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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
77d13292-6b24-4cd8-bbee-9ba14f29d46a | a-general-framework-for-density-based-time | 1612.00637 | null | http://arxiv.org/abs/1612.00637v1 | http://arxiv.org/pdf/1612.00637v1.pdf | A General Framework for Density Based Time Series Clustering Exploiting a Novel Admissible Pruning Strategy | Time Series Clustering is an important subroutine in many higher-level data
mining analyses, including data editing for classifiers, summarization, and
outlier detection. It is well known that for similarity search the superiority
of Dynamic Time Warping (DTW) over Euclidean distance gradually diminishes as
we consider ever larger datasets. However, as we shall show, the same is not
true for clustering. Clustering time series under DTW remains a computationally
expensive operation. In this work, we address this issue in two ways. We
propose a novel pruning strategy that exploits both the upper and lower bounds
to prune off a very large fraction of the expensive distance calculations. This
pruning strategy is admissible and gives us provably identical results to the
brute force algorithm, but is at least an order of magnitude faster. For
datasets where even this level of speedup is inadequate, we show that we can
use a simple heuristic to order the unavoidable calculations in a
most-useful-first ordering, thus casting the clustering into an anytime
framework. We demonstrate the utility of our ideas with both single and
multidimensional case studies in the domains of astronomy, speech physiology,
medicine and entomology. In addition, we show the generality of our clustering
framework to other domains by efficiently obtaining semantically significant
clusters in protein sequences using the Edit Distance, the discrete data
analogue of DTW. | ['Nurjahan Begum', 'Eamonn Keogh', 'Jun Wang', 'Hoang Anh Dau', 'Liudmila Ulanova'] | 2016-12-02 | null | null | null | null | ['time-series-clustering'] | ['time-series'] | [ 4.78610992e-01 -2.86007434e-01 1.65889651e-01 -2.00482294e-01
-4.26421314e-01 -9.80984390e-01 4.27229404e-01 6.03080630e-01
-6.70079827e-01 5.61552644e-01 -2.24154025e-01 -5.94926834e-01
-6.86493933e-01 -6.76031590e-01 -3.85743707e-01 -9.53179300e-01
-7.66837478e-01 5.33366621e-01 4.47529763e-01 -2.81285346e-01
5.25933206e-01 9.53028858e-01 -1.69111836e+00 -1.10966511e-01
8.51820111e-01 6.19073629e-01 -4.90684360e-02 7.45339751e-01
-1.01939082e-01 1.14023037e-01 -7.19294131e-01 -1.71043262e-01
5.81406176e-01 -4.42047298e-01 -7.09108472e-01 1.91875532e-01
-6.87641799e-02 2.08997101e-01 1.47717968e-01 8.85564804e-01
4.42888349e-01 3.85345936e-01 4.46958542e-01 -1.51915431e+00
-1.16458215e-01 2.92192191e-01 -7.46581078e-01 1.99790984e-01
4.17033315e-01 -2.36366153e-01 7.09072709e-01 -3.06885004e-01
7.57716358e-01 8.77364814e-01 8.90145302e-01 1.10749602e-01
-1.47363591e+00 -3.04916531e-01 -5.70210814e-03 2.07594082e-01
-1.31730926e+00 -9.38884616e-02 5.32354712e-01 -3.36310357e-01
1.00717449e+00 7.18923986e-01 6.70113981e-01 4.04621869e-01
1.14850596e-01 4.07785296e-01 9.91923332e-01 -3.66857976e-01
5.01253247e-01 -3.04388344e-01 1.99551538e-01 3.88681114e-01
3.73540133e-01 -3.52953412e-02 -7.04400763e-02 -6.87061250e-01
4.07171905e-01 2.63304472e-01 -3.46328348e-01 -5.97902894e-01
-1.28463042e+00 8.79363060e-01 -8.54078755e-02 5.39882898e-01
-6.06710650e-02 -6.85060769e-02 5.47489166e-01 7.88647354e-01
5.39854348e-01 7.12425828e-01 -5.53384840e-01 -3.68770599e-01
-1.07217407e+00 3.00705999e-01 1.08115339e+00 8.35880399e-01
4.03585583e-01 -3.04179639e-01 6.50046289e-01 7.07238495e-01
-4.62475032e-01 3.03335655e-02 6.62053704e-01 -1.00777280e+00
-8.07455331e-02 4.55660820e-01 3.84023823e-02 -1.13540435e+00
-6.03181601e-01 -3.22504416e-02 -7.33299255e-01 3.01760167e-01
7.00856984e-01 1.33212239e-01 -6.57223582e-01 1.79072762e+00
6.34106278e-01 1.72288284e-01 -1.07081659e-01 4.29250002e-01
-1.02133118e-01 5.74517369e-01 -2.43797332e-01 -9.24311340e-01
1.08316767e+00 -3.13781440e-01 -5.60945868e-01 5.18012643e-01
7.48569489e-01 -8.22655022e-01 7.74149716e-01 7.24570811e-01
-9.38190281e-01 -1.69673964e-01 -1.10365784e+00 4.47582686e-03
-5.80758035e-01 -5.49790621e-01 8.62258971e-01 5.80066442e-01
-1.04086292e+00 1.21688831e+00 -1.13057137e+00 -9.33060527e-01
-1.15406970e-02 5.67109823e-01 -3.25371832e-01 3.26505780e-01
-8.23472619e-01 7.56969213e-01 3.38332862e-01 -3.61219049e-01
-6.47036135e-02 -6.18605018e-01 -5.24112284e-01 -1.91772372e-01
3.43959481e-01 -2.89796114e-01 1.00744832e+00 -6.99948788e-01
-1.16296518e+00 7.58469582e-01 -2.84951985e-01 -6.48825228e-01
6.09153807e-01 2.97135353e-01 -2.53300548e-01 1.95787176e-01
-1.61695436e-01 1.65052772e-01 5.97012520e-01 -8.49270284e-01
-5.05465567e-01 -6.26131296e-01 -1.15129545e-01 1.05124013e-02
-3.24136198e-01 8.02995190e-02 -1.67125449e-01 -8.05395007e-01
3.35614294e-01 -1.02317214e+00 -5.52179456e-01 7.43194222e-02
9.75005608e-03 -3.54164422e-01 8.51413310e-01 -3.86437267e-01
1.39370584e+00 -2.19010401e+00 3.46680969e-01 6.15761757e-01
3.34021837e-01 5.05964607e-02 1.42046034e-01 7.23402143e-01
-4.40486610e-01 5.98300423e-04 -7.80180454e-01 1.29837412e-02
-1.36153921e-01 4.85873073e-01 -3.94090593e-01 8.36771548e-01
-6.90906271e-02 4.59476799e-01 -1.00999868e+00 -4.35341269e-01
1.38249785e-01 1.30296439e-01 -6.24976218e-01 -3.10180545e-01
4.74586822e-02 1.35897040e-01 -1.63043588e-01 3.78165364e-01
5.85629463e-01 1.18347565e-02 4.06276703e-01 9.32896212e-02
-4.66422915e-01 -4.07331772e-02 -1.35179913e+00 1.61131120e+00
-7.91452676e-02 5.60606122e-01 -7.78010637e-02 -1.38948834e+00
7.83143222e-01 1.64218977e-01 1.19739759e+00 -2.29566857e-01
2.50178594e-02 3.57670546e-01 1.27210110e-01 -3.84628326e-01
4.21263516e-01 -1.86361074e-01 -9.91818607e-02 6.28739476e-01
-3.51594687e-01 -1.78371191e-01 5.44145525e-01 9.66885462e-02
1.46452892e+00 -6.93393722e-02 6.15394413e-01 -4.83165830e-01
3.92998546e-01 3.78554553e-01 4.02137786e-01 4.31279331e-01
-1.76445454e-01 3.59247565e-01 6.44189537e-01 -4.11347210e-01
-1.19350135e+00 -1.01437032e+00 -3.44331026e-01 9.26463246e-01
-4.47291993e-02 -5.93651116e-01 -7.89960146e-01 -2.83517450e-01
2.41714954e-01 4.34220195e-01 -5.25082946e-01 -1.72884107e-01
-6.38851047e-01 -1.10186410e+00 5.29274881e-01 2.69395739e-01
7.55428672e-02 -7.04360783e-01 -9.86391664e-01 3.94221693e-01
1.79547250e-01 -6.99693620e-01 -3.50063890e-01 5.22845447e-01
-1.29063046e+00 -1.10069311e+00 -4.84429061e-01 -5.61513841e-01
6.12213910e-01 4.00747448e-01 6.66894078e-01 4.89686728e-02
-6.85754061e-01 2.75789350e-01 -3.76510322e-01 -3.85381430e-01
-2.74358481e-01 -1.45967245e-01 3.06084931e-01 -2.98829019e-01
6.63937092e-01 -9.91042674e-01 -3.17737460e-01 3.46769065e-01
-1.18799448e+00 -4.19712871e-01 9.24110413e-02 6.79157674e-01
6.76673591e-01 6.35629594e-01 3.74656200e-01 -7.01778293e-01
7.82613277e-01 -4.41557050e-01 -7.91104555e-01 1.60980031e-01
-4.75266337e-01 7.81021565e-02 9.36137140e-01 -6.24091804e-01
-3.13809276e-01 4.54452962e-01 3.21770422e-02 -3.98079753e-01
-2.65023187e-02 4.16542381e-01 3.06045990e-02 -2.43228078e-01
5.59866130e-01 3.01625937e-01 1.75751388e-01 -5.48197269e-01
4.00067449e-01 6.69631720e-01 6.90511227e-01 -6.17830157e-01
7.90755272e-01 8.13997805e-01 3.34650815e-01 -1.16095781e+00
-9.88004655e-02 -7.41352677e-01 -7.87306666e-01 7.49003738e-02
6.08001471e-01 -1.43543512e-01 -1.08753324e+00 6.26813173e-02
-9.79162931e-01 1.71823874e-02 -4.98175085e-01 3.12247217e-01
-8.56950939e-01 8.76210332e-01 -2.02479258e-01 -7.75195420e-01
6.90557361e-02 -9.32018340e-01 7.24462390e-01 -3.40404600e-01
-5.45195043e-01 -8.86889696e-01 3.18295360e-01 -1.99367419e-01
9.34051163e-03 7.31476605e-01 1.07273209e+00 -1.10792589e+00
-6.14711866e-02 -2.28113338e-01 2.23063618e-01 -1.06203467e-01
1.96445435e-01 3.02670926e-01 -5.83992779e-01 -3.40829611e-01
5.09059131e-01 3.48688334e-01 6.92661881e-01 1.79059073e-01
1.40355074e+00 -1.64227188e-01 -3.88085872e-01 6.32581890e-01
1.37543285e+00 5.03823578e-01 4.68341768e-01 3.22987437e-01
3.22644144e-01 9.88692045e-01 7.28601754e-01 6.03148878e-01
2.15760283e-02 7.53101885e-01 1.00551872e-02 9.99032054e-03
4.90223229e-01 3.00449401e-01 2.14496791e-01 9.52255964e-01
-3.20524305e-01 -7.34773204e-02 -9.80275035e-01 6.74467087e-01
-1.96933472e+00 -1.37254941e+00 -2.52671629e-01 2.64206457e+00
9.80812013e-01 -2.19021309e-02 5.57063341e-01 7.33699799e-01
7.21000314e-01 -2.26361349e-01 -5.37036180e-01 -8.00662458e-01
1.29096448e-01 4.71708059e-01 7.47762203e-01 5.27968705e-01
-1.05166090e+00 4.44189399e-01 6.57272863e+00 8.23294997e-01
-8.76812220e-01 -1.46294057e-01 1.33748874e-01 -3.01810056e-01
-8.16688836e-02 1.55981071e-02 -1.40745327e-01 4.86775070e-01
1.02122700e+00 -7.26783574e-01 5.73254585e-01 8.74211073e-01
3.54625165e-01 -2.02185094e-01 -1.21905470e+00 9.29018378e-01
-1.06998518e-01 -8.74824107e-01 -2.58279860e-01 2.96308815e-01
4.01429117e-01 -2.59190202e-01 -2.04996511e-01 -2.68271327e-01
2.92459220e-01 -7.50364721e-01 1.72816336e-01 -3.94671187e-02
4.31928396e-01 -1.06376171e+00 2.12057114e-01 2.69190729e-01
-1.31784701e+00 -1.59022603e-02 -4.18927372e-01 -5.80914356e-02
9.05552804e-02 7.68293321e-01 -5.52127421e-01 7.40626693e-01
4.28534418e-01 5.69821417e-01 -3.70934516e-01 1.24291432e+00
4.29142475e-01 1.02908358e-01 -8.64323199e-01 3.32236625e-02
1.65211692e-01 -4.59867626e-01 6.64255679e-01 1.15898371e+00
5.01067579e-01 4.25653607e-01 2.47805733e-02 2.67362058e-01
3.20618272e-01 1.07296228e-01 -9.01951313e-01 -1.11918181e-01
4.84026521e-01 9.90214229e-01 -1.21694052e+00 -3.39927435e-01
-3.35017025e-01 1.07811582e+00 -2.15072464e-02 3.16722915e-02
-7.80885816e-01 -9.97010827e-01 8.14899385e-01 -2.55444646e-03
3.69282961e-01 -7.23151922e-01 -3.43735516e-01 -9.26380992e-01
3.84619683e-02 -9.07531321e-01 7.35018373e-01 -2.26885840e-01
-1.18677104e+00 4.38382328e-01 3.83752406e-01 -1.38629937e+00
-4.54517335e-01 -6.35958374e-01 -3.41918677e-01 4.38711017e-01
-9.72598016e-01 -2.00793579e-01 5.79438396e-02 6.77576184e-01
4.53491062e-01 3.36803675e-01 7.86341250e-01 2.70273298e-01
-3.70883942e-01 3.96244556e-01 5.61596870e-01 -3.97507995e-01
6.04852319e-01 -1.38512945e+00 4.31863487e-01 8.87675941e-01
9.04091522e-02 9.14586246e-01 1.11105549e+00 -6.50897622e-01
-1.44689929e+00 -7.92827964e-01 1.04209685e+00 -2.93517649e-01
9.80270684e-01 -2.82002509e-01 -1.07522118e+00 4.10300732e-01
-7.66550228e-02 -4.62802440e-01 8.34902048e-01 -5.71423508e-02
-2.21008688e-01 3.59813720e-02 -1.24481273e+00 7.04737723e-01
1.28158176e+00 -2.49402434e-01 -7.69409180e-01 6.30014241e-01
5.73913038e-01 -4.42443006e-02 -1.19849634e+00 2.62312263e-01
4.78305370e-01 -8.43846142e-01 1.03161049e+00 -5.99382818e-01
-1.74454171e-02 -5.87354481e-01 -1.71824872e-01 -1.05359626e+00
-4.90066223e-02 -1.11772823e+00 2.78581884e-02 9.00939584e-01
7.01754689e-02 -7.13274658e-01 6.19245708e-01 4.48642373e-01
2.42619645e-02 -7.35870957e-01 -1.01989973e+00 -1.26187527e+00
1.99229922e-02 -4.87893879e-01 5.63453794e-01 1.35160124e+00
5.43912232e-01 -1.08365580e-01 -4.18057404e-02 -4.56886478e-02
7.43767083e-01 3.96368891e-01 6.08779490e-01 -1.48296809e+00
-4.21947300e-01 -5.45420408e-01 -7.85235643e-01 -8.79103422e-01
-1.63235173e-01 -7.60336101e-01 -1.10850289e-01 -1.00281346e+00
-1.33849934e-01 -2.37046555e-01 -1.48604810e-01 4.12007302e-01
1.28314793e-01 1.37992814e-01 5.72508993e-03 3.83271277e-01
-3.36963564e-01 2.09080074e-02 8.44884217e-01 3.09106082e-01
-4.34600383e-01 2.42750440e-02 -3.48945349e-01 8.54404449e-01
8.29968691e-01 -5.87813377e-01 -4.59208906e-01 -4.45753932e-02
3.86483707e-02 -1.28222138e-01 1.41448185e-01 -8.81611049e-01
4.61662620e-01 -2.62567848e-01 3.64697501e-02 -6.89478278e-01
1.41280904e-01 -1.05343115e+00 3.15917999e-01 7.27872491e-01
-2.48051375e-01 5.92901349e-01 2.34639436e-01 6.76879227e-01
5.94920702e-02 -2.16095909e-01 8.40532243e-01 -8.83309767e-02
-5.47729909e-01 1.46455243e-01 -4.37372029e-01 -2.13638380e-01
1.38148296e+00 -5.69958746e-01 -7.55695850e-02 -1.91403240e-01
-8.96767259e-01 2.22485021e-01 7.92093217e-01 9.66819748e-02
3.72278035e-01 -1.00938988e+00 -3.23288471e-01 1.87724650e-01
-3.44199166e-02 -2.98768371e-01 -1.48166195e-01 1.18759370e+00
-6.76943958e-01 2.40760416e-01 -2.14836955e-01 -5.98614454e-01
-1.68682766e+00 1.07872176e+00 -2.35016979e-02 -1.05112828e-02
-6.23318434e-01 3.58704776e-01 -2.04710647e-01 -1.75766632e-01
2.44519815e-01 -4.14128602e-01 2.25513354e-01 2.74011195e-01
4.97201890e-01 7.21218705e-01 2.37708062e-01 -2.72192657e-01
-4.81941193e-01 7.94139445e-01 -4.04619537e-02 -1.14534110e-01
1.56653953e+00 -8.75793844e-02 -4.88218457e-01 5.04952073e-01
1.31101215e+00 -2.40625199e-02 -8.29217196e-01 1.65368289e-01
4.33598787e-01 -5.11586547e-01 -5.08722782e-01 -1.71569020e-01
-5.12881458e-01 6.82243526e-01 3.47106814e-01 9.27344203e-01
1.60433447e+00 -1.41151518e-01 7.28591084e-01 7.16299713e-01
4.05110627e-01 -1.16441965e+00 -3.80858541e-01 1.60921782e-01
6.40363455e-01 -6.82922602e-01 3.28086227e-01 -4.11609381e-01
-3.00582111e-01 1.27428436e+00 7.32710259e-03 -2.24548116e-01
5.17692566e-01 4.58430618e-01 -3.60626668e-01 -1.79331496e-01
-7.32245088e-01 -3.18744898e-01 -2.58732378e-01 5.16508162e-01
3.00861508e-01 -1.55156702e-01 -9.00364339e-01 4.05275077e-02
-3.71578634e-01 -1.51489943e-01 5.54271579e-01 1.31397533e+00
-6.79435015e-01 -1.41096938e+00 -4.30829614e-01 3.66771460e-01
-3.77958417e-01 8.39802027e-02 -6.82399213e-01 1.03040814e+00
2.85478812e-02 8.17432463e-01 1.19948640e-01 -3.29581946e-01
3.63937527e-01 2.34805658e-01 5.58786631e-01 -3.21365654e-01
-5.47740519e-01 7.51600713e-02 -1.00471377e-01 -6.99081838e-01
-6.17406547e-01 -8.29216301e-01 -1.35651851e+00 -8.11546504e-01
-2.81575978e-01 3.09087813e-01 7.65232325e-01 7.37058401e-01
2.69674331e-01 1.73538268e-01 7.79716790e-01 -5.88712454e-01
-5.40568352e-01 -3.64890486e-01 -5.52591145e-01 4.78089511e-01
2.70194799e-01 -5.38233340e-01 -5.79324126e-01 1.62359625e-01] | [7.3002028465271, 3.506544351577759] |
f480ba65-6763-4e9b-8324-e3b7ccbc18cb | search-to-pass-messages-for-temporal | 2210.16740 | null | https://arxiv.org/abs/2210.16740v1 | https://arxiv.org/pdf/2210.16740v1.pdf | Search to Pass Messages for Temporal Knowledge Graph Completion | Completing missing facts is a fundamental task for temporal knowledge graphs (TKGs). Recently, graph neural network (GNN) based methods, which can simultaneously explore topological and temporal information, have become the state-of-the-art (SOTA) to complete TKGs. However, these studies are based on hand-designed architectures and fail to explore the diverse topological and temporal properties of TKG. To address this issue, we propose to use neural architecture search (NAS) to design data-specific message passing architecture for TKG completion. In particular, we develop a generalized framework to explore topological and temporal information in TKGs. Based on this framework, we design an expressive search space to fully capture various properties of different TKGs. Meanwhile, we adopt a search algorithm, which trains a supernet structure by sampling single path for efficient search with less cost. We further conduct extensive experiments on three benchmark datasets. The results show that the searched architectures by our method achieve the SOTA performances. Besides, the searched models can also implicitly reveal diverse properties in different TKGs. Our code is released in https://github.com/striderdu/SPA. | ['Xuelong Li', 'Quanming Yao', 'Haotong Du', 'Zhen Wang'] | 2022-10-30 | null | null | null | null | ['temporal-knowledge-graph-completion'] | ['knowledge-base'] | [-2.96133280e-01 -2.20917046e-01 -5.91556191e-01 7.82678183e-03
-3.20383161e-01 -4.00874406e-01 2.83683717e-01 -1.64730489e-01
1.05581596e-01 6.93872452e-01 1.58311486e-01 -5.85216880e-01
-6.51959062e-01 -1.14913940e+00 -6.48240745e-01 -6.27183914e-01
-2.04430223e-01 5.51780760e-01 4.21066850e-01 -3.44571888e-01
2.92774349e-01 3.25904191e-01 -1.23522401e+00 1.55844003e-01
1.05311239e+00 9.17768300e-01 3.34314495e-01 2.99254209e-01
-3.71325046e-01 7.28129268e-01 -2.23618224e-01 -1.30322531e-01
6.59008324e-02 -3.27450633e-01 -1.09125400e+00 -2.95475692e-01
1.36060089e-01 -2.33746886e-01 -1.15550840e+00 9.28890288e-01
2.96789736e-01 1.45776033e-01 1.18486047e-01 -1.30029655e+00
-9.89606440e-01 9.03964758e-01 -3.89672875e-01 3.72525036e-01
2.20747560e-01 2.96086043e-01 1.06661344e+00 -5.14209449e-01
6.31064773e-01 9.70384598e-01 4.98286486e-01 4.38631088e-01
-7.65631855e-01 -8.00025582e-01 6.09174371e-01 8.11801493e-01
-1.42123997e+00 -2.88980961e-01 1.16964769e+00 7.71712288e-02
9.30980325e-01 3.48957717e-01 1.00584698e+00 1.05313146e+00
1.03420272e-01 9.21946287e-01 7.59188473e-01 -8.64061639e-02
1.46072939e-01 -6.37617290e-01 1.07093818e-01 8.69215250e-01
2.66859263e-01 2.26538882e-01 -5.90727210e-01 8.89279172e-02
9.92169559e-01 1.40587185e-02 -4.27700847e-01 -3.70978206e-01
-1.42937768e+00 6.02944374e-01 6.68220043e-01 3.84540647e-01
-1.95169210e-01 6.67392075e-01 5.76447189e-01 4.11766708e-01
2.59636164e-01 2.72388607e-01 -5.46555459e-01 -5.70653714e-02
-6.43172562e-01 1.05456114e-01 7.56390870e-01 1.20108509e+00
6.24136925e-01 -1.34343805e-03 -2.40995422e-01 4.36475039e-01
1.42176792e-01 2.19977006e-01 3.05731148e-01 -7.21162558e-01
6.89285159e-01 1.05665481e+00 -2.35941395e-01 -1.47721004e+00
-6.20126545e-01 -5.85829616e-01 -1.21523595e+00 -6.67663574e-01
1.30436108e-01 2.97610015e-01 -1.02953649e+00 1.80365086e+00
5.31711102e-01 6.07267976e-01 -1.02307558e-01 7.89656281e-01
9.84180093e-01 9.28908408e-01 -1.83209285e-01 -2.72380799e-01
1.18376553e+00 -1.11592746e+00 -6.91953182e-01 -3.83216217e-02
1.01959383e+00 -1.28962457e-01 1.10464180e+00 3.56212318e-01
-1.02917767e+00 -3.57562661e-01 -1.00058770e+00 -1.31903484e-01
-3.99951130e-01 6.18355498e-02 1.11572707e+00 3.35627913e-01
-1.08987987e+00 5.83033204e-01 -1.05320930e+00 -2.83901513e-01
4.15229052e-01 4.03538555e-01 -7.82558694e-02 -1.69126511e-01
-1.64001405e+00 5.44464231e-01 9.65763509e-01 5.73926806e-01
-1.03739762e+00 -5.07225156e-01 -5.79260588e-01 7.24884346e-02
1.04550683e+00 -9.11785603e-01 9.66395617e-01 -1.90280139e-01
-1.27898395e+00 4.30666625e-01 5.90315834e-02 -2.18904346e-01
2.32637927e-01 2.70189375e-01 -7.27644086e-01 1.62415162e-01
-5.58634363e-02 2.62346059e-01 1.95596829e-01 -8.13052475e-01
-4.27471757e-01 -3.34578216e-01 2.63107598e-01 1.24811672e-01
-5.38990974e-01 -3.69416893e-01 -1.01294923e+00 -6.50152266e-01
3.63714606e-01 -7.77999997e-01 -3.08262914e-01 -3.18782657e-01
-5.44525206e-01 -6.26130641e-01 6.19371414e-01 -6.50185704e-01
1.95182979e+00 -1.75281179e+00 3.95502418e-01 2.66620904e-01
6.14380658e-01 1.38938174e-01 -2.64758408e-01 7.88314581e-01
2.99353153e-01 1.65372133e-01 -1.22722365e-01 -1.12511422e-02
1.54885678e-02 3.82086813e-01 -2.61266321e-01 2.33701438e-01
-3.59744996e-01 1.31647444e+00 -1.02533865e+00 -6.71119332e-01
1.28681015e-03 -1.61752328e-01 -3.70307356e-01 -3.25308084e-01
-6.66870773e-01 4.27045435e-01 -1.08440340e+00 7.18940020e-01
7.26658821e-01 -7.72826612e-01 5.10602295e-01 -3.08999956e-01
-2.20605824e-02 3.16239536e-01 -1.03629899e+00 2.25195265e+00
-2.97310650e-01 1.68696105e-01 -2.26739496e-01 -1.22072911e+00
1.01596987e+00 4.87733632e-02 3.55710268e-01 -9.21583474e-01
1.12334616e-03 4.72161144e-01 -1.80908978e-01 -6.26709461e-01
5.46323359e-01 3.45378786e-01 -9.62056369e-02 4.65385050e-01
-2.40490630e-01 3.27229500e-01 2.78316319e-01 4.18756962e-01
1.27071762e+00 1.03877589e-01 -2.22371090e-02 -2.13283211e-01
4.99983430e-01 1.22264042e-01 8.12367082e-01 6.59902275e-01
-1.17004283e-01 1.79907024e-01 5.72228432e-01 -7.95217097e-01
-9.41115201e-01 -7.58945346e-01 3.96195799e-01 8.84533942e-01
6.54296100e-01 -6.08582079e-01 -3.17520916e-01 -6.66796744e-01
-1.90588489e-01 5.26994228e-01 -5.69152296e-01 -3.92211735e-01
-9.67198074e-01 -5.21951318e-01 6.77205205e-01 5.72052002e-01
7.25164413e-01 -1.25395906e+00 -1.90720648e-01 3.64700496e-01
-4.93423074e-01 -9.96732116e-01 -6.86459661e-01 -2.27955952e-01
-1.04242527e+00 -1.26321077e+00 -4.65985358e-01 -1.07579589e+00
4.92540449e-01 4.07263547e-01 1.08721876e+00 5.80350757e-01
2.40363907e-02 2.10158795e-01 -4.61223006e-01 2.07242087e-01
1.23756908e-01 7.07404494e-01 -4.50298563e-02 -2.96570390e-01
1.58754870e-01 -1.04775977e+00 -7.65477657e-01 5.67821741e-01
-9.88463163e-01 5.49945772e-01 6.00232184e-01 6.50508344e-01
7.20801473e-01 4.52871442e-01 7.09544837e-01 -6.61204636e-01
7.91533589e-01 -6.58244550e-01 -7.59723842e-01 7.78463304e-01
-7.94390917e-01 2.94646263e-01 7.79530346e-01 -4.47344989e-01
-8.96842420e-01 -3.81295085e-01 2.26748139e-01 -7.50196576e-01
1.79665476e-01 1.14131534e+00 -1.59168959e-01 -1.91165984e-01
3.09045672e-01 8.79785836e-01 -1.89259022e-01 -4.96634483e-01
3.48225236e-01 1.88852742e-01 5.82630634e-01 -9.46108997e-01
6.03325188e-01 4.60712552e-01 2.11827606e-01 -1.92508131e-01
-6.36669874e-01 -3.06137264e-01 -2.51944602e-01 -8.75291228e-02
3.14097166e-01 -5.74642777e-01 -9.76288497e-01 5.51884413e-01
-1.11735702e+00 -5.61846972e-01 2.79685199e-01 3.29494953e-01
-4.09900367e-01 7.04740524e-01 -7.17847407e-01 -4.94416445e-01
-5.69288492e-01 -9.46278930e-01 6.48584127e-01 2.62975067e-01
3.40768337e-01 -1.03995359e+00 1.62451670e-01 2.46191368e-01
4.19931889e-01 2.24117562e-01 9.00868654e-01 -3.38122368e-01
-1.24711049e+00 1.71624854e-01 -4.22423452e-01 -4.29112524e-01
-2.93058599e-03 -1.22186616e-01 -2.80160159e-01 -2.68312246e-01
-2.05742896e-01 -2.87206382e-01 1.09474826e+00 4.21734989e-01
1.60619569e+00 -5.57053268e-01 -7.69181490e-01 9.12085354e-01
1.36103451e+00 2.52773821e-01 6.95142448e-01 4.63547528e-01
8.57020557e-01 4.02943909e-01 6.19930625e-01 3.23530346e-01
9.52571571e-01 5.97028732e-01 6.55170560e-01 2.96968251e-01
-3.92133705e-02 -6.95333898e-01 9.69319604e-03 1.22021341e+00
-2.22412676e-01 -3.71131897e-01 -1.14872217e+00 7.49173820e-01
-2.32715297e+00 -7.91185439e-01 -2.76829422e-01 1.75785625e+00
8.61843526e-01 1.46542877e-01 -6.83635026e-02 6.26147306e-03
8.37781847e-01 3.73312682e-01 -7.51320660e-01 1.37171075e-01
-6.20234199e-02 -6.82855844e-02 5.32194912e-01 1.35020420e-01
-6.71319067e-01 1.19547641e+00 5.01931477e+00 1.34931469e+00
-8.44845295e-01 -1.18908323e-01 3.86457354e-01 -6.11719973e-02
-8.10738444e-01 3.14555407e-01 -5.39975703e-01 5.85795164e-01
6.79581046e-01 -3.42347980e-01 8.84346306e-01 5.93873441e-01
-4.87215584e-03 2.75360197e-01 -7.69109368e-01 1.15899038e+00
-3.15613121e-01 -1.86551547e+00 2.35619232e-01 3.45431827e-02
7.91093647e-01 -1.60372972e-01 -3.84563319e-02 5.04709601e-01
3.12003791e-01 -9.91364062e-01 3.95427614e-01 6.69331908e-01
9.81793404e-01 -7.79262424e-01 4.90415812e-01 4.52566326e-01
-1.82626748e+00 -2.03503907e-01 -4.25686628e-01 -4.00116332e-02
2.49181926e-01 4.63728845e-01 -4.37260419e-01 1.29461956e+00
7.18378127e-01 8.28833938e-01 -5.41813672e-01 1.14619565e+00
-2.54462719e-01 5.47897875e-01 -4.21886414e-01 -4.33281660e-01
4.57814902e-01 -2.20077425e-01 4.74753529e-01 6.19456053e-01
3.46981525e-01 3.61980200e-01 2.97593564e-01 8.80236387e-01
-1.03707843e-01 -9.38715562e-02 -5.59954107e-01 -3.65878552e-01
8.97087336e-01 1.07784247e+00 -8.20894122e-01 -1.59635141e-01
-9.32704061e-02 5.77111661e-01 6.23742342e-01 4.36439633e-01
-8.72614861e-01 -4.42559302e-01 1.86733514e-01 -1.25873238e-01
1.84780627e-01 -4.07218456e-01 -2.41869882e-01 -1.27096510e+00
3.04226786e-01 -7.30117023e-01 9.01500225e-01 -7.96676338e-01
-1.13446665e+00 5.87890267e-01 1.95896655e-01 -1.06324637e+00
3.19851756e-01 -2.16681093e-01 -7.21543074e-01 4.75267231e-01
-1.39935791e+00 -1.42168617e+00 -5.77963412e-01 8.19461584e-01
2.39306092e-01 -1.53211383e-02 4.07587916e-01 3.54495436e-01
-7.70504892e-01 4.67551470e-01 -5.25105111e-02 5.83174117e-02
8.89886320e-02 -8.39914620e-01 6.76411569e-01 7.56550312e-01
-2.14074533e-02 6.83787584e-01 2.95659035e-01 -8.84817541e-01
-1.89307523e+00 -9.65500116e-01 7.14010179e-01 -1.48882374e-01
7.71877050e-01 -1.05146341e-01 -1.00102460e+00 8.45141292e-01
-8.62560351e-04 -1.85319465e-02 8.44497979e-02 3.67273062e-01
-1.84363037e-01 -3.88891578e-01 -5.48399746e-01 7.85098553e-01
1.84165859e+00 -4.12265509e-01 -3.95847708e-01 3.07259709e-01
1.35240996e+00 -6.34156108e-01 -7.52409995e-01 6.72868848e-01
2.53543079e-01 -7.46528924e-01 8.36397290e-01 -6.82549000e-01
2.86443383e-01 -6.08448446e-01 1.91411838e-01 -1.10210896e+00
-4.99639124e-01 -7.49435663e-01 -6.57770574e-01 9.69319463e-01
2.84239501e-01 -8.97003829e-01 1.08632720e+00 2.69819975e-01
-4.57776666e-01 -1.17940247e+00 -9.65881407e-01 -9.76784229e-01
-2.10351497e-01 -5.80907941e-01 1.15379858e+00 1.11418486e+00
4.70604673e-02 2.35847414e-01 -7.00439095e-01 1.23691797e-01
4.93266314e-01 5.74201703e-01 6.71037376e-01 -1.11596370e+00
-3.20193321e-02 -5.69412708e-01 -4.17549983e-02 -1.37942290e+00
3.86194587e-02 -1.15557194e+00 -3.45361590e-01 -1.91974175e+00
1.75079137e-01 -8.54153693e-01 -4.87718612e-01 7.07135737e-01
-3.55820619e-02 -5.25599062e-01 -2.68821687e-01 5.05770862e-01
-1.02284980e+00 1.01618934e+00 1.81110275e+00 -1.36314943e-01
-3.35807145e-01 -2.48564646e-01 -7.01091349e-01 2.04271913e-01
1.14763355e+00 -3.92606497e-01 -8.35106075e-01 -8.16293657e-01
8.18088293e-01 5.07707298e-01 4.70414758e-01 -8.06401074e-01
8.44159544e-01 -6.65304780e-01 -8.49446133e-02 -9.83940303e-01
3.15277874e-02 -5.18288136e-01 5.71450770e-01 6.10562503e-01
-1.91404283e-01 1.39617756e-01 2.10441709e-01 9.85045671e-01
-2.22417399e-01 2.48132929e-01 6.90045506e-02 -1.60958901e-01
-1.08632112e+00 1.05794406e+00 3.53301674e-01 1.04586825e-01
8.48863900e-01 -2.61010408e-01 -5.91089129e-01 -2.28445485e-01
-4.44416970e-01 1.00621033e+00 2.14899063e-01 2.49807507e-01
8.41680706e-01 -1.57813215e+00 -5.28682053e-01 -1.30640939e-01
2.35784009e-01 3.85186076e-01 8.08167815e-01 1.13844931e+00
-6.81339681e-01 7.69533813e-01 2.20996793e-02 -3.48376900e-01
-6.71933830e-01 9.85744834e-01 2.48106077e-01 -7.18838334e-01
-7.80066907e-01 7.44486213e-01 7.95500427e-02 -6.71830118e-01
1.10589020e-01 -3.54724109e-01 -8.66673514e-02 -3.57124895e-01
1.56795859e-01 5.32733679e-01 -2.03015015e-01 1.30711228e-01
-3.94613355e-01 4.64561731e-01 -1.61071807e-01 3.31334144e-01
1.26494336e+00 -2.23915532e-01 -2.02739254e-01 2.18886919e-02
8.77016544e-01 -3.33846033e-01 -8.25789571e-01 -4.87563699e-01
3.27266054e-03 -2.80585229e-01 8.73527676e-02 -5.07225692e-01
-1.38435161e+00 4.47370857e-01 -7.99080133e-02 3.58369082e-01
1.42383897e+00 7.11579025e-02 1.22467923e+00 5.38572252e-01
6.17555022e-01 -1.16530132e+00 1.46584913e-01 5.34431815e-01
8.85444582e-01 -8.42189968e-01 -9.25593227e-02 -5.25614142e-01
-4.16226983e-01 1.25035453e+00 9.13706958e-01 6.91968724e-02
5.28712094e-01 -4.03888553e-01 -5.64411938e-01 -6.35742009e-01
-1.10671127e+00 -8.57147425e-02 1.37959972e-01 3.25616390e-01
-3.56708378e-01 1.13444351e-01 -4.48102355e-01 6.28817260e-01
-8.50021616e-02 4.00069773e-01 2.35358134e-01 7.73021281e-01
-2.36303136e-01 -9.63009894e-01 1.62898544e-02 5.15950382e-01
8.71979296e-02 -2.57497698e-01 -1.79839149e-01 8.46721888e-01
-2.38321155e-01 7.39324033e-01 -3.43523830e-01 -7.30846763e-01
1.97867304e-02 -4.27747577e-01 4.24108654e-01 -3.30306768e-01
-1.78633720e-01 -2.58700430e-01 2.26374939e-01 -7.42324829e-01
-2.03025565e-01 -1.65047169e-01 -1.29285157e+00 -6.86611354e-01
-3.70637536e-01 2.53063977e-01 4.80266474e-02 7.73694336e-01
6.13775671e-01 6.24311328e-01 4.49729413e-01 -1.77579835e-01
-3.69212449e-01 -7.88852572e-01 -6.19142234e-01 -4.13267463e-02
4.36693244e-02 -7.72662342e-01 -1.40945345e-01 -5.36993027e-01] | [8.64083480834961, 7.886760234832764] |
f681f48f-bb83-400c-a1db-73f27e94357d | probabilistic-visual-place-recognition-for | 2105.03091 | null | https://arxiv.org/abs/2105.03091v1 | https://arxiv.org/pdf/2105.03091v1.pdf | Probabilistic Visual Place Recognition for Hierarchical Localization | Visual localization techniques often comprise a hierarchical localization pipeline, with a visual place recognition module used as a coarse localizer to initialize a pose refinement stage. While improving the pose refinement step has been the focus of much recent research, most work on the coarse localization stage has focused on improvements like increased invariance to appearance change, without improving what can be loose error tolerances. In this letter, we propose two methods which adapt image retrieval techniques used for visual place recognition to the Bayesian state estimation formulation for localization. We demonstrate significant improvements to the localization accuracy of the coarse localization stage using our methods, whilst retaining state-of-the-art performance under severe appearance change. Using extensive experimentation on the Oxford RobotCar dataset, results show that our approach outperforms comparable state-of-the-art methods in terms of precision-recall performance for localizing image sequences. In addition, our proposed methods provides the flexibility to contextually scale localization latency in order to achieve these improvements. The improved initial localization estimate opens up the possibility of both improved overall localization performance and modified pose refinement techniques that leverage this improved spatial prior. | ['Michael Milford', 'Niko Sünderhauf', 'Ming Xu'] | 2021-05-07 | null | null | null | null | ['visual-place-recognition'] | ['computer-vision'] | [ 1.51413932e-01 -2.20549807e-01 -2.32740790e-01 -4.20429587e-01
-1.13767946e+00 -8.96932960e-01 8.60290587e-01 3.78129870e-01
-8.26807082e-01 4.85976905e-01 1.08837429e-02 -9.51893926e-02
1.21525437e-01 -2.73257196e-01 -9.05795455e-01 -5.05226612e-01
-2.26029903e-01 4.26838100e-01 8.07165802e-01 -6.18044324e-02
6.28551960e-01 8.51858616e-01 -1.53904569e+00 -1.90087512e-01
4.94287997e-01 9.17098284e-01 4.35130447e-01 7.48821259e-01
2.83325762e-01 6.54433250e-01 -5.44865251e-01 3.04532826e-01
1.51522204e-01 7.83687532e-02 -7.35968649e-01 1.84998512e-02
7.35524118e-01 -2.84642518e-01 4.35340824e-03 1.04227054e+00
4.47474241e-01 3.62347275e-01 3.74795288e-01 -1.11728477e+00
-3.36191893e-01 9.94267017e-02 -7.84916699e-01 2.43931413e-01
6.55455709e-01 -8.50381255e-02 7.86764264e-01 -8.41740847e-01
7.00584769e-01 1.17257130e+00 8.66087556e-01 1.60287857e-01
-1.52567542e+00 -4.09206331e-01 3.31284642e-01 2.91395992e-01
-1.87714589e+00 -6.50205672e-01 4.00587976e-01 -3.39717180e-01
1.17721474e+00 -1.66320771e-01 2.92454541e-01 8.31930518e-01
1.93514705e-01 6.87517524e-01 1.08403778e+00 -5.70264101e-01
4.64291602e-01 1.35943502e-01 -2.73413695e-02 8.57975602e-01
2.21565768e-01 9.17571411e-02 -6.12462401e-01 -1.50607705e-01
8.37889910e-01 -8.89888182e-02 -1.16328150e-01 -1.18305445e+00
-1.43702555e+00 5.57176709e-01 8.40706170e-01 1.79869145e-01
-3.58658880e-01 5.16259432e-01 1.64751619e-01 -7.51273036e-02
2.39633873e-01 4.81905133e-01 -3.44324023e-01 -1.91319600e-01
-1.17372286e+00 1.78801462e-01 6.62277341e-01 1.09771109e+00
1.09826553e+00 -3.81935596e-01 4.68850555e-03 4.98330235e-01
6.15612626e-01 5.70830762e-01 2.80205786e-01 -1.08253741e+00
2.10396439e-01 3.20771128e-01 3.71335000e-01 -1.08108997e+00
-4.75892752e-01 -3.34073216e-01 -1.90580651e-01 3.72884929e-01
2.23433539e-01 2.61467695e-01 -1.15824199e+00 1.71321940e+00
4.27924603e-01 3.12039554e-01 -1.11922748e-01 9.26135659e-01
1.71761408e-01 4.35671628e-01 1.02697544e-01 2.16377750e-01
1.33371580e+00 -1.15101337e+00 -4.29797709e-01 -5.77474535e-01
5.28297901e-01 -8.68802965e-01 6.49282098e-01 1.22785240e-01
-7.11229086e-01 -5.79208732e-01 -1.27704787e+00 -1.78216338e-01
-3.56485605e-01 2.34367177e-01 6.28490925e-01 5.26177466e-01
-1.48654974e+00 3.81127387e-01 -1.27149928e+00 -1.02209032e+00
6.97063655e-02 6.13330543e-01 -7.37663209e-01 -1.09106809e-01
-7.02801943e-01 1.15721536e+00 4.31275457e-01 -1.42129704e-01
-8.78906727e-01 -4.55332100e-01 -1.17613554e+00 -2.38857612e-01
3.00385565e-01 -5.44383585e-01 1.38139117e+00 -4.50136691e-01
-1.47003400e+00 6.77316606e-01 -6.03310585e-01 -6.05612338e-01
3.67888361e-01 -1.98286101e-01 1.58471495e-01 2.98303038e-01
4.73174125e-01 1.16881227e+00 6.64362848e-01 -1.35444033e+00
-8.71985912e-01 -2.79626429e-01 4.98647355e-02 3.99429113e-01
2.91111410e-01 -5.91151342e-02 -8.44399273e-01 -3.41497451e-01
4.45965171e-01 -1.32116675e+00 -4.84796613e-01 4.61927056e-02
8.12406167e-02 3.56022269e-02 7.91521370e-01 -3.43206346e-01
8.19077432e-01 -2.10699105e+00 7.40563497e-02 2.58409888e-01
-6.11507520e-02 -7.18017220e-02 -1.45708501e-01 4.27730352e-01
2.13421598e-01 -1.91530585e-01 3.10117304e-02 -9.04683352e-01
9.93502066e-02 1.80812910e-01 -2.20670417e-01 1.03208804e+00
1.98076949e-01 7.88775802e-01 -9.24773455e-01 -3.90308768e-01
5.73399544e-01 6.33834243e-01 -5.49405158e-01 7.25711584e-02
-1.14330007e-02 5.90030134e-01 -3.95936966e-01 8.36419702e-01
5.44460416e-01 -3.69688839e-01 -1.49967102e-02 -1.15959942e-01
-4.07979369e-01 1.79069027e-01 -1.28405941e+00 2.25021386e+00
-3.26333940e-01 6.44661844e-01 1.94365636e-01 -5.85059524e-01
6.54226542e-01 -5.62931746e-02 3.39395791e-01 -6.33644521e-01
-2.65515652e-02 1.62976384e-01 -2.93348491e-01 8.07109624e-02
9.99260366e-01 2.60496527e-01 -2.56147236e-01 9.85951871e-02
2.59996384e-01 -7.75877610e-02 8.41396526e-02 1.45556226e-01
1.27761948e+00 7.11206198e-01 5.45390666e-01 -3.24479371e-01
5.29655099e-01 3.73989820e-01 2.32194975e-01 9.90575254e-01
-4.87176627e-01 7.53637969e-01 -6.95547611e-02 -1.73357293e-01
-1.04297900e+00 -1.01201642e+00 -1.29011780e-01 1.26874256e+00
6.47409201e-01 -5.35823166e-01 -7.08974540e-01 -5.61295211e-01
-1.00695260e-01 3.44987810e-01 -6.06616974e-01 1.38664767e-01
-6.11959398e-01 -4.59553808e-01 7.24143267e-01 7.00629652e-01
5.17790198e-01 -7.87228465e-01 -1.06522882e+00 1.09931372e-01
-1.24897867e-01 -1.17705536e+00 -3.60309035e-01 4.99225438e-01
-5.82844973e-01 -7.85259426e-01 -7.31855929e-01 -8.02223265e-01
7.73938000e-01 5.05819559e-01 7.49829352e-01 -2.34164104e-01
-2.25592494e-01 6.79903626e-01 -4.61015433e-01 -1.92541152e-01
-1.27716929e-01 4.18511659e-01 2.93186039e-01 -3.25743198e-01
1.93519175e-01 -3.28798532e-01 -6.72770977e-01 3.64811987e-01
-7.24080443e-01 -2.14306056e-01 7.86492229e-01 6.10002160e-01
8.11639547e-01 -4.60956931e-01 -5.78749143e-02 -2.82117814e-01
1.21523820e-01 -2.85365194e-01 -9.11370337e-01 1.39526695e-01
-7.43567884e-01 4.90546376e-01 4.06091437e-02 -3.72100592e-01
-7.78822958e-01 7.01417685e-01 -9.16216448e-02 -3.80868316e-01
-4.18170363e-01 3.30944270e-01 1.89116046e-01 -7.98717558e-01
7.34659135e-01 1.36570215e-01 -1.02744646e-01 -4.52786535e-01
5.87114215e-01 3.40239674e-01 6.54119015e-01 -5.21750450e-01
7.41631687e-01 5.51838875e-01 1.27469167e-01 -5.43845654e-01
-4.68680352e-01 -9.31972146e-01 -1.04146934e+00 6.81458488e-02
7.70535111e-01 -1.05640590e+00 -6.01926923e-01 1.39468685e-01
-1.17450392e+00 -3.13651413e-01 8.89474452e-02 4.73343790e-01
-6.81597233e-01 4.36234385e-01 -4.37881678e-01 -5.32355249e-01
1.25037730e-01 -1.49893463e+00 1.76647389e+00 1.96056157e-01
-2.65429318e-01 -9.43939090e-01 4.45962399e-01 5.32760751e-03
4.68621135e-01 1.88520640e-01 1.83869645e-01 -5.00666022e-01
-1.04841423e+00 -2.60754496e-01 -2.99031019e-01 -2.62527496e-01
-9.45586562e-02 -5.07223070e-01 -1.03134167e+00 -6.85765505e-01
-4.22971696e-01 -1.86093763e-01 8.64706516e-01 2.45926008e-01
3.83749485e-01 1.01486534e-01 -7.63763130e-01 6.99808478e-01
1.52687657e+00 1.05671413e-01 4.94110882e-01 9.05067742e-01
6.09700799e-01 2.39611343e-01 9.05619562e-01 1.78618580e-01
6.49285078e-01 1.11746001e+00 4.31433082e-01 -2.01070711e-01
-2.53002763e-01 -2.30843216e-01 3.97907227e-01 3.00156683e-01
1.87457487e-01 6.06426485e-02 -1.12440991e+00 7.85512865e-01
-2.29224443e+00 -6.86107159e-01 3.35646957e-01 2.20149016e+00
6.17311537e-01 -1.72495514e-01 -5.23041822e-02 -2.96815187e-01
4.92308080e-01 3.30901265e-01 -2.58205146e-01 2.96316184e-02
2.99288511e-01 -2.95131486e-02 1.01210105e+00 8.68249357e-01
-1.38878095e+00 1.37346637e+00 6.85182333e+00 5.03010750e-01
-1.25252604e+00 1.40465498e-01 2.92762332e-02 1.67675212e-01
3.58315349e-01 2.99094737e-01 -1.02153909e+00 3.82492617e-02
7.85044074e-01 3.14248145e-01 5.14825344e-01 1.12282968e+00
-4.21136245e-02 -6.64809346e-01 -1.24737716e+00 9.78906274e-01
4.14164007e-01 -1.10580897e+00 -3.30858946e-01 1.01292312e-01
5.82374454e-01 5.09220660e-01 8.47942531e-02 2.10068792e-01
2.63378531e-01 -8.39913428e-01 1.12261677e+00 3.25627834e-01
5.60853779e-01 -6.72343850e-01 7.42094696e-01 3.46787542e-01
-1.47638130e+00 2.06350416e-01 -2.32513353e-01 -6.01326711e-02
2.02811211e-01 -5.22741824e-02 -1.35431325e+00 4.08551544e-01
8.01570833e-01 4.55288023e-01 -1.13531125e+00 1.26865232e+00
-3.54807407e-01 1.38293415e-01 -5.88293314e-01 9.30594839e-03
3.26005548e-01 3.36282432e-01 4.88851637e-01 1.47387516e+00
1.55030742e-01 -3.66379768e-01 6.46386862e-01 6.13329113e-01
3.95275682e-01 -1.31037697e-01 -6.41494453e-01 4.75818843e-01
6.19858921e-01 1.24262476e+00 -1.06292737e+00 -1.82360634e-01
-1.26582891e-01 1.22582150e+00 5.85460126e-01 3.56779337e-01
-8.85217726e-01 -4.63994265e-01 5.79005480e-01 5.73484674e-02
6.39133275e-01 -7.58663356e-01 -6.23199195e-02 -1.15493739e+00
-1.25342667e-01 -5.82967281e-01 4.27962430e-02 -8.20903838e-01
-5.62242568e-01 4.65439111e-01 2.58720964e-01 -9.57060218e-01
-5.52042067e-01 -4.94474590e-01 6.69947490e-02 7.85629570e-01
-1.54726744e+00 -1.40084529e+00 -2.95357645e-01 3.31794590e-01
4.69343692e-01 2.59478539e-01 7.99540997e-01 2.15270475e-01
-1.28994346e-01 5.42419136e-01 1.11778844e-02 -2.50179209e-02
1.09029877e+00 -1.24984932e+00 4.26427186e-01 1.27730525e+00
3.57895255e-01 1.12116766e+00 7.94717193e-01 -6.65027916e-01
-1.48097718e+00 -1.02452838e+00 7.14565456e-01 -8.34546149e-01
6.12826407e-01 -5.22009373e-01 -4.50991958e-01 9.43383932e-01
3.32420617e-01 1.69082016e-01 1.56243175e-01 7.64789581e-02
-3.42579156e-01 2.53911316e-02 -1.09466493e+00 5.56119323e-01
7.97101676e-01 -6.95666909e-01 -6.00247324e-01 -7.03142509e-02
5.08734465e-01 -7.49804139e-01 -6.08492911e-01 4.76609111e-01
5.78412831e-01 -7.57655203e-01 1.08999693e+00 1.89356580e-01
-4.38009918e-01 -1.14891994e+00 -4.34004486e-01 -1.14033484e+00
-5.86815059e-01 -3.74896526e-01 3.76890860e-02 1.09201169e+00
2.18132108e-01 -3.74627560e-01 6.66707337e-01 3.77785742e-01
-1.54590607e-03 -2.05547512e-01 -1.02764606e+00 -6.87635243e-01
-5.63095272e-01 -4.12859201e-01 2.39166439e-01 5.66738904e-01
-4.23714630e-02 3.59723628e-01 -2.09877923e-01 7.52677202e-01
7.68302023e-01 -7.62095451e-02 9.89736259e-01 -8.32647681e-01
-2.79029936e-01 -2.21763641e-01 -9.66301382e-01 -1.46347463e+00
1.88282117e-01 -6.21546507e-01 7.55051196e-01 -1.61119330e+00
2.21881583e-01 -3.27227652e-01 -3.25280279e-01 6.98231339e-01
1.04227535e-01 7.57945657e-01 3.18729669e-01 3.43329072e-01
-1.33617926e+00 2.98438996e-01 4.30296183e-01 1.55857559e-02
-1.80876851e-01 -3.98724049e-01 -3.33900630e-01 6.23166561e-01
4.46796417e-01 -5.26763976e-01 -3.22811067e-01 -5.14549851e-01
-5.46374321e-02 -3.83060962e-01 6.04975402e-01 -1.35684812e+00
6.82955682e-01 7.82960579e-02 5.26707232e-01 -7.32190013e-01
3.87014896e-01 -9.30690587e-01 9.60102230e-02 4.06441420e-01
-1.72638506e-01 3.76727849e-01 4.67395216e-01 8.06635022e-01
-1.65317297e-01 2.21079420e-02 7.90055335e-01 7.28110783e-03
-1.42003560e+00 -4.04296108e-02 -4.61991340e-01 -4.52879012e-01
1.12887013e+00 -3.12596887e-01 -1.87717050e-01 -2.56745577e-01
-5.69362879e-01 1.61989331e-01 1.13993645e+00 5.85771859e-01
3.85939360e-01 -1.19103694e+00 -1.16276480e-01 2.38547951e-01
4.29565042e-01 -6.74248859e-02 -3.16296846e-01 1.13201141e+00
-8.70979726e-01 6.37859285e-01 -4.58677635e-02 -1.15102065e+00
-1.26813817e+00 6.08752012e-01 3.35940957e-01 -1.56434655e-01
-4.06750321e-01 8.08667481e-01 1.27141848e-01 -4.56250548e-01
5.03578007e-01 -5.74297786e-01 4.58048955e-02 -1.25328138e-01
5.49125791e-01 3.23288620e-01 1.40491560e-01 -9.25524533e-01
-1.03096604e+00 7.70607352e-01 -1.48582742e-01 -5.02449393e-01
1.03253281e+00 -6.23180866e-01 -4.12693880e-02 3.77952099e-01
1.10654950e+00 1.24993362e-01 -1.43531144e+00 -6.31391332e-02
2.65561223e-01 -3.63954037e-01 1.52513102e-01 -8.21386933e-01
-3.51511925e-01 5.72166622e-01 9.70335066e-01 -3.04094344e-01
7.15291381e-01 2.36818269e-01 2.41595194e-01 6.92497373e-01
1.09170675e+00 -7.72959352e-01 -1.29255116e-01 8.82857144e-01
5.19027531e-01 -1.35410237e+00 1.51930839e-01 -1.28047839e-01
-4.25584823e-01 7.66616821e-01 4.55958635e-01 -3.45174491e-01
3.87627214e-01 4.60240304e-01 1.92911118e-01 4.42461902e-03
-3.06813091e-01 -4.66701716e-01 3.74474466e-01 7.28524446e-01
3.61569315e-01 -2.78557390e-01 6.94641620e-02 -1.81337878e-01
2.17096016e-01 -2.04217359e-01 4.83932942e-02 1.28862906e+00
-6.18894100e-01 -1.15083158e+00 -6.35406315e-01 -3.84328604e-01
-3.00803959e-01 -8.18049684e-02 -4.00264055e-01 9.03673649e-01
8.76220223e-03 8.36002469e-01 -1.88872647e-02 -2.17711672e-01
2.79891849e-01 -6.62270635e-02 7.35283017e-01 -6.55849814e-01
-3.35213274e-01 1.80344731e-01 -1.06147885e-01 -1.12742710e+00
-6.26013398e-01 -8.32552969e-01 -1.34791505e+00 8.12038593e-03
-3.86328608e-01 1.76573507e-02 9.27736163e-01 1.03028262e+00
6.95629299e-01 3.85577857e-01 1.33389123e-02 -1.64862180e+00
-3.14603657e-01 -7.84993231e-01 -3.00993353e-01 4.71351780e-02
7.18620181e-01 -9.44023252e-01 5.08406535e-02 8.16637352e-02] | [7.551239490509033, -2.032731056213379] |
18477d77-5547-46a6-8016-87749ce382e9 | cread-combined-resolution-of-ellipses-and | 2105.09914 | null | https://arxiv.org/abs/2105.09914v1 | https://arxiv.org/pdf/2105.09914v1.pdf | CREAD: Combined Resolution of Ellipses and Anaphora in Dialogues | Anaphora and ellipses are two common phenomena in dialogues. Without resolving referring expressions and information omission, dialogue systems may fail to generate consistent and coherent responses. Traditionally, anaphora is resolved by coreference resolution and ellipses by query rewrite. In this work, we propose a novel joint learning framework of modeling coreference resolution and query rewriting for complex, multi-turn dialogue understanding. Given an ongoing dialogue between a user and a dialogue assistant, for the user query, our joint learning model first predicts coreference links between the query and the dialogue context, and then generates a self-contained rewritten user query. To evaluate our model, we annotate a dialogue based coreference resolution dataset, MuDoCo, with rewritten queries. Results show that the performance of query rewrite can be substantially boosted (+2.3% F1) with the aid of coreference modeling. Furthermore, our joint model outperforms the state-of-the-art coreference resolution model (+2% F1) on this dataset. | ['Hong Yu', 'Lin Li', 'Dhivya Piraviperumal', 'Joel Ruben Antony Moniz', 'Jiarui Lu', 'Shruti Bhargava', 'Bo-Hsiang Tseng'] | 2021-05-20 | null | https://aclanthology.org/2021.naacl-main.265 | https://aclanthology.org/2021.naacl-main.265.pdf | naacl-2021-4 | ['dialogue-understanding'] | ['natural-language-processing'] | [ 1.50014818e-01 8.87011707e-01 -5.01716673e-01 -5.27435720e-01
-1.27075076e+00 -8.38634908e-01 9.67071295e-01 1.79390386e-01
-5.20489633e-01 1.04710114e+00 7.43808210e-01 -1.30366003e-02
-1.24601685e-01 -6.24502301e-01 -2.31470123e-01 6.19636290e-02
3.81010860e-01 1.24157941e+00 2.82410294e-01 -8.10611367e-01
4.20513116e-02 3.54370415e-01 -1.12053335e+00 7.86505818e-01
8.30790997e-01 2.27568254e-01 1.41273871e-01 7.08353341e-01
-3.69224370e-01 1.03315735e+00 -8.01601529e-01 -6.64759219e-01
-4.23925638e-01 -5.22499323e-01 -1.79895997e+00 -2.96327680e-01
1.00809373e-01 -4.02563959e-01 -4.82760757e-01 7.62492001e-01
5.53008556e-01 3.91849935e-01 4.09094959e-01 -9.74664330e-01
5.01831584e-02 9.27059233e-01 -1.51803687e-01 1.55376330e-01
1.20952308e+00 -2.90912777e-01 1.51084042e+00 -4.97775018e-01
9.96266544e-01 1.96786904e+00 3.45586032e-01 8.15756917e-01
-1.41496480e+00 -4.27791178e-01 -3.82831618e-02 3.18669736e-01
-9.38766778e-01 -8.46406639e-01 7.62988329e-01 3.32273841e-02
1.37399030e+00 5.54986894e-01 1.08838588e-01 1.15026844e+00
-6.32001385e-02 9.54985440e-01 7.58036196e-01 -6.47597134e-01
-2.35451534e-01 3.22222084e-01 3.59349877e-01 4.63239998e-01
-5.41082859e-01 -7.33681917e-02 -6.56106532e-01 -5.94982088e-01
4.41799045e-01 -7.33639777e-01 -4.78087783e-01 -1.11652970e-01
-7.68323004e-01 1.08191347e+00 2.39652470e-01 2.80706912e-01
-2.96915770e-01 -5.23910463e-01 5.35839736e-01 5.19426525e-01
2.27686435e-01 8.95643353e-01 -3.39887381e-01 -4.10400212e-01
-3.22564662e-01 7.43252158e-01 1.62871706e+00 9.44475353e-01
3.44787806e-01 -6.75218999e-01 -4.23561901e-01 1.35100114e+00
1.96181387e-01 3.67711186e-01 2.32967958e-01 -1.57788289e+00
5.16378224e-01 6.94312990e-01 5.44516087e-01 -8.95566046e-01
-6.76423907e-01 6.20669797e-02 -2.13475317e-01 -4.27483767e-01
5.33005118e-01 -2.40961865e-01 1.70244604e-01 2.04973245e+00
3.73498946e-01 -4.15285945e-01 7.96335638e-01 9.40366447e-01
1.35819662e+00 5.58067977e-01 4.38828096e-02 -9.30163324e-01
1.74622178e+00 -1.04080260e+00 -1.20401299e+00 -3.35630596e-01
8.28554928e-01 -1.05818427e+00 8.53209555e-01 1.63364187e-01
-1.35000253e+00 -2.68210173e-01 -5.84955692e-01 -4.83086973e-01
3.42691898e-01 -1.95362225e-01 4.39470261e-01 -5.77302948e-02
-6.03157103e-01 3.30470741e-01 -5.13735414e-01 -6.35838270e-01
-2.97714412e-01 3.36769372e-01 -3.66064787e-01 2.23950595e-02
-1.64168417e+00 1.40533793e+00 4.00008678e-01 -3.42142344e-01
-4.59763527e-01 -5.13228059e-01 -8.46842945e-01 1.17293514e-01
7.54005730e-01 -5.80894232e-01 2.01742697e+00 -2.73083448e-01
-1.64998782e+00 1.02684152e+00 -3.84833246e-01 -3.79264772e-01
4.35852110e-01 -6.81792617e-01 -2.24563241e-01 9.86040011e-02
2.18620077e-01 6.70832336e-01 1.80006817e-01 -1.20720303e+00
-8.56814504e-01 -3.12628597e-01 5.69039285e-01 8.16945910e-01
3.04783583e-01 3.92718792e-01 -4.76450354e-01 1.40276004e-03
8.47515836e-02 -9.57611978e-01 1.91108838e-01 -4.87313569e-01
-4.65329856e-01 -8.63789082e-01 8.56295526e-01 -4.78378892e-01
1.30581224e+00 -1.78998756e+00 6.15484893e-01 -4.51712191e-01
5.88297062e-02 3.33059877e-01 -1.77665114e-01 5.71779370e-01
1.59986187e-02 -1.80534735e-01 -3.60424630e-02 -3.74264061e-01
7.50609115e-02 4.02216017e-01 -6.04980350e-01 -2.30663586e-02
2.07913574e-02 7.52637565e-01 -1.12752807e+00 -6.99026465e-01
-7.94275925e-02 2.11006939e-01 -5.59422135e-01 8.47606599e-01
-5.29954851e-01 2.87891924e-01 -4.86498624e-01 1.92957625e-01
1.57238722e-01 -6.50887191e-02 8.84062529e-01 -6.41373277e-01
-1.63917132e-02 7.92734087e-01 -8.42038870e-01 1.81236780e+00
-5.36005676e-01 5.83305895e-01 4.61567402e-01 -5.03588855e-01
1.03748167e+00 7.05079257e-01 1.32865265e-01 -8.39034438e-01
-2.17319746e-03 9.04122666e-02 -3.24940234e-02 -7.43515730e-01
8.52325141e-01 -2.25104988e-01 -5.59573352e-01 9.31966782e-01
1.95178628e-01 -4.80583847e-01 9.15877298e-02 5.96278965e-01
7.82155037e-01 -3.48018110e-02 5.19526124e-01 6.70052506e-03
8.06395829e-01 2.29411900e-01 5.60480177e-01 6.72019064e-01
-2.42399454e-01 -5.04842736e-02 9.44868803e-01 -3.72165084e-01
-6.41291201e-01 -8.87809992e-01 -1.77183887e-03 1.57148874e+00
1.78549275e-01 -5.35930216e-01 -6.99284256e-01 -7.88086653e-01
-1.08527102e-01 1.11172485e+00 -8.91494304e-02 -1.94462761e-01
-9.53556478e-01 -3.75368446e-01 9.60099697e-01 2.15114951e-01
4.89283293e-01 -1.32590091e+00 -6.14958107e-01 4.05702502e-01
-1.08189809e+00 -1.28704751e+00 -3.30160677e-01 -1.23428047e-01
-5.73430061e-01 -1.42209458e+00 -2.10289862e-02 -6.41676724e-01
-1.07165158e-01 9.53381881e-02 1.36112475e+00 -8.48440081e-02
6.10088184e-02 4.80542958e-01 -1.94111720e-01 -4.67184968e-02
-9.21794534e-01 2.42236093e-01 1.11650936e-01 -5.05658209e-01
5.28442502e-01 -3.50676864e-01 1.14501463e-02 2.52218217e-01
-4.76246655e-01 9.33519006e-02 1.48658380e-01 1.08946228e+00
2.89562792e-02 -6.46386147e-01 6.67731404e-01 -1.25733018e+00
1.35264635e+00 -2.56907016e-01 -2.86844999e-01 5.57033002e-01
-4.57733810e-01 2.09265113e-01 5.25182895e-02 -2.41035447e-01
-1.81536984e+00 -1.39807299e-01 6.69057891e-02 2.30321120e-02
-1.40045524e-01 5.27062654e-01 -1.86629921e-01 4.28583086e-01
8.93346488e-01 -2.54664093e-01 1.14732288e-01 -5.96224666e-01
8.14642608e-01 9.09185112e-01 1.09326220e+00 -1.07012427e+00
5.50302640e-02 -5.12045883e-02 -3.88948798e-01 -6.66203499e-01
-1.31511092e+00 -5.63771904e-01 -7.29910791e-01 -8.70150700e-02
7.00317025e-01 -8.01267087e-01 -1.05781198e+00 -8.77297968e-02
-1.83511853e+00 -1.21110767e-01 3.89250629e-02 3.53225827e-01
-8.28818321e-01 5.90498984e-01 -8.89854968e-01 -9.09508646e-01
-4.80122328e-01 -9.40403163e-01 7.71984816e-01 4.22568828e-01
-1.10937929e+00 -9.67790782e-01 4.45822626e-01 8.08202088e-01
2.29194611e-01 -1.32121667e-02 1.19287407e+00 -1.34512758e+00
6.62934929e-02 5.92390634e-02 -4.31551069e-01 -4.44662303e-01
3.41810435e-01 -3.24199975e-01 -8.88541877e-01 -3.26761931e-01
1.10556282e-01 -8.99940729e-01 2.58937329e-01 -2.50238746e-01
3.65336658e-03 -7.52784789e-01 -5.49514115e-01 4.08252552e-02
6.57562494e-01 2.05540970e-01 2.88568348e-01 2.11028367e-01
9.33091044e-02 1.13504350e+00 9.58355486e-01 4.61883485e-01
7.13953853e-01 1.15807056e+00 1.40258372e-01 2.74044275e-01
1.49374055e-02 -1.51773438e-01 4.74138856e-02 4.55896705e-01
2.98575044e-01 -1.16053671e-01 -8.39850783e-01 5.02082527e-01
-2.16014171e+00 -1.07902837e+00 -1.62859753e-01 1.75204635e+00
1.60163689e+00 4.54188399e-02 -3.96246687e-02 -2.50906408e-01
8.15424144e-01 2.54244506e-01 -3.92571449e-01 -6.92057490e-01
-1.32675588e-01 3.06640323e-02 -4.43882048e-01 1.34965265e+00
-1.08206534e+00 1.43895006e+00 6.07416916e+00 2.10090160e-01
-7.04791784e-01 -3.52911018e-02 -9.65520218e-02 -9.84496158e-03
-1.77724481e-01 1.70145676e-01 -9.39835191e-01 -2.13958159e-01
8.47554862e-01 -5.79222739e-01 4.02277172e-01 5.18528342e-01
5.02955951e-02 5.43542504e-02 -1.39704323e+00 7.02815056e-01
2.19779149e-01 -1.06824446e+00 4.22170982e-02 -4.63517934e-01
1.74002901e-01 -3.88452917e-01 -6.56542003e-01 8.94862652e-01
6.15159333e-01 -8.87959361e-01 -1.67255580e-01 5.36161184e-01
5.06688416e-01 -6.74800515e-01 8.93129528e-01 6.13478839e-01
-6.65516019e-01 -7.28898123e-02 -1.18793987e-01 -4.35684295e-03
5.55631816e-01 -2.39664108e-01 -1.11195385e+00 5.61196983e-01
3.58861029e-01 2.46705294e-01 -1.32864013e-01 4.29380774e-01
-4.83286381e-01 6.13517836e-02 -2.15062007e-01 1.94342643e-01
1.48265082e-02 7.35685304e-02 9.23362672e-01 1.51811147e+00
-4.67827648e-01 8.19261849e-01 4.41358805e-01 9.14441884e-01
-3.43050271e-01 9.35608149e-02 -3.68212402e-01 4.59513590e-02
1.13785911e+00 1.28170598e+00 2.34695837e-01 -3.92651469e-01
-1.34458899e-01 9.13137794e-01 5.87852001e-01 1.96616769e-01
-4.33345228e-01 -4.20750231e-01 4.75013405e-01 -4.40342575e-01
-2.80775249e-01 4.35395539e-01 3.46427113e-01 -9.30837810e-01
-3.69957477e-01 -1.27500117e+00 9.06912625e-01 -7.58916736e-01
-1.08910501e+00 6.65566206e-01 3.08598429e-01 -5.65393388e-01
-1.00532341e+00 -1.00305073e-01 -6.53208613e-01 8.84392440e-01
-1.29383254e+00 -1.09648371e+00 -2.26401389e-01 3.76271158e-01
7.16369390e-01 -8.79291445e-02 1.51429009e+00 -3.56735960e-02
-4.47723657e-01 4.55859721e-01 -6.05758071e-01 3.28321517e-01
1.39615881e+00 -1.25108349e+00 -2.88998690e-02 1.19125754e-01
-1.60291061e-01 7.49297380e-01 1.07427442e+00 -5.38269997e-01
-1.13702059e+00 -5.69664538e-01 1.52004516e+00 -4.90569800e-01
5.73658049e-01 1.24531455e-01 -1.41604376e+00 8.39643657e-01
5.38714945e-01 -5.49727917e-01 8.64398599e-01 5.40219009e-01
-5.77062786e-01 2.75936991e-01 -1.20621836e+00 4.76764411e-01
7.36186922e-01 -8.30669820e-01 -1.76966190e+00 5.05518794e-01
9.78064954e-01 -9.57520843e-01 -1.14200294e+00 6.31619334e-01
4.91648912e-01 -7.27060139e-01 8.39307785e-01 -1.31687248e+00
2.37675995e-01 2.36582175e-01 -3.01493198e-01 -1.03035605e+00
-9.21295509e-02 -8.65698934e-01 -4.10058618e-01 1.34823811e+00
3.95997047e-01 -4.54928428e-02 4.41999614e-01 1.11789930e+00
-3.45001779e-02 -2.85156935e-01 -1.00923872e+00 -1.15925021e-01
8.51554349e-02 3.21756629e-03 3.03505749e-01 1.05803478e+00
1.05499864e+00 1.34891355e+00 -2.72662193e-01 -1.89304482e-02
4.18659836e-01 5.25804043e-01 9.48998332e-01 -1.50617409e+00
-2.55563438e-01 -5.28733909e-01 5.13780475e-01 -1.33622456e+00
8.00782740e-01 -5.45675933e-01 1.31639823e-01 -1.42562509e+00
2.05657706e-01 -9.02580544e-02 4.13881779e-01 5.11495888e-01
-2.73813963e-01 -4.76058096e-01 2.19211429e-01 4.49769408e-01
-8.44567060e-01 6.61601841e-01 9.39727187e-01 -8.74152333e-02
-6.69198394e-01 5.53070232e-02 -5.68660259e-01 7.31529474e-01
5.46999633e-01 -2.77631193e-01 -3.68785590e-01 -2.96434790e-01
-3.42300236e-02 9.89167750e-01 -8.97130091e-03 -2.04493687e-01
7.81864822e-01 -2.10691094e-01 -3.89470071e-01 -3.56510133e-01
6.63560390e-01 -2.60827690e-01 -3.38740736e-01 2.98609674e-01
-9.21027005e-01 -2.65509337e-02 3.19859892e-01 3.30610096e-01
-2.57039458e-01 -3.50410789e-01 5.68916559e-01 -2.80371785e-01
-4.29908097e-01 -5.03723145e-01 -3.95425081e-01 5.90159237e-01
4.88563538e-01 5.36156535e-01 -6.69659436e-01 -7.90900469e-01
-1.01656079e+00 6.65640533e-01 9.87914130e-02 7.15567589e-01
5.30075669e-01 -9.90338385e-01 -8.85795891e-01 -3.01481366e-01
9.51645523e-02 1.88382670e-01 -2.52685584e-02 5.19261479e-01
4.39763591e-02 9.59795535e-01 -9.35563669e-02 -5.48331380e-01
-1.69491589e+00 2.49716416e-01 5.61044097e-01 -5.12483060e-01
-3.31813365e-01 7.33865082e-01 -6.05645739e-02 -1.03454936e+00
5.78934431e-01 3.02324861e-01 -8.77717376e-01 3.67194533e-01
5.85431695e-01 2.47857422e-01 -2.48078108e-01 -7.88953483e-01
-3.06597382e-01 1.38026536e-01 -5.84362388e-01 -4.63350326e-01
1.04306817e+00 -3.57099354e-01 -3.54401648e-01 4.22301441e-01
7.97131002e-01 5.99340387e-02 -6.90283895e-01 -8.74994814e-01
3.76487583e-01 -2.53682315e-01 -3.42548192e-01 -1.04463005e+00
-2.96697944e-01 4.99691486e-01 5.07670688e-03 2.48690516e-01
5.40064752e-01 3.93717647e-01 5.41673481e-01 1.09827268e+00
1.60316914e-01 -8.94582808e-01 2.17199534e-01 1.02121770e+00
1.32733834e+00 -1.20539498e+00 9.59778354e-02 -5.98476470e-01
-8.65924895e-01 1.11588538e+00 1.14598215e+00 3.64815265e-01
-3.11255665e-03 1.24561355e-01 4.43335086e-01 -3.95071238e-01
-1.33500803e+00 1.05641373e-01 9.67035219e-02 3.55045438e-01
8.88084710e-01 -7.34321624e-02 -4.40922797e-01 8.69397342e-01
-4.27638441e-01 -3.34247917e-01 5.72039604e-01 7.20730066e-01
-5.60444772e-01 -1.32946610e+00 -1.58181623e-01 -3.01915668e-02
-3.25846195e-01 2.16312036e-01 -1.08314502e+00 8.14994812e-01
-7.59548604e-01 1.25406897e+00 1.45105004e-01 -1.52971596e-01
8.67339969e-01 3.16136688e-01 3.64681423e-01 -8.34964514e-01
-7.07542181e-01 -6.08129539e-02 9.24981177e-01 -6.41815543e-01
-5.56321263e-01 -5.21093488e-01 -1.54617751e+00 -1.82746142e-01
-5.41572511e-01 8.67468774e-01 3.34968306e-02 1.26978016e+00
2.91497827e-01 2.17445344e-01 4.83713210e-01 -3.03589940e-01
-9.34104264e-01 -1.31995666e+00 1.34762332e-01 6.50931120e-01
2.03933433e-01 -6.24750674e-01 -2.16754466e-01 -2.55041659e-01] | [12.310928344726562, 8.157443046569824] |
32c9695e-27c0-40f9-8271-54c8e9d0f7d9 | ledits-real-image-editing-with-ddpm-inversion | 2307.00522 | null | https://arxiv.org/abs/2307.00522v1 | https://arxiv.org/pdf/2307.00522v1.pdf | LEDITS: Real Image Editing with DDPM Inversion and Semantic Guidance | Recent large-scale text-guided diffusion models provide powerful image-generation capabilities. Currently, a significant effort is given to enable the modification of these images using text only as means to offer intuitive and versatile editing. However, editing proves to be difficult for these generative models due to the inherent nature of editing techniques, which involves preserving certain content from the original image. Conversely, in text-based models, even minor modifications to the text prompt frequently result in an entirely distinct result, making attaining one-shot generation that accurately corresponds to the users intent exceedingly challenging. In addition, to edit a real image using these state-of-the-art tools, one must first invert the image into the pre-trained models domain - adding another factor affecting the edit quality, as well as latency. In this exploratory report, we propose LEDITS - a combined lightweight approach for real-image editing, incorporating the Edit Friendly DDPM inversion technique with Semantic Guidance, thus extending Semantic Guidance to real image editing, while harnessing the editing capabilities of DDPM inversion as well. This approach achieves versatile edits, both subtle and extensive as well as alterations in composition and style, while requiring no optimization nor extensions to the architecture. | ['Apolinário Passos', 'Linoy Tsaban'] | 2023-07-02 | null | null | null | null | ['image-generation'] | ['computer-vision'] | [ 6.71659172e-01 2.58317381e-01 4.18619722e-01 -3.03548336e-01
-3.58340681e-01 -6.94668055e-01 9.38463926e-01 -1.00908093e-01
-3.03205609e-01 6.58875048e-01 -5.36397770e-02 -2.39116922e-02
1.03122599e-01 -6.84159994e-01 -7.55541325e-01 -3.58478695e-01
3.35052848e-01 5.82173645e-01 1.05244428e-01 -4.34501618e-01
3.85440588e-01 5.62666059e-01 -1.48735881e+00 1.88694715e-01
1.00730491e+00 7.11085796e-01 7.31221735e-01 9.04958487e-01
-5.07616699e-01 7.39607871e-01 -7.63273060e-01 -8.75357389e-01
2.52545506e-01 -5.72600067e-01 -4.26771224e-01 4.00460213e-01
6.03486121e-01 -5.89968741e-01 -2.26363748e-01 7.90935338e-01
4.12798852e-01 1.09641969e-01 5.95213711e-01 -1.12416315e+00
-1.07392204e+00 3.01180512e-01 -5.41575491e-01 -2.82614380e-01
4.30798233e-01 3.98429185e-01 5.95633149e-01 -8.33739221e-01
1.02009106e+00 9.95732009e-01 4.61713523e-01 6.44903302e-01
-1.39304256e+00 -3.42340916e-01 1.00510933e-01 -7.22780228e-02
-1.12288260e+00 -7.92419910e-01 7.42811084e-01 -3.16649407e-01
9.99512434e-01 4.06698167e-01 8.92974496e-01 1.24666715e+00
2.67889977e-01 5.93827009e-01 9.40460205e-01 -6.41611338e-01
1.19945891e-01 4.96011883e-01 -6.50167525e-01 6.87981486e-01
-4.47435081e-02 -1.38255894e-01 -5.31671703e-01 2.05122843e-01
8.09965789e-01 1.02762267e-01 -1.48642719e-01 -4.39782917e-01
-1.17691541e+00 5.70672393e-01 1.48372620e-01 9.78698432e-02
-5.93317330e-01 4.13758963e-01 3.11817497e-01 3.90204757e-01
6.24464095e-01 5.82489848e-01 -7.25874826e-02 -3.23692232e-01
-1.48079228e+00 2.24951431e-01 7.22748935e-01 1.20897746e+00
8.11106980e-01 2.35165641e-01 -2.89443225e-01 6.79736972e-01
8.21337625e-02 4.02602643e-01 2.33432889e-01 -1.04723930e+00
2.94859439e-01 4.26414132e-01 4.91180196e-02 -1.12966716e+00
1.84362188e-01 -5.55642903e-01 -7.61440635e-01 5.98916769e-01
1.14156939e-01 3.96062881e-02 -1.02800810e+00 1.65814340e+00
2.62412041e-01 -2.76077896e-01 -3.86484772e-01 5.41723728e-01
3.87699664e-01 6.05474293e-01 4.52981181e-02 -8.74541029e-02
1.17078686e+00 -1.11492300e+00 -7.93352783e-01 -3.30740392e-01
2.79864162e-01 -1.10003662e+00 1.31173825e+00 3.89290422e-01
-1.42772961e+00 -3.44763458e-01 -1.01937556e+00 -2.73599714e-01
-4.14958417e-01 6.08572401e-02 5.16149580e-01 6.11840367e-01
-1.38912225e+00 6.70144260e-01 -5.92315912e-01 -5.14606535e-01
3.93336713e-01 1.19685374e-01 -3.17931920e-01 -8.84211138e-02
-8.48006189e-01 8.98559213e-01 2.59795070e-01 -1.18237704e-01
-8.33127201e-01 -1.02636707e+00 -6.32396638e-01 -6.30809665e-02
4.71288115e-01 -9.10843730e-01 1.26971900e+00 -1.28391218e+00
-1.78841758e+00 8.42911124e-01 -2.07250118e-01 -3.65945399e-01
1.24264085e+00 -2.11636588e-01 -2.35150531e-01 2.35786408e-01
1.47271335e-01 1.18993413e+00 1.46455503e+00 -1.31059802e+00
-1.96348801e-01 4.18339968e-02 3.01705211e-01 3.59784722e-01
-5.68264365e-01 -3.63854058e-02 -8.62045288e-01 -9.69395220e-01
-4.15013105e-01 -8.47902358e-01 -2.82874316e-01 5.83936632e-01
-4.62078810e-01 5.09743214e-01 1.16538000e+00 -8.81501138e-01
1.12189305e+00 -2.26252198e+00 1.06366523e-01 1.67019516e-01
4.12579060e-01 3.06768745e-01 -2.72988379e-01 6.91360176e-01
1.32075146e-01 1.09340660e-01 -3.50109726e-01 -9.19065714e-01
8.31416547e-02 1.71306148e-01 -3.84348452e-01 -2.25799647e-03
1.92356661e-01 1.18074501e+00 -7.27990031e-01 -4.81240869e-01
4.25979674e-01 6.84774399e-01 -7.11414874e-01 1.79950267e-01
-5.32275081e-01 3.73537123e-01 -6.72696233e-02 3.81870389e-01
6.49915218e-01 -2.01153889e-01 2.57217083e-02 -2.70386726e-01
-7.53086507e-02 -2.56838113e-01 -1.08886421e+00 2.02204251e+00
-8.26966345e-01 7.70547509e-01 9.65990797e-02 -5.77298284e-01
7.61775315e-01 3.46201472e-02 2.12131470e-01 -7.81251967e-01
-1.96317822e-01 1.40770078e-01 -4.36254382e-01 -2.56087214e-01
1.08050871e+00 -4.68330011e-02 3.36577863e-01 7.79628992e-01
-1.13405883e-01 -5.66257536e-01 4.46292937e-01 7.03197598e-01
7.26881981e-01 5.61756432e-01 1.07780300e-01 -6.08636590e-04
8.58189836e-02 -1.71923507e-02 -5.30140512e-02 8.81452978e-01
2.88915694e-01 8.76756668e-01 3.38675708e-01 -1.02660999e-01
-1.34043300e+00 -9.98544455e-01 3.00204337e-01 8.67328286e-01
-6.25750497e-02 -4.99850631e-01 -1.18234587e+00 -4.67197329e-01
-3.10864568e-01 1.03550744e+00 -6.15751088e-01 -1.02053508e-01
-2.76837796e-01 -3.92473787e-01 5.58449209e-01 2.84306228e-01
5.76187253e-01 -9.51066315e-01 -8.42618048e-01 4.12357301e-01
-2.39970423e-02 -1.04739475e+00 -9.07285750e-01 -1.47774756e-01
-7.43716717e-01 -5.56760669e-01 -1.04445219e+00 -5.00021935e-01
8.66495013e-01 3.62158060e-01 1.09664989e+00 2.23822743e-01
-6.63707256e-01 5.44073105e-01 -3.23844016e-01 -2.30348900e-01
-7.23279297e-01 -7.16153458e-02 -3.82386953e-01 4.13917825e-02
-3.84632826e-01 -7.03289211e-01 -6.95000410e-01 1.19095050e-01
-1.38896561e+00 6.53879702e-01 6.24161243e-01 5.77273428e-01
4.26873237e-01 -1.38049468e-01 3.59833360e-01 -9.75736439e-01
8.23678970e-01 -1.84176303e-02 -2.69389957e-01 3.76376510e-01
-7.85836101e-01 -6.97101504e-02 7.46368825e-01 -4.70370144e-01
-1.40756822e+00 -1.62870690e-01 2.93187089e-02 -5.25112212e-01
8.24838802e-02 2.77454913e-01 9.92376264e-03 -1.49623960e-01
6.92372441e-01 5.89782536e-01 2.15500891e-01 -2.85875112e-01
9.17272329e-01 4.52838480e-01 4.67726111e-01 -3.83732259e-01
8.76381934e-01 5.58799386e-01 -2.06820652e-01 -9.62190628e-01
-4.25411642e-01 1.81452066e-01 -5.55780113e-01 -3.17915499e-01
7.69967914e-01 -7.55022585e-01 -6.96880594e-02 6.26698077e-01
-1.25846064e+00 -5.05816996e-01 -5.60963511e-01 -1.43892765e-01
-6.35294735e-01 5.64747274e-01 -4.67233419e-01 -4.23109978e-01
-3.69961023e-01 -1.33979368e+00 9.95785177e-01 6.17049374e-02
-5.16509712e-01 -1.10091567e+00 -2.39193678e-01 2.40079060e-01
1.00837505e+00 2.28308365e-01 8.70488405e-01 -1.59341708e-01
-8.27359319e-01 -2.44910687e-01 -4.18602258e-01 2.67487764e-01
1.38634041e-01 2.68791735e-01 -7.93502212e-01 -2.55404264e-01
-2.77226776e-01 -1.58585429e-01 4.30918694e-01 1.94456920e-01
1.10603249e+00 -2.47466803e-01 -1.02162205e-01 6.68495238e-01
1.31421363e+00 1.81595474e-01 8.53197992e-01 2.74590611e-01
7.01217592e-01 4.31692541e-01 3.86960477e-01 5.68382919e-01
2.26661846e-01 8.04868996e-01 2.96709388e-01 -2.65359342e-01
-4.60066199e-01 -4.85200822e-01 2.90940762e-01 5.94805956e-01
-2.40703747e-02 -3.34260643e-01 -4.81858164e-01 3.54862630e-01
-1.59076869e+00 -1.03163230e+00 -3.06994072e-03 2.11638641e+00
9.43847656e-01 9.20307450e-03 -2.76652873e-01 -2.06842169e-01
5.60481608e-01 4.12960351e-01 -5.68360269e-01 -6.02913558e-01
-8.07187632e-02 1.97323918e-01 2.90942878e-01 4.38282907e-01
-5.92295110e-01 1.07035792e+00 6.23663282e+00 1.03164899e+00
-1.39310205e+00 2.88354963e-01 4.85592753e-01 -4.48172808e-01
-6.25454307e-01 9.19295028e-02 -3.63408148e-01 5.14733076e-01
7.14953661e-01 -2.48936921e-01 8.39095473e-01 6.56031668e-01
4.85872418e-01 -2.64843285e-01 -9.81976867e-01 1.01504886e+00
3.58307630e-01 -1.55484653e+00 5.71713626e-01 1.11150801e-01
7.95609772e-01 -4.36981380e-01 2.63425022e-01 -2.30734283e-03
2.62913853e-02 -8.34989369e-01 1.17777658e+00 7.40375280e-01
1.34518778e+00 -6.24050558e-01 8.71850178e-02 2.14025036e-01
-8.47750306e-01 3.38471949e-01 -2.14691475e-01 2.09025145e-01
4.84504074e-01 8.82452965e-01 -7.02426195e-01 3.99141133e-01
4.60041076e-01 6.26094878e-01 -6.28834665e-01 6.38065517e-01
-1.37513757e-01 4.00264934e-02 -1.98919222e-01 1.72306076e-01
1.30197808e-01 -3.30976456e-01 5.63605845e-01 1.40097690e+00
5.95443130e-01 -2.75981754e-01 -1.25303358e-01 1.21334898e+00
-2.70049334e-01 1.00575112e-01 -6.04206443e-01 -4.48466659e-01
3.33890229e-01 1.34863281e+00 -7.65173316e-01 -4.99444157e-01
-1.15597233e-01 1.89430523e+00 2.02780485e-01 4.06874835e-01
-1.07999444e+00 -5.49363077e-01 4.68434960e-01 2.65485078e-01
2.37596929e-01 -4.43828791e-01 -2.39800662e-01 -1.07478499e+00
1.86699435e-01 -1.04972613e+00 -3.73580486e-01 -1.09736645e+00
-7.94502020e-01 6.36965096e-01 -1.55397996e-01 -9.99384165e-01
-2.32922927e-01 -1.32447794e-01 -6.33524954e-01 9.95395839e-01
-1.39566314e+00 -1.46336222e+00 -5.46072960e-01 4.92404222e-01
8.74168873e-01 -1.11973226e-01 6.19754553e-01 5.00007570e-01
-2.73539424e-01 6.63989902e-01 1.25802457e-01 -4.64878231e-01
8.23861480e-01 -1.08352637e+00 6.70579970e-01 8.85052681e-01
1.59323320e-01 7.46744037e-01 9.06074226e-01 -7.55157113e-01
-1.46767318e+00 -1.14199889e+00 7.20598042e-01 -2.43540332e-01
5.03973484e-01 -3.69333357e-01 -6.38440967e-01 6.23655498e-01
6.37289643e-01 -4.50995058e-01 3.96427721e-01 -4.97819453e-01
-1.22432269e-01 9.21146423e-02 -1.10111833e+00 1.09900641e+00
1.31979012e+00 -7.15987086e-01 -1.57067224e-01 2.89827883e-01
6.90135419e-01 -4.84543532e-01 -5.68202198e-01 -2.24606931e-01
5.03666401e-01 -1.10367084e+00 9.12281454e-01 -1.46664217e-01
8.41888309e-01 -3.35676551e-01 3.60026360e-02 -1.59604704e+00
-2.50355512e-01 -1.01812267e+00 -2.36631818e-02 1.33095038e+00
4.43593025e-01 -4.71504450e-01 6.15134895e-01 8.12732577e-01
-7.37003759e-02 -4.58394974e-01 -5.13774514e-01 -5.82888246e-01
-4.63781685e-01 -5.49503148e-01 5.98036706e-01 9.00049508e-01
-4.90088642e-01 -3.13210376e-02 -6.73805177e-01 -3.09462398e-01
6.06329918e-01 -2.13594079e-01 1.01429081e+00 -7.52137601e-01
-6.26713514e-01 -5.23639321e-01 -1.89739749e-01 -1.13205016e+00
-1.77024856e-01 -7.25032628e-01 -1.64723903e-01 -1.63943517e+00
1.38478726e-01 -3.90708566e-01 3.97174835e-01 2.32949764e-01
4.42255996e-02 5.70221245e-01 6.18296325e-01 3.29509437e-01
-3.43893707e-01 5.90724170e-01 1.52553904e+00 -1.30672574e-01
-1.08654283e-01 -3.74519229e-01 -7.98791647e-01 4.81995642e-01
5.18473983e-01 -3.16219181e-01 -9.25153077e-01 -7.30156898e-01
4.60311621e-01 -1.74309611e-01 5.12351632e-01 -8.83373380e-01
2.30733454e-01 -1.15778558e-01 3.28868717e-01 -3.39521706e-01
5.29756010e-01 -7.53174961e-01 7.25462496e-01 2.36435041e-01
-3.59811842e-01 1.85527787e-01 1.74662843e-01 6.46195173e-01
-4.73114885e-02 -3.57609302e-01 7.76230097e-01 -3.66940409e-01
-8.43826473e-01 2.22963572e-01 -2.44139820e-01 -2.92504225e-02
1.25732756e+00 -6.44608617e-01 -1.97331056e-01 -6.99459970e-01
-7.63464928e-01 -1.82156309e-01 9.47858691e-01 3.96437556e-01
6.24612033e-01 -9.11174238e-01 -4.91412073e-01 3.61304134e-01
-4.66835462e-02 -1.96480796e-01 5.80784619e-01 8.50929201e-01
-8.96865487e-01 3.65250744e-02 -2.88611919e-01 -4.33626860e-01
-1.15659118e+00 4.01781976e-01 2.43730232e-01 -1.82921469e-01
-7.96119511e-01 6.41164362e-01 3.01529139e-01 -3.16797674e-01
-1.75784416e-02 8.61647725e-02 3.54995608e-01 2.36597195e-01
5.84725499e-01 1.60106003e-01 1.85758859e-01 -6.15264997e-02
8.42266232e-02 4.30562407e-01 -4.92870897e-01 -4.05157357e-01
1.14850140e+00 -5.41981339e-01 -6.93762004e-02 1.30280266e-02
8.78197312e-01 1.32862166e-01 -1.48478866e+00 -3.37610431e-02
-4.08989638e-01 -6.66590631e-01 1.18486345e-01 -1.09278905e+00
-1.06444120e+00 9.64654982e-01 3.85346323e-01 3.47949676e-02
1.01411390e+00 -4.09233093e-01 9.51826632e-01 2.06240103e-01
3.61977965e-01 -1.27116287e+00 2.72564650e-01 2.57870317e-01
1.10214448e+00 -9.46784735e-01 -7.94429332e-03 -1.74126133e-01
-9.23681855e-01 9.96289670e-01 3.26329082e-01 1.47656858e-01
2.87184685e-01 3.93384814e-01 7.97866285e-02 -4.94915135e-02
-6.00109696e-01 3.53086948e-01 -1.61560364e-02 7.38549232e-01
3.22387099e-01 -1.08887233e-01 -1.15261815e-01 -2.10667625e-01
-1.03922464e-01 2.78643847e-01 7.72355855e-01 9.72244859e-01
-1.45055726e-01 -1.20129573e+00 -2.07189918e-01 5.03104270e-01
-1.60526961e-01 -4.44197237e-01 -2.96458721e-01 6.54244423e-01
-1.52054429e-01 7.20230818e-01 6.32967800e-02 -3.04123908e-02
2.64742434e-01 1.88528728e-02 5.36741078e-01 -4.79308277e-01
-5.96137524e-01 -2.16970723e-02 9.08447355e-02 -6.03569686e-01
-2.35026389e-01 -3.91107470e-01 -1.05255389e+00 -6.17981553e-01
-1.06447071e-01 -2.20963359e-01 1.00704348e+00 8.99896026e-01
7.66505063e-01 6.10949278e-01 3.29628885e-01 -1.33446836e+00
-4.78615224e-01 -7.09062755e-01 -5.76728702e-01 4.54484135e-01
-4.36738655e-02 -3.85347813e-01 -9.60753933e-02 3.79288018e-01] | [11.399173736572266, -0.3484325408935547] |
5a00b1eb-43dd-44c1-9999-a9720bed33fa | transfusion-a-novel-slam-method-focused-on | null | null | http://openaccess.thecvf.com//content/ICCV2021/html/Zhu_Transfusion_A_Novel_SLAM_Method_Focused_on_Transparent_Objects_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Zhu_Transfusion_A_Novel_SLAM_Method_Focused_on_Transparent_Objects_ICCV_2021_paper.pdf | Transfusion: A Novel SLAM Method Focused on Transparent Objects | Recently RGB-D sensors have become very popular in the area of Simultaneous Localisation and Mapping (SLAM). The RGB-D SLAM approach relies heavily on the accuracy of the input depth map. However, refraction and reflection of transparent objects will result in false depth input of RGB-D cameras, which makes the traditional RGB-D SLAM algorithm unable to work correctly in the presence of transparent objects. In this paper, we propose a novel SLAM approach called transfusion that allows transparent object existence and recovery in the video input. Our method is composed of two parts. Transparent Objects Cut Iterative Closest Points (TC-ICP)is first used to recover camera pose, detecting and removing transparent objects from input to reduce the trajectory errors. Then Transparent Objects Reconstruction (TO-Reconstruction) is used to reconstruct the transparent objects and opaque objects separately. The opaque objects are reconstructed with the traditional method, and the transparent objects are reconstructed with the visual hull-based method. To evaluate our algorithm, we construct a new RGB-D SLAM database containing 25 video sequences. Each sequence has at least one transparent object. Experiments show that our approach can work adequately in scenes contain transparent objects while the existing approach can not handle them. Our approach significantly improves the accuracy of the camera trajectory and the quality of environment reconstruction. | ['Bo Ren', 'Jiaxiong Qiu', 'Yifan Zhu'] | 2021-01-01 | null | null | null | iccv-2021-1 | ['transparent-objects'] | ['computer-vision'] | [ 1.14563711e-01 -2.15899020e-01 3.93859029e-01 -1.77977756e-01
-3.96569908e-01 -6.08108342e-01 2.86572874e-01 -2.11220294e-01
-5.10277152e-01 4.84617323e-01 -1.42430022e-01 -1.25281617e-01
1.19668514e-01 -8.25110495e-01 -7.48781145e-01 -5.36850810e-01
1.76176012e-01 6.80422425e-01 8.18739831e-01 -1.54780865e-01
3.50681573e-01 6.68933511e-01 -1.43789792e+00 3.14173242e-03
7.52073228e-01 1.01994610e+00 6.48479223e-01 6.85897291e-01
-3.99722070e-01 4.68679398e-01 -3.26001197e-01 -1.44791692e-01
7.37963140e-01 -2.32866362e-01 -5.25689960e-01 2.09579870e-01
2.44643256e-01 -6.08453572e-01 -1.89083427e-01 9.62167621e-01
3.29971403e-01 -2.04512402e-02 2.46933714e-01 -1.34173167e+00
6.27457649e-02 -2.99606770e-01 -8.53212297e-01 -3.31196189e-01
8.85929465e-01 -1.67653307e-01 1.32944599e-01 -1.06816196e+00
6.73049629e-01 1.19069588e+00 8.85889173e-01 3.16019505e-01
-7.06518590e-01 -5.40833533e-01 3.10441176e-03 1.31306946e-01
-1.61351717e+00 -4.52148885e-01 6.61909878e-01 -3.71952295e-01
7.42595315e-01 4.36939687e-01 8.73250306e-01 3.73364985e-01
2.07649246e-01 3.78462791e-01 9.14665878e-01 -4.63947803e-01
1.98336646e-01 2.15842947e-01 -2.60115534e-01 7.57486999e-01
4.59104806e-01 7.57875964e-02 -5.89050651e-01 3.39519531e-02
9.32053328e-01 3.29584926e-01 -5.52329242e-01 -9.47433949e-01
-1.40891492e+00 5.17763674e-01 5.30245483e-01 -9.10326764e-02
-2.08949998e-01 1.57092422e-01 -1.17590511e-03 1.53318912e-01
3.59124899e-01 -9.50789154e-02 -2.45152012e-01 -2.37224475e-01
-7.50679195e-01 -6.11622818e-02 4.89662021e-01 1.30825138e+00
1.07343662e+00 -3.02869320e-01 7.81864107e-01 2.38301471e-01
5.79273403e-01 7.89980233e-01 1.97662234e-01 -9.14953530e-01
8.08910728e-01 8.07286084e-01 4.31999952e-01 -1.19515717e+00
-3.93454611e-01 2.75057316e-01 -5.49524903e-01 5.10461330e-01
1.44291475e-01 2.11493224e-01 -8.41258883e-01 9.35243785e-01
7.93492913e-01 1.27631426e-01 1.46018893e-01 1.12156606e+00
7.00530410e-01 6.93130076e-01 -7.76708543e-01 -1.85308665e-01
8.74766588e-01 -8.17843020e-01 -7.08732069e-01 -3.00918132e-01
5.33348262e-01 -1.02350390e+00 7.40194142e-01 5.13142288e-01
-7.97633469e-01 -3.63510430e-01 -9.58411276e-01 -1.13907985e-01
-1.40965000e-01 6.02330938e-02 5.09516478e-01 5.36486328e-01
-1.01208150e+00 1.89518720e-01 -8.72909844e-01 -5.27596474e-01
-9.92063507e-02 5.52686036e-01 -7.23350883e-01 -3.02436560e-01
-6.83523595e-01 9.66435134e-01 5.67336261e-01 4.48383719e-01
-6.53300643e-01 -1.73359886e-01 -7.92245924e-01 -5.92910409e-01
4.64011461e-01 -5.92397869e-01 7.82707870e-01 -8.33427131e-01
-1.55543554e+00 8.80234063e-01 -3.16754520e-01 -3.29838246e-02
7.47082233e-01 -4.24605459e-01 -1.43065760e-02 2.61460930e-01
6.50754347e-02 3.62414777e-01 5.47951639e-01 -1.75251436e+00
-6.68998897e-01 -5.24956763e-01 -1.35243446e-01 5.63254416e-01
3.09442967e-01 -1.98562220e-01 -9.15588975e-01 8.44478235e-02
1.18615103e+00 -1.07111156e+00 -1.69845194e-01 4.67435628e-01
-2.51231343e-01 4.90878522e-01 9.41568434e-01 -5.97898304e-01
7.04516888e-01 -2.21454620e+00 1.61736444e-01 4.03200626e-01
1.07414640e-01 -7.35179186e-02 2.09628746e-01 4.42265987e-01
3.32811862e-01 -2.25564495e-01 -1.68790609e-01 -7.04167724e-01
-3.58543634e-01 3.84098858e-01 -2.67994523e-01 9.17600095e-01
-5.73307931e-01 2.81480730e-01 -6.60508692e-01 -6.73197806e-01
5.27927041e-01 5.21582842e-01 -3.58565301e-01 2.07946315e-01
7.00666010e-02 5.61910510e-01 -4.32860434e-01 7.52231121e-01
1.16471255e+00 3.64037633e-01 -1.33217886e-01 -7.85853937e-02
-5.54923475e-01 4.42083590e-02 -1.65754950e+00 1.99242139e+00
-3.75714093e-01 4.79604423e-01 3.79103303e-01 -1.32549912e-01
9.48030531e-01 1.87977880e-01 4.62214917e-01 -5.55148602e-01
-5.40133491e-02 5.62128305e-01 -4.74476308e-01 -4.99580353e-01
7.01956809e-01 3.53491083e-02 1.98614046e-01 3.15165639e-01
-5.43040752e-01 -5.59592009e-01 -4.59512055e-01 1.92755193e-01
8.95232320e-01 6.42954767e-01 3.16269398e-01 1.75548896e-01
4.21916544e-01 2.99278885e-01 7.23490298e-01 2.95743316e-01
5.21224737e-03 6.89562201e-01 -2.09323138e-01 -5.34995973e-01
-9.00870562e-01 -1.16436565e+00 6.18946925e-02 3.55003297e-01
1.14474988e+00 -2.51208186e-01 -4.64039415e-01 -3.92058700e-01
1.46797493e-01 2.46043414e-01 -2.45595232e-01 2.14300118e-02
-5.57412624e-01 -4.03545737e-01 -5.45154586e-02 1.46723375e-01
7.66900778e-01 -5.87348819e-01 -1.06466889e+00 9.16864350e-03
-4.47886854e-01 -1.06106675e+00 -1.64168954e-01 -5.07425237e-03
-1.11841738e+00 -1.06976533e+00 -3.76364291e-01 -4.55030501e-01
9.59527254e-01 9.26184654e-01 4.87516791e-01 2.09184319e-01
-3.11531425e-02 3.83277148e-01 -4.56586361e-01 -3.27663332e-01
-2.40512639e-01 -5.53168774e-01 3.97106230e-01 8.99280515e-03
9.85839814e-02 -3.04971695e-01 -5.01935720e-01 6.85280263e-01
-5.27963817e-01 3.96913528e-01 4.63471115e-01 6.86863661e-02
7.60651112e-01 2.07282603e-02 -4.97353673e-01 -4.10092026e-01
-2.18900964e-01 -1.78092048e-01 -9.46244299e-01 -5.82911707e-02
-2.29135513e-01 -2.87368298e-01 1.68081179e-01 -2.73247302e-01
-8.72456729e-01 7.31023848e-01 2.57826120e-01 -6.06874168e-01
6.02612197e-02 7.11946487e-02 -4.31985140e-01 -6.23092115e-01
2.46871561e-01 2.85540402e-01 -9.52795222e-02 -6.05829358e-01
5.65197021e-02 7.99264610e-01 5.42365670e-01 -9.47990268e-02
1.09958589e+00 8.50142777e-01 1.58636153e-01 -7.95378149e-01
-3.65566373e-01 -7.54459858e-01 -9.21630561e-01 -5.09153724e-01
8.21210384e-01 -9.48197126e-01 -6.58743382e-01 4.43926871e-01
-1.37174642e+00 -7.12666959e-02 6.93197846e-02 8.65794897e-01
-6.01569653e-01 6.52310014e-01 -3.42810005e-01 -9.93731141e-01
8.03553388e-02 -1.41584122e+00 1.27748239e+00 -8.85370895e-02
2.53194541e-01 -6.08005345e-01 2.28580907e-01 2.85820574e-01
-3.67860384e-02 3.46634030e-01 1.75958976e-01 2.75767297e-01
-1.16228521e+00 -3.15181404e-01 -7.63051286e-02 -2.25375399e-01
1.87488422e-01 -9.04021934e-02 -9.66263473e-01 -2.75049835e-01
2.45163083e-01 2.39617288e-01 5.78537166e-01 4.21472676e-02
5.05546808e-01 -8.94309878e-02 -7.10389972e-01 1.01252747e+00
1.86749923e+00 2.94158250e-01 7.74599850e-01 7.96283543e-01
1.00829828e+00 5.40249109e-01 1.03492677e+00 3.14442396e-01
4.46594834e-01 9.61412549e-01 9.98050630e-01 -1.32116944e-01
1.84296712e-01 -2.19919965e-01 4.65593636e-01 8.97250652e-01
-3.69638324e-01 -2.83225328e-01 -1.09619331e+00 1.88837022e-01
-1.84983289e+00 -6.25810504e-01 -8.24046612e-01 2.68882132e+00
4.87760991e-01 4.75232862e-02 -2.33943388e-01 2.73074836e-01
5.20277321e-01 -2.62109220e-01 -1.28323361e-01 -1.73620403e-01
7.40549192e-02 -4.73034620e-01 9.58598733e-01 8.24301898e-01
-7.33415723e-01 9.22809780e-01 5.13870955e+00 1.56899333e-01
-7.79216290e-01 2.39874989e-01 -6.31738245e-01 -2.41192076e-02
-2.79448539e-01 4.63277727e-01 -8.05235207e-01 3.51612687e-01
3.69584024e-01 4.30437744e-01 2.52360582e-01 7.58921921e-01
2.50133574e-01 -9.86980319e-01 -9.08085406e-01 1.45526266e+00
2.52624571e-01 -8.72625589e-01 -7.80168027e-02 2.69223690e-01
5.85653007e-01 -2.70014722e-02 -4.51535553e-01 -3.64819646e-01
-1.07411936e-01 -4.06353056e-01 1.14575350e+00 6.53007567e-01
6.31056547e-01 -6.31746709e-01 9.27917838e-01 7.45119750e-01
-1.34920430e+00 1.11133747e-01 -6.44655645e-01 -1.52429029e-01
3.58523488e-01 6.42409503e-01 -9.62872207e-01 7.34841526e-01
8.03890526e-01 3.89343798e-01 -4.89772409e-01 1.40380442e+00
-6.30324483e-02 -5.95377982e-02 -6.70284808e-01 1.83942363e-01
7.57891536e-02 -5.46635687e-01 5.50444245e-01 6.55355871e-01
4.14690524e-01 2.36832261e-01 3.09536755e-01 5.15363336e-01
3.29499573e-01 4.23314013e-02 -8.37545931e-01 6.99998498e-01
3.64735276e-01 7.89713085e-01 -9.67204452e-01 -1.93792343e-01
-3.91161293e-01 1.59036684e+00 -1.15420641e-02 2.55611718e-01
-8.80652070e-01 -2.20654726e-01 3.40573609e-01 1.80871904e-01
-3.83153595e-02 -7.08996415e-01 -1.54405043e-01 -1.12659419e+00
2.80748904e-01 -3.01208556e-01 -1.07948005e-01 -1.24275267e+00
-5.11116087e-01 5.09268463e-01 -9.66096893e-02 -1.78994930e+00
2.09545091e-01 -4.28097457e-01 -2.53785457e-02 8.55954587e-01
-1.36610019e+00 -1.06426442e+00 -9.22172010e-01 8.03836048e-01
4.90971059e-01 3.51144254e-01 5.93585491e-01 2.29185134e-01
-8.22068304e-02 -1.82205737e-01 2.92949170e-01 -1.45694807e-01
7.16046095e-01 -8.71975839e-01 1.05471276e-01 1.04995978e+00
6.44385591e-02 4.48532492e-01 6.73985302e-01 -1.10103333e+00
-1.76077497e+00 -8.15595865e-01 5.77921212e-01 -7.39640951e-01
3.54464985e-02 -7.00119495e-01 -5.71925581e-01 9.51656282e-01
-4.76447970e-01 -2.90856492e-02 1.74742043e-01 -5.48963368e-01
2.34297626e-02 -1.84518978e-01 -1.32523882e+00 3.32464576e-01
1.07015443e+00 -4.46128249e-01 -5.27955532e-01 4.22475010e-01
7.72045732e-01 -1.00166297e+00 -4.51679796e-01 3.31577897e-01
5.37419915e-01 -1.32177496e+00 9.98088121e-01 3.09480995e-01
-1.74453527e-01 -8.61839533e-01 -4.12306219e-01 -8.90211999e-01
2.14274526e-01 -4.50706452e-01 2.45526489e-02 9.40681934e-01
-2.33261697e-02 -6.90238178e-01 6.90424681e-01 4.44133252e-01
-3.10075313e-01 -2.41411284e-01 -1.18308866e+00 -6.94715679e-01
-7.72580564e-01 -6.40021443e-01 6.15075946e-01 7.16789186e-01
-3.07883829e-01 -1.09690875e-01 -3.45345289e-01 6.99715078e-01
8.50387454e-01 3.14120650e-01 1.24201500e+00 -1.29550862e+00
2.11689863e-02 1.40535578e-01 -7.15370893e-01 -1.14434743e+00
-3.31898957e-01 -4.05715078e-01 3.76540333e-01 -1.89598143e+00
8.05263072e-02 -8.29991162e-01 5.65047145e-01 1.37705401e-01
2.02898607e-01 3.64801049e-01 2.09740520e-01 7.02364206e-01
-6.70830011e-01 3.99459302e-01 1.06716263e+00 3.24972004e-01
-4.73861217e-01 1.43756360e-01 2.54528016e-01 1.01469386e+00
4.82830584e-01 -5.80541432e-01 -2.79686481e-01 -7.91452885e-01
4.55380559e-01 3.38785231e-01 4.24248517e-01 -1.23913336e+00
2.35101268e-01 -1.96468800e-01 4.95490938e-01 -1.10050118e+00
7.28296399e-01 -1.51472402e+00 6.49592340e-01 5.69316745e-01
5.41107357e-01 1.98569521e-01 1.07899174e-01 5.95533729e-01
-1.51146399e-02 -3.32317352e-01 4.27023262e-01 -3.87267053e-01
-9.44554985e-01 3.81513536e-01 -1.29634425e-01 -5.39849877e-01
1.30205548e+00 -9.66493487e-01 6.70584813e-02 -3.26954961e-01
-4.56537217e-01 6.83204010e-02 1.21941531e+00 2.93812841e-01
1.15508306e+00 -1.23727167e+00 -3.19858879e-01 3.73162508e-01
2.08795160e-01 6.56257272e-01 -6.81881234e-03 1.02442873e+00
-1.43536270e+00 2.09833771e-01 1.37029111e-03 -9.76572216e-01
-1.47090065e+00 6.57912195e-01 3.82447481e-01 4.62105364e-01
-7.43424237e-01 7.59247661e-01 1.94588900e-01 -4.84603405e-01
2.47361109e-01 -4.37196463e-01 2.29493305e-01 -3.92604083e-01
3.76393735e-01 5.43040276e-01 1.33814678e-01 -1.03434587e+00
-6.44240677e-01 1.31968594e+00 4.69275236e-01 -3.91880691e-01
1.09902263e+00 -7.96495080e-01 -3.89086485e-01 6.26777947e-01
1.10455072e+00 5.16333699e-01 -1.10200334e+00 1.08809739e-01
-1.94210336e-01 -9.00748909e-01 8.80725831e-02 -3.14956009e-01
-7.90946245e-01 9.49160814e-01 7.59744227e-01 -2.89953619e-01
1.12896514e+00 -1.41621143e-01 7.34914720e-01 3.82415533e-01
1.31502807e+00 -7.38298357e-01 -2.61301696e-01 4.22131538e-01
6.80421472e-01 -1.16000092e+00 3.65676582e-01 -7.19101012e-01
-5.38435400e-01 1.17693377e+00 6.11718655e-01 1.91279054e-01
1.86115548e-01 3.93829107e-01 1.73720732e-01 -1.63106188e-01
7.67110940e-03 -7.22261593e-02 -1.36590183e-01 5.81720769e-01
-2.49455631e-01 -2.17122644e-01 1.28351524e-01 -7.02201650e-02
-1.98305205e-01 -6.03699684e-02 6.80039525e-01 1.30938113e+00
-8.19187105e-01 -7.68689752e-01 -1.00915742e+00 -1.84507981e-01
1.41830802e-01 1.85747877e-01 -5.03584266e-01 9.60898697e-01
3.39583278e-01 8.45754206e-01 7.58030191e-02 -5.61231136e-01
3.80941361e-01 -2.48689666e-01 7.03675747e-01 -5.49072385e-01
-2.32776806e-01 1.57187670e-01 -1.65188998e-01 -1.10041714e+00
-5.77577114e-01 -4.87491012e-01 -1.73570001e+00 -2.86089689e-01
-6.19844258e-01 1.19809106e-01 1.15090871e+00 6.49070680e-01
8.09476152e-03 -2.32065216e-01 7.11905479e-01 -1.05643058e+00
-7.64766112e-02 -5.47988951e-01 -5.75612843e-01 1.33517966e-01
6.76291883e-01 -7.36261785e-01 -4.41513360e-01 3.82445604e-02] | [7.449432849884033, -2.2115674018859863] |
b24e816e-da1d-414e-b775-eedcc3e006ab | string-kernels-for-native-language | null | null | https://aclanthology.org/J16-3005 | https://aclanthology.org/J16-3005.pdf | String Kernels for Native Language Identification: Insights from Behind the Curtains | null | ['Marius Popescu', 'Radu Tudor Ionescu', 'Aoife Cahill'] | 2016-09-01 | null | null | null | cl-2016-9 | ['native-language-identification'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.353001117706299, 3.7478058338165283] |
4ffde5b7-2a9e-4812-82f1-ca15f4fdd2fc | variational-community-partition-with-novel | 1811.04543 | null | http://arxiv.org/abs/1811.04543v1 | http://arxiv.org/pdf/1811.04543v1.pdf | Variational Community Partition with Novel Network Structure Centrality Prior | In this paper, we proposed a novel two-stage optimization method for network
community partition, which is based on inherent network structure information.
The introduced optimization approach utilizes the new network centrality
measure of both links and vertices to construct the key affinity description of
the given network, where the direct similarities between graph nodes or nodal
features are not available to obtain the classical affinity matrix. Indeed,
such calculated network centrality information presents the essential structure
of network, hence, the proper measure for detecting network communities, which
also introduces a `confidence' criterion for referencing new labeled benchmark
nodes. For the resulted challenging combinatorial optimization problem of graph
clustering, the proposed optimization method iteratively employs an efficient
convex optimization algorithm which is developed based under a new variational
perspective of primal and dual. Experiments over both artificial and real-world
network datasets demonstrate that the proposed optimization strategy of
community detection significantly improves result accuracy and outperforms the
state-of-the-art algorithms in terms of accuracy and reliability. | ['SanYang Liu', 'Yiguang Bai', 'Ke Yin', 'Jing Yuan'] | 2018-11-12 | null | null | null | null | ['network-community-partition'] | ['graphs'] | [ 1.19814843e-01 9.05107558e-02 -1.82679847e-01 1.15404032e-01
-3.24730426e-01 -6.37313843e-01 4.56301779e-01 6.64541781e-01
-3.74243349e-01 6.01264834e-01 -2.71741390e-01 -1.41738981e-01
-8.26713264e-01 -9.52783406e-01 -3.31902385e-01 -8.98910105e-01
-4.20799166e-01 6.84283435e-01 2.05937073e-01 1.63257532e-02
3.17585379e-01 3.65998179e-01 -1.26144910e+00 -2.36026287e-01
9.65855420e-01 8.41882229e-01 3.37377489e-02 3.75231743e-01
-8.14207569e-02 5.30312121e-01 -2.15887070e-01 -4.06845927e-01
2.35219911e-01 -2.49408886e-01 -7.47165859e-01 4.10214752e-01
-3.02233547e-01 4.04404879e-01 -1.07542008e-01 1.24348867e+00
3.44637722e-01 7.35293850e-02 6.58515513e-01 -1.42464697e+00
-2.79177725e-01 8.29906881e-01 -9.26716149e-01 3.20036441e-01
5.07394791e-01 -9.92687121e-02 1.48882103e+00 -7.56200790e-01
8.14226329e-01 9.14522231e-01 5.86218357e-01 -1.42383054e-01
-1.48574960e+00 -3.35027218e-01 1.74272716e-01 2.31675029e-01
-1.79395211e+00 -9.55160633e-02 1.14267087e+00 -6.53506398e-01
2.57168144e-01 3.86634260e-01 8.70390892e-01 5.56444943e-01
-9.82509851e-02 3.25571924e-01 1.08453119e+00 -2.30153039e-01
1.07708611e-01 3.74206066e-01 2.93042570e-01 8.71290445e-01
5.92468679e-01 -1.70301154e-01 -1.64002806e-01 -2.99095720e-01
3.84194285e-01 -1.89853072e-01 -3.19622844e-01 -8.06270540e-01
-1.28471756e+00 9.11351979e-01 6.60662830e-01 5.86672008e-01
-4.28570539e-01 -1.16089761e-01 2.72236228e-01 4.26202528e-02
5.36142647e-01 2.15248898e-01 1.46602407e-01 3.31028908e-01
-8.48698616e-01 -1.73192903e-01 9.91306543e-01 6.71904981e-01
8.51507962e-01 -1.58410370e-01 5.10561019e-02 4.61710930e-01
6.15346432e-01 3.25981289e-01 -2.08013296e-01 -5.70590317e-01
4.54605073e-01 1.16305411e+00 -1.41493455e-01 -1.87260163e+00
-4.69427824e-01 -8.29419136e-01 -1.28701270e+00 -2.19992399e-01
3.54620606e-01 -1.17809474e-02 -1.98679790e-01 1.54872322e+00
6.95055306e-01 3.05888265e-01 -1.12960674e-01 8.83875608e-01
7.74486005e-01 3.58460277e-01 -2.86791831e-01 -5.94270408e-01
1.19684947e+00 -6.10500515e-01 -6.27143979e-01 2.46356189e-01
2.75304317e-01 -6.36707723e-01 4.45288271e-01 1.69978932e-01
-9.60541964e-01 -1.81610212e-01 -1.05428326e+00 6.17294669e-01
-2.75708526e-01 6.82240203e-02 7.50305891e-01 4.91480500e-01
-1.05926645e+00 6.13927603e-01 -4.86503452e-01 -5.04316986e-01
1.28289118e-01 4.21026736e-01 -5.09024918e-01 9.19921473e-02
-9.80287373e-01 5.61355770e-01 5.96825361e-01 4.91655201e-01
-7.01911926e-01 -3.17545146e-01 -6.79063797e-01 2.86417991e-01
7.91921318e-01 -5.95278382e-01 1.94639042e-01 -8.50517988e-01
-1.12561429e+00 8.94917250e-01 1.57490477e-01 -1.87681630e-01
5.77904046e-01 5.70548952e-01 -2.44918272e-01 4.96795982e-01
1.63987860e-01 6.64129853e-02 7.12607563e-01 -1.55366993e+00
-3.56315285e-01 -1.85340717e-01 1.30944625e-01 1.54239252e-01
-4.45776135e-01 -2.34030232e-01 -4.21563983e-01 -4.21172708e-01
4.33475941e-01 -8.05515230e-01 -3.11596721e-01 -2.06167087e-01
-8.83797944e-01 -2.27387428e-01 4.90890354e-01 -3.95460129e-01
1.59798563e+00 -1.80149043e+00 6.57877862e-01 1.24844825e+00
8.96565914e-01 -1.30875066e-01 1.26661301e-01 6.76276445e-01
-1.25153497e-01 2.02161595e-01 -4.72656488e-01 -2.11256027e-01
-5.60618378e-02 -7.74616525e-02 3.11425805e-01 8.40284646e-01
6.51273802e-02 4.22752380e-01 -1.14990461e+00 -7.87584543e-01
1.71566769e-01 3.43661577e-01 -5.44593990e-01 7.24127665e-02
1.30159095e-01 5.22760808e-01 -5.59161067e-01 4.77341741e-01
6.60804212e-01 -7.03785181e-01 8.25042546e-01 -3.90843242e-01
-9.47387442e-02 -3.27316433e-01 -1.66066313e+00 1.36762393e+00
1.32571653e-01 2.82067269e-01 4.40555811e-01 -1.50992632e+00
9.42590892e-01 2.89063185e-01 1.00242484e+00 9.70591307e-02
4.36900467e-01 7.09253326e-02 1.88865349e-01 -3.82785648e-01
1.39675975e-01 2.50734448e-01 1.34413227e-01 5.55911601e-01
-1.11108005e-01 3.17627937e-01 6.55112445e-01 6.61331713e-01
1.33000076e+00 -3.22435111e-01 4.44939524e-01 -7.23649561e-01
1.17723238e+00 -4.81570549e-02 4.50598329e-01 5.37791789e-01
-5.55593669e-02 1.48987649e-02 9.55744743e-01 4.51367116e-03
-9.66594994e-01 -1.17797959e+00 -1.22703016e-01 6.11262023e-01
4.18286085e-01 -3.92531782e-01 -8.34955037e-01 -6.31483018e-01
6.04348034e-02 -1.39822111e-01 -5.69842875e-01 7.37457052e-02
-1.88138530e-01 -9.23967898e-01 2.02300861e-01 -1.53989270e-01
2.93152243e-01 -6.03695095e-01 -1.12375766e-02 2.89360642e-01
-3.46699357e-01 -1.13587308e+00 -4.56425428e-01 -1.27139792e-01
-8.07059228e-01 -1.69983244e+00 -4.89333481e-01 -8.61615360e-01
1.10283625e+00 2.19772890e-01 9.89355564e-01 7.12453067e-01
-3.25561076e-01 5.15830755e-01 -4.34637338e-01 4.72853661e-01
-3.74171168e-01 2.85692602e-01 6.44867942e-02 7.06514895e-01
-1.31173268e-01 -8.60226572e-01 -5.97211063e-01 1.67753592e-01
-5.93758821e-01 -1.19218655e-01 7.85238564e-01 8.57690275e-01
4.34123397e-01 4.75485295e-01 6.51452959e-01 -9.72540677e-01
8.55826080e-01 -7.68665850e-01 -7.49402225e-01 4.10921812e-01
-9.86130953e-01 1.51942924e-01 4.84100848e-01 -7.53386170e-02
-7.34008253e-01 -9.33833197e-02 3.39617133e-01 -1.59180924e-01
4.46464509e-01 1.01697302e+00 -2.33480558e-02 -3.55159491e-01
2.29225114e-01 2.17829779e-01 9.00101587e-02 -3.64583492e-01
3.30205977e-01 3.72125864e-01 2.90738046e-01 -6.10642672e-01
1.15569413e+00 6.65198743e-01 4.48758274e-01 -5.93008697e-01
-3.45949620e-01 -8.92396390e-01 -8.54875267e-01 -5.12810051e-01
6.35432661e-01 -6.17353976e-01 -1.39954758e+00 1.27643004e-01
-1.15293074e+00 5.55579722e-01 1.85484961e-01 4.03293997e-01
-6.21232577e-02 9.19846475e-01 -5.94101191e-01 -1.05934429e+00
-2.42608652e-01 -1.12299681e+00 6.32661462e-01 -2.06508234e-01
1.87863156e-01 -1.35803330e+00 2.07015470e-01 2.71910697e-01
4.37378995e-02 8.04454029e-01 8.92770827e-01 -4.80062127e-01
-8.90677512e-01 -2.26007864e-01 -5.99838018e-01 1.48724124e-01
7.94137269e-02 4.77022409e-01 -5.05042017e-01 -6.05725884e-01
-2.86982268e-01 2.59474188e-01 7.41126359e-01 2.12384984e-01
7.15263128e-01 -2.85050958e-01 -5.93260050e-01 3.58823448e-01
1.80696690e+00 -1.52300477e-01 1.83656216e-01 1.24065638e-01
8.54543388e-01 6.81739509e-01 4.75195408e-01 7.22349823e-01
4.05995637e-01 4.38745737e-01 6.71895981e-01 -2.19040215e-01
4.85668570e-01 1.18653037e-01 4.45463844e-02 1.38320637e+00
-4.45014119e-01 -2.52896011e-01 -8.14064980e-01 5.59625149e-01
-2.00851917e+00 -1.03663445e+00 -4.33886915e-01 2.18952799e+00
4.92862016e-01 2.51044601e-01 2.66211987e-01 4.01906341e-01
1.41011178e+00 3.32107246e-02 -2.35491708e-01 4.52281162e-03
-9.26297009e-02 -6.91829994e-02 4.64866549e-01 6.33089602e-01
-1.03129375e+00 4.43668514e-01 5.73341227e+00 7.80897677e-01
-6.03234529e-01 -2.36929650e-03 3.33415538e-01 3.49598080e-01
-4.29037452e-01 1.89193532e-01 -2.82771081e-01 4.66765314e-01
4.41750169e-01 -5.62164962e-01 7.04753876e-01 7.12923944e-01
2.89344937e-01 -9.62963477e-02 -9.25835848e-01 8.39466393e-01
-9.00951102e-02 -1.21130860e+00 -1.02346182e-01 3.79504442e-01
7.18819797e-01 -3.87063414e-01 -6.06115572e-02 -1.95449784e-01
1.70023426e-01 -6.96485341e-01 3.50946754e-01 5.66631794e-01
5.02223313e-01 -8.98125470e-01 8.51023495e-01 1.27200738e-01
-1.63061869e+00 1.62330773e-02 -6.30119741e-02 1.13709241e-01
1.16920173e-01 9.43193436e-01 -8.75011623e-01 1.01129150e+00
3.17345560e-01 9.73087966e-01 -6.67680442e-01 1.10123026e+00
-2.64251605e-03 4.67774630e-01 -3.49292845e-01 -7.81984404e-02
3.61380666e-01 -7.72670269e-01 1.03035247e+00 1.01282704e+00
1.10313676e-01 -1.71601996e-01 4.06327814e-01 9.44094956e-01
-7.84813985e-02 4.44202423e-01 -5.65411747e-01 -1.99171022e-01
5.55879951e-01 1.72852385e+00 -1.42928302e+00 -7.35028386e-02
-1.25544623e-01 6.89997196e-01 4.44861174e-01 3.12330097e-01
-9.15968597e-01 -2.76505500e-01 2.46282324e-01 8.59945044e-02
1.03689790e-01 -1.28719375e-01 6.28399616e-03 -1.12497246e+00
9.35829803e-02 -5.56734860e-01 4.57129091e-01 -1.68921635e-01
-1.42855394e+00 6.25671208e-01 -5.88505156e-02 -1.19095039e+00
1.26324952e-01 -5.17578304e-01 -7.11778939e-01 4.58425403e-01
-1.29919827e+00 -1.08231771e+00 -4.97439891e-01 6.09631181e-01
-2.48852313e-01 -1.26592174e-01 5.57903707e-01 5.33926427e-01
-8.81872177e-01 3.07990164e-01 4.22901869e-01 2.17497900e-01
2.92837471e-01 -1.37239754e+00 -3.10360283e-01 9.38051045e-01
-7.65445828e-02 6.53332770e-01 6.33094728e-01 -6.52079284e-01
-1.40115976e+00 -7.57952392e-01 6.27823651e-01 -6.42750338e-02
8.66191804e-01 -3.30835909e-01 -6.14311159e-01 3.00969750e-01
1.42130673e-01 -1.18465185e-01 6.91721857e-01 1.92102596e-01
-2.03827731e-02 -3.57509524e-01 -1.27237427e+00 4.20963377e-01
1.11432254e+00 -3.89753580e-01 -1.01639353e-01 5.40208399e-01
5.21751881e-01 8.26761350e-02 -1.35202932e+00 4.18031424e-01
4.91238117e-01 -9.54437733e-01 1.09925950e+00 -3.08259219e-01
1.51463047e-01 -4.78073239e-01 1.17183633e-01 -1.05064988e+00
-5.46374083e-01 -5.52752018e-01 -4.05822806e-02 1.40864718e+00
3.74665588e-01 -7.60757685e-01 7.55971372e-01 -1.43598663e-02
3.33547741e-01 -8.14709902e-01 -9.76638734e-01 -5.86595476e-01
-5.51850021e-01 -5.81156947e-02 3.78654718e-01 1.33472300e+00
1.22132719e-01 5.61207473e-01 -2.35379294e-01 3.19446176e-01
1.20068645e+00 8.52803364e-02 4.93733674e-01 -1.66249049e+00
-3.60333592e-01 -6.79171741e-01 -7.62294531e-01 -3.53386968e-01
2.18247890e-01 -1.16681015e+00 -3.05107504e-01 -1.42656815e+00
4.77343947e-01 -6.54142082e-01 -4.16509271e-01 -1.85329422e-01
-1.86666086e-01 2.91737765e-01 1.58799246e-01 1.92501783e-01
-7.56162882e-01 4.35386330e-01 1.20158923e+00 -2.93726176e-01
-7.42931440e-02 1.08099036e-01 -5.56384087e-01 4.93782610e-01
4.87982720e-01 -5.20994842e-01 -4.67916816e-01 1.15877233e-01
4.83342081e-01 1.32965699e-01 4.17051643e-01 -8.17061901e-01
4.90663737e-01 -6.70492873e-02 -6.01517074e-02 -6.27371073e-01
1.95873842e-01 -1.10266984e+00 4.44766670e-01 6.08029425e-01
-1.21425726e-01 1.24668933e-01 -2.82266170e-01 1.06270814e+00
-1.08885974e-01 -1.46250144e-01 7.34224498e-01 -4.59103882e-02
-4.45666879e-01 3.10333848e-01 -1.94631502e-01 -1.33142173e-02
1.25982761e+00 -3.02639753e-01 -4.76286449e-02 -2.63270885e-01
-1.00786352e+00 4.89826918e-01 3.71782303e-01 -4.77956831e-02
5.80783129e-01 -1.31566203e+00 -9.27145660e-01 -3.15493584e-01
5.30813402e-03 -2.33620688e-01 1.12815708e-01 1.36619985e+00
-5.37185252e-01 1.63066328e-01 3.17385159e-02 -9.07081306e-01
-1.39542007e+00 7.99681664e-01 7.91914761e-02 -7.34527588e-01
-3.42281729e-01 6.04931414e-01 -1.34947956e-01 -5.00143349e-01
5.13446378e-03 -2.30432302e-02 -5.32158911e-01 3.68334860e-01
-1.82120144e-01 6.93470180e-01 -2.21718192e-01 -8.92519414e-01
-6.81808710e-01 6.13575697e-01 2.31368780e-01 6.80210367e-02
1.25030744e+00 -3.69697481e-01 -7.13498533e-01 3.45199496e-01
1.15751040e+00 7.46433362e-02 -5.39415121e-01 -2.71337062e-01
2.89759010e-01 -4.62357759e-01 -6.54596463e-02 -2.43974045e-01
-1.18857205e+00 4.25440222e-01 2.80187756e-01 6.38455510e-01
9.65925217e-01 -1.74008295e-01 2.57359415e-01 3.40970278e-01
4.02905256e-01 -8.84849727e-01 1.28232852e-01 8.42828751e-02
5.00553489e-01 -1.22438395e+00 3.13854724e-01 -1.04661286e+00
-2.20682904e-01 1.08746374e+00 4.06999588e-01 -1.96789697e-01
9.15488899e-01 -2.41847277e-01 -5.34542978e-01 -5.08779287e-01
-4.34159517e-01 -4.03990895e-01 4.02512401e-01 3.06946158e-01
2.56900609e-01 6.29078746e-02 -7.09776580e-01 3.46891433e-01
1.86108395e-01 -5.59338152e-01 3.62081051e-01 4.03733492e-01
-3.27161908e-01 -9.82336223e-01 -2.99200982e-01 3.62441421e-01
-2.65504867e-01 -7.73579255e-02 -5.49797058e-01 8.29391181e-01
1.19610541e-01 1.02348220e+00 -1.62063822e-01 -4.33428735e-01
3.16951424e-04 -4.60057110e-01 3.34365726e-01 -5.30913472e-01
-5.94730854e-01 3.50576974e-02 9.39624235e-02 -2.57811308e-01
-8.39069784e-01 -5.85531175e-01 -9.27208900e-01 -6.13299191e-01
-8.75009537e-01 6.81779206e-01 5.28476775e-01 8.79117966e-01
1.52496323e-01 6.62271082e-01 1.16908705e+00 -7.21674860e-01
-2.92966515e-01 -7.86113799e-01 -7.62989819e-01 3.20549279e-01
1.11123301e-01 -9.23763335e-01 -6.71925366e-01 -2.17507914e-01] | [7.072240352630615, 5.261785507202148] |
ab49041f-fe77-4403-8982-052bdd3318b9 | roles-of-words-what-should-nt-be-augmented-in | null | null | https://openreview.net/forum?id=_jpxhquKzO9 | https://openreview.net/pdf?id=_jpxhquKzO9 | Roles of Words: What Should (n’t) Be Augmented in Text Augmentation on Text Classification Tasks? | Text augmentation techniques are widely used in text classification problems to improve the performance of classifiers, especially in low-resource scenarios. Previous text-editing-based methods augment the text in a non-selective manner: the words in the text are treated without difference during augmentation, which may result in unsatisfactory augmented samples. In this work, we present four kinds of roles of words (ROWs) which have different functions in text classification tasks, and design effective methods to automatically extract these ROWs based on statistical and semantic perspectives. Systematic experiments are conducted on what ROWs should (n't) be augmented during augmentation for classification tasks. Based on these experiments, we discover some interesting and instructive potential patterns that certain ROWs are especially suitable or unsuitable for certain augmentation operations. Guided by these patterns, we propose a set of Selective Text Augmentation (STA) operations, which significantly outperform traditional methods and show outstanding generalization performance. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['text-augmentation'] | ['natural-language-processing'] | [ 8.39784563e-01 1.65370256e-02 -4.18976665e-01 -4.98220921e-01
-2.08925530e-01 -2.56435424e-01 8.40099454e-01 4.70908433e-01
-5.47161341e-01 8.85623336e-01 4.80884939e-01 -4.95850652e-01
5.58836460e-02 -7.39591420e-01 -1.40853524e-01 -6.93296671e-01
2.73552299e-01 4.30732012e-01 -7.44416267e-02 -5.06252885e-01
5.82642019e-01 2.12977737e-01 -1.84549546e+00 6.03106141e-01
1.21440077e+00 5.17861366e-01 1.11886607e-02 2.05807328e-01
-8.77893269e-01 4.20301527e-01 -8.58465016e-01 -4.17126179e-01
5.38954772e-02 -5.56947589e-01 -8.91192019e-01 3.97496253e-01
1.57506198e-01 -8.51950645e-02 -6.74418807e-02 1.10308063e+00
1.65413231e-01 2.14890793e-01 7.62090862e-01 -1.11631441e+00
-6.19682193e-01 1.22281122e+00 -5.65716565e-01 1.28955245e-01
3.20236951e-01 -3.23221356e-01 1.05935872e+00 -1.20013559e+00
2.44994134e-01 1.02683103e+00 4.38354671e-01 5.40555060e-01
-9.83587205e-01 -7.09729433e-01 7.16329336e-01 4.89093922e-02
-1.27433252e+00 -4.30762768e-01 9.21007752e-01 -1.56820729e-01
5.72717965e-01 6.69683993e-01 5.48688591e-01 1.11303639e+00
-1.82147354e-01 1.06577444e+00 1.07234573e+00 -8.55700493e-01
7.85042271e-02 5.89336574e-01 4.49502945e-01 3.12876463e-01
4.73265409e-01 -5.25260925e-01 -6.06676221e-01 -3.02610099e-01
3.04982960e-01 1.16529420e-01 -2.50159413e-01 -9.13067386e-02
-1.38997340e+00 8.18181515e-01 8.92032832e-02 5.34439802e-01
-2.79394835e-01 -6.00168645e-01 5.72421491e-01 1.05875783e-01
7.43728042e-01 7.56580234e-01 -8.02828133e-01 9.93053168e-02
-5.43551326e-01 2.18608156e-01 3.50460261e-01 1.10511065e+00
6.47914290e-01 1.05816778e-02 -5.34480214e-01 1.12156260e+00
-6.60219043e-02 8.11407790e-02 8.00815284e-01 9.53663588e-02
5.78902960e-01 1.18620920e+00 6.85053244e-02 -7.12173402e-01
-3.75068128e-01 -6.26057386e-01 -1.10783398e+00 -3.67979109e-01
2.23663449e-01 2.74803787e-02 -9.49311495e-01 1.49140108e+00
3.74379784e-01 -1.04018383e-01 1.42841831e-01 3.81364614e-01
6.85885489e-01 5.05816042e-01 3.41861457e-01 -4.41529810e-01
1.67604411e+00 -9.09086764e-01 -1.14228570e+00 -3.37968081e-01
1.35198784e+00 -8.13286960e-01 1.66615105e+00 4.68387872e-01
-5.12775660e-01 -4.71733123e-01 -9.95881200e-01 2.50232577e-01
-8.42139900e-01 5.76909721e-01 7.94191301e-01 9.49844956e-01
-3.82763088e-01 4.69876438e-01 -1.71165928e-01 -1.27195776e-01
3.61183256e-01 4.12928581e-01 -1.66730136e-01 1.33817315e-01
-1.35381997e+00 6.58945918e-01 6.50727868e-01 1.93400845e-01
-2.56551560e-02 -4.51418042e-01 -8.29311132e-01 -5.57535216e-02
7.16038048e-01 -2.88462549e-01 1.03401065e+00 -1.09543550e+00
-1.10094583e+00 6.43850565e-01 -4.86716837e-01 -2.45100453e-01
2.96645164e-01 -3.27788621e-01 -6.51293099e-01 -2.61436552e-01
9.08311158e-02 3.75744820e-01 9.74379063e-01 -1.29376900e+00
-8.28395426e-01 -4.70632643e-01 -1.44982874e-01 3.54693830e-01
-1.35599649e+00 9.94896889e-02 -4.39542472e-01 -1.26914203e+00
3.95340174e-01 -8.15266669e-01 -4.13602084e-01 -3.60686570e-01
-8.30917954e-01 -4.25460428e-01 1.10330009e+00 -3.82631004e-01
1.56677568e+00 -1.91201568e+00 -1.38631538e-01 2.33781025e-01
1.95369840e-01 5.27346849e-01 -2.94719841e-02 2.42540285e-01
2.02382617e-02 5.95261276e-01 -3.10213238e-01 -4.04304713e-01
-2.39944369e-01 2.46992081e-01 -5.07579386e-01 -3.09338104e-02
2.41471305e-01 6.89172745e-01 -6.88177288e-01 -6.04578257e-01
2.38445640e-01 1.54898852e-01 -3.66934329e-01 9.19985771e-02
-2.56207347e-01 4.72726136e-01 -8.68624032e-01 6.75840080e-01
5.10901928e-01 -8.29792172e-02 1.61714390e-01 -2.19201446e-01
3.69116217e-02 5.25575697e-01 -1.05821729e+00 1.08747673e+00
-3.14789295e-01 4.16209936e-01 -4.33828115e-01 -1.03655732e+00
1.16014814e+00 1.46304280e-01 2.25546718e-01 -3.37062776e-01
3.14319521e-01 -3.12901177e-02 -1.46271354e-02 -4.85281080e-01
1.04798055e+00 3.40547669e-03 7.37181306e-02 5.67235768e-01
-3.50075483e-01 1.94913432e-01 4.61919427e-01 9.84065235e-02
7.03605950e-01 -1.85746506e-01 4.37992573e-01 -3.27436030e-01
8.65929961e-01 4.10366803e-02 4.56750065e-01 8.64190817e-01
-8.84209946e-03 3.31055641e-01 4.37376499e-01 -5.03711760e-01
-1.05624127e+00 -1.70520931e-01 -1.98402107e-01 1.39725173e+00
8.91902857e-03 -5.93726158e-01 -5.35341918e-01 -1.06467366e+00
-3.28859210e-01 9.62664187e-01 -1.01190257e+00 -2.61301190e-01
-4.88252848e-01 -1.15714204e+00 4.70602930e-01 7.07693219e-01
3.67867202e-01 -1.11813104e+00 -2.38400288e-02 2.25043353e-02
-1.09000087e-01 -1.09733963e+00 -4.89423960e-01 3.52389187e-01
-1.03112662e+00 -8.03838253e-01 -6.12372875e-01 -8.06238055e-01
1.26929355e+00 6.63515985e-01 8.01482379e-01 4.57995206e-01
2.03311220e-01 -3.04712147e-01 -1.11445582e+00 -7.70182550e-01
-5.33369839e-01 5.46102762e-01 1.64467454e-01 2.20816746e-01
5.36300421e-01 -1.44823149e-01 -1.45361587e-01 4.61313963e-01
-1.10092866e+00 3.35681796e-01 6.60389245e-01 1.05267990e+00
5.23531437e-01 3.04652989e-01 7.36576140e-01 -1.51782107e+00
1.01748872e+00 -2.56657809e-01 -3.00058007e-01 4.62575346e-01
-8.47993374e-01 2.32302040e-01 9.52466786e-01 -7.78628528e-01
-1.17092478e+00 -1.48657383e-02 -1.33300468e-01 -7.79605359e-02
-1.82160094e-01 5.53748488e-01 -4.24254537e-01 4.28937748e-02
7.51072407e-01 3.61710399e-01 -1.92875013e-01 -4.92753267e-01
1.14286810e-01 8.90064061e-01 -7.75831416e-02 -5.37606359e-01
7.03352928e-01 3.20318997e-01 -2.27415964e-01 -9.39294577e-01
-1.12988889e+00 -4.53882456e-01 -9.24045384e-01 -2.91006956e-02
2.74163216e-01 -5.48883677e-01 -1.59513459e-01 2.32205093e-01
-7.69816041e-01 8.67752451e-03 -3.17712009e-01 3.90606791e-01
1.66199103e-01 4.96537775e-01 -1.30678281e-01 -1.01625478e+00
-4.31845933e-01 -1.09894574e+00 8.72086644e-01 3.12251210e-01
-4.26073700e-01 -8.58645439e-01 -4.97248143e-01 3.73948187e-01
1.56955808e-01 -4.07945573e-01 1.22681844e+00 -1.34412014e+00
1.25703335e-01 -5.81916511e-01 -3.28213759e-02 2.08069161e-01
4.35231417e-01 8.75413604e-03 -9.63691652e-01 -1.71510443e-01
-2.11927786e-01 -6.23645028e-04 7.73785233e-01 1.15714565e-01
1.83122122e+00 -5.00279963e-01 -4.77386862e-01 4.01753753e-01
6.79066122e-01 4.13037747e-01 6.93674505e-01 4.74296659e-01
7.60623813e-01 7.82673955e-01 8.82871509e-01 5.77621281e-01
-8.79410505e-02 5.61848462e-01 1.82426572e-01 -2.28408411e-01
3.49790275e-01 -2.81380028e-01 1.74815685e-01 8.09817851e-01
1.16274014e-01 -5.08769274e-01 -6.74035192e-01 3.95009547e-01
-1.60180306e+00 -5.04542589e-01 -4.72903669e-01 2.36684966e+00
1.04235351e+00 5.15618920e-01 6.38695434e-02 7.48724937e-01
8.66492987e-01 3.51009332e-02 -3.09654981e-01 -2.47747153e-01
-3.58880073e-01 1.33885920e-01 3.10482502e-01 2.55019106e-02
-1.17262602e+00 1.05443883e+00 6.51478386e+00 1.04499388e+00
-7.74449468e-01 -1.83963284e-01 7.21624494e-01 1.92620963e-01
-5.28652489e-01 4.85431515e-02 -1.14938557e+00 2.69190282e-01
4.48347300e-01 -1.68671068e-02 -3.81401484e-03 7.87491918e-01
-3.65165845e-02 3.17926169e-03 -8.43210816e-01 5.48757493e-01
1.73846766e-01 -1.26204872e+00 8.03004384e-01 -4.99288039e-03
6.81609631e-01 -8.66341591e-01 7.04016089e-02 4.08421636e-01
1.63895283e-02 -7.38849938e-01 3.67092311e-01 4.59077805e-02
5.16143203e-01 -8.12753081e-01 9.50711071e-01 3.54762346e-01
-1.00069690e+00 -9.87125486e-02 -3.56072605e-01 -7.35417241e-03
-2.23325461e-01 7.05521047e-01 -8.72161508e-01 4.47754413e-01
4.06713426e-01 4.51943159e-01 -8.60056520e-01 5.78032494e-01
-3.92852098e-01 7.23605096e-01 3.11841276e-02 -5.21248519e-01
1.28387943e-01 -2.86776483e-01 5.48294127e-01 1.05132401e+00
2.15991214e-01 1.66035905e-01 3.28958407e-02 6.54433727e-01
-2.32708022e-01 7.76915133e-01 -6.33346796e-01 -2.95508355e-01
5.83908975e-01 1.24893057e+00 -1.11025071e+00 -7.21177518e-01
-3.80085111e-01 8.06531191e-01 -6.46845251e-02 1.66492730e-01
-5.04617155e-01 -4.55565304e-01 3.66890162e-01 7.59006739e-02
-4.00538258e-02 2.82597560e-02 -8.36739242e-01 -1.21580732e+00
1.69231474e-01 -9.36920106e-01 4.91285801e-01 -5.96946657e-01
-1.23475075e+00 8.54433656e-01 1.06055364e-02 -1.50840569e+00
2.02291623e-01 -5.09193659e-01 -5.29529274e-01 6.95516527e-01
-1.10583830e+00 -1.03751290e+00 -3.22446287e-01 4.66866255e-01
8.83096933e-01 -4.18949366e-01 8.12439799e-01 1.56488329e-01
-8.66988063e-01 1.11100399e+00 7.00039119e-02 2.17521518e-01
6.11629605e-01 -1.10711074e+00 4.62917358e-01 9.07255828e-01
3.68967295e-01 8.25518847e-01 8.07338595e-01 -8.35093319e-01
-1.10523593e+00 -1.13291681e+00 1.05116808e+00 -2.71996349e-01
4.49568182e-01 -5.17720878e-01 -1.26631045e+00 5.45926094e-01
-5.74463867e-02 -4.15316164e-01 9.33738232e-01 5.15367925e-01
-2.02544913e-01 1.76313818e-01 -9.49731290e-01 1.12465763e+00
8.79412174e-01 -2.26910800e-01 -7.20291197e-01 5.29123187e-01
6.98190629e-01 -8.47135708e-02 -2.86167622e-01 5.23020387e-01
3.52215737e-01 -4.32659090e-01 7.99776554e-01 -9.60249245e-01
4.08248395e-01 -1.85743466e-01 8.84705633e-02 -1.30758703e+00
1.51031660e-02 -4.80552495e-01 3.84400110e-03 1.33335340e+00
6.76004052e-01 -8.68885875e-01 9.72014189e-01 4.99115825e-01
-2.11777106e-01 -8.44470263e-01 -5.90022922e-01 -5.84827244e-01
-1.20501360e-02 -3.38145852e-01 8.49779665e-01 1.46004081e+00
2.08905846e-01 4.60649490e-01 -5.43799400e-01 -1.50716349e-01
1.49717256e-01 -7.49727488e-02 8.21278632e-01 -1.30207062e+00
2.98964322e-01 -6.55198574e-01 -2.08331309e-02 -1.05959296e+00
1.08690016e-01 -7.95968175e-01 -3.74607258e-02 -1.12834585e+00
2.78404862e-01 -6.72160983e-01 -3.13588202e-01 8.78978491e-01
-6.74416542e-01 1.42664909e-01 -1.95470020e-01 3.26772183e-01
-1.65684134e-01 8.30936551e-01 1.21605766e+00 -1.44803837e-01
-5.92763484e-01 3.06479871e-01 -1.08914351e+00 7.32495070e-01
1.00425470e+00 -3.57854486e-01 -4.04119849e-01 -2.17771009e-01
2.76246995e-01 -4.95369166e-01 -3.03695917e-01 -4.80057180e-01
-1.03431173e-01 -3.13123167e-01 4.51411039e-01 -7.26890326e-01
7.39043429e-02 -6.73331738e-01 -3.97051513e-01 2.81168193e-01
-8.57534111e-01 9.86191332e-02 2.82964557e-01 3.57256591e-01
-1.81918189e-01 -6.00264132e-01 5.42127073e-01 7.34298527e-02
-5.35977423e-01 1.79376185e-01 -6.40360534e-01 -2.62523174e-01
8.79405856e-01 -4.38206553e-01 -1.43041015e-01 -2.41904557e-01
-5.69266200e-01 6.77535832e-02 1.09714679e-01 5.76558173e-01
5.08892000e-01 -1.29918814e+00 -6.94272518e-01 5.79916954e-01
4.95804578e-01 -9.00723338e-02 1.00872755e-01 6.66152239e-01
3.51715833e-02 3.13970655e-01 4.68246713e-02 -1.79986209e-01
-1.51893878e+00 5.58603942e-01 -1.16736270e-01 -2.91624039e-01
-7.36592352e-01 7.23930120e-01 2.16464028e-01 -3.87723714e-01
2.85732925e-01 -3.32480103e-01 -7.27981150e-01 2.96896756e-01
6.70561314e-01 2.12690365e-02 4.02783930e-01 -5.55704415e-01
3.10765374e-02 1.54853329e-01 -7.57693112e-01 1.20101422e-01
1.04453564e+00 -1.17704779e-01 -5.18183485e-02 3.93131107e-01
6.78502262e-01 1.53874800e-01 -4.78419065e-01 -6.53591454e-01
1.68319836e-01 -7.10853815e-01 2.82061026e-02 -6.47052228e-01
-9.68795180e-01 7.61849940e-01 2.92275667e-01 5.15309036e-01
1.25073326e+00 -1.84079185e-01 5.27263403e-01 5.75849652e-01
-3.59100029e-02 -1.21449804e+00 -8.07535648e-02 5.75120628e-01
8.48197460e-01 -1.24553275e+00 3.34530592e-01 -7.81106353e-01
-7.87559748e-01 1.12073839e+00 8.89656365e-01 5.07753551e-01
3.92998546e-01 5.30884266e-02 -1.54303491e-01 -5.40211499e-02
-6.01333380e-01 -2.61072069e-01 3.57088000e-01 4.21997070e-01
6.35406375e-01 -3.19829658e-02 -6.92364156e-01 7.50862420e-01
-1.67100355e-01 -5.81133366e-01 5.44284523e-01 1.21335113e+00
-5.22936463e-01 -1.35469687e+00 -5.38729250e-01 1.06729174e+00
-2.20538273e-01 -2.64592975e-01 -7.72424936e-01 8.66183460e-01
6.51595742e-02 8.43430042e-01 -1.99590147e-01 -6.00173295e-01
3.45641911e-01 4.18704629e-01 2.10775621e-02 -7.72166193e-01
-6.49836421e-01 5.72517887e-02 2.00267687e-01 2.80574799e-01
-2.53040910e-01 -7.45705962e-01 -1.13731158e+00 -1.58812195e-01
-9.51882780e-01 3.34070235e-01 5.91713607e-01 1.30628347e+00
1.19606070e-01 6.70995533e-01 7.25572467e-01 -5.32421529e-01
-3.88891816e-01 -1.40392733e+00 -4.80963111e-01 4.94881898e-01
-1.16313547e-01 -8.57156634e-01 -3.92008245e-01 3.84460948e-02] | [10.553343772888184, 7.625919818878174] |
3b3a9b30-3e14-4e28-9baa-8e3a43b29d2e | a-corpus-for-suggestion-mining-of-german-peer | null | null | https://aclanthology.org/2022.lrec-1.593 | https://aclanthology.org/2022.lrec-1.593.pdf | A Corpus for Suggestion Mining of German Peer Feedback | Peer feedback in online education becomes increasingly important to meet the demand for feedback in large scale classes, such as e.g. Massive Open Online Courses (MOOCs). However, students are often not experts in how to write helpful feedback to their fellow students. In this paper, we introduce a corpus compiled from university students’ peer feedback to be able to detect suggestions on how to improve the students’ work and therefore being able to capture peer feedback helpfulness. To the best of our knowledge, this corpus is the first student peer feedback corpus in German which additionally was labelled with a new annotation scheme. The corpus consists of more than 600 written feedback (about 7,500 sentences). The utilisation of the corpus is broadly ranged from Dependency Parsing to Sentiment Analysis to Suggestion Mining, etc. We applied the latter to empirically validate the utility of the new corpus. Suggestion Mining is the extraction of sentences that contain suggestions from unstructured text. In this paper, we present a new annotation scheme to label sentences for Suggestion Mining. Two independent annotators labelled the corpus and achieved an inter-annotator agreement of 0.71. With the help of an expert arbitrator a gold standard was created. An automatic classification using BERT achieved an accuracy of 75.3%. | ['Roman Rietsche', 'Julius Janda', 'Eva Ritz', 'Dominik Pfütze'] | null | null | null | null | lrec-2022-6 | ['suggestion-mining'] | ['natural-language-processing'] | [-3.01423129e-02 7.71413624e-01 -3.18044163e-02 -5.95865607e-01
-7.59874523e-01 -5.34405887e-01 4.18067038e-01 9.51442420e-01
-4.24311131e-01 7.76503146e-01 1.64847657e-01 -6.21604681e-01
-2.44713247e-01 -6.58364713e-01 -3.89587194e-01 -2.66827166e-01
4.54255402e-01 3.64682406e-01 5.83734095e-01 -4.57645237e-01
8.08280766e-01 3.49746108e-01 -1.76229417e+00 5.09202659e-01
1.10244846e+00 5.57786107e-01 6.58385277e-01 7.98008621e-01
-4.71615314e-01 9.70996797e-01 -9.61534142e-01 -7.92934000e-01
-2.96230942e-01 -5.02888620e-01 -1.11365247e+00 2.05441758e-01
4.12835151e-01 -7.03430474e-02 5.45433521e-01 8.42758179e-01
3.54266226e-01 3.79624397e-01 3.10619056e-01 -7.54086673e-01
-3.88708562e-01 1.12777400e+00 -2.24637613e-01 2.01208696e-01
7.06401169e-01 -4.67994988e-01 1.09765601e+00 -8.51090908e-01
8.70795369e-01 8.07064891e-01 1.00660861e-01 5.53056598e-01
-7.94353604e-01 -3.92253101e-01 1.06906787e-01 4.45462614e-01
-6.50555551e-01 -2.18661398e-01 7.80802310e-01 -7.15093970e-01
7.21304417e-01 2.81181127e-01 5.88384986e-01 6.91490173e-01
8.98413435e-02 5.72475255e-01 1.23198783e+00 -1.02382350e+00
2.73621798e-01 1.04512441e+00 3.65307152e-01 4.66513991e-01
1.56960398e-01 -5.71809649e-01 -6.45572186e-01 4.52636704e-02
-7.67618194e-02 -2.88419157e-01 -1.25172213e-01 2.41815317e-02
-5.21856487e-01 8.80181193e-01 2.20257044e-02 7.10059941e-01
2.30251700e-02 -4.39306349e-01 4.51507449e-01 7.19393909e-01
4.78865832e-01 5.57516873e-01 -6.29451692e-01 -4.07664984e-01
-6.12304211e-01 1.47086069e-01 1.23505270e+00 6.60155177e-01
6.08063340e-01 -3.74830753e-01 1.28235981e-01 1.01016378e+00
5.99733889e-01 4.74065505e-02 9.29464161e-01 -7.88556337e-01
5.33652782e-01 1.03089154e+00 -1.21483751e-01 -9.32514489e-01
-9.75325629e-02 -3.02944481e-01 -2.85450578e-01 5.94305173e-02
3.69386584e-01 -3.61700803e-01 -2.58713752e-01 1.12449408e+00
4.88836735e-01 -3.87083352e-01 1.20008059e-01 6.06840909e-01
1.41507530e+00 3.80818874e-01 -5.35610579e-02 -4.29843068e-01
1.35667157e+00 -9.20691788e-01 -7.87389159e-01 6.66150376e-02
1.15437818e+00 -1.18743658e+00 7.35069871e-01 9.44956481e-01
-9.48805392e-01 -6.26568317e-01 -7.93796003e-01 1.71838626e-01
-1.70833439e-01 2.32221782e-01 3.50702375e-01 9.05239880e-01
-9.69300568e-01 7.34861970e-01 -3.84361982e-01 -3.91082257e-01
1.20495103e-01 3.65794957e-01 -5.44327080e-01 7.66290650e-02
-9.14459169e-01 7.76806653e-01 4.86797057e-02 -2.39162594e-01
2.30375361e-02 -6.01566613e-01 -7.44501889e-01 6.56731427e-02
4.06194866e-01 -1.01531938e-01 1.72861397e+00 -1.12901986e+00
-1.92592001e+00 1.06256461e+00 -6.28288835e-02 -3.70823026e-01
3.56474668e-01 -1.85132146e-01 -2.43046731e-01 2.27757946e-01
3.22610795e-01 2.74311185e-01 3.85766298e-01 -8.31386745e-01
-9.24796045e-01 -3.69321853e-01 -2.89372429e-02 2.26443708e-02
-6.90288723e-01 3.23556542e-01 -1.52944028e-01 -3.92511368e-01
1.46892086e-01 -7.17661917e-01 -1.83654398e-01 -8.72909665e-01
3.66223119e-02 -7.99967229e-01 6.12552941e-01 -6.18101180e-01
1.40647709e+00 -1.83175957e+00 -1.73362464e-01 1.38404399e-01
3.65356058e-01 2.70055622e-01 2.46382490e-01 6.07142270e-01
-6.42721802e-02 2.21905202e-01 4.36672196e-02 -2.40173280e-01
1.33176237e-01 8.94196779e-02 -1.32703528e-01 1.68998420e-01
1.05348542e-01 3.99864316e-01 -1.11019289e+00 -6.35672212e-01
7.02127665e-02 6.65057003e-02 -4.55489278e-01 3.79529446e-01
8.15171897e-02 3.79739195e-01 -6.15082920e-01 3.24796528e-01
1.51834577e-01 2.02834144e-01 1.97036132e-01 7.49565542e-01
-3.71152997e-01 7.60460973e-01 -1.10912681e+00 1.31374705e+00
-4.89519686e-01 8.62126052e-01 7.10244924e-02 -1.12608993e+00
1.57783985e+00 5.86041510e-01 2.85474658e-01 -6.15950704e-01
1.30289420e-01 5.81549048e-01 4.30917531e-01 -6.46383345e-01
7.60929048e-01 -1.38094844e-02 4.14051116e-02 6.66201353e-01
1.59496903e-01 -4.53435689e-01 7.89787829e-01 5.36236286e-01
1.02885461e+00 -1.01900451e-01 4.75381285e-01 -2.58884907e-01
1.00161159e+00 -1.65660620e-01 4.33755606e-01 3.96101505e-01
-3.30828577e-01 2.97081113e-01 7.75086880e-01 -2.89627045e-01
-5.84147513e-01 -2.02468768e-01 -1.64361954e-01 1.14651787e+00
-6.73410952e-01 -8.11507344e-01 -6.70894682e-01 -9.06361997e-01
-2.88244754e-01 5.59909046e-01 -3.48097324e-01 3.13953161e-01
-3.72835934e-01 -7.55988657e-02 -1.90247029e-01 -1.47439213e-02
7.33394101e-02 -1.30661964e+00 -5.19779861e-01 4.18946564e-01
1.39694232e-02 -9.25907671e-01 8.54704455e-02 3.31355870e-01
-8.71543586e-01 -1.05631542e+00 -3.09297830e-01 -7.91611969e-01
7.62963057e-01 -2.24043932e-02 1.12385440e+00 3.33681583e-01
6.41025677e-02 4.13226426e-01 -8.61488879e-01 -7.96186388e-01
-8.42514217e-01 3.58397335e-01 -1.66118946e-02 -5.82077265e-01
7.45453179e-01 -4.56647784e-01 -1.05279975e-01 2.30388969e-01
-6.81566477e-01 -2.69434988e-01 5.70797503e-01 5.20402253e-01
1.85016811e-01 -1.51494280e-01 9.48075533e-01 -1.43108904e+00
1.01974809e+00 -2.71936774e-01 -4.59314585e-01 -3.14368703e-03
-6.63233280e-01 -2.17139088e-02 6.67118132e-01 -1.47627369e-01
-1.13364112e+00 -1.11035950e-01 -5.09171963e-01 3.76255810e-01
-6.52614117e-01 6.89042389e-01 1.04701534e-01 -1.42706707e-01
7.79250383e-01 -3.76687080e-01 -3.57650518e-01 -4.39245015e-01
3.59642841e-02 1.04680288e+00 1.19320147e-01 -5.82319200e-01
2.64147371e-01 -4.21509832e-01 -9.26249251e-02 -7.80089140e-01
-1.02045643e+00 -8.97820473e-01 -7.80966043e-01 -6.38685524e-01
2.99090326e-01 -4.33886081e-01 -8.06229889e-01 -1.11266024e-01
-9.83122110e-01 -2.66694605e-01 -3.95853102e-01 6.27999961e-01
-2.31931835e-01 4.62780774e-01 -6.68194950e-01 -8.38768303e-01
-3.24813694e-01 -1.15282285e+00 3.19463938e-01 4.92062271e-01
-5.71478069e-01 -1.02594864e+00 3.65603805e-01 9.63912487e-01
1.02679012e-02 -1.57567933e-01 4.40935493e-01 -1.05253267e+00
1.52179077e-01 -3.71496320e-01 2.81171441e-01 4.50507671e-01
-8.98444280e-02 2.42683977e-01 -1.12636673e+00 2.09815979e-01
2.64453292e-01 -5.54844439e-01 5.78030407e-01 -1.46291882e-01
8.97210240e-01 -2.37700164e-01 8.05162564e-02 -5.28686702e-01
8.91064227e-01 1.14905976e-01 3.99123907e-01 6.55295849e-01
4.05182928e-01 1.31836689e+00 8.89083624e-01 3.69600862e-01
3.46660316e-01 5.66385448e-01 3.23561132e-01 8.00671756e-01
1.84723899e-01 -3.79747525e-02 7.00682998e-01 1.61098564e+00
5.90372086e-02 -4.16512303e-02 -9.97999310e-01 6.38069153e-01
-1.71253359e+00 -6.65994406e-01 -8.22994232e-01 1.99959314e+00
9.27327096e-01 5.06684303e-01 -1.40482366e-01 4.93353635e-01
4.01319653e-01 -2.76174396e-01 2.62256712e-01 -1.33789682e+00
4.77377266e-01 4.92522299e-01 7.15838671e-02 6.31271958e-01
-7.70716071e-01 6.08700931e-01 4.56422234e+00 6.24860704e-01
-7.32629418e-01 5.15709110e-02 5.91815650e-01 2.35406384e-01
-2.90115297e-01 9.70421359e-03 -1.25195670e+00 2.24001110e-01
1.37201464e+00 3.70502248e-02 -3.11331600e-01 1.06015491e+00
3.25733125e-01 -4.21846360e-01 -8.20291162e-01 5.17198861e-01
2.80285299e-01 -1.00191700e+00 -5.42958677e-01 5.18438630e-02
1.03714454e+00 -2.35251665e-01 -3.82909924e-01 3.86405081e-01
4.15820509e-01 -6.45434022e-01 6.13251090e-01 2.74731040e-01
-6.17611967e-02 -8.62313211e-01 1.30120933e+00 6.63841426e-01
-7.44755507e-01 -2.19197318e-01 -2.34643862e-01 -6.71237648e-01
-1.43618405e-01 8.26396286e-01 -1.35126269e+00 4.49439079e-01
6.82684302e-01 6.88693464e-01 -7.36889422e-01 1.00415397e+00
-7.62879193e-01 8.30163896e-01 -4.61615399e-02 -5.91121614e-01
2.44095549e-01 -3.07035655e-01 3.33844453e-01 1.07896864e+00
4.13134456e-01 1.82533681e-01 1.73691586e-01 3.19182545e-01
-2.22121805e-01 8.55485380e-01 -3.90009731e-01 -1.33659206e-02
2.58379251e-01 1.55206668e+00 -9.59922910e-01 -2.60649711e-01
-3.55687469e-01 5.46731174e-01 5.17202258e-01 -1.15430154e-01
1.27444163e-01 -4.94449675e-01 6.55763745e-02 3.69585752e-01
2.55776763e-01 1.80826098e-01 -2.97487766e-01 -6.62310779e-01
1.92947522e-01 -7.70417154e-01 1.92703515e-01 -4.00664926e-01
-1.16013277e+00 4.22305703e-01 -4.06594962e-01 -1.05550909e+00
-3.66559982e-01 -6.36677444e-01 -8.28515410e-01 8.44171047e-01
-1.14577615e+00 -3.38679880e-01 -1.40985817e-01 -1.36721075e-01
5.78287899e-01 -1.88533261e-01 7.80488670e-01 2.61860520e-01
-5.08659124e-01 2.04500109e-01 -1.48247913e-01 -9.45824757e-02
1.19000673e+00 -1.68016791e+00 -2.45761991e-01 6.31934822e-01
2.01307297e-01 4.99769300e-01 8.29177439e-01 -5.37924886e-01
-9.11234856e-01 -4.88363951e-01 1.84816730e+00 -4.85952258e-01
9.07083690e-01 7.51749352e-02 -9.55999732e-01 2.10424304e-01
3.77338111e-01 -3.84416908e-01 1.24476397e+00 1.72812656e-01
7.78975934e-02 -1.55896991e-01 -8.84752035e-01 1.52475715e-01
2.24497721e-01 -2.45737821e-01 -9.25753057e-01 4.22197610e-01
4.43793595e-01 -4.85260367e-01 -1.12989068e+00 1.61074132e-01
2.31839135e-01 -1.15368164e+00 2.75158793e-01 -3.93827945e-01
9.06339765e-01 1.84953138e-02 3.05300385e-01 -1.20553124e+00
2.39997491e-01 -4.46104795e-01 6.91005439e-02 1.69376218e+00
4.72093731e-01 -3.03961605e-01 9.27996576e-01 7.98636734e-01
-5.28492987e-01 -8.55609179e-01 -5.83802700e-01 -9.16172341e-02
-2.78508961e-02 -7.08791971e-01 2.01692209e-01 9.45973992e-01
7.61994898e-01 5.56974769e-01 2.71730036e-01 -4.89497900e-01
1.11513928e-01 5.11009283e-02 8.82543504e-01 -1.62449062e+00
-2.23654881e-01 -7.25367427e-01 -3.22226107e-01 -6.39193356e-01
2.02255368e-01 -7.23364472e-01 8.34083930e-02 -1.62456739e+00
-7.52546042e-02 -4.33351070e-01 4.76663634e-02 3.35640192e-01
-2.20001966e-01 1.15633659e-01 -4.97174226e-02 -9.40755457e-02
-6.42552197e-01 1.60532400e-01 1.25311065e+00 2.42587358e-01
-5.62726676e-01 4.53686923e-01 -9.80584979e-01 9.17219996e-01
8.82715404e-01 -6.21057570e-01 -3.45548064e-01 1.50989205e-01
8.59298706e-01 -1.43256083e-01 -2.56352961e-01 -5.51328480e-01
3.84058803e-01 -8.69657248e-02 3.88779864e-02 -4.64945763e-01
-2.94562042e-01 -7.13711500e-01 -4.93665189e-01 3.40000570e-01
-5.16319931e-01 -2.06604809e-01 7.89028630e-02 2.07868665e-01
-7.03275502e-01 -1.14702880e+00 3.34213018e-01 -1.92616537e-01
-5.04347026e-01 -4.15808737e-01 -7.02417493e-01 -2.50826985e-01
1.04313993e+00 8.33773706e-03 -2.25330442e-01 -4.42383438e-01
-7.35138655e-01 1.87512830e-01 1.60999596e-01 2.84179389e-01
5.79412341e-01 -8.98291647e-01 -7.19468653e-01 -1.43273979e-01
1.37884140e-01 2.11126566e-01 1.78631410e-01 8.70823979e-01
-6.38351858e-01 7.55985796e-01 3.35773267e-02 -3.71861428e-01
-1.79165649e+00 1.10783845e-01 -3.44414860e-01 -5.34698844e-01
-3.61551732e-01 9.99982536e-01 -4.37276274e-01 -7.72328079e-01
1.70348182e-01 -3.69797558e-01 -1.15822268e+00 5.04381835e-01
7.44186640e-01 3.79918963e-01 6.35180295e-01 -5.03815293e-01
8.53082016e-02 3.82826924e-02 -1.68387130e-01 -7.25718066e-02
1.45196462e+00 -3.91312316e-02 -3.35578740e-01 7.38930106e-01
8.09289217e-01 5.00634015e-01 -5.15318930e-01 4.74031307e-02
4.98818278e-01 -4.63269562e-01 -1.10362686e-01 -6.17312849e-01
-6.58160388e-01 8.48629653e-01 7.43405446e-02 7.59667516e-01
8.39695871e-01 1.04552945e-02 3.11063290e-01 5.48477292e-01
3.58898342e-02 -1.29092765e+00 3.45179886e-02 6.86263025e-01
6.77796841e-01 -1.53702748e+00 9.97294262e-02 -6.68116629e-01
-5.74103415e-01 1.57114136e+00 6.44469261e-01 1.59508422e-01
5.20323575e-01 -6.10124730e-02 -2.15606485e-02 -2.19530076e-01
-1.05576527e+00 -3.64756659e-02 6.65090144e-01 2.21492857e-01
1.23617041e+00 2.54467010e-01 -1.09089458e+00 8.91603231e-01
-6.39632821e-01 1.23669794e-02 1.01525319e+00 9.68220294e-01
-9.64017093e-01 -1.76393402e+00 -5.66386282e-01 5.17288268e-01
-8.13087106e-01 6.18405528e-02 -9.73865449e-01 4.34860587e-01
8.32584947e-02 1.40455985e+00 -2.27703899e-01 -1.41945168e-01
3.06805193e-01 1.99353591e-01 3.03877503e-01 -1.19759798e+00
-1.21909416e+00 -6.36471063e-02 3.01296651e-01 -5.65652661e-02
-8.95207882e-01 -7.80243933e-01 -1.07920027e+00 -1.93423972e-01
-5.40150821e-01 8.94207180e-01 1.01713920e+00 1.22540021e+00
-1.72562242e-01 6.58235908e-01 5.12098074e-01 -2.90917695e-01
-4.28051174e-01 -1.39898157e+00 -4.51327741e-01 1.85135007e-01
-1.73466861e-01 -2.00220883e-01 -3.44389290e-01 1.50630280e-01] | [11.24889087677002, 9.16128158569336] |
e9760bbf-f360-4a1a-b578-16bf477c2074 | machine-learning-to-detect-cyber-attacks-and | 2307.03323 | null | https://arxiv.org/abs/2307.03323v1 | https://arxiv.org/pdf/2307.03323v1.pdf | Machine Learning to detect cyber-attacks and discriminating the types of power system disturbances | This research proposes a machine learning-based attack detection model for power systems, specifically targeting smart grids. By utilizing data and logs collected from Phasor Measuring Devices (PMUs), the model aims to learn system behaviors and effectively identify potential security boundaries. The proposed approach involves crucial stages including dataset pre-processing, feature selection, model creation, and evaluation. To validate our approach, we used a dataset used, consist of 15 separate datasets obtained from different PMUs, relay snort alarms and logs. Three machine learning models: Random Forest, Logistic Regression, and K-Nearest Neighbour were built and evaluated using various performance metrics. The findings indicate that the Random Forest model achieves the highest performance with an accuracy of 90.56% in detecting power system disturbances and has the potential in assisting operators in decision-making processes. | ['Remy Ihabwikuzo', 'Diane Tuyizere'] | 2023-07-06 | null | null | null | null | ['decision-making'] | ['reasoning'] | [-7.29507059e-02 -3.33286792e-01 -9.24073011e-02 -8.22478831e-02
-3.23567897e-01 -6.42558336e-01 5.59104681e-01 8.73677433e-01
1.12309240e-01 8.00108373e-01 -1.60947740e-01 -7.19475448e-01
-4.75816101e-01 -9.93682742e-01 3.73240054e-01 -8.21479440e-01
-6.83061898e-01 2.51019686e-01 2.34430879e-02 1.70357719e-01
7.39029825e-01 9.72555161e-01 -1.03610456e+00 1.06568485e-01
8.62982035e-01 1.09343183e+00 -1.60759494e-01 7.22587764e-01
4.56155807e-01 9.10333753e-01 -1.36010981e+00 3.11580062e-01
2.19696984e-01 -8.21515545e-02 -6.16838932e-01 -7.68903792e-02
-9.19868946e-01 -1.56212941e-01 -9.63981599e-02 8.89714062e-01
6.22130036e-01 -3.16099301e-02 6.58422172e-01 -1.74941027e+00
-1.45119801e-01 8.13481688e-01 -6.25976741e-01 8.32056344e-01
7.50845373e-01 9.64584127e-02 6.74603224e-01 -4.44122821e-01
-2.39104807e-01 6.75512552e-01 6.71094060e-01 -1.57436699e-01
-1.10991001e+00 -9.93886948e-01 -2.85682261e-01 7.22824335e-01
-1.49368608e+00 2.70042270e-01 8.69187534e-01 -4.48433787e-01
1.45300591e+00 3.92805994e-01 6.62380517e-01 4.63953882e-01
4.38028723e-01 5.06839871e-01 1.56792200e+00 -6.15047157e-01
6.37557328e-01 3.77884120e-01 5.37672818e-01 -4.94920649e-02
4.13426876e-01 6.57652691e-02 -2.50112563e-01 -6.66124821e-01
1.85605258e-01 1.09213376e-02 -3.79155427e-01 1.89577982e-01
-6.52743757e-01 8.34261596e-01 -8.83989334e-02 6.69074535e-01
-7.20401645e-01 -8.26180696e-01 6.22368753e-01 5.26511848e-01
1.77387983e-01 5.78356743e-01 -9.39269304e-01 -3.38697731e-01
-7.80832291e-01 -3.49269927e-01 1.04730976e+00 5.64191222e-01
2.42322147e-01 5.57057083e-01 1.28889814e-01 2.95649648e-01
1.03612855e-01 2.81428456e-01 6.92384005e-01 2.48060003e-02
2.36149386e-01 8.43453646e-01 1.27372622e-01 -9.06233728e-01
-7.78660953e-01 -3.35377395e-01 -7.74615943e-01 1.21754430e-01
-1.44055903e-01 -6.48569405e-01 -4.63854313e-01 8.58316898e-01
8.71509314e-02 3.13925356e-01 2.63845950e-01 1.32570893e-01
4.53820199e-01 9.60812867e-01 1.86526850e-01 -6.78498685e-01
1.03360128e+00 -1.85875297e-01 -8.14128101e-01 5.90842903e-01
6.04547024e-01 -6.76837802e-01 5.69898188e-01 1.02609921e+00
-6.18722498e-01 -3.43986362e-01 -1.35034621e+00 1.23833644e+00
-6.05521977e-01 1.38531446e-01 4.70978707e-01 1.01147139e+00
-6.68746054e-01 4.89165068e-01 -7.91446209e-01 -4.90779608e-01
1.73712030e-01 4.02408868e-01 -5.77697977e-02 5.22889853e-01
-1.14649761e+00 1.21266544e+00 7.62293160e-01 -4.16300371e-02
-9.76325989e-01 -4.73929852e-01 -6.53385580e-01 1.24272518e-01
3.72019014e-03 8.39804560e-02 7.96579957e-01 -1.36818513e-01
-1.27432477e+00 -1.44903138e-02 2.37298742e-01 -8.18198144e-01
1.82792861e-02 -7.23311678e-02 -1.20011842e+00 2.15074793e-01
-2.82983333e-01 -7.26626039e-01 6.25321627e-01 -9.98486042e-01
-9.90372181e-01 -4.70885396e-01 -2.35466138e-01 2.45209318e-02
-3.96457732e-01 4.37252283e-01 7.84280360e-01 -4.62149888e-01
-1.28897265e-01 -2.61871189e-01 -2.35850945e-01 -1.19081438e+00
-5.74469149e-01 -3.26614916e-01 1.51470554e+00 -1.07895422e+00
1.72946787e+00 -1.57242239e+00 -5.79805732e-01 1.03807878e+00
-1.74252167e-01 5.67167699e-01 6.38572216e-01 8.63617837e-01
-4.36451733e-01 -1.24936759e-01 -1.98219996e-02 5.13687134e-01
-1.95056543e-01 1.00413688e-01 -2.62288809e-01 4.66504514e-01
1.41955853e-01 2.80666173e-01 -5.78031719e-01 9.44646299e-02
9.89636958e-01 1.70399129e-01 2.74402261e-01 4.38522190e-01
3.67016286e-01 3.90966535e-01 -4.01461393e-01 7.41181552e-01
6.40468478e-01 -2.37020273e-02 2.96567857e-01 -3.09415817e-01
9.40444227e-03 1.93071991e-01 -1.51429689e+00 4.82320219e-01
-4.74489748e-01 5.45699060e-01 -3.33422065e-01 -1.46301436e+00
1.17221951e+00 7.28404939e-01 7.20817864e-01 -4.91096318e-01
1.93759963e-01 -2.25705162e-01 3.81473936e-02 -5.17676771e-01
-1.06921762e-01 5.16158521e-01 2.03507696e-03 7.79073298e-01
1.25248775e-01 3.04464661e-02 2.96041429e-01 -6.14385828e-02
1.28791010e+00 -6.33960128e-01 1.03062391e+00 -4.06597078e-01
1.10416698e+00 -1.03214651e-01 5.23936450e-01 4.64982748e-01
-2.24514484e-01 -3.52648705e-01 4.55273002e-01 -5.56645870e-01
-5.40986836e-01 -9.57684457e-01 -3.85320425e-01 3.28428179e-01
-1.87333688e-01 -4.40380275e-01 -7.02834308e-01 -8.39595318e-01
8.86278376e-02 1.27024126e+00 -3.83155882e-01 -3.47122163e-01
-3.34252447e-01 -1.36936283e+00 4.06175911e-01 4.16818738e-01
4.84541595e-01 -1.11123204e+00 -8.28478336e-01 3.96738708e-01
3.71667147e-01 -9.21306133e-01 3.89085233e-01 6.85526073e-01
-8.45930874e-01 -1.50708926e+00 3.25632215e-01 -4.08411562e-01
8.18023860e-01 1.45297600e-02 9.62641716e-01 6.19564876e-02
-5.58442295e-01 6.87990561e-02 -6.17026746e-01 -6.35783613e-01
-3.89088839e-01 -1.57637924e-01 3.43122959e-01 -1.27228066e-01
8.77269864e-01 -8.47271681e-01 -3.41378182e-01 4.25182164e-01
-5.48695683e-01 -6.75627291e-01 5.36696315e-01 6.54155970e-01
-6.57777339e-02 1.20405722e+00 1.08146811e+00 -6.59596384e-01
1.04724824e+00 -9.02603984e-01 -9.38821256e-01 2.36844346e-01
-1.28604388e+00 -4.71299469e-01 1.03220618e+00 -2.52782464e-01
-8.47206235e-01 -2.58551091e-01 -5.81414998e-02 1.92372501e-01
-7.41601884e-01 5.27463257e-01 -3.68617058e-01 -2.94636130e-01
4.21701074e-01 4.25358772e-01 -3.18958133e-01 -5.42694747e-01
-1.57559231e-01 1.15649903e+00 7.51885846e-02 -9.90788341e-02
1.01959193e+00 2.63660215e-02 -1.04481339e-01 -1.03444052e+00
-2.27534860e-01 -5.84013641e-01 -7.56088376e-01 -2.73665011e-01
3.81058872e-01 -7.34327316e-01 -9.36413109e-01 7.04648554e-01
-5.26310027e-01 2.05757618e-01 -9.46688354e-02 6.08208358e-01
2.24806853e-02 3.07344377e-01 -7.39466906e-01 -1.08205175e+00
-7.25375831e-01 -7.52812743e-01 1.61256969e-01 4.92898226e-01
-4.76099402e-01 -1.29841328e+00 -1.11290298e-01 1.14077307e-01
5.14212370e-01 3.77534986e-01 1.19969857e+00 -1.56613255e+00
-7.73437619e-02 -6.67882919e-01 1.41595155e-01 4.74346101e-01
6.76254272e-01 2.10918710e-01 -8.36932421e-01 -7.51305819e-01
1.92187741e-01 3.28657939e-03 -3.25300902e-01 2.74275512e-01
1.08652306e+00 -4.71084148e-01 -3.32063228e-01 1.97144538e-01
1.43729317e+00 1.19396949e+00 4.28804606e-01 5.49993038e-01
1.07475512e-01 1.24957532e-01 7.56882608e-01 1.03024149e+00
3.36871117e-01 1.64246827e-01 3.00774187e-01 -1.34152457e-01
7.07226396e-01 1.34546384e-02 1.10441625e-01 8.98484588e-01
2.12962270e-01 1.06343376e-02 -9.71038640e-01 4.11509603e-01
-1.41011190e+00 -8.21636915e-01 4.37242202e-02 2.08817697e+00
3.30785006e-01 5.42766511e-01 2.99967945e-01 1.20649874e+00
4.77567703e-01 -1.63158968e-01 -2.88219273e-01 -6.18459880e-01
-1.19895197e-01 4.81090963e-01 4.02992606e-01 2.83529162e-01
-1.12729168e+00 3.07296365e-01 6.35080528e+00 5.99240243e-01
-1.05130124e+00 -8.16789046e-02 5.35310566e-01 4.35446322e-01
4.69665617e-01 -2.72721350e-02 -5.64140022e-01 5.94521582e-01
1.31329048e+00 -6.83331251e-01 1.89934447e-01 7.06382275e-01
7.24343479e-01 -4.18363810e-01 -7.50407279e-01 6.28313005e-01
-4.32714857e-02 -8.41865540e-01 1.24263102e-02 -1.17592625e-01
6.72794104e-01 -1.62186697e-01 -6.20371521e-01 1.71858937e-01
6.22952163e-01 -8.93907368e-01 3.22719291e-02 2.52886266e-01
-1.19531311e-01 -1.33793366e+00 1.17457283e+00 4.18146282e-01
-1.17161655e+00 -7.13046730e-01 1.96171224e-01 -5.01776636e-01
2.50700206e-01 7.31093228e-01 -1.30419743e+00 9.25951123e-01
9.30426538e-01 6.83259428e-01 -5.17853498e-01 1.16980743e+00
-3.23561668e-01 1.29494584e+00 -3.84944826e-01 -1.02192350e-01
-1.78984031e-01 -1.21757403e-01 2.54352510e-01 1.10883188e+00
5.60840592e-02 1.18073225e-01 2.88386106e-01 2.56404310e-01
6.74310327e-01 7.13367015e-02 -4.74470258e-01 3.34442139e-01
1.05697298e+00 1.47043920e+00 -8.49535286e-01 -1.73492044e-01
-4.71133679e-01 1.66063756e-01 -3.22336048e-01 3.23348761e-01
-7.18772709e-01 -8.76865685e-01 3.66964638e-01 -2.57802665e-01
1.89030115e-02 -2.11274877e-01 -7.53933251e-01 -7.32840598e-01
-1.32030293e-01 -9.30456698e-01 7.30658114e-01 -4.77734387e-01
-1.32703340e+00 6.47574484e-01 2.95171022e-01 -1.14133799e+00
-4.01656359e-01 -3.88568461e-01 -1.30074370e+00 8.86366606e-01
-1.27930748e+00 -7.20324516e-01 -2.41794258e-01 7.21331000e-01
5.24400830e-01 -6.12919450e-01 1.03887522e+00 2.67327011e-01
-9.62515056e-01 3.33777189e-01 1.30646884e-01 3.35053235e-01
-1.86586484e-01 -1.39245939e+00 1.75132588e-01 1.17031324e+00
-2.85107479e-03 3.49914551e-01 6.62850559e-01 -6.34655476e-01
-1.25429285e+00 -7.62067199e-01 4.96599913e-01 7.25329621e-03
8.28385234e-01 -3.90473157e-02 -8.36650431e-01 5.44248879e-01
4.69507337e-01 -3.49638224e-01 1.09609485e+00 -1.98408931e-01
3.06537211e-01 -2.44014189e-01 -1.68252337e+00 3.87491062e-02
-2.43227169e-01 -4.75514233e-01 -7.14921951e-01 1.49897426e-01
-3.67528886e-01 1.57925025e-01 -1.60991645e+00 4.86888289e-01
1.36260234e-03 -5.91350317e-01 6.90721273e-01 -4.25647229e-01
-6.90374374e-01 -5.94451189e-01 -8.08342844e-02 -1.69090497e+00
-2.24925712e-01 -8.56413484e-01 -5.44988513e-01 1.36739457e+00
6.57024205e-01 -1.06460440e+00 5.11137366e-01 3.41574878e-01
4.55792904e-01 -6.84642553e-01 -8.03216219e-01 -3.44175905e-01
-3.08515936e-01 -4.92009908e-01 1.06124473e+00 1.39057183e+00
6.53928459e-01 3.06265503e-01 -6.52031004e-02 8.40142012e-01
7.16049135e-01 -8.27532634e-02 4.79292035e-01 -1.26434052e+00
1.82381660e-01 -1.07206874e-01 -6.63385689e-01 -6.11616038e-02
-8.86638090e-02 -2.61907607e-01 -7.23128438e-01 -1.35186732e+00
-1.38962701e-01 -4.99449998e-01 -5.86946428e-01 7.34667003e-01
-1.11327633e-01 -8.10264423e-02 -6.08925186e-02 1.49628296e-01
-8.67805108e-02 2.12511286e-01 9.80555415e-02 6.48645833e-02
-2.23766923e-01 8.01430404e-01 -3.00690144e-01 7.99865603e-01
1.46165740e+00 -2.61164844e-01 -5.23251653e-01 2.57242113e-01
-4.16068763e-01 1.97995365e-01 -7.13511705e-02 -1.14278090e+00
3.79501671e-01 -6.07957914e-02 7.48832524e-01 -1.01159728e+00
-2.81407833e-01 -1.32595181e+00 2.86084026e-01 8.97509217e-01
2.25661527e-02 7.11330533e-01 2.20257640e-01 1.26583025e-01
-1.70818567e-01 -2.05012083e-01 5.79059958e-01 2.35978007e-01
-7.48971164e-01 -1.51835859e-01 -9.58393335e-01 -6.61157787e-01
1.80355322e+00 -3.49573195e-01 -2.84051239e-01 -2.39957511e-01
-8.31497192e-01 4.25180674e-01 -7.59469494e-02 5.56047022e-01
6.07827783e-01 -9.05334771e-01 -5.05380988e-01 6.47736073e-01
-8.94654393e-02 -6.32625043e-01 -1.71772227e-01 5.66462040e-01
-4.30312842e-01 5.43114066e-01 -4.47377533e-01 -5.02756894e-01
-1.30379510e+00 2.88222045e-01 4.02206004e-01 -4.31231886e-01
-4.85223055e-01 -1.67355072e-02 -1.01188910e+00 -2.64234573e-01
3.88878793e-01 -2.27724910e-01 -1.00193036e+00 2.73449309e-02
7.04311490e-01 1.08455086e+00 5.87213993e-01 -5.13859510e-01
-3.46557736e-01 2.38870963e-01 -9.83236134e-02 3.79023254e-01
1.07068253e+00 9.83750820e-03 -1.75009608e-01 3.09658766e-01
6.60824656e-01 -2.30975464e-01 -6.48469567e-01 -3.61928008e-02
6.31404996e-01 -3.64586890e-01 1.06513537e-01 -1.20872843e+00
-1.13635421e+00 4.25719380e-01 8.39031696e-01 9.25570011e-01
1.57289910e+00 -5.90114295e-01 4.48322654e-01 2.59373307e-01
7.65035391e-01 -1.09209001e+00 -5.19623697e-01 2.33150199e-01
1.01540737e-01 -9.35902417e-01 1.82764784e-01 -1.59895450e-01
-4.13152993e-01 1.10611093e+00 5.73399365e-01 8.07838887e-02
1.28106427e+00 8.63074243e-01 7.81313926e-02 -3.23656462e-02
-9.75983024e-01 3.38343471e-01 -1.88492194e-01 1.23494852e+00
6.59016520e-02 3.98788035e-01 -2.83487886e-01 6.42608464e-01
-1.86540157e-01 -2.43468024e-03 6.89217627e-01 1.34813511e+00
-5.33490896e-01 -9.67382431e-01 -4.98658776e-01 9.02412713e-01
-7.29286075e-01 1.94695577e-01 -2.06529707e-01 8.39786232e-01
-1.48589909e-01 1.63688695e+00 -1.22891463e-01 -6.60890281e-01
5.68310082e-01 -2.43351609e-01 -1.69885799e-01 -9.34598297e-02
-8.55041802e-01 -5.08339822e-01 3.65537666e-02 -2.26955667e-01
-1.08291440e-01 -8.00352454e-01 -1.12127793e+00 -2.71010846e-01
-7.40439892e-01 1.05514669e+00 7.39459276e-01 1.14084136e+00
3.30185592e-02 5.75235605e-01 1.56593895e+00 -6.15507126e-01
-6.64941192e-01 -1.33588064e+00 -7.69708991e-01 1.25606790e-01
2.61433646e-02 -5.84467351e-01 -5.93805969e-01 -1.01911731e-01] | [6.176172256469727, 2.562269687652588] |
bfe89888-3cbb-4449-929d-d87a0254f9dd | cycle-balanced-representation-learning-for | 2110.15484 | null | https://arxiv.org/abs/2110.15484v1 | https://arxiv.org/pdf/2110.15484v1.pdf | Cycle-Balanced Representation Learning For Counterfactual Inference | With the widespread accumulation of observational data, researchers obtain a new direction to learn counterfactual effects in many domains (e.g., health care and computational advertising) without Randomized Controlled Trials(RCTs). However, observational data suffer from inherent missing counterfactual outcomes, and distribution discrepancy between treatment and control groups due to behaviour preference. Motivated by recent advances of representation learning in the field of domain adaptation, we propose a novel framework based on Cycle-Balanced REpresentation learning for counterfactual inference (CBRE), to solve above problems. Specifically, we realize a robust balanced representation for different groups using adversarial training, and meanwhile construct an information loop, such that preserve original data properties cyclically, which reduces information loss when transforming data into latent representation space.Experimental results on three real-world datasets demonstrate that CBRE matches/outperforms the state-of-the-art methods, and it has a great potential to be applied to counterfactual inference. | ['Liming Zhu', 'Chen Wang', 'Xiwei Xu', 'Lina Yao', 'Guanglin Zhou'] | 2021-10-29 | null | null | null | null | ['counterfactual-inference'] | ['miscellaneous'] | [ 6.40338063e-01 -1.67776003e-01 -1.13374710e+00 -3.34184021e-01
-7.60696471e-01 -2.52144754e-01 6.04861498e-01 -2.66388834e-01
-2.30141103e-01 1.16677701e+00 7.83030272e-01 -7.78808832e-01
-4.37275112e-01 -8.49228084e-01 -8.12812686e-01 -6.82926834e-01
-2.36855313e-01 1.61229476e-01 -5.26181996e-01 -1.05951406e-01
1.74470946e-01 -2.97606997e-02 -1.19508398e+00 2.30347067e-01
1.21787417e+00 8.81640255e-01 -2.81557232e-01 -3.28932434e-01
-8.26515183e-02 8.32668424e-01 -4.03710395e-01 -4.86415684e-01
2.06001773e-01 -5.29397607e-01 -3.86590093e-01 -2.78550178e-01
5.73310331e-02 -4.31939304e-01 -5.54026783e-01 1.02082956e+00
7.50811934e-01 8.99460018e-02 8.63447130e-01 -1.27636576e+00
-1.14151764e+00 8.48277152e-01 -9.74993587e-01 8.35230667e-03
2.14242354e-01 2.90376902e-01 8.47315729e-01 -3.72746229e-01
4.66728628e-01 1.67457020e+00 4.92924541e-01 7.80188143e-01
-1.60756552e+00 -1.45066690e+00 4.31716442e-01 1.96782902e-01
-8.90840173e-01 -2.96222597e-01 1.15409207e+00 -2.81949431e-01
5.38191479e-03 2.58976191e-01 5.57211816e-01 1.87391019e+00
4.04716879e-01 8.95448327e-01 1.62714159e+00 -1.48141310e-01
5.50261915e-01 -2.16894180e-01 -4.89661843e-01 4.76148352e-02
6.69709802e-01 8.73408556e-01 2.63822116e-02 -4.71580744e-01
8.15965056e-01 4.95673656e-01 -4.56787884e-01 -7.61339724e-01
-1.34797895e+00 1.37268877e+00 5.70086360e-01 -3.35279480e-02
-6.94061756e-01 3.26191969e-02 7.16907978e-01 2.86876589e-01
5.37222505e-01 3.14663857e-01 -5.01925111e-01 3.36252391e-01
-6.51437581e-01 7.49826312e-01 3.51452410e-01 5.86507380e-01
1.25832304e-01 2.48184681e-01 -4.42918748e-01 8.18643093e-01
1.64076388e-01 6.84699655e-01 1.07063079e+00 -8.20397317e-01
8.40902627e-01 4.31269705e-01 2.04895541e-01 -9.55736995e-01
-1.99823052e-01 -3.29463392e-01 -1.38614547e+00 -7.66552836e-02
2.26458490e-01 -2.26061970e-01 -6.75946653e-01 2.07310343e+00
4.04896587e-01 5.67307949e-01 1.20463461e-01 9.34584320e-01
5.22224426e-01 4.51656431e-01 5.70769191e-01 -8.93251956e-01
1.23477423e+00 -1.81110129e-01 -9.34392393e-01 -1.91101238e-01
4.57488477e-01 -4.21216995e-01 1.03067458e+00 2.29681090e-01
-7.77031362e-01 -3.79400671e-01 -1.05950534e+00 4.55341607e-01
-7.77038783e-02 -3.86777669e-01 1.29108107e+00 7.91434944e-01
-3.70764025e-02 4.44811791e-01 -2.38819391e-01 4.21591103e-01
1.07219303e+00 2.64376879e-01 2.57192645e-03 -3.53162318e-01
-1.80395722e+00 5.17635643e-01 4.84354585e-01 -1.19959749e-01
-9.55950797e-01 -1.24479783e+00 -9.82202470e-01 -4.75395732e-02
8.16093743e-01 -1.02681828e+00 1.28791916e+00 -9.28750038e-01
-1.48355746e+00 4.57907647e-01 2.49401852e-01 -6.97807431e-01
7.88255572e-01 3.79117094e-02 -9.91115630e-01 -4.21364427e-01
3.13345909e-01 2.20961526e-01 9.32882667e-01 -1.20005584e+00
-4.65443194e-01 -6.79498553e-01 -1.33767473e-02 6.27918169e-02
-7.94399157e-02 -2.29651153e-01 4.69224066e-01 -1.18743610e+00
-1.29392102e-01 -7.19816089e-01 -6.78533018e-01 -2.11807996e-01
-2.63162553e-01 -3.19621056e-01 5.28133392e-01 -4.77867246e-01
1.39226210e+00 -1.90823507e+00 -1.68420747e-01 5.31787761e-02
1.04489192e-01 2.17113182e-01 -2.07733989e-01 -6.39176741e-02
-6.63562655e-01 2.96689868e-01 -4.82742906e-01 4.31525677e-01
-2.81132963e-02 1.27516270e-01 -8.62427950e-01 5.78849435e-01
-3.24819498e-02 9.66889501e-01 -1.11932051e+00 -4.95091170e-01
2.21865147e-01 1.12127038e-02 -6.79477096e-01 1.47926942e-01
-3.97779018e-01 5.09526432e-01 -1.02688384e+00 2.41639510e-01
1.18958330e+00 -4.09677438e-02 5.52722037e-01 8.67154896e-02
1.30349010e-01 6.52024865e-01 -1.18572605e+00 1.76191080e+00
-7.51523376e-01 -1.29722670e-01 -4.18792009e-01 -1.64976716e+00
7.72697806e-01 3.49293172e-01 4.62013125e-01 -1.14174592e+00
2.26683259e-01 2.40118802e-01 1.98090076e-01 -4.56697822e-01
-6.55728718e-03 -8.37811649e-01 -4.87610817e-01 4.77065146e-01
-4.60585684e-01 4.72976454e-02 -3.01877975e-01 -1.98318213e-01
3.98859978e-01 1.48848772e-01 6.98718190e-01 -1.19825214e-01
4.53321278e-01 -2.53056705e-01 9.37227130e-01 6.31515145e-01
-3.63958478e-02 2.38191262e-01 4.97304350e-01 -5.09084821e-01
-7.88763344e-01 -1.23666954e+00 -2.63427585e-01 6.74232423e-01
-2.70087905e-02 4.02436584e-01 -3.15029293e-01 -9.68944430e-01
3.68092477e-01 1.05669284e+00 -8.13762009e-01 -7.72933006e-01
-6.33329511e-01 -1.15041864e+00 1.90233395e-01 5.27124643e-01
6.25970483e-01 -1.08876777e+00 7.64038367e-03 3.97656560e-01
-3.08269590e-01 -4.31282699e-01 -6.52947843e-01 -5.12805343e-01
-1.21707451e+00 -1.10135472e+00 -9.39010203e-01 -3.60082626e-01
1.97610706e-01 4.07851785e-01 1.04852378e+00 -5.77646554e-01
-2.59208344e-02 -2.57048517e-01 3.02095078e-02 -6.35903001e-01
-4.33671832e-01 -4.37075764e-01 2.76578903e-01 5.16845956e-02
5.09791553e-01 -7.12280452e-01 -1.08216643e+00 2.37156585e-01
-1.02265513e+00 -1.59701973e-01 8.77513528e-01 1.25693905e+00
5.35195827e-01 3.80299464e-02 1.34308147e+00 -1.30639160e+00
9.23294485e-01 -1.04115164e+00 -7.39798963e-01 1.51735768e-01
-8.81930411e-01 1.54816613e-01 9.21427071e-01 -8.75141382e-01
-1.53353798e+00 -5.76919019e-01 1.33092180e-01 -3.79268557e-01
1.24433525e-01 4.99846399e-01 -6.04177713e-01 7.53311217e-01
6.65451467e-01 2.18757376e-01 2.05374539e-01 -4.36226398e-01
6.88907087e-01 9.27984774e-01 4.04563129e-01 -5.62091708e-01
4.89972055e-01 6.22972369e-01 -1.89087614e-02 5.04592471e-02
-1.01593471e+00 4.12575565e-02 2.29339898e-01 4.86796796e-01
3.69633019e-01 -1.10286510e+00 -1.04337299e+00 6.22975230e-02
-8.85847509e-01 5.64464480e-02 -4.81077641e-01 9.83540356e-01
-8.41201246e-01 2.25496262e-01 7.51216263e-02 -7.72594750e-01
-2.54683465e-01 -7.66753912e-01 5.93305171e-01 1.28845736e-01
-7.29030836e-03 -9.43233550e-01 2.14277238e-01 4.23235059e-01
5.06376512e-02 5.18702030e-01 1.16019535e+00 -3.41272056e-01
-3.02209526e-01 -6.30300120e-03 -2.63055027e-01 2.30541244e-01
3.70620400e-01 -6.26601577e-01 -6.27961636e-01 -3.62280965e-01
3.36693466e-01 -1.66165784e-01 9.57005441e-01 8.52298379e-01
1.90648615e+00 -9.84542847e-01 -4.88020211e-01 4.84292358e-01
1.29582644e+00 4.38532084e-01 8.22006643e-01 1.10307820e-01
2.82434613e-01 6.01979911e-01 7.62678266e-01 5.54552794e-01
3.72636855e-01 5.87800622e-01 4.20644790e-01 -4.16239388e-02
7.12791905e-02 -8.25100183e-01 1.65117875e-01 1.60570562e-01
-3.85481901e-02 9.96351764e-02 -2.16916174e-01 6.26369596e-01
-1.81267619e+00 -1.35052049e+00 2.40593642e-01 2.53844762e+00
1.09313130e+00 1.34457424e-01 3.32338512e-01 5.62355146e-02
9.10358846e-01 3.75222147e-01 -1.02801263e+00 -4.28677499e-01
5.92167005e-02 2.63169557e-01 7.42921174e-01 -2.35829949e-01
-9.42775488e-01 4.25701588e-01 5.87156343e+00 1.21497226e+00
-9.78003263e-01 1.19210482e-01 9.65894341e-01 9.08350199e-02
-8.94353092e-01 -7.43491054e-02 -1.35899916e-01 1.04020858e+00
8.25333714e-01 -7.19996631e-01 3.82203281e-01 8.61269295e-01
5.62742114e-01 4.33690220e-01 -1.00756145e+00 9.30610061e-01
-2.68026680e-01 -1.63923764e+00 4.49646205e-01 1.69707671e-01
1.14025116e+00 -4.39357489e-01 4.69735086e-01 6.88623428e-01
7.01438844e-01 -1.16960144e+00 2.84984797e-01 2.39945039e-01
1.22056067e+00 -8.28732431e-01 7.67293453e-01 3.45865130e-01
-6.33600652e-01 -4.78836477e-01 -5.84411085e-01 -1.58094555e-01
-1.13647640e-01 5.56086242e-01 -5.63216984e-01 8.22247803e-01
3.20849210e-01 8.02727461e-01 1.22847378e-01 8.06969225e-01
-2.60878980e-01 7.49042392e-01 1.76031381e-01 -9.82496813e-02
2.40811147e-03 -1.71232373e-01 2.51576126e-01 5.69184244e-01
3.29915047e-01 1.10093169e-01 7.34481290e-02 9.68628585e-01
-5.41506767e-01 2.53862739e-01 -9.40202117e-01 1.22817732e-01
6.17073238e-01 5.47699988e-01 7.49626309e-02 -2.67248482e-01
-5.50603151e-01 4.25261736e-01 1.24261761e-02 3.98186594e-01
-1.00216711e+00 1.22624426e-03 7.64187515e-01 6.89600408e-02
2.13488430e-01 5.16512513e-01 -3.42439473e-01 -1.29404926e+00
3.30839939e-02 -1.31909394e+00 8.47619653e-01 -3.10912699e-01
-1.73043215e+00 -1.65111750e-01 2.59054005e-01 -1.73543322e+00
-4.58711237e-01 -4.40785646e-01 -7.04011917e-01 7.42097437e-01
-1.70450711e+00 -9.02173460e-01 4.03274298e-01 5.65416873e-01
3.10073495e-01 -2.17087656e-01 7.34083414e-01 1.84625864e-01
-4.36020911e-01 6.86315596e-01 4.97949302e-01 -1.53810857e-02
5.64464927e-01 -8.68698359e-01 2.25190789e-01 2.64597237e-01
-1.64157450e-01 8.12461793e-01 5.26448727e-01 -5.57130098e-01
-1.20978343e+00 -1.18488991e+00 4.89037603e-01 -2.50854325e-02
7.18438864e-01 -1.22400410e-01 -6.70740902e-01 6.00466192e-01
-1.29843056e-01 -6.10766048e-03 8.08581769e-01 3.19935173e-01
-7.07501531e-01 -4.80023950e-01 -1.37898135e+00 1.00006020e+00
1.21756744e+00 -2.06599042e-01 -1.14908075e+00 3.12452823e-01
9.66745138e-01 -2.65307814e-01 -6.36741102e-01 5.86535633e-01
6.86447144e-01 -4.90865231e-01 1.29707634e+00 -1.26729453e+00
7.14836836e-01 -5.71036078e-02 1.11772224e-01 -1.62191284e+00
-2.53338069e-01 -7.83613086e-01 -1.78143501e-01 9.72017109e-01
1.84103370e-01 -1.01804531e+00 6.75335765e-01 2.87655979e-01
3.46985817e-01 -6.56060934e-01 -1.21838152e+00 -8.74638796e-01
6.07508004e-01 -3.64035040e-01 1.29617679e+00 1.18378901e+00
-8.43961090e-02 3.78053397e-01 -8.18564236e-01 -1.77708283e-01
7.66391218e-01 7.82304585e-01 6.67069018e-01 -1.21469843e+00
-3.30093980e-01 -4.80536938e-01 -1.05816320e-01 -1.06771302e+00
3.29573482e-01 -8.17114711e-01 -4.23255950e-01 -1.11716950e+00
4.56804335e-01 -5.00070632e-01 -7.00655341e-01 3.63834128e-02
-4.76639390e-01 -1.41677931e-01 -8.82034898e-02 -4.30152155e-02
-1.53163776e-01 1.03543389e+00 1.73495698e+00 -3.69688541e-01
-6.88046962e-02 3.88327092e-01 -1.49652803e+00 6.59386337e-01
8.04046690e-01 -6.47356331e-01 -6.62308216e-01 7.33631998e-02
3.17493528e-02 3.74274611e-01 4.52219695e-01 -3.09499145e-01
-5.52966237e-01 -7.74123490e-01 4.72247154e-01 -1.85805067e-01
-2.44208187e-01 -8.36671352e-01 4.11269106e-02 7.72288918e-01
-5.75159431e-01 -2.23396659e-01 7.25538880e-02 1.03255391e+00
-3.24738212e-02 6.26309589e-02 6.25032842e-01 -1.43711567e-01
-5.27964950e-01 5.76892078e-01 2.29419976e-01 4.96528655e-01
8.94068718e-01 1.59595102e-01 -4.92121845e-01 -3.42447668e-01
-2.44120553e-01 4.64743227e-01 -5.99462278e-02 7.37473249e-01
6.40509367e-01 -1.82689321e+00 -1.08836329e+00 1.14291519e-01
1.87294438e-01 -1.70917317e-01 7.90354490e-01 5.40484726e-01
2.66941160e-01 4.09598321e-01 -1.54330224e-01 -1.22002773e-02
-5.20910621e-01 1.18140364e+00 1.03618521e-02 -6.62537098e-01
-4.51142997e-01 6.73606172e-02 6.35034084e-01 -5.83251774e-01
-2.07309738e-01 -9.03444737e-02 -2.92421401e-01 -3.15885209e-02
6.95258021e-01 4.42355752e-01 -3.25372636e-01 -9.54398140e-02
-2.23349109e-01 1.34632438e-01 -3.65593106e-01 1.27412453e-01
1.40686250e+00 1.00686423e-01 2.57099509e-01 2.46119082e-01
1.25878704e+00 -1.29098371e-01 -1.18029380e+00 -2.75999814e-01
-1.46742031e-01 -8.56035292e-01 -8.15494359e-02 -8.46727252e-01
-8.11599195e-01 6.73626423e-01 7.52779305e-01 1.67422369e-01
1.19873595e+00 -1.66494787e-01 7.16803849e-01 -3.42351361e-03
3.92743260e-01 -1.05571496e+00 -2.09437817e-01 -3.18943501e-01
1.06201482e+00 -1.38449109e+00 3.89110625e-01 -1.84926495e-01
-7.14145124e-01 2.99026281e-01 2.13729978e-01 -4.20872778e-01
5.77860236e-01 -3.70032102e-01 -8.80487114e-02 1.88500345e-01
-5.97151875e-01 2.01549947e-01 2.15977654e-01 8.95478308e-01
4.79957223e-01 5.68866253e-01 -1.00735462e+00 9.83601749e-01
-1.84562072e-01 5.11515677e-01 2.92579681e-01 2.77437210e-01
3.61588746e-01 -1.14701033e+00 -3.48621964e-01 7.70660698e-01
-6.80590987e-01 -1.32537335e-01 2.40629330e-01 1.13435006e+00
1.30255535e-01 6.90289676e-01 -2.27588993e-02 3.76060419e-02
5.09287655e-01 -2.24787712e-01 3.18748504e-01 -3.54961455e-01
-4.44688536e-02 -8.69833604e-02 -1.47721440e-01 -5.60388327e-01
-6.24858201e-01 -8.33728015e-01 -8.03164661e-01 -3.60439658e-01
-2.42872626e-01 2.81920403e-01 2.76643187e-01 7.14316607e-01
3.22928250e-01 6.32976413e-01 1.28200626e+00 -2.44187921e-01
-1.40151584e+00 -9.68594611e-01 -6.40723825e-01 7.67214119e-01
4.08023536e-01 -9.46557760e-01 -3.59286815e-01 -3.60726416e-01] | [8.111063003540039, 5.465070724487305] |
531c9865-c1fa-4bd7-b58c-702545fafc28 | multi-field-models-in-neural-recipe-ranking | 2105.05710 | null | https://arxiv.org/abs/2105.05710v2 | https://arxiv.org/pdf/2105.05710v2.pdf | Evaluation of Field-Aware Neural Ranking Models for Recipe Search | Explicitly modelling field interactions and correlations in complex document structures has recently gained popularity in neural document embedding and retrieval tasks. Although this requires the specification of bespoke task-dependent models, encouraging empirical results are beginning to emerge. We present the first in-depth analyses of non-linear multi-field interaction (NL-MFI) ranking in the cooking domain in this work. Our results show that field-weighted factorisation machines models provide a statistically significant improvement over baselines in recipe retrieval tasks. Additionally, we show that sparsely capturing subsets of field interactions based on domain knowledge and feature selection heuristics offers significant advantages over baselines and exhaustive alternatives. Although field-interaction aware models are more elaborate from an architectural basis, they are often more data-efficient in optimisation and are better suited for explainability due to mirrored document and model factorisation. | ['Luis M Vaquero', 'Niall Twomey', 'Mikhail Fain', 'Kentaro Takiguchi'] | 2021-05-12 | null | null | null | null | ['document-embedding'] | ['methodology'] | [ 5.43856919e-01 -1.98450357e-01 -4.02214050e-01 -5.21412313e-01
-6.32727444e-01 -7.87123859e-01 1.11218476e+00 4.58073646e-01
-4.62881833e-01 3.08985829e-01 7.99237251e-01 -2.94406414e-01
-9.27600920e-01 -5.05292237e-01 -9.30184543e-01 -4.06702697e-01
-5.96825957e-01 7.49173880e-01 -3.76902282e-01 -5.17670155e-01
3.78315687e-01 9.82467830e-02 -1.60275924e+00 6.34814501e-01
6.89365745e-01 6.16477013e-01 5.06570399e-01 9.32087004e-01
-2.41966799e-01 8.82622361e-01 -4.02102530e-01 -3.18800449e-01
1.21467993e-01 -1.03496321e-01 -1.08369815e+00 -3.35985005e-01
8.10237944e-01 -1.92672595e-01 -2.88979769e-01 4.84275550e-01
4.08980757e-01 5.55655658e-01 8.33419204e-01 -1.05234516e+00
-1.04345047e+00 1.10927093e+00 -2.66949683e-01 2.24464744e-01
5.80354273e-01 -1.61534250e-01 1.72023690e+00 -7.66726851e-01
5.04042864e-01 1.45019364e+00 7.01666117e-01 2.59237140e-01
-1.73169661e+00 -2.45958343e-01 3.47982615e-01 2.66135961e-01
-9.69278574e-01 -2.11004972e-01 6.75921679e-01 -4.63147491e-01
1.83527875e+00 5.74071825e-01 4.36620831e-01 9.87633467e-01
2.27661535e-01 1.05260170e+00 7.45671809e-01 -9.43885982e-01
-2.26454318e-01 1.96340740e-01 5.36139727e-01 5.03055155e-01
2.74806589e-01 3.28773141e-01 -9.86426353e-01 -3.69551063e-01
4.86707985e-01 8.42499435e-02 -1.48330331e-01 -6.06987059e-01
-1.53420031e+00 1.17219365e+00 2.44900376e-01 4.62582022e-01
-5.51724732e-01 3.52811545e-01 5.05880058e-01 3.99548858e-01
7.10141301e-01 1.04372656e+00 -1.01179826e+00 -1.24551123e-02
-1.18730485e+00 8.18060696e-01 9.08503354e-01 9.74037349e-01
5.05123377e-01 -3.34781647e-01 -3.02237928e-01 1.13278711e+00
3.54970396e-01 3.76366735e-01 2.55945653e-01 -8.46173406e-01
4.89017308e-01 4.06820089e-01 -1.88085344e-02 -1.27096665e+00
-5.39403260e-01 -2.23102584e-01 -5.74477494e-01 -2.27190927e-01
2.03203604e-01 3.19331557e-01 -6.06035173e-01 1.48551285e+00
-8.48590508e-02 -2.97118723e-01 -9.36720073e-02 8.02823961e-01
6.90854669e-01 6.39137626e-01 5.50010085e-01 1.27071738e-01
1.52905154e+00 -1.32458818e+00 -8.55912328e-01 -5.55156946e-01
9.50198770e-01 -8.26768994e-01 1.10283661e+00 5.90653062e-01
-1.06579697e+00 -6.60095096e-01 -1.00845873e+00 -4.83349770e-01
-7.58723497e-01 2.09017381e-01 1.60349154e+00 5.15899062e-01
-9.92610157e-01 8.97264004e-01 -6.01871550e-01 -3.95498574e-01
-4.90251705e-02 5.85259438e-01 -4.71504003e-01 -3.29936743e-01
-1.54490697e+00 1.20483220e+00 6.60179734e-01 5.69816530e-02
-6.02166295e-01 -1.07767153e+00 -1.11669767e+00 1.28531247e-01
4.56824809e-01 -6.83692873e-01 1.14325452e+00 -5.66148698e-01
-1.09976220e+00 5.07441401e-01 -1.19782291e-01 -3.87363464e-01
-2.36282170e-01 -6.35819972e-01 -4.91598755e-01 -1.21159464e-01
-2.54868001e-01 7.39699066e-01 5.84673882e-01 -8.82455528e-01
-5.09810984e-01 -8.52022246e-02 5.99032044e-01 2.99962729e-01
-6.07501924e-01 3.88025910e-01 -2.15163559e-01 -7.39212930e-01
-3.53122205e-01 -9.01635170e-01 -2.25570053e-01 -4.98291731e-01
-2.20154658e-01 -6.04294419e-01 2.86115229e-01 -7.70810843e-01
1.68654680e+00 -1.88875175e+00 3.88699859e-01 -4.17535827e-02
2.80856162e-01 1.57904953e-01 -4.37008470e-01 8.28212500e-01
-3.04455638e-01 2.86357909e-01 1.90553531e-01 -2.61122912e-01
4.10192609e-01 2.44180918e-01 -2.01024652e-01 1.59540817e-01
3.73038322e-01 9.32317495e-01 -1.04981637e+00 -4.07278627e-01
4.76981372e-01 6.69609249e-01 -9.81931090e-01 -4.18992192e-02
-6.16252124e-01 -4.06212300e-01 -3.48335415e-01 3.91477615e-01
4.42057043e-01 -2.93115407e-01 2.92899877e-01 -3.40625733e-01
2.21838523e-02 6.90458059e-01 -1.31231666e+00 2.13684177e+00
-5.53185403e-01 6.79249585e-01 -4.56183217e-02 -9.34517324e-01
4.70268369e-01 1.86255902e-01 5.35484493e-01 -4.60290432e-01
-3.36036682e-01 -2.29694024e-02 2.35846397e-02 -3.36848199e-01
1.22545898e+00 9.84390602e-02 -1.31119922e-01 4.81170446e-01
5.77264547e-01 -2.15208139e-02 7.05117166e-01 5.21063805e-01
1.10978091e+00 5.63597679e-01 1.02944009e-01 -7.24923134e-01
2.84280956e-01 2.16377988e-01 -1.72171257e-02 1.04242623e+00
5.09259880e-01 3.76611680e-01 -4.59982753e-02 -4.23397988e-01
-8.25017869e-01 -4.57150668e-01 -1.05839573e-01 1.85404563e+00
-2.76314765e-01 -1.21477723e+00 -3.06942642e-01 -6.95289373e-01
2.39082783e-01 1.02890706e+00 -7.22859085e-01 -4.89427112e-02
-6.33859098e-01 -8.02490413e-01 3.36411774e-01 8.25776458e-01
-1.98475644e-01 -9.86695945e-01 -3.23864996e-01 5.18219471e-01
-1.41736999e-01 -4.94222343e-01 -4.90547001e-01 9.58156645e-01
-7.94631422e-01 -8.29505980e-01 -5.84391654e-01 -5.48412144e-01
2.07518339e-01 5.49320102e-01 1.69063354e+00 2.78049260e-01
-3.44071150e-01 6.28687799e-01 -5.33781052e-01 -6.30808532e-01
-2.02396303e-01 3.37778568e-01 -1.84976719e-02 -8.25363159e-01
1.01682711e+00 -1.64991692e-01 -4.84834552e-01 3.80028039e-02
-1.21867323e+00 -1.89371556e-01 5.18894553e-01 1.11477470e+00
1.21328220e-01 2.42027014e-01 -6.12194277e-02 -9.77261484e-01
1.04267514e+00 -3.51390451e-01 -3.15760225e-01 5.25663435e-01
-9.61090803e-01 5.57328463e-01 4.16252583e-01 -6.37481928e-01
-1.02684295e+00 3.38885467e-03 4.66865003e-01 -4.53143604e-02
-1.66068092e-01 8.23704898e-01 1.68295801e-01 1.76037461e-01
1.11619413e+00 -3.26406032e-01 -3.05032074e-01 -6.73778653e-01
8.53529274e-01 3.41256797e-01 1.09035961e-01 -9.64196861e-01
6.04113102e-01 -1.52731135e-01 -9.27310288e-02 -6.85589612e-01
-5.67166507e-01 -7.86440134e-01 -6.66772783e-01 1.93076149e-01
7.41200745e-01 -9.23133910e-01 -4.12514448e-01 -7.51354545e-02
-1.29498899e+00 -2.71622509e-01 -7.43671134e-02 7.10316479e-01
-3.28883380e-01 3.32605094e-01 -5.80754519e-01 -7.94390142e-01
-8.72081295e-02 -9.77228940e-01 1.29888034e+00 -1.60478473e-01
-7.19633222e-01 -1.37742913e+00 2.48319909e-01 3.02177489e-01
4.73972946e-01 -6.40477538e-02 1.29433620e+00 -9.01129961e-01
-5.05095124e-01 -1.44646659e-01 7.52765387e-02 2.59289116e-01
-5.90701662e-02 -5.46141006e-02 -8.60135436e-01 -7.63517320e-02
-4.35074657e-01 -3.12731802e-01 1.14573932e+00 4.78309631e-01
6.59044325e-01 -2.76604056e-01 -6.09448016e-01 2.97372699e-01
1.36494339e+00 6.59181774e-02 3.37417632e-01 6.64592206e-01
6.17157638e-01 9.63244796e-01 6.20198429e-01 1.28512859e-01
4.30052280e-01 7.47067094e-01 -4.78021987e-02 -2.79519081e-01
-1.00018099e-01 -3.25228870e-01 3.36648285e-01 6.13167405e-01
-7.28553981e-02 -6.46444082e-01 -8.78976882e-01 7.18297839e-01
-1.75845945e+00 -1.03102362e+00 -4.58241671e-01 1.77386475e+00
9.09964442e-01 -5.94777800e-02 1.56499535e-01 2.04529941e-01
3.44013423e-01 2.67772377e-01 -4.43212362e-03 -6.70664072e-01
-2.06479669e-01 4.31675673e-01 6.66518211e-01 5.92925251e-01
-1.20671475e+00 8.86007965e-01 7.28413486e+00 8.04271340e-01
-3.08691561e-01 -1.01130642e-01 2.65038401e-01 -3.58236611e-01
-5.99283516e-01 1.60387754e-01 -9.02379572e-01 2.47091167e-02
1.02043056e+00 3.52505207e-01 9.93384242e-01 8.02065372e-01
-2.04518080e-01 -1.64653048e-01 -1.70157075e+00 7.17864096e-01
6.06018230e-02 -1.33556306e+00 -1.45769700e-01 1.51771441e-01
4.66724187e-01 -9.14817601e-02 -8.71261358e-02 7.03225493e-01
6.89144850e-01 -1.17452765e+00 6.56085193e-01 4.59674537e-01
5.23310721e-01 -6.11253679e-01 4.49924201e-01 1.96482852e-01
-9.07781839e-01 -1.51121259e-01 -4.53458250e-01 -3.02242726e-01
8.45314115e-02 5.28629363e-01 -5.60575962e-01 7.05659509e-01
7.89775252e-01 5.14341116e-01 -7.71581948e-01 5.90581894e-01
4.57331091e-02 6.69272423e-01 -4.63338733e-01 -3.64532083e-01
3.12467247e-01 -4.35291566e-02 2.60331422e-01 1.72023380e+00
-1.65228501e-01 -1.67040929e-01 -2.13859919e-02 7.47357190e-01
5.05646467e-01 1.96211919e-01 -7.18693793e-01 -6.58322573e-01
2.31044702e-02 1.28481138e+00 -4.45695907e-01 -7.18435571e-02
-4.88659561e-01 8.16405475e-01 2.97204673e-01 2.54492909e-01
-5.80278337e-01 -4.41044569e-01 6.48190737e-01 -2.20004097e-02
5.06746233e-01 -2.87596017e-01 -1.27324358e-01 -1.07601774e+00
-6.99603781e-02 -1.17580271e+00 4.97970134e-01 -6.31397605e-01
-1.44154119e+00 3.80916774e-01 7.56726384e-01 -3.81111175e-01
-7.38403261e-01 -9.73788083e-01 -2.72312970e-03 9.49615121e-01
-1.36818635e+00 -1.38039804e+00 3.20407659e-01 2.24665046e-01
3.95904511e-01 1.22097358e-02 1.46467376e+00 3.69791955e-01
4.70266677e-02 5.08603036e-01 3.06485921e-01 -2.56207079e-01
6.80477381e-01 -1.60368097e+00 4.48527306e-01 5.23311436e-01
7.85747766e-01 1.43464720e+00 8.22452307e-01 -6.63033426e-01
-1.72636163e+00 -5.36843836e-01 1.52213836e+00 -8.32509220e-01
5.89932740e-01 -8.19045603e-01 -6.62816823e-01 7.05935299e-01
6.04307175e-01 -5.52641571e-01 9.29380536e-01 1.04390991e+00
-5.40985644e-01 1.12713583e-01 -7.49936104e-01 4.51471895e-01
1.14564824e+00 -7.78009295e-01 -6.93972707e-01 8.27371359e-01
6.47767365e-01 -1.45409852e-01 -1.08078218e+00 3.48562211e-01
6.63550258e-01 -5.09775221e-01 1.32802474e+00 -9.64204311e-01
5.98839343e-01 -9.77923274e-02 -3.48176509e-01 -1.44182336e+00
-9.98413980e-01 -6.76330745e-01 -2.60282308e-01 1.30930376e+00
6.12750232e-01 3.14892991e-03 5.58130801e-01 9.46287394e-01
1.61190182e-01 -4.72761631e-01 -1.04243726e-01 -6.51005745e-01
2.14363948e-01 -4.98923868e-01 5.63771427e-01 9.89252031e-01
1.84763193e-01 7.07640290e-01 -5.19089341e-01 -1.89726457e-01
2.80221492e-01 -1.66883379e-01 6.25649631e-01 -1.31293797e+00
-8.42953861e-01 -6.04457259e-01 -1.08778425e-01 -1.11811316e+00
3.00634265e-01 -1.22492540e+00 1.63367782e-02 -1.70487940e+00
2.95311421e-01 -1.36629567e-01 -5.61277688e-01 4.28550363e-01
-3.28451633e-01 -1.60890341e-01 1.50984034e-01 -1.98084250e-01
-6.86352849e-01 5.53393513e-02 6.51265562e-01 -4.78236318e-01
-6.99529722e-02 -2.78159887e-01 -9.17437375e-01 2.82178879e-01
5.79559863e-01 -5.62223852e-01 -7.16807723e-01 -8.87999237e-01
6.40239120e-01 -2.46177673e-01 2.34590665e-01 -4.51687306e-01
8.53023008e-02 -1.45202681e-01 5.32761931e-01 -5.32726824e-01
3.37218285e-01 -8.78121138e-01 2.57458419e-01 -8.64018574e-02
-1.00045514e+00 2.24785909e-01 4.99298662e-01 6.21229231e-01
6.81117326e-02 -6.81322932e-01 -2.29437411e-01 -4.19283211e-01
-8.02852988e-01 -1.51021853e-01 -5.24716556e-01 -1.20043285e-01
2.27146000e-01 -5.16048037e-02 -9.01861712e-02 -2.65962154e-01
-2.75644600e-01 -2.12477520e-02 8.89998451e-02 7.17739582e-01
9.40644592e-02 -1.35577798e+00 -5.92956960e-01 4.23318632e-02
2.73211449e-01 -1.56549186e-01 1.29224718e-01 4.00267482e-01
-2.10943267e-01 1.12356138e+00 3.10172200e-01 -4.24002051e-01
-1.19888675e+00 8.03810477e-01 -2.24016950e-01 -8.69582951e-01
-2.76591778e-01 1.10415328e+00 1.17003992e-01 -6.43860519e-01
4.65519309e-01 -5.19259095e-01 -2.07877025e-01 -3.44489850e-02
6.34760499e-01 3.56359273e-01 2.63491213e-01 -3.74068111e-01
-3.48882020e-01 1.75363883e-01 -4.65424746e-01 -1.93410605e-01
1.60601580e+00 3.22079957e-01 -1.01137169e-01 1.65931612e-01
1.24352884e+00 -1.62114605e-01 -8.48639667e-01 -3.82180333e-01
4.92587149e-01 -3.89120132e-01 6.03060782e-01 -1.35355031e+00
-4.84142751e-01 7.62418330e-01 3.37337852e-01 3.12293857e-01
9.59013283e-01 -6.16646148e-02 5.11036873e-01 7.46288896e-01
5.09936437e-02 -1.16982925e+00 -2.46418044e-01 5.63650787e-01
8.71987641e-01 -9.27764893e-01 4.15757567e-01 -9.71340686e-02
-6.02618873e-01 1.13259494e+00 3.26186061e-01 3.81812104e-03
7.45563745e-01 3.19784343e-01 -2.44760826e-01 -4.42675561e-01
-8.66857648e-01 -4.50511239e-02 8.11563551e-01 9.35973883e-01
9.96152639e-01 2.31364649e-02 -4.06668454e-01 5.97174823e-01
-1.73118994e-01 -2.46439025e-01 -2.58737743e-01 9.95948613e-01
-2.31852178e-02 -1.59799969e+00 -1.06200561e-01 6.03757501e-01
-5.60105443e-01 -7.69964516e-01 -4.54353243e-01 9.53313947e-01
6.83210716e-02 1.31276631e+00 -2.83940852e-01 -3.91510487e-01
2.94303417e-01 3.38490963e-01 8.42446148e-01 -6.70166194e-01
-1.08466220e+00 1.12280399e-01 4.01583374e-01 -4.83376831e-01
-6.73579991e-01 -7.11268604e-01 -7.34510362e-01 -1.37982056e-01
-9.67887342e-01 2.68567145e-01 9.95094538e-01 7.87838340e-01
4.17406052e-01 6.35360003e-01 -3.32599580e-02 -9.65186656e-01
-7.47619450e-01 -1.34735334e+00 -5.24346232e-01 4.11141902e-01
1.51536882e-01 -8.91491175e-01 -2.71439910e-01 1.04612395e-01] | [11.350522994995117, 7.73576021194458] |
237abd9b-e881-4935-bc6d-67226b25dbe3 | samanantar-the-largest-publicly-available | 2104.05596 | null | https://arxiv.org/abs/2104.05596v4 | https://arxiv.org/pdf/2104.05596v4.pdf | Samanantar: The Largest Publicly Available Parallel Corpora Collection for 11 Indic Languages | We present Samanantar, the largest publicly available parallel corpora collection for Indic languages. The collection contains a total of 49.7 million sentence pairs between English and 11 Indic languages (from two language families). Specifically, we compile 12.4 million sentence pairs from existing, publicly-available parallel corpora, and additionally mine 37.4 million sentence pairs from the web, resulting in a 4x increase. We mine the parallel sentences from the web by combining many corpora, tools, and methods: (a) web-crawled monolingual corpora, (b) document OCR for extracting sentences from scanned documents, (c) multilingual representation models for aligning sentences, and (d) approximate nearest neighbor search for searching in a large collection of sentences. Human evaluation of samples from the newly mined corpora validate the high quality of the parallel sentences across 11 languages. Further, we extract 83.4 million sentence pairs between all 55 Indic language pairs from the English-centric parallel corpus using English as the pivot language. We trained multilingual NMT models spanning all these languages on Samanantar, which outperform existing models and baselines on publicly available benchmarks, such as FLORES, establishing the utility of Samanantar. Our data and models are available publicly at https://ai4bharat.iitm.ac.in/samanantar and we hope they will help advance research in NMT and multilingual NLP for Indic languages. | ['Mitesh Shantadevi Khapra', 'Kumar Deepak', 'Srihari Nagaraj', 'Pratyush Kumar', 'Anoop Kunchukuttan', 'Vivek Raghavan', 'Aswin Pradeep', 'Navneet Kumar', 'Divyanshu Kakwani', 'Mahalakshmi J', 'Harshita Diddee', 'Sujit Sahoo', 'Ajitesh Sharma', 'Raghavan AK', 'Mayank Jobanputra', 'Aravinth Bheemaraj', 'Sumanth Doddapaneni', 'Gowtham Ramesh'] | 2021-04-12 | null | null | null | null | ['multilingual-nlp'] | ['natural-language-processing'] | [-1.78037643e-01 -5.81971526e-01 -3.42065305e-01 -3.16021800e-01
-1.62039602e+00 -1.13438869e+00 5.06573856e-01 5.35005666e-02
-5.73642373e-01 8.86346757e-01 5.20236731e-01 -5.36878824e-01
2.24599898e-01 -6.69739962e-01 -8.71437848e-01 -2.17632055e-01
-7.60103613e-02 8.55070829e-01 -7.39300251e-02 -6.73664749e-01
3.89377177e-01 1.27566829e-01 -7.46321261e-01 6.12145364e-01
1.20158410e+00 1.96092308e-01 5.58534980e-01 6.95505857e-01
-6.13642275e-01 3.13074440e-01 -6.86988652e-01 -7.39431381e-01
3.03552657e-01 -2.35907063e-01 -1.21296358e+00 -2.66948432e-01
8.06438386e-01 -7.73234293e-02 -7.76004866e-02 9.04863954e-01
4.16699320e-01 -5.31381309e-01 2.55726695e-01 -8.14993143e-01
-8.27024162e-01 1.46553075e+00 -8.92084897e-01 5.13279617e-01
4.41944778e-01 -2.22641341e-02 1.38392246e+00 -1.10548913e+00
1.22531295e+00 1.14746559e+00 5.78948796e-01 1.82015404e-01
-8.80822241e-01 -6.68223500e-01 -3.88302177e-01 2.63365328e-01
-1.44778252e+00 -6.55456424e-01 5.08283138e-01 -1.98038951e-01
1.41722953e+00 4.63276595e-01 5.78266859e-01 1.17626321e+00
2.40914121e-01 1.00164354e+00 1.36718071e+00 -9.18164849e-01
-4.71614182e-01 7.03649246e-04 3.54626447e-01 3.29100400e-01
-4.04303968e-02 -4.55312133e-01 -5.61071575e-01 -2.30131254e-01
1.68407068e-01 -5.71504235e-01 3.17393318e-02 5.47274530e-01
-1.43598711e+00 6.95698380e-01 -7.87975043e-02 6.62137032e-01
-1.49879470e-01 -4.90874231e-01 8.20017099e-01 5.62225878e-01
5.08741617e-01 3.26612204e-01 -7.36786962e-01 -2.46206135e-01
-7.59266436e-01 1.70753971e-01 8.19359541e-01 1.32308829e+00
7.22117066e-01 -1.73963130e-01 1.90265641e-01 1.43844807e+00
8.48872960e-02 1.08554268e+00 6.03676140e-01 -7.02684700e-01
1.33760679e+00 2.97941327e-01 -4.31142360e-01 -7.47735918e-01
-6.25424311e-02 -6.11579046e-02 -4.85776216e-01 -8.45585227e-01
3.61986548e-01 -6.41645938e-02 -3.00838679e-01 1.60604715e+00
2.58798540e-01 -6.44036114e-01 4.23082471e-01 4.53251034e-01
8.07318449e-01 1.10916936e+00 -3.55261713e-01 -1.61897242e-01
1.57127106e+00 -1.15069234e+00 -4.29156542e-01 -5.43949783e-01
7.89110363e-01 -1.73752046e+00 1.25504911e+00 2.41409883e-01
-1.18726456e+00 -5.14404714e-01 -9.41869974e-01 -2.98813701e-01
-3.75679612e-01 3.78894210e-01 4.85066473e-01 2.63703108e-01
-1.04360807e+00 1.49482280e-01 -5.09026945e-01 -7.57931888e-01
-8.58511552e-02 -1.12450406e-01 -6.19560063e-01 -2.69004643e-01
-1.30682313e+00 9.47092474e-01 4.51494485e-01 -3.05493940e-02
-3.15132737e-01 -4.45416868e-01 -7.86488652e-01 -4.96346116e-01
-4.97712567e-02 -2.01587871e-01 1.16529799e+00 -6.47002459e-01
-1.21484160e+00 1.31694365e+00 -2.99150586e-01 -5.10979891e-01
2.61365861e-01 -5.29672742e-01 -7.31940031e-01 6.89020976e-02
7.06017613e-01 6.73822761e-01 1.71133652e-01 -7.64907479e-01
-6.69645607e-01 -2.86700994e-01 -4.17517841e-01 9.55822170e-02
-4.13442880e-01 6.37680173e-01 -7.35346913e-01 -7.72846222e-01
6.11209050e-02 -1.05074251e+00 -8.09011012e-02 -7.96201050e-01
-8.28076839e-01 -2.81858116e-01 6.57528281e-01 -1.46585321e+00
1.14457905e+00 -1.90055788e+00 -9.92198735e-02 -7.45803490e-02
-3.06987375e-01 1.21135063e-01 -7.22126424e-01 1.13184130e+00
9.14205834e-02 2.97108710e-01 -2.18445837e-01 -3.30239475e-01
-1.77465349e-01 3.78359079e-01 -5.04589319e-01 3.97199303e-01
2.12708279e-01 8.84486318e-01 -8.11969042e-01 -8.20932984e-01
-9.24847126e-02 9.78132188e-02 -2.28616938e-01 4.11882326e-02
-6.83960551e-03 4.86980945e-01 9.98137519e-03 6.98450029e-01
6.23207092e-01 2.61625588e-01 5.27470767e-01 6.57112002e-02
-4.86630678e-01 1.13228917e+00 -5.96990764e-01 1.93227375e+00
-6.85387194e-01 8.15463662e-01 -1.30810887e-01 -6.34942770e-01
1.01933420e+00 1.91909969e-01 1.63165540e-01 -7.21770048e-01
-1.50141314e-01 6.22109354e-01 2.84739763e-01 -4.13467973e-01
9.49483573e-01 3.68589789e-01 -6.51077986e-01 5.25733054e-01
1.72018901e-01 -3.28460425e-01 9.91946816e-01 3.91215682e-01
8.75026166e-01 -7.73553271e-03 4.43799555e-01 -4.86254007e-01
5.59568524e-01 4.60534543e-01 7.52033770e-01 5.18105030e-01
8.30523819e-02 4.79852915e-01 2.75762677e-01 -3.60427499e-01
-1.64264977e+00 -1.17007148e+00 -4.67591703e-01 9.05354321e-01
-4.68317896e-01 -8.26412439e-01 -6.18833661e-01 -6.28320634e-01
-3.33485276e-01 6.01278543e-01 -1.01258777e-01 6.66253090e-01
-1.37909651e+00 -8.80422831e-01 9.86854017e-01 1.48198307e-01
4.45076227e-01 -1.10092068e+00 3.80612254e-01 1.89162746e-01
-8.92163277e-01 -1.38894343e+00 -8.29567254e-01 -8.04396942e-02
-5.88672400e-01 -1.05439770e+00 -4.49985653e-01 -1.12356174e+00
3.36564958e-01 1.44486278e-01 1.48823953e+00 -3.63136888e-01
-1.07986502e-01 -8.30686614e-02 -3.95823061e-01 -2.13412821e-01
-9.48367834e-01 4.23655361e-01 2.65566796e-01 -6.76229298e-01
5.76521754e-01 -4.61889029e-01 4.67876494e-02 -7.23252296e-02
-5.40428758e-01 1.91476986e-01 6.56965137e-01 7.58790135e-01
6.73658192e-01 -4.35227305e-01 5.63816845e-01 -9.83276188e-01
5.37183106e-01 -6.74616635e-01 -7.97230065e-01 3.90979409e-01
-1.12577520e-01 -1.95975959e-01 8.90284240e-01 -2.53310442e-01
-9.10189331e-01 -2.77693987e-01 -5.98670363e-01 3.37010264e-01
3.72139439e-02 5.96851349e-01 -1.33816466e-01 4.91617501e-01
5.16681135e-01 6.05445445e-01 -3.34782958e-01 -6.32673860e-01
5.91283679e-01 1.22347021e+00 6.91109598e-01 -9.23295438e-01
8.59676540e-01 1.20121367e-01 -5.24047732e-01 -1.06473053e+00
-6.19466305e-01 -4.15434778e-01 -8.69637311e-01 2.06876203e-01
6.09726131e-01 -9.40092444e-01 -1.90667868e-01 4.35508698e-01
-1.41400099e+00 4.46291221e-03 2.32443642e-02 4.57090050e-01
-1.99424744e-01 6.47167921e-01 -1.34585035e+00 -3.06765079e-01
-9.51736510e-01 -1.10265195e+00 9.74335611e-01 -6.55860361e-03
-6.12931311e-01 -8.31798077e-01 6.03747606e-01 7.21245050e-01
7.07855970e-02 -1.86675638e-01 1.10885596e+00 -8.28566313e-01
-2.33984753e-01 -6.03475682e-02 -6.04575593e-03 4.27482277e-01
2.44589597e-01 3.51439565e-01 -4.79968309e-01 -2.97266185e-01
-3.52899432e-01 -5.91427207e-01 3.57402653e-01 1.03994697e-01
5.39550781e-01 -5.43775499e-01 -4.79707308e-02 3.61057997e-01
1.16182756e+00 2.19696779e-02 3.13618988e-01 6.71908855e-01
5.52624345e-01 6.89052463e-01 5.37573993e-01 6.41369075e-02
7.57747948e-01 5.09698868e-01 -3.00600439e-01 5.21638617e-02
-1.91278279e-01 -4.73898739e-01 8.27311993e-01 2.35955596e+00
4.52800572e-01 9.00458079e-03 -1.30156493e+00 9.14582610e-01
-1.38749516e+00 -7.80153930e-01 -4.33096647e-01 1.88917673e+00
1.50651991e+00 5.78666776e-02 5.93301058e-02 -2.50331283e-01
7.46717215e-01 1.80159345e-01 -1.23230524e-01 -6.20437086e-01
-8.09259772e-01 3.32784534e-01 4.65874255e-01 6.38877749e-01
-1.01679242e+00 1.51114154e+00 6.47481441e+00 1.10377657e+00
-9.16135252e-01 1.93202078e-01 6.10884547e-01 -2.38999315e-02
-3.93863022e-01 2.99255371e-01 -1.34726524e+00 5.41221142e-01
1.25511158e+00 -4.70115840e-01 4.56564069e-01 7.91465580e-01
9.99526381e-02 1.56969279e-01 -8.88963044e-01 6.85966611e-01
9.82204080e-02 -1.50847673e+00 9.18702707e-02 -2.96705752e-03
8.42757881e-01 9.92256224e-01 -1.86403498e-01 4.30867612e-01
6.64827347e-01 -7.55669594e-01 7.72284329e-01 -3.14845587e-04
8.21261764e-01 -7.45014429e-01 6.73568666e-01 4.45168793e-01
-1.21425569e+00 4.68764603e-01 -5.76889217e-01 7.51494989e-02
3.36480558e-01 7.37370014e-01 -8.81091774e-01 7.24515021e-01
8.52875590e-01 9.32716370e-01 -7.34284461e-01 4.89678413e-01
-4.06860352e-01 1.00660193e+00 -5.06307483e-01 -1.22871548e-01
3.18721652e-01 -4.80539918e-01 8.76321793e-01 1.64779413e+00
2.44613037e-01 -5.46813786e-01 2.46460900e-01 5.33930004e-01
-2.79102623e-01 7.00547695e-01 -8.15018892e-01 -2.72734135e-01
9.76200461e-01 1.46343935e+00 -5.03837168e-01 -3.61655921e-01
-6.06829345e-01 9.95200455e-01 8.15353036e-01 1.11467406e-01
-4.48601246e-01 -5.20854712e-01 5.77612698e-01 -3.24210614e-01
-2.96304896e-02 -5.47818840e-01 -9.90120471e-02 -1.32663560e+00
3.76928687e-01 -1.38713872e+00 4.90120351e-01 -5.38056731e-01
-1.87837422e+00 9.86753166e-01 -1.00418709e-01 -1.22595930e+00
-5.51594257e-01 -5.29653549e-01 -5.16215086e-01 1.12557566e+00
-1.26645696e+00 -1.34939682e+00 4.97364789e-01 2.88124353e-01
9.54006851e-01 -4.48923856e-01 8.98688138e-01 5.69640934e-01
-7.33898938e-01 5.60832143e-01 3.72762769e-01 6.28846347e-01
1.01409948e+00 -1.13437462e+00 1.12095261e+00 9.81916130e-01
6.60031855e-01 9.03964937e-01 3.18477154e-01 -7.55203009e-01
-1.65217936e+00 -1.05386305e+00 1.65006948e+00 -6.09718680e-01
1.26706994e+00 -7.17730224e-01 -6.20817900e-01 9.79239404e-01
7.94030666e-01 -5.20447195e-01 8.97546232e-01 2.92010039e-01
-4.71519083e-01 -1.15536094e-01 -6.27310693e-01 9.31266189e-01
8.54791403e-01 -6.83407784e-01 -8.80415499e-01 7.35236466e-01
6.95331454e-01 -3.38145137e-01 -1.09165442e+00 1.12925254e-01
4.17419702e-01 -5.01413584e-01 7.67342269e-01 -5.13415992e-01
8.17950666e-01 -6.23732023e-02 -4.65968281e-01 -1.37013483e+00
-7.40487408e-03 -5.66259861e-01 5.29055893e-01 1.80800080e+00
7.42608786e-01 -5.40240407e-01 2.61210412e-01 -3.66662949e-01
-2.99120039e-01 -5.86952627e-01 -9.51430321e-01 -9.04263139e-01
6.86272025e-01 -5.35491109e-01 5.40846825e-01 1.17235708e+00
1.77936360e-01 8.31627309e-01 -3.72202784e-01 3.89973749e-03
4.05804425e-01 3.93430382e-01 8.57300818e-01 -5.60184598e-01
-4.78012174e-01 -3.05020541e-01 1.00861721e-01 -1.07106137e+00
5.35333812e-01 -1.53074670e+00 -8.66481662e-02 -1.39561653e+00
5.04576385e-01 -5.68116426e-01 3.87724727e-01 5.70760250e-01
-2.07033023e-01 3.56418252e-01 3.20193805e-02 5.45032084e-01
-2.44335517e-01 4.20415670e-01 7.74142623e-01 -2.07416698e-01
3.14214714e-02 -5.52913547e-01 -4.92267102e-01 6.69780314e-01
1.07715392e+00 -5.46393394e-01 1.87796414e-01 -9.04306412e-01
2.93098807e-01 -1.52023524e-01 -4.64383096e-01 -5.07672489e-01
2.61276984e-03 -2.22508878e-01 2.84812838e-01 -9.20043409e-01
1.21741436e-01 -2.58054435e-01 -1.57712579e-01 4.03653979e-01
7.17307925e-02 7.85828710e-01 4.02475893e-01 -2.70250648e-01
-3.87039423e-01 -2.12719172e-01 5.15998065e-01 -2.40045041e-01
-6.37288153e-01 -2.51767633e-04 -4.46804553e-01 4.82301623e-01
5.00119746e-01 2.96986520e-01 -8.38464797e-01 -2.88553517e-02
-6.82267081e-03 3.12771678e-01 3.52479577e-01 7.25118041e-01
1.63286567e-01 -1.25032914e+00 -1.33095014e+00 2.06418648e-01
2.67816067e-01 -4.38103855e-01 -7.62445852e-02 7.53771961e-01
-8.56149793e-01 5.55484295e-01 -2.64630169e-01 -6.32770658e-01
-1.29935336e+00 1.02999561e-01 -2.92162150e-01 -4.83286113e-01
-5.15901268e-01 5.92707336e-01 -3.69182438e-01 -1.31395006e+00
-4.56573993e-01 1.42957792e-01 -9.32678208e-02 -3.23668271e-01
4.03446019e-01 1.88290551e-01 8.90825093e-02 -1.00225425e+00
-4.57076371e-01 5.32047212e-01 -5.76459408e-01 -3.93611491e-01
1.36492836e+00 -3.16339880e-01 -9.53751862e-01 5.60837269e-01
1.16817749e+00 8.75753522e-01 -2.23655656e-01 -2.51469195e-01
4.81841236e-01 -2.67172128e-01 -6.00339592e-01 -6.36331677e-01
-5.56096196e-01 5.43551743e-01 -1.17805175e-01 -2.34777302e-01
9.31019902e-01 2.06619143e-01 1.32145560e+00 4.50744987e-01
3.69431347e-01 -1.22892010e+00 -3.51705402e-01 1.18894482e+00
8.64554882e-01 -1.22561955e+00 -2.01311603e-01 -4.42959309e-01
-4.10949826e-01 1.04749417e+00 5.03447890e-01 -1.48017509e-02
2.76888400e-01 7.03799605e-01 4.78341043e-01 1.92671314e-01
-7.81474173e-01 1.47585601e-01 1.74524307e-01 2.37105340e-01
8.92272830e-01 1.90038085e-01 -9.39638674e-01 5.69926262e-01
-1.08935928e+00 -7.11517692e-01 4.24162924e-01 8.84146988e-01
-5.54764085e-02 -1.77399254e+00 -3.48162502e-01 2.88151592e-01
-6.34544909e-01 -8.54196548e-01 -5.32653928e-01 9.65281129e-01
-7.80802891e-02 1.02312136e+00 2.51499861e-01 -1.92778319e-01
1.71409491e-02 1.09151984e-03 4.00947601e-01 -5.44704378e-01
-6.13714576e-01 2.00619146e-01 5.50300658e-01 -1.84541166e-01
-5.36663923e-03 -8.52232873e-01 -9.62499559e-01 -5.96373677e-01
-1.85064804e-02 4.48684782e-01 6.98765934e-01 9.09727991e-01
2.89682359e-01 -1.03776924e-01 7.05705822e-01 -4.45356399e-01
-3.69706482e-01 -1.42553842e+00 -1.56653300e-01 1.00814916e-01
-2.31088400e-01 3.95403683e-01 -1.16558604e-01 4.87494236e-03] | [11.448127746582031, 10.285632133483887] |
347093bc-bb43-4369-a09b-318a933e2f65 | polite-emotional-dialogue-acts-for | 2112.13572 | null | https://arxiv.org/abs/2112.13572v2 | https://arxiv.org/pdf/2112.13572v2.pdf | Polite Emotional Dialogue Acts for Conversational Analysis in Daily Dialog Data | Many socio-linguistic cues are used in the conversational analysis, such as emotion, sentiment, and dialogue acts. One of the fundamental social cues is politeness, which linguistically possesses properties useful in conversational analysis. This short article presents some of the brief findings of polite emotional dialogue acts, where we can correlate the relational bonds between these socio-linguistics cues. We found that the utterances with emotion classes Anger and Disgust are more likely to be impolite while Happiness and Sadness to be polite. Similar phenomenon occurs with dialogue acts, Inform and Commissive contain many polite utterances than Question and Directive. Finally, we will conclude on the future work of these findings. | ['Chandrakant Bothe'] | 2021-12-27 | null | null | null | null | ['emotional-dialogue-acts'] | ['natural-language-processing'] | [-3.96186084e-01 7.15963304e-01 -2.28385434e-01 -7.29350805e-01
1.09064169e-01 -4.99576956e-01 8.23533595e-01 4.19266909e-01
8.48323554e-02 8.39177728e-01 1.24901128e+00 2.72295713e-01
-9.11681131e-02 -6.14309788e-01 4.96755451e-01 -5.57603121e-01
7.16694817e-02 1.46149412e-01 -5.72831750e-01 -1.06169915e+00
6.84934914e-01 1.62556753e-01 -1.10997641e+00 6.33296728e-01
3.81980419e-01 5.69526434e-01 -4.70662594e-01 6.55742526e-01
-4.85222727e-01 1.35741925e+00 -6.76866591e-01 -8.37457061e-01
-5.47369957e-01 -6.65785849e-01 -1.20209587e+00 1.37182042e-01
-2.07165286e-01 -2.75095198e-02 9.22114775e-02 9.49400544e-01
4.29441601e-01 2.31287554e-01 8.87972593e-01 -1.36480951e+00
-4.04375285e-01 9.51090157e-01 -1.17269859e-01 1.62186265e-01
1.19525254e+00 6.66276831e-03 1.13443816e+00 -3.80445629e-01
8.39585304e-01 2.11168265e+00 4.93913174e-01 7.83607185e-01
-6.52368307e-01 -5.46560347e-01 -2.58128345e-01 -3.00638080e-01
-3.21675062e-01 -6.91723764e-01 9.59496617e-01 -5.08983433e-01
1.07774770e+00 4.65956658e-01 8.72418642e-01 1.19743299e+00
4.19168830e-01 6.80192590e-01 1.53068256e+00 -2.83188909e-01
-2.19403595e-01 8.32090318e-01 9.10487950e-01 2.84825236e-01
-4.59609449e-01 -2.47418299e-01 -3.34807158e-01 -6.71729088e-01
5.74894771e-02 -2.79188663e-01 -1.12190388e-01 5.29501259e-01
-7.74547815e-01 1.32148242e+00 2.46036481e-02 9.16898429e-01
-6.63514376e-01 -1.53697580e-01 9.66798604e-01 7.85922945e-01
7.52695918e-01 6.05040014e-01 -2.86637664e-01 -1.20415092e+00
1.25852421e-01 1.24552362e-01 1.52485418e+00 4.19448048e-01
6.41236186e-01 -1.78956553e-01 -1.53312624e-01 1.28843927e+00
4.78588551e-01 3.87586862e-01 3.56477022e-01 -1.06756747e+00
-5.30773774e-02 6.72260582e-01 2.20006313e-02 -1.58391356e+00
-4.94183034e-01 5.22834241e-01 -2.78950959e-01 -2.32943505e-01
2.24256858e-01 -4.63676661e-01 2.35205591e-01 1.70296383e+00
3.97651792e-01 -7.54123271e-01 7.29972184e-01 6.37957454e-01
1.43676710e+00 6.75089836e-01 2.94469923e-01 -6.97491765e-01
1.87572217e+00 -6.03488445e-01 -1.61690307e+00 -7.48634264e-02
6.01395309e-01 -1.42440188e+00 1.19455576e+00 -4.64222357e-02
-1.06598043e+00 1.49882972e-01 -5.46655893e-01 -3.10799591e-02
-4.92122650e-01 -3.62019777e-01 9.94754016e-01 9.53663826e-01
-5.31523883e-01 3.01807463e-01 -9.78738368e-02 -1.00785708e+00
-3.18205416e-01 1.20539434e-01 -3.80404741e-01 7.30626106e-01
-1.34313202e+00 1.25619721e+00 1.05569690e-01 -1.43137453e-02
-2.37030443e-02 1.69332430e-01 -9.12335634e-01 -3.76406044e-01
1.24151846e-02 -2.88498942e-02 1.04893041e+00 -1.16148281e+00
-1.99853909e+00 1.49633348e+00 -1.72419474e-01 4.02681343e-02
-1.93941519e-01 -1.25204891e-01 -6.64463460e-01 2.46987343e-01
-5.68540394e-02 4.33044970e-01 4.37472969e-01 -1.01972389e+00
-1.33908987e-01 -2.38997266e-01 4.78996724e-01 4.52309459e-01
-3.00855219e-01 9.52672362e-01 4.79517430e-01 -2.39434630e-01
1.21988557e-01 -7.80453384e-01 2.97776550e-01 -7.89222538e-01
-2.36955911e-01 -1.18803608e+00 8.92256796e-01 -5.10604978e-01
1.23817766e+00 -2.34386206e+00 -2.36569896e-01 -8.72551203e-02
3.23908955e-01 -3.21610384e-02 4.67285156e-01 1.18708837e+00
-1.27980366e-01 1.63218111e-01 3.12216431e-01 1.29783139e-01
4.28760320e-01 5.53485096e-01 -2.58882642e-01 4.95502025e-01
-5.85589744e-02 7.13674247e-01 -1.07214200e+00 -7.67057240e-01
1.42022416e-01 2.98603326e-01 -4.79915351e-01 6.63317561e-01
1.63115025e-01 2.19841301e-01 -6.04118645e-01 6.38196766e-01
2.76627928e-01 2.36939207e-01 2.41667807e-01 -2.65328169e-01
-2.62279838e-01 9.81733501e-01 -3.46202701e-01 6.18775427e-01
-4.58667904e-01 1.03325629e+00 5.08857489e-01 -6.39510274e-01
1.49670804e+00 7.52873600e-01 2.47814372e-01 -2.15575770e-01
7.90308714e-01 4.59627574e-03 3.63645017e-01 -1.31522655e+00
5.94230890e-01 -6.31037414e-01 -6.68869317e-01 1.03095567e+00
-3.14129084e-01 -8.22296560e-01 -1.37654230e-01 4.57506895e-01
5.18436551e-01 -4.51310754e-01 8.46068680e-01 -4.65052456e-01
7.06517160e-01 -3.74434888e-01 5.75028121e-01 1.67310107e-02
-7.41905093e-01 -4.70353216e-01 1.28443599e+00 -3.72053385e-01
-5.77912092e-01 -4.61821049e-01 -1.28218219e-01 1.29376984e+00
-1.14237674e-01 -2.50251114e-01 -7.05214024e-01 -3.25172842e-01
-3.39175791e-01 9.04681861e-01 -5.38461089e-01 9.76906717e-02
-3.51739615e-01 -5.44671834e-01 6.90023601e-01 -2.37241223e-01
4.53328907e-01 -1.67694092e+00 -4.89090294e-01 4.17722799e-02
-4.75379080e-01 -9.79330122e-01 4.74167131e-02 6.93913028e-02
-5.45622170e-01 -8.28306377e-01 1.41755983e-01 -8.41886222e-01
3.89799267e-01 -6.91199750e-02 1.06381798e+00 3.71058762e-01
1.64699003e-01 5.82627296e-01 -7.31700778e-01 -3.37746918e-01
-9.91447151e-01 -3.68157238e-01 6.27297238e-02 -2.40667149e-01
9.35140789e-01 -6.09940112e-01 -8.84211510e-02 1.83605731e-01
-5.80809295e-01 -1.81357622e-01 -2.31989503e-01 7.43893743e-01
-7.13731408e-01 -4.12640631e-01 7.33492374e-01 -1.14431190e+00
1.51565135e+00 -7.52733886e-01 6.46742046e-01 -6.44160733e-02
-1.68491915e-01 -3.02289039e-01 1.05168343e-01 -3.05498749e-01
-1.27115202e+00 -7.42214262e-01 -3.52293134e-01 6.94902956e-01
-4.68813986e-01 3.39495808e-01 1.38607919e-01 5.41694798e-02
6.46309733e-01 -5.51513672e-01 3.39412451e-01 -3.92305013e-03
-7.90425465e-02 1.42140675e+00 -9.13172364e-02 -8.37602317e-01
1.00755729e-01 2.80885756e-01 -5.74872017e-01 -1.30210006e+00
-7.89549828e-01 -6.41357124e-01 -1.24341555e-01 -9.05645132e-01
9.44092035e-01 -4.38918531e-01 -1.55772495e+00 4.48987365e-01
-1.34577835e+00 -2.06396312e-01 1.64002284e-01 6.40128136e-01
-7.25633800e-01 6.59648001e-01 -1.02842343e+00 -1.56264937e+00
-4.27537322e-01 -7.19236016e-01 3.32288682e-01 3.56252104e-01
-1.30726588e+00 -1.33843207e+00 1.44838899e-01 7.09258735e-01
4.00285989e-01 6.55777633e-01 1.02774262e+00 -1.06248891e+00
7.80242264e-01 7.32502937e-02 1.14888996e-02 1.11303970e-01
5.71523726e-01 4.02408212e-01 -7.63612688e-01 2.56984264e-01
8.14112544e-01 -1.16124439e+00 2.45511215e-02 -1.29986167e-01
-1.04693308e-01 -8.47696602e-01 -3.63754928e-02 -2.44661167e-01
7.90203094e-01 5.14814317e-01 6.92991912e-01 -4.01187092e-02
-7.71977901e-02 1.59762025e+00 1.17045832e+00 5.86516380e-01
5.35152078e-01 3.16347659e-01 1.63013726e-01 2.89292753e-01
3.28105062e-01 4.33344487e-03 8.88841152e-01 1.26955354e+00
2.05185208e-02 -1.16237998e-01 -9.43036973e-01 4.67693985e-01
-1.75751066e+00 -1.16401327e+00 -6.45407319e-01 1.21199107e+00
1.03875709e+00 -1.26246378e-01 1.35364696e-01 1.50283888e-01
5.66864967e-01 7.68116295e-01 1.44652516e-01 -1.58753085e+00
-1.48809984e-01 -2.13797361e-01 -3.19651842e-01 1.19025993e+00
-6.53079331e-01 1.28204381e+00 6.90076542e+00 2.27159753e-01
-9.49494183e-01 6.82586282e-02 7.24539816e-01 3.45847428e-01
-4.62296635e-01 -2.18395535e-02 -4.63948727e-01 1.12412237e-01
8.16683948e-01 -2.35304698e-01 3.79566878e-01 8.08398724e-01
4.66835886e-01 -6.60276115e-01 -8.94863129e-01 8.49949121e-01
2.35596880e-01 -4.92998928e-01 -7.27320373e-01 -3.51750851e-01
1.74810961e-01 -5.25285900e-01 -1.94307417e-01 4.48127449e-01
2.99344629e-01 -8.68165791e-01 -7.45627508e-02 3.31180632e-01
1.01959743e-01 -6.41183555e-01 9.10787582e-01 1.41500637e-01
-4.66024339e-01 2.87218094e-01 -4.09817994e-01 -9.08603251e-01
4.33156252e-01 3.77925992e-01 -9.42036569e-01 -2.57538050e-01
5.60015261e-01 7.70498335e-01 1.63802847e-01 2.71246452e-02
-4.07401979e-01 6.09471440e-01 -2.71007389e-01 -1.01045513e+00
4.64548260e-01 -6.93802118e-01 8.96266162e-01 1.57433832e+00
-3.92896891e-01 6.50899410e-01 -2.04040885e-01 4.29723591e-01
5.02373457e-01 6.21844709e-01 -1.04680526e+00 -4.43408579e-01
3.69669676e-01 1.43205369e+00 -7.42812395e-01 -3.37205112e-01
-1.99344873e-01 4.44213063e-01 1.19479045e-01 -6.73869997e-02
-6.65258408e-01 -2.17336133e-01 8.69422317e-01 -3.48853618e-01
-7.86120117e-01 7.36226812e-02 -2.44363397e-01 -8.55419874e-01
-5.10667205e-01 -1.13279963e+00 2.39727348e-01 -6.64625466e-01
-1.52100646e+00 6.68039441e-01 2.68390886e-02 -4.62960005e-01
-4.80806559e-01 -4.37224150e-01 -9.31420267e-01 5.61756492e-01
-7.67297447e-01 -8.78738523e-01 -1.94160163e-01 3.59763741e-01
3.46647799e-01 -1.17667563e-01 1.11043000e+00 2.84726862e-02
-3.15770239e-01 1.77939743e-01 -7.39914477e-01 2.18030781e-01
9.62653041e-01 -8.93981338e-01 -3.90426844e-01 -2.08103865e-01
-7.54028618e-01 6.36725724e-01 1.25679016e+00 -4.77086544e-01
-9.15901303e-01 1.27572089e-01 1.60650766e+00 -1.12264462e-01
1.15133166e+00 -6.65424615e-02 -6.47573292e-01 5.82739413e-01
1.02512550e+00 -9.12272453e-01 1.32139909e+00 6.07945025e-01
-6.48055822e-02 5.03491819e-01 -1.54919684e+00 8.16659510e-01
5.47125995e-01 -7.58110940e-01 -1.29526639e+00 6.78647637e-01
6.44628525e-01 -1.95623964e-01 -1.18562710e+00 -1.20985182e-02
5.58687150e-01 -1.45702338e+00 5.21006644e-01 -7.58956194e-01
5.50381303e-01 5.09086013e-01 -1.66722104e-01 -1.16222727e+00
7.67465830e-02 -1.17541826e+00 5.86702943e-01 1.57847834e+00
2.35691071e-01 -9.81632471e-01 5.51971436e-01 1.12667918e+00
-2.24613119e-02 -4.24981654e-01 -6.05704963e-01 -1.97795741e-02
1.56572506e-01 -3.27448756e-01 2.86856890e-01 1.59171617e+00
1.30621052e+00 8.47121358e-01 -5.27661443e-01 -7.38219082e-01
-1.58523336e-01 -9.24238116e-02 9.04131651e-01 -9.95618343e-01
2.89089475e-02 -4.12307471e-01 -1.80076733e-01 -8.23681355e-01
5.17105460e-01 -4.03697491e-01 -1.94869339e-02 -1.14520192e+00
1.11267669e-02 -2.39900813e-01 6.22132242e-01 2.11648211e-01
-2.23358385e-02 -1.60323411e-01 2.18717709e-01 -1.05378352e-01
-2.23288089e-01 5.39741755e-01 1.28341699e+00 2.40524754e-01
-6.19616807e-01 -1.35666966e-01 -8.62130761e-01 1.26705360e+00
1.10201859e+00 -4.76967186e-01 -1.51988026e-02 5.86146414e-01
6.39163554e-01 6.84700131e-01 -3.36939722e-01 -7.56781623e-02
-1.39948890e-01 -6.03679955e-01 -7.28956640e-01 -5.04543185e-01
6.91769660e-01 -7.35834420e-01 -1.10116400e-01 6.00238144e-01
-6.28139734e-01 7.41281658e-02 -1.16053157e-01 -9.82557535e-02
-5.59203625e-01 -5.74696004e-01 9.26692665e-01 -8.20530057e-02
9.28913150e-03 -5.17841220e-01 -1.01361752e+00 2.26862833e-01
1.17680037e+00 -2.95881450e-01 -4.74636197e-01 -1.28596079e+00
-8.43256950e-01 9.53217447e-02 1.22961290e-01 4.69996303e-01
4.52875078e-01 -1.16033280e+00 -7.04341114e-01 -5.00439227e-01
-1.43307805e-01 -7.46279716e-01 7.93402642e-02 1.27895594e+00
-6.10871315e-01 3.31594050e-01 -4.23350811e-01 -1.48454115e-01
-1.77714038e+00 2.06802294e-01 1.59174040e-01 -7.04144612e-02
-1.12963088e-01 6.49305701e-01 8.59062299e-02 -6.48347139e-01
1.45851046e-01 -8.42959881e-02 -8.21411729e-01 8.63388360e-01
3.04056823e-01 6.40292645e-01 -7.43949771e-01 -1.13585711e+00
-3.59022528e-01 2.18446299e-01 5.29245250e-02 -2.72871792e-01
1.15473557e+00 -4.04954106e-01 -1.16243899e+00 9.92061317e-01
1.39093232e+00 4.81784284e-01 -2.45805547e-01 6.08955398e-02
9.83625278e-02 -4.06266689e-01 -4.36833799e-01 -6.49627566e-01
-4.17628855e-01 7.14725792e-01 -1.34818971e-01 1.28389060e+00
5.74083447e-01 1.71787068e-01 8.16266716e-01 5.21599114e-01
2.93866489e-02 -1.43559682e+00 3.96442205e-01 1.06636262e+00
1.26228142e+00 -1.35252929e+00 -5.19887134e-02 -8.72546971e-01
-1.22823322e+00 1.36461020e+00 6.98603809e-01 -8.90260339e-02
8.51040959e-01 3.43044400e-01 6.18722856e-01 -8.47599030e-01
-1.10736275e+00 3.02091744e-02 -3.07998389e-01 4.46352363e-01
1.26554680e+00 3.64935249e-01 -1.35567796e+00 4.90306735e-01
-7.80343056e-01 -6.74773753e-01 8.33830237e-01 5.29558480e-01
-5.81926942e-01 -9.40919280e-01 -4.58449930e-01 1.89697981e-01
-8.81991327e-01 -1.11133559e-02 -1.27929580e+00 7.46981919e-01
-3.14059585e-01 1.53441560e+00 1.52859032e-01 -2.66442984e-01
2.90182009e-02 2.66863406e-01 -1.19557790e-01 -6.22540057e-01
-1.64129031e+00 5.22164349e-03 1.18609905e+00 -2.61038303e-01
-1.21293235e+00 -7.76122510e-01 -1.31734228e+00 -9.93591845e-01
-2.28392601e-01 5.68746924e-01 4.07106042e-01 1.02776456e+00
-3.08766454e-01 -7.76383504e-02 8.72839928e-01 -2.59139478e-01
-8.25188681e-02 -1.43138111e+00 -4.42551583e-01 7.40387917e-01
7.09540546e-02 -5.72868764e-01 -8.25164020e-01 -4.21657711e-01] | [13.046546936035156, 7.561603546142578] |
ef6dcc78-e560-457a-ac61-c4e05dfefb85 | gulf-arabic-linguistic-resource-building-for | null | null | https://aclanthology.org/L16-1430 | https://aclanthology.org/L16-1430.pdf | Gulf Arabic Linguistic Resource Building for Sentiment Analysis | This paper deals with building linguistic resources for Gulf Arabic, one of the Arabic variations, for sentiment analysis task using machine learning. To our knowledge, no previous works were done for Gulf Arabic sentiment analysis despite the fact that it is present in different online platforms. Hence, the first challenge is the absence of annotated data and sentiment lexicons. To fill this gap, we created these two main linguistic resources. Then we conducted different experiments: use Naive Bayes classifier without any lexicon; add a sentiment lexicon designed basically for MSA; use only the compiled Gulf Arabic sentiment lexicon and finally use both MSA and Gulf Arabic sentiment lexicons. The Gulf Arabic lexicon gives a good improvement of the classifier accuracy (90.54 {\%}) over a baseline that does not use the lexicon (82.81{\%}), while the MSA lexicon causes the accuracy to drop to (76.83{\%}). Moreover, mixing MSA and Gulf Arabic lexicons causes the accuracy to drop to (84.94{\%}) compared to using only Gulf Arabic lexicon. This indicates that it is useless to use MSA resources to deal with Gulf Arabic due to the considerable differences and conflicting structures between these two languages. | ['Richard Johansson', 'Wafia Adouane'] | 2016-05-01 | gulf-arabic-linguistic-resource-building-for-1 | https://aclanthology.org/L16-1430 | https://aclanthology.org/L16-1430.pdf | lrec-2016-5 | ['arabic-sentiment-analysis'] | ['natural-language-processing'] | [-3.65870774e-01 7.24428508e-04 2.55897760e-01 -3.01359296e-01
-5.78654766e-01 -9.05917645e-01 6.10273063e-01 6.17233336e-01
-5.63829839e-01 7.06895769e-01 -2.16169078e-02 -4.30000961e-01
4.22989652e-02 -9.29316878e-01 7.69391283e-03 -7.40819812e-01
2.88371980e-01 5.58292925e-01 3.04032117e-01 -1.23258853e+00
5.90591192e-01 4.08363849e-01 -1.63948882e+00 4.36485857e-01
9.60638046e-01 8.55722368e-01 -8.81868452e-02 3.78862083e-01
-4.04267341e-01 8.38336170e-01 -9.75774050e-01 -8.22570682e-01
1.90362990e-01 -5.10236621e-01 -7.88850665e-01 -1.89081095e-02
6.94296509e-02 7.94282854e-02 6.61735892e-01 1.10776389e+00
3.62260401e-01 1.56197427e-02 8.13667893e-01 -1.03908300e+00
-4.38867956e-01 5.94306350e-01 -6.63457513e-01 -1.08298540e-01
6.22754514e-01 -5.28580844e-01 9.06644881e-01 -9.16708052e-01
5.28062224e-01 8.97530794e-01 6.94007635e-01 2.42808312e-01
-3.32592905e-01 -5.49648345e-01 1.40925869e-02 -2.59433329e-01
-1.07365477e+00 -2.46812865e-01 5.44310510e-01 -2.58108497e-01
9.60227013e-01 1.10459641e-01 6.91732049e-01 6.33518875e-01
2.66092151e-01 2.97943771e-01 1.64985192e+00 -9.74507153e-01
1.57715067e-01 5.09195983e-01 3.87102425e-01 6.72529697e-01
3.87601614e-01 -5.99734843e-01 -3.19049180e-01 2.66629811e-02
3.49174216e-02 -3.61964226e-01 2.06344739e-01 2.84736812e-01
-7.57048786e-01 1.03555262e+00 5.77338934e-02 7.68951654e-01
-1.79399610e-01 -7.27245331e-01 5.70162892e-01 6.51466608e-01
3.05058479e-01 6.01139009e-01 -7.70073473e-01 -1.72793388e-01
-9.09737468e-01 3.62808198e-01 1.24949718e+00 8.44145000e-01
7.11641252e-01 1.07505843e-01 6.94822371e-01 1.09297895e+00
7.55605400e-01 7.41664469e-01 6.90630615e-01 -3.79469842e-01
3.95983577e-01 1.00541747e+00 1.75104246e-01 -1.20069754e+00
-7.52837121e-01 -1.86668560e-01 -4.73589659e-01 2.50755161e-01
9.09457564e-01 -5.58340251e-01 -8.60166848e-01 1.06004357e+00
7.79346377e-02 -8.76188338e-01 6.96345747e-01 7.40022898e-01
1.04141736e+00 7.22249508e-01 1.69953723e-02 -2.47267559e-01
1.67154717e+00 -9.09803092e-01 -8.45786870e-01 -4.61124480e-01
8.62128437e-01 -1.49605727e+00 1.25365448e+00 7.08301008e-01
-9.05799091e-01 -3.22535001e-02 -1.47255206e+00 2.26135164e-01
-1.12512958e+00 1.96508154e-01 5.93971193e-01 1.43207562e+00
-1.11609817e+00 5.44283306e-03 -6.39176071e-01 -6.40969217e-01
-1.88203156e-01 5.35192072e-01 -3.50969046e-01 -6.60373867e-02
-1.24135160e+00 1.31387389e+00 3.47797126e-01 -2.12892648e-02
-4.12707543e-03 -3.64113189e-02 -9.42928016e-01 -3.71357799e-01
1.71397716e-01 1.39591306e-01 9.92834985e-01 -1.37515569e+00
-1.60014427e+00 8.95341277e-01 1.68418903e-02 -2.76780389e-02
5.04721887e-02 -7.52338991e-02 -6.66360795e-01 -3.22426483e-02
2.80246973e-01 2.96807796e-01 2.71487236e-01 -1.27668285e+00
-9.61563528e-01 -5.32815576e-01 3.98458689e-01 2.18972966e-01
-4.23055202e-01 4.66864020e-01 -3.58141184e-01 -7.31387615e-01
3.68486106e-01 -1.19915974e+00 -1.32258132e-01 -9.33613420e-01
1.70658305e-01 -9.37040970e-02 5.46663880e-01 -8.67343843e-01
1.31415009e+00 -1.74515784e+00 -3.54400069e-01 5.62798977e-01
-3.31365317e-01 3.41604799e-01 7.52320662e-02 7.31052518e-01
7.25608543e-02 1.07662179e-01 -3.45303655e-01 -3.80262686e-03
-7.67600685e-02 2.76937455e-01 -8.79948512e-02 2.38771558e-01
2.79403061e-01 1.98698983e-01 -8.25630069e-01 -4.60814297e-01
5.74465059e-02 5.08998573e-01 -4.33771551e-01 -4.93056387e-01
7.85498843e-02 2.77320817e-02 -3.75474006e-01 1.14889777e+00
8.07335079e-01 3.40575635e-01 5.26106834e-01 -6.16301894e-02
-4.29956734e-01 2.54972875e-01 -1.20671713e+00 1.28422379e+00
-5.29326916e-01 3.35259765e-01 -1.35009453e-01 -7.92300522e-01
1.37242031e+00 2.73009449e-01 3.07654738e-01 -6.32800639e-01
6.92463398e-01 8.15721571e-01 2.19159603e-01 -1.76666602e-01
7.80673504e-01 -4.29842621e-01 -5.10610998e-01 6.63541317e-01
1.71230987e-01 -7.11332798e-01 6.82548821e-01 6.70597628e-02
4.98753220e-01 1.85193375e-01 3.62456203e-01 -6.14605963e-01
9.88969803e-01 5.25174618e-01 1.23004988e-01 2.15529650e-01
-2.46385619e-01 5.59179306e-01 5.99639118e-01 -5.95653057e-01
-6.09595001e-01 -4.41600055e-01 -1.49515301e-01 1.36234164e+00
-7.74294809e-02 -7.14975893e-01 -9.64761615e-01 -8.63681734e-01
-5.72462678e-01 4.54727411e-01 -5.20801723e-01 1.73835561e-01
-4.59828317e-01 -1.48985660e+00 6.00748420e-01 1.30953649e-02
5.18611431e-01 -1.06642246e+00 -5.82354486e-01 7.86160380e-02
-3.16469848e-01 -7.69578099e-01 2.04926267e-01 2.76208103e-01
-5.62656462e-01 -1.08128929e+00 -4.17813480e-01 -9.31347311e-01
3.29413474e-01 -7.56203979e-02 1.14611602e+00 3.10751081e-01
4.80805427e-01 -5.39121367e-02 -1.13011003e+00 -1.03873384e+00
-6.60049856e-01 3.62462580e-01 -1.04327656e-01 -3.30717713e-01
8.06871772e-01 7.87374005e-02 -7.33704418e-02 1.63209349e-01
-1.15262926e+00 -5.03567278e-01 5.54424942e-01 5.14469206e-01
2.08312735e-01 6.78430051e-02 8.98716033e-01 -1.01518178e+00
8.06734204e-01 -4.28812444e-01 -3.51741135e-01 1.73756838e-01
-6.88473642e-01 -1.44854560e-01 4.89801556e-01 1.52797118e-01
-9.21400309e-01 3.73859704e-02 -6.97589278e-01 8.32176507e-01
-8.24112166e-03 9.47033942e-01 -1.34941161e-01 3.48233096e-02
7.00141728e-01 -2.21854806e-01 1.74871534e-01 -1.39548257e-01
-5.77721670e-02 1.04617345e+00 -1.22269973e-01 -3.35698009e-01
1.41995192e-01 3.58251899e-01 -1.29546881e-01 -7.58390546e-01
-9.36993599e-01 -4.26769853e-01 -8.70101392e-01 -2.43801519e-01
7.31624126e-01 -7.22551644e-01 -4.69272226e-01 6.46213830e-01
-6.99460685e-01 -2.34324746e-02 8.24419409e-02 4.08772796e-01
-1.71549529e-01 3.87797207e-01 -5.20246029e-01 -7.58663237e-01
-3.75081658e-01 -1.38585937e+00 7.27033973e-01 2.71266311e-01
-4.18888062e-01 -1.07558763e+00 8.12158734e-02 4.37894911e-01
3.77381355e-01 2.70539522e-01 7.43330538e-01 -7.57305920e-01
6.43661559e-01 -4.80815709e-01 -3.46803740e-02 4.02011842e-01
4.87177789e-01 4.15215045e-01 -8.82805884e-01 -1.50662258e-01
1.96892262e-01 -3.34483057e-01 4.50464696e-01 -6.74693957e-02
-6.52915463e-02 2.27334909e-02 2.66876847e-01 -1.69456258e-01
1.44846070e+00 6.21491075e-01 6.68855906e-01 1.05959320e+00
1.86367318e-01 9.70690906e-01 1.14883626e+00 4.25401062e-01
7.05552697e-01 3.55441719e-01 3.73689532e-01 2.04952344e-01
2.81893790e-01 3.69234890e-01 8.63979578e-01 1.39306664e+00
-2.12726593e-01 -3.25286061e-01 -1.48036838e+00 6.22051120e-01
-1.45503008e+00 -3.42537642e-01 -5.30305207e-01 1.96944702e+00
9.05463517e-01 1.20279111e-01 1.91775575e-01 5.86300135e-01
1.92761511e-01 -4.77937311e-02 3.78173709e-01 -1.12238216e+00
-5.32703817e-01 6.56651199e-01 3.35959554e-01 6.45622551e-01
-1.37720907e+00 1.25411248e+00 5.47760868e+00 6.10084355e-01
-1.35401618e+00 4.11499850e-02 2.84684539e-01 3.80843729e-01
-1.43350465e-02 1.52318031e-01 -7.91852534e-01 2.97595829e-01
1.24041033e+00 2.45193407e-01 -1.49209708e-01 5.01392901e-01
-1.63004041e-01 -4.88486767e-01 -9.68350917e-02 5.74397981e-01
5.67057252e-01 -8.84327769e-01 2.94666499e-01 -2.88905621e-01
7.94167757e-01 -6.77463040e-02 -1.81352675e-01 3.08402538e-01
4.77389663e-01 -1.01575661e+00 7.56618679e-01 2.06121892e-01
2.94756830e-01 -1.32772064e+00 1.61107528e+00 2.45766595e-01
-9.46455121e-01 1.51576415e-01 -3.38934034e-01 -4.47451353e-01
-5.41878454e-02 3.49482208e-01 -7.27234662e-01 6.69027388e-01
1.10877466e+00 4.84423131e-01 -9.22047734e-01 4.91921008e-01
-2.20579714e-01 5.53397536e-01 -3.81311655e-01 -3.57854545e-01
7.68900454e-01 -6.60580516e-01 2.04584733e-01 1.31492686e+00
3.54800701e-01 -1.49380162e-01 1.94023192e-01 -3.22947711e-01
4.21873719e-01 1.09961069e+00 -4.55357522e-01 3.29106301e-02
1.52206033e-01 1.12468684e+00 -1.34514987e+00 -2.31012017e-01
-5.95866680e-01 7.50661433e-01 -1.21787660e-01 -1.44355625e-01
-6.11234605e-01 -7.87726223e-01 1.00830264e-01 -3.16959955e-02
-5.81752602e-03 -1.36895135e-01 -4.54110205e-01 -9.97371137e-01
-2.12555245e-01 -1.27561188e+00 6.15784764e-01 -8.21964681e-01
-1.19811428e+00 1.19479418e+00 -1.53741777e-01 -1.20092428e+00
-2.92810440e-01 -9.61733460e-01 8.45272616e-02 8.73976529e-01
-1.40222645e+00 -1.49292111e+00 -1.65091753e-01 4.05314744e-01
4.00858372e-01 -5.38795829e-01 1.20617473e+00 3.87401670e-01
-2.61717439e-01 5.34508288e-01 -1.87151790e-01 2.16620132e-01
1.14211380e+00 -1.28816772e+00 -3.58139962e-01 7.26588905e-01
-1.64085254e-01 6.57783270e-01 6.58620059e-01 -6.07047677e-01
-9.88050938e-01 -6.41815126e-01 1.36397135e+00 -8.23679626e-01
8.53521824e-01 -4.33085971e-02 -7.26698697e-01 5.32118499e-01
7.20977962e-01 -6.78082526e-01 1.24414468e+00 -7.86247663e-03
-2.43318975e-01 2.19709128e-02 -1.38603497e+00 7.79327571e-01
-9.08433348e-02 -1.24400213e-01 -6.72850311e-01 2.37990364e-01
2.36471146e-01 -2.57762194e-01 -9.93398607e-01 2.95416474e-01
8.00097764e-01 -1.00318825e+00 5.04489779e-01 -3.43959272e-01
4.80913818e-01 -6.09823823e-01 -3.56978476e-01 -1.27397764e+00
4.05861646e-01 -1.54315248e-01 6.84071720e-01 1.33590508e+00
9.93030131e-01 -8.94498110e-01 5.89887857e-01 1.98024392e-01
-1.44148469e-01 -4.63007212e-01 -4.94855106e-01 -5.94698638e-02
7.42078125e-01 -4.89911944e-01 6.06082976e-01 1.36335504e+00
2.83108890e-01 4.68960702e-01 -2.25438979e-02 8.75291973e-02
-2.36183271e-01 2.61007268e-02 5.16693473e-01 -1.25207913e+00
3.75826597e-01 -4.77135539e-01 -3.22024703e-01 -2.43158922e-01
-2.66399328e-02 -6.89160287e-01 -5.09512834e-02 -1.56139934e+00
-3.37825984e-01 -8.80326569e-01 -4.84122662e-03 7.18063295e-01
1.51605085e-02 9.20063019e-01 2.16792136e-01 1.44824743e-01
-2.24279389e-01 -9.31693241e-02 8.77390027e-01 1.10176645e-01
-4.15450633e-01 -1.65013865e-01 -1.17925620e+00 1.11160731e+00
1.18812883e+00 -4.67294484e-01 -1.32628620e-01 -2.24504247e-01
1.03618145e+00 -4.04901356e-01 -5.31728804e-01 -7.05191135e-01
3.76632549e-02 -1.79943442e-01 1.22373559e-01 -5.99620163e-01
9.89363715e-02 -8.06032062e-01 -2.73110598e-01 3.62958223e-01
3.81214797e-01 6.82518184e-01 5.34881115e-01 -2.97865361e-01
-6.20151579e-01 -7.32322931e-01 7.50565112e-01 -2.22094715e-01
-6.48667336e-01 -5.17427266e-01 -1.05164254e+00 -1.99388593e-01
9.47580636e-01 -3.40025872e-01 -2.44644567e-01 -3.31931919e-01
-7.13761330e-01 -6.74621612e-02 5.84568322e-01 3.73258233e-01
1.72110468e-01 -8.29612613e-01 -6.65621996e-01 1.93546802e-01
1.99685425e-01 -1.89106673e-01 -1.85469598e-01 1.00631618e+00
-1.33928061e+00 2.47419640e-01 -6.53340220e-01 -3.49963829e-02
-1.17816639e+00 -1.87866464e-02 1.14448719e-01 -4.33894485e-01
2.32736886e-01 5.18629491e-01 -6.50813222e-01 -9.93232429e-01
-2.99458295e-01 -1.61357492e-01 -1.19890916e+00 9.38276410e-01
3.87414187e-01 3.33907694e-01 7.17291832e-01 -1.34376144e+00
-5.45641065e-01 7.52945125e-01 3.36645059e-02 -4.56388444e-01
1.34312463e+00 -1.93231106e-01 -7.84801900e-01 5.95263064e-01
5.41898608e-01 9.68773186e-01 -9.05942470e-02 4.93221402e-01
3.21498930e-01 -2.14512825e-01 -1.88552484e-01 -1.12383687e+00
-8.86740208e-01 6.76840723e-01 6.24004960e-01 4.59501415e-01
1.26458633e+00 -3.23879838e-01 3.41308594e-01 5.00555396e-01
4.32588905e-01 -1.31194687e+00 -2.72214234e-01 1.24418008e+00
8.02986860e-01 -1.22601140e+00 1.66533291e-01 -6.24761105e-01
-1.06308234e+00 1.43048263e+00 4.86461729e-01 -2.95872718e-01
1.15597725e+00 3.81523401e-01 9.25253034e-01 -2.90764809e-01
-1.61116585e-01 -4.87710148e-01 7.57911578e-02 5.36485195e-01
9.45194840e-01 -5.97043987e-03 -1.11884499e+00 1.11066878e+00
-8.81133914e-01 -1.93278328e-01 1.06818473e+00 1.30113423e+00
-4.45726842e-01 -1.62283337e+00 -4.83658552e-01 3.98106933e-01
-1.04679954e+00 -2.61080563e-01 -7.22622752e-01 1.23511267e+00
3.71766776e-01 1.45506632e+00 7.91466795e-03 -3.13000709e-01
3.53228450e-01 2.38450617e-01 2.04092473e-01 -7.34152555e-01
-1.22859311e+00 2.28148505e-01 2.64023274e-01 9.07575190e-02
-1.00753677e+00 -7.24236190e-01 -1.47819221e+00 -3.97612184e-01
-5.22872865e-01 6.27874374e-01 8.96198034e-01 1.09607697e+00
-2.21375883e-01 1.35912180e-01 3.17592591e-01 -4.85754132e-01
4.45739329e-02 -1.25522840e+00 -6.13815546e-01 1.66121975e-01
-7.30766356e-02 -6.04265273e-01 -3.87863278e-01 1.85648322e-01] | [11.08513355255127, 6.93202543258667] |
81e0ba2d-9b07-43c7-ac6f-fd965a05b3ec | deepseed-3d-squeeze-and-excitation-encoder | 1904.03501 | null | https://arxiv.org/abs/1904.03501v2 | https://arxiv.org/pdf/1904.03501v2.pdf | DeepSEED: 3D Squeeze-and-Excitation Encoder-Decoder Convolutional Neural Networks for Pulmonary Nodule Detection | Pulmonary nodule detection plays an important role in lung cancer screening with low-dose computed tomography (CT) scans. It remains challenging to build nodule detection deep learning models with good generalization performance due to unbalanced positive and negative samples. In order to overcome this problem and further improve state-of-the-art nodule detection methods, we develop a novel deep 3D convolutional neural network with an Encoder-Decoder structure in conjunction with a region proposal network. Particularly, we utilize a dynamically scaled cross entropy loss to reduce the false positive rate and combat the sample imbalance problem associated with nodule detection. We adopt the squeeze-and-excitation structure to learn effective image features and utilize inter-dependency information of different feature maps. We have validated our method based on publicly available CT scans with manually labelled ground-truth obtained from LIDC/IDRI dataset and its subset LUNA16 with thinner slices. Ablation studies and experimental results have demonstrated that our method could outperform state-of-the-art nodule detection methods by a large margin. | ['Yong Fan', 'Yuemeng Li'] | 2019-04-06 | null | null | null | null | ['lung-nodule-detection', 'automated-pulmonary-nodule-detection-and'] | ['medical', 'medical'] | [ 3.79187077e-01 4.53609794e-01 -5.76162517e-01 -2.95048982e-01
-1.15124321e+00 -3.15336138e-01 2.88482785e-01 -1.19118147e-01
-3.89789641e-01 4.19267297e-01 2.34840244e-01 -6.64529920e-01
-1.22957706e-01 -7.56543100e-01 -6.13488078e-01 -7.63459086e-01
-1.90541819e-01 5.96550822e-01 6.87135398e-01 1.66837484e-01
-3.88602614e-01 3.38963360e-01 -9.89492714e-01 6.11341119e-01
7.48807728e-01 1.03479671e+00 4.86326814e-01 4.88446862e-01
2.64942199e-01 9.32108164e-01 8.40397328e-02 -2.65791565e-01
5.40154755e-01 -6.15068495e-01 -9.34228659e-01 3.18154283e-02
1.23391233e-01 -5.29620051e-01 -5.19680917e-01 9.78599429e-01
7.10752845e-01 -5.26192844e-01 7.38200963e-01 -5.67310214e-01
-2.94245690e-01 6.54219031e-01 -6.57648325e-01 4.96759921e-01
-4.36101019e-01 4.08525765e-01 7.88354933e-01 -6.81088328e-01
3.92642200e-01 5.46347201e-01 1.05769396e+00 6.37784660e-01
-8.25522840e-01 -5.49836636e-01 -3.91177505e-01 -1.65128738e-01
-1.03785980e+00 3.06452185e-01 2.07284808e-01 -3.67121696e-01
4.88070816e-01 5.09898722e-01 1.09516346e+00 1.07712829e+00
5.42978287e-01 8.09591889e-01 7.43515372e-01 -2.37793803e-01
-1.99700445e-01 2.96076685e-01 -3.49574059e-01 1.28481686e+00
6.98921561e-01 4.11916196e-01 3.16037059e-01 -1.68102875e-01
1.21464109e+00 1.42188638e-01 -3.79307508e-01 -7.38417625e-01
-1.34139633e+00 1.09084892e+00 1.27727413e+00 4.44301099e-01
-3.57980430e-01 2.21304283e-01 5.38046062e-01 -3.10225487e-01
2.72548556e-01 5.20926774e-01 1.00510819e-02 4.49379236e-01
-8.58242929e-01 -7.99062997e-02 7.98639238e-01 3.25219274e-01
6.73771948e-02 -3.42163950e-01 -8.69231343e-01 6.78566098e-01
3.00063908e-01 3.03485602e-01 8.94370854e-01 -3.34648609e-01
1.54097006e-01 7.75064468e-01 -3.48795563e-01 -3.34465772e-01
-7.51769602e-01 -1.08542156e+00 -1.11244845e+00 6.05152138e-02
2.51558810e-01 -1.04524188e-01 -1.52123833e+00 1.28393757e+00
2.11840108e-01 9.87761915e-02 -2.75693446e-01 1.13985991e+00
1.09522045e+00 -3.94170955e-02 8.59404132e-02 1.68395415e-02
1.54456818e+00 -1.08039987e+00 -2.38355279e-01 -1.88836887e-01
9.77756977e-01 -4.68367666e-01 8.52658033e-01 -2.23967820e-01
-9.77151871e-01 -3.15867305e-01 -9.79966938e-01 1.53122451e-02
1.22912452e-01 8.38279188e-01 7.84141600e-01 9.05741334e-01
-7.52156615e-01 3.03745836e-01 -1.41567445e+00 -3.92907739e-01
8.84691656e-01 6.54858112e-01 -1.48381278e-01 9.62239504e-03
-1.03225911e+00 9.21663821e-01 4.62014467e-01 7.84900505e-03
-1.15820861e+00 -1.20893431e+00 -5.81245959e-01 2.17742413e-01
5.61952829e-01 -9.80658829e-01 1.62155354e+00 -5.23889065e-01
-1.21406972e+00 8.06877196e-01 3.88801128e-01 -6.73816204e-01
8.09815109e-01 2.46165618e-01 8.84648506e-03 1.89706415e-01
2.63320714e-01 6.37857139e-01 3.92124951e-01 -7.58087993e-01
-4.95533109e-01 -2.00635299e-01 -4.13624436e-01 2.38199636e-01
-2.63386637e-01 -3.66788536e-01 -5.05137801e-01 -3.87841791e-01
3.18863809e-01 -1.08079100e+00 -7.78085113e-01 4.87989843e-01
-5.96472919e-01 1.52500093e-01 7.17865050e-01 -4.90573376e-01
1.02794504e+00 -1.74815869e+00 -2.95008034e-01 3.97993207e-01
5.60690999e-01 4.32315171e-01 2.24104792e-01 -4.63644177e-01
-1.39060378e-01 3.80304843e-01 -2.25090712e-01 2.81413347e-01
-3.26445937e-01 3.16741727e-02 5.42523921e-01 3.93896013e-01
5.34484446e-01 1.24060261e+00 -8.56068254e-01 -8.34843755e-01
3.82041156e-01 4.18017060e-01 -7.52665281e-01 6.81712925e-02
3.92245650e-02 3.48497748e-01 -7.22112894e-01 6.73834682e-01
5.14641166e-01 -9.46557403e-01 1.69013128e-01 -4.52061117e-01
6.20486364e-02 3.64694953e-01 -6.92858040e-01 1.49911332e+00
-3.53528619e-01 3.13556343e-01 -1.11479029e-01 -5.34759521e-01
4.52074856e-01 5.13797641e-01 7.54617691e-01 -6.23391330e-01
2.16814220e-01 5.20510912e-01 6.57618523e-01 -7.23958313e-01
-4.39862311e-02 -3.94507438e-01 2.24543393e-01 2.21296042e-01
-9.38209817e-02 -2.70125180e-01 2.69825407e-03 -8.39387402e-02
1.61168754e+00 -3.72790307e-01 7.12164164e-01 -4.08246815e-01
4.10187662e-01 1.30974188e-01 4.07786965e-01 8.97931635e-01
-4.02353883e-01 9.40853357e-01 7.57516265e-01 -3.53362381e-01
-1.03020453e+00 -1.08941638e+00 -6.18765533e-01 3.88962239e-01
-1.75101236e-01 1.36718169e-01 -3.83654416e-01 -1.13959241e+00
-5.42853810e-02 1.69809312e-01 -1.03332090e+00 -2.78485060e-01
-6.83796048e-01 -1.26215374e+00 6.13794625e-01 8.45985651e-01
7.46785402e-01 -7.75868058e-01 -8.55455518e-01 2.59325683e-01
-1.03654116e-01 -8.87177110e-01 -2.90780634e-01 6.91052258e-01
-9.23572540e-01 -1.40095782e+00 -9.65582132e-01 -8.65841925e-01
7.69701540e-01 1.53607219e-01 1.21400034e+00 -8.61156806e-02
-9.75754917e-01 -1.25446469e-01 -1.19824946e-01 -4.49441046e-01
-6.24944985e-01 4.96193498e-01 -5.90782285e-01 -5.77950776e-01
2.01054573e-01 -4.44618873e-02 -9.74500537e-01 3.63569170e-01
-8.00546944e-01 2.09870979e-01 1.43715417e+00 1.22664070e+00
7.76556551e-01 -3.77500504e-02 3.61781210e-01 -1.03993034e+00
2.95515507e-01 -5.22624910e-01 -4.13221031e-01 2.66448259e-01
-2.39248112e-01 7.67881870e-02 1.71412319e-01 -2.12446555e-01
-1.01961243e+00 4.25749332e-01 -1.45192906e-01 -3.02873552e-01
3.73692334e-01 4.97433394e-01 4.43847626e-01 -4.09628600e-01
1.12053263e+00 2.31134370e-02 2.57689953e-01 6.36537969e-02
-1.49110854e-01 4.38963950e-01 2.72197604e-01 9.21504125e-02
7.04999626e-01 5.41118264e-01 3.41027975e-01 -3.32907856e-01
-1.15163028e+00 -6.68276668e-01 -5.44774592e-01 -1.29812777e-01
9.76669669e-01 -1.10892379e+00 -4.89624470e-01 -5.60261793e-02
-6.06310427e-01 -1.66870639e-01 -6.03571475e-01 9.29277003e-01
-2.75466710e-01 -4.05722409e-02 -9.04275835e-01 -2.97008097e-01
-6.26437426e-01 -1.40565181e+00 1.17141056e+00 3.52472037e-01
8.28803629e-02 -7.11687922e-01 4.51376699e-02 2.51015648e-02
7.85249889e-01 3.70340496e-01 8.29743862e-01 -7.88123667e-01
-7.66053259e-01 -5.34864724e-01 -6.73986316e-01 2.28184864e-01
3.14779252e-01 -2.55273998e-01 -6.93221986e-01 -3.29670906e-01
4.05696407e-03 -4.52550828e-01 1.29804122e+00 7.36354947e-01
1.56782556e+00 1.54307336e-01 -8.47420096e-01 8.38509738e-01
1.45769644e+00 -2.59661913e-01 4.37135428e-01 1.35419264e-01
6.51265204e-01 -6.19340651e-02 2.49895677e-01 3.44528943e-01
-1.96620405e-01 1.92775503e-01 7.87548244e-01 -5.01371384e-01
-5.30644178e-01 -1.64380684e-01 -3.56913567e-01 2.53537983e-01
-1.60224438e-01 -1.58248097e-01 -1.02296245e+00 4.61514235e-01
-1.36507201e+00 -6.82876885e-01 -1.54612884e-01 1.93724263e+00
8.40924084e-01 2.58170366e-01 -2.51872092e-01 -1.71490386e-01
5.08102179e-01 -9.17236600e-03 -6.19272292e-01 3.37313652e-01
3.48172754e-01 3.20904106e-01 8.62600923e-01 2.53276266e-02
-1.62885618e+00 3.28944296e-01 6.23930788e+00 7.90153742e-01
-1.34995687e+00 2.12690338e-01 9.23156023e-01 -2.31107026e-01
-1.78836584e-01 -5.14130950e-01 -5.38333714e-01 7.46030062e-02
5.20998120e-01 1.04591511e-01 -3.82800698e-01 9.65558827e-01
-1.48413237e-02 2.36168280e-02 -1.10551107e+00 5.49182296e-01
-1.12985171e-01 -1.55497169e+00 -2.45890886e-01 3.33429188e-01
8.00514638e-01 6.95114374e-01 1.57779157e-01 5.10810494e-01
1.05794117e-01 -1.23609459e+00 -1.39639288e-01 2.65807956e-01
9.25022423e-01 -2.31870025e-01 1.25046265e+00 1.92554489e-01
-1.13874447e+00 5.42984642e-02 -4.70123500e-01 5.71738839e-01
-2.06376091e-01 6.74432397e-01 -1.80167925e+00 5.47318041e-01
6.39924109e-01 5.82526684e-01 -9.49174106e-01 1.48346245e+00
6.44406751e-02 7.36031294e-01 -5.82659900e-01 -2.58180946e-01
3.43139291e-01 3.64717215e-01 3.56285512e-01 1.04283643e+00
6.56003475e-01 9.07710195e-02 1.54617742e-01 1.08381248e+00
-2.97641486e-01 2.75357496e-02 -3.20004284e-01 7.35641122e-02
1.69214979e-01 1.44786763e+00 -9.30180669e-01 -2.04383686e-01
-4.83375967e-01 5.18309951e-01 -6.54945076e-02 -1.58998027e-01
-1.04942155e+00 1.73889145e-01 -1.70892283e-01 5.44325769e-01
4.84402180e-01 5.34829199e-01 -2.28602380e-01 -9.58453059e-01
-8.56016427e-02 -4.81610626e-01 4.35524762e-01 -3.88410538e-01
-1.32435715e+00 6.44801438e-01 -2.34565273e-01 -1.54901898e+00
-3.18311393e-01 -7.47860372e-01 -6.88908339e-01 6.59144878e-01
-1.55738497e+00 -1.23164737e+00 -6.45245433e-01 2.81590611e-01
2.15416685e-01 -1.01574443e-01 8.26586306e-01 1.92398772e-01
-3.50136071e-01 6.81294382e-01 -5.10495715e-02 3.50317538e-01
5.22890031e-01 -1.37018299e+00 3.43867913e-02 4.67797667e-01
-3.96914005e-01 -5.51831499e-02 1.05295833e-02 -6.66793287e-01
-1.07250273e+00 -1.50954711e+00 2.11040512e-01 -2.82149106e-01
5.33551931e-01 6.73752278e-02 -8.50795865e-01 7.80767977e-01
-5.61933853e-02 8.21854770e-01 5.76577365e-01 -4.14281100e-01
1.19503468e-01 3.42099704e-02 -1.26963949e+00 2.79839665e-01
8.09552014e-01 -2.81242859e-02 -2.70323813e-01 5.52708924e-01
6.22439802e-01 -8.87059510e-01 -8.41789961e-01 1.12299061e+00
3.20251763e-01 -1.05492604e+00 1.01764345e+00 -2.17430621e-01
4.40382659e-01 5.14624007e-02 3.19860011e-01 -9.43448007e-01
-6.57330334e-01 1.02786124e-01 2.92980522e-01 2.49792948e-01
7.45294392e-01 -2.86307245e-01 1.45760703e+00 3.47518712e-01
-4.67408776e-01 -1.16197193e+00 -9.21534419e-01 -4.27006781e-01
2.04023987e-01 7.48471636e-03 9.40940678e-02 5.92183650e-01
-2.13723257e-01 -6.96530566e-02 1.18737273e-01 3.65374945e-02
5.33825934e-01 -6.41143173e-02 1.26731023e-01 -1.13639784e+00
-4.36517507e-01 -6.28989339e-01 -3.90825897e-01 -7.16859460e-01
-5.76012909e-01 -1.27328563e+00 1.37892395e-01 -1.67854774e+00
9.58679497e-01 -3.06611747e-01 -4.72496003e-01 5.32268941e-01
-3.60390723e-01 3.96661758e-01 -2.07858011e-01 1.63040787e-01
-5.18389523e-01 4.79701579e-01 1.92826223e+00 -7.09026903e-02
1.34538813e-02 4.79674637e-01 -5.58170021e-01 7.01658964e-01
6.79318190e-01 -6.08465195e-01 -2.95246243e-01 -5.17744496e-02
4.27234219e-03 1.74273297e-01 7.77430534e-01 -1.28218448e+00
9.65752676e-02 1.55814722e-01 9.24459517e-01 -8.04383636e-01
1.74684480e-01 -8.44947100e-01 -1.96266472e-01 1.27605736e+00
-3.83142620e-01 -6.29277825e-01 2.12919801e-01 5.85580766e-01
-2.49498099e-01 -2.33013064e-01 9.73340094e-01 -4.98347968e-01
-2.53075600e-01 6.44045949e-01 -2.47804463e-01 -4.05070372e-02
1.16666985e+00 -2.56555118e-02 -2.59371400e-01 6.50222152e-02
-6.56526268e-01 3.25059146e-01 -1.05638029e-02 8.03273451e-03
5.04589617e-01 -1.46727800e+00 -9.92339671e-01 3.57871830e-01
1.96143508e-01 3.27885777e-01 1.59160003e-01 1.22239602e+00
-7.91948557e-01 7.39071310e-01 -2.41530724e-02 -9.93668556e-01
-9.59592938e-01 1.14088550e-01 1.10951781e+00 -1.01535010e+00
-6.96570337e-01 1.13154435e+00 4.25497949e-01 -6.59014821e-01
2.40295187e-01 -8.89067411e-01 2.05864534e-02 -4.68243927e-01
1.67115316e-01 -9.73848104e-02 3.43729228e-01 2.73717493e-01
-3.13114375e-01 1.16583832e-01 -3.15033168e-01 3.02285016e-01
1.00331914e+00 4.71628398e-01 2.59794831e-01 -1.08247124e-01
1.19072294e+00 -4.34233427e-01 -1.14129496e+00 -3.03663909e-01
-1.69805273e-01 -2.31416166e-01 3.69484395e-01 -1.02378571e+00
-1.20974743e+00 6.80282295e-01 1.23172653e+00 1.53213218e-01
9.98799682e-01 3.12712073e-01 6.71479464e-01 3.47464710e-01
-1.91571340e-01 -4.50935185e-01 4.40482467e-01 9.39749777e-02
6.59579396e-01 -2.00595236e+00 1.99039891e-01 -7.54702985e-01
-5.37226617e-01 1.08835304e+00 9.57143486e-01 -1.90318987e-01
8.68939936e-01 3.78347784e-01 -6.99121580e-02 -3.86042118e-01
-7.43587017e-01 -3.77630681e-01 5.12454093e-01 2.64578402e-01
7.63193548e-01 2.99215764e-01 -2.47104585e-01 7.29220927e-01
4.32208255e-02 2.32753769e-01 3.79919589e-01 8.13359261e-01
-6.99981093e-01 -8.92061710e-01 -8.41825381e-02 1.12947607e+00
-6.53040230e-01 -1.13705061e-01 -9.56603289e-02 1.50296307e+00
-1.19538471e-01 1.11588195e-01 -4.03853916e-02 -9.49943438e-02
2.31464088e-01 -2.24014804e-01 4.02643502e-01 -8.14733863e-01
-8.34434092e-01 3.85906309e-01 -1.08289700e-02 -4.05239940e-01
-2.35340819e-01 -5.19363284e-01 -1.30823302e+00 2.87953764e-01
-5.30035019e-01 -6.25989512e-02 4.56793278e-01 6.03104353e-01
5.96267432e-02 1.26771593e+00 4.84249204e-01 -5.17331958e-01
-9.99308228e-01 -1.16291583e+00 -3.82097542e-01 4.51196358e-02
2.95529038e-01 -3.46909046e-01 -2.55395979e-01 -3.95920455e-01] | [15.376310348510742, -2.1412734985351562] |
96023ce1-38a6-4d56-a129-dc8144814cfb | information-driven-modeling-of-biomolecular | 2103.07508 | null | https://arxiv.org/abs/2103.07508v1 | https://arxiv.org/pdf/2103.07508v1.pdf | Information-Driven Modeling of Biomolecular Complexes | Proteins play crucial roles in every cellular process by interacting with each other, with nucleic acids, metabolites, and other molecules. The resulting assemblies can be very large and intricate and pose challenges to experimental methods. In the current era of integrative modeling, it is often only by a combination of various experimental techniques and computations that 3D models of those molecular machines can be obtained. Among the various computational approaches available, molecular docking is often the method of choice when it comes to predicting 3D structures of complexes. Docking can generate particularly accurate models when taking into account the available information on the complex of interest. We review here the use of experimental and bioinformatics data in protein-protein docking, describing recent software developments and highlighting applications for the modeling of antibody-antigen complexes and membrane protein complexes, and the use of evolutionary and shape information. | ['Alexandre M. J. J. Bonvin', 'Rodrigo V. Honorato', 'Charlotte W. van Noort'] | 2021-03-12 | null | null | null | null | ['molecular-docking'] | ['medical'] | [ 1.34503558e-01 -4.32819217e-01 -1.31652534e-01 3.11944969e-02
-3.00566763e-01 -8.26820493e-01 2.60127664e-01 8.05312574e-01
-6.77215457e-01 1.31297028e+00 -2.66017914e-01 -7.22504377e-01
1.71396136e-01 -3.87097180e-01 -5.91839552e-01 -1.00665557e+00
-4.03941795e-02 9.70745385e-01 3.99634808e-01 -3.18310350e-01
3.73293877e-01 1.08280182e+00 -1.35024107e+00 -5.33371717e-02
1.08920348e+00 3.38839233e-01 5.41933894e-01 4.74602014e-01
-3.88528436e-01 3.20900790e-02 -4.49428707e-01 -6.85650170e-01
-2.57890493e-01 -4.82310474e-01 -2.43471324e-01 -1.53164968e-01
-2.76021779e-01 4.12288517e-01 4.22211468e-01 5.17879248e-01
7.29448318e-01 1.00001626e-01 7.42469490e-01 -4.31386918e-01
-9.30934027e-02 -3.32164347e-01 -2.11630493e-01 1.10305190e-01
6.38065517e-01 1.76928550e-01 6.07327402e-01 -1.20185316e+00
9.47808862e-01 1.09777880e+00 5.28965950e-01 4.17595178e-01
-1.74217844e+00 -4.25206512e-01 -1.26572326e-01 2.12696120e-01
-1.24644065e+00 -3.15890610e-01 3.38170797e-01 -6.23133302e-01
1.53025818e+00 3.63279313e-01 8.08420360e-01 7.11053252e-01
6.98529601e-01 -6.01379341e-03 9.70894694e-01 -3.77761871e-01
3.58364046e-01 -7.60196522e-02 -1.09286845e-01 4.93655413e-01
4.03837264e-01 -2.01140299e-01 -3.53779137e-01 -7.69308686e-01
3.70855480e-01 1.38296127e-01 -1.61071047e-01 -6.71068847e-01
-9.41166699e-01 7.22474456e-01 4.00369465e-02 2.82248020e-01
-5.13120770e-01 -4.09340352e-01 1.80659056e-01 -2.65132815e-01
3.85771208e-02 6.29852474e-01 -7.06171274e-01 1.05577998e-01
-3.82179946e-01 3.84205341e-01 6.57381654e-01 2.29233459e-01
6.66056335e-01 -4.65387046e-01 8.92321408e-01 6.85459614e-01
3.12515587e-01 1.41047269e-01 1.05866395e-01 -2.20588401e-01
-1.13487035e-01 7.98426807e-01 4.62038755e-01 -7.14830220e-01
-6.30182862e-01 -9.01852250e-02 -5.93483388e-01 4.29292977e-01
6.01838768e-01 7.77828619e-02 -5.59052348e-01 1.46272480e+00
6.93089008e-01 -1.37580365e-01 5.80034479e-02 6.65063560e-01
6.52396023e-01 3.73020738e-01 6.41319573e-01 -8.65946591e-01
1.39144742e+00 -4.12523508e-01 -4.46030021e-01 1.65600240e-01
4.84018177e-01 -1.11740696e+00 2.53412157e-01 3.51017147e-01
-1.02339244e+00 -2.82958806e-01 -1.01049256e+00 1.13468133e-01
-5.26647627e-01 -1.56511173e-01 6.03768766e-01 4.49040800e-01
-5.52793503e-01 6.02204084e-01 -1.04625118e+00 -4.94680345e-01
-8.02464858e-02 7.59632349e-01 -6.26837015e-01 -9.45682153e-02
-7.71443844e-01 1.46098709e+00 4.23720747e-01 3.49594615e-02
-2.72036731e-01 -3.59214932e-01 -3.52116197e-01 -1.34657025e-01
2.99327802e-02 -8.03415239e-01 6.73904300e-01 -4.47474688e-01
-1.15008163e+00 8.18767965e-01 -4.90443856e-01 -1.17931895e-01
6.96308836e-02 4.00026172e-01 -5.88901229e-02 -3.04543711e-02
-4.17796314e-01 3.29523295e-01 -6.08776277e-03 -7.82958090e-01
-1.23720124e-01 -6.13477588e-01 -3.07496667e-01 3.10784936e-01
6.90829515e-01 4.74958301e-01 6.52503073e-02 -1.57324642e-01
3.74237388e-01 -9.03418958e-01 -9.17945206e-01 -1.29767448e-01
-1.07418761e-01 -1.59571618e-01 3.29073846e-01 -5.15059292e-01
7.02965796e-01 -1.58243501e+00 8.37533236e-01 2.58971930e-01
3.89654875e-01 4.90332723e-01 2.32977256e-01 1.04688823e+00
-3.96955043e-01 8.56274888e-02 1.12227730e-01 1.62236825e-01
-3.54238749e-01 -1.33815184e-01 2.06620157e-01 4.80353475e-01
5.52577637e-02 7.23435163e-01 -5.93085110e-01 -3.71227026e-01
3.43751609e-01 8.45579207e-01 -3.34204197e-01 1.71699166e-01
-3.95922929e-01 8.16171944e-01 -6.63258016e-01 6.57614887e-01
5.85110426e-01 -3.11285138e-01 9.70393360e-01 -8.55640471e-02
-2.73678422e-01 2.57569700e-01 -7.73763061e-01 1.13984346e+00
1.07099719e-01 6.33262023e-02 2.56752744e-02 -5.86156845e-01
8.86135578e-01 5.64179480e-01 4.42663074e-01 -2.55025893e-01
1.26645193e-01 3.36974710e-01 6.47640407e-01 1.73965923e-03
-1.62351564e-01 -5.06037772e-01 3.30833822e-01 2.34374583e-01
-4.82880250e-02 1.03566319e-01 2.85425067e-01 6.69684112e-02
6.90076411e-01 1.11675814e-01 6.92823946e-01 -1.64803877e-01
6.63945079e-01 2.41124630e-01 4.77854580e-01 -6.64758086e-02
5.80430888e-02 1.72689959e-01 4.69346970e-01 -7.08785415e-01
-1.44003880e+00 -6.98984146e-01 -3.62910271e-01 6.95154488e-01
-1.85649358e-02 -4.48319167e-01 -7.00262964e-01 -5.76569140e-02
-9.68298763e-02 2.26056203e-01 -3.61782432e-01 1.03496732e-02
-2.73035616e-01 -1.29117417e+00 7.85433725e-02 -2.07317948e-01
-4.15862590e-01 -7.79063582e-01 -3.50295514e-01 7.41959214e-01
1.48346275e-01 -7.43760347e-01 1.24170780e-02 5.72936833e-01
-8.71022224e-01 -1.52630210e+00 -5.29535413e-01 -6.68066800e-01
7.51002550e-01 2.91137964e-01 8.09067607e-01 4.29535270e-01
-5.53341925e-01 -2.29894608e-01 4.11394611e-02 -6.04678273e-01
-6.19190693e-01 -2.07891539e-01 4.38772559e-01 -3.52902412e-01
3.75742674e-01 -9.04707968e-01 -5.82337141e-01 5.51330745e-01
-6.67122483e-01 2.81721413e-01 3.40164870e-01 8.97950947e-01
9.14221168e-01 -1.80152476e-01 5.56919158e-01 -5.54110467e-01
6.57712400e-01 -3.70288312e-01 -7.24280953e-01 3.45414639e-01
-2.95418739e-01 -1.11770906e-01 5.34756899e-01 -4.54742253e-01
-8.03642929e-01 4.27553564e-01 -4.60610449e-01 8.94860830e-03
-2.79987186e-01 6.01134896e-01 -5.51097453e-01 -4.49260890e-01
6.77079976e-01 4.30445284e-01 5.44503748e-01 -7.39112377e-01
-3.01816702e-01 2.36357585e-01 -1.58681441e-02 -5.23168623e-01
4.62646216e-01 1.18675746e-01 4.02764112e-01 -1.08920360e+00
-2.21024454e-01 -2.51525015e-01 -9.49664831e-01 1.29569406e-02
9.86748576e-01 -4.66769904e-01 -1.25087655e+00 3.92087877e-01
-1.07203805e+00 2.81442627e-02 6.92103922e-01 5.82400143e-01
-3.24058980e-01 6.64825499e-01 -3.99008751e-01 -7.57342577e-01
-1.60666138e-01 -1.74113047e+00 5.88131249e-01 2.30088905e-01
-2.02916875e-01 -9.79581714e-01 4.40608025e-01 4.96451765e-01
1.61491498e-01 5.06903887e-01 1.44538951e+00 -7.66163170e-01
-5.36995590e-01 -3.69132429e-01 3.62273276e-01 -2.43017793e-01
2.23218367e-01 3.42664868e-01 -3.54861438e-01 -1.26607805e-01
-3.33258927e-01 -2.98199713e-01 3.76920670e-01 2.25266829e-01
3.44213039e-01 1.30298838e-01 -6.95361495e-01 2.88601279e-01
1.28194427e+00 6.95456326e-01 7.68515706e-01 2.96676278e-01
5.10519564e-01 7.68842995e-01 7.85430908e-01 1.11390449e-01
-2.94393860e-02 1.04495287e+00 3.51136625e-01 -1.80359721e-01
5.11344016e-01 3.27451639e-02 -1.51742101e-01 4.86246735e-01
-6.66716695e-01 -1.06291093e-01 -1.04358554e+00 -1.57893211e-01
-1.75419962e+00 -1.14827311e+00 -3.51583779e-01 2.45676208e+00
9.68441904e-01 -1.56693459e-02 2.55364835e-01 -2.28788286e-01
5.95794141e-01 -5.69057286e-01 -8.08135450e-01 -4.39411759e-01
-4.75529224e-01 3.12413305e-01 8.76804292e-02 6.09420538e-01
-6.33687377e-01 6.39271259e-01 7.33242559e+00 5.81261396e-01
-1.11009967e+00 -3.10031444e-01 3.96116704e-01 1.44971684e-02
-4.56304522e-03 2.06708565e-01 -8.57089102e-01 4.45716202e-01
1.02473390e+00 -3.02536875e-01 3.18597585e-01 4.25875068e-01
3.96700770e-01 -4.93948340e-01 -8.31378818e-01 7.65351713e-01
-5.48670232e-01 -1.66497540e+00 -2.25073531e-01 4.16084200e-01
3.41983825e-01 -1.41340364e-02 -3.11425477e-01 -3.14100116e-01
7.89364651e-02 -1.20361149e+00 7.20158517e-02 5.24185836e-01
4.05001342e-01 -8.06472421e-01 6.99514449e-01 5.97019613e-01
-8.42932642e-01 4.87306982e-01 -4.88183111e-01 1.21278241e-02
2.22353369e-01 2.81863064e-01 -9.58750188e-01 3.51016551e-01
2.29232192e-01 8.11364427e-02 -2.35877439e-01 1.31536281e+00
1.80554777e-01 5.36369346e-02 -3.68331462e-01 -2.67152756e-01
-1.95305601e-01 -7.61955142e-01 2.73346543e-01 7.75777221e-01
-1.20228022e-01 3.81400824e-01 1.99380875e-01 3.81158143e-01
3.17060441e-01 4.99733537e-01 -3.48630160e-01 -1.44891098e-01
3.33284318e-01 1.15004849e+00 -9.21784341e-01 -2.02383682e-01
-4.82647896e-01 6.20556593e-01 4.75183040e-01 2.28727818e-01
-5.78925967e-01 2.10047252e-02 1.23321736e+00 1.31873831e-01
2.01062024e-01 -5.46372771e-01 5.16799510e-01 -9.05328035e-01
-2.65621811e-01 -9.27380025e-01 1.62460297e-01 -6.62489295e-01
-8.26617420e-01 3.12015682e-01 -2.70142376e-01 -6.34176314e-01
-1.14683621e-01 -7.98359513e-01 -3.19983065e-01 1.25299740e+00
-9.17937338e-01 -8.21035385e-01 3.38133067e-01 9.41716041e-03
1.83616728e-01 2.52761487e-02 1.32536376e+00 1.49953574e-01
-7.12128401e-01 -7.32077472e-03 8.13859046e-01 -8.01462173e-01
7.18920529e-01 -8.23291659e-01 1.64632887e-01 2.38279983e-01
-3.56963456e-01 9.25730705e-01 1.30400050e+00 -8.02075148e-01
-1.63770783e+00 -4.57143664e-01 7.73478687e-01 -4.79014397e-01
5.65908134e-01 -5.69827557e-01 -1.07516384e+00 3.87231290e-01
-6.12489730e-02 -3.66665334e-01 1.38634670e+00 -2.17341427e-02
-1.71323091e-01 6.13685966e-01 -1.08451188e+00 5.89884222e-01
5.31887293e-01 -1.89173117e-01 -2.96777457e-01 4.93837923e-01
2.52990276e-01 -5.67634702e-01 -1.12348187e+00 9.43698511e-02
6.76684618e-01 -1.04082108e+00 1.21417785e+00 -1.04036760e+00
-2.16923490e-01 -7.15182900e-01 2.75665112e-02 -1.02946568e+00
-5.02002597e-01 -4.57519710e-01 2.74075478e-01 5.16093433e-01
5.57323635e-01 -6.66965246e-01 8.18985164e-01 8.61768246e-01
7.12740645e-02 -9.38326538e-01 -1.02723670e+00 -3.34394813e-01
-1.91221219e-02 1.45236939e-01 5.14396787e-01 6.62046134e-01
3.08089584e-01 6.04104578e-01 -8.04671571e-02 -2.26077348e-01
5.03480256e-01 2.04565465e-01 9.35333729e-01 -1.55268180e+00
-3.41848910e-01 -3.01045746e-01 -6.22272551e-01 -3.03017676e-01
-3.91582139e-02 -6.41197860e-01 -5.66690862e-01 -1.51707637e+00
4.65145528e-01 -3.39957595e-01 4.13402207e-02 1.20186999e-01
-5.24643026e-02 9.72154886e-02 -8.87125582e-02 3.52223694e-01
-3.86359930e-01 3.90301913e-01 1.08890009e+00 1.68528363e-01
-2.00641081e-01 7.98458755e-02 -4.77977335e-01 5.50082743e-01
6.93395853e-01 -4.20580208e-01 -5.68329692e-02 4.64740425e-01
4.29404169e-01 1.16654888e-01 -1.53494850e-01 -4.24332768e-01
-1.44149484e-02 -4.43444639e-01 5.56255281e-01 -5.80898821e-01
6.49632096e-01 -8.80955338e-01 1.16385329e+00 7.99695551e-01
3.50174069e-01 1.56191185e-01 3.52978498e-01 5.84755659e-01
9.24337730e-02 -1.62060618e-01 1.01888263e+00 -1.81984931e-01
-9.43891034e-02 2.87594423e-02 -8.78073752e-01 -5.80439925e-01
1.08615851e+00 -4.84610260e-01 -1.43412665e-01 -3.72339878e-03
-1.28750122e+00 -1.35435313e-01 1.06974328e+00 -9.07573756e-03
4.42162335e-01 -8.38476539e-01 -3.06155115e-01 -5.10626063e-02
1.82944387e-01 -3.11262339e-01 2.06871450e-01 1.03943431e+00
-8.69652331e-01 8.47938538e-01 -4.48302478e-01 -4.06563163e-01
-2.00270987e+00 9.55726981e-01 5.62534034e-01 -3.01471651e-01
3.79653163e-02 2.21735850e-01 4.34374064e-01 -3.05037796e-01
-1.13000251e-01 4.32088196e-01 -2.96637237e-01 -1.10624105e-01
6.47489548e-01 1.83095895e-02 3.31288099e-01 -9.61377084e-01
-5.54896176e-01 5.07069111e-01 -2.01588944e-01 5.08764803e-01
1.61928904e+00 1.47993648e-02 -5.22827506e-01 8.67215022e-02
7.42470682e-01 5.91146275e-02 -8.10585380e-01 -1.07787098e-04
-4.94826436e-02 -1.78695768e-01 -4.36025232e-01 -7.35078275e-01
-2.92397797e-01 7.47266829e-01 3.05167437e-01 -1.92754969e-01
8.59516203e-01 7.81639442e-02 2.29531065e-01 4.88741904e-01
7.41244316e-01 -6.07751429e-01 -2.95208246e-01 1.51682958e-01
6.66154444e-01 -9.97769296e-01 1.74863786e-01 -5.78577876e-01
-2.44048476e-01 1.06832969e+00 4.44619745e-01 1.54043093e-01
4.01611656e-01 1.04081213e-01 -1.54896662e-01 -1.33219153e-01
-1.12476873e+00 -6.92100450e-02 1.24583736e-01 6.34840786e-01
8.99298429e-01 -1.31660387e-01 -9.74897265e-01 4.44351435e-01
3.45531791e-01 -3.63098234e-01 3.37122619e-01 1.12532580e+00
-6.93778992e-01 -2.04405165e+00 -6.47892237e-01 7.93954059e-02
-8.14812899e-01 3.67006194e-03 -1.05184317e+00 6.99904680e-01
-1.84845671e-01 6.99161947e-01 -3.13082397e-01 -7.34306350e-02
2.03655675e-01 4.06736434e-01 8.09413314e-01 -4.33948338e-01
-6.41870260e-01 5.39625525e-01 2.29591444e-01 -9.00973678e-02
-4.26412463e-01 -8.11256230e-01 -1.39500570e+00 -5.30511975e-01
-7.43130863e-01 6.88047469e-01 1.09624696e+00 8.05822372e-01
6.73130989e-01 2.15623155e-02 3.82717043e-01 -1.19904733e+00
1.55038625e-01 -6.68927729e-01 -5.87736487e-01 -6.34519160e-02
-5.43272533e-02 -8.86231780e-01 -4.90165688e-02 -4.70425710e-02] | [4.75002908706665, 5.304846286773682] |
b906a59e-6408-484a-85c7-246ecec3975b | motion-modulated-temporal-fragment-alignment | null | null | http://openaccess.thecvf.com//content/CVPR2022/html/Wu_Motion-Modulated_Temporal_Fragment_Alignment_Network_for_Few-Shot_Action_Recognition_CVPR_2022_paper.html | http://openaccess.thecvf.com//content/CVPR2022/papers/Wu_Motion-Modulated_Temporal_Fragment_Alignment_Network_for_Few-Shot_Action_Recognition_CVPR_2022_paper.pdf | Motion-Modulated Temporal Fragment Alignment Network for Few-Shot Action Recognition | While the majority of FSL models focus on image classification, the extension to action recognition is rather challenging due to the additional temporal dimension in videos. To address this issue, we propose an end-to-end Motion-modulated Temporal Fragment Alignment Network (MTFAN) by jointly exploring the task-specific motion modulation and the multi-level temporal fragment alignment for Few-Shot Action Recognition (FSAR). The proposed MTFAN model enjoys several merits. First, we design a motion modulator conditioned on the learned task-specific motion embeddings, which can activate the channels related to the task-shared motion patterns for each frame. Second, a segment attention mechanism is proposed to automatically discover the higher-level segments for multi-level temporal fragment alignment, which encompasses the frame-to-frame, segment-to-segment, and segment-to-frame alignments. To the best of our knowledge, this is the first work to exploit task-specific motion modulation for FSAR. Extensive experimental results on four standard benchmarks demonstrate that the proposed model performs favorably against the state-of-the-art FSAR methods. | ['Yongdong Zhang', 'Feng Wu', 'Zhe Zhang', 'Tianzhu Zhang', 'Jiamin Wu'] | 2022-01-01 | null | null | null | cvpr-2022-1 | ['few-shot-action-recognition'] | ['computer-vision'] | [ 5.01423180e-01 -2.98239410e-01 -7.46663272e-01 -4.62242693e-01
-7.53860533e-01 -8.37575197e-02 5.99400222e-01 -3.96065533e-01
-2.64101535e-01 3.65121245e-01 5.28116465e-01 1.15347378e-01
-3.30178365e-02 -2.61824667e-01 -6.28195643e-01 -8.24169397e-01
-2.78791219e-01 -2.01483682e-01 4.89188164e-01 2.69701220e-02
3.93564254e-01 2.78761744e-01 -1.39221084e+00 6.61025882e-01
4.86993700e-01 1.15752268e+00 3.28232974e-01 5.36378860e-01
-2.27739755e-02 1.19459522e+00 -1.73108220e-01 2.34993130e-01
1.46142974e-01 -7.68326998e-01 -8.88594449e-01 4.02526617e-01
6.13359928e-01 -4.98304009e-01 -6.50706768e-01 7.03605413e-01
2.07845703e-01 5.20280302e-01 4.45007175e-01 -1.18475127e+00
-6.26624107e-01 3.72898906e-01 -7.40743816e-01 7.18673646e-01
3.34064156e-01 4.97567356e-01 1.13009465e+00 -9.99499023e-01
7.41533339e-01 1.32618737e+00 1.42676324e-01 5.42576909e-01
-8.97786856e-01 -5.03157139e-01 5.58098495e-01 7.32110500e-01
-1.11235499e+00 -4.75780100e-01 9.22136068e-01 -6.16306126e-01
1.10776567e+00 -6.56109676e-02 6.24551117e-01 1.24159718e+00
4.16667044e-01 1.21237350e+00 6.73911989e-01 -2.58609176e-01
1.98799476e-01 -8.36973727e-01 3.98033559e-02 7.16720939e-01
-3.75005692e-01 -6.93524769e-03 -8.08069348e-01 2.54605591e-01
1.02356088e+00 9.48171839e-02 -4.05320495e-01 -4.59719330e-01
-1.38644958e+00 7.83810735e-01 2.47865707e-01 6.04172647e-01
-4.45839167e-01 5.34245729e-01 5.52867591e-01 -6.32533953e-02
2.88406461e-01 2.11754262e-01 -3.68910551e-01 -4.75825071e-01
-9.94065762e-01 5.94807155e-02 2.00414464e-01 6.48601711e-01
5.66799581e-01 3.57744128e-01 -6.16193056e-01 6.74180031e-01
4.76020515e-01 2.09151894e-01 7.59290218e-01 -1.18929446e+00
6.26321614e-01 4.12965894e-01 -4.22867015e-02 -8.77217293e-01
-2.73151755e-01 -1.93737239e-01 -6.45663798e-01 -1.70400925e-02
2.37617418e-01 2.44166747e-01 -9.28139389e-01 1.73203123e+00
2.63842314e-01 8.73699486e-01 1.80481132e-02 1.23099065e+00
6.21400416e-01 7.57076800e-01 2.90838033e-01 -3.57724518e-01
1.29847503e+00 -1.53598285e+00 -9.35774624e-01 -3.98631960e-01
7.34364152e-01 -5.51540315e-01 1.00284863e+00 1.98106114e-02
-1.04867721e+00 -9.28573012e-01 -9.35946524e-01 -7.86167905e-02
1.69155404e-01 2.40014702e-01 4.00625139e-01 3.39262858e-02
-5.45434356e-01 3.88795197e-01 -1.19429302e+00 -3.00099194e-01
6.93826377e-01 1.11741416e-01 -4.10968691e-01 -1.67883128e-01
-1.11051190e+00 4.79863673e-01 4.31873798e-01 2.59905219e-01
-1.05564725e+00 -4.33163732e-01 -1.14791715e+00 8.24751109e-02
5.26530504e-01 -5.06750882e-01 1.31722736e+00 -1.28145123e+00
-1.58088267e+00 4.23430502e-01 -5.32285154e-01 -6.01757228e-01
1.65873587e-01 -3.88338387e-01 -4.09279794e-01 4.97404784e-01
1.59558982e-01 8.57523561e-01 1.05104399e+00 -8.25619936e-01
-7.42982626e-01 -2.24152818e-01 8.70478600e-02 3.21094900e-01
-3.65472466e-01 7.56745040e-02 -5.97861171e-01 -1.12342894e+00
-4.29351218e-02 -7.62147665e-01 -2.83059299e-01 5.04936976e-03
-8.45986679e-02 -2.48006999e-01 1.04710793e+00 -5.98917663e-01
1.60846770e+00 -2.37586999e+00 5.27476668e-01 -3.84075761e-01
-2.19606012e-01 4.43005085e-01 -4.28766876e-01 3.15765321e-01
-1.19455330e-01 -2.49856263e-01 -2.99604893e-01 -3.52997512e-01
-1.99092954e-01 2.95349449e-01 -3.84820491e-01 4.82936531e-01
5.24336278e-01 1.13979077e+00 -1.05878961e+00 -5.63050687e-01
4.74730760e-01 4.52303529e-01 -5.18937171e-01 2.18988836e-01
-2.95257419e-01 7.55047739e-01 -7.57113695e-01 6.93342984e-01
1.21828191e-01 -3.19791794e-01 -6.08142652e-03 -3.32967669e-01
-1.60303250e-01 2.17256233e-01 -8.60550702e-01 2.18811440e+00
-1.67489201e-01 6.80948377e-01 -3.78007621e-01 -1.08842969e+00
6.33768678e-01 3.26563269e-01 8.94526422e-01 -8.52102041e-01
-5.81703614e-03 7.39158317e-02 -7.10662641e-03 -7.90998101e-01
3.61919522e-01 -3.95455770e-03 -2.43637394e-02 2.83219099e-01
1.45660102e-01 6.28317595e-01 2.96184152e-01 -7.67293498e-02
1.05371070e+00 5.54880917e-01 4.32821959e-01 7.44451955e-03
7.59078026e-01 -2.16800481e-01 8.04974258e-01 3.14682156e-01
-5.82944989e-01 5.18054307e-01 3.48798633e-01 -6.46861732e-01
-7.76159883e-01 -7.81061590e-01 2.31334880e-01 1.15963137e+00
2.36694098e-01 -3.83961976e-01 -5.96882343e-01 -7.06329226e-01
-2.21959710e-01 5.57129800e-01 -6.62808001e-01 -2.87657797e-01
-1.00426543e+00 -2.90196061e-01 2.70548820e-01 9.39081907e-01
6.73341990e-01 -1.32760775e+00 -8.64146590e-01 4.09513414e-01
-3.60901862e-01 -1.48332357e+00 -9.87321734e-01 -2.25601643e-01
-9.77734447e-01 -1.07612431e+00 -7.21897721e-01 -8.32076013e-01
2.45091587e-01 5.02511680e-01 6.65561914e-01 -2.86770821e-01
-2.84196019e-01 3.85041714e-01 -6.53768182e-01 1.51722342e-01
1.02404907e-01 -1.46895498e-01 -1.99212298e-01 5.48747778e-01
3.81246984e-01 -4.61873412e-01 -8.97915602e-01 5.09541690e-01
-1.03641653e+00 1.17192224e-01 6.90091968e-01 7.67574728e-01
8.73738885e-01 -2.22934619e-01 5.51110983e-01 -4.77805644e-01
-7.80343264e-02 -4.19044316e-01 -1.22341663e-01 1.35767356e-01
-4.94142110e-03 -2.65121721e-02 4.39949274e-01 -5.26690781e-01
-7.85096049e-01 3.56750876e-01 7.86765739e-02 -9.87563014e-01
-2.31639162e-01 4.45761681e-01 -1.90699965e-01 2.74727046e-01
2.25962460e-01 4.74166453e-01 -1.47681579e-01 -4.11340445e-01
4.35734391e-01 3.91999513e-01 5.91552556e-01 -2.23203555e-01
4.69532281e-01 7.15819240e-01 -3.25109810e-02 -8.16265464e-01
-1.02359188e+00 -7.55553067e-01 -8.59556556e-01 -3.27856779e-01
1.52115750e+00 -9.62757587e-01 -2.20555797e-01 5.16182542e-01
-1.06773055e+00 -5.85984349e-01 -2.84452230e-01 7.51304507e-01
-1.00113022e+00 6.46310329e-01 -5.13386965e-01 -6.85038269e-01
-1.47014588e-01 -1.39914441e+00 1.27982163e+00 -2.69716270e-02
-2.34882951e-01 -7.30704188e-01 8.47781450e-02 5.11449158e-01
6.00630119e-02 1.85738146e-01 6.68796539e-01 -3.05775076e-01
-8.33613276e-01 1.56852975e-01 -7.88848698e-02 3.22112799e-01
2.46591881e-01 -1.08413681e-01 -6.43101633e-01 -3.88404936e-01
-4.72665243e-02 -3.20981324e-01 1.22931099e+00 6.52852237e-01
1.35645831e+00 -1.57932281e-01 -2.41345927e-01 6.09806478e-01
1.00560117e+00 4.86365765e-01 8.48467290e-01 3.17064911e-01
1.03831351e+00 2.48799026e-01 1.14765966e+00 4.58058715e-01
2.56135792e-01 1.20558190e+00 4.13276762e-01 1.23851851e-01
-2.57486433e-01 -1.42557040e-01 7.91412175e-01 6.90197825e-01
-9.01155993e-02 -1.94875360e-01 -5.29482603e-01 6.21457577e-01
-2.14136219e+00 -1.24040663e+00 7.00119734e-02 1.93201673e+00
4.28326309e-01 8.37542713e-02 2.20325992e-01 -2.48576067e-02
5.58557689e-01 8.44840229e-01 -5.63765466e-01 -2.29506746e-01
1.00357108e-01 1.92296520e-01 1.49336830e-01 3.38093698e-01
-1.48085093e+00 1.08220470e+00 5.23513556e+00 9.58854437e-01
-1.08662200e+00 9.74301174e-02 3.57867599e-01 -2.43458584e-01
9.00477767e-02 2.12780293e-02 -7.08187938e-01 5.60844541e-01
7.48530269e-01 2.48142779e-01 2.24140018e-01 7.67307580e-01
5.75386167e-01 8.02647918e-02 -1.12344265e+00 9.53478336e-01
1.91467404e-01 -1.32804394e+00 2.55544871e-01 -9.83013026e-03
7.14178681e-01 -1.66949481e-01 -3.40293013e-02 2.57249266e-01
-4.50815231e-01 -9.68196213e-01 7.78389215e-01 5.59325337e-01
7.88773477e-01 -7.08885074e-01 3.67481083e-01 4.13006991e-02
-1.77667999e+00 -3.58223975e-01 -1.02151200e-01 4.78305742e-02
4.49914187e-01 1.26276135e-01 -2.11091176e-01 5.35053253e-01
4.30323899e-01 1.46088803e+00 -2.87539512e-01 8.13586771e-01
-1.33565739e-01 7.01546252e-01 2.98562407e-01 3.81068110e-01
7.43304968e-01 -9.55414493e-03 6.04433000e-01 1.20214415e+00
3.68231565e-01 1.08239345e-01 4.34520394e-01 4.86055464e-01
2.09764063e-01 -1.41133498e-02 -5.35318494e-01 -3.85896921e-01
1.32131562e-01 1.00451303e+00 -5.30155241e-01 -2.04515263e-01
-7.25643158e-01 1.32153726e+00 1.91738144e-01 4.09498483e-01
-9.35678482e-01 -1.13859199e-01 8.70226800e-01 7.40459934e-02
8.50360870e-01 -4.34968680e-01 2.24856943e-01 -1.12668920e+00
1.68476224e-01 -9.08609629e-01 6.43734276e-01 -5.17224252e-01
-1.06530094e+00 3.62009615e-01 -5.93197793e-02 -1.60102296e+00
-2.76537627e-01 -3.31462592e-01 -6.00512803e-01 4.22703803e-01
-1.56847334e+00 -1.28649962e+00 -1.59177005e-01 7.14036584e-01
1.31795716e+00 -3.36016297e-01 5.80994129e-01 3.30910206e-01
-6.71464384e-01 4.99054343e-01 -1.11629099e-01 3.07642817e-01
5.47946930e-01 -7.09566951e-01 4.07426208e-01 9.68001127e-01
3.78772706e-01 2.29664505e-01 2.35323593e-01 -5.76071084e-01
-1.37695408e+00 -1.38027239e+00 7.16123581e-01 -2.42321953e-01
6.92680955e-01 -5.45304269e-02 -1.01951361e+00 7.61391222e-01
4.82070148e-02 3.59315097e-01 6.36748374e-01 -3.61951709e-01
-3.11879843e-01 -2.19510049e-02 -4.68093574e-01 5.66926837e-01
1.25359535e+00 -5.33471048e-01 -5.81236780e-01 5.90927601e-02
6.49341404e-01 -2.50437796e-01 -8.07123482e-01 5.82535148e-01
6.20309174e-01 -8.13630998e-01 1.00143623e+00 -8.44386518e-01
8.28827500e-01 -5.29885530e-01 -3.96138847e-01 -8.86853456e-01
-5.18399298e-01 -6.04085803e-01 -5.60820401e-01 1.00779665e+00
-1.03741780e-01 -6.06759489e-02 5.49824059e-01 -8.29964597e-03
-5.07327020e-01 -1.07022452e+00 -1.24324155e+00 -8.32473695e-01
-2.32603461e-01 -4.27558362e-01 1.53280780e-01 9.06776726e-01
-2.40058020e-01 3.38936567e-01 -7.65255809e-01 9.82903019e-02
3.35051745e-01 4.00858909e-01 6.65623128e-01 -6.84103847e-01
-6.76942408e-01 -5.17443955e-01 -7.35169947e-01 -1.50141358e+00
3.43655795e-01 -7.16791987e-01 1.54715389e-01 -1.50153184e+00
2.65800536e-01 1.81658864e-01 -6.48642421e-01 4.16034520e-01
-3.53283256e-01 2.27946237e-01 3.75436276e-01 3.10224771e-01
-1.02146399e+00 8.32479954e-01 1.61646903e+00 -9.02602971e-02
-1.94652632e-01 -2.30912462e-01 -2.45496795e-01 6.69321656e-01
5.61515629e-01 -2.67472655e-01 -6.23138309e-01 -4.15739834e-01
-5.17245948e-01 2.33153850e-01 4.43121135e-01 -1.12762523e+00
-9.50414862e-04 -5.68396628e-01 1.96106359e-01 -6.76155448e-01
5.81205010e-01 -6.17269158e-01 -1.16931833e-01 4.89553154e-01
-5.48445284e-01 -1.71464637e-01 3.15239094e-03 9.50734317e-01
-4.51130629e-01 8.00117403e-02 8.49063694e-01 7.22577050e-02
-1.24738634e+00 5.67214191e-01 -4.64296550e-01 1.11787125e-01
1.34386289e+00 -3.93907130e-01 -1.50248781e-01 -2.21703827e-01
-6.59022689e-01 2.15426639e-01 2.13604331e-01 7.83119977e-01
7.43809402e-01 -1.67448652e+00 -5.69844425e-01 3.13958466e-01
3.60742539e-01 -1.58181936e-01 5.19205391e-01 1.09073830e+00
-1.34639040e-01 6.41516864e-01 -4.07115132e-01 -7.40076482e-01
-1.28478253e+00 6.39117420e-01 1.73821747e-01 -3.53796601e-01
-8.48364651e-01 8.07500243e-01 6.68231308e-01 3.80533993e-01
2.87517101e-01 -4.13079292e-01 -2.78443128e-01 -3.33824269e-02
6.85886562e-01 2.96594381e-01 -4.67856914e-01 -1.03964961e+00
-4.12264109e-01 8.35777164e-01 -7.56907687e-02 1.39967520e-02
1.13822830e+00 -1.40565172e-01 2.02030569e-01 6.59348488e-01
1.25618851e+00 -5.83718061e-01 -1.69336271e+00 -3.11174244e-01
-2.09842948e-03 -8.69046807e-01 -2.17616949e-02 -3.02472323e-01
-1.22249210e+00 1.05402350e+00 5.32161236e-01 -4.27543610e-01
1.25777507e+00 -1.00191221e-01 1.24032629e+00 1.94873109e-01
3.12323838e-01 -1.12467813e+00 7.70565987e-01 5.43540776e-01
7.79697537e-01 -1.02279210e+00 -1.94652483e-01 -3.08759660e-01
-9.24623728e-01 9.79216874e-01 7.71880209e-01 -9.64634195e-02
4.30280089e-01 -1.93006009e-01 -1.06231712e-01 -1.02553688e-01
-8.38872015e-01 -2.92371094e-01 5.53531408e-01 4.41412747e-01
4.42728013e-01 -2.02663645e-01 -3.97909671e-01 4.03063715e-01
5.86330175e-01 3.05987835e-01 1.10195242e-01 1.05239141e+00
-5.80358505e-01 -1.17240047e+00 9.15732421e-03 3.12545180e-01
-3.62837523e-01 1.61627233e-01 -2.25387916e-01 5.86852908e-01
9.13970843e-02 8.05313885e-01 2.20310584e-01 -2.72271276e-01
2.88818955e-01 1.89614117e-01 5.26951969e-01 -5.72776198e-01
-3.05605471e-01 4.48025346e-01 -5.82348555e-02 -1.10038793e+00
-8.60517979e-01 -8.15787196e-01 -1.49596977e+00 3.38985264e-01
6.49815872e-02 -2.32701927e-01 6.78275898e-02 1.02391756e+00
5.39136946e-01 6.91526175e-01 6.22493029e-01 -1.10983264e+00
-3.44541997e-01 -9.19269085e-01 -4.82474625e-01 5.46302497e-01
5.20096600e-01 -9.20604765e-01 -1.14009306e-02 2.42868006e-01] | [8.565207481384277, 0.5776668190956116] |
a899bcc5-1177-4240-9e2b-1b8c6d582122 | speaker-oriented-latent-structures-for | 2109.05182 | null | https://arxiv.org/abs/2109.05182v2 | https://arxiv.org/pdf/2109.05182v2.pdf | Speaker-Oriented Latent Structures for Dialogue-Based Relation Extraction | Dialogue-based relation extraction (DiaRE) aims to detect the structural information from unstructured utterances in dialogues. Existing relation extraction models may be unsatisfactory under such a conversational setting, due to the entangled logic and information sparsity issues in utterances involving multiple speakers. To this end, we introduce SOLS, a novel model which can explicitly induce speaker-oriented latent structures for better DiaRE. Specifically, we learn latent structures to capture the relationships among tokens beyond the utterance boundaries, alleviating the entangled logic issue. During the learning process, our speaker-specific regularization method progressively highlights speaker-related key clues and erases the irrelevant ones, alleviating the information sparsity issue. Experiments on three public datasets demonstrate the effectiveness of our proposed approach. | ['Wei Lu', 'Yao Xiao', 'Sicong Leng', 'Guoqing Luo', 'Guoshun Nan'] | 2021-09-11 | null | null | null | null | ['dialog-relation-extraction'] | ['natural-language-processing'] | [ 1.51398927e-01 6.55794561e-01 -4.11706328e-01 -4.34461594e-01
-8.85393441e-01 -4.68484998e-01 5.68131089e-01 1.85932711e-01
7.17437416e-02 7.10738301e-01 9.78743672e-01 -2.10119873e-01
-2.04556271e-01 -4.49262559e-01 -2.52907693e-01 -5.60988188e-01
-2.20519766e-01 3.33684593e-01 -2.40881387e-02 -3.56384695e-01
-1.22139938e-01 -1.88544080e-01 -9.38092947e-01 4.60043758e-01
1.07206333e+00 4.90615845e-01 -2.04164892e-01 3.45254183e-01
-4.83018279e-01 1.26990592e+00 -5.45659661e-01 -4.16345984e-01
-1.38584673e-01 -5.49555063e-01 -1.11259556e+00 4.66400892e-01
9.95779261e-02 -1.82947472e-01 -4.18726623e-01 9.80620921e-01
2.04202175e-01 5.99963069e-02 3.85875076e-01 -1.17880595e+00
-2.29437009e-01 1.35115182e+00 -4.76385951e-01 2.02853158e-01
6.25770211e-01 -7.17464760e-02 1.68879068e+00 -9.39095318e-01
4.42869782e-01 1.75328135e+00 2.89489448e-01 5.07090688e-01
-1.21925986e+00 -4.95326251e-01 5.93893826e-01 9.55373272e-02
-1.25469482e+00 -8.25796485e-01 1.31430018e+00 -2.94115394e-01
7.90731788e-01 5.06431580e-01 2.23115385e-01 1.17447400e+00
-2.50376731e-01 1.17421067e+00 8.90393078e-01 -3.75731319e-01
-4.66867164e-02 2.90156066e-01 6.88669682e-01 7.99898982e-01
-1.43920809e-01 -4.65964198e-01 -1.12386274e+00 -3.22002292e-01
4.23872381e-01 -3.35340053e-01 -3.24081123e-01 -1.41569987e-01
-9.00824726e-01 9.42704141e-01 -1.00631237e-01 2.31906012e-01
-3.07265818e-01 -4.87405986e-01 5.78579009e-01 5.24908483e-01
6.80304646e-01 3.32411051e-01 -6.50211990e-01 -9.52625498e-02
-3.39781791e-01 1.23891234e-01 1.08700311e+00 9.03066695e-01
7.81538010e-01 -1.16438836e-01 -3.73284787e-01 1.06803095e+00
5.23657680e-01 -1.04711160e-01 2.13189662e-01 -7.41676986e-01
1.03434920e+00 9.75866437e-01 -5.21250032e-02 -1.12150288e+00
-4.92444217e-01 -2.94104487e-01 -9.46759343e-01 -7.96930611e-01
3.93003613e-01 -5.22898316e-01 -2.76200444e-01 1.80868673e+00
5.30182719e-01 5.42777851e-02 4.25615519e-01 7.56595016e-01
1.17181742e+00 6.29554689e-01 -3.93534154e-02 -7.16289341e-01
1.65835130e+00 -1.01035297e+00 -1.22395957e+00 -4.82016444e-01
7.85019100e-01 -6.28462911e-01 1.24589574e+00 9.88923758e-02
-6.83569312e-01 -7.08052218e-02 -7.95985639e-01 -5.06383181e-02
1.92184389e-01 6.20869882e-02 8.81813824e-01 4.41507041e-01
-4.47114348e-01 -1.22859806e-01 -5.95619559e-01 7.12800473e-02
2.87655115e-01 1.89777225e-01 -2.17796877e-01 2.20855027e-01
-1.54936051e+00 4.28688198e-01 3.68225157e-01 4.20195132e-01
-5.35137534e-01 -3.77791792e-01 -1.17241001e+00 1.98663980e-01
1.20268869e+00 -2.03924879e-01 1.27887177e+00 -4.72168028e-01
-1.83512890e+00 5.36943614e-01 -5.78369856e-01 -3.97583693e-01
1.57359451e-01 -3.52806002e-01 -9.69949737e-02 -7.03937337e-02
7.01286122e-02 6.54205531e-02 7.11951315e-01 -1.14388263e+00
-3.92301232e-01 -1.52287558e-01 4.11942571e-01 4.67168331e-01
-4.60404456e-01 4.25867327e-02 -4.79940504e-01 -4.56360221e-01
4.97220784e-01 -7.24503040e-01 -1.84318587e-01 -7.82224178e-01
-1.00647068e+00 -8.19365144e-01 8.06329787e-01 -4.71411914e-01
1.45397675e+00 -2.10963011e+00 3.27317655e-01 -3.55076119e-02
6.40833139e-01 -1.97661594e-02 -7.09556192e-02 4.70726669e-01
1.52022004e-01 2.83360388e-02 -5.51884994e-02 -5.32402217e-01
3.45445611e-02 2.88393348e-01 -3.60045463e-01 1.40753508e-01
3.57952446e-01 8.42700303e-01 -9.08146739e-01 -7.74671614e-01
-9.41329598e-02 9.13679972e-02 -5.92925310e-01 4.43314701e-01
-4.62277383e-01 6.67803705e-01 -7.56573737e-01 6.00638270e-01
4.53457296e-01 -3.59522223e-01 5.64784944e-01 -1.85689181e-01
-9.52364877e-03 1.09865820e+00 -1.06417418e+00 1.56532586e+00
-4.11852747e-01 5.44404924e-01 4.41388875e-01 -1.07255483e+00
9.10944462e-01 5.29104590e-01 4.38835561e-01 -3.58171791e-01
2.97318008e-02 -1.81793004e-01 1.00068614e-01 -8.33971500e-01
3.30227941e-01 -2.52691150e-01 -3.50351185e-01 3.95792186e-01
1.36830181e-01 1.55829057e-01 -3.29793524e-03 5.51350772e-01
9.20974433e-01 -4.39504862e-01 3.22315246e-01 -1.32406518e-01
9.01706934e-01 -3.42364997e-01 9.54039931e-01 5.58257043e-01
-2.50441760e-01 -1.87194310e-02 1.22678757e+00 -2.42075250e-01
-2.78432906e-01 -6.89210713e-01 4.65365052e-02 1.25348330e+00
1.04119197e-01 -9.62958753e-01 -4.80584949e-01 -1.04931760e+00
-3.14551383e-01 4.61537153e-01 -4.91151214e-01 -1.35040745e-01
-7.02451944e-01 -6.93266988e-01 4.49929297e-01 2.62449831e-02
4.29543287e-01 -8.69345129e-01 1.34924889e-01 3.06751937e-01
-8.22434783e-01 -1.42233825e+00 -5.37788928e-01 2.53298998e-01
-5.42423308e-01 -9.98028815e-01 -7.72145540e-02 -7.89850891e-01
5.74100792e-01 2.33559772e-01 8.94802272e-01 -1.47399336e-01
3.28049660e-01 -6.25534281e-02 -3.35069567e-01 -9.73190814e-02
-5.72885454e-01 3.66822809e-01 5.18226400e-02 2.76656091e-01
6.42596722e-01 -5.00859439e-01 -1.47078216e-01 1.61168262e-01
-4.78943288e-01 2.89505929e-01 3.94244760e-01 1.12057817e+00
-4.29115407e-02 2.98273832e-01 7.50807226e-01 -1.31065571e+00
1.08959663e+00 -4.72484052e-01 -3.80078107e-01 3.71111155e-01
-2.96564311e-01 4.34851229e-01 4.73951668e-01 -4.15608853e-01
-1.45109451e+00 -6.59451261e-02 1.00146294e-01 1.16364308e-01
-9.17308107e-02 8.60673726e-01 -6.96586788e-01 4.07754898e-01
3.13753456e-01 1.69615239e-01 -3.02079588e-01 -5.60997128e-01
3.60738546e-01 7.79197633e-01 2.36640275e-01 -8.07899952e-01
6.49589002e-01 1.13493778e-01 -3.71130526e-01 -9.66185391e-01
-1.37582374e+00 -5.73633552e-01 -6.60743475e-01 -1.03076972e-01
5.23041964e-01 -9.28159416e-01 -8.75565469e-01 2.30871722e-01
-1.39132178e+00 -7.99078271e-02 -1.28606528e-01 5.08224189e-01
-1.96899876e-01 5.88735402e-01 -9.07298028e-01 -1.20698011e+00
-2.16306642e-01 -1.05331802e+00 8.33679020e-01 1.90523595e-01
-5.48091054e-01 -9.58670914e-01 8.49725083e-02 7.74479091e-01
-1.00473762e-01 -1.12324812e-01 1.11623573e+00 -8.69591832e-01
-6.10244155e-01 1.53653687e-02 -2.24484935e-01 -8.15816075e-02
5.90675175e-01 -3.35585594e-01 -1.02550101e+00 9.21903551e-02
2.52525300e-01 -5.20773649e-01 7.04900205e-01 -9.11745727e-02
6.50476933e-01 -9.81649637e-01 -1.32890165e-01 2.60146409e-02
5.89652717e-01 2.74655521e-02 -3.04797012e-03 -8.77465606e-02
9.28238630e-01 1.27360547e+00 4.84869093e-01 5.70268273e-01
8.15841496e-01 6.47879064e-01 -3.19718458e-02 7.03110779e-03
2.24792078e-01 -4.38204646e-01 4.41042155e-01 1.48447037e+00
3.95231009e-01 -1.58432633e-01 -8.03713381e-01 4.91623849e-01
-2.08055639e+00 -5.59024632e-01 -1.82917163e-01 1.71672940e+00
1.56844640e+00 4.71904427e-01 5.80473021e-02 -5.51158702e-03
5.71201026e-01 5.66503346e-01 -4.30868894e-01 -2.31105104e-01
-1.94929034e-01 -3.48075300e-01 -1.66126028e-01 9.39900458e-01
-1.17654049e+00 1.21556783e+00 5.59250498e+00 6.80656433e-01
-6.93847895e-01 1.08114347e-01 4.71690387e-01 8.16960409e-02
-5.29396057e-01 2.70294279e-01 -9.80825484e-01 2.12905079e-01
7.10871041e-01 -2.05701619e-01 5.24565727e-02 6.45725489e-01
4.00432646e-01 2.68131178e-02 -1.08399141e+00 8.01919162e-01
2.67316960e-02 -1.19763553e+00 6.21510074e-02 3.68914939e-03
4.30546850e-01 -4.00295824e-01 -2.98368335e-01 4.40159470e-01
5.43526471e-01 -7.65593350e-01 3.05268347e-01 1.53034583e-01
3.13655049e-01 -5.53547919e-01 7.52130806e-01 6.28417552e-01
-1.12988675e+00 9.07558054e-02 -2.15385277e-02 -2.44257614e-01
1.54523462e-01 5.51736951e-01 -1.10103834e+00 4.68965739e-01
2.03353330e-01 6.00520849e-01 -4.07372236e-01 2.87291378e-01
-5.12008965e-01 1.06561184e+00 -2.29895085e-01 -1.61020219e-01
3.20703000e-01 -3.39322448e-01 8.96946788e-01 1.25839853e+00
-5.60071111e-01 3.45063925e-01 4.59425718e-01 7.86812007e-01
-2.59133756e-01 3.22799295e-01 -5.84424973e-01 -1.27478972e-01
5.97671509e-01 1.20378959e+00 -4.45463419e-01 -1.90266445e-01
-5.26812136e-01 7.31619537e-01 5.26074529e-01 4.79331523e-01
-3.53357822e-01 -1.79921567e-01 5.24699152e-01 -2.00416833e-01
-5.78563698e-02 -2.61393338e-01 -2.67345130e-01 -1.48256147e+00
2.27930456e-01 -1.23670280e+00 4.66934294e-01 7.19524249e-02
-1.21781027e+00 4.62611020e-01 7.37650506e-03 -8.66535425e-01
-2.23704487e-01 4.00896855e-02 -5.87012768e-01 6.19203031e-01
-1.33373570e+00 -1.07729006e+00 1.27867818e-01 5.18473327e-01
9.05781448e-01 -1.41471714e-01 7.64605701e-01 1.22408859e-01
-1.15993118e+00 7.54399657e-01 -4.54014152e-01 4.12835360e-01
6.13467515e-01 -1.20745814e+00 7.04118237e-02 8.37979257e-01
4.14297432e-01 9.43523705e-01 7.75613844e-01 -5.65109968e-01
-1.33605170e+00 -7.00023711e-01 1.29072607e+00 -3.84871840e-01
9.35053110e-01 -9.14205611e-01 -1.31073058e+00 6.25108123e-01
2.54262418e-01 -5.06850302e-01 1.03761435e+00 9.04830813e-01
-5.04571795e-01 -4.53190319e-02 -5.59206843e-01 7.31991112e-01
8.39632571e-01 -8.80093932e-01 -9.56308603e-01 5.12877285e-01
1.15005958e+00 -3.95842969e-01 -6.19782686e-01 4.77427393e-01
2.14939043e-01 -5.08398354e-01 7.31046438e-01 -7.06560373e-01
3.09128106e-01 6.93039894e-02 1.45356044e-01 -9.72457588e-01
-6.39902800e-03 -1.32978356e+00 -5.67294300e-01 1.76548553e+00
6.22244000e-01 -3.91025603e-01 7.84987092e-01 7.88121998e-01
1.42497614e-01 -6.69279575e-01 -1.03224361e+00 -3.69110852e-01
-2.21063823e-01 -3.31933320e-01 4.57768887e-01 1.21083581e+00
6.90009475e-01 1.26728225e+00 -7.88660407e-01 3.17072332e-01
5.29953837e-01 2.70907462e-01 8.52868259e-01 -1.21825826e+00
-3.49638909e-01 -2.39026159e-01 9.53986719e-02 -1.56305826e+00
5.76735616e-01 -4.27393436e-01 3.10635537e-01 -1.12899697e+00
2.14838147e-01 -5.53947985e-01 -4.82968353e-02 5.47311723e-01
-5.90577006e-01 -4.58519042e-01 -1.38486981e-01 2.68327832e-01
-9.88278449e-01 8.87134075e-01 1.16723466e+00 -3.33439112e-01
-6.50729477e-01 1.92830622e-01 -9.45554912e-01 9.58624244e-01
7.78878391e-01 -5.52909613e-01 -6.57024980e-01 -1.47484794e-01
1.95681110e-01 2.09812105e-01 -9.92758851e-03 -2.33309329e-01
3.89987826e-01 -3.90493244e-01 -3.98193121e-01 -6.53375626e-01
3.67689788e-01 -3.87917668e-01 -5.24657726e-01 1.56828046e-01
-7.99349308e-01 -6.33949518e-01 -7.94257447e-02 6.90224767e-01
-4.59028542e-01 -2.41576694e-02 1.90934196e-01 3.74221765e-02
-2.94703811e-01 8.67639855e-02 -5.66154897e-01 4.01964009e-01
4.39706653e-01 4.23060656e-01 -1.43917769e-01 -5.96053600e-01
-6.58977151e-01 6.12761736e-01 -3.55733901e-01 5.36608934e-01
4.84635711e-01 -1.14701045e+00 -9.43588078e-01 2.70670623e-01
1.84375942e-01 3.87783617e-01 2.58423477e-01 8.92009616e-01
2.70110220e-01 6.05741560e-01 5.58981299e-01 -4.33202684e-01
-1.53124046e+00 2.64129341e-01 5.90278283e-02 -6.60776913e-01
-7.01619446e-01 1.14377999e+00 4.54296380e-01 -5.46961010e-01
5.55012107e-01 -3.93200308e-01 -6.60923183e-01 4.40311193e-01
2.86626607e-01 -1.97168370e-03 -2.01305747e-01 -6.42525852e-01
-4.52837706e-01 2.07831264e-02 -5.50856233e-01 -5.63425981e-02
1.18057072e+00 -6.41253531e-01 -3.59553576e-01 7.77736187e-01
8.99657607e-01 4.76856858e-01 -1.26102054e+00 -9.85955358e-01
5.16700983e-01 -1.53867081e-01 -5.10304123e-02 -3.24353188e-01
-6.94678009e-01 6.43979192e-01 -2.88262486e-01 5.54839790e-01
7.00836658e-01 2.88439900e-01 9.59489405e-01 6.57888949e-01
6.97391257e-02 -1.04594421e+00 3.01797688e-01 7.72027135e-01
8.48487198e-01 -1.40890849e+00 3.97112966e-03 -1.06086349e+00
-8.51926863e-01 8.56309831e-01 7.14289308e-01 4.20731664e-01
6.21387124e-01 3.20864737e-01 1.73211470e-01 -4.34723258e-01
-1.13914132e+00 -1.14826277e-01 2.94903040e-01 1.59859762e-01
6.45976126e-01 1.31823659e-01 -3.43059212e-01 9.80729282e-01
-2.38977090e-01 -5.55381417e-01 3.82648259e-01 7.70259619e-01
-2.05514595e-01 -1.21880269e+00 -1.49409235e-01 1.64428294e-01
-5.01891911e-01 -2.10665196e-01 -8.68029654e-01 2.03749254e-01
-1.70106098e-01 1.32973289e+00 -3.93490165e-01 -2.82884985e-01
2.86030442e-01 2.81235218e-01 -7.48269558e-02 -8.84069085e-01
-4.71407264e-01 4.98911321e-01 6.68766141e-01 -3.86244714e-01
-5.23532152e-01 -6.19139791e-01 -1.16767085e+00 -1.07278889e-02
-6.79451466e-01 6.56048834e-01 2.88671907e-02 1.26115632e+00
1.99005991e-01 5.45374572e-01 9.95137274e-01 -1.29474178e-01
-5.55485904e-01 -1.23007369e+00 -3.41135889e-01 2.04654872e-01
6.69805467e-01 -6.27279937e-01 -4.74434763e-01 -3.07557415e-02] | [12.528427124023438, 7.928312301635742] |
888fd3de-c4fc-4b1b-8498-1bec2da16336 | case-based-reasoning-for-rare-events | 2202.04891 | null | https://arxiv.org/abs/2202.04891v1 | https://arxiv.org/pdf/2202.04891v1.pdf | Case-based reasoning for rare events prediction on strategic sites | Satellite imagery is now widely used in the defense sector for monitoring locations of interest. Although the increasing amount of data enables pattern identification and therefore prediction, carrying this task manually is hardly feasible. We hereby propose a cased-based reasoning approach for automatic prediction of rare events on strategic sites. This method allows direct incorporation of expert knowledge, and is adapted to irregular time series and small-size datasets. Experiments are carried out on two use-cases using real satellite images: the prediction of submarines arrivals and departures from a naval base, and the forecasting of imminent rocket launches on two space bases. The proposed method significantly outperforms a random selection of reference cases on these challenging applications, showing its strong potential. | ['Tugdual Ceillier', 'Marie-Caroline Corbineau', 'Vincent Vidal'] | 2022-02-10 | null | null | null | null | ['irregular-time-series'] | ['time-series'] | [ 4.04111028e-01 -5.04287444e-02 -2.49842405e-01 -3.26360703e-01
-6.79828167e-01 -4.98776853e-01 1.17556024e+00 1.96362749e-01
-4.99569803e-01 9.40576673e-01 -1.58904269e-01 -8.50457191e-01
-6.58105671e-01 -1.04524529e+00 -2.23186105e-01 -7.74028361e-01
-7.96869576e-01 7.12278306e-01 4.20449615e-01 -5.13471663e-01
3.56116742e-01 1.12315464e+00 -1.64212132e+00 2.42314503e-01
6.14644408e-01 1.15585434e+00 1.92782700e-01 2.57127106e-01
1.71675205e-01 8.53082836e-01 -5.61377883e-01 -7.76619986e-02
5.44492304e-01 9.10626501e-02 -5.46317518e-01 3.40806961e-01
-1.13057762e-01 6.09361678e-02 -4.90500145e-02 7.33732522e-01
9.54314545e-02 4.42471296e-01 4.73500311e-01 -1.03469574e+00
3.34020555e-01 2.78067533e-02 -3.85614246e-01 7.81289160e-01
3.85083646e-01 1.19248964e-01 9.61767316e-01 -3.94223690e-01
7.36991763e-01 6.30384266e-01 7.61711717e-01 -3.01972061e-01
-8.63696814e-01 -2.11990401e-01 9.01825801e-02 5.03052533e-01
-1.58994699e+00 -5.73087297e-02 5.62484443e-01 -3.64789754e-01
1.12569129e+00 6.36021912e-01 6.49811029e-01 5.33645630e-01
2.34334439e-01 5.33627033e-01 1.33660126e+00 -4.31012511e-01
4.99717444e-01 -1.12024829e-01 -2.25921005e-01 5.58084510e-02
-2.68994775e-02 4.81631637e-01 -8.78357589e-02 -3.04892540e-01
5.48933506e-01 2.60560989e-01 -4.15554792e-01 4.04340565e-01
-1.26115429e+00 6.14703655e-01 1.66933566e-01 7.84547150e-01
-1.11267817e+00 -4.38231081e-01 1.47378102e-01 4.55744833e-01
5.25091410e-01 2.56113589e-01 -3.53768349e-01 -2.55956382e-01
-1.43346727e+00 4.49177176e-01 7.92992115e-01 6.65658236e-01
3.58614177e-01 1.97156146e-01 4.44059193e-01 2.61319637e-01
-3.06353606e-02 5.64352155e-01 3.07589680e-01 -3.34193915e-01
4.27619427e-01 5.30077159e-01 5.69334626e-01 -1.25002491e+00
-6.95677042e-01 -6.75534427e-01 -6.78751707e-01 2.19769046e-01
2.28445634e-01 -1.05390742e-01 -8.14186811e-01 8.13240826e-01
1.33172691e-01 4.69777316e-01 1.02037407e-01 9.45781052e-01
-1.56587083e-03 9.94643748e-01 -1.33682899e-02 -6.45554960e-01
1.00074899e+00 -1.19468972e-01 -5.63944638e-01 -3.08514953e-01
3.92080545e-01 -6.18445873e-01 3.11172068e-01 8.53161752e-01
-5.66856563e-01 -2.58314729e-01 -6.65899754e-01 9.37820554e-01
-3.48614573e-01 2.55079307e-02 6.66149378e-01 1.89350069e-01
-6.69522762e-01 7.58216381e-01 -9.07852471e-01 -3.90385866e-01
8.53988305e-02 4.12451774e-01 -1.83618337e-01 1.39438227e-01
-1.24162054e+00 9.81911778e-01 7.31964171e-01 4.88743991e-01
-6.94098830e-01 -2.41881907e-01 -5.40853024e-01 -1.07047409e-02
3.55599493e-01 8.14095810e-02 8.96396458e-01 -8.22978973e-01
-8.42757344e-01 7.53843665e-01 2.34828353e-01 -1.09072626e+00
4.48377401e-01 2.87646770e-01 -1.30658519e+00 -6.05581366e-02
-1.21470407e-01 -1.52977437e-01 7.81680763e-01 -9.58233595e-01
-1.22828555e+00 -3.32467943e-01 1.19767167e-01 1.58322677e-02
-1.15017466e-01 4.34365422e-02 1.38239071e-01 -7.44280159e-01
3.00996125e-01 -1.00351322e+00 -5.97361982e-01 -7.64028847e-01
1.86563861e-02 -5.66757731e-02 7.45024264e-01 -9.21724498e-01
1.47931838e+00 -1.97917891e+00 -8.68438557e-02 7.50156820e-01
-7.42497146e-01 2.82347113e-01 2.14131117e-01 9.85205591e-01
-3.22441876e-01 -1.21013813e-01 -3.42938542e-01 4.80519772e-01
-1.16216987e-01 4.73923236e-01 -1.01125741e+00 4.42146868e-01
1.19840339e-01 1.93131089e-01 -7.09870219e-01 -2.45470896e-01
2.33657524e-01 -4.98828702e-02 8.30558091e-02 -7.24565312e-02
-2.93393701e-01 5.26007116e-01 -6.36969745e-01 1.00870216e+00
4.93450761e-01 6.96173012e-02 5.44368327e-01 3.29596438e-02
-5.88797867e-01 1.15523554e-01 -1.29046905e+00 1.13734555e+00
-5.58425725e-01 4.48949486e-01 -1.87228486e-01 -1.07591677e+00
1.18325555e+00 3.93346518e-01 4.42119300e-01 -7.06021130e-01
-7.25101903e-02 3.47892255e-01 -1.59527734e-02 -5.97345948e-01
6.98955178e-01 -8.68817151e-01 -1.82942808e-01 4.34588939e-01
-5.00743926e-01 -2.84411430e-01 3.82581651e-01 -2.49014020e-01
1.11910164e+00 3.52545157e-02 7.04262376e-01 -3.04415107e-01
6.47328317e-01 6.17432952e-01 7.61134982e-01 3.61208171e-01
2.41326764e-02 1.34051785e-01 2.61866719e-01 -1.23639035e+00
-7.78147280e-01 -4.32302773e-01 -3.49434614e-01 5.76952577e-01
6.25358373e-02 -1.96483791e-01 -3.21740359e-02 -7.57641971e-01
-1.23563118e-01 1.34466028e+00 -4.02962029e-01 5.24138272e-01
-4.44129169e-01 -1.14570773e+00 7.83814043e-02 6.42717183e-02
4.51953225e-02 -9.37080860e-01 -7.32099414e-01 3.92228872e-01
-2.74860382e-01 -1.01573002e+00 5.13075233e-01 2.65795905e-02
-1.29166961e+00 -9.81626034e-01 -2.58613884e-01 -2.51026183e-01
5.56186199e-01 2.00781614e-01 9.55268264e-01 4.68963683e-02
-2.32399553e-01 3.47763479e-01 -4.16327566e-01 -6.13485515e-01
-4.29071784e-01 -1.87770367e-01 2.58630753e-01 3.09924304e-01
3.78776073e-01 -7.11782575e-01 -3.26244354e-01 7.64579594e-01
-1.27480185e+00 -2.64538705e-01 6.35376513e-01 5.76267242e-01
6.95746779e-01 8.54558706e-01 4.38323915e-01 -4.87689793e-01
2.34307885e-01 -7.63764024e-01 -9.67247248e-01 4.49410796e-01
-3.15348476e-01 -2.37678736e-01 5.15688777e-01 -6.28952608e-02
-1.09464264e+00 1.06571607e-01 7.97932409e-03 -8.16089362e-02
-6.34175897e-01 1.12168336e+00 3.45247090e-01 -3.66051011e-02
6.57910407e-01 5.19069254e-01 -2.75257736e-01 -4.97322023e-01
-4.19583678e-01 7.98396051e-01 3.96890134e-01 -1.97624624e-01
1.04366946e+00 7.47050822e-01 1.45890549e-01 -1.09334755e+00
-4.48941767e-01 -5.20540476e-01 -5.77346206e-01 -5.96877754e-01
1.54605061e-01 -9.26928043e-01 -3.45176697e-01 2.72829030e-02
-8.22995961e-01 -1.49626449e-01 -2.74017245e-01 6.14057302e-01
-4.69958544e-01 4.20205116e-01 1.77706137e-01 -1.14847279e+00
1.05506875e-01 -6.14693105e-01 8.67139459e-01 -9.19227228e-02
-1.90820217e-01 -1.18500650e+00 1.04719959e-01 2.73247600e-01
4.19108391e-01 4.65790093e-01 5.18010318e-01 -1.02027774e+00
-6.89173579e-01 -8.78625751e-01 2.58210838e-01 6.27806038e-02
8.40020701e-02 -1.62581012e-01 -5.30650675e-01 -2.97214746e-01
2.35669538e-01 1.72289222e-01 4.51375544e-01 1.70717597e-01
8.72836351e-01 -2.14892596e-01 -4.61861223e-01 4.67274599e-02
1.58770263e+00 5.74503005e-01 9.37531769e-01 7.58554280e-01
-2.78409511e-01 7.82692969e-01 1.46360397e+00 1.02553427e+00
-2.54338104e-02 9.20672596e-01 8.82414937e-01 3.13721150e-01
6.90912545e-01 4.44923669e-01 2.26235479e-01 5.89467704e-01
-1.08152008e+00 -2.04581186e-01 -1.48758459e+00 8.71748626e-01
-1.81280148e+00 -1.43192375e+00 -1.63133979e-01 2.56781578e+00
3.05858254e-01 4.37542140e-01 -1.28461182e-01 2.89419204e-01
5.90574741e-01 2.68561780e-01 -6.04130737e-02 5.69519363e-02
-1.18598066e-01 4.22210507e-02 7.91170418e-01 4.24333721e-01
-1.34300888e+00 5.26797652e-01 5.88304853e+00 1.12363732e+00
-1.21192133e+00 -4.16835397e-02 3.75744402e-01 -1.14116576e-02
-2.91299522e-01 2.15630531e-01 -8.32258761e-01 3.55620295e-01
1.12505364e+00 8.10998678e-03 1.11106768e-01 7.55147934e-01
4.92622018e-01 -2.41762608e-01 -3.76900196e-01 7.05259383e-01
-1.56967103e-01 -1.72873688e+00 -1.42799929e-01 1.47017077e-01
5.39766848e-01 1.15956053e-01 -3.92424911e-01 8.49401802e-02
-1.75003201e-01 -6.10504150e-01 6.14176929e-01 9.96135294e-01
2.77045995e-01 -7.30594754e-01 9.09914196e-01 7.35926330e-01
-8.28810394e-01 -4.24889326e-01 1.96143854e-02 -2.89784640e-01
5.62739253e-01 6.91928446e-01 -1.15983474e+00 1.02593660e+00
4.44591910e-01 6.51887417e-01 2.40688790e-02 1.17041218e+00
-1.25069708e-01 9.12558019e-01 -6.27596200e-01 2.03871801e-01
6.22002065e-01 -2.71821707e-01 9.47068691e-01 9.65261221e-01
7.19935000e-01 3.60962123e-01 1.34766072e-01 1.29098251e-01
8.75070095e-01 -1.09285213e-01 -7.70893514e-01 1.08512893e-01
1.43629715e-01 1.12758327e+00 -8.59699786e-01 -1.55247614e-01
-3.21146280e-01 4.95154202e-01 -4.29710776e-01 2.64276206e-01
-9.40626085e-01 -2.47377619e-01 3.30159813e-01 4.66810226e-01
6.25245988e-01 -6.06455147e-01 7.13278949e-02 -8.99115443e-01
-3.46930921e-02 -9.21417058e-01 5.87452590e-01 -5.37377298e-01
-1.03492701e+00 1.07545173e+00 9.84312072e-02 -2.20830154e+00
-6.47661269e-01 -5.21348596e-01 -9.47262287e-01 6.46225095e-01
-1.68513715e+00 -1.05609310e+00 4.35995273e-02 3.91770899e-01
5.48293889e-01 -3.90524775e-01 8.37394178e-01 2.20700011e-01
-3.30447733e-01 -4.22940403e-01 2.16915026e-01 -3.96603346e-01
1.78518847e-01 -8.05118680e-01 3.10152441e-01 1.07761228e+00
1.36691749e-01 2.85455763e-01 1.10457468e+00 -9.30997670e-01
-1.37537050e+00 -1.06218553e+00 1.31659079e+00 3.42449322e-02
1.20764399e+00 2.47668207e-01 -8.12862575e-01 7.10962951e-01
4.34289612e-02 -9.23911408e-02 7.40691006e-01 -3.23563851e-02
3.73202026e-01 -2.53970563e-01 -9.84398484e-01 4.24171209e-01
5.16567290e-01 -1.04444996e-01 -7.98069596e-01 1.00334489e+00
-1.61088154e-01 -9.19519439e-02 -1.26500607e+00 5.03877580e-01
1.22475177e-01 -8.81424487e-01 8.50327969e-01 -6.93685830e-01
5.16494811e-02 -6.96120918e-01 -3.17729443e-01 -1.22294486e+00
-1.40387580e-01 -7.98225462e-01 3.23639214e-01 7.63281941e-01
4.85536426e-01 -5.81885040e-01 6.00115716e-01 5.38523078e-01
-1.61484763e-01 -2.78907627e-01 -1.50285959e+00 -1.09140563e+00
-7.92227268e-01 -8.51860702e-01 6.02777481e-01 1.09662926e+00
-1.87008362e-02 -3.11563730e-01 -6.00055993e-01 7.09007442e-01
5.75666547e-01 6.43189609e-01 5.63547671e-01 -1.49681282e+00
-3.42986494e-01 -1.33657500e-01 -9.84133780e-01 -4.88035291e-01
-1.50928259e-01 -6.32452965e-01 -2.28712529e-01 -1.19178033e+00
-6.71122253e-01 -7.04873919e-01 -3.21347356e-01 6.15620077e-01
4.41388667e-01 2.74305642e-01 -8.31857920e-02 5.85813105e-01
-3.93090695e-01 3.76612395e-01 6.34861231e-01 4.93104458e-02
1.81127638e-02 8.20119619e-01 7.61497393e-02 9.61258352e-01
7.23985195e-01 -3.91558439e-01 4.35908101e-02 1.30555958e-01
6.86918199e-01 6.40778244e-01 4.58935648e-01 -1.23311794e+00
2.90071964e-01 -6.46066785e-01 -1.02326751e-01 -7.27313876e-01
1.86263904e-01 -1.37026370e+00 8.51724446e-01 5.84479809e-01
1.60025433e-01 -1.68755297e-02 4.14788306e-01 7.99953461e-01
-8.68801057e-01 -4.82724458e-01 3.14765126e-01 1.01652525e-01
-1.36654854e+00 1.59997493e-01 -7.81584263e-01 -5.44341683e-01
1.38537157e+00 -3.65554273e-01 3.13632982e-03 -4.62352306e-01
-1.20625114e+00 8.37927461e-02 1.34750813e-01 2.32053459e-01
4.23125535e-01 -1.16019940e+00 -8.23191226e-01 1.72552019e-01
3.97438966e-02 -3.95678848e-01 5.19234896e-01 1.22943449e+00
-6.99962318e-01 5.03021657e-01 -2.28076488e-01 -4.63729948e-01
-1.09614575e+00 3.07849854e-01 1.33925870e-01 -5.88468552e-01
-4.69388634e-01 1.65123865e-01 -5.12601793e-01 -2.63952911e-01
-1.30100712e-01 -3.15608352e-01 -4.73656178e-01 3.35081428e-01
7.75074840e-01 3.43621820e-01 4.05295640e-01 -6.70992613e-01
-4.49318707e-01 1.56836867e-01 2.55470686e-02 1.17498497e-02
1.86326635e+00 -3.03268749e-02 -1.71085641e-01 3.86091620e-01
2.68581688e-01 -8.22919607e-02 -1.09767509e+00 -2.80437171e-01
6.09127879e-01 -6.18043005e-01 2.11938068e-01 -9.14540291e-01
-8.46412778e-01 6.16257966e-01 3.47941935e-01 8.42301786e-01
1.44483018e+00 -3.77398193e-01 5.68164110e-01 6.02341235e-01
7.77053833e-01 -1.02310073e+00 -9.04572070e-01 4.67684627e-01
1.17638838e+00 -1.13856244e+00 3.99418890e-01 -1.96786433e-01
-9.61735308e-01 1.40458882e+00 -2.00759083e-01 1.14599774e-02
7.82836735e-01 9.45906118e-02 -1.30712464e-01 -2.62171537e-01
-8.15636694e-01 -3.04129064e-01 3.58867109e-01 3.07358056e-01
-4.18836810e-02 3.02318543e-01 -4.32485044e-01 4.41929460e-01
2.74376661e-01 1.63067222e-01 5.41831374e-01 1.38813937e+00
-6.53875411e-01 -1.02526128e+00 -7.67397165e-01 4.30651426e-01
-4.46310908e-01 -2.04447955e-01 2.19563413e-02 1.12531781e+00
1.50354058e-01 6.04791164e-01 2.48242021e-01 -3.95261228e-01
4.25030261e-01 -2.78242920e-02 -4.02810574e-02 -1.33822560e-01
-6.37447000e-01 5.31761386e-02 5.23084342e-01 -4.44203705e-01
-8.65815282e-01 -1.33227575e+00 -1.06419921e+00 8.14807042e-02
-2.39033788e-01 4.68480229e-01 9.39312875e-01 1.13436854e+00
3.36727530e-01 1.37609810e-01 8.93803835e-01 -9.72253084e-01
-6.28937900e-01 -8.62665892e-01 -1.03398824e+00 2.68258512e-01
-2.97442488e-02 -5.76692343e-01 -4.18033481e-01 1.07222684e-01] | [7.129107475280762, 3.2068912982940674] |
6740bb61-bf51-4151-bdfa-291159c49b83 | a-taxonomy-for-in-depth-evaluation-of | null | null | https://aclanthology.org/L18-1109 | https://aclanthology.org/L18-1109.pdf | A Taxonomy for In-depth Evaluation of Normalization for User Generated Content | null | ['Gertjan van Noord', 'Rob van der Goot', 'Rik van Noord'] | 2018-05-01 | a-taxonomy-for-in-depth-evaluation-of-1 | https://aclanthology.org/L18-1109 | https://aclanthology.org/L18-1109.pdf | lrec-2018-5 | ['lexical-normalization'] | ['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] | [-7.280041217803955, 3.758434772491455] |
68e0faa0-d81e-44c7-80c8-0381a28c7736 | cascaded-partial-decoder-for-fast-and | 1904.08739 | null | http://arxiv.org/abs/1904.08739v1 | http://arxiv.org/pdf/1904.08739v1.pdf | Cascaded Partial Decoder for Fast and Accurate Salient Object Detection | Existing state-of-the-art salient object detection networks rely on
aggregating multi-level features of pre-trained convolutional neural networks
(CNNs). Compared to high-level features, low-level features contribute less to
performance but cost more computations because of their larger spatial
resolutions. In this paper, we propose a novel Cascaded Partial Decoder (CPD)
framework for fast and accurate salient object detection. On the one hand, the
framework constructs partial decoder which discards larger resolution features
of shallower layers for acceleration. On the other hand, we observe that
integrating features of deeper layers obtain relatively precise saliency map.
Therefore we directly utilize generated saliency map to refine the features of
backbone network. This strategy efficiently suppresses distractors in the
features and significantly improves their representation ability. Experiments
conducted on five benchmark datasets exhibit that the proposed model not only
achieves state-of-the-art performance but also runs much faster than existing
models. Besides, the proposed framework is further applied to improve existing
multi-level feature aggregation models and significantly improve their
efficiency and accuracy. | ['Zhe Wu', 'Qingming Huang', 'Li Su'] | 2019-04-18 | cascaded-partial-decoder-for-fast-and-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Wu_Cascaded_Partial_Decoder_for_Fast_and_Accurate_Salient_Object_Detection_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Wu_Cascaded_Partial_Decoder_for_Fast_and_Accurate_Salient_Object_Detection_CVPR_2019_paper.pdf | cvpr-2019-6 | ['camouflaged-object-segmentation'] | ['computer-vision'] | [ 9.94693339e-02 -9.00187865e-02 -2.73256212e-01 -2.43460074e-01
-4.96198058e-01 2.62206942e-02 3.00543576e-01 3.97339940e-01
-4.80482787e-01 5.04143238e-01 3.33362758e-01 1.36168376e-01
1.66812509e-01 -7.98615873e-01 -7.95325220e-01 -5.80147922e-01
1.12305559e-01 -4.33752716e-01 1.10316920e+00 -2.32985377e-01
3.94965619e-01 2.95262605e-01 -1.87535512e+00 3.20370376e-01
9.39109445e-01 1.25032926e+00 8.02669346e-01 1.06190547e-01
-2.52085887e-02 9.45532978e-01 -3.21834385e-01 -1.96556635e-02
6.04979582e-02 -1.81508601e-01 -4.84185517e-01 -1.40196055e-01
3.21135908e-01 -6.27564788e-01 -3.82110387e-01 1.37629259e+00
3.70949000e-01 9.92801599e-03 1.90549582e-01 -1.07915437e+00
-9.12766337e-01 6.52504444e-01 -9.74964738e-01 5.76564729e-01
-1.16044633e-01 1.83165949e-02 1.11741292e+00 -1.16342962e+00
3.22237045e-01 1.30251253e+00 3.38814735e-01 2.76512861e-01
-7.40378141e-01 -6.83152974e-01 6.12357676e-01 3.81418914e-01
-1.36219418e+00 -2.87351519e-01 9.72692072e-01 3.61018814e-02
7.72826314e-01 5.88926785e-02 6.69257998e-01 4.71221447e-01
-2.90600937e-02 1.32405651e+00 6.01097345e-01 -1.53437957e-01
1.16264269e-01 1.13094166e-01 1.11129776e-01 1.09538889e+00
5.65526783e-01 -2.00775117e-01 -6.78906620e-01 1.26746118e-01
1.04493213e+00 3.95482570e-01 -3.71163845e-01 -3.50292742e-01
-1.34009182e+00 7.37541199e-01 1.12557697e+00 3.75342816e-01
-7.18505085e-01 1.99810177e-01 3.22239041e-01 -2.42238536e-01
2.29444683e-01 1.75195858e-01 -2.68250376e-01 2.08152220e-01
-9.78418112e-01 1.03271998e-01 2.10843645e-02 1.02284420e+00
1.04414415e+00 2.72030562e-01 -5.56694090e-01 6.91455424e-01
2.54630208e-01 1.89459831e-01 4.67671692e-01 -6.00228012e-01
5.41212738e-01 9.61026907e-01 1.75590903e-01 -1.35146594e+00
-5.16788244e-01 -1.03671122e+00 -7.56513953e-01 8.22321698e-02
-5.57399020e-02 2.66448203e-02 -9.11630690e-01 1.72499871e+00
2.83202469e-01 3.47826451e-01 6.95052743e-02 1.28815448e+00
9.74070132e-01 7.78360426e-01 1.91145316e-01 -4.79987264e-02
1.40537155e+00 -1.40514672e+00 -4.87230718e-01 -3.40546548e-01
4.07121390e-01 -6.97260797e-01 9.98092949e-01 -3.39265242e-02
-1.25303841e+00 -8.14769089e-01 -1.31186199e+00 -3.96722227e-01
-2.03525215e-01 6.49787605e-01 7.72466779e-01 8.06467757e-02
-1.04624414e+00 4.39894289e-01 -8.35751891e-01 -5.90633526e-02
8.74119401e-01 4.70473588e-01 1.51102260e-01 1.58503294e-01
-1.05457067e+00 6.69960499e-01 4.19024736e-01 1.02276273e-01
-9.35687780e-01 -7.98034668e-01 -9.26937342e-01 5.22123754e-01
3.76623243e-01 -4.85670477e-01 1.24079311e+00 -1.03923059e+00
-1.01426721e+00 2.93313265e-01 -4.31373358e-01 -3.78278345e-01
8.34518075e-02 -4.62448448e-01 -1.78243294e-01 3.31325054e-01
2.36889079e-01 1.09459853e+00 8.88712585e-01 -9.97835457e-01
-1.26227033e+00 -7.34121054e-02 1.38125747e-01 2.93844968e-01
-7.57710636e-01 1.14931293e-01 -5.39304018e-01 -6.07620537e-01
3.13049108e-01 -4.13607389e-01 -2.92313516e-01 3.53015006e-01
-3.48404109e-01 -2.60420352e-01 9.81270850e-01 -3.54829311e-01
1.19954479e+00 -2.24990249e+00 -7.94193968e-02 -1.92692369e-01
5.25901496e-01 4.72876728e-01 -1.49822429e-01 -7.07544684e-02
1.45454541e-01 -1.93163976e-01 -5.84632605e-02 -2.19636470e-01
-2.33144522e-01 -1.39814332e-01 -3.38609636e-01 2.69356996e-01
6.25171602e-01 1.03343534e+00 -9.86018181e-01 -7.25822389e-01
3.55485976e-01 4.71509993e-01 -5.69991231e-01 -2.27103442e-01
-1.44667048e-02 -4.11527278e-03 -7.72646308e-01 6.31976843e-01
9.03315604e-01 -5.25045037e-01 -4.08290386e-01 -4.73180354e-01
-4.34438199e-01 4.14494306e-01 -1.08904111e+00 1.69287241e+00
-2.29311198e-01 5.28198659e-01 -4.36977223e-02 -9.50379968e-01
1.02760363e+00 -1.25749096e-01 7.79919326e-02 -8.11617851e-01
3.50657344e-01 2.59339690e-01 1.22121565e-01 -3.44480611e-02
8.25950682e-01 3.27847272e-01 1.32068798e-01 4.36496586e-02
1.19802438e-01 6.60396993e-01 1.89647190e-02 3.06164294e-01
8.67116630e-01 -7.00048879e-02 3.34658325e-01 -5.13094366e-01
7.96346307e-01 -1.20482653e-01 8.39598060e-01 5.90260923e-01
-4.02458698e-01 5.51737666e-01 3.77734184e-01 -4.90519196e-01
-7.63998032e-01 -9.34883714e-01 1.24562934e-01 1.29172897e+00
8.08370233e-01 -2.79957950e-01 -6.71302438e-01 -4.41944242e-01
-1.04395449e-01 3.92688960e-01 -6.35181904e-01 -3.21479023e-01
-5.74998140e-01 -7.29312778e-01 2.16078922e-01 8.47042501e-01
9.05077040e-01 -1.17941225e+00 -1.04602146e+00 4.37222272e-01
-1.97136387e-01 -1.06700575e+00 -3.52245927e-01 -2.24077627e-02
-1.01063347e+00 -7.87449360e-01 -9.07064021e-01 -1.27467346e+00
7.84849703e-01 9.37701583e-01 8.35435033e-01 3.62307817e-01
-2.02525973e-01 -4.51180369e-01 -4.73176748e-01 -5.29631913e-01
3.06440204e-01 2.96985239e-01 -1.47727549e-01 4.10319418e-02
4.31498319e-01 -5.01627743e-01 -1.00743496e+00 2.10487977e-01
-7.82665849e-01 3.74410421e-01 1.02159023e+00 6.15460634e-01
7.24400163e-01 -2.68909428e-03 9.13999319e-01 -3.85413706e-01
3.08451712e-01 -3.37531507e-01 -6.70101821e-01 1.15505770e-01
-2.04138741e-01 1.62828475e-01 7.97374606e-01 -2.35698670e-01
-1.05729628e+00 9.82991233e-02 2.53527127e-02 -3.74498338e-01
9.32303444e-02 1.45909354e-01 -7.48000666e-02 -1.82070538e-01
3.23413610e-01 5.09712458e-01 -5.37578702e-01 -5.78486443e-01
5.60815297e-02 3.84859234e-01 6.04855001e-01 -2.67365038e-01
8.62599194e-01 6.23629391e-01 7.06367567e-02 -4.82738376e-01
-1.03160810e+00 -3.32824886e-01 -4.32371646e-01 1.11990757e-02
5.52511752e-01 -1.32211006e+00 -6.10705197e-01 5.10746956e-01
-1.15413845e+00 1.73050418e-01 -1.98641531e-02 4.29001242e-01
-5.32080121e-02 1.51551232e-01 -5.82513750e-01 -7.73965836e-01
-6.14442468e-01 -1.14025319e+00 1.20935774e+00 7.54052997e-01
2.85129428e-01 -3.42218071e-01 -5.58868647e-01 -2.33817875e-01
5.54542601e-01 3.94484624e-02 7.37266779e-01 -1.80950269e-01
-1.00878704e+00 8.39846879e-02 -9.32795286e-01 1.87722854e-02
1.80828482e-01 -1.30338982e-01 -9.15070415e-01 -2.25965396e-01
-6.10491298e-02 -1.41965613e-01 1.34144592e+00 4.45893258e-01
1.29641664e+00 -2.66782820e-01 -4.66848969e-01 4.74688590e-01
1.58089113e+00 -2.63061430e-02 4.22002286e-01 5.14185548e-01
7.76975036e-01 4.04008478e-01 9.06049490e-01 6.60916686e-01
4.93800581e-01 4.58965003e-01 8.22868764e-01 -5.02319694e-01
-1.58074588e-01 -3.39363456e-01 1.96645722e-01 6.35199487e-01
8.46652538e-02 2.13859841e-01 -4.96524215e-01 1.07696521e+00
-1.98361409e+00 -8.15493762e-01 -2.15543076e-01 1.88503575e+00
7.33468771e-01 3.13096017e-01 1.77589491e-01 9.90682617e-02
9.45201993e-01 4.26709741e-01 -6.90910816e-01 -1.03754047e-02
-2.48664424e-01 9.34330747e-02 4.63796318e-01 1.70990363e-01
-1.15764368e+00 1.14088941e+00 5.05334377e+00 9.06527102e-01
-1.19651496e+00 1.86713204e-01 6.13497555e-01 -4.00175571e-01
-1.78802133e-01 -1.39788255e-01 -1.01781535e+00 4.73890245e-01
2.27193490e-01 -1.49288237e-01 5.94336875e-02 1.23807979e+00
1.26940399e-01 -2.79148072e-01 -7.29479074e-01 8.64001989e-01
-4.30149361e-02 -1.47586429e+00 1.90906778e-01 -3.01569313e-01
8.73393714e-01 2.35880181e-01 2.69405395e-01 2.12131232e-01
7.95102958e-03 -5.86243808e-01 1.07092893e+00 1.45055428e-01
3.83101195e-01 -1.06420708e+00 7.14294553e-01 2.65825033e-01
-1.54032969e+00 -2.41753414e-01 -1.02764893e+00 -2.52967805e-01
1.48649607e-02 7.56647527e-01 -3.60981166e-01 3.19810629e-01
8.34113479e-01 8.67309988e-01 -8.40007305e-01 1.27824795e+00
-1.74166247e-01 4.10353839e-01 -3.21267307e-01 -3.13522995e-01
6.57828689e-01 3.92179936e-01 3.39853495e-01 1.10622633e+00
3.74638349e-01 4.42100726e-02 4.92815785e-02 9.64067996e-01
-2.07476303e-01 1.51972368e-01 -1.14512630e-01 4.50223327e-01
5.67716062e-01 1.46148288e+00 -8.39804947e-01 -5.25422335e-01
-6.88247263e-01 7.93101966e-01 5.53322792e-01 1.34506688e-01
-1.02954185e+00 -7.13083208e-01 7.34107137e-01 -1.09705061e-01
8.76483142e-01 8.64006653e-02 -4.43585664e-01 -1.12607348e+00
1.96936354e-01 -3.71855915e-01 3.82732227e-02 -7.22694457e-01
-5.86696565e-01 5.93613446e-01 -4.57078665e-01 -1.36560285e+00
3.19141269e-01 -3.03766757e-01 -6.01722062e-01 8.67007077e-01
-2.18425608e+00 -1.06257784e+00 -5.26406765e-01 6.53279543e-01
6.87645376e-01 1.68842927e-01 3.20414275e-01 1.86642081e-01
-6.79864049e-01 5.78908741e-01 2.69575939e-02 1.50342017e-01
2.95946151e-01 -8.68534744e-01 5.23655951e-01 1.22572792e+00
-9.46268439e-02 6.27892733e-01 4.86095309e-01 -4.80687022e-01
-1.11115468e+00 -1.42972720e+00 7.39008129e-01 2.26691097e-01
3.59431982e-01 -2.36394167e-01 -1.03017592e+00 3.21525991e-01
8.24110433e-02 1.67846188e-01 1.25418648e-01 -2.19222546e-01
-2.86878735e-01 -3.00922394e-01 -9.72891331e-01 6.79101825e-01
1.04312348e+00 -2.42233768e-01 -6.36580288e-01 -1.55718833e-01
9.24005687e-01 -2.56321132e-01 -2.43188888e-01 4.74366367e-01
4.38807905e-01 -1.13205278e+00 9.08752441e-01 -1.37826771e-01
6.66861415e-01 -6.54499352e-01 -4.62942384e-02 -9.10728693e-01
-7.26707458e-01 -3.64574462e-01 -1.74693719e-01 1.17288971e+00
3.34665239e-01 -4.59595442e-01 7.61818349e-01 1.85765058e-01
-3.30570102e-01 -1.03949428e+00 -6.62780643e-01 -5.78959882e-01
-4.81759399e-01 1.62245110e-02 7.40397036e-01 6.31897688e-01
-6.20130114e-02 2.88848042e-01 -3.22135746e-01 4.10913199e-01
7.16631711e-01 5.32786906e-01 3.05544287e-01 -1.22269166e+00
-2.91974209e-02 -4.90905970e-01 -5.54341614e-01 -1.46430469e+00
-1.91386610e-01 -5.71487844e-01 1.85973331e-01 -1.58451509e+00
5.59572101e-01 -4.26616460e-01 -8.85388732e-01 7.76741266e-01
-6.21930778e-01 4.97562587e-01 2.98640132e-01 1.11285359e-01
-1.05739093e+00 7.52884984e-01 1.43704104e+00 8.46285149e-02
-2.03993082e-01 -3.94913703e-01 -1.17825770e+00 8.74492109e-01
9.10654485e-01 -5.00720561e-01 -3.46466243e-01 -7.12719798e-01
-2.73002349e-02 -3.94065291e-01 6.88438177e-01 -1.35152590e+00
3.62278372e-01 -3.18954401e-02 6.92552626e-01 -9.33247089e-01
2.42060512e-01 -5.76571047e-01 -6.40570223e-01 5.65322459e-01
-2.52329409e-01 -1.49016574e-01 4.63696480e-01 4.27344620e-01
-3.60910118e-01 -1.35626853e-01 9.10920262e-01 -3.15361992e-02
-1.00864279e+00 2.09470451e-01 -2.08085492e-01 -1.02969594e-01
9.18528199e-01 -1.53887376e-01 -5.59074640e-01 -1.48814633e-01
-1.22650728e-01 1.99735582e-01 3.98037136e-01 6.41500831e-01
9.17293072e-01 -1.38065374e+00 -5.32222748e-01 2.65286624e-01
8.38540774e-03 1.82508186e-01 3.51686627e-01 8.35421562e-01
-3.01151633e-01 6.01394951e-01 -5.41711032e-01 -5.60116649e-01
-1.10111189e+00 6.21602118e-01 -1.41584883e-02 3.71239930e-02
-5.92441618e-01 1.10118628e+00 8.10925305e-01 3.12551528e-01
1.59078419e-01 -5.52197397e-01 -3.23378026e-01 -8.57909992e-02
8.93843770e-01 3.06701809e-01 -1.64478093e-01 -8.68916452e-01
-6.17878973e-01 6.19460225e-01 -4.56080854e-01 3.28285009e-01
1.44230342e+00 -1.67547777e-01 -2.82838903e-02 -8.96903500e-02
1.20614612e+00 -2.93230355e-01 -1.50854778e+00 -4.89407182e-01
-2.57077217e-01 -5.28666079e-01 4.18654114e-01 -4.29900557e-01
-1.40094256e+00 1.02826357e+00 5.24802327e-01 -1.24587141e-01
1.52869952e+00 -6.81899190e-02 1.02446008e+00 3.10783684e-01
2.93131709e-01 -1.01896131e+00 1.96178913e-01 3.27674508e-01
6.46474242e-01 -1.10616779e+00 7.92825967e-03 -6.65163577e-01
-5.34226954e-01 8.56091380e-01 9.40740287e-01 -4.81846958e-01
3.31549376e-01 4.86974642e-02 -2.60749042e-01 -1.93778072e-02
-5.97895443e-01 -6.17705524e-01 1.99132949e-01 4.21464652e-01
1.31087676e-01 -1.75178602e-01 -3.43716711e-01 7.79841602e-01
8.34774449e-02 7.30998814e-02 4.47586268e-01 9.15350080e-01
-1.11488640e+00 -4.82357889e-01 -1.45696983e-01 5.11470556e-01
-4.78026181e-01 -4.41776216e-01 3.39802355e-02 5.61674058e-01
1.69445962e-01 9.17765319e-01 2.58793570e-02 -3.44619691e-01
1.68777272e-01 -4.76365477e-01 1.08939894e-01 -4.77103055e-01
-5.32956183e-01 2.34323982e-02 -2.58600503e-01 -6.24887764e-01
-4.37813550e-01 -5.73046684e-01 -1.49567282e+00 1.28469067e-02
-5.74063182e-01 2.10811272e-02 3.90477300e-01 7.74749041e-01
6.26850188e-01 8.16199303e-01 5.93452871e-01 -9.51988876e-01
-2.86823601e-01 -8.35165620e-01 -4.99868602e-01 7.30843768e-02
5.85756719e-01 -9.32367563e-01 -3.14018503e-02 -1.62033781e-01] | [9.676881790161133, -0.47118112444877625] |
a6adfae5-d85e-40e4-b146-5b8f4955e7e5 | evaluation-of-embedding-models-for-automatic | 2206.10939 | null | https://arxiv.org/abs/2206.10939v1 | https://arxiv.org/pdf/2206.10939v1.pdf | Evaluation of Embedding Models for Automatic Extraction and Classification of Acknowledged Entities in Scientific Documents | Acknowledgments in scientific papers may give an insight into aspects of the scientific community, such as reward systems, collaboration patterns, and hidden research trends. The aim of the paper is to evaluate the performance of different embedding models for the task of automatic extraction and classification of acknowledged entities from the acknowledgment text in scientific papers. We trained and implemented a named entity recognition (NER) task using the Flair NLP-framework. The training was conducted using three default Flair NER models with two differently-sized corpora. The Flair Embeddings model trained on the larger training corpus showed the best accuracy of 0.77. Our model is able to recognize six entity types: funding agency, grant number, individuals, university, corporation and miscellaneous. The model works more precise for some entity types than the others, thus, individuals and grant numbers showed very good F1-Score over 0.9. Most of the previous works on acknowledgement analysis were limited by the manual evaluation of data and therefore by the amount of processed data. This model can be applied for the comprehensive analysis of the acknowledgement texts and may potentially make a great contribution to the field of automated acknowledgement analysis. | ['Philipp Mayr', 'Nina Smirnova'] | 2022-06-22 | null | null | null | null | ['miscellaneous'] | ['miscellaneous'] | [-5.02537072e-01 2.99639851e-01 -1.09182745e-01 -1.09894522e-01
-2.75734484e-01 -4.51038212e-01 8.13060701e-01 6.71543121e-01
-1.02320409e+00 7.63067007e-01 3.28348279e-01 -4.51771200e-01
-1.85665295e-01 -8.10830712e-01 -1.69689745e-01 -3.88427854e-01
1.68570474e-01 5.05566835e-01 6.97248578e-02 3.59083004e-02
7.78955579e-01 8.73971939e-01 -1.15113807e+00 -8.50246400e-02
6.55910492e-01 5.15616238e-01 5.14488146e-02 8.40158165e-01
-7.01089323e-01 8.71646583e-01 -9.41302538e-01 -7.66148388e-01
-1.06997021e-01 -1.31818401e-02 -7.39605367e-01 -5.89602828e-01
2.16158777e-01 1.31846771e-01 -1.64109230e-01 7.70872712e-01
6.26519799e-01 -3.51726189e-02 7.96211541e-01 -1.08539987e+00
-9.63540316e-01 9.33118224e-01 -1.35509089e-01 4.47814137e-01
4.16017562e-01 -3.09931099e-01 1.03745842e+00 -1.07946885e+00
1.17316198e+00 9.14386392e-01 7.34915912e-01 3.93747002e-01
-9.35284853e-01 -8.12082589e-01 -5.13239443e-01 5.00735521e-01
-1.30843103e+00 -2.45460480e-01 6.59790039e-01 -7.65507281e-01
1.06426549e+00 1.48642570e-01 3.33472848e-01 1.04436493e+00
3.16836923e-01 3.11782897e-01 9.40795004e-01 -5.59004366e-01
3.02684933e-01 7.86601067e-01 6.80394411e-01 3.82318050e-01
5.97328961e-01 -4.00781661e-01 -5.02468467e-01 -2.62594193e-01
5.18781781e-01 -1.69071630e-01 -1.99941054e-01 1.19163185e-01
-1.42933488e+00 9.56557453e-01 -4.51024435e-02 1.10236871e+00
-7.22496331e-01 -7.69275948e-02 4.21163440e-01 2.35763311e-01
4.37781930e-01 1.07982540e+00 -4.56293881e-01 -5.30277193e-01
-1.16751659e+00 1.59915745e-01 1.31443882e+00 5.37545502e-01
4.62374449e-01 -3.15655768e-02 -2.40569681e-01 7.05907941e-01
-3.02676298e-02 7.54200146e-02 4.82098371e-01 -9.45787907e-01
2.89918005e-01 8.19323540e-01 3.13877046e-01 -1.25090170e+00
-3.75012130e-01 -2.60378838e-01 -4.89592284e-01 -7.92867765e-02
6.39105916e-01 -4.43258733e-01 -2.68716753e-01 1.11980641e+00
1.95626676e-01 -1.30664036e-01 3.23483974e-01 6.27699614e-01
1.38092041e+00 7.75372088e-01 1.80348217e-01 4.80514914e-02
1.51065338e+00 -7.89529562e-01 -1.18963468e+00 3.34104270e-01
7.79030085e-01 -1.18342233e+00 2.67871678e-01 -6.83762133e-02
-8.05255890e-01 -5.04380405e-01 -6.87353551e-01 -6.98633343e-02
-1.04585218e+00 6.24581397e-01 7.33071625e-01 6.19490385e-01
-9.55570459e-01 7.29139626e-01 -2.77266175e-01 -5.72116077e-01
2.80712664e-01 2.98823148e-01 -7.27712333e-01 1.85562864e-01
-1.24858546e+00 1.00990903e+00 2.18570635e-01 3.00856028e-02
9.37497020e-02 -8.70187402e-01 -6.88792646e-01 2.47834370e-01
1.69329252e-02 -2.44582817e-01 5.56600690e-01 -4.18818891e-01
-1.14111042e+00 1.15552986e+00 -5.09258807e-02 -4.15369362e-01
2.92525351e-01 -6.98750168e-02 -6.48776531e-01 8.40945095e-02
6.54272735e-02 3.95669639e-01 1.50099918e-01 -6.25095546e-01
-4.17398751e-01 -3.35995525e-01 -3.16255748e-01 -4.96958405e-01
-6.95472658e-01 6.04077518e-01 6.46072403e-02 -6.34693325e-01
-5.24729073e-01 -5.65201223e-01 -1.13356531e-01 -4.16864842e-01
-1.59471959e-01 -9.13659871e-01 3.58156919e-01 -8.41724694e-01
1.42096388e+00 -2.01248074e+00 8.44362304e-02 -7.11097196e-02
4.05443788e-01 3.17995906e-01 6.37580603e-02 7.82172918e-01
-1.78054482e-01 5.81736147e-01 2.96278775e-01 7.07360804e-02
1.83922619e-01 -9.78987291e-02 4.21590209e-02 3.14476818e-01
2.35377640e-01 7.25691497e-01 -8.39780629e-01 -5.72977006e-01
-1.00141123e-01 6.09783888e-01 -5.66948839e-02 5.02064943e-01
5.41101098e-01 -2.12238103e-01 -7.17667460e-01 4.56563562e-01
6.49811506e-01 7.09627122e-02 2.13236529e-02 -7.82862008e-02
-7.49162972e-01 -6.54150173e-02 -1.16937244e+00 1.08534503e+00
-3.86907429e-01 1.08244872e+00 -2.92191859e-02 -9.22752321e-01
1.34943116e+00 3.87201518e-01 5.13047159e-01 -1.91757903e-01
2.03447074e-01 4.27495003e-01 1.89093143e-01 -8.34197581e-01
8.64668906e-01 1.32911742e-01 -1.10662408e-01 5.03218889e-01
2.79098719e-01 2.60165751e-01 3.90449584e-01 5.05098283e-01
1.29890335e+00 -1.63906172e-01 2.71663398e-01 -3.79648417e-01
7.94193864e-01 -7.22495615e-02 4.37724590e-01 6.74139082e-01
-4.20672059e-01 2.49677807e-01 7.90652335e-01 -6.52120888e-01
-1.12036824e+00 -5.51317871e-01 -3.84740889e-01 8.33872378e-01
-3.45338166e-01 -4.90037382e-01 -5.34282446e-01 -6.80109382e-01
1.99121401e-01 8.31449449e-01 -6.64680421e-01 3.03346962e-01
-4.52356309e-01 -5.09825408e-01 5.85627556e-01 3.52411628e-01
1.87061235e-01 -1.52548099e+00 -6.30286992e-01 3.38024825e-01
1.59395307e-01 -1.28298390e+00 -2.14356795e-01 6.66986555e-02
-6.01690471e-01 -1.15073526e+00 -1.04604220e+00 -8.68895710e-01
3.49671304e-01 -4.24026906e-01 1.02910173e+00 -1.52975500e-01
-4.35522646e-01 4.33672726e-01 -2.41832167e-01 -6.76082492e-01
-3.97068381e-01 2.78292060e-01 -8.43036547e-02 -2.54601240e-01
9.58928168e-01 -1.67748675e-01 -2.08941773e-01 1.00979116e-02
-5.71023822e-01 -6.54565632e-01 5.78160107e-01 8.02709877e-01
5.37669286e-02 -2.67664999e-01 7.63021708e-01 -1.10264826e+00
1.07361388e+00 -5.28456390e-01 -2.33573452e-01 2.76538372e-01
-8.99578035e-01 2.69864529e-01 6.73783123e-01 -4.46907669e-01
-8.81294131e-01 -4.25124675e-01 -2.59840071e-01 -1.37495726e-01
-3.70048493e-01 3.52721632e-01 2.83762012e-02 -1.27186209e-01
4.24699634e-01 -9.66380537e-02 -2.74007082e-01 -6.70913100e-01
1.85428694e-01 1.13371289e+00 3.28254491e-01 -3.57704073e-01
3.67094904e-01 -1.51516587e-01 -5.52189872e-02 -1.06480110e+00
-4.05609250e-01 -8.27451110e-01 -5.43244720e-01 -1.91499278e-01
8.60163212e-01 -4.21165556e-01 -7.42822111e-01 2.03410760e-01
-1.60470402e+00 2.94194698e-01 -3.89298260e-01 8.91034603e-01
2.70309448e-01 4.40892339e-01 -7.11834490e-01 -9.71752405e-01
-5.84746778e-01 -8.67010176e-01 5.40661156e-01 4.81956631e-01
-4.26947147e-01 -1.35710740e+00 3.41424763e-01 3.80789042e-01
4.83915508e-01 2.88579911e-01 7.04747975e-01 -1.29694724e+00
3.85550708e-02 -4.83277351e-01 -3.06245476e-01 1.81229085e-01
-2.39949301e-01 2.20436767e-01 -8.28172386e-01 2.93100476e-01
-1.47609875e-01 -9.52918604e-02 9.29570198e-01 -2.59390455e-02
9.13791180e-01 -3.34433407e-01 -2.95071065e-01 6.46052435e-02
1.28390241e+00 3.74515317e-02 6.44945681e-01 7.50985742e-01
5.66949010e-01 9.56968844e-01 3.11393976e-01 5.42188406e-01
2.52741665e-01 4.93962377e-01 -4.93058041e-02 1.68248415e-01
4.19453792e-02 1.62625685e-01 3.13336670e-01 8.97637904e-01
-4.00875896e-01 -2.88256586e-01 -1.02928340e+00 8.45427155e-01
-1.58037138e+00 -1.19659448e+00 -7.49500632e-01 1.89288378e+00
6.26961768e-01 -4.35432307e-02 -7.11727217e-02 3.39418538e-02
8.30425024e-01 1.18212767e-01 1.36013687e-01 -1.19388163e+00
-1.78326190e-01 4.60631102e-01 6.36536717e-01 1.55160680e-01
-7.56007612e-01 7.52856076e-01 6.36112165e+00 6.77387953e-01
-8.24431300e-01 1.34391770e-01 2.15178996e-01 3.08856755e-01
-1.90056205e-01 3.79052646e-02 -1.08995914e+00 4.94175076e-01
1.56226814e+00 -3.72800291e-01 -6.57109544e-02 7.57011712e-01
2.11790100e-01 1.11888662e-01 -8.22569072e-01 8.77161384e-01
1.21012263e-01 -1.59783518e+00 -1.80008531e-01 2.28527412e-01
3.14762950e-01 -4.75878865e-02 -4.73365903e-01 4.58050877e-01
8.43003541e-02 -9.12187576e-01 3.27495843e-01 8.63146722e-01
4.17635441e-01 -7.26414800e-01 1.48281825e+00 1.58402294e-01
-6.00706935e-01 9.93791781e-03 -7.15232074e-01 -1.43050373e-01
-6.75611123e-02 9.22607839e-01 -8.15704525e-01 5.96565187e-01
6.08919740e-01 7.41959095e-01 -5.77458322e-01 8.53477120e-01
-2.23639548e-01 7.47948408e-01 -1.07062049e-01 -5.74785888e-01
1.15926728e-01 -2.77892351e-01 5.60946345e-01 1.64154184e+00
3.82051855e-01 -7.82248471e-03 -3.50804150e-01 8.85119796e-01
-4.91574198e-01 6.64184749e-01 -6.04039788e-01 -6.63208544e-01
4.85862374e-01 1.68548346e+00 -5.54167151e-01 -2.94533998e-01
-2.34179303e-01 7.60797441e-01 4.66183156e-01 1.30489826e-01
-5.16920447e-01 -1.14547038e+00 2.97282577e-01 1.86031550e-01
4.34130609e-01 -6.96796551e-02 -1.91922724e-01 -1.06210220e+00
-1.46439999e-01 -3.10273498e-01 3.93603981e-01 -6.49642646e-01
-1.19032204e+00 6.02060616e-01 -4.32681888e-01 -5.43273985e-01
-4.30065617e-02 -9.65099275e-01 -8.67585003e-01 9.54498947e-01
-1.34288037e+00 -6.93559766e-01 -1.02031820e-01 -1.09526038e-01
1.18765056e-01 -6.53878212e-01 1.12800980e+00 2.99369067e-01
-8.22992086e-01 6.59246266e-01 3.96306574e-01 4.57103163e-01
9.34230208e-01 -1.33810794e+00 -1.69195250e-01 4.85912412e-01
1.38226390e-01 7.97221422e-01 7.27007687e-01 -3.99472177e-01
-1.22034645e+00 -7.65533924e-01 2.08706355e+00 -5.73954582e-01
1.12657952e+00 -3.35709862e-02 -9.41047549e-01 1.37809888e-01
7.05919921e-01 -1.58041894e-01 1.15275276e+00 3.61518234e-01
-2.92232662e-01 1.29626542e-01 -1.14590967e+00 2.26390034e-01
4.41555291e-01 -2.64363527e-01 -9.04387534e-01 2.97171503e-01
3.77773136e-01 1.57842755e-01 -1.69814849e+00 -2.08212331e-01
6.87565863e-01 -6.55070186e-01 7.34375119e-01 -7.87570953e-01
9.24242377e-01 2.45586544e-01 3.11297894e-01 -9.78688180e-01
-5.27032375e-01 -3.90048206e-01 -8.11540037e-02 1.77125704e+00
5.11277020e-01 -6.36869371e-01 6.21495664e-01 8.70704114e-01
1.93513766e-01 -7.41695285e-01 -9.60335016e-01 -6.47467375e-01
3.16376716e-01 -1.38551861e-01 3.77456456e-01 1.19198608e+00
1.56477198e-01 4.53660727e-01 -5.53515367e-02 -4.01544213e-01
4.36427563e-01 1.62203550e-01 6.72581315e-01 -1.66173708e+00
2.10827157e-01 -5.54702818e-01 -7.99752772e-01 -1.31705090e-01
6.05220139e-01 -9.54391778e-01 -5.00834346e-01 -1.77983308e+00
1.77071914e-01 -3.43745679e-01 -4.79018241e-01 4.39620495e-01
-1.15119234e-01 -4.87134717e-02 1.50403589e-01 3.89429450e-01
-4.01965916e-01 2.72820890e-01 8.27874124e-01 -1.28911942e-01
-6.03296943e-02 -2.24146485e-01 -6.80178165e-01 3.86189342e-01
8.31500769e-01 -8.42500746e-01 5.09853423e-01 -4.46958803e-02
2.56295621e-01 -2.96373218e-01 1.49410218e-01 -6.86478257e-01
3.81335646e-01 8.70183809e-04 4.00912136e-01 -2.06081793e-01
-2.32636645e-01 -6.90593541e-01 -7.31231496e-02 3.07416290e-01
-4.31738734e-01 1.69778213e-01 5.67928217e-02 3.75884891e-01
-1.62678316e-01 -8.87524009e-01 2.37293690e-01 -5.28106950e-02
-5.81663847e-01 -2.83084333e-01 -5.98465621e-01 7.82739446e-02
9.45995390e-01 -2.13593617e-01 -2.67981738e-01 1.82710811e-02
-5.05708873e-01 1.81135312e-02 4.34276342e-01 4.53333318e-01
4.12156403e-01 -9.99343574e-01 -8.95093143e-01 -3.67837220e-01
9.25831050e-02 -6.99101388e-01 4.99332100e-02 8.43732178e-01
-6.00120783e-01 5.72704136e-01 -3.84070128e-01 1.30508617e-01
-1.34283078e+00 4.89750028e-01 -8.11972246e-02 -5.66367745e-01
-4.77456391e-01 5.36296189e-01 -5.04328191e-01 -4.52325493e-01
1.12672076e-01 6.41002953e-02 -1.09968495e+00 4.26888078e-01
5.58233380e-01 8.87913644e-01 -1.67873338e-01 -7.86148727e-01
-5.72120368e-01 5.22018611e-01 -1.57641634e-01 8.42204615e-02
1.70957243e+00 5.20074368e-01 -4.13189948e-01 5.45347989e-01
1.30346465e+00 6.33703411e-01 -2.00710565e-01 2.11118385e-01
3.60532492e-01 -1.08027779e-01 3.63255709e-01 -6.90212786e-01
-9.62494135e-01 8.68455291e-01 4.73113686e-01 2.71966964e-01
4.08378273e-01 -1.67950392e-01 5.20730495e-01 3.74885798e-01
-6.98063374e-02 -1.32281792e+00 -4.87020344e-01 3.89059901e-01
6.29816294e-01 -1.15456223e+00 2.27060184e-01 -1.43344954e-01
-5.52336872e-01 1.77989447e+00 1.93074808e-01 7.58405626e-02
6.14099026e-01 2.41907015e-01 3.49990316e-02 -3.69082481e-01
-4.76005793e-01 1.05540887e-01 3.38137180e-01 4.09496844e-01
1.01713669e+00 -1.19838864e-02 -1.25994885e+00 1.12704444e+00
-4.09111641e-02 2.98725575e-01 1.07628274e+00 6.05648100e-01
-2.32403085e-01 -1.39026058e+00 -3.81838441e-01 4.97143239e-01
-1.16611457e+00 1.62928123e-02 -1.00984299e+00 8.31360161e-01
2.72322357e-01 8.40845466e-01 -2.18784958e-02 -1.49971977e-01
4.10019904e-01 3.89152378e-01 1.09381668e-01 -4.22559887e-01
-1.14432287e+00 -4.78280485e-01 2.45084107e-01 -1.25084773e-01
-5.44448078e-01 -7.22010136e-01 -1.15223062e+00 -4.62656856e-01
-3.38216484e-01 9.40510094e-01 9.90409374e-01 5.89815855e-01
6.98864341e-01 5.85153222e-01 4.62908119e-01 -3.06527942e-01
-2.87733406e-01 -1.19230819e+00 -7.63019741e-01 4.31441396e-01
-3.45766217e-01 -4.94575351e-01 -6.21925235e-01 -7.45173842e-02] | [9.80560302734375, 8.763481140136719] |
e0bec42f-8310-43ef-b004-fc07b8cb2453 | targeted-adversarial-attacks-against-neural | 2212.04138 | null | https://arxiv.org/abs/2212.04138v1 | https://arxiv.org/pdf/2212.04138v1.pdf | Targeted Adversarial Attacks against Neural Network Trajectory Predictors | Trajectory prediction is an integral component of modern autonomous systems as it allows for envisioning future intentions of nearby moving agents. Due to the lack of other agents' dynamics and control policies, deep neural network (DNN) models are often employed for trajectory forecasting tasks. Although there exists an extensive literature on improving the accuracy of these models, there is a very limited number of works studying their robustness against adversarially crafted input trajectories. To bridge this gap, in this paper, we propose a targeted adversarial attack against DNN models for trajectory forecasting tasks. We call the proposed attack TA4TP for Targeted adversarial Attack for Trajectory Prediction. Our approach generates adversarial input trajectories that are capable of fooling DNN models into predicting user-specified target/desired trajectories. Our attack relies on solving a nonlinear constrained optimization problem where the objective function captures the deviation of the predicted trajectory from a target one while the constraints model physical requirements that the adversarial input should satisfy. The latter ensures that the inputs look natural and they are safe to execute (e.g., they are close to nominal inputs and away from obstacles). We demonstrate the effectiveness of TA4TP on two state-of-the-art DNN models and two datasets. To the best of our knowledge, we propose the first targeted adversarial attack against DNN models used for trajectory forecasting. | ['Yiannis Kantaros', 'Jun Wang', 'Kaiyuan Tan'] | 2022-12-08 | null | null | null | null | ['trajectory-forecasting'] | ['computer-vision'] | [-1.75988659e-01 2.55016029e-01 -3.37081909e-01 -6.65373653e-02
-2.07235053e-01 -8.24073851e-01 8.07885230e-01 -3.19806486e-01
-2.14560121e-01 6.84012651e-01 4.45977002e-02 -6.58862054e-01
-1.19827064e-02 -1.08895111e+00 -1.10111403e+00 -7.33273506e-01
-2.33814061e-01 4.85007346e-01 3.53851527e-01 -4.15357023e-01
-9.16228592e-02 1.02677333e+00 -1.25870919e+00 -8.35513994e-02
8.48445475e-01 8.83903444e-01 -2.83431619e-01 8.73831034e-01
4.02924061e-01 1.07930589e+00 -6.40352249e-01 -2.75107205e-01
5.41736484e-01 -1.14197306e-01 -5.46774566e-01 -5.05786300e-01
9.61659402e-02 -8.36114287e-01 -1.17776632e+00 1.12980485e+00
1.86247513e-01 5.03170609e-01 6.65220976e-01 -1.92576742e+00
-6.64450228e-01 5.62623620e-01 4.72741514e-01 6.45331219e-02
-8.46204907e-03 7.56759226e-01 3.72268409e-01 -2.22802252e-01
4.32411164e-01 1.13797939e+00 5.60392737e-01 1.21354580e+00
-9.40941572e-01 -6.72869802e-01 5.04740298e-01 4.59165983e-02
-1.11304843e+00 -4.51753587e-01 6.83395743e-01 -4.92970884e-01
1.04974604e+00 3.92185122e-01 4.79873598e-01 1.72512555e+00
5.20821750e-01 6.80537343e-01 3.67557943e-01 2.73831517e-01
6.13180757e-01 4.19464670e-02 -3.50209773e-01 2.49196455e-01
-9.72790122e-02 8.38874400e-01 5.96631169e-02 -3.34612221e-01
5.59773147e-01 2.06341580e-01 -1.12711035e-01 -7.68849105e-02
-1.02012897e+00 6.51227176e-01 5.25244951e-01 -1.64114550e-01
-5.11169434e-01 5.70294499e-01 3.64502698e-01 2.22265378e-01
1.37462169e-01 2.95666605e-01 -2.58730382e-01 -1.62499353e-01
-3.52112979e-01 9.28948462e-01 7.06458688e-01 1.23716009e+00
9.64109078e-02 6.94301128e-01 -1.98373288e-01 4.63920645e-03
2.27367640e-01 1.13261461e+00 8.91352743e-02 -9.53257322e-01
5.21259189e-01 3.94240469e-01 5.59152305e-01 -1.14769936e+00
-4.20566916e-01 9.65328014e-04 -6.73461139e-01 6.13173366e-01
5.95141113e-01 -6.23237133e-01 -8.10832322e-01 1.74072528e+00
1.71109691e-01 6.22134268e-01 5.08489311e-01 1.10358322e+00
3.92532080e-01 9.42765355e-01 1.85213521e-01 2.42242754e-01
4.43857282e-01 -8.10195744e-01 -4.14930969e-01 -1.82586625e-01
4.98076797e-01 -3.68546039e-01 6.69728518e-01 2.17545435e-01
-8.32428217e-01 -2.49641642e-01 -8.38176787e-01 3.64550471e-01
-4.93737668e-01 -2.52881974e-01 4.11904782e-01 7.95713544e-01
-6.86778724e-01 6.74221277e-01 -1.14107156e+00 -1.18306980e-01
4.03728873e-01 4.69565779e-01 -1.90521555e-03 1.82888448e-01
-1.43418980e+00 1.10575497e+00 2.83707619e-01 1.76101238e-01
-1.64510965e+00 -7.68932462e-01 -7.07552612e-01 -1.31704167e-01
3.27219099e-01 -3.85364980e-01 1.37330425e+00 -8.53094101e-01
-1.64767742e+00 1.43923864e-01 2.05374211e-01 -1.01462078e+00
8.74400795e-01 -2.25024730e-01 -8.65569651e-01 -1.56922147e-01
-3.43775690e-01 5.54358542e-01 9.12375450e-01 -1.07276964e+00
-5.63322961e-01 7.59618124e-03 5.55643082e-01 -3.15752104e-02
-2.73384064e-01 -1.26137957e-01 1.05190929e-02 -6.85986042e-01
-5.46836972e-01 -1.27099752e+00 -5.75423419e-01 1.44296288e-01
-6.87944770e-01 4.81595211e-02 1.22959232e+00 -2.96325803e-01
1.12017846e+00 -1.89396942e+00 1.01328120e-01 3.08988333e-01
1.03865556e-01 6.84124768e-01 -1.33075505e-01 5.58549345e-01
1.84385747e-01 1.14747882e-01 1.45220086e-02 -4.62021753e-02
3.47253799e-01 4.44443405e-01 -1.27904546e+00 6.43600404e-01
-1.41221692e-03 1.03277433e+00 -9.77550983e-01 1.95204198e-01
4.08942312e-01 5.06361783e-01 -4.41059262e-01 3.77108574e-01
-6.20318413e-01 6.70066118e-01 -7.16181338e-01 4.94732738e-01
5.55103123e-01 5.91442466e-01 -2.13065192e-01 2.67176211e-01
-1.11417696e-01 8.35853741e-02 -7.30473101e-01 7.96011925e-01
-2.61631936e-01 7.23042130e-01 -1.88672677e-01 -6.40639961e-01
9.02800500e-01 3.97477806e-01 4.62342680e-01 -4.51852530e-01
3.24769229e-01 2.46829465e-02 1.16547689e-01 -3.78127664e-01
4.76446807e-01 -1.80197563e-02 -3.26520383e-01 2.83673674e-01
-5.70361733e-01 -1.39981836e-01 -3.32776338e-01 -1.33957276e-02
1.12343001e+00 5.29069342e-02 -1.39507145e-01 -7.54518881e-02
4.98381466e-01 3.20335805e-01 5.92629850e-01 8.79316390e-01
-5.99085927e-01 1.57228515e-01 4.75489318e-01 -8.94295275e-01
-1.26622295e+00 -1.10513365e+00 2.74447620e-01 6.87647581e-01
3.19976687e-01 -2.39343718e-02 -8.98301125e-01 -9.59480345e-01
1.13568410e-01 1.22762895e+00 -6.94095910e-01 -6.29528582e-01
-8.30640376e-01 -5.36354631e-02 1.27305150e+00 7.41617918e-01
2.34362766e-01 -1.19911325e+00 -6.73749328e-01 2.21548274e-01
2.35460132e-01 -1.02285385e+00 -6.13008916e-01 -3.22355449e-01
-5.70168912e-01 -1.02264404e+00 -3.27923328e-01 -4.44457799e-01
8.03497553e-01 5.79791851e-02 6.93785429e-01 9.34477970e-02
4.85442340e-01 3.01152110e-01 -1.78900704e-01 -6.39856219e-01
-8.92449439e-01 -1.43111542e-01 6.73768163e-01 -5.09810597e-02
2.66358525e-01 -4.78985608e-01 -4.23273355e-01 6.82016969e-01
-9.50368404e-01 -1.69455975e-01 -6.54788911e-02 4.14419562e-01
3.86216313e-01 2.86093593e-01 6.56908453e-01 -5.65727711e-01
7.15245426e-01 -7.38980353e-01 -9.67540979e-01 -4.49675471e-02
-2.29712591e-01 -1.21390000e-01 1.58792007e+00 -8.92525971e-01
-6.18577838e-01 4.01333161e-02 -2.58451104e-01 -8.78677726e-01
-4.84937638e-01 4.01054956e-02 -3.15203905e-01 -2.23802939e-01
7.11692929e-01 2.43553102e-01 -1.52111396e-01 1.51272982e-01
3.82894486e-01 3.80273253e-01 5.70068777e-01 -5.90526462e-01
1.21828675e+00 5.97901702e-01 1.01063699e-01 -5.76014757e-01
-3.42780858e-01 3.28495592e-01 -2.61485249e-01 -6.97703362e-01
8.07783663e-01 -4.75976944e-01 -1.35550594e+00 6.14029109e-01
-1.16357636e+00 -8.28828096e-01 -1.38889760e-01 3.42083424e-01
-7.98831344e-01 -2.25004107e-02 -3.71792942e-01 -1.02118957e+00
-7.15839937e-02 -1.38590550e+00 5.27317047e-01 1.27455443e-01
-1.81743994e-01 -1.12824762e+00 1.41193837e-01 -1.28801271e-01
7.19922900e-01 7.17514992e-01 5.42949200e-01 -8.31222177e-01
-7.95826495e-01 -4.59090948e-01 3.08616549e-01 1.52917325e-01
-1.15907989e-01 7.27360174e-02 -7.53177345e-01 -2.49580681e-01
-1.24570347e-01 -1.39178962e-01 4.06646132e-01 2.91834146e-01
1.08178020e+00 -1.00590253e+00 -5.25942862e-01 6.26793623e-01
1.01992691e+00 7.05029964e-01 7.38623559e-01 3.79963458e-01
7.77805567e-01 3.81900072e-01 7.03829229e-01 2.19966069e-01
2.09376797e-01 6.94561541e-01 1.15045607e+00 4.34669793e-01
5.18292904e-01 -4.80219215e-01 6.46619797e-01 -2.75680292e-02
-1.56927165e-02 -7.87296891e-01 -1.01694798e+00 5.50525129e-01
-2.03790998e+00 -1.20838082e+00 -8.62517059e-02 2.16343856e+00
1.13381259e-01 5.16066968e-01 2.52161801e-01 -2.33989865e-01
4.91795331e-01 1.95389345e-01 -1.23413718e+00 -5.98216116e-01
1.87835246e-01 -4.01835591e-01 8.81890595e-01 6.17367387e-01
-1.36919570e+00 1.15064502e+00 5.79913759e+00 6.44564629e-01
-1.26243353e+00 -2.96765059e-01 4.01475221e-01 -2.27523237e-01
-2.69442409e-01 -2.61171401e-01 -7.46069312e-01 6.89555347e-01
1.22789526e+00 -3.09272289e-01 7.03586578e-01 1.19324136e+00
5.85542083e-01 4.96075898e-01 -1.21474147e+00 4.11076605e-01
-3.00126553e-01 -1.44571459e+00 4.48742844e-02 7.75348470e-02
7.02489078e-01 1.43243909e-01 4.60867852e-01 5.13834715e-01
7.31105089e-01 -1.39651501e+00 9.48194146e-01 6.98629737e-01
4.59440768e-01 -1.28281868e+00 5.68668544e-01 7.12901175e-01
-1.01982892e+00 -1.95613578e-01 -2.27075562e-01 -1.49715051e-01
4.00176615e-01 -1.93407029e-01 -6.95773721e-01 2.33721301e-01
3.27824056e-01 5.82069814e-01 1.25661835e-01 6.92698002e-01
-2.60441899e-01 7.95430779e-01 -3.77543032e-01 -1.38272077e-01
7.02738762e-01 -5.77970371e-02 1.10679340e+00 7.67320931e-01
2.15263590e-01 1.33949250e-01 3.50408226e-01 1.02738488e+00
4.51528914e-02 -5.06807387e-01 -1.22403681e+00 -1.70677364e-01
5.60179234e-01 6.01570845e-01 -2.96186209e-01 -5.91742881e-02
-6.49883747e-02 7.64006853e-01 1.19324982e-01 5.16653478e-01
-1.37874293e+00 -2.24892110e-01 1.41090107e+00 -1.29663378e-01
-1.48560908e-02 -4.10360366e-01 -2.97926247e-01 -8.59148264e-01
-5.73650710e-02 -8.39742064e-01 -9.43004861e-02 -3.71266991e-01
-1.12216020e+00 8.75295103e-01 -2.33135566e-01 -1.59978604e+00
-4.88018543e-01 -5.44100881e-01 -9.91514027e-01 7.62487829e-01
-1.11281836e+00 -1.15008235e+00 6.07126541e-02 7.03532517e-01
4.13413078e-01 -4.54293281e-01 8.38259578e-01 1.34634048e-01
-7.46977210e-01 6.72715902e-01 1.26068041e-01 2.96160787e-01
2.30794355e-01 -9.67725158e-01 1.11131573e+00 1.27217269e+00
-3.88141334e-01 4.76911098e-01 1.06824696e+00 -9.32431757e-01
-1.50751901e+00 -1.76113927e+00 4.72945243e-01 -8.29567671e-01
8.37978244e-01 -2.87649006e-01 -8.82568121e-01 1.07119846e+00
-2.00233072e-01 8.62932280e-02 1.01881064e-01 -7.51510441e-01
-1.47981361e-01 1.34812132e-01 -1.37111604e+00 1.14022005e+00
8.85367393e-01 -3.32200319e-01 -1.26520887e-01 3.83086175e-01
9.13726449e-01 -8.37152600e-01 -7.30259001e-01 1.96104512e-01
5.71647167e-01 -6.55308902e-01 1.17489409e+00 -1.21273100e+00
2.01023936e-01 -4.39300478e-01 -2.63931781e-01 -1.47478652e+00
-3.61563750e-02 -8.48207235e-01 -4.31097955e-01 8.09867144e-01
3.23457807e-01 -7.98639059e-01 9.47649002e-01 1.27839386e+00
-4.06276494e-01 -8.17291319e-01 -9.21994090e-01 -1.12201405e+00
3.98898602e-01 -8.66907656e-01 9.32362676e-01 6.54350638e-01
-1.38009027e-01 -4.74965513e-01 -9.39478219e-01 8.83591652e-01
5.16690969e-01 -3.25814486e-01 9.24482346e-01 -6.14463627e-01
7.06561208e-02 -5.59448004e-01 -4.55786318e-01 -1.07593405e+00
5.69151044e-01 -6.54576063e-01 1.29946381e-01 -1.19988215e+00
-6.45222008e-01 -6.53030097e-01 9.86075997e-02 3.72815937e-01
2.12737814e-01 -2.05784827e-01 3.68284345e-01 8.51314664e-02
-3.38480920e-01 7.43251622e-01 9.45227206e-01 -3.51151556e-01
-3.15036654e-01 6.96061790e-01 -2.28835106e-01 8.83544207e-01
1.21762574e+00 -5.99130929e-01 -6.40906453e-01 -4.94625807e-01
-1.18179386e-02 1.30175024e-01 6.18284464e-01 -1.00883555e+00
4.61992621e-01 -8.46246541e-01 -2.13013347e-02 -6.30389929e-01
4.01861191e-01 -1.28951788e+00 2.82059342e-01 6.95482492e-01
-3.72350961e-01 6.58680424e-02 4.46941376e-01 7.54855096e-01
1.97876468e-01 6.38934970e-02 6.06294990e-01 2.95820236e-01
-6.28546715e-01 7.47865558e-01 -8.95710707e-01 -4.05970700e-02
1.49233878e+00 5.53741232e-02 -5.53870559e-01 -6.28141761e-01
-5.89285731e-01 6.34510159e-01 5.32454193e-01 7.82394290e-01
8.65012765e-01 -1.37678218e+00 -4.11122650e-01 9.56587419e-02
-1.92801058e-01 -2.57129297e-02 3.71241361e-01 3.66394699e-01
-6.10667706e-01 7.27574944e-01 -2.84561336e-01 -2.43474364e-01
-7.86196351e-01 1.08590531e+00 8.28713179e-01 5.18412516e-02
-7.32693017e-01 4.92217243e-01 1.49407536e-01 -7.92897224e-01
4.51165050e-01 -4.58063275e-01 -2.61812489e-02 -6.60348535e-01
6.77406728e-01 7.54143000e-01 -2.80492365e-01 -8.46304119e-01
-3.81980866e-01 3.79098132e-02 1.04580432e-01 1.38745740e-01
1.00821555e+00 1.48210838e-01 3.49995822e-01 -4.77027744e-02
7.49592721e-01 -3.70658748e-02 -1.86549091e+00 2.07595542e-01
-3.50544393e-01 -3.05607110e-01 -1.42612219e-01 -6.33234203e-01
-1.08988214e+00 5.76619089e-01 2.98766196e-01 4.12916332e-01
7.17256963e-01 -4.56180990e-01 1.31526554e+00 6.41933143e-01
5.42824447e-01 -7.97053516e-01 -3.02159697e-01 7.89561391e-01
9.57756400e-01 -9.37672079e-01 -6.08917117e-01 1.09570427e-02
-8.80300939e-01 1.01623952e+00 9.48500633e-01 -5.31256497e-01
5.59636652e-01 3.25003415e-01 1.84529111e-01 1.75069928e-01
-6.41602337e-01 1.42466620e-01 1.99854195e-01 9.07067597e-01
-6.06110811e-01 4.18441057e-01 5.16022921e-01 6.49890125e-01
-2.84523010e-01 -1.53526753e-01 6.13682985e-01 6.52130663e-01
-3.64788562e-01 -7.27057517e-01 -4.09850389e-01 1.43230364e-01
-2.52567291e-01 2.55333751e-01 -3.15628946e-01 7.76138186e-01
1.43729178e-02 1.15811312e+00 -7.66797364e-02 -7.46971190e-01
5.34088135e-01 -2.85027444e-01 -1.04105033e-01 -9.36319083e-02
-6.61675632e-01 -7.23914146e-01 8.49793628e-02 -8.03490460e-01
9.56842378e-02 -5.12715459e-01 -1.44402862e+00 -8.70440304e-01
2.62101233e-01 -1.65712237e-01 3.15616161e-01 1.00249076e+00
3.87947887e-01 5.28197408e-01 6.28513575e-01 -1.12709343e+00
-6.85914397e-01 -4.57078129e-01 -1.09913893e-01 1.54158279e-01
7.04489589e-01 -6.33437753e-01 -2.76303053e-01 -1.62706703e-01] | [5.426179885864258, 7.817234516143799] |
0e4dc39a-4151-4d5b-b02d-1e719f93c5b5 | wastewater-pipe-rating-model-using-natural | 2202.13871 | null | https://arxiv.org/abs/2202.13871v2 | https://arxiv.org/pdf/2202.13871v2.pdf | Wastewater Pipe Rating Model Using Natural Language Processing | Closed-circuit video (CCTV) inspection has been the most popular technique for visually evaluating the interior status of pipelines in recent decades. Certified inspectors prepare the pipe repair document based on the CCTV inspection. The traditional manual method of assessing sewage structural conditions from pipe repair documents takes a long time and is prone to human mistakes. The automatic identification of necessary texts has received little attention. By building an automated framework employing Natural Language Processing (NLP), this study presents an effective technique to automate the identification of the pipe defect rating of the pipe repair documents. NLP technologies are employed to break down textual material into grammatical units in this research. Further analysis entails using words to discover pipe defect symptoms and their frequency and then combining that information into a single score. Our model achieves 95.0% accuracy,94.9% sensitivity, 94.4% specificity, 95.9% precision score, and 95.7% F1 score, showing the potential of the proposed model to be used in large-scale pipe repair documents for accurate and efficient pipeline failure detection to improve the quality of the pipeline. Keywords: Sewer pipe inspection, Defect detection, Natural language processing, Text recognition | ['Dr. Hongfang Lu', 'Dr. John C. Mattews', 'Shashank Reddy Vadyala', 'Sai Nethra Betgeri'] | 2022-02-22 | null | null | null | null | ['defect-detection'] | ['computer-vision'] | [ 2.20599115e-01 -3.85714360e-02 4.14495409e-01 -2.68854916e-01
-6.00400209e-01 -7.41510451e-01 -3.64893228e-02 9.27945375e-01
-3.34387422e-02 1.20651238e-01 2.53755152e-01 -6.98332846e-01
-2.21580386e-01 -1.02834761e+00 -2.61433750e-01 -7.97880948e-01
2.63068765e-01 1.37125686e-01 4.37188178e-01 8.42562094e-02
8.10027659e-01 6.15757346e-01 -1.26915824e+00 5.67781806e-01
1.01669502e+00 6.83747828e-01 9.22146678e-01 1.05555379e+00
-2.64049321e-01 1.06200206e+00 -9.08082068e-01 -2.29266346e-01
-2.01115534e-01 -2.66310811e-01 -1.10984921e+00 5.10273993e-01
-1.92149788e-01 -5.44957221e-01 9.73761007e-02 9.86371279e-01
4.02969420e-01 -2.93553263e-01 7.05159068e-01 -6.16401732e-01
-7.64134407e-01 4.43808347e-01 -5.88639081e-01 1.99534178e-01
9.03800488e-01 1.60256669e-01 7.61304498e-01 -8.32036197e-01
1.80105388e-01 1.20249474e+00 7.10101008e-01 -1.21066324e-01
-6.27418876e-01 -8.18773508e-02 -4.97471929e-01 2.84819245e-01
-1.16218615e+00 1.92404464e-01 3.67610812e-01 -7.86136985e-01
1.03751743e+00 2.34713152e-01 7.34256089e-01 7.57474303e-02
4.34071273e-01 7.46083558e-01 8.24015677e-01 -9.18363810e-01
-8.81148055e-02 1.01310082e-01 5.19675970e-01 9.73177850e-01
2.88177878e-01 -5.72534204e-01 -4.94477078e-02 -4.58413213e-02
5.11216402e-01 1.68077201e-01 -1.95451543e-01 3.86955917e-01
-7.77539730e-01 9.07765090e-01 1.19056113e-01 4.98567402e-01
-3.55611682e-01 -4.15868223e-01 7.50951767e-01 1.57074630e-01
2.70052642e-01 2.58319080e-01 -3.80120814e-01 -8.52089003e-02
-5.16497731e-01 -2.14310482e-01 6.51472509e-01 9.17186081e-01
3.80057484e-01 -1.29791528e-01 -3.87242109e-01 8.73750806e-01
9.09927309e-01 6.90935552e-01 2.17653558e-01 -7.37555027e-01
5.33668280e-01 9.54520822e-01 1.09168932e-01 -1.16201520e+00
-4.59497780e-01 2.44864196e-01 -5.37043989e-01 -1.24885589e-02
3.05312723e-01 -4.53564748e-02 -9.98392344e-01 3.67221862e-01
9.46272910e-02 -5.94959557e-01 1.99224293e-01 5.41445315e-01
9.39675987e-01 1.19235075e+00 1.89396977e-01 -1.21982388e-01
1.90636849e+00 -5.32244682e-01 -8.64681005e-01 2.14933082e-01
7.92662561e-01 -1.18228209e+00 1.15686512e+00 6.42486334e-01
-6.85070395e-01 -4.34215486e-01 -7.42493153e-01 4.87016514e-02
5.58146881e-03 7.53348589e-01 2.79457688e-01 9.26355779e-01
-8.35019946e-01 2.69732535e-01 -8.55576754e-01 -6.39050007e-01
3.59549433e-01 2.49625593e-01 -3.50933105e-01 -4.89418834e-01
-6.76852524e-01 9.24673915e-01 3.41978103e-01 5.71306527e-01
-7.76386797e-01 -3.82397085e-01 -1.04784226e+00 1.49808303e-01
-5.04003419e-03 -2.16812398e-02 1.29551351e+00 -2.03042533e-02
-1.08407402e+00 7.82127559e-01 -2.22440913e-01 -2.87707634e-02
-3.21360677e-02 -3.02553475e-01 -1.72161281e-01 8.03548336e-01
2.65852839e-01 -1.92366451e-01 4.34835792e-01 -9.73493218e-01
-9.61018920e-01 -3.22905660e-01 -1.66057602e-01 -2.00044781e-01
-3.80395234e-01 7.20206976e-01 -2.09587246e-01 -3.43496114e-01
4.77205247e-01 -4.30144429e-01 5.23079261e-02 -7.72804320e-02
-2.41391584e-01 -9.71288860e-01 8.94595683e-01 -1.40612137e+00
1.12726688e+00 -2.08138919e+00 -6.02691054e-01 2.42107674e-01
-6.96404502e-02 5.34105718e-01 3.15171897e-01 7.99548149e-01
1.29463002e-01 1.12395130e-01 -2.05792993e-01 8.15082863e-02
-1.92851335e-01 2.21096843e-01 1.10428296e-01 4.97549057e-01
4.42581594e-01 4.03618276e-01 -1.08473516e+00 -7.35916853e-01
3.45263451e-01 3.66269112e-01 -2.10281834e-02 3.81941020e-01
4.29439127e-01 2.64487386e-01 -6.65570199e-01 9.85431790e-01
6.28891647e-01 -9.87781063e-02 2.92032827e-02 -3.57615054e-01
-3.91184002e-01 1.58524856e-01 -1.06574118e+00 1.11434698e+00
-3.12017620e-01 7.89107382e-01 -1.11709319e-01 -1.10247934e+00
1.28788006e+00 7.60472238e-01 3.98392648e-01 -4.40078348e-01
2.08170891e-01 8.33495110e-02 -4.75885034e-01 -1.79551172e+00
3.89381558e-01 6.73415661e-02 4.86034676e-02 2.48534843e-01
-1.14237584e-01 -2.48551011e-01 6.50889277e-01 1.83795780e-01
1.07433248e+00 -1.68716237e-01 2.09706545e-01 -1.91210225e-01
6.20841920e-01 3.65651548e-01 2.28217855e-01 5.22626936e-01
-1.15848910e-02 6.64997637e-01 5.96263707e-01 -3.20018411e-01
-1.15899229e+00 -7.70970523e-01 -1.24929100e-01 4.88294423e-01
-2.32192904e-01 -5.05449772e-02 -7.76050508e-01 -3.83449972e-01
-1.19227163e-01 5.76794803e-01 -2.03063011e-01 5.35085686e-02
-5.00197649e-01 -7.97206819e-01 5.53266227e-01 5.54593623e-01
3.48817885e-01 -1.22031927e+00 -5.87097645e-01 2.46338204e-01
-3.47126573e-01 -9.18432772e-01 -3.73445898e-02 -1.26401812e-01
-1.03856170e+00 -1.45018876e+00 -7.65592337e-01 -1.56449378e+00
1.20315182e+00 4.69605803e-01 7.68109500e-01 5.95097303e-01
-5.34246147e-01 2.86473542e-01 -9.56929266e-01 -2.70882457e-01
-9.28967476e-01 -4.58619773e-01 -4.04001892e-01 -2.40128875e-01
4.44496334e-01 2.32755214e-01 -5.65500319e-01 2.97289580e-01
-9.29113090e-01 -3.81505907e-01 5.90623677e-01 6.72718585e-01
1.57825798e-01 5.40592909e-01 4.86099362e-01 -6.29275143e-01
7.90785968e-01 -3.40075999e-01 -6.45572782e-01 5.23410320e-01
-4.62136924e-01 -3.82224739e-01 4.54555929e-01 1.31215662e-01
-1.31137407e+00 7.85916075e-02 -6.06500208e-01 4.57651496e-01
-5.59321046e-01 7.36515939e-01 5.77147566e-02 1.96160942e-01
4.69852686e-01 1.32556438e-01 -4.96023409e-02 -8.56324136e-01
-4.02630240e-01 1.21234977e+00 5.74835658e-01 -1.91694841e-01
5.10418952e-01 1.38393298e-01 -3.01567823e-01 -1.08624470e+00
-6.66726887e-01 -1.14744401e+00 -7.59417534e-01 -4.48645234e-01
1.22621679e+00 -7.30895638e-01 -1.08617055e+00 7.41239667e-01
-1.58475423e+00 1.23321019e-01 4.34587523e-02 6.85945511e-01
2.68344045e-01 1.17680323e+00 -9.65278089e-01 -1.25397837e+00
-6.42756939e-01 -9.78720248e-01 9.68806624e-01 3.11403275e-01
2.31322404e-02 -7.28682160e-01 -1.94814399e-01 6.66072011e-01
-5.66087887e-02 5.80291152e-02 9.60359275e-01 -2.57011652e-01
-8.32055807e-02 -5.23441195e-01 -5.54592550e-01 7.21392751e-01
5.11210442e-01 3.91581684e-01 -6.64381862e-01 -3.09615601e-02
2.82352567e-01 -4.79088016e-02 6.34532511e-01 4.70596939e-01
7.36987352e-01 -2.87100106e-01 -1.64455533e-01 -1.71581477e-01
1.49993706e+00 6.94173515e-01 9.15194452e-01 3.12729210e-01
6.89336300e-01 9.49959099e-01 1.06518340e+00 5.56901813e-01
3.63670439e-01 -1.20060064e-01 5.01773417e-01 -1.67611301e-01
2.01790079e-01 -1.81105100e-02 3.19702208e-01 9.44005132e-01
-1.27521202e-01 -4.35572714e-01 -1.29462326e+00 9.36705768e-01
-1.29989672e+00 -8.32944334e-01 -1.09083259e+00 1.75431240e+00
6.23509824e-01 7.99359232e-02 -2.66367286e-01 9.30417657e-01
7.53457248e-01 -5.55194378e-01 3.04597586e-01 -5.38719416e-01
2.04286531e-01 -1.27245300e-02 4.13459063e-01 3.03193063e-01
-9.43604887e-01 2.14462444e-01 5.50498390e+00 5.51006675e-01
-5.68993270e-01 -6.03528842e-02 3.65183920e-01 8.23642850e-01
-3.74109410e-02 -1.19076215e-01 -6.65134072e-01 4.02580678e-01
6.87977076e-01 5.08244336e-01 -1.99323222e-01 3.91229659e-01
9.63650167e-01 -6.50650084e-01 -5.19182146e-01 5.69402039e-01
7.84831867e-02 -1.02353168e+00 -1.00834623e-01 -4.62246388e-01
4.95026320e-01 -3.35912794e-01 -5.35433769e-01 -4.15790856e-01
1.09460391e-02 -6.28728449e-01 6.51258707e-01 4.77290928e-01
5.02555668e-01 -5.66457927e-01 1.42318916e+00 2.19137520e-01
-1.17917848e+00 -4.04469341e-01 -4.90393370e-01 -1.03830263e-01
5.20089447e-01 6.76107764e-01 -1.26989901e+00 5.52693307e-01
1.06046176e+00 6.06155455e-01 -6.22801363e-01 1.55276704e+00
-4.25149173e-01 8.96601737e-01 -1.61068961e-01 -3.10176879e-01
2.32548773e-01 -1.73111781e-01 5.01232706e-02 1.43774641e+00
5.48098564e-01 1.73909858e-01 7.60243982e-02 3.47981930e-01
4.46172386e-01 3.71729821e-01 -3.90471488e-01 -4.93611358e-02
1.97353691e-01 8.87216151e-01 -1.18966639e+00 -2.23773181e-01
-7.28850603e-01 6.54548585e-01 -4.60460991e-01 2.26111356e-02
-4.63031054e-01 -7.99340725e-01 -2.69524604e-01 7.52634043e-03
4.50973958e-01 -2.14076653e-01 -4.43897098e-01 -4.10612047e-01
2.00356811e-01 -4.76015776e-01 5.09565473e-01 -1.08884108e+00
-1.10530210e+00 4.72266167e-01 -4.63600941e-02 -1.37424803e+00
4.15414900e-01 -7.91626394e-01 -6.55568719e-01 8.90006721e-01
-1.43918788e+00 -1.01257229e+00 -4.54576015e-01 2.40897059e-01
9.98695850e-01 2.35412028e-02 7.55596280e-01 3.46905112e-01
-5.82181036e-01 -4.67906222e-02 2.98803389e-01 5.27172387e-01
2.20962778e-01 -1.27588773e+00 1.30552202e-01 9.70951080e-01
-4.74474967e-01 3.18403721e-01 8.15718532e-01 -8.42924237e-01
-1.32046545e+00 -7.73907304e-01 1.55853951e+00 -1.02287389e-01
5.03354013e-01 1.23286411e-01 -9.73219633e-01 8.63652527e-02
2.99274117e-01 -4.96971786e-01 5.65406024e-01 -6.33306205e-01
2.96449423e-01 -5.30786030e-02 -1.30397666e+00 -4.21733689e-03
-5.08316681e-02 -4.52985168e-01 -8.54239941e-01 7.64419734e-01
5.89432180e-01 -1.60900325e-01 -1.23283935e+00 5.78321703e-02
4.48104471e-01 -4.12465423e-01 7.17941880e-01 -8.43297020e-02
5.50260067e-01 -5.68121433e-01 3.76618892e-01 -6.77643180e-01
-2.59618580e-01 -2.19392642e-01 6.45406246e-01 1.49904943e+00
5.87287366e-01 -4.43289280e-01 4.27332103e-01 4.67375994e-01
-5.99734306e-01 -3.01997900e-01 -4.61300850e-01 -2.49820009e-01
-6.16886020e-01 -4.60907370e-01 3.03149492e-01 5.94389141e-01
3.06538492e-01 -2.58051604e-02 -1.18247896e-01 7.82952607e-01
3.45930517e-01 -2.03121558e-01 1.13937274e-01 -1.22919905e+00
1.20570011e-01 5.56073785e-02 -4.07106191e-01 -9.00858819e-01
-6.86091125e-01 -3.96008104e-01 4.81660008e-01 -2.36872220e+00
2.58816659e-01 -2.12636545e-01 1.30903721e-01 5.34860432e-01
-9.64634642e-02 2.97341555e-01 -1.98804364e-02 1.10716954e-01
-1.40881047e-01 9.61729325e-03 1.30606294e+00 -2.55013287e-01
-1.62517816e-01 3.09833854e-01 -4.34211075e-01 5.40816307e-01
9.67044234e-01 -8.48982036e-01 -2.01645955e-01 -7.62255967e-01
4.46638554e-01 1.23716190e-01 3.21175367e-01 -4.96566862e-01
3.27357441e-01 3.48860398e-02 1.06166124e-01 -8.63544881e-01
-4.39857334e-01 -1.05959320e+00 -1.51177227e-01 9.36858237e-01
1.03513347e-02 2.69248605e-01 1.38724551e-01 4.50013816e-01
-2.09479988e-01 -1.27598679e+00 4.13493425e-01 -1.57956764e-01
-6.11906826e-01 -5.05507529e-01 -1.20473111e+00 -6.15359724e-01
8.65421712e-01 -4.01717901e-01 -3.32556129e-01 6.51042163e-02
-5.59629142e-01 4.00912881e-01 1.64408647e-02 1.11327305e-01
1.01950753e+00 -7.29869783e-01 -9.28915441e-01 1.49753928e-01
9.69058946e-02 1.03655994e-01 3.60733271e-01 9.68526959e-01
-1.75782025e+00 3.70987356e-01 -2.88861059e-02 -8.26340437e-01
-1.65390563e+00 2.57480621e-01 -1.67399585e-01 -2.60393351e-01
-8.53418529e-01 5.80555260e-01 -2.70718396e-01 1.12687044e-01
1.91902652e-01 -8.08179021e-01 -1.03922796e+00 4.97651435e-02
6.26251101e-01 7.71826208e-01 2.88422465e-01 -6.43088937e-01
-2.39929900e-01 8.59824002e-01 2.58673579e-01 4.13014829e-01
1.49656057e+00 -3.87776405e-01 -5.06892443e-01 1.09901369e-01
9.77306545e-01 -1.74406633e-01 -8.43075812e-01 1.82986259e-01
3.69849563e-01 -6.07204795e-01 -1.69463262e-01 -6.55904949e-01
-1.01494551e+00 8.63327861e-01 6.79621518e-01 8.23887348e-01
1.27487957e+00 2.30379432e-01 6.73841655e-01 5.02200842e-01
-6.92072809e-02 -1.02237916e+00 -6.91341683e-02 3.29285920e-01
7.21091092e-01 -1.28584230e+00 8.59364495e-02 -7.29979277e-01
-6.47398889e-01 1.50577569e+00 1.25583276e-01 1.37126774e-01
4.86257434e-01 1.75759181e-01 8.72348025e-02 -6.31639659e-01
-5.38813546e-02 1.01705313e-01 1.04128890e-01 7.69483328e-01
5.31151235e-01 -1.00615963e-01 -7.28343904e-01 3.75944912e-01
2.57430226e-01 -2.19484955e-01 7.34012008e-01 1.40116251e+00
-9.39627767e-01 -8.32717478e-01 -8.76406848e-01 5.62179685e-01
-9.02937651e-01 1.38218582e-01 1.81047812e-01 3.36281270e-01
-9.33499485e-02 1.75031126e+00 -1.06239021e-02 4.82373312e-02
7.68449128e-01 -2.18713060e-01 2.91763619e-02 -7.83849359e-01
-7.57298410e-01 3.08959763e-02 1.78340569e-01 1.16785996e-01
-5.87568045e-01 -8.40121865e-01 -1.49630725e+00 -1.27907926e-02
-6.75883055e-01 4.41282779e-01 7.38352895e-01 1.15275681e+00
-1.99528515e-01 5.04128098e-01 7.99615741e-01 -3.74610752e-01
-2.76964366e-01 -1.25986099e+00 -6.16182506e-01 3.36404234e-01
4.11836624e-01 -2.61495203e-01 -1.49781078e-01 7.39653766e-01] | [7.481616973876953, 1.5880951881408691] |
5ce36c50-4d6c-44d0-812c-871ffbc95b29 | scene-flow-from-point-clouds-with-or-without | 2011.00320 | null | https://arxiv.org/abs/2011.00320v1 | https://arxiv.org/pdf/2011.00320v1.pdf | Scene Flow from Point Clouds with or without Learning | Scene flow is the three-dimensional (3D) motion field of a scene. It provides information about the spatial arrangement and rate of change of objects in dynamic environments. Current learning-based approaches seek to estimate the scene flow directly from point clouds and have achieved state-of-the-art performance. However, supervised learning methods are inherently domain specific and require a large amount of labeled data. Annotation of scene flow on real-world point clouds is expensive and challenging, and the lack of such datasets has recently sparked interest in self-supervised learning methods. How to accurately and robustly learn scene flow representations without labeled real-world data is still an open problem. Here we present a simple and interpretable objective function to recover the scene flow from point clouds. We use the graph Laplacian of a point cloud to regularize the scene flow to be "as-rigid-as-possible". Our proposed objective function can be used with or without learning---as a self-supervisory signal to learn scene flow representations, or as a non-learning-based method in which the scene flow is optimized during runtime. Our approach outperforms related works in many datasets. We also show the immediate applications of our proposed method for two applications: motion segmentation and point cloud densification. | ['Simon Lucey', 'James Hays', 'Jhony Kaesemodel Pontes'] | 2020-10-31 | null | null | null | null | ['motion-segmentation'] | ['computer-vision'] | [ 1.37428492e-01 -2.00433061e-01 -2.75678247e-01 -4.57214624e-01
-4.08249915e-01 -7.24703312e-01 4.14704442e-01 1.27952099e-01
-3.68295372e-01 4.31035370e-01 -9.13427919e-02 -2.10015789e-01
-7.67207295e-02 -6.64022803e-01 -8.32565606e-01 -6.87439442e-01
-3.01067561e-01 6.88769042e-01 5.59545994e-01 1.16210423e-01
5.04849613e-01 9.31623280e-01 -1.42068779e+00 -5.74590191e-02
8.37952614e-01 5.71085811e-01 3.41627032e-01 9.76288319e-01
-4.80884820e-01 8.86668205e-01 -2.65580297e-01 1.34494618e-01
5.78597605e-01 -2.42904663e-01 -1.14987695e+00 4.01399881e-01
7.73426235e-01 -5.52187599e-02 -5.34924388e-01 9.27756190e-01
1.35953218e-01 4.16333765e-01 7.11362839e-01 -1.43177521e+00
-2.02795431e-01 -1.30319327e-01 -6.32612705e-01 4.77407247e-01
2.79143631e-01 3.72152984e-01 7.11920619e-01 -6.04772151e-01
8.13444972e-01 1.06291759e+00 4.09733832e-01 3.99703771e-01
-1.34342265e+00 -3.09021413e-01 2.28556156e-01 2.05213100e-01
-1.25700414e+00 -3.20262641e-01 1.25104260e+00 -9.36389804e-01
8.12428832e-01 2.10351065e-01 7.64547706e-01 4.59346861e-01
-2.51930077e-02 7.95313060e-01 8.25059116e-01 -2.34483808e-01
3.83320272e-01 9.85871181e-02 -5.79692163e-02 9.22171950e-01
1.97216719e-01 -6.21343330e-02 -3.19908142e-01 1.11679314e-03
1.27186549e+00 5.64954393e-02 -3.46638501e-01 -8.20561290e-01
-1.46008039e+00 7.05022275e-01 5.75711548e-01 1.34045273e-01
-2.43851706e-01 3.97374213e-01 2.45712504e-01 7.59818405e-02
7.53688633e-01 3.54804575e-01 -5.52581191e-01 -2.99086273e-01
-1.10964608e+00 2.88802058e-01 8.44479263e-01 9.56073463e-01
1.35256267e+00 9.48922932e-02 3.29586476e-01 3.50507170e-01
3.54904622e-01 5.20103514e-01 3.16398710e-01 -1.40594506e+00
5.28940380e-01 7.42546856e-01 1.83651567e-01 -1.35434937e+00
-3.46035361e-01 -1.36115655e-01 -7.83517957e-01 4.37127709e-01
6.32516980e-01 3.07190721e-03 -7.95348644e-01 1.39654422e+00
6.75025046e-01 9.72761095e-01 -1.33377463e-01 1.17131615e+00
5.38628340e-01 9.32163060e-01 -1.86235577e-01 -2.51127452e-01
5.56223989e-01 -9.80277300e-01 -5.31755447e-01 -2.66499281e-01
7.81695187e-01 -6.63719058e-01 1.09665954e+00 1.51318282e-01
-9.96140659e-01 -5.68044305e-01 -7.35466659e-01 -1.09487116e-01
-1.92467004e-01 -1.60418391e-01 8.01242173e-01 4.89712805e-01
-1.01713896e+00 6.41963542e-01 -1.26014733e+00 -2.92520165e-01
4.72774327e-01 4.05309170e-01 -4.83962834e-01 2.62831226e-02
-3.42326254e-01 6.07653856e-01 2.37872273e-01 1.38788760e-01
-7.92305470e-01 -1.06923592e+00 -1.02949786e+00 -1.90148279e-01
3.19315791e-01 -8.08430135e-01 9.84927595e-01 -8.51697505e-01
-1.51938021e+00 9.57754016e-01 -4.57514763e-01 -1.12340793e-01
5.62674880e-01 -3.49427909e-01 1.13491632e-01 3.39060068e-01
2.91461766e-01 6.49413943e-01 8.72016907e-01 -1.50317764e+00
-6.48492634e-01 -2.51885235e-01 -4.15058471e-02 3.48239243e-01
5.76634407e-02 -3.43715072e-01 -5.52412927e-01 -3.96204770e-01
3.37494433e-01 -1.12918150e+00 -4.78220224e-01 3.44415545e-01
-2.85036713e-01 -9.93735865e-02 1.42098534e+00 -2.58534878e-01
8.75324070e-01 -1.86273062e+00 5.50786667e-02 8.24020877e-02
6.56020343e-02 2.05873668e-01 9.13388729e-02 1.26314806e-02
-9.78831127e-02 -2.41474155e-02 -6.39063895e-01 -2.81844974e-01
-4.53250915e-01 4.87753749e-01 -3.19521457e-01 9.62483048e-01
3.64700615e-01 7.18308389e-01 -1.29745591e+00 -7.92183042e-01
8.77540231e-01 4.19640511e-01 -6.06144965e-01 4.28110838e-01
-3.02216679e-01 9.73783374e-01 -5.08243680e-01 3.55029672e-01
6.25801921e-01 -5.08754611e-01 -3.20885479e-01 -2.97471374e-01
-2.81502038e-01 3.58346626e-02 -1.61422670e+00 2.12184262e+00
-3.73231977e-01 9.40163255e-01 -4.16483581e-02 -1.28082371e+00
8.84695292e-01 -1.05569594e-01 1.12225175e+00 -1.44902855e-01
-4.83349226e-02 -1.12026505e-01 -3.95678133e-01 -8.31102490e-01
2.63200879e-01 7.15946313e-03 1.92687020e-01 4.18823302e-01
-3.37610836e-03 -6.62661254e-01 1.85281500e-01 1.93764076e-01
1.07752550e+00 3.74373913e-01 1.01677850e-01 -3.48629147e-01
8.12093019e-01 4.47424442e-01 6.31552577e-01 5.08783638e-01
-2.07887948e-01 7.50495493e-01 1.65840745e-01 -7.17237532e-01
-9.37915087e-01 -9.14940238e-01 -9.16928239e-03 6.53690279e-01
7.25516558e-01 -1.74018845e-01 -5.16857505e-01 -7.52806425e-01
2.48445943e-02 3.55109155e-01 -2.34418839e-01 2.10843459e-01
-1.04094923e+00 -4.80962008e-01 1.25551252e-02 3.95412445e-01
4.70183492e-01 -8.48144174e-01 -8.81407917e-01 1.74294129e-01
-2.70379245e-01 -1.46944273e+00 -6.89411759e-01 -7.30019137e-02
-1.19498384e+00 -1.33066964e+00 -4.85571146e-01 -7.13087022e-01
9.68972206e-01 5.19123554e-01 1.04270208e+00 -1.72385778e-02
-3.84559661e-01 8.46672475e-01 1.89225193e-05 -1.02744348e-01
-2.63033718e-01 -4.07325625e-02 2.80997436e-02 4.36271191e-01
6.02755882e-02 -5.96263885e-01 -6.93542838e-01 3.86123657e-01
-8.53501141e-01 6.88346326e-02 2.52545401e-02 3.81723613e-01
8.96518886e-01 7.88372457e-02 4.36420403e-02 -9.44973886e-01
1.30140319e-01 -3.14744979e-01 -7.73782730e-01 -3.44446599e-02
-3.22675437e-01 2.01207355e-01 5.22397220e-01 -3.78462881e-01
-1.14264035e+00 8.09429765e-01 1.12954207e-01 -9.38457251e-01
-4.14040923e-01 6.85317740e-02 7.92318862e-03 -4.67884541e-01
6.43359423e-01 6.95597893e-03 -1.37510970e-01 -2.75380999e-01
6.86076760e-01 2.38622531e-01 8.41251254e-01 -6.69035256e-01
1.03315890e+00 1.08986354e+00 3.63515794e-01 -1.10613060e+00
-8.01538169e-01 -1.22524905e+00 -1.23943388e+00 -5.58075190e-01
1.03280020e+00 -6.90439463e-01 -6.82277262e-01 3.06727111e-01
-1.39132714e+00 -6.86630130e-01 -5.76635778e-01 6.02176785e-01
-9.19553876e-01 5.43164134e-01 -3.60200733e-01 -6.71792686e-01
-8.46980736e-02 -1.06060910e+00 1.17706227e+00 2.07765937e-01
-3.47131081e-02 -1.50220788e+00 3.88629645e-01 2.07843423e-01
-2.56250682e-03 6.11261010e-01 4.05750632e-01 -1.17868312e-01
-9.41795111e-01 6.55123144e-02 -1.42530128e-01 2.01266646e-01
4.75420892e-01 1.81548744e-01 -1.02382743e+00 -1.92013219e-01
1.50066033e-01 -8.02060496e-03 5.52111864e-01 6.67791188e-01
1.24988711e+00 -2.58667767e-01 -4.27360564e-01 1.02055383e+00
1.56226945e+00 -8.02978426e-02 3.42949897e-01 1.35907754e-01
1.23354387e+00 7.35275030e-01 5.94688118e-01 3.85139465e-01
3.97686332e-01 5.81311107e-01 5.28965056e-01 -2.14675456e-01
-1.99216962e-01 -3.67156118e-01 8.09211470e-03 9.23193157e-01
-9.82700959e-02 4.17414121e-02 -1.19640505e+00 6.23790383e-01
-2.08246112e+00 -1.00563419e+00 -5.40673912e-01 2.20490527e+00
4.35231507e-01 7.56616071e-02 -1.14997886e-02 1.59538552e-01
5.77568054e-01 2.51463026e-01 -6.97802663e-01 -1.31568229e-02
2.91620623e-02 -1.52193516e-01 6.57203138e-01 9.23975408e-01
-1.37199962e+00 1.11626637e+00 6.25271559e+00 1.90689653e-01
-1.46862435e+00 2.66282167e-02 4.03935134e-01 1.75899997e-01
-6.93836361e-02 2.38450065e-01 -6.23954415e-01 3.66030365e-01
6.05543017e-01 -2.43935257e-01 9.87849832e-02 8.81388426e-01
5.51703274e-01 -1.13403082e-01 -1.29245508e+00 1.37291443e+00
9.38102081e-02 -1.72745466e+00 -1.27556026e-01 -7.16217235e-03
8.02572489e-01 1.93002790e-01 -3.51687878e-01 -1.71628803e-01
2.42586434e-01 -6.41052008e-01 4.66569543e-01 4.99868035e-01
6.61466897e-01 -3.00819606e-01 2.18557984e-01 6.31635725e-01
-1.47504199e+00 1.68325678e-01 -2.60324776e-01 -6.65853322e-02
5.64321995e-01 6.84821665e-01 -7.95605063e-01 3.21342856e-01
7.04103887e-01 1.20384371e+00 -4.90038097e-01 1.41725051e+00
-2.58959290e-02 6.65896773e-01 -6.04473591e-01 3.45744312e-01
2.43978351e-01 -5.33818603e-01 7.07003832e-01 1.42404473e+00
1.24560399e-02 1.42424747e-01 5.62704563e-01 9.00520146e-01
1.68872118e-01 -6.18829345e-03 -8.47142160e-01 2.98130453e-01
7.70275444e-02 1.26385546e+00 -1.20473897e+00 -3.61840606e-01
-2.43804708e-01 9.31744754e-01 1.91676408e-01 5.82424104e-01
-4.70321953e-01 -7.38955438e-02 7.14095354e-01 4.27691817e-01
1.23666064e-03 -8.43770146e-01 -3.48847866e-01 -1.40671468e+00
-1.82654485e-01 -7.91755989e-02 2.32808456e-01 -7.12335765e-01
-1.21144545e+00 2.61361897e-01 2.75691986e-01 -1.55028784e+00
-2.90039599e-01 -4.90637869e-01 -7.05027759e-01 5.23407996e-01
-1.71616197e+00 -7.98978388e-01 -6.92785680e-01 9.16286170e-01
8.18461537e-01 4.98137884e-02 2.73219824e-01 2.72497654e-01
-3.09400290e-01 -1.26888976e-01 -1.70603991e-01 2.16603696e-01
4.92218614e-01 -1.35736346e+00 2.26336002e-01 9.17450905e-01
4.95205253e-01 6.86748251e-02 7.31769383e-01 -6.30374372e-01
-1.60683811e+00 -1.36086881e+00 4.61381108e-01 -7.22132087e-01
5.55837452e-01 -3.72603059e-01 -1.20703077e+00 7.25325525e-01
-2.87665457e-01 8.02671194e-01 2.86423028e-01 -4.42053378e-01
7.01908022e-02 -3.19690317e-01 -9.74830329e-01 1.96557641e-01
1.09096706e+00 -5.19683301e-01 -4.00524229e-01 5.58213174e-01
8.34554553e-01 -6.62372708e-01 -5.97559810e-01 2.81297565e-01
-7.07223043e-02 -7.44401515e-01 1.10541797e+00 -5.25583982e-01
1.47486746e-01 -6.44254446e-01 1.97365388e-01 -1.22504163e+00
-1.67884201e-01 -8.91106308e-01 -3.12946260e-01 8.45671833e-01
-7.35131651e-02 -3.46868008e-01 1.40748131e+00 6.92480087e-01
-2.01487914e-01 -3.70412439e-01 -8.72959316e-01 -7.15958655e-01
-1.04474969e-01 -6.80285871e-01 1.37838066e-01 1.35615242e+00
-3.14016700e-01 2.61539429e-01 -4.13559489e-02 5.36051869e-01
1.00015628e+00 1.37490824e-01 1.18826330e+00 -1.23713827e+00
-8.52573384e-03 -2.77037472e-01 -7.64611900e-01 -1.48670805e+00
5.94619453e-01 -9.19812322e-01 8.75751823e-02 -1.58086050e+00
-2.62341797e-01 -8.06723654e-01 3.25854301e-01 2.53284633e-01
1.76759101e-02 -1.22922644e-01 2.07771912e-01 5.49305141e-01
-5.20230174e-01 4.62935656e-01 1.37320471e+00 -1.92995220e-01
-6.35992944e-01 1.85022444e-01 6.18328415e-02 9.34201181e-01
7.40347385e-01 -4.85749424e-01 -7.93923974e-01 -5.41541874e-01
-1.01499647e-01 1.20790325e-01 3.84123743e-01 -9.64830518e-01
6.87567770e-01 -5.73286533e-01 1.12630568e-01 -6.05925918e-01
3.49340439e-01 -1.09844232e+00 5.76371215e-02 2.23801702e-01
-1.84210032e-01 1.29123032e-01 2.80927848e-02 7.79103577e-01
-2.17678726e-01 -1.52534872e-01 7.19011068e-01 -2.52226055e-01
-9.55536067e-01 6.81910992e-01 -6.18123002e-02 2.96588898e-01
1.20059574e+00 -4.64837402e-01 7.03318219e-04 -2.93324918e-01
-4.16984975e-01 1.41179830e-01 4.01895374e-01 4.34336245e-01
7.84597039e-01 -1.03842032e+00 -6.17161691e-01 3.39048654e-01
-7.83447400e-02 8.75444353e-01 -2.04137824e-02 4.77154166e-01
-1.12601376e+00 2.18983307e-01 -2.36267727e-02 -1.33147240e+00
-9.53872740e-01 5.39593399e-01 5.25289357e-01 4.20725346e-02
-1.02596998e+00 5.91488421e-01 3.34323317e-01 -4.67805207e-01
1.29051298e-01 -3.42166752e-01 -2.17700545e-02 -3.77783239e-01
2.16537029e-01 5.37948191e-01 -1.32012188e-01 -8.88981044e-01
-3.27279001e-01 1.02105927e+00 3.71523708e-01 -6.29270216e-03
1.16347969e+00 -2.70472199e-01 1.73483193e-02 7.26375103e-01
1.52506638e+00 -3.13739598e-01 -1.69289839e+00 -2.03184590e-01
1.38412669e-01 -9.06931520e-01 2.50589490e-01 7.37212002e-02
-1.35819221e+00 1.05994129e+00 5.01357853e-01 1.66131064e-01
9.07721698e-01 1.34858772e-01 6.00116968e-01 3.55438501e-01
3.76023710e-01 -8.50392520e-01 2.29360193e-01 3.84550631e-01
5.80253065e-01 -1.44552112e+00 1.01147629e-01 -7.93656409e-01
-4.95073348e-01 1.06806231e+00 6.88512743e-01 -4.55172867e-01
9.70540404e-01 3.00309598e-01 1.22240961e-01 -2.04761848e-01
-4.52093482e-01 -4.86916341e-02 3.95368785e-01 7.72569954e-01
1.95421576e-01 -2.82060236e-01 2.89469063e-01 -4.99270648e-01
-1.44129366e-01 1.06241196e-01 6.21816456e-01 1.05628395e+00
-3.51957947e-01 -8.04205120e-01 -4.05231595e-01 2.97278017e-01
4.79956083e-02 5.21794796e-01 -1.69181358e-03 6.30832493e-01
1.74105447e-02 8.07992339e-01 4.61726993e-01 5.44767780e-03
5.10329664e-01 -3.14968556e-01 4.45794523e-01 -6.44306064e-01
-1.61571264e-01 -2.12892704e-02 -4.79852617e-01 -8.90715718e-01
-1.01320517e+00 -6.96095943e-01 -1.47973990e+00 -1.85751155e-01
-5.78880534e-02 2.90430114e-02 7.46943176e-01 8.83098722e-01
2.38920361e-01 1.27528802e-01 8.80581379e-01 -1.35009134e+00
5.13890758e-02 -2.72437841e-01 -3.72777581e-01 7.15829849e-01
7.01233268e-01 -5.43095589e-01 -5.88702798e-01 7.94390023e-01] | [8.510150909423828, -2.010786771774292] |
312305b4-3c8f-478c-a7ed-5d6c24f1faa3 | enhanced-sampling-with-machine-learning-a | 2306.09111 | null | https://arxiv.org/abs/2306.09111v2 | https://arxiv.org/pdf/2306.09111v2.pdf | Enhanced Sampling with Machine Learning: A Review | Molecular dynamics (MD) enables the study of physical systems with excellent spatiotemporal resolution but suffers from severe time-scale limitations. To address this, enhanced sampling methods have been developed to improve exploration of configurational space. However, implementing these is challenging and requires domain expertise. In recent years, integration of machine learning (ML) techniques in different domains has shown promise, prompting their adoption in enhanced sampling as well. Although ML is often employed in various fields primarily due to its data-driven nature, its integration with enhanced sampling is more natural with many common underlying synergies. This review explores the merging of ML and enhanced MD by presenting different shared viewpoints. It offers a comprehensive overview of this rapidly evolving field, which can be difficult to stay updated on. We highlight successful strategies like dimensionality reduction, reinforcement learning, and flow-based methods. Finally, we discuss open problems at the exciting ML-enhanced MD interface. | ['Pratyush Tiwary', 'Ziyue Zou', 'Lukas Herron', 'Zachary Smith', 'Shams Mehdi'] | 2023-06-15 | null | null | null | null | ['dimensionality-reduction'] | ['methodology'] | [ 5.69043420e-02 -3.69949520e-01 -4.77763832e-01 2.02076554e-01
-6.56896293e-01 -5.25718927e-01 6.10567331e-01 2.59650618e-01
-5.50502598e-01 1.31293488e+00 -1.67964339e-01 -3.56315523e-01
-2.98055947e-01 -6.64480031e-01 -4.07721400e-01 -1.42621326e+00
-2.76744545e-01 4.32103276e-01 1.66202158e-01 -3.65963161e-01
6.42788827e-01 1.00576627e+00 -1.59413493e+00 -9.08853412e-02
9.70593810e-01 5.99786758e-01 2.95641243e-01 6.77776337e-01
-1.74457088e-01 2.58288413e-01 -3.28677595e-01 1.06963448e-01
-8.09683353e-02 -4.96902376e-01 -7.42764652e-01 -2.87628055e-01
6.50051534e-02 -7.17485324e-03 -4.05299604e-01 5.83116174e-01
8.01843405e-01 5.73922276e-01 6.13211572e-01 -9.36689019e-01
-5.49471229e-02 -1.86518624e-01 -2.35741332e-01 5.58213413e-01
3.74179661e-01 5.56538641e-01 7.24983871e-01 -7.60454118e-01
9.15937662e-01 1.01574278e+00 3.39328855e-01 4.74056780e-01
-1.47005498e+00 -4.14756566e-01 1.94417536e-01 3.19771737e-01
-1.10530424e+00 -2.37470716e-01 9.34428513e-01 -4.45678145e-01
1.23510695e+00 1.34757161e-02 1.00460899e+00 1.01993704e+00
5.30011714e-01 4.70129848e-01 1.51919377e+00 -6.36472046e-01
6.65563643e-01 8.96410942e-02 -1.03436582e-01 3.42379540e-01
3.54836076e-01 8.29060912e-01 -6.14305973e-01 -3.40126067e-01
7.94298410e-01 -1.64604351e-01 -1.75757825e-01 -1.01974416e+00
-9.93819118e-01 1.03555954e+00 1.44569963e-01 2.91690618e-01
-3.96602005e-01 -3.02818771e-02 4.96719867e-01 1.12611502e-01
2.19423503e-01 9.65608358e-01 -4.42739576e-01 -5.30576050e-01
-8.62058222e-01 9.58009660e-01 7.63712525e-01 3.54982615e-01
5.75482547e-01 -2.89134439e-02 4.48319405e-01 4.03856844e-01
7.45166242e-02 3.29795122e-01 2.39316933e-02 -1.29054677e+00
1.60292163e-01 2.03936249e-01 3.61180007e-01 -5.11504352e-01
-6.47665918e-01 -3.18564177e-01 -8.59270573e-01 9.20359254e-01
4.56113935e-01 -3.14234942e-01 -1.68169066e-01 1.59974504e+00
6.81032956e-01 -2.45287061e-01 1.16241612e-02 8.71006489e-01
3.74642581e-01 5.88204443e-01 1.58193439e-01 -5.32732248e-01
8.67479920e-01 -6.85602725e-01 -7.67992437e-01 2.32735097e-01
6.45289242e-01 -4.50100452e-01 7.39334166e-01 5.76024354e-01
-1.12066174e+00 -3.59823674e-01 -1.11356938e+00 1.07162535e-01
-7.56185114e-01 -4.80518043e-01 8.33513081e-01 5.46625197e-01
-7.61255264e-01 1.33172524e+00 -1.25633967e+00 -6.10202193e-01
2.39910588e-01 5.15044034e-01 -1.91752955e-01 3.59574974e-01
-1.17769718e+00 1.15429997e+00 1.75167724e-01 -3.28747451e-01
-5.60269058e-01 -8.12779784e-01 -6.36807501e-01 -5.36893666e-01
4.67427790e-01 -7.67416596e-01 9.84113932e-01 -7.94742182e-02
-1.94157672e+00 4.27863032e-01 -3.44942302e-01 -3.81167948e-01
6.45501852e-01 3.60452496e-02 -2.67705619e-01 3.27996701e-01
-1.11349247e-01 5.29820383e-01 4.76735741e-01 -8.60431731e-01
-2.58219808e-01 -3.32455516e-01 7.84748606e-03 2.62499779e-01
6.87476099e-02 -9.87093970e-02 2.73874462e-01 -4.27040935e-01
-1.38291299e-01 -8.88327777e-01 -7.55421042e-01 5.52957766e-02
4.54163663e-02 -1.58732608e-01 8.76934052e-01 -1.56077713e-01
1.30260396e+00 -1.59505200e+00 5.24112940e-01 -1.15120403e-01
3.80302101e-01 4.54744011e-01 2.67140925e-01 1.18483579e+00
-3.88687626e-02 3.17906141e-02 -3.43625456e-01 -7.80190825e-02
-2.48449311e-01 -2.66136508e-02 -1.99660838e-01 5.05447328e-01
3.44762564e-01 8.89157474e-01 -1.02661550e+00 -3.52204353e-01
7.56239712e-01 6.90312088e-01 -5.49441397e-01 1.00308605e-01
-2.52708793e-01 1.20851052e+00 -5.67471206e-01 5.08507371e-01
6.51821733e-01 -3.64802569e-01 2.90184140e-01 -1.00604035e-01
-7.85652995e-01 2.91184187e-01 -1.30372703e+00 1.50871432e+00
-1.22929074e-01 4.04983103e-01 2.83178419e-01 -1.20869398e+00
7.15774715e-01 2.11917415e-01 9.18237627e-01 -8.18291068e-01
-2.62282401e-01 3.48694175e-01 2.00029120e-01 -7.59082973e-01
4.35322881e-01 -4.47926223e-01 1.60357431e-01 5.78409612e-01
-2.45574355e-01 -4.55400050e-01 1.35659307e-01 6.66316375e-02
8.79094481e-01 5.94045758e-01 6.62746310e-01 -4.75156873e-01
5.53465307e-01 3.85738939e-01 1.66878745e-01 8.06542516e-01
-6.23318374e-01 1.04863539e-01 3.15918595e-01 -5.80131769e-01
-1.30102181e+00 -7.70640194e-01 -6.24386430e-01 5.31903863e-01
2.80558974e-01 -6.81670308e-01 -4.11144793e-01 -3.47970337e-01
8.89312923e-02 1.47731200e-01 -5.62976897e-01 -4.57834899e-02
-8.80213559e-01 -1.03254879e+00 3.19647528e-02 3.15396130e-01
2.64761865e-01 -1.13100529e+00 -8.65098596e-01 5.67000687e-01
2.94868380e-01 -7.49742746e-01 2.40932152e-01 3.11857253e-01
-1.17099941e+00 -1.10517848e+00 -7.21182048e-01 -2.84332305e-01
1.02160431e-01 3.40644121e-01 9.05140579e-01 -1.50978312e-01
-5.82720280e-01 2.34830692e-01 -2.77962595e-01 7.05102906e-02
-4.35777158e-01 3.22600245e-01 4.53239322e-01 -7.12628186e-01
2.20671609e-01 -8.84096384e-01 -7.44741499e-01 2.08893195e-01
-7.55633652e-01 -7.29116201e-02 3.68015945e-01 9.76074219e-01
5.88212669e-01 -1.42670885e-01 8.18512022e-01 -3.85977209e-01
6.03774846e-01 -3.53994608e-01 -6.33050978e-01 -1.80518731e-01
-7.82255948e-01 3.67604375e-01 6.38357759e-01 -3.44010472e-01
-9.56651092e-01 -2.35486656e-01 -3.78805727e-01 6.61231065e-03
-5.09631038e-01 3.63936245e-01 4.87562157e-02 -5.33916533e-01
6.05639756e-01 1.37942672e-01 5.06552935e-01 -6.68329716e-01
2.95507312e-01 3.89777809e-01 -1.27291843e-01 -9.18766379e-01
5.15059531e-01 6.08431756e-01 4.85308886e-01 -1.12547779e+00
-2.90262669e-01 -2.16988653e-01 -9.29375708e-01 -9.90523696e-02
7.65326977e-01 -3.58080894e-01 -8.61175954e-01 3.11316252e-01
-7.65645385e-01 -4.84640449e-01 -3.41400534e-01 6.85549498e-01
-7.34740376e-01 5.80111027e-01 -5.77569962e-01 -8.76254380e-01
6.09003156e-02 -1.45847619e+00 7.32278824e-01 2.32686251e-01
-3.34885776e-01 -1.30972767e+00 5.76713204e-01 1.18556872e-01
7.38583386e-01 5.52637339e-01 9.66304958e-01 -2.17653319e-01
-6.82781816e-01 -4.07528952e-02 1.83971673e-01 -1.60600618e-01
1.89848885e-01 2.47272208e-01 -8.49507868e-01 -5.39980412e-01
-6.95663765e-02 -3.80068243e-01 7.90845811e-01 6.62743747e-01
9.72041607e-01 1.86627805e-01 -7.30640471e-01 4.36237663e-01
1.26056123e+00 5.57632625e-01 1.89640582e-01 5.03598094e-01
2.87850440e-01 8.17846537e-01 7.74378419e-01 5.94006836e-01
8.70235041e-02 1.00664592e+00 2.47909755e-01 -2.35416412e-01
8.64590798e-03 9.17183012e-02 1.84626117e-01 7.06559896e-01
-3.14249039e-01 1.23540051e-01 -6.44544184e-01 -7.89456293e-02
-1.65311360e+00 -1.12456882e+00 -7.44761378e-02 2.16638541e+00
6.79115117e-01 3.16251144e-02 5.88414967e-01 1.58750415e-01
4.23364937e-01 2.75624782e-01 -9.87476170e-01 -5.07953167e-01
-2.54459560e-01 3.05692434e-01 5.88126592e-02 6.96934581e-01
-9.98969138e-01 8.20989847e-01 7.30180168e+00 7.30581343e-01
-1.40361583e+00 -5.12527749e-02 3.41510355e-01 -2.33816475e-01
1.17597654e-02 6.83955178e-02 -9.49985683e-01 5.30160725e-01
9.34411764e-01 -1.52930930e-01 3.82413149e-01 6.45316839e-01
7.53512502e-01 -3.57735932e-01 -8.00045192e-01 7.50560582e-01
-5.05530477e-01 -1.86009943e+00 -2.31369376e-01 5.18208981e-01
6.40131891e-01 -1.95922609e-02 3.03575415e-02 2.00081676e-01
-2.65312761e-01 -8.30036640e-01 1.88121065e-01 3.57372612e-01
7.64240682e-01 -8.41225147e-01 5.07001281e-01 2.96993911e-01
-1.06802952e+00 2.15344474e-01 -1.37755543e-01 -6.24882340e-01
4.73301023e-01 5.13656557e-01 -2.85265893e-01 4.81605917e-01
4.74860966e-01 7.39775777e-01 -1.04334675e-01 1.12991345e+00
3.52025777e-01 2.26254627e-01 -3.04539859e-01 -3.71368855e-01
2.34252498e-01 -7.41880000e-01 7.92240739e-01 9.74174500e-01
1.00839250e-01 7.85292163e-02 9.54357013e-02 9.17803586e-01
5.96790254e-01 -8.09064135e-02 -8.08383882e-01 -1.64595619e-01
4.91288930e-01 1.19717669e+00 -5.66381097e-01 -1.75883919e-01
-4.62931246e-01 4.87331718e-01 2.74375141e-01 3.41817766e-01
-7.64824390e-01 -2.82840937e-01 1.11142850e+00 2.55313486e-01
1.44912526e-01 -7.10440099e-01 4.79809344e-02 -1.14101267e+00
-1.86426327e-01 -8.14781189e-01 1.32619396e-01 -9.65758190e-02
-1.11127079e+00 3.37656230e-01 3.12698245e-01 -1.09713018e+00
-3.02922457e-01 -8.41137707e-01 -3.72087151e-01 6.95001841e-01
-1.60562527e+00 -3.75625372e-01 3.95132974e-02 1.78885609e-01
3.99209261e-01 4.10555974e-02 8.45532596e-01 1.99881598e-01
-8.18877161e-01 7.00983703e-02 6.08012259e-01 -5.70979655e-01
7.32297063e-01 -1.22980034e+00 3.13889533e-01 3.07084739e-01
-4.86165583e-01 6.66359365e-01 8.74248087e-01 -6.64116681e-01
-1.71688890e+00 -6.30244315e-01 4.96652782e-01 -3.87409300e-01
5.55449665e-01 -2.10905075e-01 -1.06633604e+00 -6.08211905e-02
1.69178516e-01 -1.04296766e-01 6.55316293e-01 -8.24410617e-02
1.41853422e-01 1.71148986e-01 -1.14262617e+00 5.37016988e-01
9.69601691e-01 -4.72548187e-01 -3.74109507e-01 3.29665571e-01
2.92693764e-01 -2.50171363e-01 -1.18576527e+00 3.52014244e-01
5.46357274e-01 -1.17501926e+00 1.11642098e+00 -8.38469148e-01
9.56954211e-02 -2.60877639e-01 1.75581187e-01 -1.24093378e+00
-3.52258503e-01 -8.86806488e-01 -5.08112788e-01 7.84209847e-01
9.07871425e-02 -9.25899029e-01 9.30136383e-01 5.98291039e-01
-5.51815331e-02 -1.17832732e+00 -1.04590404e+00 -9.16072667e-01
6.72210038e-01 -1.97340012e-01 4.31741327e-01 7.75925040e-01
5.57368875e-01 4.33861054e-02 -1.53298393e-01 -2.07394719e-01
6.94510937e-01 3.50181013e-01 5.62232196e-01 -1.16029978e+00
-2.75868595e-01 -8.03053975e-01 -1.09251969e-01 -9.27296758e-01
1.45371156e-02 -6.04587555e-01 -5.19682944e-01 -1.27127182e+00
1.24815568e-01 -4.72092122e-01 -1.79425217e-02 -2.11984858e-01
1.34575963e-02 1.05929533e-02 -7.38240927e-02 3.41966510e-01
-4.84921694e-01 8.14245820e-01 1.52306736e+00 3.89225036e-01
-6.15272880e-01 -1.06460467e-01 -3.28180343e-01 3.27723294e-01
1.07045293e+00 -2.06044406e-01 -5.69845438e-02 3.28176916e-01
2.21628919e-01 1.56474888e-01 2.54798919e-01 -1.10657072e+00
9.16151553e-02 -4.47921842e-01 2.24564716e-01 -6.12484336e-01
5.81046820e-01 -2.82905161e-01 4.84038442e-02 7.51070440e-01
-2.16836661e-01 2.51461744e-01 2.62061775e-01 5.98723292e-01
-1.55862600e-01 3.07739489e-02 1.06909621e+00 -4.35906291e-01
-5.09859145e-01 3.40359926e-01 -8.16565990e-01 1.64370760e-02
1.05675220e+00 -3.33826452e-01 -2.46304348e-01 -1.99233964e-01
-8.19369614e-01 4.08566520e-02 8.49215209e-01 -9.42445025e-02
3.25374961e-01 -1.04506135e+00 -1.40744239e-01 2.28658512e-01
-3.41780595e-02 -2.72331774e-01 4.45976675e-01 1.26088297e+00
-4.88321632e-01 8.68454337e-01 -3.22781354e-01 -5.29491782e-01
-9.09094274e-01 7.66638815e-01 4.46111262e-01 -3.87154400e-01
-6.45163059e-01 2.39640623e-01 -9.86869410e-02 -4.46866482e-01
-6.76570013e-02 -4.07504402e-02 -1.53371304e-01 4.80764210e-02
4.76199925e-01 8.73815477e-01 9.04458538e-02 -3.69431525e-01
-4.19202566e-01 8.75010848e-01 1.95304677e-02 -1.56848088e-01
1.36380863e+00 -2.64592201e-01 -2.09725723e-02 2.82941103e-01
9.45140064e-01 -1.27437487e-01 -1.65808284e+00 2.52138637e-03
3.47841270e-02 -1.94760680e-01 1.48369774e-01 -4.69167173e-01
-5.23126841e-01 1.33554614e+00 5.02347827e-01 2.55404353e-01
7.07844675e-01 -4.08877768e-02 6.03888333e-01 3.28139782e-01
6.25884891e-01 -1.00000966e+00 1.44946858e-01 4.16110665e-01
6.37293458e-01 -1.34023273e+00 1.81964785e-01 -1.61041826e-01
-2.72003502e-01 1.40291178e+00 5.96531391e-01 -1.81859881e-01
7.10176051e-01 3.89761388e-01 -3.46914381e-01 -1.56081572e-01
-8.52508307e-01 -1.75796345e-01 -2.43406296e-01 8.27332675e-01
5.15369356e-01 -2.23522812e-01 -5.58530152e-01 -7.36745372e-02
1.94300801e-01 6.79049781e-03 3.38603288e-01 1.31932986e+00
-4.87528771e-01 -1.77898574e+00 -2.81854779e-01 2.71203697e-01
-1.45246893e-01 1.74032837e-01 -1.23802952e-01 1.13452518e+00
-1.54699842e-02 6.76035702e-01 -2.15296373e-01 2.44990997e-02
-5.08046448e-02 1.38838992e-01 7.82746434e-01 -2.17943683e-01
-4.38678563e-01 1.03940971e-01 -7.90228844e-02 -7.50565588e-01
-6.64197624e-01 -1.03203750e+00 -1.06380641e+00 -5.60214341e-01
-1.99940309e-01 4.52544957e-01 6.20031834e-01 1.03753293e+00
6.51293933e-01 3.06986630e-01 4.09394175e-01 -1.40627360e+00
-5.66772699e-01 -5.33576131e-01 -5.33836901e-01 6.50052801e-02
7.51617193e-01 -1.10001910e+00 -4.82447505e-01 -4.36641902e-01] | [6.258604526519775, 3.707108736038208] |
3e872c9e-0541-436d-aac0-b62a761de4d8 | refining-query-representations-for-dense | 2205.12680 | null | https://arxiv.org/abs/2205.12680v3 | https://arxiv.org/pdf/2205.12680v3.pdf | Optimizing Test-Time Query Representations for Dense Retrieval | Recent developments of dense retrieval rely on quality representations of queries and contexts from pre-trained query and context encoders. In this paper, we introduce TOUR (Test-Time Optimization of Query Representations), which further optimizes instance-level query representations guided by signals from test-time retrieval results. We leverage a cross-encoder re-ranker to provide fine-grained pseudo labels over retrieval results and iteratively optimize query representations with gradient descent. Our theoretical analysis reveals that TOUR can be viewed as a generalization of the classical Rocchio algorithm for pseudo relevance feedback, and we present two variants that leverage pseudo-labels as hard binary or soft continuous labels. We first apply TOUR on phrase retrieval with our proposed phrase re-ranker, and also evaluate its effectiveness on passage retrieval with an off-the-shelf re-ranker. TOUR greatly improves end-to-end open-domain question answering accuracy, as well as passage retrieval performance. TOUR also consistently improves direct re-ranking by up to 2.0% while running 1.3-2.4x faster with an efficient implementation. | ['Jinhyuk Lee', 'Danqi Chen', 'Jaewoo Kang', 'Jungsoo Park', 'Mujeen Sung'] | 2022-05-25 | null | null | null | null | ['passage-retrieval'] | ['natural-language-processing'] | [ 9.11727622e-02 -3.33491832e-01 -4.67632204e-01 -4.64390695e-01
-2.14978600e+00 -9.21164930e-01 6.26895368e-01 4.55392510e-01
-6.72885537e-01 5.29232264e-01 4.87893254e-01 -1.59861416e-01
-4.15159404e-01 -6.29789770e-01 -8.92623007e-01 -1.06517211e-01
-1.27936780e-01 7.08353758e-01 1.79966182e-01 -4.42433029e-01
4.25792873e-01 -1.53114215e-01 -1.39952731e+00 7.05799401e-01
7.30706513e-01 1.15770888e+00 1.97196692e-01 1.10500681e+00
-5.23815081e-02 5.90696573e-01 -7.10768819e-01 -3.54572803e-01
8.26647580e-02 -1.05989270e-01 -1.09069586e+00 -5.23717403e-01
8.18422318e-01 -4.22921062e-01 -5.58704436e-01 5.61598003e-01
6.28473103e-01 3.53469998e-01 5.45627654e-01 -6.75771773e-01
-1.44112384e+00 6.19062245e-01 -6.82637617e-02 4.14621502e-01
6.74176216e-01 -2.39533693e-01 1.84712696e+00 -1.09120917e+00
4.56176072e-01 1.20807409e+00 2.47723043e-01 4.80034173e-01
-1.29312313e+00 -3.92756432e-01 -1.39353042e-02 3.29184055e-01
-1.57871139e+00 -3.70824456e-01 3.16548973e-01 1.05099097e-01
1.32424545e+00 5.81880391e-01 1.60619169e-01 1.07732129e+00
-2.81720813e-02 1.17458391e+00 6.27446949e-01 -5.84053814e-01
4.87716570e-02 6.66700825e-02 5.81930459e-01 6.00207210e-01
-1.22445181e-01 -1.96973607e-02 -5.82882881e-01 -5.40766835e-01
2.70786285e-01 9.65316966e-02 -5.70261836e-01 -2.35246830e-02
-9.98808861e-01 9.33854938e-01 8.31121862e-01 6.24448843e-02
-2.24419981e-01 6.10100210e-01 2.86404133e-01 7.01804042e-01
4.24528092e-01 8.34147394e-01 -6.16948724e-01 -9.93161276e-02
-1.01460433e+00 4.56980318e-01 8.39806020e-01 1.02146220e+00
8.79132926e-01 -5.45922816e-01 -9.42620337e-01 1.27370608e+00
4.54714775e-01 7.83122301e-01 7.13534176e-01 -8.37310255e-01
4.68295246e-01 3.32125187e-01 2.27452993e-01 -7.33743012e-01
-2.08504517e-02 -6.11390412e-01 -1.73972830e-01 -8.29082429e-01
-8.68077502e-02 2.99630821e-01 -1.02725232e+00 1.67112589e+00
-1.49017602e-01 9.23389941e-02 1.42375425e-01 9.44485784e-01
9.41392720e-01 8.50406110e-01 2.68856417e-02 2.18640193e-01
1.34490001e+00 -1.44113243e+00 -5.11803746e-01 -7.97674358e-02
7.96347857e-01 -7.93207526e-01 1.50894451e+00 1.40341043e-01
-1.06349385e+00 -5.15447021e-01 -8.67192805e-01 -6.76569879e-01
-4.47297513e-01 2.57002711e-01 4.44734603e-01 3.43619823e-01
-1.35317969e+00 4.51527089e-01 -5.69299877e-01 -1.52121022e-01
7.06812739e-02 2.94876724e-01 -6.19174056e-02 -5.34683526e-01
-1.45479035e+00 5.11567891e-01 -3.30950553e-03 -6.60553202e-02
-9.97296453e-01 -9.77671087e-01 -7.59272099e-01 4.26098078e-01
2.48562887e-01 -9.82633889e-01 1.75130475e+00 -2.96537727e-01
-1.37460792e+00 8.50581586e-01 -5.32757163e-01 -4.32626665e-01
-1.83317408e-01 -8.85692239e-01 -4.47812110e-01 5.37025869e-01
2.89367139e-01 7.71951139e-01 6.91310644e-01 -8.97739828e-01
-4.08134818e-01 -8.74494538e-02 4.17119056e-01 3.50152373e-01
-2.96538025e-01 -1.91339850e-01 -1.07459438e+00 -5.88985026e-01
-7.83596840e-03 -8.69966328e-01 1.04462236e-01 -2.48450994e-01
-3.04382861e-01 -6.89665079e-01 4.37139183e-01 -6.50909126e-01
1.55402744e+00 -1.99805510e+00 -1.33662447e-01 1.32922068e-01
6.28925264e-02 4.68600802e-02 -6.34401083e-01 6.70527399e-01
3.22117805e-01 2.62377501e-01 -7.78447390e-02 -4.92855012e-01
3.77287209e-01 1.14253156e-01 -7.76761413e-01 6.65264577e-02
2.82718271e-01 1.58347607e+00 -1.12389743e+00 -3.74585271e-01
-2.41504014e-01 4.28813845e-01 -7.64931500e-01 4.33272958e-01
-5.66475272e-01 -2.19609588e-01 -6.99533880e-01 8.21195424e-01
8.85356292e-02 -7.24653780e-01 -2.45617390e-01 3.18510756e-02
6.12120092e-01 9.20299888e-01 -4.63472277e-01 2.30184555e+00
-7.98538029e-01 6.89710319e-01 -1.78854004e-01 -6.36158824e-01
7.53738999e-01 3.41901422e-01 2.33458415e-01 -1.31361270e+00
-3.99610490e-01 3.78318787e-01 -8.47575963e-01 -4.88520294e-01
1.12624073e+00 3.09077501e-01 -3.26435298e-01 6.40563548e-01
2.12400928e-01 -9.24169496e-02 1.81364492e-01 5.71926236e-01
1.37877190e+00 6.26190305e-02 -2.84808964e-01 -1.64007396e-02
3.63637775e-01 8.79035331e-03 -7.75189549e-02 1.26174355e+00
2.11569309e-01 8.14640284e-01 6.76468611e-02 1.10991657e-01
-6.94397926e-01 -1.20369458e+00 -2.12786153e-01 1.85028851e+00
2.09656239e-01 -5.86999416e-01 -1.07373871e-01 -7.12244391e-01
3.22890610e-01 5.83742440e-01 -4.74190205e-01 -4.39851701e-01
-5.46148360e-01 -5.28976083e-01 6.37935042e-01 4.44362044e-01
9.69160646e-02 -1.00204456e+00 -1.65725708e-01 2.98038930e-01
-2.56808132e-01 -8.44472170e-01 -7.73435831e-01 1.44550860e-01
-8.22965622e-01 -9.54737902e-01 -1.06410456e+00 -8.85507584e-01
2.98241347e-01 5.79994202e-01 1.78942764e+00 4.06073362e-01
-2.59478360e-01 9.69928205e-01 -8.36568832e-01 1.49745584e-01
9.53878164e-02 5.33216417e-01 -3.94065201e-01 -2.84766763e-01
5.29845834e-01 -3.20257664e-01 -1.09824455e+00 1.97976649e-01
-1.18000889e+00 -6.50137961e-01 6.87605143e-01 1.14178693e+00
8.21860313e-01 -7.67851353e-01 8.65759075e-01 -7.46242344e-01
1.17887533e+00 -7.22966850e-01 -5.48105121e-01 8.76416504e-01
-9.07516062e-01 6.08089685e-01 1.64919540e-01 -2.91241288e-01
-5.93835831e-01 -6.89965606e-01 -9.11248699e-02 -5.44223666e-01
3.05455834e-01 7.61196792e-01 5.49881101e-01 2.50530779e-01
9.11009908e-01 1.46318004e-01 -5.45610011e-01 -6.87192142e-01
8.48528385e-01 7.85288513e-01 4.18025583e-01 -8.42220247e-01
4.52718377e-01 1.36861175e-01 -5.78118443e-01 -3.41321170e-01
-1.16009343e+00 -1.13803792e+00 3.52214165e-02 3.34090948e-01
4.72190261e-01 -1.05492878e+00 -5.56357980e-01 -3.42318833e-01
-8.69341850e-01 -4.49827969e-01 -3.33897442e-01 3.36420685e-01
-3.09241772e-01 2.16408700e-01 -8.77580523e-01 -5.24148881e-01
-9.03289497e-01 -9.03564095e-01 2.01136422e+00 1.42753065e-01
5.28988093e-02 -9.17646050e-01 3.68681014e-01 4.48074490e-01
6.73725188e-01 -4.95282531e-01 9.61772859e-01 -7.88066983e-01
-8.94695640e-01 -4.67545152e-01 -4.57367003e-01 2.51181155e-01
-2.35387117e-01 -5.47291338e-01 -1.17526269e+00 -5.47505379e-01
-4.74577934e-01 -9.20581222e-01 1.38251078e+00 1.25359252e-01
1.23354936e+00 -2.40548402e-01 -2.35367611e-01 4.55883741e-01
1.47413909e+00 -3.20136428e-01 3.47409487e-01 1.53985262e-01
3.97886455e-01 3.02170038e-01 6.96073949e-01 1.85515732e-01
4.59900290e-01 8.02180409e-01 7.00290799e-02 3.78377885e-01
-2.37987921e-01 -6.52145088e-01 3.42089891e-01 9.42309320e-01
7.02870846e-01 -4.46788490e-01 -6.63130343e-01 7.86922276e-01
-1.73161197e+00 -6.65094435e-01 2.91950971e-01 2.21391225e+00
1.12551033e+00 -1.08733080e-01 -2.10685581e-01 -3.21407974e-01
3.04981530e-01 2.14377567e-01 -6.34771764e-01 -1.93599120e-01
7.90526345e-02 7.73868918e-01 3.09909821e-01 8.51570070e-01
-8.68271410e-01 1.07010365e+00 6.38466024e+00 1.02144563e+00
-9.41150606e-01 1.68426692e-01 4.68469709e-01 -4.79457766e-01
-7.88335919e-01 4.47611585e-02 -8.66222203e-01 8.26385394e-02
1.21292365e+00 -3.38792503e-01 5.60077190e-01 9.67168629e-01
-4.24528390e-01 9.13600400e-02 -1.39896071e+00 1.11796331e+00
3.19163352e-01 -1.28545487e+00 1.58499613e-01 -1.77793413e-01
8.05424035e-01 4.35354829e-01 2.21656337e-01 1.18175209e+00
1.84857309e-01 -1.02321005e+00 3.44460815e-01 6.44029021e-01
8.82526159e-01 -3.86212349e-01 5.86432815e-01 9.16723236e-02
-1.08507907e+00 -8.07393938e-02 -5.57317615e-01 3.16665113e-01
1.03648119e-01 4.38049376e-01 -1.02126288e+00 4.95586365e-01
6.29590809e-01 6.76987708e-01 -8.20984721e-01 1.20570660e+00
-2.54019231e-01 7.17903256e-01 -5.12324035e-01 -2.94924200e-01
3.55407447e-01 3.16378951e-01 2.50361145e-01 1.38803160e+00
1.02418557e-01 -1.84911042e-02 1.29029766e-01 7.77774990e-01
-7.04797506e-01 1.98517650e-01 -2.33326897e-01 -2.06178173e-01
7.09063888e-01 1.01040447e+00 -8.27811752e-03 -6.59981370e-01
-2.07410142e-01 1.15157747e+00 5.36274850e-01 8.63236845e-01
-5.44170976e-01 -5.75928450e-01 4.96082127e-01 -2.65748948e-01
4.82563972e-01 2.87368912e-02 1.73843756e-01 -1.34259474e+00
3.54537427e-01 -7.50410318e-01 5.92702687e-01 -8.58628035e-01
-1.46710348e+00 5.92722654e-01 -2.53087860e-02 -9.46886659e-01
-6.29596472e-01 -3.77911329e-01 -7.22880214e-02 1.08206701e+00
-2.05894542e+00 -8.21401536e-01 -1.72919333e-01 4.72247362e-01
7.11481392e-01 9.18854773e-02 1.26850069e+00 4.81769294e-01
-1.19881608e-01 1.12980807e+00 5.73899508e-01 5.91970012e-02
9.32056308e-01 -1.31111002e+00 4.07558233e-01 1.82133496e-01
8.59058738e-01 9.91445720e-01 1.78224474e-01 -2.74904460e-01
-1.71965539e+00 -9.19596851e-01 1.17763889e+00 -8.04547668e-01
6.51298463e-01 -3.38388950e-01 -9.21133518e-01 4.75174218e-01
1.02055348e-01 1.54953003e-01 7.94665098e-01 6.32209957e-01
-8.71349990e-01 -2.38027692e-01 -8.21646154e-01 5.15629172e-01
7.87864506e-01 -1.37444568e+00 -1.06092918e+00 7.60210276e-01
1.33430493e+00 -3.52414876e-01 -8.38898599e-01 4.84735757e-01
6.27878189e-01 -4.69630271e-01 1.19698215e+00 -7.74237573e-01
1.76674441e-01 3.89372073e-02 -5.91060996e-01 -9.16900098e-01
-1.76571473e-01 -4.94920939e-01 -5.25737107e-01 7.54270256e-01
7.55638421e-01 -3.54320824e-01 6.99485004e-01 5.15539944e-01
-1.52676836e-01 -1.20174122e+00 -7.54718900e-01 -6.64235711e-01
1.42044187e-01 -6.00712180e-01 4.36595201e-01 5.51761031e-01
1.73600558e-02 6.19155765e-01 4.56213206e-02 1.60680071e-01
2.87551880e-01 5.27961850e-01 4.81723875e-01 -7.48407125e-01
-7.43159831e-01 -4.92794573e-01 -8.11128393e-02 -2.03708410e+00
1.53847426e-01 -1.25156331e+00 1.57058716e-01 -1.59965754e+00
1.54417276e-01 -5.54149389e-01 -8.25928688e-01 2.38809139e-01
-5.50654709e-01 3.61683816e-01 -9.07919630e-02 3.72758389e-01
-1.29441190e+00 8.07224691e-01 9.06485975e-01 -3.67106467e-01
-1.29501730e-01 -1.55297488e-01 -7.92779624e-01 -1.74971625e-01
4.45277870e-01 -6.22873545e-01 -5.83518088e-01 -9.95523214e-01
6.18213356e-01 2.24656805e-01 3.80215377e-01 -6.49808049e-01
2.88035810e-01 5.62969029e-01 3.77073027e-02 -6.88246727e-01
6.87975883e-01 -4.59509522e-01 -8.00525784e-01 -9.91378874e-02
-1.24289572e+00 2.28462473e-01 1.23629965e-01 9.67189491e-01
-5.68685234e-01 -3.90062600e-01 2.22112685e-02 -9.03475210e-02
-5.19216299e-01 2.14715630e-01 1.92480296e-01 5.59273183e-01
1.52859420e-01 4.45878506e-01 -4.29829180e-01 -7.11111009e-01
-5.70782542e-01 7.31330693e-01 -6.32443130e-02 6.85655057e-01
7.61676490e-01 -1.34589577e+00 -7.84417033e-01 7.58537427e-02
5.59399903e-01 -1.51341692e-01 8.29560533e-02 3.88304085e-01
-1.74235702e-01 1.21880555e+00 7.73648918e-01 -7.74547279e-01
-9.19887006e-01 4.62773114e-01 3.93932350e-02 -7.12242067e-01
-3.37793916e-01 1.42204440e+00 -3.17378417e-02 -4.10057485e-01
6.26279533e-01 -5.37668288e-01 1.84777728e-03 -1.45817563e-01
6.65627360e-01 3.57864611e-02 2.20859438e-01 5.14204055e-02
-1.43994823e-01 5.09195149e-01 -4.93810982e-01 -5.77828348e-01
9.50908124e-01 -1.63002774e-01 7.61579052e-02 4.03981686e-01
1.92163539e+00 -2.52652299e-02 -6.96124256e-01 -7.48680890e-01
2.82048464e-01 -4.53743905e-01 1.74435690e-01 -1.13460529e+00
-5.42909861e-01 7.89291739e-01 5.52245319e-01 1.69544801e-01
1.16185546e+00 2.60382652e-01 1.13026261e+00 1.14878869e+00
4.43362385e-01 -7.68763304e-01 4.84793961e-01 5.91719627e-01
1.08692026e+00 -1.33775783e+00 -1.46684617e-01 2.47752771e-01
-3.30407232e-01 7.82590985e-01 2.36429065e-01 -3.96305591e-01
5.37985027e-01 -6.35366082e-01 -4.38175723e-02 -3.48454714e-01
-1.22793686e+00 -3.29763532e-01 9.59786475e-01 -4.27011214e-02
5.54183245e-01 -3.58763188e-02 -1.89508080e-01 4.40120846e-01
-2.15628311e-01 -2.00103857e-02 -1.85777038e-01 9.35779631e-01
-5.73593438e-01 -1.12596297e+00 -1.78386614e-01 6.47051454e-01
-5.53715527e-01 -7.01681316e-01 -2.50352085e-01 3.13391775e-01
-6.56062484e-01 1.20469272e+00 5.34280837e-02 -4.24959660e-01
2.84338444e-01 4.15410846e-01 3.54173839e-01 -8.88616145e-01
-6.82834685e-01 -6.47032335e-02 1.85810983e-01 -8.45381677e-01
-5.84175400e-02 -8.01487491e-02 -9.89098608e-01 2.94491738e-01
-5.65387309e-01 8.06594372e-01 6.94714546e-01 6.82508171e-01
8.56622517e-01 3.51071268e-01 7.85086691e-01 -3.57944429e-01
-1.04808092e+00 -1.24644554e+00 -1.33444250e-01 4.24434602e-01
6.03127360e-01 -3.10095906e-01 -4.14154261e-01 -2.44499013e-01] | [11.510311126708984, 7.653527736663818] |
60be9f2b-caa6-4c58-895f-bb6fa9fd33d2 | adversarial-learning-for-chinese-ner-from | 1801.05147 | null | http://arxiv.org/abs/1801.05147v1 | http://arxiv.org/pdf/1801.05147v1.pdf | Adversarial Learning for Chinese NER from Crowd Annotations | To quickly obtain new labeled data, we can choose crowdsourcing as an
alternative way at lower cost in a short time. But as an exchange, crowd
annotations from non-experts may be of lower quality than those from experts.
In this paper, we propose an approach to performing crowd annotation learning
for Chinese Named Entity Recognition (NER) to make full use of the noisy
sequence labels from multiple annotators. Inspired by adversarial learning, our
approach uses a common Bi-LSTM and a private Bi-LSTM for representing
annotator-generic and -specific information. The annotator-generic information
is the common knowledge for entities easily mastered by the crowd. Finally, we
build our Chinese NE tagger based on the LSTM-CRF model. In our experiments, we
create two data sets for Chinese NER tasks from two domains. The experimental
results show that our system achieves better scores than strong baseline
systems. | ['Wei zhang', 'Meishan Zhang', 'Haofen Wang', 'Wenliang Chen', 'Min Zhang', 'YaoSheng Yang'] | 2018-01-16 | null | null | null | null | ['chinese-named-entity-recognition'] | ['natural-language-processing'] | [-3.11386347e-01 1.83036953e-01 2.62251109e-01 -4.59004253e-01
-1.25206661e+00 -9.31898296e-01 2.92005748e-01 -1.52936906e-01
-1.02811742e+00 1.24942374e+00 2.23246813e-01 -1.28257973e-02
1.02630281e+00 -5.31687498e-01 -7.53487766e-01 -5.68523705e-01
4.55054402e-01 6.42862082e-01 3.80818874e-01 -1.34622112e-01
-1.30942106e-01 3.14455092e-01 -6.11772716e-01 4.17694926e-01
1.17117643e+00 7.34638333e-01 3.72579575e-01 4.66860145e-01
-5.50131619e-01 1.15105963e+00 -9.29119349e-01 -8.73674452e-01
2.11897135e-01 -1.74255580e-01 -1.24709773e+00 -7.30143785e-02
-1.31709427e-01 -3.17370355e-01 -1.04437575e-01 1.19797194e+00
7.47244000e-01 2.77973950e-01 1.87513426e-01 -8.86753798e-01
-1.07479680e+00 8.90212536e-01 -2.60031998e-01 -1.55380368e-01
4.27757978e-01 4.85024489e-02 4.80778039e-01 -1.04927075e+00
8.74720216e-01 7.71833420e-01 8.08127761e-01 1.12234426e+00
-5.20872355e-01 -6.54376745e-01 3.10649514e-01 -1.41440228e-01
-1.59473443e+00 -3.06196183e-01 3.85911107e-01 -2.99736381e-01
6.68198645e-01 -1.73420701e-02 8.32865108e-03 1.33771670e+00
-2.95814127e-01 1.12705338e+00 1.16597009e+00 -5.00141382e-01
3.08835894e-01 4.18296695e-01 -1.76971018e-01 4.99323756e-01
-7.67455772e-02 -4.71443266e-01 -2.48322606e-01 -4.25866425e-01
5.19131243e-01 3.40590589e-02 -2.20587775e-01 3.25037897e-01
-1.11048317e+00 5.41946232e-01 2.69832850e-01 7.05291927e-01
-3.41042459e-01 2.01053414e-02 3.93558800e-01 1.31819412e-01
8.03925872e-01 3.68356109e-01 -8.83850276e-01 -2.76799332e-02
-6.68633699e-01 -1.46528706e-02 1.20385230e+00 1.56344604e+00
7.64690638e-01 1.01518072e-01 -5.47950387e-01 6.85073853e-01
1.13379337e-01 6.79912806e-01 6.60440087e-01 -7.81540573e-01
7.26935089e-01 3.81251514e-01 8.62047911e-01 -6.07470632e-01
-1.13174833e-01 3.38178836e-02 -7.97133923e-01 -3.13942045e-01
4.18780655e-01 -9.39419210e-01 -9.19359684e-01 1.63336051e+00
2.76900887e-01 2.26587594e-01 1.78177208e-01 8.06624055e-01
7.51022577e-01 3.57855767e-01 5.68199337e-01 6.90402910e-02
1.31883097e+00 -1.38861477e+00 -1.10539222e+00 -2.66849071e-01
9.74017441e-01 -6.55798733e-01 7.89815187e-01 3.47065143e-02
-7.50729620e-01 -5.76170087e-01 -3.84957790e-01 -2.56674886e-01
-7.39887297e-01 5.58234692e-01 2.35243425e-01 6.60841823e-01
-9.42972600e-01 5.44351995e-01 -7.82983422e-01 -2.52185494e-01
3.01734149e-01 3.85448813e-01 -5.51291227e-01 -4.19355966e-02
-1.62873745e+00 8.95295858e-01 6.00033283e-01 5.38003564e-01
-9.73867357e-01 -1.81282666e-02 -8.85753214e-01 -1.42183170e-01
4.38292384e-01 -2.15512291e-01 1.64805102e+00 -1.06643128e+00
-1.50728595e+00 9.20086145e-01 -4.79961038e-01 -2.77809292e-01
8.10202360e-01 -4.42312628e-01 -6.79765940e-01 -6.07716329e-02
3.98536265e-01 4.91905451e-01 3.47340018e-01 -1.51164055e+00
-7.88338661e-01 -1.61406144e-01 1.36958174e-02 3.60873304e-02
-2.82834977e-01 3.72582346e-01 -4.40464646e-01 -6.54633224e-01
-3.78034860e-01 -9.87410486e-01 -6.79171324e-01 -4.72276300e-01
-5.69385767e-01 -3.18644553e-01 5.05555868e-01 -1.07120395e+00
1.07025599e+00 -1.96880841e+00 -4.77406532e-01 -1.96219951e-01
2.78412640e-01 9.02400494e-01 -1.53971285e-01 4.57232922e-01
3.47160846e-01 6.13876522e-01 -2.59858876e-01 -6.79448664e-01
-6.92683160e-02 2.67511249e-01 -3.00683528e-01 1.24242194e-01
4.87701744e-01 1.11259687e+00 -1.36265731e+00 -6.93698108e-01
-4.66290355e-01 2.57966548e-01 2.32808337e-01 5.86842299e-01
-2.40597695e-01 8.98839414e-01 -9.42743659e-01 5.29285908e-01
8.22200000e-01 -3.29034626e-01 2.88860500e-01 3.99833143e-01
2.21794099e-02 -6.46352172e-02 -1.10467052e+00 1.82685089e+00
-3.21500510e-01 3.06509405e-01 7.75432289e-02 -4.21923757e-01
1.17482865e+00 9.40677106e-01 -1.34853482e-01 -3.29022884e-01
2.89617591e-02 4.87930536e-01 -5.31844020e-01 -7.80158937e-01
7.55225658e-01 -9.64415595e-02 -3.35034728e-01 3.73390436e-01
1.75724044e-01 4.01362896e-01 -2.52942778e-02 3.95871177e-02
1.02217531e+00 3.56701076e-01 2.81119674e-01 1.09932162e-01
6.03111863e-01 1.63498774e-01 1.09250998e+00 9.95571136e-01
-5.42696416e-01 4.48217273e-01 -1.23308050e-02 -5.81353605e-01
-1.01014888e+00 -5.27583480e-01 3.36087316e-01 1.26542246e+00
-9.44444984e-02 6.84532151e-02 -1.09826875e+00 -1.35382390e+00
-5.59166133e-01 4.55547810e-01 -4.01657611e-01 4.29655403e-01
-8.95793200e-01 -2.67838061e-01 1.27378941e+00 9.38197970e-01
8.97893727e-01 -1.30588853e+00 3.77628021e-02 4.16410744e-01
-6.70690298e-01 -1.58889544e+00 -7.78435647e-01 3.49464789e-02
-4.80441600e-01 -5.67692280e-01 -1.32422423e+00 -1.29876888e+00
7.67621696e-01 1.38113707e-01 1.05235112e+00 9.47340503e-02
4.37890053e-01 3.34989391e-02 -7.61113822e-01 -5.92705369e-01
-4.48976666e-01 3.55541259e-01 3.92286368e-02 6.15543537e-02
9.41073835e-01 -7.97796324e-02 -2.72785962e-01 3.93845916e-01
-9.11845684e-01 -3.28242064e-01 3.54531050e-01 6.94884419e-01
3.80684406e-01 -4.77797896e-01 6.76515758e-01 -1.11560011e+00
6.05596662e-01 -5.10954678e-01 -3.24093461e-01 7.12564766e-01
5.62173594e-03 1.86088104e-02 9.76044357e-01 -7.38194466e-01
-1.65333247e+00 4.93266433e-01 -1.39538020e-01 -2.12581515e-01
-5.47349155e-01 2.60414630e-01 -4.51765656e-01 -1.11859731e-01
6.68905735e-01 8.03039894e-02 -8.24995875e-01 -8.35424781e-01
4.85123575e-01 1.05690575e+00 4.80554491e-01 -7.34756529e-01
6.03315115e-01 5.01645148e-01 -8.16004515e-01 -4.41167891e-01
-1.12998354e+00 -4.94042873e-01 -9.20339882e-01 2.86135953e-02
1.26456404e+00 -1.36682999e+00 -6.05006695e-01 6.97927177e-01
-1.74902308e+00 -3.91698450e-01 -1.01945870e-01 4.20065850e-01
5.53670824e-02 5.32358229e-01 -1.10855746e+00 -1.04888785e+00
-4.00470763e-01 -9.85664189e-01 1.05923462e+00 7.18855023e-01
1.46132648e-01 -1.09861565e+00 1.86011031e-01 2.72449255e-01
3.47590029e-01 2.48742834e-01 -5.56289591e-03 -1.54271770e+00
-4.33195323e-01 -5.04877448e-01 -1.16233334e-01 4.74099278e-01
-2.44641915e-01 -4.04972702e-01 -1.33520508e+00 4.98034619e-03
6.76788315e-02 -6.34592533e-01 5.49576521e-01 -2.70244509e-01
7.87317634e-01 -3.62523437e-01 -4.07240778e-01 1.71180904e-01
1.36098158e+00 2.03459829e-01 5.64070642e-01 2.89112478e-01
9.91746426e-01 6.17613912e-01 7.05893278e-01 3.40626687e-01
7.36916721e-01 2.24181011e-01 -2.59241045e-01 -3.36609840e-01
4.68156636e-01 -5.48988283e-01 3.17845404e-01 1.09070718e+00
-3.37829143e-01 -5.51484644e-01 -1.12483716e+00 7.20916271e-01
-2.02600241e+00 -8.00672472e-01 -2.27647498e-01 1.85586727e+00
1.13524437e+00 -3.32318395e-01 -1.97860092e-01 -7.93398976e-01
1.34226930e+00 -6.52590021e-02 -1.96953624e-01 -1.28207847e-01
-3.78471047e-01 1.12560481e-01 7.46233046e-01 3.46870512e-01
-1.24813962e+00 1.43936598e+00 5.80286503e+00 9.43014443e-01
-7.64974892e-01 6.94292724e-01 6.85759068e-01 5.26166618e-01
-1.66009083e-01 1.64590180e-01 -1.15524220e+00 6.16420269e-01
9.18027580e-01 1.14402048e-01 1.03487477e-01 9.77491558e-01
2.89234240e-02 5.22281192e-02 -8.38545203e-01 5.64106762e-01
6.50122240e-02 -9.24699306e-01 -3.37907016e-01 -3.05822724e-03
1.17997336e+00 3.35997730e-01 -5.98100364e-01 5.37668824e-01
1.04607284e+00 -8.35695684e-01 7.04160273e-01 8.70741308e-01
7.28820205e-01 -3.97697151e-01 1.29506409e+00 8.01780224e-01
-1.19496024e+00 1.74646527e-01 -6.07223868e-01 6.39813691e-02
6.02580249e-01 4.00175899e-01 -7.01526761e-01 6.95648551e-01
5.92760921e-01 -2.00746618e-02 -4.05505300e-01 8.91608477e-01
-8.08558285e-01 5.79223275e-01 -1.08698167e-01 -3.02361131e-01
3.05233121e-01 1.60044841e-02 1.30412221e-01 1.55892503e+00
2.18796059e-01 3.27251524e-01 5.98660588e-01 7.20688224e-01
-7.22422838e-01 1.67707160e-01 -6.19794488e-01 -1.84238367e-02
7.73311734e-01 1.45918524e+00 -5.84648907e-01 -7.20709503e-01
-3.83598506e-01 1.48198617e+00 8.11220229e-01 7.22206056e-01
-7.37130642e-01 -6.65122986e-01 -3.33353803e-02 -5.10926187e-01
3.08902711e-01 -2.17405081e-01 4.17265370e-02 -1.56744182e+00
3.72220337e-01 -8.48239660e-01 1.30107969e-01 -7.52436638e-01
-1.63337731e+00 1.12414086e+00 -5.72919786e-01 -1.11380529e+00
-2.44183376e-01 -4.87652659e-01 -6.47529542e-01 1.14638591e+00
-1.45438623e+00 -1.48108435e+00 -2.02589616e-01 5.25973499e-01
3.36718470e-01 -1.21472828e-01 1.02078021e+00 3.91504824e-01
-4.68519747e-01 6.66566312e-01 1.33463487e-01 1.15110517e+00
1.00334632e+00 -1.20663643e+00 8.27646673e-01 9.92288649e-01
3.46137844e-02 6.36134803e-01 2.67333627e-01 -1.14900172e+00
-5.10285974e-01 -1.21062922e+00 1.59530997e+00 -8.49108636e-01
5.79762101e-01 -5.56421459e-01 -1.09281552e+00 9.69682336e-01
3.21941942e-01 2.07855672e-01 9.59908128e-01 -2.83515692e-01
-2.14254141e-01 4.51960474e-01 -1.30606306e+00 3.58752608e-01
8.51629555e-01 -8.18398893e-01 -6.54567897e-01 4.19782639e-01
1.16363323e+00 -4.61932689e-01 -1.02174091e+00 6.31404743e-02
6.49125203e-02 -2.78152019e-01 2.77840167e-01 -9.19533372e-01
1.01303741e-01 -5.75165629e-01 -1.05795804e-02 -1.16994321e+00
-6.42423704e-02 -8.28067601e-01 2.85999000e-01 1.82831788e+00
6.54660046e-01 -4.87845540e-01 5.82616866e-01 1.29688847e+00
4.54640612e-02 -1.50982529e-01 -7.94770300e-01 -9.89327669e-01
1.32371351e-01 -1.31467402e-01 8.01972151e-01 1.36205721e+00
-3.53712924e-02 3.61958861e-01 -7.29652941e-01 2.82350689e-01
1.68043792e-01 -4.88288015e-01 5.71145415e-01 -9.58739400e-01
-8.23808089e-02 3.60500813e-01 1.37555644e-01 -1.02257740e+00
5.63483238e-01 -6.13444567e-01 5.53828955e-01 -1.36661184e+00
2.26398081e-01 -5.22194624e-01 -1.09653130e-01 7.38069952e-01
-5.83256423e-01 1.00872450e-01 3.00101340e-01 3.00705284e-01
-1.11998641e+00 3.24995577e-01 1.05351067e+00 8.92358646e-02
-2.68742681e-01 -1.18962787e-01 -4.44248110e-01 9.40046549e-01
8.18872571e-01 -8.91359150e-01 3.34413320e-01 -1.07083595e+00
1.40642479e-01 4.81160358e-02 -3.10841892e-02 -7.38064051e-01
6.66413367e-01 6.64281333e-03 4.84980017e-01 -1.64593667e-01
-1.24099381e-01 -7.34907150e-01 -1.96484461e-01 6.67543802e-03
-4.73308921e-01 4.61580828e-02 -6.52269796e-02 7.01571822e-01
-2.95700818e-01 -7.69475818e-01 3.42283368e-01 -6.75046086e-01
-7.48888493e-01 3.44712555e-01 -4.44203466e-01 4.38531011e-01
7.90362239e-01 2.20441371e-01 -3.99171829e-01 -3.88148665e-01
-9.47400391e-01 5.45127153e-01 4.38172758e-01 2.30666995e-01
7.55939484e-02 -1.36096847e+00 -7.80108452e-01 -3.45565438e-01
1.16290838e-01 1.54492930e-01 3.91360551e-01 3.08815658e-01
-3.80827785e-01 2.31883347e-01 3.61770950e-02 7.88431522e-03
-7.90859640e-01 6.55172288e-01 1.49247542e-01 -7.34121025e-01
-6.48339689e-02 9.92061794e-01 -1.00165009e-01 -1.04906166e+00
1.81016833e-01 8.23961943e-02 -3.69194806e-01 -1.15340278e-01
8.08709741e-01 4.00677234e-01 -1.78040475e-01 -8.11814785e-01
-3.02311569e-01 1.87594190e-01 -8.85499641e-03 -4.89164084e-01
1.03890955e+00 -3.72418880e-01 -1.49516344e-01 4.34044510e-01
9.85295296e-01 4.10583079e-01 -1.09910679e+00 -6.29451871e-01
5.69730580e-01 -1.54146954e-01 -6.98554873e-01 -7.72783756e-01
-8.04773211e-01 8.72422516e-01 3.78561318e-01 2.94010360e-02
8.23865175e-01 1.12729948e-02 1.14773548e+00 8.08114946e-01
7.61162102e-01 -1.48964739e+00 -2.48306483e-01 8.02463353e-01
4.34054881e-01 -1.49999428e+00 -6.32449329e-01 -1.92691445e-01
-1.27610457e+00 8.54104877e-01 8.91658485e-01 4.48786356e-02
3.61528963e-01 2.59324431e-01 5.84137022e-01 1.88024402e-01
-3.22643518e-01 -5.40244043e-01 8.02608859e-03 9.13136840e-01
4.80310053e-01 2.02961832e-01 -2.49481589e-01 1.18122935e+00
3.47945064e-01 2.47077897e-01 5.39831936e-01 9.62356925e-01
-4.04428720e-01 -1.42063880e+00 -3.29014599e-01 -7.83832148e-02
-1.07707620e+00 -2.97929227e-01 -5.61913788e-01 4.70285028e-01
4.20140564e-01 1.29732573e+00 -2.93872237e-01 -4.16989595e-01
3.34806293e-01 4.35950667e-01 -2.38954827e-01 -6.50975049e-01
-9.43105936e-01 2.56068185e-02 4.02883142e-01 -2.39597112e-01
-6.57187223e-01 -1.48067683e-01 -1.37101889e+00 7.41485879e-02
-5.65303445e-01 5.49659789e-01 4.78253424e-01 1.23206091e+00
3.55208069e-01 -2.91927159e-02 6.11332476e-01 -5.93460321e-01
-5.63554049e-01 -1.28057849e+00 -5.69555700e-01 4.74618435e-01
-2.60595493e-02 1.98963881e-02 -7.98165947e-02 2.87591338e-01] | [9.756065368652344, 9.653223991394043] |
b1e50ccc-b509-4839-851d-3f9322af7742 | beyond-intuition-a-framework-for-applying-gps | 2307.03093 | null | https://arxiv.org/abs/2307.03093v1 | https://arxiv.org/pdf/2307.03093v1.pdf | Beyond Intuition, a Framework for Applying GPs to Real-World Data | Gaussian Processes (GPs) offer an attractive method for regression over small, structured and correlated datasets. However, their deployment is hindered by computational costs and limited guidelines on how to apply GPs beyond simple low-dimensional datasets. We propose a framework to identify the suitability of GPs to a given problem and how to set up a robust and well-specified GP model. The guidelines formalise the decisions of experienced GP practitioners, with an emphasis on kernel design and options for computational scalability. The framework is then applied to a case study of glacier elevation change yielding more accurate results at test time. | ['Richard E. Turner', 'Hong Ge', 'Ti John', 'Alex Gardner', 'Ross Viljoen', 'Jihao Andreas Lin', 'Kenza Tazi'] | 2023-07-06 | null | null | null | null | ['gaussian-processes'] | ['methodology'] | [ 5.46836015e-03 -1.39256969e-01 1.71702370e-01 -3.01469028e-01
-8.90871584e-01 -5.56455016e-01 7.45058179e-01 1.86233819e-01
-2.45396361e-01 9.65508461e-01 1.27197340e-01 -5.79437256e-01
-8.46230149e-01 -7.33192146e-01 -8.57725143e-02 -1.00843430e+00
-6.14253461e-01 5.66241086e-01 1.24704763e-01 -2.14271452e-02
3.21545631e-01 5.43501258e-01 -1.30126798e+00 -5.03978789e-01
1.10786402e+00 7.26076007e-01 1.35844409e-01 8.46237957e-01
2.31927991e-01 5.41275322e-01 -3.73639315e-01 -2.83763289e-01
2.92656630e-01 -3.18842120e-02 -3.84030819e-01 -8.08291808e-02
1.39055237e-01 8.60114396e-02 -1.09157786e-01 6.36793852e-01
8.14463794e-01 3.59234542e-01 9.73720133e-01 -1.14884043e+00
-5.86989582e-01 1.51877105e-01 -1.53875768e-01 5.51281989e-01
1.28127694e-01 3.22415769e-01 7.79973149e-01 -6.32010698e-01
1.75355583e-01 1.04191244e+00 1.28779328e+00 -8.28272253e-02
-1.74060583e+00 -1.13179609e-01 2.35111073e-01 -4.45881695e-01
-1.57414806e+00 -4.67326164e-01 1.20008275e-01 -7.47451782e-01
1.10019994e+00 2.11847097e-01 5.83563030e-01 9.58669960e-01
4.18341219e-01 3.34879488e-01 1.37017822e+00 -1.63505599e-02
6.19831324e-01 -6.90870658e-02 2.25481868e-01 -2.98670214e-02
6.11538410e-01 3.21816862e-01 -3.45515549e-01 -8.72887552e-01
6.97146714e-01 -2.56448984e-01 -3.11987609e-01 -5.40975571e-01
-6.64463460e-01 1.07155752e+00 1.14831021e-02 -1.11641720e-01
-6.88721240e-01 3.33532542e-02 1.76577687e-01 1.21792614e-01
9.24802959e-01 5.42508364e-01 -7.31798112e-01 -4.66560096e-01
-1.18082654e+00 6.59791589e-01 1.20849502e+00 9.48339760e-01
5.36409497e-01 2.04908192e-01 -2.97718197e-01 7.17954516e-01
5.63615918e-01 7.87304342e-01 -7.42777437e-02 -7.08006918e-01
5.90193570e-01 1.38054341e-01 5.13504148e-01 -1.01289868e+00
-5.67864180e-01 -3.96252424e-01 -6.89916015e-01 4.52872477e-02
4.11239684e-01 -6.38181925e-01 -8.07240188e-01 1.25801253e+00
1.85564741e-01 3.58855784e-01 3.88040811e-01 5.47847450e-01
3.47953677e-01 6.10413253e-01 5.48895955e-01 -6.21193312e-02
1.04290104e+00 -3.05744648e-01 -3.75766307e-01 -5.94954133e-01
6.78457558e-01 -4.58185405e-01 7.96386719e-01 2.10600987e-01
-8.13622117e-01 -1.89599663e-01 -5.79539955e-01 3.32517594e-01
-1.96316630e-01 -2.15288494e-02 8.19872499e-01 1.02694118e+00
-1.30684114e+00 8.38629723e-01 -1.08341169e+00 -6.14055574e-01
3.65049094e-01 4.79529023e-01 -7.95024559e-02 4.83032279e-02
-1.22510433e+00 1.10680079e+00 1.94271177e-01 5.64305425e-01
-4.41454977e-01 -9.20995235e-01 -9.20056999e-01 1.29933491e-01
-6.74355309e-03 -7.97498584e-01 9.64427292e-01 -2.75941163e-01
-1.51182818e+00 4.36020225e-01 -3.50568108e-02 -6.62522793e-01
6.50897861e-01 -3.12330902e-01 -3.46191317e-01 -2.04720691e-01
1.92902535e-01 5.17167225e-02 7.29488611e-01 -8.25407982e-01
-7.46353507e-01 -1.66833118e-01 -4.32144433e-01 3.98392379e-01
3.61913294e-01 3.37628543e-01 -2.05339417e-01 -5.72296619e-01
1.50070265e-01 -1.06416190e+00 -8.67626905e-01 -4.84330684e-01
5.18707186e-02 1.15059868e-01 5.11802375e-01 -7.92071581e-01
1.19311142e+00 -2.03998351e+00 -5.46212643e-02 4.08738405e-01
-1.30197406e-01 1.21798106e-01 2.69941211e-01 7.78402328e-01
1.14108868e-01 2.36027151e-01 -6.74556553e-01 -2.68290550e-01
1.22911274e-01 4.82755959e-01 -5.16981304e-01 7.23244071e-01
6.58009648e-01 5.49217165e-01 -1.05795169e+00 -3.35724279e-02
2.27615267e-01 5.51106155e-01 -2.98823476e-01 1.23284876e-01
2.79354960e-01 3.85125995e-01 -5.60836256e-01 7.93431640e-01
8.16696942e-01 2.71261837e-02 -3.35151926e-02 5.42577088e-01
-2.17566609e-01 1.76301077e-01 -1.44772410e+00 8.05841982e-01
-2.22004786e-01 5.59788823e-01 7.25414678e-02 -7.06385493e-01
1.30868518e+00 1.28873318e-01 4.11561430e-01 1.27257584e-02
-3.81532133e-01 3.44019145e-01 3.34540606e-02 -3.77846479e-01
4.24653769e-01 -2.83090800e-01 4.69404161e-02 1.71056256e-01
-2.22367004e-01 -6.88622832e-01 1.49259433e-01 -4.61173266e-01
1.22750270e+00 4.69202846e-01 4.52366114e-01 -9.07602489e-01
2.48202160e-01 1.61669180e-01 4.37348306e-01 9.09821570e-01
-3.01832259e-01 6.68304980e-01 6.29186630e-01 -5.23019314e-01
-7.54706562e-01 -8.43791425e-01 -6.50481403e-01 9.48564470e-01
-3.55730772e-01 -3.28722894e-01 -2.25052491e-01 -1.34651214e-01
5.35812914e-01 7.26193786e-01 -5.78304172e-01 1.44011974e-01
-1.75655171e-01 -1.68944705e+00 5.74194014e-01 5.46994388e-01
1.74426779e-01 -8.19529593e-01 -7.93201089e-01 5.12841821e-01
2.75692344e-01 -9.25535142e-01 1.18078582e-01 5.99237382e-01
-1.26920676e+00 -9.30495322e-01 -7.60786533e-01 -2.05971137e-01
4.81507719e-01 -7.24746799e-03 1.08416164e+00 -4.09218669e-01
1.02612644e-01 5.16861856e-01 -1.00788049e-01 -4.62619543e-01
-9.24870446e-02 2.49543011e-01 1.52010173e-01 -4.46289539e-01
5.35850048e-01 -4.45908099e-01 -5.29255271e-01 3.74378204e-01
-4.92497861e-01 -6.52117670e-01 4.14062977e-01 7.84724534e-01
3.50755453e-01 6.83912709e-02 3.41927409e-01 -7.34134436e-01
1.26089191e+00 -8.64483595e-01 -7.85773754e-01 5.33869080e-02
-9.30201352e-01 -6.86474517e-02 1.32701799e-01 -3.69475156e-01
-9.55989182e-01 4.23389450e-02 2.06974372e-01 -2.57860757e-02
-1.88694164e-01 9.80159581e-01 1.82741195e-01 -2.97234803e-01
8.61925066e-01 -4.40384112e-02 -1.55200794e-01 -2.99515247e-01
2.04563648e-01 3.90098184e-01 3.06065470e-01 -8.93468320e-01
5.98316073e-01 2.89603144e-01 3.56408596e-01 -1.15083694e+00
-2.37537518e-01 -5.06742597e-01 -7.67341316e-01 1.94549009e-01
3.53302896e-01 -1.08055305e+00 -3.25626731e-01 3.95447880e-01
-5.70882797e-01 -8.23133588e-01 -2.09318504e-01 6.47742748e-01
-8.03283572e-01 1.70874000e-01 -2.96740532e-01 -1.13674140e+00
-3.13564241e-01 -9.63930845e-01 1.19107854e+00 8.48915204e-02
-3.77723187e-01 -1.47446239e+00 5.02642393e-01 -1.89929828e-01
7.19157100e-01 2.76848733e-01 4.77660030e-01 -6.36583209e-01
-2.90100753e-01 -2.93133020e-01 -1.32553980e-01 2.38552168e-01
1.64767802e-01 4.57175910e-01 -1.02494287e+00 -2.26691276e-01
3.35469916e-02 1.77771792e-01 5.62516868e-01 9.49232578e-01
4.52161670e-01 -2.46422663e-01 -4.58036453e-01 9.12035525e-01
1.37658203e+00 -1.41935088e-02 6.01976931e-01 6.71173632e-01
4.43727940e-01 7.08727896e-01 6.22764468e-01 4.04083729e-01
2.74082184e-01 3.18606049e-01 -6.87567145e-02 -9.52867791e-02
7.05586910e-01 3.39261629e-03 1.22132152e-01 1.56412885e-01
-5.62883437e-01 -7.48231411e-02 -1.53016365e+00 5.53725004e-01
-2.06186032e+00 -9.03480053e-01 -6.48876846e-01 2.36631012e+00
4.29135650e-01 6.09303638e-02 2.85034254e-02 -1.78087026e-01
4.73938823e-01 1.55065998e-01 -2.02781767e-01 -5.22377253e-01
-1.95750296e-01 2.35681325e-01 1.25413787e+00 4.61982995e-01
-1.41712534e+00 1.00767493e+00 8.46910286e+00 4.38304424e-01
-7.90237188e-01 -1.81719840e-01 5.53433716e-01 2.06764221e-01
1.83292985e-01 3.57026964e-01 -9.51104045e-01 4.18161660e-01
1.36951327e+00 -4.74793792e-01 9.42646712e-02 9.40533817e-01
6.77058399e-01 -3.92691791e-01 -6.40163660e-01 6.48369372e-01
-5.20326078e-01 -9.95732605e-01 -5.76540768e-01 2.84683585e-01
8.40848267e-01 4.00524884e-01 -4.88390699e-02 3.90602231e-01
8.27933431e-01 -1.07524645e+00 2.62221962e-01 7.10206628e-01
2.33828515e-01 -6.38627708e-01 9.40252244e-01 1.56420439e-01
-1.04518211e+00 -8.17939118e-02 -5.86168170e-01 -4.62172538e-01
1.51346371e-01 4.15554821e-01 -8.62140417e-01 5.21942556e-01
8.34518731e-01 8.11600685e-01 -5.87188482e-01 1.53360140e+00
-2.67875880e-01 9.18387830e-01 -6.59506261e-01 4.59589101e-02
5.06337345e-01 -6.92636371e-01 4.40991670e-01 1.42350984e+00
7.01313078e-01 -7.37016962e-04 -2.53283195e-02 5.07601917e-01
1.01067877e+00 1.63384959e-01 -9.52487588e-01 1.64409310e-01
5.03800809e-01 1.05503690e+00 -6.45906389e-01 5.74951805e-02
-4.37766194e-01 4.31336403e-01 8.87334794e-02 7.59994507e-01
-3.61382931e-01 4.99138236e-02 1.07967305e+00 2.22404897e-01
4.58192319e-01 -4.27528501e-01 -4.11947697e-01 -7.31527388e-01
-2.04105392e-01 -7.17747748e-01 4.86734748e-01 -7.50723839e-01
-1.59193957e+00 3.82762820e-01 3.81631553e-01 -1.26295972e+00
-3.60872746e-01 -7.72674680e-01 -7.53140211e-01 1.62481022e+00
-1.35675073e+00 -9.80678678e-01 -1.48563579e-01 1.10923514e-01
1.21223740e-02 -4.14258651e-02 1.04190385e+00 -1.16515815e-01
-4.39026713e-01 -2.23933294e-01 6.13794506e-01 -4.79583293e-01
5.46616912e-01 -1.54743755e+00 9.87604082e-01 9.21016872e-01
-4.63853598e-01 7.50465989e-01 1.07595253e+00 -9.44156706e-01
-1.33868361e+00 -1.09511149e+00 6.84497237e-01 -7.55873263e-01
1.14976263e+00 -1.70275941e-01 -1.12274015e+00 7.11308062e-01
-4.12238687e-01 1.40584214e-02 7.39926934e-01 4.33597535e-01
3.52163374e-01 1.52510047e-01 -1.13395905e+00 5.91844201e-01
6.36557817e-01 -3.41312766e-01 -4.64266479e-01 1.90684155e-01
-1.17056789e-02 -6.29322171e-01 -1.07453942e+00 4.37731951e-01
2.64872849e-01 -5.94236493e-01 8.29210579e-01 -5.87467015e-01
-1.08336523e-01 -4.18915689e-01 -2.42834881e-01 -1.45849204e+00
-5.42372525e-01 -9.98974442e-01 3.40886861e-02 1.00305307e+00
5.00193417e-01 -9.58348095e-01 5.97410798e-01 1.24451041e+00
1.02007151e-01 -5.36257744e-01 -8.86701405e-01 -7.88261890e-01
1.69320941e-01 -7.15386152e-01 4.62674439e-01 8.24923038e-01
-2.58993447e-01 -6.10394925e-02 -3.63519490e-01 8.36167514e-01
6.20797217e-01 -3.28074843e-01 1.01456344e+00 -1.69830954e+00
5.59931621e-02 -3.59510124e-01 -8.17426980e-01 -5.16322076e-01
-1.80344954e-01 -2.71371782e-01 9.57177281e-02 -1.53149962e+00
-3.18547487e-01 -7.08578348e-01 1.07462958e-01 2.23565772e-01
-5.33901930e-01 -2.68771172e-01 -1.38775766e-01 3.63768756e-01
1.53985530e-01 5.07135272e-01 5.02458334e-01 2.42449895e-01
-6.67678297e-01 6.44743383e-01 -7.77553082e-01 5.04603267e-01
9.79061604e-01 -5.79700947e-01 -5.46141088e-01 1.66212190e-02
2.82474041e-01 -8.42851624e-02 4.30738807e-01 -9.39994514e-01
1.39224336e-01 -4.89061654e-01 3.85350138e-01 -4.70034242e-01
1.90686360e-01 -7.45163679e-01 6.36605442e-01 1.90149769e-01
1.42320037e-01 2.82711864e-01 6.91598892e-01 7.86730528e-01
-1.19362950e-01 -3.32791567e-01 4.34858203e-01 1.35788498e-02
-6.47493899e-01 3.63422930e-01 -5.54657161e-01 -2.19020117e-02
1.12818539e+00 -3.07833105e-01 -1.95924208e-01 -4.09766823e-01
-9.89442348e-01 6.29925787e-01 4.75145727e-01 1.62432596e-01
1.98712215e-01 -8.51657391e-01 -8.80988598e-01 2.39289641e-01
2.91735190e-03 1.64471790e-01 6.84808046e-02 8.38246346e-01
-8.12289059e-01 5.45716822e-01 -8.37179553e-03 -7.94324219e-01
-1.10309196e+00 2.98076540e-01 6.36902571e-01 -2.63884097e-01
-7.69931197e-01 5.00833809e-01 -1.22201778e-01 -4.25554425e-01
-1.90163016e-01 -7.94632494e-01 -9.97156277e-02 1.40422568e-01
3.28113079e-01 3.90751570e-01 8.50215405e-02 -5.41740477e-01
-2.19594523e-01 5.46367347e-01 5.29302537e-01 -2.53993839e-01
1.58546007e+00 -2.93020040e-01 1.42602250e-01 6.16927803e-01
5.08795857e-01 -3.37511718e-01 -1.90376794e+00 1.68136675e-02
4.62250262e-01 -5.38617074e-01 1.91375956e-01 -1.80966973e-01
-4.42043722e-01 6.07582688e-01 5.09731770e-01 3.32268655e-01
7.08627880e-01 -3.68783772e-01 -3.17178369e-01 4.71593618e-01
3.09859455e-01 -9.28377092e-01 -8.63956749e-01 6.46824419e-01
8.83382976e-01 -1.08300781e+00 4.17005152e-01 -2.19858497e-01
-1.01359069e+00 1.01194119e+00 8.21850523e-02 -1.77539468e-01
9.84194994e-01 2.49917626e-01 4.95547131e-02 -3.70876461e-01
-6.80559039e-01 -3.13638717e-01 1.33410200e-01 1.24025083e+00
2.52553940e-01 2.50080615e-01 -2.56332070e-01 3.99524510e-01
-3.11551064e-01 7.85413384e-02 3.14982176e-01 1.19428015e+00
-2.06336275e-01 -8.78846347e-01 -6.59056127e-01 8.21490526e-01
-2.91363716e-01 -3.58394772e-01 -1.29552577e-02 1.02319074e+00
-3.32430452e-01 8.18381190e-01 3.44377607e-02 2.16618761e-01
5.04755557e-01 1.70063108e-01 -1.00176940e-02 -6.67397082e-01
-4.75571066e-01 1.58494353e-01 3.37092668e-01 -1.04387820e-01
-3.85603696e-01 -1.24614525e+00 -4.72816706e-01 -3.72910708e-01
-2.46420562e-01 2.10612804e-01 4.81250197e-01 7.78147221e-01
4.36314493e-01 3.60017642e-02 3.24467808e-01 -9.26263869e-01
-5.72790682e-01 -1.15172696e+00 -9.47799802e-01 -2.10517213e-01
1.19456798e-01 -7.89324403e-01 -6.58005416e-01 -7.08851367e-02] | [6.4832282066345215, 3.5209743976593018] |
d7b2d406-0de5-42d8-90ea-0322792bbd81 | x-linear-attention-networks-for-image | 2003.14080 | null | https://arxiv.org/abs/2003.14080v1 | https://arxiv.org/pdf/2003.14080v1.pdf | X-Linear Attention Networks for Image Captioning | Recent progress on fine-grained visual recognition and visual question answering has featured Bilinear Pooling, which effectively models the 2$^{nd}$ order interactions across multi-modal inputs. Nevertheless, there has not been evidence in support of building such interactions concurrently with attention mechanism for image captioning. In this paper, we introduce a unified attention block -- X-Linear attention block, that fully employs bilinear pooling to selectively capitalize on visual information or perform multi-modal reasoning. Technically, X-Linear attention block simultaneously exploits both the spatial and channel-wise bilinear attention distributions to capture the 2$^{nd}$ order interactions between the input single-modal or multi-modal features. Higher and even infinity order feature interactions are readily modeled through stacking multiple X-Linear attention blocks and equipping the block with Exponential Linear Unit (ELU) in a parameter-free fashion, respectively. Furthermore, we present X-Linear Attention Networks (dubbed as X-LAN) that novelly integrates X-Linear attention block(s) into image encoder and sentence decoder of image captioning model to leverage higher order intra- and inter-modal interactions. The experiments on COCO benchmark demonstrate that our X-LAN obtains to-date the best published CIDEr performance of 132.0% on COCO Karpathy test split. When further endowing Transformer with X-Linear attention blocks, CIDEr is boosted up to 132.8%. Source code is available at \url{https://github.com/Panda-Peter/image-captioning}. | ['Ting Yao', 'Yingwei Pan', 'Yehao Li', 'Tao Mei'] | 2020-03-31 | x-linear-attention-networks-for-image-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Pan_X-Linear_Attention_Networks_for_Image_Captioning_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Pan_X-Linear_Attention_Networks_for_Image_Captioning_CVPR_2020_paper.pdf | cvpr-2020-6 | ['fine-grained-visual-recognition'] | ['computer-vision'] | [-4.57814559e-02 4.59375046e-02 -1.33757696e-01 -4.64798480e-01
-1.18433332e+00 -6.83391809e-01 6.46261692e-01 -2.58338124e-01
-2.75437772e-01 4.28531796e-01 2.73326457e-01 -6.05798304e-01
9.05941650e-02 -5.51526845e-01 -1.39901304e+00 -6.35667145e-01
1.26731321e-02 -2.42787935e-02 5.13478033e-02 -1.13943465e-01
1.17418110e-01 3.70247602e-01 -1.37478387e+00 9.18209493e-01
5.87812901e-01 1.22558355e+00 2.88235515e-01 9.42353427e-01
-1.21435203e-01 1.04064882e+00 -2.60920674e-01 -5.63277662e-01
1.86154276e-01 -2.49730542e-01 -1.00106835e+00 7.22881481e-02
7.74745703e-01 -5.01716733e-01 -5.87724209e-01 7.25008726e-01
3.40195119e-01 -1.66697770e-01 5.25625110e-01 -1.36156249e+00
-1.56557834e+00 4.45856631e-01 -1.01388359e+00 4.03586209e-01
2.37826273e-01 6.36887729e-01 1.40342760e+00 -1.19058907e+00
3.59960973e-01 1.28035426e+00 3.51418942e-01 4.84164923e-01
-1.08612955e+00 -6.61640525e-01 2.88987845e-01 5.26750326e-01
-1.37464714e+00 1.14968149e-02 7.13234365e-01 -3.90045881e-01
1.19097269e+00 4.47572023e-01 3.48853976e-01 9.64248359e-01
1.52698353e-01 1.07958555e+00 1.32626021e+00 -2.60737062e-01
-3.66817087e-01 2.08924025e-01 3.86045843e-01 7.46211588e-01
-2.50183940e-01 -1.79213267e-02 -4.56847310e-01 2.27501199e-01
8.30421865e-01 6.06950298e-02 -1.64440453e-01 2.34324355e-02
-1.18394160e+00 7.22263575e-01 1.18700802e+00 2.11181551e-01
-4.19091076e-01 5.60240269e-01 2.88171411e-01 7.97696114e-02
7.06239864e-02 2.84125239e-01 -3.83709162e-01 2.57980138e-01
-6.02662265e-01 3.39796506e-02 1.07692413e-01 1.14858127e+00
8.99158895e-01 -9.47028250e-02 -8.55653524e-01 7.02609777e-01
3.57716054e-01 4.48091418e-01 1.57184288e-01 -8.31898153e-01
7.24004149e-01 7.55128801e-01 -1.23635091e-01 -6.00408316e-01
-1.98819429e-01 -3.67359191e-01 -9.69113052e-01 1.23223320e-01
2.32158870e-01 8.68141055e-02 -1.30434787e+00 1.85219920e+00
2.50172857e-02 -5.05552068e-02 2.29163896e-02 1.05024827e+00
1.10091400e+00 9.32799518e-01 4.45564598e-01 3.74240011e-01
1.96898139e+00 -1.35087979e+00 -4.98365164e-01 -2.43956253e-01
2.32764468e-01 -5.92525840e-01 1.62082720e+00 -5.33925220e-02
-1.39870751e+00 -7.98812509e-01 -8.91325057e-01 -7.67392516e-01
-5.61246455e-01 2.18730971e-01 5.65344989e-01 3.07189286e-01
-1.27175927e+00 -7.95823261e-02 -4.60478902e-01 -6.27191663e-02
8.00294101e-01 5.74037611e-01 -5.55603564e-01 -3.59957516e-01
-1.22710466e+00 7.01413155e-01 4.51373458e-02 2.90438324e-01
-9.93358195e-01 -9.10275519e-01 -8.88492703e-01 2.93138742e-01
1.48871019e-01 -9.55372095e-01 1.10765934e+00 -8.90639603e-01
-1.09543502e+00 1.02124047e+00 -3.27532232e-01 -3.63223493e-01
2.94597775e-01 -2.19267070e-01 -8.01476911e-02 3.97958308e-01
1.08776353e-01 1.26758182e+00 6.64629698e-01 -1.45936823e+00
-3.41422647e-01 -3.36989701e-01 6.13423645e-01 3.53310347e-01
-2.06965327e-01 9.73678753e-02 -8.73399913e-01 -5.12730718e-01
-3.25126857e-01 -7.74786890e-01 5.95879368e-02 1.07267231e-01
-5.52025735e-01 -3.26779991e-01 8.45257521e-01 -8.01092863e-01
1.04916418e+00 -2.24871778e+00 1.01388358e-01 -6.00829050e-02
3.04535121e-01 2.05139071e-01 -4.04704243e-01 4.00466323e-01
-2.81592190e-01 2.35751525e-01 -2.28490353e-01 -4.67812628e-01
2.91193157e-01 1.97298616e-01 -3.85230601e-01 1.96472034e-01
7.28928149e-01 1.67835391e+00 -4.05168504e-01 -4.16504890e-01
2.57058322e-01 7.77309656e-01 -7.36184239e-01 3.42850477e-01
-2.71250844e-01 3.52341264e-01 -2.35617891e-01 7.42649853e-01
8.23875725e-01 -7.34366417e-01 -1.74113154e-01 -5.54719150e-01
-3.28023992e-02 -1.14807211e-01 -5.58200598e-01 1.47584951e+00
-4.63853151e-01 7.14894593e-01 8.19258168e-02 -7.63373137e-01
4.64172751e-01 1.14616275e-01 -8.88607185e-03 -1.02634120e+00
1.10156730e-01 -5.37175499e-02 -8.89795423e-02 -5.60915172e-01
2.93752521e-01 -1.09762780e-01 -2.02683225e-01 1.14799812e-01
3.42372954e-01 1.65150642e-01 -4.02214825e-02 3.80823106e-01
9.17755425e-01 2.40017269e-02 -2.09779032e-02 -1.22822933e-01
6.53286636e-01 -2.10057378e-01 -4.14414890e-02 8.78338397e-01
-1.12464949e-01 8.39373231e-01 6.73839927e-01 -1.73258707e-01
-1.20004463e+00 -1.12867355e+00 -4.88013215e-02 1.42329431e+00
2.33791143e-01 -1.09096624e-01 -7.14101493e-01 -5.58805943e-01
7.95777421e-03 6.25018477e-01 -1.08716190e+00 -4.28033322e-02
-5.16203225e-01 -6.43666744e-01 6.69356108e-01 8.22347999e-01
7.10974514e-01 -1.36919212e+00 -3.56533676e-01 -1.64523393e-01
-9.11075622e-02 -1.16946256e+00 -8.31315696e-01 2.24158064e-01
-1.69964120e-01 -8.10458481e-01 -1.00290906e+00 -8.43299747e-01
6.91085815e-01 2.83690512e-01 1.07918286e+00 1.08635664e-01
-4.64107871e-01 5.60160697e-01 -2.58688331e-01 -1.84940994e-01
2.58735836e-01 1.38594471e-02 -5.84584594e-01 2.58257806e-01
2.21379146e-01 -2.79839396e-01 -9.71151650e-01 2.37874016e-01
-1.24027765e+00 2.86051989e-01 9.10713792e-01 1.09652412e+00
5.87706327e-01 -5.81339061e-01 4.79652166e-01 -6.86761379e-01
3.65407169e-01 -6.79357827e-01 -2.34629482e-01 4.06190544e-01
-8.77524540e-02 6.74305931e-02 4.68902409e-01 -3.39171231e-01
-9.39749897e-01 -8.44145939e-02 -3.77961576e-01 -7.28021801e-01
-1.44756198e-01 3.71713489e-01 -3.27574551e-01 -4.96271327e-02
1.92912012e-01 3.36247683e-01 -4.38803256e-01 -2.07360759e-01
8.36572349e-01 5.96567631e-01 6.86070859e-01 -5.65838039e-01
6.97909236e-01 4.91732597e-01 -3.22729349e-01 -5.14603436e-01
-8.48382652e-01 -3.61157209e-01 -3.56204420e-01 -7.70554394e-02
1.43124330e+00 -9.94714856e-01 -1.16925383e+00 5.21052659e-01
-1.27444553e+00 -4.17036563e-01 -7.13035613e-02 1.54037431e-01
-5.33956528e-01 4.04860154e-02 -7.08113194e-01 -6.59357846e-01
-3.89164329e-01 -1.46399677e+00 1.25732446e+00 3.13843191e-01
2.07438186e-01 -7.28876412e-01 -4.27840143e-01 8.98503959e-01
4.53506947e-01 2.33459380e-02 9.84481692e-01 -2.81177640e-01
-1.12867880e+00 8.40113033e-03 -1.01436698e+00 3.92415196e-01
-3.43659222e-01 -3.38495195e-01 -1.18922317e+00 -2.53714412e-01
-3.62013727e-01 -6.08296037e-01 1.00650692e+00 4.31241035e-01
1.53175735e+00 -2.72189587e-01 -1.22924214e-02 7.12862909e-01
1.49306798e+00 1.17036430e-02 9.54631865e-01 2.42562905e-01
9.93853629e-01 4.40486521e-01 1.51431337e-01 1.04033135e-01
9.01928365e-01 7.33617544e-01 7.18246758e-01 -5.97510219e-01
-3.57544988e-01 -2.29934081e-01 2.69812077e-01 4.34471697e-01
8.94235298e-02 -3.01585853e-01 -8.73793066e-01 7.87551761e-01
-1.56745279e+00 -8.90202641e-01 -1.48009226e-01 1.62620831e+00
7.80765772e-01 -1.11995384e-01 -1.09208547e-01 -2.66746849e-01
5.77772796e-01 1.62157953e-01 -6.44893765e-01 -5.59328437e-01
-3.34074527e-01 2.32650191e-01 5.54930151e-01 6.41619503e-01
-1.00486076e+00 8.63814831e-01 5.09821844e+00 9.17100191e-01
-1.12742662e+00 4.04477447e-01 9.89875913e-01 -7.90831745e-02
-5.95739186e-01 -2.31400281e-01 -7.68543899e-01 3.54450226e-01
6.66737676e-01 1.86919108e-01 3.71323287e-01 4.46731657e-01
-1.22884959e-01 1.37726620e-01 -9.35180843e-01 1.07106864e+00
1.48349568e-01 -1.45883155e+00 2.23946929e-01 5.17865978e-02
7.82398164e-01 1.31211236e-01 5.68111956e-01 4.67340022e-01
9.80046391e-02 -1.34997404e+00 8.75723064e-01 4.64454979e-01
1.10622323e+00 -5.89285254e-01 6.54223502e-01 3.56413126e-02
-1.27918589e+00 -2.83427149e-01 4.39636968e-02 9.58598256e-02
2.19573840e-01 1.49788875e-02 -5.37060201e-01 5.80331504e-01
1.02710783e+00 3.18133920e-01 -7.07144082e-01 7.57266939e-01
-1.23311006e-01 4.63720262e-01 -1.16232760e-01 2.13479117e-01
7.76629925e-01 2.28492334e-01 1.98062032e-01 1.35137606e+00
1.66087851e-01 2.48635843e-01 -2.31715858e-01 9.33520854e-01
-3.63005102e-01 -6.48701787e-02 -3.23002309e-01 1.52515545e-02
1.23095717e-02 1.15465546e+00 -3.53683621e-01 -3.44474137e-01
-7.46042371e-01 1.25934863e+00 6.34446383e-01 6.70717239e-01
-1.32618713e+00 -2.83990562e-01 5.24815500e-01 1.83081850e-01
7.04654694e-01 -1.07078306e-01 -4.06102479e-01 -1.07021475e+00
1.81381062e-01 -8.14489663e-01 4.54711914e-01 -1.30687356e+00
-1.48884296e+00 9.45694983e-01 -6.98156729e-02 -7.37896383e-01
2.01793417e-01 -9.14052606e-01 -5.05775571e-01 1.18375051e+00
-1.84547925e+00 -1.92696726e+00 -3.49683315e-01 8.33591461e-01
5.70826232e-01 2.14530319e-01 5.80800533e-01 5.35916209e-01
-5.33303082e-01 9.59380746e-01 -1.40756235e-01 1.75356299e-01
5.43753803e-01 -1.27615428e+00 1.61468744e-01 6.23426259e-01
1.75323293e-01 6.57808483e-01 3.25814784e-01 -1.19598404e-01
-1.46515489e+00 -1.15609169e+00 8.21629703e-01 -5.81736088e-01
6.61176801e-01 -5.95263243e-01 -1.03862584e+00 1.02878988e+00
7.49716997e-01 3.66229564e-01 5.87727845e-01 -2.47009814e-01
-8.00743222e-01 -1.11907445e-01 -9.64603543e-01 5.38092613e-01
7.71016717e-01 -9.25333381e-01 -3.55392516e-01 2.42838711e-01
9.41326678e-01 -4.80759144e-01 -8.00684333e-01 4.80578274e-01
4.42028075e-01 -8.27293992e-01 1.29044867e+00 -6.04239225e-01
7.82519221e-01 -4.08903956e-01 -3.98503125e-01 -6.53417408e-01
-4.83486801e-01 -3.74069929e-01 -3.53012420e-02 1.20058489e+00
4.89897549e-01 -4.57960993e-01 4.62350905e-01 6.10438228e-01
-2.76583821e-01 -1.20578170e+00 -9.78918135e-01 -2.80153930e-01
2.94556499e-01 -4.12251115e-01 4.52118725e-01 7.36989558e-01
-2.93017328e-01 3.65086317e-01 -5.58887243e-01 3.94067496e-01
4.17074472e-01 3.37926261e-02 5.22787809e-01 -2.57931381e-01
-4.46556807e-01 -4.50038970e-01 -2.22392425e-01 -1.32129395e+00
7.66397864e-02 -9.11590397e-01 -1.06596708e-01 -1.63697839e+00
5.42028069e-01 -2.06519470e-01 -5.07408500e-01 7.37421393e-01
-4.66171443e-01 9.20377553e-01 5.10691047e-01 -1.04273669e-02
-6.86451077e-01 5.73502421e-01 1.49459815e+00 -2.84660012e-01
2.01494083e-01 -5.03762245e-01 -9.77808595e-01 2.79103816e-01
6.83861911e-01 -4.12223376e-02 -4.02630389e-01 -8.52780759e-01
9.52838808e-02 -1.40249701e-02 8.74852657e-01 -6.00545526e-01
2.52465338e-01 2.30760630e-02 4.60156858e-01 -6.99254751e-01
5.62719405e-01 -6.63423598e-01 4.96107712e-02 1.15552209e-01
-4.72250253e-01 2.21720591e-01 5.78218400e-01 3.37585688e-01
-4.10457224e-01 2.51168668e-01 8.28062117e-01 -1.00186959e-01
-8.36624324e-01 4.83645678e-01 6.57393485e-02 -1.57157388e-02
1.16826415e+00 -4.46791910e-02 -6.66141510e-01 -3.36529404e-01
-8.00220490e-01 4.40608859e-01 4.72393855e-02 5.57297409e-01
5.95103920e-01 -1.48354816e+00 -8.65800083e-01 3.01647037e-01
3.54469150e-01 -2.20034361e-01 9.09249067e-01 8.22546482e-01
-4.16719258e-01 7.44611442e-01 -3.62181276e-01 -6.08857632e-01
-1.18520617e+00 6.78891838e-01 2.58203506e-01 -4.14189965e-01
-3.46404701e-01 1.32214200e+00 8.56498301e-01 -2.31130585e-01
7.30150715e-02 -2.92565733e-01 -2.75800619e-02 -1.55127555e-01
5.62605739e-01 -1.61769897e-01 -2.14476421e-01 -8.67329121e-01
-4.88670290e-01 7.26063192e-01 -2.45679513e-01 -3.80110480e-02
1.21380949e+00 -2.95424998e-01 -2.13588342e-01 2.02392712e-01
1.55864620e+00 -3.94232988e-01 -1.50784516e+00 -9.87247899e-02
-5.68430245e-01 -2.81780183e-01 -6.57329038e-02 -1.08265209e+00
-1.05268002e+00 1.33229208e+00 5.61725438e-01 2.54922993e-02
1.35608804e+00 4.85528857e-01 5.52690387e-01 -1.83213890e-01
-2.09498078e-01 -3.67852181e-01 3.51127148e-01 4.30641592e-01
1.32330739e+00 -1.23395526e+00 -4.08462912e-01 -1.42594293e-01
-8.63699615e-01 6.07538760e-01 9.70533013e-01 -2.09188446e-01
4.81216103e-01 1.75678149e-01 9.64476839e-02 -2.81960070e-01
-8.83822560e-01 -2.84674942e-01 6.82644367e-01 3.19949567e-01
4.07274872e-01 3.90773378e-02 8.08816478e-02 7.13716567e-01
1.33712307e-01 -9.96194631e-02 7.05088628e-03 5.46357810e-01
-4.46080565e-02 -9.45525765e-01 -4.73331928e-01 3.16003501e-01
-5.25348485e-01 -5.83249450e-01 -5.24788201e-02 7.79390275e-01
3.51948172e-01 7.02178657e-01 2.28741214e-01 -2.57037938e-01
3.30048800e-01 -6.34888560e-03 4.75766003e-01 -2.03603074e-01
-7.97955632e-01 -7.38378540e-02 -2.98684865e-01 -5.29916883e-01
-1.88206449e-01 -3.14565927e-01 -1.22312427e+00 -2.10891560e-01
-4.23556007e-03 -1.97233468e-01 4.50464368e-01 7.16318250e-01
6.00962937e-01 7.87934661e-01 3.76425534e-01 -8.44367206e-01
-2.23980844e-01 -9.26454306e-01 -2.34472007e-01 5.46497166e-01
6.27990901e-01 -3.90358567e-01 -1.85914889e-01 1.77732766e-01] | [10.753021240234375, 1.5664310455322266] |
5eab5574-0217-46d2-9268-7b4146e89d68 | an-exploratory-literature-study-on-sharing | 2307.02443 | null | https://arxiv.org/abs/2307.02443v1 | https://arxiv.org/pdf/2307.02443v1.pdf | An Exploratory Literature Study on Sharing and Energy Use of Language Models for Source Code | Large language models trained on source code can support a variety of software development tasks, such as code recommendation and program repair. Large amounts of data for training such models benefit the models' performance. However, the size of the data and models results in long training times and high energy consumption. While publishing source code allows for replicability, users need to repeat the expensive training process if models are not shared. The main goal of the study is to investigate if publications that trained language models for software engineering (SE) tasks share source code and trained artifacts. The second goal is to analyze the transparency on training energy usage. We perform a snowballing-based literature search to find publications on language models for source code, and analyze their reusability from a sustainability standpoint. From 494 unique publications, we identified 293 relevant publications that use language models to address code-related tasks. Among them, 27% (79 out of 293) make artifacts available for reuse. This can be in the form of tools or IDE plugins designed for specific tasks or task-agnostic models that can be fine-tuned for a variety of downstream tasks. Moreover, we collect insights on the hardware used for model training, as well as training time, which together determine the energy consumption of the development process. We find that there are deficiencies in the sharing of information and artifacts for current studies on source code models for software engineering tasks, with 40% of the surveyed papers not sharing source code or trained artifacts. We recommend the sharing of source code as well as trained artifacts, to enable sustainable reproducibility. Moreover, comprehensive information on training times and hardware configurations should be shared for transparency on a model's carbon footprint. | ['Leon Moonen', 'Anastasiia Grishina', 'Max Hort'] | 2023-07-05 | null | null | null | null | ['program-repair', 'program-repair'] | ['computer-code', 'reasoning'] | [-1.46407887e-01 1.10483624e-01 -7.56131172e-01 -1.80978820e-01
-6.16044462e-01 -8.66132259e-01 7.92142898e-02 4.55881469e-02
-5.31522697e-03 4.18828800e-02 5.45275286e-02 -8.77390563e-01
-1.79368004e-01 -5.77098131e-01 -1.08676708e+00 -4.33434956e-02
2.04009518e-01 -3.64496768e-01 -2.67230272e-01 1.28282636e-01
6.37597978e-01 2.24213496e-01 -1.54234326e+00 4.20414388e-01
1.22110879e+00 3.45466554e-01 7.41681278e-01 4.34954524e-01
-4.92131412e-01 1.09847879e+00 -7.70538151e-01 -3.74193341e-01
-1.28411530e-02 -2.06441194e-01 -7.03688204e-01 -4.99511272e-01
1.61341578e-01 2.19578017e-03 2.49854788e-01 9.51065660e-01
3.59184384e-01 -3.87790650e-01 2.53605425e-01 -1.52259636e+00
-9.96926248e-01 8.91307175e-01 -4.39497024e-01 -1.87921509e-01
1.38204038e-01 -4.38727140e-02 6.07228518e-01 -9.46726739e-01
7.59044826e-01 5.31103492e-01 9.41907942e-01 6.37472510e-01
-9.62759018e-01 -1.02209604e+00 -1.59312442e-01 -3.86196338e-02
-1.17421734e+00 -6.46276057e-01 9.44906056e-01 -1.09746301e+00
1.55246437e+00 2.56259084e-01 7.90773273e-01 1.20336378e+00
6.61887825e-01 9.81788523e-03 7.70263970e-01 -7.68702507e-01
3.20827216e-01 8.50916862e-01 1.75486743e-01 9.54290807e-01
1.00721467e+00 -1.24369800e-01 -6.15693152e-01 -3.97318780e-01
2.08355933e-01 -2.17025280e-02 -1.48997304e-03 -1.12658747e-01
-1.05950212e+00 4.25046295e-01 4.96462025e-02 6.69907212e-01
6.11267649e-02 4.72740948e-01 4.09265518e-01 4.39303994e-01
4.31960821e-01 8.61826658e-01 -8.83963346e-01 -5.34424424e-01
-1.10579753e+00 -1.89157665e-01 1.03093386e+00 1.55377865e+00
9.01244640e-01 1.98851615e-01 4.15718742e-02 7.86910594e-01
6.32356524e-01 5.29793501e-01 4.57836598e-01 -1.10414827e+00
6.29147887e-01 1.19875097e+00 -1.67873904e-01 -8.89638722e-01
-2.52207726e-01 -5.89083254e-01 -3.58866513e-01 2.01983899e-01
2.53419336e-02 -3.08595002e-01 -5.10070980e-01 1.43811393e+00
-2.66796291e-01 -4.70443100e-01 -2.39169255e-01 2.17241615e-01
8.44175518e-01 5.13201594e-01 2.40854740e-01 -8.55565257e-03
1.20942461e+00 -1.24848938e+00 -5.19463658e-01 -3.58033478e-01
1.33359039e+00 -7.77989745e-01 1.34059215e+00 3.68945897e-01
-9.32798862e-01 -6.02728724e-01 -1.10161078e+00 -2.55989492e-01
-6.57240331e-01 7.81156540e-01 7.89984226e-01 1.26224542e+00
-1.04046535e+00 6.39228225e-01 -8.24602067e-01 -3.78169328e-01
3.82950455e-01 1.34090677e-01 1.20675474e-01 1.33737028e-01
-5.29508829e-01 1.03499484e+00 9.53976586e-02 -3.25705528e-01
-9.67960835e-01 -1.22748649e+00 -5.82286835e-01 1.61632061e-01
1.88359898e-02 -6.91053450e-01 1.02891731e+00 -1.24197590e+00
-1.13604522e+00 7.00615466e-01 -4.50740941e-02 1.68174803e-01
5.78663014e-02 -1.01504102e-02 -5.42137861e-01 -6.58829749e-01
5.89921325e-02 1.87123884e-02 5.17366350e-01 -1.10143554e+00
-5.81160903e-01 7.30997622e-02 1.83459148e-01 -5.63556015e-01
-1.00254643e+00 4.62661505e-01 -3.09246182e-01 -4.78696346e-01
-4.32108909e-01 -9.92325604e-01 6.03677742e-02 -6.98169693e-02
-1.09464392e-01 7.20821368e-03 4.95579362e-01 -9.06317532e-01
1.83596230e+00 -2.05398965e+00 -9.55584347e-02 2.40373500e-02
2.87259579e-01 -2.00015247e-01 -1.99092716e-01 4.62183774e-01
3.88434753e-02 1.12120724e+00 2.64937282e-01 -2.85403207e-02
1.56545967e-01 -3.34692210e-01 -3.03524430e-03 2.06630424e-01
-1.72324236e-02 7.29151368e-01 -6.19516432e-01 -4.33705568e-01
-1.71039030e-01 5.15546381e-01 -6.77699447e-01 1.01824693e-01
-1.65214106e-01 -4.17954326e-02 -4.79736835e-01 7.84932613e-01
3.05470705e-01 -5.94830275e-01 1.27855375e-01 -9.08426717e-02
-5.43770432e-01 7.00264394e-01 -8.09073567e-01 2.12078047e+00
-1.32971251e+00 9.95423973e-01 -6.83819037e-03 -3.75094116e-01
9.40666616e-01 2.81415731e-01 5.59143662e-01 -4.60079104e-01
-1.14086196e-01 5.59957743e-01 1.76426217e-01 -6.74910963e-01
4.01567757e-01 4.22364086e-01 -1.22914359e-01 7.94758022e-01
4.12729979e-02 1.63870193e-02 2.33365580e-01 -7.76519626e-02
1.32134366e+00 5.58553457e-01 -1.96118206e-02 -6.06671512e-01
2.58233938e-02 3.21858078e-01 4.05475438e-01 4.36706871e-01
4.00479972e-01 -2.38020066e-02 4.28229243e-01 -2.01331973e-01
-1.18943083e+00 -3.97247583e-01 -4.71125506e-02 1.39416778e+00
-3.91492456e-01 -1.11212802e+00 -9.76478934e-01 -7.48482227e-01
-2.25633711e-01 9.74429488e-01 -4.34434652e-01 -3.88786912e-01
-2.04520419e-01 -4.57237393e-01 6.57742202e-01 5.48043430e-01
2.23731473e-01 -7.38364637e-01 -1.07087481e+00 2.98796464e-02
-2.00394809e-01 -4.62542504e-01 -5.27224481e-01 4.65173066e-01
-1.01732969e+00 -1.15271521e+00 -2.44919807e-01 -7.01209128e-01
9.23626184e-01 2.59699315e-01 1.40888524e+00 5.49765050e-01
-2.33408794e-01 3.79664272e-01 -3.22652370e-01 -8.14907908e-01
-8.74933720e-01 3.57527882e-01 -2.48625681e-01 -9.75481331e-01
5.31179726e-01 -3.42042714e-01 -2.88743109e-01 2.14959159e-01
-4.52162474e-01 3.83601695e-01 8.61198485e-01 3.34867597e-01
1.94742069e-01 2.08714306e-01 4.98342007e-01 -9.07590806e-01
5.44873178e-01 -8.21909130e-01 -6.46664679e-01 7.54249573e-01
-1.41570473e+00 1.04878105e-01 5.68155646e-01 -3.37620646e-01
-1.18834722e+00 -1.19331121e-01 1.94345981e-01 -2.05196321e-01
3.72356415e-01 9.22637463e-01 -5.67163788e-02 -3.17390501e-01
1.25058031e+00 -1.43022761e-01 -4.60812092e-01 -6.74391747e-01
-1.66691303e-01 8.70598614e-01 -3.87328863e-01 -8.86261523e-01
7.67153203e-01 -2.27558196e-01 -3.74499798e-01 -3.75388682e-01
-1.24189138e-01 -4.56449017e-02 -2.11810127e-01 -2.18300223e-01
5.99457502e-01 -8.50614011e-01 -2.40442455e-01 -7.71309137e-02
-1.21804607e+00 -5.68567216e-01 -1.12355627e-01 3.64872366e-01
-1.44090116e-01 -2.43196785e-01 -3.04049075e-01 -7.66643524e-01
-5.08362532e-01 -1.25953948e+00 8.06915939e-01 1.28889531e-01
-4.45694208e-01 -1.07719052e+00 -4.14849222e-02 3.52382928e-01
7.43556499e-01 1.16531394e-01 1.27668083e+00 -3.84256274e-01
-5.56226015e-01 -1.03665344e-01 9.57985595e-02 3.15768987e-01
3.52193177e-01 6.19143605e-01 -9.96295929e-01 -3.75976145e-01
1.08605780e-01 6.68530837e-02 3.03147525e-01 2.63779759e-01
1.45257258e+00 -4.03450668e-01 -6.57640457e-01 5.90673029e-01
1.56660938e+00 4.36437547e-01 3.55930746e-01 5.52868485e-01
8.55875313e-01 6.59473181e-01 4.46063459e-01 3.83537143e-01
3.03663403e-01 4.29010361e-01 1.52731508e-01 -2.88213547e-02
2.12815423e-02 -4.86133188e-01 8.01589370e-01 1.21048236e+00
-2.93945402e-01 4.23229486e-02 -1.37776661e+00 5.96341908e-01
-1.32232201e+00 -5.48659384e-01 -4.79823709e-01 2.32306242e+00
7.66160429e-01 2.08637677e-02 -9.93978679e-02 -1.80270806e-01
3.70013982e-01 -3.77842098e-01 -3.95935327e-01 -4.44069564e-01
5.31687081e-01 2.70514246e-02 6.15968943e-01 -2.07720727e-01
-2.96349913e-01 2.00086832e-01 6.30948114e+00 6.31081939e-01
-1.12328351e+00 7.04118490e-01 3.53611231e-01 -2.57303625e-01
-9.26078856e-01 4.71704990e-01 -6.23648524e-01 6.89006865e-01
1.63072681e+00 -7.80729830e-01 6.28456116e-01 1.45971525e+00
4.78734106e-01 7.74326995e-02 -1.37736595e+00 7.57600069e-01
-3.43877286e-01 -1.68070006e+00 -4.67360586e-01 1.74801588e-01
1.02127075e+00 -2.89588291e-02 -1.27523854e-01 4.42111522e-01
2.36220241e-01 -9.44919825e-01 1.05388451e+00 2.90718943e-01
1.14029562e+00 -4.83514935e-01 4.68825161e-01 4.29870158e-01
-1.18200755e+00 -2.73725361e-01 -4.75983888e-01 -1.21804915e-01
-5.82387269e-01 7.58932710e-01 -5.99413157e-01 6.06359005e-01
1.05731034e+00 7.49304295e-01 -1.01895130e+00 6.46310329e-01
4.66877893e-02 9.70344841e-01 2.43207663e-02 -2.32415795e-01
-4.89774138e-01 3.43883857e-02 -9.34565589e-02 1.48349392e+00
9.07144308e-01 -6.18294477e-01 -2.89134771e-01 1.67742682e+00
-2.23230273e-01 2.44549841e-01 -7.73295820e-01 -6.65607750e-01
7.82718778e-01 1.32045841e+00 -5.27423799e-01 -6.55269250e-02
-8.30158532e-01 3.77822518e-01 1.03229649e-01 2.81924099e-01
-1.03555834e+00 -5.76046407e-01 4.09064054e-01 3.06249738e-01
-4.25145149e-01 -1.90436646e-01 -8.57408524e-01 -1.02011788e+00
4.74331468e-01 -8.00687253e-01 -1.77425444e-01 -8.90112758e-01
-6.50401235e-01 3.48622471e-01 -2.33693495e-02 -1.26787233e+00
1.09457351e-01 -3.89511883e-01 -5.23984909e-01 1.08309257e+00
-1.23128760e+00 -1.21227646e+00 -6.25510693e-01 -1.45605236e-01
5.10991931e-01 -4.60484505e-01 6.71995938e-01 5.50256133e-01
-6.19210660e-01 6.28129005e-01 1.88950807e-01 -2.29598016e-01
7.19547510e-01 -9.47177947e-01 4.93148118e-01 8.47817302e-01
-4.40518558e-03 1.37061644e+00 4.50150132e-01 -9.11846876e-01
-1.87273455e+00 -1.38423395e+00 1.07290912e+00 -9.31611896e-01
4.81776685e-01 -4.97192621e-01 -5.97923696e-01 7.14405239e-01
6.53258488e-02 -4.43896621e-01 9.94591892e-01 2.10689351e-01
-3.25215667e-01 -9.26983953e-02 -9.76953804e-01 2.98289418e-01
1.35907578e+00 -6.72595084e-01 -9.51435976e-03 4.44793463e-01
5.98690152e-01 -2.15699524e-01 -1.32280874e+00 -8.37975144e-02
6.22226596e-01 -8.04892957e-01 5.81220329e-01 -4.35349286e-01
1.04064095e+00 -1.76223248e-01 -1.42181134e-02 -1.13693643e+00
-4.89897341e-01 -4.54737127e-01 -1.71713039e-01 1.64754975e+00
8.96310806e-01 -3.72931749e-01 4.18120563e-01 1.01835573e+00
-4.85352904e-01 -5.96480668e-01 -3.52860063e-01 -1.02996469e+00
2.65333802e-01 -5.98156214e-01 8.07341993e-01 1.12660110e+00
3.17296803e-01 3.78172882e-02 4.79279421e-02 -2.02260926e-01
2.14377478e-01 -3.67895029e-02 6.72880411e-01 -1.24819446e+00
-4.84014988e-01 -4.56891984e-01 2.35456243e-01 -3.70643139e-01
3.17914963e-01 -1.37484837e+00 -1.75370842e-01 -1.82678115e+00
5.80289841e-01 -7.62934208e-01 -3.27982940e-02 9.12994087e-01
1.43795788e-01 -3.85243505e-01 -2.50281300e-02 5.89564025e-01
-1.63557678e-01 -1.19686432e-01 6.50506079e-01 -2.50363320e-01
-4.93734986e-01 -1.62255093e-01 -1.15349174e+00 6.34874344e-01
6.98786616e-01 -9.83248174e-01 -6.74467742e-01 -8.64477992e-01
8.91609013e-01 -1.82182789e-01 2.08873332e-01 -9.53994691e-01
1.82383999e-01 -4.81743366e-01 9.37647671e-02 2.03931689e-01
-6.47834241e-01 -1.19656932e+00 1.11585915e+00 6.09022081e-01
-3.54566932e-01 4.88713346e-02 4.86763120e-01 1.31152973e-01
3.36800367e-01 -8.51904213e-01 3.28832030e-01 -1.12024182e-02
-5.40609896e-01 -2.13988200e-01 -4.54467356e-01 -1.85859591e-01
1.16376972e+00 -3.05312872e-01 -6.83372557e-01 2.63786197e-01
-5.77748232e-02 -2.84615993e-01 9.96607780e-01 8.39425504e-01
1.16490476e-01 -1.22310424e+00 -3.43536705e-01 -1.41122788e-02
4.10815209e-01 -4.42009807e-01 1.05813988e-01 8.06564450e-01
-5.32635689e-01 5.08644402e-01 -2.91048586e-01 -2.03778163e-01
-1.13706243e+00 6.80110037e-01 4.02148277e-01 -8.51333663e-02
-3.23924363e-01 4.35663730e-01 -1.48823276e-01 -4.35891569e-01
-1.50450617e-02 -6.79729283e-01 2.94985682e-01 -2.00055510e-01
2.58651912e-01 8.07245135e-01 4.64521289e-01 2.22424537e-01
-5.42857945e-01 4.82280076e-01 3.19011033e-01 3.35851252e-01
1.43923783e+00 2.55975217e-01 -3.88449162e-01 7.49364436e-01
1.14555466e+00 3.34566176e-01 -9.21027243e-01 3.97384882e-01
2.79773027e-01 -3.93480778e-01 2.10278451e-01 -1.17333066e+00
-1.12958467e+00 7.33685970e-01 6.63891017e-01 1.23475485e-01
1.18395638e+00 -2.44989675e-02 3.54270637e-02 3.54321718e-01
9.92952883e-01 -1.33813727e+00 -1.08221024e-01 -4.45541739e-02
9.31397319e-01 -9.85276997e-01 4.00974341e-02 -3.67390007e-01
-1.75615028e-01 1.26641893e+00 7.77685642e-01 5.06981790e-01
6.10036433e-01 9.00405586e-01 -2.88861990e-01 -2.69312352e-01
-9.14356411e-01 7.29166210e-01 1.26986325e-01 6.02552772e-01
1.21753836e+00 -1.58112586e-01 -3.80682528e-01 1.17583072e+00
-1.07276611e-01 4.86108929e-01 8.34847689e-01 1.31314874e+00
-2.28924528e-01 -1.28241682e+00 -2.59109944e-01 9.53977585e-01
-4.35381740e-01 -2.50298917e-01 -4.92814600e-01 4.50892359e-01
3.66804510e-01 1.20410216e+00 -3.97770435e-01 -8.68834019e-01
3.58124405e-01 2.55854875e-01 3.09541941e-01 -1.08275104e+00
-1.20911109e+00 -4.38052118e-01 2.51540095e-01 -4.25984383e-01
-2.46363074e-01 -4.91265953e-01 -1.11108959e+00 -5.02934754e-01
-6.27837837e-01 1.28327370e-01 1.34534264e+00 5.53252995e-01
9.43242192e-01 1.12919903e+00 3.07072431e-01 -4.84780550e-01
-2.78688014e-01 -8.76692414e-01 -2.27044165e-01 -1.19278997e-01
-9.40669850e-02 -6.49095476e-01 -2.43230492e-01 6.44869268e-01] | [7.725892066955566, 7.786728382110596] |
44b2527d-3c72-46e1-b0c5-cda2c04301ac | building-trust-lessons-from-the-technion | 2207.14638 | null | https://arxiv.org/abs/2207.14638v2 | https://arxiv.org/pdf/2207.14638v2.pdf | Building Trust: Lessons from the Technion-Rambam Machine Learning in Healthcare Datathon Event | A datathon is a time-constrained competition involving data science applied to a specific problem. In the past decade, datathons have been shown to be a valuable bridge between fields and expertise . Biomedical data analysis represents a challenging area requiring collaboration between engineers, biologists and physicians to gain a better understanding of patient physiology and of guide decision processes for diagnosis, prognosis and therapeutic interventions to improve care practice. Here, we reflect on the outcomes of an event that we organized in Israel at the end of March 2022 between the MIT Critical Data group, Rambam Health Care Campus (Rambam) and the Technion Israel Institute of Technology (Technion) in Haifa. Participants were asked to complete a survey about their skills and interests, which enabled us to identify current needs in machine learning training for medical problem applications. This work describes opportunities and limitations in medical data science in the Israeli context. | ['Joachim A. Behar', 'Danny Eytan', 'Michal Gaziel-Yablowitz', 'Leo Anthony Celi', 'Ronit Almog', 'Jonathan A. Sobel'] | 2022-07-16 | null | null | null | null | ['machine-learning', 'machine-learning'] | ['methodology', 'miscellaneous'] | [ 3.10716391e-01 -1.20324746e-01 -2.07711980e-01 -6.72886074e-01
-5.97733855e-01 -2.41599411e-01 5.13250334e-03 9.39851701e-01
-6.41017377e-01 6.44238234e-01 9.69254002e-02 -5.42718768e-01
-4.80688542e-01 -4.31875020e-01 -2.73346275e-01 -4.60762322e-01
-2.48963386e-01 8.51671398e-01 -2.48340353e-01 5.03056124e-02
2.93080360e-01 8.34905386e-01 -1.22000539e+00 5.40896833e-01
6.82394326e-01 7.30473399e-01 1.72601148e-01 6.36156261e-01
1.86110497e-01 5.78678787e-01 -4.36642826e-01 1.00265108e-01
1.67330444e-01 -5.58001399e-01 -1.12743092e+00 -2.00836346e-01
-4.37212028e-02 2.51779288e-01 -1.99543331e-02 5.50749242e-01
6.61815941e-01 9.21267122e-02 3.15100074e-01 -9.85465646e-01
3.05228770e-01 -2.14366540e-02 -6.91505820e-02 6.29870892e-01
1.87931642e-01 2.65842885e-01 7.21562803e-01 -6.46393657e-01
8.85305762e-01 8.79211426e-01 5.19540191e-01 7.07051754e-01
-1.24579370e+00 -4.89153594e-01 -8.75416622e-02 3.40757728e-01
-1.10488975e+00 -4.99452025e-01 5.44816077e-01 -6.37322247e-01
7.23233163e-01 4.36598510e-01 8.20924342e-01 5.44585586e-01
1.99165598e-01 5.97214103e-01 1.02024603e+00 -3.40456426e-01
3.60071987e-01 3.12130064e-01 1.91443980e-01 5.35646200e-01
1.40313551e-01 7.70627856e-02 -6.37247324e-01 -2.48533815e-01
5.86724877e-01 -2.53632702e-02 -3.18640858e-01 -4.71041411e-01
-1.38055730e+00 8.22370350e-01 2.99460322e-01 7.24508822e-01
-2.80572206e-01 -2.44813770e-01 5.10098338e-01 4.41008508e-01
3.72482054e-02 5.97783148e-01 -7.10758805e-01 -1.82980105e-01
-6.83087111e-01 4.27146703e-01 6.86177254e-01 3.66299689e-01
1.68240681e-01 -3.30483913e-01 2.42788061e-01 6.95603728e-01
2.83911318e-01 1.55048013e-01 3.90832871e-01 -9.21330333e-01
1.94096431e-01 6.77776337e-01 4.29105107e-03 -8.91799033e-01
-7.94319212e-01 -1.80814654e-01 -9.80043173e-01 1.41949475e-01
6.35569632e-01 -4.47976589e-01 -6.54570758e-01 1.37397337e+00
3.73679876e-01 -9.40122753e-02 8.89731348e-02 1.14651525e+00
6.42645359e-01 3.06634694e-01 1.87705114e-01 -4.90792781e-01
1.37693870e+00 -4.84244861e-02 -5.90638816e-01 -1.01674311e-01
1.10601485e+00 -6.32058382e-01 4.62490976e-01 8.35953653e-01
-1.31259298e+00 -5.95705807e-01 -8.31287503e-01 5.17529855e-03
-2.41108418e-01 -5.13254941e-01 5.94562054e-01 5.47324777e-01
-6.57228887e-01 6.53146029e-01 -1.19525743e+00 -8.73119771e-01
6.52572036e-01 5.19683242e-01 -4.61000413e-01 -2.89392531e-01
-1.25025320e+00 1.14824653e+00 3.57848376e-01 -1.19010676e-02
-6.21678174e-01 -1.02838051e+00 -5.33116519e-01 -3.92283618e-01
-4.14508693e-02 -7.35587597e-01 7.40745842e-01 -5.17329335e-01
-5.77036262e-01 1.27948737e+00 -1.01802321e-02 -4.30506229e-01
3.72651309e-01 1.44132361e-01 -4.83358234e-01 -7.86331892e-02
-9.02776122e-02 5.43203354e-01 -7.10173771e-02 -5.76286197e-01
-8.34918678e-01 -8.23110640e-01 -2.14109570e-01 1.23971097e-01
2.95141302e-02 1.25171646e-01 -9.38801467e-03 -3.71509492e-01
1.34233028e-01 -8.02632332e-01 -4.42682058e-01 -1.97697133e-02
1.44420592e-02 -9.62264389e-02 5.44827163e-01 -6.54440522e-01
1.00976312e+00 -2.04201984e+00 3.76790434e-01 3.42971683e-01
1.32745862e-01 2.43712664e-01 1.57853872e-01 6.75888300e-01
-3.04592401e-01 1.59255669e-01 -5.37994243e-02 2.07134753e-01
-5.19789398e-01 2.72403300e-01 1.60429120e-01 6.24979973e-01
7.53825381e-02 7.21841753e-01 -8.62116933e-01 -4.77801085e-01
1.83135986e-01 1.24261834e-01 -4.75181609e-01 2.97193795e-01
-3.17582604e-03 9.69027758e-01 -6.23561621e-01 6.36834383e-01
1.41259372e-01 -2.84860373e-01 5.62625885e-01 -1.42435834e-01
-1.51067093e-01 1.67320922e-01 -8.81277144e-01 1.81748390e+00
-6.27123937e-02 5.19573390e-01 3.19327593e-01 -1.71782577e+00
8.95665586e-01 6.26178801e-01 1.11688995e+00 -5.65114439e-01
3.82478565e-01 3.44813466e-01 3.83101761e-01 -8.42172682e-01
-7.55497664e-02 -6.08221948e-01 1.55371860e-01 3.26021373e-01
-4.30164248e-01 -2.49203965e-01 2.88917363e-01 -5.42620979e-02
1.08452106e+00 -2.87009656e-01 2.87713677e-01 -6.40976429e-01
4.56468165e-01 4.29196209e-01 7.23848939e-01 2.59708911e-01
-5.27404428e-01 4.67356771e-01 3.59613091e-01 -8.82644713e-01
-1.02576470e+00 -7.65415549e-01 -5.05376101e-01 5.22911847e-01
-2.64540553e-01 -8.14654008e-02 -4.58944261e-01 -5.16994774e-01
8.47866759e-02 3.74394476e-01 -4.62931067e-01 -2.51547009e-01
-6.77896798e-01 -6.34030938e-01 1.99741095e-01 4.66371715e-01
1.62340462e-01 -1.17928755e+00 -1.07325518e+00 2.21288651e-01
-2.92315453e-01 -8.10286164e-01 -2.69966751e-01 4.56423730e-01
-1.38791299e+00 -1.23351848e+00 -6.94707215e-01 -9.39169049e-01
4.26296532e-01 -2.48250201e-01 9.41388905e-01 1.17094465e-01
-1.11534798e+00 2.38989577e-01 -2.01976880e-01 -9.24162149e-01
-5.27824938e-01 1.10465482e-01 5.16178422e-02 -3.59933585e-01
3.96273464e-01 -3.46554756e-01 -8.65075767e-01 3.96205515e-01
-9.28434610e-01 -7.78087899e-02 4.32451844e-01 8.50359976e-01
4.13233429e-01 -2.58558005e-01 8.84908557e-01 -1.18746316e+00
4.25531030e-01 -4.50809091e-01 -4.87213045e-01 2.20214844e-01
-6.87629342e-01 -2.83943146e-01 4.18611392e-02 -4.27037179e-02
-5.43932319e-01 1.42546788e-01 -4.30197753e-02 -1.49943531e-01
-2.79140621e-01 9.21878397e-01 -1.69996634e-01 1.41900316e-01
8.14323723e-01 -1.54190198e-01 4.64588761e-01 -3.32812846e-01
-1.88286826e-01 6.27997816e-01 4.90636021e-01 -6.28965139e-01
3.07641745e-01 2.39260644e-01 3.34504247e-01 -9.18766856e-01
-5.12213945e-01 -6.55876100e-01 -7.06032991e-01 -2.68355906e-01
9.81701434e-01 -6.43986881e-01 -8.05963039e-01 2.38624364e-01
-8.94283593e-01 -1.34881169e-01 -4.03039306e-01 6.99803352e-01
-4.55375969e-01 1.33810490e-02 -3.14968467e-01 -5.12305439e-01
-1.33571789e-01 -8.87966990e-01 5.78054786e-01 1.68218333e-02
-7.45493054e-01 -1.27743959e+00 1.42732114e-01 9.11122322e-01
4.29319292e-01 5.42116880e-01 1.39088774e+00 -7.22259879e-01
-4.29403365e-01 -4.29153860e-01 1.67892322e-01 3.58308226e-01
2.27987468e-01 -3.19988906e-01 -7.60720491e-01 -4.63158697e-01
1.85866982e-01 -4.98856634e-01 3.10415894e-01 4.15326297e-01
1.15913081e+00 2.21494168e-01 -5.45012951e-01 1.00006415e-02
1.15878761e+00 6.47610962e-01 4.88084644e-01 -4.03835103e-02
1.37486428e-01 9.24760401e-01 8.32281291e-01 5.33564448e-01
1.78940445e-01 4.36958700e-01 1.06754802e-01 -3.21491569e-01
2.82886595e-01 2.03872979e-01 -2.17644483e-01 6.89801157e-01
-1.67618930e-01 -7.84065798e-02 -1.36352229e+00 6.94523692e-01
-1.81810391e+00 -6.28696024e-01 -2.21549898e-01 2.43381143e+00
9.26527083e-01 2.76922017e-01 3.07159930e-01 3.10029864e-01
4.57153559e-01 -5.20939410e-01 -5.11563063e-01 -3.59572411e-01
2.17457756e-01 3.68937761e-01 1.60056010e-01 4.43947494e-01
-7.30085611e-01 3.32219720e-01 6.42144012e+00 3.70886803e-01
-9.60810244e-01 -2.90686131e-01 9.11361039e-01 -1.15798853e-01
1.93477571e-02 -2.58016020e-01 -3.28602731e-01 1.45629555e-01
1.22110343e+00 -3.75902444e-01 2.06750840e-01 3.31432402e-01
5.56444943e-01 -2.40912765e-01 -1.42935312e+00 8.69568348e-01
-2.64853477e-01 -1.45250630e+00 -5.42439342e-01 2.16830462e-01
4.06106740e-01 -4.42433059e-02 -1.75946821e-02 -1.02791831e-01
1.35052735e-02 -1.30163610e+00 -2.56624874e-02 8.14082086e-01
5.66753209e-01 -6.86761260e-01 7.83807814e-01 5.71663141e-01
-5.83639085e-01 -2.32860018e-02 1.22583590e-01 -1.50287643e-01
3.06006726e-02 6.96074009e-01 -1.09943080e+00 4.56141233e-01
5.02995491e-01 3.44380140e-01 -6.25587404e-02 1.03373790e+00
4.50092554e-01 4.91236299e-01 -2.16883421e-01 -4.70108949e-02
-9.56346095e-02 3.98649424e-02 3.23323220e-01 9.83777046e-01
-7.21019134e-02 4.48750138e-01 3.01652968e-01 3.93160284e-01
2.60586381e-01 1.07609898e-01 -4.29108292e-01 -5.27337790e-01
1.55019313e-01 1.07097089e+00 -7.00490654e-01 -9.37250480e-02
-2.90214986e-01 3.38333637e-01 -5.30159436e-02 8.33901912e-02
-5.34494460e-01 -2.17910260e-01 7.51263976e-01 6.55277550e-01
-3.23814362e-01 -7.92638361e-02 -4.93166029e-01 -6.62150741e-01
-3.58166635e-01 -1.31709683e+00 7.54287899e-01 -4.39084530e-01
-1.23606539e+00 2.57203579e-01 1.12546980e-01 -9.45645034e-01
-2.59522617e-01 -6.41030133e-01 -1.24076605e-01 1.25704753e+00
-1.10004056e+00 -6.18188977e-01 -1.39939070e-01 4.70774829e-01
3.86983663e-01 -1.64523855e-01 9.98870850e-01 5.68557382e-01
-3.64452511e-01 1.53706282e-01 1.50836417e-02 1.56532168e-01
6.96023464e-01 -9.38081920e-01 7.19445646e-02 1.35777041e-01
-9.37523767e-02 6.85856164e-01 8.18493366e-01 -6.31349862e-01
-1.46884775e+00 -8.29544842e-01 1.39167798e+00 -4.55249876e-01
3.15533280e-01 -8.82233381e-02 -8.53976250e-01 4.51680213e-01
-7.97160938e-02 -4.05564904e-02 1.11511469e+00 1.05255835e-01
3.88860732e-01 -2.66064823e-01 -1.32206750e+00 1.34749785e-01
5.91562808e-01 -3.96920294e-01 -5.15100718e-01 5.85899115e-01
3.86403911e-02 -5.60857058e-01 -1.11296618e+00 6.27935886e-01
5.66508353e-01 -6.27295136e-01 6.85537517e-01 -1.31547689e+00
1.67732954e-01 -1.19281486e-01 -3.41512868e-03 -9.37635720e-01
-1.39402673e-01 -4.61939156e-01 5.79002321e-01 6.19943082e-01
3.78577143e-01 -3.78113300e-01 1.12865007e+00 9.12912011e-01
3.74246277e-02 -1.05706632e+00 -8.51795137e-01 -3.89797926e-01
3.04300815e-01 -7.17456818e-01 -3.32187824e-02 9.62909162e-01
4.16892827e-01 2.36688688e-01 -3.31997313e-02 -1.20866872e-01
2.92427868e-01 3.12848017e-02 5.15928924e-01 -1.57546985e+00
-1.18440114e-01 -2.30844095e-01 -6.72701359e-01 -3.83529156e-01
-4.66012061e-01 -9.37386274e-01 -1.28348693e-01 -1.72657013e+00
1.16341025e-01 -5.05093992e-01 -4.66340452e-01 3.64281833e-01
3.84299085e-02 7.15168118e-02 2.92358398e-02 1.71199441e-01
-8.94246623e-02 -3.30449253e-01 1.23513365e+00 -5.44562675e-02
-1.51724055e-01 3.44498217e-01 -7.71622598e-01 4.87557173e-01
8.61496747e-01 -4.33959633e-01 -4.49001223e-01 -1.87266037e-01
7.84560144e-02 4.64758247e-01 2.82318681e-01 -1.11372006e+00
2.61517137e-01 -4.43366200e-01 5.03419757e-01 -5.22110343e-01
3.03958923e-01 -1.05732381e+00 4.71698999e-01 1.06158483e+00
-5.54700732e-01 1.13510676e-01 3.29343379e-01 1.91231608e-01
-2.25654691e-01 -1.19576328e-01 9.94343102e-01 -4.37214285e-01
-2.75811493e-01 2.62086809e-01 -4.43900615e-01 2.57649958e-01
1.40138173e+00 2.65119690e-03 2.16218662e-02 -1.81839094e-01
-1.26325560e+00 6.27681434e-01 1.35994121e-01 2.33026266e-01
5.44400394e-01 -1.00759852e+00 -8.55742812e-01 3.48852575e-01
1.49521427e-02 2.69901808e-02 4.58018124e-01 1.25651348e+00
-7.32941329e-01 5.05552173e-01 -3.34249079e-01 -5.55673242e-01
-1.49801219e+00 4.29820031e-01 2.41693139e-01 -3.90811354e-01
-4.38563704e-01 7.74650633e-01 -4.27620485e-02 -2.12171525e-01
3.34542781e-01 -3.41413647e-01 -8.30954537e-02 2.14283943e-01
4.67877984e-01 3.77042919e-01 4.34764832e-01 -2.04773948e-01
-5.98331392e-01 2.49605313e-01 -1.09233626e-03 -5.46001829e-02
1.38399923e+00 8.05929303e-02 1.45791575e-01 7.80081987e-01
1.12163067e+00 -6.49897397e-01 -4.51696128e-01 8.97484422e-02
2.88579673e-01 -2.13770628e-01 4.98392694e-02 -1.23478091e+00
-8.83337200e-01 1.24667144e+00 8.13532531e-01 3.41045618e-01
1.15134418e+00 -6.40459657e-02 6.61560118e-01 3.52540433e-01
1.95080653e-01 -1.11591625e+00 -8.30159709e-02 1.64364398e-01
9.15262997e-01 -1.11555982e+00 8.00174847e-02 -3.25529158e-01
-7.95038044e-01 1.28802979e+00 2.73011625e-01 3.31997007e-01
8.47269773e-01 2.74574578e-01 3.84063691e-01 -3.63589346e-01
-8.09566379e-01 9.88161862e-02 2.87196457e-01 7.50636041e-01
7.55931497e-01 -1.89644486e-01 -5.92039526e-01 1.94662973e-01
1.97674900e-01 6.53308034e-01 1.71980709e-01 1.41765487e+00
-5.45840561e-01 -1.37615836e+00 -3.49720180e-01 7.11184502e-01
-5.29539585e-01 1.79365113e-01 -2.91730791e-01 8.47277343e-01
2.23276228e-01 6.34940326e-01 -2.28019934e-02 -4.41914536e-02
5.99966586e-01 3.71227533e-01 4.96200562e-01 -5.94142795e-01
-6.82290673e-01 1.56137079e-01 2.17768908e-01 -2.67174363e-01
-4.01093036e-01 -9.69883502e-01 -1.49212980e+00 -5.28824814e-02
1.71688229e-01 6.37284219e-01 7.78017163e-01 8.20347607e-01
3.76593739e-01 7.00066924e-01 4.56143916e-01 -2.95612365e-01
-3.00113231e-01 -7.89201021e-01 -7.20569313e-01 3.03838015e-01
2.26099402e-01 -2.58065492e-01 1.17529303e-01 4.14173692e-01] | [8.138676643371582, 6.965848922729492] |
c116dee4-e9a4-4d08-bd65-d54e03ae267e | pliks-a-pseudo-linear-inverse-kinematic | 2211.11734 | null | https://arxiv.org/abs/2211.11734v2 | https://arxiv.org/pdf/2211.11734v2.pdf | PLIKS: A Pseudo-Linear Inverse Kinematic Solver for 3D Human Body Estimation | We introduce PLIKS (Pseudo-Linear Inverse Kinematic Solver) for reconstruction of a 3D mesh of the human body from a single 2D image. Current techniques directly regress the shape, pose, and translation of a parametric model from an input image through a non-linear mapping with minimal flexibility to any external influences. We approach the task as a model-in-the-loop optimization problem. PLIKS is built on a linearized formulation of the parametric SMPL model. Using PLIKS, we can analytically reconstruct the human model via 2D pixel-aligned vertices. This enables us with the flexibility to use accurate camera calibration information when available. PLIKS offers an easy way to introduce additional constraints such as shape and translation. We present quantitative evaluations which confirm that PLIKS achieves more accurate reconstruction with greater than 10% improvement compared to other state-of-the-art methods with respect to the standard 3D human pose and shape benchmarks while also obtaining a reconstruction error improvement of 12.9 mm on the newer AGORA dataset. | ['Bernhard Egger', 'Andreas Maier', 'Markus Kowarschik', 'Norbert Strobel', 'Srikrishna Jaganathan', 'Annette Birkhold', 'Karthik Shetty'] | 2022-11-21 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Shetty_PLIKS_A_Pseudo-Linear_Inverse_Kinematic_Solver_for_3D_Human_Body_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Shetty_PLIKS_A_Pseudo-Linear_Inverse_Kinematic_Solver_for_3D_Human_Body_CVPR_2023_paper.pdf | cvpr-2023-1 | ['camera-calibration'] | ['computer-vision'] | [ 8.09696689e-02 3.14612865e-01 1.53900132e-01 -1.21037431e-01
-5.39263010e-01 -4.16680068e-01 3.92877251e-01 -3.49236399e-01
-3.58751714e-01 4.47330356e-01 -5.58364242e-02 1.78780973e-01
1.50380835e-01 -3.41040373e-01 -9.95763659e-01 -3.08049649e-01
3.34271997e-01 1.22459555e+00 2.53479958e-01 -2.36488551e-01
-1.74831182e-01 5.33273876e-01 -1.14379203e+00 -2.20776290e-01
4.60720062e-01 7.77600348e-01 3.15499073e-03 6.00634396e-01
4.64985698e-01 1.83550090e-01 -1.25795811e-01 -3.34885389e-01
6.61669195e-01 -2.68098176e-01 -6.31985784e-01 4.05527949e-01
5.89180648e-01 -4.96594489e-01 -7.62029216e-02 5.49931347e-01
5.57010174e-01 -7.12986588e-02 5.03127813e-01 -9.30976033e-01
-2.53142770e-02 -1.73144832e-01 -7.53132105e-01 -4.36270744e-01
8.64621580e-01 1.60754517e-01 4.47365463e-01 -9.56878543e-01
1.12078869e+00 1.28771687e+00 1.13541186e+00 4.29888159e-01
-1.77544594e+00 -2.51620352e-01 -2.62273848e-01 -2.95299143e-01
-1.56814289e+00 -3.33359241e-01 7.47259319e-01 -7.38395393e-01
7.29766130e-01 3.87194961e-01 1.07706130e+00 8.39074850e-01
5.16622424e-01 4.28110749e-01 1.09001303e+00 -4.77115273e-01
-1.03634484e-01 -3.85030918e-02 -1.44267350e-01 9.13164794e-01
1.84894383e-01 2.83837080e-01 -4.96000320e-01 -2.21852899e-01
1.41450584e+00 -2.71388888e-01 -1.10264599e-01 -1.02463019e+00
-1.43834186e+00 3.72170568e-01 3.32150370e-01 -4.39301014e-01
-3.54862541e-01 3.20947021e-01 4.19581085e-02 -4.36635278e-02
3.39743137e-01 4.01050061e-01 -4.52099413e-01 -1.76056013e-01
-7.36118972e-01 5.96349061e-01 6.83119416e-01 9.91790652e-01
6.69907331e-01 -2.83186082e-02 3.98410082e-01 5.52871525e-01
6.44194961e-01 8.35829377e-01 -4.84528579e-02 -1.46357059e+00
2.41900623e-01 5.58660269e-01 2.92828470e-01 -9.61463094e-01
-7.14054763e-01 -3.04019570e-01 -5.40289521e-01 6.34252131e-01
5.02893984e-01 -1.59144387e-01 -1.01199353e+00 1.35397398e+00
9.02237773e-01 -9.25371423e-02 -3.73158783e-01 1.32745504e+00
4.61514086e-01 4.31515664e-01 -3.96205246e-01 -3.11756115e-02
1.29322958e+00 -8.09853613e-01 -4.37506407e-01 -6.01445019e-01
1.24640733e-01 -8.42385411e-01 9.40978944e-01 5.69837928e-01
-1.36380708e+00 -3.88166308e-01 -1.02540147e+00 -3.16305131e-01
3.46346021e-01 2.33045667e-01 2.85806954e-01 3.02724123e-01
-7.87509143e-01 6.52856052e-01 -1.26976705e+00 -2.77427912e-01
-2.38204166e-01 7.07453728e-01 -7.49431670e-01 1.32848188e-01
-5.76641381e-01 1.24540710e+00 8.45552236e-03 1.13916218e-01
-5.37655711e-01 -1.00316048e+00 -1.01883757e+00 -6.38025761e-01
5.85840523e-01 -1.28828490e+00 1.39022481e+00 -6.16748810e-01
-2.09551263e+00 1.21917057e+00 -5.62025979e-02 -1.57122284e-01
1.03941751e+00 -5.01793861e-01 1.72536999e-01 6.29930273e-02
-5.16769364e-02 5.03100872e-01 9.42396045e-01 -1.40948832e+00
2.03147918e-01 -4.08830762e-01 -1.29837364e-01 3.67476523e-01
4.80280995e-01 -2.18688339e-01 -9.73315597e-01 -5.83103299e-01
5.62717021e-01 -1.51551819e+00 -3.95025730e-01 6.07153416e-01
-5.29377759e-01 6.09063804e-01 6.63549006e-01 -1.15904915e+00
7.81158864e-01 -1.70295012e+00 7.26754010e-01 3.71916175e-01
9.81936306e-02 -1.72504112e-02 3.04276794e-01 2.17627794e-01
7.34064952e-02 -3.22975963e-01 -3.03885907e-01 -6.79047704e-01
-1.36985719e-01 3.35550159e-01 3.42293859e-01 8.52374971e-01
-2.12206200e-01 8.53618979e-01 -6.04943335e-01 -6.25526130e-01
4.38222647e-01 7.08735764e-01 -6.17816865e-01 2.80061334e-01
-1.60495475e-01 8.82828355e-01 -3.06693107e-01 5.99773645e-01
5.08925617e-01 -1.21310540e-01 3.43707323e-01 -5.41221142e-01
2.07282286e-02 -8.61836672e-02 -1.35333860e+00 2.09893942e+00
-4.86848980e-01 2.21157238e-01 4.54316765e-01 -4.96618569e-01
8.69006157e-01 3.48750740e-01 6.71089768e-01 -4.92606699e-01
3.44246268e-01 1.42842129e-01 -1.76622719e-01 -2.30098128e-01
1.99394763e-01 -4.12497103e-01 -4.40507196e-02 2.96338052e-01
-8.71700421e-02 -5.05857706e-01 -4.29473668e-01 -1.43574148e-01
6.96511388e-01 8.57788682e-01 4.02434975e-01 -3.64415437e-01
3.14873219e-01 2.23824412e-01 3.78999323e-01 2.90948469e-02
3.93005520e-01 1.03042030e+00 1.99082002e-01 -4.75949913e-01
-1.34421933e+00 -1.31575215e+00 -1.10055186e-01 3.05184931e-01
1.19289093e-01 -3.48441005e-01 -9.05702710e-01 -1.17784321e-01
3.53073180e-01 3.28698993e-01 -6.64066970e-01 1.33178756e-01
-8.57389033e-01 -3.63930136e-01 1.89414576e-01 5.11707544e-01
2.50208080e-01 -5.53623438e-01 -9.78773296e-01 1.71673864e-01
-5.76572679e-02 -1.32763171e+00 -5.94553471e-01 -1.81475028e-01
-1.05448568e+00 -1.06751347e+00 -6.41755581e-01 -5.37344694e-01
9.87618625e-01 -4.21634018e-01 1.15288699e+00 -1.47381816e-02
-4.86223102e-01 6.65321052e-01 1.31294951e-01 -1.69771299e-01
-4.68257904e-01 -2.10513711e-01 2.00111032e-01 -1.64721906e-01
-5.42466044e-01 -5.95315993e-01 -5.83421826e-01 6.90060735e-01
-4.98194158e-01 5.83271623e-01 1.92724422e-01 6.16833150e-01
1.12599647e+00 -4.92619216e-01 -5.63570671e-02 -5.92376292e-01
1.27763391e-01 5.63881360e-02 -7.56767333e-01 -7.86463618e-02
-4.35843199e-01 1.57438710e-01 3.05093825e-01 -5.15644908e-01
-1.02696669e+00 9.72981334e-01 -1.75667554e-01 -7.01840937e-01
8.17323029e-02 3.31097811e-01 -9.70350131e-02 -3.15445334e-01
6.16585374e-01 -1.54732183e-01 7.08751202e-01 -8.15989077e-01
3.46679062e-01 1.03045821e-01 9.39525247e-01 -7.36338258e-01
9.17595923e-01 6.84882402e-01 3.95582318e-01 -8.11873019e-01
-5.19500494e-01 -3.24907959e-01 -1.36500525e+00 -3.97771597e-01
7.35732734e-01 -8.06078196e-01 -7.69724786e-01 4.83519077e-01
-9.69558299e-01 -5.51707864e-01 -2.41908073e-01 3.79424423e-01
-9.20887113e-01 4.45584565e-01 -4.95183975e-01 -5.37310958e-01
-3.41151893e-01 -1.17868650e+00 1.39845276e+00 -3.04059356e-01
-6.82319880e-01 -8.74639034e-01 1.85500354e-01 5.81671596e-01
-3.60626243e-02 9.76182103e-01 4.08767402e-01 1.66959733e-01
-3.63631010e-01 -3.45340520e-01 5.10933697e-01 1.57040566e-01
-3.65382107e-03 3.15850489e-02 -3.99305195e-01 -4.54887956e-01
-3.08615603e-02 -8.01229551e-02 1.21280983e-01 3.53877157e-01
5.00091493e-01 -2.54864752e-01 -4.12286609e-01 8.38218987e-01
1.36387336e+00 -4.04694259e-01 5.75945497e-01 4.56928104e-01
9.15848911e-01 4.05039728e-01 6.67754114e-01 3.80836338e-01
4.25765634e-01 1.42587125e+00 4.87191319e-01 -4.84778360e-02
-2.07301602e-01 -5.28736234e-01 2.60185063e-01 6.59852207e-01
-3.43747348e-01 5.34346163e-01 -1.15487492e+00 2.39877641e-01
-1.83484638e+00 -2.20655158e-01 -4.21880245e-01 2.47971869e+00
7.71011114e-01 3.14823911e-02 3.26644659e-01 -5.93798459e-02
3.67056638e-01 -2.98957795e-01 -5.38697600e-01 -1.98510975e-01
3.80093127e-01 7.40316659e-02 6.42674446e-01 7.05920756e-01
-8.15543175e-01 7.99904585e-01 6.71996164e+00 1.63832292e-01
-1.16352057e+00 5.59253944e-03 3.36491019e-02 -2.58781493e-01
1.49532836e-02 2.76003405e-02 -7.07609415e-01 1.78382844e-01
7.32340991e-01 -6.94131032e-02 4.98507112e-01 7.22650290e-01
3.47444355e-01 -2.69343436e-01 -1.17339230e+00 1.01919615e+00
1.69464394e-01 -1.13351166e+00 -2.56018937e-01 2.97416806e-01
6.43190205e-01 -1.49618492e-01 -2.44560331e-01 -2.74482399e-01
-6.17527701e-02 -8.69681478e-01 1.23736906e+00 7.33114719e-01
9.04618919e-01 -5.78720689e-01 3.71102661e-01 7.45485842e-01
-9.62065041e-01 5.18014133e-01 -1.52381912e-01 -8.98049697e-02
6.88429773e-01 3.99362385e-01 -9.47901309e-01 5.68658173e-01
6.18975163e-01 5.97435594e-01 -4.40037251e-01 8.37137043e-01
-1.76519826e-01 3.57024729e-01 -7.74536014e-01 6.27722085e-01
-3.79703701e-01 -5.24297237e-01 7.56033003e-01 8.13278973e-01
1.94938034e-01 1.10295527e-01 3.52087766e-01 8.69998336e-01
2.07163736e-01 9.51822847e-02 -3.88068140e-01 4.10942137e-01
-9.69616510e-03 1.04765379e+00 -6.59232795e-01 -4.26517837e-02
7.19467774e-02 1.15176427e+00 1.00107826e-01 -3.13961785e-03
-8.46578181e-01 3.71684968e-01 5.32657921e-01 8.27470958e-01
1.88516214e-01 -6.36993647e-01 -4.51882482e-01 -9.32070911e-01
2.34558180e-01 -8.77599895e-01 -2.17890386e-02 -9.97178733e-01
-7.38488972e-01 3.90798241e-01 3.86368513e-01 -1.21936643e+00
-6.37512445e-01 -6.98328435e-01 -4.83790152e-02 7.49401033e-01
-8.99371862e-01 -1.36129761e+00 -2.91647732e-01 5.16353011e-01
2.91714817e-01 4.11231011e-01 7.72197604e-01 -1.24644507e-02
-1.66489720e-01 3.22693616e-01 -2.50914603e-01 -1.17122948e-01
6.33049667e-01 -1.02741742e+00 6.19263828e-01 5.50385773e-01
-1.65987328e-01 5.98597229e-01 1.00177920e+00 -8.83309841e-01
-1.98739851e+00 -6.73458397e-01 4.10208941e-01 -8.79957795e-01
2.92331666e-01 -4.43306118e-01 -6.99855566e-01 1.11566067e+00
-2.67321110e-01 2.73993462e-01 1.32518083e-01 -2.57855207e-01
1.02217318e-02 4.67314534e-02 -1.26257527e+00 4.90374357e-01
1.00206161e+00 -1.67572349e-01 -6.47864223e-01 1.63047329e-01
4.22015727e-01 -1.35482073e+00 -1.20594454e+00 4.99085039e-01
1.07662356e+00 -7.35413730e-01 1.48807669e+00 -1.12150587e-01
1.58921704e-01 -6.26938701e-01 7.87147507e-02 -1.17877638e+00
-2.17687666e-01 -8.70473564e-01 -2.18460277e-01 6.49989426e-01
2.04945743e-01 -3.98207694e-01 8.54224145e-01 7.48186767e-01
-1.54317031e-02 -7.32566714e-01 -1.07316673e+00 -8.06385994e-01
-1.23969987e-01 -3.71287137e-01 1.09999061e-01 5.71361423e-01
-1.43056586e-01 1.86806515e-01 -6.93622947e-01 3.78402174e-01
8.91859353e-01 -6.53026104e-02 1.27266443e+00 -1.32742858e+00
-4.60921675e-01 2.60195762e-01 -4.98074561e-01 -1.10431039e+00
2.94679832e-02 -5.79685867e-01 2.21067742e-01 -1.51657712e+00
-3.10684200e-02 -1.76358134e-01 5.70628166e-01 4.89029288e-01
1.71292216e-01 5.47570407e-01 2.23464444e-01 1.54286653e-01
-9.60798636e-02 2.92922586e-01 1.45635080e+00 3.09968472e-01
-2.20282868e-01 1.60462596e-02 -1.32617921e-01 1.03017426e+00
5.14092803e-01 -4.12774742e-01 -2.27379188e-01 -5.70389628e-01
1.18961588e-01 3.72318238e-01 6.29925013e-01 -1.08962083e+00
8.19735229e-02 -1.22542508e-01 4.77287918e-01 -5.59743226e-01
8.29757035e-01 -1.10334945e+00 1.15028250e+00 5.87198913e-01
7.01242536e-02 2.48903543e-01 2.54265577e-01 3.15302789e-01
3.34313512e-01 7.24802613e-02 8.53739440e-01 -2.76052117e-01
-4.09303725e-01 8.80298167e-02 1.95474830e-02 -6.71412647e-02
8.74845266e-01 -5.39749861e-01 2.69276470e-01 -3.20197880e-01
-9.88329113e-01 -2.99661551e-02 1.21152997e+00 3.85018408e-01
5.55495858e-01 -1.26548564e+00 -5.67799985e-01 4.30078685e-01
-1.55737609e-01 3.49640787e-01 1.03663905e-02 9.48450685e-01
-9.61336076e-01 2.40369484e-01 -3.41314137e-01 -9.75679636e-01
-1.55197978e+00 2.94993788e-01 7.48862982e-01 -1.50745288e-01
-1.03782141e+00 3.33944261e-01 -1.10443078e-01 -8.54072571e-01
-1.75561905e-01 -4.57744658e-01 4.94728059e-01 -5.66951096e-01
9.88665689e-03 6.23980343e-01 1.10498011e-01 -1.12641680e+00
-5.22818446e-01 1.13119578e+00 5.18198669e-01 -4.44766581e-01
1.33476520e+00 -9.57081467e-02 5.29124960e-02 2.89514393e-01
1.06621575e+00 1.96186498e-01 -1.62945962e+00 -2.77181081e-02
-2.85443783e-01 -4.42555904e-01 7.42365494e-02 -9.34381902e-01
-9.83433604e-01 5.49783885e-01 3.77509147e-01 -5.27755260e-01
6.63613439e-01 3.07539161e-02 7.89085090e-01 4.05639084e-03
6.28603518e-01 -9.47229564e-01 -1.02597773e-01 2.74539292e-01
1.36637855e+00 -7.96315908e-01 6.77207172e-01 -8.06914449e-01
-5.56103468e-01 1.03937590e+00 4.49999094e-01 -3.79070401e-01
5.69411099e-01 5.34342051e-01 2.46591285e-01 -2.26068258e-01
-2.02962756e-01 3.09688270e-01 7.67363191e-01 5.85795224e-01
2.03395635e-01 3.35677452e-02 -1.15478210e-01 2.82814473e-01
-4.77282643e-01 9.12194699e-02 3.07468981e-01 1.03752112e+00
-7.96360224e-02 -1.28664446e+00 -8.32638800e-01 4.86083403e-02
-1.14534110e-01 5.37694156e-01 -3.53158057e-01 1.09681416e+00
-1.43426657e-01 3.03436071e-01 -9.71591622e-02 -3.22706729e-01
8.27281296e-01 -4.67399396e-02 8.97062898e-01 -5.79278708e-01
-4.42096859e-01 4.57355082e-01 1.76122382e-01 -9.83193398e-01
-2.56501436e-01 -9.85071480e-01 -1.39522576e+00 -1.05345994e-01
-2.08393380e-01 -3.24822217e-01 8.83939981e-01 8.26171041e-01
4.66234386e-01 2.65294522e-01 -5.21169379e-02 -1.56487775e+00
-5.55225015e-01 -4.49201167e-01 -3.82839054e-01 3.66760224e-01
2.94028580e-01 -1.05342615e+00 -5.19448183e-02 2.74589330e-01] | [7.112146377563477, -1.1013838052749634] |
488aecce-dc31-494f-9de1-2a4aed797a97 | embedding-mental-health-discourse-for | 2307.03892 | null | https://arxiv.org/abs/2307.03892v1 | https://arxiv.org/pdf/2307.03892v1.pdf | Embedding Mental Health Discourse for Community Recommendation | Our paper investigates the use of discourse embedding techniques to develop a community recommendation system that focuses on mental health support groups on social media. Social media platforms provide a means for users to anonymously connect with communities that cater to their specific interests. However, with the vast number of online communities available, users may face difficulties in identifying relevant groups to address their mental health concerns. To address this challenge, we explore the integration of discourse information from various subreddit communities using embedding techniques to develop an effective recommendation system. Our approach involves the use of content-based and collaborative filtering techniques to enhance the performance of the recommendation system. Our findings indicate that the proposed approach outperforms the use of each technique separately and provides interpretability in the recommendation process. | ['Meng Jiang', 'Noah Ziems', 'Bang Nguyen', 'Hy Dang'] | 2023-07-08 | null | null | null | null | ['collaborative-filtering'] | ['miscellaneous'] | [ 8.72811452e-02 7.66157925e-01 -4.06620473e-01 -1.33271024e-01
-1.02870293e-01 -4.07425016e-01 5.86973369e-01 8.75978589e-01
-1.50615260e-01 4.82560009e-01 1.26032209e+00 -3.90225738e-01
-3.96087438e-01 -9.60697472e-01 4.21685070e-01 -1.55299857e-01
1.16272062e-01 1.16286375e-01 1.09216191e-01 -6.59161747e-01
6.11432195e-01 1.42835379e-01 -1.04215372e+00 7.89065480e-01
1.02092421e+00 -1.79558881e-02 -1.71288699e-01 3.04035991e-01
-5.05982041e-01 7.34667659e-01 -3.72267276e-01 -6.96123481e-01
1.56256910e-02 -4.35081631e-01 -6.86020553e-01 -1.28056243e-01
1.64155364e-01 -6.96228206e-01 -1.90582871e-01 9.34944391e-01
7.50804305e-01 1.25002772e-01 5.30912697e-01 -9.71373260e-01
-1.01982868e+00 1.06959581e+00 -2.96077847e-01 3.00702184e-01
9.34974015e-01 -5.83418369e-01 8.23830366e-01 -7.95877934e-01
1.04704988e+00 1.42228186e+00 9.40169454e-01 7.02997148e-01
-1.27441204e+00 -7.39491105e-01 7.52028301e-02 -6.04488179e-02
-9.40626740e-01 -6.71033442e-01 8.30860138e-01 -7.05574095e-01
8.73413682e-01 2.89247811e-01 9.05781984e-01 8.01963687e-01
-2.74788588e-01 2.93161850e-02 8.38517487e-01 -4.70663488e-01
1.43238574e-01 5.41220427e-01 4.54275042e-01 5.12119532e-01
6.00108206e-01 -5.08699954e-01 -5.25427759e-01 -1.14943540e+00
3.53363603e-01 3.12478900e-01 1.13491625e-01 2.70049125e-01
-8.03363800e-01 1.26239705e+00 4.71829504e-01 8.25374603e-01
-4.99898911e-01 -3.61736685e-01 6.29390121e-01 2.14839041e-01
1.13510466e+00 2.57814616e-01 4.80758607e-01 2.99121812e-02
-7.71965623e-01 2.51404464e-01 9.65841591e-01 1.00692421e-01
5.46108484e-02 -4.99771208e-01 -3.42544794e-01 8.82398963e-01
1.02992165e+00 -5.33429980e-02 4.95514125e-02 -9.50101078e-01
3.95789355e-01 1.02870071e+00 -1.43252045e-01 -1.75703537e+00
-1.11916894e-02 -5.83047867e-02 -3.22301507e-01 -1.84481964e-01
2.52328038e-01 -6.37630165e-01 9.71009582e-02 1.26286209e+00
8.38856637e-01 2.80446969e-02 -4.72746678e-02 3.91275406e-01
9.55645800e-01 2.87480503e-01 5.13383508e-01 -3.78137410e-01
1.05093277e+00 -5.74939728e-01 -1.33133340e+00 8.29517543e-02
1.00072777e+00 -7.73272395e-01 1.97990283e-01 -3.92160565e-01
-1.13293707e+00 7.98103958e-02 -7.86082506e-01 5.25925159e-02
-2.52650380e-01 -2.44019061e-01 6.92237437e-01 1.23306024e+00
-1.43678951e+00 4.83439922e-01 -7.12277055e-01 -9.10848916e-01
6.34972334e-01 2.97301382e-01 -3.19386363e-01 -3.56075950e-02
-9.04840589e-01 7.72901773e-01 -9.87240970e-02 -3.11268806e-01
3.04768503e-01 -5.85139751e-01 -6.40702724e-01 -2.86374629e-01
-8.47918242e-02 -5.36206663e-01 5.24926484e-01 -9.82321203e-01
-1.00215304e+00 5.14338255e-01 -2.58609444e-01 -1.63321137e-01
3.83190423e-01 -2.57371273e-02 -4.37158048e-01 5.73598385e-01
2.75621772e-01 2.90403336e-01 3.56918097e-01 -9.22971487e-01
-4.23297733e-01 -5.36549568e-01 3.36112320e-01 1.75669223e-01
-1.22580516e+00 9.24538851e-01 3.45221788e-01 -5.28774083e-01
1.91764265e-01 -6.09890878e-01 -1.65066898e-01 3.87511402e-01
-1.25781327e-01 -1.54961988e-01 1.05526423e+00 -1.15970254e+00
1.63891780e+00 -2.23219466e+00 3.81816961e-02 3.61332059e-01
6.39669120e-01 3.88296902e-01 2.62793005e-01 1.17922747e+00
4.03891206e-01 7.12717474e-01 2.78345257e-01 -3.48263711e-01
-3.03325474e-01 -1.82751238e-01 1.19011536e-01 4.61811662e-01
-1.66444942e-01 5.28378069e-01 -1.13811457e+00 -6.14008307e-01
-1.08787253e-01 8.86544645e-01 -1.01716566e+00 -1.06551744e-01
1.54181942e-01 1.64330557e-01 -7.03880250e-01 2.98638165e-01
4.72444683e-01 -2.94050395e-01 9.98756647e-01 1.60607561e-01
-1.01723701e-01 5.40109932e-01 -9.81614590e-01 9.33497787e-01
3.68315428e-02 7.30476201e-01 4.17368770e-01 -5.95659137e-01
9.32656109e-01 6.50884211e-01 7.91502833e-01 -2.44328920e-02
6.88396618e-02 2.07271539e-02 1.84892595e-01 -7.09292591e-01
6.92017496e-01 -2.65546590e-02 7.11141825e-01 1.12245929e+00
-4.12518024e-01 9.20477033e-01 -1.84641391e-01 7.12740779e-01
1.11834812e+00 -9.18961465e-02 3.47707242e-01 -3.08717042e-01
1.67221844e-01 -1.71095550e-01 4.28934872e-01 3.74169648e-01
-4.90119994e-01 6.11804426e-02 2.10292965e-01 6.41770959e-02
-8.64587367e-01 -3.51688355e-01 -7.63661563e-02 1.07443583e+00
-1.31865814e-01 -9.00128245e-01 -7.12087870e-01 -5.02895474e-01
1.36983156e-01 4.22842443e-01 -6.78141534e-01 8.30834955e-02
-2.36415789e-01 -8.38022649e-01 4.75252688e-01 2.53140748e-01
4.41877216e-01 -8.16357136e-01 -2.30780661e-01 5.15793383e-01
-4.75020647e-01 -4.94730711e-01 -3.48859429e-01 -9.92150247e-01
-1.04710996e+00 -1.17911661e+00 -4.88549411e-01 -6.88239813e-01
1.02881134e+00 7.35363781e-01 3.40082616e-01 7.97139704e-01
1.85815662e-01 6.48274541e-01 -6.85875595e-01 5.16776033e-02
-7.80173600e-01 -1.88556984e-01 -6.25717640e-02 6.23456277e-02
6.90630794e-01 -6.31331980e-01 -7.20573604e-01 1.12464242e-01
-8.68480921e-01 7.10171312e-02 -5.21559238e-01 2.64504254e-01
-5.69185495e-01 -3.43017668e-01 1.19922578e+00 -1.22579360e+00
1.46301877e+00 -1.26064098e+00 4.29802209e-01 8.93541425e-02
-7.61575580e-01 -8.06864381e-01 -1.52191550e-01 -5.43564260e-01
-9.77644980e-01 -5.70454121e-01 8.24409164e-03 5.42850554e-01
2.82927215e-01 9.50537980e-01 5.53639531e-01 -2.20997810e-01
9.01143730e-01 -5.68559885e-01 5.10682881e-01 -4.76251185e-01
3.02020937e-01 1.26215661e+00 -4.82268661e-01 -1.48274831e-03
3.37842643e-01 5.69763243e-01 -6.70344174e-01 -9.43581700e-01
-2.55002856e-01 -6.25881970e-01 -2.84715444e-01 -6.54654026e-01
8.67404222e-01 -8.29403639e-01 -6.36726737e-01 -1.47509918e-01
-1.30043995e+00 1.54192314e-01 6.89348057e-02 4.49903697e-01
4.35102016e-01 6.26171589e-01 -9.02178347e-01 -1.18904448e+00
-6.11705840e-01 -4.54797208e-01 2.22584441e-01 3.49030606e-02
-8.73974502e-01 -1.54469633e+00 4.23700511e-01 1.02832520e+00
8.21260154e-01 4.34995472e-01 7.42897570e-01 -1.06097448e+00
1.41584262e-01 -1.91933945e-01 -4.19986814e-01 -2.31445566e-01
5.70128083e-01 2.84336418e-01 -7.50500560e-01 -4.83409137e-01
-2.84450889e-01 5.06331995e-02 1.01395428e-01 1.12642860e-03
2.00410426e-01 -8.59148860e-01 -5.91471314e-01 -3.36711913e-01
1.19127595e+00 6.55078366e-02 7.10655034e-01 4.25529033e-01
6.12808287e-01 1.04744577e+00 1.15044892e-01 8.91964376e-01
8.53054762e-01 4.60270822e-01 5.18350899e-02 2.48343915e-01
-1.47170708e-01 -3.12231749e-01 4.34713691e-01 9.99282718e-01
-3.24983984e-01 5.48027232e-02 -8.94555390e-01 6.34368241e-01
-2.03926373e+00 -1.18127024e+00 -1.09049775e-01 1.88332951e+00
6.41051233e-01 -2.91785121e-01 3.09864879e-01 2.55634785e-01
1.14268136e+00 -7.46748522e-02 -1.32984447e-03 -5.09795070e-01
3.35649282e-01 -1.60794929e-01 2.47215535e-02 7.25180626e-01
-5.31012595e-01 4.06480521e-01 6.81160641e+00 6.34561852e-02
-1.00087714e+00 4.71808106e-01 3.44089121e-01 2.90743671e-02
-7.38155365e-01 7.68935829e-02 -5.61328888e-01 3.49842310e-01
8.98227692e-01 -3.61138105e-01 4.53067660e-01 2.85420001e-01
5.97561717e-01 1.50637165e-01 -6.29748821e-01 4.18248266e-01
3.13323051e-01 -1.80405378e+00 -2.39406362e-01 4.10271138e-01
9.12612379e-01 -1.97080061e-01 -1.27641663e-01 -2.68987715e-01
2.32438520e-01 -5.65464675e-01 2.57491261e-01 3.12088728e-01
1.79571629e-01 -3.83098781e-01 5.22831738e-01 3.76621783e-02
-8.04577529e-01 -3.70766461e-01 -2.11136475e-01 -5.51243722e-01
3.18052262e-01 2.92308122e-01 -1.20914853e+00 1.97271258e-01
3.06504309e-01 1.08939874e+00 -3.50578040e-01 1.10062194e+00
2.17677891e-01 7.83809960e-01 1.35922432e-03 -1.56571478e-01
-1.37403607e-01 -4.71131712e-01 6.71099067e-01 1.11377728e+00
2.80786514e-01 3.87427300e-01 2.91514676e-02 6.34699345e-01
-2.23584920e-02 7.01465786e-01 -8.98533940e-01 -5.89969277e-01
8.77718329e-01 1.04293728e+00 -9.30244386e-01 -1.11363634e-01
-4.19928789e-01 4.31435287e-01 4.18599844e-01 1.72595128e-01
-4.84765470e-02 9.27889347e-02 5.13260782e-01 8.79297018e-01
-2.44754571e-02 -2.03092471e-01 -2.62379199e-01 -9.25081193e-01
-3.34010631e-01 -9.34680939e-01 3.58415067e-01 -3.65233421e-01
-1.26757300e+00 1.86997503e-01 -1.19395234e-01 -7.17180192e-01
-1.81160972e-01 2.54695714e-01 -8.44776154e-01 6.76673532e-01
-8.73791456e-01 -1.22083104e+00 4.55659553e-02 3.41337949e-01
4.22888808e-02 -2.45777026e-01 1.02200437e+00 6.55633628e-01
-6.21618092e-01 4.53608632e-01 3.41082126e-01 -1.07965216e-01
7.29197741e-01 -5.61653376e-01 -9.29639265e-02 5.97022891e-01
-5.05475938e-01 1.14889646e+00 5.82079291e-01 -1.30803168e+00
-1.17923653e+00 -8.00010324e-01 1.63157964e+00 -2.52014458e-01
7.90077865e-01 -1.00835040e-01 -9.51497138e-01 5.71340263e-01
4.74521190e-01 -4.76619601e-01 1.54956257e+00 5.42362928e-01
-8.12881514e-02 4.55613106e-01 -1.98357701e+00 9.20152724e-01
1.12155032e+00 -9.64528739e-01 -8.95864487e-01 5.48347294e-01
5.49157143e-01 2.64982760e-01 -1.11697209e+00 -4.13429081e-01
5.37986934e-01 -7.34893739e-01 8.15690517e-01 -6.59329414e-01
5.98287463e-01 9.52732842e-03 -3.68205607e-02 -9.46868300e-01
-4.75270718e-01 -7.90003657e-01 1.79528221e-02 1.63651931e+00
3.72240514e-01 -9.57885921e-01 6.27319753e-01 1.36201191e+00
4.18894142e-01 -2.40397185e-01 -5.54116905e-01 3.70633930e-01
-1.73273355e-01 1.05811551e-01 3.35812390e-01 1.74446774e+00
9.46527898e-01 5.39958887e-02 -1.12334184e-01 2.83732377e-02
5.86735964e-01 -4.00182366e-01 3.01034808e-01 -1.38748682e+00
-1.36876687e-01 -1.45437405e-01 -3.80975097e-01 5.70753776e-02
-9.66931283e-02 -1.07848978e+00 -9.73812640e-01 -2.15260100e+00
4.68970239e-01 -6.46346629e-01 1.09541707e-01 4.68009025e-01
7.00552911e-02 4.18295413e-01 3.15540820e-01 2.66929299e-01
-3.57756227e-01 4.99962829e-02 9.50832486e-01 1.02257878e-01
-7.13215947e-01 -2.63201684e-01 -1.51529026e+00 4.09052253e-01
7.23765433e-01 -6.47482216e-01 -6.13659680e-01 -1.27164140e-01
9.18248296e-01 6.66157529e-02 6.80358261e-02 -3.38749945e-01
3.40825945e-01 -5.55689968e-02 -1.77215897e-02 -9.55840275e-02
8.05990100e-02 -6.70127690e-01 4.73062664e-01 6.13733411e-01
-6.98288679e-01 -1.35000095e-01 1.97048187e-01 4.47865129e-01
1.68053448e-01 -3.49506646e-01 1.25603408e-01 2.01162890e-01
2.33570501e-01 -2.81403810e-01 -8.18654954e-01 -3.59704614e-01
9.57173824e-01 -4.66908306e-01 -5.37521362e-01 -6.72896922e-01
-1.16603434e+00 -3.98369282e-02 5.59809387e-01 3.73627692e-01
6.38366699e-01 -1.37055814e+00 -8.19894791e-01 -2.13561773e-01
-9.09928754e-02 -7.58937657e-01 4.85557057e-02 7.62964070e-01
-3.70641768e-01 -4.79392372e-02 -5.24458110e-01 1.52859241e-01
-1.61032510e+00 6.68036938e-02 1.29502654e-01 2.63486832e-01
-7.25140095e-01 3.14875484e-01 -5.62831283e-01 -3.91105384e-01
9.08840373e-02 3.48549068e-01 -1.03200901e+00 6.38927102e-01
1.15357983e+00 8.54614615e-01 -3.28737259e-01 -8.12528491e-01
-2.07967207e-01 -2.01474205e-01 -1.07226610e-01 -2.97588259e-01
1.67606366e+00 -5.05560219e-01 -6.51496232e-01 5.41485250e-02
1.02261460e+00 4.19625878e-01 -5.00907838e-01 -1.11011453e-01
8.12948197e-02 -8.00269365e-01 2.95883328e-01 -5.75329900e-01
-8.46058488e-01 2.73921698e-01 5.55578947e-01 6.73703074e-01
4.53102767e-01 -3.65815908e-01 2.61015952e-01 9.48718265e-02
-9.78238806e-02 -9.31976736e-01 -9.97882262e-02 3.64830019e-04
7.95121312e-01 -1.08512092e+00 2.22494602e-01 -7.06745446e-01
-3.85275453e-01 1.14340782e+00 1.63363636e-01 3.95922698e-02
1.11138606e+00 -1.76303461e-01 -1.39816508e-01 -3.07200938e-01
-7.09184527e-01 3.34407359e-01 -2.06676386e-02 7.56900370e-01
1.11905038e+00 2.35216022e-02 -1.18194997e+00 3.44002694e-01
4.55773264e-01 7.03329071e-02 7.28376925e-01 1.12186992e+00
-4.62640643e-01 -1.44211388e+00 -3.93668592e-01 6.52402818e-01
-8.81813586e-01 -3.47242504e-02 -9.35842812e-01 7.79844895e-02
2.67044269e-02 1.50355816e+00 -1.26139268e-01 -3.24322551e-01
-2.19659895e-01 1.13925345e-01 -2.01240573e-02 -8.29645336e-01
-1.43074799e+00 -5.85719645e-02 8.07263255e-01 -1.42405242e-01
-9.37636018e-01 -9.43809152e-01 -1.11438608e+00 -7.98994243e-01
-4.31120098e-01 1.82938665e-01 8.86487305e-01 9.87079263e-01
8.80805492e-01 1.35035366e-01 5.52231610e-01 -5.16704202e-01
-2.86187362e-02 -8.79709363e-01 -2.45159790e-01 3.31308097e-01
1.07410908e-01 -4.86269027e-01 -1.01079822e-01 -1.84493810e-01] | [8.650735855102539, 10.320206642150879] |
04c22d5f-0d2f-4462-ba12-89209076bb0d | massively-deep-artificial-neural-networks-for | 1507.05053 | null | http://arxiv.org/abs/1507.05053v1 | http://arxiv.org/pdf/1507.05053v1.pdf | Massively Deep Artificial Neural Networks for Handwritten Digit Recognition | Greedy Restrictive Boltzmann Machines yield an fairly low 0.72% error rate on
the famous MNIST database of handwritten digits. All that was required to
achieve this result was a high number of hidden layers consisting of many
neurons, and a graphics card to greatly speed up the rate of learning. | ["Keiron O'Shea"] | 2015-07-17 | null | null | null | null | ['handwritten-digit-recognition'] | ['computer-vision'] | [-5.37214577e-01 -7.15469643e-02 -3.73194553e-02 -6.36488080e-01
-1.93399087e-01 -3.26648176e-01 3.51876199e-01 -2.69613326e-01
-1.13259494e+00 7.51572013e-01 -2.77366191e-01 -5.27065098e-01
3.73565853e-01 -8.11711609e-01 -4.89133388e-01 -8.10602546e-01
-1.23590626e-01 5.40924489e-01 5.96884072e-01 1.46785766e-01
5.41632771e-01 9.19936121e-01 -1.21076334e+00 2.99258173e-01
2.88724393e-01 1.04867053e+00 3.03895563e-01 9.34341967e-01
-1.97606936e-01 9.24452186e-01 -3.97556603e-01 -4.72950816e-01
3.86082351e-01 -5.92709258e-02 -5.89956164e-01 -2.66289979e-01
5.33906110e-02 -5.41884720e-01 -7.78369665e-01 9.29345548e-01
3.13048393e-01 7.02081680e-01 8.15517724e-01 -6.41867399e-01
-4.60126996e-01 6.59849942e-01 -3.90440166e-01 8.27557325e-01
-4.12709624e-01 3.25243697e-02 7.12061167e-01 -1.01899958e+00
6.47883937e-02 1.05377233e+00 7.17116892e-01 8.07694674e-01
-9.11265552e-01 -7.69668579e-01 -1.44371837e-01 6.52156249e-02
-1.32990313e+00 -6.71970606e-01 6.47199392e-01 -1.85259208e-01
1.76835299e+00 4.60166782e-02 9.53378677e-01 8.38150561e-01
3.14926028e-01 3.43695372e-01 8.13009143e-01 -4.61449504e-01
6.73791707e-01 1.69522777e-01 4.27528650e-01 8.60786796e-01
4.28033084e-01 -1.42410785e-01 -1.58980280e-01 -2.10987285e-01
1.39549208e+00 -4.10489133e-03 1.82729945e-01 -1.23722464e-01
-6.16652071e-01 9.54042077e-01 6.31376505e-01 4.38658446e-01
-3.62629533e-01 4.16095406e-01 3.90002698e-01 -2.28877857e-01
5.48926927e-02 5.92764676e-01 -2.15700254e-01 -1.97063506e-01
-7.29603827e-01 -1.05581649e-01 1.03284013e+00 5.15777230e-01
2.66098499e-01 6.66183829e-01 7.35582530e-01 8.36325407e-01
9.90668312e-02 2.39721522e-01 6.07371986e-01 -1.10854042e+00
6.94538474e-01 1.79883540e-01 2.43605580e-02 -9.60122049e-01
-5.29423118e-01 -2.32745498e-01 -1.06532252e+00 4.51014310e-01
5.62346280e-01 -1.17974197e-02 -1.25449204e+00 1.22063708e+00
-3.16125490e-02 -2.92418659e-01 6.33408278e-02 1.06198454e+00
3.55370820e-01 9.81752574e-01 3.43892246e-01 1.12428144e-01
1.04118681e+00 -7.10079730e-01 -3.57887268e-01 -7.67307818e-01
2.99896389e-01 -4.12237763e-01 8.54380786e-01 5.33904314e-01
-1.21142721e+00 -3.60809803e-01 -1.36239648e+00 -1.58532962e-01
-3.46920013e-01 -1.81125656e-01 1.13103867e+00 6.20476961e-01
-8.14959466e-01 7.13448226e-01 -1.23281169e+00 1.23485975e-01
8.49520385e-01 6.00047529e-01 -4.95749801e-01 -6.45714253e-02
-7.96524048e-01 1.32674408e+00 1.09853387e+00 2.99835831e-01
-4.82192367e-01 -1.05586864e-01 -7.01942563e-01 2.93097764e-01
-1.84119537e-01 -2.21249744e-01 9.73199189e-01 -6.65073991e-01
-1.56556654e+00 7.60644495e-01 4.86856066e-02 -4.22846913e-01
3.38368416e-01 -6.09234795e-02 -1.70539171e-01 1.60474434e-01
-6.71473682e-01 6.28137290e-01 7.21721232e-01 -6.72358334e-01
-5.89778244e-01 -4.41964179e-01 -4.45246249e-01 -5.54512739e-02
-3.29512656e-01 9.71832797e-02 -4.61999029e-01 -4.35717404e-01
6.25302732e-01 -8.29452872e-01 -3.35635453e-01 -5.34828365e-01
-1.10698398e-02 -4.41035420e-01 8.44230950e-02 -8.23664248e-01
5.99562347e-01 -2.26954556e+00 -1.85186155e-02 3.74929160e-01
1.93613797e-01 2.96360821e-01 5.09471707e-02 -3.44688475e-01
-2.69862115e-01 2.87298765e-03 7.22543895e-02 -2.80053145e-03
-1.47707507e-01 4.82422352e-01 -4.23342079e-01 7.75848806e-01
2.76949666e-02 6.43497229e-01 -5.23063958e-01 -3.08492601e-01
2.83521861e-01 4.73884195e-01 -8.40111673e-01 -9.81508046e-02
3.62607926e-01 -3.81294757e-01 -1.33469746e-01 6.20089233e-01
3.62048149e-01 -3.36137414e-02 8.14982727e-02 9.49734449e-02
4.86867391e-02 2.89339334e-01 -1.10193408e+00 1.29973543e+00
1.21608679e-03 6.50508583e-01 7.33888755e-03 -1.00805581e+00
9.22106206e-01 1.75844595e-01 2.64645785e-01 -6.37878716e-01
4.33929324e-01 1.59341484e-01 1.90910533e-01 -2.18447447e-01
3.71953428e-01 -6.73285306e-01 4.96049933e-02 4.15719241e-01
2.12015778e-01 -2.57830083e-01 -8.36262852e-02 -3.51058364e-01
8.98735702e-01 -4.27777916e-01 9.10362229e-02 -3.83150607e-01
-8.93785506e-02 -5.80124706e-02 9.07193542e-01 6.92489684e-01
-2.00406820e-01 4.52173412e-01 3.61006320e-01 -1.07609510e+00
-1.60798275e+00 -9.21243846e-01 -1.90026253e-01 1.20201159e+00
-3.09743851e-01 7.64095634e-02 -6.86473072e-01 -1.23086259e-01
4.18955088e-03 8.11493456e-01 -5.59641957e-01 -1.62859470e-01
-8.41666639e-01 -1.06689751e+00 7.94001639e-01 9.57856536e-01
4.22946572e-01 -1.19599533e+00 -8.30858350e-01 4.30667192e-01
3.04546267e-01 -8.60139191e-01 2.72032827e-01 1.00078154e+00
-1.55806005e+00 -8.84508312e-01 -6.45333707e-01 -9.42194998e-01
7.91798592e-01 -3.35883975e-01 9.32263076e-01 -7.49138370e-02
-5.01802266e-01 -4.31848645e-01 1.47338614e-01 -4.72442687e-01
-1.99830368e-01 1.32880285e-01 5.11729658e-01 -8.09382021e-01
5.59718907e-01 -5.41639328e-01 -1.95583954e-01 -8.18015561e-02
-7.34182656e-01 -5.77376224e-02 6.93556607e-01 9.37569201e-01
2.58669376e-01 3.91223043e-01 3.82506289e-02 -4.92905885e-01
3.40853631e-01 -1.03727035e-01 -6.78182662e-01 -1.60908774e-01
-5.13067186e-01 1.75153106e-01 6.30606532e-01 -8.59121203e-01
-7.74739981e-01 6.88155442e-02 -5.62388599e-01 -1.44835711e-01
-4.36457545e-02 7.23274872e-02 2.34356090e-01 -7.89051354e-02
8.02957416e-01 4.37775284e-01 -4.02134717e-01 -5.52154422e-01
-6.67066053e-02 6.38660073e-01 5.72228670e-01 -2.56350905e-01
2.61574119e-01 -2.00225320e-02 -3.57573897e-01 -7.11597085e-01
-4.69486684e-01 5.69202879e-04 -6.58431709e-01 2.53698826e-01
6.34558082e-01 -6.88816607e-01 -8.74826849e-01 6.23728693e-01
-1.16082048e+00 -6.38955057e-01 -6.51911572e-02 7.35446870e-01
-4.97540861e-01 -1.11451253e-01 -1.26619315e+00 -8.59982252e-01
-4.13648248e-01 -5.98777771e-01 -1.85479000e-01 2.94564664e-01
-1.63235247e-01 -7.57533371e-01 5.43586500e-02 3.18141073e-01
6.15421236e-01 -2.95097888e-01 1.07677579e+00 -5.25510728e-01
-5.09916954e-02 -6.70231998e-01 -2.62620538e-01 6.21523559e-01
-1.03070382e-02 -1.72222424e-02 -1.20577550e+00 -2.94365287e-01
3.87647510e-01 -4.47307378e-01 1.13520682e+00 4.01152760e-01
1.42822909e+00 -4.18602765e-01 -7.99392834e-02 6.56712294e-01
1.40185940e+00 8.81123245e-01 9.32212591e-01 4.75163281e-01
6.42528474e-01 2.40044259e-02 -2.47966081e-01 3.96082759e-01
-2.11692695e-02 1.50536001e-01 1.12517022e-01 3.43536921e-02
3.36985230e-01 1.13139816e-01 -1.41862109e-01 1.12139440e+00
-6.47614241e-01 1.79980010e-01 -1.34632194e+00 4.04355586e-01
-1.51644540e+00 -1.17474127e+00 4.62246507e-01 1.76561916e+00
1.01821935e+00 6.81061566e-01 -1.63118362e-01 2.75445998e-01
5.48079133e-01 -2.28930935e-02 -7.13114858e-01 -8.22337806e-01
1.80001944e-01 2.00190410e-01 7.80998647e-01 5.33340275e-01
-1.07118273e+00 1.04915345e+00 8.86612606e+00 4.78153408e-01
-1.23216593e+00 -1.89374134e-01 6.22798502e-01 -4.04544324e-01
3.03702921e-01 -6.46100342e-01 -8.80620956e-01 5.80499589e-01
1.24627113e+00 4.79456373e-02 6.75626576e-01 1.27387035e+00
-5.23431659e-01 -2.98646152e-01 -6.25947237e-01 1.34901595e+00
-3.83440033e-02 -1.34243619e+00 3.11975703e-02 -3.69857326e-02
3.48362744e-01 4.00944531e-01 3.43815889e-03 4.51783121e-01
5.84635615e-01 -1.29723787e+00 5.37237585e-01 4.86860722e-01
5.13993442e-01 -1.06493449e+00 8.58838499e-01 4.77129072e-01
-5.09508073e-01 -3.22669953e-01 -1.27125335e+00 -3.87967288e-01
-4.57602948e-01 6.65220559e-01 -6.93891406e-01 -4.98837322e-01
8.19552422e-01 -1.32732600e-01 -1.66043893e-01 1.03818631e+00
2.05453247e-01 7.78269231e-01 -8.15151870e-01 -3.77971590e-01
5.09353042e-01 2.59587109e-01 2.88359355e-02 1.55072534e+00
-1.15656950e-01 6.45621896e-01 -2.21888512e-01 5.91830432e-01
-1.54612705e-01 -3.25738102e-01 -3.72168988e-01 -7.61749521e-02
5.03537357e-01 1.04010665e+00 -1.05139565e+00 -4.17631149e-01
-5.13897352e-02 1.03232145e+00 7.80713737e-01 1.93169340e-01
-7.94088781e-01 -7.49389648e-01 5.22607207e-01 -3.47426161e-02
4.76257503e-01 -6.88600183e-01 -5.69533288e-01 -1.05897856e+00
-1.82820231e-01 -4.23273921e-01 1.25724107e-01 -5.68958223e-01
-9.68764603e-01 6.33840561e-01 -4.37420487e-01 -2.13193074e-01
-2.23774716e-01 -9.43146944e-01 -4.30894047e-01 9.76959050e-01
-8.31492305e-01 -4.08169866e-01 3.33359465e-02 6.18321598e-01
2.35067025e-01 -3.17766130e-01 1.21078348e+00 2.82678455e-01
-8.66376162e-01 4.82321799e-01 2.52955556e-01 8.43787968e-01
1.85372040e-01 -1.19134402e+00 6.60152435e-01 6.58815920e-01
2.82717884e-01 7.83263087e-01 7.26159275e-01 -2.82746881e-01
-1.51061475e+00 -5.05831957e-01 7.60634422e-01 -4.27026063e-01
3.14913243e-01 -3.75137627e-01 -1.10173726e+00 8.04338396e-01
-3.43598992e-01 2.66314417e-01 5.99813581e-01 3.22775036e-01
-4.37896311e-01 2.24659555e-02 -1.21684635e+00 4.49465811e-01
6.92201972e-01 -6.07213318e-01 -1.17413175e+00 -1.12243265e-01
4.14665379e-02 -5.22223532e-01 -5.73602796e-01 3.18595618e-02
6.97338462e-01 -7.69069910e-01 1.02609134e+00 -9.39910650e-01
2.66528428e-01 2.72444040e-01 -2.98520923e-01 -9.21167135e-01
-1.22571003e+00 2.35151872e-01 -3.22206527e-01 7.37122834e-01
4.36307341e-01 -6.23989880e-01 1.26459920e+00 1.57226074e+00
1.16348885e-01 -7.41939545e-01 -1.20321536e+00 -9.43914056e-01
3.49041253e-01 -4.70370799e-01 -2.29710136e-02 8.47701252e-01
1.98013678e-01 9.61898714e-02 -3.38174589e-02 -1.51096359e-01
7.71075368e-01 -3.02691966e-01 8.62754285e-02 -1.20860803e+00
-1.51482448e-01 -6.66084349e-01 -6.49130642e-01 -5.29804885e-01
5.47740608e-02 -6.18242681e-01 1.21527381e-01 -1.31603348e+00
4.44762707e-01 -7.77039111e-01 -7.43206918e-01 8.40032637e-01
3.81112657e-02 5.69434047e-01 -2.19507026e-03 3.79871160e-01
-4.63584095e-01 2.32086882e-01 4.78957832e-01 -1.22003399e-01
-3.12429637e-01 -1.97897568e-01 -6.55967593e-01 1.31021857e+00
1.13883150e+00 -8.34743083e-01 5.84705584e-02 -6.30056918e-01
9.58102867e-02 -2.36367986e-01 1.51658833e-01 -1.30659640e+00
3.75472367e-01 -2.25546360e-01 1.38882422e+00 -3.45876873e-01
7.06666231e-01 -7.51868010e-01 3.23061310e-02 9.01216626e-01
-3.31642479e-01 -1.11797176e-01 5.51905572e-01 1.40037209e-01
6.39417954e-03 -4.05593604e-01 1.32260752e+00 -3.33346009e-01
-9.08969998e-01 -2.39264104e-03 -5.92682123e-01 -7.55560100e-02
1.08688867e+00 -2.89876968e-01 -1.27195446e-02 2.26961225e-01
-5.87707877e-01 -4.84382749e-01 3.39340806e-01 1.81533620e-01
6.74689353e-01 -1.26742303e+00 -3.61021131e-01 4.50957566e-01
-5.96753299e-01 -1.29137591e-01 -1.61753185e-02 1.91821992e-01
-1.02756321e+00 7.26413608e-01 -9.02033687e-01 -2.02380955e-01
-1.17834044e+00 4.67062116e-01 6.09140456e-01 6.15994856e-02
-1.21618760e+00 1.38337219e+00 -5.01091182e-01 -1.92567214e-01
4.09511894e-01 -2.90593773e-01 2.84968894e-02 -7.18214810e-02
1.14705288e+00 3.90152246e-01 1.05005287e-01 -4.44607530e-03
-7.95748889e-01 8.27135891e-02 -2.55280584e-01 -4.15864944e-01
1.72867990e+00 5.50437808e-01 4.95841354e-02 5.21751702e-01
8.83040488e-01 -4.91988212e-01 -1.36446297e+00 -1.83261596e-02
1.53071824e-02 -3.04280788e-01 5.45521140e-01 -6.44748807e-01
-9.01733935e-01 8.73955786e-01 5.01019061e-01 1.05692245e-01
6.59543455e-01 -2.69375473e-01 5.93077779e-01 1.23527753e+00
7.88616836e-01 -1.53331685e+00 -4.47901845e-01 9.29526389e-01
5.61236322e-01 -1.27878141e+00 1.82756260e-01 3.60254794e-01
-4.93766040e-01 1.35486853e+00 3.39359611e-01 -7.40446091e-01
7.16155052e-01 7.22676277e-01 -1.32266000e-01 8.88685733e-02
-6.06064618e-01 4.08318520e-01 1.24426194e-01 4.15131062e-01
3.91595185e-01 7.03513324e-02 3.24649215e-01 7.63875067e-01
-2.04275444e-01 6.76352531e-02 8.21968541e-02 1.02061486e+00
-1.02258861e+00 -2.20815822e-01 -4.62934375e-01 4.97637868e-01
-8.13134611e-01 -2.60638654e-01 5.72811514e-02 4.36140120e-01
-3.42369258e-01 5.19000888e-01 5.45253217e-01 -3.26182932e-01
-7.48149082e-02 5.18853188e-01 8.61650825e-01 -3.15654367e-01
-4.37248498e-01 -4.42720532e-01 -2.81602949e-01 -7.23625958e-01
3.35508019e-01 6.11760877e-02 -1.46835577e+00 -9.82799649e-01
-3.06596100e-01 1.08175918e-01 1.17715406e+00 7.78704405e-01
-1.70742750e-01 2.69417405e-01 1.31643862e-01 -1.25618148e+00
-7.35508740e-01 -1.08919144e+00 -1.09643018e+00 -4.01960835e-02
1.27922073e-01 -3.63664508e-01 -5.63153982e-01 -2.43055269e-01] | [8.552042007446289, 3.0549142360687256] |
4dcc7c70-527c-4698-abcb-f85b030ae726 | robust-diversified-graph-contrastive-network | null | null | https://dl.acm.org/doi/abs/10.1145/3503161.3547894 | https://dl.acm.org/doi/pdf/10.1145/3503161.3547894 | Robust Diversified Graph Contrastive Network for Incomplete Multi-view Clustering | Incomplete multi-view clustering is a challenging task which aims to partition the unlabeled incomplete multi-view data into several clusters. The existing incomplete multi-view clustering methods neglect to utilize the diversified correlations inherent in data and handle the noise contained in different views. To address these issues, we propose a Robust Diversified Graph Contrastive Network (RDGC) for incomplete multi-view clustering, which integrates multi-view representation learning and diversified graph contrastive regularization into a unified framework. Multi-view unified and specific encoding network is developed to fuse different views into a unified representation, which can flexibly estimate the importance of views for incomplete multi-view data. Robust diversified graph contrastive regularization is proposed which captures the diversified data correlations to improve the discriminating power of the learned representation and reduce the information loss caused by the view missing problem. Moreover, our method can effectively resist the influence of noise and unreliable views by leveraging the robust contrastive learning loss. Extensive experiments conducted on four multi-view clustering datasets demonstrate the superiority of our method over the state-of-the-art methods. | ['Meiyu Liang', 'Yu Zang', 'Yunfei Long', 'Zhongchao Guan', 'Hai Zhu', 'Junping Du', 'Zhe Xue'] | 2022-10-10 | null | null | null | acm-international-conference-on-multimedia-5 | ['incomplete-multi-view-clustering'] | ['computer-vision'] | [-3.21313053e-01 -3.21323872e-01 -2.78432429e-01 -4.52797174e-01
-8.23612809e-01 -6.11321509e-01 1.12388469e-01 -3.85619909e-01
1.10365748e-01 2.79066920e-01 5.66819429e-01 3.97391707e-01
-4.38107371e-01 -3.45555604e-01 -3.33319068e-01 -1.00215900e+00
4.39307243e-01 3.41784298e-01 -1.16282292e-01 1.80119261e-01
-9.82790738e-02 1.51193574e-01 -1.28587592e+00 5.21578014e-01
1.00752759e+00 5.79627275e-01 3.68364632e-01 1.31932005e-01
6.75844848e-02 6.91844702e-01 -1.73902348e-01 -1.54997051e-01
1.55603498e-01 -3.42752665e-01 -4.19003725e-01 7.16171384e-01
1.92701519e-01 -6.30580559e-02 -3.57907176e-01 1.26262069e+00
6.14944577e-01 1.13288015e-01 5.24180055e-01 -1.36105084e+00
-8.58896792e-01 6.07254148e-01 -1.19542158e+00 2.00239345e-01
8.15283805e-02 -9.66653228e-02 1.12135541e+00 -1.03967869e+00
5.79593420e-01 1.41470098e+00 2.76483715e-01 3.90249819e-01
-1.25062108e+00 -5.26888609e-01 8.50047469e-01 2.28644848e-01
-1.40344512e+00 -1.82862982e-01 1.26102412e+00 -4.00198072e-01
9.08894315e-02 7.61219040e-02 3.49751562e-01 9.87284124e-01
-2.93086946e-01 1.04416049e+00 1.12315559e+00 1.20880976e-01
-4.07032818e-02 2.49478500e-02 1.90736577e-01 8.31218183e-01
5.15793264e-01 -1.21022910e-01 -3.81765157e-01 -3.85844633e-02
4.60415393e-01 5.99409997e-01 -6.26246154e-01 -9.74711716e-01
-1.17408562e+00 7.33393490e-01 5.00475228e-01 2.15334684e-01
-2.86344528e-01 -3.84828806e-01 6.40421867e-01 1.04647085e-01
5.79111040e-01 -2.16450512e-01 -4.87433374e-01 4.62508112e-01
-6.06332242e-01 -5.36505938e-01 3.78940910e-01 1.07803857e+00
7.63513207e-01 2.46966437e-01 5.05676456e-02 1.02537870e+00
4.78702486e-01 4.72766757e-01 1.35254219e-01 -1.14799488e+00
9.27346468e-01 1.18065739e+00 -4.46878970e-01 -1.08283436e+00
-3.36620510e-01 -6.88069642e-01 -1.47084332e+00 -1.13529652e-01
-5.14347069e-02 -1.30234510e-01 -7.47820675e-01 1.80567050e+00
3.96319330e-01 -5.01549393e-02 1.38090491e-01 1.00891173e+00
1.03799701e+00 5.55877328e-01 -3.92050236e-01 -5.45345545e-01
1.12508833e+00 -9.19684350e-01 -7.14594781e-01 -8.65958631e-02
3.15896392e-01 -4.05670196e-01 8.14907968e-01 4.50815678e-01
-9.59486187e-01 -5.33797801e-01 -1.28193855e+00 3.82704660e-02
-6.71951473e-02 2.25097835e-01 4.42370892e-01 4.33952749e-01
-6.89111590e-01 1.54880947e-02 -7.19232500e-01 1.60592467e-01
4.76246417e-01 3.21676195e-01 -6.33553565e-01 -8.89246345e-01
-8.42654407e-01 1.60961002e-01 4.10421014e-01 2.87552685e-01
-7.62545049e-01 -5.05629480e-01 -9.00302887e-01 5.72904386e-02
8.67623329e-01 -7.98555672e-01 2.64898986e-01 -8.02901328e-01
-7.92297244e-01 6.62041247e-01 -1.66682899e-01 3.81807894e-01
2.60909647e-01 3.61045338e-02 -5.43984652e-01 4.67417866e-01
3.99167061e-01 1.55575588e-01 7.35824525e-01 -2.07415223e+00
-2.11240172e-01 -1.05020702e+00 -1.69782057e-01 6.85194910e-01
-1.51777714e-01 -4.02397752e-01 -9.67001021e-01 -5.79310775e-01
8.12022269e-01 -9.14689124e-01 -2.12361023e-01 -2.05953464e-01
-3.52977484e-01 8.18802416e-02 1.10204387e+00 -7.03493118e-01
1.12626708e+00 -2.21363807e+00 7.08029091e-01 1.38777077e-01
4.80677724e-01 1.67117640e-02 -2.10382059e-01 2.77978480e-01
-1.65005669e-01 3.50534059e-02 -3.15868199e-01 -2.96072423e-01
-3.00908417e-01 4.78623331e-01 2.64200181e-01 5.57282507e-01
-3.38644832e-01 5.41933298e-01 -8.37035954e-01 -5.44306874e-01
2.07713604e-01 6.18570328e-01 -4.20739889e-01 2.51889676e-01
1.86464161e-01 7.47702599e-01 -6.69173300e-01 6.18804157e-01
1.13642550e+00 -6.43735170e-01 4.59731638e-01 -6.37592435e-01
3.01565081e-01 -4.18451965e-01 -1.37176025e+00 1.92694926e+00
-3.58938366e-01 -9.50828344e-02 5.80271184e-01 -1.10382307e+00
6.85272634e-01 3.20426971e-01 7.58463323e-01 -3.14775765e-01
-9.80373248e-02 -4.64665107e-02 -1.75662845e-01 -5.56141794e-01
6.37896582e-02 -1.30506501e-01 -7.71046951e-02 5.16924620e-01
5.77293858e-02 2.94516504e-01 -1.06396934e-03 5.98406255e-01
5.41018963e-01 2.41378229e-02 2.12721989e-01 -1.54665023e-01
9.87105072e-01 -5.64252675e-01 1.04404485e+00 1.90022022e-01
-2.02427089e-01 9.70790923e-01 4.74933863e-01 -3.39125723e-01
-7.22725272e-01 -1.02543473e+00 3.91016275e-01 9.04606879e-01
4.03874010e-01 -3.16873133e-01 -3.55601132e-01 -1.09673262e+00
-2.00957686e-01 2.01089546e-01 -5.23451090e-01 -2.91813254e-01
-3.27422261e-01 -1.02855825e+00 -1.86953068e-01 6.76853001e-01
6.62207961e-01 -4.72160727e-01 4.10362422e-01 -3.20176743e-02
-7.66652107e-01 -1.12350273e+00 -7.65456200e-01 6.48250729e-02
-9.32948530e-01 -1.32707059e+00 -6.21854126e-01 -7.75212228e-01
1.06663847e+00 1.13941026e+00 1.10921919e+00 1.29670367e-01
6.90678954e-02 8.36159289e-01 -3.79010409e-01 2.16415659e-01
-1.60686687e-01 -1.36875495e-01 3.93662136e-03 4.08823937e-01
2.45556846e-01 -6.74504042e-01 -6.29750252e-01 4.97407049e-01
-1.17228413e+00 -2.59972811e-02 4.70933735e-01 1.08075511e+00
8.84944201e-01 3.39132667e-01 5.97489238e-01 -1.08860540e+00
2.98159212e-01 -6.25924766e-01 -3.34252268e-01 6.11606956e-01
-7.60368109e-01 8.81103147e-03 7.85575569e-01 -3.98467295e-02
-1.32593000e+00 6.92736655e-02 3.76894563e-01 -9.07594144e-01
1.13870181e-01 6.24571383e-01 -9.01292741e-01 6.30391389e-02
-4.44281399e-02 4.09945458e-01 4.69304174e-02 -4.26674515e-01
7.07428098e-01 3.34515542e-01 2.29471684e-01 -5.19492507e-01
6.32017732e-01 8.30717266e-01 6.90954626e-02 -4.53742087e-01
-1.21241391e+00 -8.08189154e-01 -9.30043221e-01 -2.78783113e-01
9.38712537e-01 -1.61138201e+00 -5.87187111e-01 3.51736575e-01
-7.63501406e-01 5.03040969e-01 7.14312643e-02 5.47633529e-01
-3.19106221e-01 9.72415924e-01 -6.26795053e-01 -6.01194501e-01
-2.16778055e-01 -1.15353107e+00 9.07399178e-01 5.31107113e-02
6.48675144e-01 -1.05854809e+00 -2.02114224e-01 1.02541125e+00
-2.78591394e-01 2.22905979e-01 9.14087415e-01 -4.20171946e-01
-7.05384016e-01 5.18283583e-02 -4.29252386e-01 6.20342553e-01
3.04863781e-01 -1.54445127e-01 -7.58860588e-01 -7.11277604e-01
4.34215695e-01 -3.45669836e-01 1.14306140e+00 4.37082559e-01
1.21859765e+00 -1.81998551e-01 -2.79624015e-01 7.65593767e-01
1.73128438e+00 -2.29608957e-02 2.01447606e-01 -8.44292790e-02
1.46644175e+00 6.94410086e-01 4.39642489e-01 5.04253209e-01
7.98386216e-01 2.34005123e-01 6.75498188e-01 -4.58472455e-03
2.11228207e-01 -1.60003692e-01 9.86311957e-02 1.54562986e+00
-1.50107130e-01 -2.23063886e-01 -4.47652638e-01 4.55397338e-01
-1.90673304e+00 -1.13453567e+00 -2.33930841e-01 2.02344990e+00
3.58936675e-02 -8.83662403e-02 2.14431081e-02 -2.78314110e-03
1.16021812e+00 6.70346022e-01 -7.10547149e-01 2.84121692e-01
-4.06142443e-01 -9.09003496e-01 1.64768264e-01 1.40625402e-01
-1.11563849e+00 3.40952575e-01 4.99439192e+00 6.18336558e-01
-6.49890065e-01 1.95119113e-01 7.50639200e-01 -1.19239688e-01
-7.04988778e-01 -8.14555958e-02 -3.96308452e-01 4.99911308e-01
1.93378642e-01 1.13036737e-01 5.21859050e-01 7.50220597e-01
1.21328957e-01 1.23446450e-01 -7.78768301e-01 1.21737075e+00
5.46484768e-01 -9.09947574e-01 3.24214160e-01 2.03034103e-01
1.04886675e+00 8.81014094e-02 8.86480957e-02 7.01864138e-02
3.79078835e-01 -4.26728308e-01 1.77408665e-01 5.06006181e-01
5.84088922e-01 -1.06573868e+00 5.78704059e-01 4.08647209e-01
-1.63002241e+00 -3.55774432e-01 -4.35990423e-01 3.70488465e-01
1.78469852e-01 7.10069478e-01 -3.44632901e-02 1.38733280e+00
6.21137023e-01 1.25832593e+00 -6.32115126e-01 5.05622506e-01
1.49724679e-02 2.37794131e-01 8.17454159e-02 7.13309705e-01
1.38407201e-01 -7.51503229e-01 5.68338037e-01 4.65427518e-01
7.92506784e-02 3.29804838e-01 6.19564712e-01 6.16819620e-01
-6.38864264e-02 -1.14123851e-01 -6.95115745e-01 2.29693949e-01
3.05503786e-01 1.37704468e+00 -6.80103540e-01 -2.52677500e-01
-8.77208471e-01 9.42458212e-01 5.56514382e-01 6.02190435e-01
-4.99334544e-01 3.05904537e-01 1.48818716e-01 -2.59396285e-01
6.11187756e-01 -1.53995603e-01 -2.84813643e-01 -1.80507791e+00
2.62146711e-01 -9.57057893e-01 9.70010996e-01 -7.94109523e-01
-1.80654562e+00 5.95424712e-01 -5.48312031e-02 -1.63180423e+00
1.85856909e-01 -1.61784559e-01 -4.15722817e-01 5.20604014e-01
-1.37499285e+00 -1.48769128e+00 -3.46488297e-01 8.41274559e-01
5.54107964e-01 -3.27945024e-01 3.50952148e-01 4.09193009e-01
-8.61589432e-01 2.53103644e-01 3.80578339e-01 7.13482723e-02
6.14151895e-01 -1.16592407e+00 -3.39640737e-01 9.25458908e-01
-8.36270079e-02 5.51447988e-01 -3.38339359e-02 -5.91733158e-01
-1.44909132e+00 -1.17473209e+00 2.13001352e-02 -2.14652538e-01
2.66502678e-01 -2.20671535e-01 -1.10666239e+00 6.31015837e-01
1.73154220e-01 4.02239084e-01 1.10399806e+00 1.75738528e-01
-7.68440545e-01 -4.30247426e-01 -9.76166189e-01 3.03190678e-01
9.60371196e-01 -6.07750952e-01 -3.73160243e-01 1.07604258e-01
7.56362557e-01 4.88751344e-02 -9.73244369e-01 6.21899486e-01
1.69502944e-01 -1.34596467e+00 1.07608664e+00 -5.05446553e-01
3.50407332e-01 -4.90709811e-01 -3.85340750e-01 -1.41631007e+00
-6.83065832e-01 -2.33569056e-01 -1.65857762e-01 1.55852747e+00
1.13739066e-01 -5.90490818e-01 6.69151664e-01 4.23571765e-01
-1.44794896e-01 -7.05110490e-01 -8.26983571e-01 -5.76972902e-01
-1.53817058e-01 9.30171236e-02 4.39061821e-01 1.22628522e+00
-2.47080013e-01 7.04337597e-01 -6.55284464e-01 5.63921750e-01
1.17647707e+00 5.77168047e-01 4.13409203e-01 -1.34366727e+00
-1.39929324e-01 6.80234805e-02 -2.13605806e-01 -7.86008000e-01
2.04046637e-01 -9.75862086e-01 -3.33808303e-01 -1.80077624e+00
8.98555696e-01 -1.54390946e-01 -6.25024498e-01 5.80553263e-02
-6.32826030e-01 -6.76180124e-02 3.41360778e-01 4.32953298e-01
-1.10122478e+00 8.51591527e-01 1.62905645e+00 -2.81130135e-01
3.91120650e-02 -3.07898577e-02 -1.06770682e+00 8.27231884e-01
3.13107044e-01 -4.15219188e-01 -9.21598852e-01 -4.93972689e-01
2.21967489e-01 4.96004999e-01 1.65498793e-01 -7.90529191e-01
2.78916568e-01 -6.81254640e-03 5.45179367e-01 -9.81320977e-01
3.48726034e-01 -1.12273991e+00 1.10871345e-01 1.69698551e-01
5.67767303e-03 1.13488160e-01 -2.91006923e-01 1.32392383e+00
-4.79883909e-01 1.94795445e-01 7.63314486e-01 -6.46052957e-01
-6.17639780e-01 7.42153525e-01 1.17720731e-01 4.23706710e-01
9.05524135e-01 -5.40444888e-02 -2.94352949e-01 -4.34178084e-01
-1.06577897e+00 7.77578115e-01 5.89786708e-01 4.49622035e-01
8.07458937e-01 -1.55592299e+00 -6.67691767e-01 2.83551604e-01
2.65353292e-01 1.62602037e-01 1.12276864e+00 8.05433273e-01
3.99131589e-02 -3.65605541e-02 -1.17326126e-01 -8.50381851e-01
-1.33547604e+00 1.19702005e+00 1.41064629e-01 -4.66401428e-01
-5.25904238e-01 5.00660896e-01 6.44964337e-01 -7.09667802e-01
-4.14343253e-02 3.12332988e-01 -4.57474977e-01 3.19695264e-01
2.43855089e-01 5.86352468e-01 -2.15807423e-01 -9.22213376e-01
-4.62103784e-01 8.48353982e-01 -3.11108023e-01 2.34687835e-01
1.57507324e+00 -9.42627370e-01 -3.72826517e-01 6.40526414e-01
1.52240264e+00 -1.10464521e-01 -1.49510241e+00 -3.87185365e-01
-4.60714430e-01 -3.85796845e-01 -1.14572782e-03 -5.77096581e-01
-1.73759794e+00 8.88384223e-01 3.09307247e-01 1.09496728e-01
1.39737642e+00 5.17760962e-02 4.79308248e-01 2.08141506e-01
3.49776000e-01 -1.06180429e+00 3.72060031e-01 1.49831921e-01
7.42700934e-01 -1.57260132e+00 4.95978385e-01 -7.55783021e-01
-1.13025177e+00 7.70223737e-01 8.41447771e-01 -1.86880250e-02
8.23306024e-01 -2.41664931e-01 1.99214533e-01 -5.92686176e-01
-7.06873536e-01 -1.29548997e-01 3.53410929e-01 6.24531031e-01
2.52987836e-02 -2.25421302e-02 -3.29526998e-02 6.68894589e-01
7.19973862e-01 -4.49315280e-01 5.88352382e-01 6.07225657e-01
-3.93398553e-02 -7.74927616e-01 -3.39741230e-01 4.27762955e-01
-4.03465986e-01 1.64620236e-01 -2.95564950e-01 5.02645671e-01
7.32984394e-02 1.11488008e+00 -4.40364689e-01 -4.16850507e-01
1.43462643e-01 -2.46151850e-01 2.56332010e-01 -5.80260098e-01
-3.13906491e-01 5.25512040e-01 -2.18820289e-01 -3.79528403e-01
-1.04476690e+00 -6.18411839e-01 -9.91249323e-01 1.84455037e-01
-4.28620905e-01 1.24262668e-01 1.58179790e-01 8.19678068e-01
5.79886675e-01 6.09567761e-01 1.20489526e+00 -6.35593772e-01
-4.75900143e-01 -4.14356768e-01 -1.07949460e+00 6.90932214e-01
3.17396283e-01 -6.46180868e-01 -6.41069591e-01 -6.99862540e-02] | [8.35886287689209, 4.605549335479736] |
bb8aa14f-7382-46f6-bd4d-3494e9335622 | feature-extraction-matters-more-universal | 2303.00200 | null | https://arxiv.org/abs/2303.00200v1 | https://arxiv.org/pdf/2303.00200v1.pdf | Feature Extraction Matters More: Universal Deepfake Disruption through Attacking Ensemble Feature Extractors | Adversarial example is a rising way of protecting facial privacy security from deepfake modification. To prevent massive facial images from being illegally modified by various deepfake models, it is essential to design a universal deepfake disruptor. However, existing works treat deepfake disruption as an End-to-End process, ignoring the functional difference between feature extraction and image reconstruction, which makes it difficult to generate a cross-model universal disruptor. In this work, we propose a novel Feature-Output ensemble UNiversal Disruptor (FOUND) against deepfake networks, which explores a new opinion that considers attacking feature extractors as the more critical and general task in deepfake disruption. We conduct an effective two-stage disruption process. We first disrupt multi-model feature extractors through multi-feature aggregation and individual-feature maintenance, and then develop a gradient-ensemble algorithm to enhance the disruption effect by simplifying the complex optimization problem of disrupting multiple End-to-End models. Extensive experiments demonstrate that FOUND can significantly boost the disruption effect against ensemble deepfake benchmark models. Besides, our method can fast obtain a cross-attribute, cross-image, and cross-model universal deepfake disruptor with only a few training images, surpassing state-of-the-art universal disruptors in both success rate and efficiency. | ['Chuanxi Chen', 'Yue Xu', 'Shengshan Hu', 'Yunming Zhang', 'Zhenhao Lu', 'Dengpan Ye', 'Long Tang'] | 2023-03-01 | null | null | null | null | ['face-swapping'] | ['computer-vision'] | [-2.66632456e-02 -2.62770597e-02 2.44943887e-01 -3.80942672e-01
-5.64227998e-01 -9.67344940e-01 5.84379852e-01 -7.04578042e-01
-3.29376012e-01 4.48211342e-01 -1.10464767e-01 -2.79064953e-01
-2.66772360e-01 -6.80425406e-01 -9.47215974e-01 -9.38134611e-01
2.05076244e-02 -2.67251998e-01 -1.72309339e-01 -3.56896073e-01
-1.43938854e-01 8.71195912e-01 -1.14029527e+00 1.88234031e-01
9.04068947e-01 1.19731736e+00 -5.49980760e-01 5.85572541e-01
2.02907339e-01 2.46023193e-01 -7.93084443e-01 -1.28927529e+00
5.85509896e-01 -2.99287349e-01 -5.52492738e-01 -1.96245909e-01
6.24761343e-01 -7.42582619e-01 -6.79291189e-01 1.24331689e+00
9.83168542e-01 -3.05232644e-01 5.71779668e-01 -1.72369456e+00
-7.68718243e-01 1.27787009e-01 -6.82633877e-01 -1.17953364e-02
-8.45362470e-02 5.90399444e-01 3.45082164e-01 -8.62710416e-01
8.60239267e-02 1.57986856e+00 7.63321340e-01 8.96878004e-01
-1.13857234e+00 -1.43225968e+00 2.32565582e-01 1.00028947e-01
-1.55786669e+00 -8.64421844e-01 7.94738650e-01 -1.82046607e-01
7.83869505e-01 5.07141054e-01 3.71866584e-01 1.20269692e+00
6.22500241e-01 6.61451817e-01 8.60758364e-01 1.40982136e-01
-2.01816022e-01 -9.82311815e-02 -1.83208570e-01 8.98425221e-01
1.25380024e-01 5.60987651e-01 -2.75258243e-01 -6.85180545e-01
4.65938359e-01 1.01010039e-01 -3.54190469e-01 1.06079495e-02
-3.64901245e-01 7.68599093e-01 2.80931056e-01 -2.31518716e-01
-2.66421407e-01 3.75431687e-01 6.13876104e-01 6.05502725e-01
3.66291076e-01 1.85470209e-01 -5.80553949e-01 7.32849464e-02
-6.52052760e-01 4.26160216e-01 6.93928719e-01 4.92430747e-01
7.21420705e-01 4.18931320e-02 -2.51768023e-01 5.85974157e-01
2.31890693e-01 5.85806191e-01 2.94993520e-01 -6.52387559e-01
1.24553524e-01 3.95288199e-01 -1.63968891e-01 -1.44466031e+00
-1.17470987e-01 -4.18193370e-01 -1.13405192e+00 5.08065224e-01
5.02680540e-02 -6.02120161e-01 -9.53407228e-01 2.02027297e+00
7.53931642e-01 5.08900106e-01 5.80057018e-02 5.24684966e-01
7.87644565e-01 4.35200810e-01 -4.74416092e-02 -2.71101564e-01
1.29244077e+00 -6.99777782e-01 -6.89922333e-01 1.50666445e-01
6.02932334e-01 -7.45986521e-01 8.47631752e-01 5.49571693e-01
-8.50039959e-01 -3.34496409e-01 -1.06660473e+00 -9.64491069e-03
-3.33810836e-01 -3.58709574e-01 6.19531333e-01 1.00174546e+00
-9.48378742e-01 4.70491499e-01 -4.57670331e-01 2.92199314e-01
9.92409468e-01 8.57151628e-01 -8.38520765e-01 -5.94866986e-04
-1.43951845e+00 7.92778313e-01 9.23264250e-02 3.95519793e-01
-1.37974560e+00 -9.99292314e-01 -6.58622146e-01 5.70075437e-02
3.90382588e-01 -6.55631065e-01 9.80015635e-01 -1.03417945e+00
-1.56526911e+00 7.85982072e-01 6.12912141e-02 -2.78998941e-01
8.14233124e-01 -3.29800844e-01 -7.01200366e-01 -1.36133716e-01
-3.79424781e-01 5.43616235e-01 1.34863758e+00 -1.21828747e+00
-4.66387600e-01 -4.63340193e-01 2.67907917e-01 -9.92780849e-02
-7.41079450e-01 6.13361560e-02 -2.35694826e-01 -7.65124977e-01
-5.97550690e-01 -9.14547801e-01 -1.00553900e-01 1.94776982e-01
-8.07226479e-01 -2.71656781e-01 1.33157396e+00 -6.58099174e-01
1.35015404e+00 -2.42044044e+00 5.41169047e-02 3.08254033e-01
6.22680843e-01 8.32485259e-01 -5.12023270e-01 1.76952973e-01
-3.46920609e-01 4.94399011e-01 -2.53045022e-01 -4.02104497e-01
1.58981770e-01 1.53627858e-01 -4.46207345e-01 6.03330195e-01
1.59363657e-01 1.14592564e+00 -5.40151894e-01 -1.97759658e-01
-1.15584180e-01 5.60401857e-01 -7.08797455e-01 8.74129683e-02
1.59010351e-01 3.25913914e-02 -3.26306045e-01 6.43641591e-01
1.52809918e+00 4.72273052e-01 -3.57741982e-01 -3.96367818e-01
3.38707745e-01 -5.02225876e-01 -9.71487105e-01 1.27302587e+00
-2.88128674e-01 3.36831540e-01 1.75557986e-01 -5.24302483e-01
7.18853831e-01 1.30127594e-01 2.82941937e-01 -4.30618137e-01
4.13246512e-01 9.20487419e-02 2.28631034e-01 -5.99088907e-01
8.96377191e-02 -2.29494512e-01 -3.24360877e-01 3.20960611e-01
8.51381645e-02 2.61762738e-01 -4.85065550e-01 1.31636232e-01
1.20279419e+00 -2.86714822e-01 5.53616695e-02 -3.60028535e-01
5.69801390e-01 -7.36002207e-01 9.02624786e-01 5.47172487e-01
-5.11046231e-01 3.41880590e-01 8.64545941e-01 -5.20337760e-01
-4.78150636e-01 -9.62928951e-01 -8.97031836e-03 7.94539273e-01
1.95784554e-01 -5.50472260e-01 -1.15221345e+00 -1.42872655e+00
4.07515317e-01 3.77356976e-01 -9.84189093e-01 -1.06800067e+00
-3.01358521e-01 -8.37943852e-01 1.48239827e+00 9.12134051e-02
9.25808430e-01 -7.42452383e-01 -3.56969573e-02 -2.54050735e-02
3.67968827e-02 -6.51332498e-01 -8.29235911e-01 -1.57771111e-01
-2.24589154e-01 -1.15104258e+00 -3.92126650e-01 -5.36965013e-01
4.71942425e-01 1.87060431e-01 6.89657748e-01 3.14323932e-01
-4.11136031e-01 1.27533093e-01 7.77729750e-02 -6.62996233e-01
-2.56235808e-01 -8.15204307e-02 2.44984955e-01 4.91975427e-01
4.22371954e-01 -8.26261401e-01 -7.58859456e-01 4.03868675e-01
-1.23820484e+00 -4.96554971e-01 5.14375508e-01 7.91956663e-01
3.70759845e-01 2.87883133e-01 7.82879710e-01 -5.79320073e-01
8.73394251e-01 -3.60833585e-01 -3.86788368e-01 3.18138450e-01
-4.04429227e-01 -1.41891658e-01 6.94728673e-01 -6.10726714e-01
-9.18509543e-01 1.06612347e-01 -3.86815161e-01 -7.96013176e-01
-7.24084750e-02 2.10712299e-01 -1.12767553e+00 -5.64974368e-01
6.45247102e-01 4.27160263e-01 3.23445529e-01 -2.78652757e-01
4.50384021e-01 5.77396274e-01 5.30768692e-01 -3.78459841e-01
1.13698852e+00 3.83762896e-01 1.88583031e-01 -4.79794800e-01
-3.29708397e-01 3.04200947e-01 -8.15995038e-02 -1.59741208e-01
5.38469136e-01 -8.30356240e-01 -1.20329499e+00 1.30337536e+00
-1.27277255e+00 6.47920743e-02 1.23092107e-01 -1.34581149e-01
-1.33239135e-01 4.62646872e-01 -3.41305584e-01 -3.95020992e-01
-5.63894093e-01 -1.31836116e+00 8.02037477e-01 2.12323278e-01
1.77238107e-01 -7.11056530e-01 -1.68606304e-02 -1.02174710e-02
3.04611713e-01 6.16329014e-01 6.97801471e-01 -5.09892285e-01
-4.14561987e-01 -4.05985385e-01 -7.38022923e-02 9.10684407e-01
4.00216311e-01 1.43951118e-01 -1.12676477e+00 -4.08703774e-01
9.52498093e-02 -4.06402797e-01 9.06235397e-01 1.60676494e-01
1.57912874e+00 -7.12473392e-01 -3.23909789e-01 1.29085159e+00
1.16561115e+00 1.86165601e-01 1.11189103e+00 9.71412808e-02
1.01062322e+00 4.54154253e-01 2.49991566e-01 4.76653218e-01
2.42877871e-01 4.85231161e-01 8.38404953e-01 -2.75992274e-01
2.16346428e-01 -1.43248811e-01 5.81083894e-01 8.91469419e-02
3.39221537e-01 -2.46429384e-01 -4.63469654e-01 2.89070070e-01
-1.75956309e+00 -9.51116204e-01 2.06346467e-01 1.73065388e+00
8.59599233e-01 -9.44175646e-02 1.63618661e-02 -1.11252502e-01
6.54771864e-01 1.82567507e-01 -9.64960933e-01 -6.45629764e-01
-1.06118277e-01 -8.82767141e-02 4.13880259e-01 3.40374827e-01
-1.39376342e+00 1.21380699e+00 5.53538179e+00 1.46123970e+00
-1.30379963e+00 1.57593980e-01 8.75976980e-01 -4.92714196e-01
-3.99932802e-01 -3.49713057e-01 -8.42885911e-01 7.02559710e-01
6.55943036e-01 -3.57366443e-01 5.08774161e-01 8.10239077e-01
6.90417737e-02 5.42816103e-01 -1.01462841e+00 1.28572690e+00
5.48757464e-02 -1.24630976e+00 3.83206069e-01 6.56780064e-01
4.97516602e-01 -3.17627966e-01 4.17918742e-01 2.59270370e-01
4.42215741e-01 -1.35746205e+00 4.43611026e-01 5.66817939e-01
9.68965411e-01 -1.55316257e+00 5.36175609e-01 2.98196018e-01
-1.00831521e+00 -1.41332954e-01 -3.53219002e-01 3.24538708e-01
-8.46020505e-02 6.12055480e-01 -1.48091391e-01 6.77047968e-01
8.68223667e-01 3.71181279e-01 -4.07787621e-01 5.99161625e-01
-4.50008214e-02 2.48477012e-01 -4.48998421e-01 3.28174591e-01
3.02930504e-01 2.06170410e-01 6.21191740e-01 1.17887092e+00
1.68818682e-01 1.94952548e-01 -5.16819656e-02 2.99955994e-01
-5.86304486e-01 -1.74271762e-01 -8.79498124e-01 1.90430284e-01
5.19768775e-01 1.45220721e+00 -2.93696467e-02 2.30511986e-02
-2.08654534e-02 1.37860858e+00 3.07257682e-01 2.12969676e-01
-1.37331414e+00 -6.61992192e-01 1.62903726e+00 -2.25767970e-01
2.72124022e-01 1.99818984e-01 -4.27856520e-02 -1.10559297e+00
1.94300935e-01 -1.36087406e+00 2.68889070e-01 -1.99030414e-01
-1.40392923e+00 8.02417576e-01 -1.38073921e-01 -1.06234157e+00
9.47502479e-02 -3.99997354e-01 -7.22971022e-01 8.77534628e-01
-1.43951929e+00 -1.41821575e+00 -1.51873320e-01 1.14908504e+00
-8.99148807e-02 -3.78353596e-01 7.85285532e-01 4.47815925e-01
-9.04510975e-01 1.67946005e+00 8.97588953e-02 2.25864813e-01
7.29565322e-01 -6.58752084e-01 4.39549774e-01 1.02280331e+00
-3.47441703e-01 6.86653614e-01 3.81627351e-01 -5.13868809e-01
-1.32733059e+00 -1.40854323e+00 4.03087199e-01 -3.94478559e-01
2.33072937e-01 -5.71904659e-01 -8.03653657e-01 5.31826854e-01
1.46145076e-01 3.71102214e-01 9.62266028e-01 -1.77270398e-01
-5.47412515e-01 -6.03713751e-01 -1.68970633e+00 8.31545115e-01
1.50471771e+00 -6.33783579e-01 -1.66094780e-01 3.40927355e-02
8.38465571e-01 -3.08192670e-01 -6.31840587e-01 6.07441902e-01
6.86434984e-01 -9.39770997e-01 9.31951702e-01 -8.13438118e-01
4.56909299e-01 -2.59837747e-01 -1.16339266e-01 -1.52995908e+00
-5.34456611e-01 -1.29965878e+00 -4.17439580e-01 1.69350708e+00
4.50101495e-03 -9.90418732e-01 6.96194708e-01 6.47035480e-01
-1.03664450e-01 -1.03136003e+00 -1.25357127e+00 -9.19453621e-01
2.67221481e-01 -2.23732844e-01 1.29704070e+00 9.82209563e-01
-5.32524288e-01 -1.97268128e-01 -8.11674595e-01 4.47457254e-01
6.12366498e-01 -3.72500002e-01 1.13662899e+00 -9.13963258e-01
-1.36725172e-01 -4.89226788e-01 -5.79181314e-01 -8.28032196e-01
5.40015996e-01 -7.91307330e-01 -2.15928093e-01 -7.35657215e-01
1.51650056e-01 -2.00401008e-01 -5.52966714e-01 8.99589241e-01
-3.11133385e-01 3.14701289e-01 2.66119182e-01 -1.58751652e-01
-1.54485136e-01 1.01193345e+00 1.27449226e+00 -2.55483985e-01
6.47944212e-02 -2.74178416e-01 -1.42085350e+00 8.07517827e-01
5.79958975e-01 -7.54750133e-01 -6.28081977e-01 -4.42179263e-01
-1.18820496e-01 -4.09491241e-01 6.28245831e-01 -6.94704056e-01
5.24045229e-02 -2.90427923e-01 4.72704351e-01 -2.25841746e-01
1.80525139e-01 -9.57480133e-01 2.37778440e-01 5.50169945e-01
7.83785060e-02 1.63105696e-01 5.23727298e-01 5.73284388e-01
-1.13937937e-01 1.45184368e-01 9.49502110e-01 1.54333979e-01
-5.94381154e-01 1.03192520e+00 8.59332457e-03 -1.78711250e-01
1.32518697e+00 -1.96448699e-01 -3.88548464e-01 -1.07384562e-01
-5.33425748e-01 3.14694583e-01 2.07354769e-01 5.34843087e-01
7.97763944e-01 -1.54410589e+00 -9.52925205e-01 5.60581267e-01
-1.35337964e-01 -1.67558551e-01 7.95440793e-01 3.00163537e-01
-3.16097617e-01 -2.52603948e-01 -2.88884252e-01 -3.36105347e-01
-1.67341232e+00 6.52456820e-01 8.71551156e-01 -1.39421701e-01
-2.87748069e-01 1.36923373e+00 8.55252326e-01 -4.28734481e-01
9.83802229e-02 2.84152955e-01 1.52057394e-01 5.05713746e-03
8.22186530e-01 2.37948790e-01 3.75421681e-02 -6.15797698e-01
-6.36903763e-01 4.27044153e-01 -6.18529499e-01 3.22337896e-01
1.10917890e+00 -6.90265670e-02 -2.59874374e-01 -6.17620111e-01
1.67031598e+00 8.98328125e-02 -1.66031885e+00 1.11187929e-02
-7.20740497e-01 -6.89108193e-01 1.00373164e-01 -9.37982798e-01
-1.51210058e+00 5.71081817e-01 8.86106610e-01 -1.36610307e-02
1.61689615e+00 -3.90620202e-01 1.23172617e+00 3.21957678e-01
-1.61360614e-02 -6.34105563e-01 1.28107816e-01 3.89881641e-01
1.25634098e+00 -9.63703692e-01 -2.07266837e-01 -3.09339494e-01
-4.05041784e-01 7.32616246e-01 9.76772010e-01 -1.37469351e-01
1.25532770e+00 4.23594892e-01 2.24875882e-01 -9.28126499e-02
-6.55590296e-01 4.10630971e-01 2.45598122e-01 6.19515955e-01
-4.38248247e-01 9.26646497e-03 -1.13445260e-01 1.11509943e+00
-2.54262626e-01 2.58521847e-02 1.60441294e-01 8.51957202e-01
1.00239448e-01 -1.35213470e+00 -1.82585895e-01 3.46465319e-01
-8.42074096e-01 -1.25927284e-01 -6.01553679e-01 6.41721964e-01
6.27686799e-01 9.81667638e-01 -3.31605524e-01 -9.86344635e-01
6.09119594e-01 -2.42267027e-01 3.30823898e-01 -8.35822523e-02
-9.48606610e-01 -1.55795947e-01 -1.22694172e-01 -8.42288077e-01
2.18067974e-01 -3.91136318e-01 -7.91057169e-01 -1.08387661e+00
-4.47572201e-01 -4.92493697e-02 3.20987999e-01 1.06786275e+00
9.37742591e-01 3.62593949e-01 1.21588361e+00 -8.13419342e-01
-7.81702399e-01 -6.94732666e-01 -5.53458512e-01 3.21520269e-01
7.61227369e-01 -6.10660017e-01 -3.76966298e-01 -1.31867662e-01] | [5.607086658477783, 7.769683837890625] |
83ae5e87-de6f-4faf-b002-c4d7854d961b | flow-guided-one-shot-talking-face-generation | null | null | http://openaccess.thecvf.com//content/CVPR2021/html/Zhang_Flow-Guided_One-Shot_Talking_Face_Generation_With_a_High-Resolution_Audio-Visual_Dataset_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Zhang_Flow-Guided_One-Shot_Talking_Face_Generation_With_a_High-Resolution_Audio-Visual_Dataset_CVPR_2021_paper.pdf | Flow-Guided One-Shot Talking Face Generation With a High-Resolution Audio-Visual Dataset | One-shot talking face generation should synthesize high visual quality facial videos with reasonable animations of expression and head pose, and just utilize arbitrary driving audio and arbitrary single face image as the source. Current works fail to generate over 256 x 256 resolution realistic-looking videos due to the lack of an appropriate high-resolution audio-visual dataset, and the limitation of the sparse facial landmarks in providing poor expression details. To synthesize high-definition videos, we build a large in-the-wild high-resolution audio-visual dataset and propose a novel flow-guided talking face generation framework. The new dataset is collected from youtube and consists of about 16 hours 720P or 1080P videos. We leverage the facial 3D morphable model (3DMM) to split the framework into two cascaded modules instead of learning a direct mapping from audio to video. In the first module, we propose a novel animation generator to produce the movements of mouth, eyebrow and head pose simultaneously. In the second module, we transform animation into dense flow to provide more expression details and carefully design a novel flow-guided video generator to synthesize videos. Our method is able to produce high-definition videos and outperforms state-of-the-art works in objective and subjective comparisons. | ['Changjie Fan', 'Yu Ding', 'Lincheng Li', 'Zhimeng Zhang'] | 2021-06-19 | null | null | null | cvpr-2021-1 | ['talking-face-generation'] | ['computer-vision'] | [ 1.85858384e-02 1.28025394e-02 2.70712189e-02 -4.67698932e-01
-8.66616845e-01 -1.99689046e-01 5.03934264e-01 -9.96339321e-01
9.13606584e-02 6.64544225e-01 5.37488759e-01 3.83632034e-01
2.97256589e-01 -5.35601556e-01 -6.97189391e-01 -5.75533867e-01
-2.62616724e-02 1.62246358e-03 -4.46368903e-02 -1.90424919e-01
-1.80862218e-01 4.68698770e-01 -1.93666589e+00 3.89547646e-01
5.68958521e-01 1.16595924e+00 -1.47822313e-03 7.59133935e-01
9.62322354e-02 9.34035361e-01 -4.72958922e-01 -4.41748112e-01
3.66297781e-01 -7.92517900e-01 -4.28868264e-01 5.74368536e-01
8.79278183e-01 -9.45548058e-01 -5.98259389e-01 8.43265176e-01
8.21514308e-01 1.03034720e-01 3.90413791e-01 -1.66881466e+00
-6.36661768e-01 1.24886304e-01 -7.92715669e-01 -3.34577262e-02
8.03768992e-01 5.04551172e-01 6.31604075e-01 -1.05367649e+00
9.96703029e-01 1.64963210e+00 4.52335477e-01 1.05237687e+00
-1.08056355e+00 -9.72096801e-01 2.33774036e-02 1.52508020e-01
-1.59416795e+00 -1.11273921e+00 9.15706873e-01 -4.17884141e-01
4.23844844e-01 3.71288881e-02 9.48189139e-01 1.45385838e+00
-9.71142501e-02 4.14385259e-01 8.08856428e-01 -2.97710747e-02
1.09190326e-02 -2.63273805e-01 -8.91395628e-01 9.41853642e-01
-3.89312744e-01 2.43986949e-01 -9.71231222e-01 -6.24670610e-02
1.12224603e+00 -2.24913120e-01 -5.47488213e-01 -1.21877948e-02
-1.00600994e+00 7.98311710e-01 1.96777225e-01 -7.21102506e-02
-3.18666637e-01 4.23046052e-01 2.10067049e-01 1.36442095e-01
3.99622262e-01 3.58391032e-02 7.13976026e-02 -2.93134719e-01
-1.19476354e+00 3.71643603e-01 4.35754836e-01 9.01870847e-01
7.08111167e-01 6.40410364e-01 -2.98022836e-01 8.79055142e-01
4.27112430e-01 6.01972342e-01 3.45864862e-01 -1.68131948e+00
2.53650665e-01 2.59490274e-02 8.32785517e-02 -1.16998231e+00
-1.05554879e-01 1.95604250e-01 -6.98498428e-01 2.03354299e-01
1.22740708e-01 -3.91260594e-01 -7.87087321e-01 1.87828517e+00
5.93058586e-01 6.23700380e-01 -2.45880276e-01 1.18487060e+00
1.05383658e+00 9.23375726e-01 -1.33829892e-01 -5.86450219e-01
1.23929811e+00 -1.13287222e+00 -1.09727907e+00 1.25366643e-01
1.28517538e-01 -7.05789208e-01 1.11664200e+00 3.25189203e-01
-1.42127800e+00 -7.77669847e-01 -6.92683220e-01 -2.56704330e-01
3.12321365e-01 8.68713334e-02 4.62929994e-01 3.60622048e-01
-1.21978116e+00 2.62802839e-01 -5.91861486e-01 5.23306690e-02
4.73093867e-01 1.51801080e-01 -6.84944272e-01 -8.96795392e-02
-1.13158298e+00 4.02737319e-01 -1.95371509e-01 8.44588727e-02
-1.36249948e+00 -9.19881582e-01 -1.18274045e+00 -8.21828097e-02
1.33637756e-01 -8.56571615e-01 1.13441157e+00 -1.35022128e+00
-2.06110191e+00 6.79172397e-01 -3.31314713e-01 2.45493278e-02
6.48849666e-01 -1.07009396e-01 -3.90755236e-01 7.52546906e-01
-5.25448099e-02 1.30249619e+00 1.30740237e+00 -1.29323089e+00
-3.87910664e-01 -1.80063307e-01 -1.25167537e-02 1.45837069e-01
-3.90720069e-01 2.12195188e-01 -6.64788306e-01 -7.82644987e-01
-4.07061338e-01 -7.83334374e-01 1.26256466e-01 6.26507699e-01
-1.45444125e-01 1.93002582e-01 1.10857868e+00 -7.10290968e-01
1.04567969e+00 -2.24030042e+00 2.10611060e-01 -2.69970834e-01
2.68402070e-01 1.11802086e-01 -3.98371011e-01 6.38601035e-02
-5.08642569e-02 -2.36534625e-02 5.51496595e-02 -7.01160610e-01
-1.45959735e-01 9.93324518e-02 -3.62971783e-01 5.05823255e-01
2.66285896e-01 7.24992275e-01 -1.00064969e+00 -8.87910962e-01
2.57905066e-01 1.27131772e+00 -9.19872880e-01 4.49003577e-01
-6.61904663e-02 6.82669103e-01 -2.55530536e-01 7.87178874e-01
8.00125301e-01 5.31180352e-02 -4.43999507e-02 -3.57395232e-01
9.44951922e-03 -1.87205136e-01 -1.06138825e+00 1.94047785e+00
-5.40576100e-01 7.84276009e-01 4.77500856e-01 -4.18589532e-01
9.21588182e-01 6.42759442e-01 6.71929777e-01 -5.83130240e-01
2.50782430e-01 2.31594890e-02 -3.36855829e-01 -8.03741872e-01
2.98540443e-01 -3.92760962e-01 1.85578406e-01 2.67741561e-01
4.25032079e-01 -3.54636610e-01 8.37705731e-02 1.32300388e-02
8.35899174e-01 2.17109472e-01 -1.78930461e-01 2.28934884e-01
7.15972543e-01 -7.24919200e-01 6.79619551e-01 -1.23225629e-01
-2.22221956e-01 1.06793892e+00 5.95398188e-01 -2.99694240e-01
-9.32086527e-01 -9.97827649e-01 3.07395309e-02 1.07549834e+00
-5.17735630e-02 -7.00406432e-01 -9.32676494e-01 -2.82350838e-01
-3.18931520e-01 1.79018274e-01 -7.22205639e-01 -4.70147729e-02
-5.61050534e-01 -1.58457518e-01 6.58039570e-01 3.24626595e-01
7.49186635e-01 -1.10339475e+00 -3.63297343e-01 -7.31896982e-03
-4.91975039e-01 -1.44020081e+00 -8.72063339e-01 -8.34604144e-01
-4.29188490e-01 -9.12237823e-01 -9.74098563e-01 -6.89968705e-01
6.75325453e-01 4.86287959e-02 8.13157916e-01 -1.08349442e-01
-4.01200801e-01 2.14628100e-01 -2.13192657e-01 1.63628068e-02
-3.87789100e-01 -4.94980127e-01 3.33382010e-01 5.47368884e-01
-2.92004704e-01 -8.33299518e-01 -8.42833817e-01 3.86940271e-01
-8.29436719e-01 1.29776701e-01 2.27404520e-01 7.98952222e-01
4.94399518e-01 -2.10720256e-01 5.23987710e-01 -2.30071485e-01
4.05242383e-01 -2.73870409e-01 -4.16378945e-01 -1.59726530e-01
3.36914286e-02 -3.31953466e-01 8.97869706e-01 -5.86490214e-01
-1.14475071e+00 2.90424258e-01 -4.82597470e-01 -1.07951856e+00
1.67342601e-03 -1.99481696e-01 -4.01731849e-01 -2.04199761e-01
3.58535677e-01 1.77774906e-01 2.94113636e-01 -2.22795710e-01
4.46795940e-01 6.41022265e-01 8.00226808e-01 -4.99191731e-01
8.07133079e-01 6.69873238e-01 -2.85961479e-02 -9.97519612e-01
-5.38536549e-01 9.73275676e-02 -4.43035960e-01 -7.33486176e-01
9.21000004e-01 -1.21406019e+00 -9.56777334e-01 5.01216114e-01
-1.08101106e+00 -3.67777765e-01 -1.68543145e-01 3.60008597e-01
-9.53440905e-01 2.87488550e-01 -7.30616152e-01 -5.73967636e-01
-2.90451646e-01 -1.35069919e+00 1.53151333e+00 1.76937729e-01
-3.72084454e-02 -5.91452122e-01 1.10944659e-02 4.80532646e-01
4.15409356e-01 5.73165834e-01 2.22661629e-01 4.53808844e-01
-4.75927830e-01 2.02403605e-01 -1.26623124e-01 3.96821588e-01
3.02498639e-01 4.81747985e-01 -1.09486532e+00 -3.04111421e-01
-1.65712088e-01 -6.06732368e-01 4.96725142e-01 4.47839916e-01
1.24232364e+00 -6.27948761e-01 5.24659455e-02 1.04483747e+00
1.02230620e+00 -1.03017725e-02 7.51676917e-01 -3.64645988e-01
8.29652846e-01 7.87631989e-01 5.40187061e-01 6.61997676e-01
3.17433268e-01 1.10788965e+00 2.90505767e-01 -1.38836026e-01
-6.14013016e-01 -6.69864535e-01 8.21278214e-01 7.27248907e-01
-2.37214237e-01 7.14842826e-02 -4.44223911e-01 3.83817792e-01
-1.42958796e+00 -1.20942783e+00 4.20223534e-01 1.91396642e+00
1.01525378e+00 -3.28250557e-01 4.14537996e-01 6.16205260e-02
6.62045538e-01 3.02779973e-01 -3.74555618e-01 -1.72490239e-01
-2.73727737e-02 2.05822706e-01 -3.31664443e-01 8.52638841e-01
-7.12228119e-01 9.15720284e-01 6.12885618e+00 8.53399694e-01
-1.72191501e+00 1.10709935e-01 6.42922342e-01 -6.78543329e-01
-3.58274072e-01 -3.96184593e-01 -6.86362028e-01 6.52372181e-01
1.03784299e+00 -2.48230755e-01 2.88780570e-01 8.00454378e-01
6.83742642e-01 3.08762252e-01 -8.98393989e-01 1.39679217e+00
3.87345016e-01 -1.50433767e+00 3.47478807e-01 9.43677127e-02
6.62336826e-01 -4.89867717e-01 1.81261316e-01 -1.83506552e-02
-1.51790008e-01 -1.31059408e+00 8.29913855e-01 5.31885147e-01
1.43088913e+00 -8.62000108e-01 1.06416687e-01 -1.35488778e-01
-1.33288574e+00 -3.91503498e-02 -1.53254017e-01 1.77380443e-01
5.60664535e-01 1.40277952e-01 -3.95196348e-01 2.14654535e-01
8.19106698e-01 8.41120720e-01 -2.15504304e-01 6.43638492e-01
-2.30682328e-01 3.43971193e-01 -2.34487951e-01 4.73471552e-01
5.41414134e-02 9.56871510e-02 5.47755241e-01 1.07854080e+00
5.23114085e-01 2.16326207e-01 2.89814081e-02 8.32009196e-01
-3.09018224e-01 7.84372538e-02 -7.71607876e-01 1.52054414e-01
4.56124097e-01 1.43543124e+00 -2.92636395e-01 -1.70947611e-01
-3.61974657e-01 9.44933057e-01 -1.13418758e-01 3.22264552e-01
-1.13218784e+00 -2.60565281e-01 1.08965874e+00 5.07185340e-01
2.93416977e-01 -1.08714759e-01 4.70767140e-01 -1.27600741e+00
-6.49935156e-02 -1.08584130e+00 -5.68328202e-02 -1.05620980e+00
-9.29671943e-01 9.49985564e-01 -1.04290113e-01 -1.33028460e+00
-6.06826544e-01 -3.10904086e-01 -4.96970057e-01 5.09178221e-01
-1.54398787e+00 -1.19997513e+00 -8.06231081e-01 1.15581381e+00
7.23852098e-01 -7.34236538e-02 7.05809653e-01 6.48597956e-01
-4.78358656e-01 9.38564241e-01 -5.25745749e-01 2.23920673e-01
9.79698539e-01 -5.33185065e-01 6.77242130e-02 5.93944311e-01
3.68352234e-02 3.08248758e-01 7.33827889e-01 -2.38847643e-01
-1.48515391e+00 -1.20984280e+00 5.14120400e-01 -1.58854619e-01
3.35650146e-01 -5.79534173e-01 -6.33166075e-01 4.01743233e-01
2.04609513e-01 5.76037109e-01 6.20620608e-01 -6.94370329e-01
-3.08183074e-01 -4.17426407e-01 -1.23735189e+00 6.34579718e-01
1.17755842e+00 -5.92686296e-01 -1.57022685e-01 6.94146752e-02
6.99630916e-01 -4.13574338e-01 -9.56904650e-01 3.48991334e-01
7.96110511e-01 -1.12482131e+00 8.77697945e-01 -2.92098522e-01
8.63909125e-01 -3.86584520e-01 -7.83828422e-02 -1.16034389e+00
2.84752641e-02 -1.35950708e+00 -1.79564118e-01 1.43919027e+00
8.74034166e-02 -1.92897871e-01 7.76315749e-01 2.63651967e-01
-5.32450862e-02 -7.65138626e-01 -9.11011577e-01 -4.23936605e-01
-2.75674850e-01 -2.65021741e-01 6.93252802e-01 7.94634581e-01
-8.38261694e-02 3.97162199e-01 -8.91795635e-01 -8.75250325e-02
5.38897991e-01 7.58824348e-02 1.11215687e+00 -8.00482750e-01
-2.22721398e-01 -9.86028463e-02 -5.24694204e-01 -1.14536011e+00
4.72804397e-01 -3.72635752e-01 -1.52738765e-02 -1.11149681e+00
2.71192007e-03 -6.17243126e-02 4.55624163e-01 3.38729620e-01
1.23248622e-01 7.47568011e-01 3.13606590e-01 3.55215110e-02
-3.53627026e-01 1.04796147e+00 1.71255314e+00 1.22551046e-01
-7.14434683e-02 -3.83728892e-01 -5.77400565e-01 8.09874058e-01
2.35596269e-01 -1.52337700e-01 -6.74413919e-01 -4.66530442e-01
-3.03074419e-02 6.84893250e-01 4.34704751e-01 -9.49271262e-01
-4.40093577e-02 -2.12521255e-01 5.15835881e-01 -1.30839020e-01
9.82345581e-01 -6.01948798e-01 2.68531024e-01 1.39338166e-01
-3.14180464e-01 5.20517305e-02 7.30905533e-02 3.46172184e-01
-4.92620528e-01 4.18246508e-01 1.03793859e+00 -6.90630525e-02
-5.46246350e-01 8.28544080e-01 -1.70090690e-01 2.19316512e-01
1.10699940e+00 -2.58554816e-01 -1.63493291e-01 -9.72932935e-01
-7.29314804e-01 8.60518068e-02 6.40295923e-01 7.00030088e-01
9.08794224e-01 -1.62486386e+00 -9.36218619e-01 4.94172990e-01
-1.80128351e-01 5.17835841e-04 5.33750951e-01 6.34131074e-01
-7.36730456e-01 6.11749552e-02 -7.03729689e-01 -6.04871869e-01
-1.36271644e+00 2.91340172e-01 4.18673933e-01 2.91241169e-01
-7.10783958e-01 8.88132453e-01 4.50762928e-01 -1.06819887e-02
2.21033841e-01 8.61319005e-02 -1.58348218e-01 9.05873775e-02
9.58857894e-01 1.94651082e-01 -3.01266253e-01 -1.16687477e+00
-1.89391434e-01 8.52693498e-01 3.39319885e-01 -2.78431177e-01
1.25066745e+00 -2.29364276e-01 1.44592822e-01 7.33186826e-02
1.48043156e+00 2.00725883e-01 -1.82691598e+00 2.76144207e-01
-9.27100956e-01 -1.02065790e+00 -8.73361304e-02 -1.52934372e-01
-1.68011880e+00 1.05849624e+00 3.50956082e-01 -3.85783494e-01
1.37727427e+00 -1.56498641e-01 1.09832215e+00 -1.94716439e-01
3.33788782e-01 -8.91376972e-01 6.72980547e-01 1.03794470e-01
1.21434832e+00 -9.44110751e-01 -1.37814879e-01 -5.63307405e-01
-7.83433855e-01 1.16770959e+00 8.86279106e-01 3.08235064e-02
6.19471312e-01 5.10425031e-01 2.08588779e-01 -2.78062709e-02
-1.00165963e+00 1.36245668e-01 1.55891061e-01 8.23014796e-01
4.11955237e-01 -3.91911000e-01 6.98317401e-03 3.63079995e-01
-6.46095872e-01 3.56975913e-01 5.91892004e-01 4.44600493e-01
-1.13931715e-01 -7.51369596e-01 -2.42690399e-01 -5.40589206e-02
-5.20923972e-01 1.53177515e-01 -1.66948915e-01 6.24745607e-01
1.31070793e-01 9.42660987e-01 1.93438277e-01 -4.86419857e-01
1.62248418e-01 -1.28350317e-01 6.47161901e-01 -3.25648844e-01
-1.85777932e-01 2.65934616e-01 1.39208380e-02 -1.07388234e+00
-5.21440268e-01 -4.73906219e-01 -1.14098811e+00 -6.31471872e-01
1.46484554e-01 -3.97629179e-02 3.41021359e-01 5.40640950e-01
6.27035260e-01 3.78648877e-01 9.41218138e-01 -1.41296041e+00
-1.02951571e-01 -8.26919675e-01 -5.29346347e-01 4.97979552e-01
6.79571986e-01 -8.72950613e-01 -5.21814525e-01 4.36731577e-01] | [13.152156829833984, -0.3891148865222931] |
d40bb0f3-9859-4653-b68f-3ad761bcfe54 | slow-learning-and-fast-inference-efficient | null | null | http://proceedings.neurips.cc/paper/2021/hash/75fc093c0ee742f6dddaa13fff98f104-Abstract.html | http://proceedings.neurips.cc/paper/2021/file/75fc093c0ee742f6dddaa13fff98f104-Paper.pdf | Slow Learning and Fast Inference: Efficient Graph Similarity Computation via Knowledge Distillation | Graph Similarity Computation (GSC) is essential to wide-ranging graph applications such as retrieval, plagiarism/anomaly detection, etc. The exact computation of graph similarity, e.g., Graph Edit Distance (GED), is an NP-hard problem that cannot be exactly solved within an adequate time given large graphs. Thanks to the strong representation power of graph neural network (GNN), a variety of GNN-based inexact methods emerged. To capture the subtle difference across graphs, the key success is designing the dense interaction with features fusion at the early stage, which, however, is a trade-off between speed and accuracy. For slow learning of graph similarity, this paper proposes a novel early-fusion approach by designing a co-attention-based feature fusion network on multilevel GNN features. To further improve the speed without much accuracy drop, we introduce an efficient GSC solution by distilling the knowledge from the slow early-fusion model to the student one for fast inference. Such a student model also enables the offline collection of individual graph embeddings, speeding up the inference time in orders. To address the instability through knowledge transfer, we decompose the dynamic joint embedding into the static pseudo individual ones for precise teacher-student alignment. The experimental analysis on the real-world datasets demonstrates the superiority of our approach over the state-of-the-art methods on both accuracy and efficiency. Particularly, we speed up the prior art by more than 10x on the benchmark AIDS data. | ['Yun Fu', 'Yulun Zhang', 'Huan Wang', 'Lichen Wang', 'Handong Zhao', 'Can Qin'] | 2021-12-01 | null | https://openreview.net/forum?id=Q4SdMvWMxb | https://openreview.net/pdf?id=Q4SdMvWMxb | neurips-2021-12 | ['graph-similarity'] | ['graphs'] | [ 6.27442747e-02 -3.73196527e-02 2.76933070e-02 -2.16121599e-01
-5.33953607e-01 -3.58782172e-01 5.16178310e-01 6.13287389e-01
-3.89585942e-01 2.76104420e-01 -1.63501471e-01 -5.14719784e-01
-4.82616037e-01 -9.87534165e-01 -8.17386150e-01 -6.11736059e-01
-7.08139613e-02 3.17992717e-01 3.31566423e-01 -4.02741343e-01
3.40361983e-01 3.18049520e-01 -1.36818802e+00 -3.17230672e-01
1.47287190e+00 7.78119862e-01 2.06650570e-01 4.35119331e-01
-4.76010770e-01 4.70775902e-01 -3.63166153e-01 -8.43308628e-01
1.01805352e-01 -3.65129232e-01 -5.20921946e-01 -2.12354213e-01
5.41882455e-01 -2.26295039e-01 -5.00407398e-01 1.53440499e+00
5.42244554e-01 2.40666240e-01 4.53363806e-01 -1.52393496e+00
-1.06180632e+00 7.13567734e-01 -6.62712812e-01 9.14097726e-02
4.28326100e-01 1.35456964e-01 9.86958027e-01 -7.64985263e-01
2.92227030e-01 1.28851283e+00 8.14477146e-01 2.41833732e-01
-9.55758750e-01 -7.67391086e-01 2.92086005e-01 6.92953289e-01
-1.58039141e+00 1.82089675e-02 9.76487994e-01 -2.09467784e-01
6.76436245e-01 2.17630580e-01 8.34448636e-01 8.75132740e-01
-1.56676248e-01 5.09316921e-01 8.27933073e-01 -3.93208742e-01
2.08544731e-03 -2.73960829e-02 3.10370088e-01 1.17044055e+00
5.13301075e-01 -1.36037454e-01 -2.32977390e-01 -1.75801948e-01
5.07823288e-01 2.52433538e-01 -5.25699615e-01 -3.08719784e-01
-8.59070241e-01 9.35732305e-01 7.53876209e-01 2.88260579e-01
1.58848893e-02 1.66377604e-01 4.61957425e-01 6.61246538e-01
3.42274308e-01 3.25514674e-01 -1.23871639e-01 -4.86719608e-02
-5.95259726e-01 -1.93674956e-02 7.52407491e-01 9.18732464e-01
9.53201115e-01 -2.65637264e-02 1.20641731e-01 6.83117628e-01
1.89371303e-01 3.37736219e-01 5.87852478e-01 -2.63093352e-01
7.54208744e-01 1.13248158e+00 -5.15480518e-01 -1.69318950e+00
-5.70892543e-02 -5.21208882e-01 -1.09975255e+00 -1.04512908e-01
4.06727493e-01 2.36762896e-01 -7.44092941e-01 1.82238829e+00
8.09958518e-01 6.87558711e-01 -2.46264607e-01 6.83680892e-01
7.81364262e-01 5.77335775e-01 -1.99709594e-01 -1.71656832e-01
1.25580394e+00 -9.75691259e-01 -6.27298951e-01 -1.60638064e-01
8.78302753e-01 -6.03608251e-01 1.07654428e+00 3.06616068e-01
-5.49474239e-01 -6.40335798e-01 -1.14526021e+00 -3.32499892e-01
-6.28064573e-01 -3.13566864e-01 6.84613228e-01 5.44891536e-01
-9.03998554e-01 1.01538622e+00 -6.05236948e-01 -3.22440892e-01
2.18245789e-01 3.75249594e-01 -7.08304226e-01 -3.36932719e-01
-1.30958045e+00 7.65154600e-01 6.51440322e-01 3.52507383e-01
-3.80376667e-01 -7.65384257e-01 -8.86847019e-01 3.03531557e-01
6.96869373e-01 -5.66130996e-01 5.03792763e-01 -7.49471188e-01
-1.47445524e+00 4.47882295e-01 3.40561330e-01 -3.14804167e-01
3.51021171e-01 -1.05113886e-01 -3.46011400e-01 3.99717093e-02
-2.08903119e-01 1.36830047e-01 9.42283034e-01 -7.83464432e-01
-3.19919199e-01 -5.46182632e-01 1.69030815e-01 3.62303823e-01
-8.04572761e-01 -2.51823783e-01 -5.49547076e-01 -5.59203684e-01
6.74418733e-02 -9.17568624e-01 -1.11451879e-01 7.82423913e-02
-1.52796999e-01 -6.78436279e-01 8.04654360e-01 -7.19071805e-01
1.65202653e+00 -2.04391050e+00 3.31516922e-01 3.57919484e-01
6.60941839e-01 6.08801901e-01 -2.36418650e-01 5.31777442e-01
-2.41606638e-01 -5.24161989e-03 -2.21317992e-01 -2.62300164e-01
6.07543439e-02 2.02485099e-01 -1.16135851e-01 6.15666747e-01
2.20648989e-01 1.01883030e+00 -1.11549079e+00 -6.62056983e-01
7.13110194e-02 3.72736931e-01 -5.58833659e-01 3.60652059e-01
1.55508816e-01 -3.06269154e-02 -4.13012862e-01 3.78164500e-01
8.28259647e-01 -4.33517367e-01 2.05967411e-01 -2.71175563e-01
2.23822117e-01 -4.94831055e-02 -1.36682451e+00 1.93179440e+00
-3.46789271e-01 1.09457582e-01 1.26886955e-02 -1.45269573e+00
9.98969615e-01 -1.43282088e-02 1.13284238e-01 -5.02409458e-01
1.49229974e-01 3.48528802e-01 1.36899531e-01 -3.08964074e-01
3.61528426e-01 2.95884997e-01 -4.87092845e-02 4.41415697e-01
2.72366226e-01 2.87383422e-02 4.71294932e-02 4.65769380e-01
1.17778552e+00 -2.64064223e-01 4.30795521e-01 -2.73340970e-01
8.51806045e-01 -3.65510851e-01 5.02192557e-01 3.79520386e-01
-1.73272893e-01 2.43201539e-01 4.58941102e-01 -3.77519399e-01
-5.60768962e-01 -7.70453632e-01 3.55288893e-01 8.45485866e-01
2.64735371e-01 -7.75833726e-01 -8.82752061e-01 -1.00943661e+00
2.22771540e-01 2.22425759e-01 -5.16823590e-01 -6.85779393e-01
-5.67831218e-01 -3.82053226e-01 4.04483080e-01 3.54742318e-01
4.07385916e-01 -6.73556507e-01 9.63081196e-02 2.00476363e-01
9.96835455e-02 -1.05296552e+00 -8.11090887e-01 -2.03575015e-01
-8.09671283e-01 -1.16387522e+00 -3.61356676e-01 -8.43313217e-01
8.84912968e-01 5.20042717e-01 7.88711727e-01 6.93160057e-01
-2.30902672e-01 3.00551325e-01 -4.15103048e-01 7.94293061e-02
-1.93967044e-01 2.08078638e-01 1.22807704e-01 1.08514883e-01
2.68297851e-01 -1.02742422e+00 -5.19538403e-01 -7.72143677e-02
-8.96621108e-01 1.35666132e-01 5.77094734e-01 9.99839962e-01
3.82992685e-01 2.38195673e-01 5.77378750e-01 -9.90112901e-01
7.91550875e-01 -4.80468363e-01 -8.14850152e-01 5.04566371e-01
-1.04668176e+00 2.99461514e-01 1.11672127e+00 -5.79130769e-01
-6.73046291e-01 -1.82702541e-01 2.62695011e-02 -7.39697158e-01
3.63560736e-01 7.07234085e-01 -2.00272903e-01 -5.25482655e-01
4.38925892e-01 4.08862144e-01 2.29417846e-01 -4.71802056e-01
5.58054149e-01 5.27881682e-01 5.66005409e-01 -6.40318632e-01
1.26472580e+00 -9.09589306e-02 1.89571664e-01 -6.52465999e-01
-6.34437978e-01 -4.43755001e-01 -3.71622145e-01 -1.05342537e-01
5.20052552e-01 -8.50068927e-01 -8.55831444e-01 5.32575727e-01
-1.20446146e+00 4.90397401e-02 1.54724002e-01 3.93956780e-01
-4.94934954e-02 8.82623076e-01 -6.43957436e-01 -4.51144069e-01
-6.49474204e-01 -1.08321726e+00 7.04364896e-01 3.81363958e-01
3.22852045e-01 -1.16217709e+00 1.34452999e-01 3.24152917e-01
3.55899453e-01 2.37177745e-01 1.08654249e+00 -9.81624603e-01
-7.37740576e-01 -2.77810603e-01 -4.77009416e-01 2.38850951e-01
8.13166350e-02 -2.03789733e-02 -6.15518808e-01 -6.51285291e-01
-1.54025257e-01 -2.39305824e-01 6.77278697e-01 -4.33510393e-01
1.40291715e+00 -3.76210779e-01 -2.10874960e-01 7.46581256e-01
1.59859264e+00 -1.73464105e-01 3.08000863e-01 -6.61945567e-02
1.37411356e+00 2.47087121e-01 4.80609745e-01 1.45894244e-01
6.33438289e-01 4.45415437e-01 3.92540157e-01 1.85588837e-01
-1.82877943e-01 -6.00233257e-01 3.68035227e-01 1.66472340e+00
1.16789518e-02 -2.11544335e-01 -6.72377646e-01 4.75326627e-01
-1.92130637e+00 -6.97741568e-01 -1.77421626e-02 2.27031684e+00
8.14399004e-01 8.37962851e-02 -2.58946538e-01 2.19269350e-01
9.36006129e-01 1.26050115e-01 -3.93616796e-01 -3.24778736e-01
4.60020453e-01 3.64685625e-01 2.57227153e-01 6.26283646e-01
-8.37806106e-01 8.65207076e-01 4.38869238e+00 1.37685144e+00
-1.09043598e+00 1.33770287e-01 1.96000054e-01 4.30393606e-01
-4.38527316e-01 1.97282836e-01 -7.19032764e-01 6.76414847e-01
9.00664687e-01 -4.36207116e-01 7.74912357e-01 8.87535512e-01
-4.11114305e-01 2.65284240e-01 -1.05819464e+00 1.24471545e+00
2.24794537e-01 -1.20092452e+00 1.76969409e-01 5.27431481e-02
6.33596003e-01 -2.84198672e-01 -2.36594558e-01 6.69166863e-01
2.65524119e-01 -7.49797285e-01 1.21841379e-01 3.69307518e-01
7.50665367e-01 -8.90193939e-01 6.80783927e-01 3.71883482e-01
-1.52308619e+00 3.98622863e-02 -5.36577046e-01 8.45072493e-02
-2.11449817e-01 7.40761936e-01 -6.90905511e-01 1.11553574e+00
4.00707811e-01 7.71043658e-01 -1.03726470e+00 8.21199059e-01
-2.68013865e-01 3.57438087e-01 -4.22442883e-01 -2.02195391e-01
2.76129097e-01 -6.93797588e-01 4.40090269e-01 8.49398851e-01
4.12214428e-01 1.84750140e-01 4.09354657e-01 7.09546626e-01
-2.68435001e-01 2.22143173e-01 -8.23162258e-01 -2.81078696e-01
7.25554705e-01 1.55819631e+00 -5.47193050e-01 -2.65357971e-01
-4.79391009e-01 1.17314565e+00 1.04844964e+00 2.92806942e-02
-1.02780747e+00 -9.28743482e-01 3.10977697e-01 -7.89122358e-02
2.73472518e-01 -2.36981422e-01 2.10739896e-01 -1.27452040e+00
2.06330940e-01 -9.58053470e-01 5.48141360e-01 -2.85244554e-01
-1.45032275e+00 4.55064684e-01 -1.18538894e-01 -1.07277179e+00
-1.13246389e-01 -5.45796812e-01 -9.62026417e-01 7.19094932e-01
-1.45447004e+00 -1.12341285e+00 -5.60101867e-01 8.34794223e-01
-3.76011170e-02 6.15001284e-03 6.52101636e-01 6.23959363e-01
-8.25987637e-01 1.10994375e+00 5.63950352e-02 1.43809706e-01
6.83568060e-01 -1.31718266e+00 5.92861474e-01 1.02089703e+00
3.05206776e-01 7.15627968e-01 3.72990221e-01 -5.87221026e-01
-1.92840695e+00 -1.08948278e+00 7.77589619e-01 -1.05770148e-01
9.13563967e-01 -4.01014268e-01 -1.40634036e+00 4.38828707e-01
-3.29460390e-02 3.74084115e-01 3.64811212e-01 1.97891980e-01
-5.69706142e-01 -4.86982137e-01 -9.29486632e-01 6.22785389e-01
1.29454446e+00 -8.19102824e-01 -5.91682017e-01 2.40749985e-01
1.09060609e+00 -3.73458445e-01 -1.07637167e+00 2.45548978e-01
2.36295253e-01 -7.81402290e-01 9.07814622e-01 -5.49872994e-01
2.03215957e-01 -4.29335535e-01 1.21993527e-01 -1.37130785e+00
-3.12455982e-01 -7.52366006e-01 -4.50908870e-01 1.48197711e+00
-9.65057388e-02 -8.90890241e-01 7.36094415e-01 3.71951431e-01
-1.05079047e-01 -9.17839468e-01 -8.43329489e-01 -7.04122663e-01
-1.76083103e-01 -1.23394214e-01 8.49786580e-01 1.36924088e+00
3.82010941e-03 4.97525781e-01 -1.33783281e-01 2.28447810e-01
6.77291274e-01 3.76991808e-01 9.28590357e-01 -1.43478692e+00
-4.04751569e-01 -4.25342560e-01 -8.25843811e-01 -8.47268164e-01
3.14279050e-01 -1.27144063e+00 -2.83978343e-01 -1.25576127e+00
6.63990304e-02 -3.75331223e-01 -3.96075368e-01 4.07065421e-01
-7.53697157e-01 -2.62336075e-01 1.85039252e-01 -2.51909316e-01
-6.37356937e-01 9.11456645e-01 1.23213339e+00 -1.56552866e-01
6.40367642e-02 -3.31938356e-01 -6.91248715e-01 4.28879350e-01
5.42011261e-01 -5.82664907e-01 -7.33128726e-01 -4.27913278e-01
2.49601915e-01 1.51378075e-02 2.34600067e-01 -1.00201690e+00
7.12777197e-01 1.56760756e-02 -5.40000685e-02 -3.33013773e-01
5.82696013e-02 -8.66418958e-01 9.03968737e-02 5.87361157e-01
9.76706967e-02 4.10306007e-01 -1.17301188e-01 1.02697968e+00
-3.18181187e-01 -2.49949455e-01 5.62351048e-01 1.22837707e-01
-6.61017060e-01 7.91719317e-01 6.00086093e-01 2.19179079e-01
8.75127494e-01 -7.91920200e-02 -4.45582747e-01 -1.91545784e-01
-1.24645032e-01 4.80080098e-01 3.61241251e-01 3.91128391e-01
7.03545451e-01 -1.71300006e+00 -4.68872786e-01 3.68631005e-01
1.21350475e-01 2.38330394e-01 3.64869624e-01 9.00086045e-01
-3.94284606e-01 -2.66677383e-02 -7.13671893e-02 -4.70297664e-01
-1.22316170e+00 8.34665775e-01 3.75633617e-03 -6.49913192e-01
-5.54970264e-01 8.21266651e-01 -7.20114484e-02 -6.70420289e-01
1.95381641e-01 -1.90358967e-01 -1.52166486e-01 -2.20137592e-02
4.89716381e-01 4.81687605e-01 1.88826099e-01 -3.19778174e-01
-2.58482039e-01 8.53944302e-01 -3.27281356e-01 5.22824824e-01
1.20410001e+00 2.68098935e-02 -3.26232046e-01 1.88697129e-02
1.53368533e+00 -1.21995024e-01 -8.04920316e-01 -4.14513826e-01
-6.02708682e-02 -6.33270562e-01 1.41308859e-01 -1.46300167e-01
-1.25258863e+00 9.60758388e-01 4.78180438e-01 1.91948086e-01
1.09651339e+00 -4.75980014e-01 1.00871110e+00 6.43238842e-01
2.78420955e-01 -8.84281158e-01 9.03691798e-02 3.68757337e-01
6.94574237e-01 -1.28989625e+00 1.97409749e-01 -5.83053589e-01
-2.34557703e-01 1.23477101e+00 8.23419511e-01 -3.55594724e-01
6.69032574e-01 -2.18288139e-01 -5.38674951e-01 -2.89649844e-01
-6.68205261e-01 -1.52134895e-02 5.12965262e-01 4.32028264e-01
5.32336794e-02 3.53939421e-02 -5.00513375e-01 6.03520393e-01
-1.16831757e-01 -2.54254609e-01 3.07738543e-01 5.61179638e-01
-3.42354476e-01 -1.18413651e+00 -7.31961755e-03 4.49050099e-01
-8.40084702e-02 -1.90339118e-01 -2.51738250e-01 9.00756955e-01
1.18330149e-02 6.53972983e-01 -1.20100975e-01 -6.87733114e-01
2.82041967e-01 -6.49426281e-02 5.09649277e-01 -3.13284934e-01
-6.30602777e-01 -5.33443272e-01 -1.39586926e-01 -7.06958473e-01
-8.32343590e-04 -1.74982250e-01 -1.26037562e+00 -7.27040827e-01
-7.32031524e-01 3.28011721e-01 3.88990283e-01 8.67815495e-01
4.84921694e-01 4.01854634e-01 8.87795210e-01 -4.41019714e-01
-1.10249019e+00 -8.22006822e-01 -4.43907589e-01 5.70550203e-01
6.18222319e-02 -6.36382222e-01 -7.13043869e-01 -5.06779611e-01] | [7.213446617126465, 6.225886344909668] |
c9a0562a-d6fc-40cd-aae6-52f66fee4728 | addressing-inherent-uncertainty-risk | 2102.03119 | null | https://arxiv.org/abs/2102.03119v1 | https://arxiv.org/pdf/2102.03119v1.pdf | Addressing Inherent Uncertainty: Risk-Sensitive Behavior Generation for Automated Driving using Distributional Reinforcement Learning | For highly automated driving above SAE level~3, behavior generation algorithms must reliably consider the inherent uncertainties of the traffic environment, e.g. arising from the variety of human driving styles. Such uncertainties can generate ambiguous decisions, requiring the algorithm to appropriately balance low-probability hazardous events, e.g. collisions, and high-probability beneficial events, e.g. quickly crossing the intersection. State-of-the-art behavior generation algorithms lack a distributional treatment of decision outcome. This impedes a proper risk evaluation in ambiguous situations, often encouraging either unsafe or conservative behavior. Thus, we propose a two-step approach for risk-sensitive behavior generation combining offline distribution learning with online risk assessment. Specifically, we first learn an optimal policy in an uncertain environment with Deep Distributional Reinforcement Learning. During execution, the optimal risk-sensitive action is selected by applying established risk criteria, such as the Conditional Value at Risk, to the learned state-action return distributions. In intersection crossing scenarios, we evaluate different risk criteria and demonstrate that our approach increases safety, while maintaining an active driving style. Our approach shall encourage further studies about the benefits of risk-sensitive approaches for self-driving vehicles. | ['Alois Knoll', 'Stefan Pollok', 'Julian Bernhard'] | 2021-02-05 | null | null | null | null | ['distributional-reinforcement-learning'] | ['methodology'] | [ 1.55748889e-01 2.02440873e-01 -3.68428558e-01 -6.59922421e-01
-1.10279143e+00 -4.88437533e-01 4.52437937e-01 2.78149813e-01
-6.19901419e-01 1.00547183e+00 2.34436452e-01 -8.17539334e-01
-5.04646838e-01 -1.08112335e+00 -6.98299110e-01 -6.07525647e-01
-7.46400803e-02 6.44668162e-01 3.29586476e-01 -4.30806965e-01
2.42471054e-01 6.72263026e-01 -2.03185105e+00 -9.48041454e-02
1.07409823e+00 9.77578819e-01 -6.97316974e-02 4.82636064e-01
-4.96299490e-02 6.79271817e-01 -6.59868479e-01 -3.53960037e-01
3.22072357e-01 -2.07900554e-01 -4.20882404e-01 -4.15055931e-01
-8.02333057e-02 -6.10329926e-01 -1.14006169e-01 1.06467187e+00
2.96464115e-01 4.66649115e-01 9.06575620e-01 -1.83301950e+00
2.55575776e-01 7.17775404e-01 -1.97036475e-01 2.89051503e-01
1.40878648e-01 8.89383435e-01 6.55557692e-01 -1.20858006e-01
1.40957981e-01 1.33266020e+00 1.22476257e-01 5.15142560e-01
-1.26828253e+00 -9.00677562e-01 4.69389290e-01 3.35842788e-01
-1.25721049e+00 -1.96024954e-01 8.48753095e-01 -4.58562016e-01
6.89826071e-01 2.02539846e-01 4.58409309e-01 1.09035289e+00
7.50586212e-01 4.13351834e-01 1.14087117e+00 2.65824422e-02
6.26084685e-01 3.27244140e-02 -5.97603805e-02 -8.32770299e-03
4.92793173e-01 9.89078879e-01 -1.94513872e-01 -1.54177949e-01
-1.68376312e-01 -3.08157027e-01 4.76699203e-01 -2.14404583e-01
-6.35385334e-01 6.83135808e-01 1.31433487e-01 -2.39247456e-01
-5.29685378e-01 3.59843463e-01 3.87947381e-01 1.57053128e-01
1.03285804e-01 4.78468120e-01 -1.53369352e-01 -5.26636183e-01
-6.21967852e-01 9.15496171e-01 5.75698435e-01 8.33789110e-01
8.91803503e-01 1.91937730e-01 -7.27167904e-01 4.18220967e-01
2.46933773e-01 7.00029135e-01 -3.06227237e-01 -1.09690058e+00
7.49828577e-01 1.64429858e-01 5.15320063e-01 -6.26041293e-01
-4.05025452e-01 -2.81400174e-01 -2.66708642e-01 7.79517889e-01
4.49162662e-01 -6.29630446e-01 -8.00691724e-01 1.92641449e+00
3.40424985e-01 7.69887641e-02 1.15434043e-01 6.82583630e-01
-1.99396998e-01 5.40383577e-01 7.96339929e-01 -1.67354807e-01
1.03939223e+00 -7.51329437e-02 -8.40077221e-01 -3.18574518e-01
4.41854298e-01 -3.43799263e-01 8.09882343e-01 3.99407446e-01
-1.02659595e+00 -3.42485368e-01 -9.96210217e-01 6.50858104e-01
-3.64731342e-01 -3.96833092e-01 4.29919034e-01 8.61406863e-01
-5.19502878e-01 6.00993931e-01 -6.08708978e-01 1.50180548e-01
4.34785247e-01 3.74083072e-01 1.66842461e-01 1.24313273e-01
-1.62582576e+00 1.22756016e+00 6.00002944e-01 1.90083515e-02
-1.50499618e+00 -8.83741200e-01 -9.12032723e-01 -1.80570688e-02
8.96947503e-01 -2.78657287e-01 1.51412153e+00 -4.49677736e-01
-1.52346146e+00 1.39533326e-01 1.96797594e-01 -8.83715332e-01
7.35078216e-01 -2.17002302e-01 -6.63935602e-01 -1.76954359e-01
9.13256556e-02 6.32885873e-01 7.38968551e-01 -1.42938435e+00
-1.09361923e+00 -2.91436836e-02 3.45883429e-01 3.86764586e-01
1.94542319e-01 -1.39854088e-01 2.99434781e-01 -2.05824912e-01
-6.05094016e-01 -1.08505523e+00 -7.30739295e-01 -4.39145625e-01
-4.13648605e-01 -2.15847805e-01 5.14009476e-01 -2.96185076e-01
1.62406111e+00 -2.13897443e+00 -4.68208849e-01 6.49639130e-01
-2.30102822e-01 -3.56342159e-02 1.23443127e-01 3.28706354e-01
1.05489120e-01 6.77145505e-03 -1.94950804e-01 1.65965572e-01
2.93288022e-01 1.80809557e-01 -4.85804558e-01 2.42271915e-01
4.65270668e-01 5.34010828e-01 -1.09571433e+00 -2.75179356e-01
4.72260773e-01 -4.21490818e-02 -5.58601320e-01 4.29280698e-01
-4.46681678e-01 3.25492024e-01 -7.23895192e-01 3.35734427e-01
6.39786065e-01 7.62939274e-01 -1.71566069e-01 7.43330941e-02
-3.72700065e-01 3.08646798e-01 -1.19508386e+00 6.94529593e-01
-7.18244433e-01 1.82233334e-01 3.30834580e-03 -7.12123215e-01
8.60903025e-01 -5.00060469e-02 4.76874679e-01 -8.18762958e-01
2.91232109e-01 1.37793213e-01 2.56057352e-01 -3.10174763e-01
5.78090310e-01 -3.77359122e-01 -6.71082675e-01 4.57717299e-01
-3.54635924e-01 -5.13649821e-01 1.65551096e-01 8.29960331e-02
1.02760637e+00 1.47377104e-01 6.25747144e-02 -3.36739808e-01
2.63281167e-01 1.30629763e-01 7.14081466e-01 7.79029012e-01
-6.19554639e-01 -2.10291706e-02 9.69621241e-01 -2.14600220e-01
-7.81123996e-01 -1.30171609e+00 -1.22954115e-01 9.00722623e-01
3.70623529e-01 -2.85082553e-02 -8.42268646e-01 -7.14034855e-01
3.10918748e-01 1.72183645e+00 -4.57961053e-01 -1.05749393e+00
-5.15713573e-01 -5.67517936e-01 4.78931904e-01 5.40911555e-01
1.71708733e-01 -8.83361042e-01 -8.07253301e-01 3.95534962e-01
8.84150043e-02 -7.77839422e-01 -3.60202610e-01 3.20972145e-01
-3.75790417e-01 -8.45983744e-01 2.24896204e-02 5.47851138e-02
4.85133678e-01 -8.54723081e-02 9.43749011e-01 -2.89401442e-01
3.04362159e-02 2.89973825e-01 6.43639266e-02 -7.66305566e-01
-7.69472182e-01 -2.08630756e-01 3.12490016e-01 1.11918673e-01
3.73800844e-01 -2.65656382e-01 -6.88131094e-01 5.83152652e-01
-7.13718534e-01 -3.12385350e-01 5.83695233e-01 5.52649319e-01
6.66405618e-01 6.54743552e-01 9.57090259e-01 -5.97686112e-01
9.17063534e-01 -7.05096841e-01 -8.85661721e-01 1.16688050e-01
-7.27173209e-01 4.56928700e-01 6.28940940e-01 -4.07154888e-01
-1.44097626e+00 7.12346211e-02 -3.78202170e-01 -1.56321421e-01
-5.74144304e-01 1.73023731e-01 -5.88393033e-01 2.94951379e-01
6.56885028e-01 -1.39935911e-01 -7.80678242e-02 2.21446201e-01
3.90743375e-01 5.17054200e-01 4.35401917e-01 -1.10575521e+00
1.02620077e+00 1.13260835e-01 1.44948870e-01 -2.64059573e-01
-6.57964170e-01 -3.19307074e-02 -1.36047527e-01 -7.42859423e-01
9.07510519e-01 -7.49539256e-01 -9.85425293e-01 3.73018146e-01
-6.39200747e-01 -6.19420350e-01 -5.43668389e-01 4.58626628e-01
-9.86145377e-01 -2.18658254e-01 4.22695652e-02 -1.22526932e+00
2.63962299e-01 -1.40409875e+00 9.48166609e-01 2.87558198e-01
-3.37517172e-01 -6.97995245e-01 -1.81362182e-01 2.97291934e-01
4.77200568e-01 3.72752845e-01 1.00229883e+00 -5.67013323e-01
-6.51253819e-01 -1.14801995e-01 1.98603645e-01 2.67591387e-01
6.75860047e-03 5.81856072e-03 -8.36465418e-01 -8.42515379e-02
-4.53980774e-01 -3.08130592e-01 4.36995447e-01 3.96451056e-01
1.39235938e+00 -2.11195126e-01 -3.95767987e-01 1.28324926e-01
1.06880105e+00 7.07621813e-01 7.30720818e-01 3.25060397e-01
2.47786909e-01 1.01932573e+00 1.33461607e+00 6.41461790e-01
4.46438462e-01 7.61659384e-01 7.32937634e-01 2.74391502e-01
2.72687733e-01 -5.42344689e-01 4.86424506e-01 -4.64900076e-01
2.70552307e-01 -3.02202523e-01 -9.10282433e-01 5.48930049e-01
-1.61230779e+00 -1.17769170e+00 2.12583542e-01 2.55389881e+00
9.18095231e-01 9.20402288e-01 3.68106186e-01 1.16462760e-01
6.04678333e-01 3.11887972e-02 -8.09218228e-01 -7.64802814e-01
3.22501153e-01 -1.89340204e-01 9.08620477e-01 7.12311685e-01
-9.13339496e-01 8.55854928e-01 6.20114660e+00 9.99710977e-01
-8.70531023e-01 -2.57375091e-01 1.09272873e+00 -3.88734043e-01
-8.37262511e-01 7.47476444e-02 -1.08375549e+00 6.26626194e-01
1.38260829e+00 -7.04758346e-01 3.29112947e-01 8.37497950e-01
6.90740705e-01 -4.07048225e-01 -1.09737015e+00 4.29770082e-01
-5.45168877e-01 -9.98964608e-01 -8.78043845e-02 -2.96324701e-03
4.44527745e-01 -3.89580280e-01 9.58676189e-02 8.02798450e-01
8.39462221e-01 -9.74507928e-01 1.11525142e+00 6.51158035e-01
6.23225570e-01 -1.50653291e+00 5.85682094e-01 3.59614700e-01
-9.59296525e-01 -4.36262518e-01 1.06889524e-01 5.32381423e-02
6.21674478e-01 7.22435892e-01 -7.06908405e-01 2.91461200e-01
4.68586028e-01 6.11624531e-02 -3.28360587e-01 8.52533519e-01
-2.33275667e-01 7.70748794e-01 -4.37606633e-01 -2.93728292e-01
3.20064604e-01 -1.63289934e-01 7.00953543e-01 8.80280077e-01
2.99581915e-01 1.29480407e-01 3.44140232e-01 9.22624052e-01
4.85208899e-01 -4.08379436e-01 -9.16845083e-01 2.31788680e-01
7.19588995e-01 8.89845908e-01 -4.36890125e-01 -1.03746764e-01
1.33609250e-01 1.14618815e-01 -1.17241610e-02 4.05594647e-01
-1.20514274e+00 -3.43863219e-01 1.13792276e+00 2.68250257e-01
-1.63639069e-01 -7.52679184e-02 -5.32113314e-01 -2.13836163e-01
-7.34002516e-02 -7.44150341e-01 3.30909550e-01 -4.48428661e-01
-1.13337016e+00 4.20691997e-01 7.40039527e-01 -1.43140161e+00
-4.85661447e-01 -4.33364183e-01 -9.43249464e-01 7.92083383e-01
-1.50446343e+00 -5.21297276e-01 1.20320894e-01 4.41129506e-01
3.99237067e-01 -6.84488937e-02 8.44155550e-02 1.61162809e-01
-5.49539328e-01 6.77361131e-01 -4.23071653e-01 -4.76028115e-01
6.10942185e-01 -1.19431710e+00 2.24797219e-01 7.40127504e-01
-7.58350730e-01 1.37070686e-01 1.15317690e+00 -9.38968122e-01
-1.29451287e+00 -1.34602582e+00 4.98022646e-01 -4.48193491e-01
7.56985486e-01 -3.70904177e-01 -7.86409795e-01 2.20820248e-01
-5.10021783e-02 -3.48142356e-01 3.60977769e-01 -8.52562040e-02
1.55012414e-01 -4.28630322e-01 -1.40219164e+00 1.19818556e+00
9.75656390e-01 -1.88808769e-01 -4.45821553e-01 3.48634869e-02
8.01531851e-01 -3.17881793e-01 -5.52219093e-01 5.79960227e-01
2.69654065e-01 -6.80642962e-01 7.44670808e-01 -6.35010302e-01
6.57983571e-02 -4.06386316e-01 -1.06738202e-01 -1.62678874e+00
-2.72954434e-01 -7.53311455e-01 1.26057327e-01 1.14342809e+00
7.64531195e-01 -6.76896274e-01 4.55268770e-01 1.04920805e+00
-3.64944518e-01 -8.05448949e-01 -1.24666011e+00 -8.61656427e-01
2.63666272e-01 -1.02560031e+00 1.08669281e+00 1.14475377e-01
-1.17516570e-01 -1.12380810e-01 -2.90616184e-01 4.76188749e-01
7.78077841e-01 -2.21939653e-01 6.60661280e-01 -7.32012570e-01
-3.50842485e-03 -6.92078412e-01 -1.77162603e-01 -4.26905006e-01
5.42967916e-01 -3.17077518e-01 6.62063658e-01 -1.22118652e+00
-3.22649330e-01 -7.45522380e-01 -4.11456168e-01 2.72101790e-01
-3.45736116e-01 -4.94092166e-01 -3.55360322e-02 -5.36498129e-01
-5.29285073e-01 7.97032297e-01 1.00410628e+00 -3.04684371e-01
-3.20574462e-01 4.66310382e-01 -7.21940994e-01 4.38078433e-01
1.08688200e+00 -4.80168521e-01 -8.11248004e-01 1.98745042e-01
3.05599749e-01 3.53635043e-01 1.64567485e-01 -1.11261153e+00
-8.12135935e-02 -1.15526021e+00 -2.18269095e-01 -7.93568552e-01
3.54791246e-02 -1.06041324e+00 2.12360516e-01 6.11483693e-01
-5.30566752e-01 -2.60750409e-02 5.76878607e-01 8.45896363e-01
-7.61752129e-02 -1.61659308e-02 7.07110345e-01 3.43829930e-01
-8.07763219e-01 3.00720185e-01 -1.00018358e+00 1.17422074e-01
1.60632479e+00 -2.08625481e-01 -1.46341860e-01 -5.68946481e-01
-5.34425318e-01 8.74365866e-01 7.67931268e-02 7.14819133e-01
4.76336151e-01 -1.32550919e+00 -6.85192287e-01 2.82728016e-01
3.50945085e-01 -1.56155899e-01 5.00573337e-01 4.27557379e-01
-3.16831246e-02 1.05856434e-01 -2.19736829e-01 -3.74316037e-01
-5.78215182e-01 5.18035173e-01 5.99616051e-01 -2.73851961e-01
-2.20524445e-01 2.13680044e-01 2.08377302e-01 -2.35368639e-01
1.18594073e-01 -3.06178093e-01 -2.35406071e-01 3.13449427e-02
5.35288870e-01 5.87874830e-01 1.54644907e-01 -4.67315197e-01
-3.69680226e-01 1.24510258e-01 1.10899121e-01 -4.06343728e-01
7.08223581e-01 2.40303017e-02 5.56072891e-01 2.32000694e-01
5.42090416e-01 -1.56395376e-01 -1.86057842e+00 4.10984099e-01
1.08680628e-01 -4.98180687e-01 1.80828258e-01 -9.51960862e-01
-7.81054556e-01 5.81660092e-01 5.79045713e-01 1.26311526e-01
1.08400452e+00 -2.90622205e-01 7.12820828e-01 2.16262713e-01
7.90798306e-01 -1.54284024e+00 -1.89576551e-01 4.96850699e-01
7.55128860e-01 -1.11350977e+00 -2.30611622e-01 -1.70385212e-01
-1.06466341e+00 5.19051969e-01 1.07386291e+00 1.41795009e-01
7.94672012e-01 6.27408981e-01 1.65279105e-01 5.13024256e-02
-1.03018880e+00 -2.27886900e-01 1.92880258e-01 7.21514463e-01
-1.43231720e-01 5.39716959e-01 -3.47529411e-01 5.09524226e-01
-2.54351884e-01 -2.08308741e-01 5.19102812e-01 9.71327066e-01
-5.49787521e-01 -1.06791615e+00 -5.27375758e-01 7.13909447e-01
-1.35986060e-01 3.58216107e-01 -1.66314803e-02 6.53478861e-01
2.51959831e-01 1.22643006e+00 2.04473808e-01 -4.28402543e-01
8.34186554e-01 7.10035264e-02 2.75400840e-02 -5.14077425e-01
-3.93633157e-01 -4.61801201e-01 4.90928382e-01 -7.64221430e-01
1.91018015e-01 -9.04857516e-01 -1.48271143e+00 -2.58409619e-01
2.04917993e-02 1.21007852e-01 4.92689729e-01 1.05544817e+00
3.08983684e-01 7.52404273e-01 8.67279828e-01 -7.78945923e-01
-1.10651898e+00 -3.48922521e-01 -4.40777868e-01 3.72765929e-01
3.59561652e-01 -1.24216843e+00 -4.33193982e-01 -4.96557921e-01] | [4.8270440101623535, 1.8835680484771729] |
0510bda8-37e7-445e-a0ed-1e0815a27fc2 | joint-coreference-resolution-for-zeros-and | 2210.12169 | null | https://arxiv.org/abs/2210.12169v1 | https://arxiv.org/pdf/2210.12169v1.pdf | Joint Coreference Resolution for Zeros and non-Zeros in Arabic | Most existing proposals about anaphoric zero pronoun (AZP) resolution regard full mention coreference and AZP resolution as two independent tasks, even though the two tasks are clearly related. The main issues that need tackling to develop a joint model for zero and non-zero mentions are the difference between the two types of arguments (zero pronouns, being null, provide no nominal information) and the lack of annotated datasets of a suitable size in which both types of arguments are annotated for languages other than Chinese and Japanese. In this paper, we introduce two architectures for jointly resolving AZPs and non-AZPs, and evaluate them on Arabic, a language for which, as far as we know, there has been no prior work on joint resolution. Doing this also required creating a new version of the Arabic subset of the standard coreference resolution dataset used for the CoNLL-2012 shared task (Pradhan et al.,2012) in which both zeros and non-zeros are included in a single dataset. | ['Massimo Poesio', 'Sameer Pradhan', 'Abdulrahman Aloraini'] | 2022-10-21 | null | null | null | null | ['coreference-resolution'] | ['natural-language-processing'] | [-1.57423168e-02 5.90377092e-01 -4.00636405e-01 -4.70785290e-01
-1.15644312e+00 -1.00950551e+00 9.34844136e-01 1.51684031e-01
-5.64943314e-01 1.16121995e+00 8.66118491e-01 -2.25865111e-01
-3.68297040e-01 -5.45879781e-01 -3.06207657e-01 -4.77592021e-01
1.20366082e-01 1.26702952e+00 7.43314505e-01 -8.16087425e-01
2.24010035e-01 4.73280609e-01 -1.48160326e+00 7.16925025e-01
5.70512295e-01 7.06281662e-02 6.82714581e-02 1.92223489e-01
-2.09739581e-01 6.17169142e-01 -7.40973771e-01 -9.03318405e-01
-1.50095941e-02 -4.95110750e-02 -1.48222268e+00 -7.00951278e-01
6.66739702e-01 7.28195682e-02 -9.81074646e-02 9.45122600e-01
5.55772007e-01 5.82775213e-02 5.37446201e-01 -1.17666626e+00
-4.48773503e-01 1.39654958e+00 -5.45328379e-01 5.26872873e-01
6.27904773e-01 -4.99630898e-01 1.38010669e+00 -6.24737263e-01
1.12853813e+00 1.94619882e+00 7.00121760e-01 8.12841058e-01
-1.15346277e+00 -6.42581105e-01 -2.11187005e-02 3.24564040e-01
-1.10558331e+00 -7.16076791e-01 4.29108620e-01 -2.98916809e-02
1.52106988e+00 5.26193738e-01 -5.52250668e-02 1.17072022e+00
-1.58546373e-01 4.57416594e-01 1.02057886e+00 -8.13446105e-01
-2.28266090e-01 -8.93457159e-02 4.50176299e-01 -4.18954678e-02
5.12899697e-01 -1.27072111e-01 -4.79185373e-01 -2.83557415e-01
4.78346139e-01 -8.10864985e-01 -4.51654792e-01 -2.89619446e-01
-1.09309125e+00 7.91560769e-01 1.80313632e-01 8.22176158e-01
-1.56263515e-01 -3.84132028e-01 3.47746015e-01 2.33921289e-01
-4.99629006e-02 5.77770948e-01 -6.62433445e-01 -6.35982156e-02
-4.29464638e-01 7.10762501e-01 1.20017982e+00 9.89438534e-01
2.95156181e-01 -5.43166459e-01 7.86222517e-02 8.34458530e-01
9.42995921e-02 1.63916767e-01 2.62879342e-01 -1.30865431e+00
9.77578580e-01 4.61507589e-01 6.98479354e-01 -9.69422877e-01
-6.64290726e-01 1.49250925e-01 -2.35980656e-02 -1.22196183e-01
9.59955513e-01 -1.43108349e-02 -1.42424196e-01 2.14153671e+00
2.43219495e-01 -8.74566212e-02 7.21859276e-01 1.12038326e+00
1.22755718e+00 4.21035975e-01 3.15330774e-01 -5.31314492e-01
2.03982282e+00 -7.22697496e-01 -1.18671095e+00 -3.72684807e-01
6.60643220e-01 -1.06747973e+00 7.20845759e-01 1.06210157e-01
-1.39678550e+00 1.04301438e-01 -9.81163085e-01 -4.34453458e-01
-3.28570813e-01 -2.83622175e-01 8.36246967e-01 3.01531523e-01
-7.93769062e-01 3.55799913e-01 -6.55536652e-01 -7.16923177e-01
-1.76684827e-01 4.51768428e-01 -7.28015780e-01 -1.22200459e-01
-1.85116816e+00 1.58096838e+00 6.36427939e-01 -4.73756008e-02
-6.40498251e-02 -4.07301366e-01 -7.38388181e-01 4.04813215e-02
6.46791160e-01 -2.92880684e-01 1.58663511e+00 -8.17935169e-01
-8.96376550e-01 1.34113646e+00 -2.43587434e-01 -4.47817624e-01
9.31928903e-02 -5.58658719e-01 -6.89686775e-01 -1.13911048e-01
5.58949471e-01 5.68550944e-01 1.10909879e-01 -1.29764211e+00
-9.86632466e-01 -6.00461602e-01 6.01073861e-01 4.86218572e-01
5.00946455e-02 8.52839351e-01 -3.33939701e-01 -4.75660294e-01
1.65478215e-01 -7.95471847e-01 2.75902897e-01 -7.84412742e-01
-2.25686342e-01 -8.67266536e-01 6.70962572e-01 -6.15006387e-01
1.29576838e+00 -1.73762953e+00 4.79000419e-01 -3.33776623e-01
-1.15436442e-01 4.18696374e-01 -2.83981591e-01 7.27881551e-01
-4.18093711e-01 1.27577871e-01 -3.14571857e-01 -6.86174035e-02
1.06392466e-01 7.06800342e-01 -4.89708245e-01 2.99796581e-01
3.95617515e-01 5.80279410e-01 -1.03996027e+00 -4.70657676e-01
-3.20160359e-01 4.28420812e-01 -3.84328842e-01 -4.68343496e-02
-1.49047464e-01 9.45628285e-02 -2.19310254e-01 4.81815517e-01
7.51018524e-01 3.79725695e-01 8.75802755e-01 -4.11249250e-01
-5.77980697e-01 1.04122221e+00 -1.48345780e+00 1.44374514e+00
1.26740903e-01 3.66089642e-01 5.23108125e-01 -7.72629797e-01
5.07169366e-01 7.77396739e-01 1.75046958e-02 -6.47556782e-01
1.12510003e-01 5.25142968e-01 3.20112675e-01 -1.83400080e-01
5.25194764e-01 -3.76862526e-01 -4.16260779e-01 4.53104675e-01
1.19403675e-01 9.69477519e-02 6.87170923e-01 2.45470151e-01
1.00098193e+00 2.53059864e-01 5.65902650e-01 -6.19524181e-01
8.19294453e-01 3.43263119e-01 1.12559295e+00 5.33894360e-01
-3.04342270e-01 5.03553808e-01 7.17440128e-01 -3.03754628e-01
-7.01162040e-01 -9.09974337e-01 -6.28896415e-01 1.23418832e+00
2.06115052e-01 -4.87269998e-01 -8.28423917e-01 -6.39695466e-01
-2.60135531e-01 1.01425576e+00 -3.10927153e-01 4.68974650e-01
-1.53013480e+00 -8.19501758e-01 7.18490303e-01 4.44555283e-01
-5.86792640e-02 -1.43861997e+00 -7.18210220e-01 4.78559136e-01
-7.80322015e-01 -1.28993630e+00 -8.21482241e-02 2.79991716e-01
-4.50532824e-01 -1.52985752e+00 -7.34063685e-02 -1.00581455e+00
-1.92517471e-02 7.41767436e-02 1.50843084e+00 1.26094818e-01
1.03300907e-01 1.52339384e-01 -3.74268830e-01 -5.76564908e-01
-3.76033843e-01 1.07993297e-01 1.74294367e-01 -7.34977245e-01
1.02980256e+00 -5.44606864e-01 1.58058792e-01 1.32833244e-02
-6.40811682e-01 -4.17580694e-01 2.97226787e-01 7.59725034e-01
1.61023483e-01 -4.20798600e-01 5.59140444e-01 -1.29294467e+00
6.32213235e-01 -4.94862825e-01 -3.80192012e-01 1.92546189e-01
-1.65729046e-01 -1.49405524e-02 -2.29860982e-03 -2.06874922e-01
-1.43644881e+00 -3.07037026e-01 -1.61332279e-01 3.69133562e-01
-2.45929942e-01 4.81699228e-01 -6.04850650e-01 4.77031648e-01
6.02542937e-01 -8.11002433e-01 -2.55144060e-01 -7.65869856e-01
5.41669428e-01 6.32410884e-01 1.10741341e+00 -1.04688358e+00
5.33964932e-01 2.58215636e-01 -1.52880296e-01 -5.91318429e-01
-1.22324145e+00 -5.39738357e-01 -1.14291310e+00 5.45150101e-01
7.80320704e-01 -9.10062790e-01 -5.52701950e-01 1.03823163e-01
-1.74196053e+00 3.08395158e-02 3.56945954e-02 2.96633899e-01
-4.76827770e-01 6.47145510e-01 -9.07766044e-01 -5.56832373e-01
-4.23124909e-01 -9.20077860e-01 8.15730929e-01 1.85440168e-01
-7.55932987e-01 -9.20715332e-01 1.71765640e-01 3.17511171e-01
2.66438395e-01 1.91058189e-01 1.24003351e+00 -1.04546261e+00
-9.56992358e-02 4.75305505e-02 -3.00893009e-01 -4.39775348e-01
9.58261639e-02 -2.87535101e-01 -7.45333314e-01 -3.87771159e-01
2.28589952e-01 -3.57730836e-01 4.56140161e-01 2.55791452e-02
-8.70001465e-02 -4.40297991e-01 -5.06782234e-01 7.69647434e-02
1.14652324e+00 1.13277219e-01 8.07251275e-01 1.02101350e+00
2.35049367e-01 1.07874227e+00 1.08338594e+00 -1.27274647e-01
8.52310538e-01 1.10973454e+00 5.36484182e-01 2.78168082e-01
-3.18115979e-01 3.49830300e-01 2.60251820e-01 7.67773271e-01
-2.16742724e-01 -2.62641281e-01 -1.08111548e+00 8.93471718e-01
-2.00613356e+00 -1.23228824e+00 -6.34932518e-01 2.09201264e+00
1.33628917e+00 1.86589174e-02 -1.34931028e-01 1.21731661e-01
8.36795986e-01 4.49045785e-02 2.83883989e-01 -4.54857588e-01
-6.27424717e-01 3.73575151e-01 -2.31293142e-02 9.87743616e-01
-1.54863250e+00 1.32488191e+00 5.78094769e+00 2.50111938e-01
-6.17762387e-01 1.79305732e-01 -3.54368627e-01 1.75806373e-01
-1.51350379e-01 6.42781198e-01 -1.20016932e+00 -4.50717844e-02
1.08287072e+00 5.60615547e-02 1.99077547e-01 3.20460916e-01
-4.01280850e-01 8.32703039e-02 -1.26439810e+00 4.82376963e-01
1.63962394e-01 -1.04365420e+00 9.83090848e-02 -3.50659251e-01
2.27515340e-01 5.82197234e-02 -5.56353450e-01 4.35140014e-01
5.24497807e-01 -9.17927384e-01 7.64328361e-01 2.70502269e-01
3.02045494e-01 -6.84190094e-01 1.19614112e+00 2.69157529e-01
-8.44773650e-01 -6.29029125e-02 -4.68322456e-01 -2.76279356e-02
4.37274605e-01 -1.86672986e-01 -3.40828836e-01 7.41933763e-01
8.28756809e-01 3.72909546e-01 -2.36390278e-01 4.96868074e-01
-5.06990433e-01 3.80509615e-01 -3.08814496e-01 3.86197865e-01
1.65634051e-01 2.34478563e-02 1.09297407e+00 1.45527160e+00
1.27871819e-02 7.96413600e-01 2.07072392e-01 5.51630378e-01
1.65391952e-01 -1.07697584e-02 -8.87501240e-02 2.42672950e-01
1.16045249e+00 1.34346986e+00 -4.99692172e-01 5.37965409e-02
-6.26685023e-01 5.81714928e-01 6.44664764e-01 -1.26919681e-02
-4.49382961e-01 -3.94678265e-01 6.38055742e-01 1.13884203e-01
1.01829499e-01 7.54717812e-02 2.42917240e-02 -1.11984980e+00
-1.39946520e-01 -1.23619854e+00 1.11256516e+00 -6.53871417e-01
-1.55893350e+00 6.16070092e-01 5.53363204e-01 -6.60835385e-01
-6.44255161e-01 -6.30389094e-01 -3.69563371e-01 1.13631296e+00
-1.75139046e+00 -1.41195130e+00 5.09884488e-03 5.07261097e-01
2.03879043e-01 2.60234550e-02 1.43357348e+00 4.13410991e-01
-3.55895400e-01 5.81243753e-01 -5.05524635e-01 1.96907297e-01
1.26130521e+00 -1.36594808e+00 2.67291427e-01 8.55888426e-01
7.54128918e-02 1.22516775e+00 1.11465538e+00 -5.85647345e-01
-1.30727684e+00 -3.85992199e-01 1.81012261e+00 -1.15050626e+00
8.84833634e-01 -3.63271348e-02 -1.40631878e+00 1.15864575e+00
6.39214337e-01 -3.28487039e-01 4.48024094e-01 4.44988459e-01
-5.11659026e-01 3.43107909e-01 -1.13938177e+00 5.20153284e-01
9.83441591e-01 -2.69508123e-01 -1.62394738e+00 1.97981074e-01
7.91967928e-01 -7.43236184e-01 -1.01482713e+00 7.05535173e-01
1.38655767e-01 -9.00413096e-01 1.02221406e+00 -8.96004200e-01
1.09758444e-01 -2.58920878e-01 -5.56538284e-01 -8.24003041e-01
-4.30763513e-01 -3.42646390e-01 -1.14766225e-01 1.91826236e+00
4.22485322e-01 -5.28247058e-01 3.15363526e-01 6.69675529e-01
-3.70141745e-01 -5.61092235e-02 -1.24836409e+00 -3.51401657e-01
5.54960310e-01 -1.13083124e-01 5.69999516e-01 1.45830822e+00
4.78321135e-01 8.75184357e-01 -6.63868571e-03 3.94909829e-01
5.42641699e-01 2.52359688e-01 4.82760549e-01 -1.89479709e+00
6.39535189e-02 -3.74229014e-01 -9.15510133e-02 -2.97481328e-01
6.78506315e-01 -9.65808749e-01 -9.01083127e-02 -1.52091134e+00
3.09001476e-01 -4.41355556e-01 -9.29265097e-02 8.88941169e-01
-1.96440578e-01 6.36924664e-03 2.21960604e-01 4.96170998e-01
-3.46532971e-01 7.90099278e-02 5.64348459e-01 7.43231624e-02
-1.23913102e-01 -5.00956118e-01 -1.02918243e+00 1.01624799e+00
2.80985534e-01 -6.18079901e-01 1.66587569e-02 -6.36342347e-01
4.10124138e-02 2.18915060e-01 -1.21925026e-02 -3.06854039e-01
2.70701408e-01 -2.54270852e-01 -2.74433672e-01 -4.16820556e-01
3.83531600e-01 -6.24813557e-01 -1.12353742e-01 1.13742866e-01
-2.03977153e-01 2.84151316e-01 4.90774512e-01 1.10798925e-01
-3.11692238e-01 -6.16799831e-01 4.80433613e-01 -2.34269515e-01
-9.73064423e-01 -4.71035987e-01 -2.08026379e-01 4.40009654e-01
7.02377558e-01 2.25019351e-01 -9.07549620e-01 1.92072969e-02
-6.46200836e-01 2.93009251e-01 3.04134399e-01 8.41333926e-01
1.43056005e-01 -1.11804092e+00 -1.02185142e+00 -4.60402340e-01
8.45340639e-03 -6.67408258e-02 -1.57370538e-01 7.29295671e-01
-2.42797375e-01 7.84598827e-01 -3.05431575e-01 -1.68182045e-01
-1.53239429e+00 5.33681452e-01 1.78781614e-01 -4.68118459e-01
-6.05076790e-01 5.15365243e-01 -5.70119284e-02 -8.86571586e-01
1.98664442e-01 4.01740164e-01 -1.00823712e+00 4.19768393e-01
7.68438280e-01 1.38332516e-01 2.19495684e-01 -1.29780686e+00
-7.80221760e-01 2.52427220e-01 -2.82042205e-01 -1.64007753e-01
1.71398032e+00 -1.73725128e-01 -9.51042533e-01 3.87799531e-01
3.70945692e-01 4.47510570e-01 -3.91396791e-01 -3.27161193e-01
9.10331607e-01 -2.38398999e-01 -3.54366094e-01 -9.29246247e-01
-3.93334538e-01 5.89155078e-01 3.06288332e-01 1.71979796e-02
6.83498204e-01 3.85959119e-01 5.29573739e-01 4.79554266e-01
5.33026040e-01 -8.43064010e-01 -4.83822852e-01 1.05926645e+00
1.30910432e+00 -8.58532250e-01 7.83001408e-02 -8.31752300e-01
-5.19007742e-01 1.19818282e+00 8.60324144e-01 9.01494995e-02
-8.05412978e-02 5.96075833e-01 3.12314212e-01 -4.41583991e-01
-8.88231754e-01 -3.11569035e-01 1.24364555e-01 8.18235517e-01
1.15891695e+00 1.68288611e-02 -1.16813624e+00 1.22281408e+00
-4.30621624e-01 -4.84921575e-01 8.03214192e-01 1.02470469e+00
-3.23137909e-01 -1.60908186e+00 -5.38387537e-01 -1.68190911e-01
-1.05004001e+00 -2.05326393e-01 -7.10110426e-01 1.29256260e+00
-2.64270734e-02 1.07510817e+00 3.01490009e-01 2.77789056e-01
7.35823631e-01 1.10976189e-01 5.92992663e-01 -8.84299099e-01
-6.17974281e-01 -1.99489579e-01 9.97090995e-01 -4.58405435e-01
-9.10560489e-01 -9.99115407e-01 -1.70263171e+00 -2.02256814e-01
-5.36070824e-01 4.58631635e-01 1.33131504e-01 1.19299567e+00
-1.03238083e-01 6.27772361e-02 -2.39626780e-01 -9.43974137e-01
-6.79294765e-01 -1.07066727e+00 -4.15758878e-01 7.11838424e-01
-7.64108151e-02 -9.93916631e-01 -3.25696290e-01 -2.85502255e-01] | [9.359686851501465, 9.51179027557373] |
af27db38-133d-4b05-9836-a096cd80a1ac | deep-recurrent-neural-network-with-multi | 2112.05150 | null | https://arxiv.org/abs/2112.05150v1 | https://arxiv.org/pdf/2112.05150v1.pdf | Deep Recurrent Neural Network with Multi-scale Bi-directional Propagation for Video Deblurring | The success of the state-of-the-art video deblurring methods stems mainly from implicit or explicit estimation of alignment among the adjacent frames for latent video restoration. However, due to the influence of the blur effect, estimating the alignment information from the blurry adjacent frames is not a trivial task. Inaccurate estimations will interfere the following frame restoration. Instead of estimating alignment information, we propose a simple and effective deep Recurrent Neural Network with Multi-scale Bi-directional Propagation (RNN-MBP) to effectively propagate and gather the information from unaligned neighboring frames for better video deblurring. Specifically, we build a Multi-scale Bi-directional Propagation~(MBP) module with two U-Net RNN cells which can directly exploit the inter-frame information from unaligned neighboring hidden states by integrating them in different scales. Moreover, to better evaluate the proposed algorithm and existing state-of-the-art methods on real-world blurry scenes, we also create a Real-World Blurry Video Dataset (RBVD) by a well-designed Digital Video Acquisition System (DVAS) and use it as the training and evaluation dataset. Extensive experimental results demonstrate that the proposed RBVD dataset effectively improves the performance of existing algorithms on real-world blurry videos, and the proposed algorithm performs favorably against the state-of-the-art methods on three typical benchmarks. The code is available at https://github.com/XJTU-CVLAB-LOWLEVEL/RNN-MBP. | ['Fei Wang', 'Lean Fu', 'Yuhao Huang', 'Boyang Liang', 'Jinshan Pan', 'Hang Dong', 'Chao Zhu'] | 2021-12-09 | null | null | null | null | ['video-restoration'] | ['computer-vision'] | [-8.47529843e-02 -7.10359931e-01 -1.21808492e-01 -6.97000697e-02
-4.94530559e-01 -1.51811674e-01 2.80663162e-01 -8.75690460e-01
-1.46578506e-01 7.86278367e-01 7.51857579e-01 -2.13656187e-01
1.45711690e-01 -2.06553459e-01 -7.67069519e-01 -8.73654902e-01
2.62103928e-03 -2.97022879e-01 1.73024148e-01 8.83030072e-02
1.18736856e-01 9.76244584e-02 -1.21234560e+00 2.20693395e-01
9.31918919e-01 7.57645428e-01 5.49410284e-01 7.65202463e-01
3.73248577e-01 1.36336851e+00 -6.13144457e-01 9.03562903e-02
1.96737856e-01 -6.03670359e-01 -5.23757815e-01 1.50307491e-01
4.53279823e-01 -1.11852241e+00 -1.07584405e+00 1.33899653e+00
5.96954823e-01 2.50158161e-01 1.29269570e-01 -8.00226986e-01
-1.03621304e+00 4.89776909e-01 -8.52314949e-01 8.22945416e-01
1.87100276e-01 4.05885816e-01 4.24014747e-01 -9.24176455e-01
5.23251534e-01 1.27891636e+00 5.37505448e-01 5.43834031e-01
-7.50110507e-01 -8.12725544e-01 2.11520582e-01 7.89712071e-01
-1.31201530e+00 -7.88393676e-01 8.90780330e-01 -3.16776991e-01
6.75882936e-01 2.25733951e-01 4.00185972e-01 1.32026637e+00
2.60282695e-01 9.47687864e-01 1.11594748e+00 1.08498015e-01
2.79772338e-02 -4.14260268e-01 2.36632377e-01 4.83500957e-01
9.63354334e-02 2.63592243e-01 -2.93648928e-01 1.66819081e-01
1.23336506e+00 3.11493903e-01 -1.19562376e+00 -3.38690414e-04
-1.37312210e+00 3.08336467e-01 6.36970401e-01 4.33662027e-01
-5.18498421e-01 3.16249847e-01 3.45165700e-01 2.06182867e-01
7.04099536e-01 4.14468311e-02 -2.43363112e-01 -3.31369102e-01
-1.21301746e+00 -7.20894337e-02 3.11007828e-01 8.19032192e-01
6.34845555e-01 2.65010625e-01 -3.78262639e-01 9.35451090e-01
4.02332008e-01 3.35557699e-01 7.52601862e-01 -1.08374119e+00
5.29271960e-01 -1.22350276e-01 3.89063567e-01 -1.10516536e+00
6.88304082e-02 -5.99448681e-01 -1.36920798e+00 -1.01821817e-01
2.65371948e-01 -1.89694718e-01 -1.00374961e+00 1.71658909e+00
7.84823019e-03 1.13812566e+00 9.58992839e-02 1.60402143e+00
7.39460289e-01 1.08236480e+00 -3.64517063e-01 -6.32761121e-01
1.01034093e+00 -1.45505166e+00 -1.02711606e+00 -2.39118040e-01
6.48945197e-02 -9.24269676e-01 7.17108369e-01 3.38071823e-01
-9.94159997e-01 -9.43618655e-01 -1.04203522e+00 -3.06825966e-01
2.75220335e-01 2.38775745e-01 3.06157440e-01 6.25218302e-02
-1.12275517e+00 4.31325614e-01 -1.00847518e+00 3.31130326e-02
4.24546897e-01 1.99619718e-02 -2.26644471e-01 -6.00732565e-01
-1.43630469e+00 8.02897990e-01 2.78423261e-02 7.88302004e-01
-1.36970377e+00 -5.63587785e-01 -8.29971135e-01 1.03702452e-02
2.31568217e-01 -8.75767231e-01 1.08877373e+00 -9.97826219e-01
-1.39961457e+00 2.08571851e-01 -4.49234635e-01 -3.52534235e-01
5.92528403e-01 -7.26264656e-01 -4.23569113e-01 1.74159929e-01
-7.67073035e-02 4.28679436e-01 1.28995538e+00 -1.24159336e+00
-2.73652762e-01 1.74066722e-02 3.31307873e-02 3.05962384e-01
-3.83263715e-02 6.35359511e-02 -6.43045366e-01 -1.11196327e+00
-2.41417214e-02 -6.54798567e-01 2.34815683e-02 -2.86459297e-01
-3.00915062e-01 2.04912633e-01 1.11339605e+00 -1.33943808e+00
1.59596646e+00 -2.23146152e+00 4.59639609e-01 -6.36527240e-01
4.31952000e-01 5.92788756e-01 -2.01445058e-01 4.70960438e-02
-2.36722842e-01 -2.12697640e-01 -1.81888014e-01 -4.09028649e-01
-3.44042718e-01 6.85850680e-02 -4.45736706e-01 7.46424377e-01
-1.77735433e-01 7.10255146e-01 -1.09426200e+00 -6.17715158e-02
5.65193295e-01 8.22370887e-01 -3.90440524e-01 5.51657557e-01
7.12765157e-02 6.74739480e-01 -2.22128972e-01 4.73587811e-01
9.71534312e-01 -3.47403914e-01 -7.09121153e-02 -5.60513198e-01
-1.02045827e-01 1.10465512e-01 -8.60859573e-01 1.84371161e+00
-3.34808737e-01 9.87137616e-01 1.75661936e-01 -7.12999940e-01
4.25777435e-01 4.13615704e-01 1.69905871e-01 -3.38423789e-01
1.62220508e-01 1.30564034e-01 -5.60924821e-02 -7.86131561e-01
5.20720184e-01 2.27044418e-01 4.81554091e-01 2.90396035e-01
-1.23792939e-01 4.34539646e-01 -5.42112440e-02 1.10730395e-01
1.13927817e+00 2.32851475e-01 1.47556454e-01 -8.58008116e-03
6.65771365e-01 -6.35210693e-01 8.18147540e-01 5.87906182e-01
-6.49742305e-01 9.71916497e-01 1.18342958e-01 -4.01083976e-01
-1.01664245e+00 -9.12270367e-01 2.23210305e-01 5.66275060e-01
6.14185989e-01 -1.62325963e-01 -8.39721560e-01 -4.14027691e-01
-3.56674552e-01 5.74575067e-01 -4.11380440e-01 -6.58824295e-02
-7.59773850e-01 -9.98204470e-01 3.22123408e-01 2.43578374e-01
9.19143379e-01 -9.27047729e-01 -2.17213169e-01 1.81536332e-01
-7.18082845e-01 -1.19759274e+00 -9.08986807e-01 -5.16953290e-01
-7.34200120e-01 -9.48701441e-01 -1.23719764e+00 -7.90973663e-01
6.68577254e-01 8.60996246e-01 8.41021836e-01 1.64331160e-02
1.98223099e-01 -1.80735961e-01 -3.18910152e-01 5.35610199e-01
-3.72472167e-01 -2.60939419e-01 1.62059709e-01 1.51090652e-01
1.59915969e-01 -7.53477454e-01 -1.02617562e+00 3.54063332e-01
-1.02769339e+00 4.74249899e-01 6.02287412e-01 9.68427420e-01
2.05406830e-01 2.15260625e-01 4.63867098e-01 -3.95056278e-01
5.89380503e-01 -5.91052592e-01 -5.26881814e-01 1.45148575e-01
-2.58268982e-01 -1.46341592e-01 5.08980989e-01 -4.60579395e-01
-1.26005089e+00 -5.37251770e-01 4.65890318e-02 -1.15465462e+00
-1.97144404e-01 4.43407118e-01 -1.21439055e-01 1.73569530e-01
1.91695452e-01 5.50603330e-01 -1.48293257e-01 -6.52419448e-01
4.02213722e-01 8.53722572e-01 8.55729699e-01 -7.15262368e-02
6.71571612e-01 2.53652811e-01 -5.21617413e-01 -6.95441127e-01
-7.16252446e-01 -3.78999233e-01 -2.39207432e-01 -2.86628664e-01
7.73963332e-01 -1.42914712e+00 -4.43930656e-01 1.26495719e+00
-1.50785089e+00 -4.12810117e-01 3.53274852e-01 6.60101593e-01
-2.50718176e-01 9.41249073e-01 -1.39269066e+00 -4.18123096e-01
-4.46747512e-01 -1.57354927e+00 7.21925139e-01 3.84314328e-01
2.23706022e-01 -8.55612814e-01 -4.35235538e-02 6.33953631e-01
6.48079395e-01 -2.47267649e-01 4.04320389e-01 7.89398253e-02
-1.06735528e+00 1.57384917e-01 -5.89147449e-01 7.55925477e-01
5.70195794e-01 -2.61423528e-01 -1.00914788e+00 -5.19578218e-01
4.76664901e-01 8.45857859e-02 1.15823102e+00 7.84755051e-01
1.04412794e+00 -4.95401859e-01 -3.21385786e-02 8.61885071e-01
1.29062748e+00 -9.80138779e-04 9.75463688e-01 2.44597197e-01
1.12662864e+00 -1.95179284e-01 4.56442505e-01 2.72631019e-01
4.87762064e-01 6.66760743e-01 4.76892769e-01 3.51146907e-02
-6.36255682e-01 -4.52756956e-02 7.89313197e-01 1.42926729e+00
-1.94672897e-01 -3.98878276e-01 -4.18803006e-01 5.93459129e-01
-2.11794329e+00 -1.14220917e+00 -5.35826758e-02 1.94715142e+00
1.14254236e+00 -1.31909490e-01 -3.91380012e-01 -2.60281384e-01
9.57635760e-01 7.54452705e-01 -4.91767645e-01 2.91390002e-01
-2.60111809e-01 -3.39879751e-01 3.55289131e-01 9.07824874e-01
-1.15338695e+00 9.81312096e-01 5.19275856e+00 9.85590994e-01
-1.28075683e+00 3.68747145e-01 9.36561584e-01 -1.86043769e-01
5.72110154e-03 -1.29309922e-01 -4.87844944e-01 9.77788508e-01
7.72961080e-01 2.02348739e-01 9.83483016e-01 3.72522205e-01
8.65055203e-01 -1.83916837e-01 -7.51858175e-01 1.17882729e+00
1.25353619e-01 -1.30074775e+00 -1.04825974e-01 -3.45362723e-01
9.40318584e-01 3.73456448e-01 -2.53072418e-02 1.88122243e-02
6.33658245e-02 -7.79161870e-01 5.43778002e-01 8.71271372e-01
8.21594059e-01 -4.22716975e-01 1.00679386e+00 2.84878910e-01
-9.55260038e-01 -7.23527465e-03 -5.24717987e-01 -2.06906632e-01
4.54619288e-01 9.71816838e-01 -2.90638655e-01 6.19128227e-01
7.97931552e-01 1.40960670e+00 -3.57500196e-01 1.07937217e+00
-4.96460110e-01 7.99038589e-01 7.97949657e-02 6.28576994e-01
1.79039791e-01 -3.15711915e-01 8.11599433e-01 1.14679790e+00
4.55678016e-01 1.62603691e-01 -2.34870270e-01 8.12058866e-01
-1.85492441e-01 -6.25166714e-01 -1.59471557e-01 2.70524442e-01
3.55943710e-01 1.12598062e+00 -1.76529258e-01 -6.17665350e-01
-5.00311434e-01 1.41545415e+00 8.63658488e-02 8.83849144e-01
-1.17585945e+00 -1.27105072e-01 1.05684483e+00 -3.03024054e-01
4.69458759e-01 -3.22194129e-01 1.66650757e-01 -1.82467139e+00
5.89885227e-02 -1.14327157e+00 -7.81811550e-02 -1.35482979e+00
-1.29563594e+00 7.56144166e-01 -2.57674336e-01 -1.38820398e+00
-1.40066653e-01 -1.33650482e-01 -3.99820745e-01 1.09671533e+00
-1.80385411e+00 -9.06026006e-01 -6.16787493e-01 4.78226781e-01
9.29954350e-01 7.58630559e-02 2.48964414e-01 6.67247415e-01
-1.02185738e+00 2.64839083e-01 2.90454686e-01 2.65671402e-01
1.07416499e+00 -7.72450030e-01 5.40352046e-01 1.53455698e+00
-2.75926799e-01 7.25762963e-01 8.65709722e-01 -7.48156428e-01
-1.37595642e+00 -1.16182351e+00 5.08492708e-01 -9.31634754e-02
7.23380089e-01 3.08815613e-02 -1.09892738e+00 6.96999729e-01
6.19740009e-01 3.27417850e-01 -9.05725583e-02 -6.31670892e-01
-1.66138887e-01 -1.72368690e-01 -6.79874718e-01 5.29671371e-01
1.05225015e+00 -7.24379838e-01 -6.01746976e-01 8.48688781e-02
9.20236051e-01 -7.22563148e-01 -5.40659666e-01 5.54130018e-01
3.52427006e-01 -1.20779788e+00 1.06567836e+00 -1.06623463e-01
8.69833231e-01 -8.01379263e-01 -8.10869783e-02 -1.54739213e+00
-5.32560766e-01 -8.77723038e-01 -6.84982717e-01 1.03175902e+00
-2.42703170e-01 -6.21087492e-01 3.71482521e-01 2.65425444e-01
-3.56173187e-01 -5.93980372e-01 -7.38377869e-01 -4.46708292e-01
-4.71728116e-01 -2.68101156e-01 3.81284952e-01 1.10252988e+00
-3.14603239e-01 2.89376765e-01 -1.16273475e+00 5.62749088e-01
7.21277654e-01 9.43267494e-02 5.10638475e-01 -2.14758560e-01
-4.73106176e-01 -2.02766836e-01 -2.73511559e-01 -1.89927495e+00
1.81085486e-02 -2.58174956e-01 1.59563228e-01 -1.41918850e+00
3.13466311e-01 -1.00582585e-01 -5.49241543e-01 1.36515498e-01
-6.14514947e-01 3.15828949e-01 1.94007710e-01 5.80703974e-01
-5.80006063e-01 7.31147170e-01 1.56227481e+00 -1.72267839e-01
4.97635677e-02 -1.79490820e-01 -3.95563990e-01 6.41375124e-01
4.92270738e-01 -2.19178125e-01 -4.55590874e-01 -9.75173295e-01
-1.84874341e-01 4.35732037e-01 5.48569977e-01 -9.68166649e-01
3.83767575e-01 2.15999130e-02 4.14321303e-01 -5.55164039e-01
3.17129910e-01 -6.87272787e-01 4.34326649e-01 3.47705394e-01
-2.09698111e-01 2.95591969e-02 1.88809875e-02 5.79512596e-01
-5.01950264e-01 9.10932571e-02 7.64363289e-01 -4.76821400e-02
-8.77367258e-01 2.91367292e-01 -2.79095322e-01 -1.47363767e-01
5.88533640e-01 9.64477099e-03 -9.25081909e-01 -7.60415375e-01
-3.69154036e-01 -7.67080439e-03 6.21720016e-01 5.89004815e-01
8.05994213e-01 -1.19438636e+00 -7.88098395e-01 2.59528220e-01
-4.81102884e-01 -6.69688061e-02 7.74904490e-01 1.12147915e+00
-7.57559836e-01 2.93047786e-01 -2.39271358e-01 -5.16013443e-01
-1.13453889e+00 7.04825819e-01 4.64171290e-01 -1.19212076e-01
-5.67470968e-01 9.07743990e-01 5.70218682e-01 2.32667238e-01
9.74861160e-02 -5.22479236e-01 -3.78294550e-02 -3.01311821e-01
8.56132627e-01 5.61210036e-01 -3.69126707e-01 -7.75900602e-01
-1.78335160e-01 3.80726546e-01 -2.45927572e-01 2.47781560e-01
1.21355736e+00 -7.53662527e-01 -4.25846875e-01 2.01209977e-01
1.19715834e+00 -5.26245758e-02 -1.77288914e+00 -3.18456978e-01
-6.91946983e-01 -7.44885743e-01 4.30820227e-01 -7.00107753e-01
-1.54566574e+00 7.56251872e-01 6.70932293e-01 -1.43439710e-01
1.49037194e+00 -4.23536927e-01 1.15004706e+00 -1.50254667e-01
2.36634403e-01 -4.81895655e-01 2.50107124e-02 4.75476056e-01
9.15366948e-01 -1.19640481e+00 1.49429321e-01 -3.27618629e-01
-4.38974619e-01 9.72391367e-01 5.78216851e-01 -2.25910038e-01
5.82268715e-01 8.10563117e-02 2.96501309e-01 2.15713054e-01
-7.61346996e-01 2.83456743e-01 2.74948955e-01 1.80928335e-01
3.05174649e-01 -1.42544463e-01 1.12126733e-03 4.13376451e-01
2.59171754e-01 4.30072159e-01 8.27938378e-01 4.52165931e-01
-3.18550587e-01 -6.18490517e-01 -4.61543947e-01 1.23171516e-01
-5.20750225e-01 -5.22924066e-01 2.48947948e-01 3.72926027e-01
-3.15634608e-02 1.06193399e+00 -9.42672789e-02 -4.29330081e-01
-2.91062474e-01 -4.55943674e-01 3.28604192e-01 -3.02984901e-02
-1.76985830e-01 4.40583646e-01 -9.41341817e-02 -7.28420913e-01
-7.12966144e-01 -5.38885176e-01 -7.88406193e-01 -5.63254535e-01
-3.50918323e-01 -9.93111283e-02 4.53206822e-02 9.85643625e-01
5.98414779e-01 7.47089505e-01 6.54872775e-01 -1.26869810e+00
-4.16172951e-01 -1.38179374e+00 -3.43405426e-01 2.90812701e-01
9.72295702e-01 -4.53639686e-01 -5.96092880e-01 3.42250913e-01] | [11.449673652648926, -2.508033514022827] |
9567848e-5e58-4c14-b3fc-40ceb534a090 | cascade-luminance-and-chrominance-for-image | 2205.15999 | null | https://arxiv.org/abs/2205.15999v1 | https://arxiv.org/pdf/2205.15999v1.pdf | Cascade Luminance and Chrominance for Image Retouching: More Like Artist | Photo retouching aims to adjust the luminance, contrast, and saturation of the image to make it more human aesthetically desirable. However, artists' actions in photo retouching are difficult to quantitatively analyze. By investigating their retouching behaviors, we propose a two-stage network that brightens images first and then enriches them in the chrominance plane. Six pieces of useful information from image EXIF are picked as the network's condition input. Additionally, hue palette loss is added to make the image more vibrant. Based on the above three aspects, Luminance-Chrominance Cascading Net(LCCNet) makes the machine learning problem of mimicking artists in photo retouching more reasonable. Experiments show that our method is effective on the benchmark MIT-Adobe FiveK dataset, and achieves state-of-the-art performance for both quantitative and qualitative evaluation. | ['Xingyue Cheng', 'Chenyu Dong', 'Xi Xiao', 'Sibo Feng', 'Hailong Ma'] | 2022-05-31 | null | null | null | null | ['photo-retouching', 'image-retouching'] | ['computer-vision', 'computer-vision'] | [ 5.05872667e-01 -4.64989915e-02 -1.89148605e-01 -1.69626921e-01
-7.30248764e-02 -3.00621063e-01 3.71193558e-01 -3.46624196e-01
-2.00554311e-01 6.40062451e-01 2.48942614e-01 -2.23477632e-01
1.97219595e-01 -7.33201265e-01 -8.85369122e-01 -8.61164212e-01
5.45797765e-01 -6.37691617e-01 1.52130410e-01 -3.42459708e-01
3.50365758e-01 4.00359899e-01 -1.55869102e+00 4.91268337e-01
8.83230507e-01 9.70403135e-01 1.15468390e-01 5.81113815e-01
8.40394795e-02 8.75538409e-01 -5.04562855e-01 -6.47233725e-01
4.88136530e-01 -5.11268795e-01 -7.49332905e-01 3.24724078e-01
4.86048520e-01 -5.17223120e-01 -4.93273050e-01 1.36615539e+00
5.56247830e-01 2.31329620e-01 3.27344030e-01 -1.27333605e+00
-1.46894717e+00 7.87252247e-01 -1.25316298e+00 1.57259598e-01
1.86393023e-01 6.47881448e-01 9.20025170e-01 -6.28260076e-01
4.26356643e-01 1.31502151e+00 3.51346403e-01 5.90256453e-01
-1.25009990e+00 -9.10833001e-01 5.47952298e-03 4.41172868e-01
-1.24444914e+00 -2.17952684e-01 1.11432695e+00 4.47825231e-02
5.53825974e-01 6.32615387e-01 1.01440477e+00 9.20010746e-01
2.75840670e-01 1.11948752e+00 1.38107228e+00 -4.29365396e-01
1.21169083e-01 1.03556074e-01 -4.87856179e-01 3.02668601e-01
1.04443945e-01 3.47692251e-01 -4.02686983e-01 3.99897069e-01
1.10581863e+00 1.71233520e-01 -4.69649553e-01 -2.33578354e-01
-1.03511477e+00 2.50138909e-01 1.08630896e+00 1.38726667e-01
-3.42951775e-01 4.30988759e-01 7.99582005e-02 -1.58940628e-02
3.14008683e-01 6.08394921e-01 -1.09735094e-01 8.77314731e-02
-6.09369278e-01 -8.94712508e-02 -1.86523274e-02 6.57529056e-01
9.73449707e-01 1.56253800e-01 -4.80453163e-01 1.03703797e+00
8.76656696e-02 4.06900942e-01 4.50904250e-01 -1.11305189e+00
1.07229739e-01 7.13983715e-01 1.13771260e-01 -1.01031995e+00
-1.67074427e-01 -2.84372270e-01 -1.09867871e+00 6.24401629e-01
2.39524335e-01 -1.44383699e-01 -1.14627707e+00 1.39578795e+00
1.33436814e-01 2.93429941e-01 6.19573912e-05 1.26280296e+00
7.94180572e-01 6.84795082e-01 3.07386577e-01 -2.56917685e-01
1.24171102e+00 -1.09847784e+00 -8.75194013e-01 -2.55673677e-01
-3.02275699e-02 -7.77363181e-01 1.67677093e+00 3.17525208e-01
-1.43080747e+00 -7.51099825e-01 -1.16063142e+00 -2.57745087e-01
-2.00679824e-01 1.59664720e-01 5.57224393e-01 4.08016175e-01
-1.04520202e+00 5.88632047e-01 -3.31139117e-01 -6.97605461e-02
5.38427770e-01 3.31640244e-04 1.42913964e-02 5.09887114e-02
-1.14808249e+00 6.08333111e-01 3.43322545e-01 3.81247938e-01
-4.33921993e-01 -7.21974552e-01 -6.85693741e-01 1.89937428e-01
3.36585492e-01 -6.77334607e-01 1.35318983e+00 -1.48602295e+00
-1.80448639e+00 8.13367248e-01 2.55221099e-01 -8.85885656e-02
7.40430415e-01 1.02312379e-01 -6.67552292e-01 5.29457442e-02
-3.85139555e-01 1.05844915e+00 9.74658251e-01 -1.65608299e+00
-7.66094446e-01 -9.69768167e-02 2.15600461e-01 5.13528526e-01
-5.18880486e-01 -4.43917841e-01 -6.64003313e-01 -7.61972189e-01
-2.41223425e-01 -6.44891858e-01 -1.60623029e-01 1.03892609e-01
-8.08787227e-01 4.44369279e-02 7.69803286e-01 -3.52948606e-01
1.37498319e+00 -2.30552578e+00 -1.67031348e-01 1.20951578e-01
4.54088062e-01 4.18384761e-01 -2.61740625e-01 3.68162394e-02
-3.02948296e-01 3.37741613e-01 -6.08745515e-02 4.47368734e-02
-2.40841687e-01 3.70450690e-02 -5.86712547e-02 1.41081735e-01
1.68859422e-01 1.29280698e+00 -1.01967442e+00 -3.95786047e-01
4.59724128e-01 4.66996253e-01 -3.66878390e-01 -2.88398080e-02
-5.35164448e-03 8.66329819e-02 -1.20566964e-01 5.82156301e-01
1.05334485e+00 -2.12043718e-01 -1.01772435e-01 -6.37906492e-01
-1.77067310e-01 -3.54931176e-01 -8.03890526e-01 1.41594541e+00
-3.90457988e-01 8.83422434e-01 -4.61257428e-01 -7.21025988e-02
8.15733194e-01 -2.16913670e-01 1.40237704e-01 -1.23997498e+00
2.71240413e-01 -1.72521964e-01 -8.87689143e-02 -7.32917726e-01
7.37429619e-01 -1.53315157e-01 2.63715029e-01 3.23911071e-01
-6.96763337e-01 -1.43968076e-01 -5.09385951e-02 1.53511241e-02
7.59640098e-01 -8.70918408e-02 1.89936131e-01 2.94936690e-02
3.23999792e-01 -3.51334035e-01 4.81128782e-01 4.49118137e-01
-4.50903267e-01 6.87712371e-01 3.72094303e-01 -5.17521799e-01
-1.23249829e+00 -1.12981141e+00 1.91642284e-01 1.10077310e+00
7.69489527e-01 3.95464385e-03 -8.23616445e-01 -4.08820868e-01
-7.67007172e-02 8.83344471e-01 -9.11155045e-01 -5.03191948e-01
-2.94987977e-01 -6.04399085e-01 3.90838623e-01 4.75893766e-01
1.15972519e+00 -1.37191081e+00 -4.97584969e-01 -1.77736834e-01
-1.87660947e-01 -7.31104910e-01 -7.94631541e-01 -3.25433224e-01
-4.40943360e-01 -9.82923746e-01 -1.02531993e+00 -7.40411758e-01
7.16623843e-01 8.52980912e-01 7.40692556e-01 2.46364892e-01
-3.78805757e-01 2.57885903e-02 -3.29597503e-01 -1.88016996e-01
-4.04343337e-01 -1.78225756e-01 -3.41833502e-01 1.01529844e-01
1.61098853e-01 -5.22829056e-01 -1.25421524e+00 3.53306204e-01
-1.40523517e+00 5.17568886e-01 8.23187888e-01 6.72603965e-01
5.98556399e-01 4.65259105e-01 3.86110067e-01 -7.21215963e-01
9.08374071e-01 4.02515791e-02 -2.04867497e-01 4.34323311e-01
-7.12607563e-01 -1.48762837e-01 6.26930296e-01 -8.90759051e-01
-1.58759665e+00 -4.90025319e-02 2.97373772e-01 -4.94207144e-01
9.09495428e-02 -4.43316028e-02 -1.31856248e-01 -2.73592591e-01
7.84984231e-01 1.29315078e-01 -1.79336324e-01 -1.45734116e-01
7.39195943e-01 6.84522212e-01 8.35349381e-01 -2.56540049e-02
8.99826705e-01 5.81432700e-01 -1.88790411e-01 -4.93711799e-01
-7.36829579e-01 -8.93544629e-02 -2.73281395e-01 -6.50427580e-01
7.43716538e-01 -5.63719034e-01 -1.14232051e+00 7.25771129e-01
-8.69718850e-01 -5.54216802e-01 -5.97437203e-01 -3.30462940e-02
-2.94844896e-01 2.93216497e-01 -5.78288615e-01 -7.68760085e-01
-3.88610631e-01 -1.05606461e+00 7.67770708e-01 8.99488568e-01
2.35230103e-01 -6.03895724e-01 -4.85965610e-02 2.08412424e-01
3.21081519e-01 1.28891364e-01 8.00810754e-01 4.85767186e-01
-5.54461479e-01 4.54027206e-02 -8.47903311e-01 4.90365326e-01
3.14839929e-01 4.14336443e-01 -1.10728419e+00 2.41018645e-03
-4.95169908e-01 -1.44819975e-01 9.80390251e-01 4.88363594e-01
1.58178270e+00 -5.97391188e-01 -4.32090200e-02 6.57706678e-01
1.42496705e+00 3.36124808e-01 1.31598675e+00 5.79578221e-01
8.17134500e-01 4.31140214e-01 4.82581377e-01 3.58386040e-01
2.75892854e-01 3.67478728e-01 6.43247247e-01 -7.72135615e-01
-6.35085583e-01 -5.04867017e-01 2.05969125e-01 2.78969765e-01
-1.46545202e-01 -4.08144444e-01 -5.10367334e-01 3.57944936e-01
-1.71527708e+00 -1.09139061e+00 -5.76040894e-02 1.97496891e+00
1.03359032e+00 1.20654870e-02 3.41953672e-02 5.72202168e-03
1.10009813e+00 1.80470377e-01 -8.24391305e-01 -3.98533762e-01
-4.35651600e-01 -1.71244428e-01 6.80675268e-01 2.02023268e-01
-7.69906878e-01 1.01820731e+00 6.60450649e+00 9.88309979e-01
-1.29316878e+00 -2.63579160e-01 1.19201076e+00 -1.37000293e-01
-7.45983601e-01 -1.97312161e-01 -1.67830959e-01 6.25040770e-01
9.88878831e-02 -1.30909905e-01 9.40129876e-01 4.99864429e-01
4.95592803e-01 -3.21396232e-01 -6.39506936e-01 1.27241135e+00
-6.83875754e-02 -1.34691179e+00 1.63592204e-01 -1.32313997e-01
1.08708894e+00 -3.38879317e-01 7.89876580e-01 -8.34449567e-03
4.25593883e-01 -8.48662496e-01 7.77711451e-01 5.78772724e-01
1.10993469e+00 -7.29539037e-01 4.08903241e-01 -1.06029786e-01
-1.08059108e+00 -3.01862270e-01 -3.26466471e-01 3.00660338e-02
6.08119443e-02 3.60978931e-01 -6.63437724e-01 1.62260130e-01
8.70502710e-01 6.65038288e-01 -8.35487425e-01 1.26648462e+00
-5.56777537e-01 3.44325542e-01 1.64786801e-01 -1.22074589e-01
2.43606314e-01 -4.40686047e-02 4.32858020e-01 1.05932689e+00
6.49300069e-02 2.63609201e-01 -3.13898236e-01 1.10366011e+00
-5.54565251e-01 3.29834744e-02 -2.75226831e-01 8.08740482e-02
4.32830721e-01 1.49872565e+00 -6.81464255e-01 -2.85883307e-01
8.78312141e-02 1.28221571e+00 6.11544400e-02 7.98153579e-01
-8.96732211e-01 -5.04064143e-01 3.98454785e-01 1.39368013e-01
-1.00308090e-01 4.42209810e-01 -5.40836275e-01 -8.48989964e-01
-8.51265714e-02 -6.59845650e-01 1.57164574e-01 -1.75885844e+00
-1.29795849e+00 5.76117694e-01 -1.11105606e-01 -1.32319117e+00
3.56314510e-01 -2.98043758e-01 -8.24538529e-01 5.64901233e-01
-1.74854589e+00 -1.16866302e+00 -9.08026814e-01 4.40660805e-01
4.37408626e-01 2.60618329e-01 -1.32297026e-02 2.07093105e-01
-8.18861246e-01 6.71049654e-01 -4.44357581e-02 -1.64153785e-01
9.23180461e-01 -1.18811119e+00 4.63153660e-01 9.08156872e-01
-1.68910936e-01 2.55856574e-01 7.93012679e-01 -4.60041016e-01
-1.12242031e+00 -1.11613441e+00 2.09443018e-01 6.98521063e-02
5.17969966e-01 3.61058041e-02 -8.46187830e-01 2.78301984e-01
5.57875335e-01 -2.32528389e-01 2.23286092e-01 -3.12866807e-01
-4.39123690e-01 -3.78637761e-01 -1.08215427e+00 1.14011228e+00
9.93764877e-01 -2.35915422e-01 -1.69092268e-01 2.85078324e-02
8.29529822e-01 -3.41926426e-01 -4.58608031e-01 1.83111399e-01
7.18046248e-01 -9.45416629e-01 1.05319953e+00 -3.64678651e-01
8.35942626e-01 -3.93458366e-01 2.97709703e-01 -1.60883474e+00
-6.09075546e-01 -9.12230432e-01 3.07223231e-01 1.19458616e+00
3.01763564e-01 -4.49190825e-01 6.52386069e-01 6.72106206e-01
7.62454048e-02 -6.70452595e-01 -4.79523808e-01 -4.12393302e-01
-2.83694804e-01 -2.09167227e-01 8.78934085e-01 8.91883075e-01
-2.09453627e-02 2.89700419e-01 -7.45509028e-01 -1.54299751e-01
4.74308431e-01 -6.74565881e-02 7.59741306e-01 -6.72529757e-01
-6.32723197e-02 -9.04102087e-01 -8.08986500e-02 -9.94911015e-01
-2.81844646e-01 -5.39218783e-01 1.48515180e-01 -1.49014151e+00
6.03941202e-01 -2.88384765e-01 -4.63793546e-01 8.94225061e-01
-5.86837828e-01 7.39719450e-01 3.86382312e-01 3.02684963e-01
-5.16325235e-01 5.56614637e-01 2.23743844e+00 -4.73656952e-01
-5.42556703e-01 -3.25441837e-01 -1.15823817e+00 5.89574933e-01
7.97682524e-01 1.62541002e-01 -4.38279957e-01 -4.38625455e-01
3.76455426e-01 -2.23008752e-01 5.57951927e-01 -7.45213330e-01
8.94953683e-02 -6.47112191e-01 7.69900084e-01 -3.62908781e-01
2.87552625e-01 -6.12124324e-01 1.51735380e-01 3.01232189e-01
-5.61141968e-01 -7.35454336e-02 2.59547561e-01 5.28047919e-01
1.26843929e-01 6.94589093e-02 1.02888739e+00 1.22919492e-01
-9.73462462e-01 2.38174349e-01 -3.47395539e-02 -2.73122966e-01
1.06542730e+00 -6.24896884e-01 -8.02326322e-01 -5.49713373e-01
-1.41357511e-01 1.25913069e-01 6.67890251e-01 4.37080204e-01
7.06369102e-01 -1.55704844e+00 -4.92599308e-01 1.04461953e-01
1.01712361e-01 -2.90462762e-01 8.90869498e-01 5.35260201e-01
-5.85123897e-01 -2.37734631e-01 -5.60390592e-01 -3.28781128e-01
-1.21348476e+00 7.68267274e-01 3.83759558e-01 1.57189041e-01
-6.81240976e-01 8.57932627e-01 5.88690639e-01 2.80386120e-01
2.14546978e-01 -3.68803918e-01 -2.60489196e-01 -2.15942264e-01
6.93140030e-01 5.78226388e-01 -3.36779028e-01 -4.26479578e-01
4.60422784e-02 4.26631421e-01 -1.69310033e-01 -1.39059369e-02
1.04648244e+00 -5.48037767e-01 -3.51739228e-02 2.02378288e-01
1.08468604e+00 -9.54082310e-02 -1.66093516e+00 -1.34791166e-01
-8.07191432e-01 -7.84198105e-01 2.36608595e-01 -1.23980105e+00
-1.41550589e+00 7.69903064e-01 7.00620830e-01 5.11647046e-01
1.76380098e+00 -3.10009241e-01 9.78818536e-01 1.10606305e-01
-3.17629844e-01 -1.32544887e+00 6.79793000e-01 -1.38048187e-01
9.62824881e-01 -1.00149071e+00 -7.52069801e-02 -2.95962095e-01
-1.01612377e+00 9.35911238e-01 8.72548938e-01 1.04662925e-02
1.02101654e-01 7.72550777e-02 2.65687346e-01 -5.86549826e-02
-4.16254818e-01 -1.87792018e-01 4.56259578e-01 5.42259574e-01
2.42738035e-02 1.76028699e-01 -1.05378740e-01 2.89744437e-01
-6.04986511e-02 -3.97669673e-02 7.40961790e-01 4.46004629e-01
-5.43058157e-01 -6.07038140e-01 -3.40737015e-01 2.42613152e-01
-1.49118200e-01 -2.00564012e-01 -5.47656119e-01 7.45816231e-01
2.24982992e-01 8.80950749e-01 2.81330962e-02 -6.91942692e-01
4.03589517e-01 -6.36287153e-01 4.41863924e-01 -8.61961022e-02
-4.03609902e-01 3.26269716e-01 -3.18994075e-01 -6.48459733e-01
-4.34772789e-01 -7.46978596e-02 -1.11213982e+00 -6.94517910e-01
-4.19068426e-01 -2.58559793e-01 5.28677046e-01 6.18456244e-01
2.50174195e-01 9.95370567e-01 8.48521054e-01 -8.32069814e-01
-9.78817046e-02 -8.42783511e-01 -6.74807250e-01 6.20294154e-01
2.63429880e-01 -4.01051372e-01 -2.71465212e-01 2.94323415e-01] | [11.31021499633789, -1.2319649457931519] |
895d4ab0-d65a-47ea-9082-598c8f6137e8 | short-term-prediction-of-stream-turbidity | 2210.05821 | null | https://arxiv.org/abs/2210.05821v1 | https://arxiv.org/pdf/2210.05821v1.pdf | Short-term prediction of stream turbidity using surrogate data and a meta-model approach | Many water-quality monitoring programs aim to measure turbidity to help guide effective management of waterways and catchments, yet distributing turbidity sensors throughout networks is typically cost prohibitive. To this end, we built and compared the ability of dynamic regression (ARIMA), long short-term memory neural nets (LSTM), and generalized additive models (GAM) to forecast stream turbidity one step ahead, using surrogate data from relatively low-cost in-situ sensors and publicly available databases. We iteratively trialled combinations of four surrogate covariates (rainfall, water level, air temperature and total global solar exposure) selecting a final model for each type that minimised the corrected Akaike Information Criterion. Cross-validation using a rolling time-window indicated that ARIMA, which included the rainfall and water-level covariates only, produced the most accurate predictions, followed closely by GAM, which included all four covariates. We constructed a meta-model, trained on time-series features of turbidity, to take advantage of the strengths of each model over different time points and predict the best model (that with the lowest forecast error one-step prior) for each time step. The meta-model outperformed all other models, indicating that this methodology can yield high accuracy and may be a viable alternative to using measurements sourced directly from turbidity-sensors where costs prohibit their deployment and maintenance, and when predicting turbidity across the short term. Our findings also indicated that temperature and light-associated variables, for example underwater illuminance, may hold promise as cost-effective, high-frequency surrogates of turbidity, especially when combined with other covariates, like rainfall, that are typically measured at coarse levels of spatial resolution. | ['Catherine Leigh', 'Sevvandi Kandanaarachchi', 'Caleb Hogan', 'Bhargav Rele'] | 2022-10-11 | null | null | null | null | ['additive-models'] | ['methodology'] | [ 7.79090673e-02 -3.97868693e-01 3.49904239e-01 -2.89004922e-01
-7.02402294e-01 -6.56131387e-01 5.98608971e-01 3.37415159e-01
-6.35205865e-01 1.06722939e+00 3.20978314e-01 -8.61223459e-01
-6.48288667e-01 -1.29451966e+00 -4.18796450e-01 -1.10015094e+00
-4.54596013e-01 1.15233913e-01 -3.35151464e-01 -1.01145357e-01
-1.52887981e-02 6.81797206e-01 -1.67920732e+00 -8.59514698e-02
1.06323206e+00 6.67908370e-01 4.09611344e-01 7.85900831e-01
-3.05430859e-01 4.24403071e-01 -4.30624306e-01 1.74123123e-01
4.11804110e-01 -4.20474440e-01 -1.13154694e-01 -2.73776591e-01
4.37012821e-01 -7.66167998e-01 3.16840596e-02 5.60297430e-01
6.72291458e-01 3.79798383e-01 5.06169617e-01 -6.42316043e-01
-1.31744578e-01 4.33111787e-01 -3.52732509e-01 2.05364265e-02
-3.41664463e-01 7.52700329e-01 8.09419036e-01 -4.48084503e-01
-1.18488885e-01 1.37483084e+00 1.07462025e+00 2.55527109e-01
-1.64947987e+00 -6.35019541e-01 2.74380684e-01 -4.19712603e-01
-1.07032359e+00 -4.61556703e-01 -8.99848267e-02 -6.02872252e-01
1.30822289e+00 4.25385565e-01 9.50452387e-01 6.01197422e-01
5.10432541e-01 -1.07866675e-01 1.35752392e+00 -1.31941631e-01
2.94367641e-01 2.39387099e-02 1.62876192e-02 -6.79486766e-02
6.95185006e-01 7.21227050e-01 7.40776807e-02 -4.19797003e-01
6.85552180e-01 2.36258343e-01 -3.76106501e-01 5.79512358e-01
-4.44548070e-01 8.71447921e-01 4.38113987e-01 3.05554573e-03
-7.90870309e-01 1.69880450e-01 -1.14208713e-01 2.74035633e-01
8.46195698e-01 3.69641840e-01 -9.56166923e-01 -2.86164641e-01
-9.96739447e-01 1.00192077e-01 8.44359994e-01 1.66875482e-01
9.27051663e-01 5.72166502e-01 -1.03391055e-02 9.26073909e-01
7.31000841e-01 1.35593069e+00 -1.05559975e-02 -9.03897941e-01
1.04030512e-01 3.02518785e-01 5.15702069e-01 -6.52437925e-01
-8.52874160e-01 -1.54665291e-01 -7.34944642e-01 5.81550896e-01
5.02547383e-01 -9.27144766e-01 -1.10139394e+00 1.50080466e+00
-1.36358812e-01 4.30996507e-01 3.85115951e-01 8.23671639e-01
6.27438247e-01 1.16433406e+00 5.81498682e-01 -4.53933179e-01
9.67792809e-01 9.59530249e-02 -5.42383313e-01 -3.62693012e-01
6.32203400e-01 -3.02848607e-01 6.94344282e-01 -5.61501905e-02
-7.42105901e-01 -1.67942196e-01 -7.08331466e-01 7.03405738e-01
-6.47746742e-01 -3.30955654e-01 6.15429163e-01 8.02802444e-01
-1.24807060e+00 9.53315318e-01 -1.15967214e+00 -3.64186496e-01
9.22990218e-03 1.75854877e-01 -4.86995466e-02 2.94394583e-01
-1.22597122e+00 1.12254977e+00 -1.06188878e-01 8.22642207e-01
-8.38338256e-01 -9.10649538e-01 -9.37097728e-01 1.35361120e-01
-6.82784140e-01 -4.07553643e-01 9.31955636e-01 -8.99449229e-01
-1.62136745e+00 1.81071967e-01 -2.58153856e-01 -4.18680251e-01
2.35454336e-01 -1.49250358e-01 -3.77035618e-01 -2.88071483e-01
-4.89960751e-03 3.11379522e-01 1.77159414e-01 -1.19299090e+00
-8.21767509e-01 -4.61418837e-01 5.58776259e-02 3.69045556e-01
-2.05838934e-01 -1.35781437e-01 6.03238881e-01 6.96581826e-02
-8.10965300e-02 -7.92997599e-01 -6.65687740e-01 6.44809008e-03
3.53838354e-01 1.39143765e-01 5.09286284e-01 -6.87610984e-01
9.23190176e-01 -1.88580441e+00 -5.74748218e-01 2.13311493e-01
-3.46182555e-01 5.35821021e-01 -3.02063346e-01 6.37434959e-01
1.09465234e-01 3.34980935e-01 -2.98448026e-01 -8.01879466e-02
-1.92711771e-01 6.56771898e-01 -2.34599963e-01 7.61863530e-01
3.46479177e-01 5.43190539e-01 -1.07464218e+00 2.03570977e-01
7.43400097e-01 7.49040663e-01 2.89970972e-02 2.00942770e-01
-1.60103530e-01 4.35808510e-01 1.51242450e-01 2.28264630e-01
1.08658755e+00 1.21143505e-01 1.98209420e-01 3.84880483e-01
-7.37086058e-01 2.90315032e-01 -1.22087634e+00 7.68579304e-01
-8.28786790e-01 7.92036355e-01 2.99644470e-01 -3.55310410e-01
1.31347251e+00 6.31178737e-01 3.41227710e-01 -5.90554714e-01
-2.74911672e-01 3.06258798e-01 9.39941853e-02 -8.32387447e-01
3.65712047e-01 -6.93597078e-01 6.33623481e-01 3.61159772e-01
-6.20293498e-01 -1.42520443e-02 -1.68213189e-01 -4.84105170e-01
7.13586807e-01 1.29931375e-01 -1.22696161e-01 -5.60565174e-01
-6.32652314e-04 -8.19404237e-03 4.86139029e-01 9.92555797e-01
2.54589878e-03 4.10164595e-01 2.26665987e-04 -6.11024082e-01
-7.56863892e-01 -9.84415293e-01 -7.26943731e-01 1.06180203e+00
-2.34455764e-01 4.26250011e-01 1.79667100e-01 2.15010405e-01
4.65311021e-01 1.02703881e+00 -5.66852212e-01 1.71920925e-01
-2.99911350e-01 -1.58846664e+00 7.08258748e-01 5.32910466e-01
1.86854005e-01 -9.12532330e-01 -7.96284616e-01 8.51968825e-01
1.42446026e-01 -4.33768243e-01 4.52344120e-01 7.02469468e-01
-1.33016586e+00 -7.58891702e-01 -6.05636716e-01 -3.08815856e-02
2.92801470e-01 3.30532253e-01 1.01692009e+00 -1.05864204e-01
1.28484309e-01 2.43799850e-01 -1.40239418e-01 -6.98630929e-01
-2.87630171e-01 -5.99721968e-01 -1.54543847e-01 -2.35563114e-01
7.85162151e-01 -8.43873858e-01 -6.93149030e-01 -9.56757590e-02
-9.09170449e-01 -4.95182037e-01 4.12474662e-01 6.07009292e-01
1.80302318e-02 -2.24217653e-01 6.43996477e-01 -4.36145365e-01
3.76293838e-01 -7.63143957e-01 -8.41333926e-01 -1.37887031e-01
-5.59894383e-01 -2.52974540e-01 3.87931108e-01 -2.37799987e-01
-1.18077862e+00 -1.53655171e-01 -1.03415236e-01 7.76841566e-02
-4.18886304e-01 1.22074056e+00 1.43860430e-01 1.44787356e-01
7.57311881e-01 2.34177008e-01 2.98194468e-01 -5.93935847e-01
-1.31707385e-01 8.48013818e-01 2.56943726e-03 -4.87075180e-01
5.22204995e-01 6.27702594e-01 5.33997864e-02 -1.38457406e+00
-5.16179621e-01 -4.15273011e-01 -4.99355644e-01 -3.64390314e-01
5.22196293e-01 -1.36851287e+00 -6.16758466e-01 8.75757217e-01
-9.03631985e-01 -7.64660418e-01 -9.79212001e-02 9.09632087e-01
9.55981463e-02 -6.35595769e-02 -5.48362911e-01 -1.61241555e+00
-4.65383530e-01 -7.47440219e-01 7.00203657e-01 3.74292552e-01
6.41650409e-02 -1.53448606e+00 3.76146615e-01 -6.13212764e-01
9.73023355e-01 6.62094891e-01 5.01027763e-01 -2.97846884e-01
-2.16601282e-01 -5.94854206e-02 -3.49337339e-01 3.00996602e-01
7.06358433e-01 4.98042464e-01 -1.24925137e+00 -3.46235335e-01
-2.62919962e-01 1.32771209e-02 1.03921354e+00 1.11180198e+00
2.04406261e-01 -4.98237014e-01 -6.29536659e-02 6.35352850e-01
1.80224419e+00 4.43755209e-01 7.76313603e-01 7.21356332e-01
2.79885978e-01 7.09609210e-01 2.43983865e-02 7.74601817e-01
4.43900526e-01 1.30526289e-01 7.09520578e-01 -3.81524950e-01
4.32719827e-01 3.35210621e-01 5.29221296e-01 1.14324629e-01
-3.30632538e-01 -1.67661026e-01 -1.03669894e+00 9.06721056e-01
-1.63111126e+00 -1.23476756e+00 -6.95893407e-01 2.57565403e+00
6.72497690e-01 -2.02075884e-01 -4.34865318e-02 -2.26819575e-01
4.40639734e-01 2.82758892e-01 -3.76044542e-01 -6.03426039e-01
-3.39351058e-01 2.89697886e-01 1.15308535e+00 8.72337699e-01
-9.39413548e-01 4.58033890e-01 6.84816694e+00 -1.80870041e-01
-1.29694736e+00 -1.98714510e-01 3.73823404e-01 -1.48649216e-01
-3.69373947e-01 1.08725667e-01 -9.84251022e-01 3.70809942e-01
1.78499651e+00 6.51771277e-02 2.96644002e-01 4.38653857e-01
1.48689103e+00 -4.03879046e-01 -4.96937513e-01 -1.69662014e-01
-6.47732437e-01 -9.26480412e-01 -1.48558751e-01 2.47474909e-01
7.47211576e-01 5.79508245e-01 -3.12321037e-01 1.79509401e-01
1.03265572e+00 -1.05678570e+00 1.32515430e-01 7.99044132e-01
4.80974168e-01 -4.26598400e-01 1.10130608e+00 2.49902219e-01
-1.20920336e+00 -1.43145189e-01 -6.39489830e-01 -8.21943581e-01
2.20734760e-01 7.80059218e-01 -5.22439539e-01 2.60819525e-01
8.28643382e-01 8.00996482e-01 1.23389512e-02 1.36735797e+00
1.11047521e-01 9.56840694e-01 -1.01642334e+00 -9.34415683e-02
6.92609906e-01 -5.64604104e-01 4.10618424e-01 1.43203521e+00
6.50715768e-01 3.78603786e-01 -1.09085344e-01 5.85482478e-01
7.44510353e-01 -1.00326262e-01 -7.08061218e-01 2.60541886e-01
7.37841010e-01 9.63096261e-01 -2.07284227e-01 -3.12074214e-01
-6.44133985e-01 1.35676116e-02 -3.74419779e-01 6.26730382e-01
-4.94437546e-01 -2.37819925e-01 1.46229029e+00 3.25806737e-02
1.47273034e-01 -2.53262132e-01 -5.08534610e-01 -5.10769784e-01
-2.40022793e-01 -4.05877888e-01 3.44192296e-01 -7.74119794e-01
-1.24859118e+00 3.74915868e-01 2.39727721e-02 -1.36412013e+00
-1.88440591e-01 -6.36872172e-01 -9.85828757e-01 1.74668264e+00
-2.03478122e+00 -8.20340276e-01 -5.90537012e-01 -1.04824612e-02
1.07180834e-01 4.75871146e-01 1.21069407e+00 2.49391235e-03
-3.88605386e-01 -5.59383035e-02 1.06586063e+00 -9.37922001e-02
3.78332913e-01 -1.19992280e+00 2.17897505e-01 6.71710134e-01
-5.99491358e-01 5.63596785e-01 9.72838461e-01 -8.03706288e-01
-1.17817497e+00 -1.51126003e+00 7.68754303e-01 1.58235449e-02
8.61888528e-01 4.19749200e-01 -1.16081941e+00 6.82766497e-01
-9.88513138e-03 -2.47123122e-01 7.26442695e-01 1.56941816e-01
1.94514960e-01 -4.54420120e-01 -1.09492731e+00 3.17504674e-01
1.55206114e-01 -2.02244610e-01 -2.54843950e-01 1.90423355e-01
2.85200506e-01 -1.99955970e-01 -1.20746720e+00 5.77069044e-01
9.47327077e-01 -7.52046108e-01 5.36949337e-01 -5.20569563e-01
2.60623187e-01 -2.99259812e-01 -2.20136613e-01 -1.65787899e+00
-5.22455573e-01 -1.65803477e-01 4.73350376e-01 1.19566000e+00
7.97932744e-01 -1.08222461e+00 4.71440434e-01 1.15230000e+00
-1.48334295e-01 -1.16474830e-01 -9.39018667e-01 -8.76052797e-01
5.64911664e-01 -4.84230816e-01 6.99698746e-01 5.71585774e-01
-2.28878602e-01 -2.57532179e-01 -5.69979429e-01 8.91781569e-01
5.35262048e-01 3.24726813e-02 6.86858118e-01 -1.71985972e+00
1.02195283e-02 -4.18365180e-01 -1.57336205e-01 -5.99347711e-01
-1.18869200e-01 -2.48362303e-01 4.38423753e-01 -1.92459011e+00
-2.78937757e-01 -4.51261580e-01 -3.14555407e-01 8.51925910e-01
-4.46240693e-01 -1.28933713e-01 -2.21203238e-01 -7.55954087e-02
6.24492526e-01 6.45045578e-01 7.70457685e-01 -8.82230252e-02
-1.05211389e+00 1.68474838e-01 -4.51463252e-01 4.24098164e-01
8.46532285e-01 -5.62886477e-01 3.29109058e-02 -7.59653926e-01
-3.47354412e-02 3.01694244e-01 3.13846350e-01 -8.04557979e-01
2.54823975e-02 -8.62880945e-01 5.50219417e-01 -4.66832936e-01
4.98042077e-01 -9.79751825e-01 7.03932762e-01 6.22881174e-01
7.13205412e-02 4.14779522e-02 9.98745024e-01 3.14013332e-01
1.75772563e-01 -2.13919476e-01 7.85014987e-01 -2.75005758e-01
-5.19049764e-01 2.36044541e-01 -1.10819650e+00 -4.76909310e-01
6.89340174e-01 -3.92400742e-01 -7.37677515e-01 -5.11333704e-01
-6.21025205e-01 5.22205651e-01 3.00419718e-01 1.85599297e-01
2.87293226e-01 -5.84084034e-01 -1.23237944e+00 1.86890483e-01
-2.09036797e-01 -1.27916530e-01 2.73956776e-01 6.46860301e-01
-6.76852882e-01 2.82750696e-01 -3.06702647e-02 -6.51153862e-01
-8.07331741e-01 -1.53992891e-01 8.76622856e-01 9.19425339e-02
-6.95195198e-01 6.35034800e-01 -1.81725264e-01 -5.35289705e-01
-1.85667589e-01 -2.27929547e-01 -3.18655789e-01 4.08774436e-01
7.63281047e-01 4.89178181e-01 1.42038625e-03 -5.25399268e-01
-1.47471905e-01 5.37658393e-01 6.15318179e-01 -1.28524512e-01
1.83717978e+00 -4.78518099e-01 -2.03863233e-01 9.59115326e-01
8.76956463e-01 -4.89577055e-01 -1.81584620e+00 -1.26768216e-01
-1.63943842e-01 -4.35194045e-01 5.00985920e-01 -8.41697693e-01
-9.23756719e-01 7.33304560e-01 7.81508207e-01 3.99504870e-01
1.00716710e+00 -7.91357756e-01 1.81229666e-01 4.18392032e-01
-4.50726459e-03 -6.39816284e-01 -1.02299452e+00 6.37453854e-01
4.43996280e-01 -1.06787658e+00 6.86874613e-02 4.81692106e-01
-2.21837282e-01 1.20856416e+00 3.71008605e-01 6.37035295e-02
1.01111877e+00 4.48814005e-01 6.99301779e-01 -3.51414308e-02
-9.02985692e-01 -5.30906796e-01 -4.22273308e-01 6.87584221e-01
4.63029742e-01 3.41765881e-01 -2.16345817e-01 2.81282868e-02
-4.24894989e-02 -2.12061889e-02 7.54158199e-01 7.52719402e-01
-7.69271791e-01 -6.81125343e-01 -8.07378471e-01 9.99950111e-01
-2.87283540e-01 -4.29326296e-01 1.35140821e-01 6.26076341e-01
-2.84336001e-01 1.46416438e+00 3.86501968e-01 -4.80608866e-02
2.53248960e-01 5.50525188e-02 -3.61973017e-01 -4.15557742e-01
-7.10382223e-01 1.85694516e-01 1.65773764e-01 -4.72853780e-02
-7.71142602e-01 -9.47976172e-01 -6.57721460e-01 -6.26947701e-01
-3.64230424e-01 4.15994614e-01 7.64039516e-01 9.55844820e-01
1.36954248e-01 1.89082906e-01 8.81232083e-01 -1.46069634e+00
-5.20001173e-01 -1.59839857e+00 -1.04694450e+00 -1.66247740e-01
7.46372700e-01 -4.31314528e-01 -1.08597970e+00 -1.87751710e-01] | [6.476065158843994, 3.00111985206604] |
542f75f1-d6c3-4699-81b1-2876b0b3f6b8 | what-does-dall-e-2-know-about-radiology | 2209.13696 | null | https://arxiv.org/abs/2209.13696v1 | https://arxiv.org/pdf/2209.13696v1.pdf | What Does DALL-E 2 Know About Radiology? | Generative models such as DALL-E 2 could represent a promising future tool for image generation, augmentation, and manipulation for artificial intelligence research in radiology provided that these models have sufficient medical domain knowledge. Here we show that DALL-E 2 has learned relevant representations of X-ray images with promising capabilities in terms of zero-shot text-to-image generation of new images, continuation of an image beyond its original boundaries, or removal of elements, while pathology generation or CT, MRI, and ultrasound images are still limited. The use of generative models for augmenting and generating radiological data thus seems feasible, even if further fine-tuning and adaptation of these models to the respective domain is required beforehand. | ['Keno K. Bressem', 'Hugo JWL. Aerts', 'Marcus R. Makowski', 'Daniel Truhn', 'Felix Busch', 'Lisa C. Adams'] | 2022-09-27 | null | null | null | null | ['zero-shot-text-to-image-generation'] | ['natural-language-processing'] | [ 8.19181025e-01 8.92770588e-01 3.52763444e-01 -2.76657850e-01
-8.50402951e-01 -2.84269065e-01 8.43526125e-01 1.21355951e-01
-2.34600559e-01 7.66702354e-01 3.12943429e-01 -6.45214736e-01
-2.80280411e-01 -1.05411112e+00 -6.59171581e-01 -6.81456625e-01
-1.76680565e-01 9.23758626e-01 4.96874340e-02 -2.27858007e-01
-1.44052878e-01 7.74576485e-01 -1.18816984e+00 4.00412410e-01
6.43755734e-01 5.39181530e-01 4.52719390e-01 1.00535989e+00
-7.45068714e-02 7.14351594e-01 -6.61459208e-01 -2.55057961e-01
1.25304356e-01 -5.91583550e-01 -7.17624068e-01 5.40173352e-01
9.42596644e-02 -3.76808167e-01 -3.45889807e-01 5.98384917e-01
7.50898600e-01 1.19370684e-01 9.88840401e-01 -7.05622673e-01
-8.29437673e-01 7.13566899e-01 -3.04083169e-01 1.80234894e-01
2.86906332e-01 1.43250912e-01 -8.95064995e-02 -6.97633624e-01
1.01492286e+00 8.34725559e-01 3.88443917e-01 1.02444279e+00
-1.21394753e+00 -2.23354965e-01 -3.17358315e-01 -1.59837887e-01
-9.81704414e-01 -1.96016893e-01 6.63573265e-01 -4.05891567e-01
7.30797946e-01 6.05284691e-01 7.81355202e-01 1.12490904e+00
4.91938651e-01 6.32066190e-01 1.05663741e+00 -7.40327895e-01
3.12524945e-01 1.52436092e-01 -3.37115347e-01 8.52067828e-01
1.20320477e-01 1.83302507e-01 -2.25185454e-01 -2.41235644e-01
1.23537600e+00 -9.68541577e-02 -1.31420687e-01 -2.22496957e-01
-1.19177914e+00 7.94041932e-01 5.24034619e-01 5.99162579e-01
-8.35672140e-01 2.88883179e-01 3.04153234e-01 -1.02478065e-01
4.26289588e-01 9.96725976e-01 -3.73760005e-03 1.68366894e-01
-1.02943921e+00 2.58551061e-01 2.97129780e-01 1.08940685e+00
1.76326066e-01 3.80239815e-01 -4.03740972e-01 6.20557010e-01
-5.08189201e-02 3.02533895e-01 9.54681277e-01 -7.75426149e-01
5.64831831e-02 5.43721318e-01 -1.32934496e-01 -3.47801834e-01
-6.09301329e-01 -7.37122059e-01 -8.72027874e-01 9.28192064e-02
-1.58868864e-01 -3.39580566e-01 -1.70322001e+00 1.34738743e+00
2.74155945e-01 1.39798105e-01 1.85029373e-01 6.09283030e-01
1.14121616e+00 5.22569060e-01 2.96173066e-01 -2.31927946e-01
1.52412939e+00 -6.77090049e-01 -7.87709832e-01 -4.54468846e-01
6.38289213e-01 -7.42701888e-01 7.76099980e-01 2.00885743e-01
-1.63007867e+00 -6.22859418e-01 -7.87317276e-01 1.01966612e-01
-1.39896914e-01 9.92605612e-02 8.46637487e-01 7.73967624e-01
-1.08459604e+00 2.91283906e-01 -9.61389422e-01 -2.30502814e-01
6.28339112e-01 3.58167052e-01 -4.46466744e-01 -1.94952384e-01
-9.98708189e-01 9.60527539e-01 4.31591064e-01 -7.46582523e-02
-9.19327319e-01 -7.97554910e-01 -9.48617697e-01 -6.74173981e-02
2.53192484e-01 -1.28633177e+00 1.22418332e+00 -4.48595643e-01
-1.18954468e+00 1.04807949e+00 2.88533688e-01 -5.50734520e-01
5.49192607e-01 1.43847659e-01 -3.62104982e-01 3.02232742e-01
-4.09642048e-02 1.16842568e+00 8.67243886e-01 -1.39104080e+00
-3.74220371e-01 -4.22986716e-01 1.38857201e-01 2.39908591e-01
-8.04452077e-02 -3.17053199e-01 -4.08128649e-01 -1.02199447e+00
2.36455500e-01 -1.02608037e+00 -9.38270271e-01 -2.09129393e-01
-4.58580494e-01 3.14551234e-01 5.78223288e-01 -6.58552289e-01
1.01862466e+00 -1.81265247e+00 3.32140885e-02 2.95824528e-01
1.98790446e-01 2.19096035e-01 -2.03011885e-01 3.58200014e-01
-3.69434327e-01 2.51345634e-01 -4.33376819e-01 8.00426453e-02
-5.21345496e-01 2.88923502e-01 4.11132537e-02 1.05672725e-01
2.20818892e-01 1.16993225e+00 -7.63547719e-01 -7.08225608e-01
4.61467832e-01 5.31463206e-01 -5.89781821e-01 1.78972036e-01
-2.08294600e-01 9.49991941e-01 -4.99728143e-01 4.09713060e-01
4.49830174e-01 -3.16943437e-01 2.04827785e-01 -1.24081738e-01
2.49910936e-01 -9.29452181e-02 -8.93862486e-01 1.87868655e+00
-6.25319004e-01 3.06442708e-01 -3.93948436e-01 -7.76867509e-01
6.91710353e-01 5.24033546e-01 6.81042731e-01 -6.01376712e-01
1.27003476e-01 -4.69342880e-02 2.46411636e-01 -7.27333128e-01
5.04357278e-01 -6.10785544e-01 9.69917998e-02 6.03714287e-01
-3.69239375e-02 -6.74979687e-01 5.37334681e-01 3.69740576e-01
1.13476825e+00 8.34634379e-02 2.20474318e-01 1.03709497e-01
2.96080619e-01 1.38801455e-01 -3.30667585e-01 1.08254504e+00
3.94849896e-01 9.53522682e-01 9.53576863e-02 -1.67517751e-01
-1.26565862e+00 -1.00627613e+00 -3.78492355e-01 7.50816941e-01
-3.13634872e-01 2.09058635e-02 -7.55153775e-01 -7.03603566e-01
-5.30178666e-01 1.10067129e+00 -8.52632523e-01 -4.07558143e-01
-7.90487945e-01 -1.05474007e+00 2.05577135e-01 6.62406862e-01
1.45155052e-02 -1.43526852e+00 -9.65440810e-01 6.25706136e-01
-1.87387168e-01 -9.64271069e-01 -1.42014414e-01 2.76051253e-01
-1.31445575e+00 -9.58504200e-01 -1.27888000e+00 -8.69431436e-01
1.27378178e+00 -4.96306717e-02 1.02092755e+00 1.55521005e-01
-9.25763071e-01 5.75336337e-01 -4.58530784e-01 -4.89817083e-01
-1.13959455e+00 -1.33544415e-01 -6.83966875e-01 -5.73294282e-01
-5.10305822e-01 -3.33410054e-01 -7.51207948e-01 -3.42064500e-02
-1.59721363e+00 5.45615673e-01 7.36743808e-01 1.08182299e+00
6.95400238e-01 -1.06984504e-01 3.84243011e-01 -1.32928288e+00
8.62696528e-01 -4.18056756e-01 -5.90054393e-02 1.92395449e-01
-3.82441491e-01 2.15188697e-01 1.94828048e-01 -4.99902159e-01
-1.37998319e+00 5.83174601e-02 -4.49021965e-01 -3.18948835e-01
-3.29177111e-01 6.93435371e-01 4.80575889e-01 8.10282156e-02
1.27135432e+00 4.03866380e-01 4.03313246e-03 -2.94117332e-01
6.23462081e-01 3.13634127e-01 7.44352818e-01 -2.50071228e-01
4.92230147e-01 5.80376863e-01 1.80727154e-01 -5.52739561e-01
-4.10833687e-01 -1.18340068e-01 -5.55726111e-01 -1.23132125e-01
8.45755994e-01 -4.42067087e-01 1.86018664e-02 -3.49817723e-02
-8.01975429e-01 -1.56587422e-01 -8.80057395e-01 4.96060967e-01
-8.44160914e-01 2.47337997e-01 -6.02128148e-01 -5.90263069e-01
-5.39283693e-01 -1.45695877e+00 1.16779518e+00 1.81909308e-01
-3.59140754e-01 -1.02430940e+00 -1.00946426e-01 3.57791811e-01
3.19755256e-01 4.34593052e-01 1.33630824e+00 -5.67493081e-01
-4.71948713e-01 -6.38281047e-01 1.73661321e-01 1.92560226e-01
3.12699556e-01 -2.40581438e-01 -5.77271342e-01 -1.63050652e-01
1.45975426e-01 -9.38433483e-02 5.05182147e-01 7.39215851e-01
1.44635010e+00 -2.33221397e-01 -4.78096336e-01 3.96838248e-01
1.10354435e+00 7.00908601e-01 8.62271011e-01 2.09550843e-01
2.27554455e-01 4.49047506e-01 3.17015857e-01 5.24578333e-01
3.21294516e-02 5.40881932e-01 3.05257708e-01 -5.10163724e-01
-6.53064251e-01 -2.80974805e-02 -4.13199723e-01 3.84017885e-01
-2.49577522e-01 -2.39720792e-01 -9.96670365e-01 5.52273989e-01
-1.28077149e+00 -9.35738206e-01 -8.01064149e-02 1.88267076e+00
8.17903578e-01 7.91153610e-02 -2.68575817e-01 1.61884755e-01
6.20568573e-01 -3.21934253e-01 -4.26164120e-01 -2.11021096e-01
1.92679584e-01 7.42335379e-01 2.44761318e-01 1.64775372e-01
-7.85030663e-01 5.13738155e-01 7.52042627e+00 6.97555304e-01
-1.00505924e+00 1.64727211e-01 9.20962512e-01 6.72920123e-02
-4.85900879e-01 -1.36549085e-01 -2.63436556e-01 2.95913070e-01
7.25111067e-01 -4.88260128e-02 -7.49416742e-03 7.49804139e-01
1.36937529e-01 -3.43819588e-01 -9.57624078e-01 6.81024671e-01
1.65326849e-01 -1.86041796e+00 2.80809760e-01 4.88098636e-02
8.12889159e-01 -4.83797014e-01 1.64120927e-01 3.24953943e-01
2.85794139e-01 -9.56607878e-01 2.97669202e-01 7.54176021e-01
9.96016681e-01 -6.11279130e-01 7.45506227e-01 3.99765015e-01
-4.15140569e-01 2.35856205e-01 -1.57502443e-01 7.69005299e-01
3.88825357e-01 3.19233090e-01 -1.59281301e+00 7.03022420e-01
7.72221982e-02 -1.80636823e-01 -7.76135683e-01 1.07285130e+00
-9.88562256e-02 4.01997685e-01 -1.06319539e-01 4.21199560e-01
1.83002040e-01 1.77043021e-01 5.12336910e-01 1.12328768e+00
6.62488937e-01 3.21380377e-01 4.64285091e-02 7.50518918e-01
7.05450848e-02 1.28286272e-01 -6.16207063e-01 1.21824974e-02
1.09407745e-01 1.21429598e+00 -1.06905282e+00 -8.53476107e-01
-3.23787257e-02 8.65257144e-01 -3.05615008e-01 2.82028079e-01
-7.95571864e-01 -2.43167672e-03 -3.18296582e-01 5.88341594e-01
5.42846583e-02 1.21110648e-01 -4.45953518e-01 -7.20710337e-01
-3.19831014e-01 -8.35961342e-01 3.25455785e-01 -1.36518157e+00
-8.76383722e-01 1.02846408e+00 3.87751788e-01 -1.14331114e+00
-1.04198229e+00 -3.11505198e-01 -4.63040292e-01 6.25128746e-01
-9.90298748e-01 -1.56191456e+00 -3.75234127e-01 4.66307521e-01
7.32497394e-01 -3.60000938e-01 1.03575754e+00 8.93175676e-02
1.30381644e-01 2.98081845e-01 -6.27952814e-02 -1.10544041e-01
1.92934826e-01 -1.22163689e+00 3.36757630e-01 4.91282105e-01
2.29826957e-01 3.47470105e-01 8.79376054e-01 -9.39112782e-01
-9.90568876e-01 -9.78577077e-01 4.10474300e-01 -2.79527694e-01
1.93075895e-01 8.14709365e-02 -7.38733351e-01 5.23637235e-01
9.40821618e-02 -1.11787312e-01 6.74446702e-01 -3.39038283e-01
5.38734376e-01 3.96522939e-01 -1.30926085e+00 7.44709671e-01
8.48755598e-01 -3.98702249e-02 -5.18108487e-01 7.87049532e-01
4.60583627e-01 -9.14871514e-01 -9.42910910e-01 5.08990645e-01
1.44566253e-01 -6.62063897e-01 1.28690267e+00 -7.08416104e-01
6.55526698e-01 2.17629969e-01 2.54472405e-01 -1.33216846e+00
-2.39627540e-01 -3.64064544e-01 6.86301067e-02 6.00375533e-01
3.54871511e-01 -2.35546619e-01 9.46985126e-01 6.94073856e-01
-5.45979142e-01 -7.74700224e-01 -6.96902156e-01 -4.62479949e-01
-3.05188224e-02 -5.40431976e-01 6.24879062e-01 7.27258801e-01
-2.64492720e-01 -1.25061804e-02 -2.12624103e-01 -1.62032589e-01
2.53516197e-01 8.56513344e-03 6.80786014e-01 -8.66941631e-01
-3.10024083e-01 -3.22561264e-01 -4.01057690e-01 -4.71420497e-01
-1.97803229e-01 -1.04930341e+00 -9.94325876e-02 -2.08270144e+00
2.32498437e-01 -6.11489058e-01 1.51063666e-01 4.89876270e-01
-2.64555007e-01 3.91871125e-01 2.55489439e-01 1.90552294e-01
1.28499895e-01 3.48099351e-01 2.11835504e+00 -1.74472779e-01
-2.03088373e-01 2.56861150e-01 -7.44526565e-01 7.13636816e-01
5.52806973e-01 -5.42057633e-01 -6.02383316e-01 -6.33132532e-02
-2.94910725e-02 7.16352344e-01 3.56824785e-01 -9.29555058e-01
-8.48003253e-02 -1.94035061e-02 7.18653023e-01 -4.44692671e-01
3.89550000e-01 -6.81719601e-01 5.69687128e-01 6.62218750e-01
-4.68548179e-01 1.93995424e-02 2.84397304e-01 3.70248079e-01
-1.26121625e-01 -8.36273491e-01 7.09423006e-01 -8.48870218e-01
-4.83375698e-01 1.11835077e-01 -7.16969967e-01 -1.77711904e-01
1.21923649e+00 -4.76376355e-01 1.95454471e-02 -4.03675109e-01
-1.52286172e+00 -3.88741851e-01 1.38469443e-01 3.28478426e-01
9.54924881e-01 -1.15256834e+00 -1.08784080e+00 3.09475183e-01
1.29754478e-02 1.51287675e-01 7.11383760e-01 5.41417181e-01
-7.53018618e-01 4.28005904e-01 -3.41475248e-01 -4.39815313e-01
-1.13699305e+00 5.79430044e-01 2.04378441e-01 -7.90095031e-01
-9.14932013e-01 6.17001414e-01 6.43860102e-01 -1.33360371e-01
-2.20895603e-01 -6.94218054e-02 -2.24544436e-01 -1.67513490e-01
2.60280669e-01 7.83653557e-02 4.02793556e-01 -3.47234517e-01
2.66065449e-02 -7.71528780e-02 -2.87845522e-01 -3.91821623e-01
1.34112406e+00 7.82797784e-02 3.11650842e-01 -2.44645089e-01
4.02779967e-01 -3.19901586e-01 -7.72575319e-01 -1.33147677e-02
-3.26544493e-01 -2.96181917e-01 1.68165326e-01 -1.17235744e+00
-1.08535337e+00 8.13852847e-01 7.55838454e-01 6.14832668e-03
1.28866172e+00 3.19215536e-01 4.46207941e-01 1.05492592e-01
2.26945326e-01 -6.84592903e-01 4.64188963e-01 -6.23253509e-02
1.44811988e+00 -8.49894285e-01 2.72775054e-01 -2.86512673e-01
-9.29304719e-01 9.75476086e-01 5.34870505e-01 1.01175681e-01
4.51805860e-01 4.83015239e-01 -4.09824327e-02 -4.74413991e-01
-6.40551746e-01 -1.93965092e-01 5.63786209e-01 1.02474403e+00
6.76134169e-01 1.10280111e-01 -4.35648471e-01 8.40660185e-02
-3.05721432e-01 1.80291325e-01 8.31622422e-01 1.21458578e+00
-3.16902846e-02 -1.23860109e+00 -6.61762416e-01 7.34921277e-01
-4.97876316e-01 -3.10313821e-01 1.25359282e-01 1.03938913e+00
1.38633221e-01 3.92642796e-01 4.12313528e-02 2.82066792e-01
3.59081775e-01 1.23621464e-01 1.00188565e+00 -1.12119627e+00
-7.04483747e-01 5.61321020e-01 1.90066874e-01 3.26003619e-02
-3.05400789e-01 -5.06127775e-01 -1.34435797e+00 3.41611654e-01
-4.12782162e-01 -1.27079949e-01 1.03901577e+00 7.65685797e-01
3.16842049e-01 1.13648200e+00 2.37661272e-01 -7.74653554e-01
-2.27238655e-01 -1.20334935e+00 -3.00314784e-01 4.95473862e-01
-1.37243062e-01 -4.95565206e-01 3.22053701e-01 6.17631435e-01] | [14.250467300415039, -1.9077166318893433] |
79ce5732-1f20-4fa5-969d-e5e4a5b4cd2f | domain-generalization-via-ensemble-stacking | 2301.02145 | null | https://arxiv.org/abs/2301.02145v1 | https://arxiv.org/pdf/2301.02145v1.pdf | Domain Generalization via Ensemble Stacking for Face Presentation Attack Detection | Face presentation attack detection (PAD) plays a pivotal role in securing face recognition systems against spoofing attacks. Although great progress has been made in designing face PAD methods, developing a model that can generalize well to an unseen test domain remains a significant challenge. Moreover, due to different types of spoofing attacks, creating a dataset with a sufficient number of samples for training deep neural networks is a laborious task. This work addresses these challenges by creating synthetic data and introducing a deep learning-based unified framework for improving the generalization ability of the face PAD. In particular, synthetic data is generated by proposing a video distillation technique that blends a spatiotemporal warped image with a still image based on alpha compositing. Since the proposed synthetic samples can be generated by increasing different alpha weights, we train multiple classifiers by taking the advantage of a specific type of ensemble learning known as a stacked ensemble, where each such classifier becomes an expert in its own domain but a non-expert to others. Motivated by this, a meta-classifier is employed to learn from these experts collaboratively so that when developing an ensemble, they can leverage complementary information from each other to better tackle or be more useful for an unseen target domain. Experimental results using half total error rates (HTERs) on four PAD databases CASIA-MFSD (6.97 %), Replay-Attack (33.49%), MSU-MFSD (4.02%), and OULU-NPU (10.91%)) demonstrate the robustness of the method and open up new possibilities for advancing presentation attack detection using ensemble learning with large-scale synthetic data. | ['Mourad Oussalah', 'Djamila Romaissa Beddiar', 'Usman Muhammad'] | 2023-01-05 | null | null | null | null | ['face-presentation-attack-detection'] | ['computer-vision'] | [ 4.82301921e-01 -2.61891633e-01 5.07021472e-02 -1.58923060e-01
-6.14409685e-01 -5.88542879e-01 6.00435495e-01 -3.55833113e-01
1.12820916e-01 6.52643383e-01 -1.68333620e-01 -3.07926029e-01
6.57218248e-02 -9.22990382e-01 -7.01589763e-01 -1.00133324e+00
-3.02192360e-01 2.36023352e-01 1.34969011e-01 -4.75461155e-01
-2.24573468e-03 7.92113006e-01 -1.51156068e+00 6.89485610e-01
8.27856779e-01 1.08501720e+00 -1.76341191e-01 4.58325118e-01
1.70870364e-01 4.31818485e-01 -1.22507751e+00 -8.71417046e-01
3.26043308e-01 -5.06247938e-01 -3.14037114e-01 -3.15745808e-02
4.56418365e-01 -3.51510227e-01 -3.23817074e-01 9.61364388e-01
7.09115028e-01 -1.05590187e-01 6.09644830e-01 -1.53656065e+00
-4.70250428e-01 3.56354952e-01 -5.57413876e-01 1.60198092e-01
4.30702358e-01 2.11958691e-01 1.49466977e-01 -7.43954241e-01
5.00149786e-01 1.40206933e+00 5.24524748e-01 9.35654938e-01
-1.07990992e+00 -1.32511818e+00 -1.93355545e-01 3.19907427e-01
-1.17253065e+00 -5.63852370e-01 1.04002941e+00 -1.83985949e-01
3.01635027e-01 1.57979488e-01 1.72957271e-01 1.95462465e+00
5.31948209e-02 5.30339837e-01 1.31138039e+00 -2.40253642e-01
1.08999915e-01 4.51824337e-01 -2.45196477e-01 3.59644711e-01
3.81343305e-01 2.22246185e-01 -5.18873453e-01 -2.82541662e-01
4.31720257e-01 -1.12060651e-01 -4.51977104e-01 -3.24922204e-02
-6.43074214e-01 7.62166798e-01 3.93056810e-01 4.26693738e-01
-2.81006634e-01 -4.21463460e-01 4.15116221e-01 3.85697275e-01
3.70771676e-01 3.97517592e-01 -2.09746093e-01 3.19230676e-01
-9.19257998e-01 1.90762028e-01 8.55760336e-01 3.15719604e-01
5.21639824e-01 4.41716641e-01 1.06064999e-03 7.03695953e-01
4.72384989e-02 6.86247170e-01 5.03805161e-01 -4.52805698e-01
6.90425277e-01 3.42618436e-01 -1.00890296e-02 -1.35428810e+00
-8.81818356e-04 -6.89768255e-01 -9.11470413e-01 4.65282351e-01
4.28344548e-01 -3.71479988e-01 -8.75337422e-01 1.78231692e+00
4.01348174e-01 7.30183840e-01 2.81521678e-01 4.78026360e-01
4.73743021e-01 7.43785799e-01 5.01575880e-03 -2.98515409e-01
1.28339314e+00 -5.38795531e-01 -5.58859527e-01 -4.05952567e-03
2.77042121e-01 -7.10074842e-01 5.59031308e-01 7.26218641e-01
-6.87170565e-01 -7.67115355e-01 -1.34309447e+00 7.19951034e-01
-5.46610236e-01 -1.03501603e-01 1.06544718e-01 1.32032955e+00
-9.14828956e-01 3.76025409e-01 -2.37649485e-01 1.74652581e-04
7.61268854e-01 4.93961275e-01 -6.74234927e-01 -1.34685233e-01
-1.45863771e+00 6.30811632e-01 4.56461072e-01 1.10241054e-02
-1.23439074e+00 -6.20261610e-01 -4.38341916e-01 -4.53808904e-03
2.89336264e-01 -2.74816841e-01 6.96280122e-01 -1.14904952e+00
-1.43404901e+00 5.58262944e-01 1.00508571e-01 -5.80712318e-01
6.44046247e-01 -2.99699139e-02 -1.08839357e+00 3.77634734e-01
-1.58507377e-01 1.68764040e-01 1.62787747e+00 -1.51998806e+00
-3.04033637e-01 -6.22699976e-01 -1.03907377e-01 -2.94667989e-01
-8.48819733e-01 1.93367481e-01 3.66861112e-02 -8.64566445e-01
-4.31155056e-01 -8.28180134e-01 3.64418209e-01 -2.52877176e-01
-2.03792542e-01 -8.22906345e-02 1.61209226e+00 -9.04680729e-01
1.24247670e+00 -2.27282023e+00 4.45717461e-02 3.69912297e-01
1.67849123e-01 1.24584830e+00 -3.23989362e-01 4.87536162e-01
-4.45628554e-01 1.98911563e-01 -2.93024093e-01 -3.12953383e-01
-4.63597417e-01 -7.26269558e-02 -5.88658571e-01 2.30539516e-01
5.41084528e-01 3.36614758e-01 -8.19844782e-01 -2.02770919e-01
1.29665628e-01 7.75652707e-01 -4.81538683e-01 3.15456778e-01
-6.50985688e-02 7.29886711e-01 -4.47171092e-01 6.56622350e-01
1.29440951e+00 1.29246920e-01 1.47835702e-01 -1.63312778e-01
3.37345302e-01 -2.90461421e-01 -1.28203785e+00 1.10760987e+00
-5.39477408e-01 4.62410897e-01 1.80729046e-01 -1.22233486e+00
1.20545554e+00 4.67982948e-01 3.62206042e-01 -4.31888312e-01
2.53468037e-01 2.72190392e-01 3.60136405e-02 -6.23755217e-01
-8.84238258e-02 -3.41485790e-03 1.64055586e-01 4.72911268e-01
4.28849578e-01 2.71296114e-01 -7.61469454e-03 -3.93144786e-02
1.09923351e+00 -1.81987613e-01 -1.16278104e-01 1.41518652e-01
9.82395411e-01 -6.01756394e-01 5.08372664e-01 6.17116332e-01
-3.89520764e-01 3.65800798e-01 4.04501706e-01 -5.38751006e-01
-8.43706965e-01 -1.16211939e+00 -2.30734169e-01 8.61246467e-01
-2.38034315e-02 -3.09469312e-01 -1.00274229e+00 -1.03969002e+00
2.70598714e-04 4.96215582e-01 -7.21952975e-01 -4.52541053e-01
-8.04552197e-01 -7.68041253e-01 1.10519958e+00 2.02271193e-01
1.01399827e+00 -1.05609226e+00 -2.58603364e-01 1.17636159e-01
-1.54506236e-01 -1.26015270e+00 3.58368456e-02 -2.67504722e-01
-3.97085309e-01 -1.15892589e+00 -7.54441142e-01 -7.01296091e-01
4.15080190e-01 3.51645827e-01 5.98082483e-01 9.58532989e-02
-2.07925200e-01 2.07520705e-02 -4.27082688e-01 -4.42382097e-01
-8.11406076e-01 -2.46084452e-01 4.12343144e-01 8.27365398e-01
2.98676223e-01 -5.82488120e-01 -3.75748038e-01 4.05749410e-01
-1.11671424e+00 -3.93578321e-01 4.69229579e-01 9.10166383e-01
-2.64696896e-01 3.33332449e-01 9.03210402e-01 -6.56341672e-01
6.12230957e-01 -8.57535601e-01 -3.87896031e-01 3.78287435e-01
-7.09322840e-02 -1.87141106e-01 7.84733355e-01 -9.25364971e-01
-1.33945632e+00 -2.25676268e-01 -1.42659828e-01 -7.27243304e-01
-2.69984573e-01 2.60731783e-02 -5.09410024e-01 -3.66056591e-01
9.32266355e-01 3.02385360e-01 3.79167020e-01 -2.92027771e-01
1.50960267e-01 9.18150842e-01 5.21747291e-01 -5.17134964e-01
1.27464938e+00 3.94676596e-01 6.83961213e-02 -1.08907473e+00
-2.26044342e-01 -3.09127681e-02 -2.86451966e-01 -4.13689941e-01
6.46922231e-01 -7.49319732e-01 -7.76266038e-01 1.01881754e+00
-1.12828231e+00 1.69086456e-01 4.14548963e-01 2.02602550e-01
-5.92342690e-02 5.98703563e-01 -4.13633347e-01 -8.86273205e-01
-2.69924134e-01 -1.31356740e+00 9.02019262e-01 4.38959956e-01
2.23025680e-02 -7.36743808e-01 -5.79720549e-02 5.84357977e-01
5.40495992e-01 6.15877628e-01 6.58587158e-01 -1.04524875e+00
-2.04321280e-01 -4.88607526e-01 -7.49976933e-02 8.83331001e-01
1.84370250e-01 1.81488901e-01 -1.30365551e+00 -5.21689415e-01
4.07106489e-01 -3.90082300e-01 6.85368955e-01 -1.68251157e-01
1.27082539e+00 -4.88244116e-01 -4.64549541e-01 5.25024116e-01
1.07574010e+00 5.66891313e-01 7.34715879e-01 1.00201569e-01
4.36649859e-01 7.83958375e-01 1.56353489e-01 3.86527866e-01
-1.64775208e-01 8.01872075e-01 4.99319524e-01 2.81664938e-01
-1.40577868e-01 -2.24506453e-01 7.35744894e-01 1.01462051e-01
4.74770442e-02 -5.09969950e-01 -7.60497689e-01 7.03338236e-02
-1.19269717e+00 -1.30086601e+00 2.77488947e-01 2.24928761e+00
4.67483461e-01 3.72552052e-02 2.27652356e-01 7.07615614e-01
1.00413382e+00 2.54248440e-01 -2.99296707e-01 -3.54072630e-01
-2.83281356e-01 6.28881454e-01 -5.77908084e-02 1.03803664e-01
-1.29835272e+00 7.86502302e-01 4.44668245e+00 1.10979605e+00
-1.49709415e+00 4.43809032e-02 6.90813720e-01 2.15678319e-01
1.44501269e-01 -4.11682189e-01 -8.89174104e-01 8.28955770e-01
1.03179657e+00 1.35991856e-01 4.00899380e-01 6.43515289e-01
-2.38821879e-01 3.27731520e-01 -6.07540786e-01 1.00172043e+00
4.63673383e-01 -1.23847210e+00 2.48883709e-01 2.15447396e-01
7.11946368e-01 -4.55814093e-01 6.03253245e-01 3.25182259e-01
3.64200845e-02 -1.01111925e+00 1.81081593e-01 1.17316879e-01
7.41062284e-01 -7.91097462e-01 6.82201803e-01 3.93771470e-01
-1.04836774e+00 -4.32830662e-01 -2.11130664e-01 4.45544243e-01
-8.65652189e-02 3.60296369e-01 -1.08721864e+00 8.66580069e-01
6.25253379e-01 3.20776463e-01 -5.47092080e-01 7.56513298e-01
-1.63054150e-02 5.36071479e-01 -3.05897236e-01 1.59245834e-01
4.80065634e-03 1.64877206e-01 7.24567890e-01 1.00123560e+00
5.78666091e-01 -1.36178464e-01 -9.76639241e-02 5.18856108e-01
-2.86629349e-01 -1.07389830e-01 -1.01011097e+00 -7.66306221e-02
6.69816375e-01 1.31557691e+00 -3.19413006e-01 -1.50454134e-01
-8.98884833e-02 8.38498592e-01 1.80395603e-01 4.00069594e-01
-9.03622925e-01 -3.33879322e-01 7.29987025e-01 -2.70449240e-02
3.19224000e-01 1.33805871e-01 1.05822444e-01 -1.16261029e+00
4.42847349e-02 -1.26753616e+00 3.81150097e-01 -4.38607693e-01
-1.22908580e+00 1.08536577e+00 -5.05196489e-02 -1.44723034e+00
-4.22444433e-01 -8.48930299e-01 -8.02398384e-01 8.84760022e-01
-1.19557405e+00 -1.17512941e+00 -3.60496849e-01 8.38184595e-01
3.85120064e-01 -8.25988948e-01 9.22626436e-01 3.02387595e-01
-8.03125203e-01 1.06733501e+00 -1.63097218e-01 4.04320747e-01
7.70928085e-01 -6.26719058e-01 2.00146690e-01 9.33891654e-01
1.41893804e-01 5.76821089e-01 5.47822416e-01 -6.11582339e-01
-1.29863691e+00 -1.10520077e+00 4.88168418e-01 -3.75360817e-01
4.73534465e-01 -3.82784069e-01 -1.08877432e+00 1.64114028e-01
2.04534292e-01 2.07163051e-01 9.05741990e-01 -3.82634014e-01
-8.85736048e-01 -4.73447174e-01 -1.61472023e+00 5.81335962e-01
7.96846092e-01 -6.04250848e-01 -3.20926011e-01 2.09747538e-01
3.48946750e-01 -4.67641912e-02 -7.45293856e-01 7.06374705e-01
6.18596613e-01 -1.37371731e+00 1.13092744e+00 -7.04406619e-01
2.29841352e-01 -1.41662017e-01 -6.43219650e-02 -1.42088270e+00
1.51618436e-01 -7.76258528e-01 -3.78386825e-01 1.43991208e+00
3.50344144e-02 -7.63770580e-01 7.99209177e-01 8.01963881e-02
1.32181883e-01 -6.01071239e-01 -1.03005481e+00 -8.61150384e-01
-2.41661724e-02 -2.92195082e-01 9.44242954e-01 1.20468760e+00
-2.28221074e-01 -1.09327674e-01 -5.70567846e-01 4.55967665e-01
8.41814518e-01 -5.50114453e-01 1.00456750e+00 -1.29782367e+00
-2.48059481e-01 -2.85101831e-01 -4.93513465e-01 -4.08540726e-01
3.92181545e-01 -7.69894123e-01 -5.76896012e-01 -4.89910364e-01
-2.54755944e-01 -3.04970890e-01 -3.13849121e-01 3.21871012e-01
-1.27617449e-01 4.67586756e-01 3.72218132e-01 -4.93983701e-02
5.20086884e-02 3.28607649e-01 1.05179858e+00 -1.60348818e-01
7.57883415e-02 2.41727695e-01 -5.29873073e-01 4.41696227e-01
8.08311284e-01 -3.25808197e-01 -5.07399499e-01 -7.40584731e-02
-4.48582768e-01 3.40820163e-01 3.93515617e-01 -1.49381232e+00
5.08552305e-02 2.08902851e-01 6.94746614e-01 -2.07529634e-01
5.29443324e-01 -8.27377439e-01 1.83170423e-01 5.81259906e-01
-5.78557723e-04 -1.81748588e-02 3.11621606e-01 4.58680749e-01
-2.42720470e-01 -1.14700407e-01 9.09038782e-01 1.76492602e-01
-5.23938656e-01 3.97262156e-01 -9.73002687e-02 -1.93360910e-01
1.37825549e+00 -4.10525858e-01 -5.22228301e-01 -4.00832176e-01
-6.35808408e-01 -3.17034632e-01 1.01614662e-01 6.79886580e-01
8.00845563e-01 -1.21307147e+00 -8.75835478e-01 8.97111237e-01
5.65837327e-05 -5.27086914e-01 4.90843326e-01 2.31279120e-01
-2.22231224e-01 1.35264009e-01 -6.34079576e-01 -5.11110783e-01
-1.34532976e+00 7.04598546e-01 2.39155889e-01 -1.99302539e-01
8.21765047e-03 9.00633335e-01 4.62429449e-02 -9.73474309e-02
1.01453014e-01 5.59744537e-01 -3.53219956e-01 2.57012188e-01
9.87973928e-01 4.71440703e-01 1.87551707e-01 -8.89400482e-01
-2.88914472e-01 3.46838385e-01 -2.08119258e-01 2.35679764e-02
1.08568907e+00 3.35603178e-01 -1.59698296e-02 -2.84464180e-01
1.45377731e+00 4.88568097e-02 -1.06041670e+00 -1.47789493e-01
-1.56862259e-01 -6.31976366e-01 -3.38296920e-01 -7.06596971e-01
-1.21950471e+00 9.97479439e-01 7.72929847e-01 4.14276838e-01
1.34937596e+00 -4.28791881e-01 7.06521511e-01 1.13719247e-01
4.34309423e-01 -5.47142267e-01 5.80279529e-01 1.78419590e-01
9.18153107e-01 -1.16574347e+00 -4.94388670e-01 -3.30171496e-01
-5.19756019e-01 1.29503274e+00 7.35305548e-01 -7.58565739e-02
7.32524872e-01 2.16778353e-01 1.09536527e-02 2.44944632e-01
-4.80258077e-01 2.74908453e-01 1.46534503e-01 1.11406016e+00
7.47508882e-03 -1.31033897e-01 1.37695715e-01 4.64391053e-01
-1.15379922e-01 -9.65821147e-02 2.82441914e-01 6.84139192e-01
-1.17847741e-01 -1.52925992e+00 -7.07803726e-01 2.23611444e-01
-5.24109602e-01 3.61770749e-01 -3.49265903e-01 6.58894241e-01
3.39650154e-01 1.16995084e+00 -2.36987337e-01 -1.00708854e+00
-4.94141579e-02 1.85354441e-01 4.19558555e-01 -2.79903352e-01
-6.74041986e-01 -2.81582415e-01 -6.41194656e-02 -1.97448507e-01
-2.24091217e-01 -4.28216279e-01 -4.89928871e-01 -5.05634248e-01
-1.91861212e-01 2.60814011e-01 6.49357617e-01 8.89141500e-01
3.43087018e-01 2.83221602e-01 1.29126120e+00 -9.50302541e-01
-8.34214747e-01 -8.65789056e-01 -3.00386369e-01 5.60137212e-01
4.67060179e-01 -1.00821662e+00 -5.83081901e-01 -1.66201249e-01] | [13.065364837646484, 1.1785669326782227] |
c8e79ab7-c668-4692-a464-1d6b47a51fce | learning-dynamic-and-personalized-comorbidity | 2001.02585 | null | https://arxiv.org/abs/2001.02585v2 | https://arxiv.org/pdf/2001.02585v2.pdf | Learning Dynamic and Personalized Comorbidity Networks from Event Data using Deep Diffusion Processes | Comorbid diseases co-occur and progress via complex temporal patterns that vary among individuals. In electronic health records we can observe the different diseases a patient has, but can only infer the temporal relationship between each co-morbid condition. Learning such temporal patterns from event data is crucial for understanding disease pathology and predicting prognoses. To this end, we develop deep diffusion processes (DDP) to model "dynamic comorbidity networks", i.e., the temporal relationships between comorbid disease onsets expressed through a dynamic graph. A DDP comprises events modelled as a multi-dimensional point process, with an intensity function parameterized by the edges of a dynamic weighted graph. The graph structure is modulated by a neural network that maps patient history to edge weights, enabling rich temporal representations for disease trajectories. The DDP parameters decouple into clinically meaningful components, which enables serving the dual purpose of accurate risk prediction and intelligible representation of disease pathology. We illustrate these features in experiments using cancer registry data. | ['Alexis Bellot', 'Mihaela van der Schaar', 'Ahmed M. Alaa', 'Zhaozhi Qian', 'Jem Rashbass'] | 2020-01-08 | null | null | null | null | ['disease-trajectory-forecasting'] | ['medical'] | [-7.42272586e-02 1.56277016e-01 -5.08974016e-01 -3.20937067e-01
-1.40466779e-01 -5.44116378e-01 7.86595225e-01 7.40821481e-01
7.48951174e-03 3.47239643e-01 9.68274117e-01 -2.84327507e-01
-6.57144487e-01 -1.10500729e+00 -3.15131068e-01 -5.93881547e-01
-9.74965811e-01 9.47493732e-01 -3.71722072e-01 1.33286536e-01
-5.31297088e-01 4.08768892e-01 -5.61251581e-01 1.87442407e-01
6.82509065e-01 5.01744628e-01 -1.07774369e-01 7.10795581e-01
-8.85536745e-02 9.42712724e-01 -3.27850103e-01 -4.50851917e-01
-2.83707052e-01 -1.50839850e-01 -4.39975798e-01 -1.36343583e-01
-3.67911488e-01 5.48289306e-02 -1.11491227e+00 7.84565985e-01
1.08848736e-01 -2.85056263e-01 1.11479640e+00 -1.13629162e+00
-8.99249852e-01 6.44082725e-01 -3.52616757e-01 4.05818760e-01
2.81198174e-01 3.97239566e-01 1.05724728e+00 -5.13744839e-02
8.52153480e-01 1.22907853e+00 8.29492807e-01 5.87580681e-01
-1.56496775e+00 -2.30264381e-01 3.32713127e-01 -3.86575721e-02
-1.15696955e+00 8.74495730e-02 7.08628953e-01 -8.97850215e-01
7.51763642e-01 3.23994040e-01 1.23499060e+00 1.61051607e+00
9.44207191e-01 5.04314661e-01 5.05564630e-01 5.40566683e-01
1.05129406e-01 -7.29001284e-01 3.66869628e-01 8.86456251e-01
1.40221059e-01 1.78446695e-01 -3.23666662e-01 -6.43462956e-01
8.57859910e-01 1.10254526e+00 -2.65768468e-01 -8.83932561e-02
-1.37049174e+00 6.00012839e-01 7.62993038e-01 2.42131978e-01
-7.96077967e-01 5.26893437e-01 2.91742951e-01 3.10788155e-01
7.20797956e-01 -1.03641994e-01 -4.60967541e-01 2.11686492e-02
-6.42295778e-01 6.50422648e-02 7.38188326e-01 3.97585630e-01
-2.74050683e-02 -3.57130647e-01 -4.92271811e-01 5.58055997e-01
5.14250457e-01 6.16394818e-01 3.17833513e-01 -3.42100918e-01
2.97667950e-01 9.41398799e-01 5.01831621e-02 -1.23271918e+00
-1.03252852e+00 -4.96481329e-01 -1.42007434e+00 -5.44402361e-01
3.70072156e-01 -3.33911479e-01 -9.20401037e-01 2.12429357e+00
2.64004380e-01 5.96584976e-01 -8.07297602e-02 4.21046853e-01
4.90146846e-01 8.71076465e-01 6.79333746e-01 -1.76814333e-01
1.57858050e+00 -1.53858840e-01 -9.99649823e-01 9.00468230e-02
7.97936618e-01 7.68914744e-02 5.04525661e-01 -4.35177647e-02
-1.03204119e+00 1.93584770e-01 -4.09608871e-01 1.35842487e-01
-4.58248496e-01 -2.80964881e-01 9.64194536e-01 7.33524561e-02
-9.95105386e-01 7.68969238e-01 -1.41100955e+00 -4.17539865e-01
7.43244469e-01 2.45554388e-01 -9.72132832e-02 -2.19258144e-01
-1.44215667e+00 6.08194351e-01 3.59320790e-02 1.31970212e-01
-9.48806703e-01 -1.32008886e+00 -7.14473188e-01 3.62687707e-01
-8.69378261e-03 -1.30026925e+00 7.57237792e-01 -5.67334116e-01
-8.45155954e-01 7.96511829e-01 -9.49048623e-02 -5.26487410e-01
4.86261696e-01 1.06579639e-01 -1.12690699e+00 -1.05029240e-01
-3.17453474e-01 1.48548126e-01 5.54337859e-01 -6.28746748e-01
-2.94618398e-01 -5.90407789e-01 -4.20983166e-01 2.51454711e-01
-3.42540681e-01 -2.13966563e-01 -2.88148642e-01 -9.43523169e-01
-2.86193937e-01 -8.75460565e-01 -5.67650199e-01 4.55712408e-01
-5.19758880e-01 -2.76882857e-01 5.70342958e-01 -7.66470611e-01
1.65480900e+00 -2.22891378e+00 5.16825080e-01 1.79737806e-01
9.56809282e-01 -3.69367987e-01 -1.74760461e-01 6.73951745e-01
-1.57848880e-01 1.30957618e-01 -4.13836956e-01 -2.05708385e-01
-2.41194591e-01 4.24107730e-01 -5.32093905e-02 6.75076842e-01
4.23715979e-01 1.36016881e+00 -1.40673554e+00 -1.21788643e-01
-2.65162230e-01 8.28524351e-01 -3.24424654e-01 3.75308394e-02
-4.32593942e-01 3.05398494e-01 -5.98859549e-01 4.71839100e-01
3.06871295e-01 -9.30301785e-01 5.07297754e-01 -9.79042798e-03
2.57785738e-01 3.04690778e-01 -6.80003762e-01 1.58465385e+00
-3.70953977e-01 2.88930386e-01 -1.27444178e-01 -8.54591310e-01
5.95113218e-01 6.65850878e-01 1.00824654e+00 -1.85850263e-01
-8.84622857e-02 -2.21559554e-01 2.30849385e-01 -4.64079618e-01
-5.47651462e-02 -1.92719206e-01 -1.98468596e-01 6.43837750e-01
-1.98095217e-01 4.33641464e-01 -5.90787195e-02 3.83492291e-01
1.76087105e+00 -6.68800175e-01 1.68868616e-01 -1.93018436e-01
-1.22051954e-01 -6.29330724e-02 7.10683763e-01 4.00997132e-01
-1.61933050e-01 1.92477986e-01 1.03357744e+00 -8.49971533e-01
-6.88185513e-01 -1.72119200e+00 -2.26615638e-01 3.37068826e-01
-1.69165835e-01 -2.15021506e-01 -1.52948257e-02 -6.62178159e-01
3.77821147e-01 2.46593386e-01 -1.01625991e+00 -5.38447320e-01
-3.05658668e-01 -1.55001402e+00 3.26771617e-01 4.68190461e-01
-2.53049284e-01 -9.71175432e-01 -1.10515133e-02 4.98162597e-01
-1.25314996e-01 -5.07028341e-01 -8.29258859e-01 1.23754470e-02
-1.11244345e+00 -1.25670469e+00 -7.98187554e-01 -6.26996994e-01
6.67615712e-01 -3.91597420e-01 1.34068978e+00 3.62438485e-02
-6.71306610e-01 6.09688818e-01 1.86345741e-01 -1.78774685e-01
-3.89715761e-01 -2.08065957e-01 8.02505165e-02 3.51221919e-01
2.70164967e-01 -8.93736660e-01 -1.22684503e+00 -1.92353651e-01
-1.12636852e+00 5.33891171e-02 4.78340119e-01 5.71615815e-01
7.04712451e-01 2.19548102e-02 4.90848631e-01 -1.12131560e+00
7.31800437e-01 -1.35571671e+00 -2.67434806e-01 4.31645513e-01
-5.29318392e-01 1.50452361e-01 3.63817394e-01 -7.32873619e-01
-9.12986815e-01 -3.98091637e-02 2.52875537e-01 -2.02629566e-01
-6.01701513e-02 8.78600478e-01 1.76629260e-01 6.75737560e-01
3.84673446e-01 2.74958581e-01 5.88106439e-02 -2.96044201e-01
6.51950836e-01 1.55784637e-01 2.95747519e-01 -1.84408337e-01
3.41998428e-01 7.19058871e-01 3.49173069e-01 -4.29393709e-01
-7.20043838e-01 -2.74155974e-01 -4.02568460e-01 -1.09419830e-01
1.04239321e+00 -9.89867091e-01 -8.94452929e-01 4.71209288e-01
-9.46643651e-01 -6.08452559e-01 -4.66657609e-01 4.00763690e-01
-2.19899476e-01 -2.00943112e-01 -1.25250971e+00 -3.54757398e-01
-2.20283940e-01 -4.93736744e-01 9.10309434e-01 -1.19106613e-01
-5.56589663e-01 -1.96558297e+00 6.94635212e-01 -4.38739836e-01
2.79325753e-01 6.93095028e-01 1.54902315e+00 -2.57830858e-01
-3.57611269e-01 -3.44665229e-01 -2.11088389e-01 -5.44844747e-01
5.11575162e-01 1.39749497e-01 -1.99249536e-01 -6.52932748e-02
-3.53791475e-01 4.27505583e-01 7.93430388e-01 9.53097582e-01
1.23570824e+00 -4.91480023e-01 -9.33926165e-01 9.06329632e-01
1.12188399e+00 1.75751835e-01 3.50056291e-01 -3.86672199e-01
9.36538756e-01 5.67968845e-01 -2.90963292e-01 4.46966439e-01
7.53360510e-01 3.88665587e-01 3.49667013e-01 -2.01370940e-01
-1.92750901e-01 -3.91127676e-01 1.00348532e-01 8.08347762e-01
1.99424401e-02 -4.89185929e-01 -1.35195994e+00 5.87880075e-01
-1.92210364e+00 -1.04224813e+00 -4.99238461e-01 1.96720052e+00
9.22145247e-01 -6.80014342e-02 2.23370194e-01 -4.81119722e-01
7.47686327e-01 4.47978795e-01 -8.90304208e-01 -3.79264019e-02
-1.03724301e-01 -2.15843603e-01 3.57350141e-01 4.30246651e-01
-8.65956128e-01 4.46801394e-01 6.25880003e+00 1.28157258e-01
-1.06102896e+00 3.14725935e-02 9.37577903e-01 -4.57201421e-01
-6.46409631e-01 -3.79887849e-01 -1.77789569e-01 6.60367250e-01
1.13288510e+00 -4.57725674e-01 3.73158067e-01 3.87526080e-02
3.89852226e-01 5.00482023e-01 -1.22507989e+00 8.36304307e-01
-6.55898452e-01 -1.42121947e+00 2.22843513e-01 2.98220962e-01
7.58160591e-01 3.11224848e-01 5.06057680e-01 -5.99507354e-02
1.16914332e+00 -9.82715964e-01 2.06513628e-02 1.29166543e+00
9.64719176e-01 -5.10702848e-01 1.05084985e-01 1.20517753e-01
-1.22857130e+00 -2.16447204e-01 8.41337517e-02 2.60546267e-01
6.40263140e-01 1.01597536e+00 -6.90534890e-01 4.97661561e-01
3.19061697e-01 1.52480304e+00 -1.65828839e-01 8.93366694e-01
3.31310965e-02 8.25303495e-01 -2.23813340e-01 2.15354294e-01
5.52171730e-02 -3.73135865e-01 4.71281677e-01 1.22992349e+00
5.10939777e-01 2.88019747e-01 -2.32122447e-02 1.05825615e+00
-2.99757063e-01 -3.14216673e-01 -6.85153246e-01 -5.33209145e-01
1.58757240e-01 1.24615490e+00 -6.81143820e-01 -3.03457975e-01
-4.81967092e-01 9.32273686e-01 4.83018219e-01 5.76929033e-01
-6.58983111e-01 2.84953415e-01 1.25288475e+00 3.37205797e-01
-3.01331818e-01 -2.69505411e-01 -7.99377486e-02 -1.39318597e+00
-3.59059393e-01 -3.54803741e-01 9.44202423e-01 -3.60118181e-01
-1.89712453e+00 5.38183987e-01 -4.96644765e-01 -1.05512571e+00
9.24579985e-03 -3.27836037e-01 -9.53141332e-01 9.16612804e-01
-1.26603580e+00 -1.05933237e+00 1.16183506e-02 7.59232700e-01
-1.11622177e-02 3.65738988e-01 9.84466553e-01 2.04298809e-01
-9.16576326e-01 1.38385713e-01 1.15025379e-01 1.46264628e-01
3.24373245e-01 -1.41149545e+00 6.85234129e-01 2.42405310e-01
4.91009615e-02 4.46732253e-01 3.13327640e-01 -1.25744319e+00
-1.36905265e+00 -1.53349829e+00 1.10350728e+00 -6.76041484e-01
1.35512793e+00 -2.67303824e-01 -1.17594874e+00 7.19380081e-01
-2.67391443e-01 2.61991113e-01 9.99376953e-01 3.71995866e-01
-3.09530079e-01 1.03877895e-01 -1.00907731e+00 8.76665592e-01
1.42162192e+00 -6.73172712e-01 -2.82283306e-01 6.38217211e-01
8.34196150e-01 4.59594242e-02 -1.28221846e+00 2.73041993e-01
3.68330270e-01 -1.64882377e-01 1.14518690e+00 -1.24221957e+00
7.18655050e-01 1.77318290e-01 4.65243369e-01 -1.68738091e+00
-7.75839865e-01 -4.99754310e-01 -7.70968199e-01 6.89153194e-01
6.00346744e-01 -6.35092437e-01 5.52279413e-01 7.86096275e-01
3.33500326e-01 -8.57675672e-01 -6.61382079e-01 -4.44519192e-01
-8.76690447e-02 -2.54636377e-01 7.61343420e-01 1.36857796e+00
2.22078040e-01 7.33523667e-02 -2.27260739e-01 5.11454284e-01
4.54161048e-01 -4.30565961e-02 -1.89964220e-01 -1.48199308e+00
-3.62537801e-01 -7.65115976e-01 -3.25547159e-01 -6.69973850e-01
-1.69976979e-01 -1.19953203e+00 -5.02637386e-01 -1.68401229e+00
4.06229526e-01 -7.04843581e-01 -6.87829494e-01 2.41964579e-01
-4.97068018e-01 -4.82127130e-01 -1.65987432e-01 4.01149839e-01
-3.37474525e-01 5.54986298e-01 1.33535874e+00 -3.21243435e-01
-4.21760023e-01 1.34243041e-01 -3.48923504e-01 6.41470134e-01
7.31645167e-01 -7.20305920e-01 -4.75723118e-01 -5.52970529e-01
6.57525063e-01 8.42126548e-01 4.82327610e-01 -5.24765015e-01
2.48746797e-01 -2.77233660e-01 2.72679001e-01 -2.30763346e-01
2.11118549e-01 -8.82930338e-01 6.28425539e-01 1.04261792e+00
-5.59041202e-01 3.94892812e-01 -1.26421735e-01 1.45285153e+00
-1.43203288e-01 7.14321315e-01 1.78768888e-01 -3.19996998e-02
-4.32250462e-02 1.24897063e+00 -5.85226715e-01 4.60969582e-02
9.13663149e-01 6.19688034e-01 -3.48997325e-01 -4.86944079e-01
-1.38478768e+00 5.28128207e-01 1.19348943e-01 4.20192271e-01
5.10584533e-01 -1.44192362e+00 -6.87217295e-01 -6.29163608e-02
1.73535958e-01 -6.39427528e-02 7.55034804e-01 9.00456131e-01
-4.01228994e-01 2.42203549e-02 1.40491500e-01 -4.12424088e-01
-1.01276839e+00 9.65016186e-01 4.31019783e-01 -9.63531971e-01
-9.08730745e-01 5.87939680e-01 3.74763519e-01 -2.92214006e-01
1.59335166e-01 -4.67170477e-01 -3.70205343e-01 2.33139604e-01
4.93936062e-01 8.11731070e-02 -4.67654824e-01 -2.29534730e-01
-1.51656434e-01 2.14259818e-01 -1.48512036e-01 8.02004039e-02
1.71009946e+00 1.59133822e-02 -1.52081639e-01 6.79387987e-01
1.08760393e+00 -4.63181973e-01 -1.35344076e+00 -5.31152487e-01
2.01186880e-01 8.36939737e-02 6.18717447e-03 -9.37074304e-01
-1.26842403e+00 6.96537256e-01 4.69514966e-01 6.44804835e-01
1.04361343e+00 3.90968859e-01 9.23718214e-01 -3.99606824e-02
-2.33413018e-02 -3.08680236e-01 -1.62391156e-01 2.45521605e-01
7.51270294e-01 -9.23209548e-01 -3.98101538e-01 -3.44633311e-01
-4.88579482e-01 9.48259234e-01 -1.00354187e-01 -3.17835122e-01
1.30408943e+00 3.43367904e-01 -1.54047281e-01 -8.20535600e-01
-1.24979174e+00 2.36702561e-01 5.70809364e-01 5.11782765e-01
4.01153654e-01 7.99583018e-01 -1.32607311e-01 7.16575801e-01
2.08785743e-01 5.67204729e-02 2.71885216e-01 5.69886029e-01
1.36076957e-01 -1.01751125e+00 -4.98857945e-02 9.54594016e-01
-4.39649165e-01 -2.56664127e-01 -2.77087867e-01 3.25907290e-01
-1.59797683e-01 2.88567394e-01 5.36095560e-01 -1.38155192e-01
1.70848459e-01 1.96221247e-01 2.51510322e-01 -6.30086422e-01
-4.78557914e-01 -3.32926005e-01 -1.64044067e-01 -6.01171136e-01
-9.99887064e-02 -9.02395308e-01 -1.08949316e+00 -4.64198470e-01
5.03499031e-01 -3.06801081e-01 2.94288754e-01 5.69771111e-01
7.10272312e-01 1.08628488e+00 5.39602458e-01 -1.25611439e-01
-2.57002860e-01 -5.22981465e-01 -7.49369383e-01 6.43318236e-01
5.34762919e-01 -3.01100135e-01 -1.42772079e-01 2.21830726e-01] | [7.635952472686768, 5.974380016326904] |
96465b18-6531-4492-8ab8-766e9a65875c | delp-dar-system-for-license-plate-detection | 1910.01853 | null | https://arxiv.org/abs/1910.01853v1 | https://arxiv.org/pdf/1910.01853v1.pdf | DELP-DAR System for License Plate Detection and Recognition | Automatic License Plate detection and Recognition (ALPR) is a quite popular and active research topic in the field of computer vision, image processing and intelligent transport systems. ALPR is used to make detection and recognition processes more robust and efficient in highly complicated environments and backgrounds. Several research investigations are still necessary due to some constraints such as: completeness of numbering systems of countries, different colors, various languages, multiple sizes and varied fonts. For this, we present in this paper an automatic framework for License Plate (LP) detection and recognition from complex scenes. Our framework is based on mask region convolutional neural networks used for LP detection, segmentation and recognition. Although some studies have focused on LP detection, LP recognition, LP segmentation or just two of them, our study uses the maskr-cnn in the three stages. The evaluation of our framework is enhanced by four datasets for different countries and consequently with various languages. In fact, it tested on four datasets including images captured from multiple scenes under numerous conditions such as varied orientation, poor quality images, blurred images and complex environmental backgrounds. Extensive experiments show the robustness and efficiency of our suggested framework in all datasets. | ['Zied Selmi', 'Umapada Pal', 'M. Adel Alimi', 'Mohamed Ben Halima'] | 2019-10-04 | null | null | null | null | ['license-plate-detection'] | ['computer-vision'] | [ 8.72780196e-03 -1.13214481e+00 6.96082786e-02 -1.87484156e-02
-3.88746321e-01 -9.59583580e-01 4.75773573e-01 -4.97210652e-01
-4.86119032e-01 7.81003773e-01 -1.72031477e-01 -3.89384180e-01
2.95507550e-01 -6.38193011e-01 -5.45727611e-01 -6.00309491e-01
3.61793667e-01 1.95258737e-01 7.75459170e-01 -1.20726235e-01
9.21465635e-01 1.13854349e+00 -1.41732657e+00 6.53788969e-02
6.71395898e-01 6.50869727e-01 2.31498554e-01 8.49750519e-01
-8.56866091e-02 5.81998169e-01 -4.90028143e-01 -2.84551799e-01
7.20080256e-01 8.64432938e-03 -2.00831339e-01 8.17481935e-01
2.71659076e-01 -2.67702073e-01 -3.74462605e-01 1.11703551e+00
2.42760524e-01 2.88633615e-01 8.87851417e-01 -1.02575672e+00
-7.45326400e-01 -2.10810587e-01 -9.20484245e-01 4.33556825e-01
8.42592046e-02 3.50422978e-01 2.51430124e-01 -1.15818071e+00
3.82355720e-01 8.40747833e-01 8.63525808e-01 1.51394039e-01
-6.29219294e-01 -7.36996889e-01 -2.77759165e-01 2.60661036e-01
-1.79154873e+00 -4.89711672e-01 7.51062334e-01 -5.18276036e-01
8.58371735e-01 2.94300944e-01 7.68207386e-02 4.55958426e-01
1.67406932e-01 4.65823770e-01 1.58786440e+00 -6.31923795e-01
-1.21133164e-01 7.35385776e-01 3.62441689e-01 7.16201901e-01
4.29392159e-01 -1.84629157e-01 1.14591017e-01 3.31363887e-01
9.89855468e-01 1.89406648e-01 -9.80208218e-02 1.88174263e-01
-9.38898087e-01 6.07629478e-01 -2.76586235e-01 4.45100874e-01
-1.34943396e-01 -4.37645197e-01 1.15391269e-01 1.00629658e-01
1.48556367e-01 7.37955794e-02 -2.94764012e-01 8.61207098e-02
-1.14042747e+00 1.33859202e-01 6.83861613e-01 1.15738821e+00
6.95942998e-01 2.95403302e-01 8.92153978e-02 1.23279870e+00
5.59045255e-01 6.58996582e-01 4.66788590e-01 -4.40835804e-01
7.91697860e-01 4.78028774e-01 4.16088372e-01 -1.34225047e+00
-3.48185867e-01 -8.86070654e-02 -5.93553901e-01 5.67102432e-01
3.14041674e-01 -2.24787444e-01 -1.07087183e+00 7.02748656e-01
-2.26560995e-01 2.33319059e-01 1.57383576e-01 8.90599489e-01
8.38485241e-01 1.10351551e+00 -2.51139104e-01 -8.24001655e-02
1.49120939e+00 -1.24661219e+00 -6.54958189e-01 -2.27209508e-01
9.54754949e-02 -1.30298209e+00 6.22693002e-01 5.34360826e-01
-8.03274155e-01 -6.02588534e-01 -1.03659415e+00 1.53178856e-01
-8.19755077e-01 7.54106522e-01 1.76203415e-01 9.50726867e-01
-9.92037833e-01 -7.47720227e-02 -2.81697333e-01 -7.86927640e-01
2.48177126e-01 4.33287024e-01 -5.20123482e-01 -3.59824717e-01
-7.38072276e-01 9.29628611e-01 2.67440826e-01 3.35868061e-01
-5.03487408e-01 1.92584753e-01 -4.48017985e-01 -3.18654686e-01
2.73292899e-01 1.58221558e-01 6.36291981e-01 -9.87756670e-01
-1.46925843e+00 9.95050013e-01 6.22050799e-02 -1.48771971e-01
6.60011530e-01 2.18688846e-02 -9.19849575e-01 -8.88210535e-02
-1.29276082e-01 1.56536475e-01 8.22483301e-01 -1.02827501e+00
-8.06949735e-01 -3.92781228e-01 -2.30554149e-01 3.35706592e-01
3.46918404e-01 6.76725149e-01 -7.42890835e-01 -6.71910942e-01
1.38303816e-01 -9.91648614e-01 8.95028654e-03 -2.85938889e-01
-4.21999693e-01 5.38893044e-02 1.03859735e+00 -9.89175439e-01
6.19149923e-01 -2.18731785e+00 -6.45587862e-01 2.06112593e-01
-2.17654690e-01 6.98435247e-01 3.00631315e-01 3.54776740e-01
1.60450324e-01 2.77574211e-01 -1.12093195e-01 -5.64433746e-02
-7.84571394e-02 2.49827672e-02 -6.62838370e-02 8.87561321e-01
3.23379099e-01 3.28541428e-01 1.67788267e-01 -6.33842111e-01
5.66913903e-01 2.87996680e-01 -1.43046714e-02 2.69805938e-02
3.32967341e-01 2.09864855e-01 -4.19630051e-01 1.19135952e+00
1.22402048e+00 3.62919092e-01 -5.84879577e-01 -2.19762027e-01
-6.54593647e-01 -5.69306016e-01 -1.51316476e+00 7.39856303e-01
-2.53990263e-01 1.25367308e+00 2.45621845e-01 -9.81067836e-01
1.28185809e+00 2.80594587e-01 1.30763734e-02 -5.95470905e-01
2.21180990e-01 4.90575463e-01 -1.68897599e-01 -1.00633693e+00
7.89687157e-01 2.49287765e-02 3.42388362e-01 -2.12752298e-02
-2.58609980e-01 2.09155872e-01 4.22553390e-01 -3.29578340e-01
6.02512717e-01 -8.02881122e-02 9.69405249e-02 -7.01534152e-02
1.19058168e+00 1.46362245e-01 4.78404373e-01 6.44813120e-01
-6.25170529e-01 6.99148417e-01 2.50923503e-02 -3.93082857e-01
-1.19147122e+00 -7.96421826e-01 -2.39116222e-01 7.62219191e-01
3.80383790e-01 5.89135706e-01 -3.80246580e-01 -3.38311672e-01
-1.83217421e-01 4.31134820e-01 -4.71863411e-02 5.56985438e-01
-7.88007677e-01 -9.37081158e-01 8.18460464e-01 2.10381597e-01
1.12032843e+00 -1.07844985e+00 -8.65159407e-02 1.09137394e-01
2.64545262e-01 -1.46310449e+00 -3.61429632e-01 -3.69675666e-01
-7.41507232e-01 -1.08326840e+00 -9.60812390e-01 -1.33085203e+00
8.33561599e-01 5.80196083e-01 7.20502734e-01 -2.16558799e-02
-3.75814617e-01 1.62060574e-01 -3.34932595e-01 -3.76414478e-01
-3.96349400e-01 -5.04736662e-01 -1.01548143e-01 3.25281948e-01
4.22085673e-01 -2.21502841e-01 -2.87251234e-01 6.90300822e-01
-9.35885847e-01 -2.04330251e-01 9.45711076e-01 2.62385190e-01
1.10007070e-01 5.14295757e-01 3.11057627e-01 -8.17226827e-01
8.86434913e-01 -3.94538462e-01 -1.14293313e+00 3.16225827e-01
-2.93902606e-01 -5.08797228e-01 6.01860344e-01 -1.59595117e-01
-1.21914160e+00 -9.84763913e-03 -2.56261796e-01 -7.38615319e-02
-8.65239561e-01 6.99457601e-02 -1.27712831e-01 -5.25524616e-01
3.29852194e-01 8.78980935e-01 -1.11191571e-01 -3.96790594e-01
-6.39534965e-02 1.25004351e+00 4.97413874e-01 -2.66152740e-01
1.01195145e+00 2.59542018e-01 -9.36197564e-02 -1.26276457e+00
1.25916407e-01 -8.08510065e-01 -5.90467334e-01 -5.14269888e-01
8.55405092e-01 -8.05905759e-01 -5.49695492e-01 1.09707797e+00
-1.10074103e+00 3.78411680e-01 7.17613935e-01 6.74991190e-01
2.34179404e-02 5.92390478e-01 -6.41540647e-01 -1.05493116e+00
-1.01873867e-01 -1.49359441e+00 5.56721509e-01 5.95822036e-01
7.43600547e-01 -9.26546991e-01 -1.83933243e-01 7.44975507e-01
7.39360392e-01 1.56508774e-01 4.41450596e-01 -6.75398350e-01
-8.91505837e-01 -5.57314932e-01 -6.39264941e-01 6.27095222e-01
2.04031870e-01 5.57427526e-01 -8.14485908e-01 2.05565035e-01
-7.77754560e-02 -5.61399125e-02 8.10661256e-01 3.87946218e-01
7.27034569e-01 -3.06154698e-01 -9.60285887e-02 4.63974953e-01
2.01158667e+00 7.95076311e-01 9.41863358e-01 6.05368555e-01
6.21117711e-01 3.83982807e-01 4.19372827e-01 2.16671139e-01
4.09515314e-02 5.58633566e-01 4.61485535e-02 -2.16928333e-01
-1.38471173e-02 4.26243097e-01 6.04615152e-01 8.06331635e-01
-3.93929660e-01 -4.26104575e-01 -1.05411804e+00 2.86624193e-01
-1.28802705e+00 -9.92879868e-01 -5.24937212e-01 1.98073006e+00
2.30323628e-01 7.72339031e-02 2.20462292e-01 1.37280345e-01
1.21169043e+00 2.29662377e-02 -8.64605382e-02 -5.71947873e-01
-2.76419669e-01 -2.35067084e-01 1.09000778e+00 4.38007265e-01
-1.38236904e+00 1.00039244e+00 6.12926340e+00 7.81483114e-01
-1.47467911e+00 -7.62297437e-02 6.20651424e-01 5.98590255e-01
4.17955339e-01 -4.20709044e-01 -1.02786160e+00 5.60217738e-01
3.96657079e-01 2.00331017e-01 5.17259955e-01 5.61841071e-01
4.01249528e-01 -3.39932680e-01 -3.24050099e-01 1.10385072e+00
4.61492270e-01 -1.26024890e+00 -1.99163452e-01 3.14038247e-03
8.21449518e-01 1.81630358e-01 1.92551374e-01 2.54106879e-01
-6.26273528e-02 -9.69392419e-01 4.29714561e-01 5.77768743e-01
4.16789412e-01 -9.01644647e-01 1.05213535e+00 3.06450158e-01
-9.71869290e-01 7.76789803e-03 -5.18557549e-01 -2.25001164e-02
-1.62133992e-01 7.95170814e-02 -1.02207386e+00 3.33558649e-01
2.41238132e-01 6.06683433e-01 -7.78605521e-01 1.40137112e+00
3.35598528e-01 7.83823311e-01 -2.34453127e-01 -2.57105857e-01
5.70199966e-01 -6.65358126e-01 4.98048723e-01 1.67554712e+00
3.75286192e-01 2.21656952e-02 4.76646982e-02 9.14516985e-01
9.39421728e-02 3.89733374e-01 -7.99631357e-01 1.08249359e-01
2.57172227e-01 1.34178329e+00 -1.21345198e+00 -1.00524515e-01
-8.59428167e-01 7.21561432e-01 -3.67522329e-01 5.04520774e-01
-1.03448248e+00 -4.46366161e-01 1.14120051e-01 1.81018159e-01
3.65644187e-01 -6.40642166e-01 -1.84271216e-01 -1.08333397e+00
-6.85642287e-03 -6.78322017e-01 7.96285123e-02 -7.33329177e-01
-1.03400660e+00 4.32542980e-01 -6.96522146e-02 -1.40497315e+00
4.68434304e-01 -1.16538751e+00 -8.15502882e-01 7.98832059e-01
-1.68840230e+00 -1.11933994e+00 -1.86926588e-01 8.02652299e-01
1.20014274e+00 -8.96550536e-01 2.11147472e-01 5.58687091e-01
-1.10628998e+00 2.65236646e-01 5.97842753e-01 5.53871691e-01
5.91802359e-01 -9.58584905e-01 -1.16188936e-01 1.38320041e+00
-9.28501189e-02 4.42176402e-01 5.67951560e-01 -4.53408509e-01
-1.40016985e+00 -1.13264132e+00 3.86196375e-01 -4.33244742e-02
4.22928542e-01 -3.09678346e-01 -6.12959206e-01 5.96206725e-01
4.23500091e-01 1.13767721e-01 5.22096634e-01 -5.15960157e-01
1.30860984e-01 -3.29296917e-01 -1.33588696e+00 4.82286245e-01
1.23712204e-01 -3.63902152e-02 -5.38931310e-01 4.66955811e-01
-9.08504725e-02 -7.73728266e-02 -5.14586806e-01 -2.61467863e-02
4.69049096e-01 -1.02278340e+00 7.21852660e-01 7.16493502e-02
8.23410451e-02 -8.94324541e-01 -2.69797891e-01 -6.24538064e-01
-2.45139554e-01 -8.34324583e-02 7.08291650e-01 1.53729022e+00
5.75532794e-01 -8.71197760e-01 6.62195563e-01 7.03701735e-01
-2.68274516e-01 -2.54366398e-01 -6.42563462e-01 -6.85957134e-01
-1.70606405e-01 -4.08936322e-01 7.52534717e-02 7.28400350e-01
-7.38698661e-01 -1.02814719e-01 -6.60476983e-01 5.90999484e-01
4.89644796e-01 -8.48313347e-02 7.40224957e-01 -9.53436077e-01
6.20243810e-02 -3.79253209e-01 -8.46257627e-01 -5.18438160e-01
-1.22061901e-01 -3.97090733e-01 1.78140327e-01 -1.55138588e+00
1.10701412e-01 -4.00621444e-01 -1.62095979e-01 1.31237879e-01
2.13305488e-01 6.15806758e-01 3.01097363e-01 5.90235591e-01
-6.95668697e-01 -8.82560015e-02 1.01210129e+00 -3.23565125e-01
-2.01280136e-02 3.50955814e-01 -2.60087311e-01 8.89207900e-01
1.00957179e+00 -2.85197437e-01 -1.89398736e-01 -3.77347350e-01
-1.29759923e-01 -5.17175905e-02 2.22321272e-01 -1.30283022e+00
5.09620070e-01 -2.22318202e-01 6.62990510e-01 -8.63296568e-01
2.17069045e-01 -1.08948469e+00 1.58248141e-01 1.83150262e-01
1.28808469e-01 2.69264519e-01 3.50193739e-01 4.08905089e-01
-3.90477180e-01 -5.71266294e-01 8.93525422e-01 -5.15003562e-01
-1.40307593e+00 -4.28975262e-02 -9.41123843e-01 -2.71477014e-01
1.43825483e+00 -9.23275948e-01 -3.21697772e-01 -1.60466403e-01
-3.10415894e-01 -1.69536456e-01 5.18134534e-01 3.98545355e-01
7.41359830e-01 -1.15015340e+00 -9.74542320e-01 4.03219730e-01
-3.02851070e-02 -3.02983373e-01 1.16898105e-01 8.80751789e-01
-1.59979248e+00 7.07290590e-01 -6.97140694e-01 -3.96153897e-01
-1.51242471e+00 3.46750289e-01 3.34053785e-01 6.67796955e-02
-3.53924811e-01 4.06636178e-01 -1.11606121e-01 -3.49478275e-01
5.72234504e-02 -2.93498546e-01 -7.02941179e-01 -3.75522077e-01
4.20986682e-01 6.25663161e-01 6.79824129e-02 -1.20209682e+00
-4.32437778e-01 1.05137122e+00 -1.19703889e-01 -1.20094702e-01
1.12474144e+00 -1.38687193e-01 -1.87020823e-01 2.66754389e-01
9.98586476e-01 5.24175823e-01 -1.00730526e+00 1.74251646e-02
-4.32197424e-03 -6.63712502e-01 -1.24029666e-01 -7.61041582e-01
-1.08335257e+00 5.77177584e-01 7.23554432e-01 1.83878690e-01
1.04711366e+00 -6.58530951e-01 4.07268107e-01 4.60942924e-01
2.27787256e-01 -1.35048091e+00 -3.97418082e-01 5.35714388e-01
8.19860876e-01 -1.37706554e+00 -7.93712493e-03 -2.88236439e-01
-7.19810545e-01 1.48788679e+00 5.38896143e-01 -4.07639593e-01
5.50513923e-01 4.34973598e-01 1.95925891e-01 2.90524304e-01
-2.75665879e-01 -1.10967547e-01 1.66277379e-01 6.38182104e-01
3.80655438e-01 -3.17100771e-02 -4.82042640e-01 2.18744829e-01
4.06151831e-01 5.95935211e-02 9.80220437e-01 9.93493736e-01
-8.43359709e-01 -7.15514600e-01 -1.19368017e+00 4.64579582e-01
-9.30689216e-01 4.32079956e-02 -2.72600383e-01 1.25309646e+00
5.23215652e-01 1.18624496e+00 -7.36085474e-02 -9.22607630e-02
1.92523733e-01 -8.47662240e-02 1.65284291e-01 -1.88064888e-01
-2.32763886e-01 1.66604459e-01 1.13945410e-01 1.67371958e-01
-5.91819525e-01 -7.80417144e-01 -8.62445474e-01 -2.21614510e-01
-4.30946440e-01 -1.29859403e-01 1.07360101e+00 9.11332786e-01
-9.91043597e-02 2.27053538e-01 7.01198578e-01 -6.57741904e-01
-1.18196063e-01 -1.02065802e+00 -1.06248009e+00 2.37249434e-01
2.09440738e-01 -5.66738784e-01 -2.14686126e-01 2.92549163e-01] | [9.818549156188965, -4.972363471984863] |
86d16de6-9398-4f58-86f2-9de31f12a5c9 | procedural-content-generation-via-machine | 1702.00539 | null | http://arxiv.org/abs/1702.00539v3 | http://arxiv.org/pdf/1702.00539v3.pdf | Procedural Content Generation via Machine Learning (PCGML) | This survey explores Procedural Content Generation via Machine Learning
(PCGML), defined as the generation of game content using machine learning
models trained on existing content. As the importance of PCG for game
development increases, researchers explore new avenues for generating
high-quality content with or without human involvement; this paper addresses
the relatively new paradigm of using machine learning (in contrast with
search-based, solver-based, and constructive methods). We focus on what is most
often considered functional game content such as platformer levels, game maps,
interactive fiction stories, and cards in collectible card games, as opposed to
cosmetic content such as sprites and sound effects. In addition to using PCG
for autonomous generation, co-creativity, mixed-initiative design, and
compression, PCGML is suited for repair, critique, and content analysis because
of its focus on modeling existing content. We discuss various data sources and
representations that affect the resulting generated content. Multiple PCGML
methods are covered, including neural networks, long short-term memory (LSTM)
networks, autoencoders, and deep convolutional networks; Markov models,
$n$-grams, and multi-dimensional Markov chains; clustering; and matrix
factorization. Finally, we discuss open problems in the application of PCGML,
including learning from small datasets, lack of training data, multi-layered
learning, style-transfer, parameter tuning, and PCG as a game mechanic. | ['Christoffer Holmgård', 'Sam Snodgrass', 'Andy Nealen', 'Adam Summerville', 'Julian Togelius', 'Aaron Isaksen', 'Matthew Guzdial', 'Amy K. Hoover'] | 2017-02-02 | null | null | null | null | ['card-games'] | ['playing-games'] | [ 2.02201396e-01 3.89973521e-01 -4.27832492e-02 4.16375339e-01
-4.93381321e-01 -5.21650672e-01 6.94230556e-01 -1.45220339e-01
-3.72572578e-02 6.74246013e-01 7.57752001e-01 -3.67664099e-01
-3.76358241e-01 -1.41170299e+00 -6.93439305e-01 -1.81105450e-01
1.37971982e-01 3.09349358e-01 -2.55777389e-01 -5.28968513e-01
6.56450748e-01 1.50182517e-02 -1.63613069e+00 7.70777464e-01
8.92873704e-01 7.88332283e-01 4.64366496e-01 8.27013433e-01
-7.50189483e-01 1.53000653e+00 -8.89331937e-01 -7.18402922e-01
-1.03300400e-01 -7.81095207e-01 -9.31911349e-01 -1.36435375e-01
-2.17523798e-01 -3.04773420e-01 -1.01486444e-02 6.63128197e-01
5.12177646e-01 1.48619443e-01 7.94867814e-01 -1.05662763e+00
-1.21914828e+00 1.09819758e+00 -2.42962763e-02 -2.30053946e-01
6.76852465e-01 1.65186107e-01 9.09861088e-01 -7.17280924e-01
8.53940547e-01 9.29803729e-01 8.99384141e-01 7.02296019e-01
-9.69600618e-01 -4.86750782e-01 -2.30831310e-01 1.84310172e-02
-8.59809875e-01 -9.36837308e-03 9.23881054e-01 -8.91294301e-01
1.23046899e+00 7.79083669e-02 1.27933228e+00 1.36435401e+00
3.60093176e-01 8.85826647e-01 9.68326926e-01 -7.88341045e-01
2.79855520e-01 1.08040869e-01 -6.25532627e-01 5.57108641e-01
-1.08939417e-01 3.18763256e-01 -6.65208042e-01 -7.96412602e-02
1.20124817e+00 -4.46261406e-01 1.89050049e-01 2.43900210e-01
-7.90615618e-01 1.24722826e+00 3.84894572e-02 4.78804857e-01
-4.75199103e-01 1.94175020e-01 3.41674805e-01 1.51039466e-01
5.63865483e-01 1.18072212e+00 -3.61947268e-01 -9.95384037e-01
-9.42795813e-01 6.92566752e-01 9.87696707e-01 1.00072277e+00
4.15836960e-01 6.32275462e-01 -1.85295448e-01 1.03216720e+00
-2.63817899e-04 -3.52849811e-02 1.06696868e+00 -1.16803765e+00
2.16109857e-01 6.38290346e-01 -1.62107810e-01 -1.22328079e+00
-1.16422765e-01 -3.41418386e-01 -6.08177602e-01 3.70564342e-01
4.48150560e-02 -4.72691447e-01 -5.99305868e-01 1.45168996e+00
-3.45652491e-01 -7.59095773e-02 -2.91253366e-02 3.89990419e-01
1.26386964e+00 8.31908643e-01 6.94865882e-02 5.45703955e-02
9.02926385e-01 -1.04458165e+00 -7.02148259e-01 -3.40695918e-01
6.88412845e-01 -7.72519171e-01 1.11134493e+00 6.82176113e-01
-1.61685193e+00 -4.27454442e-01 -7.20448256e-01 7.68649504e-02
-7.40915298e-01 -5.13338074e-02 8.95482898e-01 6.87545896e-01
-1.27895153e+00 9.78351355e-01 -1.52577892e-01 -1.52078599e-01
4.41697776e-01 2.61121206e-02 7.19263405e-02 2.48331532e-01
-1.19429541e+00 9.55688477e-01 6.30483687e-01 -3.43401223e-01
-3.05691034e-01 -7.91044474e-01 -7.39366949e-01 -2.44416408e-02
2.43415102e-01 -6.40586853e-01 1.23779011e+00 -1.35627306e+00
-2.09837484e+00 6.79342806e-01 4.38960880e-01 -2.69885570e-01
1.66150719e-01 -2.83637047e-01 -2.41653889e-01 -1.65454805e-01
3.19411457e-02 6.20446205e-01 4.94168073e-01 -8.65345061e-01
-3.88820231e-01 2.32003272e-01 2.49535516e-01 3.53054971e-01
-4.65999931e-01 1.21971756e-01 -9.83808637e-02 -1.01238072e+00
-1.93423122e-01 -6.23978436e-01 -3.51721019e-01 -5.67299604e-01
-4.01566853e-04 -2.47428760e-01 3.55402261e-01 -1.10765016e+00
1.42986584e+00 -1.84003246e+00 2.30529577e-01 8.38580802e-02
1.61673561e-01 1.46411151e-01 -1.56217992e-01 8.19142997e-01
-2.43558921e-02 4.56966788e-01 3.48324239e-01 3.17095608e-01
3.25856991e-02 -1.76709145e-01 5.48794014e-05 -6.70206964e-01
7.04199076e-02 1.19088387e+00 -8.95713389e-01 -3.41986865e-01
1.34983182e-01 7.28445798e-02 -1.03229272e+00 1.77143157e-01
-5.35890937e-01 3.79660241e-02 -6.92138001e-02 4.90109593e-01
6.65400922e-02 -1.96977258e-01 7.15765059e-02 2.66907364e-01
-2.50795871e-01 3.50220859e-01 -1.13593495e+00 1.66351545e+00
-8.56317163e-01 8.39800656e-01 -3.72574747e-01 -6.89044356e-01
1.20057726e+00 4.18487042e-01 3.52016896e-01 -8.38625789e-01
1.12689465e-01 2.06627458e-01 -1.98697552e-01 -6.79160595e-01
9.78644788e-01 -2.29762465e-01 -2.41188988e-01 7.27983832e-01
3.65181863e-01 -4.34794664e-01 2.33456343e-01 1.60013005e-01
1.07243645e+00 4.93319452e-01 1.23211429e-01 1.37933537e-01
-3.02149743e-01 3.91866602e-02 2.38448679e-01 8.11535597e-01
3.96414250e-01 5.86474359e-01 5.49097240e-01 -2.23125055e-01
-1.19862914e+00 -7.63215184e-01 5.47856867e-01 1.33509171e+00
-5.09971499e-01 -7.72200286e-01 -9.96456027e-01 1.20463960e-01
-4.62839603e-01 9.25433278e-01 -4.70925927e-01 -3.65044534e-01
-4.23577905e-01 -4.78032082e-01 7.03736126e-01 5.35275400e-01
3.48304212e-01 -1.74815023e+00 -5.20490527e-01 7.87101269e-01
-3.40003431e-01 -6.15181208e-01 7.11499602e-02 -5.45357075e-03
-9.31944132e-01 -8.41501832e-01 -5.20358264e-01 -8.57757628e-01
-1.10655591e-01 -2.03458041e-01 1.45575428e+00 -1.88901410e-01
-2.11432546e-01 3.77719671e-01 -7.09885716e-01 -5.49231231e-01
-7.19380975e-01 1.29625887e-01 -3.37493062e-01 -6.16253614e-01
2.10482672e-01 -8.79745126e-01 -1.72991589e-01 -3.27000678e-01
-9.32057440e-01 5.87108850e-01 7.46123672e-01 9.26175356e-01
1.81534067e-01 2.87898809e-01 5.64224899e-01 -9.28282440e-01
1.51959145e+00 -6.14898205e-01 -9.43125412e-02 1.70091596e-02
-6.81643546e-01 -2.19955206e-01 5.26652813e-01 -6.81890070e-01
-1.23707664e+00 -6.72361732e-01 -3.11691582e-01 -2.26827472e-01
-2.88015092e-03 9.81301188e-01 -9.56016555e-02 5.04521094e-02
1.13365901e+00 3.11412781e-01 2.02713996e-01 -2.63673991e-01
8.54025126e-01 5.86375177e-01 2.52048314e-01 -8.34147990e-01
4.57370430e-01 -9.08362865e-02 -5.86142004e-01 -6.50320828e-01
-1.80851758e-01 2.72496164e-01 -1.97489336e-01 -7.33632207e-01
7.98303068e-01 -6.86958373e-01 -4.25734758e-01 5.69054067e-01
-1.06203508e+00 -7.27880478e-01 -6.88518405e-01 2.57674038e-01
-9.42629635e-01 -3.98606688e-01 -1.03125906e+00 -7.12223232e-01
-4.16190803e-01 -7.51127481e-01 3.12843621e-01 5.78945220e-01
-8.19076777e-01 -1.08100033e+00 1.54992491e-01 6.28732562e-01
5.61595201e-01 5.39096832e-01 1.31642234e+00 -3.69420230e-01
-3.63492966e-01 -2.90822327e-01 2.09298685e-01 2.80939281e-01
-6.14616871e-02 1.29484892e-01 -5.61849177e-01 3.97329867e-01
-2.32283503e-01 -6.30505204e-01 3.17046642e-01 6.54531002e-01
1.13552201e+00 -8.22794974e-01 1.65163025e-01 4.11816359e-01
1.34205770e+00 7.89267242e-01 9.77473021e-01 5.65661788e-01
6.03304207e-01 7.11164057e-01 3.59715134e-01 6.71557426e-01
2.51520485e-01 2.97082126e-01 1.18074216e-01 1.71478063e-01
-4.78602899e-03 -6.92066312e-01 3.78833383e-01 1.13941789e+00
-5.96681952e-01 6.47359118e-02 -8.21033895e-01 5.38828433e-01
-1.66573000e+00 -1.45611346e+00 -1.07897200e-01 1.60937262e+00
9.79465544e-01 2.83352733e-01 2.96892524e-01 1.58642709e-01
4.02339369e-01 6.53620660e-02 -1.78909257e-01 -1.13352299e+00
-1.19034037e-01 7.57324934e-01 9.36042443e-02 1.71335325e-01
-6.84087753e-01 1.23508513e+00 6.87868738e+00 1.34450662e+00
-8.85254622e-01 2.71187186e-01 7.72553980e-01 -2.98576951e-01
-6.65153205e-01 -1.79936454e-01 -2.73750395e-01 5.34280181e-01
9.88482058e-01 -2.40540460e-01 8.46975267e-01 1.21073174e+00
4.28735316e-01 -1.05831034e-01 -4.39781070e-01 9.18254554e-01
1.72320426e-01 -2.18366885e+00 1.16932072e-01 1.23642020e-01
1.09734499e+00 -4.69198793e-01 1.61675155e-01 6.91598654e-01
7.93896139e-01 -1.32382810e+00 1.08729219e+00 5.28634489e-01
8.45995605e-01 -9.32739854e-01 3.72101307e-01 2.49243468e-01
-9.29405630e-01 -2.79302716e-01 -1.85505182e-01 -7.45118141e-01
-7.29618371e-02 4.40703571e-01 -5.16679287e-01 2.19970748e-01
6.82277024e-01 4.64766830e-01 -4.95349973e-01 8.42400908e-01
-1.74719170e-01 6.48296654e-01 2.06527039e-01 -5.61002314e-01
2.98610687e-01 -2.98241794e-01 3.29538345e-01 1.17391086e+00
5.13973892e-01 3.13433349e-01 -1.83962300e-01 1.23324871e+00
3.40222210e-01 2.51450211e-01 -7.94885099e-01 -6.69248641e-01
3.76933128e-01 9.33494806e-01 -6.95174217e-01 -2.70814765e-02
-1.84720576e-01 7.56912053e-01 1.28746688e-01 2.87448913e-01
-6.67684615e-01 -5.01246631e-01 3.87738913e-01 4.72447097e-01
7.60129699e-03 -1.38082979e-02 -8.81034672e-01 -7.79259145e-01
-2.76235789e-01 -1.27947593e+00 1.55943232e-02 -1.07502580e+00
-1.09746540e+00 6.10716522e-01 -9.08464044e-02 -1.08962524e+00
-8.91164005e-01 -5.04345059e-01 -8.76255274e-01 8.55939329e-01
-3.84741098e-01 -1.21670222e+00 -1.62377223e-01 3.17733288e-01
7.78769016e-01 -6.66949272e-01 9.70983207e-01 3.46121527e-02
-3.66040654e-02 3.68746698e-01 3.05010647e-01 1.97380140e-01
1.06609590e-01 -1.04215419e+00 3.70227784e-01 4.48026389e-01
-6.35049026e-03 2.24419966e-01 3.66673350e-01 -8.19692016e-01
-9.60869074e-01 -9.74570096e-01 7.44695783e-01 -2.31663421e-01
4.92401600e-01 -1.99586600e-01 -3.74229550e-01 3.82823408e-01
2.17551827e-01 -1.15666616e+00 1.06035674e+00 3.34162384e-01
-3.46037373e-02 2.97683209e-01 -8.99316549e-01 8.90128672e-01
1.06818914e+00 -4.58529413e-01 -3.86917889e-01 4.89913970e-02
7.40711927e-01 -4.73220050e-01 -7.81377137e-01 -1.74693272e-01
6.35218978e-01 -9.94222403e-01 8.97397816e-01 -5.41469216e-01
1.41487908e+00 3.75162423e-01 1.31685778e-01 -1.51256287e+00
-7.90604293e-01 -7.72170782e-01 -1.39322922e-01 1.33182430e+00
4.49471325e-01 2.25367881e-02 1.14057505e+00 7.45663583e-01
-2.64995575e-01 -8.52820158e-01 -4.67322230e-01 -4.74663228e-01
4.58301067e-01 -7.71923184e-01 6.89870656e-01 1.04240882e+00
4.06109035e-01 2.10343540e-01 -5.54076433e-01 -9.38603222e-01
5.68940863e-02 -8.26168358e-02 7.49473810e-01 -1.20019615e+00
-7.76550293e-01 -1.02682996e+00 -1.21421345e-01 -5.19518733e-01
-2.67036613e-02 -8.50588083e-01 -3.43189873e-02 -1.92969954e+00
-4.69994284e-02 -2.31967732e-01 4.27135944e-01 2.43415073e-01
1.04531705e-01 9.58870798e-02 5.03126204e-01 -9.12399516e-02
-2.08087996e-01 2.82706499e-01 1.35720980e+00 -7.28390366e-02
-7.26274014e-01 -2.91853044e-02 -1.27500260e+00 8.41605067e-01
1.04169285e+00 -2.04120442e-01 -5.58067083e-01 -1.86593384e-01
9.03316975e-01 2.35727698e-01 8.88243392e-02 -1.06752920e+00
-2.85520684e-02 -6.37173831e-01 4.24145579e-01 -1.46883652e-01
2.54440576e-01 -1.34985030e-01 5.60488641e-01 2.06674471e-01
-4.06821460e-01 -2.73981392e-02 2.32476890e-01 1.10363439e-01
-9.93185863e-02 -6.75650418e-01 3.70945215e-01 -9.32969868e-01
-6.91456556e-01 -2.36247852e-01 -1.20824885e+00 2.13050529e-01
6.01112127e-01 -7.50395894e-01 -9.78370234e-02 -1.00881565e+00
-8.50703835e-01 -3.92648131e-01 1.76179245e-01 6.77165747e-01
8.40237617e-01 -1.55084383e+00 -5.66954970e-01 -8.41747597e-03
-2.56627381e-01 -2.56159827e-02 3.99317086e-01 -6.82129860e-02
-7.32849300e-01 8.60947073e-02 -4.73376691e-01 2.52984852e-01
-8.07890356e-01 2.93874860e-01 1.50256008e-01 -6.23126984e-01
-3.14790368e-01 8.95230532e-01 -1.82037979e-01 -4.57086980e-01
-1.04337178e-01 1.67393237e-02 -5.10043800e-01 2.39454955e-01
4.10510123e-01 5.42205274e-01 -1.17353000e-01 -1.83468357e-01
4.50550109e-01 2.45462134e-02 2.16472790e-01 -5.23563027e-01
1.42709029e+00 3.83355260e-01 -1.19114392e-01 6.75518811e-01
7.11560845e-01 -3.72416049e-01 -8.86099637e-01 1.51309222e-01
-1.01607494e-01 -2.34613046e-01 -5.73162287e-02 -1.13958216e+00
-8.54528427e-01 7.05885649e-01 1.69551209e-01 2.97490776e-01
1.06017971e+00 -2.10448533e-01 6.99731767e-01 5.41942380e-02
4.48406398e-01 -1.66721070e+00 7.64519930e-01 8.38609397e-01
1.04378247e+00 -4.34092790e-01 -4.11577284e-01 4.73386236e-02
-8.25625777e-01 1.17276561e+00 9.49949622e-01 -1.52919009e-01
5.85451901e-01 3.79787713e-01 -1.74483508e-01 -9.48430412e-03
-7.92057574e-01 2.84305215e-02 -1.40150869e-02 1.00637221e+00
6.23480558e-01 2.17226878e-01 -3.31556410e-01 1.27273679e+00
-1.08133245e+00 4.52504963e-01 9.55043972e-01 9.48066592e-01
-3.15705180e-01 -1.06008196e+00 -4.84438479e-01 9.79546368e-01
-5.35802841e-01 -4.91761595e-01 -5.88402748e-01 5.14967918e-01
5.74688911e-01 9.72885609e-01 1.13940723e-01 -8.14397275e-01
-9.86059755e-03 1.96073994e-01 4.60349500e-01 -7.19272375e-01
-1.01530099e+00 1.46740422e-01 2.47416183e-01 -2.96082944e-01
-1.18051469e-01 -5.20830214e-01 -7.59767890e-01 -7.44310200e-01
-2.75524110e-01 1.77497700e-01 6.12823904e-01 9.34947431e-01
3.49283874e-01 8.47781777e-01 1.03496596e-01 -9.10521865e-01
1.99814573e-01 -1.14188647e+00 -5.34711599e-01 1.40023455e-01
-7.06690192e-01 -4.01400775e-01 2.65247561e-02 3.10377896e-01] | [11.672303199768066, 8.965797424316406] |
3ea17e29-9cac-4e51-baae-3a66f0e2a86e | breaking-character-are-subwords-good-enough | null | null | https://openreview.net/forum?id=6RjSJUjpcuV | https://openreview.net/pdf?id=6RjSJUjpcuV | Breaking Character: Are Subwords Good Enough for MRLs After All? | Large pretrained language models (PLMs) typically tokenize the input string into contiguous subwords before any pretraining or inference. However, previous studies have claimed that this form of subword tokenization is inadequate for processing morphologically-rich languages (MRLs). We revisit this hypothesis by pretraining a BERT-style masked language model over character sequences instead of word-pieces. We compare the resulting model, dubbed TavBERT, against contemporary PLMs based on subwords for three highly complex and ambiguous MRLs (Hebrew, Turkish, and Arabic), testing them on both morphological and semantic tasks. Our results show, for all tested languages, that while TavBERT obtains mild improvements on surface-level tasks à la POS tagging and full morphological disambiguation, subword-based PLMs achieve significantly higher performance on semantic tasks, such as named entity recognition and extractive question answering. These results showcase and (re)confirm the potential of subword tokenization as a reasonable modeling assumption for many languages, including MRLs. | ['Anonymous'] | 2022-01-16 | null | null | null | acl-arr-january-2022-1 | ['morphological-disambiguation'] | ['natural-language-processing'] | [ 3.41775902e-02 2.34997258e-01 -8.70311558e-02 -3.71106923e-01
-8.29829037e-01 -1.02366471e+00 4.89436984e-01 6.80981696e-01
-1.23932433e+00 7.00981021e-01 2.03537628e-01 -9.13111866e-01
1.54818535e-01 -8.35167944e-01 -6.62250817e-01 -8.89987499e-02
1.22597009e-01 4.82481837e-01 3.51693243e-01 -3.21086824e-01
3.57067168e-01 6.28302634e-01 -9.19880927e-01 6.74614251e-01
1.13583767e+00 4.64829624e-01 3.33886743e-01 2.04852000e-01
-7.75741994e-01 5.46801150e-01 -7.88804591e-01 -7.26613343e-01
-1.04227409e-01 -5.98148257e-02 -1.23285508e+00 -8.27242807e-02
5.05331039e-01 3.15255582e-01 2.21573800e-01 9.14451122e-01
2.00804189e-01 1.72167376e-01 5.12308240e-01 -5.84576368e-01
-9.85968947e-01 1.27658641e+00 -1.96028307e-01 4.07857537e-01
3.73264998e-01 -6.40940070e-02 1.29794872e+00 -1.22156990e+00
6.62659287e-01 1.48708320e+00 8.59467149e-01 4.47340637e-01
-1.30601537e+00 -4.23639864e-01 4.83357161e-01 1.67446703e-01
-1.66213167e+00 -4.52976137e-01 3.76812458e-01 -1.51020288e-01
1.62889993e+00 1.18836880e-01 2.62826569e-02 6.79738283e-01
1.04544468e-01 7.34299719e-01 1.24544573e+00 -9.01433110e-01
1.38356030e-01 2.98122227e-01 3.81096005e-01 6.16409779e-01
6.48742080e-01 -3.65050852e-01 -4.11534905e-01 -7.99082965e-02
5.29370546e-01 -8.24630678e-01 1.71206277e-02 3.71070892e-01
-1.37529016e+00 7.27457941e-01 4.49233428e-02 9.28503811e-01
-4.44467485e-01 -1.81363210e-01 5.06180108e-01 3.80278677e-01
3.47149342e-01 8.22113991e-01 -9.78313327e-01 2.83897281e-01
-8.67971241e-01 -4.79189456e-02 7.07551062e-01 6.86189473e-01
7.43261397e-01 4.50961620e-01 -1.21033497e-01 1.07250893e+00
5.75535968e-02 5.02966821e-01 9.23783839e-01 -4.11610931e-01
6.11877024e-01 4.91761714e-01 5.71981296e-02 -7.02611566e-01
-3.97756726e-01 -1.23075679e-01 -1.79589018e-01 -2.12768093e-01
6.77033842e-01 -1.69600099e-01 -1.15693724e+00 1.85348237e+00
7.08388817e-03 -4.07998800e-01 3.95602018e-01 4.41764593e-01
8.73295188e-01 7.31111050e-01 9.70881641e-01 -1.03738919e-01
1.79549444e+00 -5.56623459e-01 -6.25762224e-01 -7.33049154e-01
9.59909499e-01 -8.06033015e-01 1.57602000e+00 3.53398323e-01
-1.38125050e+00 -7.86684155e-01 -8.97463381e-01 -2.66272575e-01
-8.77266705e-01 2.30064958e-01 6.56429410e-01 1.13953900e+00
-1.13864970e+00 4.30582285e-01 -6.75539374e-01 -4.73838210e-01
1.90916043e-02 2.73774087e-01 -5.94989657e-01 -2.31706902e-01
-1.37805414e+00 1.15239441e+00 1.10387623e+00 1.11546926e-01
-4.35627073e-01 -5.11916101e-01 -1.19553983e+00 -2.54994491e-03
2.38874823e-01 -1.30206868e-01 1.07426620e+00 -9.81884181e-01
-1.06456602e+00 1.47449851e+00 -3.14603448e-01 -4.01703537e-01
-1.68656722e-01 -3.47313821e-01 -6.44028366e-01 4.35676649e-02
9.07624811e-02 7.35310614e-01 5.36909580e-01 -1.23971832e+00
-5.25638282e-01 -1.42634928e-01 -6.92846328e-02 5.33625111e-03
-2.27570474e-01 7.55360484e-01 5.89118488e-02 -9.85941768e-01
7.90665150e-02 -4.98577207e-01 -1.58653080e-01 -7.23532140e-01
-1.36702418e-01 -7.81239808e-01 1.56168133e-01 -9.60692644e-01
1.29213715e+00 -1.96763015e+00 -1.88095286e-01 1.51076481e-01
-3.44588310e-01 7.20302522e-01 -5.97908974e-01 4.44482267e-01
-4.30499375e-01 6.18722856e-01 -3.44697177e-01 -2.59426653e-01
2.27229238e-01 6.20586276e-01 -3.55873436e-01 2.23946273e-01
6.95747912e-01 1.38906848e+00 -6.76413655e-01 -4.62719113e-01
-1.79448316e-03 6.90394416e-02 -4.07017201e-01 -1.35157943e-01
-4.99780476e-01 -2.89944559e-02 7.52417892e-02 5.80519557e-01
5.71489215e-01 1.40277952e-01 5.34817994e-01 1.27968062e-02
-3.31325918e-01 8.73002291e-01 -1.12098610e+00 1.42577434e+00
-5.95800102e-01 2.24278197e-01 -3.95264365e-02 -9.39958870e-01
7.68891931e-01 3.88517559e-01 -1.39775440e-01 -7.41712987e-01
-6.26951875e-03 5.58538437e-01 2.89570451e-01 -2.45795861e-01
8.17954600e-01 -5.23863852e-01 -4.74711210e-01 3.14406782e-01
3.38076204e-01 -5.00631407e-02 3.33413273e-01 1.52262794e-02
7.45219469e-01 2.43536904e-01 5.13226569e-01 -5.82822740e-01
9.01094079e-01 2.46913612e-01 5.70212901e-01 7.22639322e-01
2.26742197e-02 2.89199531e-01 2.15887114e-01 -9.47063789e-02
-9.40525413e-01 -1.22646856e+00 -2.29861259e-01 1.56388378e+00
-2.43275478e-01 -4.39155966e-01 -8.05943310e-01 -7.70909190e-01
-9.96968821e-02 1.37487233e+00 -2.12226495e-01 1.33366331e-01
-1.31685913e+00 -8.06372643e-01 1.07475781e+00 6.55231178e-01
1.61235064e-01 -1.75013483e+00 -2.10406631e-01 6.18233144e-01
-2.14374259e-01 -1.32448840e+00 -2.46418521e-01 2.58847117e-01
-6.20322764e-01 -6.96006954e-01 -4.31041420e-01 -1.38944733e+00
6.12876654e-01 -3.10519129e-01 1.36500442e+00 2.03627244e-01
-1.75527502e-02 3.89371425e-01 -6.01259887e-01 -3.96684378e-01
-5.85253477e-01 1.24178283e-01 1.36991562e-02 -2.94206023e-01
7.76700854e-01 -1.19838364e-01 1.20635055e-01 9.45528038e-03
-1.23436272e+00 -5.99708498e-01 6.64685309e-01 4.98496264e-01
6.62009597e-01 -1.84227884e-01 7.31354594e-01 -1.07893467e+00
5.56233048e-01 -3.23933244e-01 -4.23098832e-01 4.62258935e-01
-3.03402394e-01 3.70604277e-01 7.89531708e-01 -5.56081831e-01
-1.23296893e+00 -3.06310296e-01 -6.86089694e-01 3.55349839e-01
-6.54290080e-01 7.09081113e-01 -5.58563352e-01 6.27307743e-02
6.96325541e-01 3.05965900e-01 -4.55183893e-01 -7.79741764e-01
4.67875451e-01 4.42135245e-01 6.50880635e-01 -1.12638474e+00
9.04095888e-01 1.28103197e-01 -4.98486072e-01 -1.19305229e+00
-9.10711288e-01 -3.86378258e-01 -7.39822149e-01 4.16164190e-01
1.11993492e+00 -7.57928431e-01 -4.07413691e-01 3.66112798e-01
-1.38366759e+00 -4.65595990e-01 -4.45901185e-01 3.53243113e-01
-2.27200449e-01 7.24163234e-01 -9.19228017e-01 -5.41490376e-01
-2.94053167e-01 -5.77667475e-01 9.03300405e-01 -1.30374894e-01
-5.37504315e-01 -1.37963223e+00 -1.20158777e-01 5.21675535e-02
3.23022187e-01 -1.02172218e-01 1.76467419e+00 -1.23208487e+00
8.43650941e-03 -2.86515150e-02 -1.37498870e-01 5.78400314e-01
3.00418697e-02 -4.95333493e-01 -7.56376386e-01 -1.82322845e-01
-9.92162451e-02 -4.68110532e-01 7.38609850e-01 -6.68388531e-02
7.65209496e-01 -3.33145201e-01 7.23018050e-02 2.32092664e-01
1.50586641e+00 -3.76419649e-02 6.74599349e-01 4.29368079e-01
4.55484569e-01 7.51287878e-01 5.25510430e-01 -1.36473984e-01
5.26705444e-01 6.44408911e-02 -7.23585039e-02 1.10036032e-02
-2.11494803e-01 -3.41662794e-01 8.14968824e-01 1.05925083e+00
3.81950200e-01 -3.66064429e-01 -1.23531401e+00 9.46038425e-01
-1.26137125e+00 -4.12601054e-01 -6.92269430e-02 2.12832308e+00
1.17240250e+00 1.79090247e-01 -1.61305711e-01 5.40605374e-02
7.07293451e-01 1.66130036e-01 4.11595367e-02 -8.26158643e-01
-6.00743949e-01 9.82972324e-01 4.06563848e-01 6.47442877e-01
-8.83230925e-01 1.84883904e+00 6.38383341e+00 9.67972875e-01
-1.01815748e+00 1.79327264e-01 3.28129500e-01 7.94294000e-01
-5.39471626e-01 1.19271120e-02 -1.11869752e+00 1.08486256e-02
1.31512451e+00 -2.34992225e-02 1.73615426e-01 5.25916934e-01
-7.58598596e-02 -1.33460030e-01 -1.07081556e+00 6.01098001e-01
1.97989345e-01 -1.06104922e+00 5.65107882e-01 -5.13054788e-01
3.17845970e-01 4.67802733e-02 -2.06107587e-01 5.80579042e-01
3.71046752e-01 -1.29162502e+00 9.21376526e-01 3.26434225e-02
7.91654110e-01 -7.01545954e-01 7.22146332e-01 3.22226286e-01
-1.08909142e+00 3.19314867e-01 -6.66517258e-01 4.77338694e-02
2.56053150e-01 -4.37022001e-02 -7.03954995e-01 5.55186033e-01
2.41983354e-01 2.60752463e-03 -1.05424380e+00 8.55483174e-01
-6.80058360e-01 1.14307010e+00 -2.86756843e-01 -1.61984898e-02
5.30034721e-01 7.23217949e-02 3.87362033e-01 1.81764269e+00
3.94986086e-02 2.33968273e-01 3.70014399e-01 4.82240379e-01
5.16732410e-02 7.22459257e-01 -4.34092507e-02 -4.33378339e-01
5.18721342e-01 9.53872323e-01 -9.89605665e-01 -4.66591120e-01
-4.81779873e-01 9.91337597e-01 6.71052337e-01 2.98026830e-01
-4.39430356e-01 -7.21343517e-01 5.24648845e-01 1.36375979e-01
4.63773102e-01 -8.07746112e-01 -3.88638914e-01 -1.18788981e+00
2.14440841e-02 -9.49537337e-01 6.34254158e-01 -6.42849922e-01
-1.49574029e+00 8.63217473e-01 2.14961767e-02 -4.27221388e-01
-1.31275415e-01 -1.24255276e+00 -3.47787023e-01 1.16751826e+00
-1.76386774e+00 -1.27450836e+00 4.21129763e-01 4.86066043e-01
5.51319957e-01 2.70641334e-02 1.08421588e+00 1.37545943e-01
-3.51671457e-01 7.61862695e-01 -3.73668462e-01 6.15176320e-01
6.56012893e-01 -1.32955110e+00 5.92445970e-01 1.00838315e+00
6.11020803e-01 9.32118893e-01 4.94977772e-01 -6.57353818e-01
-1.08670247e+00 -1.13373399e+00 1.69698620e+00 -6.27797067e-01
8.17552447e-01 -4.14218247e-01 -1.43257654e+00 8.86929452e-01
4.98056114e-01 -1.82614297e-01 8.78418982e-01 1.69220299e-01
-5.59020340e-01 3.79987776e-01 -1.16700435e+00 5.02390146e-01
8.34283292e-01 -6.42138720e-01 -1.54007828e+00 3.23097974e-01
8.16084146e-01 6.39880970e-02 -8.93709719e-01 2.89527744e-01
-1.58259105e-02 -4.14945334e-01 8.28618467e-01 -1.22035396e+00
-9.62489992e-02 -2.85543561e-01 -2.88063496e-01 -1.26666629e+00
-2.89262235e-01 -3.85910064e-01 5.58170438e-01 1.43735647e+00
6.96384847e-01 -6.80165887e-01 5.37309825e-01 3.91589075e-01
-2.96441853e-01 -9.63926166e-02 -8.13358188e-01 -1.20936954e+00
8.53353858e-01 -8.94835591e-01 4.34409380e-01 1.12715459e+00
-2.35699397e-02 5.48628628e-01 3.86329502e-01 3.29740167e-01
3.18410337e-01 -1.74363360e-01 -3.85577343e-02 -1.08350801e+00
-1.65537536e-01 -4.14604068e-01 -1.78894684e-01 -8.25662911e-01
9.02179360e-01 -1.20525646e+00 2.72164166e-01 -1.52923882e+00
-4.54819262e-01 -6.86821043e-01 -3.33751112e-01 8.74678373e-01
-4.12758648e-01 3.81188989e-01 1.19912587e-01 -3.13455135e-01
-4.17282075e-01 2.38031343e-01 6.33125246e-01 2.80951560e-01
-1.01070493e-01 -4.81915623e-01 -6.11385465e-01 9.77982938e-01
8.59894633e-01 -4.83340979e-01 1.87100023e-01 -8.24940860e-01
3.68733943e-01 -3.07508290e-01 4.31755409e-02 -7.47224987e-01
-8.95538330e-02 -3.63671035e-01 4.79936600e-01 -3.77215236e-01
-6.74831262e-03 -5.04790306e-01 -5.41715741e-01 3.63901049e-01
-3.00544739e-01 4.38749582e-01 7.36033916e-01 -1.41689941e-01
-3.10159534e-01 -6.63742125e-01 7.79132843e-01 -4.84508961e-01
-1.21437776e+00 -6.70424663e-04 -7.57633328e-01 5.91555655e-01
5.88831782e-01 -3.85505617e-01 -8.80822986e-02 1.55359477e-01
-8.48534286e-01 6.73399195e-02 2.09447637e-01 6.31958485e-01
5.87405860e-01 -9.59047675e-01 -9.02466536e-01 2.59177566e-01
1.42344981e-01 -3.09507728e-01 -2.39547700e-01 2.54452497e-01
-6.67077303e-01 7.83712983e-01 -6.52913079e-02 6.29402548e-02
-1.06713521e+00 7.98031151e-01 1.10953547e-01 -2.34494522e-01
-2.75907874e-01 1.03734267e+00 -1.22855818e-02 -8.93389344e-01
-6.54380247e-02 -4.61238384e-01 -3.06421340e-01 1.08365521e-01
4.14867699e-01 -8.61115102e-03 3.11420143e-01 -9.41518843e-01
-6.12834096e-01 2.92881668e-01 -1.11259855e-01 -2.16946110e-01
1.12853885e+00 -8.43706727e-02 -4.52815592e-01 3.51393223e-01
7.14648306e-01 7.30947196e-01 -3.58432263e-01 -4.82573867e-01
1.00835586e+00 8.98334906e-02 -4.00341988e-01 -8.11704993e-01
-2.91841954e-01 8.14977109e-01 -1.29352182e-01 -8.19461420e-02
6.94670856e-01 1.70817688e-01 9.76725817e-01 6.82870507e-01
3.22375953e-01 -1.12336469e+00 -2.28507340e-01 1.20600069e+00
6.40210986e-01 -7.36464679e-01 -4.29395109e-01 -5.12452364e-01
-5.96301496e-01 8.58072758e-01 6.21601820e-01 -2.77920783e-01
4.71866727e-01 1.52691513e-01 2.56709188e-01 1.70790598e-01
-5.09233296e-01 -5.20148695e-01 3.21133345e-01 5.04016519e-01
9.01557326e-01 -3.59520130e-02 -7.50278950e-01 8.12600851e-01
-5.24331152e-01 -6.62677765e-01 3.58533859e-01 1.07565677e+00
-7.47184694e-01 -1.46182668e+00 -5.74171245e-01 1.14753306e-01
-8.83809626e-01 -7.93656766e-01 -3.91713232e-01 9.93219197e-01
4.02639002e-01 8.77333105e-01 3.10185134e-01 2.28052348e-01
3.60079050e-01 7.62493730e-01 5.95912039e-01 -1.32131600e+00
-1.03950334e+00 -1.87327743e-01 4.77205455e-01 -2.38269374e-01
-1.65505230e-01 -5.61909676e-01 -1.50569773e+00 1.26198769e-01
-1.88664958e-01 5.93629062e-01 4.69930291e-01 1.23366153e+00
-9.73097458e-02 1.29539490e-01 -5.03443405e-02 -3.98393899e-01
-5.44643581e-01 -9.66453671e-01 -5.89575171e-01 4.50376481e-01
-9.01310667e-02 -9.94554609e-02 -1.74705878e-01 6.02387339e-02] | [10.49616527557373, 9.983516693115234] |
857efeff-a703-421c-a830-89128a6daa77 | sentence-aware-adversarial-meta-learning-for | null | null | https://aclanthology.org/2022.coling-1.428 | https://aclanthology.org/2022.coling-1.428.pdf | Sentence-aware Adversarial Meta-Learning for Few-Shot Text Classification | Meta-learning has emerged as an effective approach for few-shot text classification. However, current studies fail to realize the importance of the semantic interaction between sentence features and neglect to enhance the generalization ability of the model to new tasks. In this paper, we integrate an adversarial network architecture into the meta-learning system and leverage cost-effective modules to build a novel few-shot classification framework named SaAML. Significantly, our approach can exploit the temporal convolutional network to encourage more discriminative representation learning and explore the attention mechanism to promote more comprehensive feature expression, thus resulting in better adaptation for new classes. Through a series of experiments on four benchmark datasets, we demonstrate that our new framework acquires considerable superiority over state-of-the-art methods in all datasets, increasing the performance of 1-shot classification and 5-shot classification by 7.15% and 2.89%, respectively. | ['Diwen Dong', 'Bo Liu', 'Xiaoyuan Liu', 'Suhe Wang'] | null | null | null | null | coling-2022-10 | ['few-shot-text-classification'] | ['natural-language-processing'] | [ 1.27942830e-01 -1.69281751e-01 -3.31326127e-01 -4.69319403e-01
-6.46233141e-01 -2.86623230e-03 7.64073491e-01 -2.55442336e-02
-5.80024958e-01 6.71441376e-01 1.21586204e-01 6.37966394e-02
9.50089768e-02 -9.33717072e-01 -4.38060462e-01 -4.43477899e-01
3.07978094e-01 -1.62859529e-01 3.71307284e-01 -6.40339077e-01
2.07382545e-01 -5.42831048e-02 -1.55154109e+00 4.46871519e-01
1.00360942e+00 1.02102447e+00 -8.16988721e-02 3.61669242e-01
-4.13282931e-01 8.46410036e-01 -8.03217769e-01 -5.29818952e-01
-6.69381171e-02 -5.31602025e-01 -6.82912111e-01 -2.32557192e-01
-2.52455715e-02 -2.72072226e-01 -7.09494174e-01 9.67890739e-01
6.60904050e-01 7.66955137e-01 5.20375550e-01 -1.04003787e+00
-1.01330805e+00 8.54985237e-01 -5.45195937e-01 4.04380381e-01
2.57338941e-01 3.70683670e-01 9.14077520e-01 -1.06732047e+00
3.96537066e-01 9.31535482e-01 5.88031054e-01 9.48020756e-01
-8.25502038e-01 -8.32670152e-01 2.55559713e-01 3.56140971e-01
-1.12135231e+00 -5.61563253e-01 1.03002453e+00 9.83790681e-03
1.16998267e+00 1.08860575e-01 3.06999326e-01 1.46521366e+00
3.35628659e-01 1.07232118e+00 9.43695605e-01 -5.13684511e-01
3.23129535e-01 6.25164062e-02 4.28012431e-01 6.63844109e-01
-2.45536328e-03 -1.30460979e-02 -5.78720629e-01 1.09628133e-01
6.61524162e-02 5.21920443e-01 -1.02053910e-01 1.76202715e-03
-7.66219079e-01 9.46104765e-01 6.90942407e-01 6.61666274e-01
-1.90549091e-01 8.56865868e-02 7.89167643e-01 2.07572445e-01
9.70178246e-01 5.52622318e-01 -3.97743702e-01 -4.08443272e-01
-7.16266215e-01 1.30764628e-02 3.98877889e-01 7.62298346e-01
4.92320478e-01 4.03472900e-01 -6.51632965e-01 1.06415737e+00
-1.79482594e-01 1.29621744e-01 1.05653417e+00 -4.79851544e-01
5.03355265e-01 9.04274821e-01 -3.52364480e-01 -7.18242586e-01
-4.76298362e-01 -6.98227763e-01 -8.27271879e-01 -1.93455353e-01
-2.62423396e-01 -2.67777860e-01 -1.05752504e+00 1.68901229e+00
7.55756646e-02 5.35388768e-01 2.38831908e-01 4.39365894e-01
1.06760168e+00 6.66981041e-01 3.38968575e-01 -1.52783662e-01
1.27917719e+00 -1.36470830e+00 -7.21887112e-01 -4.42882627e-01
8.10820341e-01 -2.65148073e-01 1.31350803e+00 -1.81706194e-02
-9.72215950e-01 -7.35544503e-01 -1.28615463e+00 6.82371780e-02
-7.14751184e-01 -4.29409564e-01 8.18001330e-01 5.55976212e-01
-3.74833882e-01 8.03680420e-01 -6.55945897e-01 -2.37727374e-01
8.11711788e-01 1.62979335e-01 -1.77073225e-01 -1.72842115e-01
-1.66077995e+00 8.41572046e-01 6.03455126e-01 -4.10269052e-01
-5.58653891e-01 -8.47258151e-01 -8.86261761e-01 3.86995286e-01
3.12597066e-01 -7.37721860e-01 1.34777164e+00 -8.98402095e-01
-1.65198958e+00 5.61758876e-01 5.05972952e-02 -4.20415670e-01
4.64820534e-01 -1.03321195e-01 -6.72454715e-01 1.90827355e-01
-3.21328640e-02 5.69427669e-01 9.29939449e-01 -8.08162689e-01
-6.62322760e-01 -2.64515847e-01 2.36125529e-01 6.66470975e-02
-1.10052979e+00 -1.82532102e-01 -1.22015908e-01 -7.93012142e-01
-3.95651400e-01 -6.39701605e-01 -1.86461955e-01 -3.42419952e-01
-5.61340004e-02 -4.67666656e-01 8.86001766e-01 -3.54333401e-01
1.37413573e+00 -2.15531659e+00 1.19206183e-01 -5.47749996e-01
5.80990948e-02 7.33688772e-01 -2.47175664e-01 4.02623534e-01
1.94455627e-02 1.42556354e-01 -1.59854278e-01 -5.02402008e-01
-1.47818588e-02 8.67888778e-02 -3.70363861e-01 1.55679613e-01
2.76631504e-01 1.51836002e+00 -9.96583700e-01 -2.68277198e-01
2.37394333e-01 3.52595001e-01 -4.80611622e-01 2.22048953e-01
-1.21660508e-01 -1.74634121e-02 -6.35166168e-01 4.93535578e-01
5.15141726e-01 -2.01295912e-01 -2.19048947e-01 1.17217571e-01
2.27986321e-01 1.22311361e-01 -4.67059076e-01 2.02570796e+00
-6.85859323e-01 3.91520619e-01 -6.29729390e-01 -1.21644843e+00
8.99957120e-01 2.68093079e-01 3.64385575e-01 -9.62635279e-01
4.30707157e-01 -1.41771525e-01 4.54021012e-03 -4.92321432e-01
5.62412500e-01 -3.34111512e-01 -2.40565404e-01 4.44748521e-01
4.60261524e-01 3.87133926e-01 7.68811181e-02 1.71344340e-01
1.17332685e+00 -5.30532934e-02 3.84073734e-01 -4.45571356e-02
4.89195168e-01 -2.64774650e-01 3.36979747e-01 7.91960835e-01
-5.01559973e-01 2.99600840e-01 4.64199968e-02 -3.87993157e-01
-7.72046089e-01 -6.23508215e-01 -1.86745554e-01 1.55975592e+00
6.40880018e-02 -3.10450464e-01 -6.34124756e-01 -9.88790214e-01
-5.13653271e-02 1.08185208e+00 -1.01987183e+00 -1.08006883e+00
-3.36331010e-01 -1.00628126e+00 4.44757342e-01 9.10152674e-01
8.36240411e-01 -1.03130531e+00 -6.15931273e-01 4.47270066e-01
-9.88302231e-02 -1.04516387e+00 -5.26111603e-01 1.53278336e-01
-9.68095720e-01 -6.54966593e-01 -7.65478849e-01 -7.38520384e-01
3.66861314e-01 6.46992326e-01 7.21425951e-01 1.68070450e-01
-3.03291500e-01 2.75009125e-01 -8.70612264e-01 -3.51760983e-01
-3.16820413e-01 5.03735363e-01 -2.10531279e-02 -2.04119515e-02
6.23403847e-01 -5.59144258e-01 -5.61012447e-01 6.57176524e-02
-1.03755915e+00 -5.95151260e-02 3.18172365e-01 1.39213014e+00
-4.83030528e-02 -2.28558891e-02 1.13284826e+00 -1.14418173e+00
7.65882254e-01 -6.46040976e-01 7.76471123e-02 1.92938417e-01
-9.20659363e-01 -4.25515212e-02 9.59469259e-01 -5.92040956e-01
-1.27653635e+00 -4.71285254e-01 -1.64349973e-01 -5.18792152e-01
1.63800865e-01 6.72507584e-01 -6.09019399e-02 1.35759143e-02
7.51830816e-01 4.75983948e-01 -8.97815675e-02 -2.93935776e-01
5.33904970e-01 9.19992387e-01 2.45891526e-01 -2.77228177e-01
7.43248641e-01 4.21239138e-01 -3.08730096e-01 -6.29901290e-01
-1.29128516e+00 -5.35706162e-01 -6.13610148e-01 9.18854177e-02
6.52731895e-01 -9.58377659e-01 -3.29316974e-01 4.97132063e-01
-8.36850822e-01 -7.39371702e-02 -3.56923074e-01 1.01355530e-01
-3.91424268e-01 3.04456770e-01 -6.86949372e-01 -6.18559361e-01
-8.27811122e-01 -8.04400623e-01 6.91688180e-01 4.83635485e-01
-5.37614804e-03 -1.03573573e+00 5.36971241e-02 3.58855873e-01
7.41888285e-01 -5.62238656e-02 7.92621195e-01 -1.02568126e+00
-2.13658977e-02 -5.31833708e-01 -5.03225848e-02 2.32100904e-01
1.35036781e-01 -2.22702324e-01 -1.32236493e+00 -4.61510330e-01
7.64318556e-02 -5.73604047e-01 1.27752137e+00 -1.58408601e-02
1.17886150e+00 -1.88388065e-01 -4.19721991e-01 7.05997050e-01
1.28650618e+00 3.26297253e-01 6.98973954e-01 4.64899749e-01
5.20597816e-01 3.27347875e-01 6.26017034e-01 5.74746728e-01
2.31599763e-01 5.93308449e-01 2.63766736e-01 1.72485664e-01
1.34980865e-02 -3.11028630e-01 3.47241729e-01 9.36493337e-01
7.68879130e-02 -1.89439669e-01 -6.21079564e-01 2.35379681e-01
-1.89585364e+00 -1.15284050e+00 5.91547132e-01 1.58724332e+00
7.07426786e-01 4.41448092e-01 -5.44215143e-02 -3.96725349e-02
6.28251493e-01 5.83766878e-01 -8.16467881e-01 -3.91339481e-01
7.04370067e-02 4.39218014e-01 -6.93446398e-02 -6.96903840e-02
-1.25103247e+00 1.18207037e+00 5.95583582e+00 9.61032271e-01
-1.05971420e+00 3.71160716e-01 6.00767314e-01 -2.53156841e-01
-1.04730055e-01 -3.31561446e-01 -8.69690299e-01 5.34951925e-01
1.15126443e+00 -5.43079913e-01 2.22846434e-01 1.10597920e+00
-3.21513832e-01 5.25134444e-01 -8.44342530e-01 7.20345497e-01
4.73895252e-01 -1.36972964e+00 1.76590651e-01 -3.93527806e-01
9.83287752e-01 -2.24171113e-02 2.43797913e-01 1.19710732e+00
1.66291147e-01 -9.82780397e-01 2.97546387e-01 4.75598276e-01
7.55048871e-01 -9.96806502e-01 8.48997474e-01 4.62709814e-01
-9.48808432e-01 -3.65918010e-01 -6.68859422e-01 -2.55730152e-01
-2.68604588e-02 5.76503538e-02 -5.40382624e-01 6.03317857e-01
4.79267806e-01 9.07690763e-01 -8.16511631e-01 6.74645305e-01
-1.24054201e-01 4.59751099e-01 3.54347318e-01 -4.32049721e-01
4.42253798e-01 3.17749739e-01 3.29798073e-01 1.09835863e+00
9.16152298e-02 2.80301452e-01 1.12737775e-01 6.51617110e-01
-4.80238438e-01 2.55854309e-01 -5.61588764e-01 -2.59907842e-01
3.46954614e-01 1.34420013e+00 -2.98615426e-01 -4.85574126e-01
-7.45351553e-01 1.10708892e+00 6.59410238e-01 2.92994052e-01
-9.04405713e-01 -9.30821061e-01 5.48374355e-01 -3.83159369e-01
4.45345998e-01 2.19462052e-01 -1.74261227e-01 -1.63589752e+00
-1.94078982e-01 -8.06513369e-01 4.66776192e-01 -5.39917171e-01
-1.55755138e+00 6.83915615e-01 -2.31884703e-01 -1.20384061e+00
-3.71371388e-01 -5.70092678e-01 -1.04482853e+00 6.41820848e-01
-1.50152314e+00 -1.36209095e+00 -3.59142184e-01 4.69850272e-01
8.85987878e-01 -6.25749350e-01 9.91221368e-01 8.38040039e-02
-8.14350307e-01 1.07572472e+00 4.50650454e-01 2.63216853e-01
6.57512248e-01 -9.21237707e-01 5.74209690e-01 8.15068483e-01
8.13315585e-02 6.01555228e-01 4.46882635e-01 -3.19131970e-01
-1.32337463e+00 -1.23487484e+00 4.72843081e-01 -5.22373855e-01
7.55359232e-01 -3.46287280e-01 -1.24387479e+00 6.88729465e-01
3.19817960e-01 1.71758115e-01 9.22409832e-01 9.41437855e-02
-5.50086737e-01 -1.81436285e-01 -1.01930547e+00 7.53009856e-01
1.06265604e+00 -5.17006516e-01 -8.83341670e-01 5.01326919e-02
1.25320196e+00 -1.58618033e-01 -8.52623641e-01 4.56849337e-01
4.37082410e-01 -8.03371847e-01 8.76959920e-01 -1.19222939e+00
8.89759481e-01 3.75143439e-01 -1.69177234e-01 -1.52092910e+00
-5.76461732e-01 -3.08701545e-01 -5.24181664e-01 1.36327040e+00
2.82970756e-01 -7.75642455e-01 6.21731579e-01 4.30337608e-01
-3.55363160e-01 -1.11801922e+00 -1.04694080e+00 -7.98858404e-01
5.45198262e-01 -2.88487256e-01 6.28573120e-01 1.28832293e+00
5.40266991e-01 7.03578293e-01 -4.63076055e-01 -4.57363725e-01
4.50635612e-01 2.23305210e-01 6.86819613e-01 -1.05394578e+00
-3.68106067e-01 -6.29441381e-01 -4.64991331e-01 -7.55457580e-01
6.92756236e-01 -1.14284921e+00 -2.10698992e-01 -1.27652097e+00
5.25212407e-01 7.19691887e-02 -9.43509758e-01 5.47623754e-01
-8.09345484e-01 1.40705675e-01 2.04171896e-01 5.35112917e-02
-8.16341162e-01 1.23545682e+00 1.27435184e+00 -5.14355540e-01
-5.63029908e-02 5.76181747e-02 -1.00945473e+00 5.76335073e-01
8.69086206e-01 -4.07420933e-01 -4.51887429e-01 -2.53049642e-01
-1.95995241e-01 -1.42600268e-01 1.13559272e-02 -9.76979256e-01
2.27785707e-01 -1.29559815e-01 6.25630915e-01 -2.56270558e-01
4.00775731e-01 -4.84626085e-01 -5.61938882e-01 6.25039399e-01
-4.92244124e-01 -2.95914888e-01 3.98998797e-01 7.81124651e-01
-1.81777850e-01 -6.02178454e-01 8.74215245e-01 -1.31678998e-01
-1.02283311e+00 5.35554945e-01 -5.31424955e-03 2.39538208e-01
1.25912905e+00 4.21042442e-02 -6.22561812e-01 -9.45419073e-02
-4.50423181e-01 2.39517346e-01 3.19107085e-01 8.68036568e-01
6.94743216e-01 -1.35179090e+00 -5.25550723e-01 2.22825810e-01
5.71366787e-01 -5.48549354e-01 7.08141267e-01 5.05024493e-01
-7.15372432e-03 4.07124162e-01 -2.64103770e-01 -2.31080562e-01
-8.63592505e-01 9.16571200e-01 2.97657996e-01 -3.59550744e-01
-8.49073946e-01 8.15157056e-01 -1.68201867e-02 -2.69426674e-01
1.57407857e-02 5.26395626e-02 -3.57192725e-01 1.86404251e-02
8.06518555e-01 3.48936617e-01 -8.47441852e-02 -3.17390621e-01
-2.51434743e-01 2.11004555e-01 -6.07283115e-01 2.16421768e-01
1.31531775e+00 -1.52787976e-02 3.72208416e-01 6.83885038e-01
1.36715066e+00 -3.99094224e-01 -1.30799019e+00 -4.84566927e-01
-4.11353968e-02 -4.01391953e-01 1.08659826e-01 -8.42469454e-01
-9.47829187e-01 1.12230325e+00 4.87675399e-01 9.71354991e-02
9.72637951e-01 -2.09881678e-01 1.18473411e+00 4.93456483e-01
2.17198223e-01 -1.06664026e+00 6.04856849e-01 7.86003768e-01
4.49737936e-01 -1.51476979e+00 -1.40657291e-01 1.15569711e-01
-7.11620271e-01 1.05098486e+00 9.27973270e-01 -1.06654897e-01
6.53825283e-01 -1.51714772e-01 -8.40005577e-02 4.80313078e-02
-1.16915441e+00 -7.98460320e-02 3.44102979e-01 3.96707445e-01
4.11078185e-01 -1.17192470e-01 -2.48677447e-01 1.13147712e+00
1.28522709e-01 1.48645937e-01 3.67736906e-01 1.02965999e+00
-7.06553161e-01 -1.01210082e+00 4.36573714e-01 6.29406989e-01
-5.62968910e-01 -3.34513396e-01 -1.64835714e-03 7.13285685e-01
-2.22760186e-01 7.99597025e-01 4.53022532e-02 -6.55118585e-01
4.83609259e-01 5.82828283e-01 3.65711898e-01 -9.48057413e-01
-6.47556007e-01 -4.16021049e-01 -1.47511467e-01 -1.87215671e-01
-2.94131845e-01 -1.76609844e-01 -1.15092230e+00 -2.58645743e-01
-5.54644167e-01 1.00060143e-01 3.43749166e-01 1.17084610e+00
5.72568238e-01 1.00447857e+00 9.21950459e-01 -4.65632498e-01
-1.06461227e+00 -1.25680435e+00 -4.77392733e-01 5.96301973e-01
-5.41242026e-03 -8.10930729e-01 -4.02047843e-01 -2.98915774e-01] | [10.224925994873047, 3.4853012561798096] |
14a92cf1-bc73-4f35-893f-6468968bb1f6 | dense-3d-object-reconstruction-from-a-single | 1802.00411 | null | http://arxiv.org/abs/1802.00411v2 | http://arxiv.org/pdf/1802.00411v2.pdf | Dense 3D Object Reconstruction from a Single Depth View | In this paper, we propose a novel approach, 3D-RecGAN++, which reconstructs
the complete 3D structure of a given object from a single arbitrary depth view
using generative adversarial networks. Unlike existing work which typically
requires multiple views of the same object or class labels to recover the full
3D geometry, the proposed 3D-RecGAN++ only takes the voxel grid representation
of a depth view of the object as input, and is able to generate the complete 3D
occupancy grid with a high resolution of 256^3 by recovering the
occluded/missing regions. The key idea is to combine the generative
capabilities of autoencoders and the conditional Generative Adversarial
Networks (GAN) framework, to infer accurate and fine-grained 3D structures of
objects in high-dimensional voxel space. Extensive experiments on large
synthetic datasets and real-world Kinect datasets show that the proposed
3D-RecGAN++ significantly outperforms the state of the art in single view 3D
object reconstruction, and is able to reconstruct unseen types of objects. | ['Hongkai Wen', 'Andrew Markham', 'Niki Trigoni', 'Bo Yang', 'Stefano Rosa'] | 2018-02-01 | null | null | null | null | ['3d-object-reconstruction'] | ['computer-vision'] | [ 8.40115845e-02 3.49055558e-01 4.45854694e-01 -1.56834841e-01
-7.75933802e-01 -6.42496109e-01 5.21488309e-01 -6.36416733e-01
-1.02756359e-01 8.02550972e-01 1.32849067e-01 8.41231868e-02
2.13229343e-01 -1.40528345e+00 -1.17876697e+00 -7.74128556e-01
3.49840552e-01 1.17168105e+00 1.54256016e-01 1.67928785e-01
-3.29470150e-02 9.98051822e-01 -1.56412804e+00 1.02004342e-01
3.13460708e-01 1.02491224e+00 1.38608515e-01 7.53618181e-01
-9.95063931e-02 7.41625488e-01 -4.54329371e-01 -1.74432769e-01
6.75181150e-01 -2.62747765e-01 -4.78440762e-01 3.02353948e-01
5.89624822e-01 -8.62803638e-01 -8.24449897e-01 7.45178342e-01
7.22613096e-01 5.99860251e-02 8.65324914e-01 -9.94886339e-01
-4.91952866e-01 1.09851435e-01 -4.04934257e-01 -1.92890629e-01
5.28701723e-01 2.83799410e-01 3.18071663e-01 -9.20651019e-01
9.47863638e-01 1.47203767e+00 5.87128758e-01 6.39306843e-01
-1.26808393e+00 -6.34683669e-01 -3.86575982e-02 -8.95130187e-02
-1.45864069e+00 -5.78764342e-02 1.06124008e+00 -6.01560056e-01
8.62032175e-01 5.35071529e-02 8.71238053e-01 1.43380058e+00
2.64005154e-01 4.35195327e-01 1.53340387e+00 -5.93075790e-02
5.07965147e-01 -3.02201003e-01 -5.84252775e-01 7.17766762e-01
1.08847748e-02 4.09986436e-01 -4.19779420e-01 -1.58253700e-01
1.51329088e+00 4.65351939e-01 -2.31993377e-01 -7.80349433e-01
-1.26653624e+00 8.46892536e-01 6.03306651e-01 -1.94709450e-01
-6.44686401e-01 2.54475206e-01 6.38674386e-03 -1.34668097e-01
5.17663360e-01 3.73437889e-02 -8.78142267e-02 1.31156400e-01
-4.29318011e-01 6.76995814e-01 6.56363726e-01 1.06989205e+00
8.66313756e-01 3.80723327e-01 -1.12535030e-01 1.87319502e-01
3.23042214e-01 1.02157366e+00 2.21345916e-01 -1.22917449e+00
3.00439000e-01 5.06444514e-01 9.50848088e-02 -5.35317481e-01
-4.12258431e-02 -1.04305558e-01 -1.01907384e+00 8.16805899e-01
1.87900394e-01 -3.73382531e-02 -1.43903005e+00 1.51904690e+00
8.56577933e-01 2.24322453e-01 1.00303479e-01 9.48919415e-01
1.21949613e+00 6.69751525e-01 -3.39882284e-01 2.13880450e-01
1.00091779e+00 -6.01055264e-01 -3.79224032e-01 -3.35728228e-01
-4.24650520e-01 -4.09547418e-01 6.71380460e-01 1.46712169e-01
-1.36469603e+00 -8.05595636e-01 -9.36629534e-01 -9.35665369e-02
-2.40923166e-01 -5.17844200e-01 5.69126964e-01 4.50454295e-01
-7.87959933e-01 3.95100117e-01 -1.14066982e+00 9.96992067e-02
6.83084071e-01 1.16761558e-01 -7.14582026e-01 -6.29135549e-01
-7.27740586e-01 5.91766834e-01 5.33315659e-01 -6.15061596e-02
-1.71152699e+00 -7.47045577e-01 -1.04066634e+00 -2.17722461e-01
3.87100339e-01 -1.31241524e+00 7.60634184e-01 -1.93071172e-01
-1.42775834e+00 8.88004601e-01 1.11738510e-01 -1.89015314e-01
7.35745430e-01 -8.86876807e-02 2.78115481e-01 1.62344038e-01
6.62226006e-02 6.58797801e-01 1.05416119e+00 -1.88439262e+00
-4.02382463e-02 -1.02759445e+00 1.05260745e-01 3.69652927e-01
6.51811004e-01 -7.47278571e-01 -3.45211864e-01 -5.75291038e-01
6.26013935e-01 -7.75353849e-01 -3.15856725e-01 9.34179276e-02
-5.78703821e-01 3.43749255e-01 8.72693241e-01 -7.35369384e-01
1.36299543e-02 -1.81473553e+00 6.37049615e-01 3.61773837e-03
3.45152676e-01 -1.67706087e-01 1.42677873e-01 2.61232167e-01
1.19940653e-01 -1.24159589e-01 -4.52256233e-01 -5.12901366e-01
-4.46066186e-02 6.65670812e-01 -2.67706990e-01 6.34210885e-01
-1.50671467e-01 1.29174948e+00 -7.87662566e-01 -2.54432917e-01
7.23578453e-01 9.53892410e-01 -5.64064860e-01 7.97040701e-01
-3.65072519e-01 8.98830235e-01 -5.38738906e-01 7.53210545e-01
1.03080559e+00 -1.31345227e-01 -7.14151114e-02 -6.41039461e-02
3.71138245e-01 -5.87645434e-02 -1.12987173e+00 2.07042575e+00
-4.07994717e-01 3.02120354e-02 9.27364081e-02 -7.92895973e-01
9.29558098e-01 2.70892203e-01 4.82427150e-01 -5.45016706e-01
6.91712648e-02 -5.16586155e-02 -5.32604516e-01 -3.81649226e-01
1.97164446e-01 -5.97748816e-01 -2.14062542e-01 4.93967861e-01
2.13717088e-01 -7.97736645e-01 -7.75875032e-01 7.75654742e-04
1.10706747e+00 5.77641010e-01 2.24680752e-01 3.98829699e-01
1.26700386e-01 -2.00075120e-01 3.16690832e-01 7.59401858e-01
2.51209140e-01 1.21773350e+00 1.19907139e-02 -5.04147470e-01
-1.47823322e+00 -1.67915356e+00 1.01033844e-01 1.62976399e-01
2.16863424e-01 3.33655357e-01 -7.22801685e-01 -7.79606402e-01
2.47821733e-01 6.36374116e-01 -1.02870488e+00 4.12402414e-02
-6.27395153e-01 -3.15401942e-01 3.57590318e-01 6.26818955e-01
6.03053987e-01 -1.35198402e+00 -8.52032542e-01 1.14922315e-01
-2.14036062e-01 -1.20027506e+00 8.87478143e-02 2.45233551e-01
-9.71732736e-01 -1.14054310e+00 -7.67372727e-01 -4.16118950e-01
8.23163629e-01 8.04307163e-02 1.25774121e+00 -3.78122449e-01
-3.49820942e-01 4.58666801e-01 -2.47270525e-01 -3.09191763e-01
-5.39209247e-01 -2.44235113e-01 -1.19374894e-01 -1.65857196e-01
-6.61912784e-02 -1.11012411e+00 -5.13278723e-01 1.93520203e-01
-1.06264877e+00 1.55910194e-01 4.57547396e-01 8.46993685e-01
1.22782481e+00 1.85170129e-01 1.42071769e-01 -8.98821414e-01
-2.00380057e-01 -4.69262093e-01 -4.99825031e-01 -1.98520333e-01
-1.93347365e-01 -5.70764281e-02 5.13165534e-01 -3.36711556e-01
-9.62347865e-01 2.47088507e-01 -5.96939921e-01 -1.11644697e+00
-6.40325010e-01 -1.82202280e-01 -5.70411801e-01 -3.19694877e-02
4.83274013e-01 6.56171978e-01 -9.06028003e-02 -5.66039205e-01
3.82907540e-01 1.84432656e-01 8.09538245e-01 -5.53391933e-01
1.15003264e+00 1.06294167e+00 1.80592582e-01 -4.08382773e-01
-7.79793918e-01 -1.25444770e-01 -9.53153551e-01 -1.63397729e-01
1.17030334e+00 -1.21306407e+00 -5.40377617e-01 6.80175841e-01
-1.15955305e+00 -4.70374674e-01 -9.18099523e-01 4.07926261e-01
-1.09175181e+00 1.53214112e-01 -3.48824292e-01 -6.68231845e-01
-2.44247675e-01 -1.14221823e+00 1.65770769e+00 2.51438282e-02
2.63299197e-01 -7.98262537e-01 1.60230368e-01 5.88242412e-01
4.28513885e-02 1.17451465e+00 8.09462368e-01 -1.20239615e-01
-1.16177940e+00 -1.47525698e-01 1.70474574e-01 3.61537695e-01
1.09542429e-01 -5.76131880e-01 -1.16598856e+00 -3.22927952e-01
3.36030304e-01 -4.73376125e-01 5.79672813e-01 4.96301323e-01
1.37214446e+00 -2.18344256e-01 -1.11703113e-01 1.01039600e+00
1.69771802e+00 2.01068148e-01 1.07027304e+00 -1.06959134e-01
1.05163455e+00 1.20084725e-01 4.21441555e-01 6.15712523e-01
3.15649629e-01 5.47518313e-01 1.24419928e+00 -3.43621662e-03
-2.77674228e-01 -7.43263483e-01 -3.45178880e-02 4.71989155e-01
-3.61167938e-01 -4.74440694e-01 -6.90615833e-01 4.20568556e-01
-1.39026535e+00 -9.30675745e-01 3.30216140e-01 2.10702085e+00
2.89050162e-01 -2.77709644e-02 -1.63798407e-01 1.37040153e-01
4.74795878e-01 3.59338254e-01 -9.75362480e-01 1.46750227e-01
-1.47461090e-02 6.25383735e-01 4.02787179e-01 5.35372913e-01
-6.04594290e-01 7.69092858e-01 6.29194593e+00 6.35411561e-01
-8.03637862e-01 2.33115405e-01 4.49563682e-01 -5.70327528e-02
-5.94055355e-01 -2.28152871e-01 -6.07310116e-01 2.80650169e-01
4.28207397e-01 2.36132577e-01 5.78255415e-01 9.93175209e-01
-3.38298500e-01 -4.73176837e-02 -1.07877576e+00 1.17522669e+00
3.65579486e-01 -1.36027408e+00 2.36636460e-01 4.55019385e-01
8.92573059e-01 1.15724370e-01 -9.67617184e-02 1.85122967e-01
5.27022600e-01 -1.37799740e+00 8.57602119e-01 8.45555365e-01
1.11502671e+00 -8.90843630e-01 6.31122828e-01 8.12757671e-01
-8.16750765e-01 4.38801706e-01 -4.55245674e-01 1.22752197e-01
3.28394383e-01 4.25793737e-01 -9.10175204e-01 8.22955966e-01
9.08433855e-01 4.47508067e-01 -2.06912503e-01 4.52859610e-01
-3.68026018e-01 1.96109563e-01 -4.14081067e-01 5.42275608e-01
-8.19745734e-02 -7.76931047e-02 6.67670488e-01 4.09249812e-01
4.89826590e-01 3.93381715e-01 1.23071000e-01 1.23752487e+00
-1.72815681e-01 -5.52024305e-01 -1.06088674e+00 2.56543338e-01
2.80567586e-01 9.27800417e-01 -5.95815122e-01 -2.89949626e-01
1.42946746e-02 1.23824978e+00 2.62260675e-01 2.11258203e-01
-8.76605749e-01 9.49838161e-02 5.03900409e-01 4.59832221e-01
7.28304148e-01 -2.34740242e-01 -2.56365061e-01 -1.04549432e+00
6.55698031e-02 -7.09731936e-01 5.80356829e-03 -1.29717934e+00
-1.24747074e+00 6.60318971e-01 -2.52517220e-02 -1.19044411e+00
-4.63078141e-01 -4.57283944e-01 -1.74813062e-01 9.87166643e-01
-1.33137298e+00 -1.34810889e+00 -8.90497386e-01 8.49264503e-01
5.28029978e-01 -1.02479689e-01 1.07550585e+00 -8.17945674e-02
2.54466325e-01 4.90664095e-02 -2.64810976e-02 5.19143045e-02
1.66447148e-01 -1.21283782e+00 6.50003076e-01 3.97416860e-01
1.24071978e-01 -2.29105335e-02 4.29261327e-01 -7.80210257e-01
-1.60664356e+00 -1.31905866e+00 9.16771367e-02 -7.63728440e-01
-1.95109427e-01 -5.81613123e-01 -8.07192028e-01 1.04684973e+00
-6.12382255e-02 5.75992644e-01 2.33226955e-01 -6.88192606e-01
-3.56660187e-01 2.70568401e-01 -1.75525904e+00 1.54105216e-01
1.37291741e+00 -5.65164268e-01 -6.38033867e-01 2.08215073e-01
8.69109809e-01 -1.21679974e+00 -1.21088648e+00 5.85687637e-01
5.56544185e-01 -1.40993547e+00 1.49186349e+00 -3.88901800e-01
8.13092768e-01 -3.62224072e-01 -5.60117006e-01 -1.37324286e+00
-1.59712553e-01 -6.94221482e-02 -4.35888797e-01 7.52234995e-01
-3.82048786e-01 -4.10709888e-01 1.11363637e+00 1.57159507e-01
-1.99898526e-01 -7.20231950e-01 -1.18738019e+00 -4.81726289e-01
1.07275330e-01 -4.77170646e-01 1.01173544e+00 6.85050607e-01
-1.18671215e+00 2.69751232e-02 -4.53598320e-01 4.49621111e-01
1.01564002e+00 4.33798522e-01 1.28892052e+00 -1.05181086e+00
-5.88745415e-01 3.53688747e-01 -7.48728514e-01 -1.10467017e+00
2.20433310e-01 -7.56759048e-01 -4.97407541e-02 -1.66855049e+00
5.19389845e-02 -3.87958288e-01 2.28157118e-01 2.78219372e-01
1.37650326e-01 6.75356925e-01 -4.44509163e-02 -6.13670088e-02
-8.73761252e-02 9.28467214e-01 1.74222171e+00 -2.05429927e-01
9.40260515e-02 9.77021921e-03 -4.38428074e-01 7.52774954e-01
2.65546530e-01 -5.09986997e-01 -3.30946535e-01 -4.97999817e-01
-1.06615521e-01 5.35933435e-01 9.75868165e-01 -9.78707910e-01
-3.67862135e-01 -2.36544594e-01 1.06471682e+00 -1.19955468e+00
8.87461603e-01 -1.13507175e+00 7.98607171e-01 2.79631317e-01
2.14896366e-01 -1.86691046e-01 1.32400692e-01 8.41456473e-01
-2.70006061e-02 1.48135215e-01 8.44832003e-01 -8.29943061e-01
-4.54404742e-01 7.76751876e-01 1.63212195e-01 1.25494629e-01
1.10873067e+00 -3.30423117e-01 8.24747458e-02 -3.60672206e-01
-1.06494081e+00 -2.47797072e-01 1.03000689e+00 2.88081646e-01
1.02395535e+00 -1.59141183e+00 -6.65898860e-01 6.60533607e-01
-4.37709540e-02 1.17312384e+00 6.38241351e-01 -1.07851297e-01
-7.64322698e-01 1.26654789e-01 -5.40145040e-01 -8.81996989e-01
-8.54456544e-01 6.39425278e-01 4.65561271e-01 -2.41826206e-01
-1.13178456e+00 7.35282242e-01 5.13518453e-01 -8.59800518e-01
-4.38359864e-02 -6.32387251e-02 2.92882085e-01 -6.02696419e-01
2.78624177e-01 2.29448557e-01 1.66272502e-02 -7.84630775e-01
-6.45004064e-02 6.87776804e-01 3.45741451e-01 -2.33166650e-01
1.69870877e+00 8.31651241e-02 8.37093592e-02 4.42568570e-01
1.00327325e+00 -2.31725723e-02 -1.74372911e+00 -1.98877364e-01
-1.16625440e+00 -7.86663532e-01 -3.83754000e-02 -5.60710192e-01
-1.32911098e+00 8.85842741e-01 6.27606988e-01 -1.15906797e-01
8.73127103e-01 2.71169037e-01 7.24603057e-01 7.91767165e-02
8.26041162e-01 -2.69260585e-01 1.80095777e-01 3.07759464e-01
1.03777099e+00 -1.02118289e+00 8.38818476e-02 -4.44853663e-01
-3.34238321e-01 7.34086633e-01 7.84443498e-01 -5.57442129e-01
5.41020572e-01 3.23872715e-01 -1.92577749e-01 -4.35980529e-01
-2.15367153e-01 5.99293113e-02 -1.15753114e-02 1.05546761e+00
-3.65718305e-01 5.69577515e-02 6.75171137e-01 2.11275578e-01
-4.15556401e-01 -1.84370294e-01 3.69766831e-01 9.49397087e-01
9.02642962e-03 -8.78113210e-01 -7.51791775e-01 1.59880787e-01
-2.50904739e-01 2.80810356e-01 -2.26055637e-01 8.98215294e-01
4.31315780e-01 1.60805389e-01 2.38285616e-01 -3.70598435e-01
4.12184715e-01 1.07894972e-01 9.99164760e-01 -7.32379079e-01
-1.69317260e-01 8.30564126e-02 -5.40086687e-01 -7.41340160e-01
-5.49272537e-01 -5.62656939e-01 -1.10699201e+00 -2.08163649e-01
1.61424249e-01 -2.93531597e-01 7.48956442e-01 7.38868356e-01
2.37607762e-01 7.89432406e-01 6.79160416e-01 -1.56484568e+00
-3.66396457e-01 -8.89833450e-01 -8.12846839e-01 5.09879589e-01
4.56810981e-01 -8.55476022e-01 -3.90084773e-01 6.88089952e-02] | [9.019465446472168, -3.3216159343719482] |
f6613c33-f90a-47a8-aeee-c19193fcbdab | a-comparative-study-of-neural-network-1 | 2104.14978 | null | https://arxiv.org/abs/2104.14978v1 | https://arxiv.org/pdf/2104.14978v1.pdf | A comparative study of neural network techniques for automatic software vulnerability detection | Software vulnerabilities are usually caused by design flaws or implementation errors, which could be exploited to cause damage to the security of the system. At present, the most commonly used method for detecting software vulnerabilities is static analysis. Most of the related technologies work based on rules or code similarity (source code level) and rely on manually defined vulnerability features. However, these rules and vulnerability features are difficult to be defined and designed accurately, which makes static analysis face many challenges in practical applications. To alleviate this problem, some researchers have proposed to use neural networks that have the ability of automatic feature extraction to improve the intelligence of detection. However, there are many types of neural networks, and different data preprocessing methods will have a significant impact on model performance. It is a great challenge for engineers and researchers to choose a proper neural network and data preprocessing method for a given problem. To solve this problem, we have conducted extensive experiments to test the performance of the two most typical neural networks (i.e., Bi-LSTM and RVFL) with the two most classical data preprocessing methods (i.e., the vector representation and the program symbolization methods) on software vulnerability detection problems and obtained a series of interesting research conclusions, which can provide valuable guidelines for researchers and engineers. Specifically, we found that 1) the training speed of RVFL is always faster than BiLSTM, but the prediction accuracy of Bi-LSTM model is higher than RVFL; 2) using doc2vec for vector representation can make the model have faster training speed and generalization ability than using word2vec; and 3) multi-level symbolization is helpful to improve the precision of neural network models. | ['Lin Yang', 'Qiang Wang', 'Weipeng Cao', 'Shuangyin Ren', 'Lianxiao Meng', 'Gaigai Tang'] | 2021-04-29 | null | null | null | null | ['vulnerability-detection'] | ['miscellaneous'] | [-1.41679376e-01 -6.02141082e-01 -2.14508489e-01 -2.09444627e-01
3.65382172e-02 -3.70993376e-01 -4.32296023e-02 2.79009640e-01
-2.63999611e-01 1.19527757e-01 -3.66548123e-03 -8.43546987e-01
1.79362334e-02 -1.05197084e+00 -2.89428711e-01 -4.28699493e-01
-6.39753742e-03 -5.24606764e-01 4.31982100e-01 -5.18836975e-01
6.52415395e-01 2.71875948e-01 -1.50310874e+00 1.67800024e-01
8.68532002e-01 8.63591671e-01 2.36436635e-01 2.00311720e-01
-8.25272083e-01 7.24422157e-01 -8.62279952e-01 -3.06828439e-01
2.42775436e-02 -2.15471447e-01 -6.03291273e-01 -6.83438361e-01
-2.02311486e-01 -1.38026416e-01 -2.31138781e-01 1.64388597e+00
2.84835905e-01 4.27081846e-02 2.01462880e-01 -1.00839162e+00
-8.47132444e-01 9.90233183e-01 -7.18595088e-01 4.20995414e-01
1.71847135e-01 1.04701361e-02 6.44600153e-01 -3.97661924e-01
4.62142378e-02 1.24423242e+00 8.03180635e-01 4.83607262e-01
-6.69895172e-01 -9.61827397e-01 3.41684669e-01 5.12740314e-01
-1.38604677e+00 7.63274282e-02 9.86857414e-01 -5.30556798e-01
1.03358150e+00 1.36211470e-01 3.10464561e-01 9.78447139e-01
4.14232612e-01 3.99608403e-01 7.43973434e-01 -4.71172214e-01
1.79941915e-02 3.47757816e-01 8.26777041e-01 4.67662245e-01
5.82053185e-01 1.70545857e-02 2.75046170e-01 -2.89367348e-01
5.45075178e-01 4.83594090e-01 -4.66954261e-01 2.58646250e-01
-8.05556595e-01 1.01422966e+00 6.54139698e-01 1.01246846e+00
-2.78593618e-02 -4.11494561e-02 1.02205849e+00 3.86975348e-01
-5.95009476e-02 5.48006296e-01 -5.76989949e-01 -2.34708831e-01
-6.16896987e-01 -1.08898930e-01 4.35033858e-01 3.59511346e-01
6.72213078e-01 6.35979474e-01 1.91848919e-01 7.64228284e-01
3.86241615e-01 3.78363192e-01 1.21564567e+00 -1.76106747e-02
5.43638527e-01 9.95257616e-01 -4.57308441e-01 -1.75089943e+00
-3.78737360e-01 -2.59644747e-01 -8.17545831e-01 2.99953967e-01
6.45751208e-02 -1.68186992e-01 -8.40764523e-01 1.56654155e+00
-1.07403740e-01 5.15860990e-02 1.94911510e-01 5.57049155e-01
1.03392935e+00 9.11177814e-01 6.23544753e-02 -2.17424497e-01
1.22574592e+00 -5.89800477e-01 -8.82684469e-01 -1.70326084e-01
9.40170050e-01 -6.22577965e-01 1.21427047e+00 4.70638543e-01
-3.45891565e-01 -8.70626032e-01 -1.33226359e+00 3.44240695e-01
-8.40641379e-01 5.54897748e-02 6.46588326e-01 1.10103893e+00
-6.88905001e-01 5.98322988e-01 -6.03989422e-01 -2.33788282e-01
-7.16418400e-02 2.63227403e-01 -2.30021104e-01 1.87900022e-01
-1.62971818e+00 8.06010604e-01 8.31712127e-01 3.16974729e-01
-4.67546582e-01 -2.56289631e-01 -9.31732655e-01 1.56683385e-01
3.76433372e-01 1.89433858e-01 9.69501972e-01 -1.15298820e+00
-1.02617896e+00 -1.42329028e-02 1.57399531e-02 -2.29000777e-01
-2.17444316e-01 -1.31352633e-01 -9.14078057e-01 -4.28089410e-01
-2.14838505e-01 -2.17831731e-01 4.75421935e-01 -8.77806246e-01
-5.17133653e-01 -3.68601948e-01 1.54805630e-01 -4.14645106e-01
-1.19291222e+00 3.18207949e-01 -1.58230618e-01 -6.39309704e-01
-3.19250077e-02 -4.86296028e-01 -2.79277146e-01 -7.20209539e-01
-3.92173886e-01 -3.69906396e-01 9.97875929e-01 -1.13447142e+00
1.96455574e+00 -2.15818238e+00 -2.69037396e-01 5.04020870e-01
1.83614239e-01 9.70651150e-01 -1.55934349e-01 2.26705313e-01
-4.38117951e-01 6.16073072e-01 -1.42758146e-01 4.50701565e-01
-3.99005294e-01 -9.87999979e-03 -5.46571791e-01 -1.74933132e-02
-1.03502564e-01 6.69997156e-01 -6.09783590e-01 -2.19430760e-01
1.38897628e-01 3.49889100e-01 -1.37878522e-01 2.49395847e-01
-5.21571673e-02 -9.58831608e-02 -8.68825316e-01 5.63862860e-01
6.12479031e-01 2.32701197e-01 6.27668872e-02 -2.17194781e-01
-2.59550273e-01 1.16637327e-01 -1.23585010e+00 9.81000602e-01
-6.78626657e-01 5.95252573e-01 -4.81827825e-01 -1.10660648e+00
1.30243087e+00 3.58726919e-01 -3.12621100e-03 -7.47513950e-01
4.74219948e-01 2.25938857e-01 2.86926776e-01 -8.79840493e-01
1.86325073e-01 2.37441570e-01 -2.26414844e-01 2.71300107e-01
-4.09599602e-01 5.74874401e-01 4.56921123e-02 -2.31012762e-01
1.02527142e+00 -2.50614822e-01 2.27636412e-01 -4.86123301e-02
9.30194139e-01 -1.76890686e-01 9.47692692e-01 3.41375023e-01
5.99800125e-02 9.37925838e-03 6.92734540e-01 -6.59497976e-01
-6.51231110e-01 -3.43799025e-01 7.30270445e-02 5.89194298e-01
6.68182224e-02 -5.05110025e-01 -8.12237561e-01 -8.45751464e-01
-2.07434207e-01 7.63451278e-01 -5.80109894e-01 -6.96200311e-01
-6.71975970e-01 -6.36610508e-01 8.00502062e-01 8.61283004e-01
7.11856723e-01 -1.11641502e+00 -5.57509422e-01 2.32134566e-01
3.09374072e-02 -6.83760285e-01 -2.54858375e-01 -7.52407759e-02
-8.94283295e-01 -1.17617512e+00 -3.55541468e-01 -7.79367328e-01
5.96515954e-01 4.64483321e-01 6.03589833e-01 6.59051657e-01
-1.12714015e-01 -3.02708924e-01 -8.26087117e-01 -3.77615184e-01
-4.73356575e-01 3.90530471e-03 1.66281939e-01 -2.10799024e-01
5.19254863e-01 -3.87264252e-01 -1.65577121e-02 2.64203936e-01
-1.14735150e+00 -4.94804323e-01 6.75827801e-01 6.56600356e-01
1.86152562e-01 7.27434397e-01 5.85084617e-01 -8.25128496e-01
1.17191494e+00 -4.38811034e-01 -7.73242593e-01 4.35529977e-01
-8.21963370e-01 2.00308412e-01 1.14816320e+00 -5.23462415e-01
-9.86709893e-01 -3.85529637e-01 -4.10208583e-01 -4.43616599e-01
-2.26142518e-02 1.13669586e+00 -5.21797001e-01 -4.54311758e-01
5.89329302e-01 3.18572342e-01 -2.37060916e-02 -6.02251709e-01
2.22680680e-02 8.84000599e-01 9.34327841e-02 -3.76935124e-01
8.58713329e-01 -3.51259649e-01 -2.85644144e-01 -7.08397150e-01
-1.16330266e-01 -2.71375868e-02 -2.80385554e-01 -5.85445669e-04
8.32821548e-01 -2.58527309e-01 -6.58208847e-01 6.13934815e-01
-1.27885509e+00 1.47613168e-01 3.95164251e-01 4.27495837e-01
4.34234560e-01 9.22130704e-01 -4.79047745e-01 -9.29037035e-01
-4.83642578e-01 -1.50473738e+00 9.79407784e-03 5.80913424e-01
-4.88311648e-02 -9.77524281e-01 -8.27919021e-02 -2.21201316e-01
7.63588727e-01 3.38027418e-01 1.42348146e+00 -7.02681661e-01
-8.70794058e-02 -4.79102433e-01 -1.88502982e-01 7.54094899e-01
2.70503283e-01 6.07435107e-01 -7.84280956e-01 -3.94693539e-02
3.75653207e-01 1.13510065e-01 8.18023384e-01 1.67998865e-01
1.50850010e+00 -3.19421977e-01 -4.40675020e-01 4.85884547e-01
1.49638498e+00 9.21463490e-01 8.15151870e-01 5.73207319e-01
8.72554779e-01 5.50900519e-01 5.50918460e-01 -2.43882425e-02
1.81931078e-01 5.60495019e-01 4.87816244e-01 1.67298261e-02
3.64556193e-01 -1.01065800e-01 5.94818294e-01 1.04457557e+00
-1.65216863e-01 8.49929601e-02 -1.15699923e+00 3.23080510e-01
-1.75820696e+00 -9.22229350e-01 -3.57544363e-01 2.45550704e+00
7.73374319e-01 3.46128196e-01 -1.96063444e-02 5.96149385e-01
8.27524602e-01 1.20936520e-01 -2.79430866e-01 -8.85244668e-01
1.64842531e-01 5.31398021e-02 3.54606599e-01 3.51855494e-02
-1.04931164e+00 7.50756621e-01 5.12260342e+00 1.11310685e+00
-1.66220939e+00 2.17437353e-02 3.89348060e-01 5.28300405e-01
-3.77316624e-01 4.05755118e-02 -8.24345529e-01 6.85959876e-01
1.12258649e+00 -2.02788487e-01 1.76296830e-01 1.02995861e+00
4.35226336e-02 3.44694763e-01 -7.15472579e-01 9.01535273e-01
2.57625617e-02 -1.00116706e+00 -6.79693669e-02 -3.52097601e-02
3.03040743e-01 -4.10549819e-01 1.70604903e-02 7.24985778e-01
1.43680852e-02 -1.03585374e+00 2.57291436e-01 2.98436105e-01
5.19661546e-01 -9.90843534e-01 1.23781240e+00 2.59719729e-01
-1.43274248e+00 -5.71847022e-01 -6.90359592e-01 -9.55601111e-02
-1.45400062e-01 6.78377092e-01 -3.40498805e-01 7.42932260e-01
8.38393509e-01 6.20137453e-01 -8.56679678e-01 1.08198845e+00
-3.23401034e-01 6.73765123e-01 4.48189899e-02 -3.85446250e-01
1.89686641e-01 7.02428147e-02 1.51417091e-01 1.02493978e+00
4.37984616e-01 -3.06335092e-01 1.27997577e-01 8.61616552e-01
2.09414020e-01 3.29218388e-01 -6.53310895e-01 -2.87388831e-01
4.38302875e-01 1.19790006e+00 -6.64294958e-01 -1.73677597e-02
-6.81073666e-01 2.25856200e-01 1.71166599e-01 2.81448960e-01
-9.30907011e-01 -1.15029824e+00 5.58504939e-01 -5.45100272e-02
1.28165230e-01 -2.54514128e-01 -3.77617031e-01 -1.08624959e+00
1.51254147e-01 -9.95455325e-01 4.17840570e-01 -4.03499424e-01
-1.04946220e+00 9.71600890e-01 -1.01883158e-01 -1.22266567e+00
-1.26605958e-01 -7.28117347e-01 -1.17323172e+00 1.00002813e+00
-1.38793468e+00 -7.73096085e-01 -3.33641678e-01 3.38043183e-01
3.02044600e-01 -4.70573097e-01 7.65394986e-01 4.98279929e-01
-1.17404652e+00 9.74753737e-01 3.41983438e-02 5.59706151e-01
4.21196461e-01 -7.55561292e-01 5.44735610e-01 1.22059333e+00
-3.40377726e-02 1.19175601e+00 3.65835011e-01 -8.26920033e-01
-1.31377172e+00 -1.15712106e+00 5.63155293e-01 -1.60054564e-01
5.90613544e-01 -5.12935780e-02 -1.46828616e+00 5.53553939e-01
-5.68851382e-02 -3.69118154e-01 7.46089697e-01 1.97503746e-01
-6.36993766e-01 -2.50888556e-01 -9.99691546e-01 6.73487425e-01
3.41778040e-01 -4.04307514e-01 -6.45821571e-01 -1.36687964e-01
9.02391195e-01 1.51973721e-02 -6.90577507e-01 5.69824100e-01
4.02086198e-01 -8.82921696e-01 8.43506396e-01 -4.69046444e-01
3.97170156e-01 -5.27774990e-01 -4.30852585e-02 -1.02688587e+00
-3.87835771e-01 8.10788125e-02 -3.21575254e-02 1.54200923e+00
3.83365065e-01 -1.04688907e+00 3.95915180e-01 6.13874078e-01
-5.05555198e-02 -8.41838956e-01 -5.65972328e-01 -9.55461562e-01
8.73377770e-02 -5.76588988e-01 8.51511002e-01 1.09512150e+00
1.24049179e-01 7.17610866e-02 -1.14975974e-01 1.18942976e-01
1.82309523e-01 -1.51477326e-02 4.90450889e-01 -1.34933817e+00
-2.57124633e-01 -8.95563841e-01 -6.49749100e-01 -5.44694960e-01
2.98723161e-01 -6.39021575e-01 -3.01951140e-01 -1.33967197e+00
-7.24828169e-02 -3.31261605e-01 -6.42542422e-01 9.40695643e-01
-4.58805829e-01 -3.24917585e-01 -1.17942132e-01 1.92071751e-01
1.58555090e-01 3.27806056e-01 6.44641519e-01 -3.64150286e-01
-4.08313751e-01 2.63642639e-01 -9.31185782e-01 7.45612979e-01
8.60105276e-01 -3.06129575e-01 -4.97708410e-01 -5.87749243e-01
3.15353572e-01 -1.45491632e-02 -1.85460672e-02 -1.01526356e+00
1.38807625e-01 -2.66715497e-01 1.55426338e-01 -1.89805076e-01
-3.42755973e-01 -7.77510583e-01 -8.20716470e-02 7.70147324e-01
-8.93300995e-02 4.42870378e-01 5.49903452e-01 2.94791430e-01
-3.54682356e-01 -8.56544375e-01 4.08435822e-01 -8.56001154e-02
-1.14693928e+00 1.70947447e-01 -3.80739242e-01 -3.71032625e-01
1.01298976e+00 -2.80106872e-01 -3.36525083e-01 -1.07714258e-01
-4.46595438e-02 6.32645711e-02 3.26507092e-01 8.15904677e-01
8.73487175e-01 -1.34750247e+00 -2.63515800e-01 3.82833421e-01
5.35177253e-02 -2.90439427e-01 3.92065108e-01 4.75658655e-01
-4.66664672e-01 4.06714588e-01 -3.43698531e-01 -1.12080825e-02
-1.41927087e+00 8.76134694e-01 3.28054488e-01 -1.47587016e-01
-3.47080082e-01 7.23055899e-01 -3.77355050e-03 -2.75432855e-01
2.77884424e-01 -4.11615968e-01 -1.02506721e+00 -8.94247815e-02
9.01971936e-01 3.62595230e-01 1.42924801e-01 -5.31034410e-01
-4.65152651e-01 8.10483873e-01 -3.58767539e-01 3.73215437e-01
1.03296876e+00 5.64638197e-01 -3.86440605e-01 3.21312845e-01
1.16522777e+00 5.44592515e-02 -2.62367547e-01 -1.15421824e-01
1.11386873e-01 -4.02468532e-01 2.81393558e-01 -6.51980519e-01
-1.44574630e+00 1.38559079e+00 6.78745687e-01 5.54503858e-01
1.15769207e+00 -5.95780373e-01 1.06060338e+00 4.17430758e-01
5.62117100e-01 -7.15683818e-01 -1.64107352e-01 5.19049823e-01
5.36251068e-01 -9.65603709e-01 -2.22233027e-01 -1.93617329e-01
-3.98534298e-01 1.58888388e+00 9.98410702e-01 4.61529084e-02
6.89066112e-01 2.72928357e-01 1.75603285e-01 9.11514238e-02
-1.82490051e-01 1.00335866e-01 2.73404121e-01 5.28076053e-01
6.76233470e-01 -1.18558072e-01 -5.58813512e-01 1.23332667e+00
7.34610774e-04 -3.45889390e-01 6.87109590e-01 8.57454360e-01
-6.26386464e-01 -1.46583939e+00 -4.54028308e-01 5.75881362e-01
-4.84425157e-01 -7.98502639e-02 -2.14550704e-01 5.28494120e-01
2.46341512e-01 1.14512563e+00 -4.39791352e-01 -1.23324716e+00
4.88797754e-01 -1.76417947e-01 -1.97011888e-01 -6.89466953e-01
-7.91315258e-01 -3.95758033e-01 -4.05735493e-01 -4.25261468e-01
1.44173533e-01 -3.85260791e-01 -1.32597649e+00 -2.92318732e-01
-7.17425346e-01 4.78194475e-01 6.76644444e-01 8.96329701e-01
1.80871323e-01 9.38099861e-01 8.30722094e-01 -3.58227193e-01
-6.44070387e-01 -9.85032082e-01 -3.02554131e-01 2.44083419e-01
-6.41433671e-02 -7.53494143e-01 -3.50567341e-01 -3.50274950e-01] | [7.046029090881348, 7.853194713592529] |
83064a7f-14b0-46b8-83ad-21c675f87b10 | wide-and-deep-graph-neural-network-with | 2107.09203 | null | https://arxiv.org/abs/2107.09203v2 | https://arxiv.org/pdf/2107.09203v2.pdf | Wide and Deep Graph Neural Network with Distributed Online Learning | Graph neural networks (GNNs) are naturally distributed architectures for learning representations from network data. This renders them suitable candidates for decentralized tasks. In these scenarios, the underlying graph often changes with time due to link failures or topology variations, creating a mismatch between the graphs on which GNNs were trained and the ones on which they are tested. Online learning can be leveraged to retrain GNNs at testing time to overcome this issue. However, most online algorithms are centralized and usually offer guarantees only on convex problems, which GNNs rarely lead to. This paper develops the Wide and Deep GNN (WD-GNN), a novel architecture that can be updated with distributed online learning mechanisms. The WD-GNN consists of two components: the wide part is a linear graph filter and the deep part is a nonlinear GNN. At training time, the joint wide and deep architecture learns nonlinear representations from data. At testing time, the wide, linear part is retrained, while the deep, nonlinear one remains fixed. This often leads to a convex formulation. We further propose a distributed online learning algorithm that can be implemented in a decentralized setting. We also show the stability of the WD-GNN to changes of the underlying graph and analyze the convergence of the proposed online learning procedure. Experiments on movie recommendation, source localization and robot swarm control corroborate theoretical findings and show the potential of the WD-GNN for distributed online learning. | ['Alejandro Ribeiro', 'Fernando Gama', 'Zhan Gao'] | 2021-07-19 | null | null | null | null | ['movie-recommendation'] | ['miscellaneous'] | [-4.57788289e-01 3.40340227e-01 -6.58295155e-02 3.68260890e-02
-1.10360449e-02 -6.42663360e-01 2.58463979e-01 1.65811092e-01
-2.11465776e-01 6.21485293e-01 -1.93241388e-01 -2.87934363e-01
-4.56574917e-01 -9.95901167e-01 -1.16588128e+00 -8.48252356e-01
-6.10979438e-01 7.65796304e-01 2.59014875e-01 -3.67218286e-01
-2.29501054e-01 5.18798947e-01 -1.03697658e+00 -3.07507038e-01
6.38947189e-01 9.63485956e-01 4.08667445e-01 7.51827657e-01
-2.80101709e-02 7.74422586e-01 -6.26695633e-01 -2.59154588e-01
6.24353230e-01 -3.19453508e-01 -4.29124534e-01 3.39234889e-01
1.06926128e-01 -4.65495497e-01 -8.27689528e-01 1.04254198e+00
5.71176648e-01 3.75971913e-01 3.44402760e-01 -1.46226227e+00
-8.21488142e-01 9.27794993e-01 -3.85239333e-01 5.75031154e-02
-3.14867273e-02 -5.60516044e-02 9.53239262e-01 -4.79540646e-01
6.39958978e-01 1.08265924e+00 7.40146518e-01 4.65186715e-01
-1.04197323e+00 -3.87517184e-01 6.41826570e-01 1.19564362e-01
-1.09233809e+00 -1.11676067e-01 8.29933465e-01 -1.97007895e-01
6.93185747e-01 -3.64450902e-01 8.46410453e-01 1.02630508e+00
2.56301731e-01 5.31592727e-01 3.91299784e-01 -1.31318659e-01
6.23388171e-01 -9.31766555e-02 -3.28401864e-01 8.82973135e-01
6.83555901e-01 4.36056368e-02 -3.95806640e-01 -1.88108131e-01
8.88674200e-01 2.10207909e-01 -4.46301103e-01 -7.33793974e-01
-5.72283804e-01 9.14595783e-01 1.02867866e+00 1.36437058e-01
-7.58673549e-01 3.46833050e-01 3.20823163e-01 9.96862829e-01
4.50771570e-01 1.17495395e-01 -5.87254584e-01 1.26173005e-01
-4.03796226e-01 -3.92983928e-02 1.29123902e+00 7.72635937e-01
9.29208100e-01 2.26147130e-01 4.02925283e-01 7.82971978e-01
2.97776908e-01 5.13051927e-01 5.80613077e-01 -5.91037214e-01
3.71398985e-01 5.95764399e-01 -9.23330411e-02 -1.35778522e+00
-6.99685693e-01 -8.46068382e-01 -1.09915888e+00 1.08659491e-01
4.26366687e-01 -7.66413987e-01 -5.64420760e-01 1.83375800e+00
6.46358550e-01 3.10040981e-01 -3.16472352e-02 8.61126423e-01
3.07787746e-01 6.72485113e-01 -5.24076760e-01 1.87520031e-02
4.59811032e-01 -1.01457214e+00 -3.78183722e-01 -3.70785713e-01
7.29228854e-01 7.03216493e-02 4.47878718e-01 3.46835136e-01
-9.26073313e-01 -2.26232648e-01 -9.73366261e-01 3.05170506e-01
-4.30418253e-01 1.47351837e-02 4.74005103e-01 2.94512272e-01
-1.57748795e+00 8.65686655e-01 -1.03316581e+00 -5.66255808e-01
1.74322441e-01 7.34704435e-01 -1.37566134e-01 -2.29164734e-01
-9.87679958e-01 5.39860070e-01 3.53849709e-01 5.19055724e-01
-1.18726158e+00 -2.36158609e-01 -6.57236695e-01 1.30887479e-01
6.64635777e-01 -7.77174950e-01 1.08814573e+00 -1.32456207e+00
-1.97656584e+00 4.22565565e-02 3.91177177e-01 -8.27668726e-01
3.68984252e-01 6.58336133e-02 -2.20009103e-01 1.51532263e-01
-1.25297725e-01 7.64477476e-02 1.16757214e+00 -1.00623083e+00
-4.61224079e-01 -1.57068729e-01 4.98661339e-01 2.10964143e-01
-6.46314740e-01 -6.04171038e-01 -3.07120651e-01 -5.08624613e-01
-8.93684402e-02 -1.07540298e+00 -3.59273463e-01 1.53449520e-01
-1.90063760e-01 -2.33747169e-01 9.76011097e-01 -4.26888555e-01
9.44273293e-01 -1.98456550e+00 5.54828167e-01 6.42773271e-01
5.02686679e-01 2.34579325e-01 -6.00539029e-01 8.32582831e-01
2.30495229e-01 -2.14840293e-01 2.28493080e-01 -3.17235947e-01
-6.79494888e-02 5.19479275e-01 -1.05920561e-01 7.19545662e-01
-1.26102060e-01 9.14475501e-01 -9.86098886e-01 2.60349870e-01
-1.30135685e-01 3.74039710e-01 -6.77452564e-01 3.13897699e-01
-4.54795867e-01 3.27154338e-01 -5.82819104e-01 2.39697039e-01
4.77277666e-01 -8.87683809e-01 5.84071219e-01 9.56004784e-02
2.07818732e-01 1.56350974e-02 -1.22768116e+00 1.56122077e+00
-5.83310544e-01 5.24251938e-01 7.37971425e-01 -1.45252061e+00
8.90637577e-01 1.95896134e-01 5.34776330e-01 -3.82535696e-01
1.80919096e-01 2.23851085e-01 1.33447960e-01 -2.98502356e-01
1.51087701e-01 2.34975219e-01 2.39102960e-01 6.26151741e-01
4.59349841e-01 1.62278146e-01 1.15532883e-01 3.59489560e-01
1.49235010e+00 -3.68133754e-01 8.06313604e-02 -1.96317732e-01
3.70502293e-01 -2.62353957e-01 5.63978016e-01 1.10303032e+00
1.08948037e-01 8.09374452e-02 6.78475142e-01 -6.45959854e-01
-8.43852997e-01 -8.27519357e-01 5.94664812e-01 1.30594099e+00
2.52681732e-01 -2.99342602e-01 -5.77587426e-01 -9.01476920e-01
1.82720855e-01 7.93608651e-02 -5.23836970e-01 -3.93195987e-01
-4.29344684e-01 -3.28347147e-01 1.18171774e-01 2.28748038e-01
3.83283645e-01 -8.60903919e-01 -3.28464508e-01 5.46352804e-01
2.95330137e-01 -8.87172520e-01 -4.44884449e-01 2.52818137e-01
-9.35611069e-01 -1.24857593e+00 -6.13350093e-01 -8.44645143e-01
9.65679467e-01 4.76560712e-01 9.10696924e-01 4.08697128e-01
1.96685329e-01 8.97682667e-01 -5.16614676e-01 -2.53244579e-01
-4.44886744e-01 3.24222565e-01 2.17356697e-01 3.62764060e-01
-3.58729959e-01 -8.89128566e-01 -4.97255147e-01 2.94043988e-01
-9.52791214e-01 -2.48328879e-01 6.35842741e-01 9.42133844e-01
3.27152342e-01 3.71490061e-01 7.28148878e-01 -9.93997097e-01
7.65882075e-01 -9.00589168e-01 -1.05329704e+00 3.17783296e-01
-5.86933732e-01 1.69556275e-01 1.37168467e+00 -7.49960899e-01
-4.77370739e-01 -9.82513949e-02 2.76536047e-01 -7.59873807e-01
3.20685893e-01 1.07484472e+00 1.12768643e-01 -5.05145609e-01
6.10838711e-01 -8.84343758e-02 3.47622365e-01 -3.47217709e-01
3.81462693e-01 4.05219644e-01 1.82482570e-01 -4.65824425e-01
1.12271643e+00 4.37773705e-01 3.01076230e-02 -9.14991975e-01
-6.47401571e-01 -1.11554116e-01 -3.33953261e-01 -3.75187814e-01
2.28569061e-01 -9.45303202e-01 -5.80337882e-01 6.74668968e-01
-9.33116853e-01 -1.08379221e+00 -3.98797750e-01 3.55543762e-01
-2.99630255e-01 1.11449905e-01 -7.29914248e-01 -6.13907993e-01
-2.72317737e-01 -5.35156548e-01 7.29364753e-01 3.38530570e-01
4.17262971e-01 -1.69174469e+00 1.64164633e-01 -2.24071309e-01
7.90719032e-01 1.83138862e-01 5.99190831e-01 -7.62853622e-01
-4.77975190e-01 -3.79876107e-01 8.21011439e-02 3.12813938e-01
1.60169527e-01 -1.52526677e-01 -3.72493505e-01 -1.11836386e+00
-7.51933157e-02 -4.76931125e-01 5.43738008e-01 4.44070995e-01
9.23233747e-01 -7.96779573e-01 -1.79194778e-01 7.15300441e-01
1.46413565e+00 -1.63468018e-01 -6.33011805e-03 2.04116300e-01
7.89682388e-01 2.75200397e-01 -2.62533221e-02 6.14402473e-01
5.78325212e-01 2.11488888e-01 8.19331944e-01 8.74864087e-02
3.81709151e-02 -3.24997395e-01 7.30804861e-01 1.18731403e+00
9.15807709e-02 -6.21290505e-01 -7.56747663e-01 4.31156158e-01
-2.14415622e+00 -6.56330287e-01 4.21019346e-01 2.17500067e+00
3.46805453e-01 -7.73661286e-02 1.79447666e-01 -3.63744289e-01
8.21651399e-01 1.47275850e-01 -1.07422209e+00 -1.90901801e-01
-1.83288634e-01 -7.70104397e-03 6.66254222e-01 4.26039487e-01
-8.55403900e-01 6.28320098e-01 5.12551594e+00 3.93778294e-01
-1.40303540e+00 4.17479798e-02 1.99521318e-01 -9.15547311e-02
-1.08847074e-01 4.02066112e-02 -4.61750329e-01 3.26705605e-01
1.20985055e+00 -3.41546178e-01 1.06884468e+00 9.81197894e-01
7.94543922e-02 3.04844260e-01 -1.05514836e+00 8.94916058e-01
3.60176479e-03 -1.36803806e+00 -2.31899992e-02 2.73192316e-01
9.73262250e-01 6.86619341e-01 -1.03249677e-01 3.02526146e-01
8.33967984e-01 -5.97605288e-01 5.48864007e-01 4.77221429e-01
3.53446960e-01 -5.95887482e-01 6.47335470e-01 6.34757280e-01
-1.22612154e+00 -6.13984644e-01 -6.95200861e-01 -6.24110810e-02
2.60749869e-02 6.53458416e-01 -6.41661942e-01 7.57998466e-01
5.61641276e-01 1.01804698e+00 -4.61838245e-01 1.23290050e+00
-3.52308273e-01 6.04609072e-01 -6.72283173e-01 -2.35646442e-01
2.66895801e-01 -4.39286113e-01 8.03896129e-01 6.11545682e-01
4.11261022e-01 -1.37186453e-01 5.57295680e-01 5.72720766e-01
-4.63550121e-01 -9.48845074e-02 -7.67248690e-01 -2.31580630e-01
4.60392892e-01 1.23838711e+00 -5.92032909e-01 -9.79466140e-02
-5.16458809e-01 1.03785408e+00 9.60241914e-01 7.18427360e-01
-5.93849361e-01 -2.99883366e-01 3.83378565e-01 9.64402687e-03
6.49338961e-01 -3.60293299e-01 6.01741374e-01 -1.32257211e+00
-3.98812294e-02 -8.61533225e-01 4.79301751e-01 -4.14726615e-01
-1.68780637e+00 5.94516933e-01 -3.71831119e-01 -1.04786444e+00
-3.83838922e-01 -5.92023909e-01 -7.22242355e-01 3.15444082e-01
-1.37446368e+00 -1.01128995e+00 -2.96651155e-01 8.27196360e-01
1.36668950e-01 -2.90167838e-01 5.62681854e-01 2.25093096e-01
-8.28565538e-01 5.67396343e-01 6.30768955e-01 2.36639321e-01
3.59501809e-01 -1.19268131e+00 3.08953691e-02 8.87873769e-01
3.24591756e-01 3.59854162e-01 5.04796863e-01 -6.71524584e-01
-1.96433055e+00 -1.41537988e+00 2.25026190e-01 3.29149067e-01
1.11443436e+00 -5.97139537e-01 -8.51812005e-01 8.09665680e-01
-4.13745679e-02 4.79715586e-01 1.85124367e-01 8.27014223e-02
-1.17775939e-01 -4.62558001e-01 -8.41575623e-01 3.71940523e-01
8.93992186e-01 -4.97784048e-01 1.09474592e-01 6.12953305e-01
6.20870292e-01 -4.20068383e-01 -7.32570529e-01 -1.09615050e-01
2.20431447e-01 -6.76645935e-01 5.13864458e-01 -5.83143055e-01
-1.76718339e-01 -1.42559826e-01 2.52054900e-01 -1.82996488e+00
-3.69302571e-01 -1.15560162e+00 -6.19772017e-01 7.57530510e-01
3.13703239e-01 -1.19135845e+00 8.85853171e-01 4.00146216e-01
-4.33696397e-02 -5.45541227e-01 -9.42707181e-01 -9.22512710e-01
8.33131932e-03 1.09026909e-01 3.50791842e-01 7.59360135e-01
-1.96230233e-01 4.29735273e-01 -3.61492306e-01 3.76491368e-01
4.97458965e-01 1.06093153e-01 9.85469401e-01 -1.29189551e+00
-7.84651160e-01 -2.61719912e-01 -4.92564380e-01 -1.37898219e+00
2.50373751e-01 -9.97527957e-01 -5.16053736e-02 -1.49623692e+00
-3.94973606e-01 -3.15960437e-01 -1.96837664e-01 4.99567628e-01
2.33755246e-01 -4.34646666e-01 1.46000445e-01 2.54825354e-01
-7.98117042e-01 7.37915039e-01 9.51621532e-01 -1.84584051e-01
-4.78026181e-01 1.67175636e-01 -4.00846481e-01 5.43977737e-01
8.62446904e-01 -5.53235948e-01 -7.70593882e-01 -8.00935745e-01
5.36605954e-01 1.00084841e-01 2.52717555e-01 -1.01277435e+00
8.87957931e-01 1.70374721e-01 1.96629744e-02 4.59197946e-02
-1.62880599e-01 -1.04037976e+00 1.49723023e-01 6.05343521e-01
-1.37658611e-01 1.86161503e-01 -1.29355073e-01 1.25150645e+00
-5.25694387e-03 4.74594459e-02 5.31955123e-01 5.37586547e-02
-3.45924288e-01 8.62945259e-01 -3.96644413e-01 -4.66593355e-02
8.42055798e-01 6.75964057e-02 -3.16393495e-01 -9.17378962e-01
-8.08521211e-01 6.45622671e-01 3.48605901e-01 2.09967703e-01
5.51834285e-01 -1.05636084e+00 -5.33205688e-01 2.22216040e-01
-4.27111149e-01 3.44586760e-01 8.77619162e-02 9.64953363e-01
-4.39313918e-01 -1.15714654e-01 5.77157289e-02 -2.95108944e-01
-3.45874339e-01 5.69228947e-01 6.55833185e-01 -4.10475910e-01
-7.56733000e-01 8.11767876e-01 -4.94352467e-02 -5.40485024e-01
4.33187634e-01 -9.62387994e-02 -3.92927378e-02 -4.23496552e-02
1.34200841e-01 3.78905416e-01 1.06351212e-01 -1.54394865e-01
-2.31209844e-02 2.11691573e-01 3.70346457e-02 2.49889597e-01
1.77451205e+00 -4.61073875e-01 -4.50430065e-01 5.44243455e-01
1.16123247e+00 -2.04905406e-01 -1.42984700e+00 -4.92535770e-01
-8.51210952e-02 4.49566804e-02 -5.11073060e-02 -3.85842919e-01
-1.67875731e+00 3.81111175e-01 3.04575205e-01 6.27978802e-01
1.11212242e+00 -4.44300622e-02 7.58605838e-01 9.57497478e-01
6.53299272e-01 -1.08207786e+00 2.80395776e-01 7.78642356e-01
8.25977981e-01 -1.00845242e+00 -1.66784123e-01 1.60137400e-01
3.38533856e-02 1.45306599e+00 5.42019248e-01 -8.30338955e-01
1.07432139e+00 1.30326062e-01 -2.85769105e-01 -2.73073882e-01
-1.02809536e+00 6.14832826e-02 1.25390381e-01 5.81130505e-01
-3.30391854e-01 -1.22923508e-01 1.02273226e-01 4.62481856e-01
-4.89186086e-02 -2.60589540e-01 6.58601105e-01 7.63936162e-01
-3.83715242e-01 -9.48106468e-01 8.92547369e-02 6.36197031e-01
-7.52380639e-02 2.55302817e-01 -4.21737939e-01 7.30326593e-01
-2.56410658e-01 1.00014496e+00 4.50021401e-03 -4.56557214e-01
1.39282301e-01 -4.64440942e-01 2.07278088e-01 -5.89316845e-01
-4.98528093e-01 -3.57706845e-02 -1.40765071e-01 -7.66606331e-01
-1.47277072e-01 -3.67880523e-01 -1.08198690e+00 -3.63832742e-01
-6.32205486e-01 4.56687540e-01 5.73245704e-01 7.45985508e-01
6.39658809e-01 4.79262024e-01 1.00736880e+00 -1.05932462e+00
-9.72566783e-01 -6.99994802e-01 -8.23086858e-01 6.64421171e-02
6.59455836e-01 -5.12429535e-01 -7.39127696e-01 -3.94815773e-01] | [6.904787063598633, 6.004111289978027] |
fe57cda4-5b98-4773-a24f-ee481790dd07 | learning-memory-guided-normality-for-anomaly | 2003.13228 | null | https://arxiv.org/abs/2003.13228v1 | https://arxiv.org/pdf/2003.13228v1.pdf | Learning Memory-guided Normality for Anomaly Detection | We address the problem of anomaly detection, that is, detecting anomalous events in a video sequence. Anomaly detection methods based on convolutional neural networks (CNNs) typically leverage proxy tasks, such as reconstructing input video frames, to learn models describing normality without seeing anomalous samples at training time, and quantify the extent of abnormalities using the reconstruction error at test time. The main drawbacks of these approaches are that they do not consider the diversity of normal patterns explicitly, and the powerful representation capacity of CNNs allows to reconstruct abnormal video frames. To address this problem, we present an unsupervised learning approach to anomaly detection that considers the diversity of normal patterns explicitly, while lessening the representation capacity of CNNs. To this end, we propose to use a memory module with a new update scheme where items in the memory record prototypical patterns of normal data. We also present novel feature compactness and separateness losses to train the memory, boosting the discriminative power of both memory items and deeply learned features from normal data. Experimental results on standard benchmarks demonstrate the effectiveness and efficiency of our approach, which outperforms the state of the art. | ['Bumsub Ham', 'Jongyoun Noh', 'Hyunjong Park'] | 2020-03-30 | learning-memory-guided-normality-for-anomaly-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Park_Learning_Memory-Guided_Normality_for_Anomaly_Detection_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Park_Learning_Memory-Guided_Normality_for_Anomaly_Detection_CVPR_2020_paper.pdf | cvpr-2020-6 | ['physical-video-anomaly-detection', 'anomaly-detection-in-surveillance-videos', 'general-action-video-anomaly-detection', 'anomaly-detection-in-surveillance-videos'] | ['computer-vision', 'computer-vision', 'computer-vision', 'methodology'] | [ 1.37674585e-01 -1.21143378e-01 1.25059143e-01 -2.62793899e-01
-1.05323739e-01 -3.35838526e-01 3.76534909e-01 8.51878151e-02
-4.33965325e-01 3.10196400e-01 4.25081365e-02 -5.14886640e-02
3.21746357e-02 -7.41375864e-01 -9.62080717e-01 -7.32985675e-01
-5.24218142e-01 1.73490599e-01 3.04747015e-01 1.05145000e-01
2.13754609e-01 6.73662543e-01 -1.75889421e+00 5.06158888e-01
6.06747866e-01 1.49317145e+00 -3.08172941e-01 6.53954148e-01
-6.68855235e-02 1.35752261e+00 -5.48144877e-01 -1.62244782e-01
2.05320850e-01 -4.21506226e-01 -4.80373532e-01 3.23612720e-01
5.53798854e-01 -9.77506518e-01 -9.05696034e-01 1.02711296e+00
1.51782528e-01 2.85427183e-01 4.78184134e-01 -1.33607876e+00
-4.65527534e-01 1.74575597e-01 -4.03785706e-01 9.35974777e-01
3.19937795e-01 5.00649333e-01 9.60588098e-01 -8.84663284e-01
4.05216098e-01 9.16796744e-01 6.20887160e-01 7.06600845e-01
-8.26698661e-01 -4.82367396e-01 4.87308860e-01 6.42883956e-01
-1.12726331e+00 -5.46541452e-01 7.07664132e-01 -5.15811384e-01
9.73394692e-01 1.98717669e-01 7.53964245e-01 1.26113224e+00
8.72209668e-02 9.94024515e-01 4.15675908e-01 -2.98797470e-02
4.17143911e-01 -3.81318450e-01 1.44751579e-01 9.22629178e-01
2.34533206e-01 1.54541016e-01 -6.00729406e-01 -2.83697277e-01
7.19194293e-01 8.12506378e-01 -5.12015939e-01 -5.02955019e-01
-1.03838873e+00 7.65428424e-01 1.91399589e-01 1.86972678e-01
-5.95084846e-01 1.28234670e-01 7.87072420e-01 6.54661298e-01
5.11944711e-01 3.05038601e-01 -3.97598296e-01 -2.29454234e-01
-8.78477514e-01 1.57583371e-01 5.52543163e-01 6.45709276e-01
4.94096547e-01 2.93501914e-01 -3.52669835e-01 4.33702558e-01
5.97627312e-02 1.81642354e-01 7.99251676e-01 -8.44426334e-01
2.25862652e-01 8.06846201e-01 -2.94124663e-01 -1.05098403e+00
-3.18477273e-01 -2.76625842e-01 -1.06647623e+00 1.15230061e-01
2.71857560e-01 1.70057658e-02 -9.91715252e-01 1.71222687e+00
3.50907743e-02 8.56020331e-01 6.72006235e-03 8.30481827e-01
4.25069094e-01 5.97399235e-01 -2.61970997e-01 -2.90872216e-01
8.44733119e-01 -6.74600124e-01 -5.98286331e-01 -1.34594336e-01
8.13543260e-01 -1.69338793e-01 9.27425623e-01 4.98280823e-01
-1.22679377e+00 -4.11019981e-01 -1.02433300e+00 2.17632860e-01
-2.74405211e-01 -1.92408100e-01 3.18932921e-01 2.41767913e-01
-9.36447203e-01 8.82399023e-01 -1.23051035e+00 -2.02902019e-01
7.76511252e-01 1.98131114e-01 -4.85790998e-01 -1.58473507e-01
-9.34450448e-01 4.15658653e-01 5.82033277e-01 1.38358563e-01
-1.17050898e+00 -7.23862588e-01 -9.34177935e-01 2.63156384e-01
5.05797267e-01 -4.97463673e-01 1.04093289e+00 -1.43876970e+00
-1.05673254e+00 6.84903264e-01 -4.11459021e-02 -8.76590252e-01
5.36805272e-01 -5.64399064e-01 -5.34292936e-01 4.45367992e-01
-2.26115555e-01 2.21897691e-01 1.09107625e+00 -7.71096706e-01
-7.33013630e-01 -3.78545552e-01 2.33291574e-02 -1.99928418e-01
-5.84508896e-01 -1.74065664e-01 -3.74895185e-01 -8.41804206e-01
2.50710219e-01 -7.16174901e-01 -2.53379583e-01 2.42706671e-01
-4.21442479e-01 3.04428171e-02 1.07473040e+00 -6.84974849e-01
1.20396006e+00 -2.54241276e+00 1.42432258e-01 5.13766646e-01
5.74115038e-01 5.05613029e-01 -2.04800233e-01 8.94512460e-02
-2.59818107e-01 -2.47269228e-01 -4.63623226e-01 -2.67481029e-01
-2.08009511e-01 6.25597715e-01 -7.50629902e-01 5.77936530e-01
6.28415346e-01 7.99710274e-01 -8.49613607e-01 -5.91067150e-02
1.21000066e-01 2.12001756e-01 -8.41239035e-01 6.07560992e-01
-2.44605646e-01 6.24379277e-01 -2.98350304e-01 7.09823251e-01
4.48109388e-01 -3.29239339e-01 -1.05309956e-01 -1.66111495e-02
3.32329392e-01 1.71582654e-01 -1.03013039e+00 1.65917158e+00
-8.03255709e-04 4.42907989e-01 -3.50618780e-01 -1.50205934e+00
5.39222419e-01 3.53053331e-01 5.70577145e-01 -7.18881249e-01
-1.47649020e-01 2.76910812e-01 6.81036189e-02 -7.11349845e-01
1.28435329e-01 2.31805325e-01 2.55630404e-01 5.33409178e-01
2.39491001e-01 8.01523089e-01 1.22426651e-01 2.30557159e-01
1.68237746e+00 -7.77518526e-02 2.00280070e-01 1.82432681e-01
7.06662774e-01 -4.55980390e-01 8.14900756e-01 9.60982621e-01
-1.50978014e-01 6.92096651e-01 9.18042898e-01 -1.08810008e+00
-9.64491904e-01 -1.12605500e+00 1.78905994e-01 1.05180466e+00
-1.80118918e-01 -4.03668523e-01 -4.73521084e-01 -1.04309070e+00
-2.18180474e-02 3.72506112e-01 -8.20856214e-01 -6.86351836e-01
-9.31732953e-01 -7.10187316e-01 5.78493118e-01 7.94373393e-01
4.50919241e-01 -1.08630359e+00 -1.01643145e+00 1.47494189e-02
-2.00696230e-01 -1.21753466e+00 -1.22733355e-01 6.91840202e-02
-1.11361027e+00 -1.19075620e+00 -5.11709452e-01 -5.27791619e-01
7.99284220e-01 -7.31753930e-02 1.22187686e+00 6.09669268e-01
-4.21280444e-01 6.76181495e-01 -5.35915494e-01 -1.10325508e-01
-2.44484678e-01 -3.54152054e-01 1.47393674e-01 5.50163686e-01
4.01210368e-01 -9.06880319e-01 -6.65147305e-01 8.16133544e-02
-1.33581734e+00 -3.60221207e-01 6.92007661e-01 9.10366118e-01
7.38659501e-01 -1.03206336e-01 3.70871723e-01 -9.75494504e-01
8.81908387e-02 -8.03684890e-01 -2.85032034e-01 -3.83801498e-02
-2.71825939e-01 3.70516330e-02 8.56858969e-01 -4.56039429e-01
-5.40741384e-01 -8.10784623e-02 -2.06848636e-01 -1.17475712e+00
-3.74219000e-01 3.10205102e-01 -1.66152745e-01 1.81601420e-01
3.79255801e-01 7.99684405e-01 2.11737618e-01 -4.34075296e-01
-1.27229601e-01 3.47224884e-02 8.07284653e-01 -3.57839882e-01
6.09736443e-01 7.24721432e-01 1.11708879e-01 -6.14773989e-01
-7.84440279e-01 -4.92999822e-01 -5.49347222e-01 -1.63279608e-01
5.89593470e-01 -9.09876585e-01 -3.58345687e-01 5.41533411e-01
-1.09568918e+00 -6.97913319e-02 -6.73254430e-01 3.49546969e-01
-7.65154481e-01 6.91122651e-01 -6.33892179e-01 -5.93870699e-01
-9.35571939e-02 -8.75394404e-01 8.28467667e-01 -2.60060817e-01
-7.45822862e-02 -7.77525365e-01 5.88200279e-02 -2.53208101e-01
4.11804408e-01 5.75878859e-01 9.80525970e-01 -1.23088205e+00
-7.72859991e-01 -4.17222083e-01 -9.17578861e-02 6.09348297e-01
-1.02841914e-01 -2.33544558e-01 -1.03031909e+00 -4.88451838e-01
1.77112624e-01 -2.24790275e-01 1.31301022e+00 1.74120054e-01
1.99324548e+00 -6.93977296e-01 -9.83641818e-02 9.61755455e-01
1.14401221e+00 7.59851784e-02 8.60762656e-01 3.18502665e-01
7.10707247e-01 3.46769810e-01 2.74080127e-01 8.56832325e-01
-1.81379486e-02 3.51902425e-01 8.63386333e-01 8.79393965e-02
1.40598491e-01 1.51039669e-02 5.84471643e-01 6.45977259e-01
-2.46073887e-01 -1.31677672e-01 -7.36103952e-01 5.01322806e-01
-2.03383875e+00 -1.17873275e+00 2.38082066e-01 2.15528035e+00
3.43404114e-01 2.84615993e-01 1.78931460e-01 4.10327166e-01
4.79454905e-01 2.67136186e-01 -7.64885724e-01 -2.11095437e-01
-1.51675284e-01 8.41917023e-02 1.07515112e-01 -1.30055040e-01
-1.31905937e+00 3.73183459e-01 5.68859148e+00 3.89070243e-01
-1.07379222e+00 3.27796349e-03 6.55971110e-01 -5.02666950e-01
1.80657670e-01 -3.38539183e-01 -2.51570880e-01 7.32823312e-01
1.07283032e+00 1.45443901e-01 3.33875030e-01 8.09335887e-01
-1.38225853e-01 2.77022809e-01 -1.61374485e+00 9.86272275e-01
3.43846411e-01 -1.28404450e+00 5.16869843e-01 -3.20093669e-02
4.09172922e-01 -2.37438846e-02 2.22113952e-01 3.60398561e-01
-2.60510147e-01 -9.65334535e-01 5.81851959e-01 9.27299142e-01
3.83945823e-01 -7.84863055e-01 8.89720738e-01 3.18702787e-01
-8.48017275e-01 -4.81328160e-01 -4.23301518e-01 -1.63087949e-01
-1.47591531e-01 6.94933176e-01 -3.79847705e-01 3.65932316e-01
9.15277362e-01 1.16913450e+00 -7.16074109e-01 1.08686602e+00
1.62299976e-01 6.79312944e-01 -2.20655188e-01 5.42525172e-01
3.52453202e-01 1.11314654e-01 8.62201393e-01 1.23187780e+00
5.12764812e-01 -5.69244511e-02 3.05211097e-01 7.02693939e-01
-1.05717763e-01 -1.51263878e-01 -8.87182176e-01 -1.89079456e-02
1.06745332e-01 7.83206940e-01 -4.77736205e-01 -3.82822156e-01
-6.95185363e-01 1.20921946e+00 3.57282639e-01 3.02260131e-01
-7.38699615e-01 -1.57808959e-01 8.53494704e-01 2.94098824e-01
5.23705781e-01 2.18761954e-02 8.83301646e-02 -1.42469466e+00
4.71599966e-01 -1.00597417e+00 8.34199488e-01 -2.80240208e-01
-1.40264440e+00 4.70717967e-01 -3.07395548e-01 -1.53224766e+00
-5.64519048e-01 -7.30021298e-01 -9.28714097e-01 3.13386470e-01
-1.49419975e+00 -8.05541337e-01 -3.35335433e-01 1.06666386e+00
5.54263830e-01 -4.84357476e-01 6.96440995e-01 3.77283037e-01
-7.78623581e-01 7.20797300e-01 8.96430090e-02 6.07628167e-01
3.69077414e-01 -1.19093847e+00 3.62192810e-01 1.28088117e+00
1.76151156e-01 4.89292294e-01 5.05119205e-01 -4.05292064e-01
-1.37517977e+00 -1.42530394e+00 3.82639050e-01 -4.04361933e-01
6.84443772e-01 -2.29111880e-01 -1.48147237e+00 1.03790081e+00
-2.83015579e-01 7.19205976e-01 8.52367520e-01 -1.65823624e-01
-5.75302482e-01 -1.06441882e-02 -9.79647875e-01 3.89355093e-01
1.16738307e+00 -5.05897880e-01 -5.59587240e-01 1.06337458e-01
4.43590730e-01 -3.07067662e-01 -4.29977715e-01 6.31172597e-01
3.40429574e-01 -1.39389706e+00 9.67195034e-01 -1.06418526e+00
6.62531853e-01 -2.23914012e-01 -2.40929604e-01 -1.14052057e+00
-2.20251411e-01 -3.16097319e-01 -1.12543309e+00 7.65451729e-01
3.60007621e-02 -6.21421099e-01 7.32389450e-01 3.64472032e-01
-4.47536588e-01 -9.16429281e-01 -1.15082562e+00 -7.60149002e-01
-2.44122908e-01 -5.21049500e-01 5.62934220e-01 8.91304851e-01
-2.04421923e-01 -3.52882117e-01 -5.21596074e-01 4.61072087e-01
3.77270013e-01 -8.40268843e-03 5.53978384e-01 -1.27159297e+00
-4.44441855e-01 -3.64318520e-01 -1.10693169e+00 -8.37174416e-01
3.16225588e-01 -6.70315087e-01 -1.30804598e-01 -7.77875960e-01
1.82130247e-01 9.83294994e-02 -7.96802819e-01 4.77958858e-01
-1.63541585e-01 2.32910246e-01 -4.22879197e-02 4.22398806e-01
-1.07034707e+00 5.18766165e-01 7.59305954e-01 -1.51772738e-01
4.83014844e-02 -1.09894663e-01 -3.92013341e-01 1.15284896e+00
5.73700249e-01 -3.33775014e-01 -3.65721345e-01 -5.43161213e-01
2.31820494e-01 -2.15418413e-01 8.88705015e-01 -1.38179171e+00
2.70885915e-01 1.91403061e-01 8.48158956e-01 -4.40520257e-01
2.03295305e-01 -9.03361499e-01 -3.17697138e-01 5.63082039e-01
-3.75631094e-01 3.71278763e-01 1.13492399e-01 1.03796327e+00
-4.42237794e-01 5.90102747e-02 6.89372659e-01 -1.54198959e-01
-9.95698273e-01 7.38687754e-01 -4.21540916e-01 1.86326027e-01
1.11624920e+00 -2.06979379e-01 2.22179554e-02 -2.81915516e-01
-8.66121054e-01 7.61652291e-02 4.42815214e-01 3.82388055e-01
1.07116437e+00 -1.42105472e+00 -5.55777967e-01 7.16968060e-01
3.72590423e-01 -3.65179009e-03 4.41887647e-01 1.05116844e+00
-4.79019910e-01 7.45083615e-02 -5.02204835e-01 -8.02308381e-01
-8.50514233e-01 8.25636923e-01 4.25043434e-01 -2.56052911e-01
-1.18803275e+00 6.94555879e-01 4.75261271e-01 7.87293017e-02
4.91737574e-01 -3.13603163e-01 -2.98109382e-01 -7.67881647e-02
9.12372589e-01 4.03574526e-01 2.03869745e-01 -5.81722736e-01
-2.39675045e-01 1.72262684e-01 -4.21288401e-01 4.36452329e-01
1.27731502e+00 1.00041002e-01 -1.46746367e-01 5.62193751e-01
1.08776486e+00 -4.70897228e-01 -1.41168833e+00 -5.41155279e-01
2.03003824e-01 -6.70254052e-01 -6.10987991e-02 -3.60185623e-01
-1.50054991e+00 8.64124537e-01 8.44681978e-01 1.07112303e-01
1.52692199e+00 -1.64452866e-01 1.10057223e+00 7.03192890e-01
-4.85748015e-02 -8.72408152e-01 4.50678676e-01 5.82584798e-01
6.36145651e-01 -1.17142308e+00 -4.62844461e-01 1.15792759e-01
-4.04987156e-01 1.26581490e+00 8.04756224e-01 -5.09524107e-01
6.31578922e-01 1.10495679e-01 -2.55479693e-01 -2.98535526e-01
-8.78782749e-01 -9.35677066e-02 2.48832718e-01 4.97140735e-01
7.34743774e-02 -3.22540969e-01 1.70671061e-01 6.92280591e-01
2.99602002e-01 -9.27890241e-02 6.04494631e-01 1.11756814e+00
-4.31554347e-01 -5.05245149e-01 -1.83231741e-01 9.52053607e-01
-8.12139511e-01 -6.49739653e-02 -1.98672548e-01 3.85697871e-01
1.95896640e-01 4.77545947e-01 6.86211288e-01 -3.39896291e-01
3.89902502e-01 3.51482570e-01 1.52905211e-01 -3.88265699e-01
-1.64233923e-01 -3.71706843e-01 -4.30274308e-01 -1.06310141e+00
-2.82955140e-01 -6.52547598e-01 -1.04026508e+00 -2.75456250e-01
1.78811595e-01 -4.63287197e-02 -4.92696203e-02 1.07038355e+00
4.88769114e-01 5.91323435e-01 7.58542418e-01 -7.18809545e-01
-6.46316946e-01 -6.84419632e-01 -6.79636657e-01 7.88039505e-01
1.03938556e+00 -6.78126037e-01 -4.99210685e-01 7.94011280e-02] | [7.817803382873535, 1.7303885221481323] |
dad6658c-b9f4-4c68-a19f-5b88c8c4b48e | zero-shot-sketch-based-image-retrieval-using | 2201.10185 | null | https://arxiv.org/abs/2201.10185v2 | https://arxiv.org/pdf/2201.10185v2.pdf | Zero-Shot Sketch Based Image Retrieval using Graph Transformer | The performance of a zero-shot sketch-based image retrieval (ZS-SBIR) task is primarily affected by two challenges. The substantial domain gap between image and sketch features needs to be bridged, while at the same time the side information has to be chosen tactfully. Existing literature has shown that varying the semantic side information greatly affects the performance of ZS-SBIR. To this end, we propose a novel graph transformer based zero-shot sketch-based image retrieval (GTZSR) framework for solving ZS-SBIR tasks which uses a novel graph transformer to preserve the topology of the classes in the semantic space and propagates the context-graph of the classes within the embedding features of the visual space. To bridge the domain gap between the visual features, we propose minimizing the Wasserstein distance between images and sketches in a learned domain-shared space. We also propose a novel compatibility loss that further aligns the two visual domains by bridging the domain gap of one class with respect to the domain gap of all other classes in the training set. Experimental results obtained on the extended Sketchy, TU-Berlin, and QuickDraw datasets exhibit sharp improvements over the existing state-of-the-art methods in both ZS-SBIR and generalized ZS-SBIR. | ['Biplab Banerjee', 'Ushasi Chaudhuri', 'Sumrit Gupta'] | 2022-01-25 | null | null | null | null | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 7.37211481e-02 -3.62594575e-01 -2.38728374e-01 -3.26887637e-01
-8.04947317e-01 -5.51229060e-01 6.52465343e-01 -2.24633187e-01
-7.71518471e-03 3.35789919e-01 3.98578644e-02 1.50361508e-01
-7.67129898e-01 -8.16752791e-01 -4.49340254e-01 -7.31912017e-01
3.48309487e-01 4.20713931e-01 4.08083946e-01 -5.20507872e-01
5.06918550e-01 4.14122611e-01 -1.39005661e+00 1.91117361e-01
6.41035914e-01 1.12537825e+00 3.82980466e-01 2.21897259e-01
-3.68642688e-01 5.20058393e-01 -2.55624533e-01 -4.09202456e-01
5.00836372e-01 -2.99364001e-01 -6.23166800e-01 1.44887522e-01
9.01974678e-01 -3.41304451e-01 -7.52980590e-01 1.06216669e+00
4.22608823e-01 3.90837222e-01 8.06368768e-01 -1.66000664e+00
-1.05542934e+00 -5.12772314e-02 -5.79199255e-01 -2.09266365e-01
2.38008976e-01 -2.69140840e-01 1.38988113e+00 -1.17868376e+00
1.14379430e+00 1.50874031e+00 3.48351687e-01 4.48314816e-01
-1.07560134e+00 -7.98889279e-01 3.59747261e-01 4.06422257e-01
-1.70429957e+00 -1.04287282e-01 1.21554482e+00 -5.10999024e-01
4.46240604e-01 8.87667388e-02 4.15947407e-01 9.14853156e-01
-1.84403941e-01 6.18017077e-01 7.30790317e-01 -3.87173295e-01
2.80430287e-01 1.76450104e-01 9.58253518e-02 7.71716118e-01
3.62667255e-02 -3.68250161e-02 -8.41892064e-01 -2.01299727e-01
1.12629652e+00 4.47424293e-01 -1.30712897e-01 -1.22218370e+00
-9.15039420e-01 9.74811316e-01 6.62322879e-01 4.96884912e-01
7.45321587e-02 1.03922635e-01 3.68655980e-01 4.96115416e-01
5.01814306e-01 2.61513859e-01 -9.43676159e-02 3.37909281e-01
-8.60409617e-01 2.06523567e-01 4.92730409e-01 1.21482909e+00
8.92888725e-01 -3.73141140e-01 -3.32387120e-01 1.30702651e+00
2.57214636e-01 4.61660862e-01 9.80431214e-02 -6.95953488e-01
5.04601717e-01 6.96324766e-01 9.73490532e-03 -1.37957454e+00
4.36051488e-01 -1.78724155e-02 -5.78938365e-01 1.70825467e-01
2.52712518e-01 6.51058137e-01 -8.21617305e-01 1.75386214e+00
2.95099616e-01 1.86346248e-01 -1.59512490e-01 1.14536166e+00
9.75152254e-01 4.91365641e-01 -1.88701272e-01 3.10391486e-01
1.24024212e+00 -9.68220472e-01 -6.00306332e-01 -1.28369957e-01
1.24506950e-01 -7.51145363e-01 1.39138079e+00 -2.82222200e-02
-8.36182475e-01 -4.28217232e-01 -1.32404292e+00 -4.42709684e-01
-6.36660874e-01 2.27634851e-02 2.33769506e-01 2.18111262e-01
-9.02571678e-01 6.10435545e-01 -2.62556970e-01 -5.40730178e-01
3.94862503e-01 6.10961914e-02 -5.77215254e-01 -5.82967281e-01
-1.21727514e+00 6.70325100e-01 6.77211285e-02 -2.76209950e-01
-6.76426888e-01 -9.79937315e-01 -9.55820024e-01 2.25409999e-01
5.96960604e-01 -5.67274094e-01 5.82609534e-01 -8.34287703e-01
-1.22963095e+00 8.79046798e-01 6.24346919e-02 2.17986420e-01
6.93751395e-01 9.93071496e-02 -1.53529644e-01 3.46455097e-01
2.39528462e-01 6.76938891e-01 1.15753675e+00 -1.39080036e+00
-3.54740888e-01 -6.66502893e-01 1.62805840e-01 3.62450629e-01
-4.87564266e-01 -3.55179131e-01 -7.03054368e-01 -8.30597997e-01
1.46661624e-01 -8.68593872e-01 1.31367102e-01 9.03750658e-01
-6.42955163e-03 -5.82337856e-01 1.33402920e+00 -5.14554322e-01
9.20290887e-01 -2.53458643e+00 3.76914412e-01 2.88894534e-01
2.02017605e-01 2.07300514e-01 -6.55491889e-01 7.57515788e-01
6.99062943e-02 -1.75260589e-01 -1.31779104e-01 -2.61353999e-01
1.14775173e-01 3.90735567e-01 -5.74821234e-01 3.80838752e-01
9.13729221e-02 8.10462594e-01 -1.18129563e+00 -5.34467161e-01
4.00198609e-01 6.37886882e-01 -3.58300805e-01 3.13828796e-01
-2.09983103e-02 8.57955441e-02 -5.93480349e-01 5.02146184e-01
7.81194210e-01 -2.78059334e-01 1.04204804e-01 -2.09372267e-01
2.30867445e-01 -7.69871324e-02 -1.15699041e+00 2.20494103e+00
-4.29945916e-01 5.49453974e-01 -6.27982393e-02 -9.65124726e-01
1.21940017e+00 -2.33435500e-02 5.23113251e-01 -9.15916443e-01
-3.69659662e-01 1.97068453e-01 -5.88733375e-01 -1.21563107e-01
3.28639448e-01 -3.57741475e-01 6.42422065e-02 2.97772020e-01
2.74236917e-01 -3.35212260e-01 -1.09909229e-01 5.73953629e-01
7.92682528e-01 3.09782475e-02 1.68114491e-02 -4.04284507e-01
4.65120703e-01 -2.20806867e-01 3.43667626e-01 6.14816844e-01
-1.23537384e-01 7.72275984e-01 3.45996141e-01 -4.32782441e-01
-1.14238107e+00 -1.43831706e+00 -9.65287443e-03 1.19685757e+00
8.89415562e-01 -4.89058435e-01 -3.13511699e-01 -8.14264119e-01
3.15412134e-01 4.46775049e-01 -6.58847094e-01 -4.19027865e-01
-1.84936211e-01 2.64257461e-01 1.76718369e-01 4.33873534e-01
5.04888177e-01 -7.36124516e-01 -1.83506593e-01 -1.19242407e-01
-1.41059965e-01 -1.01071680e+00 -9.53241229e-01 -3.37488592e-01
-5.69682598e-01 -1.19354153e+00 -9.88572776e-01 -1.00926435e+00
8.25367093e-01 9.59801853e-01 8.85617971e-01 2.72126198e-01
-6.03987932e-01 6.82319641e-01 -3.78319353e-01 9.36786085e-02
2.86834776e-01 -2.92768985e-01 -3.94532442e-01 1.11873329e-01
1.89730957e-01 -5.04730940e-01 -8.35392833e-01 6.08224511e-01
-9.58122909e-01 -1.38062164e-01 2.13176608e-01 1.17775464e+00
5.73636830e-01 -1.05899617e-01 6.01910353e-01 -5.53282619e-01
5.45982897e-01 -2.45979965e-01 -5.93369186e-01 7.96024263e-01
-6.53375864e-01 2.08381355e-01 3.20563644e-01 -5.99139154e-01
-7.86209941e-01 -1.15568921e-01 4.48007643e-01 -9.57869470e-01
4.33700889e-01 1.50838390e-01 -2.04126209e-01 -3.61197323e-01
1.99914992e-01 1.61580026e-01 1.81521177e-01 -4.22590762e-01
7.70861804e-01 5.70025027e-01 2.44179577e-01 -6.47780240e-01
8.73410702e-01 7.03153253e-01 1.18828870e-01 -1.04361486e+00
-7.16614306e-01 -8.73192191e-01 -5.64319372e-01 -9.52303335e-02
6.51043296e-01 -8.56572509e-01 -3.35559100e-01 1.42623052e-01
-9.87536252e-01 6.42265081e-02 -3.85843813e-01 2.43895706e-02
-5.68500578e-01 7.00422227e-01 -2.19577894e-01 -6.52979732e-01
-3.98063123e-01 -1.12567472e+00 1.50942087e+00 -3.83835584e-02
6.58905879e-02 -9.09312487e-01 2.43644297e-01 1.63294613e-01
2.57798314e-01 4.18478511e-02 1.21044350e+00 -3.92579019e-01
-7.33220994e-01 -2.11595148e-01 -7.82252371e-01 3.21114987e-01
2.37120882e-01 -7.59436712e-02 -6.61908925e-01 -5.07353008e-01
-4.12560672e-01 -3.15982461e-01 1.03344131e+00 1.71815082e-01
9.95474219e-01 -1.65666446e-01 -3.33761454e-01 4.37796116e-01
1.64423990e+00 6.01342097e-02 7.03623414e-01 4.84288335e-02
6.42401814e-01 6.74182355e-01 8.99756610e-01 3.55745584e-01
3.01833540e-01 9.21649218e-01 2.17369646e-01 -3.95848937e-02
-4.36074942e-01 -7.39192605e-01 7.21311290e-03 4.75437999e-01
3.65104556e-01 -4.69252765e-02 -3.91581893e-01 7.64340580e-01
-2.06218791e+00 -8.80778432e-01 4.59982336e-01 2.42541146e+00
5.49232185e-01 -3.87166977e-01 -8.53366107e-02 -7.10526016e-04
8.46678138e-01 5.08579254e-01 -4.66218889e-01 -1.35411888e-01
7.17230588e-02 1.50228381e-01 2.44996861e-01 5.24305880e-01
-9.01259065e-01 1.05142033e+00 5.25108194e+00 1.25741816e+00
-1.10689259e+00 4.16961946e-02 4.01193984e-02 -9.79264453e-02
-3.89801294e-01 1.82993561e-01 -3.87971312e-01 3.03308785e-01
-9.69460383e-02 -3.74515712e-01 6.94134712e-01 9.34008658e-01
-3.15707058e-01 1.93420365e-01 -1.27855599e+00 1.42171574e+00
3.12466770e-01 -1.25820529e+00 4.27367389e-01 3.48285697e-02
6.57730758e-01 -3.35609347e-01 2.26955935e-01 1.18391305e-01
1.06285416e-01 -7.55113304e-01 6.40017569e-01 4.29304689e-01
1.24889445e+00 -6.30701840e-01 2.06879243e-01 3.12466975e-02
-1.51902127e+00 1.07597199e-03 -7.49125957e-01 2.15424418e-01
-3.44047874e-01 2.68281311e-01 -6.40318215e-01 5.87918580e-01
4.99148518e-01 1.06269419e+00 -4.55749393e-01 7.24422991e-01
-1.19696364e-01 -2.74502873e-01 3.06046046e-02 4.53636982e-02
2.82343298e-01 -5.43957472e-01 6.44101679e-01 7.01430857e-01
3.55425805e-01 -1.45189622e-02 2.91938215e-01 1.08112657e+00
-1.35608658e-01 6.49510995e-02 -1.11417985e+00 -1.49358004e-01
5.53443730e-01 1.19479823e+00 -4.27644521e-01 -3.16483945e-01
-2.66101718e-01 1.18937218e+00 4.07948852e-01 5.29639423e-01
-5.47758996e-01 -6.15680575e-01 9.48016286e-01 1.92093998e-01
4.83406633e-01 -1.56025276e-01 -1.49517611e-01 -1.21366096e+00
9.69553813e-02 -4.11431670e-01 4.80105519e-01 -8.47392559e-01
-1.76585031e+00 2.94228911e-01 -8.41068178e-02 -1.40139246e+00
1.24197058e-01 -5.25074363e-01 -3.80558431e-01 7.42780387e-01
-1.63862360e+00 -1.41658616e+00 -4.26260054e-01 9.91920412e-01
7.56707489e-01 -1.70540884e-01 6.93257451e-01 3.37476641e-01
-1.15338370e-01 6.59414828e-01 2.43791878e-01 1.89238772e-01
9.82293487e-01 -9.34080362e-01 1.94517478e-01 3.56086552e-01
2.96925545e-01 7.65283287e-01 2.67044753e-01 -6.88219607e-01
-1.63640130e+00 -9.95146275e-01 7.28060246e-01 -1.65503219e-01
6.24356866e-01 -6.26705825e-01 -8.83018672e-01 3.51420015e-01
-1.22136265e-01 2.95252651e-01 3.04234982e-01 4.79145609e-02
-1.14533746e+00 -2.94553429e-01 -1.22158694e+00 5.95113039e-01
1.32629728e+00 -1.07102334e+00 -6.18623853e-01 3.68864179e-01
6.67406142e-01 4.83958758e-02 -8.46640229e-01 2.55392611e-01
9.81580913e-01 -6.21809602e-01 1.36174381e+00 -6.81497931e-01
4.73852128e-01 -1.55162856e-01 -4.53436702e-01 -1.26157308e+00
-3.55362594e-01 -1.36180565e-01 4.23050344e-01 1.14721835e+00
-1.71538889e-01 -4.48755413e-01 5.43997765e-01 5.62779725e-01
4.15290207e-01 -5.03947139e-01 -1.03270864e+00 -1.07040989e+00
6.12800680e-02 1.41498938e-01 4.80740726e-01 9.79589105e-01
-1.33925152e-03 4.38937694e-01 -4.68033016e-01 -1.51277021e-01
8.80706429e-01 3.85097504e-01 6.92152262e-01 -1.32619548e+00
1.08675160e-01 -3.34697604e-01 -6.87882543e-01 -8.80674958e-01
1.72369972e-01 -9.44825470e-01 -8.63515586e-02 -1.62747300e+00
4.35596824e-01 -6.06058836e-01 -5.45821071e-01 3.99724662e-01
3.78273949e-02 4.71689962e-02 5.13617814e-01 2.42948309e-01
-7.56558299e-01 8.39778662e-01 1.48409438e+00 -4.19072121e-01
1.87770396e-01 -5.24616778e-01 -4.61893052e-01 2.54185975e-01
2.10832521e-01 -3.73016238e-01 -8.72124553e-01 -3.93641114e-01
6.84942752e-02 1.46836385e-01 5.70589662e-01 -4.72868204e-01
2.70655751e-01 -2.50917941e-01 3.40741277e-02 -6.00359797e-01
5.28734088e-01 -1.06387079e+00 -2.68896278e-02 1.98952869e-01
-5.61302185e-01 -3.25389951e-01 -2.70256728e-01 9.65918124e-01
-2.09634230e-01 -3.55710573e-02 9.15700853e-01 -1.40644804e-01
-8.33621681e-01 5.59271395e-01 3.95543516e-01 1.00710645e-01
1.01628745e+00 -3.44507992e-01 -4.11612719e-01 -3.60203624e-01
-3.42321754e-01 2.30061650e-01 6.47081554e-01 9.09927249e-01
1.09685624e+00 -1.67111897e+00 -3.34434360e-01 4.75228727e-01
6.71959817e-01 -2.98962921e-01 4.29684877e-01 3.74697804e-01
-2.22298086e-01 3.96564841e-01 -4.77860570e-01 -5.60328543e-01
-1.42647469e+00 7.91386843e-01 9.70597714e-02 -1.71400353e-01
-8.35836053e-01 7.82877207e-01 7.47424603e-01 -3.98983687e-01
4.30869132e-01 4.49398607e-02 7.36548752e-02 5.78109063e-02
4.75178033e-01 4.07124102e-01 -9.39954296e-02 -4.12955761e-01
-5.15301347e-01 9.69623506e-01 -7.14011416e-02 -1.48819759e-01
1.27242517e+00 -2.38942653e-01 -1.64594516e-01 4.35739130e-01
1.53201711e+00 -3.35902601e-01 -1.28173327e+00 -4.68564421e-01
1.13234613e-02 -1.08528817e+00 7.15310872e-02 -6.57790244e-01
-1.08615005e+00 1.04319155e+00 6.27120137e-01 -1.02247283e-01
9.32580173e-01 1.25881687e-01 8.84707749e-01 2.56419480e-01
6.17146432e-01 -1.16084313e+00 6.77050591e-01 2.17258826e-01
1.20238698e+00 -1.12918317e+00 -1.12777770e-01 -5.69244981e-01
-6.41722798e-01 1.09854639e+00 3.24753791e-01 -5.62103271e-01
7.37068176e-01 -2.68579602e-01 -3.22179019e-01 -4.26226079e-01
-5.17204702e-01 -8.19684640e-02 5.84564090e-01 6.25636578e-01
4.08073561e-03 3.69110666e-02 -2.88590848e-01 3.66102010e-01
4.55183029e-01 4.70427871e-02 -5.74172400e-02 9.06677246e-01
-3.93991649e-01 -1.17111099e+00 -7.74608478e-02 1.17021069e-01
1.88325793e-01 -3.37917991e-02 -7.90916502e-01 6.76459849e-01
-1.68723524e-01 7.33253777e-01 -6.31914586e-02 -2.91224629e-01
3.39813799e-01 -8.94630328e-02 7.63967216e-01 -4.79104370e-01
-5.28475316e-03 -3.98427658e-02 -2.69719958e-01 -5.80077529e-01
-1.50378704e-01 -1.07299455e-01 -9.12100554e-01 -1.33041948e-01
-3.53944123e-01 -6.43372070e-04 6.58405900e-01 6.59733713e-01
4.23765332e-01 6.10819310e-02 7.96974897e-01 -5.47631323e-01
-7.70546377e-01 -4.94293451e-01 -9.80845332e-01 6.81358099e-01
1.73844874e-01 -1.19846463e+00 -2.71375865e-01 -2.65136659e-01] | [11.610203742980957, 0.6938369870185852] |
457a8a87-8eeb-45f9-a34a-37d23f4c2063 | convolutional-kitchen-sinks-for-transcription | 1706.00125 | null | http://arxiv.org/abs/1706.00125v1 | http://arxiv.org/pdf/1706.00125v1.pdf | Convolutional Kitchen Sinks for Transcription Factor Binding Site Prediction | We present a simple and efficient method for prediction of transcription
factor binding sites from DNA sequence. Our method computes a random
approximation of a convolutional kernel feature map from DNA sequence and then
learns a linear model from the approximated feature map. Our method outperforms
state-of-the-art deep learning methods on five out of six test datasets from
the ENCODE consortium, while training in less than one eighth the time. | [] | 2017-05-31 | null | null | null | null | ['transcription-factor-binding-site-prediction'] | ['medical'] | [ 4.56878655e-02 -3.60705494e-03 -3.25929165e-01 -6.29628897e-01
-1.34129131e+00 -7.98872113e-01 2.69894123e-01 1.81079060e-01
-8.13398063e-01 8.94756973e-01 3.16666245e-01 -3.34085196e-01
-2.17450172e-01 -5.44009030e-01 -1.19709718e+00 -1.10189140e+00
-2.05951348e-01 6.23308182e-01 3.93578917e-01 2.80405534e-03
1.04783893e-01 4.16409194e-01 -1.20222497e+00 9.42923844e-01
3.81398886e-01 7.48260975e-01 1.06086582e-02 1.03500652e+00
-2.53751855e-02 6.83652237e-02 -3.88078466e-02 -3.38964820e-01
3.23339961e-02 -1.63757071e-01 -9.13290620e-01 -6.51264668e-01
2.95791656e-01 -2.91539460e-01 -8.31439793e-01 4.47174639e-01
9.27217603e-01 1.68490440e-01 1.00391233e+00 -6.04634643e-01
-5.87014318e-01 3.67284387e-01 -2.05559611e-01 5.11349440e-01
4.63107646e-01 7.94609487e-02 1.12703037e+00 -1.18873155e+00
8.77006888e-01 9.43482816e-01 1.10700595e+00 3.64490926e-01
-1.57750905e+00 -4.45452422e-01 -6.59295380e-01 7.44939446e-02
-1.92531085e+00 -3.07071120e-01 -2.03908920e-01 -4.40403074e-01
1.81936777e+00 -2.65798345e-02 4.14546609e-01 8.79670680e-01
5.82669437e-01 8.21028352e-01 6.31529808e-01 -2.95792937e-01
2.67210752e-01 -7.58084238e-01 3.70041311e-01 1.28722417e+00
-1.42805129e-01 -2.24021986e-01 -5.43234825e-01 -1.01270914e+00
5.41543245e-01 -2.22663954e-01 -9.57719162e-02 -4.09829803e-02
-1.15506601e+00 5.95046401e-01 9.10756961e-02 2.83408552e-01
-2.10210383e-01 5.65379739e-01 4.71154362e-01 2.40880176e-01
2.95377970e-01 1.70970306e-01 -8.47591698e-01 -4.62649971e-01
-8.64847481e-01 5.74430585e-01 7.52115250e-01 4.48182464e-01
1.03463805e+00 -4.37281638e-01 -1.13402061e-01 6.74947023e-01
2.15655118e-01 2.23172352e-01 4.74705964e-01 -5.22357225e-01
-2.51777232e-01 3.10242355e-01 1.89786687e-01 -5.03128409e-01
-7.73376644e-01 -1.13989852e-01 -5.73257327e-01 -1.69977531e-01
6.97340429e-01 -1.26591712e-01 -1.13433218e+00 1.24850440e+00
2.47721761e-01 7.22422719e-01 -1.68602958e-01 8.74509394e-01
6.84575617e-01 9.14541841e-01 -1.31551668e-01 3.62387359e-01
1.25787771e+00 -5.85086048e-01 -3.90022516e-01 3.48348431e-02
9.76391017e-01 -3.60641390e-01 6.00023568e-01 5.12549758e-01
-7.72289097e-01 -1.82767972e-01 -9.21836138e-01 -5.73713601e-01
-5.01787722e-01 2.25492015e-01 9.28030908e-01 6.13660395e-01
-1.01248300e+00 7.16049135e-01 -8.96837354e-01 -3.43211561e-01
2.24360019e-01 6.65510774e-01 -9.51418817e-01 -2.60165244e-01
-1.43852425e+00 5.45455813e-01 6.00032687e-01 -4.14333381e-02
-9.22356784e-01 -1.08044875e+00 -5.31017482e-01 5.23759782e-01
-4.20201659e-01 -3.60849291e-01 1.15906692e+00 -2.94617444e-01
-1.32196856e+00 1.05755782e+00 -2.78971821e-01 -2.57964313e-01
-2.60769278e-01 -3.89517188e-01 -1.17554270e-01 -9.29306671e-02
-2.21972346e-01 4.36235905e-01 2.28793606e-01 -3.34508628e-01
-5.30615568e-01 -1.41604349e-01 -2.57052809e-01 -3.54028761e-01
-3.07214092e-02 1.45216715e-02 -6.31191373e-01 -2.64088988e-01
-2.55875379e-01 -9.72274423e-01 -2.52573848e-01 7.25921094e-02
-9.24540609e-02 -5.42763472e-01 2.27255166e-01 -7.76125073e-01
1.09884906e+00 -2.11357594e+00 2.98162043e-01 3.87762219e-01
1.03144042e-01 5.32542646e-01 -4.21097964e-01 7.61926055e-01
-4.13639814e-01 -1.26218945e-02 1.52979299e-01 2.71110445e-01
1.54339343e-01 1.77911058e-01 -1.43148765e-01 1.00766087e+00
1.34665677e-02 9.89646733e-01 -1.06090343e+00 2.51660794e-01
-4.09883916e-01 8.13268542e-01 -7.12152183e-01 1.26569554e-01
-3.66122812e-01 -5.59852272e-03 -3.11275840e-01 5.09346664e-01
5.97669184e-01 -4.43233311e-01 4.34769839e-01 -1.44831896e-01
4.02505212e-02 2.53676325e-01 -7.89666533e-01 2.27749348e+00
1.07576720e-01 7.52578914e-01 -4.20981139e-01 -1.09775841e+00
8.91120613e-01 4.74636585e-01 4.09098476e-01 1.22354180e-02
-5.78379072e-03 2.09581509e-01 2.37115212e-02 -5.68650246e-01
-1.84486106e-01 9.01209190e-03 -1.10191040e-01 5.72416604e-01
6.54186785e-01 4.87793565e-01 -1.58105448e-01 6.00649649e-03
1.79620194e+00 1.51201129e-01 5.58767021e-01 -5.70778608e-01
2.22019523e-01 -5.34114897e-01 6.12645507e-01 6.85250521e-01
-3.34509127e-02 4.96918947e-01 8.02057028e-01 -1.29806626e+00
-1.34886658e+00 -1.03068972e+00 -5.55922799e-02 1.51542091e+00
-7.03543425e-01 -6.86619997e-01 -8.67405593e-01 -8.45847368e-01
3.37944627e-01 1.98281363e-01 -1.02247024e+00 -1.77270725e-01
-4.59550530e-01 -9.59757566e-01 1.59150004e+00 7.41296351e-01
-1.48201287e-01 -4.51267153e-01 8.66345316e-02 8.80384967e-02
-5.29603250e-02 -3.57531101e-01 -5.25996864e-01 8.60266268e-01
-1.50993466e-01 -9.46751118e-01 -4.85027879e-01 -8.28708768e-01
3.71741384e-01 1.81459114e-02 8.58072996e-01 2.03399107e-01
-7.03020394e-01 -1.80681825e-01 1.39575034e-01 -1.72341615e-01
-1.12811252e-01 4.20364380e-01 1.82654828e-01 -2.80543387e-01
1.17330194e+00 -4.46684033e-01 -5.81802905e-01 -2.35496145e-02
-1.03977478e+00 -3.50622416e-01 5.98735929e-01 1.12390447e+00
6.18629575e-01 -4.49207664e-01 6.39693022e-01 -8.48030925e-01
6.13001406e-01 -7.91760325e-01 -6.90165281e-01 3.35180342e-01
-3.08091134e-01 5.95335305e-01 5.45817614e-01 -7.60250837e-02
-1.16904283e+00 8.25295210e-01 -7.22429574e-01 -2.06276197e-02
-3.94114107e-01 8.35486710e-01 1.73676834e-01 -1.31206602e-01
1.19706130e+00 4.03141111e-01 -2.61912197e-01 -6.05509579e-01
5.07708132e-01 6.87712848e-01 4.93252814e-01 -5.03700078e-01
3.40874493e-01 5.11294305e-01 2.08361760e-01 -5.46076298e-01
-8.87967527e-01 -8.88177276e-01 -1.19146466e+00 8.57412666e-02
6.58900023e-01 -6.17407203e-01 -1.02867889e+00 4.68060285e-01
-1.00441098e+00 -2.18344197e-01 6.70246184e-01 4.47790474e-01
-7.49171257e-01 4.94773462e-02 -1.23648942e+00 -2.62065530e-01
-3.49793494e-01 -8.56431425e-01 1.40202522e+00 -5.05549684e-02
-1.14934362e-01 -7.21278608e-01 9.75817978e-01 9.21598151e-02
1.66587114e-01 -8.71719494e-02 1.14920497e+00 -9.45288122e-01
-4.75718617e-01 -5.16728878e-01 -1.15421504e-01 -1.58602998e-01
-5.47006726e-02 -1.60547048e-02 -1.12967348e+00 -4.42360580e-01
-1.02602661e+00 -7.11590707e-01 1.43573928e+00 2.67599344e-01
1.13823986e+00 -7.37059396e-03 -8.70062828e-01 8.66050184e-01
1.49110782e+00 1.64680388e-02 9.62266684e-01 3.17124635e-01
2.68710285e-01 -1.60391137e-01 5.25726497e-01 3.15550864e-01
-2.64609039e-01 6.55873716e-01 6.50070012e-02 2.01654248e-02
2.87175328e-01 -2.58370161e-01 2.44774297e-02 1.95877314e-01
-2.46262014e-01 -3.06041211e-01 -1.36526632e+00 4.33765084e-01
-2.23874903e+00 -8.84848177e-01 1.49957255e-01 1.74563706e+00
1.20785844e+00 -3.98608834e-01 2.00375598e-02 -1.40412331e-01
4.40673470e-01 -2.64274687e-01 -3.63145053e-01 -4.94286418e-01
4.13452648e-02 5.88101327e-01 5.73679030e-01 4.36632246e-01
-1.20186043e+00 1.01490974e+00 8.71335793e+00 9.29729104e-01
-8.53024364e-01 -8.48421305e-02 5.92624485e-01 -1.30740792e-01
-3.29234838e-01 -1.52811393e-01 -9.89793718e-01 3.14920694e-01
1.39498639e+00 -5.79803400e-02 4.48503643e-01 9.16971743e-01
-1.57070607e-01 2.26163730e-01 -1.23142767e+00 6.96054041e-01
-6.06582500e-03 -2.02439094e+00 -3.94322604e-01 1.66538209e-01
5.28324783e-01 5.05293250e-01 -1.00751810e-01 1.94799364e-01
4.30583119e-01 -1.46228266e+00 -5.61316013e-02 9.44849730e-01
5.66164315e-01 -1.17317665e+00 8.84093225e-01 3.72154653e-01
-8.58623624e-01 6.38648272e-02 -1.03972816e+00 -1.34181425e-01
-2.13817835e-01 9.87993181e-01 -1.28696346e+00 1.19920053e-01
8.13825846e-01 2.45298937e-01 -1.82158738e-01 1.09669995e+00
-6.52514324e-02 1.10609281e+00 -2.28569001e-01 -1.40413567e-01
4.79677290e-01 2.47780308e-01 1.75003074e-02 2.11955142e+00
3.04131389e-01 4.72444266e-01 -8.10216665e-02 3.31535101e-01
-3.20819050e-01 1.22471601e-01 -5.45535386e-01 -1.70275927e-01
2.43253052e-01 1.31373835e+00 -5.21683514e-01 -2.00332925e-01
-5.69083631e-01 1.07599473e+00 1.20813131e+00 3.84361893e-01
-7.55038023e-01 -8.62407506e-01 1.19308472e+00 -7.60159492e-02
5.78242064e-01 -3.46756488e-01 2.23890588e-01 -8.83419871e-01
-5.48644781e-01 -6.30326509e-01 4.76166666e-01 -5.95415175e-01
-1.26638782e+00 1.05512990e-02 -5.21861553e-01 -4.74954933e-01
-4.83769447e-01 -9.99811292e-01 -1.44295916e-01 1.04075122e+00
-9.25370455e-01 -8.77433240e-01 3.66015702e-01 2.01798305e-01
-6.23124987e-02 -2.85109907e-01 1.47318518e+00 5.39856255e-01
-3.23807895e-01 7.67443538e-01 8.21462870e-01 2.85397112e-01
8.44388425e-01 -1.28758335e+00 1.17330396e+00 1.51670530e-01
-4.09611538e-02 1.10003293e+00 5.08523524e-01 -3.37486744e-01
-1.83484089e+00 -9.08800244e-01 1.21590590e+00 -4.86476094e-01
6.05042815e-01 -7.44353950e-01 -1.23132432e+00 1.02459681e+00
2.43399650e-01 5.17391026e-01 1.45244467e+00 3.82934511e-01
-8.44432652e-01 1.34211719e-01 -1.04124284e+00 2.11871281e-01
8.92320395e-01 -6.82241976e-01 -3.71060759e-01 3.96800846e-01
4.51754868e-01 -4.43177760e-01 -1.13992703e+00 7.59136826e-02
1.29622543e+00 -5.57172179e-01 1.12007976e+00 -1.51698136e+00
2.05302641e-01 -3.92642796e-01 -2.84131672e-02 -1.52461231e+00
-1.33427858e+00 -5.41715920e-01 5.97469974e-03 5.32599926e-01
6.15234196e-01 -8.90439302e-02 1.15275824e+00 2.63627172e-01
-2.13751748e-01 -9.03380990e-01 -1.20923734e+00 -4.47865367e-01
2.06117749e-01 -3.95539589e-02 3.24235857e-01 9.29456472e-01
6.54914677e-01 2.23150909e-01 -4.57883179e-01 1.15060791e-01
1.82705462e-01 -7.61490092e-02 7.27001369e-01 -9.43877995e-01
-4.47467446e-01 5.62733673e-02 -9.26333547e-01 -1.16192436e+00
6.10347092e-01 -1.15353143e+00 3.57526690e-01 -1.42048383e+00
6.39577627e-01 1.92713797e-01 -7.83624232e-01 8.73899758e-01
-1.86720490e-01 4.42272007e-01 -6.86667740e-01 -3.93357664e-01
-5.84609866e-01 -5.85756488e-02 4.19555157e-01 -1.67308405e-01
4.84989643e-01 -2.32456580e-01 -3.70646745e-01 4.83195186e-01
6.42731309e-01 -6.46996379e-01 -1.29534841e-01 -3.24746192e-01
6.98711216e-01 -1.60213798e-01 -5.09075895e-02 -6.35857463e-01
2.44865999e-01 -2.98055977e-01 9.11910653e-01 -5.70264220e-01
2.71635056e-01 -1.71405345e-01 2.33458489e-01 3.67531896e-01
-7.37884700e-01 1.52749754e-02 3.88434619e-01 7.63774574e-01
2.78495312e-01 -2.84244269e-01 6.08579099e-01 -2.80208942e-02
-5.53160429e-01 1.14534497e-01 -1.20457017e+00 -5.49916744e-01
4.60300356e-01 3.65381718e-01 -4.38269258e-01 -1.14675649e-01
-7.54385471e-01 -3.35112154e-01 1.75242215e-01 1.06836624e-01
4.29127306e-01 -1.36587846e+00 -6.55766487e-01 1.98829889e-01
1.98300943e-01 -5.51799119e-01 1.08637586e-01 2.59622157e-01
-7.60577083e-01 1.22523594e+00 -2.98932016e-01 -4.57312882e-01
-1.23410225e+00 5.37989676e-01 4.85105783e-01 -5.96857071e-02
-3.14145446e-01 1.27635610e+00 1.87313050e-01 -7.49455035e-01
9.85071808e-02 -2.00220108e-01 2.75918305e-01 -5.19135714e-01
1.08138084e+00 2.62825489e-01 4.90966320e-01 -2.74181515e-01
-5.76669097e-01 2.13054419e-01 -2.41600215e-01 2.73514837e-02
1.29759622e+00 4.43330914e-01 -3.16134930e-01 2.96920300e-01
1.70708632e+00 -2.73078352e-01 -1.06270540e+00 -3.36472958e-01
2.25232169e-01 -6.09246910e-01 4.10402603e-02 -8.67526531e-01
-4.41006362e-01 5.80122888e-01 6.51004553e-01 -6.50269866e-01
7.47146428e-01 1.67884484e-01 7.87375033e-01 1.27514338e+00
2.63884693e-01 -9.26865339e-01 -3.37467119e-02 7.04240739e-01
3.80345285e-01 -6.63029909e-01 -2.28333607e-01 -1.82644233e-01
9.56599861e-02 1.34263241e+00 3.43358010e-01 -2.28860438e-01
9.53044057e-01 1.02352703e+00 3.09940670e-02 -2.04505637e-01
-1.45891452e+00 8.94246101e-02 4.84495163e-01 7.31520057e-01
1.26379919e+00 -1.02253795e-01 -4.78467047e-01 9.16029871e-01
3.52101386e-01 4.74047035e-01 1.57815050e-02 9.89729345e-01
-1.00934172e+00 -1.07670295e+00 -1.37935309e-02 5.10851502e-01
-8.64130139e-01 -4.22266066e-01 -3.90511036e-01 2.58607268e-01
-2.80656125e-02 2.02121645e-01 4.93583046e-02 -5.28587461e-01
-1.32039845e-01 7.01729596e-01 5.21564603e-01 -3.73435915e-01
-3.20869118e-01 2.57503271e-01 1.44083798e-01 -8.45764518e-01
1.11389622e-01 -6.79581404e-01 -1.77782631e+00 -4.33108717e-01
-1.53570443e-01 2.12960586e-01 6.63928151e-01 1.07668447e+00
7.28907764e-01 1.44665182e-01 8.23994726e-02 -6.90132678e-01
-3.90943885e-01 -8.51264000e-01 -5.93156874e-01 -7.76434615e-02
1.96171418e-01 -5.66576123e-01 2.05009013e-01 1.03814453e-01] | [4.750124931335449, 5.6684184074401855] |
e5538ddc-a578-4b02-b5ed-70240ce18e64 | meta-personalizing-vision-language-models-to-1 | 2306.10169 | null | https://arxiv.org/abs/2306.10169v1 | https://arxiv.org/pdf/2306.10169v1.pdf | Meta-Personalizing Vision-Language Models to Find Named Instances in Video | Large-scale vision-language models (VLM) have shown impressive results for language-guided search applications. While these models allow category-level queries, they currently struggle with personalized searches for moments in a video where a specific object instance such as ``My dog Biscuit'' appears. We present the following three contributions to address this problem. First, we describe a method to meta-personalize a pre-trained VLM, i.e., learning how to learn to personalize a VLM at test time to search in video. Our method extends the VLM's token vocabulary by learning novel word embeddings specific to each instance. To capture only instance-specific features, we represent each instance embedding as a combination of shared and learned global category features. Second, we propose to learn such personalization without explicit human supervision. Our approach automatically identifies moments of named visual instances in video using transcripts and vision-language similarity in the VLM's embedding space. Finally, we introduce This-Is-My, a personal video instance retrieval benchmark. We evaluate our approach on This-Is-My and DeepFashion2 and show that we obtain a 15% relative improvement over the state of the art on the latter dataset. | ['Simon Jenni', 'Fabian Caba Heilbron', 'Josef Sivic', 'Bryan Russell', 'Chun-Hsiao Yeh'] | 2023-06-16 | meta-personalizing-vision-language-models-to | http://openaccess.thecvf.com//content/CVPR2023/html/Yeh_Meta-Personalizing_Vision-Language_Models_To_Find_Named_Instances_in_Video_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Yeh_Meta-Personalizing_Vision-Language_Models_To_Find_Named_Instances_in_Video_CVPR_2023_paper.pdf | cvpr-2023-1 | ['word-embeddings'] | ['methodology'] | [-1.49490042e-02 -3.97167355e-01 -6.37127697e-01 -3.68222505e-01
-1.21230710e+00 -6.13723099e-01 9.07032073e-01 2.67911702e-01
-7.89164066e-01 7.86475241e-02 4.82671857e-01 1.63452625e-01
-5.98171428e-02 -2.23384619e-01 -7.36347377e-01 -4.78631109e-01
-3.09558421e-01 5.97239375e-01 2.15895727e-01 1.54916212e-01
3.64894509e-01 3.00488323e-01 -1.65691555e+00 6.66804314e-01
1.37493566e-01 1.25344384e+00 5.52628458e-01 6.83726430e-01
-2.12332740e-01 6.71746969e-01 -5.87647557e-01 -3.47435087e-01
1.39696941e-01 -2.37371568e-02 -1.06626129e+00 4.18529272e-01
1.10809565e+00 -3.02991003e-01 -8.34513783e-01 7.24239171e-01
4.58775461e-01 6.37260735e-01 7.69974649e-01 -1.34457362e+00
-1.13243783e+00 3.84406805e-01 -4.30958003e-01 5.36756694e-01
4.94839281e-01 3.06937426e-01 1.32291174e+00 -1.12094975e+00
1.02411437e+00 1.31501961e+00 4.48109269e-01 6.25892758e-01
-1.15983760e+00 -3.49614799e-01 2.64061302e-01 9.43594813e-01
-1.68792534e+00 -3.89889985e-01 5.84063113e-01 -5.27278543e-01
1.25574207e+00 1.22121081e-01 5.62566578e-01 1.43427920e+00
3.03677116e-02 1.14379251e+00 5.20794332e-01 -3.11983526e-01
3.35605070e-02 2.44141653e-01 1.56469494e-01 8.60667586e-01
-1.98686019e-01 -3.57894480e-01 -9.08409297e-01 -3.75922650e-01
5.38191378e-01 2.38687862e-02 -3.13338280e-01 -6.67357087e-01
-1.44760025e+00 9.03397858e-01 1.95747912e-01 4.05685365e-01
-3.66800845e-01 5.43757737e-01 8.32818091e-01 2.94761419e-01
4.02051568e-01 6.25783145e-01 -4.63503480e-01 -2.09264308e-01
-1.03916144e+00 2.37094447e-01 6.25652134e-01 1.18099749e+00
9.36231494e-01 -5.60286522e-01 -7.18709469e-01 1.17310107e+00
2.79997706e-01 3.49284530e-01 8.86191010e-01 -1.15178657e+00
2.29985714e-01 2.47918040e-01 -1.18368119e-01 -1.01014614e+00
9.51708034e-02 -8.89778435e-02 -3.13756198e-01 -4.82526571e-01
-2.68728971e-01 4.82638896e-01 -9.84506071e-01 1.68508112e+00
6.51822798e-03 3.93452108e-01 -9.31254178e-02 7.31757224e-01
9.79983151e-01 9.18472886e-01 2.31542543e-01 -1.49324194e-01
1.49509859e+00 -1.44977641e+00 -4.81055915e-01 -2.79248625e-01
4.68964487e-01 -7.31027484e-01 1.28187025e+00 1.61803991e-01
-9.50480223e-01 -6.59048200e-01 -5.66977322e-01 -2.16333151e-01
-4.56855506e-01 9.86722410e-02 4.15167153e-01 3.51781212e-02
-1.29410386e+00 4.04259712e-01 -5.83400905e-01 -9.05774295e-01
3.51824105e-01 -1.27138440e-02 -6.28391862e-01 -2.77258784e-01
-1.15553081e+00 6.61722422e-01 3.42323542e-01 -5.55461407e-01
-1.32794988e+00 -7.44044960e-01 -8.94310951e-01 -3.86277176e-02
4.55434591e-01 -6.72182322e-01 1.33683300e+00 -6.52627110e-01
-9.88483787e-01 1.48068178e+00 -4.69907731e-01 -4.42996442e-01
2.28651557e-02 -2.57618159e-01 -3.36675286e-01 6.02420807e-01
4.62115079e-01 9.47356462e-01 1.32798982e+00 -1.18095076e+00
-8.21183026e-01 -8.51775929e-02 4.31301892e-01 1.38069347e-01
-7.94050157e-01 7.62267411e-02 -1.47718060e+00 -8.81349087e-01
-3.49962890e-01 -1.00729573e+00 -1.69902876e-01 2.04555199e-01
9.88042504e-02 -9.00832713e-01 8.89075696e-01 -4.41743076e-01
1.44219124e+00 -2.27370930e+00 4.21525359e-01 5.45894401e-03
1.81906298e-01 3.06098491e-01 -6.16767645e-01 6.69432044e-01
1.94066167e-01 1.02883287e-01 2.82216132e-01 -6.98035300e-01
2.53928930e-01 4.85950559e-01 -3.18110943e-01 4.12560701e-01
-6.16201498e-02 1.04739738e+00 -1.25070345e+00 -8.10123324e-01
1.47671208e-01 4.24976021e-01 -7.86484540e-01 4.23242480e-01
-3.87352735e-01 -3.33983521e-03 -3.01303118e-01 6.28739059e-01
3.60667035e-02 -5.21013200e-01 9.20940787e-02 -5.90971589e-01
9.38319713e-02 -3.07189506e-02 -8.92138958e-01 2.30721378e+00
-6.96967185e-01 9.13305104e-01 -1.85292155e-01 -1.14089525e+00
3.61199230e-01 3.30309361e-01 5.41319430e-01 -5.09787500e-01
-3.60660046e-01 5.91032840e-02 -7.27678597e-01 -1.00553250e+00
6.62369728e-01 4.82581735e-01 -3.06173652e-01 3.70867997e-01
7.09085941e-01 1.54621944e-01 2.68245220e-01 5.29195189e-01
1.23310304e+00 9.24196932e-03 2.17619881e-01 -1.73178256e-01
6.34056330e-01 -5.97571470e-02 3.14781040e-01 1.12912691e+00
-2.86678940e-01 4.76848811e-01 1.66615080e-02 -5.57190061e-01
-8.28013718e-01 -1.06155300e+00 5.17856590e-02 1.66847253e+00
2.05031171e-01 -1.04058886e+00 -2.80463219e-01 -6.84473455e-01
6.93327114e-02 3.89877051e-01 -6.70020163e-01 -2.34870419e-01
-4.53323692e-01 -1.22631975e-01 3.71223301e-01 3.14851820e-01
3.57864648e-01 -1.17238855e+00 -6.26152813e-01 2.19339635e-02
-4.51560348e-01 -1.30470181e+00 -1.10286725e+00 -1.50773972e-01
-4.02287275e-01 -9.99767780e-01 -7.15417087e-01 -1.10234606e+00
4.20953691e-01 5.28146148e-01 1.56083226e+00 6.37164861e-02
-8.40362787e-01 1.52165270e+00 -6.64420366e-01 1.05975911e-01
4.06683125e-02 1.17017124e-02 2.24306822e-01 9.65890661e-02
6.64895892e-01 -2.72824407e-01 -8.40551138e-01 7.53200054e-02
-1.07769918e+00 -5.29991686e-01 3.65747213e-01 9.35592949e-01
9.00834084e-01 -2.73929834e-01 -2.95894425e-02 -3.52232486e-01
7.24898398e-01 -5.78271449e-01 -1.10861272e-01 7.01022208e-01
-4.15066630e-01 1.01348989e-01 9.20393765e-02 -8.62007022e-01
-4.47718680e-01 -7.10917860e-02 1.69841081e-01 -1.20453918e+00
-1.60299525e-01 4.65682328e-01 9.81271192e-02 -8.48404467e-02
4.75861192e-01 4.78276104e-01 -3.12817782e-01 -5.43521106e-01
6.98309243e-01 5.72386086e-01 6.88705862e-01 -8.08414638e-01
6.50689423e-01 3.83228868e-01 -2.43569210e-01 -1.04492867e+00
-7.80112982e-01 -1.13575983e+00 -4.62393492e-01 -2.31376737e-01
1.07758546e+00 -9.96350408e-01 -6.82463706e-01 9.76128131e-02
-1.20628965e+00 -2.80512154e-01 -3.41880322e-01 3.74351174e-01
-8.49507451e-01 7.82122612e-01 -5.03364861e-01 -4.01756167e-01
-2.32150629e-01 -1.19986284e+00 1.27003849e+00 -3.76026928e-02
-3.00189376e-01 -1.16100395e+00 2.86089808e-01 3.04849982e-01
3.72197419e-01 -1.78656340e-01 8.30988944e-01 -7.95259953e-01
-6.68144286e-01 -2.55941749e-01 -1.33568672e-02 2.56980777e-01
1.20405868e-01 -2.98725188e-01 -7.52508938e-01 -6.41944230e-01
-3.02182734e-01 -6.24268830e-01 1.05955517e+00 8.75351485e-03
1.71390438e+00 -4.98314708e-01 -5.91438234e-01 7.50103951e-01
1.44671786e+00 -5.77072538e-02 3.17970067e-01 6.08335853e-01
5.04526556e-01 2.60691345e-01 6.88608289e-01 6.02389455e-01
4.48856086e-01 1.13489854e+00 3.81831288e-01 4.60695475e-01
-3.63037080e-01 -3.69357586e-01 4.76124376e-01 6.06952667e-01
6.44860938e-02 -3.65987241e-01 -7.16678560e-01 1.21683037e+00
-1.98123145e+00 -1.24873233e+00 5.98527849e-01 1.90120053e+00
8.59840751e-01 -5.04653335e-01 -1.50680408e-01 -5.37888229e-01
5.17524958e-01 6.05038345e-01 -5.04489303e-01 -2.63748586e-01
2.25205775e-02 3.03979069e-01 3.44719261e-01 4.69955981e-01
-1.37226355e+00 1.13893509e+00 5.68509102e+00 1.13226926e+00
-8.79857123e-01 5.39889812e-01 -2.31655855e-02 -3.18519920e-01
-2.28998363e-01 -1.49720266e-01 -9.04264808e-01 3.66276324e-01
8.01790416e-01 -2.04033196e-01 5.28794229e-01 9.04028237e-01
-1.63497865e-01 2.10202187e-01 -1.53079522e+00 1.62602842e+00
9.08754110e-01 -1.59632373e+00 6.87364995e-01 -5.65913916e-02
5.83392978e-01 7.40657523e-02 4.42536294e-01 6.37375057e-01
-1.40449792e-01 -8.19538355e-01 5.07473052e-01 6.13117576e-01
8.46742034e-01 -4.02531594e-01 2.38563433e-01 1.84080258e-01
-1.36147165e+00 -1.47187114e-01 -5.07045031e-01 6.55512154e-01
8.56525600e-02 5.04435822e-02 -8.27983379e-01 2.38355681e-01
1.11369646e+00 8.59026670e-01 -7.25223839e-01 1.07607698e+00
7.90503770e-02 3.13720167e-01 -1.11150891e-01 1.99155152e-01
6.09637558e-01 2.15040863e-01 7.90589929e-01 1.59078681e+00
3.30016017e-01 -1.06165178e-01 4.37852383e-01 4.11425859e-01
-3.02274376e-01 1.70115665e-01 -7.72373021e-01 -1.67685300e-01
4.74790722e-01 1.11790574e+00 -1.07676648e-01 -5.40585518e-01
-5.87837458e-01 1.62003732e+00 4.37187284e-01 3.55313301e-01
-6.78417921e-01 -2.63397008e-01 1.17024076e+00 4.40412154e-03
5.74962020e-01 -2.25639075e-01 7.66720235e-01 -1.40260577e+00
4.40196544e-02 -8.62892389e-01 7.00584888e-01 -7.71494031e-01
-1.58430362e+00 6.30025625e-01 2.15607166e-01 -1.21204722e+00
-5.61542273e-01 -6.53473079e-01 -1.79843560e-01 4.75799352e-01
-1.37471604e+00 -1.37424386e+00 -2.75486290e-01 9.66315210e-01
1.23108411e+00 -5.39763570e-01 9.57255006e-01 3.87699246e-01
4.01064893e-03 1.02670991e+00 1.26178309e-01 1.48994118e-01
1.15057874e+00 -9.20026302e-01 2.20187947e-01 5.18674076e-01
8.03239703e-01 8.76721740e-01 6.61544025e-01 -3.16040963e-01
-1.63809335e+00 -1.27033484e+00 1.22205937e+00 -8.60921502e-01
9.34341490e-01 -2.34424114e-01 -7.28170395e-01 9.06595588e-01
4.69114333e-01 3.04509878e-01 7.92647302e-01 8.29940289e-02
-7.47798026e-01 -4.25564758e-02 -8.62697899e-01 5.92673004e-01
1.35628629e+00 -1.05791795e+00 -9.27530766e-01 8.84998560e-01
8.81039321e-01 -2.04983011e-01 -1.05583739e+00 2.40771174e-01
5.46364903e-01 -6.43379331e-01 1.37255681e+00 -8.94049406e-01
2.92412549e-01 -1.38600945e-01 -6.68131113e-01 -1.23700869e+00
-5.67725360e-01 -5.54233134e-01 -4.65470433e-01 1.07976937e+00
6.86978102e-02 -1.94934294e-01 4.88941163e-01 3.33707958e-01
-7.63808191e-02 -7.60025799e-01 -1.01488376e+00 -9.73924875e-01
-2.20131248e-01 -5.37138939e-01 3.28675032e-01 8.24694514e-01
-1.08995721e-01 1.06491648e-01 -5.58698595e-01 2.29581490e-01
7.55725980e-01 1.71630159e-01 7.71467090e-01 -8.51058781e-01
-4.50509220e-01 -4.08370107e-01 -8.66101861e-01 -1.45069277e+00
6.42079115e-01 -1.13665819e+00 -9.56854783e-03 -1.47863972e+00
6.55745447e-01 -1.01829305e-01 -6.48254454e-01 4.28459495e-01
8.18893034e-03 5.24357855e-01 3.86184782e-01 4.66857970e-01
-1.39818001e+00 3.48205924e-01 7.07430959e-01 -6.66520774e-01
6.68920726e-02 -4.68810260e-01 -3.65476340e-01 3.69447619e-01
3.32117528e-01 -2.27070540e-01 -5.11057734e-01 -9.45590615e-01
8.92872885e-02 -1.75943837e-01 6.52886868e-01 -8.82940650e-01
4.42446023e-01 -1.46295160e-01 1.84548900e-01 -6.02001250e-01
6.68105483e-01 -7.35631645e-01 -8.20377842e-02 1.00819372e-01
-6.82978094e-01 9.26544294e-02 1.12203173e-01 8.53383064e-01
-3.90773714e-01 -2.79060394e-01 4.28738862e-01 -3.43750894e-01
-1.59431303e+00 8.02742958e-01 -3.50157231e-01 1.21313371e-01
1.01784098e+00 4.05269163e-03 -1.15518324e-01 -5.54711223e-01
-8.37056994e-01 5.22007406e-01 5.58045030e-01 9.87900138e-01
8.53218436e-01 -1.50865149e+00 -4.63341802e-01 -8.73762295e-02
9.00950909e-01 -5.44886291e-01 4.35503870e-01 5.08232951e-01
-8.93417150e-02 8.03988099e-01 1.53119713e-01 -7.87672460e-01
-1.46446824e+00 9.28344607e-01 5.36749177e-02 -1.23685807e-01
-6.71990156e-01 1.27942050e+00 3.24375123e-01 -4.09373678e-02
6.22070432e-01 -2.09667943e-02 -2.73438573e-01 2.11201757e-01
5.63836277e-01 -6.68342412e-03 -2.35887975e-01 -9.32487547e-01
-5.98126411e-01 8.28560948e-01 -1.41191632e-01 -4.99418192e-02
1.19871509e+00 -3.58697385e-01 -2.47378722e-01 6.65239811e-01
1.90887678e+00 -2.48280257e-01 -9.13146734e-01 -6.82003736e-01
3.59011302e-03 -7.56112695e-01 -1.02540918e-01 -4.16196615e-01
-9.86335158e-01 7.49557555e-01 7.20964432e-01 -2.23465413e-01
9.52524543e-01 7.54873693e-01 9.27686214e-01 8.49232078e-01
4.85027879e-01 -1.15544736e+00 5.09540200e-01 5.15462995e-01
9.75395501e-01 -1.29657102e+00 -1.29705250e-01 3.60077739e-01
-7.00131595e-01 9.84935045e-01 4.53672558e-01 -1.41261294e-01
7.15808511e-01 -6.50396168e-01 -1.29012307e-02 -2.76997000e-01
-1.12969637e+00 -1.88305452e-01 7.36556351e-01 6.78718686e-01
1.50769666e-01 -1.42792448e-01 -2.07073733e-01 2.70127416e-01
2.16746435e-01 -1.96407512e-01 3.57284285e-02 8.87962401e-01
-4.53286082e-01 -1.04239178e+00 -7.05391243e-02 4.18718308e-01
-4.25615251e-01 -2.44917572e-01 -8.50422606e-02 5.33885241e-01
1.73778266e-01 8.05005848e-01 2.77551860e-01 -4.50645059e-01
2.05696672e-01 3.35122526e-01 6.06896698e-01 -8.92606974e-01
-2.37305388e-01 -1.87144130e-01 -2.08715782e-01 -1.06903172e+00
-5.61433077e-01 -6.65743768e-01 -6.41111493e-01 1.13347895e-01
3.18328679e-01 3.80330622e-01 5.14148414e-01 6.75676644e-01
4.55384344e-01 1.32476017e-02 4.26934838e-01 -1.09588337e+00
-5.27993977e-01 -7.09185123e-01 -3.92289102e-01 7.19602942e-01
5.18930018e-01 -7.49061823e-01 -4.98126060e-01 4.42672253e-01] | [10.307475090026855, 0.9361639618873596] |
ec5d061d-905c-421d-8628-9686c840fa3c | unsupervised-learning-of-syntactic-structure | 1808.09111 | null | http://arxiv.org/abs/1808.09111v1 | http://arxiv.org/pdf/1808.09111v1.pdf | Unsupervised Learning of Syntactic Structure with Invertible Neural Projections | Unsupervised learning of syntactic structure is typically performed using
generative models with discrete latent variables and multinomial parameters. In
most cases, these models have not leveraged continuous word representations. In
this work, we propose a novel generative model that jointly learns discrete
syntactic structure and continuous word representations in an unsupervised
fashion by cascading an invertible neural network with a structured generative
prior. We show that the invertibility condition allows for efficient exact
inference and marginal likelihood computation in our model so long as the prior
is well-behaved. In experiments we instantiate our approach with both Markov
and tree-structured priors, evaluating on two tasks: part-of-speech (POS)
induction, and unsupervised dependency parsing without gold POS annotation. On
the Penn Treebank, our Markov-structured model surpasses state-of-the-art
results on POS induction. Similarly, we find that our tree-structured model
achieves state-of-the-art performance on unsupervised dependency parsing for
the difficult training condition where neither gold POS annotation nor
punctuation-based constraints are available. | ['Taylor Berg-Kirkpatrick', 'Junxian He', 'Graham Neubig'] | 2018-08-28 | unsupervised-learning-of-syntactic-structure-1 | https://aclanthology.org/D18-1160 | https://aclanthology.org/D18-1160.pdf | emnlp-2018-10 | ['constituency-grammar-induction', 'unsupervised-dependency-parsing'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.45061144e-01 6.75755739e-01 -3.27060789e-01 -7.60932028e-01
-1.18706346e+00 -7.95231879e-01 6.11234605e-01 4.14659716e-02
-4.42905456e-01 8.18289995e-01 3.62295479e-01 -7.38323390e-01
1.22654818e-01 -7.40558147e-01 -7.98161209e-01 -7.22049057e-01
-5.44319972e-02 7.94656456e-01 -7.35867694e-02 2.20899992e-02
-2.16257244e-01 1.08081199e-01 -1.09040785e+00 1.12105645e-01
7.22840071e-01 3.56958687e-01 3.10426891e-01 7.87945569e-01
-2.17867121e-01 7.62325704e-01 -3.91047835e-01 -5.10361791e-01
-2.83890329e-02 -3.74415547e-01 -8.46402287e-01 -5.98319173e-02
1.99558079e-01 -8.69624987e-02 -8.51650257e-03 1.03669262e+00
1.19431548e-01 -1.11944288e-01 7.63782144e-01 -7.65837014e-01
-8.00256968e-01 1.12193549e+00 -2.20070928e-01 1.89649388e-01
1.67269744e-02 -1.45617008e-01 1.71685183e+00 -7.23128200e-01
5.53401470e-01 1.33996987e+00 6.21746480e-01 2.98631907e-01
-1.67317927e+00 -3.30036551e-01 4.30214971e-01 -3.64318520e-01
-9.28751349e-01 -2.46806830e-01 6.72433734e-01 -3.59938711e-01
1.40323079e+00 -3.47798437e-01 4.97173555e-02 1.16953671e+00
1.90676942e-01 6.97883308e-01 9.18401897e-01 -8.77170265e-01
2.89674938e-01 -1.14884898e-01 5.24135768e-01 7.59619713e-01
4.05727953e-01 -2.19861105e-01 -3.50251228e-01 -1.82424724e-01
7.48992324e-01 -5.55339195e-02 2.19566852e-01 -1.95972413e-01
-8.16020787e-01 1.24270916e+00 -1.25593767e-01 5.24797142e-01
-3.87883902e-01 3.69749904e-01 1.49163529e-01 1.25014633e-01
5.88466763e-01 1.63570896e-01 -7.92244136e-01 -1.92283705e-01
-9.27106678e-01 -1.88654721e-01 1.27846992e+00 1.00132656e+00
8.87487471e-01 5.44857197e-02 9.81098711e-02 8.46129000e-01
5.84874690e-01 6.25191033e-01 4.11573112e-01 -8.67271841e-01
6.29160225e-01 6.72864392e-02 7.86526781e-03 -2.63462782e-01
-3.02928597e-01 -2.06105098e-01 -4.69439864e-01 -7.75820240e-02
5.74130714e-01 -6.85091019e-01 -1.19752789e+00 2.21939921e+00
7.53303245e-02 -6.21561445e-02 3.49987566e-01 2.84902275e-01
5.07434070e-01 7.70373762e-01 4.97243822e-01 -3.41536611e-01
1.45348358e+00 -6.30587637e-01 -7.30486095e-01 -4.65708226e-01
7.01229095e-01 -6.34064794e-01 8.14683497e-01 2.37248421e-01
-1.07493699e+00 -4.29878563e-01 -7.54131615e-01 -1.54105857e-01
-1.29259765e-01 1.36169389e-01 9.79517877e-01 9.02498007e-01
-1.07309973e+00 5.25505304e-01 -1.45656931e+00 -2.31756821e-01
1.01966649e-01 4.12243009e-01 -4.39404458e-01 1.80789586e-02
-1.11094773e+00 7.23061502e-01 4.90142524e-01 -7.00358823e-02
-8.83933246e-01 -3.42590958e-01 -1.20831752e+00 2.07876727e-01
1.60080388e-01 -6.73949182e-01 1.60118091e+00 -7.79668570e-01
-1.84061348e+00 8.10951829e-01 -3.39655191e-01 -5.92787564e-01
-7.55923688e-02 -4.74561095e-01 4.84062657e-02 8.93455595e-02
1.89934179e-01 5.93166769e-01 7.92306602e-01 -1.09164107e+00
-4.21367586e-01 -2.40028113e-01 1.31658152e-01 -2.96119191e-02
-1.94876760e-01 2.45866850e-01 -2.31623322e-01 -5.28792322e-01
2.45591223e-01 -8.53378773e-01 -3.88927609e-01 -8.37361097e-01
-4.72427666e-01 -5.18635154e-01 3.23984325e-01 -7.63913989e-01
1.06743824e+00 -2.02441335e+00 1.65640175e-01 7.50647560e-02
-2.63910770e-01 -2.00895853e-02 -1.38900399e-01 5.69475770e-01
-1.96481779e-01 2.41710171e-01 -4.52175796e-01 -8.52446496e-01
3.83667648e-01 9.71410155e-01 -4.96208578e-01 2.07677469e-01
6.04878962e-01 9.28243101e-01 -8.66180837e-01 -4.38288480e-01
-1.73075810e-01 5.80291808e-01 -8.32289696e-01 3.39447170e-01
-4.68798786e-01 2.09125206e-01 -3.55004758e-01 4.88821298e-01
3.41105819e-01 -3.60808969e-01 9.32591677e-01 3.11256438e-01
-5.70380390e-02 8.82076740e-01 -1.05884778e+00 1.66018832e+00
-7.41319001e-01 2.47847691e-01 1.65341899e-01 -1.08055699e+00
7.74577558e-01 4.57660526e-01 8.46784636e-02 -6.68476596e-02
-8.48566648e-03 6.35023490e-02 3.97784896e-02 -3.42743784e-01
2.13889256e-01 -6.39814317e-01 -4.59578961e-01 5.43357313e-01
7.74010837e-01 -1.14500802e-02 3.56189817e-01 1.77959204e-01
1.28021705e+00 4.72754300e-01 3.37681621e-01 -2.21985295e-01
3.11349854e-02 -2.61833072e-01 6.54365599e-01 9.37016785e-01
3.31997901e-01 4.59676951e-01 9.06403482e-01 -7.21086469e-03
-8.35220337e-01 -1.42407453e+00 -1.20579682e-01 1.55069172e+00
-5.84244311e-01 -4.38000292e-01 -7.76536345e-01 -8.79063308e-01
-2.71144003e-01 9.49449897e-01 -5.79735935e-01 3.87174815e-01
-9.31020021e-01 -1.04048502e+00 5.97176075e-01 8.92803013e-01
-5.88991567e-02 -1.26134419e+00 -2.32715577e-01 4.31936085e-01
2.62046363e-02 -1.32456958e+00 -4.23679724e-02 9.99871254e-01
-9.91749287e-01 -7.43182838e-01 -4.09080893e-01 -1.10339832e+00
8.27588677e-01 -4.13268179e-01 1.35156274e+00 -3.50418448e-01
1.65243611e-01 2.40682065e-01 -4.21833813e-01 -2.20459267e-01
-6.69080198e-01 4.00036991e-01 -2.37785771e-01 -2.59741485e-01
3.86008710e-01 -8.49532843e-01 -2.42371354e-02 -1.69950500e-01
-9.82622504e-01 -2.34792009e-01 1.02376640e+00 1.07148147e+00
5.98924518e-01 -2.23795712e-01 4.48096037e-01 -1.35953963e+00
3.36240798e-01 -5.78152001e-01 -7.25720763e-01 1.31106496e-01
-3.32871795e-01 5.54888904e-01 6.26927972e-01 -3.09611887e-01
-1.39373446e+00 3.79985839e-01 -5.42807460e-01 -7.98691511e-02
-6.33059084e-01 7.00342417e-01 -4.08148259e-01 5.80941975e-01
3.60621810e-01 1.66798718e-02 -3.62983882e-01 -8.13935459e-01
6.90265954e-01 4.14935976e-01 5.67037940e-01 -8.27777863e-01
8.37814927e-01 2.99897134e-01 -6.29248843e-02 -6.62032425e-01
-1.25257158e+00 -4.15734828e-01 -1.08163774e+00 6.57478631e-01
1.29254270e+00 -9.37578917e-01 -1.43701732e-01 1.05626121e-01
-1.43861103e+00 -5.80457926e-01 -3.01404953e-01 6.37502730e-01
-7.01609254e-01 5.39842010e-01 -1.00125122e+00 -9.35757399e-01
-1.19114019e-01 -8.61257672e-01 1.17343903e+00 -5.18233068e-02
-1.72011793e-01 -1.39762700e+00 4.77419823e-01 9.88908038e-02
-4.62264530e-02 -8.46549273e-02 1.20326650e+00 -9.92723227e-01
-5.40754855e-01 -4.86611277e-02 7.53511265e-02 5.53583443e-01
1.70675263e-01 -1.21554039e-01 -9.66742337e-01 -1.51701588e-02
-6.84299180e-03 -3.13029617e-01 1.03952932e+00 4.98324186e-01
6.04617417e-01 -3.42954725e-01 -6.81480691e-02 5.11616111e-01
1.24876690e+00 -9.85215008e-02 4.72642303e-01 -5.03237247e-02
5.57639778e-01 4.85866070e-01 3.16077709e-01 2.28483558e-01
4.45945710e-01 1.33752391e-01 2.86665648e-01 1.32512182e-01
2.32037932e-01 -4.48817998e-01 7.10659087e-01 1.03201365e+00
-1.43018842e-01 -4.38999265e-01 -1.01370215e+00 7.13243365e-01
-1.77333629e+00 -8.62636566e-01 -2.46660206e-02 1.85778630e+00
1.21311915e+00 4.86120075e-01 -1.80023760e-01 -1.59570470e-01
6.57812476e-01 1.50031582e-01 -4.80603939e-03 -5.29265225e-01
7.81529844e-02 8.96623075e-01 5.34372687e-01 8.98570120e-01
-1.39278519e+00 1.28582740e+00 6.97343445e+00 4.62058395e-01
-6.42936826e-01 3.76896620e-01 4.85596925e-01 4.22740310e-01
-5.43054760e-01 3.71523142e-01 -1.34223521e+00 1.61618114e-01
1.31749451e+00 4.99829262e-01 4.96920832e-02 1.05011415e+00
-1.35839418e-01 -7.08971396e-02 -1.26126349e+00 5.02639472e-01
-9.82629042e-03 -1.00073028e+00 -3.11299443e-01 1.60328537e-01
7.17781842e-01 3.28420281e-01 -7.62015507e-02 5.22445560e-01
1.18293107e+00 -1.05652261e+00 4.13648754e-01 1.89011633e-01
5.44966042e-01 -4.77387339e-01 7.68519819e-01 5.46283305e-01
-8.07450712e-01 2.19769150e-01 -3.15826297e-01 -1.84308365e-01
5.80520928e-01 6.66576684e-01 -8.88632834e-01 3.51205915e-01
2.84358501e-01 3.84835124e-01 -1.80120543e-01 4.03946936e-01
-1.09906101e+00 1.42894471e+00 -5.91279209e-01 5.01837203e-05
4.22445804e-01 -1.59185976e-01 2.59434044e-01 1.72120094e+00
3.36078182e-02 8.39718357e-02 3.83767724e-01 7.36199558e-01
-1.68875724e-01 -1.71009582e-02 -3.81245047e-01 -1.81782603e-01
2.49767646e-01 1.25135231e+00 -9.29868460e-01 -3.76074195e-01
-6.54579639e-01 7.19730914e-01 6.83591664e-01 4.80786622e-01
-7.02494860e-01 -2.33585298e-01 4.08672392e-01 -1.47658229e-01
1.10281968e+00 -7.90374339e-01 -3.12682867e-01 -1.23199451e+00
-1.78155437e-01 -4.20168400e-01 4.83263195e-01 -2.98013300e-01
-1.49579465e+00 3.69904965e-01 3.19394588e-01 -4.19242650e-01
-9.96321976e-01 -9.19798791e-01 -7.02720940e-01 1.00000966e+00
-1.69034445e+00 -1.25417495e+00 3.98400873e-01 1.32511914e-01
4.47403759e-01 6.27953932e-02 1.19896495e+00 -8.06364715e-02
-5.33384204e-01 4.30339575e-01 3.23883891e-02 6.24339044e-01
5.17406881e-01 -1.63613904e+00 6.02546215e-01 1.03105462e+00
7.14195669e-01 9.66991246e-01 6.49483800e-01 -7.07164049e-01
-1.27689624e+00 -8.83220196e-01 1.27051771e+00 -6.41610146e-01
1.00990927e+00 -6.55260801e-01 -8.70481014e-01 1.34216797e+00
4.18937117e-01 4.48270105e-02 1.03310597e+00 7.03064263e-01
-5.39219379e-01 3.43449026e-01 -6.56614184e-01 1.49098858e-01
8.38393509e-01 -4.65029031e-01 -1.15522444e+00 4.08710420e-01
8.21265697e-01 -1.73722774e-01 -9.01153445e-01 1.85893133e-01
3.31007957e-01 -6.53351426e-01 6.67826116e-01 -8.84584665e-01
4.43012774e-01 9.01272148e-02 -4.01339591e-01 -1.11482906e+00
-3.77417833e-01 -8.62858057e-01 -1.27527818e-01 1.42434299e+00
8.65701973e-01 -6.32349968e-01 7.16294944e-01 5.57189345e-01
-3.21535736e-01 -4.79695767e-01 -8.65131974e-01 -7.75213301e-01
4.57838625e-01 -5.31627953e-01 6.52588904e-02 6.61019564e-01
-6.07067440e-03 7.24092484e-01 -8.44446495e-02 4.36948687e-01
4.92110133e-01 9.28751156e-02 5.34257352e-01 -1.15613317e+00
-9.64860439e-01 4.80543561e-02 -2.10305303e-01 -1.31859589e+00
7.18409657e-01 -8.17888856e-01 4.61276442e-01 -1.52182293e+00
2.00875089e-01 -3.24107021e-01 -2.97212690e-01 8.73687804e-01
-2.28292421e-01 -1.33975996e-02 -2.53380425e-02 2.65929326e-02
-6.37301028e-01 2.85815597e-01 6.61070883e-01 2.79288530e-01
-8.36089402e-02 -5.55941723e-02 -7.37775207e-01 1.05148029e+00
7.96416283e-01 -8.25989187e-01 -2.39618748e-01 -5.36727607e-01
3.70181233e-01 1.05683908e-01 1.65133044e-01 -4.57596481e-01
5.72346784e-02 -9.46560279e-02 1.58636764e-01 -4.41724300e-01
3.14919144e-01 -2.97386557e-01 -3.27211678e-01 8.55709985e-02
-2.96769738e-01 2.19601812e-03 -6.62443861e-02 5.92790365e-01
-2.97744066e-01 -6.80817425e-01 5.26218772e-01 -3.07000548e-01
-4.50426012e-01 -6.68325415e-03 -4.57415253e-01 4.30698276e-01
4.78088677e-01 2.60557503e-01 -3.03599928e-02 -2.69915551e-01
-1.11734164e+00 -2.07858339e-01 1.14103027e-01 7.98189938e-02
1.74978584e-01 -8.56269538e-01 -7.27419019e-01 2.92841494e-01
-3.79176646e-01 1.58961713e-01 -2.00260237e-01 6.30086362e-01
-1.43785715e-01 5.23751616e-01 2.24908635e-01 -5.75189352e-01
-9.74248230e-01 3.37612867e-01 -2.05299601e-01 -8.49511266e-01
-3.96393627e-01 9.87230182e-01 2.83430040e-01 -4.76220727e-01
1.11199670e-01 -7.94811666e-01 -8.11141506e-02 9.39057320e-02
9.30762663e-02 -2.72474438e-01 -2.85292584e-02 -5.25557697e-01
-1.53598085e-01 2.06008285e-01 -2.00716570e-01 -4.07645106e-01
1.58813083e+00 7.37720430e-02 -1.01853654e-01 5.87659419e-01
9.73336220e-01 2.48637736e-01 -1.41341829e+00 -1.60275087e-01
2.13974506e-01 1.82742506e-01 -1.95416823e-01 -5.85289896e-01
-5.19780219e-01 1.11827540e+00 -1.02712847e-01 2.72233218e-01
6.19912684e-01 4.68690842e-01 6.47672653e-01 7.03879714e-01
3.24875563e-01 -9.75242436e-01 1.23304799e-02 1.03180993e+00
2.72681803e-01 -1.13723886e+00 -2.42107078e-01 -6.19243979e-01
-5.61694682e-01 9.81665075e-01 1.27758250e-01 -4.11718518e-01
8.18356752e-01 7.28211045e-01 1.47456601e-02 -6.70893397e-03
-8.72206569e-01 -4.52678025e-01 1.19927354e-01 5.49131989e-01
7.74262071e-01 1.71094596e-01 -1.14555731e-01 1.01013863e+00
-4.15034801e-01 -3.51329416e-01 2.57716298e-01 1.10711014e+00
-6.04336858e-01 -1.55386937e+00 -1.62938207e-01 9.42004323e-02
-9.42432404e-01 -6.41090214e-01 -1.13578022e-01 8.29595864e-01
-1.68721572e-01 8.39985549e-01 1.76801130e-01 2.99418807e-01
-5.20234481e-02 6.92338824e-01 6.37766182e-01 -1.35258937e+00
-3.22426468e-01 4.18966860e-01 3.48311394e-01 -2.33117238e-01
-4.39788759e-01 -9.49600279e-01 -1.46796262e+00 4.22899067e-01
-3.49974543e-01 3.38932186e-01 7.30688453e-01 1.19974840e+00
1.01167560e-01 3.29116017e-01 2.90588766e-01 -7.46001184e-01
-9.66965616e-01 -1.22804141e+00 -4.58986461e-01 2.40041912e-01
7.55187497e-02 -4.20873642e-01 -3.23935151e-01 4.81634796e-01] | [10.387995719909668, 9.623323440551758] |
c610b793-388a-4d25-afd8-ba796cddaff7 | efficient-directed-graph-sampling-via | 2210.14263 | null | https://arxiv.org/abs/2210.14263v1 | https://arxiv.org/pdf/2210.14263v1.pdf | Efficient Directed Graph Sampling via Gershgorin Disc Alignment | Graph sampling is the problem of choosing a node subset via sampling matrix $\mathbf{H} \in \{0,1\}^{K \times N}$ to collect samples $\mathbf{y} = \mathbf{H} \mathbf{x} \in \mathbb{R}^K$, $K < N$, so that the target signal $\mathbf{x} \in \mathbb{R}^N$ can be reconstructed in high fidelity. While sampling on undirected graphs is well studied, we propose the first sampling scheme tailored specifically for directed graphs, leveraging a previous undirected graph sampling method based on Gershgorin disc alignment (GDAS). Concretely, given a directed positive graph $\mathcal{G}^d$ specified by random-walk graph Laplacian matrix $\mathbf{L}_{rw}$, we first define reconstruction of a smooth signal $\mathbf{x}^*$ from samples $\mathbf{y}$ using graph shift variation (GSV) $\|\mathbf{L}_{rw} \mathbf{x}\|^2_2$ as a signal prior. To minimize worst-case reconstruction error of the linear system solution $\mathbf{x}^* = \mathbf{C}^{-1} \mathbf{H}^\top \mathbf{y}$ with symmetric coefficient matrix $\mathbf{C} = \mathbf{H}^\top \mathbf{H} + \mu \mathbf{L}_{rw}^\top \mathbf{L}_{rw}$, the sampling objective is to choose $\mathbf{H}$ to maximize the smallest eigenvalue $\lambda_{\min}(\mathbf{C})$ of $\mathbf{C}$. To circumvent eigen-decomposition entirely, we maximize instead a lower bound $\lambda^-_{\min}(\mathbf{S}\mathbf{C}\mathbf{S}^{-1})$ of $\lambda_{\min}(\mathbf{C})$ -- smallest Gershgorin disc left-end of a similarity transform of $\mathbf{C}$ -- via a variant of GDAS based on Gershgorin circle theorem (GCT). Experimental results show that our sampling method yields smaller signal reconstruction errors at a faster speed compared to competing schemes. | ['Gene Cheung', 'Hong Vicky Zhao', 'Yuejiang Li'] | 2022-10-25 | null | null | null | null | ['graph-sampling'] | ['graphs'] | [ 4.52872097e-01 8.83268774e-01 1.15828775e-01 -4.78103533e-02
-8.88256848e-01 -6.19445324e-01 -3.32124472e-01 2.46691927e-02
-1.08274847e-01 8.20580840e-01 -5.21253943e-01 -7.51481652e-01
-9.86388624e-01 -1.28947473e+00 -8.46846581e-01 -9.37929630e-01
-9.76923823e-01 1.14766553e-01 -5.28780557e-03 -4.13888246e-01
-1.08441211e-01 4.34993982e-01 -1.44720006e+00 -2.01209664e-01
9.20001507e-01 1.41979456e+00 -1.11384124e-01 5.65917194e-01
1.98288307e-01 5.29731631e-01 -5.89173019e-01 -4.65278745e-01
6.15137041e-01 -1.15135348e+00 -6.81337893e-01 -2.34056748e-02
3.34736645e-01 9.90858823e-02 -5.98884642e-01 1.72751236e+00
2.54945904e-01 3.40104997e-01 1.07177305e+00 -8.47952127e-01
-4.74209905e-01 6.08523488e-01 -1.09679711e+00 4.52042073e-02
3.16250414e-01 1.78902879e-01 1.01944995e+00 -4.61550027e-01
6.36262059e-01 7.80693293e-01 8.15274358e-01 2.56203920e-01
-1.63343525e+00 -1.12926781e+00 1.19989060e-01 -5.28846860e-01
-1.91046369e+00 4.47123013e-02 6.68069780e-01 -4.40502226e-01
3.72958660e-01 7.82570004e-01 6.75275922e-01 1.66987166e-01
-2.66853534e-02 8.85611996e-02 1.10200000e+00 -6.66474760e-01
1.48523375e-01 -3.27886581e-01 3.36034358e-01 1.33357882e+00
2.86713719e-01 -7.59479851e-02 -8.75934064e-02 -3.65158655e-02
1.35180259e+00 -7.33828545e-02 -7.08699942e-01 3.01354825e-01
-8.93404007e-01 1.08126378e+00 4.28182632e-01 2.45544896e-01
-1.03475943e-01 4.99557972e-01 -2.62566268e-01 3.66619319e-01
1.82003573e-01 3.12259793e-01 -1.34882286e-01 3.79323214e-01
-1.17760706e+00 2.54940540e-01 6.27084911e-01 1.56575108e+00
1.25559580e+00 2.49917865e-01 -1.94396794e-01 7.46155679e-01
4.07897681e-02 9.21938300e-01 -5.84414005e-01 -1.45819604e+00
4.50091004e-01 6.63668036e-01 -1.41504765e-01 -1.01281548e+00
-4.39335763e-01 -4.09925044e-01 -1.26209271e+00 2.64618039e-01
7.74634361e-01 -1.93069875e-01 -5.09172797e-01 2.05923676e+00
8.29228759e-02 -2.47286409e-01 -5.34939826e-01 7.92922437e-01
4.87916946e-01 7.94239044e-01 -3.34934741e-01 -9.36793029e-01
1.13161922e+00 -1.66632861e-01 7.44816065e-02 1.47021800e-01
4.12868261e-01 -7.49642849e-01 1.28177845e+00 4.10449862e-01
-1.55762029e+00 -3.53436530e-01 -1.02708709e+00 4.59444731e-01
2.62062013e-01 -1.88419998e-01 3.33349437e-01 1.04117107e+00
-1.05376077e+00 7.14701176e-01 -4.84183401e-01 2.07879409e-01
1.85805261e-01 5.65759063e-01 -2.06283912e-01 -1.71027660e-01
-9.10155535e-01 -1.20933190e-01 -2.72821248e-01 2.33135894e-02
-7.79022694e-01 -8.79769802e-01 -6.15310371e-01 -2.42556185e-02
5.39131165e-01 -5.64954579e-02 4.28524196e-01 -4.89304543e-01
-9.26471353e-01 8.35558176e-01 1.41739324e-01 6.68878108e-03
3.05015922e-01 7.82119393e-01 -4.22540784e-01 3.33269477e-01
3.36628675e-01 2.19267562e-01 7.50663280e-01 -1.43603086e+00
-2.69953668e-01 -8.51542354e-01 6.23293296e-02 -3.87897313e-01
-1.07331030e-01 -6.73009530e-02 -2.92411774e-01 -8.42309177e-01
8.34961295e-01 -9.68455613e-01 -1.93933949e-01 -4.18816656e-01
-4.94149387e-01 3.23180735e-01 3.76506507e-01 -6.12008393e-01
1.74666500e+00 -2.02305079e+00 1.77516907e-01 1.38528872e+00
6.97449803e-01 -1.36005431e-01 5.46650052e-01 4.75829929e-01
-1.49603292e-01 2.07873628e-01 -5.35021186e-01 3.04069400e-01
-1.38378948e-01 -6.31044284e-02 5.91563284e-02 8.21793199e-01
-5.43307424e-01 1.19188718e-01 -6.96246028e-01 -4.12547924e-02
-9.91953760e-02 2.91204631e-01 -5.69281697e-01 -1.88257754e-01
-3.85930598e-01 5.74835658e-01 -7.07427263e-01 4.74376023e-01
1.02185321e+00 -5.22072792e-01 4.17356968e-01 -4.04313840e-02
-1.88522279e-01 -3.72971445e-01 -1.83570242e+00 1.34665358e+00
2.58500159e-01 -1.39463082e-01 1.06985044e+00 -1.62447858e+00
1.07868838e+00 -9.79724973e-02 1.10201240e+00 -7.96511889e-01
1.81920752e-01 2.22804248e-01 -2.20911741e-01 6.50797933e-02
-1.07166551e-01 -4.14270520e-01 -1.72463343e-01 7.03141928e-01
-3.84185202e-02 -6.23851538e-01 2.19730541e-01 4.43215370e-01
1.73467708e+00 -1.39302418e-01 -1.80563688e-01 -8.53822947e-01
6.69523001e-01 -1.05837651e-01 3.64337564e-01 8.65008950e-01
1.25441328e-03 4.37745899e-01 9.57283497e-01 6.01448268e-02
-9.05102491e-01 -1.16806078e+00 -3.95413011e-01 1.14363265e+00
4.40430075e-01 -5.02718687e-01 -1.34056807e+00 -3.65012407e-01
-2.03412339e-01 7.26273477e-01 -5.17668366e-01 -6.06504753e-02
-6.16593361e-01 -8.96090508e-01 7.79862821e-01 2.31802255e-01
5.49960136e-01 -8.88604522e-01 -3.23081851e-01 1.99075993e-02
-3.99982445e-02 -5.30612290e-01 -5.89215279e-01 3.83662462e-01
-5.70690095e-01 -1.11223495e+00 -7.47258484e-01 -3.17150056e-01
1.17060030e+00 6.27437681e-02 8.57656002e-01 3.73180479e-01
-6.93734527e-01 5.41109920e-01 -2.22893491e-01 -2.35589981e-01
-6.74025044e-02 -2.85367489e-01 -5.16801327e-02 -1.23344250e-02
-3.20403278e-01 -1.01361156e+00 -9.71128702e-01 5.07904649e-01
-9.24182296e-01 -2.24639416e-01 -1.56995490e-01 8.35101128e-01
1.01061082e+00 3.39295030e-01 2.77935207e-01 -5.23832560e-01
2.96156973e-01 -5.51117118e-04 -1.21777713e+00 -1.68136522e-01
-6.99250698e-01 -1.78327858e-01 8.41095686e-01 -7.08682388e-02
-2.09578067e-01 -4.20225978e-01 -6.02850989e-02 -5.80598652e-01
2.95331061e-01 4.61242497e-01 -2.27372292e-02 -1.82143390e-01
9.77039695e-01 2.80223697e-01 -3.61630991e-02 -3.47598732e-01
3.64659846e-01 9.85576659e-02 5.71869969e-01 -9.73644614e-01
6.02445483e-01 5.07888019e-01 7.41101086e-01 -9.50992942e-01
-5.60813546e-01 9.04513896e-02 -6.05914183e-02 -4.18908775e-01
8.01491678e-01 -5.48749149e-01 -1.65927887e+00 -3.36190343e-01
-2.34427646e-01 -3.88599813e-01 -8.02198052e-01 4.68048424e-01
-7.74424136e-01 4.10426229e-01 -5.42654157e-01 -1.11621702e+00
-2.84953296e-01 -1.34864044e+00 8.26927423e-01 6.62171692e-02
-2.73618221e-01 -8.21841955e-01 -5.97715616e-01 5.33381939e-01
2.68144727e-01 4.70119148e-01 1.39310265e+00 -1.86037481e-01
-8.28520298e-01 -4.73742664e-01 -5.42000830e-01 7.77744949e-01
-3.53536040e-01 -2.56845891e-01 -4.47567523e-01 -5.94445586e-01
6.46100119e-02 1.54748023e-01 5.15505791e-01 5.12952745e-01
1.40546298e+00 -6.86375797e-01 -4.07842964e-01 5.53424954e-01
1.57373512e+00 4.77850810e-02 8.59904766e-01 -3.09725195e-01
2.23990098e-01 1.70240611e-01 4.99913581e-02 6.08611822e-01
5.15740290e-02 4.76822972e-01 4.92866069e-01 2.55488932e-01
-1.50474563e-01 1.11473411e-01 2.59602606e-01 3.76715809e-01
-4.18037742e-01 -1.14502795e-02 -7.76953101e-01 4.08097744e-01
-1.30415559e+00 -1.01054382e+00 -3.31409901e-01 2.77623415e+00
6.95697963e-01 1.88771784e-01 2.51519144e-01 4.08850491e-01
9.18098748e-01 1.25218332e-01 -7.24548623e-02 6.84326887e-03
-1.46531295e-02 1.46112239e+00 7.65415013e-01 6.67878509e-01
-4.25028175e-01 4.93577600e-01 2.41789412e+00 1.40334332e+00
-9.60287631e-01 -1.22171894e-01 5.13937950e-01 -3.09300900e-01
-6.22893929e-01 4.40420210e-01 -7.43562937e-01 8.00720453e-01
5.46043515e-01 1.73378214e-01 7.79928923e-01 6.12551689e-01
1.40566871e-01 -3.93739760e-01 -9.47151959e-01 7.74504781e-01
-4.43071574e-02 -1.50305986e+00 -5.79006314e-01 3.23778450e-01
8.25831532e-01 -6.80538476e-01 1.09147772e-01 1.21180505e-01
7.78032064e-01 -1.39943600e+00 7.28841662e-01 1.91593692e-01
1.54749382e+00 -5.87998211e-01 -3.29821438e-01 3.65449786e-01
-1.55275702e+00 -1.00232311e-01 -1.04440048e-01 1.05563179e-01
1.08929155e-02 1.06440902e+00 -2.29512453e-01 9.10331607e-01
1.02079654e+00 -3.01045552e-02 2.42733747e-01 3.64614576e-01
9.34462696e-02 7.16300786e-01 -4.88090843e-01 2.10431203e-01
1.54414743e-01 -6.93312407e-01 7.46726811e-01 7.86292672e-01
6.65213585e-01 1.02516747e+00 7.16472745e-01 1.19381177e+00
-2.87211597e-01 2.98075438e-01 -3.21009219e-01 1.53057814e-01
3.40771139e-01 8.23730528e-01 -1.02562642e+00 1.39389545e-01
-1.33509189e-01 3.10498714e-01 7.41789043e-02 6.09128475e-01
-8.88729215e-01 -1.04865932e+00 5.11555672e-01 1.04197657e+00
2.07023650e-01 -3.34187388e-01 -4.75801468e-01 -6.37329340e-01
1.87230229e-01 -6.96145296e-01 5.11264265e-01 -5.15091300e-01
-1.01153088e+00 2.75330454e-01 1.14842646e-01 -1.13223982e+00
5.45000210e-02 -3.68301183e-01 4.52176295e-02 1.26345217e+00
-4.53223944e-01 -7.13024616e-01 -3.15876365e-01 1.09718573e+00
-3.90478253e-01 -3.02628160e-01 7.93220222e-01 4.28952485e-01
-1.01519294e-01 8.66803944e-01 3.26869547e-01 1.54290259e-01
8.05231184e-02 -5.15434086e-01 -4.75440979e-01 6.24544024e-01
-4.58643675e-01 5.75442672e-01 6.30173028e-01 -6.41828477e-01
-1.67810905e+00 -8.04380059e-01 1.57928243e-01 1.77541524e-01
3.91123116e-01 -3.74324143e-01 -6.83287024e-01 9.82588291e-01
-1.71775609e-01 1.48208976e-01 3.97090852e-01 -4.57291454e-01
-2.88183212e-01 -2.81524360e-01 -1.78309178e+00 6.83279037e-01
1.18196929e+00 -2.74966896e-01 4.80304003e-01 3.87636065e-01
3.89167964e-01 -5.66902637e-01 -1.38498807e+00 7.48957038e-01
1.82817698e-01 -1.14098310e+00 1.02099979e+00 -1.32960171e-01
1.33276016e-01 -3.40195358e-01 -7.54629612e-01 -6.24803305e-01
-2.66802907e-01 -1.23910224e+00 3.43844473e-01 8.90579939e-01
5.14553189e-01 -6.52521133e-01 1.08991659e+00 5.62363565e-01
-3.74648541e-01 -5.70374012e-01 -1.32782459e+00 -5.74358404e-01
4.55195874e-01 -3.20141703e-01 1.73087940e-01 9.65374529e-01
1.20430991e-01 -2.38702163e-01 7.26432651e-02 1.61805123e-01
7.60532498e-01 -4.26609442e-02 5.91864824e-01 -1.12506509e+00
-7.98268437e-01 -5.03672183e-01 -1.16375349e-01 -8.46917868e-01
-2.81730175e-01 -1.19934058e+00 -4.56413686e-01 -1.31163585e+00
-2.90077422e-02 -1.26157415e+00 1.21961623e-01 3.83776546e-01
6.50944769e-01 9.92518514e-02 1.77498069e-02 -5.73500432e-02
-1.87602099e-02 -6.41234666e-02 1.61865973e+00 -2.35598788e-01
-1.76468581e-01 -1.27287125e-02 -8.40556562e-01 6.05612159e-01
1.27361879e-01 -2.52877206e-01 -5.73954344e-01 -1.56755418e-01
7.07188725e-01 1.06207669e+00 4.39460188e-01 -9.31015432e-01
1.08787075e-01 -2.72136033e-01 2.20211804e-01 -3.12313557e-01
5.31028748e-01 -5.49484968e-01 5.37745893e-01 7.48344697e-03
-4.74552996e-02 5.36549054e-02 -1.21311486e-01 3.32976341e-01
2.10999683e-01 -2.70460427e-01 1.18724656e+00 -4.31208372e-01
3.14476728e-01 2.13552013e-01 -1.06576413e-01 2.42420167e-01
1.24068260e+00 -3.77493471e-01 -3.10096979e-01 -7.69583285e-01
-9.56788182e-01 -1.02681816e-02 4.04195368e-01 -5.81388235e-01
1.54615581e-01 -1.10829926e+00 -7.22574949e-01 6.03354931e-01
-1.26729161e-01 7.38503635e-01 5.95838547e-01 1.10479236e+00
-6.41963542e-01 -1.64346129e-01 3.69602263e-01 -6.58286572e-01
-4.93242264e-01 1.47153050e-01 4.84949917e-01 -2.93962181e-01
-3.86147141e-01 1.56400251e+00 -2.23518625e-01 -2.14020044e-01
1.97121829e-01 -1.84804499e-01 6.49762392e-01 -2.61803329e-01
-1.85938571e-02 1.00649190e+00 3.21905643e-01 -6.57644808e-01
-1.87649682e-01 4.55207735e-01 3.22394699e-01 -2.91730344e-01
1.14635336e+00 -1.28527924e-01 -8.21825624e-01 -2.50185370e-01
1.33556867e+00 3.64785701e-01 -8.91073823e-01 2.87917316e-01
-6.00971818e-01 -3.86929750e-01 -6.31643057e-01 -6.91138923e-01
-1.23948348e+00 7.81421542e-01 6.23609066e-01 6.41382933e-01
1.14841747e+00 8.45827162e-02 3.93568933e-01 -1.72721773e-01
8.29276800e-01 -1.47376406e+00 2.15837151e-01 4.26381767e-01
6.01928592e-01 -5.53069651e-01 1.68377712e-01 -5.39851487e-01
-1.96286559e-01 9.64973032e-01 1.38620257e-01 -4.13447618e-01
1.11982393e+00 7.34676346e-02 -3.54989886e-01 -5.86338341e-01
3.58930945e-01 1.23280562e-01 1.01192325e-01 -8.41815863e-03
6.33516073e-01 2.04941571e-01 -4.94929165e-01 1.02779937e+00
-6.19498074e-01 -2.56373942e-01 4.19556588e-01 9.80037451e-01
-3.65047902e-01 -1.14005256e+00 -5.05735278e-01 8.65392804e-01
-3.70175004e-01 4.04680297e-02 2.29120642e-01 7.56268501e-01
4.07716036e-01 9.07639027e-01 -1.83472604e-01 -8.52703676e-02
3.17257166e-01 2.24957615e-01 1.04856229e+00 -4.07090485e-01
-3.82279813e-01 6.30687654e-01 -1.76492538e-02 -2.61061668e-01
1.46012723e-01 -6.13703251e-01 -1.92395711e+00 -5.72699428e-01
3.79549190e-02 2.66750038e-01 2.67098218e-01 5.80409467e-01
-4.71545197e-03 2.57381320e-01 6.80507302e-01 -4.39768642e-01
-5.50707281e-01 -5.66525638e-01 -1.49319243e+00 5.38837910e-01
-1.49946451e-01 -5.50593674e-01 -4.46561337e-01 -9.53039825e-02] | [6.544025897979736, 4.70329475402832] |
0ff03764-faa4-44e9-8648-bb6e92b95ef5 | small-temperature-is-all-you-need-for | 2306.06855 | null | https://arxiv.org/abs/2306.06855v1 | https://arxiv.org/pdf/2306.06855v1.pdf | Small Temperature is All You Need for Differentiable Architecture Search | Differentiable architecture search (DARTS) yields highly efficient gradient-based neural architecture search (NAS) by relaxing the discrete operation selection to optimize continuous architecture parameters that maps NAS from the discrete optimization to a continuous problem. DARTS then remaps the relaxed supernet back to the discrete space by one-off post-search pruning to obtain the final architecture (finalnet). Some emerging works argue that this remap is inherently prone to mismatch the network between training and evaluation which leads to performance discrepancy and even model collapse in extreme cases. We propose to close the gap between the relaxed supernet in training and the pruned finalnet in evaluation through utilizing small temperature to sparsify the continuous distribution in the training phase. To this end, we first formulate sparse-noisy softmax to get around gradient saturation. We then propose an exponential temperature schedule to better control the outbound distribution and elaborate an entropy-based adaptive scheme to finally achieve the enhancement. We conduct extensive experiments to verify the efficiency and efficacy of our method. | ['Zhiming Ding', 'Jiuling Zhang'] | 2023-06-12 | null | null | null | null | ['architecture-search'] | ['methodology'] | [ 2.57787168e-01 2.35864848e-01 -9.03194919e-02 -5.85458457e-01
-7.30108798e-01 -4.33725446e-01 9.21288356e-02 -4.43934888e-01
-6.59391224e-01 7.62428403e-01 -2.08197251e-01 -3.52761835e-01
-2.37622902e-01 -4.94864553e-01 -6.85049951e-01 -1.02823877e+00
1.82874724e-01 4.70739156e-01 1.55566111e-01 -6.66781291e-02
1.81018740e-01 4.03357238e-01 -1.33746052e+00 5.53536080e-02
1.21143377e+00 1.12812603e+00 6.03182435e-01 4.65995818e-01
-2.05597818e-01 3.87315929e-01 -6.43172026e-01 -2.70859808e-01
5.42562425e-01 -6.22143030e-01 -7.06277847e-01 -9.99208242e-02
2.32289612e-01 -1.27123967e-01 -6.89574108e-02 1.35541844e+00
4.85637456e-01 3.04886878e-01 2.47857854e-01 -1.13940799e+00
-4.62196916e-01 9.27385569e-01 -7.03128099e-01 3.68336648e-01
-5.78336537e-01 1.37371793e-01 9.83857036e-01 -7.42450118e-01
2.73344010e-01 1.11712348e+00 4.50900853e-01 6.35171413e-01
-1.03369009e+00 -7.15269089e-01 6.30208969e-01 5.85545488e-02
-1.41509235e+00 -5.24668336e-01 8.95593584e-01 2.12218776e-01
8.49668324e-01 4.75142390e-01 7.65190661e-01 7.45426178e-01
-2.43166983e-01 7.98664212e-01 8.31987321e-01 -4.81033266e-01
5.61331451e-01 1.61544219e-01 3.07226349e-02 9.41559494e-01
2.13927567e-01 6.76217899e-02 -2.11228490e-01 -1.45333380e-01
7.23946273e-01 -1.43977523e-01 -1.45832986e-01 -2.00237408e-02
-7.00477719e-01 6.86065257e-01 8.03421378e-01 3.15961599e-01
-6.33093536e-01 3.64440769e-01 2.32783243e-01 4.14425224e-01
1.46864593e-01 6.60028100e-01 -6.47152483e-01 1.15422852e-01
-1.41094494e+00 1.04360737e-01 4.80191052e-01 5.88469267e-01
6.53272808e-01 5.23252964e-01 -1.83484703e-01 1.11026669e+00
4.16697592e-01 1.20115921e-01 7.69024372e-01 -9.78066564e-01
5.52248001e-01 6.15937471e-01 -3.46498132e-01 -5.82032561e-01
-2.61536121e-01 -1.27331376e+00 -8.65053654e-01 1.74153432e-01
1.45870462e-01 -3.74899268e-01 -1.11824453e+00 2.02853560e+00
4.48573649e-01 3.81269991e-01 -1.82687372e-01 1.11755323e+00
4.42515373e-01 8.29483986e-01 3.67433056e-02 -2.34296560e-01
1.10096836e+00 -1.26309013e+00 -4.18577582e-01 -5.15057087e-01
5.11341751e-01 -8.90958309e-02 1.18382454e+00 3.27733546e-01
-1.37853026e+00 -3.51172626e-01 -1.34696341e+00 1.19274415e-01
1.00272559e-01 3.58401120e-01 3.70047152e-01 6.05169475e-01
-1.20045662e+00 5.09995759e-01 -1.20046616e+00 2.51815617e-01
5.65303683e-01 7.31597006e-01 1.15859449e-01 2.06448629e-01
-1.14756727e+00 5.89232087e-01 7.34614074e-01 5.23825347e-01
-8.50731552e-01 -5.89610755e-01 -5.49965918e-01 2.74317145e-01
5.06166101e-01 -9.36039746e-01 9.71472561e-01 -1.16463780e+00
-1.77084422e+00 4.78719711e-01 -1.06931470e-01 -5.78036547e-01
2.01193497e-01 2.78102249e-01 -1.07508391e-01 -2.46106401e-01
-4.00951058e-01 8.79230261e-01 9.90080774e-01 -1.17522824e+00
-5.96204042e-01 -5.72994173e-01 -3.37540768e-02 6.43394589e-01
-7.56998360e-01 -1.94810480e-01 -5.11151373e-01 -6.13761842e-01
5.47866523e-01 -1.07561779e+00 -5.21583378e-01 -4.41888988e-01
-5.73587775e-01 -1.53342903e-01 3.08094293e-01 -5.08178949e-01
1.78401923e+00 -1.99592495e+00 4.49612796e-01 4.86303300e-01
3.12843502e-01 2.94325441e-01 -3.35572243e-01 -2.60933936e-01
-6.96047172e-02 2.07664862e-01 -4.63165253e-01 -8.11333477e-01
-3.78774963e-02 3.91817033e-01 -1.82341114e-01 2.92864799e-01
6.06256165e-02 8.37978959e-01 -4.06178802e-01 -5.39367259e-01
-3.59301984e-01 2.78265059e-01 -8.16674113e-01 -2.62680626e-03
-2.01600835e-01 1.48856059e-01 -7.61231184e-01 6.03797376e-01
7.62743235e-01 -4.05751586e-01 1.30605072e-01 -1.03903122e-01
-1.56923234e-01 4.04276937e-01 -1.14485610e+00 1.69555557e+00
-4.00803387e-01 2.03976810e-01 4.89824384e-01 -1.12121296e+00
9.73805785e-01 -6.97621778e-02 2.67742962e-01 -5.22813082e-01
2.50329971e-01 2.67782271e-01 1.33536696e-01 -1.93225935e-01
4.94529992e-01 -9.73241329e-02 1.75171420e-01 2.71929383e-01
-6.36298656e-02 1.09991275e-01 -3.07285070e-01 -1.45258039e-01
9.22398984e-01 -6.86976984e-02 -2.18263432e-01 -2.91194826e-01
5.71963370e-01 -1.81311220e-01 8.32096517e-01 5.51698506e-01
-1.80262014e-01 4.69730854e-01 4.35585350e-01 -1.79812729e-01
-1.02507484e+00 -8.00870717e-01 1.05369456e-01 1.23950815e+00
-8.63639489e-02 3.25987339e-02 -1.19236422e+00 -7.16693342e-01
-3.33945721e-01 7.30991840e-01 -5.80176950e-01 -4.40074742e-01
-8.98278594e-01 -1.14698803e+00 5.60026467e-01 3.16500008e-01
7.57858038e-01 -1.02766204e+00 -6.03209555e-01 -3.73713709e-02
-3.21259163e-02 -5.63648880e-01 -6.69855237e-01 6.62042677e-01
-1.23760843e+00 -2.77524710e-01 -7.56459475e-01 -1.12661612e+00
8.84546041e-01 -4.48653758e-01 8.04489493e-01 4.21773374e-01
6.60429150e-02 -6.69013798e-01 5.85887916e-02 -6.09590299e-02
-2.65186280e-01 7.63383687e-01 -5.93511164e-02 -1.04096927e-01
-2.07731146e-02 -7.93694377e-01 -8.18236649e-01 1.23589374e-01
-9.14382279e-01 2.21057702e-02 8.54492903e-01 9.99827981e-01
9.48828697e-01 1.93930611e-01 6.83075905e-01 -4.53687012e-01
8.49859476e-01 -5.28779030e-01 -8.26035678e-01 4.40916210e-01
-1.12287343e+00 7.01101780e-01 7.78148055e-01 -6.04902267e-01
-8.96676838e-01 1.09398350e-01 -2.52096146e-01 -7.21058607e-01
2.36458048e-01 6.59123838e-01 -1.19706094e-01 -7.65252560e-02
5.38191557e-01 3.45590740e-01 -4.64464501e-02 -4.67465878e-01
5.06847277e-02 3.87666643e-01 5.38476765e-01 -4.83321816e-01
7.21973300e-01 5.31447195e-02 -1.77541316e-01 -3.05812001e-01
-7.98328698e-01 1.20721005e-01 -1.27684653e-01 2.41601601e-01
5.16519964e-01 -5.76967120e-01 -6.24437451e-01 1.24910966e-01
-8.41129720e-01 -4.64617372e-01 -2.88482040e-01 4.48643535e-01
-1.71728447e-01 7.50198215e-02 -6.20362878e-01 -8.78657103e-01
-7.59808421e-01 -1.29852676e+00 7.06393063e-01 4.24101889e-01
1.69551045e-01 -9.17264760e-01 2.08191574e-01 1.71315193e-01
4.65516329e-01 -1.03050195e-01 9.41168427e-01 -7.04541564e-01
-3.96358579e-01 8.84536728e-02 -8.22580606e-03 4.19371039e-01
-4.02117729e-01 -2.90032923e-01 -5.76982737e-01 -4.53661114e-01
4.06149447e-01 -3.03846717e-01 1.15917778e+00 6.23620272e-01
1.40556073e+00 -4.73273814e-01 -2.66523033e-01 1.22330892e+00
1.57011986e+00 3.51372540e-01 6.43447340e-01 6.46747708e-01
4.85413045e-01 2.31226653e-01 4.13479835e-01 4.39378768e-01
2.18255982e-01 4.71091717e-01 6.64105535e-01 -3.96278948e-02
-8.03660452e-02 -2.10037395e-01 3.90752286e-01 1.14328897e+00
5.24705090e-02 -2.24298537e-01 -6.38134539e-01 4.47879463e-01
-1.64335382e+00 -6.71123683e-01 4.44909692e-01 2.10815144e+00
1.30068374e+00 2.14562416e-01 -3.81344408e-02 -1.02926210e-01
8.26178133e-01 9.97122079e-02 -1.03264499e+00 -4.74611908e-01
-7.75383785e-02 2.21630171e-01 6.49682045e-01 5.43413162e-01
-6.54809892e-01 9.85834122e-01 6.12097597e+00 1.08367252e+00
-1.27994204e+00 1.85319737e-01 9.25390482e-01 -5.20209730e-01
-5.49773455e-01 8.12620744e-02 -1.02642953e+00 6.00197434e-01
9.77288067e-01 2.41802379e-01 1.00590003e+00 8.59733462e-01
8.37025139e-03 3.16033483e-01 -7.09789395e-01 9.67804670e-01
-2.23734766e-01 -1.11835718e+00 -2.67938524e-01 -3.55616137e-02
8.29187453e-01 1.85421795e-01 4.45374489e-01 4.32483286e-01
2.43603513e-01 -1.02769089e+00 8.64512503e-01 3.11842620e-01
5.68024695e-01 -8.78167748e-01 3.71709824e-01 5.43106318e-01
-1.02739847e+00 -5.10591924e-01 -5.60257077e-01 5.52306831e-01
2.00170428e-02 4.74608481e-01 -8.98926020e-01 1.03040442e-01
4.56387550e-01 -1.53701320e-01 -5.55164397e-01 1.02054679e+00
-4.10432778e-02 6.33207619e-01 -6.77552164e-01 -2.43331388e-01
4.75356311e-01 -4.79282349e-01 7.23768890e-01 7.20948398e-01
4.33478147e-01 -7.56799430e-02 2.16903873e-02 1.28608894e+00
-2.17395470e-01 -1.85171217e-01 1.36899650e-01 4.57731895e-02
9.03588891e-01 1.11520708e+00 -6.35754406e-01 -1.80758178e-01
1.40361503e-01 9.20961320e-01 6.92704380e-01 5.88980794e-01
-1.14235985e+00 -5.02047956e-01 3.11953545e-01 -1.28989577e-01
4.79081154e-01 -7.76533931e-02 -6.38993502e-01 -9.09626126e-01
2.02436730e-01 -7.83418477e-01 3.21093857e-01 -3.80200654e-01
-6.90197766e-01 1.21576178e+00 -1.34350777e-01 -6.72860324e-01
-2.25274548e-01 -1.00528948e-01 -6.74146533e-01 7.94538319e-01
-1.46357000e+00 -8.24217319e-01 -1.27958983e-01 5.05524933e-01
6.28175378e-01 -3.09541702e-01 3.90896976e-01 4.34921086e-01
-1.17040300e+00 1.25536144e+00 1.76075861e-01 -3.14387023e-01
2.54410565e-01 -1.05535388e+00 3.41553748e-01 7.23104060e-01
-1.00419410e-01 6.86697304e-01 7.02384233e-01 -5.34358442e-01
-1.17451334e+00 -9.87488091e-01 4.43914503e-01 2.76559889e-01
3.95660400e-01 -2.42655605e-01 -9.26530838e-01 5.08034229e-01
5.30503355e-02 -1.56659350e-01 2.01751247e-01 -3.03666107e-02
-3.66706401e-02 -3.47560674e-01 -1.13107979e+00 6.82927430e-01
1.05348158e+00 -1.33554548e-01 -3.84610385e-01 8.75858366e-02
9.58579063e-01 -3.53952944e-01 -6.01529539e-01 4.39887673e-01
3.43333393e-01 -5.78077018e-01 8.49542737e-01 -6.09470725e-01
2.08037809e-01 -2.35034540e-01 -1.19567148e-01 -1.55326378e+00
-3.29794616e-01 -8.36176217e-01 -2.31234208e-01 1.05464029e+00
9.16310251e-01 -7.83601344e-01 1.24721014e+00 6.40299559e-01
-5.82977831e-01 -1.29513741e+00 -1.04100227e+00 -5.90284705e-01
1.78108811e-01 -3.30130570e-02 9.30039942e-01 6.81810617e-01
-1.72904506e-01 1.37588143e-01 -1.76253229e-01 1.60524100e-01
5.17603219e-01 6.42672479e-02 -3.93246114e-03 -9.12841678e-01
-6.29396737e-01 -6.72657430e-01 1.35295108e-01 -1.31698477e+00
1.62560910e-01 -9.17638183e-01 2.35723779e-01 -1.17729795e+00
9.96054411e-02 -1.00231397e+00 -7.31703103e-01 5.42574704e-01
-2.58711010e-01 -7.27783591e-02 4.48765792e-03 3.06277364e-01
-5.52468419e-01 8.11519861e-01 1.35480356e+00 9.74942520e-02
-6.39660478e-01 1.35006487e-01 -7.74981976e-01 3.90771747e-01
9.40969527e-01 -7.03750372e-01 -7.39062786e-01 -6.99072182e-01
3.87935579e-01 5.75133488e-02 5.87065704e-02 -8.88874054e-01
4.80007678e-01 -1.44201249e-01 1.36322826e-01 -3.80355388e-01
2.91850239e-01 -6.44768238e-01 4.69238497e-02 4.43153709e-01
-8.61411035e-01 2.53314018e-01 8.61446112e-02 3.92410845e-01
-8.90882611e-02 -7.05182493e-01 8.87242317e-01 -2.07564794e-02
-2.70378828e-01 4.79938626e-01 2.05220401e-01 8.34428519e-02
7.54685104e-01 -2.44323581e-01 -1.37372047e-01 -7.71241039e-02
-8.67256403e-01 5.60508370e-01 2.80862987e-01 -6.59552068e-02
6.91541255e-01 -1.33377457e+00 -5.35950363e-01 3.93036157e-01
-6.44869566e-01 3.72496367e-01 2.19495073e-01 9.90060627e-01
-4.83262748e-01 3.37927312e-01 -6.90591112e-02 -3.41727287e-01
-1.06037390e+00 3.37663978e-01 5.80993414e-01 -5.05343497e-01
-2.98867434e-01 1.34097099e+00 1.00072928e-01 -4.46310341e-01
5.65726399e-01 -2.88241714e-01 -1.74875744e-02 -3.25400740e-01
2.64226615e-01 2.33453438e-01 1.58936590e-01 -6.40169485e-03
-3.61021280e-01 1.18145391e-01 -3.14138770e-01 -3.27632248e-01
1.39759636e+00 -2.85954237e-01 -2.07226261e-01 -5.13271475e-03
1.45445323e+00 -2.96416044e-01 -1.51139033e+00 -1.08883478e-01
-1.89680189e-01 -9.51402914e-03 5.52152216e-01 -7.59129763e-01
-1.53874695e+00 5.71412802e-01 8.18167806e-01 2.20772162e-01
1.44092214e+00 -6.09080531e-02 1.13386786e+00 5.56900918e-01
-2.46122614e-01 -1.34630728e+00 -1.13066100e-01 4.80167657e-01
7.58977473e-01 -7.47594535e-01 -9.86067578e-02 6.34008795e-02
-6.46547794e-01 9.65365946e-01 8.17916036e-01 -2.90959805e-01
5.22003055e-01 4.71044689e-01 -3.76194715e-01 -8.81094187e-02
-1.01776874e+00 2.21487433e-01 1.21434614e-01 8.41795653e-02
-4.13627252e-02 -7.09787831e-02 -2.41335347e-01 8.57894957e-01
-4.50073183e-01 -1.35889679e-01 -8.49615857e-02 6.97487593e-01
-6.49204969e-01 -1.09907317e+00 -2.19585508e-01 4.53339487e-01
-5.01490593e-01 -4.02690530e-01 -1.19777516e-01 3.25759828e-01
1.47775281e-02 4.95848805e-01 1.48773029e-01 -5.14095128e-01
-3.59424055e-02 2.05462873e-01 3.79065603e-01 -2.69441456e-01
-7.84506440e-01 2.24363148e-01 -1.15329795e-01 -3.36778134e-01
2.00275749e-01 -5.05535841e-01 -1.52186513e+00 -9.18083489e-02
-7.73908675e-01 3.27709526e-01 9.15749013e-01 8.85798514e-01
4.95220691e-01 7.58151829e-01 7.55482316e-01 -5.28259635e-01
-1.33453083e+00 -8.57921541e-01 -3.69934499e-01 -1.24865092e-01
5.00534236e-01 -3.84306818e-01 -6.10629499e-01 -3.57583702e-01] | [8.62331485748291, 3.3578200340270996] |
fed4ae41-4c41-41b3-9ed3-626c149c213e | a-temporal-graph-neural-network-for-cyber | 2212.03390 | null | https://arxiv.org/abs/2212.03390v1 | https://arxiv.org/pdf/2212.03390v1.pdf | A Temporal Graph Neural Network for Cyber Attack Detection and Localization in Smart Grids | This paper presents a Temporal Graph Neural Network (TGNN) framework for detection and localization of false data injection and ramp attacks on the system state in smart grids. Capturing the topological information of the system through the GNN framework along with the state measurements can improve the performance of the detection mechanism. The problem is formulated as a classification problem through a GNN with message passing mechanism to identify abnormal measurements. The residual block used in the aggregation process of message passing and the gated recurrent unit can lead to improved computational time and performance. The performance of the proposed model has been evaluated through extensive simulations of power system states and attack scenarios showing promising performance. The sensitivity of the model to intensity and location of the attacks and model's detection delay versus detection accuracy have also been evaluated. | ['Mia Naeini', 'Md Abul Hasnat', 'Seyed Hamed Haghshenas'] | 2022-12-07 | null | null | null | null | ['smart-grid-prediction', 'cyber-attack-detection'] | ['miscellaneous', 'miscellaneous'] | [-1.66664198e-01 -1.69024631e-01 2.17472449e-01 1.85693413e-01
-1.01222798e-01 -5.77437937e-01 5.19151270e-01 5.82310915e-01
2.39223748e-01 7.29089737e-01 -2.05633268e-01 -7.53779829e-01
-4.49358046e-01 -9.01726484e-01 -1.29635870e-01 -1.00241959e+00
-1.09303713e+00 3.55667435e-02 9.20209661e-02 -2.32070431e-01
9.73586664e-02 7.79901922e-01 -6.93664074e-01 -1.91564545e-01
4.43521410e-01 1.06005013e+00 -2.29597256e-01 9.40513253e-01
4.41992193e-01 8.73165429e-01 -1.47186041e+00 5.90563297e-01
3.57494563e-01 -3.69176209e-01 -3.72972906e-01 -8.72432441e-02
-6.58530056e-01 -3.19389492e-01 -6.17150128e-01 1.46470261e+00
7.82248020e-01 1.82859182e-01 3.74947876e-01 -1.44086945e+00
-1.55243538e-02 8.63723457e-01 -4.36158299e-01 8.73753786e-01
2.97845840e-01 1.82119012e-01 2.98597187e-01 -3.69480252e-01
9.74965021e-02 9.30867255e-01 7.94948637e-01 -2.81903028e-01
-1.03963220e+00 -7.02787459e-01 -4.70766351e-02 3.48551035e-01
-1.44032264e+00 3.23767900e-01 8.95552337e-01 -8.46258998e-02
1.26080060e+00 2.12473422e-01 7.61763752e-01 4.49008316e-01
5.41962206e-01 2.49818742e-01 6.46090209e-01 -2.62849331e-01
5.55646122e-01 -3.52603614e-01 5.06961703e-01 4.51254904e-01
4.95784730e-01 2.85881996e-01 -2.15888303e-02 -4.11739290e-01
6.26704156e-01 -1.22903258e-01 -2.77401984e-01 2.50709981e-01
-3.53331625e-01 9.18867230e-01 5.89360178e-01 5.20723403e-01
-7.49617517e-01 1.97111636e-01 8.84461522e-01 4.75417882e-01
5.68084121e-01 2.05429390e-01 -3.06365669e-01 -1.56665500e-03
-8.10643673e-01 -4.44434673e-01 9.58279133e-01 5.26492596e-01
7.82534406e-02 1.35649025e+00 1.49470538e-01 -2.98569314e-02
2.27578029e-01 6.76881850e-01 4.27420050e-01 -1.48287248e-02
3.16043884e-01 7.90447772e-01 -1.18378572e-01 -1.38211751e+00
-9.84081447e-01 -9.27760541e-01 -1.19169509e+00 2.10280880e-01
2.23148301e-01 -7.49178112e-01 -8.33992720e-01 1.34043229e+00
3.63505781e-01 6.51911676e-01 1.01491369e-01 4.58577514e-01
2.15186849e-01 1.30246520e+00 -1.80527300e-01 -4.92113829e-01
9.85725284e-01 -1.86686128e-01 -1.01369691e+00 3.07566971e-01
1.01929832e+00 -4.11541432e-01 -2.34383002e-01 3.64857197e-01
-8.82281005e-01 -2.56671846e-01 -1.32375216e+00 9.38310385e-01
-5.54162562e-01 2.00708866e-01 2.66422778e-01 7.54118025e-01
-1.07824743e+00 7.53281653e-01 -9.85988796e-01 -4.20767665e-01
-8.71697366e-02 7.46385634e-01 2.42841184e-01 5.61391592e-01
-1.53462696e+00 1.16921473e+00 7.71814167e-01 8.27262819e-01
-1.01585650e+00 -4.13483113e-01 -6.79719269e-01 7.84271732e-02
1.24859825e-01 1.45213446e-02 7.12402463e-01 -4.59274679e-01
-1.14081872e+00 -3.75603378e-01 4.80654329e-01 -1.22938287e+00
2.78089583e-01 3.46544117e-01 -1.03580177e+00 2.28428409e-01
-2.56911874e-01 -7.78369129e-01 6.21243060e-01 -5.09375036e-01
-5.87184787e-01 -3.46769035e-01 -3.16351920e-01 -1.20776318e-01
9.07036737e-02 -3.09756473e-02 7.38545775e-01 -4.85888511e-01
2.20595941e-01 -5.22459090e-01 -5.35193384e-01 -1.06264198e+00
-6.33541763e-01 -8.95974338e-02 1.56024861e+00 -1.00458598e+00
1.39552772e+00 -1.94689393e+00 -4.53894436e-01 1.00321937e+00
-1.32136345e-01 6.96975589e-01 2.31422529e-01 1.19124877e+00
-4.65664685e-01 -3.14742744e-01 2.78558314e-01 2.75246203e-01
-1.62384883e-01 2.49128088e-01 -5.80685318e-01 1.11859941e+00
-1.09216779e-01 7.33657122e-01 -7.01549232e-01 4.63224143e-01
6.96268797e-01 5.86293817e-01 2.51325488e-01 1.12490848e-01
2.11968794e-01 2.65564471e-01 -5.98882675e-01 2.97481507e-01
5.46564758e-01 -3.57707411e-01 2.44422600e-01 -3.41614753e-01
-5.49556501e-03 3.63338500e-01 -1.50517964e+00 4.81250942e-01
-1.86751321e-01 5.73256433e-01 1.09423064e-01 -1.43556738e+00
1.09168231e+00 6.55425906e-01 5.54228663e-01 -6.96408391e-01
5.69454730e-01 -2.49788254e-01 1.64332598e-01 -4.69902977e-02
1.35252118e-01 4.79796916e-01 -1.76119879e-01 8.28390598e-01
1.64069831e-01 2.18625993e-01 2.49483660e-01 3.82828087e-01
1.30392933e+00 -6.44520938e-01 5.36194086e-01 -5.63483715e-01
6.08258963e-01 -5.34979589e-02 4.96412009e-01 7.65487790e-01
-1.80230364e-02 -6.24701202e-01 7.09204733e-01 -6.64336741e-01
-8.06893229e-01 -7.78442204e-01 1.03884950e-01 2.73102939e-01
-1.33943371e-02 -2.51039922e-01 -4.76661921e-01 -4.76210058e-01
-1.68722436e-01 1.05885422e+00 -5.52930951e-01 -6.66929364e-01
-5.98159552e-01 -1.52408266e+00 6.32266283e-01 4.08384353e-01
4.44544107e-01 -9.39263403e-01 -6.18628621e-01 6.77597821e-01
4.38354194e-01 -1.02068698e+00 2.08585411e-01 8.15520227e-01
-9.06061411e-01 -1.20435107e+00 1.59581099e-02 -4.79434311e-01
9.66803372e-01 -2.56791145e-01 7.34412730e-01 2.41639361e-01
-2.46466413e-01 2.24792585e-01 -3.42269063e-01 -2.44886488e-01
-8.21025014e-01 -2.53944933e-01 3.95491987e-01 -7.64899999e-02
-6.05637841e-02 -7.46949375e-01 -3.80760044e-01 1.71254218e-01
-6.75681472e-01 -5.21754205e-01 4.68459651e-02 7.94907033e-01
-2.10771725e-01 1.00038683e+00 8.35950673e-01 -5.50949156e-01
1.00773597e+00 -5.43384075e-01 -1.47024632e+00 6.12130985e-02
-8.70997846e-01 -1.14043526e-01 9.61874902e-01 -3.82467002e-01
-4.44531113e-01 -2.14300483e-01 2.67776936e-01 -1.82032540e-01
1.35682523e-01 6.57551050e-01 1.25109985e-01 -5.34035742e-01
2.82898873e-01 3.61103565e-01 -7.47055709e-02 -2.61578768e-01
7.62920361e-03 3.51275474e-01 5.46485782e-01 8.66168588e-02
9.81911719e-01 2.59062439e-01 5.00500023e-01 -9.73378241e-01
-1.38336673e-01 -2.06569895e-01 -4.19954956e-01 -4.49466586e-01
2.70353526e-01 -7.34784245e-01 -1.19501340e+00 7.43365884e-01
-1.08751178e+00 -3.38803455e-02 -9.54998359e-02 4.90958035e-01
9.38791931e-02 3.40174437e-01 -1.17726481e+00 -1.21680868e+00
-7.36362755e-01 -8.92548859e-01 2.07167402e-01 2.52857864e-01
3.06989312e-01 -1.57270944e+00 -5.31649366e-02 -6.03185236e-01
6.45205796e-01 5.62087417e-01 8.65377724e-01 -1.20894146e+00
-5.75413942e-01 -8.20639849e-01 1.86260864e-02 3.17536294e-01
2.40176514e-01 2.08501071e-02 -5.28794587e-01 -9.05506968e-01
5.13308167e-01 4.15709674e-01 -1.42548859e-01 6.85868323e-01
1.83026031e-01 -6.61323547e-01 -3.35614055e-01 4.42457914e-01
1.77979457e+00 7.50016272e-01 2.51164675e-01 9.74714085e-02
4.72702801e-01 -2.07380448e-02 1.40217915e-01 7.44950712e-01
-2.15716407e-01 1.55091941e-01 6.43504322e-01 -1.84966132e-01
4.44915861e-01 2.33079325e-02 4.59360629e-01 1.24452186e+00
3.40334356e-01 -4.68255997e-01 -9.11387086e-01 3.43672216e-01
-1.51777124e+00 -9.59326684e-01 -2.83757031e-01 1.93393397e+00
-1.41166165e-01 3.36286783e-01 6.72012046e-02 8.43578875e-01
1.08811975e+00 1.90419734e-01 -4.32401747e-01 -7.19480872e-01
-2.31907010e-01 4.27764505e-02 8.10699165e-01 5.65988183e-01
-7.75817871e-01 2.78147995e-01 6.58876562e+00 7.04529107e-01
-1.33213699e+00 -5.39924875e-02 4.75300014e-01 3.61839771e-01
7.32728481e-01 -8.85750651e-02 -7.48097658e-01 4.86324489e-01
1.57026815e+00 -6.01593494e-01 4.02528584e-01 4.31312561e-01
6.67398214e-01 -2.64645398e-01 -5.36684275e-01 5.81839085e-01
-2.87470352e-02 -1.07513034e+00 -4.77765016e-02 9.37449187e-02
6.82366729e-01 4.51862603e-01 -2.55116522e-01 2.30116546e-02
7.11608887e-01 -7.09169447e-01 2.07061946e-01 1.20369963e-01
6.31036088e-02 -1.21055448e+00 1.06373155e+00 3.41195792e-01
-1.43233156e+00 -5.37416577e-01 2.35123727e-02 -3.27883393e-01
7.35570729e-01 7.84088671e-01 -1.43333852e+00 9.60653961e-01
3.33141744e-01 6.48674726e-01 -5.12845337e-01 1.17237318e+00
-3.39347690e-01 1.18787944e+00 -8.27844739e-01 -3.02355140e-01
5.10063291e-01 -2.79265285e-01 8.18594277e-01 8.51339221e-01
2.35207051e-01 5.48540913e-02 4.63251621e-01 6.20478988e-01
4.88764226e-01 -3.84030372e-01 -7.46341586e-01 1.51372060e-01
5.14229834e-01 1.16795170e+00 -1.14618170e+00 -3.94774854e-01
-1.13825150e-01 1.17647856e-01 -2.67230362e-01 6.25823021e-01
-6.93203390e-01 -4.37034547e-01 2.73189461e-03 -1.80522323e-01
2.84225523e-01 -1.36348680e-01 -2.10985541e-01 -5.82022607e-01
-2.10063756e-01 -4.44834173e-01 7.23584294e-01 -6.66056693e-01
-1.02627361e+00 8.03520024e-01 -2.44196430e-02 -1.35597444e+00
-7.68644273e-01 -2.07249641e-01 -1.01901841e+00 9.49803293e-01
-8.24903011e-01 -6.07478857e-01 3.69436294e-01 8.00207615e-01
-1.16564310e-03 -2.71872044e-01 6.94332778e-01 2.23973185e-01
-9.57702756e-01 1.02169320e-01 2.74528176e-01 6.08733475e-01
-6.37486458e-01 -1.05860519e+00 4.63702798e-01 1.62216163e+00
-8.23983625e-02 2.75047749e-01 1.12822652e+00 -1.01308274e+00
-1.32142866e+00 -1.09881163e+00 2.17054054e-01 2.11873457e-01
1.11732948e+00 -4.08094138e-01 -8.56364369e-01 7.43708968e-01
4.65984315e-01 1.94395602e-01 -1.08953908e-01 -4.53759670e-01
3.68755072e-01 -9.75543931e-02 -1.23539233e+00 2.40354225e-01
-4.00413685e-02 -5.08578837e-01 -4.40194815e-01 3.66560996e-01
1.09063201e-01 -2.71198660e-01 -9.17403698e-01 3.55472088e-01
-3.06498587e-01 -3.86670679e-01 6.08723581e-01 -3.57986569e-01
-8.58055949e-01 -6.12682641e-01 1.94317833e-01 -1.69534004e+00
-1.48327529e-01 -1.14334357e+00 -3.73300284e-01 8.74360204e-01
4.36012417e-01 -1.09625280e+00 5.39636075e-01 1.05896136e-02
1.79576606e-01 -4.45840955e-01 -1.31720912e+00 -6.52599990e-01
-4.29484665e-01 -2.24344090e-01 5.63554645e-01 9.29994047e-01
3.45154107e-01 3.57053578e-01 -3.18475991e-01 1.11002433e+00
7.53896236e-01 -1.08560920e-01 2.25276455e-01 -9.18079019e-01
5.00203110e-03 2.13858653e-02 -8.72810006e-01 -1.54423997e-01
-2.83732593e-01 -5.96958160e-01 -5.02038598e-01 -1.22504389e+00
-5.66077113e-01 4.60777730e-02 -5.27560711e-01 4.73643839e-01
2.63675749e-01 -1.29827186e-01 -1.16922017e-02 -1.76832855e-01
-3.65648121e-01 2.79588044e-01 2.87870198e-01 9.00363177e-02
-1.34040698e-01 2.70830423e-01 4.55113024e-01 5.26123464e-01
1.17293096e+00 -5.82232356e-01 -4.09160286e-01 -4.18817475e-02
1.36459246e-01 8.50526094e-01 2.71919042e-01 -1.32051909e+00
6.53664768e-01 6.35840952e-01 4.50524271e-01 -1.17694271e+00
-7.00532633e-04 -1.01341820e+00 4.69786704e-01 1.10093498e+00
1.17710851e-01 1.05548227e+00 3.99070591e-01 6.50162458e-01
-1.08506925e-01 1.46184072e-01 6.28018677e-01 2.52586991e-01
-3.95433635e-01 4.05685371e-03 -7.95755625e-01 -4.13239509e-01
1.16085672e+00 1.63716413e-02 -4.79635715e-01 -5.70385039e-01
-8.55157673e-01 4.40311760e-01 -3.93578172e-01 1.39659166e-01
3.46571922e-01 -1.10224485e+00 -4.51869607e-01 5.14111102e-01
-7.05077708e-01 -6.55813396e-01 2.72032525e-02 9.20301855e-01
-3.53955239e-01 6.29392147e-01 -1.49898157e-01 -6.53493106e-01
-1.08006740e+00 4.82666641e-01 9.99397516e-01 -5.61641276e-01
-6.13267839e-01 9.05969590e-02 -5.24553537e-01 1.18561827e-01
1.84172451e-01 -3.94494593e-01 -3.85660082e-01 -8.91792923e-02
4.74531740e-01 8.29332232e-01 3.83861184e-01 -5.46072364e-01
-2.13431284e-01 1.01681583e-01 -8.26252624e-02 2.26998404e-01
1.31075907e+00 -3.20851773e-01 -3.80456239e-01 4.64765280e-01
9.32959378e-01 -5.69300711e-01 -7.83375323e-01 2.88033448e-02
1.19433768e-01 2.74663806e-01 4.04740930e-01 -9.79435384e-01
-1.53331196e+00 4.16754454e-01 7.88228452e-01 1.03960669e+00
1.13616347e+00 -6.32951319e-01 4.30037081e-01 1.30151048e-01
5.29576063e-01 -9.21695650e-01 -5.41955471e-01 4.84764099e-01
2.01165125e-01 -4.21538234e-01 1.18838340e-01 -9.85712651e-03
1.19962111e-01 1.08848619e+00 1.38864070e-01 -6.87500060e-01
1.05564535e+00 8.27511907e-01 -3.14643197e-02 -4.23388332e-01
-1.02453017e+00 4.24081802e-01 -2.08741710e-01 5.21093845e-01
-3.63299131e-01 1.81083202e-01 -2.72281229e-01 2.23626956e-01
-6.06896691e-02 -2.29990512e-01 1.01645887e+00 9.65270877e-01
-1.99775204e-01 -3.33690464e-01 -4.97841507e-01 4.96240675e-01
-8.43627751e-01 9.98041853e-02 8.48571584e-02 8.48591864e-01
-4.85584974e-01 1.17454946e+00 -6.75129369e-02 -3.64087313e-01
3.21124911e-01 -2.77872115e-01 -3.10801268e-01 -1.00832969e-01
-9.76449788e-01 1.03873380e-01 4.30511646e-02 -5.55394232e-01
1.71488449e-01 -3.98994118e-01 -1.31230307e+00 -2.84570277e-01
-9.06067669e-01 9.19862509e-01 7.83793807e-01 8.65050077e-01
1.56254187e-01 8.33121717e-01 9.58291352e-01 -5.35425663e-01
-8.11404049e-01 -1.08237791e+00 -1.07141209e+00 -1.08210385e-01
6.13091946e-01 -3.89102548e-01 -9.09266293e-01 -6.74626887e-01] | [6.099612236022949, 2.584961414337158] |
b2b5b672-3d11-4f4f-b9c8-07805a448003 | unsupervised-image-segmentation-by-mutual | 2107.00691 | null | https://arxiv.org/abs/2107.00691v1 | https://arxiv.org/pdf/2107.00691v1.pdf | Unsupervised Image Segmentation by Mutual Information Maximization and Adversarial Regularization | Semantic segmentation is one of the basic, yet essential scene understanding tasks for an autonomous agent. The recent developments in supervised machine learning and neural networks have enjoyed great success in enhancing the performance of the state-of-the-art techniques for this task. However, their superior performance is highly reliant on the availability of a large-scale annotated dataset. In this paper, we propose a novel fully unsupervised semantic segmentation method, the so-called Information Maximization and Adversarial Regularization Segmentation (InMARS). Inspired by human perception which parses a scene into perceptual groups, rather than analyzing each pixel individually, our proposed approach first partitions an input image into meaningful regions (also known as superpixels). Next, it utilizes Mutual-Information-Maximization followed by an adversarial training strategy to cluster these regions into semantically meaningful classes. To customize an adversarial training scheme for the problem, we incorporate adversarial pixel noise along with spatial perturbations to impose photometrical and geometrical invariance on the deep neural network. Our experiments demonstrate that our method achieves the state-of-the-art performance on two commonly used unsupervised semantic segmentation datasets, COCO-Stuff, and Potsdam. | ['Hamid Rezatofighi', 'Ali Royat', 'S. Ehsan Mirsadeghi'] | 2021-07-01 | null | null | null | null | ['unsupervised-semantic-segmentation'] | ['computer-vision'] | [ 6.21632218e-01 3.19545597e-01 1.76074907e-01 -4.42210436e-01
-6.35391235e-01 -5.48739016e-01 5.11049092e-01 -5.72662763e-02
-5.95012844e-01 4.38233882e-01 -3.45157683e-01 -1.82080790e-01
1.74245119e-01 -8.59795630e-01 -7.08105922e-01 -7.91008353e-01
2.38646209e-01 4.18723345e-01 5.78234673e-01 -6.00110330e-02
2.58649230e-01 4.76360291e-01 -1.37482369e+00 -8.40287283e-02
1.24516737e+00 1.04373717e+00 3.04980367e-01 5.12247741e-01
-2.77764291e-01 7.05234706e-01 -4.94301170e-01 -2.37444177e-01
5.54884136e-01 -4.27381575e-01 -1.11933672e+00 4.68734831e-01
1.35517552e-01 6.44988120e-02 -9.33996215e-02 1.33617496e+00
7.37381652e-02 4.01856184e-01 7.09211171e-01 -1.11470687e+00
-4.74681824e-01 2.19938681e-01 -5.91571927e-01 -2.12025046e-01
-1.35087788e-01 -1.14319446e-02 8.02958488e-01 -5.53885460e-01
5.42859316e-01 1.03391683e+00 5.45328796e-01 3.45029384e-01
-1.13681734e+00 -3.67623478e-01 2.08480611e-01 7.53172413e-02
-1.26371431e+00 -1.08017333e-01 1.19200695e+00 -4.68463838e-01
4.73232895e-01 3.41923051e-02 4.25572366e-01 5.82098544e-01
-2.47791205e-02 7.42998958e-01 1.12668908e+00 -4.04993832e-01
5.28990805e-01 7.78990015e-02 -1.04795158e-01 6.91172719e-01
-1.25642493e-01 -1.37180269e-01 1.12828285e-01 1.65527970e-01
8.25156868e-01 4.03124914e-02 3.15351933e-02 -5.66877484e-01
-8.69655669e-01 9.56534863e-01 7.94075549e-01 1.60383075e-01
-3.87517750e-01 4.66580363e-03 2.02287197e-01 -1.20447166e-01
5.63308001e-01 4.42407787e-01 -3.43307585e-01 3.17738563e-01
-9.20953810e-01 -2.84424182e-02 6.22301757e-01 7.05647707e-01
1.02526355e+00 8.49883556e-02 2.63292521e-01 1.03688419e+00
3.21558744e-01 2.61654973e-01 4.80386049e-01 -1.15146446e+00
1.36031166e-01 8.40973675e-01 -5.68896160e-02 -9.80717599e-01
-4.10001487e-01 -4.03519094e-01 -8.38266373e-01 5.13891280e-01
4.34014797e-01 -1.57411784e-01 -1.39699590e+00 1.68033183e+00
5.42459667e-01 5.52288108e-02 3.38028461e-01 9.49027479e-01
6.01927936e-01 5.15075743e-01 2.80334204e-01 1.50248602e-01
1.18410921e+00 -1.25489926e+00 -3.87349606e-01 -6.46959782e-01
2.05516487e-01 -7.81727731e-01 8.54044437e-01 3.37468296e-01
-8.33804309e-01 -6.31530881e-01 -1.07603562e+00 -7.00841993e-02
-5.99455059e-01 -1.15440942e-01 7.35681057e-01 6.02050841e-01
-7.59458423e-01 5.25030315e-01 -1.01835299e+00 -4.19622183e-01
7.05482066e-01 2.53380090e-01 -3.69928062e-01 6.47825450e-02
-8.57118845e-01 5.99651515e-01 6.76441371e-01 -6.05270714e-02
-9.39616561e-01 -2.61696428e-01 -9.64682817e-01 -2.70332128e-01
5.60887992e-01 -5.64312458e-01 9.32121158e-01 -1.39072740e+00
-1.58996868e+00 1.08077145e+00 5.11120781e-02 -5.73415637e-01
5.03623724e-01 -1.62840724e-01 -3.45138907e-02 3.39116395e-01
2.46937767e-01 9.73340929e-01 9.22713399e-01 -1.67520201e+00
-5.61233640e-01 -4.63784009e-01 1.63001776e-01 2.82767594e-01
-4.72870432e-02 -1.49223179e-01 -7.19972849e-01 -8.39373469e-01
5.68918407e-01 -9.18203175e-01 -6.85166597e-01 -1.63049266e-01
-7.10431159e-01 2.79727834e-03 9.98631060e-01 -7.18630314e-01
4.95009184e-01 -2.32126856e+00 2.74486870e-01 1.75937623e-01
3.98704503e-03 2.78334618e-01 -2.26184428e-02 1.27241462e-01
2.01738253e-02 -3.79877687e-02 -9.67240036e-01 -5.16223371e-01
-1.65560156e-01 4.76388335e-01 -7.04317912e-02 5.32558024e-01
1.90464646e-01 7.43713796e-01 -7.84525871e-01 -5.33269227e-01
5.34014881e-01 3.89018685e-01 -5.59132099e-01 3.43688428e-01
-5.24615705e-01 1.01282704e+00 -5.86224437e-01 4.53903377e-01
8.29523921e-01 -1.13973744e-01 -1.14966862e-01 3.39366607e-02
-2.70529762e-02 -2.68063962e-01 -1.05490100e+00 1.96861231e+00
-1.74819112e-01 3.65866512e-01 1.88857943e-01 -1.47545469e+00
8.96136165e-01 -1.65034443e-01 6.41679585e-01 -4.71598804e-01
2.94534415e-01 1.26928523e-01 -2.30673566e-01 -3.74090999e-01
4.62160587e-01 -1.02830879e-01 -1.61588252e-01 1.86305583e-01
5.08175651e-03 -4.87894475e-01 7.53492117e-02 1.15010202e-01
7.38981485e-01 2.80166119e-01 1.10105656e-01 -3.11150521e-01
7.11749434e-01 2.56192237e-01 7.30658472e-01 5.55055678e-01
-2.85682529e-01 9.12820160e-01 2.67700285e-01 -2.02757865e-01
-1.09409297e+00 -1.16824841e+00 -1.13439344e-01 9.87455189e-01
5.47649920e-01 1.15936801e-01 -1.38781559e+00 -7.52571285e-01
-2.24831715e-01 6.03312850e-01 -5.36158204e-01 -1.48230987e-02
-3.73207062e-01 -7.54804909e-01 5.14235973e-01 4.86068904e-01
1.14338350e+00 -1.30635226e+00 -6.04302466e-01 1.34740397e-01
-1.35393828e-01 -1.43776846e+00 -2.40152553e-01 2.33630717e-01
-6.42161012e-01 -1.24304163e+00 -7.16298223e-01 -1.02042007e+00
8.79691958e-01 1.96390271e-01 8.79935801e-01 -2.08682612e-01
-4.05032694e-01 2.55295992e-01 -3.88665408e-01 -3.30367744e-01
-3.98301929e-01 5.07506169e-03 -3.04960549e-01 2.41015732e-01
1.11831285e-01 -6.36516571e-01 -6.59843147e-01 2.61679888e-01
-1.27361751e+00 1.80160165e-01 6.84505463e-01 5.35366595e-01
9.24086809e-01 3.12460929e-01 2.82295674e-01 -1.17396975e+00
1.14152908e-01 -2.89615571e-01 -6.45108759e-01 3.28464620e-02
-3.33218902e-01 -1.97628904e-02 7.59517789e-01 -1.73552111e-02
-1.26410615e+00 3.47091526e-01 -3.46491694e-01 -2.44933665e-01
-6.78654611e-01 3.73780251e-01 -3.68579060e-01 -2.45365724e-01
4.85558003e-01 2.34877229e-01 -1.72028348e-01 -2.92221010e-01
6.83786929e-01 6.56662881e-01 9.13655281e-01 -4.48265612e-01
8.11377764e-01 8.93633366e-01 3.57762575e-02 -1.04534042e+00
-9.84754264e-01 -6.52539134e-01 -9.29949522e-01 -8.75297263e-02
1.49023104e+00 -6.87804401e-01 -2.38590345e-01 8.62746954e-01
-9.01999712e-01 -5.22188127e-01 -2.79221028e-01 1.81550995e-01
-7.83410251e-01 5.80602884e-01 -3.18502009e-01 -6.45855129e-01
-2.29330122e-01 -1.29714108e+00 9.26560521e-01 4.97900218e-01
2.61577547e-01 -9.97446716e-01 -1.23614848e-01 7.66070366e-01
2.01526493e-01 5.86418808e-01 6.34011507e-01 -7.58842707e-01
-6.26320601e-01 -1.81898940e-02 -4.35586542e-01 7.30198324e-01
1.16184600e-01 -2.95102239e-01 -1.04760015e+00 4.57155704e-02
1.47654518e-01 -2.99693644e-01 9.49734151e-01 4.41286981e-01
1.38400972e+00 -8.87305196e-03 -1.74964994e-01 8.47831428e-01
1.51729941e+00 3.39287221e-01 7.40998745e-01 4.40633506e-01
1.06474257e+00 6.83300912e-01 5.33895552e-01 1.82915553e-01
3.91092211e-01 3.96287411e-01 6.61160111e-01 -4.46869195e-01
-1.78967603e-02 -1.15396298e-01 -3.40836905e-02 4.27967548e-01
1.20954923e-01 -1.45952061e-01 -8.01048458e-01 5.64140439e-01
-1.94785964e+00 -5.81569135e-01 -2.26580910e-02 2.09703422e+00
5.66293001e-01 1.78520486e-01 -1.29568443e-01 1.31423026e-01
7.30185270e-01 2.97996640e-01 -6.97475791e-01 -2.81986952e-01
-1.32057965e-01 2.19367817e-01 7.76797354e-01 4.29416090e-01
-1.58845675e+00 1.42301309e+00 5.36896706e+00 8.60315382e-01
-9.48611081e-01 6.40352890e-02 8.42287838e-01 6.87281132e-01
-5.70536451e-03 -4.77528200e-02 -3.15003663e-01 3.00916225e-01
4.50317264e-01 3.45624655e-01 4.62827235e-01 9.79211152e-01
1.96858004e-01 -3.77124161e-01 -6.04429543e-01 7.60231197e-01
1.91482991e-01 -8.52353632e-01 -1.60341978e-01 -1.41316816e-01
1.05729640e+00 1.17849305e-01 8.53649154e-02 3.16629224e-02
5.13427794e-01 -1.00712168e+00 5.47461689e-01 3.40687186e-01
4.22455221e-01 -8.42981160e-01 7.08729327e-01 3.09438407e-01
-1.04863024e+00 1.23345733e-01 -2.48220980e-01 1.67403817e-01
9.49148089e-02 5.65600157e-01 -4.43457961e-01 5.38251936e-01
7.02580690e-01 6.34889841e-01 -5.64149022e-01 8.22194517e-01
-5.77894270e-01 4.51317191e-01 -3.47273499e-01 5.31019807e-01
5.88281095e-01 -6.61270320e-01 3.87825519e-01 9.91931975e-01
-2.27269635e-01 1.33047447e-01 5.26286602e-01 9.83051360e-01
-1.35390148e-01 8.88665468e-02 -4.23463434e-01 -1.27712556e-03
1.73310027e-01 1.28126740e+00 -1.29171097e+00 -3.28104973e-01
-3.44794631e-01 1.30591488e+00 2.16454253e-01 4.80108857e-01
-7.90658891e-01 -3.74338537e-01 5.51130116e-01 -3.00388902e-01
4.37232554e-01 -3.46494406e-01 -6.58081055e-01 -8.68852794e-01
-7.95780420e-02 -6.19139016e-01 2.02435687e-01 -6.01592600e-01
-1.09579885e+00 4.37061787e-01 -1.41417772e-01 -9.32082176e-01
1.47989541e-02 -5.07352471e-01 -7.33545661e-01 6.97719038e-01
-1.60536182e+00 -1.38325489e+00 -4.97656077e-01 6.85691535e-01
7.88935602e-01 -2.09162682e-02 6.53434277e-01 8.02146420e-02
-6.08896196e-01 1.09455973e-01 2.36655965e-01 4.56050217e-01
4.25683141e-01 -1.44151616e+00 3.31534982e-01 1.17392969e+00
5.04483171e-02 1.80576861e-01 6.21005058e-01 -5.29429197e-01
-8.12208116e-01 -1.38787103e+00 2.79155433e-01 -2.58318279e-02
5.42006135e-01 -2.54679650e-01 -8.44434798e-01 6.70860648e-01
2.42348477e-01 -2.78871227e-02 4.78373498e-01 -3.35141987e-01
-1.35828316e-01 3.39909196e-02 -1.37938023e+00 6.41448975e-01
8.22964013e-01 -4.67237085e-01 -6.90101922e-01 4.22138900e-01
7.17571318e-01 -3.92971337e-01 -6.52226806e-01 5.55808365e-01
1.65230840e-01 -1.14128923e+00 1.08285546e+00 -2.95834392e-01
5.00222981e-01 -4.65696275e-01 -1.77794710e-01 -1.27072453e+00
7.24930614e-02 -4.32374448e-01 5.25465906e-01 1.08896136e+00
2.12105945e-01 -6.84504271e-01 1.15191853e+00 4.98317540e-01
-4.28818047e-01 -5.31824112e-01 -5.81264973e-01 -5.40458560e-01
1.46414831e-01 -4.78622645e-01 2.26693586e-01 8.75679314e-01
-5.32285094e-01 1.11436293e-01 -1.11300573e-02 4.74068463e-01
9.60530579e-01 1.33590534e-01 8.61734807e-01 -1.08951724e+00
-2.05705896e-01 -4.40786779e-01 -5.78512609e-01 -9.95695114e-01
2.90743828e-01 -6.25624895e-01 4.00653869e-01 -1.62420523e+00
8.37312639e-02 -4.40303057e-01 -1.66424170e-01 4.73497003e-01
-1.96643785e-01 5.78011751e-01 1.90745473e-01 2.49959841e-01
-7.19248891e-01 5.97140789e-01 1.23104346e+00 -2.97489047e-01
-2.74859369e-01 -3.94529849e-02 -5.44302046e-01 1.16525126e+00
9.13801014e-01 -3.31116885e-01 -4.60451782e-01 -3.26687157e-01
-3.24494362e-01 -2.97189981e-01 4.33981180e-01 -1.19877481e+00
1.96848184e-01 -1.30023301e-01 2.16939762e-01 -4.94437456e-01
3.14458609e-01 -8.52202535e-01 4.80644684e-03 3.08140308e-01
-1.92033201e-01 -4.51039940e-01 2.24527150e-01 5.57294369e-01
-3.68059158e-01 -2.89926261e-01 1.23333919e+00 -3.71628314e-01
-1.11471474e+00 2.41486743e-01 -2.10886016e-01 2.10334420e-01
1.30019009e+00 -2.30676264e-01 -6.73563257e-02 -1.57046661e-01
-8.71375680e-01 1.94502801e-01 7.23307371e-01 2.98763931e-01
5.23889899e-01 -8.15746605e-01 -4.26764697e-01 3.09913248e-01
5.86146861e-02 5.95754147e-01 2.76935279e-01 5.70426106e-01
-8.58543992e-01 1.47623330e-01 -2.25284502e-01 -7.00775564e-01
-1.00908673e+00 5.26123345e-01 3.60418975e-01 -1.72544673e-01
-5.29285491e-01 8.91907394e-01 5.88179588e-01 -6.51426673e-01
1.25412196e-01 -6.55183718e-02 -2.87960112e-01 -2.71904886e-01
-5.29674208e-03 9.08985063e-02 -1.33624062e-01 -8.04635167e-01
-1.59944773e-01 7.93068111e-01 7.49252588e-02 -5.68909757e-02
1.15910268e+00 -2.80434310e-01 -2.29638100e-01 2.39106208e-01
1.19198751e+00 -1.69231266e-01 -1.43743360e+00 -1.32139802e-01
-1.20034397e-01 -2.31028855e-01 9.16855708e-02 -6.13554120e-01
-1.29913521e+00 7.82436609e-01 4.86652642e-01 3.43160838e-01
1.25762153e+00 2.01052092e-02 9.90975857e-01 1.61554307e-01
3.52274746e-01 -1.21642017e+00 7.01196939e-02 5.07911384e-01
4.50200260e-01 -1.51256680e+00 -9.42544192e-02 -7.98659086e-01
-7.62429833e-01 7.49092460e-01 4.37950343e-01 -4.29434419e-01
6.24827147e-01 -7.57552907e-02 3.79013717e-01 -1.20643474e-01
2.37754956e-01 -5.42393684e-01 2.82384902e-01 5.68462670e-01
3.79135609e-02 1.35786578e-01 -1.46713540e-01 3.39700639e-01
-1.93002656e-01 -3.80821019e-01 2.09751174e-01 8.46962154e-01
-5.87814033e-01 -9.98118043e-01 -4.02391285e-01 1.19010225e-01
-5.53237498e-01 6.11476265e-02 -4.63592798e-01 7.14491069e-01
3.36948931e-01 1.07493818e+00 4.63981815e-02 -2.37538114e-01
1.18719846e-01 -1.01100206e-01 2.56841362e-01 -6.29519403e-01
-2.91514516e-01 9.19883624e-02 -2.68231720e-01 -5.48572123e-01
-7.40185738e-01 -5.47281682e-01 -1.58860791e+00 1.56567842e-01
-1.32330563e-02 5.92781380e-02 8.93224001e-01 1.29644310e+00
8.01473483e-03 6.86755002e-01 6.56544924e-01 -1.03837252e+00
-8.91512930e-02 -8.61205697e-01 -5.36062717e-01 7.26721287e-01
4.94181588e-02 -6.23996615e-01 -3.16736698e-01 3.55212867e-01] | [9.78630542755127, 0.75559002161026] |
463bfdf3-5191-4c49-a44f-a68243b0d065 | unidirectional-video-denoising-by-mimicking | 2204.05532 | null | https://arxiv.org/abs/2204.05532v3 | https://arxiv.org/pdf/2204.05532v3.pdf | Unidirectional Video Denoising by Mimicking Backward Recurrent Modules with Look-ahead Forward Ones | While significant progress has been made in deep video denoising, it remains very challenging for exploiting historical and future frames. Bidirectional recurrent networks (BiRNN) have exhibited appealing performance in several video restoration tasks. However, BiRNN is intrinsically offline because it uses backward recurrent modules to propagate from the last to current frames, which causes high latency and large memory consumption. To address the offline issue of BiRNN, we present a novel recurrent network consisting of forward and look-ahead recurrent modules for unidirectional video denoising. Particularly, look-ahead module is an elaborate forward module for leveraging information from near-future frames. When denoising the current frame, the hidden features by forward and look-ahead recurrent modules are combined, thereby making it feasible to exploit both historical and near-future frames. Due to the scene motion between non-neighboring frames, border pixels missing may occur when warping look-ahead feature from near-future frame to current frame, which can be largely alleviated by incorporating forward warping and proposed border enlargement. Experiments show that our method achieves state-of-the-art performance with constant latency and memory consumption. Code is avaliable at https://github.com/nagejacob/FloRNN. | ['Zhenxing Niu', 'WangMeng Zuo', 'Xiaohe Wu', 'Junyi Li'] | 2022-04-12 | null | null | null | null | ['video-denoising', 'video-restoration'] | ['computer-vision', 'computer-vision'] | [ 7.20751062e-02 -2.71726042e-01 -1.00657612e-01 -1.63237706e-01
-5.19244313e-01 -2.19702780e-01 3.06268930e-01 -2.99977362e-01
-3.77414882e-01 5.79793036e-01 5.65750420e-01 -2.43568420e-01
1.68349266e-01 -7.25016415e-01 -7.43874609e-01 -8.99494171e-01
-1.95524707e-01 -6.39788508e-01 4.59110945e-01 -2.86219090e-01
2.74769422e-02 3.06833386e-01 -1.21618044e+00 4.60382551e-01
5.34143507e-01 8.79499435e-01 5.07133782e-01 6.79991663e-01
5.99244051e-02 9.78525221e-01 -2.64437914e-01 -2.22548798e-01
3.39068234e-01 -5.75388432e-01 -4.91431534e-01 -4.14228216e-02
1.75998390e-01 -8.63398254e-01 -9.23843801e-01 1.00889850e+00
3.94531935e-01 2.72277206e-01 -1.50622919e-01 -1.06582868e+00
-5.69364130e-01 5.60301721e-01 -6.87587440e-01 4.17294592e-01
2.63742179e-01 2.26013273e-01 6.57686293e-01 -8.78089309e-01
7.12169886e-01 1.10791659e+00 7.23007679e-01 5.47325730e-01
-8.99080455e-01 -7.65684247e-01 4.34740245e-01 5.74375331e-01
-1.05375528e+00 -5.98461509e-01 8.93962026e-01 -9.07047465e-02
8.25590909e-01 2.26925299e-01 7.46661425e-01 1.11657691e+00
5.16214430e-01 8.49531353e-01 6.87546968e-01 -9.03099701e-02
-3.69328521e-02 -5.95937669e-01 -2.87559256e-02 5.41554332e-01
-3.17779958e-01 1.69780612e-01 -6.90340817e-01 1.14828400e-01
9.87763166e-01 5.15405953e-01 -6.35924816e-01 7.07259700e-02
-1.24420762e+00 5.77573776e-01 6.40769064e-01 2.27767274e-01
-6.65474176e-01 3.15071911e-01 6.19092762e-01 4.56179053e-01
4.65885848e-01 -2.80391872e-01 -3.28618824e-01 -2.66385466e-01
-1.18405354e+00 -1.31465524e-01 3.53703499e-01 7.46411204e-01
8.38426948e-01 3.16342384e-01 -9.39339548e-02 7.28346765e-01
2.13101745e-01 1.80106476e-01 5.19410431e-01 -1.33802974e+00
5.19276679e-01 1.37647510e-01 -5.71635328e-02 -1.26900363e+00
-2.04694927e-01 -3.27582359e-01 -1.34421802e+00 1.22287244e-01
4.12620641e-02 -2.17337757e-01 -1.02947283e+00 1.61000550e+00
2.35641912e-01 5.49546778e-01 -2.28276174e-03 1.09515214e+00
8.05576861e-01 1.14449096e+00 -8.51397067e-02 -6.29488289e-01
1.08743143e+00 -1.15275586e+00 -9.16905999e-01 -1.91519454e-01
4.20286238e-01 -9.54676449e-01 5.38274944e-01 2.47142360e-01
-1.21438670e+00 -6.25623167e-01 -1.00686717e+00 -3.04651916e-01
2.05545977e-01 -5.97963855e-02 3.33910555e-01 -4.76051830e-02
-1.19111013e+00 7.13952601e-01 -1.17284620e+00 -1.10772900e-01
1.61262900e-01 1.71482146e-01 -4.77136075e-01 -2.83285916e-01
-1.16430068e+00 2.88394928e-01 5.57134449e-02 7.36389995e-01
-9.15793359e-01 -5.70013642e-01 -9.20144856e-01 -1.37500865e-02
4.73307014e-01 -6.80955470e-01 1.02388287e+00 -9.66270924e-01
-1.44832408e+00 -2.12714989e-02 -5.58146000e-01 -5.64341366e-01
6.63664043e-01 -4.59929436e-01 -5.26662529e-01 1.92787603e-01
8.32627863e-02 5.80322146e-01 1.08201849e+00 -9.36175704e-01
-5.28733552e-01 -2.32568249e-01 4.14565764e-02 1.35931939e-01
-4.57989305e-01 -3.29149105e-02 -7.59682178e-01 -1.16243696e+00
5.54627180e-01 -8.84461582e-01 -2.37215966e-01 1.46549150e-01
-2.84859776e-01 1.96480975e-01 1.34039330e+00 -1.04191375e+00
1.40583384e+00 -2.45309591e+00 9.67230350e-02 -2.30491132e-01
1.91205069e-01 4.59398836e-01 -2.73762852e-01 6.41310811e-01
-2.13073164e-01 -1.51424576e-02 -1.26533419e-01 -3.89015824e-01
-5.81232607e-01 3.27310622e-01 -6.63468480e-01 4.17191833e-01
-7.16843316e-03 6.14779174e-01 -8.58901322e-01 -1.70150012e-01
3.79799515e-01 9.41382468e-01 -5.23562908e-01 1.68937728e-01
1.77170724e-01 5.64012349e-01 -2.84528464e-01 5.66197932e-01
9.21399295e-01 2.30683526e-03 8.77073333e-02 -4.92026389e-01
-3.67678314e-01 2.82488048e-01 -1.00920808e+00 1.78261638e+00
-4.31444079e-01 8.76692891e-01 2.75038451e-01 -7.99785614e-01
7.16261089e-01 4.40481812e-01 3.54290634e-01 -8.74063909e-01
-7.63115957e-02 7.76404291e-02 -2.78102577e-01 -4.50963110e-01
6.75300360e-01 8.15922841e-02 3.30436379e-01 1.71126798e-01
-1.42232776e-01 5.65822482e-01 2.66642988e-01 1.67161912e-01
1.23514175e+00 4.05711353e-01 -6.74467012e-02 1.37289166e-01
6.78105235e-01 -2.76429832e-01 1.26027310e+00 5.39841950e-01
-3.33585560e-01 8.90840948e-01 3.17341089e-01 -6.68242633e-01
-8.35932612e-01 -9.83330131e-01 3.75706106e-01 9.13212299e-01
3.55430603e-01 -7.29967475e-01 -4.13479596e-01 -3.74149024e-01
-4.28834051e-01 1.89533204e-01 -4.40943182e-01 -1.57461911e-01
-1.10426795e+00 -4.25912112e-01 1.54697746e-01 5.56559622e-01
8.81926179e-01 -9.28989768e-01 -6.53147042e-01 5.53539515e-01
-5.08744717e-01 -1.04587162e+00 -7.59204090e-01 -1.86158985e-01
-1.25704575e+00 -7.43891060e-01 -1.06717432e+00 -6.48130178e-01
4.81887817e-01 8.44449937e-01 7.95833468e-01 2.81519979e-01
-2.38353778e-02 -1.27450079e-01 -4.96996224e-01 4.21099275e-01
-1.24129727e-01 -3.34950984e-01 -2.63595581e-01 1.26825958e-01
-1.53284058e-01 -8.23589563e-01 -1.12223470e+00 3.90067369e-01
-1.04614031e+00 4.47769672e-01 3.39419037e-01 1.11515093e+00
6.66011035e-01 2.83609666e-02 2.56590694e-01 -5.73142171e-01
1.94595948e-01 -4.26297307e-01 -4.08172101e-01 1.24911424e-02
-1.40466854e-01 -2.90685833e-01 7.85257757e-01 -2.70345390e-01
-1.09606886e+00 -8.95730332e-02 -3.71821254e-01 -7.41067111e-01
2.35650137e-01 4.32039142e-01 -5.34817167e-02 2.05479711e-01
7.95861408e-02 3.93391520e-01 6.03175461e-02 -6.18317664e-01
8.53008404e-02 4.63883400e-01 6.72394395e-01 -1.58049334e-02
5.05828857e-01 7.00401366e-01 -1.59861982e-01 -7.45582342e-01
-6.77504897e-01 -2.74995476e-01 -2.41544768e-01 -3.54986995e-01
8.09663355e-01 -1.23363638e+00 -6.86506331e-01 7.05816865e-01
-1.27730417e+00 -2.66317070e-01 -5.92774041e-02 4.78487700e-01
-2.00311512e-01 7.38883674e-01 -1.16544485e+00 -4.41924483e-01
-6.01684332e-01 -1.43647254e+00 7.58767903e-01 3.53096008e-01
-4.32784259e-02 -8.23073328e-01 -2.59867400e-01 1.79086968e-01
4.21953142e-01 8.21238533e-02 4.72626835e-01 1.37421861e-01
-8.20512235e-01 -5.81420213e-02 -2.16507703e-01 4.49175119e-01
1.61661729e-01 9.46056917e-02 -6.32304907e-01 -5.79079807e-01
2.43798509e-01 1.79601118e-01 1.23008418e+00 5.14014840e-01
1.01284361e+00 -4.03337449e-01 -1.29050925e-01 9.67188716e-01
1.21473002e+00 3.03713083e-01 8.28876436e-01 6.37704432e-01
8.14298332e-01 2.53341675e-01 6.61516845e-01 4.71265882e-01
5.16071320e-01 5.53435683e-01 5.34687400e-01 -1.94683999e-01
-4.15428221e-01 -1.43255025e-01 7.46125162e-01 1.12540579e+00
-1.26759380e-01 -2.09948048e-01 -6.85752451e-01 6.67358518e-01
-2.12365818e+00 -1.12517989e+00 -1.45895869e-01 2.04678965e+00
5.49368680e-01 8.73457082e-03 -2.51094520e-01 1.02670476e-01
8.28804731e-01 6.37540221e-01 -5.32403767e-01 -2.27699533e-01
-2.24712536e-01 -1.94498405e-01 3.67226869e-01 4.58609045e-01
-1.01424694e+00 6.66709781e-01 5.44572973e+00 7.37701833e-01
-1.31772399e+00 1.98132887e-01 7.01757848e-01 -2.05204889e-01
-2.35669285e-01 2.47345224e-01 -4.87000197e-01 5.88835895e-01
7.40547597e-01 2.01013982e-01 3.52998078e-01 3.98140371e-01
7.02084839e-01 -2.71715462e-01 -6.48846388e-01 9.81288612e-01
-1.79093659e-01 -1.52056503e+00 1.40407309e-01 -4.42276075e-02
7.02660799e-01 3.29220928e-02 -5.77955246e-02 1.74087865e-04
-6.26138523e-02 -4.77760255e-01 6.91071630e-01 5.21928966e-01
4.67410684e-01 -1.02381682e+00 7.43906081e-01 3.30217481e-01
-1.39782071e+00 -2.63375193e-01 -2.44641647e-01 -1.35602772e-01
5.63874781e-01 9.53351378e-01 -8.94993842e-02 6.57000899e-01
1.13579488e+00 1.22698009e+00 -1.55675888e-01 7.80589700e-01
-2.45093644e-01 5.62541366e-01 -1.13274977e-01 7.88273513e-01
3.96046311e-01 -4.41133499e-01 6.92707896e-01 1.06341469e+00
7.14214921e-01 7.66990408e-02 8.48330706e-02 3.06502074e-01
-1.45643398e-01 -4.18021977e-01 -3.81883085e-01 2.70853162e-01
2.80812860e-01 1.08950651e+00 -5.38855195e-01 -3.79361898e-01
-5.36159337e-01 1.25061381e+00 -8.49269181e-02 6.32358730e-01
-8.62429202e-01 -2.92830020e-01 8.57586920e-01 6.63489401e-02
4.40558970e-01 -4.27550644e-01 -1.37488753e-01 -1.35386407e+00
3.08919162e-01 -7.02512562e-01 4.11873817e-01 -8.86710525e-01
-9.18379962e-01 8.41995299e-01 -4.67610210e-01 -1.53085434e+00
-2.71569490e-01 6.45570233e-02 -6.48404002e-01 7.04557002e-01
-1.62331390e+00 -1.02094734e+00 -4.71974313e-01 7.61607230e-01
8.73158872e-01 1.42507270e-01 4.83841002e-01 4.69401687e-01
-8.13030064e-01 4.04079944e-01 1.80202693e-01 2.13745430e-01
8.61540556e-01 -5.37290037e-01 5.02286494e-01 1.45962226e+00
-1.38056129e-01 5.81518829e-01 7.15736568e-01 -7.40300238e-01
-1.53949654e+00 -1.18780410e+00 7.73331165e-01 5.47853291e-01
5.93375325e-01 -1.03351377e-01 -1.13865578e+00 6.91999733e-01
3.46558183e-01 3.45093012e-01 2.97285408e-01 -3.24333191e-01
-3.61104369e-01 -3.75265062e-01 -6.91808045e-01 8.28278959e-01
1.02555239e+00 -6.06261194e-01 -1.81997120e-01 8.42008889e-02
6.73955441e-01 -4.88246262e-01 -6.99565411e-01 5.86249888e-01
5.11520028e-01 -1.46490073e+00 9.59343433e-01 2.54131675e-01
6.65364444e-01 -5.97457290e-01 -2.56354772e-02 -1.06812024e+00
-2.37827897e-01 -1.07350659e+00 -4.03779536e-01 1.20505786e+00
-3.47477458e-02 -5.46909750e-01 7.69617140e-01 3.62948596e-01
-2.57679760e-01 -7.41728604e-01 -1.04932046e+00 -6.22593164e-01
-3.42931092e-01 -4.96375829e-01 3.11124682e-01 7.48666465e-01
-3.04078370e-01 2.09203959e-02 -9.51932251e-01 1.89498633e-01
4.83231008e-01 2.92325705e-01 5.39000988e-01 -5.41518927e-01
-2.18265414e-01 -2.27463059e-02 -3.70023429e-01 -1.66181481e+00
-8.72349516e-02 -4.64585662e-01 8.02688301e-02 -1.55278218e+00
2.07204819e-01 -8.87956172e-02 -2.81030864e-01 5.50901651e-01
-8.92216936e-02 5.65418720e-01 5.02370000e-01 4.12807107e-01
-3.08321238e-01 7.49490976e-01 1.43737257e+00 -5.22430874e-02
-2.39317060e-01 -1.89472809e-01 -2.82581568e-01 7.94668853e-01
6.91235423e-01 -3.74613374e-01 -3.33133817e-01 -9.53997672e-01
-7.84303877e-04 6.99712694e-01 4.88136023e-01 -9.94027197e-01
6.18409693e-01 1.08791441e-01 4.85700756e-01 -8.00203323e-01
6.96510315e-01 -8.00689459e-01 4.69834745e-01 5.02378345e-01
-1.34960953e-02 4.46693361e-01 9.32352543e-02 7.04262257e-01
-7.02301502e-01 1.35981709e-01 5.97696006e-01 2.97307651e-02
-9.91802335e-01 3.76247108e-01 -5.83370924e-01 -3.76218945e-01
9.09563363e-01 -4.12770152e-01 -3.04024816e-01 -5.94480515e-01
-9.07575071e-01 1.70246214e-01 4.76838619e-01 4.58231598e-01
8.75348628e-01 -1.15947640e+00 -5.71895838e-01 3.38426650e-01
-4.33065921e-01 -7.34833330e-02 8.78281474e-01 1.07867432e+00
-6.02379143e-01 1.18820893e-03 -1.79693416e-01 -6.17628455e-01
-1.46076596e+00 4.19290125e-01 8.58617723e-02 -1.62437126e-01
-1.09488285e+00 7.99790084e-01 3.05609107e-01 2.73175776e-01
1.02180742e-01 -5.94084598e-02 6.99716881e-02 1.31582841e-01
7.74858534e-01 5.37053406e-01 2.45481730e-03 -7.00118423e-01
-1.82272211e-01 4.85668480e-01 -2.61785150e-01 5.46690188e-02
1.48585367e+00 -5.23540735e-01 -3.54351342e-01 2.43484929e-01
1.32084894e+00 -2.04324022e-01 -1.67649329e+00 -1.64504543e-01
-3.98790002e-01 -6.52153432e-01 3.81828964e-01 -3.12540412e-01
-1.70670915e+00 8.32390666e-01 5.55890560e-01 -1.93592340e-01
1.69350278e+00 -5.80484211e-01 1.44995809e+00 3.89266163e-02
2.87538320e-01 -9.28054094e-01 -7.61196688e-02 6.93796456e-01
8.48028123e-01 -1.07457888e+00 7.93255717e-02 -5.28475702e-01
-3.39269072e-01 1.33017087e+00 3.76181394e-01 -1.62933901e-01
8.20012748e-01 3.00214767e-01 1.82275653e-01 1.90858245e-01
-8.58421087e-01 1.49105176e-01 -7.69451587e-03 1.69880688e-01
4.36103761e-01 -2.19827950e-01 -3.42250437e-01 1.28208101e-01
1.79833844e-01 9.63304490e-02 7.24715352e-01 1.19976282e+00
-1.19465888e-01 -1.17646062e+00 -3.33856463e-01 1.94341034e-01
-5.46947837e-01 -2.62316018e-01 2.88997650e-01 5.17303467e-01
-9.94470790e-02 1.17463410e+00 1.31900206e-01 -4.32687134e-01
6.61953613e-02 -4.00968909e-01 -3.00306715e-02 -1.08915173e-01
-6.02351308e-01 6.13647878e-01 5.50013036e-02 -1.03881955e+00
-4.96873885e-01 -5.76638162e-01 -1.22982168e+00 -5.99707842e-01
-8.99519995e-02 -5.94415031e-02 4.49176848e-01 7.29588807e-01
6.39281750e-01 6.51091158e-01 7.03674912e-01 -1.12879944e+00
-9.39450711e-02 -7.11640000e-01 -2.23847970e-01 7.82571211e-02
7.70797193e-01 -2.43078306e-01 -3.68600965e-01 2.42070079e-01] | [11.219730377197266, -2.0326130390167236] |
76f42513-dad0-47ee-86d2-180f9706646d | exploiting-temporal-attention-features-for | 2008.02344 | null | https://arxiv.org/abs/2008.02344v2 | https://arxiv.org/pdf/2008.02344v2.pdf | Exploiting Temporal Attention Features for Effective Denoising in Videos | Video Denoising is one of the fundamental tasks of any videoprocessing pipeline. It is different from image denoising due to the tem-poral aspects of video frames, and any image denoising approach appliedto videos will result in flickering. The proposed method makes use oftemporal as well as spatial dimensions of video frames as part of a two-stage pipeline. Each stage in the architecture named as Spatio-TemporalNetwork uses a channel-wise attention mechanism to forward the encodersignal to the decoder side. The Attention Block used in this paper usessoft attention to ranks the filters for better training. | ['Aryansh Omray', 'Utsav Krishnan', 'Samyak Jain', 'Pratik Chattopadhyay'] | 2020-08-05 | null | null | null | null | ['video-denoising'] | ['computer-vision'] | [ 6.03197962e-02 -2.73090720e-01 6.89930841e-02 -4.07182634e-01
-3.54856730e-01 -2.85174519e-01 4.98937130e-01 -2.84355044e-01
-6.81451499e-01 4.00610656e-01 3.44836056e-01 -2.38343596e-01
3.04262906e-01 -4.48147357e-01 -8.62144470e-01 -8.44766200e-01
1.26862582e-02 -4.98365849e-01 6.01255894e-01 -1.54383585e-01
3.12416613e-01 4.74881738e-01 -1.37812912e+00 7.57082939e-01
4.80303288e-01 1.18465698e+00 3.74081880e-01 1.17207503e+00
-8.46860409e-02 1.37160182e+00 -5.95500708e-01 -3.20651412e-01
1.27967060e-01 -7.19736636e-01 -4.63450074e-01 1.51006296e-01
4.03349280e-01 -5.66241741e-01 -6.91105008e-01 1.33942342e+00
1.45534948e-01 3.16457242e-01 3.38689297e-01 -8.71109128e-01
-5.37212133e-01 4.39695269e-01 -5.08539557e-01 9.16760623e-01
1.01849690e-01 -8.95770192e-02 4.09156263e-01 -9.96975660e-01
4.84708607e-01 1.37681901e+00 6.49830401e-01 4.52047586e-01
-9.37047005e-01 -5.52555442e-01 4.35350955e-01 7.36353397e-01
-9.04234946e-01 -8.07517707e-01 7.07324266e-01 -3.10437620e-01
1.05800545e+00 -2.54078978e-03 5.47747552e-01 9.18286264e-01
7.33727336e-01 6.70352995e-01 7.80051887e-01 -1.56582654e-01
1.60665140e-01 -4.81841803e-01 2.16528237e-01 8.03725272e-02
2.12125815e-02 1.73871368e-01 -5.21875143e-01 4.74169225e-01
9.08585608e-01 1.41883790e-01 -2.22120076e-01 1.93007395e-01
-6.16100132e-01 5.29903889e-01 3.89299333e-01 4.59297657e-01
-6.74234092e-01 6.90912902e-01 5.99579990e-01 6.40291989e-01
3.69605154e-01 -9.15990248e-02 -2.90780872e-01 -2.74680763e-01
-1.16259480e+00 -1.12828419e-01 4.49579328e-01 7.76320279e-01
4.91733104e-01 5.32012284e-01 -9.10673887e-02 4.21674639e-01
3.62587363e-01 1.58598244e-01 5.84274828e-01 -1.23158813e+00
3.18048567e-01 -3.04614734e-02 -1.00030392e-01 -9.35057640e-01
-6.75163567e-02 -2.97949500e-02 -8.87647331e-01 4.79223937e-01
2.60057002e-01 -2.98272222e-01 -1.39419007e+00 1.26263630e+00
-1.91146391e-03 4.49640334e-01 -2.20510572e-01 1.13437891e+00
7.67625809e-01 8.96547794e-01 2.79919475e-01 -3.06589812e-01
1.04503059e+00 -9.86745477e-01 -1.30257821e+00 -1.71202227e-01
5.61200567e-02 -8.73945832e-01 3.63434553e-01 6.70599818e-01
-1.37937987e+00 -8.50975633e-01 -1.02637005e+00 -4.48823988e-01
-1.92030340e-01 -3.23760092e-01 3.27237487e-01 1.84226766e-01
-1.27815318e+00 6.12075448e-01 -1.05049169e+00 -2.35085979e-01
4.45821166e-01 2.62681603e-01 -3.65261406e-01 2.13180929e-02
-1.07218349e+00 8.37668419e-01 3.54230404e-01 5.42084515e-01
-1.03843129e+00 -2.69747823e-01 -8.75939608e-01 1.52454600e-01
3.54342490e-01 -2.53346950e-01 1.42433834e+00 -1.56402481e+00
-1.41335368e+00 2.66278952e-01 -4.35613543e-01 -5.47138333e-01
5.01986086e-01 -5.83157659e-01 -4.20702755e-01 4.54124779e-01
-2.31139049e-01 5.70414841e-01 1.51921737e+00 -9.14986372e-01
-7.86703587e-01 -7.87916929e-02 1.42121032e-01 1.17763303e-01
9.55635589e-03 2.40652546e-01 -8.44639003e-01 -1.11010337e+00
2.35500485e-01 -4.06473249e-01 -3.18789840e-01 -1.00131996e-01
-1.35186866e-01 1.28767863e-01 1.28288388e+00 -1.09713542e+00
1.33212161e+00 -2.39272141e+00 3.31435025e-01 5.02771176e-02
1.33459747e-01 3.21366251e-01 -5.83075583e-02 3.24435383e-01
-5.04423141e-01 1.86816584e-02 1.96768522e-01 -1.96993336e-01
-4.67727780e-01 1.44540921e-01 -1.00291945e-01 6.25960112e-01
3.69495362e-01 5.86686373e-01 -9.06752944e-01 -2.78067082e-01
5.01394153e-01 7.05355883e-01 -6.49938881e-01 2.05342531e-01
4.30090316e-02 4.69995618e-01 -1.94724619e-01 3.44275534e-01
5.97635150e-01 2.71395743e-01 1.43911997e-02 -4.31597292e-01
-3.81995708e-01 2.63636678e-01 -9.42432463e-01 1.65279019e+00
-2.05625877e-01 1.05840445e+00 5.33047616e-01 -9.60370064e-01
2.89218158e-01 6.31277442e-01 3.91471595e-01 -6.70475304e-01
5.84183335e-01 -1.19108796e-01 2.01509386e-01 -9.12198961e-01
4.98295754e-01 -1.46691680e-01 4.20083702e-01 5.15135610e-03
4.25566405e-01 2.36577690e-01 5.58480084e-01 7.28102401e-02
1.05978286e+00 3.09053242e-01 -8.90237000e-03 -2.68100113e-01
6.30253017e-01 -2.52556115e-01 7.07978189e-01 7.30430007e-01
-4.02318805e-01 6.95673168e-01 6.71033204e-01 -5.75750053e-01
-1.25870371e+00 -4.86252844e-01 1.01942420e-01 9.81222212e-01
8.27888027e-02 -3.79988700e-01 -8.34485233e-01 -3.72117192e-01
-3.35771888e-01 4.12329614e-01 -6.97384834e-01 -2.30138630e-01
-7.37211823e-01 -3.22258711e-01 1.77697256e-01 6.83039963e-01
5.94190836e-01 -1.00934196e+00 -6.61028445e-01 3.07325512e-01
-6.72555491e-02 -1.05372202e+00 -6.15026891e-01 5.09974778e-01
-9.43934321e-01 -1.19497871e+00 -5.98030686e-01 -8.92411113e-01
5.29675126e-01 3.90404254e-01 8.41822803e-01 1.52300298e-01
9.24549550e-02 -5.10078706e-02 -4.77574080e-01 -2.93104976e-01
-1.95161030e-01 -2.56540030e-01 -3.95909935e-01 2.46014580e-01
6.33217096e-01 -3.85824889e-01 -5.87277830e-01 -1.60958692e-01
-1.13873339e+00 -1.08925328e-01 3.46623659e-01 8.19932103e-01
4.24559772e-01 4.92735028e-01 -1.65640712e-02 -7.01244950e-01
6.71681643e-01 -3.24395537e-01 -7.61879802e-01 -1.26243606e-01
2.02315167e-01 -8.83451030e-02 7.78194249e-01 -2.84504920e-01
-9.49445307e-01 2.06973627e-01 -3.25797141e-01 -6.67513788e-01
-1.82413444e-01 3.94311458e-01 -3.96758020e-01 3.83532718e-02
2.54394799e-01 -2.56770756e-02 -2.24331226e-02 -5.57639360e-01
1.34507969e-01 4.58088219e-01 4.75730896e-01 2.66353875e-01
5.87981045e-01 4.21688497e-01 -1.57190472e-01 -1.20091307e+00
-5.62712491e-01 -3.64735484e-01 -7.37211406e-01 -4.36549455e-01
1.32306218e+00 -9.59670246e-01 -6.46984816e-01 8.97404253e-01
-1.45782733e+00 -6.32445961e-02 -6.05810285e-02 6.00531936e-01
-1.73935756e-01 2.47323424e-01 -9.82276738e-01 -8.64017725e-01
-1.22722007e-01 -1.29881382e+00 4.98008430e-01 3.28390568e-01
7.82531425e-02 -8.03245902e-01 -2.86348403e-01 -2.48690173e-01
3.33477736e-01 -1.50231108e-01 5.54476380e-01 -2.25021079e-01
-6.14511192e-01 -3.17139447e-01 -1.10859647e-01 8.64470363e-01
-1.06023759e-01 2.33682111e-01 -1.10831356e+00 -1.15701482e-01
6.37556434e-01 1.67059954e-02 1.31738997e+00 1.02529800e+00
1.43739188e+00 -1.48353383e-01 1.61180526e-01 7.94193864e-01
1.47440612e+00 6.21531129e-01 1.14165699e+00 3.32779884e-01
6.35410964e-01 4.26008970e-01 3.30995828e-01 1.02711320e-01
9.61923972e-02 2.77951807e-01 6.54474616e-01 -2.60347009e-01
-2.46735245e-01 1.92512915e-01 6.10996544e-01 7.46076286e-01
-2.53086895e-01 -3.77016097e-01 -6.96706414e-01 3.99490237e-01
-1.65583670e+00 -1.46942914e+00 -4.85175759e-01 1.99421048e+00
4.01538253e-01 1.30214363e-01 -5.49339391e-02 3.09402227e-01
7.69436359e-01 2.90705204e-01 -3.79677266e-01 -7.55721807e-01
-7.05224723e-02 5.11730462e-02 6.57593548e-01 7.37286627e-01
-1.36189258e+00 7.15300202e-01 7.10729313e+00 2.87560701e-01
-1.16401470e+00 1.58582062e-01 7.11147249e-01 -1.59567907e-01
2.19940245e-01 -2.33957946e-01 -4.00735259e-01 7.11748004e-01
1.02780855e+00 2.02317536e-01 6.35946572e-01 3.93633336e-01
7.30828941e-01 -4.41371888e-01 -9.00624633e-01 1.04501343e+00
1.58486590e-01 -1.00814307e+00 -9.18306783e-02 -4.00849283e-01
4.17073518e-01 2.98668388e-02 -6.96731508e-02 -7.71394670e-02
1.20269887e-01 -9.76622939e-01 8.95302415e-01 5.86155236e-01
4.47127938e-01 -9.64542329e-01 9.94140267e-01 -3.09409499e-02
-1.14353740e+00 -2.90468752e-01 -3.34816664e-01 -1.28995955e-01
5.56373715e-01 4.75203693e-01 -1.79931268e-01 1.78366035e-01
1.28983617e+00 1.12342119e+00 -1.92087740e-01 1.37400365e+00
-4.60834533e-01 6.53624773e-01 3.24483551e-02 4.91437048e-01
3.78676027e-01 -2.83922315e-01 3.95594001e-01 1.25235951e+00
3.16277146e-01 2.81453907e-01 -9.72461700e-02 8.60389248e-02
-7.43675157e-02 -3.53901774e-01 -4.93716210e-01 -3.13014507e-01
6.93925321e-02 9.57953036e-01 -7.48144448e-01 -5.70243061e-01
-7.89173424e-01 1.15512061e+00 -1.92193627e-01 6.93216503e-01
-8.82726073e-01 -6.13756537e-01 6.65406704e-01 1.60030231e-01
6.80740237e-01 -2.93644637e-01 -3.15695465e-01 -1.07942140e+00
-6.89653531e-02 -9.11841393e-01 1.89893737e-01 -8.86725485e-01
-8.72487545e-01 7.73184180e-01 -3.40433657e-01 -1.01391673e+00
-8.52445215e-02 -6.40593052e-01 -6.48353934e-01 9.76821780e-01
-1.64898479e+00 -7.02445984e-01 -4.19150651e-01 7.24046588e-01
9.80228007e-01 1.00880057e-01 3.82962584e-01 5.66974103e-01
-6.96811736e-01 6.89724684e-02 1.13684885e-01 2.95183212e-01
6.40769362e-01 -1.14453828e+00 4.11245197e-01 1.67484593e+00
-3.77350301e-01 4.10775959e-01 9.98594820e-01 -7.54884601e-01
-1.30504632e+00 -9.87903237e-01 9.77072835e-01 6.83520213e-02
6.97451055e-01 -3.66054952e-01 -1.03955662e+00 7.10679531e-01
8.78403664e-01 8.58384091e-03 2.48849705e-01 -2.66519547e-01
-1.50537267e-02 -1.79537639e-01 -8.14126670e-01 4.01206583e-01
6.35703087e-01 -5.71832478e-01 -5.97069621e-01 1.50161339e-02
4.57417220e-01 -4.64972764e-01 -4.87192094e-01 -9.68973413e-02
4.18603778e-01 -1.04752791e+00 7.13622332e-01 -5.58928549e-01
6.22713745e-01 -5.36913335e-01 -2.42724102e-02 -1.17336488e+00
-6.14376366e-01 -5.94618261e-01 -3.18905830e-01 8.83582950e-01
5.05174771e-02 -1.66726351e-01 7.31887102e-01 3.98386717e-01
-4.57233071e-01 -1.42179117e-01 -7.45766640e-01 -4.71304923e-01
-9.77155194e-02 -5.94625533e-01 1.19382679e-01 5.46198785e-01
-2.86331058e-01 2.73330808e-01 -6.27179980e-01 2.86189884e-01
4.45654154e-01 -5.93248248e-01 2.53916532e-01 -8.29985380e-01
-1.36556506e-01 -2.61550099e-01 -5.97038746e-01 -1.19551313e+00
-2.69416690e-01 -3.33604008e-01 5.07126570e-01 -1.54837108e+00
-7.79140517e-02 4.84106630e-01 -3.92152667e-01 2.76188016e-01
-7.42847696e-02 3.33773017e-01 3.24395835e-01 -8.09130818e-02
-6.33964062e-01 3.86286169e-01 1.06856632e+00 -7.72442156e-03
-1.28684808e-02 -2.77831644e-01 -2.66693771e-01 7.21583545e-01
6.48587286e-01 -3.69917840e-01 -4.33313012e-01 -1.12902117e+00
-1.19439729e-01 4.20161709e-02 4.13405150e-01 -1.06402445e+00
4.67339456e-01 -2.45597214e-02 7.90784955e-01 -8.71475399e-01
4.54500288e-01 -1.29430485e+00 2.63132513e-01 6.31988525e-01
-7.32941926e-02 5.33633769e-01 6.22689389e-02 4.99344140e-01
-7.00879037e-01 -2.27617621e-01 9.66971874e-01 -2.68496871e-01
-1.07476568e+00 6.05572984e-02 -9.95988965e-01 -2.02753067e-01
1.07458901e+00 -5.57677865e-01 -8.13686475e-02 -5.37221909e-01
-7.12096035e-01 7.01697618e-02 3.27915221e-01 3.63083601e-01
6.61499262e-01 -1.02940965e+00 -4.72088248e-01 3.56714457e-01
-5.97650886e-01 -1.74847484e-01 5.17641962e-01 9.88842368e-01
-9.62983847e-01 4.09812123e-01 -3.25337827e-01 -5.14481485e-01
-1.21679401e+00 6.58294797e-01 5.78013301e-01 2.85790294e-01
-6.59265876e-01 1.11444724e+00 2.14842483e-01 6.21320069e-01
5.65319657e-01 -3.75147074e-01 -4.57728118e-01 1.71281293e-01
1.14257097e+00 4.54455286e-01 2.10442901e-01 -7.71205723e-01
-2.73254216e-01 2.18496457e-01 -2.65203506e-01 -1.18714459e-02
1.47410738e+00 -2.16039747e-01 -2.17844829e-01 3.27000409e-01
1.05481625e+00 -4.67033267e-01 -1.71555483e+00 3.08284461e-01
6.56388775e-02 -6.30130649e-01 5.75500250e-01 -5.15459418e-01
-1.48218846e+00 9.65669751e-01 6.87866032e-01 1.43114775e-01
1.51189697e+00 -5.74033976e-01 6.01415575e-01 -1.18145965e-01
3.54785076e-03 -1.44592667e+00 -6.36945665e-02 7.20330358e-01
8.17636669e-01 -1.25095916e+00 -1.73668712e-01 -2.67964303e-01
-4.45336044e-01 1.28280926e+00 5.69276690e-01 -4.18850899e-01
9.66507792e-01 5.60973346e-01 1.93231508e-01 2.01830976e-02
-8.95865202e-01 -7.31900930e-02 1.49122864e-01 6.12391472e-01
8.20877075e-01 -7.00869977e-01 -4.51409191e-01 3.76490295e-01
3.65237057e-01 1.52839154e-01 6.00211143e-01 1.15178216e+00
-6.43558919e-01 -1.03188396e+00 -3.65853190e-01 3.76468480e-01
-8.97168756e-01 -1.23868950e-01 -9.51264128e-02 4.62541133e-01
3.47546577e-01 1.24975073e+00 2.51594663e-01 -4.13035065e-01
3.07221353e-01 -3.13794881e-01 2.13274062e-01 -4.50822711e-01
-9.43041682e-01 5.32150865e-01 -1.24238156e-01 -9.79899228e-01
-5.89283943e-01 -6.19764030e-01 -1.26647127e+00 -3.94413918e-01
-1.55376837e-01 1.28677860e-01 6.77435040e-01 8.69014978e-01
1.26826063e-01 9.27091241e-01 3.39221478e-01 -1.11290205e+00
1.61198305e-03 -8.43487561e-01 -3.69420290e-01 5.02610028e-01
7.15865850e-01 -2.55127430e-01 -1.16831385e-01 4.68083978e-01] | [11.393436431884766, -2.220376968383789] |
c344fe5b-895b-4f32-9ed7-7b83cc672e87 | new-insights-on-target-speaker-extraction | 2202.00733 | null | https://arxiv.org/abs/2202.00733v1 | https://arxiv.org/pdf/2202.00733v1.pdf | New Insights on Target Speaker Extraction | In recent years, researchers have become increasingly interested in speaker extraction (SE), which is the task of extracting the speech of a target speaker from a mixture of interfering speakers with the help of auxiliary information about the target speaker. Several forms of auxiliary information have been employed in single-channel SE, such as a speech snippet enrolled from the target speaker or visual information corresponding to the spoken utterance. Many SE studies have reported performance improvement compared to speaker separation (SS) methods with oracle selection, arguing that this is due to the use of auxiliary information. However, such works have not considered state-of-the-art SS methods that have shown impressive separation performance. In this paper, we revise and examine the role of the auxiliary information in SE. Specifically, we compare the performance of two SE systems (audio-based and video-based) with SS using a common framework that utilizes the state-of-the-art dual-path recurrent neural network as the main learning machine. In addition, we study how much the considered SE systems rely on the auxiliary information by analyzing the systems' output for random auxiliary signals. Experimental evaluation on various datasets suggests that the main purpose of the auxiliary information in the considered SE systems is only to specify the target speaker in the mixture and that it does not provide consistent extraction performance gain when compared to the uninformed SS system. | ['Emanuël A. P. Habets', 'Soumitro Chakrabarty', 'Wolfgang Mack', 'Mohamed Elminshawi'] | 2022-02-01 | null | null | null | null | ['target-speaker-extraction', 'speaker-separation'] | ['audio', 'speech'] | [ 4.74669069e-01 2.78616101e-01 3.30705754e-02 -8.98389071e-02
-1.07960486e+00 -4.25750554e-01 6.67851448e-01 -5.31644523e-02
-4.56733048e-01 6.07539415e-01 2.44128302e-01 -3.33956808e-01
-1.51950896e-01 -1.68512911e-02 -5.22749722e-01 -1.15689600e+00
9.99765545e-02 2.71616876e-01 1.01681471e-01 -2.15819329e-01
1.79737341e-02 4.08609182e-01 -1.66319013e+00 1.19917914e-01
5.18668592e-01 8.04650545e-01 2.78813958e-01 7.10088372e-01
-1.55329794e-01 4.08304691e-01 -1.11387861e+00 1.11701205e-01
5.64596131e-02 -7.95334458e-01 -4.76532966e-01 1.08585991e-01
-4.16819900e-02 6.09477237e-02 -1.79099627e-02 7.85910428e-01
6.62603080e-01 4.22020927e-02 6.55180573e-01 -1.19739640e+00
1.74576938e-01 9.51761425e-01 -2.59017050e-01 3.54599774e-01
3.50617111e-01 -2.09457457e-01 8.29461753e-01 -1.00480556e+00
3.52240235e-01 9.71166849e-01 2.66837895e-01 4.23559815e-01
-1.20970953e+00 -7.36879587e-01 1.63506210e-01 2.08595663e-01
-1.55060947e+00 -1.12686610e+00 1.07859266e+00 -2.42695794e-01
7.43423581e-01 4.92199004e-01 2.61587560e-01 1.37469995e+00
-2.97972858e-01 1.03447700e+00 1.09860623e+00 -7.58679509e-01
1.23491608e-01 8.30415726e-01 4.76658851e-01 2.47481823e-01
-4.02139239e-02 1.36163980e-01 -6.54000401e-01 -2.10322097e-01
2.69572884e-01 -5.82956195e-01 -6.65332258e-01 -1.60391659e-01
-9.96027410e-01 6.58488870e-01 -1.47950828e-01 7.19952345e-01
-4.61017966e-01 -3.69538337e-01 3.25160503e-01 3.14215273e-01
4.39383209e-01 9.40075070e-02 -1.36148974e-01 -1.55505285e-01
-1.20104623e+00 -1.01331338e-01 1.15203381e+00 7.28933573e-01
2.75703877e-01 5.15837312e-01 -1.73565939e-01 7.66955972e-01
5.11947751e-01 4.55669791e-01 6.20417178e-01 -3.19935590e-01
6.38364911e-01 6.86813742e-02 -3.91652845e-02 -4.69498277e-01
-2.42267281e-01 -8.90408874e-01 -7.38872111e-01 1.22471765e-01
3.78294349e-01 -3.64983708e-01 -7.84538865e-01 1.71671820e+00
2.05987647e-01 2.34858245e-01 5.50836563e-01 7.49900520e-01
9.47527468e-01 7.34045804e-01 -3.59591275e-01 -5.48460007e-01
1.10343373e+00 -9.17728066e-01 -9.20334220e-01 -1.53882548e-01
2.03744650e-01 -7.35926032e-01 5.06706178e-01 6.86308563e-01
-8.36895347e-01 -5.38475394e-01 -1.11431146e+00 5.61423898e-01
-2.46374711e-01 6.06587827e-01 -4.00463939e-02 1.01785946e+00
-9.02491689e-01 1.31977633e-01 -7.32301354e-01 -1.39911592e-01
-2.43849277e-01 5.16416013e-01 -4.54984158e-01 5.53814411e-01
-1.18484688e+00 7.97759593e-01 1.03001736e-01 4.25046533e-01
-8.01406145e-01 -1.13901190e-01 -8.13682795e-01 3.32327753e-01
5.94448686e-01 -1.55089140e-01 1.10757959e+00 -1.35082877e+00
-1.93651390e+00 4.80201930e-01 -4.83903259e-01 -5.51730275e-01
4.62538242e-01 -2.82579094e-01 -7.20104575e-01 2.49142632e-01
-1.56217471e-01 3.96053270e-02 1.37847948e+00 -1.28686428e+00
-5.49721003e-01 -4.22277153e-01 -4.87232268e-01 1.85634166e-01
-4.06277925e-02 4.29552756e-02 -2.85001069e-01 -6.35893881e-01
9.89465863e-02 -9.01341379e-01 1.70097768e-01 -7.41004705e-01
-8.22231650e-01 -1.64032236e-01 9.50307429e-01 -7.83796847e-01
1.28884315e+00 -2.50953841e+00 4.29513425e-01 3.93945187e-01
-1.08001694e-01 5.14262676e-01 -2.82648914e-02 5.19753754e-01
-3.39134395e-01 -2.17037573e-01 -1.31736338e-01 -7.58185923e-01
1.13072820e-01 -1.94720849e-01 -4.77647215e-01 4.73371685e-01
4.81908098e-02 3.16699415e-01 -2.57337540e-01 -3.86224389e-01
2.32297644e-01 6.59381390e-01 -5.06759025e-02 3.08154732e-01
2.57340491e-01 5.19963562e-01 -2.91932911e-01 3.44540089e-01
3.40619415e-01 2.33534053e-01 1.20294623e-01 -1.28638268e-01
-1.97497904e-01 5.83244920e-01 -1.45775819e+00 1.13962042e+00
-3.49739164e-01 7.15463817e-01 6.26310349e-01 -1.10308146e+00
9.24530685e-01 8.88718307e-01 1.72647670e-01 -2.54459202e-01
1.94907174e-01 3.49929780e-01 3.22038352e-01 -3.03226858e-01
3.40702742e-01 -1.45839127e-02 1.95438772e-01 3.76225501e-01
1.58037543e-01 1.51137978e-01 8.07980672e-02 -5.01355231e-02
6.69672430e-01 -1.37695968e-02 4.52603608e-01 -4.08263057e-02
9.67696130e-01 -5.01987457e-01 2.32587337e-01 7.74724483e-01
4.01550569e-02 5.19403636e-01 3.04428160e-01 5.79129219e-01
-5.73470652e-01 -8.95473182e-01 -1.58008449e-02 8.75562668e-01
-1.28177732e-01 -1.67532876e-01 -7.66799092e-01 -4.14244533e-01
-4.52245593e-01 8.70100796e-01 -1.88677251e-01 7.25523755e-02
-4.92426783e-01 -7.49526620e-01 8.44729424e-01 2.22347587e-01
3.01297754e-01 -9.33980942e-01 -5.76063871e-01 2.99074918e-01
-1.63430780e-01 -1.13817978e+00 -3.46104264e-01 6.97409868e-01
-6.00753903e-01 -6.21417642e-01 -9.27998841e-01 -4.53734905e-01
4.05287743e-01 7.20022023e-02 5.24733722e-01 -4.43007380e-01
2.77597249e-01 4.94582802e-01 -4.44333225e-01 -4.21770245e-01
-7.31430054e-01 2.94889450e-01 1.48304120e-01 5.99733233e-01
3.14900786e-01 -5.80714881e-01 4.45587449e-02 3.36389959e-01
-7.69251883e-01 -1.47055641e-01 8.37045550e-01 6.90097451e-01
1.11759156e-01 1.53703004e-01 7.66404986e-01 -6.92248225e-01
7.41892874e-01 -4.46413249e-01 -4.68894064e-01 1.22382306e-01
-3.98308754e-01 4.20984328e-01 4.81981099e-01 -5.26499331e-01
-1.18440866e+00 1.68454096e-01 -5.01747131e-01 -4.47920650e-01
-4.52478945e-01 4.87182945e-01 -4.55905199e-01 1.70452610e-01
3.99070531e-01 7.50542819e-01 7.53398314e-02 -6.08695805e-01
-1.05762826e-02 1.16408789e+00 3.52156609e-01 -1.97176918e-01
5.89706540e-01 7.83623233e-02 -1.66256219e-01 -1.35622752e+00
-3.86071742e-01 -7.61684716e-01 -4.51763779e-01 -3.91227975e-02
4.84223634e-01 -6.59412563e-01 -5.93359053e-01 4.98658717e-01
-1.01155472e+00 4.95387949e-02 -1.22697234e-01 8.63810241e-01
-3.08252454e-01 4.83021945e-01 -3.20476264e-01 -1.38955426e+00
-3.14217687e-01 -1.29616761e+00 1.04724944e+00 -1.28849782e-02
-2.06997097e-01 -6.39325023e-01 8.17631278e-03 3.54948103e-01
3.60420674e-01 -2.16199309e-01 6.25404835e-01 -1.25285769e+00
-4.19849128e-01 -1.92854136e-01 3.60564470e-01 5.46402514e-01
3.68234068e-01 -2.41561279e-01 -1.42115855e+00 -1.88695356e-01
4.17927384e-01 2.40683079e-01 9.13556218e-01 5.29804289e-01
3.63615602e-01 -2.20042273e-01 -4.02815163e-01 2.13727430e-01
9.43515301e-01 4.58803713e-01 3.58191103e-01 -4.94760536e-02
4.40399170e-01 8.02309871e-01 3.70527565e-01 1.16793215e-01
-1.26980633e-01 9.27878916e-01 6.10150248e-02 -4.34611691e-03
1.29269306e-02 2.37635430e-03 9.27486777e-01 8.68294179e-01
9.26905125e-02 -5.77867270e-01 -5.59980989e-01 4.33380663e-01
-1.67921317e+00 -9.10363078e-01 -6.63070977e-02 2.43847585e+00
3.87786031e-01 4.26979959e-01 3.19917530e-01 8.13330114e-01
9.12503242e-01 1.45861343e-01 -2.38500848e-01 -2.61488169e-01
-1.52550012e-01 3.95315774e-02 2.81250089e-01 6.37975693e-01
-1.04023015e+00 3.22210014e-01 5.70899343e+00 9.35909331e-01
-1.45791471e+00 1.11393414e-01 1.67396680e-01 -6.47238120e-02
7.48831779e-02 -3.50928605e-01 -1.15785944e+00 3.44544739e-01
1.30356991e+00 7.83290491e-02 8.17092210e-02 5.50816000e-01
3.28827232e-01 -2.33319715e-01 -1.31501591e+00 1.04238486e+00
4.88762677e-01 -6.56379938e-01 -1.74631938e-01 2.23634124e-01
-1.28515316e-02 -1.83245748e-01 1.42420575e-01 3.77129734e-01
-5.33708632e-01 -8.91630948e-01 9.18557227e-01 4.25410181e-01
3.42443645e-01 -6.48798764e-01 9.65272546e-01 6.08235657e-01
-1.14808810e+00 -2.70468853e-02 3.88817012e-01 1.80560112e-01
2.70267993e-01 4.51550335e-01 -1.08653736e+00 7.75365293e-01
2.32416287e-01 1.02317467e-01 -3.68860334e-01 7.88620055e-01
-3.36455643e-01 1.01174772e+00 -5.94855011e-01 -3.26966345e-02
2.04495564e-01 -1.15749180e-01 1.19924998e+00 1.26183569e+00
3.85971636e-01 -1.58963427e-01 -1.06059946e-01 7.99533486e-01
3.07713926e-01 2.66280174e-01 -6.41521394e-01 -2.22957358e-01
3.25334907e-01 9.24643278e-01 -6.72988534e-01 -1.92924514e-01
-3.17680985e-01 8.38355839e-01 -2.68828392e-01 5.10189772e-01
-4.77951437e-01 -4.40469831e-01 2.02729300e-01 6.67589158e-02
6.18381321e-01 -2.93131471e-02 2.16776863e-01 -8.41395736e-01
2.13980272e-01 -8.51235092e-01 5.76009899e-02 -5.30494750e-01
-7.63105392e-01 9.71059620e-01 1.39714330e-01 -1.19487882e+00
-9.51126099e-01 -3.09415728e-01 -4.56414044e-01 1.11364853e+00
-1.22204173e+00 -9.24634278e-01 2.10974529e-01 5.12147486e-01
7.23780394e-01 -5.42185843e-01 7.92626143e-01 2.19466373e-01
-6.42791510e-01 5.84562182e-01 1.99673682e-01 1.33109167e-01
6.07378006e-01 -1.04471469e+00 -9.71245170e-02 9.35000658e-01
4.48734403e-01 5.74984252e-01 1.00267947e+00 -3.43332142e-01
-1.20362127e+00 -5.32222450e-01 9.16332424e-01 -2.79856533e-01
1.94536358e-01 -5.41896820e-01 -8.50642443e-01 5.76404750e-01
3.29895675e-01 -3.90564412e-01 5.88629246e-01 7.57846981e-02
-8.84220656e-03 -9.13383663e-02 -8.08858275e-01 2.64601886e-01
2.99024552e-01 -6.64420962e-01 -9.37208116e-01 -4.60260779e-01
4.95061398e-01 -2.34568834e-01 -1.79892644e-01 2.69721538e-01
4.26288545e-01 -1.02136850e+00 8.40884626e-01 -8.30274671e-02
-1.96450979e-01 -3.61104429e-01 -5.38248792e-02 -1.28333056e+00
1.33335441e-01 -5.37410975e-01 -9.47567821e-02 1.47221565e+00
5.98125339e-01 -7.18702257e-01 5.94173431e-01 2.24257007e-01
-4.65790704e-02 -4.04611826e-01 -1.08540261e+00 -8.50228786e-01
-5.01787186e-01 -3.71563733e-01 2.20501617e-01 5.17816067e-01
-1.24954760e-01 9.04465318e-01 -5.19280910e-01 4.04941559e-01
4.64336783e-01 3.19427885e-02 7.01183438e-01 -1.12739754e+00
-5.19101501e-01 -4.27239805e-01 -2.33171433e-01 -8.94419193e-01
4.29278612e-01 -5.10379076e-01 2.24145278e-01 -1.02661121e+00
-2.83627778e-01 -1.39599562e-01 -3.78551543e-01 1.16976090e-01
-5.88921197e-02 -9.24693570e-02 1.03108585e-01 -3.08459420e-02
-9.26173404e-02 6.37752056e-01 4.74241078e-01 -2.57139176e-01
-6.12351716e-01 6.74578905e-01 -4.44223851e-01 6.01533175e-01
4.97423738e-01 -4.92360532e-01 -4.97101992e-01 2.43926778e-01
-8.09918717e-02 3.95236194e-01 3.54055434e-01 -9.06127036e-01
4.07012910e-01 5.49204648e-01 -7.26649445e-03 -7.95454443e-01
7.97380090e-01 -1.01853251e+00 3.93093467e-01 2.66065031e-01
-3.64089310e-01 -4.18623358e-01 2.56144315e-01 5.02799869e-01
-6.27044439e-01 -6.70389175e-01 4.20353413e-01 2.05869585e-01
-3.56584579e-01 -4.54940051e-01 -7.17023551e-01 -3.90978873e-01
6.58026218e-01 -2.68152207e-01 2.05589548e-01 -7.19979644e-01
-1.04716480e+00 -3.19644213e-01 -3.74287426e-01 3.07677269e-01
4.26263750e-01 -8.16828191e-01 -8.02708924e-01 2.83519745e-01
-7.46317282e-02 -4.12675947e-01 3.16278413e-02 1.12760615e+00
3.51509035e-01 8.72808456e-01 6.79057240e-02 -6.42512321e-01
-1.80359924e+00 4.80226547e-01 2.65656531e-01 -8.48803073e-02
-4.14324194e-01 6.25369847e-01 2.78633296e-01 -4.40228879e-02
6.08750224e-01 -3.11644942e-01 -5.06986856e-01 3.57864857e-01
3.40302736e-01 3.86501580e-01 2.71687716e-01 -8.40841472e-01
-4.04024094e-01 2.35795334e-01 1.09991387e-01 -6.74690187e-01
1.14435446e+00 -3.00308347e-01 2.04681233e-01 9.52405751e-01
9.46323991e-01 5.30008197e-01 -6.59412324e-01 -2.63735175e-01
2.07458884e-02 7.79059064e-03 2.04208151e-01 -5.55737674e-01
-7.85673380e-01 9.15758133e-01 6.24096036e-01 3.79013151e-01
1.18972552e+00 -1.12281278e-01 1.83788612e-01 3.40436369e-01
2.87384838e-01 -9.25664127e-01 -4.40300882e-01 1.92099199e-01
8.46756160e-01 -1.04854333e+00 -2.80487567e-01 -5.47355592e-01
-6.32322848e-01 1.02813983e+00 9.41789672e-02 2.67424703e-01
6.99220777e-01 2.80898988e-01 1.06980450e-01 4.13360111e-02
-5.81566095e-01 -3.38929743e-01 4.35115486e-01 3.58134657e-01
3.41318071e-01 -8.48628581e-02 -1.90261349e-01 9.51279223e-01
-1.50375947e-01 -2.21323684e-01 5.17249644e-01 7.15812147e-01
-2.63767600e-01 -1.18909574e+00 -8.38894188e-01 2.67345876e-01
-5.36014855e-01 -2.11703971e-01 -4.28113461e-01 7.75763273e-01
-1.82635352e-01 1.15912414e+00 -2.59538829e-01 -8.33391696e-02
2.52758354e-01 5.44991314e-01 2.63272941e-01 -6.75388694e-01
-7.28396237e-01 5.90995431e-01 3.14037263e-01 -1.00655802e-01
-6.94831848e-01 -7.03949988e-01 -9.03940618e-01 4.22090501e-01
-6.85690105e-01 7.35008597e-01 9.27539468e-01 1.25615621e+00
1.07343301e-01 6.86350405e-01 5.84637761e-01 -9.44713414e-01
-4.94146585e-01 -1.23062396e+00 -7.77477741e-01 -1.28996730e-01
8.51516366e-01 -4.33498830e-01 -7.63384402e-01 -7.42548183e-02] | [14.748710632324219, 5.825908184051514] |
9456886d-b0c2-4a7f-a28f-601cd812656c | radar-shadow-detection-in-sar-images-using | 1309.1830 | null | http://arxiv.org/abs/1309.1830v1 | http://arxiv.org/pdf/1309.1830v1.pdf | Radar shadow detection in SAR images using DEM and projections | Synthetic aperture radar (SAR) images are widely used in target recognition
tasks nowadays. In this letter, we propose an automatic approach for radar
shadow detection and extraction from SAR images utilizing geometric projections
along with the digital elevation model (DEM) which corresponds to the given
geo-referenced SAR image. First, the DEM is rotated into the radar geometry so
that each row would match that of a radar line of sight. Next, we extract the
shadow regions by processing row by row until the image is covered fully. We
test the proposed shadow detection approach on different DEMs and a simulated
1D signals and 2D hills and volleys modeled by various variance based Gaussian
functions. Experimental results indicate the proposed algorithm produces good
results in detecting shadows in SAR images with high resolution. | ['V. B. S. Prasath', 'O. Haddad'] | 2013-09-07 | null | null | null | null | ['shadow-detection', 'detecting-shadows'] | ['computer-vision', 'computer-vision'] | [ 8.29192936e-01 -5.17180443e-01 6.45465016e-01 -3.80009174e-01
-4.09546882e-01 -4.06485140e-01 5.90511620e-01 -4.39035088e-01
-3.69760126e-01 9.28683341e-01 1.92501083e-01 -5.30881226e-01
-1.64588362e-01 -9.46947932e-01 8.18130821e-02 -9.19447005e-01
-3.24480206e-01 5.69963813e-01 4.15789247e-01 -6.33596778e-02
3.45475316e-01 1.14824545e+00 -1.01371109e+00 -1.18403777e-01
9.81480718e-01 7.00211763e-01 5.62818944e-01 8.93166959e-01
4.47900385e-01 2.65858620e-01 -7.20616639e-01 2.50760376e-01
6.44879520e-01 -2.56983727e-01 2.00352222e-01 5.22346914e-01
2.86906540e-01 -4.79794174e-01 -4.33702469e-01 1.05014646e+00
3.13887805e-01 -1.44396976e-01 9.62846994e-01 -6.65331542e-01
-4.35766906e-01 1.06308766e-01 -1.10574281e+00 4.04436558e-01
-8.43100809e-03 -2.23080516e-01 1.53351724e-01 -1.13395548e+00
4.44073260e-01 1.09793448e+00 5.18914759e-01 -4.20146167e-01
-9.50131059e-01 -6.76426411e-01 -3.20042819e-01 1.88122377e-01
-1.62643540e+00 -3.15545976e-01 5.71259558e-01 -2.83946365e-01
2.92955875e-01 3.69388431e-01 6.85425282e-01 2.23812148e-01
8.35230827e-01 3.15237075e-01 1.87026978e+00 -4.73610401e-01
3.41728628e-02 -1.34273261e-01 6.23383462e-01 3.76549691e-01
1.06514072e+00 3.53790343e-01 -1.32488072e-01 -3.42177361e-01
4.81743842e-01 1.35051548e-01 -6.35852039e-01 -3.67145926e-01
-9.61773038e-01 9.04609859e-01 2.76484042e-01 2.42783770e-01
-7.20823348e-01 -2.87913889e-01 -4.35949922e-01 -9.25917625e-02
2.17446968e-01 1.75892591e-01 2.26618517e-02 8.76241326e-01
-1.28714418e+00 3.06367338e-01 5.84572792e-01 7.37095833e-01
5.14016986e-01 7.49972522e-01 1.36904150e-01 3.15747589e-01
4.71389771e-01 1.62585723e+00 -8.16758871e-02 -1.77532151e-01
1.63264245e-01 3.17415774e-01 5.81426501e-01 -1.14046991e+00
-3.36114258e-01 -8.66649747e-01 -9.08302307e-01 6.96869195e-01
1.51509061e-01 -2.30788842e-01 -1.53471351e+00 7.81321704e-01
6.67893738e-02 3.09700072e-01 8.48568261e-01 8.12905848e-01
5.12822986e-01 8.96567643e-01 -3.49821031e-01 -4.52739120e-01
1.68838418e+00 -3.01780015e-01 -7.48743534e-01 -7.91710675e-01
-5.69312498e-02 -1.12102425e+00 3.34990531e-01 6.05977952e-01
-1.88571602e-01 -2.41663083e-01 -1.46139646e+00 9.67717409e-01
-1.50912821e-01 6.27921283e-01 4.22431618e-01 7.79380500e-01
-5.12739241e-01 6.00906741e-03 -5.77106118e-01 -1.40964776e-01
1.79669604e-01 -5.02163768e-02 2.56858543e-02 -4.39542621e-01
-9.52272177e-01 1.11644769e+00 1.45804107e-01 6.78436399e-01
-9.35887873e-01 -3.49903166e-01 -7.43948817e-01 -1.59292102e-01
1.53056189e-01 -2.18588173e-01 5.93194485e-01 -5.96030295e-01
-7.50421762e-01 5.02565503e-01 -1.46515304e-02 -7.16424048e-01
1.86421931e-01 -5.95723927e-01 -1.06401527e+00 9.91088822e-02
-3.38036902e-02 -2.84123808e-01 1.02742326e+00 -1.63737583e+00
-3.22353214e-01 -7.92670310e-01 -7.81153142e-01 4.48207766e-01
6.60706282e-01 5.99345490e-02 4.58116293e-01 -4.92080688e-01
6.26348436e-01 -8.96045506e-01 -3.88575912e-01 -4.77892935e-01
-6.12134457e-01 1.04704213e+00 1.36526537e+00 -9.71025527e-01
9.59782541e-01 -1.97556794e+00 -5.35176575e-01 8.12318027e-01
-1.70816377e-01 3.79736423e-01 2.50542372e-01 4.82401490e-01
1.11611806e-01 -7.23798871e-01 -7.62021065e-01 5.03327489e-01
-4.47794855e-01 -1.31549286e-02 -7.99427688e-01 9.28480089e-01
-4.68968868e-01 3.76501590e-01 -1.00764714e-01 -3.88167441e-01
7.84493983e-03 5.06443977e-01 4.60840374e-01 1.05356708e-01
1.40252650e-01 2.01804325e-01 -7.19952941e-01 8.34387839e-01
1.76001918e+00 3.98236305e-01 2.14221418e-01 -2.13025704e-01
-5.69789290e-01 -4.62770790e-01 -1.35000038e+00 4.34093267e-01
-2.93741212e-03 8.92313242e-01 2.73620397e-01 -5.00777304e-01
1.47677493e+00 -4.46063690e-02 1.56943616e-03 -5.58640122e-01
-7.78068975e-02 -2.08621863e-02 1.97524861e-01 -1.61654964e-01
7.00400531e-01 -4.17077959e-01 -3.25299054e-02 2.99709171e-01
-1.06829476e+00 -3.71894896e-01 -4.16096330e-01 2.22845618e-02
8.89487028e-01 -2.71422595e-01 6.49668634e-01 -4.20318395e-01
4.08969402e-01 6.93303943e-01 4.55238730e-01 5.26241422e-01
1.49773329e-01 3.36831719e-01 -1.58174485e-01 -3.33144993e-01
-7.36215174e-01 -1.40541232e+00 -4.50063765e-01 1.86536416e-01
3.14798057e-01 4.90320295e-01 -3.31707925e-01 -1.39480338e-01
-1.97189301e-02 1.01506329e+00 -4.23604339e-01 3.82258385e-01
-5.53228259e-01 -1.42480826e+00 3.88544798e-01 6.16496019e-02
9.20275748e-01 -8.60190749e-01 -1.31939018e+00 2.97021903e-02
3.59336473e-02 -1.17280531e+00 1.44715592e-01 5.95206618e-02
-9.53736544e-01 -1.08120263e+00 -8.56388569e-01 -5.06449163e-01
8.30892682e-01 1.10686469e+00 7.76138186e-01 -3.56498778e-01
-8.47445786e-01 7.09098652e-02 -3.65692437e-01 -5.59889972e-01
1.53112978e-01 -9.76653755e-01 -8.00554827e-02 1.54625162e-01
1.79640070e-01 -6.11180425e-01 -7.10941136e-01 3.67586881e-01
-6.37171507e-01 6.76843598e-02 1.20403266e+00 5.34552455e-01
2.96175867e-01 4.97247368e-01 -4.22385298e-02 -6.66344941e-01
3.44418585e-01 -1.05683036e-01 -1.18384314e+00 1.65601760e-01
-3.87654722e-01 -1.90890342e-01 1.43652111e-01 1.38618410e-01
-1.56945622e+00 1.60503522e-01 5.91987312e-01 4.91610430e-02
-4.07109559e-01 5.00513196e-01 -2.42460489e-01 -3.09413135e-01
7.38366365e-01 6.07792079e-01 -4.43192452e-01 -2.89940417e-01
5.74013144e-02 8.28635275e-01 7.07280457e-01 1.07942834e-01
1.60143578e+00 8.66755843e-01 2.68512011e-01 -1.46065986e+00
-7.58196831e-01 -4.25104052e-01 -6.74844384e-01 -1.66653186e-01
8.57806385e-01 -8.50183368e-01 6.39795931e-03 4.78746474e-01
-8.38384390e-01 -7.83816576e-02 3.56509715e-01 1.03694975e+00
-1.21604457e-01 5.17767191e-01 1.21609293e-01 -1.37964201e+00
-5.85966170e-01 -4.88385946e-01 7.21708179e-01 2.60925174e-01
2.23378152e-01 -5.03939271e-01 2.86631733e-01 1.54851988e-01
3.74184281e-01 6.24308288e-01 6.94263756e-01 -4.09642518e-01
-9.59795058e-01 -2.74227917e-01 -7.04124451e-01 -2.15183329e-02
1.28102750e-01 -3.93775642e-01 -5.48070312e-01 -4.10729617e-01
3.57769012e-01 3.01625282e-01 1.00043786e+00 7.90234566e-01
1.02302052e-01 -1.03112489e-01 -6.40323937e-01 6.66542470e-01
2.10662651e+00 6.45245969e-01 1.08735096e+00 4.09006596e-01
1.54755011e-01 3.39426875e-01 1.29366755e+00 6.29475772e-01
-2.17340961e-01 3.61776441e-01 2.38285139e-01 -2.62933463e-01
-2.27802601e-02 3.36160123e-01 3.17751825e-01 -2.01277179e-03
-6.80718943e-02 -3.01468521e-01 -1.00616693e+00 4.09618944e-01
-1.24720466e+00 -1.00725532e+00 -8.63293350e-01 2.17281675e+00
-1.77013636e-01 5.50703704e-02 -8.08243811e-01 -2.20192205e-02
6.41944766e-01 3.36485296e-01 -1.82115704e-01 1.64952334e-02
-4.73412693e-01 5.14334559e-01 1.22427940e+00 9.47169960e-01
-8.85268688e-01 9.11428213e-01 5.78803492e+00 3.74138296e-01
-9.88107562e-01 -3.18539053e-01 -1.49193838e-01 7.24824071e-01
-3.06245536e-01 4.19732422e-01 -9.22252536e-01 6.14303201e-02
5.67581654e-01 -1.37746781e-01 -1.80479020e-01 4.25174177e-01
4.39000845e-01 -1.14560032e+00 -1.26676008e-01 6.83625638e-01
2.42658764e-01 -9.11876678e-01 1.49787888e-01 3.58815134e-01
5.93166530e-01 -1.33598983e-01 5.92617728e-02 -2.69633174e-01
7.11154103e-01 -9.77487445e-01 1.53174728e-01 9.55141425e-01
6.03589773e-01 -6.40908837e-01 8.82039905e-01 2.83905119e-01
-1.26858544e+00 1.68045193e-01 -4.86117303e-01 -4.87036258e-02
4.25284892e-01 8.41307998e-01 -1.41351187e+00 8.39284658e-01
2.12064698e-01 -1.59295183e-02 -3.85251462e-01 1.31682146e+00
-1.30588964e-01 5.30817151e-01 -2.96029270e-01 2.83993006e-01
4.62533712e-01 -8.10743988e-01 8.65636885e-01 1.40401268e+00
1.00671470e+00 8.35515618e-01 1.84385598e-01 3.58308733e-01
7.45387912e-01 -7.29247108e-02 -9.83774900e-01 1.91402644e-01
4.06864047e-01 1.10320365e+00 -9.21392381e-01 -3.90263677e-01
3.63492407e-02 8.13374519e-01 -9.11963761e-01 5.75499237e-01
-8.14746439e-01 -6.15702391e-01 1.70792267e-01 3.27628404e-01
6.10101819e-01 -7.80149341e-01 -3.61602455e-01 -5.94066024e-01
-3.70699525e-01 -6.17441833e-01 1.98535681e-01 -1.34012854e+00
-5.20278513e-01 9.09745216e-01 3.21092606e-01 -1.51799715e+00
3.12249273e-01 -5.76943099e-01 -7.69145072e-01 1.10518181e+00
-1.49246752e+00 -1.30090237e+00 -8.77770841e-01 5.76415062e-01
3.80795240e-01 -4.63865131e-01 8.47073078e-01 -3.12441379e-01
-4.67202440e-03 -4.44798529e-01 3.62925500e-01 3.42820771e-02
6.17785752e-01 -8.00095916e-01 2.37327546e-01 1.34664428e+00
-4.36763614e-02 2.22946480e-01 1.43086028e+00 -1.23456454e+00
-1.26315212e+00 -9.23030198e-01 6.68726802e-01 1.87497020e-01
5.12488604e-01 -8.64103287e-02 -6.14408553e-01 7.59494126e-01
3.37287456e-01 -3.90394479e-01 5.39087296e-01 -4.37860548e-01
-1.72399133e-02 -1.88883662e-01 -1.23308623e+00 4.52563494e-01
4.36796516e-01 1.55767292e-01 -1.01554358e+00 3.65415365e-01
-9.90650579e-02 -2.86969602e-01 -2.54790872e-01 7.55347788e-01
5.72705507e-01 -1.09683478e+00 9.30149078e-01 9.53562334e-02
-2.67855585e-01 -7.51963258e-01 -7.94385850e-01 -1.40883052e+00
-3.40532750e-01 -1.74429446e-01 5.60979962e-01 5.56219101e-01
1.48977831e-01 -7.57574022e-01 8.25531781e-01 -3.37634027e-01
-2.81575263e-01 -1.91792592e-01 -6.39200509e-01 -7.49289215e-01
-6.22213900e-01 1.06609846e-02 1.02395818e-01 4.68195438e-01
-7.45059490e-01 3.75897974e-01 -5.38404346e-01 1.32449698e+00
1.53086543e+00 8.57714951e-01 9.50103879e-01 -1.61748517e+00
-3.83744910e-02 4.50597167e-01 -3.12665641e-01 -7.92545199e-01
-1.62105918e-01 -3.83460075e-01 2.46931180e-01 -1.88012362e+00
2.12015644e-01 -5.09344995e-01 2.03916848e-01 -1.74033061e-01
2.76492566e-01 2.44322658e-01 -3.00191548e-02 3.79669100e-01
3.17939818e-01 4.57425624e-01 1.02536988e+00 1.60394516e-02
-1.50793076e-01 5.27444005e-01 -2.87299901e-01 1.02320838e+00
7.40928113e-01 -5.23269534e-01 -2.80359715e-01 -1.89498022e-01
-4.23978269e-01 5.19784927e-01 5.42666554e-01 -1.32661712e+00
1.44797489e-01 -3.60148370e-01 8.57816637e-01 -1.56504524e+00
8.53862345e-01 -1.00976384e+00 4.82975930e-01 8.66855323e-01
6.00162208e-01 2.90102046e-02 1.52220652e-01 7.75097549e-01
-2.88749039e-02 -1.98198467e-01 1.07458103e+00 -7.84748048e-02
-7.54095733e-01 7.95668643e-03 -7.74141252e-01 -5.19846499e-01
1.15770626e+00 -6.05342567e-01 -4.49405909e-01 -4.89876568e-01
-7.24441111e-01 -4.42881696e-02 3.15153450e-01 -4.30166006e-01
1.19541728e+00 -8.41967344e-01 -1.11312664e+00 3.16897452e-01
3.56911272e-02 -6.69423640e-01 3.10693711e-01 8.22945297e-01
-7.70803332e-01 5.29833853e-01 -5.53460956e-01 -3.34682226e-01
-1.86572695e+00 2.03741655e-01 2.07387879e-01 -3.64972264e-01
-9.25319672e-01 2.26884604e-01 4.12724763e-01 1.38851330e-01
-2.69332737e-01 -1.65997475e-01 -2.36816987e-01 -1.29950419e-01
6.23791397e-01 2.46471986e-01 -1.84693679e-01 -7.62899816e-01
-4.19119120e-01 9.27584529e-01 1.55437589e-01 -7.36680329e-01
1.40224004e+00 2.06615049e-02 1.09910622e-01 1.08670563e-01
1.90825924e-01 8.18669617e-01 -9.84078646e-01 -3.46737683e-01
-1.67326909e-02 -8.98815453e-01 1.33130908e-01 -8.91672969e-01
-5.18725574e-01 5.44778585e-01 8.39966238e-01 -1.26161546e-01
1.18129826e+00 -3.73571485e-01 5.48743308e-01 6.20643914e-01
4.05363917e-01 -6.10397160e-01 -4.44256127e-01 4.12626594e-01
9.63925362e-01 -6.88568950e-01 8.29079628e-01 -6.44847572e-01
-9.18348670e-01 1.06470549e+00 -8.94332752e-02 -5.03023267e-01
8.00922573e-01 6.06237292e-01 2.03014106e-01 -4.72021848e-01
-2.19239399e-01 -5.61938286e-01 -2.25063898e-02 9.90935504e-01
-1.40107825e-01 4.90488380e-01 -3.44950885e-01 9.07340199e-02
-3.27316731e-01 -2.65401632e-01 9.93197978e-01 1.05488193e+00
-1.45847297e+00 -5.60806155e-01 -1.32388210e+00 4.17274863e-01
-4.18292694e-02 -3.48108619e-01 -4.58997399e-01 1.11821067e+00
-1.91938058e-01 8.24844182e-01 4.89390939e-02 -5.69185168e-02
2.79975921e-01 -2.02618301e-01 4.78879690e-01 -4.08717424e-01
2.36264259e-01 2.31856123e-01 4.62984085e-01 1.53753996e-01
-2.42413860e-02 -7.87801862e-01 -1.20838141e+00 1.68782622e-01
-3.63321394e-01 5.67148268e-01 7.17325091e-01 7.42059648e-01
-2.22763523e-01 2.15727016e-01 6.85161829e-01 -5.24457037e-01
-3.71680379e-01 -1.08178449e+00 -1.09645808e+00 -5.42768896e-01
2.80037314e-01 -8.36044192e-01 -4.51791435e-01 2.73116250e-02] | [6.783083915710449, 1.02863609790802] |
49e15c1b-ca89-48df-8697-67ff5433b6a8 | online-risk-bounded-motion-planning-for | 1904.02341 | null | http://arxiv.org/abs/1904.02341v1 | http://arxiv.org/pdf/1904.02341v1.pdf | Online Risk-Bounded Motion Planning for Autonomous Vehicles in Dynamic Environments | A crucial challenge to efficient and robust motion planning for autonomous
vehicles is understanding the intentions of the surrounding agents. Ignoring
the intentions of the other agents in dynamic environments can lead to risky or
over-conservative plans. In this work, we model the motion planning problem as
a partially observable Markov decision process (POMDP) and propose an online
system that combines an intent recognition algorithm and a POMDP solver to
generate risk-bounded plans for the ego vehicle navigating with a number of
dynamic agent vehicles. The intent recognition algorithm predicts the
probabilistic hybrid motion states of each agent vehicle over a finite horizon
using Bayesian filtering and a library of pre-learned maneuver motion models.
We update the POMDP model with the intent recognition results in real time and
solve it using a heuristic search algorithm which produces policies with
upper-bound guarantees on the probability of near colliding with other dynamic
agents. We demonstrate that our system is able to generate better motion plans
in terms of efficiency and safety in a number of challenging environments
including unprotected intersection left turns and lane changes as compared to
the baseline methods. | ['Brian C. Williams', 'Sungkweon Hong', 'Andreas Hofmann', 'Xin Huang'] | 2019-04-04 | null | null | null | null | ['intent-recognition'] | ['natural-language-processing'] | [ 4.24837582e-02 8.55872989e-01 -2.25514904e-01 -3.15096915e-01
-8.47545385e-01 -6.13213718e-01 8.98279786e-01 -1.61908433e-01
-5.61418772e-01 7.75725007e-01 2.29790792e-01 -8.00889850e-01
-1.75023332e-01 -8.20071936e-01 -6.41152442e-01 -6.07113302e-01
-5.23290038e-01 1.10211194e+00 6.99591517e-01 -1.76666200e-01
-1.25614125e-02 6.49196684e-01 -1.40442121e+00 -3.72748911e-01
5.58730662e-01 5.59255540e-01 2.93827862e-01 1.00090742e+00
3.61305356e-01 9.94524360e-01 7.60505944e-02 -5.71035184e-02
5.57813585e-01 -4.25507240e-02 -7.05728889e-01 9.14826095e-02
-2.57449806e-01 -8.58539462e-01 -4.76735681e-01 8.77324641e-01
1.06160832e-03 6.15237057e-01 6.43613875e-01 -1.82318068e+00
8.11460912e-01 2.90109813e-01 2.61276998e-02 6.22982979e-02
2.86722898e-01 8.38258147e-01 5.97675204e-01 -1.08044684e-01
8.52162182e-01 1.61965024e+00 2.65266925e-01 4.72645879e-01
-9.02496099e-01 -2.49780372e-01 5.15969217e-01 5.20961285e-01
-1.13839602e+00 -3.34164500e-01 1.85817793e-01 -2.67078161e-01
1.08336604e+00 1.99993744e-01 6.56889737e-01 8.96449208e-01
8.21378469e-01 7.65866995e-01 6.61946714e-01 1.82951495e-01
6.97698653e-01 -2.69063205e-01 -5.27509749e-02 6.87986076e-01
1.72548637e-01 6.68248117e-01 1.75322089e-02 -5.05596101e-01
-2.60910615e-02 -2.71065030e-02 3.04127961e-01 -5.49986064e-01
-1.00742900e+00 9.32737172e-01 -1.20163172e-01 -5.67159832e-01
-9.86308873e-01 3.39560807e-01 1.28906071e-01 1.74106926e-01
-1.88393429e-01 5.30437231e-02 -1.58297241e-01 -3.26845795e-01
-6.05916679e-01 9.41297352e-01 1.00900114e+00 1.05047810e+00
7.47515857e-01 1.02457449e-01 -2.40960687e-01 -2.13305429e-01
7.46788144e-01 9.23649609e-01 -4.29057926e-01 -1.70607090e+00
5.43804824e-01 1.90127790e-01 8.90057504e-01 -8.01286995e-01
-5.87435305e-01 3.16871077e-01 -6.94083348e-02 8.81738842e-01
2.13765621e-01 -4.99300510e-01 -1.09404242e+00 1.62089002e+00
7.55880177e-01 2.02599958e-01 4.56122994e-01 7.80451953e-01
4.80264379e-03 9.31894481e-01 1.44529298e-01 -4.56352592e-01
9.57581997e-01 -7.75615811e-01 -7.87720323e-01 -6.35542154e-01
4.43536997e-01 -4.27527785e-01 8.38713720e-02 2.69508064e-01
-1.09828639e+00 7.81928599e-02 -1.06950498e+00 3.04885715e-01
1.23490259e-01 -6.35538697e-01 2.49355406e-01 4.45684701e-01
-1.03206468e+00 3.38054985e-01 -1.43735802e+00 -3.68387014e-01
2.38827497e-01 5.01684129e-01 -1.08160533e-01 -2.77616084e-01
-9.36429799e-01 1.38207924e+00 4.85924512e-01 -2.07386017e-01
-1.92098355e+00 -3.96629348e-02 -1.03005934e+00 -1.83714375e-01
1.04800200e+00 -6.55702174e-01 1.51640010e+00 -1.90500826e-01
-1.74062145e+00 5.57859652e-02 -2.83447474e-01 -9.54869330e-01
9.03780341e-01 -3.63262929e-02 -1.80674940e-01 1.51380599e-02
3.34613293e-01 7.62910664e-01 7.02433050e-01 -1.16070914e+00
-1.19296443e+00 -1.68424711e-01 2.00582623e-01 8.18134844e-01
8.21890950e-01 -2.39095792e-01 -3.81971449e-01 2.52986372e-01
5.92336580e-02 -1.65355623e+00 -9.20548081e-01 -2.44511366e-01
-2.22252041e-01 6.42981613e-03 9.07594323e-01 -3.51877689e-01
7.13968992e-01 -1.81335843e+00 6.37426451e-02 3.54717940e-01
-1.12592429e-01 -4.10471596e-02 -1.39698595e-01 6.70107663e-01
7.15724885e-01 -2.94164717e-01 -2.15083198e-03 -3.25502694e-01
1.97643280e-01 8.08018863e-01 -6.96936011e-01 7.04380870e-01
-2.27615729e-01 6.06858075e-01 -9.12532151e-01 -3.59155685e-01
3.99375707e-01 1.12390712e-01 -4.59420532e-01 2.60502279e-01
-6.39895737e-01 5.34696639e-01 -8.24837923e-01 2.05086306e-01
6.10240161e-01 5.60793698e-01 3.70402575e-01 7.14932859e-01
-5.43529809e-01 4.12512749e-01 -1.40884411e+00 9.75757062e-01
-1.33495510e-01 4.16083574e-01 4.57056999e-01 -1.89557403e-01
4.96704161e-01 2.45476797e-01 4.17741567e-01 -3.21412563e-01
3.34456742e-01 -9.98924673e-02 -1.40584931e-01 -2.75172889e-01
7.84012437e-01 -1.42847642e-01 -5.13701797e-01 6.99164152e-01
-4.12329972e-01 -4.05490279e-01 -1.43916816e-01 1.71744943e-01
1.40417755e+00 -1.52250677e-01 3.08085561e-01 -5.62487543e-02
2.97308147e-01 7.02063382e-01 8.01496565e-01 1.15116429e+00
-6.57139599e-01 -2.16168165e-01 4.59751159e-01 -6.51229143e-01
-8.03817868e-01 -9.07085240e-01 4.97281164e-01 7.94957519e-01
7.76029646e-01 -9.80703607e-02 -6.09371781e-01 -3.42064202e-01
-9.34280083e-02 1.43009603e+00 -3.92587751e-01 -1.39025852e-01
-7.56407261e-01 -2.97593743e-01 2.32413992e-01 1.43463150e-01
2.99708992e-01 -7.33193934e-01 -1.23484874e+00 3.94493133e-01
-5.09768128e-02 -1.23771620e+00 -1.97138995e-01 -1.24557070e-01
-4.63283777e-01 -9.53255475e-01 1.72820702e-01 -2.77492911e-01
5.20375848e-01 4.57716316e-01 3.93458068e-01 -2.26900578e-01
4.11762774e-01 3.60099286e-01 3.81303132e-02 -7.50807405e-01
-9.55282390e-01 -3.60081404e-01 2.77933627e-01 -6.15683757e-02
6.13445118e-02 -5.32663614e-02 -3.26319188e-01 5.34217000e-01
-4.44385171e-01 4.35668916e-01 2.46736154e-01 2.04950035e-01
6.61069453e-01 5.45478940e-01 -1.31675184e-01 -2.30491310e-01
4.37530100e-01 -6.95054531e-01 -1.21479988e+00 -1.40059814e-01
-3.40752900e-01 2.80951887e-01 2.83831090e-01 -1.04187690e-01
-1.13997376e+00 5.32124579e-01 6.08693771e-02 -1.10858038e-01
-3.33815008e-01 1.21219419e-01 -3.77049476e-01 5.12042679e-02
1.04562260e-01 1.67995095e-01 1.94490984e-01 1.08613618e-01
2.87378967e-01 1.84979811e-01 3.75676423e-01 -3.49268019e-01
8.60284269e-01 9.54735041e-01 4.59102124e-01 -6.46964967e-01
-1.63011849e-01 -3.48116159e-01 -1.48457170e-01 -7.16473579e-01
9.11423385e-01 -1.01600599e+00 -9.99865353e-01 3.14990491e-01
-1.14252436e+00 -7.21263707e-01 -1.36748180e-01 5.83767593e-01
-1.13680899e+00 2.44052887e-01 -1.73409507e-01 -1.29466844e+00
2.29299709e-01 -1.52776515e+00 9.04340088e-01 2.13995427e-01
-1.48060173e-01 -6.17839754e-01 2.86064148e-01 2.99330235e-01
1.81962803e-01 3.40709865e-01 4.21702862e-01 -5.04284978e-01
-1.23561633e+00 -1.51316509e-01 3.78169417e-01 -6.46550953e-01
-5.54804802e-01 -1.34473637e-01 -4.09787118e-01 -3.14407289e-01
2.62538344e-02 1.34094805e-01 4.59759414e-01 4.61227000e-01
-6.05792515e-02 -8.86982560e-01 -8.13295007e-01 7.07975328e-02
1.26897264e+00 7.36703873e-01 5.45431197e-01 5.23081362e-01
1.47920489e-01 8.57023299e-01 1.25683808e+00 5.46220779e-01
9.77196336e-01 7.71455705e-01 1.06030762e+00 7.08059728e-01
5.71271002e-01 -3.55369568e-01 7.37168193e-01 -2.89642274e-01
1.26027847e-02 -3.69413793e-01 -1.18746054e+00 7.90223360e-01
-2.26150417e+00 -1.15582120e+00 -6.61371052e-02 2.15311003e+00
6.76723942e-02 4.55275208e-01 3.14150333e-01 -3.24204147e-01
4.50700849e-01 1.02724433e-01 -7.53879845e-01 -5.51233470e-01
4.47897434e-01 -7.64784336e-01 1.26891124e+00 1.22589326e+00
-1.13108683e+00 1.21187699e+00 6.49323034e+00 4.43746090e-01
-6.10304415e-01 2.32774109e-01 2.38800839e-01 -9.01010185e-02
-2.74804741e-01 5.09660065e-01 -1.11289382e+00 1.57545969e-01
1.26504862e+00 -3.73419851e-01 5.45787632e-01 9.77179289e-01
7.32775331e-01 -6.92707300e-01 -7.33837426e-01 2.22267643e-01
-4.53548253e-01 -1.36621821e+00 -3.77148986e-01 3.18394601e-01
7.73333251e-01 4.82230544e-01 -2.35069364e-01 3.20491195e-01
1.39776015e+00 -8.35688651e-01 1.22670281e+00 5.40852785e-01
-1.42141268e-01 -1.12419939e+00 6.10802829e-01 1.11315441e+00
-1.02809918e+00 -4.63138878e-01 1.07446574e-01 -3.64792943e-01
1.10068095e+00 -1.54909119e-01 -1.28059340e+00 3.47661763e-01
3.43309999e-01 7.71039352e-02 2.66921759e-01 8.72526348e-01
-3.41713458e-01 3.37249249e-01 -7.45576680e-01 -7.75883421e-02
7.55584002e-01 -2.08426818e-01 1.36730015e+00 5.44122040e-01
1.27759472e-01 4.02426690e-01 7.48543084e-01 5.71493208e-01
8.19807410e-01 -6.89714670e-01 -9.04479861e-01 2.24114746e-01
4.64511842e-01 7.07054794e-01 -6.35282457e-01 -3.39895368e-01
-1.94110572e-02 3.63929600e-01 -1.08904675e-01 4.14431125e-01
-9.79315281e-01 3.37544322e-01 1.14443088e+00 -6.32126778e-02
3.42701346e-01 -6.30895793e-01 -1.40613645e-01 -4.52981532e-01
-3.09742153e-01 -5.34998000e-01 2.51770347e-01 -5.87517381e-01
-2.40758732e-01 4.78509933e-01 3.69039565e-01 -1.18434882e+00
-8.39359522e-01 -2.02925965e-01 -6.59351110e-01 6.18777394e-01
-1.34454262e+00 -8.13519120e-01 1.13844082e-01 2.27317989e-01
7.36365318e-01 -8.38067308e-02 4.27146673e-01 -4.28001523e-01
-2.18497857e-01 -4.25479859e-01 -1.42524883e-01 -4.85907912e-01
-1.12321891e-01 -7.63116300e-01 6.79373324e-01 1.13466728e+00
-5.84569991e-01 -6.49245381e-02 1.44708383e+00 -1.14181471e+00
-1.82633734e+00 -1.19074929e+00 7.62682498e-01 -4.05608773e-01
4.81462896e-01 1.32742181e-01 -4.78721201e-01 1.03763044e+00
6.61937967e-02 -4.26708788e-01 -4.34996597e-02 -6.20514750e-01
3.99704009e-01 4.05738086e-01 -1.22899902e+00 1.10635352e+00
7.91467428e-01 7.04431161e-02 -5.68940580e-01 1.14043549e-01
5.71981251e-01 -6.31173491e-01 -1.42160997e-01 5.53136528e-01
4.04899091e-01 -6.67015553e-01 6.96329415e-01 -5.23649037e-01
-5.41329622e-01 -6.62786484e-01 -1.88775375e-01 -1.01619720e+00
-1.49382472e-01 -1.09810364e+00 4.71850261e-02 3.38138372e-01
3.72007310e-01 -5.64409018e-01 8.61828804e-01 1.19359064e+00
-1.83658957e-01 -2.40137979e-01 -1.56529140e+00 -7.59231329e-01
-3.45039785e-01 -8.86091411e-01 4.85970467e-01 2.52568014e-02
-1.30356818e-01 -3.25956307e-02 -5.05694449e-01 9.35649812e-01
8.62879336e-01 -3.08997110e-02 1.08607936e+00 -7.37831533e-01
2.48683058e-02 -7.33450353e-02 -3.51736158e-01 -1.05855501e+00
4.53230470e-01 -3.08717936e-01 8.63891482e-01 -1.68372476e+00
-1.42266080e-02 -4.18136001e-01 5.78982115e-01 4.63019937e-01
2.37728760e-01 -7.23629653e-01 4.02957439e-01 1.39593586e-01
-1.05429757e+00 5.99309623e-01 1.08224356e+00 -1.76461443e-01
-3.44894826e-01 5.12543142e-01 -7.88325071e-02 8.98804009e-01
7.81903267e-01 -6.87568605e-01 -5.53743601e-01 -2.25901484e-01
1.56360835e-01 1.00676775e+00 2.40999371e-01 -1.04043043e+00
5.18003583e-01 -1.01586211e+00 -7.05436945e-01 -1.08132720e+00
6.74358785e-01 -9.55787063e-01 8.63626003e-01 9.19568062e-01
-1.94595546e-01 6.90293238e-02 3.52486193e-01 1.14141393e+00
1.90221176e-01 -7.08387420e-02 6.20598495e-01 -3.74774635e-02
-1.16361439e+00 3.65428180e-01 -1.53629279e+00 -2.06718832e-01
1.65592003e+00 -1.62424549e-01 -8.04233253e-02 -9.74843383e-01
-7.76149154e-01 1.09069490e+00 5.72981060e-01 3.95488530e-01
5.31783402e-01 -9.06401992e-01 -5.82588851e-01 1.67743593e-01
-3.70041758e-01 -7.10952356e-02 4.19317663e-01 6.94743335e-01
-5.71393549e-01 5.71541190e-01 -3.59828115e-01 -4.20674831e-01
-1.26015520e+00 4.70279515e-01 4.28937614e-01 -4.06689167e-01
-5.68174541e-01 1.87020838e-01 7.53364936e-02 -3.77984852e-01
1.77753642e-01 -8.25002268e-02 -3.12800527e-01 -3.59600246e-01
6.03812218e-01 7.49756694e-01 -4.06254321e-01 -1.14727700e+00
-3.68725300e-01 1.77133977e-02 1.14676856e-01 -8.11585963e-01
1.00899971e+00 -5.99972188e-01 3.97575498e-01 1.42086804e-01
4.93498117e-01 -3.02539617e-01 -2.06932545e+00 1.17342085e-01
1.13654539e-01 -3.05122316e-01 1.73274383e-01 -4.33023244e-01
-3.36497277e-01 1.61930993e-01 3.11595529e-01 1.22734029e-02
3.20325613e-01 -7.45468540e-03 8.45691502e-01 5.02182841e-01
1.07678723e+00 -1.15611494e+00 -4.58923757e-01 9.82938349e-01
5.37560165e-01 -1.13830924e+00 -9.69861671e-02 -3.09596330e-01
-1.13898802e+00 6.09603167e-01 3.93199325e-01 -6.21756911e-02
4.38885778e-01 3.86040449e-01 7.99935237e-02 9.52687860e-03
-1.25513804e+00 -3.82427156e-01 -2.15767905e-01 5.92529118e-01
-1.16490841e+00 4.60644364e-01 -1.06109016e-01 1.43928438e-01
-1.13542661e-01 -1.04136743e-01 8.39023173e-01 1.27428472e+00
-9.88203108e-01 -8.29002440e-01 -5.92702627e-01 5.60804605e-02
-4.89441939e-02 5.41951895e-01 -1.03422530e-01 6.68430865e-01
-1.36710539e-01 1.30597365e+00 9.16422009e-02 -3.14693600e-01
1.45355061e-01 -5.69916889e-02 -6.23674989e-02 -3.94810617e-01
9.97173563e-02 -1.01545282e-01 6.73167169e-01 -9.96101022e-01
-1.71162382e-01 -1.13393819e+00 -1.60370886e+00 -4.01596993e-01
1.04923032e-01 -5.07033477e-03 5.99475205e-01 1.36917686e+00
3.94207478e-01 8.17636698e-02 5.28330445e-01 -1.20573497e+00
-7.45946646e-01 -2.75130540e-01 -8.46457854e-02 -2.15464547e-01
5.56431711e-01 -7.43021607e-01 -2.91255742e-01 -4.64014113e-01] | [4.996975421905518, 1.6919682025909424] |
79004082-60ba-4705-8982-2fe1686b4ca0 | fast-and-accurate-decoding-of-finger | null | null | https://iopscience.iop.org/article/10.1088/1741-2552/ac4ed1 | https://iopscience.iop.org/article/10.1088/1741-2552/ac4ed1/pdf | Fast and accurate decoding of finger movements from ECoG through Riemannian features and modern machine learning techniques | Objective.Accurate decoding of individual finger movements is crucial for advanced prosthetic control. In this work, we introduce the use of Riemannian-space features and temporal dynamics of electrocorticography (ECoG) signal combined with modern machine learning (ML) tools to improve the motor decoding accuracy at the level of individual fingers.
Approach. We selected a set of informative biomarkers that correlated with finger movements and evaluated the performance of state-of-the-art ML algorithms on the brain-computer interface (BCI) competition IV dataset (ECoG, three subjects) and a second ECoG dataset with a similar recording paradigm (Stanford, nine subjects). We further explored the temporal concatenation of features to effectively capture the history of ECoG signal, which led to a significant improvement over single-epoch decoding in both classification (p < 0.01) and regression tasks (p < 0.01).
Main results. Using feature concatenation and gradient boosted trees (the top-performing model), we achieved a classification accuracy of 77.0% in detecting individual finger movements (six-class task, including rest state), improving over the state-of-the-art conditional random fields by 11.7% on the three BCI competition subjects. In continuous decoding of movement trajectory, our approach resulted in an average Pearson's correlation coefficient (r) of 0.537 across subjects and fingers, outperforming both the BCI competition winner and the state-of-the-art approach reported on the same dataset (CNN + LSTM). Furthermore, our proposed method features a low time complexity, with only<17.2 s required for training and<50 ms for inference. This enables about 250× speed-up in training compared to previously reported deep learning method with state-of-the-art performance.
Significance.The proposed techniques enable fast, reliable, and high-performance prosthetic control through minimally-invasive cortical signals. | ['Mahsa Shoaran', 'Bingzhao Zhu', 'Lin Yao'] | 2022-02-25 | null | null | null | journal-of-neural-engineering-2022-2 | ['brain-decoding', 'brain-decoding'] | ['medical', 'miscellaneous'] | [ 3.95295233e-01 -1.89026564e-01 4.94658276e-02 8.82567316e-02
-9.34085608e-01 -2.61917800e-01 5.20826638e-01 -4.38554406e-01
-7.51583457e-01 1.08507848e+00 6.32016435e-02 -1.37031838e-01
-5.69277704e-01 -2.75414050e-01 -8.24480534e-01 -7.42087185e-01
-6.96156085e-01 1.48075655e-01 1.68638647e-01 -3.29250768e-02
3.95833880e-01 5.55942714e-01 -1.51326263e+00 4.10228789e-01
8.46226513e-01 1.42197740e+00 6.31880939e-01 3.96882206e-01
3.67017716e-01 2.86694705e-01 -6.50488198e-01 -8.62362981e-02
-6.72768950e-02 -3.22901875e-01 -4.39326525e-01 -6.68140709e-01
1.74634889e-01 -1.66771039e-01 -5.30825138e-01 6.61397219e-01
1.10268927e+00 -3.70540053e-01 6.99431419e-01 -7.97432721e-01
-2.95005023e-01 4.96046871e-01 -1.44474834e-01 2.59499699e-01
1.19664729e-01 4.16974872e-01 5.42573988e-01 -6.19123816e-01
7.49611318e-01 4.81199294e-01 6.33936822e-01 6.76605284e-01
-1.32500327e+00 -1.01616764e+00 -5.94467185e-02 5.72423577e-01
-1.30901361e+00 -3.68370980e-01 6.52710974e-01 -6.76024973e-01
1.34468615e+00 2.96655327e-01 1.22011840e+00 1.59739232e+00
7.25914776e-01 5.98051250e-01 1.36738086e+00 -3.52450520e-01
1.75938800e-01 -1.97605491e-01 -1.71366394e-01 1.87485859e-01
9.87639427e-02 3.31736833e-01 -7.85302103e-01 6.03256859e-02
9.33613598e-01 -2.74766952e-01 -4.15541738e-01 1.24382138e-01
-1.61059105e+00 3.14845175e-01 3.66261750e-01 8.43460739e-01
-9.99326885e-01 3.89037699e-01 2.70349860e-01 1.31726369e-01
2.58425027e-01 4.88426268e-01 -6.41608059e-01 -7.70529330e-01
-1.02452743e+00 1.62470713e-01 5.78917384e-01 6.76164031e-01
-8.29738155e-02 -3.68156098e-02 -7.72911668e-01 6.93856776e-01
9.08268914e-02 5.47599912e-01 6.00380421e-01 -6.80555105e-01
6.08112931e-01 3.01288456e-01 -6.10353462e-02 -6.25737131e-01
-8.04414630e-01 -9.22086060e-01 -9.24410820e-01 4.57012862e-01
5.76650381e-01 -2.60298133e-01 -7.69105613e-01 1.89939785e+00
-5.30996084e-01 9.22485907e-03 -5.24278462e-01 8.84833932e-01
5.47571220e-02 -1.25345930e-01 7.29998872e-02 -2.73453891e-01
1.24720287e+00 -1.07989006e-01 -6.94736958e-01 -1.72988877e-01
4.19235021e-01 -3.20641577e-01 8.27274799e-01 9.46138442e-01
-8.75610948e-01 -4.31711018e-01 -1.03063154e+00 6.95368409e-01
-8.84717479e-02 5.52083075e-01 7.83616960e-01 5.08981168e-01
-9.72161889e-01 1.08392048e+00 -1.22149694e+00 -3.50749135e-01
7.87407577e-01 1.01005423e+00 -6.44166231e-01 3.05284441e-01
-1.26315653e+00 1.33124685e+00 -7.37546533e-02 3.08090687e-01
-8.68531525e-01 -7.13785708e-01 8.95889327e-02 -2.61748552e-01
-3.07495385e-01 -4.59789276e-01 5.16678274e-01 -3.63541782e-01
-1.75281000e+00 8.19840670e-01 -9.43708420e-02 -3.85736972e-01
6.09218180e-01 -6.47610962e-01 -3.74638617e-01 2.52429061e-02
-1.63699865e-01 4.41971183e-01 7.41649806e-01 -5.48996687e-01
-2.05581814e-01 -9.03867781e-01 -5.40366471e-01 -2.62400985e-01
-2.89210379e-01 7.86532313e-02 -1.64833993e-01 -6.78857446e-01
2.05708891e-01 -9.17489231e-01 6.30693361e-02 -2.25165367e-01
-3.44073683e-01 -8.56227651e-02 2.24655420e-02 -1.16350341e+00
1.26491165e+00 -1.96469295e+00 5.53828597e-01 2.94419676e-01
3.25410724e-01 1.71066999e-01 -8.52794796e-02 3.48493576e-01
-1.72847211e-01 -2.20812783e-01 -2.63903379e-01 6.40032142e-02
-1.10401794e-01 -3.74474347e-01 1.20126635e-01 6.64074540e-01
1.69373885e-01 1.22527611e+00 -4.52415437e-01 -6.62278011e-03
2.92268217e-01 6.68395400e-01 -4.27995384e-01 1.05429396e-01
2.13839844e-01 9.40451145e-01 -1.48899317e-01 4.89318371e-01
2.55984485e-01 1.35194376e-01 2.47508407e-01 -5.08658588e-01
-1.62075952e-01 3.58160764e-01 -8.15781236e-01 2.19923377e+00
-4.98358697e-01 8.27310801e-01 -2.34351113e-01 -1.02484274e+00
9.36416209e-01 4.57590014e-01 7.44878054e-01 -9.64386702e-01
6.19877398e-01 5.03051460e-01 5.84225416e-01 -5.30963123e-01
-6.24994457e-01 4.82155429e-03 1.01853617e-01 1.07003368e-01
3.89239848e-01 2.87639439e-01 -2.61680454e-01 -2.18666792e-01
1.40175092e+00 5.52226722e-01 -6.14342652e-02 -2.98131227e-01
2.54316390e-01 -3.74868691e-01 7.85148591e-02 9.34157252e-01
-1.12278424e-01 4.91873980e-01 4.14441317e-01 9.05368179e-02
-4.17916685e-01 -1.02730322e+00 -4.09594208e-01 6.34168804e-01
-2.45196566e-01 -2.68170655e-01 -8.31459165e-01 4.77221012e-02
2.01216385e-01 5.44426739e-01 -7.99999952e-01 -4.47572440e-01
-6.28371477e-01 -6.83009923e-01 6.02845788e-01 6.75850391e-01
3.37196887e-01 -1.42365932e+00 -8.37205529e-01 6.55133784e-01
-1.67339176e-01 -1.13731611e+00 3.41218680e-01 5.69331467e-01
-1.13260925e+00 -8.71060133e-01 -9.61449802e-01 -5.10078192e-01
-2.20177555e-03 -8.39253128e-01 4.69413102e-01 -6.62317932e-01
-6.76264524e-01 2.18941689e-01 -1.83324888e-01 -3.13211471e-01
2.67972171e-01 8.24837536e-02 2.21963033e-01 -8.81161243e-02
4.41456705e-01 -1.14636874e+00 -6.73272252e-01 -3.06589026e-02
-7.18080178e-02 3.32597420e-02 1.19781816e+00 8.75390232e-01
4.75200832e-01 -5.44637203e-01 7.98522174e-01 -3.17446403e-02
6.95887506e-01 -2.30656326e-01 -2.35395744e-01 2.48298943e-01
-6.76518500e-01 1.26075342e-01 5.61590314e-01 -8.12119484e-01
-4.83117461e-01 4.96851541e-02 -2.36660484e-02 -2.65850127e-01
-1.01893686e-01 1.96295679e-01 -8.00521523e-02 -2.82128930e-01
6.98977888e-01 5.52048802e-01 8.30950737e-02 -6.08410478e-01
2.60046986e-03 8.04999709e-01 7.36568689e-01 -4.45866883e-01
9.81469005e-02 1.54845193e-01 1.69388086e-01 -7.41082788e-01
6.87368214e-03 -4.44353223e-02 -1.05125928e+00 -4.70759153e-01
8.86447906e-01 -6.80748522e-01 -1.03236711e+00 1.00067449e+00
-1.29169548e+00 -4.64106500e-01 2.93256313e-01 1.19315624e+00
-8.63382578e-01 -1.18963234e-01 -3.22098404e-01 -8.60068321e-01
-6.70410752e-01 -9.12837684e-01 8.80018473e-01 -3.52384090e-01
-5.02797425e-01 -3.55138779e-01 4.75021079e-02 -1.31701455e-01
7.03661263e-01 5.32253623e-01 6.20411813e-01 -4.31116641e-01
-2.40699440e-01 -5.69022655e-01 1.54952321e-03 2.77749538e-01
-2.20095981e-02 -6.71091676e-01 -1.07806206e+00 -2.21790597e-01
-9.17986333e-02 -8.97625759e-02 6.03283763e-01 6.88777924e-01
1.15325427e+00 2.20224991e-01 -5.39045274e-01 5.22748947e-01
1.27084637e+00 3.68127763e-01 1.05418205e+00 1.52810976e-01
4.48070914e-01 4.03992563e-01 1.27358660e-01 4.01887894e-01
-1.81361437e-01 1.08755374e+00 2.74084270e-01 2.79610097e-01
-2.46526614e-01 4.36621904e-02 3.38274866e-01 6.91846848e-01
-7.62405932e-01 3.75640333e-01 -7.15914309e-01 4.49371040e-01
-1.72465575e+00 -7.75879383e-01 -2.11867616e-01 2.48499060e+00
7.95972407e-01 3.35671961e-01 2.86040250e-02 4.49362934e-01
4.87360179e-01 -3.83945167e-01 -8.14595401e-01 -3.37398751e-03
-4.66633365e-02 1.10904980e+00 5.14380932e-01 1.32457420e-01
-7.34394372e-01 8.61614108e-01 5.61722183e+00 6.80062771e-01
-1.42043579e+00 4.09466863e-01 -1.90048575e-01 -5.61462462e-01
4.69561011e-01 -3.81859273e-01 -6.02813303e-01 7.98516154e-01
1.24876726e+00 1.06393352e-01 9.43727434e-01 3.07878643e-01
2.75404900e-01 -1.26482248e-01 -1.09271014e+00 1.41307878e+00
-7.41023645e-02 -1.23217463e+00 -5.55139780e-01 2.25407675e-01
2.24076182e-01 5.47553003e-01 -2.38678053e-01 2.06032932e-01
-6.93956316e-01 -1.22120988e+00 8.82306874e-01 1.09309113e+00
1.35712898e+00 -2.81433880e-01 5.62835455e-01 3.11401188e-01
-1.03291368e+00 -2.86815524e-01 1.19251668e-01 -1.96450248e-01
6.14824370e-02 3.79132122e-01 -2.05836967e-01 4.52700943e-01
9.12359893e-01 8.98812294e-01 -5.62260985e-01 1.00162351e+00
-2.70958811e-01 7.17241287e-01 -4.59177554e-01 -5.11205316e-01
-1.91444322e-01 1.81989118e-01 6.30634606e-01 1.14498210e+00
4.40651506e-01 5.52571826e-02 -7.37619102e-01 9.98684883e-01
3.16750556e-01 -1.80262208e-01 -3.54137242e-01 -1.19537354e-01
3.73917669e-01 9.53401148e-01 -2.50232369e-01 4.63912375e-02
-1.28793836e-01 1.18322086e+00 2.24488318e-01 3.40381235e-01
-7.10609078e-01 -8.99095476e-01 3.92468542e-01 7.19430819e-02
1.78228796e-01 -3.62109393e-01 -7.98600376e-01 -9.62319195e-01
6.62302256e-01 -5.97054064e-01 -2.20558986e-01 -6.57961488e-01
-1.00440419e+00 7.23826528e-01 -1.31123632e-01 -1.28857505e+00
-4.45721358e-01 -1.10302925e+00 -3.14739913e-01 1.24241114e+00
-9.03831720e-01 -8.37559402e-01 -3.50536376e-01 9.32095408e-01
-9.15830508e-02 -2.60606289e-01 1.24760711e+00 4.32567745e-01
-2.52808869e-01 5.40930867e-01 1.53248236e-01 -5.13979718e-02
5.61670423e-01 -8.83511961e-01 1.50146306e-01 5.94081104e-01
-1.28053159e-01 8.60384941e-01 2.17329562e-01 -4.97888207e-01
-1.57172179e+00 -5.79785764e-01 8.48348081e-01 -3.18615437e-01
5.14992416e-01 -6.73417509e-01 -4.01552022e-01 2.99541354e-01
-2.12766021e-01 -3.02322656e-01 4.14423794e-01 4.28553447e-02
5.17289601e-02 -3.42347205e-01 -1.04785717e+00 5.12259007e-01
1.58472836e+00 -6.40759706e-01 -7.66593993e-01 1.97835684e-01
-8.37097540e-02 -4.18023616e-02 -1.13640749e+00 6.17138505e-01
1.42092836e+00 -8.15299213e-01 8.28556001e-01 -5.48705876e-01
1.86364874e-01 1.84385434e-01 -1.78164318e-01 -1.22222507e+00
-5.30172825e-01 -6.68911457e-01 -3.85642976e-01 7.55302727e-01
3.36230278e-01 -8.13038766e-01 4.49634999e-01 4.65718448e-01
-1.44726798e-01 -1.19930041e+00 -1.18129694e+00 -9.74677742e-01
1.63786054e-01 -9.48535681e-01 -3.37182656e-02 1.42296478e-01
4.67464626e-01 1.32684233e-02 -3.90498638e-01 -3.15211058e-01
6.35615408e-01 -2.92270482e-01 1.29818559e-01 -1.38248420e+00
-3.33370507e-01 -6.27726018e-01 -9.38758731e-01 -6.23105466e-01
7.48973787e-02 -1.21038866e+00 -6.11500815e-02 -1.79507732e+00
3.89202125e-02 -5.00869080e-02 -7.35614061e-01 6.45148158e-01
1.23552494e-01 3.23490769e-01 2.08970290e-02 2.58657545e-01
1.06385276e-01 4.57440466e-01 9.61549282e-01 -9.64427963e-02
-3.37850034e-01 9.64881331e-02 -5.41179836e-01 2.42511816e-02
9.53258753e-01 -4.63588208e-01 7.13613629e-02 -1.10502645e-01
-2.59824306e-01 5.06460518e-02 6.63604498e-01 -1.60060644e+00
3.22766691e-01 6.32002771e-01 8.80089164e-01 -3.45799059e-01
4.79325235e-01 -5.39029121e-01 2.55309016e-01 9.25731361e-01
-1.89429551e-01 -4.46415275e-01 3.74838173e-01 3.07949692e-01
1.08516566e-01 3.90486658e-01 7.17378914e-01 1.19022422e-01
-4.68667626e-01 1.36334643e-01 -5.77523649e-01 -1.05257563e-01
8.97914410e-01 -4.31257725e-01 -3.06839570e-02 3.93966287e-02
-1.01710713e+00 -4.44201738e-01 -3.37663800e-01 5.80505192e-01
5.61352849e-01 -1.28147209e+00 -7.83598661e-01 4.16170567e-01
2.80147493e-02 -9.06273007e-01 2.50423819e-01 1.56298637e+00
-8.76498595e-02 8.88526857e-01 -9.18909252e-01 -8.06581616e-01
-8.38017404e-01 -9.45646837e-02 4.51831490e-01 2.33964669e-03
-9.05394912e-01 9.02903199e-01 -3.53952587e-01 1.39699141e-02
6.22375906e-01 -3.13917428e-01 -2.66222298e-01 -4.58951741e-02
3.73965055e-01 3.86028469e-01 5.46886146e-01 -2.16329396e-01
-9.08862174e-01 8.33405793e-01 3.00802290e-01 -4.37926024e-01
1.66770136e+00 3.82906169e-01 -9.37940180e-03 5.72489023e-01
1.20308936e+00 -3.20499539e-01 -1.25475729e+00 2.96120375e-01
5.79087017e-03 -8.73515531e-02 2.07935467e-01 -1.66318583e+00
-9.11452353e-01 1.17240632e+00 1.42557228e+00 -5.90588212e-01
1.19693923e+00 -9.83725935e-02 5.72862744e-01 1.98032245e-01
1.18183017e+00 -8.66576672e-01 -3.99607599e-01 9.30898115e-02
1.37533498e+00 -7.09212720e-01 -1.10346429e-01 1.36010215e-01
-5.57657599e-01 1.16149926e+00 3.59353840e-01 -3.80298674e-01
9.24795806e-01 2.74391592e-01 -3.35223019e-01 -1.89751074e-01
-4.15239185e-01 -1.45353451e-01 6.80402339e-01 7.32700586e-01
5.25675237e-01 3.80908459e-01 -7.69515455e-01 1.25565517e+00
-1.79918602e-01 6.81273639e-01 -3.45721155e-01 8.62806380e-01
-1.60847791e-02 -8.27814341e-01 3.79283838e-02 9.99368250e-01
-4.98972863e-01 -2.88095683e-01 -2.36665830e-01 9.47168231e-01
1.77059591e-01 8.83969486e-01 -1.59605861e-01 -1.02231991e+00
5.61243594e-01 3.51127595e-01 1.19393444e+00 -2.77499080e-01
-8.06489766e-01 -5.24119101e-02 -6.35863617e-02 -1.20751405e+00
-1.62708044e-01 -7.46260762e-01 -1.12129629e+00 2.80762196e-01
-2.57549554e-01 -3.85359824e-01 1.08819497e+00 1.06035924e+00
6.49311423e-01 7.71256328e-01 2.61115789e-01 -1.29485035e+00
-5.33686340e-01 -1.63154590e+00 -8.53373587e-01 4.18139063e-02
-4.27297652e-02 -1.10534728e+00 -3.76470804e-01 -2.07452431e-01] | [12.960762977600098, 3.398362159729004] |
711258b1-3738-44df-b995-647b768f69d9 | robust-video-segment-proposals-with-painless | null | null | http://openaccess.thecvf.com/content_cvpr_2015/html/Wu_Robust_Video_Segment_2015_CVPR_paper.html | http://openaccess.thecvf.com/content_cvpr_2015/papers/Wu_Robust_Video_Segment_2015_CVPR_paper.pdf | Robust Video Segment Proposals With Painless Occlusion Handling | We propose a robust algorithm to generate video segment proposals. The proposals generated by our method can start from any frame in the video and are robust to complete occlusions. Our method does not assume specific motion models and even has a limited capability to generalize across videos. We build on our previous least squares tracking framework, where image segment proposals are generated and tracked using learned appearance models. The innovation in our new method lies in the use of two efficient moves, the merge move and free addition, to efficiently start segments from any frame and track them through complete occlusions, without much additional computation. Segment size interpolation is used for effectively detecting occlusions. We propose a new metric for evaluating video segment proposals on the challenging VSB-100 benchmark and present state-of-the-art results. Preliminary results are also shown for the potential use of our framework to track segments across different videos. | ['Zhengyang Wu', 'Rahul Sukthankar', 'James M. Rehg', 'Fuxin Li'] | 2015-06-01 | null | null | null | cvpr-2015-6 | ['occlusion-handling'] | ['computer-vision'] | [ 1.98539749e-01 -2.56802917e-01 -4.75926042e-01 -3.84423405e-01
-8.68862510e-01 -6.44530892e-01 5.82245648e-01 -1.43869892e-01
-2.29153365e-01 5.51286876e-01 -1.77870348e-01 -5.17763980e-02
3.67237896e-01 -2.75743067e-01 -9.59273100e-01 -5.73138595e-01
-6.09925240e-02 4.21869844e-01 1.05407488e+00 -5.99002466e-02
1.86502561e-01 6.21289849e-01 -1.35517383e+00 3.13768566e-01
5.28498113e-01 9.42580402e-01 1.51697412e-01 7.98818886e-01
5.98987564e-03 8.51721764e-01 -3.39294195e-01 -7.02297017e-02
5.11664450e-01 -2.87397712e-01 -8.16639006e-01 6.21477306e-01
1.03105068e+00 -6.32367253e-01 -4.66888785e-01 5.53758621e-01
2.46784389e-01 7.92379081e-02 5.27516901e-01 -1.27364564e+00
1.05499908e-01 2.35897750e-01 -7.90477812e-01 2.58056521e-01
6.47624135e-01 1.53553709e-01 5.44379413e-01 -9.41621065e-01
1.16438365e+00 9.79563951e-01 9.19989586e-01 5.75742841e-01
-1.42062986e+00 -5.44359446e-01 5.15094638e-01 2.04444110e-01
-1.42478740e+00 -7.19600797e-01 5.25203407e-01 -6.56627595e-01
8.01653564e-01 3.76042753e-01 7.71375179e-01 7.69282639e-01
-1.79531071e-02 1.15109468e+00 6.44527972e-01 -5.63820601e-01
-2.43239999e-02 -4.32047620e-02 1.19763643e-01 9.28097486e-01
1.00121863e-01 7.08967149e-02 -3.24125558e-01 -1.56994998e-01
1.14235580e+00 -1.24115162e-01 -2.91628152e-01 -9.96085525e-01
-1.37305462e+00 5.85901260e-01 2.00377882e-01 3.97333838e-02
-1.86276555e-01 5.23204267e-01 4.45681274e-01 3.57563645e-02
2.46881589e-01 -9.44589153e-02 -3.81957412e-01 4.73050736e-02
-1.66366816e+00 3.91945571e-01 5.93527377e-01 1.39103639e+00
7.21452892e-01 3.02786753e-02 -3.63757402e-01 6.48392379e-01
5.05320609e-01 3.54422957e-01 8.60695466e-02 -1.51292408e+00
2.91972548e-01 2.15174496e-01 4.05891776e-01 -7.62270212e-01
-3.15675199e-01 -1.41036093e-01 -1.14550643e-01 4.93460745e-01
4.88921821e-01 -1.51151299e-01 -1.01081967e+00 1.41219044e+00
5.19437373e-01 6.89629912e-01 -5.41004658e-01 6.35523260e-01
7.44881809e-01 5.61225593e-01 2.85385381e-02 -3.39772195e-01
1.03674293e+00 -1.36349165e+00 -3.78934354e-01 -3.42956744e-02
4.41608638e-01 -1.02054751e+00 2.61024117e-01 4.24622536e-01
-1.53328454e+00 -7.63278604e-01 -9.78390634e-01 -4.15562876e-02
2.42481783e-01 2.44406864e-01 4.73571777e-01 5.30779183e-01
-1.22468448e+00 5.63952088e-01 -1.06791377e+00 -4.06102806e-01
5.29182792e-01 5.71398556e-01 -3.86960745e-01 1.09146321e-02
-4.66230839e-01 7.74175823e-01 3.21897328e-01 1.29011357e-02
-8.56659949e-01 -5.84748626e-01 -8.31885517e-01 -2.10994661e-01
2.34908238e-01 -9.29646790e-01 1.23036289e+00 -1.45511401e+00
-1.33349538e+00 7.57878006e-01 -7.01523721e-01 -5.03621101e-01
8.47214162e-01 -1.35846734e-01 -8.26574340e-02 3.87525290e-01
-5.34386784e-02 1.21340084e+00 9.52646255e-01 -1.32567060e+00
-9.40105677e-01 3.50558639e-01 -5.11101820e-02 2.70749740e-02
1.25393391e-01 4.49751079e-01 -1.19178092e+00 -8.61621380e-01
1.39187276e-01 -1.33858132e+00 -3.86648357e-01 5.11056006e-01
-8.96215588e-02 -1.98462754e-01 9.41865623e-01 -6.57384276e-01
1.29740071e+00 -1.82572615e+00 8.36307555e-02 2.54788935e-01
-2.09965762e-02 4.45419908e-01 -2.79728830e-01 1.12525128e-01
-1.64625086e-02 -2.35427797e-01 -1.21484928e-01 -5.51824868e-01
-1.97313890e-01 1.90031931e-01 -6.32058159e-02 6.60065055e-01
2.02800795e-01 7.50893772e-01 -8.63400578e-01 -9.31514680e-01
4.18723553e-01 4.97220069e-01 -5.87346435e-01 -9.29519907e-02
-1.98467165e-01 3.28603268e-01 -3.28963727e-01 6.95039153e-01
7.38613665e-01 -2.24561945e-01 -1.75002068e-01 -2.90263027e-01
-4.51099873e-01 1.16137275e-02 -1.50852334e+00 1.93293989e+00
8.78370702e-02 9.86991465e-01 3.23798992e-02 -7.99705565e-01
5.32427490e-01 4.05156940e-01 8.48997295e-01 -1.46346495e-01
-1.63691770e-02 8.03845301e-02 -2.19307572e-01 -3.71872634e-01
4.69698101e-01 3.18749547e-01 3.93462092e-01 2.57542133e-01
1.05592750e-01 1.17333248e-01 7.87091374e-01 2.60447741e-01
9.01386559e-01 8.19639027e-01 2.44347900e-01 -1.88488483e-01
6.37297511e-01 2.41005525e-01 8.01210046e-01 8.37944984e-01
-5.12449265e-01 1.05316603e+00 2.28136599e-01 -6.55768692e-01
-1.18685865e+00 -1.23030007e+00 -3.12766939e-01 1.07154620e+00
4.48404431e-01 -2.31004730e-01 -6.76834166e-01 -8.01142156e-01
-1.72493786e-01 2.41183892e-01 -4.85204190e-01 5.44733226e-01
-1.07603848e+00 -2.41653726e-01 1.41343191e-01 7.51416922e-01
-4.79842722e-02 -1.08961070e+00 -5.68946004e-01 4.36891377e-01
-3.53896558e-01 -1.13792646e+00 -8.28663111e-01 -2.58624852e-01
-9.94558454e-01 -1.21530616e+00 -9.64991271e-01 -1.12238240e+00
8.38399887e-01 4.66470093e-01 1.22835803e+00 2.52208263e-01
-5.05424142e-01 5.98395646e-01 -2.19960436e-01 -1.39720976e-01
-4.79289532e-01 -2.08644122e-01 -1.67402267e-01 -1.43594950e-01
9.18354318e-02 7.68770278e-02 -9.53300536e-01 7.97027171e-01
-6.22344732e-01 1.71747208e-01 3.45882207e-01 5.49051046e-01
8.65811646e-01 -6.13663077e-01 2.54503012e-01 -6.66983783e-01
-2.44202286e-01 -2.39191294e-01 -8.48754704e-01 3.76664758e-01
-8.99414942e-02 -2.35299826e-01 1.06083825e-01 -4.80847299e-01
-9.96073008e-01 6.93885088e-01 -2.79915985e-02 -6.64040983e-01
-1.59484237e-01 -1.59175619e-01 2.44055033e-01 -5.92034459e-01
4.85451430e-01 -5.14268596e-03 -1.04437023e-02 -2.19751745e-01
4.98467475e-01 1.29239559e-01 6.54423892e-01 -4.15116727e-01
9.02529418e-01 8.11375678e-01 -1.48872593e-02 -9.21954036e-01
-2.11019754e-01 -1.03104854e+00 -9.40834999e-01 -4.94128913e-01
7.31533945e-01 -1.06765974e+00 -4.15819913e-01 2.78527737e-01
-1.24678349e+00 -3.70723695e-01 -2.07518935e-01 5.22383630e-01
-9.25766170e-01 8.63694429e-01 -7.28412271e-01 -3.67089987e-01
-2.95872003e-01 -1.38464153e+00 1.06578231e+00 2.06596628e-01
-4.38545227e-01 -1.00098979e+00 1.96813643e-01 1.34271562e-01
2.27612361e-01 4.06642884e-01 1.71863899e-01 -2.13073328e-01
-1.21419418e+00 -1.86813250e-01 -3.09093088e-01 1.65208995e-01
5.82774580e-02 6.23478949e-01 -6.16593301e-01 -6.60591304e-01
-4.13423777e-01 -5.63583039e-02 8.82249713e-01 8.43392551e-01
8.07563186e-01 -2.30708495e-02 -7.81894207e-01 7.10623264e-01
1.49316633e+00 1.87318474e-01 7.29264915e-01 4.52802241e-01
7.24495530e-01 3.67036194e-01 8.56956899e-01 3.20372730e-01
1.47073120e-01 1.23339581e+00 1.73273742e-01 -3.88063312e-01
-4.06591356e-01 1.69407025e-01 4.79887158e-01 4.95224953e-01
-4.16105539e-01 -6.89130500e-02 -6.05747640e-01 6.71666563e-01
-2.00851536e+00 -1.26789713e+00 -4.53345388e-01 2.20956087e+00
6.25404954e-01 9.91179347e-02 5.09768367e-01 -2.51787364e-01
7.83432543e-01 -1.74316708e-02 -3.91573459e-01 -1.67235121e-01
2.95671765e-02 1.85210973e-01 6.67032719e-01 4.64841425e-01
-1.53389990e+00 9.74249184e-01 7.66495466e+00 8.65023911e-01
-9.25852716e-01 -7.30527714e-02 4.39104527e-01 -3.71784389e-01
7.02999011e-02 2.63042718e-01 -1.04438102e+00 3.20335567e-01
3.74793828e-01 1.40247613e-01 1.33924916e-01 8.14395189e-01
4.58617300e-01 -1.24622934e-01 -1.08222425e+00 7.99890041e-01
2.48276874e-01 -1.56090200e+00 -2.30752990e-01 -1.94450900e-01
1.16609657e+00 2.23715618e-01 -1.00871481e-01 -1.40063271e-01
1.86659753e-01 -4.92968410e-01 1.09384632e+00 5.73407829e-01
5.14608324e-01 -4.36633676e-01 3.46118271e-01 -1.14872791e-01
-1.45051384e+00 1.84751034e-01 -3.72686505e-01 3.30063522e-01
6.66766524e-01 -1.20743206e-02 -8.85386765e-01 1.13749668e-01
4.18570876e-01 7.34942317e-01 -6.36195779e-01 1.82427931e+00
1.04299728e-02 4.81396824e-01 -4.34565037e-01 3.93368363e-01
1.42248064e-01 -1.59268200e-01 5.95931649e-01 1.57455122e+00
3.79837632e-01 -1.99090421e-01 5.48307180e-01 4.26072538e-01
1.99686199e-01 1.82250455e-01 -2.70118892e-01 7.03301907e-01
3.45384330e-01 1.24415672e+00 -1.09146142e+00 -6.41793609e-01
-8.13363731e-01 9.51612830e-01 -4.07701693e-02 5.59951842e-01
-1.22164321e+00 -1.59193814e-01 4.73120451e-01 3.76311898e-01
9.34548855e-01 -2.77430952e-01 1.05532795e-01 -1.13952851e+00
5.31200366e-03 -7.79803157e-01 2.70052195e-01 -6.56295955e-01
-9.39699173e-01 4.04733479e-01 1.13361411e-01 -1.68013704e+00
-3.72041255e-01 -4.20591503e-01 -6.79211676e-01 5.61256468e-01
-1.50784290e+00 -1.36766720e+00 -2.75478333e-01 6.18696511e-01
9.86393809e-01 1.21022396e-01 3.55121464e-01 4.27094668e-01
-3.39807689e-01 5.49747705e-01 2.14409649e-01 2.11766362e-01
9.48603451e-01 -1.01733243e+00 6.71675503e-01 1.06342280e+00
3.89757395e-01 6.59910738e-01 7.89330363e-01 -6.30643249e-01
-9.13710713e-01 -1.06375396e+00 6.98533118e-01 -5.81545115e-01
4.51141059e-01 -1.83105126e-01 -7.95588553e-01 9.49239314e-01
3.03967625e-01 4.28044349e-01 4.17774290e-01 -2.42516041e-01
-1.93525419e-01 6.46936446e-02 -8.94373596e-01 4.80995536e-01
9.81054306e-01 -8.39283783e-03 -4.70281243e-01 3.21643710e-01
4.17816579e-01 -6.41962826e-01 -6.26148283e-01 4.20654982e-01
9.39415038e-01 -1.02865624e+00 1.34510517e+00 -5.05190730e-01
1.08508412e-02 -8.01224291e-01 1.91258162e-01 -7.04374969e-01
-4.92608309e-01 -9.27735209e-01 -2.34524220e-01 1.22365546e+00
3.35004568e-01 2.46336162e-02 1.16622102e+00 7.22964168e-01
-2.09711671e-01 -5.55840671e-01 -9.59426880e-01 -8.94848168e-01
-2.38108650e-01 -2.78790474e-01 5.61847910e-02 5.68342030e-01
-1.30539268e-01 -2.74845421e-01 -4.68361259e-01 -4.54456732e-03
7.98247695e-01 2.39435017e-01 1.05815113e+00 -1.01955712e+00
-3.72106880e-01 -4.45773512e-01 -6.07916236e-01 -1.46827030e+00
-2.16309894e-02 -5.42296827e-01 2.30457857e-01 -1.56406188e+00
2.30417013e-01 -6.04196489e-01 -2.43781269e-01 3.62675190e-01
-1.01247981e-01 6.57089472e-01 5.09737253e-01 2.71142930e-01
-1.07229447e+00 -1.87525731e-02 1.10840964e+00 -6.54857531e-02
-2.56810695e-01 2.04384893e-01 -3.77715379e-02 9.84223008e-01
4.42222863e-01 -5.30031741e-01 -2.21802354e-01 -3.80869031e-01
-3.02517682e-01 4.68454510e-02 2.59262651e-01 -1.11816394e+00
2.20175877e-01 -2.90389010e-03 6.20029688e-01 -7.78207242e-01
3.70705128e-01 -7.76257455e-01 4.08601582e-01 5.30779779e-01
2.30650743e-03 1.55299708e-01 2.10259348e-01 5.20769894e-01
1.95954014e-02 -5.08638084e-01 9.83399630e-01 -4.09023315e-02
-1.05253720e+00 3.88139457e-01 -3.84101301e-01 -1.67594969e-01
1.38240719e+00 -6.43520057e-01 -3.06658428e-02 -1.04469195e-01
-9.30209398e-01 2.05051169e-01 8.84958744e-01 4.96571630e-01
5.24625242e-01 -1.35331249e+00 -7.52294481e-01 5.72007224e-02
3.80549692e-02 -3.15925688e-01 1.44697890e-01 9.24309492e-01
-1.08756006e+00 5.29655993e-01 -1.90184996e-01 -1.06650233e+00
-1.72377372e+00 7.46048927e-01 4.87918742e-02 -3.66850011e-02
-5.48663259e-01 7.48958826e-01 3.29794109e-01 3.34108531e-01
2.28052974e-01 -3.74902964e-01 -1.47294745e-01 5.01261726e-02
6.20053649e-01 5.65520883e-01 -1.47102982e-01 -1.12157965e+00
-3.47024262e-01 1.03040206e+00 -2.14727581e-01 4.76851687e-02
1.04825211e+00 -2.77682662e-01 2.66434193e-01 1.38306752e-01
1.05665791e+00 6.30630106e-02 -1.84444201e+00 -7.03468248e-02
-8.65656883e-02 -7.11159766e-01 -2.26328045e-01 -3.28629494e-01
-1.12284577e+00 4.14876699e-01 7.23281860e-01 -9.74985063e-02
8.86047721e-01 -5.64186722e-02 8.14637423e-01 -2.86999717e-02
5.73111355e-01 -9.50880527e-01 -6.81577772e-02 2.84826100e-01
5.82479060e-01 -1.28150308e+00 2.60037422e-01 -6.06405556e-01
-4.15807664e-01 1.25337005e+00 4.19718295e-01 -2.64840335e-01
3.53668571e-01 2.97327280e-01 2.60417372e-01 1.94855854e-01
-5.49416006e-01 -1.36521056e-01 5.19443870e-01 6.73424244e-01
5.88113487e-01 -4.16787654e-01 -4.54020739e-01 -3.13822985e-01
6.07888401e-01 2.14558169e-01 5.36198497e-01 9.56160545e-01
-5.68133175e-01 -1.39330637e+00 -5.27669132e-01 2.42292792e-01
-4.31548387e-01 1.77107260e-01 7.75875673e-02 9.08028245e-01
1.53886393e-01 7.85082936e-01 2.46378016e-02 2.44256303e-01
1.84619814e-01 -8.47033039e-02 7.48991370e-01 -3.94527346e-01
-3.37334991e-01 5.76410353e-01 -6.23381846e-02 -8.83517981e-01
-9.60961103e-01 -1.25670159e+00 -1.06506848e+00 4.51442413e-02
-7.05822885e-01 -5.45556769e-02 5.96193671e-01 6.57067060e-01
2.04939678e-01 3.24726760e-01 4.54219013e-01 -1.35095811e+00
5.28026111e-02 -5.24852157e-01 -1.49448246e-01 4.96751845e-01
4.39176142e-01 -7.61343777e-01 -9.10155848e-03 7.73137987e-01] | [9.073062896728516, -0.2648884654045105] |
9d40cc97-b169-4a9b-8316-0c031172d481 | learning-to-track-with-object-permanence | 2103.14258 | null | https://arxiv.org/abs/2103.14258v2 | https://arxiv.org/pdf/2103.14258v2.pdf | Learning to Track with Object Permanence | Tracking by detection, the dominant approach for online multi-object tracking, alternates between localization and association steps. As a result, it strongly depends on the quality of instantaneous observations, often failing when objects are not fully visible. In contrast, tracking in humans is underlined by the notion of object permanence: once an object is recognized, we are aware of its physical existence and can approximately localize it even under full occlusions. In this work, we introduce an end-to-end trainable approach for joint object detection and tracking that is capable of such reasoning. We build on top of the recent CenterTrack architecture, which takes pairs of frames as input, and extend it to videos of arbitrary length. To this end, we augment the model with a spatio-temporal, recurrent memory module, allowing it to reason about object locations and identities in the current frame using all the previous history. It is, however, not obvious how to train such an approach. We study this question on a new, large-scale, synthetic dataset for multi-object tracking, which provides ground truth annotations for invisible objects, and propose several approaches for supervising tracking behind occlusions. Our model, trained jointly on synthetic and real data, outperforms the state of the art on KITTI and MOT17 datasets thanks to its robustness to occlusions. | ['Adrien Gaidon', 'Wolfram Burgard', 'Jie Li', 'Pavel Tokmakov'] | 2021-03-26 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Tokmakov_Learning_To_Track_With_Object_Permanence_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Tokmakov_Learning_To_Track_With_Object_Permanence_ICCV_2021_paper.pdf | iccv-2021-1 | ['online-multi-object-tracking'] | ['computer-vision'] | [-1.12325832e-01 -2.99829483e-01 1.60330925e-02 7.87270591e-02
-4.30639207e-01 -8.04106891e-01 6.03020370e-01 1.57769680e-01
-6.29353762e-01 6.56087816e-01 -1.95344225e-01 3.79578546e-02
6.85559362e-02 -4.79453564e-01 -1.15140259e+00 -4.69251305e-01
-1.45888001e-01 7.14097977e-01 1.02402699e+00 8.34989920e-02
-2.28331104e-01 5.58759868e-01 -1.86190665e+00 2.27182105e-01
2.42953748e-01 8.72032523e-01 3.09739769e-01 1.01712871e+00
2.22892627e-01 1.01509631e+00 -4.92934436e-01 -4.10366327e-01
6.29722401e-02 -8.36299509e-02 -6.03691638e-01 3.42728645e-01
9.95722711e-01 -4.25204158e-01 -4.10085618e-01 8.04037929e-01
1.99749753e-01 -1.39075173e-02 2.53636718e-01 -1.39588284e+00
-5.00115573e-01 3.14347118e-01 -4.09123719e-01 4.71685678e-01
3.58831614e-01 4.10915583e-01 9.30865347e-01 -6.72924042e-01
1.01526225e+00 1.26905203e+00 8.40177178e-01 6.66548193e-01
-1.14911664e+00 -3.32384616e-01 5.76868474e-01 8.54726359e-02
-1.11894739e+00 -5.28086782e-01 3.37767005e-01 -6.88265264e-01
6.66150212e-01 1.16061151e-01 8.50943208e-01 1.34434867e+00
1.57711551e-01 9.72862542e-01 8.91255081e-01 -3.33037198e-01
1.12091921e-01 2.01970875e-01 1.27971396e-01 7.30861604e-01
5.17769396e-01 4.13459867e-01 -5.27033865e-01 -5.42991161e-02
5.97175658e-01 2.29301780e-01 -1.94421545e-01 -7.14209318e-01
-1.44295418e+00 2.91714132e-01 4.46949214e-01 5.28224766e-01
-4.46111619e-01 4.59739357e-01 1.39647067e-01 2.78481990e-01
2.74049848e-01 -6.97629750e-02 -4.28541332e-01 9.14491490e-02
-1.05854726e+00 4.17863429e-01 6.81827545e-01 9.61780310e-01
5.00005841e-01 -1.23065375e-01 -3.76308203e-01 -5.93002327e-02
4.56340671e-01 5.72641551e-01 1.66244194e-01 -7.96365917e-01
4.31004353e-02 4.79685932e-01 6.01347983e-01 -7.41896570e-01
-4.26129073e-01 -7.28682756e-01 -4.01653439e-01 6.37437582e-01
8.07681739e-01 3.47231403e-02 -7.82195330e-01 2.08281541e+00
6.76255345e-01 5.87698340e-01 -5.15855812e-02 9.87438619e-01
4.33476448e-01 1.44575775e-01 2.63903677e-01 -1.27038345e-01
1.47771120e+00 -1.03358877e+00 -5.68854034e-01 -3.65288079e-01
4.15212393e-01 -5.44328570e-01 4.07103062e-01 2.68779755e-01
-1.07736671e+00 -8.75833690e-01 -8.04241478e-01 1.47563756e-01
-5.41591167e-01 2.12003157e-01 5.30573249e-01 4.95205492e-01
-1.19169927e+00 6.85020924e-01 -1.28256357e+00 -7.49276519e-01
4.34146374e-01 3.20765734e-01 -3.74842048e-01 8.02336410e-02
-8.93134058e-01 1.04944468e+00 6.14858568e-01 2.14467645e-01
-1.18490887e+00 -4.89441484e-01 -6.15481138e-01 -1.04137436e-01
6.01388633e-01 -9.94121790e-01 1.16883016e+00 -8.58415902e-01
-1.01566434e+00 8.18271041e-01 -3.19508761e-01 -8.20896208e-01
9.64182854e-01 -4.88260448e-01 -4.34256822e-01 -1.65788811e-02
7.88545609e-02 6.87422454e-01 1.10243964e+00 -1.29435658e+00
-8.46679807e-01 -4.77634877e-01 2.38736540e-01 -2.20794275e-01
-1.45439833e-01 2.01354027e-01 -6.22133911e-01 -3.41455460e-01
-4.26958539e-02 -1.11194766e+00 -1.58452809e-01 6.73551381e-01
-1.69031948e-01 -1.19669951e-01 1.08552420e+00 -4.94337678e-01
8.25049818e-01 -1.99044013e+00 2.37900466e-01 -2.22929046e-01
2.90998340e-01 2.71853566e-01 -1.04287369e-02 1.96755365e-01
2.34004468e-01 -4.05969679e-01 1.21970303e-01 -8.05684626e-01
9.71709341e-02 1.65134162e-01 -4.41558093e-01 7.65451372e-01
2.31662393e-01 1.22502315e+00 -1.08471203e+00 -6.84784532e-01
3.77352983e-01 7.28067875e-01 -3.41349125e-01 5.98722510e-02
-6.88873887e-01 7.81953752e-01 -3.80329818e-01 7.00895488e-01
4.99272734e-01 -5.36548734e-01 1.39441133e-01 -5.07091507e-02
-4.00628358e-01 -2.81767964e-01 -1.47962463e+00 1.81738102e+00
-2.07796738e-01 7.38792300e-01 1.41198393e-02 -4.37881976e-01
4.27043438e-01 4.79122400e-01 5.16950607e-01 -4.38288212e-01
5.37366271e-02 1.82082504e-01 -7.84226954e-02 -3.16565186e-01
5.76317728e-01 3.22778225e-01 2.45393679e-01 3.17794532e-01
1.54431209e-01 5.94138384e-01 4.86015528e-01 2.97431350e-01
1.32094038e+00 7.13830769e-01 6.45449385e-02 1.39645085e-01
4.84234631e-01 2.72345059e-02 4.37647164e-01 9.87277269e-01
-3.33648324e-01 4.82978731e-01 6.60535172e-02 -6.82984769e-01
-8.84484887e-01 -1.11840963e+00 1.39554322e-01 1.30691445e+00
3.52207541e-01 -3.16569269e-01 -4.56820667e-01 -6.39881551e-01
1.34324774e-01 3.14522803e-01 -7.79538274e-01 1.33538663e-01
-7.91404307e-01 -2.67054439e-01 3.52198452e-01 5.10500789e-01
1.49736613e-01 -1.11427140e+00 -1.30758810e+00 4.62118030e-01
-1.59181431e-01 -1.44486928e+00 -2.76521087e-01 -4.23942059e-02
-7.18167782e-01 -1.23962796e+00 -6.17880464e-01 -3.23153615e-01
3.74924600e-01 2.62307435e-01 1.27185702e+00 4.23223317e-01
-3.41300309e-01 6.43628061e-01 -1.07964605e-01 -3.29306245e-01
-4.74189103e-01 -1.50554270e-01 1.84530511e-01 2.14316204e-01
-3.27030569e-02 -4.35463756e-01 -4.94927168e-01 2.99134791e-01
-7.41068661e-01 -5.27904443e-02 7.73184538e-01 4.41951960e-01
5.66750169e-01 -2.31752917e-01 8.36762562e-02 -6.32305384e-01
-2.85894394e-01 -2.46228829e-01 -9.44291532e-01 5.62259018e-01
-1.03779379e-02 3.38759534e-02 2.99249738e-01 -7.64425933e-01
-7.46490777e-01 5.33088207e-01 1.44902512e-01 -1.00192332e+00
-1.78832754e-01 -1.66779086e-01 2.60392837e-02 -2.63204724e-01
5.49589813e-01 2.82508403e-01 -2.51798898e-01 -4.86268908e-01
6.02969944e-01 -8.43480453e-02 7.85176575e-01 -3.20937961e-01
1.00296891e+00 9.72040415e-01 -2.67599840e-02 -6.48929000e-01
-1.06972134e+00 -6.09733224e-01 -9.68880534e-01 -6.13433599e-01
9.64649916e-01 -1.05073202e+00 -1.20791185e+00 3.87919128e-01
-1.42418659e+00 -2.48329073e-01 -3.71527910e-01 4.66547936e-01
-5.77536345e-01 2.28766829e-01 -4.27026242e-01 -1.07662404e+00
2.36758403e-02 -8.51043940e-01 1.39150679e+00 8.92790854e-02
-1.58040021e-02 -1.05901337e+00 1.48678750e-01 -7.08872899e-02
4.72521752e-01 5.17220616e-01 2.05414761e-02 -4.89011317e-01
-1.21374691e+00 -2.81004488e-01 -1.73488352e-02 -2.08529204e-01
-2.53379345e-01 -4.02404070e-02 -1.09941328e+00 -6.12295866e-01
-6.77602040e-03 -2.76294678e-01 1.14148164e+00 3.06084514e-01
6.39229059e-01 -2.13038683e-01 -8.86746883e-01 2.78707266e-01
1.41441774e+00 -2.91797370e-01 2.26368219e-01 3.88358235e-01
5.74462295e-01 3.13984215e-01 6.46664858e-01 3.04693967e-01
4.61738855e-01 9.69283998e-01 8.57226312e-01 -4.36592065e-02
-4.58726108e-01 -1.72323197e-01 4.27469939e-01 9.60852951e-02
-1.80932030e-01 -3.77292752e-01 -6.13447905e-01 6.14122510e-01
-2.12085080e+00 -1.33081508e+00 -2.62537330e-01 2.19186974e+00
3.80972087e-01 4.67584699e-01 4.66147214e-01 -1.58704355e-01
7.40001142e-01 2.08429411e-01 -6.40636325e-01 5.29603899e-01
-2.15521500e-01 -2.68532246e-01 5.30198932e-01 4.39929873e-01
-1.43066859e+00 9.62972224e-01 5.45017672e+00 2.87642598e-01
-9.43975806e-01 4.90241587e-01 -7.04108998e-02 -1.25290692e-01
2.59100676e-01 2.06789419e-01 -1.16724479e+00 4.31334525e-01
8.70957196e-01 2.69789219e-01 1.49005368e-01 7.90434182e-01
-1.59137562e-01 -1.82113200e-01 -1.43362844e+00 6.82493389e-01
2.74498034e-02 -1.22176707e+00 -2.85151333e-01 -6.02850765e-02
4.68392253e-01 1.80617675e-01 1.30237695e-02 2.34522298e-01
4.04947072e-01 -5.62261105e-01 1.30241501e+00 7.69412041e-01
1.37949318e-01 -5.39998300e-02 3.56043100e-01 6.29598856e-01
-1.60780478e+00 -6.93944395e-02 -7.92186633e-02 2.93986630e-02
4.69151288e-01 2.01602757e-01 -4.76977974e-01 6.43584490e-01
5.81851006e-01 7.45266259e-01 -9.60707963e-01 1.28658426e+00
-1.54374331e-01 1.47280604e-01 -5.41281044e-01 1.97289720e-01
8.96057189e-02 3.36779654e-01 7.85907924e-01 1.17369688e+00
1.50495395e-01 -2.80178696e-01 5.37591457e-01 8.87518287e-01
1.99653804e-01 -5.18415749e-01 -5.74408472e-01 2.70795166e-01
2.75988847e-01 1.28683913e+00 -9.44165230e-01 -4.59591538e-01
-5.84973156e-01 9.84351218e-01 5.38540006e-01 2.16541678e-01
-1.17548549e+00 3.05481970e-01 5.10132551e-01 5.86039387e-02
7.85536647e-01 -2.92055190e-01 4.17926788e-01 -1.21038413e+00
3.01498771e-01 -4.95529801e-01 5.51743388e-01 -7.35016346e-01
-9.14860010e-01 7.71585882e-01 -2.78959572e-01 -1.35376287e+00
-2.71019876e-01 -6.26369953e-01 -3.02138090e-01 4.99495536e-01
-1.65166259e+00 -1.65719235e+00 -3.34319681e-01 5.31793475e-01
3.50612760e-01 3.76708001e-01 5.63312769e-01 5.30954182e-01
-3.11755180e-01 1.86405510e-01 -1.39884263e-01 1.81059420e-01
5.95881879e-01 -1.13964617e+00 5.40567100e-01 1.12459826e+00
7.48697996e-01 4.87836570e-01 1.03285491e+00 -7.41330564e-01
-1.59474528e+00 -1.33998775e+00 8.81595612e-01 -1.19085050e+00
8.11585069e-01 -5.92959642e-01 -1.09282339e+00 1.20477819e+00
-1.86106227e-02 6.13561511e-01 4.03995113e-03 3.64867076e-02
-3.19930583e-01 1.78234816e-01 -7.05714405e-01 4.20244813e-01
1.34939468e+00 -3.45966876e-01 -5.74966192e-01 5.98427415e-01
6.92604482e-01 -7.37573147e-01 -7.02062666e-01 2.19871521e-01
7.42960870e-01 -1.09781027e+00 1.17817831e+00 -5.84262490e-01
-3.00764650e-01 -6.52755260e-01 1.60176661e-02 -6.77502811e-01
-2.57444918e-01 -6.58996940e-01 -6.77834272e-01 1.07426834e+00
1.57110944e-01 -3.49065661e-01 7.81015277e-01 3.65216315e-01
6.20497428e-02 -3.51752579e-01 -1.00930643e+00 -1.07324517e+00
-4.89815712e-01 -4.58421975e-01 2.30553359e-01 5.59631646e-01
-6.51044190e-01 1.29098460e-01 -6.47355437e-01 5.97054064e-01
9.59133625e-01 4.76699084e-01 1.01371443e+00 -1.55930173e+00
-5.74942231e-01 -2.08827496e-01 -5.74680746e-01 -1.07660818e+00
1.40620455e-01 -6.00984395e-01 9.02426466e-02 -1.32914770e+00
2.16713458e-01 -2.56117970e-01 -2.34740973e-01 5.11688054e-01
-7.95144215e-02 3.88734132e-01 5.76687276e-01 3.99181038e-01
-1.50165093e+00 3.31879705e-01 9.86839294e-01 -1.08447997e-02
-5.05733602e-02 2.00767323e-01 -8.21845829e-02 8.18578482e-01
3.26613754e-01 -6.96534932e-01 1.87978104e-01 -4.66046244e-01
1.07500166e-01 7.00830743e-02 1.07712650e+00 -1.50015044e+00
7.74835706e-01 1.35547861e-01 5.01696408e-01 -9.26102400e-01
6.99680269e-01 -1.15431511e+00 5.58195531e-01 8.31203640e-01
-1.86461240e-01 2.02162236e-01 3.36451620e-01 9.14945424e-01
1.25924334e-01 1.25704572e-01 7.13161826e-01 -2.42744505e-01
-9.54083145e-01 4.88456368e-01 3.74550112e-02 -7.16422349e-02
1.20442545e+00 -2.09922403e-01 -2.88896948e-01 9.65999886e-02
-1.01049101e+00 3.70159537e-01 6.73645377e-01 6.14365458e-01
4.01964635e-01 -1.24627042e+00 -6.58531964e-01 5.30408286e-02
1.39175281e-01 -2.59314716e-01 3.12110871e-01 1.12637591e+00
-2.73162499e-02 5.21947324e-01 -1.00282930e-01 -1.05227363e+00
-1.31494653e+00 1.04220414e+00 3.24486822e-01 -3.16764802e-01
-9.11280572e-01 4.95648891e-01 9.06699374e-02 5.80417961e-02
4.65599298e-01 -4.21495140e-01 -1.00775160e-01 3.94065911e-03
7.27468133e-01 2.04878733e-01 -1.21703327e-01 -8.56419742e-01
-5.27161300e-01 7.61199713e-01 -4.65617143e-02 -2.24121347e-01
1.06829560e+00 -2.35502809e-01 1.29796207e-01 7.19086289e-01
6.46770060e-01 -1.14462405e-01 -1.70421422e+00 -5.00704050e-01
2.33419776e-01 -6.14421964e-01 -4.08850074e-01 -6.61827147e-01
-8.92914832e-01 6.63237572e-01 7.97999024e-01 2.26274788e-01
6.69030964e-01 3.09853584e-01 5.20379186e-01 4.89230424e-01
5.30262232e-01 -5.35485268e-01 1.62081391e-01 3.65423530e-01
5.33941090e-01 -1.43126655e+00 -1.20933972e-01 2.17939001e-02
-3.00813675e-01 9.48906839e-01 6.68717623e-01 -2.24843323e-02
3.88888329e-01 2.99538970e-01 -1.33060694e-01 -3.29031676e-01
-1.00526392e+00 -7.33300388e-01 2.30313927e-01 5.43004096e-01
5.38985282e-02 -3.09233099e-01 3.90653342e-01 -5.28402142e-02
2.27081001e-01 2.00329930e-01 1.05868578e-01 1.08004379e+00
-5.29206574e-01 -8.11200976e-01 -5.39044440e-01 -1.27433091e-01
-4.35855925e-01 2.46409789e-01 -3.12659204e-01 1.11542642e+00
3.52368951e-01 7.61208475e-01 7.59158656e-02 9.15412009e-02
3.62138450e-01 -9.50688776e-03 6.53444707e-01 -3.67728591e-01
-5.76663792e-01 -7.52573535e-02 -1.82359532e-01 -7.46863902e-01
-8.60394478e-01 -9.75912154e-01 -9.92575407e-01 -2.12100521e-03
-5.16322017e-01 -1.14333339e-01 5.70616722e-01 1.02010071e+00
2.79190063e-01 8.04702401e-01 5.23495674e-02 -1.04510641e+00
-4.30896610e-01 -6.94522798e-01 -2.49711633e-01 5.17913759e-01
8.70575547e-01 -9.52425361e-01 -9.45291817e-02 1.61922798e-01] | [6.340550899505615, -2.055274248123169] |
c456c77e-5b06-4a36-805e-dd7d4ee45520 | dual-view-correlation-hybrid-attention | 2306.10676 | null | https://arxiv.org/abs/2306.10676v1 | https://arxiv.org/pdf/2306.10676v1.pdf | Dual-view Correlation Hybrid Attention Network for Robust Holistic Mammogram Classification | Mammogram image is important for breast cancer screening, and typically obtained in a dual-view form, i.e., cranio-caudal (CC) and mediolateral oblique (MLO), to provide complementary information. However, previous methods mostly learn features from the two views independently, which violates the clinical knowledge and ignores the importance of dual-view correlation. In this paper, we propose a dual-view correlation hybrid attention network (DCHA-Net) for robust holistic mammogram classification. Specifically, DCHA-Net is carefully designed to extract and reinvent deep features for the two views, and meanwhile to maximize the underlying correlations between them. A hybrid attention module, consisting of local relation and non-local attention blocks, is proposed to alleviate the spatial misalignment of the paired views in the correlation maximization. A dual-view correlation loss is introduced to maximize the feature similarity between corresponding strip-like regions with equal distance to the chest wall, motivated by the fact that their features represent the same breast tissues, and thus should be highly-correlated. Experimental results on two public datasets, i.e., INbreast and CBIS-DDSM, demonstrate that DCHA-Net can well preserve and maximize feature correlations across views, and thus outperforms the state-of-the-arts for classifying a whole mammogram as malignant or not. | ['Xin Yang', 'Qiang Li', 'Xin Li', 'Kangyi Liu', 'Junlin Xian', 'Zhiwei Wang'] | 2023-06-19 | null | null | null | null | ['clinical-knowledge'] | ['miscellaneous'] | [-2.35200450e-02 1.89085782e-01 -3.71285975e-01 -6.46114290e-01
-6.69842482e-01 -1.16200708e-01 2.73557931e-01 4.50413078e-02
-3.14861573e-02 2.42743045e-01 5.20597577e-01 -8.91348198e-02
-5.20117819e-01 -7.01641440e-01 -6.11573815e-01 -9.10056889e-01
1.74393475e-01 1.01785116e-01 2.78691528e-04 -6.53774068e-02
-2.74214059e-01 3.99698347e-01 -9.89408314e-01 4.99190897e-01
5.49363554e-01 1.21440184e+00 2.23466218e-01 4.37898070e-01
3.91704977e-01 7.74323404e-01 4.74504903e-02 -6.36892140e-01
1.66001603e-01 -6.26778483e-01 -4.47408587e-01 3.59398425e-01
3.07380855e-01 -4.01114553e-01 -6.64781988e-01 1.10269105e+00
5.16470730e-01 -4.23583120e-01 4.40752983e-01 -8.85331869e-01
-7.87671804e-01 3.84548128e-01 -1.38671196e+00 4.50888813e-01
-8.42835456e-02 -7.01929852e-02 9.79797900e-01 -7.50763893e-01
3.38344574e-01 9.95736599e-01 6.29513383e-01 3.01618844e-01
-9.82765615e-01 -6.63149178e-01 1.50774404e-01 2.19417512e-01
-1.27423358e+00 -2.90441185e-01 8.88473690e-01 -1.50175020e-01
2.99973130e-01 5.42423427e-01 8.95885408e-01 8.75237226e-01
7.07302034e-01 9.25411046e-01 9.89446878e-01 -1.48350403e-01
-3.55060607e-01 2.76156943e-02 -2.59586629e-02 7.08984971e-01
3.06619704e-01 1.37633428e-01 -3.62835050e-01 -1.12628579e-01
7.48948276e-01 5.36798596e-01 -4.24087346e-01 -8.05414379e-01
-1.43961775e+00 6.66498601e-01 8.02534819e-01 4.24316913e-01
-4.79391813e-01 -2.72918642e-01 4.68617618e-01 -2.40766004e-01
3.19441736e-01 8.72389600e-02 -1.44854024e-01 4.77977276e-01
-6.89056754e-01 -3.55727412e-02 2.03895211e-01 7.89426267e-01
3.95239741e-01 -4.96248424e-01 -3.53359133e-01 8.55062425e-01
4.30218250e-01 5.12037218e-01 6.45411432e-01 -3.72275084e-01
6.37113512e-01 8.02515030e-01 -2.63940990e-01 -1.48255861e+00
-6.54979646e-01 -8.23433816e-01 -1.38064826e+00 -4.72940445e-01
-3.91959995e-02 -1.18054431e-02 -8.51820529e-01 1.63111436e+00
4.82770681e-01 5.04867248e-02 1.33507019e-02 1.11033487e+00
1.04533768e+00 1.40833810e-01 -1.85505629e-01 -2.33308211e-01
1.61861908e+00 -8.56606066e-01 -7.67519653e-01 -3.14273536e-01
6.81907594e-01 -5.26112497e-01 6.07580423e-01 3.94170545e-02
-1.06149364e+00 -4.58519667e-01 -1.14939964e+00 -2.23925300e-02
1.86436161e-01 2.95653313e-01 6.16774261e-01 4.24165159e-01
-5.46102524e-01 2.43566141e-01 -1.28669381e+00 -4.71128039e-02
7.34678388e-01 2.11934119e-01 -5.63682616e-01 -4.31779325e-01
-9.10066485e-01 5.52061021e-01 1.30661041e-01 4.65689600e-01
-5.45579076e-01 -6.96031749e-01 -8.41050684e-01 1.79162279e-01
2.49242425e-01 -7.08777368e-01 8.41145098e-01 -9.32961583e-01
-8.77171934e-01 9.22249973e-01 -2.13735074e-01 -2.87746817e-01
5.44594884e-01 -1.63023602e-02 -5.38918257e-01 2.79729962e-01
3.17064494e-01 5.68231463e-01 4.46926028e-01 -1.31495202e+00
-5.65825105e-01 -8.87317836e-01 -2.54013240e-01 4.36782569e-01
-6.17045462e-01 -1.51785865e-01 -9.55806255e-01 -7.36716509e-01
7.88831890e-01 -8.59670162e-01 -2.70047575e-01 8.22014660e-02
-6.24847651e-01 3.61113220e-01 7.16889799e-01 -9.26783800e-01
1.17566180e+00 -2.28998971e+00 1.28851637e-01 2.59711474e-01
3.84208977e-01 -9.75187197e-02 4.99784388e-02 -1.42969862e-01
-4.32059616e-01 -4.42593358e-02 -2.51201957e-01 1.90251488e-02
-5.31084180e-01 2.55741328e-01 2.34840587e-01 7.94716835e-01
2.37679631e-01 1.14160132e+00 -1.00506032e+00 -6.08417153e-01
5.19713722e-02 3.30941826e-01 -5.18266141e-01 7.44832829e-02
3.72540951e-01 5.80136597e-01 -6.56698883e-01 7.40613043e-01
1.02813232e+00 -6.33966684e-01 3.89746726e-01 -5.60600996e-01
2.65138954e-01 -2.18640879e-01 -6.52725935e-01 1.69733155e+00
-5.28341889e-01 3.56851727e-01 -1.03446916e-02 -9.96844769e-01
7.18149364e-01 2.80017227e-01 1.01781726e+00 -1.04119122e+00
5.21328999e-03 1.62829950e-01 4.35438365e-01 -6.50187671e-01
2.22171686e-04 -1.58592418e-01 2.27233633e-01 1.93746492e-01
-1.96851358e-01 2.20996901e-01 -1.07915387e-01 6.19673468e-02
7.65492797e-01 -3.19805384e-01 7.73375452e-01 -3.54680181e-01
6.06323779e-01 -4.73392636e-01 8.97600234e-01 3.72766286e-01
-1.34223029e-01 9.41824019e-01 8.08730125e-01 -3.80905449e-01
-6.81679487e-01 -9.18630242e-01 -3.75066042e-01 5.06078959e-01
6.87368095e-01 -1.75801143e-01 -3.97357553e-01 -1.04787850e+00
2.91674025e-02 2.43810490e-01 -1.05391300e+00 -3.92038375e-01
-8.06665659e-01 -1.03753734e+00 2.13780403e-01 9.18468654e-01
6.11950517e-01 -4.85443830e-01 -7.52407670e-01 -8.84139352e-03
-2.79859096e-01 -9.40397203e-01 -7.32935965e-01 1.50688857e-01
-8.42256725e-01 -1.27288806e+00 -9.68890131e-01 -5.65415502e-01
9.12636399e-01 6.88618124e-01 9.24396694e-01 5.51823415e-02
-3.51194590e-01 -5.72254322e-02 -3.77427042e-01 -1.87695771e-01
9.97329354e-02 -1.76095426e-01 -5.32772064e-01 3.39885414e-01
3.91607046e-01 -5.96711218e-01 -9.12982464e-01 5.82380414e-01
-9.74359155e-01 3.71763498e-01 9.58899021e-01 1.55222297e+00
8.51244032e-01 -1.64384857e-01 3.17488641e-01 -9.43212390e-01
3.17572169e-02 -7.49681115e-01 -1.66017532e-01 3.86441141e-01
-2.02417284e-01 -2.37838656e-01 4.35425133e-01 -6.42442405e-02
-8.72726381e-01 3.75296064e-02 -8.37862045e-02 -5.87131083e-01
1.32737011e-01 6.72154248e-01 -5.29175699e-01 1.81550123e-02
1.88899323e-01 2.64982551e-01 -1.41553923e-01 -1.46018445e-01
-1.54366627e-01 3.32771122e-01 4.09669489e-01 1.32453069e-01
4.53417718e-01 8.07986677e-01 1.23776071e-01 -3.78949255e-01
-1.10757279e+00 -5.60512900e-01 -5.82412243e-01 6.67029992e-02
8.81029427e-01 -9.42607760e-01 -4.32417989e-01 3.45004857e-01
-8.07749927e-01 2.75993109e-01 -1.52085543e-01 8.04542363e-01
-2.76473671e-01 3.73535991e-01 -4.48185951e-01 -3.24438363e-01
-2.20222786e-01 -1.31126940e+00 1.24813735e+00 4.95750606e-01
2.65133530e-01 -7.51515865e-01 -3.40162814e-01 4.20062900e-01
3.90303582e-02 2.94191480e-01 9.66866970e-01 -5.68035543e-01
-2.07279310e-01 -2.76528090e-01 -7.49398768e-01 2.47959644e-01
3.89674455e-01 -3.51418823e-01 -9.11295891e-01 -3.74258369e-01
3.00727785e-01 -1.69499621e-01 1.03436065e+00 6.67859197e-01
1.57336926e+00 -1.20868094e-01 -6.73879802e-01 1.03256071e+00
1.16135693e+00 2.95692593e-01 5.39349020e-01 6.70476332e-02
7.45366573e-01 4.74133939e-01 5.31248331e-01 4.98330414e-01
4.57339108e-01 6.85024858e-01 7.79677033e-01 -5.16656637e-01
2.72219647e-02 -1.68674573e-01 -2.23112881e-01 8.65414202e-01
2.02972963e-01 -2.61091769e-01 -9.16153014e-01 5.14500499e-01
-1.69816530e+00 -7.16305435e-01 9.35250446e-02 2.08850527e+00
4.34707254e-01 -5.61518818e-02 -2.56815851e-01 -2.12352082e-01
7.55638957e-01 3.81814241e-01 -7.66109109e-01 1.79432139e-01
-1.75487250e-01 -1.36624977e-01 5.57537973e-01 6.54698163e-02
-1.36269271e+00 4.14726026e-02 5.13254547e+00 8.13545942e-01
-1.32966256e+00 2.20425382e-01 1.31274891e+00 -2.82154173e-01
-3.99168104e-01 -2.58216172e-01 -5.00085354e-01 4.66384947e-01
1.58938110e-01 1.17634080e-01 -2.26704627e-01 7.40758479e-01
-1.06619313e-01 -5.04151918e-02 -1.00894678e+00 8.66374552e-01
3.25447768e-01 -1.07611895e+00 -9.37565267e-02 2.29313865e-01
8.41967225e-01 -1.26185790e-01 2.37330303e-01 -1.71215311e-02
-2.08684564e-01 -1.05021596e+00 3.72835726e-01 5.95801950e-01
7.57213235e-01 -8.99547458e-01 1.11789823e+00 2.30615795e-01
-1.30148256e+00 -4.19839397e-02 -2.96819866e-01 7.85540462e-01
-1.39059767e-01 7.54776359e-01 -3.49433273e-01 1.20848656e+00
7.79138863e-01 1.07169354e+00 -7.90955782e-01 9.05331135e-01
2.95357794e-01 1.62828043e-01 -9.73263681e-02 2.51112282e-01
2.77716696e-01 -5.56288809e-02 4.02608126e-01 9.52489316e-01
4.61709708e-01 1.84242353e-01 1.64868429e-01 6.59933388e-01
2.71798447e-02 7.67274499e-02 -2.90424943e-01 3.82719010e-01
1.15965389e-01 1.34981096e+00 -6.28519356e-01 -9.15910751e-02
-7.36846387e-01 8.09924424e-01 1.58288404e-01 1.05199784e-01
-9.00898278e-01 -6.43570423e-02 1.53892681e-01 2.24558547e-01
4.33390588e-01 3.92406136e-01 -3.37038338e-01 -1.18206573e+00
3.87380451e-01 -8.10349405e-01 4.40403223e-01 -4.72207397e-01
-1.40424931e+00 6.95036173e-01 -4.99458872e-02 -1.65723014e+00
9.45024639e-02 -3.25043559e-01 -6.70185089e-01 7.94017971e-01
-1.56641150e+00 -1.37381804e+00 -6.67029440e-01 4.52524960e-01
1.64837733e-01 -3.23904455e-02 5.19023478e-01 3.93856823e-01
-5.88033199e-01 9.06915307e-01 2.37976626e-01 2.39118055e-01
6.73560441e-01 -1.01942682e+00 -6.64779842e-02 5.95863998e-01
-2.26515323e-01 4.55004811e-01 2.42520627e-02 -4.82992113e-01
-1.39598835e+00 -1.28225338e+00 4.89714861e-01 4.03690301e-02
2.75042504e-01 -2.51552388e-02 -8.51053536e-01 4.61782128e-01
1.13144428e-01 8.64289641e-01 7.59887755e-01 -1.14434250e-01
-2.66655833e-01 -3.77882302e-01 -8.42340052e-01 4.60100889e-01
9.17437077e-01 -1.86396912e-01 -5.41830882e-02 4.54498798e-01
3.33673477e-01 -6.92672193e-01 -9.99708295e-01 1.00192857e+00
7.61322021e-01 -1.36755192e+00 9.76463675e-01 -5.12461662e-01
8.76289010e-01 4.55272477e-03 -3.17488700e-01 -1.17900014e+00
-4.57906246e-01 -5.58016188e-02 -3.16535234e-02 8.67262185e-01
2.91664094e-01 -5.10462105e-01 8.37069571e-01 -7.01251253e-02
-2.25413129e-01 -1.52735567e+00 -1.08899438e+00 -4.78228897e-01
-2.54737809e-02 4.75500990e-03 5.81308603e-01 1.02018213e+00
-1.57421321e-01 2.66420931e-01 -3.08073431e-01 3.19504887e-01
2.46306270e-01 4.80287850e-01 5.18827975e-01 -8.37962091e-01
-5.39727628e-01 -5.21923006e-01 -5.92969954e-01 -1.09516966e+00
-2.58411050e-01 -8.79686654e-01 -9.91663486e-02 -1.23545122e+00
8.94251585e-01 -4.46140766e-01 -5.69454253e-01 1.79062873e-01
-6.46438599e-01 1.44898921e-01 -6.09450787e-02 1.63160712e-01
-2.90525228e-01 6.98569477e-01 1.85717773e+00 -1.33923456e-01
6.64631575e-02 8.75827745e-02 -1.07016265e+00 9.11261737e-01
4.31454659e-01 -4.05379534e-01 -2.21847877e-01 -5.19236505e-01
8.38406682e-02 5.30912042e-01 4.68152046e-01 -9.41459119e-01
1.91835538e-01 -9.82500538e-02 9.34772134e-01 -9.45272505e-01
3.06910008e-01 -9.00901079e-01 -3.15918326e-02 4.59431469e-01
-3.51146400e-01 2.09029466e-02 4.90868092e-02 9.00687456e-01
-5.52539587e-01 3.88313718e-02 9.40960050e-01 -2.61586338e-01
-3.43121812e-02 5.73092461e-01 2.64311612e-01 -1.14115663e-01
1.13000000e+00 -1.68579102e-01 -2.12373927e-01 -1.52591467e-01
-6.06436968e-01 5.55421829e-01 2.63107121e-01 3.32707524e-01
6.35412693e-01 -1.64258873e+00 -1.03910470e+00 4.13515508e-01
3.76866728e-01 2.58452863e-01 8.26885760e-01 1.57313621e+00
-2.64213294e-01 4.75009352e-01 -1.86422467e-01 -9.06133413e-01
-1.27789485e+00 5.72451055e-01 5.68581283e-01 -9.40493166e-01
-6.96720302e-01 8.94180417e-01 1.01965117e+00 -3.66997391e-01
-2.40682848e-02 -1.63353682e-01 -2.84567535e-01 -1.62500460e-02
5.38583219e-01 -1.33992121e-01 2.42653579e-01 -7.29631484e-01
-4.54496801e-01 6.34393752e-01 -5.05042553e-01 2.44517803e-01
1.28901339e+00 -6.98033720e-02 -8.86830464e-02 5.59752509e-02
1.43257689e+00 1.03262298e-01 -1.11141384e+00 -4.72197384e-01
-3.56374741e-01 -7.15273321e-01 2.16150552e-01 -5.17153502e-01
-1.69768691e+00 1.09562600e+00 7.85511017e-01 -4.26598303e-02
1.47314847e+00 -1.95133220e-02 7.84377873e-01 9.19155404e-02
-2.26691648e-01 -5.90609014e-01 2.76925713e-01 8.04459006e-02
9.52153742e-01 -1.49680936e+00 4.01936769e-01 -5.41288137e-01
-9.59232807e-01 1.13308811e+00 9.76865232e-01 4.38330397e-02
9.37527299e-01 1.07996017e-01 -2.91151762e-01 -4.10160482e-01
-6.20380282e-01 1.87466100e-01 5.75379968e-01 3.66559595e-01
7.27255583e-01 2.05965251e-01 -2.09224328e-01 1.02193975e+00
2.19999284e-01 -4.17970628e-01 5.09706289e-02 9.39368069e-01
1.74868722e-02 -5.00120163e-01 -3.56887728e-01 9.25092220e-01
-6.33299828e-01 5.20501137e-02 -2.64321029e-01 9.57176566e-01
4.51567292e-01 5.82416058e-01 1.65589452e-01 -5.44380307e-01
3.01673442e-01 -6.22992218e-01 2.92707533e-01 -4.88990307e-01
-5.19259393e-01 6.02014720e-01 -3.26071650e-01 -5.16058803e-01
-3.71561289e-01 -7.89321423e-01 -9.81278539e-01 1.21203952e-01
-7.59682655e-01 -1.45457715e-01 1.10553153e-01 8.16706777e-01
2.91407168e-01 8.28631878e-01 1.06448138e+00 -5.08336782e-01
-5.89113951e-01 -9.17385280e-01 -5.69703341e-01 3.16207647e-01
4.48510051e-01 -5.86330056e-01 1.37928411e-01 -1.08548790e-01] | [14.953553199768066, -2.3201065063476562] |
71450cd7-e7a3-4b0e-b983-c115c03a73f9 | modeling-hierarchical-syntax-structure-with | null | null | https://openreview.net/forum?id=w8g-gf1wjjl | https://openreview.net/pdf?id=w8g-gf1wjjl | Modeling Hierarchical Syntax Structure with Triplet Position for Source Code Summarization | Automatic code summarization, which aims to describe the source code in natural language, has become an essential task in software maintenance. Our fellow researchers have attempted to achieve such a purpose through various machine learning-based approaches. One key challenge keeping these approaches from being practical lies in the lacking of retaining the semantic structure of source code, which has unfortunately been overlooked by the state-of-the-art. Existing approaches resort to representing the syntax structure of code by modeling the Abstract Syntax Trees (ASTs). However, the hierarchical structures of ASTs have not been well explored. In this paper, we propose CODESCRIBE to model the hierarchical syntax structure of code by introducing a novel triplet position for code summarization. Specifically, CODESCRIBE leverages the graph neural network and Transformer to preserve the structural and sequential information of code, respectively. In addition, we propose a pointer-generator network that pays attention to both the structure and sequential tokens of code for a better summary generation. Experiments on two real-world datasets in Java and Python demonstrate the effectiveness of our proposed approach when compared with several state-of-the-art baselines. | ['Anonymous'] | 2021-11-16 | null | null | null | acl-arr-november-2021-11 | ['code-summarization'] | ['computer-code'] | [ 2.60583252e-01 1.98542252e-01 -3.77891272e-01 -2.02547684e-01
-5.96110582e-01 -4.47963953e-01 4.09542859e-01 5.09163380e-01
3.54843408e-01 2.37197146e-01 5.94320774e-01 -5.23697078e-01
2.36988097e-01 -5.54151475e-01 -6.82252705e-01 -1.83021545e-01
-6.08050898e-02 -4.29057717e-01 4.56859469e-01 -2.09582657e-01
8.02139163e-01 -2.15284660e-01 -1.64038992e+00 4.28269088e-01
1.35666239e+00 5.82871616e-01 3.20477635e-01 2.46977940e-01
-8.18114281e-01 1.16304696e+00 -5.84580183e-01 -6.03991032e-01
-1.36891633e-01 -5.49779654e-01 -8.89422536e-01 -2.21291110e-02
4.31536257e-01 -3.14078256e-02 -2.23736748e-01 1.37264764e+00
1.53083101e-01 -3.85200500e-01 3.28545094e-01 -1.27797687e+00
-6.77740216e-01 1.04672408e+00 -9.66694832e-01 -7.65504539e-02
4.10770983e-01 7.50200078e-02 1.29089093e+00 -5.95680296e-01
4.31350678e-01 1.03661764e+00 7.05491662e-01 4.65234697e-01
-1.04314566e+00 -3.75081688e-01 1.26943052e-01 8.24391097e-02
-1.14233506e+00 -3.84453803e-01 1.05043495e+00 -5.89934528e-01
1.09103990e+00 1.13313869e-01 3.91912580e-01 6.54160738e-01
5.50566673e-01 7.31156707e-01 6.33333266e-01 -4.27719772e-01
5.69832921e-02 6.34727404e-02 4.60131705e-01 1.11135411e+00
3.76012444e-01 -4.79310960e-01 -2.90205449e-01 -3.81996095e-01
1.65718317e-01 7.20214918e-02 -2.05972821e-01 -6.47289097e-01
-1.07521200e+00 7.04655647e-01 4.26543802e-01 2.82976180e-01
-1.37046188e-01 2.90595055e-01 9.17263806e-01 6.66353703e-02
1.99669883e-01 3.75107884e-01 -2.50976801e-01 -2.97484785e-01
-1.04488420e+00 2.13634655e-01 7.71744788e-01 1.31227541e+00
8.19418609e-01 1.28490746e-01 -1.36271715e-01 6.84523761e-01
4.34807241e-01 4.42324728e-02 5.65084696e-01 -7.85856247e-01
8.09517443e-01 1.48509490e+00 -2.72138268e-01 -1.19921327e+00
-9.26491395e-02 -5.49564064e-01 -8.20412159e-01 -1.41436905e-01
-1.59768417e-01 3.08327526e-01 -4.42907125e-01 1.47149444e+00
4.31944616e-02 -8.66268110e-03 1.87572073e-02 2.38502994e-01
8.47314715e-01 6.49285555e-01 -2.52626449e-01 -1.20330364e-01
1.27746665e+00 -1.33395076e+00 -5.57743609e-01 -2.81084418e-01
9.01071846e-01 -6.83368206e-01 1.16642439e+00 9.87600684e-02
-1.04220796e+00 -5.30107677e-01 -1.10252333e+00 -1.69459835e-01
-2.79659890e-02 2.43744522e-01 6.57713413e-01 6.62044227e-01
-1.05725992e+00 4.85663563e-01 -9.99665320e-01 -3.49854916e-01
4.29238230e-01 5.23960255e-02 -2.06945941e-01 1.94927007e-01
-5.77029526e-01 4.10784513e-01 5.91067374e-01 -3.45971547e-02
-5.42890549e-01 -6.74064279e-01 -1.05613899e+00 4.72078979e-01
5.77592731e-01 -7.55623460e-01 1.47742128e+00 -7.62324929e-01
-1.25015795e+00 4.56527978e-01 -3.71424079e-01 -3.18609565e-01
-8.32292661e-02 -4.29506414e-02 -1.32824108e-01 -1.48915946e-01
1.87076762e-01 1.20698735e-01 5.80407441e-01 -1.43578994e+00
-6.68168306e-01 -2.73723394e-01 2.51353472e-01 -5.91161922e-02
-6.34833574e-01 2.42546842e-01 -5.18462896e-01 -7.30067074e-01
-2.80861650e-02 -7.51936197e-01 -9.32409540e-02 -2.26781204e-01
-5.98363042e-01 -3.98398399e-01 7.58372545e-01 -8.78726184e-01
2.12115622e+00 -2.28479648e+00 3.16102743e-01 -1.26965746e-01
4.51441169e-01 2.13062420e-01 -6.61959127e-02 8.98328006e-01
1.10318348e-01 3.33569914e-01 -7.09687889e-01 -3.94862086e-01
2.07414478e-01 2.90942769e-02 -5.73857665e-01 1.72099248e-01
1.65557861e-01 8.44241858e-01 -9.93697464e-01 -8.00420523e-01
-1.96058854e-01 4.47130576e-02 -1.13766348e+00 3.25186700e-01
-5.01852393e-01 8.69708210e-02 -6.00935698e-01 6.65459991e-01
6.33041322e-01 -3.66027236e-01 4.00753260e-01 -9.82296839e-02
-3.43529969e-01 5.91023862e-01 -8.51499557e-01 2.11766267e+00
-4.57976401e-01 1.73342749e-01 -1.34935811e-01 -1.01203370e+00
8.75179410e-01 6.67974874e-02 3.07436228e-01 -5.83617806e-01
-2.14683130e-01 2.70325273e-01 9.12048505e-04 -7.24869192e-01
7.72335172e-01 2.79056132e-01 -4.47811037e-01 5.86096287e-01
-4.04629588e-01 -1.39459759e-01 3.75629961e-01 7.11710155e-01
1.35586941e+00 5.71252286e-01 5.13778210e-01 -2.63575822e-01
7.22656369e-01 9.42163914e-02 6.52510643e-01 6.29417241e-01
1.93418950e-01 4.30722505e-01 9.58666921e-01 -1.90752417e-01
-9.61323321e-01 -6.90518677e-01 2.47303680e-01 9.17431891e-01
6.81705251e-02 -1.15391922e+00 -1.03369617e+00 -9.99908209e-01
-2.25263730e-01 7.20868587e-01 -4.90627229e-01 -3.55912209e-01
-8.92005563e-01 -5.10923564e-01 6.74377978e-01 4.61109042e-01
4.82328206e-01 -8.68344843e-01 -8.00268292e-01 2.39291564e-01
-3.95355135e-01 -7.76689351e-01 -8.28613162e-01 -1.39757782e-01
-9.14963603e-01 -1.05408621e+00 -2.77709246e-01 -8.62814486e-01
7.39090204e-01 4.10451174e-01 1.08024383e+00 7.09175289e-01
-5.92790954e-02 -1.38311600e-02 -6.87647820e-01 -5.78680038e-02
-8.85885775e-01 5.04693449e-01 -5.90895534e-01 -2.37699598e-01
1.08270288e-01 -7.65268087e-01 -4.13517088e-01 -1.28616542e-02
-1.22758281e+00 3.63013268e-01 7.18844414e-01 6.26532555e-01
9.95650068e-02 1.82450384e-01 3.23668361e-01 -1.18468785e+00
5.46991289e-01 -6.70557678e-01 -5.97640574e-01 4.54525888e-01
-5.81673443e-01 5.38240850e-01 9.33073342e-01 6.08605333e-02
-1.16794276e+00 -6.77270070e-02 -1.32581070e-01 6.68816641e-02
2.30694607e-01 1.00966728e+00 -2.36335799e-01 1.39453784e-01
3.51184011e-01 5.87501109e-01 3.66231659e-04 -5.77712834e-01
2.55997390e-01 8.71412575e-01 4.55942601e-01 -9.17239904e-01
8.38046372e-01 2.50014484e-01 -1.58829659e-01 -5.22434950e-01
-6.14417851e-01 -3.52414519e-01 -4.72020179e-01 3.00837427e-01
5.10719359e-01 -6.52481318e-01 -3.28798145e-01 4.70684826e-01
-1.48700166e+00 8.40886906e-02 5.92014566e-02 -2.75818020e-01
-4.31389362e-01 1.06768882e+00 -4.42848086e-01 -4.65566099e-01
-4.92126793e-01 -1.52449346e+00 1.23373747e+00 2.36565806e-02
-2.06647113e-01 -7.01517403e-01 2.30565786e-01 2.22479731e-01
6.78169966e-01 1.78594500e-01 1.53664362e+00 -3.62739444e-01
-7.89623320e-01 4.42417301e-02 -3.66230279e-01 1.99333593e-01
3.84388059e-01 3.66839230e-01 -4.59263384e-01 -2.53645986e-01
-4.72950153e-02 -6.90103248e-02 7.07990706e-01 -1.19262561e-01
1.29412174e+00 -6.96578741e-01 -4.67531055e-01 6.16955757e-01
1.45688951e+00 1.97833985e-01 6.70494020e-01 3.83719206e-01
8.74691665e-01 6.75264180e-01 1.99131250e-01 6.29305243e-01
9.05564725e-01 6.75520241e-01 7.58949876e-01 2.85847902e-01
-2.42058396e-01 -5.26033938e-01 4.68371570e-01 1.46080554e+00
4.24447030e-01 -3.55880223e-02 -1.05386472e+00 7.73206890e-01
-1.90591705e+00 -8.84930313e-01 -4.65593904e-01 1.95073760e+00
9.61429656e-01 8.69000554e-02 -2.17430703e-02 7.14049935e-02
6.68618381e-01 2.21443862e-01 -3.28652233e-01 -4.06520814e-01
2.37599179e-01 -1.52743071e-01 1.70067534e-01 4.13001850e-02
-7.98200607e-01 7.08305717e-01 5.12003899e+00 5.91844857e-01
-8.90236080e-01 -3.26372571e-02 -9.48896185e-02 4.59307909e-01
-5.81914544e-01 6.90343559e-01 -6.85555220e-01 6.10353172e-01
8.42644811e-01 -6.78029180e-01 4.01018232e-01 9.42149162e-01
9.81279239e-02 1.18650300e-02 -1.14201605e+00 5.73620319e-01
2.04631835e-01 -1.32316911e+00 3.06869447e-01 -1.68981537e-01
7.89112449e-01 -2.24204049e-01 -2.27050513e-01 5.60543954e-01
2.16985956e-01 -5.66783369e-01 9.94046271e-01 3.65212351e-01
5.04826665e-01 -4.77069557e-01 6.47655189e-01 4.46985155e-01
-1.51030707e+00 -1.50895491e-01 -2.50140041e-01 3.27831320e-02
-8.81171152e-02 5.20003140e-01 -6.06468797e-01 9.94647622e-01
5.80535114e-01 9.89657462e-01 -1.09744811e+00 1.19294989e+00
-1.06983095e-01 5.47111750e-01 1.85121715e-01 -7.91990682e-02
2.22468331e-01 1.69838276e-02 3.74284416e-01 1.44201398e+00
4.85306770e-01 -3.43773007e-01 2.96984464e-01 1.15839696e+00
-2.69188285e-01 2.21736044e-01 -6.32824659e-01 -3.82385522e-01
4.98628706e-01 1.08181429e+00 -6.00940049e-01 -3.13145250e-01
-7.97474861e-01 6.84756577e-01 3.42554837e-01 7.23765939e-02
-8.34007621e-01 -8.59451175e-01 2.29639068e-01 2.02027723e-01
3.80731702e-01 -2.29922488e-01 -3.23915422e-01 -1.45062363e+00
4.85545188e-01 -1.16974568e+00 2.93940455e-01 -6.09969437e-01
-8.79318357e-01 6.87449932e-01 1.49755120e-01 -1.24991941e+00
-1.97364762e-02 -6.22445494e-02 -8.32434297e-01 6.10507548e-01
-1.52971721e+00 -1.19806898e+00 -3.34359020e-01 1.24886483e-01
6.71561778e-01 -1.10147260e-01 5.72917998e-01 3.23832244e-01
-7.51552463e-01 5.55157065e-01 3.89764607e-02 1.53252512e-01
4.44939137e-01 -1.28544617e+00 9.11016166e-01 1.25495982e+00
-1.80268943e-01 1.45506167e+00 6.49633646e-01 -6.99470878e-01
-1.77878606e+00 -1.15839052e+00 1.01786685e+00 -1.87863812e-01
9.11291778e-01 -3.23598176e-01 -1.23767567e+00 7.59352207e-01
5.56446791e-01 -2.52636880e-01 4.49718833e-01 -2.41347596e-01
-6.28461957e-01 -1.18713178e-01 -5.70081890e-01 6.04818523e-01
1.03046644e+00 -5.19563675e-01 -9.16861236e-01 3.25705968e-02
9.11872685e-01 -3.59894991e-01 -5.66139638e-01 2.07948819e-01
3.83257985e-01 -1.14149475e+00 5.48729777e-01 -3.62081259e-01
1.06468630e+00 -5.93191624e-01 -2.99772806e-02 -1.23807573e+00
-3.92869301e-02 -7.09522605e-01 -2.20788598e-01 1.68879497e+00
1.40275657e-01 -4.82938349e-01 6.20608628e-01 1.24802351e-01
-6.36212051e-01 -6.36877716e-01 -5.37601709e-01 -6.75201893e-01
1.28953354e-04 -1.81589916e-01 7.97741711e-01 8.32777977e-01
4.54278409e-01 4.73513037e-01 -2.68487036e-01 -1.21992780e-03
6.08796597e-01 5.27469099e-01 1.00075948e+00 -1.19770849e+00
-3.44158202e-01 -6.62074566e-01 -2.04388991e-01 -1.10369599e+00
5.16760349e-01 -1.23775041e+00 4.00704518e-02 -1.91879976e+00
6.65065527e-01 -2.64484793e-01 3.74058336e-02 5.88113248e-01
-3.00149590e-01 -4.86417353e-01 4.41046841e-02 3.99493128e-01
-6.75411105e-01 7.19583690e-01 7.69252062e-01 -3.78066093e-01
-3.26584056e-02 -9.87430289e-02 -1.05438101e+00 6.13117516e-01
7.55392194e-01 -7.88807929e-01 -5.18731236e-01 -8.15161526e-01
4.34257030e-01 3.12114745e-01 1.49029061e-01 -9.17174518e-01
4.98510897e-01 -1.45978212e-01 -6.98543489e-01 -6.01613343e-01
-2.95198381e-01 -7.51619279e-01 1.92493841e-01 5.57792187e-01
-3.27719659e-01 5.07790267e-01 1.75304696e-01 4.58762407e-01
-4.23398018e-01 -6.43584967e-01 5.83938956e-01 -2.96002656e-01
-7.03104675e-01 8.29843953e-02 -1.32435635e-01 2.83636123e-01
7.89128482e-01 -1.79589480e-01 -6.95492804e-01 1.55582920e-01
2.11437061e-01 1.92347929e-01 9.02781904e-01 6.28329873e-01
6.20338082e-01 -1.11114347e+00 -5.73255360e-01 2.24566117e-01
5.82195759e-01 -4.70449999e-02 7.48147890e-02 7.62327850e-01
-6.98068798e-01 4.30648863e-01 -1.22416154e-01 -3.76291424e-01
-1.26610875e+00 7.04012275e-01 5.29599153e-02 -4.88833547e-01
-6.87825620e-01 2.39020377e-01 2.29654625e-01 -4.86502707e-01
1.70314744e-01 -5.46904504e-01 -1.53258443e-01 -3.35731983e-01
4.17047650e-01 2.40824863e-01 1.31842211e-01 -4.55906540e-01
-4.18745458e-01 5.68681121e-01 -3.47210467e-01 4.48354244e-01
1.37466753e+00 -1.43611148e-01 -8.19611132e-01 2.21005917e-01
1.32740617e+00 3.26742828e-01 -8.36318254e-01 -3.11311334e-01
6.33537173e-01 -4.13551807e-01 -1.52250037e-01 -4.38956439e-01
-9.44631040e-01 9.65743065e-01 -5.11817262e-02 3.27803433e-01
9.45976794e-01 -1.59293041e-01 9.71847653e-01 3.99108499e-01
5.47043920e-01 -5.79995036e-01 1.84941098e-01 5.17476082e-01
6.58828557e-01 -9.18849885e-01 -2.48781387e-02 -5.09225070e-01
-4.19312865e-01 1.19519150e+00 6.00939810e-01 1.81382880e-01
1.74128368e-01 3.38064224e-01 -3.60770971e-01 -2.58498877e-01
-7.91373849e-01 1.31489038e-01 1.37484938e-01 2.56945044e-01
6.60585880e-01 -3.76127481e-01 -5.15768707e-01 6.29075110e-01
2.82406937e-02 1.12199031e-01 1.05268776e+00 1.49642563e+00
-5.01484871e-01 -1.44894898e+00 -9.21890065e-02 4.88904208e-01
-6.55568540e-01 -4.85903800e-01 -2.87462980e-01 5.76837003e-01
-1.33800358e-01 7.99436271e-01 -3.47194791e-01 -3.06345910e-01
4.92097169e-01 -2.51797363e-02 1.93044752e-01 -1.14724779e+00
-8.44007850e-01 -2.24871948e-01 -2.15150103e-01 -4.33538288e-01
-4.42316085e-01 -5.77647090e-01 -1.35863459e+00 -2.13493943e-01
-2.45524749e-01 4.32610750e-01 5.41260421e-01 5.14605582e-01
6.76103473e-01 7.47476339e-01 5.90690970e-01 -3.74714285e-01
-7.17172503e-01 -6.90773129e-01 -2.78560668e-01 4.03298289e-01
3.53135616e-01 -3.90762717e-01 -3.21458250e-01 4.04119283e-01] | [7.569317817687988, 7.96443510055542] |
8b3f5b0e-e98b-4a62-9026-1d690ebb2348 | iterative-data-programming-for-expanding-text | 2002.01412 | null | https://arxiv.org/abs/2002.01412v1 | https://arxiv.org/pdf/2002.01412v1.pdf | Iterative Data Programming for Expanding Text Classification Corpora | Real-world text classification tasks often require many labeled training examples that are expensive to obtain. Recent advancements in machine teaching, specifically the data programming paradigm, facilitate the creation of training data sets quickly via a general framework for building weak models, also known as labeling functions, and denoising them through ensemble learning techniques. We present a fast, simple data programming method for augmenting text data sets by generating neighborhood-based weak models with minimal supervision. Furthermore, our method employs an iterative procedure to identify sparsely distributed examples from large volumes of unlabeled data. The iterative data programming techniques improve newer weak models as more labeled data is confirmed with human-in-loop. We show empirical results on sentence classification tasks, including those from a task of improving intent recognition in conversational agents. | ['Neil Mallinar', 'Ayush Gupta', 'Abhishek Shah', 'Tin Kam Ho', 'Rajendra Ugrani'] | 2020-02-04 | null | null | null | null | ['intent-recognition'] | ['natural-language-processing'] | [ 2.25963399e-01 3.90883386e-01 -1.80718184e-01 -9.30822432e-01
-8.14841688e-01 -3.80996972e-01 5.79137683e-01 2.33630627e-01
-4.00913388e-01 7.82396972e-01 3.88436139e-01 -3.90368670e-01
2.41142929e-01 -6.34113669e-01 -5.72006702e-01 -7.37612426e-01
2.89823741e-01 8.96410644e-01 -3.01429212e-01 -3.77642274e-01
2.87548184e-01 1.22332163e-02 -1.56435764e+00 6.63118780e-01
1.32549715e+00 5.16085625e-01 1.21391587e-01 8.01843524e-01
-5.11419535e-01 1.12217140e+00 -6.23731494e-01 -3.14409494e-01
1.55494273e-01 -3.85142952e-01 -9.11049187e-01 2.28880882e-01
5.38501799e-01 -4.12694484e-01 3.42461728e-02 9.35783565e-01
2.12075889e-01 4.51233804e-01 5.87927043e-01 -1.04091275e+00
-8.51090252e-01 8.09526563e-01 -4.15179521e-01 3.80507484e-02
2.73447514e-01 -3.98784280e-02 8.89500558e-01 -1.18590629e+00
4.41030741e-01 1.34013796e+00 9.22843158e-01 9.04671490e-01
-1.47742915e+00 -4.08398271e-01 3.78635883e-01 8.44917148e-02
-8.65147889e-01 -3.97260040e-01 1.00584340e+00 -3.94658744e-01
8.89081836e-01 2.88472652e-01 4.78328258e-01 1.09826851e+00
-3.10511410e-01 1.23503411e+00 1.14012229e+00 -8.28124404e-01
3.65614831e-01 4.37788934e-01 8.81476760e-01 8.86976361e-01
-1.71220109e-01 -6.07503317e-02 -6.81032419e-01 -4.95519191e-01
2.47537658e-01 9.25526470e-02 -1.72947258e-01 -1.67172924e-01
-9.04474318e-01 1.18638718e+00 1.91987008e-01 1.23336695e-01
-4.41824406e-01 -2.47194633e-01 3.95933926e-01 6.44052148e-01
1.15315187e+00 2.53633410e-01 -7.39084482e-01 -2.75850534e-01
-6.87020838e-01 1.60230830e-01 1.22971857e+00 8.61090839e-01
8.71813238e-01 -1.56763904e-02 1.11335948e-01 1.21738517e+00
5.13250113e-01 3.31230253e-01 7.14084268e-01 -9.01691854e-01
6.58416152e-01 8.24615479e-01 -1.40321136e-01 -6.30173206e-01
-2.52606452e-01 -2.09512368e-01 -8.74936640e-01 1.83932692e-01
4.84139383e-01 -4.11458045e-01 -9.49383199e-01 1.62400532e+00
6.03662670e-01 2.44349778e-01 2.33005971e-01 3.52873057e-01
8.55803967e-01 9.38201308e-01 -9.98327043e-03 -4.09269094e-01
8.81408274e-01 -1.34510171e+00 -7.60361493e-01 -3.14297974e-01
1.16575503e+00 -4.86722410e-01 1.41275942e+00 8.51387441e-01
-9.09925401e-01 -5.67468882e-01 -6.66245639e-01 -2.50847936e-01
-2.72544533e-01 -1.51331857e-01 7.96776474e-01 7.37980485e-01
-1.06642604e+00 5.41584432e-01 -7.79057443e-01 -3.81557383e-02
6.40543044e-01 4.59122390e-01 -2.49630705e-01 -1.60252541e-01
-6.65492058e-01 1.02450681e+00 3.55438054e-01 -3.81734297e-02
-8.96162570e-01 -9.21740413e-01 -1.02949977e+00 -3.26758958e-02
1.24663547e-01 -6.20183289e-01 1.64579737e+00 -1.32252383e+00
-1.55972445e+00 8.84251714e-01 -3.43477011e-01 -4.71799314e-01
3.26487869e-01 -4.12787408e-01 1.40696019e-01 -1.05212569e-01
-1.41204670e-01 6.87552392e-01 9.27176178e-01 -1.32089484e+00
-5.68925202e-01 -5.28409004e-01 -1.72987983e-01 4.32300955e-01
-8.77681315e-01 -2.65417825e-02 2.27248311e-01 -7.28368044e-01
9.54053104e-02 -7.53289759e-01 -6.19502008e-01 -2.02488571e-01
-3.26441586e-01 -7.78077364e-01 1.11729205e+00 -6.35329485e-01
9.46909010e-01 -1.87173033e+00 2.40333185e-01 2.35355332e-01
4.01656568e-01 2.05670744e-01 -2.48663142e-01 1.23765312e-01
-1.21686094e-01 3.52452733e-02 -3.60605836e-01 -8.72236371e-01
-1.36799246e-01 5.15796065e-01 -4.19244021e-01 5.13717942e-02
1.49918377e-01 6.36590600e-01 -1.01128626e+00 -4.84834194e-01
2.17249990e-01 1.75874814e-01 -8.48394156e-01 5.62310934e-01
-7.37867832e-01 4.71301287e-01 -3.96306306e-01 3.04230869e-01
4.94626254e-01 -3.15426141e-01 1.91088632e-01 1.61285400e-01
1.53988883e-01 5.19095719e-01 -9.20748472e-01 1.73941553e+00
-5.24184406e-01 5.48691988e-01 2.80101448e-01 -1.51375282e+00
1.01273632e+00 2.71080911e-01 3.12325150e-01 -2.49070209e-02
1.46993876e-01 -1.25737220e-01 -2.66227275e-01 -6.31823421e-01
4.09077734e-01 -1.95540935e-01 1.07834928e-01 1.07663965e+00
2.77520776e-01 -4.98391867e-01 2.83686221e-01 4.46494132e-01
1.04673326e+00 -8.84113740e-03 1.46152273e-01 -3.58550847e-01
3.02243561e-01 3.03760678e-01 3.79416138e-01 7.80613363e-01
-6.67584911e-02 4.32348102e-01 2.34395057e-01 -7.75733531e-01
-1.02712131e+00 -6.61480367e-01 -1.88683614e-01 1.73313510e+00
-4.65646982e-01 -4.27637547e-01 -8.98092031e-01 -1.17571831e+00
-1.47359252e-01 8.65426481e-01 -5.71091473e-01 7.85762370e-02
-5.88733554e-01 -9.91525114e-01 9.20015872e-02 5.89416683e-01
2.07149476e-01 -1.14287663e+00 5.10715842e-02 2.55940080e-01
-1.91589668e-01 -6.80122018e-01 -4.64275599e-01 6.85853839e-01
-1.32695043e+00 -8.75164509e-01 -3.51127982e-01 -1.42906594e+00
1.22217953e+00 4.02151853e-01 1.44000757e+00 5.99960268e-01
-6.77385107e-02 3.64921421e-01 -3.33997399e-01 -7.40015984e-01
-8.52090299e-01 -1.35684386e-01 2.05012619e-01 -2.79615492e-01
7.68431544e-01 -6.21479571e-01 2.78164111e-02 -2.28016563e-02
-6.86138213e-01 3.61947507e-01 2.21947700e-01 1.45394802e+00
3.81102353e-01 -9.62161794e-02 8.22995603e-01 -1.44740045e+00
9.40885663e-01 -5.70891559e-01 -4.02689695e-01 2.72721618e-01
-4.99652088e-01 2.37334713e-01 7.78702736e-01 -6.82593465e-01
-1.55710101e+00 2.57477462e-01 -3.68508697e-01 -6.47697598e-02
-4.10969555e-01 7.68183172e-01 1.28810123e-01 1.88575345e-04
1.24738395e+00 1.68719739e-01 3.29718560e-01 -4.75501329e-01
5.83577991e-01 1.02082527e+00 2.14555353e-01 -8.80639911e-01
5.83414495e-01 2.25682139e-01 -7.37464488e-01 -1.14992654e+00
-1.28041196e+00 -4.74460810e-01 -8.96563232e-01 -1.24436475e-01
4.90208656e-01 -8.51997495e-01 -4.01987910e-01 5.31454444e-01
-1.23466110e+00 -6.28898621e-01 -4.75035727e-01 4.66937333e-01
-4.13476020e-01 4.73535210e-01 -8.81051481e-01 -8.13786626e-01
-5.35833597e-01 -7.96251237e-01 7.47142494e-01 1.51472166e-01
-4.08964634e-01 -1.32048035e+00 4.54168379e-01 8.74448717e-01
1.33949757e-01 -4.36755806e-01 1.06921029e+00 -1.24528658e+00
-2.69948125e-01 -2.38332257e-01 3.49690467e-01 8.48363817e-01
5.93897514e-02 -3.37792113e-02 -1.31984794e+00 -1.58320725e-01
4.99991298e-01 -9.93190289e-01 6.88112438e-01 3.60872120e-01
1.40092170e+00 -5.12515068e-01 -2.42805704e-01 3.24252725e-01
7.20600188e-01 6.77559450e-02 1.75890625e-01 -2.27089569e-01
8.89137447e-01 8.70936096e-01 3.73192519e-01 3.70135367e-01
3.67485344e-01 1.31104812e-01 6.35464967e-04 -1.85436144e-01
2.10385665e-01 -1.59916565e-01 4.48634267e-01 1.37844920e+00
1.62792265e-01 -2.43234355e-02 -8.23698401e-01 3.62503648e-01
-2.00244737e+00 -1.02769220e+00 -1.63324043e-01 1.73770380e+00
1.37127221e+00 5.23329750e-02 9.52140763e-02 1.36536181e-01
6.38363421e-01 -1.98960617e-01 -4.57406670e-01 -3.30105722e-01
1.81330413e-01 2.70522743e-01 -2.58619577e-01 8.16764534e-01
-1.11195767e+00 8.12308609e-01 7.16968012e+00 5.16785204e-01
-5.72245717e-01 8.11259896e-02 1.20374608e+00 -4.08209302e-02
-5.84489107e-01 -3.49286377e-01 -8.43862951e-01 2.34619483e-01
9.69942510e-01 -1.37771204e-01 3.12181860e-01 1.29043937e+00
3.46125633e-01 2.44004335e-02 -1.38674545e+00 8.61146271e-01
3.27860385e-01 -1.57048774e+00 -8.36189762e-02 -1.94808140e-01
1.19378424e+00 1.45737052e-01 -1.15924172e-01 6.22153878e-01
1.04807675e+00 -7.82166302e-01 1.12493455e-01 7.86428005e-02
-7.39311054e-02 -4.95566189e-01 6.87067091e-01 9.12252486e-01
-5.01918137e-01 -6.94929361e-02 -3.48394483e-01 -2.92939812e-01
-1.44227326e-01 6.65859997e-01 -1.08144069e+00 -2.15890542e-01
5.86623847e-01 8.32062185e-01 -4.48209316e-01 6.52924359e-01
-2.83584893e-01 9.77648735e-01 -6.18151426e-01 -4.42728370e-01
1.47447214e-01 -4.91234511e-01 2.43449330e-01 1.06497300e+00
-8.13044682e-02 4.10637051e-01 5.14794350e-01 7.14900792e-01
-1.83724940e-01 2.30849147e-01 -8.22647691e-01 2.38105401e-01
2.82055497e-01 1.26044762e+00 -5.12857497e-01 -5.60977280e-01
-5.44517159e-01 5.71798861e-01 6.69082522e-01 2.28174239e-01
-3.06480020e-01 7.49457181e-02 2.45480567e-01 -4.08852160e-01
-1.21711999e-01 -2.13041067e-01 -8.31299245e-01 -1.32214022e+00
-5.50087094e-02 -1.36835873e+00 5.03593147e-01 -6.31339073e-01
-1.75335932e+00 5.36267817e-01 -4.53605950e-02 -8.65028799e-01
-4.84032959e-01 -7.09179819e-01 -9.18649733e-01 6.94297254e-01
-9.56626117e-01 -9.27558005e-01 -3.91005844e-01 7.43293166e-01
9.39994991e-01 -3.75833780e-01 1.21910810e+00 -1.93726812e-02
-6.01387441e-01 2.95950085e-01 3.20671767e-01 8.21181312e-02
4.83548105e-01 -1.59584844e+00 2.84218967e-01 6.24319792e-01
4.90817249e-01 6.13353074e-01 6.79993927e-01 -6.70638025e-01
-1.20497215e+00 -6.89019144e-01 9.23411071e-01 -6.35581195e-01
6.92165077e-01 -6.28969967e-01 -1.34200704e+00 9.16433096e-01
4.64902788e-01 -1.53166458e-01 1.14252257e+00 7.51995683e-01
-1.86672583e-01 3.96105051e-02 -1.34063053e+00 5.04085660e-01
7.64329374e-01 -5.45366168e-01 -1.13964081e+00 1.00898945e+00
7.06966281e-01 -1.94764793e-01 -5.41220605e-01 9.53684822e-02
5.87293059e-02 -6.33877993e-01 8.18465233e-01 -1.12842810e+00
4.16521966e-01 1.60233259e-01 1.29568383e-01 -1.77572119e+00
7.62720779e-02 -6.98862433e-01 -2.31960535e-01 9.97200131e-01
6.96688652e-01 -4.58943576e-01 1.49428284e+00 1.01902020e+00
-3.63835394e-01 -5.35885334e-01 -6.02229059e-01 -2.59238571e-01
2.00303003e-01 -5.87893069e-01 1.15624117e-02 1.25629640e+00
5.19486248e-01 7.26183176e-01 -2.17578739e-01 -2.04840750e-01
8.41386080e-01 1.66193005e-02 8.11027408e-01 -1.44588542e+00
-1.78548887e-01 -8.51397887e-02 7.41155669e-02 -1.43415987e+00
6.17482901e-01 -9.44405973e-01 5.38912594e-01 -1.32937706e+00
2.71735132e-01 -7.09720135e-01 6.52584285e-02 7.14905620e-01
-5.10712922e-01 -3.83759513e-02 -2.16618612e-01 1.98876023e-01
-6.26389146e-01 7.91249812e-01 9.90547836e-01 -3.33634406e-01
-4.25401360e-01 2.96516955e-01 -6.54849410e-01 1.19529235e+00
8.51759493e-01 -5.67759991e-01 -6.82735562e-01 -5.24662733e-01
-1.02302833e-02 -2.09136575e-01 -9.11919028e-02 -5.20425379e-01
2.84975588e-01 -2.27876812e-01 3.58358413e-01 -7.35099673e-01
3.68550271e-01 -8.56840968e-01 -5.51918626e-01 3.77589732e-01
-7.92648435e-01 5.87148555e-02 8.43932927e-02 4.82004791e-01
-2.76067466e-01 -7.50589192e-01 7.95241356e-01 -2.39245981e-01
-5.58917284e-01 8.17619264e-02 -5.37484586e-01 1.95584461e-01
1.06617594e+00 1.96587205e-01 -3.71858656e-01 -6.28631294e-01
-1.14164913e+00 3.92832756e-01 -2.92197857e-02 2.10384414e-01
7.05647409e-01 -1.15987182e+00 -8.93519938e-01 4.36089724e-01
-1.45334825e-01 4.14020061e-01 5.28298169e-02 4.54835057e-01
-4.03374344e-01 2.72911545e-02 1.80256158e-01 -8.10066819e-01
-1.68809903e+00 5.08780241e-01 1.35177329e-01 -3.31466734e-01
-5.23250699e-01 1.26484156e+00 1.38180971e-01 -1.16230631e+00
7.23170877e-01 -3.77321005e-01 -1.80489644e-01 8.14252272e-02
8.55831146e-01 1.73354506e-01 1.20533943e-01 5.70738222e-03
8.80938843e-02 -3.85792144e-02 -5.46861470e-01 7.78638422e-02
1.69788265e+00 4.56735902e-02 -1.34858534e-01 4.36493814e-01
1.05714202e+00 -3.34825993e-01 -1.22969806e+00 -5.05406082e-01
1.78801149e-01 -2.11667061e-01 6.97316974e-02 -6.69270337e-01
-6.91986144e-01 8.74693274e-01 3.55977595e-01 6.14548922e-01
8.49721670e-01 -1.83573328e-02 4.06746417e-01 1.17098367e+00
3.88813801e-02 -1.28680670e+00 5.66798091e-01 7.62485921e-01
7.66186416e-01 -1.74654365e+00 -1.76191494e-01 -4.68336642e-01
-7.25674093e-01 1.22229671e+00 8.52176368e-01 4.94490080e-02
8.59266222e-01 5.80126166e-01 4.14141029e-01 -1.72276780e-01
-1.08832252e+00 2.93676913e-01 4.56349589e-02 6.99330151e-01
5.75259805e-01 -9.56546441e-02 7.32804388e-02 4.07626510e-01
-2.37384111e-01 7.77350040e-03 4.15191263e-01 1.08929694e+00
-7.74624109e-01 -1.23304164e+00 -4.29723293e-01 8.33880484e-01
-2.28234798e-01 -5.45284688e-01 -3.77243251e-01 3.74326617e-01
-3.11688900e-01 1.24399805e+00 1.53017715e-01 -2.87629992e-01
1.34827614e-01 4.72686857e-01 3.59638035e-01 -1.20695794e+00
-8.36156726e-01 -2.56221086e-01 3.73181760e-01 -1.45628816e-02
-6.13245964e-01 -5.34648895e-01 -1.19126892e+00 -4.54318017e-01
-5.16722977e-01 5.47970474e-01 6.13957226e-01 1.25884771e+00
1.63830787e-01 1.42114714e-01 8.07522118e-01 -9.54522967e-01
-1.02012336e+00 -1.35550714e+00 -1.32174656e-01 5.93788445e-01
2.36632511e-01 -1.27549559e-01 -5.71796834e-01 5.45629740e-01] | [10.893287658691406, 8.284479141235352] |
080bfee0-1f33-48a3-b3cd-f1f1be93f9b7 | speech-separation-based-on-multi-stage | 2008.03149 | null | https://arxiv.org/abs/2008.03149v1 | https://arxiv.org/pdf/2008.03149v1.pdf | Speech Separation Based on Multi-Stage Elaborated Dual-Path Deep BiLSTM with Auxiliary Identity Loss | Deep neural network with dual-path bi-directional long short-term memory (BiLSTM) block has been proved to be very effective in sequence modeling, especially in speech separation. This work investigates how to extend dual-path BiLSTM to result in a new state-of-the-art approach, called TasTas, for multi-talker monaural speech separation (a.k.a cocktail party problem). TasTas introduces two simple but effective improvements, one is an iterative multi-stage refinement scheme, and the other is to correct the speech with imperfect separation through a loss of speaker identity consistency between the separated speech and original speech, to boost the performance of dual-path BiLSTM based networks. TasTas takes the mixed utterance of two speakers and maps it to two separated utterances, where each utterance contains only one speaker's voice. Our experiments on the notable benchmark WSJ0-2mix data corpus result in 20.55dB SDR improvement, 20.35dB SI-SDR improvement, 3.69 of PESQ, and 94.86\% of ESTOI, which shows that our proposed networks can lead to big performance improvement on the speaker separation task. We have open sourced our re-implementation of the DPRNN-TasNet here (https://github.com/ShiZiqiang/dual-path-RNNs-DPRNNs-based-speech-separation), and our TasTas is realized based on this implementation of DPRNN-TasNet, it is believed that the results in this paper can be reproduced with ease. | ['Jiqing Han', 'Rujie Liu', 'Ziqiang Shi'] | 2020-08-06 | null | null | null | null | ['speaker-separation'] | ['speech'] | [-1.10436566e-01 -1.63906217e-01 1.77534446e-01 -2.10539281e-01
-1.18888688e+00 -3.29291075e-01 9.63596106e-02 -6.08958781e-01
-2.64908075e-01 4.34182882e-01 3.77073139e-01 -6.59432948e-01
1.36473149e-01 -2.40433719e-02 -5.64568520e-01 -1.04052162e+00
5.66125512e-02 1.99732155e-01 2.35154822e-01 -4.43264663e-01
-1.44584119e-01 2.34382495e-01 -1.42669332e+00 3.72245252e-01
7.47866631e-01 9.24212277e-01 3.15942585e-01 1.11113870e+00
-2.08097160e-01 7.79946208e-01 -8.62928808e-01 -3.92035693e-01
2.29739860e-01 -5.89381337e-01 -7.33824253e-01 -5.35192609e-01
3.46193135e-01 -5.57523705e-02 -4.26719397e-01 1.05289698e+00
1.08433294e+00 1.71111926e-01 3.12541813e-01 -9.63850260e-01
-4.45236385e-01 1.11060655e+00 -6.56059206e-01 5.52690208e-01
-9.54875052e-02 8.53200778e-02 9.01417971e-01 -7.05169857e-01
-2.33025059e-01 1.33128524e+00 8.07697415e-01 6.92040503e-01
-8.37597370e-01 -1.19291854e+00 -6.61645159e-02 3.77061844e-01
-1.33684897e+00 -1.05718625e+00 4.97994810e-01 1.56385433e-02
1.30013895e+00 7.20327914e-01 3.93163115e-01 1.11629546e+00
-4.04441543e-02 8.45379055e-01 9.24917758e-01 -5.19429505e-01
-1.68123960e-01 -7.01813996e-02 5.81210911e-01 3.25310677e-01
-2.71934986e-01 3.11656058e-01 -6.92641020e-01 2.07329348e-01
4.30309206e-01 -4.54790801e-01 -3.48057032e-01 6.27489150e-01
-9.47784901e-01 4.11930352e-01 3.06725234e-01 7.33044147e-01
-1.23318687e-01 9.21730548e-02 5.25406778e-01 4.66312170e-01
5.39799213e-01 7.81254247e-02 -4.71932769e-01 -4.36210543e-01
-1.12632847e+00 2.12330185e-02 7.35721707e-01 8.99914145e-01
1.85655847e-01 8.02415967e-01 -3.01157087e-01 1.31804729e+00
2.69132584e-01 8.86244774e-01 9.89801943e-01 -7.86182284e-01
4.93984491e-01 -3.34394544e-01 -2.91428924e-01 -5.18118322e-01
-2.42666945e-01 -8.03857744e-01 -8.52444053e-01 -2.56073307e-02
1.43300965e-01 -3.71823519e-01 -1.17412138e+00 1.73053253e+00
9.31328442e-03 5.06515205e-01 3.79677922e-01 9.07212675e-01
1.11188972e+00 1.08859944e+00 -3.15274864e-01 -3.52088302e-01
1.42181170e+00 -1.46247160e+00 -1.08717918e+00 -1.90985769e-01
2.38863498e-01 -1.10615277e+00 7.37905324e-01 5.68669200e-01
-1.20837569e+00 -7.89382577e-01 -1.08639336e+00 1.44780219e-01
-1.22857906e-01 1.56284243e-01 -2.09186394e-02 9.77155924e-01
-1.19870138e+00 3.44810426e-01 -7.15177655e-01 -1.03601247e-01
-7.96029493e-02 2.76473671e-01 -7.94541091e-02 2.87504971e-01
-1.39632058e+00 7.83647478e-01 2.34997123e-01 4.25461084e-01
-8.91920805e-01 -6.13670111e-01 -5.99505842e-01 2.30761975e-01
3.01632971e-01 -3.74201894e-01 1.67174399e+00 -9.57959235e-01
-1.97925901e+00 5.22249281e-01 -5.59113741e-01 -7.92907178e-01
1.38485610e-01 -3.11665386e-01 -9.53571141e-01 -2.06240207e-01
-9.52903926e-02 3.43543917e-01 8.95479918e-01 -1.11943793e+00
-7.84565806e-01 -2.36697122e-01 -5.95884025e-01 2.97649324e-01
-1.93278164e-01 5.41871369e-01 -3.24874848e-01 -8.37939739e-01
1.64464429e-01 -9.30642664e-01 1.25941008e-01 -9.95585561e-01
-6.35790944e-01 -6.60490170e-02 7.61609316e-01 -1.27865636e+00
1.22477484e+00 -2.31158209e+00 -5.13352342e-02 -1.27092078e-01
-1.54870823e-01 1.05005884e+00 -2.14962751e-01 1.19903229e-01
-2.50737160e-01 8.29703882e-02 -1.19077891e-01 -8.13783944e-01
2.37738322e-02 1.76621042e-02 -5.60854077e-01 1.90578967e-01
-2.25802913e-01 6.89733684e-01 -4.84425813e-01 -7.78471231e-02
1.59642950e-01 6.91924691e-01 -3.61017808e-02 2.13242605e-01
2.51612782e-01 3.69238377e-01 3.04786861e-01 2.50518262e-01
9.33395565e-01 5.11377454e-01 -1.39170378e-01 -1.03077658e-01
-3.51885408e-01 8.63117695e-01 -1.07547438e+00 1.64534104e+00
-4.49208975e-01 8.91009808e-01 3.74263495e-01 -7.54011869e-01
9.42727983e-01 8.77777934e-01 -1.35983145e-02 -7.45124280e-01
2.97976613e-01 4.31533724e-01 3.42373371e-01 -2.79376835e-01
3.51659805e-01 -4.64059770e-01 2.14043647e-01 1.70893788e-01
2.69666940e-01 -2.40553431e-02 -1.39695331e-01 -8.80315453e-02
7.73082614e-01 -3.12007904e-01 -2.25105695e-02 -1.09368071e-01
5.66777706e-01 -5.39551735e-01 6.76522136e-01 6.29509568e-01
-3.39602172e-01 7.50880361e-01 4.92487065e-02 2.11837813e-01
-8.21915209e-01 -1.12014198e+00 1.68997765e-01 1.21185696e+00
-2.58582115e-01 -1.34908840e-01 -9.35066104e-01 -5.68599887e-02
-4.55031991e-01 9.99999821e-01 3.50146592e-02 -1.14363626e-01
-7.72701204e-01 -6.26474977e-01 1.28691888e+00 3.94850582e-01
5.40446520e-01 -9.94776309e-01 6.09122254e-02 2.72089183e-01
-4.58969027e-01 -1.12836885e+00 -8.62161040e-01 3.81844997e-01
-5.64652443e-01 -3.41165811e-01 -1.05820310e+00 -1.04587209e+00
-9.29202661e-02 5.23289025e-01 7.45353222e-01 -3.50957751e-01
2.08513975e-01 -2.31942460e-01 -3.83588731e-01 -4.92006809e-01
-6.58378959e-01 9.98349786e-02 4.03998673e-01 5.75190503e-03
3.69045138e-01 -6.92901909e-01 -2.06565619e-01 2.68294394e-01
-4.28457201e-01 -4.67364378e-02 4.26861227e-01 7.50522733e-01
2.84695983e-01 2.61394024e-01 6.73375964e-01 -2.85572559e-01
5.85586011e-01 -1.41250744e-01 -3.61633122e-01 8.48364905e-02
-3.74784529e-01 -6.52414262e-02 6.26143396e-01 -3.45456809e-01
-1.36834395e+00 -3.84008199e-01 -9.93895531e-01 -6.48591101e-01
-1.89946011e-01 3.37988883e-01 -3.50518018e-01 2.11137816e-01
2.66215473e-01 5.04933536e-01 -5.46136834e-02 -7.40069151e-01
2.35881940e-01 1.31709123e+00 7.27178514e-01 -2.04992622e-01
5.07063627e-01 5.77525757e-02 -7.68439651e-01 -1.03930640e+00
-6.69725299e-01 -6.83997989e-01 -2.24594399e-01 1.89818308e-01
7.65668273e-01 -1.03199673e+00 -8.19844902e-01 9.75570560e-01
-1.27562523e+00 -4.19498950e-01 -2.66105658e-03 7.53313661e-01
-2.15756997e-01 3.07426870e-01 -9.52158689e-01 -1.06578732e+00
-8.73648465e-01 -1.22997093e+00 6.21177375e-01 3.19159329e-01
3.07038496e-03 -6.24702632e-01 1.68860182e-01 6.25816464e-01
6.52026951e-01 -6.36220038e-01 4.24761325e-01 -9.17562544e-01
-2.02262640e-01 1.74219579e-01 9.76312831e-02 1.04248428e+00
1.79006398e-01 -1.63693763e-02 -1.63191628e+00 -3.01272303e-01
4.96863395e-01 2.45968983e-01 9.10790622e-01 7.94220686e-01
6.86204672e-01 -3.75671864e-01 -8.60197172e-02 6.66122377e-01
8.78507316e-01 6.47935331e-01 8.09954584e-01 5.76359732e-03
8.61792386e-01 3.19202274e-01 4.25885916e-01 -1.30552843e-01
3.19223940e-01 7.50297666e-01 1.09079033e-01 -3.11789155e-01
-6.33233607e-01 1.18592799e-01 8.40726614e-01 1.76955986e+00
5.50747775e-02 -3.98822457e-01 -8.51539254e-01 5.64924300e-01
-1.54604936e+00 -1.04177666e+00 -4.91972029e-01 2.20169425e+00
9.59432185e-01 1.97509587e-01 2.82890379e-01 4.33794439e-01
9.89306748e-01 2.52385318e-01 -1.93801001e-01 -7.99745321e-01
-2.22136840e-01 4.52946544e-01 5.33232212e-01 8.07546318e-01
-8.64863217e-01 1.07009101e+00 5.53025103e+00 1.47996569e+00
-1.52069032e+00 5.75192511e-01 5.29617369e-01 -2.80743539e-01
9.30191576e-02 -3.51512998e-01 -9.55004692e-01 5.33929884e-01
1.68722677e+00 -6.81645982e-03 4.58243102e-01 3.20847988e-01
3.99786502e-01 2.20968574e-01 -7.69085407e-01 1.19223559e+00
2.23964512e-01 -8.53395522e-01 -1.45086691e-01 -2.41429791e-01
2.87666768e-01 3.03446501e-01 2.60094523e-01 5.05915642e-01
3.00241172e-01 -9.77767229e-01 9.88065422e-01 8.57521519e-02
6.62193716e-01 -9.77423906e-01 9.23599064e-01 3.66846740e-01
-1.26687288e+00 -6.79511800e-02 -1.45294383e-01 1.30781144e-01
2.89220303e-01 6.41330659e-01 -9.83940661e-01 7.83969522e-01
9.55848515e-01 2.75326520e-01 -5.38145676e-02 1.01194787e+00
-2.43356362e-01 1.23900735e+00 -1.34734422e-01 1.69722438e-01
2.51887292e-01 -4.20163013e-02 1.05434906e+00 1.56453049e+00
4.60862041e-01 -1.04284272e-01 -3.29467654e-01 6.51762903e-01
1.66490257e-01 -7.42413849e-02 -1.08269624e-01 1.19859673e-01
5.47154486e-01 1.04015541e+00 -3.50707561e-01 -4.36685711e-01
-8.20277929e-02 9.60270405e-01 -1.23378076e-01 4.60734338e-01
-1.08796632e+00 -5.74992061e-01 9.29412305e-01 -3.53385091e-01
5.65316975e-01 -1.34662151e-01 -3.12798172e-01 -7.91198373e-01
-4.19130325e-02 -1.19646442e+00 1.61399856e-01 -8.65557909e-01
-1.04777300e+00 1.13787138e+00 -4.03549850e-01 -9.40298200e-01
-1.69408932e-01 -4.00448412e-01 -6.72177136e-01 1.31627834e+00
-1.48665845e+00 -1.10781395e+00 1.53695434e-01 5.10723710e-01
8.80675316e-01 -3.55716288e-01 7.26086795e-01 7.11526215e-01
-8.40466380e-01 9.53009367e-01 3.10047954e-01 2.65757889e-01
8.14024508e-01 -1.19225585e+00 7.82240331e-01 1.19946313e+00
2.94164181e-01 4.91299331e-01 9.40484226e-01 -2.92902797e-01
-9.42714036e-01 -9.99955535e-01 8.69091749e-01 -9.86299738e-02
6.03856504e-01 -5.19066334e-01 -9.66804624e-01 6.72591627e-01
5.74988961e-01 -4.94629413e-01 6.72160983e-01 4.03988473e-02
-3.32861632e-01 -5.74743927e-01 -7.63244569e-01 5.99972785e-01
7.04744101e-01 -6.00586951e-01 -7.48494685e-01 -1.34097561e-01
1.31744087e+00 -4.62945461e-01 -4.17944849e-01 2.89933592e-01
4.26974148e-01 -1.09561396e+00 9.46484447e-01 -1.02680139e-01
-2.04891339e-01 -4.22845751e-01 -2.31201872e-01 -1.66296089e+00
-1.13447875e-01 -9.23316777e-01 1.05038784e-01 1.68261063e+00
5.23402333e-01 -8.28015745e-01 2.04910889e-01 -7.32168481e-02
-8.07518125e-01 -1.99572846e-01 -1.16457331e+00 -1.13554883e+00
9.14348513e-02 -6.72503948e-01 5.63970804e-01 7.30909109e-01
-1.14955857e-01 5.00608504e-01 -7.03653038e-01 4.31665480e-01
3.86702508e-01 -3.73746037e-01 4.82244819e-01 -6.65138662e-01
-5.21220446e-01 -5.70437253e-01 -2.73697153e-02 -1.16199756e+00
1.55639917e-01 -7.25064635e-01 4.47990507e-01 -1.30750346e+00
-2.36165956e-01 -3.13154817e-01 -6.25599980e-01 5.13115287e-01
-2.42023975e-01 1.23033397e-01 2.71155626e-01 -9.21076015e-02
-2.37712249e-01 5.64218462e-01 7.98654854e-01 -1.77384466e-01
-4.37868267e-01 4.10385966e-01 -6.32258236e-01 5.09846926e-01
1.00913668e+00 -6.07236922e-01 -2.29293272e-01 -4.71526444e-01
-5.53135693e-01 2.28063345e-01 9.33116227e-02 -1.29549217e+00
3.17936301e-01 3.79226267e-01 -3.00127178e-01 -7.09092975e-01
7.85611987e-01 -4.93092597e-01 2.93375582e-01 5.95759511e-01
-1.57147199e-01 -3.35703790e-01 5.95675111e-01 8.45882222e-02
-5.17830193e-01 -2.03657985e-01 8.08226347e-01 -4.94812848e-03
-4.69946980e-01 -1.36425346e-01 -5.61218083e-01 -1.55968592e-01
4.50847626e-01 -5.89930788e-02 -3.34659487e-01 -4.11615729e-01
-6.99873209e-01 -1.85383530e-03 -3.30872744e-01 4.89473909e-01
3.20568651e-01 -1.23408997e+00 -1.00946963e+00 1.85632631e-01
-4.51384276e-01 -2.57955939e-01 5.86083412e-01 1.02946997e+00
-2.85243750e-01 6.86242938e-01 3.22512514e-03 -6.76735163e-01
-1.85901403e+00 2.73183972e-01 6.76804483e-01 2.32356228e-02
-4.47708040e-01 1.43090808e+00 2.29778796e-01 -5.12754381e-01
5.15766442e-01 -4.75919306e-01 -1.12641878e-01 -1.08183421e-01
6.01333499e-01 5.95861852e-01 4.68915135e-01 -8.15176427e-01
-4.51457232e-01 3.17554951e-01 -1.32097289e-01 -4.30271417e-01
1.16770327e+00 -3.54157716e-01 -1.91491634e-01 8.38824749e-01
1.12955570e+00 3.51834029e-01 -7.52225935e-01 -2.88241535e-01
-2.38738507e-01 -1.02136835e-01 2.73272067e-01 -9.80832756e-01
-1.23291409e+00 1.16478539e+00 8.67765307e-01 3.14658880e-01
1.24143577e+00 -2.18503714e-01 1.25770128e+00 -1.27552068e-02
-4.93561216e-02 -8.46228063e-01 -1.55961171e-01 9.48309183e-01
8.12916636e-01 -9.89432991e-01 -7.09604263e-01 -1.39767915e-01
-7.40068436e-01 9.17925000e-01 3.88672888e-01 2.35034987e-01
4.74513263e-01 5.54923415e-01 5.87759793e-01 1.82626054e-01
-6.63638532e-01 -2.97087938e-01 2.07219794e-01 4.09565121e-01
5.42089403e-01 3.59100342e-01 9.00542364e-02 8.17254245e-01
-6.95613980e-01 -4.74321216e-01 3.87204885e-01 4.17473316e-01
-4.89982873e-01 -9.74510193e-01 -8.19391191e-01 3.38072069e-02
-7.17180192e-01 -6.44073546e-01 -7.15160519e-02 2.58670807e-01
1.23308320e-02 1.40392339e+00 3.38909663e-02 -6.03323996e-01
3.45568389e-01 3.15293700e-01 1.44357860e-01 -4.45843428e-01
-8.06940675e-01 6.86272621e-01 1.81398183e-01 -1.40486628e-01
-2.41999999e-01 -5.57604551e-01 -1.32880425e+00 -4.57629561e-01
-4.63310838e-01 4.31754142e-01 8.45868170e-01 9.33689296e-01
2.64305502e-01 1.12290001e+00 5.19804657e-01 -7.79964387e-01
-4.53906834e-01 -1.41925621e+00 -7.18173683e-01 -2.27661043e-01
8.41737866e-01 -2.12091625e-01 -5.15868008e-01 -4.60840911e-02] | [14.86534309387207, 6.041489601135254] |
1de34d0d-4de0-4480-9310-f49c76de60d2 | learning-hand-held-object-reconstruction-from | 2305.03036 | null | https://arxiv.org/abs/2305.03036v1 | https://arxiv.org/pdf/2305.03036v1.pdf | Learning Hand-Held Object Reconstruction from In-The-Wild Videos | Prior works for reconstructing hand-held objects from a single image rely on direct 3D shape supervision which is challenging to gather in real world at scale. Consequently, these approaches do not generalize well when presented with novel objects in in-the-wild settings. While 3D supervision is a major bottleneck, there is an abundance of in-the-wild raw video data showing hand-object interactions. In this paper, we automatically extract 3D supervision (via multiview 2D supervision) from such raw video data to scale up the learning of models for hand-held object reconstruction. This requires tackling two key challenges: unknown camera pose and occlusion. For the former, we use hand pose (predicted from existing techniques, e.g. FrankMocap) as a proxy for object pose. For the latter, we learn data-driven 3D shape priors using synthetic objects from the ObMan dataset. We use these indirect 3D cues to train occupancy networks that predict the 3D shape of objects from a single RGB image. Our experiments on the MOW and HO3D datasets show the effectiveness of these supervisory signals at predicting the 3D shape for real-world hand-held objects without any direct real-world 3D supervision. | ['Saurabh Gupta', 'Matthew Jin', 'Matthew Chang', 'Aditya Prakash'] | 2023-05-04 | null | null | null | null | ['object-reconstruction'] | ['computer-vision'] | [ 1.43201232e-01 -1.88892670e-02 -1.72061220e-01 -3.50357890e-01
-7.26108849e-01 -7.03332067e-01 3.61013204e-01 -4.29454029e-01
-2.96072274e-01 3.64012331e-01 1.70422956e-01 -7.31282542e-03
3.58345918e-02 -2.90836245e-01 -1.27093995e+00 -6.32615447e-01
2.87243187e-01 9.13504004e-01 3.99823993e-01 -5.28440773e-02
3.78166474e-02 8.09085429e-01 -1.82204986e+00 2.64063179e-01
3.22308168e-02 1.07113373e+00 5.68156421e-01 8.56767297e-01
2.78014302e-01 5.43007612e-01 -2.34380901e-01 1.43976271e-01
7.09899306e-01 -2.45488454e-02 -5.04213512e-01 7.87095726e-01
7.80331314e-01 -8.95191371e-01 -5.55211961e-01 5.65047324e-01
5.19360483e-01 1.12278081e-01 4.40491259e-01 -1.25143600e+00
-2.75845081e-01 -1.94821432e-02 -5.81885874e-01 -8.36029500e-02
5.95603347e-01 4.64989364e-01 8.82385433e-01 -1.07887065e+00
8.79908741e-01 1.40892422e+00 5.08316219e-01 5.31014144e-01
-1.43601215e+00 -2.68721551e-01 3.64328176e-01 -2.03445852e-02
-1.31587052e+00 -5.28706908e-01 9.28378165e-01 -6.21291459e-01
8.64031613e-01 2.52943248e-01 9.98696685e-01 1.55978441e+00
-1.87279791e-01 9.75925744e-01 1.03645432e+00 -4.00965303e-01
2.78193414e-01 1.46567851e-01 -1.85532391e-01 6.45535171e-01
1.30078748e-01 7.12203681e-02 -9.16335225e-01 -7.62806162e-02
1.23986018e+00 5.39487898e-01 -4.27293330e-01 -1.13667715e+00
-1.32617354e+00 2.72143275e-01 4.55031574e-01 -2.66941249e-01
-3.55175197e-01 3.03832769e-01 -1.71708539e-01 -1.09177418e-01
3.70750666e-01 7.36660287e-02 -6.26176357e-01 -1.19080357e-01
-6.52889550e-01 5.30588865e-01 5.78517854e-01 1.34113717e+00
7.41310358e-01 -1.78083777e-01 -8.30956362e-03 3.97105813e-01
6.28576994e-01 8.36084247e-01 1.22108031e-02 -1.17807472e+00
6.17734730e-01 4.30680871e-01 4.03482288e-01 -6.39858186e-01
-2.73099780e-01 -1.77243561e-01 -2.27142632e-01 4.35722589e-01
6.02594495e-01 3.37806433e-01 -1.14887369e+00 1.48887515e+00
6.05269253e-01 2.57802494e-02 -3.51815015e-01 1.28468442e+00
5.92077851e-01 3.20112050e-01 -5.56024969e-01 3.00667696e-02
1.02574694e+00 -9.16241348e-01 -5.33760786e-01 -5.35877883e-01
1.11772470e-01 -3.93323570e-01 1.15014982e+00 5.07778347e-01
-9.03543234e-01 -4.39438969e-01 -8.74918401e-01 -2.04640925e-01
-1.18168853e-01 -1.07426852e-01 4.64997917e-01 2.24449769e-01
-6.40273094e-01 3.06310862e-01 -1.32330465e+00 -3.60991150e-01
4.07405257e-01 3.90616655e-01 -6.55199707e-01 -2.78102428e-01
-4.09809977e-01 7.91058183e-01 1.08085483e-01 3.55872303e-01
-1.31121707e+00 -5.29871225e-01 -9.66825962e-01 -3.30137104e-01
8.60599101e-01 -6.58377886e-01 1.21590745e+00 -5.18018007e-01
-1.33748555e+00 9.99865472e-01 -1.00626446e-01 -5.14011160e-02
6.73637509e-01 -4.23587859e-01 4.07861859e-01 1.92374423e-01
-5.69912232e-02 7.05083787e-01 1.08274329e+00 -1.84047711e+00
-2.26257011e-01 -1.09773171e+00 2.27364853e-01 2.71740824e-01
3.22168204e-03 -4.82597172e-01 -6.08594477e-01 -4.97244447e-01
5.41953266e-01 -1.17937183e+00 -2.58350130e-02 6.02672756e-01
-4.08201665e-01 1.60690039e-01 1.05237579e+00 -6.65101051e-01
3.08617324e-01 -1.99777412e+00 5.57846129e-01 3.95856681e-04
6.50058314e-02 -1.71969160e-01 7.50681609e-02 1.42305195e-01
2.15495840e-01 -2.94170082e-01 -5.70039488e-02 -8.27200294e-01
3.50170769e-02 5.28572500e-01 -4.02786851e-01 5.59173763e-01
2.28664979e-01 8.88317108e-01 -8.72350276e-01 -3.94169986e-01
3.40234607e-01 7.96645164e-01 -8.25015366e-01 7.39347696e-01
-6.03672683e-01 8.19357038e-01 -3.64832401e-01 9.89891171e-01
4.80036169e-01 -2.48863220e-01 2.15032790e-02 -2.72131592e-01
7.03190416e-02 8.44881237e-02 -1.27018738e+00 2.25474501e+00
-3.79918009e-01 2.37153336e-01 3.35491359e-01 -6.81018889e-01
5.77750266e-01 3.07193995e-01 4.91175145e-01 -1.48730978e-01
1.77067190e-01 5.55559881e-02 -2.02666610e-01 -6.65654480e-01
2.72790790e-01 -1.87717244e-01 1.47683248e-01 4.57440823e-01
4.03897971e-01 -5.53272307e-01 -4.76987600e-01 7.30162300e-03
1.08448732e+00 7.50451148e-01 2.96731368e-02 1.62502497e-01
-1.51497170e-01 -1.32341161e-01 3.85384828e-01 6.89362109e-01
-1.30100355e-01 1.18013942e+00 1.40454248e-01 -4.62787032e-01
-1.15672469e+00 -1.24888051e+00 1.49412593e-02 9.15718794e-01
1.91009000e-01 -2.11752713e-01 -4.56726164e-01 -7.10526824e-01
1.99963614e-01 2.42044419e-01 -5.45589268e-01 1.85423240e-01
-4.91475701e-01 -1.97325125e-01 -1.26264086e-02 7.46379137e-01
1.45714328e-01 -9.84491587e-01 -1.23292387e+00 2.87242085e-02
-2.81579971e-01 -1.41850376e+00 -4.64533597e-01 5.39336562e-01
-9.63687360e-01 -1.09052324e+00 -6.58264041e-01 -4.28274125e-01
7.44881451e-01 5.47185123e-01 7.75684237e-01 -1.65459067e-01
-5.09443879e-01 9.28595781e-01 -2.63856590e-01 -4.86678183e-01
1.13775961e-01 -2.30741769e-01 4.08126801e-01 1.20491281e-01
-5.46754710e-03 -8.19949687e-01 -5.13215423e-01 3.67629766e-01
-7.59214520e-01 4.26402688e-02 4.21089977e-01 7.66823530e-01
8.86240304e-01 -4.05166149e-01 -4.01360430e-02 -3.66119742e-01
-1.50805399e-01 -2.50699604e-03 -6.39954567e-01 1.57682270e-01
-1.75026339e-02 9.17560235e-02 2.64032036e-01 -6.53063238e-01
-9.39614654e-01 7.70974815e-01 1.46286920e-01 -1.14405417e+00
-3.72472644e-01 2.15855446e-02 -4.46327627e-01 1.56848207e-01
6.34335697e-01 9.21765044e-02 6.56884760e-02 -7.84300327e-01
1.65555745e-01 6.58096313e-01 6.76612020e-01 -7.74417579e-01
8.63008082e-01 8.81258905e-01 6.21657185e-02 -7.55325198e-01
-1.18976653e+00 -4.80696112e-01 -1.25376761e+00 -3.25881898e-01
9.10899937e-01 -1.00787675e+00 -7.81154394e-01 3.65589857e-01
-1.21930182e+00 -4.88302797e-01 -4.14313227e-01 4.91421461e-01
-9.69621599e-01 1.49165869e-01 -2.63482124e-01 -1.11317194e+00
8.95419493e-02 -1.45333135e+00 1.80944836e+00 -2.44138688e-01
-4.94208857e-02 -5.59862912e-01 -1.55207798e-01 7.57853746e-01
-1.56798795e-01 4.20863092e-01 4.67490911e-01 -1.99120536e-01
-1.08857644e+00 -3.15200448e-01 2.98408389e-01 3.11535418e-01
1.04527794e-01 -3.59719783e-01 -1.22669911e+00 -3.29011112e-01
2.59949982e-01 -7.95518160e-01 5.51255286e-01 2.10027784e-01
1.28103244e+00 -1.84757754e-01 -1.53072119e-01 7.37432301e-01
1.08685350e+00 -1.80232257e-01 2.05262646e-01 1.55487418e-01
1.02700520e+00 5.54662108e-01 7.52151251e-01 7.15750337e-01
4.93405223e-01 9.77851927e-01 8.83404911e-01 2.62831330e-01
-9.84545648e-02 -6.05723023e-01 3.12824667e-01 5.07628322e-01
-3.91290754e-01 -3.81821543e-02 -1.00229001e+00 3.07301044e-01
-1.87674916e+00 -5.46447515e-01 1.48390859e-01 2.26982641e+00
7.13998020e-01 1.03129342e-01 1.58638462e-01 1.86811432e-01
3.42078865e-01 9.70225856e-02 -9.48929071e-01 5.77943981e-01
1.28835171e-01 -1.39166061e-02 4.54261810e-01 5.23906767e-01
-8.78179908e-01 7.90543318e-01 5.22228765e+00 5.83467260e-02
-1.07371330e+00 1.33607715e-01 7.79518262e-02 -6.20390892e-01
-6.70900196e-02 8.82283077e-02 -9.33127522e-01 1.40173093e-03
2.79282928e-01 6.86436594e-01 6.79455042e-01 8.85923207e-01
5.97518273e-02 -1.97842807e-01 -1.72755420e+00 1.36363637e+00
3.78043026e-01 -8.53713512e-01 -8.82045329e-02 3.92505944e-01
4.72592175e-01 1.19077437e-01 -6.45457255e-03 -6.19621202e-02
7.48560503e-02 -7.61639059e-01 1.19155049e+00 5.52853703e-01
6.70060158e-01 -1.39568970e-01 3.15653175e-01 9.18382168e-01
-7.63536692e-01 4.61725751e-03 -1.57741189e-01 -1.44746140e-01
3.70026171e-01 3.34387243e-01 -8.33547413e-01 1.61088511e-01
8.86006236e-01 7.53095746e-01 -5.37866533e-01 6.78287506e-01
-2.18390331e-01 2.00333908e-01 -7.60231733e-01 4.01710749e-01
-1.38844233e-02 9.97998789e-02 6.04378998e-01 6.26041949e-01
1.21357240e-01 1.97154477e-01 2.89049059e-01 9.45448875e-01
-4.89197522e-02 -6.34719312e-01 -8.36534142e-01 1.24846213e-01
1.89753532e-01 9.48818743e-01 -6.68635011e-01 -7.49905631e-02
-1.82991207e-01 1.13691175e+00 3.64550948e-01 3.53649527e-01
-7.18246579e-01 1.86159298e-01 5.49846530e-01 4.66406316e-01
6.36281669e-01 -6.44208729e-01 -8.35248530e-02 -1.46182919e+00
5.68259478e-01 -7.65712202e-01 -1.17476635e-01 -1.11369956e+00
-1.18930221e+00 2.60709256e-01 2.87610799e-01 -1.26984882e+00
-2.59697109e-01 -9.66783047e-01 1.05208352e-01 5.71236372e-01
-1.20048213e+00 -1.24891531e+00 -7.03690648e-01 7.86532819e-01
7.40019798e-01 1.79400355e-01 7.10443258e-01 -9.81285125e-02
-1.78282082e-01 1.95799664e-01 -2.63705313e-01 -1.12069398e-01
7.67618775e-01 -1.09835458e+00 1.91252083e-01 3.88213664e-01
5.26635408e-01 4.70407784e-01 6.36695862e-01 -4.92613256e-01
-2.27481914e+00 -7.33618975e-01 8.43833461e-02 -1.43078470e+00
2.87831962e-01 -1.13312173e+00 -6.33486807e-01 1.09345305e+00
-4.49309617e-01 7.10659087e-01 2.18501627e-01 1.60738323e-02
-5.98297477e-01 -1.25674114e-01 -1.07402515e+00 3.20831805e-01
1.59904838e+00 -7.30482876e-01 -5.59519351e-01 4.89562154e-01
6.74818397e-01 -7.80790567e-01 -7.65248716e-01 3.48649591e-01
9.01473701e-01 -8.22256684e-01 1.25868368e+00 -6.89931452e-01
4.22415912e-01 -4.26076084e-01 -6.85554445e-01 -1.05332637e+00
1.43881038e-01 -3.00023466e-01 -5.39414883e-01 7.49558866e-01
-5.10427691e-02 -7.69774169e-02 9.77226257e-01 7.96936512e-01
1.05027340e-01 -6.56333089e-01 -9.17241037e-01 -9.34707046e-01
-4.10241246e-01 -6.18501365e-01 2.96500742e-01 5.74303389e-01
-2.33502015e-01 3.97074431e-01 -5.43944240e-01 3.73733938e-01
9.70424950e-01 1.31092682e-01 1.26175058e+00 -1.21172452e+00
-6.50343657e-01 2.38436922e-01 -5.56427717e-01 -1.47859633e+00
2.18888208e-01 -7.45077789e-01 5.27581930e-01 -1.28868222e+00
3.27137887e-01 -3.13350797e-01 1.40087396e-01 5.05671382e-01
1.89545229e-01 2.51627326e-01 4.12102848e-01 3.04101497e-01
-5.62437892e-01 7.69210875e-01 1.34945488e+00 -1.37170136e-01
-1.34818122e-01 -9.08543766e-02 -1.42236307e-01 8.99288833e-01
3.00689310e-01 -5.38326085e-01 -4.06516373e-01 -8.13873172e-01
-4.88734525e-03 3.09922308e-01 8.81256402e-01 -8.26371253e-01
1.06129043e-01 -2.89891064e-01 4.74062771e-01 -9.88774061e-01
1.04812312e+00 -1.41654265e+00 1.71734124e-01 1.06306612e-01
-1.93329126e-01 -1.91563398e-01 -4.41014804e-02 9.42638516e-01
1.69815019e-01 -2.37663109e-02 3.61076295e-01 -4.67948109e-01
-4.41773683e-01 5.10345519e-01 5.16521744e-02 1.35298604e-02
8.40564072e-01 -3.68692279e-01 9.91511866e-02 -3.49184871e-01
-9.62450445e-01 5.16802408e-02 6.87275827e-01 6.51249468e-01
7.53809869e-01 -1.19009900e+00 -3.19286346e-01 5.33560216e-01
3.04877251e-01 8.43385100e-01 2.01049037e-02 6.97159767e-01
-3.31889331e-01 4.01091367e-01 -1.62209317e-01 -1.29223919e+00
-1.27612591e+00 6.57107890e-01 3.37511092e-01 2.08350345e-01
-8.33507717e-01 7.71008134e-01 3.40119302e-01 -7.24177778e-01
7.66827524e-01 -5.07956922e-01 5.77609718e-01 -2.65884906e-01
4.09453362e-01 7.92993233e-02 1.41649857e-01 -6.15384996e-01
-3.17768812e-01 6.14834368e-01 9.00109485e-02 -4.25002843e-01
1.77368379e+00 4.26831329e-03 6.66671917e-02 1.02207446e+00
1.11491311e+00 -2.26926386e-01 -2.00982642e+00 -4.90714341e-01
-2.03421131e-01 -9.84442294e-01 3.74851003e-02 -6.33327246e-01
-9.94644105e-01 1.18715847e+00 5.98666608e-01 -4.57589120e-01
6.33139968e-01 4.30942208e-01 3.91370386e-01 9.64490831e-01
9.03461874e-01 -9.12648857e-01 4.76913810e-01 4.22157556e-01
1.34365106e+00 -1.41455626e+00 4.81975451e-02 -3.31135154e-01
-4.09128636e-01 9.50892687e-01 8.25196207e-01 8.42608288e-02
7.24287868e-01 3.27869624e-01 -1.53627247e-01 -2.86464155e-01
-6.21374846e-01 -3.90362889e-02 1.08965285e-01 6.46522760e-01
-6.23291768e-02 -5.57491593e-02 7.61298537e-01 4.19753194e-01
-7.28017241e-02 -1.37153380e-02 5.10694265e-01 1.37892950e+00
-1.54761478e-01 -8.49714100e-01 -6.77945554e-01 2.31385157e-01
2.94455010e-02 2.80662417e-01 -4.82625365e-01 6.48636699e-01
5.30502759e-02 5.43960631e-01 -1.54002029e-02 -3.88161540e-01
6.93913877e-01 2.03707412e-01 1.28488278e+00 -9.85010087e-01
-6.38300478e-02 3.04645598e-01 -3.79021853e-01 -6.63559198e-01
-5.73619545e-01 -8.90612423e-01 -9.67004061e-01 1.43964693e-01
-3.29555035e-01 -6.10181689e-01 8.21843207e-01 9.54869092e-01
2.41002277e-01 3.44928652e-01 3.30273420e-01 -1.82250464e+00
-7.20979929e-01 -9.30929124e-01 -6.74597383e-01 4.02507305e-01
8.47531497e-01 -1.09042263e+00 -4.53434765e-01 2.73654997e-01] | [6.604733943939209, -1.2184679508209229] |
98f24aae-c242-4c29-af69-9fef68f3dd67 | contrastive-regularization-for-semi | 2201.06247 | null | https://arxiv.org/abs/2201.06247v2 | https://arxiv.org/pdf/2201.06247v2.pdf | Contrastive Regularization for Semi-Supervised Learning | Consistency regularization on label predictions becomes a fundamental technique in semi-supervised learning, but it still requires a large number of training iterations for high performance. In this study, we analyze that the consistency regularization restricts the propagation of labeling information due to the exclusion of samples with unconfident pseudo-labels in the model updates. Then, we propose contrastive regularization to improve both efficiency and accuracy of the consistency regularization by well-clustered features of unlabeled data. In specific, after strongly augmented samples are assigned to clusters by their pseudo-labels, our contrastive regularization updates the model so that the features with confident pseudo-labels aggregate the features in the same cluster, while pushing away features in different clusters. As a result, the information of confident pseudo-labels can be effectively propagated into more unlabeled samples during training by the well-clustered features. On benchmarks of semi-supervised learning tasks, our contrastive regularization improves the previous consistency-based methods and achieves state-of-the-art results, especially with fewer training iterations. Our method also shows robust performance on open-set semi-supervised learning where unlabeled data includes out-of-distribution samples. | ['Wook-Shin Han', 'Minsu Cho', 'Yeongjae Cheon', 'Ildoo Kim', 'Sungwoong Kim', 'Doyup Lee'] | 2022-01-17 | null | null | null | null | ['semi-supervised-image-classification'] | ['computer-vision'] | [ 9.31872800e-02 3.12400639e-01 -6.86627448e-01 -9.04240191e-01
-8.94522369e-01 -4.71445739e-01 3.16340148e-01 1.67613968e-01
-1.82200938e-01 8.52683604e-01 -4.76364829e-02 9.03921947e-02
-1.15551520e-02 -4.11194116e-01 -7.93035567e-01 -8.78726184e-01
2.45767638e-01 8.31323028e-01 1.83640838e-01 2.85506278e-01
-1.96770415e-01 1.46054864e-01 -1.47720265e+00 3.44044715e-01
9.97742712e-01 1.02334583e+00 -1.27148420e-01 -1.01157099e-01
-1.37912542e-01 9.07827020e-01 -1.87722310e-01 -9.56707299e-02
3.31310391e-01 -5.06274641e-01 -7.79615343e-01 5.60826778e-01
5.27551293e-01 -7.02453628e-02 7.27061480e-02 1.34610450e+00
-1.19937912e-01 4.08304296e-02 7.94487536e-01 -1.22257638e+00
-7.08942533e-01 8.47795069e-01 -7.96941817e-01 -3.84215385e-01
-2.92371213e-01 -3.04532796e-01 1.07105339e+00 -1.16166139e+00
5.53461254e-01 1.01494431e+00 7.09201992e-01 5.61843455e-01
-1.40566659e+00 -9.50467825e-01 4.86570001e-01 4.04343521e-03
-1.43220627e+00 -2.73950934e-01 7.03864276e-01 -4.44496483e-01
3.21354121e-01 3.05298805e-01 3.08294833e-01 6.67034030e-01
-4.39806700e-01 9.39957738e-01 1.11964071e+00 -5.39606631e-01
3.56082618e-01 6.76208675e-01 8.39970112e-01 7.30050445e-01
3.51291239e-01 3.08058336e-02 -3.89151067e-01 -2.87882030e-01
4.05127406e-01 3.04672301e-01 -1.11502662e-01 -5.27260602e-01
-9.81314480e-01 1.00197589e+00 3.54361027e-01 1.90562770e-01
-1.13314316e-01 -1.51098087e-01 2.67937213e-01 1.59786895e-01
1.00855899e+00 6.58540353e-02 -8.79673183e-01 4.25998241e-01
-1.05345166e+00 -2.29071349e-01 5.98290920e-01 1.16161430e+00
1.25422251e+00 -3.43801588e-01 -3.73066992e-01 1.11747253e+00
5.91530979e-01 3.28640163e-01 3.90170157e-01 -1.03831279e+00
3.17235410e-01 8.80216479e-01 4.83344644e-02 -4.59298551e-01
-2.90614456e-01 -7.44436204e-01 -1.03572190e+00 7.73742795e-03
3.85235548e-01 -1.89717814e-01 -1.00000727e+00 1.81510401e+00
6.56477809e-01 3.65551591e-01 -6.71984330e-02 8.24708343e-01
4.35694337e-01 5.69603086e-01 -1.00393314e-02 -8.19590986e-01
1.04948092e+00 -1.32424843e+00 -9.10135210e-01 -2.21495822e-01
1.10412729e+00 -5.92779577e-01 8.21012616e-01 2.88639665e-01
-6.51104391e-01 -6.40416861e-01 -9.70579624e-01 1.47806361e-01
-1.09928496e-01 3.19661438e-01 6.37423396e-01 4.29471374e-01
-5.65944135e-01 7.02907026e-01 -9.46681321e-01 9.27226990e-02
7.00247824e-01 2.92115986e-01 -3.62390846e-01 -3.01743090e-01
-8.89648378e-01 4.60522145e-01 4.77497131e-01 -2.10927483e-02
-4.07544822e-01 -8.97759676e-01 -8.88526142e-01 4.98589426e-02
5.18432081e-01 -1.29849270e-01 1.13797712e+00 -1.06605768e+00
-1.18398070e+00 9.42773938e-01 -3.94738406e-01 -2.81538874e-01
5.48488259e-01 -1.69105887e-01 -3.91588271e-01 -1.53432176e-01
3.02688032e-01 6.64977968e-01 8.37497890e-01 -1.47778237e+00
-8.84548604e-01 -4.43194777e-01 -4.76902097e-01 1.23307839e-01
-5.90356767e-01 -2.54437685e-01 -6.27116263e-01 -5.10417938e-01
7.37744689e-01 -1.37699389e+00 -1.86592087e-01 3.33773680e-02
-5.07762671e-01 -3.63433480e-01 9.75702882e-01 -4.64758389e-02
1.13352728e+00 -2.33142400e+00 -2.00092643e-01 3.49660426e-01
4.39369857e-01 2.84809947e-01 9.25363526e-02 -1.84982464e-01
-1.19684681e-01 1.20325543e-01 -3.21511626e-01 -7.05009460e-01
1.44716520e-02 4.28608298e-01 -2.99386352e-01 6.58117771e-01
1.24190345e-01 6.06749117e-01 -1.02761924e+00 -7.62599051e-01
7.65450597e-02 2.26283729e-01 -4.09276098e-01 1.78448647e-01
-2.40220115e-01 6.29131258e-01 -4.44054395e-01 4.92171526e-01
1.01041377e+00 -1.00348866e+00 2.98941970e-01 -8.95071253e-02
2.75937647e-01 9.66999978e-02 -1.40672100e+00 1.33923602e+00
-1.34003475e-01 1.35608926e-01 5.87468222e-02 -9.11317229e-01
7.75400519e-01 1.16776161e-01 6.02437377e-01 -2.09570944e-01
-1.43740594e-01 3.85468096e-01 -3.29856813e-01 -5.02847768e-02
1.05222948e-01 -6.22939132e-02 1.98056281e-01 6.80352628e-01
1.44081250e-01 1.32474273e-01 3.38035554e-01 3.48038375e-01
5.58030367e-01 5.44177778e-02 4.25518975e-02 -3.49998146e-01
4.68242615e-01 -1.18846238e-01 1.01943624e+00 6.28301144e-01
-1.27670601e-01 5.21216214e-01 2.47686446e-01 -2.70816207e-01
-7.60517895e-01 -9.33892310e-01 -6.01446211e-01 1.29242122e+00
6.90215379e-02 -4.44387496e-01 -5.77062786e-01 -1.41660738e+00
1.61555067e-01 5.34211397e-01 -7.66938448e-01 -1.96389705e-01
-2.32627273e-01 -8.71631861e-01 6.43358082e-02 5.21990418e-01
2.00054973e-01 -6.42134249e-01 2.84722835e-01 1.48812726e-01
-6.48929477e-02 -9.38026845e-01 -5.71200013e-01 8.11277270e-01
-1.00855589e+00 -1.13658488e+00 -4.50268000e-01 -1.02062535e+00
1.44357383e+00 2.66680747e-01 9.17694211e-01 2.21408740e-01
1.96131647e-01 -2.17910767e-01 -5.59759617e-01 -2.30662495e-01
-6.06567442e-01 -3.96620072e-02 1.79247096e-01 1.80485293e-01
5.07761419e-01 -1.27122670e-01 -1.25159115e-01 6.77950263e-01
-7.70279050e-01 -1.80149823e-02 1.91390395e-01 1.40168786e+00
1.13580632e+00 1.60295200e-02 6.40479982e-01 -1.74038112e+00
-1.08845763e-01 -5.98925650e-01 -7.09870756e-01 4.08882737e-01
-1.15302682e+00 3.43553632e-01 8.10275257e-01 -5.59186459e-01
-1.06211782e+00 4.16802526e-01 2.25051120e-01 -5.66550672e-01
-1.71175878e-02 3.17043006e-01 -2.14650054e-02 3.44307497e-02
7.43902922e-01 -4.21126634e-02 2.00663581e-02 -5.50318301e-01
3.07286620e-01 6.92251682e-01 1.14977129e-01 -4.58759278e-01
7.99814284e-01 6.50568187e-01 3.77923548e-02 -2.99148619e-01
-1.66401303e+00 -6.94561183e-01 -8.57468188e-01 7.91514143e-02
2.47106165e-01 -1.10399806e+00 -2.24888235e-01 4.63958144e-01
-5.63453734e-01 -3.24169934e-01 -4.76577878e-01 6.83363497e-01
-1.69913858e-01 5.86552203e-01 -7.89438188e-01 -6.41288102e-01
-1.27176464e-01 -9.88081336e-01 9.18844402e-01 8.25697780e-02
-1.17335580e-01 -1.06309474e+00 1.39519554e-02 3.21747243e-01
9.66736581e-03 -2.61162788e-01 6.90019250e-01 -1.14828253e+00
-2.37820789e-01 -4.09818530e-01 -2.03533798e-01 7.86311448e-01
3.59731376e-01 -5.58860786e-02 -1.14845562e+00 -4.00263369e-01
5.21559417e-02 -8.35837245e-01 1.13264394e+00 4.24384475e-01
1.18554199e+00 -2.59510159e-01 -6.40713215e-01 5.89348614e-01
1.15363240e+00 -3.52919698e-01 -9.48910229e-03 -1.29721567e-01
7.98255384e-01 6.93832219e-01 1.13370204e+00 6.30793929e-01
9.20375139e-02 2.76222616e-01 2.57829756e-01 -1.38941020e-01
1.35997787e-01 -2.64896899e-01 2.97270361e-02 9.62686718e-01
3.31152618e-01 3.39155570e-02 -4.58658218e-01 2.95097232e-01
-2.18685246e+00 -7.21157789e-01 -5.70020199e-01 2.33069110e+00
1.25403380e+00 1.39024213e-01 -2.47742146e-01 1.40231401e-01
1.09266865e+00 -4.09704149e-02 -7.09094465e-01 2.80876607e-01
-1.04371503e-01 -9.82963741e-02 5.17370105e-01 5.66013217e-01
-1.43122125e+00 9.05970395e-01 6.04365158e+00 1.03478324e+00
-7.96032667e-01 2.81213135e-01 1.04077446e+00 -1.05189539e-01
-2.54030913e-01 -2.04108898e-02 -1.29862320e+00 6.14070773e-01
6.48577392e-01 2.24829733e-01 1.90390453e-01 1.17197263e+00
-1.87026635e-02 -9.66910571e-02 -1.46952903e+00 7.56731629e-01
-5.13995718e-03 -1.20981598e+00 -3.05507958e-01 2.41098627e-02
1.43469059e+00 6.05492145e-02 9.85268205e-02 4.97906178e-01
6.23880029e-01 -6.54149890e-01 6.49714231e-01 1.26848608e-01
7.84048915e-01 -5.68251550e-01 8.61258447e-01 6.38322890e-01
-9.67394233e-01 -1.88373104e-02 -6.10802948e-01 1.56156179e-02
3.08380555e-02 1.32525349e+00 -5.69224954e-01 3.31899375e-01
4.03110951e-01 1.01993823e+00 -5.32504499e-01 5.97136140e-01
-3.84962082e-01 1.02428532e+00 -4.85415787e-01 2.13125929e-01
1.48720279e-01 -3.89133662e-01 -1.57110706e-01 1.11924255e+00
-5.08185290e-02 2.98108663e-02 6.44209087e-01 8.61707628e-01
-2.84423947e-01 6.67271167e-02 -1.74797267e-01 3.01612616e-01
7.73252547e-01 1.36712539e+00 -6.94375813e-01 -7.41913676e-01
-3.98056626e-01 6.92665637e-01 7.72917390e-01 2.91131437e-01
-9.18611050e-01 1.66959301e-01 2.62076050e-01 -1.57644507e-02
1.70306101e-01 1.98351175e-01 -4.50655520e-01 -1.18300283e+00
2.94575077e-02 -4.09552932e-01 5.85366726e-01 -3.01685005e-01
-1.68848503e+00 3.07527095e-01 -2.26226926e-01 -1.42544281e+00
-2.39070132e-01 -3.52788717e-01 -1.54474124e-01 7.41341710e-01
-1.58515942e+00 -1.08172679e+00 -1.13382623e-01 4.35791045e-01
2.97521263e-01 3.30721959e-02 8.17369282e-01 2.93692887e-01
-7.05023825e-01 8.04220140e-01 6.45417631e-01 1.42372563e-01
1.21999705e+00 -1.30155766e+00 -2.12613598e-01 7.09342420e-01
2.73902029e-01 6.06742799e-01 2.81432927e-01 -7.84543216e-01
-6.62165999e-01 -1.51360309e+00 8.43259692e-01 -2.84764320e-01
6.00901842e-01 -2.61626810e-01 -1.22105014e+00 9.93796229e-01
-3.74474734e-01 8.82370889e-01 9.61368561e-01 3.07854623e-01
-6.30562484e-01 -1.46863744e-01 -1.16615796e+00 1.51690796e-01
8.70916903e-01 -3.37386101e-01 -3.45513701e-01 8.78232777e-01
7.51697659e-01 -2.91856647e-01 -7.82111704e-01 5.12531221e-01
1.40765518e-01 -7.61934519e-01 5.32180548e-01 -5.42278707e-01
2.48978883e-01 -4.24723268e-01 1.71416756e-02 -1.26133394e+00
-5.83862484e-01 -3.61522347e-01 -2.62667924e-01 1.45971966e+00
7.49299169e-01 -6.09964311e-01 1.07319736e+00 8.21456134e-01
-2.31413513e-01 -6.63530827e-01 -5.91362715e-01 -8.62860560e-01
8.05468038e-02 -3.04299295e-01 1.33313790e-01 1.42062879e+00
1.87574893e-01 2.55055755e-01 -4.60711926e-01 1.36324674e-01
8.81039083e-01 3.43093961e-01 4.47879374e-01 -1.58463323e+00
-3.16506475e-01 1.48320571e-02 5.38984612e-02 -1.14034009e+00
6.17138565e-01 -1.16376591e+00 2.88943499e-01 -1.04653871e+00
6.26419663e-01 -1.02119052e+00 -4.14821297e-01 9.11272526e-01
-5.69253266e-01 3.65515143e-01 -3.53969708e-02 6.03060901e-01
-1.09366369e+00 3.69137228e-01 1.25481284e+00 -1.34978876e-01
-2.48555347e-01 1.72205538e-01 -5.23953319e-01 8.88077259e-01
6.56185627e-01 -9.78200078e-01 -3.64355177e-01 -5.03383130e-02
2.87210960e-02 -3.66306514e-01 -5.20301536e-02 -7.34335721e-01
2.16578275e-01 -2.22824752e-01 5.67803800e-01 -6.24039888e-01
1.14524297e-01 -1.01036298e+00 1.38422260e-02 4.45458919e-01
-7.65793085e-01 -6.88206971e-01 -2.46808439e-01 7.72324145e-01
-2.39440650e-01 -4.50370342e-01 1.13134849e+00 5.78057137e-04
-3.15131962e-01 3.39139819e-01 -1.56136854e-02 2.90521383e-01
1.06757975e+00 -2.28402000e-02 -2.25681007e-01 -1.61193628e-02
-9.03750479e-01 5.17971516e-01 5.75548828e-01 -2.22405186e-03
1.80436626e-01 -1.38526464e+00 -6.27644598e-01 4.17058200e-01
3.30716431e-01 5.68546414e-01 2.16514841e-01 7.94323564e-01
-3.83069809e-03 2.50108510e-01 2.12253079e-01 -9.66753364e-01
-1.17143941e+00 7.67172575e-01 -4.60683890e-02 -4.73354936e-01
-3.47173154e-01 9.57291663e-01 3.38132530e-01 -7.55339980e-01
4.30439383e-01 -3.91727895e-01 -7.05931112e-02 -2.35998407e-02
5.30235946e-01 3.46473753e-01 5.18144257e-02 -4.35009062e-01
-2.18134731e-01 4.01642263e-01 -5.77526867e-01 2.11166650e-01
1.16618979e+00 -1.49573505e-01 -2.31957182e-01 6.78183019e-01
1.50135076e+00 1.56532869e-01 -1.69810331e+00 -6.43042147e-01
5.81589639e-02 -3.62485826e-01 -5.77482209e-02 -7.78620183e-01
-1.13396800e+00 5.28542936e-01 3.43203366e-01 8.37331340e-02
7.51066208e-01 2.77122527e-01 3.46500397e-01 4.45778102e-01
3.51388574e-01 -1.41969061e+00 2.76801009e-02 5.83245695e-01
2.34655574e-01 -1.73150241e+00 2.28312656e-01 -8.98999572e-01
-8.54286432e-01 8.70233893e-01 7.56699204e-01 -4.11515087e-02
9.86528337e-01 2.79548436e-01 1.75463453e-01 1.00415900e-01
-7.19567835e-01 9.54218581e-02 2.87011445e-01 3.88053358e-01
4.60670203e-01 1.76431358e-01 -2.63884693e-01 7.94243336e-01
3.08911443e-01 -1.64432883e-01 3.03847700e-01 8.77603173e-01
-5.19246221e-01 -1.06897426e+00 -3.66410017e-01 7.87936628e-01
-3.42584670e-01 -1.30563341e-02 -2.05516592e-01 5.47330618e-01
1.33669347e-01 9.53513086e-01 2.98869833e-02 -1.31335273e-01
-1.70905575e-01 2.81218767e-01 1.32983699e-01 -1.02853060e+00
-3.46441776e-01 2.93972433e-01 -1.31690904e-01 -2.91282922e-01
-5.75304866e-01 -6.01284444e-01 -1.69296336e+00 -2.50198506e-02
-9.07958865e-01 4.43646848e-01 4.45793480e-01 1.01746571e+00
3.52362394e-01 3.20234746e-01 1.00803113e+00 -4.90081847e-01
-1.07620502e+00 -1.11159229e+00 -1.01583838e+00 7.60089099e-01
1.45461604e-01 -5.99996388e-01 -7.32906580e-01 2.29749620e-01] | [9.441054344177246, 3.798900842666626] |
6182c04d-f21f-4292-a684-2d65392f3503 | revisiting-weak-to-strong-consistency-in-semi | 2208.09910 | null | https://arxiv.org/abs/2208.09910v2 | https://arxiv.org/pdf/2208.09910v2.pdf | Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation | In this work, we revisit the weak-to-strong consistency framework, popularized by FixMatch from semi-supervised classification, where the prediction of a weakly perturbed image serves as supervision for its strongly perturbed version. Intriguingly, we observe that such a simple pipeline already achieves competitive results against recent advanced works, when transferred to our segmentation scenario. Its success heavily relies on the manual design of strong data augmentations, however, which may be limited and inadequate to explore a broader perturbation space. Motivated by this, we propose an auxiliary feature perturbation stream as a supplement, leading to an expanded perturbation space. On the other, to sufficiently probe original image-level augmentations, we present a dual-stream perturbation technique, enabling two strong views to be simultaneously guided by a common weak view. Consequently, our overall Unified Dual-Stream Perturbations approach (UniMatch) surpasses all existing methods significantly across all evaluation protocols on the Pascal, Cityscapes, and COCO benchmarks. Its superiority is also demonstrated in remote sensing interpretation and medical image analysis. We hope our reproduced FixMatch and our results can inspire more future works. Code and logs are available at https://github.com/LiheYoung/UniMatch. | ['Yinghuan Shi', 'Wayne Zhang', 'Litong Feng', 'Lei Qi', 'Lihe Yang'] | 2022-08-21 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Yang_Revisiting_Weak-to-Strong_Consistency_in_Semi-Supervised_Semantic_Segmentation_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Yang_Revisiting_Weak-to-Strong_Consistency_in_Semi-Supervised_Semantic_Segmentation_CVPR_2023_paper.pdf | cvpr-2023-1 | ['semi-supervised-change-detection'] | ['computer-vision'] | [ 3.69756401e-01 2.15039514e-02 -1.94375619e-01 -3.25922638e-01
-9.85507131e-01 -6.08215213e-01 8.36856067e-01 5.34042977e-02
-1.83775172e-01 5.76122999e-01 2.81987727e-01 -2.27187574e-01
-1.05762973e-01 -3.87979150e-01 -7.52406836e-01 -8.94112229e-01
1.07973441e-01 2.99288988e-01 2.76795179e-01 -4.06085253e-01
-1.46753108e-02 2.20215172e-01 -1.43311107e+00 2.53054351e-01
9.16349411e-01 6.78443491e-01 1.35593228e-02 4.92490619e-01
3.76708150e-01 5.05584776e-01 -5.10108583e-02 -3.80537182e-01
5.09222686e-01 -1.98025703e-01 -9.73203897e-01 1.81606576e-01
6.41153574e-01 -2.01116756e-01 -1.64378688e-01 1.13628876e+00
6.75278962e-01 5.32437526e-02 3.31139147e-01 -1.29989707e+00
-8.96911979e-01 6.57056689e-01 -8.04633200e-01 2.39984408e-01
2.13869676e-01 4.71340209e-01 1.06851232e+00 -9.74820673e-01
6.31630778e-01 9.69855249e-01 8.18124115e-01 4.75328475e-01
-1.43654561e+00 -4.17625159e-01 6.43243670e-01 2.03643471e-01
-1.33960259e+00 -4.71184939e-01 7.43342578e-01 -4.06804681e-01
7.70994484e-01 5.22120535e-01 3.80245358e-01 1.28767633e+00
-3.86541098e-01 1.05692875e+00 1.24524307e+00 -2.08826035e-01
1.61403820e-01 -1.82073005e-02 1.40262678e-01 5.54434657e-01
2.03008810e-03 1.53012365e-01 -3.32604527e-01 -3.71107668e-01
9.11970586e-02 2.35791989e-02 -7.59972036e-01 -3.15864950e-01
-1.35844111e+00 5.18297255e-01 5.02547026e-01 1.66383654e-01
-3.69630665e-01 -1.36288553e-01 4.00376171e-01 2.94587821e-01
8.15552354e-01 4.12857652e-01 -5.62062144e-01 -1.20445400e-01
-9.25135612e-01 4.69913840e-01 5.70805132e-01 8.00518513e-01
8.33113253e-01 -2.61454701e-01 -3.70428503e-01 7.11498022e-01
1.81862503e-01 5.20289540e-01 2.69705415e-01 -1.01194620e+00
2.99888730e-01 4.51316386e-01 2.40729704e-01 -9.43775654e-01
-4.62328643e-01 -7.27674663e-01 -1.02505052e+00 -8.09214711e-02
2.84537047e-01 2.66429354e-02 -9.04884040e-01 1.77934635e+00
7.08117604e-01 5.87897480e-01 -7.13766813e-02 1.03287923e+00
7.13063836e-01 3.76153439e-01 -2.15040937e-01 8.70942511e-03
1.24203336e+00 -1.30686092e+00 -3.93554240e-01 -1.67610735e-01
4.57715183e-01 -5.36263525e-01 1.53557324e+00 5.05659759e-01
-8.00419748e-01 -3.77020180e-01 -8.36523354e-01 1.23214334e-01
-1.58203527e-01 -7.89689198e-02 4.07556355e-01 2.57288992e-01
-1.04441893e+00 6.33169413e-01 -1.21281016e+00 -5.34645319e-01
6.08731627e-01 -1.11373700e-01 -3.54743749e-01 -2.15631351e-01
-1.02042818e+00 5.99997282e-01 9.49773565e-02 2.88703978e-01
-6.92784488e-01 -1.05782604e+00 -8.35407436e-01 -3.15136403e-01
4.60530877e-01 -7.24816144e-01 1.27981365e+00 -6.77440763e-01
-1.28268349e+00 9.00462151e-01 -2.27317307e-02 -5.44697285e-01
9.23027873e-01 -4.11466211e-01 -1.77228138e-01 1.49645448e-01
2.32988194e-01 6.49060130e-01 5.88704824e-01 -1.43752670e+00
-4.40485537e-01 -4.05306607e-01 7.02141039e-03 2.85575181e-01
-3.11006755e-01 -2.17174292e-01 -8.28542233e-01 -7.40758121e-01
1.96384057e-01 -1.05637050e+00 -5.57404995e-01 1.22702152e-01
-5.67504764e-01 1.32390112e-01 6.97479963e-01 -3.52287501e-01
1.05546999e+00 -2.29018497e+00 2.95360863e-01 1.06111355e-01
2.35465676e-01 3.13563913e-01 -3.11270863e-01 3.39514583e-01
-2.28947565e-01 2.57452130e-01 -8.15044820e-01 -7.23625898e-01
7.49750882e-02 4.01936918e-01 -4.22077894e-01 7.89443433e-01
3.55623960e-01 9.73217130e-01 -1.20849502e+00 -2.36263782e-01
2.48339236e-01 3.21326256e-01 -6.44524515e-01 -1.50454966e-02
-2.18037695e-01 8.45388293e-01 -2.58143842e-01 8.49187911e-01
1.00859654e+00 -5.92794657e-01 -2.06236392e-01 -3.17469209e-01
6.78936690e-02 5.35517447e-02 -1.17041576e+00 1.85023105e+00
-1.76993813e-02 1.40320092e-01 1.53282747e-01 -1.12837088e+00
6.27319753e-01 1.35344565e-01 6.08343601e-01 -5.64591706e-01
-3.52333009e-01 1.83414996e-01 -2.40154728e-01 -5.41287541e-01
4.08458173e-01 -9.70139205e-02 8.94704312e-02 3.69515955e-01
-2.08255231e-01 -1.81430262e-02 1.19011454e-01 3.03124696e-01
1.19933736e+00 4.93611097e-01 3.42297226e-01 -2.40429148e-01
4.80578333e-01 1.67035997e-01 7.01100588e-01 9.77850795e-01
-6.63007379e-01 1.11748815e+00 4.49687362e-01 -3.79096419e-01
-7.75085211e-01 -1.04564106e+00 -4.99277204e-01 9.42028761e-01
2.89768577e-01 -4.80042398e-01 -4.97784525e-01 -9.08748090e-01
8.90891105e-02 3.16338092e-01 -8.76592875e-01 1.13131151e-01
-3.60678315e-01 -1.17953861e+00 7.89750159e-01 4.51814920e-01
4.69361186e-01 -8.19739461e-01 -2.76552051e-01 -1.99242365e-02
-3.19335401e-01 -1.19811797e+00 -3.22891444e-01 7.61546269e-02
-8.15490603e-01 -1.00790548e+00 -7.69059539e-01 -2.99853325e-01
4.80453253e-01 3.06128204e-01 1.07858396e+00 1.65729776e-01
3.64951268e-02 4.22416955e-01 -6.41837418e-01 -1.98379442e-01
-2.55135328e-01 2.19901919e-01 1.07081726e-01 -3.90530229e-02
1.85095310e-01 -8.11818719e-01 -8.95336449e-01 4.02699798e-01
-9.85579729e-01 1.45385608e-01 4.07693207e-01 1.02395213e+00
8.18302393e-01 -4.07074571e-01 3.44968647e-01 -1.21264911e+00
2.18625933e-01 -7.39712656e-01 -2.42035672e-01 9.22630355e-02
-7.30652094e-01 -9.80070233e-02 6.52758777e-01 -1.82148904e-01
-9.32889938e-01 1.60808079e-02 -3.15544546e-01 -5.02734423e-01
-3.14020097e-01 6.02369368e-01 2.71489210e-02 1.84152141e-01
1.01936603e+00 3.00728709e-01 3.78704742e-02 -6.98642969e-01
4.65187639e-01 6.59760714e-01 7.15244472e-01 -6.60200715e-01
9.58968818e-01 9.39498842e-01 -3.73356134e-01 -6.07322931e-01
-9.94589031e-01 -7.36276388e-01 -5.14134586e-01 7.39661604e-02
3.83712113e-01 -1.11269534e+00 -3.05259079e-01 5.60444415e-01
-8.41827691e-01 -4.86264378e-01 -3.99395645e-01 2.74699450e-01
-3.44788015e-01 6.55435920e-01 -5.74700296e-01 -4.29389328e-01
-3.06835055e-01 -1.14367008e+00 1.48155510e+00 -8.81625488e-02
-7.88521320e-02 -8.57307255e-01 3.21359277e-01 4.63074267e-01
4.12328094e-01 5.00224769e-01 3.61735284e-01 -6.03410244e-01
-2.75192142e-01 -9.79960430e-03 -1.52140677e-01 4.02769297e-01
2.05717057e-01 1.95798025e-01 -1.17285609e+00 -4.69543874e-01
-7.44667649e-02 -5.34987271e-01 1.14352977e+00 1.56946599e-01
1.23652112e+00 -7.61892647e-02 -2.09567115e-01 1.03277230e+00
1.24249649e+00 -6.78608179e-01 4.76832420e-01 5.54554164e-01
8.28162611e-01 4.78368342e-01 6.65758550e-01 5.40326118e-01
6.08402193e-01 5.86604297e-01 6.25365078e-01 -5.01592398e-01
-4.94640991e-02 3.52106690e-02 2.38093257e-01 5.85936368e-01
-2.01918736e-01 -1.07977115e-01 -1.05946469e+00 5.90922952e-01
-2.14329410e+00 -8.82820070e-01 -1.87276557e-01 1.91279590e+00
8.67204726e-01 -7.41246045e-02 6.36642240e-03 6.65069669e-02
4.78067577e-01 4.94840175e-01 -7.31041491e-01 1.98816806e-01
-2.80935884e-01 2.11489713e-03 3.13988715e-01 6.39184177e-01
-1.31716633e+00 7.80654132e-01 5.81823874e+00 5.94933331e-01
-1.22141683e+00 2.57734805e-01 7.17197120e-01 -2.64458239e-01
-5.15440524e-01 -9.07033086e-02 -5.21781325e-01 4.33650613e-01
6.11319900e-01 4.56732512e-02 3.52169961e-01 7.48265445e-01
3.02253425e-01 4.38980274e-02 -1.16163921e+00 7.74700165e-01
-6.66457489e-02 -1.26421416e+00 2.37274375e-02 -1.24833599e-01
8.57013881e-01 7.93989956e-01 2.30380267e-01 2.95524478e-01
2.68686265e-01 -8.01642120e-01 5.69379270e-01 2.46964648e-01
5.58916986e-01 -3.13910425e-01 7.24016666e-01 5.23854136e-01
-1.10805631e+00 2.33005315e-01 -1.63283035e-01 -1.55554404e-02
5.84075265e-02 7.31107891e-01 -4.61465716e-01 1.04040420e+00
9.86748159e-01 1.09770799e+00 -7.79283941e-01 7.85588861e-01
-2.18450069e-01 8.25665951e-01 -5.35689116e-01 5.65984130e-01
3.90951514e-01 -9.53213498e-02 7.41975427e-01 1.34883714e+00
6.04293495e-02 6.05614334e-02 4.64853406e-01 8.00296128e-01
-7.61782527e-02 3.81840654e-02 -5.27416825e-01 3.74716967e-01
3.82959008e-01 1.40161157e+00 -5.45022607e-01 -2.91045338e-01
-5.03758013e-01 1.07228458e+00 4.61018473e-01 4.27106023e-01
-9.35958564e-01 2.42998302e-01 7.64360964e-01 -7.96706155e-02
2.53616452e-01 1.32406279e-01 -3.38874251e-01 -1.57159185e+00
4.37477082e-01 -1.13465393e+00 5.28586507e-01 -4.95089620e-01
-1.55372179e+00 6.11650527e-01 6.15757965e-02 -1.46462202e+00
-1.36531389e-03 -3.49444360e-01 -5.86197615e-01 8.21522236e-01
-1.91563106e+00 -1.25770795e+00 -5.50554335e-01 5.45655310e-01
3.62595618e-01 2.44296983e-01 7.90898144e-01 2.90213436e-01
-8.85761976e-01 5.84886849e-01 9.02855545e-02 -7.26864161e-03
9.63778615e-01 -1.42679107e+00 5.63607216e-01 1.25324786e+00
2.81759530e-01 4.54663366e-01 8.04567277e-01 -4.59926814e-01
-1.27712810e+00 -1.42235446e+00 3.60938221e-01 -8.02465558e-01
7.01621711e-01 -1.99761495e-01 -1.22918510e+00 7.63907671e-01
1.85397238e-01 6.83024347e-01 7.42258430e-01 1.80589169e-01
-3.72422516e-01 -8.89423415e-02 -9.72463369e-01 7.38325715e-01
1.15365553e+00 -2.75078297e-01 -4.41989809e-01 4.37029570e-01
8.20747972e-01 -6.28456950e-01 -8.84816229e-01 7.20113873e-01
3.32141787e-01 -1.22400689e+00 7.64463603e-01 -5.44596791e-01
5.86235821e-01 -5.64878702e-01 -4.12984908e-01 -1.25920677e+00
-3.55167091e-01 -6.83228850e-01 -1.69388875e-01 1.10294557e+00
4.63170469e-01 -7.09977329e-01 8.23464692e-01 3.07399541e-01
-2.94925183e-01 -1.14100051e+00 -7.36874461e-01 -7.93249249e-01
7.79163986e-02 -7.01467037e-01 6.84637129e-01 1.25538719e+00
-1.39839903e-01 7.23475143e-02 -4.22229797e-01 5.93133867e-01
7.16762722e-01 2.80058056e-01 9.15271342e-01 -8.48811388e-01
-7.03552365e-01 -4.13230807e-01 -1.46796927e-01 -1.13612771e+00
5.60599677e-02 -1.05078626e+00 1.63684160e-01 -1.18954957e+00
3.21837604e-01 -5.04208624e-01 -5.32441914e-01 6.69515729e-01
-7.16229558e-01 7.61653483e-01 1.50529072e-01 5.61464787e-01
-5.44481575e-01 7.62180030e-01 1.19521630e+00 -1.75339252e-01
-2.07904115e-01 -6.44613579e-02 -1.13510382e+00 8.14605534e-01
7.84874976e-01 -4.18934017e-01 -2.59128124e-01 -5.63589573e-01
1.70545191e-01 -3.68320435e-01 4.75780874e-01 -7.30907738e-01
6.89271465e-02 -9.52236876e-02 2.97087114e-02 -3.43614012e-01
4.04684013e-03 -5.34388125e-01 -2.62031034e-02 4.41318214e-01
-2.21646205e-01 -3.92495357e-02 2.91338980e-01 7.97378957e-01
-3.16698492e-01 2.09414110e-01 8.31693649e-01 -6.92380667e-02
-6.33153975e-01 5.70372224e-01 1.66276768e-01 2.11013407e-01
9.40729201e-01 -4.89633232e-02 -4.19869214e-01 -1.56084418e-01
-8.49261343e-01 6.00062788e-01 7.31846511e-01 3.13579321e-01
2.63628870e-01 -1.20998096e+00 -1.11425805e+00 1.81122482e-01
5.22908390e-01 4.51328635e-01 3.17167789e-01 1.42868078e+00
-3.73506606e-01 -6.90623820e-02 2.18214452e-01 -1.14814878e+00
-1.09514725e+00 4.23714936e-01 3.57674211e-01 -3.97278190e-01
-8.93732309e-01 8.17364931e-01 1.42170444e-01 -8.16665769e-01
1.24361590e-01 -5.00123680e-01 7.36705959e-02 -2.30331257e-01
6.28858864e-01 3.15619439e-01 2.94255674e-01 -5.13010979e-01
-6.53863609e-01 4.44134116e-01 -1.86658457e-01 -5.70549108e-02
1.58083749e+00 -2.44412988e-01 -9.31083038e-02 3.91279459e-01
9.64599252e-01 1.47124276e-01 -1.51413310e+00 -6.49414778e-01
-1.47836745e-01 -3.99703503e-01 -1.43828049e-01 -8.02159846e-01
-1.02851999e+00 5.72130263e-01 4.85393614e-01 6.02146760e-02
1.39286613e+00 2.01539978e-01 5.68005681e-01 4.98371959e-01
6.89324513e-02 -7.45744765e-01 -2.03316107e-01 3.93886626e-01
9.33546007e-01 -1.63967597e+00 1.05135806e-01 -3.87318134e-01
-7.07790911e-01 6.52557135e-01 6.07895076e-01 -5.70608974e-02
6.75980866e-01 4.43423957e-01 2.49253646e-01 -2.57267296e-01
-9.14692104e-01 -2.78589308e-01 2.60281771e-01 5.16150296e-01
4.28002477e-01 -4.03954089e-02 -1.26695916e-01 4.46961850e-01
-8.90972391e-02 -5.63533381e-02 3.04334432e-01 8.73600304e-01
-1.13442190e-01 -1.11114657e+00 -4.69898611e-01 3.96943033e-01
-4.13687795e-01 -3.07086438e-01 -2.82426149e-01 7.26702034e-01
1.71067536e-01 7.03603864e-01 -4.68111448e-02 -3.37342948e-01
5.30363798e-01 -2.05277279e-01 3.20879102e-01 -6.94817126e-01
-6.27161801e-01 1.52879348e-02 -1.13982148e-01 -9.15682137e-01
-7.03318000e-01 -8.95808041e-01 -1.00346947e+00 -7.87175894e-02
-8.86993036e-02 4.49663810e-02 1.48071587e-01 9.44842577e-01
5.75951874e-01 3.05320889e-01 6.31079316e-01 -9.44474220e-01
-8.04487288e-01 -1.10991526e+00 -3.04288298e-01 7.18731165e-01
5.91352105e-01 -4.31718737e-01 -4.46944475e-01 9.93902907e-02] | [9.311583518981934, 1.3514041900634766] |
b5aa88e0-f0dc-4511-b969-40d4833821de | why-should-i-turn-left-towards-active | null | null | https://aclanthology.org/2021.reinact-1.9 | https://aclanthology.org/2021.reinact-1.9.pdf | Why Should I Turn Left? Towards Active Explainability for Spoken Dialogue Systems. | In this paper we argue that to make dialogue systems able to actively explain their decisions they can make use of enthymematic reasoning. We motivate why this is an appropriate strategy and integrate it within our own proof-theoretic dialogue manager framework based on linear logic. In particular, this enables a dialogue system to provide reasonable answers to why-questions that query information previously given by the system. | ['Jean-Philippe Bernardy', 'Christine Howes', 'Ellen Breitholtz', 'Vladislav Maraev'] | null | null | null | null | reinact-2021-10 | ['spoken-dialogue-systems'] | ['speech'] | [-1.26115263e-01 1.37423944e+00 -1.41261965e-01 -4.43193942e-01
-6.11362159e-01 -8.38397264e-01 1.03308105e+00 2.42081583e-01
-2.51169026e-01 1.03808069e+00 3.34734976e-01 -1.22093832e+00
-2.46166617e-01 -9.85109568e-01 -1.54018521e-01 2.78220385e-01
3.59528184e-01 5.57672262e-01 5.99678516e-01 -1.07158995e+00
5.25236309e-01 2.72368878e-01 -1.02989054e+00 4.17863518e-01
7.61798441e-01 3.01064998e-01 2.72649731e-02 1.06183076e+00
-7.28826344e-01 1.74456811e+00 -3.85938197e-01 -6.19245231e-01
-1.25478014e-01 -5.75997293e-01 -1.65609181e+00 -1.21868342e-01
-1.87812075e-01 -5.91742635e-01 -2.52668649e-01 8.41345191e-01
-1.75732955e-01 -2.35966653e-01 5.27156234e-01 -1.43655622e+00
-7.17540681e-02 1.02505660e+00 5.53021550e-01 -4.58662659e-02
9.96180475e-01 3.14409584e-01 1.05123007e+00 -3.28957081e-01
7.64152527e-01 1.54605567e+00 5.22633493e-01 9.71554220e-01
-1.18364084e+00 1.60524085e-01 6.71286881e-02 2.32311487e-01
-8.08955014e-01 -6.12326801e-01 6.09342873e-01 -2.96749979e-01
1.01187074e+00 7.14623868e-01 6.13513410e-01 4.95658547e-01
4.31941420e-01 7.05866218e-01 1.25181270e+00 -8.34583700e-01
3.09082747e-01 8.73574376e-01 5.25563657e-01 7.87096202e-01
9.55629498e-02 -2.70380497e-01 -4.13892865e-01 -2.84400165e-01
8.17491531e-01 -5.81287324e-01 -1.23542391e-01 -3.76674831e-01
-8.05207670e-01 1.24182463e+00 -1.27232999e-01 6.21312797e-01
-2.72081077e-01 2.49887109e-02 4.94197696e-01 7.93535173e-01
-9.31642056e-02 9.73584414e-01 -5.16575873e-01 -3.44707608e-01
-3.23267400e-01 5.82908511e-01 1.87912369e+00 7.18345642e-01
4.19467866e-01 -2.32406661e-01 1.53897807e-01 2.86503792e-01
8.19838703e-01 1.08602837e-01 -5.80558404e-02 -1.90104961e+00
1.52602494e-02 8.02269101e-01 5.57441413e-01 -5.95780551e-01
-2.00892583e-01 2.34877422e-01 -1.29134282e-01 1.86246529e-01
6.09720170e-01 -1.28153712e-01 1.89637050e-01 1.43386400e+00
3.18859220e-01 -5.99462152e-01 8.64215970e-01 5.13118327e-01
6.86301589e-01 4.71556544e-01 -9.60513055e-02 -5.51623940e-01
1.30582201e+00 -5.26308358e-01 -1.13656950e+00 6.86311945e-02
1.01116920e+00 -4.70260710e-01 1.02555263e+00 5.10342002e-01
-1.20991862e+00 2.74264608e-02 -9.23055232e-01 -2.68418908e-01
-3.66689861e-01 -2.82615542e-01 9.38309789e-01 5.36970019e-01
-1.20868659e+00 3.41148317e-01 -4.95044947e-01 -6.15420341e-01
-4.48434889e-01 1.66961491e-01 -1.58253945e-02 2.15999320e-01
-1.71807921e+00 1.31574261e+00 5.17695725e-01 1.79882362e-01
-4.52941597e-01 -2.07216576e-01 -7.97214746e-01 -1.83518380e-01
9.48718190e-01 -8.00273657e-01 2.09144807e+00 -2.86069095e-01
-2.11995316e+00 8.57734203e-01 2.45408174e-02 -8.63472462e-01
7.66531467e-01 -1.34650812e-01 2.91346665e-02 2.90114909e-01
1.34984702e-01 4.45133686e-01 8.92267227e-02 -9.90953565e-01
-5.59635043e-01 -1.78863794e-01 1.15575838e+00 4.83477972e-02
4.28133816e-01 2.63892889e-01 6.31984398e-02 2.98303694e-01
1.74669102e-01 -7.73375332e-01 -2.17382744e-01 8.85741878e-03
-5.11605859e-01 -5.86964369e-01 4.80838060e-01 -3.78930777e-01
1.29428387e+00 -1.43986106e+00 -8.67588520e-02 -3.66944224e-02
2.88041264e-01 1.81701958e-01 4.74603236e-01 1.11265290e+00
3.19748670e-01 4.19504285e-01 3.31091076e-01 2.90327996e-01
8.06668818e-01 5.77420890e-01 -6.15534663e-01 -8.34198594e-02
9.25515965e-02 9.03575838e-01 -6.15437746e-01 -7.44931400e-01
5.52300513e-01 -2.05535576e-01 -5.22125840e-01 6.18660867e-01
-1.01799321e+00 1.91444159e-01 -8.00743759e-01 3.35013866e-01
3.08016300e-01 -2.16812804e-01 6.37587547e-01 4.47151482e-01
-4.54629630e-01 7.87579298e-01 -1.07225776e+00 1.38564634e+00
-4.67364311e-01 4.63640332e-01 5.26138902e-01 -7.36657560e-01
7.25951493e-01 5.36535740e-01 -1.24492534e-01 -3.10163438e-01
1.19143546e-01 2.58077592e-01 4.25006486e-02 -8.52402866e-01
3.53342175e-01 -4.97101545e-01 -1.15290754e-01 8.13862383e-01
-2.61408478e-01 -8.08513284e-01 8.98626894e-02 8.31159353e-01
9.60143089e-01 8.40822980e-02 6.61770284e-01 -3.55573982e-01
1.39646173e+00 5.31002283e-01 7.41654187e-02 1.06375146e+00
-1.09547094e-01 -4.59869534e-01 1.05725586e+00 -5.80842853e-01
-1.08348823e+00 -5.86563408e-01 -9.77716595e-02 8.92030478e-01
-9.64908451e-02 -8.90917599e-01 -9.78672445e-01 -6.48004293e-01
-3.07132632e-01 1.25653899e+00 -2.24393100e-01 3.99636686e-01
-3.92764032e-01 2.25561380e-01 7.89777160e-01 6.53985217e-02
6.25330865e-01 -1.05713677e+00 -9.50890362e-01 2.99742550e-01
-3.15598547e-01 -9.18159783e-01 4.30634648e-01 2.19428286e-01
-7.12979019e-01 -1.34381330e+00 -1.73695087e-02 -1.70716003e-01
1.29455879e-01 -1.43996142e-02 9.20513511e-01 6.17035806e-01
3.41192424e-01 8.97229910e-01 -3.78358513e-01 -3.10983330e-01
-1.31140733e+00 7.19627514e-02 -4.63925034e-01 -8.18938076e-01
3.66872370e-01 -2.32021958e-01 -9.88537632e-03 1.55860350e-01
-7.07835078e-01 2.92454869e-01 -3.24715525e-02 6.21216178e-01
-2.72938490e-01 -3.11690979e-02 3.38491350e-01 -1.27051306e+00
1.47972131e+00 -2.07259715e-01 -7.08411038e-01 7.73854196e-01
-9.38515782e-01 5.09294391e-01 6.19711161e-01 2.27037713e-01
-1.10356534e+00 -3.08929533e-01 -2.17769384e-01 4.71560597e-01
-3.23836207e-01 6.36551917e-01 -1.36963665e-01 -8.15224349e-02
5.84425330e-01 1.48513868e-01 4.87946033e-01 -5.59156351e-02
7.51839638e-01 6.31404519e-01 2.91762888e-01 -9.78995204e-01
5.53713441e-01 -8.56339261e-02 7.53016993e-02 -5.71410656e-01
-4.92159486e-01 -1.69225097e-01 -4.98718709e-01 -3.90333384e-01
6.25954330e-01 -3.74033540e-01 -1.47601712e+00 -1.37028530e-01
-1.52107191e+00 -5.27734995e-01 -2.43987292e-01 7.42080212e-02
-1.07114947e+00 6.36420846e-01 -7.11558223e-01 -1.63517106e+00
-1.28480092e-01 -9.29083884e-01 5.42392015e-01 1.97161973e-01
-7.20591605e-01 -1.24484265e+00 8.77640918e-02 3.38540912e-01
5.14351606e-01 -6.34084120e-02 1.12819803e+00 -1.11293519e+00
-5.61987519e-01 -3.00054066e-02 -7.01639652e-02 1.77871417e-02
-2.86179818e-02 1.44967243e-01 -6.64325416e-01 3.69245887e-01
4.23397541e-01 -5.01323044e-01 -1.08735688e-01 -2.33287662e-01
3.53468746e-01 -8.65180075e-01 1.99298095e-02 -3.99391353e-01
1.22317910e+00 -3.75566557e-02 5.56853831e-01 5.43427587e-01
-1.70915917e-01 1.20318902e+00 7.09672570e-01 1.88796461e-01
9.02745664e-01 8.68524373e-01 5.22124916e-02 4.11437303e-01
3.57372195e-01 -3.35468769e-01 2.18723774e-01 2.82366991e-01
2.20209673e-01 1.19912036e-01 -1.01729596e+00 1.18048027e-01
-2.35459924e+00 -1.01398730e+00 -6.47818565e-01 1.50524580e+00
1.25268281e+00 3.35358590e-01 2.18097255e-01 7.31148422e-02
3.02260846e-01 1.45213027e-02 -2.40888327e-01 -1.17215192e+00
2.29075909e-01 -8.02968070e-02 1.34952456e-01 1.30808425e+00
-5.02605796e-01 1.01207745e+00 7.26652718e+00 2.00543180e-01
-5.72548389e-01 -7.46952370e-02 4.37409729e-02 6.80342555e-01
-6.49320662e-01 6.01584792e-01 -4.45390433e-01 -3.86719108e-02
1.24923360e+00 -5.37450254e-01 5.12150466e-01 7.69329250e-01
4.49590176e-01 -8.55361283e-01 -1.45817506e+00 2.34516218e-01
-2.79540449e-01 -1.49377549e+00 -1.27223849e-01 2.66471747e-02
-1.91067278e-01 -8.16922426e-01 -5.91525972e-01 3.54679257e-01
7.45284140e-01 -1.00360799e+00 3.47063005e-01 8.95925879e-01
-1.26552284e-02 -4.77627635e-01 9.21032608e-01 1.17374313e+00
-4.94080275e-01 -2.43336558e-02 -1.45580143e-01 -5.92566550e-01
2.26913542e-01 7.16035068e-03 -1.42811298e+00 6.35865808e-01
-2.47720882e-01 -1.07242569e-01 -3.83293867e-01 5.54211557e-01
-7.19291568e-01 4.58958626e-01 -3.26768756e-01 -5.44734299e-01
2.50148177e-01 -3.26023400e-01 3.89782548e-01 9.61015940e-01
-2.57445127e-01 7.69864738e-01 1.27032101e-01 9.65001762e-01
7.12323546e-01 6.32138997e-02 -8.58783185e-01 7.65391067e-02
3.83266777e-01 9.12059724e-01 -3.79227221e-01 -5.65740228e-01
-2.03259766e-01 5.28951406e-01 7.75879249e-03 -1.46437153e-01
-3.98480296e-01 -1.06727645e-01 1.18931476e-02 4.59368639e-02
-3.22297007e-01 -2.75405824e-01 -1.89723238e-01 -1.17491245e+00
-1.11772813e-01 -1.19808781e+00 2.91082412e-01 -1.20604372e+00
-9.13736641e-01 3.69609356e-01 4.44114208e-01 -4.19002473e-01
-9.85521734e-01 -5.37635446e-01 -6.45245433e-01 8.88565183e-01
-1.30800331e+00 -1.06960726e+00 2.42892042e-01 2.96737283e-01
4.58502889e-01 2.16266468e-01 1.26862013e+00 -4.33869183e-01
-1.16690263e-01 -9.80959758e-02 -8.15542758e-01 -1.80358469e-01
4.23898876e-01 -1.59443665e+00 1.37606710e-01 5.43118715e-01
-5.49741685e-02 1.17139888e+00 1.42165816e+00 -4.90123898e-01
-1.77053213e+00 -2.25009248e-01 1.37879264e+00 -6.33268893e-01
1.00923407e+00 -1.12879068e-01 -1.14177430e+00 5.43147624e-01
7.51089096e-01 -7.71157980e-01 4.64051068e-01 1.25953788e-02
-1.54303148e-01 3.19698751e-01 -1.08837056e+00 5.79171658e-01
5.07891476e-01 -9.49155033e-01 -1.57007527e+00 6.06292129e-01
7.40752876e-01 -3.96393001e-01 -6.49606168e-01 4.30296324e-02
4.09330398e-01 -1.08220840e+00 5.26051342e-01 -8.89080584e-01
1.94722340e-01 -3.87119591e-01 -1.97234765e-01 -6.42845094e-01
4.58502233e-01 -1.22793603e+00 1.42966509e-02 1.07998526e+00
5.26585639e-01 -9.32197273e-01 5.79880357e-01 1.75280726e+00
3.35440516e-01 -4.65643197e-01 -8.08252931e-01 -3.38027507e-01
4.13956404e-01 -7.34703839e-01 3.64056110e-01 6.84666693e-01
1.21410835e+00 5.79772174e-01 -2.00310290e-01 -1.07873835e-01
1.71009675e-01 -5.02600595e-02 1.09913516e+00 -1.25292563e+00
-2.52072603e-01 -3.61533284e-01 1.82371482e-01 -1.20145369e+00
4.29719567e-01 -6.38901889e-01 1.60473421e-01 -1.72880459e+00
-3.68361473e-01 -3.27306926e-01 4.99835044e-01 5.03570795e-01
3.64673883e-01 -6.06337130e-01 3.47804785e-01 1.60800472e-01
-8.84230614e-01 1.73205450e-01 9.53819156e-01 2.71322429e-01
-1.16866849e-01 1.96811765e-01 -9.63647842e-01 8.58527780e-01
8.27990592e-01 -2.71841735e-01 -4.43754405e-01 2.41956055e-01
8.32367718e-01 9.62513387e-01 5.23254216e-01 -3.85079116e-01
7.83538163e-01 -6.32063150e-01 -6.57520533e-01 -2.88922071e-01
5.52453566e-04 -7.64156222e-01 1.09399445e-01 7.23734617e-01
-1.04206979e+00 -1.23332724e-01 1.31000280e-01 1.22702993e-01
-2.13853806e-01 -8.06530893e-01 1.80633351e-01 -4.56466764e-01
-6.50842309e-01 -7.46690214e-01 -1.12736630e+00 -9.85968038e-02
8.94400954e-01 -2.34847739e-02 -4.77507293e-01 -9.35495317e-01
-7.44260728e-01 6.18224978e-01 6.68116331e-01 -1.81816265e-01
5.33745706e-01 -5.91595948e-01 -3.21134239e-01 -3.37238133e-01
-3.09718046e-02 -5.09374917e-01 -2.27751702e-01 7.16631472e-01
-7.93379009e-01 1.21175313e+00 -4.40839306e-02 -1.38514236e-01
-1.19624496e+00 4.20120329e-01 6.98538780e-01 -2.12952778e-01
-5.83255827e-01 3.02823782e-01 -2.06983328e-01 -6.80255115e-01
1.29273519e-01 -2.75106549e-01 -4.60025728e-01 -3.39652956e-01
5.36209643e-01 -3.13772000e-02 -3.45474422e-01 -4.62543964e-03
-4.18577254e-01 1.33178040e-01 1.16192237e-01 -8.63853335e-01
1.07543910e+00 -5.38851559e-01 -5.47248065e-01 7.89451599e-01
2.74399161e-01 1.64754450e-01 -6.01710498e-01 -2.79232860e-01
3.39418083e-01 -2.38282993e-01 -2.65140593e-01 -1.15095246e+00
2.68675953e-01 7.56935716e-01 -2.09764019e-01 1.41236663e+00
3.80484819e-01 1.38019219e-01 3.02332044e-01 1.05421090e+00
5.67902505e-01 -1.02408683e+00 -2.73285747e-01 5.97971082e-01
1.12481093e+00 -1.03371263e+00 2.41494969e-01 -7.15893805e-01
-7.90638328e-01 1.88287890e+00 5.15088081e-01 2.21886545e-01
1.36147946e-01 2.70755976e-01 2.64155179e-01 -5.47651410e-01
-1.45151687e+00 -2.15780929e-01 -4.79668587e-01 3.36891860e-01
4.40690815e-01 -1.57628190e-02 -9.20715690e-01 4.64626551e-01
-3.79651815e-01 4.34899151e-01 1.13583004e+00 1.16340923e+00
-7.71184921e-01 -1.75324225e+00 -4.07329232e-01 -1.57990515e-01
-3.67930442e-01 1.59019276e-01 -1.16035998e+00 1.21119559e+00
-4.67058212e-01 1.51046634e+00 -6.65388584e-01 3.93559486e-02
1.81674063e-01 4.29685384e-01 4.66977745e-01 -8.31476510e-01
-7.72398591e-01 -1.43681183e-01 9.75006282e-01 -4.19989198e-01
-4.83768910e-01 -4.39609349e-01 -1.43943620e+00 -5.28432608e-01
-3.19854230e-01 8.90940249e-01 5.19002616e-01 1.26701069e+00
-9.05640423e-03 1.33871481e-01 4.70806837e-01 1.69808909e-01
-1.03033781e+00 -7.61337519e-01 -2.67680734e-01 -3.16677094e-01
1.90700620e-01 -1.45284146e-01 -3.90836567e-01 -1.77267998e-01] | [12.533985137939453, 8.01231861114502] |
5f948349-3bdb-43c8-bd1e-4cbd6e8c5937 | promptner-a-prompting-method-for-few-shot | 2305.12217 | null | https://arxiv.org/abs/2305.12217v1 | https://arxiv.org/pdf/2305.12217v1.pdf | PromptNER: A Prompting Method for Few-shot Named Entity Recognition via k Nearest Neighbor Search | Few-shot Named Entity Recognition (NER) is a task aiming to identify named entities via limited annotated samples. Recently, prototypical networks have shown promising performance in few-shot NER. Most of prototypical networks will utilize the entities from the support set to construct label prototypes and use the query set to compute span-level similarities and optimize these label prototype representations. However, these methods are usually unsuitable for fine-tuning in the target domain, where only the support set is available. In this paper, we propose PromptNER: a novel prompting method for few-shot NER via k nearest neighbor search. We use prompts that contains entity category information to construct label prototypes, which enables our model to fine-tune with only the support set. Our approach achieves excellent transfer learning ability, and extensive experiments on the Few-NERD and CrossNER datasets demonstrate that our model achieves superior performance over state-of-the-art methods. | ['Xipeng Qiu', 'Yaqian Zhou', 'Hang Yan', 'Mozhi Zhang'] | 2023-05-20 | null | null | null | null | ['few-shot-ner', 'named-entity-recognition-ner'] | ['natural-language-processing', 'natural-language-processing'] | [-2.98486054e-01 -6.09400384e-02 -4.17882264e-01 -6.08573437e-01
-7.38466442e-01 -6.04460776e-01 4.02715445e-01 2.17633516e-01
-9.23213780e-01 4.61565465e-01 2.67652184e-01 2.34846100e-01
-1.70973897e-01 -8.07690024e-01 -1.19056180e-01 -3.41921419e-01
1.21004321e-01 5.44930756e-01 4.50024843e-01 -2.58501858e-01
1.08845256e-01 5.54069102e-01 -1.44275892e+00 1.79351568e-01
8.37344766e-01 9.21224713e-01 1.36667356e-01 2.50785083e-01
-7.05946505e-01 8.87077212e-01 -4.12529260e-01 -4.23947930e-01
1.95453823e-01 -1.10121399e-01 -9.03448045e-01 -5.65469563e-01
2.97771662e-01 -9.24416855e-02 -4.58588362e-01 9.21784341e-01
9.60395992e-01 9.37840700e-01 6.68797731e-01 -9.50479209e-01
-9.23414230e-01 7.85152912e-01 4.58120443e-02 4.44150269e-01
1.91193819e-01 3.59092690e-02 1.23575449e+00 -1.30351937e+00
8.36802959e-01 9.27494526e-01 1.00719738e+00 1.02894580e+00
-1.06712115e+00 -7.88899422e-01 7.06223473e-02 3.64147246e-01
-1.60592735e+00 -6.79091513e-01 5.13013601e-01 -1.34833634e-01
1.24188554e+00 -1.10779107e-01 1.05450965e-01 1.02276039e+00
-5.21573305e-01 7.57731795e-01 6.33851051e-01 -4.52942461e-01
5.83858073e-01 4.60503474e-02 7.36550689e-01 6.11122847e-01
2.02246732e-03 9.15997699e-02 -4.41596538e-01 -3.82017672e-01
4.60155874e-01 2.76741117e-01 -1.39414757e-01 -2.60306716e-01
-1.10784340e+00 8.29294741e-01 7.26105213e-01 6.70067430e-01
-5.96397638e-01 -2.77606606e-01 3.93981546e-01 1.62396595e-01
3.32089663e-01 8.57956469e-01 -6.24332309e-01 -1.14997648e-01
-1.01716793e+00 -2.54849911e-01 1.16080403e+00 1.16951537e+00
8.73320878e-01 -1.21194214e-01 -7.35662103e-01 1.29054511e+00
-1.10740922e-01 1.26501366e-01 7.92897582e-01 -6.94489896e-01
5.02784789e-01 6.63339913e-01 2.20288023e-01 -6.82954729e-01
-5.44195116e-01 8.06412622e-02 -8.10256898e-01 -4.35903192e-01
1.82146147e-01 -4.30619329e-01 -1.16824734e+00 1.62487853e+00
4.80471224e-01 7.76471019e-01 2.30710536e-01 8.08080971e-01
1.27558970e+00 7.35089660e-01 6.10409617e-01 -1.96920469e-01
1.35197663e+00 -1.12757266e+00 -4.88131255e-01 -1.74588829e-01
1.02341127e+00 -2.51141965e-01 1.17117631e+00 -2.28296623e-01
-4.78740275e-01 -4.75244015e-01 -7.67084837e-01 2.82053519e-02
-7.03725278e-01 1.29284590e-01 4.40969586e-01 6.09433949e-01
-6.66990280e-01 1.00814807e+00 -5.71162999e-01 -6.36349678e-01
3.77963692e-01 1.95443213e-01 -4.58726019e-01 -2.77858406e-01
-1.58924937e+00 9.40502644e-01 8.38531792e-01 -2.92327136e-01
-6.11303508e-01 -1.00127900e+00 -8.88291121e-01 5.45914650e-01
3.93565565e-01 -5.01160145e-01 1.44897711e+00 -1.62097752e-01
-1.36699283e+00 8.42510521e-01 -3.68888713e-02 -3.61726761e-01
-8.77108052e-02 -2.26683971e-02 -8.74326468e-01 8.23354349e-02
2.40276873e-01 6.43831611e-01 2.27582961e-01 -6.92099035e-01
-6.54728234e-01 1.27897933e-02 -4.01068218e-02 1.50395557e-01
-9.06511545e-01 2.99885899e-01 -2.78217584e-01 -4.90432829e-01
-1.10876605e-01 -6.49811327e-01 -4.82780963e-01 -3.59835923e-01
-3.91338617e-01 -8.67218316e-01 5.80399752e-01 -9.71857607e-02
1.50759459e+00 -2.14325786e+00 -5.48554361e-01 3.80566418e-02
1.67029664e-01 6.51824951e-01 -4.89405364e-01 5.15820265e-01
-2.12478563e-01 2.33874738e-01 7.62486458e-02 -3.06988079e-02
2.57053643e-01 4.36241291e-02 -3.29840809e-01 1.42026454e-01
1.48032948e-01 1.09469783e+00 -1.20147026e+00 -6.63912654e-01
-5.89634553e-02 2.53498524e-01 -2.05019891e-01 5.03607512e-01
5.51625155e-02 -3.86465602e-02 -6.30566597e-01 4.79439706e-01
2.75340915e-01 -4.16281730e-01 2.37780716e-02 -3.90838176e-01
-1.19019397e-01 2.73037493e-01 -1.28781164e+00 1.70415974e+00
-3.01999599e-01 6.21653385e-02 -4.88587260e-01 -8.19392502e-01
1.12021124e+00 5.67637861e-01 3.55620652e-01 -6.54848158e-01
8.81853476e-02 1.83783561e-01 -3.72448295e-01 -4.83083278e-01
4.47191536e-01 -3.97608161e-01 -2.76387602e-01 4.29542810e-01
5.87507308e-01 3.50418001e-01 4.51981068e-01 2.27887094e-01
1.47054994e+00 -3.62719625e-01 7.61192381e-01 8.65039900e-02
1.97733700e-01 1.44410893e-01 1.00436914e+00 1.04323697e+00
-5.53219616e-01 4.11772490e-01 -6.46803677e-02 -5.61915874e-01
-9.93229568e-01 -1.04579377e+00 6.25668885e-03 1.61740637e+00
1.50553793e-01 -4.71972317e-01 -5.69712162e-01 -1.07166946e+00
-2.12502748e-01 1.07288682e+00 -5.96161067e-01 -3.28564376e-01
-4.63093698e-01 -4.05626655e-01 9.39742029e-01 7.07314551e-01
4.36146080e-01 -1.36234605e+00 -2.75803357e-01 3.78516287e-01
-6.66347295e-02 -9.23270881e-01 -8.32772076e-01 3.94301176e-01
-8.28488827e-01 -9.54310000e-01 -9.20698881e-01 -1.16580212e+00
4.58770394e-01 2.94311076e-01 1.10645127e+00 -2.36735001e-01
-3.04204613e-01 3.61681402e-01 -5.83932579e-01 -6.02981560e-02
-6.14105947e-02 3.81669044e-01 2.74675965e-01 -2.01710090e-01
9.21414733e-01 -5.76146066e-01 -3.96076709e-01 4.87768918e-01
-6.99546278e-01 -5.49442470e-01 6.02945626e-01 1.13518810e+00
6.79289639e-01 -7.68314525e-02 7.55143821e-01 -1.18198740e+00
6.33546948e-01 -7.75592029e-01 -2.05072328e-01 6.32104397e-01
-7.48423934e-01 1.77386299e-01 8.78617942e-01 -8.40866327e-01
-1.41351986e+00 1.46610424e-01 8.20105299e-02 -7.14974105e-01
-5.45365095e-01 5.65837741e-01 -1.63082838e-01 1.01963140e-01
1.20921957e+00 8.55916459e-03 -7.71665215e-01 -8.24723542e-01
6.99576020e-01 8.86949658e-01 6.24838173e-01 -5.93079567e-01
6.10636473e-01 1.43814191e-01 -4.61135119e-01 -5.86568236e-01
-1.39700913e+00 -1.19577396e+00 -6.53633893e-01 1.41589075e-01
6.76623762e-01 -9.31868613e-01 -3.78858924e-01 -3.17085758e-02
-1.16515207e+00 -3.27093229e-02 -6.18979394e-01 5.44890106e-01
-3.80783707e-01 -4.84214537e-03 -9.11010444e-01 -6.33037627e-01
-6.87782824e-01 -3.91492754e-01 7.03324854e-01 8.01969945e-01
-5.29166684e-02 -8.84825706e-01 5.65023065e-01 -8.05026293e-02
3.91816705e-01 -3.44022036e-01 6.83723092e-01 -1.76858759e+00
-4.69620951e-04 -3.90276641e-01 -2.74514198e-01 3.97882387e-02
4.35015708e-02 -5.13999522e-01 -1.03806448e+00 -1.58585608e-01
-3.13521326e-01 -4.66246247e-01 8.72506678e-01 -1.60400316e-01
9.43430007e-01 -2.27264762e-01 -6.42853081e-01 5.42122304e-01
1.28044438e+00 8.06337595e-02 4.05148476e-01 1.92655161e-01
7.46311545e-01 4.31898713e-01 8.72883081e-01 5.77976108e-01
2.82566398e-01 4.57691133e-01 -2.15903699e-01 1.44188672e-01
2.55916622e-02 -5.07542253e-01 9.68860369e-03 1.08721316e+00
1.20892331e-01 -1.90112889e-01 -1.19757009e+00 7.39904463e-01
-1.88033712e+00 -1.30014670e+00 4.55585748e-01 1.88856161e+00
9.98687685e-01 -2.31444329e-01 -8.26533213e-02 -5.25294662e-01
1.26771808e+00 9.44228843e-02 -7.86271036e-01 7.93788508e-02
-4.67619561e-02 2.86358505e-01 3.44034404e-01 -2.26588339e-01
-1.41792464e+00 1.31855512e+00 6.21244192e+00 1.20823824e+00
-7.70077348e-01 3.27874482e-01 2.77624875e-01 6.35860711e-02
6.78571761e-02 -3.46347922e-03 -1.35924447e+00 3.90228361e-01
1.33314621e+00 -5.46034157e-01 1.71181723e-01 1.22712243e+00
-1.84660748e-01 5.52368939e-01 -1.18057752e+00 1.02284694e+00
4.74788621e-02 -1.30375266e+00 -2.20968515e-01 -4.72860932e-01
9.44292665e-01 4.30071771e-01 -5.15776336e-01 9.44063365e-01
8.35443437e-01 -7.08235323e-01 1.39613599e-01 7.38064110e-01
9.23659086e-01 -7.25972593e-01 6.80125296e-01 5.64702034e-01
-1.49845123e+00 -3.86915594e-01 -7.51393676e-01 3.47627044e-01
3.24801505e-01 5.46257079e-01 -1.14990199e+00 3.50409776e-01
6.80048823e-01 6.70873940e-01 -3.44452590e-01 1.51115119e+00
-2.60117084e-01 6.82415068e-01 -3.40748250e-01 -3.41657311e-01
2.16658831e-01 2.00195596e-01 3.88607204e-01 1.48724461e+00
4.34673697e-01 6.04138851e-01 4.77729291e-01 6.85195148e-01
-5.48384905e-01 5.23821473e-01 -5.54198205e-01 -4.12586838e-01
1.20914412e+00 1.53820610e+00 -6.74790859e-01 -6.08447075e-01
-5.06128907e-01 1.04301703e+00 9.88260210e-01 2.35845163e-01
-5.28631747e-01 -1.02434456e+00 5.48029542e-01 -3.55106860e-01
5.97216368e-01 2.42078245e-01 3.33774760e-02 -1.41531193e+00
-4.04085279e-01 -4.13101077e-01 9.85580742e-01 -6.48306310e-01
-2.10972595e+00 5.33505499e-01 -2.07859471e-01 -1.30962849e+00
-1.76143363e-01 -4.51216519e-01 -8.28032255e-01 6.39771283e-01
-1.26048183e+00 -1.04207861e+00 -2.19005525e-01 6.27181947e-01
4.68521118e-01 -3.09505552e-01 1.19902527e+00 4.77422237e-01
-6.02761507e-01 8.49874616e-01 2.90967792e-01 6.54520869e-01
1.04028404e+00 -1.14790368e+00 4.74620312e-01 6.66467130e-01
4.43077475e-01 8.99971664e-01 1.99471623e-01 -7.62411058e-01
-8.26616764e-01 -1.26229000e+00 1.16138327e+00 -3.07066828e-01
7.09305525e-01 -2.29409859e-02 -9.63652432e-01 5.60189009e-01
-2.67212123e-01 5.53895712e-01 1.18421853e+00 3.84544492e-01
-8.80743861e-01 1.39201796e-02 -1.17251599e+00 5.40535808e-01
1.26945519e+00 -6.64425671e-01 -1.12275112e+00 1.97636113e-01
1.08520639e+00 -6.67584240e-02 -1.13657355e+00 3.12117279e-01
3.42514038e-01 -5.12424827e-01 1.01141369e+00 -1.06118011e+00
-1.42628223e-01 -3.40543628e-01 -9.59038138e-02 -1.59270668e+00
-7.31908500e-01 -2.31768623e-01 -2.53287643e-01 1.72172022e+00
4.66436267e-01 -3.39169443e-01 7.54438162e-01 8.74452412e-01
-2.12098286e-01 -4.28082615e-01 -7.95609653e-01 -1.07286906e+00
-2.03789219e-01 -1.15044311e-01 7.86210001e-01 1.57370377e+00
2.81360090e-01 5.91002464e-01 -2.85873353e-01 6.90829307e-02
3.65036219e-01 1.53443798e-01 2.46798724e-01 -1.74547768e+00
-7.70164728e-02 -1.11697972e-01 -3.15297753e-01 -8.48586142e-01
4.70212460e-01 -1.06590092e+00 4.57623512e-01 -1.39256763e+00
3.18783045e-01 -7.37331867e-01 -9.01080787e-01 1.04082656e+00
-4.67722535e-01 -4.73297425e-02 2.30197772e-01 4.21567440e-01
-1.25796092e+00 5.15361547e-01 6.51835442e-01 -1.55791372e-01
-3.95120353e-01 -1.98433757e-01 -6.34986877e-01 5.85755467e-01
6.52197182e-01 -7.97336578e-01 -2.01284245e-01 -1.77100956e-01
-6.65412564e-03 -1.55811116e-01 -1.91458538e-01 -1.05388451e+00
9.02303338e-01 -3.19454521e-01 3.18121493e-01 -3.94362032e-01
1.25082105e-01 -5.31940520e-01 -1.83917638e-02 -1.29213231e-02
-6.94322348e-01 -2.91296005e-01 -2.68659025e-01 7.38571584e-01
-1.09136090e-01 -6.90861166e-01 8.16089749e-01 -2.34719768e-01
-1.49710739e+00 7.93388546e-01 5.85675575e-02 6.03558958e-01
9.01474535e-01 2.54529137e-02 -4.29257303e-01 1.18351623e-01
-8.04093421e-01 3.44302833e-01 3.06361288e-01 4.86694247e-01
5.35916805e-01 -1.65624070e+00 -5.40497005e-01 -2.77736634e-01
7.16301560e-01 -2.25343913e-01 4.38285083e-01 3.84921610e-01
8.70169401e-02 2.72362888e-01 -2.38596573e-02 -1.21874539e-02
-8.99585247e-01 9.19477522e-01 2.59128779e-01 -4.02720094e-01
-5.98944724e-01 1.02853894e+00 3.40236872e-02 -9.71566379e-01
2.07501262e-01 2.33003348e-01 -6.71681345e-01 2.85389751e-01
9.78557169e-01 4.93264794e-01 -1.04986303e-01 -4.38988090e-01
-3.57691824e-01 1.90267786e-01 -2.23915383e-01 9.24865752e-02
1.47245800e+00 -3.80141437e-02 1.62632123e-01 5.81832647e-01
1.23475611e+00 -3.57417732e-01 -9.02300119e-01 -8.40931535e-01
5.94329178e-01 -1.13930076e-01 -7.83254281e-02 -7.12970972e-01
-7.72579312e-01 6.38179064e-01 5.40882945e-01 -3.43091972e-02
8.27478230e-01 8.78883079e-02 1.01630247e+00 1.18713403e+00
4.56831038e-01 -1.43287539e+00 -6.46980405e-02 9.70527589e-01
6.67693242e-02 -1.12878668e+00 -5.00869215e-01 -2.31306911e-01
-6.00182354e-01 9.94234741e-01 8.42643678e-01 -6.32423908e-02
6.92013621e-01 -4.01206352e-02 7.66641498e-02 -1.21952780e-01
-8.23146820e-01 -6.36938751e-01 3.56225371e-01 6.25431776e-01
3.19492549e-01 -6.22399226e-02 -3.10900629e-01 1.28106391e+00
2.14937612e-01 4.77543138e-02 5.96592091e-02 8.54886115e-01
-9.32773113e-01 -1.03505576e+00 -6.34689257e-02 7.08702385e-01
-2.71118432e-01 -4.66318190e-01 -1.90754592e-01 2.72572309e-01
7.76620805e-02 9.38308895e-01 -1.19563541e-03 -4.03307766e-01
6.39707148e-01 6.33164585e-01 -1.24435805e-01 -1.06855083e+00
-7.06387639e-01 -6.07131898e-01 3.22758287e-01 -5.92472970e-01
-1.58141091e-01 -1.88857734e-01 -1.43929172e+00 -5.49869090e-02
-7.07056522e-01 6.83465898e-01 3.33568275e-01 9.84340668e-01
6.65182889e-01 8.75116512e-02 7.30257690e-01 -5.82529962e-01
-9.57383215e-01 -1.08076179e+00 -8.25521588e-01 6.32119119e-01
-1.95011973e-01 -7.65792608e-01 -2.45724380e-01 -1.13857992e-01] | [9.638422012329102, 9.303979873657227] |
646f4a57-8fbc-4405-9564-09d0bc3f9df1 | the-potential-of-visual-chatgpt-for-remote | 2304.13009 | null | https://arxiv.org/abs/2304.13009v2 | https://arxiv.org/pdf/2304.13009v2.pdf | The Potential of Visual ChatGPT For Remote Sensing | Recent advancements in Natural Language Processing (NLP), particularly in Large Language Models (LLMs), associated with deep learning-based computer vision techniques, have shown substantial potential for automating a variety of tasks. One notable model is Visual ChatGPT, which combines ChatGPT's LLM capabilities with visual computation to enable effective image analysis. The model's ability to process images based on textual inputs can revolutionize diverse fields. However, its application in the remote sensing domain remains unexplored. This is the first paper to examine the potential of Visual ChatGPT, a cutting-edge LLM founded on the GPT architecture, to tackle the aspects of image processing related to the remote sensing domain. Among its current capabilities, Visual ChatGPT can generate textual descriptions of images, perform canny edge and straight line detection, and conduct image segmentation. These offer valuable insights into image content and facilitate the interpretation and extraction of information. By exploring the applicability of these techniques within publicly available datasets of satellite images, we demonstrate the current model's limitations in dealing with remote sensing images, highlighting its challenges and future prospects. Although still in early development, we believe that the combination of LLMs and visual models holds a significant potential to transform remote sensing image processing, creating accessible and practical application opportunities in the field. | ['José Marcato Junior', 'Ana Paula Marques Ramos', 'Wesley Nunes Gonçalves', 'Eduardo Lopes de Lemos', 'Lucas Prado Osco'] | 2023-04-25 | null | null | null | null | ['line-detection'] | ['computer-vision'] | [ 4.89359379e-01 3.14208120e-03 -1.49609186e-02 -5.05744182e-02
-5.61388850e-01 -8.47754657e-01 6.97210193e-01 1.08933531e-01
-3.54023457e-01 2.14244708e-01 -3.31625938e-01 -1.10223854e+00
-3.47296260e-02 -1.06384146e+00 -3.66860092e-01 -7.14017630e-01
-3.74287128e-01 2.39200786e-01 -3.78435515e-02 -1.90622360e-01
2.70948648e-01 9.56322968e-01 -1.57204151e+00 4.63948339e-01
8.74525070e-01 7.92664170e-01 7.93476582e-01 7.56609797e-01
-7.22913980e-01 7.28966057e-01 -4.69008327e-01 -3.54353152e-02
2.44285807e-01 7.11864531e-02 -7.78045952e-01 4.66863304e-01
2.12430328e-01 -2.31339216e-01 1.89742967e-01 9.55901861e-01
3.70100528e-01 -2.27319431e-02 5.90150833e-01 -1.14588416e+00
-8.95441651e-01 2.75397956e-01 -7.09035873e-01 3.74305278e-01
1.04503736e-01 6.17519796e-01 1.02542293e+00 -9.31741476e-01
4.80987847e-01 1.31833386e+00 7.21805394e-01 1.52034014e-01
-1.05477405e+00 -2.48940051e-01 2.29835436e-01 1.04571596e-01
-1.39126456e+00 3.11913770e-02 4.30193841e-01 -6.81730866e-01
1.25307333e+00 3.98601860e-01 8.24425638e-01 4.06588525e-01
2.45208740e-01 9.67018485e-01 1.35884964e+00 -6.88884676e-01
1.78703457e-01 1.33441240e-02 -9.26174819e-02 7.53733635e-01
-1.53895289e-01 1.19731590e-01 -2.23604098e-01 1.74366355e-01
7.53257632e-01 -9.85317156e-02 -2.36919865e-01 1.47499815e-01
-1.21282506e+00 9.98839080e-01 7.98380971e-01 3.09541911e-01
-3.91838700e-01 6.30006148e-03 1.59961239e-01 5.14500774e-02
7.06541240e-01 3.05821419e-01 -2.40618840e-01 3.46695393e-01
-1.17430675e+00 6.80691078e-02 6.28685296e-01 7.05586553e-01
8.64459813e-01 2.63112754e-01 -7.11354613e-02 5.77234864e-01
6.23969674e-01 8.68340909e-01 1.87636539e-01 -7.25600600e-01
5.65002024e-01 7.72300124e-01 -2.09866390e-01 -1.22245252e+00
-5.33551991e-01 -1.34075955e-01 -8.01533997e-01 7.04836309e-01
1.59105003e-01 -1.49409771e-01 -1.05161417e+00 8.55940342e-01
9.43394974e-02 -2.61799663e-01 1.02201886e-01 9.07926202e-01
1.10922360e+00 1.41568947e+00 6.02311373e-01 1.73307583e-01
1.36543775e+00 -7.34792769e-01 -2.70658076e-01 -7.74072886e-01
4.39109832e-01 -6.00248158e-01 1.29899728e+00 2.24482000e-01
-6.27057254e-01 -6.02205992e-01 -7.48867691e-01 -1.29313722e-01
-7.93032825e-01 4.95659262e-01 7.56147802e-01 5.08087099e-01
-1.37870336e+00 2.11895213e-01 -7.86988676e-01 -7.17992008e-01
6.35934234e-01 1.25729233e-01 -2.44056195e-01 8.91491398e-03
-9.69861031e-01 9.75817204e-01 6.56930506e-01 6.90036297e-01
-7.75754213e-01 -5.00798464e-01 -1.02927637e+00 2.11336628e-01
3.88504565e-01 -5.40287495e-01 1.11724472e+00 -9.06715512e-01
-1.19118631e+00 1.14351118e+00 3.63258459e-02 -5.09788275e-01
4.79243934e-01 1.17990868e-02 -1.13953888e-01 5.01644969e-01
7.22979233e-02 1.06998610e+00 7.90533721e-01 -1.13775897e+00
-6.13310337e-01 -3.53742748e-01 -1.21809602e-01 4.17787343e-01
-1.31969288e-01 1.98410377e-01 -4.85423923e-01 -4.33933705e-01
8.07137415e-03 -7.04817533e-01 -6.17081404e-01 7.77938604e-01
-1.98586851e-01 -5.07061593e-02 9.70741034e-01 -7.17504263e-01
7.46388674e-01 -2.15396070e+00 -1.65528819e-01 2.06680760e-01
1.72465459e-01 8.03392887e-01 -3.46771181e-01 7.98520982e-01
2.50847042e-01 5.88095486e-01 -4.66312021e-01 -1.79112609e-02
-6.39928132e-02 4.08939004e-01 -4.74731117e-01 3.25359523e-01
5.09833038e-01 1.36250830e+00 -7.06107318e-01 -6.54367030e-01
8.21402848e-01 4.86056417e-01 4.14229371e-03 3.87429669e-02
-5.31414211e-01 3.51360530e-01 -5.02263188e-01 5.68166137e-01
7.23670483e-01 -2.40223840e-01 1.81747630e-01 1.46834359e-01
-6.86929584e-01 -3.25773835e-01 -7.54542649e-01 1.10366142e+00
-4.03347254e-01 1.19594169e+00 2.44120300e-01 -1.02487314e+00
1.07851470e+00 6.80830106e-02 5.99671062e-03 -8.71371388e-01
-5.50193116e-02 -5.46266697e-02 -3.99963588e-01 -1.00377834e+00
5.29536009e-01 -2.32302025e-01 2.91031897e-01 4.58696574e-01
-2.77184993e-01 -5.09465814e-01 2.10971549e-01 6.11634739e-02
1.70422301e-01 3.99260312e-01 6.22073472e-01 -1.42423660e-01
5.19223213e-01 6.04149461e-01 -1.75161436e-01 9.08960879e-01
-1.29493371e-01 4.16224808e-01 1.95421517e-01 -6.15410447e-01
-1.01673472e+00 -8.23853552e-01 7.26093799e-02 1.14506078e+00
-1.58402145e-01 -3.54607850e-02 -5.04171669e-01 -2.74125725e-01
-1.35654047e-01 5.23676634e-01 -3.56582552e-01 5.06995559e-01
-2.65197307e-01 -1.01920700e+00 6.27789021e-01 4.69174922e-01
8.06665003e-01 -1.51052427e+00 -1.04574788e+00 1.63826793e-01
-8.30001980e-02 -1.30442035e+00 3.66178244e-01 5.44093996e-02
-1.06865251e+00 -9.83061254e-01 -7.02967644e-01 -8.54470909e-01
6.59282267e-01 7.19684303e-01 1.01913631e+00 1.45527065e-01
-5.54397464e-01 4.50774640e-01 -4.60667610e-01 -5.00241518e-01
-5.73612750e-01 3.22566777e-02 -9.28294539e-01 -3.37940514e-01
2.65817791e-01 -3.14165324e-01 -4.05448794e-01 -1.57814279e-01
-1.35076416e+00 3.53092939e-01 8.26315284e-01 3.78350109e-01
4.20500457e-01 1.72125608e-01 5.86060109e-03 -7.53362775e-01
4.57780808e-01 -4.41328317e-01 -7.07985163e-01 4.84207332e-01
-3.54519129e-01 -1.50959462e-01 5.54968238e-01 8.36676806e-02
-1.19275963e+00 1.87236011e-01 -2.26079091e-01 1.14139728e-01
-5.81398904e-01 1.22740209e+00 5.75210042e-02 -2.12556362e-01
6.48590684e-01 5.36813736e-01 1.30057812e-01 -4.34396982e-01
7.22427964e-01 8.01901281e-01 5.49584925e-01 -1.99974820e-01
8.25169802e-01 6.94788337e-01 -1.29877806e-01 -1.50423026e+00
-6.32234573e-01 -5.81799388e-01 -5.98760366e-01 -3.11344296e-01
9.41608965e-01 -9.23108697e-01 -5.14281034e-01 5.87743282e-01
-1.04978192e+00 -4.90061402e-01 8.12284723e-02 7.16177151e-02
-2.11200029e-01 7.73454309e-01 -4.74411219e-01 -1.01809299e+00
-7.88075924e-01 -8.60620499e-01 1.31530035e+00 3.90942246e-01
1.42447963e-01 -1.17289090e+00 -3.88266385e-01 4.30856168e-01
4.32031304e-01 2.11831927e-01 1.02073717e+00 -2.70886421e-01
-6.52490497e-01 3.32982130e-02 -5.90688348e-01 3.35510880e-01
-1.41162470e-01 4.42422777e-01 -9.60501134e-01 -2.57796980e-02
-2.11628124e-01 -1.36648968e-01 8.56626570e-01 3.82217854e-01
8.61148298e-01 -3.22200418e-01 -2.83882201e-01 5.90580940e-01
1.79142368e+00 1.84659734e-01 8.33858669e-01 6.81881607e-01
7.35066473e-01 7.54730940e-01 5.25097847e-01 3.59273553e-01
4.30472881e-01 3.10798079e-01 7.54719436e-01 -7.40973711e-01
2.47726077e-03 -1.24597892e-01 3.22019935e-01 2.60762632e-01
-1.37866974e-01 -3.93129617e-01 -1.49767935e+00 3.92573059e-01
-1.67038953e+00 -9.47521508e-01 -4.86651808e-01 1.54539919e+00
3.43736589e-01 -3.03447634e-01 -2.88332313e-01 7.80519396e-02
4.98869985e-01 4.03880507e-01 -2.57685065e-01 -4.05821800e-01
-3.18247020e-01 2.06696287e-01 4.17920887e-01 3.80968660e-01
-1.27626443e+00 1.32724369e+00 6.49302292e+00 6.82993889e-01
-1.54652834e+00 -3.06486785e-01 6.26906455e-01 5.85754633e-01
-1.43618628e-01 -3.51621397e-02 -7.10101843e-01 3.00726779e-02
7.24836171e-01 1.25294134e-01 2.60062724e-01 7.62946546e-01
8.75353932e-01 -4.74793315e-01 -4.70804334e-01 7.21037388e-01
7.52399862e-02 -1.53049850e+00 5.44602096e-01 4.42533493e-02
4.84017670e-01 2.80030847e-01 1.31310642e-01 1.06462933e-01
1.54805198e-01 -1.16694927e+00 7.41180897e-01 2.55852997e-01
6.54036701e-01 -5.07211685e-01 6.33401573e-01 5.76869905e-01
-1.40022123e+00 -1.56272575e-01 -4.55552697e-01 -2.91830599e-01
1.60844564e-01 3.66175264e-01 -1.20149219e+00 6.50656402e-01
7.85697222e-01 8.47771764e-01 -9.00971293e-01 1.07679987e+00
-4.54090774e-01 5.25292814e-01 -1.58980668e-01 -8.24474357e-03
7.03421056e-01 -4.62003350e-01 4.30444658e-01 1.50762105e+00
2.77540267e-01 8.50151032e-02 4.50453103e-01 1.05204320e+00
4.90556180e-01 3.15240920e-01 -7.34338343e-01 -4.48803246e-01
1.35793820e-01 1.38590622e+00 -1.20023465e+00 -3.59159201e-01
-4.37778324e-01 7.72464395e-01 1.77894849e-02 5.05882740e-01
-6.48600042e-01 -1.82082415e-01 1.40250593e-01 -8.29019248e-02
3.21452558e-01 -7.17293143e-01 -5.11651695e-01 -9.77252424e-01
-8.74682367e-02 -9.70618546e-01 3.67603540e-01 -1.37710321e+00
-9.20322955e-01 6.63680375e-01 3.98681834e-02 -1.17521679e+00
-3.55530493e-02 -1.00547111e+00 -5.51419497e-01 8.88637066e-01
-2.03546643e+00 -1.80436087e+00 -4.88476813e-01 3.78961951e-01
8.18239689e-01 1.38973460e-01 7.66681910e-01 -1.42629504e-01
-2.19892368e-01 -3.48547071e-01 1.40161857e-01 2.56974578e-01
-4.46201302e-02 -1.05270207e+00 7.71929801e-01 1.05298793e+00
4.55960691e-01 2.86994427e-01 5.70625663e-01 -3.38985056e-01
-1.38044071e+00 -1.31321871e+00 8.93943191e-01 -9.03092846e-02
6.13480985e-01 -1.66607812e-01 -9.19042587e-01 6.28842294e-01
1.48632035e-01 -1.75233290e-01 6.15457773e-01 -6.51540160e-01
-1.53078988e-01 1.59792423e-01 -8.66463184e-01 7.18126535e-01
4.42056239e-01 -6.40431881e-01 -6.13291025e-01 5.06126642e-01
2.43688554e-01 -1.80329025e-01 -3.33689839e-01 1.80667769e-02
2.59676933e-01 -9.43855703e-01 1.12318826e+00 -3.14482450e-01
5.38186669e-01 -5.60323596e-01 1.01478845e-01 -1.05735624e+00
-1.11862898e-01 -2.97303855e-01 5.19129872e-01 1.06046212e+00
4.22233760e-01 -4.36913669e-01 4.51196194e-01 3.21327031e-01
-1.89045116e-01 -2.64016628e-01 -4.68279988e-01 -5.67725837e-01
1.53462201e-01 -7.96435833e-01 2.68843025e-01 6.09010935e-01
-3.56268466e-01 2.29628488e-01 -2.55766183e-01 4.49008673e-01
3.48308861e-01 4.00242358e-01 7.20546782e-01 -1.39584351e+00
-9.93082896e-02 -4.89402950e-01 -2.55531639e-01 -1.00667429e+00
1.51629180e-01 -8.87523413e-01 1.67919934e-01 -2.13935661e+00
8.93107876e-02 -2.06693470e-01 4.45634156e-01 9.37720478e-01
-3.54438834e-03 6.46708012e-01 5.84526539e-01 2.88058162e-01
-2.83583462e-01 2.30940387e-01 1.25510120e+00 -4.72682476e-01
-2.62465417e-01 -2.05332711e-01 -6.03793442e-01 7.87862539e-01
9.51147079e-01 -2.14941978e-01 -2.58376598e-01 -6.97876036e-01
3.82350236e-01 -1.38115436e-01 7.84481883e-01 -8.25393677e-01
1.94342852e-01 -3.73755753e-01 3.61310869e-01 -4.94374454e-01
-8.68063793e-02 -7.66309619e-01 6.17480762e-02 4.41431463e-01
-1.21079460e-01 -9.46082361e-03 6.67869151e-01 3.72660249e-01
-3.22443157e-01 -1.53169215e-01 5.94588757e-01 -6.41625941e-01
-1.57527471e+00 9.80185717e-02 -1.13006735e+00 -3.19383770e-01
1.09764147e+00 -5.04217029e-01 -4.46484745e-01 -2.04436123e-01
-8.54216516e-01 2.49063686e-01 1.45781994e-01 3.66491348e-01
7.12501407e-01 -5.06195307e-01 -8.38461578e-01 2.96398342e-01
2.41143242e-01 3.52823362e-02 3.35512698e-01 6.13149703e-01
-1.21893406e+00 4.72523987e-01 -3.69626671e-01 -8.02663863e-01
-1.31356168e+00 3.88716727e-01 2.16218337e-01 -2.26360485e-01
-9.86381888e-01 4.87111568e-01 2.14704081e-01 -3.52718443e-01
-8.74419808e-02 -5.08940041e-01 -5.64632952e-01 1.62307516e-01
6.69289649e-01 1.71268374e-01 2.63994206e-02 -7.37482190e-01
-2.22901031e-01 6.66196346e-01 3.77681553e-01 -3.17385137e-01
1.52044022e+00 -4.10484076e-01 -3.23892623e-01 2.88801253e-01
7.00320601e-01 -3.71730685e-01 -1.11928546e+00 -7.41756484e-02
1.87519416e-01 -2.84734249e-01 2.87581414e-01 -9.10447061e-01
-1.12632477e+00 1.43172967e+00 6.39491141e-01 2.94740200e-01
1.18480575e+00 -5.22927381e-02 2.96029329e-01 5.30431509e-01
2.49837160e-01 -6.86216414e-01 -2.36001849e-01 6.86238766e-01
9.00280416e-01 -1.41272068e+00 5.66158928e-02 -4.37029213e-01
-8.45918655e-01 1.54540396e+00 2.15028614e-01 3.30312192e-01
3.81328255e-01 1.87133312e-01 6.01882994e-01 -4.47761238e-01
-3.17885429e-01 -6.01071358e-01 3.43033135e-01 1.03762352e+00
3.68417054e-01 1.10898651e-01 -2.15422332e-01 -1.72676325e-01
2.28805970e-02 5.73517606e-02 4.21693921e-01 1.01551116e+00
-8.62516880e-01 -7.87589967e-01 -6.29412532e-01 1.18254542e-01
-3.08426023e-01 -6.06389225e-01 -4.48624462e-01 9.35247004e-01
3.01232766e-02 9.83209610e-01 -3.23495939e-02 1.77035943e-01
-1.00012757e-01 1.16876913e-02 5.75108975e-02 -7.77517736e-01
-7.34521449e-01 6.03924207e-02 -9.84865353e-02 -2.14633763e-01
-6.99378490e-01 -3.40264618e-01 -1.25419009e+00 -1.44743472e-01
8.06234926e-02 -1.90464463e-02 8.50358069e-01 1.17199814e+00
2.60438770e-01 9.82179046e-02 1.33846626e-01 -1.18805611e+00
-2.53925145e-01 -6.93880618e-01 -6.07111931e-01 -1.05740689e-01
2.94434339e-01 1.40002361e-02 -2.03742266e-01 4.48722273e-01] | [9.659543991088867, -1.2276585102081299] |
67f8fd4a-765a-44a6-bd75-2c9c9446a890 | unified-recurrence-modeling-for-video-action-1 | 2206.01009 | null | https://arxiv.org/abs/2206.01009v1 | https://arxiv.org/pdf/2206.01009v1.pdf | Unified Recurrence Modeling for Video Action Anticipation | Forecasting future events based on evidence of current conditions is an innate skill of human beings, and key for predicting the outcome of any decision making. In artificial vision for example, we would like to predict the next human action before it happens, without observing the future video frames associated to it. Computer vision models for action anticipation are expected to collect the subtle evidence in the preamble of the target actions. In prior studies recurrence modeling often leads to better performance, the strong temporal inference is assumed to be a key element for reasonable prediction. To this end, we propose a unified recurrence modeling for video action anticipation via message passing framework. The information flow in space-time can be described by the interaction between vertices and edges, and the changes of vertices for each incoming frame reflects the underlying dynamics. Our model leverages self-attention as the building blocks for each of the message passing functions. In addition, we introduce different edge learning strategies that can be end-to-end optimized to gain better flexibility for the connectivity between vertices. Our experimental results demonstrate that our proposed method outperforms previous works on the large-scale EPIC-Kitchen dataset. | ['Oswald Lanz', 'Simon See', 'Cheng-Kuang Lee', 'Giuseppe Fiameni', 'Tsung-Ming Tai'] | 2022-06-02 | unified-recurrence-modeling-for-video-action | https://openreview.net/forum?id=6j9YOwh8itH | https://openreview.net/pdf?id=6j9YOwh8itH | null | ['action-anticipation'] | ['computer-vision'] | [ 2.77631462e-01 1.76099986e-02 -4.60584670e-01 -3.99625063e-01
-1.41591588e-02 -9.99618471e-02 7.91246116e-01 2.24907324e-01
-1.18999481e-01 3.05747449e-01 7.50791073e-01 -1.18549272e-01
-5.80304079e-02 -6.75388932e-01 -8.20044160e-01 -4.94510055e-01
-4.88053828e-01 7.87764564e-02 3.34108651e-01 9.42462757e-02
1.72333807e-01 1.57074675e-01 -1.18968713e+00 4.93067294e-01
4.40759569e-01 1.06680501e+00 3.30330223e-01 9.15893495e-01
2.91982979e-01 1.55658758e+00 3.59510094e-01 -3.40014160e-01
1.09069958e-01 -2.96025395e-01 -5.44606328e-01 3.67091805e-01
2.28850141e-01 -6.05435193e-01 -1.23790777e+00 6.87968314e-01
1.06647223e-01 2.22105920e-01 5.27038991e-01 -1.27024293e+00
-3.09032798e-01 7.41459548e-01 -6.79757655e-01 5.21554708e-01
3.46403748e-01 6.28039181e-01 1.03717470e+00 -6.16830230e-01
6.93545401e-01 1.23386037e+00 5.61255395e-01 2.25001320e-01
-8.05890143e-01 -3.70261282e-01 8.46530735e-01 9.60265338e-01
-1.02782536e+00 -4.89272058e-01 1.04351199e+00 -4.88805324e-01
8.34945261e-01 -3.39462385e-02 1.07283056e+00 1.16645491e+00
5.69645524e-01 1.20550561e+00 4.57620829e-01 2.51663309e-02
1.27959788e-01 -5.40772974e-01 4.04785238e-02 8.93821180e-01
-1.41102493e-01 1.81025222e-01 -8.62190306e-01 1.82572126e-01
8.78421724e-01 6.09467804e-01 -3.11275423e-01 -2.41864324e-01
-1.38430738e+00 5.98130584e-01 5.90037942e-01 -2.69442588e-01
-7.89595068e-01 7.00134993e-01 4.89874601e-01 1.52026350e-03
4.97669488e-01 -2.57931173e-01 -3.48898709e-01 -4.61456627e-01
-4.25623327e-01 9.59805623e-02 4.66125160e-01 7.58771777e-01
5.59548438e-01 -1.63527444e-01 -5.64041376e-01 1.65518865e-01
3.82330954e-01 4.47776429e-02 1.65779680e-01 -1.08427107e+00
5.34064829e-01 6.46943986e-01 2.39844233e-01 -1.37512922e+00
-3.56945843e-01 -3.33584219e-01 -8.46049309e-01 -1.95347339e-01
3.45950514e-01 -2.67619282e-01 -6.56794250e-01 1.75893438e+00
4.51600790e-01 1.29387891e+00 -2.14250833e-01 8.00133407e-01
4.80587333e-01 8.03683102e-01 1.40905350e-01 -2.86725968e-01
1.11241102e+00 -1.12354946e+00 -5.55070162e-01 -5.16287506e-01
6.23870909e-01 -3.09187412e-01 6.30581915e-01 1.22247614e-01
-9.20071840e-01 -6.61944747e-01 -5.86297035e-01 -1.38517499e-01
2.63210028e-01 7.57911652e-02 9.49139953e-01 -1.57540947e-01
-7.57815421e-01 7.28962541e-01 -1.21001875e+00 -3.04102778e-01
5.64597845e-01 -4.83913068e-03 -1.97160125e-01 -2.57121861e-01
-1.02529609e+00 4.92746115e-01 2.85868019e-01 4.93198901e-01
-1.13456142e+00 -5.23755312e-01 -8.97922575e-01 2.59916097e-01
6.13157153e-01 -8.00858498e-01 1.07911420e+00 -8.98983359e-01
-1.21107912e+00 3.52448046e-01 -4.01751071e-01 -9.98954356e-01
6.72971308e-01 -3.32121074e-01 -2.04707906e-01 1.68150961e-01
-1.39430046e-01 8.00229192e-01 8.93653870e-01 -7.77833164e-01
-8.51479828e-01 -1.44392535e-01 2.32007131e-01 3.73211920e-01
-1.50894925e-01 -2.40644202e-01 -8.22508037e-01 -6.05733633e-01
4.80642579e-02 -1.01978374e+00 -4.57374841e-01 1.52767047e-01
-1.92735836e-01 -2.75274247e-01 7.13759720e-01 -6.25596941e-01
1.38260031e+00 -2.27028990e+00 1.63208410e-01 -5.49886823e-02
3.56755823e-01 -1.21638417e-01 -9.36372057e-02 4.07596856e-01
1.33637208e-02 -3.09771359e-01 1.60914779e-01 -3.03000271e-01
-1.12266876e-01 1.91116601e-01 -5.11944830e-01 6.33516431e-01
1.97613865e-01 1.16448021e+00 -1.31846261e+00 -4.63299394e-01
3.21709126e-01 6.92971706e-01 -5.90212882e-01 1.01295494e-01
-4.41872478e-01 5.84845960e-01 -7.26099432e-01 4.00373399e-01
2.78400294e-02 -6.33858800e-01 1.35609910e-01 -2.35267699e-01
8.09495524e-02 2.46297434e-01 -9.77391601e-01 1.61296546e+00
-1.05776459e-01 7.68236458e-01 -4.87563163e-01 -1.00687754e+00
4.15771812e-01 3.14059049e-01 7.22797513e-01 -5.38711190e-01
-2.15129387e-02 -5.30328691e-01 7.18806237e-02 -7.47862637e-01
3.65937799e-01 4.18045312e-01 1.87925875e-01 4.42141145e-01
-4.68128115e-01 3.76676530e-01 2.48364076e-01 6.49526834e-01
1.27843118e+00 3.50425839e-01 2.57730871e-01 3.21889371e-01
2.26271957e-01 -1.15279831e-01 1.03574324e+00 6.96723521e-01
-3.90767276e-01 4.16058719e-01 7.17354596e-01 -8.61273527e-01
-7.77232587e-01 -8.09533834e-01 4.17018563e-01 8.87205184e-01
2.06415772e-01 -4.57523018e-01 -4.51956093e-01 -5.65268517e-01
-1.67323366e-01 6.68962061e-01 -7.77983308e-01 -3.51803601e-01
-6.05163634e-01 -4.05846775e-01 -1.87514529e-01 8.05456877e-01
5.60571790e-01 -1.20408082e+00 -8.38036001e-01 3.79706800e-01
-4.85763758e-01 -1.24778879e+00 -8.00379276e-01 -4.11960036e-01
-9.34777558e-01 -1.15057468e+00 -3.89607757e-01 -4.99098271e-01
5.67098200e-01 4.92002934e-01 1.16283548e+00 3.48242372e-01
-2.59706020e-01 8.53892267e-01 -3.40247869e-01 -1.63752541e-01
5.40193915e-02 -4.07918215e-01 1.04248077e-01 6.77879751e-01
3.20329100e-01 -6.55919731e-01 -1.11482823e+00 8.27872828e-02
-6.06973469e-01 5.10889113e-01 5.91809034e-01 5.99753380e-01
6.73414648e-01 8.13777000e-02 2.41807297e-01 -6.26179695e-01
2.48698778e-02 -7.58566737e-01 -3.27376932e-01 3.42574418e-01
-2.14588061e-01 -1.24270670e-01 3.40467542e-01 -5.46354592e-01
-1.13013434e+00 4.69175965e-01 3.25085789e-01 -5.74383557e-01
-2.98707969e-02 6.79903805e-01 8.86404887e-02 3.27787787e-01
6.11090958e-02 4.00784671e-01 -3.26846361e-01 -1.10720135e-01
4.12737131e-01 1.00305699e-01 5.46794415e-01 -3.83713603e-01
1.19539075e-01 8.38313997e-01 9.02957767e-02 -7.15913653e-01
-9.41248178e-01 -5.11432528e-01 -5.55885673e-01 -8.46491277e-01
7.31897473e-01 -1.23625910e+00 -1.07162976e+00 6.39983892e-01
-1.12636602e+00 -6.14626944e-01 -1.82798624e-01 6.14556909e-01
-6.39192820e-01 5.11324823e-01 -6.75133944e-01 -8.16307068e-01
9.54193994e-03 -9.88306344e-01 9.68561471e-01 2.37346321e-01
-2.51955956e-01 -1.19129872e+00 2.20247917e-02 2.34215289e-01
-1.49004966e-01 1.99952036e-01 5.75642645e-01 -5.05345576e-02
-1.02482319e+00 -2.54534662e-01 -2.58625746e-01 -2.08333552e-01
2.98907962e-02 9.98481065e-02 -5.74505925e-01 -6.93106055e-02
-4.34242114e-02 -4.00125422e-02 1.28261983e+00 7.66368508e-01
1.20901120e+00 -3.73150796e-01 -6.70713723e-01 4.15362179e-01
1.02342534e+00 -6.05763905e-02 8.12303662e-01 -4.18698825e-02
9.31179643e-01 5.33603609e-01 7.28689611e-01 7.71252930e-01
9.31606293e-01 5.00825763e-01 6.33851945e-01 1.34717584e-01
-2.02546611e-01 -5.06725132e-01 7.57349372e-01 5.63863099e-01
-2.60068953e-01 -3.91822845e-01 -8.78298581e-01 6.96517706e-01
-2.46154618e+00 -1.45060825e+00 -3.00021023e-01 2.07982302e+00
4.07813132e-01 2.83016205e-01 5.03016934e-02 -2.33433634e-01
6.16634429e-01 4.75737274e-01 -9.72187638e-01 2.28531927e-01
1.44985795e-01 -5.99466681e-01 2.66581714e-01 5.18680632e-01
-1.17971778e+00 9.60692108e-01 5.91116619e+00 4.15258318e-01
-9.64701295e-01 -1.48371965e-01 1.02039266e+00 -3.51331711e-01
-2.38940958e-02 2.42948428e-01 -6.25581503e-01 6.67847872e-01
7.33141065e-01 -9.05692205e-02 3.18863511e-01 5.00300884e-01
6.61909223e-01 -4.07645702e-01 -1.35799313e+00 8.89125109e-01
-1.43403150e-02 -1.49965036e+00 -9.11170989e-03 -1.47065878e-01
6.58942103e-01 1.67433456e-01 8.15239176e-02 2.66699772e-02
5.22047997e-01 -6.84385955e-01 6.00034058e-01 1.09004080e+00
1.90166146e-01 -4.91233617e-01 1.99736774e-01 3.56697470e-01
-1.52614701e+00 -3.05868596e-01 -2.18317360e-01 -5.61501443e-01
5.76093793e-01 6.21736288e-01 -7.02959716e-01 2.99631387e-01
5.15303969e-01 1.58380747e+00 -4.48550969e-01 1.06354880e+00
-3.40225101e-01 8.64348054e-01 -2.22639948e-01 1.10082179e-01
2.63285220e-01 -2.97636509e-01 5.50055802e-01 8.90416861e-01
2.77308285e-01 4.79872465e-01 4.86992747e-01 3.50229561e-01
5.81883900e-02 -2.96162724e-01 -6.23649597e-01 -8.30000192e-02
1.12191051e-01 1.02036130e+00 -8.13580036e-01 -5.89297652e-01
-7.14772582e-01 1.18440163e+00 5.38633287e-01 5.60437262e-01
-1.02116430e+00 3.68370920e-01 7.96813011e-01 4.03795503e-02
4.71123904e-01 -4.00314331e-01 -1.17878787e-01 -1.22282326e+00
1.23994760e-01 -4.83518779e-01 6.25936806e-01 -8.38188529e-01
-1.09017456e+00 1.06436819e-01 -3.30381781e-01 -1.33297765e+00
-2.64629066e-01 -2.27549866e-01 -1.01549649e+00 3.24188620e-01
-1.43164015e+00 -1.11829674e+00 -3.08445334e-01 5.39599776e-01
8.90584230e-01 2.68473804e-01 2.07553461e-01 4.81454283e-03
-7.27941275e-01 -6.92847893e-02 -1.87705234e-01 2.46016145e-01
4.55190778e-01 -8.28405678e-01 7.01994598e-01 1.20303321e+00
5.42541265e-01 1.31979123e-01 7.30868697e-01 -1.09451032e+00
-1.47352159e+00 -1.28542721e+00 1.00035584e+00 -5.17447114e-01
8.53680313e-01 -5.10983393e-02 -8.57244849e-01 1.11638570e+00
5.59529439e-02 2.19814271e-01 3.12362969e-01 6.39168993e-02
-5.20098545e-02 -1.57029182e-02 -3.57206553e-01 8.94067049e-01
1.42533743e+00 -4.11416709e-01 -2.43281171e-01 5.50208092e-01
6.53130651e-01 -3.71508509e-01 -5.24870276e-01 2.82294571e-01
6.75490439e-01 -9.65887487e-01 1.06313741e+00 -8.03292394e-01
8.23643148e-01 -3.05128932e-01 1.96439937e-01 -1.10199511e+00
-6.52103782e-01 -7.20169246e-01 -6.80306017e-01 9.36145365e-01
2.04558566e-01 -2.02753961e-01 8.75112236e-01 9.71555233e-01
-4.94975820e-02 -8.34776938e-01 -6.25080585e-01 -3.99980813e-01
-7.08070815e-01 -7.76157439e-01 3.82153630e-01 6.47737801e-01
8.29414427e-02 2.70110428e-01 -9.60515797e-01 4.11330372e-01
5.51229537e-01 2.09948704e-01 8.47860754e-01 -9.31144059e-01
-4.47680950e-01 -3.15834820e-01 -5.40968776e-01 -1.58910108e+00
1.28879458e-01 -6.13230407e-01 1.07006736e-01 -1.56973505e+00
5.18319607e-01 -2.26504341e-01 -4.08771604e-01 2.70139992e-01
-5.97934902e-01 -5.74365966e-02 3.14424425e-01 1.44328475e-01
-1.04583538e+00 7.74121761e-01 1.28163230e+00 -1.66314572e-01
-1.68540925e-01 9.61897001e-02 -3.66919458e-01 1.04825640e+00
6.02594197e-01 -3.01780045e-01 -6.24502599e-01 -6.45266294e-01
5.14414608e-01 4.15463597e-01 6.60792768e-01 -9.88711715e-01
6.34520531e-01 -4.48338926e-01 3.36407363e-01 -6.03923202e-01
5.45608521e-01 -8.42598498e-01 2.02025861e-01 4.59600091e-01
-4.57875699e-01 5.22004254e-02 -3.19908470e-01 1.37744701e+00
-3.43251787e-02 3.04283530e-01 2.19047219e-01 -4.96739447e-02
-1.23460925e+00 1.04321921e+00 -2.74006933e-01 -4.20944802e-02
1.26095033e+00 -9.63934734e-02 2.44156849e-02 -7.90755212e-01
-6.87801421e-01 6.88974082e-01 3.46553981e-01 4.74428475e-01
7.32074559e-01 -1.33058321e+00 -8.01051736e-01 -1.31611705e-01
6.61493242e-02 -9.02552158e-02 6.71620309e-01 1.18293464e+00
-2.06911415e-01 8.27834308e-02 3.38216610e-02 -6.95387840e-01
-1.27318537e+00 7.74359584e-01 -1.96481328e-02 -3.34703475e-01
-1.09231770e+00 8.51918459e-01 5.11023700e-01 4.23943758e-01
4.46734220e-01 -4.61193889e-01 -3.65985215e-01 -8.04611519e-02
7.64515102e-01 4.94015068e-01 -5.94483852e-01 -6.52294636e-01
-1.65277973e-01 3.39682668e-01 -2.33574867e-01 1.35944590e-01
1.43682837e+00 -4.37847137e-01 1.49203837e-01 5.69185495e-01
7.50946164e-01 -4.75837380e-01 -2.11606312e+00 -5.13252258e-01
-2.60905474e-01 -6.12011015e-01 2.33511895e-01 -3.47237706e-01
-1.26671743e+00 8.16244721e-01 2.28878930e-01 -4.76039909e-02
1.11987317e+00 6.83811903e-02 1.10706317e+00 3.59033495e-01
3.42476547e-01 -1.04040384e+00 1.62865952e-01 4.26975965e-01
6.65511429e-01 -1.44059646e+00 4.89674583e-02 -4.88526046e-01
-9.26356137e-01 1.02113521e+00 3.85598570e-01 -1.81330025e-01
7.13101029e-01 -6.45624027e-02 -3.36139143e-01 -2.94334948e-01
-1.33293164e+00 -1.70587808e-01 3.32310438e-01 2.61265635e-01
1.88035205e-01 4.41729240e-02 1.69218387e-02 3.74057025e-01
4.96383458e-01 2.98894823e-01 3.18824261e-01 6.44264698e-01
-3.26211482e-01 -6.80712998e-01 8.99848435e-03 6.26554012e-01
-1.75686792e-01 -5.78075722e-02 -2.15992779e-01 2.07705185e-01
-7.60412291e-02 9.24632668e-01 3.54655027e-01 -2.84016401e-01
5.65246120e-03 -1.21663667e-01 4.16677922e-01 -2.94845700e-01
-1.07012585e-01 -1.03987426e-01 -3.68635333e-03 -1.09355783e+00
-5.53623855e-01 -1.18329179e+00 -1.25668204e+00 -3.95005822e-01
8.67744908e-02 -3.45696837e-01 2.47016810e-02 1.13191569e+00
7.26579070e-01 6.86313152e-01 6.63159490e-01 -8.62350881e-01
-1.57006755e-02 -5.94578445e-01 -2.08886936e-01 6.22170687e-01
3.38844359e-01 -6.81296527e-01 -5.56080267e-02 5.53789556e-01] | [8.23252010345459, 0.4354982078075409] |
975e1b95-a647-48c4-a487-f7a207082d5f | a-multi-platform-annotation-ecosystem-for | null | null | https://aclanthology.org/W19-4021 | https://aclanthology.org/W19-4021.pdf | A Multi-Platform Annotation Ecosystem for Domain Adaptation | This paper describes an ecosystem consisting of three independent text annotation platforms. To demonstrate their ability to work in concert, we illustrate how to use them to address an interactive domain adaptation task in biomedical entity recognition. The platforms and the approach are in general domain-independent and can be readily applied to other areas of science. | ['Keith Suderman', 'Jan-Christoph Klie', 'Jin-Dong Kim', 'Richard Eckart de Castilho', 'Nancy Ide'] | 2019-08-01 | null | null | null | ws-2019-8 | ['text-annotation'] | ['natural-language-processing'] | [ 3.62896234e-01 5.16934037e-01 -8.71050432e-02 -4.86251891e-01
-6.11373603e-01 -7.63900638e-01 5.32269597e-01 5.72520494e-01
-6.69220209e-01 1.12533438e+00 -2.93449610e-02 -3.86800408e-01
-1.94426894e-01 -2.04027519e-01 -2.54325449e-01 -3.88102651e-01
2.53707729e-03 1.05546820e+00 5.67270875e-01 -5.01577377e-01
5.88027611e-02 3.25703204e-01 -1.25232601e+00 4.23113763e-01
6.02972746e-01 9.01723430e-02 -1.50705591e-01 6.14483953e-01
-4.89904881e-01 3.06675434e-01 -8.75342488e-01 -3.12571526e-01
-1.72829390e-01 -1.56305179e-01 -1.33366168e+00 -3.11619461e-01
-7.05457926e-02 1.82179391e-01 2.03526258e-01 7.81746924e-01
9.82560813e-01 -1.75196938e-02 7.09751785e-01 -1.00663090e+00
-1.90835506e-01 4.52705264e-01 -3.72724142e-03 3.29891205e-01
7.64809132e-01 -4.89708334e-01 6.49524629e-01 -9.17365670e-01
1.32805419e+00 7.65401959e-01 9.42838430e-01 7.45693803e-01
-1.17332304e+00 -4.56282854e-01 -5.68034127e-03 -6.08807839e-02
-1.43197513e+00 -5.44194281e-01 3.35137725e-01 -6.46065116e-01
1.26751697e+00 3.43390942e-01 4.53071773e-01 1.06505036e+00
-1.11171871e-01 5.73478222e-01 7.94700086e-01 -7.40898669e-01
2.47382492e-01 6.06603801e-01 3.16448927e-01 6.16019607e-01
5.65983891e-01 -4.76764053e-01 -5.51545262e-01 -6.30888939e-01
4.74228919e-01 -4.10246313e-01 -1.05076164e-01 -4.23738360e-01
-1.43400919e+00 5.51431775e-01 -2.46498764e-01 8.92608464e-01
-3.17287445e-02 -3.40014100e-01 9.20538306e-01 3.51681471e-01
5.75695932e-01 7.65646935e-01 -7.86942840e-01 7.52142742e-02
-9.37167227e-01 2.84174234e-01 1.38544810e+00 1.26313162e+00
2.44164050e-01 -5.03562808e-01 -2.49118507e-02 6.59778714e-01
2.66395599e-01 -7.29479194e-02 6.76142931e-01 -4.41426337e-01
1.37052625e-01 6.06095254e-01 3.43228519e-01 -3.47992450e-01
-9.23979461e-01 -1.09666046e-02 -4.34172839e-01 -1.21837407e-01
4.50500995e-01 -3.66985202e-01 -7.95390069e-01 1.13903856e+00
5.64755201e-01 1.24821678e-01 3.21089149e-01 4.60396975e-01
1.40366697e+00 4.82838079e-02 7.71107256e-01 3.49880010e-02
1.82144058e+00 -5.23130000e-01 -1.11632240e+00 -1.04787543e-01
1.06797147e+00 -9.97486115e-01 6.37642622e-01 1.84304059e-01
-8.77444863e-01 -3.16651493e-01 -9.07117009e-01 -1.39950484e-01
-1.22336149e+00 3.37203071e-02 5.87342501e-01 7.46647477e-01
-1.05771303e+00 2.88652897e-01 -8.08856428e-01 -9.97593582e-01
2.55675852e-01 4.70228881e-01 -6.49657369e-01 3.82550955e-01
-1.59925783e+00 1.11228466e+00 1.00148058e+00 -4.00685459e-01
-1.98984906e-01 -5.24517179e-01 -5.81835568e-01 -3.06581080e-01
1.56686872e-01 -9.94259298e-01 1.39431381e+00 -6.53195262e-01
-1.39456606e+00 1.53582251e+00 1.01586223e-01 -1.88839197e-01
5.55587173e-01 -1.87729239e-01 -8.23123872e-01 1.31535172e-01
1.66418090e-01 4.14250433e-01 2.31436133e-01 -6.37559056e-01
-7.27834940e-01 -2.80113131e-01 -3.66877466e-01 2.68110752e-01
-6.45017862e-01 4.81931210e-01 -6.60518110e-01 -7.95162439e-01
-3.43083680e-01 -7.14845896e-01 -4.28820014e-01 -1.87470376e-01
-2.91421473e-01 -2.98907846e-01 6.43955171e-01 -5.19190669e-01
1.30675507e+00 -2.00574255e+00 6.94538355e-02 1.44045815e-01
2.30100691e-01 3.06040585e-01 2.95969248e-01 8.86645675e-01
-1.35691896e-01 4.18494582e-01 -3.77205014e-01 -9.74481255e-02
3.82278599e-02 3.58926058e-01 1.96737945e-01 2.90792465e-01
2.25775436e-01 8.66080880e-01 -1.10336149e+00 -7.66889691e-01
-1.13491252e-01 3.89212430e-01 -2.00427309e-01 -2.12283283e-02
-1.70640469e-01 5.99219322e-01 -8.92196596e-01 5.87406635e-01
2.78519124e-01 -4.23295230e-01 6.70484424e-01 -2.15775054e-02
-2.23039612e-01 2.23355442e-01 -1.10986280e+00 2.03940678e+00
1.21256799e-01 7.27022529e-01 2.64585316e-01 -1.13732672e+00
7.98481703e-01 1.09105504e+00 5.92152536e-01 9.53661799e-02
3.15077603e-03 4.56592351e-01 -3.14636111e-01 -8.18616152e-01
5.05757809e-01 -2.35726640e-01 -3.67154270e-01 5.00007510e-01
5.70046186e-01 -1.48732429e-02 1.39333427e-01 1.59833819e-01
1.14025545e+00 2.71078050e-01 1.15441072e+00 -4.32876825e-01
7.93706298e-01 4.30536658e-01 2.45834142e-01 6.19536459e-01
-3.64430249e-01 1.52008131e-01 2.19378904e-01 -5.79905748e-01
-9.52109993e-01 -5.90624094e-01 -5.22153974e-01 1.38359499e+00
-2.36311127e-02 -4.53511924e-01 -7.26732612e-01 -1.09997439e+00
-1.41814440e-01 2.88211703e-01 -7.01786160e-01 4.03526366e-01
-2.73813516e-01 -7.78080702e-01 1.01716781e+00 4.09934402e-01
1.32485643e-01 -9.72506702e-01 -6.26538217e-01 4.20533597e-01
-8.53241086e-02 -1.13190055e+00 3.81371081e-02 4.23226982e-01
-6.89297318e-01 -1.05671048e+00 -1.06292784e+00 -9.75607753e-01
6.01035297e-01 -3.13164413e-01 1.30687785e+00 -9.67913121e-02
-4.87461597e-01 7.95894325e-01 -4.10115093e-01 -1.08372533e+00
-6.81033969e-01 7.53440261e-01 -1.17236093e-01 -4.77676213e-01
9.57224131e-01 -3.42699349e-01 -3.70374560e-01 5.44769585e-01
-1.06734121e+00 -2.30420500e-01 3.15839350e-01 8.30270827e-01
5.47018647e-01 -5.09084284e-01 7.68192708e-01 -1.69332886e+00
9.35943067e-01 -5.66713810e-01 -2.53882021e-01 5.72646856e-01
-6.56547010e-01 -2.09490076e-01 2.66257059e-02 -3.30335021e-01
-1.03932691e+00 6.92134142e-01 -4.66129899e-01 3.32049221e-01
-6.33179188e-01 7.48238087e-01 -9.41753834e-02 -1.65739745e-01
1.03917527e+00 -1.18580967e-01 -2.18538091e-01 -6.86702013e-01
5.71428955e-01 8.89281273e-01 4.70956087e-01 -1.79614514e-01
4.56814557e-01 3.06751758e-01 -2.79248416e-01 -5.83828449e-01
-2.44017020e-01 -1.01803470e+00 -1.06369781e+00 -5.76360077e-02
1.05128121e+00 -9.62157726e-01 -2.70220220e-01 2.96936780e-02
-1.15504539e+00 -3.68983448e-02 -3.58054966e-01 3.66685301e-01
-3.61295044e-01 3.86135578e-01 -3.16011012e-01 -3.31766069e-01
-3.12273204e-01 -6.64372444e-01 1.09104490e+00 1.69816375e-01
-6.14319563e-01 -1.55146241e+00 3.68277758e-01 1.14085212e-01
1.68331742e-01 2.84031808e-01 5.17482877e-01 -1.85216916e+00
4.61838663e-01 -4.59148943e-01 1.90027371e-01 -3.41811478e-01
2.20762685e-01 -1.06177934e-01 -1.11112225e+00 -3.57974023e-02
-2.31176168e-01 -3.60466950e-02 3.65284503e-01 -2.14936346e-01
7.11593628e-01 2.41875462e-03 -1.13994503e+00 5.74790500e-03
1.01483679e+00 8.53975117e-02 4.74921346e-01 6.14635110e-01
3.33861530e-01 7.67104864e-01 6.38131678e-01 2.97860146e-01
1.55218512e-01 9.04418111e-01 -2.98116714e-01 -5.98604918e-01
2.38739356e-01 3.54623616e-01 -2.66813755e-01 5.48308253e-01
-1.57754809e-01 -1.44257754e-01 -1.60686851e+00 5.30786514e-01
-2.01183844e+00 -9.01882887e-01 -2.96399564e-01 1.69508111e+00
1.09217167e+00 -9.78624448e-02 2.26210684e-01 -2.51010597e-01
7.36537516e-01 -5.23740649e-01 -3.67045552e-01 -3.93606663e-01
-8.74676704e-02 3.08466852e-01 5.89489281e-01 1.90886647e-01
-1.38628638e+00 9.89550948e-01 8.92955971e+00 6.21216774e-01
-6.53572142e-01 4.27954316e-01 3.06890905e-01 3.44065398e-01
-2.05615629e-02 -3.06504905e-01 -7.62603462e-01 1.15731582e-02
1.19201982e+00 -2.91028023e-01 -3.99032056e-01 8.07564437e-01
-1.46760494e-01 1.72402382e-01 -1.22145367e+00 5.47771215e-01
-1.53747112e-01 -1.44980729e+00 -2.24130452e-01 -1.90899983e-01
3.76762480e-01 1.63046837e-01 -3.21835905e-01 2.15657324e-01
4.82368529e-01 -9.59574342e-01 1.46578103e-01 4.70517904e-01
8.91780913e-01 -4.28320840e-04 1.00669730e+00 7.73717836e-02
-8.93869758e-01 2.51511544e-01 2.24701036e-02 2.20436722e-01
2.73189425e-01 3.19851428e-01 -1.32448196e+00 1.01449180e+00
8.21591139e-01 5.50566733e-01 -3.78528148e-01 1.06032991e+00
-2.11178109e-01 2.69171983e-01 -2.59721518e-01 -6.99640214e-02
-1.81264281e-01 3.75227273e-01 6.79915428e-01 2.09302139e+00
1.37791336e-01 1.81605205e-01 6.68812096e-02 2.10337937e-01
-3.53599489e-02 6.72988772e-01 -8.15836787e-01 -1.47794783e-01
5.37073374e-01 1.32219577e+00 -1.13593936e+00 -7.17683017e-01
-6.01706147e-01 1.23018181e+00 1.98133856e-01 1.94393039e-01
-5.03479242e-01 -8.02740455e-01 4.23977494e-01 -1.63509250e-01
3.37654829e-01 2.20484100e-02 -3.02094877e-01 -1.13485360e+00
-3.16940606e-01 -9.19643164e-01 9.14189458e-01 -4.95218039e-01
-1.39369035e+00 6.63144886e-01 1.31529987e-01 -1.27446055e+00
-1.41736597e-01 -6.07715487e-01 -2.49936745e-01 8.49707484e-01
-1.19131374e+00 -9.87801254e-01 -1.11380294e-01 5.81799865e-01
1.62811399e-01 -3.89831066e-01 1.63470078e+00 7.33185291e-01
-5.40048301e-01 5.90765595e-01 4.46540654e-01 2.35013574e-01
1.36554432e+00 -1.37845898e+00 7.17894435e-01 2.94093102e-01
6.63357899e-02 7.39958227e-01 8.17701936e-01 -7.07303286e-01
-8.89417410e-01 -1.03318155e+00 1.17254806e+00 -9.72488225e-01
6.61071181e-01 -5.54032803e-01 -1.09796929e+00 8.28379691e-01
5.03907323e-01 8.89902711e-02 1.52834761e+00 2.43728399e-01
-1.80530787e-01 4.21194851e-01 -1.36787570e+00 2.72358179e-01
9.80025768e-01 -2.74278820e-01 -8.69228780e-01 7.23596275e-01
2.86645412e-01 -6.24926805e-01 -1.26846921e+00 2.09958330e-01
5.70585430e-01 -5.45883060e-01 9.92352009e-01 -9.97977853e-01
-2.98028708e-01 -2.75260627e-01 2.87839174e-01 -1.04703689e+00
-1.77842587e-01 -8.26326728e-01 6.84258193e-02 1.31986642e+00
8.86964381e-01 -9.87269878e-01 4.44614351e-01 7.12408543e-01
-9.45315659e-02 -1.26821399e-01 -1.22652435e+00 -6.30123734e-01
1.04125932e-01 -3.37329395e-02 5.84430814e-01 1.48343635e+00
8.67852330e-01 5.75648665e-01 1.14984535e-01 -4.46889549e-03
-1.69771194e-01 -3.70805800e-01 6.86415195e-01 -1.79774177e+00
-1.09733164e-01 -5.18689394e-01 -5.19274712e-01 -6.10451341e-01
-1.54392183e-01 -8.93379211e-01 3.58923227e-02 -1.58474731e+00
3.58728245e-02 -5.17544568e-01 -5.73629141e-01 8.37775946e-01
-1.95733681e-01 3.93942982e-01 -3.12798083e-01 5.40418744e-01
-8.90677929e-01 -3.28616828e-01 3.83638024e-01 -6.67455569e-02
-4.03199613e-01 -2.32589871e-01 -8.17233443e-01 6.85732245e-01
8.46229017e-01 -7.22720027e-01 -1.72723293e-01 -1.47828788e-01
4.28937346e-01 -3.33148897e-01 -1.78594708e-01 -7.52197146e-01
3.62239897e-01 8.76607895e-02 3.99746299e-01 -7.95237944e-02
-1.92674562e-01 -8.17824602e-01 5.07023454e-01 3.53198886e-01
-4.22598511e-01 -1.44437999e-01 7.50005007e-01 6.39619708e-01
-2.15135753e-01 -4.00076181e-01 3.60543668e-01 -3.24241012e-01
-8.41444314e-01 -1.67625144e-01 -9.81269956e-01 -1.06048390e-01
1.38570285e+00 -3.06799203e-01 -1.54530868e-01 1.61668360e-01
-1.29220819e+00 2.22392350e-01 4.49696302e-01 2.42893204e-01
2.07424946e-02 -9.39451337e-01 -7.03032315e-01 -1.21047899e-01
7.10652053e-01 -3.91313136e-01 -6.50668368e-02 8.11637998e-01
-6.68552399e-01 6.60477281e-01 -3.54840279e-01 -4.79010284e-01
-1.62008917e+00 5.14535785e-01 5.54660082e-01 -6.28815055e-01
-5.00544429e-01 7.01536059e-01 -1.34245262e-01 -6.50040865e-01
1.25063896e-01 1.45430237e-01 -7.08661199e-01 3.40300739e-01
6.16446972e-01 1.88099239e-02 3.70198220e-01 -3.53855580e-01
-8.00890684e-01 1.69893920e-01 -5.70841059e-02 -2.01162413e-01
1.37485266e+00 -3.04921031e-01 -2.52099186e-02 6.63191080e-01
6.73625231e-01 6.09559454e-02 -3.02054167e-01 -1.69897422e-01
6.24156594e-01 3.03069472e-01 -2.81391561e-01 -1.17119718e+00
-3.55278254e-01 5.87852061e-01 8.67756963e-01 5.85895002e-01
1.05363071e+00 2.96893060e-01 2.37304866e-01 5.93174279e-01
2.39497535e-02 -1.38279009e+00 -6.68013632e-01 3.52976620e-01
5.65214097e-01 -1.22940803e+00 3.25681090e-01 -6.03770137e-01
-5.36420107e-01 1.47537434e+00 2.25593120e-01 5.22888184e-01
8.25061262e-01 3.58824968e-01 2.63510406e-01 -6.16074920e-01
-4.48739350e-01 -2.53101498e-01 4.34060872e-01 1.02347207e+00
1.13157964e+00 -2.01342762e-01 -9.09494519e-01 6.96011782e-01
3.84163290e-01 3.99473965e-01 1.52921662e-01 1.38548505e+00
-2.76736796e-01 -1.66841221e+00 -3.60282183e-01 2.84839660e-01
-9.59561110e-01 -1.17346644e-07 -9.48226333e-01 9.68647718e-01
1.06540158e-01 5.59237719e-01 -2.52942145e-01 -2.22159475e-02
5.38210034e-01 7.05346465e-01 1.63845062e-01 -9.19190109e-01
-1.00517225e+00 4.18177396e-02 5.81414700e-01 -2.50027031e-01
-7.55398810e-01 -8.58514905e-01 -1.24939978e+00 1.57468393e-01
-2.28504092e-01 4.61593717e-01 6.38514280e-01 6.22859955e-01
7.41871417e-01 4.41696316e-01 -2.18735337e-01 -4.91236359e-01
1.30756244e-01 -1.03110373e+00 -4.21204865e-01 3.82219762e-01
-8.23046267e-02 -5.12965322e-01 1.79933280e-01 6.48700953e-01] | [8.60906982421875, 8.777655601501465] |
a6609e50-71dd-4f6c-8809-0f7b030c3b1f | hierarchical-spatial-sum-product-networks-for | 1511.05292 | null | http://arxiv.org/abs/1511.05292v3 | http://arxiv.org/pdf/1511.05292v3.pdf | Hierarchical Spatial Sum-Product Networks for Action Recognition in Still Images | Recognizing actions from still images is popularly studied recently. In this
paper, we model an action class as a flexible number of spatial configurations
of body parts by proposing a new spatial SPN (Sum-Product Networks). First, we
discover a set of parts in image collections via unsupervised learning. Then,
our new spatial SPN is applied to model the spatial relationship and also the
high-order correlations of parts. To learn robust networks, we further develop
a hierarchical spatial SPN method, which models pairwise spatial relationship
between parts inside sub-images and models the correlation of sub-images via
extra layers of SPN. Our method is shown to be effective on two benchmark
datasets. | ['Jinghua Wang', 'Gang Wang'] | 2015-11-17 | null | null | null | null | ['action-recognition-in-still-images'] | ['computer-vision'] | [ 1.34936318e-01 9.98507887e-02 -3.08472395e-01 -3.46147895e-01
1.57331213e-01 -5.65860391e-01 5.60469687e-01 -3.12821776e-01
5.55800237e-02 2.97864735e-01 4.45476711e-01 6.22927547e-01
-6.64846122e-01 -6.87465966e-01 -1.01134658e+00 -5.86931765e-01
-4.73291487e-01 3.32204282e-01 4.66532290e-01 -5.91962785e-02
-5.45384139e-02 6.13691330e-01 -1.41316819e+00 1.88555047e-01
5.52960813e-01 8.92628789e-01 -6.68443441e-02 2.60521263e-01
2.36364558e-01 6.96083069e-01 -3.36364150e-01 -1.65595472e-01
5.93358517e-01 -4.66574579e-01 -7.77459979e-01 4.96321082e-01
6.49453402e-01 9.31467786e-02 -4.18075472e-01 1.15722525e+00
4.25824411e-02 3.50853473e-01 7.21554577e-01 -1.26553679e+00
-6.40722394e-01 7.84751534e-01 -5.91902077e-01 -1.55904979e-01
1.69941559e-01 -1.17621897e-02 1.19533765e+00 -5.57677329e-01
6.76139832e-01 1.43307567e+00 6.27221823e-01 3.99475157e-01
-1.37695277e+00 -4.89676148e-01 1.65246099e-01 1.40620768e-01
-1.55844510e+00 -6.04802556e-02 1.00987530e+00 -4.26521629e-01
5.66319585e-01 6.06882274e-02 1.04036856e+00 9.72029865e-01
4.11635756e-01 9.45140243e-01 8.15156102e-01 6.95576705e-03
1.66518450e-01 -5.34708440e-01 1.74586803e-01 7.45652378e-01
2.80133039e-01 -1.47458777e-01 -9.07978356e-01 -5.85738346e-02
1.17362285e+00 4.74025667e-01 1.40964419e-01 -9.89142597e-01
-1.18634081e+00 3.02617580e-01 7.86632419e-01 3.95170510e-01
-4.42227751e-01 1.03482448e-01 -3.30289677e-02 -2.75375187e-01
2.55546629e-01 4.58191127e-01 -3.86548311e-01 3.63447994e-01
-8.36030960e-01 2.54774183e-01 5.66307962e-01 9.59727407e-01
6.01014376e-01 -5.08935153e-01 -9.87759382e-02 9.82655644e-01
3.69202316e-01 3.60677749e-01 1.85477495e-01 -1.27756917e+00
3.42750698e-01 9.62282181e-01 -1.63267806e-01 -1.37806475e+00
-6.94904625e-01 -4.54398841e-01 -1.40595937e+00 -2.23292157e-01
2.76539981e-01 1.54182598e-01 -8.01195204e-01 1.87953293e+00
2.40039021e-01 1.52955696e-01 -3.29327464e-01 6.53804541e-01
4.52024758e-01 3.27810109e-01 -2.32980981e-01 1.04827378e-02
1.08013833e+00 -1.07611907e+00 -5.88297665e-01 -1.11600861e-01
3.18707734e-01 -1.20046988e-01 4.50547218e-01 4.46849227e-01
-1.23742855e+00 -6.77769959e-01 -9.21350479e-01 -3.38947736e-02
-3.86305571e-01 2.18666464e-01 5.54871261e-01 7.57667795e-02
-1.08115375e+00 1.00592768e+00 -8.17828953e-01 -3.89218807e-01
7.55614340e-01 7.16146529e-01 -6.45583630e-01 -6.80225119e-02
-8.36252451e-01 4.27829176e-01 4.61452574e-01 3.11616659e-01
-5.93129635e-01 -5.42027354e-01 -1.21787190e+00 1.60211902e-02
5.32767117e-01 -8.85200858e-01 7.99229205e-01 -1.07059884e+00
-1.37687480e+00 4.76528674e-01 1.08948864e-01 -2.66089469e-01
2.10465580e-01 -2.89278328e-01 -2.90427834e-01 2.21627191e-01
1.70448944e-01 9.60784256e-01 6.92239523e-01 -1.13403654e+00
-2.25160882e-01 -6.09061837e-01 1.43462002e-01 3.06730896e-01
-2.04022035e-01 -1.16114408e-01 -5.03410935e-01 -5.38552225e-01
4.91160959e-01 -9.58492875e-01 -4.02312011e-01 2.28007168e-01
-7.95536876e-01 -3.22902411e-01 5.63633442e-01 -4.62923616e-01
8.70230079e-01 -2.38380456e+00 5.71521401e-01 5.30009270e-01
3.64977211e-01 -6.20242115e-03 -2.82559782e-01 1.46425515e-01
-1.11310147e-01 -3.02977152e-02 -3.66381288e-01 -3.02414089e-01
6.92194281e-03 7.47687221e-01 1.84861749e-01 4.87920851e-01
3.08054537e-01 1.03479028e+00 -8.95372987e-01 -6.98410392e-01
1.92941204e-02 4.70777214e-01 -5.93384206e-01 -1.71978518e-01
-6.88671321e-02 4.63891298e-01 -2.74523616e-01 6.55600607e-01
7.96408057e-01 -4.39040989e-01 2.47321218e-01 -3.35543036e-01
9.93662328e-02 -9.94393006e-02 -1.24154234e+00 1.93251586e+00
1.35220572e-01 3.19120996e-02 -1.10052593e-01 -9.89077568e-01
6.20200276e-01 -1.43620923e-01 8.30516636e-01 -5.60474217e-01
6.38943687e-02 -2.07397789e-01 2.20814615e-01 -3.90279770e-01
3.77710283e-01 -6.88222945e-02 -1.42454684e-01 2.28579521e-01
4.15059954e-01 2.51069605e-01 3.52805823e-01 1.97417304e-01
1.13455045e+00 3.01114053e-01 3.45124722e-01 -5.22816777e-01
4.73786741e-01 -2.56978244e-01 7.70051539e-01 5.99698603e-01
-1.68280810e-01 6.76107883e-01 6.52153254e-01 -6.33596659e-01
-7.84980834e-01 -1.37408161e+00 6.01646444e-03 1.00691950e+00
4.50141728e-01 -5.59410393e-01 -6.18019760e-01 -8.56246829e-01
3.07953008e-03 3.82651458e-04 -9.68838871e-01 -2.08556373e-02
-5.64624608e-01 -3.68006647e-01 3.84630919e-01 9.84875023e-01
7.36195564e-01 -1.17947054e+00 -3.11523914e-01 -2.93440104e-01
-1.67668805e-01 -9.41153824e-01 -7.91823685e-01 -5.17788483e-03
-8.97899270e-01 -1.28557205e+00 -8.07874262e-01 -8.29486251e-01
9.49295282e-01 3.11657697e-01 9.66209054e-01 4.76953792e-05
-3.40678722e-01 4.66140628e-01 -2.45954022e-01 -7.41243511e-02
3.58985901e-01 -1.56603098e-01 4.11498636e-01 3.56075794e-01
-1.90541539e-02 -9.17705417e-01 -6.84246421e-01 6.51922941e-01
-9.39598083e-01 4.41929661e-02 8.14347208e-01 7.09462285e-01
9.08568859e-01 3.23534548e-01 -2.35046387e-01 -5.14547944e-01
1.18254840e-01 -3.38451594e-01 -4.11260664e-01 3.11440766e-01
3.07294011e-01 2.56303340e-01 4.61979866e-01 -5.32080293e-01
-8.56330931e-01 4.75023240e-01 4.51359183e-01 -6.91538393e-01
-1.26945168e-01 3.16840947e-01 -4.69760180e-01 1.79998036e-02
3.17291796e-01 4.28224474e-01 2.15784132e-01 -4.99103606e-01
4.24130291e-01 -1.38413846e-01 6.27491891e-01 -5.12884080e-01
6.88026488e-01 7.26989031e-01 6.42252803e-01 -8.67305994e-01
-1.04231477e+00 -5.98989844e-01 -1.46267498e+00 -1.21639028e-01
1.07088864e+00 -6.60993218e-01 -7.31795192e-01 5.43185115e-01
-1.00798774e+00 -3.64183605e-01 -2.90428936e-01 4.35898870e-01
-5.58605134e-01 4.95950222e-01 -5.17143607e-01 -5.13065219e-01
1.08603254e-01 -7.19852805e-01 1.05531502e+00 1.05924085e-01
-1.73904717e-01 -8.44373345e-01 3.02849680e-01 1.95324451e-01
-1.96956888e-01 1.36016876e-01 4.56368566e-01 -5.10533094e-01
-6.71488345e-01 -4.51097712e-02 -1.92385495e-01 4.87955511e-01
2.63603002e-01 7.82032907e-02 -2.04858437e-01 1.36768252e-01
-3.37826341e-01 -8.37078691e-02 8.37161422e-01 7.11000383e-01
1.44929600e+00 -4.93953586e-01 -5.40235996e-01 6.22446477e-01
1.05895603e+00 -7.10901096e-02 6.12388134e-01 -1.33607447e-01
1.00148618e+00 9.35136318e-01 3.01407456e-01 2.84487575e-01
2.51005650e-01 6.30737066e-01 4.67184544e-01 1.92079574e-01
6.80759698e-02 -5.17607749e-01 2.07769156e-01 6.84002280e-01
-5.07110238e-01 9.72345099e-02 -7.48788416e-01 4.77859616e-01
-2.12857676e+00 -1.15571821e+00 -6.22406648e-03 1.98887086e+00
2.39814863e-01 4.34395410e-02 4.85291123e-01 -3.37818682e-01
6.54501259e-01 2.55421847e-01 -4.89523560e-01 2.90305138e-01
-2.68632382e-01 6.07310794e-02 5.23983121e-01 1.19419567e-01
-1.36855984e+00 6.86292231e-01 7.22279835e+00 6.06174529e-01
-4.04493481e-01 -1.69467911e-01 4.83398885e-01 -3.13231587e-01
1.23275317e-01 -1.83470011e-01 -6.61895216e-01 4.01691198e-01
2.16670275e-01 3.09905976e-01 2.60576069e-01 7.12249875e-01
3.54850963e-02 -6.67891726e-02 -1.24860144e+00 9.66548085e-01
3.33788484e-01 -1.16858065e+00 3.17803681e-01 3.59841138e-01
8.70519817e-01 -1.78531483e-01 7.73651078e-02 -1.17289253e-01
5.39526582e-01 -1.04453802e+00 6.62679017e-01 9.36924160e-01
9.21540931e-02 -6.89883769e-01 5.78944504e-01 5.70591450e-01
-1.37053156e+00 -2.75746137e-01 -4.63923901e-01 -2.60520667e-01
7.13411346e-02 2.95993239e-01 -9.62282345e-02 7.00964212e-01
9.08686459e-01 1.32468772e+00 -5.90565026e-01 1.09183669e+00
-3.35085720e-01 1.16935454e-01 -5.25580943e-01 7.61004612e-02
6.79771602e-02 -3.35798711e-01 3.18214089e-01 7.66421854e-01
1.80954665e-01 1.94564983e-01 1.73484668e-01 9.97569859e-01
-7.04449937e-02 -7.97175243e-02 -8.79526854e-01 1.18199058e-01
9.97653417e-03 1.29674017e+00 -8.41498196e-01 -2.16384754e-01
-1.58822015e-01 9.96813774e-01 3.60381961e-01 1.57325402e-01
-6.03881180e-01 2.60561526e-01 5.71364939e-01 1.78702831e-01
4.81434971e-01 -4.01267886e-01 -8.55610594e-02 -1.05471933e+00
2.24858850e-01 -5.44217885e-01 4.28006589e-01 -7.70508468e-01
-1.47896039e+00 1.33148834e-01 3.91486317e-01 -1.43821311e+00
2.37896219e-01 -8.00168216e-01 -5.14914274e-01 3.36906493e-01
-7.08737016e-01 -1.22898054e+00 -5.74917793e-02 6.97124481e-01
3.69191319e-01 1.51857436e-01 3.80448759e-01 -1.14262208e-01
-6.02038205e-01 3.50862741e-01 -9.96863917e-02 5.07293820e-01
3.67743880e-01 -1.20257592e+00 6.68211933e-03 7.24294186e-01
5.56522667e-01 9.44637120e-01 2.97200263e-01 -7.88753927e-01
-1.15921152e+00 -1.16583383e+00 8.64583015e-01 -5.33529043e-01
5.95831513e-01 -5.62731206e-01 -6.45846725e-01 8.63836288e-01
1.57112479e-01 4.34325039e-01 6.20268762e-01 2.82060474e-01
-5.97987175e-01 -2.05214813e-01 -9.29933190e-01 6.29454732e-01
1.67707157e+00 -2.95212567e-01 -5.89195609e-01 3.94724429e-01
5.23739159e-01 -2.13840425e-01 -7.90307641e-01 5.68477869e-01
7.77346492e-01 -1.09582782e+00 1.18523240e+00 -5.24690688e-01
6.83406234e-01 -4.07761455e-01 -1.66621715e-01 -1.24571383e+00
-7.30511963e-01 -3.59989345e-01 -4.41128872e-02 8.54096472e-01
3.29711884e-01 -3.68344069e-01 8.80353868e-01 3.64325136e-01
6.71924353e-02 -6.69426203e-01 -7.82860935e-01 -1.17487919e+00
1.79243982e-02 -2.14294225e-01 4.31063712e-01 7.89220750e-01
2.43910670e-01 2.14937374e-01 -5.38770437e-01 2.71075785e-01
8.62168014e-01 -4.94160093e-02 6.49379253e-01 -1.28365040e+00
-6.29777312e-01 -6.18201971e-01 -8.21792185e-01 -1.22067571e+00
6.45830110e-02 -6.77476764e-01 2.65908837e-01 -1.33876586e+00
6.88829482e-01 1.67025805e-01 -5.00184298e-01 6.22803509e-01
1.99717328e-01 3.13054562e-01 1.54301435e-01 2.32497200e-01
-1.32730365e+00 6.43673301e-01 1.38998652e+00 -2.44750693e-01
-2.61423141e-01 8.32832456e-02 -4.06072706e-01 1.04646409e+00
6.58653140e-01 -2.61562437e-01 -3.64328563e-01 -1.83844030e-01
3.45157444e-01 -1.62186414e-01 7.77744114e-01 -1.34328508e+00
3.99255008e-01 -1.63917765e-01 7.33064950e-01 -1.04849374e+00
4.42943513e-01 -8.52358937e-01 2.54574925e-01 2.67933846e-01
-4.18692589e-01 -7.53608346e-02 -1.56896934e-01 6.99840188e-01
-2.79951930e-01 1.69068530e-01 5.26429117e-01 -3.70354146e-01
-3.95910174e-01 4.85164046e-01 -1.28992081e-01 -3.57955933e-01
1.04157734e+00 -2.32051268e-01 4.27349545e-02 -1.55232713e-01
-1.08622169e+00 3.71911138e-01 3.69905055e-01 3.29024762e-01
5.13592243e-01 -1.65431595e+00 -1.06993079e-01 2.43422359e-01
1.49791732e-01 3.30113679e-01 6.78821325e-01 1.14473045e+00
-2.78274417e-01 3.35891336e-01 -4.95419919e-01 -7.15119720e-01
-1.53143430e+00 6.94224477e-01 4.47162837e-01 -2.76492536e-01
-5.64238608e-01 9.58498836e-01 7.30780423e-01 -5.61328530e-01
1.80630386e-01 -4.49622512e-01 -3.41550261e-01 -1.71523377e-01
3.28980267e-01 3.86678725e-01 -6.90622211e-01 -1.05290353e+00
-4.06033605e-01 8.75824571e-01 2.78837737e-02 1.05249800e-01
1.47860897e+00 1.26077369e-01 -4.84657884e-01 5.45532644e-01
1.14065659e+00 -3.33666623e-01 -1.47562778e+00 -4.43747967e-01
-1.32290915e-01 -2.77072132e-01 -5.00461221e-01 -4.97772872e-01
-1.17912984e+00 6.75100088e-01 2.35513523e-01 1.43098891e-01
1.13561559e+00 5.79809487e-01 2.72375733e-01 4.18215036e-01
3.47211748e-01 -1.23736060e+00 5.16861975e-01 6.32730663e-01
9.42529202e-01 -9.01439250e-01 1.62646055e-01 -7.26855397e-01
-4.62231219e-01 8.47254157e-01 6.24357641e-01 -5.38945436e-01
1.08576131e+00 -8.40686858e-02 -5.51927030e-01 -5.04923046e-01
-4.04171318e-01 -2.42087543e-01 8.01382065e-01 5.72249532e-01
-1.27228294e-02 3.00973147e-01 -2.86763191e-01 7.09424138e-01
-1.96057484e-01 -9.62854475e-02 -6.08197711e-02 9.02895510e-01
-1.11872278e-01 -1.09585798e+00 -3.08502465e-01 4.85831767e-01
-2.01009274e-01 1.18721992e-01 -8.87840152e-01 7.03766644e-01
6.92109585e-01 4.81465727e-01 3.72332156e-01 -5.96355736e-01
3.06074500e-01 -1.15508765e-01 7.93186009e-01 -6.56237423e-01
-2.26522937e-01 1.39256537e-01 -3.19481015e-01 -9.51361179e-01
-9.15905476e-01 -9.65938628e-01 -9.39511418e-01 2.05147624e-01
7.21255764e-02 -2.79880255e-01 1.12878636e-01 1.17076647e+00
3.94248754e-01 3.35152119e-01 4.76200491e-01 -9.68057394e-01
-1.90695569e-01 -9.32176471e-01 -1.19336700e+00 5.73101640e-01
5.32752201e-02 -1.01420748e+00 -1.97828144e-01 4.56657484e-02] | [7.880404949188232, 0.19102603197097778] |
5fc153fa-0cad-41e0-add5-77f0385c37d5 | independent-natural-policy-gradient-methods | 2204.05466 | null | https://arxiv.org/abs/2204.05466v2 | https://arxiv.org/pdf/2204.05466v2.pdf | Independent Natural Policy Gradient Methods for Potential Games: Finite-time Global Convergence with Entropy Regularization | A major challenge in multi-agent systems is that the system complexity grows dramatically with the number of agents as well as the size of their action spaces, which is typical in real world scenarios such as autonomous vehicles, robotic teams, network routing, etc. It is hence in imminent need to design decentralized or independent algorithms where the update of each agent is only based on their local observations without the need of introducing complex communication/coordination mechanisms. In this work, we study the finite-time convergence of independent entropy-regularized natural policy gradient (NPG) methods for potential games, where the difference in an agent's utility function due to unilateral deviation matches exactly that of a common potential function. The proposed entropy-regularized NPG method enables each agent to deploy symmetric, decentralized, and multiplicative updates according to its own payoff. We show that the proposed method converges to the quantal response equilibrium (QRE) -- the equilibrium to the entropy-regularized game -- at a sublinear rate, which is independent of the size of the action space and grows at most sublinearly with the number of agents. Appealingly, the convergence rate further becomes independent with the number of agents for the important special case of identical-interest games, leading to the first method that converges at a dimension-free rate. Our approach can be used as a smoothing technique to find an approximate Nash equilibrium (NE) of the unregularized problem without assuming that stationary policies are isolated. | ['Yuejie Chi', 'Fan Chen', 'Shicong Cen'] | 2022-04-12 | null | null | null | null | ['policy-gradient-methods'] | ['methodology'] | [-2.24895701e-01 4.92917329e-01 7.55704567e-02 3.40272576e-01
-2.81451464e-01 -5.79692602e-01 5.51181793e-01 4.94922131e-01
-1.04019177e+00 1.16465735e+00 -1.90739006e-01 -2.94345737e-01
-4.75187927e-01 -8.29228222e-01 -5.38518488e-01 -1.06952190e+00
-2.49645829e-01 7.58939087e-01 2.23653525e-01 -5.76785624e-01
1.61462456e-01 2.46025473e-01 -1.18346405e+00 -6.59024775e-01
1.09501410e+00 9.75283444e-01 2.60363638e-01 8.31321537e-01
1.11358643e-01 7.51522839e-01 -4.35657322e-01 -5.47237732e-02
6.66508377e-01 -5.71659863e-01 -5.36120951e-01 1.40980527e-01
-5.91573894e-01 -2.01861337e-01 -6.52475432e-02 1.41206491e+00
4.66550797e-01 3.87858868e-01 6.43948019e-01 -1.32155466e+00
-6.32295161e-02 7.02362120e-01 -8.29163134e-01 1.73674468e-02
-1.81151420e-01 -2.46480070e-02 9.15372193e-01 -1.15596555e-01
6.40964687e-01 1.07363665e+00 2.02640221e-01 7.02343464e-01
-1.21279991e+00 -3.73862237e-01 2.92475581e-01 -5.12217656e-02
-1.02928400e+00 5.80410566e-03 4.21294153e-01 -2.76846796e-01
4.85342413e-01 1.08167231e-01 7.54103720e-01 4.21667963e-01
1.34883285e-01 4.25280690e-01 1.06905222e+00 -3.31356227e-01
7.59784818e-01 1.52189314e-01 -2.66497940e-01 5.12950897e-01
5.93768299e-01 -1.98484138e-02 1.67932063e-01 -3.39509934e-01
6.33706391e-01 -1.19207017e-01 -1.95662975e-01 -5.92094719e-01
-8.99498582e-01 9.86656487e-01 3.04301202e-01 2.13904649e-01
-8.59439075e-01 1.86926574e-01 1.68082803e-01 6.66498363e-01
5.72848499e-01 5.71029007e-01 -3.10668856e-01 -1.70454651e-01
-4.87021059e-01 4.70860720e-01 1.11043978e+00 5.24476469e-01
7.21734643e-01 9.46811810e-02 1.74439326e-01 4.82940465e-01
8.18731859e-02 5.31476200e-01 2.52071470e-01 -1.15068364e+00
5.03745973e-01 5.49516499e-01 7.95415282e-01 -8.47420096e-01
-5.91047049e-01 -4.78149205e-01 -9.88642156e-01 5.93164623e-01
7.95089185e-01 -8.17900717e-01 -1.28340721e-01 2.14677763e+00
5.38272202e-01 -3.21224511e-01 2.88485050e-01 8.58821988e-01
-1.64580971e-01 7.70007193e-01 -2.87342697e-01 -9.09612656e-01
1.03585136e+00 -5.65260589e-01 -4.80001301e-01 -1.50359467e-01
6.43668771e-01 -2.87682056e-01 6.39469922e-01 1.18652277e-01
-1.15257144e+00 4.01263356e-01 -7.79814720e-01 7.74818480e-01
8.72204639e-03 -3.07453871e-01 7.63716400e-02 4.34879661e-01
-1.10782611e+00 5.58201790e-01 -6.64902866e-01 -4.35602814e-01
-5.20963338e-04 5.73108792e-01 -1.76048279e-01 3.35405409e-01
-9.40470099e-01 9.41090465e-01 5.27774453e-01 3.67650650e-02
-7.95241892e-01 -4.80359852e-01 -3.72483462e-01 2.91134894e-01
9.66101587e-01 -5.26398122e-01 1.25186777e+00 -1.21464896e+00
-1.81320870e+00 1.35311007e-01 1.78225517e-01 -4.87415016e-01
1.03877747e+00 4.38278258e-01 3.35095972e-01 5.85015155e-02
1.89340547e-01 1.55582845e-01 6.33973539e-01 -1.07118070e+00
-8.08962584e-01 -3.50275099e-01 5.39367497e-01 7.57644951e-01
-3.80800843e-01 -2.23294929e-01 2.55544603e-01 -4.58538048e-02
-3.31622124e-01 -1.18603146e+00 -8.81819904e-01 -1.69563115e-01
-1.02983579e-01 -3.13562214e-01 4.93839324e-01 -1.63635209e-01
7.27181196e-01 -1.92882907e+00 5.58339238e-01 2.45846242e-01
3.09443295e-01 -5.63567691e-02 -1.45391107e-01 7.22547352e-01
2.81740844e-01 7.80629218e-02 -9.28053930e-02 -1.70798823e-01
1.06759176e-01 9.24888626e-02 1.17347248e-01 6.90949619e-01
-2.52919227e-01 3.55431437e-01 -1.12918758e+00 -2.34547600e-01
5.16224578e-02 -1.72116876e-01 -6.36580527e-01 -1.02337651e-01
-2.12492600e-01 4.26760167e-01 -8.21397245e-01 -2.70296603e-01
5.85506082e-01 -2.31268868e-01 4.80476648e-01 4.73704576e-01
-5.56906581e-01 -2.55815089e-01 -1.43916678e+00 8.96027923e-01
-5.74513435e-01 3.29995662e-01 7.07928479e-01 -1.35408294e+00
6.09358370e-01 4.07425910e-01 8.89187276e-01 -3.84639800e-01
3.44086975e-01 4.15925592e-01 2.16376930e-01 1.12684384e-01
3.90593439e-01 -2.06510305e-01 -1.73018917e-01 6.38582230e-01
-2.14840129e-01 1.21551372e-01 6.29703879e-01 3.05955678e-01
1.06837881e+00 -6.71740949e-01 6.27905786e-01 -6.42379940e-01
6.35575473e-01 -1.57729775e-01 6.41416013e-01 8.57867062e-01
-1.50212362e-01 -2.47115374e-01 1.10749221e+00 -5.82180358e-02
-1.16321433e+00 -6.47887468e-01 3.68481636e-01 7.69690573e-01
4.90038663e-01 -1.09925531e-01 -7.81422019e-01 -3.97844255e-01
5.63946664e-02 3.90964180e-01 -6.36299551e-01 -2.10286342e-02
-3.31003904e-01 -8.91989410e-01 -1.38076916e-01 -2.59035558e-01
8.22930813e-01 -8.85078073e-01 -8.05200934e-01 4.64426875e-01
-2.18382888e-02 -9.00557935e-01 -4.12113011e-01 5.15751075e-03
-5.27129829e-01 -1.08493292e+00 -9.62380111e-01 -2.53443688e-01
8.31065714e-01 1.07587434e-01 3.02719206e-01 -2.22859979e-01
3.24383736e-01 4.93016899e-01 -1.60203293e-01 -3.43889207e-01
-6.61342621e-01 1.82958618e-01 4.01037157e-01 3.58528614e-01
-3.55753958e-01 -5.79360068e-01 -7.66027808e-01 2.44069785e-01
-8.41428220e-01 -7.20669478e-02 2.58752078e-01 6.16927326e-01
2.17039332e-01 2.85378814e-01 7.86405742e-01 -4.18851465e-01
1.19409454e+00 -5.61658680e-01 -1.41350174e+00 1.28149465e-01
-5.89236438e-01 3.23870599e-01 1.17242372e+00 -5.40115535e-01
-8.38706911e-01 7.34491050e-02 4.82629061e-01 1.74870446e-01
3.44532788e-01 4.79159147e-01 7.81254843e-02 -3.16325516e-01
4.19488400e-01 2.54226893e-01 4.47439402e-01 -1.83911577e-01
3.62033606e-01 6.35697365e-01 5.26351966e-02 -5.73508680e-01
6.60411954e-01 4.04533863e-01 5.01953721e-01 -9.41198647e-01
-1.70342430e-01 -7.97049254e-02 3.28559056e-02 -4.25061971e-01
5.54760277e-01 -5.70975959e-01 -1.60585475e+00 5.44596970e-01
-1.06232929e+00 -3.06750745e-01 -5.93495607e-01 5.47712982e-01
-7.61760890e-01 4.68637437e-01 -5.43908775e-01 -1.31648898e+00
-1.58365741e-01 -9.98251438e-01 1.13004744e-01 5.41575968e-01
4.18711782e-01 -9.57465351e-01 2.49782443e-01 -7.52267987e-02
6.63543582e-01 1.47799253e-01 5.76949239e-01 -6.31444037e-01
-4.63434279e-01 -2.67631441e-01 8.06517377e-02 1.42117485e-01
-4.68177758e-02 -2.28064299e-01 -4.18022983e-02 -5.35679102e-01
2.02635854e-01 -2.28854686e-01 2.47399479e-01 6.16962016e-01
2.13097051e-01 -8.03909481e-01 -1.12457149e-01 -8.02510381e-02
1.50058603e+00 3.14237118e-01 1.03910811e-01 3.29317480e-01
8.82873759e-02 5.94771266e-01 4.00783449e-01 9.87786829e-01
4.15858597e-01 6.01526797e-01 7.28845835e-01 3.03967386e-01
6.61908925e-01 7.73706511e-02 4.08959776e-01 5.39619386e-01
-4.33921933e-01 -2.58097529e-01 -5.17923594e-01 4.59071636e-01
-2.31189895e+00 -1.09281039e+00 1.71819478e-01 2.69288611e+00
6.74274087e-01 -8.83118734e-02 5.16236007e-01 -2.48964399e-01
8.62830341e-01 -7.54676387e-02 -7.92561352e-01 -5.73676169e-01
-6.35670200e-02 -3.81417781e-01 9.73514318e-01 7.99228489e-01
-6.71554029e-01 5.61513186e-01 5.34895468e+00 7.83047497e-01
-1.07201970e+00 2.97731012e-01 5.48396945e-01 -2.47312263e-01
8.91353041e-02 8.15271661e-02 -4.02877420e-01 6.89650714e-01
7.83992052e-01 -9.48342443e-01 8.82316232e-01 9.19995487e-01
7.26134300e-01 -5.52425623e-01 -5.03117621e-01 6.57751858e-01
-5.67524195e-01 -1.11927581e+00 -5.64243376e-01 6.48309171e-01
9.43405092e-01 1.46119058e-01 -1.27557456e-01 -2.34250203e-02
7.66021609e-01 -3.27272385e-01 5.45511425e-01 9.30576921e-02
2.18975946e-01 -1.05604970e+00 7.36552477e-01 7.87602365e-01
-1.22874713e+00 -5.36687136e-01 -3.60500783e-01 -4.27200943e-01
3.53341311e-01 2.64702469e-01 -5.11156559e-01 5.37526011e-01
1.77135333e-01 8.32626149e-02 2.28340641e-01 1.09124768e+00
1.77409232e-01 1.71843052e-01 -8.70311916e-01 -6.14694953e-01
5.53283513e-01 -7.56548107e-01 7.93974221e-01 3.20553631e-01
3.43267560e-01 2.60350972e-01 3.42404336e-01 4.97081131e-01
-6.48042262e-02 4.06538576e-01 -4.48045820e-01 -1.14108197e-01
3.21484357e-01 1.23896909e+00 -1.10568571e+00 -2.29107738e-01
-2.97439873e-01 7.46335387e-01 3.26308995e-01 4.25983191e-01
-4.85252649e-01 -3.55398059e-01 7.58654654e-01 -8.45112950e-02
1.83643833e-01 -2.18512908e-01 1.89451933e-01 -9.97121155e-01
1.34174272e-01 -5.16872466e-01 2.39020914e-01 -4.49626483e-02
-1.07163310e+00 4.50704515e-01 -2.83719581e-02 -1.12522674e+00
-6.50314689e-01 -1.26522556e-01 -6.55940533e-01 5.32582998e-01
-1.05473578e+00 -2.74344862e-01 2.40652844e-01 5.95665216e-01
1.57361269e-01 -2.99295694e-01 4.64573950e-01 1.38328016e-01
-6.25128388e-01 1.99689329e-01 8.59985471e-01 -3.06816131e-01
2.43743751e-02 -1.13802421e+00 -3.62550765e-01 7.55767703e-01
-4.99429256e-01 1.94212385e-02 1.10218322e+00 -3.46133113e-01
-1.34340620e+00 -6.51110113e-01 4.66519654e-01 3.05546224e-01
1.17551100e+00 -2.32300878e-01 -3.19061071e-01 2.15699241e-01
1.96570382e-01 8.67361724e-02 -6.58091158e-02 -1.49197906e-01
4.39449102e-01 -4.43667352e-01 -1.17627430e+00 8.20693731e-01
6.20215774e-01 7.45308697e-02 8.05885196e-02 4.27442491e-01
5.18626988e-01 -1.86970979e-02 -4.22375977e-01 -7.95304552e-02
2.41926625e-01 -7.24260628e-01 3.01475257e-01 -5.13389349e-01
-8.08861405e-02 -2.30464771e-01 1.26045659e-01 -1.71123564e+00
-1.97238117e-01 -1.08234739e+00 1.35744095e-01 7.43213773e-01
3.14850152e-01 -1.21923840e+00 8.12066436e-01 5.90686202e-01
4.13659453e-01 -5.94551086e-01 -1.47095418e+00 -9.26159918e-01
2.15177581e-01 2.32601181e-01 1.43480003e-01 6.14443958e-01
3.71988088e-01 2.87641406e-01 -4.79787171e-01 1.21041745e-01
8.66398752e-01 -1.17664322e-01 6.34734213e-01 -1.12828302e+00
-5.10879040e-01 -6.79203629e-01 -2.67754018e-01 -9.87928927e-01
9.05572549e-02 -3.57276320e-01 -4.51868810e-02 -1.35534358e+00
2.19846562e-01 -5.71036339e-01 -9.18659344e-02 1.13885194e-01
2.38480374e-01 -2.82852948e-01 5.98290205e-01 6.36544973e-02
-8.06186557e-01 7.84056783e-01 1.12821794e+00 -7.37835914e-02
-6.09498024e-01 5.04720569e-01 -6.67747259e-01 6.12438381e-01
8.45584750e-01 -5.26006401e-01 -6.14058018e-01 2.96682529e-02
4.82621610e-01 5.16437471e-01 1.06620505e-01 -6.18355215e-01
4.04444724e-01 -5.66300631e-01 -6.21332467e-01 -6.92839129e-03
1.26572609e-01 -7.31417120e-01 8.89229923e-02 8.84737670e-01
-3.12903434e-01 9.14383829e-02 -2.77384490e-01 8.15201759e-01
2.70742457e-02 -5.53249240e-01 8.31177354e-01 -2.11630926e-01
-4.39786687e-02 2.75134295e-01 -9.38487411e-01 1.51887104e-01
1.22167051e+00 1.53969377e-01 -2.32787073e-01 -1.04844630e+00
-5.64463079e-01 6.34905457e-01 4.09094691e-01 -8.10464099e-02
1.07873693e-01 -1.03622878e+00 -7.16050148e-01 -4.12847310e-01
-4.49339837e-01 -1.93513453e-01 3.51812929e-01 1.02836037e+00
-2.32894897e-01 2.33061090e-01 -2.60581493e-01 -1.56468719e-01
-9.24761295e-01 4.19683009e-01 6.54418230e-01 -5.76296985e-01
-2.72344321e-01 1.77276433e-01 2.69612789e-01 -1.47429690e-01
-2.56677922e-02 3.66984727e-03 -2.20708344e-02 7.40247071e-02
3.99883777e-01 5.95155895e-01 -4.12899911e-01 -3.26379269e-01
-1.38172403e-01 3.28100950e-01 -4.25757170e-02 -5.00374496e-01
1.29997373e+00 -2.74492919e-01 -2.88233727e-01 1.20696485e-01
8.03852677e-01 -1.40132144e-01 -1.22711086e+00 -1.65996209e-01
-1.14258058e-01 -5.78894205e-02 -8.32625031e-02 -2.60450095e-01
-1.06623530e+00 2.58077681e-01 3.88804227e-01 8.26149702e-01
9.20694113e-01 -9.60427523e-02 8.64135101e-02 6.52477622e-01
7.40180492e-01 -1.30407822e+00 -1.26910478e-01 6.39459491e-01
6.20336711e-01 -9.27808046e-01 -1.13359563e-01 4.84471321e-02
-5.75762212e-01 9.35689747e-01 3.73715401e-01 -3.42130691e-01
6.80761814e-01 1.24446064e-01 -4.08192843e-01 2.28871942e-01
-8.31638277e-01 -3.10124964e-01 -4.15035188e-01 4.12484556e-01
-2.23671496e-01 3.03554177e-01 -1.02618301e+00 3.55378866e-01
1.64439350e-01 -2.00993001e-01 1.01584280e+00 7.20503151e-01
-6.32554591e-01 -1.16704059e+00 -1.65783882e-01 4.33236122e-01
-2.60214090e-01 2.17844456e-01 -2.01910645e-01 7.86360621e-01
-3.61977369e-01 9.91089046e-01 5.11311479e-02 2.22756222e-01
1.22627512e-01 -4.14097309e-01 2.99535751e-01 -2.91159093e-01
-1.77618921e-01 3.61679830e-02 -1.44570753e-01 -4.04916227e-01
-3.83375913e-01 -6.59563839e-01 -1.19759023e+00 -6.06904626e-01
-3.82276416e-01 6.74394667e-01 5.92577040e-01 1.09429407e+00
3.30057174e-01 5.34595456e-03 1.07220066e+00 -7.42097199e-01
-1.15984559e+00 -7.58869171e-01 -8.27259719e-01 -1.06542530e-02
2.71960914e-01 -5.43622375e-01 -6.90091074e-01 -7.96057582e-01] | [4.364416122436523, 2.6993346214294434] |
a28f1fae-c900-4d3d-8786-5d67c358773f | personalized-federated-learning-with-local | 2304.01783 | null | https://arxiv.org/abs/2304.01783v2 | https://arxiv.org/pdf/2304.01783v2.pdf | Personalized Federated Learning with Local Attention | Federated Learning (FL) aims to learn a single global model that enables the central server to help the model training in local clients without accessing their local data. The key challenge of FL is the heterogeneity of local data in different clients, such as heterogeneous label distribution and feature shift, which could lead to significant performance degradation of the learned models. Although many studies have been proposed to address the heterogeneous label distribution problem, few studies attempt to explore the feature shift issue. To address this issue, we propose a simple yet effective algorithm, namely \textbf{p}ersonalized \textbf{Fed}erated learning with \textbf{L}ocal \textbf{A}ttention (pFedLA), by incorporating the attention mechanism into personalized models of clients while keeping the attention blocks client-specific. Specifically, two modules are proposed in pFedLA, i.e., the personalized single attention module and the personalized hybrid attention module. In addition, the proposed pFedLA method is quite flexible and general as it can be incorporated into any FL method to improve their performance without introducing additional communication costs. Extensive experiments demonstrate that the proposed pFedLA method can boost the performance of state-of-the-art FL methods on different tasks such as image classification and object detection tasks. | ['Yu Zhang', 'Shujun Yang', 'Junchao Tian', 'Sicong Liang'] | 2023-04-02 | null | null | null | null | ['personalized-federated-learning'] | ['methodology'] | [-5.37291057e-02 -2.31629401e-01 -4.43482012e-01 -6.15196884e-01
-8.25592756e-01 -2.31351614e-01 2.13317215e-01 -3.54612708e-01
-3.89324650e-02 6.21158719e-01 -1.34658203e-01 -1.61168188e-01
-1.84298828e-01 -5.46606302e-01 -6.62835717e-01 -1.15092838e+00
4.48830515e-01 5.67007363e-01 3.96113008e-01 2.74731189e-01
4.03848588e-02 6.32817447e-01 -1.56335533e+00 5.77084482e-01
9.52810884e-01 1.48125529e+00 4.22245353e-01 2.18293712e-01
-5.19958556e-01 1.20161533e+00 -3.53340209e-01 -4.75143552e-01
1.61628723e-01 -1.93833441e-01 -8.20766091e-01 2.06046522e-01
3.74588341e-01 -6.24995351e-01 -3.64796609e-01 8.38313401e-01
6.28853500e-01 7.05138966e-03 4.72914636e-01 -1.52404106e+00
-6.80397093e-01 6.23287797e-01 -6.65363252e-01 2.54200548e-01
-2.21000448e-01 1.78292647e-01 9.77173865e-01 -1.07444119e+00
2.30941787e-01 1.42484295e+00 4.62084770e-01 5.47194600e-01
-1.14597261e+00 -9.66446102e-01 6.55908287e-01 6.75838828e-01
-1.38207757e+00 -5.49847245e-01 5.98160446e-01 -1.05522804e-01
7.21430540e-01 3.15603673e-01 -1.08259603e-01 7.64751136e-01
1.02422647e-01 1.22591925e+00 1.25330782e+00 -3.36478353e-01
5.64102493e-02 3.51951838e-01 3.29089016e-01 6.66594803e-01
1.58295836e-02 -3.38926911e-01 -5.53075135e-01 -3.84757727e-01
4.39825773e-01 5.00441074e-01 -2.88226068e-01 -2.95204192e-01
-9.31312442e-01 7.05330968e-01 5.30888855e-01 1.59386292e-01
-3.78762156e-01 1.66509002e-01 4.30978686e-01 1.70845449e-01
5.21939278e-01 -2.92816103e-01 -6.31909728e-01 2.50234365e-01
-5.11128783e-01 -2.65762955e-02 6.71217442e-01 1.05424297e+00
1.12403882e+00 -2.23845318e-01 -4.57696080e-01 8.88995767e-01
4.30086762e-01 4.44941133e-01 7.11440086e-01 -9.32642043e-01
6.04891479e-01 7.32692719e-01 -1.03663296e-01 -8.10953259e-01
-1.59507036e-01 -3.82951438e-01 -6.85884774e-01 -6.58588707e-02
-4.37013134e-02 -2.89900243e-01 -5.35270631e-01 1.66030204e+00
5.59107959e-01 4.12665993e-01 -2.10534513e-01 7.35428512e-01
7.09465623e-01 5.76525986e-01 6.75212368e-02 -1.51523992e-01
1.04700911e+00 -1.64334071e+00 -5.41554570e-01 1.55675653e-02
6.31200254e-01 -7.39701092e-01 9.24776495e-01 2.42176965e-01
-8.38892758e-01 -3.25101674e-01 -5.44954717e-01 3.88144627e-02
-1.31506205e-01 2.63599791e-02 5.82581997e-01 5.29510379e-01
-1.04650295e+00 1.56637549e-01 -6.19531453e-01 -3.88191909e-01
7.40378201e-01 6.08672619e-01 -1.00373410e-01 -3.74759495e-01
-8.17323267e-01 4.48470771e-01 2.40769923e-01 -3.01954057e-02
-9.98227537e-01 -3.24184328e-01 -2.13922009e-01 4.56063986e-01
6.06406510e-01 -4.88555551e-01 1.62659764e+00 -1.03401780e+00
-1.52395225e+00 3.62258404e-01 -3.35036099e-01 -1.34828955e-01
4.27425027e-01 1.11582287e-01 -3.24238509e-01 -4.06153277e-02
-2.35666558e-02 4.38167572e-01 8.08278263e-01 -1.27969229e+00
-1.11160409e+00 -6.48730934e-01 -7.65748173e-02 2.38731995e-01
-7.55908191e-01 8.22806358e-02 -7.11118579e-01 -4.12864774e-01
1.02463306e-03 -6.74832106e-01 9.96175483e-02 8.58011246e-02
-1.78254396e-01 -6.88346207e-01 1.32719636e+00 -2.15664983e-01
1.22091115e+00 -2.17298603e+00 -3.95245254e-01 1.36455283e-01
2.54998535e-01 3.23517174e-01 -3.09471905e-01 4.42644209e-01
2.46068269e-01 -4.33605686e-02 3.38225394e-01 -5.30296326e-01
1.13602109e-01 4.08552110e-01 -2.80355066e-01 4.43851054e-01
-1.10559396e-01 8.10096264e-01 -6.46691024e-01 -7.89832532e-01
-4.48639281e-02 3.81813258e-01 -5.73913813e-01 3.88142884e-01
-4.15601939e-01 3.93910170e-01 -8.73988688e-01 7.78082013e-01
8.68974686e-01 -6.26619756e-01 2.38906398e-01 -2.15113476e-01
-6.27989471e-02 1.77130237e-01 -1.10712290e+00 1.20089662e+00
-5.60453713e-01 2.97587411e-03 4.21124667e-01 -9.74750340e-01
8.77780437e-01 4.21729892e-01 7.14186847e-01 -6.96513176e-01
2.99989551e-01 2.55981743e-01 -3.41568738e-01 -5.30927718e-01
4.53109369e-02 8.24718028e-02 3.68806571e-01 8.67747068e-01
8.74425024e-02 9.11879718e-01 -2.41008237e-01 -3.63291334e-03
1.12754524e+00 -1.39709458e-01 -3.14016730e-01 -8.47899914e-02
7.00311422e-01 -2.39934415e-01 7.45664358e-01 8.33513618e-01
-3.19173485e-01 1.92625225e-01 2.00231507e-01 -4.53038335e-01
-5.37879109e-01 -5.07816076e-01 -8.17482919e-02 1.94547975e+00
3.71449262e-01 -1.97734758e-01 -7.47622371e-01 -1.23383582e+00
1.31141067e-01 4.34295237e-01 -3.37831944e-01 -8.44548494e-02
-4.53702360e-01 -6.84235215e-01 2.86100000e-01 5.10858357e-01
8.78011882e-01 -1.19085515e+00 -3.64255577e-01 2.72196531e-01
-3.01434338e-01 -8.24292362e-01 -7.93150365e-01 2.60673374e-01
-6.79461360e-01 -9.63521898e-01 -4.64909434e-01 -8.91393244e-01
5.31666100e-01 7.76525140e-01 8.19039762e-01 6.85957968e-02
5.16974963e-02 2.64867932e-01 -4.67372268e-01 -3.02289337e-01
9.26638916e-02 4.28896219e-01 -2.65089989e-01 7.06799448e-01
4.60879505e-01 -4.65010762e-01 -7.98147559e-01 8.22050869e-01
-9.13386524e-01 -1.19631328e-01 8.31118762e-01 1.01310551e+00
6.32505298e-01 8.36753845e-02 8.27521443e-01 -1.18954730e+00
4.94068831e-01 -8.92866373e-01 -2.98482835e-01 6.53444827e-01
-1.06242526e+00 -5.45745641e-02 8.97714615e-01 -3.89914691e-01
-1.35903668e+00 -7.22097680e-02 -7.21706301e-02 -6.49328053e-01
-1.50868222e-01 2.49676853e-01 -4.56584454e-01 -2.99774468e-01
2.70762742e-01 3.33249718e-01 -7.28866011e-02 -1.02672935e+00
2.57206976e-01 1.31774211e+00 2.89439499e-01 -7.06299365e-01
1.58936590e-01 4.43725169e-01 -4.05021369e-01 -8.88172686e-02
-8.30952525e-01 -5.25985956e-01 -1.81990936e-01 -1.71219483e-01
3.93595517e-01 -7.12545931e-01 -1.11263680e+00 6.41853452e-01
-9.73499894e-01 -2.81913549e-01 -3.78426053e-02 1.50774747e-01
-4.84604925e-01 3.16789687e-01 -6.08487129e-01 -6.37236297e-01
-6.88986659e-01 -1.44729948e+00 9.81042027e-01 3.82712096e-01
3.62882495e-01 -6.22533977e-01 -1.87159464e-01 6.11661494e-01
7.26886868e-01 -4.82466936e-01 1.17227387e+00 -9.60031271e-01
-8.48532140e-01 -1.63161427e-01 -6.78556859e-01 4.23893809e-01
2.70037204e-01 -4.19503331e-01 -1.17002165e+00 -5.52762806e-01
5.19183688e-02 -6.15879953e-01 6.75759256e-01 -9.96548384e-02
1.59541750e+00 -7.57685483e-01 -4.77229118e-01 5.22230685e-01
1.37063932e+00 3.36081743e-01 1.74293369e-01 1.57282665e-01
7.64651835e-01 2.35497281e-01 3.52905422e-01 7.04886615e-01
6.08515024e-01 7.72987306e-01 7.14700699e-01 1.04028262e-01
-2.15159193e-01 -2.33185235e-02 3.23540866e-01 8.02546442e-01
3.74963880e-01 -5.04059017e-01 -6.74449623e-01 3.81414741e-01
-2.22528911e+00 -7.04324186e-01 1.84641719e-01 1.94586432e+00
5.86847186e-01 -3.73910636e-01 -6.08976036e-02 -2.83619165e-01
9.37856019e-01 5.08987345e-02 -1.12469184e+00 -3.36563110e-01
1.41600162e-01 -7.51026794e-02 4.93691325e-01 -1.38102770e-02
-8.85606408e-01 8.02710176e-01 5.49553633e+00 1.14185870e+00
-1.44859385e+00 6.23593509e-01 7.13762760e-01 -1.63977265e-01
-2.26894706e-01 -1.94284111e-01 -1.10431075e+00 7.69135356e-01
7.73465395e-01 -1.34751171e-01 5.86375654e-01 1.19880593e+00
4.87445621e-03 3.07524770e-01 -9.23622072e-01 9.15876389e-01
6.75276443e-02 -1.11754310e+00 2.12610632e-01 7.82422200e-02
6.93389297e-01 4.79038775e-01 7.68027678e-02 5.10687709e-01
3.84553939e-01 -6.31573737e-01 4.53058958e-01 5.93733668e-01
7.41676211e-01 -8.63904119e-01 7.49615848e-01 6.73821926e-01
-1.09386790e+00 -5.84121406e-01 -3.60373378e-01 4.75263238e-01
-1.88640252e-01 2.08041340e-01 -6.44004822e-01 5.79403281e-01
1.03362048e+00 4.15856779e-01 -5.74014544e-01 1.10848045e+00
3.21896225e-01 6.83255255e-01 -3.26230019e-01 9.56817567e-02
3.31202537e-01 7.83594847e-02 8.95584226e-02 8.80440176e-01
2.24049300e-01 -4.92258705e-02 4.94881690e-01 4.92869586e-01
-4.48413849e-01 5.50385892e-01 -1.38454139e-01 3.91052037e-01
6.12152457e-01 1.28301823e+00 -2.54540890e-01 -2.72244900e-01
-7.94436038e-01 8.47987711e-01 7.75477648e-01 3.69101703e-01
-7.80275524e-01 -2.01530352e-01 4.30219710e-01 1.09493628e-01
6.03970945e-01 3.71205688e-01 2.88457982e-02 -1.10354829e+00
-6.13670005e-03 -1.03335392e+00 6.62672758e-01 -6.69823647e-01
-1.74615610e+00 8.11008096e-01 -3.40672255e-01 -7.79839575e-01
7.13423714e-02 -2.17009202e-01 -5.55647075e-01 8.46937537e-01
-1.71416831e+00 -1.28280532e+00 -5.34687579e-01 1.15470290e+00
6.16029441e-01 -2.66369283e-01 6.87533617e-01 6.19472682e-01
-1.03411698e+00 9.81470883e-01 6.71832621e-01 -2.37023890e-01
9.48062778e-01 -7.56401062e-01 -2.80153394e-01 4.12354827e-01
-1.63204461e-01 4.56621885e-01 -2.74881180e-02 -4.07781273e-01
-1.46569061e+00 -1.59980118e+00 7.61154830e-01 7.49845430e-02
3.44476730e-01 -1.38190746e-01 -1.00291288e+00 8.80484760e-01
2.21405551e-01 4.37908649e-01 6.68044209e-01 -2.61333119e-02
-4.06836003e-01 -6.41862988e-01 -1.32615483e+00 3.36304128e-01
8.84121060e-01 -2.94536144e-01 4.34810854e-02 5.86209953e-01
6.55509531e-01 -8.10759515e-02 -7.24866867e-01 3.68363798e-01
4.49783772e-01 -1.10691702e+00 4.71545666e-01 -5.54932773e-01
-1.79159090e-01 -8.65141675e-02 -2.75911152e-01 -6.68603480e-01
-5.85545897e-01 -4.85454679e-01 -4.47344273e-01 1.59592259e+00
1.00593917e-01 -1.11063254e+00 9.00798202e-01 7.90349543e-01
-1.62808672e-01 -1.10127151e+00 -1.05295277e+00 -6.70371473e-01
-1.06890209e-01 -1.14892639e-01 1.06687284e+00 6.84850097e-01
-2.72339493e-01 1.79468453e-01 -4.10665542e-01 1.12187140e-01
5.00943840e-01 4.92717922e-01 7.66500235e-01 -1.13553560e+00
-3.56072724e-01 -3.36944163e-01 2.71439310e-02 -1.44713533e+00
2.92216003e-01 -1.05262816e+00 3.43036354e-02 -1.34348452e+00
5.46348274e-01 -9.87885356e-01 -7.96230316e-01 9.55533266e-01
-3.41368169e-02 3.67319062e-02 3.47303420e-01 8.21580648e-01
-1.11190093e+00 7.83076346e-01 1.02355540e+00 -8.98566917e-02
1.08123653e-01 2.48380661e-01 -7.80494988e-01 4.74660307e-01
6.59883380e-01 -6.31479859e-01 -4.17694509e-01 -6.52369678e-01
-4.26102638e-01 -1.18681669e-01 1.48165807e-01 -7.60438323e-01
4.94645536e-01 -2.11265355e-01 1.42101690e-01 -5.85378885e-01
2.19556596e-02 -1.21747267e+00 3.04646790e-02 6.01530559e-02
-1.86026633e-01 6.86590895e-02 -1.57210127e-01 6.34118438e-01
-1.35115519e-01 -2.11577609e-01 7.47548819e-01 -1.13716558e-01
-6.31233454e-01 7.18554616e-01 -2.59752542e-01 -3.11777145e-01
1.17633224e+00 1.68019563e-01 -6.58917606e-01 -3.20525944e-01
-3.36842418e-01 6.07946277e-01 3.62148762e-01 3.46663326e-01
2.14686394e-01 -1.34397483e+00 -4.93782252e-01 2.87124068e-01
1.05390929e-01 6.65789917e-02 5.78975260e-01 1.00957596e+00
-2.36242652e-01 6.89823568e-01 6.76345900e-02 -5.47852457e-01
-1.10808170e+00 7.85691679e-01 4.21016812e-01 -3.42175871e-01
-4.22747672e-01 7.25266337e-01 4.49397713e-01 -5.77334821e-01
5.79600394e-01 4.82226983e-02 5.89242391e-02 -1.01895235e-01
5.88989794e-01 6.43203199e-01 2.55461186e-01 -7.16897309e-01
-3.76761168e-01 2.97968894e-01 -5.71988940e-01 5.30786693e-01
1.22343314e+00 -5.74595094e-01 -2.01430142e-01 1.50683612e-01
1.26454139e+00 -3.46546859e-01 -1.38279331e+00 -7.82055438e-01
-2.15308040e-01 -5.67486823e-01 2.30247676e-01 -9.67245519e-01
-1.62093699e+00 7.48655736e-01 8.38225186e-01 1.44231200e-01
1.34346211e+00 5.11809066e-02 1.03745258e+00 3.49986881e-01
7.07851768e-01 -7.68189311e-01 4.00784053e-02 2.87948430e-01
6.21244133e-01 -1.08803153e+00 -3.38086665e-01 -2.47124761e-01
-5.28109133e-01 8.85689199e-01 9.66745496e-01 2.68003702e-01
7.23363221e-01 -2.71659531e-03 1.42805427e-01 -4.85169850e-02
-1.04570568e+00 5.15106926e-03 -1.16424216e-02 9.81699228e-02
6.61066100e-02 -2.31990829e-01 -2.32180357e-01 7.73152947e-01
6.15179658e-01 3.23760003e-01 -1.10177658e-01 1.16001403e+00
-3.46666902e-01 -1.38040888e+00 -4.12517190e-01 5.87891698e-01
-5.59934795e-01 1.51457220e-01 -6.70951679e-02 2.73926765e-01
5.14039695e-01 1.09530747e+00 -1.10408343e-01 -3.73514920e-01
2.10562125e-01 3.42846274e-01 7.02843145e-02 -5.29818118e-01
-7.38065422e-01 2.91875988e-01 -3.16082805e-01 -6.70465231e-01
-2.54225075e-01 -5.52128196e-01 -1.21925044e+00 -4.48519409e-01
-7.49563873e-01 1.83396310e-01 5.06132305e-01 9.05765951e-01
9.13579702e-01 2.47432783e-01 1.15856707e+00 -7.20811129e-01
-7.35486627e-01 -9.25023615e-01 -7.58499444e-01 2.04169735e-01
1.44635737e-01 -6.10155165e-01 -2.25013077e-01 -2.04308674e-01] | [5.812575817108154, 6.256898880004883] |
fae00122-b9ac-4990-9874-a1103ad10fb0 | web-table-classification-based-on-visual | 2103.05110 | null | https://arxiv.org/abs/2103.05110v1 | https://arxiv.org/pdf/2103.05110v1.pdf | Web Table Classification based on Visual Features | Tables on the web constitute a valuable data source for many applications, like factual search and knowledge base augmentation. However, as genuine tables containing relational knowledge only account for a small proportion of tables on the web, reliable genuine web table classification is a crucial first step of table extraction. Previous works usually rely on explicit feature construction from the HTML code. In contrast, we propose an approach for web table classification by exploiting the full visual appearance of a table, which works purely by applying a convolutional neural network on the rendered image of the web table. Since these visual features can be extracted automatically, our approach circumvents the need for explicit feature construction. A new hand labeled gold standard dataset containing HTML source code and images for 13,112 tables was generated for this task. Transfer learning techniques are applied to well known VGG16 and ResNet50 architectures. The evaluation of CNN image classification with fine tuned ResNet50 (F1 93.29%) shows that this approach achieves results comparable to previous solutions using explicitly defined HTML code based features. By combining visual and explicit features, an F-measure of 93.70% can be achieved by Random Forest classification, which beats current state of the art methods. | ['Heiko Paulheim', 'Babette Bühler'] | 2021-02-25 | null | null | null | null | ['table-extraction'] | ['miscellaneous'] | [-3.09151150e-02 5.29158831e-01 -4.07536030e-01 -3.05179894e-01
-9.61073637e-01 -8.21482658e-01 8.68309081e-01 6.88485086e-01
-3.17805231e-01 7.44651079e-01 8.29935670e-02 -2.59218603e-01
2.76392866e-02 -1.27784348e+00 -1.14302409e+00 -1.57727256e-01
-1.46853039e-02 6.52885795e-01 5.00361919e-01 -4.24602151e-01
7.97402263e-02 3.91975075e-01 -1.78748548e+00 8.89318526e-01
6.01430655e-01 1.72901225e+00 -1.34861082e-01 5.92525780e-01
-7.17699111e-01 1.15717566e+00 -7.34932184e-01 -9.09128308e-01
1.92910373e-01 -2.10066989e-01 -9.99632537e-01 1.32642698e-03
8.88022184e-01 -2.03779399e-01 -3.09101254e-01 9.76876497e-01
4.71053869e-02 -3.80135775e-01 5.47359645e-01 -1.04332221e+00
-7.06740618e-01 8.88568580e-01 -2.65957594e-01 5.59489019e-02
5.62294424e-01 -1.21649951e-01 1.36196172e+00 -7.25955188e-01
1.22992098e+00 9.08200502e-01 7.46998727e-01 2.28025336e-02
-1.34553874e+00 -4.74902183e-01 -3.36141102e-02 4.60632771e-01
-1.38032544e+00 -1.83659136e-01 7.41582990e-01 -5.73512912e-01
9.01634574e-01 3.32601428e-01 6.37409151e-01 1.03674793e+00
2.86031049e-02 7.76485860e-01 1.04918492e+00 -4.46033061e-01
1.13230571e-01 6.82783544e-01 4.37425114e-02 9.91988301e-01
6.52094603e-01 -4.63994354e-01 -6.30768538e-01 1.74639583e-01
5.66335499e-01 -2.38701940e-01 -6.37916774e-02 -9.77303743e-01
-1.24556851e+00 7.70875514e-01 1.01800799e+00 3.42700481e-01
-3.63475949e-01 6.95583150e-02 5.37489593e-01 3.59526753e-01
2.87926227e-01 4.15436447e-01 -6.34655774e-01 -8.90428375e-04
-8.77786815e-01 2.68938154e-01 1.11819255e+00 1.06069505e+00
8.19235325e-01 -1.23679474e-01 -2.06186771e-01 6.74040437e-01
7.11506903e-02 5.03456473e-01 1.55350506e-01 -5.64192057e-01
1.00365102e+00 1.37381971e+00 -1.30960718e-01 -1.11731875e+00
-1.49657845e-01 -3.17317039e-01 -8.06636989e-01 3.82071674e-01
8.74520600e-01 4.67800975e-01 -1.20948493e+00 9.96950030e-01
3.19766283e-01 -9.33980465e-01 -7.33378232e-02 4.22833472e-01
1.24593782e+00 2.12349221e-01 -2.01187924e-01 4.08230424e-01
1.63668573e+00 -5.43285429e-01 -5.89194775e-01 -1.29635688e-02
3.84690285e-01 -5.43845177e-01 1.17875397e+00 5.13743520e-01
-6.66227877e-01 -4.68712091e-01 -1.36472487e+00 -3.06149453e-01
-1.28993511e+00 2.56298304e-01 6.63779318e-01 7.52755761e-01
-9.06442046e-01 5.70415378e-01 -4.09572035e-01 -4.32970315e-01
9.26492751e-01 4.14859563e-01 -8.28238487e-01 -1.36220213e-02
-1.01951659e+00 8.03041279e-01 5.61869860e-01 -4.71494831e-02
-4.98680234e-01 -4.98485267e-01 -9.21372354e-01 3.20544720e-01
6.78019106e-01 -1.95309773e-01 1.00130868e+00 -8.60818148e-01
-9.31046426e-01 1.13381755e+00 2.54001707e-01 -7.63522625e-01
9.39058781e-01 -6.77814186e-02 -2.27879584e-01 2.77708769e-01
3.09378468e-03 4.69775826e-01 5.95729172e-01 -1.54480553e+00
-4.15298581e-01 -2.15693265e-01 2.86613017e-01 -2.84149349e-01
-3.27700794e-01 -4.07570004e-01 -6.28912449e-01 -4.67110038e-01
3.37572880e-02 -5.88721216e-01 1.51294798e-01 1.33682564e-01
-7.52427936e-01 5.73892817e-02 4.87432122e-01 -1.03844440e+00
1.03170860e+00 -1.73918736e+00 -2.92681932e-01 5.55256665e-01
5.67354620e-01 2.03768581e-01 1.99480057e-01 3.76872033e-01
-3.08568384e-02 2.51487553e-01 -1.97765023e-01 3.17267418e-01
2.77120948e-01 -1.74011767e-01 -2.61619866e-01 8.08382854e-02
1.29200742e-01 1.22867429e+00 -5.06170452e-01 -7.75219619e-01
2.25271463e-01 5.16254008e-01 -4.30050671e-01 -3.37463617e-02
-4.93813246e-01 -2.54911005e-01 -2.78460801e-01 9.49276268e-01
5.00478268e-01 -5.64424872e-01 5.45285463e-01 -3.08651179e-01
3.44645947e-01 5.11476934e-01 -1.08766830e+00 1.39079273e+00
-4.92632210e-01 7.78748333e-01 -1.41697913e-01 -8.04026723e-01
9.57176268e-01 2.96300594e-02 2.61871815e-01 -8.89609396e-01
1.66436374e-01 2.16588393e-01 -1.34453580e-01 -2.25567639e-01
3.73136759e-01 4.84522104e-01 -1.69885978e-01 -4.15046811e-02
3.73799145e-01 -8.30594525e-02 4.45133299e-01 4.47041690e-01
1.29238188e+00 2.53807485e-01 4.66788799e-01 -2.05939010e-01
5.91541350e-01 3.71781915e-01 -1.16304018e-01 7.72991836e-01
1.16369113e-01 6.97716594e-01 9.33931112e-01 -9.05767858e-01
-1.16855943e+00 -1.03292561e+00 -1.23233445e-01 7.73357928e-01
-1.44490585e-01 -7.26384819e-01 -9.07102466e-01 -1.22855067e+00
3.42800349e-01 2.92777181e-01 -1.08878517e+00 2.01285407e-01
-3.93953979e-01 -5.29628217e-01 3.26804429e-01 5.69801152e-01
8.39046657e-01 -1.23042166e+00 -3.43190908e-01 1.63748056e-01
-2.72706032e-01 -1.28641176e+00 2.70672768e-01 4.79930282e-01
-6.29456460e-01 -1.48102200e+00 -3.13271433e-01 -4.19438839e-01
5.24153352e-01 -2.24949166e-01 1.66558540e+00 1.65298775e-01
-5.93946576e-01 1.14276238e-01 -1.80130750e-01 -2.18370393e-01
-4.02749866e-01 6.50606811e-01 -5.24775982e-01 -9.10677016e-02
4.58217800e-01 -4.09562618e-01 -3.44165683e-01 -5.90400212e-02
-7.44567633e-01 1.15810625e-01 9.22686279e-01 6.68296397e-01
5.77586472e-01 -1.18213385e-01 9.81263071e-02 -1.33395433e+00
2.37039328e-01 -1.96668729e-01 -8.28791440e-01 3.61638606e-01
-6.88790560e-01 4.40753728e-01 6.62579417e-01 -1.20443150e-01
-8.62635314e-01 2.74500757e-01 -1.49611831e-01 -1.38021156e-01
-2.02253118e-01 4.10585523e-01 -4.94376570e-01 -5.74981570e-02
8.83055389e-01 1.77862227e-01 -1.58994824e-01 -5.70916653e-01
4.22063977e-01 5.38679659e-01 4.27549124e-01 -3.22121918e-01
1.12941146e+00 5.66861808e-01 5.36963008e-02 -6.38358653e-01
-8.38621914e-01 -1.42826736e-01 -9.20763135e-01 -1.71226829e-01
8.50728393e-01 -7.73992538e-01 -1.06143761e+00 1.51566658e-02
-7.80697167e-01 -3.08034003e-01 -2.22292215e-01 -4.73757200e-02
-3.86263132e-01 -3.60307954e-02 -4.23303545e-01 -5.08071780e-01
-1.99306414e-01 -8.07117879e-01 1.06227112e+00 -2.71745920e-01
-1.88399255e-01 -8.14398766e-01 -3.60110372e-01 6.23360872e-01
4.37530398e-01 7.06408322e-01 1.03324270e+00 -9.53496814e-01
-1.14279854e+00 -6.47991002e-01 -6.63337350e-01 8.44913647e-02
-1.40890274e-02 3.41175012e-02 -1.17321408e+00 2.30390757e-01
-5.99280000e-01 -4.82550174e-01 1.18731034e+00 -1.97973266e-01
1.23030829e+00 -4.69589651e-01 -3.89545262e-01 4.99154836e-01
1.81403017e+00 1.10742256e-01 7.61712790e-01 9.02700126e-01
1.15896797e+00 5.20664930e-01 3.77873063e-01 2.23238200e-01
4.83280033e-01 6.13933206e-01 7.65178323e-01 -1.83441620e-02
-4.03723210e-01 -5.11597276e-01 -2.69625708e-02 2.73341686e-01
-1.33178011e-01 -7.03934506e-02 -1.21884561e+00 4.28460270e-01
-1.48744631e+00 -8.44585478e-01 -2.76915252e-01 2.08444476e+00
8.64580452e-01 7.68674970e-01 -1.91089697e-02 5.79764605e-01
5.15586674e-01 1.33375555e-01 -3.02731037e-01 -1.00817271e-01
-1.56600296e-01 3.04316819e-01 8.29142988e-01 2.41457626e-01
-1.38587105e+00 7.08174407e-01 5.92113161e+00 8.43517721e-01
-9.25332546e-01 -2.11275846e-01 7.11697280e-01 1.09178752e-01
-2.21450895e-01 -2.50822932e-01 -9.92671371e-01 1.79160684e-01
8.95833373e-01 1.04038998e-01 4.11960453e-01 1.18530548e+00
-5.89258671e-01 -1.77152812e-01 -1.07820809e+00 9.51449633e-01
1.21781863e-01 -1.72443378e+00 3.50559443e-01 2.91032881e-01
4.66097772e-01 -1.67170033e-01 -8.53266641e-02 3.60823065e-01
4.81467962e-01 -1.16322637e+00 9.49434936e-01 3.78690869e-01
1.09001589e+00 -7.74040103e-01 8.68504167e-01 -2.03066230e-01
-1.26153743e+00 1.44463241e-01 -9.88658220e-02 3.28581542e-01
-6.95600390e-01 7.09498525e-01 -1.06207240e+00 6.29208744e-01
1.06086242e+00 4.93343472e-01 -1.39757240e+00 6.34199202e-01
-2.75542647e-01 4.09972131e-01 -3.04646939e-01 -2.29721323e-01
4.21433523e-03 2.92399377e-01 8.15702528e-02 1.24017799e+00
-1.10223494e-01 -7.43656278e-01 -1.27251506e-01 8.06329548e-01
-6.68150187e-01 3.16581160e-01 -8.45662773e-01 -1.62534669e-01
5.15359379e-02 1.50851023e+00 -1.33347070e+00 -5.32317460e-01
-4.12754416e-01 7.44187951e-01 5.79413712e-01 1.15818985e-01
-6.94458604e-01 -9.23809111e-01 3.18763971e-01 4.24321830e-01
8.67349684e-01 1.16340779e-01 -5.50188959e-01 -1.19998670e+00
5.04698634e-01 -9.21086371e-01 3.61602813e-01 -7.15076864e-01
-8.38000059e-01 9.76241589e-01 -2.13048667e-01 -1.18702304e+00
-2.88208455e-01 -9.83253419e-01 7.31807500e-02 5.73417068e-01
-1.36118400e+00 -1.37582386e+00 -7.82152653e-01 4.59530473e-01
2.47938305e-01 -2.92049617e-01 7.79988647e-01 1.97677270e-01
-6.37870729e-02 6.19094729e-01 8.52223337e-02 7.85360396e-01
4.88897443e-01 -1.86121094e+00 5.64266503e-01 6.99111938e-01
4.95816499e-01 4.31990504e-01 5.87778032e-01 -6.89433813e-01
-1.48771834e+00 -1.05208528e+00 9.60635245e-01 -8.55464697e-01
7.65644789e-01 -1.07623827e+00 -1.01337540e+00 6.64678633e-01
2.31201082e-01 1.77291840e-01 3.47242147e-01 1.32599697e-01
-9.46058452e-01 -5.38624704e-01 -1.18152893e+00 4.16439772e-01
8.99806261e-01 -6.03538454e-01 -4.50048804e-01 3.35970014e-01
4.21400666e-01 -4.70167488e-01 -9.60305810e-01 2.32543126e-01
7.95359969e-01 -1.08394635e+00 1.15683067e+00 -3.49433452e-01
9.50403988e-01 -1.93249404e-01 -2.27275744e-01 -8.38904202e-01
1.01543441e-01 -4.93259132e-02 -3.32251251e-01 1.37745047e+00
7.09304154e-01 -2.47974455e-01 1.22277963e+00 2.90191650e-01
4.24502611e-01 -5.51267743e-01 -6.70185030e-01 -7.44540036e-01
-1.16088629e-01 -3.22723418e-01 6.14254177e-01 8.95172834e-01
-1.36717960e-01 3.52116734e-01 6.64959699e-02 -2.80323416e-01
6.20965064e-01 2.60933161e-01 9.32512403e-01 -1.59474301e+00
-1.22171929e-02 -4.55672652e-01 -6.44401610e-01 -3.08418661e-01
-5.59320971e-02 -1.16982639e+00 -1.75862357e-01 -1.97128868e+00
1.87015742e-01 -2.61809796e-01 -1.38791710e-01 6.40900970e-01
2.71534808e-02 5.89323282e-01 1.43809617e-01 -1.35584606e-03
-7.02612400e-01 1.41128466e-01 1.00224495e+00 -5.38450003e-01
2.58058429e-01 -3.33456695e-01 -6.73919499e-01 6.07348144e-01
6.73226237e-01 -4.48536932e-01 -1.00016557e-01 8.52352157e-02
6.02895319e-01 -1.48706004e-01 3.68359089e-01 -1.25482512e+00
-1.07061286e-02 1.93714231e-01 1.02642465e+00 -7.20279813e-01
2.96656668e-01 -1.07454455e+00 -3.38000543e-02 4.39802051e-01
-3.57976347e-01 -1.28427848e-01 2.88771421e-01 5.78950703e-01
-4.84904319e-01 1.09658875e-02 4.19414520e-01 -5.11240423e-01
-6.92959249e-01 3.14701162e-02 -3.00886989e-01 8.93549547e-02
6.83709979e-01 -2.33019799e-01 -4.08480525e-01 -3.37519377e-01
-6.64061785e-01 -1.31722733e-01 4.00638103e-01 4.82911468e-01
4.61936444e-01 -1.31203222e+00 -3.72549266e-01 7.83161297e-02
5.06124914e-01 -9.80577692e-02 -1.88785076e-01 3.16931218e-01
-9.80677307e-01 6.42106891e-01 -3.75492007e-01 -4.96231526e-01
-1.29525626e+00 7.98093259e-01 1.01421818e-01 -5.87610781e-01
-6.37376070e-01 6.30309224e-01 -1.28642455e-01 -5.01527965e-01
2.83886731e-01 -5.09432614e-01 -4.20007080e-01 5.09365261e-01
3.12952727e-01 2.50819847e-02 7.94676721e-01 -3.99276823e-01
-4.98106241e-01 2.82115102e-01 -2.39108488e-01 -1.62976626e-02
1.48767483e+00 2.69675255e-01 -5.30174421e-03 5.68733633e-01
1.17936516e+00 2.71941900e-01 -9.27578866e-01 -1.65071487e-01
4.70350057e-01 -4.73043084e-01 -2.24618524e-01 -1.21718252e+00
-1.20585573e+00 8.48026872e-01 3.95176232e-01 6.85673654e-01
7.75843441e-01 4.14929576e-02 2.77771443e-01 8.46812248e-01
3.84226829e-01 -8.79404724e-01 5.78798093e-02 3.52103055e-01
1.00719357e+00 -1.61466181e+00 1.06820300e-01 -7.15226352e-01
-5.00314713e-01 1.26577151e+00 6.06958985e-01 -1.51729554e-01
5.35612881e-01 3.29781830e-01 1.68171018e-01 -4.38432008e-01
-6.86423898e-01 -4.51866239e-01 5.80818892e-01 6.18880451e-01
5.44782817e-01 -1.28656581e-01 4.75716479e-02 3.99154037e-01
-6.00119352e-01 -9.10411924e-02 5.34594715e-01 1.00362146e+00
-2.91152090e-01 -9.49225664e-01 -3.65267545e-01 8.43990147e-01
-7.28192151e-01 -2.47086972e-01 -6.59401298e-01 1.27710140e+00
2.10511461e-02 4.68941480e-01 7.35522360e-02 -4.14402783e-01
4.11634535e-01 3.95382553e-01 4.57062870e-01 -4.37001765e-01
-8.61109853e-01 -4.08590406e-01 3.98492128e-01 -6.54793620e-01
-1.84908181e-01 -4.23609525e-01 -7.97763586e-01 -3.75762254e-01
5.59561700e-02 -3.82203758e-02 8.68639231e-01 6.55552268e-01
1.02879100e-01 7.85783350e-01 5.29557392e-02 -5.13059020e-01
-1.07070632e-01 -8.95542026e-01 -4.36532825e-01 6.42832756e-01
3.15125346e-01 -7.49583483e-01 -1.53080404e-01 3.69037658e-01] | [11.683211326599121, 2.9726779460906982] |
7e808ad6-9290-42ea-a902-9a4493333902 | exploiting-multiple-sources-for-open-domain | null | null | https://aclanthology.org/D13-1122 | https://aclanthology.org/D13-1122.pdf | Exploiting Multiple Sources for Open-Domain Hypernym Discovery | null | ['Ting Liu', 'Bing Qin', 'Ruiji Fu'] | 2013-10-01 | null | null | null | emnlp-2013-10 | ['hypernym-discovery'] | ['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] | [-7.428530216217041, 3.6492881774902344] |
ad869481-9164-498f-9143-85d9209fe1a4 | subgraph-aware-few-shot-inductive-link | 2108.00954 | null | https://arxiv.org/abs/2108.00954v1 | https://arxiv.org/pdf/2108.00954v1.pdf | Subgraph-aware Few-Shot Inductive Link Prediction via Meta-Learning | Link prediction for knowledge graphs aims to predict missing connections between entities. Prevailing methods are limited to a transductive setting and hard to process unseen entities. The recent proposed subgraph-based models provided alternatives to predict links from the subgraph structure surrounding a candidate triplet. However, these methods require abundant known facts of training triplets and perform poorly on relationships that only have a few triplets. In this paper, we propose Meta-iKG, a novel subgraph-based meta-learner for few-shot inductive relation reasoning. Meta-iKG utilizes local subgraphs to transfer subgraph-specific information and learn transferable patterns faster via meta gradients. In this way, we find the model can quickly adapt to few-shot relationships using only a handful of known facts with inductive settings. Moreover, we introduce a large-shot relation update procedure to traditional meta-learning to ensure that our model can generalize well both on few-shot and large-shot relations. We evaluate Meta-iKG on inductive benchmarks sampled from NELL and Freebase, and the results show that Meta-iKG outperforms the current state-of-the-art methods both in few-shot scenarios and standard inductive settings. | ['Yuedong Yang', 'Haifeng Hu', 'Ya Sun', 'Sijie Mai', 'Shuangjia Zheng'] | 2021-07-26 | null | null | null | null | ['inductive-link-prediction'] | ['graphs'] | [-1.54555216e-01 6.95538640e-01 -8.95572543e-01 -2.67231643e-01
-5.99727333e-01 -2.55718797e-01 6.81420386e-01 5.03412187e-01
-3.34164128e-02 1.23488045e+00 1.65011376e-01 -5.75656109e-02
-3.27270001e-01 -1.37245417e+00 -1.16433167e+00 -9.92548615e-02
-4.47225481e-01 8.59982014e-01 4.88799334e-01 -5.56808889e-01
-2.68783689e-01 -1.04201347e-01 -1.38206542e+00 4.05681938e-01
1.18170106e+00 5.97422957e-01 -2.39294484e-01 4.75463808e-01
-3.77576560e-01 1.13187861e+00 9.47424099e-02 -1.03915322e+00
3.10559571e-02 -5.78352809e-01 -1.23928165e+00 -3.96995634e-01
5.04371345e-01 -7.59426430e-02 -6.77791476e-01 8.72696102e-01
3.26612413e-01 3.94118816e-01 6.80536926e-01 -1.32324409e+00
-9.92237389e-01 1.08404434e+00 -3.01576376e-01 5.89177370e-01
6.96096480e-01 -1.75072357e-01 1.59715104e+00 -1.08433735e+00
1.20831811e+00 1.19434011e+00 7.35556245e-01 4.55449432e-01
-1.23533106e+00 -5.61859369e-01 5.38910568e-01 7.03529775e-01
-1.23941398e+00 -1.95260912e-01 5.44640839e-01 -2.15391129e-01
1.39782071e+00 3.04021291e-03 8.59114766e-01 1.14440191e+00
2.70119030e-02 6.80071175e-01 6.15879714e-01 -5.05279720e-01
1.39870182e-01 9.00347903e-02 4.58438247e-01 1.00987506e+00
5.48376799e-01 -2.78170228e-01 -1.00348401e+00 -3.13820153e-01
4.05290544e-01 -7.60780573e-02 -2.83755839e-01 -5.85861623e-01
-1.18113089e+00 7.79655933e-01 7.67420709e-01 2.17216209e-01
-2.11249635e-01 1.10516541e-01 2.80597061e-01 6.54225826e-01
8.58245909e-01 4.85988826e-01 -5.08549452e-01 2.40041181e-01
-4.07420069e-01 -3.30277793e-02 1.22526467e+00 1.40561616e+00
1.26094890e+00 -5.46147645e-01 -3.62910688e-01 6.92240119e-01
9.54501331e-02 5.57641238e-02 2.50215828e-01 -3.88797820e-01
9.86379862e-01 9.13244247e-01 -4.32382375e-02 -1.01782465e+00
-1.58230573e-01 -5.00655890e-01 -6.19840205e-01 -6.11848235e-01
2.67341919e-02 -1.71080291e-01 -9.75082397e-01 1.66353858e+00
6.83472335e-01 8.05207551e-01 7.78096542e-02 4.52509612e-01
1.22663355e+00 5.99603891e-01 2.15075254e-01 -4.15784389e-01
8.89636934e-01 -1.18712080e+00 -5.05584002e-01 -2.87685096e-01
1.07324922e+00 -1.60640273e-02 9.26775575e-01 -2.26935089e-01
-9.85798597e-01 -3.64342839e-01 -9.37083602e-01 -1.38754636e-01
-8.22954595e-01 -5.48893988e-01 1.09762394e+00 2.51322776e-01
-9.17750299e-01 8.55961263e-01 -5.18653214e-01 -8.30825508e-01
5.27452707e-01 3.36144537e-01 -3.92512172e-01 -3.20031255e-01
-1.82563305e+00 1.06608236e+00 8.21087301e-01 -4.66428250e-02
-7.25604415e-01 -1.05558848e+00 -9.06380653e-01 2.93909132e-01
8.17799687e-01 -1.28627992e+00 7.30654478e-01 -3.82746369e-01
-1.14732993e+00 8.48488152e-01 -1.16013162e-01 -6.59857213e-01
3.16407859e-01 -2.30812296e-01 -6.35466516e-01 5.61441220e-02
1.02880456e-01 3.67947578e-01 5.49507678e-01 -1.18762970e+00
-6.68575585e-01 -1.41076297e-01 3.72921735e-01 3.55986148e-01
-4.08952028e-01 -7.07288802e-01 -4.88956481e-01 -1.99936435e-01
1.50373969e-02 -6.73707962e-01 3.27714458e-02 -4.07715172e-01
-5.95652819e-01 -4.66815412e-01 4.60746735e-01 -2.85218954e-02
1.20648372e+00 -1.54760587e+00 2.83900350e-01 2.81173527e-01
4.13636029e-01 4.23887938e-01 -2.71499813e-01 8.10796916e-01
-5.00527676e-03 3.03371586e-02 1.94023028e-01 -2.75291596e-02
-1.31502330e-01 3.03238362e-01 -2.85080791e-01 1.73218846e-01
-5.53155225e-03 1.42640698e+00 -1.56740534e+00 -6.99127078e-01
-7.15110227e-02 6.26108944e-02 -4.48697895e-01 1.58351019e-01
-4.41575885e-01 -7.92893395e-03 -6.06490850e-01 7.34165967e-01
2.68080503e-01 -6.99925244e-01 3.54759187e-01 -2.07810402e-01
4.83991772e-01 2.55094886e-01 -7.67702818e-01 1.90733445e+00
-5.41970372e-01 2.17091516e-01 -7.41539061e-01 -1.13053203e+00
8.66335928e-01 2.11866513e-01 2.20948607e-01 -3.43570232e-01
-4.20146063e-02 2.19353929e-01 -1.30369306e-01 -6.35389507e-01
2.57623285e-01 -1.71586514e-01 2.17231527e-01 2.51260102e-01
7.94387043e-01 4.47533168e-02 5.33290803e-01 8.25416982e-01
1.32291615e+00 2.52592921e-01 6.15692377e-01 1.59961805e-01
2.47248858e-01 -3.08467019e-02 5.84676504e-01 9.64400113e-01
1.40068144e-01 4.56395373e-02 4.47164506e-01 -3.17794532e-01
-5.89454949e-01 -1.12733185e+00 3.60385925e-01 1.27802026e+00
5.26644588e-01 -8.21186125e-01 -1.87722579e-01 -1.02206588e+00
1.38540298e-01 8.70599210e-01 -9.22396302e-01 -5.67989469e-01
-1.52107254e-01 -6.96170330e-01 6.75015822e-02 4.90791112e-01
3.13187122e-01 -8.80779445e-01 1.96902677e-01 3.37644964e-01
-9.66890752e-02 -1.17946637e+00 -1.69778317e-01 1.45441040e-01
-1.06605899e+00 -1.29331183e+00 -3.16026568e-01 -8.56494725e-01
6.40283406e-01 1.61091477e-01 1.53466284e+00 1.98310494e-01
-3.40414464e-01 5.39174736e-01 -6.87562644e-01 4.33618948e-03
3.08931638e-02 4.74897623e-01 -9.57263336e-02 -9.17148665e-02
5.64583063e-01 -1.00991499e+00 -4.19129699e-01 -5.20549789e-02
-2.99960226e-01 3.41783799e-02 5.63247979e-01 1.12148142e+00
4.38056648e-01 -2.44489148e-01 8.37196171e-01 -1.69991171e+00
5.31591952e-01 -1.01592684e+00 -6.43109754e-02 9.47204709e-01
-8.29361677e-01 1.48791581e-01 5.59481800e-01 -4.15763646e-01
-1.28960288e+00 -3.20571899e-01 4.94277745e-01 -3.23771507e-01
2.80846059e-01 8.82776499e-01 -4.34208065e-02 -7.79026449e-02
8.81644487e-01 -8.31853971e-02 -6.67731762e-01 -1.63142249e-01
9.02817667e-01 4.10864614e-02 3.49744797e-01 -7.23505080e-01
9.76231396e-01 4.56350058e-01 1.20459184e-01 -5.08546233e-01
-1.57072175e+00 -7.28153825e-01 -8.84994090e-01 -1.71957970e-01
4.75998193e-01 -9.96232867e-01 -5.57283521e-01 -1.32565409e-01
-9.26219821e-01 -4.23791021e-01 -5.19860744e-01 5.60198724e-01
-4.78572071e-01 1.04179315e-01 -7.76313126e-01 -4.91830587e-01
-4.64058906e-01 -1.77513435e-01 5.76770663e-01 2.24504575e-01
-5.05130962e-02 -1.46850514e+00 5.24003386e-01 1.70840859e-01
6.39789030e-02 3.25719804e-01 1.06927311e+00 -8.38586032e-01
-7.73969710e-01 -1.49606422e-01 -1.51740626e-01 -3.77003640e-01
3.02711278e-01 -1.30673662e-01 -8.21502268e-01 -1.89972371e-01
-8.04171741e-01 -8.00644994e-01 1.27142465e+00 -6.29347488e-02
6.95412338e-01 -3.78540188e-01 -9.44970787e-01 5.81218541e-01
1.43905115e+00 -3.38691801e-01 4.95709151e-01 2.52373368e-01
9.68225300e-01 3.93676877e-01 8.82319033e-01 2.90818721e-01
7.02544391e-01 3.07323486e-01 1.40587538e-01 1.05028145e-01
-2.47134462e-01 -8.62037957e-01 -3.89210060e-02 9.84340668e-01
-4.49272454e-01 -2.93219090e-01 -7.73643076e-01 8.05418491e-01
-2.36531925e+00 -1.18460441e+00 -5.34679443e-02 1.95628500e+00
1.17098784e+00 2.70937324e-01 9.80030047e-04 -3.63903433e-01
9.05597627e-01 1.05352208e-01 -6.63915157e-01 -2.64553558e-02
-6.44100085e-02 1.86637595e-01 2.16247141e-01 5.52341938e-01
-9.25128639e-01 1.42106485e+00 5.73956633e+00 7.21006632e-01
-3.64040047e-01 2.49742791e-01 3.67689818e-01 -9.98838097e-02
-6.83361888e-01 4.50619549e-01 -9.48270738e-01 7.09991008e-02
7.61133492e-01 -5.76126695e-01 1.73062474e-01 8.63082528e-01
-5.70639729e-01 8.25089514e-02 -1.46775413e+00 7.70728350e-01
1.01982631e-01 -1.50679255e+00 2.61861742e-01 -3.22875798e-01
1.25023508e+00 6.18536174e-02 -3.08022857e-01 8.69898379e-01
6.60755336e-01 -7.84174860e-01 7.20039068e-04 8.52136791e-01
5.36982894e-01 -6.77517474e-01 7.34777331e-01 2.47947752e-01
-1.47647417e+00 1.38447376e-05 -6.30701303e-01 7.46265892e-03
2.25891069e-01 7.98616052e-01 -1.13801229e+00 1.06565106e+00
4.77225661e-01 1.14651787e+00 -6.01722598e-01 9.79221225e-01
-5.80497205e-01 6.26505435e-01 -1.76826775e-01 -2.48541921e-01
8.19809735e-02 3.09821479e-02 4.24404830e-01 1.05093002e+00
2.04191133e-01 4.93292361e-01 3.03609133e-01 6.29789352e-01
-5.52218139e-01 2.59883314e-01 -8.58313024e-01 -3.21618728e-02
7.32253432e-01 1.17518961e+00 -5.93480170e-01 -5.78001022e-01
-6.24178052e-01 8.67845774e-01 1.11299217e+00 5.96886039e-01
-7.08823979e-01 -3.13650519e-01 8.26710165e-02 1.94695853e-02
4.18256164e-01 2.23918319e-01 6.21002540e-02 -1.59246957e+00
-1.27105907e-01 -2.32155889e-01 1.01059377e+00 -6.35025740e-01
-1.63104212e+00 5.06831765e-01 2.67014146e-01 -9.78219926e-01
-3.37011814e-01 -2.11493522e-01 -7.53800392e-01 4.32375163e-01
-1.63489211e+00 -1.53851187e+00 -4.33883280e-01 6.70667529e-01
2.53356516e-01 -1.42985299e-01 7.74116337e-01 4.08358462e-02
-3.90639216e-01 5.97411215e-01 -1.67699456e-01 6.71661124e-02
7.61014163e-01 -1.35246503e+00 5.37571967e-01 6.37890756e-01
4.56875265e-01 6.97873831e-01 6.85026467e-01 -1.05479443e+00
-1.36424065e+00 -1.17817044e+00 1.07607484e+00 -3.70373070e-01
1.00205755e+00 -2.17120275e-01 -1.08376682e+00 1.18400156e+00
7.64385313e-02 6.17582083e-01 8.21285367e-01 1.04019606e+00
-6.16407871e-01 -2.05310270e-01 -9.43228483e-01 6.22131288e-01
1.74254870e+00 -5.44982851e-01 -1.01466572e+00 5.42420089e-01
8.63404930e-01 -2.58672744e-01 -1.04434681e+00 5.35030961e-01
2.22661927e-01 -6.89263403e-01 1.05378246e+00 -1.17725122e+00
4.25399125e-01 -5.75301833e-02 1.45755038e-01 -1.67331624e+00
-4.92152452e-01 -6.57970250e-01 -1.05095959e+00 1.29190898e+00
9.90052223e-01 -6.93034351e-01 8.45799267e-01 5.19100249e-01
1.11422516e-01 -9.38209832e-01 -5.74905336e-01 -9.58886802e-01
-2.91058570e-01 8.96894932e-02 5.12134254e-01 1.35782695e+00
7.68622220e-01 9.17371869e-01 -4.59437668e-01 5.39216399e-02
7.15848505e-01 5.26539207e-01 6.72198296e-01 -1.42686868e+00
-4.42949742e-01 2.17509702e-01 -5.59775352e-01 -5.60103476e-01
6.08726740e-01 -1.23071897e+00 -2.37224981e-01 -1.91987133e+00
6.40625238e-01 -4.05725121e-01 -4.63565558e-01 5.75369954e-01
-7.88130701e-01 -1.39267510e-02 -1.15275733e-01 3.93557400e-02
-1.17900229e+00 8.09531510e-01 1.24837899e+00 -3.65241826e-01
-3.36150795e-01 2.53226771e-03 -4.38294560e-01 5.00244737e-01
6.12994850e-01 -4.86422092e-01 -9.97469425e-01 -5.49570136e-02
8.16742718e-01 -1.01483539e-01 1.93498999e-01 -6.31849229e-01
6.83574915e-01 -1.14238650e-01 2.59986103e-01 -4.45634812e-01
3.51278752e-01 -4.41437274e-01 5.11785001e-02 2.07932189e-01
-5.25539517e-01 -6.48525119e-01 -2.31017396e-01 1.18159080e+00
-9.50571820e-02 -1.53199196e-01 2.42938310e-01 -3.03115666e-01
-1.17460310e+00 7.17294991e-01 4.27495658e-01 7.07234859e-01
1.07536042e+00 -3.53919640e-02 -5.38377702e-01 -4.56450343e-01
-1.03940880e+00 5.16450226e-01 1.71291590e-01 3.17786425e-01
6.10172033e-01 -1.42964005e+00 -6.31862640e-01 -5.05460203e-01
5.59895277e-01 9.50939581e-02 2.54293144e-01 8.19286168e-01
-3.89223658e-02 2.41594657e-01 6.35807440e-02 -2.23305419e-01
-9.75713015e-01 9.78770733e-01 9.61938351e-02 -4.70964223e-01
-8.39976907e-01 1.28566468e+00 -1.61384776e-01 -3.77131343e-01
1.46095976e-01 -2.10586414e-01 -1.60270303e-01 3.48494411e-01
3.04145724e-01 4.04570132e-01 -3.71798016e-02 -1.35094911e-01
-3.26391250e-01 3.90293330e-01 -4.27831143e-01 1.96155995e-01
1.45022118e+00 -1.19566366e-01 3.98778431e-02 8.33558559e-01
1.06811881e+00 -2.83338487e-01 -9.37384546e-01 -6.29046142e-01
2.37400293e-01 -4.74175513e-01 -3.03622156e-01 -5.83247602e-01
-7.29655802e-01 4.69888300e-01 -2.13972971e-01 2.25977853e-01
6.43879890e-01 5.05650043e-01 7.39456236e-01 8.73513579e-01
6.91307127e-01 -1.04931402e+00 6.84741735e-02 4.72124636e-01
5.08716106e-01 -1.34119546e+00 3.63048196e-01 -9.66121972e-01
-6.07442677e-01 9.21451926e-01 9.78455126e-01 8.96724686e-03
8.34127426e-01 -2.72958934e-01 -5.27661264e-01 -4.95465755e-01
-1.33553302e+00 -6.61240578e-01 5.29754221e-01 6.61335945e-01
4.52668905e-01 -6.92444667e-02 -1.71121910e-01 2.64307052e-01
8.76950026e-02 1.96544826e-01 2.35446155e-01 8.54696929e-01
-5.14210224e-01 -9.94692504e-01 3.54807258e-01 7.77317226e-01
3.90007123e-02 -3.69487494e-01 -5.35191476e-01 8.74748945e-01
-2.30888743e-02 7.48694360e-01 -2.87854373e-01 -3.91306669e-01
1.93494856e-01 2.13488087e-01 7.15141416e-01 -1.04976928e+00
-3.00049275e-01 -6.04568183e-01 5.73092341e-01 -5.53572953e-01
-6.24756932e-01 -3.13468516e-01 -1.08542776e+00 -2.44836658e-01
-6.85363770e-01 3.89462203e-01 -2.25991324e-01 1.18275344e+00
4.07827049e-01 4.90540504e-01 4.70439225e-01 -3.60297620e-01
-2.99012482e-01 -1.02884102e+00 -6.62804723e-01 4.77764428e-01
2.53130961e-02 -8.98506582e-01 -3.18176985e-01 -1.50316149e-01] | [8.787701606750488, 7.9650774002075195] |
5c9e6278-1e5d-4047-8e55-c356ca8f1d9f | bnpc-a-corpus-for-paraphrase-detection-in | null | null | https://openreview.net/forum?id=4xNbfr01sn2 | https://openreview.net/pdf?id=4xNbfr01sn2 | BnPC: A Corpus for Paraphrase Detection in Bangla | In this paper, we present the first benchmark dataset for paraphrase detection in Bangla language. Despite being the sixth most spoken language in the world, paraphrase identification in the Bangla language is barely explored. Our dataset contains 8,787 human-annotated sentence pairs collected from a total of 23 newspaper outlets' headlines on four categories. We explore different linguistic features and pre-trained language models to benchmark the dataset. We perform a human evaluation experiment to obtain a better understanding of the task's constraints, which reveals intriguing insights. We make our dataset and code publicly available. | ['Anonymous'] | 2021-12-17 | null | null | null | acl-arr-december-2022-12 | ['paraphrase-identification'] | ['natural-language-processing'] | [-1.76740617e-01 -3.03512454e-01 -5.24981976e-01 -6.20877922e-01
-1.06376731e+00 -9.31070209e-01 7.68510282e-01 4.03327584e-01
-5.47453821e-01 5.37833989e-01 7.92800665e-01 -4.99979198e-01
1.80882543e-01 -4.59010750e-01 -3.78988594e-01 -2.09551543e-01
2.28503555e-01 2.03379378e-01 -1.42014354e-01 -5.25930405e-01
7.68878579e-01 4.95022595e-01 -9.61303115e-01 9.45589125e-01
9.14137423e-01 6.30855143e-01 8.63484964e-02 4.25256222e-01
7.63975456e-02 8.42355371e-01 -6.50881350e-01 -1.15186691e+00
2.21365362e-01 -6.19655669e-01 -1.31216478e+00 1.96756758e-02
4.17930901e-01 -2.30022922e-01 -7.31199384e-01 9.84789014e-01
5.78328669e-01 -3.19513768e-01 5.49914598e-01 -8.40099931e-01
-1.12565780e+00 7.90910065e-01 -6.74329698e-01 7.11891770e-01
8.58693302e-01 9.61918086e-02 1.46870899e+00 -1.02095294e+00
7.94735789e-01 1.15832245e+00 6.50336325e-01 4.37333018e-01
-7.57866263e-01 -5.15696287e-01 -4.24319446e-01 4.42082226e-01
-1.30672550e+00 -7.34043717e-01 1.04944623e+00 -2.25148708e-01
1.14524233e+00 2.07396418e-01 4.79280204e-01 1.49861288e+00
1.76456317e-01 1.09075356e+00 1.44009101e+00 -6.58372581e-01
-1.52810782e-01 4.31424409e-01 5.25876939e-01 5.50119340e-01
2.86951005e-01 -5.19119620e-01 -9.32171166e-01 -2.35268787e-01
2.44676113e-01 -2.23726153e-01 -9.24957693e-02 3.13552208e-02
-1.39495111e+00 1.08908963e+00 6.78524226e-02 3.51551712e-01
7.72558078e-02 -5.95595241e-01 9.55396116e-01 9.25152063e-01
3.65110368e-01 7.75778055e-01 -2.00671077e-01 -5.03031671e-01
-8.14799309e-01 1.93220168e-01 9.80542719e-01 1.09983087e+00
3.05881977e-01 -5.30968547e-01 -3.55617814e-02 1.48096955e+00
-9.83829126e-02 3.97962004e-01 8.32893789e-01 -7.19612241e-01
9.79862273e-01 6.98776722e-01 -6.02997877e-02 -1.04423332e+00
-1.77141562e-01 -2.02001736e-01 -7.09311604e-01 -6.98821902e-01
6.02876604e-01 4.43131709e-03 -1.39708474e-01 1.26301455e+00
-4.53435332e-01 -6.76704347e-01 1.06544010e-01 8.14887822e-01
1.05017066e+00 5.28829992e-01 -4.85953212e-01 -1.55983433e-01
1.49537623e+00 -1.11230338e+00 -4.46703106e-01 -2.35743538e-01
8.22784364e-01 -1.21909702e+00 1.43623888e+00 3.65462840e-01
-1.17188048e+00 -4.40614849e-01 -1.07728148e+00 -4.55325276e-01
-3.60460490e-01 1.97235867e-01 6.40720248e-01 5.15957236e-01
-6.35925233e-01 2.53027558e-01 -3.80904853e-01 -9.05385315e-01
3.58600020e-01 -3.21437776e-01 -5.55704534e-01 -1.62010580e-01
-1.36868405e+00 1.32019687e+00 3.79347056e-02 -9.79396105e-02
-4.01340902e-01 -2.27978066e-01 -7.04646885e-01 -3.56990725e-01
-1.10263407e-01 -2.48456687e-01 1.18771780e+00 -7.05047488e-01
-1.23610950e+00 1.79520571e+00 -4.14752513e-01 -6.29819930e-01
5.88165581e-01 -2.52776861e-01 -5.75963080e-01 7.91771933e-02
1.15092769e-01 2.24661738e-01 5.12909532e-01 -6.26693010e-01
-3.87520492e-01 -2.41922021e-01 2.33293205e-01 4.50270437e-02
-4.16461676e-01 8.72263849e-01 -3.36912841e-01 -9.81895030e-01
1.39786405e-02 -8.65510166e-01 1.69617385e-01 -3.34759831e-01
-4.29672539e-01 -4.00289267e-01 3.62847477e-01 -1.08055425e+00
1.61025262e+00 -2.01912355e+00 2.56736949e-02 -1.97962120e-01
-1.20871828e-03 1.16771646e-01 -1.53923899e-01 1.07429147e+00
4.75284420e-02 -1.51776765e-02 -1.17500447e-01 -3.19327116e-01
-5.77507243e-02 1.99650023e-02 -6.08570695e-01 5.45133293e-01
1.79588690e-01 1.24333799e+00 -9.38592911e-01 -5.16822934e-01
-1.79728284e-01 -1.88517317e-01 -4.22889203e-01 1.25417784e-01
3.58650029e-01 1.58544302e-01 -1.69586897e-01 9.45429623e-01
5.23585260e-01 5.11211716e-03 -1.12230211e-01 -1.41982585e-02
6.82557523e-02 8.94623697e-01 -3.38413239e-01 1.85803294e+00
-5.84634304e-01 1.12630379e+00 -1.97109118e-01 -9.82827842e-01
1.02657890e+00 -2.17299573e-02 3.17746447e-03 -8.71046543e-01
2.10305348e-01 3.41145039e-01 1.47268280e-01 -7.11091697e-01
8.17162752e-01 -1.63604959e-03 -7.50025332e-01 5.64647615e-01
-1.39200687e-01 -3.54038656e-01 3.45954210e-01 4.79084343e-01
1.10610604e+00 -5.28300464e-01 5.80298245e-01 -5.90493202e-01
7.44695902e-01 1.87078491e-01 1.97001949e-01 9.22901571e-01
-5.39316833e-01 7.98191607e-01 6.87762439e-01 -5.14101326e-01
-1.27249086e+00 -1.05415595e+00 -4.47836101e-01 1.01256824e+00
-8.45149904e-03 -6.16123617e-01 -5.33462882e-01 -7.17933118e-01
9.10080746e-02 6.20442867e-01 -4.71679181e-01 -2.09082678e-01
-7.02043176e-01 -6.35338306e-01 1.06541550e+00 2.79801100e-01
7.33890295e-01 -1.06069505e+00 -1.31020278e-01 -1.74397349e-01
-6.21831477e-01 -1.32748806e+00 -7.53480911e-01 -4.41201925e-02
-6.48811758e-01 -9.65479136e-01 -6.67127371e-01 -1.20285261e+00
3.13616246e-01 3.63553077e-01 1.37872958e+00 -1.05275653e-01
-2.79047102e-01 -5.89970611e-02 -5.78299642e-01 -3.74660283e-01
-9.13881838e-01 2.90972948e-01 -1.60023142e-02 -3.99584025e-01
1.09838641e+00 -3.18229973e-01 -2.88243949e-01 1.84572861e-01
-3.49094063e-01 9.88907367e-02 5.36366999e-01 9.81235921e-01
-1.03533953e-01 -3.59806269e-01 8.02554250e-01 -8.98109376e-01
1.20099425e+00 -6.32748187e-01 -3.13768864e-01 3.83523077e-01
-1.73628464e-01 -1.68138281e-01 7.24113703e-01 -2.23791942e-01
-7.37757921e-01 -2.03340411e-01 -2.53539324e-01 2.62866199e-01
-1.00410329e-02 7.85099149e-01 2.84674913e-02 2.13770583e-01
7.15061784e-01 4.30981547e-01 -1.01433761e-01 -5.87549627e-01
2.85689950e-01 1.57612836e+00 8.51305962e-01 -3.57345343e-01
5.04602015e-01 2.23605782e-01 -5.86095393e-01 -1.12339365e+00
-1.04646277e+00 -1.00012600e+00 -9.26358402e-01 -5.38674146e-02
3.49042386e-01 -1.16029382e+00 -3.46204072e-01 7.29097903e-01
-1.36482573e+00 1.92698881e-01 8.33310410e-02 2.14788392e-01
-3.99761915e-01 7.64149249e-01 -8.98435771e-01 -2.96818316e-01
-4.58766401e-01 -7.37908721e-01 8.17947805e-01 -1.68455765e-01
-7.81699181e-01 -7.65307009e-01 2.25873291e-01 8.93253207e-01
2.76180893e-01 -2.45013073e-01 8.43507469e-01 -9.12030101e-01
7.01129362e-02 -2.69773126e-01 -3.71987194e-01 5.23618042e-01
2.94420034e-01 -1.38920859e-01 -7.37080157e-01 -3.97146136e-01
3.24363083e-01 -7.02409506e-01 8.80560875e-01 -2.12796926e-01
1.08085668e+00 -4.12492454e-01 1.10218190e-01 1.74323320e-01
9.19612706e-01 -2.56058097e-01 4.88262087e-01 5.86903274e-01
4.43622261e-01 7.87141263e-01 5.83819389e-01 5.30855119e-01
4.18770492e-01 6.04602695e-01 -2.40125924e-01 4.18165922e-01
-1.98103115e-01 -4.43382382e-01 5.50072551e-01 1.52815366e+00
3.38844180e-01 -1.33068100e-01 -1.04486287e+00 8.81965816e-01
-1.50498879e+00 -1.15947318e+00 -1.31717592e-01 2.00027204e+00
1.25306487e+00 2.69937605e-01 3.83645415e-01 1.81177422e-01
6.71215594e-01 2.30024412e-01 -2.89647728e-01 -7.87873507e-01
-5.92235267e-01 -8.39530304e-03 3.38179559e-01 3.82138669e-01
-1.05674660e+00 9.70516145e-01 6.74314928e+00 7.21512377e-01
-8.43466759e-01 4.61150706e-02 4.00274366e-01 -1.86513767e-01
-2.11428151e-01 -1.04316249e-01 -6.88161314e-01 7.73700237e-01
8.83954525e-01 -5.46127498e-01 3.56194198e-01 7.15402365e-01
3.18633795e-01 -5.95261380e-02 -1.31944335e+00 1.22359610e+00
4.99818534e-01 -1.24533784e+00 1.87405959e-01 -2.33175069e-01
6.34177089e-01 3.49284530e-01 -1.31178781e-01 3.56379837e-01
8.00519362e-02 -1.01171792e+00 6.09302461e-01 1.79796502e-01
6.06248498e-01 -7.69237280e-01 7.60697663e-01 5.60183465e-01
-4.65640694e-01 -1.59557194e-01 -6.03759706e-01 -4.01384681e-01
4.84079123e-02 5.02965987e-01 -4.64968443e-01 2.60165989e-01
7.45653033e-01 1.17805505e+00 -1.07581210e+00 8.89451802e-01
-2.48034686e-01 9.53165054e-01 -1.49451094e-02 -4.65235591e-01
2.56422400e-01 -2.87852913e-01 5.87406337e-01 1.79245663e+00
-1.27985179e-01 -8.94861966e-02 -3.76431465e-01 6.00117922e-01
-3.88200909e-01 3.03557545e-01 -8.68807733e-01 -3.09613794e-01
5.84889829e-01 1.05099571e+00 -4.53712791e-01 -2.51289696e-01
-8.24367762e-01 1.34202671e+00 7.28608668e-01 2.85526756e-02
-7.97244549e-01 -8.23586881e-01 4.86674041e-01 -5.32970056e-02
-3.09331745e-01 -2.29494274e-01 -3.87600154e-01 -1.64024472e+00
2.57174313e-01 -1.20054269e+00 3.70677650e-01 -6.35628223e-01
-1.94267702e+00 2.70705551e-01 2.83783451e-02 -1.11310744e+00
-1.55914962e-01 -5.63329101e-01 -7.24742770e-01 9.94396448e-01
-1.17067611e+00 -1.06716967e+00 -1.22606352e-01 3.89275104e-01
1.02476871e+00 -5.75465143e-01 6.54390275e-01 3.89023185e-01
-5.82424402e-01 1.00814259e+00 4.15873051e-01 5.06246209e-01
1.04489136e+00 -9.85675871e-01 9.84777331e-01 7.73013532e-01
2.30314389e-01 9.57838297e-01 7.24708438e-01 -2.84037530e-01
-1.43246639e+00 -5.41991889e-01 1.80942523e+00 -8.58884633e-01
1.18570769e+00 -7.21063375e-01 -9.91605461e-01 4.43015993e-01
5.89429140e-01 -5.51901698e-01 6.67838871e-01 1.42122373e-01
-6.71970904e-01 1.36233404e-01 -1.01181424e+00 5.16527593e-01
1.24700081e+00 -1.24460089e+00 -1.30246568e+00 5.72166502e-01
4.22443897e-01 -2.68554449e-01 -6.90650940e-01 1.13047764e-01
6.32354975e-01 -9.07150149e-01 7.99076378e-01 -8.37482035e-01
9.84658599e-01 2.83889115e-01 -1.55509144e-01 -1.24230540e+00
-2.80666113e-01 -4.26425457e-01 6.87922388e-02 1.35776055e+00
4.83237118e-01 -5.92493355e-01 7.66155422e-01 3.29818487e-01
-1.79465432e-02 -9.18239951e-01 -9.09976125e-01 -9.45259154e-01
6.19577527e-01 -1.75739974e-01 3.39942485e-01 1.17620778e+00
7.47627914e-01 6.12013042e-01 -4.25518274e-01 -3.92281353e-01
2.59839416e-01 4.21780974e-01 7.07008481e-01 -6.36734664e-01
-3.04072261e-01 -6.21067643e-01 -2.02995792e-01 -1.26046109e+00
5.01186848e-01 -1.30013466e+00 -3.94791663e-01 -1.36176825e+00
8.95546794e-01 1.11793494e-02 -6.55797310e-03 3.02381694e-01
7.83217698e-03 3.33397418e-01 -3.50670479e-02 6.14507854e-01
-4.09226954e-01 3.10542464e-01 9.34978247e-01 -3.36846143e-01
6.82333410e-02 1.28981039e-01 -8.82998705e-01 5.49192965e-01
1.09138405e+00 -2.28356868e-01 -2.33820692e-01 -6.38640106e-01
2.95471936e-01 -1.91958085e-01 1.39177203e-01 -4.05359626e-01
1.16472587e-01 -1.82727754e-01 3.58349271e-02 -7.32495785e-01
-5.91028435e-03 -2.73285747e-01 -4.64874834e-01 4.94613916e-01
-8.44064474e-01 5.55902600e-01 8.73071030e-02 2.23132253e-01
-4.27719295e-01 -5.94444811e-01 9.47278142e-01 -1.36415794e-01
-4.70934123e-01 -2.84353793e-01 -5.97764552e-01 6.07805073e-01
7.36723840e-01 -1.73595220e-01 -4.93146539e-01 -5.82483530e-01
-1.67369321e-02 2.00382635e-01 6.96574628e-01 9.46575820e-01
4.93191034e-01 -1.23411477e+00 -1.20435059e+00 1.22284859e-01
5.75964689e-01 -7.22017884e-01 -8.29111710e-02 6.01545751e-01
-6.81772470e-01 6.47262335e-01 -3.69106829e-01 -2.05785334e-01
-1.38809323e+00 5.62594473e-01 2.28725560e-02 -1.66168272e-01
-3.69286329e-01 7.54389405e-01 -5.02403438e-01 -5.30979574e-01
8.35915953e-02 1.51604619e-02 -2.26170391e-01 2.05708265e-01
3.92638594e-01 5.31382024e-01 1.06964909e-01 -9.76506352e-01
-5.21167278e-01 2.24890396e-01 -5.65176249e-01 -2.68075597e-02
1.05510855e+00 -4.64533180e-01 -4.44566786e-01 5.28275192e-01
1.63460207e+00 5.06406069e-01 -8.21774229e-02 -2.91441172e-01
3.15332234e-01 -6.61926866e-01 -4.29109782e-01 -7.48615086e-01
-4.03703153e-01 7.91536570e-01 -1.89347550e-01 1.74772650e-01
7.95612395e-01 2.84023374e-01 1.05181813e+00 7.43549764e-01
4.16480392e-01 -1.30998063e+00 2.15879828e-01 9.46181417e-01
1.35728109e+00 -1.43883705e+00 1.58686459e-01 -1.33611009e-01
-8.60786259e-01 9.65028226e-01 3.64489257e-01 -2.18106970e-01
4.42296654e-01 -2.56799281e-01 3.05141747e-01 -8.75341799e-03
-6.20838404e-01 3.74779105e-01 2.62063503e-01 1.96042418e-01
8.68568599e-01 -2.14991167e-01 -8.77934694e-01 7.97210932e-01
-9.82339025e-01 -3.53790790e-01 7.26496518e-01 8.02079976e-01
-2.34514907e-01 -1.09401786e+00 3.34881917e-02 7.18585849e-01
-7.47604430e-01 -3.07523400e-01 -9.74662364e-01 5.44150054e-01
-5.93226612e-01 1.14800406e+00 -1.36933148e-01 -1.86452657e-01
4.06126767e-01 7.34241307e-02 5.83363950e-01 -6.57654524e-01
-6.31859660e-01 -6.74013197e-01 3.77338618e-01 -2.87723809e-01
-2.64422923e-01 -9.87429321e-01 -6.98829710e-01 -7.23937869e-01
-2.17529297e-01 1.10018209e-01 3.26200902e-01 7.42510617e-01
1.89688668e-01 -4.15891618e-01 9.64164317e-01 -3.19929928e-01
-9.60897326e-01 -1.13246047e+00 -4.92117494e-01 8.58267426e-01
9.92629975e-02 -3.30527537e-02 -5.62095582e-01 -1.07953437e-01] | [11.304984092712402, 9.195080757141113] |
cd8825cd-ca8f-4acd-83e7-5f5a2ad79d73 | per-vis-person-retrieval-in-video | 2012.02408 | null | https://arxiv.org/abs/2012.02408v1 | https://arxiv.org/pdf/2012.02408v1.pdf | PeR-ViS: Person Retrieval in Video Surveillance using Semantic Description | A person is usually characterized by descriptors like age, gender, height, cloth type, pattern, color, etc. Such descriptors are known as attributes and/or soft-biometrics. They link the semantic gap between a person's description and retrieval in video surveillance. Retrieving a specific person with the query of semantic description has an important application in video surveillance. Using computer vision to fully automate the person retrieval task has been gathering interest within the research community. However, the Current, trend mainly focuses on retrieving persons with image-based queries, which have major limitations for practical usage. Instead of using an image query, in this paper, we study the problem of person retrieval in video surveillance with a semantic description. To solve this problem, we develop a deep learning-based cascade filtering approach (PeR-ViS), which uses Mask R-CNN [14] (person detection and instance segmentation) and DenseNet-161 [16] (soft-biometric classification). On the standard person retrieval dataset of SoftBioSearch [6], we achieve 0.566 Average IoU and 0.792 %w $IoU > 0.4$, surpassing the current state-of-the-art by a large margin. We hope our simple, reproducible, and effective approach will help ease future research in the domain of person retrieval in video surveillance. The source code and pretrained weights available at https://parshwa1999.github.io/PeR-ViS/. | ['Vandit Gajjar', 'Arpit Garg', 'Parshwa Shah'] | 2020-12-04 | null | null | null | null | ['person-retrieval'] | ['computer-vision'] | [ 8.03511441e-02 -5.76618195e-01 -8.13616589e-02 -4.70939785e-01
-6.36850059e-01 -5.60823679e-01 6.56966150e-01 4.89109047e-02
-5.65283775e-01 4.93044406e-01 4.27787825e-02 2.32228264e-01
-1.20784836e-02 -9.27349627e-01 -6.60855770e-01 -6.55576527e-01
1.12746909e-01 2.47877195e-01 1.19172752e-01 -1.34077087e-01
-1.60092875e-01 1.95491046e-01 -1.74014819e+00 1.94561124e-01
3.98749799e-01 1.36079741e+00 -1.07532009e-01 6.55628324e-01
2.45950624e-01 2.83462614e-01 -6.89355493e-01 -9.62433338e-01
4.71517503e-01 -2.31376424e-01 -7.10862100e-01 -9.21560898e-02
7.67558992e-01 -7.62816310e-01 -6.18564487e-01 1.12500525e+00
7.80998707e-01 2.39562653e-02 6.56557620e-01 -1.09828556e+00
-7.85350382e-01 4.47198637e-02 -4.63059753e-01 2.84476727e-01
8.62501085e-01 2.62263536e-01 6.96040094e-01 -4.69992697e-01
4.63671386e-01 1.42275846e+00 7.48382390e-01 1.01322198e+00
-6.30384862e-01 -7.84358144e-01 -1.41467061e-02 3.21794152e-01
-1.75319803e+00 -4.03244853e-01 3.07133824e-01 -4.19537723e-01
7.39620149e-01 5.44180334e-01 9.04844224e-01 1.31082475e+00
-1.86657347e-02 8.79235625e-01 8.41970146e-01 2.98564956e-02
-4.03610505e-02 2.35099152e-01 2.56675810e-01 8.92025650e-01
4.81271535e-01 1.18573867e-01 -3.71855885e-01 -3.27123284e-01
7.16724575e-01 6.88962877e-01 -3.84127676e-01 4.60364856e-02
-1.10863721e+00 6.69352472e-01 6.09078765e-01 1.63798645e-01
-3.84211719e-01 2.58987486e-01 4.31844354e-01 4.79718685e-01
4.91564214e-01 1.55354038e-01 -1.39024511e-01 6.40140176e-02
-1.09834445e+00 5.83448529e-01 7.32173026e-01 1.08368182e+00
5.01216531e-01 -3.62140387e-01 -4.85327393e-01 8.06189418e-01
3.46731633e-01 1.10843945e+00 1.99496284e-01 -8.37303936e-01
1.70173906e-02 6.43063545e-01 2.31699616e-01 -1.16212606e+00
-1.01065740e-01 -2.16692761e-01 -9.43098068e-01 -4.19707060e-01
6.36774242e-01 -4.03265506e-02 -9.28840458e-01 1.62810981e+00
3.29033703e-01 1.38024375e-01 -1.18511520e-01 1.31588471e+00
1.26436901e+00 6.39014184e-01 1.23814374e-01 4.98397984e-02
1.87046516e+00 -7.31470406e-01 -5.04272521e-01 -7.49870539e-02
7.28427991e-02 -6.07763886e-01 6.35478735e-01 1.20214783e-01
-1.03041828e+00 -6.45885825e-01 -6.91811979e-01 1.55137759e-02
-7.90850103e-01 2.00308040e-01 5.50216913e-01 9.39960897e-01
-1.20145488e+00 2.53960460e-01 -6.01365983e-01 -1.10654044e+00
4.78920668e-01 6.10043824e-01 -2.70220160e-01 -2.76173979e-01
-1.36708939e+00 4.65989560e-01 8.20346177e-03 3.18345539e-02
-1.06686199e+00 -5.05851328e-01 -7.28276372e-01 -1.92683339e-02
5.75372696e-01 -1.20906997e+00 7.97626734e-01 -9.80333209e-01
-1.19104385e+00 1.43332183e+00 -1.41052932e-01 -4.44081485e-01
5.45834899e-01 -2.79085219e-01 -4.98156369e-01 4.22230721e-01
9.60182622e-02 8.92023444e-01 1.15320611e+00 -8.63721669e-01
-6.03033066e-01 -6.84256136e-01 3.89035046e-01 -3.86806130e-02
-5.89186251e-01 3.59983861e-01 -1.12511623e+00 -5.41901529e-01
-4.68073100e-01 -8.73906732e-01 7.11214244e-02 3.26791734e-01
-1.77747771e-01 -4.23235416e-01 4.90637094e-01 -9.05122042e-01
1.02384436e+00 -2.03799629e+00 -8.69957265e-03 1.56369463e-01
2.12768659e-01 5.08346021e-01 -1.50601804e-01 1.86741143e-01
4.27366287e-01 2.47210194e-03 -1.35606572e-01 -2.51623660e-01
1.43229347e-02 -3.26172292e-01 6.19325265e-02 7.15209723e-01
8.91973078e-02 1.08285558e+00 -8.02751124e-01 -6.50907636e-01
1.57818630e-01 8.65178883e-01 -4.33043629e-01 4.77698326e-01
1.06896862e-01 1.14807986e-01 -6.70099139e-01 1.24362719e+00
5.95203161e-01 -2.01803923e-01 -1.64891735e-01 -3.99921268e-01
1.99739933e-01 -4.35260415e-01 -1.01623571e+00 1.61478794e+00
7.24969655e-02 4.78979737e-01 3.51209849e-01 -9.76472557e-01
7.14184999e-01 3.20288539e-01 6.39200211e-01 -5.69946229e-01
3.62084687e-01 7.06433803e-02 -3.67727071e-01 -7.01024234e-01
4.77520287e-01 3.03269356e-01 -1.50464088e-01 9.65123847e-02
-1.27484538e-02 2.88527936e-01 2.77410597e-01 1.05101325e-01
1.05018485e+00 -1.97578245e-03 1.52740732e-01 -3.33493263e-01
7.43061304e-01 -7.03388751e-02 2.49215558e-01 9.95069027e-01
-5.90543091e-01 6.75008178e-01 2.63788234e-02 -6.54231191e-01
-8.68074417e-01 -9.79236245e-01 -1.07224442e-01 1.07820332e+00
4.07499343e-01 -1.98020726e-01 -1.15333116e+00 -4.49740529e-01
1.88653424e-01 -2.06120089e-01 -6.36560321e-01 -1.27038702e-01
-5.32626510e-01 -8.49602222e-01 6.89618051e-01 3.24576169e-01
9.46263909e-01 -1.12243545e+00 -7.77873039e-01 -1.52056843e-01
-2.35894531e-01 -1.08446515e+00 -7.09779561e-01 -6.61933780e-01
-4.63804513e-01 -1.29192972e+00 -1.58816028e+00 -6.90223217e-01
7.52403140e-01 3.71657342e-01 1.13597703e+00 5.13358831e-01
-5.82554698e-01 1.07933795e+00 -3.44348699e-01 -2.86047161e-01
1.50055438e-01 1.78898707e-01 1.57207668e-01 2.80148894e-01
8.36308062e-01 2.97464848e-01 -1.21268559e+00 3.26054156e-01
-9.47122157e-01 -3.94392431e-01 3.16739827e-01 6.58522844e-01
3.02660555e-01 -2.07899690e-01 2.44349152e-01 -4.55914587e-01
6.03741527e-01 -4.43873376e-01 -5.45955241e-01 5.06649792e-01
-3.85403723e-01 -4.66121614e-01 1.53976187e-01 -4.10603076e-01
-7.84103215e-01 -1.12946086e-01 -9.04834643e-02 -4.28775787e-01
-3.63971055e-01 1.37414113e-01 -1.03217535e-01 -1.28526874e-02
3.39162171e-01 3.50954622e-01 5.08046299e-02 -2.26633340e-01
-8.33988562e-02 8.52701545e-01 4.36896592e-01 -3.75991166e-01
7.47826040e-01 7.30252802e-01 -4.07365292e-01 -7.75373638e-01
-6.85071349e-01 -8.95268619e-01 -2.84198552e-01 -4.48552012e-01
1.16423857e+00 -1.16271281e+00 -1.25386274e+00 7.60404587e-01
-1.10660994e+00 -4.64732833e-02 1.21674374e-01 2.64868587e-01
-1.79126635e-01 3.67671132e-01 -6.75261676e-01 -8.99188459e-01
-8.33307147e-01 -1.10584664e+00 1.49741852e+00 5.02969503e-01
-5.56166396e-02 -7.17370510e-01 -3.82347792e-01 8.48360837e-01
6.46365643e-01 1.59434795e-01 1.94024146e-01 -6.11090720e-01
-8.14726412e-01 -5.75863898e-01 -5.57053208e-01 2.05488458e-01
-2.85021141e-02 -3.58486682e-01 -9.58043575e-01 -6.94685698e-01
-2.32199878e-01 -2.74844348e-01 1.11403406e+00 5.66236496e-01
1.26747859e+00 -1.92159101e-01 -5.78830898e-01 6.56965852e-01
1.41267037e+00 1.67308778e-01 5.58851421e-01 2.11646140e-01
5.15060544e-01 6.16351008e-01 5.04850686e-01 4.63734806e-01
4.42648202e-01 7.58362174e-01 1.38691798e-01 -1.94399521e-01
-1.80751532e-01 1.57446101e-01 3.92745912e-01 1.66537508e-01
-4.64454889e-01 -4.89983290e-01 -8.45532477e-01 4.47131425e-01
-1.83159864e+00 -1.14357901e+00 2.84039021e-01 2.15278101e+00
6.10183954e-01 -1.76421970e-01 3.65181178e-01 -2.83552885e-01
9.40488338e-01 1.18258089e-01 -4.71072793e-01 1.41548902e-01
-3.30917910e-02 -4.38365526e-02 5.73779285e-01 2.11557269e-01
-1.44414473e+00 9.08058286e-01 5.68798494e+00 9.32262540e-01
-9.61224020e-01 1.61014631e-01 7.52250195e-01 -2.67745465e-01
1.85323700e-01 -7.11805880e-01 -1.18118370e+00 6.07551277e-01
7.05417693e-01 1.63875312e-01 6.80540383e-01 5.26061237e-01
-6.49114102e-02 -7.28641599e-02 -1.14978707e+00 1.44937885e+00
5.53856909e-01 -1.07403874e+00 9.17197168e-02 -5.15210889e-02
4.55113411e-01 -2.29802519e-01 2.36463323e-01 3.43330145e-01
-2.93857455e-01 -8.86166036e-01 5.35897970e-01 8.05356085e-01
9.26975131e-01 -4.24388140e-01 1.02704632e+00 -1.19192759e-02
-1.26132691e+00 -6.84011355e-02 -3.87832612e-01 2.78478980e-01
-8.39971527e-02 2.50831634e-01 -4.18374062e-01 3.82276505e-01
1.39704716e+00 7.61889398e-01 -5.61719298e-01 1.00351453e+00
3.60778511e-01 3.58761072e-01 -3.52131128e-01 -2.65827149e-01
2.66075283e-02 -1.31715506e-01 4.95475769e-01 1.42122447e+00
3.96208763e-01 3.15180331e-01 3.05543274e-01 8.65224242e-01
-2.42543042e-01 1.96031928e-02 -5.77161431e-01 7.12697534e-03
1.36925444e-01 1.12861145e+00 -6.43221378e-01 -6.54686868e-01
-5.33769488e-01 1.12586558e+00 -2.75094718e-01 5.07661641e-01
-8.88530314e-01 -1.57824829e-01 7.88528740e-01 2.99844205e-01
2.28281647e-01 4.82297242e-02 3.55610698e-01 -1.29105699e+00
6.26348183e-02 -9.91743624e-01 5.88508189e-01 -5.37535310e-01
-1.47679877e+00 5.31180441e-01 2.41268709e-01 -1.03627455e+00
-7.69746453e-02 -7.26564109e-01 -2.44833916e-01 6.98903322e-01
-1.42941082e+00 -1.34998345e+00 -6.37059331e-01 9.08951104e-01
5.85934997e-01 -4.95106131e-01 7.91405141e-01 6.43308520e-01
-6.24288261e-01 7.25084186e-01 -1.71273842e-01 5.14499545e-01
8.53826761e-01 -7.76212513e-01 2.01747388e-01 5.77251852e-01
-2.89305806e-01 8.87185335e-01 4.81339484e-01 -7.02501357e-01
-1.65622020e+00 -1.02600145e+00 7.22466588e-01 -6.31903827e-01
2.44174853e-01 -3.81190896e-01 -5.20570278e-01 2.84862578e-01
2.26121202e-01 -1.52219906e-02 7.22016096e-01 -2.33845770e-01
-2.64029026e-01 -4.41304833e-01 -1.46122634e+00 6.29986227e-01
1.21726549e+00 -5.98432183e-01 -2.34266400e-01 5.13615012e-01
4.39395666e-01 -2.62663215e-01 -9.59665060e-01 3.78892243e-01
9.05270875e-01 -8.31638873e-01 1.48469687e+00 -2.71026224e-01
5.17438091e-02 -7.77006075e-02 -1.57131121e-01 -5.80733836e-01
-1.96798950e-01 -3.37745428e-01 -6.12569489e-02 1.09850287e+00
4.50370759e-02 -5.70193470e-01 7.99233675e-01 8.06314647e-01
3.46918762e-01 -6.81871831e-01 -7.36632109e-01 -7.16674209e-01
-3.00129205e-01 -1.34176999e-01 7.52569556e-01 6.11066222e-01
-5.03272414e-01 -1.72158271e-01 -5.46344697e-01 5.26189245e-02
8.70149791e-01 1.04716979e-01 6.67032242e-01 -1.21764064e+00
-1.22422896e-01 -3.85690838e-01 -4.99255091e-01 -9.57332671e-01
9.14992988e-02 -7.66822934e-01 -1.61253601e-01 -1.49154460e+00
7.24405944e-01 -8.91702920e-02 -4.48308915e-01 3.58844846e-01
-2.05307811e-01 6.95343733e-01 2.66704053e-01 3.20037931e-01
-9.73605454e-01 2.27971181e-01 1.11900592e+00 -6.82352245e-01
1.10512130e-01 1.05321340e-01 -4.95687068e-01 4.79862750e-01
7.95700490e-01 -2.25705087e-01 4.03700769e-03 -4.46301639e-01
-3.93792018e-02 4.26816009e-03 9.36592102e-01 -1.10542440e+00
3.54409277e-01 9.45636481e-02 6.96824312e-01 -3.64338160e-01
7.67669499e-01 -9.48028982e-01 7.60747418e-02 7.93374777e-01
-3.08581322e-01 -1.85087752e-02 -1.00233242e-01 5.23264229e-01
-1.75766587e-01 -1.38304278e-01 4.38953042e-01 -4.29470718e-01
-8.25608790e-01 8.16330075e-01 -1.83025166e-01 -1.79818884e-01
8.93655360e-01 -2.92832404e-01 -4.43194747e-01 -4.22335327e-01
-5.50527692e-01 3.13224137e-01 4.73697722e-01 5.25149107e-01
8.02718699e-01 -1.14450336e+00 -8.17402601e-01 2.52307743e-01
2.26378679e-01 -4.41929996e-01 2.04501852e-01 7.07370698e-01
-3.82511526e-01 5.85453868e-01 -2.43411317e-01 -6.97971880e-01
-1.53741062e+00 5.38546085e-01 4.32094961e-01 1.16878383e-01
-3.52820039e-01 7.68596768e-01 1.52388841e-01 -6.99739829e-02
5.45531809e-01 3.38511765e-02 -2.76160538e-01 1.06998444e-01
6.94520295e-01 3.92005086e-01 -2.91473031e-01 -7.83075333e-01
-6.45887911e-01 8.74440789e-01 5.08775786e-02 2.16612577e-01
1.15306532e+00 -6.67136088e-02 -1.60526544e-01 -2.37253457e-01
1.27279043e+00 -5.04652739e-01 -9.18145299e-01 -2.29608208e-01
-2.59301990e-01 -5.20330369e-01 -1.35889679e-01 -6.36566997e-01
-1.36558330e+00 5.87189138e-01 1.24185622e+00 4.20851409e-01
1.19634318e+00 2.83224046e-01 1.17538238e+00 5.91269970e-01
4.24072564e-01 -1.03358436e+00 -4.34380956e-03 1.54699996e-01
9.02692020e-01 -1.73111701e+00 1.02168046e-01 -3.98913711e-01
-2.51347095e-01 8.26241553e-01 5.41753292e-01 -1.39749572e-01
6.98050261e-01 -1.39326587e-01 3.59965935e-02 -4.37598735e-01
-2.64851123e-01 -5.54666698e-01 6.27403438e-01 6.76149964e-01
4.19409662e-01 1.86376616e-01 -2.02188626e-01 4.91958708e-01
1.22634567e-01 1.52102485e-01 -2.01082036e-01 7.21156180e-01
-2.58619338e-01 -8.68117273e-01 -7.12157965e-01 5.92916787e-01
-7.09281147e-01 -1.97339058e-02 -3.97044897e-01 4.74332750e-01
2.54614919e-01 1.14341319e+00 4.97139208e-02 -3.10014069e-01
3.39469969e-01 -5.97843938e-02 4.18180287e-01 -2.78082967e-01
-1.00877059e+00 4.67545390e-02 -1.38430774e-01 -6.65704727e-01
-9.06379104e-01 -6.30619526e-01 -6.53480947e-01 -5.60493588e-01
-1.54625410e-02 -1.68725044e-01 4.28887367e-01 7.05746412e-01
1.93818703e-01 1.44055635e-01 1.39218703e-01 -6.34885490e-01
-3.76611769e-01 -8.28671694e-01 -3.24435949e-01 7.58789957e-01
2.91986048e-01 -5.14738083e-01 -8.58482048e-02 3.15197051e-01] | [14.629817008972168, 0.886760413646698] |
d9122e7d-bd94-44f4-8c48-64c3ac4ed332 | continual-learning-for-neural-semantic | 2010.07865 | null | https://arxiv.org/abs/2010.07865v2 | https://arxiv.org/pdf/2010.07865v2.pdf | Update Frequently, Update Fast: Retraining Semantic Parsing Systems in a Fraction of Time | Currently used semantic parsing systems deployed in voice assistants can require weeks to train. Datasets for these models often receive small and frequent updates, data patches. Each patch requires training a new model. To reduce training time, one can fine-tune the previously trained model on each patch, but naive fine-tuning exhibits catastrophic forgetting - degradation of the model performance on the data not represented in the data patch. In this work, we propose a simple method that alleviates catastrophic forgetting and show that it is possible to match the performance of a model trained from scratch in less than 10% of a time via fine-tuning. The key to achieving this is supersampling and EWC regularization. We demonstrate the effectiveness of our method on multiple splits of the Facebook TOP and SNIPS datasets. | ['Rushin Shah', 'Anna Rumshisky', 'Andrey Simanovsky', 'Rahul Goel', 'Vladislav Lialin'] | 2020-10-15 | null | null | null | null | ['goal-oriented-dialogue-systems'] | ['natural-language-processing'] | [ 1.44191131e-01 6.73833668e-01 -1.37387738e-01 -7.20039248e-01
-9.53924835e-01 -5.12008786e-01 -6.54184595e-02 2.52241138e-02
-4.59773481e-01 8.23928654e-01 2.70616353e-01 -2.71187991e-01
-4.28466685e-02 -6.47152722e-01 -1.08659685e+00 -2.21804827e-01
3.82668763e-01 8.22761059e-01 3.82917613e-01 1.04836576e-01
-1.89546749e-01 2.00127229e-01 -1.53867412e+00 5.46986341e-01
8.54914248e-01 5.57418704e-01 5.24715185e-01 8.08696330e-01
-3.13217849e-01 5.99403203e-01 -6.62885129e-01 -3.46857250e-01
1.77917480e-01 -1.07313819e-01 -1.00980771e+00 1.85489431e-01
7.19038486e-01 -4.25969183e-01 -2.02839121e-01 6.06496394e-01
4.60964441e-01 1.76901907e-01 -5.69277117e-03 -9.66233253e-01
-2.96950132e-01 8.96280229e-01 -1.71348125e-01 8.64619017e-02
-1.67407677e-01 -2.00652793e-01 8.49776864e-01 -5.51542282e-01
6.85213208e-01 1.29926467e+00 1.11132967e+00 9.36032951e-01
-1.46508312e+00 -7.14233100e-01 3.96765083e-01 -3.51462007e-01
-1.08481359e+00 -9.04886186e-01 4.19656396e-01 -1.80197194e-01
1.18536913e+00 6.42714798e-02 5.77560723e-01 1.22397697e+00
-2.32311755e-01 6.10434055e-01 7.29224503e-01 -5.08727849e-01
4.88466293e-01 3.86270493e-01 3.93430650e-01 7.60557652e-01
2.89750785e-01 -1.99581936e-01 -9.25025046e-01 -5.56763828e-01
4.21037227e-01 5.95248118e-03 1.60012200e-01 -2.05640331e-01
-4.87245530e-01 6.58864439e-01 -1.13328546e-02 1.38799071e-01
-2.24596575e-01 1.18356690e-01 3.12333941e-01 3.47130865e-01
5.60324073e-01 3.79636109e-01 -1.18753922e+00 -5.75043201e-01
-1.00742662e+00 1.32764533e-01 1.21241879e+00 1.15628445e+00
9.11077201e-01 -2.83412158e-01 5.40600605e-02 1.26342726e+00
3.48453261e-02 2.73991853e-01 8.27204287e-01 -1.28563654e+00
3.95756483e-01 3.29077601e-01 9.60030127e-04 -3.09824109e-01
-4.34302598e-01 -3.64181578e-01 -2.40276188e-01 -2.85651714e-01
5.43330789e-01 -2.88519174e-01 -1.09334016e+00 1.94691586e+00
5.24087012e-01 2.81428128e-01 -1.39279932e-01 1.98229700e-01
4.48252141e-01 3.76112878e-01 3.57646286e-01 -2.70048559e-01
9.88914013e-01 -1.13152635e+00 -5.90273798e-01 -7.10240066e-01
6.85881913e-01 -5.31139791e-01 1.56741416e+00 3.63060534e-01
-1.03637469e+00 -3.70082378e-01 -8.59622061e-01 -2.51937956e-01
-1.09908842e-01 -7.12672994e-02 8.20095718e-01 7.82269657e-01
-9.88200247e-01 8.30267012e-01 -1.13672614e+00 -7.88369417e-01
5.77352881e-01 3.84666115e-01 -3.10666293e-01 -1.52433440e-01
-6.85645580e-01 6.35407090e-01 1.09699175e-01 -4.29177493e-01
-7.27582633e-01 -1.17058575e+00 -4.53328341e-01 2.41606027e-01
4.42997545e-01 -7.25806475e-01 1.78129637e+00 -7.50867128e-01
-1.53297901e+00 8.07704389e-01 -5.42970300e-01 -5.41003644e-01
4.67940331e-01 -5.55470943e-01 -2.01993302e-01 -2.77995378e-01
1.73567861e-01 5.75285912e-01 1.20897114e+00 -9.87619996e-01
-6.65273666e-01 -4.86843824e-01 -7.40696816e-03 -2.52921525e-02
-5.39291382e-01 -2.22411528e-01 -7.90071428e-01 -4.33458745e-01
2.73953766e-01 -1.02130461e+00 -1.25307962e-01 -2.00292557e-01
-1.35274336e-01 4.23874445e-02 5.84538221e-01 -7.48767257e-01
1.32234108e+00 -2.35050297e+00 -2.45757818e-01 -8.46459419e-02
1.20011859e-01 2.77000099e-01 -2.63190776e-01 1.21269807e-01
2.59392381e-01 2.53932416e-01 -2.48623401e-01 -6.98427796e-01
-2.30854064e-01 7.50152588e-01 -2.41291195e-01 2.48233974e-02
1.15487412e-01 4.57129806e-01 -6.92783415e-01 -1.98927075e-01
-1.72237560e-01 3.27751875e-01 -1.18120027e+00 1.65962562e-01
-4.95342940e-01 1.92177862e-01 -3.26472521e-01 4.04634744e-01
6.74459815e-01 -3.40616375e-01 6.02082908e-01 5.61179519e-02
1.95048615e-01 7.55960584e-01 -1.12848699e+00 1.80655372e+00
-6.46609187e-01 2.39103734e-01 3.17701697e-01 -8.08543980e-01
6.52774453e-01 5.40949889e-02 3.70034993e-01 -5.65465689e-01
-3.28488976e-01 2.83615023e-01 -4.07949239e-01 -4.31510091e-01
3.54379475e-01 -3.09037507e-01 -1.38909817e-02 6.42862558e-01
2.38474235e-01 -9.58258361e-02 1.00761406e-01 1.41109377e-01
1.40305495e+00 -9.33867041e-03 8.39405209e-02 -6.72615021e-02
-2.54074574e-01 1.33550137e-01 9.91327047e-01 1.16697872e+00
-1.56568810e-01 5.50432444e-01 4.25632000e-01 -5.93389988e-01
-1.30642176e+00 -8.32210422e-01 -1.67091414e-01 1.75710094e+00
-4.37770784e-01 -7.16796100e-01 -1.15620446e+00 -9.77312863e-01
3.90663534e-01 7.83401310e-01 -4.65759605e-01 -2.54050195e-01
-6.95788801e-01 -8.80141616e-01 6.57207608e-01 4.81329858e-01
3.05989563e-01 -8.65164638e-01 -3.91408324e-01 5.80319107e-01
-1.65305942e-01 -1.11724877e+00 -3.49754900e-01 5.82188129e-01
-1.35219860e+00 -8.45652699e-01 -8.20330754e-02 -6.74456954e-01
5.70601761e-01 3.53824139e-01 1.44543386e+00 2.90757447e-01
-1.95348278e-01 2.40960762e-01 -3.60381275e-01 -4.40049171e-01
-6.02726638e-01 6.59920096e-01 1.73156574e-01 -3.63783479e-01
6.16818786e-01 -7.99880028e-01 -2.89483592e-02 6.36762753e-02
-6.72708273e-01 -2.69719083e-02 3.18977177e-01 1.13264489e+00
5.26376843e-01 -5.66012226e-02 5.90135455e-01 -1.65730155e+00
5.79730928e-01 -4.03933883e-01 -4.24096406e-01 3.85375798e-01
-9.21318054e-01 2.95365095e-01 4.41431850e-01 -4.71597999e-01
-1.09580779e+00 2.78263599e-01 -2.43509118e-03 -2.38779575e-01
-3.23594771e-02 2.70249426e-01 4.18618917e-02 2.78768897e-01
7.87071109e-01 -3.92850608e-01 -4.56895530e-02 -1.08747637e+00
4.23831701e-01 8.94311428e-01 4.08831924e-01 -8.16788077e-01
4.96267885e-01 3.12931538e-01 -7.78389096e-01 -8.05923283e-01
-1.17516530e+00 -8.18550363e-02 -5.84639192e-01 2.96006292e-01
3.95078927e-01 -1.01126528e+00 -3.53055745e-01 3.53228301e-01
-9.18144047e-01 -7.27509737e-01 -6.66394770e-01 1.29249930e-01
-3.44434083e-01 4.67959009e-02 -4.43317086e-01 -6.34304702e-01
-3.17453831e-01 -6.35250509e-01 1.01079166e+00 2.36104608e-01
-4.46911573e-01 -6.70385420e-01 1.50077254e-01 5.49004972e-01
4.93487686e-01 -5.69117963e-01 9.70931053e-01 -7.30880737e-01
-3.00146759e-01 -1.11214720e-01 2.84476578e-02 3.95516694e-01
1.70902088e-01 1.27941780e-02 -1.38134873e+00 -5.35323501e-01
9.18358043e-02 -3.49430382e-01 9.33432877e-01 3.68755192e-01
1.19766736e+00 -3.48087907e-01 -4.30592597e-01 5.91734469e-01
1.00622714e+00 1.16854608e-01 3.13482612e-01 3.61964315e-01
4.56167042e-01 3.71082366e-01 4.10074115e-01 5.88553786e-01
3.52600425e-01 5.10223687e-01 -1.08611695e-01 1.83923885e-01
-2.94123024e-01 -6.04291201e-01 2.33094618e-01 8.72176528e-01
3.93598348e-01 4.44539152e-02 -7.83961117e-01 6.36972845e-01
-1.88143766e+00 -6.19743884e-01 5.06617486e-01 2.30955791e+00
1.04866815e+00 3.43920946e-01 -7.64919147e-02 -1.46342918e-01
5.64436615e-01 -9.19560269e-02 -8.09425712e-01 -4.26046818e-01
2.64425948e-02 5.55115044e-01 4.99239177e-01 9.24364209e-01
-8.97882760e-01 1.41269410e+00 7.62352371e+00 3.56642574e-01
-8.53747785e-01 5.85818887e-01 5.13906837e-01 -6.68269634e-01
-3.36074531e-01 2.20035404e-01 -1.02862954e+00 4.65317935e-01
1.38796496e+00 3.77090164e-02 9.43137944e-01 1.24958372e+00
1.55850410e-01 -1.08607210e-01 -1.05118477e+00 7.86244690e-01
-1.65234402e-01 -1.28696775e+00 -1.12663344e-01 -1.54758602e-01
6.10876083e-01 3.39972824e-01 -2.19114766e-01 6.09436929e-01
6.17373645e-01 -6.52150273e-01 4.63251531e-01 3.68487626e-01
5.50284266e-01 -5.18922210e-01 3.55150551e-01 4.64532405e-01
-5.71147561e-01 -3.27473521e-01 -6.27183914e-01 3.95098841e-03
-2.88260859e-02 8.24320614e-01 -1.14254737e+00 -1.90786719e-01
1.12471998e+00 2.78401852e-01 -7.29473829e-01 7.34782755e-01
1.17963672e-01 9.80684996e-01 -7.43793726e-01 2.24525258e-01
-2.06745967e-01 -3.54032442e-02 2.32743278e-01 1.01157582e+00
3.91234398e-01 3.32067385e-02 -3.04104704e-02 5.00892937e-01
-2.67031401e-01 -1.04559764e-01 -4.69738573e-01 -3.56307507e-01
9.38164413e-01 7.98176169e-01 -1.19525194e-01 -4.66646880e-01
-4.12587285e-01 1.00016117e+00 1.01709020e+00 2.42634058e-01
-5.60611844e-01 -9.22058374e-02 8.00977468e-01 2.76395977e-01
5.11345506e-01 -1.37520567e-01 -4.57933933e-01 -1.18700516e+00
2.52970278e-01 -1.02547121e+00 5.69333911e-01 -5.90598464e-01
-1.19363010e+00 4.47600394e-01 -4.31536764e-01 -4.69493717e-01
-3.11605215e-01 -3.81310076e-01 -1.64406046e-01 6.96826696e-01
-1.17154860e+00 -8.71657073e-01 -4.19787526e-01 3.38181615e-01
6.66384161e-01 -2.41122618e-01 1.14474857e+00 4.66213316e-01
-7.59639680e-01 9.46745276e-01 3.67180079e-01 -2.90252328e-01
9.10293758e-01 -1.06734359e+00 6.18767083e-01 5.96018553e-01
2.02657312e-01 7.70234466e-01 8.01682472e-01 -7.03202963e-01
-1.19369590e+00 -9.98612702e-01 1.21604455e+00 -5.27128577e-01
4.28271025e-01 -4.23051387e-01 -1.18027508e+00 1.03955531e+00
-3.35801244e-01 -8.96397009e-02 6.10513508e-01 7.44352341e-01
-5.34631371e-01 -4.40776736e-01 -1.37405133e+00 2.92201817e-01
1.24585581e+00 -7.36644506e-01 -4.92723018e-01 4.07982409e-01
8.96291494e-01 -4.83278096e-01 -7.33356059e-01 2.98050076e-01
6.20758593e-01 -8.31193089e-01 6.58140182e-01 -1.04324865e+00
-7.52665251e-02 1.90613329e-01 -3.20529610e-01 -1.19690096e+00
-2.86417454e-01 -9.86360788e-01 -3.34769279e-01 1.31023085e+00
6.47777855e-01 -5.62424064e-01 1.15294981e+00 1.14109433e+00
-2.28100549e-02 -3.48230481e-01 -1.01896453e+00 -6.52206540e-01
8.69066268e-02 -6.76277280e-01 7.33308673e-01 7.56065249e-01
-1.53435078e-02 5.49218178e-01 -3.26433539e-01 2.05184236e-01
4.90875155e-01 -2.43623883e-01 9.29215252e-01 -1.44545066e+00
-5.82029819e-01 3.78434472e-02 1.83478445e-01 -9.69315827e-01
9.28143859e-02 -6.39881253e-01 -6.34282976e-02 -1.29723430e+00
3.58160764e-01 -8.35622072e-01 -2.13551268e-01 8.76456738e-01
-2.99127609e-01 -1.93204015e-01 4.72039618e-02 2.00167015e-01
-3.86386096e-01 2.15755150e-01 7.27625132e-01 3.63456607e-02
-5.88404596e-01 1.16202913e-01 -9.93723631e-01 6.62721634e-01
7.95082390e-01 -1.01899838e+00 -4.47732747e-01 -6.73607647e-01
2.01882049e-01 -2.95025170e-01 -1.42876012e-02 -1.02705765e+00
1.11970343e-01 7.64289200e-02 1.27899319e-01 -2.65231162e-01
3.96167725e-01 -7.25188076e-01 2.97468066e-01 2.06981346e-01
-3.94553304e-01 -1.74627423e-01 4.38511670e-01 7.05712914e-01
1.80784017e-01 -5.33233762e-01 9.59939480e-01 -2.13713154e-01
-4.46200490e-01 7.96204358e-02 -2.39308611e-01 2.30671778e-01
5.08731782e-01 1.83523312e-01 -3.36492300e-01 -1.97064891e-01
-1.16783357e+00 -4.57346104e-02 4.30304617e-01 5.76861143e-01
5.22053577e-02 -9.72446322e-01 -3.27348500e-01 5.38737774e-01
-5.27555384e-02 1.02377445e-01 1.75898448e-01 4.63634729e-01
-2.83590138e-01 2.09706128e-01 1.04099698e-01 -3.99308890e-01
-1.30130184e+00 3.19041550e-01 3.93495739e-01 -2.07201630e-01
-7.41026223e-01 1.06430650e+00 -1.70341417e-01 -9.24143553e-01
5.69934607e-01 -4.07908469e-01 3.80116314e-01 -1.55881993e-04
4.70494330e-01 2.51643986e-01 4.04756874e-01 9.91949886e-02
-2.75263011e-01 2.08141670e-01 -5.74822962e-01 -2.91578472e-01
1.64285743e+00 -1.66022703e-01 4.58445735e-02 5.18585265e-01
9.75036800e-01 7.68243298e-02 -1.58456731e+00 -5.99932015e-01
-1.37305800e-02 -3.85485768e-01 1.57644711e-02 -9.54804063e-01
-7.73503542e-01 6.32916510e-01 5.13502300e-01 2.44076446e-01
8.69136930e-01 9.26691145e-02 1.00876856e+00 8.12042892e-01
5.03124058e-01 -1.64658463e+00 -1.41628489e-01 7.04957426e-01
4.97919142e-01 -1.15655899e+00 -1.63147196e-01 -3.50969762e-01
-5.56548953e-01 8.93767059e-01 5.92791855e-01 1.60994977e-01
7.56710887e-01 4.75656241e-01 6.02976456e-02 -1.19786779e-03
-1.09156799e+00 1.65533170e-01 -3.29039454e-01 6.51966214e-01
3.58759105e-01 2.92209446e-01 1.83469709e-02 1.02984428e+00
-5.00554264e-01 2.16484264e-01 2.92172700e-01 1.14525402e+00
-7.08770096e-01 -1.23942816e+00 -4.30048779e-02 7.28973746e-01
-3.15362483e-01 -2.12011695e-01 -4.46456224e-01 4.98045772e-01
1.95101125e-03 1.02760696e+00 1.40320271e-01 -4.14980829e-01
5.58601081e-01 7.08304107e-01 4.18580741e-01 -8.83523464e-01
-5.64130485e-01 -1.02297924e-02 4.34889674e-01 -7.31272280e-01
1.26686576e-03 -7.35272944e-01 -1.26165485e+00 -5.07901132e-01
-2.10770041e-01 2.30986774e-01 6.41511261e-01 8.55392814e-01
8.46412003e-01 4.43196505e-01 4.94063616e-01 -3.38221669e-01
-8.52727950e-01 -1.19862497e+00 -5.60310364e-01 3.22411656e-01
2.09626228e-01 -5.13990462e-01 -4.94357914e-01 1.11685403e-01] | [10.767254829406738, 8.341317176818848] |
f4178178-cac4-43f3-a142-02b479861cba | graph-based-matrix-completion-applied-to | 2306.08627 | null | https://arxiv.org/abs/2306.08627v1 | https://arxiv.org/pdf/2306.08627v1.pdf | Graph-Based Matrix Completion Applied to Weather Data | Low-rank matrix completion is the task of recovering unknown entries of a matrix by assuming that the true matrix admits a good low-rank approximation. Sometimes additional information about the variables is known, and incorporating this information into a matrix completion model can lead to a better completion quality. We consider the situation where information between the column/row entities of the matrix is available as a weighted graph. In this framework, we address the problem of completing missing entries in air temperature data recorded by weather stations. We construct test sets by holding back data at locations that mimic real-life gaps in weather data. On such test sets, we show that adequate spatial and temporal graphs can significantly improve the accuracy of the completion obtained by graph-regularized low-rank matrix completion methods. | ['Michel Journée', 'P. -A. Absil', 'Benoît Loucheur'] | 2023-06-14 | null | null | null | null | ['low-rank-matrix-completion', 'matrix-completion'] | ['methodology', 'methodology'] | [ 2.26838991e-01 8.06284770e-02 8.11684504e-02 -2.94728223e-02
-7.45902419e-01 -8.75410616e-01 2.80368507e-01 2.30279669e-01
-2.22870573e-01 6.89301252e-01 5.39721787e-01 -3.42321962e-01
-6.02495790e-01 -7.59850562e-01 -9.96809304e-01 -4.88112599e-01
-7.43687332e-01 5.07464349e-01 -3.30396533e-01 -3.65170121e-01
-2.33851328e-01 2.62840480e-01 -9.83767748e-01 1.26198709e-01
8.25063109e-01 3.07512343e-01 2.50929832e-01 9.25426900e-01
1.84741274e-01 5.35132647e-01 -2.61829019e-01 -1.13115855e-01
6.27556264e-01 -4.96520437e-02 -6.05277836e-01 3.44438285e-01
4.26250100e-01 -1.40132412e-01 -7.82945573e-01 9.92441535e-01
-1.05380662e-01 2.79530615e-01 5.33781171e-01 -1.32664931e+00
-1.42500028e-01 5.15932441e-01 -6.89827561e-01 6.88422918e-02
7.31855690e-01 -2.42070287e-01 1.30426204e+00 -1.09814751e+00
5.44680774e-01 9.78124022e-01 9.66196418e-01 -8.24095085e-02
-1.92668569e+00 -3.99711400e-01 2.07889855e-01 -3.25659337e-03
-1.76934218e+00 -5.23414195e-01 7.55309165e-01 -6.78192139e-01
5.20862222e-01 5.53688467e-01 4.67783540e-01 5.72773337e-01
6.01499714e-02 2.07616463e-01 8.45729172e-01 -4.93875369e-02
2.72882581e-01 -1.91340819e-01 2.96001285e-01 4.98986483e-01
8.90945613e-01 9.54551846e-02 -5.63852251e-01 -5.99548101e-01
3.60446066e-01 5.01495719e-01 -5.24511397e-01 -5.24420798e-01
-1.15622818e+00 6.42148137e-01 5.69273591e-01 3.92241068e-02
-5.41820467e-01 -4.73111682e-02 -1.73350513e-01 4.06034797e-01
4.72745776e-01 4.37334329e-01 -2.04645872e-01 2.87635058e-01
-1.00669682e+00 1.38837352e-01 1.07694221e+00 9.09395874e-01
1.21583259e+00 8.10021460e-02 2.09486559e-01 4.23705757e-01
9.26151425e-02 1.03418064e+00 -4.14075881e-01 -8.40458274e-01
1.13160372e+00 5.54979980e-01 4.35917258e-01 -1.48598826e+00
-6.67373657e-01 -3.71661723e-01 -1.15515757e+00 -2.34634399e-01
6.61071539e-01 -4.51495230e-01 -7.61725366e-01 1.78568220e+00
3.51845443e-01 6.12246752e-01 1.30522743e-01 7.69886672e-01
1.85590744e-01 9.21934545e-01 -5.77549398e-01 -4.15744245e-01
9.21078086e-01 -2.96047628e-01 -8.68616462e-01 -4.14264470e-01
5.34725904e-01 -4.05196548e-01 7.00092435e-01 2.37653658e-01
-5.45434654e-01 -2.73536116e-01 -9.97651458e-01 3.42836440e-01
-3.18969756e-01 -1.37831703e-01 5.78390121e-01 5.01437962e-01
-9.50345993e-01 4.85580057e-01 -8.51566732e-01 9.28123519e-02
-2.44950056e-01 3.88419360e-01 -8.69682610e-01 -8.56991649e-01
-1.17627335e+00 4.48752820e-01 -1.96322247e-01 6.12744093e-01
-9.27015245e-01 -8.67421925e-01 -1.07499766e+00 5.66492276e-03
6.89538538e-01 -4.01715577e-01 5.56849122e-01 -2.13661015e-01
-2.65743107e-01 1.97197244e-01 -4.85494107e-01 -2.23129123e-01
1.13379017e-01 -1.12104543e-01 -4.82962817e-01 -2.38953102e-02
2.37126738e-01 -2.36822054e-01 9.76097405e-01 -1.38968134e+00
-1.66257218e-01 -6.51511252e-01 8.80602896e-02 -6.85875788e-02
-4.65928689e-02 -7.30554879e-01 -3.66430670e-01 -4.23395276e-01
3.54802161e-01 -1.15228033e+00 -7.80089855e-01 -1.57790229e-01
-5.91387808e-01 4.75630820e-01 3.12870085e-01 -1.22652304e+00
1.38037193e+00 -2.15664220e+00 4.80635792e-01 1.00414479e+00
5.65269589e-01 -2.89506942e-01 -5.64994037e-01 1.19192421e+00
-2.25658268e-01 -1.43548176e-02 -7.31444716e-01 -4.15630400e-01
-2.13184685e-01 8.18205178e-01 -3.62680167e-01 8.69103968e-01
-8.28822106e-02 3.02140445e-01 -1.00183105e+00 1.16688125e-01
-1.45289544e-02 2.59059072e-01 -5.17229140e-01 3.43525894e-02
1.36100009e-01 3.56555641e-01 -4.20916468e-01 2.90222466e-01
9.27567005e-01 -2.50688493e-01 6.21478260e-01 -1.74454048e-01
3.78220528e-02 3.49679627e-02 -1.81129432e+00 1.59666824e+00
-4.39367324e-01 3.93834710e-01 7.25447714e-01 -7.97160208e-01
4.73695964e-01 2.94703841e-01 8.08854342e-01 -4.13940489e-01
-5.18428564e-01 -1.46106109e-01 -1.06516946e-02 -1.90240651e-01
6.50702655e-01 2.12110221e-01 3.59105617e-02 4.55063850e-01
-4.98510867e-01 -1.65068641e-01 4.32790786e-01 7.85388887e-01
1.40250421e+00 -4.35096145e-01 1.01203471e-01 -2.72355199e-01
2.29464695e-01 1.68178737e-01 7.85259187e-01 6.79347873e-01
4.08527046e-01 5.44313729e-01 7.37100482e-01 -2.82132119e-01
-1.16514671e+00 -9.45605993e-01 -2.76599219e-03 5.73430777e-01
-3.15087199e-01 -8.57970715e-01 -3.36982548e-01 -2.10214466e-01
2.44307771e-01 2.71622807e-01 -9.58522141e-01 1.36861578e-01
-2.47220173e-01 -1.01887357e+00 9.28041935e-02 1.83028370e-01
-2.75840666e-02 1.64056316e-01 2.96258718e-01 1.83877900e-01
-4.60909158e-01 -1.18813074e+00 -6.11785829e-01 1.92823589e-01
-1.15043402e+00 -1.44895101e+00 -4.05460566e-01 -2.97019422e-01
1.41660261e+00 7.03159451e-01 1.10926342e+00 1.41551256e-01
-1.74604401e-01 5.50421953e-01 -2.44781002e-01 2.07320198e-01
-1.43739590e-02 -1.95951909e-01 2.86020249e-01 4.01771754e-01
-2.89435029e-01 -7.45868623e-01 -3.76509666e-01 1.76726073e-01
-1.19774938e+00 -1.35360464e-01 3.78495246e-01 8.04467618e-01
6.24505758e-01 2.15432942e-01 7.33385459e-02 -1.04936969e+00
6.34574473e-01 -8.40665042e-01 -8.37275803e-01 3.11855435e-01
-5.67078233e-01 3.18599790e-01 5.98115683e-01 -1.87794402e-01
-4.25340533e-01 7.21804917e-01 5.14963865e-01 -3.28872383e-01
5.60864985e-01 1.11368191e+00 -6.47327527e-02 -9.06597376e-02
7.24437773e-01 7.27810934e-02 -1.00254565e-01 -6.54997766e-01
5.27575493e-01 4.51406315e-02 4.59731996e-01 -6.70134723e-01
1.38575661e+00 5.45280576e-01 5.74238360e-01 -1.03824353e+00
-7.22575247e-01 -7.05017507e-01 -7.71512806e-01 -2.71776374e-02
1.78584263e-01 -1.38716447e+00 -4.05913651e-01 -1.03386991e-01
-7.27593243e-01 -3.41803163e-01 -2.12545514e-01 6.53667212e-01
1.39854653e-02 5.02519906e-01 -3.51356030e-01 -8.22227895e-01
3.28685403e-01 -6.14677250e-01 8.62954676e-01 -5.40326774e-01
-1.97344944e-02 -1.20532382e+00 7.22402573e-01 9.93777588e-02
1.05354495e-01 3.77698779e-01 6.87307358e-01 -3.63314509e-01
-6.28019750e-01 -5.14464617e-01 -4.13229428e-02 -4.67281006e-02
2.65794009e-01 -1.54672265e-01 -7.24003553e-01 -6.24664009e-01
-2.62136430e-01 2.38227203e-01 8.27818751e-01 3.28746140e-01
6.80942237e-01 -7.04970121e-01 -3.18816036e-01 5.82289279e-01
1.52303576e+00 -4.96877998e-01 3.92322719e-01 -4.15179640e-01
9.95973051e-01 7.61021852e-01 3.34691197e-01 8.68415236e-01
6.98280811e-01 5.09027481e-01 4.25321996e-01 -2.70871460e-01
1.92273617e-01 -4.62588727e-01 2.05694169e-01 9.54384565e-01
5.55622252e-03 -3.17021683e-02 -1.18823564e+00 8.40105712e-01
-2.03270364e+00 -9.16399896e-01 -5.57576656e-01 2.59602261e+00
4.94660020e-01 -3.45018595e-01 4.91621047e-02 2.29735538e-01
5.98147333e-01 4.28270608e-01 -4.85908300e-01 2.82662243e-01
-2.71648318e-01 8.73870682e-04 1.04724813e+00 1.22278154e+00
-8.24519694e-01 1.62942305e-01 7.49759197e+00 1.90572426e-01
-3.15527469e-01 -1.76481873e-01 1.60057247e-01 2.28253584e-02
-7.87311018e-01 3.10296506e-01 -4.00226682e-01 5.36133582e-03
1.04184127e+00 -2.39306614e-01 1.16466284e+00 3.43629956e-01
5.06977797e-01 -1.67366371e-01 -1.24491966e+00 8.12299252e-01
-1.36841647e-02 -1.06341410e+00 -1.99762374e-01 5.90098023e-01
1.14420569e+00 1.15024082e-01 -8.61845985e-02 6.79058675e-03
8.32261741e-01 -9.05449212e-01 -1.16939666e-02 7.38901734e-01
7.01961756e-01 -6.44806027e-01 5.01078665e-01 3.35793495e-01
-1.50464821e+00 3.89620252e-02 -5.71322262e-01 -4.57082689e-01
1.21773086e-01 1.11673248e+00 -8.59447896e-01 8.28547537e-01
3.20653200e-01 1.06190884e+00 -5.51047087e-01 1.04732859e+00
-3.88742425e-03 7.82982051e-01 -7.38178074e-01 5.27098298e-01
1.80550907e-02 -8.92076552e-01 6.35891974e-01 6.60230219e-01
2.43481427e-01 3.59744936e-01 4.94207054e-01 5.48738837e-01
5.33183990e-03 2.22891048e-02 -1.31049895e+00 -7.03741685e-02
2.30442166e-01 1.18912423e+00 -3.05714786e-01 -8.78658295e-02
-7.22936392e-01 6.52979374e-01 2.85068154e-01 9.33883071e-01
-2.89016277e-01 -3.47402357e-02 9.98379290e-01 8.97581279e-02
1.26425534e-01 -8.41985703e-01 -1.90047294e-01 -1.34085417e+00
1.98848099e-01 -8.54980290e-01 4.79004592e-01 -5.81554711e-01
-1.30682409e+00 1.90769091e-01 2.21942309e-02 -1.21639800e+00
-3.43984544e-01 -3.72555345e-01 -4.49560970e-01 9.01447654e-01
-1.05990922e+00 -7.88141012e-01 -2.72958159e-01 8.32873285e-01
-1.09074779e-01 2.56274670e-01 6.71324968e-01 4.02443171e-01
-6.97670043e-01 8.08303524e-03 6.22227371e-01 1.48500070e-01
3.20838571e-01 -1.21742105e+00 5.34964323e-01 1.43868279e+00
4.43584830e-01 7.96384096e-01 8.44968140e-01 -8.75936508e-01
-2.21202350e+00 -1.06951427e+00 7.92730749e-01 -5.40164590e-01
9.82535243e-01 -6.33923113e-01 -9.33301866e-01 9.42568779e-01
-2.74693668e-01 2.14048356e-01 7.64408648e-01 5.12725830e-01
-4.22590971e-01 -3.22010070e-01 -8.37080717e-01 5.73713601e-01
9.14312363e-01 -9.57960665e-01 -2.37077713e-01 7.48213768e-01
5.60344458e-01 -1.80642024e-01 -8.64492357e-01 1.20299228e-01
2.13378131e-01 -2.56414592e-01 1.05207586e+00 -8.31092179e-01
-9.47242305e-02 -7.83323109e-01 -6.27134264e-01 -1.64938188e+00
-4.52321142e-01 -7.87412047e-01 -2.90686395e-02 7.91110694e-01
8.00533593e-01 -5.04968882e-01 7.87291646e-01 1.20104337e+00
1.29417107e-01 -1.31678116e-02 -8.48677516e-01 -7.08755851e-01
-2.46127024e-01 -6.31949961e-01 5.72967291e-01 1.03998125e+00
1.37532473e-01 3.92829359e-01 -1.05142224e+00 8.60487521e-01
8.27692449e-01 5.41501977e-02 9.82461512e-01 -1.40888798e+00
-3.91510487e-01 4.47388083e-01 -9.43633094e-02 -1.03578901e+00
1.30684540e-01 -7.86142945e-01 -7.93749541e-02 -1.65349925e+00
1.57888219e-01 -5.08014321e-01 1.02343909e-01 5.08655846e-01
-1.64230466e-01 6.75350875e-02 1.55678103e-02 1.02602609e-01
-4.10675257e-01 3.82163793e-01 1.04762173e+00 -3.40447277e-01
-2.34512284e-01 9.25386474e-02 -5.35519183e-01 2.92605340e-01
4.03559387e-01 -7.53635526e-01 -6.17486537e-01 -4.41257536e-01
9.01833057e-01 7.86671937e-01 1.64346695e-01 -7.77548313e-01
5.23355663e-01 -3.36322576e-01 1.51868418e-01 -6.52550757e-01
5.84358692e-01 -1.38265419e+00 8.13166440e-01 4.75167215e-01
-1.39807612e-01 5.59111297e-01 1.70246258e-01 1.19768476e+00
-2.01641336e-01 -6.58345446e-02 -4.74065877e-02 1.05047584e-01
-3.18721861e-01 5.95298648e-01 -5.06760359e-01 2.21085578e-01
5.75351000e-01 1.50803309e-02 -2.52932519e-01 -8.29512537e-01
-1.09560585e+00 5.76995373e-01 4.56926316e-01 6.77293465e-02
6.98917449e-01 -1.23933673e+00 -9.72552299e-01 3.13853979e-01
2.78992891e-01 -8.21042135e-02 3.63222063e-01 8.83696198e-01
-2.55881518e-01 2.57394284e-01 2.73076504e-01 -3.32121253e-01
-1.11977506e+00 6.94000840e-01 1.42475665e-01 -4.19340849e-01
-4.24516946e-01 1.90018147e-01 2.19925344e-01 -4.98586982e-01
1.98347475e-02 -3.94237399e-01 5.76106720e-02 1.13306835e-01
6.22276783e-01 5.83535731e-01 3.91508155e-02 -6.56648815e-01
-2.80240387e-01 4.99126881e-01 4.07540560e-01 -3.58466804e-01
1.49905491e+00 -4.82820481e-01 -2.89428949e-01 5.29874027e-01
9.77239251e-01 6.42949760e-01 -1.21970856e+00 -5.84616184e-01
-7.19765648e-02 -6.68081403e-01 -4.89057489e-02 -1.95377812e-01
-1.22708011e+00 5.73406279e-01 8.12084302e-02 2.09591404e-01
9.56889749e-01 -4.27151173e-01 2.64931768e-01 9.41375911e-01
3.39025557e-01 -6.99002862e-01 -3.99819136e-01 4.61528629e-01
9.94331837e-01 -1.14381409e+00 4.48782086e-01 -5.37184656e-01
-4.60959285e-01 7.55101025e-01 2.37104390e-02 -2.25965872e-01
1.11526823e+00 7.71061182e-02 -1.89897329e-01 -2.90808886e-01
-9.10046160e-01 -4.56516802e-01 5.22303224e-01 7.78808892e-01
-3.32961641e-02 5.06131887e-01 1.14822187e-01 1.33904904e-01
-1.38091817e-01 -6.40487850e-01 8.58807743e-01 5.59116483e-01
-1.28614694e-01 -1.06700301e+00 -8.28856707e-01 6.69444025e-01
-1.24532513e-01 -3.33144188e-01 -4.22279567e-01 5.18570065e-01
-5.04798412e-01 1.35441816e+00 -1.29898965e-01 -5.94986796e-01
3.99185151e-01 -2.45368019e-01 6.72147572e-02 -5.27922750e-01
-7.95509517e-02 -1.17873020e-01 3.98779690e-01 -6.95499182e-01
-2.74849087e-01 -8.36724997e-01 -8.53474498e-01 -6.38332307e-01
7.76801035e-02 5.98136425e-01 5.96009672e-01 7.54008114e-01
3.44615847e-01 3.91527355e-01 9.59119439e-01 -7.52643824e-01
-3.11167717e-01 -6.97020888e-01 -1.04676008e+00 4.00634050e-01
9.00350213e-01 -3.68114948e-01 -6.30362988e-01 -9.73129459e-03] | [6.911013603210449, 4.730090618133545] |
b8f5b3c0-ac9f-4388-840d-c009e15fabf3 | multi-level-protein-representation-learning | 2306.04899 | null | https://arxiv.org/abs/2306.04899v1 | https://arxiv.org/pdf/2306.04899v1.pdf | Multi-level Protein Representation Learning for Blind Mutational Effect Prediction | Directed evolution plays an indispensable role in protein engineering that revises existing protein sequences to attain new or enhanced functions. Accurately predicting the effects of protein variants necessitates an in-depth understanding of protein structure and function. Although large self-supervised language models have demonstrated remarkable performance in zero-shot inference using only protein sequences, these models inherently do not interpret the spatial characteristics of protein structures, which are crucial for comprehending protein folding stability and internal molecular interactions. This paper introduces a novel pre-training framework that cascades sequential and geometric analyzers for protein primary and tertiary structures. It guides mutational directions toward desired traits by simulating natural selection on wild-type proteins and evaluates the effects of variants based on their fitness to perform the function. We assess the proposed approach using a public database and two new databases for a variety of variant effect prediction tasks, which encompass a diverse set of proteins and assays from different taxa. The prediction results achieve state-of-the-art performance over other zero-shot learning methods for both single-site mutations and deep mutations. | ['Liang Hong', 'Yu Guang Wang', 'Yuanhong Jiang', 'Bingxin Zhou', 'Yang Tan'] | 2023-06-08 | null | null | null | null | ['protein-folding'] | ['natural-language-processing'] | [ 5.95774233e-01 -6.66935071e-02 3.84828672e-02 -4.06527430e-01
-3.63068342e-01 -6.41168654e-01 2.53968835e-01 4.81849074e-01
-3.70750070e-01 1.29410338e+00 -1.48246408e-01 -6.24378145e-01
-6.92902878e-02 -5.69461703e-01 -1.01955271e+00 -1.00748503e+00
-4.46545929e-02 7.64376104e-01 4.46368694e-01 -5.65589666e-01
4.16413724e-01 4.11545336e-01 -1.69711173e+00 3.97345036e-01
1.06360686e+00 2.25233018e-01 5.96976578e-01 8.21220100e-01
-3.05883855e-01 3.58401388e-01 -5.06343007e-01 -4.81283665e-01
-2.44466346e-02 -6.60618663e-01 -4.06773001e-01 -2.31200352e-01
-5.74102551e-02 2.28608191e-01 1.27270162e-01 8.17225575e-01
8.46544445e-01 -1.07424147e-03 8.36063623e-01 -3.80835384e-01
-1.01322627e+00 1.78795010e-01 -3.28502178e-01 3.72059345e-01
5.36054313e-01 8.21432948e-01 1.18835807e+00 -8.40733349e-01
9.15028155e-01 1.00780332e+00 5.87704957e-01 3.69581282e-01
-1.52586019e+00 -1.00562096e-01 -1.89471543e-01 4.93061572e-01
-9.25956964e-01 -1.56556338e-01 4.30047303e-01 -4.92584050e-01
1.70669734e+00 7.14011118e-02 6.38431370e-01 9.92957652e-01
7.21763372e-01 2.82845974e-01 6.28359914e-01 -3.62332523e-01
5.29057920e-01 -4.09716457e-01 1.80023640e-01 1.00230587e+00
1.83708593e-01 1.96698070e-01 -8.62336457e-01 -7.11270452e-01
-6.19641393e-02 1.45080268e-01 -1.83212057e-01 -6.65186882e-01
-9.17664886e-01 6.59525275e-01 -7.25942571e-03 -5.35102524e-02
-5.96316040e-01 -4.10115719e-01 3.56216878e-01 2.63186514e-01
2.42005035e-01 8.09969664e-01 -1.11328781e+00 -3.40083569e-01
-4.96119499e-01 3.10499579e-01 6.16642952e-01 5.47034383e-01
8.79895627e-01 -2.18664557e-01 -6.56145364e-02 7.86973715e-01
3.59003805e-02 2.76737303e-01 5.37847638e-01 -1.65789306e-01
-1.65460944e-01 7.68285930e-01 2.28562996e-01 -5.48406601e-01
-4.43781018e-01 -1.40866369e-01 -3.68077755e-01 3.45844537e-01
3.18027765e-01 -8.83038118e-02 -9.25166547e-01 1.62535489e+00
4.35476571e-01 -7.42330626e-02 1.58360243e-01 6.72960401e-01
5.65856278e-01 6.74449980e-01 2.87180841e-01 -5.54255843e-01
1.24374878e+00 -5.32936633e-01 -5.01553416e-01 1.78393975e-01
6.54244125e-01 -6.30968213e-01 1.18113744e+00 3.53311419e-01
-8.02439392e-01 -5.22262931e-01 -1.36595464e+00 1.25945717e-01
-5.64101100e-01 -2.11231038e-01 6.86062276e-01 5.33851326e-01
-4.62836087e-01 1.02675939e+00 -7.54772484e-01 -4.61328477e-01
2.94270217e-01 3.43057483e-01 -1.73003599e-01 2.87131965e-01
-1.33765113e+00 1.13840461e+00 5.45656741e-01 -4.17073488e-01
-9.04032469e-01 -1.12723613e+00 -4.51984942e-01 3.77074033e-02
2.81294733e-01 -5.11168480e-01 8.59728813e-01 -6.44914627e-01
-1.65323925e+00 8.54575038e-01 -3.97453517e-01 -5.91026425e-01
1.25838056e-01 -9.38617736e-02 -2.09508836e-01 -2.49490872e-01
-3.59537542e-01 2.99402297e-01 5.51886141e-01 -4.76961195e-01
-3.98754954e-01 -6.38707161e-01 -2.10965395e-01 1.63390428e-01
1.24644987e-01 -7.31006712e-02 2.22533658e-01 -5.69506347e-01
-2.57573187e-01 -9.08205152e-01 -5.22902071e-01 6.95676357e-02
-1.16426185e-01 -1.13728136e-01 6.85790300e-01 -5.20616293e-01
9.46524322e-01 -1.58563173e+00 6.55256987e-01 -4.97861467e-02
1.43473819e-01 7.09362328e-01 -5.34111857e-02 8.39053810e-01
-2.92875916e-01 -2.40603745e-01 -5.19575596e-01 5.88747740e-01
-3.12561750e-01 1.03774369e-01 1.52407348e-01 5.48345864e-01
2.54613072e-01 9.63512897e-01 -7.29197919e-01 -1.20373644e-01
4.14300054e-01 4.16384220e-01 -7.05096245e-01 2.84319341e-01
-8.46566617e-01 4.33285087e-01 -4.67746198e-01 6.35166705e-01
4.74025935e-01 -3.66272479e-01 7.19623685e-01 -1.34488553e-01
-1.61614373e-01 2.86692027e-02 -4.15884167e-01 1.76062715e+00
3.15548219e-02 9.48345810e-02 -5.94889522e-01 -1.00730217e+00
1.05639517e+00 9.18162689e-02 5.27558446e-01 -4.28575993e-01
-8.37749466e-02 -1.38654802e-02 4.94877964e-01 -6.41306877e-01
2.23180791e-03 -3.02221954e-01 2.38464057e-01 3.43144268e-01
1.30104005e-01 3.13373953e-01 3.59762877e-01 -6.70365319e-02
1.29143178e+00 6.19955301e-01 7.43957102e-01 -2.51321167e-01
5.95367432e-01 4.19504285e-01 7.96798706e-01 4.56188977e-01
-1.77030325e-01 2.75465816e-01 5.51753700e-01 -8.24793041e-01
-1.54839945e+00 -9.57045078e-01 -1.03092395e-01 1.55867994e+00
-6.92291632e-02 -4.04757738e-01 -7.12829709e-01 -5.28984487e-01
1.44066781e-01 7.92639434e-01 -5.46591103e-01 -6.12126529e-01
-4.14074272e-01 -1.39034855e+00 5.93601704e-01 1.90965366e-02
-8.24266449e-02 -1.18901277e+00 -8.44452322e-01 4.21983570e-01
1.05167083e-01 -3.78223687e-01 -2.15474695e-01 6.75111175e-01
-5.67605376e-01 -1.55295265e+00 -6.32084846e-01 -7.94661224e-01
5.06563604e-01 -1.63380444e-01 9.51447248e-01 -1.02030456e-01
-9.47191000e-01 -6.00582480e-01 -3.37732106e-01 -5.28082371e-01
-6.77710712e-01 -1.25654668e-01 1.89333946e-01 -3.21203947e-01
9.46996033e-01 -7.48510540e-01 -4.13593113e-01 3.93276900e-01
-6.85520053e-01 2.85642128e-02 7.60170519e-01 1.23943818e+00
7.67505825e-01 -1.93718970e-01 5.52810907e-01 -1.11766243e+00
7.63144255e-01 -3.48341733e-01 -6.43941581e-01 8.63515437e-01
-7.05616772e-01 8.37343931e-01 8.52756202e-01 -3.55572104e-01
-1.35905612e+00 4.16424632e-01 -3.44696254e-01 1.60418332e-01
-1.76090315e-01 3.89792681e-01 -3.03542703e-01 7.57583678e-02
9.51112151e-01 6.51931465e-01 3.24147284e-01 -5.85527241e-01
4.13280100e-01 4.74191308e-01 2.94547766e-01 -6.76641703e-01
2.82529891e-01 1.10042460e-01 1.14538178e-01 -9.24755156e-01
-3.68913025e-01 -2.96284676e-01 -9.49486554e-01 1.45130351e-01
7.54140794e-01 -3.48266572e-01 -1.08910835e+00 4.73983973e-01
-9.14596617e-01 -1.91748235e-02 7.55654573e-02 1.09475762e-01
-6.71676934e-01 7.12404668e-01 -5.50862730e-01 -4.64049637e-01
-5.40963411e-01 -1.29797769e+00 9.65483546e-01 1.67642068e-02
-2.42893040e-01 -7.15825677e-01 4.87111628e-01 1.65071070e-01
1.58498690e-01 1.03789628e-01 1.70081270e+00 -7.59655535e-01
-2.65877306e-01 2.09790602e-01 3.21483999e-01 -1.05108149e-01
3.28254491e-01 2.84527212e-01 -3.54584098e-01 -1.61382124e-01
-4.29743379e-01 -5.33326268e-01 8.95791054e-01 2.69836158e-01
7.74534941e-01 7.98878297e-02 -3.89214426e-01 5.95954359e-01
1.43426156e+00 5.79850733e-01 7.54149020e-01 2.94919997e-01
4.38476443e-01 4.29538190e-01 8.91729951e-01 7.95894086e-01
-3.83005679e-01 7.40250409e-01 3.78880650e-01 1.37887150e-01
3.30687821e-01 -2.76375592e-01 2.46306464e-01 3.72667387e-02
-2.73200214e-01 -3.63109052e-01 -8.70318472e-01 1.19640134e-01
-1.79955208e+00 -1.05488074e+00 8.34710710e-03 2.04516864e+00
1.21077192e+00 1.07059300e-01 5.77354878e-02 -3.22621912e-01
7.05928683e-01 3.28179412e-02 -1.43897855e+00 -4.79700953e-01
-4.03756678e-01 4.73910004e-01 5.32832801e-01 3.40603143e-01
-7.72897363e-01 1.17816472e+00 7.08499098e+00 6.37008905e-01
-8.70723426e-01 -3.05543840e-01 5.03710747e-01 -1.34505913e-01
-1.92803696e-01 1.11044966e-01 -8.59199405e-01 4.66161370e-01
1.14563942e+00 -2.45302260e-01 3.18239480e-01 7.08490551e-01
4.04930651e-01 1.25810653e-01 -9.74574447e-01 5.78616023e-01
-1.24298751e-01 -1.86704695e+00 1.04285978e-01 -1.33934945e-01
4.73821282e-01 5.03761582e-02 -2.24797219e-01 2.97611672e-03
4.71156538e-01 -1.02004755e+00 7.11356103e-02 7.22300529e-01
6.95751250e-01 -9.63751495e-01 5.21840692e-01 4.35325980e-01
-6.93957388e-01 1.25629619e-01 -8.52196336e-01 -1.46460518e-01
1.98387533e-01 5.67405939e-01 -1.03109825e+00 1.16579071e-01
3.39170039e-01 6.67351007e-01 -4.53819573e-01 4.57004368e-01
-6.14521615e-02 4.24457431e-01 -3.48914415e-02 -6.17003858e-01
-1.60562351e-01 -4.86102730e-01 5.24533033e-01 9.80252564e-01
-8.59201327e-02 3.57257456e-01 -2.94699445e-02 8.26307178e-01
1.54872075e-01 3.06702971e-01 -4.31440115e-01 -3.72363985e-01
3.60951662e-01 8.40268254e-01 -3.33831608e-01 -3.48543882e-01
-4.97012466e-01 1.17954957e+00 6.24374390e-01 2.11171210e-01
-1.09833336e+00 -5.14178336e-01 1.46527016e+00 1.48878783e-01
3.67520571e-01 2.13306863e-02 2.63101548e-01 -8.66050363e-01
-1.74449816e-01 -1.21681857e+00 3.21176827e-01 -5.62159777e-01
-1.26855755e+00 1.38762459e-01 -3.95952582e-01 -7.37703741e-01
-1.57586455e-01 -1.00140667e+00 -3.71151656e-01 8.64425540e-01
-9.85517800e-01 -7.17543304e-01 2.31075898e-01 1.44455591e-02
9.27575409e-01 -5.47444224e-01 1.09716415e+00 -5.17506376e-02
-6.89149141e-01 4.47584540e-01 6.02154613e-01 -6.29652202e-01
8.27400088e-01 -1.06968641e+00 9.07045066e-01 6.00537121e-01
-2.35309124e-01 8.81511569e-01 1.18484139e+00 -1.09345520e+00
-1.66306531e+00 -8.71869028e-01 4.59941924e-01 -2.86520541e-01
4.09676194e-01 -1.90273106e-01 -1.07274783e+00 4.09766763e-01
-1.40285388e-01 -2.96400487e-01 1.07590365e+00 1.15778521e-01
-1.51997954e-01 3.81317854e-01 -1.19042349e+00 5.81940413e-01
1.34206998e+00 -2.12250084e-01 -7.32778072e-01 3.76052171e-01
6.92478001e-01 1.51594705e-03 -7.41416276e-01 4.26230669e-01
6.97712421e-01 -1.13654709e+00 1.08170652e+00 -1.62113023e+00
3.57656926e-01 -4.59047765e-01 -1.97839245e-01 -1.31484318e+00
-6.94562197e-01 -4.89673793e-01 -1.33306712e-01 6.10333681e-01
8.14750016e-01 -5.12558043e-01 1.01885772e+00 2.01175869e-01
-1.56239972e-01 -9.28869486e-01 -4.96186078e-01 -4.85295057e-01
-9.57351923e-02 1.37722716e-01 5.99027872e-01 6.71670556e-01
2.54287452e-01 5.71835697e-01 -5.14266610e-01 -5.81454597e-02
5.42744160e-01 7.18450844e-02 6.18479729e-01 -1.13575661e+00
-6.70782447e-01 -1.80275902e-01 -6.60487592e-01 -8.51735413e-01
1.12426341e-01 -7.30983853e-01 -5.91889955e-02 -1.12669706e+00
4.99576598e-01 3.18617314e-01 -2.83867657e-01 3.16544384e-01
-4.33971465e-01 -3.47825974e-01 -6.74971104e-01 2.09788829e-02
-4.69239652e-01 6.64477706e-01 7.94946730e-01 -1.05474308e-01
-1.91031799e-01 -1.18776739e-01 -4.83994037e-01 6.14113927e-01
7.69445658e-01 -2.17766717e-01 -4.99977618e-01 2.20550403e-01
2.84345359e-01 -1.26799211e-01 -1.00302331e-01 -7.02003777e-01
-2.13411808e-01 -5.36324859e-01 4.83512580e-01 -5.56598544e-01
2.18260214e-01 -3.11594278e-01 3.94139886e-01 9.60727930e-01
-5.58397830e-01 1.73934132e-01 -6.04222156e-02 9.03513789e-01
1.96186498e-01 2.26997398e-02 1.16213310e+00 -4.14995104e-01
-9.04854476e-01 2.61143029e-01 -6.05752230e-01 -1.44109324e-01
1.29933560e+00 -3.43990296e-01 -2.02276170e-01 1.43005192e-01
-9.36645627e-01 -2.41027653e-01 8.47815037e-01 4.61595893e-01
7.72142112e-01 -5.65117657e-01 -6.79426014e-01 4.65284646e-01
4.58201319e-01 -7.63972521e-01 1.52052552e-01 4.36796725e-01
-1.10511863e+00 6.17988586e-01 -4.45940435e-01 -4.93426293e-01
-1.77580798e+00 8.94103289e-01 4.82635617e-01 -8.54070187e-02
-5.18854678e-01 8.46556425e-01 1.03974700e-01 -5.45355499e-01
-8.22596923e-02 6.40216991e-02 5.29398816e-03 -4.24769014e-01
5.51008344e-01 2.37840414e-01 1.25633642e-01 -4.06969130e-01
-3.62725943e-01 4.25987750e-01 -3.67777139e-01 5.64598322e-01
1.50415516e+00 5.37663288e-02 -1.63072452e-01 3.76216769e-01
9.31342125e-01 -4.34487492e-01 -1.35099721e+00 -2.29166180e-01
3.23893875e-01 -3.93198073e-01 -5.17774880e-01 -1.07933581e+00
-2.72641391e-01 5.51368594e-01 8.01735699e-01 -6.90950036e-01
7.86841273e-01 2.56171469e-02 7.61889815e-01 9.49784696e-01
7.49061644e-01 -1.08573639e+00 3.54055353e-02 5.58282137e-01
2.81009167e-01 -1.26550317e+00 8.58143941e-02 -1.82354882e-01
-6.15330458e-01 8.95669103e-01 4.60459620e-01 2.35176444e-01
3.64292413e-01 1.51878119e-01 -1.11973695e-01 -2.55173266e-01
-1.16426635e+00 7.15108030e-03 3.30933295e-02 6.87097073e-01
8.77381444e-01 1.06303148e-01 -7.73346841e-01 3.77183139e-01
1.21715471e-01 -1.23849563e-01 1.55265126e-02 1.12285697e+00
-1.01730752e+00 -1.52067363e+00 -8.46645012e-02 4.45940644e-01
-5.02894878e-01 -2.52405792e-01 -9.88240004e-01 3.00969124e-01
9.64155495e-02 6.22459292e-01 -4.26058739e-01 -2.65179843e-01
2.25104883e-01 5.44607401e-01 6.43256068e-01 -5.35425007e-01
-3.16229999e-01 -3.25270802e-01 1.54103965e-01 -2.55550534e-01
9.24941059e-03 -6.80061579e-01 -1.61731339e+00 -2.95647085e-01
-4.37153250e-01 3.15024674e-01 5.34354270e-01 8.06728423e-01
8.88343215e-01 6.02634907e-01 2.78300434e-01 -3.94066155e-01
-7.43081510e-01 -9.45092916e-01 -5.04069030e-01 5.77581704e-01
-1.14866696e-01 -6.73999667e-01 6.65376708e-02 2.77560681e-01] | [4.746790885925293, 5.588354587554932] |
7d3bafa7-9198-4bb6-bba5-7b74d3cdc300 | artificial-and-beneficial-exploiting | 2104.03054 | null | https://arxiv.org/abs/2104.03054v1 | https://arxiv.org/pdf/2104.03054v1.pdf | Artificial and beneficial -- Exploiting artificial images for aerial vehicle detection | Object detection in aerial images is an important task in environmental, economic, and infrastructure-related tasks. One of the most prominent applications is the detection of vehicles, for which deep learning approaches are increasingly used. A major challenge in such approaches is the limited amount of data that arises, for example, when more specialized and rarer vehicles such as agricultural machinery or construction vehicles are to be detected. This lack of data contrasts with the enormous data hunger of deep learning methods in general and object recognition in particular. In this article, we address this issue in the context of the detection of road vehicles in aerial images. To overcome the lack of annotated data, we propose a generative approach that generates top-down images by overlaying artificial vehicles created from 2D CAD drawings on artificial or real backgrounds. Our experiments with a modified RetinaNet object detection network show that adding these images to small real-world datasets significantly improves detection performance. In cases of very limited or even no real-world images, we observe an improvement in average precision of up to 0.70 points. We address the remaining performance gap to real-world datasets by analyzing the effect of the image composition of background and objects and give insights into the importance of background. | ['Ribana Roscher', 'Jens Bongartz', 'Immanuel Weber'] | 2021-04-07 | null | null | null | null | ['object-detection-in-aerial-images'] | ['computer-vision'] | [ 5.37228167e-01 -1.36856541e-01 2.87263244e-01 -4.96510416e-02
-2.33706594e-01 -5.58175683e-01 6.24113262e-01 -1.11986473e-02
-5.83021820e-01 5.72711110e-01 -2.99537152e-01 -4.03486520e-01
8.66211280e-02 -1.00428975e+00 -8.08318675e-01 -5.65816343e-01
1.98700745e-02 1.49143398e-01 6.78376973e-01 -2.31418237e-01
4.25779670e-02 7.05255091e-01 -1.95910954e+00 1.60618231e-01
6.44193411e-01 8.65818799e-01 5.26628196e-01 7.07400858e-01
-1.17544634e-02 5.00936031e-01 -6.98499739e-01 -4.65591222e-01
7.01088548e-01 -1.65151179e-01 -4.20576751e-01 4.61069673e-01
8.53956819e-01 -5.32786071e-01 -2.26510912e-01 1.17177045e+00
3.87942374e-01 -1.54942289e-01 6.01181149e-01 -1.18982792e+00
-4.01425242e-01 2.02129006e-01 -8.97486627e-01 3.84103030e-01
2.11309642e-02 3.69422406e-01 6.23901010e-01 -7.50072420e-01
5.35950720e-01 1.16526568e+00 6.31161809e-01 2.79028267e-01
-1.26627982e+00 -5.31110287e-01 1.69586360e-01 3.44851352e-02
-1.38507950e+00 -3.32490563e-01 5.24995625e-01 -6.05593026e-01
5.04045069e-01 1.82144850e-01 4.86141860e-01 7.68532157e-01
-1.56597942e-01 7.78282464e-01 8.20445418e-01 -5.17585874e-01
1.04372106e-01 3.68746430e-01 -1.71962619e-01 6.60715342e-01
7.17587650e-01 1.74508721e-01 1.69645444e-01 1.81933865e-01
7.52294123e-01 7.11447373e-02 -1.40785471e-01 -3.98221225e-01
-1.07135856e+00 7.70755529e-01 5.29368818e-01 2.93198198e-01
-3.58312130e-01 1.39272183e-01 1.19000994e-01 8.95215794e-02
3.78831744e-01 4.18772638e-01 -2.75782615e-01 3.71521771e-01
-9.81531560e-01 1.10390142e-01 6.58423066e-01 8.26767385e-01
7.25167215e-01 2.23739803e-01 1.07209973e-01 6.47893786e-01
2.14923196e-03 6.14723444e-01 -2.07587960e-03 -6.20816708e-01
3.64068002e-01 8.88467729e-01 3.75661433e-01 -1.06512702e+00
-3.28628719e-01 -6.35503232e-01 -6.50634050e-01 6.38637960e-01
7.59896278e-01 -2.50743061e-01 -1.01682603e+00 1.38043690e+00
3.28719407e-01 1.19947791e-01 2.61360221e-02 8.98624539e-01
7.06522763e-01 5.43667018e-01 -6.00242577e-02 1.37505844e-01
1.43791902e+00 -6.66438878e-01 -1.89447135e-01 -5.46504438e-01
2.49353930e-01 -7.66375422e-01 7.70967185e-01 2.33786643e-01
-6.26658022e-01 -5.13187289e-01 -9.22795653e-01 1.90545022e-01
-5.94215214e-01 7.29829192e-01 5.55871308e-01 6.91730261e-01
-8.42794657e-01 3.93727630e-01 -4.74355161e-01 -7.19609201e-01
6.34272873e-01 3.10899377e-01 -3.33273768e-01 -1.42683476e-01
-7.82330811e-01 9.69470739e-01 3.17521185e-01 1.04517750e-01
-9.62631822e-01 -5.83173573e-01 -6.36988640e-01 3.82575244e-02
6.57731891e-01 -3.59179884e-01 9.21756029e-01 -1.15142751e+00
-9.33707654e-01 9.70067739e-01 3.40583324e-01 -6.31444991e-01
7.97796965e-01 -2.39960968e-01 -2.73013860e-01 -1.07946023e-01
5.00958301e-02 7.02202260e-01 8.61708760e-01 -1.24027824e+00
-1.01388109e+00 -3.68799746e-01 3.22051316e-01 -2.42336933e-02
-5.62481843e-02 4.26578857e-02 -3.47428352e-01 -3.15000445e-01
-3.83143097e-01 -1.01171565e+00 -3.03714216e-01 2.17380509e-01
-3.14728111e-01 2.66903788e-02 1.10547018e+00 -4.65868235e-01
6.69356942e-01 -2.07922554e+00 -2.03816950e-01 -7.11158803e-03
1.27266562e-02 6.57734215e-01 -3.23114067e-01 2.08635375e-01
5.35934307e-02 -2.21132115e-02 -2.91396350e-01 8.72335732e-02
-3.12336832e-01 1.45565839e-02 -2.72477984e-01 3.87937635e-01
5.42855859e-01 6.82872891e-01 -9.40402150e-01 -2.70092338e-01
4.65200245e-01 5.42886019e-01 -1.61704838e-01 -9.97502804e-02
-1.87306195e-01 6.85012192e-02 -3.24553192e-01 7.44312704e-01
5.25400043e-01 -7.77639672e-02 -1.05228173e-02 -1.63462535e-01
-3.45745683e-01 -1.64080262e-01 -1.29675519e+00 1.07745755e+00
-3.59143525e-01 1.18267035e+00 1.26463324e-01 -9.11197424e-01
8.46715391e-01 -6.95332810e-02 1.93994850e-01 -5.02925277e-01
2.39427656e-01 6.09122328e-02 2.46894404e-01 -4.03616011e-01
6.97395504e-01 1.04668766e-01 2.52437651e-01 6.94663972e-02
-1.90561935e-01 -3.52395982e-01 5.46746731e-01 -2.56974827e-02
1.23228431e+00 -8.95301625e-03 2.93347031e-01 -9.66327861e-02
3.49047214e-01 2.21316546e-01 2.09907740e-01 6.25750840e-01
-1.45874575e-01 6.89041495e-01 4.56774890e-01 -6.82470679e-01
-1.16879392e+00 -5.70932925e-01 -2.05249533e-01 8.80760849e-01
1.81315720e-01 1.07694948e-02 -8.33932042e-01 -5.45194209e-01
1.07927710e-01 5.53449929e-01 -7.01396525e-01 1.27874032e-01
-3.30532402e-01 -1.11427987e+00 4.99877661e-01 4.78785306e-01
5.20850599e-01 -1.06750321e+00 -1.22171295e+00 1.25311553e-01
9.74460244e-02 -1.46323931e+00 1.80478469e-01 5.98275252e-02
-7.74825454e-01 -1.27506292e+00 -7.00115800e-01 -5.67536294e-01
7.42479444e-01 6.70907855e-01 1.12986100e+00 1.59101233e-01
-9.52149928e-01 3.57166201e-01 -1.31800100e-01 -6.80120051e-01
-3.41128647e-01 8.07016250e-03 -4.00615372e-02 1.27102286e-01
4.55545813e-01 -2.17377111e-01 -5.88325560e-01 4.76968557e-01
-9.24839079e-01 2.02033781e-02 8.52745593e-01 6.91594899e-01
2.84236401e-01 3.20043236e-01 2.95608789e-01 -6.35110199e-01
3.55506301e-01 -2.29384452e-01 -1.23687863e+00 1.78113118e-01
-1.77058652e-01 -6.21858798e-02 2.77476370e-01 -4.60191071e-01
-6.77883923e-01 6.08622253e-01 1.94972336e-01 -3.38369071e-01
-5.28176308e-01 1.23325542e-01 -5.70998602e-02 -2.68192261e-01
8.01016688e-01 4.05150764e-02 -2.30753854e-01 -2.46674106e-01
1.89921364e-01 5.26403427e-01 5.46775281e-01 -1.63440689e-01
9.74361360e-01 5.93954086e-01 1.69203654e-01 -1.32841361e+00
-6.23301804e-01 -4.68748957e-01 -6.80089653e-01 -3.37749630e-01
8.02691877e-01 -8.80567610e-01 -4.09196943e-01 3.43607157e-01
-1.17043674e+00 -4.47767049e-01 -3.96595895e-01 3.91590685e-01
-2.15561241e-01 1.87596604e-01 -1.51530862e-01 -1.03188825e+00
-4.13740762e-02 -1.18906283e+00 1.20466304e+00 3.42026383e-01
2.67875075e-01 -5.30464113e-01 -2.28405461e-01 -5.30212373e-03
4.17544156e-01 4.69658703e-01 6.17033184e-01 -2.90136397e-01
-1.00668514e+00 -5.55597246e-01 -5.89444041e-01 4.60779071e-01
3.14656831e-02 2.61339754e-01 -1.00091195e+00 4.87689348e-03
-4.89691675e-01 -6.89840689e-02 1.05063796e+00 3.65499675e-01
8.61910522e-01 1.40631959e-01 -4.62894529e-01 1.64300069e-01
1.55364919e+00 1.66167066e-01 6.71669066e-01 2.36647159e-01
5.88285387e-01 9.59484994e-01 5.84326625e-01 2.13323638e-01
-2.67532840e-02 7.19553888e-01 8.20140660e-01 -4.26666051e-01
-3.37457597e-01 -2.39636116e-02 1.99560851e-01 -1.30866170e-01
-3.71469587e-01 -3.34877253e-01 -1.02635252e+00 7.58212209e-01
-1.64811027e+00 -1.11639500e+00 -2.97844082e-01 2.31413245e+00
2.64392287e-01 9.58113000e-02 1.38702884e-01 1.49483800e-01
1.00157142e+00 -2.52510190e-01 -4.63723302e-01 2.16114298e-01
-1.41883388e-01 -1.60520077e-01 8.86963129e-01 1.33812889e-01
-1.37266076e+00 9.24484253e-01 6.40204000e+00 5.36711276e-01
-1.19200385e+00 -1.82668641e-01 6.45469248e-01 5.96815683e-02
3.46142828e-01 -3.03992748e-01 -8.61691594e-01 5.56100726e-01
5.44472754e-01 2.44570971e-01 2.05155224e-01 1.05199552e+00
8.50700289e-02 -4.06066656e-01 -1.02285063e+00 9.87087607e-01
2.42579859e-02 -1.12933660e+00 -1.15370445e-01 2.67176867e-01
6.05563998e-01 2.17257082e-01 3.15493681e-02 1.22788131e-01
3.56615186e-01 -9.40652549e-01 6.27320886e-01 1.67786077e-01
6.23762786e-01 -5.35816908e-01 8.82228971e-01 3.64587367e-01
-1.09134543e+00 -2.62145370e-01 -7.03415275e-01 -1.36891678e-01
-7.67772794e-02 5.53519547e-01 -1.01816690e+00 1.70822397e-01
6.14551723e-01 3.93014193e-01 -8.42413902e-01 1.42871237e+00
-2.98320740e-01 4.86550301e-01 -5.01811504e-01 -8.27057734e-02
3.13560188e-01 -2.05704346e-01 4.23157662e-01 1.23315549e+00
4.68476772e-01 -7.67164081e-02 9.14426427e-03 8.88315141e-01
-1.64423853e-01 -1.28845289e-01 -1.08612084e+00 4.19096537e-02
2.51750678e-01 1.63329756e+00 -1.16953957e+00 -1.59510851e-01
-5.51165044e-01 8.61071706e-01 1.22342026e-02 2.97678709e-01
-7.55973160e-01 -3.53056610e-01 5.77879250e-01 5.34907758e-01
6.03826284e-01 -2.35065684e-01 1.63827315e-01 -8.23526144e-01
-1.10096164e-01 -6.76286995e-01 -8.86926427e-03 -8.22405994e-01
-9.51980233e-01 5.65306723e-01 -1.91638675e-02 -1.20863950e+00
-7.35922232e-02 -1.06527913e+00 -6.01399183e-01 5.42793691e-01
-1.45994556e+00 -1.20141840e+00 -7.47808933e-01 1.52186483e-01
6.98911607e-01 -2.10022181e-01 4.24838543e-01 4.48086739e-01
-4.23593223e-01 6.70740232e-02 1.04216233e-01 3.56681287e-01
3.85477811e-01 -1.06939948e+00 6.45746648e-01 1.17055130e+00
4.23886150e-01 1.50541514e-01 7.43511319e-01 -3.59622419e-01
-1.26176095e+00 -1.25947523e+00 4.03632402e-01 -7.35938370e-01
3.77683192e-01 -4.52668607e-01 -6.85244918e-01 3.73048633e-01
2.20093206e-02 2.59745717e-01 2.50526518e-01 -2.17917934e-01
-4.88292277e-02 -1.79177254e-01 -1.02204156e+00 5.61949432e-01
9.26378727e-01 -1.94727644e-01 -2.77114421e-01 2.47361213e-01
4.22648728e-01 -2.49978974e-01 -1.75275967e-01 4.81114894e-01
6.65872157e-01 -9.79315400e-01 1.11534476e+00 -4.13012713e-01
4.22898084e-01 -7.07455397e-01 -1.99363247e-01 -1.17951119e+00
-9.00516436e-02 -1.32434621e-01 2.27728933e-01 1.06399357e+00
3.78970325e-01 -2.59486139e-01 8.69625926e-01 3.96093965e-01
1.64439037e-01 -3.70040685e-01 -5.86607516e-01 -7.28297770e-01
-2.84206420e-01 -3.16713542e-01 4.05494809e-01 6.31556869e-01
-8.08912694e-01 3.53468120e-01 -3.34953457e-01 2.39384130e-01
5.13243318e-01 1.19570114e-01 1.14700961e+00 -1.50269794e+00
2.98169963e-02 -3.09233576e-01 -8.87203932e-01 -7.04620898e-01
-1.84083641e-01 -4.35771912e-01 1.53541461e-01 -1.67362356e+00
1.05794407e-01 -3.63877386e-01 -6.74176496e-04 3.17798108e-01
-1.04552120e-01 6.15866542e-01 3.29322249e-01 -1.00253545e-01
-4.82749373e-01 9.85810980e-02 8.27723920e-01 -3.02956820e-01
-8.14511776e-02 1.11456618e-01 -5.02671421e-01 9.30395305e-01
7.78214991e-01 -3.60928804e-01 -1.75594419e-01 -6.01721525e-01
1.21103026e-01 -4.86766338e-01 7.56851137e-01 -1.34064209e+00
1.35377809e-01 9.63841528e-02 5.93543768e-01 -6.97900772e-01
2.64815003e-01 -1.02923906e+00 6.07434586e-02 6.45571828e-01
6.37055263e-02 -1.53716095e-02 4.27259237e-01 6.82913244e-01
-5.07284440e-02 -3.48360747e-01 8.08144271e-01 -2.63387918e-01
-1.01653504e+00 -4.60943440e-03 -4.58095133e-01 -2.34290823e-01
1.20660520e+00 -4.48435307e-01 -3.98662657e-01 -2.72624284e-01
-2.95066178e-01 -1.22785764e-02 5.29575527e-01 5.46005011e-01
4.85473007e-01 -8.56173933e-01 -8.81323338e-01 2.17593968e-01
2.16281891e-01 8.43945146e-02 -5.58205098e-02 7.03538895e-01
-6.44974470e-01 4.03846681e-01 -3.81588578e-01 -7.39374995e-01
-1.49839270e+00 5.77680349e-01 3.34798276e-01 5.51684527e-03
-4.63122547e-01 6.27260804e-01 6.60108864e-01 -4.97052334e-02
1.95011020e-01 -5.16776085e-01 -1.74490452e-01 1.96747169e-01
6.40770257e-01 5.99249005e-01 1.54254392e-01 -6.69863284e-01
-2.30124429e-01 5.41439533e-01 1.11480758e-01 2.37895191e-01
1.28381789e+00 1.35795981e-01 1.52571425e-01 5.70728593e-02
5.82250595e-01 -6.50805160e-02 -1.44493341e+00 -1.18192218e-01
9.06838700e-02 -5.45154452e-01 1.52145594e-01 -6.99815810e-01
-1.12543595e+00 1.08602428e+00 9.94787276e-01 3.43466789e-01
9.71807897e-01 -3.60904895e-02 2.98622817e-01 5.57410121e-01
5.53520739e-01 -8.99140298e-01 1.56234512e-02 2.67064542e-01
7.45587826e-01 -1.51409137e+00 2.20858157e-01 -3.39258581e-01
-3.66617948e-01 1.09873354e+00 5.15621126e-01 -1.14019513e-01
1.31548792e-01 4.01526570e-01 5.19810637e-05 -1.72683373e-01
-3.23590547e-01 -7.79289424e-01 2.28854418e-01 7.35072494e-01
1.63590640e-01 5.54863140e-02 1.46867437e-02 1.35522693e-01
1.89227834e-01 -7.55464733e-02 6.16652131e-01 8.30605209e-01
-6.61184490e-01 -8.37274790e-01 -5.17091870e-01 4.81901228e-01
-4.75106955e-01 -7.22846314e-02 -7.08521545e-01 1.07663381e+00
3.72786164e-01 8.36086392e-01 2.12736458e-01 -1.65518358e-01
3.29443902e-01 -2.47628391e-01 4.90968347e-01 -6.53109968e-01
-2.75550157e-01 -1.88164786e-01 9.42321320e-04 -2.83110768e-01
-5.39027274e-01 -6.27427459e-01 -6.98971212e-01 -1.04099661e-02
-7.28405476e-01 -2.61944145e-01 9.77112055e-01 6.41095161e-01
2.76832283e-01 6.23393595e-01 2.36204877e-01 -1.06111789e+00
-2.53519148e-01 -7.79973865e-01 -5.58657765e-01 2.37486497e-01
3.42391908e-01 -8.44185948e-01 -2.87076652e-01 3.10281634e-01] | [8.5631685256958, -0.8910019993782043] |
8435c560-1737-4443-863b-ae90afa4bf47 | voxelembed-3d-instance-segmentation-and | 2106.11480 | null | https://arxiv.org/abs/2106.11480v1 | https://arxiv.org/pdf/2106.11480v1.pdf | VoxelEmbed: 3D Instance Segmentation and Tracking with Voxel Embedding based Deep Learning | Recent advances in bioimaging have provided scientists a superior high spatial-temporal resolution to observe dynamics of living cells as 3D volumetric videos. Unfortunately, the 3D biomedical video analysis is lagging, impeded by resource insensitive human curation using off-the-shelf 3D analytic tools. Herein, biologists often need to discard a considerable amount of rich 3D spatial information by compromising on 2D analysis via maximum intensity projection. Recently, pixel embedding-based cell instance segmentation and tracking provided a neat and generalizable computing paradigm for understanding cellular dynamics. In this work, we propose a novel spatial-temporal voxel-embedding (VoxelEmbed) based learning method to perform simultaneous cell instance segmenting and tracking on 3D volumetric video sequences. Our contribution is in four-fold: (1) The proposed voxel embedding generalizes the pixel embedding with 3D context information; (2) Present a simple multi-stream learning approach that allows effective spatial-temporal embedding; (3) Accomplished an end-to-end framework for one-stage 3D cell instance segmentation and tracking without heavy parameter tuning; (4) The proposed 3D quantification is memory efficient via a single GPU with 12 GB memory. We evaluate our VoxelEmbed method on four 3D datasets (with different cell types) from the ISBI Cell Tracking Challenge. The proposed VoxelEmbed method achieved consistent superior overall performance (OP) on two densely annotated datasets. The performance is also competitive on two sparsely annotated cohorts with 20.6% and 2% of data-set having segmentation annotations. The results demonstrate that the VoxelEmbed method is a generalizable and memory-efficient solution. | ['Yuankai Huo', 'Bryan A. Millis', 'Matthew J. Tyska', 'Anita Mahadevan-Jansen', 'Tianyuan Yao', 'Ruining Deng', 'Aadarsh Jha', 'Quan Liu', 'Mengyang Zhao'] | 2021-06-22 | null | null | null | null | ['3d-instance-segmentation-1'] | ['computer-vision'] | [ 1.40967667e-02 -3.41471344e-01 -3.58848721e-02 9.37404409e-02
-1.00221920e+00 -5.42539954e-01 3.40852082e-01 5.70380151e-01
-7.68710315e-01 6.46265209e-01 -1.78564057e-01 -2.11399943e-01
3.65261286e-02 -4.94664103e-01 -6.01476371e-01 -1.09508312e+00
-4.72650200e-01 6.08275771e-01 3.41258705e-01 1.68352529e-01
1.69132948e-01 7.75465429e-01 -1.05986738e+00 -2.23959330e-02
3.33175689e-01 9.59460378e-01 2.15619147e-01 1.10387301e+00
-1.09322742e-01 2.89510667e-01 -1.82969049e-01 1.84722215e-01
2.96234876e-01 -2.34510288e-01 -5.87792456e-01 2.67528296e-02
4.89106745e-01 -4.06876653e-01 -1.23122983e-01 5.36873937e-01
9.01655018e-01 -2.20180690e-01 7.28727758e-01 -9.40424383e-01
-2.68560559e-01 -7.30412900e-02 -8.34949911e-01 7.08101392e-01
1.27069309e-01 3.83426189e-01 5.55357397e-01 -9.07304466e-01
1.09216678e+00 8.96711111e-01 9.40976799e-01 6.35391116e-01
-1.56122541e+00 -3.78938556e-01 -3.57528739e-02 -2.34981820e-01
-1.41218138e+00 -8.96566510e-02 6.10759377e-01 -8.44417512e-01
1.06654847e+00 3.02396327e-01 1.18698120e+00 7.70009100e-01
5.38400173e-01 6.33420110e-01 1.18985891e+00 1.39257070e-02
2.95012981e-01 -1.97332114e-01 2.66007811e-01 9.87237155e-01
2.27510840e-01 6.20883657e-03 -4.15277213e-01 -3.29101801e-01
1.08875394e+00 2.46388704e-01 -3.73543322e-01 -4.12194610e-01
-1.79623747e+00 5.28865576e-01 2.76469171e-01 3.54483634e-01
-2.74746656e-01 5.30884087e-01 7.72742569e-01 -3.69129749e-03
6.76453650e-01 1.01089835e-01 -6.13901317e-01 -2.59254277e-01
-1.16561663e+00 2.97547758e-01 5.61875165e-01 6.07915580e-01
4.75145996e-01 -1.33139834e-01 -1.10237271e-01 3.90023828e-01
2.50766337e-01 5.17089367e-01 3.50938022e-01 -1.02074039e+00
-1.34609237e-01 7.21730530e-01 1.87369343e-02 -1.00025988e+00
-7.36952245e-01 -1.56979039e-01 -1.00015402e+00 2.23164856e-01
7.42322683e-01 7.98760951e-02 -7.58883595e-01 1.54305398e+00
8.42688382e-01 6.40997231e-01 -2.64458865e-01 1.11534512e+00
7.76136041e-01 6.05894268e-01 8.16111863e-02 -2.96258420e-01
1.46752787e+00 -5.31499565e-01 -5.23520291e-01 5.19367635e-01
1.04769409e+00 -3.14609259e-01 1.01379359e+00 1.90609843e-01
-8.66257668e-01 -1.99777618e-01 -8.73303473e-01 -2.98894346e-01
-4.40399587e-01 -2.21281305e-01 7.11829543e-01 5.74026763e-01
-1.11902118e+00 4.35066760e-01 -1.41876400e+00 -4.01733160e-01
9.93621528e-01 5.79869747e-01 -6.54958665e-01 -4.37892880e-03
-5.04492760e-01 2.71456748e-01 1.47249848e-01 -1.59038931e-01
-9.64993834e-01 -1.46288323e+00 -6.31080687e-01 -2.98304826e-01
-6.35246262e-02 -1.04084432e+00 5.23805261e-01 -1.62512153e-01
-1.26160121e+00 1.40776908e+00 -2.16655061e-01 -2.85334021e-01
7.39558399e-01 6.37125522e-02 3.56945843e-01 5.58756709e-01
1.64002746e-01 8.62105846e-01 2.36170754e-01 -1.10165346e+00
-4.07644689e-01 -7.34188497e-01 -5.90256095e-01 -7.68710449e-02
-2.20113114e-01 -2.50292629e-01 -6.11692727e-01 -5.80692768e-01
2.42994562e-01 -9.37017024e-01 -2.65884221e-01 7.99539387e-01
-1.59106731e-01 1.02170363e-01 9.16190326e-01 -5.63988209e-01
6.98430121e-01 -2.03047371e+00 3.02221030e-01 -2.65641689e-01
6.93495929e-01 7.17331842e-02 2.01394066e-01 1.14185266e-01
2.26664320e-01 2.06333786e-01 -3.50782514e-01 -5.05519211e-01
-1.92920029e-01 -4.78795581e-02 -4.35920879e-02 1.01708043e+00
-1.04082339e-02 1.12362158e+00 -1.16206658e+00 -9.64797556e-01
2.61980653e-01 9.75303411e-01 -7.82880187e-01 -4.59756441e-02
3.38197090e-02 7.91821897e-01 -4.13329065e-01 1.10023534e+00
5.32131553e-01 -4.25151378e-01 -2.46081017e-02 -2.75353521e-01
-2.20224351e-01 -4.47468519e-01 -7.65644491e-01 1.85994506e+00
9.51974317e-02 5.60356855e-01 1.58726856e-01 -9.58149850e-01
6.30819261e-01 3.14610660e-01 1.25826192e+00 -3.41645420e-01
1.88649446e-01 3.87065977e-01 -6.09150589e-01 -4.64427859e-01
-8.70103613e-02 -3.55442226e-01 -1.35111079e-01 3.38806987e-01
9.89565700e-02 -1.84781924e-01 3.29691954e-02 2.64975756e-01
1.35874796e+00 3.30013186e-01 4.37562019e-02 -5.77828288e-01
4.35715854e-01 2.23717764e-01 7.45810568e-01 4.06098604e-01
-8.72584343e-01 4.81696039e-01 6.58393919e-01 -7.64421642e-01
-1.16914392e+00 -1.11334014e+00 -3.28509301e-01 6.56180441e-01
2.20643997e-01 -5.37886173e-02 -5.23032367e-01 -5.34058511e-01
4.10085499e-01 -3.19398701e-01 -7.75252402e-01 3.35441470e-01
-7.25638151e-01 -1.05796027e+00 7.37070739e-01 4.43767935e-01
1.66508511e-01 -3.59521896e-01 -9.25479829e-01 2.89442271e-01
1.55926362e-01 -1.02898049e+00 -6.20980680e-01 1.95198029e-01
-1.25965285e+00 -1.02623260e+00 -1.12634063e+00 -8.29529703e-01
6.95415080e-01 3.22302997e-01 7.67178297e-01 1.49316460e-01
-8.18469167e-01 5.47481656e-01 -4.08607814e-03 -3.87019068e-02
7.03166425e-02 -1.82853878e-01 2.27811426e-01 -2.84335762e-01
2.20544621e-01 -6.82106018e-01 -8.96654844e-01 1.86724141e-01
-7.26131141e-01 9.76573005e-02 1.89280242e-01 1.13355458e+00
1.51396108e+00 -3.25442374e-01 4.69675869e-01 -8.87521029e-01
9.03866664e-02 -3.14745158e-01 -8.05993080e-01 -1.20442241e-01
-4.70986366e-01 -2.59508252e-01 5.10332704e-01 -4.28338587e-01
-4.53498989e-01 2.50914663e-01 -1.37155503e-01 -6.22684896e-01
3.85161042e-02 2.87419140e-01 2.89639980e-01 -3.87255043e-01
3.31329435e-01 4.50409085e-01 4.21168894e-01 -2.48720273e-01
1.31962318e-02 2.06579849e-01 4.97475207e-01 -4.60068941e-01
4.07421768e-01 1.18366766e+00 5.12973428e-01 -9.68414783e-01
-2.92372286e-01 -7.59585381e-01 -8.89772236e-01 -3.81096691e-01
1.00522780e+00 -8.91188622e-01 -1.17051923e+00 5.54279029e-01
-8.58935654e-01 -7.15777874e-01 -1.74393177e-01 5.00595808e-01
-7.47303009e-01 6.44079447e-01 -1.11770523e+00 -6.64183915e-01
-5.72968721e-01 -1.26252687e+00 1.52902687e+00 -4.04680483e-02
-1.55825362e-01 -1.17569482e+00 4.90081728e-01 4.83529359e-01
2.33152553e-01 9.44628656e-01 8.71723533e-01 -2.58405477e-01
-8.08115900e-01 -2.87888736e-01 -3.15563411e-01 -4.35119271e-01
5.88585436e-03 3.25814486e-02 -9.21243131e-01 -4.67297524e-01
-2.45076731e-01 -2.01452449e-01 9.18568671e-01 7.91305482e-01
1.04791045e+00 4.09466401e-02 -6.92087531e-01 1.16073012e+00
1.59063852e+00 -2.76624471e-01 3.32655013e-01 1.84558541e-01
8.49258721e-01 3.47989321e-01 4.01255369e-01 4.59556222e-01
1.60473093e-01 6.32788777e-01 3.23313087e-01 -2.68407077e-01
-2.80137450e-01 -3.67370509e-02 1.76180527e-01 1.06145656e+00
-5.34001216e-02 -2.66724117e-02 -9.83494818e-01 8.00029039e-01
-1.71835709e+00 -8.92386615e-01 -2.54421622e-01 2.01276326e+00
9.01698589e-01 -1.19072177e-01 3.92062157e-01 1.26952454e-01
3.67988586e-01 -8.04653689e-02 -6.51549816e-01 1.58683941e-01
-1.78734049e-01 -1.83895193e-02 6.25742972e-01 5.84749341e-01
-1.17798543e+00 7.55784750e-01 5.73235130e+00 7.30549991e-01
-1.15268266e+00 3.34714264e-01 8.48704278e-01 -5.01498282e-01
-7.58857280e-02 -3.55927020e-01 -9.22444105e-01 5.58000088e-01
6.62881970e-01 2.25299709e-02 -5.98469004e-02 5.18152177e-01
4.76151288e-01 -1.03667848e-01 -1.12099802e+00 1.25500762e+00
-2.77842373e-01 -1.83899438e+00 3.84281240e-02 5.25467575e-01
5.20763934e-01 7.28351697e-02 8.65508094e-02 -7.77176544e-02
-2.40974560e-01 -9.25382018e-01 4.47477788e-01 4.65350091e-01
1.06800902e+00 -5.63815117e-01 7.01755941e-01 3.20602566e-01
-1.22119284e+00 3.07790220e-01 -2.65458792e-01 2.49314502e-01
4.00603592e-01 8.10831785e-01 -7.82177925e-01 1.89492300e-01
7.30471075e-01 8.47179234e-01 -2.37604693e-01 1.07503557e+00
6.29428148e-01 5.61075211e-01 -4.88058925e-01 1.12877684e-02
1.99359924e-01 -2.47800380e-01 8.06621909e-01 1.61199343e+00
2.77389407e-01 2.82294333e-01 1.33513033e-01 8.73424113e-01
-9.09432769e-02 6.88153040e-03 -5.52962124e-01 -8.55858997e-02
3.35300386e-01 1.33465195e+00 -1.33206499e+00 -1.38829082e-01
-2.22557500e-01 9.48777735e-01 1.69474691e-01 9.66001600e-02
-7.98610628e-01 -1.70711115e-01 7.53249049e-01 3.24026704e-01
2.91370898e-01 -5.39783418e-01 -3.20227712e-01 -9.39649045e-01
-4.08040375e-01 -2.28891164e-01 3.84316325e-01 -3.53004813e-01
-1.20235002e+00 1.20317519e-01 -2.87413716e-01 -1.05096662e+00
3.64459097e-01 -8.40650380e-01 -1.10787399e-01 5.93255103e-01
-1.66681838e+00 -1.13487244e+00 -4.00189102e-01 4.40547138e-01
3.14488351e-01 7.60815889e-02 8.89132023e-01 5.83703756e-01
-5.83817959e-01 3.93118471e-01 1.68582350e-01 1.98196188e-01
4.64197755e-01 -1.31246126e+00 -8.71325061e-02 4.26281184e-01
-7.45289549e-02 4.97090369e-01 4.38757271e-01 -6.49630785e-01
-2.01869226e+00 -1.19187927e+00 7.25686669e-01 -7.03571260e-01
5.43607593e-01 -5.03394246e-01 -9.02218401e-01 5.14950752e-01
-2.84110069e-01 6.39073431e-01 1.09203172e+00 -3.88417095e-01
-4.62236628e-02 -3.40052433e-02 -1.46607363e+00 5.63947260e-01
9.81793582e-01 -3.11128229e-01 -7.71507025e-02 3.84049177e-01
7.32092798e-01 -5.75754821e-01 -1.27702296e+00 2.07221046e-01
6.47879004e-01 -7.11186409e-01 1.09792292e+00 -2.97077417e-01
1.51867434e-01 -5.60660243e-01 -2.24337652e-01 -5.12695432e-01
-3.00345838e-01 -6.63471580e-01 -4.95174378e-01 7.22061515e-01
9.11480561e-02 -4.56262946e-01 1.14555073e+00 3.12460691e-01
-2.80077845e-01 -1.26673186e+00 -1.41889048e+00 -7.04673231e-01
2.92185545e-01 -2.75657713e-01 2.10170761e-01 9.37697947e-01
3.66411507e-02 -5.80027811e-02 -5.59246168e-03 -7.45684728e-02
9.81647193e-01 1.34253740e-01 8.42767298e-01 -1.17116952e+00
-1.77364707e-01 -5.04233122e-01 -8.55172634e-01 -1.19309652e+00
-1.19800396e-01 -1.14862800e+00 -2.85164118e-01 -1.15625775e+00
6.40734613e-01 -4.28417593e-01 -4.44318056e-01 2.23837316e-01
-2.45179646e-02 6.51443660e-01 -6.47660345e-02 3.69469464e-01
-7.45044351e-01 4.14967865e-01 1.47814810e+00 -2.95916140e-01
-2.21061632e-01 -6.12142086e-01 -2.06765428e-01 3.84512305e-01
4.70757365e-01 -4.15535212e-01 -1.02030896e-01 -3.01868021e-01
3.19361202e-02 1.26088575e-01 7.86095142e-01 -9.74186540e-01
3.03840607e-01 1.21814325e-01 6.08920217e-01 -7.82395899e-01
5.04678726e-01 -7.11432755e-01 1.96852863e-01 8.56161058e-01
2.99265459e-02 -2.51109689e-01 3.81812572e-01 9.17565525e-01
1.26490682e-01 4.63246167e-01 1.00274396e+00 -1.72936961e-01
-5.27320743e-01 6.67028964e-01 -4.55115229e-01 1.19047880e-01
1.31814718e+00 -7.30978191e-01 -6.46106750e-02 2.64590085e-01
-6.22765481e-01 1.23425446e-01 8.12473595e-01 -3.24102193e-01
6.86007082e-01 -1.30050969e+00 -6.42406225e-01 3.41592892e-03
-3.44736539e-02 1.25065118e-01 4.31216031e-01 1.43432760e+00
-1.19198799e+00 6.00602686e-01 -2.14398861e-01 -1.26875937e+00
-1.22931314e+00 4.79603678e-01 4.36836600e-01 -3.91661227e-01
-9.01857495e-01 1.09645391e+00 2.06879944e-01 -5.05896330e-01
-9.26078856e-02 -5.34395576e-01 1.63410261e-01 -1.53691517e-02
4.48552996e-01 4.73855704e-01 -1.58901632e-01 -7.16276824e-01
-6.05271459e-01 9.95098233e-01 5.72950691e-02 5.25912270e-02
1.57151592e+00 -7.24227950e-02 -3.10466643e-02 7.46122360e-01
1.56655073e+00 -2.86502540e-01 -1.56557453e+00 1.35534823e-01
-2.09984079e-01 -4.59404647e-01 3.64629626e-01 -4.05336708e-01
-1.17437327e+00 8.72584045e-01 8.67367506e-01 -1.21395476e-01
8.68682504e-01 3.11206039e-02 1.30449307e+00 4.84081125e-03
6.39186203e-01 -8.77770066e-01 2.59594582e-02 2.91732192e-01
3.09196621e-01 -1.20395195e+00 2.27967247e-01 -4.06204581e-01
-1.36591598e-01 8.80917490e-01 2.03516200e-01 -8.06461945e-02
7.56945014e-01 4.62250620e-01 -2.68621296e-02 -6.19463980e-01
-7.74289608e-01 1.56841338e-01 -1.46197096e-01 6.90471470e-01
5.53129196e-01 -1.44238263e-01 -2.21386284e-01 4.87630576e-01
3.86901110e-01 2.25715756e-01 1.97427943e-01 1.04144704e+00
-3.26570660e-01 -5.40228963e-01 -1.52340323e-01 4.13996607e-01
-5.85291505e-01 2.16789171e-01 -8.75740498e-02 9.06995773e-01
3.29235792e-02 1.73451617e-01 6.23371713e-02 -2.37454101e-03
-1.00317456e-01 -1.15909405e-01 7.15243638e-01 -3.29496861e-01
-3.99642169e-01 3.50947052e-01 -6.35634065e-01 -7.71100581e-01
-3.67710680e-01 -8.77435267e-01 -1.67400229e+00 -3.88292819e-01
-8.62495005e-02 -2.86277413e-01 4.98240143e-01 5.96971571e-01
7.14291036e-01 3.90876740e-01 2.51495957e-01 -1.23122370e+00
6.73680333e-03 -3.85549873e-01 -7.97994614e-01 3.09605360e-01
4.67826724e-01 -6.32473350e-01 -4.02072489e-01 5.22434413e-01] | [14.484166145324707, -3.1832711696624756] |
a97cb2be-0fc4-4d89-bc05-ca516fc8a5e8 | improving-the-generation-of-personalised | null | null | https://aclanthology.org/W17-3536 | https://aclanthology.org/W17-3536.pdf | Improving the generation of personalised descriptions | Referring expression generation (REG) models that use speaker-dependent information require a considerable amount of training data produced by every individual speaker, or may otherwise perform poorly. In this work we propose a simple personalised method for this task, in which speakers are grouped into profiles according to their referential behaviour. Intrinsic evaluation shows that the use of speaker{'}s profiles generally outperforms the personalised method found in previous work. | ["r{\\'e}", 'Thiago Castro Ferreira', 'Iv Paraboni'] | 2017-09-01 | null | null | null | ws-2017-9 | ['referring-expression-generation'] | ['computer-vision'] | [ 2.56400496e-01 4.83805597e-01 -2.40071490e-02 -8.49082947e-01
-1.19952714e+00 -5.48239172e-01 9.78237092e-01 1.20497029e-02
-3.34003419e-01 8.41338217e-01 6.86673641e-01 -2.08879113e-02
7.67741278e-02 -4.36972022e-01 -2.19220221e-01 -4.52845544e-01
-5.16495816e-02 5.70234954e-01 2.02872038e-01 -8.13949764e-01
4.15083408e-01 3.06958020e-01 -1.57961118e+00 3.51633251e-01
7.89198875e-01 5.94237328e-01 -4.98813875e-02 4.82013166e-01
-5.40829957e-01 7.52606332e-01 -1.35431373e+00 -5.18384516e-01
-2.99495488e-01 -7.41225660e-01 -1.18306088e+00 1.03866622e-01
1.87164709e-01 2.82649666e-01 4.06789958e-01 8.54527473e-01
7.02883065e-01 7.73771226e-01 4.63489830e-01 -8.88023734e-01
-6.45535290e-01 1.14284074e+00 -6.17702119e-03 5.52079141e-01
8.74403059e-01 -3.85867536e-01 9.34853494e-01 -4.91392404e-01
6.37265146e-01 1.34606123e+00 5.29443324e-01 1.09356880e+00
-1.58157814e+00 -5.66818178e-01 3.55803251e-01 4.04864624e-02
-1.71360600e+00 -1.01362836e+00 1.18320692e+00 -3.32020760e-01
8.61472905e-01 3.63055736e-01 1.59190759e-01 1.30342543e+00
-5.02188683e-01 4.95809287e-01 1.18387139e+00 -7.06177115e-01
3.49786997e-01 6.86474979e-01 9.60999876e-02 2.56921589e-01
-3.18236113e-01 -1.90169021e-01 -9.19486105e-01 -3.00416380e-01
3.90171170e-01 -9.30473328e-01 -4.28844482e-01 -1.25492096e-01
-1.00914156e+00 9.89984393e-01 3.28376740e-02 9.61626649e-01
-5.75390697e-01 -1.97266594e-01 4.81277823e-01 3.75229955e-01
6.95878386e-01 4.75528598e-01 -1.95546567e-01 -3.21372002e-01
-1.04167938e+00 3.21912110e-01 9.13495064e-01 1.07835853e+00
6.63574457e-01 2.22052678e-01 -3.76655459e-01 1.22628307e+00
4.60183829e-01 -5.30440845e-02 8.91870916e-01 -9.83730316e-01
2.85970479e-01 4.85668629e-01 3.05349171e-01 -5.98148942e-01
-1.65725037e-01 -7.49712512e-02 -2.52298951e-01 -2.79994518e-01
1.46121919e-01 -1.75793290e-01 -5.14577389e-01 2.00554013e+00
3.75571400e-01 -1.95243791e-01 5.61161458e-01 4.55670953e-01
1.09600079e+00 6.95541799e-01 5.86187840e-01 -5.07370889e-01
1.31133819e+00 -4.53613997e-01 -1.09818149e+00 4.83391993e-02
4.85448956e-01 -8.17341924e-01 1.03611112e+00 2.19411090e-01
-1.38340461e+00 -5.01487792e-01 -9.05783653e-01 2.11124271e-01
-3.38085562e-01 -2.63743073e-01 7.25749433e-01 1.41935539e+00
-1.45071256e+00 4.94594038e-01 -2.97881246e-01 -4.94753093e-01
2.21714556e-01 3.60261112e-01 -2.33954102e-01 5.59648216e-01
-1.34624553e+00 1.10675883e+00 2.80594617e-01 -3.22295241e-02
-3.97642672e-01 -1.89118788e-01 -9.50524449e-01 -2.29273707e-01
-5.12394235e-02 -5.43978333e-01 1.67692852e+00 -1.09597790e+00
-2.35672307e+00 1.09525323e+00 -5.83865225e-01 -4.65046138e-01
3.20918560e-01 -1.70084964e-02 -6.10059798e-01 9.73598659e-02
2.48073339e-02 5.56938946e-01 7.59629786e-01 -1.55654657e+00
-4.12790090e-01 -3.46210122e-01 1.53339490e-01 2.21734643e-01
-1.14412218e-01 6.44119143e-01 -1.39532328e-01 -6.15322232e-01
4.56844736e-03 -5.58835864e-01 -1.03456946e-02 -9.14509594e-01
-2.24389702e-01 -1.08033288e+00 5.41203678e-01 -3.21207255e-01
1.47857165e+00 -1.96314788e+00 1.52017817e-01 2.09688514e-01
-4.80260044e-01 2.55558223e-01 2.29549363e-01 5.25946081e-01
-3.83850098e-01 2.13031769e-01 -2.22818553e-01 -7.14656115e-01
1.15660198e-01 4.23836857e-01 -1.14782386e-01 1.91046298e-01
1.23949118e-01 4.67690498e-01 -8.33017230e-01 -8.74771833e-01
-6.42194003e-02 5.78910589e-01 -2.88247138e-01 2.52550840e-01
-4.50849049e-02 4.60011244e-01 -4.75796044e-01 3.61510962e-01
2.55502939e-01 2.13047385e-01 8.91295895e-02 2.62755066e-01
-1.64739594e-01 6.31339490e-01 -1.12071979e+00 1.59379828e+00
-7.34444439e-01 4.10151333e-01 4.96956818e-02 -8.30515862e-01
1.37966466e+00 9.12333488e-01 4.02324758e-02 -2.45864287e-01
1.69458717e-01 2.18265444e-01 -8.97687078e-02 -4.90854055e-01
4.99769092e-01 -8.28595877e-01 -2.80558556e-01 4.41453218e-01
7.00659528e-02 -2.78376877e-01 1.46758094e-01 6.63660690e-02
7.79043674e-01 3.05752277e-01 6.47223294e-01 -3.59867781e-01
1.16629624e+00 -2.09488958e-01 4.34763342e-01 3.91645223e-01
-5.42590141e-01 2.48131201e-01 3.41801226e-01 8.37577134e-02
-4.59640056e-01 -7.03967631e-01 -2.68140912e-01 1.54857552e+00
-1.76334396e-01 -3.28596532e-01 -9.05880570e-01 -5.60451090e-01
-6.14120424e-01 1.53595471e+00 -7.59485185e-01 -4.39502932e-02
-6.56272173e-01 -4.84982640e-01 6.94260955e-01 3.55246723e-01
3.53981942e-01 -1.43600047e+00 -5.74819922e-01 4.60851043e-01
-4.24004853e-01 -6.90267563e-01 -3.73791099e-01 9.39584747e-02
-7.55035818e-01 -2.77702034e-01 -7.08813071e-01 -6.73249602e-01
4.04491782e-01 -2.70841300e-01 1.17689466e+00 -2.25842789e-01
5.25062084e-01 4.67644006e-01 -7.09950149e-01 -6.77340031e-01
-9.46878850e-01 1.13140374e-01 -2.44909912e-01 2.27263302e-01
6.35345161e-01 -6.64501965e-01 -8.44560266e-02 2.71565646e-01
-6.76234961e-01 -5.09454668e-01 2.62580216e-01 6.10545874e-01
2.10783631e-01 -2.60614365e-01 9.48220432e-01 -1.17338932e+00
1.04252326e+00 -3.94720346e-01 -9.96246412e-02 1.14838764e-01
-3.07025671e-01 1.71984866e-01 2.28889361e-01 -2.08661407e-01
-1.67504847e+00 8.55094120e-02 -4.41607535e-01 2.22087055e-01
-6.30948186e-01 1.11594133e-01 -5.25392652e-01 1.25679016e-01
1.05969715e+00 9.55861658e-02 5.21082617e-02 -4.43841100e-01
2.92822719e-01 8.62877607e-01 5.20585060e-01 -7.82350242e-01
4.75361139e-01 3.52560654e-02 -2.68106818e-01 -8.39567959e-01
-7.21109927e-01 -6.27884686e-01 -4.96351808e-01 -3.72882396e-01
6.09032571e-01 -6.23189509e-01 -5.33529401e-01 -2.41007693e-02
-1.51208031e+00 -1.43319502e-01 -4.87086087e-01 3.06080610e-01
-7.35984087e-01 2.75756359e-01 -1.35760114e-01 -1.28483546e+00
-1.74960226e-01 -6.32836044e-01 1.09545529e+00 2.01959372e-01
-9.68470991e-01 -1.30404997e+00 1.70187190e-01 4.69098926e-01
6.45972073e-01 2.32466280e-01 6.48489892e-01 -1.09443784e+00
9.59832966e-02 -3.16966265e-01 3.52728635e-01 3.91628265e-01
4.40496743e-01 -1.17253006e-01 -1.22704315e+00 3.95208836e-01
1.89320624e-01 -9.57815349e-02 4.26861256e-01 1.50401453e-02
6.95159316e-01 -3.24045151e-01 -1.42158106e-01 2.82383878e-02
1.00861895e+00 4.75202262e-01 5.08616865e-01 3.96793574e-01
1.52684957e-01 1.30909038e+00 3.44115704e-01 1.66386008e-01
5.67438722e-01 9.09038365e-01 2.63560651e-04 4.15275574e-01
3.16734463e-02 1.09549994e-02 5.59493244e-01 5.15653312e-01
-5.10462582e-01 -3.77642214e-01 -7.73764670e-01 6.65301144e-01
-1.68036640e+00 -1.34590495e+00 -7.44881704e-02 2.05369902e+00
1.09460235e+00 -1.52047098e-01 3.34079534e-01 2.80563176e-01
1.05618310e+00 2.32414380e-01 5.89481071e-02 -9.30290937e-01
-3.41121584e-01 3.25966358e-01 -6.26052991e-02 7.46492267e-01
-8.29704344e-01 1.06088519e+00 7.53533792e+00 4.82419670e-01
-8.25262725e-01 2.63428301e-01 1.58187971e-01 1.82712272e-01
-4.63717878e-01 -8.25880393e-02 -9.51688707e-01 3.44148725e-01
1.51549792e+00 -5.65431893e-01 9.49589238e-02 7.64694929e-01
2.35555604e-01 -1.87575161e-01 -1.19509006e+00 9.05740976e-01
5.80294669e-01 -7.42986262e-01 1.59463733e-01 -2.11699158e-01
3.86129916e-01 -4.79303241e-01 -1.74154609e-01 3.79546225e-01
4.81019258e-01 -8.10674191e-01 7.78397024e-01 5.79266548e-01
1.79593056e-01 -6.76100433e-01 8.07052374e-01 2.70027936e-01
-6.96013689e-01 2.87705302e-01 8.67035463e-02 2.17818066e-01
5.13466656e-01 -9.53828916e-02 -8.76402497e-01 6.34893656e-01
5.44908643e-01 8.25520009e-02 -4.68603402e-01 6.79076731e-01
-3.68930340e-01 8.31871748e-01 -3.86212051e-01 -2.96848297e-01
2.46551067e-01 1.21459052e-01 8.60247076e-01 1.48692405e+00
2.51356333e-01 1.15043506e-01 -3.66514683e-01 6.63067818e-01
2.42580950e-01 5.78617036e-01 -6.01701379e-01 3.64794850e-01
4.90427494e-01 9.96406972e-01 -4.81260091e-01 -5.19465506e-01
-1.32061914e-01 1.11378121e+00 1.81785412e-02 1.98221311e-01
-4.35714096e-01 -3.48431617e-01 2.15243921e-01 9.27552506e-02
3.25995713e-01 1.79814190e-01 -3.92854065e-02 -5.93482256e-01
-1.37652501e-01 -5.74645042e-01 6.59857213e-01 -8.49809349e-01
-1.37000060e+00 1.14751220e+00 4.64895725e-01 -8.92066002e-01
-1.08518171e+00 -3.82216603e-01 -6.95354402e-01 1.13541901e+00
-1.20876181e+00 -1.03652453e+00 -8.13059509e-02 7.78605163e-01
6.44997835e-01 -2.47236252e-01 1.25477052e+00 -4.73583024e-03
-4.42360997e-01 6.54478908e-01 -6.96067631e-01 -7.46676326e-02
6.02705598e-01 -1.39295578e+00 6.79854006e-02 4.76855308e-01
2.07555622e-01 1.13957322e+00 1.36100054e+00 -1.98078111e-01
-5.10124147e-01 -8.27997863e-01 1.80702782e+00 -7.76054204e-01
5.27844965e-01 5.22081088e-03 -1.17605877e+00 8.30506504e-01
6.40621245e-01 -4.04949278e-01 1.33036804e+00 5.05884945e-01
-2.90167809e-01 -9.44310427e-02 -1.47656584e+00 3.64788532e-01
9.90680754e-01 -5.14682353e-01 -1.40385425e+00 2.32604891e-01
3.39095116e-01 -1.07578255e-01 -8.01006258e-01 -9.36426781e-03
-1.36681292e-02 -1.13381743e+00 5.01072824e-01 -3.87212753e-01
-1.95385560e-01 -3.10710728e-01 -2.96084464e-01 -1.57779396e+00
-6.36806339e-02 -1.08518267e+00 2.24573046e-01 1.92969251e+00
5.62451184e-01 -8.61267984e-01 4.82595652e-01 8.26174736e-01
-2.70464003e-01 -1.47010788e-01 -1.41162837e+00 -8.83181036e-01
1.79495648e-01 -4.17516083e-01 1.03231847e+00 8.82040501e-01
3.87072593e-01 9.19273913e-01 1.34330764e-01 -6.64745793e-02
3.79746675e-01 -2.99508810e-01 5.72223604e-01 -1.43135142e+00
7.38438815e-02 -8.38728428e-01 -5.00969231e-01 -6.53831899e-01
6.84916496e-01 -7.07089126e-01 1.25188008e-01 -1.34469128e+00
-4.32080030e-01 -4.20837015e-01 -3.25228900e-01 3.89962614e-01
-2.01570868e-01 1.75803035e-01 -1.85113192e-01 1.41669169e-01
-5.30620933e-01 5.94978571e-01 6.10916257e-01 2.50097901e-01
-5.92192888e-01 3.74241590e-01 -1.09551990e+00 7.68457115e-01
9.24620867e-01 -3.76915008e-01 -3.23481888e-01 -8.61044787e-03
4.58689779e-02 -5.17057255e-02 2.57450402e-01 -8.13213766e-01
4.15179245e-02 1.18976504e-01 -1.08405635e-01 -3.94811094e-01
5.43527544e-01 -4.62977976e-01 2.28371993e-01 -2.03166977e-01
-4.99474555e-01 7.27936700e-02 1.79316923e-01 2.94109792e-01
-4.77001101e-01 -6.40592337e-01 7.67884731e-01 -3.06465685e-01
-7.02085257e-01 -3.07091534e-01 -6.61942422e-01 -1.84719950e-01
1.01483071e+00 -4.78743136e-01 1.33195713e-01 -6.25894964e-01
-9.08077478e-01 -1.92707524e-01 1.59602523e-01 2.99294084e-01
2.52729595e-01 -1.30304384e+00 -7.94476449e-01 -2.79801428e-01
5.02272666e-01 -1.33194298e-01 2.94794053e-01 5.17459333e-01
-1.06374465e-01 4.00308818e-01 -6.67796209e-02 -3.14965457e-01
-1.36062932e+00 4.85507965e-01 4.39236403e-01 1.79752596e-02
-3.69886100e-01 1.18421352e+00 -4.33317274e-01 -2.83979774e-01
-6.19268753e-02 1.78954795e-01 -9.29409385e-01 6.25295043e-01
5.52079856e-01 2.45207980e-01 7.40190968e-02 -1.57138908e+00
-3.49264085e-01 3.39087814e-01 -2.93236990e-02 -7.93260098e-01
1.15530384e+00 -3.82266194e-01 -5.36611900e-02 1.07417774e+00
8.28872323e-01 3.26457322e-01 -7.24157095e-01 -3.83381337e-01
3.58092338e-01 -2.50828713e-01 -5.00311283e-03 -5.20333111e-01
-5.01331449e-01 4.82588023e-01 2.52419442e-01 7.44277179e-01
1.05389786e+00 4.96066511e-01 1.92595944e-01 2.10359842e-01
5.25757015e-01 -1.31416750e+00 -3.80677819e-01 2.79600978e-01
1.29256904e+00 -8.86541843e-01 -1.72589645e-01 -4.83493894e-01
-9.62705672e-01 8.81811976e-01 2.93755412e-01 2.24263132e-01
4.63540286e-01 -2.30704620e-01 6.85075641e-01 -1.73975751e-01
-7.99044669e-01 -3.44367862e-01 7.67351836e-02 1.11497760e+00
1.00674891e+00 8.21472928e-02 -6.50100648e-01 7.63075352e-01
-8.26570749e-01 -3.14201564e-01 4.64798063e-01 9.32764292e-01
-2.47101545e-01 -1.43564129e+00 -6.63553655e-01 -1.33738577e-01
-6.25046670e-01 9.96889081e-04 -7.01021373e-01 8.11399400e-01
1.84187517e-01 1.42044699e+00 1.59231089e-02 1.91968918e-01
8.44303131e-01 8.36441934e-01 3.89461964e-01 -8.60135019e-01
-1.06860542e+00 9.11866352e-02 6.85356915e-01 -3.22921425e-01
-1.13999951e+00 -1.31083393e+00 -1.05125213e+00 -1.17224239e-01
-3.65326941e-01 5.77844262e-01 7.21341670e-01 8.35391223e-01
3.71699296e-02 4.51439351e-01 7.13214815e-01 -8.70255589e-01
-5.24083614e-01 -1.31531143e+00 -5.87518573e-01 5.23671985e-01
1.91844136e-01 -7.34145582e-01 -6.06499970e-01 3.84633541e-01] | [10.564905166625977, 9.16638469696045] |
30d40dd1-fe3c-437e-ac12-272d7c0da3f3 | soundclr-contrastive-learning-of | 2103.01929 | null | https://arxiv.org/abs/2103.01929v1 | https://arxiv.org/pdf/2103.01929v1.pdf | SoundCLR: Contrastive Learning of Representations For Improved Environmental Sound Classification | Environmental Sound Classification (ESC) is a challenging field of research in non-speech audio processing. Most of current research in ESC focuses on designing deep models with special architectures tailored for specific audio datasets, which usually cannot exploit the intrinsic patterns in the data. However recent studies have surprisingly shown that transfer learning from models trained on ImageNet is a very effective technique in ESC. Herein, we propose SoundCLR, a supervised contrastive learning method for effective environment sound classification with state-of-the-art performance, which works by learning representations that disentangle the samples of each class from those of other classes. Our deep network models are trained by combining a contrastive loss that contributes to a better probability output by the classification layer with a cross-entropy loss on the output of the classifier layer to map the samples to their respective 1-hot encoded labels. Due to the comparatively small sizes of the available environmental sound datasets, we propose and exploit a transfer learning and strong data augmentation pipeline and apply the augmentations on both the sound signals and their log-mel spectrograms before inputting them to the model. Our experiments show that our masking based augmentation technique on the log-mel spectrograms can significantly improve the recognition performance. Our extensive benchmark experiments show that our hybrid deep network models trained with combined contrastive and cross-entropy loss achieved the state-of-the-art performance on three benchmark datasets ESC-10, ESC-50, and US8K with validation accuracies of 99.75\%, 93.4\%, and 86.49\% respectively. The ensemble version of our models also outperforms other top ensemble methods. The code is available at https://github.com/alireza-nasiri/SoundCLR. | ['Jianjun Hu', 'Alireza Nasiri'] | 2021-03-02 | null | null | null | null | ['environmental-sound-classification', 'sound-classification'] | ['audio', 'audio'] | [ 3.88797939e-01 -4.71575588e-01 2.96469867e-01 -3.16481739e-01
-1.07010508e+00 -3.42552453e-01 1.88194677e-01 -1.74018711e-01
-4.79876071e-01 4.85293746e-01 4.12860550e-02 -2.46841341e-01
1.70950294e-01 -7.20279038e-01 -8.23773265e-01 -8.39173198e-01
-2.16503829e-01 -3.02247740e-02 1.56492904e-01 -1.42177746e-01
-1.03744686e-01 1.04763202e-01 -1.91148102e+00 7.26875484e-01
6.64888442e-01 1.45405257e+00 1.05817728e-01 1.10995650e+00
2.57140994e-01 7.36795664e-01 -6.36729896e-01 -1.38501078e-01
3.22872460e-01 -4.52157944e-01 -4.43279654e-01 -6.45603001e-01
5.24430275e-01 -1.41351506e-01 -4.26278055e-01 1.15664959e+00
9.84012723e-01 1.69141337e-01 6.74689889e-01 -1.21141469e+00
-3.66547287e-01 9.35597301e-01 -3.69679153e-01 3.51633340e-01
6.43918738e-02 2.60063171e-01 1.17356038e+00 -9.45708990e-01
-1.28391996e-01 1.00324380e+00 9.02493417e-01 5.89677095e-01
-1.12397349e+00 -1.37951887e+00 -7.33474642e-02 5.07680416e-01
-1.43429875e+00 -7.46217191e-01 8.30313087e-01 -3.84383112e-01
1.04833388e+00 5.49064279e-01 4.18559849e-01 1.25951517e+00
-1.93284839e-01 6.51414871e-01 1.21766746e+00 -6.21651411e-01
2.86005139e-01 2.01933160e-01 1.46062383e-02 3.92372549e-01
-3.72864366e-01 4.14245188e-01 -9.47904289e-01 -6.75642043e-02
1.83435395e-01 -4.62792605e-01 -4.01036203e-01 3.18167120e-01
-6.40647650e-01 5.84135234e-01 5.20423353e-01 2.03028545e-01
-2.72535592e-01 3.58565718e-01 4.55440223e-01 3.21509212e-01
6.35905206e-01 3.11235577e-01 -6.24570370e-01 -2.12049276e-01
-9.90347505e-01 1.40491202e-01 5.34218192e-01 4.19974178e-01
5.46199262e-01 4.88628089e-01 -1.41004801e-01 1.17556047e+00
2.83034563e-01 6.22864127e-01 6.51137292e-01 -6.00165546e-01
4.83550519e-01 8.20930451e-02 -3.61461848e-01 -7.69703567e-01
-1.26835808e-01 -9.48685884e-01 -8.70464981e-01 3.04204077e-01
-1.01386577e-01 -1.12531535e-01 -1.02614486e+00 1.97395134e+00
6.86750934e-02 7.86207736e-01 1.36735171e-01 7.69424260e-01
9.03322697e-01 8.80002975e-01 1.73727170e-01 1.00443713e-01
1.18931532e+00 -1.06658590e+00 -6.45854294e-01 -3.04568589e-01
1.88358501e-01 -7.63721406e-01 1.23964083e+00 4.95431453e-01
-8.42661381e-01 -8.80786002e-01 -1.29818404e+00 1.87836558e-01
-5.47402084e-01 4.40103948e-01 2.93536156e-01 8.19378376e-01
-8.57283533e-01 6.08949840e-01 -8.53910983e-01 3.12774897e-01
6.83579862e-01 4.47025269e-01 -1.31617457e-01 3.29756737e-01
-1.28050029e+00 3.14040035e-01 3.21607351e-01 3.53108421e-02
-1.23616374e+00 -9.44615304e-01 -5.61971068e-01 1.90377176e-01
1.35618195e-01 -3.57707411e-01 1.37392867e+00 -1.05636549e+00
-1.63832247e+00 5.77726960e-01 1.45515919e-01 -7.85714030e-01
4.81163114e-01 -5.44971704e-01 -7.13074565e-01 2.84823142e-02
-5.66997379e-02 6.50782228e-01 9.80714440e-01 -1.15916336e+00
-7.29099035e-01 -2.39968807e-01 -3.46717119e-01 7.05761611e-02
-6.42867029e-01 9.64937732e-02 3.84195372e-02 -8.67727399e-01
-2.53591776e-01 -8.78078341e-01 1.75036073e-01 -3.04558754e-01
-4.34665382e-01 -2.77629010e-02 7.95901716e-01 -8.51085186e-01
1.40410876e+00 -2.52455807e+00 -6.45082593e-02 -1.41648963e-01
-1.18119955e-01 6.93152905e-01 -2.84489214e-01 1.38027087e-01
-2.96255648e-01 1.14171080e-01 -5.41168571e-01 -4.96719927e-01
1.20616697e-01 -1.94977298e-01 -6.35129929e-01 1.63598403e-01
3.59178483e-01 3.85693461e-01 -8.32541823e-01 5.91362342e-02
6.31586686e-02 7.91261137e-01 -6.28468633e-01 4.17024642e-01
-1.69886112e-01 3.55775952e-01 6.21922947e-02 4.00965273e-01
7.81078875e-01 3.34815979e-01 -1.80019721e-01 -4.87888679e-02
3.98632847e-02 7.37020314e-01 -1.19854271e+00 1.61379933e+00
-7.90562987e-01 5.90312600e-01 1.32824138e-01 -8.51114511e-01
8.00469756e-01 5.22599339e-01 1.31310254e-01 -4.95627075e-01
1.26459152e-01 5.76686084e-01 2.94951051e-01 -3.65949303e-01
2.94635355e-01 -2.17881635e-01 9.98062193e-02 2.30142415e-01
2.86931962e-01 -1.44888133e-01 -2.65391797e-01 -2.39010751e-01
8.94631147e-01 2.74640676e-02 -1.05868198e-01 -5.42338416e-02
5.09709537e-01 -5.50225377e-01 4.70242709e-01 7.11790621e-01
-1.60122812e-01 6.91995800e-01 -2.84791421e-02 -1.08283386e-01
-7.17080772e-01 -1.00114834e+00 -3.61326575e-01 1.41414106e+00
-3.96532804e-01 -3.54286402e-01 -8.98309231e-01 -5.87135255e-01
-1.01202913e-01 8.08577120e-01 -5.17687440e-01 -5.24718523e-01
-4.15939420e-01 -9.51708138e-01 1.10703075e+00 7.07649827e-01
7.03259468e-01 -1.14962220e+00 -5.32634020e-01 2.27247253e-01
-1.84192315e-01 -1.07101953e+00 -1.72763318e-01 6.86514139e-01
-4.51766878e-01 -6.76000953e-01 -4.34640080e-01 -7.15360403e-01
-2.14311797e-02 -1.54376239e-01 9.04823720e-01 -8.13647658e-02
-1.67954013e-01 2.23060250e-02 -4.26512957e-01 -8.86018753e-01
-5.93637764e-01 2.14380443e-01 2.19388947e-01 3.90580505e-01
3.59447479e-01 -9.07948554e-01 -5.95018268e-01 9.22234133e-02
-8.54012311e-01 -7.54387528e-02 2.60931373e-01 9.28626776e-01
5.84990919e-01 2.04420879e-01 9.86957312e-01 -4.34098154e-01
4.32635546e-01 -4.97784644e-01 -2.64656156e-01 -2.23879755e-01
-2.87978560e-01 -1.79080188e-01 7.81237483e-01 -4.30673838e-01
-8.86030734e-01 4.90970947e-02 -6.61567509e-01 -4.62895006e-01
-3.86828214e-01 2.19593182e-01 -1.75966159e-01 3.25670242e-01
7.68060088e-01 2.03586832e-01 -3.05962890e-01 -8.08224082e-01
1.62854552e-01 1.36512113e+00 6.98791146e-01 -4.80527133e-01
5.30574858e-01 2.69909948e-01 -2.67037451e-01 -7.98827291e-01
-8.76323342e-01 -3.62332284e-01 -2.44306073e-01 -1.52604803e-01
8.51433754e-01 -1.22776198e+00 -3.93196970e-01 8.28307271e-01
-7.78320491e-01 -5.11414886e-01 -3.26371610e-01 5.78908026e-01
-3.17204177e-01 -4.06917706e-02 -4.74802375e-01 -1.18185318e+00
-5.80183566e-01 -1.24412429e+00 1.02708399e+00 6.42308742e-02
4.33755443e-02 -5.25201201e-01 1.57947242e-01 2.63140917e-01
6.81983709e-01 -5.11361845e-02 8.68965566e-01 -9.17391121e-01
-2.51739621e-01 -1.32713348e-01 1.67370364e-01 1.03710556e+00
1.34135202e-01 1.23355095e-03 -2.00078607e+00 -1.07711382e-01
-1.05265528e-01 -4.24847126e-01 1.46058369e+00 2.14730725e-01
1.53373301e+00 -2.74814904e-01 1.13543823e-01 6.81010187e-01
1.25585377e+00 3.93953860e-01 6.29468322e-01 1.88406825e-01
5.26100993e-01 2.83870012e-01 2.48257577e-01 2.92386681e-01
7.84565434e-02 7.71838665e-01 5.05508423e-01 1.59226269e-01
-6.13065898e-01 -3.79887998e-01 6.92854524e-01 1.16903186e+00
6.47213385e-02 -4.04659569e-01 -7.59961843e-01 5.86417794e-01
-1.50345862e+00 -8.23628128e-01 2.21519060e-02 2.14658713e+00
1.04833603e+00 3.09295598e-02 1.43394312e-02 8.71748149e-01
6.41303480e-01 1.79737106e-01 -3.96493793e-01 -4.21320468e-01
-1.83992669e-01 8.81258190e-01 4.41455096e-02 3.32388580e-01
-1.63830018e+00 8.45668495e-01 5.23671818e+00 1.16979992e+00
-1.39819264e+00 2.73370475e-01 7.70447135e-01 -2.83992976e-01
1.82943106e-01 -4.93541837e-01 -7.31273890e-01 5.63727498e-01
1.41132689e+00 3.07724684e-01 6.72660470e-01 6.75057173e-01
-4.56823446e-02 2.94288725e-01 -1.05415022e+00 9.73919272e-01
-6.82322681e-02 -9.20305550e-01 -6.27208799e-02 -2.13597029e-01
6.63248003e-01 2.92589366e-01 6.72124505e-01 5.03510118e-01
6.00703917e-02 -1.18669236e+00 1.06405532e+00 2.91306049e-01
1.00525653e+00 -8.33335817e-01 7.26088762e-01 1.26452804e-01
-1.27137339e+00 -3.53898466e-01 -3.43306631e-01 -1.90364450e-01
-2.33310148e-01 4.82080191e-01 -8.46813679e-01 5.88916004e-01
1.14898622e+00 4.68216360e-01 -4.63255972e-01 9.62938726e-01
-2.39592224e-01 1.28800309e+00 -4.23473746e-01 3.40109095e-02
1.53947426e-02 1.67428106e-01 6.72144353e-01 1.55632985e+00
4.79937553e-01 -2.68593013e-01 -1.45910755e-01 8.39110970e-01
-2.89242655e-01 2.07154691e-01 -3.59995157e-01 3.95984314e-02
5.29818118e-01 1.02464998e+00 -2.45891914e-01 -2.73185909e-01
-5.84568530e-02 5.40555775e-01 2.04403609e-01 1.50544420e-01
-8.91576767e-01 -6.46961749e-01 8.49869967e-01 -1.34571210e-01
4.51954544e-01 2.34664664e-01 -3.50305498e-01 -8.36927950e-01
8.10274407e-02 -1.12704396e+00 3.06276947e-01 -6.34985864e-01
-1.11961102e+00 1.01893532e+00 -2.41617978e-01 -1.26736557e+00
-1.03345916e-01 -6.76604688e-01 -7.01220989e-01 8.85979295e-01
-1.76832461e+00 -1.00840294e+00 -2.40515202e-01 2.99992323e-01
5.06101906e-01 -5.24872780e-01 1.09781992e+00 6.47942364e-01
-6.16803408e-01 1.00382137e+00 9.03407186e-02 2.31396109e-01
6.99578345e-01 -1.31917369e+00 3.23345900e-01 7.43541062e-01
5.20613432e-01 1.60198420e-01 5.02153635e-01 -1.09076150e-01
-9.29445565e-01 -1.32714498e+00 4.66309100e-01 -2.06081003e-01
5.14863312e-01 -5.55848420e-01 -1.05022728e+00 3.22352141e-01
3.64805400e-01 2.18191251e-01 9.24610674e-01 -1.82643071e-01
-6.30524993e-01 -5.05605578e-01 -8.66297245e-01 2.85777152e-01
9.51789856e-01 -8.75870705e-01 -3.81489456e-01 2.87735835e-02
8.34726572e-01 -2.75634587e-01 -6.03597999e-01 6.45281911e-01
6.39306545e-01 -9.18830454e-01 9.05236244e-01 -7.08759010e-01
5.57450175e-01 -3.51655930e-01 -7.08941281e-01 -1.50612342e+00
-3.80328968e-02 -4.16250497e-01 -6.81786612e-02 1.41935503e+00
5.66141427e-01 -5.68447471e-01 2.90825814e-01 -8.55373591e-02
-4.70158815e-01 -6.85968995e-01 -1.16813517e+00 -7.72425354e-01
2.06010535e-01 -1.06503892e+00 8.64295721e-01 7.35249043e-01
-4.29117858e-01 3.57962579e-01 -2.72187859e-01 4.42852795e-01
3.01551402e-01 -2.11633340e-01 6.12906694e-01 -1.13421738e+00
-5.05836546e-01 -4.31512713e-01 -4.43129122e-01 -6.96479082e-01
3.34941506e-01 -1.14832509e+00 2.56596655e-01 -9.61412668e-01
2.02603191e-02 -5.35451233e-01 -9.40909624e-01 6.30521476e-01
-1.66239753e-01 5.51562726e-01 1.79138109e-01 -2.42217571e-01
-1.97658241e-01 7.70452440e-01 7.66540587e-01 -3.33184481e-01
-3.09692383e-01 1.43881023e-01 -5.10818481e-01 7.23969936e-01
1.05320907e+00 -6.74510121e-01 -1.93097055e-01 -5.07963777e-01
1.15294732e-01 -3.76369059e-01 6.21031761e-01 -1.46103680e+00
2.75754072e-02 3.89291048e-01 1.84091434e-01 -2.85538822e-01
6.69904530e-01 -6.27278149e-01 1.39509097e-01 3.38990599e-01
-6.25521302e-01 -4.54797029e-01 5.26269674e-01 5.45551062e-01
-4.44378078e-01 -1.38908267e-01 9.70757544e-01 9.53874514e-02
-3.42413276e-01 8.59760344e-02 -2.21956357e-01 5.26776575e-02
4.73256499e-01 2.90765166e-01 -2.57111907e-01 -3.27147871e-01
-7.38000572e-01 -3.25693607e-01 -2.69338757e-01 5.21402419e-01
4.86293316e-01 -1.39813077e+00 -1.00465477e+00 5.24677038e-01
-5.18908538e-02 -8.98006707e-02 4.93064314e-01 6.35993242e-01
-1.97293743e-01 2.32901216e-01 -6.54070973e-02 -6.45554781e-01
-1.34745502e+00 1.02246195e-01 7.08404720e-01 -2.71774810e-02
-3.26901138e-01 1.20580363e+00 3.51831555e-01 -6.35075629e-01
5.47969878e-01 -4.82861847e-01 -2.35547677e-01 -3.75053436e-02
6.12608671e-01 4.76784021e-01 4.63495702e-01 -6.98915899e-01
-4.18545961e-01 3.69528443e-01 1.23595603e-01 -2.27213994e-01
1.66474771e+00 3.58503610e-01 9.18766409e-02 5.96551180e-01
1.45767224e+00 1.20288014e-01 -1.22429943e+00 -2.55084097e-01
-4.22455877e-01 -2.93149322e-01 4.03398752e-01 -1.11715496e+00
-1.26335859e+00 1.37715220e+00 1.13550389e+00 2.54423022e-01
1.57124138e+00 -2.90342182e-01 7.58327007e-01 1.09581612e-01
-3.19221988e-02 -1.00998640e+00 1.16009422e-01 5.04965961e-01
1.07172704e+00 -9.80393469e-01 -4.51133937e-01 -2.38037318e-01
-4.55737948e-01 8.88832688e-01 4.96222466e-01 -2.51888745e-02
9.75483179e-01 5.22021592e-01 2.26016164e-01 2.37865642e-01
-9.28257823e-01 -1.95765033e-01 3.73564482e-01 5.57714999e-01
3.27074200e-01 3.72067571e-01 2.80779392e-01 1.23284113e+00
-5.89566588e-01 -2.11751834e-01 2.24918455e-01 6.58933818e-01
-3.81221414e-01 -8.77387404e-01 -3.12197924e-01 3.67297053e-01
-9.73474920e-01 -3.13255429e-01 -3.73720586e-01 2.85611689e-01
3.03544879e-01 1.20242691e+00 7.86965862e-02 -9.16883051e-01
2.83301353e-01 4.43943918e-01 1.56151220e-01 -5.27113855e-01
-8.53935480e-01 2.25752428e-01 8.79546478e-02 -1.00605957e-01
-2.99495280e-01 -5.04743636e-01 -1.10887241e+00 2.75159836e-01
-4.64559287e-01 1.56054229e-01 7.95723677e-01 5.54507375e-01
4.71711755e-01 1.01086807e+00 8.52008462e-01 -8.06373179e-01
-7.94242322e-01 -1.29509115e+00 -5.65746605e-01 1.62540883e-01
6.99986041e-01 -6.02977276e-01 -6.76710606e-01 1.27132803e-01] | [15.130382537841797, 5.224027156829834] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.