 -1.60129055e-01 -9.94797885e-01 3.73486131e-01
7.91652679e-01 -4.97463286e-01 -6.56116486e-01 2.35641986e-01
6.25596717e-02 -7.35386431e-01 -6.95209622e-01 4.78292048e-01
8.07356536e-01 -1.17478633e+00 1.04705656e+00 -5.42922914e-01
4.66494203e-01 -5.15515022e-02 -1.08183116e-01 -1.49794793e+00
-3.04952830e-01 -1.88515767e-01 -2.69604683e-01 8.27069998e-01
5.05474329e-01 -6.34867609e-01 9.09970999e-01 6.77578390e-01
-1.22376017e-01 -9.08637464e-01 -1.22217500e+00 -4.02445346e-01
4.12287086e-01 -7.00880408e-01 2.12982725e-02 4.42435831e-01
-2.35594496e-01 4.93904620e-01 -2.56711960e-01 2.91111842e-02
7.31893063e-01 -4.60764885e-01 1.04685807e+00 -1.58526301e+00
-2.20806435e-01 -6.79346621e-01 -6.40184939e-01 -9.63756680e-01
6.99463010e-01 -6.66866362e-01 2.23357379e-01 -1.36630642e+00
-1.69693250e-02 -1.14630260e-01 -1.06756002e-01 3.30911696e-01
1.26343191e-01 4.08581614e-01 2.09750816e-01 -1.54243827e-01
-3.27061504e-01 5.06847739e-01 9.70096290e-01 6.59973174e-02
-2.27676928e-01 4.80157018e-01 -4.94928360e-01 9.77850735e-01
9.14442539e-01 -5.61134040e-01 -2.71971021e-02 -2.05339566e-01
-4.79557812e-02 -1.03045844e-01 3.32963675e-01 -9.54637289e-01
4.86908644e-01 4.29542482e-01 7.16231465e-01 -4.83933657e-01
7.24847734e-01 -9.75835443e-01 -1.43284291e-01 4.33828413e-01
-3.28792185e-01 -3.99374813e-02 1.28968462e-01 1.33665860e-01
7.07741920e-03 -3.60586375e-01 1.00991344e+00 -8.61545205e-02
-4.41242993e-01 3.39790612e-01 -2.71670401e-01 -2.02769771e-01
9.66333628e-01 -1.61685809e-01 1.90508123e-02 -5.23265123e-01
-9.47343051e-01 -8.45681131e-02 8.27723667e-02 4.91936088e-01
5.76247156e-01 -1.00108433e+00 -9.04908895e-01 3.33436698e-01
2.29370221e-02 -7.17051178e-02 -1.96106166e-01 1.44808137e+00
-3.61445606e-01 6.04649782e-01 -3.37919034e-02 -8.18568766e-01
-1.48202980e+00 5.39855838e-01 5.29792786e-01 -1.30900875e-01
-4.30741191e-01 6.89300001e-01 1.61156833e-01 -7.10037768e-01
5.18134117e-01 -1.81216076e-01 -6.01672709e-01 1.54224426e-01
3.91427666e-01 4.33349758e-01 3.56350780e-01 -1.01150537e+00
-5.53954124e-01 7.96507239e-01 -5.13437055e-02 -2.27849856e-01
1.95457888e+00 6.22724518e-02 -2.30593681e-01 3.39788973e-01
1.35793161e+00 -3.44342649e-01 -1.10187137e+00 -1.39680460e-01
5.61627857e-02 3.00693847e-02 4.55490887e-01 -3.68288189e-01
-1.18270779e+00 1.14875579e+00 7.95762300e-01 -1.20567575e-01
9.36303616e-01 6.20941333e-02 1.53203949e-01 4.32192057e-01
3.24608415e-01 -1.08127260e+00 -1.64740697e-01 5.58511376e-01
1.10645044e+00 -1.46044779e+00 2.97066510e-01 -1.68834761e-01
-4.01138246e-01 1.33519912e+00 6.24677896e-01 -6.18792474e-01
9.01180625e-01 3.59037966e-01 1.13874478e-02 4.81201671e-02
-4.52823609e-01 -6.20201170e-01 6.54145956e-01 7.12816000e-01
1.06061995e+00 3.59160751e-02 -4.40613776e-01 3.56537223e-01
-5.09621680e-01 1.74298584e-01 2.21405208e-01 6.06614769e-01
-2.70525962e-01 -1.10642767e+00 -5.95858157e-01 2.94240266e-01
-5.93226731e-01 5.22779189e-02 -2.79932857e-01 1.09763718e+00
9.86948684e-02 7.35229611e-01 2.81601939e-02 -1.23686329e-01
5.21661222e-01 -4.70331758e-02 9.56261039e-01 -4.54560518e-01
-5.02434909e-01 2.05256745e-01 -1.37526050e-01 -4.95405227e-01
-5.23785293e-01 -8.11915219e-01 -9.01831388e-01 -1.61500543e-01
-4.17328119e-01 -2.44386777e-01 1.18484771e+00 1.23838663e+00
-1.23003595e-01 4.53788042e-01 2.75097638e-01 -1.66231906e+00
-4.62150842e-01 -1.42501247e+00 -5.34695387e-01 1.94661587e-01
6.05362475e-01 -9.71937060e-01 -2.69363016e-01 -2.90046424e-01] | [13.670577049255371, 0.2902841567993164] |
ee6a46ae-0a97-487c-8620-d1f8d56bf550 | modet-learning-deformable-image-registration | 2306.05688 | null | https://arxiv.org/abs/2306.05688v1 | https://arxiv.org/pdf/2306.05688v1.pdf | ModeT: Learning Deformable Image Registration via Motion Decomposition Transformer | The Transformer structures have been widely used in computer vision and have recently made an impact in the area of medical image registration. However, the use of Transformer in most registration networks is straightforward. These networks often merely use the attention mechanism to boost the feature learning as the segmentation networks do, but do not sufficiently design to be adapted for the registration task. In this paper, we propose a novel motion decomposition Transformer (ModeT) to explicitly model multiple motion modalities by fully exploiting the intrinsic capability of the Transformer structure for deformation estimation. The proposed ModeT naturally transforms the multi-head neighborhood attention relationship into the multi-coordinate relationship to model multiple motion modes. Then the competitive weighting module (CWM) fuses multiple deformation sub-fields to generate the resulting deformation field. Extensive experiments on two public brain magnetic resonance imaging (MRI) datasets show that our method outperforms current state-of-the-art registration networks and Transformers, demonstrating the potential of our ModeT for the challenging non-rigid deformation estimation problem. The benchmarks and our code are publicly available at https://github.com/ZAX130/SmileCode. | ['Yi Wang', 'Dong Ni', 'Haiqiao Wang'] | 2023-06-09 | null | null | null | null | ['image-registration', 'medical-image-registration'] | ['computer-vision', 'medical'] | [ 5.02584875e-02 9.10203606e-02 -2.25116447e-01 -5.65185666e-01
-7.86762476e-01 -2.30783254e-01 5.30019283e-01 -4.74702924e-01
-5.18813312e-01 1.91487372e-01 4.98040348e-01 4.52446844e-03
-7.02329054e-02 -4.82150257e-01 -5.85026324e-01 -9.15617645e-01
2.09641326e-02 5.80422342e-01 2.95409977e-01 -3.00929636e-01
-2.58922335e-02 2.80497253e-01 -8.53493631e-01 8.07575881e-02
8.98531139e-01 7.68343568e-01 3.99066478e-01 -6.23420142e-02
1.32808641e-01 3.59640330e-01 -6.33905875e-03 -2.38314733e-01
2.88253456e-01 -1.81156829e-01 -1.12387168e+00 -1.59415469e-01
5.40670276e-01 -4.44284588e-01 -5.98086536e-01 1.10116792e+00
7.75547504e-01 2.32608140e-01 4.57456887e-01 -9.67333674e-01
-7.94233620e-01 7.28330791e-01 -7.42833972e-01 5.25509894e-01
6.72264397e-03 4.62099761e-02 8.94907534e-01 -1.04640770e+00
7.05037773e-01 1.12848628e+00 7.18287289e-01 8.33337426e-01
-1.13522768e+00 -7.26185501e-01 1.25326142e-01 3.48856121e-01
-1.25959349e+00 -4.20550138e-01 9.45903003e-01 -4.67496991e-01
8.78030658e-01 8.77857432e-02 6.45874560e-01 1.03717411e+00
3.26313257e-01 8.49850297e-01 9.78861392e-01 9.65086073e-02
-2.46274635e-01 -6.85506761e-01 2.26682499e-01 6.76436424e-01
-2.27837581e-02 -1.10180870e-01 -3.07304204e-01 -8.25441182e-02
1.11972594e+00 1.42050907e-01 -5.84626555e-01 -3.49694312e-01
-1.52442813e+00 7.18045950e-01 9.43417370e-01 6.58559859e-01
-3.27927709e-01 3.26806277e-01 2.36794412e-01 -4.92163897e-02
7.44605064e-01 1.26432881e-01 -1.94632322e-01 1.52821049e-01
-8.16232920e-01 1.92322403e-01 1.58615142e-01 5.66795468e-01
4.21098888e-01 -4.18835655e-02 -3.13431054e-01 9.60490644e-01
5.99299908e-01 2.84253508e-01 8.52418482e-01 -9.00896251e-01
5.13281167e-01 5.23812771e-01 -3.82496357e-01 -8.37657154e-01
-7.71908998e-01 -4.04805422e-01 -1.10181940e+00 -1.43091738e-01
4.00233984e-01 9.96437594e-02 -9.35387671e-01 2.03492475e+00
4.74513829e-01 4.93585587e-01 -4.81075197e-01 1.15973008e+00
1.01174664e+00 9.80996117e-02 2.38842331e-02 7.84936026e-02
1.37333894e+00 -1.09380901e+00 -7.15602875e-01 -3.37396473e-01
6.61089361e-01 -6.20257139e-01 8.98043096e-01 -1.84738219e-01
-1.25926137e+00 -3.93499881e-01 -6.82522595e-01 -2.84627914e-01
1.22855797e-01 -7.23319650e-02 7.15675175e-01 2.59989589e-01
-1.15596533e+00 6.57442927e-01 -1.42805612e+00 -1.26543520e-02
8.53834867e-01 6.14650786e-01 -5.28968036e-01 -6.60987571e-02
-1.19348860e+00 1.06905377e+00 2.54896004e-02 5.13086140e-01
-5.93708038e-01 -8.46595168e-01 -9.28378820e-01 -2.00409979e-01
8.93023517e-03 -8.64312291e-01 1.13084173e+00 -6.90944970e-01
-1.36631048e+00 1.06273949e+00 -1.96454003e-01 -1.28692687e-01
5.70025980e-01 -2.72529393e-01 -8.23727325e-02 1.42234489e-01
1.59061179e-01 8.12991321e-01 7.13992596e-01 -7.29180217e-01
1.68836102e-01 -6.41757369e-01 -3.69430259e-02 2.49357685e-01
-3.05760294e-01 1.83318064e-01 -4.99357134e-01 -1.05343771e+00
4.41993892e-01 -1.12349629e+00 -4.14433271e-01 4.14507948e-02
-3.22490484e-01 -2.97007263e-01 6.01238251e-01 -7.53175855e-01
1.02187109e+00 -1.91715264e+00 6.24237537e-01 6.92734197e-02
4.96745884e-01 1.17485441e-01 -2.57162929e-01 -1.66325495e-01
-2.82339782e-01 -1.28999017e-02 -5.87084532e-01 -4.81541932e-01
-8.43476430e-02 1.57537118e-01 4.08207579e-03 8.02153766e-01
8.57855007e-02 1.33842981e+00 -8.21077347e-01 -4.09772635e-01
2.64784712e-02 8.31991494e-01 -7.99723625e-01 1.55960247e-01
2.00791076e-01 9.41370010e-01 -5.82893133e-01 4.27584589e-01
7.63011694e-01 -3.69407445e-01 1.59904078e-01 -6.70210183e-01
2.68469691e-01 3.07972610e-01 -7.20071137e-01 2.21835685e+00
-1.62988320e-01 3.13410729e-01 1.06828719e-01 -1.07616365e+00
5.25177896e-01 4.30221915e-01 1.04471791e+00 -4.70153511e-01
4.27843779e-01 2.20390320e-01 3.78831029e-01 -5.20956576e-01
2.96638608e-02 -1.47441149e-01 2.11099878e-01 5.35953462e-01
2.47894257e-01 3.74996550e-02 -8.50992352e-02 -6.90751821e-02
9.00279880e-01 3.78537863e-01 2.90341415e-02 -4.63312626e-01
5.66622794e-01 -4.69198436e-01 9.21166420e-01 2.95141131e-01
-3.58743668e-01 8.95956218e-01 9.91729274e-02 -6.19557798e-01
-6.46078289e-01 -9.53959763e-01 -3.70976001e-01 9.14342046e-01
9.88394395e-02 -9.09980610e-02 -1.06895912e+00 -7.10505724e-01
-1.58796951e-01 1.65758673e-02 -7.85459578e-01 -2.55195290e-01
-1.04703033e+00 -1.14705956e+00 5.06373823e-01 7.42582917e-01
6.09779179e-01 -1.14571166e+00 -4.64070022e-01 2.19583035e-01
-6.47210419e-01 -1.16109300e+00 -1.20444953e+00 -3.24016124e-01
-1.03154504e+00 -9.21821415e-01 -1.00146115e+00 -8.40812027e-01
7.74181128e-01 1.79418400e-01 8.08592021e-01 1.57377258e-01
-2.39812732e-01 3.13737929e-01 -1.32410944e-01 8.56648982e-02
-7.50894248e-02 3.10324311e-01 8.43144208e-02 1.39100000e-01
1.22795671e-01 -8.86085868e-01 -8.30474079e-01 3.18280011e-01
-8.93096149e-01 2.11831599e-01 4.75156963e-01 8.48235011e-01
5.96069753e-01 -7.27848887e-01 4.90330338e-01 -8.33574116e-01
4.17499572e-01 -4.93387491e-01 -2.28696331e-01 1.27570987e-01
-4.49032128e-01 2.43710890e-01 3.59075032e-02 -6.75510466e-01
-8.80956948e-01 -1.91715658e-02 -5.41736901e-01 -5.30508637e-01
6.47591576e-02 6.15189373e-01 -1.79565594e-01 -3.62515122e-01
2.79012173e-01 8.33990499e-02 1.30965590e-01 -6.15201771e-01
2.31250718e-01 5.41820936e-02 6.16113782e-01 -5.49776971e-01
8.67206395e-01 5.77572703e-01 4.71029915e-02 -2.50940025e-01
-8.17034602e-01 -3.26857120e-01 -9.58770156e-01 -1.13831319e-01
1.22035646e+00 -8.08925450e-01 -4.64413911e-01 7.69200623e-01
-1.23436451e+00 -5.12189448e-01 1.49616087e-03 5.74747801e-01
-5.04239917e-01 4.59787756e-01 -7.96559572e-01 -3.40766422e-02
-7.29562044e-01 -1.68402469e+00 1.11394024e+00 6.53633475e-02
-1.18107922e-01 -1.09401464e+00 7.80265480e-02 5.54605722e-01
7.52768040e-01 3.32009852e-01 7.83909738e-01 -4.57132608e-01
-4.38889325e-01 6.53616432e-03 2.66265823e-03 1.25177158e-02
2.09146291e-01 -5.96302688e-01 -8.94741952e-01 -4.71696883e-01
1.05992064e-01 -1.47197813e-01 1.08156478e+00 6.83056474e-01
1.38818741e+00 -9.54020172e-02 -2.09370136e-01 1.17641771e+00
1.04875326e+00 -1.45455137e-01 7.21180737e-01 2.71802336e-01
1.22419071e+00 4.01988536e-01 4.32956479e-02 1.56515893e-02
8.22885156e-01 1.09421051e+00 5.16234457e-01 -2.40133494e-01
-2.83320427e-01 2.21533760e-01 4.13206905e-01 1.21580267e+00
-5.87797582e-01 3.80806744e-01 -1.12899017e+00 4.77444857e-01
-1.83208704e+00 -9.46878254e-01 -2.05206834e-02 2.00249434e+00
1.00667775e+00 -2.37352028e-01 2.59119347e-02 -3.26943666e-01
6.09742582e-01 3.16142976e-01 -5.38213730e-01 1.52903035e-01
4.24693711e-02 3.24922144e-01 2.56260931e-01 5.24294376e-01
-1.27844477e+00 7.76437044e-01 5.57120657e+00 6.26476288e-01
-1.16718173e+00 7.40487158e-01 5.11881053e-01 7.78691024e-02
-3.67993683e-01 -1.49341136e-01 -4.72903132e-01 3.71540993e-01
4.88416940e-01 4.16918062e-02 4.21111465e-01 4.05776918e-01
1.55386463e-01 4.51895565e-01 -9.58430529e-01 8.87609243e-01
2.52802353e-02 -1.27763188e+00 3.43969911e-02 1.27029046e-01
5.48238695e-01 5.56578517e-01 1.35395706e-01 3.43297608e-02
8.65789317e-03 -1.09211433e+00 6.85440779e-01 5.30896842e-01
6.59698427e-01 -3.87632459e-01 7.27110028e-01 4.39390317e-02
-1.37664020e+00 1.60604879e-01 -1.95515752e-01 4.16932344e-01
3.30627561e-01 3.66850197e-01 -3.15509588e-01 6.52303755e-01
8.29101145e-01 1.00637901e+00 -5.60018241e-01 9.91648257e-01
-2.35057518e-01 4.45404857e-01 -1.60827130e-01 6.83373451e-01
1.04191557e-01 -2.82857627e-01 6.32946610e-01 9.98053908e-01
2.67487913e-01 6.68302774e-02 1.76122576e-01 9.93399382e-01
-2.78365642e-01 2.40421370e-02 -2.47858644e-01 3.53041589e-01
1.82082877e-01 1.50021172e+00 -6.61757410e-01 -3.67050022e-02
-5.01316249e-01 9.09569979e-01 4.01107728e-01 2.57288724e-01
-9.39782679e-01 8.45552161e-02 7.91703045e-01 1.57528639e-01
1.46957293e-01 -1.62237912e-01 -1.10166781e-01 -1.51439512e+00
1.72240332e-01 -7.26318717e-01 2.24068940e-01 -5.35993814e-01
-1.46310043e+00 8.78087759e-01 1.43473759e-01 -1.09209621e+00
-2.25468315e-02 -4.57999349e-01 -7.57302225e-01 8.63083005e-01
-1.43885052e+00 -1.53477216e+00 -4.32070524e-01 7.89878905e-01
3.40917796e-01 3.45356539e-02 7.82028139e-01 6.62259817e-01
-6.50538445e-01 6.90600276e-01 -3.11304957e-01 4.90857244e-01
7.97731996e-01 -1.01016402e+00 6.06078982e-01 8.48974347e-01
-4.01016101e-02 9.02729690e-01 1.37845650e-01 -4.78498459e-01
-1.34787178e+00 -1.12025332e+00 6.73952699e-01 -4.67931271e-01
6.84401512e-01 -3.56724381e-01 -1.27017069e+00 7.92603731e-01
2.12997437e-01 5.75531125e-01 5.10594487e-01 -2.71495610e-01
-4.74269569e-01 -1.61204115e-02 -1.08319616e+00 3.66681308e-01
1.30942345e+00 -4.61344153e-01 -6.78899765e-01 2.83874214e-01
6.61988914e-01 -7.77182221e-01 -1.28417301e+00 5.94504058e-01
5.28805375e-01 -5.50710559e-01 1.19403744e+00 -5.13116181e-01
4.22314048e-01 -1.98196247e-01 6.87993467e-02 -1.30347800e+00
-5.92066586e-01 -6.02814257e-01 -1.08840279e-01 1.04923332e+00
8.44263285e-02 -8.47035587e-01 4.43996459e-01 7.68244565e-01
-4.16881859e-01 -9.99746144e-01 -1.39767385e+00 -4.77923810e-01
5.52409291e-01 -2.91024685e-01 6.25697315e-01 1.29141569e+00
-1.42045930e-01 1.88136056e-01 -2.68481672e-01 -3.41013446e-02
7.53160119e-01 4.45228927e-02 1.75025746e-01 -1.25108552e+00
-3.02040726e-01 -7.42688715e-01 -3.92046869e-01 -9.94586766e-01
4.35443729e-01 -1.53894353e+00 -1.36063010e-01 -1.52073050e+00
5.03982782e-01 -3.69251043e-01 -3.81767511e-01 9.40642655e-01
-2.58417994e-01 4.97310847e-01 1.95829868e-01 4.14760411e-01
-2.66476035e-01 6.39954686e-01 1.69208372e+00 -1.75802901e-01
6.09999858e-02 -7.91547149e-02 -7.12673664e-01 8.94517541e-01
8.10995102e-01 -4.42206711e-01 -3.18063736e-01 -9.07777607e-01
-9.04390737e-02 -4.19492275e-02 5.80692589e-01 -7.28157759e-01
2.26869807e-01 6.50415644e-02 2.08562970e-01 -2.13896677e-01
1.47773921e-01 -6.89242303e-01 5.69965318e-02 4.31185603e-01
-2.15489402e-01 3.81668836e-01 5.58858439e-02 2.15207100e-01
-1.31754816e-01 1.34532958e-01 9.19516861e-01 -5.83180599e-02
-3.86605531e-01 1.09887731e+00 -7.25882575e-02 3.00770432e-01
5.85477233e-01 1.02712713e-01 -2.76024759e-01 -2.60482766e-02
-8.94836843e-01 1.47940189e-01 2.51861542e-01 5.99605322e-01
6.90913022e-01 -1.71032536e+00 -8.17965984e-01 1.50641590e-01
-1.29679844e-01 1.89481139e-01 3.71567249e-01 1.68021393e+00
-2.14807957e-01 1.15091801e-01 -3.79903615e-01 -6.49507999e-01
-9.69278336e-01 1.37074903e-01 7.08739698e-01 -4.19738740e-01
-1.02968764e+00 7.90124357e-01 5.06649673e-01 -5.50202668e-01
-6.11536875e-02 -6.12369359e-01 -3.79692197e-01 -1.65917560e-01
4.25877005e-01 5.07734753e-02 8.20034146e-02 -1.18255532e+00
-6.76110566e-01 9.18125868e-01 -1.24988578e-01 -4.15645540e-02
1.76033151e+00 -4.42113407e-04 -4.79584068e-01 -1.53038595e-02
1.25534749e+00 -3.40484232e-01 -1.21096444e+00 -6.12738848e-01
1.74048971e-02 -2.02711448e-01 1.64054409e-01 -4.01592910e-01
-1.83078706e+00 9.57765400e-01 7.87466645e-01 -4.09598470e-01
1.21115041e+00 1.42556652e-01 1.10761678e+00 6.90008551e-02
4.52387989e-01 -6.14935040e-01 -5.71241714e-02 4.93622214e-01
1.06501400e+00 -1.15276742e+00 -1.52747437e-01 -3.74481797e-01
-5.45834899e-01 9.97600257e-01 6.56198144e-01 -3.40809822e-01
8.90872717e-01 3.79246980e-01 -4.61164005e-02 -3.73227149e-01
-3.02246630e-01 -9.44128633e-03 8.37403715e-01 4.79383916e-01
7.98361778e-01 -2.60188803e-02 -4.24139112e-01 8.05016816e-01
6.63151266e-03 -5.36887236e-02 2.30926216e-01 6.59906268e-01
-4.93605472e-02 -1.40553069e+00 -1.90675482e-01 3.97539467e-01
-7.34492183e-01 -1.75194547e-01 -7.82944262e-02 5.40767372e-01
9.20224562e-02 3.86662990e-01 -1.06116064e-01 -3.52227300e-01
2.17193663e-01 -2.62363218e-02 8.85178268e-01 -4.74642456e-01
-8.10280204e-01 3.33712459e-01 -3.78081262e-01 -9.27268863e-01
-7.86510229e-01 -8.39613557e-01 -1.46161234e+00 -6.68457225e-02
-8.06905478e-02 -2.49678969e-01 1.63690776e-01 1.05716586e+00
3.90518308e-01 5.47581673e-01 3.45962465e-01 -1.32367408e+00
-4.36360061e-01 -1.09296978e+00 -1.47249803e-01 6.87568247e-01
2.65030503e-01 -1.05060160e+00 2.84815375e-02 1.18089365e-02] | [14.006367683410645, -2.5897305011749268] |
9ec2e938-c024-4a5f-ba42-37e1814d28b9 | on-the-confidence-of-neural-network | null | null | https://openreview.net/forum?id=HJf2ds2ssm | https://openreview.net/pdf?id=HJf2ds2ssm | On the Confidence of Neural Network Predictions for some NLP Tasks | Neural networks are known to produce unexpected results on inputs that are far from the training distribution. One approach to tackle this problem is to detect the samples on which the trained network can not answer reliably. ODIN is a recently proposed method for out-of-distribution detection that does not modify the trained network and achieves good performance for various image classification tasks. In this paper we adapt ODIN for sentence classification and word tagging tasks. We show that the scores produced by ODIN can be used as a confidence measure for the predictions on both in-distribution and out-of-distribution datasets. | ['Anonymous'] | 2018-10-22 | null | null | null | null | ['sentence-classification'] | ['natural-language-processing'] | [ 6.51389807e-02 6.64519891e-02 -3.21161598e-01 -9.32542086e-01
-7.63371468e-01 -6.28086388e-01 4.76520896e-01 4.20657098e-01
-4.76092547e-01 1.10164726e+00 -3.12298745e-01 -2.49990955e-01
6.14158250e-02 -7.27060139e-01 -8.46496820e-01 -7.67890394e-01
1.04915805e-01 6.44976556e-01 7.72946298e-01 3.39460820e-01
3.05940002e-01 6.45398736e-01 -1.40197933e+00 6.80267930e-01
4.99833494e-01 1.09162438e+00 8.69422033e-02 9.75906372e-01
-3.18402857e-01 9.49230313e-01 -1.30522346e+00 -4.27862853e-01
-1.48559988e-01 -5.09001911e-01 -7.31156290e-01 -2.10941672e-01
4.76591170e-01 -3.77681889e-02 -3.12822647e-02 1.50408590e+00
5.59646606e-01 -7.50840157e-02 1.19626558e+00 -1.12750936e+00
-5.61849535e-01 5.35719395e-01 -1.11975625e-01 6.91911340e-01
2.14202479e-01 -2.73239404e-01 7.23522663e-01 -8.97583365e-01
5.28923631e-01 1.09164405e+00 6.32447720e-01 6.29179478e-01
-1.20177734e+00 -5.24827898e-01 -2.32328922e-01 1.33286923e-01
-1.25935733e+00 -1.20993704e-01 5.15643060e-01 -6.28489375e-01
1.09997499e+00 1.29393131e-01 2.69628782e-02 1.13116384e+00
5.28237820e-01 8.17874432e-01 1.06945336e+00 -6.47601843e-01
7.17820883e-01 5.98755121e-01 3.70269954e-01 4.02156115e-01
1.72393784e-01 1.82121024e-01 -3.87704521e-01 -2.46148661e-01
3.11973959e-01 -2.96504736e-01 -7.88400471e-02 -9.60617885e-02
-6.03500426e-01 9.39569771e-01 3.07139844e-01 3.44476193e-01
-1.82506636e-01 -2.73431633e-02 5.50437868e-01 4.09081489e-01
9.06603158e-01 5.01249492e-01 -8.24338138e-01 -1.19152464e-01
-8.76976848e-01 1.31410897e-01 8.08264732e-01 7.11774945e-01
4.97475505e-01 3.66905262e-03 -4.76559728e-01 1.07667208e+00
-5.49566373e-02 5.34210086e-01 7.73441911e-01 -3.09614003e-01
7.76459500e-02 2.41606563e-01 -6.05206862e-02 -5.78707159e-01
-4.35411274e-01 -4.12888557e-01 -6.93693042e-01 3.02057534e-01
6.15785480e-01 -3.66200894e-01 -1.22790515e+00 1.54781365e+00
4.77424311e-03 -8.65342617e-02 1.56851113e-01 5.10106087e-01
7.33783007e-01 6.58636570e-01 4.38979790e-02 7.26162419e-02
9.02629077e-01 -5.56878150e-01 -4.34515208e-01 -1.83774546e-01
6.88935637e-01 -7.97516406e-01 8.51190507e-01 5.79352498e-01
-5.71965575e-01 -6.45012617e-01 -9.05374587e-01 5.18572628e-01
-4.26762581e-01 2.55086124e-01 2.61236161e-01 6.62882328e-01
-8.85758758e-01 7.80784786e-01 -4.24443573e-01 -1.31391346e-01
6.78783000e-01 4.08777833e-01 -1.66103452e-01 -8.92024413e-02
-9.80414748e-01 1.03307521e+00 7.26006269e-01 -3.65509778e-01
-8.03478837e-01 -2.39221618e-01 -6.22450769e-01 2.83640791e-02
-1.90823361e-01 5.34884864e-03 1.40852332e+00 -1.22703278e+00
-9.97791886e-01 8.84993374e-01 -8.13048407e-02 -6.65318727e-01
2.63817877e-01 4.66910601e-02 -5.98830163e-01 -6.87239915e-02
4.20372374e-02 6.64059818e-01 9.74034250e-01 -8.53302240e-01
-6.35466158e-01 -2.98998266e-01 -5.05619347e-01 -4.24192607e-01
-2.91772723e-01 9.54305753e-02 1.04554921e-01 -5.94469607e-01
-1.56667188e-01 -7.21851885e-01 -8.49176422e-02 -3.12598735e-01
-6.73218548e-01 -8.24899256e-01 6.04951084e-01 -3.68851066e-01
9.90930319e-01 -2.03460932e+00 -5.86882114e-01 3.01702172e-01
-5.37056960e-02 5.37469149e-01 -1.70024946e-01 2.90487520e-02
-2.52654046e-01 -3.25607248e-02 8.17984864e-02 -4.32201847e-02
-1.81879364e-02 4.15033609e-01 -4.43402052e-01 4.12008196e-01
6.05179071e-01 3.11576128e-01 -7.99055994e-01 -2.85859972e-01
-1.77990422e-01 1.82969108e-01 -2.31543392e-01 5.01060367e-01
-5.04109502e-01 2.18712211e-01 -7.03373104e-02 4.78966415e-01
7.70997047e-01 -1.22319467e-01 -8.20044279e-02 1.83585003e-01
4.32253897e-01 3.93900782e-01 -6.75527394e-01 8.09183776e-01
-2.16544539e-01 1.06832528e+00 -6.70698524e-01 -1.18361866e+00
1.26124942e+00 8.77212361e-02 -2.79488862e-01 -6.58001244e-01
3.63312781e-01 3.41233045e-01 2.71783084e-01 -6.34925127e-01
4.37247187e-01 -4.78587508e-01 -1.05063990e-01 2.07466111e-01
6.57645881e-01 3.71450782e-02 2.15987608e-01 -1.19682148e-01
1.19021165e+00 -3.00904900e-01 3.03242594e-01 -3.21404010e-01
2.09129915e-01 -1.32284507e-01 4.44719911e-01 1.03725660e+00
-1.30454168e-01 6.61091506e-01 1.05773306e+00 -3.70422751e-01
-1.19328785e+00 -1.21811366e+00 -5.70592761e-01 1.09020007e+00
-3.42983156e-01 5.48341908e-02 -8.02928507e-01 -1.31758463e+00
5.23958057e-02 9.21310425e-01 -6.02880180e-01 -2.59964347e-01
1.15903944e-01 -8.38882148e-01 6.80365682e-01 6.74732745e-01
-7.27933869e-02 -1.20142603e+00 -1.32336155e-01 -1.01869293e-01
2.96863794e-01 -9.88421917e-01 -5.69132566e-02 9.77143645e-01
-8.33238721e-01 -9.17976558e-01 -8.76256764e-01 -9.65849161e-01
9.21359539e-01 -3.36334288e-01 1.38683867e+00 -1.88269868e-01
-1.64835185e-01 -1.49838522e-01 -4.33990657e-01 -8.01458657e-01
-7.59542763e-01 2.07337886e-02 1.12486973e-01 -3.00784647e-01
9.64193940e-01 -1.53249040e-01 -3.59374136e-02 3.72408897e-01
-1.00332904e+00 -7.19514966e-01 5.19010961e-01 1.03739750e+00
4.07094419e-01 3.18755597e-01 1.12046289e+00 -1.25333226e+00
8.82170796e-01 -6.96378708e-01 -6.93408847e-01 1.67253256e-01
-6.74334228e-01 3.90499830e-01 7.73533463e-01 -7.10701704e-01
-5.46219051e-01 5.41305766e-02 -7.26581514e-01 -3.70962352e-01
-6.10399604e-01 4.00533319e-01 1.86317451e-02 2.29441673e-01
1.01561153e+00 9.63281989e-02 -1.58346757e-01 -5.35309672e-01
-3.17545414e-01 1.05441868e+00 3.19774389e-01 -1.49637133e-01
3.13698202e-01 -2.75585949e-01 -1.78930238e-01 -8.48166049e-01
-1.31439853e+00 -5.34557521e-01 -5.55218458e-01 -1.92094922e-01
6.31980717e-01 -6.38628781e-01 -2.21042365e-01 6.73679292e-01
-1.32839727e+00 -4.27272558e-01 -2.94227362e-01 5.98370314e-01
-2.14262620e-01 9.13496614e-02 -3.25231165e-01 -1.08847022e+00
3.73634212e-02 -8.43246579e-01 7.36676633e-01 4.13080424e-01
-5.82956493e-01 -1.22523046e+00 1.65710509e-01 -3.46110374e-01
2.19943777e-01 -2.06127778e-01 9.94328916e-01 -1.54227328e+00
1.55652110e-02 -6.50001228e-01 -1.93598971e-01 1.10938478e+00
-7.84447938e-02 -4.95855324e-02 -1.32869661e+00 -9.35001522e-02
-7.65684247e-02 -4.67848510e-01 1.04703259e+00 5.41722119e-01
1.32596564e+00 -9.89642441e-02 -1.87367573e-01 6.88749701e-02
1.30811751e+00 2.82768190e-01 7.70150900e-01 9.29602087e-02
9.67776328e-02 4.96685624e-01 5.92813492e-01 3.65433156e-01
-3.60146165e-01 5.47724485e-01 2.46238425e-01 6.52569607e-02
1.80619105e-03 -1.78175449e-01 4.89589721e-01 3.23226750e-01
7.18722761e-01 -7.17162609e-01 -8.97620082e-01 8.09957564e-01
-1.60459018e+00 -9.61046338e-01 -2.98339456e-01 2.09866095e+00
9.55350220e-01 6.82785988e-01 2.47359022e-01 1.04402035e-01
8.21685314e-01 -1.72870770e-01 -5.11978209e-01 -8.28713179e-01
-2.00439915e-01 4.80642945e-01 5.08956671e-01 3.23521763e-01
-1.14398193e+00 6.23139739e-01 7.55253553e+00 1.11733449e+00
-1.13974583e+00 9.96666402e-02 1.06311893e+00 2.97198057e-01
-1.21952184e-01 -4.98906136e-01 -1.10517991e+00 7.79920936e-01
1.26579523e+00 2.28685111e-01 -3.77797216e-01 1.31779611e+00
-2.87372023e-01 -5.48834682e-01 -1.17358935e+00 9.23764884e-01
3.04194868e-01 -1.06755006e+00 -2.10572794e-01 -1.43445030e-01
8.37747455e-01 3.12224686e-01 1.65833712e-01 3.72163087e-01
3.32290471e-01 -1.00153148e+00 5.20323098e-01 5.45230985e-01
6.57113850e-01 -9.82493639e-01 1.42442548e+00 6.65051281e-01
-1.38462842e-01 2.91967511e-01 -9.67461228e-01 -4.76182103e-02
-4.01649475e-01 1.25508666e+00 -1.70348799e+00 -3.00759971e-01
5.22992551e-01 3.67573023e-01 -6.03764117e-01 1.41520727e+00
-4.35530752e-01 9.72378373e-01 -1.51372388e-01 -5.62901497e-01
4.17613015e-02 4.80083436e-01 2.30905920e-01 1.45142686e+00
4.11022276e-01 -5.77663720e-01 5.54680638e-02 6.67952836e-01
-1.95102021e-01 -1.98455036e-01 -8.69242668e-01 -4.90599647e-02
1.05772771e-01 7.67860174e-01 -6.81882441e-01 -5.06491899e-01
-5.58904046e-03 1.13963950e+00 5.16300261e-01 1.31962806e-01
-6.06110394e-01 -6.74512327e-01 5.89990199e-01 -4.42415290e-02
6.24917805e-01 1.35799453e-01 -1.67370602e-01 -1.03537786e+00
-1.77715737e-02 -2.77492225e-01 3.28209668e-01 -7.32043326e-01
-1.97525465e+00 9.53476667e-01 -1.23010166e-01 -1.21058977e+00
-6.57692432e-01 -1.18013632e+00 -5.74643373e-01 9.55756307e-01
-1.41843808e+00 -2.38243774e-01 8.01471546e-02 2.56743908e-01
3.69367927e-01 -5.11667907e-01 1.09433568e+00 1.80427164e-01
-2.59907663e-01 6.69382274e-01 4.92772847e-01 4.86982822e-01
7.86999464e-01 -1.49892700e+00 1.57345429e-01 6.90329790e-01
3.71158957e-01 1.42524689e-02 9.19429183e-01 -5.31268537e-01
-2.23062381e-01 -1.35220051e+00 1.30839646e+00 -4.90083367e-01
3.73829871e-01 -2.95839250e-01 -9.95550990e-01 4.27080870e-01
-2.62677018e-02 2.96548396e-01 9.41715121e-01 2.95924097e-01
-2.78090268e-01 -6.10892475e-02 -1.26667225e+00 -7.59462342e-02
2.72550732e-01 -5.79126000e-01 -6.13256991e-01 6.28525257e-01
2.45731562e-01 -3.99207294e-01 -5.53061664e-01 1.10596836e-01
2.81939030e-01 -1.15644109e+00 5.58994830e-01 -8.93673658e-01
5.71058333e-01 -7.24315038e-03 -2.88688719e-01 -1.58271945e+00
-9.68954414e-02 -8.92923176e-02 9.47273970e-02 1.18264186e+00
7.16867149e-01 -5.43655217e-01 9.81090486e-01 2.15770528e-01
6.77831247e-02 -7.40743697e-01 -1.01949918e+00 -1.12484920e+00
8.94417614e-02 -6.47307336e-01 2.45674834e-01 5.95259905e-01
-1.32383361e-01 5.37722170e-01 -2.72069305e-01 1.42012075e-01
5.01620471e-01 -2.89176822e-01 4.02932793e-01 -1.32521546e+00
-2.81595707e-01 -2.13125855e-01 -9.90959823e-01 -1.19555533e+00
4.86080706e-01 -9.06078815e-01 5.70566237e-01 -1.13620806e+00
2.16073662e-01 -3.79356652e-01 -5.55592239e-01 3.64532977e-01
-4.05977219e-02 5.51496208e-01 -2.39970520e-01 8.02224278e-02
-7.02791512e-01 3.08251262e-01 7.15003729e-01 -1.48320645e-01
2.78570443e-01 4.23418969e-01 -3.28975290e-01 9.68557596e-01
1.03294063e+00 -1.13425362e+00 -3.09462577e-01 2.10742746e-02
1.19504035e-01 -2.17142329e-01 3.42032105e-01 -1.24003029e+00
-1.23509943e-01 3.18669565e-02 8.28383327e-01 -8.50675821e-01
-6.28923401e-02 -6.35006368e-01 -4.44358379e-01 3.35453451e-01
-7.27545500e-01 5.17840963e-03 2.04458505e-01 5.60543120e-01
-4.60345089e-01 -9.69129741e-01 1.08279598e+00 8.99247080e-02
-6.99870765e-01 -1.31948054e-01 -6.71585858e-01 1.49148449e-01
9.58941281e-01 1.77660912e-01 -4.39921170e-01 -3.62986475e-01
-9.35288012e-01 -2.69129246e-01 1.01009369e-01 2.26691946e-01
6.50919437e-01 -1.24294651e+00 -5.25573552e-01 3.54923159e-01
2.07381532e-01 -2.14004487e-01 -3.02725453e-02 5.97306609e-01
-4.88000512e-01 5.67943871e-01 -1.05331376e-01 -8.32129478e-01
-1.50019908e+00 4.08791274e-01 4.16334063e-01 -4.20085847e-01
8.65531806e-03 1.28049219e+00 4.28313203e-02 -5.26875973e-01
3.19669515e-01 -4.97686446e-01 -1.54645070e-01 3.70334787e-03
7.60013759e-01 -1.61542669e-01 4.04233575e-01 -2.10532531e-01
-3.95212650e-01 -8.56719986e-02 -3.43824893e-01 1.12775676e-01
1.28382194e+00 2.97537982e-01 -2.99971066e-02 8.55943978e-01
1.47479045e+00 -1.88779429e-01 -1.02753639e+00 -2.47772247e-01
2.79837668e-01 -5.65207958e-01 1.36681078e-02 -1.22245133e+00
-6.66388988e-01 1.19209898e+00 8.03473949e-01 7.55726457e-01
9.58253443e-01 4.04399395e-01 6.04171395e-01 5.08423865e-01
1.57934025e-01 -1.24840415e+00 2.79651642e-01 8.57443750e-01
6.01351321e-01 -1.32964146e+00 -4.17643815e-01 -1.22371793e-01
-6.38895631e-01 1.39488769e+00 7.12774575e-01 -2.97207355e-01
8.76965940e-01 5.27483463e-01 1.35492608e-01 3.98027301e-02
-8.81773353e-01 -1.11553580e-01 4.44238842e-01 9.93443906e-01
5.35962760e-01 9.07866880e-02 -1.03794988e-02 5.25419354e-01
-4.88716103e-02 -5.68338446e-02 6.74047947e-01 6.49132490e-01
-7.78255880e-01 -1.17793810e+00 -2.40490094e-01 9.48343456e-01
-7.39372253e-01 -1.31175727e-01 -5.18405914e-01 2.54382074e-01
1.82651490e-01 7.30613589e-01 3.80843312e-01 -5.62030196e-01
1.95429176e-01 3.56504142e-01 3.88091058e-01 -8.94290447e-01
-3.17095339e-01 -1.37350991e-01 3.52949589e-01 -1.22053780e-01
-2.53652394e-01 -5.00883222e-01 -1.15012312e+00 1.64104596e-01
-5.70466638e-01 2.52240688e-01 6.00801349e-01 8.98932755e-01
2.40779757e-01 5.86332202e-01 6.66853309e-01 -4.05658036e-01
-8.37626100e-01 -1.19819355e+00 -9.58081305e-01 4.04602379e-01
6.01929724e-01 -5.27722657e-01 -5.05011082e-01 -1.09697886e-01] | [9.315056800842285, 3.04284930229187] |
fdf7e52d-ac8c-4eab-8b45-622e92a3bead | neural-sequential-phrase-grounding-seqground | 1903.07669 | null | http://arxiv.org/abs/1903.07669v1 | http://arxiv.org/pdf/1903.07669v1.pdf | Neural Sequential Phrase Grounding (SeqGROUND) | We propose an end-to-end approach for phrase grounding in images. Unlike
prior methods that typically attempt to ground each phrase independently by
building an image-text embedding, our architecture formulates grounding of
multiple phrases as a sequential and contextual process. Specifically, we
encode region proposals and all phrases into two stacks of LSTM cells, along
with so-far grounded phrase-region pairs. These LSTM stacks collectively
capture context for grounding of the next phrase. The resulting architecture,
which we call SeqGROUND, supports many-to-many matching by allowing an image
region to be matched to multiple phrases and vice versa. We show competitive
performance on the Flickr30K benchmark dataset and, through ablation studies,
validate the efficacy of sequential grounding as well as individual design
choices in our model architecture. | ['Leonid Sigal', 'Pelin Dogan', 'Markus Gross'] | 2019-03-18 | neural-sequential-phrase-grounding-seqground-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Dogan_Neural_Sequential_Phrase_Grounding_SeqGROUND_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Dogan_Neural_Sequential_Phrase_Grounding_SeqGROUND_CVPR_2019_paper.pdf | cvpr-2019-6 | ['phrase-grounding'] | ['natural-language-processing'] | [ 4.58602011e-01 3.42539757e-01 -3.50316316e-01 -5.47584832e-01
-1.16974640e+00 -7.10478246e-01 7.85128534e-01 2.78892875e-01
-4.04578894e-01 3.20490390e-01 6.85073197e-01 -3.96144867e-01
2.90105492e-01 -7.76550770e-01 -1.10992634e+00 -4.91267562e-01
9.26198065e-02 2.12931678e-01 1.43417194e-01 -2.99517941e-02
1.46718770e-01 1.11017726e-01 -1.01715529e+00 8.60275924e-01
1.80521950e-01 8.66837084e-01 2.78106093e-01 6.06936932e-01
-1.74363166e-01 4.49143469e-01 -2.93310165e-01 -6.41010404e-01
3.82070690e-01 -3.16299498e-01 -7.94137001e-01 1.81872413e-01
1.11265171e+00 -5.77848673e-01 -5.50178289e-01 8.80467474e-01
2.55929798e-01 9.97673795e-02 3.51699889e-01 -9.85304296e-01
-1.36603403e+00 8.91585112e-01 -7.77282476e-01 1.86816990e-01
3.44033837e-01 1.31385893e-01 1.87810290e+00 -1.17478907e+00
6.42171085e-01 1.05056322e+00 6.83453143e-01 4.05090332e-01
-1.57928097e+00 -5.25916219e-01 6.23547554e-01 -2.56254107e-01
-1.52603281e+00 -1.99296027e-01 4.38416272e-01 -4.05774742e-01
1.28456259e+00 -7.43792132e-02 8.17047834e-01 8.12961400e-01
3.83814275e-01 7.96609104e-01 7.80073464e-01 -4.24209744e-01
-1.31003838e-02 -1.78450093e-01 2.27236748e-01 9.79690373e-01
5.67587651e-02 -4.81202044e-02 -7.87763834e-01 -1.95797145e-01
1.03294826e+00 1.35495067e-01 1.39094796e-02 -3.51556182e-01
-1.55778372e+00 9.60287392e-01 1.00076139e+00 2.53300852e-04
-4.41406906e-01 9.05419350e-01 9.03678760e-02 -9.36245453e-03
3.79880726e-01 3.92812490e-01 -1.63808689e-01 5.83089232e-01
-1.24079895e+00 5.59855402e-01 2.27166221e-01 1.07844615e+00
1.11687279e+00 -5.23260891e-01 -7.54961133e-01 4.34132636e-01
4.64750290e-01 3.56049895e-01 8.61770362e-02 -6.22692406e-01
6.71539783e-01 3.38309407e-01 1.62675902e-01 -1.02392089e+00
-2.07878068e-01 -5.94615817e-01 -5.47520638e-01 -4.15132016e-01
-9.36557874e-02 1.06460735e-01 -1.23543227e+00 1.79536963e+00
-2.01625098e-02 6.67995751e-01 -1.99755967e-01 8.02383721e-01
9.65115786e-01 8.07177663e-01 3.63656282e-01 3.64327580e-01
1.44183469e+00 -1.37479663e+00 -3.50732237e-01 -5.39255857e-01
5.55934370e-01 -5.14951169e-01 1.18108141e+00 -7.29010031e-02
-1.23508310e+00 -6.15708232e-01 -7.83251345e-01 -3.21677089e-01
-2.11480916e-01 -1.89704075e-01 5.65889120e-01 1.03720047e-01
-1.70815253e+00 2.70861298e-01 -5.72877467e-01 -3.62042129e-01
4.52462971e-01 5.57784200e-01 -3.44951659e-01 -4.26301099e-02
-1.10375738e+00 5.40329278e-01 3.83783668e-01 2.55066752e-01
-8.60647440e-01 -8.22486520e-01 -8.44254494e-01 1.55949444e-01
1.07451394e-01 -1.30687308e+00 1.21118367e+00 -6.72006130e-01
-8.26839924e-01 1.32137907e+00 -4.65126783e-01 -8.87044787e-01
-3.15618329e-02 -3.06967705e-01 -1.69111162e-01 2.21487224e-01
1.97095782e-01 1.62611961e+00 1.00001609e+00 -1.35050666e+00
-8.82283747e-01 2.07296871e-02 5.04621208e-01 2.53602475e-01
-9.01908055e-02 5.13149872e-02 -8.03426981e-01 -7.74795055e-01
2.64335543e-01 -9.35727000e-01 -4.31560785e-01 6.21274374e-02
-6.40798450e-01 -2.72202399e-02 3.58460069e-01 -6.55780911e-01
1.26459646e+00 -2.12389350e+00 3.12017530e-01 3.64981860e-01
4.49020565e-01 -3.83375645e-01 -5.04250288e-01 3.51810217e-01
-1.37669310e-01 5.03525972e-01 -7.98054114e-02 -6.62389517e-01
2.94138640e-01 3.10846508e-01 -8.16823661e-01 3.27241838e-01
6.55518651e-01 1.62063622e+00 -9.52093661e-01 -4.88345087e-01
9.99924610e-04 3.90224934e-01 -8.63704562e-01 1.12452144e-02
-4.13707912e-01 -5.17946221e-02 -3.10321242e-01 5.25011778e-01
2.66218096e-01 -6.00337446e-01 -1.86854184e-01 -4.34116364e-01
1.75130337e-01 2.46876180e-01 -6.09119058e-01 2.13744783e+00
-4.41198766e-01 5.80296040e-01 -1.24196172e-01 -5.09589612e-01
7.37329364e-01 2.89769471e-01 3.05775493e-01 -6.93251610e-01
-3.00900906e-01 6.58999383e-02 -3.10657918e-01 -1.04917571e-01
9.72962737e-01 -2.55394876e-01 -3.35472792e-01 4.52865183e-01
1.05553567e-01 -9.27690044e-02 -1.84941486e-01 4.43835795e-01
1.11104906e+00 2.45489836e-01 1.33115098e-01 -3.09641689e-01
8.16236660e-02 -5.16957194e-02 1.86819941e-01 1.04069602e+00
-7.88994506e-02 9.70963836e-01 2.71742731e-01 -5.90473354e-01
-1.33005226e+00 -1.25240648e+00 1.18809491e-01 1.39058745e+00
2.68099874e-01 -4.66300309e-01 -5.77672720e-01 -6.82029247e-01
6.75628409e-02 4.57416177e-01 -8.59493911e-01 9.91742238e-02
-7.42217362e-01 -3.17051113e-01 3.82953852e-01 6.95539594e-01
2.43167847e-01 -1.06502891e+00 -3.99991393e-01 2.72867292e-01
-2.69791603e-01 -1.37200367e+00 -9.36672330e-01 3.35257649e-01
-7.00451016e-01 -4.94327188e-01 -8.41812849e-01 -1.16591704e+00
7.76146770e-01 5.07467330e-01 1.65930736e+00 3.56548995e-01
7.46884942e-02 3.80300701e-01 -3.03335041e-01 -9.54359919e-02
-1.15891052e-02 2.69637585e-01 -3.38856757e-01 -5.57712577e-02
3.04595023e-01 -3.15172434e-01 -8.43563199e-01 2.27102622e-01
-9.65864718e-01 3.13420683e-01 7.05188274e-01 9.62871850e-01
1.09610260e+00 -1.85886785e-01 2.40839183e-01 -7.43729353e-01
4.89108115e-01 -2.66334414e-01 -4.98396188e-01 3.25393200e-01
-4.74043638e-02 8.69632363e-02 1.31925747e-01 -1.02737986e-01
-4.69584167e-01 3.01845018e-02 -1.04774356e-01 -4.52759355e-01
-9.77516398e-02 6.00904584e-01 1.33006126e-01 -5.32711260e-02
4.32905287e-01 2.79271245e-01 -4.61105257e-01 -4.99034040e-02
8.70113373e-01 9.08688232e-02 7.78964996e-01 -7.02334821e-01
9.51501310e-01 4.50160503e-01 -2.63437539e-01 -4.64773238e-01
-1.06650269e+00 -5.81823647e-01 -6.45818114e-01 1.83364423e-03
1.30037081e+00 -1.35221124e+00 -8.41804221e-02 5.22629060e-02
-1.31295741e+00 -5.09762764e-01 -2.83954442e-01 -1.13960266e-01
-5.49655616e-01 -9.24594235e-03 -6.07589900e-01 -3.36169600e-01
-4.28923726e-01 -1.12839437e+00 1.66143990e+00 1.22463359e-02
-2.95202911e-01 -1.01311398e+00 -7.97988921e-02 1.25235811e-01
9.79356319e-02 7.08841607e-02 1.01775372e+00 -3.66417497e-01
-9.10056412e-01 -1.62091717e-01 -4.89503711e-01 -1.88491896e-01
-2.57787019e-01 -4.30013090e-02 -9.43012357e-01 -3.59042317e-01
-6.30417705e-01 -4.05794382e-01 1.35096741e+00 5.33396006e-01
1.06362641e+00 -3.86703283e-01 -4.79386032e-01 6.99757636e-01
1.67677116e+00 -2.92353541e-01 7.33165026e-01 4.58112121e-01
9.50955451e-01 3.35971147e-01 4.30833727e-01 3.97312865e-02
3.85471374e-01 7.69747913e-01 6.35396838e-01 -5.61191857e-01
-6.71795607e-02 -6.54608071e-01 2.26685971e-01 4.35169458e-01
4.54718322e-01 -4.80862230e-01 -8.45080137e-01 7.86783814e-01
-1.86231494e+00 -9.66369987e-01 1.98770657e-01 1.87881565e+00
6.82183266e-01 3.89977187e-01 7.56106228e-02 -5.30098438e-01
7.06324339e-01 4.82447743e-01 -2.10719928e-01 -3.81918669e-01
-1.67542800e-01 4.11377102e-01 6.33505642e-01 6.29454792e-01
-1.24584055e+00 1.31242990e+00 6.94063759e+00 5.95671833e-01
-1.08540511e+00 4.10465747e-02 8.39574099e-01 -1.66482851e-01
-1.08596373e+00 2.90955678e-02 -1.00932121e+00 9.64147504e-03
5.16037524e-01 3.48669618e-01 3.92275125e-01 3.43427449e-01
1.06590949e-01 2.15675354e-01 -1.34136772e+00 7.74383903e-01
-1.84314236e-01 -1.82148004e+00 4.82215285e-01 2.78728247e-01
9.86899018e-01 3.25250119e-01 3.43220353e-01 1.32067502e-01
3.67732853e-01 -1.25951767e+00 1.17181540e+00 2.25326508e-01
6.94693029e-01 -4.37714219e-01 5.71407616e-01 -1.19136319e-01
-1.40587354e+00 -7.97082856e-02 -4.01667535e-01 1.22334503e-01
3.13459814e-01 3.04978162e-01 -7.85656452e-01 4.97570693e-01
4.70082581e-01 7.37403333e-01 -5.51861823e-01 8.96706581e-01
-2.90360004e-01 5.57692468e-01 -1.36232331e-01 1.80606559e-01
1.06811941e+00 2.40224097e-02 2.21650660e-01 1.28230321e+00
2.57278860e-01 8.46605673e-02 4.56840426e-01 1.12413895e+00
-2.14881852e-01 -1.80286393e-02 -6.22789323e-01 -1.76405028e-01
5.87349236e-01 1.20579433e+00 -8.87129366e-01 -2.68605709e-01
-6.52237177e-01 1.02373242e+00 4.18459505e-01 7.56001472e-01
-9.24086511e-01 1.31841883e-01 5.68330050e-01 1.64162412e-01
7.18003035e-01 -3.26020211e-01 -4.39724147e-01 -9.22703981e-01
-3.73456837e-03 -6.09773517e-01 4.02683169e-01 -1.08523583e+00
-1.17295539e+00 5.94899893e-01 -2.17344537e-02 -8.15957487e-01
1.53575763e-01 -5.05276024e-01 -7.79133856e-01 1.13082683e+00
-1.53473043e+00 -1.72771096e+00 -8.17943886e-02 3.76339972e-01
4.64087635e-01 1.33499950e-01 7.81078517e-01 -2.06619781e-03
-4.28928792e-01 6.97489381e-01 -3.48764300e-01 2.66408831e-01
4.70401198e-01 -1.17802131e+00 1.08250105e+00 1.00407851e+00
8.79135132e-01 1.03952086e+00 6.84340417e-01 -4.89537925e-01
-9.54711914e-01 -1.37027574e+00 1.00683999e+00 -5.41416883e-01
8.46684813e-01 -6.58386290e-01 -7.93682575e-01 9.84714031e-01
5.64000368e-01 8.41861367e-02 6.92661166e-01 3.75979185e-01
-6.73933089e-01 5.15235588e-02 -7.71071076e-01 7.34823346e-01
1.01896858e+00 -8.94158363e-01 -6.89655304e-01 2.54035801e-01
1.30808520e+00 -3.46847892e-01 -6.82003975e-01 1.44221038e-01
5.97375631e-01 -7.27654874e-01 1.17122841e+00 -6.86526000e-01
8.05112958e-01 -3.01012933e-01 -6.13466561e-01 -1.09478462e+00
-7.98938453e-01 -4.04856414e-01 1.79607496e-02 8.99787366e-01
7.82389104e-01 -1.47887677e-01 1.16231346e+00 4.25557762e-01
-2.90999919e-01 -1.13443220e+00 -6.10840857e-01 -3.54780614e-01
1.60055861e-01 -3.75364542e-01 7.83100843e-01 7.24020123e-01
-1.77891612e-01 4.28538054e-01 -2.38797098e-01 4.57994699e-01
4.80921060e-01 2.97706217e-01 5.82717538e-01 -5.29027820e-01
-5.75304985e-01 -5.59473276e-01 -2.74198592e-01 -1.51570106e+00
2.31960759e-01 -9.05151069e-01 4.52237755e-01 -1.95345807e+00
4.90410507e-01 -5.09023607e-01 -6.26489341e-01 7.45329559e-01
-4.61841643e-01 7.84559309e-01 2.99857706e-01 1.96534783e-01
-7.80329525e-01 2.15661019e-01 1.12812400e+00 -4.37025666e-01
-1.12672761e-01 -5.18403113e-01 -9.03368235e-01 4.11943287e-01
5.57491243e-01 -4.09446061e-01 -4.06177282e-01 -1.05252576e+00
7.14787722e-01 -9.63538215e-02 7.75921166e-01 -7.23773003e-01
2.08408490e-01 -5.70631959e-02 2.61117399e-01 -9.24875200e-01
2.65552759e-01 -7.20328450e-01 1.16619505e-01 2.14969680e-01
-7.87565827e-01 3.16045552e-01 3.26053321e-01 8.14117789e-01
-2.94662625e-01 2.86302529e-02 4.29345489e-01 -3.15251380e-01
-1.02174854e+00 5.89559853e-01 -3.98220159e-02 -7.08797872e-02
7.43085206e-01 -3.79357189e-01 -2.66660362e-01 -3.59924227e-01
-7.88346052e-01 4.17190880e-01 5.95553219e-01 4.05332834e-01
7.09594965e-01 -1.40240121e+00 -7.53579259e-01 7.44650215e-02
3.99490416e-01 2.34010905e-01 1.14235863e-01 4.83978927e-01
-5.69094419e-01 6.90903902e-01 6.81376681e-02 -8.55437398e-01
-1.12359154e+00 6.24738276e-01 3.32645893e-01 -3.36854458e-01
-7.74089992e-01 1.63822150e+00 8.11614037e-01 -1.42974719e-01
2.41280347e-01 -6.95991218e-01 2.14020863e-01 -2.10319281e-01
4.04288322e-01 -6.16557837e-01 -4.42512780e-02 -6.57368362e-01
-3.03347468e-01 7.35120058e-01 -3.36427987e-01 -3.62557828e-01
1.05277693e+00 -2.28389740e-01 -2.07536325e-01 2.21548542e-01
1.18920243e+00 -2.08214760e-01 -1.30095911e+00 -4.67518926e-01
-2.97637507e-02 -3.71249825e-01 1.45492718e-01 -5.38900614e-01
-9.22357023e-01 7.65363812e-01 2.84845769e-01 -5.97129911e-02
8.94208491e-01 2.90788978e-01 8.42270434e-01 3.84779394e-01
4.45429444e-01 -5.77517688e-01 3.79156381e-01 3.89135450e-01
7.92362690e-01 -1.10076547e+00 -7.13533089e-02 -2.38258123e-01
-4.57968265e-01 8.15958738e-01 5.19578278e-01 -2.84976214e-01
4.01347548e-01 8.34813341e-02 -8.28853846e-02 -3.13880324e-01
-8.34109545e-01 -1.66171014e-01 5.99567533e-01 2.59378195e-01
4.13647801e-01 4.73178253e-02 8.61956179e-02 3.58067960e-01
-8.67787972e-02 -1.27395019e-01 2.16345116e-01 7.87282407e-01
-6.11477852e-01 -9.48238850e-01 -2.89225727e-01 4.27041590e-01
-3.59394193e-01 -6.42624676e-01 -4.30031508e-01 5.86707711e-01
2.32884288e-01 6.10318840e-01 4.57265526e-01 -5.68633735e-01
1.02621324e-01 -2.36130822e-02 4.38549966e-01 -1.07165289e+00
-8.65975738e-01 2.48743176e-01 1.34253008e-02 -5.99841774e-01
-4.07642633e-01 -3.66442055e-01 -1.29493308e+00 -3.98713537e-02
-2.95498848e-01 -7.15025738e-02 2.41735965e-01 9.87158775e-01
4.18827713e-01 5.68704367e-01 1.91132471e-01 -8.37219596e-01
-1.85545340e-01 -5.25832117e-01 -3.26857328e-01 3.66597146e-01
4.77542609e-01 -2.57402986e-01 1.97472394e-01 2.15085983e-01] | [10.552774429321289, 1.4528958797454834] |
9b045687-fd82-471b-b182-143563135f21 | mixing-specific-data-augmentation-techniques | 2008.02480 | null | https://arxiv.org/abs/2008.02480v1 | https://arxiv.org/pdf/2008.02480v1.pdf | Mixing-Specific Data Augmentation Techniques for Improved Blind Violin/Piano Source Separation | Blind music source separation has been a popular and active subject of research in both the music information retrieval and signal processing communities. To counter the lack of available multi-track data for supervised model training, a data augmentation method that creates artificial mixtures by combining tracks from different songs has been shown useful in recent works. Following this light, we examine further in this paper extended data augmentation methods that consider more sophisticated mixing settings employed in the modern music production routine, the relationship between the tracks to be combined, and factors of silence. As a case study, we consider the separation of violin and piano tracks in a violin piano ensemble, evaluating the performance in terms of common metrics, namely SDR, SIR, and SAR. In addition to examining the effectiveness of these new data augmentation methods, we also study the influence of the amount of training data. Our evaluation shows that the proposed mixing-specific data augmentation methods can help improve the performance of a deep learning-based model for source separation, especially in the case of small training data. | ['Alvin Wen-Yu Su', 'Yi-Hsuan Yang', 'Yin-Cheng Yeh', 'Wen-Yi Hsiao', 'Ching-Yu Chiu'] | 2020-08-06 | null | null | null | null | ['music-source-separation'] | ['music'] | [ 4.50354874e-01 -2.88802981e-01 -5.47641851e-02 6.67783767e-02
-8.75662506e-01 -7.55199194e-01 7.20487416e-01 1.63975060e-01
-2.95515865e-01 5.44035852e-01 4.72419083e-01 2.49237847e-02
-5.31472862e-01 -1.92759439e-01 -4.45437402e-01 -8.73212278e-01
7.31661096e-02 2.26930141e-01 -4.68821451e-02 -1.27369672e-01
9.90607962e-02 5.13571620e-01 -1.69664741e+00 9.48042199e-02
8.68965685e-01 7.53705025e-01 -1.72949404e-01 6.50049508e-01
1.88363213e-02 5.79483867e-01 -9.10209119e-01 -1.94026604e-01
3.74799311e-01 -8.68000448e-01 -2.31896162e-01 1.16749860e-01
5.96766055e-01 1.35812908e-01 -2.05659479e-01 8.18978608e-01
8.44998598e-01 4.13518101e-01 4.56657052e-01 -9.80122864e-01
1.29342809e-01 1.12217867e+00 -4.95742500e-01 4.07781333e-01
3.40559594e-02 -1.31049499e-01 8.76185954e-01 -5.17548621e-01
1.56919211e-01 8.61452460e-01 7.38033175e-01 7.05966204e-02
-1.36709392e+00 -9.77557659e-01 -1.96800992e-01 3.19198668e-01
-1.13616085e+00 -7.70455062e-01 1.15118885e+00 -4.75502014e-01
3.93361658e-01 3.42100173e-01 4.77687210e-01 1.00759339e+00
-5.07722497e-01 7.92345762e-01 1.03039443e+00 -7.96535015e-01
2.51933277e-01 2.78916862e-02 2.31268704e-01 -4.48720306e-02
2.74437577e-01 2.24363521e-01 -8.69878650e-01 -1.80745721e-01
5.74717283e-01 -4.96286958e-01 -3.66157979e-01 -2.51488626e-01
-1.08268428e+00 7.41803765e-01 1.19592920e-02 6.51077986e-01
-3.31898928e-01 1.30245194e-01 3.22270066e-01 1.74260095e-01
5.74358106e-01 8.14251125e-01 -2.58001298e-01 -2.90660411e-01
-1.49793267e+00 3.03062737e-01 7.66932309e-01 3.21310639e-01
2.26192608e-01 5.98798633e-01 -4.19849068e-01 1.09872150e+00
1.28103524e-01 3.47169846e-01 5.67090333e-01 -1.05381024e+00
3.75758469e-01 3.42454463e-01 -3.33450325e-02 -6.64025724e-01
-3.75780135e-01 -1.15656936e+00 -6.77255094e-01 2.68034667e-01
6.47642314e-01 -3.13371569e-01 -6.37800515e-01 1.79519570e+00
2.09443688e-01 5.87277174e-01 -1.90544128e-02 7.76641905e-01
3.98931652e-01 2.97738254e-01 -3.42719764e-01 -4.40772802e-01
1.08823836e+00 -9.09574509e-01 -9.64179337e-01 -1.59704536e-01
1.51263282e-01 -1.06547177e+00 6.70847774e-01 8.70806456e-01
-9.45164561e-01 -7.82656789e-01 -1.15630555e+00 3.61577094e-01
-5.27746975e-02 5.47964633e-01 3.97977561e-01 9.90868270e-01
-4.11582947e-01 7.64981508e-01 -6.53232396e-01 -1.75444409e-01
2.78710872e-01 2.90005594e-01 7.37508945e-03 9.56218392e-02
-9.65637863e-01 5.33113837e-01 1.82488918e-01 4.98734564e-02
-6.41992211e-01 -6.15310669e-01 -3.75049472e-01 2.56384969e-01
3.76210153e-01 -4.16031331e-01 1.17216766e+00 -1.10023081e+00
-1.46795511e+00 3.96732807e-01 4.18691412e-02 -7.01543927e-01
4.40966040e-01 -6.10929728e-01 -6.35046124e-01 -7.54573494e-02
-1.70689270e-01 2.53595740e-01 1.05897141e+00 -1.26200533e+00
-4.75093156e-01 -3.81261975e-01 -2.09029898e-01 1.97525769e-01
-3.33336323e-01 3.94481122e-02 -3.73789102e-01 -1.22556734e+00
5.16395643e-02 -1.16184211e+00 -7.55090341e-02 -7.37014651e-01
-4.29558337e-01 2.60455489e-01 4.11195070e-01 -7.49228060e-01
1.26077807e+00 -2.40201020e+00 2.50946999e-01 4.42637920e-01
-2.36364529e-01 4.79621172e-01 -1.39202476e-01 3.56363058e-01
-2.86847085e-01 -1.80537373e-01 -3.46660167e-01 -5.72436690e-01
-6.18252764e-03 -1.04593650e-01 -2.61432439e-01 2.19915092e-01
-8.05749837e-03 3.78590286e-01 -3.91295880e-01 -6.52945563e-02
1.47749588e-01 6.87525868e-01 -4.51781571e-01 -5.57951145e-02
-2.66965455e-03 7.04990923e-01 1.25006557e-01 2.53860235e-01
3.82220954e-01 1.22043349e-01 1.55032948e-01 -2.90662467e-01
-1.08156286e-01 4.26060855e-01 -1.63664496e+00 1.79122543e+00
-3.51772636e-01 9.16581213e-01 1.63849562e-01 -5.69536626e-01
8.74718070e-01 5.70831239e-01 6.62314594e-01 -3.62234414e-01
2.90832907e-01 2.99342066e-01 5.73804975e-01 -2.15805948e-01
5.51402152e-01 -2.86195159e-01 2.63188958e-01 4.85696942e-01
1.58866957e-01 -7.25759380e-03 5.47950327e-01 -1.23560563e-01
9.74700630e-01 3.50744324e-03 -1.01695634e-01 4.10465077e-02
4.03722972e-01 -1.80875435e-01 3.06731969e-01 7.25135922e-01
2.40882441e-01 8.95182729e-01 1.49776071e-01 3.28387886e-01
-6.70584083e-01 -7.59252369e-01 -7.83525929e-02 9.76678014e-01
-3.24299186e-01 -6.04730368e-01 -7.73606062e-01 -1.86253399e-01
-1.29080504e-01 6.75981522e-01 -2.70278037e-01 -2.47826979e-01
-6.10945284e-01 -1.09982908e+00 9.04339314e-01 3.30405235e-01
2.37863094e-01 -9.33824539e-01 -4.09388155e-01 2.53021777e-01
-4.29394871e-01 -1.13145804e+00 -3.38128418e-01 3.95246893e-01
-9.85490501e-01 -9.69994605e-01 -7.21595407e-01 -3.70927423e-01
-2.33886149e-02 6.39542043e-02 6.98502481e-01 -3.19820017e-01
2.96346210e-02 3.49465638e-01 -5.90352178e-01 -7.32569396e-01
-6.64330184e-01 3.11163217e-01 1.23391949e-01 4.83470619e-01
1.32542336e-02 -8.62291634e-01 -2.89835483e-01 2.10968509e-01
-1.11189413e+00 -8.95527825e-02 5.90015173e-01 4.21577007e-01
1.39644191e-01 2.89375365e-01 6.42957807e-01 -7.64828920e-01
9.53111947e-01 -2.19337046e-01 -4.00676608e-01 -1.16969548e-01
-4.48161542e-01 9.14179683e-02 3.13771576e-01 -7.17516661e-01
-7.82603741e-01 6.65069371e-02 -1.63844675e-01 -5.43834507e-01
-1.68756306e-01 5.90022862e-01 -1.43494397e-01 4.04507779e-02
7.72552073e-01 -1.88774355e-02 2.60597132e-02 -9.25046444e-01
2.24503443e-01 5.54264665e-01 6.07777715e-01 -3.02734643e-01
8.49813998e-01 3.41272265e-01 -2.73457542e-03 -8.12227726e-01
-6.80748641e-01 -6.16118789e-01 -5.60765207e-01 -1.49062932e-01
4.92504954e-01 -7.76454628e-01 -3.22444111e-01 5.26474416e-01
-8.56417120e-01 -2.00546846e-01 -6.23071969e-01 9.59920228e-01
-4.01122242e-01 2.82899886e-01 -3.21682334e-01 -9.75763619e-01
-1.78885505e-01 -9.59308743e-01 5.94737291e-01 8.68303403e-02
-3.44801009e-01 -6.59804344e-01 5.03438830e-01 4.27702397e-01
4.65720743e-01 6.58280998e-02 6.89172566e-01 -9.25755680e-01
-3.11359257e-01 -3.79914373e-01 3.55171204e-01 6.60926104e-01
3.38472217e-01 -2.17316285e-01 -1.48235297e+00 -1.37811765e-01
2.29841143e-01 2.58658230e-01 9.64129210e-01 4.20097291e-01
6.01893485e-01 -8.48399550e-02 3.17983516e-02 3.53079975e-01
1.03749871e+00 1.36603892e-01 4.89113539e-01 1.97941810e-01
6.33722067e-01 5.99810779e-01 3.20597410e-01 4.06332582e-01
-2.19586521e-01 1.00832415e+00 2.10171685e-01 -1.77301258e-01
-4.89244819e-01 3.02086622e-02 3.46593291e-01 8.89523745e-01
-1.71399355e-01 -2.13145137e-01 -6.66165352e-01 4.17930752e-01
-1.62327969e+00 -9.44728613e-01 -1.97283149e-01 2.43311644e+00
6.78099334e-01 9.83423293e-02 4.47347075e-01 1.06922305e+00
6.56958759e-01 7.82220215e-02 -2.98826993e-01 1.72789712e-02
-3.45991671e-01 4.64340895e-01 3.68219763e-01 2.82150209e-01
-1.12248313e+00 4.15660262e-01 5.62632895e+00 9.56659615e-01
-1.25461888e+00 1.57939017e-01 4.53147814e-02 -5.91546416e-01
-4.97913547e-02 -6.56557381e-02 -3.91272128e-01 4.52117324e-01
9.05220091e-01 2.81161964e-01 6.15718663e-01 1.91717282e-01
2.97295153e-01 -1.47256494e-01 -1.01957297e+00 1.12305760e+00
3.03955942e-01 -1.01056218e+00 -2.33869031e-01 1.98366582e-01
6.46366060e-01 3.51972692e-02 2.00208753e-01 2.51167506e-01
-1.99378908e-01 -8.18092108e-01 8.72949362e-01 4.49665010e-01
1.07977025e-01 -6.38974667e-01 6.16762996e-01 3.29064399e-01
-7.69591630e-01 -4.14925247e-01 3.67080241e-01 8.76213089e-02
1.75907165e-01 7.24460542e-01 -6.96989715e-01 7.87452459e-01
3.78052562e-01 3.13761175e-01 -6.51516497e-01 1.56956840e+00
-5.55581674e-02 1.15944219e+00 -5.77039361e-01 4.54028368e-01
-1.39249906e-01 -1.89335018e-01 8.63186955e-01 1.09245777e+00
3.83542299e-01 -2.42214844e-01 -1.61691323e-01 7.35680997e-01
2.03179512e-02 2.38355190e-01 -2.63640702e-01 -1.41066447e-01
2.94303119e-01 1.13236749e+00 -7.25974798e-01 -7.41709992e-02
-3.62909213e-02 5.44863105e-01 -2.39323646e-01 4.46972102e-01
-7.32722819e-01 -7.68855885e-02 6.22930408e-01 2.98547119e-01
3.74599248e-01 -3.22073787e-01 -2.79737383e-01 -9.15254414e-01
-1.94771122e-02 -1.21204591e+00 2.44739056e-01 -4.42253977e-01
-9.72670197e-01 5.22233605e-01 5.95978685e-02 -1.45982027e+00
-3.82833004e-01 -2.68515706e-01 -6.39246285e-01 7.55812645e-01
-1.23340654e+00 -7.76048481e-01 2.65522692e-02 3.35362077e-01
3.64933044e-01 -4.60181832e-01 6.04814291e-01 7.67304659e-01
-6.49710357e-01 4.80444849e-01 4.00111020e-01 3.04931886e-02
8.07349503e-01 -9.06148195e-01 1.95122704e-01 1.02350068e+00
1.03175020e+00 4.62571949e-01 9.92953479e-01 -2.81398535e-01
-9.05420065e-01 -6.77393436e-01 4.42900062e-01 -3.18031460e-01
3.21389258e-01 -2.04442248e-01 -8.11115444e-01 2.26710975e-01
2.08915055e-01 -3.67609799e-01 9.61804926e-01 4.20704514e-01
-1.71267405e-01 -3.43333513e-01 -6.13529980e-01 4.64681655e-01
5.77462554e-01 -4.34276760e-01 -3.92462134e-01 -2.07315609e-01
2.11048380e-01 -1.26396492e-01 -5.00026166e-01 5.24942219e-01
6.32320464e-01 -1.08596516e+00 8.94681633e-01 -1.84777185e-01
-8.82693008e-02 -4.72116232e-01 -1.70015357e-02 -1.47183692e+00
-1.75172612e-01 -6.49949551e-01 -1.97235256e-01 1.66074896e+00
3.55430394e-01 -1.51774019e-01 7.59354055e-01 2.12872207e-01
-5.86423054e-02 -1.87261179e-01 -8.36263895e-01 -7.74373353e-01
-1.43914461e-01 -6.83335364e-01 3.58079761e-01 9.05915618e-01
-2.34620959e-01 5.82468987e-01 -5.76281965e-01 4.23131250e-02
3.19805473e-01 9.32067633e-02 9.04025257e-01 -1.44400048e+00
-7.83078015e-01 -5.98221183e-01 -1.94944620e-01 -6.14719629e-01
-9.15231630e-02 -9.23041940e-01 -1.26603376e-02 -1.25394535e+00
-2.62338281e-01 -4.97515649e-01 -5.98986089e-01 2.15131164e-01
-1.82464302e-01 4.04048085e-01 7.03133225e-01 2.29901865e-01
-1.76919222e-01 4.53559339e-01 7.12086618e-01 -1.98333383e-01
-5.04185915e-01 4.33066696e-01 -6.48966670e-01 7.35861063e-01
8.18299532e-01 -6.25150144e-01 -4.82050419e-01 -2.16084540e-01
1.64423585e-01 -1.13518417e-01 2.04521030e-01 -1.64873481e+00
1.33504272e-01 4.10825193e-01 1.86094731e-01 -2.74430126e-01
7.43270874e-01 -7.99503148e-01 4.64561075e-01 1.00505471e-01
-4.32042450e-01 -2.99253732e-01 5.41248858e-01 4.38519478e-01
-3.85133028e-01 -2.76143342e-01 4.79345262e-01 2.79368341e-01
-8.89269114e-02 -2.84942925e-01 -3.54180187e-01 -4.54418957e-02
4.10733819e-01 -1.81271687e-01 9.42983702e-02 -5.61529636e-01
-8.06906402e-01 -3.26707155e-01 -5.20417728e-02 6.10850096e-01
7.91564360e-02 -1.19125187e+00 -8.64050210e-01 2.34653860e-01
-5.69881797e-02 -4.60177988e-01 2.28834018e-01 1.08449268e+00
-4.98369820e-02 1.59563467e-01 -1.04381837e-01 -3.69725168e-01
-1.43388355e+00 4.12270069e-01 2.52025455e-01 -1.76152766e-01
-1.90053806e-01 6.10016286e-01 -6.15361659e-03 -1.07608117e-01
6.51341379e-01 -4.36131060e-01 -4.83251959e-01 4.56417859e-01
2.99020410e-01 6.85675144e-01 3.80608946e-01 -6.47038758e-01
-1.11638457e-01 4.27602798e-01 3.77430111e-01 -5.93496799e-01
1.09615290e+00 4.76471111e-02 8.94468576e-02 8.18979919e-01
6.59000158e-01 7.25197077e-01 -8.13872695e-01 -1.50814936e-01
1.87538654e-01 -2.33647481e-01 1.17453732e-01 -9.15141225e-01
-1.02824485e+00 8.63578141e-01 8.71880829e-01 4.10795808e-01
1.32136774e+00 -3.37502062e-01 3.63674343e-01 7.31356218e-02
-1.27180619e-02 -8.13817024e-01 -5.75501844e-02 3.54416400e-01
7.61972606e-01 -8.38021636e-01 -1.15305170e-01 -1.76504806e-01
-4.07588184e-01 8.38276386e-01 5.52171692e-02 1.61588732e-02
6.15661204e-01 2.74152309e-01 3.73029292e-01 7.32755587e-02
-2.97531188e-01 -5.16552091e-01 5.79722941e-01 1.90857664e-01
5.99552512e-01 -2.11277694e-01 -2.01287299e-01 7.71545410e-01
-3.75604689e-01 3.05806529e-02 3.66778374e-01 6.32210135e-01
-1.41873062e-01 -1.45940351e+00 -7.92132556e-01 3.68923545e-01
-7.46080518e-01 -2.53732055e-01 -4.84164089e-01 5.33650577e-01
3.48741919e-01 1.24305081e+00 -1.66667610e-01 -2.91550130e-01
3.56861800e-01 3.91148239e-01 5.80720723e-01 -5.82552135e-01
-1.06595385e+00 5.95811129e-01 1.20458513e-01 1.27019286e-01
-7.62579799e-01 -6.74930036e-01 -8.29392374e-01 1.57547846e-01
-6.02341175e-01 3.14780325e-01 8.66140604e-01 9.69481051e-01
2.63175845e-01 1.00911129e+00 3.15774024e-01 -9.27144647e-01
-4.17028457e-01 -1.39191341e+00 -6.69723451e-01 4.14919019e-01
4.80155557e-01 -4.87368196e-01 -3.22352648e-01 1.48962155e-01] | [15.541414260864258, 5.5316362380981445] |
18a59629-9940-44f6-b3d7-dd1f225204c4 | mutual-clustering-on-comparative-texts-via | 1903.03762 | null | http://arxiv.org/abs/1903.03762v1 | http://arxiv.org/pdf/1903.03762v1.pdf | Mutual Clustering on Comparative Texts via Heterogeneous Information Networks | Currently, many intelligence systems contain the texts from multi-sources,
e.g., bulletin board system (BBS) posts, tweets and news. These texts can be
``comparative'' since they may be semantically correlated and thus provide us
with different perspectives toward the same topics or events. To better
organize the multi-sourced texts and obtain more comprehensive knowledge, we
propose to study the novel problem of Mutual Clustering on Comparative Texts
(MCCT), which aims to cluster the comparative texts simultaneously and
collaboratively. The MCCT problem is difficult to address because 1)
comparative texts usually present different data formats and structures and
thus they are hard to organize, and 2) there lacks an effective method to
connect the semantically correlated comparative texts to facilitate clustering
them in an unified way. To this aim, in this paper we propose a Heterogeneous
Information Network-based Text clustering framework HINT. HINT first models
multi-sourced texts (e.g. news and tweets) as heterogeneous information
networks by introducing the shared ``anchor texts'' to connect the comparative
texts. Next, two similarity matrices based on HINT as well as a transition
matrix for cross-text-source knowledge transfer are constructed. Comparative
texts clustering are then conducted by utilizing the constructed matrices.
Finally, a mutual clustering algorithm is also proposed to further unify the
separate clustering results of the comparative texts by introducing a
clustering consistency constraint. We conduct extensive experimental on three
tweets-news datasets, and the results demonstrate the effectiveness and
robustness of the proposed method in addressing the MCCT problem. | ['Fei-Yue Wang', 'Philip S. Yu', 'Danyan Wen', 'Zhaohui Peng', 'Senzhang Wang', 'Jianping Cao'] | 2019-03-09 | null | null | null | null | ['text-clustering'] | ['natural-language-processing'] | [-3.72761518e-01 -2.28859439e-01 -1.83197215e-01 -1.63414642e-01
-7.21487761e-01 -4.48338121e-01 7.19028413e-01 5.02226830e-01
-3.65854204e-01 4.03439045e-01 5.49432099e-01 -4.14264165e-02
-3.72659624e-01 -8.07417691e-01 -1.80186257e-01 -7.45730340e-01
1.04942113e-01 6.26067400e-01 4.89933610e-01 -3.35616589e-01
2.63950914e-01 -2.73117840e-01 -1.31593049e+00 4.55861926e-01
1.15374017e+00 5.43067992e-01 5.58688939e-01 3.69736999e-02
-7.91526377e-01 8.64978433e-01 -6.90484822e-01 -3.54884177e-01
4.33673002e-02 -6.89552546e-01 -9.88174856e-01 1.59491837e-01
-4.43964869e-01 3.61947417e-01 -1.76982224e-01 1.26593471e+00
2.64378339e-01 4.58854854e-01 7.53659725e-01 -1.60490334e+00
-4.38910276e-01 1.22395897e+00 -9.35119748e-01 8.55602100e-02
3.39740515e-01 -5.50689280e-01 1.09228718e+00 -6.79388523e-01
7.28533685e-01 1.29872525e+00 5.64421892e-01 7.36407191e-02
-8.42016876e-01 -7.32279420e-01 2.88704067e-01 4.73454863e-01
-1.60485578e+00 -1.34846419e-01 1.06185937e+00 -4.16290045e-01
1.64347455e-01 4.88459945e-01 5.06209850e-01 8.42638195e-01
-2.30289444e-01 7.87115276e-01 1.03991759e+00 -3.44367146e-01
8.73648971e-02 4.36647087e-01 3.44889998e-01 2.33075947e-01
2.07042485e-01 -7.60727227e-01 -3.47785413e-01 -8.97312611e-02
1.41841158e-01 1.73236609e-01 -4.44668204e-01 -2.76443046e-02
-1.39962363e+00 8.87562633e-01 4.45464402e-01 1.08981586e+00
-1.10489048e-01 -6.33681476e-01 5.61738610e-01 2.77981281e-01
5.57158649e-01 4.71113920e-02 -3.48628581e-01 1.91975102e-01
-7.57003427e-01 -4.19168800e-01 9.40230608e-01 1.23428094e+00
9.63785172e-01 -5.79391241e-01 1.71047643e-01 9.41239059e-01
7.09932506e-01 3.09140086e-01 7.81104684e-01 -3.04352582e-01
1.02672279e+00 8.76182616e-01 -5.64690493e-02 -1.97381675e+00
-5.56570053e-01 -3.57607543e-01 -1.19836223e+00 -7.36657798e-01
1.62105769e-01 -2.23533034e-01 -1.65829644e-01 1.49503219e+00
6.00041211e-01 3.54144394e-01 3.43642354e-01 7.61305809e-01
1.21846402e+00 1.01887822e+00 -4.65910137e-02 -5.35108447e-01
1.45073795e+00 -8.77798796e-01 -1.02478778e+00 2.60620862e-01
6.24837279e-01 -1.05426717e+00 7.14022934e-01 1.51118832e-02
-7.48116076e-01 -4.10017073e-01 -7.65759945e-01 1.90460399e-01
-5.43819726e-01 1.16958320e-01 3.77659574e-02 3.22330087e-01
-8.22587848e-01 2.07815036e-01 -6.66946828e-01 -6.01383388e-01
-3.66439670e-02 6.35914952e-02 -2.07241490e-01 1.01117074e-01
-1.59908700e+00 5.71549296e-01 8.24531376e-01 -5.58744520e-02
-6.58073323e-03 -3.48269224e-01 -5.55588543e-01 -7.91178122e-02
5.54838300e-01 -4.95039076e-01 7.31531739e-01 -9.91808712e-01
-1.20075488e+00 5.50107956e-01 -1.10661462e-01 5.32419458e-02
3.89337361e-01 4.03753638e-01 -8.80692363e-01 3.61144990e-01
5.21944702e-01 1.37682676e-01 3.51518035e-01 -1.63244712e+00
-1.10882056e+00 -3.53057384e-01 -3.55577469e-01 5.77079773e-01
-8.30359161e-01 2.32984141e-01 -9.84799504e-01 -7.55316317e-01
4.21702385e-01 -8.14787328e-01 -1.89150348e-01 -7.95189500e-01
-6.93820059e-01 -3.92470926e-01 9.94112194e-01 -5.86605430e-01
1.51944721e+00 -2.08397269e+00 2.53457069e-01 5.92832983e-01
5.00947297e-01 -1.12773366e-01 2.33295262e-01 6.81598902e-01
1.73596531e-01 2.45900095e-01 -9.52607170e-02 -2.67145157e-01
-2.43887659e-02 1.25231683e-01 2.81732529e-02 2.57019490e-01
-4.98176396e-01 6.07749820e-01 -1.00192165e+00 -1.09510052e+00
9.04343501e-02 2.69205481e-01 2.44995598e-02 -1.01401145e-03
3.43313441e-02 5.98627388e-01 -8.70938420e-01 1.66924462e-01
5.98091662e-01 -5.32379568e-01 5.18409669e-01 -5.48183501e-01
-2.54439712e-01 -1.13688380e-01 -1.38472366e+00 1.46499884e+00
-2.92876661e-01 5.65613329e-01 1.18534580e-01 -1.22533691e+00
9.36027288e-01 4.09842700e-01 7.84981310e-01 -3.60936016e-01
5.84792554e-01 -2.49579661e-02 -6.87668175e-02 -5.31635761e-01
5.23310006e-01 -1.13564320e-01 -2.44524524e-01 6.32624149e-01
-1.80322096e-01 2.90011913e-01 3.41401339e-01 6.80142879e-01
7.44188070e-01 -5.29274702e-01 1.59622058e-01 -4.26520646e-01
7.58725584e-01 2.34492555e-01 6.51450574e-01 2.04752728e-01
5.45795225e-02 2.85803556e-01 3.08384210e-01 1.51937231e-01
-6.70530379e-01 -5.92134953e-01 -1.11223020e-01 9.73394632e-01
6.90473378e-01 -6.91203892e-01 -5.57849526e-01 -6.68797433e-01
-3.21417302e-01 4.47039932e-01 -5.97449124e-01 9.06826258e-02
-1.28595233e-01 -8.42418730e-01 3.99973989e-01 1.29945666e-01
8.28754604e-01 -6.27944171e-01 1.41245022e-01 2.79544890e-01
-9.12443995e-01 -9.80283082e-01 -6.37456179e-01 -2.14870140e-01
-4.66473252e-01 -1.08054757e+00 -4.33682501e-01 -1.12343395e+00
7.46542990e-01 8.81713092e-01 7.81197608e-01 1.78530037e-01
4.33304191e-01 3.06739271e-01 -9.82337415e-01 -6.08886629e-02
-4.68660891e-01 3.06175739e-01 6.62861019e-02 4.14451033e-01
4.33734357e-01 -5.01798093e-01 -4.28390652e-01 7.57833064e-01
-1.20394254e+00 3.79207343e-01 4.27131981e-01 6.27305448e-01
2.75947064e-01 9.60597813e-01 7.44928002e-01 -9.14554298e-01
7.88942218e-01 -1.08390331e+00 -1.33753374e-01 4.26171064e-01
-4.36624348e-01 -3.16388786e-01 9.26391184e-01 -4.07729626e-01
-1.33964372e+00 -3.89557779e-01 2.60446966e-01 -3.44242871e-01
-1.48936629e-01 1.14409769e+00 -5.27821362e-01 2.49710917e-01
3.57159227e-01 2.38979086e-01 -1.63966656e-01 -4.22508597e-01
4.05633181e-01 1.10346985e+00 3.17825496e-01 -6.14088178e-01
9.54110563e-01 6.20708704e-01 -3.14654231e-01 -7.47363985e-01
-7.42062688e-01 -9.43368375e-01 -7.23726213e-01 -4.10272300e-01
1.01160872e+00 -8.06518495e-01 -6.56641066e-01 3.16448301e-01
-1.01705396e+00 1.81796208e-01 2.66017705e-01 7.11771786e-01
6.68744594e-02 6.83288395e-01 -7.62181938e-01 -4.23546791e-01
-1.08461432e-01 -9.41362262e-01 5.25172353e-01 3.50198269e-01
-9.64548662e-02 -1.44222367e+00 2.88976263e-02 4.45166528e-01
1.36258334e-01 1.21675059e-01 8.27534199e-01 -1.13292539e+00
-9.35559794e-02 1.47845909e-01 -2.76395500e-01 -2.00727671e-01
6.24914527e-01 2.30183452e-01 -4.10982966e-01 -2.99950242e-01
2.10138321e-01 1.26903638e-01 5.17853320e-01 4.88238260e-02
6.41231537e-01 -4.77706462e-01 -7.89605141e-01 8.61534774e-02
1.17342722e+00 5.46881616e-01 3.14883322e-01 5.63525736e-01
9.97688115e-01 1.13849902e+00 6.96472049e-01 4.86494064e-01
8.79950345e-01 4.11330700e-01 -1.47818312e-01 -2.15677217e-01
3.36858243e-01 -2.45075047e-01 1.41314253e-01 2.08839154e+00
3.26958508e-03 -4.16351706e-01 -9.51924145e-01 3.77031296e-01
-2.18032885e+00 -1.13975370e+00 -6.16560578e-01 1.83181596e+00
8.83518219e-01 -5.82510158e-02 8.17981511e-02 1.40023381e-01
1.55465925e+00 -8.92618895e-02 -2.83377767e-01 5.17803252e-01
-2.77959943e-01 -5.35381734e-01 2.05388200e-02 2.41717279e-01
-1.13135612e+00 7.15728641e-01 4.59876251e+00 1.30031693e+00
-7.75756538e-01 2.98790872e-01 4.71005023e-01 2.26704806e-01
-4.94949877e-01 2.17195123e-01 -6.12607419e-01 1.00316858e+00
6.59975410e-01 -5.80141306e-01 4.60733175e-02 4.53441650e-01
1.92224920e-01 -5.86464955e-03 -4.78370398e-01 9.64319348e-01
9.52723622e-02 -1.23065948e+00 8.13588202e-02 -1.75618336e-01
9.67764020e-01 -9.25613567e-02 -1.91725686e-01 1.51523575e-01
6.51048005e-01 -2.18021646e-01 4.48249936e-01 4.43883389e-01
3.01183134e-01 -9.84678924e-01 7.86605000e-01 5.77412009e-01
-1.80308628e+00 2.73009092e-01 -3.14860702e-01 4.10892010e-01
1.53134197e-01 6.59830570e-01 -3.61668468e-01 1.39675820e+00
7.90612996e-01 1.11011660e+00 -5.80148101e-01 8.04125845e-01
4.41909814e-03 5.45200348e-01 -2.95487761e-01 -2.98642606e-01
3.75431061e-01 -7.41392732e-01 3.73163640e-01 1.18132305e+00
4.52799857e-01 4.52076107e-01 5.46048105e-01 5.80512464e-01
-1.49098814e-01 8.67388666e-01 -2.98627764e-01 2.73812085e-01
8.67464423e-01 1.56382632e+00 -1.34757054e+00 -6.45709872e-01
-6.07696831e-01 7.25634992e-01 1.96755022e-01 3.49957287e-01
-8.12616527e-01 -7.19229996e-01 -6.73025250e-02 -2.83054978e-01
-2.59944536e-02 -3.41290124e-02 5.01774251e-02 -1.53070819e+00
-1.31542236e-01 -7.20713913e-01 7.40065098e-01 -6.47647858e-01
-1.75168741e+00 7.14360118e-01 6.59653395e-02 -1.72141552e+00
2.66941935e-01 9.78946537e-02 -6.34241819e-01 4.28310812e-01
-1.20731723e+00 -9.94239330e-01 -3.90505105e-01 1.20053935e+00
3.55905861e-01 -2.34929696e-01 2.94577211e-01 5.81380069e-01
-8.87967408e-01 3.22798580e-01 6.90139413e-01 4.00633097e-01
8.89849007e-01 -1.03126884e+00 -2.59057850e-01 7.50318170e-01
8.56498405e-02 9.32086825e-01 4.31523472e-01 -9.89834607e-01
-1.17305136e+00 -1.23097467e+00 6.45346224e-01 -1.43838450e-01
1.17235088e+00 -1.05089195e-01 -1.09366047e+00 5.73794603e-01
6.47875845e-01 -6.05170250e-01 1.14139926e+00 1.63302451e-01
-1.73446268e-01 -3.53324064e-03 -9.25765574e-01 6.98854446e-01
7.93038487e-01 -3.38069171e-01 -1.04344308e+00 5.70094943e-01
8.09840322e-01 2.65178531e-02 -1.06504881e+00 1.32771647e-02
-1.27377838e-01 -7.92491674e-01 5.24627984e-01 -2.19464704e-01
2.71486223e-01 -7.30341375e-01 -1.84533857e-02 -1.55132866e+00
-4.13950145e-01 -5.13305128e-01 4.05512094e-01 1.92884278e+00
2.93416828e-01 -7.94692159e-01 4.47950393e-01 3.29816759e-01
-1.21995956e-01 -9.10082757e-02 -7.50007927e-01 -7.43466020e-01
-1.04600519e-01 -3.99713010e-01 7.34115124e-01 1.88507390e+00
7.27385223e-01 8.22948694e-01 -2.99758554e-01 2.04082653e-01
4.47834522e-01 3.89396250e-01 5.43202043e-01 -1.53518760e+00
2.16330793e-02 -7.22268879e-01 -1.41166955e-01 -9.90479708e-01
2.60773599e-01 -1.17986941e+00 2.50753872e-02 -1.50739169e+00
5.50686359e-01 -7.51708746e-01 -1.52140319e-01 2.01321334e-01
-3.62304956e-01 -1.87861249e-01 3.52416337e-02 9.36462820e-01
-1.02928483e+00 6.91250980e-01 1.10174263e+00 -1.39739797e-01
-3.67081374e-01 -1.22170486e-01 -7.26300299e-01 6.21998489e-01
7.56945610e-01 -4.47827458e-01 -5.40926158e-01 -1.81649119e-01
2.30852187e-01 2.10983962e-01 -3.46814781e-01 -7.49787271e-01
8.79905164e-01 -1.56233206e-01 9.53352675e-02 -8.30656707e-01
-2.09836364e-01 -1.09543502e+00 3.81871939e-01 2.17204347e-01
-1.97614327e-01 1.38023898e-01 -1.24407090e-01 8.01569819e-01
-5.53123355e-01 -9.20802876e-02 4.44966108e-01 -6.78462163e-02
-5.38234770e-01 3.32832754e-01 -6.52721643e-01 2.26993740e-01
1.17702472e+00 5.14921807e-02 -5.24119854e-01 -4.93150026e-01
-7.56950140e-01 7.08973467e-01 3.57295632e-01 4.66290444e-01
2.84992069e-01 -1.56403589e+00 -6.68268442e-01 -3.80924642e-01
1.42981350e-01 1.96686555e-02 5.08424819e-01 1.08496320e+00
-1.55582115e-01 2.80868202e-01 4.81793880e-02 -5.80548823e-01
-1.37060022e+00 8.25161457e-01 -1.80582002e-01 -2.75022954e-01
-3.85919064e-01 3.15388471e-01 5.15305176e-02 -5.25176704e-01
-3.15051675e-02 -3.75804976e-02 -7.88708568e-01 7.18727112e-01
4.32568014e-01 3.75774086e-01 -9.07408074e-02 -1.09488106e+00
-2.17907697e-01 6.85873389e-01 -5.05865039e-03 -1.70537740e-01
1.11832869e+00 -8.03927183e-01 -6.48702025e-01 6.99924886e-01
1.40636384e+00 1.99368387e-01 -3.72755587e-01 -6.70045972e-01
2.80386597e-01 -3.79556507e-01 -9.96246859e-02 -1.40680969e-01
-1.32377803e+00 4.39741611e-01 -1.03039145e-01 8.93033624e-01
1.05924428e+00 2.14485630e-01 8.15840721e-01 3.71241838e-01
3.04027855e-01 -1.27127397e+00 1.66004449e-02 3.46453995e-01
3.62639248e-01 -1.07177246e+00 -1.24021612e-01 -6.73061430e-01
-8.70284319e-01 1.09379137e+00 5.01272559e-01 4.59298491e-01
9.92343307e-01 8.60243477e-03 2.39959300e-01 -3.51118088e-01
-5.04392266e-01 -4.02237743e-01 3.88893247e-01 2.69077957e-01
2.76324540e-01 2.66659521e-02 -5.34479380e-01 7.86371589e-01
-1.97844580e-01 -7.21955299e-01 3.91956985e-01 7.94850528e-01
-4.77625996e-01 -1.14761782e+00 -6.54999971e-01 4.30685937e-01
-2.75462002e-01 8.32617581e-02 -4.34583694e-01 6.82286322e-01
1.21970601e-01 1.50422037e+00 1.97672425e-03 -8.42837632e-01
2.86773629e-02 -1.45637050e-01 -2.99027324e-01 -5.50029874e-01
-5.99202812e-01 5.05754113e-01 -1.38296083e-01 1.45986781e-01
-1.02987635e+00 -7.19656885e-01 -1.47418571e+00 -5.25378346e-01
-7.18667209e-01 9.61230874e-01 3.60075474e-01 1.12935114e+00
3.66433561e-01 4.79719073e-01 1.15503621e+00 -4.87719178e-01
2.98020363e-01 -8.63392115e-01 -8.01569462e-01 7.95452893e-01
-3.51091444e-01 -6.31531835e-01 -5.15216231e-01 8.27097446e-02] | [10.280964851379395, 6.783207893371582] |
840ed9d4-aad6-4568-8c44-5f4a0c39e0b0 | botshape-a-novel-social-bots-detection | 2303.10214 | null | https://arxiv.org/abs/2303.10214v2 | https://arxiv.org/pdf/2303.10214v2.pdf | BotShape: A Novel Social Bots Detection Approach via Behavioral Patterns | An essential topic in online social network security is how to accurately detect bot accounts and relieve their harmful impacts (e.g., misinformation, rumor, and spam) on genuine users. Based on a real-world data set, we construct behavioral sequences from raw event logs. After extracting critical characteristics from behavioral time series, we observe differences between bots and genuine users and similar patterns among bot accounts. We present a novel social bot detection system BotShape, to automatically catch behavioral sequences and characteristics as features for classifiers to detect bots. We evaluate the detection performance of our system in ground-truth instances, showing an average accuracy of 98.52% and an average f1-score of 96.65% on various types of classifiers. After comparing it with other research, we conclude that BotShape is a novel approach to profiling an account, which could improve performance for most methods by providing significant behavioral features. | ['Chengjie Mou', 'Xuesong Ye', 'Jun Wu'] | 2023-03-17 | null | null | null | null | ['misinformation'] | ['miscellaneous'] | [-1.43759742e-01 -2.85526097e-01 -5.04568696e-01 -3.59215625e-02
-4.64346521e-02 -7.99843669e-01 6.51376963e-01 2.27861777e-01
-3.55652004e-01 6.03501081e-01 -2.26577744e-01 -4.56305683e-01
3.30645323e-01 -7.84525275e-01 1.18863195e-01 -3.49217981e-01
-1.76147029e-01 3.28084886e-01 7.65320063e-01 -8.16252530e-02
8.73677313e-01 6.72600031e-01 -1.07180870e+00 2.92241275e-01
5.27710140e-01 6.68119371e-01 -2.05668062e-01 7.79843092e-01
-8.89718607e-02 1.24829674e+00 -1.30128336e+00 -7.24565983e-01
7.79488608e-02 -3.69701743e-01 -9.40294087e-01 2.31984839e-01
-1.33646913e-02 -5.73884845e-01 -3.68628234e-01 1.33782113e+00
1.24597460e-01 -3.46922934e-01 5.43343842e-01 -1.59908187e+00
-4.55133408e-01 7.17637241e-01 -8.35233152e-01 7.04619706e-01
3.86117458e-01 5.06036580e-01 9.67873096e-01 -3.09146523e-01
5.40350139e-01 1.14108670e+00 9.34917510e-01 7.54613519e-01
-1.41093302e+00 -8.47626388e-01 -2.56778061e-01 2.86561430e-01
-9.16911900e-01 -3.21421564e-01 4.13176626e-01 -7.89903283e-01
6.88632548e-01 2.41793796e-01 3.68407428e-01 1.39622653e+00
7.33598098e-02 4.94899333e-01 1.07541168e+00 4.15969528e-02
3.72217000e-02 5.90981305e-01 6.89406753e-01 7.61252999e-01
7.64135897e-01 -1.68997556e-01 -8.49925429e-02 -9.94674444e-01
3.90703857e-01 2.91849583e-01 2.43180633e-01 4.38479841e-01
-6.56274259e-01 1.15092587e+00 2.28857070e-01 3.94894898e-01
-5.27493834e-01 -2.38611817e-01 7.75477469e-01 1.39983088e-01
5.68717539e-01 5.38696349e-01 -1.48009747e-01 -1.92491785e-01
-6.07464015e-01 2.35377878e-01 9.39453602e-01 5.39395690e-01
8.27066779e-01 1.65135920e-01 -1.97476968e-01 7.86219954e-01
-4.96294573e-02 6.77649736e-01 7.40746796e-01 -7.85465360e-01
-1.13325834e-03 9.43871140e-01 2.27509245e-01 -1.43703198e+00
-3.28017682e-01 -1.42760187e-01 -5.92184782e-01 -1.98220685e-01
5.72624922e-01 -1.65674448e-01 -4.10531223e-01 1.45307577e+00
2.47301221e-01 3.06175739e-01 -6.80441618e-01 3.04415077e-01
3.82705837e-01 2.35354066e-01 1.28714979e-01 -3.94989997e-01
1.52499032e+00 -5.05813837e-01 -6.09589338e-01 -2.84958184e-01
6.11008644e-01 -4.74160582e-01 1.08683562e+00 9.50198714e-03
-5.60302258e-01 -3.29798795e-02 -4.97728139e-01 7.27569997e-01
-4.71640348e-01 -2.13783503e-01 5.71311533e-01 1.14465141e+00
-7.96535075e-01 7.44561195e-01 -6.45153582e-01 -6.11631691e-01
4.56017941e-01 1.14219218e-01 7.49298558e-02 5.77880919e-01
-9.25562620e-01 6.21784210e-01 1.20492905e-01 -5.88456929e-01
-9.71305788e-01 -3.83480459e-01 -1.81871220e-01 -6.96957782e-02
3.14524084e-01 1.18206970e-01 1.57584059e+00 -4.86462921e-01
-1.03441072e+00 1.07170665e+00 -2.82341212e-01 -7.95852423e-01
4.58901256e-01 3.45050424e-01 -5.51494002e-01 1.82379186e-01
5.02613246e-01 -3.75070840e-01 1.11066961e+00 -1.10508513e+00
-6.56533122e-01 -4.30984706e-01 -5.77393696e-02 -9.28405166e-01
-8.60523880e-01 7.54227400e-01 2.78512985e-01 -4.12126154e-01
-1.01389408e-01 -9.85242903e-01 -1.82239801e-01 -9.57815707e-01
-7.81231880e-01 -4.70941275e-01 1.27116585e+00 -8.81205320e-01
1.51445067e+00 -1.66929722e+00 -6.07693732e-01 2.16353536e-01
7.51260519e-01 8.48553836e-01 1.67422533e-01 4.61263806e-01
1.02085702e-01 7.09631205e-01 -6.53314451e-03 -3.70419353e-01
-2.77089089e-01 -1.71722203e-01 -6.79953456e-01 5.72522640e-01
1.23599172e-01 7.70811021e-01 -1.15571654e+00 -3.11173946e-01
-4.38018925e-02 -1.27445668e-01 -5.87989509e-01 4.83412832e-01
9.58345905e-02 3.15968663e-01 -4.61435497e-01 8.48551512e-01
4.61069286e-01 -5.26086330e-01 3.80875111e-01 4.55461234e-01
1.38895988e-01 6.55107439e-01 -5.50007761e-01 -3.29182968e-02
-2.41628721e-01 8.49756896e-01 1.11366935e-01 -6.14758313e-01
9.53929961e-01 1.86079234e-01 5.09310901e-01 -1.59719303e-01
5.54257751e-01 2.19456270e-01 -1.19760938e-01 -5.66253603e-01
6.41379118e-01 3.36211294e-01 -8.66545439e-02 1.18940187e+00
-2.24519163e-01 4.61000741e-01 3.73691797e-01 4.68320072e-01
1.81071794e+00 -1.02423573e+00 6.31585419e-01 -6.39499500e-02
6.62354887e-01 1.23311222e-01 3.74744833e-01 1.08203280e+00
-8.96276951e-01 -1.74522385e-01 9.90902960e-01 -3.72614652e-01
-1.04182410e+00 -6.84967935e-01 2.61646837e-01 1.18754101e+00
-7.48166740e-02 -5.95118582e-01 -1.04027224e+00 -1.19909453e+00
1.39309093e-01 3.50547552e-01 -4.33221370e-01 -2.23133981e-01
-7.35027790e-01 -1.04546535e+00 8.79819751e-01 9.48242471e-03
5.84980130e-01 -1.29626703e+00 -1.17995627e-01 2.85972297e-01
-5.24587870e-01 -1.49648702e+00 -4.65011925e-01 -3.59795481e-01
-6.67672098e-01 -1.53265417e+00 1.88785475e-02 -3.34625006e-01
4.45746362e-01 8.78467739e-01 1.03350818e+00 6.47410810e-01
-3.23699743e-01 -1.58214346e-01 -3.47496331e-01 -2.53995955e-01
-1.02143455e+00 3.18440199e-01 3.25303108e-01 3.41358423e-01
8.11321318e-01 -7.73757994e-01 -2.13189721e-01 6.97268724e-01
-4.55772370e-01 -7.75399208e-01 1.23609461e-01 4.05130774e-01
-4.56126869e-01 1.47684351e-01 8.54601860e-01 -1.34065807e+00
1.12369275e+00 -1.03700912e+00 -5.94534636e-01 -3.45474333e-01
-7.21120834e-01 -5.18600225e-01 7.74996996e-01 -8.89415383e-01
-6.34562016e-01 -3.88355672e-01 1.40697911e-01 -1.75262943e-01
-2.02498078e-01 -6.88197687e-02 4.02143091e-01 -2.22695936e-02
1.13330865e+00 3.33752990e-01 1.78389281e-01 -4.30990428e-01
-2.10741997e-01 1.21402633e+00 3.11314076e-01 -9.32619348e-02
9.74179327e-01 6.28522933e-01 -3.56678843e-01 -1.09006763e+00
-8.78296554e-01 -1.07128739e+00 -3.74608248e-01 -3.60844076e-01
4.20099109e-01 -3.65974426e-01 -1.55917668e+00 9.61632729e-01
-1.34217107e+00 -1.48102269e-01 3.41258168e-01 -3.95774618e-02
-1.58720359e-01 7.43149698e-01 -1.19710302e+00 -1.25574374e+00
-3.18829298e-01 -6.51489019e-01 5.74299753e-01 8.44510868e-02
-6.54362381e-01 -9.70275700e-01 1.62541628e-01 4.63723809e-01
5.97004831e-01 1.78163275e-01 4.91572052e-01 -1.24582028e+00
-2.93072551e-01 -7.65819013e-01 -6.07325077e-01 3.04506391e-01
3.89402181e-01 2.32992291e-01 -9.66213644e-01 -2.63521314e-01
2.08926767e-01 -1.26469776e-01 7.33577311e-01 -1.30384207e-01
1.02265525e+00 -9.21945810e-01 -5.40941298e-01 2.01779115e-03
1.02462649e+00 2.10552484e-01 4.71298426e-01 2.58905590e-01
6.02752447e-01 6.83275163e-01 1.85664460e-01 7.38370180e-01
-2.43372843e-03 4.86518949e-01 6.23816609e-01 4.81802791e-01
2.43543610e-01 -3.92419696e-01 6.88037753e-01 4.24518287e-01
-8.21821839e-02 -1.41412914e-01 -9.31391656e-01 4.53218102e-01
-1.76865327e+00 -1.58670294e+00 -5.87536871e-01 1.97239578e+00
6.16537571e-01 3.13865721e-01 1.24003088e+00 4.18114156e-01
1.48005617e+00 2.83326775e-01 -3.29808325e-01 -4.35395598e-01
3.65180761e-01 -1.63897306e-01 9.01378930e-01 4.94907141e-01
-1.05463028e+00 1.06512439e+00 7.02611923e+00 8.19101095e-01
-1.06034863e+00 3.68162930e-01 4.22086507e-01 3.05266291e-01
4.92363542e-01 -9.36064050e-02 -1.19090128e+00 8.06782961e-01
1.44397902e+00 -4.28106666e-01 8.38162899e-01 1.12993979e+00
5.30399740e-01 1.30675167e-01 -4.24725294e-01 6.00468993e-01
1.05998330e-02 -1.24171686e+00 -2.59083837e-01 5.82053840e-01
5.93055904e-01 -1.50987096e-02 -2.30260268e-01 3.96536767e-01
8.36831212e-01 -7.74956346e-01 4.15017933e-01 -1.03365937e-02
2.64688075e-01 -4.17954236e-01 7.46539295e-01 7.64559746e-01
-1.02684653e+00 -6.18432105e-01 -2.45049700e-01 -3.36455107e-01
1.36107415e-01 7.75870681e-01 -1.59843481e+00 -1.08120620e-01
3.98889631e-01 9.13681328e-01 -8.01550806e-01 8.04369926e-01
-2.13024244e-01 1.19803488e+00 -6.88250512e-02 -5.77691972e-01
4.25731875e-02 1.31811976e-01 6.92786098e-01 1.32833683e+00
-2.18107626e-01 -2.16029108e-01 2.22106010e-01 9.14580405e-01
-6.77161962e-02 -1.59912720e-01 -7.65258968e-01 -4.93074059e-01
7.45491564e-01 1.57086432e+00 -9.56094325e-01 -5.23181915e-01
4.08239551e-02 7.59829521e-01 3.54108930e-01 -4.69603091e-02
-9.67032433e-01 -4.54986721e-01 8.43978822e-01 5.83436906e-01
-1.74497917e-01 -2.16616273e-01 -5.46687186e-01 -1.22839630e+00
-2.90706038e-01 -9.80116904e-01 3.64020526e-01 -5.68739697e-02
-1.73016632e+00 6.78083837e-01 -2.01346695e-01 -9.27122474e-01
-5.19712210e-01 -6.37874424e-01 -9.51299489e-01 6.66531086e-01
-8.38491261e-01 -7.39265323e-01 -3.49287838e-01 4.70319629e-01
2.72815406e-01 -4.87237543e-01 4.67345089e-01 1.65091872e-01
-9.39983189e-01 4.06196713e-01 -2.36819997e-01 6.69207335e-01
3.35043222e-01 -1.03114223e+00 7.07139134e-01 8.25898111e-01
-2.66089201e-01 7.37419307e-01 7.78052688e-01 -9.84022677e-01
-7.02782452e-01 -1.22809184e+00 1.15467763e+00 -8.92703593e-01
1.32760739e+00 -4.08490539e-01 -8.83673191e-01 7.78834462e-01
-3.15889806e-01 -3.13559353e-01 2.42180526e-01 6.26097918e-02
-8.17500174e-01 3.57596099e-01 -1.55904651e+00 6.03971779e-01
1.08423424e+00 -7.85384417e-01 -6.96955249e-02 6.03256762e-01
3.37775260e-01 4.68057573e-01 -3.75194043e-01 -3.16438116e-02
4.66042995e-01 -1.33280730e+00 7.89477527e-01 -9.22491908e-01
3.03790182e-01 2.29983822e-01 1.72571376e-01 -8.98735821e-01
-4.50736642e-01 -8.76371503e-01 -3.64329249e-01 1.28629255e+00
5.75854033e-02 -1.14332211e+00 9.91299689e-01 3.19977924e-02
6.33957505e-01 -1.90732807e-01 -4.48638022e-01 -1.00906336e+00
-2.26090074e-01 -8.54442939e-02 5.77109993e-01 1.22054040e+00
3.67213279e-01 3.69558811e-01 -6.20610833e-01 1.25867799e-01
8.82304490e-01 -2.02539414e-01 1.11898994e+00 -1.40577042e+00
-1.32599935e-01 -6.84099793e-01 -5.33399999e-01 -4.98003066e-01
4.26880896e-01 -5.67464292e-01 -3.29066515e-01 -7.93563366e-01
5.96099257e-01 -2.27268234e-01 1.37958989e-01 4.64851677e-01
-2.18100563e-01 7.47132480e-01 -3.32006700e-02 7.16046154e-01
-2.51010597e-01 -2.45282382e-01 6.43690646e-01 -5.46586141e-02
-2.72920281e-01 6.05955899e-01 -6.32089198e-01 9.74019289e-01
1.30804944e+00 -8.58595073e-01 4.30649780e-02 5.16400695e-01
-9.36006978e-02 -6.99503273e-02 5.19372165e-01 -6.04660451e-01
-9.92381871e-02 -5.86896539e-01 -3.46410751e-01 -3.15314054e-01
1.33679003e-01 -2.57080615e-01 -3.77387971e-01 9.94835436e-01
1.09156181e-05 1.46235421e-01 -2.09261209e-01 6.74294233e-01
2.81965673e-01 -4.57689673e-01 1.21071911e+00 -4.27630007e-01
-2.63663560e-01 2.97149926e-01 -1.00625002e+00 1.13596156e-01
1.03104877e+00 3.51757673e-03 -8.71823907e-01 -5.39544582e-01
-3.99217904e-01 -2.68789887e-01 4.73840147e-01 3.43824416e-01
1.12909779e-01 -9.60679471e-01 -6.78835690e-01 2.08941698e-01
2.46848576e-02 -1.30728889e+00 -1.89146072e-01 7.33086109e-01
-4.37938064e-01 3.66750866e-01 -2.31270611e-01 -5.35530388e-01
-1.72532451e+00 5.33593714e-01 3.17469060e-01 -5.60890138e-01
-1.90744534e-01 3.41906637e-01 -5.45878351e-01 -3.38030130e-01
6.65915981e-02 4.60965484e-01 -3.44527990e-01 2.02853158e-01
9.00683939e-01 9.45347369e-01 7.48484256e-03 -8.25318515e-01
-5.23175716e-01 -2.71345139e-01 -2.47289896e-01 9.05994773e-02
9.18325305e-01 -1.32465949e-02 -5.80412149e-01 2.87290007e-01
1.00882316e+00 4.81580228e-01 -4.00165200e-01 -2.78433681e-01
6.49484575e-01 -8.63827467e-01 -5.23037493e-01 -3.47711176e-01
-8.52437854e-01 6.02334142e-01 5.12592085e-02 1.49634182e+00
4.95794028e-01 -1.85016040e-02 9.91591096e-01 3.01895589e-01
4.78303432e-01 -5.59569359e-01 6.58684850e-01 8.19637358e-01
2.38712415e-01 -1.25120258e+00 -1.92564607e-01 -7.20047235e-01
-3.70795578e-01 7.45431006e-01 6.50811911e-01 -4.26684052e-01
5.61789095e-01 -1.17517784e-01 -3.67222428e-02 -3.16426396e-01
-6.09684825e-01 -1.33617193e-01 -4.28803384e-01 7.33192265e-01
1.53631940e-01 2.76244670e-01 -4.73972738e-01 3.75478774e-01
-2.31902868e-01 -4.81963038e-01 1.11930656e+00 6.97702646e-01
-8.83396745e-01 -9.04349804e-01 -3.78849834e-01 6.41345441e-01
-9.45660055e-01 1.43206209e-01 -1.20448887e+00 4.86936241e-01
-2.76599467e-01 1.50063288e+00 -8.86678472e-02 -1.04158938e+00
2.55454313e-02 6.88418522e-02 -2.46393248e-01 -9.18991685e-01
-1.07335353e+00 -6.50084734e-01 3.82717520e-01 -4.10572410e-01
2.91018952e-02 -5.58242202e-01 -6.74259424e-01 -1.51992977e+00
-5.13771236e-01 2.19655678e-01 5.15878856e-01 7.47669876e-01
3.31531852e-01 1.33265257e-01 1.36664689e+00 -6.73714638e-01
-1.02856123e+00 -1.38892853e+00 -7.07244039e-01 6.56820178e-01
2.57019132e-01 -5.85111082e-01 -9.93545234e-01 -1.40856296e-01] | [8.135527610778809, 10.143219947814941] |
801c234f-fa4f-4a25-b410-98de33d1a972 | geospark-sparking-up-point-cloud-segmentation | 2303.08274 | null | https://arxiv.org/abs/2303.08274v1 | https://arxiv.org/pdf/2303.08274v1.pdf | GeoSpark: Sparking up Point Cloud Segmentation with Geometry Clue | Current point cloud segmentation architectures suffer from limited long-range feature modeling, as they mostly rely on aggregating information with local neighborhoods. Furthermore, in order to learn point features at multiple scales, most methods utilize a data-agnostic sampling approach to decrease the number of points after each stage. Such sampling methods, however, often discard points for small objects in the early stages, leading to inadequate feature learning. We believe these issues are can be mitigated by introducing explicit geometry clues as guidance. To this end, we propose GeoSpark, a Plug-in module that incorporates Geometry clues into the network to Spark up feature learning and downsampling. GeoSpark can be easily integrated into various backbones. For feature aggregation, it improves feature modeling by allowing the network to learn from both local points and neighboring geometry partitions, resulting in an enlarged data-tailored receptive field. Additionally, GeoSpark utilizes geometry partition information to guide the downsampling process, where points with unique features are preserved while redundant points are fused, resulting in better preservation of key points throughout the network. We observed consistent improvements after adding GeoSpark to various backbones including PointNet++, KPConv, and PointTransformer. Notably, when integrated with Point Transformer, our GeoSpark module achieves a 74.7% mIoU on the ScanNetv2 dataset (4.1% improvement) and 71.5% mIoU on the S3DIS Area 5 dataset (1.1% improvement), ranking top on both benchmarks. Code and models will be made publicly available. | ['Ioannis Brilakis', 'Georgios Hadjidemetriou', 'Shujun Wang', 'Lei Zhu', 'Hengshuang Zhao', 'Xiaoyang Wu', 'Zhening Huang'] | 2023-03-14 | null | null | null | null | ['point-cloud-segmentation'] | ['computer-vision'] | [-3.12664032e-01 -2.88252592e-01 -2.16178194e-01 -4.05704767e-01
-6.88692212e-01 -7.65167594e-01 5.41158199e-01 5.18113315e-01
-3.00065309e-01 2.87359208e-01 3.26836072e-02 -5.52330501e-02
-3.91469784e-02 -1.14873624e+00 -8.53321791e-01 -4.44544494e-01
-4.90094982e-02 4.47463602e-01 5.97539723e-01 -3.06293219e-01
2.63581306e-01 1.04022443e+00 -1.56035423e+00 1.67179465e-01
9.44128931e-01 1.15272045e+00 2.49661803e-02 2.02475056e-01
-3.33566695e-01 5.97731508e-02 -3.19680691e-01 2.47637760e-02
6.76158845e-01 3.79699379e-01 -6.29679859e-01 -2.50600398e-01
7.64437497e-01 -4.98902440e-01 -2.19925493e-01 7.28995740e-01
3.79466951e-01 2.71915972e-01 3.73609364e-01 -1.20591843e+00
-3.69682729e-01 3.47997129e-01 -8.10773075e-01 -4.33581658e-02
-2.97972187e-02 3.69766921e-01 1.18210161e+00 -1.23235154e+00
4.04969156e-01 1.10492122e+00 9.67764556e-01 1.35330170e-01
-1.45771253e+00 -8.12009513e-01 3.10688645e-01 -2.64114700e-02
-1.61044395e+00 -2.70168513e-01 9.87734258e-01 -2.22286135e-01
1.16988254e+00 3.48370910e-01 9.35839951e-01 3.77624303e-01
-2.17367813e-01 6.96270823e-01 6.24026239e-01 8.81132185e-02
2.37034321e-01 -1.72742546e-01 2.56860852e-01 4.31774020e-01
2.55502820e-01 -1.51337618e-02 -6.00678682e-01 -2.16950387e-01
1.05164981e+00 5.10028660e-01 -9.91487503e-02 -6.20713472e-01
-1.07527256e+00 8.59427929e-01 1.24845886e+00 -2.71797143e-02
-5.63715518e-01 7.30473846e-02 1.20796762e-01 9.78673398e-02
5.30214369e-01 6.33640945e-01 -6.52002990e-01 -5.64508177e-02
-1.22079492e+00 4.77154166e-01 5.21605849e-01 1.12008977e+00
1.51010263e+00 -1.96185559e-01 -2.41581816e-02 7.96605408e-01
1.00401610e-01 5.20304859e-01 -3.73668261e-02 -1.02467144e+00
5.14755726e-01 1.15922022e+00 -1.55669093e-01 -1.08018172e+00
-6.93733752e-01 -7.14290082e-01 -6.41079724e-01 4.09609377e-01
2.55266339e-01 1.18481770e-01 -1.05709219e+00 1.37573326e+00
4.90814418e-01 2.10574627e-01 -2.47830763e-01 8.35040689e-01
8.77456546e-01 5.50777555e-01 -2.03425825e-01 4.46906447e-01
1.21346617e+00 -9.24933970e-01 2.44436517e-01 -1.17160603e-01
6.80920303e-01 -6.73054099e-01 1.23263884e+00 2.86049873e-01
-9.68428075e-01 -5.41193783e-01 -9.54064906e-01 -2.97149181e-01
-3.57412964e-01 -2.43634939e-01 6.92067742e-01 1.68466464e-01
-1.12819111e+00 8.83107126e-01 -1.08067489e+00 -1.82837173e-01
8.47375810e-01 6.75024152e-01 -2.76153356e-01 -2.59359062e-01
-4.83774096e-01 2.74325877e-01 7.81321004e-02 -2.10436061e-01
-3.85218948e-01 -1.38125753e+00 -8.29681695e-01 2.83099681e-01
2.74362206e-01 -7.32371569e-01 9.90121186e-01 -4.67594922e-01
-7.93422699e-01 3.46499890e-01 -2.92569131e-01 -4.06623483e-01
4.16844577e-01 -3.45819473e-01 -6.59497008e-02 3.34974110e-01
3.04765731e-01 1.17495799e+00 5.77530742e-01 -1.28733444e+00
-8.84869635e-01 -5.96676469e-01 1.87229320e-01 2.49674246e-01
-2.50417352e-01 -5.16694188e-01 -6.55219972e-01 -6.50213480e-01
5.10632873e-01 -9.44662929e-01 -3.70819330e-01 2.75981337e-01
-2.30595797e-01 -3.01329643e-01 1.21382701e+00 -1.56152323e-01
8.27984512e-01 -2.26597571e+00 -3.04482490e-01 5.21834612e-01
5.24651766e-01 4.83766943e-02 -2.00972185e-01 2.59994149e-01
-4.61581023e-03 2.27225348e-01 -1.59418598e-01 -4.81102943e-01
-1.96180105e-01 1.74059212e-01 -2.80418068e-01 4.21193659e-01
4.68414098e-01 9.38228786e-01 -7.57314205e-01 -2.29728654e-01
5.63947618e-01 6.99224234e-01 -1.08001697e+00 -3.60477924e-01
-2.09092304e-01 3.07325870e-01 -5.66218555e-01 8.93142760e-01
1.10085583e+00 -3.73685539e-01 -5.89057148e-01 -4.71747458e-01
-3.50827545e-01 3.32812279e-01 -1.16233456e+00 1.93478560e+00
-5.10742605e-01 4.28120852e-01 1.56490542e-02 -3.60211730e-01
9.55550730e-01 -1.82130620e-01 9.14324999e-01 -4.98377800e-01
-1.19658425e-01 1.27043843e-01 -1.69686094e-01 1.80389732e-01
8.20532143e-01 2.25653410e-01 1.77425355e-01 2.43499070e-01
-7.75138959e-02 -3.27657789e-01 -1.41614547e-03 2.49861300e-01
1.21774364e+00 7.98828155e-02 -1.03050366e-01 -2.82163292e-01
7.72933215e-02 4.22319591e-01 5.80389082e-01 6.45282567e-01
1.50965899e-01 1.01689136e+00 2.24228054e-01 -6.32487535e-01
-9.09564734e-01 -1.11948299e+00 -4.37623233e-01 9.23159838e-01
3.53287905e-01 -7.04078138e-01 -4.86246228e-01 -6.00987375e-01
3.70319486e-01 5.27580559e-01 -3.71843606e-01 -1.24437936e-01
-6.25443935e-01 -4.30080801e-01 3.08322519e-01 8.45719874e-01
5.35420299e-01 -8.06591809e-01 -4.52326715e-01 2.28161752e-01
1.47046030e-01 -8.70311141e-01 -4.67044383e-01 3.56696904e-01
-1.19207633e+00 -9.15541172e-01 -4.17060375e-01 -4.85767990e-01
8.12534273e-01 7.56572902e-01 1.06839728e+00 3.52375865e-01
-7.26141334e-02 4.95757125e-02 -4.87920016e-01 -2.24555194e-01
2.92820543e-01 6.14995599e-01 5.01486696e-02 -3.60185415e-01
5.12211204e-01 -9.31392491e-01 -9.62374091e-01 4.46053684e-01
-8.28751147e-01 5.72597347e-02 5.47302246e-01 6.82970941e-01
8.49554777e-01 -1.90865442e-01 3.06527466e-01 -5.43276668e-01
5.30305058e-02 -4.77560908e-01 -6.01618588e-01 -2.95766115e-01
-5.21531284e-01 3.26718166e-02 7.63015151e-01 -1.15668520e-01
-5.69786370e-01 2.35469192e-01 -2.49886751e-01 -6.73153222e-01
-1.77263245e-01 2.90261805e-01 -1.18616611e-01 -5.29116213e-01
7.13791847e-01 1.13420799e-01 1.58816818e-02 -6.27265036e-01
4.00559008e-01 4.08982724e-01 4.05787975e-01 -7.86060989e-01
1.02116764e+00 7.69031107e-01 -6.73482660e-03 -7.55666196e-01
-5.78725159e-01 -7.12907016e-01 -6.77287638e-01 1.66883826e-01
5.70519269e-01 -1.08064318e+00 -6.26272321e-01 3.19805145e-01
-9.55470145e-01 -2.37856150e-01 -6.23588920e-01 3.35821122e-01
-2.55166113e-01 5.08666486e-02 -4.61678416e-01 -1.84843227e-01
-4.04974550e-01 -1.23089039e+00 1.28202915e+00 4.14326191e-01
-2.62446254e-01 -7.05876052e-01 -3.09444219e-01 1.11195132e-01
4.28862840e-01 1.72342390e-01 7.46540666e-01 -7.36206710e-01
-9.51686859e-01 -2.30619192e-01 -4.93947715e-01 2.37307176e-01
2.97063023e-01 -1.08034359e-02 -1.02958655e+00 -3.92535955e-01
-2.72163361e-01 -1.03955623e-02 9.70545590e-01 2.44954661e-01
1.35960877e+00 4.00373600e-02 -4.95956868e-01 1.16677344e+00
1.34326947e+00 -1.68102488e-01 3.27281326e-01 4.05570686e-01
1.05041313e+00 3.28037351e-01 5.62867403e-01 4.89725411e-01
4.90794718e-01 5.46039164e-01 7.16306984e-01 -3.52574050e-01
-1.75259084e-01 -3.09477776e-01 -1.07156530e-01 5.21419942e-01
-9.81347412e-02 1.45696715e-01 -1.02976215e+00 5.87349892e-01
-1.71464348e+00 -6.85171604e-01 -2.03661621e-01 2.12257385e+00
5.34568131e-01 2.44092673e-01 1.63244307e-01 -9.10684988e-02
4.41803604e-01 2.54088938e-01 -8.33743572e-01 1.76707789e-01
-1.05050452e-01 3.92933100e-01 8.46774757e-01 2.95922965e-01
-1.08508039e+00 1.13259161e+00 5.34887552e+00 9.37728941e-01
-1.37888777e+00 -1.98423535e-01 5.15070915e-01 -5.21963179e-01
-4.56052750e-01 1.98231101e-01 -9.64083850e-01 3.02001089e-01
3.49529147e-01 9.23680365e-02 2.64890164e-01 1.12705886e+00
2.19105780e-01 -3.16610970e-02 -1.03031540e+00 9.42704201e-01
-3.36473167e-01 -1.68747246e+00 1.41522095e-01 2.90898800e-01
7.05715775e-01 7.27897763e-01 -3.50140445e-02 2.32102752e-01
2.02338785e-01 -8.64963114e-01 6.37446105e-01 2.05204338e-01
7.17116356e-01 -1.23804486e+00 4.61098582e-01 2.14892939e-01
-1.51071894e+00 8.13743547e-02 -5.55205882e-01 -3.45779955e-02
6.19163886e-02 7.73816407e-01 -8.29885244e-01 6.42574251e-01
9.50658321e-01 8.83749843e-01 -7.38068283e-01 1.25473011e+00
1.30702794e-01 5.09852469e-01 -9.80287015e-01 3.14115316e-01
4.22765285e-01 -1.12587050e-01 6.19482100e-01 9.16108072e-01
3.69856864e-01 -8.00690725e-02 4.11839604e-01 1.00239277e+00
-1.17648043e-01 8.00368860e-02 -5.26044607e-01 4.70522553e-01
8.21435511e-01 1.56344521e+00 -1.04652011e+00 -1.59825444e-01
-5.44941962e-01 6.39007688e-01 3.01405132e-01 2.21101969e-01
-6.37217283e-01 -5.25647223e-01 1.14086616e+00 5.51861644e-01
4.94424552e-01 -4.05404508e-01 -7.40178525e-01 -9.96363342e-01
1.29965693e-01 -5.53915739e-01 2.82633547e-02 -4.87249404e-01
-1.14924765e+00 5.89405775e-01 7.93138612e-03 -1.41631329e+00
1.23022839e-01 -4.12766427e-01 -7.00008571e-01 8.52517724e-01
-1.50464809e+00 -1.25239539e+00 -6.18788838e-01 5.89232147e-01
4.95486826e-01 7.23229647e-02 4.01903600e-01 3.65304589e-01
-3.22266012e-01 6.09187901e-01 -1.26287863e-01 2.70165578e-02
6.35686934e-01 -1.20066333e+00 8.22627366e-01 6.51614666e-01
2.17502654e-01 9.38478231e-01 3.29494953e-01 -7.60016203e-01
-1.26875651e+00 -1.39575136e+00 3.43873203e-01 -4.50766653e-01
4.66107994e-01 -3.69773269e-01 -1.20492351e+00 5.34971952e-01
-3.14017892e-01 4.98211592e-01 3.71358037e-01 3.09279561e-01
-3.75571460e-01 -3.61029536e-01 -1.21495962e+00 6.66092277e-01
1.22996187e+00 -3.51127833e-01 -3.07687402e-01 1.40631780e-01
9.58698511e-01 -4.71578389e-01 -1.08520067e+00 6.32780910e-01
2.30044812e-01 -9.62155640e-01 1.16782224e+00 -1.07769437e-01
2.16919437e-01 -6.27569377e-01 -2.01208144e-01 -1.29617500e+00
-5.32462537e-01 -4.12742674e-01 1.27977105e-02 1.29532671e+00
4.00827825e-01 -6.96572423e-01 1.22202075e+00 6.20386183e-01
-4.23221052e-01 -9.71889138e-01 -8.52919102e-01 -7.11161256e-01
1.44308135e-02 -7.41559803e-01 1.23506474e+00 8.51280749e-01
-4.82162416e-01 2.10315347e-01 2.01668054e-01 3.14880967e-01
4.69682336e-01 2.40498334e-01 1.02994895e+00 -1.52860832e+00
6.20051175e-02 -6.56304061e-01 -4.90201712e-01 -1.37433875e+00
-2.42468417e-01 -9.78314161e-01 -1.58087775e-01 -1.55077565e+00
-9.28238183e-02 -1.01194501e+00 -1.07092641e-01 7.93596625e-01
-1.34821653e-01 5.16111553e-01 3.36773902e-01 4.33029622e-01
-3.74701947e-01 5.78323722e-01 1.07342339e+00 9.72076729e-02
-5.29389620e-01 4.84283008e-02 -8.46208572e-01 9.00039434e-01
8.98615837e-01 -4.69245017e-01 -2.43915305e-01 -5.70947051e-01
8.34641755e-02 -5.38220048e-01 6.05921745e-01 -1.38424456e+00
3.42841864e-01 -9.67603177e-03 6.80604637e-01 -1.17252958e+00
5.76049685e-01 -9.57438588e-01 1.72953531e-02 1.29921690e-01
1.95577353e-01 2.80340672e-01 4.82549489e-01 2.77893603e-01
-1.82139829e-01 1.17450982e-01 6.33549094e-01 2.33776476e-02
-7.19179451e-01 7.40289450e-01 2.75933325e-01 -1.87198326e-01
9.22508776e-01 -5.76904714e-01 -3.61657739e-01 -6.52690157e-02
-4.68507618e-01 4.81394708e-01 1.09858227e+00 3.41322541e-01
5.93386352e-01 -1.29584014e+00 -3.57309073e-01 5.70365667e-01
1.97343782e-01 1.06174052e+00 3.60321492e-01 8.30105901e-01
-7.58633614e-01 7.74214044e-02 -4.55890186e-02 -9.82535779e-01
-1.01855958e+00 2.45912135e-01 7.60102794e-02 8.22051764e-02
-8.89317334e-01 8.14154506e-01 2.81552821e-01 -5.04806280e-01
-1.12916180e-03 -7.10857272e-01 1.28640726e-01 3.56806703e-02
3.84303719e-01 4.19331610e-01 3.60912651e-01 -6.07521534e-01
-5.65457642e-01 8.01297903e-01 -3.77615780e-01 1.80647597e-01
1.62711668e+00 7.75159672e-02 -1.40280843e-01 1.10770062e-01
1.32643235e+00 3.71687651e-01 -1.67055655e+00 -4.40723836e-01
-1.83828041e-01 -4.80046183e-01 1.61703557e-01 -4.64929461e-01
-1.45716906e+00 6.34819984e-01 3.81653279e-01 1.02124289e-01
9.99865174e-01 2.33488306e-01 9.74641562e-01 4.39511567e-01
5.74534059e-01 -8.55493724e-01 -2.69892544e-01 5.28509200e-01
7.89209425e-01 -1.19783092e+00 1.61743402e-01 -5.35466969e-01
-3.66694272e-01 1.07389450e+00 8.11815083e-01 -5.17301977e-01
7.47200906e-01 3.69168133e-01 -2.06750855e-02 -3.91716212e-01
-5.00457346e-01 -9.64695960e-02 3.77921462e-01 4.47342515e-01
1.44304097e-01 5.01049962e-03 3.42890441e-01 4.04271662e-01
-6.29022419e-01 -2.38846064e-01 6.87317923e-02 9.42201674e-01
-5.41578948e-01 -9.58798170e-01 -4.90815759e-01 8.08750570e-01
-1.02007404e-01 -2.54094839e-01 -2.14118466e-01 8.31459045e-01
2.60076970e-01 6.36571050e-01 4.97759849e-01 -5.18915355e-01
5.62217355e-01 -3.39332849e-01 5.56999743e-02 -7.59059846e-01
-7.01784492e-01 1.12713374e-01 -3.66274744e-01 -9.97157872e-01
-3.10929157e-02 -7.53404319e-01 -1.62953877e+00 -4.78209317e-01
-2.67681688e-01 5.49239554e-02 6.34398282e-01 6.63268745e-01
8.51761818e-01 5.25267363e-01 5.00496447e-01 -1.30965149e+00
-3.19171697e-01 -7.13303685e-01 -3.62451732e-01 9.02054161e-02
4.13715988e-01 -6.12185121e-01 -2.98218697e-01 -4.52719599e-01] | [7.88925313949585, -3.469700574874878] |
e6da7e66-c66d-4bb1-be55-d80944320030 | egocentric-video-description-based-on | 1704.02163 | null | http://arxiv.org/abs/1704.02163v3 | http://arxiv.org/pdf/1704.02163v3.pdf | Egocentric Video Description based on Temporally-Linked Sequences | Egocentric vision consists in acquiring images along the day from a first
person point-of-view using wearable cameras. The automatic analysis of this
information allows to discover daily patterns for improving the quality of life
of the user. A natural topic that arises in egocentric vision is storytelling,
that is, how to understand and tell the story relying behind the pictures. In
this paper, we tackle storytelling as an egocentric sequences description
problem. We propose a novel methodology that exploits information from
temporally neighboring events, matching precisely the nature of egocentric
sequences. Furthermore, we present a new method for multimodal data fusion
consisting on a multi-input attention recurrent network. We also publish the
first dataset for egocentric image sequences description, consisting of 1,339
events with 3,991 descriptions, from 55 days acquired by 11 people.
Furthermore, we prove that our proposal outperforms classical attentional
encoder-decoder methods for video description. | ['Francisco Casacuberta', 'Álvaro Peris', 'Marc Bolaños', 'Petia Radeva', 'Sergi Soler'] | 2017-04-07 | null | null | null | null | ['video-description'] | ['computer-vision'] | [ 2.81903774e-01 -1.18091032e-01 9.17144418e-02 -3.63151193e-01
-6.42506063e-01 -3.54665786e-01 8.32585871e-01 1.27113223e-01
-2.91361481e-01 7.15071738e-01 9.43267405e-01 7.34788597e-01
-2.06663758e-01 -3.91042262e-01 -8.80698144e-01 -7.11385131e-01
2.77273148e-01 2.57603824e-01 -1.61511570e-01 -1.67401150e-01
3.87276441e-01 1.07873499e-01 -1.92556489e+00 5.34871697e-01
4.11674939e-02 6.71211720e-01 6.09123588e-01 1.07481217e+00
3.04468483e-01 1.10322821e+00 -2.43534073e-01 -5.42442977e-01
-3.98795098e-01 -5.72238684e-01 -6.01770520e-01 4.61993217e-01
4.30773169e-01 -4.83026147e-01 -6.64476335e-01 1.05827701e+00
5.83893776e-01 4.37836230e-01 5.05079687e-01 -1.01682746e+00
-4.44319218e-01 6.97822452e-01 -3.19550663e-01 2.42580026e-01
8.64798963e-01 -5.52986488e-02 8.93694699e-01 -8.66401732e-01
1.07636154e+00 9.30737913e-01 4.05445457e-01 5.34403741e-01
-8.66317034e-01 -1.09255746e-01 -8.69040862e-02 9.71683264e-01
-1.38542581e+00 -8.09710920e-01 9.01899695e-01 -5.70970416e-01
8.91050100e-01 1.42616346e-01 9.47806835e-01 1.60795975e+00
1.70263737e-01 8.29729795e-01 5.18150091e-01 -3.50105047e-01
7.28850514e-02 2.17234075e-01 2.64822841e-01 2.86586374e-01
9.09872428e-02 -3.77917439e-01 -8.81279945e-01 2.34814942e-01
5.88910520e-01 4.25371706e-01 -4.43533272e-01 -6.16442084e-01
-1.69432390e+00 5.09874523e-01 -4.25125994e-02 5.81351161e-01
-8.22064877e-01 4.31159735e-01 5.94440877e-01 -1.39843389e-01
3.61204654e-01 9.47914496e-02 1.50741816e-01 -5.09154320e-01
-7.32587039e-01 2.75068939e-01 5.25211096e-01 1.15737736e+00
5.32473922e-01 -1.50324181e-01 -3.01155627e-01 5.64965308e-01
-3.26880477e-02 5.51653624e-01 5.78139186e-01 -1.27940214e+00
2.05026507e-01 1.26156449e-01 2.23354161e-01 -1.07425737e+00
-4.42613721e-01 -1.99544668e-01 -7.20090806e-01 -5.09900987e-01
1.72673434e-01 -1.19196624e-01 -2.34019011e-01 1.73693895e+00
-8.13413784e-02 2.56687909e-01 1.62368461e-01 1.15157497e+00
9.11805451e-01 6.01574183e-01 -2.42473006e-01 -4.13681507e-01
1.81992149e+00 -7.72090971e-01 -1.11422479e+00 2.33259857e-01
3.18805963e-01 -5.72513044e-01 5.62074125e-01 4.29285467e-01
-1.47866416e+00 -5.26164353e-01 -8.11708987e-01 -1.75730214e-01
-2.97680438e-01 1.08344622e-01 1.55560166e-01 1.57505214e-01
-1.06118286e+00 3.52126420e-01 -2.61709809e-01 -1.02449942e+00
-1.49589255e-01 1.90092120e-02 -7.01084316e-01 -9.40728188e-02
-1.01340985e+00 8.26020241e-01 8.45611870e-01 -1.82082474e-01
-1.07185686e+00 -3.29063386e-01 -8.69161010e-01 3.50681603e-01
4.18034852e-01 -1.16861773e+00 1.21200132e+00 -9.47651744e-01
-1.19354653e+00 9.77775931e-01 -5.84261298e-01 -7.61104345e-01
2.68057257e-01 -5.05780935e-01 -3.50857943e-01 5.21663904e-01
1.54241011e-01 7.40203798e-01 9.90882933e-01 -1.23320413e+00
-6.84727967e-01 -4.43819255e-01 1.95989877e-01 4.99550760e-01
-3.78115654e-01 -6.80475086e-02 -7.24042535e-01 -4.68472809e-01
-1.46907046e-01 -6.54559851e-01 1.93082109e-01 -5.53039551e-01
-3.68820816e-01 -6.87207505e-02 6.71806812e-01 -7.46061683e-01
1.00531137e+00 -2.22555995e+00 5.93958497e-01 -2.90258288e-01
4.97703850e-01 -1.75455108e-01 1.42639056e-01 6.50861263e-01
-3.42346519e-01 -4.77502584e-01 -1.80343077e-01 -7.27794468e-01
-6.95545599e-02 1.30183950e-01 -4.41677332e-01 4.58528161e-01
-1.79059193e-01 8.21135521e-01 -1.07771111e+00 -5.62731683e-01
4.45949793e-01 5.69760859e-01 -4.58859682e-01 3.32463533e-01
-1.86495390e-02 5.14964521e-01 -2.09182724e-01 4.29205507e-01
2.24032521e-01 -1.64942354e-01 2.16166340e-02 -3.12218130e-01
-3.65340441e-01 -2.80134737e-01 -6.55957282e-01 2.41120291e+00
-3.11549306e-01 1.07917356e+00 -3.51922005e-01 -9.46086228e-01
6.55696273e-01 7.39328086e-01 6.52347565e-01 -5.64351678e-01
3.28878462e-01 -2.07454517e-01 -6.57549620e-01 -1.06550694e+00
9.46747422e-01 -7.48315826e-02 -2.70348221e-01 5.07791877e-01
2.22256705e-01 2.41184950e-01 5.71647644e-01 2.06190631e-01
1.00140965e+00 1.83511749e-01 6.07243896e-01 2.61874735e-01
6.63892806e-01 -5.33216968e-02 2.81849265e-01 8.14015090e-01
-1.77386448e-01 1.05034244e+00 4.21333998e-01 -5.91318071e-01
-1.32848418e+00 -1.11038065e+00 2.37668484e-01 8.05569351e-01
2.07769543e-01 -4.26593542e-01 -8.78136277e-01 -1.53902277e-01
-5.03313780e-01 9.47962284e-01 -6.61106408e-01 -1.40063062e-01
-4.02923822e-01 -3.34613502e-01 3.39442670e-01 2.94793546e-01
7.75465071e-01 -1.17889273e+00 -9.81294334e-01 1.59449592e-01
-9.36372995e-01 -1.50201583e+00 -5.70049822e-01 -3.91437560e-01
-5.83810627e-01 -1.00929248e+00 -1.23146880e+00 -4.12426502e-01
4.50593799e-01 7.25148499e-01 1.10955250e+00 -7.12458313e-01
-2.44544357e-01 1.23962295e+00 -5.58075607e-01 -9.10963267e-02
-3.36745903e-02 -2.28927568e-01 2.71245629e-01 6.09842539e-01
4.79414463e-01 -7.67066777e-01 -4.78052884e-01 -4.31122184e-02
-5.63458681e-01 2.98473150e-01 4.51275676e-01 5.91454744e-01
4.01461601e-01 -2.32622907e-01 2.71413058e-01 -5.43100774e-01
4.00592744e-01 -8.40322256e-01 -4.15043794e-02 3.39733839e-01
8.26732889e-02 -6.35893410e-03 3.43947470e-01 -2.07148343e-01
-1.31333268e+00 3.16166729e-01 2.31933862e-01 -9.37950432e-01
-4.02956873e-01 1.00490630e-01 -1.08771084e-03 4.86259311e-01
3.91377896e-01 6.54897511e-01 -3.40672344e-01 -3.14765871e-01
6.32864296e-01 4.86050189e-01 1.09538949e+00 -9.80071798e-02
3.43145043e-01 8.47902000e-01 -2.69189328e-01 -1.30552042e+00
-5.50637484e-01 -9.94602799e-01 -7.07904160e-01 -7.88291991e-01
1.31023562e+00 -1.11418962e+00 -1.16325915e+00 1.84031323e-01
-1.76853740e+00 4.57492113e-01 -4.10490155e-01 7.58504272e-01
-1.37045300e+00 6.11041784e-01 -1.95734337e-01 -9.55033064e-01
-1.47981882e-01 -7.21744001e-01 1.13593364e+00 1.62399799e-01
-4.70959991e-01 -8.58517349e-01 2.37888396e-01 5.88039160e-01
4.59931046e-02 4.00945902e-01 3.19541126e-01 -3.54393631e-01
-7.40502000e-01 -8.25735852e-02 -3.27343345e-01 -2.19408855e-01
2.87238061e-02 -4.17550534e-01 -1.18045163e+00 6.39266893e-02
4.25425649e-01 6.92959428e-02 9.84714210e-01 6.85613692e-01
9.36947763e-01 -2.44248614e-01 -1.18661165e-01 4.32182670e-01
1.32535350e+00 2.75590032e-01 7.80009031e-01 1.20100178e-01
8.55032504e-01 9.34168518e-01 4.98339206e-01 9.93574321e-01
5.47395527e-01 8.19947422e-01 5.64432263e-01 3.32746387e-01
9.55334008e-02 -1.65345624e-01 5.73643148e-01 8.02140534e-01
-5.00999093e-01 -4.37607884e-01 -5.84560871e-01 9.43716526e-01
-2.34062815e+00 -1.65568197e+00 -6.03475831e-02 2.07046413e+00
-2.60352120e-02 -2.61203915e-01 1.97809875e-01 2.61371657e-02
1.10685277e+00 2.80397207e-01 -3.81119996e-01 -3.39814395e-01
-3.29093218e-01 -6.61934853e-01 3.71142477e-01 2.08941311e-01
-1.10440910e+00 6.50389910e-01 5.98778248e+00 4.70851213e-01
-6.70493662e-01 2.81602979e-01 1.11088581e-01 -4.46668893e-01
-1.50684491e-01 -2.36307934e-01 -5.88915348e-01 4.52231050e-01
9.81213629e-01 -4.50237334e-01 3.36396486e-01 7.52732694e-01
7.04900384e-01 -3.71481001e-01 -1.10245311e+00 1.68971002e+00
1.01255894e+00 -1.39992940e+00 2.56997257e-01 -1.64999992e-01
6.72936082e-01 -1.96381718e-01 -2.51230132e-02 -2.86368281e-01
-1.87403798e-01 -5.62249184e-01 8.94931257e-01 1.29361653e+00
5.06874382e-01 -7.97203183e-01 7.91367292e-01 3.66491765e-01
-1.16779065e+00 -2.19696373e-01 -7.43454844e-02 -1.99447162e-02
7.26979136e-01 3.62930626e-01 -8.32660377e-01 6.67079568e-01
8.24172795e-01 1.52018130e+00 -2.87288487e-01 9.90514576e-01
-5.27708908e-04 3.14673893e-02 1.81192160e-01 1.38356723e-02
1.06279366e-01 -2.89362490e-01 1.18158317e+00 1.34084404e+00
6.94574356e-01 2.12599233e-01 -6.49254858e-01 8.37163448e-01
-5.33386059e-02 -1.60060510e-01 -1.36778951e+00 -3.30812745e-02
-1.16349891e-01 9.87568557e-01 -5.53016067e-01 -5.47683418e-01
-3.39498252e-01 1.25364959e+00 -2.68196724e-02 3.48231882e-01
-9.31888819e-01 -1.42992809e-01 5.85239649e-01 -3.42718065e-02
3.45176041e-01 -2.94012070e-01 1.63153604e-01 -1.48326075e+00
1.62680242e-02 -3.51834059e-01 3.48395944e-01 -1.55570555e+00
-8.54667604e-01 4.81722265e-01 3.94276619e-01 -1.38824856e+00
-7.77809203e-01 -1.33134425e-01 -5.00544667e-01 1.97898418e-01
-1.04836738e+00 -1.07233846e+00 -7.19545424e-01 8.16050172e-01
1.16242898e+00 -2.00682089e-01 5.53105295e-01 4.08167094e-01
-3.77747238e-01 -7.34086111e-02 1.56251505e-01 -1.55134186e-01
8.34374070e-01 -1.06547797e+00 1.90981328e-01 8.56378555e-01
3.74935865e-01 5.00133097e-01 1.06659770e+00 -4.57224071e-01
-1.38163912e+00 -7.79863894e-01 1.33855391e+00 -4.91910726e-01
4.56197143e-01 -1.55704096e-01 -4.76235956e-01 8.18428397e-01
5.50278842e-01 -6.13797545e-01 4.04850394e-01 -3.47925127e-01
-5.58804208e-03 9.09555890e-03 -7.31195271e-01 7.56862223e-01
1.27520740e+00 -7.97066212e-01 -7.91069150e-01 3.44654441e-01
8.18498909e-01 -1.05934389e-01 -5.25974452e-01 -3.21088061e-02
7.06460416e-01 -1.30401731e+00 1.07099545e+00 -3.26149672e-01
6.47779524e-01 -2.27446139e-01 -4.11361009e-01 -9.73483980e-01
-1.60941541e-01 -6.86734438e-01 -1.68069050e-01 1.20263243e+00
-2.46755138e-01 -1.53622061e-01 5.60286880e-01 1.92154944e-01
-9.20743570e-02 9.29737166e-02 -8.01300287e-01 -2.87068844e-01
-9.27481353e-01 -5.54118574e-01 3.11453313e-01 7.00464249e-01
1.05610453e-01 2.75322825e-01 -1.08546686e+00 -2.51430333e-01
8.37004960e-01 -6.98939711e-02 7.37510622e-01 -9.41393197e-01
-6.92060217e-02 -1.47226647e-01 -7.71523774e-01 -9.35807407e-01
1.84206262e-01 -5.86176634e-01 -4.26113121e-02 -1.31821489e+00
6.38214052e-01 8.24866176e-01 -5.85142709e-03 -9.84645560e-02
3.94325554e-01 3.98056269e-01 2.91235417e-01 2.49987766e-01
-1.10599077e+00 6.89121366e-01 8.75642657e-01 -1.53475076e-01
1.79835353e-02 -3.11265111e-01 -3.97056878e-01 8.41230929e-01
4.52763915e-01 -3.15989494e-01 -3.80405188e-01 -3.78686875e-01
2.99190849e-01 4.64775085e-01 7.48843610e-01 -1.25618958e+00
6.26670063e-01 1.86498955e-01 1.74651265e-01 -1.04720640e+00
9.81205881e-01 -9.11468506e-01 4.29967582e-01 3.37379754e-01
-3.87726575e-01 3.09107721e-01 -1.61044568e-01 9.01939511e-01
-4.14419949e-01 -3.18163961e-01 2.82800794e-01 -3.01228762e-01
-1.17114210e+00 -5.09349331e-02 -7.16215253e-01 -1.32692739e-01
1.23281884e+00 -4.42358077e-01 -2.06990823e-01 -9.62750435e-01
-1.03524017e+00 9.50350836e-02 3.51971686e-01 4.47964728e-01
1.03063405e+00 -1.41868663e+00 -8.04137349e-01 -1.32178888e-01
4.84209925e-01 -5.24506152e-01 8.28069985e-01 8.30778539e-01
-3.99314314e-01 7.27580190e-01 -5.93249023e-01 -7.23999977e-01
-1.59325957e+00 7.24830031e-01 9.84147042e-02 1.34102523e-01
-8.80297422e-01 2.72237062e-01 5.43875992e-01 5.23405850e-01
5.74599281e-02 1.18002377e-01 -7.94267774e-01 4.92291033e-01
6.57822311e-01 6.23824239e-01 -4.25893605e-01 -1.12312639e+00
-1.81142718e-01 9.89949763e-01 1.64117903e-01 -5.27912676e-01
1.28917229e+00 -8.88508379e-01 -9.54758562e-03 1.06090117e+00
1.18957019e+00 -4.92752880e-01 -1.04409564e+00 -2.53534436e-01
-1.19592436e-01 -3.54159594e-01 -2.16806650e-01 -8.18057582e-02
-6.91779017e-01 8.36041868e-01 4.66253847e-01 3.95477377e-02
1.22044349e+00 -6.05531484e-02 8.13388944e-01 6.71843469e-01
3.53389055e-01 -1.02478850e+00 2.24686339e-01 4.35314089e-01
1.07598782e+00 -1.26860332e+00 -1.58807412e-01 1.89181529e-02
-9.45973217e-01 1.33135116e+00 -2.40762383e-02 -5.89036234e-02
3.66562426e-01 -4.22248542e-01 -4.82039511e-01 -3.85354340e-01
-8.96448612e-01 -5.97234368e-01 3.25428210e-02 4.57876682e-01
7.44172856e-02 -5.22755794e-02 -1.56017572e-01 5.40860355e-01
-9.12409276e-02 1.36703715e-01 8.43844831e-01 5.12131393e-01
-3.64991367e-01 -5.68093896e-01 -6.31630123e-01 -1.49864048e-01
-2.36957371e-01 1.08459026e-01 -4.05511737e-01 5.22604942e-01
2.49554843e-01 9.62269008e-01 3.05652767e-01 -3.38388622e-01
2.78685212e-01 2.19182059e-01 5.34916818e-01 -2.29472995e-01
-2.94549704e-01 -9.80880298e-03 2.46956572e-01 -6.02988243e-01
-9.17583466e-01 -9.90979552e-01 -7.99150944e-01 -3.60132396e-01
3.59046161e-01 8.31041113e-02 7.64975905e-01 1.07015026e+00
1.75898105e-01 6.89109921e-01 6.25521421e-01 -1.10728872e+00
2.09497184e-01 -7.87861049e-01 -7.07129657e-01 5.06955981e-01
5.55204988e-01 -6.15635693e-01 -3.52173507e-01 7.43116915e-01] | [10.28252124786377, 0.6605015993118286] |
0ac07271-39a4-484c-a991-6cb0e3219554 | multilingual-sense-intersection-in-a-parallel | null | null | https://aclanthology.org/2016.gwc-1.8 | https://aclanthology.org/2016.gwc-1.8.pdf | Multilingual Sense Intersection in a Parallel Corpus with Diverse Language Families | Supervised methods for Word Sense Disambiguation (WSD) benefit from high-quality sense-annotated resources, which are lacking for many languages less common than English. There are, however, several multilingual parallel corpora that can be inexpensively annotated with senses through cross-lingual methods. We test the effectiveness of such an approach by attempting to disambiguate English texts through their translations in Italian, Romanian and Japanese. Specifically, we try to find the appropriate word senses for the English words by comparison with all the word senses associated to their translations. The main advantage of this approach is in that it can be applied to any parallel corpus, as long as large, high-quality inter-linked sense inventories exist for all the languages considered. | ['Francis Bond', 'Giulia Bonansinga'] | null | null | null | null | gwc-2016-1 | ['word-sense-disambiguation'] | ['natural-language-processing'] | [-7.68276751e-02 -3.07483166e-01 -4.27025050e-01 -1.86818223e-02
-8.23934972e-01 -9.92565930e-01 6.57354653e-01 6.38990045e-01
-1.04200423e+00 1.25274515e+00 3.59342903e-01 -3.29077035e-01
-2.08773836e-01 -7.02086866e-01 8.47411230e-02 -3.59884322e-01
3.11745286e-01 7.72818804e-01 3.35303783e-01 -7.63536096e-01
1.81063503e-01 1.66985780e-01 -1.37771368e+00 -1.15136467e-02
1.17775774e+00 3.00489336e-01 7.70812809e-01 -7.76483640e-02
-7.46784568e-01 4.53028791e-02 -6.17398560e-01 -5.94306946e-01
-9.15681571e-02 -1.14323668e-01 -1.26273584e+00 -1.34493873e-01
-1.17593147e-02 6.69543326e-01 2.57662594e-01 1.40866745e+00
3.72949868e-01 7.26031214e-02 2.33880237e-01 -6.39778137e-01
-4.67345595e-01 9.53687727e-01 -2.62054443e-01 2.10073307e-01
8.23774755e-01 -3.21017057e-01 1.44533372e+00 -8.82452190e-01
1.25266087e+00 1.20002282e+00 4.79662269e-01 3.13742906e-01
-1.20085657e+00 -4.57646996e-01 3.55387665e-02 1.41786903e-01
-1.60299516e+00 -1.47808641e-01 5.67522705e-01 -1.63763449e-01
1.33597469e+00 5.87407500e-02 7.60116756e-01 1.04995739e+00
3.63964811e-02 4.14191186e-01 1.31279850e+00 -1.08407152e+00
8.72189701e-02 2.97081828e-01 2.58413285e-01 2.65866250e-01
7.11850107e-01 -4.20941323e-01 -5.08861423e-01 -2.60045022e-01
2.91627526e-01 -3.62807512e-01 -1.23773649e-01 -1.93983451e-01
-1.58266628e+00 6.65057480e-01 -1.27483651e-01 1.37549675e+00
-3.42946172e-01 -5.89357316e-01 8.08451533e-01 4.10742790e-01
7.14166164e-01 1.08071935e+00 -9.55679417e-01 -1.31642416e-01
-8.09106767e-01 2.68043041e-01 8.51598620e-01 9.53171790e-01
9.01501894e-01 -3.10899705e-01 3.88850033e-01 1.10954630e+00
1.74330056e-01 6.88539207e-01 1.07960200e+00 -4.01219100e-01
3.70456696e-01 8.69003475e-01 2.77486354e-01 -1.01104736e+00
-4.49398309e-01 -1.41987920e-01 -3.50813091e-01 -2.94022858e-01
2.49674678e-01 -7.24655241e-02 -4.77117449e-01 1.72779953e+00
3.68084222e-01 -2.27262139e-01 6.86025679e-01 5.93106151e-01
6.21174812e-01 2.36363932e-01 3.63339990e-01 -4.83528078e-01
1.91769493e+00 -2.55260050e-01 -8.00791740e-01 -6.52580440e-01
7.57794917e-01 -1.27038121e+00 1.30620992e+00 1.97313488e-01
-6.67208552e-01 -3.23810011e-01 -1.06804705e+00 -3.05507868e-03
-9.61575508e-01 -2.03013912e-01 7.30139077e-01 5.96949458e-01
-9.68543708e-01 2.35968888e-01 -5.79648852e-01 -8.57831299e-01
-3.32022160e-01 7.65013620e-02 -6.40418172e-01 -3.67985703e-02
-1.92053425e+00 1.40686834e+00 1.09867918e+00 -4.43773597e-01
5.61746471e-02 -2.71005213e-01 -1.02046311e+00 -2.77428597e-01
4.64547724e-01 -4.51450795e-01 9.72043753e-01 -1.16511679e+00
-8.43159080e-01 1.51337814e+00 -3.00995231e-01 -2.03910992e-01
-9.19786766e-02 -6.17758892e-02 -1.03893244e+00 6.17652461e-02
7.86408365e-01 2.96669573e-01 9.27152112e-02 -1.08516991e+00
-9.62528288e-01 -4.12058383e-01 7.30076209e-02 3.32748324e-01
-4.34548348e-01 3.72358561e-01 -3.35989326e-01 -8.56483757e-01
8.96513760e-02 -8.77363920e-01 -4.25175637e-01 -7.38934696e-01
-1.73217833e-01 -5.31273425e-01 4.34541225e-01 -5.15562713e-01
1.27701521e+00 -2.05904222e+00 -7.77644739e-02 1.84652761e-01
-6.82202652e-02 3.96494001e-01 -1.63061067e-01 7.60608852e-01
-5.56778461e-02 3.27881128e-01 -1.02559507e-01 -4.43326123e-02
7.55095705e-02 8.50185931e-01 -1.78001150e-01 7.56244138e-02
9.59639251e-02 6.53890967e-01 -1.50157964e+00 -7.46489167e-01
2.06770286e-01 2.02242151e-01 -7.81390071e-02 -2.46724784e-01
-1.12680919e-01 1.05034336e-01 -7.08956778e-01 3.89246255e-01
1.69623613e-01 7.43113682e-02 8.98858845e-01 3.67647794e-04
-4.32923585e-01 1.00157595e+00 -1.23765075e+00 1.86522663e+00
-6.81762397e-01 4.08256978e-01 -2.54918635e-01 -8.11154246e-01
9.20086980e-01 6.44388139e-01 5.32554269e-01 -5.95234692e-01
-1.20708145e-01 8.68974924e-01 -1.61755323e-01 -2.81918794e-01
9.90380645e-01 -5.98497629e-01 -5.21107912e-01 3.01218241e-01
3.50542098e-01 -3.25747162e-01 8.99326801e-01 9.31067690e-02
6.58989549e-01 -8.47942308e-02 1.19114244e+00 -8.37090313e-01
6.36040986e-01 5.28545916e-01 7.70680130e-01 3.03852379e-01
1.77521601e-01 6.16301969e-02 1.10612184e-01 -2.64882088e-01
-9.46650684e-01 -9.69233096e-01 -3.22925240e-01 9.24028754e-01
2.48650998e-01 -8.24359417e-01 -6.01095259e-01 -5.27482986e-01
-3.76743436e-01 8.11713576e-01 -1.55037209e-01 4.48025525e-01
-4.70902294e-01 -6.26552284e-01 5.92893362e-01 9.96858105e-02
1.77245989e-01 -1.18112242e+00 -4.36374724e-01 5.86136162e-01
-5.14258802e-01 -1.28748202e+00 -2.14070022e-01 2.30595708e-01
-5.87755084e-01 -1.19747317e+00 -3.01738799e-01 -9.07689512e-01
1.66308448e-01 2.28032067e-01 1.56564832e+00 7.13450909e-02
-4.05684561e-02 1.79429412e-01 -6.63650811e-01 -4.93021190e-01
-5.27246535e-01 4.11349088e-01 4.35106754e-01 -5.84713221e-01
9.95682180e-01 -4.15490985e-01 1.77439809e-01 2.56273717e-01
-1.12238038e+00 -4.65704232e-01 2.25756586e-01 7.69761860e-01
7.78153896e-01 4.08507846e-02 5.96988261e-01 -1.11589503e+00
7.81447351e-01 -3.66164565e-01 -4.83443648e-01 5.03498137e-01
-8.97075117e-01 2.55193591e-01 3.91560555e-01 -2.99018502e-01
-1.03297913e+00 -1.02025613e-01 -4.90058184e-01 4.65346426e-01
-3.15661311e-01 8.86666894e-01 -3.48465234e-01 1.51425347e-01
7.58092463e-01 6.06762953e-02 -6.11720681e-01 -4.93161052e-01
6.00868404e-01 6.86100662e-01 4.73903507e-01 -8.31692636e-01
6.38117015e-01 7.16156662e-02 -3.47249955e-01 -1.02935481e+00
-9.54833806e-01 -9.74216640e-01 -9.21043396e-01 1.22383259e-01
9.34273481e-01 -8.39655280e-01 -1.16791852e-01 6.92166016e-03
-1.20896018e+00 3.07696164e-01 -4.00800377e-01 7.82602608e-01
-2.34236151e-01 6.33665383e-01 -2.85254121e-01 -4.69974786e-01
-3.73809248e-01 -9.20864284e-01 8.82415593e-01 9.14929956e-02
-9.23272491e-01 -1.51478827e+00 3.68759871e-01 -1.95201691e-02
1.67965576e-01 6.66080713e-02 1.12133884e+00 -1.14156008e+00
2.75625139e-01 -1.48213813e-02 1.83046639e-01 3.31667550e-02
5.30677378e-01 -1.58305928e-01 -4.80970055e-01 -1.86115056e-01
-1.01818018e-01 -3.61510813e-01 4.29140925e-01 -1.57821357e-01
-1.88718550e-02 -1.75004780e-01 -5.27512670e-01 -9.59308892e-02
1.73870647e+00 1.96621373e-01 3.45059901e-01 7.78302014e-01
4.80033726e-01 5.96689463e-01 8.79430532e-01 4.46024500e-02
4.31983352e-01 6.86387002e-01 -2.88671583e-01 -1.24242730e-01
1.44345343e-01 -1.36349231e-01 1.87869415e-01 1.34348917e+00
7.08241910e-02 -1.48692042e-01 -1.24885082e+00 1.07264853e+00
-1.58794665e+00 -6.54787421e-01 -1.45786241e-01 2.10678363e+00
1.30927622e+00 1.87470049e-01 -2.33136257e-03 1.24829836e-01
6.56705856e-01 1.74772978e-01 5.26922494e-02 -2.41982728e-01
-6.14923835e-01 6.34548008e-01 4.49105084e-01 5.97859561e-01
-9.18620169e-01 1.43022501e+00 6.75894451e+00 1.05481565e+00
-7.63164818e-01 1.89407304e-01 -7.50598609e-02 3.95834565e-01
-6.59526348e-01 2.58398414e-01 -8.37274730e-01 2.32295379e-01
6.78569436e-01 -4.45182145e-01 2.51346737e-01 7.60640204e-01
-3.31716525e-04 -1.02332316e-01 -7.12570012e-01 8.82169187e-01
-7.44283050e-02 -1.10241389e+00 8.34076107e-02 -3.32725167e-01
8.11114967e-01 2.66130358e-01 -6.37478530e-01 -6.88837022e-02
6.50406718e-01 -4.83070970e-01 6.74582839e-01 1.37656834e-02
8.84490550e-01 -8.81538272e-01 9.87942278e-01 9.24548730e-02
-1.35412228e+00 5.89173615e-01 -5.41346788e-01 2.23336369e-03
1.91811696e-01 7.90100217e-01 -6.84129596e-01 9.65711057e-01
5.40933788e-01 6.86838984e-01 -4.10466194e-01 5.33414602e-01
-7.13865280e-01 5.40674448e-01 -2.07147747e-01 -2.84950733e-01
3.55517894e-01 -2.40762025e-01 7.83512652e-01 1.56334031e+00
4.90226537e-01 -3.64538915e-02 5.57687283e-01 2.99697727e-01
1.97104126e-01 7.59438276e-01 -7.20753968e-01 -3.55712295e-01
7.71846592e-01 1.11498463e+00 -9.95037019e-01 -5.00282824e-01
-5.46930254e-01 8.03071260e-01 3.19804132e-01 1.47470027e-01
9.12755951e-02 -5.59893250e-01 1.07781148e+00 -1.62496865e-01
-8.87586325e-02 -4.44273680e-01 -1.97962672e-01 -1.18597031e+00
4.96918187e-02 -1.01807225e+00 7.43448555e-01 -6.61450326e-01
-1.79075825e+00 9.59494650e-01 7.25861266e-02 -1.07309628e+00
-6.64419711e-01 -8.99426699e-01 -5.59010170e-02 1.04861796e+00
-1.45787561e+00 -1.05814886e+00 4.28555608e-01 6.12888813e-01
4.57853764e-01 -1.60441175e-01 1.37263060e+00 2.26719514e-01
1.42276369e-03 2.25558266e-01 5.05157523e-02 1.40591070e-01
9.21465278e-01 -1.47221160e+00 5.41122615e-01 8.59823287e-01
6.63024247e-01 9.89888489e-01 9.26470935e-01 -8.25269997e-01
-8.24276388e-01 -6.28613293e-01 1.94727135e+00 -3.32424223e-01
1.33541524e+00 -1.63269848e-01 -8.83329690e-01 4.77250069e-01
5.79379499e-01 -4.34258431e-01 8.97683620e-01 6.58928752e-01
-3.70066792e-01 1.52347639e-01 -9.23579812e-01 6.14082754e-01
1.05625427e+00 -7.76102066e-01 -1.31663179e+00 3.53873760e-01
7.64988482e-01 -2.42887601e-01 -1.12374723e+00 -1.48984984e-01
2.43802816e-01 -4.80362952e-01 9.74537134e-01 -5.91577411e-01
7.94201531e-03 -5.11927605e-01 -2.93760031e-01 -1.72878993e+00
-5.33117987e-02 -5.79124272e-01 9.17516887e-01 1.53296292e+00
7.41161466e-01 -9.15053904e-01 3.32224593e-02 1.29853249e-01
-3.18686701e-02 -1.31908894e-01 -1.03418601e+00 -9.55112219e-01
1.03312209e-01 -5.93493640e-01 8.31659973e-01 1.52891219e+00
6.36061311e-01 7.32985377e-01 -4.70243692e-02 -8.93658772e-02
1.53645843e-01 2.31164545e-01 1.18941829e-01 -1.47788894e+00
-8.53844360e-02 -3.62683862e-01 -4.90924060e-01 -5.64307988e-01
6.21161342e-01 -9.80326116e-01 1.75746456e-01 -1.46849203e+00
5.97822182e-02 -7.26457834e-01 -3.95884782e-01 6.99594855e-01
-6.13229036e-01 2.05975652e-01 -9.89642367e-03 1.46760285e-01
-5.62658072e-01 5.74342348e-02 8.70207310e-01 3.65716219e-02
-1.76687241e-01 -4.37040210e-01 -7.80743122e-01 7.84770310e-01
6.79323196e-01 -7.88780093e-01 -1.04694694e-01 -4.56476301e-01
7.13780046e-01 -2.42757156e-01 -2.59966910e-01 -5.59349358e-01
-1.44759910e-02 -5.25776327e-01 -8.89404491e-02 -1.78364366e-01
-2.41448283e-01 -6.79232657e-01 1.58677697e-01 2.72799432e-01
-9.26490799e-02 6.30584121e-01 2.16735691e-01 1.44472733e-01
-6.27523005e-01 -6.50080144e-01 6.21251941e-01 -5.47439277e-01
-1.30063593e+00 -2.06114247e-01 -6.37488782e-01 4.82974142e-01
7.25858986e-01 -7.86061659e-02 3.34264487e-02 1.75657451e-01
-3.25539619e-01 -3.26972865e-02 8.76534402e-01 6.13041699e-01
1.00722440e-01 -1.41661274e+00 -6.76575422e-01 -1.35762483e-01
6.54696345e-01 -4.30747479e-01 -2.54653901e-01 3.08395833e-01
-3.21846932e-01 5.68323314e-01 -3.18514884e-01 -9.99292657e-02
-1.11769712e+00 8.19153786e-01 -1.13451459e-01 -5.88523149e-01
-3.45537871e-01 2.15161189e-01 -3.73642564e-01 -6.39295220e-01
-4.72061366e-01 -1.07881710e-01 -5.09213269e-01 2.30874375e-01
5.10395706e-01 -1.68565400e-02 2.51277924e-01 -1.05261695e+00
-7.16732562e-01 4.65791732e-01 2.82162100e-01 -4.24112201e-01
1.15565956e+00 -4.73130226e-01 -5.60809731e-01 6.17724180e-01
7.48696506e-01 6.52337492e-01 -7.12437332e-02 -4.15138781e-01
8.70624900e-01 -3.24435055e-01 -2.44012937e-01 -6.38398230e-01
-5.24673998e-01 3.14530343e-01 3.14047247e-01 3.55300456e-01
1.08777821e+00 1.89842790e-01 7.40479708e-01 5.20001054e-01
7.89134920e-01 -1.26155090e+00 -5.54292202e-01 9.18003976e-01
3.94856155e-01 -1.01683879e+00 -2.73383379e-01 -5.29858291e-01
-6.23758793e-01 9.64928985e-01 -1.73435993e-02 9.27061364e-02
4.29611564e-01 2.55279183e-01 5.09954393e-01 -2.81142771e-01
-4.70042825e-01 -7.37056375e-01 4.11212772e-01 7.48375356e-01
9.59174812e-01 3.04345518e-01 -1.39102864e+00 5.14466166e-01
-4.17419791e-01 -3.74022216e-01 3.83802861e-01 8.08497608e-01
-4.36510175e-01 -1.90145350e+00 -3.21332544e-01 7.23178387e-02
-7.73787439e-01 -6.23987436e-01 -5.78160703e-01 1.04027522e+00
2.97821075e-01 1.24680948e+00 -1.86320007e-01 -1.11506656e-01
2.30141938e-01 2.08430320e-01 2.60510594e-01 -9.07233000e-01
-5.44362247e-01 8.93203765e-02 5.24219096e-01 -2.84667403e-01
-9.22222912e-01 -7.85092831e-01 -1.23131406e+00 2.59021204e-02
-3.28166336e-01 7.55146801e-01 4.57230538e-01 1.52312660e+00
-7.62090310e-02 -2.61929408e-02 2.76772499e-01 -4.20930952e-01
-1.29148513e-02 -9.45275486e-01 -7.62859285e-01 5.88395178e-01
-2.14742690e-01 -7.32015967e-01 9.63269770e-02 6.59034997e-02] | [10.363567352294922, 9.383463859558105] |
27f59828-0d97-490e-9de6-420a4ff5d430 | multi-output-gaussian-processes-for | 1812.08739 | null | https://arxiv.org/abs/1812.08739v2 | https://arxiv.org/pdf/1812.08739v2.pdf | Multi-Output Gaussian Processes for Crowdsourced Traffic Data Imputation | Traffic speed data imputation is a fundamental challenge for data-driven transport analysis. In recent years, with the ubiquity of GPS-enabled devices and the widespread use of crowdsourcing alternatives for the collection of traffic data, transportation professionals increasingly look to such user-generated data for many analysis, planning, and decision support applications. However, due to the mechanics of the data collection process, crowdsourced traffic data such as probe-vehicle data is highly prone to missing observations, making accurate imputation crucial for the success of any application that makes use of that type of data. In this article, we propose the use of multi-output Gaussian processes (GPs) to model the complex spatial and temporal patterns in crowdsourced traffic data. While the Bayesian nonparametric formalism of GPs allows us to model observation uncertainty, the multi-output extension based on convolution processes effectively enables us to capture complex spatial dependencies between nearby road segments. Using 6 months of crowdsourced traffic speed data or "probe vehicle data" for several locations in Copenhagen, the proposed approach is empirically shown to significantly outperform popular state-of-the-art imputation methods. | ['Filipe Rodrigues', 'Francisco C. Pereira', 'Kristian Henrickson'] | 2018-12-20 | null | null | null | null | ['traffic-data-imputation'] | ['time-series'] | [-2.16228604e-01 -1.44843593e-01 -1.16635062e-01 -5.76273024e-01
-1.02097714e+00 -3.04533213e-01 6.95962548e-01 5.08484066e-01
-6.64817870e-01 8.88366580e-01 4.62425649e-01 -4.79856014e-01
-3.03551584e-01 -1.10128391e+00 -8.37327242e-01 -6.29266202e-01
2.42509827e-01 7.65104532e-01 2.01321319e-01 -2.54784733e-01
4.03839312e-02 6.23976290e-01 -1.53564131e+00 -1.39717057e-01
9.94956493e-01 6.52590275e-01 1.68717355e-01 4.93167579e-01
-4.96957481e-01 5.40878832e-01 -1.76498547e-01 -5.76237619e-01
1.74058184e-01 2.34309569e-01 -1.26025423e-01 -1.03734344e-01
-2.40310151e-02 -2.32730195e-01 -4.46969002e-01 5.36147118e-01
4.00218785e-01 4.61399019e-01 6.23836935e-01 -1.52245390e+00
-4.80790198e-01 9.31638330e-02 -3.98582518e-01 9.09231901e-02
1.90092713e-01 3.92381787e-01 3.14437360e-01 -5.79276562e-01
2.09250793e-01 1.12555718e+00 8.26182246e-01 5.94497360e-02
-1.46133852e+00 -4.97735560e-01 5.44937775e-02 -1.04419790e-01
-1.45633709e+00 -6.27931416e-01 4.38493907e-01 -8.63036036e-01
5.87948978e-01 8.43095034e-02 3.04085761e-01 1.31576192e+00
2.31685966e-01 4.90208387e-01 9.84552443e-01 4.50483561e-02
5.12035489e-01 2.93572158e-01 1.34548366e-01 6.21512718e-03
1.63991004e-01 1.70248330e-01 -4.86235887e-01 -4.62302178e-01
4.54328328e-01 4.24686790e-01 4.55884695e-01 -1.34957343e-01
-8.62706959e-01 7.26906300e-01 9.25134644e-02 -1.94751978e-01
-7.48789787e-01 3.74912530e-01 3.81166562e-02 -2.41336524e-01
8.73698413e-01 -4.27839994e-01 -2.01490760e-01 -5.81636965e-01
-9.85485852e-01 8.58010054e-01 5.92521489e-01 1.02239811e+00
8.43644202e-01 -6.78047240e-02 -4.70591456e-01 7.34439731e-01
3.14877063e-01 1.11575055e+00 -1.37861803e-01 -1.10984254e+00
9.45713997e-01 5.74756742e-01 7.60349989e-01 -9.16001320e-01
-3.20594639e-01 9.47820861e-03 -8.98398817e-01 -7.13184401e-02
9.16354299e-01 -3.91956091e-01 -7.61539817e-01 1.45181692e+00
4.35006887e-01 1.80689692e-01 -2.77158052e-01 8.88628781e-01
2.94077456e-01 3.68732005e-01 6.94910765e-01 3.12237352e-01
1.09921253e+00 -1.04001597e-01 -6.89136624e-01 1.10680632e-01
4.36079949e-01 -3.53865832e-01 5.86809814e-01 -1.07583605e-01
-8.41250598e-01 -5.47018409e-01 -1.91428810e-01 -1.15118615e-01
-6.17501199e-01 -3.12853813e-01 5.98285794e-01 8.58930349e-01
-6.63425326e-01 2.61404693e-01 -8.84105623e-01 -3.18997949e-01
7.45139360e-01 2.34032422e-01 -4.79838908e-01 -3.38464856e-01
-1.04546940e+00 8.17044914e-01 -3.14389646e-01 4.49104846e-01
-5.60683548e-01 -9.12266552e-01 -7.40317762e-01 4.20574285e-03
2.43811071e-01 -5.12263298e-01 1.12044489e+00 -2.86196321e-01
-1.01300156e+00 2.58694261e-01 -8.19566846e-01 -4.79336619e-01
9.08343494e-01 8.44158605e-02 -5.94285548e-01 -3.57059151e-01
4.54988539e-01 3.52846444e-01 6.14742339e-01 -9.95890915e-01
-7.86010087e-01 -4.33098644e-01 -2.89605349e-01 -1.98798925e-01
4.66996849e-01 4.76703569e-02 -1.43364027e-01 -3.49807292e-01
-6.23760968e-02 -9.72987056e-01 -5.46825469e-01 -2.63740867e-01
-3.39063674e-01 -1.27766892e-01 4.98670727e-01 -6.87020063e-01
8.38886023e-01 -2.09057212e+00 -3.58660281e-01 4.73168910e-01
1.03861280e-01 -3.54398303e-02 2.54907180e-02 7.16464281e-01
4.59508508e-01 9.45618078e-02 -3.16548914e-01 -8.13555241e-01
3.26565623e-01 4.94772702e-01 -3.59129518e-01 5.50800264e-01
2.89533913e-01 1.09190691e+00 -9.83782589e-01 -2.90491164e-01
6.50665820e-01 5.70239902e-01 -1.68911755e-01 -2.44869143e-01
-7.04017505e-02 9.01883960e-01 -5.99505424e-01 5.13079762e-01
8.02598238e-01 3.07547003e-01 -3.15551072e-01 4.72177476e-01
-4.22794282e-01 1.87379241e-01 -1.16660082e+00 1.25938213e+00
-5.03207982e-01 6.65484011e-01 -1.88290086e-02 -4.76813108e-01
8.10154736e-01 3.35720837e-01 4.94719118e-01 -9.87391651e-01
-6.39433563e-02 1.50847845e-02 -3.62894386e-01 -6.60931885e-01
9.05352771e-01 -1.69946969e-01 -2.73703516e-01 3.18904608e-01
-3.43614757e-01 1.17637232e-01 8.63524228e-02 -2.48402171e-02
1.05808616e+00 2.31760368e-01 -5.15555561e-01 5.65941781e-02
8.69115666e-02 2.19453827e-01 7.10405052e-01 8.26503813e-01
-1.72440365e-01 6.54301584e-01 4.20640320e-01 -5.89102566e-01
-1.07688522e+00 -1.28030217e+00 -2.21209258e-01 8.35918605e-01
-1.73464671e-01 2.18413249e-01 -5.46994328e-01 -1.72009870e-01
6.38621628e-01 8.97815943e-01 -5.67135692e-01 2.83914685e-01
-3.00237298e-01 -9.81815219e-01 5.23620248e-01 5.47415078e-01
2.98931539e-01 -6.98839843e-01 -1.64788991e-01 5.92714369e-01
-4.09311652e-01 -1.12490475e+00 -4.09944542e-02 -3.33340049e-01
-7.74685860e-01 -8.64099145e-01 -5.40980995e-01 1.26638502e-01
4.87257689e-01 3.82511616e-01 7.55558491e-01 -2.78681159e-01
1.06247939e-01 3.94626319e-01 -1.80727243e-01 -7.31405735e-01
-2.86859035e-01 9.18845460e-02 5.37172556e-02 5.00393093e-01
9.33283329e-01 -6.97153032e-01 -3.71758103e-01 2.41140336e-01
-7.66407907e-01 -6.13673925e-02 1.80216894e-01 8.85318667e-02
5.37662804e-01 -2.67294403e-02 8.21538210e-01 -5.35558701e-01
6.38538718e-01 -1.18066180e+00 -6.17542684e-01 -1.40112594e-01
-1.65975153e-01 -1.46131054e-01 5.61763383e-02 -2.20636129e-01
-1.08850574e+00 1.10975001e-02 -5.90177402e-02 -8.85604247e-02
-7.55784094e-01 3.90541285e-01 -1.97954863e-01 2.89462447e-01
4.08675492e-01 -3.50235701e-02 1.45563409e-01 -6.17636204e-01
2.37484649e-01 9.25211906e-01 3.74092788e-01 -5.87554455e-01
6.17940843e-01 7.87341833e-01 2.27475628e-01 -1.16854179e+00
-1.75888568e-01 -5.81806779e-01 -5.53208888e-01 -2.87192702e-01
9.75461125e-01 -1.10823882e+00 -9.80787992e-01 4.54438627e-01
-1.08739614e+00 -3.43755543e-01 -3.83161157e-01 6.00715518e-01
-4.59482849e-01 -4.46240008e-02 -1.71306178e-01 -1.46234024e+00
3.99958253e-01 -1.12007368e+00 1.05974627e+00 3.01391575e-02
-2.59503126e-01 -9.75603044e-01 6.02909923e-02 5.61889231e-01
6.03999496e-01 4.59927529e-01 5.60237706e-01 -4.38912183e-01
-8.72424126e-01 -7.87768722e-01 -2.50909209e-01 1.06458686e-01
6.50351420e-02 -1.06614068e-01 -1.16343677e+00 5.02930760e-01
-2.59016335e-01 4.97939199e-01 3.79879832e-01 5.78997910e-01
1.00055969e+00 -7.06495270e-02 -2.19385251e-01 -3.99097917e-05
1.10309172e+00 -2.52796561e-01 8.07245374e-01 6.51033968e-02
6.97450221e-01 9.97534394e-01 3.78497303e-01 4.19708550e-01
1.22324562e+00 7.59095967e-01 3.08143020e-01 8.43449980e-02
3.35898310e-01 -4.59095418e-01 -1.90118507e-01 1.47774279e-01
-2.37664923e-01 -2.74647713e-01 -1.14739156e+00 9.87838924e-01
-2.27197266e+00 -1.24976182e+00 -1.04917812e+00 2.56314826e+00
6.85104579e-02 -4.48484421e-02 3.63050282e-01 -9.68498886e-02
6.99611187e-01 -1.54924661e-01 -1.10053927e-01 -3.60562116e-01
-1.07860163e-01 -6.06802441e-02 1.08728409e+00 4.88655657e-01
-7.28582442e-01 3.52754951e-01 5.63339615e+00 7.41431713e-01
-4.35945660e-01 4.27715987e-01 5.24759948e-01 -2.11787596e-01
-4.25619870e-01 -1.19248837e-01 -8.10018599e-01 1.05580699e+00
1.36557066e+00 1.41882926e-01 3.65247279e-01 2.75987059e-01
1.26142263e+00 -7.42882907e-01 -8.46814573e-01 6.71479344e-01
-7.19463229e-01 -1.25061762e+00 -4.19632435e-01 4.93842423e-01
6.95279896e-01 3.50737870e-01 -1.05359010e-01 2.19766736e-01
3.88652861e-01 -1.00078797e+00 7.99432397e-01 1.14631855e+00
3.04935485e-01 -8.39309990e-01 8.80507469e-01 6.84019268e-01
-8.26961637e-01 8.31637718e-03 -1.48797497e-01 -3.66678864e-01
6.83116674e-01 9.44526076e-01 -4.11463946e-01 3.80729079e-01
4.87112015e-01 2.00755715e-01 -3.10249746e-01 1.28494990e+00
-6.15060853e-04 7.71214187e-01 -6.33111179e-01 6.78801462e-02
1.99045509e-01 -4.65063423e-01 4.56127942e-01 1.03272867e+00
4.07475501e-01 -4.82739061e-02 -9.98324826e-02 9.07639086e-01
7.65720457e-02 -3.69863808e-01 -7.61698186e-01 2.13111877e-01
5.89415014e-01 8.66440415e-01 -2.92661428e-01 -7.78266694e-03
-5.15776753e-01 3.98316115e-01 6.26129508e-02 7.87984371e-01
-8.36290002e-01 6.97500780e-02 1.10731530e+00 7.42931426e-01
1.47257715e-01 -7.52548516e-01 -4.25056517e-01 -4.95711356e-01
1.66494146e-01 -2.32651427e-01 -4.06632155e-01 -7.41225898e-01
-1.37621355e+00 -1.19165353e-01 1.96751714e-01 -1.01424909e+00
-2.58880168e-01 -1.72788948e-01 -5.61749279e-01 1.52491581e+00
-1.37589335e+00 -1.15129030e+00 -1.91495419e-01 3.75602722e-01
2.13642046e-01 1.85183257e-01 6.46561027e-01 3.86245817e-01
-3.54360253e-01 -4.75720204e-02 6.26276493e-01 3.23151872e-02
3.21854383e-01 -8.55267704e-01 1.02131689e+00 5.37488997e-01
-9.17147845e-02 4.24686223e-01 7.00993896e-01 -9.29799557e-01
-1.39192629e+00 -1.33670020e+00 1.31159496e+00 -1.08616698e+00
6.54275298e-01 -5.30137002e-01 -8.73728454e-01 5.98534524e-01
-5.58034062e-01 8.95287469e-02 7.63275325e-01 1.66248545e-01
9.10982862e-02 -2.53494501e-01 -1.45501363e+00 4.69521523e-01
7.29398608e-01 -6.24866426e-01 -2.88516194e-01 1.99204177e-01
4.13841724e-01 -2.11115137e-01 -6.50276124e-01 -1.04206234e-01
4.98329639e-01 -8.73126686e-01 7.55089164e-01 -4.92960095e-01
5.34449518e-02 -4.78131205e-01 -1.94511071e-01 -1.15536392e+00
-1.57705635e-01 -4.88332719e-01 7.44784698e-02 1.39347255e+00
4.58787620e-01 -8.30157161e-01 7.83657491e-01 1.65056455e+00
1.58120617e-01 -1.30277604e-01 -1.40346503e+00 -5.20240963e-01
1.11932531e-02 -1.29557168e+00 1.08965957e+00 4.04960722e-01
-4.04649258e-01 -3.65997314e-01 -2.72138149e-01 4.78140354e-01
8.73005033e-01 -3.50693107e-01 1.08505762e+00 -1.45517159e+00
1.78304926e-01 -1.33794155e-02 -5.03392518e-01 -7.35118628e-01
1.61161005e-01 -3.67016822e-01 1.70732081e-01 -1.57006800e+00
-1.83329999e-01 -1.11180258e+00 1.70702890e-01 2.66610365e-02
3.49848308e-02 2.55550206e-01 -1.85303241e-02 -1.04647111e-02
-3.02605480e-01 5.29970944e-01 6.43009961e-01 7.91495070e-02
-3.32134753e-01 5.27076602e-01 -4.45484221e-01 4.12591457e-01
8.23113561e-01 -7.89538920e-01 -1.44507438e-01 -6.95042133e-01
3.75213951e-01 2.26848096e-01 9.72759545e-01 -8.59167993e-01
3.39721650e-01 -3.81492943e-01 3.18375319e-01 -7.01468825e-01
4.76128936e-01 -1.05093706e+00 6.19113982e-01 -2.43551001e-01
1.47974547e-02 5.43800630e-02 2.56276906e-01 1.06875312e+00
6.51858374e-02 5.70847839e-02 -1.18854076e-01 1.29686207e-01
-3.13992232e-01 4.07148927e-01 -9.65667069e-01 -6.01179153e-02
8.09253216e-01 -5.08917391e-01 -8.13866332e-02 -4.64822680e-01
-6.51076972e-01 4.46747333e-01 5.49367189e-01 4.60167080e-01
2.34100491e-01 -1.23266518e+00 -7.17342734e-01 3.13175231e-01
2.02944167e-02 1.88541248e-01 6.56591356e-01 1.06981337e+00
-1.39197372e-02 2.79059768e-01 1.53373376e-01 -5.03955483e-01
-4.47872519e-01 1.64100543e-01 2.87329475e-03 3.28943163e-01
-4.49800968e-01 1.47201136e-01 -3.27986807e-01 -5.00527382e-01
-1.16594620e-01 -4.39475417e-01 2.86022991e-01 3.92146483e-02
4.84760582e-01 1.21631432e+00 2.39109650e-01 -9.37107742e-01
-8.24208185e-02 5.73498718e-02 6.18423760e-01 -3.60196888e-01
1.19777286e+00 -5.42267442e-01 2.15532765e-01 7.36227632e-01
6.21167183e-01 7.15252012e-02 -1.55581939e+00 -6.68335706e-02
9.12504792e-02 -5.18048286e-01 -6.10648394e-02 -5.14947057e-01
-5.26411295e-01 9.06581819e-01 4.14606214e-01 3.29727024e-01
3.07157040e-01 -3.90320688e-01 6.82341039e-01 3.85934263e-02
6.88227415e-01 -9.96937275e-01 -1.02607954e+00 3.23003381e-01
4.19529527e-01 -1.51491785e+00 -3.64482731e-01 -2.04289988e-01
-5.40075004e-01 4.84807312e-01 -1.38401642e-01 -2.50000358e-02
7.06864417e-01 1.87428489e-01 -1.99102178e-01 6.34152591e-02
-3.12247902e-01 -4.15162921e-01 -5.02232881e-03 1.12579870e+00
-5.74839413e-02 3.53159159e-01 1.55081213e-01 6.80820167e-01
-4.68926691e-03 3.85450393e-01 5.53195894e-01 7.63144314e-01
-1.89087883e-01 -1.02078712e+00 -5.91103613e-01 6.45538330e-01
-3.29308361e-01 4.92232852e-02 2.13129625e-01 8.48678887e-01
2.76705235e-01 1.39606249e+00 4.66467440e-01 2.03215793e-01
6.41643822e-01 2.46413574e-01 8.34501982e-02 -3.36278111e-01
-4.09542114e-01 -6.51300073e-01 2.48779848e-01 -5.12714982e-01
-2.84755051e-01 -1.01207948e+00 -9.63148415e-01 -9.99882102e-01
1.77760497e-01 8.79996493e-02 1.28005338e+00 1.28384328e+00
6.21022999e-01 1.41282335e-01 3.55970532e-01 -1.06873846e+00
-2.24558875e-01 -8.72912884e-01 -6.74223781e-01 3.28371733e-01
6.14128053e-01 -8.45424533e-01 -1.39706254e-01 -6.67855889e-02] | [6.649166584014893, 2.0679574012756348] |
f0ae0cc4-7737-42d7-8ba0-938ef33d6c56 | siamese-network-for-rgb-d-salient-object | 2008.12134 | null | https://arxiv.org/abs/2008.12134v2 | https://arxiv.org/pdf/2008.12134v2.pdf | Siamese Network for RGB-D Salient Object Detection and Beyond | Existing RGB-D salient object detection (SOD) models usually treat RGB and depth as independent information and design separate networks for feature extraction from each. Such schemes can easily be constrained by a limited amount of training data or over-reliance on an elaborately designed training process. Inspired by the observation that RGB and depth modalities actually present certain commonality in distinguishing salient objects, a novel joint learning and densely cooperative fusion (JL-DCF) architecture is designed to learn from both RGB and depth inputs through a shared network backbone, known as the Siamese architecture. In this paper, we propose two effective components: joint learning (JL), and densely cooperative fusion (DCF). The JL module provides robust saliency feature learning by exploiting cross-modal commonality via a Siamese network, while the DCF module is introduced for complementary feature discovery. Comprehensive experiments using five popular metrics show that the designed framework yields a robust RGB-D saliency detector with good generalization. As a result, JL-DCF significantly advances the state-of-the-art models by an average of ~2.0% (max F-measure) across seven challenging datasets. In addition, we show that JL-DCF is readily applicable to other related multi-modal detection tasks, including RGB-T (thermal infrared) SOD and video SOD, achieving comparable or even better performance against state-of-the-art methods. We also link JL-DCF to the RGB-D semantic segmentation field, showing its capability of outperforming several semantic segmentation models on the task of RGB-D SOD. These facts further confirm that the proposed framework could offer a potential solution for various applications and provide more insight into the cross-modal complementarity task. | ['Qijun Zhao', 'Ge-Peng Ji', 'Deng-Ping Fan', 'Ce Zhu', 'Keren Fu', 'Jianbing Shen'] | 2020-08-26 | null | null | null | null | ['rgb-d-salient-object-detection', 'salient-object-detection'] | ['computer-vision', 'computer-vision'] | [ 3.21793109e-01 3.86989140e-03 -4.89256442e-01 -2.34770522e-01
-9.98695731e-01 -4.75989163e-01 5.48338234e-01 9.94884670e-02
-5.32498777e-01 2.63552457e-01 -2.46119639e-03 8.43914151e-02
-1.30830705e-01 -5.56912422e-01 -6.82743192e-01 -8.96888554e-01
2.17529118e-01 -1.97675079e-02 7.91789532e-01 -3.74492556e-01
2.70053923e-01 5.16602635e-01 -1.90067101e+00 1.35880962e-01
8.61127615e-01 1.46597779e+00 6.09137297e-01 1.72961012e-01
-7.28839934e-02 5.07778049e-01 -2.78672278e-01 -2.13878483e-01
3.44492584e-01 -2.27914259e-01 -7.31920302e-01 -1.69757269e-02
3.30440283e-01 -2.25527272e-01 -1.72150850e-01 1.03654301e+00
5.08349538e-01 9.42589715e-02 5.45420825e-01 -1.38699818e+00
-5.39262116e-01 2.89881974e-01 -8.72251809e-01 4.02011037e-01
4.22752798e-01 6.29010275e-02 1.16644883e+00 -9.74637151e-01
4.97865826e-01 9.56097186e-01 6.09084964e-01 5.16210377e-01
-1.06108940e+00 -4.40751702e-01 1.67942628e-01 1.48085311e-01
-1.21874249e+00 -1.09998830e-01 1.26840150e+00 -1.76277742e-01
7.85421431e-01 2.87425727e-01 6.46210492e-01 9.55709755e-01
-1.12323731e-01 1.41715455e+00 1.20315623e+00 -2.54842013e-01
2.16513008e-01 9.99824330e-02 -1.17619075e-01 8.44141901e-01
1.80293173e-01 2.88200472e-03 -9.48187470e-01 9.39161107e-02
7.78109670e-01 2.82045782e-01 -3.23132873e-01 -8.13138306e-01
-1.23451042e+00 6.68196261e-01 1.08872783e+00 2.55051494e-01
-1.65085584e-01 7.02712610e-02 2.94652462e-01 -1.56809062e-01
3.61103743e-01 3.49090666e-01 -3.67853612e-01 2.75219232e-01
-1.01610267e+00 5.61988614e-02 1.89415663e-01 8.40122879e-01
1.02032113e+00 -9.97900516e-02 -6.10340908e-02 6.05368197e-01
4.61058617e-01 6.61205649e-01 5.48908353e-01 -7.02927411e-01
3.80859196e-01 8.98378551e-01 -2.63879038e-02 -8.70570302e-01
-7.44864047e-01 -6.70371890e-01 -6.27704620e-01 2.55657315e-01
3.86992693e-01 2.30051309e-01 -9.10493553e-01 1.73995221e+00
4.58960682e-01 5.91520639e-03 1.47455394e-01 1.33848941e+00
1.12693703e+00 1.97174713e-01 1.12476043e-01 1.59695894e-01
1.31398582e+00 -9.27791119e-01 -3.49009395e-01 -3.43763232e-01
4.93567288e-01 -6.56318903e-01 1.13978982e+00 1.12984397e-01
-9.78309751e-01 -5.85518241e-01 -1.04888368e+00 -3.31541568e-01
-4.49412584e-01 4.02881235e-01 9.69366670e-01 4.50957447e-01
-9.55994546e-01 3.42038542e-01 -8.66880238e-01 -4.33594704e-01
6.55749083e-01 4.41615462e-01 -4.12450254e-01 2.99441838e-03
-9.80636358e-01 7.57285893e-01 2.87576407e-01 2.00139463e-01
-8.47326279e-01 -6.35737956e-01 -9.04493332e-01 -1.95344254e-01
4.96991962e-01 -7.42801785e-01 9.74379659e-01 -9.27691460e-01
-1.37063313e+00 1.10451174e+00 -7.87963793e-02 -2.46338397e-01
2.89434820e-01 -2.40800440e-01 -1.63656339e-01 5.96526980e-01
3.50739688e-01 8.04852307e-01 9.49629486e-01 -1.41690600e+00
-7.84541428e-01 -5.09439051e-01 1.38574839e-01 3.01250011e-01
-4.56499159e-01 -1.29350409e-01 -5.31995714e-01 -6.65641785e-01
6.20423734e-01 -7.56202281e-01 -8.51684511e-02 2.72936404e-01
-5.46159089e-01 -2.13297606e-01 9.27443862e-01 -2.55118936e-01
9.72748041e-01 -2.08887410e+00 4.84177947e-01 -1.40798953e-03
4.16197687e-01 2.30211809e-01 -5.49220778e-02 7.83197284e-02
1.26574278e-01 -2.61644393e-01 -4.53158766e-01 -5.84330201e-01
-1.83696707e-03 4.17952761e-02 -8.45506936e-02 6.60516858e-01
3.84685934e-01 1.16605628e+00 -1.10225224e+00 -5.17546475e-01
4.27871585e-01 3.43714714e-01 -2.86159813e-01 1.43776238e-01
-9.25691798e-02 2.96235025e-01 -6.44042552e-01 1.20045984e+00
6.13577306e-01 -4.04106200e-01 -2.09323734e-01 -5.69177747e-01
-1.58969223e-01 6.18398637e-02 -9.31641757e-01 2.19798923e+00
-1.48015946e-01 4.40462708e-01 2.40801275e-02 -1.17519045e+00
9.68365669e-01 -8.71076882e-02 5.59941232e-01 -8.98229301e-01
3.16497505e-01 5.42112768e-01 -4.00085002e-01 -4.91620988e-01
4.91020590e-01 -1.83508292e-01 -3.22295576e-01 3.08552682e-01
3.33681911e-01 -2.52593607e-01 -7.90391937e-02 2.61114776e-01
7.98988640e-01 1.95453793e-01 1.74226701e-01 -2.19445437e-01
6.44320786e-01 -4.55998071e-02 6.10628605e-01 7.03867376e-01
-5.66450417e-01 7.22257435e-01 3.51451635e-01 -7.22344890e-02
-6.18213892e-01 -1.39416611e+00 -8.97437707e-02 1.07715988e+00
9.34457242e-01 -8.55393335e-02 -5.16723573e-01 -1.00687182e+00
1.76121831e-01 1.93680033e-01 -7.91776717e-01 -3.25344265e-01
-2.29142591e-01 -6.69736862e-01 4.18916196e-01 7.36145437e-01
8.60995591e-01 -7.66062915e-01 -8.63093019e-01 -5.86848967e-02
-1.19869992e-01 -1.35026550e+00 -2.55512148e-01 5.34234762e-01
-9.58179653e-01 -1.01733434e+00 -9.65385675e-01 -8.27014565e-01
3.63545746e-01 9.27204847e-01 7.68729150e-01 -1.34269074e-01
-1.97638571e-01 5.38393021e-01 -4.85206723e-01 -2.32644871e-01
1.78497255e-01 2.06859857e-01 -1.13626286e-01 3.38402167e-02
3.66498649e-01 -4.87942040e-01 -8.75909030e-01 2.77142435e-01
-1.00922418e+00 1.31750181e-01 9.24826145e-01 7.10610449e-01
6.77755415e-01 -2.65621632e-01 5.00440001e-01 -2.70330280e-01
4.87008207e-02 -4.58096802e-01 -3.79220515e-01 2.19564602e-01
-3.78577769e-01 -2.71416665e-03 2.52651572e-01 -4.90461662e-02
-1.02596021e+00 3.25649291e-01 6.74585849e-02 -5.49159050e-01
-2.13063374e-01 1.93034813e-01 -2.67624289e-01 -3.87877375e-01
4.53077078e-01 3.52605194e-01 1.61236525e-02 -5.45337856e-01
3.60416412e-01 4.83549535e-01 7.14006543e-01 -3.14950913e-01
8.80756855e-01 8.92068565e-01 1.69507608e-01 -6.09364450e-01
-1.23207128e+00 -8.61179411e-01 -7.47752428e-01 -2.60003477e-01
1.02557969e+00 -1.19873452e+00 -4.63237673e-01 6.85461104e-01
-8.19867909e-01 -2.20844850e-01 -1.92065373e-01 4.51370597e-01
-5.61166525e-01 3.39907646e-01 -4.28606629e-01 -5.91610312e-01
-1.79280296e-01 -1.21104431e+00 1.79105365e+00 5.12578905e-01
2.35610217e-01 -9.33045566e-01 -2.61195511e-01 6.79140270e-01
2.65455693e-01 3.12473804e-01 5.22161365e-01 -2.80084968e-01
-6.85510039e-01 -6.77700788e-02 -6.61326408e-01 2.61759758e-01
1.85081631e-01 -3.10083389e-01 -1.16089034e+00 -1.71091288e-01
-1.02391683e-01 -4.70226020e-01 1.26752079e+00 2.70262569e-01
9.48882163e-01 3.59854490e-01 -3.49977225e-01 7.09604025e-01
1.62343419e+00 -3.47378820e-01 2.99198449e-01 5.93638361e-01
9.18314219e-01 4.20456141e-01 7.27041602e-01 3.51673812e-01
6.14413381e-01 8.00351202e-01 8.18159282e-01 -4.99181837e-01
-3.41866434e-01 -1.16915219e-01 2.70300955e-01 4.64295059e-01
-1.45245001e-01 1.22220777e-01 -8.48402381e-01 5.13934374e-01
-1.93381310e+00 -7.78939486e-01 -1.34985700e-01 1.93956959e+00
6.80996895e-01 1.39226288e-01 4.40222085e-01 3.03424865e-01
4.71721977e-01 3.10518950e-01 -6.71795249e-01 8.95923451e-02
-6.12541556e-01 2.03370333e-01 6.46129489e-01 4.15801853e-02
-1.46130717e+00 9.98511493e-01 5.50703239e+00 9.61524189e-01
-1.27691114e+00 3.60025048e-01 4.03303206e-01 -2.26792783e-01
-3.50578278e-01 -1.61518082e-01 -7.28931785e-01 2.97170788e-01
2.25717917e-01 2.40523845e-01 4.23409678e-02 8.51189971e-01
-3.00528333e-02 -5.72671831e-01 -7.05110610e-01 9.98938143e-01
1.78711653e-01 -1.21643817e+00 -1.31690055e-01 -1.87299967e-01
8.88296783e-01 2.84230620e-01 3.00509989e-01 1.62363071e-02
-2.40417570e-01 -5.78146398e-01 1.01051605e+00 4.31179732e-01
5.81036329e-01 -6.09094441e-01 7.70589709e-01 1.01591885e-01
-1.41564274e+00 -3.53607565e-01 -2.54605383e-01 1.25452176e-01
9.13650021e-02 7.03146756e-01 -2.44367421e-01 8.41571093e-01
9.66505587e-01 1.12027681e+00 -7.70049989e-01 1.08334959e+00
-3.29249889e-01 8.62189829e-02 -3.62240374e-01 1.78512128e-03
5.36849737e-01 1.38428509e-01 5.89255691e-01 1.14089680e+00
1.56188339e-01 -1.59741208e-01 2.07509547e-01 8.89628291e-01
1.01295516e-01 -9.75127667e-02 -2.79426694e-01 2.44369105e-01
2.58194000e-01 1.43551254e+00 -1.15112197e+00 -1.58422530e-01
-5.60160160e-01 1.17248154e+00 3.22047353e-01 4.00275111e-01
-9.66939330e-01 -2.87082016e-01 6.97088778e-01 -2.39151925e-01
7.00647950e-01 -2.27009758e-01 -4.28787947e-01 -1.12753057e+00
9.60134044e-02 -3.99089098e-01 2.26503402e-01 -8.88833761e-01
-1.08484674e+00 4.86706436e-01 -1.33103788e-01 -1.43025458e+00
2.29719877e-01 -7.32995272e-01 -3.93036187e-01 6.47172987e-01
-2.12398458e+00 -1.46716845e+00 -5.88250875e-01 7.99223900e-01
2.68736899e-01 -3.14039364e-02 4.32267249e-01 1.71716318e-01
-7.46597528e-01 5.21247685e-01 -1.11288121e-02 -5.49442209e-02
4.83235031e-01 -1.32574975e+00 -1.85617402e-01 9.50125396e-01
1.14155769e-01 3.35139215e-01 4.00382489e-01 -3.21175516e-01
-1.57886338e+00 -9.96970415e-01 4.30042535e-01 -3.60076398e-01
6.00538552e-01 -3.59682649e-01 -6.22408271e-01 1.56616405e-01
-2.24365562e-01 3.69754583e-01 5.61866045e-01 -2.53471404e-01
-4.76883322e-01 -2.27667555e-01 -1.11538792e+00 3.34200054e-01
1.17439878e+00 -8.97950292e-01 -6.40285790e-01 1.66435435e-01
8.39931369e-01 -3.20103586e-01 -7.27610350e-01 5.81740260e-01
4.50785011e-01 -1.44197583e+00 1.22894537e+00 -8.37698206e-02
5.18726468e-01 -6.48590147e-01 -4.39789474e-01 -8.33417296e-01
-1.08942434e-01 -2.91756481e-01 -3.66611123e-01 1.07366669e+00
1.31404057e-01 -3.59709561e-01 8.89928818e-01 2.86816269e-01
-3.86044562e-01 -1.02541423e+00 -1.01006246e+00 -8.08809578e-01
-3.17075074e-01 -5.66103995e-01 3.98991227e-01 7.16548860e-01
-1.11958772e-01 6.47304952e-02 -2.19801050e-02 2.69316435e-01
8.72581482e-01 4.97178584e-01 4.46054250e-01 -1.19441736e+00
-1.43079683e-01 -8.19045007e-01 -6.35873854e-01 -1.18734264e+00
1.39080733e-01 -1.02610934e+00 5.87822460e-02 -1.44617021e+00
2.33007640e-01 -4.93598878e-01 -5.76786160e-01 5.86041749e-01
-3.77428025e-01 6.81775033e-01 1.84950396e-01 3.30978274e-01
-9.82848406e-01 8.85899246e-01 1.30661130e+00 -5.85028231e-02
-1.95685565e-01 -1.27380505e-01 -7.93820858e-01 7.39004672e-01
6.79337740e-01 -2.23765433e-01 -3.05143654e-01 -3.26765090e-01
3.70791666e-02 -2.66609609e-01 7.70940185e-01 -1.12996817e+00
3.34560484e-01 2.19344366e-02 4.14041072e-01 -5.22528410e-01
4.69294041e-01 -7.68587768e-01 -4.94966477e-01 3.02483410e-01
-8.62094574e-03 -4.42055285e-01 1.63563430e-01 7.62413442e-01
-3.82531255e-01 4.32560295e-02 7.50299037e-01 -2.67515592e-02
-1.25999081e+00 1.90743789e-01 -8.67417529e-02 -5.00866957e-02
1.07395577e+00 -5.91034293e-01 -2.70125061e-01 3.01844962e-02
-5.77081144e-01 1.02822699e-01 5.38826287e-01 4.22332406e-01
7.42631197e-01 -1.27386463e+00 -7.71469846e-02 3.50534767e-01
4.58565205e-01 1.56361073e-01 3.53886485e-01 1.28992116e+00
-1.26285017e-01 3.93956751e-01 -2.53985703e-01 -1.06604683e+00
-9.98002291e-01 4.99245405e-01 1.86325952e-01 4.38162461e-02
-3.90769064e-01 1.18987703e+00 2.47018859e-01 -3.38135570e-01
2.70313829e-01 -4.29060787e-01 -1.05513714e-01 2.76156962e-01
2.31237754e-01 3.28451395e-01 4.91125993e-02 -9.14627671e-01
-6.62164211e-01 9.13516879e-01 2.22872600e-01 1.85831472e-01
1.42651618e+00 -4.01220441e-01 -3.79098975e-03 4.30902094e-01
1.31967497e+00 -3.22190672e-01 -1.53751636e+00 -5.09867072e-01
3.26175727e-02 -2.94986725e-01 3.74673635e-01 -5.68498135e-01
-1.32395792e+00 8.01399350e-01 7.14999914e-01 1.30882319e-02
1.54448032e+00 4.35405970e-01 8.35191011e-01 1.78988233e-01
5.24959207e-01 -1.05098176e+00 4.58900899e-01 1.84783652e-01
6.84045374e-01 -1.69279373e+00 8.56845155e-02 -5.38047552e-01
-6.95214152e-01 9.91240025e-01 5.90607107e-01 -1.12426341e-01
5.95680237e-01 -6.76116198e-02 7.68755004e-03 -3.70538622e-01
-1.67703807e-01 -7.86877990e-01 6.29896343e-01 5.40407121e-01
1.67632088e-01 -1.80338719e-03 7.44442567e-02 7.02457130e-01
1.32391363e-01 -1.61383525e-01 1.20416693e-01 1.04475701e+00
-6.28250182e-01 -8.35658729e-01 -2.51326621e-01 2.55298376e-01
-1.72180891e-01 8.87231380e-02 -3.63535255e-01 9.16395545e-01
3.25572461e-01 8.19087744e-01 -1.08166829e-01 -4.72618997e-01
1.66798174e-01 -2.37862140e-01 5.50357938e-01 -4.35503483e-01
-6.37136698e-01 2.04375520e-01 -3.89051229e-01 -9.89404500e-01
-1.01697838e+00 -8.03758144e-01 -1.35253811e+00 1.19104020e-01
-5.36692441e-01 -2.95183092e-01 5.77725172e-01 1.09312606e+00
2.77185410e-01 4.39185947e-01 5.88557422e-01 -1.36969519e+00
-2.06978872e-01 -5.70567310e-01 -7.75262237e-01 3.43643367e-01
5.62120736e-01 -1.08652556e+00 -4.73330975e-01 -2.99163103e-01] | [9.662993431091309, -0.7715610265731812] |
c006d2a2-bcc7-4adb-9957-a3624bbb4f8c | collaborative-regression-of-expressive-bodies | 2105.05301 | null | https://arxiv.org/abs/2105.05301v2 | https://arxiv.org/pdf/2105.05301v2.pdf | Collaborative Regression of Expressive Bodies using Moderation | Recovering expressive humans from images is essential for understanding human behavior. Methods that estimate 3D bodies, faces, or hands have progressed significantly, yet separately. Face methods recover accurate 3D shape and geometric details, but need a tight crop and struggle with extreme views and low resolution. Whole-body methods are robust to a wide range of poses and resolutions, but provide only a rough 3D face shape without details like wrinkles. To get the best of both worlds, we introduce PIXIE, which produces animatable, whole-body 3D avatars with realistic facial detail, from a single image. For this, PIXIE uses two key observations. First, existing work combines independent estimates from body, face, and hand experts, by trusting them equally. PIXIE introduces a novel moderator that merges the features of the experts, weighted by their confidence. All part experts can contribute to the whole, using SMPL-X's shared shape space across all body parts. Second, human shape is highly correlated with gender, but existing work ignores this. We label training images as male, female, or non-binary, and train PIXIE to infer "gendered" 3D body shapes with a novel shape loss. In addition to 3D body pose and shape parameters, PIXIE estimates expression, illumination, albedo and 3D facial surface displacements. Quantitative and qualitative evaluation shows that PIXIE estimates more accurate whole-body shape and detailed face shape than the state of the art. Models and code are available at https://pixie.is.tue.mpg.de. | ['Michael J. Black', 'Dimitrios Tzionas', 'Timo Bolkart', 'Vasileios Choutas', 'Yao Feng'] | 2021-05-11 | null | null | null | null | ['3d-human-reconstruction', '3d-multi-person-mesh-recovery'] | ['computer-vision', 'computer-vision'] | [-3.20855290e-01 2.99270302e-01 -7.62153342e-02 -5.61362684e-01
-4.34969276e-01 -6.45610869e-01 2.88298309e-01 -6.79144740e-01
1.85234696e-01 3.56782019e-01 4.25533682e-01 5.34766138e-01
3.58891189e-01 -4.97323364e-01 -5.44407904e-01 -5.46408772e-01
1.66666925e-01 8.67313623e-01 -1.79349542e-01 -1.65635452e-01
-2.58798897e-01 5.02411544e-01 -1.50599611e+00 1.19062491e-01
2.47012034e-01 1.03168046e+00 -7.50887454e-01 7.05487907e-01
2.53989786e-01 2.05233291e-01 -3.43931764e-01 -1.09799504e+00
5.36371589e-01 -3.28677416e-01 -4.05563712e-01 3.71474952e-01
1.08663750e+00 -7.09105134e-01 6.36195466e-02 7.06984103e-01
6.88277781e-01 -4.12539281e-02 7.28383422e-01 -1.46255302e+00
-2.36548424e-01 5.92876151e-02 -1.10825729e+00 -6.95573032e-01
7.35485137e-01 4.08458859e-01 6.37917221e-01 -9.79821086e-01
7.23476171e-01 1.86748576e+00 1.11221325e+00 1.01762712e+00
-1.27368975e+00 -1.06240642e+00 1.46825582e-01 -5.89655280e-01
-1.53823221e+00 -8.25390458e-01 7.85320878e-01 -6.27791405e-01
4.40744609e-01 3.29216450e-01 1.07104170e+00 1.22203553e+00
-1.41240701e-01 6.23955965e-01 1.12555599e+00 -1.04904108e-01
-1.98583677e-01 1.32883400e-01 -2.65151650e-01 1.15088427e+00
1.45517111e-01 1.04464792e-01 -8.07067692e-01 -4.60272878e-01
1.04829860e+00 -9.96018648e-02 -1.98568761e-01 -6.61555052e-01
-7.72506475e-01 5.04563153e-01 7.75469095e-03 -3.08630049e-01
-1.22010358e-01 3.62114906e-01 -1.98721867e-02 -1.84092149e-01
6.38722122e-01 -2.25562821e-04 -2.92779058e-01 -1.12516284e-01
-9.08834636e-01 6.81206167e-01 1.07553923e+00 8.34607542e-01
6.81121230e-01 -2.59117726e-02 6.40918836e-02 7.23560631e-01
5.39315999e-01 1.18313062e+00 -2.07764208e-01 -1.56472123e+00
-1.88059062e-02 5.27220488e-01 7.43954405e-02 -1.33632708e+00
-4.33305234e-01 -4.96157743e-02 -6.27271295e-01 7.83493102e-01
7.15043366e-01 -2.35944912e-01 -9.49142694e-01 1.96621764e+00
1.01124847e+00 -2.63530612e-01 -5.09074569e-01 1.25746024e+00
8.82130921e-01 2.13669479e-01 2.22087353e-02 1.42432004e-01
1.61112595e+00 -6.86441362e-01 -4.87112015e-01 -4.17184860e-01
-2.75122792e-01 -7.21367121e-01 8.07762146e-01 6.33559883e-01
-1.59747040e+00 -4.03256267e-01 -6.75432384e-01 -2.01533780e-01
1.16053179e-01 2.05860823e-01 7.41090536e-01 1.10527170e+00
-1.08851588e+00 6.07272089e-01 -9.25544143e-01 -4.06585842e-01
5.21418631e-01 5.42405248e-01 -8.32872033e-01 1.21274188e-01
-5.99808693e-01 9.06967580e-01 -4.93758559e-01 -1.83483418e-02
-6.97848439e-01 -7.93283224e-01 -9.82836962e-01 -2.50693381e-01
2.45027512e-01 -1.11575377e+00 1.17970812e+00 -1.22816074e+00
-1.61428750e+00 1.46851611e+00 -2.99941152e-01 2.07167611e-01
8.83281946e-01 -3.97664160e-01 3.45297121e-02 2.23237216e-01
-2.13209927e-01 1.06422698e+00 1.27480769e+00 -1.54786813e+00
5.59214056e-02 -9.70440805e-01 -1.87722638e-01 5.33902824e-01
-6.70047998e-02 1.42950058e-01 -7.79507339e-01 -4.09997642e-01
3.85950767e-02 -9.52391624e-01 1.91770062e-01 8.49679112e-01
-1.95174918e-01 1.09714054e-01 6.38267756e-01 -1.01066673e+00
7.17779517e-01 -2.01429820e+00 2.85906434e-01 2.95507014e-01
6.17237151e-01 -1.28536284e-01 8.22963938e-02 -8.65383074e-02
1.33637622e-01 1.35990575e-01 -7.46378079e-02 -8.80236506e-01
3.21318388e-01 -1.33231292e-02 1.56618029e-01 7.37506330e-01
-2.14815084e-02 8.38676870e-01 -5.48605800e-01 -7.92618215e-01
8.13107267e-02 8.97688866e-01 -7.57236123e-01 2.96129435e-02
6.46041334e-02 4.87237573e-01 -1.65204152e-01 1.11963630e+00
9.19533432e-01 -7.60281235e-02 3.03454459e-01 -3.76751363e-01
2.97641158e-01 -4.35026318e-01 -1.21899033e+00 1.52836120e+00
-1.18796282e-01 1.97833911e-01 9.69149053e-01 -2.35131815e-01
8.97436440e-01 3.35312545e-01 6.11452818e-01 -1.82939351e-01
4.06358063e-01 -1.30202100e-02 -4.05777901e-01 -2.69687682e-01
1.13488555e-01 -6.11186683e-01 -4.13559265e-02 5.26666641e-01
4.97126468e-02 -5.68434536e-01 -3.25416476e-01 -3.19038816e-02
5.70198655e-01 6.46152794e-01 3.02118480e-01 -3.07699461e-02
-9.25503075e-02 -3.48006934e-01 5.28591871e-01 1.48443222e-01
-3.75218540e-01 1.03799284e+00 5.00826836e-01 -4.88279819e-01
-1.02452135e+00 -1.27368093e+00 7.15136454e-02 1.09787226e+00
6.20308630e-02 -3.20157439e-01 -1.08971429e+00 -5.40706575e-01
5.32356262e-01 1.45355403e-01 -8.78008723e-01 1.16624504e-01
-5.27474880e-01 -4.10934120e-01 5.75371623e-01 5.31828701e-01
2.17911512e-01 -6.80147767e-01 -6.57569885e-01 -3.00075859e-01
-3.30209941e-01 -9.39319551e-01 -7.90096045e-01 -6.07149005e-01
-6.10049009e-01 -1.06371129e+00 -7.93543577e-01 -1.68764427e-01
8.17463219e-01 -1.46301612e-01 1.47990298e+00 2.17961878e-01
-5.15268266e-01 8.82785976e-01 -3.45922075e-02 -5.17174006e-01
-1.85844541e-01 -4.76273179e-01 4.23534125e-01 -1.20664500e-02
1.82226330e-01 -7.25242615e-01 -7.33969510e-01 5.80097318e-01
-1.27539366e-01 1.99763641e-01 2.57727891e-01 5.08624792e-01
4.95175451e-01 -4.05698419e-01 -1.44540906e-01 -6.70966804e-01
1.07350849e-01 -1.48797929e-01 -2.69410282e-01 3.06044370e-02
-2.11947948e-01 -3.39208961e-01 -3.01970560e-02 -5.58496237e-01
-1.26891065e+00 2.52525866e-01 -1.51536256e-01 -7.51537085e-01
-2.08323836e-01 -2.97776222e-01 -4.59014058e-01 -1.83265403e-01
7.76238441e-01 -3.05133760e-01 5.53014755e-01 -4.99228537e-01
3.60252589e-01 2.93404430e-01 7.07097173e-01 -9.83292997e-01
8.09223950e-01 8.64492238e-01 9.10969004e-02 -6.75007641e-01
-7.39885390e-01 -1.39169276e-01 -8.11857998e-01 -6.40939534e-01
7.57274687e-01 -1.04193974e+00 -1.25795293e+00 5.24116576e-01
-1.02500391e+00 -4.44006473e-01 -3.76676112e-01 2.96462119e-01
-5.55885196e-01 2.52828628e-01 -5.03469706e-01 -1.07395291e+00
-4.94215310e-01 -7.74333000e-01 1.68868995e+00 1.41170412e-01
-8.66043329e-01 -6.97741389e-01 -2.20938679e-02 7.06519306e-01
3.92089076e-02 7.09622860e-01 3.37980688e-01 9.44666117e-02
-1.77855760e-01 -1.98442966e-01 -9.78559926e-02 2.33315393e-01
-1.15638018e-01 1.74441159e-01 -1.25277424e+00 -2.60855228e-01
-1.35110542e-01 -6.21443748e-01 4.57583994e-01 5.78997016e-01
7.18366325e-01 -5.21757901e-01 -3.52292985e-01 7.91198730e-01
9.16872799e-01 -3.65760922e-01 3.83674800e-01 -4.78511184e-01
8.54414999e-01 9.91899848e-01 3.77920836e-01 7.71972775e-01
4.93730366e-01 6.50556684e-01 5.00528753e-01 -2.28835091e-01
-4.22806621e-01 -4.63674814e-01 5.35243511e-01 3.00402880e-01
-5.90939164e-01 6.48009032e-02 -7.95854092e-01 1.51503950e-01
-1.34309256e+00 -1.01122081e+00 1.24484055e-01 2.18598056e+00
8.54313731e-01 -2.71971315e-01 5.34939051e-01 -1.68789461e-01
5.38747847e-01 1.11348676e-02 -6.49944067e-01 -1.61644116e-01
5.52280759e-03 2.08391279e-01 2.23241329e-01 5.83078444e-01
-9.56840515e-01 8.86888683e-01 6.56639814e+00 5.95295608e-01
-9.64651287e-01 4.91349399e-02 8.40190470e-01 -5.76963663e-01
-2.73696721e-01 -2.10590914e-01 -8.73333335e-01 1.06913708e-01
2.35597491e-01 1.70503721e-01 5.46124101e-01 9.10826683e-01
-6.90565705e-02 -8.31292942e-02 -1.04250991e+00 1.28717899e+00
5.46271265e-01 -6.16975427e-01 -1.63415000e-01 1.85999438e-01
5.88593602e-01 -5.62302589e-01 2.22316250e-01 8.91242325e-02
5.14920056e-01 -1.38517356e+00 1.08621502e+00 7.42324114e-01
1.21125031e+00 -5.57112873e-01 3.44144285e-01 2.81831205e-01
-1.03394186e+00 2.21674353e-01 -2.46483795e-02 -9.94511396e-02
1.54649824e-01 3.87556314e-01 -4.98459011e-01 1.09488674e-01
9.10919487e-01 4.18301672e-01 -3.01842988e-01 3.82081866e-01
-2.57647991e-01 2.59658992e-01 -7.12366819e-01 2.88010448e-01
-5.23211658e-01 -1.73658192e-01 5.12320220e-01 9.61035252e-01
3.39549303e-01 5.11269629e-01 2.41108805e-01 9.11612391e-01
-1.43966496e-01 5.02219386e-02 -3.82477850e-01 8.00420344e-02
2.89861798e-01 1.50583780e+00 -4.95706379e-01 -1.73493087e-01
-1.26418471e-01 1.00703406e+00 1.39846206e-01 1.16338260e-01
-7.18304455e-01 3.20086718e-01 1.05695534e+00 7.53617883e-01
-1.13688990e-01 -1.00938447e-01 -3.02223682e-01 -1.02208734e+00
-2.66307779e-02 -1.05408323e+00 2.00877547e-01 -1.06726086e+00
-1.25983393e+00 2.37695083e-01 1.82027832e-01 -8.71890008e-01
-2.95170844e-01 -6.45921469e-01 -3.30259502e-01 7.48593867e-01
-7.47992635e-01 -1.84523749e+00 -5.72562754e-01 5.71683824e-01
1.71323344e-01 3.84460568e-01 7.99334884e-01 7.34068379e-02
-2.66817331e-01 8.06424141e-01 -7.75086641e-01 2.37363547e-01
1.12032795e+00 -1.12136817e+00 4.32858258e-01 2.02825710e-01
-1.18111119e-01 6.87784195e-01 7.40737319e-01 -7.64536142e-01
-1.62867677e+00 -6.01142585e-01 5.14528513e-01 -1.16160083e+00
5.10842586e-03 -5.54835439e-01 -4.18205112e-01 8.04036140e-01
-1.25748083e-01 2.57290840e-01 6.69712305e-01 3.49997520e-01
-6.60757124e-01 -3.25080119e-02 -1.58184242e+00 6.17135525e-01
1.32622981e+00 -2.27140248e-01 -2.10409358e-01 -1.08102389e-01
6.19912297e-02 -8.15562725e-01 -1.06661308e+00 4.79938537e-01
1.52310002e+00 -1.30037796e+00 1.21580064e+00 -2.05109194e-01
3.00492644e-01 -1.78389490e-01 -9.36424732e-02 -1.09680748e+00
-7.71332011e-02 -7.07587361e-01 -2.16565281e-01 1.14397013e+00
-4.30588936e-03 -2.95564204e-01 1.15922630e+00 1.07415164e+00
3.31661046e-01 -7.31784344e-01 -6.69622242e-01 -4.48674917e-01
1.20190628e-01 -2.66432524e-01 6.81436300e-01 7.74653375e-01
-3.25688645e-02 1.72290668e-01 -7.53925145e-01 -9.78202149e-02
9.12694573e-01 1.71322301e-01 1.41358173e+00 -1.48956811e+00
-3.66897255e-01 -3.81222188e-01 -2.76086211e-01 -7.26082563e-01
1.66508794e-01 -5.42698264e-01 -4.98483703e-02 -1.04909086e+00
5.92020631e-01 -2.41988465e-01 6.28939450e-01 9.05350447e-01
-1.25231937e-01 8.38204145e-01 2.34065041e-01 3.69923674e-02
-2.84404486e-01 3.42841536e-01 1.47421861e+00 1.05797768e-01
6.53603449e-02 -6.46504909e-02 -7.82727480e-01 1.30020869e+00
5.64126670e-01 -1.57929376e-01 -2.03111872e-01 -3.56580615e-01
3.20657790e-01 2.06646249e-01 8.50601137e-01 -8.37358534e-01
-1.65649012e-01 -1.48538619e-01 1.14163959e+00 -4.21831131e-01
1.09056342e+00 -6.98774517e-01 6.84794843e-01 1.02712221e-01
2.09210798e-01 -2.61833906e-01 1.92663938e-01 1.97211638e-01
2.44714573e-01 2.55552471e-01 9.43941534e-01 -5.30480504e-01
1.28645655e-02 5.48603058e-01 1.57276034e-01 1.17417634e-01
7.34899580e-01 -5.04730761e-01 2.20687594e-02 -8.17178726e-01
-9.61219311e-01 7.06043541e-02 1.07228136e+00 1.89155072e-01
4.38243747e-01 -1.23933411e+00 -8.22855175e-01 3.79509866e-01
-1.99863330e-01 -1.06532797e-01 4.27952021e-01 8.34522069e-01
-4.16919887e-01 -3.52013350e-01 -2.38222793e-01 -6.15518868e-01
-1.86661017e+00 6.21939935e-02 4.97459948e-01 1.36587963e-01
-3.43009233e-01 1.05360556e+00 4.56816375e-01 -7.39001513e-01
4.80904430e-02 1.08760871e-01 2.46749386e-01 1.12728760e-01
5.37098587e-01 4.56990123e-01 -4.22323197e-01 -1.13449287e+00
-3.92510235e-01 1.39803767e+00 4.90193993e-01 -3.73118609e-01
9.72222507e-01 -1.90114453e-01 -8.80881175e-02 1.10817909e-01
9.28786814e-01 5.16261041e-01 -1.60717905e+00 1.76329315e-01
-8.69786024e-01 -7.16670156e-01 -3.07123601e-01 -9.36719835e-01
-1.22087765e+00 1.05826139e+00 4.31865484e-01 -4.92457926e-01
9.37067032e-01 1.54774830e-01 7.00964570e-01 -1.79072469e-01
4.98461276e-01 -9.55914259e-01 2.35027134e-01 1.05089642e-01
1.12407720e+00 -1.19869053e+00 5.94296515e-01 -6.88951075e-01
-7.95040250e-01 8.47923398e-01 8.96414161e-01 1.47075266e-01
6.08036518e-01 6.81395531e-01 1.98203370e-01 -3.52544129e-01
-4.03676510e-01 -2.91761290e-03 4.87353534e-01 7.32534170e-01
3.84058058e-01 3.22570026e-01 1.36639446e-01 8.03387880e-01
-6.51786745e-01 -7.91746303e-02 -1.18411787e-01 5.66882432e-01
-2.53314495e-01 -8.74429703e-01 -9.37994957e-01 1.81560650e-01
-4.39457357e-01 3.53507787e-01 -7.28667915e-01 8.27844203e-01
3.92345488e-01 5.83553314e-01 2.55351160e-02 -3.43068451e-01
3.14544111e-01 1.86138541e-01 1.04234362e+00 -4.19178128e-01
-5.14614701e-01 3.11831236e-01 7.55094141e-02 -8.35711181e-01
-3.14943492e-01 -8.05622339e-01 -1.08250356e+00 -7.60522246e-01
-1.13833278e-01 -2.84193784e-01 7.42234766e-01 3.93744975e-01
2.64874429e-01 -2.34952167e-01 3.40700507e-01 -1.55278111e+00
-3.65094036e-01 -7.81537652e-01 -6.21495128e-01 6.04620695e-01
1.80990800e-01 -9.26273823e-01 -4.96539593e-01 3.19853783e-01] | [7.160050392150879, -1.1093348264694214] |
4de95214-4be5-4590-939f-90b6c95c5573 | towards-effective-image-manipulation | 2210.08529 | null | https://arxiv.org/abs/2210.08529v2 | https://arxiv.org/pdf/2210.08529v2.pdf | Towards Effective Image Manipulation Detection with Proposal Contrastive Learning | Deep models have been widely and successfully used in image manipulation detection, which aims to classify tampered images and localize tampered regions. Most existing methods mainly focus on extracting global features from tampered images, while neglecting the relationships of local features between tampered and authentic regions within a single tampered image. To exploit such spatial relationships, we propose Proposal Contrastive Learning (PCL) for effective image manipulation detection. Our PCL consists of a two-stream architecture by extracting two types of global features from RGB and noise views respectively. To further improve the discriminative power, we exploit the relationships of local features through a proxy proposal contrastive learning task by attracting/repelling proposal-based positive/negative sample pairs. Moreover, we show that our PCL can be easily adapted to unlabeled data in practice, which can reduce manual labeling costs and promote more generalizable features. Extensive experiments among several standard datasets demonstrate that our PCL can be a general module to obtain consistent improvement. The code is available at https://github.com/Sandy-Zeng/PCL. | ['Shu-Tao Xia', 'Tao Dai', 'Shanzhao Qiu', 'Bowen Zhao', 'Yuyuan Zeng'] | 2022-10-16 | null | null | null | null | ['image-manipulation-detection', 'image-manipulation'] | ['computer-vision', 'computer-vision'] | [ 1.92713112e-01 -6.55084252e-01 -4.03174430e-01 -1.05769280e-02
-9.20760274e-01 -6.45792425e-01 6.26422346e-01 -1.65037170e-01
-3.00621092e-01 3.21406603e-01 -5.50240129e-02 -1.37241080e-01
4.11245406e-01 -6.55512393e-01 -6.11480832e-01 -1.12157071e+00
-8.90906714e-03 -3.59853119e-01 3.65996480e-01 -8.38837847e-02
5.42862296e-01 6.00421667e-01 -1.09387910e+00 3.74690056e-01
5.47177136e-01 9.75999475e-01 3.30388725e-01 4.17749852e-01
2.87676573e-01 8.23042870e-01 -6.11497104e-01 -3.75844650e-02
4.28315490e-01 -2.85194367e-01 -5.93044221e-01 4.03837323e-01
1.75184950e-01 -6.95976675e-01 -6.32606030e-01 1.27960360e+00
4.16577339e-01 -1.46360040e-01 4.86586243e-01 -1.40853977e+00
-8.36162150e-01 3.00360590e-01 -9.83146012e-01 2.46311605e-01
2.34339461e-01 4.98469144e-01 6.35481060e-01 -1.12458360e+00
4.10974234e-01 1.11408758e+00 4.54905719e-01 3.42941225e-01
-8.24469149e-01 -8.73154104e-01 1.12480789e-01 4.79559571e-01
-1.53228080e+00 -4.75811571e-01 1.28297269e+00 -1.26997642e-02
5.25711954e-01 2.59955496e-01 3.49236757e-01 1.35605800e+00
7.05139711e-02 1.36159742e+00 1.14877772e+00 -2.62178510e-01
-5.71113601e-02 1.76635921e-01 -5.91011941e-02 7.42984831e-01
8.52809101e-02 3.15934829e-02 -4.88419116e-01 -1.53781816e-01
1.00122070e+00 2.24491611e-01 -6.70661867e-01 -5.82431018e-01
-1.48979032e+00 7.08862185e-01 7.58633316e-01 5.50896764e-01
7.04427017e-03 2.26900727e-01 3.83336067e-01 1.58503518e-01
9.59407240e-02 3.32091391e-01 -3.44404966e-01 -2.83373203e-02
-7.52434611e-01 -5.81006333e-02 2.23596901e-01 9.21801925e-01
1.08052802e+00 -1.62439629e-01 1.48515850e-01 6.65376186e-01
2.17308313e-01 6.04505539e-01 4.81461257e-01 -9.54517782e-01
5.22150397e-01 5.53286672e-01 3.18232387e-01 -1.55363226e+00
-1.77935928e-01 -2.30166852e-01 -9.01151776e-01 -4.13300209e-02
3.04877311e-01 2.71121204e-01 -7.04887033e-01 1.26633561e+00
3.26999456e-01 3.36652696e-01 -2.03247458e-01 9.72209036e-01
4.57764328e-01 5.83508909e-01 -1.46681726e-01 -1.98630437e-01
1.18010938e+00 -1.13069177e+00 -5.36035478e-01 -2.96281010e-01
8.25686276e-01 -7.60700107e-01 1.19527066e+00 3.86963040e-01
-7.18729913e-01 -3.17065954e-01 -9.56220090e-01 -1.64485291e-01
-5.57529926e-01 3.98528397e-01 4.66908008e-01 6.81533337e-01
-7.11738884e-01 4.31619942e-01 -7.89382935e-01 -1.90851048e-01
8.39608788e-01 2.50231564e-01 -5.04963100e-01 -1.79991856e-01
-1.04834843e+00 5.83931327e-01 4.66137767e-01 3.34932655e-01
-1.05157113e+00 -6.12693131e-02 -7.76583672e-01 -7.65974820e-02
4.26908821e-01 -9.79642496e-02 6.71287537e-01 -9.28399622e-01
-1.32913697e+00 8.16853344e-01 -2.94610858e-02 -8.76767859e-02
5.17274737e-01 -1.56718060e-01 -9.37414467e-02 5.26417255e-01
1.73531651e-01 7.79095531e-01 1.27850139e+00 -1.73863661e+00
-4.65629101e-01 -3.81005079e-01 -4.07137387e-02 7.51124471e-02
-6.54636621e-01 2.30590895e-01 -6.76703632e-01 -8.49524260e-01
1.50649697e-01 -7.84330010e-01 -2.15861380e-01 3.26124668e-01
-3.47262353e-01 -1.97113212e-02 1.39787650e+00 -6.79839492e-01
1.01537764e+00 -2.36076140e+00 -7.62160793e-02 1.08943628e-02
1.66304544e-01 6.14592850e-01 -4.25321102e-01 3.79462779e-01
1.81481555e-01 2.09256187e-01 -3.90857816e-01 -2.36915961e-01
-3.26248765e-01 -2.52640117e-02 -4.68747139e-01 9.47982490e-01
4.60911453e-01 1.13218081e+00 -1.19015110e+00 -6.45669699e-01
4.99846101e-01 4.02840734e-01 -1.43328696e-01 8.29383582e-02
1.03718221e-01 6.22586250e-01 -6.91841722e-01 1.12754416e+00
1.22723532e+00 -3.14360440e-01 3.94876599e-02 -2.95856625e-01
1.37188986e-01 -4.08867262e-02 -8.33551645e-01 1.54158616e+00
-3.46402526e-01 5.75819731e-01 2.77468503e-01 -1.17191768e+00
8.26187134e-01 -9.02429223e-02 3.19461435e-01 -5.65095782e-01
4.99547362e-01 2.67612398e-01 -4.53901559e-01 -7.90074706e-01
6.38478220e-01 3.94406766e-01 -1.11829340e-01 4.23945695e-01
-2.16609538e-01 -6.81819618e-02 -2.15879142e-01 1.41974762e-01
8.87840211e-01 1.75653696e-01 3.71152848e-01 2.89358422e-02
6.38433039e-01 -2.66226619e-01 5.71407020e-01 8.58498812e-01
-7.57240772e-01 6.48086309e-01 4.99914587e-01 -3.11244845e-01
-8.07769477e-01 -7.94058084e-01 -3.53159308e-02 6.91182077e-01
9.14335787e-01 -3.78422678e-01 -5.12209713e-01 -1.21395290e+00
-2.06996128e-01 1.89444065e-01 -4.93857771e-01 -1.45248979e-01
-8.00369024e-01 -7.18556345e-01 5.97121954e-01 5.00335276e-01
9.51196730e-01 -9.92614686e-01 -5.26350558e-01 -1.58182383e-01
-4.66537446e-01 -1.12634635e+00 -3.96619886e-01 3.33006680e-02
-6.61985338e-01 -1.09302509e+00 -8.31946313e-01 -1.01412642e+00
7.11078525e-01 1.01515055e+00 4.36207920e-01 3.90816629e-01
-2.68335760e-01 3.58778745e-01 -5.33245683e-01 7.07869902e-02
-2.23370090e-01 -6.94042444e-02 -8.09355290e-04 2.45687217e-01
1.84877336e-01 -3.38896722e-01 -8.48754287e-01 7.20408797e-01
-1.06503510e+00 -9.32170302e-02 6.05973601e-01 1.06909120e+00
2.70215571e-01 6.01553768e-02 3.93687725e-01 -2.34799042e-01
2.58203357e-01 -4.25773561e-01 -3.41774940e-01 1.72174156e-01
-1.06723391e-01 -2.63805985e-01 4.31152552e-01 -5.38172662e-01
-7.05319405e-01 2.14172721e-01 1.93443626e-01 -6.80199742e-01
-1.66639522e-01 3.11508868e-02 -3.69895279e-01 -6.73427999e-01
3.89718443e-01 6.32509291e-01 1.43897776e-02 -1.26080647e-01
4.28747296e-01 1.00498664e+00 2.73427337e-01 -3.32482964e-01
1.02120340e+00 8.61062467e-01 -1.78904146e-01 -8.04681063e-01
-1.16973750e-01 -6.99684024e-01 -5.90293646e-01 -2.37670898e-01
3.90806407e-01 -8.52940261e-01 -7.07299054e-01 9.67778623e-01
-1.12767875e+00 -2.89840877e-01 1.31571144e-01 1.70145020e-01
-4.38754350e-01 1.19898283e+00 -7.04928935e-01 -8.29921663e-01
-1.04856566e-01 -1.41938984e+00 1.50858116e+00 2.65022844e-01
4.27972645e-01 -8.59055161e-01 -4.85321194e-01 2.88620323e-01
4.28509653e-01 -3.59620079e-02 5.14413655e-01 -3.16974163e-01
-1.00837457e+00 -5.17135382e-01 -4.88507986e-01 4.59669799e-01
2.23547041e-01 -1.64148182e-01 -9.40038919e-01 -4.24215168e-01
1.98233008e-01 -6.70907915e-01 1.20737219e+00 -2.30559465e-02
1.50314319e+00 -1.97959334e-01 -7.84778833e-01 5.52388251e-01
1.30790901e+00 -3.08652408e-02 8.12793314e-01 5.03372669e-01
7.78667986e-01 4.15277511e-01 9.35310066e-01 2.26188585e-01
1.92354217e-01 6.16318345e-01 7.44019568e-01 -1.93020508e-01
1.31266743e-01 -2.09651068e-01 6.43255949e-01 6.09845877e-01
1.76757604e-01 -2.10858881e-01 -8.05282712e-01 6.57220900e-01
-1.90667462e+00 -8.23399127e-01 9.67977643e-02 2.02564263e+00
5.53698123e-01 -2.14446276e-01 -2.12273952e-02 1.22664832e-01
9.43498850e-01 6.10600412e-01 -3.80417377e-01 3.02564472e-01
-2.43196383e-01 -2.70420223e-01 6.43912792e-01 2.40899652e-01
-1.65905881e+00 1.11439741e+00 5.77456570e+00 1.42201114e+00
-1.39199078e+00 1.13047965e-01 9.92650330e-01 2.98301756e-01
-1.11962579e-01 1.11569941e-01 -5.70201457e-01 5.16010344e-01
1.46130502e-01 3.11522305e-01 3.41483504e-01 8.64238203e-01
2.03511178e-01 -1.68958679e-01 -6.52468741e-01 1.26490283e+00
1.38723671e-01 -1.37603033e+00 2.94915587e-02 5.45986332e-02
6.10787749e-01 -1.42223597e-01 3.88937414e-01 1.12722091e-01
-2.76571035e-01 -7.31819630e-01 4.88277197e-01 1.73359767e-01
5.78203559e-01 -4.86583501e-01 6.41498983e-01 4.26750332e-01
-1.28968143e+00 -3.77530426e-01 -4.11020428e-01 2.15488642e-01
2.27024173e-03 2.64203042e-01 -7.05631733e-01 6.16566837e-01
6.96885288e-01 1.06937611e+00 -6.42275155e-01 8.39636266e-01
-3.64741504e-01 4.05203193e-01 -1.89825892e-01 1.01611661e-02
3.12673658e-01 -1.18342958e-01 4.74224627e-01 1.10859919e+00
1.62842691e-01 -3.38501126e-01 1.94091469e-01 8.33744287e-01
-3.71202864e-02 1.20954290e-01 -7.17570305e-01 4.29582186e-02
5.92062950e-01 1.57632101e+00 -9.74986553e-01 -2.32312605e-01
-4.33921158e-01 1.33164084e+00 3.11303318e-01 4.50972617e-01
-1.10877895e+00 -5.49474299e-01 3.14478338e-01 -9.49282944e-02
5.40338993e-01 -3.63246322e-01 -1.10096477e-01 -1.57885540e+00
1.24844261e-01 -1.14138782e+00 -5.38449809e-02 -7.04346836e-01
-1.17058611e+00 2.46651724e-01 -2.64391869e-01 -1.67595923e+00
4.90296632e-01 -7.10665524e-01 -5.79533756e-01 4.61759418e-01
-1.65075696e+00 -1.39443088e+00 -4.49015796e-01 6.53230131e-01
4.54055369e-01 4.07964475e-02 5.23350298e-01 1.25023305e-01
-7.70628512e-01 6.63624465e-01 1.80798277e-01 4.50235933e-01
9.08053696e-01 -7.93331087e-01 3.63900661e-01 9.87607002e-01
-3.86527516e-02 7.85371602e-01 1.38045982e-01 -4.16547865e-01
-1.50375676e+00 -1.24562609e+00 3.49969953e-01 -3.91672164e-01
7.46165216e-01 -3.68654579e-01 -9.24076259e-01 4.87778068e-01
1.96729243e-01 3.82564217e-01 2.34287828e-01 -4.82321709e-01
-5.33171535e-01 -1.10701621e-01 -1.18062901e+00 5.94737172e-01
9.07483816e-01 -7.29385257e-01 7.19778659e-03 4.34053212e-01
3.11061025e-01 -3.09308618e-01 -5.92337728e-01 4.48394209e-01
5.36543071e-01 -9.26449656e-01 1.02733278e+00 -3.81563120e-02
3.74030173e-01 -5.77580273e-01 -1.61762312e-01 -8.33022356e-01
-2.47903064e-01 -7.63418913e-01 -2.23846152e-01 1.08814263e+00
-7.05980659e-02 -7.14232802e-01 7.35001445e-01 -8.56441632e-02
2.68799931e-01 -9.07528877e-01 -1.01889861e+00 -7.81114519e-01
2.09118783e-01 -3.10096681e-01 4.35747653e-01 1.12804246e+00
1.51325792e-01 -1.32410631e-01 -7.51562595e-01 5.01012504e-01
5.74104369e-01 -3.47558744e-02 7.90111482e-01 -4.42520112e-01
-1.43476680e-01 -4.27580655e-01 -6.83625996e-01 -1.60730493e+00
1.19108059e-01 -7.49264419e-01 3.19009759e-02 -1.01054037e+00
4.39360917e-01 -6.79880381e-01 -3.97787154e-01 5.23803830e-01
-5.36102831e-01 6.87599778e-01 2.37115845e-01 5.70767462e-01
-7.86548197e-01 6.71402335e-01 1.31213462e+00 -3.67431849e-01
3.53988893e-02 -2.72755772e-01 -5.08632064e-01 6.24933004e-01
1.10534918e+00 -4.11284357e-01 -6.15422353e-02 -2.65900224e-01
-1.50810316e-01 -3.24185826e-02 7.02060044e-01 -8.85501921e-01
1.80000737e-02 -1.78399578e-01 5.05677521e-01 -5.04984617e-01
3.75938892e-01 -8.21156025e-01 -4.17291850e-01 6.98142946e-01
-1.41856909e-01 -2.90693760e-01 -2.18706913e-02 8.51396739e-01
-2.18267545e-01 -2.17440352e-01 8.80543590e-01 -2.48654902e-01
-7.95437694e-01 3.54071438e-01 -3.85961294e-01 -4.97608513e-01
1.36518061e+00 -1.81829929e-01 -6.42560363e-01 -4.43584025e-01
-2.47854382e-01 1.15462773e-01 8.37445974e-01 4.95654881e-01
8.45196724e-01 -1.35282242e+00 -3.52429509e-01 4.11325902e-01
2.79343903e-01 -2.74959564e-01 2.06632137e-01 1.19491339e+00
-7.36066043e-01 3.06452423e-01 -1.30807012e-01 -8.07632506e-01
-1.34519601e+00 1.02778172e+00 3.25128824e-01 -1.25413820e-01
-5.05666018e-01 8.70462835e-01 2.79120415e-01 -1.55579776e-01
1.70312867e-01 -1.23805575e-01 1.31104523e-02 -2.37267509e-01
6.64118409e-01 3.08035821e-01 -3.40615302e-01 -7.94419408e-01
-5.18408895e-01 7.24535644e-01 -4.02928650e-01 2.33609095e-01
9.43928063e-01 -2.91144669e-01 -2.32205853e-01 9.18141752e-02
1.45760846e+00 1.32333368e-01 -1.32111037e+00 -1.67271599e-01
-2.58529156e-01 -1.02932179e+00 -3.13033275e-02 -2.53990382e-01
-1.30154645e+00 9.98131156e-01 6.63977802e-01 2.25348160e-01
1.36292648e+00 3.60231623e-02 8.47068310e-01 3.36113691e-01
6.24527276e-01 -8.16038847e-01 5.05270898e-01 7.18282983e-02
8.49812329e-01 -1.46177936e+00 5.10306060e-02 -6.15405202e-01
-6.12857878e-01 1.20210636e+00 4.90100026e-01 -3.61142308e-01
4.04140264e-01 1.48433536e-01 1.21249206e-01 1.31619751e-01
-2.01842338e-01 -4.00553457e-02 -2.80154705e-01 8.63778532e-01
1.35717437e-01 -7.29738250e-02 -2.71323860e-01 2.13802069e-01
6.59474552e-01 -2.13690966e-01 5.33574402e-01 1.18958056e+00
-4.17178631e-01 -1.14802778e+00 -4.78908867e-01 1.15711875e-01
-2.88385779e-01 -7.64221996e-02 -4.72020179e-01 7.91310132e-01
1.32291675e-01 9.79682684e-01 -3.23475033e-01 -6.40458286e-01
-3.00798982e-01 -4.06786054e-01 5.70798934e-01 -1.75534755e-01
-3.16663653e-01 2.30861321e-01 -3.26519996e-01 -7.57309198e-01
-5.59119046e-01 -6.71190917e-01 -8.60921979e-01 -1.03023544e-01
-8.80497873e-01 -1.78256825e-01 3.61345410e-01 6.97514892e-01
5.18446505e-01 -2.00635251e-02 1.18887377e+00 -1.31119597e+00
-5.86080611e-01 -9.00650442e-01 -5.17467439e-01 2.61889428e-01
5.25931180e-01 -5.93804777e-01 -4.82123405e-01 9.39316824e-02] | [12.292232513427734, 0.9306294918060303] |
2251e2d1-09c2-4b7d-9279-9bca8f7416fd | event-based-structured-light-for-depth | 1811.10771 | null | http://arxiv.org/abs/1811.10771v1 | http://arxiv.org/pdf/1811.10771v1.pdf | Event-Based Structured Light for Depth Reconstruction using Frequency Tagged Light Patterns | This paper presents a new method for 3D depth estimation using the output of
an asynchronous time driven image sensor. In association with a high speed
Digital Light Processing projection system, our method achieves real-time
reconstruction of 3D points cloud, up to several hundreds of hertz. Unlike
state of the art methodology, we introduce a method that relies on the use of
frequency tagged light pattern that make use of the high temporal resolution of
event based sensors. This approch eases matching as each pattern unique
frequency allow for any easy matching between displayed patterns and the event
based sensor. Results are show on real scenes. | ['S. -H. Ieng', 'R. Benosman', 'T. Leroux'] | 2018-11-27 | null | null | null | null | ['3d-depth-estimation'] | ['computer-vision'] | [ 4.70342100e-01 -3.31012607e-01 7.33237922e-01 -2.85873294e-01
-3.22644651e-01 -3.73559475e-01 7.71845281e-01 3.78823094e-02
-6.26621425e-01 5.41935384e-01 -2.95388371e-01 -3.45905200e-02
-1.80078335e-02 -1.13268697e+00 -5.70748568e-01 -5.13052762e-01
1.59167379e-01 5.77645779e-01 9.54880118e-01 2.08074916e-02
3.68319988e-01 9.51328695e-01 -2.22249341e+00 1.38682678e-01
1.44793570e-01 1.35899615e+00 3.83745313e-01 8.02523553e-01
-3.74720991e-01 3.20107937e-01 -4.70177144e-01 2.05049351e-01
4.06227797e-01 4.26182337e-02 9.34272632e-02 3.24267782e-02
3.94351065e-01 -6.90465748e-01 -9.97978002e-02 6.77826941e-01
4.85520303e-01 -1.45033643e-01 4.02862102e-01 -9.90220547e-01
5.67060709e-01 -3.01756948e-01 -5.15830278e-01 1.23986728e-01
1.06948459e+00 2.07918555e-01 1.64141566e-01 -1.01428938e+00
1.11333954e+00 6.55091465e-01 4.95260149e-01 6.11722656e-02
-1.15291059e+00 -3.91299993e-01 -5.28041065e-01 1.60537809e-01
-1.19243395e+00 -4.23799723e-01 9.17008281e-01 -4.78888422e-01
1.42222834e+00 1.11393087e-01 1.11339235e+00 3.55650663e-01
5.99599600e-01 -1.21277012e-01 1.43514764e+00 -7.69732535e-01
4.57926184e-01 1.30367130e-01 3.55317928e-02 5.26521504e-01
3.59293148e-02 5.00525773e-01 -9.75129426e-01 -1.12655446e-01
1.04948938e+00 2.33822361e-01 -2.66727298e-01 -4.04967517e-01
-1.09209478e+00 1.54956222e-01 2.39738114e-02 1.83646038e-01
-7.09701598e-01 4.41046625e-01 -8.34470242e-02 1.33597150e-01
4.35991675e-01 -4.59571928e-02 -3.27744074e-02 -4.36533928e-01
-6.86540365e-01 6.22468777e-02 6.67124093e-01 9.39078629e-01
1.06549740e+00 -1.61356777e-01 2.41439402e-01 -1.24020956e-01
5.51377594e-01 9.13140059e-01 1.64904624e-01 -8.21679831e-01
1.16915237e-02 6.83623731e-01 3.76619220e-01 -4.53195184e-01
-2.97694325e-01 3.69376212e-01 -1.52101800e-01 1.24914575e+00
2.80813336e-01 7.88676143e-02 -6.56061709e-01 7.57637441e-01
5.62866688e-01 3.75493199e-01 2.21697288e-03 6.59104586e-01
6.37354136e-01 6.85873151e-01 -6.48828268e-01 -5.41046262e-01
1.08470941e+00 9.28427503e-02 -7.98654020e-01 4.21176076e-01
-1.31922185e-01 -9.66089427e-01 8.48473251e-01 8.15002680e-01
-1.24395025e+00 -3.39028090e-01 -1.02400386e+00 3.84293944e-02
-2.12063283e-01 -4.05568779e-01 5.48760831e-01 4.61895525e-01
-1.13270330e+00 3.06322694e-01 -8.68576527e-01 -4.98146653e-01
-7.15019628e-02 5.09026647e-01 -2.14546919e-01 2.11826175e-01
-3.92175406e-01 7.34973311e-01 -1.03050992e-01 -2.40151837e-01
-1.52258217e-01 -5.74581444e-01 -3.84616017e-01 -1.91188350e-01
-3.80317084e-02 -5.76688945e-01 1.31760550e+00 -4.20195341e-01
-2.17950630e+00 1.05209339e+00 -3.60325575e-01 -3.81238908e-01
3.68871719e-01 -2.97164526e-02 -1.99064210e-01 6.69135332e-01
-9.60120559e-02 2.74358869e-01 6.87840700e-01 -9.75672424e-01
-7.94247210e-01 -3.54604751e-01 -1.47846565e-01 2.89920252e-02
4.30709459e-02 -1.25643220e-02 -1.36786267e-01 3.09567839e-01
5.14671922e-01 -4.21397269e-01 -7.20063671e-02 4.13158745e-01
1.48986369e-01 -1.17972925e-01 9.95198071e-01 5.11581719e-01
5.07090092e-01 -2.04093981e+00 -4.35226113e-01 2.57954150e-01
-1.16526201e-01 -2.95220107e-01 5.02366602e-01 8.57023001e-01
1.99677661e-01 -7.27870166e-01 9.11850631e-02 -5.03537238e-01
-2.75180966e-01 1.63263574e-01 -5.51464915e-01 5.05727708e-01
-2.60395259e-01 1.84997410e-01 -5.60854018e-01 -3.57553750e-01
1.03233111e+00 6.07873201e-01 -1.83263451e-01 1.41246721e-01
-3.60810935e-01 3.25598985e-01 -1.74311608e-01 6.44023359e-01
7.32040167e-01 -1.11638643e-01 -1.23302013e-01 -1.18236341e-01
-1.06646430e+00 3.73996824e-01 -1.47006607e+00 1.85470366e+00
-4.14901108e-01 7.70805418e-01 -1.13915488e-01 -1.89304240e-02
1.15373218e+00 6.19816124e-01 7.01180577e-01 -9.33998525e-01
2.19892010e-01 2.72927672e-01 -8.17871451e-01 -5.06976724e-01
5.67494929e-01 -4.13160622e-01 5.01162350e-01 6.08508825e-01
-3.32820237e-01 -7.42548227e-01 -1.17280714e-01 5.29150653e-04
1.16320872e+00 4.74734843e-01 3.58664006e-01 -8.11460242e-02
2.04465196e-01 1.91031978e-01 1.41234741e-01 4.46941584e-01
7.52610341e-02 6.61821067e-01 -1.46099404e-01 -5.76715112e-01
-8.72359216e-01 -1.18444967e+00 -5.62433124e-01 9.60188657e-02
5.75043976e-01 -4.48369235e-01 -1.40922159e-01 1.82551786e-01
6.97503835e-02 3.90583694e-01 -2.51666725e-01 7.12440193e-01
-4.52943295e-01 -3.33220631e-01 -3.08574349e-01 2.57632673e-01
2.83250391e-01 -8.37936699e-01 -1.69964790e+00 3.69962931e-01
4.28747714e-01 -1.11467218e+00 3.68786335e-01 2.26471201e-01
-1.16953981e+00 -9.52904284e-01 -1.21079013e-01 -1.33463994e-01
4.12023664e-01 4.69629854e-01 1.16612625e+00 -3.38413954e-01
-4.99914229e-01 1.19215262e+00 -2.53024191e-01 -9.59129214e-01
5.38641065e-02 -6.96021020e-01 3.30356732e-02 -2.28008162e-02
5.54677725e-01 -1.06749833e+00 -7.12137938e-01 5.86458445e-02
-8.13319147e-01 2.79931277e-01 -1.14523046e-01 -6.16604947e-02
1.04566383e+00 -1.73957676e-01 -2.94610441e-01 -5.61013699e-01
-5.41022010e-02 5.00774756e-03 -1.63400674e+00 -3.06640804e-01
-7.79201865e-01 -2.19083667e-01 1.00550465e-01 -3.23908955e-01
-1.05486262e+00 6.29675627e-01 9.02490392e-02 -4.05867726e-01
-1.85762465e-01 -8.02442618e-03 2.60914356e-01 -4.46653873e-01
8.41306984e-01 1.50106132e-01 -9.40947793e-03 -2.93452024e-01
-3.02003343e-02 4.43017662e-01 6.35175884e-01 -2.14838386e-01
7.66064703e-01 1.27697253e+00 6.51521385e-01 -8.60764384e-01
1.40460670e-01 -6.51214838e-01 -5.80185235e-01 -8.68606448e-01
7.39429772e-01 -7.58252084e-01 -1.04710126e+00 6.11522317e-01
-1.57277179e+00 -1.14821739e-01 -8.31542611e-01 8.05994749e-01
-8.22106957e-01 -1.03835361e-02 -2.98578143e-01 -1.20990288e+00
-7.80684426e-02 -6.68290019e-01 1.29629946e+00 4.73825097e-01
5.05146943e-02 -7.12575734e-01 5.76653242e-01 -3.43771160e-01
2.03980073e-01 5.28914690e-01 1.07216939e-01 6.46033049e-01
-1.48041987e+00 -4.80575293e-01 -3.03575806e-02 -4.03803706e-01
-8.18015933e-02 4.00885016e-01 -1.36310518e+00 2.32130677e-01
5.11567056e-01 1.23112172e-01 2.45987073e-01 5.14102876e-01
5.45649707e-01 5.23195922e-01 -5.61659098e-01 6.25210822e-01
2.42478275e+00 4.43697482e-01 9.62166429e-01 4.33435261e-01
2.43834704e-01 2.79554814e-01 6.55017972e-01 7.75952101e-01
3.07924092e-01 1.01152062e+00 7.45189488e-01 2.10899010e-01
-1.67122960e-01 8.73287246e-02 2.14651510e-01 2.20980629e-01
-4.95788813e-01 -1.67037621e-01 -8.38182509e-01 2.31328145e-01
-1.55500841e+00 -9.64321077e-01 -7.03186154e-01 2.52855253e+00
2.64670312e-01 1.75527185e-01 -5.51477820e-02 4.59448785e-01
3.08045238e-01 -2.20508158e-01 -1.21531837e-01 -3.31979960e-01
-1.41374907e-02 6.74085379e-01 8.82435858e-01 6.22150958e-01
-2.54667789e-01 3.90349865e-01 6.97549295e+00 7.12439343e-02
-1.30655444e+00 1.82416737e-01 -4.13465261e-01 -3.51041585e-01
-4.82054949e-01 2.05221191e-01 -9.41733658e-01 5.77467084e-01
9.01116014e-01 -2.91370720e-01 1.52839139e-01 4.44190890e-01
3.78087044e-01 -1.08156180e+00 -1.01140642e+00 1.36061978e+00
-1.15502268e-01 -1.34278893e+00 -3.50875050e-01 2.43247047e-01
3.95604461e-01 2.04915524e-01 -3.71082127e-01 -6.99546695e-01
-1.14906929e-01 -2.67154932e-01 7.44087756e-01 9.91286159e-01
1.00060201e+00 -4.81737167e-01 1.75383314e-01 3.86084050e-01
-1.23508477e+00 4.08711255e-01 -2.88625896e-01 -6.01570725e-01
7.29223311e-01 9.47645485e-01 -1.00175357e+00 4.11121279e-01
7.30565608e-01 4.69520420e-01 1.13216341e-01 1.16612411e+00
-1.43363886e-02 1.72432542e-01 -1.07201195e+00 -1.33786887e-01
-2.52472669e-01 -2.72308856e-01 7.42638767e-01 8.69190633e-01
8.06579709e-01 6.03779137e-01 -2.53899634e-01 5.66159427e-01
5.47725499e-01 -1.46782652e-01 -1.25587654e+00 6.19725227e-01
2.74914622e-01 1.05399263e+00 -6.93641126e-01 -2.45580018e-01
-5.39991438e-01 8.60731065e-01 -2.43811876e-01 6.40737340e-02
-3.24814588e-01 -3.86172771e-01 1.35592669e-01 7.22806871e-01
1.37316227e-01 -6.76711798e-01 -5.50317347e-01 -8.85348320e-01
1.19047284e-01 2.22736105e-01 -3.86138563e-03 -1.38090944e+00
-9.27089930e-01 4.09791350e-01 4.38182056e-02 -1.70149243e+00
-5.66215515e-01 -6.92129612e-01 -6.55080557e-01 9.27075744e-01
-1.57384038e+00 -8.15443873e-01 -5.91942370e-01 9.84287918e-01
1.46745279e-01 9.52903181e-02 9.57566142e-01 1.04325145e-01
2.09163159e-01 -4.54500824e-01 -1.07541621e-01 -6.64663136e-01
4.89116609e-01 -1.06435776e+00 4.40962538e-02 8.36926103e-01
7.56772235e-02 -1.18172005e-01 7.83375204e-01 -6.03305042e-01
-1.74241769e+00 -2.54670441e-01 1.07638991e+00 -6.29072309e-01
3.24615657e-01 -3.69473904e-01 -5.38672864e-01 4.35145736e-01
1.26243234e-01 1.76832348e-01 2.82525957e-01 -2.57079035e-01
6.94298968e-02 -6.52463078e-01 -1.41331422e+00 5.43321855e-02
9.31757927e-01 -7.15474904e-01 -3.93036366e-01 4.63836104e-01
3.51342499e-01 -7.65635192e-01 -3.59730482e-01 3.57085943e-01
6.72598362e-01 -1.69376600e+00 6.81810558e-01 6.57573104e-01
-1.01067387e-02 -5.77101171e-01 -2.55486786e-01 -3.68442118e-01
7.00847059e-02 -8.00459981e-01 -7.75541132e-03 6.50712192e-01
-2.15113685e-01 -9.56706703e-01 9.31287169e-01 6.33842170e-01
-9.08037722e-02 -9.26864147e-03 -1.36801052e+00 -6.42366409e-01
-1.00722790e+00 -6.83090329e-01 4.11317587e-01 4.07667696e-01
1.06214896e-01 1.54613584e-01 2.05399573e-01 2.93830127e-01
9.17278647e-01 4.01194304e-01 8.08359385e-01 -1.45666552e+00
-3.34408343e-01 5.96733317e-02 -9.52296555e-01 -7.41089225e-01
-6.30273879e-01 -3.63110930e-01 2.00655665e-02 -1.48215294e+00
-3.88886720e-01 -3.80960226e-01 7.27569833e-02 -7.85894468e-02
6.07158005e-01 5.53147078e-01 -9.87142175e-02 2.65413672e-01
-2.15117767e-01 7.66355917e-02 6.63384795e-01 7.08051682e-01
-3.67894530e-01 5.82666621e-02 6.08433723e-01 6.93277776e-01
2.76086390e-01 -5.24986804e-01 -4.27664399e-01 -4.74935144e-01
7.55079269e-01 4.09769028e-01 5.70189714e-01 -1.58799934e+00
6.15397513e-01 -3.00640035e-02 3.96406591e-01 -1.10856915e+00
8.78396153e-01 -1.44324398e+00 7.97688544e-01 4.97649580e-01
5.92081249e-01 3.43252689e-01 2.78046668e-01 3.80705118e-01
-7.07729757e-02 -2.04793870e-01 7.33482361e-01 -3.22741270e-01
-6.50124311e-01 1.72253951e-01 -4.68638480e-01 -9.66902196e-01
1.27447677e+00 -9.32271421e-01 -2.18163610e-01 -2.36451924e-01
-4.19240206e-01 -5.04061222e-01 9.20269251e-01 -4.21127558e-01
7.36449718e-01 -1.15712810e+00 -1.40644923e-01 5.54715514e-01
5.55784814e-02 2.13266145e-02 1.34757444e-01 4.67926711e-01
-8.17162693e-01 3.00829560e-01 -3.64359438e-01 -1.19755304e+00
-1.27997327e+00 1.97971433e-01 3.87997955e-01 3.20361495e-01
-1.13680124e+00 5.08228064e-01 -2.30232388e-01 4.00702327e-01
-1.22540712e-01 -3.84007990e-01 1.72487095e-01 -1.33876100e-01
6.44205987e-01 3.61593604e-01 3.34129363e-01 3.42813358e-02
-3.91622216e-01 1.20772421e+00 5.94202578e-01 -9.20432627e-01
1.39977121e+00 -3.40439141e-01 -2.70435307e-02 9.23576534e-01
6.22966647e-01 3.72514784e-01 -1.41294730e+00 -1.60385773e-01
-5.14488399e-01 -6.97280705e-01 2.31726497e-01 -5.15568912e-01
-3.64345402e-01 8.45856190e-01 1.13303280e+00 4.67525810e-01
1.49365354e+00 -2.10889921e-01 4.96536702e-01 3.14881712e-01
1.01188815e+00 -8.25269997e-01 -2.72940636e-01 2.94275463e-01
3.78482610e-01 -9.51934695e-01 1.56188071e-01 -5.14578521e-01
2.81936020e-01 1.43717515e+00 1.09354280e-01 -2.82854110e-01
8.07136714e-01 8.64487827e-01 1.82032421e-01 -5.80935121e-01
-8.85262549e-01 -2.83459902e-01 -2.97785878e-01 6.47468626e-01
1.23887502e-01 -2.11764440e-01 -4.21136737e-01 -3.66009027e-01
8.63586515e-02 6.70574665e-01 8.58437240e-01 1.31176782e+00
-7.52433598e-01 -1.30263531e+00 -6.48125350e-01 5.71803972e-02
-1.67588051e-02 1.41447574e-01 1.39238715e-01 6.82129145e-01
1.14707515e-01 7.06411541e-01 6.53764844e-01 -9.37478840e-02
5.83269119e-01 9.41595361e-02 1.04007673e+00 -6.98761880e-01
-4.33544248e-01 1.77900821e-01 -1.83234721e-01 -9.43706095e-01
-8.24681580e-01 -7.75672853e-01 -1.34557223e+00 -6.61154315e-02
-8.28741416e-02 -3.14404696e-01 1.41623831e+00 5.87146878e-01
3.64050597e-01 -1.13817856e-01 8.23206484e-01 -1.09771729e+00
2.66744435e-01 -4.58026081e-01 -7.68612921e-01 8.92887563e-02
3.32121730e-01 -6.04620099e-01 -3.03324282e-01 1.81421980e-01] | [9.07041072845459, -2.3610386848449707] |
7a91b71e-18c0-496e-9c23-7594539be16e | artificial-cognitively-inspired-generation-of | 2112.02457 | null | https://arxiv.org/abs/2112.02457v1 | https://arxiv.org/pdf/2112.02457v1.pdf | Artificial Cognitively-inspired Generation of the Notion of Topological Group in the Context of Artificial Mathematical Intelligence | The new computational paradigm of conceptual computation has been introduced in the research program of Artificial Mathematical Intelligence. We provide the explicit artificial generation (or conceptual computation) for the fundamental mathematical notion of topological groups. Specifically, we start with two basic notions belonging to topology and abstract algebra, and we describe recursively formal specifications in the Common Algebraic Specification Language (CASL). The notion of conceptual blending between such conceptual spaces can be materialized computationally in the Heterogeneous Tool Set (HETS). The fundamental notion of topological groups is explicitly generated through three different artificial specifications based on conceptual blending and conceptual identification, starting with the concepts of continuous functions and mathematical groups (described with minimal set-theoretical conditions). This constitutes in additional heuristic evidence for the third pillar of Artificial Mathematical Intelligence. | ['Florian Geismann', 'Yoe A. Herrera-Jaramillo', 'Danny A. J. Gomez-Ramirez'] | 2021-12-05 | null | null | null | null | ['abstract-algebra'] | ['reasoning'] | [-7.36453235e-02 6.48545682e-01 2.08937332e-01 -1.17268145e-01
6.66799664e-01 -9.67307985e-01 1.46081352e+00 9.93070751e-02
2.34051362e-01 1.64622381e-01 8.76636710e-03 -6.83458924e-01
-1.05684769e+00 -1.31940091e+00 -1.81602374e-01 -2.08280772e-01
-6.13472879e-01 7.49146760e-01 -1.99904572e-02 -8.24565887e-01
4.51474220e-01 6.97551727e-01 -1.90578175e+00 -2.32703611e-02
1.20204735e+00 8.54319394e-01 -3.00304424e-02 5.82236052e-01
-9.90476847e-01 6.92750812e-01 -4.31191415e-01 -4.01052237e-01
1.85457706e-01 -2.61678964e-01 -1.08091557e+00 5.92678227e-02
1.27838645e-02 5.08268356e-01 7.04695135e-02 9.81286108e-01
-4.94655341e-01 -1.27994418e-01 6.87237918e-01 -1.88609397e+00
-9.29762840e-01 1.01595449e+00 5.48463106e-01 -5.43197751e-01
5.36876798e-01 -3.73358764e-02 8.83569539e-01 -1.01868808e+00
8.89092207e-01 1.53596187e+00 7.85103261e-01 9.88307148e-02
-1.58220863e+00 8.97752792e-02 -2.05510080e-01 -1.76428735e-01
-1.58769953e+00 9.16030779e-02 8.54596198e-01 -9.08618331e-01
2.20920265e-01 6.52206540e-01 1.15697598e+00 1.55218557e-01
-2.02192023e-01 4.67146575e-01 9.01306570e-01 -8.77766013e-01
5.32964528e-01 2.98577070e-01 7.97364473e-01 7.35993087e-01
6.22774243e-01 2.19901800e-01 -1.26192719e-01 -1.84411794e-01
1.33904719e+00 -1.26917139e-01 4.69933860e-02 -1.08225834e+00
-1.55341649e+00 7.44165897e-01 7.35877901e-02 1.05840600e+00
-1.42686591e-01 3.74945134e-01 4.13913012e-01 6.37666285e-01
-5.88300191e-02 9.19834495e-01 -3.03316504e-01 1.06622890e-01
-4.78963226e-01 4.75334406e-01 9.86557186e-01 1.46851254e+00
6.64748371e-01 2.61304379e-01 4.66262907e-01 1.95110053e-01
3.16536933e-01 1.44861639e-01 2.33175993e-01 -1.11962390e+00
-6.07036464e-02 1.07700253e+00 2.21308976e-01 -1.21885383e+00
-4.05416876e-01 -3.24866563e-01 -7.22258985e-01 3.58577847e-01
2.84646809e-01 3.04903686e-01 -4.78277326e-01 1.63296974e+00
3.89975607e-02 6.22399263e-02 3.26084316e-01 1.80096954e-01
5.29327989e-01 2.81895399e-01 -6.37993962e-02 -1.15020119e-01
1.06307113e+00 -4.05210614e-01 -9.84486580e-01 7.67497480e-01
9.76445675e-01 -1.91395134e-01 1.10989869e+00 2.18693823e-01
-1.40745890e+00 -6.36999130e-01 -1.44318461e+00 1.19221829e-01
-1.45485675e+00 -1.41969651e-01 1.50545454e+00 1.01995373e+00
-1.61690676e+00 4.50318009e-01 -4.26770627e-01 -5.53228140e-01
-1.23457626e-01 2.29448482e-01 -1.33635223e-01 5.47946334e-01
-1.18529809e+00 1.00850797e+00 8.16659749e-01 3.94293293e-02
-1.53888568e-01 -1.08803344e+00 -9.33358729e-01 -1.59966201e-01
7.68819302e-02 -1.25096762e+00 8.24341595e-01 -9.38260317e-01
-1.53018236e+00 1.08838141e+00 6.72562063e-01 -7.44873941e-01
6.60944462e-01 2.09015861e-01 -5.63619912e-01 -1.33722303e-02
-5.93505315e-02 4.39945489e-01 4.33761388e-01 -1.50176299e+00
-2.80310422e-01 -9.45572183e-02 6.18524075e-01 -1.45682216e-01
-8.30378756e-02 -5.71919568e-02 1.48959756e-01 -9.32366133e-01
6.29339933e-01 -4.48679537e-01 -6.05295636e-02 5.00704013e-02
-1.61684632e-01 -3.18142295e-01 7.56455302e-01 1.91732034e-01
1.27649033e+00 -2.00944304e+00 6.74138367e-01 5.58930635e-01
4.43849295e-01 -2.96309944e-02 2.44683504e-01 5.26327014e-01
-4.02476549e-01 6.41937256e-01 -2.83644348e-01 4.10313219e-01
7.87742555e-01 -4.33108956e-02 -4.81895059e-01 1.82120427e-01
8.44793245e-02 1.05897045e+00 -1.00801289e+00 -3.90891194e-01
7.75466502e-01 -8.23969170e-02 -5.85902750e-01 -2.79576421e-01
-6.35139644e-01 3.87698933e-02 -5.32931507e-01 5.09594619e-01
8.04953396e-01 2.07984507e-01 5.88266730e-01 -1.89819410e-01
-6.69663727e-01 -6.74527586e-02 -1.74015582e+00 1.95995474e+00
-4.72542197e-01 3.41462016e-01 2.32592419e-01 -1.01570678e+00
1.04829907e+00 2.54874021e-01 4.09414887e-01 -1.51138768e-01
-9.83958971e-03 2.94243813e-01 -7.90208727e-02 -1.45728797e-01
6.09508753e-01 -8.25213194e-02 -5.86847425e-01 4.86345768e-01
1.28189325e-01 -1.16263092e+00 3.07536632e-01 1.98520586e-01
6.18507028e-01 4.03934300e-01 4.76631403e-01 -1.21318114e+00
8.28650773e-01 1.21183015e-01 -1.02305345e-01 7.73645461e-01
3.05530578e-01 4.61496785e-02 4.85245913e-01 -7.86343515e-01
-1.20281518e+00 -1.75889945e+00 -3.31180423e-01 6.78428411e-01
2.63464093e-01 -9.65301514e-01 -8.48041117e-01 6.03358634e-02
-7.60275349e-02 8.25535357e-01 -6.62252426e-01 6.33803681e-02
-4.13382173e-01 -2.54937373e-02 7.86695361e-01 1.83170021e-01
3.38108480e-01 -6.86296105e-01 -9.33540881e-01 1.64915342e-02
3.80452335e-01 -7.95598328e-01 3.67232800e-01 -1.02758653e-01
-1.11046648e+00 -9.59388018e-01 -5.45328707e-02 -6.50960445e-01
5.03923893e-01 -2.27284625e-01 1.49548590e+00 7.17935145e-01
-3.93317461e-01 8.33823442e-01 -3.03064615e-01 -5.60236156e-01
-5.23732960e-01 -4.50647444e-01 2.62203366e-01 -2.05286428e-01
-3.00467998e-01 -9.70633924e-01 1.81307241e-01 7.15534538e-02
-1.33517611e+00 6.06805384e-01 -1.10849954e-01 5.53360760e-01
1.09770009e-02 4.54366744e-01 1.16875194e-01 -4.19696450e-01
7.49525368e-01 -2.25804761e-01 -8.13231945e-01 4.76512462e-01
6.23939037e-02 7.31124461e-01 5.23221552e-01 -1.83865681e-01
-8.20053697e-01 -5.94402075e-01 5.88011563e-01 -9.91504788e-02
-2.35547200e-01 7.31688261e-01 -4.85894263e-01 4.92227403e-03
2.70750970e-01 1.02624394e-01 3.39846350e-02 -1.37510866e-01
1.08972287e+00 1.33124352e-01 7.69160271e-01 -1.46612632e+00
9.40679848e-01 4.01583999e-01 7.45667160e-01 -1.00546920e+00
2.72837222e-01 3.63930553e-01 -1.18166280e+00 -8.05460662e-02
6.10483527e-01 -8.84599462e-02 -7.54706860e-01 3.93894136e-01
-1.12273026e+00 -2.00180486e-02 -9.05230880e-01 1.92305446e-01
-1.18418682e+00 5.93597218e-02 4.53249253e-02 -1.04177463e+00
1.79755285e-01 -9.35260952e-01 5.36956489e-01 -3.21404815e-01
-4.35374230e-01 -1.47363770e+00 -2.16692343e-01 -3.74901146e-01
4.36651349e-01 6.95045590e-01 1.56510925e+00 -5.28991342e-01
-6.61353767e-01 6.78444952e-02 1.34206535e-02 -1.26861334e-01
-4.96922694e-02 6.09501362e-01 -3.65610987e-01 4.90721494e-01
-1.58794764e-02 3.45237792e-01 -9.12882462e-02 6.22898377e-02
8.94818723e-01 -7.02404305e-02 -1.85051665e-01 3.87573272e-01
1.72923601e+00 3.59322488e-01 7.35651195e-01 1.25242651e-01
2.55205154e-01 9.07836735e-01 4.27944213e-01 1.89902678e-01
2.46182993e-01 8.70583594e-01 -6.32855967e-02 6.16420470e-02
-7.15351477e-02 -2.43510589e-01 -1.77825242e-01 8.62286210e-01
-4.02322888e-01 4.95004833e-01 -1.35703540e+00 2.52335340e-01
-1.97034431e+00 -1.09549153e+00 -3.57407391e-01 2.18617082e+00
4.99685049e-01 1.09634861e-01 2.08714485e-01 4.52609420e-01
9.40717936e-01 -3.94538820e-01 3.73341858e-01 -7.32317328e-01
-1.91809639e-01 4.71942961e-01 -6.78108931e-02 6.69182718e-01
-7.78701127e-01 9.34240639e-01 7.71125793e+00 6.11393034e-01
-4.56333429e-01 -8.77906010e-02 -4.56877798e-03 6.00622594e-01
-6.98836565e-01 4.32431996e-01 -7.36432499e-04 4.80210721e-01
8.59448791e-01 -6.56792283e-01 5.84058583e-01 5.92027247e-01
5.74739836e-02 1.89516559e-01 -1.46237183e+00 8.70404780e-01
-3.92994046e-01 -2.12904882e+00 6.82699442e-01 -3.94416489e-02
3.97283971e-01 -9.06229019e-01 -1.64544508e-02 1.79081447e-02
7.71832764e-01 -1.24149168e+00 1.41302633e+00 1.15217757e+00
7.17567801e-01 -6.64044201e-01 3.67014557e-01 4.11891155e-02
-1.70003927e+00 -1.69416294e-01 1.74884260e-01 -3.05465192e-01
4.86632884e-02 1.81186080e-01 -9.86154601e-02 1.28678691e+00
1.75248787e-01 3.23007852e-01 -4.06170160e-01 8.80521953e-01
3.42933327e-01 -2.62854815e-01 -4.12951976e-01 -8.83638263e-02
1.48618132e-01 -9.66641009e-01 6.31177068e-01 9.60428119e-01
5.15949905e-01 1.93885952e-01 -1.01328725e-02 1.79023767e+00
3.67526352e-01 -1.16578192e-01 -1.05165601e+00 -3.81485075e-01
6.93331122e-01 7.60276854e-01 -8.62671375e-01 -6.24214172e-01
-8.90622362e-02 1.72641143e-01 -3.25839490e-01 1.07662857e-01
-8.15629125e-01 -7.13153839e-01 5.86275518e-01 2.12651297e-01
-4.92055416e-01 -8.39717329e-01 -9.70586479e-01 -8.31169784e-01
-2.96669036e-01 -4.58102196e-01 4.68387753e-02 -1.13855159e+00
-8.29031408e-01 2.65396833e-01 7.16745555e-01 -1.17674053e+00
-2.90162981e-01 -8.57822001e-01 -8.14918637e-01 8.14413607e-01
-1.81315169e-01 -1.52254856e+00 -1.23929773e-02 4.72111404e-01
7.36767426e-02 -2.12626219e-01 1.19137490e+00 -2.03417301e-01
8.29256549e-02 3.27026844e-02 -2.07233846e-01 -1.55654207e-01
-4.63212132e-01 -1.63675487e+00 4.87159491e-01 5.15649021e-01
-7.95950070e-02 1.14876091e+00 9.68681216e-01 -3.13194990e-01
-1.80435514e+00 -5.77780902e-01 8.48370373e-01 -8.36858749e-01
1.12109029e+00 -5.13501585e-01 -4.29951668e-01 7.82579362e-01
1.25681803e-01 -3.52537274e-01 3.29895318e-01 -6.98471069e-02
-4.06961858e-01 1.56006455e-01 -1.32820666e+00 9.99470592e-01
1.58894181e+00 -6.47158206e-01 -1.22249389e+00 1.45509362e-01
1.33188951e+00 1.52544335e-01 -1.24366212e+00 5.59832275e-01
4.39812571e-01 -1.01041842e+00 1.43538821e+00 -5.50108075e-01
-4.42167431e-01 -8.06689203e-01 -3.76174629e-01 -9.21899319e-01
-2.30792776e-01 -1.06044650e+00 4.44656983e-02 1.01859307e+00
4.40680571e-02 -8.77560735e-01 1.81540355e-01 6.90873384e-01
-4.05922711e-01 -2.02602953e-01 -4.63466823e-01 -1.03260255e+00
3.82980555e-01 -7.89659739e-01 1.39967477e+00 1.31390238e+00
9.46885705e-01 -3.90364230e-02 6.34940386e-01 -4.52661783e-01
6.31226003e-01 1.49527773e-01 6.08619332e-01 -2.07158113e+00
1.43802419e-01 -1.10413539e+00 -1.13538611e+00 -1.81043133e-01
1.67897165e-01 -1.15223312e+00 -7.78763413e-01 -1.38004208e+00
-4.99854445e-01 -8.02258790e-01 1.64846525e-01 -8.01817998e-02
9.77853179e-01 -3.79314452e-01 3.87626976e-01 -8.14589730e-04
-3.26413065e-01 3.20303857e-01 9.46400821e-01 6.59808144e-02
-2.78361320e-01 -4.88745630e-01 -5.27950168e-01 1.03931749e+00
6.98083878e-01 3.94573390e-01 -7.65626609e-01 -4.05493006e-02
6.05123222e-01 8.02131668e-02 7.55233109e-01 -1.25135732e+00
1.71855271e-01 -5.55888712e-01 -2.87475705e-01 -6.14385724e-01
1.50251806e-01 -9.04443741e-01 1.03151906e+00 6.07944250e-01
-2.97236383e-01 3.58256489e-01 6.72815293e-02 -9.21066925e-02
-1.01695552e-01 -2.25579277e-01 5.80948770e-01 -2.92848378e-01
-7.26990044e-01 -2.79130518e-01 -2.98112601e-01 -2.12021530e-01
1.07139266e+00 -8.01133394e-01 -3.41984153e-01 2.69255728e-01
-1.02535415e+00 -2.57773638e-01 9.03602362e-01 2.79837549e-01
3.57085079e-01 -1.60791814e+00 -1.60447747e-01 1.60613686e-01
1.29551038e-01 -2.50878900e-01 -3.17927659e-01 4.50463027e-01
-9.83598650e-01 8.86220336e-01 -5.61407864e-01 -3.78842503e-01
-4.73198086e-01 8.34609985e-01 5.47177315e-01 1.68895766e-01
-5.85286736e-01 1.52391687e-01 1.90815136e-01 -7.59084404e-01
1.06013350e-01 -9.37524498e-01 3.95109206e-02 -2.14839861e-01
1.88957900e-01 5.36686122e-01 -2.70960659e-01 -4.23366755e-01
-1.79772824e-01 6.88602626e-01 8.91048968e-01 -5.44445574e-01
1.04262125e+00 5.61369620e-02 -1.10216916e+00 8.43203187e-01
6.81931496e-01 -1.85696617e-01 -3.14538106e-02 2.01885149e-01
6.27156079e-01 -8.32805410e-02 -4.89888817e-01 -6.80286765e-01
-1.39733315e-01 5.34229338e-01 3.42719704e-01 1.07093394e+00
6.09648466e-01 5.17396741e-02 -2.35824242e-01 4.19283152e-01
9.22213614e-01 -1.10005009e+00 -1.47149518e-01 6.01891756e-01
1.30143869e+00 5.47130033e-03 -3.23922276e-01 -8.58306766e-01
4.96197492e-02 1.47464991e+00 1.77798212e-01 -2.12610602e-01
1.02427387e+00 5.65588415e-01 -5.52614152e-01 -7.07402766e-01
-5.00381291e-01 -2.07537293e-01 1.36185765e-01 1.00679588e+00
4.32758898e-01 6.09308422e-01 -7.33298361e-01 7.48255253e-01
-6.78512275e-01 1.41582951e-01 6.88882947e-01 1.19368434e+00
-4.40510839e-01 -1.07791913e+00 -5.95211267e-01 -4.82016727e-02
2.62939245e-01 -1.57658324e-01 -5.73636770e-01 1.23531115e+00
6.76890671e-01 7.00830340e-01 5.29222310e-01 -4.37328905e-01
1.42423406e-01 2.27055356e-01 7.67298222e-01 -3.88096929e-01
-3.55056465e-01 -7.21539855e-01 1.72222868e-01 -3.61208498e-01
-5.71828723e-01 -4.07410949e-01 -1.49790347e+00 -5.98556161e-01
2.59480383e-02 6.33584857e-01 8.15438032e-01 8.22763264e-01
2.15226695e-01 3.81706029e-01 2.15542406e-01 -7.88912773e-01
-8.35630372e-02 -5.14537811e-01 -7.20212102e-01 2.94444144e-01
-1.12237602e-01 -9.72304642e-01 -2.39636779e-01 2.65291542e-01] | [9.023005485534668, 6.94024133682251] |
4109adb1-a92c-433e-83be-9685c069f661 | interactive-learning-from-natural-language | 2207.00627 | null | https://arxiv.org/abs/2207.00627v1 | https://arxiv.org/pdf/2207.00627v1.pdf | Interactive Learning from Natural Language and Demonstrations using Signal Temporal Logic | Natural language is an intuitive way for humans to communicate tasks to a robot. While natural language (NL) is ambiguous, real world tasks and their safety requirements need to be communicated unambiguously. Signal Temporal Logic (STL) is a formal logic that can serve as a versatile, expressive, and unambiguous formal language to describe robotic tasks. On one hand, existing work in using STL for the robotics domain typically requires end-users to express task specifications in STL, a challenge for non-expert users. On the other, translating from NL to STL specifications is currently restricted to specific fragments. In this work, we propose DIALOGUESTL, an interactive approach for learning correct and concise STL formulas from (often) ambiguous NL descriptions. We use a combination of semantic parsing, pre-trained transformer-based language models, and user-in-the-loop clarifications aided by a small number of user demonstrations to predict the best STL formula to encode NL task descriptions. An advantage of mapping NL to STL is that there has been considerable recent work on the use of reinforcement learning (RL) to identify control policies for robots. We show we can use Deep Q-Learning techniques to learn optimal policies from the learned STL specifications. We demonstrate that DIALOGUESTL is efficient, scalable, and robust, and has high accuracy in predicting the correct STL formula with a few number of demonstrations and a few interactions with an oracle user. | ['Jyotirmoy V. Deshmukh', 'Jesse Thomason', 'Sara Mohammadinejad'] | 2022-07-01 | null | null | null | null | ['formal-logic'] | ['reasoning'] | [ 2.35423610e-01 5.18747389e-01 -2.47465014e-01 -5.54117680e-01
-8.11510921e-01 -8.57264400e-01 6.49882734e-01 -1.67071611e-01
-1.73002005e-01 9.76762891e-01 -7.58793727e-02 -6.23050630e-01
-1.72423646e-01 -3.53582233e-01 -9.92662430e-01 -1.70307495e-02
4.70287129e-02 7.71462142e-01 5.43378592e-01 -4.70517218e-01
1.25906497e-01 5.36795735e-01 -1.50326467e+00 3.86072606e-01
7.37653911e-01 9.32688236e-01 7.29950547e-01 9.63211954e-01
-3.01539630e-01 1.29157722e+00 -3.71150613e-01 8.73140022e-02
3.13298106e-01 -3.74450386e-01 -1.20594049e+00 -1.71693608e-01
-2.39169016e-01 -5.25987327e-01 -2.20151376e-02 9.23600018e-01
-5.00374380e-03 -8.55854899e-02 4.43943381e-01 -1.83111632e+00
-1.91961899e-01 8.60009074e-01 1.80875108e-01 -5.43668330e-01
9.76128697e-01 8.13948274e-01 1.12304974e+00 -1.12736858e-01
6.67041361e-01 1.85068429e+00 5.25003076e-01 7.33918250e-01
-1.18707538e+00 -5.06417394e-01 1.84169918e-01 2.97516119e-02
-9.02847886e-01 -2.63822645e-01 6.59346461e-01 -6.15619957e-01
1.26380515e+00 -9.87190008e-02 7.41296351e-01 1.14840782e+00
5.23643315e-01 9.28075373e-01 1.17480457e+00 -5.35447598e-01
3.00214380e-01 1.38750359e-01 -1.25508144e-01 8.40220809e-01
-1.67107314e-01 4.69523549e-01 -3.58337551e-01 -1.65732726e-01
8.78193200e-01 -2.26785421e-01 6.62964955e-02 -8.84868741e-01
-1.30542243e+00 8.04180443e-01 1.35517552e-01 7.89613277e-02
-2.56112099e-01 7.39710450e-01 7.32296288e-01 7.19525754e-01
-4.31297630e-01 1.00511241e+00 -8.63203585e-01 -5.32912552e-01
-6.73458576e-02 5.14380574e-01 1.23512197e+00 1.12259090e+00
5.75166523e-01 -6.03308231e-02 -9.77978706e-02 3.90030205e-01
4.02913660e-01 5.40423512e-01 4.10069644e-01 -1.59572065e+00
3.45664084e-01 5.53178489e-01 7.11274564e-01 -4.47470129e-01
-4.28039253e-01 2.36059323e-01 -6.77701905e-02 5.46764493e-01
3.06974113e-01 -2.97888666e-01 -6.46780849e-01 1.84500027e+00
1.65287435e-01 -1.76249444e-01 6.18034780e-01 8.12643707e-01
8.09609964e-02 8.73632312e-01 1.52267769e-01 -4.77583744e-02
1.18424308e+00 -7.32119203e-01 -7.09013879e-01 -4.56281424e-01
7.77555943e-01 -3.47762018e-01 1.49534023e+00 5.11520207e-01
-9.48611021e-01 -2.99701154e-01 -1.03002822e+00 -1.00318611e-01
-1.58735439e-01 -1.55860716e-02 8.31735253e-01 7.70700257e-03
-1.09483171e+00 6.71412885e-01 -1.08512640e+00 -4.95515317e-01
7.68553019e-02 6.78366184e-01 -2.97149360e-01 -1.27926365e-01
-1.22947061e+00 1.11505663e+00 7.37092674e-01 -8.15089568e-02
-1.45117831e+00 -3.43265623e-01 -1.45470607e+00 -1.38125271e-01
9.02936816e-01 -5.98518014e-01 2.28830814e+00 -9.45048571e-01
-2.23981810e+00 2.98449308e-01 1.15384974e-01 -7.76640952e-01
4.36314553e-01 -2.02164471e-01 3.70516069e-02 1.06741212e-01
2.50823677e-01 9.70519185e-01 5.74204504e-01 -1.01525664e+00
-8.09717596e-01 -1.27195388e-01 7.29166865e-01 2.49565303e-01
3.77737939e-01 1.15794875e-01 4.56974050e-03 1.70890465e-01
-2.20305100e-01 -1.13094878e+00 -3.07822317e-01 2.02956274e-01
-2.46344820e-01 -6.58356190e-01 7.66115308e-01 -2.83908308e-01
4.76278991e-01 -2.00534749e+00 1.81231573e-01 -8.00639093e-02
-2.09973514e-01 -1.11579159e-02 -2.53225446e-01 5.18504560e-01
4.37096775e-01 -1.24483235e-01 -1.26386538e-01 -1.51391793e-02
7.45709181e-01 6.83982551e-01 -6.27123654e-01 1.58422664e-01
5.08956075e-01 1.05117393e+00 -1.30867040e+00 -4.54631597e-01
3.36547554e-01 4.24409695e-02 -6.21418595e-01 6.93781435e-01
-9.96959388e-01 5.69506586e-01 -7.13848948e-01 2.79894203e-01
-1.51669785e-01 1.79342981e-02 4.65081424e-01 2.95583695e-01
-1.45736456e-01 6.96045220e-01 -1.11730325e+00 1.81678557e+00
-7.58037388e-01 4.75926936e-01 9.88792628e-02 -1.01075947e+00
7.69526660e-01 5.64060390e-01 1.37699798e-01 -4.34556335e-01
1.63182646e-01 4.64936763e-01 -8.29991773e-02 -7.52366126e-01
7.09745884e-02 -3.79102439e-01 -6.81744516e-01 6.48710489e-01
5.99667802e-02 -1.00233531e+00 1.50072649e-01 -7.33993053e-02
1.17438936e+00 7.90453434e-01 6.47227526e-01 2.50384808e-02
3.24477792e-01 4.22318548e-01 5.79008222e-01 9.42840397e-01
-2.88497508e-01 -2.02000186e-01 7.66685963e-01 -6.29410982e-01
-9.40944552e-01 -8.61006200e-01 6.41317904e-01 1.07515657e+00
2.59936541e-01 -1.81931138e-01 -7.23399639e-01 -5.99059880e-01
-6.16545826e-02 1.15522349e+00 -1.03183091e-01 -2.81816721e-01
-6.29916072e-01 5.07263184e-01 5.94990849e-01 5.31492949e-01
4.27833885e-01 -1.67260921e+00 -1.40004528e+00 3.13645184e-01
-6.52769804e-02 -1.37249637e+00 -5.04163802e-01 5.53248286e-01
-7.66358852e-01 -1.13433838e+00 9.21409018e-03 -9.08417046e-01
4.45338219e-01 -1.37804642e-01 8.74225795e-01 -3.46968532e-01
-5.71132898e-02 6.46682560e-01 -2.66698897e-01 -6.66556597e-01
-9.44225132e-01 -2.37210006e-01 1.92987442e-01 -6.11651540e-01
9.78635624e-02 -2.94181526e-01 1.04938865e-01 3.17312181e-01
-7.01059639e-01 4.32677507e-01 5.47783852e-01 8.65559638e-01
3.51494551e-01 1.06063791e-01 3.40976328e-01 -6.71051025e-01
9.05707896e-01 1.36616603e-01 -9.46350217e-01 3.42488378e-01
-4.77034748e-01 8.05098295e-01 8.80752206e-01 -7.04938233e-01
-9.94774282e-01 6.90160155e-01 1.76739305e-01 -3.36632282e-01
-4.35549736e-01 4.97916460e-01 -1.05697177e-01 1.33489877e-01
6.61695242e-01 2.76641667e-01 2.68634766e-01 2.09754519e-02
4.92703855e-01 6.44746006e-01 5.78390718e-01 -1.18676579e+00
6.36911273e-01 -1.21748976e-01 -6.43804520e-02 -4.01189566e-01
-6.70565546e-01 -1.46217272e-01 -5.36273837e-01 9.30334628e-02
7.34798908e-01 -6.26676440e-01 -1.21329582e+00 4.81056236e-02
-1.37703121e+00 -1.07457578e+00 -4.32082832e-01 3.81759852e-01
-1.39489818e+00 5.66501729e-02 -4.33630288e-01 -1.25052214e+00
-6.28767237e-02 -1.61922359e+00 1.32915807e+00 -1.01193100e-01
-7.42941380e-01 -4.86972898e-01 -1.80422813e-01 2.94890609e-02
1.91726804e-01 1.15998253e-01 9.66221213e-01 -5.67990363e-01
-6.66186452e-01 -1.68844521e-01 2.17471439e-02 2.01946795e-01
1.42220393e-01 -3.78373951e-01 -6.37790620e-01 -1.63103849e-01
-1.38317361e-01 -9.50989187e-01 -1.66880652e-01 2.18875423e-01
8.19652796e-01 -6.19574785e-01 -2.55615681e-01 3.14476937e-02
1.03075445e+00 7.23548770e-01 4.05737042e-01 2.74398088e-01
1.96358219e-01 6.82100892e-01 1.00872934e+00 2.87738383e-01
6.07114017e-01 6.25533700e-01 3.19232047e-01 4.10964936e-01
2.71537185e-01 -6.46905363e-01 7.86061645e-01 2.87469141e-02
3.89044464e-01 2.54875809e-01 -9.63724911e-01 4.13524598e-01
-2.08077168e+00 -7.77845383e-01 6.18243277e-01 1.67463779e+00
1.21130288e+00 3.19416255e-01 7.74766207e-02 5.14119165e-04
4.14768666e-01 -4.42814708e-01 -7.18264937e-01 -8.09620321e-01
4.62286323e-01 -2.76148245e-02 2.13996708e-01 8.35507333e-01
-9.23422039e-01 1.28055656e+00 6.16290903e+00 3.07311296e-01
-1.16292465e+00 -2.23903149e-01 -5.12107164e-02 3.71778637e-01
-7.05177262e-02 2.07429230e-01 -4.97147709e-01 7.61271417e-02
8.78383696e-01 -1.61972195e-01 1.00049543e+00 1.02081466e+00
3.39461863e-01 -1.18224032e-01 -1.83050597e+00 9.40232813e-01
-4.16022241e-01 -1.00248098e+00 -8.97058770e-02 -4.57637429e-01
2.02676639e-01 -5.93617409e-02 -2.68310487e-01 7.13350773e-01
1.05777788e+00 -1.05508459e+00 1.20218158e+00 2.61572033e-01
7.75744557e-01 -4.46422279e-01 5.22141635e-01 9.84598398e-01
-1.05768442e+00 -6.65962696e-01 -1.28506243e-01 -4.19440418e-01
1.33677915e-01 -2.72593141e-01 -1.46443009e+00 2.44424522e-01
5.25180519e-01 4.88264948e-01 1.46996230e-01 4.88589287e-01
-7.44677603e-01 1.92256704e-01 -3.82824779e-01 -6.60895228e-01
4.53635514e-01 1.47455394e-01 4.57933754e-01 1.04972100e+00
2.33895391e-01 2.50173062e-01 6.79264605e-01 1.09595394e+00
4.13377345e-01 -5.37037432e-01 -1.08604395e+00 -4.44564313e-01
2.24996313e-01 6.37297511e-01 -2.69652039e-01 -1.32413298e-01
-1.50389910e-01 9.12353814e-01 2.60717422e-01 2.66077965e-01
-6.17138565e-01 -5.16288340e-01 6.35187745e-01 -4.27457318e-02
4.62718531e-02 -5.54127336e-01 9.88720879e-02 -7.31725574e-01
-3.85410003e-02 -1.10998178e+00 1.42979920e-01 -1.36601174e+00
-9.49624360e-01 4.13898826e-01 4.21379358e-01 -1.10823309e+00
-9.87346470e-01 -8.45076084e-01 -1.92501739e-01 5.89758217e-01
-1.46311057e+00 -9.93451715e-01 1.17194310e-01 3.53566825e-01
7.42298007e-01 -1.50020555e-01 1.03469706e+00 -4.74040538e-01
1.01796560e-01 3.00704520e-02 -6.68178678e-01 -7.96918273e-02
3.86058122e-01 -1.35226095e+00 2.13508010e-01 4.06173199e-01
-2.08737686e-01 5.70768893e-01 1.12340105e+00 -6.12543523e-01
-1.88870859e+00 -9.76661444e-01 8.35867941e-01 -4.97564465e-01
7.71671236e-01 -3.99854273e-01 -3.86798471e-01 1.14957368e+00
1.07295819e-01 -1.82657257e-01 -6.49296045e-02 -3.26184690e-01
-2.53267795e-01 -8.46413001e-02 -1.19642210e+00 9.35279012e-01
9.18050110e-01 -7.72581935e-01 -1.14544392e+00 3.18229765e-01
1.27696574e+00 -6.31511271e-01 -4.94428098e-01 3.36018264e-01
6.04168177e-01 -4.65404719e-01 6.87363863e-01 -6.71676397e-01
1.58951789e-01 -6.71807528e-01 -1.77720994e-01 -1.31210327e+00
1.26422182e-01 -1.08337903e+00 1.37795433e-01 5.24975061e-01
4.36163992e-01 -6.14058256e-01 3.62901896e-01 9.63912725e-01
-2.86074668e-01 -5.63475728e-01 -6.91716969e-01 -9.19450760e-01
-1.80130675e-01 -7.24187970e-01 6.19870007e-01 4.14121479e-01
5.50482512e-01 6.19618773e-01 -1.99632511e-01 1.85648322e-01
2.08643526e-01 2.37989902e-01 9.95283604e-01 -1.19770062e+00
-4.10917252e-01 -1.55355975e-01 -1.22503206e-01 -1.34294474e+00
8.38889062e-01 -8.21805060e-01 8.92883003e-01 -1.65013528e+00
-1.67013437e-01 -4.21814561e-01 9.36148092e-02 9.63768244e-01
6.01546347e-01 -7.26341069e-01 2.36054555e-01 -1.70401648e-01
-9.34315324e-01 5.34405589e-01 1.49732673e+00 -4.78172600e-01
-4.39334273e-01 2.94774286e-02 -5.41985095e-01 6.86399937e-01
7.68424571e-01 -3.00690889e-01 -7.98623800e-01 -4.30626303e-01
2.69824564e-01 4.90434051e-01 3.93226743e-01 -9.05725360e-01
3.73316944e-01 -7.79440582e-01 -2.41260961e-01 -1.14350237e-01
4.39177632e-01 -1.16375649e+00 1.32566735e-01 7.21854150e-01
-8.03672731e-01 4.21737395e-02 3.01603258e-01 3.80989164e-01
-5.10472991e-02 -3.12968671e-01 5.19726098e-01 -4.42309797e-01
-1.05200756e+00 -1.02347165e-01 -6.76740825e-01 -8.89759436e-02
1.16121733e+00 -6.76254258e-02 1.10966004e-01 -6.38364732e-01
-4.75163579e-01 6.57439232e-01 3.29614699e-01 4.05698597e-01
7.25665808e-01 -9.92825270e-01 -9.65595543e-02 7.54010454e-02
1.95416912e-01 3.20775628e-01 -4.68309730e-01 4.16838467e-01
-4.98402297e-01 8.16380858e-01 -4.23141897e-01 -6.04474068e-01
-8.71326327e-01 4.81490314e-01 5.51477849e-01 -3.19641948e-01
-7.43262827e-01 4.85850632e-01 1.99315637e-01 -1.07772648e+00
4.65406090e-01 -9.38358247e-01 3.07951979e-02 -7.52966166e-01
4.03340995e-01 -2.51541644e-01 -5.66202700e-01 -1.04455620e-01
-3.26817811e-01 4.13780868e-01 3.46285790e-01 -5.13909042e-01
1.26856530e+00 7.06203580e-02 -3.79613340e-02 6.86602056e-01
7.71585643e-01 -7.03314841e-01 -1.52926838e+00 -1.63505077e-01
4.33036149e-01 1.26839399e-01 -3.90201032e-01 -9.02277529e-01
-7.00070336e-02 8.24492514e-01 1.96754351e-01 2.75341943e-02
7.08999574e-01 1.90351531e-01 6.62178814e-01 1.23126161e+00
1.08625865e+00 -1.11527681e+00 3.85675907e-01 9.65399086e-01
1.13415837e+00 -1.11720812e+00 -4.40765798e-01 -2.56107479e-01
-9.76848722e-01 1.13350499e+00 6.88798845e-01 8.02546516e-02
1.77887663e-01 6.11500800e-01 1.03271261e-01 -1.64842047e-02
-1.05321872e+00 1.06335804e-01 -2.36138463e-01 8.37861121e-01
-2.39835456e-02 1.97215885e-01 2.26212695e-01 6.08712316e-01
-2.80735642e-01 4.13379908e-01 4.74525154e-01 1.36392319e+00
-6.09510839e-01 -1.17540860e+00 -2.50116199e-01 1.03986092e-01
-1.14307925e-01 3.78417999e-01 -4.32643592e-01 8.82101834e-01
-2.43197411e-01 1.03196526e+00 -2.35820368e-01 -1.91613257e-01
5.22387445e-01 2.97031373e-01 7.75759876e-01 -9.24844921e-01
-2.98317552e-01 -2.72070289e-01 2.58908480e-01 -1.03792620e+00
-2.02906787e-01 -5.49393296e-01 -1.89015317e+00 2.85808235e-01
1.31733611e-01 2.43838578e-01 5.76732934e-01 1.30663037e+00
1.71717122e-01 4.35791641e-01 3.65368664e-01 -9.71847296e-01
-1.06795979e+00 -7.12818861e-01 -2.77431905e-01 1.08534306e-01
5.91053963e-01 -7.55046666e-01 -7.44018555e-02 1.04250208e-01] | [4.438179969787598, 0.9375279545783997] |
a00277ed-c1f7-4e4e-be09-56b0b0a35e09 | degree-quant-quantization-aware-training-for | 2008.05000 | null | https://arxiv.org/abs/2008.05000v3 | https://arxiv.org/pdf/2008.05000v3.pdf | Degree-Quant: Quantization-Aware Training for Graph Neural Networks | Graph neural networks (GNNs) have demonstrated strong performance on a wide variety of tasks due to their ability to model non-uniform structured data. Despite their promise, there exists little research exploring methods to make them more efficient at inference time. In this work, we explore the viability of training quantized GNNs, enabling the usage of low precision integer arithmetic during inference. We identify the sources of error that uniquely arise when attempting to quantize GNNs, and propose an architecturally-agnostic method, Degree-Quant, to improve performance over existing quantization-aware training baselines commonly used on other architectures, such as CNNs. We validate our method on six datasets and show, unlike previous attempts, that models generalize to unseen graphs. Models trained with Degree-Quant for INT8 quantization perform as well as FP32 models in most cases; for INT4 models, we obtain up to 26% gains over the baselines. Our work enables up to 4.7x speedups on CPU when using INT8 arithmetic. | ['Nicholas D. Lane', 'Javier Fernandez-Marques', 'Shyam A. Tailor'] | 2020-08-11 | null | https://openreview.net/forum?id=NSBrFgJAHg | https://openreview.net/pdf?id=NSBrFgJAHg | iclr-2021-1 | ['graph-regression'] | ['graphs'] | [ 1.26876608e-01 2.66599059e-01 -3.28097224e-01 -5.13066053e-01
-5.47258139e-01 -4.62812364e-01 6.64617360e-01 2.97081262e-01
-5.09237409e-01 7.21581399e-01 1.29801527e-01 -8.49251509e-01
8.97102877e-02 -1.33423853e+00 -9.33350503e-01 -3.20885688e-01
-2.68154174e-01 6.09491885e-01 2.98978865e-01 -9.72817987e-02
2.65509397e-01 5.25054455e-01 -1.32118118e+00 3.58353168e-01
5.25402009e-01 9.46913421e-01 -3.26182485e-01 9.89368737e-01
-2.14749917e-01 9.61928129e-01 -7.78313577e-01 -6.32032156e-01
3.42807949e-01 -1.24625839e-01 -9.42456543e-01 -5.16352177e-01
9.95471001e-01 -6.99234843e-01 -5.19317687e-01 9.88405466e-01
3.93904746e-01 2.30331719e-02 5.26140273e-01 -1.24870098e+00
-8.32976401e-01 9.59490776e-01 -3.39265376e-01 3.66972834e-01
-7.29519725e-02 6.46596700e-02 1.24180424e+00 -3.74665737e-01
5.82758963e-01 1.51220560e+00 1.00839424e+00 4.43340868e-01
-1.48879325e+00 -7.95298100e-01 -2.21276760e-01 1.13207161e-01
-1.53488743e+00 -3.46317053e-01 3.03049952e-01 1.21703014e-01
1.54107308e+00 2.04536449e-02 5.18812895e-01 7.63139427e-01
4.01219904e-01 5.27793825e-01 7.93023765e-01 -4.63997364e-01
2.19893798e-01 -4.60619897e-01 1.11133538e-01 9.47815537e-01
5.65264881e-01 -5.60747795e-02 -3.85679483e-01 -1.77356929e-01
8.93058896e-01 -3.64536643e-01 8.94642323e-02 -8.20443258e-02
-9.32572722e-01 9.37550783e-01 9.03399646e-01 1.82649404e-01
-3.38144638e-02 1.07650912e+00 7.34307945e-01 3.54360968e-01
5.77933192e-01 5.46363711e-01 -4.54862744e-01 -1.83881491e-01
-9.20783281e-01 3.99962544e-01 1.02842844e+00 1.02857184e+00
9.10257339e-01 2.67775983e-01 -1.48740828e-01 5.22172213e-01
4.61230949e-02 -8.80106613e-02 1.76808655e-01 -1.19260466e+00
5.34317672e-01 6.00862563e-01 -4.37614381e-01 -1.11675751e+00
-4.17124957e-01 -5.04125774e-01 -1.06998134e+00 1.09947063e-01
5.76788247e-01 -8.41075089e-03 -1.15985560e+00 1.73023760e+00
1.77540034e-02 3.22333217e-01 -1.17361583e-01 2.95417100e-01
7.11860478e-01 6.00104332e-01 2.49421746e-01 4.03299868e-01
1.15919292e+00 -9.24176037e-01 -3.91836673e-01 -8.16393346e-02
1.14298582e+00 -4.46343690e-01 9.19878244e-01 5.88911474e-01
-1.12337625e+00 -3.93199891e-01 -1.25212872e+00 -4.43146735e-01
-6.26978457e-01 -2.69880950e-01 1.06570554e+00 9.48469281e-01
-1.88807714e+00 1.19170344e+00 -1.03360033e+00 -2.00147510e-01
6.20744944e-01 6.01842999e-01 -2.08404392e-01 -2.26046041e-01
-1.20871949e+00 7.04182506e-01 8.20018411e-01 -1.19746901e-01
-6.03705704e-01 -7.40965307e-01 -8.07299554e-01 4.09681499e-01
2.52059758e-01 -7.74090886e-01 1.35737288e+00 -6.36979699e-01
-1.16403449e+00 5.44207573e-01 -1.84581783e-02 -1.11729264e+00
1.68900162e-01 1.13677815e-01 -1.77282199e-01 1.78085655e-01
-2.53560573e-01 1.03257453e+00 3.64274621e-01 -7.06213832e-01
-3.72823238e-01 -8.23409036e-02 5.40179491e-01 -1.20579042e-01
-6.67559206e-01 -2.38454700e-01 -4.72093731e-01 -6.11644506e-01
-1.48420617e-01 -8.07931483e-01 -3.04836780e-01 1.77527238e-02
-4.72455382e-01 -5.67562222e-01 6.54175401e-01 -4.04421568e-01
1.17026329e+00 -1.71112454e+00 -3.64589155e-01 3.10879350e-01
7.18761563e-01 5.58666706e-01 -1.09514862e-01 3.73676330e-01
1.71150938e-01 6.95905149e-01 -7.63312262e-03 -3.89205426e-01
3.23405683e-01 6.49560928e-01 -7.62765780e-02 1.60778061e-01
5.63648880e-01 1.16296315e+00 -8.45549226e-01 -5.90115607e-01
8.84002224e-02 6.40476227e-01 -8.30522180e-01 -2.49029458e-01
-5.84518433e-01 -2.55020201e-01 -2.11901054e-01 5.50439298e-01
5.79296291e-01 -8.18147361e-01 4.08198118e-01 -2.55556673e-01
2.74229079e-01 7.41118491e-01 -8.35247397e-01 1.47986352e+00
-3.04952711e-01 7.95152664e-01 -1.25301957e-01 -9.82082009e-01
7.41297483e-01 5.38031273e-02 -3.08094528e-02 -8.40814114e-01
1.29626215e-01 1.23446830e-01 1.15464456e-01 2.27844268e-01
8.40492904e-01 1.50035545e-01 1.68702349e-01 3.28669161e-01
2.18547106e-01 -8.81263986e-02 4.58111674e-01 3.97193491e-01
1.56267416e+00 -2.28278667e-01 1.73579484e-01 -4.10201967e-01
6.58877715e-02 -5.06730750e-02 6.81833327e-02 1.05983877e+00
-1.92414671e-01 3.57385725e-01 7.46617913e-01 -8.70808840e-01
-1.27699673e+00 -8.08938205e-01 -3.31495851e-02 1.26602161e+00
-2.22577095e-01 -8.86845708e-01 -8.47387791e-01 -5.62699854e-01
-5.60369045e-02 7.09446788e-01 -5.50860882e-01 -1.46423936e-01
-7.23764837e-01 -8.47664654e-01 1.25671709e+00 9.50745523e-01
7.31836438e-01 -7.38586366e-01 -2.11726129e-01 4.53246921e-01
3.70161861e-01 -1.18979096e+00 -3.87900099e-02 3.65891963e-01
-1.13472807e+00 -7.78474033e-01 -2.97460556e-01 -5.74758828e-01
3.13251674e-01 -3.62878405e-02 1.72445047e+00 5.72190940e-01
-2.44196832e-01 -2.69979723e-02 -7.95861110e-02 -1.53137475e-01
-5.32368779e-01 5.85617185e-01 -3.12832296e-01 -8.22004080e-01
4.43919867e-01 -7.92404234e-01 -6.44060969e-01 -1.33348122e-01
-9.47583497e-01 -7.80688832e-03 5.56430578e-01 7.40004539e-01
3.55679572e-01 2.32914910e-01 3.42121154e-01 -1.13723528e+00
7.13040173e-01 -3.60047847e-01 -6.53849781e-01 -8.51912573e-02
-1.01439512e+00 4.66743588e-01 1.01243973e+00 -1.34334594e-01
-4.42764223e-01 -3.17836732e-01 -3.63153577e-01 -3.05663854e-01
6.33385312e-03 5.86013675e-01 3.12785506e-01 -5.21044731e-01
9.08512712e-01 -2.80772746e-01 -1.31614029e-01 -8.33919719e-02
5.34340024e-01 1.12190232e-01 4.28718895e-01 -8.71985495e-01
5.87241769e-01 2.76635915e-01 3.57597083e-01 -6.11839116e-01
-5.46095133e-01 -1.13408431e-01 -2.93725044e-01 2.63195932e-01
7.06593752e-01 -8.24540317e-01 -8.34942341e-01 3.39294463e-01
-1.19147646e+00 -7.39878356e-01 1.27629653e-01 5.28050959e-02
-2.72547603e-01 3.24751616e-01 -1.03829312e+00 -5.13349771e-01
-5.64740300e-01 -1.08613884e+00 1.12464702e+00 -4.05258499e-02
-2.65366197e-01 -1.47355616e+00 -1.77956030e-01 2.53954455e-02
7.73635268e-01 3.70262116e-01 1.21598029e+00 -7.15651631e-01
-9.90879893e-01 5.41423745e-02 -7.57189810e-01 4.10030484e-01
-3.29646677e-01 1.99911520e-01 -8.21352184e-01 -3.98003072e-01
-7.30257630e-01 -5.95893085e-01 1.04575408e+00 1.74684197e-01
1.58386397e+00 -4.97035593e-01 -2.85237044e-01 9.06072199e-01
1.66747344e+00 -1.86334878e-01 8.99115801e-01 2.64018208e-01
9.55222428e-01 -1.19348556e-01 -6.79036528e-02 4.83247191e-02
6.05582654e-01 3.53344679e-01 7.28203535e-01 -1.14873566e-01
-3.92320931e-01 -2.58273959e-01 -6.04135543e-02 7.65217364e-01
-1.53201148e-01 -6.39651597e-01 -1.13357210e+00 5.83537936e-01
-1.37298727e+00 -6.32459939e-01 -1.78409010e-01 1.86849546e+00
8.10768843e-01 5.99570513e-01 -1.24156877e-01 1.69917658e-01
4.06836480e-01 3.09934914e-01 -4.45203245e-01 -8.51433873e-01
1.57521829e-01 9.15729523e-01 1.07533216e+00 4.46742743e-01
-9.45313513e-01 1.01978922e+00 7.22761202e+00 1.04209352e+00
-1.14898300e+00 -1.46734342e-01 9.50507283e-01 1.12838559e-01
-3.17722112e-01 -3.63805331e-02 -9.08471763e-01 1.55342698e-01
1.59407949e+00 3.93982492e-02 6.42818451e-01 9.28029358e-01
-4.11953241e-01 2.44467303e-01 -1.21721363e+00 6.29472733e-01
-8.22938532e-02 -1.72323191e+00 1.98060110e-01 1.47878587e-01
8.18762541e-01 3.20112377e-01 -2.25426629e-03 5.73066294e-01
9.47173953e-01 -1.47659683e+00 4.00325537e-01 -7.56272953e-03
9.10158932e-01 -7.77999818e-01 6.83235347e-01 1.61715597e-01
-1.28809595e+00 2.81152159e-01 -6.33643270e-01 -4.35315222e-01
-2.63377577e-01 6.03260159e-01 -1.56966269e+00 4.59470212e-01
6.75580978e-01 5.39807379e-01 -8.41827810e-01 6.83770061e-01
-1.49492562e-01 8.92977118e-01 -6.24111831e-01 -4.52976972e-01
5.81126332e-01 1.27297744e-01 -1.05711654e-01 1.34481227e+00
2.18680963e-01 -2.04194486e-01 3.58235911e-02 9.96944785e-01
-6.24493718e-01 -1.50640711e-01 -5.70267558e-01 -3.12351316e-01
7.04837203e-01 1.07074964e+00 -8.38735819e-01 -5.60491085e-01
-3.02978396e-01 7.33747721e-01 8.73537660e-01 2.70550758e-01
-7.40898371e-01 -5.28825641e-01 3.09391588e-01 1.38079613e-01
5.35113215e-01 -5.32272041e-01 -3.92837465e-01 -7.92999685e-01
-6.65832832e-02 -9.10372555e-01 4.20480549e-01 -7.07736671e-01
-1.20394111e+00 5.27545750e-01 5.89409992e-02 -3.80791843e-01
-4.79461312e-01 -9.67201591e-01 -4.57511961e-01 8.65150690e-01
-1.54616928e+00 -1.08672595e+00 -2.73698866e-01 3.31939101e-01
-9.22349691e-02 2.62290686e-01 1.02755225e+00 4.16700661e-01
-1.90747038e-01 9.73047018e-01 1.10456161e-01 4.37788457e-01
4.47905153e-01 -1.59166241e+00 1.20103216e+00 7.12458611e-01
2.52277881e-01 9.93281245e-01 5.51970780e-01 -5.68805993e-01
-1.50677538e+00 -1.32854807e+00 7.31951535e-01 -2.23812655e-01
6.99043155e-01 -4.59430307e-01 -9.27229285e-01 9.75312769e-01
1.87705144e-01 2.62186974e-01 3.19612503e-01 4.25211221e-01
-8.09273243e-01 -6.41614646e-02 -1.04141450e+00 6.93719566e-01
1.21496630e+00 -5.40244877e-01 -7.82322660e-02 4.43980187e-01
1.05875075e+00 -6.86989486e-01 -1.25002182e+00 3.09564501e-01
4.50316399e-01 -1.01757205e+00 1.14893866e+00 -7.01051891e-01
5.43864608e-01 3.54394764e-02 -1.11640371e-01 -1.16102898e+00
-4.12754238e-01 -4.58053768e-01 -4.38354939e-01 9.09627557e-01
3.06522131e-01 -7.76559711e-01 1.28358734e+00 4.87351090e-01
-1.91310897e-01 -7.88963974e-01 -8.57540727e-01 -6.92173183e-01
4.61940020e-01 -4.23930943e-01 7.58192718e-01 8.58512640e-01
-4.99656379e-01 2.78214753e-01 -3.40046324e-02 8.14950392e-02
6.28746271e-01 -3.71978551e-01 9.10106361e-01 -1.07688701e+00
-3.98571998e-01 -7.10188150e-01 -8.41580570e-01 -1.16146994e+00
3.21930587e-01 -1.05347013e+00 -2.18925714e-01 -1.66183698e+00
-1.22814849e-01 -6.48218393e-01 -2.49969736e-01 7.46989071e-01
-1.01982936e-01 6.38192236e-01 -5.81069328e-02 -1.26229674e-01
-6.88182712e-01 1.57175824e-01 1.04056180e+00 -1.32165089e-01
3.33842069e-01 -5.54966450e-01 -7.93475747e-01 4.92274255e-01
1.00326848e+00 -2.91925699e-01 -5.20769119e-01 -8.23392212e-01
4.98051882e-01 -2.47705758e-01 5.27350247e-01 -1.30548143e+00
3.59496117e-01 2.06571624e-01 5.42286694e-01 -6.42172039e-01
3.30996811e-01 -3.79533917e-01 9.45794880e-02 4.68934953e-01
-3.16634029e-01 4.12593752e-01 5.34204960e-01 4.67456937e-01
-4.16807495e-02 -1.57271996e-01 5.63314736e-01 -2.01427624e-01
-6.51467264e-01 4.82180685e-01 -6.89809769e-02 3.16321641e-01
4.46072787e-01 -6.92123771e-02 -7.68142045e-01 -5.11266053e-01
-2.28483647e-01 -2.08096001e-02 5.41617930e-01 -5.37837408e-02
2.78050125e-01 -1.26433563e+00 -4.97879475e-01 3.45357321e-02
-1.88074008e-01 2.99858898e-01 -5.47351763e-02 3.28682989e-01
-1.09644198e+00 5.63825130e-01 -8.60923901e-02 -6.09696627e-01
-8.90444636e-01 5.11389196e-01 3.76415491e-01 -7.21528888e-01
-4.81413841e-01 9.70267594e-01 -2.88200021e-01 -4.75775093e-01
2.80501723e-01 -7.24137366e-01 4.32171375e-01 -5.14546037e-01
4.95157212e-01 3.63772601e-01 6.34472907e-01 -6.69816360e-02
-1.04644708e-01 1.01066425e-01 -3.09379458e-01 2.18213618e-01
1.23755014e+00 4.75719780e-01 -3.34280401e-01 3.93947028e-02
1.59157121e+00 -4.34436411e-01 -1.09726930e+00 -1.73257247e-01
-5.37681915e-02 -1.86838344e-01 1.97443172e-01 -5.46940327e-01
-1.01691830e+00 9.71042454e-01 2.90081114e-01 6.33273959e-01
9.10549164e-01 -2.70489275e-01 9.48353648e-01 7.23327756e-01
4.91633475e-01 -6.59310102e-01 -5.19571677e-02 8.75855744e-01
2.60795832e-01 -9.78168786e-01 2.32252881e-01 -4.25258100e-01
1.79135233e-01 1.25034082e+00 4.98820841e-01 -5.37336230e-01
4.14457262e-01 4.57253426e-01 -2.82826304e-01 -2.50091434e-01
-1.07808936e+00 1.74583107e-01 1.20313376e-01 7.21045136e-01
6.34667158e-01 5.41687869e-02 1.01633128e-02 -1.59348845e-01
-6.03192091e-01 1.99505296e-02 4.60185766e-01 7.95497000e-01
-2.66537964e-01 -1.18317008e+00 -1.15819070e-02 8.29248548e-01
-6.06565893e-01 -5.66025913e-01 -2.99417645e-01 1.12842417e+00
-1.46719739e-01 6.87561572e-01 2.88953245e-01 -4.73216563e-01
-7.89542645e-02 9.86338221e-03 6.56945407e-01 -5.59877336e-01
-7.36312509e-01 -5.30914307e-01 4.56598282e-01 -7.58300066e-01
-1.66807085e-01 -9.76364762e-02 -1.25682533e+00 -1.08355629e+00
-2.51640201e-01 -2.20050171e-01 6.21113181e-01 6.55652940e-01
4.63798165e-01 8.85159791e-01 -1.68440446e-01 -7.43472934e-01
-7.79928744e-01 -8.61136913e-01 -2.06571251e-01 2.17055678e-01
2.10968390e-01 -4.12340999e-01 -2.77595371e-01 -2.31642172e-01] | [6.995605945587158, 5.9654693603515625] |
c83548ba-3d88-48bb-bb1b-b45adbf85099 | machine-learning-for-early-prediction-of | 1904.07990 | null | http://arxiv.org/abs/1904.07990v2 | http://arxiv.org/pdf/1904.07990v2.pdf | Machine learning for early prediction of circulatory failure in the intensive care unit | Intensive care clinicians are presented with large quantities of patient
information and measurements from a multitude of monitoring systems. The
limited ability of humans to process such complex information hinders
physicians to readily recognize and act on early signs of patient
deterioration. We used machine learning to develop an early warning system for
circulatory failure based on a high-resolution ICU database with 240 patient
years of data. This automatic system predicts 90.0% of circulatory failure
events (prevalence 3.1%), with 81.8% identified more than two hours in advance,
resulting in an area under the receiver operating characteristic curve of 94.0%
and area under the precision-recall curve of 63.0%. The model was externally
validated in a large independent patient cohort. | ['Gunnar Rätsch', 'Karsten Borgwardt', 'Cristóbal Esteban', 'Matthias Hüser', 'Martin Faltys', 'Tobias M. Merz', 'Stephanie L. Hyland', 'Max Horn', 'Xinrui Lyu', 'Marc Zimmermann', 'Bastian Rieck', 'Michael Moor', 'Dean Bodenham', 'Christian Bock', 'Thomas Gumbsch'] | 2019-04-16 | null | null | null | null | ['circulatory-failure'] | ['medical'] | [ 9.53442007e-02 3.91848981e-02 -1.13522328e-01 -3.68819773e-01
-5.72484195e-01 -3.25474501e-01 -1.21486127e-01 1.16799843e+00
-7.04598606e-01 8.33867550e-01 1.18633471e-01 -7.02394307e-01
-4.21231091e-01 -5.78082383e-01 1.32013485e-01 -1.52952954e-01
-5.61928868e-01 8.90769422e-01 -3.34347002e-02 4.31656808e-01
1.41548663e-01 6.85837448e-01 -5.05272806e-01 4.92127270e-01
7.84713566e-01 1.05830324e+00 -2.25389861e-02 1.07977438e+00
1.03203617e-01 1.08396184e+00 -7.07041383e-01 2.84879178e-01
3.34699750e-01 -5.14281273e-01 -3.49477023e-01 -2.41023421e-01
-2.06038415e-01 -6.79627478e-01 -2.34778583e-01 4.33958508e-02
7.95902431e-01 -2.63378531e-01 7.79674768e-01 -9.72464144e-01
7.72716105e-02 2.28150845e-01 1.19164310e-01 7.40763545e-01
2.16452733e-01 4.80249345e-01 3.46704364e-01 -3.93680841e-01
1.04213305e-01 5.97197235e-01 7.20645070e-01 3.23163986e-01
-1.21379507e+00 -2.88161904e-01 -3.66864979e-01 -3.49941343e-01
-1.29559636e+00 -2.36636430e-01 -1.38322776e-02 -9.59073246e-01
1.06630516e+00 -9.07159820e-02 9.45482492e-01 3.44538182e-01
8.86383653e-01 -2.27355748e-01 9.60253596e-01 -3.21779609e-01
-1.22672036e-01 2.25989088e-01 2.47583166e-01 6.39070034e-01
5.67991972e-01 3.12094122e-01 -3.74598622e-01 -6.02490544e-01
9.54240501e-01 8.15242410e-01 -3.65484357e-01 -8.15915763e-02
-1.32808101e+00 5.08865178e-01 4.08121245e-03 -3.85957994e-02
-6.74124837e-01 -4.51300204e-01 5.57318687e-01 3.18570673e-01
-6.78694323e-02 4.99336153e-01 -9.10618305e-01 -4.25199330e-01
-5.62278569e-01 -3.77082109e-01 1.11907947e+00 5.46580911e-01
-4.66516688e-02 -1.48993447e-01 -7.86844194e-02 6.11422360e-01
8.20229575e-02 6.51162803e-01 3.54916483e-01 -9.29554582e-01
3.43225241e-01 8.28175664e-01 4.39693332e-01 -5.62660992e-01
-1.13531125e+00 -5.48635304e-01 -1.17592645e+00 1.09211504e-01
6.56978965e-01 -7.34019399e-01 -4.88742501e-01 1.20876741e+00
-2.07464278e-01 -4.23724085e-01 2.67234534e-01 6.60819948e-01
1.94943950e-01 2.04905510e-01 5.12671232e-01 -5.86231232e-01
1.27695751e+00 -2.64836729e-01 -4.67230111e-01 -7.75216743e-02
8.27713192e-01 -2.24587262e-01 4.55780953e-01 6.94060385e-01
-9.33692992e-01 -3.01822752e-01 -7.46370852e-01 5.97032785e-01
-1.84255838e-02 5.83909173e-03 1.50271326e-01 2.99468726e-01
-7.54596591e-01 8.07404101e-01 -1.08514726e+00 -5.02595782e-01
5.37974119e-01 3.32452238e-01 -2.80706048e-01 -1.06481068e-01
-1.15144026e+00 1.44224072e+00 5.73874891e-01 -6.86938241e-02
-4.89674658e-01 -9.23272550e-01 -5.69881678e-01 3.69462483e-02
-7.84916282e-02 -1.38815153e+00 1.01733661e+00 -3.44772398e-01
-8.03308129e-01 7.07156897e-01 -4.86841314e-02 -5.88339686e-01
8.16333532e-01 -6.02543771e-01 -5.45508087e-01 5.98378778e-01
-2.21499354e-01 -1.11686006e-01 3.03070009e-01 -4.88457590e-01
-8.50769043e-01 -4.94446307e-01 -6.22381449e-01 2.87330717e-01
-4.84263785e-02 1.65070355e-01 3.94120634e-01 -3.21848154e-01
8.44483301e-02 -5.70158780e-01 -7.45788991e-01 1.32475328e-02
1.74648836e-01 2.94540197e-01 4.27519560e-01 -8.10126841e-01
1.50935686e+00 -1.85058105e+00 -4.89413738e-01 1.76670000e-01
8.50117028e-01 1.37335047e-01 5.35248935e-01 6.40990496e-01
-2.42721170e-01 7.81334266e-02 -2.00677842e-01 3.07374775e-01
-5.51342368e-01 5.55730332e-03 -6.42378032e-02 5.11865377e-01
1.87054396e-01 8.82142365e-01 -1.01946759e+00 -3.61963779e-01
7.30353534e-01 3.18071187e-01 -1.66261494e-01 7.76935339e-01
6.15333915e-01 6.71661139e-01 -3.39598686e-01 4.75290895e-01
-1.45148799e-01 -7.53726542e-01 2.56266475e-01 3.66690904e-01
2.36492511e-02 4.66843434e-02 -5.55493832e-01 1.18545234e+00
-4.76480350e-02 2.87716597e-01 -2.73641407e-01 -7.86855698e-01
1.03158128e+00 6.33635938e-01 1.16922677e+00 -3.42746049e-01
2.76942968e-01 3.53903808e-02 2.55860925e-01 -7.62574852e-01
-5.50925732e-01 -7.68456936e-01 -1.43340006e-02 5.16323388e-01
-4.29896563e-01 -1.29173575e-02 -1.23934686e-01 2.32089430e-01
1.67316151e+00 -5.66406190e-01 1.02958930e+00 -2.85731226e-01
1.96665049e-01 1.45468906e-01 7.22984254e-01 9.77465928e-01
-5.21805167e-01 6.35091186e-01 4.53536898e-01 -9.27947998e-01
-8.88497651e-01 -1.27474058e+00 -5.26292145e-01 2.18123525e-01
-4.91431326e-01 -3.66894007e-01 -8.53058547e-02 -4.50502336e-01
4.93250847e-01 4.95061576e-01 -2.18014404e-01 -4.19172466e-01
-4.12122339e-01 -7.89009273e-01 3.49398196e-01 9.15180206e-01
-6.34699687e-02 -8.18483174e-01 -1.50832045e+00 8.85936379e-01
6.88089430e-02 -8.43124390e-01 1.47707194e-01 2.13533625e-01
-1.34350324e+00 -1.59409726e+00 -6.91305399e-01 -3.74197185e-01
5.18038332e-01 -1.87726751e-01 1.27522433e+00 1.58945426e-01
-6.62353814e-01 3.91665101e-01 -8.05677101e-02 -7.10307896e-01
-6.27750576e-01 -3.08235526e-01 4.87367272e-01 -5.17025411e-01
6.14122093e-01 -4.67025667e-01 -9.66610789e-01 2.97947615e-01
-4.48708862e-01 -4.56245877e-02 7.17132330e-01 8.91513169e-01
4.21612650e-01 -3.85742605e-01 1.14760816e+00 -8.00841987e-01
6.40772164e-01 -6.18607581e-01 -4.44077462e-01 1.90933749e-01
-1.41122413e+00 -3.98400575e-01 6.35818660e-01 -3.58311236e-01
-3.31958145e-01 2.37351432e-01 3.36286277e-01 -3.60442013e-01
-5.00064850e-01 4.34969097e-01 4.96419400e-01 6.33556843e-01
7.87285924e-01 -7.95962736e-02 4.55331266e-01 -3.05287868e-01
-4.29950029e-01 7.67444968e-01 4.89283115e-01 -2.43660688e-01
3.92625988e-01 -6.04343414e-02 4.02435213e-01 -4.71145540e-01
-6.36938751e-01 -7.55735934e-01 -1.09053624e+00 -1.90427452e-01
7.60133982e-01 -1.02835262e+00 -1.00080061e+00 2.53160834e-01
-6.90319657e-01 -3.97147417e-01 -3.28544766e-01 1.06664824e+00
-3.46924484e-01 2.68395394e-01 -6.52951539e-01 -8.64255428e-01
-7.25671947e-01 -5.21522582e-01 3.25419009e-01 -6.09740205e-02
-9.88576055e-01 -1.08055925e+00 3.49765390e-01 -1.18596174e-01
6.40745163e-01 8.72881174e-01 8.09303641e-01 -1.05544710e+00
-2.49745860e-03 -7.41296351e-01 -2.46970460e-01 3.40345293e-01
8.05573761e-01 1.98032577e-02 -4.70611781e-01 -4.17161971e-01
1.04728535e-01 4.81219664e-02 5.48133492e-01 5.28819323e-01
8.90074074e-01 -1.42925441e-01 -5.24830461e-01 3.15780848e-01
1.28316057e+00 7.28497982e-01 2.60184735e-01 1.58401102e-01
4.11205202e-01 2.33766854e-01 3.35998982e-01 1.04887617e+00
3.18151951e-01 2.51295567e-02 -7.44344592e-02 -9.79217291e-02
6.94331825e-01 1.08643554e-01 -2.47320369e-01 6.19860947e-01
-9.18644890e-02 4.32996489e-02 -1.66473997e+00 3.31458956e-01
-1.68253303e+00 -6.93463087e-01 -1.86619654e-01 2.43319488e+00
6.57081962e-01 3.98322314e-01 1.91414639e-01 -6.53553233e-02
5.96900046e-01 -5.60595214e-01 -7.30126679e-01 -3.30445975e-01
8.19274187e-02 2.14038361e-02 5.58978379e-01 3.07143450e-01
-8.62877965e-01 1.08977973e-01 7.28412390e+00 -7.44244814e-01
-7.38888502e-01 -3.99952382e-01 7.97816217e-01 -1.71909109e-01
8.12369585e-01 -8.95068124e-02 -3.56808752e-01 5.36251724e-01
1.66281605e+00 -6.91538274e-01 -1.14405360e-02 7.45810211e-01
6.27531230e-01 -2.93053240e-01 -1.35690892e+00 1.05276382e+00
-1.52393863e-01 -1.13397825e+00 -4.54351127e-01 -2.24735230e-01
2.16639787e-01 -2.87769316e-03 -5.19910336e-01 4.05503064e-02
1.43244237e-01 -1.10845315e+00 1.10352904e-01 1.12781394e+00
1.38614452e+00 -3.72111827e-01 9.10070002e-01 4.22462404e-01
-7.32844412e-01 -3.38861823e-01 2.23798081e-02 -5.38117349e-01
5.41426599e-01 8.43129635e-01 -1.44764745e+00 9.44942161e-02
4.49015260e-01 6.45458043e-01 -3.50038677e-01 1.30637395e+00
3.13978270e-02 6.14008486e-01 -5.42378962e-01 3.99535626e-01
-3.99637818e-01 6.68753460e-02 3.31202745e-01 1.11999655e+00
1.61889911e-01 9.28717375e-01 3.46546769e-01 3.02071065e-01
1.99055225e-01 1.96810365e-01 -8.81630480e-01 -2.86778063e-02
4.41309571e-01 8.28583658e-01 -4.53590453e-01 -6.13876760e-01
-5.70828617e-01 4.32438433e-01 9.79339108e-02 1.13772780e-01
-2.99283266e-01 -4.82573837e-01 3.18362385e-01 5.40832341e-01
-4.65617746e-01 -1.50608957e-01 -7.97430813e-01 -1.01918447e+00
-2.24868551e-01 -6.08873010e-01 9.56851244e-01 -5.47631204e-01
-1.40364575e+00 6.22510612e-01 -2.33636588e-01 -1.31072426e+00
-5.77830374e-01 -5.52129328e-01 -4.63875175e-01 1.34482396e+00
-1.25321412e+00 -1.01545297e-01 -4.82178241e-01 2.95141399e-01
9.12266001e-02 -1.20154031e-01 1.37946463e+00 -6.42080158e-02
-5.77605605e-01 1.21401764e-01 4.71417569e-02 3.47287118e-01
9.78971660e-01 -1.20045519e+00 -1.30170444e-02 2.93778598e-01
-1.14315701e+00 8.61191511e-01 4.02441502e-01 -8.48202825e-01
-1.01924634e+00 -1.05601943e+00 9.56920266e-01 -1.16086960e+00
4.78220075e-01 4.64371443e-01 -1.10330439e+00 4.67370063e-01
-3.09044302e-01 -4.48302254e-02 1.23909318e+00 -1.69590145e-01
7.60160238e-02 -6.39620051e-02 -1.29763544e+00 -2.04231031e-02
4.56471801e-01 -2.69391119e-01 -1.01628101e+00 1.41921967e-01
1.36523768e-01 -3.90865594e-01 -1.70657527e+00 9.32265937e-01
8.57208967e-01 -5.93059182e-01 6.75391138e-01 -9.74458039e-01
2.20112622e-01 -1.81121565e-02 1.93145797e-01 -1.03972113e+00
-5.33345461e-01 -5.86503267e-01 -2.51429707e-01 4.46820766e-01
4.44521606e-01 -1.24773574e+00 3.16817790e-01 1.38830018e+00
-1.18260451e-01 -1.02294624e+00 -5.04600167e-01 -4.59938109e-01
5.61114140e-02 -6.83107823e-02 1.47892535e-01 9.82369125e-01
6.78971648e-01 3.54384899e-01 -4.24891561e-02 4.31928597e-02
5.43193042e-01 1.35983482e-01 3.91908705e-01 -1.70936775e+00
6.82086274e-02 -2.30617657e-01 -3.33879411e-01 -2.72152990e-01
-6.03414059e-01 -5.47370732e-01 -2.97910392e-01 -1.93916142e+00
5.71440399e-01 -5.69800496e-01 -1.03221393e+00 6.20587468e-01
-4.02086765e-01 -4.18391228e-01 -1.72198460e-01 4.10431087e-01
-2.69660562e-01 2.05844939e-01 7.35317588e-01 4.08580333e-01
-5.79828382e-01 1.67641729e-01 -5.54474711e-01 6.83626711e-01
1.34494579e+00 -6.71037316e-01 -2.19475716e-01 -1.24105923e-02
-2.49005765e-01 8.26549888e-01 1.34521082e-01 -1.00008237e+00
5.37957177e-02 -4.89696950e-01 1.20259237e+00 -6.50719166e-01
-2.63669956e-02 -8.80631030e-01 4.31503594e-01 1.20132923e+00
-1.67707846e-01 5.79067767e-01 4.58756268e-01 4.17748332e-01
-2.19057966e-02 3.97140384e-01 8.64779174e-01 -2.67723888e-01
-2.42524758e-01 3.50774288e-01 -7.27660835e-01 3.11640799e-01
1.05751503e+00 -1.55762695e-02 -3.76160890e-01 -1.94412082e-01
-9.93193328e-01 4.01096910e-01 4.97676164e-01 3.47938955e-01
7.07225561e-01 -8.07158291e-01 -9.04626012e-01 1.74061000e-01
3.34005922e-01 -1.74589321e-01 2.15278983e-01 1.10067713e+00
-8.78874123e-01 6.47331417e-01 -3.07721019e-01 -5.65187871e-01
-9.93915319e-01 7.12858796e-01 5.97265124e-01 -1.67431489e-01
-1.16650593e+00 2.50164002e-01 1.26656145e-03 3.75820339e-01
2.10287198e-01 -2.51997679e-01 -2.00901404e-02 -8.91534016e-02
7.98720479e-01 3.46645623e-01 -1.53187647e-01 -2.47760803e-01
-6.59915805e-01 1.75739646e-01 -3.51714194e-02 1.15384646e-01
1.16160619e+00 2.48548072e-02 -2.51074992e-02 6.55617893e-01
9.48092341e-01 -5.55322528e-01 -1.07549369e+00 -1.34630710e-01
-1.00638783e-02 -4.76629525e-01 -3.40143383e-01 -1.17834210e+00
-4.38112348e-01 8.72325420e-01 8.99084747e-01 1.48969561e-01
1.03307962e+00 -2.03308076e-01 5.57045996e-01 5.01099169e-01
2.66621709e-01 -7.71105528e-01 -2.75629491e-01 2.74608999e-01
4.80051041e-01 -1.30965674e+00 2.04650968e-01 1.30062580e-01
-7.90317118e-01 1.14410770e+00 3.88579577e-01 1.99036703e-01
8.72618616e-01 2.10777849e-01 5.75619936e-01 -1.95349738e-01
-1.21768582e+00 4.98118162e-01 -1.32928029e-01 3.83671850e-01
3.51133406e-01 3.93215150e-01 -3.46467793e-01 8.10322523e-01
2.79563725e-01 4.68031764e-01 4.93002355e-01 1.31025457e+00
-8.31653237e-01 -7.10363030e-01 -1.26132712e-01 1.12795055e+00
-7.94434726e-01 -1.14678979e-01 -1.35846764e-01 3.92953455e-01
-4.53906775e-01 1.14031339e+00 -7.52237812e-02 -5.62706776e-02
6.73911393e-01 8.51368725e-01 1.45374224e-01 -6.96261287e-01
-3.54331017e-01 -1.08898893e-01 1.05408058e-01 -2.73421973e-01
2.82755960e-02 -6.89316392e-01 -1.55203497e+00 1.77484810e-01
1.97623149e-01 1.54802650e-01 4.06090409e-01 6.30598068e-01
6.52830243e-01 7.69689858e-01 7.64972031e-01 -1.71646625e-01
-7.36877501e-01 -1.04964495e+00 -4.95808333e-01 2.14664191e-01
6.63094342e-01 -4.96494919e-01 -5.30278385e-01 3.42754602e-01] | [7.978334426879883, 6.159065246582031] |
a69765fd-e7a2-43b5-a6fa-4e47218a2619 | bayesian-sparsification-of-deep-c-valued | null | null | https://proceedings.icml.cc/static/paper_files/icml/2020/6728-Paper.pdf | https://proceedings.icml.cc/static/paper_files/icml/2020/6728-Paper.pdf | Bayesian Sparsification of Deep C-valued Networks | With continual miniaturization ever more applications of deep learning can be found in embedded systems, where it is common to encounter data with natural representation in the complex domain. To this end we extend Sparse Variational Dropout to complex-valued neural networks and verify the proposed Bayesian technique by conducting a large numerical study of the performance-compression trade-off of C-valued networks on two tasks: image recognition on MNIST-like and CIFAR10 datasets and music transcription on MusicNet. We replicate the state-of-the-art result by Trabelsi et al. (2018) on MusicNet with a complex-valued network compressed by 50-100x at a small performance penalty. | ['Evgeny Burnaev', 'Ivan Nazarov'] | null | null | https://proceedings.icml.cc/static/paper_files/icml/2020/6728-Paper.pdf | https://proceedings.icml.cc/static/paper_files/icml/2020/6728-Paper.pdf | icml-2020-1 | ['music-transcription'] | ['music'] | [ 1.54032245e-01 3.39079946e-01 -1.46664366e-01 -4.91652995e-01
-7.19895840e-01 -3.62360895e-01 4.31307524e-01 -3.23505223e-01
-7.40011215e-01 8.33673954e-01 -1.32187352e-01 -9.78957936e-02
-6.48094416e-01 -3.56124520e-01 -9.97886300e-01 -7.26740897e-01
-3.26583862e-01 6.08105958e-01 -1.98980033e-01 3.34150344e-01
-1.22426234e-01 4.14271563e-01 -1.26208150e+00 2.87407964e-01
3.57402980e-01 1.24475050e+00 1.78943537e-02 4.85636801e-01
5.56743503e-01 6.83990419e-01 -3.91375601e-01 -6.26127481e-01
3.48958373e-01 -1.44804612e-01 -4.92229313e-01 -1.92164823e-01
6.07798100e-01 3.47999334e-02 -7.46606648e-01 1.49692476e+00
4.99902844e-01 2.11179644e-01 9.68542099e-01 -1.20533109e+00
-3.93011510e-01 1.14363062e+00 -4.11279917e-01 1.01723023e-01
-6.52623355e-01 2.12359965e-01 1.01877177e+00 -9.15114701e-01
2.37093478e-01 1.34947586e+00 9.29231048e-01 4.50027317e-01
-1.31770611e+00 -8.74907732e-01 -1.29490554e-01 3.02185059e-01
-1.58750534e+00 -6.20975554e-01 6.69858158e-01 -2.54901618e-01
9.40950215e-01 -2.94948876e-01 6.05720878e-01 1.44799912e+00
-8.80351961e-02 6.09015465e-01 6.25211775e-01 -1.30897611e-01
5.33964336e-01 1.44534245e-01 1.65571153e-01 6.58657730e-01
4.79789466e-01 7.06460401e-02 -7.54456341e-01 -3.76734976e-03
7.46661365e-01 1.20920114e-01 -6.32079877e-03 -2.48947993e-01
-1.27170289e+00 9.01478231e-01 4.27659184e-01 1.13857269e-01
-5.46750605e-01 9.80138540e-01 5.55652201e-01 5.29046416e-01
3.50915790e-01 3.20655853e-01 -5.47742903e-01 -5.45444727e-01
-1.32035279e+00 2.34855220e-01 1.06009257e+00 1.05986094e+00
3.90882909e-01 6.43626750e-01 -1.33079112e-01 6.74088418e-01
3.40621680e-01 6.41651094e-01 4.24351364e-01 -1.32925200e+00
4.04050976e-01 -2.00985745e-01 -2.93134242e-01 -6.46268725e-01
-2.40500912e-01 -1.00954294e+00 -1.55981922e+00 3.04902375e-01
4.82969820e-01 -4.75847214e-01 -7.40445137e-01 1.77358520e+00
-3.38792205e-01 5.10435283e-01 2.45835453e-01 9.42161918e-01
5.84181249e-01 5.28482258e-01 -2.26682037e-01 1.79189574e-02
1.13122511e+00 -7.44955897e-01 -7.61688352e-01 3.63777354e-02
3.11187774e-01 -4.43448126e-01 9.28307593e-01 9.84281361e-01
-1.00602520e+00 -3.70534658e-01 -1.31271470e+00 -1.36990706e-02
3.85259613e-02 4.00397390e-01 6.90904260e-01 7.64021993e-01
-9.64502990e-01 1.16299522e+00 -1.10491395e+00 1.53280139e-01
1.01046264e+00 6.18745625e-01 2.94515789e-02 3.19414251e-02
-9.13429201e-01 6.15025342e-01 4.07078743e-01 1.80364370e-01
-1.49607682e+00 -1.04330361e+00 -5.26334405e-01 3.18520188e-01
3.09748799e-01 -6.63395464e-01 1.25672245e+00 -9.18945968e-01
-1.61484087e+00 5.33072174e-01 6.03399910e-02 -1.10132718e+00
5.04073858e-01 -2.70232022e-01 -1.33209407e-01 8.57933089e-02
-4.17762309e-01 7.04337239e-01 1.15258861e+00 -6.78244650e-01
-2.52732962e-01 -2.97004431e-01 -1.25894830e-01 -2.50936300e-01
-5.52076399e-01 -4.51125324e-01 3.95352617e-02 -7.42292166e-01
4.64987643e-02 -1.01701963e+00 -1.78165138e-01 2.31337026e-01
-5.02235532e-01 -1.89523414e-01 6.51815832e-01 -3.08857203e-01
6.62208617e-01 -2.46099687e+00 4.63802695e-01 1.54202297e-01
3.40573490e-01 2.25956932e-01 -2.40058184e-01 9.19213369e-02
-6.50999099e-02 8.87043104e-02 -2.84788340e-01 -7.38657773e-01
4.15040195e-01 1.10961035e-01 -4.73453432e-01 8.17218065e-01
3.41355741e-01 8.45334291e-01 -6.57558739e-01 -2.95752615e-01
-7.49812350e-02 8.65439117e-01 -7.47926533e-01 -1.73799857e-01
-3.56546372e-01 -1.78767759e-02 -1.20480925e-01 5.16253650e-01
3.94327581e-01 -3.52749616e-01 5.61625510e-02 -3.94844741e-01
3.03301752e-01 1.32553801e-01 -1.15187788e+00 2.00453591e+00
-2.82513231e-01 1.05757749e+00 2.59734184e-01 -1.28862071e+00
7.91243494e-01 5.90270460e-01 2.33103916e-01 -2.91104347e-01
2.69256979e-01 6.64268881e-02 1.99179396e-01 -1.79505825e-01
2.08503976e-01 -4.02406693e-01 4.81147282e-02 2.31757104e-01
4.16757017e-01 -1.67965889e-01 -4.84537967e-02 1.25124514e-01
1.04922152e+00 -2.95407087e-01 -1.66023090e-01 -6.48908079e-01
3.01303901e-02 -2.79659450e-01 3.62643242e-01 9.74014044e-01
-1.76191673e-01 6.47973776e-01 8.21010232e-01 -1.57667443e-01
-1.25991344e+00 -1.28090203e+00 -2.65859157e-01 9.30019438e-01
-4.11255747e-01 -1.51586562e-01 -7.02104330e-01 -2.03578010e-01
-1.33732781e-01 6.11396968e-01 -6.00132883e-01 -2.25494549e-01
-2.33541310e-01 -6.84232175e-01 9.48398948e-01 4.32605624e-01
5.37911475e-01 -1.09423125e+00 -5.12789309e-01 -2.59595364e-02
3.18365425e-01 -1.20805490e+00 -1.75094336e-01 6.49444997e-01
-1.08475661e+00 -7.47384846e-01 -8.14195216e-01 -9.97474313e-01
2.07433954e-01 -1.91205844e-01 1.16968918e+00 -5.17516732e-01
-1.05779976e-01 1.80642992e-01 7.43911862e-02 -5.70994794e-01
-6.04908206e-02 5.81080139e-01 3.64126265e-01 4.31945175e-02
7.52892122e-02 -1.09329259e+00 -3.08747649e-01 -1.52082071e-01
-1.06977463e+00 -1.05557188e-01 7.87841439e-01 1.04529619e+00
6.57479048e-01 1.70614600e-01 8.31743658e-01 -7.56540120e-01
5.26389778e-01 -6.39074862e-01 -7.36880958e-01 -1.08560219e-01
-4.16308641e-01 3.86928320e-01 8.13257039e-01 -9.42443609e-01
-5.57547927e-01 2.56545097e-01 -5.60826994e-02 -1.01289594e+00
1.21891655e-01 7.52581298e-01 6.18197545e-02 1.73550427e-01
6.84193373e-01 -2.43457723e-02 -1.22686371e-01 -4.25028682e-01
3.92525226e-01 4.03518707e-01 7.60673463e-01 -6.09362602e-01
7.83310354e-01 5.04522979e-01 3.28680128e-01 -9.92735922e-01
-9.61411834e-01 1.14045769e-01 -4.57866400e-01 2.93260477e-02
7.52091050e-01 -1.12804377e+00 -1.02677011e+00 3.47298265e-01
-1.10614717e+00 -5.37172616e-01 -5.81061959e-01 9.00824547e-01
-6.24421418e-01 -1.76315054e-01 -7.38053858e-01 -8.36554229e-01
-3.46802056e-01 -1.07313132e+00 7.93130457e-01 7.43966177e-02
6.73113912e-02 -8.06549668e-01 -6.53052479e-02 -2.61325210e-01
5.24513006e-01 5.55015635e-03 1.04432952e+00 -5.34842193e-01
-6.65045202e-01 6.73911721e-02 -3.85565072e-01 6.59004271e-01
-5.31311095e-01 -2.23874107e-01 -1.28155982e+00 -5.27825058e-01
1.95224628e-01 -7.49199808e-01 1.30376375e+00 6.39058948e-01
1.44473743e+00 -2.13809803e-01 2.65377671e-01 1.09529185e+00
1.39775121e+00 -1.09653264e-01 4.99928713e-01 -1.36284484e-02
6.98898554e-01 3.90482843e-01 3.92280333e-02 5.32476664e-01
-2.40701765e-01 3.33505392e-01 7.64524698e-01 2.35387787e-01
-2.95193046e-02 -9.01480839e-02 2.81736821e-01 1.03883386e+00
1.99742153e-01 4.78460751e-02 -7.04535842e-01 5.67387104e-01
-1.90680850e+00 -8.68109286e-01 2.00117648e-01 1.99559057e+00
8.69116783e-01 3.40483278e-01 -5.13833500e-02 3.06712151e-01
3.74356508e-01 9.89887938e-02 -8.87635648e-01 -3.82038653e-02
-1.75747916e-01 4.38954592e-01 5.65781891e-01 2.08006099e-01
-9.77358401e-01 6.60559773e-01 5.94760799e+00 1.17112064e+00
-1.27871287e+00 3.46458226e-01 9.11647141e-01 -7.38557756e-01
-7.32178390e-02 -4.78695869e-01 -6.95246637e-01 1.77046970e-01
1.53990734e+00 -9.33921859e-02 7.74182022e-01 9.10514772e-01
7.88105372e-03 3.09318244e-01 -1.54559302e+00 1.61250567e+00
-9.29196551e-02 -1.60765469e+00 -2.41989136e-01 1.21922433e-01
8.09657812e-01 5.02060711e-01 5.97364545e-01 4.80235904e-01
2.50307947e-01 -1.64326406e+00 9.22723353e-01 5.08173645e-01
1.01953053e+00 -9.09026921e-01 6.04772210e-01 2.34638512e-01
-7.80295372e-01 -1.39595037e-02 -8.49608123e-01 -2.67176609e-02
-7.25287423e-02 8.39932024e-01 -5.95480084e-01 -3.67039889e-02
7.99675405e-01 1.15979624e+00 -2.76857823e-01 1.08791208e+00
1.04607798e-01 1.23593926e+00 -4.97713387e-01 -1.83961242e-01
4.76192772e-01 -2.73746997e-01 5.09392321e-01 1.09004653e+00
3.44921738e-01 -1.84172243e-01 -4.87777054e-01 1.35198987e+00
-6.13432586e-01 -5.21192014e-01 -5.57592332e-01 -4.06151772e-01
4.75807458e-01 1.01180804e+00 -6.58142984e-01 -3.23769033e-01
-9.98794809e-02 6.30870223e-01 3.32496077e-01 7.00190425e-01
-8.47391129e-01 -4.16763574e-01 6.74265265e-01 -1.12308472e-01
6.50293112e-01 -3.93955767e-01 -5.34976304e-01 -9.85223591e-01
-1.18488890e-04 -8.49309444e-01 -1.48737982e-01 -4.99614805e-01
-1.31417167e+00 6.27263010e-01 8.70578140e-02 -1.04860246e+00
-3.19539428e-01 -7.72616506e-01 -2.51657486e-01 7.04653263e-01
-1.20213163e+00 -5.67969441e-01 9.45373774e-02 5.32781899e-01
3.97243351e-01 -6.00883722e-01 5.60177565e-01 5.21010935e-01
-5.46041012e-01 8.31685424e-01 7.07800984e-01 2.64150560e-01
2.67600745e-01 -9.77727532e-01 3.14510167e-01 3.85604024e-01
5.99273920e-01 4.39373046e-01 7.13366687e-01 -7.08665922e-02
-1.67017436e+00 -1.26999784e+00 2.62343168e-01 -5.75461537e-02
8.46590638e-01 -5.90556502e-01 -8.84687126e-01 6.92215264e-01
2.28044614e-01 4.02127057e-02 1.83434874e-01 1.66789845e-01
-6.03114069e-01 -3.14800978e-01 -1.04880965e+00 4.47509557e-01
8.62036049e-01 -8.11384976e-01 -4.12713468e-01 2.54704088e-01
9.06733572e-01 2.32690126e-02 -7.61465311e-01 4.02381897e-01
7.08675146e-01 -7.47700095e-01 9.80465591e-01 -5.56430936e-01
4.98550087e-01 1.49463102e-01 -5.51054358e-01 -1.22268057e+00
-2.30713971e-02 -7.65168011e-01 -3.09802681e-01 8.22367132e-01
4.95105773e-01 -3.72938335e-01 1.11360753e+00 -7.53904954e-02
-2.66924888e-01 -9.10812736e-01 -1.43731976e+00 -8.44871223e-01
3.64734799e-01 -7.09201455e-01 5.79951927e-02 8.11877012e-01
-3.00140262e-01 5.10197699e-01 -5.12957036e-01 9.89776254e-02
9.67343569e-01 -4.97581095e-01 3.20351630e-01 -1.30152273e+00
-6.72919393e-01 -6.61150217e-01 -4.35358137e-01 -9.54820395e-01
4.83420074e-01 -9.84768093e-01 5.29294554e-03 -8.66238356e-01
2.73706675e-01 -2.83830196e-01 -4.12849545e-01 3.03904384e-01
6.61078274e-01 4.98469412e-01 4.19692814e-01 1.53108239e-01
-7.72049963e-01 9.55631495e-01 7.15726137e-01 -4.36757594e-01
2.81987414e-02 -3.28497626e-02 -3.30311298e-01 6.82809234e-01
7.42054880e-01 -9.21516716e-01 -6.40742242e-01 -5.47150671e-01
4.11395371e-01 2.06991479e-01 4.14154917e-01 -1.52444613e+00
4.15015936e-01 3.54315609e-01 2.05221221e-01 -5.84068239e-01
8.20772588e-01 -7.94117332e-01 1.46058559e-01 5.81508815e-01
-7.77409315e-01 -1.44544765e-02 1.80918589e-01 7.70294487e-01
-2.86061227e-01 -5.03256023e-01 8.78574312e-01 1.61556497e-01
-1.11894928e-01 4.83383626e-01 -4.49914604e-01 3.92838538e-01
4.89303976e-01 2.28703618e-01 -1.86514989e-01 -5.59459209e-01
-7.93928981e-01 -1.50641590e-01 -2.57427618e-02 1.13562278e-01
7.17109323e-01 -1.34691811e+00 -8.47514868e-01 2.72341490e-01
-4.81170982e-01 2.05081314e-01 -3.77262235e-02 8.96722674e-01
-4.15848494e-01 5.19461095e-01 -1.40281811e-01 -8.03604841e-01
-7.87414610e-01 9.91134718e-02 5.43854296e-01 -2.63171811e-02
-4.98200774e-01 9.25145626e-01 2.00590361e-02 -1.32055506e-01
9.85446870e-01 -5.61352789e-01 3.26579213e-01 7.02982768e-03
2.68169135e-01 3.18816870e-01 5.37053458e-02 -1.64143905e-01
-1.85333446e-01 1.45864725e-01 1.90784588e-01 -5.79954982e-01
1.62986875e+00 3.34155291e-01 3.64183597e-02 9.60159838e-01
1.53287888e+00 -6.44302309e-01 -1.63359356e+00 -5.05670249e-01
-1.93248410e-02 1.24241509e-01 3.35006535e-01 -4.16412115e-01
-1.48787487e+00 1.22390544e+00 8.04702520e-01 -3.89208682e-02
8.50519061e-01 5.50027154e-02 4.86927509e-01 1.19946563e+00
3.84381562e-01 -7.44037092e-01 3.29423368e-01 6.99964941e-01
8.12046468e-01 -1.16921604e+00 -2.56022885e-02 2.50522792e-01
-4.78699505e-01 9.14598107e-01 9.41397697e-02 -6.18166268e-01
1.17932105e+00 3.21838379e-01 -2.84032047e-01 -2.10763425e-01
-1.06559002e+00 2.80586064e-01 2.71937311e-01 5.55264056e-01
2.63481557e-01 2.88201720e-02 1.26368076e-01 8.78209233e-01
-3.97728860e-01 -4.74662893e-02 5.02403975e-01 4.39460784e-01
-1.77075982e-01 -4.97765124e-01 -1.79308653e-01 8.07819605e-01
-6.92689955e-01 -4.52354372e-01 -2.23662350e-02 6.44201279e-01
-2.83571810e-01 7.67475724e-01 1.62156522e-01 -2.98275411e-01
-3.98315266e-02 -2.38264166e-02 3.92913848e-01 -4.15057242e-01
-5.80381870e-01 2.89610308e-02 -5.27346097e-02 -5.17331123e-01
-5.56490481e-01 -5.47804773e-01 -1.22541511e+00 -3.89776260e-01
1.87251583e-01 3.98161635e-02 8.90220821e-01 8.00797522e-01
1.63580194e-01 6.86888456e-01 1.01512186e-01 -1.22265542e+00
-1.24125445e+00 -1.23581350e+00 -9.96174991e-01 2.22607315e-01
6.44217312e-01 -4.66370165e-01 -5.87684393e-01 1.06794508e-02] | [8.548210144042969, 3.21889066696167] |
b340a49e-916f-47db-866d-1a763af11146 | distributed-optimization-in-distribution | 2205.09981 | null | https://arxiv.org/abs/2205.09981v1 | https://arxiv.org/pdf/2205.09981v1.pdf | Distributed Optimization in Distribution Systems with Grid-Forming and Grid-Supporting Inverters | With massive penetrations of active grid-edge technologies, distributed computing and optimization paradigm has gained significant attention to solve distribution-level optimal power flow (OPF) problems. However, the application of generic distributed optimization techniques to OPF problems leads to a very large number of macro-iterations or communication rounds among the distributed computing agents delaying the decision-making process or resulting in suboptimal solutions. Moreover, the existing distribution-level OPF problems typically model inverter-interfaced distributed energy resources (DERs) as grid-following inverters; grid-supporting and grid-forming functionalities have not been explicitly considered. The added complexities introduced by different inverter models require further attention to developing an appropriate model for new types of inverter-based DERs and computationally-tractable OPF algorithms. In this paper, we expand the distribution-level OPF model to include a combination of the grid-forming, grid-supporting, grid-following inverter-based DERs and also present the application of a domain-specific problem decomposition and distributed algorithm for the topologically radial power distribution systems to efficiently solve distribution-level OPF problem. | ['Anamika Dubey', 'Rabayet Sadnan'] | 2022-05-20 | null | null | null | null | ['distributed-optimization', 'problem-decomposition'] | ['methodology', 'miscellaneous'] | [-6.95427358e-01 -2.39049241e-01 2.88849056e-01 2.54735295e-02
-3.06923300e-01 -1.12804139e+00 2.59298861e-01 4.95340139e-01
3.80335480e-01 1.28838086e+00 -2.69566089e-01 -2.78692663e-01
-7.95097947e-01 -1.09224761e+00 -1.30044669e-01 -9.40591753e-01
-6.73677862e-01 9.40504491e-01 -2.28150666e-01 -3.80777210e-01
2.81720851e-02 8.72423470e-01 -1.39165711e+00 -3.62476051e-01
1.39055777e+00 1.13327587e+00 1.64869741e-01 3.32536310e-01
3.41356367e-01 6.41981542e-01 -1.05763268e+00 3.67510647e-01
4.89128381e-01 -5.65048814e-01 -3.49740803e-01 2.11495206e-01
-4.87350971e-01 -2.76786447e-01 1.70361668e-01 1.37003243e+00
8.30268264e-01 4.85832691e-01 5.31766534e-01 -1.93997872e+00
-5.69058299e-01 3.98668319e-01 -6.50140285e-01 4.16596174e-01
3.13075364e-01 -8.22929069e-02 8.10119152e-01 -6.97353303e-01
5.18079042e-01 7.77132392e-01 2.92653114e-01 -3.39157254e-01
-1.11074984e+00 -2.03415066e-01 2.35477239e-01 8.39389682e-01
-1.72351754e+00 1.81348607e-01 7.83006251e-01 -2.68199027e-01
1.50119162e+00 3.14649433e-01 1.33683515e+00 -3.80180836e-01
2.01633036e-01 3.38823915e-01 1.19998980e+00 -1.35929331e-01
9.84539986e-01 -3.50074917e-01 -5.89811467e-02 -1.71984702e-01
7.66563594e-01 -1.16158769e-01 -7.16909319e-02 -2.74462048e-02
3.96075398e-01 -4.41997945e-01 -5.46629786e-01 -5.11366606e-01
-5.56279540e-01 1.13312173e+00 6.45767629e-01 5.60603976e-01
-7.51006246e-01 -3.49467844e-01 2.70918638e-01 3.61027360e-01
5.64061522e-01 3.79143804e-01 -4.62458819e-01 2.74740942e-02
-1.07625020e+00 3.94624591e-01 1.12442172e+00 1.02514923e+00
6.10115945e-01 7.56881118e-01 2.69839078e-01 2.75357425e-01
9.50195417e-02 3.70524287e-01 -9.46894959e-02 -8.90525222e-01
2.99869869e-02 5.96903622e-01 4.07087177e-01 -8.85480285e-01
-5.92951298e-01 -8.52249205e-01 -1.28694212e+00 7.10203290e-01
2.12619752e-01 -5.53388655e-01 7.55951628e-02 1.19167793e+00
6.35886073e-01 -3.09607029e-01 -1.41871318e-01 1.10192037e+00
-1.10444836e-01 1.02886426e+00 -6.35665178e-01 -8.30665648e-01
1.45990181e+00 -1.11192966e+00 -1.08182478e+00 4.20063674e-01
5.72118342e-01 -6.04927540e-01 -7.03823268e-02 6.83534443e-01
-1.51133716e+00 -1.08589880e-01 -1.19334435e+00 2.22205222e-01
-7.17450917e-01 -2.58472741e-01 1.71454966e-01 5.15891910e-01
-1.55015552e+00 3.89664799e-01 -2.59816498e-01 -1.37222037e-01
3.71105224e-01 4.43393677e-01 1.50416186e-02 -1.69451743e-01
-9.90235567e-01 1.52559900e+00 4.77372020e-01 5.20959198e-01
-7.50482261e-01 -8.95052135e-01 -4.75406349e-01 5.23021877e-01
4.67099041e-01 -9.92559910e-01 7.93542147e-01 -8.53420794e-01
-1.63004255e+00 -2.46607795e-01 1.13317646e-01 -5.31435728e-01
4.19807941e-01 4.46459591e-01 -3.14366311e-01 1.25899866e-01
1.92435402e-02 -2.55763024e-01 4.47735131e-01 -9.68229771e-01
-9.07291710e-01 -3.08495671e-01 5.00133969e-02 5.77754378e-01
-8.89749900e-02 -1.47032393e-02 6.84607029e-01 -5.58192372e-01
-3.52523834e-01 -2.37877950e-01 -6.39857531e-01 -3.74225587e-01
-3.77019197e-02 -5.98650336e-01 1.17821908e+00 -7.97426820e-01
1.15498281e+00 -1.71479976e+00 5.81716180e-01 4.31150079e-01
9.60625485e-02 -1.31063774e-01 1.20744936e-01 1.06308639e+00
-2.71050960e-01 -1.39584243e-01 7.89074078e-02 -8.98560975e-03
5.25380433e-01 5.90555966e-01 1.99876562e-01 7.33249784e-01
-2.48820886e-01 5.54900825e-01 -9.28091884e-01 9.08424035e-02
5.30382156e-01 2.87209064e-01 -5.91208220e-01 -4.11599018e-02
-2.40974456e-01 4.50471580e-01 -3.98598313e-01 5.42298496e-01
1.02512240e+00 -1.86094820e-01 3.81545812e-01 -3.77758205e-01
-6.05530083e-01 -1.37546003e-01 -1.53989148e+00 1.56485093e+00
-4.98916000e-01 2.75213778e-01 9.51518714e-01 -1.68579209e+00
4.45149809e-01 4.04469937e-01 8.51284802e-01 -6.76242769e-01
4.56383862e-02 3.86790991e-01 2.53191978e-01 1.59154460e-01
2.55955338e-01 1.96549013e-01 8.19981694e-02 5.78852475e-01
1.87941477e-01 -4.26413655e-01 9.35257077e-01 3.49363163e-02
8.74318182e-01 -2.90685683e-01 4.29449052e-01 -1.16895998e+00
8.40387285e-01 1.50909528e-01 8.67890596e-01 8.79474729e-02
-1.69535615e-02 -1.26030758e-01 5.20974815e-01 -2.31894121e-01
-8.64118814e-01 -8.98223996e-01 -2.19586417e-01 3.90801013e-01
2.08932281e-01 -3.48539323e-01 -6.32215083e-01 -4.11144257e-01
1.04829304e-01 1.20768321e+00 2.36239061e-02 4.14533466e-01
-4.14045751e-01 -1.19068158e+00 -4.09605056e-01 1.56977266e-01
3.42685610e-01 -2.10716516e-01 -3.82665813e-01 6.27561271e-01
1.70308992e-01 -9.03972447e-01 -2.14886606e-01 4.24480706e-01
-5.30701160e-01 -1.14264643e+00 -9.10729647e-01 -8.65095675e-01
1.11599934e+00 -2.75104046e-01 1.27987659e+00 -6.89048916e-02
-2.57610738e-01 1.47801891e-01 -1.66874900e-01 -2.74354845e-01
1.57728493e-01 -1.68382689e-01 1.39702052e-01 -1.93736583e-01
-3.99330765e-01 -9.35094357e-01 -7.15264142e-01 3.60917628e-01
-4.79807466e-01 -6.97678998e-02 -1.72643960e-01 7.47690976e-01
4.43488389e-01 1.35705316e+00 1.23907268e+00 1.28577039e-01
8.23273599e-01 -7.79356241e-01 -1.52490652e+00 1.48226187e-01
-6.78340495e-01 -4.65400547e-01 1.24750912e+00 -1.01232737e-01
-7.01664150e-01 -3.53662103e-01 1.28508016e-01 -1.68965250e-01
2.59910405e-01 7.53718615e-01 -4.88082975e-01 -3.88165593e-01
-2.31695578e-01 8.58448818e-02 -1.34938031e-01 -3.67714614e-01
3.44979733e-01 2.33581498e-01 -9.36370566e-02 -5.69027662e-01
9.65882063e-01 1.20281592e-01 5.09668231e-01 -4.82481420e-01
-1.64648876e-01 9.02100950e-02 -1.64180547e-01 -3.98224086e-01
5.95592082e-01 -8.24721873e-01 -1.14568317e+00 5.47989845e-01
-1.34822202e+00 -1.83442488e-01 -8.73116314e-01 1.49919495e-01
-3.16491663e-01 3.40866059e-01 -5.00403821e-01 -8.41103375e-01
-4.31698501e-01 -1.35144603e+00 4.05431360e-01 4.62343961e-01
2.14859322e-01 -1.34988213e+00 -1.08080223e-01 -2.19300110e-02
9.31357145e-01 5.71320593e-01 1.06402469e+00 -4.59309340e-01
-9.51481462e-01 9.73592326e-02 -2.93345395e-02 3.83703262e-01
2.18540266e-01 -1.93633392e-01 -1.30474404e-01 -9.09593880e-01
1.92534313e-01 2.17451572e-01 -7.61871517e-01 4.61377323e-01
7.86680222e-01 -4.66494590e-01 -7.49619082e-02 3.22602361e-01
2.24138451e+00 5.30330002e-01 1.62741303e-01 1.05464317e-01
2.20240310e-01 3.94469410e-01 4.46812779e-01 7.77924955e-01
8.17855000e-01 7.75202274e-01 6.66671693e-01 -1.27165645e-01
2.69428849e-01 4.41079319e-01 1.06444150e-01 1.08811390e+00
-2.46430725e-01 -6.55719519e-01 -5.44560552e-01 6.75589204e-01
-1.96961606e+00 -6.38370931e-01 -2.34031156e-01 1.87336302e+00
2.33916864e-01 -4.09660965e-01 1.68666705e-01 3.90477747e-01
9.09437835e-01 -2.61449754e-01 -4.89190400e-01 -6.22094631e-01
-3.97606254e-01 3.96661788e-01 4.60334688e-01 4.49623644e-01
-3.38169008e-01 -4.91262563e-02 5.25104046e+00 1.10769904e+00
-7.14430273e-01 4.91376102e-01 5.48847258e-01 -2.24371940e-01
-2.48808593e-01 1.28743112e-01 -3.14060658e-01 8.87278557e-01
1.05747294e+00 -1.20091820e+00 1.21971941e+00 4.58319187e-01
9.08602476e-01 -5.47450721e-01 -9.82807577e-01 1.01681817e+00
-3.13943893e-01 -1.13778400e+00 -4.16952670e-01 4.22401011e-01
1.45243311e+00 1.63659409e-01 -7.32311904e-01 -1.77656502e-01
5.14019191e-01 -6.73623562e-01 8.70784521e-01 2.32001781e-01
5.31940982e-02 -1.22992682e+00 8.59669983e-01 3.38319957e-01
-1.44682407e+00 -4.30593550e-01 -2.73605809e-02 -5.46872020e-01
1.06847835e+00 1.16186726e+00 -8.66706446e-02 1.50971723e+00
9.52524245e-01 7.88231254e-01 7.01427311e-02 1.21774459e+00
-3.43048960e-01 1.50332853e-01 -7.82325149e-01 2.26692460e-03
1.32276744e-01 -9.66282904e-01 4.96795595e-01 2.92984009e-01
8.40488017e-01 8.92426372e-02 4.60888147e-01 9.27904010e-01
2.49010310e-01 -1.13504450e-03 -2.78072685e-01 1.65307358e-01
7.03496099e-01 1.70221734e+00 -9.62574601e-01 -3.87589335e-01
-4.73570853e-01 4.45482135e-01 -1.94404975e-01 6.86575949e-01
-6.94851696e-01 -4.29877698e-01 9.07942772e-01 -1.06917374e-01
3.66874486e-01 -3.76433730e-01 -5.17330050e-01 -1.04606616e+00
-1.58255324e-01 -7.44250417e-01 4.10404027e-01 -1.00632370e+00
-1.74751019e+00 2.63626009e-01 1.31689936e-01 -9.96649384e-01
-5.00907719e-01 -3.43865067e-01 -9.82716858e-01 1.38887060e+00
-1.88177109e+00 -7.81398594e-01 -7.43267536e-02 7.82673538e-01
4.78024960e-01 -8.02125856e-02 7.60514557e-01 7.85206020e-01
-8.53017628e-01 -2.01058403e-01 5.97228050e-01 -3.37467581e-01
-3.25144738e-01 -1.49478209e+00 -5.42866290e-01 1.18097031e+00
-7.04140484e-01 -1.28005132e-01 7.82905638e-01 -2.66524017e-01
-1.99410164e+00 -6.03785872e-01 7.52035618e-01 5.92233956e-01
1.13182354e+00 -3.28710437e-01 -3.92543614e-01 4.42864299e-01
1.45387292e+00 1.26688525e-01 3.47593844e-01 -4.22959536e-01
7.36028850e-01 -4.40549433e-01 -1.65340054e+00 2.40195736e-01
5.88820815e-01 -1.77376673e-01 -3.18677813e-01 6.15422308e-01
-2.02277675e-01 -1.55704230e-01 -1.36519015e+00 2.30157316e-01
-4.37840790e-01 -4.23601568e-01 8.08290541e-01 2.44345382e-01
-3.33214164e-01 -1.07102251e+00 -8.01742971e-02 -2.09704304e+00
-3.72760028e-01 -8.54726434e-01 -4.90950167e-01 1.20289421e+00
-1.69876903e-01 -1.58365798e+00 2.74041981e-01 6.35643816e-03
-4.54644173e-01 -6.04872406e-01 -1.59893763e+00 -8.06857705e-01
2.87639767e-01 2.13486716e-01 1.19624674e+00 1.13789320e+00
4.73319381e-01 2.51565427e-01 4.15302753e-01 7.22576976e-01
1.02865803e+00 4.55945820e-01 1.17628597e-01 -9.72032785e-01
-1.37784153e-01 -6.51078284e-01 -2.83114076e-01 -5.61060846e-01
-1.20380558e-01 -7.44408429e-01 -4.86460775e-01 -2.02814579e+00
-7.12021172e-01 -3.71927410e-01 4.89085689e-02 2.58584172e-01
5.07848918e-01 1.94222052e-02 5.42212427e-01 -7.02619702e-02
-3.31740946e-01 6.95273936e-01 1.19624221e+00 -1.86080709e-01
4.36713696e-01 -3.36463094e-01 -2.32176930e-01 4.71193284e-01
8.52438629e-01 -3.29548419e-02 -7.51888156e-01 -4.09976691e-01
5.57615995e-01 5.21948338e-01 4.31162603e-02 -7.79964507e-01
6.47119105e-01 3.48832365e-03 8.75797421e-02 -1.00601232e+00
6.13873899e-02 -1.46622264e+00 8.31440389e-01 6.80607557e-01
7.87518263e-01 6.57591999e-01 -1.33092448e-01 -2.59463191e-02
-5.84451914e-01 -3.03932011e-01 5.53248167e-01 -3.45875286e-02
-5.03553867e-01 1.01387493e-01 -1.13481784e+00 -1.00193821e-01
1.78455901e+00 -8.30182955e-02 -4.54630941e-01 -3.32580090e-01
-1.05158699e+00 1.01746953e+00 4.22736526e-01 4.22613099e-02
-8.23902115e-02 -1.12069142e+00 -8.04106414e-01 6.36266991e-02
-8.41793180e-01 1.87857836e-01 2.26928711e-01 9.83954549e-01
-8.06915224e-01 3.61574680e-01 -1.78757161e-01 -5.46879411e-01
-4.54369694e-01 5.76260805e-01 7.07942247e-01 -5.42098522e-01
-2.82249957e-01 9.22351032e-02 -3.97448212e-01 -3.35430913e-02
-4.19434667e-01 -2.98892796e-01 7.42641166e-02 7.65712440e-01
9.52044502e-02 1.25430191e+00 4.77319390e-01 -2.72985131e-01
-3.42337042e-01 2.99709350e-01 5.90767980e-01 2.18412444e-01
1.60654926e+00 -6.34904206e-01 -8.05111885e-01 -1.97076932e-01
7.91660488e-01 -3.01230758e-01 -6.91525578e-01 5.41279316e-01
-1.70772761e-01 -4.63063223e-03 6.15064859e-01 -1.13791919e+00
-1.62578106e+00 2.89671749e-01 3.03419828e-01 1.00515056e+00
1.68412626e+00 -5.95979154e-01 3.53592068e-01 -2.83855408e-01
1.31705105e+00 -1.25553560e+00 -7.35150695e-01 5.13854623e-01
8.09865057e-01 -1.91715583e-01 1.71481296e-01 -4.58355159e-01
3.21775615e-01 1.11822760e+00 2.67259389e-01 -4.83267784e-01
9.16040599e-01 8.64628255e-01 -6.36139810e-01 2.09717620e-02
-6.35549366e-01 -2.70504486e-02 -2.11212009e-01 6.57997549e-01
-9.65760574e-02 6.10624850e-02 -1.03109133e+00 5.34806788e-01
-1.77024916e-01 3.08395535e-01 6.54011786e-01 1.16227615e+00
1.27876535e-01 -1.19721508e+00 -6.55906081e-01 4.07045960e-01
-3.05026293e-01 1.00893997e-01 5.98124206e-01 6.58404887e-01
4.72642571e-01 1.12740445e+00 4.31812346e-01 7.03135014e-01
4.03740972e-01 -2.79881001e-01 5.66429973e-01 -2.37250239e-01
-8.89324546e-01 2.60003269e-01 1.04142465e-01 -1.91650838e-01
-2.31385425e-01 -8.65027487e-01 -1.22679329e+00 -6.38309896e-01
-6.33940756e-01 8.78173351e-01 6.97181344e-01 9.35483873e-01
3.72882009e-01 7.49705970e-01 9.82167900e-01 -1.37253904e+00
-4.86648142e-01 -6.33494139e-01 -1.27342415e+00 -5.59216380e-01
-1.87372997e-01 -4.71737593e-01 -6.11766934e-01 -5.74284136e-01] | [5.663600921630859, 2.5455076694488525] |
fcabab28-193b-4436-a096-9c50d0661b98 | mining-social-science-publications-for-survey | null | null | https://aclanthology.org/W17-2907 | https://aclanthology.org/W17-2907.pdf | Mining Social Science Publications for Survey Variables | Research in Social Science is usually based on survey data where individual research questions relate to observable concepts (variables). However, due to a lack of standards for data citations a reliable identification of the variables used is often difficult. In this paper, we present a work-in-progress study that seeks to provide a solution to the variable detection task based on supervised machine learning algorithms, using a linguistic analysis pipeline to extract a rich feature set, including terminological concepts and similarity metric scores. Further, we present preliminary results on a small dataset that has been specifically designed for this task, yielding a significant increase in performance over the random baseline. | ['Peter Mutschke', 'Andrea Zielinski'] | 2017-08-01 | null | null | null | ws-2017-8 | ['variable-detection'] | ['natural-language-processing'] | [ 9.85265598e-02 3.54311280e-02 -5.90798736e-01 -4.67471033e-01
-9.11457360e-01 -6.19364679e-01 1.03737617e+00 6.96818292e-01
-4.93211627e-01 7.97947645e-01 4.85935658e-01 -5.72532117e-01
-3.39294910e-01 -4.22571361e-01 -6.41346648e-02 -1.50914699e-01
6.84059337e-02 4.81459796e-01 -3.77313524e-01 8.25356971e-03
8.98132563e-01 4.30125177e-01 -1.73433828e+00 -6.58704266e-02
6.30219698e-01 6.63865626e-01 3.35714594e-02 3.18513274e-01
-7.91184306e-01 8.11170101e-01 -6.92555547e-01 -4.21791852e-01
-1.23577192e-03 -5.45532346e-01 -1.08586764e+00 -1.64762855e-01
7.34410703e-01 5.07752538e-01 3.84773850e-01 1.13014507e+00
1.14827424e-01 5.52520119e-02 8.57175469e-01 -1.28018022e+00
-8.56698275e-01 7.87805438e-01 -3.32114756e-01 3.64029348e-01
6.81037784e-01 -1.31778568e-01 1.84598541e+00 -8.26757193e-01
1.07301855e+00 1.37448764e+00 4.49335992e-01 3.07824790e-01
-1.59416866e+00 -7.99423575e-01 1.52190894e-01 2.14305997e-01
-1.08204651e+00 -3.60052705e-01 7.92155862e-01 -8.35484982e-01
1.02159548e+00 2.58627653e-01 4.60521460e-01 1.07059932e+00
4.31702510e-02 3.22091520e-01 1.48313582e+00 -6.61861300e-01
4.34211075e-01 5.11208355e-01 6.54360652e-01 3.55577499e-01
3.21253359e-01 -1.83075964e-02 -5.38543522e-01 -2.96296835e-01
3.52623820e-01 -1.77888691e-01 1.62681967e-01 -3.55432123e-01
-1.28643930e+00 1.16456592e+00 1.29798651e-01 8.00141335e-01
-4.90051597e-01 -2.21923083e-01 1.68434203e-01 7.41314411e-01
7.72247612e-01 1.07992125e+00 -7.67234206e-01 -4.04166222e-01
-1.23199773e+00 5.70507407e-01 1.15798593e+00 6.93800151e-01
9.78049457e-01 -4.34921145e-01 -3.31441164e-01 6.71564877e-01
4.13158566e-01 3.64329457e-01 4.56039280e-01 -9.33217943e-01
4.44189489e-01 1.01647735e+00 -2.77743451e-02 -1.20027721e+00
-6.09084606e-01 -2.32360661e-01 -1.22747935e-01 1.49399653e-01
3.74481618e-01 -2.13253677e-01 -6.26494169e-01 1.58197701e+00
3.24982613e-01 -9.51732248e-02 -1.26788110e-01 6.98586524e-01
1.13707662e+00 -6.99574500e-03 3.46907616e-01 -3.15713406e-01
1.31059885e+00 -5.51006377e-01 -1.00956488e+00 -1.47269815e-01
8.22920263e-01 -6.93234563e-01 1.09582901e+00 2.56069601e-01
-7.89260745e-01 -2.34686568e-01 -5.97219944e-01 -2.61848569e-01
-7.51258671e-01 -3.83603334e-01 8.52366090e-01 7.41966248e-01
-1.18785882e+00 1.62512153e-01 -4.19164598e-01 -6.72955513e-01
4.78783011e-01 2.98655126e-02 -5.13475180e-01 3.87354255e-01
-9.43146944e-01 9.01142538e-01 1.37852222e-01 -6.05309784e-01
-2.24984765e-01 -8.19870710e-01 -8.74059975e-01 -7.57216364e-02
3.81441534e-01 -4.73743260e-01 9.69916403e-01 -8.73328447e-01
-1.13960767e+00 1.08604097e+00 -4.61212426e-01 -2.54657030e-01
7.41337836e-02 8.58898908e-02 -4.67322588e-01 1.76847577e-02
6.17317259e-01 3.42409700e-01 4.56975877e-01 -1.08127201e+00
-7.80669034e-01 -5.26468992e-01 -1.34555280e-01 -1.85072005e-01
-7.18511224e-01 7.64393866e-01 2.54270844e-02 -7.35346138e-01
-5.73927239e-02 -7.79772758e-01 -3.67510378e-01 -3.47599864e-01
7.36460090e-02 -5.51909566e-01 4.22499955e-01 -5.82400739e-01
1.41889799e+00 -1.66518509e+00 3.47880185e-01 3.03741544e-01
5.75347066e-01 -1.87304050e-01 -1.04136281e-02 3.52633178e-01
-2.75774837e-01 4.79992896e-01 -2.43018791e-01 -4.37969863e-01
-9.49011445e-02 -2.92927958e-02 -1.58038706e-01 4.95929569e-01
3.81227911e-01 1.01739073e+00 -1.09803772e+00 -8.81182075e-01
2.27128983e-01 -1.65724792e-02 -5.22022784e-01 7.16025084e-02
-2.49814585e-01 1.19795680e-01 -6.97881699e-01 8.38872194e-01
1.45719454e-01 -5.86061850e-02 4.75014513e-03 4.42746937e-01
-5.24510860e-01 6.30121112e-01 -1.03029072e+00 1.50200343e+00
-3.72572124e-01 1.05706739e+00 1.63817555e-01 -1.36873353e+00
1.08248758e+00 2.23356292e-01 7.55183816e-01 -6.04430497e-01
3.34502280e-01 1.97921261e-01 -2.02080458e-01 -7.29931772e-01
5.65002263e-01 -2.02492565e-01 -2.78875470e-01 4.03005302e-01
9.95382071e-02 -3.32738131e-01 3.31745863e-01 1.49330348e-01
1.20150793e+00 -2.43731719e-02 4.91314948e-01 -6.58560574e-01
6.67784274e-01 3.69717747e-01 5.04025877e-01 7.31440187e-01
-3.82316351e-01 5.03262639e-01 6.10281587e-01 -1.19790830e-01
-6.92360520e-01 -5.67911506e-01 -4.01859522e-01 1.00733209e+00
-2.84122229e-01 -7.25040257e-01 -5.53856850e-01 -6.99509740e-01
2.33362123e-01 9.14916992e-01 -9.39838529e-01 3.10365409e-01
-1.60258576e-01 -4.35304403e-01 8.19893777e-02 1.63222671e-01
-2.45578274e-01 -1.13269734e+00 -4.77556348e-01 1.45906970e-01
1.56820059e-01 -1.04128718e+00 2.49413878e-01 6.00560829e-02
-8.52778673e-01 -9.88539100e-01 -2.90670455e-01 -7.98276424e-01
2.89944768e-01 3.45940471e-01 1.48795652e+00 6.42585456e-02
-1.92526698e-01 2.46884063e-01 -4.91318882e-01 -8.08184922e-01
-1.33721709e-01 4.32336301e-01 -2.44982123e-01 -4.26719666e-01
1.34605634e+00 -2.51251698e-01 1.83031484e-02 -1.77344024e-01
-6.49084687e-01 -4.85193968e-01 4.84300435e-01 5.94157577e-01
1.67582422e-01 -3.69568229e-01 6.59319878e-01 -1.09267724e+00
9.10512567e-01 -1.05412376e+00 -5.66989899e-01 1.40108522e-02
-1.00481617e+00 9.12397355e-02 3.46335232e-01 -3.06431830e-01
-7.50003457e-01 -4.45608556e-01 1.55554652e-01 7.51855746e-02
-4.96533811e-01 1.18495524e+00 1.79654539e-01 -3.27635929e-02
8.02919507e-01 -5.58429718e-01 2.59384904e-02 -6.29321098e-01
4.48426872e-01 1.10406816e+00 1.83436304e-01 -6.13563001e-01
8.45298171e-01 1.84297919e-01 4.92378585e-02 -9.93966937e-01
-1.01684678e+00 -7.54253626e-01 -5.93690515e-01 -6.50867224e-02
6.21783912e-01 -8.25467706e-01 -3.47617805e-01 -1.69242352e-01
-7.69139409e-01 5.90926819e-02 -2.25169897e-01 8.50263596e-01
-1.84464782e-01 1.08721226e-01 -3.34161110e-02 -7.99310923e-01
-1.25086755e-01 -9.47639465e-01 9.33821082e-01 9.37011838e-02
-8.90138745e-01 -1.48947287e+00 4.01147127e-01 5.90121746e-01
4.61604744e-01 3.42158377e-01 6.64271116e-01 -9.07144964e-01
-9.50136483e-02 -1.74789563e-01 -1.10564619e-01 -1.45334959e-01
1.07455015e-01 2.06647992e-01 -6.40436769e-01 -7.67288655e-02
4.47754413e-02 -2.15909421e-01 7.86594152e-01 4.03300345e-01
1.09824920e+00 1.02432936e-01 -3.04857373e-01 3.52543056e-01
1.28580868e+00 -2.69616306e-01 1.70697272e-01 9.65798140e-01
6.17561340e-01 9.76068377e-01 6.26575351e-01 2.94393063e-01
7.37664342e-01 5.09639442e-01 -5.55545129e-02 1.13009643e-02
2.92050168e-02 2.19916046e-01 1.42362580e-01 5.58792949e-01
1.82659402e-01 1.51241764e-01 -1.33204210e+00 1.03469157e+00
-1.66015482e+00 -9.67508078e-01 -6.74799442e-01 1.85529745e+00
9.38198388e-01 1.54089212e-01 1.85826913e-01 1.49400365e-02
5.65630317e-01 1.34799823e-01 -1.15497157e-01 -6.73003197e-01
-1.59533769e-01 5.69244921e-01 5.22171855e-01 4.72586364e-01
-8.47715616e-01 1.06765878e+00 7.71496058e+00 3.69134933e-01
-9.48710501e-01 -5.38158342e-02 6.90876693e-02 -1.76287591e-01
-7.47370005e-01 3.05989623e-01 -6.52934253e-01 3.99643183e-01
1.29002500e+00 -4.89161640e-01 3.06434989e-01 5.71112990e-01
2.73574293e-01 -1.30374134e-01 -1.06570137e+00 7.98161745e-01
1.46572381e-01 -1.20110452e+00 -1.33886874e-01 1.05652079e-01
7.74643600e-01 6.56512827e-02 -7.31337965e-02 3.17024946e-01
4.77542460e-01 -1.12405694e+00 7.92051435e-01 4.47123408e-01
5.75511813e-01 -3.88515264e-01 5.95515370e-01 2.33825594e-02
-7.38599062e-01 -1.80372775e-01 -1.17599294e-02 -7.25737572e-01
-2.02769384e-01 8.10720623e-01 -7.92822123e-01 3.89088094e-01
7.16360331e-01 1.11308241e+00 -7.98133016e-01 8.45408738e-01
-3.23043168e-01 1.11287236e+00 8.70075449e-02 -5.29855132e-01
1.45127580e-01 -2.69550204e-01 6.77612662e-01 1.36519766e+00
2.09007040e-02 -1.66303217e-01 8.13385099e-03 1.15993989e+00
-9.15771723e-02 4.11930770e-01 -5.34743786e-01 -4.49544042e-01
6.69315875e-01 1.45247769e+00 -6.08240008e-01 -1.03338197e-01
-1.02845299e+00 1.14854351e-01 7.28379011e-01 3.26134443e-01
-1.97043568e-01 -4.83653843e-01 9.33000088e-01 -2.70962119e-02
1.53496653e-01 -3.25585961e-01 -9.21812952e-01 -1.29549623e+00
1.06572229e-02 -9.60721791e-01 4.54869777e-01 -3.70173484e-01
-1.44753194e+00 3.34628820e-02 1.06135048e-02 -5.05423903e-01
-4.68225956e-01 -4.82331455e-01 -3.13302934e-01 1.29205012e+00
-1.41715837e+00 -9.12644863e-01 -1.94429189e-01 3.10860991e-01
2.31591582e-01 -5.17403364e-01 9.40206885e-01 1.48473918e-01
-6.52907431e-01 4.89187062e-01 1.81350529e-01 -4.19031754e-02
8.65168095e-01 -1.53825021e+00 4.33333039e-01 6.88259304e-01
2.31112525e-01 9.85750616e-01 1.09305358e+00 -7.31991768e-01
-1.35653973e+00 -6.50818110e-01 1.53246975e+00 -9.19440866e-01
1.31552458e+00 -3.48392308e-01 -9.67605233e-01 7.59520650e-01
1.45935237e-01 -4.86056268e-01 1.09386492e+00 9.70364869e-01
-2.16154680e-01 3.41772974e-01 -1.21543849e+00 4.06798542e-01
9.20860589e-01 -7.58068264e-01 -1.22687268e+00 -2.62387767e-02
4.18575168e-01 7.87889585e-02 -1.17691326e+00 -6.44472688e-02
3.80734473e-01 -4.71351355e-01 6.54090226e-01 -8.29164267e-01
7.32469320e-01 -1.44744571e-02 2.12450065e-02 -1.29631686e+00
-5.37152052e-01 -5.04070044e-01 1.04657061e-01 1.70333517e+00
5.20593882e-01 -6.40824914e-01 7.30298400e-01 1.03097844e+00
2.49714330e-01 -2.90169209e-01 -8.96327019e-01 -3.73506963e-01
5.87358356e-01 -5.26276946e-01 6.35931790e-01 1.56110668e+00
2.99336970e-01 7.07566738e-01 2.86159426e-01 -2.46939123e-01
5.61898589e-01 3.47196519e-01 7.28462160e-01 -1.97921157e+00
3.80608439e-01 -8.67194653e-01 -7.24885941e-01 -2.17905492e-01
7.73411393e-01 -1.08061194e+00 -2.02786699e-01 -1.61158180e+00
3.95467520e-01 -3.83814871e-01 -3.87615174e-01 1.87731653e-01
-6.15323663e-01 -7.55872875e-02 -4.27383855e-02 1.20156661e-01
-4.80922997e-01 1.89742893e-01 6.81334376e-01 5.66365346e-02
-4.19864655e-01 -3.66507858e-01 -1.36164331e+00 3.08776587e-01
7.76573539e-01 -6.24241054e-01 1.68239940e-02 -4.48006779e-01
3.54716241e-01 -2.83787906e-01 2.77248591e-01 -5.96143067e-01
1.05025202e-01 -6.40239537e-01 -3.45733501e-02 -3.13366532e-01
-2.79930055e-01 -6.69096589e-01 1.90141238e-02 -1.37598798e-01
-8.42400193e-01 3.95065621e-02 6.94277789e-03 1.85100690e-01
-3.07414830e-01 -4.01793778e-01 2.66816914e-01 -1.06100217e-01
-6.44495487e-01 -4.36855704e-02 -3.51942182e-01 3.63529474e-01
9.72026646e-01 -6.42044246e-02 -1.86337292e-01 -5.14117926e-02
-2.21140414e-01 3.26353431e-01 6.88596368e-01 7.21585870e-01
1.22484639e-01 -1.09705150e+00 -1.13482308e+00 -4.37919460e-02
3.49047303e-01 -3.72853696e-01 -5.03125548e-01 5.10092378e-01
-1.60737097e-01 6.36463404e-01 -2.00407565e-01 -2.03819215e-01
-1.23500586e+00 8.53617311e-01 -9.70775411e-02 -7.99513757e-02
-4.64518756e-01 5.72199464e-01 -4.50851142e-01 -4.11399513e-01
1.10859923e-01 -2.17835665e-01 -7.93958306e-01 6.95237935e-01
4.97123778e-01 4.84363168e-01 7.91135430e-03 -9.05773342e-01
-4.83320653e-01 2.62567163e-01 8.48891437e-02 -4.03582871e-01
1.69887435e+00 -2.86654621e-01 -4.06829715e-01 8.02950740e-01
1.46626627e+00 1.84791610e-01 -3.54676157e-01 -5.12526870e-01
6.90528214e-01 -7.01731443e-01 7.35833764e-01 -6.60492718e-01
-8.50673258e-01 5.20010054e-01 3.17637801e-01 7.21220613e-01
6.06687844e-01 2.39598632e-01 2.16498494e-01 1.86189115e-01
1.83167771e-01 -1.32123256e+00 -3.79355311e-01 5.78048170e-01
6.91807985e-01 -1.57140279e+00 1.89975381e-01 -1.70751870e-01
-3.46004546e-01 9.47297454e-01 4.44084629e-02 1.46632090e-01
9.46135163e-01 1.09970421e-01 2.74138689e-01 -7.24700689e-01
-6.74206316e-01 -4.95616168e-01 3.11072826e-01 5.52404940e-01
1.02701652e+00 1.58659726e-01 -1.09917760e+00 5.66251636e-01
-6.50292039e-01 -2.51550693e-03 5.19342184e-01 8.46047044e-01
-3.48395258e-01 -1.23388147e+00 -4.99326766e-01 8.51659656e-01
-9.13149297e-01 -2.14965194e-01 -9.36776996e-01 6.35555267e-01
-1.42718121e-01 1.51794839e+00 1.22301072e-01 -5.79420328e-02
2.87364900e-01 -8.30812603e-02 -2.31616553e-02 -7.61483788e-01
-9.03084219e-01 -3.62462252e-01 2.14273572e-01 -3.61201763e-01
-9.87418175e-01 -1.41663396e+00 -9.48856115e-01 -2.49137670e-01
-2.53189921e-01 4.10914302e-01 9.31336522e-01 1.08676636e+00
2.66685933e-01 4.86037135e-01 5.73513508e-01 -4.33339894e-01
-3.00767452e-01 -1.11121750e+00 -4.80838537e-01 8.66072893e-01
4.12330985e-01 -9.37794805e-01 -6.44800961e-01 -5.46315918e-03] | [9.981464385986328, 8.963281631469727] |
7c9e0c29-b940-46f8-abbf-f9ed54ed40da | invariant-representations-in-deep-learning | 2305.00335 | null | https://arxiv.org/abs/2305.00335v1 | https://arxiv.org/pdf/2305.00335v1.pdf | Invariant Representations in Deep Learning for Optoacoustic Imaging | Image reconstruction in optoacoustic tomography (OAT) is a trending learning task highly dependent on measured physical magnitudes present at sensing time. The large number of different settings, and also the presence of uncertainties or partial knowledge of parameters, can lead to reconstructions algorithms that are specifically tailored and designed to a particular configuration which could not be the one that will be ultimately faced in a final practical situation. Being able to learn reconstruction algorithms that are robust to different environments (e.g. the different OAT image reconstruction settings) or invariant to such environments is highly valuable because it allows to focus on what truly matters for the application at hand and discard what are considered spurious features. In this work we explore the use of deep learning algorithms based on learning invariant and robust representations for the OAT inverse problem. In particular, we consider the application of the ANDMask scheme due to its easy adaptation to the OAT problem. Numerical experiments are conducted showing that, when out-of-distribution generalization (against variations in parameters such as the location of the sensors) is imposed, there is no degradation of the performance and, in some cases, it is even possible to achieve improvements with respect to standard deep learning approaches where invariance robustness is not explicitly considered. | ['Leonardo Rey Vega', 'Martin G. Gonzalez', 'Matias Vera'] | 2023-04-29 | null | null | null | null | ['image-reconstruction'] | ['computer-vision'] | [ 3.75920773e-01 -3.33748966e-01 3.31652254e-01 -2.03593537e-01
-5.33707142e-01 -3.89637589e-01 5.09532809e-01 -4.42646816e-02
-5.14258206e-01 9.06288266e-01 -1.04553588e-01 -1.41231596e-01
-8.88609529e-01 -5.71990669e-01 -8.05923581e-01 -1.03354478e+00
-4.59469622e-03 6.78423345e-01 2.23491773e-01 -1.72615543e-01
2.09159970e-01 1.04214430e+00 -1.66950488e+00 -1.06856875e-01
6.06207132e-01 1.14793217e+00 3.62125248e-01 4.72466618e-01
2.42292788e-02 5.76476634e-01 -5.09829521e-01 2.36828759e-01
4.73712891e-01 -1.48778468e-01 -3.40245217e-01 1.87168568e-01
1.81283593e-01 -7.59939179e-02 -6.57155644e-03 1.08112895e+00
5.16274512e-01 2.02663049e-01 7.74680912e-01 -5.14671981e-01
-1.27947852e-01 3.60627502e-01 -2.33331695e-01 9.44395885e-02
2.95749694e-01 2.04675481e-01 4.55488324e-01 -6.56617045e-01
5.23006618e-01 9.16318774e-01 5.17084241e-01 9.22384635e-02
-1.42865372e+00 -2.98598975e-01 -8.25557187e-02 2.86551923e-01
-1.24787974e+00 -4.67358083e-01 8.39242995e-01 -5.50687134e-01
4.10293102e-01 2.14129001e-01 3.13489318e-01 1.19523144e+00
1.90931693e-01 2.66877115e-01 1.45182717e+00 -6.10418200e-01
5.28900385e-01 3.72344166e-01 -3.24192196e-01 2.02403307e-01
3.38534981e-01 2.10097358e-01 -1.09152511e-01 2.84508437e-01
7.02046812e-01 -5.30048192e-01 -7.18228161e-01 -9.30394530e-01
-1.21783614e+00 5.26333690e-01 5.10486960e-01 5.44760764e-01
-4.79000986e-01 -7.65024498e-02 2.89215773e-01 6.66085839e-01
6.75807968e-02 7.52186060e-01 -3.35383803e-01 -2.46957559e-02
-8.28174710e-01 1.69322446e-01 6.84381008e-01 5.49040020e-01
6.50411963e-01 4.11557257e-01 8.82490426e-02 5.70895731e-01
-1.02905624e-01 4.62483376e-01 6.15339637e-01 -6.60140991e-01
1.32725686e-01 1.69331908e-01 4.22074795e-01 -8.59913945e-01
-5.89564085e-01 -8.55006993e-01 -7.73424327e-01 6.82575107e-01
5.57459116e-01 -1.19560100e-01 -9.49022710e-01 1.73176765e+00
3.19659948e-01 1.38272136e-01 2.64450312e-01 1.10109568e+00
4.08330232e-01 2.54627645e-01 -4.16991472e-01 -2.63119400e-01
8.09160590e-01 -4.22220379e-02 -6.27777576e-01 -2.47644216e-01
2.87733108e-01 -1.03981066e+00 9.34094489e-01 6.69583619e-01
-5.93701422e-01 -5.26762128e-01 -1.20510733e+00 3.57538015e-01
-4.40009087e-01 4.00942378e-02 2.04760104e-01 5.51606297e-01
-7.21977770e-01 9.25947905e-01 -5.38123608e-01 -4.00281131e-01
-8.20718408e-02 4.38322634e-01 -3.87381673e-01 -6.80080578e-02
-9.83697474e-01 1.32331562e+00 4.28126842e-01 3.96418810e-01
-5.97088397e-01 -5.73112786e-01 -4.00545061e-01 -2.78215483e-02
5.90693057e-01 -5.76755166e-01 9.99997497e-01 -1.04402661e+00
-1.73037183e+00 4.45398688e-01 1.73336938e-01 -4.56286609e-01
9.77588892e-01 -9.53418091e-02 -3.42408985e-01 9.97540876e-02
-7.51028135e-02 1.02599062e-01 1.23141670e+00 -1.22504711e+00
-1.33106723e-01 -1.83930039e-01 6.61297515e-02 -1.25665301e-02
1.59798369e-01 -2.93002069e-01 3.82853486e-02 -3.72708350e-01
5.49931467e-01 -1.13661969e+00 -2.47192323e-01 7.44084716e-02
-3.05839360e-01 4.08057362e-01 7.12445319e-01 -3.27442050e-01
2.73899078e-01 -2.10061240e+00 4.77083176e-01 3.37287605e-01
-3.36400032e-01 1.53911650e-01 2.44916659e-02 4.73689526e-01
-4.80490834e-01 -2.86149561e-01 -2.24330723e-01 1.05663978e-01
-9.76500362e-02 1.83096215e-01 -2.79922098e-01 7.68029988e-01
4.96776924e-02 4.17033374e-01 -4.64357257e-01 -1.59552217e-01
7.72053838e-01 4.42876428e-01 -2.71386921e-01 1.74121931e-01
-2.75228173e-01 1.09278738e+00 -5.47773838e-01 2.15193391e-01
5.20735979e-01 5.39409257e-02 -2.63715684e-02 -2.70681858e-01
-2.38467127e-01 1.02524243e-01 -1.54392695e+00 1.27080929e+00
-1.00600421e+00 6.21291280e-01 9.78526473e-02 -1.30292332e+00
9.50825691e-01 3.85922134e-01 5.86758554e-01 -8.67138088e-01
3.11448306e-01 6.76115513e-01 3.73423934e-01 -7.46767104e-01
7.61878192e-02 -6.37585163e-01 1.44531459e-01 2.08721086e-01
-8.38574860e-03 -5.73744893e-01 -1.46739930e-01 -3.59959394e-01
7.47081578e-01 2.08337829e-01 3.47785264e-01 -4.36740428e-01
6.16851091e-01 -4.02375162e-01 4.14332449e-01 6.91879630e-01
9.53117833e-02 6.99210703e-01 3.02488744e-01 -4.66886878e-01
-9.53148186e-01 -8.77142727e-01 -5.73006392e-01 3.20045382e-01
9.82431099e-02 3.91804188e-01 -3.06588799e-01 -2.32669041e-01
-1.26998544e-01 7.37696648e-01 -4.74395275e-01 -2.13450640e-01
-5.07179797e-01 -6.08017743e-01 -8.63005966e-02 -6.79662898e-02
4.29476887e-01 -9.72808838e-01 -9.25153434e-01 2.24331319e-01
-4.83229943e-02 -1.36823738e+00 3.71773362e-01 5.18929124e-01
-8.40149105e-01 -1.03266644e+00 -5.74831069e-01 -1.41805872e-01
2.82643914e-01 -6.92567602e-02 7.07354426e-01 -6.07071042e-01
-1.34748057e-01 6.45493925e-01 -3.28580618e-01 -4.47622865e-01
-5.21701038e-01 -1.34259254e-01 2.62749285e-01 3.48747730e-01
-8.71368647e-02 -1.04060090e+00 -3.81601125e-01 1.88720077e-01
-1.05767584e+00 -3.08821708e-01 7.05313206e-01 7.03704417e-01
4.66399312e-01 3.02308112e-01 3.97510678e-01 -7.55508900e-01
3.73489738e-01 -2.59205699e-01 -9.11484718e-01 1.89923406e-01
-3.59545887e-01 3.10667604e-01 9.03467834e-01 -5.51155686e-01
-9.61190045e-01 1.59173712e-01 -1.45212129e-01 -4.98406827e-01
-4.49655771e-01 4.50007319e-01 -3.33370030e-01 -6.56303585e-01
7.75585115e-01 2.38810092e-01 -1.92146912e-01 -4.56064194e-01
1.55643951e-02 3.92602473e-01 3.04848254e-01 -5.95725060e-01
1.02757382e+00 3.48235130e-01 5.55240870e-01 -1.22815692e+00
-5.08145273e-01 -1.35482579e-01 -7.47319102e-01 -4.03718978e-01
5.66008151e-01 -6.31861329e-01 -3.82305473e-01 2.15208694e-01
-8.66073370e-01 -2.14808419e-01 -6.26878440e-01 7.24783838e-01
-7.90806711e-01 3.58561456e-01 1.19574711e-01 -9.02495921e-01
2.11579502e-01 -1.51229322e+00 7.42122293e-01 1.09621614e-01
-6.94661066e-02 -1.01467729e+00 -2.12178096e-01 1.44394964e-01
6.60776675e-01 3.81554484e-01 1.04644263e+00 -3.90526295e-01
-7.95410633e-01 -3.39218080e-01 1.06818236e-01 4.82562810e-01
1.24438614e-01 -1.91066891e-01 -1.13737762e+00 -3.93584102e-01
6.64186418e-01 -2.29696721e-01 7.15560198e-01 5.22190332e-01
1.18643785e+00 7.67738968e-02 7.35154971e-02 5.99345505e-01
1.65921795e+00 -1.05408439e-02 6.83788478e-01 4.42528754e-01
3.43218565e-01 7.12624252e-01 4.46517527e-01 3.49346846e-01
-3.95760745e-01 1.25688946e+00 7.33454406e-01 8.10845271e-02
-2.02492326e-02 2.23678038e-01 3.85623835e-02 3.38152200e-01
-1.41734391e-01 -1.30381644e-01 -5.32025337e-01 4.41164255e-01
-1.52891707e+00 -8.08251560e-01 -3.22797149e-02 2.72271037e+00
2.72076786e-01 2.43337840e-01 -2.38463089e-01 5.34926414e-01
5.47978520e-01 1.61152035e-01 -5.61506093e-01 -4.75461364e-01
-1.44411027e-01 4.05148447e-01 5.80281317e-01 3.81986409e-01
-8.10474634e-01 2.50300080e-01 5.41030598e+00 6.49977446e-01
-1.64903760e+00 -9.80711654e-02 1.65522993e-01 1.86061244e-02
-1.92097232e-01 -9.42529067e-02 -8.22506249e-02 4.11558419e-01
7.02221572e-01 1.50796756e-01 5.22613883e-01 5.18443942e-01
3.69841546e-01 -3.33298177e-01 -1.01890743e+00 9.29757833e-01
-1.94498599e-01 -8.81359339e-01 -1.70418218e-01 5.92396548e-03
4.50605541e-01 9.99493804e-03 2.94934332e-01 -4.14883811e-03
-1.91414431e-01 -9.91703391e-01 7.82884777e-01 6.91771090e-01
4.62220043e-01 -5.60683131e-01 8.03390503e-01 5.04384756e-01
-5.49610078e-01 -2.54640967e-01 -3.01306605e-01 -3.27244252e-02
1.71950370e-01 9.26290095e-01 -8.45089078e-01 7.81558931e-01
6.70371652e-01 4.13538903e-01 -2.05450714e-01 1.36212671e+00
-2.25854471e-01 3.29226881e-01 -6.67746842e-01 -8.37620348e-02
1.16866499e-01 -2.04508439e-01 1.07456052e+00 8.36253643e-01
7.39668012e-01 -8.25261250e-02 -4.75403517e-02 8.03045869e-01
2.01603085e-01 2.48536635e-02 -8.46848130e-01 3.74156982e-01
3.37591432e-02 1.08550763e+00 -5.98515391e-01 2.87213713e-01
-2.02121824e-01 6.52296484e-01 1.28441200e-01 6.11758053e-01
-5.23565471e-01 -1.48356080e-01 5.39680421e-01 5.10167718e-01
6.10887229e-01 -2.64067739e-01 -9.40521732e-02 -1.05437183e+00
2.56230742e-01 -7.54402936e-01 3.49582657e-02 -8.28621984e-01
-1.07376564e+00 5.32114267e-01 7.69234449e-02 -1.43840086e+00
-2.58447349e-01 -7.87980974e-01 -5.36295116e-01 7.62097478e-01
-1.68732321e+00 -9.02428687e-01 -5.74446805e-02 5.75989842e-01
2.49454066e-01 -1.29281997e-03 6.98303342e-01 3.19565952e-01
-9.74732190e-02 1.07078791e-01 3.65265071e-01 -4.62890148e-01
6.13682747e-01 -1.11023247e+00 -4.61523145e-01 9.83022332e-01
2.28594109e-01 4.01544631e-01 1.42370141e+00 -1.89670041e-01
-1.38829052e+00 -8.52839947e-01 4.15439576e-01 -1.53861269e-01
5.74635625e-01 -2.79266953e-01 -1.06884217e+00 6.34160876e-01
-9.72009674e-02 2.59886652e-01 9.30106044e-02 1.60713848e-02
-1.46491826e-01 -4.46082264e-01 -1.19564104e+00 4.94279683e-01
4.46752101e-01 -5.15100062e-01 -4.75865275e-01 4.08667296e-01
2.25908726e-01 -4.97248471e-01 -6.81409180e-01 6.42410636e-01
4.95219469e-01 -1.34139872e+00 1.03642678e+00 -3.19079697e-01
7.69805312e-02 -3.07373613e-01 -2.72744268e-01 -1.39440429e+00
-1.38437405e-01 -4.42494273e-01 4.95825917e-01 9.11006451e-01
2.79187322e-01 -8.84487748e-01 5.88619411e-01 3.56923848e-01
-1.07008694e-02 -4.57714051e-01 -1.48332429e+00 -8.63983154e-01
5.86568601e-02 -6.63345933e-01 5.63736200e-01 6.56107306e-01
-4.80472237e-01 7.62610510e-02 -5.78811884e-01 5.79133570e-01
6.21578693e-01 4.69368726e-01 6.83160484e-01 -1.35045195e+00
-5.54431319e-01 -2.85232872e-01 -6.50409043e-01 -6.54792070e-01
-8.05026479e-03 -5.88879585e-01 2.52717614e-01 -1.09747779e+00
-4.97151911e-01 -7.39217103e-01 -4.37140822e-01 -1.12646013e-01
3.25781524e-01 1.17790299e-02 2.40736920e-02 1.36041585e-02
-5.57986125e-02 6.00898325e-01 1.15828323e+00 -5.02160415e-02
-1.63018197e-01 5.02212822e-01 -1.59036443e-01 6.69359505e-01
7.59812415e-01 -4.86221731e-01 -4.15644407e-01 -3.99384230e-01
2.33601615e-01 2.62561411e-01 5.10469794e-01 -1.52943468e+00
1.84836775e-01 -9.43596512e-02 3.58811080e-01 -5.58317564e-02
5.63370526e-01 -1.26869190e+00 5.12826622e-01 3.56657535e-01
-2.46831268e-01 -4.09825355e-01 2.61391908e-01 4.93645519e-01
-1.59239024e-01 -6.81435943e-01 1.02796733e+00 -3.16362828e-01
-6.07861817e-01 -2.45711040e-02 -3.81512254e-01 -3.30983281e-01
7.53018498e-01 -3.24773192e-01 1.78293645e-01 -6.41210675e-01
-6.92407370e-01 -3.05061996e-01 5.39723158e-01 1.77100375e-01
3.88837337e-01 -9.23078001e-01 -5.94773054e-01 1.91344738e-01
1.34153172e-01 -1.55878454e-01 3.31685513e-01 9.58808005e-01
-2.96736687e-01 3.75144720e-01 -2.50599265e-01 -6.10330939e-01
-1.05044651e+00 7.09456682e-01 7.28351712e-01 -6.57607391e-02
-6.86721087e-01 4.19355541e-01 6.15679845e-02 -3.96521479e-01
3.42393182e-02 -4.36970890e-01 -2.14035764e-01 -1.00551099e-01
4.80509400e-02 6.58661649e-02 5.03226697e-01 -6.20646119e-01
-1.49830475e-01 9.48403835e-01 2.84443438e-01 -6.52560359e-03
1.46239424e+00 -5.92385195e-02 1.68805242e-01 5.15323400e-01
9.36107695e-01 3.12961012e-01 -1.14996922e+00 -1.53878510e-01
2.63154823e-02 -4.94681835e-01 2.88480222e-01 -8.08619142e-01
-1.17515302e+00 1.08349311e+00 8.89908195e-01 1.27001017e-01
1.19694555e+00 -2.80218184e-01 2.42028445e-01 4.51185405e-01
6.07623756e-01 -8.59724283e-01 -1.32910475e-01 1.54779494e-01
1.09732413e+00 -1.23975110e+00 2.11415768e-01 -1.88060209e-01
-1.25543639e-01 1.35885692e+00 1.24665096e-01 -2.15576947e-01
5.37081897e-01 1.74739257e-01 1.72990754e-01 2.58736406e-02
-2.67886281e-01 -3.18211347e-01 1.34467229e-01 7.44696498e-01
7.37502351e-02 -4.72916327e-02 -1.48852229e-01 -6.52242452e-02
-1.73015222e-02 -1.63286664e-02 7.56466568e-01 5.59935331e-01
-3.40494722e-01 -1.07900536e+00 -6.52065158e-01 2.53727853e-01
-3.12669843e-01 5.21823227e-01 -1.33829251e-01 1.18457496e+00
3.33585322e-01 7.42322266e-01 -3.71716946e-01 -2.09403764e-02
5.30990601e-01 -4.99780253e-02 6.92791760e-01 -3.35778385e-01
-2.10096613e-01 9.75136906e-02 2.47858819e-02 -4.43536878e-01
-4.63669479e-01 -8.02605867e-01 -9.32797015e-01 9.24223363e-02
-2.46397644e-01 2.70109847e-02 8.25588644e-01 1.20342922e+00
-3.32198553e-02 5.75719416e-01 3.30035806e-01 -9.76273775e-01
-5.78866839e-01 -7.52344549e-01 -7.60035396e-01 2.61353284e-01
6.75628245e-01 -1.00019050e+00 -6.04303837e-01 -5.11398435e-01] | [11.455086708068848, -2.3258068561553955] |
34d65a77-0c1e-4efb-b776-e03249f25d3f | development-and-evaluation-of-conformal | 2304.00970 | null | https://arxiv.org/abs/2304.00970v1 | https://arxiv.org/pdf/2304.00970v1.pdf | Development and Evaluation of Conformal Prediction Methods for QSAR | The quantitative structure-activity relationship (QSAR) regression model is a commonly used technique for predicting biological activities of compounds using their molecular descriptors. Predictions from QSAR models can help, for example, to optimize molecular structure; prioritize compounds for further experimental testing; and estimate their toxicity. In addition to the accurate estimation of the activity, it is highly desirable to obtain some estimate of the uncertainty associated with the prediction, e.g., calculate a prediction interval (PI) containing the true molecular activity with a pre-specified probability, say 70%, 90% or 95%. The challenge is that most machine learning (ML) algorithms that achieve superior predictive performance require some add-on methods for estimating uncertainty of their prediction. The development of these algorithms is an active area of research by statistical and ML communities but their implementation for QSAR modeling remains limited. Conformal prediction (CP) is a promising approach. It is agnostic to the prediction algorithm and can produce valid prediction intervals under some weak assumptions on the data distribution. We proposed computationally efficient CP algorithms tailored to the most advanced ML models, including Deep Neural Networks and Gradient Boosting Machines. The validity and efficiency of proposed conformal predictors are demonstrated on a diverse collection of QSAR datasets as well as simulation studies. | ['Vladimir Svetnik', 'Robert P. Sheridan', 'Andy Liaw', 'Yuting Xu'] | 2023-04-03 | null | null | null | null | ['prediction-intervals'] | ['miscellaneous'] | [ 2.35054150e-01 -3.85347188e-01 -6.13242805e-01 -5.03226459e-01
-8.93888891e-01 -6.81070685e-01 4.05130476e-01 8.76563907e-01
-1.25079289e-01 1.37038112e+00 -2.36535758e-01 -5.83711863e-01
-3.28321427e-01 -8.57663631e-01 -8.59324276e-01 -8.73326480e-01
-4.66178596e-01 6.18147731e-01 1.35885686e-01 3.98574919e-02
4.48909312e-01 8.70261014e-01 -1.21911752e+00 2.83931047e-01
1.14723682e+00 1.40515876e+00 -1.17745079e-01 5.12616515e-01
-7.22152963e-02 7.25014567e-01 -6.89254165e-01 -3.71739417e-01
-9.31032896e-02 -2.44247913e-01 -4.60749894e-01 -6.93823338e-01
1.97761178e-01 1.29417568e-01 3.59073520e-01 8.32908332e-01
3.90944541e-01 1.50010288e-01 1.10143638e+00 -1.20778370e+00
-3.10270339e-01 4.35169369e-01 -4.05688494e-01 1.37193397e-01
4.76946115e-01 -4.02298748e-01 8.04561138e-01 -9.12367702e-01
1.24198414e-01 8.42478573e-01 8.66739035e-01 7.74288028e-02
-1.35908735e+00 -8.92366648e-01 -6.69258013e-02 3.89229238e-01
-1.62619901e+00 -1.28250003e-01 4.76634473e-01 -6.16912961e-01
8.13290238e-01 3.64856660e-01 4.71202254e-01 7.86944270e-01
9.78664994e-01 3.09040934e-01 1.26613486e+00 -2.27310002e-01
9.14465845e-01 3.01423967e-01 -8.43663979e-03 4.10716325e-01
2.52351403e-01 3.59549612e-01 -6.50171220e-01 -6.14076614e-01
1.96995839e-01 3.43288220e-02 -2.52726704e-01 -4.26333994e-01
-4.14471000e-01 1.20967770e+00 5.62408030e-01 2.57263333e-02
-4.84310150e-01 1.71317413e-01 2.70248115e-01 -9.01246741e-02
7.38092899e-01 6.45512521e-01 -8.25058401e-01 1.64146855e-01
-9.77011859e-01 4.56554204e-01 8.02418530e-01 5.86007595e-01
6.67684019e-01 -8.14670920e-02 -1.89689785e-01 5.04182935e-01
3.20090950e-01 2.08487019e-01 1.80543512e-01 -3.04780304e-01
1.30208671e-01 2.85026252e-01 3.65572244e-01 -7.33441412e-01
-4.76682961e-01 -6.95711374e-01 -7.48398602e-01 2.48456508e-01
3.44379455e-01 -7.11959004e-02 -8.52228343e-01 1.56008911e+00
3.54219824e-01 2.44847074e-01 -3.96682844e-02 5.49516857e-01
6.16932094e-01 1.08243787e+00 6.07596934e-01 -5.27222812e-01
9.19562638e-01 -2.16156691e-01 -3.71628553e-01 2.34090924e-01
6.65528715e-01 -6.54461563e-01 6.25613689e-01 7.22281218e-01
-5.72174251e-01 -1.49939686e-01 -1.21625161e+00 4.95695502e-01
-5.72390079e-01 -6.69400617e-02 8.56007397e-01 9.24506247e-01
-4.82629716e-01 1.15548539e+00 -7.70721495e-01 2.39552140e-01
4.86966223e-01 8.84479582e-01 -3.66598070e-01 6.49824589e-02
-1.32907295e+00 1.02944052e+00 4.26932484e-01 -3.69467363e-02
-7.55473316e-01 -1.23103511e+00 -7.40920126e-01 1.46578878e-01
-2.97955722e-02 -3.72103602e-01 9.10119295e-01 -9.35767114e-01
-1.51734877e+00 2.65941806e-02 -1.78852841e-01 -7.85078645e-01
1.82359561e-01 -1.28033325e-01 -1.98890835e-01 -2.52717912e-01
-2.46666316e-02 4.11117136e-01 4.18757826e-01 -8.97849262e-01
-4.99597132e-01 -6.39777362e-01 -3.18905443e-01 4.32196781e-02
6.90393597e-02 -5.25706634e-02 3.22843790e-01 -3.50120604e-01
-2.68145949e-01 -8.65218520e-01 -4.33275193e-01 -3.42447519e-01
-4.09862071e-01 -3.03267837e-01 6.24212205e-01 -6.50994122e-01
1.13586843e+00 -1.52893484e+00 -8.00696462e-02 5.76311767e-01
-6.52214978e-03 1.98367983e-01 1.03311658e-01 7.26249635e-01
-4.20569181e-01 7.85988122e-02 -3.49047214e-01 3.41988534e-01
-4.63545322e-01 -3.05117041e-01 -4.97716695e-01 6.56387568e-01
3.23870659e-01 5.35205305e-01 -6.55379832e-01 -3.11417669e-01
2.83123165e-01 6.20492935e-01 -4.72491086e-01 2.91580617e-01
-3.99488300e-01 5.13476372e-01 -5.91588199e-01 6.09446228e-01
6.98028982e-01 -3.08944225e-01 1.26806572e-01 -2.16990933e-01
-2.49557123e-01 2.49989375e-01 -1.13163435e+00 8.96578431e-01
-3.49377722e-01 2.53129721e-01 -7.91927576e-01 -1.24224901e+00
1.02965689e+00 2.59781659e-01 7.72621691e-01 -6.43389106e-01
-6.86251000e-02 1.92887753e-01 9.81136113e-02 2.53603965e-01
-1.00353137e-01 -4.86851335e-01 9.51152220e-02 -9.31744426e-02
-1.40327923e-02 1.26243820e-02 5.07090241e-02 -2.55842328e-01
9.14382219e-01 2.68676251e-01 8.38266671e-01 -4.64896142e-01
6.49007261e-01 1.29289299e-01 6.53344035e-01 4.12936807e-01
1.04167633e-01 1.29961565e-01 4.14243370e-01 -6.61514401e-01
-8.32700968e-01 -9.52729344e-01 -8.72375131e-01 9.10165250e-01
-2.75949035e-02 -3.95627230e-01 -6.17576063e-01 -5.57250500e-01
5.50891459e-02 6.54986501e-01 -7.76922703e-01 -1.92619011e-01
-7.84077346e-02 -1.16771984e+00 2.29445294e-01 5.96928895e-01
2.98734717e-02 -5.67029834e-01 -1.00907370e-01 5.34642756e-01
4.61505622e-01 -6.12503052e-01 8.77548382e-03 9.15315628e-01
-7.69701004e-01 -1.23501194e+00 -5.49097002e-01 -2.77048588e-01
3.48526418e-01 -2.01416612e-01 9.95904267e-01 -2.79939145e-01
-1.25934958e-01 -3.14148754e-01 -2.84928799e-01 -9.90973055e-01
-2.89757490e-01 -1.67366102e-01 9.33577493e-02 -2.20407784e-01
2.43343696e-01 -5.34589410e-01 -7.36427605e-01 3.98324192e-01
-6.18446410e-01 -4.06173468e-01 5.31035960e-01 8.67160082e-01
1.09533834e+00 1.25176370e-01 9.03213561e-01 -1.01008546e+00
6.38849795e-01 -7.02578664e-01 -1.00687742e+00 3.84240955e-01
-8.73734832e-01 3.34235489e-01 1.04839587e+00 -3.71579379e-01
-7.79628634e-01 4.47929859e-01 -4.05088097e-01 2.36331280e-02
-1.20606460e-02 9.14041400e-01 -9.54213440e-02 -4.47953701e-01
9.11395550e-01 4.16675620e-02 -4.48533148e-01 -1.78387195e-01
-3.90369520e-02 4.57348824e-01 -9.61262584e-02 -6.78846359e-01
4.32192147e-01 1.25276800e-02 6.85013294e-01 -8.12020481e-01
-8.81571472e-01 -4.15123999e-01 -4.18042034e-01 3.01694721e-02
5.81298232e-01 -7.63885498e-01 -1.00916159e+00 -1.18124038e-01
-1.01721513e+00 -1.08103462e-01 1.73280671e-01 6.85783088e-01
-5.21114945e-01 1.99546427e-01 -1.12289786e-01 -7.96634853e-01
-6.40589535e-01 -1.27265680e+00 8.28175247e-01 1.71687990e-01
-3.84183019e-01 -1.09921598e+00 3.80096287e-01 1.05300434e-01
1.72957450e-01 5.39644122e-01 1.22322094e+00 -1.07626331e+00
-2.52867788e-01 -6.19039834e-01 -7.35028763e-04 2.01452181e-01
6.90171719e-02 1.68770015e-01 -1.01787698e+00 -1.80141598e-01
-3.24284703e-01 -3.93035531e-01 6.32601380e-01 1.02218223e+00
1.58512998e+00 -3.23344231e-01 -4.78526622e-01 5.72488725e-01
1.44979942e+00 7.74947166e-01 6.11357987e-01 -5.45508675e-02
2.14762554e-01 3.94758821e-01 9.05486226e-01 6.46250665e-01
-9.24795717e-02 8.65387321e-01 5.33557653e-01 7.82678872e-02
3.87790084e-01 -3.91029596e-01 1.01206638e-01 2.92293960e-03
-1.93738714e-01 -2.73445129e-01 -8.09032381e-01 -1.94367111e-01
-1.61663413e+00 -9.03383493e-01 -2.49549270e-01 2.72212005e+00
1.05167294e+00 -8.94069374e-02 2.34249502e-01 2.44325340e-01
4.33810711e-01 -3.24287951e-01 -7.71028161e-01 -7.50209093e-01
1.08689614e-01 6.68996036e-01 6.90832436e-01 5.05215168e-01
-1.10436141e+00 7.35201716e-01 6.77953768e+00 1.22203219e+00
-1.30163550e+00 -3.30810785e-01 1.18919516e+00 4.20916885e-01
-1.25494227e-01 1.01242363e-01 -1.12567067e+00 5.76875687e-01
1.32618976e+00 -3.75547051e-01 1.07007183e-01 1.03589082e+00
3.04182589e-01 -3.18182409e-01 -1.25158060e+00 6.45085752e-01
-3.01663429e-01 -1.71566963e+00 9.71997604e-02 2.01386362e-01
7.68284619e-01 -2.63011545e-01 1.18450917e-01 1.63397655e-01
3.30587149e-01 -1.50825942e+00 2.85362720e-01 4.07133937e-01
7.97051847e-01 -1.32958555e+00 9.36673939e-01 3.75160158e-01
-1.10618734e+00 -5.65593205e-02 -6.76325858e-01 1.05136558e-01
-3.57128590e-01 7.59175837e-01 -1.19110906e+00 4.23239589e-01
4.97650981e-01 5.13217807e-01 -3.42003226e-01 1.36650634e+00
5.96904382e-02 6.74231231e-01 -3.65730107e-01 -5.04247665e-01
1.39344707e-01 -2.70211011e-01 -5.12866303e-02 9.92796302e-01
5.06743670e-01 8.01064521e-02 4.75645997e-02 7.32237518e-01
5.19523919e-02 5.02075255e-01 -5.94817400e-01 -1.07194297e-01
5.96834719e-01 8.38127315e-01 -5.21662951e-01 3.82023901e-02
-2.15523854e-01 3.60690892e-01 3.59149843e-01 2.68536121e-01
-8.80288899e-01 -2.93439656e-01 6.50237560e-01 2.83919632e-01
4.13233861e-02 -1.88850798e-02 -4.07506317e-01 -6.07856929e-01
-3.79674435e-01 -7.24415839e-01 5.01280427e-01 -5.15928864e-01
-1.16296601e+00 4.54707950e-01 7.91609883e-02 -1.27614319e+00
-3.20184618e-01 -9.17091191e-01 -5.29833078e-01 1.11313438e+00
-1.36745560e+00 -7.33092546e-01 6.65177330e-02 4.19703811e-01
2.35737354e-01 -2.35612631e-01 1.04470706e+00 6.09061792e-02
-4.35459018e-01 4.23000365e-01 4.19635981e-01 -4.48636979e-01
6.17178619e-01 -1.30992758e+00 -2.66285151e-01 2.93782860e-01
1.73668519e-01 4.86085117e-01 1.15073550e+00 -7.07678139e-01
-1.30081975e+00 -1.17994142e+00 6.16107285e-01 -4.59600300e-01
5.77895820e-01 1.58284325e-02 -8.36439252e-01 2.17690811e-01
-3.96347553e-01 1.48680314e-01 1.36468756e+00 2.83564121e-01
-7.79351965e-02 -4.02835637e-01 -1.15242529e+00 2.49469206e-01
2.40147486e-01 -1.22656617e-02 -8.20935294e-02 5.90968192e-01
3.65913272e-01 -3.87811869e-01 -1.22272635e+00 6.95338070e-01
7.41017520e-01 -8.32170844e-01 1.17402089e+00 -9.37248290e-01
3.22163910e-01 -3.24133635e-01 -2.73162663e-01 -1.44027495e+00
-2.50244588e-01 -1.36182770e-01 -1.35931954e-01 6.70572698e-01
7.48237669e-01 -3.90105426e-01 1.04224682e+00 7.20488489e-01
2.65652779e-02 -1.24691260e+00 -1.04818487e+00 -8.43415737e-01
3.07132572e-01 -5.67398071e-01 6.46256447e-01 6.86312377e-01
-4.00766172e-02 3.25596780e-01 -4.44961905e-01 3.73233527e-01
5.90411067e-01 1.30451143e-01 5.24441123e-01 -1.47727633e+00
-2.26810202e-01 -1.97797164e-01 -6.47157311e-01 -4.00254905e-01
1.89835593e-01 -6.10968709e-01 -1.71259031e-01 -1.29556763e+00
1.95082530e-01 -4.85623688e-01 -4.61176962e-01 2.65862286e-01
8.85861292e-02 5.11450507e-02 -4.14940894e-01 -7.48805841e-03
-2.34497666e-01 6.15040958e-01 8.08161676e-01 -8.57986286e-02
-4.31217700e-01 5.65631747e-01 -6.51571453e-01 5.34281969e-01
7.89912939e-01 -7.03566492e-01 -5.46403527e-01 5.17471969e-01
5.05927980e-01 4.71031964e-01 1.13876857e-01 -1.00318444e+00
1.10029101e-01 -6.24819636e-01 1.12106049e+00 -6.86499119e-01
2.69238889e-01 -7.36589730e-01 4.20737743e-01 6.52723014e-01
-4.64654624e-01 -1.53033555e-01 3.25406611e-01 7.92446852e-01
-3.70756447e-01 -1.81656271e-01 9.43545103e-01 2.73144543e-01
-2.65704244e-01 6.31648958e-01 -2.12993890e-01 -4.32965487e-01
1.22710586e+00 -3.03820282e-01 -1.63196158e-02 -3.78247231e-01
-5.99830329e-01 -9.67052951e-02 3.75116058e-02 -1.24000236e-01
6.47076726e-01 -1.24666858e+00 -5.45919657e-01 -1.65534511e-01
3.04328978e-01 -3.01669210e-01 -1.12291664e-01 5.44164777e-01
-6.20404243e-01 8.70482445e-01 -9.55580324e-02 -3.85799021e-01
-1.20498335e+00 5.89964032e-01 5.43402851e-01 -3.46097887e-01
1.24919698e-01 7.74749517e-01 4.15839195e-01 -3.39219719e-02
2.30690427e-02 -1.60185352e-01 -3.38198304e-01 -1.69554278e-01
6.39973760e-01 3.55135679e-01 4.68821287e-01 -3.14008981e-01
-7.02234507e-01 4.82928306e-01 -4.74941656e-02 3.56425703e-01
1.39623058e+00 6.37576938e-01 1.43023551e-01 1.97658867e-01
1.10492003e+00 -1.70836285e-01 -1.13198113e+00 9.99635085e-02
1.51336461e-01 -2.99321741e-01 3.99269223e-01 -1.22768593e+00
-4.42333162e-01 9.98806059e-01 6.84463799e-01 -4.29503694e-02
8.99089277e-01 -7.61578754e-02 2.20245674e-01 5.04511714e-01
3.70017648e-01 -8.51449788e-01 -1.39316365e-01 1.53623089e-01
1.05854714e+00 -1.27279520e+00 4.98596102e-01 -4.31385756e-01
-4.47759926e-01 1.32632089e+00 3.72116894e-01 -7.29561746e-02
8.05590808e-01 1.55282766e-01 -5.32380939e-01 -3.43078673e-02
-7.66873538e-01 2.14728028e-01 8.56550813e-01 6.85674012e-01
1.05337310e+00 2.45444894e-01 -5.29989898e-01 7.02113450e-01
-2.37489250e-02 1.41488696e-02 1.17297664e-01 6.84816957e-01
-5.69414377e-01 -1.23957849e+00 -2.79585391e-01 6.04105771e-01
-6.73348129e-01 -2.21341655e-01 -5.17398715e-01 6.09190762e-01
-1.63850978e-01 8.55444968e-01 -2.83778399e-01 -2.98821002e-01
1.03313155e-01 -1.41559303e-01 5.07120669e-01 -5.97015142e-01
-2.10738480e-01 -1.40132159e-01 1.79768845e-01 -4.07546937e-01
-1.58392608e-01 -4.17556763e-01 -1.20952892e+00 -2.99829811e-01
-5.58781326e-01 6.73901916e-01 9.87940490e-01 1.09315550e+00
2.13028103e-01 6.35503381e-02 9.22014296e-01 -5.58985770e-01
-5.47179103e-01 -6.26734674e-01 -6.31411433e-01 -9.09382552e-02
1.32333502e-01 -8.55572581e-01 -5.42122945e-02 -1.51152894e-01] | [5.072172164916992, 5.553449630737305] |
bcfde3c1-3724-40db-a972-a30c3f20af39 | online-multi-object-tracking-with-instance | 1902.08231 | null | http://arxiv.org/abs/1902.08231v1 | http://arxiv.org/pdf/1902.08231v1.pdf | Online Multi-Object Tracking with Instance-Aware Tracker and Dynamic Model Refreshment | Recent progresses in model-free single object tracking (SOT) algorithms have
largely inspired applying SOT to \emph{multi-object tracking} (MOT) to improve
the robustness as well as relieving dependency on external detector. However,
SOT algorithms are generally designed for distinguishing a target from its
environment, and hence meet problems when a target is spatially mixed with
similar objects as observed frequently in MOT. To address this issue, in this
paper we propose an instance-aware tracker to integrate SOT techniques for MOT
by encoding awareness both within and between target models. In particular, we
construct each target model by fusing information for distinguishing target
both from background and other instances (tracking targets). To conserve
uniqueness of all target models, our instance-aware tracker considers response
maps from all target models and assigns spatial locations exclusively to
optimize the overall accuracy. Another contribution we make is a dynamic model
refreshing strategy learned by a convolutional neural network. This strategy
helps to eliminate initialization noise as well as to adapt to the variation of
target size and appearance. To show the effectiveness of the proposed approach,
it is evaluated on the popular MOT15 and MOT16 challenge benchmarks. On both
benchmarks, our approach achieves the best overall performances in comparison
with published results. | ['Chiu C. Tan', 'Heng Fan', 'Haibin Ling', 'Peng Chu'] | 2019-02-21 | null | null | null | null | ['online-multi-object-tracking'] | ['computer-vision'] | [ 1.99461401e-01 -3.79191786e-01 -7.56670907e-02 -5.60842417e-02
-4.59994286e-01 -5.98060071e-01 6.26228929e-01 1.96600467e-01
-4.10800636e-01 4.20911849e-01 -3.71752262e-01 1.87783912e-01
5.67557253e-02 -6.09840333e-01 -9.07954991e-01 -8.41068685e-01
-1.97841033e-01 6.88458383e-01 1.15653133e+00 1.07407734e-01
-4.23256680e-03 6.39391005e-01 -1.67960083e+00 1.13713130e-01
6.58110619e-01 1.10340714e+00 5.89380920e-01 6.79224014e-01
-5.37405498e-02 8.16140294e-01 -9.60924685e-01 5.73060364e-02
6.34007633e-01 -4.06590514e-02 -1.47656888e-01 1.46421802e-03
1.04902017e+00 -1.53665662e-01 -1.99486881e-01 9.35522735e-01
5.90117335e-01 6.45947605e-02 4.19621497e-01 -1.30097497e+00
-1.71225443e-01 5.43138146e-01 -8.11618268e-01 6.36628866e-01
-1.57147706e-01 5.48191845e-01 4.41975623e-01 -4.82156783e-01
3.66405398e-01 1.35707688e+00 8.12187791e-01 6.63061857e-01
-1.19516540e+00 -8.78907204e-01 5.52518845e-01 2.78432369e-01
-1.32968295e+00 -5.32270312e-01 4.88294750e-01 -4.02448326e-01
7.74914920e-01 5.02035201e-01 6.25117838e-01 9.85632479e-01
3.92109632e-01 7.60870159e-01 9.35269833e-01 -1.11141510e-03
1.70865178e-01 1.84228405e-01 1.74364835e-01 3.36685807e-01
5.03883600e-01 3.72142106e-01 -3.77090812e-01 -1.02920309e-01
3.61698002e-01 2.82092001e-02 -1.78023443e-01 -5.89794755e-01
-1.39490402e+00 2.55645633e-01 7.51644194e-01 2.66098380e-01
-3.86171490e-01 6.06065929e-01 3.14380884e-01 2.49575265e-02
2.59180367e-01 -4.71966788e-02 -2.65699327e-01 1.11165553e-01
-9.80773687e-01 9.12760124e-02 3.66500050e-01 1.10719776e+00
6.25613868e-01 2.55222976e-01 -6.99061513e-01 5.26503026e-01
4.65402365e-01 9.86191213e-01 3.24074894e-01 -6.26939595e-01
3.09127033e-01 7.61076450e-01 3.27591538e-01 -8.16803455e-01
-4.92319822e-01 -8.96946788e-01 -5.27990758e-01 3.87934685e-01
2.97845572e-01 2.66190786e-02 -1.17428017e+00 1.82700622e+00
9.25176084e-01 6.68780327e-01 -1.38300315e-01 8.36399615e-01
9.03949261e-01 4.40860152e-01 2.46350884e-01 -8.64590108e-02
1.56251061e+00 -9.67391431e-01 -6.73128247e-01 -5.55730939e-01
5.75604439e-01 -7.09420145e-01 2.80846953e-01 1.25884980e-01
-5.85675061e-01 -8.17662418e-01 -9.60518062e-01 5.69710851e-01
-4.82111812e-01 2.87860185e-01 1.94366127e-01 7.17585862e-01
-1.06120932e+00 3.04995120e-01 -1.06081104e+00 -6.25039935e-01
5.85667431e-01 6.31313384e-01 -4.59580645e-02 1.47648051e-01
-7.51099527e-01 9.22749043e-01 5.23778319e-01 1.07152648e-01
-1.32922149e+00 -7.66651750e-01 -4.99223650e-01 -7.18773529e-02
6.31651998e-01 -5.69547772e-01 1.16291535e+00 -9.79484320e-01
-1.17162609e+00 4.55639303e-01 -2.03585982e-01 -8.23759198e-01
4.63363171e-01 -1.84891403e-01 -4.98106092e-01 -2.94827670e-01
1.84818834e-01 7.88807631e-01 9.86372292e-01 -1.37055802e+00
-1.15408361e+00 -2.94166297e-01 6.02336004e-02 5.78520298e-02
-3.62450957e-01 1.69609934e-01 -6.78012133e-01 -5.23298442e-01
-1.05267897e-01 -1.02668834e+00 -1.87349215e-01 2.34932989e-01
-1.91555679e-01 -1.94238484e-01 1.54791510e+00 -1.20719247e-01
1.12731779e+00 -2.08288097e+00 -1.60344541e-01 -2.42937282e-01
3.04851651e-01 5.56335568e-01 -1.77903801e-01 4.59452011e-02
2.99452573e-01 -5.50200343e-01 1.28231660e-01 -5.92436075e-01
1.56873018e-02 3.84363472e-01 -1.46627083e-01 6.05314314e-01
2.05174312e-01 8.94498885e-01 -8.75747621e-01 -6.73649311e-01
4.35400695e-01 6.00454569e-01 -1.49065107e-01 8.13994333e-02
-5.48018157e-01 6.20704353e-01 -5.44742703e-01 8.07874858e-01
1.02973354e+00 -3.60944480e-01 1.18846864e-01 -1.95955604e-01
-3.70923460e-01 3.85765210e-02 -1.43962049e+00 1.13223684e+00
-2.84508973e-01 5.30715883e-01 2.33659148e-01 -5.06965041e-01
6.95870399e-01 1.13829181e-01 6.96571112e-01 -7.74951339e-01
2.00904727e-01 5.87899946e-02 2.36485064e-01 -5.79862250e-03
5.81148982e-01 3.49213451e-01 1.13858342e-01 1.52188867e-01
-2.48717323e-01 4.87418920e-01 1.95234329e-01 1.67388842e-01
1.27848709e+00 2.24617332e-01 7.75188878e-02 -3.45250309e-01
4.37531710e-01 2.65579700e-01 9.03804958e-01 1.11009395e+00
-4.76015896e-01 1.75863490e-01 -3.30708116e-01 -5.63424706e-01
-2.98246235e-01 -9.74882901e-01 -1.62736192e-01 1.12483370e+00
7.03130066e-01 -2.52824605e-01 -4.92172062e-01 -7.47362137e-01
2.33094811e-01 3.78582388e-01 -7.50866055e-01 -1.53937697e-01
-8.85538936e-01 -7.71289468e-01 3.76552016e-01 6.38647139e-01
4.47952390e-01 -9.27426517e-01 -1.27586854e+00 4.79748935e-01
-9.21837091e-02 -1.29328263e+00 -4.88782614e-01 4.36664999e-01
-7.13759124e-01 -1.07391167e+00 -4.57856447e-01 -3.07022601e-01
3.47935498e-01 8.39102387e-01 1.09899974e+00 1.95073977e-01
-3.10367882e-01 5.40338516e-01 -2.21545368e-01 -6.72829509e-01
-3.89799774e-01 -5.65209910e-02 2.74499148e-01 2.76349336e-01
2.24228993e-01 -2.26302549e-01 -5.15000165e-01 8.11542630e-01
-8.61808658e-01 -2.60958165e-01 6.14674270e-01 3.30118775e-01
6.90018177e-01 1.34893462e-01 1.21735074e-01 -4.47320879e-01
-2.72050381e-01 -5.37218153e-01 -1.12354255e+00 3.36375713e-01
-3.33592534e-01 -1.81557015e-01 4.03710723e-01 -9.71382082e-01
-7.60822713e-01 3.50094289e-01 4.20877695e-01 -8.54368746e-01
-5.54478988e-02 -3.72047395e-01 -2.59167582e-01 -5.46858728e-01
5.84348381e-01 2.82117367e-01 -3.68284374e-01 -5.21798730e-01
-1.72929795e-04 2.80414760e-01 4.08196270e-01 -4.26399231e-01
1.17045617e+00 7.27680683e-01 1.28876105e-01 -5.96327066e-01
-7.19682515e-01 -6.88101470e-01 -5.26193202e-01 -6.73025608e-01
7.73261368e-01 -1.08545423e+00 -8.15370500e-01 4.67108458e-01
-1.00003481e+00 -5.18593550e-01 -2.38304526e-01 3.49058628e-01
-1.79447278e-01 2.47923657e-01 -1.05366267e-01 -9.46963906e-01
-2.08476365e-01 -1.23047984e+00 1.34250450e+00 5.41400731e-01
2.81301349e-01 -8.63288939e-01 1.98798954e-01 -6.29165247e-02
7.54636407e-01 2.98334897e-01 -7.98893496e-02 -5.94289184e-01
-1.26127458e+00 -5.13386168e-02 -2.14083791e-01 -2.93403774e-01
1.88137382e-01 -1.93478227e-01 -9.87211823e-01 -7.38417923e-01
-1.80525839e-01 1.34924904e-01 1.04508114e+00 4.72818524e-01
6.92164838e-01 2.15566605e-02 -1.13241875e+00 5.62355518e-01
1.44725645e+00 1.86506450e-01 2.12951630e-01 5.37759066e-01
8.55657876e-01 2.18947917e-01 1.02145398e+00 4.70913291e-01
3.19941789e-01 1.37287164e+00 1.06912100e+00 7.38167316e-02
-5.88623941e-01 2.04126626e-01 6.30238295e-01 6.31089264e-04
1.99311197e-01 -4.74101871e-01 -7.36098230e-01 7.05956399e-01
-1.96935189e+00 -9.83450174e-01 -4.19683456e-01 2.30810785e+00
2.45992988e-01 2.99874514e-01 3.92026305e-01 -3.91496986e-01
8.30087900e-01 1.84384361e-01 -7.10771978e-01 4.21142995e-01
-1.59271464e-01 -7.34433681e-02 9.43359971e-01 2.83114821e-01
-1.36952031e+00 8.41633797e-01 5.55215359e+00 9.99295592e-01
-1.21197510e+00 3.90189260e-01 -2.25658730e-01 -3.88275176e-01
4.31655526e-01 -4.82315160e-02 -1.49868417e+00 6.73396409e-01
9.60926950e-01 2.16025144e-01 5.18636331e-02 7.97732294e-01
7.57883564e-02 -1.39350832e-01 -9.33584511e-01 6.36854351e-01
-1.71732590e-01 -1.05682516e+00 -2.89725155e-01 1.78632468e-01
4.13627654e-01 4.68540698e-01 8.76461864e-02 4.91915911e-01
5.32128155e-01 -4.82128948e-01 1.10049677e+00 2.86303371e-01
1.89500019e-01 -1.82365000e-01 6.03851616e-01 4.58647668e-01
-2.06647825e+00 -2.07293600e-01 -3.55092466e-01 1.96030691e-01
1.71883926e-02 3.24436277e-01 -8.49111497e-01 7.63128579e-01
1.04856396e+00 7.57287920e-01 -1.01630902e+00 1.50859964e+00
3.83137107e-01 4.46374655e-01 -8.56422961e-01 8.73744190e-02
2.55125374e-01 4.71316010e-01 1.04414487e+00 1.30872643e+00
1.50190786e-01 -3.42870057e-01 6.68570161e-01 6.25933886e-01
3.79840910e-01 -2.50607342e-01 -4.79531765e-01 4.02353913e-01
6.19341433e-01 1.40108049e+00 -1.00845301e+00 -4.72762108e-01
-1.93702415e-01 5.49226999e-01 1.85227662e-01 2.23134328e-02
-1.39729977e+00 2.83438474e-01 8.06156933e-01 2.52968818e-01
9.39713240e-01 -8.24303403e-02 2.84042865e-01 -8.89373004e-01
1.26468077e-01 -5.67089915e-01 5.86619139e-01 -3.46969813e-01
-9.82186139e-01 8.15532923e-01 1.34479269e-01 -1.64456916e+00
2.01410502e-01 -4.68293726e-01 -5.42025864e-01 6.04904175e-01
-1.63369536e+00 -1.45571959e+00 -6.25313401e-01 5.73731005e-01
4.74518478e-01 1.90468371e-01 1.70981154e-01 6.76847219e-01
-6.80886686e-01 6.28864050e-01 5.24237305e-02 -2.08462059e-01
7.54556537e-01 -1.10595560e+00 4.00091499e-01 1.00195491e+00
1.34080634e-01 4.09443378e-01 8.28178287e-01 -9.04523432e-01
-1.54045796e+00 -1.60477209e+00 3.28252792e-01 -8.56157243e-01
5.13585150e-01 -4.25410122e-01 -9.25565958e-01 7.74133444e-01
1.10614888e-01 5.38333178e-01 7.57178888e-02 -2.58815587e-01
-3.21071088e-01 -3.18899572e-01 -1.06675029e+00 3.83419782e-01
1.05690229e+00 1.91858947e-01 -1.09050192e-01 3.17202210e-01
6.25066578e-01 -6.87771976e-01 -7.53665149e-01 6.90570414e-01
4.09915835e-01 -9.41988289e-01 1.07075918e+00 -1.63276922e-02
-6.96305454e-01 -1.19057953e+00 -2.41468102e-01 -8.81532431e-01
-4.98277903e-01 -5.50841570e-01 -5.99868298e-01 1.22134924e+00
-8.26808345e-03 -5.58207572e-01 8.30833077e-01 -3.38674411e-02
-3.39736998e-01 -3.14730763e-01 -1.31354749e+00 -1.22836494e+00
-4.86847073e-01 -1.63181096e-01 6.49705708e-01 5.71799755e-01
-8.80855381e-01 2.17317827e-02 -6.58741832e-01 7.35012293e-01
1.11939991e+00 2.15797365e-01 1.17088747e+00 -1.41651487e+00
-3.54474097e-01 -3.70207489e-01 -5.41043639e-01 -1.22247100e+00
-2.55333513e-01 -5.82869589e-01 4.06094104e-01 -1.21887827e+00
2.20897570e-01 -7.83738077e-01 -5.74752152e-01 5.05318165e-01
-2.13065878e-01 5.90558827e-01 5.08101702e-01 3.51734221e-01
-1.18208134e+00 3.79180014e-01 9.66577470e-01 -4.26902503e-01
-1.22896262e-01 2.78130978e-01 -3.49819392e-01 3.46552134e-01
5.30182004e-01 -9.78352785e-01 -6.64079413e-02 -5.21947205e-01
-3.52471620e-01 -2.10837513e-01 8.30106795e-01 -1.75660408e+00
4.51527625e-01 -1.80338353e-01 3.77827525e-01 -1.19667315e+00
5.50122261e-01 -1.40860093e+00 7.21729279e-01 7.82616794e-01
2.08803788e-01 -6.83291480e-02 6.82803333e-01 8.71936679e-01
2.80771434e-01 6.45145774e-03 1.06613684e+00 5.13232164e-02
-9.01584923e-01 3.73015225e-01 -1.75488532e-01 -3.25586885e-01
1.37990320e+00 -5.00639498e-01 -4.95511323e-01 2.23370209e-01
-3.69105160e-01 6.60044849e-01 6.01577282e-01 6.58971190e-01
1.39878973e-01 -1.28301990e+00 -4.66648072e-01 -2.56058313e-02
4.05194968e-01 -2.25734398e-01 3.74746174e-01 1.27967989e+00
1.15272971e-02 5.01284719e-01 -1.94851831e-01 -1.22460401e+00
-1.59731507e+00 7.73535192e-01 6.90163016e-01 -3.92983288e-01
-7.69774079e-01 6.04425967e-01 5.45699060e-01 -1.67910352e-01
5.23176849e-01 -3.83726805e-01 -2.38322556e-01 -2.51337796e-01
7.38804698e-01 2.15579554e-01 -1.10115875e-02 -9.31701720e-01
-7.32092440e-01 7.03501821e-01 -2.24962041e-01 4.33621049e-01
9.04019952e-01 -1.09078638e-01 3.07915539e-01 2.64926344e-01
4.00339007e-01 7.99367577e-03 -1.60908413e+00 -3.93512458e-01
1.35540128e-01 -5.42349398e-01 -3.43489274e-02 -7.12240040e-01
-1.15134573e+00 4.97902721e-01 1.21968913e+00 2.64484942e-01
9.63190913e-01 7.73972049e-02 6.40834093e-01 2.14993149e-01
6.69336498e-01 -6.18595064e-01 3.73743884e-02 4.24370021e-01
2.80003816e-01 -1.20214283e+00 -7.97552466e-02 -1.73754573e-01
-1.78306326e-01 7.48012006e-01 1.09687400e+00 -1.99411456e-02
3.91815126e-01 5.53450286e-01 1.50008425e-01 -2.48236224e-01
-7.48571336e-01 -6.18642092e-01 2.29667172e-01 8.70405793e-01
-1.30465239e-01 -5.30650243e-02 1.79678753e-01 2.84492150e-02
4.43798214e-01 -1.80283993e-01 9.30605382e-02 1.16325116e+00
-7.97941089e-01 -1.02390504e+00 -9.27155495e-01 2.58453220e-01
-3.11102599e-01 2.19177485e-01 -3.02941114e-01 8.72272491e-01
3.87811124e-01 9.37585831e-01 1.27095534e-02 -3.48681539e-01
4.15089905e-01 -5.09482324e-01 5.06696165e-01 -4.72525567e-01
-1.15951419e+00 2.84968048e-01 -1.58548638e-01 -6.56622112e-01
-8.07699502e-01 -6.81203067e-01 -1.04639912e+00 -9.02731642e-02
-6.82513535e-01 -5.18803345e-03 5.42493701e-01 7.88570046e-01
5.07314384e-01 1.00291181e+00 2.73670167e-01 -1.12198973e+00
-3.17818075e-01 -1.05942667e+00 -2.47125328e-01 5.37365712e-02
5.68252385e-01 -1.34703374e+00 -1.48289308e-01 -2.40856096e-01] | [6.304345607757568, -2.074509620666504] |
ea0d32f3-a9b8-4f40-a8ab-d667e7e8d2b7 | analysing-diffusion-based-generative | 2211.02397 | null | https://arxiv.org/abs/2211.02397v2 | https://arxiv.org/pdf/2211.02397v2.pdf | Analysing Diffusion-based Generative Approaches versus Discriminative Approaches for Speech Restoration | Diffusion-based generative models have had a high impact on the computer vision and speech processing communities these past years. Besides data generation tasks, they have also been employed for data restoration tasks like speech enhancement and dereverberation. While discriminative models have traditionally been argued to be more powerful e.g. for speech enhancement, generative diffusion approaches have recently been shown to narrow this performance gap considerably. In this paper, we systematically compare the performance of generative diffusion models and discriminative approaches on different speech restoration tasks. For this, we extend our prior contributions on diffusion-based speech enhancement in the complex time-frequency domain to the task of bandwith extension. We then compare it to a discriminatively trained neural network with the same network architecture on three restoration tasks, namely speech denoising, dereverberation and bandwidth extension. We observe that the generative approach performs globally better than its discriminative counterpart on all tasks, with the strongest benefit for non-additive distortion models, like in dereverberation and bandwidth extension. Code and audio examples can be found online at https://uhh.de/inf-sp-sgmsemultitask | ['Timo Gerkmann', 'Simon Welker', 'Julius Richter', 'Jean-Marie Lemercier'] | 2022-11-04 | null | null | null | null | ['bandwidth-extension', 'bandwidth-extension', 'speech-denoising', 'speech-dereverberation'] | ['audio', 'speech', 'speech', 'speech'] | [ 2.1260795e-01 8.9674205e-02 2.2906917e-01 -2.3485799e-01
-9.5092815e-01 -2.5753736e-01 8.4782857e-01 -1.3655451e-01
-4.0893605e-01 4.9706522e-01 6.6843814e-01 -2.4814512e-01
-3.9915726e-01 -4.7570491e-01 -2.3909862e-01 -1.2174311e+00
9.2308842e-02 5.6803610e-02 9.2931338e-02 -3.9724740e-01
9.3054362e-03 3.8962013e-01 -1.5978180e+00 8.1734031e-02
9.3682581e-01 7.8104615e-01 4.5141479e-01 7.9791015e-01
4.0205380e-01 4.0387338e-01 -7.6157981e-01 -3.5755384e-01
1.2568069e-02 -4.2484507e-01 -4.1066250e-01 2.2285478e-01
2.5998548e-01 -2.1987452e-01 -7.7697504e-01 1.1465470e+00
1.1529394e+00 4.8603562e-01 6.0348457e-01 -8.8333589e-01
-1.0179824e+00 5.2913547e-01 -3.2010689e-01 5.2164966e-01
1.8460487e-01 1.4857978e-02 7.6560420e-01 -1.1170713e+00
5.7204443e-01 1.3249419e+00 6.4974791e-01 4.1496512e-01
-1.2500663e+00 -3.2558817e-01 -2.0814694e-01 5.6084472e-01
-1.0059032e+00 -9.3540281e-01 9.6159917e-01 -2.6561373e-01
9.0657902e-01 2.7834287e-01 2.9298380e-01 1.2678568e+00
-6.3409269e-02 9.0235066e-01 1.1662608e+00 -4.0262961e-01
1.8975161e-01 -1.4442708e-01 -1.5198855e-01 1.0189011e-02
-1.4944765e-01 5.1748240e-01 -6.5041792e-01 2.0279641e-01
5.7923079e-01 -3.5521495e-01 -6.9119304e-01 1.4291761e-02
-8.5007298e-01 9.5329505e-01 3.5718730e-01 7.9229307e-01
-4.7155613e-01 1.0180469e-01 4.1417816e-01 5.6486303e-01
1.1444917e+00 8.3239593e-02 -1.6558142e-01 -2.0192970e-01
-1.1025903e+00 2.9241157e-01 5.7666028e-01 5.5568415e-01
3.7830970e-01 6.0546368e-01 -3.6276120e-01 1.4544533e+00
2.8812954e-01 5.5554461e-01 4.0281838e-01 -8.8792348e-01
3.3035412e-01 -3.9250803e-01 -2.0530437e-01 -7.3276985e-01
-3.4236270e-01 -8.8391113e-01 -1.3191469e+00 2.5422901e-01
1.4757721e-01 -1.7961064e-01 -8.8093674e-01 1.6674545e+00
2.2683010e-01 8.2456782e-02 7.0834309e-02 1.0468109e+00
6.9071841e-01 8.7139112e-01 -1.4742662e-01 -4.6160567e-01
9.0450478e-01 -8.5995030e-01 -1.0759048e+00 -1.9066857e-01
1.8710808e-01 -1.2272811e+00 6.4241832e-01 6.3359207e-01
-1.2183288e+00 -7.4881887e-01 -7.8469127e-01 -2.9343209e-01
-1.0203409e-01 7.5974971e-02 3.0848941e-01 9.5005876e-01
-1.6170484e+00 7.6569724e-01 -6.5501869e-01 -2.8129950e-01
3.9663443e-01 -7.2706193e-02 -1.6244113e-01 -2.4115901e-01
-1.1609366e+00 8.8171536e-01 -4.8121624e-03 1.7418525e-01
-9.0171522e-01 -6.5279514e-01 -7.8009802e-01 -1.9462092e-02
2.1366832e-01 -6.5197450e-01 1.2069112e+00 -7.5149554e-01
-1.4942364e+00 6.7428184e-01 -1.2722103e-01 -7.5070083e-01
5.8768928e-01 -3.1531298e-01 -5.9311908e-01 -3.3270169e-02
-1.9536130e-01 4.2138445e-01 1.2597661e+00 -1.0089453e+00
-2.8652835e-01 -3.1932446e-01 -1.8586259e-01 1.2592547e-01
-3.6097264e-01 8.3022267e-02 -2.2719574e-01 -1.3395851e+00
1.0028764e-01 -7.0943022e-01 -1.4759725e-01 -1.8757544e-01
-3.4201795e-01 -1.4107747e-01 1.0563096e+00 -1.1771443e+00
1.0947033e+00 -2.3679113e+00 3.9250687e-01 -1.8372303e-01
-1.2241913e-02 5.6673652e-01 -2.0798305e-01 6.6376436e-01
-2.4664541e-01 -2.1631451e-01 -5.0831616e-01 -8.5409373e-01
1.9395283e-01 2.3701410e-01 -3.7714463e-01 6.2616950e-01
6.8712592e-02 5.8958882e-01 -6.0484904e-01 5.0046604e-02
3.7366253e-01 1.2161152e+00 -6.7055136e-01 8.0174156e-02
1.6222458e-01 6.0401565e-01 1.7407323e-01 4.4386780e-01
7.3080850e-01 2.6467267e-01 -1.8761773e-01 -1.1375414e-01
-8.1278339e-02 2.2097212e-01 -1.1068075e+00 1.8508176e+00
-5.8378643e-01 1.0194966e+00 5.0863647e-01 -1.1401628e+00
8.8626653e-01 6.1364996e-01 4.2301607e-01 -7.9106802e-01
-1.7925797e-01 3.2699871e-01 1.6055229e-01 -4.0818143e-01
6.4366692e-01 -5.1852423e-01 4.8661643e-01 2.8479138e-01
3.6256421e-01 -3.6022496e-01 2.1287140e-01 1.7078729e-01
9.1160870e-01 -7.1246433e-03 9.2479803e-02 -2.2526404e-01
4.4011715e-01 -6.0965830e-01 1.8338206e-01 5.2619779e-01
-1.6367984e-01 8.9782691e-01 1.1571305e-01 2.7231658e-01
-1.1819727e+00 -1.2178055e+00 -4.1583583e-01 1.2077115e+00
-2.1489079e-01 -1.8229769e-01 -8.2967383e-01 -2.2752465e-01
-1.9697133e-01 7.8418022e-01 -2.6444402e-01 -2.6031801e-01
-4.5536458e-01 -8.8853782e-01 4.4021967e-01 3.5917157e-01
3.7869200e-01 -9.9180788e-01 2.5130993e-02 3.9725992e-01
-2.2140853e-01 -9.5618290e-01 -5.7370710e-01 2.6028791e-01
-8.1487459e-01 -5.0603354e-01 -1.3279294e+00 -5.8042336e-01
1.7410749e-01 2.9852962e-01 8.7009668e-01 -1.4926165e-01
-2.6471797e-01 6.9852042e-01 -5.4039139e-01 -3.7676433e-01
-7.7094263e-01 -9.0954900e-02 7.0819542e-02 1.6074175e-01
-1.1728713e-01 -1.0143752e+00 -7.9157215e-01 2.5671974e-01
-1.2562022e+00 -2.4125300e-01 4.6830237e-01 9.3001866e-01
3.4755936e-01 -5.2620951e-02 9.3039811e-01 -4.4331935e-01
1.1162980e+00 -4.2733273e-01 -2.6493374e-01 -3.3916664e-01
-7.1037728e-01 -1.7568386e-01 4.2345026e-01 -4.1821009e-01
-1.1995173e+00 -4.1234851e-01 -8.8755339e-01 -3.3825731e-01
-2.5854155e-01 5.4520917e-01 -1.5567607e-01 1.3517131e-01
7.5084078e-01 4.2091399e-01 -7.3521353e-02 -9.0625072e-01
3.7014279e-01 7.5126797e-01 6.6018587e-01 -2.2971171e-01
7.6013702e-01 4.4941041e-01 -1.7764169e-01 -1.3865263e+00
-3.9946631e-01 -5.5635446e-01 -2.2509553e-01 -3.2847521e-01
7.2221226e-01 -8.0493629e-01 -2.3665875e-02 7.2687167e-01
-1.1569359e+00 -4.1310680e-01 -5.4555237e-01 4.8448747e-01
-6.9498646e-01 5.7518750e-01 -6.8615019e-01 -8.7295204e-01
-3.3749068e-01 -1.1035906e+00 1.0489080e+00 1.5412121e-01
1.9114168e-01 -1.2309130e+00 1.4772004e-01 1.4077689e-01
9.2039025e-01 -1.5344939e-01 5.8402181e-01 -5.4469544e-01
-2.0544097e-01 -9.1115020e-02 6.4887375e-02 8.3491296e-01
1.4373019e-01 -4.0170962e-01 -1.4224049e+00 -4.5796752e-01
2.8735599e-01 1.2830459e-01 1.3129247e+00 7.9711103e-01
9.2899352e-01 -7.9622522e-02 7.6262832e-02 6.2788910e-01
1.0397004e+00 1.5863021e-01 1.0545534e+00 -3.2751717e-02
3.1536657e-01 6.7235237e-01 2.4108036e-01 3.7999356e-01
-7.7643514e-02 9.0183604e-01 3.4448129e-01 -3.3304334e-01
-8.5459358e-01 8.7945230e-02 5.3118467e-01 1.0642687e+00
-2.9044941e-01 -5.7627344e-01 -6.9866544e-01 7.4403733e-01
-1.5201497e+00 -1.1160519e+00 -2.2262786e-01 2.0269303e+00
6.4592820e-01 -1.7722519e-01 7.3088184e-02 6.7969245e-01
7.2511953e-01 5.3208959e-01 -3.1130734e-01 -3.7569463e-01
-5.8186644e-01 4.7142974e-01 7.1208104e-02 7.4972856e-01
-1.0005587e+00 5.9716624e-01 5.6890898e+00 1.2084885e+00
-9.9530125e-01 7.1005678e-01 7.2323519e-01 -9.0665378e-02
-3.0989933e-01 -3.0420846e-01 -4.4149664e-01 3.3030042e-01
1.0465685e+00 -1.1750374e-01 4.4726101e-01 5.1831561e-01
5.6965804e-01 -8.8713735e-02 -8.0056322e-01 1.0795230e+00
7.0283100e-02 -1.0657318e+00 -3.1903529e-01 2.7009064e-01
8.1734681e-01 2.0252357e-01 4.3782443e-01 2.0199692e-01
-7.2505072e-02 -9.3295336e-01 6.8282008e-01 5.2812266e-01
6.5515572e-01 -6.1654198e-01 5.5623478e-01 2.9707143e-01
-8.2163221e-01 -2.5470380e-02 -3.7346813e-01 1.9344611e-01
6.3491380e-01 1.1440079e+00 -4.8927739e-01 7.0262259e-01
6.0005182e-01 7.1384865e-01 -8.8039353e-02 1.2846565e+00
-4.3370342e-01 8.7608463e-01 -1.4999261e-01 4.7558194e-01
9.1523498e-02 -3.3048722e-01 1.0930370e+00 1.4659534e+00
7.4305236e-01 -1.7947201e-01 -3.2878584e-01 6.8646026e-01
5.2620012e-02 -2.2060281e-02 -4.7921112e-01 -1.2613113e-02
1.3983130e-01 1.0202994e+00 -3.4801447e-01 -1.3991697e-01
-2.1785688e-01 1.0567584e+00 -2.4821712e-01 6.5954459e-01
-8.0738837e-01 -2.8990862e-01 6.8928361e-01 2.1671422e-01
4.6858689e-01 -3.2505742e-01 2.9963229e-02 -8.6751282e-01
3.1817898e-02 -8.2441241e-01 4.8398588e-02 -9.1698217e-01
-1.2478467e+00 7.0778662e-01 -1.4741907e-01 -9.9456847e-01
-5.0744474e-01 -4.4857347e-01 -4.4396096e-01 1.0888605e+00
-1.6156642e+00 -9.9597102e-01 -5.9147587e-05 6.4797467e-01
8.8245463e-01 -4.3818921e-02 5.7387513e-01 7.5981557e-01
-4.1232106e-01 2.2982298e-01 5.0096923e-01 -2.8566346e-01
8.4488297e-01 -1.1939102e+00 4.3877867e-01 1.2326710e+00
4.3466640e-01 2.7773279e-01 1.1167994e+00 -4.4999841e-01
-1.1681103e+00 -9.5805109e-01 8.2949680e-01 1.2061725e-01
5.9852552e-01 -2.8144550e-01 -1.0642265e+00 3.5133189e-01
5.1135492e-01 -5.7430852e-02 3.8092700e-01 8.6713016e-02
-1.8710840e-01 -5.1436327e-02 -9.0011650e-01 3.9825833e-01
1.1116010e+00 -7.0884120e-01 -3.0935419e-01 4.2317039e-01
4.6810642e-01 -3.7350518e-01 -8.9178479e-01 2.3105834e-01
1.3148290e-01 -1.2771231e+00 1.2410964e+00 -1.0809446e-01
4.9891070e-01 -1.5853657e-01 -3.1349328e-01 -1.7852577e+00
-1.7327784e-01 -9.6928960e-01 -2.2372864e-01 1.3570412e+00
3.3491549e-01 -8.4020811e-01 2.4035192e-01 -1.7518365e-01
-6.4534426e-01 -4.1701731e-01 -1.3055325e+00 -9.2509097e-01
8.4840126e-02 -8.3058101e-01 1.5971270e-01 7.1360272e-01
-2.8532279e-01 2.3226093e-01 -7.1562088e-01 1.1855556e-01
5.8658636e-01 -1.6908571e-01 4.9024352e-01 -9.5624846e-01
-5.6365222e-01 -6.5297461e-01 -2.7624211e-01 -1.2908967e+00
-3.0161791e-02 -9.3610829e-01 2.5707254e-01 -1.6866533e+00
-2.3882115e-01 -1.8735632e-02 -6.9841653e-02 9.0797588e-02
-1.3924558e-01 2.8579602e-01 1.5669361e-01 8.5131399e-02
-7.2222114e-02 7.8588963e-01 1.3044834e+00 -1.8558852e-01
-1.2961589e-01 2.1458702e-01 -6.0188156e-01 4.4530517e-01
7.5480020e-01 -3.7993565e-01 -4.7882006e-01 -6.1164767e-01
-1.8220739e-02 8.7316602e-02 5.1599467e-01 -1.0730565e+00
1.9561931e-01 4.0391478e-01 4.4677816e-02 -1.4663999e-01
6.8057185e-01 -4.4184384e-01 1.1814536e-01 3.1621847e-01
-1.7439194e-01 -2.3823573e-01 2.3729882e-01 6.4142233e-01
-5.9547573e-01 -1.5255217e-01 8.7102973e-01 3.5913956e-01
-5.5184782e-01 2.3451687e-01 -6.0782987e-01 -1.6830516e-01
4.4702083e-01 -2.2286461e-01 -2.6785886e-01 -8.3034199e-01
-1.0876918e+00 -5.8348972e-01 3.2259509e-02 4.9266505e-01
7.2177070e-01 -1.1113222e+00 -1.1687099e+00 1.0869646e-01
-3.2551935e-01 -3.6595345e-01 8.1207490e-01 1.3665423e+00
2.9379616e-02 3.2089791e-01 1.8984491e-01 -5.8406335e-01
-1.2600528e+00 4.8988673e-01 3.2423395e-01 -1.5360501e-01
-7.3400903e-01 9.1389126e-01 3.3643547e-01 7.2361283e-02
1.6065170e-01 -2.1768510e-01 -5.8779500e-02 1.5218565e-01
5.4522544e-01 6.4343280e-01 4.7570425e-01 -7.7912241e-01
-4.6686497e-02 2.1117748e-01 1.9271551e-01 -3.5680920e-01
1.6638234e+00 -3.5961953e-01 1.4954418e-01 1.9140209e-01
1.1246104e+00 2.6419020e-01 -1.2599291e+00 -2.4119344e-01
-1.7658824e-01 -2.9591203e-01 5.0664943e-01 -8.4041148e-01
-1.1926345e+00 1.1179674e+00 8.6378312e-01 7.5354701e-01
1.5721166e+00 6.8097264e-02 6.0720551e-01 -3.2267389e-01
8.6515792e-02 -9.5785308e-01 1.3168685e-01 5.0757700e-01
1.4878048e+00 -1.0513726e+00 -1.8097159e-01 -3.3041689e-01
-3.3626717e-01 9.2330199e-01 -2.0821919e-01 -6.1523006e-03
6.8616480e-01 4.7066802e-01 -6.2282678e-02 1.2718652e-02
-3.8716498e-01 -4.9056426e-01 4.3719107e-01 1.1401687e+00
6.2882960e-01 -1.5950462e-02 -2.2755966e-01 2.2149052e-01
-2.0516035e-01 -3.1775543e-01 3.6634150e-01 4.8574692e-01
-4.8718747e-01 -1.2937672e+00 -6.6729999e-01 6.9777265e-02
-5.0193399e-01 -4.5813406e-01 -1.4087497e-01 5.7565612e-01
-6.5356173e-02 1.3481060e+00 -1.7894539e-01 -9.7734123e-02
2.8670126e-01 -6.4232871e-02 4.5321468e-01 -3.4718919e-01
-5.9344447e-01 6.9850785e-01 2.4454854e-01 -3.5851005e-01
-6.4039052e-01 -8.3343947e-01 -6.6956037e-01 -3.1805840e-01
-2.6007438e-01 -4.8039161e-02 8.9159250e-01 8.2025748e-01
2.4186979e-01 8.7431651e-01 4.4941625e-01 -9.9174547e-01
-6.0005170e-01 -1.4705139e+00 -7.9602712e-01 2.7179193e-01
4.9927467e-01 -4.1927567e-01 -5.2291089e-01 1.9154546e-01] | [15.165258407592773, 5.984804630279541] |
0223373b-d432-45d2-b35e-0b396249f3da | frugalml-how-to-use-ml-prediction-apis-more | 2006.07512 | null | https://arxiv.org/abs/2006.07512v1 | https://arxiv.org/pdf/2006.07512v1.pdf | FrugalML: How to Use ML Prediction APIs More Accurately and Cheaply | Prediction APIs offered for a fee are a fast-growing industry and an important part of machine learning as a service. While many such services are available, the heterogeneity in their price and performance makes it challenging for users to decide which API or combination of APIs to use for their own data and budget. We take a first step towards addressing this challenge by proposing FrugalML, a principled framework that jointly learns the strength and weakness of each API on different data, and performs an efficient optimization to automatically identify the best sequential strategy to adaptively use the available APIs within a budget constraint. Our theoretical analysis shows that natural sparsity in the formulation can be leveraged to make FrugalML efficient. We conduct systematic experiments using ML APIs from Google, Microsoft, Amazon, IBM, Baidu and other providers for tasks including facial emotion recognition, sentiment analysis and speech recognition. Across various tasks, FrugalML can achieve up to 90% cost reduction while matching the accuracy of the best single API, or up to 5% better accuracy while matching the best API's cost. | ['James Zou', 'Matei Zaharia', 'Lingjiao Chen'] | 2020-06-12 | null | http://proceedings.neurips.cc/paper/2020/hash/789ba2ae4d335e8a2ad283a3f7effced-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/789ba2ae4d335e8a2ad283a3f7effced-Paper.pdf | neurips-2020-12 | ['facial-emotion-recognition'] | ['computer-vision'] | [-7.08003156e-03 -1.30927294e-01 -7.07677186e-01 -9.04549420e-01
-8.97819281e-01 -6.37290239e-01 2.54897922e-01 -1.15210600e-01
-2.13784561e-01 1.53041542e-01 -9.19077471e-02 -2.27503598e-01
-2.02031627e-01 -2.55949765e-01 -5.42139292e-01 -3.70543480e-01
-1.40891135e-01 4.06643242e-01 -1.17738374e-01 1.52221784e-01
1.13801144e-01 4.74902898e-01 -1.68218625e+00 5.23026347e-01
6.37244463e-01 1.61404669e+00 1.43842120e-02 3.52123260e-01
-2.16818795e-01 6.85449719e-01 -2.41428986e-01 -7.64986455e-01
6.49421811e-01 2.09340319e-01 -5.66707969e-01 2.42355362e-01
4.93926585e-01 -3.96828890e-01 7.16969594e-02 8.71023655e-01
2.15235740e-01 -2.46293664e-01 3.74298096e-01 -1.77583683e+00
-7.90654570e-02 4.61505532e-01 -5.88488460e-01 9.39458981e-02
4.17176038e-01 6.63005710e-02 1.50405824e+00 -6.10305488e-01
3.14511895e-01 9.76825833e-01 4.28168476e-01 5.22083759e-01
-1.12809956e+00 -9.19609070e-01 4.22037870e-01 -1.38833299e-01
-1.19522321e+00 -8.96691382e-01 6.82946801e-01 -7.23846331e-02
1.17254508e+00 5.46954274e-01 3.98726702e-01 8.30909729e-01
5.94412126e-02 1.10130179e+00 9.79140341e-01 -1.31386578e-01
1.37724593e-01 4.20413435e-01 2.58672774e-01 7.72103369e-01
-2.32401248e-02 -4.19443309e-01 -9.75247979e-01 -6.94051981e-01
5.62560298e-02 3.18935037e-01 -8.69048461e-02 -3.05347860e-01
-4.91511464e-01 8.00955117e-01 1.04379848e-01 -2.02832390e-02
-3.94701958e-01 1.04511492e-01 2.96646625e-01 4.65542018e-01
4.20498490e-01 4.95692581e-01 -1.01599574e+00 -5.46244860e-01
-9.15603936e-01 2.20754519e-01 1.20673704e+00 9.22877014e-01
7.35632420e-01 -2.93917328e-01 3.03515613e-01 1.05896449e+00
3.49660695e-01 3.66260678e-01 4.80665833e-01 -1.14070594e+00
4.40337628e-01 7.48973012e-01 1.41456664e-01 -8.50819051e-01
-1.33029774e-01 -1.28938090e-02 -2.32649446e-01 -1.10872500e-01
3.45566064e-01 -2.75676191e-01 -4.64810610e-01 1.60207820e+00
3.25152934e-01 6.92230463e-02 -3.80273372e-01 6.38209760e-01
2.74785727e-01 5.37791729e-01 -9.02515054e-02 -2.44320780e-01
1.39101791e+00 -9.27977681e-01 -2.40306824e-01 -4.25489604e-01
6.08605146e-01 -7.74132073e-01 1.29696381e+00 7.72415996e-01
-1.01594901e+00 5.58857135e-02 -8.99427176e-01 7.59604052e-02
-2.27054656e-01 1.28518501e-02 1.38013005e+00 8.48141968e-01
-7.05108643e-01 6.09972656e-01 -9.96590912e-01 -4.43024561e-02
6.97039366e-01 8.29539716e-01 -3.93070757e-01 -2.23558247e-01
-4.46948081e-01 4.54937756e-01 -2.57332176e-01 -2.87770718e-01
-5.83267927e-01 -8.35337222e-01 -5.62780797e-01 2.46788234e-01
4.07611579e-01 -2.43330956e-01 1.48593581e+00 -1.42225027e+00
-1.50874889e+00 8.20286155e-01 -3.87834132e-01 -4.76595432e-01
2.43712530e-01 -3.52270067e-01 -3.71740609e-01 -3.38531524e-01
-5.56800887e-02 5.73929667e-01 1.00649595e+00 -5.27177393e-01
-9.23856676e-01 -4.88475025e-01 3.41430724e-01 2.18196407e-01
-9.64127421e-01 3.36407334e-01 -6.31371796e-01 -1.54398456e-01
-1.55172944e-01 -1.07577074e+00 -1.21169381e-01 -3.41514796e-02
-1.93994835e-01 -4.88916993e-01 7.09987342e-01 -3.90930235e-01
1.12319314e+00 -2.26124907e+00 -7.85402954e-02 3.64242494e-01
6.10399507e-02 -1.50336623e-01 -2.22262487e-01 1.20669879e-01
2.19084680e-01 1.67094365e-01 1.09429918e-01 -5.72236538e-01
1.59045890e-01 2.72669673e-01 -2.59645343e-01 3.84692311e-01
1.42754614e-01 4.91825730e-01 -5.36178887e-01 -1.46310583e-01
-2.04389185e-01 4.35516149e-01 -9.03268039e-01 4.21809644e-01
-2.42873758e-01 -2.03179166e-01 -5.24807990e-01 1.26200724e+00
4.17244494e-01 -6.00584745e-01 2.72255898e-01 -2.30818734e-01
1.38872311e-01 6.21591508e-01 -1.20823610e+00 1.44516611e+00
-8.32436085e-01 4.59602505e-01 5.05788445e-01 -8.72814178e-01
6.28390849e-01 2.85054475e-01 8.31481218e-01 -6.28418684e-01
1.10386149e-03 4.10566330e-01 -3.20185833e-02 -4.10992205e-01
6.97525293e-02 2.16151804e-01 -9.84151959e-02 6.84651613e-01
-6.83298633e-02 1.42491460e-01 1.65850341e-01 -1.62151903e-01
1.21632302e+00 -1.89966023e-01 2.23876655e-01 -1.52166829e-01
2.61918485e-01 -2.11939678e-01 6.74544990e-01 5.78763604e-01
-1.76004693e-01 1.57135323e-01 5.45553327e-01 -7.72026181e-01
-7.39226758e-01 -5.31253755e-01 -1.50666460e-01 1.68598342e+00
-3.41092795e-01 -4.92991686e-01 -7.44945407e-01 -1.02849686e+00
3.45362276e-01 3.73640448e-01 -2.24298835e-01 1.09884106e-01
-2.17613146e-01 -1.02235126e+00 1.63930908e-01 2.10316971e-01
3.67882758e-01 -7.77545631e-01 -5.80868959e-01 4.56853099e-02
1.64632410e-01 -1.28707337e+00 -8.48661125e-01 1.55466899e-01
-7.05914557e-01 -9.24288273e-01 -1.90471351e-01 -3.03920537e-01
6.24144852e-01 3.04063529e-01 1.24929476e+00 8.68606940e-02
-3.69073421e-01 4.81086195e-01 -7.16766492e-02 -5.94863236e-01
2.61164736e-02 3.05779040e-01 1.26427084e-01 3.14055413e-01
6.05483592e-01 -6.35686100e-01 -5.49983621e-01 2.83427954e-01
-6.09033227e-01 -1.97348237e-01 5.01403272e-01 4.48566467e-01
4.33278769e-01 9.65879261e-02 4.24583316e-01 -1.08861780e+00
7.19357669e-01 -6.50908709e-01 -8.21686089e-01 1.43300951e-01
-1.11700761e+00 -3.01178377e-02 6.65371299e-01 -5.23898065e-01
-6.30259633e-01 4.00071472e-01 -1.48253918e-01 -4.74582821e-01
-2.77790185e-02 2.77433097e-01 -4.08607274e-01 -2.86236346e-01
2.91389793e-01 -8.68166797e-03 4.07031998e-02 -6.45353436e-01
1.93213046e-01 1.04063368e+00 7.36165908e-04 -6.58756733e-01
2.74536759e-01 5.17711937e-01 -4.00312781e-01 -5.71289003e-01
-1.07135403e+00 -5.47371507e-01 -6.97114840e-02 9.85193104e-02
3.13541353e-01 -8.47079098e-01 -1.10575116e+00 2.06265032e-01
-7.80653000e-01 -1.84019938e-01 6.42708838e-02 -3.85637246e-02
-2.36667797e-01 4.33468111e-02 -3.81232023e-01 -8.09349179e-01
-7.60864198e-01 -1.50954270e+00 1.11903834e+00 2.97610730e-01
-4.24556136e-01 -5.84516108e-01 -4.63025540e-01 7.90935099e-01
6.07544959e-01 -1.85420424e-01 7.49005497e-01 -9.61471021e-01
-6.14136279e-01 -6.38712347e-01 -2.72673666e-01 5.52253246e-01
3.53309005e-01 1.04990385e-01 -1.01962173e+00 -3.30329746e-01
-1.89457424e-02 -4.47058439e-01 3.69354665e-01 2.32882887e-01
1.65576947e+00 -7.77329683e-01 -2.81208962e-01 7.89330542e-01
1.27130580e+00 2.90738136e-01 -9.77504067e-03 2.56328821e-01
5.40325105e-01 4.41818237e-01 4.98402208e-01 6.53828919e-01
4.20908391e-01 8.68450403e-01 5.37461042e-01 2.43976533e-01
3.65371436e-01 3.09637077e-02 6.37106478e-01 6.07105792e-01
1.07602187e-01 8.77763256e-02 -7.65809357e-01 2.36179635e-01
-1.68969512e+00 -6.42094910e-01 5.39890647e-01 2.40259457e+00
7.59577692e-01 3.95754665e-01 3.28356475e-01 -4.08677272e-02
1.99438091e-02 3.96555364e-02 -9.34983492e-01 -7.10711718e-01
1.15482621e-01 2.83358186e-01 6.08772993e-01 1.74424186e-01
-1.01744819e+00 7.27118909e-01 6.87992764e+00 8.97893369e-01
-1.39404094e+00 9.76542532e-02 1.09408009e+00 -5.90918839e-01
-2.12487385e-01 -1.11246541e-01 -9.72619951e-01 8.31689894e-01
1.31765366e+00 -2.61541635e-01 1.15358686e+00 1.40290356e+00
5.50052486e-02 1.89207152e-01 -1.39181948e+00 1.17250419e+00
-4.08851206e-02 -1.42934358e+00 -4.57611382e-02 3.47648978e-01
5.33024609e-01 4.36505854e-01 2.15018302e-01 1.22587785e-01
1.54462680e-01 -1.00486720e+00 4.38908964e-01 1.89218774e-01
4.39299047e-01 -9.05117452e-01 4.68452245e-01 3.81327510e-01
-9.28743601e-01 -4.35545236e-01 -1.31289467e-01 -1.52523264e-01
-2.08735153e-01 5.35739243e-01 -7.81730175e-01 -8.93526524e-02
9.44743991e-01 5.59721053e-01 -2.27416411e-01 6.58317566e-01
1.55154645e-01 8.12075019e-01 -6.71085894e-01 -1.58483252e-01
4.57075909e-02 -1.55349448e-01 2.43436351e-01 1.11935699e+00
2.03946844e-01 -1.69760495e-01 3.51002514e-01 4.14692491e-01
-6.83702826e-01 4.24212843e-01 -4.21209008e-01 -4.78561848e-01
6.44635141e-01 1.53751266e+00 -3.71311337e-01 -2.04155430e-01
-7.58199394e-01 9.09107089e-01 3.72651070e-01 5.66087663e-02
-7.72868037e-01 -2.23303005e-01 1.38141406e+00 4.83848527e-02
9.59237963e-02 1.71418507e-02 -1.09487906e-01 -1.18508351e+00
2.60131031e-01 -1.39907467e+00 5.88592529e-01 -5.02397418e-01
-1.45465553e+00 5.03273726e-01 -3.17068070e-01 -9.57874000e-01
-2.51992255e-01 -7.93700874e-01 -3.41334671e-01 6.15117252e-01
-1.55960846e+00 -9.57599759e-01 -4.22837809e-02 4.99549061e-01
7.28047192e-01 -4.56481308e-01 9.48191345e-01 5.60768604e-01
-7.70912170e-01 8.87064219e-01 -1.48515165e-01 -1.20829239e-01
5.45365334e-01 -8.74099076e-01 1.37375712e-01 3.76559287e-01
4.81204152e-01 7.27457583e-01 5.52930593e-01 -4.92512286e-02
-2.09675002e+00 -6.83930099e-01 7.79523492e-01 -4.47780609e-01
8.04622114e-01 -4.73805815e-01 -6.45516992e-01 7.38905728e-01
1.68424904e-01 1.53392956e-01 1.05352366e+00 4.43561614e-01
-6.91456914e-01 -8.04109633e-01 -1.35029387e+00 5.13556302e-01
8.00562680e-01 -5.63599527e-01 8.76808316e-02 7.48771966e-01
6.98858142e-01 -4.53902870e-01 -1.01764238e+00 3.35000873e-01
8.75234842e-01 -9.63288963e-01 7.25785434e-01 -8.14663708e-01
3.46175462e-01 2.00781912e-01 -5.10301352e-01 -8.02557051e-01
2.35515982e-01 -1.08043110e+00 -4.22840714e-01 1.26711047e+00
6.52672112e-01 -8.29279780e-01 1.08697081e+00 1.48125207e+00
3.30919534e-01 -1.29697478e+00 -8.78817677e-01 -6.36192024e-01
-5.09641886e-01 -8.58955681e-01 9.03282702e-01 8.02999020e-01
-5.45936115e-02 7.37370849e-02 -4.41078901e-01 1.50336012e-01
5.35787880e-01 4.88378316e-01 8.51151943e-01 -1.06841111e+00
-5.47946572e-01 -5.29473543e-01 -1.76718995e-01 -1.06186521e+00
3.65613252e-01 -9.26069200e-01 -4.52275723e-01 -8.82908404e-01
3.00145984e-01 -8.53350580e-01 -5.50132811e-01 9.54626858e-01
2.33262166e-01 9.19561237e-02 1.68046981e-01 3.42010349e-01
-5.80226362e-01 -5.52989244e-02 4.86795425e-01 -4.48741280e-02
-2.98123211e-01 3.66230816e-01 -1.13372564e+00 7.93492794e-01
6.44083321e-01 -4.95428026e-01 -3.47884715e-01 -5.34729898e-01
4.96922940e-01 -2.06215173e-01 -2.70286143e-01 -5.83654404e-01
1.40866920e-01 -4.46581662e-01 -4.24697623e-03 1.78774163e-01
5.44561088e-01 -1.09743106e+00 -5.70949540e-02 6.93469262e-03
-4.28506494e-01 1.91697046e-01 2.69236207e-01 3.02034259e-01
7.44067505e-03 -2.62673348e-02 5.33335924e-01 6.43142238e-02
-4.53876764e-01 6.32423937e-01 -1.37079535e-02 4.18926589e-03
8.41431916e-01 1.07007928e-01 4.27830070e-02 -4.65297103e-01
-2.87307024e-01 1.95369512e-01 3.65153849e-01 5.93813837e-01
3.38903576e-01 -1.02028668e+00 -2.37363413e-01 3.83616477e-01
1.07438691e-01 -3.97464544e-01 -1.80123106e-01 7.43541598e-01
-7.86508620e-02 2.76358455e-01 3.61464098e-02 -4.84435797e-01
-1.58245623e+00 2.66249061e-01 3.04557174e-01 -2.46760964e-01
-1.95862591e-01 1.06438434e+00 -2.35354215e-01 -3.26787442e-01
4.53296036e-01 -2.63876230e-01 2.65755922e-01 -1.12739839e-01
6.70231283e-01 1.76687554e-01 3.95367235e-01 -3.73057485e-01
-3.96617323e-01 2.18406647e-01 -2.67087579e-01 1.34188592e-01
1.50965035e+00 9.72702876e-02 -1.89883411e-01 2.34307289e-01
1.33609438e+00 1.90679133e-01 -1.32116246e+00 -3.26443762e-02
1.91502392e-01 -6.19493425e-01 2.35712409e-01 -7.78319061e-01
-1.44485104e+00 5.63630939e-01 4.47759688e-01 3.79185736e-01
1.34835219e+00 -9.52245817e-02 1.02146578e+00 4.41362888e-01
6.42434180e-01 -9.42081332e-01 -1.42523333e-01 -8.77853017e-03
5.43806136e-01 -1.41143179e+00 6.40518293e-02 -3.58396262e-01
-6.84662163e-01 9.52938795e-01 3.65898639e-01 -1.37684941e-01
8.06899130e-01 6.12311482e-01 -4.01100479e-02 -8.64319503e-02
-1.24660707e+00 2.16130853e-01 4.06945825e-01 8.98349583e-02
5.47911525e-01 2.70044595e-01 -4.99089248e-03 7.34136224e-01
-9.18971598e-02 1.28905838e-02 1.10446364e-01 9.19891357e-01
-1.46092474e-01 -1.39492905e+00 3.47191729e-02 1.04163611e+00
-9.78008270e-01 -7.86952302e-02 -2.13574052e-01 4.38206643e-01
-1.23644248e-02 9.31028068e-01 6.99761286e-02 -4.05654848e-01
2.75929689e-01 3.43324929e-01 2.40098372e-01 -6.33147955e-01
-6.29444957e-01 1.10036798e-01 3.41997355e-01 -1.13838303e+00
-2.24873424e-01 -9.35149848e-01 -1.11494958e+00 -2.26746589e-01
-2.08495393e-01 1.20163094e-02 1.01787078e+00 1.00898290e+00
9.20852363e-01 -1.32518396e-01 9.43828166e-01 -6.26320004e-01
-7.61586070e-01 -4.25435215e-01 -4.21765655e-01 4.03461993e-01
1.24451548e-01 -4.77699488e-01 -5.53011775e-01 -3.85281332e-02] | [9.075872421264648, 4.396977424621582] |
51e06212-d8d5-4238-8fa7-0390ae5ab90d | on-expert-behaviors-and-question-types-for | 2001.05952 | null | https://arxiv.org/abs/2001.05952v2 | https://arxiv.org/pdf/2001.05952v2.pdf | On Expert Behaviors and Question Types for Efficient Query-Based Ontology Fault Localization | We challenge existing query-based ontology fault localization methods wrt. assumptions they make, criteria they optimize, and interaction means they use. We find that their efficiency depends largely on the behavior of the interacting expert, that performed calculations can be inefficient or imprecise, and that used optimization criteria are often not fully realistic. As a remedy, we suggest a novel (and simpler) interaction approach which overcomes all identified problems and, in comprehensive experiments on faulty real-world ontologies, enables a successful fault localization while requiring fewer expert interactions in 66 % of the cases, and always at least 80 % less expert waiting time, compared to existing methods. | ['Patrick Rodler'] | 2020-01-16 | null | null | null | null | ['fault-localization'] | ['computer-code'] | [-2.16556102e-01 3.19438368e-01 8.06946084e-02 -3.18178982e-01
-4.16381687e-01 -6.17629528e-01 1.62416250e-01 5.11764228e-01
-4.31129515e-01 8.45055103e-01 -2.28367120e-01 -3.04070741e-01
-9.32768822e-01 -7.08860397e-01 -4.15347457e-01 3.31641436e-02
-1.58596337e-01 9.94102120e-01 8.11412990e-01 -4.49358165e-01
3.51174831e-01 3.12271953e-01 -1.99129212e+00 6.24224730e-02
1.10388017e+00 1.05571961e+00 -2.94492066e-01 3.30521584e-01
4.02804650e-02 1.09397280e+00 -9.01166797e-01 -2.07356393e-01
-3.97456884e-02 -1.12625370e-02 -1.28884947e+00 -2.53909886e-01
-8.17644596e-02 -4.41905744e-02 -1.55439100e-03 1.06325376e+00
2.63419002e-01 2.39412598e-02 3.66489291e-01 -1.63049519e+00
-3.48357528e-01 5.97847700e-01 2.72070885e-01 1.09741062e-01
9.91681397e-01 -1.76645089e-02 5.59076071e-01 -4.85087901e-01
5.73929608e-01 1.18992829e+00 8.63602400e-01 2.67573893e-01
-1.12929916e+00 -1.80392563e-01 9.74415988e-03 4.97534126e-01
-1.92360580e+00 -6.62898719e-01 2.26722494e-01 -3.30663979e-01
1.43778968e+00 6.42121315e-01 4.14311439e-01 6.61943018e-01
4.61535990e-01 9.90413129e-02 6.24439359e-01 -6.77165926e-01
6.04606867e-01 2.49620765e-01 2.15277895e-01 6.96376026e-01
6.52261615e-01 -2.53240049e-01 -4.71649349e-01 -7.98601151e-01
1.86815768e-01 -1.32889658e-01 -4.21162844e-01 -3.43750983e-01
-9.29114103e-01 2.01981932e-01 -1.81371927e-01 7.03045726e-01
-1.92305237e-01 1.53558552e-01 3.41032952e-01 6.78337395e-01
1.74528629e-01 7.34761775e-01 -6.85667157e-01 -3.35105032e-01
-4.69945908e-01 4.65769261e-01 1.35147965e+00 8.68376374e-01
7.72609055e-01 -2.21311152e-01 2.31694505e-01 5.63454986e-01
3.32650393e-01 1.80238843e-01 3.41914982e-01 -1.19911528e+00
1.42874926e-01 1.17333090e+00 5.49168527e-01 -8.91856968e-01
-4.28029329e-01 -8.64119604e-02 -3.70228365e-02 6.06506169e-01
2.27126598e-01 1.97324827e-01 -5.62820792e-01 1.28176391e+00
1.32586777e-01 -2.90004522e-01 1.01508819e-01 5.30151129e-01
1.12432748e-01 -8.27271044e-02 4.52048071e-02 -3.09984684e-01
1.24535775e+00 -5.27618647e-01 -9.30901408e-01 -1.37794897e-01
8.99764776e-01 -6.23084247e-01 1.01499784e+00 5.76480210e-01
-1.07426250e+00 -1.49433404e-01 -1.16089857e+00 2.62867868e-01
-6.43376172e-01 -3.53876710e-01 4.89005774e-01 6.77608013e-01
-1.22340322e+00 8.22708488e-01 -7.58274138e-01 -6.90551758e-01
-8.69931281e-02 6.23973191e-01 -3.26756239e-01 -1.53323740e-01
-1.12521601e+00 1.30744457e+00 6.17311776e-01 -1.86116993e-01
-5.40942967e-01 -5.46013474e-01 -6.44728243e-01 1.32310718e-01
7.84737170e-01 -6.05556309e-01 1.35688138e+00 -6.72804594e-01
-1.31627953e+00 3.79861623e-01 -2.01933999e-02 -3.05881083e-01
7.31243312e-01 -3.60197872e-01 -8.33665907e-01 -6.30074739e-02
2.04413518e-01 5.10932617e-02 3.15230489e-01 -1.15237451e+00
-6.13210917e-01 -2.99684733e-01 4.74000961e-01 1.67319953e-01
-2.81500936e-01 7.29792565e-02 -2.07417503e-01 -1.29176423e-01
2.11451560e-01 -5.29649734e-01 -2.94860780e-01 6.34846538e-02
-1.66285127e-01 -4.32050943e-01 7.63431013e-01 -3.83453578e-01
1.46049464e+00 -2.27305102e+00 1.09119006e-01 5.13584435e-01
1.59800753e-01 -3.62731256e-02 2.80301929e-01 9.41411197e-01
-4.89941835e-02 3.26544106e-01 -2.44509235e-01 1.22127399e-01
2.54963636e-01 5.13272583e-01 -6.91423342e-02 2.84858823e-01
-1.19916454e-01 2.73946434e-01 -8.18059385e-01 -7.05539644e-01
2.02202842e-01 1.53874740e-01 -4.72790241e-01 6.86679259e-02
-3.05801719e-01 5.68429939e-02 -4.57961529e-01 9.82957780e-01
1.03521571e-01 -2.87494689e-01 5.61831176e-01 -1.02127708e-01
-1.60075743e-02 2.30714798e-01 -1.59204876e+00 1.30728245e+00
-3.52554172e-01 9.99785066e-02 -2.78210878e-01 -9.07428682e-01
5.32422543e-01 7.90442765e-01 5.60938656e-01 -7.74351597e-01
4.12958115e-03 6.63488746e-01 -9.35281441e-02 -9.41928089e-01
1.66945562e-01 4.35156494e-01 2.06562802e-02 4.73718554e-01
-5.50222360e-02 -6.98934719e-02 4.03733552e-01 -1.43406942e-01
1.77456689e+00 2.08072904e-02 6.29851222e-01 -4.94217336e-01
4.54943985e-01 3.04548949e-01 6.10315442e-01 8.64369094e-01
-2.10443705e-01 2.95919329e-01 5.21329582e-01 -6.94229782e-01
-5.16082942e-01 -5.72068989e-01 -1.82163239e-01 8.75504255e-01
4.28525954e-01 -8.17296445e-01 -6.45740449e-01 -8.49309266e-01
9.18796882e-02 8.91943634e-01 -3.83372486e-01 -2.84879714e-01
-3.22687477e-01 -3.99081498e-01 7.89322853e-01 3.76424521e-01
2.94890888e-02 -9.07346845e-01 -9.20608997e-01 5.02566993e-01
-8.55960771e-02 -1.08264792e+00 2.29098335e-01 1.88940287e-01
-8.82387519e-01 -1.62183487e+00 1.60185352e-01 -2.09136844e-01
7.47095406e-01 -1.89166769e-01 1.41829765e+00 8.90401602e-01
-3.64640117e-01 5.12015462e-01 -5.21680832e-01 -2.36803994e-01
-3.68920207e-01 -1.34760663e-01 9.01901871e-02 -4.41363126e-01
3.83478045e-01 -6.80609047e-01 -1.93434089e-01 4.93438065e-01
-9.68831301e-01 -6.25397682e-01 3.38766932e-01 5.30311346e-01
5.92860729e-02 7.74181187e-01 4.67809290e-01 -1.08005571e+00
7.00566053e-01 -5.82040191e-01 -5.77019811e-01 6.85245991e-01
-1.42891312e+00 2.46315911e-01 3.17789316e-01 -1.15933239e-01
-8.32390726e-01 -1.16598569e-01 -5.68357594e-02 3.26975398e-02
-3.20877463e-01 5.32766223e-01 -1.68296248e-01 -5.01757443e-01
1.15218782e+00 -2.72973090e-01 -2.41566464e-01 -4.27270263e-01
-1.15668535e-01 4.67965037e-01 2.58268535e-01 -8.48792434e-01
5.70616484e-01 2.99737126e-01 -2.27871403e-01 -3.73763889e-01
-1.22709014e-01 -1.27790347e-01 -2.89347082e-01 -1.31436944e-01
3.42890650e-01 -4.64202285e-01 -1.01485825e+00 1.32323503e-01
-1.01486278e+00 -2.07651615e-01 -4.20065373e-01 1.67961746e-01
-3.43401223e-01 6.05772913e-01 -3.06502342e-01 -1.08199978e+00
3.71237472e-02 -1.19459927e+00 9.08702672e-01 -1.00299865e-01
-6.63904250e-01 -8.41147900e-01 -8.80409926e-02 2.12870613e-01
6.13160729e-01 2.23003060e-01 1.18121409e+00 -8.43777359e-01
-4.87339556e-01 -5.75881124e-01 1.37741491e-02 1.53685704e-01
2.96169162e-01 2.47667864e-01 -6.28877342e-01 -5.05877584e-02
-1.50901496e-01 -9.23321769e-02 -1.33084893e-01 -3.04121524e-01
8.52875352e-01 -2.79213369e-01 -6.05701983e-01 -1.88168526e-01
1.59673357e+00 3.18870306e-01 5.84809959e-01 5.83663702e-01
2.26602882e-01 4.90190476e-01 7.56468058e-01 4.67009157e-01
4.71386582e-01 8.16608965e-01 5.21536589e-01 5.32095671e-01
2.02281907e-01 4.37581778e-01 2.48902664e-01 4.34239715e-01
-3.22977334e-01 -2.96957076e-01 -1.22395790e+00 6.31187916e-01
-2.31940866e+00 -6.09043181e-01 -7.16477707e-02 2.14859247e+00
7.93353856e-01 3.73421997e-01 -1.87955096e-01 6.22227848e-01
3.81406546e-01 -4.22179401e-01 -3.23824495e-01 -3.71745005e-02
1.86780572e-01 1.94967985e-01 3.52901071e-01 6.85739219e-01
-6.62185431e-01 5.08729696e-01 7.71915388e+00 2.24656865e-01
-4.71033633e-01 2.72394121e-01 -2.74238557e-01 1.36145622e-01
-3.59625757e-01 3.06239188e-01 -1.59816787e-01 4.56415504e-01
9.85034525e-01 -2.68340677e-01 7.00379372e-01 9.63138163e-01
-2.00786635e-01 -4.94109511e-01 -1.34668863e+00 6.58563197e-01
-1.01206675e-01 -8.95648777e-01 -3.15122396e-01 -9.93190706e-02
2.07561642e-01 -2.55509406e-01 -6.90626800e-01 9.25000086e-02
3.81000280e-01 -8.36494684e-01 1.02097273e+00 9.52479839e-01
1.43003002e-01 -6.40117705e-01 1.21737182e+00 2.72653550e-01
-8.98229718e-01 -4.77381855e-01 -1.83862578e-02 -2.84170061e-01
-6.82317764e-02 8.80459964e-01 -4.16926056e-01 1.09282362e+00
1.22929800e+00 6.11351104e-04 -6.92612648e-01 1.14964032e+00
-1.09855987e-01 1.26066282e-01 -7.28245378e-01 1.64628595e-01
-3.96903753e-01 3.59466136e-01 4.44372535e-01 8.01242113e-01
3.35120738e-01 -1.31189981e-02 3.02931547e-01 3.77638936e-01
4.68167692e-01 9.69427824e-02 -4.99368876e-01 1.60059314e-02
9.89106417e-01 7.69731939e-01 -6.36141717e-01 -3.81559819e-01
-2.46222451e-01 8.35596144e-01 2.08036602e-01 4.51420784e-01
-6.67608500e-01 -6.27428412e-01 6.26069367e-01 1.88887492e-01
-2.38719910e-01 -5.69043085e-02 -2.85802577e-02 -9.26705182e-01
5.11062026e-01 -1.12074745e+00 4.94285226e-01 -7.70933628e-01
-1.06272399e+00 7.79287457e-01 1.72333449e-01 -8.89636219e-01
-5.02119899e-01 -5.03653765e-01 -3.49880517e-01 4.64690715e-01
-9.23113883e-01 -6.17668629e-01 -4.11964029e-01 4.69968110e-01
4.45816442e-02 9.10511166e-02 1.15147126e+00 7.23471045e-01
-4.88859266e-01 2.80442685e-01 -4.17962193e-01 -5.82014024e-01
7.12180257e-01 -1.19128537e+00 -1.28025144e-01 7.51485825e-01
-1.57559648e-01 5.64717174e-01 1.03615141e+00 -6.28068686e-01
-1.34398246e+00 -6.56398952e-01 1.09705555e+00 -7.21209347e-01
4.99861777e-01 4.92662638e-02 -9.60344791e-01 7.92337596e-01
1.55728266e-01 -5.83401732e-02 4.71646816e-01 1.43655628e-01
-8.50290135e-02 -2.10406989e-01 -1.54406381e+00 4.71275240e-01
1.23647714e+00 -6.25490487e-01 -6.64609730e-01 7.09721029e-01
5.58968186e-01 -3.40682447e-01 -1.19092333e+00 7.08262742e-01
4.11534607e-01 -1.21374023e+00 7.03715026e-01 -4.07357544e-01
-4.87508029e-01 -7.62799144e-01 -4.03603107e-01 -8.69689226e-01
-1.77554995e-01 -3.46065283e-01 -2.84174412e-01 1.16024673e+00
3.47380131e-01 -1.06098902e+00 2.02180386e-01 1.08602417e+00
-1.86696738e-01 -6.16565764e-01 -9.53258216e-01 -8.73857081e-01
-9.15112495e-01 -5.23975790e-01 7.81558514e-01 1.12761068e+00
3.53710771e-01 -2.31697574e-01 1.54327214e-01 6.44437909e-01
5.10118484e-01 -3.97839934e-01 3.22393596e-01 -1.66861784e+00
-1.83631539e-01 -1.39941379e-01 -6.33680582e-01 -2.45378390e-02
3.54779921e-02 -2.66905457e-01 1.28000587e-01 -1.53607404e+00
-3.05756658e-01 -6.08874202e-01 -3.69673431e-01 9.33750153e-01
7.33918995e-02 3.33898850e-02 -5.12817383e-01 3.82666945e-01
-8.22336376e-01 8.84759799e-02 2.89189667e-01 6.68049902e-02
7.21351355e-02 -2.88043410e-01 -6.20251715e-01 7.43501544e-01
4.90419924e-01 -6.57936633e-01 -5.53703189e-01 -4.60031420e-01
7.67179787e-01 -1.47503773e-02 4.49592173e-01 -1.54359305e+00
5.89036644e-01 -2.34019995e-01 -3.68733555e-02 1.24153597e-02
-1.20933738e-03 -1.35329306e+00 7.66941130e-01 5.05998254e-01
9.89823788e-02 2.95732617e-01 2.35904171e-03 3.42257231e-01
-2.58423477e-01 -7.79806137e-01 2.47533828e-01 -1.13140285e-01
-6.88365638e-01 -3.33308518e-01 -6.03914559e-01 -2.57556379e-01
1.12809968e+00 -1.39521241e-01 -3.33250135e-01 -2.09488943e-02
-8.35810542e-01 2.01356158e-01 9.09585953e-01 3.79563779e-01
5.84409200e-02 -9.76690710e-01 -7.78404996e-02 -2.24248488e-02
3.81789327e-01 -2.00649366e-01 -8.28929432e-03 9.23994839e-01
-7.31870770e-01 2.47436345e-01 -5.99960536e-02 -3.91928792e-01
-9.23299551e-01 4.99876350e-01 5.43250084e-01 1.55607136e-02
-4.74221230e-01 4.40646857e-01 -5.73612511e-01 -3.91979754e-01
4.29018825e-01 -2.83567756e-01 -1.64152328e-02 -3.32686424e-01
2.73172110e-01 7.13658869e-01 6.42710090e-01 -6.74272776e-02
-7.71454096e-01 4.75801378e-01 1.74153164e-01 1.38137797e-02
1.09941149e+00 -7.74551928e-02 -4.85449493e-01 5.10329723e-01
1.45147622e-01 3.49899791e-02 -5.53342044e-01 6.07018126e-03
5.14834106e-01 -4.17076260e-01 -2.30023324e-01 -9.57573295e-01
-7.03861892e-01 1.16220295e-01 8.14757824e-01 9.55988109e-01
1.06626177e+00 -4.41112369e-02 2.63608962e-01 9.51708138e-01
9.45747495e-01 -1.21640313e+00 -3.60132605e-01 1.25397086e-01
7.84042478e-01 -8.94155920e-01 1.46030560e-01 -6.94756627e-01
-1.84325784e-01 1.04946101e+00 7.97053218e-01 7.96635523e-02
6.50695145e-01 4.68573242e-01 1.88064218e-01 -5.32375097e-01
-8.99559915e-01 9.64592583e-03 -2.75664002e-01 5.61978698e-01
2.52069473e-01 -2.03167230e-01 -5.22429109e-01 5.27235985e-01
2.12853312e-01 2.97221214e-01 3.02905142e-01 1.57294381e+00
-4.32293653e-01 -1.47547829e+00 -4.77506548e-01 3.54386985e-01
-3.82896155e-01 3.59373957e-01 -2.97967225e-01 9.57991779e-01
3.94224674e-01 1.18362713e+00 -2.72829026e-01 -3.32143635e-01
9.55423117e-01 4.81802881e-01 4.50554550e-01 -6.28406048e-01
-5.12556553e-01 -6.70233130e-01 5.82079053e-01 -9.80923712e-01
-2.40938202e-01 -4.72033620e-01 -1.44998312e+00 -2.40476117e-01
-5.59727728e-01 4.92206454e-01 3.74560684e-01 1.39582610e+00
6.93478644e-01 6.68438077e-01 1.85464963e-01 -3.43431473e-01
-5.05512536e-01 -8.28454673e-01 -3.73274416e-01 2.47765735e-01
-1.26970202e-01 -1.08327866e+00 -3.96376610e-01 -3.33505054e-03] | [5.49169921875, 2.808230400085449] |
cfea139a-4165-4f65-8495-e3b1678fd2cc | learning-to-execute-efficient-learning-of | 2111.07908 | null | https://arxiv.org/abs/2111.07908v1 | https://arxiv.org/pdf/2111.07908v1.pdf | Learning to Execute: Efficient Learning of Universal Plan-Conditioned Policies in Robotics | Applications of Reinforcement Learning (RL) in robotics are often limited by high data demand. On the other hand, approximate models are readily available in many robotics scenarios, making model-based approaches like planning a data-efficient alternative. Still, the performance of these methods suffers if the model is imprecise or wrong. In this sense, the respective strengths and weaknesses of RL and model-based planners are. In the present work, we investigate how both approaches can be integrated into one framework that combines their strengths. We introduce Learning to Execute (L2E), which leverages information contained in approximate plans to learn universal policies that are conditioned on plans. In our robotic manipulation experiments, L2E exhibits increased performance when compared to pure RL, pure planning, or baseline methods combining learning and planning. | ['Marc Toussaint', 'Ozgur S. Oguz', 'Danny Driess', 'Ingmar Schubert'] | 2021-11-15 | null | http://proceedings.neurips.cc/paper/2021/hash/0e9b734aa25ca8096cb7b56dc0dd8929-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/0e9b734aa25ca8096cb7b56dc0dd8929-Paper.pdf | neurips-2021-12 | ['learning-to-execute'] | ['computer-code'] | [-8.60256106e-02 5.24991274e-01 -6.45806372e-01 -7.91110620e-02
-7.09220529e-01 -4.95523959e-01 8.54784727e-01 2.12949499e-01
-5.47893703e-01 1.02771616e+00 2.86246628e-01 -1.54893205e-01
-3.30834538e-01 -9.19659734e-01 -7.97668099e-01 -2.94639587e-01
-2.85906792e-02 8.24342728e-01 3.23588043e-01 -3.42706382e-01
3.96014512e-01 7.70163953e-01 -1.53384995e+00 4.79160957e-02
7.89012671e-01 8.14561188e-01 4.47324455e-01 2.76981920e-01
-1.20733380e-01 1.05271757e+00 -1.26014456e-01 2.65814751e-01
5.01612663e-01 -1.97250605e-01 -8.93061578e-01 -3.41861807e-02
-2.28858575e-01 -6.89920127e-01 -4.20083165e-01 7.96226382e-01
2.28339493e-01 1.89501300e-01 6.66698337e-01 -1.17951405e+00
8.19799155e-02 6.83217943e-01 -1.79533198e-01 -4.39417899e-01
8.36355507e-01 4.78637874e-01 7.09057868e-01 -5.39759934e-01
7.66324639e-01 1.41283536e+00 5.29700458e-01 5.14406860e-01
-1.29944861e+00 -1.53724268e-01 2.25562230e-01 1.60003081e-01
-1.00716674e+00 -4.43161666e-01 7.08031476e-01 -3.07258278e-01
1.08663607e+00 -2.57083207e-01 6.87975526e-01 9.95994449e-01
3.68645340e-01 1.02350724e+00 1.41476524e+00 -4.11816478e-01
6.59177601e-01 -1.17173210e-01 -2.90873021e-01 6.49499893e-01
2.88051188e-01 8.50886583e-01 -5.41556776e-01 -6.28931150e-02
8.92951131e-01 -1.16850242e-01 -4.62894700e-02 -7.11182654e-01
-1.33892965e+00 6.71853125e-01 4.49948639e-01 7.47033358e-02
-6.95189655e-01 3.32064182e-01 2.68583566e-01 3.79369140e-01
-1.22291602e-01 9.11764085e-01 -2.62817532e-01 -4.47073251e-01
-7.69082129e-01 8.62581134e-01 1.03164423e+00 1.15840220e+00
1.01275539e+00 -4.14709710e-02 -9.67918038e-02 3.50693375e-01
3.56990159e-01 1.71728700e-01 5.26572824e-01 -1.47261071e+00
2.80501038e-01 5.05088210e-01 5.42495072e-01 -6.75176680e-01
-7.11688042e-01 1.80281445e-01 -2.98939228e-01 6.47085786e-01
4.96049523e-01 -1.00597873e-01 -9.18870628e-01 1.55743837e+00
2.63983399e-01 5.15277162e-02 2.31070295e-01 7.26368368e-01
2.30800152e-01 4.37078595e-01 -2.26782039e-02 -2.15595007e-01
6.84609771e-01 -1.14521945e+00 -6.49849951e-01 -5.24573207e-01
5.54909408e-01 -4.28914666e-01 1.07547987e+00 6.39725506e-01
-1.09660923e+00 -4.76599813e-01 -1.05555916e+00 7.19075575e-02
-2.83100843e-01 -2.68290639e-01 7.32886732e-01 2.69857287e-01
-1.03052568e+00 1.14342487e+00 -1.24682546e+00 -3.57544124e-01
2.49256492e-01 4.37799484e-01 -2.87873328e-01 -3.64874393e-01
-7.02192605e-01 1.53524387e+00 9.35505986e-01 -2.06940114e-01
-1.19892454e+00 -2.20180929e-01 -9.03004289e-01 -1.96712971e-01
9.08421814e-01 -3.71140659e-01 1.79534352e+00 -3.65127325e-01
-2.00030112e+00 1.02115281e-01 1.96119711e-01 -6.80595517e-01
7.76911259e-01 -5.35405576e-01 2.46167690e-01 -7.30744377e-02
-1.07514933e-01 7.06625462e-01 4.64987189e-01 -1.44450939e+00
-5.73518455e-01 -1.37210518e-01 4.84787077e-01 1.80896595e-01
3.22775990e-01 -3.97176623e-01 -2.24522222e-02 -1.80655435e-01
1.84951618e-01 -1.02162075e+00 -7.09789336e-01 2.19504703e-02
-1.44600719e-01 -3.13883930e-01 5.73101759e-01 -2.54922688e-01
8.98377895e-01 -1.73153985e+00 4.22616839e-01 1.38362065e-01
-2.85051525e-01 1.88137755e-01 -3.29594046e-01 9.82956707e-01
4.49916780e-01 -1.21511929e-01 -1.57800272e-01 -8.02156553e-02
3.91081125e-01 9.21901345e-01 -3.53894919e-01 3.20231289e-01
1.64792642e-01 9.05991256e-01 -1.40425360e+00 -4.84519571e-01
5.31332254e-01 6.64548390e-03 -6.19971752e-01 3.35553527e-01
-8.32704544e-01 7.57999003e-01 -7.50636697e-01 6.49415493e-01
2.07494125e-01 3.14648330e-01 5.25212586e-01 2.49174193e-01
-2.95411885e-01 5.69060326e-01 -1.28472316e+00 2.09343004e+00
-6.28309488e-01 1.81361625e-03 -1.30683571e-01 -9.23996747e-01
8.45149457e-01 3.66156250e-01 5.40491283e-01 -5.62916815e-01
-1.08352564e-01 5.12776554e-01 -7.75548816e-02 -5.04934549e-01
5.46356142e-01 -1.76059067e-01 -1.54775515e-01 5.28254569e-01
1.64636225e-01 -8.58271778e-01 2.76288033e-01 -2.32903302e-01
1.26136220e+00 1.14433813e+00 8.84258389e-01 -2.67436057e-02
2.16188639e-01 3.44745994e-01 6.46259367e-01 1.02635098e+00
-2.22543344e-01 2.49699011e-01 5.49755037e-01 -5.11848629e-01
-9.24528718e-01 -9.42114532e-01 1.88361883e-01 6.33857071e-01
2.29591236e-01 -5.92266023e-01 -5.11628151e-01 -7.37484515e-01
-5.90575114e-03 9.12709951e-01 -2.66963810e-01 -8.10975209e-02
-8.19728494e-01 -1.31160423e-01 3.59592438e-01 4.32499260e-01
4.42859977e-01 -1.47603846e+00 -1.20624495e+00 5.97693443e-01
-3.99814434e-02 -1.03250456e+00 2.09137052e-01 2.41229892e-01
-1.12107384e+00 -1.12252223e+00 -3.74918699e-01 -3.48271519e-01
3.37917984e-01 8.18540528e-02 1.16009486e+00 -7.93702155e-02
3.30326229e-01 5.74361384e-01 -6.17863238e-01 -5.84419668e-01
-6.52563214e-01 1.30336091e-01 2.69869119e-01 -7.01241016e-01
-5.40077537e-02 -7.93568909e-01 -1.38031945e-01 6.85186237e-02
-7.82786489e-01 -1.99370813e-02 9.70881164e-01 8.07464063e-01
6.16136074e-01 6.34215213e-03 5.37228048e-01 -6.84074521e-01
7.02686906e-01 -3.23398411e-01 -6.70269012e-01 1.76619872e-01
-8.28177571e-01 4.71436203e-01 5.52092552e-01 -3.51411998e-01
-1.00740230e+00 3.40795100e-01 -1.32355601e-01 -2.64472574e-01
-5.35958588e-01 8.29589546e-01 -4.81396690e-02 -1.97306201e-02
6.75965786e-01 1.03254646e-01 1.02056265e-01 -4.84381557e-01
4.21969742e-01 3.54679167e-01 4.06893134e-01 -1.04565692e+00
6.01820171e-01 1.98020816e-01 1.93122685e-01 -5.30687690e-01
-7.01522171e-01 -1.52768925e-01 -7.40633905e-01 -2.49762058e-01
3.64359409e-01 -5.22572577e-01 -5.31046748e-01 2.27989614e-01
-1.17049491e+00 -8.82945657e-01 -7.29369104e-01 6.93301976e-01
-1.43114614e+00 3.62900704e-01 -4.59565043e-01 -1.07301557e+00
2.89470822e-01 -1.47079837e+00 9.09816027e-01 1.11561202e-01
-3.28917444e-01 -5.84301233e-01 1.68350458e-01 -2.12704360e-01
2.93166578e-01 5.79183340e-01 7.15497971e-01 -6.15423858e-01
-6.67569876e-01 -1.62880480e-01 1.03088923e-01 1.10278353e-02
6.44349903e-02 -4.16750878e-01 -7.99625814e-01 -1.90401286e-01
-1.32310823e-01 -7.86845624e-01 4.41012949e-01 2.52058655e-01
8.16438198e-01 -4.82147068e-01 -2.88415074e-01 8.07945579e-02
1.44823146e+00 2.00523302e-01 5.94643474e-01 7.06843793e-01
1.92891479e-01 8.94802511e-01 1.24181890e+00 5.80289841e-01
3.30871075e-01 8.85729969e-01 7.09500790e-01 5.86654007e-01
7.71056348e-03 -6.44796968e-01 4.75857854e-01 4.53276396e-01
-4.27542180e-01 1.86051264e-01 -1.08622932e+00 3.80485624e-01
-2.34129167e+00 -8.81743729e-01 2.84598529e-01 2.18105698e+00
8.28031540e-01 2.14222968e-01 2.18318492e-01 6.39786646e-02
1.17915839e-01 7.66499490e-02 -7.23198473e-01 -3.93048972e-01
3.54521513e-01 2.39783019e-01 4.07150090e-01 4.89349097e-01
-8.76681387e-01 9.52130139e-01 6.81045246e+00 6.21948123e-01
-7.53263474e-01 -9.86452922e-02 -1.54135853e-01 9.78902876e-02
-1.33593544e-01 2.35577613e-01 -4.57198709e-01 2.27642015e-01
1.00284290e+00 -7.65404031e-02 8.48110795e-01 1.14319777e+00
2.77360886e-01 -4.60270017e-01 -1.50935423e+00 5.74201763e-01
-3.99389088e-01 -1.18874729e+00 -1.09969586e-01 2.77014285e-01
5.52468121e-01 1.75878108e-01 -2.29700208e-01 7.07852900e-01
8.21659803e-01 -1.08896530e+00 9.57871616e-01 5.77713549e-01
3.50224584e-01 -7.68779218e-01 7.24475265e-01 1.05386543e+00
-8.19733799e-01 -3.90258849e-01 -4.56906646e-01 -5.07532597e-01
4.20539111e-01 2.42902443e-01 -9.65891600e-01 8.83138776e-01
3.62645835e-01 6.16690576e-01 -6.75918236e-02 1.07488215e+00
-6.68128908e-01 1.89407468e-01 -4.33936685e-01 -6.47389293e-02
5.13448000e-01 -3.03258836e-01 5.35317481e-01 8.30278933e-01
2.75766820e-01 -1.83607321e-02 7.59714782e-01 8.62913132e-01
3.42934966e-01 -2.37651065e-01 -1.01591730e+00 -6.11418076e-02
5.99832296e-01 8.56194377e-01 -5.78498244e-01 -1.39084116e-01
-3.47539783e-01 5.73608935e-01 6.30104244e-01 2.22531185e-01
-5.37707925e-01 5.65188825e-02 4.30414081e-01 -1.44855440e-01
2.18898170e-02 -6.04896665e-01 -1.34874597e-01 -8.12212348e-01
3.60863693e-02 -9.98451769e-01 4.93266135e-02 -6.36402190e-01
-8.34352374e-01 2.71724194e-01 4.00712252e-01 -1.30206847e+00
-8.24513197e-01 -6.54051423e-01 -2.73053408e-01 3.95277083e-01
-1.85974228e+00 -1.03766453e+00 -1.12716697e-01 1.25002831e-01
6.33364022e-01 1.84093919e-02 1.08147216e+00 -3.85464042e-01
-1.05280004e-01 7.12232839e-04 -9.22541767e-02 -2.08872825e-01
4.23305839e-01 -1.30740786e+00 1.59265593e-01 5.95852673e-01
1.72125712e-01 3.71112972e-01 7.92487442e-01 -6.33895993e-01
-1.59611762e+00 -8.51506352e-01 5.23474813e-01 -2.60954857e-01
5.56196153e-01 1.24991938e-01 -7.29792416e-01 8.74529362e-01
-7.11722067e-03 -1.28847390e-01 1.49991691e-01 5.87172434e-02
-1.23625576e-01 3.34527850e-01 -1.30433726e+00 7.82038629e-01
9.17232573e-01 -2.91644573e-01 -9.96719897e-01 2.98928261e-01
6.27627134e-01 -7.82559872e-01 -1.02094650e+00 5.69394410e-01
7.05743670e-01 -1.21454179e+00 8.67464066e-01 -5.46457708e-01
5.16910911e-01 -2.65841395e-01 -2.18446761e-01 -1.48437202e+00
-6.06773570e-02 -6.93450630e-01 -4.47117507e-01 8.44617486e-01
1.99895099e-01 -5.77340841e-01 7.05505848e-01 6.25484586e-01
-3.31649840e-01 -8.81237268e-01 -9.95950699e-01 -1.08800304e+00
2.14932173e-01 -5.49717367e-01 7.88581133e-01 5.21261454e-01
4.39595103e-01 -6.88900128e-02 -4.18708235e-01 4.09109108e-02
3.38077784e-01 1.14678368e-01 9.69507456e-01 -1.11378229e+00
-4.64426577e-01 -4.02313709e-01 -2.81257153e-01 -1.10393262e+00
4.13620830e-01 -6.54664278e-01 5.59448421e-01 -1.79082763e+00
-3.58030915e-01 -8.52865338e-01 -1.30702779e-01 6.26018465e-01
3.01616609e-01 -4.79726523e-01 5.36406636e-01 2.87900090e-01
-4.71128404e-01 7.48503387e-01 1.32682931e+00 -2.62571909e-02
-5.24090111e-01 6.28885254e-02 -1.79111615e-01 1.06377923e+00
1.13852274e+00 -3.58871132e-01 -3.69540364e-01 -3.31753671e-01
1.75261766e-01 4.38031793e-01 7.63447881e-02 -1.25966656e+00
4.64356512e-01 -6.24440134e-01 -1.13414526e-01 -3.25886846e-01
4.42590326e-01 -9.14198160e-01 2.18205107e-03 7.03726292e-01
-4.05617267e-01 9.58734229e-02 1.60880566e-01 9.17601049e-01
-2.37057418e-01 -5.55320561e-01 5.07641077e-01 -5.25976419e-01
-9.58174706e-01 9.18102711e-02 -5.70595026e-01 -1.77945033e-01
1.03193760e+00 -1.47730768e-01 7.07257614e-02 -4.32433605e-01
-4.93158966e-01 1.88195542e-01 5.88647485e-01 2.10387811e-01
6.02514684e-01 -1.21595395e+00 -4.09659207e-01 -1.46212354e-01
8.34024176e-02 4.67851579e-01 -2.67022491e-01 8.63777637e-01
-4.70088303e-01 4.51963723e-01 -4.99617726e-01 -3.17941248e-01
-4.35075551e-01 7.82918453e-01 2.71488845e-01 -5.96430838e-01
-6.88638389e-01 1.87686995e-01 -3.12966526e-01 -1.06384182e+00
3.68470877e-01 -5.28144538e-01 -7.51341209e-02 -3.44275624e-01
3.49638641e-01 5.75249135e-01 -2.16371089e-01 -2.92347401e-01
-7.76233226e-02 2.70518303e-01 2.02107400e-01 -4.40014392e-01
1.43531704e+00 8.61140341e-02 6.25456795e-02 6.02312922e-01
2.48888701e-01 -2.47922540e-01 -1.58498049e+00 -3.50197345e-01
6.46166801e-01 -3.17222208e-01 -1.36633590e-01 -8.03652167e-01
-5.30487061e-01 6.74448550e-01 1.32898018e-01 -3.97654884e-02
7.96414852e-01 -8.73571411e-02 4.79483992e-01 8.84178519e-01
1.23132455e+00 -1.26456928e+00 7.86335096e-02 6.76128626e-01
1.04667020e+00 -1.13546872e+00 2.68447995e-01 -2.33264223e-01
-6.13593400e-01 1.31248116e+00 6.61246777e-01 -4.14302647e-01
2.85222977e-01 3.62537146e-01 -1.55870482e-01 4.43601273e-02
-7.37023592e-01 -5.50610006e-01 -1.28234819e-01 9.50546801e-01
4.36635464e-02 -2.27120388e-02 -3.48937660e-01 3.29848617e-01
-2.24844575e-01 4.38874900e-01 7.17618942e-01 1.57049131e+00
-7.07943201e-01 -1.43009007e+00 -3.42840821e-01 2.19810292e-01
-5.29679842e-02 4.20779109e-01 -2.29289159e-01 1.13330758e+00
-1.05113439e-01 1.09106410e+00 -3.34632754e-01 -3.46063495e-01
5.34843087e-01 2.80676447e-02 8.30999196e-01 -6.75648928e-01
-3.26979339e-01 -1.59431100e-01 2.84073830e-01 -1.12920356e+00
-6.34705365e-01 -7.19802678e-01 -1.41479731e+00 -2.13956251e-03
-1.06401160e-01 -7.67893046e-02 5.30915856e-01 1.21044433e+00
1.89294994e-01 3.54273677e-01 2.43376374e-01 -1.37838984e+00
-1.18389428e+00 -8.35532784e-01 -4.98235673e-01 -1.10127494e-01
2.07327574e-01 -1.04204869e+00 3.95856611e-02 -2.99683452e-01] | [4.319046497344971, 1.3514196872711182] |
2f5d1ae3-65f3-4001-81bd-d37967a781a1 | experiments-on-hybrid-corpus-based-sentiment | null | null | https://aclanthology.org/W12-0501 | https://aclanthology.org/W12-0501.pdf | Experiments on Hybrid Corpus-Based Sentiment Lexicon Acquisition | null | ["Bojana Dalbelo Ba{\\v{s}}i{\\'c}", 'Jan {\\v{S}}najder', 'Goran Glava{\\v{s}}'] | 2012-04-01 | null | null | null | ws-2012-4 | ['subjectivity-analysis'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.391997814178467, 3.721400499343872] |
c0d6d828-55e5-470a-a6f1-f61f8f9e85a3 | score-and-lyrics-free-singing-voice-1 | 1912.11747 | null | https://arxiv.org/abs/1912.11747v2 | https://arxiv.org/pdf/1912.11747v2.pdf | Score and Lyrics-Free Singing Voice Generation | Generative models for singing voice have been mostly concerned with the task of ``singing voice synthesis,'' i.e., to produce singing voice waveforms given musical scores and text lyrics. In this work, we explore a novel yet challenging alternative: singing voice generation without pre-assigned scores and lyrics, in both training and inference time. In particular, we outline three such generation schemes, and propose a pipeline to tackle these new tasks. Moreover, we implement such models using generative adversarial networks and evaluate them both objectively and subjectively. | ['Yi-Hsuan Yang', 'Yin-Cheng Yeh', 'Yu-Hua Chen', 'Jen-Yu Liu'] | 2019-12-26 | null | https://openreview.net/forum?id=HygcdeBFvr | https://openreview.net/pdf?id=HygcdeBFvr | null | ['audio-generation', 'singing-voice-synthesis'] | ['audio', 'speech'] | [ 1.25684112e-01 2.76662529e-01 2.99201697e-01 1.78603679e-02
-1.14837289e+00 -1.03397775e+00 3.73343557e-01 -7.48396277e-01
1.17680840e-01 7.35691309e-01 3.13133657e-01 5.71669973e-02
1.19126335e-01 -6.89793468e-01 -4.46562678e-01 -6.40240014e-01
3.49950254e-01 5.30707002e-01 -2.04568624e-01 -2.80031234e-01
-2.02991039e-01 1.29201949e-01 -1.63835883e+00 1.57742277e-01
6.71857536e-01 6.77421451e-01 -3.11690748e-01 1.27863586e+00
3.80445510e-01 8.46543252e-01 -1.22566736e+00 -6.89708233e-01
-9.79580451e-03 -8.89066398e-01 -3.85323226e-01 -1.55987009e-01
5.26276708e-01 -3.08775336e-01 -2.58016557e-01 9.94040430e-01
1.02763951e+00 4.71603513e-01 5.71310580e-01 -1.20845819e+00
-6.19318485e-01 9.17378664e-01 3.42093378e-01 -3.10011566e-01
5.82517684e-01 5.65522850e-01 1.34143770e+00 -9.43830967e-01
5.48912168e-01 9.25677657e-01 7.94511080e-01 1.16437399e+00
-1.46892226e+00 -6.66009843e-01 -6.60881996e-01 -9.79258679e-03
-1.19047749e+00 -8.32297444e-01 1.16403604e+00 -3.96679461e-01
4.76861507e-01 7.95988977e-01 5.37430644e-01 1.38846111e+00
-4.95128840e-01 9.30306494e-01 9.36876059e-01 -3.24480683e-01
3.02799523e-01 -2.30058342e-01 -7.60921419e-01 3.07696134e-01
-4.51751262e-01 4.52384502e-01 -8.69413853e-01 3.25961970e-02
5.56871057e-01 -6.74006999e-01 -1.85452342e-01 2.73914695e-01
-1.21486485e+00 7.05905259e-01 -4.57578264e-02 3.16459894e-01
-4.92234439e-01 4.33601469e-01 2.03980476e-01 2.01713309e-01
2.76215494e-01 7.54751444e-01 1.81237478e-02 -2.13070974e-01
-1.55061162e+00 7.74701774e-01 1.01559627e+00 6.22790456e-01
1.98882595e-01 7.87368774e-01 -5.36464870e-01 9.67820525e-01
1.99483469e-01 6.49232686e-01 5.08408189e-01 -1.25685143e+00
1.07688956e-01 -3.16489518e-01 3.67573947e-02 -5.94512463e-01
-1.81737557e-01 -5.13550878e-01 -5.35230935e-01 3.19240838e-01
3.52729917e-01 -5.15192688e-01 -7.01516032e-01 2.02124977e+00
1.43578321e-01 7.26338446e-01 8.92219469e-02 1.07065713e+00
1.26246202e+00 6.52089536e-01 -1.83190912e-01 -2.95317292e-01
9.12394702e-01 -1.13240492e+00 -1.12750483e+00 1.93970408e-02
-2.77959973e-01 -1.05421913e+00 1.45073175e+00 6.79891348e-01
-1.76824129e+00 -8.61098707e-01 -1.05161023e+00 -1.67240143e-01
1.31104887e-01 2.74808019e-01 3.52672309e-01 8.61261010e-01
-1.30715573e+00 1.06743896e+00 -4.87505257e-01 3.61841887e-01
-2.31384416e-03 2.00337484e-01 9.50080156e-03 6.31684005e-01
-1.44513142e+00 4.61137891e-01 4.50354442e-02 2.95310095e-02
-1.18023503e+00 -6.97044909e-01 -6.97432160e-01 1.68932900e-02
8.89368355e-02 -8.11932087e-01 1.86100376e+00 -5.14584899e-01
-2.19930148e+00 7.64182210e-01 -2.34442670e-02 -3.69961888e-01
5.12397587e-01 -3.66035849e-01 -6.57204449e-01 -4.66811061e-02
-1.70776635e-01 5.20150602e-01 1.22029030e+00 -1.18828619e+00
-7.97448680e-02 8.16865340e-02 -1.25681669e-01 1.78648174e-01
-2.16011956e-01 1.06931761e-01 -1.82283983e-01 -1.14559150e+00
-1.78311914e-01 -8.77468705e-01 -1.04578771e-02 -3.63063842e-01
-7.06224322e-01 -2.40011439e-01 4.21060830e-01 -9.75000679e-01
1.41017568e+00 -2.10753679e+00 4.81001496e-01 1.11818751e-02
1.22919761e-01 4.17847514e-01 -2.64754802e-01 2.87573785e-01
-3.79292928e-02 9.20089614e-03 -2.63186276e-01 -7.07974970e-01
5.60989797e-01 1.76993132e-01 -7.36619055e-01 -1.42573044e-01
4.27355587e-01 1.19874310e+00 -1.02127516e+00 -3.48139048e-01
1.88192949e-01 6.01778448e-01 -9.26471591e-01 7.15103567e-01
-4.21493441e-01 7.81773686e-01 1.26516551e-01 6.93096995e-01
1.29370734e-01 5.18638194e-01 1.40720174e-01 -4.39871922e-02
-5.63000180e-02 7.66775012e-01 -1.10376239e+00 1.77905881e+00
-7.56387055e-01 5.30727863e-01 8.00418332e-02 -4.33361888e-01
1.11284244e+00 9.21263576e-01 2.17286095e-01 -6.44163787e-02
1.32900000e-01 3.49005491e-01 8.69586095e-02 -4.30173665e-01
6.20926440e-01 -7.68979609e-01 -2.24183932e-01 5.98387659e-01
4.09999013e-01 -1.07071483e+00 1.48990110e-01 -4.65346843e-01
9.40968215e-01 4.44740772e-01 1.32815406e-01 4.61675018e-01
4.41792130e-01 -4.11423266e-01 3.86619568e-01 5.13687968e-01
-2.04929579e-02 1.23137367e+00 3.18299174e-01 3.82348001e-02
-9.15502965e-01 -1.42508447e+00 2.10107535e-01 1.13571215e+00
-5.33942997e-01 -4.44119662e-01 -1.27803171e+00 -3.49563390e-01
-2.36798033e-01 1.10000861e+00 -3.43778849e-01 -1.60355330e-01
-7.62413681e-01 -3.69932771e-01 1.12233889e+00 6.03492498e-01
-2.68530309e-01 -1.76700234e+00 -2.40976065e-01 5.40327907e-01
-3.35572898e-01 -7.61644542e-01 -8.25021088e-01 6.55031875e-02
-4.57302004e-01 -7.43171990e-01 -8.82631481e-01 -7.00486660e-01
-4.81478274e-02 -5.38340807e-01 1.38258004e+00 6.40501454e-02
-1.56855717e-01 2.29579695e-02 -2.63530761e-01 -5.96265137e-01
-8.67506564e-01 4.07118946e-02 3.80050451e-01 2.26854160e-02
-8.75599757e-02 -1.00248599e+00 -5.56172013e-01 1.62861064e-01
-9.35139716e-01 -1.37022600e-01 1.49833456e-01 7.50626504e-01
7.01059997e-01 -8.41361582e-02 8.73178005e-01 -6.52501523e-01
1.18033814e+00 -1.74498796e-01 -1.89215362e-01 -1.23448439e-01
-2.10741475e-01 -2.00593635e-01 9.92141724e-01 -5.67904055e-01
-6.24538243e-01 -4.85593677e-02 -1.08932626e+00 -6.60891771e-01
-3.54485720e-01 1.35860220e-01 -3.73517185e-01 3.48019302e-01
8.31804454e-01 1.84970409e-01 -2.16579214e-01 -5.81349671e-01
8.20167065e-01 6.86468422e-01 1.39647281e+00 -6.88977897e-01
1.37361872e+00 1.42003596e-01 -1.73717201e-01 -5.22229910e-01
-8.05788577e-01 -9.09011289e-02 -6.05150878e-01 -3.10309082e-01
8.24645817e-01 -5.69306433e-01 -7.76753902e-01 3.90882581e-01
-1.14267540e+00 -5.56689501e-01 -9.28835630e-01 3.16914886e-01
-1.25396752e+00 1.00409262e-01 -7.10593522e-01 -9.05739605e-01
-6.98737323e-01 -9.25026715e-01 1.10124242e+00 2.63263166e-01
-7.74594307e-01 -9.14637804e-01 5.07681251e-01 3.99717480e-01
4.85595465e-01 5.49583733e-01 4.36418682e-01 -5.20805657e-01
-6.53764084e-02 -3.50538105e-01 4.43520635e-01 7.39434659e-01
9.64249820e-02 1.32033840e-01 -1.44557858e+00 -4.95979041e-02
2.53112335e-02 -6.58363223e-01 4.35071856e-01 1.86515853e-01
1.04304111e+00 -4.57297295e-01 5.45517862e-01 8.13002229e-01
5.98554075e-01 2.04224437e-01 6.61453307e-01 -3.58496606e-01
5.65523684e-01 6.40124500e-01 4.57437545e-01 3.65512222e-01
3.06541193e-02 9.13866460e-01 2.32984871e-01 -9.75735933e-02
-7.72224426e-01 -8.29277933e-01 4.76288050e-01 1.36427808e+00
-4.73941594e-01 -3.16888690e-01 -3.87692153e-01 4.23577845e-01
-1.38804674e+00 -1.30247772e+00 1.89354401e-02 1.93681931e+00
1.25588322e+00 -8.89425054e-02 2.97834635e-01 6.71206057e-01
5.86494565e-01 5.27746439e-01 -4.89206791e-01 -5.61376870e-01
-8.00831169e-02 1.20914829e+00 -4.29141790e-01 5.60493350e-01
-1.01218927e+00 1.00047052e+00 7.03402948e+00 7.99432814e-01
-1.20453274e+00 3.00988138e-01 1.13713168e-01 -3.09847176e-01
-6.45576060e-01 -2.24025011e-01 -3.72078001e-01 4.61945385e-01
1.01473224e+00 -1.93569109e-01 1.16622508e+00 6.84903264e-01
2.49470398e-01 7.01031804e-01 -1.09231055e+00 8.60163569e-01
3.20605487e-01 -1.03446782e+00 -5.03465794e-02 -4.60288614e-01
8.46234977e-01 -4.82719839e-01 3.76565963e-01 4.87002879e-01
1.12435840e-01 -1.42067254e+00 1.16795766e+00 6.38959944e-01
1.24019563e+00 -6.69969976e-01 3.84461999e-01 1.78291127e-01
-1.05559576e+00 4.38012332e-01 1.99011013e-01 7.16247559e-02
5.12796342e-01 1.43549606e-01 -9.13040757e-01 3.27035755e-01
2.11146519e-01 2.27748156e-01 -1.13625534e-01 8.99639666e-01
-9.93458211e-01 1.13317311e+00 -1.09611921e-01 1.49297521e-01
-1.77223757e-01 5.11056595e-02 8.58938575e-01 1.01915789e+00
6.70496047e-01 -9.55122262e-02 -1.58969045e-01 1.37525547e+00
-2.33348578e-01 1.03200719e-01 -3.52904797e-01 -2.28042737e-01
4.68444556e-01 1.28944707e+00 -1.45468935e-01 -1.21916652e-01
3.76247577e-02 9.79118884e-01 -2.85317823e-02 2.61954546e-01
-8.68161738e-01 -4.83176291e-01 6.60650074e-01 2.13790953e-01
5.33025675e-02 -9.75830033e-02 -3.93875599e-01 -9.70470369e-01
-1.78370491e-01 -9.70323324e-01 -9.53878392e-04 -1.02648401e+00
-1.11680019e+00 7.64546692e-01 -5.66946983e-01 -1.28591025e+00
-8.94334853e-01 -4.81607527e-01 -1.26460004e+00 9.62614894e-01
-1.38734794e+00 -1.03297341e+00 -1.50708124e-01 3.93190771e-01
5.92378259e-01 -2.96196401e-01 1.05383945e+00 4.93079275e-01
-3.17173064e-01 8.10657680e-01 -2.13363886e-01 -2.75628641e-02
8.51290822e-01 -1.77422965e+00 8.82287681e-01 6.32605612e-01
7.47994483e-01 4.63828593e-01 1.02309787e+00 -2.13795230e-01
-1.08065021e+00 -1.00625014e+00 1.04754055e+00 -7.31202066e-01
6.38507247e-01 -2.56858230e-01 -7.63938129e-01 2.27278560e-01
1.87622175e-01 -1.08126149e-01 9.52512681e-01 -3.12467329e-02
-6.74003363e-02 1.13560267e-01 -1.02503026e+00 6.05012834e-01
1.03327453e+00 -7.60915875e-01 -7.72897124e-01 1.23616897e-01
7.56654263e-01 -7.16226757e-01 -9.50050116e-01 5.55134654e-01
5.83172441e-01 -9.10049856e-01 1.15305209e+00 -7.08972335e-01
6.04715109e-01 -5.36658525e-01 -3.31663564e-02 -1.67244709e+00
-1.93495914e-01 -1.32355583e+00 -4.29882556e-01 1.35531187e+00
3.63977402e-01 5.94759956e-02 6.87189400e-01 -6.73931241e-02
-4.94331568e-01 -3.79502416e-01 -8.53857219e-01 -6.82249546e-01
1.52705655e-01 -6.87756181e-01 9.63613093e-01 7.21287489e-01
-3.06030661e-01 3.71501327e-01 -8.82856727e-01 -2.09690053e-02
4.79852766e-01 2.66710430e-01 9.97853398e-01 -1.01419735e+00
-8.29665124e-01 -5.40438175e-01 4.28369455e-02 -5.66625118e-01
3.54866445e-01 -1.02794242e+00 5.69286406e-01 -1.19138610e+00
-3.42514575e-01 -7.60727227e-02 -4.00451988e-01 3.81888121e-01
-4.81046885e-01 9.52429175e-01 4.14811999e-01 5.08195199e-02
-3.03710938e-01 6.15971863e-01 1.45802736e+00 -1.95107739e-02
-3.30989629e-01 4.07460183e-01 -4.01081771e-01 7.74863243e-01
1.03001392e+00 -3.60237896e-01 -2.51870155e-01 -1.34501219e-01
2.60508478e-01 3.94108266e-01 6.50284588e-01 -1.11093092e+00
2.69483514e-02 -5.27211055e-02 5.37839569e-02 -5.17536819e-01
6.24320865e-01 -8.37013125e-02 3.68111640e-01 1.77840382e-01
-4.54729021e-01 -3.52874219e-01 -2.05001589e-02 8.66929721e-03
-4.94114786e-01 -5.24499238e-01 7.44214594e-01 9.78292674e-02
-1.23701915e-01 3.69820856e-02 -1.33917868e-01 3.34335566e-01
5.04210114e-01 1.96170121e-01 2.75213104e-02 -7.22082555e-01
-1.31531668e+00 -2.90338099e-01 1.65662140e-01 5.23654938e-01
6.12764537e-01 -1.71958649e+00 -9.02927339e-01 2.20901564e-01
-9.13716033e-02 -3.89230549e-02 2.64128149e-01 5.25916874e-01
-4.03439283e-01 1.10899262e-01 -5.74513376e-02 -4.85135950e-02
-1.13347363e+00 3.25353771e-01 4.01019365e-01 -1.30871549e-01
-3.11497927e-01 8.06206584e-01 -8.73796046e-02 -8.25432420e-01
2.29376569e-01 -4.91370633e-02 -2.97738791e-01 7.71583021e-02
4.26262945e-01 4.94385600e-01 8.04311316e-03 -6.16094708e-01
9.02294442e-02 3.06545764e-01 6.95714653e-01 -4.42789167e-01
1.06828225e+00 3.58649045e-01 7.20491633e-02 5.07476032e-01
6.18391871e-01 6.68218195e-01 -1.17094028e+00 1.64475128e-01
-3.78532968e-02 -2.68173575e-01 -7.70280138e-02 -9.58123684e-01
-9.25167263e-01 9.57748294e-01 2.42246434e-01 2.83621162e-01
1.03017092e+00 3.54145430e-02 1.30231333e+00 6.18559308e-02
2.61874571e-02 -1.04011369e+00 2.04151317e-01 4.40958828e-01
1.28544068e+00 -5.73457658e-01 -5.63383460e-01 -1.09770842e-01
-6.05195642e-01 9.19908226e-01 3.87619674e-01 -1.77381948e-01
2.47773513e-01 3.70137274e-01 4.98690724e-01 1.62684202e-01
-7.13039994e-01 -4.34630752e-01 7.85713077e-01 8.17935050e-01
7.42797732e-01 4.25257385e-01 -2.97318161e-01 1.15087497e+00
-1.22310627e+00 6.80526942e-02 1.97866887e-01 6.96825609e-02
-1.05568156e-01 -1.30559349e+00 -5.75456440e-01 4.44761757e-03
-6.20290637e-01 -3.35824579e-01 -7.61098325e-01 2.10066050e-01
5.59349060e-01 1.10981953e+00 -1.67361036e-01 -7.56901622e-01
6.08501375e-01 3.55086893e-01 4.16143686e-01 -8.22463155e-01
-8.76073182e-01 2.61410445e-01 2.23122045e-01 -1.75060242e-01
-1.45749345e-01 -6.91991389e-01 -1.06424069e+00 -7.73853883e-02
-1.81574881e-01 2.48267934e-01 5.92076778e-01 5.82006693e-01
-1.65984273e-01 1.02109098e+00 9.77132618e-01 -9.67575848e-01
-9.01871979e-01 -1.03810287e+00 -5.97118139e-01 5.48668623e-01
1.30071625e-01 -2.53927171e-01 -4.22934592e-01 1.31793454e-01] | [15.58244800567627, 6.0288801193237305] |
8567c628-aabc-45c5-ac02-84fcf9411375 | dependency-parsing-in-a-morphological-rich | null | null | https://aclanthology.org/2021.pail-1.3 | https://aclanthology.org/2021.pail-1.3.pdf | Dependency Parsing in a Morphological rich language, Tamil | Dependency parsing is the process of analysing the grammatical structure of a sentence based on the dependencies between the words in a sentence. The annotation of dependency parsing is done using different formalisms at word-level namely Universal Dependencies and chunk-level namely AnnaCorra. Though dependency parsing is deeply dealt in languages such as English, Czech etc the same cannot be adopted for the morphologically rich and agglutinative languages. In this paper, we discuss the development of a dependency parser for Tamil, a South Dravidian language. The different characteristics of the language make this task a challenging task. Tamil, a morphologically rich and agglutinative language, has copula drop, accusative and genitive case drop and pro-drop. Coordinative constructions are introduced by affixation of morpheme ‘um’. Embedded clausal structures are common in relative participle and complementizer clauses. In this paper, we have discussed our approach to handle some of these challenges. We have used Malt parser, a supervised learning- approach based implementation. We have obtained an accuracy of 79.27% for Unlabelled Attachment Score, 73.64% for Labelled Attachment Score and 68.82% for Labelled Accuracy. | ['Sobha Lalitha Devi', 'Vijay Sundar Ram'] | null | null | null | null | pail-icon-2021-12 | ['dependency-parsing'] | ['natural-language-processing'] | [-5.48424959e-01 2.08942562e-01 2.73964614e-01 -5.70476830e-01
-3.75654429e-01 -8.28776181e-01 2.13227138e-01 7.02869356e-01
-6.66075408e-01 1.09233713e+00 4.51376379e-01 -5.75134516e-01
9.16968882e-02 -6.82067215e-01 4.39605862e-02 -4.69265878e-01
-1.09195188e-01 7.91029930e-01 4.46723819e-01 -5.98839879e-01
3.14671993e-01 4.61875856e-01 -1.05660903e+00 4.16531354e-01
6.47109985e-01 -6.69181347e-02 4.15739834e-01 6.62483156e-01
-2.49525994e-01 7.32470691e-01 -7.72476792e-01 -5.80464125e-01
-1.86097726e-01 -3.41173649e-01 -1.28255761e+00 -2.56246448e-01
-3.39313269e-01 3.28856349e-01 7.61199951e-01 8.36737990e-01
2.48816997e-01 -2.27840677e-01 7.80463755e-01 -6.43672347e-01
-2.83224612e-01 8.69535863e-01 -5.63502252e-01 5.57728410e-01
4.80673075e-01 -6.16509438e-01 1.18870986e+00 -8.46282482e-01
7.44320095e-01 1.50848460e+00 3.00272673e-01 4.85431522e-01
-7.59655476e-01 -4.12380815e-01 -2.30222754e-02 -1.18070886e-01
-8.89081240e-01 -1.03681020e-01 3.41888636e-01 -4.56327796e-01
1.57744777e+00 3.36384680e-03 2.80788034e-01 2.32803226e-01
5.10270774e-01 3.90220523e-01 1.28849483e+00 -9.66714919e-01
9.94650051e-02 1.81069627e-01 6.83427155e-01 3.72885168e-01
6.13972664e-01 -4.61586207e-01 3.06080319e-02 1.02821521e-01
2.05643952e-01 -5.97376406e-01 3.01652670e-01 4.79781806e-01
-6.68082356e-01 1.11302912e+00 7.08257034e-02 1.00079715e+00
-2.11801991e-01 -3.92127573e-01 8.04477930e-01 1.35716930e-01
2.13013679e-01 -8.28653276e-02 -7.94394076e-01 -1.46743938e-01
-1.55545905e-01 2.15038061e-01 1.10317600e+00 8.66641819e-01
3.58446926e-01 -3.20816226e-02 6.93910837e-01 8.76824379e-01
6.69047713e-01 3.15162212e-01 3.82883251e-01 -3.89807671e-01
6.53588057e-01 7.25347877e-01 -4.51172851e-02 -4.49923903e-01
-5.79040527e-01 1.26548812e-01 -2.27021843e-01 3.45026284e-01
5.44300675e-01 -5.34863770e-01 -7.57552683e-01 1.67691398e+00
5.43398559e-01 -1.01465428e+00 7.92376697e-01 5.75409412e-01
9.03289557e-01 7.10149229e-01 8.79010797e-01 -4.17193770e-01
1.92103660e+00 -5.43636501e-01 -9.10178721e-01 9.23157483e-03
9.70069706e-01 -1.24614370e+00 8.22098434e-01 1.51339069e-01
-1.41279948e+00 -1.61308199e-01 -9.47155535e-01 -1.21591017e-01
-6.25124991e-01 -1.34705678e-01 4.91247684e-01 8.94628584e-01
-7.96721935e-01 1.80393636e-01 -9.35149670e-01 -5.82769573e-01
-1.83676302e-01 6.58815980e-01 -8.06130230e-01 1.54680327e-01
-9.70120609e-01 1.34096432e+00 9.68112051e-01 -7.42033264e-03
1.54936701e-01 1.04148217e-01 -9.69724000e-01 -2.37134486e-01
-1.31891102e-01 1.03544533e-01 1.15004194e+00 -1.10639584e+00
-1.08586287e+00 1.45821369e+00 8.23418126e-02 -2.72407174e-01
-2.26770461e-01 -1.62349463e-01 -5.43873847e-01 4.77315336e-02
3.43334079e-01 2.24811122e-01 9.53907296e-02 -9.14729238e-01
-1.06212497e+00 -7.84968555e-01 2.24923581e-01 2.63693929e-01
3.90291899e-01 1.15572715e+00 2.37558469e-01 -4.21028465e-01
2.42733002e-01 -8.32650781e-01 1.53633714e-01 -1.26180446e+00
-2.24688370e-03 -7.26759374e-01 8.36834729e-01 -1.09064281e+00
1.32789195e+00 -1.92567801e+00 8.04667026e-02 -1.39683962e-01
-2.18204975e-01 5.23771763e-01 3.95810127e-01 9.41112816e-01
-3.47783267e-01 1.82056263e-01 -4.56548274e-01 -2.86350120e-02
-4.56483252e-02 1.01561964e+00 3.21303636e-01 4.17452842e-01
5.35595000e-01 4.74025935e-01 -6.25218689e-01 -8.60812306e-01
-1.17777260e-02 4.39507991e-01 -3.84483457e-01 1.68676451e-01
1.00524001e-01 2.58381546e-01 -4.17111546e-01 8.71787429e-01
8.77735734e-01 6.91752553e-01 8.25977027e-01 1.05883554e-01
-6.94880784e-01 6.40393853e-01 -1.17781961e+00 1.20400190e+00
-4.59823161e-01 1.16293199e-01 3.70587319e-01 -1.08224440e+00
9.13522243e-01 6.13396645e-01 -1.97176322e-01 -3.88721943e-01
3.96525711e-01 7.55437315e-01 6.87335074e-01 -6.18245006e-01
3.91713589e-01 -6.45800889e-01 -5.27730823e-01 2.01797277e-01
2.55546719e-01 8.76215473e-02 6.93288624e-01 1.73848957e-01
7.55965948e-01 3.57493758e-01 1.11606443e+00 -7.93287516e-01
1.03061879e+00 4.23131436e-02 8.85022223e-01 -3.47306043e-01
-2.34605014e-01 2.74512261e-01 7.96056807e-01 -3.63612473e-01
-8.75878692e-01 -9.86495554e-01 -6.86736047e-01 1.15681028e+00
-5.73300719e-01 -3.63146871e-01 -9.07947481e-01 -8.58781397e-01
-5.36228299e-01 7.03738153e-01 -4.10332918e-01 6.02430880e-01
-1.38341677e+00 -1.00381601e+00 2.81865656e-01 5.20395815e-01
1.77604079e-01 -1.83784640e+00 -8.84062231e-01 6.71498716e-01
1.27179027e-01 -1.18236482e+00 2.88040668e-01 6.97149873e-01
-8.22964549e-01 -1.06934178e+00 -1.11780325e-02 -1.33431745e+00
4.89551246e-01 -6.09106481e-01 1.46006691e+00 5.03313124e-01
-1.28774300e-01 1.67324278e-03 -9.62398410e-01 -9.03088748e-01
-6.84265912e-01 1.14329038e-02 -2.11452991e-01 -6.51691854e-01
7.20598578e-01 -5.74925423e-01 8.88244249e-03 -3.77819926e-01
-8.69464338e-01 -5.21020830e-01 3.65419745e-01 5.08020699e-01
1.25110194e-01 -3.51064622e-01 5.49558580e-01 -1.53345633e+00
4.09633487e-01 -4.93010223e-01 -4.00799096e-01 -1.93500556e-02
-1.05807155e-01 1.71528056e-01 5.02570331e-01 2.19066635e-01
-1.20439112e+00 7.42392465e-02 -8.85021269e-01 9.17691708e-01
-3.85708153e-01 5.81389725e-01 -6.71902120e-01 2.92093635e-01
3.05889904e-01 -5.12479663e-01 -2.36357391e-01 -6.53348565e-01
-1.39307454e-01 7.09255695e-01 4.28824186e-01 -7.66958773e-01
3.99751276e-01 -9.59371924e-02 1.64656445e-01 -9.54711258e-01
-6.03909433e-01 -7.33109295e-01 -1.26259601e+00 1.05299719e-01
1.44051743e+00 -7.46724606e-01 -5.07952154e-01 2.93857932e-01
-1.43878508e+00 5.35941496e-02 -6.11142069e-02 3.98577631e-01
-1.35389894e-01 3.66883636e-01 -8.90293598e-01 -1.01862574e+00
-6.20423555e-01 -7.10447192e-01 6.11534297e-01 2.33837023e-01
-3.95802140e-01 -1.35625923e+00 4.21628445e-01 7.88965635e-03
4.10006531e-02 6.40680611e-01 1.33068061e+00 -1.01425028e+00
3.23401183e-01 -7.15824291e-02 -5.63367940e-02 4.36350912e-01
1.89968929e-01 1.87758237e-01 -5.62633276e-01 -5.54146990e-02
1.56005487e-01 -2.50158459e-01 1.65192306e-01 2.92902961e-02
-3.94001395e-01 -6.50885999e-02 1.08180068e-01 -2.35776920e-02
1.87475562e+00 5.19765615e-01 7.17630863e-01 5.53869426e-01
2.94812471e-01 1.06215572e+00 9.92403328e-01 1.66746810e-01
5.92000186e-01 2.05145970e-01 3.78374517e-01 3.50641310e-01
-2.59855064e-04 3.34212184e-01 5.49032807e-01 1.25116193e+00
-3.33110392e-01 -2.48223066e-01 -1.25886023e+00 7.77139604e-01
-1.53891349e+00 -4.79729205e-01 -1.19668436e+00 1.85254645e+00
9.50984120e-01 3.94969076e-01 2.03537032e-01 5.49957633e-01
7.51166224e-01 -2.99859345e-01 5.56902647e-01 -1.28884363e+00
-1.90265939e-01 9.83322740e-01 1.42056435e-01 1.06775653e+00
-1.04409349e+00 1.27798593e+00 4.81115580e+00 1.25968307e-01
-8.06550980e-01 4.76390481e-01 -3.89522240e-02 5.81225634e-01
1.43064946e-01 2.89863288e-01 -1.10110116e+00 3.34777594e-01
1.41092730e+00 3.90569985e-01 -3.62230033e-01 4.23442990e-01
6.99629486e-02 -6.67418063e-01 -7.60745227e-01 4.19697165e-01
-2.52369605e-02 -7.73461699e-01 -1.47531122e-01 -1.13234155e-01
1.28400534e-01 -9.33701359e-03 -6.00016415e-01 3.33144575e-01
1.25439629e-01 -9.06030476e-01 6.92065179e-01 -2.56005883e-01
2.63358057e-01 -1.09542060e+00 1.19398367e+00 2.64443845e-01
-1.32327008e+00 1.24661572e-01 -5.55087864e-01 -5.27776122e-01
3.91024888e-01 2.22473145e-01 -4.27035570e-01 5.77122211e-01
7.16648698e-01 1.99673712e-01 -4.42386538e-01 5.09741545e-01
-6.88772082e-01 6.91063046e-01 -4.56252813e-01 -1.55521080e-01
5.34994125e-01 -5.15762150e-01 3.78048837e-01 1.85635710e+00
-8.64721462e-02 4.95460510e-01 -1.11612774e-01 -6.34872988e-02
5.02312601e-01 7.82325864e-01 -3.93737048e-01 4.18809444e-01
2.61127025e-01 1.38814974e+00 -1.20506072e+00 -1.65128440e-01
-5.02726793e-01 6.97126925e-01 5.56038320e-01 -2.28193209e-01
-3.68167222e-01 -4.24320102e-01 2.44150937e-01 1.42032996e-01
3.94078225e-01 -3.11742902e-01 -2.04510428e-03 -8.19070637e-01
1.26906902e-01 -6.33897185e-01 8.70311797e-01 -2.75003105e-01
-1.01412761e+00 9.54445601e-01 2.93601930e-01 -5.37337601e-01
-5.28546751e-01 -1.03325140e+00 -8.08386087e-01 1.21298146e+00
-1.33570206e+00 -1.43724895e+00 2.99052298e-01 4.56981212e-01
5.01159668e-01 -1.36714429e-01 1.34034300e+00 1.74421981e-01
-6.29210114e-01 2.08075307e-02 -2.90295124e-01 4.45567578e-01
4.26828653e-01 -1.68462658e+00 1.01437323e-01 9.43768382e-01
-2.10452974e-01 5.40124416e-01 8.80617499e-01 -5.09173632e-01
-8.28648150e-01 -5.55438042e-01 1.84985328e+00 -5.17542541e-01
7.36653745e-01 -2.97965854e-01 -9.65475142e-01 7.91795969e-01
8.79487097e-01 5.77482469e-02 1.04284251e+00 5.28307185e-02
-1.86743230e-01 2.06699058e-01 -1.36468720e+00 1.00542285e-01
4.70706612e-01 -8.82181898e-02 -1.20085073e+00 2.48398200e-01
4.50177342e-01 -2.50708640e-01 -9.97563124e-01 3.84528339e-02
2.08461508e-01 -1.23092604e+00 5.03185451e-01 -6.43247724e-01
1.27702206e-01 -1.64747819e-01 -2.97191262e-01 -8.18231225e-01
-2.71096915e-01 -3.09042186e-01 3.93659860e-01 1.74012291e+00
7.29860723e-01 -6.72373593e-01 4.44268137e-01 5.44992030e-01
-3.73097628e-01 -9.34604332e-02 -9.55024481e-01 -3.23167294e-01
7.80335844e-01 -1.24831870e-01 1.39564887e-01 8.66736412e-01
3.42430562e-01 1.11828637e+00 4.11351174e-01 1.38304299e-02
5.00887454e-01 -1.48743615e-01 5.17304987e-02 -1.66238630e+00
-1.19429350e-01 7.51157850e-02 -4.05816525e-01 1.47915870e-01
3.60360116e-01 -7.19429970e-01 -2.45128438e-01 -1.55514371e+00
-2.09750429e-01 -6.31161153e-01 1.81191027e-01 3.67243230e-01
-9.72318575e-02 2.22125128e-01 2.75048822e-01 7.59969354e-02
-1.49246722e-01 -3.05359930e-01 6.86387241e-01 5.50344706e-01
-2.18188539e-01 -2.25368917e-01 -4.48437929e-01 1.05739534e+00
1.18297029e+00 -8.78027439e-01 -1.25571787e-01 -3.53111506e-01
4.86705691e-01 1.40595838e-01 -4.05408531e-01 -6.84544384e-01
-2.06897795e-01 -3.68037522e-02 -1.26656055e-01 -6.42578423e-01
-9.35752541e-02 -9.02040660e-01 -2.86294185e-02 8.04284692e-01
3.44804674e-01 9.05857623e-01 3.46616954e-01 -1.99529901e-01
-2.19029173e-01 -1.19087207e+00 1.01649678e+00 -5.06409824e-01
-6.09316349e-01 -3.31745237e-01 -7.72934973e-01 4.39490020e-01
1.25414360e+00 -1.05643012e-01 6.01939522e-02 4.70118165e-01
-9.27211225e-01 -1.26706362e-01 2.23104596e-01 -1.68358069e-02
2.70123512e-01 -7.28241444e-01 -9.90339518e-01 -4.28868532e-02
-5.74528836e-02 2.65309997e-02 -1.66108802e-01 6.12975597e-01
-1.23306954e+00 4.07857835e-01 -6.76059067e-01 3.23161595e-02
-1.76455891e+00 5.16658545e-01 4.09572162e-02 -6.70732379e-01
-3.64317209e-01 5.74172318e-01 -7.40578398e-02 -5.19166768e-01
-2.94384569e-01 -3.36952507e-01 -1.10653460e+00 3.12324405e-01
1.87739626e-01 8.93553719e-02 3.21588010e-01 -1.40605116e+00
-7.20406651e-01 5.24685204e-01 -1.44496590e-01 -2.32393786e-01
1.38110685e+00 -5.78424819e-02 -8.94086361e-01 5.08431673e-01
1.14453650e+00 6.25335097e-01 -2.85886019e-01 5.71885169e-01
5.88998973e-01 2.03057423e-01 -4.01204884e-01 -8.29564154e-01
-3.86147946e-01 8.47101986e-01 1.33605391e-01 4.61379379e-01
9.15870428e-01 1.51229903e-01 4.52590466e-01 5.05670458e-02
2.25989163e-01 -1.11953926e+00 -6.52945042e-01 1.08367085e+00
8.08774292e-01 -9.44252193e-01 -6.43985569e-02 -9.06412542e-01
-6.07115209e-01 1.21875179e+00 5.36795259e-01 -5.09193838e-01
7.25348830e-01 7.85051286e-01 2.70433605e-01 -3.22790474e-01
-5.99959552e-01 -4.88283187e-01 -3.43989640e-01 9.63506520e-01
1.37994790e+00 3.45224679e-01 -1.55999088e+00 5.96708834e-01
-5.09552717e-01 -6.45428240e-01 8.33231926e-01 1.46660697e+00
-5.55433512e-01 -1.83860815e+00 -6.29837394e-01 -7.74614140e-02
-1.38838220e+00 -3.15959454e-01 -3.33125383e-01 1.56966770e+00
3.80517781e-01 1.08123529e+00 2.74965465e-01 1.66810095e-01
4.06992167e-01 2.07346067e-01 7.28463948e-01 -1.19492042e+00
-1.36060166e+00 2.12859914e-01 7.48116851e-01 1.24104261e-01
-8.92399132e-01 -9.75079000e-01 -2.12417269e+00 -2.20795617e-01
-1.19907603e-01 7.83429921e-01 7.00740516e-01 1.07175195e+00
-5.59794486e-01 1.65353134e-01 1.50172740e-01 -4.99611020e-01
-8.21367130e-02 -1.09548223e+00 -8.55442286e-01 2.66879916e-01
-7.78133273e-02 -5.23295343e-01 -1.74169451e-01 1.72100797e-01] | [10.344305992126465, 10.068018913269043] |
46790817-e70d-4d7b-8d8f-7705664837f3 | end-to-end-annotator-bias-approximation-on | 2111.02326 | null | https://arxiv.org/abs/2111.02326v1 | https://arxiv.org/pdf/2111.02326v1.pdf | End-to-End Annotator Bias Approximation on Crowdsourced Single-Label Sentiment Analysis | Sentiment analysis is often a crowdsourcing task prone to subjective labels given by many annotators. It is not yet fully understood how the annotation bias of each annotator can be modeled correctly with state-of-the-art methods. However, resolving annotator bias precisely and reliably is the key to understand annotators' labeling behavior and to successfully resolve corresponding individual misconceptions and wrongdoings regarding the annotation task. Our contribution is an explanation and improvement for precise neural end-to-end bias modeling and ground truth estimation, which reduces an undesired mismatch in that regard of the existing state-of-the-art. Classification experiments show that it has potential to improve accuracy in cases where each sample is annotated only by one single annotator. We provide the whole source code publicly and release an own domain-specific sentiment dataset containing 10,000 sentences discussing organic food products. These are crawled from social media and are singly labeled by 10 non-expert annotators. | ['Georg Groh', 'Hannah Danner', 'Maximilian Wich', 'Christian Widmer', 'Maria Luisa Ripoll Dominguez', 'Andreas Koch', 'David Szabo', 'Gerhard Hagerer'] | 2021-11-03 | null | https://aclanthology.org/2021.icnlsp-1.1 | https://aclanthology.org/2021.icnlsp-1.1.pdf | icnlsp-2021-11 | ['misconceptions'] | ['miscellaneous'] | [ 1.95470124e-01 7.00132430e-01 -2.46390328e-01 -9.50146794e-01
-8.23128462e-01 -1.07543468e+00 1.15487099e-01 7.12945700e-01
-4.76424128e-01 8.58478785e-01 2.89084613e-01 -5.88384718e-02
7.14679122e-01 -2.67866999e-01 -9.94515300e-01 -5.76940536e-01
6.50592446e-01 7.62641072e-01 8.06967244e-02 -3.99153054e-01
2.16077238e-01 -2.56475270e-01 -1.00841594e+00 7.48780668e-01
8.88150394e-01 8.90796483e-01 -2.02660099e-01 5.00733435e-01
-4.94149253e-02 1.22004974e+00 -7.09990144e-01 -1.29946697e+00
4.82413322e-02 -2.16388971e-01 -9.38801110e-01 7.14936927e-02
7.23489046e-01 -1.25847757e-01 4.78631735e-01 1.56871247e+00
4.80735391e-01 -5.43582976e-01 5.10732710e-01 -1.18300664e+00
-1.20043004e+00 9.04328287e-01 -5.35786152e-01 -1.63164183e-01
2.96096116e-01 1.34761944e-01 1.04169512e+00 -6.56947076e-01
6.45154357e-01 1.06323206e+00 1.19748569e+00 8.26420009e-01
-1.24457610e+00 -5.33095002e-01 2.78131932e-01 -1.99519917e-01
-1.00527561e+00 -3.30209553e-01 5.77027678e-01 -1.00991547e+00
4.18674976e-01 -1.69673737e-03 3.27010304e-01 1.46008086e+00
-3.75668965e-02 8.38318408e-01 1.28149724e+00 -3.77701581e-01
5.62645674e-01 7.47345865e-01 5.45321643e-01 4.05274898e-01
6.04354620e-01 -3.04947019e-01 -7.71749973e-01 -4.49199438e-01
-3.32367457e-02 -3.50830078e-01 -2.00810090e-01 -4.03948069e-01
-9.29616451e-01 1.08090472e+00 4.81764436e-01 1.19496576e-01
-3.91614318e-01 1.07421547e-01 8.06007743e-01 2.36792773e-01
1.12442863e+00 7.10738063e-01 -1.03625870e+00 1.59777209e-01
-8.48506153e-01 4.54772234e-01 1.16927457e+00 1.09654772e+00
6.13441110e-01 -1.10670507e-01 -2.44090915e-01 7.09927738e-01
4.31895554e-01 6.46265388e-01 4.57959145e-01 -8.08978319e-01
2.31950358e-01 7.44321585e-01 8.31586778e-01 -1.20667994e+00
-4.49549228e-01 -1.71030730e-01 -5.00337660e-01 9.04877633e-02
8.89678359e-01 -4.97002989e-01 -4.45972860e-01 1.55439579e+00
3.72636676e-01 -6.52545214e-01 -8.11198354e-03 1.19804442e+00
9.04673934e-01 -3.37828957e-02 5.46197712e-01 1.51561081e-01
1.84476995e+00 -1.02202392e+00 -9.68678534e-01 -6.75752640e-01
8.77753675e-01 -9.48619843e-01 1.04895949e+00 4.98191744e-01
-8.21538091e-01 -3.17887336e-01 -9.42888618e-01 -2.05295339e-01
-5.60151577e-01 3.71946990e-01 5.19980788e-01 1.04911256e+00
-4.79391098e-01 5.33995569e-01 -6.09612644e-01 -3.02017719e-01
8.46774220e-01 2.05652088e-01 -5.10083973e-01 9.61521119e-02
-1.32933629e+00 1.15020669e+00 4.80233207e-02 2.82843173e-01
-8.76714647e-01 -7.12643445e-01 -8.15505147e-01 -3.98040861e-01
4.29349035e-01 -1.28842503e-01 1.89100254e+00 -1.75110173e+00
-1.25728977e+00 1.54901385e+00 -3.62497717e-01 -5.29059529e-01
7.74249136e-01 -4.68105078e-01 -3.22894901e-01 -3.34600210e-01
6.25558257e-01 5.12689412e-01 6.09504521e-01 -1.44936526e+00
-4.46771860e-01 -6.17591381e-01 -4.82038781e-02 -1.59915358e-01
-2.79125243e-01 2.04100728e-01 4.35187906e-01 -5.38591683e-01
-2.41246700e-01 -1.10062122e+00 -1.98503137e-01 1.70015752e-01
-4.99143451e-01 -2.22253799e-01 1.28277421e-01 -7.36726999e-01
8.16968918e-01 -1.92048585e+00 -3.72733146e-01 -4.51184362e-01
1.92005888e-01 5.77597380e-01 1.16751730e-01 3.10878664e-01
4.45217127e-03 3.91573668e-01 -1.41513839e-01 -4.26116824e-01
1.66208804e-01 1.11399859e-01 -3.01784426e-01 7.61905372e-01
2.93501019e-01 7.77896702e-01 -1.35886014e+00 -2.21840769e-01
-2.57059455e-01 1.96943939e-01 -2.76004761e-01 3.44569176e-01
-3.71193320e-01 4.69795614e-01 -3.41940761e-01 5.65201819e-01
6.88173056e-01 -2.59078205e-01 1.60274625e-01 -3.98758322e-01
5.29238470e-02 3.67283106e-01 -8.41945350e-01 1.60198426e+00
-2.55678897e-03 6.67485893e-01 3.35997075e-01 -6.97234929e-01
1.03618693e+00 4.50549722e-01 -1.99597374e-01 -1.72962248e-01
4.02709663e-01 6.41211867e-01 -2.71455228e-01 -8.20659339e-01
6.32208943e-01 -3.35904956e-01 -3.81556690e-01 3.94470602e-01
1.81070626e-01 -1.81761518e-01 1.12348422e-01 -7.17833415e-02
5.61936736e-01 6.50161430e-02 4.74564970e-01 -7.43023694e-01
3.80112380e-01 6.41482592e-01 6.59430444e-01 8.89892876e-01
-7.36938655e-01 6.61515653e-01 8.00305843e-01 -1.15972745e+00
-1.24842656e+00 -2.50756294e-01 -1.32875349e-02 1.51490450e+00
-1.35271594e-01 1.08478546e-01 -1.11860812e+00 -1.14234400e+00
6.88995719e-02 7.94151604e-01 -1.16325557e+00 7.72548392e-02
-9.64794382e-02 -9.18559253e-01 7.60414302e-01 5.07793903e-01
2.43610188e-01 -8.05249155e-01 -4.57789272e-01 1.84608117e-01
-3.16433668e-01 -1.33831954e+00 -4.50122684e-01 5.04661918e-01
-3.24803472e-01 -1.26513815e+00 -6.61014378e-01 -6.88812077e-01
8.36301744e-01 -1.11038275e-02 1.61996520e+00 -9.81314257e-02
2.17492983e-01 -2.71617800e-01 -4.96048093e-01 -1.16108429e+00
-9.54705179e-01 2.92984515e-01 -1.24853440e-01 -7.69345984e-02
1.26020014e+00 1.23294450e-01 -5.42217851e-01 5.79732656e-01
-6.67561591e-01 -2.49937311e-01 4.85449359e-02 6.39587879e-01
2.09727153e-01 -6.62383258e-01 7.74914920e-01 -1.56486821e+00
5.71143270e-01 -5.73929071e-01 -6.11804962e-01 9.60972607e-02
-5.86120546e-01 -8.57748240e-02 4.94802833e-01 -3.51227611e-01
-9.30885911e-01 3.57842326e-01 -7.60622323e-02 3.55316430e-01
-5.87441206e-01 1.87224239e-01 1.63916662e-01 1.53303668e-01
1.26394463e+00 -5.41744709e-01 5.60650863e-02 -3.44924688e-01
3.24831098e-01 9.81190562e-01 2.64196098e-01 -4.29375887e-01
1.96557656e-01 5.36562502e-01 -5.75680554e-01 -1.81738496e-01
-1.85267639e+00 -4.27724600e-01 -7.63053894e-01 -1.14940375e-01
1.12317610e+00 -1.41487992e+00 -8.06690216e-01 6.43872857e-01
-1.56304204e+00 -3.13066125e-01 -1.76168025e-01 -3.98604292e-03
-7.91437402e-02 6.57216683e-02 -6.78858876e-01 -8.09613347e-01
-4.01937217e-01 -1.13485360e+00 1.15220678e+00 1.30298480e-01
-9.17937338e-01 -8.81808341e-01 1.83426023e-01 8.28402400e-01
3.70238990e-01 2.69934684e-01 2.94403017e-01 -1.37534392e+00
3.96131605e-01 -7.07167029e-01 -2.12291971e-01 6.93479776e-01
-1.67626128e-01 -9.01879296e-02 -1.43282306e+00 1.52121112e-02
2.28879318e-01 -9.21577215e-01 4.71927226e-01 1.09213427e-01
5.59670925e-01 -3.60739321e-01 1.69402305e-02 -3.43407065e-01
1.28553438e+00 -3.70692074e-01 2.92990476e-01 3.44249338e-01
7.21727908e-01 1.00792587e+00 5.74902773e-01 1.78733706e-01
6.32977366e-01 3.82777333e-01 6.60641611e-01 -6.50851950e-02
3.24625701e-01 -1.81017831e-01 3.19466680e-01 3.66836190e-01
2.19002590e-01 -3.58356208e-01 -8.90098393e-01 7.09372640e-01
-2.12042594e+00 -6.78586066e-01 -7.68814802e-01 1.87214351e+00
1.19512081e+00 1.42552387e-02 7.00505227e-02 2.55154185e-02
1.06232500e+00 1.06095932e-01 -4.34361041e-01 -6.04959130e-01
-1.99594915e-01 -5.92045963e-01 8.51020634e-01 5.14257669e-01
-1.34133744e+00 1.05351055e+00 6.14757824e+00 3.64255220e-01
-6.10153794e-01 5.47913730e-01 9.51742947e-01 1.03454411e-01
-1.21963426e-01 -2.20802039e-01 -1.02335155e+00 5.31234503e-01
7.06314862e-01 4.32277709e-01 2.10144579e-01 1.33632696e+00
1.59709692e-01 -3.21613222e-01 -1.26351154e+00 4.08875972e-01
4.43551093e-01 -9.09582615e-01 -4.79170561e-01 -1.74144477e-01
1.11673307e+00 3.08523685e-01 -4.39207673e-01 1.40315190e-01
7.45537579e-01 -8.18545938e-01 1.23629475e+00 1.57216117e-01
2.60391682e-01 -2.92170078e-01 1.40665233e+00 3.43547374e-01
-2.51499623e-01 8.76413509e-02 -4.67897981e-01 -3.66769195e-01
2.06988290e-01 1.10420263e+00 -8.19630563e-01 -7.20421001e-02
8.32594812e-01 3.93010139e-01 -6.36040330e-01 3.75156671e-01
-7.25961745e-01 6.57046974e-01 2.37255674e-02 -6.08848155e-01
2.64207035e-01 3.82203281e-01 2.65062094e-01 1.35997891e+00
-2.75431633e-01 -9.94959772e-02 9.63462070e-02 9.81809378e-01
-3.79517376e-01 4.94119674e-02 -4.91246313e-01 -1.39848515e-01
1.63551033e-01 1.41453791e+00 -7.15140104e-01 -3.56485188e-01
-3.32730591e-01 1.08558321e+00 5.77582002e-01 1.80188924e-01
-6.38665915e-01 -3.99425589e-02 4.37589198e-01 -2.77507911e-03
1.29661679e-01 4.13866460e-01 -7.12973237e-01 -1.13113904e+00
1.12473786e-01 -1.13117552e+00 1.07096285e-01 -1.04585350e+00
-1.91225433e+00 6.12735808e-01 -4.90958124e-01 -8.24287236e-01
1.10411927e-01 -1.18649900e+00 -4.54988852e-02 8.49665225e-01
-1.34256983e+00 -1.15916765e+00 -4.30525362e-01 -5.21998592e-02
4.45254207e-01 1.56618208e-02 1.14476144e+00 2.04751432e-01
-4.05733883e-01 4.57888097e-01 -2.58175194e-01 5.26176095e-01
1.12461245e+00 -1.42648280e+00 3.77445817e-01 4.73588139e-01
-1.76472932e-01 3.92541707e-01 1.37919307e+00 -7.39411294e-01
-7.65451252e-01 -9.96296525e-01 1.36272061e+00 -1.25546968e+00
8.95049274e-01 -6.08654022e-01 -9.62163985e-01 6.79941475e-01
3.61301154e-01 2.25544944e-01 1.25780058e+00 1.90256879e-01
-7.51202464e-01 1.16211183e-01 -1.27704823e+00 2.08060727e-01
5.45838475e-01 -6.67240679e-01 -6.12379789e-01 8.58625472e-01
5.54892123e-01 -7.12360322e-01 -6.34092927e-01 1.84860080e-01
5.43409109e-01 -9.18915570e-01 2.59530485e-01 -8.82948101e-01
8.75933409e-01 -4.05072868e-01 -1.62031665e-01 -1.44239569e+00
-2.12616473e-02 -2.87558913e-01 4.97168571e-01 1.47416341e+00
7.88534522e-01 -4.20080692e-01 7.64109015e-01 1.27457666e+00
8.86988919e-03 -4.07552481e-01 -5.24184704e-01 -2.24569932e-01
2.60965616e-01 -3.93099070e-01 5.65253496e-01 1.34461832e+00
1.40587121e-01 5.68687379e-01 -4.95500028e-01 2.62839347e-01
4.00559515e-01 -2.69074202e-01 7.24282086e-01 -1.36299789e+00
1.31407077e-03 -2.09639743e-01 -9.74804237e-02 -4.93407369e-01
2.78067440e-01 -6.40653014e-01 5.32746553e-01 -1.14676714e+00
1.61501616e-01 -1.03628270e-01 2.04600751e-01 5.54363906e-01
-4.63329613e-01 5.95786393e-01 -1.58123478e-01 7.55820498e-02
-8.50612104e-01 1.08967453e-01 9.24629509e-01 -8.49733725e-02
2.67220974e-01 -1.48894504e-01 -1.26086009e+00 1.17292929e+00
6.64846063e-01 -1.23579347e+00 1.18459873e-02 -8.23831797e-01
1.13763821e+00 -6.31159902e-01 6.21543884e-01 -4.06623989e-01
1.65862694e-01 1.90834045e-01 2.35906065e-01 3.56543362e-02
-3.23170781e-01 -8.91682863e-01 -9.51385275e-02 2.20085904e-01
-8.95769835e-01 -1.07678697e-01 1.30081236e-01 6.49624228e-01
-1.03268832e-01 -9.27825034e-01 6.18313551e-01 -6.38724566e-01
-1.78846747e-01 -2.38228247e-01 -4.19067979e-01 4.32811826e-01
8.07241976e-01 3.20453405e-01 -6.34548783e-01 -2.05200657e-01
-6.85093760e-01 1.52783290e-01 6.04091883e-01 2.31699586e-01
-4.06308681e-01 -1.10645342e+00 -1.03478682e+00 -2.44804636e-01
5.00137269e-01 -1.40230469e-02 7.27430061e-02 6.25205994e-01
-5.00185966e-01 6.71961084e-02 1.15495652e-01 -5.07465959e-01
-9.71227586e-01 5.50131202e-01 3.59552205e-01 -2.10765213e-01
2.65396953e-01 1.12246633e+00 -2.03069642e-01 -7.63302863e-01
1.38044313e-01 -2.05046222e-01 -2.08643928e-01 3.82406533e-01
8.13673019e-01 2.62954414e-01 3.43721271e-01 -7.83970773e-01
-3.17765832e-01 2.82671243e-01 -2.58002639e-01 1.24116547e-01
1.08171308e+00 -1.84214577e-01 -2.42515981e-01 7.55129397e-01
8.88725460e-01 -6.45684749e-02 -9.88884509e-01 -7.63072893e-02
4.07943092e-02 -2.24595636e-01 -1.52657598e-01 -1.30894887e+00
-8.72154415e-01 9.31961358e-01 4.67723548e-01 6.49981678e-01
4.13077861e-01 -1.32948458e-01 4.08705354e-01 3.83237302e-01
3.89947981e-01 -1.32179296e+00 -1.36763096e-01 3.40014011e-01
8.68543088e-01 -1.99444187e+00 -4.99137752e-02 -5.64991772e-01
-1.28796661e+00 8.01953912e-01 6.66482389e-01 -2.66426176e-01
4.28465128e-01 1.37156233e-01 7.20045030e-01 -4.18433368e-01
-4.15544271e-01 2.90730745e-01 1.55113846e-01 6.31434202e-01
1.13056827e+00 2.48435289e-01 -3.65204543e-01 1.07661176e+00
-3.34735885e-02 7.92978704e-02 6.01733565e-01 6.93949521e-01
-2.32041910e-01 -8.31261992e-01 -2.71428496e-01 1.57495886e-01
-9.68622506e-01 -1.52970314e-01 -7.40933836e-01 2.87444711e-01
2.14336008e-01 1.31874573e+00 -3.57665181e-01 -1.81860402e-02
4.84838367e-01 3.28616917e-01 -5.27206510e-02 -7.53121972e-01
-1.28644109e+00 -2.62000114e-01 6.53181553e-01 -3.79147679e-01
-7.93271005e-01 -5.62148869e-01 -8.38524938e-01 -7.99789727e-02
-5.72394311e-01 2.61904478e-01 7.91284561e-01 1.10070753e+00
3.53510737e-01 1.57423228e-01 1.86889499e-01 -8.52016091e-01
-8.06082785e-01 -1.24501181e+00 -3.27885211e-01 9.17229831e-01
3.72473180e-01 -2.52726823e-01 -5.99457800e-01 4.84612435e-01] | [9.6477632522583, 4.687412261962891] |
428b8136-b0e7-4aee-b578-447e36fe1710 | accurate-and-robust-scale-recovery-for | 2101.05995 | null | https://arxiv.org/abs/2101.05995v2 | https://arxiv.org/pdf/2101.05995v2.pdf | Accurate and Robust Scale Recovery for Monocular Visual Odometry Based on Plane Geometry | Scale ambiguity is a fundamental problem in monocular visual odometry. Typical solutions include loop closure detection and environment information mining. For applications like self-driving cars, loop closure is not always available, hence mining prior knowledge from the environment becomes a more promising approach. In this paper, with the assumption of a constant height of the camera above the ground, we develop a light-weight scale recovery framework leveraging an accurate and robust estimation of the ground plane. The framework includes a ground point extraction algorithm for selecting high-quality points on the ground plane, and a ground point aggregation algorithm for joining the extracted ground points in a local sliding window. Based on the aggregated data, the scale is finally recovered by solving a least-squares problem using a RANSAC-based optimizer. Sufficient data and robust optimizer enable a highly accurate scale recovery. Experiments on the KITTI dataset show that the proposed framework can achieve state-of-the-art accuracy in terms of translation errors, while maintaining competitive performance on the rotation error. Due to the light-weight design, our framework also demonstrates a high frequency of 20Hz on the dataset. | ['Dermot Kerr', 'Sonya Coleman', 'Shiwen Liang', 'Delong Zhu', 'Yunzhou Zhang', 'Rui Tian'] | 2021-01-15 | null | null | null | null | ['loop-closure-detection', 'monocular-visual-odometry'] | ['computer-vision', 'robots'] | [-1.33632779e-01 -2.11337402e-01 -3.57306510e-01 -4.41680163e-01
-6.22980893e-01 -2.48896718e-01 2.02539474e-01 1.42837852e-01
-5.09123623e-01 3.77644211e-01 -3.59511971e-01 -1.68972909e-01
-1.02137029e-01 -7.30418026e-01 -8.64682376e-01 -5.68571448e-01
1.47216335e-01 5.77346146e-01 3.83849770e-01 -3.09635758e-01
4.15738940e-01 7.16780186e-01 -1.60699093e+00 -6.42276168e-01
1.19534814e+00 1.17316771e+00 4.22053516e-01 3.33734512e-01
1.98006481e-01 2.93894678e-01 -1.77703857e-01 -8.28976333e-02
5.59078872e-01 1.53811201e-01 -1.11698881e-01 2.99528033e-01
7.24681854e-01 -4.32398736e-01 -2.98422903e-01 1.20235753e+00
3.72342885e-01 2.08489463e-01 2.72065043e-01 -1.25771999e+00
8.30610171e-02 -3.37681413e-01 -8.95842433e-01 -1.01544835e-01
5.16840458e-01 -5.66126853e-02 9.36616480e-01 -1.14106464e+00
7.31197834e-01 8.36965442e-01 6.18523955e-01 -1.55492142e-01
-1.00013912e+00 -6.83899879e-01 1.61802676e-02 2.85480231e-01
-1.82386780e+00 -3.99571806e-01 7.18687236e-01 -4.19501662e-01
8.16815436e-01 7.06739947e-02 7.30986893e-01 2.78844148e-01
4.28944558e-01 2.18588129e-01 8.82385135e-01 -3.06971312e-01
1.59639388e-01 4.05377112e-02 -8.15595984e-02 8.38258505e-01
8.67672563e-01 1.41204512e-02 -8.88985753e-01 3.71654481e-02
8.52536917e-01 2.85940409e-01 -9.98824611e-02 -1.06854236e+00
-1.30029273e+00 7.28435457e-01 7.76121020e-01 -2.63912976e-01
-3.39140236e-01 -9.87482723e-03 -1.09757908e-01 9.14635956e-02
3.23436260e-01 3.21849257e-01 -5.96948341e-02 -3.73648852e-02
-9.65898335e-01 2.93187052e-01 4.81529176e-01 1.28796268e+00
1.18645942e+00 -1.89602971e-01 4.75359529e-01 3.66195410e-01
5.37723482e-01 1.06107426e+00 2.14329466e-01 -8.09152544e-01
9.16832626e-01 8.80180776e-01 4.23487872e-01 -1.31476772e+00
-6.46218419e-01 -4.70978767e-01 -6.89035773e-01 4.36982572e-01
1.71076670e-01 4.78154607e-03 -8.26067030e-01 1.27804267e+00
8.92299175e-01 1.08395055e-01 -7.55763724e-02 1.26110995e+00
3.62094551e-01 2.77957916e-01 -6.08670950e-01 -1.20762497e-01
1.38531947e+00 -6.75443709e-01 -7.19326079e-01 -7.07018137e-01
5.35024762e-01 -1.02289152e+00 5.98776281e-01 3.99509490e-01
-4.59270477e-01 -5.47903121e-01 -1.56695807e+00 -3.84275496e-01
-3.26037407e-03 5.06568789e-01 5.39418459e-01 3.48433584e-01
-6.00760877e-01 4.14965302e-01 -1.00814962e+00 -4.80470955e-01
-2.39862993e-01 3.39097202e-01 -5.56766391e-01 3.55917518e-03
-7.88984001e-01 1.12962604e+00 2.86854953e-01 2.15528071e-01
-1.49223283e-01 -3.16812754e-01 -1.08969510e+00 -3.53945047e-01
6.43051565e-01 -7.65715837e-01 9.47395265e-01 -2.76365936e-01
-1.57019126e+00 9.60242748e-01 -4.01244909e-01 -6.20328724e-01
6.51183426e-01 -6.10729575e-01 -2.59393722e-01 2.67364144e-01
3.40296000e-01 2.69179314e-01 9.40401018e-01 -1.12831163e+00
-8.90604913e-01 -7.30478942e-01 -1.57859027e-01 5.83172560e-01
1.50399342e-01 -5.27863681e-01 -5.64203560e-01 -1.03483222e-01
1.24206173e+00 -1.41969776e+00 -3.48515868e-01 9.64148343e-02
-2.22187504e-01 2.51794249e-01 7.55636990e-01 -6.85403883e-01
1.16856229e+00 -2.13802648e+00 8.02593082e-02 4.60907370e-01
2.47183040e-01 -1.62535101e-01 5.21518767e-01 1.82556868e-01
3.43720853e-01 -5.77314556e-01 9.02607962e-02 -3.96274239e-01
-1.72168106e-01 1.49506003e-01 -2.19164893e-01 9.86035824e-01
-4.16091830e-02 5.22494555e-01 -7.70804226e-01 -3.04931909e-01
4.53111291e-01 3.39604437e-01 -4.50201482e-01 9.89360288e-02
1.94249272e-01 3.30551952e-01 -5.49066842e-01 6.92918837e-01
8.75954211e-01 -1.19048119e-01 7.30457855e-03 -2.49757543e-01
-5.25496423e-01 3.41987938e-01 -1.86512637e+00 2.00913429e+00
-3.66679132e-01 4.37306345e-01 1.55511677e-01 -4.86766607e-01
1.35853720e+00 -1.64153352e-01 3.92921627e-01 -5.80835700e-01
8.57737139e-02 6.58328295e-01 -1.91560879e-01 -2.73298174e-01
9.13269579e-01 7.94953704e-02 -9.31606144e-02 -1.14974640e-01
-3.48667532e-01 -6.30499303e-01 5.68549000e-02 -7.93645158e-02
7.28851497e-01 2.93039024e-01 7.07603037e-01 -1.75590917e-01
5.02301753e-01 1.12898491e-01 8.25020373e-01 1.71671212e-01
-8.02674219e-02 4.95500654e-01 -1.23760551e-01 -4.53372091e-01
-1.01187313e+00 -8.56717944e-01 -1.31986916e-01 2.73602605e-01
9.55108643e-01 -3.89122754e-01 -4.16792184e-01 -1.37466360e-02
3.66095752e-01 1.57713771e-01 -1.55959010e-01 -9.53368321e-02
-7.03713357e-01 -3.54240477e-01 -2.29759172e-01 4.17233109e-01
7.47541845e-01 -2.44282678e-01 -1.13775325e+00 1.56821489e-01
-3.47033471e-01 -1.44483376e+00 -2.44499102e-01 3.12543586e-02
-1.16960931e+00 -1.03757668e+00 -2.37606212e-01 -6.41758144e-01
6.96330428e-01 9.59460497e-01 6.36002004e-01 -1.16266444e-01
-6.08833618e-02 -1.18370406e-01 1.38803516e-02 -3.11537236e-01
1.49254307e-01 1.15086801e-01 4.86672342e-01 1.56109810e-01
2.04507008e-01 -4.91996467e-01 -6.53185189e-01 5.13809979e-01
-2.15597928e-01 2.56918162e-01 6.92401648e-01 4.66962546e-01
9.87878382e-01 -4.42379303e-02 -1.07480243e-01 -2.02676743e-01
-1.91949695e-01 -1.24808699e-01 -1.18727124e+00 -2.18694508e-01
-6.60686433e-01 1.02103002e-01 1.56931788e-01 -2.01193854e-01
-8.51333141e-01 5.56338608e-01 3.19823772e-01 -5.20564497e-01
1.81454271e-01 4.22011822e-01 -1.03620164e-01 -4.26533103e-01
7.55432069e-01 -3.78101766e-02 2.22137213e-01 -4.45068240e-01
3.43382329e-01 5.50994396e-01 5.94627917e-01 -2.27989435e-01
1.25127268e+00 9.59407985e-01 5.32712817e-01 -1.00081837e+00
-6.86474562e-01 -1.03161013e+00 -9.10033643e-01 -1.33009106e-01
7.80929506e-01 -1.34581423e+00 -6.58460796e-01 2.70698667e-01
-8.55109215e-01 2.39802241e-01 6.49838289e-03 9.27565813e-01
-6.62986755e-01 4.12656844e-01 -1.67644903e-01 -7.41890192e-01
-3.69177639e-01 -1.23534584e+00 1.21638489e+00 3.19516748e-01
1.12893285e-04 -3.85020077e-01 1.04646012e-01 4.57318753e-01
-9.28916633e-02 4.61067408e-01 1.43898070e-01 1.52215227e-01
-1.16809237e+00 -4.88371462e-01 -6.01558909e-02 -2.22132102e-01
9.76374298e-02 -4.12522443e-02 -7.89073229e-01 -4.91936743e-01
4.36655581e-02 8.45626369e-02 5.99132597e-01 2.27832139e-01
2.85817832e-01 1.03966042e-01 -5.99616766e-01 9.76319134e-01
1.55575156e+00 -8.99475515e-02 3.02703530e-01 7.62399614e-01
8.44139636e-01 3.13035667e-01 1.28634536e+00 5.25645971e-01
6.74761713e-01 1.13099539e+00 6.94332302e-01 -4.51726988e-02
1.91819325e-01 -2.93297738e-01 2.01946765e-01 8.82529318e-01
-1.61276087e-01 4.68327492e-01 -9.06814456e-01 3.92233938e-01
-2.04416060e+00 -5.74289978e-01 -2.43946925e-01 2.61957645e+00
3.16416502e-01 2.75976330e-01 -9.38019529e-02 1.84533536e-01
6.01308346e-01 1.09137103e-01 -6.61770582e-01 -4.61323336e-02
9.19411331e-02 -3.17901969e-01 1.14807463e+00 6.84958875e-01
-1.09895396e+00 8.48273456e-01 4.93567419e+00 3.45113009e-01
-1.40926254e+00 -1.90356150e-01 -1.81644842e-01 2.69025355e-03
2.03019366e-01 2.23795265e-01 -1.40408170e+00 1.46745414e-01
5.15576899e-01 -2.71371931e-01 3.98043513e-01 1.07451451e+00
1.51204765e-01 -3.62422138e-01 -6.24144197e-01 1.18422365e+00
1.51099294e-01 -9.48207378e-01 -4.45274830e-01 4.20274377e-01
7.93540180e-01 3.48438472e-01 -1.34968072e-01 -1.72618911e-01
1.11215059e-02 -3.17160368e-01 1.00379324e+00 2.98675537e-01
8.35442245e-01 -8.19216609e-01 5.47273815e-01 8.13843191e-01
-1.63923538e+00 -3.96087430e-02 -5.47550261e-01 -3.15703958e-01
1.76754937e-01 9.00548279e-01 -9.17568088e-01 1.02655661e+00
6.55676305e-01 7.00062752e-01 -4.47081715e-01 1.17554677e+00
-4.70662475e-01 -1.87068269e-01 -7.73915231e-01 1.67284682e-01
-1.03582852e-01 -5.48172593e-01 7.32068419e-01 6.04859173e-01
8.21295500e-01 5.14838621e-02 5.26191413e-01 3.48038226e-01
1.82534114e-01 2.60222107e-01 -6.70757771e-01 5.51596820e-01
5.55283487e-01 1.45745993e+00 -7.02039361e-01 -1.44013360e-01
-4.18821961e-01 6.95678294e-01 3.69920194e-01 1.04234055e-01
-7.55321562e-01 -2.81954706e-01 9.32374358e-01 3.50773454e-01
3.56310248e-01 -7.65704095e-01 -5.39945364e-01 -1.39471018e+00
6.09632671e-01 -7.34591246e-01 8.79335850e-02 -6.61236048e-01
-4.00947094e-01 4.23578858e-01 -2.04430968e-01 -1.97515547e+00
-2.86257833e-01 -5.62371910e-01 -1.33128896e-01 7.54687190e-01
-1.58029497e+00 -9.08081651e-01 -7.55858123e-01 3.88156444e-01
2.43930578e-01 1.89020291e-01 4.48575169e-01 9.72653031e-02
-2.98861384e-01 1.44393414e-01 -3.82030308e-02 -2.29153916e-01
7.61561096e-01 -9.81132269e-01 4.20016438e-01 1.20454562e+00
1.87851474e-01 7.73883939e-01 9.07464385e-01 -8.20412219e-01
-1.82195163e+00 -8.28798950e-01 9.82398331e-01 -2.73219198e-01
4.54646707e-01 -3.06172758e-01 -7.05151737e-01 6.60846889e-01
-4.47582573e-01 1.51992276e-01 -2.22790577e-02 -4.08502892e-02
-2.18140781e-01 -4.46186006e-01 -1.04311001e+00 5.02999246e-01
1.09627283e+00 -4.38526422e-01 -5.13753235e-01 1.49399608e-01
5.29965043e-01 -1.34618688e+00 -6.48869872e-01 6.46999717e-01
5.97283542e-01 -8.55939627e-01 1.09848285e+00 2.28606641e-01
-1.69055745e-01 -9.40371931e-01 -1.55461133e-01 -1.04073465e+00
-2.55750299e-01 -6.40238285e-01 -2.33197734e-01 7.73336470e-01
1.95628814e-02 -7.92933106e-01 9.02278841e-01 3.31201226e-01
-5.41225187e-02 -5.34138322e-01 -1.24071920e+00 -8.00365269e-01
-8.09477329e-01 -3.42189521e-01 5.30930281e-01 6.32311583e-01
-1.46836668e-01 3.98391098e-01 -3.49002540e-01 8.95440340e-01
9.01006460e-01 5.33375263e-01 1.51191342e+00 -1.36788702e+00
-1.59406796e-01 3.41595151e-02 -8.58608186e-01 -1.49520767e+00
-1.86355963e-01 -5.50676405e-01 2.45706022e-01 -1.44478190e+00
-2.79449999e-01 -3.65957201e-01 1.63886473e-01 9.55312327e-03
-1.56377092e-01 2.30741337e-01 1.70001417e-01 4.39577222e-01
-4.28545356e-01 4.55917329e-01 1.02038348e+00 2.70920843e-01
-3.67376715e-01 1.34345755e-01 -2.78842658e-01 9.22354221e-01
5.83985925e-01 -3.18138570e-01 -3.06169748e-01 -4.17574584e-01
4.99054879e-01 1.50359407e-01 2.01867104e-01 -1.36150265e+00
5.44552982e-01 -2.68327087e-01 2.42467210e-01 -1.04136169e+00
6.96741879e-01 -9.58335638e-01 2.64694601e-01 5.24012327e-01
4.60386783e-01 1.61324754e-01 6.22051135e-02 5.11981308e-01
-2.99721241e-01 2.76608597e-02 7.17467725e-01 2.40184054e-01
-9.29401219e-01 3.33115488e-01 2.28376016e-01 -3.67659450e-01
1.05644727e+00 -5.48056424e-01 -1.77124724e-01 -2.64466047e-01
-2.34180704e-01 2.93752909e-01 1.02643108e+00 4.63252008e-01
5.71537256e-01 -1.38087654e+00 -3.54629129e-01 5.63999832e-01
5.28473675e-01 4.52938676e-01 8.06540996e-03 1.06965184e+00
-7.95897126e-01 4.38003540e-01 2.01588357e-03 -1.12129736e+00
-1.29394507e+00 3.41165721e-01 2.18237236e-01 -7.59006366e-02
-6.91039264e-01 4.39308822e-01 2.08568331e-02 -2.99555928e-01
-2.19016999e-01 -7.14694500e-01 6.14602305e-02 3.50163840e-02
3.80989194e-01 5.43226063e-01 4.51369703e-01 -1.03548551e+00
-5.31776190e-01 1.25546622e+00 2.30470359e-01 -2.07427859e-01
1.00399280e+00 -5.13066351e-01 2.24869102e-01 3.85845065e-01
1.00515270e+00 3.03904325e-01 -1.48072708e+00 -3.92528087e-01
-3.37727647e-03 -6.88132465e-01 8.54747817e-02 8.45357105e-02
-6.48020625e-01 7.03745246e-01 5.81067920e-01 -1.70142695e-01
8.84193957e-01 -3.93858254e-01 6.70156538e-01 5.58588266e-01
8.94931912e-01 -1.29320192e+00 -2.35959336e-01 6.03291512e-01
8.11630487e-01 -1.55663276e+00 6.47059500e-01 -8.39486778e-01
-3.33355159e-01 1.06961834e+00 4.71919000e-01 -2.86832929e-01
3.71255368e-01 5.13093062e-02 2.68716305e-01 -2.17120126e-02
-2.58159399e-01 -3.41591090e-01 3.89550269e-01 2.71824956e-01
-5.04156575e-02 2.38131702e-01 -2.02861637e-01 -8.16800967e-02
-5.99844337e-01 -1.24103419e-01 3.21082830e-01 1.07057881e+00
-8.88607800e-01 -7.58110404e-01 -7.83556998e-01 3.88044268e-02
1.85468048e-02 3.24084908e-01 7.88202733e-02 8.06294203e-01
-6.63793879e-03 8.60130548e-01 6.53383136e-02 -3.75178903e-01
6.41511679e-01 -1.65980294e-01 4.43641186e-01 -5.36314666e-01
5.92503473e-02 2.04944745e-01 -5.33065759e-03 -1.04701245e+00
-2.48879477e-01 -6.47066236e-01 -1.47242701e+00 -2.47882590e-01
-6.09506249e-01 -1.70896724e-01 1.05806649e+00 8.38396430e-01
4.69836831e-01 -7.35133812e-02 7.72107601e-01 -1.02903128e+00
-6.44185841e-01 -6.08111799e-01 -7.07879186e-01 3.12267244e-01
5.75103521e-01 -9.90179718e-01 -3.78257155e-01 -2.36245200e-01] | [7.494079113006592, -2.129279375076294] |
f468c8aa-181d-4b53-bf3c-2362a0e40680 | context-driven-satirical-news-generation | null | null | https://aclanthology.org/2020.figlang-1.5 | https://aclanthology.org/2020.figlang-1.5.pdf | Context-Driven Satirical News Generation | While mysterious, humor likely hinges on an interplay of entities, their relationships, and cultural connotations. Motivated by the importance of context in humor, we consider methods for constructing and leveraging contextual representations in generating humorous text. Specifically, we study the capacity of transformer-based architectures to generate funny satirical headlines, and show that both language models and summarization models can be fine-tuned to regularly generate headlines that people find funny. Furthermore, we find that summarization models uniquely support satire-generation by enabling the generation of topical humorous text. Outside of our formal study, we note that headlines generated by our model were accepted via a competitive process into a satirical newspaper, and one headline was ranked as high or better than 73{\%} of human submissions. As part of our work, we contribute a dataset of over 15K satirical headlines paired with ranked contextual information from news articles and Wikipedia. | ['Nam Do', 'Zachary Horvitz', 'Michael L. Littman'] | 2020-07-01 | null | null | null | ws-2020-7 | ['news-generation'] | ['natural-language-processing'] | [-4.98025306e-02 4.35435206e-01 5.79964295e-02 -6.92561790e-02
-7.32958198e-01 -7.69392610e-01 1.13728416e+00 3.26308131e-01
-1.59139737e-01 9.14656699e-01 1.43819427e+00 -5.20344079e-02
2.48856664e-01 -8.31432045e-01 -5.15839875e-01 1.01429403e-01
3.44389647e-01 4.51611489e-01 -1.21720940e-01 -1.14459836e+00
5.74618518e-01 -8.46918896e-02 -1.31031835e+00 9.26351011e-01
1.00453818e+00 3.51509482e-01 -5.41624799e-02 8.35884154e-01
-2.05456436e-01 1.46456599e+00 -9.92531419e-01 -8.66947412e-01
-1.32949620e-01 -8.68649602e-01 -8.28984678e-01 -1.51248425e-01
6.82080209e-01 -1.27480641e-01 -4.88737643e-01 7.47462153e-01
6.48339152e-01 1.24064334e-01 7.17475176e-01 -8.23059261e-01
-1.13277996e+00 1.42359316e+00 -6.23476133e-02 1.70853093e-01
6.64161980e-01 1.61131561e-01 1.52695251e+00 -1.04443300e+00
1.11273670e+00 1.20840502e+00 6.90386713e-01 6.63935781e-01
-1.16267586e+00 -2.61920273e-01 -2.56891638e-01 1.04770325e-01
-7.91909277e-01 -5.08444428e-01 8.96185696e-01 -5.49216628e-01
8.52361739e-01 5.92998803e-01 6.22274578e-01 1.62165236e+00
-5.06498180e-02 9.88950789e-01 9.30449247e-01 -2.93788075e-01
3.20927538e-02 1.70421645e-01 2.26223022e-02 4.60932702e-01
8.24157745e-02 -4.78976697e-01 -9.06832099e-01 -2.44925454e-01
4.15875852e-01 -3.48790139e-01 -4.53884661e-01 2.19166741e-01
-1.33000278e+00 1.12504888e+00 6.31742656e-01 3.61552894e-01
-3.72987658e-01 9.77219269e-02 6.29418731e-01 2.69304127e-01
5.16917825e-01 1.35318446e+00 6.74236566e-02 -2.13376671e-01
-1.09489787e+00 7.51111150e-01 1.16559196e+00 8.26213300e-01
3.06978256e-01 -4.37429808e-02 -5.96022725e-01 9.40906703e-01
-3.65102351e-01 5.66915154e-01 5.73357761e-01 -7.61536479e-01
3.90293270e-01 5.10273695e-01 3.26619834e-01 -1.46015942e+00
-3.81596386e-01 -5.48619092e-01 -6.15737677e-01 -4.28088009e-01
1.90473497e-01 -2.58779228e-01 -2.53634423e-01 1.70243526e+00
-3.14202398e-01 -4.75832850e-01 6.45065913e-03 9.47274446e-01
1.24798548e+00 8.19790661e-01 -7.52174556e-02 -1.80892482e-01
1.29338944e+00 -8.09696198e-01 -6.43116236e-01 -3.42174262e-01
5.20003796e-01 -1.18168879e+00 1.36878431e+00 3.17911863e-01
-1.48051679e+00 -1.54672831e-01 -9.27219212e-01 -5.09743452e-01
-1.21814556e-01 2.18971595e-02 2.46030867e-01 9.14699435e-02
-9.18531954e-01 7.13433743e-01 -4.58119735e-02 -4.10260588e-01
2.29247048e-01 -5.07875681e-01 -1.00673981e-01 2.98857003e-01
-1.47716153e+00 1.19324768e+00 -1.06315643e-01 -3.58934820e-01
-7.16365814e-01 -6.88888729e-01 -6.49864435e-01 -1.35703474e-01
-4.07355204e-02 -9.37153876e-01 1.47163737e+00 -8.40692341e-01
-1.14595711e+00 8.62347722e-01 -7.05771074e-02 -7.02466905e-01
6.63799822e-01 -4.68962312e-01 -2.11012080e-01 1.39124542e-01
1.98037639e-01 3.77218843e-01 6.10123038e-01 -1.05584145e+00
-2.54730821e-01 9.44244489e-02 2.63151258e-01 3.32634360e-01
-5.95644653e-01 1.68374971e-01 6.52864426e-02 -8.08643162e-01
-4.04989421e-01 -6.22096181e-01 -1.01420119e-01 -6.89167917e-01
-7.59105504e-01 -2.61379957e-01 4.14815247e-01 -8.65393519e-01
1.69088376e+00 -1.73306882e+00 3.40750992e-01 -6.71074027e-03
4.72281337e-01 -2.58483022e-01 -1.10337064e-01 1.07207310e+00
3.50235939e-01 3.16292256e-01 8.58711228e-02 -2.81609893e-01
3.74618977e-01 -2.43379474e-01 -1.00520790e+00 8.41685086e-02
2.66719684e-02 1.15472698e+00 -1.19711828e+00 -2.85383314e-01
-1.63486049e-01 2.33803362e-01 -6.93062186e-01 1.48799181e-01
-5.98078251e-01 -4.13691960e-02 -3.65910947e-01 2.79680461e-01
-2.50427835e-02 -4.82811362e-01 1.17278486e-01 -1.39120072e-01
-1.10181883e-01 9.98559415e-01 -3.38234037e-01 1.21705735e+00
-6.81211472e-01 8.07647228e-01 -2.49261737e-01 9.58232023e-03
1.14512134e+00 1.28092170e-01 3.99502777e-02 -6.99975491e-01
5.61980344e-02 3.13156635e-01 -3.34716558e-01 -2.79249579e-01
1.44928575e+00 -5.33302188e-01 -4.96723175e-01 8.49526465e-01
-1.46690756e-01 -4.52609926e-01 4.09392774e-01 8.80266547e-01
1.26477611e+00 -1.99938238e-01 2.25163072e-01 -3.01476181e-01
1.61075294e-01 3.52393806e-01 6.07018769e-02 8.90809178e-01
1.83483511e-01 8.61060321e-01 7.47958422e-01 -4.26649004e-01
-1.53715515e+00 -8.70271325e-01 -5.32634482e-02 1.28804469e+00
-1.04559496e-01 -1.03261590e+00 -6.78859234e-01 -4.43336397e-01
-4.25637439e-02 1.32212508e+00 -6.72089994e-01 -4.73613217e-02
-7.11718142e-01 -6.58500195e-01 6.98195338e-01 1.79314211e-01
5.05022556e-02 -1.38628423e+00 -6.73947990e-01 4.14524853e-01
-8.02625895e-01 -8.13651800e-01 -5.05333364e-01 -7.95690119e-02
-3.80247235e-01 -9.13379967e-01 -7.07084835e-01 -4.40016717e-01
2.84713328e-01 2.65480012e-01 1.76873720e+00 2.07509667e-01
-1.13169849e-01 1.32647872e-01 -7.28504539e-01 -2.14514256e-01
-7.10531414e-01 2.79299468e-01 -7.75303394e-02 -4.78324294e-01
1.61344260e-01 -6.56918049e-01 -5.23030639e-01 -5.76376617e-02
-8.46258998e-01 4.19895142e-01 1.40218943e-01 9.74631488e-01
-2.27860630e-01 -5.15212715e-01 8.54223967e-01 -1.25922596e+00
1.30581057e+00 -8.01904857e-01 1.51267007e-01 -1.02853641e-01
-2.96650529e-01 -3.86990840e-03 9.37724888e-01 -7.89629966e-02
-8.51110995e-01 -3.89782369e-01 8.69399831e-02 -1.27723306e-01
4.54758614e-01 5.98727167e-01 5.04626632e-01 5.37228107e-01
1.55467165e+00 -1.30722048e-02 -1.35590374e-01 -4.52423126e-01
9.87730563e-01 5.54369211e-01 7.52511203e-01 -7.32541203e-01
8.42406631e-01 2.24266797e-01 -4.57795799e-01 -7.65352011e-01
-1.23480272e+00 -3.41502041e-01 -1.32076338e-01 -2.61490256e-01
5.48962355e-01 -9.63590086e-01 -4.27115083e-01 6.42805696e-02
-1.45485580e+00 -2.20767453e-01 -4.63926822e-01 -2.09366873e-01
-5.32788455e-01 3.44263077e-01 -9.46242571e-01 -5.35824776e-01
-8.47074449e-01 -4.35197920e-01 7.29159772e-01 1.27997234e-01
-1.01065481e+00 -9.09400165e-01 3.09260398e-01 5.93091786e-01
6.67415380e-01 4.09896135e-01 1.06791067e+00 -8.54461849e-01
-2.39499837e-01 -2.70004630e-01 -1.70190036e-01 -7.02596232e-02
-2.04617232e-01 -5.98978177e-02 -8.15622807e-01 3.05163041e-02
-2.47364327e-01 -8.42748702e-01 1.00197828e+00 -6.54998571e-02
6.01294279e-01 -1.26354575e+00 2.41088957e-01 3.58633101e-02
9.25014138e-01 -5.99407256e-01 7.44446337e-01 5.23308873e-01
6.33136511e-01 7.43724346e-01 1.87333405e-01 9.54784453e-01
5.96897900e-01 4.50404435e-01 3.01969379e-01 1.51300848e-01
-3.17503065e-01 -8.79207790e-01 5.83350122e-01 9.51502085e-01
1.27434954e-01 -1.91387326e-01 -6.98563993e-01 8.22257519e-01
-1.79345465e+00 -1.71049190e+00 -4.14178938e-01 1.89497805e+00
1.30817127e+00 3.92328128e-02 3.92695159e-01 -3.09697598e-01
5.46536088e-01 5.74666440e-01 -3.85325290e-02 -5.99438667e-01
-6.68926537e-01 2.59615719e-01 1.12684160e-01 8.10248137e-01
-6.86149597e-01 1.14223850e+00 7.02103567e+00 6.23100758e-01
-8.62332106e-01 -1.95274159e-01 5.18098772e-01 -5.32175601e-01
-1.20958984e+00 -7.77303204e-02 -6.25300467e-01 4.09435123e-01
7.49711275e-01 -6.55795038e-01 5.72780669e-01 8.99642110e-01
2.48959616e-01 1.84192002e-01 -1.07567918e+00 6.62337899e-01
5.02231240e-01 -1.80154514e+00 4.54389453e-01 -3.63392651e-01
1.10289824e+00 -2.67440509e-02 1.33757234e-01 3.70496571e-01
8.04203391e-01 -1.32095921e+00 1.08813477e+00 5.44662535e-01
4.43100572e-01 -6.13603294e-01 5.69112122e-01 2.36573279e-01
-4.10918325e-01 -2.55845778e-04 -3.39183927e-01 -3.77328277e-01
5.01286566e-01 8.12597811e-01 -1.16258347e+00 1.10002786e-01
3.03710133e-01 7.78329313e-01 -8.99515450e-01 7.20309913e-01
-5.52684546e-01 6.11029744e-01 -7.81930797e-03 -6.01615548e-01
2.67508209e-01 2.77886569e-01 8.23221147e-01 1.70183802e+00
2.61792153e-01 1.16540223e-01 3.09270397e-02 1.11907089e+00
-4.55988646e-01 3.36186588e-01 -6.16493881e-01 -3.44252557e-01
5.52596748e-01 1.56688261e+00 -1.50392234e-01 -3.68765980e-01
1.50964439e-01 9.54703510e-01 4.51622576e-01 1.05863571e-01
-5.81701577e-01 -2.57609874e-01 2.50950158e-01 3.79347235e-01
-6.44645989e-02 -3.53151336e-02 -8.09959710e-01 -1.29141140e+00
-1.04925610e-01 -1.21917307e+00 3.09316605e-01 -1.15989351e+00
-1.71704829e+00 7.35669792e-01 -3.96970779e-01 -7.27146447e-01
-4.70572054e-01 -1.59098700e-01 -9.54057336e-01 7.53461123e-01
-9.09556568e-01 -1.16119921e+00 -2.90342093e-01 1.67887345e-01
6.33770466e-01 -1.88091904e-01 5.71142435e-01 -1.21893428e-01
1.45925032e-02 3.73218447e-01 -1.62861869e-01 5.32538369e-02
8.35451365e-01 -1.35067260e+00 6.27837002e-01 7.30511010e-01
3.43571812e-01 8.65293205e-01 1.45246613e+00 -7.18399644e-01
-1.07671249e+00 -9.28913951e-01 1.55272305e+00 -9.64377582e-01
1.11563396e+00 -1.49724096e-01 -6.74245298e-01 5.40802062e-01
6.83300495e-01 -8.58351111e-01 6.42989218e-01 4.69595551e-01
-8.50385189e-01 4.71690416e-01 -7.88781583e-01 9.78458107e-01
1.04937887e+00 -7.17001021e-01 -9.66251373e-01 6.14632964e-01
8.05699289e-01 -3.62895280e-01 -4.50279087e-01 7.08634257e-02
2.55677998e-01 -9.16213751e-01 6.72533035e-01 -9.38717484e-01
1.70237792e+00 -1.41559213e-01 -1.82356000e-01 -1.46560061e+00
-5.10171950e-01 -8.89962435e-01 -7.95792565e-02 1.19784033e+00
6.22515261e-01 9.08057392e-02 4.65852767e-01 5.16328275e-01
-3.66255790e-01 -5.27079761e-01 -4.67932820e-01 -1.86643749e-01
3.63861918e-01 -1.03099063e-01 2.82506585e-01 1.02821970e+00
8.38146031e-01 9.51716185e-01 -7.67947912e-01 -7.41923153e-01
2.95029759e-01 2.80056775e-01 8.52725267e-01 -8.06603193e-01
-4.33430523e-01 -7.59842396e-01 1.00917414e-01 -9.53519583e-01
1.15198363e-02 -1.19572115e+00 7.46348351e-02 -1.63337493e+00
6.79504991e-01 -1.28039062e-01 7.05374107e-02 4.57241833e-01
-2.32271060e-01 4.78091925e-01 4.60985005e-01 4.08126920e-01
-7.62594283e-01 5.73290467e-01 1.26313961e+00 5.67705138e-03
-2.60683805e-01 -5.25848985e-01 -1.31549108e+00 4.82950985e-01
6.96813822e-01 2.75118239e-02 -1.70399517e-01 -2.73400962e-01
1.04798365e+00 -4.91031520e-02 5.31115651e-01 -7.13295579e-01
5.54593578e-02 -3.14898401e-01 2.63513803e-01 -4.38424230e-01
2.01615795e-01 1.11910418e-01 -9.39671025e-02 1.43680140e-01
-9.44968283e-01 2.36473620e-01 -9.44596082e-02 2.17784464e-01
-2.46026460e-02 -1.23552501e-01 7.05009282e-01 -3.51317525e-01
-3.20850134e-01 -2.69811511e-01 -4.23629314e-01 6.43805444e-01
2.64533877e-01 2.69846767e-01 -1.00547528e+00 -1.09103680e+00
-2.97601610e-01 -1.71095114e-02 7.70187557e-01 6.27688825e-01
4.70264852e-01 -1.32499135e+00 -1.20303392e+00 -4.92911309e-01
2.63562024e-01 -4.46678460e-01 1.24447607e-01 5.63400984e-01
-5.34859061e-01 2.09262550e-01 -1.51255041e-01 1.30303562e-01
-8.39276791e-01 2.99093902e-01 -5.03992289e-02 -1.97392523e-01
-7.66320467e-01 8.09030592e-01 1.25755727e-01 -3.81618649e-01
-1.83312625e-01 2.35960960e-01 -2.87352473e-01 2.56644368e-01
7.75085270e-01 3.45655710e-01 -1.01803705e-01 -7.30758607e-01
5.19956648e-02 -1.93504035e-01 -1.88995689e-01 -4.44064736e-01
1.30440485e+00 -2.24817842e-02 -3.42394739e-01 4.86024022e-01
8.27946424e-01 6.82370961e-01 -6.82576239e-01 -9.86008346e-02
-2.92364880e-02 -2.15025708e-01 -3.42432261e-01 -1.23129392e+00
-3.07554483e-01 4.84647930e-01 -6.41346157e-01 4.56856012e-01
5.95068276e-01 1.46022484e-01 1.21960354e+00 4.24297422e-01
2.87973613e-01 -1.17934525e+00 6.05551481e-01 1.06545317e+00
1.44797182e+00 -7.64537752e-01 2.10407395e-02 1.88911706e-02
-1.05735302e+00 1.13641071e+00 5.16842604e-01 -3.02965909e-01
-2.04403415e-01 -5.96652329e-02 1.13278329e-01 -3.16042304e-01
-1.17758024e+00 1.90197423e-01 3.96881878e-01 2.46829212e-01
9.67875361e-01 1.63000256e-01 -7.63462305e-01 8.66905451e-01
-1.24998963e+00 -2.23752856e-01 1.06440699e+00 2.05542013e-01
-9.16858673e-01 -5.72146237e-01 -2.19842613e-01 2.87577868e-01
-4.29808319e-01 -5.14496684e-01 -1.10043943e+00 2.13181913e-01
-3.89692664e-01 8.21643531e-01 -1.92092434e-01 -6.08334184e-01
6.98946193e-02 -4.29516546e-02 2.75669128e-01 -6.43562496e-01
-1.14011943e+00 -3.77417296e-01 8.43416393e-01 -2.34356865e-01
2.75711834e-01 -4.97564435e-01 -1.09846234e+00 -9.71428096e-01
2.51889139e-01 3.30103368e-01 3.62736464e-01 7.34572768e-01
3.34728092e-01 1.90514386e-01 7.70801842e-01 -5.66789269e-01
-7.93743312e-01 -1.10458279e+00 -2.85796642e-01 9.27240133e-01
1.61815569e-01 1.24822054e-02 -5.23869455e-01 6.95610493e-02] | [8.93290901184082, 10.967108726501465] |
ab57ff6a-e5dd-4f40-badf-c5c0c970a2f9 | online-multi-object-tracking-with-historical | 1805.10916 | null | http://arxiv.org/abs/1805.10916v4 | http://arxiv.org/pdf/1805.10916v4.pdf | Online Multi-Object Tracking with Historical Appearance Matching and Scene Adaptive Detection Filtering | In this paper, we propose the methods to handle temporal errors during
multi-object tracking. Temporal error occurs when objects are occluded or noisy
detections appear near the object. In those situations, tracking may fail and
various errors like drift or ID-switching occur. It is hard to overcome
temporal errors only by using motion and shape information. So, we propose the
historical appearance matching method and joint-input siamese network which was
trained by 2-step process. It can prevent tracking failures although objects
are temporally occluded or last matching information is unreliable. We also
provide useful technique to remove noisy detections effectively according to
scene condition. Tracking performance, especially identity consistency, is
highly improved by attaching our methods. | ['Young-min Song', 'Young-chul Yoon', 'Kwangjin Yoon', 'Moongu Jeon', 'Abhijeet Boragule'] | 2018-05-28 | null | null | null | null | ['online-multi-object-tracking'] | ['computer-vision'] | [-1.17520444e-01 -8.39854240e-01 3.85249592e-03 5.09013794e-02
-2.54013330e-01 -6.25607133e-01 3.49501818e-01 -2.69214004e-01
-4.88876551e-01 8.46259713e-01 -3.83634567e-01 2.78294116e-01
1.28462380e-02 -2.64363348e-01 -6.54387236e-01 -7.09891796e-01
1.19563388e-02 4.60573554e-01 8.59809160e-01 1.02927007e-01
1.89450502e-01 6.98723853e-01 -1.68464911e+00 -7.99612775e-02
6.70519948e-01 7.64552653e-01 2.55341023e-01 8.06519449e-01
-2.26605490e-01 4.45757508e-01 -8.17283094e-01 -1.42499641e-01
4.46168602e-01 -2.44370863e-01 -3.11807930e-01 9.31259468e-02
7.67082989e-01 -4.67983931e-01 -1.77006140e-01 1.32764781e+00
3.81917387e-01 2.66115278e-01 4.96713638e-01 -1.62562692e+00
-6.46414220e-01 4.61254753e-02 -9.93542850e-01 5.92238426e-01
2.42998242e-01 2.42038459e-01 2.18619570e-01 -8.42357039e-01
7.00111270e-01 1.46737576e+00 1.04033279e+00 8.22466910e-01
-8.21175277e-01 -8.55791688e-01 5.86686790e-01 3.21545690e-01
-1.46268129e+00 -5.32777488e-01 6.12106442e-01 -4.51559752e-01
3.90781611e-01 1.60646692e-01 4.59753871e-01 1.10006583e+00
3.86247605e-01 9.84344244e-01 6.41674399e-01 -1.18922934e-01
-2.72644758e-01 4.52669561e-02 1.53054580e-01 4.61000383e-01
6.66678488e-01 3.82173538e-01 -2.42936447e-01 -1.90322801e-01
7.76872754e-01 4.77048069e-01 2.29079723e-02 -1.81390911e-01
-1.48318565e+00 2.56351262e-01 2.05150545e-01 3.33153218e-01
-2.83268124e-01 2.94446945e-01 4.93694663e-01 4.57927078e-01
1.52069196e-01 -1.45402700e-01 -2.88271576e-01 1.13775879e-01
-9.83445644e-01 3.77132267e-01 2.90663362e-01 1.31680727e+00
3.84868652e-01 2.85824597e-01 -2.99271733e-01 4.98733729e-01
4.69650656e-01 8.55066419e-01 4.57516611e-01 -8.04955125e-01
1.87699735e-01 2.78581977e-01 6.26178503e-01 -1.00096834e+00
-4.13787782e-01 -3.64087909e-01 -8.86084020e-01 6.29008412e-01
4.69094843e-01 -1.38389364e-01 -1.12348187e+00 1.62105584e+00
7.35601783e-01 5.27466953e-01 -1.55071199e-01 1.09758830e+00
7.65060663e-01 3.66217017e-01 2.95989722e-01 -6.44559920e-01
1.17618120e+00 -9.51139867e-01 -1.40180635e+00 2.47105137e-02
7.48112872e-02 -1.25501907e+00 2.21548095e-01 3.53868037e-01
-1.12564194e+00 -1.12978208e+00 -1.05359864e+00 4.47274804e-01
-4.21842158e-01 1.30882189e-01 3.05548072e-01 3.46909016e-01
-8.41947496e-01 7.05473185e-01 -8.58797193e-01 -4.40194100e-01
3.44941527e-01 5.75072467e-01 -2.36756727e-01 2.65196443e-01
-1.03628576e+00 1.04928386e+00 3.49775910e-01 5.08847117e-01
-8.30058455e-01 -2.68586129e-01 -4.52998698e-01 -5.38906872e-01
4.14042264e-01 -7.89247632e-01 1.14783931e+00 -1.19156229e+00
-1.14281344e+00 5.31716883e-01 -3.45136613e-01 -2.29310408e-01
9.63219583e-01 -4.05909091e-01 -8.65049303e-01 -2.73587704e-01
2.52822071e-01 4.21892434e-01 1.02225029e+00 -1.22800434e+00
-9.36415017e-01 -3.44345480e-01 -6.07262373e-01 1.93290427e-01
4.07903343e-02 3.27736169e-01 -6.00514829e-01 -7.50453413e-01
3.48772734e-01 -1.04744375e+00 -2.36808375e-01 5.91294885e-01
-9.81035680e-02 -3.25871885e-01 1.62218630e+00 -5.49470603e-01
1.13281655e+00 -2.16148806e+00 -9.56825092e-02 -3.61426398e-02
1.93858966e-02 4.93185759e-01 -4.43939157e-02 1.58668142e-02
2.07462564e-01 -9.97549593e-02 2.65867531e-01 -4.67210472e-01
-2.98743010e-01 1.16400398e-01 -1.63203895e-01 6.82483494e-01
1.06828526e-01 5.67150831e-01 -7.54577100e-01 -8.54032159e-01
3.17150384e-01 3.42692226e-01 3.01989354e-02 1.30031466e-01
-2.22701102e-01 7.34564304e-01 -6.37724221e-01 9.72970963e-01
1.25234294e+00 -1.50113463e-01 -3.87272507e-01 -3.67011428e-01
-3.67995054e-01 -5.49388289e-01 -1.83338487e+00 1.62247765e+00
3.50113958e-02 5.10614872e-01 2.24660560e-01 -4.54753369e-01
7.80836761e-01 4.74641502e-01 6.39693022e-01 -3.56113613e-01
1.52797133e-01 1.05003431e-01 -1.06321253e-01 -6.73127472e-01
6.72347605e-01 2.10760638e-01 4.24662441e-01 9.12190601e-02
-3.32442433e-01 5.15696883e-01 1.13785341e-01 -4.62475792e-02
9.97171581e-01 5.95152736e-01 -7.58314645e-03 1.17602207e-01
6.82440579e-01 1.13238908e-01 1.11888957e+00 8.76747847e-01
-8.16587090e-01 5.06271660e-01 -4.55374956e-01 -6.04828775e-01
-9.31039393e-01 -1.20454514e+00 -2.48679981e-01 8.01950157e-01
7.97908902e-01 2.17475057e-01 -5.83912581e-02 -5.79378963e-01
2.69306928e-01 -1.10395141e-01 -4.01933521e-01 -9.10534859e-02
-8.88310015e-01 -5.84399343e-01 3.31718177e-01 6.60864294e-01
6.41145289e-01 -9.74671721e-01 -2.74210304e-01 5.60480893e-01
1.02006897e-01 -1.02430999e+00 -8.09419215e-01 -3.78156066e-01
-1.03938961e+00 -1.09829021e+00 -7.25355089e-01 -7.87331998e-01
6.97901785e-01 6.14155114e-01 6.92335844e-01 6.09809875e-01
-2.63952762e-01 3.09995413e-02 -2.09478259e-01 -4.15457487e-01
-2.36916542e-01 -3.91534060e-01 5.41079402e-01 -2.26252358e-02
2.67219543e-01 -2.80692846e-01 -6.31169260e-01 6.14471972e-01
-5.71469009e-01 -3.14864844e-01 4.83812034e-01 7.25583613e-01
6.35152400e-01 -2.64612492e-03 6.01916730e-01 -3.84154201e-01
1.96713328e-01 -1.61029831e-01 -1.01119924e+00 4.24615473e-01
-3.82058799e-01 -1.34889856e-01 4.90086108e-01 -1.10711455e+00
-1.02741361e+00 4.59549546e-01 1.45011768e-01 -1.04312277e+00
-1.48952439e-01 -2.81689376e-01 -5.12388125e-02 -4.79780257e-01
3.16738486e-01 7.02440366e-02 1.35970619e-02 -5.35582662e-01
4.88224961e-02 5.34680426e-01 7.34804094e-01 -3.63255441e-01
1.05309558e+00 6.33172274e-01 7.66465887e-02 -4.36282963e-01
-4.87794489e-01 -6.87479138e-01 -5.84246218e-01 -5.09955525e-01
9.02130425e-01 -1.00334513e+00 -1.08310819e+00 6.01562083e-01
-1.50098348e+00 3.93457174e-01 1.73595488e-01 7.16077745e-01
-7.92060643e-02 5.87354004e-01 -6.30045474e-01 -1.22838128e+00
-3.27669621e-01 -9.12918210e-01 9.14262056e-01 6.68376625e-01
2.45140344e-01 -7.64626145e-01 9.25091505e-02 -2.13330403e-01
4.20732975e-01 2.57766515e-01 -2.43016616e-01 -3.35046381e-01
-1.01640117e+00 -1.49928778e-01 -1.42284140e-01 -1.00044303e-01
2.47334495e-01 5.07607758e-01 -7.73260593e-01 -5.42502642e-01
-3.34979743e-02 1.61609426e-01 8.45750570e-01 6.88155890e-01
8.26737761e-01 -7.00687524e-03 -1.04908872e+00 4.53916669e-01
1.42183852e+00 5.35676122e-01 3.69941354e-01 4.99842972e-01
6.97859526e-01 1.95086643e-01 1.05446708e+00 1.34866804e-01
-3.21993530e-02 9.05664086e-01 2.15345368e-01 -5.30925877e-02
-2.70552069e-01 -4.91476804e-02 3.08242589e-01 6.28191769e-01
-1.54134199e-01 -2.52710164e-01 -4.96751785e-01 5.72212338e-01
-2.31493855e+00 -1.35488522e+00 -6.25685632e-01 2.28348327e+00
6.20700598e-01 3.66135567e-01 1.62334368e-01 -3.58430535e-01
1.22876513e+00 -1.65848151e-01 -6.80880189e-01 3.45132440e-01
-2.54228354e-01 -5.36668479e-01 6.78802609e-01 4.33967918e-01
-1.30221820e+00 8.89735043e-01 6.98218966e+00 6.34767115e-01
-8.05879474e-01 3.41753542e-01 -1.12333044e-01 -1.28697932e-01
2.88389504e-01 2.32197586e-02 -1.42535460e+00 8.14431190e-01
3.82964164e-01 9.85108465e-02 -4.85957712e-02 6.92807674e-01
3.30795169e-01 -2.53441602e-01 -8.32997978e-01 1.07629919e+00
4.77705561e-02 -1.05398035e+00 -1.85460836e-01 -3.99825394e-01
6.99493945e-01 -3.33527684e-01 1.98050626e-02 1.32250130e-01
2.99030334e-01 -3.41026723e-01 6.39895737e-01 9.05552983e-01
3.97922873e-01 -3.21327299e-01 6.06922150e-01 2.67163098e-01
-1.79380000e+00 9.46153998e-02 -4.48882788e-01 1.13740906e-01
4.79955792e-01 3.50263000e-01 -3.18656266e-01 7.15734184e-01
7.57993639e-01 7.97523320e-01 -5.48271358e-01 1.72663224e+00
1.17928512e-01 -6.13935143e-02 -4.50498223e-01 4.36849259e-02
-8.23414996e-02 -3.65654156e-02 9.54417586e-01 9.29068923e-01
2.99526364e-01 -2.64361724e-02 6.63156688e-01 6.27361536e-01
3.46120298e-01 -1.92606896e-01 -6.89143717e-01 4.82306689e-01
7.10363090e-01 1.23554897e+00 -7.38702774e-01 -6.52785718e-01
-4.36669171e-01 1.06619334e+00 -1.24239713e-01 3.92179489e-01
-1.27019048e+00 -2.19007865e-01 6.26311362e-01 -8.70830715e-02
2.95146585e-01 -2.84026444e-01 6.51740730e-02 -1.15737963e+00
1.89571559e-01 -4.32023525e-01 6.02029324e-01 -9.03305590e-01
-1.41491699e+00 4.67311233e-01 -1.95227817e-01 -1.76584971e+00
1.94029257e-01 -3.37797552e-01 -7.00375915e-01 7.10491836e-01
-1.21138144e+00 -1.33938336e+00 -3.14541072e-01 5.77485085e-01
7.63264060e-01 -1.21695027e-01 1.79549739e-01 8.53439748e-01
-7.56385028e-01 5.75495481e-01 2.12889731e-01 1.86126446e-03
1.14998722e+00 -8.73560488e-01 9.36358869e-02 1.11427653e+00
-1.18453816e-01 7.27043092e-01 9.96010005e-01 -1.23357427e+00
-1.20460415e+00 -1.11052692e+00 5.70529461e-01 -4.81778949e-01
5.00713408e-01 7.64385015e-02 -1.13881719e+00 8.35769713e-01
1.68393597e-01 2.55014241e-01 1.90994162e-02 -6.78258836e-02
1.55472934e-01 -1.86804995e-01 -1.16760063e+00 5.43177426e-01
1.21288037e+00 -4.74011078e-02 -7.27222741e-01 5.34101903e-01
7.13011622e-01 -6.70274615e-01 -5.59210241e-01 4.89117831e-01
6.88375175e-01 -3.89782399e-01 1.00243402e+00 -5.64603567e-01
-6.27154887e-01 -1.20744121e+00 2.60287344e-01 -7.92712092e-01
-7.07057834e-01 -6.10688746e-01 -3.05805326e-01 1.32580996e+00
1.04547076e-01 -3.52565616e-01 7.69446194e-01 4.60478127e-01
-1.26174930e-02 1.27939635e-03 -9.89008248e-01 -1.24937320e+00
-5.00297070e-01 1.35480434e-01 4.40740436e-01 9.64994848e-01
-7.14508176e-01 3.55925784e-02 -8.20851445e-01 5.64682424e-01
8.57290149e-01 9.12626460e-02 8.18333983e-01 -1.30853856e+00
1.61412191e-02 -2.65574515e-01 -4.58573192e-01 -1.13283443e+00
-8.28129500e-02 -2.23728448e-01 1.98561803e-01 -1.13041306e+00
2.73753673e-01 -4.98928517e-01 -4.87878144e-01 2.59137481e-01
-3.76167029e-01 7.40967318e-02 1.66229621e-01 6.18482828e-01
-1.05194628e+00 4.36658770e-01 1.38270497e+00 -2.57542133e-01
-1.93364620e-01 3.68733197e-01 1.52492031e-01 5.34205854e-01
5.83998084e-01 -7.23204851e-01 1.10490531e-01 -4.43672121e-01
-1.24882743e-01 1.62635669e-01 5.39945126e-01 -1.28177679e+00
8.57276082e-01 -2.90579677e-01 9.05376792e-01 -1.30510056e+00
3.13441128e-01 -1.16883385e+00 8.60780239e-01 7.93756604e-01
1.88902736e-01 6.13067091e-01 4.34486717e-01 8.26846600e-01
-9.79363620e-02 -2.65265614e-01 7.48712778e-01 -1.36320993e-01
-9.30987120e-01 5.41613638e-01 -1.83641717e-01 -4.00649905e-01
1.28393662e+00 -6.08089924e-01 -2.43120372e-01 6.18711561e-02
-9.95154977e-01 6.18488371e-01 4.95190620e-01 6.15357280e-01
5.20005405e-01 -1.90986145e+00 -5.66291869e-01 1.04541041e-01
-1.89681575e-01 -2.84778267e-01 4.30440575e-01 1.03183734e+00
-2.49116525e-01 7.19364211e-02 -3.38298619e-01 -8.89650881e-01
-1.74684131e+00 9.06962276e-01 3.90069366e-01 1.48240566e-01
-6.08137071e-01 6.49498463e-01 -1.16267890e-01 9.17070732e-02
3.16422105e-01 -1.30889177e-01 -2.39225611e-01 -1.65272325e-01
7.54934192e-01 4.51760441e-01 -4.23076153e-01 -5.72463572e-01
-6.56708300e-01 1.06977010e+00 -2.44567975e-01 -3.18062305e-02
6.79326296e-01 -4.28391516e-01 7.15963319e-02 5.60039043e-01
7.31764078e-01 -9.18226615e-02 -1.47035730e+00 -2.73910761e-01
-9.01639313e-02 -8.80405962e-01 -2.34868690e-01 -4.92475837e-01
-9.75142002e-01 3.67216408e-01 1.21141362e+00 2.00344354e-01
8.13581765e-01 -2.71403432e-01 7.39220560e-01 2.56416291e-01
3.06870252e-01 -1.38286102e+00 9.25110206e-02 3.86783391e-01
7.06691265e-01 -1.65502894e+00 1.74259648e-01 -2.25846827e-01
-3.33656043e-01 9.43062186e-01 1.13663805e+00 -8.52304175e-02
5.69376111e-01 3.37186068e-01 2.31519923e-01 2.32160967e-02
-6.54969990e-01 -3.43097597e-01 2.37907946e-01 6.27795935e-01
1.10533059e-01 -3.86634767e-01 -4.55806434e-01 1.15489438e-02
6.19167984e-01 -4.94500448e-04 1.22707851e-01 1.14921522e+00
-7.22429633e-01 -1.00798285e+00 -1.01866746e+00 2.04438657e-01
-4.82918918e-01 4.09564823e-01 -5.39338291e-02 9.43182349e-01
4.36579883e-01 8.39131236e-01 1.50178388e-01 -1.91980988e-01
3.92081559e-01 -1.65997759e-01 5.59529185e-01 -8.46450254e-02
-5.31364739e-01 2.96596169e-01 -2.40316182e-01 -4.87502724e-01
-8.54384422e-01 -1.01339781e+00 -1.23020148e+00 -4.12380964e-01
-9.01741922e-01 -1.34785429e-01 3.64269614e-01 7.30925322e-01
1.83200777e-01 6.23380423e-01 3.79152507e-01 -6.20411277e-01
-5.03423929e-01 -9.48555827e-01 -3.83487254e-01 5.09858251e-01
6.21199012e-01 -1.07579446e+00 -2.96880543e-01 2.18774542e-01] | [6.5064005851745605, -2.0052082538604736] |
c4d08fc7-9224-4721-abd6-69b12e4aaa88 | accelerated-parallel-mri-using-memory | 2304.01351 | null | https://arxiv.org/abs/2304.01351v1 | https://arxiv.org/pdf/2304.01351v1.pdf | Accelerated parallel MRI using memory efficient and robust monotone operator learning (MOL) | Model-based deep learning methods that combine imaging physics with learned regularization priors have been emerging as powerful tools for parallel MRI acceleration. The main focus of this paper is to determine the utility of the monotone operator learning (MOL) framework in the parallel MRI setting. The MOL algorithm alternates between a gradient descent step using a monotone convolutional neural network (CNN) and a conjugate gradient algorithm to encourage data consistency. The benefits of this approach include similar guarantees as compressive sensing algorithms including uniqueness, convergence, and stability, while being significantly more memory efficient than unrolled methods. We validate the proposed scheme by comparing it with different unrolled algorithms in the context of accelerated parallel MRI for static and dynamic settings. | ['Mathews Jacob', 'Aniket Pramanik'] | 2023-04-03 | null | null | null | null | ['compressive-sensing'] | ['computer-vision'] | [ 5.31027019e-01 1.02572694e-01 -1.19703086e-02 -3.50759029e-01
-7.35327184e-01 6.28883541e-02 3.79449278e-01 1.00329787e-01
-8.41799200e-01 6.59772277e-01 4.30955201e-01 -3.83274764e-01
-4.99013364e-01 -2.03790754e-01 -8.31401348e-01 -8.76185775e-01
-6.07213914e-01 2.95959264e-01 1.01751097e-01 3.84408352e-03
1.45340860e-01 5.99450290e-01 -6.36164725e-01 2.13273302e-01
5.70849180e-01 1.01552343e+00 2.17285126e-01 5.82174718e-01
3.49239767e-01 1.01271224e+00 7.21040741e-02 1.93404645e-01
6.36251807e-01 -2.87550211e-01 -8.01654577e-01 6.95147291e-02
2.68215865e-01 -4.27717149e-01 -4.76833791e-01 8.29633117e-01
8.86724710e-01 2.78623432e-01 3.34907144e-01 -7.37429976e-01
-1.79882959e-01 4.68611598e-01 -7.77669191e-01 7.09536314e-01
1.88444424e-02 -9.55434889e-02 5.05251944e-01 -1.06279898e+00
5.34208298e-01 6.18208766e-01 1.10233307e+00 3.47550511e-01
-1.23373115e+00 -4.27824467e-01 -3.12629580e-01 1.18677318e-01
-1.12805259e+00 -5.11515081e-01 9.21139777e-01 -3.20847571e-01
6.83764577e-01 2.08163992e-01 6.28985882e-01 8.61044466e-01
4.73821819e-01 7.45980442e-01 1.26576984e+00 -5.60447395e-01
5.95102131e-01 -1.49996698e-01 1.13058224e-01 6.73588753e-01
1.36142701e-01 2.55437255e-01 -5.00985980e-01 -5.23418367e-01
9.20703232e-01 4.22147335e-03 -5.67109942e-01 -7.97239244e-01
-1.21970618e+00 1.11950386e+00 6.23073161e-01 2.90663630e-01
-7.69613981e-01 2.11652905e-01 5.83036184e-01 2.30235737e-02
5.36578953e-01 4.30778563e-01 -4.78018448e-02 2.11134300e-01
-1.40413260e+00 2.86477149e-01 6.19039953e-01 4.79049534e-01
2.06553489e-01 2.43747041e-01 -1.11142159e-01 6.60584748e-01
2.88010180e-01 2.50560343e-01 6.73377335e-01 -9.57971752e-01
-2.30140542e-03 -2.26833418e-01 -1.58896118e-01 -1.06214225e+00
-7.78585136e-01 -9.11701620e-01 -1.08772457e+00 2.73278743e-01
6.19257949e-02 -3.18568766e-01 -6.17681742e-01 1.59791493e+00
4.85104918e-01 5.92406631e-01 -1.55530736e-01 1.11869919e+00
4.23981845e-01 3.52859795e-01 -1.25059247e-01 -4.78636950e-01
8.53496492e-01 -7.90631473e-01 -6.84683979e-01 2.75673643e-02
6.34139717e-01 -7.76293933e-01 8.39212120e-01 4.09786075e-01
-1.32327139e+00 -2.38766372e-01 -1.09777057e+00 2.36054763e-01
3.25676352e-01 -1.14180326e-01 6.80234432e-01 6.11051977e-01
-1.25699818e+00 7.36139178e-01 -1.23852110e+00 8.80041569e-02
4.42983955e-01 4.21964020e-01 -1.23831049e-01 -1.63492426e-01
-9.97026801e-01 7.97500312e-01 1.90711930e-01 2.06206933e-01
-7.56518662e-01 -1.20423424e+00 -7.68233299e-01 -2.51723230e-01
1.36331916e-01 -5.49406171e-01 1.10604644e+00 -7.82211959e-01
-1.56210673e+00 7.99374521e-01 4.73740548e-02 -9.12971973e-01
9.28599894e-01 -2.72334129e-01 -1.75443411e-01 4.64674562e-01
-4.73565906e-02 4.40412790e-01 1.02639818e+00 -9.95790541e-01
-4.10764366e-02 -2.40282804e-01 -3.78662050e-01 1.82714805e-01
-9.94649678e-02 -7.38499388e-02 -4.94257659e-02 -5.53634644e-01
4.66918677e-01 -1.16622448e+00 -5.93451321e-01 1.61190540e-01
-9.66443643e-02 4.91422445e-01 6.23065114e-01 -8.28809798e-01
8.18177640e-01 -2.03005219e+00 -5.58636747e-02 5.28700113e-01
3.08037549e-01 2.00938750e-02 7.97936842e-02 1.03269465e-01
-4.77668911e-01 -6.98516905e-01 -7.13226974e-01 -4.55128029e-02
-6.40722632e-01 1.40745312e-01 -8.26794282e-02 9.61782575e-01
3.34232375e-02 7.89107621e-01 -9.58686233e-01 -4.88583028e-01
2.76168972e-01 5.80984116e-01 -7.47214794e-01 -8.20304006e-02
3.35504055e-01 1.09422600e+00 -1.32567197e-01 2.04046443e-01
8.46822023e-01 -3.38390678e-01 3.07515174e-01 -5.07634401e-01
-1.57525942e-01 -1.24111585e-02 -1.25449979e+00 1.86901939e+00
-5.76359868e-01 4.44133699e-01 5.96212447e-01 -1.43131065e+00
4.80559438e-01 3.59395683e-01 1.26277280e+00 -8.51889312e-01
9.24867392e-02 3.43823612e-01 6.67546466e-02 -6.51539385e-01
1.86821610e-01 -5.39403260e-01 4.86859262e-01 3.92674536e-01
-9.46766734e-02 -6.35329727e-03 -1.74701184e-01 3.15913916e-01
1.19271576e+00 -8.23749229e-02 7.72461370e-02 -8.79928052e-01
4.32302922e-01 -6.08473420e-02 5.56554139e-01 8.88766885e-01
-3.40574265e-01 6.61440372e-01 1.00521304e-01 -5.51070392e-01
-1.11333203e+00 -1.03505170e+00 -5.58478177e-01 5.75706959e-01
-1.14931889e-01 5.04463166e-02 -4.53256488e-01 -3.81507009e-01
-3.00034255e-01 3.62493515e-01 -6.33802414e-01 4.88587432e-02
-1.05751514e+00 -1.21221960e+00 2.48604089e-01 5.15255630e-01
4.40756887e-01 -7.12820411e-01 -1.00836825e+00 3.67889851e-01
-2.62728810e-01 -1.16498649e+00 -3.76053810e-01 2.46286601e-01
-1.13791120e+00 -1.01417387e+00 -9.97739732e-01 -5.45335114e-01
6.50427580e-01 1.26028866e-01 9.27845299e-01 -1.39723718e-01
-3.37392926e-01 5.94790399e-01 -8.24003816e-02 1.22886850e-02
-2.87940979e-01 -1.83902740e-01 1.92223325e-01 1.31031722e-01
-4.39128906e-01 -7.84126997e-01 -9.84332144e-01 -1.75478399e-01
-1.19178712e+00 9.15561318e-02 7.35157073e-01 1.00208569e+00
7.55758584e-01 -2.13952765e-01 2.64808446e-01 -1.10958540e+00
6.56921506e-01 -6.31580830e-01 -4.55107331e-01 -6.59280568e-02
-7.41567850e-01 1.21101126e-01 3.42780530e-01 -3.43330830e-01
-8.74476314e-01 4.40934360e-01 -3.57653648e-01 -3.20016950e-01
3.48225385e-01 7.20868468e-01 6.95153296e-01 -6.18188083e-01
8.45584512e-01 3.60516936e-01 2.95692503e-01 -2.45519385e-01
2.19525024e-01 1.60480052e-01 7.00008690e-01 -6.96270227e-01
4.05935168e-01 1.02881253e+00 4.15642232e-01 -1.08332145e+00
-7.31331229e-01 -7.16580868e-01 -6.00964189e-01 -3.45778614e-01
7.61099398e-01 -7.08663583e-01 -4.51798081e-01 2.24696040e-01
-8.44006002e-01 -3.94020677e-01 -3.98762226e-01 1.02639747e+00
-7.43448138e-01 7.73295820e-01 -1.05357802e+00 -5.34920931e-01
-4.70916867e-01 -1.22368407e+00 6.15172803e-01 -1.46703631e-01
-8.00015777e-02 -1.17990696e+00 2.81376719e-01 1.87063903e-01
8.25879574e-01 4.91297126e-01 7.40304232e-01 -4.13041949e-01
-1.48058444e-01 -1.34199217e-01 -1.41121417e-01 2.82364368e-01
-8.77701938e-02 -6.96028590e-01 -8.49351585e-01 -6.86417043e-01
7.73563623e-01 -2.57707477e-01 7.25488365e-01 9.73763406e-01
1.18209016e+00 -2.31317073e-01 2.53972635e-02 8.80288303e-01
1.49299121e+00 -1.87878892e-01 4.33101714e-01 2.51347333e-01
4.39355075e-01 3.82188022e-01 -1.08187996e-01 5.08824587e-01
-1.03334174e-01 4.56326574e-01 3.49026203e-01 -3.61006945e-01
-2.36525133e-01 2.03630880e-01 -8.54957476e-02 1.11513257e+00
2.22814381e-01 6.63970292e-01 -1.12004709e+00 5.53425014e-01
-1.78030348e+00 -6.15526974e-01 -2.52030879e-01 2.15380073e+00
7.30469286e-01 5.45885675e-02 1.56030372e-01 1.82338178e-01
5.34913599e-01 3.93017456e-02 -4.42779034e-01 -1.68746755e-01
-1.11456560e-02 5.36661029e-01 8.20091069e-01 7.04555571e-01
-1.05200839e+00 8.01936463e-02 7.18865728e+00 5.14832139e-01
-1.43427515e+00 6.06324136e-01 6.66186631e-01 -1.15915716e-01
-1.48815243e-02 -1.21009856e-01 -1.19235083e-01 2.08674058e-01
7.82253087e-01 3.20629537e-01 4.27343547e-01 5.70545375e-01
4.67266053e-01 -1.88584432e-01 -9.51435924e-01 1.01385474e+00
1.49414465e-01 -1.69661248e+00 -3.25370252e-01 -8.96932706e-02
8.52930188e-01 4.28574860e-01 1.61869854e-01 -9.88892466e-02
-1.25343591e-01 -8.63006592e-01 5.06742597e-01 3.07106823e-01
4.59346414e-01 -4.63483334e-01 7.16444731e-01 1.75190866e-01
-7.56884575e-01 2.21271347e-02 -1.17447712e-01 -4.11801785e-02
3.68465990e-01 1.02742994e+00 -7.02979624e-01 6.03605151e-01
7.02587128e-01 3.66913229e-01 -9.51679870e-02 1.17311859e+00
2.26053089e-01 7.48665452e-01 -4.79966223e-01 5.45310438e-01
5.49358428e-01 -1.75065815e-01 6.88918531e-01 1.32653809e+00
1.20894685e-02 6.83664531e-02 3.22743207e-01 7.64230251e-01
2.63750881e-01 1.88519016e-01 -4.13629442e-01 4.98374224e-01
-1.21035978e-01 1.15466499e+00 -9.32808101e-01 -2.10422814e-01
-5.93078792e-01 8.64963174e-01 2.20147282e-01 2.58994132e-01
-6.25749588e-01 1.50825471e-01 -1.02827989e-01 3.94468933e-01
2.25827396e-01 -5.06057203e-01 -4.40614492e-01 -9.59695578e-01
-2.15685163e-02 -6.69473767e-01 5.92462420e-01 -5.63013375e-01
-1.32256639e+00 4.25466061e-01 2.29228616e-01 -1.07669365e+00
-2.18262207e-02 -5.57836354e-01 -6.86722100e-01 6.14854157e-01
-1.66294420e+00 -8.01182985e-01 -2.77291715e-01 7.88709939e-01
2.95471311e-01 2.18974471e-01 5.01729250e-01 5.07795215e-01
-2.35683575e-01 2.37969801e-01 2.96372563e-01 6.21766560e-02
3.00862938e-01 -1.09550321e+00 -3.75117570e-01 9.01122272e-01
-5.79067580e-02 5.15446782e-01 7.51854420e-01 -5.61698616e-01
-1.45660615e+00 -8.82399738e-01 3.31096202e-01 2.92701870e-01
7.56791294e-01 -5.70771657e-02 -8.04669201e-01 5.75150847e-01
1.95651963e-01 5.19752860e-01 6.25246108e-01 -1.58247530e-01
1.00837745e-01 -8.18522926e-03 -1.25865304e+00 3.52351367e-01
5.60780525e-01 -3.71788085e-01 -3.53649497e-01 7.73095369e-01
2.38720611e-01 -8.04147601e-01 -8.72093499e-01 5.26221633e-01
3.66598248e-01 -9.32342649e-01 1.16394520e+00 -2.39433944e-01
4.01708782e-01 -4.07135114e-02 -1.40177429e-01 -1.11443508e+00
-3.94755960e-01 -8.17914486e-01 -2.10954055e-01 2.01767504e-01
1.37506798e-01 -6.17979705e-01 9.62102056e-01 6.03598058e-01
-2.98698366e-01 -8.97709429e-01 -1.25406122e+00 -7.25611389e-01
2.95876384e-01 -5.97746432e-01 -2.31094286e-01 1.15385878e+00
5.31918369e-02 -3.76557792e-03 -5.93269289e-01 3.37523371e-01
1.00100017e+00 -6.10390790e-02 4.47571054e-02 -6.56419158e-01
-7.31631398e-01 -2.31400996e-01 -2.97542989e-01 -8.59257638e-01
-9.27979499e-02 -1.16647041e+00 -4.14802320e-02 -1.09594095e+00
3.06791097e-01 -5.49403012e-01 -6.98590636e-01 1.50984675e-01
4.60035875e-02 4.41903681e-01 -2.25926656e-02 3.99239659e-01
-3.08664620e-01 4.28693742e-01 1.10158908e+00 -6.05622074e-03
-4.56526759e-04 7.38132820e-02 -2.61198610e-01 6.62907124e-01
7.22588181e-01 -5.11290014e-01 -4.20857310e-01 -5.37137508e-01
1.17871076e-01 4.43322808e-01 4.27496314e-01 -1.20775700e+00
4.39183772e-01 3.25081021e-01 3.69411111e-01 -2.30625257e-01
1.50883511e-01 -8.20617855e-01 -5.18811680e-03 1.13594282e+00
-5.12875676e-01 2.82170951e-01 1.78527340e-01 5.66862285e-01
-9.87770036e-02 -3.33585650e-01 1.13884926e+00 -2.90638655e-01
-3.28790098e-01 2.90442795e-01 -3.34401786e-01 6.98425248e-02
6.24792993e-01 -1.13880038e-01 6.06851041e-01 -3.72715235e-01
-1.02162349e+00 -2.03393832e-01 -1.42007042e-02 -1.77019387e-01
8.23106110e-01 -1.10859609e+00 -9.29974496e-01 3.69567394e-01
-2.57758349e-01 -2.10243002e-01 4.25835073e-01 1.57124281e+00
-8.38948786e-01 3.07910562e-01 -2.09558278e-01 -1.00369668e+00
-8.01131248e-01 4.59764421e-01 5.83133459e-01 -5.15919328e-01
-1.21099329e+00 7.89721072e-01 -1.66833028e-02 -5.46362221e-01
1.22487478e-01 -1.22297354e-01 3.00499231e-01 -6.57665491e-01
5.10815442e-01 4.62227017e-01 5.17018974e-01 -3.59609336e-01
-4.28966492e-01 2.54732400e-01 -1.23313136e-01 -4.20599431e-01
1.49152648e+00 -2.26160828e-02 -1.19482666e-01 1.84980765e-01
1.29828882e+00 9.88312736e-02 -1.28502059e+00 -4.73459780e-01
1.77469462e-01 -2.27204338e-01 6.59512460e-01 -4.59581256e-01
-1.29179800e+00 5.54604709e-01 1.08384860e+00 -7.87178203e-02
1.02331913e+00 -3.14203799e-01 9.68239665e-01 1.97296411e-01
1.49109021e-01 -8.78975987e-01 1.25629529e-01 1.56762689e-01
9.50181127e-01 -1.18593729e+00 3.40705603e-01 -3.02724421e-01
-4.09971476e-01 1.08824837e+00 6.14140816e-02 -5.46270549e-01
1.01449275e+00 5.48076332e-01 5.95187210e-02 -3.69282067e-01
-1.27301276e-01 3.07411253e-01 2.05171496e-01 4.64768112e-01
6.09412432e-01 -1.32276252e-01 -6.53149664e-01 -1.06596678e-01
1.66986749e-01 2.32819200e-01 4.11829382e-01 1.32771659e+00
-9.91491601e-04 -7.90074050e-01 -2.58574069e-01 4.95667309e-01
-5.75803161e-01 -1.27559066e-01 4.12153125e-01 6.52793407e-01
-1.73593178e-01 6.71911180e-01 -2.85213023e-01 2.02236310e-01
-3.69481966e-02 -1.67882696e-01 8.87653351e-01 -2.77192503e-01
-5.32227218e-01 1.28329709e-01 -3.26133639e-01 -8.11144352e-01
-7.12799609e-01 -7.73601174e-01 -1.25444329e+00 -5.00348099e-02
-1.70844719e-01 1.20158456e-01 8.04576278e-01 1.06495118e+00
2.00410232e-01 4.71068025e-01 6.31237507e-01 -8.82001758e-01
-8.17609668e-01 -6.67175055e-01 -4.47985888e-01 5.99492490e-01
2.68612504e-01 -4.81966615e-01 -1.67990297e-01 -2.07602665e-01] | [13.42078971862793, -2.392184019088745] |
071463ae-639f-4b3e-b4a9-f52da0135ec6 | self-supervised-visual-place-recognition-by | 2208.09315 | null | https://arxiv.org/abs/2208.09315v1 | https://arxiv.org/pdf/2208.09315v1.pdf | Self-Supervised Visual Place Recognition by Mining Temporal and Feature Neighborhoods | Visual place recognition (VPR) using deep networks has achieved state-of-the-art performance. However, most of them require a training set with ground truth sensor poses to obtain positive and negative samples of each observation's spatial neighborhood for supervised learning. When such information is unavailable, temporal neighborhoods from a sequentially collected data stream could be exploited for self-supervised training, although we find its performance suboptimal. Inspired by noisy label learning, we propose a novel self-supervised framework named \textit{TF-VPR} that uses temporal neighborhoods and learnable feature neighborhoods to discover unknown spatial neighborhoods. Our method follows an iterative training paradigm which alternates between: (1) representation learning with data augmentation, (2) positive set expansion to include the current feature space neighbors, and (3) positive set contraction via geometric verification. We conduct comprehensive experiments on both simulated and real datasets, with either RGB images or point clouds as inputs. The results show that our method outperforms our baselines in recall rate, robustness, and heading diversity, a novel metric we propose for VPR. Our code and datasets can be found at https://ai4ce.github.io/TF-VPR/. | ['Chen Feng', 'Ruoyu Wang', 'Li Ding', 'Yiming Li', 'Xuchu Xu', 'Xinhao Liu', 'Chao Chen'] | 2022-08-19 | null | null | null | null | ['visual-place-recognition'] | ['computer-vision'] | [ 8.86517018e-02 -2.21834555e-01 -4.72669899e-01 -5.32182813e-01
-7.24223077e-01 -7.08278656e-01 5.68738163e-01 1.71021059e-01
-5.90950429e-01 6.95905685e-01 3.21468152e-02 -1.86821520e-01
-7.19946623e-02 -9.37718809e-01 -1.02482140e+00 -7.17394769e-01
-2.75993407e-01 3.03230494e-01 1.63017422e-01 -6.95405453e-02
2.11085483e-01 6.89274728e-01 -1.79154944e+00 -1.28209248e-01
5.87225616e-01 1.35563898e+00 2.30950013e-01 2.64048964e-01
1.57038108e-01 6.80045485e-01 -2.26689488e-01 4.88288343e-01
5.85165322e-01 9.01209339e-02 -5.61316967e-01 1.11200906e-01
1.87135801e-01 -2.31501967e-01 -5.97546995e-01 9.77469921e-01
4.26650137e-01 4.84152883e-01 5.24195015e-01 -1.47144091e+00
-8.34507644e-01 7.66757652e-02 -6.56799555e-01 -6.20897561e-02
2.78561592e-01 4.07104939e-01 7.12352455e-01 -9.66461420e-01
6.23685837e-01 8.42668056e-01 6.48968220e-01 5.02443492e-01
-1.17002201e+00 -9.49561954e-01 2.73721606e-01 1.55404866e-01
-1.94858420e+00 -4.63413566e-01 9.87672627e-01 -1.91732213e-01
8.67831647e-01 6.30017668e-02 7.33659983e-01 1.02986455e+00
-4.98063773e-01 8.67209911e-01 9.08305883e-01 -1.04201727e-01
6.58216834e-01 -1.24542952e-01 -1.36102498e-01 6.61213994e-01
2.56253127e-02 4.39938813e-01 -7.57139146e-01 -9.19402093e-02
9.66517687e-01 3.16267014e-01 -3.35498631e-01 -6.22403204e-01
-1.34182703e+00 7.25339949e-01 1.16889668e+00 3.80226076e-02
-3.14729512e-01 2.90452391e-01 1.92943007e-01 1.70972005e-01
3.17652881e-01 2.25315720e-01 -4.50759649e-01 8.04324895e-02
-6.91361129e-01 1.20807961e-01 3.85007530e-01 1.20085073e+00
1.21382570e+00 -1.66972578e-02 2.04424873e-01 8.44642937e-01
5.78419209e-01 9.13086176e-01 5.97525775e-01 -1.05378187e+00
4.48133081e-01 6.90747738e-01 2.37885490e-01 -1.21528387e+00
-4.25198376e-01 -1.75231069e-01 -9.75403309e-01 2.07247064e-01
1.31698430e-01 -1.06661297e-01 -1.33115208e+00 1.61865103e+00
3.87407273e-01 4.78237003e-01 2.44202152e-01 1.22721195e+00
1.09466147e+00 7.16409266e-01 -2.54797012e-01 4.73887101e-02
6.53051019e-01 -8.98857892e-01 -4.25251275e-01 -4.02917534e-01
5.59921384e-01 -8.47239196e-02 8.05509388e-01 1.30093828e-01
-4.51604038e-01 -5.12117207e-01 -1.24051309e+00 2.91177625e-04
-6.13133013e-01 3.00115347e-01 5.67283034e-01 1.82811379e-01
-1.23101091e+00 5.24621069e-01 -9.18911338e-01 -4.60676670e-01
7.40516067e-01 4.41506326e-01 -6.20365798e-01 -1.78568184e-01
-8.55436027e-01 2.94361740e-01 3.87161732e-01 2.37537310e-01
-1.16723168e+00 -3.33478004e-01 -1.26783681e+00 -3.33075583e-01
3.35094124e-01 -2.55454868e-01 9.83254194e-01 -7.05062687e-01
-1.14439201e+00 7.58414388e-01 -4.33862001e-01 -5.21550357e-01
4.15466487e-01 -9.56239924e-02 -2.23141104e-01 -4.29120511e-02
3.48322213e-01 1.04048884e+00 6.23023272e-01 -1.65369093e+00
-6.96799338e-01 -5.57205319e-01 -8.54656175e-02 3.28020781e-01
-2.75166705e-02 -6.36733115e-01 -5.18443465e-01 -4.83013630e-01
8.37643325e-01 -1.00445950e+00 -4.49346900e-01 4.30053025e-01
-4.01766330e-01 -1.59162119e-01 7.89810896e-01 -2.24150226e-01
7.35671580e-01 -2.37025952e+00 -3.41238976e-01 5.88056862e-01
1.85850054e-01 -8.43594433e-04 -2.43432194e-01 1.49658978e-01
1.24899475e-02 9.40615535e-02 -2.70773500e-01 -4.42016155e-01
-1.67061940e-01 4.20521110e-01 -6.01867378e-01 8.95608246e-01
1.16633274e-01 8.67267787e-01 -1.22501588e+00 -3.01446140e-01
5.34269810e-01 4.36854571e-01 -1.93745837e-01 -3.51374364e-03
-2.06414118e-01 5.90677381e-01 -4.58289742e-01 1.01363957e+00
6.11582339e-01 -4.74778980e-01 -1.20485276e-01 -1.53618455e-02
-2.75576472e-01 2.45028213e-01 -1.39542747e+00 1.83283508e+00
-1.36486545e-01 7.30849564e-01 -2.50299335e-01 -7.97586083e-01
1.33751404e+00 -7.99114704e-02 4.63054001e-01 -1.01299500e+00
7.51567772e-04 2.90566444e-01 -5.38856387e-01 -2.66626328e-01
5.88500917e-01 3.58485639e-01 5.72881103e-02 2.21651182e-01
3.61198075e-02 2.66159624e-01 -1.89904958e-01 -4.74660769e-02
1.12616730e+00 2.84564704e-01 2.81223804e-01 9.79164392e-02
2.89970428e-01 2.52815127e-01 9.12798047e-01 8.45264673e-01
-4.85937387e-01 9.46534216e-01 -1.04209393e-01 -6.51728153e-01
-7.80428112e-01 -1.11586058e+00 -1.39623269e-01 6.46280766e-01
7.79329598e-01 -1.24610744e-01 -8.18304121e-02 -6.65591598e-01
1.52884662e-01 4.23210531e-01 -7.86735058e-01 -2.12952252e-02
-4.72869396e-01 -4.47019696e-01 4.90049422e-01 7.97966719e-01
8.32349956e-01 -1.18516731e+00 -7.23279715e-01 -1.56094283e-01
-2.78320581e-01 -8.69327605e-01 -7.99700320e-02 4.45806742e-01
-8.27243090e-01 -1.07322907e+00 -5.20472467e-01 -8.59281242e-01
1.01511073e+00 6.61292434e-01 4.94321853e-01 9.39554349e-02
4.01777402e-03 3.88154745e-01 -5.46841860e-01 -3.23964149e-01
2.74283528e-01 -4.25979607e-02 4.07619327e-01 -6.26531709e-03
4.19178307e-01 -6.80936217e-01 -7.14791298e-01 3.78206700e-01
-6.86746061e-01 -1.63380042e-01 4.54740882e-01 7.09088683e-01
1.15234792e+00 -2.20338598e-01 3.16369146e-01 -2.69454479e-01
5.26351072e-02 -6.36225402e-01 -7.54997075e-01 6.41334131e-02
-4.07956034e-01 -4.48110774e-02 4.60928440e-01 -4.48918134e-01
-5.25927365e-01 5.34289300e-01 7.67298192e-02 -9.98154879e-01
-5.49960852e-01 3.22207361e-01 -2.93239336e-02 -3.53537172e-01
8.76029491e-01 6.00485742e-01 -1.25850588e-01 -2.65484720e-01
2.53896028e-01 6.42404139e-01 6.54823184e-01 -3.74464393e-01
1.02447534e+00 9.38610554e-01 -1.51229426e-01 -7.77123928e-01
-5.59778750e-01 -7.90311873e-01 -5.83159626e-01 -3.28844637e-01
5.61962247e-01 -1.22989643e+00 -7.21856117e-01 3.22877973e-01
-9.19781148e-01 -5.11115551e-01 -4.24009413e-01 4.45439726e-01
-5.44337094e-01 5.15369549e-02 -1.49624035e-01 -9.64130878e-01
-1.94986314e-01 -8.64092410e-01 1.26844513e+00 3.86453122e-01
1.39739454e-01 -6.74729943e-01 1.01980425e-01 5.64425327e-02
2.35934302e-01 4.62170362e-01 3.33376345e-03 -7.04190731e-01
-8.85212481e-01 -2.11852878e-01 -2.04865918e-01 1.18007071e-01
3.26896727e-01 -3.52482796e-01 -1.14558899e+00 -4.08866227e-01
-2.82234639e-01 -6.02432907e-01 8.83647263e-01 2.93670058e-01
1.33231473e+00 -3.25778484e-01 -6.41162813e-01 1.00858486e+00
1.50595748e+00 2.97034830e-01 4.55086797e-01 6.79630876e-01
8.38711560e-01 1.92332104e-01 8.65438461e-01 5.37129581e-01
6.93172336e-01 2.93271452e-01 7.86868453e-01 -1.52196303e-01
2.02693745e-01 -5.55735767e-01 4.11231406e-02 2.41956577e-01
8.89447331e-02 -1.33513600e-01 -1.24505734e+00 9.86698925e-01
-2.14738798e+00 -7.21743464e-01 2.60999739e-01 2.30775952e+00
3.34721416e-01 -5.38503192e-02 -6.46013021e-02 2.25262299e-01
6.92974508e-01 3.80043417e-01 -1.06309283e+00 4.37391013e-01
-1.99302360e-01 -2.85046846e-01 8.88840139e-01 2.58561105e-01
-1.45167649e+00 9.74640787e-01 5.12353039e+00 3.69123340e-01
-1.31603611e+00 -1.71167314e-01 6.33105636e-01 -1.82770610e-01
-1.83628686e-02 -1.54207587e-01 -6.21020377e-01 2.72580236e-01
3.38510782e-01 1.47601977e-01 6.76893055e-01 1.08370912e+00
1.02769598e-01 -1.59651458e-01 -1.03757071e+00 1.50674105e+00
1.31719857e-01 -1.52967930e+00 -1.36257336e-01 4.56805415e-02
7.08118796e-01 6.68653488e-01 1.20546676e-01 3.09935778e-01
4.72991496e-01 -8.76866877e-01 6.93946302e-01 3.58451933e-01
8.20657194e-01 -6.06054962e-01 5.31689346e-01 4.25344735e-01
-1.47019362e+00 -3.24820757e-01 -3.30142856e-01 -1.75021924e-02
-1.03882775e-01 2.64084250e-01 -8.61004829e-01 5.38710356e-01
1.12686753e+00 1.24643338e+00 -5.91391206e-01 1.24770319e+00
-3.20847154e-01 3.74325126e-01 -7.83757508e-01 -4.33610752e-02
2.66827792e-01 -9.10341218e-02 5.90056419e-01 6.37511015e-01
2.70813674e-01 2.12367848e-01 3.83750647e-01 9.07103896e-01
-1.66231230e-01 -2.17034772e-01 -1.05812895e+00 3.42249840e-01
1.05040932e+00 1.08180439e+00 -8.44771802e-01 -1.44846842e-01
-1.53533041e-01 7.70049393e-01 4.98211861e-01 5.60174763e-01
-6.28123701e-01 -3.49488348e-01 7.93487072e-01 -1.82769597e-01
3.87001753e-01 -3.13087732e-01 -3.83478463e-01 -1.07676554e+00
2.61086226e-01 -4.39705282e-01 2.65910149e-01 -8.13614666e-01
-1.18300891e+00 4.37303990e-01 -2.10621506e-01 -1.77504766e+00
-1.01849236e-01 -3.58186543e-01 -2.78576374e-01 4.25656646e-01
-1.62630880e+00 -1.06190991e+00 -6.44244134e-01 6.96933508e-01
3.33682537e-01 -1.21263778e-02 6.18618965e-01 1.60075709e-01
-4.13575917e-01 4.93862003e-01 1.85446277e-01 5.93662500e-01
4.77127194e-01 -1.03423262e+00 4.79689568e-01 7.49871016e-01
2.43529871e-01 4.72850919e-01 2.71580070e-01 -6.87933147e-01
-1.42642546e+00 -1.66004658e+00 4.85849530e-01 -2.51585484e-01
3.85073960e-01 -4.05948400e-01 -7.19782352e-01 9.25063908e-01
-4.08169836e-01 6.59925222e-01 4.61040556e-01 -1.33264154e-01
-3.66607040e-01 -3.43284041e-01 -1.32262957e+00 6.01014256e-01
1.35179508e+00 -4.74213868e-01 -3.15396488e-01 3.25430244e-01
9.29194391e-01 -7.23715484e-01 -5.29841542e-01 7.33325660e-01
2.54701376e-01 -6.31973386e-01 9.36576068e-01 -7.21420124e-02
6.83925003e-02 -9.21275795e-01 -5.56022584e-01 -1.14421308e+00
-1.98257849e-01 -1.54985085e-01 -1.27702981e-01 9.73095119e-01
3.06720078e-01 -7.67309785e-01 9.12285328e-01 4.67323005e-01
-4.31005880e-02 -6.87382877e-01 -1.03029871e+00 -9.51108575e-01
-3.71474355e-01 -6.65337920e-01 7.41841137e-01 1.18529677e+00
-1.83228612e-01 -1.48292882e-02 -1.33309707e-01 7.51878500e-01
7.45887280e-01 2.39083856e-01 7.80440927e-01 -1.04297888e+00
2.49655902e-01 -1.21351723e-02 -6.48225188e-01 -1.29001606e+00
1.01158448e-01 -7.75939643e-01 4.83334929e-01 -1.69083953e+00
-1.61541745e-01 -8.36134911e-01 -2.96199322e-01 1.02947378e+00
3.34848881e-01 3.60190570e-01 1.48986757e-01 4.39604402e-01
-1.06941688e+00 9.23402965e-01 7.77542055e-01 -2.70208120e-01
-6.26253843e-01 -2.19399050e-01 -2.74828881e-01 6.33626521e-01
1.04926479e+00 -3.90870184e-01 -3.73925388e-01 -5.81390619e-01
3.16577591e-03 -1.06536187e-01 5.75305879e-01 -1.54469848e+00
5.35905361e-01 -2.80863583e-01 7.69291222e-01 -8.56139123e-01
4.75597829e-01 -9.33190167e-01 -4.54969816e-02 2.74045020e-01
-3.71405602e-01 2.44443491e-02 2.67913371e-01 8.60642910e-01
-1.20533481e-01 1.83508590e-01 5.02359807e-01 -1.11877494e-01
-1.15375125e+00 6.20411634e-01 -1.52089477e-01 -1.40174717e-01
1.07232416e+00 -3.48569632e-01 -4.72683400e-01 -3.33059311e-01
-5.99493384e-01 6.39800131e-01 8.14135015e-01 7.25319207e-01
1.10269475e+00 -1.53017306e+00 -2.73200393e-01 3.73800159e-01
5.35036325e-01 7.96655893e-01 1.27035081e-01 4.10145611e-01
-4.67611521e-01 2.64314592e-01 9.95161235e-02 -1.02252710e+00
-9.02731001e-01 5.96393347e-01 3.48925233e-01 3.29522133e-01
-6.95300996e-01 7.30885863e-01 1.16587653e-04 -9.79573846e-01
6.07996523e-01 -3.38070482e-01 -1.75278142e-01 -2.46328875e-01
5.49058139e-01 2.23790497e-01 -1.49954811e-01 -8.17889512e-01
-6.40254438e-01 4.89930183e-01 2.93317139e-01 -6.95770383e-02
1.49410677e+00 -2.93180525e-01 1.17700368e-01 6.88021243e-01
1.29427326e+00 -5.79971910e-01 -1.59042227e+00 -5.38378358e-01
-1.25860050e-01 -5.82774937e-01 1.07339635e-01 -6.07869864e-01
-1.18701267e+00 4.38874453e-01 1.03260589e+00 -2.34395638e-01
1.13775599e+00 1.25395015e-01 6.02416873e-01 8.95365477e-01
9.13688362e-01 -9.93825972e-01 7.29501620e-02 6.37140930e-01
8.50711942e-01 -1.71341205e+00 -2.45180860e-01 -7.00999051e-02
-4.66765136e-01 6.92873955e-01 8.91716480e-01 -3.58973175e-01
8.15238237e-01 -8.29298608e-03 2.35700265e-01 -1.49298072e-01
-5.21963000e-01 -5.44667006e-01 1.02233090e-01 8.12879682e-01
-2.47438535e-01 4.03053053e-02 3.91613424e-01 2.46691391e-01
-1.73771009e-01 1.49514630e-01 1.66692555e-01 1.27978373e+00
-3.80790263e-01 -4.21976417e-01 -4.86958236e-01 4.21023220e-01
2.86799967e-01 1.31763384e-01 -2.21660003e-01 7.06156909e-01
1.29363418e-01 9.25434291e-01 2.54361838e-01 -7.98225105e-01
3.05545270e-01 -3.26608896e-01 8.17568004e-02 -4.26959634e-01
-9.31655020e-02 -2.85244584e-01 -2.89521992e-01 -8.41154218e-01
-4.94241893e-01 -7.19285965e-01 -1.56222749e+00 -3.11821103e-02
-2.74438411e-03 -1.28431529e-01 7.01366961e-01 8.83434474e-01
4.89871621e-01 1.74782351e-01 8.30956578e-01 -1.33696580e+00
-1.74700484e-01 -8.04037392e-01 -3.78931075e-01 1.99504390e-01
9.27910924e-01 -7.85552382e-01 -4.66090888e-01 -1.22520246e-01] | [7.532192230224609, -2.016571044921875] |
06683ca9-94cb-4a1f-97ee-e63fe6f46379 | sequence-level-knowledge-distillation-for | 1811.04531 | null | http://arxiv.org/abs/1811.04531v1 | http://arxiv.org/pdf/1811.04531v1.pdf | Sequence-Level Knowledge Distillation for Model Compression of Attention-based Sequence-to-Sequence Speech Recognition | We investigate the feasibility of sequence-level knowledge distillation of
Sequence-to-Sequence (Seq2Seq) models for Large Vocabulary Continuous Speech
Recognition (LVSCR). We first use a pre-trained larger teacher model to
generate multiple hypotheses per utterance with beam search. With the same
input, we then train the student model using these hypotheses generated from
the teacher as pseudo labels in place of the original ground truth labels. We
evaluate our proposed method using Wall Street Journal (WSJ) corpus. It
achieved up to $ 9.8 \times$ parameter reduction with accuracy loss of up to
7.0\% word-error rate (WER) increase | ["Raden Mu'az Mun'im", 'Nakamasa Inoue', 'Koichi Shinoda'] | 2018-11-12 | null | null | null | null | ['sequence-to-sequence-speech-recognition'] | ['speech'] | [ 6.98009133e-01 7.30535030e-01 8.45754445e-02 -4.73113924e-01
-1.49255240e+00 -4.93456364e-01 3.88442427e-01 5.36768995e-02
-8.25730741e-01 1.36642981e+00 2.77436852e-01 -7.58532047e-01
3.85956168e-01 -4.05821383e-01 -8.59696209e-01 -5.13453066e-01
2.38214374e-01 5.50181687e-01 3.01061362e-01 -3.29063416e-01
4.74842489e-02 8.75667576e-03 -1.18975556e+00 4.59853560e-01
8.05209458e-01 6.54063165e-01 6.67186320e-01 1.12097633e+00
-2.86326855e-01 1.01477981e+00 -1.05822110e+00 -3.21105301e-01
6.36298433e-02 -6.35168076e-01 -1.08019102e+00 6.01301976e-02
3.18936467e-01 -1.54633969e-02 -4.16269720e-01 1.00466323e+00
6.91527963e-01 5.07275403e-01 2.41160840e-01 -4.95547950e-01
-4.73349541e-01 1.10195661e+00 -3.25333187e-03 4.52860415e-01
3.29050720e-01 2.49443173e-01 1.04093671e+00 -8.61696362e-01
5.75501859e-01 1.20665503e+00 1.32663667e-01 8.41578960e-01
-1.19465756e+00 -7.34415531e-01 8.64182934e-02 2.11709246e-01
-1.60397208e+00 -9.58111882e-01 2.86847144e-01 2.83533875e-02
1.72488272e+00 3.56031090e-01 2.23293468e-01 1.14720535e+00
-3.31225604e-01 8.53012741e-01 9.99941707e-01 -8.67314458e-01
2.52296627e-01 2.00446740e-01 9.51052681e-02 7.88961530e-01
-2.93312460e-01 1.39122084e-01 -6.23237848e-01 1.23061448e-01
2.96726525e-01 -7.85605371e-01 -3.51297885e-01 4.14908201e-01
-1.06988645e+00 9.12609518e-01 -1.05427384e-01 8.25574696e-02
-1.98932454e-01 1.85978249e-01 3.89644057e-01 3.58835578e-01
3.19793701e-01 5.58572590e-01 -8.75629485e-01 -4.66835022e-01
-9.72367704e-01 -4.95920554e-02 8.28933716e-01 1.26147079e+00
4.39797282e-01 7.67692685e-01 -4.35073137e-01 1.14611685e+00
2.39128277e-01 7.78327644e-01 9.93535519e-01 -5.15948296e-01
7.33883440e-01 -1.70862973e-01 -3.94349322e-02 -3.16911131e-01
9.47272778e-02 -6.79148376e-01 -5.48638284e-01 -4.94645864e-01
7.55795417e-03 -3.46792996e-01 -1.19984317e+00 1.72194195e+00
1.17916591e-01 4.43275899e-01 6.20079279e-01 3.82753521e-01
8.23644280e-01 1.12577093e+00 1.72735870e-01 -5.10638714e-01
1.05447733e+00 -1.24301612e+00 -8.67632151e-01 -3.75354826e-01
8.25821042e-01 -8.11833084e-01 9.48532641e-01 4.23624933e-01
-1.37568486e+00 -6.30604744e-01 -8.30538154e-01 2.57131130e-01
-1.08880766e-01 3.39343876e-01 -2.51706004e-01 7.51985013e-01
-1.12616765e+00 2.17235342e-01 -5.09771585e-01 5.96425347e-02
2.03695908e-01 3.36450875e-01 -4.17397969e-04 9.98332500e-02
-1.34520328e+00 9.59895551e-01 6.95607185e-01 -1.34536415e-01
-1.16710973e+00 -7.75070548e-01 -9.49504793e-01 -3.53032947e-02
5.54992318e-01 -1.18033469e-01 1.74290836e+00 -8.23321819e-01
-2.21917152e+00 5.94274938e-01 -3.95493686e-01 -8.67433071e-01
1.30770309e-02 -1.34558514e-01 -6.28249168e-01 -9.41421986e-02
-3.50387275e-01 7.07872927e-01 6.09562039e-01 -9.66414273e-01
-7.03261197e-01 2.99547911e-01 -2.48975277e-01 1.85396373e-01
1.00489631e-01 1.63164794e-01 -1.56460181e-01 -7.73516893e-01
-3.44991207e-01 -8.90741825e-01 -1.85689777e-01 -1.00510073e+00
-3.93189132e-01 -4.26356405e-01 4.33547139e-01 -1.10552859e+00
1.41945088e+00 -1.98331320e+00 -5.63529842e-02 5.78840375e-02
-2.41908178e-01 9.27375317e-01 -4.69705105e-01 3.76440734e-01
5.39324656e-02 1.42688066e-01 -2.28623375e-01 -3.24572027e-01
-2.29435250e-01 3.26133668e-01 -4.61474091e-01 6.93632364e-02
3.44133824e-01 9.38023448e-01 -1.00505269e+00 -1.42541379e-01
1.50249511e-01 2.97819674e-01 -4.25196856e-01 5.67869365e-01
-6.09225273e-01 3.77469540e-01 -8.07885230e-02 5.70901632e-02
3.13571602e-01 9.07256007e-02 3.04675102e-01 3.33986968e-01
2.21615747e-01 8.60460639e-01 -9.56068337e-01 1.65456688e+00
-8.11676800e-01 6.51557684e-01 -1.60469711e-01 -1.19861770e+00
1.15484774e+00 7.04558969e-01 -2.58462667e-01 -7.94902980e-01
3.16533446e-02 2.16967285e-01 2.44720370e-01 -3.88956040e-01
4.21451002e-01 -4.40409362e-01 4.11571143e-03 1.51677698e-01
4.57490534e-01 -2.18957007e-01 -2.26166889e-01 -3.27207856e-02
1.07848716e+00 -3.03160548e-02 3.87855560e-01 -1.22362107e-01
7.73161650e-01 -3.11332103e-02 3.27305585e-01 8.39527845e-01
1.23088069e-01 3.83724362e-01 1.16130412e-02 2.59405058e-02
-1.13279688e+00 -9.06696022e-01 1.78689361e-01 1.23540354e+00
-4.04432178e-01 -1.77189603e-01 -8.48235905e-01 -6.96660817e-01
-3.80865663e-01 1.37671828e+00 -9.96720418e-02 -1.02776915e-01
-8.88990879e-01 -1.21005982e-01 1.10633671e+00 3.84287000e-01
2.52331585e-01 -1.35546625e+00 -1.18258305e-01 4.14118469e-01
-1.21546507e-01 -1.42661309e+00 -6.03154123e-01 2.73944497e-01
-5.79757273e-01 -5.23238122e-01 -5.64587891e-01 -9.79028523e-01
5.76923192e-01 -3.17291915e-01 9.96754408e-01 -1.26976132e-01
-1.52290255e-01 -6.94546849e-02 -5.21963596e-01 -3.47932488e-01
-1.09753954e+00 4.58434969e-02 8.46193135e-02 -1.79442719e-01
2.06341982e-01 3.32922675e-02 -6.44689724e-02 -3.36920959e-03
-6.94422245e-01 6.70031682e-02 6.12323761e-01 1.07107055e+00
4.81303662e-01 -1.40244141e-01 9.41360235e-01 -8.34838212e-01
7.10230887e-01 -3.01238894e-01 -7.69024134e-01 4.10150468e-01
-4.36247200e-01 3.94020140e-01 9.14264202e-01 -5.34166574e-01
-1.26907420e+00 -1.18653672e-02 -6.92328155e-01 -3.94424766e-01
-3.88164222e-01 5.17256677e-01 -7.42028654e-02 3.10303688e-01
5.03048956e-01 7.40363598e-01 -8.13837722e-02 -5.21621943e-01
6.16637170e-01 1.15485895e+00 5.70515275e-01 -4.84976232e-01
4.72185820e-01 -3.75410497e-01 -5.66633284e-01 -1.20959079e+00
-9.28349197e-01 -3.49942356e-01 -3.94477725e-01 2.08508223e-01
7.77787149e-01 -1.03502846e+00 -4.91247892e-01 1.45560741e-01
-1.27032769e+00 -6.84724510e-01 -4.11720157e-01 5.22901177e-01
-5.08455634e-01 2.21776798e-01 -4.29945767e-01 -1.03183103e+00
-5.91065466e-01 -1.08149064e+00 8.87692034e-01 2.39889976e-02
-2.44496182e-01 -8.48896682e-01 -1.23190910e-01 5.96536934e-01
4.82583225e-01 -5.25573730e-01 8.71903658e-01 -1.33883595e+00
-4.17211771e-01 -3.22460905e-02 2.57622570e-01 8.34677935e-01
-9.40794200e-02 -6.13771021e-01 -9.77224350e-01 -3.21582556e-01
-1.79035529e-01 -5.72720647e-01 4.64770943e-01 2.51694676e-02
1.25147319e+00 -7.21829116e-01 -7.63842911e-02 1.55111566e-01
1.19577324e+00 5.57149172e-01 5.70841551e-01 -2.33873710e-01
4.08826828e-01 3.63061070e-01 3.64157110e-01 2.46505708e-01
7.33489692e-02 7.58762538e-01 -3.55958134e-01 3.03345412e-01
-5.21749735e-01 -5.62449455e-01 6.26323283e-01 1.61014104e+00
5.26859641e-01 -6.69320166e-01 -9.64362323e-01 6.72287107e-01
-1.37786567e+00 -7.50150681e-01 2.28770807e-01 2.31480074e+00
1.50142765e+00 2.81265706e-01 -1.37855604e-01 -1.41489059e-01
6.06643200e-01 -6.17143437e-02 -1.83814541e-01 -7.15497375e-01
1.07853718e-01 9.03715670e-01 5.35622418e-01 1.21923435e+00
-5.49225211e-01 1.56669366e+00 6.35392475e+00 1.41428733e+00
-1.15628910e+00 2.59865761e-01 6.27981186e-01 -1.22245692e-01
-2.82859564e-01 -1.35264471e-01 -1.21003330e+00 4.14816231e-01
1.95518470e+00 -2.01759949e-01 5.57414889e-01 6.38823867e-01
2.63944775e-01 2.04502046e-02 -8.72057676e-01 7.46710956e-01
2.15852857e-01 -1.35217929e+00 1.00219645e-01 -2.54833460e-01
7.49106586e-01 6.64715245e-02 -3.00943047e-01 7.24330068e-01
7.36637235e-01 -1.29476547e+00 5.48892140e-01 2.10361511e-01
1.17936051e+00 -7.68063664e-01 7.39159226e-01 7.56269455e-01
-7.66928852e-01 1.79040253e-01 -3.31981987e-01 5.00900857e-02
3.85866314e-01 2.37523064e-01 -1.83295977e+00 3.03955466e-01
-1.33473342e-02 -7.35361055e-02 1.83667615e-03 6.52239501e-01
-4.53909427e-01 1.43900084e+00 -3.42210919e-01 -5.06015658e-01
3.93885583e-01 2.19937071e-01 4.24992621e-01 1.64683318e+00
1.80076823e-01 4.38922435e-01 1.69362858e-01 4.16222543e-01
-2.74999350e-01 1.93528011e-01 -2.49308228e-01 -1.83137268e-01
7.69612014e-01 5.31269968e-01 -3.52703094e-01 -9.03084457e-01
-3.14734459e-01 1.06801140e+00 4.50356394e-01 4.11405891e-01
-6.50701702e-01 -4.42670494e-01 5.05809069e-01 -1.51096284e-01
6.98970795e-01 -2.99559981e-01 1.30665287e-01 -7.27845907e-01
-1.45267636e-01 -1.22264576e+00 3.31905350e-04 -7.42098689e-01
-8.88294935e-01 8.37558389e-01 -5.39866462e-02 -6.38931870e-01
-5.98137021e-01 -2.78374970e-01 -3.02161604e-01 1.32070053e+00
-1.57358313e+00 -6.84383392e-01 3.71641368e-01 1.15597948e-01
1.42694354e+00 -4.96609658e-01 1.18906188e+00 -1.84771977e-02
-4.99300599e-01 1.04400837e+00 3.98608148e-01 2.52619475e-01
2.19673634e-01 -1.12648666e+00 7.86765814e-01 6.32735193e-01
4.24090922e-01 5.02615690e-01 6.17235422e-01 -8.00399303e-01
-1.24969828e+00 -1.15590441e+00 1.30911613e+00 -3.11454743e-01
6.01444781e-01 -5.27333736e-01 -8.71823490e-01 5.89541733e-01
4.58408684e-01 -1.88735262e-01 7.38205671e-01 -2.11569354e-01
-1.91305667e-01 1.72458500e-01 -1.03741419e+00 5.04903555e-01
8.57761323e-01 -5.43881357e-01 -7.87281692e-01 2.18745053e-01
1.36882222e+00 -6.54110610e-01 -6.94942176e-01 3.17484677e-01
6.41219541e-02 -7.24089742e-02 8.42475533e-01 -9.58729327e-01
1.16180763e-01 -5.26461788e-02 -3.34257543e-01 -1.58595407e+00
-4.77755954e-03 -8.83230686e-01 -1.41590834e-01 1.02524173e+00
9.80924904e-01 -3.40302974e-01 5.61966062e-01 1.78911358e-01
-4.57659215e-01 -8.66839588e-01 -1.14069080e+00 -1.06066847e+00
1.78093493e-01 -6.99642599e-01 4.98027503e-01 5.36099792e-01
1.73804805e-01 7.02573955e-01 -2.88621068e-01 2.82520801e-01
1.73403814e-01 -5.33773303e-01 5.31347275e-01 -4.38500017e-01
-6.02900863e-01 -1.75224528e-01 -2.04394758e-01 -1.46481431e+00
3.39187592e-01 -1.02093077e+00 6.47199810e-01 -1.17445159e+00
-2.19881743e-01 -3.36722136e-01 -3.22684050e-01 4.10721034e-01
-3.69967967e-01 -7.68213794e-02 1.41201138e-01 -4.59638953e-01
-6.29771054e-01 7.19238281e-01 8.98443222e-01 2.49010008e-02
-9.42614228e-02 -2.34047370e-03 -3.27237099e-01 2.09002778e-01
8.66869926e-01 -5.01532912e-01 -5.18395424e-01 -4.07376051e-01
-2.86095291e-01 4.13558513e-01 -2.53946215e-01 -7.52314746e-01
9.49337631e-02 -2.18884751e-01 -1.75869673e-01 -3.37700069e-01
3.62679392e-01 -3.67389381e-01 -2.47506559e-01 4.64068055e-01
-9.18170810e-01 -2.11700872e-01 5.83156884e-01 6.64745986e-01
-1.69831708e-01 -5.52288771e-01 7.20103800e-01 -1.95978448e-01
-6.63308680e-01 -2.30347976e-01 -5.88479578e-01 3.85513186e-01
8.48937333e-01 7.05322549e-02 -1.84548646e-01 -4.97081369e-01
-7.74386525e-01 1.14494182e-01 -3.90516877e-01 5.70318520e-01
6.86697006e-01 -8.67205620e-01 -1.04302871e+00 3.75104249e-01
1.50390957e-02 2.81152432e-03 -2.55225487e-02 3.34385365e-01
-2.10319519e-01 9.72020268e-01 4.92888600e-01 -2.89421767e-01
-1.43991065e+00 3.56507182e-01 2.24149019e-01 -1.51676819e-01
-4.94802624e-01 1.46981752e+00 -2.37758979e-01 -7.07629800e-01
5.36924541e-01 -3.44242990e-01 1.78909060e-02 -5.54441988e-01
5.19000709e-01 2.31499776e-01 2.18220696e-01 -5.65655351e-01
-2.42618501e-01 -8.69360566e-02 -1.32381469e-01 -5.40374756e-01
1.12536597e+00 -7.80820176e-02 4.89659369e-01 4.06221569e-01
1.27867234e+00 1.66359648e-01 -8.42162132e-01 -5.95764816e-01
1.65113911e-01 -1.65459052e-01 -9.98865999e-03 -1.16314077e+00
-4.56937879e-01 7.69176126e-01 3.80539507e-01 -1.59047723e-01
5.85004091e-01 1.15660399e-01 1.12139094e+00 5.53882957e-01
1.94742382e-01 -1.21682143e+00 -4.19671851e-04 8.72246504e-01
6.37856305e-01 -1.08541596e+00 -4.20366973e-01 -2.94844300e-01
-8.69285822e-01 8.21692288e-01 4.90400583e-01 1.74383953e-01
4.52418596e-01 3.49227399e-01 2.18870431e-01 1.55079782e-01
-1.01371133e+00 -2.80279011e-01 2.55955428e-01 4.31922346e-01
5.59328496e-01 4.09805417e-01 -2.91147679e-01 4.02755916e-01
-5.82931995e-01 -8.75987783e-02 4.81419355e-01 7.63866484e-01
-8.44795108e-01 -1.15993690e+00 -1.26784936e-01 4.24500585e-01
-3.78486484e-01 -6.58069193e-01 1.44947767e-02 1.74381748e-01
-2.74324417e-01 1.17359018e+00 3.33528221e-02 -4.21496391e-01
2.95616776e-01 7.32390702e-01 2.43561253e-01 -1.09587479e+00
-4.36361849e-01 1.04584053e-01 6.44620538e-01 -2.53031850e-01
2.21023142e-01 -5.18422127e-01 -1.48147726e+00 3.05857658e-01
-4.02715147e-01 4.91785139e-01 8.96699667e-01 1.19522774e+00
3.29265445e-01 6.65584803e-01 5.01548588e-01 -4.07674015e-02
-1.00354934e+00 -1.39772081e+00 -1.01875611e-01 1.30918652e-01
3.89484525e-01 -8.24287683e-02 -2.11608917e-01 3.98978114e-01] | [14.431419372558594, 6.870495319366455] |
aaeaddd5-6b9a-4a76-b967-ed62aca04ee4 | 1st-place-solution-for-youtubevos-challenge | 2106.06649 | null | https://arxiv.org/abs/2106.06649v2 | https://arxiv.org/pdf/2106.06649v2.pdf | 1st Place Solution for YouTubeVOS Challenge 2021:Video Instance Segmentation | Video Instance Segmentation (VIS) is a multi-task problem performing detection, segmentation, and tracking simultaneously. Extended from image set applications, video data additionally induces the temporal information, which, if handled appropriately, is very useful to identify and predict object motions. In this work, we design a unified model to mutually learn these tasks. Specifically, we propose two modules, named Temporally Correlated Instance Segmentation (TCIS) and Bidirectional Tracking (BiTrack), to take the benefit of the temporal correlation between the object's instance masks across adjacent frames. On the other hand, video data is often redundant due to the frame's overlap. Our analysis shows that this problem is particularly severe for the YoutubeVOS-VIS2021 data. Therefore, we propose a Multi-Source Data (MSD) training mechanism to compensate for the data deficiency. By combining these techniques with a bag of tricks, the network performance is significantly boosted compared to the baseline, and outperforms other methods by a considerable margin on the YoutubeVOS-VIS 2019 and 2021 datasets. | ['Masao Yamanaka', 'Masayuki Yamazaki', 'Chuong H. Nguyen', 'Nam LH. Phan', 'Tuan N. Tang', 'Thuy C. Nguyen'] | 2021-06-12 | null | null | null | null | ['video-instance-segmentation'] | ['computer-vision'] | [ 6.12079725e-02 -3.87744576e-01 -4.61505413e-01 -2.59174943e-01
-7.27368414e-01 -3.45831126e-01 4.23815906e-01 -2.74507701e-01
-4.70463663e-01 6.53911173e-01 4.25227322e-02 4.65457030e-02
7.33021870e-02 -2.75976121e-01 -8.62082899e-01 -8.58129561e-01
7.54550248e-02 -3.10255978e-02 7.24856973e-01 1.57075584e-01
5.91474213e-02 4.11615819e-02 -1.58254981e+00 4.89895403e-01
6.60906911e-01 1.20860398e+00 6.03335261e-01 3.80787641e-01
-2.08836034e-01 1.05201316e+00 -4.47835386e-01 -1.37415007e-01
4.22188699e-01 -3.24920893e-01 -7.73371041e-01 3.85173768e-01
5.55998087e-01 -6.00602746e-01 -3.44905674e-01 8.90925646e-01
2.68627644e-01 2.09459960e-01 3.54010284e-01 -1.42010188e+00
-9.18537304e-02 3.87028277e-01 -1.00455403e+00 5.45578003e-01
9.25664306e-02 3.19148928e-01 8.56108785e-01 -7.84316957e-01
7.20915616e-01 1.19251215e+00 4.73711044e-01 5.75116634e-01
-8.12533259e-01 -7.82215297e-01 5.39744377e-01 3.40089589e-01
-1.14251244e+00 -4.35966909e-01 7.83457458e-01 -5.78795254e-01
4.46057826e-01 1.86407045e-01 7.30919600e-01 1.29585898e+00
-2.22538963e-01 1.33509851e+00 7.74969935e-01 -2.03417856e-02
-6.87313005e-02 1.53522734e-02 1.47962809e-01 4.80040938e-01
-6.87381849e-02 7.42605403e-02 -5.15650153e-01 4.00838703e-01
8.03660750e-01 2.00604796e-01 -4.13773686e-01 -3.69447947e-01
-1.24252915e+00 2.60239840e-01 2.28436515e-01 2.89141089e-01
-2.58126616e-01 2.83683926e-01 6.04360640e-01 1.59051090e-01
5.87918222e-01 -4.18503545e-02 -6.76068425e-01 -2.27926701e-01
-1.15368438e+00 1.60133541e-01 4.01936650e-01 1.09296823e+00
5.53661764e-01 1.73321143e-01 -4.41010267e-01 7.70635188e-01
2.44670838e-01 2.17993051e-01 4.22649086e-01 -1.00915539e+00
7.53320694e-01 4.15048510e-01 1.87595949e-01 -9.19023216e-01
-1.55195922e-01 -3.22722971e-01 -7.67089725e-01 -1.92720443e-02
5.48019052e-01 -2.36275375e-01 -9.65119302e-01 1.69706178e+00
5.90810537e-01 8.14210117e-01 -1.78799003e-01 1.19816828e+00
9.88819718e-01 6.99103773e-01 1.98985577e-01 -4.12413299e-01
1.25060391e+00 -1.26266181e+00 -8.99284661e-01 -2.81801641e-01
5.27619958e-01 -5.54436862e-01 9.07005548e-01 2.46589139e-01
-9.36733603e-01 -7.82496929e-01 -8.87256503e-01 8.82176235e-02
-1.41168728e-01 2.16055751e-01 5.67680299e-01 3.32426727e-01
-6.53411031e-01 5.80860913e-01 -9.11345601e-01 3.46147791e-02
6.49194717e-01 2.40080640e-01 -1.80554911e-01 -4.40524630e-02
-1.01527178e+00 4.27787334e-01 5.47278702e-01 1.72063574e-01
-8.83078992e-01 -5.94957471e-01 -6.89828157e-01 -8.82187858e-02
8.74243736e-01 -3.42199594e-01 1.11641526e+00 -1.28912902e+00
-1.16412294e+00 5.73119581e-01 -2.61217356e-01 -4.60184306e-01
8.24895144e-01 -4.03097361e-01 -2.40355134e-01 2.41841108e-01
1.89040452e-01 7.60494173e-01 1.13904405e+00 -1.16929007e+00
-9.09272194e-01 -2.30023324e-01 5.44947349e-02 2.09632352e-01
-4.12301809e-01 1.24917477e-01 -1.27146316e+00 -8.72256875e-01
-1.26293063e-01 -8.83952677e-01 -1.30369380e-01 -1.10524017e-02
-4.11877781e-01 -2.42027834e-01 1.27822661e+00 -7.84741282e-01
1.40741885e+00 -2.41866446e+00 2.44608656e-01 -2.20478252e-01
1.67296216e-01 4.88924950e-01 -1.83417693e-01 -1.39258474e-01
-1.00062519e-01 1.77446101e-02 -8.70779231e-02 -4.51412320e-01
-4.85190064e-01 2.53275543e-01 -4.45984975e-02 2.64246196e-01
1.83904782e-01 9.24144328e-01 -8.94832969e-01 -7.46762693e-01
2.55183548e-01 8.62576962e-02 -4.46203560e-01 1.53288424e-01
-4.98692065e-01 7.62261391e-01 -5.74230015e-01 7.24212945e-01
5.58158875e-01 -4.23781663e-01 9.51203853e-02 -4.24560994e-01
-8.55231956e-02 -1.15584508e-01 -1.30042028e+00 1.71811628e+00
-1.45323783e-01 7.01424122e-01 8.63372162e-02 -1.07494926e+00
4.68127400e-01 4.70579326e-01 9.99187946e-01 -8.79183054e-01
7.01682642e-02 3.15398462e-02 -1.34299889e-01 -8.66034389e-01
3.66045803e-01 4.30674493e-01 2.50003457e-01 1.48846209e-01
-9.93428677e-02 4.21142191e-01 5.20438552e-01 1.62540540e-01
8.93393874e-01 5.22472262e-01 4.05705757e-02 -6.88279374e-03
5.48253715e-01 -9.38749984e-02 1.06985605e+00 4.51135308e-01
-4.38670188e-01 6.05482817e-01 4.54744756e-01 -3.95398051e-01
-7.98249483e-01 -5.84150851e-01 1.11154541e-01 9.95355964e-01
6.12650394e-01 -4.86801624e-01 -7.23245323e-01 -9.13662136e-01
2.19722856e-02 2.57900357e-01 -5.28875887e-01 6.67313710e-02
-8.29525352e-01 -6.77876294e-01 2.11943224e-01 7.33958304e-01
7.38976955e-01 -1.03770399e+00 -5.83914101e-01 2.94900209e-01
-4.06425327e-01 -1.53854024e+00 -8.86164188e-01 -2.16997582e-02
-8.54551017e-01 -1.29770470e+00 -9.51293230e-01 -6.25478208e-01
3.32086921e-01 6.95217490e-01 8.55501354e-01 1.37663022e-01
-1.42873585e-01 1.70320779e-01 -4.77503151e-01 -7.85158724e-02
-5.55677004e-02 -1.41067609e-01 -1.61269233e-01 2.53113627e-01
2.51754254e-01 -3.52255076e-01 -6.33812487e-01 5.31237900e-01
-9.84551013e-01 3.88099849e-01 4.24018443e-01 6.93145454e-01
6.76453173e-01 -9.94588807e-02 4.89364445e-01 -7.68382430e-01
6.13421425e-02 -5.19248843e-01 -5.86806118e-01 2.23672673e-01
-3.80390555e-01 -3.06880176e-01 4.53769356e-01 -6.47706389e-01
-9.80438590e-01 2.93605924e-01 2.02103823e-01 -1.05227077e+00
-1.11994475e-01 2.78338999e-01 -2.62714535e-01 6.28685430e-02
1.68532029e-01 1.83187053e-01 4.97350916e-02 -6.08146906e-01
1.26996696e-01 5.14295280e-01 4.16372061e-01 -2.08458528e-01
4.97103333e-01 4.44855005e-01 -2.94875056e-01 -7.27809608e-01
-9.73176241e-01 -7.24692345e-01 -5.44856310e-01 -4.77353722e-01
1.17259204e+00 -1.19497252e+00 -5.56620359e-01 5.84278524e-01
-1.11411655e+00 -4.85375881e-01 -5.92870042e-02 4.43852603e-01
-3.35963756e-01 5.18376112e-01 -7.13716388e-01 -6.61254942e-01
-9.27470922e-02 -1.36449277e+00 1.04042661e+00 2.13489771e-01
1.24176197e-01 -6.61453485e-01 -3.61390114e-01 5.50752223e-01
3.24863680e-02 3.23324054e-01 4.01469052e-01 -5.00327706e-01
-1.01212108e+00 1.15924619e-01 -4.94712442e-01 5.49346328e-01
2.33687624e-01 1.29380852e-01 -7.82470167e-01 -1.69256374e-01
-7.91689288e-03 -7.78916925e-02 1.10434043e+00 6.35924339e-01
1.61091971e+00 -1.98680133e-01 -4.38179433e-01 6.70753360e-01
1.23701370e+00 4.92628902e-01 4.64538693e-01 4.06694919e-01
1.08478749e+00 4.57025439e-01 9.36308265e-01 3.70604903e-01
4.05491650e-01 9.90921497e-01 5.92439473e-01 -1.51754200e-01
-3.23742211e-01 1.68932453e-01 3.96848708e-01 6.52833521e-01
-1.84652701e-01 -3.38385701e-01 -7.11075902e-01 5.89540184e-01
-2.21586418e+00 -1.11622310e+00 -3.61409396e-01 2.01180840e+00
6.20644093e-01 1.37692660e-01 3.21490079e-01 2.39391457e-02
8.66282165e-01 4.45854515e-01 -6.83074892e-01 3.35709929e-01
-1.19916201e-01 -4.49746042e-01 6.48270369e-01 7.75626600e-02
-1.35223305e+00 9.58999276e-01 5.73631144e+00 1.00894856e+00
-1.08599424e+00 2.05170512e-01 8.14827025e-01 -3.45964968e-01
9.98312905e-02 -1.38762549e-01 -8.21195483e-01 1.01873529e+00
4.91559029e-01 1.36571154e-01 2.85840809e-01 6.57231867e-01
4.48665380e-01 -1.14769466e-01 -9.90824521e-01 1.11828148e+00
-3.59603167e-02 -1.31778085e+00 -1.62084743e-01 -8.78432542e-02
6.96403444e-01 1.79434624e-02 -1.47115709e-02 2.51729459e-01
-1.72432572e-01 -5.19790649e-01 8.95917058e-01 2.87535101e-01
6.16814435e-01 -4.49423522e-01 5.22656918e-01 3.37362498e-01
-1.34716880e+00 -2.31870368e-01 -4.36516525e-03 2.61556417e-01
4.58173573e-01 5.57446897e-01 -3.21249068e-01 6.63287580e-01
9.61139083e-01 1.28542292e+00 -4.98734087e-01 1.18663323e+00
-5.23677375e-03 5.59455812e-01 -1.89064398e-01 3.20481241e-01
3.93242955e-01 -2.99214095e-01 5.65585613e-01 1.12051845e+00
1.74213991e-01 5.43225482e-02 4.26472276e-01 6.40792072e-01
-9.63753313e-02 -1.09168544e-01 -2.78384209e-01 3.15438807e-02
2.72753298e-01 1.23300219e+00 -8.02250504e-01 -6.54274166e-01
-6.81412101e-01 1.00691998e+00 6.07654639e-02 5.07589042e-01
-1.23378968e+00 1.36427164e-01 6.86003983e-01 6.78019002e-02
6.22474313e-01 -2.23016143e-01 -5.61937038e-03 -1.32302642e+00
2.45466560e-01 -8.20806205e-01 2.94664085e-01 -6.19442880e-01
-9.50875878e-01 3.47683281e-01 2.79348139e-02 -1.53195870e+00
-9.04532298e-02 -3.45891327e-01 -5.36769688e-01 3.73698950e-01
-1.63730180e+00 -9.65622008e-01 -5.01438737e-01 7.19860911e-01
1.05613077e+00 -1.41884554e-02 3.71352583e-02 8.00434947e-01
-1.04483664e+00 3.77004534e-01 4.10629511e-02 2.89922625e-01
7.22872496e-01 -1.01857436e+00 2.22425282e-01 9.78072047e-01
1.13178901e-01 1.61618114e-01 5.02426863e-01 -7.15549707e-01
-1.24287379e+00 -1.39130044e+00 3.63901347e-01 -2.59763509e-01
4.44070607e-01 -1.42859861e-01 -1.00059688e+00 6.12215519e-01
3.84661630e-02 2.93695122e-01 3.34013909e-01 -2.55544782e-01
-1.94985464e-01 -2.24164397e-01 -8.28659952e-01 4.89943832e-01
1.21017146e+00 -2.08058953e-01 -2.94054240e-01 5.29402375e-01
9.47384357e-01 -5.72644770e-01 -6.43543482e-01 4.92248714e-01
4.77184981e-01 -1.07936645e+00 1.06199527e+00 -5.25142908e-01
5.59728384e-01 -4.31280851e-01 -2.21407115e-02 -8.73138547e-01
-6.57088384e-02 -5.65706372e-01 -5.21549106e-01 1.36711144e+00
5.92845641e-02 -1.25390828e-01 8.78095448e-01 4.85912234e-01
-1.37719780e-01 -7.05671310e-01 -8.49940538e-01 -8.86055470e-01
-5.16946673e-01 -4.74504024e-01 3.98709297e-01 9.33905900e-01
-4.06576484e-01 1.82579428e-01 -7.35509098e-01 2.44053230e-02
5.40346742e-01 1.77488819e-01 8.60419929e-01 -8.73702943e-01
-3.43289375e-01 -3.44818741e-01 -2.26644978e-01 -1.48803961e+00
4.16754298e-02 -5.52192211e-01 1.25151679e-01 -1.16575217e+00
1.54364318e-01 -4.34488505e-01 -3.50722224e-01 4.89464492e-01
-5.02804697e-01 1.75187305e-01 4.97852206e-01 5.04031122e-01
-1.03157878e+00 4.75858867e-01 1.60033989e+00 -6.49332926e-02
-2.96474308e-01 1.87479272e-01 -3.95174563e-01 6.91029489e-01
5.71551442e-01 -3.30123484e-01 -3.50821227e-01 -6.24125957e-01
-2.32968286e-01 3.16462606e-01 4.84789819e-01 -9.30134892e-01
2.72702813e-01 -2.60515332e-01 2.23521754e-01 -8.47153366e-01
4.63746428e-01 -9.71479416e-01 2.54440814e-01 3.28084588e-01
-1.40305772e-01 -1.79543421e-02 2.16193333e-01 8.55188847e-01
-4.59569931e-01 2.23254338e-02 6.90326571e-01 -2.07200781e-01
-1.24292994e+00 7.08709180e-01 -1.81540221e-01 8.91558826e-02
1.29510105e+00 -4.47721928e-01 -1.07980311e-01 -1.38526335e-01
-5.87958097e-01 6.34149551e-01 2.22122893e-01 7.01051772e-01
5.25749922e-01 -1.30049801e+00 -4.39693213e-01 1.02882355e-01
-4.83068489e-02 2.69376576e-01 4.94880199e-01 1.20695567e+00
-1.52070314e-01 2.12432936e-01 -5.57019413e-02 -8.84993017e-01
-1.41189492e+00 6.16553783e-01 7.74012506e-02 -1.39446974e-01
-7.63160586e-01 8.76518607e-01 3.18295270e-01 1.86920658e-01
5.16501009e-01 -3.67992938e-01 -4.31817472e-01 3.93371105e-01
5.68841934e-01 3.93397301e-01 -2.25955904e-01 -6.25225604e-01
-2.86454290e-01 5.55765986e-01 -1.70424297e-01 2.83658564e-01
1.24163342e+00 -3.86783570e-01 2.24207208e-01 3.78682375e-01
1.14316988e+00 -2.75985062e-01 -1.83373833e+00 -3.19514871e-01
4.83065136e-02 -6.52493179e-01 -4.89708185e-02 -4.39946979e-01
-1.59823394e+00 6.71824694e-01 4.70565945e-01 2.22385466e-01
1.17363417e+00 -1.13898605e-01 1.08024895e+00 -4.75500822e-02
2.15019062e-01 -1.26244211e+00 2.72640109e-01 3.39463979e-01
5.38832128e-01 -1.54069626e+00 -1.65871561e-01 -6.16627753e-01
-9.29372549e-01 8.43349516e-01 8.99400115e-01 7.25202262e-02
4.66029018e-01 2.02509806e-01 -5.13007678e-02 -1.26810593e-03
-7.50189543e-01 -3.24607730e-01 3.08903426e-01 2.86298662e-01
1.82188824e-01 -3.09824288e-01 -1.96686864e-01 5.35581172e-01
5.30101359e-01 1.75671950e-01 2.36568674e-01 8.09990108e-01
-3.60832810e-01 -9.00483966e-01 -2.72947371e-01 5.40791273e-01
-5.60307562e-01 4.63113301e-02 5.55516332e-02 8.97485614e-01
3.03867638e-01 7.49994159e-01 -1.06365711e-03 -3.75396043e-01
8.20029825e-02 -2.32782722e-01 1.98726460e-01 -3.30121249e-01
-5.73064744e-01 3.21238816e-01 -8.17345083e-02 -9.92407680e-01
-9.56263423e-01 -7.84079492e-01 -1.14692342e+00 4.55886386e-02
-4.35088277e-01 -2.20275670e-01 4.41995353e-01 1.16420567e+00
3.72743964e-01 8.71417344e-01 5.55616260e-01 -9.98130441e-01
-2.67403841e-01 -6.80541635e-01 -5.11678874e-01 6.47469223e-01
4.31131870e-01 -7.94859231e-01 -2.73714304e-01 2.64983982e-01] | [9.23218822479248, 0.02550787292420864] |
556f86ac-8bbe-4422-b1ac-9442d7e564cf | a-system-for-real-time-twitter-sentiment | null | null | https://aclanthology.org/P12-3020 | https://aclanthology.org/P12-3020.pdf | A System for Real-time Twitter Sentiment Analysis of 2012 U.S. Presidential Election Cycle | null | ['Fran{\\c{c}}ois Bar', 'Abe Kazemzadeh', 'Hao Wang', 'Shrikanth Narayanan', 'Dogan Can'] | 2012-07-01 | null | null | null | acl-2012-7 | ['twitter-sentiment-analysis'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.312845706939697, 3.64819598197937] |
f87d9a12-74f9-461c-acd5-d353dc177192 | group-collaborative-learning-for-co-salient | 2104.01108 | null | https://arxiv.org/abs/2104.01108v2 | https://arxiv.org/pdf/2104.01108v2.pdf | Group Collaborative Learning for Co-Salient Object Detection | We present a novel group collaborative learning framework (GCoNet) capable of detecting co-salient objects in real time (16ms), by simultaneously mining consensus representations at group level based on the two necessary criteria: 1) intra-group compactness to better formulate the consistency among co-salient objects by capturing their inherent shared attributes using our novel group affinity module; 2) inter-group separability to effectively suppress the influence of noisy objects on the output by introducing our new group collaborating module conditioning the inconsistent consensus. To learn a better embedding space without extra computational overhead, we explicitly employ auxiliary classification supervision. Extensive experiments on three challenging benchmarks, i.e., CoCA, CoSOD3k, and Cosal2015, demonstrate that our simple GCoNet outperforms 10 cutting-edge models and achieves the new state-of-the-art. We demonstrate this paper's new technical contributions on a number of important downstream computer vision applications including content aware co-segmentation, co-localization based automatic thumbnails, etc. | ['Yu-Wing Tai', 'Ling Shao', 'Chi Keung Tang', 'Huazhu Fu', 'Deng-Ping Fan', 'Qi Fan'] | 2021-03-15 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Fan_Group_Collaborative_Learning_for_Co-Salient_Object_Detection_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Fan_Group_Collaborative_Learning_for_Co-Salient_Object_Detection_CVPR_2021_paper.pdf | cvpr-2021-1 | ['co-saliency-detection'] | ['computer-vision'] | [-1.75756812e-02 6.59924326e-03 -3.14302742e-01 -1.88551083e-01
-9.01815236e-01 -2.90337890e-01 5.44067144e-01 4.76192653e-01
-2.67415375e-01 1.49761587e-01 1.62373975e-01 6.27878085e-02
-4.46174473e-01 -2.96453059e-01 -6.40121877e-01 -8.42164755e-01
-1.82954043e-01 3.02882075e-01 5.26211619e-01 1.70626506e-01
4.77417499e-01 -2.37022284e-02 -1.86482692e+00 4.35023576e-01
1.49726534e+00 1.04813957e+00 3.65027934e-01 2.88393378e-01
-2.08726581e-02 7.51947582e-01 -2.13898316e-01 -2.18835860e-01
3.57105285e-01 -2.42939070e-01 -6.41743422e-01 6.75925136e-01
7.56136596e-01 1.75535977e-01 2.09540084e-01 1.25680411e+00
4.91133660e-01 1.32053763e-01 7.08175242e-01 -1.43879175e+00
-8.25840712e-01 9.67200935e-01 -1.18175626e+00 2.61374265e-01
-2.74881031e-02 -3.84434983e-02 1.52871227e+00 -1.30744886e+00
7.02871442e-01 1.18601549e+00 5.35272777e-01 1.27652779e-01
-1.29845405e+00 -8.03122818e-01 7.87834704e-01 7.47084796e-01
-1.43630433e+00 -2.44207799e-01 1.00376451e+00 -2.21378818e-01
4.84980702e-01 3.24908018e-01 6.32188141e-01 7.80283153e-01
-1.63588807e-01 1.47311664e+00 1.05113471e+00 -3.78785789e-01
3.32668573e-01 1.22469338e-02 3.73119324e-01 8.66938412e-01
3.03667307e-01 -5.28455853e-01 -7.92641878e-01 -1.30607098e-01
4.85282391e-01 1.37744904e-01 -1.87460363e-01 -8.43798995e-01
-1.42957664e+00 7.52605498e-01 6.70505941e-01 4.26696241e-01
-1.48525938e-01 9.85111892e-02 4.29999948e-01 2.49691337e-01
6.84949458e-01 8.15216973e-02 -1.45179093e-01 4.44514573e-01
-1.06746483e+00 -3.15698832e-02 4.97031868e-01 1.19895518e+00
7.07524776e-01 -2.46987700e-01 -1.87667996e-01 9.96608853e-01
4.58857805e-01 1.15042485e-01 5.81533432e-01 -8.64653945e-01
6.64734006e-01 8.20976317e-01 -2.74620876e-02 -1.19969487e+00
-1.02103144e-01 -7.25629449e-01 -8.88313830e-01 -1.20982714e-02
9.33508947e-02 -1.21231377e-02 -7.65938759e-01 1.26105523e+00
5.58992445e-01 5.59670687e-01 -2.01812357e-01 1.15939093e+00
8.96232665e-01 4.46461856e-01 1.72771722e-01 -2.64326066e-01
9.48526442e-01 -1.60993886e+00 -6.02605522e-01 -2.47545376e-01
5.79374015e-01 -7.50406027e-01 7.87336588e-01 2.82891005e-01
-9.32041109e-01 -8.13390672e-01 -1.08991516e+00 6.14507496e-03
-3.44303012e-01 3.44334036e-01 1.02759433e+00 2.36429378e-01
-8.84630680e-01 6.63660169e-01 -6.79039299e-01 -5.15031554e-02
7.49764442e-01 3.10472041e-01 -2.46692732e-01 1.27198696e-01
-6.57439947e-01 2.48078883e-01 3.07862759e-01 1.98578071e-02
-5.76912642e-01 -7.74806261e-01 -5.90564966e-01 1.06715195e-01
7.91636467e-01 -7.03255773e-01 6.67362332e-01 -1.21955073e+00
-1.12396491e+00 9.53947067e-01 -1.83143746e-02 -3.83516848e-01
7.96053648e-01 -5.34317970e-01 -3.03050905e-01 1.21672973e-01
4.58242565e-01 7.65925825e-01 1.01550770e+00 -1.52248418e+00
-9.30038750e-01 -5.23754060e-01 -5.27053893e-01 4.76786375e-01
-5.15309751e-01 -7.62550458e-02 -9.31502700e-01 -8.19503486e-01
5.00209272e-01 -9.47649837e-01 -4.61407244e-01 2.92447299e-01
-4.73752677e-01 -6.75480843e-01 9.69823360e-01 -2.98493087e-01
1.13028634e+00 -2.31794834e+00 4.01356280e-01 2.41961136e-01
6.18858993e-01 7.11969882e-02 -1.30746394e-01 1.03692047e-01
-8.85899886e-02 -3.47240046e-02 -1.32073388e-01 -9.74222541e-01
7.75110498e-02 -1.95046850e-02 -1.02230020e-01 6.63873732e-01
7.30878562e-02 9.93428409e-01 -1.11007535e+00 -1.08912647e+00
3.41905653e-01 3.66127566e-02 -4.11287665e-01 2.40789093e-02
1.70671381e-02 8.92426521e-02 -3.80341470e-01 7.13831723e-01
6.28134608e-01 -6.66238129e-01 1.59988180e-01 -3.62715334e-01
-1.55669779e-01 -2.42753744e-01 -1.67225862e+00 1.82982290e+00
-1.18713833e-01 4.36751306e-01 2.65311837e-01 -1.05984747e+00
7.59216309e-01 -1.46796644e-01 7.42151976e-01 -3.71127248e-01
1.95317119e-01 9.04983133e-02 -1.85904771e-01 -2.07675382e-01
4.59912181e-01 7.77683020e-01 1.51424333e-01 5.00826359e-01
1.67332053e-01 4.01897699e-01 1.93035737e-01 6.34625673e-01
6.14741743e-01 -2.64805049e-01 1.45019349e-02 -7.55767226e-01
4.29187119e-01 -2.53322631e-01 7.34405160e-01 9.00149405e-01
-4.90595579e-01 8.03218067e-01 8.80796909e-02 -6.81297183e-02
-6.81600809e-01 -1.05752552e+00 2.43174106e-01 1.27181339e+00
7.49617100e-01 -5.97480178e-01 -3.83370996e-01 -1.03480828e+00
2.49390557e-01 9.51574743e-02 -7.76890337e-01 5.36965299e-03
-3.54540735e-01 -7.58797348e-01 -1.41954005e-01 7.66475141e-01
6.06059313e-01 -6.31147861e-01 -2.50600755e-01 1.02738872e-01
-2.25018486e-01 -1.00093627e+00 -9.73945677e-01 2.41455406e-01
-8.11914623e-01 -1.06155598e+00 -7.25329280e-01 -1.27052927e+00
9.46157575e-01 8.83480132e-01 8.55712295e-01 1.66797534e-01
-2.52828896e-01 3.48014683e-01 -4.77781087e-01 -4.36939120e-01
3.47879529e-01 -7.69622624e-02 -5.27573004e-02 4.51294154e-01
2.38162309e-01 -4.48078334e-01 -7.53619373e-01 4.41822320e-01
-5.15877366e-01 3.87931108e-01 5.32408297e-01 8.15344691e-01
9.39113736e-01 -5.11314385e-02 5.78705728e-01 -8.61756325e-01
4.04287905e-01 -3.38755220e-01 -3.95831078e-01 3.67908359e-01
-8.12263548e-01 -1.50840044e-01 4.05668199e-01 -6.08000934e-01
-8.85026038e-01 9.72726792e-02 5.29419065e-01 -7.42708623e-01
2.91240484e-01 1.95879951e-01 -2.20120996e-01 -6.39775349e-03
3.08541834e-01 1.72972307e-01 -1.68448105e-01 -4.67294157e-01
6.58181250e-01 5.39563298e-01 8.13233018e-01 -4.28409129e-01
7.93993175e-01 8.62403572e-01 -3.21761400e-01 -5.24076343e-01
-9.31140840e-01 -1.25505316e+00 -8.01180422e-01 -2.65840769e-01
6.92485332e-01 -1.22582161e+00 -4.26970124e-01 3.34158480e-01
-7.68621683e-01 1.10465668e-01 -2.00975582e-01 3.99006248e-01
-3.89427453e-01 7.17683077e-01 -4.77241158e-01 -6.67137682e-01
-7.69274831e-01 -9.39238846e-01 1.13675714e+00 3.92910033e-01
-5.90936877e-02 -8.35095286e-01 -2.89426446e-01 4.54253614e-01
-4.88648564e-02 1.12549707e-01 4.25340563e-01 -6.30771875e-01
-7.48601437e-01 1.80613562e-01 -4.72977400e-01 3.07112426e-01
2.23082244e-01 2.33808666e-01 -7.77761161e-01 -4.49060172e-01
-4.59772825e-01 -1.51122332e-01 1.34554136e+00 2.19412178e-01
1.54175055e+00 -2.82365642e-02 -5.95947325e-01 5.88508308e-01
1.38460290e+00 -2.74181336e-01 2.41331652e-01 1.28171608e-01
1.01421666e+00 4.41777050e-01 9.32577074e-01 5.22935867e-01
4.26218241e-01 3.44468266e-01 4.20228720e-01 -2.74102122e-01
-3.02280217e-01 -2.37191021e-01 1.66274592e-01 1.18254352e+00
-1.30202860e-01 -1.45275835e-02 -4.71665949e-01 1.02321208e+00
-2.58616185e+00 -7.87260711e-01 -3.61946553e-01 1.90807283e+00
7.35959888e-01 1.87657759e-01 2.66101509e-01 2.87725896e-01
9.28636253e-01 2.78491378e-01 -4.51003879e-01 1.81050330e-01
-2.73349553e-01 -1.55599937e-01 6.07737243e-01 1.58521473e-01
-1.61650097e+00 1.04021180e+00 4.99956179e+00 1.08002949e+00
-7.48443842e-01 2.32631013e-01 6.91549182e-01 2.06747159e-01
-1.26495630e-01 -1.67723149e-01 -8.80972505e-01 6.67925596e-01
1.76098838e-01 -1.93582401e-01 -1.20437331e-01 1.07265997e+00
-2.71267891e-02 -1.43118098e-01 -1.11010432e+00 1.25599515e+00
3.07977825e-01 -1.66212535e+00 -3.13781500e-02 -1.43627912e-01
1.21220911e+00 1.41533643e-01 2.79502660e-01 1.05720110e-01
4.25908774e-01 -4.36441332e-01 8.50504994e-01 3.30995589e-01
1.79766029e-01 -7.44344354e-01 5.12405634e-01 2.16163605e-01
-1.31830013e+00 2.02216930e-03 -2.92455107e-01 4.03774232e-01
-1.02604933e-01 9.06306803e-01 -5.38809359e-01 8.07759166e-01
9.98225093e-01 1.27827120e+00 -8.58546138e-01 1.29412436e+00
-3.09069976e-02 4.88311619e-01 -4.08919662e-01 -7.93172717e-02
3.92689049e-01 -1.32250071e-01 6.37796760e-01 1.41300821e+00
-2.38074377e-01 -1.91260278e-01 5.01681447e-01 8.85829806e-01
-1.22055724e-01 3.01774591e-01 3.62066403e-02 1.96814731e-01
5.95649302e-01 1.34634078e+00 -1.29333007e+00 -5.78292847e-01
-2.57968426e-01 1.34692729e+00 3.09250027e-01 1.32434785e-01
-9.07405138e-01 -3.47974300e-01 5.08357584e-01 -4.21370864e-01
7.62224197e-01 -2.92566150e-01 -5.41091621e-01 -1.29483914e+00
1.25584215e-01 -6.59842074e-01 6.35878623e-01 -3.60251158e-01
-1.53220129e+00 1.09848119e-01 -1.89774469e-01 -1.32293701e+00
1.80080593e-01 -2.05445856e-01 -7.04364896e-01 3.91269982e-01
-1.50218785e+00 -1.35561895e+00 -4.19994205e-01 6.98169827e-01
8.32365334e-01 1.43655604e-02 3.39549661e-01 3.48431408e-01
-7.84162402e-01 7.19550431e-01 2.98885733e-01 8.38836581e-02
1.00899577e+00 -1.44776809e+00 2.67459899e-01 8.43967378e-01
6.26202166e-01 3.86550426e-01 5.19736767e-01 -6.67441249e-01
-1.23722184e+00 -1.27936840e+00 6.81668520e-01 -3.43441457e-01
7.91504323e-01 -5.25463402e-01 -9.45307553e-01 4.54513252e-01
1.41152591e-01 4.33439344e-01 6.06233716e-01 1.87911659e-01
-5.25580823e-01 -2.16661558e-01 -7.40571082e-01 5.92144728e-01
1.42897856e+00 -3.59558463e-01 -6.26069307e-01 5.23874640e-01
6.90642357e-01 -9.41241235e-02 -6.91104770e-01 3.70667666e-01
3.15065026e-01 -1.01541388e+00 1.01970494e+00 -3.58471543e-01
3.01401198e-01 -6.08798027e-01 -3.65432464e-02 -1.20125723e+00
-7.38573611e-01 -5.61590672e-01 -2.69461602e-01 1.45915639e+00
2.04222903e-01 -2.33037636e-01 9.79608595e-01 2.25002185e-01
-1.92213908e-01 -9.62138593e-01 -6.99908614e-01 -7.79072642e-01
-3.07996333e-01 -2.71666378e-01 2.58504838e-01 1.10183918e+00
-3.93407159e-02 1.86040238e-01 -3.07545841e-01 3.82192552e-01
1.05159807e+00 6.24544680e-01 7.41346836e-01 -1.24855638e+00
-2.91795611e-01 -5.44871986e-01 -4.07998443e-01 -1.16420615e+00
-2.13619694e-02 -9.84558523e-01 -3.00550573e-02 -1.50612414e+00
4.57406104e-01 -6.20075703e-01 -6.71147823e-01 4.11552161e-01
-5.37609220e-01 3.17455590e-01 4.35263485e-01 5.12169480e-01
-1.62020063e+00 6.02945209e-01 1.08394718e+00 -3.99149805e-01
-2.21793771e-01 -2.70004302e-01 -8.38044822e-01 8.17552984e-01
4.03747350e-01 -1.94682315e-01 -2.83803463e-01 -4.68807548e-01
-1.51339918e-01 -8.49521756e-01 3.09344530e-01 -9.88889813e-01
7.16026664e-01 -1.06188044e-01 4.51634496e-01 -7.73708284e-01
-6.01348914e-02 -8.02251637e-01 -1.81309149e-01 2.22267672e-01
-3.23160768e-01 -1.66441992e-01 -6.67225122e-02 1.00607550e+00
-2.78358757e-01 1.27031997e-01 8.62174630e-01 3.03877834e-02
-1.27387524e+00 3.31120223e-01 2.92402282e-02 -9.76466388e-02
1.31485927e+00 -2.54721820e-01 -1.68645546e-01 -7.05667213e-02
-8.14759374e-01 8.50591540e-01 3.16754818e-01 6.66759193e-01
6.66617572e-01 -1.30108702e+00 -6.31232440e-01 1.34446010e-01
3.47983211e-01 1.23705581e-01 3.37586999e-01 9.55867231e-01
-8.83160383e-02 -1.68811139e-02 3.10462210e-02 -1.00356972e+00
-1.67357266e+00 5.09507895e-01 -8.43219608e-02 -1.73916191e-01
-8.62089276e-01 1.09988081e+00 1.97628707e-01 -3.72524722e-03
5.38585663e-01 -2.44700119e-01 -3.00805598e-01 3.41366917e-01
4.65291917e-01 5.37967443e-01 7.97790736e-02 -5.06255984e-01
-4.91211891e-01 6.14056468e-01 -4.61734653e-01 4.21542853e-01
1.27669334e+00 -4.66700792e-01 -2.20548213e-02 4.39573020e-01
1.16931665e+00 -7.28169829e-02 -1.34139860e+00 -5.35856485e-01
2.30186388e-01 -5.82262218e-01 2.46565685e-01 -4.92772639e-01
-1.31604326e+00 4.16681319e-01 8.09557974e-01 -1.19318165e-01
9.35994565e-01 4.35488492e-01 5.40259361e-01 2.49268085e-01
3.30521584e-01 -1.67444956e+00 4.34634656e-01 4.50971983e-02
9.33054924e-01 -1.50598943e+00 2.05473006e-01 -8.31172407e-01
-7.94539750e-01 7.27014959e-01 6.37475431e-01 -4.33827639e-01
7.99207270e-01 -6.55624121e-02 -1.50954902e-01 -1.43744990e-01
-7.48874605e-01 -4.44501758e-01 5.14654696e-01 3.58323127e-01
1.24932952e-01 2.13412911e-01 -4.53651547e-01 6.69130147e-01
2.17108428e-01 -3.71481389e-01 1.51427194e-01 7.72054851e-01
-5.25479138e-01 -7.58035660e-01 -2.52525330e-01 5.74212849e-01
-2.81237990e-01 1.80891201e-01 -6.06141925e-01 6.16986096e-01
4.60490853e-01 9.55662966e-01 2.67431885e-01 -4.67097104e-01
-5.22619039e-02 -2.25760564e-01 2.26809278e-01 -6.33603036e-01
-5.37092090e-01 4.28333551e-01 -1.91870704e-01 -6.22423172e-01
-7.87902534e-01 -8.95402193e-01 -1.21412659e+00 3.86379473e-02
-7.78533876e-01 2.25606769e-01 3.77115309e-01 6.89283073e-01
6.65885568e-01 4.33136135e-01 7.09085345e-01 -9.61464286e-01
-3.52035463e-01 -8.09063494e-01 -6.71535671e-01 6.97806358e-01
1.81006730e-01 -5.57543159e-01 -5.20435035e-01 9.01522115e-02] | [9.730093955993652, -0.05714463070034981] |
d4cbb30c-e91d-40ab-925b-16bc8160ed38 | penalizing-the-hard-example-but-not-too-much | null | null | https://ieeexplore.ieee.org/abstract/document/9956020 | https://ieeexplore.ieee.org/abstract/document/9956020 | Penalizing the Hard Example But Not Too Much: A Strong Baseline for Fine-Grained Visual Classification | Though significant progress has been achieved on fine-grained visual classification (FGVC), severe overfitting still hinders model generalization. A recent study shows that hard samples in the training set can be easily fit, but most existing FGVC methods fail to classify some hard examples in the test set. The reason is that the model overfits those hard examples in the training set, but does not learn to generalize to unseen examples in the test set. In this article, we propose a moderate hard example modulation (MHEM) strategy to properly modulate the hard examples. MHEM encourages the model to not overfit hard examples and offers better generalization and discrimination. First, we introduce three conditions and formulate a general form of a modulated loss function. Second, we instantiate the loss function and provide a strong baseline for FGVC, where the performance of a naive backbone can be boosted and be comparable with recent methods. Moreover, we demonstrate that our baseline can be readily incorporated into the existing methods and empower these methods to be more discriminative. Equipped with our strong baseline, we achieve consistent improvements on three typical FGVC datasets, i.e., CUB-200-2011, Stanford Cars, and FGVC-Aircraft. We hope the idea of moderate hard example modulation will inspire future research work toward more effective fine-grained visual recognition. | ['Yi Yang', 'Xiaohan Wang', 'Linchao Zhu', 'Yuanzhi Liang'] | 2022-11-21 | null | null | null | ieee-transactions-on-neural-networks-and-8 | ['fine-grained-visual-recognition', 'fine-grained-image-classification'] | ['computer-vision', 'computer-vision'] | [ 1.75930440e-01 -1.36549711e-01 -4.64335173e-01 -5.51677167e-01
-7.24164784e-01 -7.53131032e-01 6.03594184e-01 -4.78824049e-01
-2.39068553e-01 7.92615294e-01 -4.35164720e-02 -1.21111661e-01
7.43816197e-02 -6.08659804e-01 -7.21240401e-01 -6.24480844e-01
3.69250208e-01 4.40534204e-01 3.64075214e-01 -9.03781727e-02
-8.98954868e-02 5.21504164e-01 -1.59911144e+00 6.38134658e-01
9.00412321e-01 1.25396478e+00 2.37910509e-01 5.18492579e-01
1.92133754e-01 5.24646580e-01 -7.81480491e-01 -5.29384077e-01
2.76962608e-01 -2.18366221e-01 -8.03916037e-01 2.14631289e-01
1.01833427e+00 -2.49916777e-01 -3.16140860e-01 8.84492517e-01
4.24353957e-01 1.11584745e-01 9.88466144e-01 -1.42560148e+00
-9.38806772e-01 1.00316860e-01 -3.97139817e-01 1.37451693e-01
-1.31717131e-01 3.61647159e-01 1.03682625e+00 -1.13708603e+00
5.00456750e-01 1.30998755e+00 6.70560300e-01 8.18987310e-01
-1.17722738e+00 -7.03563750e-01 5.17221153e-01 4.12387729e-01
-1.45331347e+00 -2.85444587e-01 5.60184121e-01 -4.23108727e-01
7.55042911e-01 5.30431867e-01 4.54850584e-01 1.37012172e+00
-1.27182990e-01 9.68781948e-01 1.20287097e+00 -3.16663980e-01
1.58312112e-01 2.02920437e-01 3.20120692e-01 5.88451922e-01
2.72110552e-01 5.23009181e-01 -6.88363239e-02 3.59310098e-02
5.72432041e-01 8.69705230e-02 -4.45445985e-01 -4.62028205e-01
-6.93272412e-01 8.89177144e-01 7.79401779e-01 1.84423149e-01
-1.05692841e-01 1.87634323e-02 2.35191703e-01 2.10162073e-01
3.43209118e-01 6.65078580e-01 -3.66004288e-01 1.33789152e-01
-9.67733979e-01 2.06726640e-01 5.37418544e-01 9.33065712e-01
8.16311061e-01 1.41634092e-01 -6.31157517e-01 1.08482647e+00
1.15906268e-01 5.13195276e-01 4.37187791e-01 -9.73863304e-01
3.98286253e-01 4.39513534e-01 -1.25914350e-01 -8.11162293e-01
-7.18610436e-02 -9.23987389e-01 -8.44056964e-01 2.81123966e-01
2.42806718e-01 2.26484746e-01 -1.19009256e+00 1.73172021e+00
2.13993695e-02 3.07790905e-01 -2.01801240e-01 1.00991428e+00
8.68747652e-01 5.19823909e-01 5.98734766e-02 -3.57173309e-02
1.06068277e+00 -1.37043047e+00 -3.54277343e-01 -5.03310978e-01
3.81947458e-01 -4.66310114e-01 1.57865822e+00 5.28648496e-01
-8.03699553e-01 -9.10438716e-01 -1.19278049e+00 1.52514726e-01
-4.54741925e-01 1.67819709e-01 8.02288234e-01 6.92151964e-01
-9.28835094e-01 5.58659494e-01 -5.85256398e-01 -3.60940099e-02
7.88291276e-01 1.74657241e-01 -3.04799944e-01 -4.66770619e-01
-9.56439912e-01 6.86474621e-01 3.91222566e-01 9.32411477e-02
-1.12373328e+00 -6.47791266e-01 -7.62274802e-01 4.46011201e-02
3.20982903e-01 -7.68864095e-01 1.15891838e+00 -1.00739551e+00
-1.03810656e+00 8.94711673e-01 -1.75747052e-02 -3.72645706e-01
6.61170840e-01 -7.34283626e-02 -3.75950694e-01 9.80494469e-02
3.60641107e-02 8.99499118e-01 1.13069391e+00 -1.62324095e+00
-5.40028214e-01 -1.57223403e-01 1.11636870e-01 3.30367684e-02
-5.44455707e-01 -3.54461282e-01 -7.60178804e-01 -8.63429010e-01
-3.83931130e-01 -9.06800807e-01 -1.30905077e-01 5.87074794e-02
-4.29201066e-01 -2.67771035e-01 8.23590696e-01 -1.25599399e-01
1.31801534e+00 -2.25985551e+00 3.25904451e-02 8.58002305e-02
3.38265777e-01 7.22421408e-01 -4.73382175e-01 7.31535256e-02
-1.19559290e-02 2.89969146e-01 -9.82764140e-02 -3.85747612e-01
2.03804776e-01 4.64693785e-01 -5.70836306e-01 1.41572103e-01
3.78339112e-01 1.20537972e+00 -6.49685502e-01 -2.45497033e-01
1.66719586e-01 3.22012395e-01 -7.85489321e-01 2.19004303e-01
-2.08288699e-01 1.51231363e-01 -3.45056713e-01 7.53867209e-01
7.96372950e-01 -6.84965074e-01 -9.75896269e-02 -4.64380234e-01
3.15341145e-01 -1.43408298e-01 -8.98949027e-01 9.63023901e-01
-2.73264349e-01 4.32689637e-01 -1.40314206e-01 -1.06727993e+00
7.32994854e-01 -1.46854267e-01 -2.31126249e-01 -7.83313990e-01
-6.60155118e-02 7.37947598e-02 -1.06497876e-01 -2.99546808e-01
3.90622288e-01 -1.84628814e-01 -5.70870638e-02 -1.94934420e-02
4.12780717e-02 -2.53935933e-01 1.03880815e-01 -1.48298976e-03
7.36277580e-01 4.58880626e-02 3.00113708e-01 5.68305142e-02
3.21764499e-01 -5.08244224e-02 5.22417367e-01 1.14667618e+00
-5.07682979e-01 7.21917987e-01 3.26859802e-02 -4.45686042e-01
-8.42439115e-01 -1.13704872e+00 -3.38182092e-01 1.32874560e+00
1.52504012e-01 -3.83993119e-01 -6.42290235e-01 -1.23193610e+00
2.32238680e-01 5.92881322e-01 -7.63084233e-01 -3.99609834e-01
-3.77860188e-01 -8.17023754e-01 5.84208727e-01 9.70784962e-01
6.54114425e-01 -8.00810635e-01 1.44726261e-02 -1.26382902e-01
-1.24717792e-02 -1.11499226e+00 -4.94926155e-01 2.71083444e-01
-6.52554929e-01 -9.40441430e-01 -5.51248014e-01 -8.22427332e-01
5.79159439e-01 4.26774770e-01 1.37814248e+00 4.38728392e-01
-4.03691560e-01 1.91889107e-01 -3.03688258e-01 -3.06791186e-01
-1.98541895e-01 -4.64289896e-02 -1.55921401e-02 -6.78878650e-02
3.40835243e-01 -2.01471522e-01 -5.97009122e-01 6.33989811e-01
-6.58398211e-01 -3.71155851e-02 6.09773636e-01 1.33702803e+00
8.52226079e-01 4.94298749e-02 6.97967410e-01 -9.93739545e-01
2.90057063e-01 -3.19819182e-01 -3.04635108e-01 2.99408734e-01
-7.14960873e-01 -1.40656397e-01 8.74132454e-01 -6.50455177e-01
-7.80442238e-01 -1.22068703e-01 -3.66616696e-01 -9.03785765e-01
-2.98814476e-01 2.73208290e-01 -2.25925699e-01 -3.83591920e-01
7.51358926e-01 1.57810107e-01 -4.35067028e-01 -4.56463486e-01
4.62410510e-01 6.43806100e-01 5.87524295e-01 -6.90436482e-01
8.65227520e-01 2.75418222e-01 -2.66422302e-01 -6.82147264e-01
-1.16012275e+00 -3.11264217e-01 -3.45368087e-01 1.72401480e-02
5.08449554e-01 -9.87983584e-01 -4.13904816e-01 1.70431480e-01
-6.18516386e-01 -6.31903470e-01 -3.06496352e-01 1.41985863e-01
-4.69748139e-01 4.02040154e-01 -5.70419312e-01 -5.73988259e-01
4.80726063e-02 -1.20906425e+00 1.30819738e+00 1.73117429e-01
-1.30203307e-01 -1.01674616e+00 -3.10117304e-01 5.27453184e-01
4.13546056e-01 -6.45553321e-02 8.86086702e-01 -5.40861845e-01
-6.18047118e-01 -1.93460047e-01 -4.57314461e-01 7.40040898e-01
-4.92314287e-02 1.19384795e-01 -1.18856037e+00 -6.73811615e-01
-3.32450449e-01 -8.69045436e-01 1.43401575e+00 2.47884005e-01
1.65392303e+00 -3.49908233e-01 -5.43373168e-01 9.65264261e-01
1.28585267e+00 -5.03694452e-02 6.97883487e-01 1.74026370e-01
7.92419434e-01 2.37256810e-01 7.43965149e-01 8.98035169e-02
3.44141603e-01 9.32500780e-01 4.95408833e-01 -3.80929381e-01
-5.50730646e-01 -2.94984579e-01 9.79804397e-02 5.57151496e-01
-1.72249407e-01 -3.56931061e-01 -6.17775559e-01 3.22618634e-01
-1.60947335e+00 -9.58496332e-01 1.35223746e-01 1.91199446e+00
8.90577734e-01 1.97081760e-01 1.10826112e-01 1.88163862e-01
5.33009291e-01 1.49444446e-01 -5.54696798e-01 -2.04655230e-01
-3.06480646e-01 4.09625024e-01 2.50917375e-01 4.94688272e-01
-1.37902975e+00 1.11938095e+00 7.05664062e+00 1.30917943e+00
-1.25759280e+00 6.78678006e-02 8.69555473e-01 -7.16198161e-02
-2.89301544e-01 -2.73679972e-01 -1.20468235e+00 6.15938306e-01
5.56955814e-01 2.64993697e-01 4.19066995e-01 1.14937806e+00
-2.29257658e-01 4.29901958e-01 -1.21304095e+00 1.02943420e+00
1.05772771e-01 -1.35217381e+00 3.30991894e-01 -1.45806829e-02
8.83516729e-01 -1.29751801e-01 4.48111176e-01 9.66831625e-01
1.51053429e-01 -1.34477127e+00 5.89133382e-01 3.32026094e-01
1.16799116e+00 -4.90336448e-01 6.23670876e-01 4.09469634e-01
-1.12377048e+00 -2.81574726e-01 -7.28573978e-01 5.58738373e-02
-2.95627624e-01 3.48360568e-01 -8.07354569e-01 4.71844524e-01
6.71251595e-01 6.59455717e-01 -9.80086863e-01 1.11126482e+00
-2.77524680e-01 7.93148160e-01 -1.83165446e-02 -5.08833230e-02
3.48995209e-01 2.22445279e-01 2.43422508e-01 1.32251132e+00
6.31981492e-02 -1.36895314e-01 5.55843115e-01 8.07308435e-01
-1.70597419e-01 -9.39280763e-02 -4.08225149e-01 -1.91882312e-01
4.23140436e-01 1.14102519e+00 -3.48048687e-01 -4.61720407e-01
-5.22751331e-01 1.07209814e+00 7.71475434e-01 6.23249948e-01
-9.22350466e-01 -1.28726617e-01 5.49971700e-01 -1.36407297e-02
6.40882671e-01 1.25484049e-01 -2.61667699e-01 -1.34865427e+00
-6.55555576e-02 -1.22681201e+00 5.11158168e-01 -7.24846363e-01
-1.70655668e+00 7.69289851e-01 -1.42321125e-01 -1.24498022e+00
-2.07107976e-01 -1.03726387e+00 -5.26104271e-01 7.42816150e-01
-1.66331410e+00 -1.25033176e+00 -5.67220032e-01 6.66700006e-01
6.05091333e-01 -2.43527651e-01 7.71478474e-01 3.58301550e-01
-5.23566961e-01 1.22557724e+00 -1.12775072e-01 1.05917603e-02
8.57336581e-01 -1.24099100e+00 2.14524657e-01 6.72958553e-01
2.38854572e-01 6.49329603e-01 5.12946963e-01 -5.76338112e-01
-1.00703108e+00 -1.55846727e+00 4.81411994e-01 -5.66281497e-01
4.83422518e-01 -4.28315133e-01 -1.00576341e+00 7.48162091e-01
-1.15182579e-01 4.98378873e-01 5.60208738e-01 1.23179786e-01
-7.60701299e-01 -1.50879353e-01 -1.04823124e+00 5.61042011e-01
1.18262863e+00 -5.16692519e-01 -6.64716363e-01 3.27848643e-01
6.64958656e-01 -2.57398605e-01 -6.19749427e-01 7.66777754e-01
4.94300693e-01 -9.01239395e-01 1.25412226e+00 -9.43512440e-01
2.14855328e-01 -2.91763544e-01 -4.35978532e-01 -1.50595236e+00
-7.71042705e-01 -1.62137091e-01 -4.10322666e-01 1.03653097e+00
3.59679431e-01 -5.85044682e-01 8.98074090e-01 2.53251910e-01
-4.16305423e-01 -1.06450355e+00 -6.69443071e-01 -1.11747038e+00
3.83077383e-01 -4.51806903e-01 4.96911466e-01 8.03477228e-01
-5.21930218e-01 3.28952342e-01 -5.14006913e-01 1.13787428e-01
5.97725570e-01 3.20375323e-01 7.95561969e-01 -1.24745178e+00
-7.93712318e-01 -5.44733524e-01 -4.33575183e-01 -1.39957094e+00
1.20120287e-01 -9.52639878e-01 1.49725014e-02 -1.29226303e+00
4.99911577e-01 -6.76673770e-01 -3.66821349e-01 7.10256875e-01
-5.73651969e-01 8.66228402e-01 4.05258864e-01 2.02592686e-01
-8.06594908e-01 6.17427289e-01 1.67242265e+00 -3.80275428e-01
1.70380622e-01 3.88485119e-02 -9.62581992e-01 5.55459142e-01
5.35478354e-01 -1.73038974e-01 -4.28654373e-01 -7.72870407e-02
-2.27226838e-01 -3.70786279e-01 6.03215635e-01 -8.72406065e-01
-1.14396133e-01 -2.61397243e-01 8.82833958e-01 -4.33312148e-01
4.31435436e-01 -5.96061707e-01 -5.46802543e-02 2.75555342e-01
-2.78524041e-01 -4.55049217e-01 2.74838567e-01 6.37547195e-01
-2.86404997e-01 -2.05368236e-01 1.07010067e+00 1.41610047e-02
-9.20816839e-01 6.07841313e-01 3.89248282e-02 3.84471893e-01
9.75180149e-01 -3.14567983e-01 -7.56131053e-01 -1.93752080e-01
-7.39346385e-01 3.20803523e-01 6.12527251e-01 4.77151394e-01
6.57447517e-01 -1.64785886e+00 -5.98237455e-01 4.16857690e-01
3.73957992e-01 -3.92112076e-01 3.12376529e-01 6.36180282e-01
-6.28445223e-02 4.31848764e-01 -1.25442460e-01 -7.73941576e-01
-1.10326362e+00 8.66958439e-01 3.64178151e-01 -2.60355055e-01
-3.65996867e-01 1.05164492e+00 8.26079726e-01 -3.25093806e-01
4.37140524e-01 -2.15034142e-01 -1.93939909e-01 -1.83913440e-01
7.65390038e-01 5.94123118e-02 8.11434165e-02 -4.77824241e-01
-4.38368350e-01 6.30999863e-01 -3.46798897e-01 5.40944517e-01
1.12763774e+00 5.44234528e-04 4.68826294e-01 1.86954394e-01
1.31078660e+00 -3.90832797e-02 -1.56486869e+00 -1.55079603e-01
-3.29091549e-01 -6.24490082e-01 -1.09283753e-01 -1.02066159e+00
-1.12737525e+00 9.80634630e-01 6.42888486e-01 2.80568123e-01
1.24192142e+00 1.36357009e-01 4.81268585e-01 4.53626841e-01
3.21934730e-01 -8.38096619e-01 3.41938853e-01 5.93118846e-01
9.73814249e-01 -1.41100287e+00 -2.05137283e-01 -6.21864080e-01
-7.50568271e-01 9.27266419e-01 8.71453106e-01 -1.47942543e-01
3.73482257e-01 3.43176544e-01 1.98261179e-02 -4.97034155e-02
-8.49248230e-01 -3.05540442e-01 7.43650079e-01 8.21727812e-01
2.49724776e-01 1.32771611e-01 4.44383509e-02 9.19278204e-01
-1.42098650e-01 -2.10873876e-02 6.96059912e-02 4.53352392e-01
-5.61060786e-01 -9.87840950e-01 -2.50136375e-01 5.60157716e-01
-2.31982633e-01 -2.28854284e-01 -4.29168075e-01 8.88922751e-01
3.95141989e-01 8.72833014e-01 -1.07284086e-02 -5.89811862e-01
5.17544806e-01 4.08866722e-03 6.59566998e-01 -7.44880319e-01
-3.77024680e-01 -1.18818861e-02 9.85898599e-02 -7.34069467e-01
-1.00385129e-01 -2.33442977e-01 -7.93289125e-01 -1.26738474e-01
-2.95089602e-01 6.61815256e-02 1.24906339e-02 8.73867631e-01
3.38227004e-01 6.03328586e-01 5.94233274e-01 -9.29628313e-01
-8.62204790e-01 -7.26236820e-01 -4.40985769e-01 5.56112945e-01
4.64806199e-01 -1.08652890e+00 -5.28791070e-01 -6.75708801e-02] | [9.643061637878418, 2.1889684200286865] |
468e8e5a-2b7a-4981-bae9-997bf5e6dc24 | magnushammer-a-transformer-based-approach-to | 2303.04488 | null | https://arxiv.org/abs/2303.04488v1 | https://arxiv.org/pdf/2303.04488v1.pdf | Magnushammer: A Transformer-based Approach to Premise Selection | Premise selection is a fundamental problem of automated theorem proving. Previous works often use intricate symbolic methods, rely on domain knowledge, and require significant engineering effort to solve this task. In this work, we show that Magnushammer, a neural transformer-based approach, can outperform traditional symbolic systems by a large margin. Tested on the PISA benchmark, Magnushammer achieves $59.5\%$ proof rate compared to a $38.3\%$ proof rate of Sledgehammer, the most mature and popular symbolic-based solver. Furthermore, by combining Magnushammer with a neural formal prover based on a language model, we significantly improve the previous state-of-the-art proof rate from $57.0\%$ to $71.0\%$. | ['Yuhuai Wu', 'Piotr Miłoś', 'Łukasz Kuciński', 'Christian Szegedy', 'Jin Peng Zhou', 'Albert Qiaochu Jiang', 'Szymon Tworkowski', 'Szymon Antoniak', 'Maciej Mikuła'] | 2023-03-08 | null | null | null | null | ['automated-theorem-proving', 'automated-theorem-proving'] | ['miscellaneous', 'reasoning'] | [ 7.75638297e-02 2.94943571e-01 -3.38359326e-01 1.58057854e-01
-9.73105371e-01 -6.87058508e-01 2.31568530e-01 1.67198882e-01
-9.88588929e-02 1.25159824e+00 -8.55316341e-01 -1.16613615e+00
-2.44905636e-01 -1.15994060e+00 -1.21634960e+00 1.15369130e-02
-3.53679568e-01 4.27210569e-01 4.17597920e-01 -4.08295423e-01
2.06692591e-01 9.01587382e-02 -1.48767090e+00 1.90956533e-01
1.04881024e+00 1.05847526e+00 -5.17737389e-01 5.99284887e-01
-1.06261827e-01 9.28958476e-01 -7.14012682e-01 -3.79784018e-01
2.62784332e-01 -4.32352245e-01 -8.54066849e-01 -9.21851337e-01
5.73415577e-01 -3.50363880e-01 -1.60983726e-01 1.39936090e+00
-3.08091380e-02 -2.33813480e-01 9.40883085e-02 -1.68568575e+00
-1.25854695e-02 1.32813776e+00 -5.18763661e-01 -4.09768261e-02
4.63935912e-01 6.95009679e-02 1.06839204e+00 -4.49547529e-01
6.03739321e-01 9.95842159e-01 8.62277448e-01 4.25956398e-01
-1.45457232e+00 -1.38152003e+00 3.63684967e-02 3.74001682e-01
-1.85997295e+00 -1.64405107e-01 6.09878659e-01 -2.15735421e-01
1.54246962e+00 1.66138783e-01 5.39043307e-01 4.83191967e-01
3.79455954e-01 3.75035524e-01 1.03366351e+00 -5.00631988e-01
1.09737091e-01 -4.18275706e-02 1.45778120e-01 1.30063605e+00
6.49930060e-01 1.16543464e-01 -3.36056799e-01 -4.50379878e-01
5.46431959e-01 -4.60836053e-01 -9.34171453e-02 -3.03857643e-02
-9.38346028e-01 8.97196054e-01 1.57039035e-02 8.54468346e-02
1.27642274e-01 7.20000923e-01 6.47888720e-01 3.74625534e-01
-1.25967219e-01 7.96580732e-01 -4.87296104e-01 -3.38081717e-01
-1.33825886e+00 7.42690206e-01 1.20438588e+00 8.32905829e-01
5.21290600e-01 4.52919394e-01 -3.66824656e-03 1.07546479e-01
1.29036248e-01 8.75954151e-01 -1.27516076e-01 -1.26216161e+00
5.22276521e-01 8.11422467e-01 6.69124946e-02 -1.04188108e+00
-1.23002827e-01 -5.68499982e-01 -6.92718029e-01 3.85067940e-01
5.18684030e-01 -3.00873935e-01 -5.74857175e-01 1.85915399e+00
1.02914430e-01 2.67639488e-01 1.72734275e-01 4.14830804e-01
4.48179364e-01 9.28707242e-01 -4.31634963e-01 -2.42537767e-01
9.05393362e-01 -7.47037232e-01 -4.21974093e-01 -2.44667642e-02
8.94520521e-01 -4.02848333e-01 6.06641769e-01 1.15583909e+00
-1.60114539e+00 5.88620044e-02 -1.49012113e+00 1.17169626e-01
-1.42115042e-01 7.25615621e-02 8.19887221e-01 3.63473028e-01
-1.06301510e+00 5.81264853e-01 -7.26743817e-01 4.89536047e-01
1.98398754e-01 7.38203049e-01 -4.52745594e-02 -2.97206372e-01
-1.31355464e+00 8.35236430e-01 5.26500642e-01 8.66867676e-02
-9.91155148e-01 -1.01214468e+00 -1.08221400e+00 2.54958600e-01
1.13881505e+00 -4.32144076e-01 1.78709090e+00 -2.97583610e-01
-1.50234985e+00 1.97171748e-01 -1.13662675e-01 -1.13633311e+00
4.01792735e-01 -4.82041463e-02 -3.69432062e-01 1.66384116e-01
9.50655267e-02 3.34716111e-01 2.64074266e-01 -8.57097685e-01
-4.91572022e-01 2.87366897e-01 5.36923885e-01 -7.96944320e-01
2.75590062e-01 1.59724727e-02 7.79177099e-02 -2.32266068e-01
-1.66063845e-01 -1.00602245e+00 -1.65751845e-01 -4.06793766e-02
-4.83215153e-01 -4.15587902e-01 5.88622212e-01 -4.73683953e-01
1.38023400e+00 -1.75662708e+00 1.22708641e-01 6.80756211e-01
4.84924167e-01 6.15418851e-01 2.42679358e-01 3.64335418e-01
7.96912313e-02 1.60435170e-01 -2.05266923e-01 2.34262824e-01
5.89915216e-01 7.18771992e-03 -4.19846535e-01 2.42266998e-01
4.02497947e-01 9.44375873e-01 -9.84602928e-01 -7.53308654e-01
1.44743584e-02 -8.94511177e-05 -8.79111409e-01 -2.53522515e-01
-8.17097604e-01 -5.76309323e-01 -3.17978740e-01 7.18540728e-01
5.51030397e-01 -4.65521932e-01 5.88477552e-01 2.09639639e-01
-1.04498260e-01 4.59833443e-01 -1.29494333e+00 1.58837247e+00
-5.15497923e-01 7.51094520e-01 3.18177700e-01 -9.67321157e-01
7.46918976e-01 2.64820635e-01 -2.70612333e-02 -7.48872340e-01
4.06166315e-01 5.92429101e-01 3.20375830e-01 1.41807586e-01
4.35480058e-01 -2.27776259e-01 -3.49087983e-01 5.95904887e-01
-9.82074961e-02 -3.95124614e-01 8.36740971e-01 4.41876441e-01
1.45013857e+00 2.15941861e-01 1.94797978e-01 -3.29655945e-01
6.45164132e-01 4.67252284e-01 7.65604079e-01 7.25757539e-01
3.31914186e-01 -6.41972795e-02 1.20886898e+00 -3.08974326e-01
-7.25392103e-01 -7.11906195e-01 1.67055741e-01 5.46832919e-01
-7.08587095e-02 -9.01855290e-01 -8.43189001e-01 -5.97541451e-01
9.60888416e-02 1.08734608e+00 -3.23972076e-01 -4.40620989e-01
-7.49051571e-01 1.02797210e-01 1.25605941e+00 8.40076804e-01
5.09661317e-01 -8.16340685e-01 -6.52902603e-01 2.70480275e-01
-5.39043285e-02 -1.03774381e+00 2.14070782e-01 2.35379025e-01
-5.46452284e-01 -1.20714390e+00 -1.53341889e-01 -4.85873848e-01
5.13406515e-01 -2.02290341e-01 1.10114336e+00 3.44083875e-01
-1.66039973e-01 -3.88502747e-01 -9.67169851e-02 -4.71444964e-01
-7.03866184e-01 1.80063978e-01 3.86924855e-02 -8.10048878e-01
-3.43664922e-02 -5.10128617e-01 1.46839783e-01 8.98010209e-02
-7.21581578e-01 1.02421112e-01 4.90481347e-01 9.39053118e-01
4.76248503e-01 5.30389667e-01 3.51488471e-01 -8.75155509e-01
4.04537708e-01 -8.50619003e-02 -1.55149186e+00 2.98853427e-01
-1.02674830e+00 3.74883473e-01 1.18005168e+00 -3.65944743e-01
-4.16811049e-01 -2.60751426e-01 1.96267098e-01 -7.66392469e-01
4.02110249e-01 8.55699718e-01 2.52435595e-01 -2.62601763e-01
8.21518123e-01 1.25379279e-01 4.41782847e-02 2.19078019e-01
-4.17183526e-03 -3.75352539e-02 7.58158505e-01 -1.07634902e+00
1.19024670e+00 -6.12342730e-02 5.47457159e-01 9.44485217e-02
-7.22797513e-01 3.12789351e-01 2.51033664e-01 1.61074102e-01
-5.66101074e-03 -5.32594681e-01 -1.39180219e+00 1.13280170e-01
-1.19040751e+00 -8.48059177e-01 -1.76927596e-01 1.98646590e-01
-2.63744593e-01 1.33648336e-01 -5.23039997e-01 -9.73490953e-01
-5.47616720e-01 -1.21014500e+00 8.21719110e-01 3.06317303e-02
-6.22319937e-01 -4.41035986e-01 -7.46644437e-02 1.27894416e-01
3.36944968e-01 6.07910633e-01 1.08867359e+00 -4.69723374e-01
-6.51560724e-01 -3.96412224e-01 -5.62460184e-01 5.15112579e-01
-4.07072425e-01 3.44893932e-01 -3.84733379e-01 2.57519595e-02
-4.20832455e-01 -3.73003691e-01 3.79448205e-01 1.68445170e-01
1.04320049e+00 -6.67768598e-01 -3.49473357e-01 2.03023508e-01
1.36182415e+00 1.64864898e-01 4.88184035e-01 3.17774594e-01
2.25363791e-01 -5.11210002e-02 5.96348405e-01 3.63526791e-01
2.44445235e-01 4.06640321e-01 4.41081911e-01 3.35756034e-01
2.28451863e-01 -3.17077875e-01 3.28746706e-01 2.83939451e-01
-1.23021610e-01 1.65351659e-01 -1.39640450e+00 4.26750898e-01
-1.86229336e+00 -8.33213508e-01 -2.91221999e-02 1.90214872e+00
1.38293636e+00 8.12521100e-01 3.43233570e-02 7.91541398e-01
3.89346212e-01 -4.22623336e-01 -2.20485121e-01 -7.87335575e-01
1.81937128e-01 1.14314187e+00 5.46237409e-01 6.50991142e-01
-6.66923046e-01 1.01344872e+00 6.27151251e+00 9.23097968e-01
-1.29144776e+00 -3.76443684e-01 -3.41256142e-01 -3.82414967e-01
-9.53796133e-02 9.90720466e-02 -7.00716734e-01 1.31395504e-01
1.19925761e+00 -7.99663901e-01 7.05449164e-01 1.02694201e+00
-2.88583130e-01 -3.42859775e-01 -1.37427473e+00 5.75137079e-01
3.12874243e-02 -1.59918535e+00 -3.60837162e-01 -6.14267290e-02
7.15069354e-01 -3.39042038e-01 1.18851885e-02 7.34838009e-01
4.87318158e-01 -1.44104850e+00 8.60257745e-01 1.55430660e-01
8.63855243e-01 -1.11341310e+00 1.07995629e+00 2.76802778e-01
-8.98804247e-01 -1.65284246e-01 2.05541700e-01 -6.13690197e-01
2.43665818e-02 4.79148865e-01 -1.09855235e+00 6.41169727e-01
5.26773930e-01 3.31458569e-01 -2.57834047e-01 9.02102053e-01
-6.75499260e-01 8.06825995e-01 -6.40312850e-01 -4.29035664e-01
4.69503045e-01 3.02292019e-01 3.91992867e-01 1.08961809e+00
2.37900585e-01 9.41166803e-02 1.21649340e-01 1.20751321e+00
-1.42343968e-01 -5.51701307e-01 -4.58281159e-01 -3.16197515e-01
3.32741022e-01 9.89447415e-01 -3.93098176e-01 -5.42778492e-01
-3.75165767e-03 1.86152890e-01 1.81484688e-02 6.29983097e-02
-1.52416015e+00 -8.98882270e-01 2.30085820e-01 1.07987531e-01
4.61235881e-01 -4.51073259e-01 -1.71669260e-01 -8.30794752e-01
3.86688948e-01 -1.23086810e+00 -1.83815211e-02 -8.53592455e-01
-3.85013759e-01 3.84850472e-01 4.05029386e-01 -7.97498703e-01
-8.71799886e-01 -6.04830503e-01 -2.97292590e-01 8.54916632e-01
-1.52201128e+00 -8.28780949e-01 7.52229840e-02 2.63544768e-01
-1.57529235e-01 5.64368404e-02 9.87503052e-01 4.27630752e-01
-5.70583463e-01 8.92265320e-01 -3.05571109e-01 1.75862744e-01
2.24378347e-01 -1.06581712e+00 8.39128792e-02 9.89070654e-01
-2.73316115e-01 9.33204174e-01 8.51849973e-01 -4.18831348e-01
-2.07875299e+00 -9.43554163e-01 9.63779867e-01 1.75438374e-02
1.09742129e+00 -2.38202035e-01 -6.76435709e-01 7.92462587e-01
-8.00826326e-02 1.45976627e-02 1.36005506e-01 2.98815966e-01
-7.97943115e-01 -3.04072857e-01 -1.16884410e+00 6.98855937e-01
8.62529635e-01 -5.03715694e-01 -6.96591914e-01 1.69078037e-01
7.35962391e-01 -8.70170474e-01 -9.29548562e-01 4.20828938e-01
6.72450304e-01 -6.90160096e-01 6.41701043e-01 -4.38951880e-01
7.39014089e-01 -5.63559055e-01 -4.63917516e-02 -6.76094055e-01
-4.55371290e-02 -1.05462086e+00 -7.05411851e-01 9.11780477e-01
7.15516210e-01 -9.73932266e-01 5.77971578e-01 5.36205709e-01
-3.11670363e-01 -1.03081203e+00 -1.09209096e+00 -1.06390905e+00
3.94072771e-01 -6.54462099e-01 5.86675644e-01 7.36652613e-01
6.10934138e-01 2.13393405e-01 1.40101075e-01 7.11327866e-02
4.35249150e-01 3.64710331e-01 7.81911671e-01 -1.15106106e+00
-4.28451478e-01 -8.56501281e-01 -2.51929492e-01 -2.38080248e-01
6.75456882e-01 -9.50893283e-01 1.57239605e-02 -1.41042900e+00
-5.42189740e-02 -3.17550182e-01 -2.97119319e-01 1.17462969e+00
4.87177342e-01 2.98158348e-01 8.36725384e-02 -5.43191791e-01
-7.10073113e-01 2.00618897e-02 9.76936877e-01 -4.82051700e-01
4.18313853e-02 -4.92726266e-01 -7.94356465e-01 7.24517643e-01
7.64953315e-01 -3.94781411e-01 -1.47133395e-01 -2.35616639e-01
9.68207300e-01 3.04854155e-01 6.06744945e-01 -1.16648889e+00
3.51093978e-01 -3.39437187e-01 -4.18380827e-01 -4.19529796e-01
2.95118421e-01 -4.88855422e-01 2.06305191e-01 8.76431882e-01
-1.26280934e-01 2.00471565e-01 8.57845187e-01 -5.92516959e-02
-2.99544603e-01 -1.85073718e-01 3.26889217e-01 2.06548244e-01
-4.28749919e-01 -1.79379851e-01 -2.93493062e-01 1.89940289e-01
1.02503097e+00 9.32205245e-02 -5.16905010e-01 -5.63451201e-02
-9.71887484e-02 4.74977672e-01 3.33394706e-01 -2.39590332e-01
3.83822918e-01 -9.24100518e-01 -4.94355887e-01 -3.83285545e-02
-2.60199815e-01 3.43768567e-01 -7.93970376e-02 1.00204742e+00
-7.43157268e-01 7.24174201e-01 -2.20865048e-02 -3.48090053e-01
-1.12632275e+00 4.05841410e-01 3.94109011e-01 -5.09171963e-01
-4.15930301e-01 6.37932599e-01 -4.05418068e-01 -2.42112875e-01
9.49381739e-02 -1.07077479e+00 5.98223269e-01 -6.08581185e-01
4.35992002e-01 6.34781420e-01 1.13892615e-01 1.26519918e-01
-7.14653134e-01 3.90049249e-01 2.71444857e-01 -2.89436638e-01
1.24376798e+00 9.97949660e-01 -5.21204174e-01 6.31647781e-02
7.70081758e-01 7.69691467e-02 -3.91272753e-01 -1.64519325e-01
-4.23658967e-01 1.64800242e-01 8.85882899e-02 -1.21046197e+00
-8.10405374e-01 7.63884544e-01 -3.32893044e-01 3.85030121e-01
1.05974889e+00 -2.55476177e-01 9.35157299e-01 8.96288633e-01
6.76852584e-01 -9.68984365e-01 -6.26397431e-01 8.16964447e-01
6.27738416e-01 -6.69573486e-01 3.83343071e-01 -5.06925166e-01
1.21536799e-01 9.78342116e-01 5.76935172e-01 -3.94797355e-01
1.00211717e-01 9.07702029e-01 -5.03949404e-01 5.72970137e-02
-9.49896157e-01 3.14722687e-01 8.49392116e-02 1.01608679e-01
1.39842987e-01 9.00406914e-04 -2.13516995e-01 1.02753496e+00
-6.78291321e-01 6.83881104e-01 5.38479924e-01 1.18394423e+00
-1.19771324e-01 -1.31825089e+00 -4.83304709e-01 2.38314211e-01
-6.11950457e-01 -2.46479645e-01 -3.32521677e-01 1.29227245e+00
8.12361687e-02 8.32312882e-01 -3.16814363e-01 -2.58739144e-01
8.99496526e-02 1.47247463e-01 1.05656052e+00 -5.19902468e-01
-7.12242126e-01 -2.97878593e-01 6.40761673e-01 -7.42644310e-01
5.19160740e-02 -4.35098380e-01 -1.73690462e+00 -9.39872444e-01
-4.24656242e-01 4.55641180e-01 4.43918854e-01 9.65883315e-01
3.23534787e-01 6.59524798e-01 -3.67310755e-02 -4.18127120e-01
-6.72072411e-01 -4.32961792e-01 -2.56039172e-01 -6.11954749e-01
2.22264603e-01 -7.60910988e-01 -2.66080976e-01 -3.42114449e-01] | [8.89887523651123, 7.021091938018799] |
5a4ea988-1f12-4df1-8a3e-c691fa31b84f | learning-interpretable-deep-state-space-model | 2102.00397 | null | https://arxiv.org/abs/2102.00397v1 | https://arxiv.org/pdf/2102.00397v1.pdf | Learning Interpretable Deep State Space Model for Probabilistic Time Series Forecasting | Probabilistic time series forecasting involves estimating the distribution of future based on its history, which is essential for risk management in downstream decision-making. We propose a deep state space model for probabilistic time series forecasting whereby the non-linear emission model and transition model are parameterized by networks and the dependency is modeled by recurrent neural nets. We take the automatic relevance determination (ARD) view and devise a network to exploit the exogenous variables in addition to time series. In particular, our ARD network can incorporate the uncertainty of the exogenous variables and eventually helps identify useful exogenous variables and suppress those irrelevant for forecasting. The distribution of multi-step ahead forecasts are approximated by Monte Carlo simulation. We show in experiments that our model produces accurate and sharp probabilistic forecasts. The estimated uncertainty of our forecasting also realistically increases over time, in a spontaneous manner. | ['Yaohui Jin', 'Xiaokang Yang', 'Junchi Yan', 'Longyuan Li'] | 2021-01-31 | null | null | null | null | ['probabilistic-time-series-forecasting'] | ['time-series'] | [-1.72991708e-01 2.91503102e-01 -9.45818126e-02 -4.22318310e-01
-8.32075596e-01 -7.02451766e-01 9.57040310e-01 -7.11233392e-02
-5.62122092e-02 8.99609506e-01 9.09624577e-01 -7.33674765e-01
-3.84965450e-01 -9.14305270e-01 -4.25034940e-01 -7.73135960e-01
-4.05453861e-01 3.98870707e-01 1.70427531e-01 -1.16360344e-01
-5.15627638e-02 7.52021015e-01 -1.03839993e+00 -1.85189232e-01
5.13934195e-01 1.00339103e+00 -4.54535820e-02 6.77724004e-01
-2.14743391e-01 9.80131328e-01 -6.02657855e-01 -2.07819983e-01
2.84818590e-01 -1.89219519e-01 -7.24283978e-02 -4.80296820e-01
-8.86137843e-01 -4.90884423e-01 -3.85828972e-01 7.81974971e-01
1.94710776e-01 5.08707881e-01 1.02167559e+00 -1.16858196e+00
-2.34805033e-01 1.16451585e+00 -2.85139740e-01 2.80295223e-01
-4.36442465e-01 3.24341118e-01 5.45766532e-01 -4.40618426e-01
4.58912551e-02 1.38845813e+00 8.21678162e-01 3.22334766e-01
-9.01101530e-01 -7.18863904e-01 6.58135712e-01 -2.99150109e-01
-1.23861945e+00 -2.21575052e-01 1.01488590e+00 -5.30698597e-01
8.41000676e-01 1.31886870e-01 3.56193483e-01 1.20159209e+00
8.38620067e-01 3.76014769e-01 4.57972944e-01 -2.12737635e-01
5.08159280e-01 -1.47128552e-01 2.07121015e-01 -2.64536347e-02
-1.76729709e-01 9.11267400e-01 -2.45365620e-01 -3.98875564e-01
4.89711404e-01 5.29152632e-01 2.30474547e-01 4.78703320e-01
-7.04175234e-01 7.06508517e-01 6.69393390e-02 1.14535853e-01
-8.31267655e-01 5.22267222e-01 1.28355607e-01 8.51082206e-02
1.02090704e+00 1.24661721e-01 -5.50020099e-01 -2.46748272e-02
-8.04644644e-01 1.41343340e-01 7.45102167e-01 7.90967286e-01
2.35947296e-01 5.53224027e-01 -5.69585323e-01 4.00393128e-01
4.31452066e-01 8.02386403e-01 3.67201120e-01 -1.06703627e+00
2.99483716e-01 -8.44479352e-02 5.98066509e-01 -8.54844391e-01
-5.66405237e-01 -7.17608571e-01 -1.12885880e+00 2.08636835e-01
4.24032748e-01 -6.87314630e-01 -1.19303226e+00 1.86843693e+00
-1.64646972e-02 6.33954823e-01 2.25244492e-01 4.21723902e-01
-5.07723307e-03 1.15036070e+00 2.88266957e-01 -5.86702824e-01
8.26471627e-01 -5.23944557e-01 -8.55911732e-01 2.64600310e-02
3.69901985e-01 -2.34315231e-01 1.48392186e-01 1.83979183e-01
-8.30646336e-01 -2.24766850e-01 -5.49640059e-01 5.00219524e-01
-4.40233946e-01 -1.50246486e-01 2.90691257e-01 3.25699955e-01
-9.30522561e-01 7.63356149e-01 -1.08122754e+00 3.62383783e-01
2.51934081e-02 8.41531679e-02 3.58874470e-01 5.09603202e-01
-1.73227751e+00 9.47392941e-01 3.79156440e-01 6.24428988e-01
-1.30102479e+00 -7.98680961e-01 -6.28401637e-01 4.83399242e-01
1.69420227e-01 -4.80475664e-01 1.45343983e+00 -3.86932313e-01
-1.73966575e+00 -3.08286548e-01 -2.96457112e-01 -8.86940658e-01
6.35740101e-01 -1.05945896e-02 -8.34209979e-01 -3.13655227e-01
-1.55668899e-01 1.35221094e-01 9.02553737e-01 -8.90096545e-01
-7.37212718e-01 -2.40658242e-02 -3.04391593e-01 -5.52136339e-02
-1.29898623e-01 7.47627299e-03 8.58880877e-02 -8.22612405e-01
-1.09123379e-01 -8.99228036e-01 -8.96122873e-01 -5.18643320e-01
-4.43576813e-01 -1.43225759e-01 4.13899571e-01 -8.30883205e-01
1.52586615e+00 -2.06526947e+00 -3.92819911e-01 6.44565165e-01
4.07862216e-02 -2.10345045e-01 1.19107179e-01 7.73177445e-01
-2.04155311e-01 2.48788431e-01 -4.27026421e-01 -4.35689598e-01
4.97452989e-02 2.14649588e-01 -1.23894274e+00 2.81595051e-01
2.40614846e-01 8.71289849e-01 -8.50116551e-01 1.19834363e-01
1.81907147e-01 4.30958688e-01 1.88824818e-01 3.12747359e-01
-4.16391551e-01 5.85161328e-01 -6.10162437e-01 2.93480337e-01
5.69984317e-01 -1.62199721e-01 5.05341850e-02 3.73266459e-01
-4.34906900e-01 2.22705409e-01 -8.60768616e-01 7.43776441e-01
-5.27304232e-01 3.44245404e-01 -3.64589870e-01 -6.53886914e-01
9.81942952e-01 5.09645581e-01 2.77473986e-01 -2.48875439e-01
4.74790260e-02 -6.78693950e-02 -2.47639865e-01 8.81434679e-02
4.67645913e-01 -4.84039158e-01 -2.44475394e-01 7.18337536e-01
-4.97784257e-01 4.66216281e-02 -2.83427149e-01 -4.67814282e-02
8.80498886e-01 -1.14812486e-01 1.44500092e-01 -1.81308135e-01
9.90027487e-02 -1.05034702e-01 7.69048393e-01 9.33698535e-01
8.43576491e-02 2.08366096e-01 8.42648208e-01 -6.48787379e-01
-7.27735639e-01 -1.07464921e+00 1.31482974e-01 7.51998425e-01
-2.12006390e-01 1.64393261e-01 -5.31274267e-02 -5.39801478e-01
1.11059442e-01 1.57349336e+00 -7.44652629e-01 -3.99447948e-01
-4.53526556e-01 -1.02368224e+00 1.97752029e-01 7.96511948e-01
-2.83383906e-01 -9.26605105e-01 -1.10634029e-01 7.05047488e-01
3.60620201e-01 -5.72979629e-01 -2.69088805e-01 4.63270873e-01
-7.84822345e-01 -4.17149395e-01 -7.93042839e-01 7.75888264e-02
4.19370562e-01 -1.68993294e-01 8.93351078e-01 -6.89926803e-01
6.46319807e-01 -5.27318614e-03 1.63681969e-01 -9.74154532e-01
-7.53602505e-01 -6.92882910e-02 6.69138953e-02 -3.69115248e-02
3.94096710e-02 -9.23626065e-01 -5.16589105e-01 9.96876433e-02
-8.11379552e-01 -1.91682726e-01 1.39563024e-01 7.45148838e-01
5.29679358e-01 3.70335370e-01 1.11724508e+00 -7.93377638e-01
9.26579773e-01 -9.12680209e-01 -1.14826453e+00 3.72049958e-01
-7.62617946e-01 3.27527285e-01 8.07585955e-01 -4.21371549e-01
-1.42360795e+00 1.21354803e-01 -1.48041416e-02 -5.28452039e-01
6.01214729e-02 8.66718411e-01 1.97266966e-01 7.22899854e-01
2.81287998e-01 1.66885197e-01 -1.81713477e-01 -5.19901395e-01
3.84799361e-01 3.99432778e-01 6.46549225e-01 -3.59802753e-01
8.18852186e-01 4.51182693e-01 1.96858838e-01 -1.31051838e-01
-9.34353054e-01 -1.62318319e-01 -3.38097334e-01 -4.03866708e-01
4.46634114e-01 -1.05778062e+00 -7.82535434e-01 4.78753686e-01
-1.26109374e+00 -2.27220625e-01 -4.81450707e-01 6.96681499e-01
-2.44608879e-01 -1.54703990e-01 -5.73029160e-01 -1.77921331e+00
-2.19413146e-01 -8.89515758e-01 7.79370606e-01 1.89510733e-01
-1.60975769e-01 -1.32923436e+00 3.72068137e-01 -7.53985822e-01
6.52928770e-01 3.65510255e-01 8.61558616e-01 -7.86750436e-01
-3.73210698e-01 -4.96971130e-01 3.38473055e-03 7.97490180e-02
1.00618161e-01 1.99231371e-01 -1.01513112e+00 2.26073965e-01
1.90319926e-01 6.26066923e-01 1.13932800e+00 8.70246172e-01
1.20023966e+00 -4.34829623e-01 -5.80710649e-01 5.26835203e-01
9.61624026e-01 6.76783621e-01 4.07877505e-01 2.24635750e-02
4.01338398e-01 7.76434481e-01 3.28849822e-01 7.49860108e-01
3.72865498e-01 -4.34855521e-02 6.26360655e-01 2.45009109e-01
5.78303277e-01 -5.45852423e-01 3.64602894e-01 7.79855072e-01
4.35794294e-02 -7.18457997e-01 -1.18440461e+00 7.19002724e-01
-1.94723475e+00 -1.16875660e+00 8.98159668e-03 2.08676648e+00
5.99767089e-01 4.08808172e-01 1.13732360e-01 -3.33277076e-01
8.17572355e-01 1.67112857e-01 -8.50584686e-01 -3.65420401e-01
-6.56591728e-02 -2.34150290e-01 9.37049389e-01 6.61661685e-01
-1.01561642e+00 6.05497241e-01 7.46687222e+00 5.91390431e-01
-1.02405262e+00 -1.12978414e-01 1.25304639e+00 1.01548389e-01
-8.44176352e-01 3.11454143e-02 -9.48640704e-01 7.57134557e-01
1.77129436e+00 -7.14418948e-01 1.25505894e-01 7.22042203e-01
6.85844123e-01 9.13551003e-02 -6.91618085e-01 2.90468127e-01
-7.41951287e-01 -1.58936524e+00 -8.29941556e-02 -6.29154593e-02
7.53618479e-01 2.72140652e-01 1.87726617e-01 3.79481345e-01
1.25099504e+00 -9.65545654e-01 8.17071557e-01 1.40129471e+00
4.91442263e-01 -1.24328828e+00 8.85439575e-01 6.60131276e-01
-9.54488277e-01 -5.04927218e-01 -9.44861919e-02 -1.76320881e-01
8.22862625e-01 1.13487399e+00 -8.26566815e-01 3.42052132e-01
3.98596227e-01 6.48374856e-01 1.62156537e-01 7.05232143e-01
-3.11733514e-01 1.18513048e+00 -5.70518792e-01 -3.93152423e-02
3.64706755e-01 -3.06719661e-01 7.37455130e-01 9.56971288e-01
7.48012424e-01 2.25628763e-02 -1.06367487e-02 8.51448238e-01
1.43595830e-01 -5.90270936e-01 -5.69343448e-01 -1.20842017e-01
6.31073415e-01 7.56601572e-01 -5.56969404e-01 -3.89916062e-01
9.72091034e-02 2.78029323e-01 -1.37555957e-01 8.75286460e-01
-7.92547047e-01 -1.62566304e-01 5.44849753e-01 -2.25717202e-01
3.30387056e-01 -2.78331518e-01 -4.24758911e-01 -8.15235019e-01
-2.32289538e-01 -3.23031470e-02 4.12437856e-01 -6.36991084e-01
-1.71513808e+00 7.43299961e-01 1.97568253e-01 -1.33966088e+00
-1.23110938e+00 -2.38682777e-01 -1.07794511e+00 1.49153352e+00
-1.73582983e+00 -8.15390289e-01 5.80410659e-01 1.92388017e-02
1.99596375e-01 -1.80173352e-01 7.21541107e-01 -4.61561024e-01
-7.61645794e-01 9.27458927e-02 7.48756349e-01 -1.00161463e-01
2.72584826e-01 -1.04332638e+00 9.35780108e-01 9.81309533e-01
-5.11000633e-01 5.22790790e-01 8.87077510e-01 -9.40310359e-01
-7.96145141e-01 -1.61321843e+00 9.32131290e-01 -4.80656385e-01
1.11893404e+00 -1.52035519e-01 -1.03048778e+00 5.61544836e-01
-1.45532444e-01 1.24020744e-02 4.21642751e-01 -3.28599736e-02
-1.39163807e-01 -8.56640041e-02 -9.22734201e-01 4.81874406e-01
4.86309856e-01 -4.72897023e-01 -3.83208215e-01 4.27090585e-01
1.18327856e+00 -9.33138728e-02 -6.31125152e-01 2.11130932e-01
6.03415430e-01 -4.88607764e-01 7.71730602e-01 -8.05277884e-01
3.23121488e-01 -1.51961073e-01 1.04915701e-01 -1.58918238e+00
-5.55373907e-01 -1.02788877e+00 -5.27101338e-01 1.24629915e+00
7.31684923e-01 -1.02043748e+00 4.06451464e-01 1.03917027e+00
-2.51846939e-01 -4.03003514e-01 -1.16943467e+00 -9.83619690e-01
2.85894871e-01 -9.74286675e-01 1.23115373e+00 5.73794305e-01
-3.41720074e-01 -1.27643690e-01 -7.22727180e-01 6.92161500e-01
5.78626335e-01 2.19171658e-01 7.38731548e-02 -1.21003234e+00
-9.21302512e-02 -4.55661148e-01 1.74468055e-01 -7.02844918e-01
4.01527911e-01 -4.61056650e-01 3.11893553e-01 -1.10293782e+00
-1.85420960e-01 -2.69912541e-01 -9.60290790e-01 4.00133461e-01
-2.67943263e-01 -4.58359867e-01 -9.94694382e-02 1.74869806e-01
-1.15474768e-01 9.02088344e-01 7.32803464e-01 -4.27667275e-02
-4.55650598e-01 6.41614079e-01 -7.52053559e-01 7.42751420e-01
1.01405716e+00 -8.44213188e-01 -4.88142550e-01 -2.33926848e-01
2.51447737e-01 7.39223063e-01 2.89488494e-01 -5.84006429e-01
3.28470230e-01 -7.40589678e-01 3.65559518e-01 -1.17349982e+00
2.21131355e-01 -6.97166741e-01 5.78192115e-01 3.37700844e-01
-7.07514167e-01 2.34715149e-01 2.32457981e-01 1.15775502e+00
-1.44689113e-01 4.04815413e-02 1.77306145e-01 -3.50690889e-03
-3.96828800e-01 5.15909076e-01 -8.89573753e-01 -3.22180688e-01
6.98934257e-01 3.71432811e-01 -1.85648322e-01 -1.09344864e+00
-9.52727437e-01 6.28279150e-01 -1.96094111e-01 2.01213092e-01
3.53703767e-01 -1.15226567e+00 -7.15589225e-01 -7.42812604e-02
-2.10107058e-01 -1.99207336e-01 4.66123939e-01 3.79681528e-01
6.56579360e-02 4.41831499e-01 4.32956755e-01 -1.99594542e-01
-1.87230080e-01 5.95694423e-01 5.19366801e-01 -6.03057027e-01
-4.46282059e-01 6.98587120e-01 3.37264776e-01 -2.05285817e-01
1.51750773e-01 -5.71137309e-01 -4.29372549e-01 1.41669109e-01
9.02274191e-01 4.21655715e-01 -2.10256845e-01 -2.71779478e-01
-2.45285809e-01 -1.50784597e-01 1.23472147e-01 -5.54707468e-01
1.67327368e+00 -3.35573137e-01 1.16226999e-02 8.95779014e-01
7.92700887e-01 -1.92571640e-01 -1.66484356e+00 -1.67116001e-01
3.96785110e-01 1.06050856e-01 2.69897968e-01 -1.04855561e+00
-8.40556800e-01 8.49777400e-01 3.89550447e-01 4.89622921e-01
1.04799652e+00 -2.77930498e-01 8.18907619e-01 2.19351903e-01
-1.72882024e-02 -9.34567213e-01 -9.20129299e-01 9.73867238e-01
1.01841414e+00 -1.02091765e+00 -3.00255835e-01 1.01275101e-01
-6.55183911e-01 1.18797112e+00 -6.96304664e-02 -6.53440282e-02
1.30073309e+00 6.26249731e-01 1.74431473e-01 1.67805150e-01
-1.27662253e+00 3.93808307e-03 2.97762334e-01 5.61110318e-01
-4.16494943e-02 4.71175462e-01 1.77283615e-01 1.27370727e+00
-1.99013114e-01 -1.96390357e-02 4.75755513e-01 5.00707865e-01
-2.42753580e-01 -7.56100357e-01 -3.42974991e-01 5.49179792e-01
-4.20260012e-01 -2.18292832e-01 1.21744975e-01 2.66066879e-01
-5.00207126e-01 9.32941079e-01 2.34227270e-01 -3.90342176e-01
1.21029101e-01 1.75349891e-01 -7.04488873e-01 -3.77171874e-01
-5.96709371e-01 3.02455604e-01 1.61398813e-01 -2.65406072e-01
1.82619527e-01 -6.17946625e-01 -1.14963007e+00 -1.14746295e-01
-1.63352042e-01 2.19758034e-01 7.21918285e-01 1.09310853e+00
5.89351714e-01 8.31773639e-01 1.14479899e+00 -7.41642773e-01
-1.02647853e+00 -9.13772225e-01 -6.24750614e-01 -4.37032789e-01
5.97564876e-01 -4.78208005e-01 -7.59228408e-01 -1.62665650e-01] | [6.750422477722168, 3.120046377182007] |
cc2674f5-0452-4203-9339-dd4d59d519cd | semantic-segmentation-using-super-resolution | 2306.15218 | null | https://arxiv.org/abs/2306.15218v1 | https://arxiv.org/pdf/2306.15218v1.pdf | Semantic Segmentation Using Super Resolution Technique as Pre-Processing | Combining high-level and low-level visual tasks is a common technique in the field of computer vision. This work integrates the technique of image super resolution to semantic segmentation for document image binarization. It demonstrates that using image super-resolution as a preprocessing step can effectively enhance the results and performance of semantic segmentation. | ['Tingkai Chang', 'Chun-Tse Chien', 'Jen-Shiun Chiang', 'Wei-Han Chen', 'Chih-Chia Chen'] | 2023-06-27 | null | null | null | null | ['image-super-resolution', 'super-resolution'] | ['computer-vision', 'computer-vision'] | [ 8.88135374e-01 -1.10226177e-01 -3.60004932e-01 -2.93068916e-01
-6.59598112e-01 -4.34802026e-01 4.22625840e-01 1.67391613e-01
-4.02571052e-01 3.64660174e-01 6.03293851e-02 -1.37571454e-01
7.97542557e-02 -8.68757665e-01 -2.78495193e-01 -5.33763647e-01
8.47483039e-01 2.38090515e-01 8.71862769e-01 -3.93867612e-01
7.86210060e-01 7.94162571e-01 -1.58678710e+00 8.27037871e-01
1.06745100e+00 3.28282982e-01 5.40374875e-01 8.83447409e-01
-8.21056664e-01 2.24982157e-01 -8.53263021e-01 1.54148504e-01
6.21344037e-02 -3.77138346e-01 -1.18417299e+00 6.33536696e-01
8.46080065e-01 -6.97720572e-02 1.69055820e-01 1.63277757e+00
2.57568657e-01 2.96718001e-01 4.58127618e-01 -4.67838049e-01
-7.52475679e-01 5.98857105e-01 -9.40374374e-01 8.29762280e-01
4.74477738e-01 -5.58613777e-01 5.53447545e-01 -6.34823561e-01
7.19510853e-01 1.51050484e+00 4.72739726e-01 2.71683276e-01
-1.21140885e+00 -3.84893864e-01 -1.07803285e-01 2.24613488e-01
-1.22557604e+00 5.75550310e-02 6.60055637e-01 -3.91176969e-01
7.13373244e-01 3.02117050e-01 1.92947492e-01 1.80133849e-01
2.13868454e-01 7.27080822e-01 1.42815590e+00 -7.98332870e-01
2.53973186e-01 9.90537405e-02 8.52916181e-01 3.60348344e-01
3.42723548e-01 -1.86897501e-01 -4.53194886e-01 4.60479051e-01
1.20283747e+00 4.24352884e-02 1.24087729e-01 8.23961645e-02
-6.62667572e-01 5.23584664e-01 3.94317418e-01 1.10572433e+00
-3.46510500e-01 -1.49848297e-01 1.57593831e-01 -2.82430440e-01
5.86883187e-01 5.46079338e-01 1.16694666e-01 2.30448022e-01
-1.58367872e+00 2.38672141e-02 -7.08796605e-02 5.64022779e-01
4.93900180e-01 4.99061406e-01 -3.66840839e-01 7.55943775e-01
-5.58467992e-02 2.37923488e-01 5.53869724e-01 -1.36584425e+00
1.45131350e-01 7.63740182e-01 -5.98739907e-02 -8.54257643e-01
-3.01992923e-01 -2.86431700e-01 -3.59278351e-01 7.20688343e-01
4.21069086e-01 5.85005879e-01 -1.69978857e+00 7.25930929e-01
3.66195887e-01 -1.05182827e-01 1.60503108e-02 7.95390666e-01
9.44003284e-01 8.02677631e-01 1.49759680e-01 -2.15877414e-01
1.51700175e+00 -9.25673127e-01 -1.12956262e+00 -3.08587760e-01
-7.43857473e-02 -8.59112501e-01 9.36307132e-01 3.84877712e-01
-1.49032426e+00 -9.05901432e-01 -1.14557350e+00 -2.73570120e-01
-7.84533083e-01 -2.87880003e-01 6.18894219e-01 8.45088780e-01
-1.07888329e+00 5.48885584e-01 -6.38875663e-01 -3.57946306e-01
5.17212033e-01 3.39142561e-01 -4.40903515e-01 -1.46119595e-01
-8.35780799e-01 8.21574390e-01 1.01727355e+00 -3.24233025e-01
-2.77298152e-01 -3.72282028e-01 -8.02527070e-01 1.44897193e-01
1.44139424e-01 -1.81371629e-01 8.86942267e-01 -8.55489492e-01
-1.11003685e+00 1.13671494e+00 -4.07605201e-01 -2.48925135e-01
8.44610780e-02 -1.31176636e-01 -1.32866845e-01 8.26819599e-01
3.28629702e-01 4.23821330e-01 1.14422941e+00 -1.55057132e+00
-1.11631393e+00 -7.85309672e-01 -3.57205212e-01 4.72667575e-01
-6.74272375e-03 4.02001262e-01 -8.18548918e-01 -5.37110925e-01
6.10248089e-01 -1.72929540e-01 -1.78064287e-01 -6.87801421e-01
-1.21960424e-01 -1.78742424e-01 1.23253524e+00 -1.18271863e+00
8.37995172e-01 -1.94566643e+00 7.09806383e-02 1.91138163e-01
8.27964172e-02 4.20055002e-01 1.42154872e-01 -3.94736022e-01
-5.66326715e-02 3.65556955e-01 -2.27295309e-01 4.48391885e-02
-5.75667977e-01 -2.62888763e-02 1.33249879e-01 1.04348041e-01
-3.89575571e-01 8.68045866e-01 -5.72343588e-01 -1.05748224e+00
8.50514829e-01 5.95730662e-01 -7.14604482e-02 -2.60466307e-01
-3.63905705e-03 4.00435179e-01 -3.44630659e-01 7.91027248e-01
1.04262447e+00 -2.80318502e-02 -1.05476618e-01 -3.48598391e-01
-3.47436637e-01 -2.51519620e-01 -1.07201767e+00 1.50699461e+00
-8.16570148e-02 8.65260661e-01 3.01100820e-01 -9.72654343e-01
7.17882216e-01 6.96086586e-02 3.96065623e-01 -1.12901855e+00
1.91054180e-01 -2.03544617e-01 -5.56781828e-01 -2.99131155e-01
9.82884705e-01 -1.19863860e-01 -9.02176052e-02 3.30283828e-02
-5.47026470e-02 -5.94592869e-01 4.53527361e-01 1.50461018e-01
3.89741987e-01 1.46434218e-01 2.55193919e-01 -2.44195580e-01
6.56834781e-01 5.06662428e-01 2.48408273e-01 7.79520869e-01
-2.40327716e-01 6.98803484e-01 -3.04236822e-02 4.34041396e-03
-1.46879578e+00 -1.22749841e+00 -2.86392063e-01 1.44570124e+00
6.33048952e-01 1.36834741e-01 -1.57150733e+00 -1.75018802e-01
-1.72616988e-01 6.62383556e-01 -4.97430116e-01 2.86228955e-01
-8.18084478e-01 -9.48235512e-01 1.71611309e-01 7.07878292e-01
9.44196045e-01 -1.04570115e+00 -5.00025272e-01 -1.04546525e-01
-1.49353415e-01 -1.47682774e+00 -2.39691764e-01 1.45120099e-01
-1.10955501e+00 -7.25615144e-01 -8.94716322e-01 -1.10488915e+00
8.35012436e-01 5.93443155e-01 7.60873973e-01 -1.15528181e-02
-7.64870048e-01 2.01168507e-01 -4.16198641e-01 1.79478936e-02
-5.74822783e-01 -2.47979790e-01 -6.54352248e-01 -3.33034515e-01
3.17146748e-01 3.77368480e-02 -3.93855155e-01 8.00070167e-02
-1.18611169e+00 1.54189020e-01 4.10210997e-01 5.34033358e-01
9.26775157e-01 8.33063006e-01 -3.85406949e-02 -9.15319562e-01
7.38447666e-01 4.56915915e-01 -5.60099065e-01 2.07012087e-01
-6.01109445e-01 -6.31759986e-02 3.28146487e-01 8.41886457e-03
-1.46969450e+00 1.77535098e-02 1.19360916e-01 -2.36453414e-02
-5.54978251e-01 -4.64586765e-02 -1.52438417e-01 -3.69143397e-01
6.28295422e-01 4.48956996e-01 -7.33094588e-02 -6.46870673e-01
4.98789638e-01 8.18021297e-01 9.64692593e-01 1.33318929e-02
5.70160449e-01 8.84900153e-01 -9.16713849e-02 -1.09253561e+00
-9.30790603e-01 -8.39872599e-01 -1.11902273e+00 -6.08096048e-02
1.42437196e+00 -6.16947889e-01 -2.09182799e-01 4.04634982e-01
-7.76329994e-01 -1.75935954e-01 -4.77116525e-01 1.18586630e-01
-2.27759004e-01 5.54471791e-01 -6.99815392e-01 -5.84375262e-01
-4.55171853e-01 -9.43847179e-01 1.14182413e+00 8.10831785e-01
1.01334557e-01 -9.31864440e-01 -3.65482211e-01 1.16447043e+00
2.85639584e-01 3.96496654e-01 6.63733602e-01 -2.77356565e-01
-4.89320368e-01 -2.04505593e-01 -7.41922379e-01 5.75021684e-01
1.73907235e-01 -5.08497143e-03 -8.94266009e-01 3.64627540e-02
1.07851930e-01 3.53075176e-01 1.13236189e+00 9.43296611e-01
9.64696586e-01 2.25908637e-01 -4.39271241e-01 6.32648289e-01
1.72141588e+00 4.21097964e-01 9.34057236e-01 7.79337943e-01
1.00877309e+00 5.11866987e-01 9.03660119e-01 -2.88467944e-01
-4.99143302e-02 7.02220142e-01 -5.21129817e-02 -6.14668131e-01
-6.83466613e-01 2.13804811e-01 -1.64825499e-01 2.35761553e-01
-2.78662354e-01 2.58750051e-01 -9.52728629e-01 5.25626481e-01
-1.60840452e+00 -1.22870588e+00 -3.44608068e-01 1.57098365e+00
7.34275937e-01 2.11213946e-01 9.76285115e-02 5.75413525e-01
1.41113317e+00 -8.10858328e-03 -1.77551910e-01 -9.28944528e-01
-2.60594189e-01 3.28867644e-01 8.33701491e-01 8.83220613e-01
-1.33637738e+00 1.72294259e+00 7.59375858e+00 1.42194033e+00
-8.94388497e-01 1.76342160e-01 5.18080592e-01 5.36607802e-01
-3.59530151e-02 -3.14328760e-01 -1.00086236e+00 4.90695596e-01
4.64325488e-01 -2.14527518e-01 3.14298749e-01 7.10045516e-01
2.45984852e-01 -5.88151574e-01 -3.48530412e-01 8.62925589e-01
4.92295742e-01 -1.31217670e+00 5.85813224e-01 -2.36986041e-01
9.55809116e-01 -6.21349156e-01 3.07429433e-01 -1.39720827e-01
1.07457079e-01 -1.36556160e+00 3.66765052e-01 1.28954798e-01
7.30411112e-01 -8.55691731e-01 6.11672282e-01 4.06701602e-02
-1.26702011e+00 -4.13184240e-02 -3.49610537e-01 2.59364873e-01
2.63412654e-01 1.45103395e-01 -7.98414409e-01 2.07244232e-01
8.53202641e-01 2.65149534e-01 -8.02405357e-01 1.00735593e+00
-8.84931684e-02 1.11506082e-01 -1.08002722e-02 5.54428041e-01
1.85963765e-01 -2.55517691e-01 4.60558385e-01 1.53027833e+00
-2.36024857e-01 1.51185840e-01 1.65418878e-01 8.22403789e-01
2.48321831e-01 1.23084404e-01 -2.00385183e-01 2.72017363e-02
1.88032970e-01 1.13728487e+00 -1.62872040e+00 -6.25464559e-01
-4.29292917e-02 1.29043162e+00 -4.69211042e-01 3.97787511e-01
-3.51630062e-01 -5.32091916e-01 -9.46785847e-04 1.28541768e-01
3.71413529e-01 -2.96508849e-01 -1.29041231e+00 -6.08073294e-01
-4.16174859e-01 -6.58636868e-01 5.90805590e-01 -8.82773638e-01
-3.09383899e-01 4.47622567e-01 2.09719867e-01 -6.40701950e-01
-3.09390873e-02 -5.72675288e-01 -3.79672110e-01 8.73722613e-01
-1.40428984e+00 -1.42621017e+00 -5.48991978e-01 6.61584198e-01
1.17151892e+00 8.38162377e-03 5.02118170e-01 -1.17395520e-01
-2.66238004e-01 1.51837185e-01 4.35911328e-01 1.14898644e-01
1.49201527e-01 -1.51976061e+00 2.15681404e-01 1.33748150e+00
5.99576347e-02 1.71415210e-01 9.25202608e-01 -7.68199742e-01
-4.28729266e-01 -5.89750528e-01 3.41050625e-01 -4.17190492e-01
1.70741528e-02 -9.93734226e-02 -1.05279398e+00 1.53576344e-01
8.37417394e-02 -5.72527707e-01 3.31437618e-01 -4.29392487e-01
8.42117295e-02 2.06920162e-01 -1.65826821e+00 4.45561171e-01
2.55157739e-01 -4.63338286e-01 -1.14121938e+00 7.19575062e-02
7.57783413e-01 -4.71935391e-01 -7.29709983e-01 5.14620245e-01
2.16579527e-01 -8.07874262e-01 1.61181104e+00 -1.32351965e-01
1.05292097e-01 -3.16057593e-01 1.16227031e-01 -6.62666321e-01
-4.36926454e-01 -2.41314489e-02 3.03910762e-01 1.11719656e+00
2.05091566e-01 -8.59821960e-02 1.08897758e+00 5.57094991e-01
1.21713392e-01 1.20817594e-01 -7.74430990e-01 -5.64778984e-01
8.87548178e-02 -2.52175003e-01 3.46898437e-02 6.49343789e-01
-1.90734014e-01 1.93482593e-01 1.49717882e-01 1.44097194e-01
1.02431226e+00 3.46203834e-01 2.05006644e-01 -1.19884646e+00
2.91439414e-01 -8.07245016e-01 -5.49310505e-01 -4.77548927e-01
8.62047225e-02 -5.57192981e-01 -2.77306116e-03 -2.23023653e+00
3.35586220e-01 2.82793254e-01 -4.08603549e-01 9.33160186e-02
-4.56593752e-01 8.16089213e-01 3.23950380e-01 1.73311293e-01
-6.38341427e-01 -2.37177774e-01 1.58194304e+00 -1.74851552e-01
-4.84282285e-01 -1.11808702e-01 -7.85364926e-01 9.73129094e-01
9.66970146e-01 -2.07583055e-01 -1.32175207e-01 -1.85026862e-02
-4.29538488e-01 -1.62006691e-01 7.47912154e-02 -7.09869444e-01
1.69728845e-01 -2.85031706e-01 7.41625547e-01 -8.16156745e-01
3.40773493e-01 -6.44171119e-01 -6.52482748e-01 2.74084777e-01
-2.82879859e-01 -3.22841614e-01 4.21329767e-01 4.22471464e-01
-4.69970226e-01 -5.36655784e-01 1.25607371e+00 -6.36626422e-01
-1.50855839e+00 -3.93035263e-01 -5.03724575e-01 -1.42600015e-01
1.16846168e+00 -1.09140110e+00 -4.43723172e-01 1.48984224e-01
-9.30476964e-01 -6.22668639e-02 6.68042541e-01 3.45920593e-01
9.93582487e-01 -5.25261104e-01 -5.19324541e-01 1.40482381e-01
-4.76265907e-01 5.96904382e-02 3.99311066e-01 3.44641745e-01
-1.09083176e+00 6.03114069e-01 -7.35252380e-01 -6.99206352e-01
-2.02822161e+00 5.46921134e-01 2.29791060e-01 2.54245289e-02
-9.19301212e-01 8.22771251e-01 1.26226813e-01 3.56976002e-01
1.51378229e-01 -4.62220013e-02 -8.74954343e-01 -5.05957566e-02
7.52464294e-01 6.57806277e-01 -2.60440230e-01 -8.71714056e-01
-3.34337085e-01 1.14302695e+00 -3.43689919e-01 -3.90307069e-01
9.22766864e-01 -5.49437523e-01 -3.69586825e-01 2.88737148e-01
7.41222680e-01 -9.98001993e-02 -9.70211864e-01 -1.17424250e-01
3.20949763e-01 -7.90338397e-01 6.10651076e-01 -6.37843668e-01
-6.58213317e-01 9.30243433e-01 9.84069109e-01 2.50180930e-01
1.29383528e+00 2.63148218e-01 6.09658957e-01 -1.01156250e-01
-4.59126337e-03 -1.87281418e+00 3.19117874e-01 3.08473527e-01
4.29617584e-01 -1.26733708e+00 4.21211153e-01 -8.33751500e-01
-6.93447530e-01 1.04336858e+00 5.55318415e-01 -3.60071093e-01
1.15547098e-01 4.33090806e-01 6.08873889e-02 -2.41037548e-01
2.71896660e-01 -6.13866687e-01 5.27738154e-01 7.62969613e-01
3.10009539e-01 -1.00493550e-01 -4.83627647e-01 1.89350799e-01
-4.24621217e-02 -2.52778441e-01 7.22328365e-01 9.17066574e-01
-1.06086099e+00 -8.33396196e-01 -1.10785520e+00 9.14457589e-02
-1.04162550e+00 -9.53049436e-02 -3.64684105e-01 4.91104066e-01
2.74217099e-01 8.87809098e-01 3.82051975e-01 1.69090927e-02
1.30891455e-02 1.68375075e-02 5.76560140e-01 -7.97231495e-01
-4.09758002e-01 2.23942682e-01 -4.09164548e-01 -3.85838300e-01
-6.40129745e-01 -3.18908751e-01 -1.65204811e+00 1.33495599e-01
-3.29810143e-01 1.13108121e-01 1.12341070e+00 9.15099621e-01
-2.34474331e-01 9.85671043e-01 7.70120416e-03 -9.86724973e-01
3.05928618e-01 -6.30013525e-01 -8.16605926e-01 6.96812928e-01
3.39809120e-01 -2.32927352e-01 -1.79385260e-01 6.97822273e-01] | [9.571313858032227, 0.19211427867412567] |
55d5c81c-62d3-4dc3-8c0f-6d3227635812 | streaming-multi-talker-speech-recognition | 2104.02109 | null | https://arxiv.org/abs/2104.02109v1 | https://arxiv.org/pdf/2104.02109v1.pdf | Streaming Multi-talker Speech Recognition with Joint Speaker Identification | In multi-talker scenarios such as meetings and conversations, speech processing systems are usually required to transcribe the audio as well as identify the speakers for downstream applications. Since overlapped speech is common in this case, conventional approaches usually address this problem in a cascaded fashion that involves speech separation, speech recognition and speaker identification that are trained independently. In this paper, we propose Streaming Unmixing, Recognition and Identification Transducer (SURIT) -- a new framework that deals with this problem in an end-to-end streaming fashion. SURIT employs the recurrent neural network transducer (RNN-T) as the backbone for both speech recognition and speaker identification. We validate our idea on the LibrispeechMix dataset -- a multi-talker dataset derived from Librispeech, and present encouraging results. | ['Yifan Gong', 'Jinyu Li', 'Naoyuki Kanda', 'Liang Lu'] | 2021-04-05 | null | null | null | null | ['speaker-identification'] | ['speech'] | [ 4.52456564e-01 -1.74721070e-02 1.48311064e-01 -6.53822005e-01
-1.46572852e+00 -6.89069211e-01 4.46475804e-01 -1.37947917e-01
-1.66873038e-01 2.71051645e-01 5.28874516e-01 -5.65409899e-01
1.13864392e-01 6.75110668e-02 -3.29500347e-01 -8.50229442e-01
4.34161909e-02 6.19883120e-01 -9.28015187e-02 -1.80003896e-01
-2.66325235e-01 3.55810612e-01 -1.44947374e+00 6.20330751e-01
3.81103158e-01 7.74937987e-01 7.08461702e-02 1.28625047e+00
-2.53691852e-01 8.96267056e-01 -7.01013982e-01 -2.43808374e-01
1.86428607e-01 -7.77536809e-01 -6.91764832e-01 1.84531689e-01
2.75654763e-01 -1.87133461e-01 -3.52429956e-01 5.97999573e-01
7.91237891e-01 1.84035957e-01 4.11976904e-01 -9.55891371e-01
3.56262803e-01 1.20410216e+00 -4.14350718e-01 4.73431736e-01
2.73671657e-01 -2.38594487e-01 1.09796262e+00 -1.03052509e+00
6.73530996e-02 1.42996395e+00 6.17922246e-01 5.43955505e-01
-1.19684339e+00 -7.57449985e-01 1.34399697e-01 1.16272774e-02
-1.37015736e+00 -1.40817153e+00 7.81271696e-01 -1.80096507e-01
9.30386007e-01 6.53107703e-01 2.50203609e-01 1.24816394e+00
-5.66746473e-01 1.11649501e+00 6.38120353e-01 -5.74748874e-01
9.22371820e-02 -3.41045409e-02 3.19119573e-01 -3.57264355e-02
-5.08691907e-01 -7.58216009e-02 -8.44666839e-01 -1.78905576e-01
5.95284283e-01 -1.10458359e-01 -2.42550656e-01 3.26582909e-01
-1.25698948e+00 4.94068384e-01 -2.34022304e-01 4.71856773e-01
-7.22124696e-01 -1.21235229e-01 6.78732276e-01 6.84031427e-01
7.20452309e-01 -1.72997862e-01 -2.02543885e-01 -3.84680003e-01
-1.32318556e+00 -7.54999369e-02 1.23515701e+00 8.70371342e-01
2.44966328e-01 3.55102032e-01 -9.54632387e-02 1.36008072e+00
3.42606664e-01 3.45175773e-01 7.57637262e-01 -5.15740991e-01
7.84478247e-01 -6.37999848e-02 -7.98984841e-02 -2.43545994e-01
-1.42108142e-01 -3.43852758e-01 -1.02844620e+00 -2.91055262e-01
1.80332839e-01 -4.90164846e-01 -9.10487533e-01 1.40910745e+00
4.12572891e-01 4.77637500e-01 4.65813845e-01 7.67175257e-01
9.05873835e-01 1.12549019e+00 -2.27503538e-01 -5.82350194e-01
1.37149715e+00 -1.15746903e+00 -8.89357150e-01 -1.67336911e-01
4.42554802e-01 -9.72771585e-01 5.74682295e-01 4.07108217e-01
-1.30242181e+00 -4.41611409e-01 -5.98894894e-01 2.08103761e-01
1.42779082e-01 1.48658767e-01 -1.18593231e-01 7.55980730e-01
-1.08140039e+00 1.78859048e-02 -7.58747518e-01 -2.99434602e-01
-1.60290360e-01 5.82019389e-01 -2.24709943e-01 1.84825927e-01
-1.03316677e+00 3.94959956e-01 2.57485926e-01 4.95851278e-01
-1.00918555e+00 -3.83481532e-01 -6.51842952e-01 3.05231124e-01
1.46535575e-01 -2.82915145e-01 1.82351339e+00 -1.20754457e+00
-2.13561249e+00 5.24020016e-01 -7.37877309e-01 -6.70816779e-01
2.52079517e-01 -1.41761333e-01 -6.90724254e-01 5.97454570e-02
-3.14786822e-01 6.57921955e-02 1.29973352e+00 -9.99757528e-01
-6.97898030e-01 -2.49993175e-01 -5.94518542e-01 5.08009553e-01
-2.17898399e-01 6.94039047e-01 -3.01363140e-01 -7.41968095e-01
4.28635925e-01 -8.64794910e-01 -6.31659776e-02 -9.99452710e-01
-6.34596407e-01 -1.98924974e-01 1.07759368e+00 -9.40715313e-01
1.16335738e+00 -2.63770580e+00 2.73908466e-01 2.32030988e-01
-4.91412058e-02 5.90145648e-01 -8.81650671e-02 7.12141037e-01
-4.85204339e-01 -3.19185495e-01 -1.31090477e-01 -1.15765083e+00
-1.97099522e-01 1.28430784e-01 -6.34797275e-01 3.53170782e-01
-7.04570999e-03 4.99175102e-01 -5.91002703e-01 -2.43657187e-01
9.67447311e-02 4.72047269e-01 -3.89681272e-02 6.42786860e-01
6.14250526e-02 5.37092388e-01 -4.72866558e-02 5.70587814e-01
4.38709259e-01 2.34112978e-01 3.12606633e-01 1.15967222e-01
-2.99682736e-01 7.19123006e-01 -1.32374704e+00 1.43194067e+00
-6.42362416e-01 8.37804794e-01 1.04140890e+00 -1.16883552e+00
9.70076323e-01 1.17005122e+00 3.27240676e-01 -5.01921438e-02
1.11225910e-01 4.46847379e-01 6.63459376e-02 -5.52110791e-01
4.76442963e-01 -5.60309529e-01 -9.25388113e-02 4.78730083e-01
1.78099439e-01 -5.50754629e-02 -1.21395774e-01 7.03112483e-02
9.79340315e-01 -5.75767994e-01 1.85127184e-01 1.67047709e-01
6.50068879e-01 -4.03642029e-01 4.25797760e-01 5.71879983e-01
-9.15187597e-02 8.16766381e-01 1.28495291e-01 6.63754791e-02
-8.79544199e-01 -9.03758705e-01 3.58644456e-01 1.45886469e+00
-5.14619887e-01 -2.41062850e-01 -7.77327359e-01 -2.56871730e-01
-4.32971448e-01 4.92973566e-01 2.74573639e-02 3.56497318e-01
-8.31384838e-01 -5.23660600e-01 1.04516971e+00 3.98464859e-01
1.36247918e-01 -8.76829028e-01 1.64160088e-01 6.16076827e-01
-4.07613009e-01 -1.28000546e+00 -9.62135196e-01 5.42895496e-01
-6.69939160e-01 -3.45642447e-01 -9.88750398e-01 -1.23075545e+00
3.00263137e-01 4.58733350e-01 7.10077584e-01 -4.62515116e-01
1.47623092e-01 2.04755902e-01 -3.84692192e-01 -2.98116088e-01
-9.97717321e-01 3.44101310e-01 2.93522477e-01 7.98453271e-01
7.69907534e-02 -7.86023080e-01 -7.60266408e-02 3.82994503e-01
-7.73014128e-01 -2.12688521e-01 5.57075202e-01 7.48282433e-01
8.28706771e-02 -5.39705791e-02 9.95267212e-01 -5.78012049e-01
7.28505433e-01 -3.69720757e-01 -4.43286955e-01 3.14880401e-01
1.89167351e-01 -3.65852565e-01 6.33225322e-01 -6.36497617e-01
-1.20673072e+00 3.14785749e-01 -5.27583122e-01 -7.07738876e-01
-4.12574947e-01 5.30003309e-01 -3.77339065e-01 2.89983690e-01
3.60935718e-01 6.24015152e-01 2.41399910e-02 -7.98045218e-01
2.89510280e-01 1.67679501e+00 7.40786076e-01 -1.48117915e-01
5.22719026e-01 5.89872263e-02 -6.44999146e-01 -1.53098762e+00
-4.35662746e-01 -1.29288673e+00 -5.04817367e-01 -6.49854466e-02
6.38381064e-01 -1.38832664e+00 -8.46804321e-01 7.07897902e-01
-1.47845173e+00 -1.53568938e-01 -7.91426376e-02 7.74033308e-01
-2.62947261e-01 3.05080473e-01 -8.25012743e-01 -1.53331256e+00
-6.29126787e-01 -1.05002713e+00 1.12099338e+00 -1.53182540e-02
-4.08872396e-01 -8.10532749e-01 1.15007482e-01 5.71792245e-01
4.02968079e-01 -4.05162036e-01 3.49835694e-01 -1.25568199e+00
-2.41739064e-01 -2.15910599e-01 2.20420420e-01 6.81155026e-01
3.31388652e-01 -2.23966122e-01 -1.34997857e+00 -3.37652653e-01
2.42754579e-01 -1.82923406e-01 7.11112440e-01 4.18661475e-01
7.03204930e-01 -4.52750385e-01 -1.02938041e-01 3.57619435e-01
6.90233052e-01 3.78325075e-01 3.04334790e-01 -5.00703275e-01
6.98065460e-01 9.02345598e-01 3.74802724e-02 1.93974435e-01
2.26448938e-01 6.13462806e-01 -1.55148193e-01 -8.99153352e-02
-3.24570239e-02 -1.13116667e-01 8.94831002e-01 1.78361571e+00
2.91417658e-01 -4.52670366e-01 -8.69832814e-01 7.43509233e-01
-1.76826346e+00 -1.11944282e+00 -1.22760326e-01 2.19719791e+00
7.95615613e-01 -2.00421154e-01 5.68789244e-01 5.21536350e-01
1.06555176e+00 1.61165118e-01 -2.50252038e-01 -4.49460864e-01
-1.14994310e-01 2.62958314e-02 2.02330813e-01 7.18700767e-01
-9.67981339e-01 8.13936710e-01 6.86217213e+00 9.23679471e-01
-1.36447752e+00 2.56052256e-01 5.73995650e-01 -3.88057649e-01
5.00919782e-02 -1.47476360e-01 -9.37012434e-01 3.48002821e-01
1.65055645e+00 5.09819575e-02 5.13606429e-01 3.94102454e-01
4.98335123e-01 1.91865712e-01 -1.24102044e+00 1.34577537e+00
1.50790617e-01 -1.00428200e+00 -2.60460049e-01 -1.06082849e-01
2.91960478e-01 2.36108646e-01 -1.06317967e-01 2.95031220e-01
2.92133331e-01 -8.59168530e-01 7.02964365e-01 -1.00860551e-01
5.64185739e-01 -6.02115333e-01 5.34599006e-01 6.71941102e-01
-1.35602307e+00 -1.57173604e-01 2.99284458e-01 7.63190761e-02
6.87216520e-01 6.53432369e-01 -1.40447116e+00 6.92256629e-01
3.49976093e-01 4.00850505e-01 3.16297382e-01 8.96229923e-01
6.44110888e-02 1.31571114e+00 -5.38926840e-01 1.24770731e-01
1.84964776e-01 -2.01887608e-01 8.62965703e-01 1.59368789e+00
4.24662501e-01 8.00744966e-02 1.66703358e-01 2.15346783e-01
-2.30437517e-01 5.50411418e-02 -2.54944801e-01 -2.41019696e-01
5.87574303e-01 1.17768550e+00 -5.50295293e-01 -5.60887337e-01
-2.33442008e-01 9.97985005e-01 -5.27967550e-02 4.57240403e-01
-4.26974386e-01 -4.17133659e-01 7.00682461e-01 -1.24053493e-01
4.89069849e-01 -2.62493134e-01 1.53105959e-01 -1.14556777e+00
1.18475556e-01 -1.22122955e+00 2.51384020e-01 -3.17075729e-01
-1.11680448e+00 9.33352172e-01 -9.33915451e-02 -1.07965052e+00
-7.63126433e-01 -6.86716139e-02 -8.86099935e-01 1.06325436e+00
-1.23048902e+00 -1.18186963e+00 1.88045412e-01 6.01826847e-01
1.10353875e+00 -2.82879233e-01 8.07747245e-01 4.89771277e-01
-9.53975976e-01 4.46204215e-01 3.99437964e-01 2.60376304e-01
5.18263638e-01 -1.00505924e+00 6.68487012e-01 8.03914607e-01
4.31683004e-01 4.83902723e-01 6.65874898e-01 -2.09797129e-01
-1.29023719e+00 -9.08810973e-01 1.12779176e+00 -1.01084504e-02
6.96016014e-01 -8.32648218e-01 -9.65924203e-01 8.38874161e-01
3.61974359e-01 -2.85414994e-01 9.90013599e-01 1.94148690e-01
-8.90122727e-02 -3.34069848e-01 -6.93896115e-01 3.06488186e-01
7.10468411e-01 -9.41595793e-01 -6.58831000e-01 1.72940478e-01
7.95648754e-01 -3.84739667e-01 -7.11709738e-01 -1.87393546e-01
3.66434276e-01 -7.28267968e-01 7.92382836e-01 -2.31050447e-01
-2.38686994e-01 -1.63417995e-01 -2.80279517e-01 -1.44904995e+00
1.55031443e-01 -1.45675099e+00 -9.55804531e-03 1.74185753e+00
5.36374390e-01 -6.12884343e-01 6.56658769e-01 1.44051835e-01
-3.09819490e-01 -1.31211802e-01 -1.39045620e+00 -8.19240153e-01
-4.07867789e-01 -5.52807987e-01 6.39021695e-01 7.58302331e-01
1.96243659e-01 8.45349073e-01 -7.19454408e-01 4.68995064e-01
4.87709373e-01 -1.24588124e-01 7.08908975e-01 -9.26795483e-01
-5.36636055e-01 -2.63660610e-01 -7.81924725e-02 -1.49184024e+00
3.06271851e-01 -6.25146627e-01 6.02424562e-01 -1.12772703e+00
-3.30958992e-01 -1.52782828e-01 -6.47141412e-02 2.08614260e-01
6.13774322e-02 -2.48943001e-01 9.23460200e-02 4.85162675e-01
-3.38983089e-01 4.60766703e-01 4.55468833e-01 -1.25033885e-01
-5.35165012e-01 5.57857394e-01 -3.42833906e-01 4.69052315e-01
6.46108866e-01 -3.97350401e-01 -3.76967013e-01 -3.97153050e-01
-4.92204726e-01 6.68877959e-01 1.54214576e-02 -8.17928553e-01
5.73663652e-01 3.63904387e-01 -2.74280041e-01 -7.56828785e-01
7.33292043e-01 -7.06806660e-01 6.11509979e-02 8.84138048e-02
-5.31904638e-01 -1.15532242e-01 7.51584172e-02 4.54350859e-01
-7.14129627e-01 -1.36415303e-01 4.47344124e-01 8.17197114e-02
-1.82468370e-01 1.19982958e-01 -9.24521208e-01 -2.25281268e-01
6.67448819e-01 1.02095410e-01 1.63455158e-01 -7.58399248e-01
-9.94551420e-01 1.09888569e-01 -2.77484208e-01 3.44424963e-01
5.29223680e-01 -9.47911561e-01 -1.16900492e+00 5.72902739e-01
-1.50968209e-01 3.12963933e-01 2.16813967e-01 9.57754314e-01
4.17884020e-03 3.01798940e-01 6.18607223e-01 -7.45444477e-01
-1.89156461e+00 2.12752014e-01 3.91594678e-01 -1.55563578e-01
-3.68462265e-01 1.08005106e+00 9.54526886e-02 -5.13624370e-01
5.49451888e-01 -3.97096664e-01 -7.53408447e-02 3.33588004e-01
6.99210763e-01 2.93602139e-01 3.44402552e-01 -9.62805033e-01
-3.12719047e-01 4.48277555e-02 -2.39699125e-01 -7.12080717e-01
1.29567099e+00 -4.46831524e-01 -2.37609774e-01 9.38937187e-01
1.28257847e+00 5.80678731e-02 -9.34620857e-01 -4.13155526e-01
9.35220346e-02 4.51188833e-02 9.52219591e-02 -3.04821759e-01
-7.48501062e-01 8.54095280e-01 2.16424182e-01 7.18430042e-01
1.01460946e+00 1.49596542e-01 9.71780419e-01 4.66160089e-01
-1.27117574e-01 -9.24150229e-01 -2.49019459e-01 6.45842969e-01
8.20337296e-01 -9.77853775e-01 -6.35782838e-01 -4.55651790e-01
-5.68700612e-01 1.04806662e+00 -3.96982841e-02 3.90096962e-01
7.59455323e-01 5.05976379e-01 4.25226510e-01 2.35490844e-01
-9.60377872e-01 -7.57106766e-02 2.16054507e-02 3.15519303e-01
6.72786653e-01 5.89793101e-02 5.03116131e-01 4.57366109e-01
-2.21190527e-01 -2.51537830e-01 2.99907655e-01 9.70174670e-01
-2.95019895e-01 -1.34569132e+00 -1.00243080e+00 2.32197940e-01
-5.34148514e-01 -2.89031386e-01 -4.64960217e-01 -9.21801478e-02
-3.79102170e-01 1.54580462e+00 1.33852124e-01 -3.63609076e-01
2.95255512e-01 4.00313050e-01 2.85364240e-02 -7.49789894e-01
-1.10465229e+00 8.07503402e-01 4.77322668e-01 -3.71984392e-03
-5.50231636e-01 -8.95111442e-01 -1.04328179e+00 -8.79529789e-02
-5.44886470e-01 6.00857496e-01 8.74550402e-01 1.07675803e+00
3.04941893e-01 6.80846453e-01 1.04433393e+00 -1.08847404e+00
-7.33151793e-01 -1.36365163e+00 -9.12626266e-01 1.15481086e-01
9.52929318e-01 1.56487525e-01 -4.83294934e-01 3.07145417e-01] | [14.617280960083008, 6.333874225616455] |
6865a6a0-1b44-4c83-86ff-d88cc86be734 | tensorize-factorize-and-regularize-robust | null | null | http://openaccess.thecvf.com/content_cvpr_2018/html/Hwang_Tensorize_Factorize_and_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Hwang_Tensorize_Factorize_and_CVPR_2018_paper.pdf | Tensorize, Factorize and Regularize: Robust Visual Relationship Learning | Visual relationships provide higher-level information of objects and their relations in an image â this enables a semantic understanding of the scene and helps downstream applications. Given a set of localized objects in some training data, visual relationship detection seeks to detect the most likely ârelationshipâ between objects in a given image. While the specific objects may be well represented in training data, their relationships may still be infrequent. The empirical distribution obtained from seeing these relationships in a dataset does not model the underlying distribution well â a serious issue for most learning methods. In this work, we start from a simple multi-relational learning model, which in principle, offers a rich formalization for deriving a strong prior for learning visual relationships. While the inference problem for deriving the regularizer is challenging, our main technical contribution is to show how adapting recent results in numerical linear algebra lead to efficient algorithms for a factorization scheme that yields highly informative priors. The factorization provides sample size bounds for inference (under mild conditions) for the underlying [[object, predicate, object]] relationship learning task on its own and surprisingly outperforms (in some cases) existing methods even without utilizing visual features. Then, when integrated with an end to-end architecture for visual relationship detection leveraging image data, we substantially improve the state-of-the-art. | ['Sathya N. Ravi', 'Vikas Singh', 'Seong Jae Hwang', 'Maxwell D. Collins', 'Hyunwoo J. Kim', 'Zirui Tao'] | 2018-06-01 | null | null | null | cvpr-2018-6 | ['visual-relationship-detection'] | ['computer-vision'] | [ 1.71117276e-01 1.36348903e-01 -3.70237917e-01 -3.61869603e-01
-7.07730353e-01 -7.00448871e-01 6.60157859e-01 4.51878786e-01
-9.98770520e-02 3.99976432e-01 8.69862363e-03 -1.88267261e-01
-4.48420137e-01 -6.61087632e-01 -1.01551199e+00 -7.37771630e-01
-2.27116212e-01 6.08951032e-01 2.75570422e-01 1.21672057e-01
-8.98810755e-03 6.37457907e-01 -1.61684787e+00 4.81761456e-01
3.12867433e-01 9.84979153e-01 3.86173099e-01 7.52475679e-01
1.67957157e-01 9.52701569e-01 -2.05997184e-01 -4.71807092e-01
3.03585768e-01 -1.97804987e-01 -8.75341177e-01 4.21197951e-01
9.29655135e-01 -2.79143691e-01 -3.16810727e-01 1.02748358e+00
1.15947939e-01 2.46386528e-01 1.01711571e+00 -1.42853880e+00
-7.08525658e-01 4.79039937e-01 -9.79722261e-01 5.14220715e-01
1.88204467e-01 -1.23876065e-01 1.68077779e+00 -1.02854609e+00
6.82685852e-01 1.42927158e+00 4.26637620e-01 -7.14396685e-03
-1.59317875e+00 -2.50852346e-01 3.85277629e-01 2.66091704e-01
-1.43650746e+00 -3.15921754e-01 5.58006763e-01 -5.46475112e-01
9.27051663e-01 2.18669295e-01 5.58845043e-01 7.12029219e-01
-8.30561817e-02 1.05762911e+00 8.38678002e-01 -5.00816882e-01
-1.07963346e-01 3.54937047e-01 2.78869271e-01 8.03238511e-01
3.06746125e-01 -1.87479079e-01 -8.30884457e-01 -2.15192899e-01
8.11290562e-01 3.87922190e-02 -1.71697542e-01 -8.36530447e-01
-1.17026138e+00 7.80284405e-01 6.80657864e-01 8.80490541e-02
-2.13908628e-01 2.94444472e-01 1.04833119e-01 1.87125951e-01
2.40590632e-01 4.90604304e-02 -3.22817206e-01 3.39624077e-01
-6.46139085e-01 4.24672842e-01 7.37224460e-01 1.11218786e+00
1.02431560e+00 -4.83647823e-01 -1.30451694e-01 6.61909044e-01
4.52698678e-01 2.78367043e-01 -1.47139683e-01 -8.99915814e-01
4.31851953e-01 4.20324683e-01 9.10831243e-03 -1.25231683e+00
-1.84788272e-01 -3.79293323e-01 -6.01997554e-01 2.22544447e-01
6.76032126e-01 1.81328088e-01 -5.29110849e-01 1.81973958e+00
4.19487625e-01 1.55873105e-01 -6.49316087e-02 7.96536863e-01
8.04911911e-01 6.18268192e-01 -9.95219126e-02 -2.13722602e-01
1.52302122e+00 -7.13413179e-01 -4.67908025e-01 -4.10021544e-01
4.54760224e-01 -7.48894274e-01 9.87778306e-01 3.08560461e-01
-9.96347070e-01 -5.73418558e-01 -7.83461332e-01 -4.51196581e-01
-2.23136455e-01 2.09123030e-01 1.11655891e+00 1.75282180e-01
-1.04331982e+00 3.01121205e-01 -5.93773305e-01 -6.51865840e-01
7.34882593e-01 2.89887995e-01 -5.33204854e-01 -5.93000129e-02
-6.66468143e-01 9.25660491e-01 2.79754102e-01 -2.30846982e-02
-1.01939547e+00 -7.34240413e-01 -8.48303795e-01 8.74352157e-02
6.96248174e-01 -9.00984883e-01 1.16475248e+00 -8.45143795e-01
-6.18983865e-01 1.17614651e+00 -5.19970417e-01 -4.58439440e-01
4.38115537e-01 -3.19923341e-01 1.28550753e-01 4.63884920e-01
2.56987572e-01 8.00614238e-01 9.44534302e-01 -1.61015141e+00
-7.99659193e-01 -4.36421216e-01 4.25319076e-01 3.92852724e-01
-8.52344781e-02 1.16868727e-01 -7.64986515e-01 -4.04526681e-01
1.19838834e-01 -7.66677618e-01 -1.62723690e-01 7.16433704e-01
-4.02841419e-01 -3.34329039e-01 7.59265065e-01 -1.96248874e-01
6.35362923e-01 -2.15747070e+00 1.24387741e-01 3.23386461e-01
4.05884832e-01 -3.82440053e-02 7.74656832e-02 3.26337397e-01
-3.30561638e-01 -1.29913449e-01 2.02725064e-02 -4.55153912e-01
-1.49378881e-01 3.40283513e-01 -6.91767931e-01 7.92855561e-01
3.64605665e-01 9.22156334e-01 -8.76671076e-01 -7.13425934e-01
2.78419435e-01 4.75528926e-01 -5.92914581e-01 3.87931883e-01
-2.69258559e-01 2.10568190e-01 -3.19019288e-01 4.53732729e-01
4.74153370e-01 -7.85888433e-01 4.34193090e-02 -5.87578773e-01
1.76692292e-01 -2.40254533e-02 -1.33523715e+00 1.54685140e+00
-3.29129279e-01 7.61767447e-01 7.05529228e-02 -1.24100435e+00
7.03999579e-01 4.04586345e-02 4.69154865e-01 -2.38460630e-01
-1.17847852e-01 -2.61214405e-01 8.50606635e-02 -6.67090356e-01
2.68267572e-01 -2.52877206e-01 2.57474184e-01 2.65415102e-01
1.50352716e-01 -3.60879786e-02 3.23136419e-01 6.03365362e-01
9.05140579e-01 2.11227641e-01 5.28701305e-01 -2.09085882e-01
3.01807433e-01 -4.26733084e-02 2.52761155e-01 9.60925221e-01
-4.67677880e-03 6.00906372e-01 8.14601898e-01 -1.58401281e-01
-7.81536341e-01 -1.34230828e+00 -2.60449767e-01 1.09211802e+00
3.02562892e-01 -6.62643850e-01 -8.29856396e-02 -5.96974969e-01
2.46735498e-01 3.21322143e-01 -7.37732470e-01 1.10074006e-01
-3.13727796e-01 -6.43914163e-01 1.47664219e-01 6.88856781e-01
1.85218960e-01 -6.57833517e-01 -6.05653584e-01 -2.13690117e-01
2.71338839e-02 -1.52535737e+00 -2.82510906e-01 4.02585238e-01
-6.27214134e-01 -1.26302969e+00 -3.62043321e-01 -7.86910653e-01
8.87782574e-01 5.42189002e-01 1.24419534e+00 -5.45062274e-02
-4.89218563e-01 6.46108627e-01 1.12288876e-03 -3.12455863e-01
-1.07797682e-01 -5.88873982e-01 -1.54710919e-01 1.64440334e-01
2.89767176e-01 -6.11092865e-01 -4.17550594e-01 1.96968347e-01
-6.52253568e-01 7.49144554e-02 7.41580427e-01 8.00411403e-01
7.70913184e-01 2.43482873e-01 7.34582618e-02 -9.18148100e-01
1.00683831e-01 -4.78070140e-01 -6.77140534e-01 4.45650637e-01
-2.77308226e-01 1.77557215e-01 4.37770545e-01 -4.37445164e-01
-9.38657939e-01 3.35369378e-01 4.34193611e-01 -7.54517972e-01
-1.95170194e-01 4.69182789e-01 -2.13580564e-01 2.30973229e-01
7.47511625e-01 -4.61234264e-02 4.43596058e-02 -2.91186601e-01
8.25874031e-01 1.10155717e-01 6.15015149e-01 -7.86140621e-01
9.38406408e-01 9.56534922e-01 5.20168364e-01 -8.93986702e-01
-1.33781552e+00 -9.08115864e-01 -8.79993856e-01 -1.43613219e-01
8.22469413e-01 -1.13954306e+00 -9.80027318e-01 -2.72479385e-01
-1.20540547e+00 -1.93015397e-01 -2.43131414e-01 3.42557013e-01
-7.06065834e-01 3.86023998e-01 -3.01147640e-01 -9.41244185e-01
2.87129104e-01 -8.59797835e-01 1.12926555e+00 4.29293178e-02
-2.28667051e-01 -1.14133596e+00 -2.30269641e-01 3.28292757e-01
-2.37603337e-01 1.09795302e-01 1.15244961e+00 -4.50948864e-01
-1.04219186e+00 8.77585113e-02 -6.76718533e-01 1.69545963e-01
7.88043141e-02 1.19548582e-01 -1.01440775e+00 -1.11489795e-01
-1.78886428e-01 -5.35475671e-01 1.10811973e+00 5.33431947e-01
1.21657181e+00 -1.11534834e-01 -5.83729625e-01 4.82860327e-01
1.44752157e+00 -3.16634804e-01 2.22443983e-01 -1.25671089e-01
9.58480477e-01 9.28878069e-01 7.59962559e-01 4.14266437e-01
3.90170425e-01 7.23587871e-01 6.05476558e-01 -1.85851380e-01
-2.36319989e-01 -2.06740171e-01 1.77048489e-01 -2.42408980e-02
-3.21630649e-02 -1.06998503e-01 -7.77016878e-01 6.86023355e-01
-2.15534925e+00 -1.09044135e+00 -3.27833325e-01 2.23316503e+00
9.54294801e-01 8.71320516e-02 2.93742299e-01 -7.87208825e-02
6.71168685e-01 3.94715033e-02 -5.79666615e-01 9.86792445e-02
-2.36206010e-01 -1.87241621e-02 4.38454866e-01 5.83635867e-01
-1.19147348e+00 8.04137111e-01 6.56507349e+00 6.68155253e-01
-6.95633531e-01 -1.86447933e-01 6.15868270e-01 -9.85248387e-02
-3.94968987e-01 3.02975774e-01 -1.21302485e+00 -1.78026572e-01
4.16394413e-01 -1.95773557e-01 2.97565699e-01 8.62420797e-01
1.42463610e-01 -3.31000239e-01 -1.65191376e+00 1.25888503e+00
2.75864214e-01 -1.36419284e+00 9.11799744e-02 1.49503469e-01
4.33899850e-01 -1.39391974e-01 2.64546722e-01 -3.98614481e-02
4.48421299e-01 -1.16188085e+00 7.58271396e-01 3.68672431e-01
4.76630002e-01 -6.25639379e-01 5.74085377e-02 3.91418755e-01
-1.22345781e+00 9.45867132e-03 -5.16103148e-01 -7.93243274e-02
4.75915745e-02 7.90615201e-01 -9.55557168e-01 5.40183723e-01
6.70726895e-01 1.13222098e+00 -6.93917871e-01 9.23708797e-01
-3.00675392e-01 3.23567659e-01 -5.00821292e-01 6.18678741e-02
4.61698994e-02 -8.56292173e-02 5.39356530e-01 1.16999137e+00
-1.01592608e-01 -2.46334150e-02 1.90979272e-01 1.13431025e+00
2.15987861e-03 -4.19809893e-02 -8.92902076e-01 1.25998929e-01
1.54332876e-01 1.53896856e+00 -8.66948962e-01 -1.37157917e-01
-7.01834023e-01 7.15460479e-01 8.06983054e-01 5.61247706e-01
-7.63568759e-01 1.99312821e-01 6.13355815e-01 1.86557651e-01
6.57458246e-01 -2.50411332e-01 -1.52970910e-01 -1.15931320e+00
1.97095782e-01 -7.50839293e-01 7.21595287e-01 -8.39235306e-01
-1.52223539e+00 -1.13210606e-03 4.26706940e-01 -1.09738791e+00
-2.98835725e-01 -9.45984423e-01 -3.64590347e-01 8.27818573e-01
-1.50351095e+00 -1.35947239e+00 -1.58927009e-01 7.10092425e-01
3.61662477e-01 7.22479373e-02 6.61830544e-01 -1.69995725e-02
-2.69955516e-01 2.76989132e-01 -2.77544200e-01 9.80799422e-02
7.49237061e-01 -1.52590549e+00 -2.19158828e-01 1.06808674e+00
7.20646262e-01 8.64308119e-01 8.87023628e-01 -4.54391807e-01
-1.49673271e+00 -8.91116917e-01 7.85269439e-01 -6.59838498e-01
9.73377109e-01 -4.69083548e-01 -8.01006973e-01 9.71631527e-01
6.97491467e-02 6.69066191e-01 6.93822503e-01 5.47423065e-01
-7.52556145e-01 -1.78277463e-01 -7.50813723e-01 7.40112662e-01
1.18090355e+00 -7.37148583e-01 -5.32571137e-01 6.24549329e-01
3.79672557e-01 -4.00763363e-01 -6.57287121e-01 1.62710384e-01
4.26895678e-01 -1.15054417e+00 1.31650662e+00 -8.04942250e-01
6.66303933e-01 -5.75783789e-01 -4.61766005e-01 -8.37212384e-01
-2.62332618e-01 -5.56432545e-01 -4.31998968e-01 1.32180166e+00
2.68757910e-01 -1.72170296e-01 7.50048697e-01 6.64048851e-01
2.70300001e-01 -8.30709159e-01 -6.09312236e-01 -7.72148073e-01
-1.41083479e-01 -5.73676527e-01 -1.69645175e-01 7.34713256e-01
-1.34248987e-01 7.05416024e-01 -2.47874975e-01 6.48911774e-01
8.81798327e-01 5.29699028e-01 8.91695201e-01 -1.06156445e+00
-8.32758427e-01 -4.07589197e-01 -6.28522575e-01 -1.36735559e+00
4.20967102e-01 -9.66379702e-01 -1.10700503e-01 -1.60035348e+00
6.23257518e-01 -4.17327404e-01 -7.24823475e-02 5.78626335e-01
-3.05696100e-01 8.20483118e-02 2.18189240e-01 1.97268575e-01
-7.88217664e-01 2.88195789e-01 1.22899485e+00 -1.69070646e-01
1.55067757e-01 1.71220943e-01 -8.66862535e-01 7.74226606e-01
2.29563132e-01 -2.57085860e-01 -7.81490624e-01 -2.75012910e-01
5.45894802e-01 -5.29613905e-02 9.49501574e-01 -4.27405149e-01
3.78042907e-01 -9.84705389e-02 4.95565504e-01 -7.83258915e-01
4.80769902e-01 -9.52378869e-01 -3.56271453e-02 1.56796455e-01
-6.77829444e-01 -1.17678486e-01 7.77344778e-02 9.17490661e-01
-9.60725993e-02 -2.91831106e-01 7.41166115e-01 -8.49898309e-02
-7.93397725e-01 4.31317508e-01 -4.36173193e-02 1.97353944e-01
1.08737433e+00 -8.74274671e-02 -2.62988091e-01 -5.28829575e-01
-8.27631950e-01 1.89539209e-01 2.46247679e-01 1.59458339e-01
5.59679389e-01 -1.20007205e+00 -5.72745383e-01 -1.22568347e-01
3.61271054e-01 3.53280336e-01 -1.39503879e-02 8.15383911e-01
-2.42831167e-02 5.50018772e-02 8.49387646e-02 -9.60940421e-01
-1.54708719e+00 7.57475555e-01 2.17721522e-01 -5.01433089e-02
-6.71492040e-01 1.12123775e+00 7.99699128e-01 2.19826967e-01
3.83311391e-01 -2.87312269e-01 -2.15353921e-01 3.04520518e-01
5.99664688e-01 -2.01205537e-02 -2.02060789e-01 -7.27434695e-01
-4.77110416e-01 5.30421913e-01 -2.35612795e-01 -1.36691824e-01
1.34288263e+00 -1.82332397e-01 -2.03761682e-01 8.02552521e-01
1.32283759e+00 1.79502293e-02 -1.49501145e+00 -5.54536164e-01
-1.54365465e-01 -7.83333063e-01 -2.08595331e-04 -4.13820446e-01
-1.00830841e+00 8.85249019e-01 2.30160028e-01 3.45080823e-01
8.96713495e-01 7.27280915e-01 1.30424956e-02 5.77040493e-01
1.00674957e-01 -5.28039157e-01 3.88891995e-01 1.74818635e-01
9.22993362e-01 -1.54144013e+00 4.49875832e-01 -9.58007276e-01
-6.66887701e-01 1.20928514e+00 4.47290421e-01 -1.11992404e-01
8.18971515e-01 4.06487137e-01 -1.70950502e-01 -4.56846565e-01
-9.29255426e-01 -7.05395937e-01 5.06945550e-01 8.09736311e-01
3.35016578e-01 -2.96726525e-01 2.68372059e-01 9.12180915e-02
6.31974777e-03 -4.29891109e-01 2.66671926e-01 5.39239764e-01
-3.03828239e-01 -9.60872054e-01 -1.24808855e-01 4.13294792e-01
-4.21529591e-01 -1.85642630e-01 -2.52668291e-01 8.28003705e-01
1.59308508e-01 9.01711643e-01 2.21602887e-01 1.86129972e-01
1.34978846e-01 -4.09790903e-01 9.36634302e-01 -7.83976793e-01
-1.55634284e-01 1.98526025e-01 1.21334240e-01 -5.83529770e-01
-8.86542916e-01 -1.02236009e+00 -1.41026974e+00 2.47304887e-02
-3.02524418e-01 -3.46085131e-01 1.60600528e-01 1.10185468e+00
1.17523365e-01 2.34901279e-01 3.91671985e-01 -6.93363965e-01
-3.60694617e-01 -4.54183936e-01 -7.12394178e-01 5.81713498e-01
4.89563584e-01 -7.89450467e-01 -3.30585927e-01 3.88911307e-01] | [10.217864036560059, 1.6609879732131958] |
1223fc86-6165-494c-a817-982083c269dd | posegpt-quantization-based-3d-human-motion | 2210.10542 | null | https://arxiv.org/abs/2210.10542v1 | https://arxiv.org/pdf/2210.10542v1.pdf | PoseGPT: Quantization-based 3D Human Motion Generation and Forecasting | We address the problem of action-conditioned generation of human motion sequences. Existing work falls into two categories: forecast models conditioned on observed past motions, or generative models conditioned on action labels and duration only. In contrast, we generate motion conditioned on observations of arbitrary length, including none. To solve this generalized problem, we propose PoseGPT, an auto-regressive transformer-based approach which internally compresses human motion into quantized latent sequences. An auto-encoder first maps human motion to latent index sequences in a discrete space, and vice-versa. Inspired by the Generative Pretrained Transformer (GPT), we propose to train a GPT-like model for next-index prediction in that space; this allows PoseGPT to output distributions on possible futures, with or without conditioning on past motion. The discrete and compressed nature of the latent space allows the GPT-like model to focus on long-range signal, as it removes low-level redundancy in the input signal. Predicting discrete indices also alleviates the common pitfall of predicting averaged poses, a typical failure case when regressing continuous values, as the average of discrete targets is not a target itself. Our experimental results show that our proposed approach achieves state-of-the-art results on HumanAct12, a standard but small scale dataset, as well as on BABEL, a recent large scale MoCap dataset, and on GRAB, a human-object interactions dataset. | ['Grégory Rogez', 'Philippe Weinzaepfel', 'Fabien Baradel', 'Thomas Lucas'] | 2022-10-19 | null | null | null | null | ['human-object-interaction-detection'] | ['computer-vision'] | [ 3.34624499e-01 8.27401280e-02 -1.65510714e-01 -3.01575452e-01
-9.32781756e-01 -3.21394235e-01 7.53076375e-01 -7.36286521e-01
-3.57027769e-01 6.80703461e-01 7.61153102e-01 2.41265863e-01
1.67810738e-01 -8.45427513e-01 -8.27378094e-01 -9.17856514e-01
-2.34728858e-01 7.52027452e-01 2.79975176e-01 -4.52447720e-02
-2.26842657e-01 1.51655167e-01 -1.50451970e+00 5.27988672e-01
3.50930780e-01 7.83283710e-01 4.44916606e-01 9.81974065e-01
5.36413550e-01 1.17427742e+00 -5.04613817e-01 -2.51190722e-01
1.14384644e-01 -9.68383789e-01 -5.96520901e-01 1.77048326e-01
2.59727269e-01 -5.03037870e-01 -8.32892060e-01 6.15283847e-01
3.91239852e-01 3.78057271e-01 8.18955243e-01 -1.38890052e+00
-6.91679239e-01 7.88072824e-01 -1.22969739e-01 5.46039082e-02
5.27410984e-01 6.06364429e-01 1.20577574e+00 -6.64335072e-01
7.86721826e-01 1.43806887e+00 5.35572469e-01 8.79078448e-01
-1.26596940e+00 -4.52730149e-01 1.31328732e-01 3.08230162e-01
-1.18099487e+00 -3.17794472e-01 7.12554753e-01 -5.77782035e-01
1.15882075e+00 2.99171120e-01 8.25542510e-01 1.92316139e+00
4.75681633e-01 1.11052442e+00 6.00331128e-01 -1.03630526e-02
3.35968792e-01 -5.91293514e-01 -3.93491805e-01 4.08416957e-01
-3.72806817e-01 3.79282802e-01 -8.61233950e-01 -2.82557551e-02
8.21978092e-01 7.47496188e-02 -3.56524318e-01 -1.49021342e-01
-1.67116678e+00 7.87115574e-01 3.02980185e-01 3.62411924e-02
-6.72797143e-01 5.05290985e-01 7.77997449e-02 -7.60825723e-02
2.17830271e-01 1.38336331e-01 -1.76634490e-01 -4.86433417e-01
-1.11701918e+00 5.74913323e-01 6.44777119e-01 1.01648259e+00
5.69839478e-01 2.67260015e-01 -6.19076014e-01 4.49841529e-01
3.15929621e-01 7.83357918e-01 7.23636091e-01 -1.04884696e+00
6.15367532e-01 -1.94959775e-01 1.96805075e-01 -8.43514144e-01
-3.01718026e-01 -3.81674498e-01 -8.61733258e-01 -3.50937396e-02
3.29559654e-01 -2.59580135e-01 -1.22679639e+00 2.15657544e+00
-1.13905333e-02 4.87584263e-01 -4.76528099e-03 1.15630364e+00
2.81438917e-01 1.03287542e+00 1.75856188e-01 -1.58907130e-01
1.03724742e+00 -1.05136395e+00 -7.21293449e-01 -4.49929833e-01
4.20288920e-01 -4.06321615e-01 9.26933587e-01 4.86699224e-01
-9.70378637e-01 -8.90918434e-01 -8.49205434e-01 -2.69139968e-02
8.78643021e-02 1.75754532e-01 4.88407254e-01 2.19591096e-01
-9.60888088e-01 8.23920488e-01 -1.34579468e+00 -1.94426417e-01
1.29815210e-02 1.00364655e-01 -2.48454928e-01 4.73184027e-02
-1.22118676e+00 7.09772229e-01 4.41354275e-01 1.25671819e-01
-1.28330648e+00 -4.62729633e-01 -9.56534564e-01 -2.25728396e-02
1.83936179e-01 -9.63315785e-01 1.32083666e+00 -4.39563453e-01
-1.52007449e+00 3.52075487e-01 -3.04482490e-01 -9.59974587e-01
6.94001913e-01 -5.81611991e-01 -4.11307573e-01 1.32988781e-01
2.78629243e-01 1.04929984e+00 9.93738115e-01 -7.44069338e-01
-6.05953693e-01 -1.79398935e-02 -3.34604055e-01 2.74026155e-01
-4.56193015e-02 -4.18035090e-01 -4.76702303e-01 -9.72245812e-01
-2.05383208e-02 -1.24923718e+00 -3.45755428e-01 -4.52002466e-01
-5.10218322e-01 -1.20773055e-01 8.50236118e-01 -7.62309909e-01
1.24279702e+00 -2.05289674e+00 7.44679868e-01 -6.31762668e-02
-1.69951051e-01 -2.13846818e-01 -2.18012750e-01 5.29033005e-01
-4.99853939e-02 -1.56729832e-01 -2.91597575e-01 -6.55858099e-01
2.01788172e-01 4.45852429e-01 -7.60385931e-01 2.73192823e-01
3.37051719e-01 1.11326218e+00 -1.03006804e+00 -3.80619645e-01
2.96475083e-01 5.29037237e-01 -7.23675609e-01 4.07592386e-01
-5.49708426e-01 8.15753520e-01 -2.77455866e-01 3.82908016e-01
1.60110489e-01 -2.57886231e-01 -2.01957161e-03 -9.10202861e-02
1.29541576e-01 3.61204058e-01 -8.67942810e-01 2.00593114e+00
-2.25663275e-01 6.04735792e-01 -7.23933220e-01 -7.50630736e-01
8.07150066e-01 4.96586025e-01 7.25224018e-01 -4.49065745e-01
-3.71055268e-02 -1.50991097e-01 -3.52540128e-02 -4.54526752e-01
5.20434678e-01 -2.05228731e-01 -4.26798195e-01 3.61811936e-01
3.03408414e-01 -8.27442948e-03 1.22727178e-01 9.39774588e-02
1.35066330e+00 7.62223005e-01 4.66559967e-03 3.68186235e-01
1.30515471e-01 -4.75517809e-02 6.20356679e-01 6.64424539e-01
1.03995658e-01 1.19402492e+00 3.57988089e-01 -2.56345183e-01
-1.16118836e+00 -1.27275884e+00 4.37869996e-01 9.46777582e-01
-1.46995202e-01 -4.92007762e-01 -7.10217237e-01 -5.35043597e-01
-2.36421779e-01 1.02299094e+00 -8.01046908e-01 -3.32143337e-01
-1.06255937e+00 -6.82031274e-01 3.99231791e-01 7.63642490e-01
2.93993860e-01 -1.49269295e+00 -1.01929283e+00 5.75914860e-01
-8.01684856e-01 -1.34104168e+00 -6.32668972e-01 7.71863163e-02
-6.82619929e-01 -5.78025281e-01 -9.23848152e-01 -4.72339541e-01
2.38717750e-01 -1.81940034e-01 1.05643833e+00 -5.47727108e-01
-7.19717070e-02 2.85172671e-01 -4.67976928e-01 -4.78745475e-02
-3.90442163e-01 -3.59242931e-02 9.61025432e-03 2.31459752e-01
2.22303063e-01 -6.88549578e-01 -6.80326521e-01 3.34990501e-01
-6.79426968e-01 3.22182178e-01 6.64046407e-01 9.80987132e-01
6.11366928e-01 -1.48255080e-01 2.70555526e-01 -4.63908881e-01
1.40078828e-01 -6.26388252e-01 -2.63229042e-01 -1.98598802e-02
4.57913242e-02 2.71932662e-01 4.90779281e-01 -8.36802125e-01
-1.09777129e+00 4.35140163e-01 -2.92620569e-01 -9.92077827e-01
-1.86761603e-01 3.79935890e-01 -1.23154633e-01 7.29467332e-01
7.40494668e-01 6.26055419e-01 -2.53442943e-01 -3.91779512e-01
4.73648340e-01 1.57808304e-01 1.15555823e+00 -3.64538670e-01
7.41428077e-01 5.33107460e-01 8.88332874e-02 -7.25413322e-01
-6.35856032e-01 -3.39421600e-01 -6.89979970e-01 -2.49599159e-01
1.29366136e+00 -1.02984798e+00 -4.97604787e-01 4.45796639e-01
-1.25824368e+00 -6.63838983e-01 -5.07561982e-01 7.97092736e-01
-1.27536988e+00 2.65028059e-01 -8.00143540e-01 -9.25938964e-01
-8.16863552e-02 -9.76724446e-01 1.40215898e+00 -1.58571661e-01
-8.30062926e-01 -7.74824858e-01 2.74930149e-01 2.08893120e-01
4.91287783e-02 5.99004090e-01 4.66837943e-01 -4.17285055e-01
-9.11545277e-01 -2.46591941e-01 5.57450771e-01 1.86107099e-01
-3.50383334e-02 -2.72165865e-01 -8.08461547e-01 -2.15658292e-01
8.19130242e-02 -3.41407984e-01 1.01218665e+00 5.16829669e-01
9.33168590e-01 -3.34834516e-01 -4.30895418e-01 6.03430033e-01
9.25514579e-01 2.57260978e-01 8.69824171e-01 6.14530891e-02
7.64876664e-01 3.84349644e-01 8.95102918e-01 6.11090541e-01
2.83076912e-01 9.02883410e-01 3.78824055e-01 1.99592561e-01
-2.97350109e-01 -8.46302092e-01 7.55900562e-01 7.21560121e-01
-3.30122024e-01 -5.59264183e-01 -7.69828856e-01 6.39308512e-01
-2.13859177e+00 -1.42347705e+00 6.34697229e-02 1.99040389e+00
6.87647462e-01 1.73091829e-01 3.94692451e-01 -4.10098955e-03
3.93450439e-01 2.93125033e-01 -4.29946005e-01 2.57009298e-01
-1.28458053e-01 2.51824707e-01 3.07341427e-01 3.92015964e-01
-1.29980004e+00 9.90458250e-01 6.11090374e+00 7.95602918e-01
-1.18336165e+00 9.99100432e-02 3.53980064e-01 -2.78181791e-01
-1.25692382e-01 -7.64835700e-02 -9.40856099e-01 7.16718078e-01
1.15529668e+00 -3.92832458e-02 2.75545120e-01 8.52523625e-01
2.52009749e-01 8.87320861e-02 -1.51283836e+00 9.39186513e-01
1.03875861e-01 -1.15582657e+00 1.54759094e-01 2.19767183e-01
6.23437703e-01 4.65034805e-02 2.98685133e-01 5.19186199e-01
4.07560319e-01 -1.04727995e+00 1.31991422e+00 8.54067802e-01
6.36433601e-01 -5.56944251e-01 4.76525992e-01 7.01193810e-01
-1.21902013e+00 -4.56423014e-02 -1.60900071e-01 -4.49371710e-02
8.35245311e-01 1.73422694e-01 -8.50081384e-01 4.76736367e-01
6.48253679e-01 9.50887024e-01 -2.56230086e-01 8.50915790e-01
-3.84122401e-01 8.20857942e-01 -2.83434242e-01 1.68272227e-01
2.67334819e-01 -4.27697711e-02 7.13873267e-01 1.33775520e+00
7.03115046e-01 2.63174206e-01 2.50789791e-01 9.22416151e-01
4.36134845e-01 -6.08143389e-01 -6.23633087e-01 -6.87700137e-02
2.06577241e-01 9.21557844e-01 -5.70598066e-01 -4.23312485e-01
-2.15768740e-01 1.34336734e+00 1.18914768e-02 5.77798903e-01
-1.25180984e+00 8.09119940e-02 4.48512226e-01 1.07890740e-01
7.55272269e-01 -4.94755238e-01 4.04119715e-02 -1.25036001e+00
-3.85890789e-02 -8.15852225e-01 4.77870971e-01 -8.81784141e-01
-1.06473351e+00 7.72733927e-01 3.95731121e-01 -1.70759690e+00
-1.21171105e+00 -2.78612286e-01 -4.28574055e-01 8.19282055e-01
-8.15722823e-01 -1.29740667e+00 -9.37363878e-02 5.95434546e-01
8.61532331e-01 -5.00180833e-02 7.51415014e-01 7.80018233e-03
-9.58332345e-02 3.50588620e-01 -4.10357028e-01 2.31778800e-01
4.59667563e-01 -1.09322202e+00 8.07985365e-01 9.51725781e-01
5.84621549e-01 3.75849038e-01 1.03777361e+00 -9.18036103e-01
-1.18382335e+00 -1.35096049e+00 8.22547853e-01 -8.96621346e-01
5.08841515e-01 -4.91863668e-01 -8.02449167e-01 1.05977583e+00
1.41975209e-01 -1.58521011e-01 3.73858362e-01 -4.01019692e-01
2.11958308e-02 3.08260202e-01 -5.39039195e-01 6.84066951e-01
1.36088324e+00 -4.61183637e-01 -7.87625790e-01 2.65313447e-01
7.77002633e-01 -6.00393414e-01 -7.30830014e-01 5.13732612e-01
6.14394546e-01 -9.86176789e-01 1.14599955e+00 -4.76237446e-01
5.76317251e-01 -3.57614696e-01 -2.94253230e-01 -1.39924455e+00
-6.02647841e-01 -7.49893427e-01 -4.18051600e-01 8.02660227e-01
4.05546367e-01 -9.90000442e-02 1.04100621e+00 3.86129946e-01
-2.63986856e-01 -5.89688122e-01 -8.63580644e-01 -9.25862014e-01
-1.87157020e-01 -5.89687109e-01 4.19754595e-01 5.36124825e-01
-2.81603605e-01 3.77726287e-01 -1.06535077e+00 1.18633963e-01
6.69567704e-01 4.96330718e-03 1.00089431e+00 -7.30805397e-01
-1.02909029e+00 -1.22584037e-01 -5.29585063e-01 -1.55487859e+00
1.33891432e-02 -5.19887924e-01 5.32424152e-01 -1.27196193e+00
2.40522176e-02 8.59085843e-02 2.48927809e-02 4.24702138e-01
-1.79266196e-03 1.81960613e-01 4.44049835e-01 3.10982287e-01
-5.10667622e-01 9.21048820e-01 1.27016580e+00 -1.52075171e-01
-9.53647792e-02 1.64569899e-01 2.97526009e-02 6.93843424e-01
3.60703260e-01 -5.70221364e-01 -7.02069283e-01 -2.73015290e-01
-1.11216202e-01 6.01563752e-01 6.76357806e-01 -1.49428344e+00
3.23044360e-01 -3.54538709e-01 4.98406380e-01 -1.10286069e+00
9.21392262e-01 -5.42807817e-01 7.24969387e-01 5.19853652e-01
-3.26870620e-01 -3.33711803e-02 -1.59771025e-01 7.33630180e-01
-2.98023105e-01 2.79908419e-01 3.32078815e-01 -1.80683911e-01
-8.32114041e-01 4.95821923e-01 -3.94126058e-01 -5.83488569e-02
8.87112141e-01 -2.01691583e-01 2.70865113e-02 -8.03442180e-01
-1.19162965e+00 7.23003149e-02 1.35870710e-01 7.20792890e-01
7.34106243e-01 -1.59107804e+00 -7.12297201e-01 2.27600157e-01
-7.16213882e-02 1.12732565e-02 4.20567423e-01 6.82823181e-01
-1.65413946e-01 5.41094184e-01 -3.55595976e-01 -9.87187624e-01
-9.04549539e-01 5.92816353e-01 6.17728606e-02 -2.63795733e-01
-8.72141302e-01 8.48303080e-01 4.03894365e-01 8.46858025e-02
1.96324974e-01 -5.56850314e-01 -6.97782040e-02 -1.44107372e-01
4.04086858e-01 1.18868932e-01 -4.34394687e-01 -9.73031402e-01
-8.25485960e-02 4.87237036e-01 2.44525611e-01 -7.59152293e-01
1.20634937e+00 -1.55897653e-02 3.97962153e-01 8.70553374e-01
1.11552250e+00 -2.85829484e-01 -1.68735611e+00 -1.30045548e-01
-1.33939788e-01 -3.50683689e-01 -4.62126911e-01 -4.55186248e-01
-7.55086362e-01 9.68766987e-01 2.26446122e-01 -5.83011061e-02
1.07218230e+00 7.33743757e-02 1.04100597e+00 2.55378187e-01
6.58399224e-01 -8.65245521e-01 4.73394871e-01 5.55386722e-01
1.11726522e+00 -7.71835804e-01 -3.26378018e-01 -2.39281461e-01
-1.08435965e+00 8.11100721e-01 6.23614371e-01 -2.25347772e-01
4.33859169e-01 9.62819606e-02 -2.19786465e-01 8.16554725e-02
-1.14092481e+00 -2.20630184e-01 4.56512094e-01 6.55871511e-01
2.03678682e-01 2.46183574e-01 1.24584891e-01 5.86952984e-01
-7.81077385e-01 2.08471522e-01 2.22039416e-01 8.30442190e-01
-2.56237000e-01 -8.55981648e-01 -4.36927468e-01 1.02553144e-01
-1.96466610e-01 1.06799632e-01 -3.22984792e-02 6.16617918e-01
1.49476454e-01 6.45565689e-01 3.19273025e-01 -7.12848186e-01
2.83327729e-01 1.20527469e-01 4.12304699e-01 -5.46988606e-01
-8.86105932e-03 4.99941587e-01 9.89056677e-02 -8.54080439e-01
-3.13759774e-01 -9.15631771e-01 -1.23125088e+00 4.56349105e-02
1.61930650e-01 -7.10210949e-03 1.30161092e-01 9.00910378e-01
1.49858296e-01 6.37580037e-01 3.75077873e-01 -1.24096739e+00
-6.99833870e-01 -1.23430562e+00 -3.68824810e-01 7.98429370e-01
4.70150203e-01 -6.85901165e-01 -2.70729989e-01 5.88303626e-01] | [7.322096347808838, -0.12120560556650162] |
a6628893-5355-4dbf-a8e5-a2ac34130246 | rms-flownet-efficient-and-robust-multi-scale | 2204.00354 | null | https://arxiv.org/abs/2204.00354v1 | https://arxiv.org/pdf/2204.00354v1.pdf | RMS-FlowNet: Efficient and Robust Multi-Scale Scene Flow Estimation for Large-Scale Point Clouds | The proposed RMS-FlowNet is a novel end-to-end learning-based architecture for accurate and efficient scene flow estimation which can operate on point clouds of high density. For hierarchical scene flow estimation, the existing methods depend on either expensive Farthest-Point-Sampling (FPS) or structure-based scaling which decrease their ability to handle a large number of points. Unlike these methods, we base our fully supervised architecture on Random-Sampling (RS) for multiscale scene flow prediction. To this end, we propose a novel flow embedding design which can predict more robust scene flow in conjunction with RS. Exhibiting high accuracy, our RMS-FlowNet provides a faster prediction than state-of-the-art methods and works efficiently on consecutive dense point clouds of more than 250K points at once. Our comprehensive experiments verify the accuracy of RMS-FlowNet on the established FlyingThings3D data set with different point cloud densities and validate our design choices. Additionally, we show that our model presents a competitive ability to generalize towards the real-world scenes of KITTI data set without fine-tuning. | ['Didier Stricker', 'Mohammad-Ali Nikouei Mahani', 'René Schuster', 'Ramy Battrawy'] | 2022-04-01 | null | null | null | null | ['scene-flow-estimation'] | ['computer-vision'] | [-4.20189679e-01 -3.92858177e-01 -1.13361321e-01 -2.99641162e-01
-4.28934008e-01 -4.00581300e-01 4.67961371e-01 6.47943690e-02
-4.08295989e-01 5.18561780e-01 1.74867719e-01 -1.76388863e-02
-1.89037785e-01 -1.01488543e+00 -6.48038805e-01 -2.63571024e-01
-5.56763828e-01 5.09225726e-01 6.73024178e-01 -3.19170326e-01
1.69197768e-01 7.74341166e-01 -1.73802876e+00 -4.29472402e-02
9.21389580e-01 1.01221478e+00 2.11334571e-01 9.16866124e-01
-2.00317830e-01 9.18717802e-01 -4.26170677e-01 -2.42941216e-01
6.59908950e-01 1.17393747e-01 -7.98826158e-01 -1.89841375e-01
1.12895501e+00 -8.08442116e-01 -5.98703742e-01 4.33337063e-01
4.01259899e-01 2.52956688e-01 6.76302910e-01 -1.33554518e+00
-3.01020145e-01 -6.93096826e-03 -4.42428440e-01 4.44993377e-01
1.92606226e-01 5.59189379e-01 9.27969635e-01 -9.98278439e-01
4.79705215e-01 1.30701518e+00 9.23291147e-01 2.58228600e-01
-1.23357749e+00 -7.68071890e-01 1.64992824e-01 3.14498633e-01
-1.36996162e+00 -3.04902375e-01 8.00980091e-01 -6.89324737e-01
1.19195032e+00 -1.06504284e-01 1.12816679e+00 7.36363411e-01
-1.63180530e-02 6.54954195e-01 5.15603185e-01 2.53769964e-01
3.87106866e-01 -1.72744378e-01 -2.78283328e-01 7.88274884e-01
2.76552975e-01 2.30243638e-01 -6.28408372e-01 -4.54196669e-02
9.78945792e-01 9.80610326e-02 -4.52645302e-01 -7.18104124e-01
-1.46555591e+00 9.39890742e-01 1.18253183e+00 -1.81774914e-01
-2.91438639e-01 5.53500712e-01 4.83363479e-01 1.29847646e-01
5.82533538e-01 4.88042504e-01 -5.90488434e-01 -3.27472746e-01
-1.22084999e+00 4.33538258e-01 8.06733370e-01 1.06851697e+00
1.13002241e+00 1.99780360e-01 7.35016391e-02 4.02755946e-01
2.56218553e-01 6.78752244e-01 2.14924328e-02 -1.20191658e+00
5.81797302e-01 6.65046751e-01 1.78167984e-01 -1.39832580e+00
-4.78944570e-01 -4.34177727e-01 -9.28872168e-01 4.39992458e-01
3.67028952e-01 5.89629635e-02 -5.66033304e-01 1.42630732e+00
5.03238678e-01 8.29355955e-01 -2.29107235e-02 1.13792837e+00
6.86261117e-01 8.84495437e-01 -8.63042846e-03 2.10244462e-01
8.10861886e-01 -1.14866781e+00 -1.19387813e-01 -1.94212228e-01
8.13114583e-01 -3.76622975e-01 1.23441827e+00 1.27449721e-01
-8.69887948e-01 -7.80720830e-01 -9.65937436e-01 -3.24201971e-01
-2.54178464e-01 1.60995618e-01 9.33141887e-01 3.41460645e-01
-1.21312296e+00 8.23233426e-01 -1.05858815e+00 -3.75028372e-01
8.26723218e-01 3.50373417e-01 -5.04462481e-01 -1.92834660e-02
-6.33959293e-01 6.58211589e-01 1.34638473e-01 6.18002601e-02
-8.87210310e-01 -1.51171958e+00 -9.02956724e-01 2.17550501e-01
-1.18907563e-01 -1.23166990e+00 8.24748456e-01 -3.23513925e-01
-1.51221418e+00 5.76146960e-01 -5.25952019e-02 -7.32902169e-01
7.06430912e-01 -4.73874360e-01 5.93919009e-02 5.54986775e-01
1.51757941e-01 1.16061592e+00 7.69341826e-01 -1.02980757e+00
-7.04824805e-01 -1.25056058e-01 1.47445336e-01 1.35613650e-01
-3.45237732e-01 -5.91358900e-01 -8.56878534e-02 -2.44012177e-01
-1.53428584e-01 -7.98965514e-01 -3.12833995e-01 7.42232740e-01
7.55323330e-03 -1.62939742e-01 9.81417120e-01 -1.79068726e-02
9.15822029e-01 -2.11854935e+00 1.43309116e-01 -3.80788088e-01
5.83330750e-01 3.15660566e-01 -7.66457170e-02 4.38170552e-01
2.65618950e-01 -6.26673847e-02 -3.25401574e-01 -5.42046547e-01
-8.00417364e-02 1.73129678e-01 -3.79713327e-01 7.43296027e-01
5.66408277e-01 7.59781539e-01 -1.15294790e+00 -7.33005404e-01
1.03987062e+00 7.11243987e-01 -1.09909093e+00 3.64805043e-01
-1.11252464e-01 4.96711761e-01 -4.02159899e-01 3.33440602e-01
9.51746464e-01 -5.02796352e-01 -5.24051964e-01 -3.72449875e-01
-3.05437118e-01 2.40571573e-01 -1.07833612e+00 2.16002560e+00
-7.83708870e-01 7.98623621e-01 -1.81907266e-01 -5.61187923e-01
9.82180357e-01 -3.00440639e-01 7.26685464e-01 -2.22600654e-01
-2.37562675e-02 1.35812372e-01 -2.78020054e-01 -3.61524135e-01
6.26211643e-01 1.83694083e-02 1.30805030e-01 -9.79957730e-02
3.68851632e-01 -4.94028211e-01 2.68581539e-01 3.01679999e-01
1.22939670e+00 1.67546839e-01 1.21390522e-01 -4.88659859e-01
6.54275775e-01 3.24047297e-01 4.33673322e-01 5.39810300e-01
-4.65254813e-01 6.75217867e-01 1.78974167e-01 -8.74285221e-01
-1.03195322e+00 -1.04408717e+00 -2.38809451e-01 7.83159792e-01
5.70616186e-01 -7.62582958e-01 -1.80095702e-01 -7.37269521e-01
2.85821348e-01 3.04357439e-01 -4.31296378e-01 2.39148568e-02
-5.40103853e-01 -4.13988829e-01 3.95199150e-01 5.83834231e-01
7.49096870e-01 -7.32642949e-01 -9.95623946e-01 3.05655897e-01
-3.09301727e-02 -1.36193216e+00 -2.82900572e-01 -1.58257321e-01
-9.19397414e-01 -1.13583505e+00 -5.69629014e-01 -6.03851259e-01
3.62095207e-01 5.52665651e-01 1.36306190e+00 -4.11629304e-02
-2.74675131e-01 4.34438467e-01 -3.78853381e-01 -9.67003629e-02
9.60099860e-05 4.18083429e-01 -2.30919272e-02 -2.62938619e-01
2.95030743e-01 -8.91490877e-01 -1.18990254e+00 2.31742531e-01
-6.02580786e-01 7.89243281e-02 3.70895296e-01 5.98271489e-01
4.77935106e-01 -4.34164219e-02 2.42686689e-01 -3.57803553e-01
1.41824529e-01 -4.34053034e-01 -7.54791796e-01 -1.86997548e-01
-2.13274777e-01 4.40318659e-02 9.86499727e-01 -2.03443825e-01
-6.20055079e-01 2.49469504e-01 -2.25715965e-01 -1.01703668e+00
-2.11347163e-01 7.34083867e-03 4.81115192e-01 -6.29894376e-01
8.68167877e-01 7.49328285e-02 -1.25322893e-01 -2.51630634e-01
5.50229013e-01 1.74803093e-01 4.86384600e-01 -2.82184839e-01
1.12242639e+00 9.41907048e-01 3.66908431e-01 -1.01865017e+00
-6.92318499e-01 -7.11167216e-01 -9.98393536e-01 -4.20584708e-01
8.61724734e-01 -1.32143545e+00 -1.07593381e+00 3.98149908e-01
-1.08075047e+00 -4.69452083e-01 -3.81648839e-01 6.47318900e-01
-7.26432204e-01 2.91569561e-01 -6.19749188e-01 -7.27484584e-01
-4.62842315e-01 -1.05004692e+00 1.60316479e+00 8.88575241e-02
5.37159741e-02 -1.10193276e+00 3.31637591e-01 3.78595889e-02
5.68105340e-01 4.62153137e-01 3.44332367e-01 6.82292730e-02
-1.10220182e+00 2.51867007e-02 -3.99460793e-01 7.28671998e-02
1.21644307e-02 2.83695757e-01 -8.98831189e-01 -5.42597473e-01
-2.96705723e-01 -4.50891733e-01 1.09268844e+00 3.17682922e-01
1.17175567e+00 -1.35758862e-01 -1.92142576e-01 1.31216884e+00
1.71614516e+00 -5.33535540e-01 3.49149287e-01 2.41782591e-01
1.04312849e+00 4.10187364e-01 5.39848387e-01 6.72227979e-01
7.57166922e-01 5.41616201e-01 8.88907194e-01 -1.74666673e-01
-2.04783633e-01 -5.45622468e-01 1.61159828e-01 8.10241878e-01
4.09217812e-02 -1.55057475e-01 -9.43851769e-01 7.22505033e-01
-1.76987958e+00 -9.76080954e-01 -2.97872752e-01 2.14197230e+00
2.71438807e-01 7.98803791e-02 2.79793650e-01 1.10071108e-01
1.36450663e-01 5.56953073e-01 -5.73236525e-01 1.42757341e-01
6.48724511e-02 9.87090096e-02 7.77682304e-01 5.24016500e-01
-1.23354554e+00 1.25431442e+00 6.13202620e+00 5.01463771e-01
-1.31157577e+00 -8.51468667e-02 2.26717308e-01 -3.88106406e-01
-1.26883149e-01 4.11422737e-02 -8.51957560e-01 4.77520883e-01
8.07014823e-01 -1.27515420e-01 2.80014426e-01 1.05993927e+00
1.81993842e-01 1.56585529e-01 -9.81235921e-01 1.26749551e+00
-2.22022697e-01 -1.81199884e+00 1.29633799e-01 4.65025939e-02
4.70371544e-01 6.09371006e-01 -2.34282047e-01 3.07412386e-01
4.25205797e-01 -7.97474802e-01 6.79403484e-01 2.14119866e-01
7.66509473e-01 -5.92040896e-01 4.61978287e-01 2.66988516e-01
-1.63004553e+00 -1.09132789e-01 -8.22580218e-01 -2.63910919e-01
4.56608325e-01 6.04149699e-01 -7.93717980e-01 5.13836682e-01
1.02835894e+00 1.39193618e+00 -7.22141683e-01 1.36678457e+00
1.13582611e-01 4.83340889e-01 -7.27894187e-01 2.23385505e-02
3.99705052e-01 -1.13102250e-01 6.10377192e-01 1.24918222e+00
3.63968313e-01 -1.25978336e-01 3.69143665e-01 1.01479554e+00
3.88362706e-02 6.84202164e-02 -7.31045187e-01 4.01685536e-01
5.01294494e-01 1.37149167e+00 -6.07102871e-01 -2.76419759e-01
-4.22478378e-01 7.10750878e-01 7.25893438e-01 1.36458769e-01
-8.37106586e-01 -2.27034226e-01 1.11307669e+00 3.23223412e-01
4.82470840e-01 -5.97316384e-01 -4.22587618e-02 -1.41656315e+00
-9.56186727e-02 -3.37593332e-02 1.26705274e-01 -7.34340668e-01
-1.45137835e+00 7.77015924e-01 6.56998763e-03 -1.73633909e+00
-9.14608389e-02 -6.16049290e-01 -6.55675530e-01 5.61995268e-01
-2.03400874e+00 -1.09575939e+00 -9.12328064e-01 7.13449717e-01
4.48669970e-01 6.73317239e-02 5.39566398e-01 3.08985919e-01
-3.59395385e-01 1.95186868e-01 -2.53405541e-01 9.31629911e-02
4.89947736e-01 -1.36829817e+00 7.17324317e-01 8.87714565e-01
1.25546992e-01 1.66112915e-01 5.88184416e-01 -3.40491235e-01
-1.41183925e+00 -1.48638952e+00 3.78095150e-01 -5.49721599e-01
7.31701493e-01 -4.71701801e-01 -9.07222927e-01 4.55730557e-01
-1.35447994e-01 6.18470967e-01 3.83170635e-01 -9.62171331e-02
-2.76710093e-01 -4.62298036e-01 -1.07292533e+00 4.45850581e-01
1.42098331e+00 -3.63023490e-01 -3.07894379e-01 4.29735988e-01
1.01792407e+00 -4.41304684e-01 -1.05932760e+00 6.18249774e-01
1.93382844e-01 -1.20785677e+00 1.21180713e+00 -2.27170348e-01
6.44699275e-01 -6.41641021e-01 8.09239969e-02 -1.33401120e+00
-6.67010128e-01 -5.64770877e-01 -3.04724872e-01 7.27373481e-01
1.72812324e-02 -5.08683980e-01 1.10198939e+00 2.15527177e-01
-3.33211899e-01 -5.82969248e-01 -9.66125190e-01 -7.75762320e-01
5.63950576e-02 -4.10991937e-01 7.10528016e-01 7.34873891e-01
-2.75541902e-01 3.07722539e-01 -3.30260009e-01 3.77704978e-01
8.05708885e-01 2.92412192e-01 1.35061824e+00 -1.35795009e+00
-4.06166948e-02 -4.24103081e-01 -1.04583132e+00 -1.58292019e+00
1.93825722e-01 -7.83780932e-01 -1.58565730e-01 -1.50599563e+00
-5.06251097e-01 -5.70167720e-01 1.80774331e-01 1.53712153e-01
-1.14945926e-01 2.17833459e-01 3.86010468e-01 4.35223579e-01
-7.69328713e-01 1.17864859e+00 1.33300841e+00 -2.12827735e-02
-4.23042506e-01 -2.07539961e-01 -1.24591179e-01 5.96964717e-01
6.07531905e-01 -3.14638108e-01 -6.02344036e-01 -6.69667423e-01
1.40069246e-01 3.92112844e-02 6.53825223e-01 -1.70873785e+00
3.41682732e-01 4.10706038e-03 4.06654745e-01 -7.87367761e-01
5.46686530e-01 -9.57245946e-01 -8.07937980e-02 3.56273502e-01
-8.05543959e-02 4.17326242e-02 3.43708932e-01 7.67944336e-01
-1.89575568e-01 3.35358560e-01 7.76378751e-01 -1.02356456e-01
-8.44940603e-01 9.40760434e-01 3.06701642e-02 1.59281522e-01
1.07599378e+00 -2.48631492e-01 -5.01425505e-01 -2.42533132e-01
-2.29313657e-01 3.81387144e-01 7.22904086e-01 4.07030702e-01
7.42592752e-01 -1.20317650e+00 -6.87879443e-01 3.81079078e-01
3.98130238e-01 7.01292396e-01 3.73012602e-01 7.62126267e-01
-1.29857051e+00 2.99448699e-01 -2.93237865e-01 -1.15609312e+00
-6.16563201e-01 5.09967744e-01 3.86908233e-01 -5.42941168e-02
-1.01970530e+00 9.24699306e-01 2.73549944e-01 -4.81282741e-01
-5.78898005e-02 -6.62146986e-01 -9.17257890e-02 -2.75040001e-01
3.94945681e-01 4.07293111e-01 -7.49739408e-02 -7.25530148e-01
-4.68975961e-01 8.44694674e-01 2.26352364e-01 2.42841169e-01
1.42487967e+00 -1.95227191e-02 2.10823163e-01 4.27354932e-01
1.41503334e+00 -2.65798122e-01 -1.93349135e+00 1.57980770e-02
-5.00989795e-01 -1.08269083e+00 3.19240868e-01 -5.09363189e-02
-1.34753263e+00 1.11315191e+00 5.13514757e-01 1.27064452e-01
9.14281607e-01 -1.67193636e-01 8.93644989e-01 3.77198786e-01
7.32285619e-01 -4.61629450e-01 4.73371372e-02 3.84546667e-01
6.95014715e-01 -1.46649241e+00 4.83346172e-02 -4.94508862e-01
-5.10428488e-01 1.14500558e+00 8.48663747e-01 -7.16211975e-01
8.20762455e-01 1.57458380e-01 -1.62954703e-01 -2.83380121e-01
-9.88127172e-01 -3.73787016e-01 1.13477603e-01 7.11899102e-01
-6.68125004e-02 -1.78698242e-01 4.63557333e-01 -3.43304068e-01
-6.39093459e-01 2.49764845e-01 4.40497607e-01 6.87593937e-01
-4.28318590e-01 -4.22492474e-01 9.17042121e-02 4.53862906e-01
1.35895744e-01 1.74865350e-02 -3.78007554e-02 8.85970056e-01
-1.64028361e-01 7.81988978e-01 4.34049964e-01 -5.78623414e-01
4.64248031e-01 -6.01700127e-01 3.80736202e-01 -3.64604503e-01
-5.44758499e-01 -4.15986449e-01 -3.87331963e-01 -1.05878246e+00
-7.46296167e-01 -4.56781715e-01 -9.52343047e-01 -8.42698693e-01
-4.83549312e-02 -1.36725992e-01 3.76267791e-01 5.57161748e-01
7.25078762e-01 2.74652094e-01 7.20241129e-01 -1.46543038e+00
-1.18168093e-01 -7.55172908e-01 -3.79315078e-01 5.48724174e-01
6.73616350e-01 -8.22947145e-01 -6.86116159e-01 -2.89602298e-02] | [8.538954734802246, -2.041053295135498] |
b8e140d5-d81a-4b03-a79e-92e8f0918007 | relvit-concept-guided-vision-transformer-for-1 | 2204.11167 | null | https://arxiv.org/abs/2204.11167v2 | https://arxiv.org/pdf/2204.11167v2.pdf | RelViT: Concept-guided Vision Transformer for Visual Relational Reasoning | Reasoning about visual relationships is central to how humans interpret the visual world. This task remains challenging for current deep learning algorithms since it requires addressing three key technical problems jointly: 1) identifying object entities and their properties, 2) inferring semantic relations between pairs of entities, and 3) generalizing to novel object-relation combinations, i.e., systematic generalization. In this work, we use vision transformers (ViTs) as our base model for visual reasoning and make better use of concepts defined as object entities and their relations to improve the reasoning ability of ViTs. Specifically, we introduce a novel concept-feature dictionary to allow flexible image feature retrieval at training time with concept keys. This dictionary enables two new concept-guided auxiliary tasks: 1) a global task for promoting relational reasoning, and 2) a local task for facilitating semantic object-centric correspondence learning. To examine the systematic generalization of visual reasoning models, we introduce systematic splits for the standard HICO and GQA benchmarks. We show the resulting model, Concept-guided Vision Transformer (or RelViT for short) significantly outperforms prior approaches on HICO and GQA by 16% and 13% in the original split, and by 43% and 18% in the systematic split. Our ablation analyses also reveal our model's compatibility with multiple ViT variants and robustness to hyper-parameters. | ['Anima Anandkumar', 'Song-Chun Zhu', 'Yuke Zhu', 'Chaowei Xiao', 'Huaizu Jiang', 'Zhiding Yu', 'Weili Nie', 'Xiaojian Ma'] | 2022-04-24 | relvit-concept-guided-vision-transformer-for | https://openreview.net/forum?id=afoV8W3-IYp | https://openreview.net/pdf?id=afoV8W3-IYp | iclr-2022-4 | ['systematic-generalization'] | ['reasoning'] | [-1.56862102e-02 1.48886561e-01 -1.10801160e-01 -1.99196339e-01
-2.74325788e-01 -7.16501951e-01 9.13568258e-01 1.80771768e-01
-3.40638161e-01 2.14816809e-01 3.25783253e-01 -4.22087997e-01
-2.43891016e-01 -8.57941568e-01 -8.31336796e-01 -3.52684498e-01
9.20664072e-02 6.28234088e-01 3.57321918e-01 -2.84555942e-01
1.17506266e-01 4.94137824e-01 -1.64129615e+00 5.60305774e-01
7.28033662e-01 1.14581311e+00 2.03950405e-01 3.96930367e-01
-3.07036072e-01 1.20135200e+00 -4.11072105e-01 -6.36196434e-01
3.63725722e-01 -2.92233020e-01 -1.21859121e+00 1.95863187e-01
7.16103017e-01 -2.73324192e-01 -4.40835208e-01 9.57254231e-01
9.21234041e-02 3.32492948e-01 6.13330603e-01 -1.58398938e+00
-1.22733068e+00 6.45680130e-01 -5.12629032e-01 3.49996865e-01
3.37574631e-01 3.72652411e-01 1.42870533e+00 -1.06328630e+00
7.85511076e-01 1.53345609e+00 4.27604139e-01 4.52886254e-01
-1.30046654e+00 -5.03124058e-01 4.15283084e-01 6.10741317e-01
-1.39088202e+00 -4.33070779e-01 5.56060791e-01 -6.26623988e-01
1.24699807e+00 8.51532668e-02 7.18201160e-01 7.42404521e-01
-3.43604386e-01 7.79609799e-01 1.09345067e+00 -2.62356520e-01
8.70066434e-02 6.63924143e-02 2.78607219e-01 8.74553740e-01
2.81909674e-01 -7.04950606e-03 -6.45875692e-01 2.41043210e-01
7.53327429e-01 2.53468044e-02 -2.74681002e-01 -7.03092337e-01
-1.47728610e+00 8.15545678e-01 9.60812032e-01 1.18848145e-01
-3.27487856e-01 3.59316170e-01 3.00257564e-01 3.52155238e-01
-5.30504547e-02 7.43105292e-01 -3.58504891e-01 2.34772831e-01
-3.57069463e-01 3.38785201e-01 5.88470101e-01 1.30624819e+00
9.03508723e-01 -1.90292031e-01 -3.63877237e-01 8.53392005e-01
2.87480325e-01 4.41484779e-01 1.46328762e-01 -1.15818036e+00
6.27773643e-01 9.35686767e-01 -1.63238585e-01 -1.01478767e+00
-4.55255240e-01 -4.00157958e-01 -6.27017200e-01 3.30099165e-02
4.50964034e-01 3.69523108e-01 -1.07297778e+00 1.91965616e+00
2.77855039e-01 -9.32923779e-02 2.63076276e-01 8.52405608e-01
1.36747265e+00 3.76536638e-01 3.27268749e-01 9.60340500e-02
1.79212654e+00 -1.19604313e+00 -4.04729694e-01 -3.28689277e-01
5.91013014e-01 -5.35908699e-01 1.41008115e+00 3.96476910e-02
-1.14524961e+00 -6.04624152e-01 -7.73445070e-01 -6.84770882e-01
-7.03650594e-01 -4.10376824e-02 1.02763772e+00 1.25637323e-01
-1.14622843e+00 8.37115292e-03 -5.13650835e-01 -6.62461758e-01
8.72283459e-01 2.59890735e-01 -4.05249685e-01 -2.99913079e-01
-1.14567065e+00 9.80293095e-01 4.71410424e-01 -1.71863571e-01
-8.18128109e-01 -8.40964913e-01 -1.00823092e+00 4.36557770e-01
6.58068597e-01 -1.28724623e+00 1.15454090e+00 -6.88832641e-01
-8.31635058e-01 1.25329387e+00 -2.09533483e-01 -5.11596203e-01
2.37040505e-01 -1.98216513e-01 -1.55843034e-01 4.23795223e-01
2.80950218e-01 1.04439533e+00 6.79716349e-01 -1.51276338e+00
-6.81526840e-01 -2.68652588e-01 7.26590157e-01 4.87301201e-01
-1.41297013e-01 -2.75086075e-01 -9.27190304e-01 -5.88892221e-01
3.02465409e-01 -8.72035503e-01 2.72626132e-01 4.09412235e-01
-4.26847935e-01 -5.22039890e-01 5.27055979e-01 -4.15369421e-01
7.74409354e-01 -2.03943849e+00 2.44020998e-01 2.21743450e-01
7.73474872e-01 1.37827516e-01 -2.29893968e-01 2.68349856e-01
-1.27919912e-01 -2.69984361e-02 -5.17263152e-02 -2.50469476e-01
6.20137863e-02 3.36137265e-01 -5.17599881e-01 -6.62778094e-02
2.51322716e-01 1.36305511e+00 -1.06645918e+00 -5.26811659e-01
2.12869063e-01 2.43025690e-01 -6.94432199e-01 7.80242309e-03
-2.46415585e-01 1.27359882e-01 -2.33350202e-01 7.96104193e-01
4.64489579e-01 -7.50453353e-01 1.75669461e-01 -7.34084010e-01
2.21085921e-01 3.12111855e-01 -9.12162483e-01 1.68485701e+00
-5.83880007e-01 7.46627629e-01 -3.58731419e-01 -9.64405477e-01
8.03428411e-01 -7.39737377e-02 2.14483514e-01 -1.10805666e+00
-1.04441188e-01 -1.92326590e-01 6.59902692e-02 -5.94633281e-01
5.08432567e-01 -9.89645943e-02 2.16781646e-01 4.16262031e-01
2.28641450e-01 -4.78259102e-02 2.37585649e-01 7.91805923e-01
9.37034130e-01 7.95353204e-02 3.65793496e-01 -2.05003515e-01
4.14322317e-01 3.81082185e-02 3.11848670e-01 8.22484851e-01
-1.50713235e-01 4.90314096e-01 7.09690094e-01 -4.21034247e-01
-7.64946878e-01 -1.36276758e+00 3.74097489e-02 1.17267323e+00
6.16814315e-01 -6.05731249e-01 -2.49291673e-01 -8.55652213e-01
2.64275640e-01 7.27804363e-01 -8.68181765e-01 -2.57268816e-01
-4.22047228e-01 -3.59439701e-01 4.00035620e-01 9.16665375e-01
7.83519924e-01 -1.16017604e+00 -6.48238838e-01 -2.91760862e-01
-3.22460055e-01 -1.51137125e+00 -2.82321692e-01 1.24949396e-01
-4.59788829e-01 -1.43467796e+00 -4.48543012e-01 -7.56092131e-01
6.78567827e-01 6.97019935e-01 1.47941184e+00 4.00926083e-01
-1.19413987e-01 7.94970751e-01 -3.22460562e-01 -1.86570823e-01
-1.71425685e-01 -1.57777846e-01 -1.68744832e-01 -1.12071313e-01
2.46630684e-01 -4.75876361e-01 -6.35859728e-01 2.50178784e-01
-7.97799170e-01 4.71862048e-01 6.04087889e-01 8.74322295e-01
4.53492165e-01 -2.15596557e-01 3.01982909e-01 -9.24492717e-01
4.28940475e-01 -3.62165153e-01 -3.96404028e-01 6.81386888e-01
-5.43479145e-01 2.83187687e-01 4.56809789e-01 -3.20081592e-01
-1.06734538e+00 -1.91844314e-01 2.71038145e-01 -6.38734818e-01
1.31717801e-01 4.18332189e-01 -2.48656079e-01 -4.38705124e-02
8.40913892e-01 2.41723210e-01 -2.87099153e-01 -1.86759189e-01
8.82609665e-01 1.05585486e-01 8.22576284e-01 -8.16379726e-01
1.03164959e+00 6.24919951e-01 1.47921965e-01 -5.96902549e-01
-1.10097754e+00 -5.73498487e-01 -5.80297709e-01 2.97795143e-02
1.02527618e+00 -1.22166908e+00 -1.03121853e+00 2.14162588e-01
-1.22214162e+00 -4.21027750e-01 -3.55731070e-01 1.76919237e-01
-5.81274033e-01 3.16745549e-01 -4.72981572e-01 -2.14233026e-01
-2.49635473e-01 -1.11382496e+00 9.84701216e-01 1.81655660e-01
-6.93661049e-02 -9.86761570e-01 -2.90573478e-01 7.51900315e-01
2.44351879e-01 -5.29572181e-02 1.28708279e+00 -4.30860251e-01
-9.87972677e-01 4.82762188e-01 -9.78196919e-01 2.46664569e-01
4.29987833e-02 -3.23630005e-01 -1.00846171e+00 -1.90036252e-01
-4.26726311e-01 -6.01621151e-01 1.25980341e+00 -7.71738067e-02
1.24347055e+00 -3.90158556e-02 -3.72292876e-01 9.38405037e-01
1.31745470e+00 7.43730515e-02 6.04001522e-01 3.68465543e-01
1.09494507e+00 5.39537966e-01 4.83727932e-01 1.21532522e-01
1.04027534e+00 7.36465156e-01 4.90628123e-01 -1.59244791e-01
-6.08958364e-01 -3.41734260e-01 -5.35864122e-02 4.07969654e-01
-2.36652225e-01 -1.29806548e-01 -1.19276500e+00 6.69994831e-01
-2.06346369e+00 -9.14216816e-01 9.84241813e-03 1.84318376e+00
7.15552747e-01 -2.57028546e-02 3.69961746e-02 -2.47745499e-01
4.04076964e-01 1.55886076e-02 -6.05352759e-01 -4.49107438e-02
-2.28054881e-01 2.14154795e-01 2.30720863e-01 3.18557501e-01
-1.01149559e+00 1.30638874e+00 5.96463633e+00 4.38201338e-01
-7.60017216e-01 2.12257672e-02 2.70607024e-01 1.96023323e-02
-5.05670846e-01 2.47146264e-01 -6.07292175e-01 -5.83316498e-02
7.36262649e-02 -1.00564487e-01 6.66404307e-01 6.26408339e-01
-3.88724476e-01 -1.02183789e-01 -1.40398777e+00 1.39972937e+00
2.82849520e-01 -1.57850933e+00 6.10922635e-01 -1.93465292e-01
6.31648302e-01 -1.10869743e-01 8.06062445e-02 5.29157400e-01
4.24856663e-01 -1.08834100e+00 8.96625400e-01 5.74827552e-01
7.08792627e-01 -4.64945883e-01 3.92936498e-01 -2.94198185e-01
-1.42054141e+00 -1.74489513e-01 -3.17454427e-01 8.11486691e-02
-1.54404789e-01 2.01789722e-01 -6.40298545e-01 4.89967257e-01
9.19509888e-01 1.00421476e+00 -9.31137204e-01 8.21942508e-01
-4.93181854e-01 2.39845112e-01 -1.31591514e-01 3.28664303e-01
2.53517896e-01 2.14294776e-01 4.04536754e-01 9.64834750e-01
-5.20716906e-02 9.77756605e-02 5.81570230e-02 1.35207105e+00
-3.37325990e-01 -2.85524577e-01 -3.70312154e-01 2.96519790e-02
6.65175557e-01 1.19520497e+00 -8.42554569e-01 -5.16388297e-01
-6.34332657e-01 9.30758595e-01 7.70460427e-01 7.32909083e-01
-8.44829023e-01 -2.21863106e-01 7.98822820e-01 9.60923079e-03
6.48228168e-01 -1.85433805e-01 -2.15609446e-01 -1.44601929e+00
5.64353466e-02 -9.27112937e-01 6.62841499e-01 -1.16929710e+00
-1.34862924e+00 4.96553302e-01 2.12684110e-01 -8.69132876e-01
2.34864540e-02 -7.97525525e-01 -3.85776132e-01 8.43249798e-01
-1.67436409e+00 -1.71815181e+00 -8.25958133e-01 1.11305869e+00
3.72021198e-01 -1.44392267e-01 5.71303189e-01 1.79293856e-01
-2.70000637e-01 6.29203200e-01 -5.36702275e-01 1.55399308e-01
6.45130277e-01 -1.40373075e+00 5.31661451e-01 7.53916800e-01
4.61135715e-01 9.77742672e-01 4.69624043e-01 -3.41623217e-01
-1.33979559e+00 -1.02696478e+00 8.19251239e-01 -7.46026516e-01
6.75752878e-01 -4.67017859e-01 -1.01294768e+00 1.01162374e+00
3.97941843e-02 2.56957233e-01 3.62539411e-01 5.15395284e-01
-1.03577626e+00 -1.83895752e-01 -8.06778133e-01 8.71927738e-01
1.56554031e+00 -9.33771789e-01 -8.58597100e-01 4.09270138e-01
9.15268600e-01 -2.96124905e-01 -7.29439735e-01 3.24339569e-01
5.17525852e-01 -1.02413619e+00 1.33609998e+00 -6.91644490e-01
4.25348908e-01 -5.68781972e-01 -3.42430651e-01 -1.02430868e+00
-5.95172405e-01 -2.11333245e-01 -3.50420296e-01 1.11798608e+00
2.48651117e-01 -5.78966856e-01 3.82120162e-01 4.96951520e-01
-9.66148525e-02 -7.69564867e-01 -5.75552404e-01 -7.94258773e-01
-7.72152692e-02 -4.46168691e-01 5.86839676e-01 1.21415687e+00
-1.73733890e-01 6.51141524e-01 -3.13691999e-04 2.13934869e-01
4.60799217e-01 4.21581596e-01 8.69656026e-01 -1.02085364e+00
-4.60724056e-01 -6.43730164e-01 -4.90491629e-01 -1.29744565e+00
1.21165864e-01 -1.14344215e+00 -3.12834501e-01 -1.97155178e+00
5.35642684e-01 -5.55326581e-01 -2.93743908e-01 9.21461701e-01
-2.94615299e-01 1.77656338e-01 6.73424542e-01 3.85935903e-01
-9.54238057e-01 5.31768262e-01 1.45693278e+00 -3.42084318e-01
-2.69309767e-02 -4.26085681e-01 -1.06744313e+00 5.29853761e-01
4.04918134e-01 -1.04616478e-01 -8.88624787e-01 -6.07594013e-01
5.52299559e-01 -1.79308504e-01 1.08715510e+00 -8.30147624e-01
2.06319302e-01 -2.03630015e-01 2.42530569e-01 -4.38288212e-01
3.20310920e-01 -7.15404570e-01 -2.09208637e-01 1.89536184e-01
-3.98875684e-01 1.92758009e-01 2.46522292e-01 5.44767737e-01
-2.04946414e-01 2.64085233e-01 5.87157547e-01 -1.70108378e-01
-1.33238935e+00 2.79574037e-01 1.05971210e-01 3.94645721e-01
9.04622614e-01 -2.11713329e-01 -9.40322518e-01 -3.38868856e-01
-7.53401875e-01 4.76030201e-01 4.24481243e-01 6.51747584e-01
7.36694455e-01 -1.42625153e+00 -3.88606578e-01 -2.48921551e-02
7.20055699e-01 1.09605588e-01 2.62823552e-01 8.73782694e-01
-4.37949836e-01 3.97940665e-01 -2.85501510e-01 -6.04601502e-01
-1.25531888e+00 8.51750672e-01 4.22765702e-01 -1.50101632e-01
-8.42415333e-01 1.05217075e+00 9.16626096e-01 -2.63606757e-01
2.10268423e-01 -5.52025259e-01 -1.84257343e-01 -6.60653338e-02
3.44322234e-01 -2.73738168e-02 -1.30098403e-01 -6.55121505e-01
-5.00738204e-01 5.28113306e-01 -1.41681492e-01 1.57580182e-01
1.01872897e+00 -2.57143587e-01 -1.56755656e-01 3.04909557e-01
1.01995265e+00 -3.25878620e-01 -1.09624851e+00 -5.96058190e-01
-1.06504351e-01 -3.73469025e-01 -1.21932052e-01 -1.07754529e+00
-1.19845748e+00 8.99360657e-01 3.33596528e-01 -1.30574340e-02
1.22577941e+00 5.69590509e-01 4.94340271e-01 6.85407460e-01
2.86446095e-01 -5.24112165e-01 3.21844220e-01 5.13248444e-01
1.05306542e+00 -1.28765023e+00 9.67746377e-02 -5.92371106e-01
-8.03832471e-01 1.00456953e+00 8.97250235e-01 1.90515593e-01
4.27582026e-01 -2.59038389e-01 6.75151646e-02 -6.87311471e-01
-9.13395643e-01 -7.05616832e-01 7.17046440e-01 8.16754758e-01
1.97639316e-01 -6.40426725e-02 1.75108835e-01 3.47094268e-01
-1.49661273e-01 -2.88071394e-01 6.43660054e-02 6.44898415e-01
-1.42322227e-01 -7.54391015e-01 -6.69293702e-02 3.15298617e-01
2.26332724e-01 -4.19857234e-01 -3.59686673e-01 1.07373941e+00
3.97749990e-01 7.43873596e-01 2.24311993e-01 -2.90208936e-01
3.74476075e-01 -9.74110290e-02 6.44144118e-01 -6.34580433e-01
-4.40034568e-01 -4.15235966e-01 1.75681055e-01 -7.82537341e-01
-5.86186111e-01 -5.77271700e-01 -1.46845424e+00 -1.88948020e-01
-5.50286360e-02 -3.44656616e-01 2.23930359e-01 1.18809474e+00
4.27402407e-01 7.03311443e-01 6.01018444e-02 -3.07760149e-01
-2.47403696e-01 -5.82051516e-01 -3.11623335e-01 9.67854500e-01
2.93348312e-01 -1.03179049e+00 -6.71224296e-02 1.38519675e-01] | [10.564416885375977, 1.6552205085754395] |
30387e97-6144-44a7-be64-06c62d8c23d9 | contextual-explanation-networks | 1705.10301 | null | https://arxiv.org/abs/1705.10301v4 | https://arxiv.org/pdf/1705.10301v4.pdf | Contextual Explanation Networks | Modern learning algorithms excel at producing accurate but complex models of the data. However, deploying such models in the real-world requires extra care: we must ensure their reliability, robustness, and absence of undesired biases. This motivates the development of models that are equally accurate but can be also easily inspected and assessed beyond their predictive performance. To this end, we introduce contextual explanation networks (CEN)---a class of architectures that learn to predict by generating and utilizing intermediate, simplified probabilistic models. Specifically, CENs generate parameters for intermediate graphical models which are further used for prediction and play the role of explanations. Contrary to the existing post-hoc model-explanation tools, CENs learn to predict and to explain simultaneously. Our approach offers two major advantages: (i) for each prediction valid, instance-specific explanation is generated with no computational overhead and (ii) prediction via explanation acts as a regularizer and boosts performance in data-scarce settings. We analyze the proposed framework theoretically and experimentally. Our results on image and text classification and survival analysis tasks demonstrate that CENs are not only competitive with the state-of-the-art methods but also offer additional insights behind each prediction, that can be valuable for decision support. We also show that while post-hoc methods may produce misleading explanations in certain cases, CENs are consistent and allow to detect such cases systematically. | ['Maruan Al-Shedivat', 'Avinava Dubey', 'Eric P. Xing'] | 2017-05-29 | contextual-explanation-networks-1 | https://openreview.net/forum?id=HJUOHGWRb | https://openreview.net/pdf?id=HJUOHGWRb | iclr-2018-1 | ['interpretability-techniques-for-deep-learning'] | ['miscellaneous'] | [ 5.20040512e-01 7.46035695e-01 -4.88813519e-01 -6.41951680e-01
-6.68028831e-01 -2.39061996e-01 9.00337219e-01 4.18210894e-01
7.40863457e-02 8.11968267e-01 1.19137645e-01 -6.75729334e-01
-6.55915201e-01 -5.28499722e-01 -7.03729808e-01 -6.98720694e-01
-2.36102179e-01 7.71001816e-01 7.88219646e-02 2.11210355e-01
2.93386161e-01 3.57392371e-01 -1.81190813e+00 4.50653523e-01
1.24854708e+00 9.16565657e-01 8.90287664e-03 5.81621587e-01
-1.12640463e-01 8.85700107e-01 -2.15157866e-01 -7.59868622e-01
-2.73421735e-01 -4.81704205e-01 -5.42140841e-01 2.10883901e-01
2.19028905e-01 -1.59588888e-01 -7.24840490e-03 7.48306453e-01
1.44744754e-01 -2.36126155e-01 1.10444260e+00 -1.67495167e+00
-7.74985075e-01 9.33050871e-01 -2.83384681e-01 -1.26425803e-01
1.01406448e-01 6.35925978e-02 1.14188766e+00 -9.15347815e-01
2.82125801e-01 1.05616605e+00 8.48297238e-01 6.90612435e-01
-1.30797815e+00 -5.35151005e-01 4.32211071e-01 2.67878532e-01
-9.50364530e-01 -3.40922415e-01 6.44341469e-01 -3.32276762e-01
5.45555890e-01 5.67283034e-01 4.22260195e-01 1.38478875e+00
4.63355809e-01 9.69491065e-01 9.84899700e-01 -5.23191333e-01
3.04822952e-01 4.58857596e-01 4.77272719e-01 8.34264100e-01
4.42329168e-01 3.36860508e-01 -6.95540965e-01 -2.82161534e-01
4.24099892e-01 2.83359826e-01 -3.95122975e-01 -5.24463594e-01
-1.02975953e+00 9.96721625e-01 4.00169224e-01 3.20943482e-02
-6.02130532e-01 1.85753405e-01 7.45813996e-02 -6.79879338e-02
4.40241426e-01 2.95847952e-01 -4.74396646e-01 1.38051346e-01
-1.09172690e+00 1.59498453e-01 7.78404653e-01 7.95886695e-01
5.87794542e-01 -1.35389958e-02 -2.72608995e-01 5.02900362e-01
4.33221668e-01 2.60309160e-01 4.86999691e-01 -5.91071546e-01
9.38618183e-02 7.37165987e-01 8.50511566e-02 -9.38026428e-01
-5.48509240e-01 -9.62379694e-01 -1.11175323e+00 8.15897807e-02
3.78424674e-01 2.56833971e-01 -9.01303351e-01 1.67293334e+00
1.70342341e-01 2.11617351e-01 -9.70867742e-03 7.38277376e-01
5.31934381e-01 2.71229535e-01 2.27195933e-01 -2.66456902e-01
1.39165699e+00 -8.98247004e-01 -7.92956769e-01 -5.48436403e-01
5.18021584e-01 -4.31478173e-01 1.18762922e+00 5.97935021e-01
-8.72525215e-01 -3.71052951e-01 -9.25715864e-01 2.68837720e-01
-2.47484162e-01 2.98671335e-01 9.53901470e-01 7.14797258e-01
-9.45236266e-01 6.23110890e-01 -8.86484802e-01 -2.12693781e-01
4.73775417e-01 4.37779665e-01 -2.94055849e-01 5.03719412e-02
-8.89464915e-01 8.37656081e-01 2.11988837e-01 2.47630388e-01
-7.95644760e-01 -6.09623611e-01 -7.58795798e-01 3.99162382e-01
3.42839479e-01 -9.14433956e-01 1.07636094e+00 -1.25518382e+00
-1.10280955e+00 4.10259485e-01 -3.86124611e-01 -7.42040038e-01
6.79497838e-01 -4.25205916e-01 -1.98148936e-01 3.81313660e-03
7.48285465e-03 5.03723025e-01 8.78664196e-01 -1.56279993e+00
-4.91555184e-01 -3.35698068e-01 -1.47911355e-01 -2.73908794e-01
-2.88547248e-01 -4.30988818e-01 -4.68215376e-01 -5.01473606e-01
3.77196670e-01 -8.23590517e-01 -4.22844291e-01 3.29685062e-02
-8.01416099e-01 -1.27915934e-01 6.05510473e-01 -4.81361091e-01
1.13279235e+00 -1.87148809e+00 -7.50733539e-02 4.20740753e-01
4.92981613e-01 -1.01147378e-02 1.19991206e-01 3.07653934e-01
-2.16853991e-01 2.70917296e-01 -4.28506166e-01 -7.09928989e-01
9.08354744e-02 4.14796323e-01 -4.00991470e-01 3.54623437e-01
3.47647071e-01 8.88315320e-01 -6.76211476e-01 -5.91409802e-01
3.39385480e-01 5.98440409e-01 -6.37788892e-01 1.91301152e-01
-2.91490436e-01 2.85523057e-01 -5.55930376e-01 4.71016079e-01
3.39075267e-01 -6.99403286e-01 3.55789870e-01 1.00387320e-01
1.99433237e-01 1.50935367e-01 -1.02180350e+00 9.83126342e-01
-3.75474244e-01 5.91083944e-01 -4.60064083e-01 -1.21359837e+00
8.88608873e-01 1.60334334e-01 -1.15222791e-02 -2.17029750e-01
8.95160437e-02 2.52667934e-01 -1.61157638e-01 -3.93922567e-01
3.07773590e-01 -1.85614675e-01 2.34470323e-01 4.41591918e-01
-5.28739728e-02 2.77424216e-01 -9.33956802e-02 2.21221760e-01
9.40072775e-01 1.45664196e-02 6.05125189e-01 -1.03270806e-01
5.06798744e-01 -1.70989782e-01 5.59828103e-01 1.17451799e+00
1.10176444e-01 6.20180070e-01 9.23335850e-01 -5.91966748e-01
-7.53069520e-01 -1.09107959e+00 -2.15161279e-01 8.62887084e-01
-1.11623317e-01 -2.07952678e-01 -5.75363994e-01 -1.17822266e+00
-2.20130160e-02 1.17004085e+00 -9.55813885e-01 -1.14543110e-01
1.24794664e-02 -8.96129608e-01 1.43364727e-01 7.18268394e-01
9.72404778e-02 -7.91778862e-01 -4.40795898e-01 5.66486828e-02
-2.16424391e-01 -9.21928585e-01 -2.51767156e-03 4.88431185e-01
-1.10835350e+00 -1.27803087e+00 -3.95723701e-01 -2.01981232e-01
9.85943258e-01 5.87477125e-02 1.26514232e+00 5.30170023e-01
4.89019006e-02 3.69838923e-01 -3.36056858e-01 -6.07139587e-01
-5.08643031e-01 1.33494601e-01 -9.43421498e-02 2.22073033e-01
2.09212705e-01 -6.61187053e-01 -6.41878128e-01 4.30098712e-01
-1.02543783e+00 3.41674149e-01 1.00336671e+00 1.13611186e+00
6.87732577e-01 -4.08198871e-02 7.93415904e-01 -1.39533901e+00
4.04181540e-01 -6.30941570e-01 -5.25918603e-01 3.86823326e-01
-1.30685079e+00 4.27366406e-01 7.69958854e-01 -2.23117262e-01
-1.06043792e+00 1.58859879e-01 8.99219587e-02 -1.74024150e-01
-2.11944312e-01 6.09564006e-01 -9.17216390e-03 3.82174253e-01
8.35323572e-01 2.46133089e-01 4.74266969e-02 -3.73734891e-01
2.78548747e-01 5.11617303e-01 4.61911798e-01 -3.09041321e-01
8.44597578e-01 4.88540530e-01 1.87918335e-01 -4.23558950e-01
-1.02007103e+00 -1.44380301e-01 -5.46069682e-01 -2.52004564e-01
5.18607378e-01 -5.46414495e-01 -6.01984143e-01 -1.32633939e-01
-9.80808556e-01 -1.22614123e-01 2.23540422e-02 4.35667723e-01
-5.48982382e-01 3.72030377e-01 -2.22943693e-01 -1.20587540e+00
-2.43109405e-01 -1.00319028e+00 9.38918471e-01 2.58441180e-01
-3.88304114e-01 -1.21459079e+00 -3.17774326e-01 2.97805220e-01
2.48414278e-01 2.36055523e-01 1.19100344e+00 -1.08096099e+00
-7.97342598e-01 -4.56706434e-01 -2.05692440e-01 1.09255880e-01
-1.25118539e-01 5.20966537e-02 -1.17577469e+00 1.14509329e-01
-4.51309651e-01 -6.21709302e-02 9.44078028e-01 5.13998270e-01
1.57499361e+00 -4.75386739e-01 -6.01291716e-01 2.64140576e-01
1.07349050e+00 -2.61914164e-01 5.46230853e-01 1.77547336e-01
2.29367256e-01 8.23687792e-01 5.14029682e-01 4.99619544e-01
4.95867044e-01 6.02582574e-01 8.39110851e-01 -2.90461779e-01
1.01049803e-01 -4.75446671e-01 1.62135735e-01 3.12041223e-01
-4.55202833e-02 -4.23283070e-01 -1.04162598e+00 5.57587683e-01
-2.30237389e+00 -8.88834476e-01 -6.49271607e-01 2.31293726e+00
5.51657021e-01 2.66483575e-01 -8.23953152e-02 3.91835481e-01
4.88052577e-01 -1.66045934e-01 -5.08474350e-01 -2.66153663e-01
-1.33913711e-01 2.93259770e-02 2.26079270e-01 4.64298874e-01
-8.59633565e-01 5.31223118e-01 6.15438557e+00 5.72454751e-01
-9.13919210e-01 -1.53646329e-02 9.15090859e-01 1.17048837e-01
-6.87663674e-01 9.35499966e-02 -5.80313027e-01 2.09266245e-01
9.74841058e-01 1.38116077e-01 1.44198969e-01 1.11672902e+00
2.79131919e-01 -1.29776567e-01 -1.36781418e+00 5.75175345e-01
-2.03953795e-02 -1.46102381e+00 3.88501525e-01 1.35753781e-01
5.25442779e-01 -3.85472596e-01 3.13580930e-01 2.19879761e-01
2.85437077e-01 -1.24065721e+00 8.71713340e-01 7.78834105e-01
3.54875535e-01 -7.50925183e-01 1.12689412e+00 5.45826197e-01
-7.05762625e-01 -2.84717381e-01 -3.30627054e-01 5.36265858e-02
-5.79158515e-02 1.00242257e+00 -1.03953469e+00 8.30099344e-01
3.82257462e-01 5.64521551e-01 -8.18121016e-01 9.73646104e-01
-6.83557332e-01 9.14138615e-01 -7.52426982e-02 -1.03767879e-01
1.71698213e-01 2.99437076e-01 2.10987136e-01 1.14449835e+00
3.64555478e-01 -6.21903837e-02 -1.36240363e-01 9.39543128e-01
2.41426945e-01 1.02720879e-01 -5.17008305e-01 2.17506096e-01
4.27772194e-01 1.17496836e+00 -7.96804547e-01 -3.40644151e-01
-2.85949707e-01 8.06252539e-01 4.16542470e-01 3.29499871e-01
-9.52751935e-01 1.21571004e-01 2.29070291e-01 1.37708068e-01
1.22402303e-01 1.55910015e-01 -7.31070340e-01 -1.09096766e+00
-8.12493712e-02 -8.58513176e-01 6.24297082e-01 -8.42508793e-01
-1.42325914e+00 8.27752352e-01 2.26794071e-02 -1.04641390e+00
-6.20055497e-01 -7.22232342e-01 -6.83123589e-01 6.12536848e-01
-1.73645318e+00 -1.31931794e+00 -4.26212311e-01 4.53256607e-01
5.28104901e-01 -6.33103261e-03 9.06279266e-01 -1.33072868e-01
-5.79855084e-01 6.47353709e-01 3.88945416e-02 -2.34105557e-01
4.66913611e-01 -1.40514112e+00 1.62126627e-02 7.32815802e-01
4.44125295e-01 7.42990077e-01 1.07609558e+00 -5.65947294e-01
-9.53717828e-01 -1.07804203e+00 1.27011609e+00 -6.02151453e-01
4.05927420e-01 -2.28879943e-01 -9.91033614e-01 7.62307525e-01
-7.91977644e-02 -7.95670822e-02 9.04743016e-01 6.84978247e-01
-3.62578630e-01 1.66551098e-02 -9.54837084e-01 6.75470769e-01
8.76179516e-01 -2.56965607e-01 -5.16609967e-01 3.24010134e-01
1.79498941e-01 -8.57891291e-02 -2.56188005e-01 2.93350995e-01
8.17456782e-01 -1.48967791e+00 8.69379401e-01 -7.69142032e-01
7.81311572e-01 -8.73481706e-02 4.44409959e-02 -1.12904513e+00
-7.33887255e-02 -5.43700278e-01 -1.55425772e-01 1.16245401e+00
9.11033928e-01 -6.47275865e-01 8.64101052e-01 9.12139475e-01
-1.40395418e-01 -9.63814914e-01 -9.45208490e-01 -6.69938564e-01
-1.87149331e-01 -9.70959127e-01 6.91994488e-01 7.78817832e-01
1.63412653e-02 1.13872737e-01 -5.55996954e-01 5.38819909e-01
8.46765637e-01 1.74148336e-01 7.87005901e-01 -1.51343477e+00
-3.21293563e-01 -2.96052903e-01 -3.66081774e-01 -8.65889311e-01
2.62637496e-01 -7.57765293e-01 9.12917720e-04 -1.53696012e+00
4.54535067e-01 -4.60726082e-01 -3.23695004e-01 7.02651918e-01
-5.35152435e-01 -1.87418200e-02 -5.78734018e-02 3.11520010e-01
-4.94025558e-01 5.63686073e-01 6.01164341e-01 3.11913788e-01
1.55998264e-02 2.98198730e-01 -9.29279923e-01 9.64511812e-01
6.70419037e-01 -6.77875459e-01 -4.83685136e-01 -7.96311945e-02
1.87617719e-01 1.63047314e-01 8.42739105e-01 -7.60542214e-01
1.36929438e-01 -1.04073305e-02 5.52952945e-01 -5.91605604e-01
1.27837941e-01 -9.04481173e-01 1.40349627e-01 5.06659448e-01
-5.60274243e-01 -3.73683535e-02 -4.51410674e-02 9.95302081e-01
-4.87358458e-02 -2.32149333e-01 4.85339016e-01 1.85198665e-01
-2.47588143e-01 1.19054846e-01 -4.51139003e-01 -3.71240675e-01
8.00909042e-01 -1.88156635e-01 -2.94488460e-01 -8.24834228e-01
-7.86493123e-01 2.53274262e-01 1.39441177e-01 2.02239826e-01
6.21709287e-01 -1.08011961e+00 -5.46852350e-01 1.73402891e-01
2.74848640e-01 -3.70743871e-01 2.25721896e-01 9.25063848e-01
2.51728781e-02 6.12177312e-01 2.10361362e-01 -8.28074634e-01
-1.14591932e+00 5.76122582e-01 2.10107744e-01 -6.87014997e-01
-4.98339951e-01 5.39285123e-01 5.17668545e-01 -3.65789026e-01
2.84732312e-01 -2.86520809e-01 -2.58184463e-01 -2.17909813e-02
4.11085248e-01 6.11767843e-02 -1.16591811e-01 -6.92945570e-02
-2.53322154e-01 1.52082101e-01 -6.09895177e-02 9.80110839e-02
1.40292633e+00 -2.01830611e-01 3.30309004e-01 4.03877228e-01
3.71493101e-01 -1.61423638e-01 -1.28948998e+00 -1.29712269e-01
3.57461125e-01 -4.32615757e-01 5.01776598e-02 -1.15120339e+00
-9.10432458e-01 1.01637113e+00 3.52405518e-01 5.54270804e-01
1.13912201e+00 2.01198682e-01 9.45038572e-02 1.83669746e-01
8.01013485e-02 -5.69290578e-01 1.15911663e-02 -3.74443224e-03
1.00852406e+00 -1.32347214e+00 -5.18026948e-03 -5.60458004e-01
-7.56716430e-01 1.25751984e+00 3.74299079e-01 2.79559165e-01
3.90708297e-01 9.22435056e-03 4.08081617e-03 -3.28516304e-01
-1.11002648e+00 -9.88809764e-02 6.63611472e-01 7.28718877e-01
5.59293807e-01 -7.19909519e-02 -1.99373499e-01 1.30659509e+00
-1.89080447e-01 2.23196931e-02 3.85718286e-01 4.42141116e-01
-3.47589493e-01 -1.02787471e+00 -4.92708474e-01 5.79935312e-01
-3.85184437e-01 -2.47617021e-01 -4.31657553e-01 1.09935772e+00
-1.45430982e-01 1.04772973e+00 -2.99031258e-01 -2.49585375e-01
2.06826061e-01 4.29281920e-01 -8.23009238e-02 -3.64806294e-01
-4.22766924e-01 -2.11402372e-01 1.96695611e-01 -5.29620588e-01
-2.98263669e-01 -7.44858205e-01 -1.09113336e+00 -1.36784524e-01
-5.87051213e-01 3.36187333e-01 6.63394094e-01 1.16088247e+00
4.61084366e-01 6.93555474e-01 5.23590088e-01 -6.09099150e-01
-7.11196601e-01 -7.55483806e-01 -3.33822787e-01 8.68211091e-02
3.50902200e-01 -8.77752602e-01 -7.07551718e-01 1.73338041e-01] | [8.774094581604004, 5.557348728179932] |
5bb8b974-bd04-4ea8-9dad-d454c2a0ccfa | iplan-interactive-and-procedural-layout | 2203.14412 | null | https://arxiv.org/abs/2203.14412v2 | https://arxiv.org/pdf/2203.14412v2.pdf | iPLAN: Interactive and Procedural Layout Planning | Layout design is ubiquitous in many applications, e.g. architecture/urban planning, etc, which involves a lengthy iterative design process. Recently, deep learning has been leveraged to automatically generate layouts via image generation, showing a huge potential to free designers from laborious routines. While automatic generation can greatly boost productivity, designer input is undoubtedly crucial. An ideal AI-aided design tool should automate repetitive routines, and meanwhile accept human guidance and provide smart/proactive suggestions. However, the capability of involving humans into the loop has been largely ignored in existing methods which are mostly end-to-end approaches. To this end, we propose a new human-in-the-loop generative model, iPLAN, which is capable of automatically generating layouts, but also interacting with designers throughout the whole procedure, enabling humans and AI to co-evolve a sketchy idea gradually into the final design. iPLAN is evaluated on diverse datasets and compared with existing methods. The results show that iPLAN has high fidelity in producing similar layouts to those from human designers, great flexibility in accepting designer inputs and providing design suggestions accordingly, and strong generalizability when facing unseen design tasks and limited training data. | ['He Wang', 'Yanlong Huang', 'Feixiang He'] | 2022-03-27 | null | http://openaccess.thecvf.com//content/CVPR2022/html/He_iPLAN_Interactive_and_Procedural_Layout_Planning_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/He_iPLAN_Interactive_and_Procedural_Layout_Planning_CVPR_2022_paper.pdf | cvpr-2022-1 | ['layout-design'] | ['computer-vision'] | [ 1.12116322e-01 2.63345316e-02 1.09766237e-01 -2.04403296e-01
-1.91355318e-01 -7.15033948e-01 8.17983270e-01 -3.83070916e-01
6.61765039e-02 5.32491684e-01 2.81701624e-01 -3.55128527e-01
-1.63480029e-01 -9.15745318e-01 -4.98556763e-01 -4.66302693e-01
4.61490840e-01 6.15633368e-01 -2.17986450e-01 -3.67183357e-01
4.03692782e-01 6.11928105e-01 -1.42499161e+00 1.86759070e-01
9.87872243e-01 5.81457675e-01 4.04018074e-01 5.07995009e-01
-4.35950279e-01 4.23894674e-01 -4.35090691e-01 -6.51053071e-01
1.32597461e-01 -4.28453803e-01 -3.89837593e-01 4.47938889e-01
3.20393831e-01 -3.62045527e-01 -1.73604578e-01 7.99135327e-01
7.28989899e-01 -9.60015599e-03 8.15661311e-01 -1.15928066e+00
-1.23379648e+00 7.74267554e-01 -4.99037296e-01 -7.41686404e-01
5.52010477e-01 7.20913529e-01 1.01923406e+00 -1.12030184e+00
6.42438233e-01 1.29816449e+00 4.87573385e-01 6.25257194e-01
-1.40045547e+00 -7.22329617e-01 3.17052156e-01 -2.48402804e-01
-1.35361183e+00 -4.25893098e-01 1.36559916e+00 -5.78466237e-01
3.83544236e-01 1.85910329e-01 9.41043735e-01 1.33632672e+00
-6.81030080e-02 1.26618278e+00 5.95931470e-01 -2.77168483e-01
4.65879023e-01 1.83656260e-01 -5.75409174e-01 7.08363593e-01
6.09535873e-02 3.76213132e-03 -2.59713054e-01 1.40132189e-01
1.21489799e+00 9.31325555e-02 -1.21452361e-01 -5.97069860e-01
-1.28471422e+00 6.76526070e-01 7.02354550e-01 1.40756622e-01
-5.64349115e-01 2.33001173e-01 2.16958195e-01 4.79211062e-02
7.23438039e-02 1.00772226e+00 -8.86229873e-02 2.08857674e-02
-8.84747684e-01 6.37177289e-01 6.16756797e-01 1.27301693e+00
5.77264607e-01 2.64090687e-01 -2.15534881e-01 8.57775390e-01
1.86487347e-01 5.11908472e-01 2.06188411e-01 -9.46445525e-01
3.60339612e-01 9.29935634e-01 2.16584012e-01 -1.18666828e+00
-1.40894324e-01 -5.92173636e-01 -1.28189552e+00 3.78137350e-01
-6.74285889e-02 -2.32615963e-01 -8.95015597e-01 1.52249610e+00
1.51933506e-01 -3.64231765e-01 -3.18721294e-01 9.91430938e-01
7.21831381e-01 6.11377120e-01 5.14832884e-02 2.38239020e-01
1.20194137e+00 -1.10212493e+00 -6.95332050e-01 -4.23541367e-01
3.10696393e-01 -7.39290714e-01 1.51662576e+00 7.38776088e-01
-1.05545616e+00 -9.41097438e-01 -9.55796599e-01 -4.41926271e-02
-2.20221505e-01 4.71833706e-01 9.71192420e-01 5.40920556e-01
-8.76785100e-01 5.77316344e-01 -2.09582284e-01 -1.93868980e-01
7.82776415e-01 2.00642124e-01 -2.16503099e-01 -2.04487592e-01
-8.48183692e-01 3.43392700e-01 2.70542294e-01 4.45059389e-01
-7.65446186e-01 -7.74687231e-01 -6.82917833e-01 -5.16871326e-02
5.45476317e-01 -1.26598930e+00 1.33287370e+00 -9.35756326e-01
-1.68069720e+00 2.18173891e-01 2.74248779e-01 7.57864043e-02
7.97837198e-01 -4.15441692e-01 -2.65089810e-01 -5.11048794e-01
8.11708346e-02 1.16622984e+00 8.99148881e-01 -1.53282356e+00
-3.40000927e-01 -2.16908623e-02 1.29690051e-01 1.22319721e-01
-1.52582839e-01 -5.78518391e-01 -4.68606144e-01 -8.23486209e-01
1.02829672e-01 -8.85077715e-01 -5.64042807e-01 3.73974562e-01
-6.28804624e-01 -3.21959406e-01 9.21795905e-01 -3.52205515e-01
1.38403404e+00 -2.04705930e+00 2.22927928e-01 2.25690275e-01
1.62997186e-01 3.77973825e-01 -1.76672190e-01 7.71653771e-01
1.40485793e-01 2.32271984e-01 -1.39666751e-01 -3.01618874e-01
4.16394889e-01 7.08613470e-02 -3.03689778e-01 -8.45086053e-02
2.29689032e-01 1.40352774e+00 -1.01526725e+00 -1.86736897e-01
2.96041697e-01 3.65574390e-01 -8.60773683e-01 3.74255478e-01
-6.00608528e-01 3.80560517e-01 -5.38190305e-01 6.70560777e-01
4.56572473e-01 -2.82247335e-01 1.26465216e-01 -2.97054112e-01
-1.55553892e-01 -1.24233119e-01 -1.26057589e+00 2.10503221e+00
-6.63953066e-01 6.54424012e-01 -1.42068043e-01 -6.41947031e-01
1.27357793e+00 1.41361132e-01 -1.33830234e-01 -9.22255099e-01
1.43679351e-01 1.33906215e-01 1.35709196e-01 -5.83182991e-01
6.05503201e-01 1.95862755e-01 -2.82173246e-01 5.66698611e-01
-4.87019032e-01 -4.01738167e-01 1.22685671e-01 1.79830045e-01
9.58475292e-01 4.87487108e-01 8.37327614e-02 -6.88687712e-02
3.22632581e-01 -2.03781694e-01 2.68549770e-01 5.60049176e-01
3.47365111e-01 5.98525167e-01 3.83852541e-01 -6.66237175e-01
-1.30961072e+00 -1.04517233e+00 4.50476319e-01 8.13265443e-01
2.73507996e-03 -4.71526921e-01 -6.70367181e-01 -5.56527317e-01
-7.58904219e-02 8.93787801e-01 -5.83068669e-01 -1.27065018e-01
-6.11692786e-01 6.48528934e-02 3.36859375e-01 7.01719344e-01
5.88128686e-01 -1.52599943e+00 -6.66913867e-01 4.64141846e-01
1.01784103e-01 -6.70687199e-01 -6.79837465e-01 -2.87372142e-01
-6.50739729e-01 -5.75115681e-01 -9.23993528e-01 -8.87560427e-01
1.15093231e+00 2.01524556e-01 1.11034763e+00 1.35532752e-01
-5.48065305e-01 6.13408312e-02 -2.19631866e-01 -2.33694002e-01
-3.29742014e-01 2.43858710e-01 -2.64059484e-01 -1.46507949e-01
-3.15804899e-01 -8.66551518e-01 -9.56185162e-01 5.00557840e-01
-8.58128071e-01 8.80923688e-01 1.16111624e+00 7.98621893e-01
2.98804432e-01 9.82808918e-02 7.87893414e-01 -1.13080347e+00
9.16041255e-01 -1.72572598e-01 -3.06297511e-01 1.88007310e-01
-5.01771092e-01 2.35499069e-01 1.04910827e+00 -5.50135672e-01
-1.10879219e+00 1.91537142e-01 -1.66059256e-01 -3.86482328e-01
-4.00763124e-01 3.81727189e-01 -6.88813150e-01 2.85070926e-01
5.76588988e-01 1.48026273e-01 -2.20183328e-01 -5.16489327e-01
8.20991278e-01 4.34108108e-01 6.05017483e-01 -7.73232818e-01
1.03281999e+00 -6.74545094e-02 -1.28043860e-01 -6.13649189e-01
-2.94544071e-01 1.29756689e-01 -6.47282839e-01 -3.29846799e-01
4.08378541e-01 -4.60231394e-01 -6.17744207e-01 1.95212826e-01
-1.24711025e+00 -3.87521923e-01 -3.10673088e-01 -1.67587340e-01
-4.74921018e-01 5.66750355e-02 -9.94424894e-02 -7.33364165e-01
-4.85353917e-01 -1.31928515e+00 1.13787389e+00 4.14320529e-01
-6.43750608e-01 -8.36889386e-01 -1.90768361e-01 1.44593403e-01
5.40540159e-01 3.01128238e-01 1.21387935e+00 -5.87452091e-02
-9.04483318e-01 -2.95186073e-01 -2.79198408e-01 5.91231845e-02
3.57511044e-01 1.82434350e-01 -7.92792678e-01 1.04917347e-01
-8.42958808e-01 -1.74073175e-01 3.15118164e-01 1.08540520e-01
1.37023699e+00 -4.92550790e-01 -4.00045127e-01 5.22382200e-01
1.12440991e+00 4.08500284e-01 7.97721744e-01 1.24828242e-01
9.51112330e-01 6.49797797e-01 4.53339905e-01 5.53245723e-01
3.26242387e-01 5.73651731e-01 1.94830790e-01 -3.25863093e-01
-1.96032301e-01 -8.94059598e-01 1.74342599e-02 5.49116135e-01
7.14108646e-02 -4.93276715e-01 -7.48628914e-01 4.82075959e-01
-1.98209262e+00 -7.82752156e-01 -3.49290334e-02 1.70500934e+00
6.92744493e-01 2.28778169e-01 7.97116160e-02 2.21151099e-01
4.93592262e-01 2.23790985e-02 -6.28652036e-01 -4.75216210e-01
2.11729571e-01 2.22527906e-01 -8.49192888e-02 3.26405652e-02
-7.61821508e-01 1.09041321e+00 5.36614466e+00 8.27744186e-01
-1.08853769e+00 -6.25384152e-01 7.25181699e-01 1.64269507e-01
-8.58402371e-01 4.41913679e-02 -3.91209930e-01 4.83072072e-01
2.66804427e-01 -3.73933911e-02 6.14653170e-01 1.04559183e+00
4.08213258e-01 1.41307473e-01 -1.33063018e+00 1.12898457e+00
-1.55062884e-01 -1.54010415e+00 4.03774381e-01 1.55130297e-01
8.71292949e-01 -9.37513053e-01 2.42260247e-01 3.10253531e-01
3.06912005e-01 -1.09538960e+00 9.35474515e-01 6.33783519e-01
8.00746441e-01 -9.99162078e-01 3.47172290e-01 4.68719751e-01
-1.37007523e+00 -1.28699997e-02 -9.34160203e-02 -5.88618480e-02
4.36914027e-01 4.91544724e-01 -1.05703211e+00 4.01501238e-01
1.65897161e-01 4.62474942e-01 -7.11268187e-01 1.09788561e+00
-5.09371340e-01 1.58926547e-01 3.62442695e-02 -2.96149403e-01
3.41123879e-01 -1.97044402e-01 1.98127195e-01 1.13296926e+00
5.55824399e-01 -2.37922817e-02 3.58884007e-01 1.39674950e+00
-9.68257263e-02 1.62631392e-01 -8.27511966e-01 -3.58969539e-01
5.42536855e-01 1.31295741e+00 -8.56197298e-01 -7.90072232e-02
7.51528665e-02 1.08725667e+00 2.10991845e-01 4.41208512e-01
-8.98889601e-01 -6.28257990e-01 3.88087541e-01 4.72087711e-01
2.39160046e-01 -3.55250359e-01 -5.89688778e-01 -5.49188375e-01
2.31693760e-02 -1.01363826e+00 -4.36209053e-01 -1.10828996e+00
-1.12235165e+00 5.28417110e-01 -3.73193622e-01 -1.23807633e+00
-1.92732722e-01 -4.63034749e-01 -9.64482307e-01 6.68487906e-01
-9.20574307e-01 -1.42380178e+00 -5.03088892e-01 1.25055179e-01
1.02300751e+00 -2.43609279e-01 5.55146098e-01 3.75594914e-01
-6.32312417e-01 6.82714403e-01 -4.58035201e-01 8.02117512e-02
5.62031269e-01 -1.14122450e+00 9.06824291e-01 6.47798300e-01
1.21581189e-01 9.69473124e-01 6.14434540e-01 -8.56120467e-01
-1.80619192e+00 -1.01725209e+00 4.15780842e-01 -1.17595233e-01
3.77419740e-01 -6.31693602e-01 -6.48120105e-01 1.98255673e-01
3.41034561e-01 -5.07531822e-01 4.64230567e-01 2.12750193e-02
-1.08755939e-01 4.88089658e-02 -7.03908682e-01 1.17232811e+00
1.45957935e+00 -2.35541448e-01 -3.81125033e-01 -4.04149108e-03
7.19411373e-01 -2.04902560e-01 -4.18774903e-01 1.74297124e-01
9.01551008e-01 -8.96932662e-01 9.40648377e-01 -1.98237851e-01
7.12878168e-01 -4.83994901e-01 2.37204999e-01 -1.31847894e+00
-7.59433508e-01 -9.88445997e-01 1.18044987e-01 1.54955006e+00
5.60955167e-01 -1.48213401e-01 8.93234432e-01 7.56877780e-01
-4.01679158e-01 -8.05897534e-01 -6.34742007e-02 -4.91852254e-01
-4.73670930e-01 -5.16574681e-01 1.06032443e+00 5.77625632e-01
-3.43490988e-01 5.70972919e-01 -4.71556365e-01 -2.25284055e-01
4.79992121e-01 2.34893575e-01 1.41149306e+00 -1.41856098e+00
-3.58168244e-01 -8.82431746e-01 4.61407043e-02 -1.27404237e+00
-4.24219295e-02 -7.66955972e-01 3.30704927e-01 -1.90006149e+00
-2.28542596e-01 -6.48246408e-01 3.43313545e-01 3.76800865e-01
-9.31609347e-02 -4.92397286e-02 1.73454121e-01 3.34364213e-02
-4.57724929e-01 8.51657093e-01 1.86713886e+00 -3.05634052e-01
-4.51646030e-01 1.05252154e-01 -1.11618280e+00 7.41960287e-01
5.43927729e-01 -5.25564440e-02 -7.79588103e-01 -4.36109066e-01
5.80942929e-01 -8.44898075e-02 3.82575899e-01 -1.04814529e+00
4.17099416e-01 -2.70869762e-01 6.89242125e-01 -6.36337817e-01
5.14110513e-02 -8.11609566e-01 6.34168565e-01 3.54531020e-01
-3.41142863e-01 1.98979333e-01 6.07063919e-02 4.97720540e-01
1.17375486e-01 -3.20682600e-02 3.74495596e-01 -1.71325594e-01
-8.58716905e-01 3.66813123e-01 -1.20342351e-01 -2.56319255e-01
7.75523245e-01 -4.17031944e-01 -5.46573214e-02 -3.55809897e-01
-4.52672839e-01 2.19381794e-01 6.11305416e-01 5.99703908e-01
7.96235383e-01 -1.64548969e+00 -4.16867375e-01 5.03223360e-01
1.90934718e-01 3.12447220e-01 3.19391191e-01 2.38498807e-01
-6.78376853e-01 2.43259937e-01 -1.49393499e-01 -4.62330252e-01
-7.46018171e-01 6.55323446e-01 -2.78426051e-01 -5.83169535e-02
-8.11690867e-01 7.57721305e-01 4.67347831e-01 -3.65753323e-01
3.38476896e-01 -3.30961049e-01 1.33193687e-01 -1.87305117e-03
4.85843748e-01 2.23001063e-01 -3.07995498e-01 4.25908528e-02
5.33182845e-02 5.98181427e-01 -1.27255231e-01 -1.51974365e-01
1.31507087e+00 1.29693121e-01 1.37738138e-01 9.54093635e-02
7.05796540e-01 -1.44360587e-01 -1.48810422e+00 1.40512027e-02
-7.34558478e-02 -5.98890007e-01 -7.39672258e-02 -1.01668406e+00
-1.02300572e+00 9.61045742e-01 2.51358151e-01 6.89931363e-02
9.45333362e-01 -2.31937215e-01 9.16430175e-01 5.80750763e-01
5.65195322e-01 -9.70473528e-01 5.36691129e-01 1.59503832e-01
1.30252993e+00 -9.10461307e-01 -9.69249159e-02 -1.94035321e-01
-9.22794938e-01 1.05734742e+00 7.97482014e-01 -1.25846505e-01
1.61016241e-01 2.54396260e-01 -2.47422278e-01 -1.37944922e-01
-7.49699831e-01 4.77503948e-02 5.44783890e-01 6.43157184e-01
5.41650414e-01 8.02475587e-02 6.02700002e-02 6.35849714e-01
-3.95454198e-01 1.77146628e-01 9.61890146e-02 8.23930919e-01
-3.28323245e-01 -1.33465719e+00 -6.34009242e-02 1.80561751e-01
4.45917070e-01 3.06412503e-02 -6.21295333e-01 7.88959026e-01
3.12501371e-01 5.28032780e-01 -2.12264135e-01 -4.44832563e-01
6.69418395e-01 -9.81076881e-02 2.44484991e-01 -4.29560781e-01
-6.48125648e-01 1.91486970e-01 2.82449797e-02 -3.66543174e-01
1.51883900e-01 -3.27347964e-01 -1.01429677e+00 -3.72317344e-01
-1.08543649e-01 2.79127099e-02 6.23740911e-01 6.50447726e-01
5.25887311e-01 8.07031214e-01 6.52586639e-01 -1.24220288e+00
-3.40643637e-02 -6.09254122e-01 -2.25944921e-01 2.49585122e-01
-9.15064514e-02 -6.90432727e-01 7.42240697e-02 6.58616237e-03] | [11.345625877380371, -0.23379634320735931] |
780f90e6-e555-4ace-bd29-6fc2d7244d84 | escaping-saddle-points-with-bias-variance | 2202.06083 | null | https://arxiv.org/abs/2202.06083v3 | https://arxiv.org/pdf/2202.06083v3.pdf | Escaping Saddle Points with Bias-Variance Reduced Local Perturbed SGD for Communication Efficient Nonconvex Distributed Learning | In recent centralized nonconvex distributed learning and federated learning, local methods are one of the promising approaches to reduce communication time. However, existing work has mainly focused on studying first-order optimality guarantees. On the other side, second-order optimality guaranteed algorithms, i.e., algorithms escaping saddle points, have been extensively studied in the non-distributed optimization literature. In this paper, we study a new local algorithm called Bias-Variance Reduced Local Perturbed SGD (BVR-L-PSGD), that combines the existing bias-variance reduced gradient estimator with parameter perturbation to find second-order optimal points in centralized nonconvex distributed optimization. BVR-L-PSGD enjoys second-order optimality with nearly the same communication complexity as the best known one of BVR-L-SGD to find first-order optimality. Particularly, the communication complexity is better than non-local methods when the local datasets heterogeneity is smaller than the smoothness of the local loss. In an extreme case, the communication complexity approaches to $\widetilde \Theta(1)$ when the local datasets heterogeneity goes to zero. Numerical results validate our theoretical findings. | ['Taiji Suzuki', 'Tomoya Murata'] | 2022-02-12 | null | null | null | null | ['distributed-optimization'] | ['methodology'] | [-7.46697187e-01 -1.05236694e-02 -2.41274208e-01 -1.37640357e-01
-1.51574218e+00 -2.94577390e-01 -2.50922382e-01 2.33248964e-01
-3.77471030e-01 1.17922843e+00 1.22109577e-01 -2.44819194e-01
-5.54151773e-01 -5.66399693e-01 -1.01872420e+00 -1.24845839e+00
-2.50302166e-01 1.56692654e-01 -7.89785534e-02 -6.98290439e-03
-9.42990333e-02 1.02609925e-01 -8.33784223e-01 -3.66332948e-01
1.26546979e+00 1.22412312e+00 5.94215021e-02 1.81704924e-01
7.35439500e-03 4.71873492e-01 -3.15558463e-01 -2.76185066e-01
6.95159733e-01 -5.58200896e-01 -4.23422873e-01 -1.38396770e-01
4.97538924e-01 -2.10202336e-01 3.47553380e-03 1.56587327e+00
8.39796126e-01 2.52136201e-01 1.59310937e-01 -1.29770184e+00
-2.26149827e-01 7.50225127e-01 -9.05654252e-01 2.01048508e-01
-4.71842960e-02 -3.22166979e-02 8.21848214e-01 -9.26270723e-01
5.02868652e-01 1.24495912e+00 1.03279364e+00 2.31335267e-01
-1.02874231e+00 -6.12478077e-01 4.91532207e-01 -5.98569103e-02
-1.59506118e+00 -1.11270539e-01 5.93685091e-01 -3.26123357e-01
1.33746982e-01 3.19460988e-01 2.89468288e-01 3.07012528e-01
2.45530114e-01 8.76555204e-01 1.15110791e+00 -2.02742890e-01
5.22707939e-01 1.64615214e-01 -4.90505248e-02 8.45986664e-01
6.99420989e-01 -1.08124837e-01 -5.20177543e-01 -5.51584303e-01
4.04925644e-01 -7.37734959e-02 -5.29136598e-01 -4.71639097e-01
-9.89256561e-01 8.04408431e-01 5.47649980e-01 2.36267895e-01
-4.90974754e-01 1.14010170e-01 3.94933730e-01 4.43753392e-01
1.04304183e+00 -1.99526772e-02 -6.78182721e-01 -1.07668586e-01
-8.39224517e-01 2.32418939e-01 1.00731337e+00 1.01375759e+00
8.88706326e-01 2.27189764e-01 -2.94017404e-01 6.61575198e-01
3.32814366e-01 5.84904253e-01 1.60015151e-01 -9.02666390e-01
9.63415265e-01 4.04665619e-01 2.82784909e-01 -1.24279666e+00
-3.08833539e-01 -1.07871199e+00 -1.12839758e+00 -2.52228342e-02
5.32041669e-01 -7.47328937e-01 2.51828820e-01 1.57632184e+00
9.66393173e-01 1.08539030e-01 -1.77466664e-02 1.33261502e+00
3.73830587e-01 6.71799183e-01 -4.16540951e-01 -6.81782603e-01
7.50825405e-01 -1.10792363e+00 -6.15392923e-01 2.96871066e-01
9.29991007e-01 -5.59004664e-01 9.35947716e-01 2.87990183e-01
-1.12414551e+00 1.75001957e-02 -8.04142654e-01 3.87861341e-01
1.70727521e-01 6.52291626e-02 3.53213161e-01 6.40704215e-01
-1.03263199e+00 5.05000234e-01 -7.72815108e-01 -4.29845974e-02
6.54229641e-01 1.80699021e-01 5.64057380e-02 -2.66658634e-01
-6.68747544e-01 1.83322936e-01 -1.07527189e-01 5.36536127e-02
-1.05659139e+00 -1.21682000e+00 -3.23080599e-01 5.63594438e-02
7.99097598e-01 -6.99414015e-01 9.70584035e-01 -7.80930281e-01
-1.68774390e+00 2.72423297e-01 -7.09328726e-02 -3.09612423e-01
9.35204983e-01 -2.21416995e-01 1.94248691e-01 -1.37424886e-01
1.35239348e-01 -2.95656443e-01 5.47087073e-01 -1.24428248e+00
-7.36522853e-01 -6.77014649e-01 -4.76370826e-02 4.65145946e-01
-7.87136376e-01 -2.18157202e-01 -6.78768978e-02 -5.80280781e-01
-2.97122952e-02 -7.59257734e-01 -4.76306558e-01 1.94502279e-01
-2.98015863e-01 -2.93044835e-01 8.30012679e-01 -5.69956720e-01
1.22516799e+00 -2.19367886e+00 1.07491560e-01 2.83765376e-01
2.73525685e-01 3.05933226e-02 -1.80886767e-03 3.68828088e-01
6.01397097e-01 1.91405073e-01 -6.65286556e-02 -4.17237341e-01
1.12854674e-01 -1.45459205e-01 6.73077479e-02 1.19484913e+00
-7.12181807e-01 5.80775917e-01 -1.06870389e+00 -4.41473484e-01
-2.08458900e-01 2.23226845e-01 -6.88293695e-01 -2.42716447e-02
4.66584563e-02 4.92990643e-01 -9.36631024e-01 5.04210413e-01
9.18490112e-01 -3.33204031e-01 1.11019492e-01 -1.03200927e-01
-4.22560662e-01 -2.14764714e-01 -1.51838732e+00 1.59738481e+00
-7.45181739e-01 2.24858657e-01 1.05714476e+00 -1.36054063e+00
8.02598655e-01 1.72165841e-01 1.00032890e+00 -2.48962343e-01
-6.70161843e-02 7.14662313e-01 -4.88869905e-01 -3.05537343e-01
-4.69402112e-02 -7.67624602e-02 7.84228966e-02 4.43165481e-01
-3.56789619e-01 3.00731391e-01 -9.63543877e-02 7.56894946e-02
1.09492099e+00 -1.15400024e-01 2.07003042e-01 -8.82553101e-01
7.29756474e-01 -2.01450333e-01 9.96140301e-01 7.80397296e-01
-2.32558370e-01 2.56317347e-01 2.52690792e-01 -2.11588204e-01
-6.94215059e-01 -8.14276159e-01 5.08208806e-03 1.04298103e+00
4.58754897e-01 -3.42570603e-01 -8.25142860e-01 -7.77888179e-01
3.08073401e-01 3.34940702e-01 -5.23730218e-02 6.40536174e-02
-4.30217862e-01 -8.20501268e-01 2.51086831e-01 -5.12061864e-02
9.53574359e-01 -1.29530668e-01 -1.73478633e-01 3.16450536e-01
-4.73554097e-02 -6.95182502e-01 -1.00259316e+00 -1.09656207e-01
-1.02821016e+00 -9.16751027e-01 -1.04027152e+00 -7.41587520e-01
9.23568964e-01 3.73927683e-01 7.71096468e-01 -3.98926660e-02
-8.99749175e-02 5.66760421e-01 -2.20685124e-01 -3.23665291e-01
2.01491952e-01 2.82117546e-01 9.08981711e-02 4.41311717e-01
-2.56134480e-01 -4.16000098e-01 -8.18130791e-01 5.69805026e-01
-5.87492168e-01 -5.43120086e-01 4.28752929e-01 9.14177120e-01
1.02547050e+00 2.02648267e-01 7.56224334e-01 -7.36469746e-01
9.08318996e-01 -7.95948207e-01 -9.86687481e-01 3.08818877e-01
-8.66247416e-01 -3.47197428e-02 1.30180979e+00 -3.09708536e-01
-9.73718822e-01 -4.65670899e-02 4.34246004e-01 -5.70998609e-01
4.23516870e-01 7.84981430e-01 -2.53841132e-01 -4.48845565e-01
4.79295433e-01 3.01932186e-01 1.48660894e-02 -6.44811273e-01
3.24068606e-01 5.80458283e-01 8.70178193e-02 -8.63648415e-01
7.19712496e-01 6.74320400e-01 1.92889646e-01 -7.09887624e-01
-8.77471328e-01 -6.19809091e-01 9.57470611e-02 -2.44425341e-01
2.67120421e-01 -8.36208820e-01 -9.01373386e-01 3.71350199e-01
-6.80420637e-01 -3.30901504e-01 -5.66492558e-01 7.61834502e-01
-4.76992548e-01 4.71248567e-01 -1.80145070e-01 -1.09002769e+00
-6.88420415e-01 -8.56318176e-01 8.25237155e-01 2.15653539e-01
5.48875451e-01 -1.27101016e+00 1.42982423e-01 2.13685766e-01
7.66492128e-01 3.84664774e-01 4.02982026e-01 -3.94151360e-01
-6.08996451e-01 -1.03039108e-01 -6.56962916e-02 3.59160304e-01
1.07187279e-01 -5.61632037e-01 -1.19039603e-01 -8.90767634e-01
3.13867331e-01 -2.09521174e-01 4.06212330e-01 6.37024939e-01
1.22402489e+00 -1.09164762e+00 -3.46149832e-01 9.44546998e-01
1.77897286e+00 -2.60860085e-01 -7.15406537e-02 1.60036832e-01
4.76599991e-01 1.44692078e-01 7.83690572e-01 1.04583037e+00
4.76453155e-01 5.79571486e-01 4.98210430e-01 -4.75828499e-02
-1.20061021e-02 -1.47285357e-01 4.41244096e-01 1.03737116e+00
9.25711095e-02 -1.77935481e-01 -6.81120992e-01 5.01904547e-01
-2.23023033e+00 -6.74864769e-01 -2.78070688e-01 2.45396709e+00
8.53719532e-01 -4.61778164e-01 1.25972718e-01 -3.13148946e-01
1.02893019e+00 6.14543855e-02 -8.43238533e-01 -1.35983124e-01
-2.06756413e-01 -3.15343380e-01 9.15156484e-01 5.33525884e-01
-7.49227107e-01 4.70236272e-01 4.72235632e+00 1.17074502e+00
-9.26409423e-01 5.91331303e-01 6.81770802e-01 -3.08014959e-01
-3.36145192e-01 9.48209018e-02 -8.14692497e-01 7.08192468e-01
5.05592883e-01 -5.88989139e-01 7.41332531e-01 1.12830603e+00
6.03273809e-01 -4.88981558e-03 -6.13571942e-01 1.19886446e+00
-1.48259342e-01 -1.20402551e+00 -4.90444213e-01 3.27370584e-01
1.38849139e+00 1.17171466e-01 -2.67345347e-02 -4.76779742e-03
4.42015827e-01 -4.97698247e-01 3.71264011e-01 5.73186815e-01
6.75231934e-01 -1.00308168e+00 8.37122679e-01 6.96522355e-01
-1.35659397e+00 -3.80988687e-01 -6.24010682e-01 1.30912066e-01
1.29482508e-01 1.26207221e+00 -1.16781123e-01 9.28706825e-01
9.16753054e-01 7.80369282e-01 -2.65889298e-02 1.49631977e+00
1.41512141e-01 7.57995188e-01 -8.89420033e-01 -3.16827536e-01
3.14829499e-01 -6.20240748e-01 1.02816522e+00 6.75659418e-01
5.66954911e-01 3.97182778e-02 6.94370747e-01 5.40697753e-01
-3.86767089e-01 7.74551928e-01 -3.73593658e-01 5.12792051e-01
6.21065259e-01 1.18794250e+00 -2.08963946e-01 -1.14602938e-01
-3.74597639e-01 5.61335266e-01 4.46235716e-01 2.95533389e-01
-6.75394714e-01 -5.22540808e-01 5.39574087e-01 -4.80096154e-02
4.04867589e-01 -2.06836745e-01 -1.66841105e-01 -1.23738074e+00
5.16010523e-01 -7.73744524e-01 5.70073903e-01 4.51301225e-02
-1.52642310e+00 2.45347559e-01 -1.86286122e-01 -1.22108459e+00
3.08932662e-01 2.77613779e-03 -5.76544702e-01 5.74471354e-01
-1.44221175e+00 -7.82944262e-01 -3.85254711e-01 7.95251966e-01
3.96515816e-01 -2.27184370e-01 4.34186906e-01 5.74874640e-01
-7.53124774e-01 9.72933769e-01 1.00841343e+00 -3.76856893e-01
5.70768714e-01 -9.94189978e-01 -7.51222312e-01 7.46677458e-01
-3.09262812e-01 3.88611913e-01 6.43100798e-01 -4.56441969e-01
-1.80174994e+00 -1.23414898e+00 5.34718454e-01 3.98536801e-01
5.75963378e-01 -5.28963022e-02 -6.10674441e-01 2.19064936e-01
4.95488048e-02 6.27337456e-01 4.81150955e-01 -3.76268551e-02
1.31873742e-01 -9.95851815e-01 -1.39846122e+00 3.39640081e-01
1.12079144e+00 -4.34517525e-02 3.85890633e-01 8.21910441e-01
4.87507850e-01 -4.16055650e-01 -1.12083387e+00 2.24756211e-01
3.62759158e-02 -7.31590390e-01 6.35104299e-01 -3.90202165e-01
-1.64745882e-01 -2.30042607e-01 -3.49024564e-01 -1.44803894e+00
4.21474427e-02 -1.31242239e+00 -2.69214392e-01 1.27628469e+00
2.74475694e-01 -1.11067140e+00 9.47784007e-01 1.43373221e-01
-2.04268470e-01 -1.16727173e+00 -1.47652304e+00 -1.31387746e+00
2.65255302e-01 -1.59431752e-02 4.72693563e-01 8.49340379e-01
-1.89480454e-01 -7.04817399e-02 -4.05126423e-01 3.28107983e-01
1.16336834e+00 2.78044611e-01 1.00330007e+00 -8.04601312e-01
-4.54230547e-01 -2.72556096e-01 -5.86711355e-02 -1.56403291e+00
9.00280774e-02 -1.09137630e+00 1.28116578e-01 -1.46365225e+00
2.81914286e-02 -8.50785971e-01 -3.11017692e-01 2.18097940e-01
-1.30826429e-01 -3.12753707e-01 1.73341379e-01 4.77700949e-01
-9.61316943e-01 8.83885682e-01 1.29541254e+00 -8.26570764e-02
-4.31900680e-01 2.79398441e-01 -7.39385962e-01 5.85857987e-01
8.29645336e-01 -7.99928546e-01 -3.90297025e-01 -4.93964732e-01
3.80669862e-01 2.27228194e-01 1.12257019e-01 -9.10651922e-01
5.78623235e-01 -4.27525878e-01 -3.58310401e-01 -4.05128926e-01
-1.39689788e-01 -9.83224213e-01 2.14250460e-02 5.80807805e-01
-3.06362599e-01 -3.88492160e-02 -3.16474795e-01 9.63095784e-01
-2.49571443e-01 -1.16619527e-01 7.77817547e-01 -4.78288122e-02
-7.03861564e-02 6.59394681e-01 -1.64993778e-01 5.91273963e-01
1.28632569e+00 1.35226980e-01 -3.88682902e-01 -6.52237535e-01
-2.97505528e-01 6.93289399e-01 1.09450780e-01 -1.16670758e-01
4.67762709e-01 -1.36729789e+00 -9.85670984e-01 -3.86397600e-01
-5.12179494e-01 3.48864853e-01 3.30950677e-01 1.57503784e+00
-3.76438737e-01 3.37758869e-01 5.52866220e-01 -7.07163036e-01
-1.06817210e+00 3.21831465e-01 5.94922185e-01 -3.77472758e-01
-5.42572916e-01 8.87748480e-01 -1.06181167e-01 -5.29787004e-01
3.29972744e-01 -5.33744954e-02 5.35340905e-01 -2.80231625e-01
1.52179912e-01 1.14561725e+00 1.39700547e-01 -1.32903099e-01
-4.47231919e-01 6.12491667e-01 2.43121266e-01 2.12781444e-01
1.55126381e+00 -4.85260248e-01 -3.59618366e-01 2.68962979e-01
1.76411176e+00 3.61152589e-01 -1.40390229e+00 -4.42744374e-01
-2.00609148e-01 -6.64659262e-01 3.49702895e-01 -5.31062007e-01
-1.64164340e+00 3.63257527e-01 6.54404104e-01 1.29695043e-01
1.13263404e+00 -1.33471802e-01 9.08947051e-01 3.91104847e-01
9.39583063e-01 -1.25074482e+00 6.88197138e-03 2.83840835e-01
8.34001720e-01 -1.09520316e+00 3.72850657e-01 -3.55288595e-01
-3.15376580e-01 9.56189036e-01 5.93183100e-01 -4.26400661e-01
9.79422152e-01 8.29433426e-02 -2.27804929e-01 -5.29182376e-03
-5.02498567e-01 1.75240308e-01 2.75773145e-02 1.47097439e-01
5.81582598e-02 5.57593815e-02 -1.01288652e+00 6.46764040e-01
9.67132021e-03 -1.30313158e-01 8.44856352e-02 7.91827202e-01
-3.52146864e-01 -7.06338823e-01 -4.89303648e-01 5.00826359e-01
-7.68274724e-01 1.58462718e-01 3.24093527e-03 6.02867723e-01
-3.24422773e-03 9.81825292e-01 -3.64049971e-01 1.26491711e-01
1.43696964e-01 -5.13202012e-01 1.99166924e-01 -2.42167905e-01
-5.94246924e-01 1.70496285e-01 -1.03715144e-01 -7.56276190e-01
-3.05225730e-01 -6.27874315e-01 -1.26108742e+00 -3.18659306e-01
-5.63939691e-01 5.93280375e-01 8.20054889e-01 7.01235294e-01
6.77733958e-01 8.02881643e-02 1.23313034e+00 -4.59776849e-01
-1.32059264e+00 -5.05440176e-01 -7.52500236e-01 9.98502448e-02
2.65519261e-01 -2.45749146e-01 -8.09369743e-01 -5.35415173e-01] | [6.30286169052124, 4.901219367980957] |
0a21e90d-3551-4ae8-a2c9-23472f553e1f | fine-grained-session-recommendations-in-e | 2210.15451 | null | https://arxiv.org/abs/2210.15451v1 | https://arxiv.org/pdf/2210.15451v1.pdf | Fine-Grained Session Recommendations in E-commerce using Deep Reinforcement Learning | Sustaining users' interest and keeping them engaged in the platform is very important for the success of an e-commerce business. A session encompasses different activities of a user between logging into the platform and logging out or making a purchase. User activities in a session can be classified into two groups: Known Intent and Unknown intent. Known intent activity pertains to the session where the intent of a user to browse/purchase a specific product can be easily captured. Whereas in unknown intent activity, the intent of the user is not known. For example, consider the scenario where a user enters the session to casually browse the products over the platform, similar to the window shopping experience in the offline setting. While recommending similar products is essential in the former, accurately understanding the intent and recommending interesting products is essential in the latter setting in order to retain a user. In this work, we focus primarily on the unknown intent setting where our objective is to recommend a sequence of products to a user in a session to sustain their interest, keep them engaged and possibly drive them towards purchase. We formulate this problem in the framework of the Markov Decision Process (MDP), a popular mathematical framework for sequential decision making and solve it using Deep Reinforcement Learning (DRL) techniques. However, training the next product recommendation is difficult in the RL paradigm due to large variance in browse/purchase behavior of the users. Therefore, we break the problem down into predicting various product attributes, where a pattern/trend can be identified and exploited to build accurate models. We show that the DRL agent provides better performance compared to a greedy strategy. | ['Sreekanth Vempati', 'Saif Jawaid', 'Lakshya Kumar', 'Diddigi Raghu Ram Bharadwaj'] | 2022-10-20 | null | null | null | null | ['product-recommendation'] | ['miscellaneous'] | [ 1.02392465e-01 -1.19291529e-01 -7.58281887e-01 -4.55321878e-01
-2.82699853e-01 -6.52261555e-01 3.48454446e-01 4.00158197e-01
-2.92267799e-01 3.30191851e-01 -3.04581691e-02 -5.68700314e-01
-3.81931633e-01 -8.05549502e-01 -6.70025408e-01 -8.43141913e-01
-3.90918255e-01 4.92072970e-01 -3.01405728e-01 -2.45352671e-01
2.52142936e-01 2.77769774e-01 -1.39322555e+00 5.41295782e-02
5.93398035e-01 1.28309608e+00 5.80159426e-01 5.80399513e-01
-1.86806485e-01 6.96516335e-01 -1.49190398e-02 -2.88316488e-01
5.87229073e-01 -3.69811624e-01 -6.38379872e-01 5.16718805e-01
-3.48189056e-01 -4.27858412e-01 1.15837052e-01 8.46295774e-01
-5.35633834e-03 5.42132974e-01 4.68289703e-01 -1.38814235e+00
-3.55498135e-01 7.31982529e-01 -5.34609258e-01 1.16553187e-01
5.44584870e-01 -9.40721203e-03 1.39765775e+00 -2.39581525e-01
3.83605480e-01 7.08080947e-01 2.24379674e-01 2.22326294e-01
-1.33725011e+00 -5.52794456e-01 9.09332752e-01 2.44588882e-01
-9.99610603e-01 -2.29474425e-01 7.16708243e-01 -4.64469999e-01
5.04440486e-01 4.10334915e-02 8.41696262e-01 1.04055846e+00
3.39318007e-01 1.20688236e+00 6.67992890e-01 -1.03239223e-01
8.56860578e-01 2.57310510e-01 2.57773370e-01 1.22751743e-01
9.98530388e-02 5.83672784e-02 -4.70282346e-01 -1.99632086e-02
6.14403903e-01 7.26221681e-01 -2.92406287e-02 -5.49206316e-01
-7.56457627e-01 1.21214402e+00 3.77627075e-01 2.88221091e-01
-1.25150561e+00 -2.35436797e-01 -4.08400446e-02 6.24318600e-01
-2.32749023e-02 3.41250360e-01 -4.43165272e-01 -2.78917670e-01
-9.50330496e-01 4.23356652e-01 1.30841255e+00 9.06971931e-01
7.04393744e-01 -3.27787161e-01 2.67645810e-02 2.97826856e-01
5.09090841e-01 -3.32904644e-02 3.71071011e-01 -7.60886669e-01
2.90948123e-01 4.73795056e-01 7.77089953e-01 -8.01131368e-01
-2.80847102e-01 -3.56652588e-01 -6.95725977e-01 1.46208137e-01
6.37307584e-01 -2.85723895e-01 -6.21414721e-01 1.63745177e+00
2.22483113e-01 8.33506063e-02 1.41235041e-02 9.52351570e-01
2.97056064e-02 7.88402081e-01 3.16601753e-01 -7.52915621e-01
1.46572721e+00 -7.43139327e-01 -6.20836139e-01 -3.20071816e-01
3.75499845e-01 -6.95607126e-01 8.82106543e-01 7.70236969e-01
-9.10812378e-01 -4.58336771e-01 -9.26688910e-01 3.84955555e-01
-1.36751890e-01 -8.47478062e-02 8.49826396e-01 3.69968623e-01
-6.01151228e-01 7.63616741e-01 -8.31185281e-01 -4.91581649e-01
1.85251921e-01 5.79893887e-01 1.58004120e-01 -2.34898806e-01
-1.13070571e+00 5.20373881e-01 7.00127631e-02 7.27804527e-02
-8.64650428e-01 -6.02107584e-01 -6.42486572e-01 3.77546221e-01
8.57911825e-01 -4.35706377e-01 1.60722435e+00 -1.19346368e+00
-1.62478542e+00 3.93050104e-01 -2.47516558e-01 -7.76817918e-01
5.65636992e-01 -1.87961072e-01 -3.39549214e-01 -3.44646335e-01
-1.40952198e-02 2.56822884e-01 1.04970503e+00 -9.32375133e-01
-1.22975886e+00 -5.72658122e-01 5.62402070e-01 5.05973935e-01
-1.83743555e-02 -1.69284359e-01 -3.91223550e-01 -1.60795018e-01
7.15039298e-02 -1.12111378e+00 -4.57256168e-01 -3.72502089e-01
-5.36367781e-02 -3.83297801e-01 6.86973989e-01 -6.44106090e-01
1.13880992e+00 -2.05236959e+00 2.27249004e-02 1.82715356e-01
-6.05570972e-02 -1.56807125e-01 1.07261129e-01 7.01113284e-01
2.51398951e-01 -3.50502506e-03 3.82476628e-01 -2.88427144e-01
6.75672740e-02 2.92134434e-01 -2.68972486e-01 2.12152407e-01
-2.87925661e-01 7.57260799e-01 -8.22779357e-01 3.27931158e-02
-1.23599339e-02 1.63219035e-01 -5.17260611e-01 5.19010186e-01
-4.48601097e-01 5.34743249e-01 -9.48627770e-01 5.14022708e-01
4.69980866e-01 -4.01803434e-01 5.18379807e-01 1.04439169e-01
-9.55254361e-02 4.07613844e-01 -1.46457589e+00 1.31746662e+00
-9.22005475e-01 -3.69316824e-02 3.81249547e-01 -1.00182867e+00
5.04005373e-01 2.17567652e-01 6.99884057e-01 -6.59242809e-01
2.02584770e-02 -2.18202829e-01 2.20202610e-01 -3.46731544e-01
5.83970964e-01 -3.25167179e-01 -6.54889643e-02 9.27787244e-01
-3.47240448e-01 4.98132676e-01 3.66946727e-01 1.14909731e-01
8.88245404e-01 1.24879196e-01 7.17459261e-01 -1.04785040e-01
2.18906656e-01 -2.32947290e-01 5.89483678e-01 1.07351148e+00
-2.25230187e-01 -1.06156565e-01 5.48277497e-01 -3.40726852e-01
-6.10105991e-01 -8.84123802e-01 3.03082198e-01 1.49396300e+00
4.29774284e-01 -8.60942453e-02 -2.96573251e-01 -6.44084513e-01
7.23726153e-02 1.00637054e+00 -5.04075170e-01 -1.22469828e-01
-2.39642277e-01 -5.41842043e-01 -5.50563753e-01 3.81876886e-01
5.19593954e-01 -1.19351745e+00 -7.74173200e-01 5.75908124e-01
-2.38586470e-01 -7.99850821e-01 -7.69963384e-01 3.00863594e-01
-7.91920662e-01 -8.09279561e-01 -4.42285329e-01 -6.07821941e-01
3.95714879e-01 3.67650449e-01 1.03079855e+00 -2.16894671e-01
2.07138583e-01 3.13451320e-01 -4.33216631e-01 -3.32290530e-01
-3.13923150e-01 2.00231493e-01 2.69930542e-01 5.09309709e-01
5.84337592e-01 -5.42171240e-01 -1.07420075e+00 5.81488550e-01
-8.09832633e-01 -1.06675975e-01 7.11651087e-01 4.82363224e-01
5.43066978e-01 6.86905444e-01 5.21945536e-01 -7.59887516e-01
5.85488260e-01 -8.29149306e-01 -6.27533317e-01 1.52654648e-01
-8.34053457e-01 -1.37601748e-01 5.93528152e-01 -7.49628782e-01
-1.02712774e+00 2.58324057e-01 -1.61351070e-01 -7.15582073e-02
-1.80289119e-01 6.50007665e-01 -2.34406367e-01 4.09386396e-01
-4.34625149e-02 3.06803852e-01 -5.13745286e-02 -5.52479208e-01
1.42854899e-01 6.12831712e-01 -1.67311460e-01 -2.57340044e-01
3.23984265e-01 1.87078878e-01 -2.60042071e-01 -6.81906462e-01
-9.42288458e-01 -9.11916375e-01 -3.81444544e-01 -9.79571715e-02
5.37275910e-01 -6.99351013e-01 -1.44069886e+00 4.54668179e-02
-4.89167720e-01 -4.36153293e-01 -3.72152120e-01 5.14545262e-01
-6.55203342e-01 2.43963543e-02 -6.56427562e-01 -1.11136580e+00
-2.58613303e-02 -1.17979121e+00 7.62900651e-01 4.40083921e-01
-2.81288236e-01 -1.07046044e+00 -2.67618448e-01 4.76369709e-01
2.58534104e-01 -4.07187998e-01 8.27450335e-01 -1.09623730e+00
-8.47532690e-01 -3.54957670e-01 3.20406944e-01 5.60222268e-02
4.30429608e-01 -4.89548206e-01 -3.90124261e-01 -4.78338331e-01
3.16082567e-01 -4.94453050e-02 3.62546414e-01 7.09672749e-01
7.92336106e-01 -5.28601527e-01 -2.54141986e-01 -2.55601108e-02
1.40514827e+00 7.58577824e-01 3.14839870e-01 3.38092864e-01
1.22095913e-01 6.90936625e-01 1.16135991e+00 6.72740400e-01
4.96725827e-01 7.79117167e-01 4.77282375e-01 2.59272277e-01
8.01743090e-01 -5.14357567e-01 5.81908643e-01 1.38526022e-01
1.27865393e-02 -3.92737538e-01 -3.35755646e-01 1.09012462e-01
-2.07077336e+00 -9.06649351e-01 4.28048730e-01 2.63272572e+00
5.06053329e-01 2.19639599e-01 5.18296599e-01 -3.31383832e-02
5.78582644e-01 -2.74427515e-02 -1.04137683e+00 -3.56231958e-01
5.82110822e-01 -3.93086523e-01 3.54048967e-01 4.93291616e-01
-1.00561798e+00 7.56888807e-01 5.18149137e+00 5.79155445e-01
-9.90242600e-01 -5.58798872e-02 8.75652075e-01 -1.72822446e-01
-1.06444983e-02 4.98619564e-02 -1.25142610e+00 4.70177293e-01
1.01209188e+00 -1.36881232e-01 7.80424595e-01 9.54599202e-01
7.30244040e-01 -3.90044004e-01 -1.58187234e+00 8.14347267e-01
-2.70546764e-01 -7.98055708e-01 -2.32593507e-01 5.14336228e-01
4.49695528e-01 -3.77509534e-01 1.65015772e-01 5.84863424e-01
5.59428334e-01 -7.37863243e-01 5.94950378e-01 4.18840379e-01
-6.35600388e-02 -7.75629759e-01 5.75665951e-01 8.55481505e-01
-1.12528980e+00 -6.07116222e-01 -2.80610397e-02 -3.48906696e-01
5.54013193e-01 4.86538470e-01 -8.28071296e-01 3.66106927e-01
5.66854298e-01 7.17294693e-01 1.34271219e-01 8.41461957e-01
1.14257261e-02 5.51406980e-01 -2.59934217e-01 -2.12170422e-01
3.92127395e-01 -8.10689569e-01 3.22677135e-01 4.38698858e-01
4.90301013e-01 1.45701513e-01 5.66749275e-01 7.84524679e-01
9.89881158e-03 2.67592788e-01 -4.40524608e-01 -3.30885708e-01
2.48075724e-02 1.30507016e+00 -9.01359499e-01 6.43956533e-04
-5.10474205e-01 1.19478261e+00 -2.14806385e-02 4.58425701e-01
-6.22474015e-01 1.12741180e-01 8.66685569e-01 4.42207575e-01
6.81361258e-01 -9.41017196e-02 -1.62968356e-02 -9.75688815e-01
2.60527432e-02 -8.88665020e-01 4.75504488e-01 -3.71596158e-01
-1.08075500e+00 2.07292408e-01 -3.68500292e-01 -1.15171814e+00
-4.86941248e-01 -2.99632549e-01 -5.88359356e-01 7.36339688e-01
-1.47549176e+00 -8.48722875e-01 9.91203785e-02 4.49716777e-01
8.87646794e-01 1.26910314e-01 4.33310866e-01 1.55229092e-01
-4.77082878e-01 3.17570210e-01 1.09290272e-01 -3.17871809e-01
2.56794989e-01 -1.17473054e+00 -5.78192137e-02 4.96702850e-01
2.01999992e-01 8.56483698e-01 8.03767860e-01 -7.52592385e-01
-1.69586968e+00 -7.60741532e-01 9.10552979e-01 -8.95393491e-02
6.86337173e-01 -4.07029241e-01 -6.44769728e-01 9.42228079e-01
-7.15145022e-02 -5.48602879e-01 8.28429043e-01 4.33932334e-01
3.32172871e-01 -1.18833318e-01 -1.01951706e+00 7.45805621e-01
6.81903839e-01 -2.79982779e-02 -3.11991781e-01 2.98304230e-01
5.29339075e-01 1.11962304e-01 -8.16703379e-01 -1.11141369e-01
5.78089356e-01 -7.85309196e-01 8.58623385e-01 -6.96556151e-01
2.11591795e-01 8.72268900e-03 -5.11187911e-02 -1.34941816e+00
-3.82527769e-01 -9.71716404e-01 -3.24499995e-01 1.07206619e+00
4.12611693e-01 -3.95654917e-01 1.17811441e+00 1.09056866e+00
5.69664001e-01 -7.87878335e-01 -5.46940088e-01 -4.76513922e-01
-3.45567286e-01 -4.07162666e-01 6.70918643e-01 4.90472913e-01
1.24985747e-01 3.66887212e-01 -5.88935792e-01 2.36542270e-01
5.82793355e-01 7.78276026e-01 4.87988234e-01 -1.24663329e+00
-6.40980184e-01 -2.27515310e-01 1.28207207e-01 -1.75885832e+00
-2.13856846e-01 -6.82602227e-01 1.40837282e-01 -1.34416258e+00
2.61024207e-01 -6.45959198e-01 -4.84126568e-01 1.65910542e-01
1.36952400e-01 -2.86856055e-01 2.57791400e-01 4.05627489e-01
-8.98383379e-01 2.08520457e-01 1.02242351e+00 7.66335949e-02
-9.10488367e-01 1.23270273e+00 -1.13427162e+00 5.96070826e-01
7.57445753e-01 -1.71492502e-01 -6.22571826e-01 4.00550336e-01
6.08664036e-01 7.15414584e-01 -1.21839069e-01 -3.98366451e-01
2.48963460e-01 -4.48426515e-01 1.26549929e-01 -5.60266972e-01
3.65390062e-01 -1.19652295e+00 1.34968653e-01 5.87672234e-01
-6.25491679e-01 -1.87834352e-01 -3.63130599e-01 9.62331951e-01
1.93514451e-02 -4.46576267e-01 5.05257845e-01 -2.16427147e-01
-8.59073222e-01 6.30006433e-01 -7.61216462e-01 -4.27871525e-01
1.20880783e+00 -2.10167825e-01 5.83568692e-01 -1.04639566e+00
-1.23921883e+00 5.62148988e-01 2.04524904e-01 6.34941101e-01
2.74513721e-01 -9.69885230e-01 -2.25148261e-01 1.80741981e-01
-4.62077223e-02 -3.36370707e-01 2.89306045e-01 8.45026612e-01
1.29924372e-01 5.67085981e-01 -3.05082407e-02 -3.53148520e-01
-1.21851003e+00 8.19878578e-01 1.11277692e-01 -7.57270575e-01
-3.83618116e-01 4.18125451e-01 3.65994573e-01 5.08002304e-02
5.05894780e-01 -3.49385470e-01 -6.31114841e-01 3.43169630e-01
6.17416501e-01 2.27036744e-01 -7.51713738e-02 -2.58822411e-01
1.23922497e-01 4.44361269e-02 -6.83980346e-01 4.63340506e-02
1.32540572e+00 -7.16880500e-01 3.88782501e-01 6.38912261e-01
9.57862556e-01 -2.79635817e-01 -1.49510372e+00 -3.86486977e-01
2.60302834e-02 -3.85712266e-01 1.91794455e-01 -7.56330788e-01
-1.01519680e+00 5.31945407e-01 6.09488785e-01 7.34016955e-01
1.07015896e+00 3.37371714e-02 9.48722064e-01 3.84240180e-01
5.92850506e-01 -1.02207029e+00 1.97511166e-02 1.08059213e-01
5.47549069e-01 -1.46517062e+00 -2.51091480e-01 -2.81900734e-01
-1.09499609e+00 7.71699071e-01 2.94725150e-01 8.95236433e-02
9.75857675e-01 -7.46511891e-02 -1.64638877e-01 3.57001983e-02
-7.52336800e-01 -2.56332964e-01 -6.97214007e-02 2.37074360e-01
2.71209061e-01 2.01311871e-01 -1.31350666e-01 8.83723319e-01
-5.34986556e-02 3.21433544e-01 4.54349637e-01 1.00455368e+00
-3.29899937e-01 -1.18994296e+00 1.03406377e-01 6.93440616e-01
-6.63932681e-01 -2.25071739e-02 6.70326799e-02 4.52789217e-01
-1.61238089e-01 1.14864302e+00 1.69972703e-01 -9.51239616e-02
3.24787706e-01 1.85045794e-01 1.42679349e-01 -5.54152429e-01
-5.35633862e-01 3.38042825e-01 -1.12759471e-01 -6.19474351e-01
-3.54411721e-01 -1.01553667e+00 -9.32355821e-01 -3.50225806e-01
-2.71280706e-01 3.80746722e-01 7.65660763e-01 1.07964396e+00
3.02954137e-01 1.88052714e-01 9.43567872e-01 -8.22057545e-01
-8.59914780e-01 -6.52653813e-01 -1.13323915e+00 3.52593690e-01
3.77602220e-01 -6.12898171e-01 -3.97561282e-01 2.52780672e-02] | [4.849436283111572, 3.0987398624420166] |
e5f504f8-cd32-4e8c-85e8-b8a8f9ea87e4 | connecting-a-french-dictionary-from-the | 2206.11022 | null | https://arxiv.org/abs/2206.11022v3 | https://arxiv.org/pdf/2206.11022v3.pdf | Connecting a French Dictionary from the Beginning of the 20th Century to Wikidata | The \textit{Petit Larousse illustr\'e} is a French dictionary first published in 1905. Its division in two main parts on language and on history and geography corresponds to a major milestone in French lexicography as well as a repository of general knowledge from this period. Although the value of many entries from 1905 remains intact, some descriptions now have a dimension that is more historical than contemporary. They are nonetheless significant to analyze and understand cultural representations from this time. A comparison with more recent information or a verification of these entries would require a tedious manual work. In this paper, we describe a new lexical resource, where we connected all the dictionary entries of the history and geography part to current data sources. For this, we linked each of these entries to a wikidata identifier. Using the wikidata links, we can automate more easily the identification, comparison, and verification of historically-situated representations. We give a few examples on how to process wikidata identifiers and we carried out a small analysis of the entities described in the dictionary to outline possible applications. The resource, i.e. the annotation of 20,245 dictionary entries with wikidata links, is available from GitHub url{https://github.com/pnugues/petit_larousse_1905/ | ['Pierre Nugues'] | 2022-06-22 | null | https://aclanthology.org/2022.lrec-1.272 | https://aclanthology.org/2022.lrec-1.272.pdf | lrec-2022-6 | ['general-knowledge'] | ['miscellaneous'] | [-3.54687750e-01 -2.06968077e-02 -3.98617506e-01 -1.40744090e-01
-6.13521814e-01 -9.93564188e-01 1.01358545e+00 5.88755965e-01
-5.56694329e-01 1.00587547e+00 7.44638920e-01 -4.58055049e-01
-3.33010525e-01 -8.17393124e-01 -2.38569990e-01 -2.42619932e-01
8.81510302e-02 5.42171419e-01 8.96304399e-02 -6.38592839e-01
4.36287284e-01 1.34502634e-01 -1.31383598e+00 4.04739305e-02
7.04819202e-01 5.02333701e-01 3.83721083e-01 2.78689116e-01
-3.92093956e-01 4.10148740e-01 -5.76221585e-01 -5.87465107e-01
-5.27890436e-02 -4.11293954e-01 -9.70285773e-01 -5.71429133e-01
3.20814222e-01 4.42436263e-02 -3.50000441e-01 9.56936598e-01
3.75612557e-01 -8.85380954e-02 4.94405299e-01 -6.64601922e-01
-7.06151485e-01 1.12497509e+00 -1.67664424e-01 4.05736268e-01
9.07742500e-01 -5.15750587e-01 1.02411103e+00 -1.01918030e+00
1.30184174e+00 8.15520823e-01 8.00839305e-01 1.84107751e-01
-6.64085448e-01 -5.95325708e-01 1.55514687e-01 1.25747949e-01
-1.72075343e+00 -6.36014521e-01 3.55185509e-01 -7.16494322e-01
8.38829339e-01 2.54572481e-01 9.49069440e-01 9.96656120e-01
-2.20810734e-02 2.95328259e-01 9.07383859e-01 -6.86978281e-01
-2.37902030e-01 4.22364146e-01 4.33214791e-02 6.94798887e-01
5.20311177e-01 -2.14204937e-01 -6.50324881e-01 -2.99623400e-01
8.08876216e-01 -1.41897976e-01 -4.11631823e-01 -6.48636967e-02
-1.38657629e+00 5.81151724e-01 1.56949520e-01 8.73692930e-01
-2.50336411e-03 -7.34897852e-02 7.46811867e-01 2.82368928e-01
4.39386517e-01 4.36120898e-01 -4.65844780e-01 -5.64939022e-01
-9.34373915e-01 1.65931672e-01 1.10303319e+00 1.19385219e+00
4.34073567e-01 -1.75260037e-01 7.58969128e-01 1.09258676e+00
2.96870023e-01 4.28915948e-01 4.21788216e-01 -5.70370078e-01
2.87132204e-01 5.85253060e-01 2.13054046e-01 -1.14711797e+00
-3.64697784e-01 -2.77323663e-01 -3.14750403e-01 -2.85414428e-01
7.70105064e-01 4.47956286e-02 -5.98159373e-01 1.58430636e+00
1.09405823e-01 -5.22672474e-01 -3.93727273e-02 6.22910857e-01
1.25069141e+00 5.21025717e-01 1.04318351e-01 -1.91134930e-01
1.65077829e+00 -3.10106963e-01 -8.20169270e-01 2.39889234e-01
7.86551952e-01 -1.08161724e+00 8.85643721e-01 2.27506131e-01
-1.05164516e+00 -8.32744315e-02 -9.76164758e-01 -2.78471112e-01
-1.06403816e+00 1.37890056e-01 6.98582113e-01 6.61005378e-01
-9.68603134e-01 5.40883899e-01 -7.75861382e-01 -1.17082560e+00
-1.17427878e-01 2.92872218e-03 -4.81346518e-01 4.26429272e-01
-1.57452786e+00 1.21794033e+00 5.99615991e-01 -5.65383434e-02
-3.49333227e-01 -4.00241226e-01 -1.04823220e+00 -4.22433168e-01
5.22760749e-01 -1.89598441e-01 9.88547444e-01 -6.45274162e-01
-8.96264136e-01 1.32963336e+00 -6.22181036e-02 -3.74617912e-02
3.87442946e-01 6.78285807e-02 -1.22096455e+00 5.86481728e-02
5.81705987e-01 2.28110909e-01 1.67728309e-02 -9.39341605e-01
-8.95144045e-01 -1.38379812e-01 1.64512947e-01 1.05714150e-01
-3.47455174e-01 4.71696377e-01 -6.82106853e-01 -1.13199222e+00
1.58478558e-01 -6.09016478e-01 2.67670184e-01 -1.69916466e-01
-5.67368977e-02 -8.97792950e-02 1.25721067e-01 -1.20147157e+00
2.03661442e+00 -2.10456824e+00 1.03100978e-01 2.34478816e-01
1.22364290e-01 -4.80226129e-02 3.41005713e-01 1.13405466e+00
3.74953486e-02 4.75838155e-01 -2.70081609e-01 1.81036130e-01
1.03517301e-01 3.24642360e-01 -2.75645524e-01 6.38134480e-01
-5.95867276e-01 5.64249575e-01 -1.36601222e+00 -4.21903431e-01
1.02934048e-01 5.17973125e-01 -3.04940194e-01 -7.73543656e-01
1.44230023e-01 2.16202140e-01 -2.57442653e-01 8.38963211e-01
1.91777959e-01 1.72795057e-02 6.18365943e-01 -5.37018776e-02
-7.21616089e-01 7.56491005e-01 -1.17454541e+00 2.04043961e+00
-4.10142809e-01 7.54271448e-01 5.29219359e-02 -3.88609141e-01
9.38898325e-01 5.70128322e-01 4.84066874e-01 -3.72880340e-01
1.74573839e-01 9.26969349e-01 -1.03704453e-01 -2.24107727e-01
1.00079095e+00 -1.76654682e-02 -6.10963106e-01 5.10227740e-01
9.61610395e-03 6.27921969e-02 7.07177043e-01 4.19171244e-01
7.56073356e-01 4.71057475e-01 9.89494681e-01 -8.02643061e-01
5.97620368e-01 4.05519038e-01 5.54711878e-01 3.40161383e-01
-3.31817046e-02 2.40461752e-01 1.94351748e-01 -8.24895382e-01
-1.21910703e+00 -1.07585764e+00 -8.96536291e-01 8.41866970e-01
1.00540593e-01 -1.36181879e+00 -4.31292921e-01 -3.40099275e-01
-4.95168231e-02 6.92377329e-01 -7.79357493e-01 4.07410473e-01
-5.30157149e-01 -5.14311314e-01 6.28539324e-01 2.41886109e-01
3.05714309e-01 -9.15569246e-01 -5.80545247e-01 2.07544357e-01
-2.88822651e-01 -6.57519460e-01 -2.24573702e-01 -5.49246408e-02
-2.27879465e-01 -1.04116201e+00 -6.92492962e-01 -8.60126615e-01
5.22490323e-01 -1.78670704e-01 1.09753358e+00 2.85062879e-01
-2.36506641e-01 5.08274794e-01 -6.65378809e-01 -2.45409161e-01
-4.40043658e-01 5.29440403e-01 3.19840819e-01 -7.67090261e-01
6.49333656e-01 -4.83102918e-01 -2.33742639e-01 1.51172787e-01
-7.09482670e-01 -2.00761810e-01 -6.23786151e-02 6.82793260e-01
3.22958469e-01 -3.36178392e-01 5.52025974e-01 -1.03806448e+00
4.77592200e-01 -8.45559657e-01 -4.95238483e-01 2.33716041e-01
-4.60409164e-01 -3.01991403e-01 4.90060866e-01 1.26497939e-01
-7.80072451e-01 -2.77281910e-01 -3.24797213e-01 4.27379042e-01
-5.00720702e-02 1.08633888e+00 1.11450173e-01 2.38408968e-01
6.60818398e-01 2.58265026e-02 -3.28637600e-01 -7.28649795e-01
4.02134389e-01 9.16639507e-01 5.96402466e-01 -8.89192104e-01
5.19681394e-01 3.54345739e-01 -5.81609368e-01 -9.06726658e-01
-4.02035803e-01 -5.03359795e-01 -1.08559060e+00 -3.65357786e-01
6.31645977e-01 -9.90487397e-01 -3.81321132e-01 7.91808367e-02
-8.46604049e-01 -1.73451453e-01 -3.45486313e-01 4.58580643e-01
-2.86433041e-01 9.45126638e-02 -3.43043059e-01 -4.20482546e-01
1.10280290e-01 -7.29542553e-01 5.41557848e-01 -6.51354194e-02
-7.74830639e-01 -1.67358267e+00 3.04345965e-01 -1.00930333e-01
5.49708642e-02 4.80896175e-01 8.88729393e-01 -6.69610381e-01
-1.06465980e-01 -1.95794344e-01 1.06640361e-01 -4.62166876e-01
4.59416896e-01 1.41248748e-01 -4.53295410e-01 -3.19321990e-01
-4.37782884e-01 1.29262120e-01 5.48452556e-01 -1.42767370e-01
4.43472773e-01 -2.65261829e-01 -4.63777930e-01 5.78344524e-01
1.73461711e+00 4.44632322e-01 4.94614959e-01 9.17595088e-01
4.36226815e-01 5.04612446e-01 5.01237452e-01 5.44679224e-01
6.40364408e-01 7.47517347e-01 -1.43159166e-01 2.38895774e-01
-1.96276188e-01 -3.56899291e-01 2.48294309e-01 1.37726331e+00
-4.91151124e-01 -1.63121715e-01 -1.39677656e+00 1.08400548e+00
-1.65654480e+00 -1.02217090e+00 -1.54328495e-01 2.20925474e+00
9.59171414e-01 -1.39696524e-01 2.24127427e-01 -1.71038900e-02
6.77311778e-01 2.26230979e-01 -2.54913028e-02 -4.80740726e-01
-2.79278666e-01 1.70395955e-01 5.57443082e-01 8.23050499e-01
-8.33721995e-01 1.19188225e+00 7.08154821e+00 7.07705081e-01
-8.79568040e-01 1.77257955e-01 -1.23523757e-01 -1.46506324e-01
-4.86926466e-01 3.18589389e-01 -9.41670895e-01 5.53718507e-01
7.60315955e-01 -7.28300512e-01 3.53248745e-01 5.99862158e-01
2.16333382e-02 -3.65804553e-01 -7.67438889e-01 8.40607941e-01
2.70125240e-01 -1.53524232e+00 3.95346768e-02 1.81855977e-01
4.85992223e-01 1.36965197e-02 -1.89851940e-01 7.22842962e-02
4.63659376e-01 -8.21067452e-01 1.20725060e+00 3.39973181e-01
1.29966164e+00 -5.73486090e-01 4.91502404e-01 -2.02113897e-01
-1.55826068e+00 1.32081553e-01 -3.47043604e-01 -2.33616784e-01
5.30196130e-01 3.18627238e-01 -3.69007736e-01 8.56356740e-01
6.82100475e-01 1.12378275e+00 -4.95193601e-01 9.89723921e-01
-4.02306020e-01 3.27496260e-01 -2.74196595e-01 -1.58279330e-01
2.39473253e-01 -3.99949759e-01 8.24693501e-01 1.36938500e+00
5.65319240e-01 2.29539171e-01 5.12662046e-02 4.33193713e-01
-7.21057039e-03 6.69956863e-01 -6.77939415e-01 -2.04240263e-01
9.56118882e-01 1.07763910e+00 -1.09771395e+00 -4.62814331e-01
-5.16286254e-01 7.41290748e-01 2.77608186e-01 7.81731457e-02
-6.18357778e-01 -7.99640238e-01 6.60308838e-01 5.75613856e-01
4.90640476e-02 -5.57251990e-01 -6.33530989e-02 -1.30640531e+00
-1.75996013e-02 -8.16811740e-01 7.41546273e-01 -5.77867806e-01
-8.98145854e-01 6.50401175e-01 3.89188886e-01 -9.59110975e-01
-2.83610165e-01 -5.87795854e-01 -2.68443059e-02 8.19525719e-01
-7.62461901e-01 -1.08892715e+00 6.82219639e-02 4.04262155e-01
3.13737929e-01 -3.01874936e-01 1.10426736e+00 6.37362719e-01
-4.53083724e-01 3.92673910e-01 3.87789071e-01 3.86953115e-01
8.93062532e-01 -1.32649624e+00 4.96275365e-01 6.72581136e-01
2.28937909e-01 1.09731615e+00 6.36201262e-01 -1.14785147e+00
-7.72285998e-01 -4.48754460e-01 1.46993220e+00 -8.27429473e-01
1.25763035e+00 -4.40700322e-01 -6.09769106e-01 1.13282681e+00
6.07026219e-01 -7.64162481e-01 9.13103700e-01 3.00347209e-01
-1.24744140e-01 1.95196003e-01 -8.44474196e-01 8.02169979e-01
1.33785200e+00 -6.77652597e-01 -7.39354074e-01 4.19697285e-01
6.44007772e-02 -4.60722774e-01 -1.44600916e+00 -1.11426264e-01
9.12937462e-01 -5.17534077e-01 6.50680244e-01 -8.42486024e-02
1.01424180e-01 -5.06349683e-01 -7.54789412e-02 -1.13053310e+00
-2.42700979e-01 -7.49534726e-01 5.31907976e-01 1.42785573e+00
7.44461000e-01 -9.10204411e-01 1.56611234e-01 3.20006937e-01
-2.81849891e-01 -2.31802136e-01 -8.23784530e-01 -7.20172226e-01
3.43010604e-01 -3.03481877e-01 6.08690202e-01 1.59391975e+00
7.86564112e-01 -7.69510567e-02 -2.89906472e-01 -3.42128605e-01
2.11342111e-01 -2.61457950e-01 4.82432365e-01 -1.23640108e+00
2.44494289e-01 -5.00746250e-01 -4.76985186e-01 -7.56408155e-01
-1.63638219e-01 -1.38506198e+00 -3.92154992e-01 -1.83124614e+00
-1.09260559e-01 -6.83183849e-01 1.20775253e-01 5.78557849e-01
3.20908844e-01 3.35958093e-01 3.95564064e-02 6.53643608e-01
-2.58725882e-01 -1.16209976e-01 7.80774891e-01 2.99635589e-01
-2.27652252e-01 -7.34080791e-01 -7.13942468e-01 8.55479717e-01
7.79748797e-01 -4.90221709e-01 -1.46165341e-01 -4.87023056e-01
7.67563522e-01 -2.99377620e-01 -1.21026568e-01 -8.70417893e-01
3.37996393e-01 -3.54217365e-02 2.07869858e-01 -6.04641378e-01
2.11127982e-01 -5.93114614e-01 7.55390644e-01 4.69948202e-01
8.30369443e-02 4.18736368e-01 2.80749232e-01 8.52463469e-02
-2.80634880e-01 -4.32790846e-01 3.20537657e-01 -6.30868852e-01
-1.08256388e+00 -4.68933471e-02 -7.52681017e-01 1.24426782e-01
9.02468026e-01 -5.10443270e-01 -3.87034148e-01 -1.62338302e-01
-1.10446346e+00 -2.48311227e-03 1.15690672e+00 4.29390252e-01
1.66320547e-01 -1.46492350e+00 -6.50907159e-01 3.16475742e-02
3.47024918e-01 -4.97865409e-01 -7.97025263e-02 7.24737585e-01
-1.08533061e+00 4.96601909e-01 -5.09308219e-01 -3.16391103e-02
-9.37207460e-01 4.60610151e-01 8.37065876e-02 9.34357643e-02
-8.08661699e-01 3.92856121e-01 -3.82717997e-01 -4.15180832e-01
1.98411080e-03 -1.55625865e-01 -5.64319074e-01 6.68954730e-01
4.47910249e-01 4.10985053e-01 -1.78124413e-01 -1.20976317e+00
-7.09903300e-01 7.94503391e-01 -1.46784559e-01 -4.99308795e-01
1.34290230e+00 -5.64814627e-01 -4.65237886e-01 9.50666785e-01
8.48258317e-01 6.69614494e-01 -2.46804759e-01 -1.23272106e-01
3.00633252e-01 -5.09473741e-01 -3.38233918e-01 -1.00929368e+00
-7.64903665e-01 2.05098107e-01 7.86870047e-02 3.53540540e-01
6.72019243e-01 1.26112625e-01 6.20674670e-01 1.66377470e-01
5.66777289e-01 -1.20320475e+00 -7.71433771e-01 7.48443246e-01
1.08002257e+00 -6.44378424e-01 3.36755991e-01 -4.34385628e-01
-3.93217117e-01 1.19160521e+00 7.35125178e-03 1.46858960e-01
1.00268757e+00 2.78503060e-01 3.54003131e-01 -4.74866182e-01
-3.84315997e-01 -2.11849213e-01 4.02131788e-02 4.07670110e-01
8.84066582e-01 2.74958145e-02 -1.14410424e+00 5.17500639e-01
-8.35139751e-01 -1.48589760e-01 7.46161938e-01 1.12458789e+00
-3.53476912e-01 -1.44287157e+00 -4.07759279e-01 1.91621572e-01
-5.23338139e-01 -4.36279953e-01 -6.04729533e-01 1.30026305e+00
3.45458716e-01 5.90086460e-01 2.78038144e-01 -2.22631633e-01
1.05574064e-01 2.00260699e-01 4.94711578e-01 -6.22371733e-01
-8.97449613e-01 1.64774865e-01 7.47220278e-01 -1.21224247e-01
-6.65593624e-01 -1.13212788e+00 -1.16732800e+00 -8.53246510e-01
-5.65343983e-02 4.20996398e-01 6.92746699e-01 6.63976133e-01
-3.41829434e-02 7.83940181e-02 -8.75228494e-02 -5.75551510e-01
3.40684414e-01 -7.54002810e-01 -9.18102860e-01 1.53197259e-01
-2.48728245e-01 -8.09073746e-01 -3.18055451e-01 1.00591861e-01] | [10.196975708007812, 10.089967727661133] |
9bf99b6e-3587-4b5f-bbca-35fb4e9125f6 | retinex-based-image-denoising-contrast | 2307.02625 | null | https://arxiv.org/abs/2307.02625v1 | https://arxiv.org/pdf/2307.02625v1.pdf | Retinex-based Image Denoising / Contrast Enhancement using Gradient Graph Laplacian Regularizer | Images captured in poorly lit conditions are often corrupted by acquisition noise. Leveraging recent advances in graph-based regularization, we propose a fast Retinex-based restoration scheme that denoises and contrast-enhances an image. Specifically, by Retinex theory we first assume that each image pixel is a multiplication of its reflectance and illumination components. We next assume that the reflectance and illumination components are piecewise constant (PWC) and continuous piecewise planar (PWP) signals, which can be recovered via graph Laplacian regularizer (GLR) and gradient graph Laplacian regularizer (GGLR) respectively. We formulate quadratic objectives regularized by GLR and GGLR, which are minimized alternately until convergence by solving linear systems -- with improved condition numbers via proposed preconditioners -- via conjugate gradient (CG) efficiently. Experimental results show that our algorithm achieves competitive visual image quality while reducing computation complexity noticeably. | ['Xianming Liu', 'Gene Cheung', 'Yeganeh Gharedaghi'] | 2023-07-05 | null | null | null | null | ['image-denoising'] | ['computer-vision'] | [ 7.06344783e-01 -1.43529415e-01 2.09041998e-01 -1.12005793e-01
-6.81844175e-01 -3.74454886e-01 2.31719404e-01 -2.38155156e-01
-2.42998660e-01 7.43346691e-01 1.91996112e-01 -5.87575287e-02
2.04236843e-02 -4.31653172e-01 -9.07858431e-01 -9.35241878e-01
-1.22948643e-02 -3.64852011e-01 -3.89365166e-01 1.32997409e-02
1.56803682e-01 4.90189940e-01 -1.15536916e+00 -1.35623693e-01
1.18496096e+00 7.31517673e-01 9.65739936e-02 8.85836244e-01
2.51988947e-01 1.00222182e+00 -9.78829488e-02 -5.39515279e-02
6.20241225e-01 -5.64616144e-01 -5.84104359e-01 7.24468350e-01
6.55144751e-01 -3.45313698e-01 -6.15337372e-01 1.30551505e+00
3.55303258e-01 4.60367352e-01 6.23684525e-01 -8.77476990e-01
-1.10722637e+00 -1.51926354e-01 -1.34506309e+00 -1.44339055e-01
3.34137052e-01 2.43477240e-01 5.68513751e-01 -1.09291303e+00
5.41479945e-01 1.25973606e+00 6.85493767e-01 3.32103044e-01
-1.74172378e+00 -1.33321166e-01 3.45892519e-01 -9.48212575e-03
-1.45757210e+00 -6.67881846e-01 1.03422618e+00 -1.62898645e-01
7.81000614e-01 4.80653256e-01 6.23255968e-01 4.36409920e-01
2.17424169e-01 5.11642694e-01 1.63455284e+00 -5.27528226e-01
1.03713430e-01 -1.49050862e-01 4.64238226e-02 9.42614377e-01
1.80056661e-01 -7.19652027e-02 -4.45240408e-01 -1.19555511e-01
1.08972788e+00 5.04667498e-02 -9.30311143e-01 -2.22505033e-01
-9.33254004e-01 5.28785408e-01 5.27451396e-01 -5.49418926e-02
-6.70291305e-01 2.77812004e-01 -2.46773645e-01 2.84507781e-01
5.57291090e-01 5.69369383e-02 1.30634770e-01 6.18454814e-01
-8.21466029e-01 -1.57091364e-01 5.85304976e-01 9.09965158e-01
1.32098198e+00 4.12769943e-01 -1.16735540e-01 1.15137708e+00
6.40842199e-01 9.92290676e-01 -1.10095493e-01 -1.21030521e+00
1.79906577e-01 3.22488338e-01 3.01751792e-01 -1.24896801e+00
-1.34286866e-01 -5.65335155e-01 -1.19929254e+00 4.73801762e-01
2.74913430e-01 -4.14193384e-02 -1.07983208e+00 1.56791329e+00
3.30228180e-01 4.33685899e-01 2.26775724e-02 1.05974400e+00
5.03620386e-01 8.96199942e-01 -5.37732281e-02 -8.16874385e-01
1.08471143e+00 -8.67414773e-01 -9.61803555e-01 -3.68497699e-01
-1.11994304e-01 -9.56698954e-01 1.18869698e+00 5.05895138e-01
-1.51799858e+00 -3.71670127e-01 -1.00735140e+00 -3.98624033e-01
2.27571145e-01 2.21211717e-01 1.55344695e-01 5.17545938e-01
-1.44006681e+00 3.29293042e-01 -7.41156995e-01 -3.44664678e-02
2.81182796e-01 2.07963556e-01 -2.27135718e-01 -4.69829112e-01
-4.21751678e-01 5.78878939e-01 -2.97051668e-01 4.87733781e-01
-9.62815285e-01 -7.43356168e-01 -8.68499815e-01 -2.45690867e-01
2.38335371e-01 -7.52220869e-01 4.65777665e-01 -1.14616120e+00
-1.85783374e+00 1.09134316e+00 -4.78581548e-01 -9.13692936e-02
4.22382802e-01 -1.58553734e-01 -3.13776225e-01 3.38868052e-01
-2.19574794e-01 9.58881900e-02 1.43774164e+00 -2.02319884e+00
-1.63829920e-03 -3.70325625e-01 -1.76566169e-01 5.79882979e-01
2.71325884e-03 -4.22246419e-02 -7.86221147e-01 -4.27706391e-01
5.35711706e-01 -7.43659019e-01 -4.03917313e-01 3.52521479e-01
-6.65901721e-01 6.49874091e-01 6.87558949e-01 -1.14555204e+00
1.01839793e+00 -2.16882086e+00 1.52567029e-01 4.49130774e-01
5.14440775e-01 2.54527777e-01 -3.55190903e-01 2.87416816e-01
-1.82566524e-01 -3.04094464e-01 -7.30064929e-01 -5.75988710e-01
-3.58847171e-01 3.04155409e-01 -2.53958881e-01 1.16791821e+00
-7.61896968e-02 9.34979796e-01 -9.53376532e-01 -2.93568730e-01
5.16210079e-01 1.04504728e+00 -4.46632683e-01 2.90300637e-01
1.30483270e-01 6.36239052e-01 -1.88459337e-01 5.10161698e-01
1.26289284e+00 -2.82131016e-01 1.69746310e-01 -7.14829862e-01
-2.49897659e-01 -3.39124739e-01 -1.26916921e+00 1.40458405e+00
-7.17949808e-01 4.50267553e-01 8.16839814e-01 -9.66985047e-01
8.92573178e-01 -1.20395847e-01 4.22791451e-01 -5.49449444e-01
-1.00124717e-01 -1.90275237e-02 -7.05031157e-01 -3.65406692e-01
3.10718894e-01 -2.15824753e-01 7.18725145e-01 2.42366299e-01
-2.42070675e-01 -2.54863173e-01 -1.28272176e-01 3.11221480e-01
8.72165740e-01 1.28334910e-01 2.99878836e-01 -2.43235052e-01
8.16393077e-01 -2.51432478e-01 4.56442744e-01 6.66736424e-01
-7.64714777e-02 8.46085370e-01 1.82159051e-01 -1.05768247e-02
-7.51124084e-01 -1.13589931e+00 3.07000615e-02 7.15967000e-01
4.23389763e-01 -7.30055347e-02 -6.45697713e-01 -7.44752064e-02
-3.65665078e-01 5.80828428e-01 -3.65083128e-01 1.02916338e-01
-5.73952854e-01 -9.11000788e-01 -9.41467658e-03 -4.66356389e-02
7.19151258e-01 -4.41293508e-01 -2.49679416e-01 1.19080372e-01
-2.21540764e-01 -9.90738094e-01 -7.19379842e-01 -9.31516811e-02
-7.27419019e-01 -1.05208290e+00 -1.00356972e+00 -7.54353046e-01
1.30220664e+00 9.33130562e-01 1.13548863e+00 1.97085738e-01
-3.97101313e-01 9.50731158e-01 1.33444369e-01 2.51422316e-01
-3.36936936e-02 -1.00237906e+00 -1.22843340e-01 6.48928642e-01
-5.61996400e-01 -7.25635529e-01 -9.12199676e-01 1.80544138e-01
-9.70441759e-01 2.02255949e-01 3.77003282e-01 8.03317606e-01
1.30864739e+00 -2.72511388e-03 -8.80704895e-02 -1.04695511e+00
5.73568165e-01 -1.20885335e-01 -8.77977252e-01 4.44469929e-01
-6.47122264e-01 -3.15579444e-01 5.77920914e-01 -2.63191342e-01
-1.31921756e+00 1.94234610e-01 2.59195030e-01 -5.24192035e-01
1.47826329e-01 3.41145664e-01 -1.25682622e-01 -8.83598208e-01
6.41352475e-01 4.18617934e-01 3.03452043e-03 -3.59211355e-01
7.47367084e-01 2.75484383e-01 8.95965576e-01 -4.33139056e-01
1.06500745e+00 9.28391993e-01 2.90290147e-01 -1.30371821e+00
-7.28492677e-01 -5.29126823e-01 -1.43503934e-01 -4.90824193e-01
8.29247773e-01 -1.09434593e+00 -8.40522230e-01 5.87980390e-01
-8.13748538e-01 -5.41310966e-01 -4.37937647e-01 3.44408691e-01
-5.62517703e-01 8.43711078e-01 -7.09999442e-01 -1.05547702e+00
-3.17007124e-01 -8.16970766e-01 8.11876655e-01 3.73456329e-01
6.49852157e-01 -1.29152966e+00 1.26784861e-01 3.83122116e-01
6.44194961e-01 4.83693749e-01 4.62349743e-01 5.75074553e-01
-7.65745997e-01 -1.48121826e-03 -6.50635481e-01 6.16805434e-01
1.79433912e-01 -1.31799862e-01 -1.01255620e+00 -6.33073866e-01
4.69378650e-01 -1.02322824e-01 7.72195578e-01 8.65427077e-01
9.75767195e-01 -5.85333407e-01 -7.48106167e-02 1.29818785e+00
2.05404782e+00 -3.07455629e-01 8.04070890e-01 -5.14207296e-02
1.03012526e+00 3.30995560e-01 1.31410539e-01 3.85870755e-01
3.61524552e-01 4.10139292e-01 4.95899349e-01 -7.75428057e-01
-5.82211673e-01 -1.28662273e-01 4.19186264e-01 7.82636106e-01
-4.24858332e-01 -3.11733007e-01 -5.48929513e-01 4.28789943e-01
-1.74535048e+00 -7.03775823e-01 -6.73936188e-01 2.42005229e+00
7.06568956e-01 -6.38735533e-01 -1.84668541e-01 -4.77402583e-02
6.65771008e-01 3.37117970e-01 -5.08446455e-01 -1.42200179e-02
-5.96692860e-01 1.27323672e-01 9.59653258e-01 1.04972935e+00
-8.07907760e-01 6.29648447e-01 6.98131037e+00 4.25624460e-01
-1.02780449e+00 4.19562012e-02 7.61320233e-01 1.81293249e-01
-5.27992487e-01 1.55266136e-01 -1.64838284e-01 7.71814436e-02
4.32528675e-01 -8.06314051e-02 1.15706658e+00 1.84403598e-01
4.65759277e-01 -2.43102536e-01 -5.01384676e-01 1.18311965e+00
2.13357344e-01 -9.27072287e-01 -7.70156011e-02 -8.06433111e-02
1.21834052e+00 -6.31143525e-02 3.36354017e-01 -3.40441287e-01
4.72334266e-01 -8.37403536e-01 2.93112785e-01 8.64799857e-01
8.79561603e-01 -4.04896855e-01 4.22323570e-02 -7.00298995e-02
-1.13325596e+00 2.26646170e-01 -6.20582938e-01 1.30108029e-01
2.53942966e-01 8.94454896e-01 -1.22672111e-01 6.34470046e-01
4.28359568e-01 8.94365668e-01 -1.34879440e-01 9.75436389e-01
-5.43868184e-01 6.34574115e-01 -3.11870098e-01 6.85923278e-01
-7.11775720e-02 -1.11222851e+00 8.29561710e-01 1.08339834e+00
6.18702471e-02 5.22437394e-01 3.32454503e-01 1.00644553e+00
-7.00747892e-02 1.77320212e-01 -2.82468915e-01 4.01714087e-01
-9.82614160e-02 1.50791550e+00 -5.93656957e-01 -1.56983256e-01
-5.90934813e-01 1.29256058e+00 9.57820714e-02 1.31613934e+00
-5.06454110e-01 -1.04672529e-01 5.09362221e-01 7.05204979e-02
7.23766237e-02 -3.88707101e-01 -1.91372469e-01 -1.38139582e+00
5.20358644e-02 -7.93080389e-01 1.26263231e-01 -1.08919823e+00
-1.30721200e+00 4.24736053e-01 -3.25065941e-01 -1.00190401e+00
4.27543163e-01 -3.83135974e-01 -5.72283685e-01 1.26061320e+00
-2.19693828e+00 -1.27526939e+00 -6.44826412e-01 1.15998852e+00
1.23262912e-01 4.32860285e-01 4.40933526e-01 2.01554865e-01
-6.67453527e-01 2.24369153e-01 4.29830104e-01 -3.01486343e-01
5.46974659e-01 -1.20201135e+00 -4.38296422e-02 1.17458391e+00
-2.03870192e-01 6.02873325e-01 6.85483575e-01 -6.51585162e-01
-1.93934214e+00 -1.09096205e+00 5.26225746e-01 3.24680865e-01
4.64769483e-01 5.58482520e-02 -1.09641898e+00 7.69911766e-01
2.96943575e-01 4.72048014e-01 3.07460606e-01 -5.07642984e-01
-3.05029660e-01 -1.59524143e-01 -1.35939527e+00 5.06324828e-01
9.45868075e-01 -6.99877739e-01 4.99124750e-02 7.33358800e-01
2.03219920e-01 -5.82390487e-01 -7.01563179e-01 5.30483797e-02
1.84391007e-01 -9.09107327e-01 1.21855557e+00 -6.45735189e-02
8.25669020e-02 -7.98663557e-01 -2.94630557e-01 -1.29112148e+00
-3.99100840e-01 -1.14480317e+00 -7.63401240e-02 1.06037629e+00
1.01000192e-02 -8.67807150e-01 4.04969394e-01 6.70314133e-01
-9.05308723e-02 -3.20602387e-01 -6.39891326e-01 -6.51393652e-01
-4.37869996e-01 -7.28452876e-02 -5.40306466e-03 9.53770101e-01
-3.07236195e-01 1.07109323e-01 -8.23879361e-01 4.93331909e-01
1.32494998e+00 -2.38152687e-02 6.24306858e-01 -6.86280966e-01
-3.94322723e-01 -1.28043136e-02 -6.93048835e-02 -1.34570122e+00
-2.40158252e-02 -7.72635102e-01 3.33668828e-01 -1.74538946e+00
2.06441090e-01 -1.81387499e-01 -2.37791598e-01 3.98287356e-01
-4.11521524e-01 6.87135637e-01 5.31498827e-02 3.29809308e-01
-3.32975417e-01 5.69172740e-01 1.45530879e+00 -1.29635096e-01
-3.39588940e-01 -1.91212937e-01 -6.58693433e-01 7.28788793e-01
3.08853447e-01 -1.52847484e-01 -4.93538499e-01 -6.89450681e-01
2.61284679e-01 2.31898054e-01 4.12293077e-01 -7.18017101e-01
1.49414435e-01 -2.82337457e-01 3.57940346e-01 -8.89874026e-02
4.66215342e-01 -7.72988677e-01 3.86616915e-01 1.50052547e-01
-1.57848135e-01 -2.80290902e-01 -1.73932258e-02 8.64530981e-01
-1.64035354e-02 1.52902320e-01 1.15628386e+00 1.16658472e-02
-3.33703250e-01 3.97329181e-01 -3.79219234e-01 -6.21282719e-02
6.35725975e-01 -1.86215907e-01 -3.44129503e-01 -7.95351803e-01
-6.20146692e-01 5.94874658e-02 7.94151008e-01 -4.48993415e-01
9.42070723e-01 -1.21520150e+00 -9.90269959e-01 3.24225664e-01
-2.62547284e-01 -3.17807943e-01 4.75214928e-01 1.18262160e+00
-7.45824814e-01 -3.34584475e-01 8.27852339e-02 -5.17984688e-01
-1.17956901e+00 4.76181865e-01 5.09533286e-01 -1.08629383e-01
-8.00773919e-01 1.02204132e+00 4.49999899e-01 -3.90435494e-02
4.49129604e-02 3.09483390e-02 -2.62296852e-02 -5.29191494e-01
5.77000737e-01 6.59040689e-01 -1.94146156e-01 -8.59118700e-01
-2.11119398e-01 1.09821606e+00 2.63081431e-01 -1.33583084e-01
1.41611826e+00 -6.98026240e-01 -5.13136327e-01 -1.88509114e-02
1.24357021e+00 3.96550208e-01 -1.66921461e+00 -3.46673310e-01
-5.98603666e-01 -8.68438363e-01 6.19490087e-01 -6.00735068e-01
-1.42356944e+00 5.57830453e-01 7.34252334e-01 1.23057917e-01
1.66383696e+00 -4.77143258e-01 4.36641514e-01 4.97869514e-02
-5.80120832e-02 -7.80287445e-01 -8.09797198e-02 1.88968450e-01
1.07881510e+00 -9.24463332e-01 5.36995888e-01 -6.87698901e-01
-4.53982145e-01 8.87902439e-01 -1.29931048e-02 -5.03513157e-01
5.92001915e-01 9.77789536e-02 3.15629780e-01 -1.37897953e-01
-2.46175244e-01 -2.52653778e-01 7.02887237e-01 6.99478149e-01
4.56401497e-01 -2.68033929e-02 -2.94112176e-01 -3.08913291e-01
4.98810738e-01 -1.93820849e-01 5.19198060e-01 6.74878895e-01
-3.61629546e-01 -8.54212999e-01 -6.55798733e-01 2.64584988e-01
-1.19790137e-01 -2.67857164e-01 -1.59894243e-01 2.42288306e-01
-2.73696929e-01 1.26817024e+00 -3.48441750e-01 8.92124176e-02
2.43663773e-01 -5.28718531e-01 6.59004569e-01 -2.96390325e-01
-2.44123548e-01 6.10413373e-01 -2.50222594e-01 -7.58364201e-01
-7.25047231e-01 -4.19021606e-01 -1.12939930e+00 -2.49693334e-01
-3.89195323e-01 -1.26545042e-01 7.35259533e-01 6.57001019e-01
1.30154625e-01 3.97640496e-01 9.11521137e-01 -1.04844999e+00
-2.18960181e-01 -3.94164830e-01 -8.18819821e-01 3.94541442e-01
6.57392740e-01 -1.61181390e-01 -6.91976786e-01 4.27961320e-01] | [10.450117111206055, -2.751959800720215] |
148aa2bd-d813-48c1-9462-324f4e7b3b92 | learning-surface-parameterization-for | null | null | https://openreview.net/forum?id=PGGjnBiQ84G | https://openreview.net/pdf?id=PGGjnBiQ84G | Learning Surface Parameterization for Document Image Unwarping | In this paper, we present a novel approach to learn texture mapping for a 3D surface and apply it to document image unwarping. We propose an efficient method to learn surface parameterization by learning a continuous bijective mapping between 3D surface positions and 2D texture-space coordinates. Our surface parameterization network can be conveniently plugged into a differentiable rendering pipeline and trained using multi-view images and rendering loss. Recent work on differentiable rendering techniques for implicit surfaces has shown high-quality 3D scene reconstruction and view synthesis results. However, these methods typically learn the appearance color as a function of the surface points and lack explicit surface parameterization. Thus they do not allow texture map extraction or texture editing. By introducing explicit surface parameterization and learning with a recent differentiable renderer for implicit surfaces, we demonstrate state-of-the-art document-unwarping via texture extraction. We show that our approach can reconstruct high-frequency textures for arbitrary document shapes in both synthetic and real scenarios. We also demonstrate the usefulness of our system by applying it to document texture editing. | ['Dimitris Samaras', 'Zhixin Shu', 'Ke Ma', 'Sagnik Das'] | 2021-09-29 | null | null | null | null | ['3d-scene-reconstruction'] | ['computer-vision'] | [ 7.35431075e-01 7.47028440e-02 4.06961679e-01 -3.59143704e-01
-1.03665876e+00 -8.47068012e-01 8.86756182e-01 -4.08904821e-01
2.41224453e-01 1.82697326e-01 -1.22201197e-01 -1.84974670e-01
1.38932660e-01 -9.44839060e-01 -1.14071023e+00 -4.76735175e-01
1.99098647e-01 7.69490123e-01 1.86491594e-01 -3.87684286e-01
4.38961893e-01 8.29542458e-01 -1.62367010e+00 6.23515904e-01
5.42374909e-01 8.10818017e-01 -5.12542278e-02 1.02816880e+00
-2.85553277e-01 1.50590599e-01 -4.23348159e-01 -1.41200930e-01
4.45635974e-01 -1.52646646e-01 -5.69194198e-01 4.22484726e-01
1.10085964e+00 -6.82262182e-01 1.51179850e-01 5.69865108e-01
3.74749839e-01 -2.05432504e-01 8.96581590e-01 -6.95367098e-01
-7.35435247e-01 -4.22882766e-01 -8.43304396e-01 -6.49763763e-01
4.68680859e-01 -2.91158408e-01 7.45983303e-01 -1.11836064e+00
8.85630012e-01 1.41625273e+00 8.34004521e-01 3.96568716e-01
-1.69949245e+00 -1.45567179e-01 1.05483599e-01 -3.10288817e-01
-1.08445752e+00 -5.74791253e-01 1.11986673e+00 -4.90124762e-01
1.02029204e+00 6.69430256e-01 6.03087604e-01 7.81401753e-01
2.02433869e-01 5.42149961e-01 1.24100220e+00 -8.43700290e-01
-5.52949309e-02 -6.38765246e-02 -3.75212669e-01 1.06215000e+00
-2.91693151e-01 -3.31031159e-02 -5.93488038e-01 -2.25178480e-01
1.65193486e+00 -2.54570186e-01 -2.18307957e-01 -9.00362909e-01
-1.30000484e+00 4.78882670e-01 -6.09005727e-02 -3.76453638e-01
5.09114824e-02 4.48011249e-01 2.01623634e-01 5.68033576e-01
1.06050682e+00 3.72183830e-01 -2.82129794e-01 -9.15008858e-02
-6.71085477e-01 2.33183071e-01 8.98645818e-01 1.07130659e+00
1.02121246e+00 3.67737301e-02 5.46099506e-02 9.99390900e-01
1.10174164e-01 8.30335259e-01 -3.67699742e-01 -1.33473921e+00
2.63709694e-01 2.99886346e-01 4.06100482e-01 -1.08034456e+00
-4.89325784e-02 2.26705343e-01 -5.16090631e-01 8.78598809e-01
2.19661027e-01 2.58826524e-01 -9.98987734e-01 1.09713840e+00
4.71484184e-01 1.11529551e-01 -4.58566427e-01 6.80591464e-01
4.04471993e-01 4.60530996e-01 -7.47221768e-01 2.24561259e-01
1.11940503e+00 -7.66870201e-01 -5.48251033e-01 3.13179761e-01
5.16173899e-01 -1.17241251e+00 1.62622559e+00 6.60282075e-01
-1.35039687e+00 -2.31129304e-01 -1.05276215e+00 -5.78288734e-01
-7.55585805e-02 3.78815234e-01 6.58220351e-01 6.19501531e-01
-1.32895648e+00 6.72932625e-01 -1.02096808e+00 -2.32320473e-01
2.57508576e-01 3.83687288e-01 -2.24893242e-01 1.77334454e-02
-4.79774833e-01 5.34554005e-01 -5.98685741e-01 -6.52303845e-02
-6.43247068e-01 -8.39736521e-01 -5.47507167e-01 -2.07810387e-01
2.28369445e-01 -1.08556151e+00 1.17071784e+00 -8.45870018e-01
-2.30925417e+00 1.24540687e+00 -2.55739152e-01 2.67192304e-01
7.68706143e-01 -4.66741979e-01 -9.44360048e-02 8.67837518e-02
-1.01441965e-01 2.42711142e-01 1.33706927e+00 -1.71022534e+00
-2.04474702e-01 -1.80509403e-01 1.04750395e-01 4.41235960e-01
-5.79595976e-02 -3.94139409e-01 -6.96173549e-01 -7.48173654e-01
3.98617983e-01 -6.04889452e-01 1.61521047e-01 7.62700975e-01
-3.84810835e-01 2.96080858e-01 1.13662469e+00 -8.15065145e-01
5.83775640e-01 -2.14215207e+00 2.38210842e-01 5.51375568e-01
9.19332430e-02 -3.75694335e-01 -2.67037064e-01 3.62685472e-01
1.80052802e-01 5.80856428e-02 -2.65449464e-01 -6.96991682e-01
2.10869536e-01 1.86923310e-01 -5.08973002e-01 4.98754233e-01
2.26515532e-01 6.15605414e-01 -6.17124498e-01 3.08297221e-02
4.01212811e-01 1.08688247e+00 -9.65525270e-01 1.55238435e-01
-4.85307813e-01 5.54427385e-01 -3.67731273e-01 5.84274590e-01
8.63978505e-01 -9.03599262e-02 1.67863309e-01 -4.84515488e-01
-2.58344680e-01 3.55874151e-01 -1.04718888e+00 2.07758498e+00
-9.36762094e-01 6.28092170e-01 4.39194173e-01 -6.18922532e-01
1.10450172e+00 1.58431903e-01 4.85446244e-01 -6.62789881e-01
-2.13367179e-01 3.51020277e-01 -9.04993057e-01 -1.31788090e-01
6.15262091e-01 -5.99741898e-02 2.46655986e-01 7.44565129e-01
-4.26722288e-01 -1.03438616e+00 -4.70431030e-01 -1.53608233e-01
8.61882508e-01 9.11768079e-01 -3.84067953e-01 -4.10536975e-01
4.22982424e-01 -1.63582548e-01 -1.73448965e-01 5.31754673e-01
8.46404672e-01 1.02828884e+00 5.91675520e-01 -6.80887759e-01
-1.65437531e+00 -1.25952208e+00 -2.83824950e-01 1.00661623e+00
-2.50878818e-02 -3.26205134e-01 -8.81564736e-01 -2.20993713e-01
1.09244652e-01 5.45851350e-01 -6.72976017e-01 3.65168959e-01
-9.36867654e-01 -3.79613608e-01 2.40911543e-01 1.98283941e-01
2.57346720e-01 -5.20435870e-01 -4.62604195e-01 6.97083399e-02
1.25968456e-01 -8.40643764e-01 -6.14527524e-01 -1.48494646e-01
-9.73314345e-01 -1.00052166e+00 -7.47111440e-01 -7.54814386e-01
9.97912347e-01 3.64856750e-01 1.27903128e+00 1.09632559e-01
-3.15447330e-01 9.78075027e-01 1.36198565e-01 -1.37408689e-01
-6.44911766e-01 2.41055842e-02 -2.98695743e-01 2.35849291e-01
-2.54348695e-01 -8.90190423e-01 -6.43596113e-01 4.75683451e-01
-8.54283392e-01 6.94009602e-01 -2.40895823e-02 6.83064997e-01
1.08996403e+00 -4.52139437e-01 -8.65733176e-02 -1.28085351e+00
5.58138013e-01 3.71502280e-01 -9.96348858e-01 2.64397830e-01
-3.02884936e-01 2.64145166e-01 4.15218115e-01 -2.39660040e-01
-1.26937032e+00 -6.15955051e-03 -7.83109143e-02 -3.88311267e-01
5.83432950e-02 8.72737989e-02 -7.10754469e-02 -4.41311449e-01
7.02560842e-01 6.68713376e-02 2.19803154e-01 -8.02279711e-01
6.16325080e-01 2.76617616e-01 4.92992729e-01 -1.05714560e+00
9.25364912e-01 1.15151668e+00 2.74093777e-01 -1.03889000e+00
-5.01086950e-01 2.43678577e-02 -7.95759261e-01 -1.91361308e-01
6.28785491e-01 -7.36613572e-01 -8.61688673e-01 6.37173772e-01
-1.26114762e+00 -8.80213737e-01 -3.58285815e-01 -1.01493828e-01
-1.08849871e+00 4.26275879e-01 -4.32905674e-01 -4.58242953e-01
-2.15351671e-01 -1.00276458e+00 1.97522366e+00 -6.01368189e-01
-8.36280063e-02 -1.09533525e+00 1.44253746e-01 2.21685935e-02
6.16173625e-01 4.52474743e-01 1.19594407e+00 7.55700052e-01
-1.11123478e+00 3.31779867e-02 -3.53980809e-01 5.99741936e-02
4.84005213e-01 2.54385918e-01 -1.26552951e+00 -2.45387718e-01
1.55791007e-02 -7.65843317e-02 6.52855814e-01 3.49917620e-01
1.42025506e+00 -2.22529143e-01 -1.75903231e-01 1.31396627e+00
1.44775081e+00 -2.74730802e-01 6.92439497e-01 4.66757059e-01
1.15032709e+00 5.31318963e-01 1.93634346e-01 4.05562013e-01
3.07870507e-01 1.04725814e+00 2.06079826e-01 -3.64393353e-01
-3.20310473e-01 -2.23407254e-01 2.27209702e-01 9.28449869e-01
-5.21536648e-01 -1.27525032e-01 -7.48782516e-01 4.15760502e-02
-1.51089823e+00 -5.42051733e-01 -2.32939556e-01 2.37313843e+00
7.79488981e-01 -1.27779290e-01 -1.78470135e-01 -8.86642188e-02
4.64094400e-01 1.19766705e-01 -4.85385925e-01 -7.06535101e-01
-2.13175461e-01 4.29013044e-01 5.60158730e-01 1.05514336e+00
-9.25678372e-01 9.46490526e-01 6.72300243e+00 6.18896365e-01
-1.24184680e+00 3.56547423e-02 4.70373422e-01 -1.43621713e-02
-1.14874613e+00 -8.83766785e-02 -3.91472429e-01 -2.40099236e-01
3.32830101e-01 2.31031850e-01 8.68522882e-01 3.62460166e-01
1.79210320e-01 1.26457050e-01 -1.19277084e+00 1.10627592e+00
5.82072772e-02 -1.67207170e+00 2.92022109e-01 -9.75416973e-03
7.76250839e-01 -1.66071832e-01 2.87266463e-01 -4.01793510e-01
1.96309581e-01 -9.38499153e-01 7.58656919e-01 7.29919136e-01
1.71124184e+00 -5.07416606e-01 -2.64525384e-01 4.32326905e-02
-1.09088838e+00 6.85735583e-01 -2.10472107e-01 2.80380487e-01
1.54089481e-01 4.92327452e-01 -6.77605331e-01 4.08129632e-01
4.92636025e-01 7.96913087e-01 -2.84958541e-01 4.81815785e-01
-1.43720314e-01 2.90165395e-01 -6.54281914e-01 3.46669048e-01
-1.56184167e-01 -5.55977702e-01 5.53089678e-01 1.02542531e+00
5.41613162e-01 -1.18636996e-01 -1.22023411e-01 1.07717586e+00
4.05923463e-02 2.90694982e-02 -6.03906631e-01 2.71061957e-01
3.20930406e-02 1.07590830e+00 -7.17153132e-01 -1.48941129e-01
-2.87724525e-01 1.44192648e+00 3.29622924e-01 6.82824910e-01
-5.10632157e-01 -2.39830390e-01 5.56447983e-01 5.61565101e-01
3.00185680e-01 -6.81633770e-01 -5.78233957e-01 -1.28323138e+00
1.83350414e-01 -6.41981483e-01 -4.79523629e-01 -9.82813537e-01
-1.26857173e+00 4.15276557e-01 -1.50616184e-01 -1.11814427e+00
-5.01472596e-03 -9.27113652e-01 -3.07416856e-01 1.26646161e+00
-1.49431598e+00 -1.37924457e+00 -2.95214921e-01 5.46294451e-01
3.38115722e-01 2.38948734e-03 1.11332655e+00 1.10200599e-01
2.79379517e-01 4.53074783e-01 4.10916537e-01 -3.85253072e-01
7.59929001e-01 -1.32374275e+00 1.12144840e+00 3.03634703e-01
-1.59516975e-01 4.35433596e-01 3.95101607e-01 -5.54444313e-01
-1.93478417e+00 -7.87158847e-01 3.00557822e-01 -7.81573117e-01
2.74103552e-01 -9.56745505e-01 -9.63523805e-01 7.65198290e-01
-7.72589934e-04 -1.39816508e-01 2.59042501e-01 8.76359418e-02
-4.48001742e-01 1.39627874e-01 -1.09423685e+00 8.10635924e-01
1.29968667e+00 -9.27734911e-01 -2.09325217e-02 4.45983201e-01
4.14641708e-01 -1.03122139e+00 -8.31154108e-01 2.16070756e-01
8.89183521e-01 -9.85543966e-01 1.24391496e+00 -1.54076323e-01
3.71054620e-01 -3.34470063e-01 -1.05685800e-01 -1.22701836e+00
-3.64544243e-01 -1.01550806e+00 -6.65105656e-02 8.08887184e-01
1.96217895e-01 -7.70469427e-01 9.14702713e-01 4.53839570e-01
-2.99179524e-01 -6.16710007e-01 -7.31477559e-01 -3.94101053e-01
2.58677937e-02 -3.04087579e-01 6.57381415e-01 9.79231060e-01
-6.80911005e-01 -1.28737548e-02 -3.74636590e-01 1.72639906e-01
8.00385475e-01 3.96427482e-01 1.10757494e+00 -1.05146921e+00
-5.15804529e-01 -3.41829985e-01 -2.08201315e-02 -1.52741671e+00
-5.19559272e-02 -8.31015587e-01 -9.84379128e-02 -1.55108166e+00
-1.52593926e-01 -8.60387266e-01 4.58864838e-01 1.60707876e-01
2.48203248e-01 3.66116405e-01 -2.41271332e-01 1.69205904e-01
1.35964438e-01 4.79214460e-01 1.72369015e+00 -8.08330625e-03
-2.44646579e-01 -2.10417897e-01 -2.52407402e-01 7.58702576e-01
5.80463111e-01 -4.59554829e-02 -5.89583516e-01 -1.05733025e+00
6.06922388e-01 -2.45103203e-02 6.29690468e-01 -4.22166079e-01
-2.91974455e-01 -8.08916017e-02 3.39359164e-01 -6.54987216e-01
6.92098618e-01 -7.64210105e-01 4.18698728e-01 -2.20388442e-01
-1.94774225e-01 6.25925884e-02 3.22701305e-01 4.92852598e-01
2.06982076e-01 1.37464315e-01 5.28280079e-01 -3.63606773e-02
-2.59099245e-01 3.45931411e-01 -1.15249023e-01 -7.69349337e-02
3.59989971e-01 -5.03367543e-01 -1.93802059e-01 -3.51235002e-01
-7.17347801e-01 -3.42790067e-01 1.18083000e+00 2.23463431e-01
7.32228816e-01 -1.45901072e+00 -7.01181650e-01 8.92424822e-01
-1.90002784e-01 2.94526994e-01 6.80792630e-02 4.11657304e-01
-1.28021371e+00 -1.40998969e-02 -9.52634886e-02 -9.85040307e-01
-1.06750655e+00 -5.85647374e-02 5.70539355e-01 5.91419674e-02
-1.00751245e+00 6.88877344e-01 7.15069532e-01 -7.12163925e-01
-7.60295242e-02 -5.18495798e-01 4.86956388e-01 -4.57385540e-01
1.76185817e-01 1.58425778e-01 3.64309669e-01 -3.01644772e-01
7.97865093e-02 1.27342486e+00 9.00171846e-02 -5.02281427e-01
1.59405124e+00 -8.14866498e-02 -2.33850390e-01 4.89947200e-01
1.38539898e+00 4.56576914e-01 -1.71152425e+00 2.94549614e-02
-5.04976809e-01 -7.99478233e-01 5.93300946e-02 -5.67367792e-01
-7.78050423e-01 1.07375264e+00 4.56689984e-01 1.47533834e-01
8.96774709e-01 -2.83469528e-01 6.23167455e-01 4.50068355e-01
5.29435694e-01 -1.11824763e+00 1.70033097e-01 6.89723134e-01
1.32936561e+00 -7.20197201e-01 7.47776702e-02 -8.39635611e-01
-1.08231694e-01 1.46657038e+00 8.52576271e-02 -4.92064416e-01
7.67207861e-01 6.17548525e-01 2.16254279e-01 -3.79412204e-01
-5.15173852e-01 5.53302646e-01 5.22080064e-01 6.17244184e-01
6.26717269e-01 4.26102132e-02 2.82651454e-01 -3.98925424e-01
-2.32199609e-01 -1.83673590e-01 4.08830523e-01 9.02600408e-01
-3.17738280e-02 -1.47077012e+00 -4.84448224e-01 3.56483608e-01
-1.52669504e-01 -1.05283856e-01 -3.13645780e-01 6.22427821e-01
-3.95845145e-01 2.60059852e-02 2.78779209e-01 2.77270842e-02
6.07822239e-01 -5.38357627e-03 1.26343250e+00 -6.93517029e-01
-1.49423257e-01 4.95654106e-01 2.70047307e-01 -7.17954576e-01
-3.68118346e-01 -7.05796480e-01 -8.89564097e-01 -3.16996455e-01
-2.04041228e-01 -4.84047979e-01 7.98582852e-01 3.79740089e-01
5.86496949e-01 4.37059551e-01 7.05854297e-01 -1.39931214e+00
-1.70422629e-01 -3.11381072e-01 -6.49782538e-01 4.25820231e-01
4.50486571e-01 -6.84403419e-01 -3.07498783e-01 4.39745069e-01] | [9.270092010498047, -3.254979133605957] |
a6327c21-7958-44d8-aac0-90ec45b9af80 | unsupervised-learning-of-accurate-siamese | 2204.01475 | null | https://arxiv.org/abs/2204.01475v1 | https://arxiv.org/pdf/2204.01475v1.pdf | Unsupervised Learning of Accurate Siamese Tracking | Unsupervised learning has been popular in various computer vision tasks, including visual object tracking. However, prior unsupervised tracking approaches rely heavily on spatial supervision from template-search pairs and are still unable to track objects with strong variation over a long time span. As unlimited self-supervision signals can be obtained by tracking a video along a cycle in time, we investigate evolving a Siamese tracker by tracking videos forward-backward. We present a novel unsupervised tracking framework, in which we can learn temporal correspondence both on the classification branch and regression branch. Specifically, to propagate reliable template feature in the forward propagation process so that the tracker can be trained in the cycle, we first propose a consistency propagation transformation. We then identify an ill-posed penalty problem in conventional cycle training in backward propagation process. Thus, a differentiable region mask is proposed to select features as well as to implicitly penalize tracking errors on intermediate frames. Moreover, since noisy labels may degrade training, we propose a mask-guided loss reweighting strategy to assign dynamic weights based on the quality of pseudo labels. In extensive experiments, our tracker outperforms preceding unsupervised methods by a substantial margin, performing on par with supervised methods on large-scale datasets such as TrackingNet and LaSOT. Code is available at https://github.com/FlorinShum/ULAST. | ['Wanli Ouyang', 'Wei Wu', 'Weihao Gan', 'Weitao Feng', 'Bo Li', 'Xin Li', 'Peixia Li', 'Jinyang Guo', 'Lei Qiao', 'Qiuhong Shen'] | 2022-04-04 | null | http://openaccess.thecvf.com//content/CVPR2022/html/Shen_Unsupervised_Learning_of_Accurate_Siamese_Tracking_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Shen_Unsupervised_Learning_of_Accurate_Siamese_Tracking_CVPR_2022_paper.pdf | cvpr-2022-1 | ['visual-object-tracking'] | ['computer-vision'] | [ 6.15603896e-03 -3.09072077e-01 -5.33728898e-01 -2.59539187e-01
-5.55658400e-01 -6.45535648e-01 4.85132456e-01 -3.22504222e-01
-5.89578211e-01 7.96701670e-01 -1.01820841e-01 2.08461061e-01
-6.25605360e-02 -2.12933481e-01 -7.75298476e-01 -9.29974198e-01
-1.59879643e-02 2.73503691e-01 5.55902123e-01 3.44098002e-01
-4.73564444e-03 4.59930211e-01 -1.23593593e+00 -4.64868307e-01
8.75919878e-01 9.08558369e-01 2.54871339e-01 4.31597084e-01
2.02643976e-01 7.78825760e-01 -2.58015782e-01 -3.23199540e-01
4.08859283e-01 -3.99120539e-01 -3.50778520e-01 5.44968188e-01
7.92948306e-01 -1.92117244e-01 -5.85947096e-01 1.38798654e+00
1.89254239e-01 3.47738653e-01 5.20298421e-01 -1.39201701e+00
-3.22192907e-01 4.14745033e-01 -7.11667359e-01 2.79536843e-01
-3.74440402e-02 5.79537511e-01 9.16204929e-01 -1.12275279e+00
8.49320650e-01 9.11947608e-01 8.80460024e-01 7.07066476e-01
-1.30056119e+00 -8.68172288e-01 5.02377152e-01 1.08955711e-01
-1.29114556e+00 -5.61344087e-01 9.51270461e-01 -7.89245963e-01
2.51644760e-01 -1.02724455e-01 7.52365887e-01 1.08650279e+00
5.31989783e-02 8.79765451e-01 7.31824696e-01 -9.77504477e-02
-8.38605836e-02 -4.95783612e-02 -1.09430760e-01 1.00690246e+00
2.71309078e-01 6.97793901e-01 -6.92032337e-01 5.23774810e-02
7.93476462e-01 9.74292308e-02 -1.69441104e-01 -8.25635135e-01
-1.49830985e+00 6.68833971e-01 6.16813242e-01 1.08158670e-01
-2.55389601e-01 4.34699237e-01 2.20845893e-01 3.59182239e-01
4.80230957e-01 1.64881438e-01 -3.74143481e-01 1.38631174e-02
-1.20712113e+00 8.84118676e-02 3.59320104e-01 1.10873175e+00
7.14544773e-01 3.00620943e-01 -4.78312910e-01 4.82670069e-01
7.01096296e-01 6.60823286e-01 1.40710235e-01 -1.08729839e+00
3.49079132e-01 3.12428743e-01 4.04782593e-01 -8.72582853e-01
-2.42497340e-01 -8.11827719e-01 -7.20956266e-01 3.46646905e-01
7.60778844e-01 -2.01768875e-01 -9.00113344e-01 1.89051032e+00
7.26582050e-01 5.17035425e-01 -3.58261496e-01 1.15876567e+00
3.71965319e-01 2.36259654e-01 1.12642378e-01 -4.53938782e-01
1.04725254e+00 -1.20302749e+00 -7.46980488e-01 -7.84270912e-02
5.04023969e-01 -6.91110671e-01 5.63712060e-01 1.82422012e-01
-9.85550880e-01 -9.06431794e-01 -9.07775283e-01 3.40589553e-01
2.26513995e-03 3.73339504e-01 5.02140105e-01 4.15228039e-01
-7.72451699e-01 6.72244132e-01 -1.33280098e+00 -3.00238967e-01
6.85147643e-01 3.13613385e-01 -2.19868839e-01 1.88263476e-01
-8.12099814e-01 5.89167535e-01 3.48452508e-01 3.83224905e-01
-1.09825695e+00 -6.37600005e-01 -9.48014379e-01 -2.86898732e-01
6.01459622e-01 -5.77146947e-01 9.73271847e-01 -9.60623085e-01
-1.50478435e+00 6.55988514e-01 -1.79644868e-01 -6.01843655e-01
8.81987512e-01 -3.55310649e-01 -3.67597371e-01 9.72107723e-02
2.57799685e-01 8.33468080e-01 1.23711658e+00 -1.09938717e+00
-7.65263259e-01 -1.63804099e-01 -1.41271591e-01 1.02229126e-01
-3.08443725e-01 2.52549462e-02 -7.44373322e-01 -9.79524672e-01
8.83520022e-02 -1.25251210e+00 -2.72294074e-01 6.77394331e-01
-2.31197610e-01 -1.89143494e-01 1.00939989e+00 -3.88246983e-01
1.03800225e+00 -2.19087458e+00 1.64475694e-01 1.72666430e-01
3.73698443e-01 2.27442130e-01 2.02854071e-02 -1.71739962e-02
2.30413511e-01 -4.43309873e-01 -3.82082790e-01 -5.55058241e-01
-2.04934105e-02 2.25439057e-01 -2.10760519e-01 1.00876701e+00
3.14700574e-01 9.41168189e-01 -1.22030056e+00 -8.19992244e-01
1.79935113e-01 4.43882495e-01 -4.89304274e-01 1.12271130e-01
-3.37461323e-01 1.06669819e+00 -6.26193523e-01 7.36975193e-01
5.00138879e-01 -5.97974122e-01 -1.15441494e-01 -1.87766179e-01
-3.43419045e-01 -5.48973531e-02 -1.09262002e+00 1.94870079e+00
7.89428055e-02 7.90349960e-01 3.41761671e-02 -9.64467466e-01
7.91006029e-01 2.03172952e-01 8.00638258e-01 -2.82722443e-01
1.10205263e-01 2.19890550e-02 6.17326945e-02 -2.02071995e-01
3.11642081e-01 5.21549061e-02 2.06791312e-01 2.16310307e-01
2.07441568e-01 3.67222279e-01 4.28318202e-01 1.56328619e-01
9.12856996e-01 7.26694643e-01 -1.28971681e-01 -1.74106449e-01
6.94272637e-01 1.58506379e-01 1.02658010e+00 7.61958718e-01
-6.38068497e-01 4.76077527e-01 -6.45017996e-02 -1.81475893e-01
-7.22943604e-01 -9.57420111e-01 -1.60673037e-01 1.09021175e+00
4.06473577e-01 -8.74651968e-02 -4.40362811e-01 -9.33839381e-01
1.23222150e-01 3.04429550e-02 -5.55513322e-01 -1.07751846e-01
-6.81623161e-01 -2.96328723e-01 4.32614475e-01 7.32597768e-01
4.55371082e-01 -1.10777831e+00 -5.89924455e-01 2.65909493e-01
1.52133897e-01 -1.21092677e+00 -1.11975241e+00 1.92024529e-01
-9.55624700e-01 -1.09627783e+00 -7.88551927e-01 -7.79251277e-01
9.13950861e-01 2.02799052e-01 8.09278011e-01 2.07716167e-01
-2.05961332e-01 5.48185647e-01 -8.42831656e-02 -1.93112697e-02
-3.85030103e-03 5.09421900e-02 3.74088287e-01 2.79661357e-01
1.97570249e-01 -3.88538897e-01 -6.85451090e-01 6.03779852e-01
-4.16715592e-01 -9.98951420e-02 4.01835233e-01 1.00187695e+00
7.21289217e-01 -3.07997912e-01 3.70818287e-01 -6.96998239e-01
-9.56886560e-02 -2.31188819e-01 -1.22411168e+00 1.46237269e-01
-5.96145272e-01 6.48929477e-02 4.56380546e-01 -7.67008245e-01
-9.67961371e-01 5.35167515e-01 2.11317748e-01 -1.13213372e+00
1.13047518e-01 1.97523206e-01 2.73805615e-02 -4.45967764e-01
4.26031977e-01 3.19039106e-01 2.82899290e-02 -2.92501599e-01
3.64352763e-01 -1.32472515e-01 6.86971247e-01 -5.37309110e-01
1.53244507e+00 6.27800167e-01 -9.76140276e-02 -3.85444552e-01
-9.67942953e-01 -7.50071347e-01 -8.88230979e-01 -7.07184911e-01
8.27502728e-01 -9.00871933e-01 -7.29482710e-01 3.30331266e-01
-8.63918245e-01 -4.27678585e-01 -4.43314105e-01 8.49297166e-01
-5.31616449e-01 4.17118371e-01 -4.38352674e-01 -7.79541790e-01
-1.08214781e-01 -9.60068583e-01 8.80770326e-01 4.37902153e-01
7.77433366e-02 -1.10367703e+00 1.63520128e-01 3.55356634e-02
1.94225505e-01 2.63442039e-01 -2.49638677e-01 -3.48583549e-01
-9.31126475e-01 -4.91935648e-02 -2.43818194e-01 1.83685198e-01
1.66607574e-01 -1.44963162e-02 -6.62206888e-01 -6.87457085e-01
-9.63323563e-02 -3.78292233e-01 1.03753805e+00 4.87394750e-01
9.87424135e-01 -9.66650695e-02 -5.72487831e-01 9.93028820e-01
1.12494481e+00 -6.96510426e-04 9.41456854e-02 2.49569952e-01
9.67944443e-01 3.59672189e-01 1.07686841e+00 1.98576316e-01
1.15489304e-01 7.55813658e-01 1.93139032e-01 8.54402632e-02
-3.55603397e-01 -3.98382813e-01 5.77804327e-01 8.10396850e-01
-9.13185105e-02 1.50084734e-01 -6.42830729e-01 5.96080542e-01
-2.05937958e+00 -1.00945997e+00 5.00960983e-02 2.31463218e+00
8.82628858e-01 4.23267424e-01 2.55884588e-01 -2.95852840e-01
8.65948737e-01 3.23237032e-01 -8.58033180e-01 7.72545457e-01
5.20103015e-02 -2.57549942e-01 7.78161764e-01 5.10196567e-01
-1.41961706e+00 1.09696627e+00 5.18840933e+00 5.84271550e-01
-1.18796206e+00 3.01671535e-01 1.02538213e-01 -1.86619669e-01
1.05248079e-01 1.64606437e-01 -1.08756196e+00 7.47981250e-01
3.95881861e-01 -1.30587980e-01 1.57330111e-01 7.53785789e-01
3.31302375e-01 1.84339225e-01 -1.13225317e+00 9.61594701e-01
-1.00998186e-01 -1.22754025e+00 -4.18167204e-01 3.02594528e-03
8.70073020e-01 2.08267018e-01 1.17838338e-01 2.43145168e-01
3.94298851e-01 -4.88654256e-01 7.47534692e-01 6.75451636e-01
6.51520491e-01 -2.90877402e-01 1.71928763e-01 3.17186326e-01
-1.64243829e+00 1.34634018e-01 -3.12752783e-01 3.94737780e-01
3.07841390e-01 2.95075685e-01 -3.04640859e-01 4.26847160e-01
6.72702610e-01 1.28535414e+00 -5.09231865e-01 1.55721343e+00
-3.61720622e-01 7.30179012e-01 -4.25710857e-01 2.70874560e-01
3.17835599e-01 -2.95116574e-01 8.71180415e-01 1.12283576e+00
1.60698593e-01 -1.06285170e-01 6.63738370e-01 8.24773908e-01
-1.11718729e-01 -2.14009568e-01 -4.03566331e-01 6.09135628e-02
4.21644717e-01 1.31278110e+00 -9.08369839e-01 -2.62336552e-01
-3.97966415e-01 8.43917549e-01 3.15604389e-01 5.42078197e-01
-1.24410188e+00 -6.25312328e-02 4.30509120e-01 2.96185748e-03
7.08300173e-01 -4.36875015e-01 5.20782173e-02 -1.40746295e+00
1.37627110e-01 -3.67434055e-01 4.23875183e-01 -4.63597327e-01
-1.24556935e+00 3.92560750e-01 -1.85087308e-01 -1.87967587e+00
-6.96803182e-02 -2.61266977e-01 -6.13080800e-01 4.87854123e-01
-1.73447013e+00 -1.13869870e+00 -3.57503474e-01 7.64023483e-01
6.40590966e-01 -1.26062706e-01 2.96058077e-02 5.71444452e-01
-6.41595781e-01 7.63992012e-01 1.55607745e-01 4.28626537e-01
9.45352674e-01 -1.12767482e+00 1.07124142e-01 1.07654583e+00
4.10991132e-01 6.81254089e-01 6.66407526e-01 -8.06122959e-01
-1.32280242e+00 -1.38695049e+00 5.53297341e-01 -4.48636472e-01
1.00453103e+00 -3.20984095e-01 -1.05033612e+00 8.25638831e-01
-1.15141785e-02 7.03726530e-01 2.11480990e-01 -1.25007913e-01
-2.40874559e-01 -1.57966599e-01 -7.40087211e-01 4.81886387e-01
1.42060173e+00 -3.28829616e-01 -3.66889745e-01 4.47012275e-01
5.06757021e-01 -5.46488225e-01 -8.11798394e-01 3.59641492e-01
5.42988539e-01 -4.05931681e-01 8.41126800e-01 -4.84252691e-01
-1.02069333e-01 -8.56182992e-01 3.62375259e-01 -9.19893801e-01
-3.91001672e-01 -9.89267290e-01 -6.45716548e-01 1.27867031e+00
2.81369001e-01 -5.33076286e-01 1.08714175e+00 2.21075609e-01
-8.47097710e-02 -4.93611187e-01 -8.86599422e-01 -1.23258555e+00
-3.24357837e-01 -3.11491162e-01 -1.06707714e-01 9.13308680e-01
-4.77436513e-01 1.13317044e-02 -6.25410676e-01 3.21987867e-01
1.43264055e+00 1.81729257e-01 8.50843906e-01 -1.26820004e+00
-2.99328625e-01 -4.01965767e-01 -3.64990354e-01 -1.58612657e+00
3.11856836e-01 -8.77426147e-01 5.04496276e-01 -9.49112117e-01
1.40436605e-01 -6.44084930e-01 -5.79232752e-01 3.81252289e-01
-2.53003627e-01 4.27855790e-01 3.43645066e-01 7.36176312e-01
-1.14541948e+00 8.18626881e-01 1.30676723e+00 -3.04930627e-01
-8.28408152e-02 1.57221586e-01 -1.21496737e-01 7.71903038e-01
5.35829008e-01 -9.41509485e-01 -2.08863273e-01 -3.13893646e-01
-2.62569994e-01 2.46163923e-02 5.98433733e-01 -1.13227999e+00
7.97067046e-01 -1.45745501e-01 4.62058663e-01 -7.97050118e-01
1.84979588e-01 -9.67629194e-01 1.08484656e-01 5.68399072e-01
-3.69091451e-01 -1.34923100e-01 -4.17647287e-02 9.18444693e-01
-2.13601336e-01 -1.17228337e-01 8.80776048e-01 2.37748563e-01
-7.39905298e-01 8.08891833e-01 -5.26148416e-02 1.40236273e-01
9.97758269e-01 -3.98061752e-01 -1.12827476e-02 -1.41455866e-02
-9.04829383e-01 6.82687104e-01 5.77026010e-01 5.44469774e-01
4.30943310e-01 -1.56936240e+00 -5.35723269e-01 8.21037889e-02
1.73923850e-01 -8.28449801e-02 -8.61080363e-03 1.38168991e+00
-1.62830338e-01 2.51697540e-01 -1.10503323e-01 -1.07115710e+00
-1.13779199e+00 5.69507718e-01 3.59974504e-01 -1.46507889e-01
-9.90672350e-01 8.94212246e-01 3.25297087e-01 -5.73821105e-02
7.07592010e-01 -1.22575007e-01 -9.08821747e-02 -2.83997040e-02
4.14462835e-01 4.85710576e-02 -4.78248984e-01 -7.98845828e-01
-4.39722359e-01 8.20282578e-01 -1.62341297e-01 7.52423778e-02
1.02305722e+00 -2.50368178e-01 3.74103785e-01 4.25315142e-01
9.05570507e-01 -9.11188424e-02 -2.19745350e+00 -6.26948416e-01
3.32630008e-01 -6.76322639e-01 -2.40649860e-02 -2.93699533e-01
-1.54468644e+00 4.11705047e-01 7.80320168e-01 -4.43478934e-02
8.96439075e-01 -5.17701730e-02 6.68998361e-01 3.23189497e-01
2.92522192e-01 -1.11181343e+00 1.95366696e-01 4.61663753e-01
5.18367290e-01 -1.53089225e+00 6.54739812e-02 -3.10120434e-01
-5.19313633e-01 8.00462544e-01 7.79038429e-01 -4.00775909e-01
6.80774450e-01 2.46538252e-01 2.52495538e-02 -2.49971539e-01
-6.76288605e-01 -5.04287004e-01 5.22198141e-01 5.11445284e-01
2.46744618e-01 -3.49355519e-01 -2.28687838e-01 2.08887067e-02
2.39843190e-01 2.15225711e-01 -1.63312480e-02 9.85460579e-01
-3.34632486e-01 -8.78876567e-01 -4.36138809e-01 2.18449712e-01
-3.19295079e-01 1.81002587e-01 2.13399734e-02 7.66675413e-01
4.88507785e-02 6.09908223e-01 -1.86177175e-02 -1.06288433e-01
1.60564318e-01 -2.46935099e-01 5.78304708e-01 -4.70112622e-01
-4.71305788e-01 4.32431370e-01 -2.17161030e-01 -6.43606901e-01
-9.29111898e-01 -9.85112548e-01 -1.11750567e+00 3.59201506e-02
-6.43850207e-01 1.64147854e-01 1.75471023e-01 9.63058412e-01
6.74174726e-02 5.62406123e-01 5.85801244e-01 -9.37064111e-01
-5.06793618e-01 -8.67513895e-01 -4.30333048e-01 4.01119679e-01
4.60893661e-01 -1.25818706e+00 -3.74039680e-01 5.20214319e-01] | [6.379184722900391, -2.113546133041382] |
0cf90b6b-1f38-408f-b9c1-2b695cf56a9a | im-loss-information-maximization-loss-for | null | null | https://openreview.net/forum?id=Jw34v_84m2b | https://openreview.net/pdf?id=Jw34v_84m2b | IM-Loss: Information Maximization Loss for Spiking Neural Networks | Spiking Neural Network (SNN), recognized as a type of biologically plausible architecture, has recently drawn much research attention. It transmits information by 0/1 spikes. This bio-mimetic mechanism of SNN demonstrates extreme energy efficiency since it avoids any multiplications on neuromorphic hardware. However, the forward-passing 0/1 spike quantization will cause information loss and accuracy degradation. To deal with this problem, the Information maximization loss (IM-Loss) that aims at maximizing the information flow in the SNN is proposed in the paper. The IM-Loss not only enhances the information expressiveness of an SNN directly but also plays a part of the role of normalization without introducing any additional operations (e.g., bias and scaling) in the inference phase. Additionally, we introduce a novel differentiable spike activity estimation, Evolutionary Surrogate Gradients (ESG) in SNNs. By appointing automatic evolvable surrogate gradients for spike activity function, ESG can ensure sufficient model updates at the beginning and accurate gradients at the end of the training, resulting in both easy convergence and high task performance. Experimental results on both popular non-spiking static and neuromorphic datasets show that the SNN models trained by our method outperform the current state-of-the-art algorithms. | ['Zhe Ma', 'Xuhui Huang', 'YingLei Wang', 'Xiaode Liu', 'Liwen Zhang', 'Yuanpei Chen', 'Yufei Guo'] | 2022-10-31 | null | null | null | neurips-2022-10 | ['event-data-classification'] | ['computer-vision'] | [ 5.04310846e-01 -3.28247666e-01 3.97681236e-01 -2.45263934e-01
-6.60477765e-03 -8.35346654e-02 4.85622197e-01 -4.60100845e-02
-8.27182412e-01 9.96722281e-01 -2.83275157e-01 1.59196138e-01
-8.19137990e-02 -7.82570899e-01 -9.99172926e-01 -1.05470228e+00
4.77019325e-02 -9.01071280e-02 4.94254500e-01 -1.53147712e-01
6.21552706e-01 3.77650976e-01 -1.57183981e+00 -1.42999366e-01
1.05450606e+00 1.48179257e+00 5.44664919e-01 1.59869134e-01
-2.49806300e-01 4.94159788e-01 -6.14349842e-01 -3.37705582e-01
3.65438461e-01 -6.45812631e-01 3.74594331e-02 -5.27013063e-01
-2.10597485e-01 6.02043048e-03 -5.00873804e-01 1.30263805e+00
7.41329789e-01 -1.28435761e-01 5.09914637e-01 -1.34965253e+00
-4.62661684e-01 7.59435892e-01 -3.51964384e-01 4.00400192e-01
-4.50258881e-01 4.95003879e-01 3.41115326e-01 -5.78200459e-01
3.63181472e-01 9.08707023e-01 7.48918533e-01 6.99582458e-01
-1.20879543e+00 -1.05916166e+00 -8.70936364e-02 2.40956992e-01
-1.48999381e+00 -4.60093915e-01 7.66391993e-01 -7.32368752e-02
8.93007994e-01 5.86356036e-03 1.02251244e+00 9.69816267e-01
6.24589801e-01 5.91603279e-01 8.65746021e-01 8.29625577e-02
7.51043379e-01 -2.09833235e-01 -9.96954590e-02 3.99302423e-01
4.68148947e-01 4.35098186e-02 -8.16079855e-01 8.95150974e-02
1.06128097e+00 -1.17517568e-01 -3.94938171e-01 -2.76332404e-02
-9.42797840e-01 2.50469387e-01 7.42577195e-01 2.29864433e-01
-4.88486290e-01 6.16136968e-01 3.86644959e-01 -3.26590985e-02
-2.30963621e-02 3.20095569e-01 -2.66499929e-02 -4.45183843e-01
-9.06942248e-01 3.58605832e-02 4.38156068e-01 7.24669576e-01
5.89679420e-01 4.66693640e-01 -3.18979621e-01 7.88444102e-01
1.30840182e-01 4.40683097e-01 8.84822488e-01 -8.62305164e-01
1.53196841e-01 8.59633803e-01 -1.73549607e-01 -8.81550431e-01
-3.11641395e-01 -7.39444137e-01 -1.30061924e+00 1.50041237e-01
1.55445397e-01 -5.00394497e-03 -8.20250273e-01 2.06821609e+00
-1.17065050e-02 4.54388320e-01 -1.10640720e-01 8.68778110e-01
3.73035312e-01 7.91362882e-01 1.25668019e-01 -3.95893067e-01
1.04141212e+00 -5.50221145e-01 -8.71539354e-01 -3.76962483e-01
-7.07952976e-02 -2.19449207e-01 7.72243679e-01 1.98796570e-01
-1.24811268e+00 -4.79757249e-01 -1.39842558e+00 -9.91825201e-03
-3.13167393e-01 -2.03969568e-01 6.07730925e-01 4.50939655e-01
-1.12101412e+00 7.84010589e-01 -1.05754364e+00 -1.65965520e-02
6.61284089e-01 5.85780919e-01 7.63431862e-02 4.51584518e-01
-9.96389985e-01 7.84230471e-01 4.35574234e-01 2.68313318e-01
-6.11123145e-01 -7.86680162e-01 -4.04871196e-01 3.25256765e-01
-2.36721426e-01 -5.96161246e-01 9.06165123e-01 -8.82656693e-01
-1.69919825e+00 4.40358311e-01 -2.45782107e-01 -6.55128002e-01
3.68777514e-01 3.46887380e-01 -2.20002711e-01 -2.10400060e-01
-2.94126630e-01 8.70576262e-01 5.21256328e-01 -9.07911003e-01
-3.31094563e-01 -4.91282105e-01 -5.11991024e-01 -6.16018884e-02
-7.57076800e-01 -4.00694191e-01 -3.16717923e-01 -8.43436241e-01
3.23031545e-01 -5.96598029e-01 -5.83338030e-02 5.52698612e-01
-3.86622921e-02 -1.46740839e-01 4.92402673e-01 -4.81330007e-01
1.31356728e+00 -2.15921402e+00 4.52985317e-02 1.66461319e-01
9.41210426e-03 3.77976686e-01 -2.82351859e-02 5.94459809e-02
4.02282625e-01 -2.06110269e-01 -8.52941155e-01 -6.17479458e-02
-6.89493641e-02 2.11086467e-01 -1.22896835e-01 2.95663744e-01
4.68048334e-01 1.02716255e+00 -7.65101016e-01 -2.31565788e-01
-7.08899349e-02 7.41700768e-01 -4.65061486e-01 4.15915325e-02
-9.41978842e-02 3.19550455e-01 -6.64612502e-02 4.78010178e-01
8.20715308e-01 -2.03737646e-01 -9.32912901e-02 -1.32042110e-01
-3.87258321e-01 1.17802672e-01 -1.01206732e+00 1.77962959e+00
-2.65527606e-01 5.56241155e-01 -1.42787039e-01 -1.10127652e+00
1.34066081e+00 -1.68806359e-01 3.06555122e-01 -1.19676602e+00
4.00128156e-01 5.42495012e-01 2.58281708e-01 -9.40025151e-02
3.21016237e-02 2.06557155e-01 2.59608060e-01 1.91397280e-01
-5.06954640e-03 2.18035117e-01 9.01766866e-02 -1.45104706e-01
9.91766691e-01 2.67207362e-02 -1.50782010e-03 -6.39138937e-01
4.70126629e-01 -4.27455246e-01 9.74192262e-01 5.32808483e-01
-2.65063763e-01 2.94740468e-01 1.75746262e-01 -1.74374551e-01
-1.13779783e+00 -1.17932844e+00 -4.40427423e-01 5.25652289e-01
5.44438839e-01 1.68554425e-01 -1.02211726e+00 2.53709793e-01
-2.83088796e-02 4.40847039e-01 -4.83753562e-01 -6.43557787e-01
-5.64811587e-01 -1.11546993e+00 8.36471915e-01 6.78561509e-01
1.05482030e+00 -1.26231766e+00 -1.02252448e+00 6.07171655e-01
-9.32758376e-02 -8.00899029e-01 -4.30356741e-01 6.23765588e-01
-1.17512846e+00 -4.23803687e-01 -6.04951859e-01 -8.55850101e-01
7.36701012e-01 -3.37381244e-01 6.92817032e-01 -1.17092043e-01
-5.56070089e-01 -5.22543073e-01 1.83965117e-01 -4.36367780e-01
8.90310332e-02 -1.31470531e-01 -2.29294598e-02 9.08338279e-02
5.57578444e-01 -1.09699774e+00 -9.35746253e-01 1.02933712e-01
-9.76054788e-01 2.73972332e-01 6.03025556e-01 8.66902053e-01
9.52715278e-01 -2.93648928e-01 8.89809489e-01 -2.61429489e-01
5.78206122e-01 -3.40901792e-01 -7.44797468e-01 1.54725239e-01
-8.29688013e-01 5.75245619e-01 9.01492715e-01 -5.87224007e-01
-7.89990604e-01 -5.10675833e-02 -1.60694540e-01 -1.43513352e-01
4.68676984e-01 9.43699852e-05 -8.83860663e-02 -3.85467142e-01
5.41313887e-01 9.73181367e-01 1.09598689e-01 -2.69069582e-01
-2.77590692e-01 5.34138441e-01 7.57632852e-01 -2.53773808e-01
4.05491710e-01 4.33633566e-01 2.34893084e-01 -5.54680586e-01
5.81560936e-03 4.50364277e-02 3.34335193e-02 -3.22418481e-01
4.82358217e-01 -6.59149945e-01 -1.12165427e+00 1.03334606e+00
-1.13429749e+00 -1.40965015e-01 -2.21587792e-01 2.99521744e-01
-5.11227429e-01 4.97614406e-02 -6.60058260e-01 -1.02625203e+00
-8.06296825e-01 -8.89142632e-01 5.42335391e-01 7.10477412e-01
1.05874240e-01 -5.02584934e-01 -1.34051353e-01 -3.15459698e-01
8.32597256e-01 2.61387855e-01 9.05527234e-01 -2.32086480e-01
-6.76148474e-01 1.32913366e-01 -2.41756022e-01 2.52225012e-01
-8.68696347e-02 -1.47148132e-01 -1.10009897e+00 -2.18864530e-01
3.67988020e-01 -2.44536594e-01 9.46144879e-01 5.05607009e-01
1.34851396e+00 -2.90559292e-01 -2.16603369e-01 1.01176882e+00
1.69492316e+00 5.50703049e-01 9.44518983e-01 3.47974509e-01
2.61058301e-01 1.79403365e-01 -1.74059987e-01 6.15382850e-01
1.91275552e-01 3.45680296e-01 5.41077971e-01 2.64649510e-01
-1.38328344e-01 -3.52808475e-01 2.64422983e-01 1.25374174e+00
-2.02521775e-03 -1.21711038e-01 -5.71452737e-01 3.61315250e-01
-1.80957747e+00 -7.52263308e-01 5.60866073e-02 2.25457239e+00
1.16924644e+00 3.27395290e-01 -2.55169600e-01 3.98509324e-01
7.92473137e-01 -2.77882874e-01 -1.46904767e+00 -2.24299639e-01
-6.23150110e-01 4.16331857e-01 5.40016770e-01 -2.32649334e-02
-3.81796569e-01 4.69837368e-01 5.50829983e+00 8.22908103e-01
-1.35044611e+00 -9.65764076e-02 6.19235814e-01 -2.82410860e-01
-2.76534319e-01 -3.78995389e-01 -7.44299889e-01 1.22843945e+00
9.70225573e-01 -3.16710413e-01 9.58941698e-01 5.06469846e-01
1.74772650e-01 -1.24964416e-01 -8.07569444e-01 1.29589307e+00
-3.00285637e-01 -1.44210935e+00 1.92952558e-01 -1.05955437e-01
6.08970404e-01 1.08040590e-02 1.07370261e-02 1.14861742e-01
-2.26133108e-01 -7.65272915e-01 9.65817809e-01 8.02900851e-01
6.23993754e-01 -9.12306666e-01 6.31950855e-01 4.22016203e-01
-1.14028645e+00 -2.61273891e-01 -7.14762866e-01 -1.56930506e-01
5.76581806e-03 6.83889806e-01 -1.81279525e-01 6.12691343e-02
6.85147405e-01 5.17039716e-01 -4.16972786e-01 1.47306848e+00
1.05288327e-01 3.32252920e-01 -5.83386779e-01 -6.73571825e-01
5.27104959e-02 -3.67045373e-01 3.52555901e-01 1.02630079e+00
5.77919781e-01 1.87013596e-01 -5.50124288e-01 1.39254403e+00
-3.95488739e-01 -4.99545038e-02 -2.48937100e-01 -6.42939061e-02
9.73044038e-01 9.46041644e-01 -7.82848537e-01 -2.85047069e-02
1.93369538e-01 9.74032938e-01 3.12841833e-01 2.21963167e-01
-7.37180471e-01 -8.60645056e-01 6.58563852e-01 1.22375771e-01
1.00009419e-01 1.37022650e-02 -9.32740271e-01 -6.97672844e-01
4.01414931e-01 -2.68849730e-01 -2.06319109e-01 -6.01576388e-01
-1.05769539e+00 6.27836525e-01 -6.31636858e-01 -1.14081144e+00
-1.46630079e-01 -4.89486367e-01 -6.49556041e-01 8.44771743e-01
-1.53769135e+00 -3.91685426e-01 -4.64879572e-01 3.97840768e-01
2.74747998e-01 4.68736887e-02 5.35863936e-01 4.29128498e-01
-7.48755097e-01 8.06025088e-01 3.73580128e-01 -1.04421973e-01
1.55708000e-01 -7.66778946e-01 5.63059330e-01 8.37701261e-01
-4.14706498e-01 6.24901056e-01 5.73889613e-01 -3.72423649e-01
-1.52701688e+00 -9.34890270e-01 8.00485492e-01 2.52263099e-01
4.18045461e-01 -4.92756426e-01 -1.05154037e+00 3.03223878e-02
4.90970984e-02 -1.72193781e-01 3.42712283e-01 -7.25893259e-01
-1.56890333e-01 -5.99063873e-01 -1.37288904e+00 6.90321922e-01
1.41282845e+00 -1.99903041e-01 -2.38562137e-01 -1.23144820e-01
4.71567273e-01 -2.43060850e-02 -6.13705158e-01 5.34854174e-01
6.17851436e-01 -9.37760592e-01 7.04359889e-01 -2.09224820e-02
3.52317870e-01 -4.18704659e-01 1.70612317e-02 -1.07874155e+00
-3.57221901e-01 -5.78681111e-01 -2.35922396e-01 1.15641904e+00
2.26305783e-01 -8.29666615e-01 5.82517684e-01 5.45050323e-01
-2.62370586e-01 -1.05439317e+00 -1.24618661e+00 -1.02270126e+00
-7.11994767e-02 -1.72207057e-02 5.50835371e-01 3.82949769e-01
1.24699421e-01 -9.72643271e-02 3.49574583e-03 -1.33373022e-01
9.78596628e-01 -2.70385146e-01 1.53780431e-01 -1.36063933e+00
-2.89043635e-02 -8.94122720e-01 -7.08891809e-01 -9.76224661e-01
-1.63476020e-01 -8.71987581e-01 4.56663132e-01 -1.35375571e+00
2.72346586e-01 -3.83335829e-01 -6.91487193e-01 3.57189417e-01
-1.23986229e-01 1.80327743e-01 -4.59089763e-02 2.39402592e-01
-2.35180572e-01 8.58781934e-01 1.05124187e+00 -6.30208105e-02
-1.59763992e-02 -4.42405939e-01 -5.06314576e-01 4.08643693e-01
8.70196939e-01 -5.74513078e-01 -4.55858171e-01 -5.82707465e-01
2.88188197e-02 -2.84408867e-01 2.64445245e-01 -1.53829956e+00
8.00501704e-01 1.58235967e-01 6.49354219e-01 -1.80467874e-01
3.85362208e-01 -5.36365986e-01 4.39617157e-01 1.00851405e+00
-2.57122576e-01 1.05277449e-01 2.19575137e-01 5.18344164e-01
-9.46212485e-02 -3.09408367e-01 1.17054987e+00 4.23777476e-02
-5.14982104e-01 3.10331404e-01 -3.02064210e-01 1.95662528e-01
8.12399089e-01 -7.50439823e-01 -3.46440166e-01 1.48474693e-01
3.64066511e-02 1.82074890e-01 3.63322914e-01 2.67056704e-01
7.16645658e-01 -1.39619005e+00 -5.16964912e-01 5.41032791e-01
-8.93744379e-02 -1.00346252e-01 9.63932052e-02 6.47417188e-01
-4.68387067e-01 2.27546483e-01 -8.29922497e-01 -5.96393704e-01
-4.04262185e-01 1.31487787e-01 3.77604306e-01 1.51372716e-01
-3.65179181e-01 1.01958060e+00 -8.22715536e-02 7.96416700e-02
6.21101499e-01 -2.98008531e-01 7.70042986e-02 -2.50141978e-01
5.79582810e-01 3.77644897e-01 1.15958795e-01 2.40425766e-03
-4.59794700e-01 4.34161544e-01 1.79331303e-01 -1.10853270e-01
1.48647237e+00 8.51176977e-02 -3.48318368e-01 5.75821400e-01
9.93836343e-01 -8.33565831e-01 -1.68265033e+00 -7.71300048e-02
6.49690032e-02 -1.16943642e-01 1.43196091e-01 -8.07311773e-01
-1.15981019e+00 9.38797832e-01 9.94747400e-01 -1.69394791e-01
1.43667030e+00 -6.34038687e-01 1.14763331e+00 4.06260401e-01
6.89861655e-01 -1.26476693e+00 3.14475447e-02 4.61963594e-01
6.90616965e-01 -7.01203823e-01 -4.94756699e-01 1.28023073e-01
-2.44237259e-01 1.06595230e+00 8.21977913e-01 -2.78926015e-01
5.32497525e-01 7.28182018e-01 -3.12055647e-01 2.11916983e-01
-7.64861107e-01 2.79974103e-01 -2.41725534e-01 4.69850421e-01
1.21564142e-01 -2.10474223e-01 -7.13082612e-01 8.23928833e-01
-1.27012670e-01 3.89109731e-01 1.80993706e-01 9.84686494e-01
-7.70000577e-01 -6.42917693e-01 1.18661016e-01 5.40067554e-01
-2.14599669e-01 -3.79334927e-01 -9.54821408e-02 1.97921544e-01
6.64799213e-02 5.42324901e-01 3.99643689e-01 -4.85817254e-01
3.11265767e-01 6.73001036e-02 5.30722976e-01 1.65291160e-01
-7.71013618e-01 -2.68500745e-01 -6.90223336e-01 -4.06230390e-01
-7.54252523e-02 -3.03110540e-01 -1.83206916e+00 -3.84939611e-01
-2.48850614e-01 3.11264358e-02 1.18053198e+00 6.18033648e-01
6.64969444e-01 7.59294093e-01 5.99512815e-01 -7.14025915e-01
-7.76139677e-01 -8.63480687e-01 -5.15781999e-01 3.50821257e-01
-1.29597187e-02 -5.00213087e-01 -4.53027964e-01 8.20981264e-02] | [8.233723640441895, 2.4896819591522217] |
5cf46312-815f-4e87-926f-770d82c14ee3 | calibrated-and-partially-calibrated-semi | 2103.06535 | null | https://arxiv.org/abs/2103.06535v3 | https://arxiv.org/pdf/2103.06535v3.pdf | Calibrated and Partially Calibrated Semi-Generalized Homographies | In this paper, we propose the first minimal solutions for estimating the semi-generalized homography given a perspective and a generalized camera. The proposed solvers use five 2D-2D image point correspondences induced by a scene plane. One of them assumes the perspective camera to be fully calibrated, while the other solver estimates the unknown focal length together with the absolute pose parameters. This setup is particularly important in structure-from-motion and image-based localization pipelines, where a new camera is localized in each step with respect to a set of known cameras and 2D-3D correspondences might not be available. As a consequence of a clever parametrization and the elimination ideal method, our approach only needs to solve a univariate polynomial of degree five or three. The proposed solvers are stable and efficient as demonstrated by a number of synthetic and real-world experiments. | ['Zuzana Kukelova', 'Janne Heikkila', 'Patrik Beliansky', 'Daniel Barath', 'Torsten Sattler', 'Snehal Bhayani'] | 2021-03-11 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Bhayani_Calibrated_and_Partially_Calibrated_Semi-Generalized_Homographies_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Bhayani_Calibrated_and_Partially_Calibrated_Semi-Generalized_Homographies_ICCV_2021_paper.pdf | iccv-2021-1 | ['image-based-localization'] | ['computer-vision'] | [ 4.78535481e-02 4.89618815e-02 6.06303252e-02 -1.24998279e-01
-6.10609353e-01 -9.66476679e-01 4.56263244e-01 -7.23063648e-02
-3.86334598e-01 5.02398789e-01 -4.73135442e-01 1.34669140e-01
1.43581787e-02 -3.53139788e-01 -8.03052247e-01 -4.59697604e-01
2.29338899e-01 9.00300741e-01 1.92748234e-01 1.02989659e-01
5.27079582e-01 7.50314534e-01 -1.13531065e+00 -5.68980515e-01
4.82730567e-01 8.93274128e-01 3.83992642e-01 7.83171713e-01
2.87978202e-01 2.69843459e-01 -2.67391741e-01 -2.72602588e-01
7.55783260e-01 -2.05218762e-01 -5.66514313e-01 6.96465969e-01
7.35862911e-01 -5.41407526e-01 -5.51327616e-02 1.17814219e+00
1.77845567e-01 6.43521920e-02 1.74147069e-01 -1.15633321e+00
3.32874715e-01 -8.69931355e-02 -7.37933099e-01 -3.73964339e-01
8.35110724e-01 1.24708377e-01 7.58819580e-01 -1.04655850e+00
9.51345742e-01 1.02727163e+00 6.97241962e-01 -7.51742199e-02
-1.10074914e+00 -1.16339467e-01 -6.89562410e-02 -1.32858127e-01
-1.58363831e+00 -4.06457722e-01 8.05548847e-01 -5.14135182e-01
6.04365468e-01 3.58287878e-02 7.00173914e-01 6.48508012e-01
6.76645190e-02 -7.27672502e-02 8.92535627e-01 -4.98584151e-01
2.11402848e-01 3.23263764e-01 -1.71923707e-03 7.06930220e-01
6.08503044e-01 -8.32949653e-02 -3.22174639e-01 -2.75975913e-01
1.20658565e+00 1.07713453e-01 -3.61965925e-01 -1.07928526e+00
-1.54359317e+00 6.42379999e-01 1.86149493e-01 7.40865071e-04
-3.01958293e-01 4.59592678e-02 -2.17350826e-01 2.46193744e-02
7.17177615e-02 6.08605981e-01 -2.43892387e-01 -1.64024919e-01
-8.32620502e-01 1.09123932e-02 9.85634685e-01 1.42621303e+00
1.25438225e+00 -1.20461941e-01 7.66413867e-01 2.74106205e-01
1.58588603e-01 7.11315334e-01 6.42166659e-02 -1.30887294e+00
4.16775286e-01 5.21290183e-01 4.61981267e-01 -1.29341245e+00
-3.08735490e-01 -4.94468570e-01 -6.71595275e-01 1.19976521e-01
5.80302179e-01 -1.05879735e-02 -4.20795947e-01 1.25751567e+00
5.11108935e-01 2.95771241e-01 -9.99857485e-02 1.11057472e+00
4.13022339e-02 3.43312562e-01 -1.02143335e+00 -3.72973531e-01
1.19429302e+00 -7.91837811e-01 -3.31580222e-01 -7.70400584e-01
3.34892184e-01 -1.20878959e+00 4.85578895e-01 4.17686194e-01
-1.23307920e+00 -2.69609839e-01 -1.13225508e+00 -1.59678180e-02
5.59067540e-02 4.62777644e-01 2.05927372e-01 2.75770783e-01
-8.79026294e-01 3.90533566e-01 -8.47418487e-01 -4.79066700e-01
-4.54100370e-01 5.54057360e-01 -7.09638476e-01 -3.55914570e-02
-6.05602622e-01 1.04806364e+00 2.12623373e-01 3.29319328e-01
-6.08059704e-01 -3.56702685e-01 -7.65736997e-01 2.74578985e-02
6.86972797e-01 -8.76227558e-01 1.09898317e+00 -7.04819024e-01
-1.69138122e+00 9.53663468e-01 -2.45536029e-01 -1.71739563e-01
8.13451886e-01 -3.16600114e-01 2.65557915e-01 4.56368804e-01
3.10343504e-02 1.74357176e-01 9.33421850e-01 -1.35318327e+00
-3.20578903e-01 -5.07829070e-01 3.45286995e-01 4.56399918e-01
1.13364391e-01 -3.40941012e-01 -8.28638852e-01 -1.55103216e-02
8.32181573e-01 -1.35493112e+00 -4.82769251e-01 -2.40227003e-02
-3.97201270e-01 7.69150734e-01 7.34948575e-01 -6.31293774e-01
5.44107318e-01 -1.89016569e+00 4.35803354e-01 4.17941272e-01
1.43242940e-01 -1.82883516e-01 1.71216622e-01 4.88134116e-01
8.70411657e-03 -3.83720189e-01 1.39169618e-01 -4.57212031e-01
-4.31054175e-01 -1.91611368e-02 -6.62266538e-02 1.00596046e+00
-3.24413359e-01 2.33492464e-01 -8.79704893e-01 -2.33575150e-01
5.07491946e-01 4.84943420e-01 -3.71632397e-01 3.56255293e-01
7.97444806e-02 8.62293780e-01 -4.22742456e-01 3.63541067e-01
1.00667059e+00 -2.01413020e-01 3.29006910e-01 -4.54499513e-01
-5.10623753e-01 2.36720309e-01 -1.80774009e+00 1.90745008e+00
-5.25567114e-01 4.62248176e-01 3.29280317e-01 -5.73887885e-01
8.40804279e-01 2.24365473e-01 6.62208080e-01 -8.21520835e-02
4.00725693e-01 3.11537027e-01 -2.48348042e-01 -4.94880006e-02
6.07032180e-01 1.35385081e-01 1.23139024e-01 2.43808419e-01
1.59981381e-02 -6.15168512e-01 1.91911817e-01 8.51191282e-02
9.62737679e-01 1.25296041e-01 6.67032123e-01 -2.75124669e-01
7.38575637e-01 -6.90839766e-03 7.61065960e-01 3.30583811e-01
3.31487477e-01 8.89772952e-01 5.79062760e-01 -4.00874853e-01
-1.26178777e+00 -6.47859275e-01 9.17215422e-02 -1.35564163e-01
5.20480871e-01 -3.81129980e-01 -8.53278875e-01 -2.28754386e-01
-1.59342989e-01 2.51654029e-01 -1.89085633e-01 1.20727427e-01
-6.67425275e-01 7.53794834e-02 -1.93164170e-01 1.30289420e-01
5.33644497e-01 4.46656831e-02 -1.02864301e+00 -3.53219919e-02
-1.76043540e-01 -1.65662050e+00 -6.62269115e-01 -1.08496463e-02
-9.49737549e-01 -1.43210006e+00 -6.07304752e-01 -5.69605052e-01
1.16375279e+00 7.83190131e-01 7.94631362e-01 -1.40832245e-01
1.21138476e-01 6.19778216e-01 -5.96286803e-02 1.15730263e-01
-1.64964736e-01 -4.79313247e-02 2.66665816e-01 2.50247836e-01
-1.21432953e-01 -5.24576426e-01 -4.90432173e-01 7.93520689e-01
-4.19350654e-01 1.62147835e-01 4.25491869e-01 6.90268457e-01
7.92479396e-01 -4.42894083e-03 -5.51400602e-01 -6.43483043e-01
-2.56033186e-02 -6.14835583e-02 -1.65721476e+00 9.49498191e-02
-2.61724979e-01 -1.32942528e-01 5.75838923e-01 -5.18588662e-01
-8.71120811e-01 1.00903904e+00 4.65379208e-01 -9.22052383e-01
-1.11098625e-02 4.41552430e-01 -4.21024710e-01 -5.54120004e-01
4.68701839e-01 -7.61038586e-02 -3.62842940e-02 -3.97180319e-01
2.16937408e-01 1.67311028e-01 6.64827704e-01 -3.94384444e-01
1.25493491e+00 7.07661688e-01 4.36487794e-01 -1.06966174e+00
-5.45220375e-01 -8.42680275e-01 -1.40706837e+00 -1.28391325e-01
7.32159495e-01 -1.01587546e+00 -8.55367064e-01 4.28749204e-01
-1.55617630e+00 2.17372149e-01 -5.21618947e-02 8.80787730e-01
-5.91873229e-01 7.28197575e-01 -1.30481213e-01 -6.75794661e-01
8.33595619e-02 -1.59388542e+00 1.20900273e+00 7.66955987e-02
-9.22667310e-02 -8.15469444e-01 1.07275330e-01 1.31745785e-01
1.15920156e-02 3.49646866e-01 9.51679200e-02 3.45665915e-03
-1.18560362e+00 -5.52448571e-01 8.47427696e-02 9.90138873e-02
2.88827717e-03 7.85167217e-02 -7.76993036e-01 -4.29285824e-01
5.73802054e-01 3.01538378e-01 1.42187908e-01 2.76039928e-01
3.29741687e-01 -1.30943149e-01 -2.78833508e-01 1.10092342e+00
1.62428272e+00 1.76276252e-01 2.50835717e-01 4.48401392e-01
8.12012076e-01 4.07402068e-01 8.20772290e-01 4.95917082e-01
3.27053696e-01 1.00998735e+00 7.32832789e-01 3.78192216e-02
2.80027807e-01 -2.92059839e-01 1.75087973e-01 8.30135345e-01
-1.32735148e-01 1.81849599e-01 -8.80707681e-01 3.48221779e-01
-1.74254978e+00 -5.92762530e-01 -5.00710905e-01 2.86488938e+00
1.26432672e-01 -1.02053232e-01 -7.31471479e-02 -4.94536348e-02
7.01205194e-01 -3.77132632e-02 -5.34207225e-01 1.41299218e-01
2.85696629e-02 -3.18900436e-01 1.05694914e+00 8.92106235e-01
-7.99100637e-01 8.90711427e-01 5.70227766e+00 -1.22089550e-01
-1.16592836e+00 -1.01525612e-01 -7.74321109e-02 1.80785775e-01
-4.38465811e-02 8.77995849e-01 -1.02306163e+00 1.84746817e-01
4.88063514e-01 -2.54124165e-01 5.37560880e-01 1.02377141e+00
-2.83222320e-03 -4.88380641e-01 -1.30807972e+00 1.39268517e+00
3.70011002e-01 -1.22766840e+00 -4.01114762e-01 3.74258667e-01
6.59789860e-01 -9.92364138e-02 -2.47826770e-01 -5.84326506e-01
-1.15857095e-01 -3.36382180e-01 7.11419165e-01 3.95410448e-01
6.92757249e-01 -5.52648127e-01 5.54817915e-01 7.14255035e-01
-1.14618683e+00 2.28234753e-01 -6.25177622e-01 -1.93470821e-01
4.60346818e-01 7.27540672e-01 -1.03000641e+00 6.40353143e-01
2.94964969e-01 6.16372108e-01 -3.34876120e-01 1.13018537e+00
-3.73402417e-01 -6.13584034e-02 -6.31746411e-01 4.28333908e-01
-2.23061126e-02 -7.73148656e-01 8.06514859e-01 6.88458323e-01
6.88254952e-01 1.92012280e-01 2.31160998e-01 7.56394565e-01
6.97974861e-02 -1.61622480e-01 -8.33496869e-01 4.19916898e-01
3.43943745e-01 1.57070839e+00 -7.32563198e-01 -6.19511679e-02
-3.78577709e-01 1.15681469e+00 -5.09568900e-02 3.61948997e-01
-7.43534148e-01 -1.78795680e-01 4.97971296e-01 2.60935992e-01
6.25780895e-02 -9.32849169e-01 -8.67952630e-02 -1.78278244e+00
2.40762785e-01 -7.51592040e-01 -1.46648020e-01 -8.58948648e-01
-5.29210508e-01 6.62699282e-01 1.86395898e-01 -1.57588243e+00
-4.83247042e-01 -6.00764215e-01 -3.51259828e-01 8.64200115e-01
-1.21437716e+00 -1.02452993e+00 -5.88104308e-01 5.08913398e-01
3.47094268e-01 7.35569522e-02 6.05952203e-01 2.14156181e-01
-3.21336865e-01 7.30633512e-02 -2.41889283e-02 -1.46929577e-01
7.26277232e-01 -1.11278784e+00 3.36707145e-01 1.20242083e+00
1.23402797e-01 8.47346485e-01 8.31395209e-01 -4.52493757e-01
-1.97146523e+00 -5.64594209e-01 7.56509960e-01 -5.73172748e-01
5.65751195e-01 -5.19552350e-01 -5.64738035e-01 1.12327158e+00
-5.00273407e-02 -1.48050273e-02 -4.96104695e-02 -3.19906145e-01
-2.02376209e-02 -3.02455842e-01 -9.83346999e-01 9.82901528e-02
7.81786501e-01 -4.93137151e-01 -2.63417006e-01 5.62679827e-01
2.77727902e-01 -1.08288527e+00 -6.56047940e-01 7.82243609e-02
5.26818931e-01 -1.04689145e+00 1.06212473e+00 2.18000725e-01
-1.77117035e-01 -6.70735240e-01 -2.20493272e-01 -1.25785303e+00
-1.49971619e-01 -9.41261888e-01 1.98899433e-01 9.71636653e-01
-5.86159714e-02 -7.67086506e-01 8.06670964e-01 4.98565227e-01
1.19528465e-01 -1.15582637e-01 -8.52708399e-01 -8.69914353e-01
-6.43325388e-01 -6.01165369e-02 4.20838982e-01 9.72775877e-01
-2.98148274e-01 6.88075662e-01 -5.99748850e-01 8.48645449e-01
9.24772501e-01 2.76006877e-01 1.46528757e+00 -1.22892451e+00
-6.01391196e-01 6.54185638e-02 -7.04438925e-01 -1.49434030e+00
5.06409705e-02 -2.87050515e-01 -3.43068466e-02 -1.10251749e+00
7.78196380e-02 -1.43692985e-01 5.53157151e-01 -1.32470474e-01
3.94947916e-01 -1.97082534e-01 1.62630901e-01 3.14406633e-01
-2.82901615e-01 9.15059969e-02 7.92686522e-01 1.92951486e-01
-2.24948645e-01 4.80826318e-01 -1.87652528e-01 1.16635716e+00
5.02725601e-01 -3.19485784e-01 -5.44823647e-01 -7.59447217e-01
3.75400484e-01 5.39697707e-01 3.95707905e-01 -9.59343910e-01
5.90674341e-01 -2.77284205e-01 1.50422573e-01 -8.27879369e-01
7.61323452e-01 -1.23949039e+00 6.51905656e-01 2.00534627e-01
3.10123414e-01 4.18946952e-01 -2.16140389e-01 3.01850259e-01
-2.80161589e-01 -4.28938597e-01 8.34167659e-01 -2.93562442e-01
-4.73001331e-01 1.96215108e-01 1.72674254e-01 -3.47226501e-01
1.15324092e+00 -4.83031571e-01 -1.24996394e-01 -7.03317106e-01
-5.24232924e-01 -1.87611729e-02 1.22145402e+00 4.94934358e-02
7.46141911e-01 -1.07619786e+00 -2.04824418e-01 3.84204239e-01
1.77194612e-04 4.20153856e-01 1.01728573e-01 1.03086257e+00
-1.11952031e+00 4.91965771e-01 -1.02240667e-01 -1.02901018e+00
-1.34831953e+00 6.29402280e-01 5.53829908e-01 -9.32796896e-02
-4.45003986e-01 4.21418309e-01 3.14555258e-01 -5.11901319e-01
1.10163905e-01 -4.00911450e-01 3.52529049e-01 -3.39014798e-01
2.72697657e-01 6.32468462e-01 5.29183634e-02 -1.08472681e+00
-3.88151228e-01 1.27645385e+00 2.13015765e-01 -3.41877669e-01
1.15612924e+00 -3.49272966e-01 -3.05586696e-01 3.48449856e-01
1.30952287e+00 4.29989070e-01 -1.29017985e+00 -7.54953548e-02
-1.22509122e-01 -8.62457633e-01 -2.88841482e-02 -1.16803594e-01
-1.02006960e+00 9.64530468e-01 2.60637015e-01 -2.47586861e-01
9.24683511e-01 -2.59852976e-01 5.66206753e-01 6.44811034e-01
7.49871850e-01 -8.51299822e-01 -2.14957044e-01 6.07289732e-01
7.96379507e-01 -1.03328812e+00 4.60691869e-01 -8.05849969e-01
-3.43323857e-01 1.30128431e+00 5.62725246e-01 -1.97782442e-01
4.19430673e-01 1.07727170e-01 4.86429892e-02 8.94947052e-02
-2.73490101e-01 1.20316811e-01 2.35786125e-01 3.14850986e-01
8.43726173e-02 -2.49044985e-01 -5.35514019e-02 -2.19395518e-01
-2.25547150e-01 -2.78852791e-01 9.21411693e-01 7.59482861e-01
-2.92854846e-01 -1.35750902e+00 -7.32508779e-01 -3.62485945e-01
5.82585745e-02 3.61146569e-01 -4.67572242e-01 9.04324412e-01
-1.59527004e-01 6.91267133e-01 -1.15773261e-01 -1.53083786e-01
5.27405500e-01 -4.62098747e-01 7.02573478e-01 -6.47360146e-01
-1.62853897e-01 6.10710144e-01 -1.41092047e-01 -7.89675474e-01
-3.01932484e-01 -7.71435320e-01 -1.16309476e+00 -9.67681780e-02
-5.84322095e-01 9.47933719e-02 1.05687642e+00 6.08954310e-01
3.90112549e-01 -1.60240665e-01 8.41487288e-01 -1.15872645e+00
-5.90466380e-01 -5.84982574e-01 -7.04573214e-01 3.64926681e-02
4.52317774e-01 -6.02832794e-01 -6.06090724e-01 1.04754366e-01] | [7.872624397277832, -2.325680732727051] |
baac7915-21d8-4e91-bcfc-3e6d30342cfb | rescue-conversations-from-dead-ends-efficient | 2305.03262 | null | https://arxiv.org/abs/2305.03262v1 | https://arxiv.org/pdf/2305.03262v1.pdf | Rescue Conversations from Dead-ends: Efficient Exploration for Task-oriented Dialogue Policy Optimization | Training a dialogue policy using deep reinforcement learning requires a lot of exploration of the environment. The amount of wasted invalid exploration makes their learning inefficient. In this paper, we find and define an important reason for the invalid exploration: dead-ends. When a conversation enters a dead-end state, regardless of the actions taken afterward, it will continue in a dead-end trajectory until the agent reaches a termination state or maximum turn. We propose a dead-end resurrection (DDR) algorithm that detects the initial dead-end state in a timely and efficient manner and provides a rescue action to guide and correct the exploration direction. To prevent dialogue policies from repeatedly making the same mistake, DDR also performs dialogue data augmentation by adding relevant experiences containing dead-end states. We first validate the dead-end detection reliability and then demonstrate the effectiveness and generality of the method by reporting experimental results on several dialogue datasets from different domains. | ['Shihan Wang', 'Mehdi Dastani', 'Zhenyu Wang', 'Yangyang Zhao'] | 2023-05-05 | null | null | null | null | ['efficient-exploration'] | ['methodology'] | [ 3.12889507e-03 6.05429709e-01 -2.22043931e-01 -4.11483914e-01
-3.76834542e-01 -7.62516856e-01 7.00877607e-01 1.03829745e-02
-4.97754395e-01 1.28151572e+00 5.02412677e-01 -4.93915439e-01
1.05361760e-01 -7.26679027e-01 -3.69314283e-01 -6.10372663e-01
-7.20396563e-02 6.59392536e-01 8.09285417e-03 -5.60212314e-01
3.67401779e-01 2.89419711e-01 -1.27611256e+00 1.62093297e-01
9.27591085e-01 3.46143067e-01 9.93733332e-02 6.40209734e-01
-3.76036942e-01 1.16095746e+00 -1.05567777e+00 7.22343251e-02
1.92449227e-01 -7.79482722e-01 -1.29589534e+00 6.17375858e-02
-4.27557141e-01 -9.73877251e-01 -2.08868682e-01 1.02495074e+00
4.48490411e-01 5.64927995e-01 1.81305736e-01 -1.17900360e+00
-1.04241604e-02 7.31772006e-01 -1.97514027e-01 3.09850574e-02
6.04134679e-01 5.04235685e-01 6.86571002e-01 -4.22585309e-01
6.25367939e-01 1.31152332e+00 4.39220637e-01 1.18813241e+00
-8.62030029e-01 -4.94908154e-01 4.17553097e-01 -1.22247577e-01
-5.54514885e-01 -6.09590173e-01 8.19217801e-01 -4.47211005e-02
1.03748763e+00 2.93370754e-01 7.24784195e-01 1.33346128e+00
4.59834263e-02 8.84438217e-01 1.19362104e+00 -4.76272017e-01
6.27117395e-01 1.06013231e-01 -6.36488125e-02 7.70189464e-01
-2.72533506e-01 2.92113125e-01 -5.75044811e-01 -2.69149393e-01
6.46432877e-01 -3.88052762e-01 -1.73366666e-01 -1.20235443e-01
-8.82920265e-01 9.26158011e-01 1.11945514e-02 2.13575348e-01
-4.46021080e-01 -2.88377166e-01 7.05683172e-01 8.07643771e-01
7.80044124e-02 9.31380689e-01 -5.65612912e-01 -8.51160288e-01
-3.47286761e-01 2.73666382e-01 1.17650580e+00 6.46707892e-01
4.94530976e-01 3.86485644e-02 -2.62623876e-01 8.50309193e-01
5.51847275e-03 -4.77141440e-02 6.97532892e-01 -1.29882193e+00
2.98875630e-01 8.51642489e-01 7.18139470e-01 -5.19359052e-01
-6.20281816e-01 -2.50174016e-01 -4.90181208e-01 5.76502919e-01
6.28942490e-01 -6.80316329e-01 -4.17792022e-01 2.02998662e+00
6.40263259e-01 -3.11389387e-01 7.11145282e-01 9.21490610e-01
6.14717484e-01 7.08230913e-01 7.34503269e-02 -6.50091231e-01
1.06522477e+00 -1.08486581e+00 -1.02117634e+00 -3.77438039e-01
9.88854647e-01 -2.96898723e-01 1.24143350e+00 5.13664722e-01
-9.98877764e-01 -2.79958457e-01 -9.23151672e-01 3.11880767e-01
-5.23061398e-03 -1.15313418e-01 6.06118619e-01 1.86868340e-01
-8.14552128e-01 8.14686358e-01 -6.24944091e-01 -3.51168424e-01
-2.38212064e-01 3.81067470e-02 -1.83743775e-01 5.24944127e-01
-1.52492452e+00 1.04547846e+00 5.89929163e-01 8.26838762e-02
-1.02316201e+00 -1.87080979e-01 -6.39682770e-01 -1.40574589e-01
7.06782103e-01 -2.32829288e-01 1.80415821e+00 -1.07653618e+00
-2.11564374e+00 5.47321916e-01 2.50767265e-02 -5.68376958e-01
9.75369990e-01 -4.30740625e-01 -1.99549451e-01 -1.99846253e-01
-7.81172961e-02 4.32737887e-01 5.43713927e-01 -1.27200186e+00
-7.34558105e-01 -3.41897100e-01 3.93414557e-01 7.63831377e-01
-2.61291981e-01 -3.19735020e-01 -1.24536373e-01 -2.11148068e-01
4.79274355e-02 -8.93971026e-01 -2.85106391e-01 -4.63499576e-01
-6.29855812e-01 -4.53395575e-01 7.74632275e-01 -4.05155540e-01
1.47099066e+00 -1.90020418e+00 6.71704859e-02 1.34081841e-02
2.19569564e-01 4.09113914e-01 -3.04753423e-01 6.72268212e-01
9.56030488e-02 -9.45063308e-02 -1.84449807e-01 -1.67989969e-01
-1.25515625e-01 3.90720695e-01 -5.15162528e-01 1.82105079e-01
-2.24991009e-01 5.97903132e-01 -1.35935092e+00 -1.96266741e-01
2.84488291e-01 -9.99015719e-02 -6.19262636e-01 7.92440474e-01
-4.56496239e-01 8.53419662e-01 -5.88000834e-01 2.30060220e-01
2.19922453e-01 7.24457437e-03 4.66532528e-01 2.63379008e-01
-4.13827926e-01 8.01416039e-01 -9.15473700e-01 1.25850248e+00
-4.91844565e-01 4.92151439e-01 1.68087021e-01 -6.87976718e-01
9.76684749e-01 2.37284496e-01 2.82498658e-01 -8.97303760e-01
1.55229300e-01 9.81769189e-02 8.45963135e-02 -6.68003976e-01
7.63076544e-01 1.29632667e-01 -7.36131147e-02 8.41340542e-01
-3.17639619e-01 2.03537866e-01 1.50549039e-01 7.26804808e-02
1.13436580e+00 3.28058690e-01 4.80836242e-01 6.58757165e-02
4.97058183e-01 3.37855145e-02 7.50094056e-01 9.68513548e-01
-6.15232944e-01 -1.19907863e-01 8.64592731e-01 -5.08752763e-01
-9.27727818e-01 -7.12114513e-01 4.53567654e-01 1.45900786e+00
1.26779974e-01 -2.44804218e-01 -7.29094505e-01 -1.06993639e+00
-2.54982024e-01 1.27236581e+00 -7.08671272e-01 -6.21764004e-01
-6.87166214e-01 -4.81684566e-01 5.33130348e-01 2.92383969e-01
9.73027349e-01 -1.59716415e+00 -9.95897889e-01 3.52840960e-01
-6.19229555e-01 -6.79560840e-01 -1.68833405e-01 3.78755420e-01
-8.39388549e-01 -1.03386903e+00 -4.45869118e-01 -5.16191185e-01
6.11364424e-01 -5.25082350e-02 8.62685859e-01 1.93613797e-01
3.35841775e-01 -8.09615571e-03 -5.04047692e-01 -3.41300517e-02
-1.17027247e+00 -4.77722920e-02 2.28441864e-01 -4.25453424e-01
3.23950022e-01 -3.44117671e-01 -5.18644929e-01 3.37692827e-01
-3.68952751e-01 1.79274917e-01 1.66060582e-01 9.57926214e-01
-1.70312151e-02 -7.78765157e-02 1.02864218e+00 -8.68740857e-01
1.42077041e+00 -3.28082263e-01 -4.41355735e-01 1.20594777e-01
-7.62106478e-01 3.99218440e-01 7.98811734e-01 -3.65664065e-01
-1.23076391e+00 -3.30110103e-01 -3.14180791e-01 -1.12475894e-01
-1.92911088e-01 2.69430757e-01 2.99580768e-02 4.74578977e-01
7.29391396e-01 5.86108983e-01 3.28217208e-01 -3.65321577e-01
2.49753654e-01 7.78034508e-01 4.22535598e-01 -5.52623808e-01
2.52466470e-01 -5.98134473e-03 -5.55913329e-01 -6.49256766e-01
-7.39059150e-01 -8.87417868e-02 -1.95818007e-01 -5.12322545e-01
3.98042887e-01 -5.68619668e-01 -1.16727114e+00 4.68379855e-01
-1.03568935e+00 -8.68114173e-01 -3.91984910e-01 3.18065882e-02
-5.63969195e-01 6.03577375e-01 -5.85872412e-01 -1.33161473e+00
-4.31821734e-01 -9.32450533e-01 3.24800670e-01 5.09659111e-01
-6.71054661e-01 -7.84154952e-01 2.88936615e-01 8.96919444e-02
1.69302776e-01 -6.06108122e-02 8.17022920e-01 -1.04319584e+00
8.80248398e-02 1.58590093e-01 2.26932764e-01 1.64312810e-01
2.62087524e-01 -1.12793311e-01 -7.41157711e-01 -3.40508908e-01
1.74530391e-02 -7.12614834e-01 6.01454318e-01 -1.40438244e-01
8.11178446e-01 -8.79370689e-01 -5.65850101e-02 2.20384479e-01
7.15255082e-01 7.82718301e-01 4.33535993e-01 8.27749789e-01
1.00735545e-01 7.77009666e-01 1.04145896e+00 9.68396246e-01
4.35338378e-01 4.38608766e-01 7.25194752e-01 -4.50709043e-03
1.67744815e-01 -5.15069067e-01 6.11584246e-01 4.60599840e-01
2.03069210e-01 -3.56163263e-01 -6.55193627e-01 5.33516109e-01
-1.94337642e+00 -9.04914439e-01 4.09668922e-01 2.25392509e+00
1.48910153e+00 3.42074275e-01 2.31226832e-01 1.09622315e-01
5.38967550e-01 3.29601377e-01 -9.01643693e-01 -9.31972146e-01
1.49073452e-01 -4.41756815e-01 2.46408321e-02 8.54588389e-01
-7.87687778e-01 1.33758759e+00 6.12342739e+00 1.72348261e-01
-1.02966046e+00 -2.23785594e-01 5.15016556e-01 1.34946600e-01
-1.65929466e-01 -4.65572700e-02 -5.74667215e-01 3.71322751e-01
6.04587972e-01 1.10231146e-01 7.46602595e-01 8.24946702e-01
4.02577490e-01 -4.31245834e-01 -1.04349542e+00 4.35397834e-01
-4.18876320e-01 -9.09182429e-01 -1.92593038e-01 -3.67063135e-01
3.14852893e-01 -4.33804244e-01 -2.37712964e-01 6.69630706e-01
8.32497835e-01 -5.31144321e-01 4.93523896e-01 1.67577729e-01
6.02612615e-01 -1.03160465e+00 3.61965150e-01 9.34713900e-01
-4.00077015e-01 -4.02640641e-01 -4.27770875e-02 -4.03210461e-01
-9.70572531e-02 3.33765298e-02 -1.48115802e+00 1.02258541e-01
3.01067472e-01 9.31674987e-02 -4.22348594e-03 4.37300354e-01
-7.20940948e-01 4.67103183e-01 -6.63880482e-02 -4.25742626e-01
5.10286987e-01 -2.07960427e-01 9.30697620e-01 9.18745518e-01
-1.43659115e-01 2.80891836e-01 1.49061158e-01 8.19833875e-01
3.34531367e-02 -2.29045182e-01 -5.61950028e-01 -1.42363757e-01
8.75460029e-01 1.02902281e+00 -3.98223728e-01 -2.66635776e-01
9.52433571e-02 1.10799992e+00 4.89529699e-01 3.90503258e-01
-6.99095786e-01 -3.64318848e-01 6.03449821e-01 -4.28767949e-01
-2.73707360e-01 -1.21505791e-02 -1.85343608e-01 -9.09287632e-01
-1.06636040e-01 -1.01654792e+00 9.76612270e-02 -5.72823763e-01
-8.10208142e-01 7.68764198e-01 -2.75075167e-01 -1.04620051e+00
-8.23468983e-01 9.47292000e-02 -7.33163536e-01 4.86841917e-01
-1.31926394e+00 -2.39541307e-01 -2.05578044e-01 2.56158143e-01
9.02234018e-01 -2.39933938e-01 1.12077796e+00 -2.86891818e-01
-6.69955790e-01 6.57402694e-01 4.56611551e-02 2.90379733e-01
6.80865228e-01 -1.34348428e+00 3.20553958e-01 4.91307110e-01
-5.59978604e-01 6.37158036e-01 1.17638290e+00 -8.61367583e-01
-1.06432652e+00 -8.41196716e-01 7.87075400e-01 9.08184424e-02
3.09633195e-01 -1.09558880e-01 -1.00363517e+00 6.69031322e-01
2.29299203e-01 -7.70334482e-01 3.49442422e-01 1.66704029e-01
-5.86238056e-02 2.64389694e-01 -1.24226820e+00 1.09646475e+00
8.08607221e-01 -2.16105148e-01 -9.62549388e-01 2.88666189e-01
5.83424687e-01 -6.06350720e-01 -2.91179240e-01 4.54327345e-01
4.98572022e-01 -1.24907815e+00 4.38726395e-01 -8.13168645e-01
4.67058212e-01 6.93978593e-02 3.72643977e-01 -1.60732532e+00
-6.69525191e-03 -1.25022566e+00 -4.01806474e-01 9.92811322e-01
2.01120660e-01 -5.80950022e-01 7.89970934e-01 6.73059583e-01
-2.98678428e-02 -9.86004770e-01 -7.91951418e-01 -5.28992176e-01
-9.50463116e-02 6.92664832e-03 5.91698110e-01 1.07979989e+00
5.52916408e-01 3.88479859e-01 -6.80621266e-01 -1.66108608e-01
3.54678094e-01 2.89912105e-01 9.29935277e-01 -7.88568735e-01
-2.68215723e-02 -3.67183656e-01 5.55190861e-01 -1.36106730e+00
2.66344279e-01 -2.24788144e-01 3.61222208e-01 -1.47818601e+00
-1.72111586e-01 -4.87474531e-01 -4.31543551e-02 6.48086548e-01
-2.01959714e-01 -5.49471498e-01 6.51115626e-02 1.51211306e-01
-7.83449829e-01 8.26462030e-01 1.38451791e+00 2.82584310e-01
-8.93927574e-01 1.96033001e-01 -5.61415851e-01 9.04930532e-01
1.17308640e+00 -1.31685928e-01 -5.54551065e-01 5.57163730e-02
1.43168882e-01 6.45434320e-01 -1.93801727e-02 -6.28132522e-01
1.27733439e-01 -4.59444821e-01 1.15954228e-01 -3.92914385e-01
3.30675185e-01 -5.11740208e-01 -4.56698060e-01 7.61317670e-01
-1.03503084e+00 3.64013389e-02 1.94518104e-01 4.99688804e-01
1.18833944e-01 -4.78792846e-01 8.52882922e-01 -2.88825095e-01
-8.16449642e-01 -3.58315349e-01 -8.97693813e-01 1.60913244e-01
1.15160632e+00 -4.39197272e-01 -2.06589952e-01 -8.00988197e-01
-8.26466620e-01 8.02886307e-01 4.29869413e-01 5.66110969e-01
6.99063838e-01 -9.19577360e-01 -4.25945312e-01 3.90619606e-01
-1.35919124e-01 -1.76235259e-01 4.28393856e-02 3.34821254e-01
-2.60992706e-01 1.80645913e-01 -2.52671778e-01 -1.65392116e-01
-1.31008244e+00 3.18277031e-01 7.43364513e-01 -4.84172463e-01
-8.48846376e-01 7.24963725e-01 -8.35561156e-02 -6.01722658e-01
8.01508844e-01 -4.16109599e-02 -6.09035313e-01 5.96501417e-02
5.38262784e-01 4.28160161e-01 -1.70234755e-01 1.45764789e-03
-1.61220387e-01 -2.76709855e-01 -3.63517165e-01 -4.46899414e-01
1.03860939e+00 -3.88028353e-01 1.36351541e-01 5.69152772e-01
6.36242211e-01 -2.25441247e-01 -1.63356030e+00 -9.84805077e-02
-7.31483251e-02 -3.65091920e-01 -2.64087468e-01 -1.41812360e+00
-5.86078942e-01 4.93064195e-01 3.42856079e-01 5.13983071e-01
8.58247042e-01 -3.33913207e-01 7.72261858e-01 8.87777865e-01
3.94890547e-01 -1.60808432e+00 3.24467629e-01 1.05930316e+00
1.10150433e+00 -1.16958034e+00 -2.05137327e-01 1.68264061e-01
-1.21218884e+00 1.03626204e+00 1.20727670e+00 2.28851840e-01
-3.46946381e-02 2.87930369e-01 2.45365188e-01 -4.15423932e-03
-1.19242013e+00 8.20137188e-02 -6.57983959e-01 5.57218254e-01
2.39249736e-01 -7.93327242e-02 -4.80977386e-01 3.45319837e-01
-2.86546916e-01 -8.67307037e-02 8.34627986e-01 1.24671328e+00
-8.04305553e-01 -9.66942012e-01 -1.55081794e-01 1.10495940e-01
-1.49793848e-01 3.06531578e-01 -8.19041669e-01 7.43438542e-01
-5.14769077e-01 1.05014026e+00 1.03525095e-01 -4.26961154e-01
3.38717580e-01 2.90863037e-01 2.00984716e-01 -3.74255687e-01
-7.75941849e-01 -9.06638801e-02 5.91745019e-01 -7.47612417e-01
6.47933409e-02 -3.76763552e-01 -2.01555300e+00 -3.58960122e-01
-2.63142526e-01 4.90270823e-01 3.31061065e-01 8.55736494e-01
3.32863927e-01 6.71118498e-01 1.01004159e+00 -2.71049827e-01
-1.25386441e+00 -1.12934721e+00 -1.78631410e-01 2.82896131e-01
7.56078005e-01 -7.39900947e-01 -5.02958357e-01 -5.53117454e-01] | [13.050880432128906, 8.04122543334961] |
a534004a-7c78-4699-b802-049f231bd0c9 | learn-from-each-other-to-classify-better | null | null | https://www.sciencedirect.com/science/article/pii/S0031320323002509 | https://reader.elsevier.com/reader/sd/pii/S0031320323002509?token=58FF76D12E1144ADB66AD959AB1396F4468B3FE8F47CB2AB90846908B89FA214B4EC69F4E1AD384E9CF4F40D51740518&originRegion=us-east-1&originCreation=20230416125734 | Learn from Each Other to Classify Better: Cross-layer Mutual Attention Learning for Fine-grained Visual Classification | Fine-grained visual classification (FGVC) is valuable yet challenging. The difficulty of FGVC mainly lies in its intrinsic inter-class similarity, intra-class variation, and limited training data. Moreover, with the popularity of deep convolutional neural networks, researchers have mainly used deep, abstract, semantic information for FGVC, while shallow, detailed information has been neglected. This work proposes a cross-layer mutual attention learning network (CMAL-Net) to solve the above problems. Specifically, this work views the shallow to deep layers of CNNs as “experts” knowledgeable about different perspectives. We let each expert give a category prediction and an attention region indicating the found clues. Attention regions are treated as information carriers among experts, bringing three benefits: (ⅰ) helping the model focus on discriminative regions; (ⅱ) providing more training data; (ⅲ) allowing experts to learn from each other to improve the overall performance. CMAL-Net achieves state-of-the-art performance on three competitive datasets: FGVC-Aircraft, Stanford Cars, and Food-11. | ['Jien Kato', 'Yu Wang', 'Longjiao Zhao', 'Dichao Liu'] | 2023-03-22 | null | null | null | pattern-recognition-2023-3 | ['fine-grained-image-classification'] | ['computer-vision'] | [-3.94702643e-01 -1.25888884e-01 -3.60805064e-01 -4.37907308e-01
-3.45325530e-01 -4.42368120e-01 2.31470257e-01 9.53573287e-02
-2.81263232e-01 3.80918056e-01 3.48685145e-01 -6.57332316e-02
1.27444208e-01 -7.33694673e-01 -6.55989408e-01 -4.63480830e-01
2.72322029e-01 1.27303138e-01 5.26398003e-01 6.66265562e-03
1.76753595e-01 3.29771906e-01 -1.42853832e+00 7.90576041e-01
1.01418459e+00 1.49967504e+00 4.87685323e-01 2.52018631e-01
-2.04202071e-01 1.33732641e+00 -4.42691118e-01 -4.93822694e-01
-7.85136148e-02 -1.43989697e-01 -9.50827897e-01 7.93660283e-02
1.95721179e-01 -3.88333559e-01 -2.38941282e-01 1.10150802e+00
3.01055104e-01 1.35131553e-01 7.12704599e-01 -1.31039608e+00
-1.24092972e+00 3.58592421e-01 -7.01517582e-01 2.29782030e-01
-5.73087111e-02 2.79630005e-01 1.30303454e+00 -1.01455021e+00
2.90620714e-01 1.28997076e+00 5.70282936e-01 5.15146494e-01
-8.89397323e-01 -7.63711095e-01 6.92726016e-01 5.80893099e-01
-1.35958374e+00 -1.22761331e-01 8.60034108e-01 -7.02652633e-01
8.45575035e-01 5.44694141e-02 5.98317742e-01 1.07703245e+00
-1.35142431e-01 1.20348871e+00 6.67723119e-01 -1.14397980e-01
1.34564908e-02 3.14804643e-01 2.05450714e-01 7.46697009e-01
5.50211631e-02 7.38394484e-02 -1.81528419e-01 2.26686344e-01
8.38254452e-01 4.11766469e-01 -4.25347865e-01 -2.91414887e-01
-9.29464936e-01 9.55224991e-01 1.09721279e+00 4.47553009e-01
-3.56421560e-01 9.58092883e-02 4.07979816e-01 1.48199677e-01
6.01949215e-01 4.47599888e-01 -6.21201396e-01 2.96588242e-01
-5.20522594e-01 7.01559195e-03 4.78512287e-01 9.13663805e-01
7.71245837e-01 3.81621718e-02 -4.84007329e-01 9.07804012e-01
3.26088905e-01 1.10394567e-01 4.68175739e-01 -7.49837399e-01
5.65550029e-01 8.78870308e-01 4.71739471e-02 -1.38450813e+00
-2.16668472e-01 -9.56585884e-01 -1.01025403e+00 2.74032559e-02
2.04080880e-01 8.68656263e-02 -8.25658143e-01 1.57064497e+00
-3.05075310e-02 9.18178111e-02 -2.11388692e-01 1.14115441e+00
1.37497127e+00 5.92630625e-01 3.24613750e-01 3.28203589e-01
1.33312654e+00 -1.46523058e+00 -5.22424757e-01 -5.17935216e-01
2.03630418e-01 -6.73615754e-01 1.01038742e+00 3.80271226e-01
-6.71729803e-01 -1.05303729e+00 -8.75167608e-01 -2.28431597e-01
-4.76498783e-01 5.00017703e-01 6.22916162e-01 6.09356090e-02
-9.20004606e-01 4.66489196e-01 -4.68809426e-01 -2.99382536e-03
9.82679784e-01 6.12228736e-02 -2.35794798e-01 -3.76713455e-01
-1.14756167e+00 6.75459087e-01 3.74363512e-01 1.51406690e-01
-1.04930007e+00 -7.96145439e-01 -6.38873458e-01 3.42469931e-01
4.18653101e-01 -5.52891791e-01 1.06829512e+00 -1.39708018e+00
-1.06541562e+00 8.79568577e-01 -3.82745489e-02 -2.13561177e-01
5.52169502e-01 -3.76917779e-01 -4.27402765e-01 1.60254359e-01
3.21479380e-01 8.38543892e-01 7.85990238e-01 -1.28985870e+00
-7.52960682e-01 -3.31228107e-01 2.21812278e-01 1.25418976e-02
-3.24118972e-01 -2.11193308e-01 -7.32448220e-01 -8.25873733e-01
-3.37433927e-02 -5.15366912e-01 -1.03221722e-01 9.73954424e-02
-4.22195673e-01 -7.43882895e-01 8.46976757e-01 -7.30861068e-01
1.15447974e+00 -2.37130475e+00 1.36467636e-01 -2.68195122e-01
5.40031731e-01 4.97850925e-01 -2.35829040e-01 1.04852021e-01
1.66344773e-02 9.70093682e-02 2.46661872e-01 -1.51116252e-01
-1.39629856e-01 5.93278110e-02 -1.98944435e-01 1.68951333e-01
6.05292678e-01 1.25267780e+00 -9.97153103e-01 -3.67465317e-01
1.65186986e-01 3.71880800e-01 -5.48562467e-01 3.68235767e-01
-2.91436046e-01 4.17375118e-01 -6.61373198e-01 8.34973752e-01
7.06447124e-01 -7.48845339e-01 -1.73825204e-01 -7.66626716e-01
-2.15827916e-02 -2.64131248e-01 -7.08815038e-01 1.49961591e+00
-3.35613579e-01 6.10392690e-01 7.56503493e-02 -1.09606874e+00
8.61689150e-01 1.44409329e-01 3.65915708e-02 -8.66346300e-01
2.68842489e-01 2.99273692e-02 -8.19292367e-02 -6.56314850e-01
3.20891768e-01 8.17619935e-02 1.23196252e-01 8.47698227e-02
2.43571594e-01 4.48476911e-01 -1.08356744e-01 1.23759076e-01
6.53939307e-01 -7.66967386e-02 2.79554665e-01 -3.20437849e-01
4.03964192e-01 -1.48366988e-01 7.54256725e-01 3.99664551e-01
-6.60097778e-01 5.32354653e-01 4.03949052e-01 -6.49005771e-01
-5.77537775e-01 -8.37956548e-01 2.05766737e-01 1.33697402e+00
4.18788463e-01 -3.10785800e-01 -4.01139706e-01 -9.61714029e-01
2.27989301e-01 5.51166117e-01 -1.11676371e+00 -3.71510118e-01
-3.99269313e-01 -2.01316312e-01 2.62623072e-01 1.06180990e+00
7.50271678e-01 -1.21146560e+00 -4.77117121e-01 1.52012244e-01
-3.82796705e-01 -9.71935689e-01 -4.72350061e-01 2.63087451e-01
-6.73517883e-01 -1.32341015e+00 -9.38037753e-01 -9.16699767e-01
6.50728703e-01 6.30189180e-01 1.47776484e+00 2.30571166e-01
-3.28065902e-01 4.33114804e-02 -4.82396513e-01 -3.77390832e-01
2.21140578e-01 -4.64488789e-02 -4.41133946e-01 1.33906186e-01
5.58177590e-01 -8.78002569e-02 -9.18798864e-01 6.35350347e-01
-4.12397385e-01 6.97424589e-03 6.53367400e-01 1.03223002e+00
6.06733024e-01 1.39615819e-01 5.63365757e-01 -8.42638910e-01
3.15338343e-01 -5.98811328e-01 -4.05804485e-01 2.57441342e-01
-4.06653881e-01 -2.08688438e-01 7.23334968e-01 -2.85582542e-01
-1.00187492e+00 -2.77957261e-01 -1.54605180e-01 -8.98565173e-01
-2.95255661e-01 3.30759227e-01 -3.29313815e-01 1.15343025e-02
3.85378450e-01 1.51047066e-01 -3.17110509e-01 -6.50069714e-01
3.22537988e-01 4.66798812e-01 4.32019383e-01 -4.49556231e-01
3.01409453e-01 1.46769658e-01 -5.65993190e-01 -2.81758457e-01
-1.35133004e+00 -6.54493511e-01 -6.60812497e-01 -1.14407472e-01
1.17538917e+00 -1.25366402e+00 -6.67412102e-01 4.35940325e-01
-1.14276898e+00 -2.85155237e-01 -1.14602581e-01 1.62276313e-01
2.55159177e-02 7.20156133e-02 -5.71117640e-01 -4.50427502e-01
-1.53293550e-01 -1.28995824e+00 1.00250971e+00 4.21942353e-01
1.48874633e-02 -1.11277258e+00 -4.20665026e-01 5.16888857e-01
5.87962449e-01 1.03248037e-01 9.40868378e-01 -5.44184029e-01
-6.74696803e-01 -1.50062531e-01 -9.50308442e-01 5.64151049e-01
2.63123810e-01 -1.39615005e-02 -1.19852066e+00 -2.52005428e-01
-4.34220552e-01 -6.07275307e-01 1.27062500e+00 4.82090950e-01
1.76524842e+00 -2.85060614e-01 -5.11257529e-01 6.27369761e-01
1.29480672e+00 1.79648757e-01 3.13887507e-01 1.62861928e-01
1.02808917e+00 6.86613083e-01 6.00800216e-01 2.89965570e-01
5.73499560e-01 5.36996722e-01 8.35735321e-01 -4.00916696e-01
-3.43134135e-01 -2.98106760e-01 -1.20004363e-01 6.94940150e-01
-1.24452278e-01 -1.98412299e-01 -5.93155563e-01 7.85546839e-01
-1.96437311e+00 -9.18036759e-01 -1.68091636e-02 1.61012721e+00
5.19131720e-01 -3.65056507e-02 4.99535576e-02 -9.84374732e-02
7.20719993e-01 2.55284309e-01 -7.76615202e-01 6.97973669e-02
-1.45541439e-02 -8.42167586e-02 1.54637441e-01 -2.33357586e-02
-1.51909745e+00 9.70700622e-01 5.22386789e+00 1.15149140e+00
-1.22301996e+00 1.85361907e-01 9.99878228e-01 1.89710125e-01
-1.88410088e-01 -4.83232677e-01 -6.88337564e-01 5.78612506e-01
1.80006996e-01 2.63024598e-01 1.30166635e-01 1.14849734e+00
-2.91762859e-01 1.70669690e-01 -1.00912571e+00 1.14305878e+00
2.38552839e-02 -1.40399420e+00 2.16958493e-01 -1.18050791e-01
8.07923734e-01 9.37922895e-02 1.70058966e-01 5.25999606e-01
4.19208199e-01 -1.18542826e+00 8.74339759e-01 4.54416633e-01
8.58930171e-01 -1.00009418e+00 9.81094301e-01 3.33080232e-01
-1.49312067e+00 -3.91630173e-01 -6.99321985e-01 1.27229765e-01
-2.13766858e-01 6.78578556e-01 -3.44031453e-01 6.65939450e-01
1.02631593e+00 1.14788043e+00 -6.40843093e-01 9.93604898e-01
-4.31937337e-01 5.85263550e-01 3.35814327e-01 -2.47955825e-02
5.75233281e-01 1.37105450e-01 -9.04937834e-02 1.18733370e+00
9.67693403e-02 1.00500241e-01 5.41809559e-01 9.45142329e-01
-2.43505672e-01 -1.06982894e-01 -1.48412958e-01 -1.14625610e-01
3.54054987e-01 1.26461852e+00 -8.42984140e-01 -3.18862140e-01
-5.47575176e-01 1.04099107e+00 6.48338616e-01 3.71745914e-01
-6.57926381e-01 -4.34735656e-01 7.84572184e-01 -8.75012353e-02
7.59342372e-01 2.96730161e-01 -3.30037773e-01 -1.13488770e+00
-3.04512560e-01 -7.74515510e-01 5.02954543e-01 -7.77434886e-01
-1.73519135e+00 8.43343258e-01 -5.38389981e-01 -1.27037263e+00
2.40296274e-01 -7.98091829e-01 -6.64115965e-01 9.74385858e-01
-1.58913004e+00 -1.26210308e+00 -6.82596743e-01 6.27129138e-01
8.62013638e-01 -3.58430356e-01 6.05828762e-01 4.75481361e-01
-5.83149374e-01 7.32067525e-01 -2.05203533e-01 5.32244503e-01
6.17212296e-01 -1.15310657e+00 1.58677608e-01 6.33583844e-01
-7.59947449e-02 5.51176012e-01 2.13025153e-01 -4.60830688e-01
-9.48385417e-01 -1.54518330e+00 6.39345884e-01 -4.47142154e-01
5.24783134e-01 -2.70593017e-01 -9.76051629e-01 5.23982108e-01
2.38541871e-01 4.69427586e-01 7.12545097e-01 2.32045069e-01
-6.64241016e-01 -2.74739593e-01 -1.02705443e+00 1.82805896e-01
9.66651678e-01 -6.33894801e-01 -4.92931634e-01 2.05981910e-01
1.05726445e+00 -1.00180082e-01 -4.92118657e-01 3.32320303e-01
5.88088691e-01 -1.03912473e+00 1.05934596e+00 -7.61709392e-01
7.41196156e-01 -3.12536091e-01 -2.37700611e-01 -1.48176610e+00
-9.79383171e-01 2.30588630e-01 -1.90005913e-01 1.22490084e+00
2.21729547e-01 -5.15612304e-01 6.41408026e-01 2.56442964e-01
-3.28997403e-01 -1.00590730e+00 -4.13495481e-01 -4.83307779e-01
-3.75789707e-03 -3.94816428e-01 6.90391660e-01 1.13841867e+00
-5.28261423e-01 4.27650452e-01 -5.24202228e-01 2.99978480e-02
3.55309278e-01 4.51216161e-01 4.57531512e-01 -1.51528764e+00
-3.96623522e-01 -5.87401927e-01 -2.09711909e-01 -1.22884202e+00
5.69090880e-02 -8.13916087e-01 -6.52667060e-02 -1.62423611e+00
4.90051329e-01 -4.83568519e-01 -6.61625922e-01 6.54166102e-01
-4.06350881e-01 2.99091280e-01 3.16541284e-01 2.49562681e-01
-9.90856588e-01 6.29824996e-01 1.60956359e+00 -2.96856612e-01
2.23768070e-01 1.34479105e-01 -1.14280653e+00 7.87509859e-01
5.14137745e-01 -1.58695132e-01 -5.18250644e-01 -7.90113628e-01
-7.22256256e-03 -1.83797240e-01 5.61316192e-01 -8.66040945e-01
1.36874303e-01 -1.59204088e-03 9.27479804e-01 -8.84719431e-01
2.56466456e-02 -8.34108233e-01 -2.06683531e-01 4.88005430e-01
-3.12877327e-01 -1.22665562e-01 1.21556498e-01 7.28451788e-01
-6.16243005e-01 -1.95359364e-02 8.35749507e-01 -4.80517566e-01
-1.27143419e+00 6.16170645e-01 3.35947052e-02 1.60968035e-01
9.16927576e-01 5.54751307e-02 -4.41266298e-01 -1.88139647e-01
-6.72981322e-01 4.22525615e-01 2.90024906e-01 5.54144621e-01
6.30411506e-01 -1.41091597e+00 -4.83373851e-01 2.03008428e-01
4.17712986e-01 8.67375731e-02 6.36609316e-01 5.95095158e-01
-2.76240110e-01 6.83348775e-01 -3.17893088e-01 -6.84483588e-01
-9.74677682e-01 8.70754004e-01 1.62769929e-01 -3.08836579e-01
-2.27718160e-01 1.52997327e+00 8.70491326e-01 -3.36101472e-01
4.23261672e-01 -3.28527778e-01 -5.86893439e-01 3.59557331e-01
6.46682501e-01 3.29501569e-01 -6.09271526e-02 -6.10688746e-01
-5.38660467e-01 6.60911620e-01 -3.05002719e-01 7.80762970e-01
1.17976034e+00 -7.55724609e-02 8.20916072e-02 1.98635265e-01
1.34670556e+00 -2.45826289e-01 -1.58371496e+00 -4.87990737e-01
-2.75991291e-01 -4.39921886e-01 1.97617754e-01 -1.21262276e+00
-1.55347097e+00 1.37613547e+00 5.84147453e-01 5.54156229e-02
1.20620918e+00 2.05548078e-01 6.55098021e-01 7.30621666e-02
3.22018772e-01 -8.58804762e-01 4.87952709e-01 5.30592084e-01
1.00305533e+00 -1.55772448e+00 -1.94627702e-01 -4.58941519e-01
-8.69492888e-01 9.85170543e-01 1.00483811e+00 -2.16022551e-01
7.87211120e-01 -1.57824963e-01 7.30636790e-02 -2.47952819e-01
-7.78211951e-01 -3.29253912e-01 7.57824183e-01 6.42493606e-01
5.18213093e-01 2.20069349e-01 2.81499803e-01 1.18888474e+00
2.49498531e-01 8.31926242e-02 -2.81519771e-01 4.97423500e-01
-5.77416182e-01 -6.21463597e-01 3.54630314e-02 4.37977076e-01
-4.92171824e-01 -1.43350005e-01 -3.71297419e-01 7.01972723e-01
6.31236315e-01 1.02482188e+00 1.58608079e-01 -5.19056320e-01
2.94332623e-01 -3.81203771e-01 1.26264766e-01 -5.40381670e-01
-6.28458798e-01 1.22531988e-01 -1.33698002e-01 -7.49359131e-01
-2.75053859e-01 -2.44137406e-01 -8.22693586e-01 -9.07690749e-02
-1.81234419e-01 2.78949648e-01 1.54687613e-01 9.46881294e-01
5.12079775e-01 1.06964958e+00 6.40696704e-01 -6.65371120e-01
-3.07617188e-01 -9.16397333e-01 -4.05458778e-01 3.30079019e-01
3.86654019e-01 -9.27954018e-01 -1.06023252e-01 -1.56575605e-01] | [9.537590980529785, 1.8556393384933472] |
dbd8e2f3-b710-4ace-99c1-c24e031d5a9f | toward-super-resolution-for-appearance-based | 2303.10151 | null | https://arxiv.org/abs/2303.10151v1 | https://arxiv.org/pdf/2303.10151v1.pdf | Toward Super-Resolution for Appearance-Based Gaze Estimation | Gaze tracking is a valuable tool with a broad range of applications in various fields, including medicine, psychology, virtual reality, marketing, and safety. Therefore, it is essential to have gaze tracking software that is cost-efficient and high-performing. Accurately predicting gaze remains a difficult task, particularly in real-world situations where images are affected by motion blur, video compression, and noise. Super-resolution has been shown to improve image quality from a visual perspective. This work examines the usefulness of super-resolution for improving appearance-based gaze tracking. We show that not all SR models preserve the gaze direction. We propose a two-step framework based on SwinIR super-resolution model. The proposed method consistently outperforms the state-of-the-art, particularly in scenarios involving low-resolution or degraded images. Furthermore, we examine the use of super-resolution through the lens of self-supervised learning for gaze prediction. Self-supervised learning aims to learn from unlabelled data to reduce the amount of required labeled data for downstream tasks. We propose a novel architecture called SuperVision by fusing an SR backbone network to a ResNet18 (with some skip connections). The proposed SuperVision method uses 5x less labeled data and yet outperforms, by 15%, the state-of-the-art method of GazeTR which uses 100% of training data. | ['Majid Komeili', "Galen O'Shea"] | 2023-03-17 | null | null | null | null | ['gaze-estimation', 'eye-tracking', 'marketing'] | ['computer-vision', 'computer-vision', 'miscellaneous'] | [ 6.50318146e-01 1.69040576e-01 -3.52197707e-01 -4.59489763e-01
-3.25707585e-01 -6.22076206e-02 2.33327389e-01 -2.91212082e-01
-5.50043166e-01 8.23038936e-01 1.47578955e-01 -1.85913831e-01
-9.90791097e-02 -2.94167191e-01 -8.37571859e-01 -6.08663976e-01
2.47471884e-01 -2.41742685e-01 4.56989735e-01 -3.00961018e-01
5.25540173e-01 2.94336081e-01 -2.29267049e+00 1.35838747e-01
9.79395509e-01 9.27390397e-01 4.65919018e-01 5.11403203e-01
1.22597381e-01 8.14747930e-01 -4.19176966e-01 -1.53669640e-01
1.23575166e-01 -4.28619057e-01 -6.27525449e-01 -5.27506247e-02
1.02174211e+00 -2.68062174e-01 7.28320628e-02 1.11148667e+00
3.82434309e-01 2.31589958e-01 2.35552624e-01 -1.09571087e+00
-6.95700109e-01 2.70902347e-02 -1.11377597e+00 7.63370931e-01
3.50743532e-01 -3.30504738e-02 6.65948987e-01 -6.75827146e-01
7.06726611e-01 1.02641487e+00 2.89923340e-01 8.62670362e-01
-1.24640143e+00 -1.05096245e+00 1.13903157e-01 4.78571981e-01
-1.16536319e+00 -7.66092539e-01 5.63762128e-01 -1.05701499e-01
6.60795331e-01 1.67976290e-01 1.17456451e-01 9.21353221e-01
2.77338862e-01 4.36230004e-01 1.47367823e+00 -5.97227871e-01
1.67429857e-02 3.95775110e-01 2.04615518e-01 5.89504302e-01
2.12227315e-01 2.38446593e-01 -1.12903547e+00 3.77739340e-01
6.74191117e-01 5.75494692e-02 -6.78664148e-01 -2.74031639e-01
-8.49308431e-01 4.46498185e-01 6.72055900e-01 2.78845072e-01
-3.51399153e-01 -1.01304442e-01 -1.90413222e-01 2.25832254e-01
8.85669351e-01 2.04468921e-01 -3.23602974e-01 -3.96644361e-02
-1.14403605e+00 -2.22618490e-01 1.37351930e-01 6.88673377e-01
5.68306923e-01 4.76013534e-02 2.21935194e-02 6.60748422e-01
5.40806651e-01 4.69755590e-01 5.38041115e-01 -1.02002549e+00
1.56248212e-01 4.39426780e-01 1.02615841e-01 -7.63456881e-01
-4.69175398e-01 -5.75324774e-01 -6.71924055e-01 6.77672267e-01
3.37997258e-01 1.25123188e-01 -9.13901210e-01 1.89267194e+00
4.13901955e-01 5.01066089e-01 -1.91599116e-01 1.14444327e+00
7.37416506e-01 3.03388953e-01 1.90114930e-01 -6.73574924e-01
1.20500457e+00 -8.80141020e-01 -9.22511399e-01 -1.45120665e-01
2.80717105e-01 -6.58876538e-01 1.07002854e+00 4.59859729e-01
-1.15670705e+00 -6.98583245e-01 -1.04162312e+00 -2.21658915e-01
-4.32999097e-02 -1.01512671e-01 2.58514345e-01 8.16033781e-01
-1.59676254e+00 5.43223619e-01 -6.79404736e-01 -6.31415904e-01
5.42400360e-01 7.28067756e-01 -3.88130367e-01 -5.92519045e-02
-7.71220446e-01 1.09626007e+00 -9.09073129e-02 -1.29536152e-01
-2.14756936e-01 -6.97653115e-01 -8.42430532e-01 1.18279591e-01
3.50495666e-01 -6.32081985e-01 1.05550563e+00 -1.21122980e+00
-1.53553247e+00 7.42895067e-01 -9.01226282e-01 -4.26180661e-01
1.25158235e-01 -4.71475244e-01 -5.75481355e-01 2.56275266e-01
-1.18063755e-01 8.23065758e-01 1.18914938e+00 -1.19051981e+00
-9.69416916e-01 -7.53531098e-01 7.02818856e-02 4.87759441e-01
-3.46337080e-01 2.09899768e-01 -2.90033877e-01 -1.94629282e-01
-7.13931844e-02 -9.34410393e-01 1.64045677e-01 4.81579155e-02
4.18860242e-02 -1.02835923e-01 1.03358424e+00 -6.86946034e-01
1.20718741e+00 -2.17682672e+00 2.92043891e-02 -5.99636175e-02
5.55404842e-01 3.94544750e-01 -1.82439834e-02 -3.73045772e-01
-3.67682844e-01 -4.11977395e-02 9.10878256e-02 -5.17259896e-01
-6.19987905e-01 -2.83613205e-01 -5.54462373e-02 6.10237062e-01
3.95274423e-02 6.20646715e-01 -7.49578774e-01 -4.92532313e-01
2.94689864e-01 8.93368721e-01 -2.04992622e-01 5.19459173e-02
4.46417183e-02 7.75971174e-01 1.24977805e-01 3.95132512e-01
6.69449031e-01 -6.61832273e-01 -9.39439163e-02 -1.99472874e-01
-2.94287205e-01 1.12842768e-01 -9.09598589e-01 1.66167808e+00
-4.10032302e-01 1.05159044e+00 -6.01884946e-02 -3.26501608e-01
6.66720867e-01 8.70412961e-02 3.02906096e-01 -1.17030621e+00
1.95504263e-01 -1.47011787e-01 -2.54373066e-02 -4.93715227e-01
7.46827960e-01 3.71626467e-02 6.79610908e-01 5.89159250e-01
-7.29739666e-02 7.67800689e-01 -7.36679975e-03 2.62686033e-02
8.90218139e-01 4.93902713e-01 1.84997022e-01 6.91247359e-02
6.28710330e-01 -4.09385115e-01 3.21175635e-01 1.93982735e-01
-3.57964516e-01 5.76838374e-01 1.21642508e-01 3.31438519e-02
-7.62674809e-01 -6.34686053e-01 -1.82353526e-01 1.51535261e+00
3.98946762e-01 1.64946876e-02 -9.23422933e-01 -4.94977474e-01
-4.00819719e-01 8.72222364e-01 -7.17069745e-01 -7.38163367e-02
-4.82097417e-01 -5.36510408e-01 -2.25189235e-02 2.63608068e-01
6.25559926e-01 -1.07565784e+00 -9.52782393e-01 -4.30552721e-01
-1.16048470e-01 -1.19474208e+00 -4.41550255e-01 -2.54405051e-01
-9.22938645e-01 -1.05689180e+00 -8.09826791e-01 -5.72455108e-01
7.79106736e-01 8.90765309e-01 8.68910849e-01 1.41339704e-01
5.20501807e-02 1.57967880e-01 -2.80505478e-01 -2.82207340e-01
-1.77393258e-01 1.70622587e-01 1.50373340e-01 2.29744270e-01
5.67395449e-01 -3.28368157e-01 -8.36279571e-01 3.20955813e-01
-6.27901316e-01 1.40502453e-01 5.76150239e-01 5.95664561e-01
5.45417964e-01 -1.28632754e-01 4.74493980e-01 -9.60773766e-01
4.77568030e-01 -2.67991662e-01 -6.70586050e-01 1.92490622e-01
-1.01828539e+00 1.35745585e-01 1.73028871e-01 -3.00598085e-01
-1.41104221e+00 -8.53312537e-02 2.23243237e-01 -7.34750688e-01
-3.61411244e-01 3.73625904e-02 1.81581348e-01 -4.90514278e-01
9.30480540e-01 -5.44573069e-02 3.19595546e-01 -4.52710450e-01
2.49924242e-01 8.60141456e-01 3.69360626e-01 3.71183455e-01
6.21502161e-01 5.71977317e-01 2.25697488e-01 -8.53060305e-01
-9.84363854e-01 -6.32837415e-01 -5.82411110e-01 -3.68546933e-01
8.76866043e-01 -8.16433668e-01 -9.34557378e-01 2.99632162e-01
-8.53180707e-01 -1.11506499e-01 -1.29237995e-02 4.02898401e-01
-3.27853352e-01 1.84317425e-01 -2.71902710e-01 -9.16003346e-01
-5.10568023e-01 -9.58186328e-01 9.25774276e-01 7.88571239e-01
4.70533222e-02 -6.41840100e-01 -7.05480278e-02 5.98980963e-01
5.40965080e-01 -7.71902949e-02 3.30999792e-01 -2.55086124e-01
-7.81858802e-01 3.24562788e-01 -5.32810092e-01 2.18219206e-01
2.19034150e-01 -2.20889017e-01 -1.34002829e+00 -3.55101615e-01
-5.71726486e-02 -1.19288005e-01 7.50057876e-01 6.16312504e-01
9.27035570e-01 9.13261324e-02 -3.27198595e-01 6.67276144e-01
1.37838936e+00 3.17627788e-02 6.54092312e-01 3.70734781e-01
8.06503952e-01 9.26274002e-01 8.70761275e-01 -1.22100189e-01
5.79171062e-01 6.45443857e-01 5.61044395e-01 -1.58951595e-01
-4.43146139e-01 1.38813883e-01 2.08194837e-01 2.31039509e-01
-4.28528547e-01 -4.78084981e-02 -6.88965619e-01 3.38339686e-01
-1.71825731e+00 -1.16444635e+00 -2.35986948e-01 2.44795251e+00
7.04975963e-01 1.17554046e-01 2.83661306e-01 1.36947051e-01
8.76561701e-01 -2.06646118e-02 -8.32919061e-01 -1.86286777e-01
1.09882772e-01 3.54365826e-01 6.21933520e-01 3.12581539e-01
-8.26034963e-01 8.39772940e-01 6.02646542e+00 2.83001870e-01
-1.46362555e+00 4.19410408e-01 4.73321855e-01 -4.24009293e-01
4.22510728e-02 -4.71639603e-01 -9.36293900e-01 5.65502882e-01
1.30836570e+00 9.85015407e-02 5.65976262e-01 4.96333867e-01
4.79414672e-01 -5.48888028e-01 -9.06858921e-01 1.15434873e+00
6.37162387e-01 -1.15940893e+00 -5.04595876e-01 9.47392359e-02
5.27039707e-01 2.28940755e-01 5.39982378e-01 -1.44476846e-01
-1.71454951e-01 -1.04288936e+00 3.32394272e-01 5.99742949e-01
1.17499387e+00 -6.76452994e-01 5.89569747e-01 2.76581079e-01
-8.25706005e-01 -1.46355197e-01 -3.10868807e-02 1.30805895e-01
2.83091050e-02 4.42487337e-02 -6.52366042e-01 2.13729993e-01
1.12783813e+00 7.85708070e-01 -7.99594939e-01 1.10996377e+00
-1.97924480e-01 4.07982200e-01 -3.79973240e-02 2.04354689e-01
-3.34440559e-01 4.71771993e-02 3.94056231e-01 6.53820157e-01
1.86546907e-01 1.29213363e-01 -6.29450560e-01 5.57210863e-01
-1.19362392e-01 -1.20235361e-01 -3.87343138e-01 4.70202595e-01
2.77842939e-01 1.09104109e+00 -5.22116899e-01 9.53792110e-02
-5.35980105e-01 8.99900138e-01 2.47193277e-01 5.12254477e-01
-5.65791667e-01 -8.06578621e-02 5.34097016e-01 5.23353100e-01
3.85140151e-01 1.22640051e-01 -4.02753979e-01 -9.77029979e-01
-7.23884925e-02 -7.64544964e-01 1.35068879e-01 -1.25071168e+00
-7.70690620e-01 7.99646974e-01 -2.19789483e-02 -1.30376661e+00
-4.17901814e-01 -3.03665996e-01 -1.31815031e-01 1.11046088e+00
-2.20421934e+00 -1.11569583e+00 -7.01078892e-01 7.69967377e-01
6.35892391e-01 -6.51712716e-02 5.95494449e-01 1.88792601e-01
-5.52436113e-01 6.63245440e-01 -2.94342707e-03 -3.96460027e-01
1.14570737e+00 -1.04200327e+00 5.42831682e-02 9.93273437e-01
6.73692226e-02 6.03191197e-01 9.28831220e-01 -4.55258965e-01
-9.59680796e-01 -8.83501351e-01 8.81060183e-01 -4.06994015e-01
3.47729146e-01 -7.18502328e-02 -1.12736392e+00 5.89163542e-01
5.02068937e-01 1.51368305e-01 6.85223520e-01 2.55802810e-01
-1.69619650e-01 -3.91381413e-01 -1.24764824e+00 4.39888626e-01
9.82395589e-01 -4.60685104e-01 -5.57361126e-01 -1.04550324e-01
6.47052169e-01 -5.96333802e-01 -5.21927416e-01 2.80497015e-01
5.67992747e-01 -1.24618125e+00 7.04603851e-01 -1.93849355e-01
2.69136667e-01 -2.86862046e-01 2.99676299e-01 -1.12493241e+00
-3.12664658e-01 -5.49162090e-01 -4.78776693e-01 9.03865635e-01
2.31224671e-01 -5.29601276e-01 9.12902594e-01 6.36995614e-01
2.62448877e-01 -5.28281093e-01 -7.09004939e-01 -2.99130559e-01
-5.08379996e-01 -5.90887889e-02 2.93187797e-01 7.30149209e-01
-1.62774608e-01 7.15131700e-01 -3.87291253e-01 2.10761040e-01
9.37034667e-01 -1.23066463e-01 6.27183497e-01 -1.40492439e+00
1.32278919e-01 -2.00290516e-01 -4.25372928e-01 -8.41795683e-01
1.67777151e-01 -1.98696449e-01 -1.20516516e-01 -1.28740573e+00
9.75382179e-02 -1.81840822e-01 -6.41265094e-01 2.81338036e-01
-3.04421932e-01 6.51950181e-01 2.91002125e-01 4.68435138e-01
-7.80519128e-01 1.74876079e-01 1.23831618e+00 3.38888615e-01
-4.49076355e-01 2.20761940e-01 -8.40229809e-01 5.71333051e-01
7.08919525e-01 -3.67405683e-01 -7.03809917e-01 -2.86611080e-01
3.16064149e-01 3.73229152e-03 2.46871173e-01 -9.22256529e-01
6.01514161e-01 8.53577703e-02 3.56605977e-01 -5.64224303e-01
4.53471929e-01 -9.60776031e-01 -1.69956893e-01 1.45984488e-02
-3.88682902e-01 1.24312723e-02 1.82170346e-01 6.39620066e-01
-4.51274030e-02 5.24934381e-03 9.68855798e-01 2.42612183e-01
-7.98278630e-01 7.74502307e-02 1.86803833e-01 -2.26743847e-01
1.12590599e+00 -5.81627131e-01 -6.58393562e-01 -4.08884525e-01
-6.54545128e-01 7.81297684e-02 7.25969493e-01 5.85771918e-01
7.52938807e-01 -7.10063100e-01 -3.75748187e-01 3.31078231e-01
3.55667397e-02 -7.05108419e-02 3.72399330e-01 1.02347255e+00
-8.66812766e-02 3.89565974e-01 -4.94277388e-01 -7.95874178e-01
-1.78442430e+00 6.15312815e-01 1.26902834e-01 1.33334666e-01
-4.35822487e-01 7.37166822e-01 1.84961885e-01 3.61132950e-01
3.69700342e-01 -3.92231643e-02 -9.13354576e-01 -6.48464710e-02
9.24063146e-01 4.21750605e-01 7.18707517e-02 -1.08537126e+00
-1.99886486e-01 7.14354575e-01 -3.38099897e-01 5.35797104e-02
1.32089174e+00 -7.25726426e-01 -1.90651440e-03 5.28014600e-01
8.35245252e-01 -1.66111276e-01 -1.34112990e+00 -4.26276654e-01
-1.22824199e-01 -5.60293734e-01 5.24728298e-01 -8.73528779e-01
-1.14845824e+00 8.23425531e-01 1.20755756e+00 3.92151922e-02
1.58174551e+00 -6.35233372e-02 5.02569377e-01 -5.91756515e-02
3.68271142e-01 -9.41815436e-01 -1.87957805e-04 8.15347731e-02
5.17892778e-01 -1.65928042e+00 9.93178710e-02 -4.31868255e-01
-8.35258842e-01 6.28860891e-01 8.39292288e-01 1.75641645e-02
6.73346102e-01 -9.35800523e-02 2.12127045e-01 -1.18883997e-01
-7.20166147e-01 -4.89470720e-01 6.20546639e-01 7.38188028e-01
4.20919329e-01 -4.25172985e-01 -5.52557446e-02 6.47662580e-02
-6.93471134e-02 4.41708386e-01 5.24626791e-01 7.11385369e-01
-3.83417845e-01 -7.82925725e-01 -5.07333815e-01 5.25855064e-01
-7.57972300e-01 -1.47698984e-01 -1.05367042e-03 6.62498057e-01
1.07250556e-01 1.10755754e+00 2.93973267e-01 -3.75774413e-01
1.21467754e-01 -1.45231783e-01 4.93444949e-01 -5.82984209e-01
-4.78414714e-01 1.09055042e-01 -1.49638042e-01 -8.06894422e-01
-1.24040949e+00 -6.96576536e-01 -8.90661299e-01 -3.93385053e-01
-4.86345053e-01 -3.94134402e-01 8.44077766e-01 9.41882610e-01
6.83937609e-01 7.03479290e-01 5.32417715e-01 -8.65360081e-01
-1.05275624e-01 -1.05441928e+00 -5.22880137e-01 2.06808835e-01
8.27358007e-01 -8.77692461e-01 -3.08198512e-01 3.09184402e-01] | [14.097604751586914, 0.07312560081481934] |
cd8dc0c3-4a38-4a8b-a395-209e983f95dc | point-m2ae-multi-scale-masked-autoencoders | 2205.14401 | null | https://arxiv.org/abs/2205.14401v2 | https://arxiv.org/pdf/2205.14401v2.pdf | Point-M2AE: Multi-scale Masked Autoencoders for Hierarchical Point Cloud Pre-training | Masked Autoencoders (MAE) have shown great potentials in self-supervised pre-training for language and 2D image transformers. However, it still remains an open question on how to exploit masked autoencoding for learning 3D representations of irregular point clouds. In this paper, we propose Point-M2AE, a strong Multi-scale MAE pre-training framework for hierarchical self-supervised learning of 3D point clouds. Unlike the standard transformer in MAE, we modify the encoder and decoder into pyramid architectures to progressively model spatial geometries and capture both fine-grained and high-level semantics of 3D shapes. For the encoder that downsamples point tokens by stages, we design a multi-scale masking strategy to generate consistent visible regions across scales, and adopt a local spatial self-attention mechanism during fine-tuning to focus on neighboring patterns. By multi-scale token propagation, the lightweight decoder gradually upsamples point tokens with complementary skip connections from the encoder, which further promotes the reconstruction from a global-to-local perspective. Extensive experiments demonstrate the state-of-the-art performance of Point-M2AE for 3D representation learning. With a frozen encoder after pre-training, Point-M2AE achieves 92.9% accuracy for linear SVM on ModelNet40, even surpassing some fully trained methods. By fine-tuning on downstream tasks, Point-M2AE achieves 86.43% accuracy on ScanObjectNN, +3.36% to the second-best, and largely benefits the few-shot classification, part segmentation and 3D object detection with the hierarchical pre-training scheme. Code is available at https://github.com/ZrrSkywalker/Point-M2AE. | ['Rongyao Fang', 'Hongsheng Li', 'Yu Qiao', 'Dong Wang', 'Bin Zhao', 'Peng Gao', 'Ziyu Guo', 'Renrui Zhang'] | 2022-05-28 | null | null | null | null | ['3d-point-cloud-classification', '3d-point-cloud-linear-classification', 'point-cloud-pre-training', 'few-shot-3d-point-cloud-classification'] | ['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision'] | [-6.99189678e-03 2.55651236e-01 -1.01773053e-01 -3.95078778e-01
-1.00886667e+00 -3.99433076e-01 4.63868678e-01 -1.76794186e-01
1.26324249e-02 1.91521347e-02 3.70536447e-02 -2.48488829e-01
3.05715263e-01 -1.07789075e+00 -1.33410275e+00 -5.93627334e-01
-8.09880793e-02 4.80780363e-01 3.40748847e-01 -1.38103679e-01
-1.57638267e-01 9.12674189e-01 -1.78471780e+00 6.43816531e-01
4.97305661e-01 1.19081771e+00 5.91233611e-01 5.88479400e-01
-4.98886526e-01 6.83488369e-01 -3.04759204e-01 -2.57759422e-01
3.30920994e-01 5.85641079e-02 -6.23076141e-01 4.35483366e-01
7.28898346e-01 -4.01402175e-01 -4.17887926e-01 8.30923736e-01
4.46600199e-01 -1.95739731e-01 6.23866081e-01 -9.00869548e-01
-1.02972245e+00 5.20045102e-01 -6.95261240e-01 7.23799691e-02
-9.69174802e-02 3.62050354e-01 9.62028444e-01 -1.43142688e+00
4.98860538e-01 1.24369955e+00 8.88033628e-01 6.26474321e-01
-1.28545082e+00 -8.08967650e-01 1.24778874e-01 -9.97559652e-02
-1.47807717e+00 -1.80835277e-01 8.53103042e-01 -5.07829726e-01
1.48113549e+00 -1.70239527e-02 7.78408229e-01 1.02613413e+00
-1.07524190e-02 9.44643080e-01 9.71352696e-01 -2.48787314e-01
1.71661064e-01 -1.07720895e-02 -1.73338622e-01 8.51733506e-01
-1.13206461e-01 3.23208630e-01 -5.39874017e-01 7.15474412e-02
1.35315025e+00 1.78001910e-01 -9.78890341e-03 -5.10990620e-01
-1.08097708e+00 8.26295972e-01 1.09437633e+00 1.85897782e-01
-4.25275236e-01 2.10368603e-01 1.31049171e-01 1.81639940e-01
7.11616278e-01 2.76847899e-01 -5.51108658e-01 2.68858373e-01
-9.88293707e-01 4.49436568e-02 2.27777839e-01 1.28557348e+00
1.12459338e+00 3.98215175e-01 -6.47173747e-02 8.26196909e-01
3.31857681e-01 7.13138223e-01 2.47144133e-01 -7.75180042e-01
3.52736950e-01 8.18713725e-01 -2.42016643e-01 -4.38615263e-01
-1.39102459e-01 -8.20000350e-01 -9.90608871e-01 6.39287174e-01
-7.49593973e-02 3.90590638e-01 -1.46480882e+00 1.47688162e+00
4.72460181e-01 3.19955677e-01 1.27201912e-03 8.96950006e-01
1.10979080e+00 8.45955133e-01 7.48759434e-02 3.88605297e-01
1.32871640e+00 -9.36446011e-01 7.18289167e-02 -4.81988579e-01
4.30629104e-01 -6.07091069e-01 1.22212684e+00 -1.08019315e-01
-1.32212031e+00 -9.43730533e-01 -9.67480719e-01 -4.63338673e-01
-2.56585628e-01 1.65015727e-01 6.17531359e-01 1.61701679e-01
-1.09452927e+00 5.50423443e-01 -1.07193708e+00 -1.83621287e-01
1.01761961e+00 2.50996113e-01 -3.99325907e-01 -2.13510051e-01
-8.32496881e-01 6.90146804e-01 1.03151098e-01 -1.60656065e-01
-1.23572206e+00 -1.15125453e+00 -1.08320248e+00 1.94635227e-01
-1.13990039e-01 -9.81317282e-01 1.32568443e+00 -7.22231209e-01
-1.34888935e+00 1.42147624e+00 -2.30380952e-01 -5.99484146e-01
2.29360610e-01 -6.76467866e-02 -1.05263203e-01 1.79795444e-01
4.72768426e-01 1.33457279e+00 1.10664356e+00 -1.47531366e+00
-6.12780035e-01 -4.59876120e-01 -1.07830130e-01 2.67598122e-01
-8.82931519e-03 -2.79563516e-01 -5.89509904e-01 -7.59719312e-01
4.18615073e-01 -7.12216437e-01 -2.02568263e-01 2.91031092e-01
-2.41319776e-01 -2.66108692e-01 8.10276568e-01 -3.26371729e-01
5.21890461e-01 -2.37823319e+00 -3.74856405e-03 -5.26392534e-02
2.72687525e-01 7.12316483e-02 -2.90126264e-01 4.60430980e-02
-2.97618240e-01 -9.50866342e-02 -5.49472094e-01 -7.30360806e-01
-1.00275949e-01 2.66993642e-01 -5.91673434e-01 3.36306125e-01
7.01283514e-01 1.33999741e+00 -5.95838308e-01 -3.11278790e-01
6.02478266e-01 5.87241292e-01 -6.48953676e-01 2.69563377e-01
-4.60543364e-01 2.34590560e-01 -2.92637020e-01 1.02257955e+00
9.37896192e-01 -7.53163576e-01 -5.45858681e-01 -3.52689743e-01
-1.71217546e-01 3.33099693e-01 -6.68397963e-01 2.15802717e+00
-7.45585859e-01 4.94281918e-01 1.40993595e-01 -8.12113523e-01
9.36900556e-01 1.88313257e-02 4.37839776e-01 -6.58185482e-01
-1.32501749e-02 2.12414145e-01 -3.26781839e-01 -1.33793384e-01
2.97646135e-01 -2.86078513e-01 -1.31088486e-02 9.62034464e-02
4.12067533e-01 -5.75816631e-01 -3.94467771e-01 1.03208937e-01
1.05718887e+00 3.06054235e-01 2.33357474e-02 -3.98700796e-02
2.21329674e-01 2.58349925e-01 8.86193663e-02 7.15665162e-01
1.62363783e-01 1.07631814e+00 -7.51041546e-02 -3.93923879e-01
-1.34145880e+00 -1.36547148e+00 -2.53807724e-01 9.14722800e-01
1.15155824e-01 -1.93180442e-01 -4.57359999e-01 -5.85360467e-01
4.11566615e-01 7.16990054e-01 -6.63027883e-01 -1.78290188e-01
-5.03785431e-01 -3.34480613e-01 4.64671046e-01 7.91081607e-01
5.72985232e-01 -1.04224622e+00 -5.28378725e-01 2.37806037e-01
1.96111068e-01 -1.26135528e+00 -2.30294019e-01 6.56212986e-01
-1.11988878e+00 -6.42532647e-01 -9.03128922e-01 -1.06547654e+00
7.69047737e-01 5.35435975e-01 1.32970917e+00 -6.78718388e-02
-2.52431154e-01 2.78715670e-01 -2.29466140e-01 -3.33634585e-01
-3.79638880e-01 1.85591266e-01 -2.57195264e-01 -3.32450449e-01
4.37352002e-01 -8.87444377e-01 -5.63205838e-01 2.97651678e-01
-6.97427034e-01 2.83497304e-01 1.00659943e+00 8.23753655e-01
1.04331553e+00 -2.93307543e-01 1.27257645e-01 -4.93668884e-01
-1.26981229e-01 -4.24091518e-01 -5.66586018e-01 -1.32952541e-01
-1.69466823e-01 3.53000574e-02 4.08812493e-01 -4.08548653e-01
-8.92855525e-01 2.28939906e-01 -6.40685141e-01 -1.28611493e+00
-4.39941913e-01 2.60564443e-02 -1.63379192e-01 -2.12183774e-01
6.91765010e-01 6.24415934e-01 -1.52753219e-01 -5.75517297e-01
5.20864129e-01 5.49646795e-01 5.88454306e-01 -4.58441794e-01
1.10165954e+00 8.06009829e-01 -2.35870719e-01 -9.28814769e-01
-1.03121018e+00 -4.43444788e-01 -5.79669952e-01 4.47300263e-02
1.10555577e+00 -1.48516881e+00 -4.93963063e-01 4.26962644e-01
-1.21799338e+00 -7.01530397e-01 -8.30552280e-01 1.38333753e-01
-6.91842556e-01 2.04680320e-02 -6.79787755e-01 -5.07883668e-01
-3.96283209e-01 -9.77094233e-01 1.67027354e+00 -1.08763784e-01
8.55921954e-02 -5.49373567e-01 -2.54179716e-01 3.14216942e-01
2.91447043e-01 2.22251583e-02 8.62541854e-01 -2.45757103e-01
-1.05033219e+00 5.59349470e-02 -4.14011538e-01 4.95368659e-01
1.51877673e-02 -3.94662321e-01 -1.21521246e+00 -1.80684686e-01
1.46751195e-01 -5.12503147e-01 9.63039279e-01 4.79437888e-01
1.60391724e+00 -1.44465089e-01 -3.34240526e-01 1.10351396e+00
1.40008986e+00 -2.63956845e-01 5.62682986e-01 2.55699694e-01
9.14336562e-01 1.33139566e-01 2.83536017e-01 3.26095074e-01
5.02072215e-01 3.73508126e-01 6.62890911e-01 -2.80400395e-01
-5.65706193e-01 -6.95039034e-01 1.67815432e-01 7.67048180e-01
5.11705875e-02 5.31965494e-02 -9.66323256e-01 6.84078395e-01
-1.38227224e+00 -9.26124334e-01 2.78471168e-02 1.84828722e+00
7.24565744e-01 3.11416358e-01 -1.32359639e-01 -1.10558517e-01
5.72466254e-01 3.24600995e-01 -7.25657940e-01 3.53459232e-02
-3.23946029e-01 4.71866757e-01 6.71305954e-01 3.78428638e-01
-1.16640806e+00 1.30817485e+00 5.19270420e+00 1.10357356e+00
-1.28316092e+00 2.88390487e-01 6.49992049e-01 -1.01806544e-01
-4.39638615e-01 -2.06298694e-01 -9.75167871e-01 3.49740982e-01
6.24325037e-01 3.62672687e-01 1.97255611e-01 1.24566329e+00
-1.08558975e-01 3.99255067e-01 -1.03602862e+00 1.14943063e+00
-2.47326810e-02 -1.80916417e+00 1.58569366e-01 5.07307332e-03
8.89539003e-01 7.84168541e-01 9.70180482e-02 4.22516853e-01
3.90468538e-01 -1.01623046e+00 1.05612922e+00 1.38367787e-01
1.18204796e+00 -5.00517368e-01 2.06638351e-01 4.39573586e-01
-1.38297975e+00 2.87505006e-03 -6.13902867e-01 1.18904456e-01
1.50957346e-01 6.41432166e-01 -7.80972779e-01 3.70519459e-01
1.03747261e+00 8.44622552e-01 -4.14095074e-01 7.71950603e-01
-3.02404106e-01 5.76842129e-01 -6.09182715e-01 1.63464844e-01
4.66472507e-01 2.34636828e-01 6.02422953e-01 1.17502081e+00
4.87233877e-01 1.62992895e-01 -5.27456738e-02 1.25848985e+00
-2.74725705e-01 -2.78916329e-01 -6.53053522e-01 2.18686447e-01
5.10666668e-01 1.02446461e+00 -5.78745484e-01 -4.28579241e-01
-5.27357638e-01 1.18526340e+00 5.58781683e-01 2.62451470e-01
-8.05195034e-01 -1.03899471e-01 8.15069616e-01 4.69562680e-01
9.44506407e-01 -3.85502726e-01 -6.08199298e-01 -1.15343261e+00
-2.28119344e-02 -6.74282372e-01 1.68988824e-01 -1.11102998e+00
-1.46958184e+00 8.20349336e-01 -1.06601693e-01 -1.54033506e+00
5.24939746e-02 -7.84947395e-01 -5.34563184e-01 7.56126642e-01
-1.64691985e+00 -1.64519310e+00 -4.66842741e-01 6.19865656e-01
8.53104174e-01 -8.93267170e-02 7.85221577e-01 2.06475884e-01
-1.04587944e-02 3.86871964e-01 -1.72221065e-01 9.63042900e-02
2.37607986e-01 -1.22269940e+00 9.48721945e-01 6.45007730e-01
4.52269524e-01 4.23682690e-01 2.41656452e-01 -6.75321698e-01
-1.30061114e+00 -1.46567476e+00 6.92388177e-01 -5.85660160e-01
5.30609667e-01 -7.43819773e-01 -1.17577219e+00 7.59973049e-01
-7.55399466e-02 3.66845399e-01 3.57052714e-01 -2.13536784e-01
-5.92157185e-01 -7.28893699e-03 -1.03286612e+00 4.08201426e-01
1.47536254e+00 -8.67450714e-01 -7.97914982e-01 3.90977323e-01
9.83022273e-01 -7.74848461e-01 -8.60051870e-01 5.72962642e-01
8.67717490e-02 -1.02719545e+00 1.39614034e+00 -3.04133028e-01
7.87662983e-01 -3.61872852e-01 -4.07641023e-01 -1.06477702e+00
-5.94584644e-01 -1.75598189e-01 -3.30124676e-01 9.30421472e-01
3.53063434e-01 -3.15246433e-01 1.21001685e+00 8.64400417e-02
-6.90497279e-01 -1.02256918e+00 -9.60220516e-01 -7.97596872e-01
2.46157467e-01 -7.91365147e-01 8.16262007e-01 7.64768779e-01
-7.16451108e-01 3.12579274e-01 -5.59262466e-03 5.95682800e-01
8.26093972e-01 5.54889798e-01 7.91108966e-01 -8.73716295e-01
-3.07728171e-01 -4.85505134e-01 -4.09923226e-01 -1.74611425e+00
1.54899329e-01 -1.20936275e+00 -5.45266643e-02 -1.61635685e+00
-1.41404793e-01 -5.90775669e-01 -1.44966722e-01 5.51516235e-01
1.45608723e-01 4.93901610e-01 7.11044669e-02 3.15691471e-01
-2.92544544e-01 9.34809089e-01 1.31627071e+00 -3.51829022e-01
-9.24133733e-02 3.16918939e-02 -4.92624819e-01 6.65897548e-01
6.35474622e-01 -4.47002381e-01 -2.29766160e-01 -8.93931806e-01
-1.39715180e-01 -1.01335928e-01 8.24347138e-01 -1.13207006e+00
1.36334121e-01 1.33890212e-01 7.50204563e-01 -1.14287293e+00
6.37523472e-01 -8.73629093e-01 -1.17160217e-03 2.68588483e-01
-8.13095495e-02 -5.47791645e-02 4.17987406e-01 5.41566312e-01
-2.84043878e-01 1.15011027e-02 8.88799429e-01 -4.71291006e-01
-9.08650637e-01 6.99317515e-01 -6.07481897e-02 -1.06004298e-01
7.76572049e-01 -4.39225167e-01 -2.32172497e-02 -1.64755117e-02
-8.29182029e-01 1.32810280e-01 7.40424991e-01 4.81815487e-01
1.00735569e+00 -1.36544991e+00 -6.57706976e-01 6.36948287e-01
2.71641016e-01 7.74058580e-01 6.27597511e-01 3.33807379e-01
-5.00108361e-01 2.32039750e-01 -1.60958767e-01 -1.17316794e+00
-8.89595926e-01 5.44828832e-01 4.57731545e-01 4.86027859e-02
-1.06593072e+00 1.40166318e+00 5.87290883e-01 -6.06691062e-01
1.79075330e-01 -5.44918776e-01 3.69973332e-01 -3.26048285e-01
1.70847625e-01 -1.29633024e-01 2.41183147e-01 -6.39819026e-01
-4.24786270e-01 9.18545008e-01 5.09429164e-02 1.33837417e-01
1.64320314e+00 5.80367632e-02 1.68805540e-01 4.44523484e-01
1.38695335e+00 -1.60506979e-01 -1.58917272e+00 -3.70363683e-01
-4.06901181e-01 -4.21545029e-01 2.33020678e-01 -5.70330083e-01
-1.03905249e+00 1.18517303e+00 6.41530991e-01 -7.58195296e-02
9.25966084e-01 6.33737147e-01 7.45912969e-01 1.98595509e-01
3.73295099e-01 -4.43238020e-01 2.22840160e-01 5.80660880e-01
1.07938230e+00 -1.38789904e+00 -1.71486512e-01 -4.82726961e-01
-6.00485027e-01 8.47165644e-01 9.05841172e-01 -5.05842865e-01
7.54775763e-01 3.99320990e-01 1.32321296e-02 -4.86904383e-01
-7.48403609e-01 -4.24182445e-01 3.92112315e-01 6.44585013e-01
9.06941220e-02 2.11939942e-02 7.72406459e-01 5.41108727e-01
-3.99347484e-01 -1.41270682e-01 -4.83941659e-02 7.41846859e-01
-4.04332429e-01 -6.38045192e-01 -3.04403514e-01 4.10295457e-01
8.37340206e-02 -1.85859427e-01 -3.59052047e-02 7.40609765e-01
3.47863406e-01 3.07681769e-01 5.74201941e-01 -4.03296739e-01
3.93707991e-01 -1.41559429e-02 5.63983142e-01 -8.48065853e-01
-4.08098787e-01 9.74401981e-02 -3.19523275e-01 -5.30918658e-01
-2.56333172e-01 -5.35618007e-01 -1.37729692e+00 -2.90868543e-02
-2.22803190e-01 -2.27737039e-01 5.47052979e-01 5.52528977e-01
7.18555629e-01 6.52264237e-01 7.01025188e-01 -1.32444930e+00
-5.14493763e-01 -9.08312738e-01 -3.08137745e-01 2.44607568e-01
5.08467078e-01 -6.20037079e-01 -3.01534176e-01 -4.69980836e-02] | [8.0751314163208, -3.427947759628296] |
0f9a58cf-b8f4-42ab-8c4b-37aaa77f7c79 | chatgpt-a-blessing-or-a-curse-for | 2304.14993 | null | https://arxiv.org/abs/2304.14993v2 | https://arxiv.org/pdf/2304.14993v2.pdf | ChatGPT -- a Blessing or a Curse for Undergraduate Computer Science Students and Instructors? | ChatGPT is an AI language model developed by OpenAI that can understand and generate human-like text. It can be used for a variety of use cases such as language generation, question answering, text summarization, chatbot development, language translation, sentiment analysis, content creation, personalization, text completion, and storytelling. While ChatGPT has garnered significant positive attention, it has also generated a sense of apprehension and uncertainty in academic circles. There is concern that students may leverage ChatGPT to complete take-home assignments and exams and obtain favorable grades without genuinely acquiring knowledge. This paper adopts a quantitative approach to demonstrate ChatGPT's high degree of unreliability in answering a diverse range of questions pertaining to topics in undergraduate computer science. Our analysis shows that students may risk self-sabotage by blindly depending on ChatGPT to complete assignments and exams. We build upon this analysis to provide constructive recommendations to both students and instructors. | ['Harshal D. Akolekar', 'Dhruv Kumar', 'Sayan Mitra', 'M. Osama Ataullah', 'Jahnvi Kadia', 'Harshal Dev', 'Ritvik Budhiraja', 'Ishika Joshi'] | 2023-04-28 | null | null | null | null | ['text-summarization'] | ['natural-language-processing'] | [ 6.55089393e-02 8.75699043e-01 8.13467577e-02 -2.23097041e-01
-9.86734748e-01 -9.21749175e-01 6.02903187e-01 7.06761003e-01
1.42092124e-01 9.07825112e-01 2.91950494e-01 -9.46965218e-01
-1.04765862e-01 -6.34253502e-01 -4.22613919e-01 -9.25073326e-02
6.54956460e-01 2.37632275e-01 -2.68850736e-02 -7.35119760e-01
8.88258159e-01 1.65656671e-01 -1.32262397e+00 2.98193425e-01
1.33732545e+00 3.43814641e-02 1.04000717e-01 8.47872674e-01
-6.04560494e-01 1.51798856e+00 -1.32589328e+00 -8.14295709e-01
-4.68528718e-01 -7.05458105e-01 -1.46542811e+00 -2.10630178e-01
3.21472615e-01 1.52292550e-01 6.95478693e-02 9.30673420e-01
5.23636520e-01 2.60171711e-01 2.06482649e-01 -1.31041276e+00
-9.20523524e-01 9.68576193e-01 -4.13514316e-01 2.42497325e-01
9.10966337e-01 2.05502689e-01 8.11532736e-01 -2.37354770e-01
8.02645028e-01 1.11624432e+00 6.43895388e-01 5.78211427e-01
-9.76055086e-01 -6.07826173e-01 -2.49945983e-01 2.03999691e-02
-9.22990382e-01 -6.20078743e-02 5.98692775e-01 -5.62455893e-01
7.40561962e-01 5.65747261e-01 8.68237138e-01 1.10403633e+00
6.61589384e-01 8.14013660e-01 1.04784203e+00 -5.77714086e-01
2.10796580e-01 8.27211559e-01 7.99870789e-02 5.71414769e-01
2.08263442e-01 -8.27587903e-01 -7.44643211e-01 -2.51337975e-01
4.23886806e-01 -3.46859097e-01 4.55504321e-02 4.28719401e-01
-1.29015028e+00 9.27253962e-01 -1.04195386e-01 4.55068529e-01
-4.03161272e-02 -6.18093275e-02 1.22155368e-01 7.11529732e-01
2.19210297e-01 1.36842191e+00 -3.36383998e-01 -1.00456953e+00
-5.73111832e-01 4.70997036e-01 1.51247311e+00 1.19489980e+00
3.01868916e-01 -9.50289443e-02 -2.68319994e-01 8.12365413e-01
1.78635851e-01 2.58532166e-01 9.13774014e-01 -1.18910635e+00
3.98286343e-01 8.75701368e-01 -6.16046041e-02 -1.11952651e+00
2.55537212e-01 -6.69645518e-02 -2.38607079e-01 1.81950733e-01
3.37907046e-01 -7.25241780e-01 -4.33413804e-01 1.33818066e+00
-7.21529350e-02 -2.59998411e-01 3.16711903e-01 4.44260687e-01
1.17931962e+00 6.59609735e-01 8.13966095e-02 8.59883204e-02
1.40786839e+00 -8.21480334e-01 -8.01503956e-01 -3.42780411e-01
8.44443977e-01 -1.39503253e+00 1.04622340e+00 5.75554669e-01
-1.30725706e+00 -4.68357950e-02 -9.75431979e-01 -2.71020442e-01
-2.76819527e-01 -2.99209297e-01 3.63508701e-01 9.06426251e-01
-8.34830284e-01 8.07889819e-01 -3.54776591e-01 -6.32534325e-01
3.80104810e-01 -1.15933143e-01 -1.72515467e-01 -1.80221558e-01
-9.10149574e-01 1.10332346e+00 -1.64911851e-01 -4.63465035e-01
-3.12458932e-01 -9.26567912e-01 -5.71588814e-01 1.64323598e-01
3.98639858e-01 -6.35901570e-01 1.78281546e+00 -4.15141046e-01
-1.80217564e+00 8.22622836e-01 2.56416231e-01 -2.42478758e-01
5.65059721e-01 -1.49613380e-01 1.55574560e-01 -1.17619790e-01
2.23029301e-01 4.41947699e-01 3.45518649e-01 -5.95709980e-01
-1.15985893e-01 -8.79376475e-03 2.68014580e-01 4.26111549e-01
-2.91355371e-01 6.06682035e-04 3.72139126e-01 -5.10416925e-01
2.71069929e-02 -8.54717612e-01 -3.00405592e-01 -5.81590474e-01
-6.65342808e-01 -4.40768331e-01 7.91437447e-01 -3.81057739e-01
1.06104004e+00 -1.39520168e+00 -2.95999140e-01 -2.17098147e-02
3.90335768e-01 2.79186338e-01 8.23196173e-02 1.09059632e+00
6.18490577e-02 9.71295178e-01 3.48041117e-01 1.37717351e-01
1.55483454e-01 -1.31718263e-01 -3.89015675e-01 -1.72895595e-01
1.28691554e-01 8.20957243e-01 -1.21044958e+00 -5.60672283e-01
1.10256061e-01 1.21219661e-02 -3.95825684e-01 3.01822841e-01
-3.71991903e-01 2.72253156e-01 -8.02573979e-01 5.39299488e-01
5.05758598e-02 -4.30468231e-01 -2.30181627e-02 1.04423618e+00
-3.43766928e-01 7.14766979e-01 -7.38016248e-01 1.54935539e+00
-6.28258348e-01 1.18974745e+00 -1.57964692e-01 -6.82335794e-01
1.04638910e+00 5.62789261e-01 1.65123507e-01 -1.47641718e-01
3.25115323e-01 -4.59324345e-02 4.96670976e-02 -5.78819513e-01
9.51792300e-01 -2.26818044e-02 -2.01555759e-01 1.29696202e+00
9.32030976e-02 -1.03854704e+00 3.16706091e-01 8.30787301e-01
1.24683928e+00 -9.09968466e-02 9.24108699e-02 -3.96490991e-01
2.31338575e-01 2.75215179e-01 -2.06390217e-01 8.91640067e-01
-4.94609401e-02 4.10490960e-01 8.73901188e-01 -1.77286416e-01
-7.11221397e-01 -5.48238933e-01 2.64835387e-01 1.08180308e+00
-2.23720878e-01 -8.16988528e-01 -8.16734493e-01 -2.83191949e-01
-4.32073087e-01 1.21579480e+00 -1.42090276e-01 -4.40728486e-01
6.71177208e-02 -1.75500959e-01 6.75968826e-01 -9.13333520e-02
3.57158989e-01 -1.23559797e+00 -7.36377954e-01 3.66332918e-01
-4.15117770e-01 -9.35917675e-01 -2.98757195e-01 -7.14163929e-02
-4.97345537e-01 -6.00272298e-01 -5.68883598e-01 -7.88007200e-01
7.38082290e-01 4.04390633e-01 1.12812197e+00 3.50063562e-01
-4.48969007e-01 9.64591384e-01 -4.14168745e-01 -1.15501142e+00
-1.07593608e+00 3.81439358e-01 -3.86470646e-01 -8.25092137e-01
4.13155138e-01 -7.27746367e-01 -2.15488076e-01 1.93819840e-04
-8.86126399e-01 4.30111021e-01 3.77131611e-01 5.27361572e-01
-2.75386006e-01 -2.93687314e-01 8.89791489e-01 -1.18343663e+00
1.53656185e+00 -6.12857640e-01 -2.56623745e-01 3.99551928e-01
-6.95014775e-01 -3.51422722e-03 4.22611088e-01 -4.91904914e-01
-1.11332798e+00 -6.05591655e-01 -1.05410941e-01 1.19825847e-01
-2.74807543e-01 7.13411391e-01 3.27460378e-01 -3.36127341e-01
1.09693968e+00 6.34582154e-03 2.49742493e-01 -1.51210539e-02
2.45900631e-01 6.14820004e-01 1.52680933e-01 -7.05120146e-01
8.71623397e-01 -3.69255185e-01 -4.45543766e-01 -1.01691759e+00
-8.82388413e-01 -3.36600125e-01 -5.44869751e-02 -5.69392264e-01
5.39633393e-01 -6.67746305e-01 -1.25432122e+00 1.55624688e-01
-1.16234672e+00 -3.80212098e-01 -4.03898180e-01 1.46330357e-01
-9.26651731e-02 1.23527870e-01 -7.04965115e-01 -6.59086049e-01
-2.26283193e-01 -1.01944101e+00 3.53637844e-01 1.07055783e+00
-1.04460585e+00 -1.06409156e+00 -2.02420726e-02 1.18453622e+00
5.11651158e-01 1.62695900e-01 7.60144055e-01 -9.90006864e-01
-6.06427372e-01 -5.72565079e-01 3.32440943e-01 6.17823824e-02
-1.84848055e-01 1.59925893e-01 -8.77670884e-01 1.92255527e-01
-8.70684534e-02 -8.50396395e-01 2.50446727e-03 -2.21053466e-01
7.70457745e-01 -7.52649724e-01 3.40244509e-02 -2.22816870e-01
8.94022703e-01 1.58923730e-01 4.61836100e-01 4.41771358e-01
4.83723015e-01 9.45783317e-01 2.44502798e-01 4.22894120e-01
4.89839375e-01 4.72026207e-02 6.28128648e-02 5.80071688e-01
2.25468010e-01 -5.18026888e-01 3.86988550e-01 1.13625479e+00
2.36258969e-01 -3.85482758e-01 -1.19474626e+00 6.49191976e-01
-1.57893443e+00 -1.05159330e+00 -4.48480368e-01 1.80785751e+00
1.03509343e+00 1.95877194e-01 -2.99466521e-01 -1.46601826e-01
3.82298231e-01 -4.30036783e-02 -1.25246197e-01 -1.20725751e+00
3.48476678e-01 3.20728034e-01 2.08117947e-01 5.65013289e-01
1.19116670e-02 9.24906492e-01 6.13172340e+00 6.37801051e-01
-9.49029088e-01 -1.77817807e-01 8.48519325e-01 2.55341142e-01
-6.89803600e-01 2.15183184e-01 -5.95625877e-01 1.11652583e-01
1.02827406e+00 -1.29777348e+00 2.20962465e-01 8.64942551e-01
-4.68788520e-02 -3.54865760e-01 -7.96916187e-01 4.26167548e-01
3.09606731e-01 -1.65562725e+00 -1.23842865e-01 -2.98316982e-02
1.18052971e+00 -5.45930564e-01 -7.46258646e-02 5.27756870e-01
1.01801991e+00 -1.20736694e+00 4.17537421e-01 1.92624152e-01
8.33066702e-02 -7.42872357e-01 7.08028734e-01 7.15504408e-01
-2.82893449e-01 9.50101838e-02 -1.24272510e-01 -7.94141412e-01
-2.23945171e-01 2.75354028e-01 -1.76540196e+00 3.18480015e-01
1.91807508e-01 1.35502487e-01 -7.38340318e-01 9.44586754e-01
-4.02533948e-01 8.32885265e-01 -2.62711216e-02 -7.16392040e-01
1.13394387e-01 -2.32999146e-01 6.85229361e-01 8.63014162e-01
5.78629673e-01 5.22341192e-01 1.77779302e-01 1.05776083e+00
-7.97409415e-02 3.50105733e-01 -6.78988039e-01 -7.32185841e-01
6.74777031e-01 1.30091023e+00 -7.82583654e-01 -2.74378419e-01
5.76543063e-02 6.11365616e-01 -1.06097691e-01 1.89545378e-01
-2.24007919e-01 -9.55157161e-01 4.07184422e-01 1.18443005e-01
-3.66837680e-01 -1.49331957e-01 -8.39697957e-01 -8.35299969e-01
-2.41353020e-01 -1.04967916e+00 -1.48465261e-01 -1.24706936e+00
-1.06214762e+00 4.04495090e-01 -1.59040540e-01 -8.82908463e-01
-5.70683718e-01 -1.65938422e-01 -1.35629845e+00 9.40598011e-01
-7.39708900e-01 -5.86449325e-01 -3.72259468e-01 2.43811339e-01
5.86293578e-01 -2.65016258e-01 8.77753735e-01 -2.85431385e-01
-4.79830980e-01 5.40611625e-01 -3.29516709e-01 3.78747750e-03
9.55387235e-01 -1.31114864e+00 3.13634753e-01 6.66372597e-01
-2.19252650e-02 8.30247939e-01 1.20919800e+00 -5.56663692e-01
-1.48031402e+00 -6.81772292e-01 1.31254590e+00 -7.58993626e-01
8.12045813e-01 5.98598421e-02 -7.10792840e-01 6.79460406e-01
6.51262403e-01 -7.58955121e-01 1.02964294e+00 1.13047175e-01
7.93196335e-02 3.76738518e-01 -8.57682943e-01 9.44900215e-01
2.98205405e-01 -4.48797166e-01 -1.06358218e+00 7.94685304e-01
6.98940933e-01 -7.96096623e-01 -9.12703633e-01 -6.52661622e-01
3.10529590e-01 -8.14182162e-01 6.31675541e-01 -5.32635093e-01
1.13976729e+00 3.66541952e-01 6.48431599e-01 -1.42195463e+00
2.76873201e-01 -1.48252547e+00 4.36241388e-01 1.56659722e+00
5.32472432e-01 -5.13396323e-01 8.31042230e-01 1.30815685e+00
-5.15834987e-01 -4.06142294e-01 -5.08883834e-01 -2.56579310e-01
3.80452514e-01 -3.92747700e-01 3.27627718e-01 1.36616564e+00
1.03645742e+00 8.70221555e-01 -8.94739404e-02 -3.98931593e-01
2.55306810e-01 7.22117499e-02 1.07220471e+00 -1.40496051e+00
1.69852570e-01 -6.45444751e-01 -2.89102226e-01 -6.93363130e-01
-1.44340530e-01 -8.51287007e-01 1.70598820e-01 -1.73222995e+00
-1.45029724e-01 -1.67894930e-01 4.74166840e-01 6.14279211e-01
-8.13266039e-02 -2.68055080e-03 2.62352139e-01 -1.35265991e-01
-6.40626073e-01 4.88111936e-02 1.65355778e+00 6.57063797e-02
-4.72587138e-01 1.01491235e-01 -1.59318042e+00 8.41794670e-01
8.84660482e-01 -3.81210625e-01 -6.50703430e-01 1.99743547e-03
8.02111566e-01 4.49717224e-01 1.08855985e-01 -9.93470609e-01
5.20733118e-01 -6.71403348e-01 1.10596806e-01 1.27301430e-02
-2.04241633e-01 -2.78448015e-01 -3.00167292e-01 2.21358150e-01
-9.47554350e-01 1.66900337e-01 3.72443914e-01 1.62483573e-01
-1.98201850e-01 -6.39674902e-01 4.03622538e-01 -3.94885838e-01
8.13109130e-02 -2.43413031e-01 -1.33509469e+00 4.43459868e-01
1.11527932e+00 -3.15799743e-01 -6.60463393e-01 -1.19226182e+00
-3.75282735e-01 5.18020153e-01 1.76295370e-01 7.02519476e-01
4.73806560e-01 -9.21678126e-01 -6.89912558e-01 -1.31505057e-01
3.78866568e-02 1.16881132e-01 3.39583635e-01 4.97027397e-01
-8.13142776e-01 5.44465840e-01 -3.39804143e-01 -2.53114581e-01
-1.37970638e+00 -2.35440925e-01 -2.89557189e-01 -4.37915385e-01
-5.83156347e-01 1.11955976e+00 -4.52410251e-01 -4.52964485e-01
-2.29641963e-02 -1.52191997e-01 -2.39824414e-01 9.20965523e-02
6.77063286e-01 2.27686167e-01 5.45529947e-02 -2.40727030e-02
2.55320728e-01 -2.91231036e-01 -3.53157878e-01 -2.17850670e-01
1.14118373e+00 -1.51964545e-01 1.30041420e-01 6.52835250e-01
7.16948271e-01 3.48884493e-01 -4.33848172e-01 1.79809049e-01
1.62859410e-01 -2.94926822e-01 -2.36244723e-01 -1.25034654e+00
-2.30316386e-01 9.05768991e-01 -2.13380978e-01 6.97675169e-01
1.79147094e-01 -2.05015346e-01 7.50011683e-01 8.33269417e-01
5.65135777e-02 -1.04266167e+00 6.31143272e-01 8.59470010e-01
8.71046543e-01 -1.12925625e+00 5.89086451e-02 -1.11291297e-01
-1.07856119e+00 1.29760063e+00 1.04112077e+00 3.21033627e-01
1.27389848e-01 5.00194244e-02 2.38945693e-01 -2.13657185e-01
-1.29695082e+00 5.66364229e-01 9.47087184e-02 4.49181616e-01
1.05110300e+00 5.95506001e-03 -3.49211067e-01 5.92956603e-01
-9.75272179e-01 -1.55925273e-03 1.84082663e+00 1.24661338e+00
-5.87097943e-01 -1.10100245e+00 -5.94257116e-01 3.84765655e-01
-7.57391572e-01 -2.46013761e-01 -1.05594683e+00 4.96615231e-01
-4.02837396e-01 1.24826658e+00 -2.81158060e-01 -4.12783474e-01
-1.98835526e-02 5.91290534e-01 2.21603096e-01 -1.08859217e+00
-1.03608525e+00 -5.75721145e-01 3.16825718e-01 1.91423833e-01
5.64575382e-02 -5.54408073e-01 -1.10384870e+00 -8.10387552e-01
-4.15306181e-01 8.46219182e-01 7.74217188e-01 1.12354279e+00
4.51062351e-01 6.43415630e-01 1.48582160e-01 -2.61849165e-01
-6.45285487e-01 -1.24603951e+00 -1.22079417e-01 -4.01545800e-02
8.56751949e-03 1.16046399e-01 -2.12966517e-01 1.93284929e-01] | [10.307145118713379, 7.4398908615112305] |
72ad05f4-18f2-45fc-a86c-b7acf8ea2947 | material-recognition-cnns-and-hierarchical | 1706.08685 | null | http://arxiv.org/abs/1706.08685v1 | http://arxiv.org/pdf/1706.08685v1.pdf | Material Recognition CNNs and Hierarchical Planning for Biped Robot Locomotion on Slippery Terrain | In this paper we tackle the problem of visually predicting surface friction
for environments with diverse surfaces, and integrating this knowledge into
biped robot locomotion planning. The problem is essential for autonomous robot
locomotion since diverse surfaces with varying friction abound in the real
world, from wood to ceramic tiles, grass or ice, which may cause difficulties
or huge energy costs for robot locomotion if not considered. We propose to
estimate friction and its uncertainty from visual estimation of material
classes using convolutional neural networks, together with probability
distribution functions of friction associated with each material. We then
robustly integrate the friction predictions into a hierarchical (footstep and
full-body) planning method using chance constraints, and optimize the same
trajectory costs at both levels of the planning method for consistency. Our
solution achieves fully autonomous perception and locomotion on slippery
terrain, which considers not only friction and its uncertainty, but also
collision, stability and trajectory cost. We show promising friction prediction
results in real pictures of outdoor scenarios, and planning experiments on a
real robot facing surfaces with different friction. | ['Yukitoshi Minami Shiguematsu', 'Martim Brandao', 'Kenji Hashimoto', 'Atsuo Takanishi'] | 2017-06-27 | null | null | null | null | ['material-recognition'] | ['computer-vision'] | [ 1.59641609e-01 1.79252788e-01 1.34014159e-01 -7.93041289e-02
-2.70373195e-01 -2.81506300e-01 2.74897993e-01 1.13505997e-01
-3.21891606e-01 1.02887762e+00 -3.03240716e-01 1.90500945e-01
-1.85275644e-01 -1.14762497e+00 -9.72862005e-01 -6.03213251e-01
-7.10226178e-01 1.05732834e+00 6.42545700e-01 -6.91424072e-01
1.64042443e-01 2.48694152e-01 -2.10333157e+00 9.22583640e-02
1.20469928e+00 8.73683929e-01 8.41696501e-01 7.74602652e-01
3.73456120e-01 1.31183341e-01 -7.99254999e-02 -1.04224317e-01
8.81285369e-02 1.81855127e-01 -8.39457870e-01 -1.25458077e-01
5.48982713e-03 -2.16072574e-01 1.24667935e-01 6.97628915e-01
1.89023882e-01 3.23515862e-01 1.24082613e+00 -1.27268887e+00
-9.87542346e-02 1.72335505e-01 -5.15250981e-01 -6.21121585e-01
3.46366823e-01 1.80994153e-01 8.37909222e-01 -4.72052068e-01
8.28162611e-01 1.48036265e+00 8.74173760e-01 4.46653128e-01
-9.37270164e-01 1.71005409e-02 1.78482130e-01 6.40568137e-01
-1.11135983e+00 -2.47370660e-01 4.09301013e-01 -5.16013324e-01
1.15478015e+00 -7.14679509e-02 8.28514516e-01 1.02381468e+00
8.51130366e-01 5.78559220e-01 4.98174220e-01 -1.45928621e-01
6.97737455e-01 -5.98918557e-01 -5.18867970e-01 9.17942524e-01
6.71766579e-01 -4.84636538e-02 -4.05732304e-01 5.96134365e-02
5.81162691e-01 -5.00227273e-01 -5.05531877e-02 -8.30945492e-01
-1.16231191e+00 3.34508300e-01 6.98055089e-01 -3.88899952e-01
-3.15587819e-01 5.34312129e-01 4.18411911e-01 -1.85396403e-01
1.26929268e-01 3.51537526e-01 -5.45576274e-01 -2.44012505e-01
-2.90602446e-01 8.47309470e-01 8.27282727e-01 8.96517098e-01
8.76591921e-01 -1.04256347e-01 5.62451303e-01 8.97891700e-01
4.38559145e-01 1.02148604e+00 1.56088034e-02 -1.40592563e+00
7.56508887e-01 3.98758441e-01 6.97366476e-01 -1.22935104e+00
-9.88934517e-01 4.88366395e-01 -9.19819057e-01 8.70453775e-01
4.93224889e-01 -7.10053891e-02 -1.14052463e+00 1.59626341e+00
3.25760692e-01 -3.97084802e-01 -2.96400487e-03 1.16090310e+00
2.17590198e-01 5.33244193e-01 6.11902513e-02 1.32410109e-01
1.24057341e+00 -8.69772792e-01 -3.63727361e-01 -6.15685940e-01
6.00151896e-01 -1.25138551e-01 1.01933718e+00 5.49630582e-01
-1.14064765e+00 -2.77898729e-01 -1.18847346e+00 -2.42469624e-01
-2.49146506e-01 2.65006870e-01 5.48660457e-01 5.41239325e-03
-8.53327513e-01 1.14288902e+00 -1.36443412e+00 -4.71438944e-01
3.25557441e-01 2.21461013e-01 -2.64865607e-01 2.43983977e-03
-1.22161567e+00 1.37974477e+00 4.63403389e-02 3.25362861e-01
-7.34393418e-01 -8.61116499e-02 -1.02690637e+00 -9.49560851e-03
2.49211118e-01 -9.65368152e-01 1.15303564e+00 -4.75206375e-01
-1.62649810e+00 4.94156480e-01 -9.42260548e-02 -4.06188786e-01
8.14610302e-01 -5.77438116e-01 4.55917209e-01 -4.61431779e-02
2.97098666e-01 8.06732714e-01 5.49150467e-01 -1.54017651e+00
-5.01194298e-01 -2.00262278e-01 -2.80210041e-02 6.41689420e-01
3.86939079e-01 -1.18029833e+00 -2.90091187e-01 1.18381836e-01
1.97525695e-01 -1.02647531e+00 -7.12610662e-01 4.94917274e-01
-3.32335174e-01 -8.39897767e-02 3.55893850e-01 -5.98340034e-01
5.51480711e-01 -1.54264903e+00 6.89085543e-01 2.82333463e-01
-2.62928665e-01 -2.49267399e-01 9.71672013e-02 4.76104289e-01
9.18021679e-01 5.58424070e-02 -8.09829473e-01 -2.21447214e-01
1.65900260e-01 6.03260219e-01 1.05200469e-01 3.49406958e-01
3.42287242e-01 5.83647013e-01 -1.03514493e+00 -5.03610253e-01
1.99842542e-01 4.49906826e-01 -5.46143532e-01 -8.68551284e-02
-5.99209905e-01 1.81559175e-01 -5.90270638e-01 1.00471389e+00
7.91053534e-01 1.44294173e-01 4.45698172e-01 -1.35683417e-01
-3.73462319e-01 1.09736316e-01 -1.12743461e+00 1.66416717e+00
-6.97944224e-01 4.30648744e-01 4.80747968e-01 -8.45345855e-01
8.87776673e-01 -3.49195093e-01 3.16919714e-01 -6.35471165e-01
-1.29634393e-02 5.41763723e-01 -2.53207356e-01 -6.85549796e-01
8.10037494e-01 8.63473117e-02 -2.74936020e-01 -3.07032615e-01
-3.37978005e-01 -8.34332466e-01 -2.18924414e-02 -3.84499013e-01
1.05442286e+00 9.59250629e-01 1.81504443e-01 -5.35587192e-01
1.09816000e-01 3.93573940e-01 5.37802756e-01 4.55067843e-01
-9.63157415e-02 6.24493420e-01 3.10073733e-01 -4.92830068e-01
-1.43174767e+00 -1.44792652e+00 -1.47849724e-01 7.55181432e-01
1.04658675e+00 4.07650210e-02 -5.93455970e-01 9.25127417e-02
4.16402072e-01 2.14882135e-01 -7.08990693e-01 -1.26197040e-01
-9.59050119e-01 -7.06809223e-01 7.27384984e-02 5.44994950e-01
4.56889570e-01 -1.22661936e+00 -1.49808836e+00 5.01976728e-01
-5.64255178e-01 -8.80047321e-01 3.10533077e-01 5.61536431e-01
-4.60826159e-01 -1.08633661e+00 -5.13458788e-01 -6.08805954e-01
3.39205265e-01 1.77383423e-01 1.11895108e+00 2.92874426e-01
-7.73999393e-01 4.26415801e-02 -2.60593683e-01 -1.95527628e-01
-1.28531292e-01 -2.87415963e-02 2.94573605e-01 -7.49299288e-01
-5.70067346e-01 -6.16485238e-01 -6.82902634e-01 6.49734199e-01
-1.89610392e-01 4.37877238e-01 2.25359544e-01 7.88073838e-01
7.12899983e-01 8.15801471e-02 2.62692004e-01 -1.29237458e-01
7.83506483e-02 -6.78216636e-01 -5.36687076e-01 1.28226280e-01
-1.46390259e-01 9.66812111e-03 6.34647489e-01 -1.32769436e-01
-1.07124758e+00 2.48487025e-01 -1.43360868e-01 9.00431126e-02
-1.03828847e-01 2.66303509e-01 -2.50511736e-01 1.23284526e-01
7.70286381e-01 -2.72977799e-01 -2.38527089e-01 -1.14143148e-01
3.83215189e-01 2.83946961e-01 5.80967724e-01 -8.96178186e-01
3.49149704e-01 9.43052232e-01 7.29884267e-01 -1.04483771e+00
-1.98156938e-01 6.10146075e-02 -8.16882789e-01 -4.28251266e-01
9.57509637e-01 -7.49549329e-01 -1.04632556e+00 7.28437066e-01
-1.22296715e+00 -9.23854947e-01 1.59851797e-02 2.38246366e-01
-1.29248023e+00 4.25415039e-01 -3.43426317e-01 -1.11272514e+00
-1.39213189e-01 -1.15653145e+00 1.36229885e+00 2.87300855e-01
-3.63845319e-01 -7.29945362e-01 1.03261858e-01 -6.73708171e-02
1.92017004e-01 8.49705517e-01 7.78528452e-01 5.95280290e-01
-5.32952428e-01 2.91843772e-01 -1.29029319e-01 -2.84946918e-01
1.41243398e-01 3.11108589e-01 -5.47823727e-01 -7.16698766e-02
-5.76873660e-01 -6.35682344e-01 1.11812532e+00 6.55177176e-01
7.48516798e-01 -1.10286519e-01 -7.68285751e-01 5.85718274e-01
1.46829271e+00 7.90784359e-02 5.87791681e-01 7.91634083e-01
7.26982653e-01 1.15687323e+00 1.05106175e+00 6.94053829e-01
7.45677829e-01 8.91916156e-01 1.16271210e+00 1.82251126e-01
1.23859175e-01 -1.39792208e-02 3.77967775e-01 3.19608510e-01
-5.70600808e-01 -6.39843762e-01 -1.20411694e+00 8.14203262e-01
-2.06893349e+00 -7.62600660e-01 -3.40320855e-01 2.07860756e+00
4.80580717e-01 1.81509659e-01 -6.43244311e-02 1.62325390e-02
5.33595920e-01 -3.10474008e-01 -8.25397730e-01 -5.74613214e-01
-4.04309258e-02 -1.35954931e-01 7.37216055e-01 8.30656052e-01
-1.09066522e+00 9.94291425e-01 6.17323971e+00 4.46114153e-01
-9.19766486e-01 -5.57550669e-01 2.36564949e-01 1.36585489e-01
-2.78737992e-01 -1.43784434e-01 -6.11094177e-01 3.99632990e-01
5.61571181e-01 1.48282215e-01 5.53173542e-01 1.02640915e+00
1.39061749e-01 -7.24693835e-01 -8.39406908e-01 3.83521825e-01
-4.41123039e-01 -1.00487244e+00 -7.50973374e-02 -2.34052435e-01
4.83524561e-01 9.43470001e-02 -2.24893048e-01 -1.45438448e-01
5.94611406e-01 -1.00480306e+00 1.06420827e+00 7.96510160e-01
6.92442656e-01 -7.57367492e-01 6.53104126e-01 4.30445433e-01
-1.60412061e+00 -4.77984510e-02 -7.39717782e-01 -4.06156063e-01
7.14705288e-01 3.96527410e-01 -7.39591002e-01 6.53333902e-01
1.04666376e+00 6.79863811e-01 4.53718156e-02 1.02190828e+00
-1.01325214e-01 -8.31237063e-02 -7.48019278e-01 -3.51988941e-01
-2.00259443e-02 -3.15929174e-01 4.48177993e-01 8.64873052e-01
6.02002561e-01 -7.18498603e-02 3.96383218e-02 6.77083910e-01
4.50049937e-01 -1.49965852e-01 -6.15934014e-01 4.58020568e-01
5.09801686e-01 9.22357321e-01 -9.98211384e-01 -6.24360256e-02
1.87075555e-01 1.07605147e+00 5.36794424e-01 1.91808939e-02
-8.04332137e-01 -4.34244633e-01 8.85662556e-01 1.85371920e-01
1.96550906e-01 -6.11182332e-01 -6.24837518e-01 -8.11753333e-01
3.07063580e-01 1.39422834e-01 -1.92870021e-01 -9.96809900e-01
-1.17566121e+00 1.85122952e-01 2.91721313e-03 -1.26270962e+00
-6.35770895e-03 -8.48030984e-01 -4.66780484e-01 6.59688413e-01
-1.44148147e+00 -1.05052984e+00 -5.05324304e-01 -7.45538920e-02
5.66639543e-01 4.10843998e-01 7.52139449e-01 -2.16506630e-01
-1.07952744e-01 3.97566594e-02 2.19443515e-01 -8.31781268e-01
4.46121365e-01 -1.28652644e+00 8.69716346e-01 2.06547186e-01
-8.00940990e-01 -1.10033780e-01 1.00896525e+00 -1.09579301e+00
-1.44812572e+00 -9.25963342e-01 4.74637985e-01 -1.26117766e-01
5.60528278e-01 -3.26585919e-01 -1.04386771e+00 -7.42442459e-02
-3.90817761e-01 -3.43861490e-01 -2.13918477e-01 -4.25348282e-02
2.79766232e-01 2.87378758e-01 -1.29834807e+00 8.47593129e-01
1.56661594e+00 2.10222438e-01 -3.32238972e-01 2.02720016e-01
5.05815387e-01 -6.44969046e-01 -5.47732711e-01 8.59708726e-01
1.19455945e+00 -8.77641320e-01 1.13743603e+00 -3.02853525e-01
6.83730543e-01 -2.87030965e-01 -3.17465156e-01 -1.25117028e+00
-3.60961616e-01 -1.63865864e-01 1.45077929e-01 6.25076234e-01
3.66239876e-01 -8.90300199e-02 9.80068326e-01 2.99163938e-01
-5.99425256e-01 -9.21990752e-01 -1.49471056e+00 -7.71488965e-01
2.53251433e-01 -2.75503308e-01 2.12055042e-01 4.38498318e-01
2.34519109e-01 -1.93003878e-01 -5.26074946e-01 4.62980181e-01
7.34402418e-01 5.27500026e-02 7.00823367e-01 -1.30467129e+00
4.80165705e-02 -3.40037763e-01 -4.68772769e-01 -8.52390468e-01
1.27314389e-01 -4.50485975e-01 1.08998978e+00 -2.15300846e+00
-3.30977440e-01 -7.17767835e-01 5.36485434e-01 4.38132763e-01
-9.13316309e-02 -7.24888891e-02 -9.70506147e-02 2.34804407e-01
-3.73496532e-01 9.62436974e-01 1.24065840e+00 -1.40831277e-01
-3.13238919e-01 -1.97156414e-01 1.97084025e-01 1.14705431e+00
8.22827160e-01 -4.42883298e-02 -2.59426504e-01 -6.35941803e-01
5.04834354e-01 2.79220313e-01 4.02429551e-01 -1.41312623e+00
-4.99551222e-02 -6.86177731e-01 2.62350649e-01 -5.33449829e-01
7.68350780e-01 -7.14732826e-01 8.35658014e-02 8.32406461e-01
7.16852471e-02 -2.95499206e-01 2.72347242e-01 7.86034465e-01
3.50832731e-01 -3.58623974e-02 9.88635242e-01 -3.30215305e-01
-7.75096238e-01 -6.27274737e-02 -7.41513550e-01 -2.50911385e-01
8.29012215e-01 -3.85637432e-01 -4.65383112e-01 -7.69472569e-02
-9.29662347e-01 7.89579391e-01 1.05744457e+00 2.72862643e-01
7.03533292e-01 -1.23952818e+00 -5.07902801e-01 -9.31781083e-02
8.56033294e-04 5.21046579e-01 5.71881354e-01 3.02070796e-01
-1.25083327e+00 -1.86694309e-01 -8.71481657e-01 -8.55883241e-01
-7.84969628e-01 6.21230043e-02 3.58489692e-01 -9.29656178e-02
-5.39836586e-01 8.89659643e-01 2.09322646e-01 -4.71107751e-01
-1.17063291e-01 -6.14584982e-01 -1.90475062e-01 -1.61669150e-01
-1.67503580e-01 7.90532649e-01 -5.71986921e-02 -5.70151865e-01
-6.66537225e-01 1.29015052e+00 5.57030916e-01 -1.81225702e-01
1.26544762e+00 -4.75862354e-01 1.86606213e-01 4.67292905e-01
5.44828057e-01 -3.96061122e-01 -1.76382720e+00 5.31833649e-01
-4.78795134e-02 -1.29232287e-01 -5.71542799e-01 -5.83381891e-01
-3.33722204e-01 9.43621039e-01 4.02847260e-01 -8.06677565e-02
6.70352161e-01 -1.94179639e-01 8.21901917e-01 9.02137041e-01
1.13729727e+00 -1.46520734e+00 -9.25084725e-02 7.54259884e-01
1.07766390e+00 -1.13910544e+00 1.80397838e-01 -6.04460478e-01
-7.11100757e-01 1.23278308e+00 9.11604643e-01 -5.38118243e-01
3.80831361e-01 4.13751125e-01 -1.80755869e-01 1.07928015e-01
-9.21581626e-01 -1.24885008e-01 -3.09986770e-02 9.66850758e-01
-2.28344575e-01 3.32572877e-01 -3.11385453e-01 4.59461391e-01
-3.84091765e-01 -3.68948191e-01 6.11045659e-01 1.23035502e+00
-1.21345246e+00 -7.39347160e-01 -3.06149423e-01 2.78447270e-01
3.35003793e-01 4.41459984e-01 -2.09342882e-01 7.40783632e-01
5.57143927e-01 5.77394664e-01 1.89849228e-01 -5.28622568e-01
4.61843342e-01 -2.75336593e-01 5.30889988e-01 -3.32508624e-01
1.24404982e-01 -3.69067729e-01 4.85546678e-01 -7.71128774e-01
-1.62249446e-01 -9.39824700e-01 -1.88847208e+00 3.66989002e-02
-1.33070275e-01 -1.03314027e-01 8.22348177e-01 8.76537502e-01
6.53893873e-02 3.66689563e-01 3.14930141e-01 -1.88545644e+00
-1.72253609e-01 -7.46681690e-01 -5.42012155e-01 1.79496035e-01
3.61667871e-01 -1.46453798e+00 -3.96226078e-01 1.71010807e-01] | [4.827092170715332, 1.0360214710235596] |
9e66ab64-7e16-4091-b51c-d6ca2af6cc1f | generative-adversarial-exploration-for | 2201.11685 | null | https://arxiv.org/abs/2201.11685v1 | https://arxiv.org/pdf/2201.11685v1.pdf | Generative Adversarial Exploration for Reinforcement Learning | Exploration is crucial for training the optimal reinforcement learning (RL) policy, where the key is to discriminate whether a state visiting is novel. Most previous work focuses on designing heuristic rules or distance metrics to check whether a state is novel without considering such a discrimination process that can be learned. In this paper, we propose a novel method called generative adversarial exploration (GAEX) to encourage exploration in RL via introducing an intrinsic reward output from a generative adversarial network, where the generator provides fake samples of states that help discriminator identify those less frequently visited states. Thus the agent is encouraged to visit those states which the discriminator is less confident to judge as visited. GAEX is easy to implement and of high training efficiency. In our experiments, we apply GAEX into DQN and the DQN-GAEX algorithm achieves convincing performance on challenging exploration problems, including the game Venture, Montezuma's Revenge and Super Mario Bros, without further fine-tuning on complicate learning algorithms. To our knowledge, this is the first work to employ GAN in RL exploration problems. | ['Peng Sun', 'Yong Yu', 'Ming Zhou', 'Weinan Zhang', 'Minghuan Liu', 'Menghui Zhu', 'Weijun Hong'] | 2022-01-27 | null | null | null | null | ['montezumas-revenge'] | ['playing-games'] | [ 1.06208794e-01 4.47648942e-01 -4.20590043e-01 7.85056651e-02
-8.12753379e-01 -7.60011852e-01 6.68295860e-01 -1.84040830e-01
-4.85103101e-01 1.21342635e+00 -9.86783653e-02 -6.20512843e-01
1.08374804e-01 -9.73963439e-01 -8.26838493e-01 -9.02143419e-01
-2.70353675e-01 6.63193226e-01 -1.08340189e-01 -3.65390062e-01
2.02629134e-01 4.35209751e-01 -1.09865415e+00 -3.14161599e-01
1.06684542e+00 8.44440758e-01 5.55277150e-03 7.23480403e-01
2.79008150e-01 7.99811125e-01 -8.35595012e-01 -1.66690141e-01
6.16348684e-01 -1.00838292e+00 -8.13569367e-01 -2.65567154e-01
-2.24923640e-01 -4.81014848e-01 -2.48370305e-01 1.15631378e+00
5.53391695e-01 2.70750523e-01 4.82950121e-01 -1.30533469e+00
-6.60630703e-01 8.77966940e-01 -3.92571390e-01 3.71478945e-01
2.51351744e-01 7.09085524e-01 7.78111517e-01 -1.65886745e-01
5.64652562e-01 1.21073043e+00 1.35820165e-01 9.80605602e-01
-1.23688936e+00 -9.15414095e-01 2.08526149e-01 6.36095032e-02
-8.71249080e-01 -1.79621145e-01 9.57991540e-01 -7.13562146e-02
4.07272398e-01 1.28031656e-01 9.01530385e-01 1.57547617e+00
7.60736242e-02 1.00339127e+00 1.68112242e+00 -2.63836950e-01
1.05665803e+00 1.31460177e-02 -6.78300023e-01 8.12452734e-01
1.27201751e-01 8.59038591e-01 -4.00197476e-01 -1.73534855e-01
9.30332124e-01 -2.55542547e-01 -6.84397593e-02 -5.82615852e-01
-8.91741455e-01 1.21934664e+00 5.56077898e-01 -7.47895613e-02
-4.73276526e-01 4.03842837e-01 2.06900641e-01 5.52499950e-01
5.48682325e-02 1.09297717e+00 -1.43010452e-01 -7.07482636e-01
-6.71448827e-01 3.03133041e-01 6.21995389e-01 6.44735396e-01
6.89967930e-01 4.52820301e-01 -3.25591862e-01 2.01405436e-01
3.23928781e-02 4.67323303e-01 6.92848265e-01 -1.28632855e+00
3.04544181e-01 3.80540967e-01 3.00390720e-01 -4.98668700e-01
-9.76157486e-02 -5.07376730e-01 -5.84601879e-01 8.76785278e-01
2.94105679e-01 -6.99039280e-01 -9.88005698e-01 2.01393700e+00
2.93554604e-01 4.76405444e-03 4.20867771e-01 8.63536537e-01
2.18134969e-01 4.99321133e-01 -1.31246656e-01 -8.34685490e-02
8.39866042e-01 -9.70733345e-01 -4.39435840e-01 -4.79622096e-01
3.55675429e-01 -1.17851533e-01 1.29721045e+00 3.84516120e-01
-1.19684768e+00 -4.10482049e-01 -1.20710003e+00 6.17142022e-01
-3.60681415e-01 2.97961477e-02 1.00752294e+00 6.69866145e-01
-8.41795444e-01 8.26805413e-01 -1.09380531e+00 -1.14465527e-01
7.07192838e-01 3.14180732e-01 8.11599642e-02 2.29478478e-01
-1.36000514e+00 9.40295935e-01 6.76284850e-01 -4.62745391e-02
-1.75277019e+00 -2.39611343e-01 -8.14435542e-01 -4.08295467e-02
8.44542205e-01 -4.05170649e-01 1.28805709e+00 -1.07747567e+00
-1.99298596e+00 2.84012318e-01 2.65880913e-01 -8.24578583e-01
8.76822054e-01 1.28496274e-01 -2.79820532e-01 -6.63491040e-02
8.52390304e-02 7.97045112e-01 1.05037320e+00 -1.29095435e+00
-4.99859840e-01 -1.27357364e-01 4.69882995e-01 2.86521375e-01
3.88524774e-03 -5.94449520e-01 -1.11570591e-02 -6.80132627e-01
-1.36963546e-01 -1.14683461e+00 -4.92957830e-01 -2.38007903e-01
-6.52107656e-01 -2.17996180e-01 6.83337510e-01 -1.76327288e-01
8.75217974e-01 -1.83993673e+00 1.12025663e-01 3.39403242e-01
-6.13788739e-02 4.44371998e-01 -2.53417850e-01 5.29964328e-01
3.52366745e-01 1.23327099e-01 -2.74593860e-01 -5.26223667e-02
2.64046162e-01 5.84116697e-01 -7.19371915e-01 1.95361435e-01
1.17661692e-01 1.20788682e+00 -1.29761183e+00 -7.09812641e-02
-7.21147358e-02 6.37482107e-02 -5.78458607e-01 4.37754631e-01
-5.50523758e-01 8.52799535e-01 -7.49603152e-01 6.58107579e-01
3.25549126e-01 -7.60083720e-02 1.11485556e-01 4.64325011e-01
1.76301599e-01 4.43222940e-01 -8.98293138e-01 1.68892360e+00
-3.70567322e-01 3.50292116e-01 -2.05693141e-01 -8.30330789e-01
1.06907380e+00 -1.70752719e-01 -5.70488423e-02 -8.54668260e-01
2.24759117e-01 4.55920659e-02 9.96411666e-02 -7.29713067e-02
5.83867073e-01 -1.27765164e-01 -3.25876147e-01 7.99279988e-01
-2.11200014e-01 -3.64593476e-01 1.99308872e-01 2.07579985e-01
1.22610915e+00 4.20539796e-01 2.66018540e-01 8.94987117e-03
2.08870798e-01 -3.35354134e-02 6.75606608e-01 1.18658340e+00
-3.61517429e-01 1.21800974e-01 8.73809755e-01 -3.33182991e-01
-8.71137917e-01 -1.26044989e+00 3.11274052e-01 9.23238516e-01
3.01096171e-01 -2.06468016e-01 -6.82917476e-01 -1.27965295e+00
-7.81433657e-02 1.01156533e+00 -1.10175133e+00 -6.49795055e-01
-4.35252100e-01 -2.81932086e-01 6.33776605e-01 4.78980809e-01
8.14002633e-01 -1.73043072e+00 -1.07110453e+00 2.63216384e-02
2.24381909e-01 -5.31024694e-01 -4.71730262e-01 6.05381429e-01
-7.73869753e-01 -1.12042165e+00 -5.96216500e-01 -4.31443393e-01
6.67714477e-01 -3.51130098e-01 9.63025630e-01 -2.96546519e-01
-4.17979760e-03 3.81180719e-02 -3.89642060e-01 -2.67609030e-01
-7.39744246e-01 2.65144408e-01 9.99010652e-02 -4.56335872e-01
2.04527125e-01 -5.47188640e-01 -6.06000245e-01 3.21745127e-01
-7.02708066e-01 -6.09400086e-02 7.63366580e-01 1.08957124e+00
5.68317533e-01 1.72630638e-01 8.03614914e-01 -7.98881173e-01
9.05098379e-01 -4.34476078e-01 -8.21159959e-01 7.08973035e-02
-8.52290928e-01 5.89356959e-01 9.62179482e-01 -6.44450724e-01
-9.02489781e-01 -1.64297014e-01 -9.46782678e-02 -4.24813062e-01
-8.84741172e-02 1.44449279e-01 -5.59257232e-02 -9.15079564e-02
7.52709508e-01 4.49447095e-01 5.64124901e-03 -2.03265488e-01
4.24232423e-01 2.24667698e-01 5.45711935e-01 -7.95980096e-01
9.39344704e-01 3.27046722e-01 7.29662478e-02 -1.40529215e-01
-7.73367107e-01 3.12635303e-01 -1.02252096e-01 3.20814066e-02
5.81051648e-01 -6.46511614e-01 -8.94582808e-01 8.05293843e-02
-5.14861882e-01 -9.45330143e-01 -8.73547375e-01 2.53176749e-01
-9.59406555e-01 7.15457350e-02 -3.06381494e-01 -9.47005153e-01
-1.64707556e-01 -1.26291609e+00 6.67031527e-01 6.33480728e-01
-4.59650159e-02 -7.56060600e-01 3.73988897e-01 9.17130485e-02
4.02834713e-01 5.04136860e-01 6.46091402e-01 -5.64581871e-01
-6.30539477e-01 1.72789052e-01 2.96531439e-01 3.08999479e-01
2.19992906e-01 -3.56117785e-01 -6.80313647e-01 -7.34388590e-01
-1.15109622e-01 -9.20795202e-01 8.70221674e-01 1.00482702e-01
1.37741363e+00 -5.73955119e-01 -1.98900774e-01 7.42720068e-01
1.08410490e+00 7.71068275e-01 7.81196296e-01 6.12171710e-01
2.96788394e-01 2.80739423e-02 8.85628223e-01 6.18716657e-01
1.00510076e-01 4.34828728e-01 8.31070900e-01 2.65506580e-02
2.03651339e-01 -8.86895299e-01 7.17272937e-01 9.64507163e-02
7.11288601e-02 -2.17110217e-01 -4.02628362e-01 4.34606016e-01
-1.68421245e+00 -9.98526931e-01 9.60976362e-01 2.07220793e+00
1.06516016e+00 3.80439758e-01 1.43712133e-01 -7.31496438e-02
4.68539149e-01 3.12673450e-01 -1.32033324e+00 -7.38564968e-01
3.21990848e-02 5.62709272e-01 4.96387333e-01 3.84860188e-01
-9.93609846e-01 1.18874681e+00 5.95559406e+00 1.04656541e+00
-9.89790142e-01 -6.27905428e-02 7.65604913e-01 -2.00132936e-01
-4.74455357e-01 2.06078485e-01 -6.54392838e-01 6.03612244e-01
6.67598426e-01 3.65718119e-02 1.02047265e+00 9.73561406e-01
2.47269142e-02 -2.68178433e-01 -9.64294076e-01 5.66597700e-01
-1.50947243e-01 -1.12950039e+00 -1.85576335e-01 2.08393678e-01
9.84414220e-01 -2.04947054e-01 4.79193777e-01 7.63354599e-01
1.02421677e+00 -1.23792672e+00 5.87422431e-01 2.93810368e-01
7.53073275e-01 -1.27215612e+00 3.67066979e-01 4.33744699e-01
-5.99433541e-01 -2.64654726e-01 -2.87942827e-01 9.76611748e-02
-3.64936069e-02 8.22867975e-02 -1.03621864e+00 1.36904776e-01
4.91774708e-01 3.05598795e-01 -4.55851316e-01 5.89975238e-01
-9.50445175e-01 7.77345181e-01 -1.17408454e-01 -2.15595722e-01
7.41911352e-01 -3.78882021e-01 6.65098190e-01 5.00567675e-01
3.54270071e-01 -1.25902802e-01 1.77560687e-01 1.37329125e+00
-1.37732863e-01 -5.01420140e-01 -7.16915905e-01 -2.97598809e-01
4.48901057e-01 1.07072747e+00 -5.90071619e-01 -1.54519811e-01
3.43312442e-01 1.11486423e+00 5.06839991e-01 2.96403080e-01
-9.78938639e-01 -3.36705506e-01 6.09486759e-01 -2.41804481e-01
5.43905675e-01 4.15873947e-03 7.28079155e-02 -9.89999115e-01
-2.27353364e-01 -1.14319193e+00 3.19776833e-01 -6.25759065e-01
-1.03014362e+00 7.69610822e-01 -2.99956024e-01 -1.09956110e+00
-7.03724325e-01 -1.46007389e-01 -7.87124515e-01 7.37484276e-01
-1.59129059e+00 -7.07543135e-01 -6.23026937e-02 5.50292909e-01
3.74006927e-01 -5.10119915e-01 7.50427663e-01 -5.16659677e-01
-5.57935417e-01 8.81233811e-01 2.63139009e-01 6.02877811e-02
3.01504254e-01 -1.57896745e+00 5.41595399e-01 9.28560972e-01
1.09461226e-01 3.58724385e-01 7.11350024e-01 -7.95699954e-01
-1.27980506e+00 -1.08349895e+00 5.52188940e-02 -1.85451582e-01
5.26494324e-01 -3.34957480e-01 -6.01622164e-01 6.60789192e-01
1.96881369e-01 -2.45925318e-02 2.80532122e-01 -1.59123927e-01
-1.82376519e-01 1.82693452e-01 -1.38028216e+00 7.98845768e-01
9.82544482e-01 -4.19073075e-01 -4.44019765e-01 1.08086586e-01
5.63913226e-01 -5.94308376e-01 -5.87755620e-01 1.86061233e-01
1.78195491e-01 -9.92027164e-01 7.42599130e-01 -8.22117925e-01
3.83598536e-01 -3.67472529e-01 1.83549136e-01 -1.79321480e+00
-3.97009030e-02 -1.18925333e+00 -4.87789243e-01 8.71834099e-01
3.39756995e-01 -7.69162714e-01 1.23444653e+00 8.88424143e-02
2.30308902e-02 -1.01143456e+00 -9.46879089e-01 -1.17597485e+00
2.88710266e-01 9.94978175e-02 9.02845860e-01 7.69508421e-01
-5.23188449e-02 -1.04518905e-02 -5.58014512e-01 -8.05551857e-02
5.90707481e-01 4.18219775e-01 8.97437036e-01 -6.17419124e-01
-7.42962241e-01 -4.43876892e-01 -1.20610965e-03 -1.19804347e+00
1.95151687e-01 -7.56568432e-01 1.72769800e-01 -1.09349215e+00
-2.67361607e-02 -7.79516697e-01 -3.30736011e-01 6.57012224e-01
-1.74482882e-01 -5.15934415e-02 2.80750334e-01 -1.43960984e-02
-7.52736568e-01 1.10034692e+00 1.73881316e+00 -2.32074499e-01
-4.62965578e-01 2.09446713e-01 -9.71971512e-01 3.56197357e-01
1.17141259e+00 -5.38972318e-01 -6.99897170e-01 8.21457058e-02
1.04658924e-01 1.82062626e-01 4.08102334e-01 -9.93774772e-01
-3.88267487e-02 -3.22292775e-01 3.94601315e-01 -3.39435309e-01
3.56977761e-01 -4.26447004e-01 4.82949764e-02 8.36484134e-01
-6.54983222e-01 1.34782255e-01 1.31687090e-01 7.19306171e-01
3.96838300e-02 -3.05337578e-01 8.45713913e-01 -2.34004498e-01
-6.47362173e-01 3.90800327e-01 -4.46760803e-01 5.02093732e-01
1.22384655e+00 -5.21139242e-02 -3.24017763e-01 -6.13655627e-01
-6.15049183e-01 4.55943614e-01 6.72439277e-01 2.58552104e-01
5.56078911e-01 -1.37744749e+00 -3.19360167e-01 4.02543157e-01
-9.87329558e-02 9.73594710e-02 -6.32462353e-02 3.66849095e-01
-2.05556333e-01 1.67616010e-01 -4.18947816e-01 -2.07870215e-01
-5.84024906e-01 7.72767305e-01 3.56404692e-01 -7.21752286e-01
-6.18669212e-01 7.76612997e-01 -1.76276609e-01 -3.65468621e-01
2.41277680e-01 -2.16075927e-01 -5.86963026e-03 -2.22917572e-01
2.77876437e-01 3.78443182e-01 -3.67934644e-01 2.06785630e-02
-1.04903340e-01 6.31939769e-02 -1.53798208e-01 -2.84413248e-01
1.08813465e+00 1.98749125e-01 4.01216000e-01 1.24932788e-01
8.10393214e-01 -1.18472502e-01 -2.03211498e+00 4.72025760e-02
-3.92882794e-01 -4.49926645e-01 5.38245700e-02 -1.25578845e+00
-1.21466398e+00 6.02666974e-01 6.53999865e-01 2.62101650e-01
1.11152220e+00 -1.06069736e-01 7.26028860e-01 5.15910327e-01
6.97949767e-01 -1.15049267e+00 3.19302320e-01 4.09218311e-01
7.87971973e-01 -1.31470108e+00 -2.58965194e-01 4.98983234e-01
-1.02515721e+00 7.45428085e-01 8.06089938e-01 -3.80663127e-01
1.36377245e-01 3.44982781e-02 -1.32586718e-01 -1.69109747e-01
-6.53122663e-01 -3.39670271e-01 -1.55192882e-01 7.59286404e-01
-5.06032050e-01 2.16374412e-01 -6.34330288e-02 2.59507477e-01
-4.67627048e-01 -8.82653445e-02 5.88206410e-01 1.10202408e+00
-3.50723654e-01 -1.24902594e+00 -1.45036653e-01 4.01899397e-01
-1.86140090e-01 4.91391011e-02 -3.58346611e-01 8.69320571e-01
4.22210805e-02 7.38569081e-01 6.24069609e-02 -3.46592307e-01
1.43902318e-03 -2.36408338e-01 4.96868402e-01 -3.06119651e-01
-6.41489863e-01 -2.45249942e-01 -1.10088497e-01 -8.44721437e-01
6.54821191e-03 -6.23307943e-01 -1.25361311e+00 -2.07601786e-01
-5.56863733e-02 6.32547855e-01 4.83112574e-01 6.86364651e-01
3.84573519e-01 6.66047633e-01 1.02859616e+00 -6.36941791e-01
-9.47791517e-01 -6.60390496e-01 -6.82225823e-01 1.69285253e-01
3.84975523e-01 -9.12450790e-01 -4.45669591e-01 -6.36842608e-01] | [3.971839666366577, 1.8104054927825928] |
f3558275-c040-46d5-92d7-69c76d84833a | autoremover-automatic-object-removal-for | 1911.12588 | null | https://arxiv.org/abs/1911.12588v1 | https://arxiv.org/pdf/1911.12588v1.pdf | AutoRemover: Automatic Object Removal for Autonomous Driving Videos | Motivated by the need for photo-realistic simulation in autonomous driving, in this paper we present a video inpainting algorithm \emph{AutoRemover}, designed specifically for generating street-view videos without any moving objects. In our setup we have two challenges: the first is the shadow, shadows are usually unlabeled but tightly coupled with the moving objects. The second is the large ego-motion in the videos. To deal with shadows, we build up an autonomous driving shadow dataset and design a deep neural network to detect shadows automatically. To deal with large ego-motion, we take advantage of the multi-source data, in particular the 3D data, in autonomous driving. More specifically, the geometric relationship between frames is incorporated into an inpainting deep neural network to produce high-quality structurally consistent video output. Experiments show that our method outperforms other state-of-the-art (SOTA) object removal algorithms, reducing the RMSE by over $19\%$. | ['Baoquan Chen', 'Ruigang Yang', 'Rong Zhang', 'Weiwei Xu', 'Peng Wang', 'Chenye Guan', 'Wei Li', 'Jinhui Yu', 'Jin Fang', 'Yuhang Song'] | 2019-11-28 | null | null | null | null | ['video-inpainting'] | ['computer-vision'] | [ 3.01297903e-01 1.36502430e-01 4.91218090e-01 -3.35427016e-01
-4.42504466e-01 -3.74817878e-01 1.70782521e-01 -8.22807252e-01
-1.65996492e-01 7.61559188e-01 4.83422652e-02 -8.67367815e-03
4.82766479e-01 -6.43938720e-01 -1.25768530e+00 -7.20899701e-01
1.66581407e-01 2.59760559e-01 6.79937541e-01 -3.18192273e-01
1.75417066e-01 4.56710160e-01 -1.86354256e+00 1.31148085e-01
1.11973166e+00 8.64773154e-01 7.92906582e-01 6.39558494e-01
5.65855764e-02 9.77894843e-01 -5.85364759e-01 -2.06419349e-01
4.56271857e-01 -3.48005414e-01 -1.17045939e-01 3.83271843e-01
9.00336564e-01 -9.60950732e-01 -8.60244930e-01 8.37422609e-01
5.00877857e-01 3.40133131e-01 3.17047268e-01 -1.35151064e+00
-8.09177384e-02 1.99020013e-01 -6.66974962e-01 1.46140292e-01
1.66832462e-01 3.59808654e-01 2.64600277e-01 -1.20055258e+00
9.30706680e-01 1.33596730e+00 6.11781120e-01 4.86845076e-01
-7.37180114e-01 -7.81489134e-01 3.19526196e-01 4.22798067e-01
-1.23972225e+00 -6.18617237e-01 1.14550042e+00 -3.69788975e-01
4.99210596e-01 1.04334041e-01 7.71540701e-01 1.27708697e+00
4.43864197e-01 1.02405751e+00 8.32870424e-01 -6.22385330e-02
2.97648907e-02 -2.35104263e-02 -2.20688388e-01 8.06725502e-01
2.80366242e-01 3.33129197e-01 -7.15366304e-01 1.34360567e-01
5.02584755e-01 1.25329420e-01 -4.92666125e-01 -5.13072610e-01
-1.10358119e+00 4.99311984e-01 3.03630382e-01 -2.55014032e-01
-1.79198802e-01 5.29334486e-01 1.71309039e-01 -1.22577168e-01
5.05581856e-01 -6.61487877e-02 -2.24830538e-01 1.25415146e-01
-1.03733587e+00 5.06614149e-01 4.52098697e-01 1.26780248e+00
9.30088937e-01 5.64430892e-01 -7.32483109e-03 5.57793856e-01
9.85248089e-02 9.16327596e-01 1.43902183e-01 -1.43463981e+00
6.14599586e-01 1.91578671e-01 3.80555183e-01 -1.05341983e+00
-2.40608275e-01 -3.59409750e-01 -7.53347933e-01 3.57243985e-01
3.12165450e-03 -2.34124064e-01 -1.11205792e+00 1.65547585e+00
4.82893795e-01 4.63494033e-01 2.44703740e-01 1.00021458e+00
9.92507458e-01 7.68440843e-01 -4.59115446e-01 -2.42200240e-01
9.91094470e-01 -1.31476855e+00 -9.11348879e-01 -7.19197273e-01
1.85459718e-01 -7.47242630e-01 7.69219577e-01 1.88277587e-01
-1.02379203e+00 -7.69841254e-01 -1.13165629e+00 -3.01090717e-01
-1.99352503e-01 6.44957125e-02 4.44874078e-01 1.45019695e-01
-7.92077184e-01 5.18279612e-01 -6.53040588e-01 -1.27103582e-01
3.35484862e-01 1.05933314e-02 -2.73565739e-01 -4.81968254e-01
-1.11888742e+00 7.67123818e-01 2.07018524e-01 3.04269820e-01
-1.31824803e+00 -7.40448296e-01 -9.87445593e-01 -7.39042908e-02
7.98257947e-01 -8.02761436e-01 1.00557303e+00 -1.10347235e+00
-1.20748127e+00 4.15491849e-01 -4.79071736e-01 -3.37308139e-01
1.02554917e+00 -5.16263008e-01 -2.69882023e-01 5.28665334e-02
2.79775649e-01 9.10299301e-01 1.05855274e+00 -1.94220388e+00
-6.94666028e-01 -1.67645708e-01 -3.37772332e-02 4.21121716e-01
1.62346393e-01 -3.72078180e-01 -8.52863431e-01 -5.94893456e-01
-8.52888301e-02 -1.27000868e+00 -1.60752177e-01 -6.11149520e-02
-4.05992627e-01 2.98368543e-01 1.51644933e+00 -7.86989272e-01
8.18266153e-01 -2.26578212e+00 6.64402097e-02 -3.04534018e-01
2.60709047e-01 1.58329353e-01 -3.85776758e-02 2.72883773e-01
1.92658693e-01 -5.67929327e-01 -4.01078492e-01 -5.78280985e-01
-1.99465379e-01 5.29881775e-01 -5.99417508e-01 5.04176915e-01
5.11040092e-02 7.26585925e-01 -9.73600328e-01 -5.55567145e-01
4.26427752e-01 4.71857667e-01 -3.58140200e-01 4.15537477e-01
-3.13282400e-01 4.55963999e-01 -3.35467964e-01 7.09371865e-01
1.26087308e+00 1.40544832e-01 -1.47598609e-01 -1.79806367e-01
-2.63805389e-01 -2.15939768e-02 -1.13053524e+00 1.71453464e+00
-1.77369282e-01 1.12689161e+00 3.94544840e-01 -4.22027200e-01
6.07223392e-01 -5.11439517e-02 5.43002129e-01 -6.16124928e-01
7.08013177e-02 2.40386382e-01 -2.17666328e-01 -8.01743031e-01
8.45320761e-01 2.14678451e-01 2.00811878e-01 1.05257995e-01
-3.04208368e-01 -2.41498008e-01 1.88478425e-01 4.10858452e-01
1.13269258e+00 4.37625200e-01 -3.63149881e-01 -1.22878574e-01
2.56863922e-01 8.78278911e-02 1.06877518e+00 5.43347597e-01
-7.16405511e-02 1.14359665e+00 2.35991746e-01 -6.02945507e-01
-1.05451691e+00 -9.72533584e-01 2.27089047e-01 7.83642292e-01
7.09022522e-01 -9.94810611e-02 -8.83208573e-01 -5.68679214e-01
7.90781900e-02 7.90452480e-01 -5.58126032e-01 -6.08498603e-02
-9.25983071e-01 -4.44091260e-01 1.57876730e-01 5.62943697e-01
6.57314122e-01 -1.03020787e+00 -9.55136597e-01 2.47358412e-01
-6.47170067e-01 -1.56739676e+00 -6.83401823e-01 3.62428948e-02
-5.70651412e-01 -1.01571882e+00 -7.19354391e-01 -4.39866573e-01
6.32788956e-01 1.04931116e+00 1.18123901e+00 1.44071907e-01
-2.54373938e-01 1.07497245e-01 -1.57991692e-01 -5.67064881e-01
-4.08516943e-01 -4.09008890e-01 1.13572754e-01 1.23823188e-01
-5.80989346e-02 -4.51076210e-01 -8.64191949e-01 4.33100998e-01
-1.19335949e+00 6.59340978e-01 7.32382417e-01 6.17840886e-01
4.25931036e-01 1.49813712e-01 8.40230472e-03 -8.18592966e-01
-1.49467871e-01 -2.69608378e-01 -6.82728529e-01 -1.03493668e-01
-2.08805069e-01 -1.21843435e-01 5.18621087e-01 -3.20881307e-01
-1.28275621e+00 4.38811243e-01 7.98729807e-02 -9.38726425e-01
1.79621764e-02 -1.85028717e-01 -3.97374481e-01 -8.10097903e-02
3.50915521e-01 3.83989245e-01 -1.52580246e-01 -1.27659842e-01
2.42222458e-01 3.89044195e-01 7.55410850e-01 -2.15443835e-01
1.09980059e+00 1.02978253e+00 1.09337838e-02 -9.60784733e-01
-8.43704045e-01 -2.66442031e-01 -5.96891403e-01 -6.88385844e-01
8.46859753e-01 -1.32366586e+00 -3.17101389e-01 6.47932231e-01
-1.42446959e+00 -6.01157606e-01 9.65178479e-03 3.69084448e-01
-7.72236764e-01 5.64349651e-01 -2.72283792e-01 -7.98085034e-01
-1.89964488e-01 -1.35616708e+00 1.32731950e+00 2.78507560e-01
2.43715122e-01 -3.55456442e-01 -1.92547411e-01 5.00807762e-01
1.75618544e-01 4.65690732e-01 3.47486556e-01 2.30493635e-01
-1.25725746e+00 2.23068386e-01 -3.52871895e-01 3.92756850e-01
1.65024772e-02 4.03016359e-02 -1.13191247e+00 -2.07125977e-01
-1.68941878e-02 -1.71644967e-02 1.15296519e+00 4.98318046e-01
1.08829689e+00 -4.77764495e-02 -6.00068212e-01 9.54473972e-01
1.23669088e+00 2.03328863e-01 7.85145998e-01 1.12436749e-01
1.24647856e+00 7.32003212e-01 1.10044003e+00 3.72369677e-01
4.18674648e-01 5.79533994e-01 6.64427340e-01 -2.41869450e-01
-5.31576931e-01 -2.72783309e-01 6.33139074e-01 5.54341614e-01
-1.02903783e-01 -6.69385612e-01 -5.98699689e-01 5.66956937e-01
-2.10736728e+00 -1.02426374e+00 -4.43367004e-01 1.87026334e+00
3.68114293e-01 1.01064302e-01 -4.07780498e-01 -2.40181983e-01
6.95895195e-01 4.42557812e-01 -8.42144191e-01 1.35407820e-01
-2.77368456e-01 -4.42762285e-01 9.49954033e-01 5.51221669e-01
-9.51544285e-01 1.15653348e+00 5.34736156e+00 7.59617329e-01
-1.09905720e+00 1.30785452e-02 4.81534511e-01 -2.22510800e-01
-3.38620842e-01 -2.49940343e-02 -8.48664522e-01 7.02717543e-01
5.50552309e-01 2.38717064e-01 3.89809638e-01 9.10325170e-01
5.97941041e-01 -4.78319764e-01 -8.06276858e-01 9.34401870e-01
4.47444618e-01 -1.28665650e+00 -1.76281616e-01 5.44214733e-02
1.02195203e+00 3.28285098e-01 -1.29911557e-01 7.75454193e-02
1.93419129e-01 -7.17189848e-01 1.11417544e+00 6.81522965e-01
6.73139215e-01 -7.98213542e-01 6.24160588e-01 2.70505726e-01
-1.23979437e+00 -7.81643763e-02 -4.28537875e-01 2.63559222e-01
5.30091822e-01 8.47415328e-01 -5.25182962e-01 5.37533462e-01
9.80467379e-01 7.57148504e-01 -3.71403396e-01 9.04774427e-01
-1.33374825e-01 1.37612149e-01 -3.47852111e-01 3.91586781e-01
1.88927025e-01 -2.54312545e-01 7.02713072e-01 1.18009675e+00
5.34309864e-01 2.11446717e-01 1.80558398e-01 8.93778145e-01
-1.12975471e-01 -5.24616897e-01 -8.99667442e-01 2.60561377e-01
2.08445176e-01 1.29559028e+00 -7.21439600e-01 -6.21548712e-01
-3.73431474e-01 1.24098933e+00 -2.28454713e-02 5.43788791e-01
-1.25236690e+00 -2.49100298e-01 7.68680811e-01 3.85802180e-01
5.57542086e-01 -4.03070062e-01 3.62624228e-02 -1.18606842e+00
3.74122351e-01 -6.71552002e-01 -2.82378972e-01 -1.37505162e+00
-7.56856382e-01 5.30922472e-01 -4.34308350e-02 -1.44161355e+00
-1.60347462e-01 -1.96818233e-01 -5.66190779e-01 5.78365684e-01
-1.82212269e+00 -1.23947215e+00 -1.04980814e+00 4.51341748e-01
9.88124609e-01 1.16727807e-01 1.15736499e-01 4.66299236e-01
-4.02834982e-01 -1.44148311e-02 2.16698036e-01 -5.73269762e-02
1.02220726e+00 -1.00300777e+00 6.05970442e-01 1.18603075e+00
-3.46409827e-01 1.37467682e-01 9.19558823e-01 -9.33000743e-01
-1.71211863e+00 -1.47774637e+00 4.79948372e-01 -3.73425394e-01
1.98452190e-01 -5.35611331e-01 -8.82733166e-01 5.36624014e-01
2.99586564e-01 1.90520331e-01 -3.40505242e-02 -7.66374707e-01
3.28362547e-02 -2.94591695e-01 -8.94205928e-01 6.37217164e-01
1.44473755e+00 -1.52411133e-01 -2.45714515e-01 3.74213874e-01
9.68392313e-01 -9.04983342e-01 -8.77967402e-02 5.86311996e-01
5.12415290e-01 -1.40997100e+00 1.03050113e+00 9.64729302e-03
5.97940326e-01 -6.54858708e-01 -1.29778311e-01 -1.18603468e+00
1.36068324e-02 -7.96429932e-01 -2.82734454e-01 9.83692408e-01
-1.20739199e-01 -1.63291097e-01 9.68499362e-01 4.83233660e-01
-7.49880016e-01 -4.22765851e-01 -7.87088811e-01 -5.94299614e-01
-5.54033637e-01 -3.66911799e-01 3.44025642e-01 5.07166922e-01
-7.95120180e-01 2.51507193e-01 -7.73721874e-01 4.10695374e-01
7.93837011e-01 3.24329644e-01 1.29646862e+00 -9.21408653e-01
-1.68829486e-01 1.11910455e-01 -9.34993476e-02 -1.29515684e+00
1.64507255e-01 -9.11132693e-02 7.08018243e-01 -1.45908272e+00
1.44858658e-01 -2.78662145e-01 2.15177968e-01 4.28439602e-02
-2.69662559e-01 2.78568625e-01 1.28588006e-01 5.17877936e-02
-5.00815690e-01 8.62186193e-01 1.45387101e+00 -1.71237022e-01
-1.82115152e-01 -7.76894912e-02 -2.71335870e-01 9.76510406e-01
3.84696573e-01 -5.38804054e-01 -4.38613802e-01 -7.52167523e-01
-9.80229303e-02 2.78081954e-01 5.24111450e-01 -1.28876746e+00
1.20978214e-01 -3.89956683e-01 5.28565645e-01 -1.25401032e+00
8.93449903e-01 -9.33062196e-01 2.99764872e-01 4.46821332e-01
2.58805573e-01 1.05291292e-01 1.83360487e-01 8.92910659e-01
-9.20246616e-02 -9.99618843e-02 8.76614809e-01 -2.67265737e-01
-1.00161743e+00 3.60967457e-01 -4.14984345e-01 -1.00558214e-02
1.02129722e+00 -2.68642694e-01 -3.13214302e-01 -5.31962574e-01
-1.80541694e-01 3.29660714e-01 7.48728096e-01 5.51410198e-01
8.54812324e-01 -1.04788721e+00 -5.99520445e-01 3.05851638e-01
1.24324575e-01 5.10809064e-01 6.38922513e-01 6.83126688e-01
-7.85575926e-01 2.06406683e-01 -1.24391109e-01 -6.17090940e-01
-1.31098866e+00 6.11130416e-01 2.14748532e-01 3.05162370e-01
-1.02120852e+00 7.03180611e-01 8.15333009e-01 -6.71302229e-02
2.62374192e-01 -4.32628661e-01 2.78418511e-01 -2.54693300e-01
4.99799758e-01 3.78280789e-01 4.44377922e-02 -7.93641925e-01
-2.11377546e-01 5.36387026e-01 2.04697013e-01 -7.45629594e-02
1.32757080e+00 -3.29025567e-01 1.76357403e-01 2.46188998e-01
1.06294215e+00 2.55986303e-01 -2.02132654e+00 1.17850089e-02
-6.63049757e-01 -7.56972790e-01 4.37419377e-02 -3.31053317e-01
-1.36557472e+00 6.34715617e-01 5.86775005e-01 -1.80784315e-01
9.12122428e-01 -3.24364364e-01 1.33492076e+00 4.51251030e-01
3.19310784e-01 -1.32738006e+00 1.07567564e-01 6.12337291e-01
8.73525500e-01 -1.42926884e+00 1.54001638e-01 -6.14924431e-01
-7.96254218e-01 1.05925286e+00 8.62955451e-01 -2.68683255e-01
2.72561699e-01 3.84642124e-01 1.78485677e-01 -9.25544947e-02
-5.80683649e-01 -2.28884444e-01 -6.04226105e-02 6.79582059e-01
-2.62090415e-01 -2.54775673e-01 -8.71486031e-03 9.16520357e-02
-1.73818037e-01 -2.02500764e-02 1.00226521e+00 9.14783597e-01
-5.71857929e-01 -7.36216605e-01 -5.26910186e-01 1.29650563e-01
5.99382743e-02 -4.09600651e-03 -3.82883251e-01 8.11737239e-01
5.07287562e-01 9.35466051e-01 8.41457695e-02 -4.41593319e-01
2.36764580e-01 -2.45778710e-01 3.01476300e-01 -1.85448900e-01
3.92971421e-03 1.22313671e-01 1.02291450e-01 -8.65946770e-01
-3.59162539e-01 -6.39287233e-01 -1.30504036e+00 -4.83733833e-01
-2.72263467e-01 -2.37546206e-01 7.79173195e-01 8.67106140e-01
5.84403276e-01 8.73506427e-01 6.44058704e-01 -1.45235074e+00
1.78128570e-01 -7.91188359e-01 -4.73077446e-01 3.55589449e-01
6.43144429e-01 -1.00656390e+00 -4.41913575e-01 2.12586924e-01] | [8.990234375, -2.4878571033477783] |
4f4833bf-7d49-4a65-835f-cd68c2fd37f0 | implicitly-normalized-forecaster-with | 2305.06743 | null | https://arxiv.org/abs/2305.06743v2 | https://arxiv.org/pdf/2305.06743v2.pdf | Implicitly normalized forecaster with clipping for linear and non-linear heavy-tailed multi-armed bandits | The Implicitly Normalized Forecaster (INF) algorithm is considered to be an optimal solution for adversarial multi-armed bandit (MAB) problems. However, most of the existing complexity results for INF rely on restrictive assumptions, such as bounded rewards. Recently, a related algorithm was proposed that works for both adversarial and stochastic heavy-tailed MAB settings. However, this algorithm fails to fully exploit the available data. In this paper, we propose a new version of INF called the Implicitly Normalized Forecaster with clipping (INF-clip) for MAB problems with heavy-tailed reward distributions. We establish convergence results under mild assumptions on the rewards distribution and demonstrate that INF-clip is optimal for linear heavy-tailed stochastic MAB problems and works well for non-linear ones. Furthermore, we show that INF-clip outperforms the best-of-both-worlds algorithm in cases where it is difficult to distinguish between different arms. | ['Nikita Kornilov', 'Alexander Gasnikov', 'Eduard Gorbunov', 'Alexander Nazin', 'Nikolay Kutuzov', 'Yuriy Dorn'] | 2023-05-11 | null | null | null | null | ['multi-armed-bandits'] | ['miscellaneous'] | [-8.16404745e-02 3.05243246e-02 -6.53720379e-01 -3.09442729e-01
-1.02549720e+00 -1.07090557e+00 2.90482819e-01 -7.78333396e-02
-4.37158495e-01 1.29606545e+00 -1.23963188e-02 -6.38866305e-01
-4.47686553e-01 -7.30728805e-01 -1.14219272e+00 -9.09515560e-01
1.29798427e-01 8.75702858e-01 -1.16932988e-01 -3.30144674e-01
-3.18182111e-02 5.57654142e-01 -7.12400258e-01 1.76283851e-01
8.95617545e-01 1.14030147e+00 -2.03236878e-01 5.76378882e-01
1.52922822e-02 8.30325246e-01 -5.39870381e-01 -9.91227567e-01
6.64819121e-01 -3.94754946e-01 -7.27196872e-01 -2.14231536e-01
2.67063886e-01 -7.12941587e-01 -1.81560561e-01 1.14085042e+00
3.72357965e-01 2.92284101e-01 6.09260976e-01 -1.47841048e+00
-4.77153122e-01 1.05131030e+00 -7.75571883e-01 3.14884394e-01
2.87990142e-02 -1.40793279e-01 1.05020010e+00 -2.58157134e-01
2.67136872e-01 1.30982471e+00 4.87527519e-01 6.03357732e-01
-1.21399856e+00 -6.56714320e-01 5.63246131e-01 -8.14057142e-02
-6.54382110e-01 -2.41874754e-02 6.17180169e-01 -3.96466106e-01
2.01180473e-01 6.84543550e-01 2.76978523e-01 1.08537126e+00
-2.45445207e-01 1.29104757e+00 1.59812987e+00 -3.09313416e-01
4.37626779e-01 -1.04683405e-02 1.82856545e-01 2.93638647e-01
2.83545583e-01 4.51122046e-01 -1.79123417e-01 -7.45695233e-01
4.95865464e-01 3.07260036e-01 -2.24361464e-01 -3.35442007e-01
-9.54160333e-01 1.13467121e+00 3.04890484e-01 1.09127974e-02
-7.58558631e-01 2.27135196e-01 1.00150585e-01 5.43681026e-01
6.77600443e-01 2.96110511e-01 -3.89928341e-01 2.91526057e-02
-8.85212183e-01 9.06476259e-01 7.68148959e-01 9.29015040e-01
2.64851391e-01 1.52805626e-01 -8.47873688e-01 4.86520737e-01
-8.91560912e-02 8.90929997e-01 -6.56131953e-02 -1.00879371e+00
9.12246525e-01 -3.34423304e-01 1.11374247e+00 -7.34471440e-01
-2.52652228e-01 -7.89069474e-01 -6.22964799e-01 2.80717075e-01
9.19266880e-01 -5.33628106e-01 -7.28688598e-01 2.03075552e+00
3.67391050e-01 -1.33965230e-02 -1.34739354e-01 1.15093577e+00
6.89760142e-04 5.07800639e-01 -1.41054884e-01 -5.96313238e-01
8.25682461e-01 -1.12190259e+00 -7.75170684e-01 -3.13657522e-01
3.08301926e-01 -5.39199769e-01 7.24631965e-01 4.91006196e-01
-1.18246245e+00 1.97951838e-01 -7.43645847e-01 5.69803417e-01
-5.58657087e-02 -4.81066853e-01 7.20645964e-01 1.14771307e+00
-4.25623238e-01 5.07562399e-01 -5.58888674e-01 4.17384207e-01
5.59832692e-01 3.54438633e-01 -9.26151052e-02 -3.61121893e-01
-1.10925293e+00 7.25585878e-01 1.05933219e-01 -2.07954310e-02
-1.23311651e+00 -5.48716307e-01 -3.69971305e-01 -9.45501328e-02
8.88394594e-01 -7.24644363e-01 1.57908177e+00 -1.24268687e+00
-1.66203761e+00 3.58032674e-01 -9.42065865e-02 -8.17741454e-01
1.17156518e+00 -3.38656455e-01 9.74896178e-02 -5.22325858e-02
-4.84058596e-02 -1.57767043e-01 9.41570222e-01 -1.24332023e+00
-5.68223000e-01 -6.07217014e-01 8.71239901e-01 3.48579846e-02
-3.58427651e-02 1.59632146e-01 3.30414236e-01 -8.72466862e-01
-3.21147203e-01 -1.21096742e+00 -6.03164196e-01 -2.32387334e-01
-6.09385014e-01 2.37610862e-01 2.52913088e-01 -4.89185661e-01
8.69732559e-01 -1.88886440e+00 1.21890232e-01 3.26132178e-01
-1.99524194e-01 2.99819946e-01 -5.55512942e-02 4.51844484e-01
1.45793036e-01 2.48396955e-02 -4.21817265e-02 -4.24934179e-01
4.27565962e-01 3.74901503e-01 -7.47597873e-01 7.31170356e-01
-4.23799813e-01 7.58035064e-01 -9.40374911e-01 -1.17797963e-01
-1.44275010e-01 -3.04894239e-01 -5.48241496e-01 2.10201487e-01
-5.46091139e-01 6.44974768e-01 -7.01409698e-01 5.73304951e-01
9.90367830e-01 5.73898330e-02 3.78750749e-02 3.98152560e-01
2.89846450e-01 -3.99450839e-01 -1.00111341e+00 1.16347671e+00
-5.18702984e-01 -9.86826047e-02 5.67518175e-01 -1.27891874e+00
4.06837434e-01 3.28108817e-01 2.28310242e-01 -3.16649526e-01
3.12744230e-01 3.81403834e-01 -7.71068633e-02 -2.10880116e-01
3.40369374e-01 -8.90236735e-01 -2.01377466e-01 6.17456317e-01
-4.21725005e-01 -4.86145618e-05 1.54361069e-01 2.32552096e-01
8.50092471e-01 -2.54249960e-01 1.36182576e-01 -6.34392649e-02
3.07565808e-01 -2.07049668e-01 4.88004327e-01 1.43870950e+00
-3.13007563e-01 3.96320373e-01 7.29558349e-01 -3.04279834e-01
-6.67442679e-01 -1.06517327e+00 2.53207177e-01 1.57992613e+00
2.23416433e-01 3.79996032e-01 -5.55348039e-01 -1.01644313e+00
5.10576129e-01 1.01426768e+00 -7.52650619e-01 3.06967348e-01
-1.25808418e-01 -7.61238098e-01 3.06201071e-01 3.64110172e-01
2.48845249e-01 -4.40510929e-01 -8.02269876e-02 3.58190417e-01
-1.89315185e-01 -9.80756044e-01 -6.19412780e-01 1.71987116e-02
-6.04314387e-01 -9.17826712e-01 -1.25652039e+00 -5.54591082e-02
4.36103284e-01 2.78166205e-01 7.95942008e-01 -5.16796470e-01
5.19682527e-01 3.46086383e-01 -3.70112777e-01 -7.27589786e-01
-3.62462401e-01 -1.83759898e-01 9.02734101e-02 4.27462757e-01
-2.99035430e-01 -4.01679188e-01 -6.34485066e-01 6.39990568e-01
-9.01513457e-01 -2.67582953e-01 4.77925599e-01 1.12894464e+00
6.33534074e-01 -2.30045319e-01 9.11193669e-01 -1.17090702e+00
7.11117506e-01 -6.80340409e-01 -8.61109972e-01 4.17275280e-01
-2.56087005e-01 1.72430560e-01 9.56869841e-01 -7.61677980e-01
-1.09062696e+00 -1.15290970e-01 4.92567644e-02 -6.62336767e-01
1.48407593e-01 4.10584062e-01 -3.88154313e-02 -2.17381626e-01
4.33118045e-01 2.78001800e-02 4.89864964e-03 -4.29833770e-01
5.94728231e-01 6.71333134e-01 3.62955064e-01 -1.07678568e+00
8.04703891e-01 7.36318886e-01 3.07530731e-01 7.01111555e-02
-1.55450392e+00 -4.80072061e-03 1.44862801e-01 -3.80068310e-02
2.51504034e-01 -5.44085741e-01 -9.42209303e-01 3.28450561e-01
-8.26087952e-01 -4.03202623e-01 -3.13735098e-01 6.30475581e-01
-1.04066765e+00 2.75048912e-01 -5.18663168e-01 -1.55697238e+00
-2.81631380e-01 -1.13465285e+00 4.51250851e-01 4.58835438e-02
3.87425691e-01 -7.88843334e-01 -7.63234170e-03 7.89720893e-01
5.98675609e-01 4.72612679e-01 5.72150707e-01 -9.58905816e-01
-4.07618105e-01 -4.95040148e-01 7.51191825e-02 3.63134086e-01
-2.75510579e-01 -6.23324454e-01 -6.25699818e-01 -4.60976809e-01
1.19935885e-01 -5.38025737e-01 7.62653410e-01 7.09937572e-01
1.26000559e+00 -9.00912166e-01 -1.06939785e-01 3.90265435e-01
1.07203519e+00 1.86647087e-01 1.81147143e-01 5.30034304e-01
1.27164602e-01 3.60435784e-01 1.13151419e+00 7.67315805e-01
8.24613050e-02 5.80435336e-01 9.94146228e-01 2.44262770e-01
7.89366066e-01 -1.52635753e-01 3.10255468e-01 -1.75494626e-01
-2.49904469e-01 -4.70268935e-01 -3.24817806e-01 5.87036192e-01
-2.17579055e+00 -1.15701437e+00 8.88532475e-02 2.44475198e+00
8.63316774e-01 2.68278927e-01 5.82295656e-01 7.15113711e-04
8.70551229e-01 1.43741176e-01 -8.18384886e-01 -7.24889874e-01
-2.89714605e-01 1.40850082e-01 9.16779160e-01 5.78388810e-01
-1.17155313e+00 5.91267467e-01 6.36717272e+00 1.28908622e+00
-5.76848686e-01 5.74696362e-01 8.66911054e-01 -5.71744263e-01
-3.58464330e-01 -9.31757465e-02 -5.77552736e-01 7.60704875e-01
5.89358330e-01 -4.34657753e-01 7.63432682e-01 1.16368341e+00
4.36336733e-02 -1.90283265e-03 -9.21919823e-01 7.45830774e-01
-2.81803071e-01 -1.14588571e+00 -2.96787530e-01 9.68911797e-02
1.21656656e+00 -1.85452968e-01 5.36582649e-01 3.28586876e-01
9.16239500e-01 -9.55012977e-01 9.37619209e-01 4.33176070e-01
5.66244185e-01 -1.25022149e+00 1.04459977e+00 8.38394821e-01
-4.10211593e-01 -5.83067238e-01 -2.08797410e-01 -2.03695148e-01
2.43933573e-01 6.82007015e-01 -2.30178148e-01 7.40506411e-01
8.27364549e-02 -1.04308069e-01 2.87398309e-01 1.01899147e+00
-2.31945083e-01 7.37583399e-01 -5.82877517e-01 -1.28742248e-01
7.68727660e-01 -1.33227974e-01 6.99071109e-01 8.36939573e-01
3.22132826e-01 2.19686180e-01 4.02677387e-01 5.22967100e-01
-1.20685801e-01 7.11545125e-02 -3.69901448e-01 9.79377888e-04
3.14913750e-01 8.82581830e-01 -1.73292100e-01 -2.32007578e-01
-2.98891515e-01 7.37781584e-01 2.64971584e-01 4.20147091e-01
-1.10695744e+00 -1.33821080e-02 6.50713146e-01 -6.12934791e-02
5.14060676e-01 8.04702714e-02 -6.63858503e-02 -9.18986857e-01
9.48124528e-02 -8.68670106e-01 8.21016014e-01 -3.81975085e-01
-1.69748509e+00 2.33815745e-01 8.79788622e-02 -1.02118123e+00
-3.55752409e-01 -4.58104551e-01 -4.08850491e-01 8.22185934e-01
-1.47980201e+00 -1.29135847e+00 3.82077605e-01 9.44741607e-01
3.10994416e-01 -1.20616361e-01 5.26918650e-01 -3.59846056e-02
-4.05117065e-01 8.71571183e-01 9.54163432e-01 -2.37107098e-01
5.01370847e-01 -1.39699113e+00 -3.58184665e-01 7.15205848e-01
-3.37478593e-02 3.58534962e-01 1.14073241e+00 -3.20038080e-01
-1.46979690e+00 -9.78775024e-01 6.74626902e-02 -8.88745561e-02
9.61275160e-01 -4.89248931e-02 -2.47468397e-01 9.00637865e-01
1.22952461e-01 3.33447248e-01 6.98687673e-01 5.26233539e-02
-4.57229942e-01 -2.73914546e-01 -1.58535647e+00 3.85178179e-01
7.99747765e-01 -8.87270197e-02 -3.64517421e-01 9.41101551e-01
6.18530869e-01 -7.01100707e-01 -7.50256717e-01 1.13353141e-01
4.85184848e-01 -8.34659517e-01 9.85873103e-01 -1.18626952e+00
2.76421070e-01 1.11040600e-01 -4.46302503e-01 -1.51358581e+00
-2.68448919e-01 -1.18873608e+00 -4.25362110e-01 8.61133635e-01
3.89549285e-01 -1.11599290e+00 7.71592498e-01 6.94783151e-01
1.82557568e-01 -8.59916210e-01 -1.33554995e+00 -1.44283283e+00
5.23220658e-01 -4.67654943e-01 7.66909838e-01 7.08539426e-01
-3.36735919e-02 -2.45783508e-01 -1.07697558e+00 2.53918499e-01
8.34574044e-01 7.27668285e-01 8.87576103e-01 -7.20503926e-01
-9.17338908e-01 -3.60680193e-01 6.15552776e-02 -1.20127070e+00
4.17042583e-01 -6.12686336e-01 -2.58626968e-01 -1.11759412e+00
2.86383897e-01 -5.60422719e-01 -5.49613953e-01 3.66220176e-01
-2.55227476e-01 5.23002120e-03 6.05548561e-01 5.87071963e-02
-6.86371326e-01 4.69432414e-01 1.38393390e+00 -3.31978351e-01
7.75440112e-02 9.67257619e-01 -8.43328893e-01 5.01809239e-01
8.60376954e-01 -5.95429778e-01 -2.66703814e-01 -2.02828288e-01
2.19727844e-01 6.52043879e-01 1.83564454e-01 -4.41133857e-01
-4.31403294e-02 -9.23760355e-01 -7.20037818e-02 -7.42579937e-01
4.16506559e-01 -9.39572453e-01 2.05476746e-01 5.15638530e-01
-5.58850825e-01 -3.42507064e-01 -1.42052099e-01 1.06023765e+00
1.19564958e-01 -5.96283019e-01 8.22960138e-01 -2.63860017e-01
2.80435562e-01 5.55041552e-01 -2.77187556e-01 3.13321471e-01
1.22679687e+00 4.40313637e-01 -6.33483171e-01 -1.05438137e+00
-9.78313446e-01 4.22318548e-01 3.18161584e-02 -1.59706622e-01
1.60162121e-01 -1.31467760e+00 -7.99090505e-01 -4.09263968e-01
-3.17566723e-01 -1.96184367e-01 4.54428077e-01 1.02397370e+00
-1.43121228e-01 4.67930853e-01 1.52805567e-01 -5.79458140e-02
-1.08750308e+00 1.21065915e+00 2.37649679e-01 -7.86219478e-01
-3.21611464e-02 8.01062703e-01 1.69369400e-01 -1.21376224e-01
1.60511732e-01 -1.20989773e-02 2.82849282e-01 -7.84290731e-02
6.15540028e-01 5.24721861e-01 -2.87150532e-01 -5.15817463e-01
-3.76277305e-02 3.99189107e-02 -2.30270460e-01 -5.26101887e-01
1.37490606e+00 -9.77997184e-02 1.28411204e-01 2.08222568e-01
6.41359091e-01 2.91183233e-01 -1.24424934e+00 -5.70352077e-01
-3.06126475e-01 -9.61357594e-01 -8.75812471e-02 -8.93020689e-01
-1.24071920e+00 6.94214463e-01 2.76908845e-01 5.27137339e-01
1.00379908e+00 -1.93675756e-01 8.23478699e-01 3.78849834e-01
8.04995656e-01 -8.14138472e-01 -2.35125974e-01 3.63870323e-01
9.38658178e-01 -1.08882332e+00 6.96320236e-02 -1.71164960e-01
-8.72363150e-01 8.48334372e-01 1.81339055e-01 -8.21766555e-02
4.74425793e-01 1.55595168e-01 -2.78244972e-01 3.80804747e-01
-4.39130992e-01 -3.74509156e-01 -2.97186729e-02 6.45298421e-01
-4.22661245e-01 4.97296631e-01 -5.89750469e-01 1.05860364e+00
-2.26462021e-01 -5.74685168e-04 5.00118971e-01 1.06415200e+00
-2.49275133e-01 -1.04872370e+00 -7.75178790e-01 5.68079114e-01
-1.12361062e+00 1.43975869e-01 -2.83658981e-01 5.61833262e-01
-2.52829731e-01 1.19142294e+00 -4.64664817e-01 -6.99543133e-02
1.85938656e-01 -2.15030566e-01 6.52834535e-01 -7.98234567e-02
-6.77639604e-01 7.52193853e-02 1.03135578e-01 -4.04753476e-01
-4.68157589e-01 -5.60183287e-01 -5.21046340e-01 -6.32931828e-01
-5.94479918e-01 3.07859838e-01 5.00072777e-01 1.06935763e+00
1.21260330e-01 3.42974365e-02 1.14341462e+00 -6.27278805e-01
-1.43765724e+00 -9.17672396e-01 -9.68026340e-01 2.97746629e-01
5.76963186e-01 -7.52949595e-01 -5.64058661e-01 -5.84143281e-01] | [4.5424394607543945, 3.3402926921844482] |
22b549ea-892d-4ab8-a2a2-54633c7f3629 | classification-specific-parts-for-improving | 1909.07075 | null | https://arxiv.org/abs/1909.07075v1 | https://arxiv.org/pdf/1909.07075v1.pdf | Classification-Specific Parts for Improving Fine-Grained Visual Categorization | Fine-grained visual categorization is a classification task for distinguishing categories with high intra-class and small inter-class variance. While global approaches aim at using the whole image for performing the classification, part-based solutions gather additional local information in terms of attentions or parts. We propose a novel classification-specific part estimation that uses an initial prediction as well as back-propagation of feature importance via gradient computations in order to estimate relevant image regions. The subsequently detected parts are then not only selected by a-posteriori classification knowledge, but also have an intrinsic spatial extent that is determined automatically. This is in contrast to most part-based approaches and even to available ground-truth part annotations, which only provide point coordinates and no additional scale information. We show in our experiments on various widely-used fine-grained datasets the effectiveness of the mentioned part selection method in conjunction with the extracted part features. | ['Dimitri Korsch', 'Paul Bodesheim', 'Joachim Denzler'] | 2019-09-16 | null | null | null | null | ['fine-grained-visual-categorization'] | ['computer-vision'] | [ 2.48690784e-01 1.36072144e-01 -1.25591174e-01 -3.69836122e-01
-8.33091438e-01 -7.84817576e-01 8.45521867e-01 8.22147012e-01
-5.56274891e-01 6.34350359e-01 5.20925149e-02 4.06674385e-01
-3.34768891e-01 -7.35520184e-01 -5.61640620e-01 -7.62009323e-01
2.16122605e-02 7.36468911e-01 8.09964001e-01 -7.66685512e-03
6.04088962e-01 9.95918572e-01 -2.10551906e+00 3.83731365e-01
6.55317843e-01 1.30086660e+00 2.71900982e-01 4.99262005e-01
-4.43738490e-01 5.63149035e-01 -4.85078752e-01 -7.32847601e-02
2.71579057e-01 -1.44736424e-01 -8.91836584e-01 6.44404769e-01
4.54545408e-01 9.22605395e-02 4.59456563e-01 1.02171910e+00
1.65952444e-01 2.14907855e-01 9.47859943e-01 -9.20316517e-01
1.57365538e-02 2.63561815e-01 -5.47421157e-01 5.90805672e-02
2.18159884e-01 -6.85812309e-02 9.27433312e-01 -8.48184049e-01
6.88377321e-01 9.63783383e-01 6.91383123e-01 2.53746539e-01
-1.33632803e+00 -1.47375628e-01 5.54455578e-01 2.31947318e-01
-1.64407098e+00 -3.56288761e-01 1.00011325e+00 -7.04408109e-01
7.72474766e-01 3.92016679e-01 5.83941817e-01 4.08239543e-01
7.23472622e-04 5.02776384e-01 1.40482473e+00 -5.50120592e-01
5.50010145e-01 4.19058830e-01 3.61594945e-01 5.97828567e-01
2.98554450e-01 -1.11212216e-01 -3.83899242e-01 -2.18507066e-01
5.57327092e-01 1.64875418e-01 -3.26351136e-01 -8.08465123e-01
-1.21224129e+00 6.75188482e-01 8.15127730e-01 6.02511048e-01
-7.98247039e-01 3.50038745e-02 2.12854609e-01 -1.73270047e-01
5.90366066e-01 1.24612562e-01 -5.61833560e-01 2.37932369e-01
-1.23766899e+00 -1.62470359e-02 6.34622395e-01 6.78394616e-01
1.41662550e+00 -4.86112803e-01 -3.63231272e-01 8.38376701e-01
2.64690369e-01 -1.48029169e-02 5.58748245e-01 -6.10459805e-01
1.18403420e-01 1.12299228e+00 1.99854687e-01 -8.89183998e-01
-3.80530775e-01 -6.08016193e-01 -8.09992135e-01 7.20333695e-01
6.51057363e-01 4.88508314e-01 -1.10740650e+00 1.42281342e+00
7.12801516e-01 -4.16706741e-01 -6.35436952e-01 9.34310496e-01
4.77706909e-01 6.95559159e-02 3.12325090e-01 -1.78799331e-02
1.67393875e+00 -8.45797062e-01 -1.96228996e-01 -2.06999883e-01
2.71944404e-01 -6.64229751e-01 7.39246488e-01 5.27211189e-01
-7.58799374e-01 -7.36867130e-01 -9.23004806e-01 8.47817399e-03
-7.04149604e-01 5.07968605e-01 4.24272984e-01 4.99903798e-01
-1.06447911e+00 6.14338338e-01 -6.78394556e-01 -3.81842852e-01
7.10220933e-01 2.78208733e-01 -7.10403264e-01 1.91841245e-01
-3.57887894e-01 7.35166669e-01 4.68053162e-01 8.39492157e-02
-6.07216120e-01 -4.72137421e-01 -6.78042352e-01 2.39200398e-01
3.31890285e-01 -6.02495253e-01 8.69389236e-01 -1.29864287e+00
-1.21215284e+00 1.01612318e+00 -3.26385677e-01 -4.04171675e-01
6.12601101e-01 1.18328676e-01 2.85080642e-01 5.59327185e-01
2.36918643e-01 7.65453696e-01 1.10647678e+00 -1.55006194e+00
-8.63789737e-01 -6.05297029e-01 8.75067711e-02 2.21166417e-01
-9.55156982e-02 -2.33583987e-01 -4.10478801e-01 -6.60345733e-01
6.06328249e-01 -6.27309620e-01 -3.98752004e-01 3.44641417e-01
-4.45350826e-01 -3.28116953e-01 5.54208696e-01 -4.44780380e-01
8.52678597e-01 -1.83660758e+00 1.18246377e-01 3.83579642e-01
3.82446676e-01 -1.64326698e-01 2.34429896e-01 3.52416635e-01
1.03691839e-01 -9.46742967e-02 -2.75262386e-01 -2.05509692e-01
-5.15631679e-03 -2.66750515e-01 9.22023579e-02 7.36562252e-01
3.73868227e-01 7.18846440e-01 -5.98289728e-01 -8.64282429e-01
6.42565727e-01 5.55145681e-01 -3.09262514e-01 -6.05240799e-02
-9.96982530e-02 2.48676151e-01 -6.44513965e-01 5.99231839e-01
6.93721533e-01 -2.71304667e-01 -8.23828280e-02 -4.73341048e-01
-2.70878226e-01 -7.90435001e-02 -1.39533567e+00 1.45964348e+00
-5.01775861e-01 3.56502980e-01 2.61491984e-01 -9.98302460e-01
7.80910373e-01 9.06735659e-02 5.07843375e-01 -3.33101511e-01
2.03799680e-01 2.72049040e-01 -2.38204286e-01 1.16192602e-01
5.21981955e-01 -1.93881646e-01 7.07743317e-02 2.61277497e-01
1.07258558e-01 -6.12557456e-02 2.60558993e-01 -1.44336373e-02
9.27060008e-01 4.64877695e-01 8.18621576e-01 -5.30336678e-01
8.68322372e-01 3.39989126e-01 1.44439533e-01 6.29602313e-01
-1.91338390e-01 9.17869866e-01 2.66077995e-01 -2.30560049e-01
-6.68440521e-01 -7.02158153e-01 -3.59965593e-01 8.81867290e-01
4.52127904e-01 -2.83866853e-01 -9.63101625e-01 -1.07888031e+00
3.00165378e-02 3.98240328e-01 -1.14362860e+00 1.20339185e-01
-2.43427336e-01 -4.23863411e-01 -1.82370186e-01 4.17099237e-01
4.46217149e-01 -1.21494555e+00 -9.67367291e-01 1.87979728e-01
-2.27701329e-02 -6.77247763e-01 -3.20756257e-01 6.81026459e-01
-9.39805508e-01 -1.24881828e+00 -9.23945129e-01 -5.45920670e-01
1.11399817e+00 3.39729548e-01 1.19240880e+00 6.18715659e-02
-5.25028527e-01 4.15653586e-01 -4.97959524e-01 -5.11558875e-02
-1.60376638e-01 -1.14923425e-03 -3.96007061e-01 4.01674598e-01
2.39539281e-01 -3.96754473e-01 -7.25436866e-01 5.12389123e-01
-5.76163590e-01 -1.89223617e-01 1.07234859e+00 7.83988476e-01
1.10184503e+00 2.57617623e-01 1.99306399e-01 -7.70248175e-01
8.61743391e-02 2.72290315e-02 -7.33269274e-01 3.21096629e-01
-4.61286426e-01 2.09730208e-01 4.38178480e-01 -2.72503018e-01
-1.11173713e+00 4.36890334e-01 -1.81727782e-02 -3.24302733e-01
-8.45878065e-01 4.35224883e-02 -2.21327856e-01 -3.55656743e-01
7.30865836e-01 2.84105390e-01 -2.90672272e-01 -6.68511391e-01
2.15575799e-01 5.24889052e-01 1.47605658e-01 -4.24200922e-01
6.69671357e-01 8.05320978e-01 6.36139736e-02 -8.31895351e-01
-6.18040264e-01 -9.62297618e-01 -1.24290919e+00 -2.35500634e-01
8.60244334e-01 -6.60050690e-01 -5.18217981e-01 2.42966622e-01
-9.19641674e-01 -2.97514945e-01 -6.59375548e-01 2.26616725e-01
-5.35881162e-01 4.12282199e-01 -2.13627845e-01 -8.26795042e-01
-2.39057302e-01 -1.07434785e+00 1.75510657e+00 7.03709722e-02
7.68389646e-03 -8.02621067e-01 -2.46298626e-01 1.89891875e-01
2.17072338e-01 7.50793889e-02 4.81343299e-01 -4.78284329e-01
-7.87907541e-01 -4.95734930e-01 -4.86175418e-01 1.88226566e-01
6.43034801e-02 -1.59019947e-01 -1.08773541e+00 -1.13418013e-01
-1.87053725e-01 -5.32451160e-02 1.08500361e+00 6.51876688e-01
1.13901114e+00 -5.43540195e-02 -6.94227755e-01 1.63118705e-01
1.88113010e+00 -2.03418538e-01 3.48245442e-01 2.67297357e-01
5.94539583e-01 8.42591405e-01 9.73944902e-01 5.15427470e-01
-3.97279114e-03 9.95018244e-01 5.23830712e-01 8.69224072e-02
-4.65660393e-01 5.94390593e-02 -1.72515631e-01 -2.95432545e-02
-2.29235634e-01 4.71821353e-02 -5.94154596e-01 6.02445364e-01
-1.74429512e+00 -9.83761370e-01 -3.25846612e-01 2.59511757e+00
5.63510299e-01 2.81281829e-01 2.65865088e-01 3.63057375e-01
8.92420113e-01 -7.66589642e-02 -2.22697303e-01 -1.70672219e-02
3.38898189e-02 1.43586516e-01 6.93593383e-01 5.95777333e-01
-1.32289052e+00 6.86219394e-01 5.40857887e+00 1.20099092e+00
-9.27470088e-01 1.29263416e-01 6.16120219e-01 2.72281677e-01
-7.54033849e-02 1.26612306e-01 -8.46518934e-01 2.64762908e-01
3.06008488e-01 4.10915405e-01 1.17401816e-01 1.06835508e+00
-8.70944634e-02 -6.27900660e-01 -8.51342022e-01 8.41587663e-01
2.25869333e-03 -1.10648692e+00 -1.46520078e-01 2.64149100e-01
4.05882746e-01 -1.73193246e-01 -6.02046728e-01 -3.16949300e-02
-9.24000889e-02 -5.76895595e-01 1.21302581e+00 6.19822323e-01
6.31089211e-01 -7.19168365e-01 6.04806483e-01 4.35652941e-01
-1.58915412e+00 -1.06810004e-01 -4.19373095e-01 -2.30455399e-02
-8.20490569e-02 8.70440841e-01 -6.88965499e-01 5.89998901e-01
6.43044591e-01 4.57465142e-01 -9.31222498e-01 1.35359943e+00
-3.01886082e-01 4.27949399e-01 -3.88340563e-01 -9.43889394e-02
1.34676009e-01 -5.74185289e-02 4.42984402e-01 1.22846365e+00
4.91188280e-02 -2.75626391e-01 2.96643704e-01 6.86264873e-01
3.94083679e-01 2.59192854e-01 -1.26210034e-01 4.34633076e-01
7.93528184e-02 1.77463257e+00 -1.69433427e+00 -5.95233381e-01
-1.61732331e-01 1.17835915e+00 4.42818016e-01 1.75524086e-01
-4.28721189e-01 -6.48512423e-01 3.97713035e-01 4.43169028e-01
7.27778256e-01 -1.10158026e-01 -3.75386775e-01 -1.03991103e+00
6.88991696e-02 -2.86746740e-01 3.69970173e-01 -6.55570209e-01
-1.20170295e+00 8.89286935e-01 1.53881073e-01 -1.44511080e+00
-2.81925797e-01 -8.30069721e-01 -2.49564275e-01 9.31058824e-01
-1.46767259e+00 -1.35999715e+00 -6.36225700e-01 4.00260895e-01
6.62889063e-01 3.49447101e-01 7.14957535e-01 3.00602484e-02
-1.49467424e-01 1.50851279e-01 -2.10202798e-01 -9.53180715e-02
3.57727796e-01 -1.61722004e+00 -1.91445246e-01 7.78284907e-01
2.36213684e-01 4.02494997e-01 7.30728209e-01 -6.25779808e-01
-7.08883822e-01 -9.56858158e-01 8.83750439e-01 -4.76983011e-01
2.76360959e-01 -4.77016926e-01 -7.66029298e-01 2.27298126e-01
-5.74279428e-02 4.40505058e-01 1.04722068e-01 8.41501132e-02
-3.02107275e-01 -1.73908532e-01 -1.38276410e+00 1.64815456e-01
9.43965077e-01 -3.49291295e-01 -5.26174724e-01 2.29321480e-01
1.31744638e-01 -1.90320592e-02 -6.41480803e-01 1.57434449e-01
6.50849342e-01 -1.39221060e+00 1.05024421e+00 -2.70423889e-02
2.52586380e-02 -5.45068741e-01 -9.16449726e-02 -1.03857064e+00
-7.18125761e-01 1.19927190e-01 3.00912231e-01 1.19718432e+00
1.13663122e-01 -5.40911734e-01 9.84830379e-01 2.35768124e-01
-1.68557331e-01 -7.07181036e-01 -9.31363285e-01 -5.70255697e-01
-3.66476834e-01 -4.15743530e-01 3.85975778e-01 4.09231126e-01
-2.95407832e-01 1.21086225e-01 3.55632424e-01 1.63933977e-01
8.22800159e-01 8.30847979e-01 5.86485445e-01 -1.65166795e+00
-2.25328892e-01 -7.81339705e-01 -1.01275003e+00 -5.64524710e-01
4.50611524e-02 -6.37673736e-01 3.89903069e-01 -1.81600869e+00
3.65301132e-01 -5.37775755e-01 -2.80132681e-01 6.11950755e-01
-1.54478043e-01 8.48468065e-01 -1.63673148e-01 3.68675202e-01
-6.83756649e-01 2.84457803e-01 9.18730497e-01 -1.11465856e-01
-6.96185902e-02 2.16648951e-01 -5.07342935e-01 7.15290129e-01
5.84012210e-01 -3.45775992e-01 -1.60748884e-01 2.82250285e-01
-2.50914216e-01 -4.00197059e-01 7.76642442e-01 -1.23775554e+00
-2.30510924e-02 -1.23606347e-01 6.70374393e-01 -8.20299506e-01
4.58495408e-01 -1.05720186e+00 6.14732131e-02 4.26351726e-01
-1.84662566e-01 -6.27088189e-01 5.94917126e-02 7.18782008e-01
-3.41279477e-01 -4.98745352e-01 9.13997650e-01 -3.52349877e-01
-1.01395857e+00 1.99366674e-01 -3.35815310e-01 -3.81415486e-01
1.07969940e+00 -6.41801298e-01 3.28929350e-02 -1.20701365e-01
-8.24665070e-01 -4.04907316e-01 6.95726514e-01 -3.63760628e-02
3.81894320e-01 -1.09153962e+00 -4.93689865e-01 8.57035890e-02
4.71754074e-01 -1.63761228e-01 4.81250197e-01 1.02045810e+00
-2.56621659e-01 5.06382465e-01 -3.18270028e-01 -7.60671377e-01
-1.35828233e+00 9.39784527e-01 4.01473254e-01 -2.59620041e-01
-6.02074027e-01 8.53794932e-01 6.22689962e-01 -2.20625356e-01
9.84993484e-03 -5.21775365e-01 -4.75026011e-01 4.61688727e-01
3.67358059e-01 2.37189621e-01 5.07050097e-01 -1.03347206e+00
-7.43552208e-01 8.83678019e-01 2.75443316e-01 -1.21758454e-01
1.04239082e+00 -3.17100465e-01 -7.01076537e-02 2.74026871e-01
1.08727193e+00 4.14929390e-01 -1.34338284e+00 -8.05907995e-02
1.13444358e-01 -6.61458611e-01 1.78281918e-01 -9.76225436e-01
-8.52529645e-01 8.65760446e-01 6.49294853e-01 5.40012777e-01
1.25735056e+00 4.15851921e-01 -9.28685963e-02 -4.90039214e-02
7.32077599e-01 -9.48649347e-01 -1.25024065e-01 -3.42546180e-02
1.03527033e+00 -1.21712112e+00 3.40309739e-01 -7.71028578e-01
-4.00737107e-01 1.12721527e+00 4.87574697e-01 -2.71599531e-01
7.80817449e-01 5.15957829e-03 -1.58745408e-01 -2.15562806e-01
-5.40468633e-01 -8.74154627e-01 7.54007161e-01 6.85074329e-01
3.92932773e-01 -1.37368878e-02 -4.68318284e-01 4.66718316e-01
1.51348352e-01 -2.10134983e-01 3.73458639e-02 8.49276364e-01
-8.48174870e-01 -1.10000908e+00 -6.74165308e-01 6.07351005e-01
-1.92018077e-01 2.08680466e-01 -5.90076745e-01 8.22011054e-01
3.79267484e-01 8.24767590e-01 1.92682967e-01 -3.80859971e-02
2.72040278e-01 -9.85584874e-03 6.35078788e-01 -6.29453480e-01
-6.98840320e-01 2.48378009e-01 -4.63016406e-02 -7.46225417e-01
-4.97973442e-01 -8.50531459e-01 -1.13440192e+00 4.39075351e-01
-6.58580482e-01 2.54843950e-01 9.51741576e-01 8.35452020e-01
3.59258592e-01 5.39609373e-01 5.06773114e-01 -1.53545177e+00
-2.86213696e-01 -9.97828364e-01 -8.20398152e-01 3.75163883e-01
2.91130751e-01 -9.21947837e-01 -5.84599257e-01 -6.41888231e-02] | [9.535089492797852, 1.1396801471710205] |
cbb30a20-70a4-4f95-99b2-7a429fc98f79 | depth-attentional-features-for-single-image | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Hu_Depth-Attentional_Features_for_Single-Image_Rain_Removal_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Hu_Depth-Attentional_Features_for_Single-Image_Rain_Removal_CVPR_2019_paper.pdf | Depth-Attentional Features for Single-Image Rain Removal | Rain is a common weather phenomenon, where object visibility varies with depth from the camera and objects faraway are visually blocked more by fog than by rain streaks. Existing methods and datasets for rain removal, however, ignore these physical properties, thereby limiting the rain removal efficiency on real photos. In this work, we first analyze the visual effects of rain subject to scene depth and formulate a rain imaging model collectively with rain streaks and fog; by then, we prepare a new dataset called RainCityscapes with rain streaks and fog on real outdoor photos. Furthermore, we design an end-to-end deep neural network, where we train it to learn depth-attentional features via a depth-guided attention mechanism, and regress a residual map to produce the rain-free image output. We performed various experiments to visually and quantitatively compare our method with several state-of-the-art methods to demonstrate its superiority over the others.
| [' Pheng-Ann Heng', ' Lei Zhu', ' Chi-Wing Fu', 'Xiaowei Hu'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['single-image-deraining'] | ['computer-vision'] | [-4.26822379e-02 -5.32082021e-01 5.36813676e-01 -6.52328074e-01
-3.33203338e-02 -4.60848421e-01 -3.73636633e-02 -4.68390375e-01
-2.03049079e-01 9.81093049e-01 1.80111397e-02 -3.68246526e-01
4.57332194e-01 -8.86269093e-01 -6.89465284e-01 -9.97948587e-01
-8.05662200e-02 -1.92695856e-01 3.63883734e-01 -1.56588271e-01
1.55939981e-01 5.04675448e-01 -1.57627594e+00 5.31423977e-03
1.33638692e+00 4.90156949e-01 7.05797493e-01 8.59704554e-01
1.65250916e-02 1.26988983e+00 -6.47409856e-01 7.74594396e-02
3.11973155e-01 -6.20648921e-01 -2.03221843e-01 1.90440118e-01
9.65605855e-01 -8.34294081e-01 -5.84984243e-01 1.00523067e+00
4.11675632e-01 1.87687948e-01 4.66357738e-01 -5.95770180e-01
-1.15209556e+00 -1.15997538e-01 -7.26479948e-01 8.01721513e-01
1.46101221e-01 4.31671560e-01 5.47015429e-01 -1.26224065e+00
2.86773562e-01 1.21461654e+00 4.27941978e-01 4.31798339e-01
-6.34742975e-01 -6.89787626e-01 6.19375050e-01 1.42329872e-01
-9.86913860e-01 -2.91103482e-01 4.70045418e-01 -2.25502342e-01
6.49529994e-01 3.42728555e-01 9.98122931e-01 5.87203562e-01
3.47812444e-01 6.93617165e-01 1.46908045e+00 -2.04921708e-01
2.08663657e-01 -1.57108635e-01 4.90398198e-01 9.02286053e-01
7.81802297e-01 3.08703691e-01 -7.39801601e-02 4.64319885e-01
7.65274048e-01 4.42896128e-01 -9.42820609e-01 4.06750366e-02
-4.31965858e-01 7.03187764e-01 1.09880781e+00 -2.29917735e-01
-2.88661420e-01 4.99007851e-02 -5.43310404e-01 3.41741264e-01
7.34759808e-01 6.28615171e-02 -1.72352895e-01 6.25087500e-01
-1.02337885e+00 5.57964623e-01 6.05535328e-01 6.53141856e-01
1.04871023e+00 4.58410710e-01 -1.19008154e-01 5.97330987e-01
5.80371082e-01 1.35121119e+00 -1.54651687e-01 -7.44377136e-01
3.11051458e-01 5.93149513e-02 6.86842024e-01 -7.97317266e-01
-4.45827633e-01 -3.58324826e-01 -8.42804372e-01 9.21575665e-01
2.41759315e-01 -3.16162676e-01 -1.71477723e+00 1.16664326e+00
2.68092692e-01 4.67999727e-01 1.68524534e-01 1.73887336e+00
1.17093384e+00 9.11411226e-01 -1.52806779e-02 -4.68112946e-01
1.03759611e+00 -1.12998521e+00 -1.04364252e+00 -7.88711250e-01
-1.28754147e-03 -6.12699091e-01 1.29165149e+00 4.16074723e-01
-1.00466335e+00 -5.02972722e-01 -1.19706690e+00 -1.13278314e-01
-2.07427859e-01 -2.25370169e-01 6.37054741e-01 5.12071013e-01
-9.51404512e-01 3.84420305e-01 -8.76337171e-01 -1.52427629e-01
5.08059323e-01 -2.48377621e-01 1.77741572e-01 -7.52705991e-01
-1.16712892e+00 1.04215074e+00 -1.06000572e-01 7.91015029e-01
-1.32540166e+00 -5.98954141e-01 -7.24669576e-01 -5.98106422e-02
-1.85230017e-01 -8.41501296e-01 9.08225060e-01 -1.09662771e+00
-1.34957981e+00 6.74768269e-01 -4.71997201e-01 -1.73477054e-01
2.85879552e-01 -8.51212144e-01 -5.29481471e-01 1.19298734e-01
-1.95085570e-01 1.72484890e-01 1.15376699e+00 -1.96395230e+00
-7.93460131e-01 -2.74981678e-01 5.52113652e-01 4.60700691e-01
4.22865450e-01 -1.41157538e-01 -3.94724041e-01 -3.44309181e-01
4.80447374e-02 -7.21374810e-01 -3.11644584e-01 6.85057719e-04
-2.53757507e-01 6.23208880e-01 9.26598430e-01 -8.19884300e-01
1.08024144e+00 -1.90986991e+00 6.48424774e-02 -1.70366853e-01
5.74434996e-01 5.72590113e-01 -1.27652034e-01 -6.86873645e-02
1.05233744e-01 -5.39418831e-02 -6.11683488e-01 -3.96605968e-01
-5.71451306e-01 4.43209022e-01 -6.38114870e-01 7.51230299e-01
3.09994042e-01 7.06571400e-01 -9.09315526e-01 -2.63722777e-01
3.12882215e-01 5.84422469e-01 -1.75424561e-01 7.17861474e-01
-2.25603387e-01 5.00884235e-01 -3.23654860e-01 9.27135408e-01
1.56704485e+00 9.95208323e-03 -1.98320717e-01 1.34253234e-01
-3.93671572e-01 -7.74036348e-02 -6.68729305e-01 8.50335896e-01
-6.60754263e-01 7.96870112e-01 1.43669158e-01 -2.13856071e-01
8.24172139e-01 -1.84743032e-01 -5.01359701e-01 -8.54054093e-01
3.15851904e-02 9.35388729e-02 -1.83366790e-01 -9.48992789e-01
4.37606722e-01 -3.56882691e-01 7.06197619e-01 3.38141508e-02
-4.50099260e-01 -2.18787208e-01 -3.05228263e-01 1.31244138e-01
8.97971272e-01 -6.54771104e-02 -1.84089541e-01 -2.39625033e-02
2.36222446e-02 -1.89751133e-01 5.66633284e-01 9.39830124e-01
-1.61150843e-01 1.13441980e+00 -1.03036113e-01 -7.60956049e-01
-6.61429346e-01 -1.43785095e+00 -7.57480487e-02 7.71732628e-01
7.92879939e-01 3.43986601e-01 -5.32181680e-01 -5.09542346e-01
9.81263518e-02 7.20500529e-01 -8.18632424e-01 2.11750511e-02
-6.72625661e-01 -1.16742265e+00 2.95722298e-02 3.05795014e-01
8.27956200e-01 -1.39114344e+00 -7.29914486e-01 -6.01119883e-02
-1.35804027e-01 -1.06833518e+00 -8.03809464e-02 7.67483190e-02
-8.66349220e-01 -1.16413021e+00 -7.36692250e-01 -6.48996294e-01
5.96951842e-01 1.16713095e+00 1.49908543e+00 4.31318641e-01
-5.15539527e-01 -1.38694122e-01 -6.59853399e-01 -6.99609935e-01
3.78870547e-01 -4.82452363e-01 -2.57525295e-01 -1.96583927e-01
1.56090215e-01 -7.55711496e-01 -1.21351969e+00 -1.49271652e-01
-1.11491573e+00 -2.14964807e-01 6.29291475e-01 4.97719049e-01
4.35050875e-01 -1.61861509e-01 2.01647937e-01 -1.11827266e+00
2.75154889e-01 -5.42418957e-01 -8.70451987e-01 -3.09962052e-04
-2.44213909e-01 -4.09373134e-01 3.16339135e-01 9.07111242e-02
-1.35939670e+00 3.48533653e-02 1.58722162e-01 -6.92853570e-01
-1.56072353e-03 3.85085642e-01 -3.93091291e-01 -3.56952757e-01
5.15867651e-01 2.53881752e-01 -6.07267618e-01 -3.90474677e-01
5.06630123e-01 4.21382666e-01 4.55531150e-01 2.04031289e-01
1.14376736e+00 9.90350306e-01 -3.02158713e-01 -9.96976495e-01
-1.59135664e+00 -2.63684481e-01 -3.65581602e-01 -3.41056168e-01
1.02036071e+00 -1.28433895e+00 -3.57428193e-01 7.86747456e-01
-1.15702987e+00 -6.41960859e-01 1.82147861e-01 4.50818002e-01
1.03646047e-01 2.93581814e-01 -6.04520679e-01 -1.27545786e+00
-4.64597672e-01 -6.04425550e-01 7.52039015e-01 7.49922752e-01
6.27093554e-01 -7.94601262e-01 2.57774323e-01 1.36613011e-01
4.63476986e-01 2.99825341e-01 4.76889431e-01 6.29483879e-01
-1.19701326e+00 2.79276878e-01 -6.53228164e-01 4.29855019e-01
3.55140805e-01 2.27514550e-01 -1.35663247e+00 -4.06446338e-01
2.35274822e-01 -9.87394229e-02 1.60597181e+00 7.24740386e-01
8.59108806e-01 -2.05243111e-01 -1.09690659e-01 1.24353933e+00
1.95438981e+00 5.07986248e-02 1.08071446e+00 4.13054615e-01
1.13457251e+00 2.86700904e-01 8.34470332e-01 2.00723693e-01
2.82115251e-01 -1.14095882e-02 1.08715224e+00 -7.10007846e-01
-3.64968240e-01 2.69473791e-01 3.24925214e-01 6.20327055e-01
-5.89223564e-01 -6.99826479e-01 -5.52716136e-01 6.74314022e-01
-1.50070643e+00 -9.08123076e-01 -4.84655976e-01 2.09021735e+00
5.25222957e-01 1.19655520e-01 -4.85335767e-01 -4.37969953e-01
3.12341601e-01 5.82370758e-01 -6.48345828e-01 -1.12387240e-01
-4.73164499e-01 4.47107732e-01 7.70412028e-01 9.71583188e-01
-1.23907077e+00 1.26091397e+00 6.43418455e+00 -7.63494596e-02
-1.35776949e+00 4.61915657e-02 3.11505377e-01 -3.79903883e-01
-4.28457618e-01 1.95112766e-03 -7.75777042e-01 4.87298608e-01
4.45342243e-01 4.53916252e-01 4.78440553e-01 4.72907215e-01
6.66965663e-01 -3.75632763e-01 -4.62736934e-01 7.61505783e-01
1.64560020e-01 -7.84002364e-01 -1.51640433e-03 -3.13318610e-01
9.39680696e-01 4.82795805e-01 7.03736693e-02 1.83383137e-01
7.87042499e-01 -1.25384331e+00 5.43208480e-01 9.55089152e-01
4.84319091e-01 -4.33147103e-01 7.50756264e-01 -5.78891784e-02
-8.71673107e-01 9.64340195e-03 -8.81860912e-01 -3.77779037e-01
2.76624799e-01 1.21937382e+00 -1.65960416e-01 5.76426208e-01
1.10586667e+00 6.90146804e-01 -4.78479832e-01 1.35656166e+00
-8.43128026e-01 8.18029344e-01 -1.25114009e-01 3.49243730e-01
2.22658724e-01 -3.37737083e-01 5.04250407e-01 1.41695166e+00
2.34265551e-01 5.62576711e-01 3.32056172e-02 7.87520707e-01
-3.86609733e-02 -4.84660059e-01 -5.97943664e-01 5.60443282e-01
3.24565470e-01 1.19359004e+00 -4.16134864e-01 -2.76086390e-01
-4.91521627e-01 1.25321591e+00 2.49632001e-01 9.75490153e-01
-1.10530519e+00 -5.74233532e-01 1.13919139e+00 2.29556561e-02
6.38195813e-01 -2.95691878e-01 -1.97298229e-01 -1.45817029e+00
7.19647780e-02 -5.61562181e-01 -2.04515740e-01 -1.28115118e+00
-1.33512235e+00 7.53495574e-01 -4.67919409e-01 -1.16915810e+00
5.43420911e-01 -5.81341207e-01 -9.77303684e-01 1.32154024e+00
-2.54025650e+00 -9.10492241e-01 -1.25666320e+00 4.91341233e-01
5.02663374e-01 5.40436387e-01 4.43627775e-01 3.50342274e-01
-5.34027278e-01 1.32438749e-01 4.61623490e-01 -1.97614625e-01
8.12595963e-01 -1.30309153e+00 3.49491000e-01 1.29831147e+00
-2.08541915e-01 2.28525370e-01 1.09201515e+00 -6.33290708e-01
-1.29250383e+00 -1.64972389e+00 4.69113261e-01 -3.97946298e-01
1.81896597e-01 -2.83780128e-01 -1.42096841e+00 6.64053380e-01
2.55400538e-01 5.75447798e-01 -6.47909790e-02 1.13114323e-02
-4.98060197e-01 -2.32249245e-01 -9.54884171e-01 5.08653998e-01
1.01989555e+00 -2.01118588e-01 -6.61204994e-01 4.32186604e-01
9.89769697e-01 -6.27544641e-01 -2.60279588e-02 4.06206667e-01
4.69711065e-01 -1.37253451e+00 8.93905461e-01 -4.87786323e-01
7.32068717e-01 -6.30737126e-01 -2.00304762e-01 -1.69184971e+00
-6.41248107e-01 -9.05974135e-02 -2.84090012e-01 9.41530585e-01
1.22446954e-01 -5.48954844e-01 6.28254652e-01 -1.13646165e-01
-4.15477633e-01 -4.02579844e-01 -4.25051093e-01 -4.44613427e-01
9.81986970e-02 1.11959599e-01 3.38427842e-01 9.43233669e-01
-8.28855872e-01 2.07546249e-01 -6.82100177e-01 1.12858796e+00
7.46336818e-01 6.77196443e-01 6.44845426e-01 -1.07019925e+00
-6.06788024e-02 5.02408370e-02 4.05547172e-02 -1.13720238e+00
-2.04605177e-01 -2.05669731e-01 7.09402502e-01 -1.85849822e+00
1.98584616e-01 -2.90539593e-01 -4.19808805e-01 2.27184296e-01
-9.19762969e-01 3.91192824e-01 2.07873687e-01 5.30396700e-01
-5.78525126e-01 7.73868620e-01 1.62386596e+00 -8.78934041e-02
-3.40568900e-01 -1.18798673e-01 -4.52042043e-01 9.52225924e-01
8.66453111e-01 -5.81600666e-01 -2.05398172e-01 -1.14690626e+00
5.80766238e-02 -6.40042722e-02 4.77278680e-01 -1.06159687e+00
-2.90933758e-01 -4.60169375e-01 7.31173754e-01 -5.21026433e-01
4.56454992e-01 -5.91799736e-01 -3.88119221e-01 3.08546990e-01
3.10045481e-01 -3.18854243e-01 9.89239067e-02 7.45657563e-01
-2.41701528e-01 -9.84261557e-03 1.09647787e+00 -2.85059422e-01
-6.61203504e-01 5.18831551e-01 -5.88387489e-01 -1.05383284e-01
5.64485610e-01 -4.58690859e-02 -8.31876218e-01 -3.77318770e-01
-6.77240908e-01 3.13262582e-01 4.74192351e-01 1.63808569e-01
9.49677646e-01 -6.76030397e-01 -9.84756947e-01 2.04728514e-01
-4.66762744e-02 2.54235119e-01 5.97324073e-01 5.52580714e-01
-1.11276042e+00 -6.86821807e-03 -1.90101624e-01 -2.77311891e-01
-1.32915139e+00 5.78632474e-01 6.12646043e-01 1.05689049e-01
-9.25781012e-01 1.04315233e+00 9.33703184e-01 -1.07033111e-01
6.83474615e-02 -7.13973403e-01 -2.47415498e-01 -4.16361153e-01
8.46003592e-01 7.57248253e-02 -1.89463440e-02 -3.10622379e-02
-1.55477762e-01 5.64496219e-01 -1.07045211e-01 3.32008392e-01
1.33054316e+00 -5.43033659e-01 -1.19093366e-01 6.23367012e-01
6.14700496e-01 1.85608879e-01 -1.55738521e+00 -1.56010568e-01
-7.31068373e-01 -9.36408520e-01 2.63738245e-01 -9.03833508e-01
-1.74209297e+00 1.06658483e+00 1.02716875e+00 1.42229751e-01
1.38071859e+00 -2.34082490e-01 8.03466976e-01 4.61985737e-01
1.93230473e-02 -3.16171974e-01 -1.14199296e-01 8.72074127e-01
7.84133732e-01 -1.53319788e+00 2.31438294e-01 -6.09738886e-01
-5.94083667e-01 7.86307633e-01 8.10136020e-01 -6.61676288e-01
8.14858437e-01 2.60719419e-01 7.41735637e-01 -3.75381589e-01
-4.63863850e-01 -5.64323723e-01 -2.56026179e-01 7.60145485e-01
2.45361820e-01 2.80600227e-02 -7.93794394e-02 3.41072947e-01
7.83315003e-02 2.12399542e-01 7.59595335e-01 9.46059346e-01
-1.14946580e+00 -2.77940363e-01 -6.51074052e-01 2.80900866e-01
-3.03119093e-01 -5.72390616e-01 -1.52060702e-01 7.28838146e-01
4.17849332e-01 1.06124282e+00 1.61848038e-01 -1.24924779e-01
3.39231133e-01 -6.05661452e-01 5.59313118e-01 -5.68507135e-01
-2.45994002e-01 -1.71413124e-01 -1.74493268e-01 -3.82505625e-01
-6.22974873e-01 -3.50162357e-01 -1.08627355e+00 -3.00011784e-01
-2.72055477e-01 -5.13228439e-02 3.28184217e-01 8.88667643e-01
1.40277771e-02 7.99482763e-01 7.89046943e-01 -1.17626286e+00
2.75159091e-01 -1.05897307e+00 -1.13929689e+00 1.28407747e-01
1.05904734e+00 -6.54383004e-01 -7.99224377e-01 2.30792314e-01] | [10.923750877380371, -3.236196756362915] |
46143e7e-493e-46f6-bc83-0dd8b7c6c20b | facial-soft-biometrics-for-recognition-in-the | 2210.13129 | null | https://arxiv.org/abs/2210.13129v1 | https://arxiv.org/pdf/2210.13129v1.pdf | Facial Soft Biometrics for Recognition in the Wild: Recent Works, Annotation, and COTS Evaluation | The role of soft biometrics to enhance person recognition systems in unconstrained scenarios has not been extensively studied. Here, we explore the utility of the following modalities: gender, ethnicity, age, glasses, beard, and moustache. We consider two assumptions: 1) manual estimation of soft biometrics and 2) automatic estimation from two commercial off-the-shelf systems (COTS). All experiments are reported using the labeled faces in the wild (LFW) database. First, we study the discrimination capabilities of soft biometrics standalone. Then, experiments are carried out fusing soft biometrics with two state-of-the-art face recognition systems based on deep learning. We observe that soft biometrics is a valuable complement to the face modality in unconstrained scenarios, with relative improvements up to 40%/15% in the verification performance when using manual/automatic soft biometrics estimation. Results are reproducible as we make public our manual annotations and COTS outputs of soft biometrics over LFW, as well as the face recognition scores. | ['Fernando Alonso-Fernandez', 'Ruben Vera-Rodriguez', 'Julian Fierrez', 'Ester Gonzalez-Sosa'] | 2022-10-24 | null | null | null | null | ['person-recognition'] | ['computer-vision'] | [ 3.83040607e-02 8.59363824e-02 6.66799620e-02 -8.30571949e-01
-4.14915472e-01 -4.22965646e-01 6.95514560e-01 -5.99628806e-01
-4.15950269e-01 6.98910415e-01 -6.98931813e-02 -9.48395357e-02
1.39170647e-01 -3.48981440e-01 -4.47906286e-01 -9.25350845e-01
1.81000471e-01 3.17796528e-01 -2.80160367e-01 -2.29392171e-01
-6.24652095e-02 6.75714374e-01 -1.77150846e+00 1.40012056e-01
4.01863605e-01 1.48116028e+00 -8.39883506e-01 5.76719999e-01
2.42066562e-01 -1.05412547e-02 -7.00807095e-01 -1.18651485e+00
7.56739616e-01 1.81605190e-01 -3.11722100e-01 -1.34024039e-01
1.14609861e+00 -6.31510615e-01 -1.52660623e-01 9.65208709e-01
9.40168858e-01 -5.46184778e-01 7.10803151e-01 -1.19905806e+00
-8.18671584e-01 3.09033960e-01 -5.91537714e-01 -3.05558681e-01
7.31512547e-01 5.68276525e-01 2.86109120e-01 -9.98693705e-01
1.48320958e-01 1.22338951e+00 1.06295264e+00 9.82850969e-01
-9.50162649e-01 -9.57121313e-01 -3.13649416e-01 1.89617157e-01
-1.77663779e+00 -1.27958846e+00 2.79185861e-01 -4.84796017e-01
7.10139751e-01 3.28626603e-01 2.04394355e-01 1.22877061e+00
3.56493667e-02 2.59623617e-01 1.59424531e+00 -6.78944111e-01
1.02420889e-01 6.04475260e-01 2.21009657e-01 8.18664312e-01
5.51432848e-01 4.94883776e-01 -8.69970500e-01 -1.84176713e-01
3.25505495e-01 -9.42508429e-02 -1.66382268e-01 2.11329773e-01
-5.76708138e-01 1.34320140e-01 -2.43120372e-01 2.04721913e-01
-2.98414081e-01 -8.36422965e-02 4.18231264e-02 5.95880389e-01
4.59295303e-01 -6.58499002e-02 -5.72722614e-01 1.05337873e-02
-1.15493143e+00 -1.49141759e-01 1.07827210e+00 8.96413624e-01
5.44541836e-01 2.97965795e-01 -9.96643677e-02 8.19598198e-01
8.03657413e-01 1.18711901e+00 1.62551984e-01 -5.30515313e-01
5.22430837e-02 5.28665066e-01 1.86519817e-01 -6.07617021e-01
-5.34919873e-02 4.53470461e-02 -6.79306626e-01 5.76192975e-01
8.30006957e-01 -3.51198822e-01 -1.18830192e+00 1.47542870e+00
1.95454508e-01 1.97392777e-01 -1.72362477e-02 4.81831074e-01
9.13036287e-01 -8.25065225e-02 6.94439039e-02 -3.94251645e-02
1.59673321e+00 -2.65859723e-01 -7.35070288e-01 2.15631023e-01
-7.95362592e-02 -8.78082514e-01 7.87028611e-01 4.39832985e-01
-1.07302737e+00 -8.32334995e-01 -1.00038612e+00 3.20083827e-01
-8.38964224e-01 6.55041397e-01 2.10197210e-01 2.11937904e+00
-1.44228590e+00 5.16995788e-01 -8.09445083e-01 -5.29499352e-01
4.62113231e-01 1.11255670e+00 -6.23771906e-01 1.62104696e-01
-8.92664552e-01 1.05164075e+00 -2.16628045e-01 2.57747799e-01
-6.43971264e-01 -6.05316281e-01 -8.65820110e-01 -1.44481272e-01
7.11088032e-02 -2.17210829e-01 8.24388742e-01 -1.02510059e+00
-1.81923342e+00 1.44407213e+00 -3.99317086e-01 -2.01717630e-01
5.46243250e-01 -3.64397168e-01 -6.77543879e-01 -2.22027972e-01
-5.88525355e-01 1.13565892e-01 1.08472526e+00 -1.10816765e+00
6.49558054e-03 -8.36214244e-01 -2.63430744e-01 -5.51550806e-01
-7.36828446e-01 5.30409336e-01 -1.38803214e-01 -2.55907089e-01
-3.80833894e-01 -1.00360250e+00 5.79425693e-01 2.16808259e-01
-4.53161418e-01 -2.50963010e-02 7.80380845e-01 -1.12000430e+00
8.69834065e-01 -2.02503848e+00 -3.02786708e-01 5.17609894e-01
1.35960445e-01 6.91691875e-01 -7.63272047e-02 -1.19419888e-01
-3.51527408e-02 8.03774372e-02 1.90360676e-02 -7.97819316e-01
2.83282250e-01 -3.38420831e-02 9.56666544e-02 6.65854394e-01
1.18355088e-01 8.12499702e-01 -3.11001185e-02 -4.36492026e-01
2.14058265e-01 8.20547163e-01 4.71900078e-03 3.06711078e-01
5.64491391e-01 1.44686431e-01 1.91746548e-01 1.60998321e+00
1.16856265e+00 3.86615932e-01 1.50119156e-01 -5.72074354e-01
2.41434611e-02 -2.91312307e-01 -1.36851704e+00 7.86794245e-01
-2.22604319e-01 3.18125278e-01 2.72264242e-01 -3.83090436e-01
1.00897026e+00 6.03790879e-01 2.03933135e-01 -3.10543388e-01
4.85354513e-01 3.99976760e-01 1.66454315e-01 -4.92711037e-01
1.15060024e-01 -1.14439569e-01 5.27831554e-01 2.66644001e-01
6.44299865e-01 5.16089261e-01 -1.08892873e-01 -6.24372780e-01
5.80409229e-01 3.41677815e-01 3.21712911e-01 -4.63221341e-01
8.54138494e-01 -1.03746891e+00 2.80980319e-01 5.08925974e-01
-4.92532223e-01 4.64224339e-01 9.36321169e-02 -2.40178421e-01
-7.69355893e-01 -9.99056637e-01 -4.36061919e-01 1.15795946e+00
-2.24127010e-01 -7.05032870e-02 -1.25640035e+00 -6.69627666e-01
4.68171000e-01 7.68735399e-03 -9.10469472e-01 -4.32370342e-02
-3.08929741e-01 -1.13912368e+00 1.06029177e+00 7.53044844e-01
6.69021308e-01 -5.60708344e-01 -2.94187516e-01 -5.01761377e-01
5.40225029e-01 -1.13826108e+00 -3.64732355e-01 -2.16819555e-01
-3.13885808e-01 -1.18117511e+00 -1.19927096e+00 -3.08963627e-01
6.05786979e-01 -3.78480524e-01 9.34269011e-01 2.43429258e-01
-2.31360450e-01 7.72846997e-01 2.06726231e-03 -6.29716694e-01
-4.81806487e-01 -9.59000289e-02 8.86477053e-01 7.42553711e-01
8.65117371e-01 -3.15543056e-01 -5.87736130e-01 6.68161154e-01
-1.36043295e-01 -6.19087219e-01 5.58467448e-01 6.32431746e-01
-5.68206273e-02 -5.44262171e-01 5.11340022e-01 -8.31295490e-01
2.10703015e-01 6.32545678e-03 -6.63718104e-01 6.94631457e-01
-9.03515220e-01 -1.80929437e-01 8.18173364e-02 -4.31495100e-01
-1.31769705e+00 2.83788174e-01 -4.41261142e-01 -1.93213582e-01
-6.34612024e-01 -2.38096286e-02 -5.31387150e-01 -1.01920986e+00
7.19854593e-01 1.47284314e-01 1.94580287e-01 -5.79101086e-01
-2.28144124e-01 1.39247668e+00 5.68120062e-01 -7.46251225e-01
9.17720854e-01 3.79852086e-01 -7.71746933e-02 -9.91569757e-01
-2.91057408e-01 -3.22521776e-01 -1.11161971e+00 -5.30377686e-01
6.08919561e-01 -6.86758459e-01 -1.63872898e+00 1.20889258e+00
-7.62247205e-01 -8.47350210e-02 1.69190746e-02 1.68545932e-01
-3.48159149e-02 4.62021500e-01 -5.07601202e-01 -1.64539731e+00
-5.82376063e-01 -1.03308856e+00 1.44904542e+00 5.67386746e-01
-7.17472211e-02 -8.58698547e-01 -2.48831004e-01 6.62726402e-01
7.63470352e-01 3.21763992e-01 1.31459996e-01 -6.69979692e-01
-1.28738701e-01 -7.61364758e-01 -2.42291093e-01 7.07787037e-01
3.77645612e-01 2.77050674e-01 -1.89483619e+00 -4.42898959e-01
-2.21494332e-01 -2.46463701e-01 7.05730796e-01 3.04907233e-01
8.50837588e-01 -3.13419640e-01 2.29522716e-02 6.95713699e-01
1.27535164e+00 1.83065996e-01 7.80704856e-01 -2.36895368e-01
5.15406549e-01 7.18957365e-01 -1.05983727e-02 3.28728288e-01
2.25703150e-01 6.51000977e-01 1.50657177e-01 -2.01543435e-01
-2.28439793e-01 3.76479954e-01 6.73402607e-01 1.55800238e-01
-1.09337449e+00 6.61652312e-02 -1.13615894e+00 6.77586067e-03
-1.27112913e+00 -8.74059021e-01 1.31175533e-01 2.51064253e+00
9.72005665e-01 -1.00835674e-01 3.54962051e-01 2.24971592e-01
7.50185311e-01 -2.71619856e-01 -3.22247475e-01 -1.53453752e-01
-2.49924287e-01 5.96365094e-01 6.51727736e-01 4.86219078e-01
-1.10512733e+00 6.67835593e-01 6.62177563e+00 3.18045706e-01
-1.40518093e+00 1.64793655e-01 8.26499701e-01 -2.08354555e-02
5.25789380e-01 -5.55091083e-01 -1.41373909e+00 6.22138560e-01
1.23762429e+00 4.13141668e-01 4.35768783e-01 7.40002632e-01
-2.44040757e-01 -7.25480616e-02 -1.30202603e+00 1.04905260e+00
5.13964236e-01 -8.64221811e-01 -2.52190977e-01 4.62230057e-01
6.13761961e-01 -3.24830025e-01 4.35341746e-01 1.34024933e-01
-1.06334649e-01 -1.22279871e+00 5.75237155e-01 9.34942126e-01
1.52544153e+00 -2.70719647e-01 1.37358153e+00 -2.83373863e-01
-1.07860231e+00 5.64940423e-02 -1.79791972e-01 2.81976312e-01
-3.28940928e-01 2.08034277e-01 -6.05901122e-01 4.56106007e-01
6.67116404e-01 2.47004092e-01 -1.00270832e+00 6.54761970e-01
2.42967173e-01 6.94484949e-01 -4.11689252e-01 1.87279791e-01
-5.66801488e-01 1.55223235e-01 6.00282848e-02 1.43895876e+00
3.37811947e-01 -1.02475658e-02 -1.53763399e-01 6.02453470e-01
-2.04815656e-01 -1.44666865e-01 -2.76754946e-01 2.90239267e-02
2.08949715e-01 1.22727013e+00 -2.37981945e-01 -4.05124307e-01
-4.43911344e-01 6.95639789e-01 -4.68495935e-01 3.04810494e-01
-3.42208743e-01 -1.58094153e-01 8.55062068e-01 3.94564927e-01
-2.00004250e-01 -2.68204254e-03 -5.55386961e-01 -1.21956825e+00
-1.35258203e-02 -1.03898323e+00 1.25601441e-01 -4.17086005e-01
-1.38101125e+00 6.71605170e-01 -1.56111300e-01 -8.65892828e-01
-2.97923863e-01 -1.33484519e+00 -5.39548814e-01 1.46554661e+00
-1.44562805e+00 -1.73594952e+00 -6.89222217e-01 5.28339982e-01
2.49668444e-03 -9.85712290e-01 1.06692183e+00 5.17491758e-01
-5.05343556e-01 1.46749485e+00 -5.20822592e-02 4.79317576e-01
1.05493808e+00 -1.21783078e+00 3.01339746e-01 7.93553710e-01
-2.27353081e-01 9.97601807e-01 4.88327950e-01 -5.38296640e-01
-1.56636715e+00 -4.43289429e-01 5.19979000e-01 -7.83180654e-01
8.62116367e-02 -3.37804228e-01 -7.21452594e-01 5.05418003e-01
2.35499218e-01 1.86869204e-01 1.35970032e+00 3.32753748e-01
-7.12527573e-01 -4.50116783e-01 -1.79596007e+00 2.27459326e-01
6.92436635e-01 -7.34434605e-01 -1.73074096e-01 -6.54908046e-02
-3.04116160e-01 -3.10385227e-01 -1.17351401e+00 6.41344905e-01
1.60809302e+00 -8.96810234e-01 1.13944411e+00 -5.53498089e-01
6.97861984e-02 -2.22379178e-01 -5.00306904e-01 -6.16075277e-01
8.69145989e-02 -5.01218677e-01 -4.66311157e-01 1.69757998e+00
2.78331041e-01 -8.93265724e-01 9.72060025e-01 1.19226575e+00
4.35459644e-01 -6.93901777e-01 -8.62341464e-01 -9.66246545e-01
-2.65888840e-01 -6.66250437e-02 8.50830793e-01 7.97982275e-01
-4.78029490e-01 -3.85950625e-01 -5.88775575e-01 3.72881591e-01
1.00480485e+00 -3.88884574e-01 8.89488101e-01 -1.53516924e+00
-3.08237761e-01 -2.64327168e-01 -7.69951344e-01 -1.65428251e-01
1.23575069e-01 -3.64322662e-01 -7.98090622e-02 -5.49637854e-01
1.51093453e-01 -3.42851691e-02 -5.46073020e-01 9.59964991e-01
3.13467309e-02 9.88856792e-01 1.26377940e-01 -1.26041789e-02
6.79656789e-02 6.47677779e-02 5.94107926e-01 -3.50510865e-01
9.89581794e-02 2.58079737e-01 -7.57807970e-01 8.83749485e-01
5.58389962e-01 2.02102631e-01 4.03381348e-01 -1.01726772e-02
-2.36653820e-01 -2.49812812e-01 5.01872778e-01 -1.07385862e+00
1.99164718e-01 -1.40195608e-01 9.97451782e-01 -7.64464363e-02
7.02864885e-01 -8.23978126e-01 -3.32592316e-02 3.80509228e-01
-2.13879463e-03 -3.31464946e-01 4.30377543e-01 -5.15023097e-02
1.23152949e-01 -1.44684613e-01 1.18227553e+00 2.41613925e-01
-2.44188339e-01 3.23442400e-01 -1.42824491e-02 -5.40679574e-01
6.79433942e-01 -7.49912262e-01 -3.64218384e-01 -2.16621190e-01
-8.56406152e-01 -3.61972600e-01 4.46935117e-01 1.72786653e-01
4.35185581e-01 -1.11580181e+00 -9.32852983e-01 7.72329926e-01
2.58359518e-02 -9.80435371e-01 7.93405995e-03 8.93285692e-01
-2.44669929e-01 2.24788025e-01 -6.32580280e-01 -5.32355011e-01
-2.10857224e+00 1.20718181e-01 6.62054360e-01 4.30465490e-01
2.72974640e-01 8.77502859e-01 -4.29651141e-01 -4.19198573e-01
6.49837434e-01 6.75077885e-02 -5.82598038e-02 5.80655858e-02
8.61858726e-01 4.47285265e-01 3.82774651e-01 -9.42854702e-01
-7.11003780e-01 8.79504383e-01 -7.28994012e-02 -4.24678251e-02
1.24266350e+00 1.07580259e-01 -1.08455233e-01 1.43039122e-01
7.99134076e-01 2.31145576e-01 -9.09536839e-01 -1.15092620e-01
3.52765135e-02 -8.04438174e-01 -2.33281329e-01 -1.26202786e+00
-1.22938049e+00 9.14071798e-01 1.48056936e+00 2.70973742e-01
1.10340869e+00 -2.14975595e-01 4.69709784e-01 4.94975299e-01
3.14139903e-01 -1.00628805e+00 -4.46288049e-01 1.83307067e-01
8.57443929e-01 -1.45870328e+00 9.81759131e-02 -4.70043987e-01
-3.05964559e-01 1.32148337e+00 6.23613596e-01 2.17599407e-01
1.10110712e+00 7.24472642e-01 2.75022507e-01 6.44961149e-02
-1.97251156e-01 -2.36590207e-01 7.65951931e-01 9.12808597e-01
8.72395635e-01 -5.77755040e-03 1.89459901e-02 7.60815442e-01
-2.08890811e-01 2.38465130e-01 1.70653939e-01 6.33327007e-01
-2.29664682e-03 -1.24846554e+00 -8.45696509e-01 5.86538255e-01
-7.89263606e-01 6.18942827e-02 -7.54207373e-01 3.87428641e-01
5.32133579e-01 7.48928070e-01 -2.93655008e-01 -7.00450420e-01
2.64119983e-01 6.77796423e-01 7.21020222e-01 -3.05339694e-01
-9.19923127e-01 -3.36364239e-01 -4.06905226e-02 -2.75521874e-01
-6.25860870e-01 -8.91978502e-01 -4.53172892e-01 -7.07813382e-01
-4.56955314e-01 -5.43279290e-01 9.69251812e-01 1.22156358e+00
3.07812423e-01 3.60484957e-03 4.83847022e-01 -1.08262467e+00
-7.04513133e-01 -1.19569135e+00 -6.16003811e-01 8.93042237e-02
4.11781311e-01 -7.70448089e-01 -1.23315431e-01 6.04366660e-01] | [13.338809967041016, 1.032435417175293] |
46d39929-07d3-4de3-ba07-aab6ade54779 | uncertainty-quantification-via-spatial | 2306.09882 | null | https://arxiv.org/abs/2306.09882v1 | https://arxiv.org/pdf/2306.09882v1.pdf | Uncertainty Quantification via Spatial-Temporal Tweedie Model for Zero-inflated and Long-tail Travel Demand Prediction | crucial for transportation management. However, traditional spatial-temporal deep learning models grapple with addressing the sparse and long-tail characteristics in high-resolution O-D matrices and quantifying prediction uncertainty. This dilemma arises from the numerous zeros and over-dispersed demand patterns within these matrices, which challenge the Gaussian assumption inherent to deterministic deep learning models. To address these challenges, we propose a novel approach: the Spatial-Temporal Tweedie Graph Neural Network (STTD). The STTD introduces the Tweedie distribution as a compelling alternative to the traditional 'zero-inflated' model and leverages spatial and temporal embeddings to parameterize travel demand distributions. Our evaluations using real-world datasets highlight STTD's superiority in providing accurate predictions and precise confidence intervals, particularly in high-resolution scenarios. | ['Xiaowei Gao', 'Jiayuan Luo', 'Hao Chen', 'Xianghui Zhang', 'Dingyi Zhuang', 'Xinke Jiang'] | 2023-06-16 | null | null | null | null | ['management'] | ['miscellaneous'] | [-4.64480847e-01 -2.69483387e-01 -5.59586048e-01 -3.06394815e-01
-5.74705422e-01 -4.97735411e-01 7.63595462e-01 3.86591434e-01
-5.00435494e-02 7.41209924e-01 7.36280560e-01 -9.01516914e-01
-8.31386089e-01 -1.01606476e+00 -6.98740005e-01 -5.01712978e-01
-7.31691062e-01 4.90857899e-01 -8.21348950e-02 -2.68832535e-01
-1.08549520e-01 5.30584514e-01 -1.10744238e+00 -1.08707868e-01
1.03638732e+00 1.15481174e+00 -2.64295936e-01 3.61736864e-01
-2.83233374e-01 5.74633121e-01 -3.01524222e-01 -5.50329864e-01
2.35121593e-01 4.11213934e-01 -6.82137609e-02 -3.42985541e-01
3.13904285e-01 -5.28766155e-01 -8.25999200e-01 7.45677471e-01
2.34131664e-01 4.40728039e-01 1.22606385e+00 -1.82835627e+00
-1.19082773e+00 6.51674807e-01 -7.77058125e-01 5.36425650e-01
2.43877992e-02 1.97665915e-01 9.49312687e-01 -8.11471879e-01
6.79307655e-02 1.38118196e+00 1.24995172e+00 -1.91994011e-01
-1.58946931e+00 -5.72380781e-01 3.81565601e-01 -2.92627495e-02
-1.74571371e+00 -8.53661820e-02 7.23272860e-01 -8.57084453e-01
9.86070156e-01 4.32680324e-02 3.97263438e-01 1.42630816e+00
5.79419315e-01 7.00136721e-01 5.29706001e-01 3.40701044e-01
1.73771188e-01 -1.25134349e-01 7.35299662e-02 2.69238025e-01
5.48270822e-01 6.55155480e-01 -2.18482763e-01 -1.05213463e-01
6.05703115e-01 4.05689210e-01 3.11228335e-01 -4.37051326e-01
-9.16090667e-01 9.75195348e-01 7.27569818e-01 -1.74133390e-01
-6.13444388e-01 5.56944966e-01 2.88787365e-01 9.07563120e-02
7.96825409e-01 -2.34581847e-02 -3.26545388e-01 -2.52530813e-01
-1.01300859e+00 4.45764959e-01 5.30569732e-01 8.69665325e-01
6.37778461e-01 4.38694417e-01 -3.08677733e-01 5.31656086e-01
4.10884023e-01 8.16439152e-01 -1.78022906e-01 -5.57743251e-01
8.58054399e-01 3.08439165e-01 3.98780167e-01 -1.70436168e+00
-6.91105425e-01 -5.61544716e-01 -1.20381558e+00 -3.43830884e-02
4.62553233e-01 -5.82553804e-01 -8.73789608e-01 1.75754106e+00
2.25521117e-01 4.09217983e-01 -1.81103274e-01 7.70934582e-01
4.78915185e-01 9.80276167e-01 4.23563570e-01 4.58857268e-01
7.70089984e-01 -4.50047940e-01 -7.48014927e-01 -5.38708121e-02
4.13982749e-01 -2.83998609e-01 8.65493894e-01 -4.19611186e-02
-6.02719247e-01 -3.93309742e-01 -6.36756420e-01 3.46663259e-02
-8.23552966e-01 -1.27831548e-01 8.96440506e-01 5.84096193e-01
-8.62596214e-01 2.25551158e-01 -8.06193292e-01 3.34312953e-02
5.16561687e-01 7.76599944e-02 -1.62799641e-01 -1.92853257e-01
-1.52115774e+00 9.04213667e-01 1.80977583e-02 4.14344162e-01
-7.06759274e-01 -1.40422595e+00 -1.01096046e+00 3.14236701e-01
2.54644305e-01 -4.77278054e-01 8.57163310e-01 -3.94998603e-02
-8.77848923e-01 -6.33828994e-03 1.28052637e-01 -6.18593872e-01
4.76374090e-01 6.03943272e-03 -9.87655461e-01 -4.71690953e-01
2.16046140e-01 3.07586730e-01 6.31167054e-01 -1.11963344e+00
-6.33488894e-01 -5.77855334e-02 -4.50562723e-02 -2.88909048e-01
-5.86592704e-02 -7.06623256e-01 -3.03022824e-02 -8.98530185e-01
-2.84025043e-01 -8.63505840e-01 -4.06593889e-01 -2.61259135e-02
-4.15121228e-01 -3.60314786e-01 3.82797629e-01 -3.68338317e-01
1.78918111e+00 -2.20576262e+00 -2.83741236e-01 5.14163256e-01
4.83641446e-01 3.57782911e-03 -1.49390757e-01 9.34615791e-01
7.91317523e-02 1.73421770e-01 -5.90583012e-02 -1.98881850e-01
7.27276087e-01 3.98693293e-01 -5.37379265e-01 4.82656837e-01
5.22710502e-01 1.14092755e+00 -1.00550175e+00 -8.61631185e-02
3.67734760e-01 6.87004030e-01 -4.93184477e-01 -2.30655223e-01
-2.88773924e-01 -8.99369270e-02 -3.86136711e-01 3.25553477e-01
1.02091026e+00 -3.45368624e-01 -1.28328949e-01 -1.98619753e-01
-4.08760697e-01 2.50664622e-01 -1.07431519e+00 1.11189520e+00
-3.79461467e-01 8.59789848e-01 -2.37946361e-01 -8.20133746e-01
8.24818850e-01 -9.35853422e-02 5.67090869e-01 -8.96492124e-01
-7.30155036e-03 -3.75709273e-02 -1.67474687e-01 -3.78634542e-01
7.31809080e-01 -7.79336616e-02 -4.40082341e-01 3.67009193e-01
-4.35566068e-01 1.19217016e-01 4.16020043e-02 3.03972006e-01
7.45135963e-01 -2.20133528e-01 -4.59445447e-01 -4.19892877e-01
-2.34675169e-01 3.31124887e-02 4.10968751e-01 6.00279987e-01
-2.03490570e-01 3.56489480e-01 8.63314331e-01 -8.23442340e-01
-9.45512950e-01 -1.64316273e+00 -2.77610570e-01 9.53616679e-01
-9.34898183e-02 -9.56103876e-02 -6.56822547e-02 -3.15493137e-01
1.05229402e+00 1.00640726e+00 -1.08924460e+00 -1.99119881e-01
-9.47041065e-02 -7.89876997e-01 5.09859920e-01 8.05136442e-01
-1.05085388e-01 -6.01656437e-02 -6.85440376e-02 3.65405351e-01
-8.84568691e-02 -9.10813034e-01 -6.04950249e-01 2.13982433e-01
-2.79892474e-01 -6.48682475e-01 -4.69757885e-01 -2.35436007e-01
2.33396098e-01 2.80472338e-01 1.14696455e+00 -5.64453125e-01
1.43790094e-03 2.93949068e-01 3.55590954e-02 -4.86870050e-01
9.39701125e-02 9.56834331e-02 3.37714851e-01 1.80694789e-01
7.49575436e-01 -8.19504619e-01 -7.42707133e-01 1.41318202e-01
-8.95412385e-01 -4.73001242e-01 4.52691793e-01 6.65256858e-01
2.77692437e-01 3.87994736e-01 1.00521314e+00 -4.04199153e-01
1.07012081e+00 -1.32103300e+00 -7.15718627e-01 -7.02973157e-02
-8.77974808e-01 1.00932576e-01 4.02828991e-01 -4.95745301e-01
-7.20568538e-01 -5.79291701e-01 -3.90809737e-02 -5.95438063e-01
-6.57745032e-03 1.03173625e+00 4.23101723e-01 3.17331582e-01
4.24262583e-01 -6.77036075e-03 6.85122889e-03 -3.03496838e-01
4.88677084e-01 3.97249341e-01 4.94299859e-01 -5.79068840e-01
1.03286064e+00 3.99486512e-01 1.35369599e-01 -8.54578257e-01
-7.13610530e-01 -1.19959265e-01 -6.25243545e-01 -1.37848422e-01
8.10141981e-01 -1.00837493e+00 -8.63985062e-01 1.01265945e-02
-8.73908162e-01 -5.34873128e-01 -3.50616962e-01 7.30139852e-01
-1.75391629e-01 -2.04594079e-02 -4.16677594e-01 -1.14212525e+00
2.28409767e-01 -7.42308080e-01 9.63890135e-01 -8.93819705e-02
-1.54205933e-01 -1.61927998e+00 2.61188537e-01 -3.35422993e-01
9.28701818e-01 7.42273211e-01 1.27892935e+00 -2.78791398e-01
-4.94475186e-01 -3.19296300e-01 -6.96882844e-01 -9.88630299e-03
1.23310491e-01 1.84794977e-01 -5.57614744e-01 -5.93885221e-02
-7.20051765e-01 6.80721924e-02 7.54313469e-01 8.23111653e-01
1.09803689e+00 -3.04513454e-01 -3.20511967e-01 7.04177320e-01
1.28647292e+00 -9.53455716e-02 3.00290406e-01 1.40405342e-01
7.43625939e-01 6.13379955e-01 2.48286992e-01 7.94862986e-01
1.25927567e+00 3.33043665e-01 5.71808815e-01 -5.40834852e-02
7.26832077e-02 -8.30551445e-01 3.42835672e-02 5.96722603e-01
4.91150290e-01 -6.17068291e-01 -1.35033464e+00 9.44048703e-01
-1.75959599e+00 -1.08664262e+00 -4.46220815e-01 1.85052204e+00
4.49888617e-01 2.95869619e-01 3.60053867e-01 -2.65929341e-01
3.60699564e-01 5.74647248e-01 -7.30609238e-01 -6.38345718e-01
-2.40180001e-01 -3.89982194e-01 8.07734311e-01 6.30698919e-01
-1.02934754e+00 6.93735361e-01 6.98948669e+00 6.42187357e-01
-1.04636240e+00 -8.12808424e-02 6.90438569e-01 -2.27588564e-01
-8.89982998e-01 -6.68082416e-01 -7.15501130e-01 7.63060570e-01
1.42803812e+00 -4.05915439e-01 3.27007473e-01 5.24259567e-01
5.53884089e-01 6.53481111e-02 -1.02203870e+00 9.30073261e-01
-3.66139770e-01 -1.63127708e+00 7.81340674e-02 2.60288447e-01
8.17684948e-01 4.94710743e-01 9.33760643e-01 5.56806743e-01
9.27583575e-01 -1.36250508e+00 5.76670527e-01 7.36075282e-01
9.30658400e-01 -1.13401985e+00 4.66030180e-01 1.47117451e-01
-1.36222541e+00 -3.62066835e-01 -4.82014000e-01 -6.81612715e-02
4.90047961e-01 6.96265817e-01 -5.52322030e-01 2.84534037e-01
6.40809357e-01 9.22982812e-01 -3.34055126e-01 9.00737703e-01
3.87323439e-01 5.64415216e-01 -6.69488847e-01 2.02522017e-02
9.33381259e-01 -4.08200473e-01 3.04064304e-01 1.37149906e+00
4.77720261e-01 -2.03119576e-01 1.59816533e-01 1.03945661e+00
4.37165797e-02 -4.83597219e-01 -9.95063901e-01 -2.70378917e-01
6.84672773e-01 6.93127573e-01 -1.14917912e-01 4.32078056e-02
-6.23200417e-01 2.62162000e-01 3.28794926e-01 8.42099786e-01
-1.04598200e+00 -3.07663321e-01 1.34511936e+00 3.30030650e-01
5.36468625e-01 -8.16591084e-01 -3.43963772e-01 -8.49382222e-01
-8.35706964e-02 -2.89532453e-01 3.01852256e-01 -5.64970195e-01
-2.07569337e+00 3.88434082e-01 2.45723754e-01 -1.24564052e+00
-1.42425895e-01 -6.56537652e-01 -5.69218934e-01 9.63552415e-01
-1.93423998e+00 -1.27879620e+00 2.01450605e-02 7.39037275e-01
4.60743532e-02 9.37089920e-02 4.00327891e-01 6.82680011e-01
-6.81489527e-01 7.89110601e-01 7.44466066e-01 7.32374266e-02
3.12746644e-01 -1.40105629e+00 8.76307487e-01 6.12782001e-01
-2.55611598e-01 6.16704345e-01 6.82383776e-01 -7.15479493e-01
-1.59488499e+00 -1.50302100e+00 8.85059595e-01 -8.54155898e-01
1.46837807e+00 -4.68183964e-01 -7.74458766e-01 8.26172173e-01
1.36599066e-02 6.37180582e-02 9.86284435e-01 4.25856322e-01
-6.15874112e-01 -4.44799960e-01 -9.77705479e-01 6.91733897e-01
7.85545111e-01 -6.61628664e-01 -3.83716404e-01 3.72344583e-01
9.80058849e-01 -2.86469911e-03 -1.25693941e+00 1.36836201e-01
8.12654316e-01 -6.01827204e-01 1.03309011e+00 -9.21557426e-01
3.60657662e-01 -8.72536004e-02 -4.20433670e-01 -1.69830060e+00
-7.22211301e-01 -5.61052144e-01 -1.83284521e-01 9.84671831e-01
7.49371350e-01 -8.28140914e-01 7.44214952e-01 1.09472263e+00
-1.80601135e-01 -6.03911757e-01 -1.09550524e+00 -9.14978623e-01
5.45436382e-01 -9.28129852e-01 1.24377990e+00 1.06174862e+00
-2.34794527e-01 3.23820449e-02 -6.98904812e-01 6.33978248e-01
7.83242047e-01 1.57544091e-02 7.27463484e-01 -1.36713219e+00
2.30103269e-01 -6.17685318e-01 -3.84888738e-01 -1.23555124e+00
1.06315687e-01 -7.57501841e-01 -1.58537582e-01 -1.57327974e+00
-4.39539373e-01 -8.79369497e-01 -5.45279503e-01 4.68405746e-02
1.85281813e-01 -1.75768565e-02 -2.15402573e-01 -3.03878367e-01
-4.63665605e-01 9.01124179e-01 7.78277099e-01 -3.08907062e-01
-7.16458708e-02 5.73787466e-02 -6.84531033e-01 3.19750458e-01
6.37019813e-01 -3.58254015e-01 -7.02718198e-01 -1.12030280e+00
5.86327374e-01 9.90304351e-02 4.32388723e-01 -6.94133103e-01
4.32952017e-01 -5.21959603e-01 4.88805175e-01 -1.24732733e+00
3.41919929e-01 -1.05073249e+00 2.97369182e-01 1.41636938e-01
-2.77427495e-01 7.49175370e-01 5.59083521e-01 1.23672485e+00
-1.20789543e-01 8.04074287e-01 1.56756520e-01 4.93158817e-01
-6.02097869e-01 8.88660491e-01 -7.73350835e-01 6.95659667e-02
8.02940786e-01 -2.40919054e-01 -5.92800081e-01 -5.72905660e-01
-6.91809237e-01 8.19347322e-01 8.79873522e-03 7.17309654e-01
6.26094937e-01 -1.85688567e+00 -7.07241654e-01 3.97091776e-01
2.30876893e-01 -2.30052426e-01 5.86791158e-01 8.08616400e-01
-3.62432212e-01 7.12849915e-01 -1.41781822e-01 -6.04647696e-01
-2.31244676e-02 7.78007507e-01 2.39742965e-01 -4.76579443e-02
-4.47488666e-01 7.53414214e-01 9.06105489e-02 -6.41510665e-01
2.09270447e-01 -8.04317713e-01 5.73936030e-02 4.12004650e-01
2.02057645e-01 6.66448295e-01 -2.46417448e-01 -6.99109197e-01
-3.68781835e-01 3.29572320e-01 3.25122327e-01 6.26354739e-02
1.63277042e+00 -5.46971321e-01 2.33918473e-01 7.56053448e-01
1.31182158e+00 -1.16344474e-01 -1.58681619e+00 -3.97012740e-01
2.19547823e-01 -6.25846028e-01 3.70179892e-01 -7.81650305e-01
-8.67395043e-01 1.15798056e+00 4.48038429e-01 7.65511096e-01
5.39938331e-01 -3.85216177e-01 1.09199357e+00 1.61010832e-01
1.11945570e-01 -9.69282568e-01 -1.90994307e-01 6.77577853e-01
7.68794060e-01 -1.41273415e+00 -2.61470497e-01 1.73171550e-01
-7.11594582e-01 8.10907900e-01 3.69747400e-01 -2.40537748e-01
1.39746332e+00 3.03121328e-01 -1.34768441e-01 -2.49206290e-01
-9.90737081e-01 -2.35992029e-01 5.25641203e-01 9.29854572e-01
4.66767587e-02 5.84251642e-01 2.83627421e-01 6.13830924e-01
-2.90290236e-01 -2.23578408e-01 3.33722264e-01 3.86503845e-01
-1.23213157e-01 -5.85011840e-01 1.07897427e-02 7.02439785e-01
-1.40684634e-01 -1.21438704e-01 2.20286682e-01 9.02005851e-01
-1.63734406e-01 1.01562333e+00 6.51109576e-01 -7.20423579e-01
4.20466214e-01 -2.15137601e-01 -1.79506838e-01 -1.89519376e-01
-1.07777752e-01 -2.03483537e-01 -1.66849494e-02 -6.86179221e-01
3.12578499e-01 -6.40142620e-01 -7.35329092e-01 -1.07594776e+00
2.13040322e-01 2.12571189e-01 6.70314372e-01 6.10912025e-01
7.55953372e-01 5.43943346e-01 6.62423193e-01 -7.57330000e-01
-8.42241764e-01 -6.97664261e-01 -8.38711381e-01 3.17169696e-01
1.09978855e+00 -1.08926475e+00 -3.67264360e-01 -5.42243242e-01] | [6.540089130401611, 2.278674602508545] |
6ece5270-5411-4ee4-bf33-d46ed0cf004e | wild-face-anti-spoofing-challenge-2023 | 2304.05753 | null | https://arxiv.org/abs/2304.05753v3 | https://arxiv.org/pdf/2304.05753v3.pdf | Wild Face Anti-Spoofing Challenge 2023: Benchmark and Results | Face anti-spoofing (FAS) is an essential mechanism for safeguarding the integrity of automated face recognition systems. Despite substantial advancements, the generalization of existing approaches to real-world applications remains challenging. This limitation can be attributed to the scarcity and lack of diversity in publicly available FAS datasets, which often leads to overfitting during training or saturation during testing. In terms of quantity, the number of spoof subjects is a critical determinant. Most datasets comprise fewer than 2,000 subjects. With regard to diversity, the majority of datasets consist of spoof samples collected in controlled environments using repetitive, mechanical processes. This data collection methodology results in homogenized samples and a dearth of scenario diversity. To address these shortcomings, we introduce the Wild Face Anti-Spoofing (WFAS) dataset, a large-scale, diverse FAS dataset collected in unconstrained settings. Our dataset encompasses 853,729 images of 321,751 spoof subjects and 529,571 images of 148,169 live subjects, representing a substantial increase in quantity. Moreover, our dataset incorporates spoof data obtained from the internet, spanning a wide array of scenarios and various commercial sensors, including 17 presentation attacks (PAs) that encompass both 2D and 3D forms. This novel data collection strategy markedly enhances FAS data diversity. Leveraging the WFAS dataset and Protocol 1 (Known-Type), we host the Wild Face Anti-Spoofing Challenge at the CVPR2023 workshop. Additionally, we meticulously evaluate representative methods using Protocol 1 and Protocol 2 (Unknown-Type). Through an in-depth examination of the challenge outcomes and benchmark baselines, we provide insightful analyses and propose potential avenues for future research. The dataset is released under Insightface. | ['Zhen Lei', 'Jiankang Deng', 'Jun Wan', 'Hugo Jair Escalante', 'Sergio Escalera', 'Ajian Liu', 'Chuanbao Xiao', 'Zhian Chen', 'Haochi He', 'Qiqi Shao', 'Jia Guo', 'Dong Wang'] | 2023-04-12 | null | null | null | null | ['face-anti-spoofing'] | ['computer-vision'] | [ 6.73804402e-01 -4.12585646e-01 -2.10961521e-01 -1.64286882e-01
-3.12379390e-01 -8.25652063e-01 6.02400780e-01 -3.56278092e-01
-4.39380929e-02 5.10421515e-01 1.68342084e-01 -3.15471202e-01
-8.02851394e-02 -5.80426455e-01 -6.06706560e-01 -6.74986541e-01
-4.03566599e-01 -1.95115840e-03 9.83897597e-02 -1.62113607e-01
-5.95376478e-04 8.53358924e-01 -1.75386059e+00 2.40420163e-01
2.82881021e-01 1.02873123e+00 -5.93415499e-01 5.70649385e-01
5.39515197e-01 1.72374159e-01 -1.02840233e+00 -7.24303842e-01
4.94909972e-01 -1.96641713e-01 -4.06923532e-01 -3.37396725e-03
1.04041159e+00 -5.04624963e-01 -4.36033100e-01 9.29276109e-01
8.50599229e-01 -2.33728632e-01 2.30012208e-01 -1.86415923e+00
-4.15017843e-01 1.10025972e-01 -7.00449526e-01 2.41780072e-01
5.97742081e-01 3.58554900e-01 5.56257248e-01 -7.18390286e-01
7.69272983e-01 1.60333920e+00 8.35456133e-01 7.97628462e-01
-1.19863009e+00 -1.17960525e+00 -3.51223111e-01 3.71252522e-02
-1.29856372e+00 -1.41692352e+00 8.54767859e-01 -2.79792100e-01
6.16797149e-01 2.35996857e-01 5.41903198e-01 1.88995731e+00
1.72710828e-02 3.62690508e-01 1.34568560e+00 -1.31754398e-01
1.05638109e-01 1.38394237e-01 -2.51885373e-02 4.69705343e-01
8.58242929e-01 5.93358159e-01 -8.87649477e-01 -8.06590855e-01
3.26878518e-01 -4.37583923e-02 -3.16566736e-01 -4.23945963e-01
-7.89813995e-01 7.33615994e-01 -1.05410758e-02 -9.05759484e-02
-8.24665930e-03 -2.71595418e-01 4.90256965e-01 3.23441625e-01
8.61906633e-02 9.61308628e-02 -3.55805755e-01 -4.17366847e-02
-9.69515860e-01 3.75829428e-01 9.70679700e-01 6.51919782e-01
5.03764808e-01 1.98152214e-01 3.96478146e-01 6.75227344e-01
4.63437557e-01 1.18198514e+00 1.07802875e-01 -9.72596824e-01
2.95983672e-01 -6.68035895e-02 -3.55664194e-02 -1.80803728e+00
-1.12566069e-01 -1.46632701e-01 -4.79109436e-01 1.40241951e-01
3.59974951e-01 -7.30949715e-02 -7.57176638e-01 1.92922819e+00
5.01258254e-01 2.10423306e-01 -1.02564745e-01 5.44467747e-01
7.49997973e-01 1.46562159e-01 -4.74351794e-02 -4.25835878e-01
1.46150851e+00 -4.09523815e-01 -6.29621804e-01 -2.22818315e-01
1.34034872e-01 -8.98899198e-01 9.23725128e-01 5.79965413e-01
-5.28018653e-01 -2.39738718e-01 -1.03103805e+00 4.21423852e-01
-4.23257917e-01 -3.69968772e-01 5.32535255e-01 1.59713078e+00
-1.09204614e+00 3.04903090e-01 -5.95626354e-01 -7.32424438e-01
9.39903021e-01 4.73445356e-01 -7.98551857e-01 -3.27400833e-01
-1.04560888e+00 2.64860094e-01 -1.08010978e-01 -1.83832482e-01
-1.31951702e+00 -8.66196632e-01 -8.47643435e-01 -3.15869391e-01
2.92725712e-01 -1.64550737e-01 6.77038252e-01 -3.91006798e-01
-1.21550786e+00 8.68620276e-01 -2.32637197e-01 -2.72830486e-01
5.61046779e-01 -2.23208770e-01 -1.04649103e+00 5.38336098e-01
-3.16075906e-02 3.89737010e-01 1.26689625e+00 -1.39844155e+00
-5.66214025e-02 -5.69180012e-01 -2.30776191e-01 -6.65228069e-01
-6.19701385e-01 4.30887789e-01 -2.53086016e-02 -7.40710020e-01
-1.81997329e-01 -1.02946615e+00 4.17784363e-01 -4.51165019e-03
-5.08273005e-01 2.96459705e-01 1.63351011e+00 -6.52743816e-01
9.40872788e-01 -2.34059548e+00 -5.14240861e-01 2.95133501e-01
2.95042783e-01 8.07434082e-01 -3.77216488e-01 3.90703589e-01
-1.70233652e-01 5.74402988e-01 -2.92042941e-01 -4.90054011e-01
-1.69615254e-01 -6.29970944e-03 -4.72014606e-01 8.84815156e-01
2.97367275e-01 4.38281268e-01 -8.62650514e-01 -5.62487900e-01
2.23436713e-01 7.66535342e-01 -8.92928302e-01 1.59425288e-01
9.70869884e-02 4.12281573e-01 -2.33984306e-01 1.25986481e+00
1.25379121e+00 1.28098294e-01 1.25734910e-01 -3.92160088e-01
4.01632041e-01 2.13060603e-01 -9.83603358e-01 1.25809348e+00
-7.39269555e-02 7.28585005e-01 5.01071930e-01 -6.60279632e-01
7.99506843e-01 2.87145913e-01 8.35459948e-01 -4.23883408e-01
2.21045047e-01 2.15656072e-01 -1.67484254e-01 -6.73734546e-01
2.39833027e-01 8.69616792e-02 2.20172301e-01 5.96383691e-01
7.71195740e-02 -4.04893747e-03 1.01125941e-01 1.27625108e-01
1.49576712e+00 -2.64402419e-01 1.00913141e-02 -2.83212155e-01
3.60121578e-01 -2.98899829e-01 7.07215071e-01 6.60336733e-01
-1.09446084e+00 4.02709365e-01 3.27322245e-01 -2.47167796e-01
-7.82636702e-01 -1.21948433e+00 -5.92282832e-01 7.40815163e-01
3.36659737e-02 -6.92063034e-01 -6.46176577e-01 -8.22831810e-01
4.41154867e-01 2.30407212e-02 -6.39227986e-01 -1.02370083e-01
-6.01160586e-01 -8.75667632e-01 1.25820899e+00 -1.08125418e-01
7.73446202e-01 -9.30771410e-01 -4.62471306e-01 -2.31899157e-01
-1.96177930e-01 -1.48395145e+00 -1.92282394e-01 -4.99688357e-01
-4.00995344e-01 -1.66299760e+00 -1.49157494e-01 -3.13363403e-01
3.71729940e-01 6.89582765e-01 1.12015641e+00 3.32265705e-01
-5.67541420e-01 4.17584896e-01 -2.28104785e-01 -4.40577596e-01
-5.12539804e-01 -3.16007733e-01 7.43057609e-01 2.33716592e-01
5.87424099e-01 -6.36579096e-01 -5.64368129e-01 6.05092824e-01
-8.98934960e-01 -7.04219580e-01 2.32855946e-01 6.30526721e-01
-6.46960959e-02 1.17194429e-01 6.08696461e-01 -7.57097304e-01
3.18425655e-01 -7.97196865e-01 -3.29117835e-01 5.58371767e-02
-5.48761606e-01 -3.94893110e-01 2.83479095e-01 -4.54209834e-01
-7.63489962e-01 -3.36229563e-01 -2.23171432e-02 -4.78600174e-01
-3.63181442e-01 -2.40834534e-01 -4.90477920e-01 -4.83768523e-01
7.59847999e-01 -6.66574314e-02 4.62084889e-01 -3.54695380e-01
-2.53397319e-02 9.98351038e-01 6.14881158e-01 -6.07392192e-01
1.10156763e+00 7.81818569e-01 3.54134403e-02 -1.45399261e+00
-1.55240089e-01 -1.53043807e-01 -1.83947086e-01 -1.81365922e-01
2.61913955e-01 -8.00128758e-01 -1.17105389e+00 1.00818586e+00
-8.86532724e-01 -1.46719068e-01 1.34205490e-01 1.68321997e-01
-5.82730956e-02 6.70834422e-01 -4.74232584e-01 -9.33424950e-01
-2.42846310e-01 -1.26814079e+00 1.14109218e+00 8.65577608e-02
-1.87906042e-01 -7.26643205e-01 -6.41280040e-02 6.93535805e-01
5.66259027e-01 6.82577014e-01 3.64165664e-01 -7.50689685e-01
-1.06905121e-02 -2.28012040e-01 -2.30304927e-01 3.90049607e-01
5.46645224e-01 3.40388983e-01 -1.34505320e+00 -9.29660380e-01
1.01291426e-01 -5.88299036e-01 7.32668579e-01 -3.65645140e-02
9.10238922e-01 -3.86380404e-01 -4.06784743e-01 7.05489755e-01
1.02507424e+00 1.01389922e-02 5.56470931e-01 1.14701897e-01
6.23855472e-01 8.87324154e-01 2.96205133e-01 5.92248082e-01
-1.74566098e-02 6.65216565e-01 6.37979925e-01 3.65512967e-01
-3.62583607e-01 -3.29921246e-01 7.22494483e-01 2.24243715e-01
3.40941519e-01 -4.72436070e-01 -8.46381426e-01 3.72237742e-01
-8.60674620e-01 -1.22004294e+00 2.33259663e-01 2.21052527e+00
6.21193051e-01 -1.30359262e-01 4.23993826e-01 5.16490161e-01
9.57034469e-01 6.61281228e-01 -4.76994932e-01 -5.21951169e-02
-4.15008366e-01 1.32091060e-01 4.44921136e-01 3.45220923e-01
-1.19024551e+00 8.33154678e-01 6.90820312e+00 6.65213227e-01
-1.31096613e+00 4.55477573e-02 4.22065556e-01 -2.73133188e-01
-8.69875401e-02 -2.95357764e-01 -9.93530452e-01 7.67371595e-01
1.16784835e+00 1.11782365e-01 7.18371570e-01 5.75197518e-01
1.24423601e-01 1.54523715e-01 -8.32988560e-01 1.12352252e+00
5.09410143e-01 -1.10477960e+00 -7.99131542e-02 5.61172187e-01
5.83518505e-01 3.81266959e-02 3.73584181e-01 -2.37925991e-01
3.35454464e-01 -1.10218668e+00 4.83164966e-01 -3.14432770e-01
1.14487636e+00 -4.47829098e-01 3.63092184e-01 -1.85579211e-01
-1.08369732e+00 -3.56752694e-01 -2.18819901e-01 3.13573688e-01
3.88472639e-02 7.33206153e-01 -6.65583253e-01 4.29073900e-01
9.76673663e-01 6.85343981e-01 -5.76157749e-01 5.37176609e-01
2.09169369e-02 9.96972084e-01 -7.56347835e-01 4.17074353e-01
-4.45346206e-01 3.64181101e-01 7.97127664e-01 1.04663813e+00
2.15153322e-01 -1.66238353e-01 -7.34902397e-02 5.08002698e-01
-2.76826411e-01 -3.56615126e-01 -1.02810884e+00 -4.04574633e-01
1.10676444e+00 1.07066894e+00 -3.15078110e-01 8.68893713e-02
-1.63910642e-01 5.25488734e-01 -1.93383798e-01 3.73014212e-01
-6.17389977e-01 -1.69679314e-01 1.40229595e+00 2.61503369e-01
-1.00702408e-03 -2.66530305e-01 3.90140787e-02 -1.09223413e+00
4.58999872e-02 -1.52004373e+00 4.72222000e-01 -1.76100597e-01
-1.39464784e+00 5.11989892e-01 -3.49183716e-02 -7.89262652e-01
-1.89833701e-01 -7.09658980e-01 -2.27705911e-01 5.24686396e-01
-1.42993915e+00 -1.10590255e+00 -5.53580344e-01 7.52751350e-01
1.23881593e-01 -5.87360382e-01 7.31775522e-01 5.39435089e-01
-8.60540152e-01 1.09158933e+00 -2.66995430e-01 2.03027561e-01
1.02269471e+00 -3.25991303e-01 4.37790841e-01 1.00313580e+00
-1.14385799e-01 1.04657292e+00 8.40809047e-01 -8.26605856e-01
-1.87608874e+00 -9.69291568e-01 5.12612700e-01 -6.11153960e-01
6.12245679e-01 -5.46466649e-01 -7.47935712e-01 5.10787606e-01
-1.46955669e-01 6.10181749e-01 8.73989582e-01 -3.04668248e-01
-9.16432619e-01 -2.94852406e-01 -1.87442994e+00 3.63124281e-01
1.33507276e+00 -7.71580696e-01 -2.68464275e-02 1.78540379e-01
5.84940672e-01 -8.48085061e-02 -8.85932863e-01 7.78956473e-01
1.01565552e+00 -1.32831347e+00 1.16412532e+00 -6.10462308e-01
1.63521081e-01 -7.59244338e-02 -4.89107281e-01 -8.81179035e-01
1.83889210e-01 -1.20172012e+00 -5.23673356e-01 1.64711916e+00
-2.03274325e-01 -1.00831437e+00 1.05226123e+00 3.29771072e-01
4.14343178e-01 -4.30493653e-01 -9.18278635e-01 -1.00435913e+00
-1.03113376e-01 -3.38057309e-01 1.13623512e+00 1.05268610e+00
-3.54841679e-01 -4.55545098e-01 -5.44979930e-01 3.87445688e-01
1.20588136e+00 -4.40403253e-01 1.12383115e+00 -1.15995646e+00
-1.98576257e-01 -9.37998816e-02 -6.46860063e-01 -6.54664278e-01
3.83991182e-01 -4.92120922e-01 -2.81838089e-01 -3.61415148e-01
1.47275686e-01 -6.49554968e-01 -9.76396054e-02 4.33242768e-01
2.98848916e-02 7.92849898e-01 2.09717244e-01 3.93372089e-01
8.51681307e-02 2.16831997e-01 8.16099346e-01 -1.50027469e-01
1.41293749e-01 -2.43706703e-01 -6.60238147e-01 4.00613457e-01
8.15665245e-01 -4.04692322e-01 -3.92631650e-01 -2.78626502e-01
-2.45955572e-01 -2.10180610e-01 6.97202981e-01 -1.15234661e+00
-8.15897211e-02 -1.90978304e-01 3.77738655e-01 -1.48324281e-01
5.48494756e-01 -8.81472945e-01 1.89246744e-01 6.66279793e-01
-2.83930842e-02 4.23815213e-02 3.07106197e-01 4.39062268e-01
1.87724486e-01 3.67043108e-01 9.29485202e-01 3.18378627e-01
-3.61576349e-01 5.69318891e-01 5.28832637e-02 2.70820796e-01
9.60421622e-01 -4.52209979e-01 -1.05551124e+00 -1.14526406e-01
-1.16879322e-01 -2.91917831e-01 9.10152078e-01 5.91123104e-01
4.93620306e-01 -1.13806283e+00 -7.12208271e-01 9.20140266e-01
1.42633721e-01 -5.20263314e-01 9.34294686e-02 4.88256007e-01
-4.87324715e-01 3.19797665e-01 -4.34188426e-01 -6.60427988e-01
-1.52969229e+00 5.21900892e-01 1.06974758e-01 1.36069953e-01
-2.98820406e-01 6.95756197e-01 -2.37640738e-01 -3.07789654e-01
5.21285720e-02 6.56272411e-01 1.66653112e-01 -1.02633856e-01
8.96607459e-01 6.65497065e-01 1.37815312e-01 -1.18143702e+00
-7.75339842e-01 3.19145381e-01 9.14002284e-02 -3.30573246e-02
1.16246748e+00 -9.11732689e-02 -1.61352187e-01 -3.28394860e-01
1.48683310e+00 5.44841588e-01 -1.20209026e+00 -9.77574755e-03
-2.87931979e-01 -1.07412875e+00 -3.31139714e-01 -5.53155243e-01
-1.31967521e+00 8.18513155e-01 8.16793263e-01 2.12935850e-01
1.04041541e+00 -1.33672446e-01 8.75915885e-01 -5.64364344e-03
6.42650843e-01 -3.74398679e-01 2.51297176e-01 1.55448318e-01
7.47671127e-01 -1.31640637e+00 4.53843102e-02 -7.63840735e-01
-3.03320517e-03 7.62700856e-01 5.05029619e-01 2.74769902e-01
7.55416453e-01 5.20646751e-01 1.47296011e-01 -1.95798650e-01
-3.48947257e-01 4.86969858e-01 -3.75177562e-01 1.22065842e+00
1.22490481e-01 -7.62664825e-02 1.67330652e-01 1.76534981e-01
-5.52799881e-01 -1.17225766e-01 2.96760589e-01 1.22642708e+00
-3.73399258e-02 -1.24243987e+00 -8.00760448e-01 3.23677957e-01
-6.92409873e-01 3.56674343e-01 -4.98293072e-01 4.95633602e-01
1.74450383e-01 1.71565378e+00 -2.26978362e-01 -8.10721517e-01
-6.95368275e-02 -1.86509505e-01 3.33429635e-01 -2.55114436e-01
-2.87595302e-01 -5.67350507e-01 8.02632794e-02 -8.47460032e-01
-3.75741661e-01 -9.23621178e-01 -5.20553231e-01 -1.21074963e+00
-1.40774727e-01 -8.08403268e-02 7.89788425e-01 6.44777536e-01
7.52891898e-01 -4.80543338e-02 1.03634608e+00 -9.13491607e-01
-6.92001402e-01 -7.06850350e-01 -4.92779911e-01 6.85553968e-01
7.75268376e-01 -1.01017094e+00 -7.50700355e-01 1.35607734e-01] | [13.017666816711426, 1.133664846420288] |
5be4780f-c9c3-4d36-aaa7-65ee0676c659 | spatiotemporal-augmentation-on-selective | 2204.03865 | null | https://arxiv.org/abs/2204.03865v2 | https://arxiv.org/pdf/2204.03865v2.pdf | Frequency Selective Augmentation for Video Representation Learning | Recent self-supervised video representation learning methods focus on maximizing the similarity between multiple augmented views from the same video and largely rely on the quality of generated views. However, most existing methods lack a mechanism to prevent representation learning from bias towards static information in the video. In this paper, we propose frequency augmentation (FreqAug), a spatio-temporal data augmentation method in the frequency domain for video representation learning. FreqAug stochastically removes specific frequency components from the video so that learned representation captures essential features more from the remaining information for various downstream tasks. Specifically, FreqAug pushes the model to focus more on dynamic features rather than static features in the video via dropping spatial or temporal low-frequency components. To verify the generality of the proposed method, we experiment with FreqAug on multiple self-supervised learning frameworks along with standard augmentations. Transferring the improved representation to five video action recognition and two temporal action localization downstream tasks shows consistent improvements over baselines. | ['Junmo Kim', 'Dongyoon Wee', 'Dongyoon Han', 'Minho Shim', 'Taeoh Kim', 'Jinhyung Kim'] | 2022-04-08 | null | null | null | null | ['action-localization'] | ['computer-vision'] | [ 4.39548254e-01 -5.15458314e-03 -5.78506768e-01 -3.33591074e-01
-7.87757576e-01 -4.71558005e-01 5.86880744e-01 -2.31602564e-01
-1.50680274e-01 5.62813342e-01 9.99352992e-01 3.06824833e-01
2.67499387e-02 -4.10401255e-01 -8.39079380e-01 -7.07100093e-01
-3.36419076e-01 -1.46913290e-01 1.48226693e-01 -1.34224370e-01
8.49078894e-02 7.25312307e-02 -1.60476041e+00 8.39256287e-01
3.43340904e-01 7.52080023e-01 1.66119009e-01 5.88095009e-01
3.11611980e-01 1.09814513e+00 -4.92308557e-01 3.00634891e-01
4.77185398e-01 -6.54407203e-01 -7.54252970e-01 3.30508560e-01
8.26552033e-01 -5.90548456e-01 -8.07840765e-01 7.66871691e-01
1.16002761e-01 5.47167718e-01 6.14393950e-01 -1.42083061e+00
-7.21367657e-01 4.54296261e-01 -7.91307390e-01 6.96011126e-01
5.85973263e-01 9.85909104e-02 1.07100415e+00 -8.48926127e-01
8.01477015e-01 1.28872490e+00 5.97029567e-01 6.50665760e-01
-1.18013120e+00 -4.46578473e-01 7.15753198e-01 3.29335511e-01
-1.03576803e+00 -5.64903975e-01 8.62666845e-01 -5.17620742e-01
1.11129558e+00 3.33956182e-02 5.50906122e-01 1.40194142e+00
1.78132746e-02 9.15525019e-01 8.60917091e-01 -3.78662467e-01
2.29642205e-02 -2.59263277e-01 2.71692257e-02 6.57062292e-01
-1.33317068e-01 1.30321935e-01 -1.04547572e+00 -1.34319454e-01
1.00322509e+00 2.85055637e-01 -5.13142169e-01 -6.85683608e-01
-1.36057091e+00 7.95518517e-01 2.90222406e-01 3.08282167e-01
-3.49767327e-01 4.78982031e-01 5.49626946e-01 2.34256819e-01
5.28553247e-01 3.39051992e-01 -6.62516057e-01 -4.38197374e-01
-7.94014454e-01 1.30923897e-01 1.66493971e-02 7.83936799e-01
6.60717726e-01 2.81913519e-01 -4.83982176e-01 6.13177776e-01
2.68165708e-01 3.31318289e-01 8.71786714e-01 -1.25956249e+00
4.81752545e-01 5.19339025e-01 -1.50343418e-01 -8.13383698e-01
-1.07072577e-01 -9.09807459e-02 -3.47794563e-01 8.51538032e-02
3.32553864e-01 -1.64534375e-01 -1.07680202e+00 2.21202803e+00
1.07684605e-01 4.77345765e-01 5.33787869e-02 9.10617530e-01
6.88556433e-01 7.16022909e-01 1.28956109e-01 -2.32062042e-01
1.09276533e+00 -1.14493525e+00 -7.89354265e-01 -2.57139355e-01
7.75364637e-01 -4.11071002e-01 1.16089690e+00 1.96710199e-01
-8.28981340e-01 -7.63114333e-01 -9.90038335e-01 -8.90460517e-03
3.24961208e-02 2.23683923e-01 6.89839602e-01 1.91777438e-01
-1.03005183e+00 6.26674831e-01 -9.61398542e-01 -4.52744812e-01
3.87214243e-01 2.04942510e-01 -8.12413812e-01 -2.11118460e-01
-1.10878611e+00 5.05769730e-01 2.89867818e-01 -3.75095516e-01
-9.52251196e-01 -7.12678552e-01 -1.22840011e+00 2.11376846e-02
4.30951297e-01 -4.77597415e-01 1.15920269e+00 -1.48336029e+00
-1.29171133e+00 5.46004772e-01 -3.29779029e-01 -5.24685681e-01
1.69474274e-01 -6.95506871e-01 -2.98260212e-01 5.12704372e-01
1.84605986e-01 8.36004198e-01 1.24943304e+00 -1.03874147e+00
-5.32715321e-01 -3.34431082e-01 3.26508105e-01 3.82989734e-01
-4.54294413e-01 -3.99574041e-01 -3.53462070e-01 -1.23234868e+00
4.50813733e-02 -1.05351901e+00 1.17055871e-01 -2.45258640e-02
7.10922480e-02 -1.06630147e-01 1.34137821e+00 -5.44404268e-01
1.07828426e+00 -2.45719337e+00 3.09779614e-01 -7.53505751e-02
-2.20019966e-02 1.07667826e-01 -5.71329832e-01 3.89782697e-01
-5.99414170e-01 -1.77570492e-01 4.21555378e-02 -1.15601026e-01
-3.97119731e-01 2.36249968e-01 -3.70248795e-01 5.19426703e-01
3.61974627e-01 7.83277094e-01 -1.15632415e+00 -1.89012423e-01
3.06422144e-01 5.54222584e-01 -9.31709707e-01 1.83473334e-01
-2.71112293e-01 5.79948127e-01 -3.81630033e-01 5.11561871e-01
2.11202472e-01 -2.66748637e-01 1.18719473e-01 -5.13262749e-01
2.66129136e-01 2.76317090e-01 -8.74348938e-01 2.16971302e+00
-3.01247805e-01 7.59954870e-01 -2.64013320e-01 -1.07592392e+00
4.99539912e-01 4.18974817e-01 1.14051628e+00 -5.81045032e-01
-1.84722260e-01 -3.33368093e-01 -4.16334420e-02 -6.09407604e-01
4.33960676e-01 -2.22387891e-02 1.17496200e-01 5.80155849e-01
4.84957606e-01 4.57377583e-01 2.64513940e-01 4.28415954e-01
1.29505348e+00 8.27808559e-01 2.63106018e-01 -6.35252660e-03
3.43375295e-01 -1.32827550e-01 6.17229700e-01 4.79174167e-01
-4.44910705e-01 8.42115462e-01 2.49420658e-01 -2.46493146e-01
-6.90930605e-01 -8.97456586e-01 3.03712457e-01 1.54780221e+00
-1.16993468e-02 -9.69200730e-01 -5.50110102e-01 -1.18135834e+00
1.39634265e-02 4.24032003e-01 -9.66666996e-01 -6.27501130e-01
-6.11576796e-01 -3.41606617e-01 2.52183259e-01 9.54613507e-01
5.23375392e-01 -9.99162853e-01 -6.50710940e-01 7.46005308e-03
-4.41137671e-01 -1.09360492e+00 -8.17406714e-01 8.62481222e-02
-1.04169750e+00 -1.19222176e+00 -7.04431355e-01 -3.89774472e-01
5.63407660e-01 8.59157026e-01 9.13162827e-01 -1.27688348e-01
-9.64033157e-02 9.19617176e-01 -7.37652063e-01 1.33112609e-01
-9.46983024e-02 -1.39937788e-01 1.01470746e-01 1.22429647e-01
4.36840534e-01 -6.47072256e-01 -5.41094720e-01 2.23591983e-01
-8.34048450e-01 -1.57614537e-02 3.87410879e-01 9.56257284e-01
4.56618339e-01 -1.48911372e-01 6.15411699e-01 -7.41518736e-01
1.49748430e-01 -5.25214553e-01 2.85266228e-02 -8.29089209e-02
-1.79893434e-01 1.41895995e-01 3.71222079e-01 -5.72425604e-01
-1.03678298e+00 3.36434662e-01 2.66674191e-01 -9.75529671e-01
-5.62666766e-02 4.22740549e-01 -1.70972303e-01 2.11932480e-01
7.05798626e-01 1.88268214e-01 9.41427648e-02 -3.32225561e-01
3.92238766e-01 1.04303010e-01 4.60882992e-01 -5.01754940e-01
6.79545522e-01 5.76604068e-01 -2.83440232e-01 -6.47850752e-01
-8.92395616e-01 -5.46647608e-01 -7.17501402e-01 -3.02180499e-01
8.27375054e-01 -1.22318840e+00 -2.97478884e-01 1.92367092e-01
-6.97668016e-01 -3.68863404e-01 -7.44192183e-01 7.64490724e-01
-8.82123828e-01 3.42900634e-01 -3.54804963e-01 -6.99218452e-01
2.40003377e-01 -9.47448432e-01 1.37250566e+00 3.37696774e-03
-4.03144360e-01 -8.77991438e-01 2.74596184e-01 4.63856459e-01
1.42284319e-01 2.58231252e-01 5.79439223e-01 -4.49051768e-01
-4.23976511e-01 -5.70400022e-02 2.17688885e-02 4.50879306e-01
6.58356965e-01 -4.27239053e-02 -1.22173142e+00 -3.81128728e-01
-1.16658956e-02 -6.13310695e-01 1.28981304e+00 5.34419000e-01
1.20800567e+00 -3.72013599e-01 -3.06915790e-01 5.56680799e-01
1.08676481e+00 2.00420395e-01 5.86481452e-01 3.77189279e-01
8.59399080e-01 4.66334045e-01 7.77632117e-01 4.86832917e-01
1.07868411e-01 8.15835118e-01 3.22878897e-01 -1.67445256e-03
-2.68656433e-01 -4.77192402e-01 8.99843156e-01 2.57985532e-01
-3.43576550e-01 -1.49786368e-01 -4.80067581e-01 5.50637841e-01
-2.09474349e+00 -1.34796798e+00 5.20000994e-01 2.07878280e+00
6.16564631e-01 3.26189809e-02 3.68094385e-01 1.89023197e-01
6.28494859e-01 5.65360010e-01 -5.55521071e-01 1.83435589e-01
-2.51364727e-02 -1.15967393e-01 3.15235764e-01 1.69919163e-01
-1.46272242e+00 7.72122264e-01 6.34747267e+00 4.55138683e-01
-1.04539597e+00 1.55534018e-02 4.26794946e-01 -5.10187328e-01
-1.14052013e-01 -1.56333879e-01 -3.33339155e-01 3.26955020e-01
7.32130527e-01 -8.99029374e-02 1.91808060e-01 1.02299821e+00
3.06054145e-01 -6.17853180e-02 -1.41215456e+00 9.84915912e-01
2.18682289e-01 -1.27344346e+00 2.22274780e-01 3.24317400e-04
7.11932600e-01 -1.07485401e-02 1.41732529e-01 5.20410419e-01
2.09364206e-01 -9.27790165e-01 5.72512269e-01 4.46416438e-01
7.23376930e-01 -6.13035142e-01 4.67780173e-01 4.23393212e-03
-1.28481376e+00 -2.40576312e-01 -4.29620072e-02 -7.45660439e-02
2.05234930e-01 6.65069968e-02 -5.59836864e-01 3.81828219e-01
7.21122503e-01 1.30710661e+00 -6.25010967e-01 6.43140793e-01
3.04713156e-02 6.39497697e-01 1.31456973e-02 8.27128351e-01
2.55146742e-01 -1.00643001e-02 5.71625352e-01 1.13794959e+00
1.73163280e-01 1.84276715e-01 4.39039826e-01 2.50822097e-01
-3.94746289e-02 -1.60077214e-01 -9.16322649e-01 -3.47094119e-01
1.78461775e-01 8.95625234e-01 -4.62659031e-01 -3.43811810e-01
-6.85332477e-01 1.06031930e+00 1.86938867e-01 5.05610466e-01
-8.76261652e-01 2.29441747e-01 9.25528526e-01 2.66141415e-01
4.00781512e-01 -3.23610604e-01 3.12286288e-01 -1.47514141e+00
5.24145812e-02 -1.10407698e+00 7.62922347e-01 -7.97562242e-01
-1.08448029e+00 3.95181149e-01 2.65797257e-01 -1.54437757e+00
-6.01211727e-01 -3.99609357e-01 -4.13696259e-01 4.01363015e-01
-1.26812720e+00 -1.10445189e+00 -3.58167768e-01 9.69806492e-01
1.09666598e+00 -3.99984956e-01 7.94045448e-01 1.95948496e-01
-3.87263715e-01 5.61914265e-01 -1.44237950e-01 9.28793699e-02
9.94816601e-01 -1.12147284e+00 1.70013130e-01 8.99448097e-01
4.61138189e-01 6.11573458e-01 5.21668315e-01 -6.77312136e-01
-1.35279477e+00 -1.14793503e+00 1.97508305e-01 -4.51878369e-01
4.82540637e-01 -2.00627699e-01 -1.02863574e+00 9.21291053e-01
1.52389780e-01 4.61591840e-01 7.24977195e-01 7.70267919e-02
-6.75787628e-01 9.83478874e-02 -9.70571935e-01 2.91637540e-01
1.31632090e+00 -7.58655310e-01 -6.07325792e-01 3.92799288e-01
6.66766167e-01 -2.08419800e-01 -5.54862916e-01 5.67415953e-01
5.55812836e-01 -9.81575966e-01 1.01244032e+00 -1.07662880e+00
5.56616187e-01 -3.94312769e-01 -4.38102603e-01 -1.36297417e+00
-5.21263540e-01 -4.91994381e-01 -7.01391637e-01 1.11314297e+00
-1.89950708e-02 -1.14513390e-01 9.42419410e-01 1.94758981e-01
-1.66224867e-01 -4.39519495e-01 -7.58695722e-01 -7.35939026e-01
-3.24668020e-01 -2.51760960e-01 9.31495577e-02 1.04093230e+00
1.46795124e-01 3.97252232e-01 -7.09188342e-01 -8.05140659e-03
3.37702841e-01 -4.40779999e-02 6.96845412e-01 -8.57409716e-01
-5.06983280e-01 -1.37241215e-01 -6.88612580e-01 -1.04787278e+00
3.00410807e-01 -6.66964650e-01 2.36817095e-02 -1.30858731e+00
3.86478961e-01 1.08099975e-01 -4.79688883e-01 7.01939702e-01
-3.41550857e-01 3.43499511e-01 3.08441430e-01 3.28435093e-01
-6.19585991e-01 8.46675038e-01 1.10510397e+00 -1.54776812e-01
-1.98570490e-01 -1.82215750e-01 -5.99350274e-01 8.21211576e-01
6.03822768e-01 -1.85098588e-01 -1.12375736e+00 -4.72130686e-01
-3.00478518e-01 7.79151693e-02 3.10839623e-01 -1.02604175e+00
-3.01884562e-01 -2.41141453e-01 6.66426599e-01 -3.09369683e-01
5.30773938e-01 -7.47491956e-01 -1.25751883e-01 2.40806967e-01
-5.20925999e-01 1.90709293e-01 2.39383891e-01 9.09737349e-01
-3.63933086e-01 1.31871492e-01 7.28770792e-01 -1.75725982e-01
-1.05700111e+00 3.16501081e-01 -5.50983727e-01 1.03892289e-01
1.04812217e+00 -3.48281115e-01 -2.04869330e-01 -6.21508002e-01
-1.09191406e+00 -3.04281376e-02 5.43197036e-01 7.53776789e-01
6.68494582e-01 -1.53807306e+00 -5.44416368e-01 3.95277530e-01
3.86528492e-01 -3.36162984e-01 4.39311445e-01 6.17584288e-01
5.38784936e-02 4.41245109e-01 -4.64568168e-01 -6.84749842e-01
-1.42887378e+00 6.94118738e-01 2.73454964e-01 -1.07385769e-01
-9.15616035e-01 8.75708461e-01 5.97973466e-01 1.16654277e-01
3.39073986e-01 -3.17068964e-01 -3.86657238e-01 2.11277023e-01
6.92871153e-01 1.82540700e-01 -8.96433815e-02 -9.87623870e-01
-4.33215708e-01 3.99239451e-01 -2.70201594e-01 -2.28839502e-01
1.41515052e+00 -1.03921890e-01 5.55149198e-01 5.47193706e-01
1.25664222e+00 -4.07944992e-02 -1.91032171e+00 -2.96554565e-01
-9.39245448e-02 -8.54088306e-01 1.54931009e-01 -4.76469249e-01
-1.42776239e+00 5.74532390e-01 6.93915904e-01 -1.63525522e-01
1.15561175e+00 -2.54386198e-02 4.78371263e-01 3.67903948e-01
3.44029993e-01 -9.75135803e-01 7.35585809e-01 4.27360922e-01
1.01104200e+00 -1.27687109e+00 1.78894535e-01 -3.28835785e-01
-9.61702824e-01 1.01077139e+00 9.64361370e-01 -2.92122990e-01
4.95067835e-01 2.60372944e-02 3.61327827e-02 -2.38376617e-01
-9.79559124e-01 -2.47652948e-01 4.72190052e-01 8.48107934e-01
7.91487634e-01 -4.92981583e-01 1.38913348e-01 3.34120840e-01
1.98304981e-01 3.71801816e-02 5.45128405e-01 1.24974763e+00
-1.83630496e-01 -9.13875043e-01 -1.99508652e-01 4.67463255e-01
-3.64010572e-01 6.29657432e-02 -3.38206321e-01 8.17130864e-01
1.37925729e-01 7.09966183e-01 1.49947032e-01 -4.45433289e-01
3.03140759e-01 1.59882918e-01 6.78371727e-01 -8.64308357e-01
-1.79955304e-01 4.60580677e-01 2.59474888e-02 -1.02146184e+00
-8.77217293e-01 -9.34596479e-01 -1.29046881e+00 2.80797243e-01
-1.79849669e-01 4.70457077e-02 4.92426455e-02 8.87727857e-01
4.18808639e-01 7.95248389e-01 5.73177457e-01 -1.00977802e+00
-2.70162195e-01 -1.08196354e+00 -5.31016707e-01 7.08317578e-01
5.94616115e-01 -1.23977911e+00 -3.02595049e-01 5.26330650e-01] | [8.677473068237305, 0.7462456822395325] |
3f3c75be-21de-4e7d-9a60-e9c547b0d8c1 | an-extractive-abstractive-approach-for-multi | null | null | https://aclanthology.org/2022.sdp-1.25 | https://aclanthology.org/2022.sdp-1.25.pdf | An Extractive-Abstractive Approach for Multi-document Summarization of Scientific Articles for Literature Review | Research in the biomedical domain is con- stantly challenged by its large amount of ever- evolving textual information. Biomedical re- searchers are usually required to conduct a lit- erature review before any medical interven- tion to assess the effectiveness of the con- cerned research. However, the process is time- consuming, and therefore, automation to some extent would help reduce the accompanying information overload. Multi-document sum- marization of scientific articles for literature reviews is one approximation of such automa- tion. Here in this paper, we describe our pipelined approach for the aforementioned task. We design a BERT-based extractive method followed by a BigBird PEGASUS-based ab- stractive pipeline for generating literature re- view summaries from the abstracts of biomedi- cal trial reports as part of the Multi-document Summarization for Literature Review (MSLR) shared task1 in the Scholarly Document Pro- cessing (SDP) workshop 20222. Our proposed model achieves the best performance on the MSLR-Cochrane leaderboard3 on majority of the evaluation metrics. Human scrutiny of our automatically generated summaries indicates that our approach is promising to yield readable multi-article summaries for conducting such lit- erature reviews. | ['Tirthankar Ghosal', 'Trinita Roy', 'Kartik Shinde'] | null | null | null | null | sdp-coling-2022-10 | ['document-summarization'] | ['natural-language-processing'] | [ 0.33524302 0.16554677 -0.3953433 -0.05593806 -1.6207607 -0.7711609
0.62020934 0.9571796 -0.45398402 1.0779637 0.74065673 -0.6726878
-0.29708368 -0.29835343 -0.6783738 -0.3054492 0.39717847 0.43665144
-0.16740069 0.11379215 1.044635 0.4944595 -1.1034858 0.4625229
1.3576516 0.19559538 0.3585941 1.0179737 -0.24182716 0.57974064
-0.98553795 -0.38561994 -0.41273102 -0.5127107 -0.78952014 -0.25763
0.41179088 0.17672396 0.10116072 0.78916836 0.71417326 -0.2443676
0.7808479 -0.8371021 -0.3604811 0.99453545 -0.71819174 0.54257095
0.605878 -0.1885803 0.99747515 -0.8578153 1.0129875 1.257958
0.33671138 0.33207446 -0.8756436 -0.61220235 -0.0782644 0.03130426
-0.93020433 -0.6668087 0.57248515 -0.5971127 0.95779836 0.53228873
0.55535996 0.9219567 0.9041185 0.61280686 0.8525656 -0.47707325
0.28572214 -0.02785719 0.47399032 0.362719 0.9364693 -0.7451542
-0.70419025 -0.5993178 -0.03119506 -0.1756907 -0.27124208 0.5968914
-1.4249723 0.45605114 -0.08930745 0.2998866 -0.77273285 -0.07443424
0.8007558 0.01493938 0.6511173 0.80427045 -0.19882147 -0.22720262
-1.5541807 0.5323429 0.9179896 0.82447505 -0.01670499 -0.47029164
-0.49511057 0.7509572 0.31078404 0.3840836 0.5128531 -0.72556025
0.5767903 0.6248164 -0.04173621 -1.1141778 -0.41495213 -0.2985225
-0.9085833 -0.35225186 -0.11813465 -0.3553669 -0.7188556 1.0025172
0.1685399 -0.25984216 0.26382115 0.52174795 1.4296764 0.8148501
0.31207156 -0.5630328 1.8452296 -0.72519135 -1.1653161 0.03298406
0.6622472 -1.0976889 0.45277333 0.5974736 -1.2513305 0.01800613
-1.2719865 -0.30738455 -0.37376657 0.2591638 0.2145039 0.18740611
-0.92896974 0.39938328 -0.6688231 -0.5925956 0.6089761 0.04780081
-0.4284476 -0.14570467 -0.7941617 0.99583936 0.34949854 0.1995641
-0.6151781 -1.0984644 -0.5425762 -0.03336581 0.45311347 -0.7345716
1.1153759 0.09098765 -1.0151263 0.96329874 -0.36827493 -0.36099026
0.27391648 -0.14607127 -0.33813787 0.4704973 0.44849533 0.45726287
0.11864296 -0.681077 -0.50357854 -0.52795774 -0.3725619 0.16867587
-0.16090669 0.5730442 -0.33859867 -0.82059306 -0.35345247 -0.721312
-0.36836419 -0.38203874 -0.7911979 -0.3724438 0.5804396 -0.9964815
1.7285098 -1.5819031 -0.01255927 -0.26627314 0.27667361 0.2173076
-0.02416612 0.92038506 -0.05682554 0.6374378 -0.27188098 -0.08446942
-0.38098162 -0.18381414 -0.30072317 0.34535718 0.23281813 0.9260208
-1.1027578 -0.84496266 -0.27854156 0.14845538 -0.19464883 0.17795596
-0.3156105 0.18960106 -0.6475816 0.6496457 0.35421583 -0.24754402
0.05144372 -0.21596694 -0.59185576 0.65730995 -0.68881184 1.7208799
-0.28331506 0.67969185 0.05377835 -0.9694845 0.7895685 0.5433531
0.4700923 -0.14722954 0.03610056 0.40513828 0.00986968 -0.5855207
0.6445902 0.12199033 -0.10491142 0.4461841 -0.13732088 -0.36645848
0.5803171 0.7003306 1.2137144 0.00671009 0.92523134 -0.56141317
0.5673373 0.47149995 0.44903824 0.7156643 0.09174363 0.6929172
0.79035753 -0.13172998 -0.7906941 -0.24704702 -0.25370273 0.3450266
-0.6096085 -0.83947027 -0.71018356 -0.4917551 -0.21419847 0.9513055
-0.53889304 0.05132437 -0.32714337 -0.84241414 0.79770845 -0.08169215
0.11884075 -1.2358096 -1.0892203 0.38499618 -0.14450943 -0.89975005
-0.52767247 0.15644206 -0.6730939 -1.1661288 -1.185708 -0.48276046
0.52296966 -0.0261053 0.85991806 -0.03468302 -0.54732305 0.20966211
-0.24200101 -1.022988 -0.75693756 0.14531498 -0.25300494 -0.720396
0.49315953 -0.11899414 -0.6667277 -0.45131332 -1.021682 0.13274087
0.87263787 0.68133914 0.68221587 -0.3074295 1.1278275 -0.9526086
1.3235406 -0.61904544 -0.47672775 0.5999599 -0.7462329 0.03912256
0.45904112 -0.08890291 -0.8155902 -0.34149247 -0.07387426 0.18330073
0.02212917 1.2715461 0.06522737 0.53817594 0.64377445 0.0278785
-0.02668448 -0.61248165 0.28033072 1.005896 0.49461094 -0.3247492
0.08517646 -0.10855991 0.12861365 -0.93005013 -0.72099936 -0.532517
-0.16344641 -0.09926722 0.8845402 -0.72062916 -0.6312856 0.0679649
-1.6227795 0.3071469 0.06888256 0.43857586 0.08326251 0.5593971
-0.35278344 -0.72472686 -1.137055 -1.2555772 1.3044138 0.33557403
-0.6648324 -0.59388286 0.37483475 0.39177716 0.11953074 0.39153036
1.018146 -1.1091692 -0.06349526 -0.34284347 -0.15172674 -0.03981536
0.2562402 0.38265562 -0.569374 0.15393317 -0.1760742 0.04667815
0.750275 0.76853794 1.151342 -0.5872662 -0.52326703 -0.07753906
0.97755075 0.3051329 0.39861757 0.38777936 0.5041518 0.6925659
0.6424461 0.48826498 0.4906138 0.30560645 -0.13501112 0.10697481
-0.14034575 -0.05417804 0.17409258 1.1252273 0.1807426 -0.37344503
-1.2001251 0.833227 -1.6969889 -0.7863845 -0.37578455 1.858648
1.2487284 0.12809807 -0.03779375 -0.24433899 0.46415344 -0.01992891
-0.39272147 -0.86371076 -0.01188059 0.11059915 0.4968633 0.20708531
-0.59719974 0.57379055 5.984097 0.767899 -0.88826627 -0.15452364
0.6661865 -0.06005784 -0.37654385 -0.0813146 -0.95207995 0.5564494
1.5643283 -1.0458486 -0.29769173 0.5643396 0.6780567 -0.50791955
-1.0235085 0.64687246 0.1810887 -1.9714462 0.3041452 0.06073099
0.47125223 0.03156915 -0.4760137 -0.08330265 0.15425895 -0.9466513
0.33641377 0.690297 0.9278878 -0.78659445 0.82708824 0.28467724
-0.6445989 0.41416982 -0.43085805 0.42194793 0.19380344 1.1213969
-1.2092327 1.0336343 0.4331483 0.89427763 -0.67451906 1.3042431
-0.07181235 0.76620287 0.11151973 -0.49147943 0.06511669 -0.08756213
0.7991059 1.7678457 0.47498044 0.29951176 -0.05268655 0.66085935
-0.4177672 0.35125703 -0.51738375 -0.7473122 0.4644643 1.4047328
-0.8166347 -0.59873897 -0.07796787 0.4701823 0.02325215 0.0172096
-0.3067613 -0.8017421 -0.13237393 -0.13880719 -0.05620927 0.14196548
-0.54552776 -0.88497335 -0.10715229 -1.2078748 0.5325852 -0.8941355
-0.9210977 0.7178018 0.25266144 -0.9848455 -0.32435647 -0.12575023
-0.4652803 1.0411847 -1.309223 -0.89131474 0.21275327 -0.25146636
0.87456053 -0.21290438 0.7774809 -0.06664567 -0.78395647 0.23873019
0.05214244 -0.38240364 0.9688472 -1.260909 0.2876791 0.8104259
-0.3219526 1.1409035 1.0121706 -1.396539 -1.788968 -0.85259795
1.3041949 -0.43956274 0.8457365 -0.0499316 -1.0041357 0.17339316
0.5884838 -0.76455617 0.8952285 -0.10344742 0.19677237 0.13700637
-0.78859115 0.7227588 0.39209583 -0.09983314 -1.1620361 0.63500124
0.74873066 -0.3520524 -1.0434511 0.15634137 0.52986366 0.03343743
0.68376005 -0.52794945 1.1546525 -0.19512531 0.4743617 -1.2549127
0.05486728 -0.78101104 0.20017955 1.2359201 0.57667285 -0.19874889
0.29916158 0.57608074 -0.39958102 -0.8518917 -0.7332885 -0.226809
0.19283763 -0.11992886 0.15877639 0.717911 0.4649822 0.725376
0.03335353 -0.16356304 0.62609094 -0.12482695 0.594313 -1.2197067
0.2806564 -0.5935589 0.13152643 -0.40429142 -0.01577532 -1.0292182
0.14084114 -2.2789795 0.5661854 0.18615596 -0.10514394 0.41045892
-0.5956172 -0.48908395 -0.19671166 0.35240024 -0.683396 0.17381912
1.0361588 -0.17512353 -0.2652345 -0.2648924 -1.151878 0.29641888
0.42478776 -0.76897573 -0.20422748 0.0961234 0.41789716 0.28205505
0.03871773 -0.41771415 0.567388 -0.2582647 0.12569985 -1.2441761
-0.47251952 -0.23035912 0.18634407 0.52643806 -0.68759423 0.46872488
0.6139835 0.45044395 -0.17507221 -0.45057544 0.19373198 -0.3123654
0.05665415 -0.13871807 -0.76204014 0.04000722 0.5655198 0.10339154
-0.60727787 -0.07322245 0.04830215 0.5921155 0.09343296 0.3193532
0.52381796 -0.5456875 -1.3489649 -0.5523038 0.13333496 0.01601238
0.26315555 0.9650284 -0.74158084 0.9689142 -0.03695028 -0.22046359
-1.4622253 0.44727895 -0.33653647 -0.77382255 -0.81659156 0.47053102
-0.0226901 -0.11176991 0.20462626 -0.50681204 -0.57995164 0.26588127
1.0056893 0.5945744 0.5058666 -0.29745394 -0.59658706 0.10536785
-0.49229544 -0.36355147 1.7422734 0.099898 -0.51461184 0.5368022
1.0537426 0.17817496 -0.1705219 0.2480613 0.49716836 0.28565732
0.24783465 -1.3663033 -0.3565744 0.6288242 -0.1994869 -0.05218513
0.8536705 -0.09266564 0.42067036 0.54115593 -0.28902146 -0.96204144
-0.38450688 0.06359091 1.4270089 -1.0781219 0.90105253 -0.1112029
-0.68391293 1.3439872 0.08901853 0.15543714 0.41718748 0.20328265
-0.16412714 -0.79348123 -1.0500531 0.59816724 0.8377455 0.09456962
0.7985926 -0.20448002 -1.2636026 0.744821 -0.04012833 0.2928797
0.88015604 1.1137191 -0.26374996 -0.9472511 -0.6568946 0.7343918
-1.1604486 -0.35590175 -0.65111697 0.5179989 -0.64033383 1.1183372
-0.4199428 0.23865752 0.33058676 -0.02746604 0.08103298 -0.85606366
-0.8184249 0.45339882 0.4192935 -0.11266825 -0.5712879 -0.7035093
-1.2148527 0.01184611 -0.1778997 0.5237583 1.165767 1.0026574
0.8670127 0.90456295 0.17334968 -0.47420666 -0.26167428 -1.0604023
-0.13084652 -0.19212645 0.27275333 -0.20654768 -0.21514277 0.28601182] | [12.289118766784668, 9.601730346679688] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.