Split (2)

train · 44.6k rows

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
46df8c80-a48f-462d-8276-359c25e159f0	capsulevos-semi-supervised-video-object	1910.00132	null	https://arxiv.org/abs/1910.00132v1	https://arxiv.org/pdf/1910.00132v1.pdf	CapsuleVOS: Semi-Supervised Video Object Segmentation Using Capsule Routing	In this work we propose a capsule-based approach for semi-supervised video object segmentation. Current video object segmentation methods are frame-based and often require optical flow to capture temporal consistency across frames which can be difficult to compute. To this end, we propose a video based capsule network, CapsuleVOS, which can segment several frames at once conditioned on a reference frame and segmentation mask. This conditioning is performed through a novel routing algorithm for attention-based efficient capsule selection. We address two challenging issues in video object segmentation: 1) segmentation of small objects and 2) occlusion of objects across time. The issue of segmenting small objects is addressed with a zooming module which allows the network to process small spatial regions of the video. Apart from this, the framework utilizes a novel memory module based on recurrent networks which helps in tracking objects when they move out of frame or are occluded. The network is trained end-to-end and we demonstrate its effectiveness on two benchmark video object segmentation datasets; it outperforms current offline approaches on the Youtube-VOS dataset while having a run-time that is almost twice as fast as competing methods. The code is publicly available at https://github.com/KevinDuarte/CapsuleVOS.	['Yogesh S Rawat', 'Mubarak Shah', 'Kevin Duarte']	2019-09-30	capsulevos-semi-supervised-video-object-1	http://openaccess.thecvf.com/content_ICCV_2019/html/Duarte_CapsuleVOS_Semi-Supervised_Video_Object_Segmentation_Using_Capsule_Routing_ICCV_2019_paper.html	http://openaccess.thecvf.com/content_ICCV_2019/papers/Duarte_CapsuleVOS_Semi-Supervised_Video_Object_Segmentation_Using_Capsule_Routing_ICCV_2019_paper.pdf	iccv-2019-10	['one-shot-visual-object-segmentation']	['computer-vision']	[ 2.91629676e-02 -2.34066606e-01 -4.43308711e-01 -7.77366087e-02 -6.46607637e-01 -7.56859899e-01 -8.53486918e-03 -5.98351918e-02 -5.37830651e-01 3.92413050e-01 -1.54215470e-01 -7.57643059e-02 1.90948009e-01 -3.30660164e-01 -8.82576406e-01 -5.95920622e-01 -2.65285552e-01 2.82538354e-01 7.27174640e-01 3.62141043e-01 1.59467384e-01 3.71902764e-01 -1.08916616e+00 2.25617722e-01 6.53180242e-01 1.02099442e+00 4.66038138e-01 8.89329135e-01 5.51125668e-02 1.01992023e+00 -3.83826435e-01 1.00342683e-01 4.42703396e-01 -4.52089131e-01 -8.35528374e-01 5.65953135e-01 4.32756573e-01 -4.53227222e-01 -4.04701769e-01 9.94058847e-01 3.49636227e-01 3.14907819e-01 2.04771206e-01 -1.17923152e+00 -1.33930281e-01 5.23801446e-01 -4.69216436e-01 5.77024579e-01 2.80571312e-01 1.96134448e-01 5.22481620e-01 -7.40797997e-01 1.06369138e+00 9.43307877e-01 4.73419547e-01 7.86199272e-01 -1.02293003e+00 -3.71264488e-01 6.11699402e-01 1.41446695e-01 -1.34818029e+00 -4.50257093e-01 8.26523423e-01 -5.32564521e-01 6.31165028e-01 1.88395068e-01 9.54863906e-01 7.12236643e-01 2.17979848e-01 1.25186324e+00 5.72201014e-01 3.21426764e-02 3.73772681e-01 -3.33007909e-02 1.16987772e-01 6.18165255e-01 -1.28111407e-01 -1.16589077e-01 -3.94592226e-01 1.51965261e-01 1.01959956e+00 3.15551072e-01 -7.73941159e-01 -5.12547553e-01 -1.21382093e+00 7.92957842e-01 3.51430297e-01 1.69966325e-01 -3.13858449e-01 4.64291781e-01 4.43573058e-01 7.07769319e-02 4.55509692e-01 1.12333044e-01 -5.08384764e-01 2.35546473e-02 -1.48124635e+00 1.27797037e-01 9.95137990e-01 1.09625459e+00 3.14568788e-01 6.46187663e-02 -3.71211261e-01 5.12886286e-01 5.22073686e-01 6.07754588e-02 3.58845919e-01 -1.14141262e+00 1.10649720e-01 3.23059678e-01 5.87675609e-02 -7.94765353e-01 -2.81712055e-01 -2.68003315e-01 -5.19591153e-01 2.71236688e-01 4.51851219e-01 -1.85434431e-01 -1.21441555e+00 1.40080822e+00 6.91441357e-01 7.87228107e-01 -1.48388743e-01 1.29377317e+00 1.12806547e+00 7.58828342e-01 1.74420193e-01 -4.30831075e-01 1.26182306e+00 -1.58487737e+00 -7.53221929e-01 7.61353597e-03 3.78732413e-01 -8.51987541e-01 4.37693805e-01 4.15223718e-01 -1.35454917e+00 -3.89281005e-01 -9.55774248e-01 4.97599356e-02 -2.72160202e-01 -1.32978270e-02 5.23961127e-01 4.36738372e-01 -1.10127008e+00 7.11374342e-01 -1.25901997e+00 -1.42411694e-01 7.32791066e-01 7.06458390e-01 3.38129839e-03 5.55403810e-03 -6.61019206e-01 2.84461945e-01 5.40692449e-01 3.28866035e-01 -1.25877583e+00 -7.56552041e-01 -9.49707389e-01 -9.09667462e-02 7.99985051e-01 -6.39378548e-01 1.19707382e+00 -1.41023695e+00 -1.61262834e+00 4.74860609e-01 -3.83830041e-01 -6.50872290e-01 8.30404460e-01 -2.26287067e-01 -6.77942038e-02 6.77896202e-01 7.37666339e-02 9.45837975e-01 1.02342284e+00 -1.02839816e+00 -5.90642631e-01 2.97663677e-02 -2.57990621e-02 2.86086470e-01 1.02545373e-01 2.26408526e-01 -1.35602915e+00 -9.40892696e-01 -3.15635614e-02 -1.06465185e+00 -5.04650831e-01 3.34293276e-01 -6.50338888e-01 -2.12768704e-01 1.06471336e+00 -6.34568274e-01 1.12553275e+00 -1.93963552e+00 3.55646938e-01 1.27329513e-01 4.23029393e-01 4.03337181e-01 -1.41206875e-01 -1.59675777e-01 4.55866158e-02 4.34865914e-02 -2.53472209e-01 -4.94615465e-01 -4.72643882e-01 -9.09929946e-02 7.62892142e-02 7.86594748e-01 9.83598754e-02 1.03543866e+00 -8.33612382e-01 -7.20561922e-01 5.72466016e-01 6.63078964e-01 -6.97390199e-01 3.11351240e-01 -5.40067077e-01 7.89028704e-01 -4.77689803e-01 1.00403714e+00 7.36713111e-01 -5.10021806e-01 9.60777029e-02 -2.52144754e-01 -2.40651444e-01 -1.79826230e-01 -1.32240307e+00 2.01779413e+00 5.80742732e-02 6.02230668e-01 2.80605644e-01 -8.82454276e-01 2.08944723e-01 3.86814773e-01 9.44612801e-01 -1.53699473e-01 4.41243082e-01 8.41597468e-02 -3.19308788e-01 -6.29454851e-01 3.29918146e-01 5.01033425e-01 3.25029790e-01 1.56570762e-01 4.04296480e-02 6.87953606e-02 7.42908001e-01 3.11279804e-01 9.50602770e-01 5.83183408e-01 -2.39017069e-01 -3.69252950e-01 5.65307021e-01 2.07116902e-01 8.92266810e-01 6.16458416e-01 -3.75989735e-01 8.33821177e-01 3.47085357e-01 -3.95686150e-01 -7.15247512e-01 -8.55230510e-01 -2.54710130e-02 7.74337053e-01 7.86101520e-01 -5.26436210e-01 -8.84645164e-01 -7.84596920e-01 -2.75426388e-01 5.33160940e-02 -4.94130731e-01 2.65189618e-01 -7.84921706e-01 -4.69724208e-01 -1.04703531e-01 5.65882742e-01 2.52328187e-01 -1.06398630e+00 -8.73576939e-01 5.12659848e-01 -1.04005225e-01 -1.45316386e+00 -9.52128947e-01 1.78568646e-01 -9.57059383e-01 -1.40020013e+00 -9.83751416e-01 -9.84454989e-01 8.36593628e-01 3.94482464e-01 1.01233482e+00 2.89464325e-01 -4.04269099e-01 5.42154253e-01 -2.79199213e-01 3.35110053e-02 -7.53434375e-02 2.63309926e-02 -2.35133737e-01 -1.00451432e-01 2.92799678e-02 -1.81421995e-01 -1.05048299e+00 4.27165717e-01 -1.09801531e+00 9.51507688e-02 2.06540540e-01 5.29317617e-01 1.00614572e+00 -1.89461380e-01 2.20043749e-01 -8.62342119e-01 -4.69188578e-02 -5.39882541e-01 -9.68114913e-01 2.31049553e-01 -3.33341095e-03 -3.66138011e-01 4.98450935e-01 -7.53380835e-01 -5.72584212e-01 6.17700517e-01 8.31615627e-02 -9.87282515e-01 1.92106012e-02 1.75126716e-01 1.26974523e-01 -3.66079122e-01 -1.41867861e-01 1.01127312e-01 9.26492661e-02 -3.13443333e-01 1.25665709e-01 1.53591663e-01 4.61801589e-01 -2.01142892e-01 6.15600824e-01 7.74842858e-01 -3.99018198e-01 -7.42229402e-01 -6.20950043e-01 -7.92019248e-01 -5.94484448e-01 -5.76997340e-01 1.25758445e+00 -1.15842783e+00 -9.05715883e-01 4.14079428e-01 -1.18351519e+00 -6.79052472e-01 -1.63654566e-01 5.80163300e-01 -4.70824450e-01 4.47570324e-01 -9.51935410e-01 -4.82671678e-01 -3.97848159e-01 -1.61166668e+00 9.72890794e-01 5.80575347e-01 -1.34508803e-01 -1.06219697e+00 -2.56937534e-01 3.02809924e-01 3.49747598e-01 4.68907863e-01 6.39737993e-02 -4.59757566e-01 -1.31887078e+00 9.02649388e-02 2.75037754e-02 3.37176770e-02 -3.06294281e-02 1.55229077e-01 -5.63916326e-01 -5.55661738e-01 4.23380621e-02 -3.56466398e-02 1.03242838e+00 1.05962610e+00 1.34286952e+00 -9.48792696e-02 -5.39237142e-01 9.71389413e-01 1.53625619e+00 3.96467477e-01 4.39555973e-01 2.21683323e-01 9.30422962e-01 2.96554178e-01 5.91600537e-01 3.16301823e-01 1.66704938e-01 6.94196999e-01 5.11853337e-01 -2.84321219e-01 -2.98355550e-01 2.00041339e-01 3.93033653e-01 6.62667871e-01 1.48539990e-01 -4.77184296e-01 -5.14025211e-01 7.49406695e-01 -1.98705578e+00 -6.85616374e-01 -1.01967432e-01 1.97933042e+00 6.21908128e-01 9.27115604e-02 2.17893586e-01 -2.98484147e-01 8.93109262e-01 1.63040444e-01 -7.44878590e-01 1.27455443e-01 3.32275070e-02 -5.91403358e-02 5.21045983e-01 5.93245208e-01 -1.59223902e+00 1.17387021e+00 5.31597567e+00 6.71286821e-01 -1.38040888e+00 1.95154384e-01 8.58708918e-01 -5.09610534e-01 4.92372736e-02 -4.49517034e-02 -8.80525231e-01 6.56894863e-01 5.13765037e-01 1.95562541e-01 3.43953013e-01 7.78577209e-01 4.50012594e-01 -4.09205228e-01 -1.11956620e+00 9.81225908e-01 1.93368092e-01 -1.72562563e+00 -1.85025796e-01 -2.36700758e-01 7.29220212e-01 2.01227456e-01 -1.38496608e-01 -1.83943346e-01 1.33912954e-02 -8.72708917e-01 1.04831147e+00 3.37727100e-01 5.34056008e-01 -6.61798894e-01 4.17049974e-01 -1.46595031e-01 -1.48201632e+00 1.19540177e-01 -1.45249903e-01 4.71074730e-01 6.23389006e-01 4.71016586e-01 -3.42484951e-01 3.54217827e-01 8.59371841e-01 1.06745565e+00 -4.74564552e-01 1.58055425e+00 -1.01136919e-02 4.68465179e-01 -5.06526887e-01 2.21554667e-01 3.98213595e-01 -2.02024788e-01 7.52120376e-01 1.22876525e+00 1.07779913e-01 5.44502810e-02 7.49440789e-01 8.28521192e-01 -5.37317097e-02 -7.40977526e-02 -1.70869365e-01 -5.41274296e-03 2.28333116e-01 1.37296307e+00 -1.48493421e+00 -4.95413244e-01 -4.19678509e-01 1.17194295e+00 -9.10565406e-02 6.02656484e-01 -1.16029632e+00 -9.78322849e-02 3.14823419e-01 1.25984559e-02 9.83902335e-01 -2.02342331e-01 2.14176312e-01 -1.30048978e+00 1.27223864e-01 -5.86484671e-01 3.99093926e-01 -5.10629475e-01 -7.51935244e-01 7.71036208e-01 -2.92145222e-01 -1.39235508e+00 -3.45798931e-03 -4.63157356e-01 -6.15623176e-01 2.10816085e-01 -1.67771804e+00 -9.11625683e-01 -3.48376870e-01 7.42898941e-01 1.02111411e+00 1.22356921e-01 1.97550252e-01 5.47145247e-01 -8.21821272e-01 1.57385454e-01 -1.00618474e-01 3.60526860e-01 5.22351205e-01 -9.84550774e-01 2.18415167e-03 1.03023541e+00 1.07083812e-01 5.09111881e-01 5.69092393e-01 -8.69838893e-01 -1.60009265e+00 -1.42729771e+00 2.68368363e-01 -2.40932077e-01 3.40041459e-01 -3.85092467e-01 -6.76868200e-01 7.02766895e-01 4.69943523e-01 5.54581285e-01 5.15764475e-01 -7.18757153e-01 2.46143430e-01 1.75104350e-01 -9.32704866e-01 4.56840217e-01 7.74896204e-01 -2.72681960e-03 -2.87139416e-02 6.82829618e-01 8.75639200e-01 -8.64972532e-01 -7.95390785e-01 3.49283308e-01 2.70961970e-01 -7.56769180e-01 9.29468453e-01 -3.32691699e-01 3.21136504e-01 -8.19198012e-01 3.14106613e-01 -6.82208598e-01 1.44262254e-01 -9.79617715e-01 -3.72145295e-01 9.37530160e-01 1.79648548e-01 -1.96167052e-01 1.10001206e+00 6.55029774e-01 -7.72271231e-02 -9.88900125e-01 -8.21622968e-01 -6.73131287e-01 -4.17818934e-01 -2.92405903e-01 -5.12815826e-02 7.75728524e-01 -3.74014914e-01 -1.53110266e-01 -2.70853192e-01 1.71440795e-01 7.76670992e-01 2.98735589e-01 4.31066573e-01 -7.65544057e-01 -3.37564260e-01 -3.37912858e-01 -3.23629439e-01 -1.41690910e+00 9.86271501e-02 -7.70116091e-01 1.37292936e-01 -1.38789499e+00 2.43935168e-01 -2.61043966e-01 -1.43424183e-01 1.80062145e-01 -2.37775110e-02 5.23499370e-01 4.81755674e-01 2.51390696e-01 -1.31770933e+00 3.26717615e-01 1.24314392e+00 -2.80867040e-01 -5.27767897e-01 8.77298638e-02 -1.78411514e-01 8.57803047e-01 6.92311823e-01 -6.15849793e-01 -2.13596493e-01 -4.43231106e-01 -3.71391058e-01 3.41003597e-01 4.65350360e-01 -9.74041820e-01 6.25331998e-01 1.86033338e-01 5.95561504e-01 -8.22056234e-01 3.22057307e-01 -8.38374197e-01 2.20307991e-01 7.35595107e-01 -1.19583003e-01 -3.06599587e-02 2.04902202e-01 7.34994233e-01 -2.90821582e-01 -2.39695936e-01 8.68904114e-01 -3.22729826e-01 -9.57973719e-01 7.19147682e-01 -5.81808269e-01 8.73435587e-02 1.44794369e+00 -3.16072345e-01 1.70126945e-01 -1.76825732e-01 -1.00482833e+00 6.57280982e-01 4.94861990e-01 5.23790717e-01 7.30267525e-01 -1.02635837e+00 -4.01376188e-01 -2.22210530e-02 -4.09296006e-01 3.80597264e-01 3.76630813e-01 1.21549976e+00 -8.92403901e-01 2.46738032e-01 1.62692338e-01 -1.00104499e+00 -1.43255424e+00 6.66445792e-01 5.29170454e-01 -6.62680492e-02 -7.66883135e-01 1.03480971e+00 2.31216416e-01 1.34976983e-01 5.30362248e-01 -4.81290996e-01 -3.50320518e-01 3.26550845e-03 5.55918396e-01 1.33585498e-01 -3.34158480e-01 -6.24396622e-01 -4.54178184e-01 8.29277098e-01 -2.96467334e-01 2.00551406e-01 1.33011913e+00 -2.19374865e-01 -2.09520534e-01 2.37890244e-01 1.32080090e+00 -1.54563457e-01 -1.86444223e+00 -6.15621768e-02 -7.89751112e-02 -4.72002894e-01 5.95420189e-02 -4.32600439e-01 -1.70741725e+00 3.46953332e-01 7.50312984e-01 -1.97130945e-02 1.07777011e+00 3.00156046e-03 1.07104957e+00 -1.29881963e-01 1.94413532e-02 -1.12135708e+00 1.12267919e-01 2.58431494e-01 5.60043991e-01 -1.31638610e+00 1.66231528e-01 -8.18614542e-01 -5.79173505e-01 1.21663582e+00 5.26323438e-01 -4.18816417e-01 7.35754430e-01 5.40483177e-01 6.91166222e-02 -2.47757494e-01 -5.52999079e-01 -1.04380451e-01 3.45900327e-01 1.53365806e-01 3.62394780e-01 -3.43620777e-01 -2.11025834e-01 2.20487759e-01 4.41738039e-01 2.33121589e-01 5.35224438e-01 1.00181973e+00 -1.46784961e-01 -9.38031256e-01 -2.84104317e-01 2.70214170e-01 -9.48944330e-01 -1.96220446e-02 1.20115750e-01 6.63404346e-01 -6.06529368e-03 8.58612537e-01 1.16699204e-01 1.71074614e-01 -1.74771219e-01 -4.57817823e-01 4.09476578e-01 -4.54656154e-01 -7.52378106e-01 7.64202476e-01 -2.70816356e-01 -1.10372150e+00 -9.57860112e-01 -7.15248287e-01 -1.46261299e+00 1.37979880e-01 -3.89145374e-01 1.48076996e-01 4.71453398e-01 8.02644730e-01 3.22625369e-01 8.23581278e-01 4.16154891e-01 -1.22623110e+00 1.61917523e-01 -3.56589586e-01 -3.19090545e-01 3.06431830e-01 5.05192339e-01 -4.60949630e-01 -1.93066210e-01 5.25931060e-01]	[9.17383098602295, -0.08460091054439545]
b6f2fbe4-53c9-4d57-904d-4949617dd9ed	real-time-hyperspectral-imaging-in-hardware	2204.02084	null	https://arxiv.org/abs/2204.02084v1	https://arxiv.org/pdf/2204.02084v1.pdf	Real-time Hyperspectral Imaging in Hardware via Trained Metasurface Encoders	Hyperspectral imaging has attracted significant attention to identify spectral signatures for image classification and automated pattern recognition in computer vision. State-of-the-art implementations of snapshot hyperspectral imaging rely on bulky, non-integrated, and expensive optical elements, including lenses, spectrometers, and filters. These macroscopic components do not allow fast data processing for, e.g real-time and high-resolution videos. This work introduces Hyplex, a new integrated architecture addressing the limitations discussed above. Hyplex is a CMOS-compatible, fast hyperspectral camera that replaces bulk optics with nanoscale metasurfaces inversely designed through artificial intelligence. Hyplex does not require spectrometers but makes use of conventional monochrome cameras, opening up the possibility for real-time and high-resolution hyperspectral imaging at inexpensive costs. Hyplex exploits a model-driven optimization, which connects the physical metasurfaces layer with modern visual computing approaches based on end-to-end training. We design and implement a prototype version of Hyplex and compare its performance against the state-of-the-art for typical imaging tasks such as spectral reconstruction and semantic segmentation. In all benchmarks, Hyplex reports the smallest reconstruction error. We additionally present what is, to the best of our knowledge, the largest publicly available labeled hyperspectral dataset for semantic segmentation.	['Andrea Fratalocchi', 'Bernard Ghanem', 'Silvio Giancola', 'Fedor Getman', 'Qizhou Wang', 'Arturo Burguete-Lopez', 'Maksim Makarenko']	2022-04-05	null	http://openaccess.thecvf.com//content/CVPR2022/html/Makarenko_Real-Time_Hyperspectral_Imaging_in_Hardware_via_Trained_Metasurface_Encoders_CVPR_2022_paper.html	http://openaccess.thecvf.com//content/CVPR2022/papers/Makarenko_Real-Time_Hyperspectral_Imaging_in_Hardware_via_Trained_Metasurface_Encoders_CVPR_2022_paper.pdf	cvpr-2022-1	['spectral-reconstruction']	['computer-vision']	[ 1.05742371e+00 -3.14796776e-01 1.14999905e-01 -5.95888793e-02 -4.31116313e-01 -7.77803302e-01 1.54545859e-01 -9.95048732e-02 -4.27891254e-01 3.37799013e-01 -6.03843272e-01 -2.99781650e-01 -2.36550614e-01 -7.66537964e-01 -7.20028937e-01 -9.63601410e-01 2.71691710e-01 5.22555232e-01 2.73945600e-01 7.63600543e-02 2.80844420e-01 5.02225041e-01 -1.98678529e+00 4.79911208e-01 1.08287811e+00 1.51618028e+00 6.47135794e-01 4.62881237e-01 -5.62511235e-02 1.20813422e-01 -1.88028693e-01 1.48108914e-01 6.86799288e-01 -2.63641477e-01 -3.46986473e-01 3.65586996e-01 8.88675511e-01 -2.15915412e-01 -1.78418115e-01 1.46083760e+00 3.76924872e-01 7.51035986e-03 5.44343174e-01 -7.55705953e-01 -8.76088440e-01 2.05134809e-01 -6.56466126e-01 -9.33660343e-02 1.65913075e-01 5.27928650e-01 7.30670393e-01 -8.28812659e-01 5.86758852e-01 6.18434012e-01 6.55515134e-01 4.08166111e-01 -1.27043021e+00 -3.94766837e-01 -4.98786643e-02 7.19849244e-02 -1.17955673e+00 -4.83007014e-01 5.51772118e-01 -6.06281042e-01 1.09712553e+00 4.12489861e-01 9.88881946e-01 6.38232052e-01 -6.94729239e-02 1.64313927e-01 1.65603530e+00 -4.02859211e-01 5.96071422e-01 4.63110209e-02 3.26719731e-02 9.44895029e-01 5.14778435e-01 2.00751618e-01 -5.42816281e-01 1.62810218e-02 4.80070144e-01 2.70315737e-01 -8.10125828e-01 -1.08094282e-01 -1.14088476e+00 1.50342494e-01 5.83550692e-01 6.51490092e-02 -2.63073266e-01 -8.13744888e-02 -2.41439462e-01 1.26979291e-01 6.20763481e-01 9.39869165e-01 -2.25111157e-01 4.66065377e-01 -1.13229561e+00 -4.80717510e-01 5.65734982e-01 6.26866162e-01 1.03831482e+00 4.21185121e-02 1.34551600e-01 6.41000330e-01 3.32087010e-01 8.81122112e-01 2.78137565e-01 -9.63147342e-01 -1.66646793e-01 8.17101061e-01 2.11537421e-01 -3.87641698e-01 -5.22287667e-01 -4.72758055e-01 -4.98595148e-01 6.96466029e-01 6.25422299e-02 2.43124552e-02 -1.12652743e+00 9.22768891e-01 1.90457582e-01 3.64094317e-01 1.54666618e-01 1.27440393e+00 8.56603980e-01 8.93102348e-01 -5.20764053e-01 -3.02868485e-01 1.03954613e+00 -1.15056503e+00 -2.19779387e-01 -3.17966282e-01 1.37600437e-01 -6.49981856e-01 1.23967302e+00 7.72263348e-01 -9.01926816e-01 5.12939580e-02 -1.26435733e+00 -4.92755584e-02 -6.17718816e-01 4.01507229e-01 7.23771572e-01 8.79402518e-01 -7.98358977e-01 7.90371776e-01 -8.75869215e-01 -4.40365881e-01 5.90312421e-01 3.88383359e-01 -2.02852190e-01 -6.00976087e-02 -2.46525660e-01 3.62888843e-01 4.23649907e-01 -7.30336364e-03 -6.70014262e-01 -1.02191484e+00 -3.88055384e-01 -4.25374135e-02 4.85906392e-01 -7.79376030e-01 1.02651775e+00 -1.27344108e+00 -2.17524958e+00 1.14313948e+00 -1.27833828e-01 -3.46961141e-01 5.39642870e-02 -4.55812626e-02 -2.45761767e-01 6.07400417e-01 -3.39338183e-01 4.23957855e-01 8.54280531e-01 -1.33007312e+00 -3.11299235e-01 -6.38793945e-01 -1.57279387e-01 1.37247756e-01 -5.10643959e-01 -1.55036211e-01 -1.27781212e-01 1.05705969e-01 3.83946598e-01 -9.73620832e-01 -1.44712046e-01 5.13647676e-01 -6.24090970e-01 6.44447207e-01 1.02873802e+00 -2.41027310e-01 4.41808730e-01 -1.96887994e+00 9.25477743e-02 6.79225028e-02 2.47479990e-01 5.95728278e-01 -9.28174928e-02 2.99988478e-01 3.97549458e-02 -4.68004495e-02 -6.65027082e-01 -3.12919647e-01 -2.89683044e-01 -3.01693738e-01 -6.07636422e-02 7.55424201e-01 -1.75900638e-01 6.95212841e-01 -6.87130868e-01 1.46150440e-01 5.64711630e-01 3.41034800e-01 -3.42879951e-01 6.20823614e-02 -7.56522357e-01 6.58585131e-01 -1.85877651e-01 1.10316122e+00 8.50202978e-01 -6.05371892e-01 3.48144560e-03 -4.97102678e-01 -9.29969788e-01 1.69032663e-02 -8.53495240e-01 1.90584230e+00 -2.01674372e-01 6.65214479e-01 6.08174264e-01 -8.85387003e-01 6.83732629e-01 1.32042216e-02 5.46698809e-01 -7.79825211e-01 6.48268834e-02 6.49181306e-01 -5.74972391e-01 -5.86028218e-01 2.47189388e-01 -4.00461107e-01 7.19495118e-01 1.88523695e-01 -2.07333356e-01 -6.25519872e-01 1.36748269e-01 -4.06616211e-01 7.74229228e-01 6.44898862e-02 1.05111308e-01 -4.18669105e-01 2.55677074e-01 4.58362669e-01 1.47309572e-01 5.91798723e-01 3.56635660e-01 7.24708676e-01 -1.27616167e-01 -2.48270512e-01 -9.31376815e-01 -1.04565573e+00 -2.45575637e-01 6.74693346e-01 4.92455423e-01 -5.25838174e-02 -7.82098591e-01 1.20686844e-01 6.90173591e-03 2.85570800e-01 -1.60447285e-01 3.14886808e-01 2.53385574e-01 -1.07750618e+00 1.51402161e-01 -1.49141312e-01 6.78791463e-01 -6.59349203e-01 -9.24606860e-01 1.14165798e-01 3.27582061e-01 -1.42082465e+00 2.10533246e-01 7.19410554e-02 -9.37003314e-01 -1.25491595e+00 -4.00688231e-01 -4.97286290e-01 6.55985355e-01 8.59110475e-01 6.75677717e-01 -2.40083337e-01 -9.07689154e-01 4.15828824e-01 -1.27941117e-01 -5.22545457e-01 -4.63978527e-03 -1.22900404e-01 -1.09225467e-01 2.78318971e-01 -1.66069835e-01 -7.38416255e-01 -9.62990344e-01 1.60344422e-01 -9.94448781e-01 5.11100352e-01 6.78972721e-01 6.00408733e-01 9.30604517e-01 6.55962452e-02 -7.31616393e-02 -1.06160033e+00 -6.03312999e-02 -5.66679798e-02 -1.37808192e+00 3.21947902e-01 -7.76247859e-01 -3.12716961e-01 8.17666113e-01 -5.47509044e-02 -8.95386994e-01 2.76072115e-01 1.68124542e-01 -4.67240006e-01 -3.59825850e-01 6.55252099e-01 1.20520815e-01 -7.65124023e-01 9.41135049e-01 2.56465793e-01 1.01850666e-01 -4.32991922e-01 1.37145102e-01 8.92704487e-01 5.94626009e-01 -2.32439250e-01 7.26573527e-01 1.20381069e+00 2.71062315e-01 -1.56896210e+00 -1.25299168e+00 -9.21406209e-01 -2.60794193e-01 -2.29534686e-01 9.19864178e-01 -1.04382277e+00 -7.90265262e-01 7.64292836e-01 -8.80242407e-01 -7.32754111e-01 -2.50714839e-01 3.67151737e-01 -4.04167145e-01 2.50254452e-01 -3.91141623e-01 -8.28518271e-01 -6.18163347e-01 -1.11406493e+00 1.30540335e+00 4.61098939e-01 6.09340310e-01 -6.61769509e-01 -2.47741923e-01 8.18061233e-01 5.80088615e-01 4.07727212e-01 6.66819572e-01 3.02913755e-01 -1.22373962e+00 1.26302272e-01 -6.85215056e-01 3.53884190e-01 6.39487505e-02 2.77829558e-01 -1.45689142e+00 -2.74773866e-01 4.44075614e-02 -4.77674961e-01 1.29068100e+00 5.94748318e-01 1.52456331e+00 -2.31521321e-03 -4.47475463e-01 1.41353536e+00 1.99787045e+00 1.59629568e-01 6.21892989e-01 3.35891485e-01 8.65794718e-01 6.10386789e-01 2.69623876e-01 2.88401365e-01 -1.00314967e-01 4.95061666e-01 8.37317884e-01 -4.10985857e-01 3.38753266e-03 4.22577083e-01 2.92127430e-01 4.27403480e-01 -1.56749591e-01 -3.51974845e-01 -9.12039340e-01 1.07044116e-01 -1.61103892e+00 -9.87770617e-01 -4.01956052e-01 2.29004788e+00 5.22060037e-01 -4.57721710e-01 -4.12985414e-01 -1.48101345e-01 5.82520783e-01 1.43284634e-01 -7.88841188e-01 3.20618391e-01 -4.00868535e-01 5.27822137e-01 8.51139545e-01 5.44218123e-01 -1.17950070e+00 8.70577276e-01 5.42696381e+00 4.12498266e-01 -1.80510843e+00 2.19683811e-01 3.19386512e-01 -4.24217880e-01 -3.02279949e-01 6.46755844e-02 -4.28043723e-01 3.47797483e-01 7.03392804e-01 2.14094713e-01 1.05729699e+00 4.52717364e-01 2.42867008e-01 -4.35398608e-01 -7.85987139e-01 1.31047356e+00 9.00895800e-03 -1.82397950e+00 2.85511371e-02 1.32215261e-01 9.94128406e-01 6.65037632e-01 2.80409485e-01 -7.46335924e-01 -4.37363595e-01 -1.12895942e+00 7.86784708e-01 5.48822045e-01 1.07776368e+00 -2.14104161e-01 1.69989154e-01 2.52876192e-01 -1.01703823e+00 -3.94959807e-01 -5.52965760e-01 -2.08431445e-02 -1.24676913e-01 1.14465714e+00 -3.69116157e-01 5.03571987e-01 7.34623671e-01 8.49147379e-01 -3.33316207e-01 1.22953463e+00 9.24740359e-02 6.08365893e-01 -5.02846658e-01 9.54219177e-02 3.46355855e-01 -7.91118443e-01 6.54757380e-01 1.01142919e+00 6.55615628e-01 3.22792649e-01 4.82577205e-01 1.28923512e+00 -1.92233250e-02 -1.71960831e-01 -5.89584768e-01 -5.82871139e-01 1.51377723e-01 1.49275064e+00 -9.34500873e-01 -1.07328467e-01 -4.72215474e-01 9.70667899e-01 -4.34347130e-02 3.81011516e-01 -4.13956493e-01 -2.56403565e-01 6.75127089e-01 3.98458332e-01 8.82757679e-02 -4.37769175e-01 -4.38879281e-01 -1.38028407e+00 -3.86714027e-03 -5.93902111e-01 2.37820167e-02 -1.20513666e+00 -1.17017555e+00 2.92535037e-01 -4.89584625e-01 -9.28737581e-01 7.06973433e-01 -1.34218431e+00 -3.40854675e-01 7.86292493e-01 -2.12895703e+00 -9.96721029e-01 -9.73053992e-01 5.30079067e-01 2.07559749e-01 9.81541872e-02 9.66094494e-01 -2.19030189e-03 -6.77128613e-01 -3.03101629e-01 6.72086656e-01 -4.35769528e-01 5.62886834e-01 -1.14534187e+00 8.42063278e-02 7.64706492e-01 -5.04135601e-02 2.97753036e-01 4.17650014e-01 -4.35539961e-01 -2.03630161e+00 -1.28418362e+00 -2.06199005e-01 1.46929145e-01 7.18758881e-01 -2.18988061e-01 -7.40061164e-01 2.37627015e-01 1.43906966e-01 2.08752260e-01 9.25836980e-01 -5.25932133e-01 -3.72879088e-01 -3.09495300e-01 -1.14772666e+00 4.40069348e-01 1.33681667e+00 -7.20173776e-01 -3.33924405e-02 1.00280046e+00 4.37512577e-01 -4.36025023e-01 -4.65966105e-01 4.54061151e-01 5.39608598e-01 -1.45815456e+00 9.42371964e-01 1.84945807e-01 4.94736522e-01 -5.14064908e-01 -2.02870313e-02 -1.20665014e+00 -1.30990744e-01 -7.32504368e-01 2.99754500e-01 4.32784200e-01 6.52678192e-01 -1.08059418e+00 1.07835579e+00 1.37075290e-01 -7.46484399e-01 -5.77150583e-01 -6.91210985e-01 -9.45385993e-01 -4.99549717e-01 -7.99046606e-02 4.40986156e-01 9.80094910e-01 -2.38963872e-01 9.68643129e-02 2.31014878e-01 7.51740873e-01 1.13536894e+00 8.82810533e-01 1.85778171e-01 -1.58138609e+00 -4.95331496e-01 -6.36057258e-01 -1.95489764e-01 -7.90491164e-01 1.73836932e-01 -1.02858770e+00 2.68546138e-02 -1.69851029e+00 1.44318432e-01 -2.71161586e-01 8.20264220e-02 3.28635037e-01 3.67469251e-01 6.59321904e-01 -4.93944213e-02 3.17475885e-01 -4.68742698e-01 4.50097263e-01 1.28537202e+00 -5.23801923e-01 -3.74846220e-01 -4.03095692e-01 -4.90575880e-01 7.31594920e-01 8.88763785e-01 -3.58265162e-01 -3.94948840e-01 -6.56603873e-01 3.94569635e-01 -2.57148951e-01 7.20508397e-01 -1.42939889e+00 5.13171673e-01 -4.00926799e-01 2.65800089e-01 -5.81884906e-02 7.18763947e-01 -8.51791084e-01 3.19483876e-01 5.03851593e-01 1.92063183e-01 -7.59331703e-01 1.21314816e-01 4.67481136e-01 -3.22466977e-02 -3.04277480e-01 1.16332984e+00 -5.09868145e-01 -7.99758136e-01 4.01670337e-01 -3.44872206e-01 -3.76153976e-01 1.07668865e+00 -5.97257435e-01 -9.55139279e-01 2.65637338e-01 -3.40487838e-01 -2.50727832e-01 1.10952353e+00 -3.69944870e-01 6.47216916e-01 -4.39574242e-01 -3.51800084e-01 1.95738286e-01 3.03093225e-01 2.16840997e-01 3.09012383e-01 7.92112410e-01 -1.10146463e+00 2.81899095e-01 -1.63209453e-01 -7.74843395e-01 -1.20032060e+00 1.78602636e-01 8.32009494e-01 2.95392007e-01 -7.17748582e-01 8.90378296e-01 1.45137891e-01 -3.87531102e-01 -2.67469704e-01 -4.08829600e-01 2.80280769e-01 -1.85659871e-01 5.21668911e-01 4.00721014e-01 3.49953771e-01 -1.22427456e-01 -1.39759719e-01 9.03247058e-01 4.66248661e-01 -8.18067268e-02 1.54493129e+00 -2.68146936e-02 -6.71961725e-01 4.79236871e-01 7.50295579e-01 -1.77786857e-01 -1.42920327e+00 1.61520466e-02 -4.52552319e-01 -5.67383170e-01 5.63961744e-01 -9.59990442e-01 -1.09227252e+00 1.06262720e+00 7.23806620e-01 1.89055815e-01 1.36426127e+00 -2.49020562e-01 6.22210026e-01 7.61736691e-01 6.08371258e-01 -9.92997050e-01 1.59438774e-02 4.24777210e-01 5.54440856e-01 -1.28248799e+00 2.90690988e-01 -9.37391758e-01 -1.13365360e-01 1.18872452e+00 4.04984087e-01 2.26607203e-01 5.15759766e-01 2.92011470e-01 1.29268438e-01 -6.63109779e-01 -4.35594350e-01 -2.71032691e-01 3.05488110e-01 6.39434218e-01 1.78529307e-01 2.57096648e-01 3.05936188e-01 1.09433513e-02 5.37789166e-02 -7.98618346e-02 5.62974572e-01 8.42155874e-01 -8.32697093e-01 -6.06990278e-01 -3.49392593e-01 7.30145276e-01 -1.60036936e-01 -3.70139778e-01 -5.62926650e-01 -1.33001162e-02 1.80719644e-01 9.55523908e-01 -2.57988740e-02 -3.61596197e-02 1.57405496e-01 1.13109387e-02 6.89785600e-01 -9.25808072e-01 -4.60262895e-01 -6.86899051e-02 -1.99032366e-01 -6.23507142e-01 -7.94191837e-01 -4.14161503e-01 -1.22598052e+00 2.93101892e-02 -5.30021489e-01 -1.57417729e-01 1.11355233e+00 7.96400070e-01 6.01042807e-01 1.75155848e-01 3.89618486e-01 -1.38810468e+00 -2.98047692e-01 -5.02989411e-01 -8.82679939e-01 -1.02861010e-01 3.53164673e-01 -4.43902820e-01 -3.85142595e-01 9.85019654e-03]	[10.20314884185791, -2.434542417526245]
baf486a7-d016-4692-92ad-dd4c39cbd2eb	end-to-end-video-instance-segmentation-via-1	2203.03145	null	https://arxiv.org/abs/2203.03145v1	https://arxiv.org/pdf/2203.03145v1.pdf	End-to-end video instance segmentation via spatial-temporal graph neural networks	Video instance segmentation is a challenging task that extends image instance segmentation to the video domain. Existing methods either rely only on single-frame information for the detection and segmentation subproblems or handle tracking as a separate post-processing step, which limit their capability to fully leverage and share useful spatial-temporal information for all the subproblems. In this paper, we propose a novel graph-neural-network (GNN) based method to handle the aforementioned limitation. Specifically, graph nodes representing instance features are used for detection and segmentation while graph edges representing instance relations are used for tracking. Both inter and intra-frame information is effectively propagated and shared via graph updates and all the subproblems (i.e. detection, segmentation and tracking) are jointly optimized in an unified framework. The performance of our method shows great improvement on the YoutubeVIS validation dataset compared to existing methods and achieves 35.2% AP with a ResNet-50 backbone, operating at 22 FPS. Code is available at http://github.com/lucaswithai/visgraph.git .	['Weiyao Lin', 'Kean Chen', 'Ning Xu', 'Tao Wang']	2022-03-07	end-to-end-video-instance-segmentation-via	http://openaccess.thecvf.com//content/ICCV2021/html/Wang_End-to-End_Video_Instance_Segmentation_via_Spatial-Temporal_Graph_Neural_Networks_ICCV_2021_paper.html	http://openaccess.thecvf.com//content/ICCV2021/papers/Wang_End-to-End_Video_Instance_Segmentation_via_Spatial-Temporal_Graph_Neural_Networks_ICCV_2021_paper.pdf	iccv-2021-1	['video-instance-segmentation']	['computer-vision']	[-8.55025277e-03 -3.12457010e-02 -6.64464355e-01 -1.48277298e-01 -5.78841448e-01 -5.92927754e-01 1.05821416e-01 -8.15142225e-03 -4.17887598e-01 4.87011552e-01 -1.99245855e-01 -2.86749572e-01 1.88529655e-01 -6.51090622e-01 -6.31271720e-01 -4.03884470e-01 -2.08618030e-01 2.60305852e-01 9.25816536e-01 1.36357903e-01 -3.46579170e-03 3.94333363e-01 -1.31856668e+00 4.85085743e-03 5.59886575e-01 1.14628351e+00 2.76226014e-01 9.24046934e-01 -1.28919140e-01 1.17351103e+00 -3.67965907e-01 -1.87986463e-01 5.72646201e-01 -2.00486466e-01 -8.28274608e-01 4.58882123e-01 6.93182170e-01 -7.08083808e-01 -8.02958906e-01 1.10022521e+00 4.92703825e-01 2.51630127e-01 5.38579449e-02 -1.45597994e+00 -2.37842336e-01 5.88270485e-01 -9.31118190e-01 5.18266380e-01 1.13096133e-01 2.20549211e-01 9.53612030e-01 -5.57788253e-01 8.54040325e-01 9.58465993e-01 5.32920539e-01 5.17945349e-01 -9.52310920e-01 -7.82175958e-01 5.16854167e-01 3.58499646e-01 -1.28412747e+00 -4.64478225e-01 6.54848158e-01 -3.62518370e-01 7.36369848e-01 -2.94801290e-03 9.25580740e-01 5.33117950e-01 -5.17491363e-02 1.02511728e+00 3.98163676e-01 3.68417874e-02 -4.11433354e-03 -2.69563705e-01 2.55063742e-01 8.62076521e-01 1.55141920e-01 -4.87418398e-02 -4.59811300e-01 4.22303110e-01 1.14449537e+00 1.55590534e-01 -3.31810445e-01 -3.68818432e-01 -1.05977511e+00 5.59794009e-01 6.51421249e-01 2.01600622e-02 -4.06882107e-01 6.16072655e-01 5.86831093e-01 1.38751090e-01 4.98166531e-01 -8.23026896e-02 -5.38000882e-01 -2.66035318e-01 -1.31675291e+00 2.59728706e-03 8.49455535e-01 1.36329675e+00 7.77242243e-01 2.42633611e-01 -3.25733572e-01 5.42652965e-01 1.38436735e-01 3.44272494e-01 1.88041762e-01 -1.35318136e+00 4.04601723e-01 6.37931883e-01 -1.87273771e-01 -1.17807400e+00 -4.85747039e-01 -3.12106699e-01 -5.80141246e-01 -8.05264562e-02 5.97663581e-01 -4.02451247e-01 -1.15221584e+00 1.48600638e+00 6.62144244e-01 8.50666165e-01 -3.07360470e-01 1.05012465e+00 1.22448969e+00 6.14451110e-01 2.01851279e-01 -1.70664668e-01 1.30766320e+00 -1.56039023e+00 -5.71631908e-01 -1.91147730e-01 6.67612195e-01 -6.79394066e-01 6.36959016e-01 1.72589973e-01 -1.42333913e+00 -5.94035089e-01 -7.39409208e-01 -1.58261389e-01 -1.29002348e-01 1.20828420e-01 4.80004936e-01 3.54953289e-01 -1.35570586e+00 4.78961617e-01 -1.09175682e+00 -3.05881530e-01 7.00787246e-01 6.80629790e-01 -1.77965537e-01 -6.63979491e-03 -8.28968167e-01 1.29628658e-01 6.09847069e-01 9.26267654e-02 -8.45962584e-01 -6.58426285e-01 -8.01822484e-01 1.35565832e-01 9.07475889e-01 -5.06682813e-01 1.29127765e+00 -8.72023046e-01 -1.31805277e+00 6.29752219e-01 -1.33344263e-01 -6.10043466e-01 5.72878540e-01 -2.02939972e-01 -1.10294081e-01 4.17015195e-01 2.22132698e-01 8.96938264e-01 7.10786402e-01 -1.08360255e+00 -1.08162713e+00 -1.28950119e-01 3.77188832e-01 2.84256786e-01 -2.68871129e-01 -1.04374010e-02 -1.62855709e+00 -5.99702299e-01 3.16437408e-02 -9.91190612e-01 -1.94702119e-01 9.94754061e-02 -6.06927514e-01 -1.66631922e-01 1.19882846e+00 -8.24898124e-01 1.55196667e+00 -1.94153845e+00 1.67124197e-01 2.08716821e-02 6.99052095e-01 5.36113381e-01 -1.50869608e-01 1.21761560e-01 1.82705849e-01 5.11394069e-02 5.40117472e-02 -5.25204659e-01 -1.94682688e-01 2.26760864e-01 2.34614283e-01 4.99489963e-01 -6.68016821e-02 1.02417958e+00 -9.77712393e-01 -8.12496305e-01 4.93087202e-01 5.33614695e-01 -5.93242764e-01 6.78435862e-02 -3.70839596e-01 5.54890752e-01 -5.36949515e-01 8.17337990e-01 6.87947154e-01 -6.67009532e-01 3.02008152e-01 -3.64194751e-01 -2.46830862e-02 -1.08317941e-01 -1.23697758e+00 1.70822239e+00 -3.84434089e-02 7.97075331e-01 3.01228791e-01 -1.04966390e+00 3.65606964e-01 1.88285917e-01 1.01822460e+00 -7.31414497e-01 3.04625124e-01 -4.61875945e-02 -1.74922615e-01 -3.86868447e-01 5.78998864e-01 6.27472341e-01 2.62295991e-01 2.94327170e-01 7.78938383e-02 3.47893745e-01 7.00788021e-01 4.93151248e-01 1.25291014e+00 4.29175854e-01 4.47870716e-02 4.01970893e-02 5.49193978e-01 1.50170505e-01 9.32023168e-01 6.99973702e-01 -6.28950179e-01 5.53999007e-01 5.81146300e-01 -3.76224369e-01 -8.12985122e-01 -7.86775649e-01 3.23679745e-01 1.12954938e+00 7.20811486e-01 -6.44713402e-01 -8.14989567e-01 -8.38094175e-01 -9.57056955e-02 2.87079304e-01 -3.77132595e-01 1.34772584e-01 -8.90713215e-01 -2.48529553e-01 2.83561736e-01 7.11926818e-01 7.71561623e-01 -9.62153077e-01 -6.29615128e-01 4.25586581e-01 -2.27435291e-01 -1.68222630e+00 -7.31476903e-01 -2.16544032e-01 -8.92430663e-01 -1.41970265e+00 -6.75195277e-01 -7.41632044e-01 5.31071723e-01 6.43296838e-01 1.20771277e+00 5.49243569e-01 -3.43310505e-01 5.40761769e-01 -3.77402216e-01 6.97604194e-02 -1.82183217e-02 2.44596183e-01 -4.14069355e-01 -6.43754229e-02 1.55093372e-01 -3.62059802e-01 -8.33389282e-01 3.68898064e-01 -7.56852031e-01 2.01755702e-01 2.46137559e-01 4.89852220e-01 8.49984705e-01 1.13512650e-02 2.45851040e-01 -9.61159229e-01 1.18361413e-01 -4.88932222e-01 -9.22246099e-01 1.22074813e-01 -3.92244577e-01 -5.14948845e-01 3.51922423e-01 -3.19308639e-01 -6.61982954e-01 2.56390691e-01 9.40609053e-02 -8.50903988e-01 -4.77232896e-02 3.27449352e-01 6.48937840e-03 -3.06844860e-01 1.38929650e-01 2.49778271e-01 3.02709248e-02 -1.86173171e-01 3.36454332e-01 1.88006088e-01 4.23044175e-01 -3.71289849e-01 7.57897735e-01 5.46441972e-01 4.06559296e-02 -7.58631408e-01 -7.88827658e-01 -9.07639742e-01 -6.33173168e-01 -6.24526858e-01 1.05919790e+00 -1.06677806e+00 -7.90033460e-01 4.83650267e-01 -9.94209766e-01 -8.47627819e-01 -2.82182366e-01 3.95154297e-01 -5.18327713e-01 6.43515885e-01 -8.92126143e-01 -5.33263862e-01 -3.41061801e-01 -1.43906450e+00 1.13943660e+00 4.63767588e-01 2.73151934e-01 -1.06818318e+00 -3.61585140e-01 3.67310703e-01 3.46065491e-01 2.95496523e-01 1.32481918e-01 -4.97134358e-01 -1.24201107e+00 -1.36045232e-01 -7.75407732e-01 2.30444625e-01 7.82349929e-02 1.86073020e-01 -5.29813409e-01 -4.17611003e-01 -4.19495374e-01 -9.78104994e-02 9.22647178e-01 8.10067296e-01 1.32572317e+00 -8.84604603e-02 -4.78799999e-01 9.95390296e-01 1.48258483e+00 1.77801624e-01 4.77444291e-01 2.91339099e-01 1.24846756e+00 2.20979601e-01 5.99636257e-01 4.54859108e-01 6.93864167e-01 7.64343143e-01 6.01696610e-01 -2.45435640e-01 -5.78982294e-01 1.27397388e-01 2.23358646e-01 6.73043191e-01 -2.08988458e-01 -6.18461311e-01 -9.36063588e-01 4.72090334e-01 -2.27184319e+00 -9.32739735e-01 -2.65189260e-01 1.87049091e+00 4.60864842e-01 4.40540053e-02 4.13407534e-01 -1.87956959e-01 1.06572366e+00 3.56236935e-01 -7.83618450e-01 3.10926527e-01 1.86745778e-01 -1.06762320e-01 8.80703449e-01 4.27880526e-01 -1.42608023e+00 1.38827264e+00 5.14189768e+00 8.11355472e-01 -1.05783749e+00 1.24425024e-01 6.34985626e-01 -2.67928064e-01 3.71559829e-01 -1.52346399e-02 -8.29295993e-01 4.98549819e-01 6.91662848e-01 -1.56538580e-02 4.74357873e-01 6.69140458e-01 3.81562114e-01 -1.39431223e-01 -8.20318401e-01 1.09501994e+00 -1.84405789e-01 -1.52156496e+00 -6.96025044e-02 8.08173791e-02 6.08321726e-01 4.03320253e-01 -5.25592417e-02 2.32977599e-01 3.00241649e-01 -5.14240205e-01 7.63899565e-01 2.76800275e-01 7.19526827e-01 -6.45756781e-01 4.93882060e-01 4.78191525e-02 -1.77512491e+00 2.87607349e-02 -1.00358292e-01 1.55716509e-01 5.57480156e-01 3.23343039e-01 -5.73512316e-01 5.87906539e-01 7.62441933e-01 1.22877800e+00 -4.84772652e-01 1.22443330e+00 -1.73555002e-01 7.44292378e-01 -4.40296501e-01 3.75161678e-01 3.68771672e-01 -1.24817476e-01 6.86291635e-01 1.19912696e+00 7.79911652e-02 -4.39176522e-03 8.55191648e-01 5.90829253e-01 -4.21023190e-01 -1.18732145e-02 -1.72324896e-01 -2.50286912e-03 4.71324742e-01 1.41455615e+00 -1.33487284e+00 -5.91104627e-01 -7.24590123e-01 9.43859696e-01 2.21083373e-01 6.11950517e-01 -1.32277858e+00 -2.33225435e-01 5.43316126e-01 1.91771120e-01 6.61616683e-01 -4.60714489e-01 4.80146669e-02 -1.38875592e+00 -7.44427815e-02 -5.72378039e-01 6.13166034e-01 -4.76149857e-01 -8.00355256e-01 4.15042430e-01 -8.38282257e-02 -1.22392452e+00 -1.04281399e-02 -3.58345658e-01 -5.35803914e-01 2.90140927e-01 -1.81097662e+00 -1.15569639e+00 -7.16804266e-01 8.85275483e-01 7.85212457e-01 6.81379288e-02 1.56014264e-01 7.02988088e-01 -1.01241398e+00 4.94312674e-01 -3.23251903e-01 6.23322189e-01 4.57380027e-01 -1.20296514e+00 4.49792296e-01 1.20930207e+00 9.38226134e-02 1.57271206e-01 2.31686547e-01 -8.47518921e-01 -1.47512758e+00 -1.46906638e+00 2.93537349e-01 -8.56986269e-02 7.20299006e-01 -1.79280490e-01 -7.38794744e-01 8.27335894e-01 2.02712104e-01 4.69904780e-01 3.02402347e-01 -2.86867291e-01 -6.37883767e-02 4.80865166e-02 -8.70280266e-01 6.26381576e-01 1.31334627e+00 -1.71822757e-01 2.26927236e-01 5.21010458e-01 8.96254003e-01 -8.84185553e-01 -9.41947937e-01 1.94682091e-01 3.04824889e-01 -9.36957240e-01 1.04370129e+00 -3.86015862e-01 2.75972396e-01 -6.34750426e-01 -2.14289706e-02 -6.19816303e-01 -2.62452751e-01 -8.27527821e-01 -6.21322632e-01 1.08514905e+00 2.13190898e-01 -4.21111912e-01 1.14615369e+00 5.66182077e-01 -1.32648855e-01 -7.43335426e-01 -7.45526016e-01 -6.90106452e-01 -5.63632846e-01 -4.95029479e-01 3.81941319e-01 7.65073597e-01 -4.73369986e-01 2.56569117e-01 -3.77008766e-01 2.69616425e-01 8.44426990e-01 8.21553841e-02 1.05455971e+00 -8.90976369e-01 -3.16234171e-01 -4.72412795e-01 -6.38518810e-01 -1.52233350e+00 9.00051892e-02 -9.19869006e-01 -4.05654944e-02 -1.78695595e+00 9.25314426e-02 -3.60299081e-01 -2.54406959e-01 5.43375075e-01 -1.28448382e-01 4.34342861e-01 6.59677207e-01 3.53930891e-01 -1.29784656e+00 2.18704224e-01 1.45285225e+00 -4.84203063e-02 -3.40222776e-01 -2.24019065e-02 -3.34233075e-01 7.54416466e-01 9.20264840e-01 -3.64047527e-01 -5.04578888e-01 -6.45525813e-01 -2.27893442e-01 3.50886226e-01 5.35938144e-01 -1.24623168e+00 6.30786598e-01 -2.58262008e-02 2.18758762e-01 -6.87696099e-01 3.08337480e-01 -8.92122924e-01 -1.80361699e-02 4.35375392e-01 -2.75046024e-02 1.80195481e-01 2.76003361e-01 8.49769235e-01 -2.86353976e-01 1.11510947e-01 6.57060087e-01 -2.41199255e-01 -1.26867414e+00 9.83230174e-01 -6.11142293e-02 3.29229802e-01 1.32772243e+00 -5.17441630e-01 -2.95160472e-01 -5.00918448e-01 -1.00160491e+00 6.45515203e-01 5.43179691e-01 2.88827926e-01 5.11609137e-01 -1.01463962e+00 -5.04531622e-01 -7.72656277e-02 -2.18547478e-01 2.32196033e-01 4.54306006e-01 1.18202615e+00 -7.78094590e-01 2.55424321e-01 -2.33748332e-02 -9.32601988e-01 -1.35699308e+00 4.57205236e-01 4.35455531e-01 -3.28094214e-01 -8.63356709e-01 9.72360313e-01 2.02920616e-01 -4.86571202e-03 4.94832873e-01 -2.73231387e-01 -2.35596627e-01 4.79630232e-02 3.04341376e-01 5.45572281e-01 -2.37913311e-01 -9.02000308e-01 -2.31348827e-01 6.63434386e-01 -1.61631003e-01 3.16487998e-01 1.27022171e+00 -4.00530815e-01 -6.24795370e-02 -4.68364060e-02 1.14759254e+00 -3.27261537e-01 -1.60431099e+00 -3.72826427e-01 -1.03984594e-01 -4.99748707e-01 2.97599554e-01 -3.31370652e-01 -1.91825306e+00 6.19943798e-01 4.23755914e-01 3.54251474e-01 1.26692152e+00 -5.98414093e-02 1.24484229e+00 3.06056011e-02 2.94515640e-01 -1.12362218e+00 6.49157986e-02 4.83452201e-01 2.81106979e-01 -1.19372857e+00 1.64608464e-01 -7.19709814e-01 -4.46252853e-01 1.17705905e+00 8.73586655e-01 -2.81040788e-01 6.58684015e-01 2.48320594e-01 -8.31559021e-03 -3.67988974e-01 -6.31072938e-01 -4.14381295e-01 2.86048651e-01 3.95893216e-01 3.83071840e-01 -8.99291262e-02 -7.77278692e-02 8.43765959e-03 3.71586829e-01 2.05466285e-01 5.26734054e-01 9.57163095e-01 -3.40600401e-01 -9.04231131e-01 7.95185715e-02 5.59405923e-01 -5.65965354e-01 3.47653106e-02 1.00610524e-01 1.03759456e+00 -1.88427011e-03 8.77507925e-01 1.39266461e-01 -3.89625818e-01 2.12124154e-01 -4.52684313e-01 2.48261720e-01 -6.50126040e-01 -6.15143239e-01 3.05845499e-01 1.41573235e-01 -1.25755715e+00 -8.02150190e-01 -5.29257059e-01 -1.54861176e+00 -4.87775296e-01 -4.43609387e-01 -1.86979100e-01 3.63891214e-01 6.42929912e-01 6.12451375e-01 8.81529510e-01 2.16324449e-01 -1.17916644e+00 5.88778742e-02 -4.27903831e-01 -4.18609947e-01 3.21698874e-01 2.63604373e-01 -6.25448108e-01 -1.21950381e-01 2.76684165e-01]	[9.153003692626953, -0.08170034736394882]
f191aeff-e3d9-47c3-8d99-53a6f01aa8a9	sparse-adversarial-video-attacks-with-spatial	2111.05468	null	https://arxiv.org/abs/2111.05468v1	https://arxiv.org/pdf/2111.05468v1.pdf	Sparse Adversarial Video Attacks with Spatial Transformations	In recent years, a significant amount of research efforts concentrated on adversarial attacks on images, while adversarial video attacks have seldom been explored. We propose an adversarial attack strategy on videos, called DeepSAVA. Our model includes both additive perturbation and spatial transformation by a unified optimisation framework, where the structural similarity index (SSIM) measure is adopted to measure the adversarial distance. We design an effective and novel optimisation scheme which alternatively utilizes Bayesian optimisation to identify the most influential frame in a video and Stochastic gradient descent (SGD) based optimisation to produce both additive and spatial-transformed perturbations. Doing so enables DeepSAVA to perform a very sparse attack on videos for maintaining human imperceptibility while still achieving state-of-the-art performance in terms of both attack success rate and adversarial transferability. Our intensive experiments on various types of deep neural networks and video datasets confirm the superiority of DeepSAVA.	['Qiang Ni', 'Leandro Soriano Marcolino', 'Wenjie Ruan', 'Ronghui Mu']	2021-11-10	null	null	null	null	['bayesian-optimisation']	['methodology']	[ 2.91254401e-01 -1.83647543e-01 3.19026589e-01 4.48371656e-02 -5.92893541e-01 -7.01931357e-01 7.03038275e-01 -3.59653026e-01 -4.20921981e-01 3.92902821e-01 2.01547563e-01 -3.84889424e-01 5.88604361e-02 -8.05728853e-01 -8.45030427e-01 -9.78697479e-01 -3.13730538e-01 -3.24215323e-01 4.77438718e-01 -4.26970035e-01 5.86066879e-02 6.30678296e-01 -9.02722418e-01 -7.83304945e-02 7.53968120e-01 9.38811421e-01 -1.69944942e-01 8.22049856e-01 4.37311262e-01 9.44384933e-01 -8.69315684e-01 -1.02089798e+00 7.20680177e-01 -3.58041942e-01 -6.12524390e-01 -3.41491215e-02 3.80103111e-01 -4.89001870e-01 -1.17371988e+00 1.35757554e+00 9.60948467e-01 1.33596465e-01 6.03832066e-01 -1.05130172e+00 -5.78434527e-01 5.19089282e-01 -5.91620743e-01 5.50060570e-01 3.51035416e-01 3.45097929e-01 5.64810157e-01 -4.18380737e-01 3.00125957e-01 1.36530888e+00 6.75702929e-01 5.91984510e-01 -1.01226056e+00 -7.98589230e-01 7.33360127e-02 4.58908498e-01 -1.46067321e+00 -3.78153026e-01 1.02606547e+00 -1.58861920e-01 4.38585043e-01 5.08989334e-01 3.50932002e-01 1.37147009e+00 5.45451641e-01 6.38706684e-01 8.80746901e-01 -2.38661155e-01 1.35964602e-01 -2.02236548e-01 -7.37495065e-01 6.06445312e-01 -6.85064569e-02 3.01319003e-01 -1.87599391e-01 -2.54869610e-01 9.14378941e-01 -3.47009413e-02 -4.87688869e-01 -3.33886921e-01 -1.03067064e+00 1.02393448e+00 5.23624897e-01 7.03544617e-02 -4.01006520e-01 2.02298567e-01 5.30714035e-01 4.66994882e-01 2.71766841e-01 3.31342667e-01 -1.14625387e-01 2.08582933e-04 -5.42883456e-01 3.88508856e-01 6.09784067e-01 5.85067272e-01 1.45405143e-01 4.68381405e-01 -2.95093298e-01 6.84566975e-01 3.26797038e-01 6.96677744e-01 4.54625249e-01 -9.47206259e-01 4.50280964e-01 1.49001807e-01 -2.61631906e-01 -1.69574964e+00 1.26998127e-01 -4.25523251e-01 -1.09778309e+00 4.75173324e-01 2.40327358e-01 -4.55428213e-01 -8.20191503e-01 1.73406124e+00 3.58688325e-01 6.82899296e-01 1.74293205e-01 9.00798082e-01 6.15158439e-01 6.92045450e-01 1.92172173e-02 -2.18891487e-01 1.01350033e+00 -7.48304963e-01 -5.57936609e-01 -9.70259011e-02 3.40927452e-01 -8.70263219e-01 5.59575140e-01 3.72649044e-01 -1.29752576e+00 -4.69374120e-01 -1.18113244e+00 4.87389177e-01 -9.48408991e-02 -6.80925786e-01 3.19436491e-01 9.29630756e-01 -7.48123825e-01 5.96352935e-01 -8.09767544e-01 1.43111601e-01 5.87255657e-01 5.07007062e-01 -4.96544868e-01 -9.39010084e-02 -1.49942541e+00 7.39016771e-01 3.16955924e-01 3.22752930e-02 -1.42438316e+00 -5.43236136e-01 -8.49939406e-01 -2.11879820e-01 4.08417821e-01 -5.98361969e-01 7.57373631e-01 -1.10001266e+00 -1.66609204e+00 7.44717836e-01 4.34909135e-01 -8.63020599e-01 7.25761831e-01 -1.75698072e-01 -7.24331379e-01 3.80498201e-01 -4.19998854e-01 4.08053905e-01 1.41833270e+00 -1.06143188e+00 -3.62379491e-01 -2.56528825e-01 3.68847758e-01 2.90449381e-01 -6.87809825e-01 3.94557327e-01 -4.81957376e-01 -1.43406236e+00 -2.93603152e-01 -1.08265162e+00 -3.05622101e-01 -9.19941515e-02 -3.70171458e-01 1.93476483e-01 1.17903924e+00 -8.15898776e-01 1.46499300e+00 -2.23726869e+00 5.12498856e-01 2.66831130e-01 3.05766016e-01 9.31360245e-01 -3.10845166e-01 4.32088107e-01 -3.19048136e-01 3.68264019e-02 -3.28599364e-01 1.07107610e-02 -1.07680984e-01 7.87229985e-02 -5.52530348e-01 6.94393277e-01 1.20844081e-01 8.45611811e-01 -8.65554452e-01 -3.27272713e-01 1.63521111e-01 6.99818254e-01 -7.68195271e-01 5.00318646e-01 1.48216233e-01 4.25704747e-01 -4.83761519e-01 4.91329730e-01 8.35634053e-01 2.22167686e-01 -3.82486694e-02 -1.74952209e-01 4.82654542e-01 -3.41802001e-01 -1.06076062e+00 1.41108358e+00 -1.78273976e-01 6.30199373e-01 1.31635563e-02 -9.62632060e-01 7.32239902e-01 3.33183289e-01 3.67999524e-01 -4.87020254e-01 4.46381837e-01 -2.22117435e-02 1.17089383e-01 -4.21571374e-01 1.34444252e-01 7.71978199e-02 -4.35372330e-02 5.08581325e-02 -7.02052936e-02 8.43988732e-02 -2.17411906e-01 3.72020304e-01 1.39278090e+00 -3.28810588e-02 1.24919780e-01 -1.65353760e-01 8.49366724e-01 -4.65627044e-01 3.52984935e-01 7.22854793e-01 -3.57395798e-01 5.16722977e-01 3.61490726e-01 -4.76765573e-01 -9.36091363e-01 -1.15733778e+00 1.09196693e-01 9.81534064e-01 3.82497877e-01 -3.07156682e-01 -1.02744603e+00 -1.06983054e+00 -6.89513683e-02 2.99595982e-01 -6.10381603e-01 -6.19989336e-01 -7.76173592e-01 -6.52262628e-01 9.07089710e-01 2.99286395e-01 8.03796411e-01 -9.87976730e-01 -2.23217264e-01 5.85427396e-02 -3.38663384e-02 -1.12229192e+00 -6.39125288e-01 -3.15280348e-01 -5.94277322e-01 -9.83482718e-01 -9.22334790e-01 -7.44064152e-01 6.18054152e-01 2.02210709e-01 7.80276656e-01 -5.60706444e-02 -1.90984279e-01 1.41900674e-01 -4.82379466e-01 -3.17022860e-01 -7.15923250e-01 -2.80417114e-01 2.23715827e-01 1.95109740e-01 4.34788316e-02 -6.68737411e-01 -7.41472065e-01 6.28830016e-01 -1.29555607e+00 -5.94072104e-01 5.71632564e-01 8.51999819e-01 1.67380512e-01 4.23789233e-01 2.09993958e-01 -5.34700811e-01 6.23205721e-01 -4.63000238e-01 -5.39387584e-01 -9.87327322e-02 -7.56309368e-03 -1.87464803e-01 8.73711228e-01 -7.47286558e-01 -7.11015761e-01 -2.75110602e-01 -4.34710950e-01 -9.14027691e-01 -6.43969104e-02 3.37181747e-01 -2.87780106e-01 -7.41191268e-01 8.08871031e-01 3.18515360e-01 1.63278028e-01 -8.25203508e-02 2.26332307e-01 5.17185986e-01 7.59940505e-01 -2.12296292e-01 1.40535903e+00 5.85626721e-01 1.11443229e-01 -6.74557924e-01 -5.19852698e-01 -1.81379113e-02 -3.05052549e-01 -2.86666781e-01 7.94998646e-01 -8.79329801e-01 -7.60993123e-01 9.63361621e-01 -8.80167425e-01 -9.49023664e-02 9.49269384e-02 3.70051056e-01 -4.15265769e-01 8.94857287e-01 -5.35217226e-01 -4.35950041e-01 -3.57173383e-01 -1.37634206e+00 6.26391768e-01 -2.33511608e-02 2.11397603e-01 -1.09648073e+00 4.45805825e-02 3.37280601e-01 3.99915665e-01 7.68927217e-01 4.67585176e-01 -6.52444363e-01 -6.32255316e-01 -6.02268577e-01 -1.93440113e-02 7.22458243e-01 1.27066687e-01 -1.37317702e-01 -7.00565159e-01 -7.80943811e-01 2.28966042e-01 -2.21476033e-01 7.50815094e-01 2.21754223e-01 1.30221486e+00 -6.68850899e-01 -1.05800137e-01 1.04739630e+00 1.30517411e+00 3.61250669e-01 1.06512022e+00 4.80067700e-01 9.08608615e-01 1.79514304e-01 4.01991725e-01 4.28218096e-01 -6.28744736e-02 8.75595748e-01 9.74478424e-01 4.46916148e-02 2.64465474e-02 1.36480238e-02 7.07553804e-01 7.37010479e-01 -8.94976258e-02 -5.68242252e-01 -7.05123723e-01 4.23011214e-01 -1.65678680e+00 -1.18733215e+00 4.31400001e-01 2.14031410e+00 6.85263753e-01 3.26500386e-01 1.75646037e-01 3.07762980e-01 7.41881609e-01 5.81806123e-01 -3.81508619e-01 -4.10006136e-01 -1.28095508e-01 3.77795883e-02 7.54136682e-01 3.25385213e-01 -1.60237920e+00 9.49448705e-01 6.74084759e+00 1.31338477e+00 -1.06709552e+00 -1.55506246e-02 5.84073544e-01 -3.03004449e-03 -2.63874214e-02 -3.99050266e-01 -4.12387311e-01 6.45508170e-01 8.61094654e-01 -9.99073908e-02 5.81632614e-01 6.23171151e-01 -3.60778570e-02 5.02199411e-01 -4.83060598e-01 9.92541730e-01 3.05063307e-01 -1.27462566e+00 2.93905199e-01 2.28697494e-01 7.41667867e-01 -2.27300256e-01 4.03940588e-01 6.39139041e-02 4.91724283e-01 -1.05915618e+00 4.73539203e-01 2.97276169e-01 7.76619554e-01 -1.20281124e+00 8.06920767e-01 9.37810317e-02 -9.43138838e-01 -2.28691384e-01 -3.27316046e-01 2.23996684e-01 1.69772923e-01 2.63502519e-03 -4.56120700e-01 5.73392272e-01 6.87411368e-01 5.45135677e-01 -4.05983210e-01 9.85909760e-01 -2.29217738e-01 7.09919453e-01 -1.29007310e-01 2.51889765e-01 5.55112422e-01 -4.04633433e-02 1.06124496e+00 1.04808640e+00 1.09111950e-01 8.01022574e-02 9.61006954e-02 8.67998153e-02 -1.70454457e-01 -8.22338983e-02 -7.07684100e-01 2.29163706e-01 4.05283660e-01 9.43557382e-01 -3.88366342e-01 5.49196079e-02 -1.94466129e-01 1.25868917e+00 -9.49898362e-02 2.40038514e-01 -1.27011013e+00 -6.35598898e-01 9.34286356e-01 -1.13611221e-01 6.56161845e-01 -1.77104175e-01 2.39850238e-01 -1.07668352e+00 -5.31093590e-02 -1.42730939e+00 4.07072634e-01 -3.62588257e-01 -1.17469168e+00 8.11929107e-01 -3.41150314e-02 -1.41283619e+00 -2.31415331e-01 -4.84427214e-01 -7.51728833e-01 7.05319643e-01 -1.14398670e+00 -9.50345039e-01 -2.00163230e-01 1.01096773e+00 4.43862915e-01 -7.54514873e-01 6.44707382e-01 3.43341053e-01 -7.41212428e-01 1.10191464e+00 1.54479071e-01 4.45755333e-01 6.12666130e-01 -9.25317347e-01 6.65263951e-01 1.26951337e+00 -2.66634896e-02 3.47082108e-01 9.60591078e-01 -5.09371579e-01 -1.42855871e+00 -1.26956201e+00 3.77569832e-02 -3.28779250e-01 8.67516398e-01 -1.52488083e-01 -7.55630493e-01 4.57368553e-01 4.03874546e-01 1.34048909e-01 4.32897836e-01 -6.28276885e-01 -5.15065193e-01 -1.64929032e-01 -1.26917243e+00 1.08166206e+00 1.03542662e+00 -5.53103387e-01 -5.09307802e-01 3.22462052e-01 8.77768219e-01 -5.39874971e-01 -1.02302456e+00 5.74073434e-01 4.03885126e-01 -1.00180197e+00 1.41420770e+00 -5.57502687e-01 3.96600783e-01 -3.54096234e-01 -2.23122954e-01 -1.33047271e+00 -5.14969707e-01 -1.24909461e+00 -4.93113607e-01 8.63239586e-01 -1.24642625e-01 -6.10261261e-01 7.27822602e-01 6.08264804e-02 3.89716402e-02 -8.09702039e-01 -9.28055882e-01 -9.19710279e-01 9.14609730e-02 -3.05636734e-01 4.80187297e-01 9.32938516e-01 -4.43625093e-01 -3.27971786e-01 -8.70650113e-01 5.07266104e-01 7.45185196e-01 -7.20062971e-01 8.90256464e-01 -5.82091749e-01 -3.71350080e-01 -3.94005507e-01 -1.24009347e+00 -8.73710573e-01 2.14181945e-01 -4.97570813e-01 -2.07769543e-01 -7.70486057e-01 -7.41772130e-02 3.91819403e-02 -4.38352287e-01 1.74848661e-01 -3.68297994e-01 7.89459884e-01 2.00476512e-01 9.55270529e-02 -4.65376586e-01 5.50212502e-01 1.13503087e+00 -2.11617932e-01 7.58202150e-02 -7.77844992e-03 -6.88864291e-01 8.37327063e-01 6.42167270e-01 -4.71251667e-01 -4.00660783e-01 -5.60240149e-01 -1.12255499e-01 -1.04434088e-01 4.58401352e-01 -1.21242726e+00 5.11837713e-02 -9.02652144e-02 2.91196823e-01 -1.23101838e-01 3.16911697e-01 -8.98130000e-01 1.41810194e-01 5.95199943e-01 -2.20633641e-01 1.86587512e-01 3.38827550e-01 8.67721736e-01 -3.36259723e-01 8.01472068e-02 1.10476506e+00 1.27908170e-01 -6.76819026e-01 6.94433212e-01 -1.79774970e-01 -7.98001885e-02 1.45028293e+00 -2.22978175e-01 -1.58171251e-01 -5.46628416e-01 -6.25606477e-01 6.51242910e-04 5.18106103e-01 5.71571290e-01 9.11836684e-01 -1.49956846e+00 -9.99908924e-01 2.22471282e-01 -2.86891729e-01 -3.87136310e-01 4.94963855e-01 5.64170003e-01 -8.86870027e-01 2.58679520e-02 -3.02136034e-01 -3.67805392e-01 -1.57792735e+00 8.63425374e-01 3.10656369e-01 -2.70138681e-01 -5.08163273e-01 1.19904172e+00 3.81052256e-01 8.58717188e-02 3.55127960e-01 5.42254269e-01 -2.45799363e-01 -3.03045958e-01 7.33896077e-01 3.82648498e-01 -1.68893874e-01 -1.00407040e+00 -3.36401999e-01 4.77310002e-01 -2.06111997e-01 1.51743025e-01 1.04694963e+00 -7.60526434e-02 7.49446675e-02 -3.64708155e-01 1.44520712e+00 2.60888338e-01 -1.54425061e+00 -3.24309140e-01 -5.60009778e-01 -9.13899243e-01 2.19138712e-01 -4.55746710e-01 -1.37174428e+00 5.91290653e-01 7.30798542e-01 4.59039003e-01 1.43218040e+00 -4.35806930e-01 1.12424612e+00 2.22013950e-01 1.73689976e-01 -5.65988302e-01 2.97111213e-01 4.42882001e-01 9.78660524e-01 -9.84999955e-01 -7.38829933e-03 -3.64169717e-01 -7.36296296e-01 7.45778501e-01 3.25049877e-01 -6.01099432e-01 7.73771822e-01 1.61719933e-01 1.80354431e-01 1.34933636e-01 -3.94149452e-01 1.66528910e-01 4.36046720e-01 6.98092878e-01 -8.62089768e-02 -2.53228396e-01 -6.98590726e-02 7.15834200e-02 -6.16842099e-02 -4.51618224e-01 3.28063816e-01 8.83379042e-01 -1.34863123e-01 -1.01905572e+00 -5.27793407e-01 8.67737830e-02 -1.09795403e+00 -9.54278484e-02 -1.33603111e-01 5.45249760e-01 -5.10787740e-02 8.91926408e-01 -1.92275062e-01 -8.50847006e-01 2.22552598e-01 -5.55262268e-01 4.44404513e-01 -1.19570807e-01 -7.03919351e-01 4.65163961e-02 -1.98902845e-01 -6.89003885e-01 -3.87642115e-01 -4.95349765e-01 -4.89196092e-01 -7.48722255e-01 -3.24393094e-01 2.93576829e-02 4.00108516e-01 8.47194135e-01 3.61315995e-01 5.40216327e-01 1.26557577e+00 -9.32448745e-01 -7.30867147e-01 -6.57106221e-01 -4.20929134e-01 6.38808727e-01 3.77767265e-01 -5.22438407e-01 -4.77530688e-01 1.02647021e-02]	[5.4308271408081055, 7.943722248077393]
073d1313-339a-4248-b8b3-b55b46281d7f	disturbing-target-values-for-neural-network	2110.05003	null	https://arxiv.org/abs/2110.05003v1	https://arxiv.org/pdf/2110.05003v1.pdf	Disturbing Target Values for Neural Network Regularization	Diverse regularization techniques have been developed such as L2 regularization, Dropout, DisturbLabel (DL) to prevent overfitting. DL, a newcomer on the scene, regularizes the loss layer by flipping a small share of the target labels at random and training the neural network on this distorted data so as to not learn the training data. It is observed that high confidence labels during training cause the overfitting problem and DL selects disturb labels at random regardless of the confidence of labels. To solve this shortcoming of DL, we propose Directional DisturbLabel (DDL) a novel regularization technique that makes use of the class probabilities to infer the confident labels and using these labels to regularize the model. This active regularization makes use of the model behavior during training to regularize it in a more directed manner. To address regression problems, we also propose DisturbValue (DV), and DisturbError (DE). DE uses only predefined confident labels to disturb target values. DV injects noise into a portion of target values at random similar to DL. In this paper, 6 and 8 datasets are used to validate the robustness of our methods in classification and regression tasks respectively. Finally, we demonstrate that our methods are either comparable to or outperform DisturbLabel, L2 regularization, and Dropout. Also, we achieve the best performance in more than half the datasets by combining our methods with either L2 regularization or Dropout.	['Mofassir ul Islam Arif', 'Klavdia Zavalich', 'Paweena Tarepakdee', 'Hanna Lukashonak', 'Yongho Kim']	2021-10-11	null	null	null	null	['l2-regularization']	['methodology']	[ 1.13501102e-01 3.61413419e-01 -3.80506426e-01 -6.92269385e-01 -7.93851376e-01 -4.25028503e-01 3.07671517e-01 3.59701850e-02 -6.97567523e-01 1.05395985e+00 1.14055984e-01 1.46139842e-02 5.61466534e-03 -5.47850490e-01 -1.07462394e+00 -9.09586132e-01 1.87745079e-01 1.39884517e-01 3.28126550e-01 2.85438269e-01 2.16301352e-01 4.89158660e-01 -1.43737447e+00 4.12778646e-01 1.04388177e+00 7.23532736e-01 1.23724146e-02 2.66709507e-01 -7.67133257e-04 1.24463058e+00 -7.07346499e-01 -7.83338100e-02 2.76499599e-01 -2.86514699e-01 -4.40811396e-01 2.10125938e-01 5.90039313e-01 -2.11770609e-01 -1.04996793e-01 1.03500545e+00 4.11129415e-01 2.65686780e-01 6.26141310e-01 -9.58466351e-01 -6.20476782e-01 5.36504865e-01 -6.78641379e-01 1.00789666e-02 5.10087721e-02 1.35887429e-01 7.74023473e-01 -9.43711221e-01 5.52594423e-01 1.25318491e+00 7.89030969e-01 6.88225150e-01 -1.52264380e+00 -7.73541689e-01 6.50613308e-01 -1.48843303e-01 -1.25461519e+00 -4.72503245e-01 8.95650744e-01 -4.28887397e-01 5.55753052e-01 1.51509300e-01 1.18591689e-01 1.52036655e+00 1.33058801e-02 9.00491059e-01 1.27227938e+00 -5.22262394e-01 4.04415607e-01 7.22420335e-01 6.45611644e-01 6.51304722e-01 3.08741659e-01 2.44504601e-01 -4.25581068e-01 -1.24343976e-01 5.71801245e-01 -2.01112464e-01 -4.63028103e-01 -2.52561849e-02 -6.30189359e-01 9.64256883e-01 4.49754804e-01 2.37775818e-02 -9.27606374e-02 2.08885923e-01 2.42450237e-01 4.03541416e-01 6.78863764e-01 4.90428329e-01 -5.21025360e-01 2.40604192e-01 -6.41892493e-01 -2.03816667e-02 5.87281585e-01 7.40183055e-01 1.01985025e+00 1.31991118e-01 -7.56771266e-01 1.01434171e+00 4.72048491e-01 1.23070747e-01 2.68130600e-01 -9.72844958e-01 4.75941539e-01 6.84588790e-01 2.22683996e-02 -7.75204480e-01 -2.74856955e-01 -6.88082278e-01 -6.17368996e-01 6.77428782e-01 3.04428995e-01 -3.19895983e-01 -1.26567972e+00 2.03341603e+00 8.78893882e-02 3.98558706e-01 -1.44696981e-01 8.20640743e-01 7.99166203e-01 4.94002163e-01 2.09255293e-01 -2.05647856e-01 5.85248649e-01 -9.45603430e-01 -6.86598837e-01 -3.17650348e-01 8.65500867e-01 -4.80076164e-01 1.42933321e+00 5.58257401e-01 -7.97575712e-01 -6.26864254e-01 -1.04895127e+00 1.15828283e-01 -2.08645359e-01 4.71882313e-01 4.66060907e-01 6.32727742e-01 -7.94185698e-01 7.57080078e-01 -8.45817864e-01 3.61379683e-02 7.11928248e-01 4.50854152e-01 -2.26454198e-01 -1.82490602e-01 -9.27034497e-01 7.57061005e-01 6.70933649e-02 1.15958899e-01 -1.11743557e+00 -7.38211811e-01 -9.00438249e-01 -3.00886065e-01 2.89205760e-01 -1.67899728e-01 7.01916218e-01 -1.27786314e+00 -1.47527397e+00 9.67898011e-01 -1.10686600e-01 -3.26018810e-01 7.00062573e-01 -5.36360681e-01 -7.05784336e-02 -3.56092691e-01 1.24483695e-02 6.55693650e-01 8.25431824e-01 -1.68392909e+00 -3.70251238e-01 -2.10662633e-01 4.29212525e-02 1.32418588e-01 -2.88369387e-01 -2.82528400e-01 -4.10589516e-01 -4.39688206e-01 -1.28951278e-02 -8.16662788e-01 -2.52613634e-01 2.64601260e-02 -8.26996028e-01 -1.13968648e-01 8.64932299e-01 -3.00639272e-01 1.23180676e+00 -2.36204767e+00 6.67028278e-02 1.91236570e-01 2.40464494e-01 4.78556484e-01 -3.94751847e-01 -1.40420780e-01 -2.65012886e-02 3.14492553e-01 -3.96085024e-01 -7.57602751e-01 -3.18168491e-01 6.25309289e-01 -2.10157931e-01 6.64779902e-01 3.41820747e-01 4.78575677e-01 -7.12310314e-01 -1.78767011e-01 1.51091933e-01 6.55709863e-01 -7.93499649e-01 4.59430963e-01 -2.28700936e-01 7.30147243e-01 -5.17592788e-01 3.12831640e-01 9.50547695e-01 3.49841150e-03 -8.68874341e-02 -1.70624942e-01 -1.63370967e-01 2.45502159e-01 -1.26820064e+00 1.22659075e+00 -3.77338409e-01 5.65658748e-01 9.18007195e-02 -1.12174165e+00 1.06953025e+00 1.48855090e-01 2.34763860e-03 -4.95063484e-01 9.01444480e-02 1.61898479e-01 -2.58244157e-01 -7.37734973e-01 -3.23205143e-02 4.75474186e-02 1.89350486e-01 8.40771273e-02 -5.71323857e-02 1.61084414e-01 3.81108299e-02 -3.55426222e-02 1.17611921e+00 2.47264951e-01 -2.14585781e-01 -1.17999561e-01 4.84854698e-01 -2.48937711e-01 1.05129158e+00 1.15562141e+00 -1.41659915e-01 8.88659358e-01 7.61447012e-01 -3.18743467e-01 -7.83472657e-01 -9.58324850e-01 -3.44152331e-01 9.58420634e-01 7.55073829e-03 2.58735754e-03 -5.69322228e-01 -1.28309953e+00 1.00359015e-01 1.09253812e+00 -9.03725803e-01 -5.38955986e-01 -5.76269805e-01 -1.05325282e+00 5.24568498e-01 3.66558909e-01 6.25972986e-01 -1.09517014e+00 -6.28814250e-02 -8.81740998e-04 1.82525367e-01 -7.32766807e-01 -8.30188319e-02 8.28252316e-01 -7.82724857e-01 -1.14880979e+00 -2.68923581e-01 -8.10758114e-01 1.12311625e+00 -1.03105545e-01 1.07628834e+00 2.00962558e-01 -9.15080532e-02 5.73810935e-02 -3.67990017e-01 -1.17798388e-01 -3.63865793e-01 4.89046425e-02 -1.04011640e-01 1.41113728e-01 2.51296729e-01 -5.62048793e-01 -2.56888807e-01 2.50889957e-01 -8.91163468e-01 -3.29303294e-01 4.14020598e-01 9.03287530e-01 6.65883720e-01 -5.90257198e-02 6.17993474e-01 -1.62832820e+00 6.24556065e-01 -6.36445820e-01 -6.29780054e-01 1.36370480e-01 -7.10506320e-01 2.39664868e-01 6.78283215e-01 -8.35485637e-01 -1.02460063e+00 -3.14701945e-02 -1.07446007e-01 -6.56098843e-01 -2.83516914e-01 4.18671578e-01 -1.31176487e-01 -2.56300997e-02 7.57548392e-01 -1.73290163e-01 -3.05427790e-01 -7.88288593e-01 1.28219366e-01 5.87421536e-01 5.40590025e-02 -4.97899234e-01 6.05037987e-01 3.35833549e-01 -2.13442966e-01 -3.73622060e-01 -1.51017427e+00 -9.28982273e-02 -4.48249519e-01 -8.17010403e-02 7.61988223e-01 -9.34968114e-01 -5.07474303e-01 4.93623555e-01 -1.05326617e+00 -6.12022817e-01 -6.25448108e-01 5.87496400e-01 -1.02122284e-01 -1.03455894e-01 -4.79860783e-01 -8.60123038e-01 2.17111915e-01 -1.09839404e+00 6.65757537e-01 3.19494247e-01 2.89315507e-02 -1.08663797e+00 1.36980191e-01 1.78582683e-01 3.15858185e-01 5.38473368e-01 8.13897371e-01 -8.38513553e-01 -4.82862175e-01 -2.56608605e-01 -1.11975476e-01 9.69894111e-01 1.81194231e-01 -3.28678750e-02 -1.26831913e+00 -4.36447293e-01 2.12845325e-01 -8.70183289e-01 1.21687222e+00 4.04035807e-01 1.31553805e+00 -2.51861840e-01 -2.05857247e-01 8.18219423e-01 1.41374004e+00 4.63276841e-02 6.80125177e-01 4.04774100e-01 8.08669567e-01 5.04613101e-01 3.56337130e-01 4.14642334e-01 6.41705142e-03 4.02737290e-01 5.69487691e-01 -3.89475673e-01 -2.80709565e-01 -3.37804347e-01 5.07005572e-01 6.02178991e-01 1.87357232e-01 -3.47960562e-01 -4.66450661e-01 3.03050727e-01 -1.78320527e+00 -5.89399636e-01 -4.13991719e-01 2.54579735e+00 1.12825012e+00 5.32334507e-01 -4.75396365e-01 5.84019907e-02 6.80805981e-01 1.52551144e-01 -6.61635756e-01 -3.64762694e-01 -4.03543919e-01 2.33345199e-02 6.72984898e-01 9.63346362e-01 -1.09100664e+00 9.72841084e-01 6.35520983e+00 8.34585667e-01 -1.19629848e+00 1.49760723e-01 1.03622806e+00 -1.70013070e-01 -3.49929929e-01 1.71543926e-01 -1.17570257e+00 5.63744724e-01 4.93497074e-01 4.46988046e-01 4.30272877e-01 7.62971580e-01 5.64448714e-01 -3.06798011e-01 -1.20617616e+00 5.89350164e-01 1.47899017e-01 -1.00754726e+00 2.63375267e-02 -3.01389903e-01 8.82140756e-01 7.20634311e-02 1.42624959e-01 5.32589555e-01 5.56026757e-01 -1.18654108e+00 6.09844923e-01 7.16535509e-01 5.98616838e-01 -6.55233860e-01 8.01963568e-01 5.24611056e-01 -6.10660732e-01 -1.11557536e-01 -5.93345821e-01 -1.88652232e-01 -6.47407696e-02 1.08696234e+00 -6.54391766e-01 1.49406463e-01 5.99026084e-01 8.36112916e-01 -7.58801639e-01 1.01891100e+00 -8.06620359e-01 1.12224567e+00 -4.08564657e-01 3.59942049e-01 2.58094013e-01 -2.54056931e-01 5.08772016e-01 1.12711060e+00 -9.29992571e-02 -3.06986779e-01 3.64250183e-01 1.16355979e+00 -2.74705827e-01 -1.74873278e-01 -5.22367358e-01 5.12388945e-01 3.95577192e-01 9.53794122e-01 -3.27517450e-01 -1.58783630e-01 -3.38486165e-01 8.41405630e-01 8.19709659e-01 6.62683368e-01 -9.07402456e-01 -2.92044312e-01 3.33696842e-01 1.54662848e-01 8.12240317e-02 1.82410032e-01 -4.36612189e-01 -9.89070177e-01 1.57850578e-01 -6.70431912e-01 1.69649258e-01 -5.34432471e-01 -1.39283907e+00 6.19626224e-01 -1.17746376e-01 -1.05611897e+00 3.13357502e-01 -3.69517863e-01 -6.85438395e-01 8.35629761e-01 -1.84645045e+00 -6.70081437e-01 -3.70254248e-01 5.38709104e-01 4.51792091e-01 -1.05790615e-01 4.58521217e-01 4.89943981e-01 -1.02602446e+00 7.09025741e-01 2.60980904e-01 1.42005712e-01 9.44916070e-01 -1.24809456e+00 -3.03834498e-01 5.57320833e-01 -1.55831456e-01 3.41827780e-01 5.38035452e-01 -6.91119075e-01 -7.87392139e-01 -1.47031045e+00 7.69897759e-01 -3.78760070e-01 1.65728495e-01 -6.67242050e-01 -1.25320709e+00 9.31995213e-01 -1.55788928e-01 4.95528370e-01 4.91834342e-01 6.97392272e-03 -4.07825291e-01 -2.05557287e-01 -1.45758426e+00 4.11286384e-01 8.53075027e-01 -8.41760933e-02 -4.21027482e-01 5.10423422e-01 8.44678521e-01 -2.82588243e-01 -3.57876390e-01 5.43485701e-01 -1.88533834e-03 -1.11041081e+00 6.16677642e-01 -4.88881022e-01 1.61649004e-01 -3.68346065e-01 -9.33478922e-02 -1.29219365e+00 -3.12676787e-01 -3.90884966e-01 1.32750228e-01 1.60731471e+00 7.71188796e-01 -6.34771705e-01 1.01178539e+00 5.36836803e-01 -1.63474277e-01 -8.32386792e-01 -9.26736057e-01 -7.45006740e-01 1.41877845e-01 -3.14992547e-01 -5.95213026e-02 1.14500117e+00 -6.83759749e-01 2.40250170e-01 -5.97591996e-01 2.35185176e-01 7.75411546e-01 -5.08686841e-01 4.99789238e-01 -1.15739048e+00 -2.34352484e-01 2.44273734e-03 -1.42074861e-02 -1.07971668e+00 4.18323904e-01 -8.89175117e-01 1.28998831e-01 -1.25517964e+00 9.51198116e-02 -8.89979482e-01 -4.27050859e-01 6.65563583e-01 -2.92719066e-01 3.25545281e-01 -1.44909561e-01 3.17910135e-01 -5.36025167e-01 7.39215195e-01 1.17037690e+00 -1.58674955e-01 -5.43115318e-01 1.54469445e-01 -5.98483145e-01 9.92860615e-01 6.79496408e-01 -1.07310951e+00 -4.56891924e-01 -6.08711720e-01 8.83146375e-02 -4.33309972e-01 2.06601501e-01 -9.07734990e-01 8.61353129e-02 1.53282052e-02 4.99747783e-01 -1.72357500e-01 1.77753195e-01 -7.74151742e-01 -2.32483283e-01 3.25472862e-01 -7.34908223e-01 -4.64914829e-01 1.54409677e-01 4.87452537e-01 -1.60519853e-01 -5.27221382e-01 1.20239902e+00 -4.80429977e-02 -1.88281730e-01 -3.00760306e-02 -4.41188991e-01 2.84473985e-01 9.52829301e-01 -1.15034558e-01 -2.07740009e-01 -7.76857287e-02 -1.03010559e+00 4.47894812e-01 2.70002156e-01 1.39349699e-01 5.36492765e-01 -1.26254213e+00 -8.00078690e-01 3.22722733e-01 -1.18673399e-01 3.18009704e-01 -1.36313751e-01 8.76307726e-01 -2.14503586e-01 -1.70519665e-01 2.36907899e-01 -5.59410632e-01 -9.68894601e-01 4.48456436e-01 5.53170979e-01 -4.05064136e-01 -5.69096088e-01 1.17536891e+00 3.13553095e-01 -7.51451015e-01 8.00272584e-01 -2.47634262e-01 -3.73490363e-01 -2.01402660e-02 3.30781370e-01 2.95742184e-01 -1.21136397e-01 -5.33337444e-02 -2.93639809e-01 6.49103940e-01 -3.69218200e-01 2.46313512e-01 1.40890861e+00 1.74978953e-02 2.53614951e-02 7.55217850e-01 1.31176949e+00 2.28937209e-01 -1.86946261e+00 -1.58988982e-01 1.34932488e-01 -3.85470539e-01 2.26589546e-01 -1.04016364e+00 -1.28611898e+00 5.72717905e-01 5.98163724e-01 -5.25410697e-02 1.04528069e+00 -1.12791643e-01 1.49365216e-01 3.67266983e-01 1.17887734e-02 -1.12637758e+00 2.88282692e-01 5.00142634e-01 7.73377657e-01 -1.49027550e+00 -1.29375467e-02 -4.26203370e-01 -6.50802553e-01 7.54320443e-01 9.46374893e-01 -5.79368472e-01 6.14469171e-01 4.36243027e-01 2.51822114e-01 6.44914666e-03 -8.07991505e-01 1.22354096e-02 -1.06462827e-02 6.13275230e-01 4.81636405e-01 -4.36832517e-01 -2.41629049e-01 5.39267063e-01 4.80200768e-01 1.10371552e-01 5.58607638e-01 8.68357599e-01 -5.34433305e-01 -1.10707831e+00 -4.39111620e-01 3.95481020e-01 -4.83167350e-01 -6.97549060e-02 -3.22557867e-01 7.04207480e-01 5.38944781e-01 9.57083225e-01 -6.13573380e-02 -2.22460017e-01 5.14367223e-01 -6.83660284e-02 1.34194285e-01 -8.06318045e-01 -8.01042080e-01 -3.52538489e-02 1.87913924e-01 -5.73541880e-01 -2.70890325e-01 -2.70370334e-01 -1.24352455e+00 -1.81036964e-02 -5.44314682e-01 1.56400174e-01 3.53549838e-01 8.76676440e-01 2.19865844e-01 6.79707885e-01 9.88767564e-01 -5.99016845e-01 -6.93289042e-01 -1.03710723e+00 -6.29939139e-01 6.66037321e-01 5.04582942e-01 -8.48297417e-01 -1.13826406e+00 -5.24295904e-02]	[9.260207176208496, 3.817368268966675]
87aea591-f201-49d6-8124-1c84a6c49ab1	collaboratively-learning-preferences-from	1506.07947	null	http://arxiv.org/abs/1506.07947v1	http://arxiv.org/pdf/1506.07947v1.pdf	Collaboratively Learning Preferences from Ordinal Data	In applications such as recommendation systems and revenue management, it is important to predict preferences on items that have not been seen by a user or predict outcomes of comparisons among those that have never been compared. A popular discrete choice model of multinomial logit model captures the structure of the hidden preferences with a low-rank matrix. In order to predict the preferences, we want to learn the underlying model from noisy observations of the low-rank matrix, collected as revealed preferences in various forms of ordinal data. A natural approach to learn such a model is to solve a convex relaxation of nuclear norm minimization. We present the convex relaxation approach in two contexts of interest: collaborative ranking and bundled choice modeling. In both cases, we show that the convex relaxation is minimax optimal. We prove an upper bound on the resulting error with finite samples, and provide a matching information-theoretic lower bound.	['Sewoong Oh', 'Jiaming Xu', 'Kiran K. Thekumparampil']	2015-06-26	collaboratively-learning-preferences-from-1	http://papers.nips.cc/paper/5818-collaboratively-learning-preferences-from-ordinal-data	http://papers.nips.cc/paper/5818-collaboratively-learning-preferences-from-ordinal-data.pdf	neurips-2015-12	['collaborative-ranking']	['graphs']	[ 2.53059745e-01 6.03289455e-02 -7.10524678e-01 -7.53420532e-01 -9.15758789e-01 -6.72724664e-01 2.25188255e-01 3.04265499e-01 -6.71928644e-01 6.28854394e-01 7.52999544e-01 -2.81609088e-01 -9.22157407e-01 -6.39699399e-01 -8.32409501e-01 -5.30293226e-01 -3.59906614e-01 9.04176176e-01 -5.34397304e-01 -9.62290019e-02 3.68837684e-01 1.77715942e-01 -1.31257081e+00 6.30615473e-01 6.14197195e-01 1.17376816e+00 9.59635824e-02 3.79769623e-01 1.93936035e-01 5.31976759e-01 3.62302959e-01 -4.73289102e-01 6.44754469e-01 4.43551727e-02 -7.91787684e-01 2.98520654e-01 2.85036296e-01 -4.80157107e-01 -1.28727078e-01 1.16161466e+00 1.90166757e-01 4.53246087e-01 1.00401759e+00 -1.16629016e+00 -3.54570448e-01 9.58192885e-01 -3.62430692e-01 -3.06295633e-01 5.22930980e-01 -5.35047352e-01 1.68581617e+00 -9.31828797e-01 5.36560357e-01 1.25336826e+00 3.27528179e-01 5.99831417e-02 -1.95377159e+00 -2.40771681e-01 2.50328451e-01 7.05285296e-02 -1.02381814e+00 -3.29016894e-01 5.36080301e-01 -5.33555090e-01 1.16240725e-01 6.05430901e-01 2.53064036e-01 1.23283398e+00 -2.05673739e-01 9.00020123e-01 1.52830803e+00 -3.72543514e-01 5.05648434e-01 3.54038507e-01 3.86118174e-01 4.30752695e-01 3.53785843e-01 1.32766917e-01 -5.28376818e-01 -8.91479135e-01 5.07637501e-01 5.51866591e-01 -3.12400818e-01 -6.68399096e-01 -1.07792056e+00 1.19056988e+00 1.77426785e-01 -3.04108173e-01 -7.32890308e-01 -1.38252497e-01 -7.94720054e-02 6.83676243e-01 3.29518199e-01 2.92326927e-01 -5.73437989e-01 1.98053613e-01 -9.11555588e-01 2.86162972e-01 1.25004590e+00 8.19375515e-01 6.77272558e-01 -6.94346309e-01 -1.10107824e-01 8.09544802e-01 7.61633575e-01 4.52314138e-01 1.12887904e-01 -1.08995044e+00 6.61491275e-01 4.97747846e-02 6.86298788e-01 -9.74123359e-01 -1.80504128e-01 -2.24359199e-01 -7.24486172e-01 2.24385753e-01 8.35453868e-01 -1.00964025e-01 -3.89852136e-01 1.76770532e+00 1.95781723e-01 -2.53865808e-01 -2.20885590e-01 1.05287004e+00 -5.14824018e-02 5.77755153e-01 -2.23600879e-01 -5.13864815e-01 9.24092889e-01 -3.79946619e-01 -3.97076964e-01 1.18225940e-01 6.35935068e-01 -6.02299750e-01 9.89994347e-01 8.65045726e-01 -1.04187131e+00 -3.63017842e-02 -6.59707427e-01 -8.13054200e-03 1.26501963e-01 -1.58385374e-02 6.63051724e-01 6.12050295e-01 -7.18259215e-01 9.85306919e-01 -4.31286842e-01 -5.93981668e-02 1.18247211e-01 7.42638707e-01 -2.54128098e-01 -3.71282965e-01 -9.29432988e-01 4.59747046e-01 -1.04768246e-01 -7.94446617e-02 -6.57616019e-01 -5.26079357e-01 -3.77665550e-01 1.96606606e-01 7.46548772e-01 -5.67184865e-01 1.34052277e+00 -7.96691358e-01 -1.22480881e+00 4.89647955e-01 -5.61233461e-02 -1.92829967e-01 7.30048060e-01 -1.61715358e-01 -1.48695931e-01 -5.36513686e-01 6.39864504e-02 -1.27656087e-01 6.28191411e-01 -1.10773563e+00 -9.66178656e-01 -6.93942428e-01 4.04358566e-01 3.88333499e-01 -2.92454809e-01 -1.60181150e-01 -1.20071298e-03 -3.48503083e-01 2.60415554e-01 -1.08124781e+00 -6.39289737e-01 2.07989328e-02 -3.21854830e-01 -6.61260262e-02 -4.85183299e-02 -5.83597243e-01 9.57253993e-01 -2.00474167e+00 1.13477588e-01 8.76401901e-01 1.06142789e-01 -6.71758056e-01 -1.88555449e-01 6.24246120e-01 2.26315960e-01 2.43535116e-01 4.54649106e-02 -3.51431251e-01 6.26161814e-01 3.55978280e-01 -5.73009551e-01 8.00869703e-01 -6.32499814e-01 5.59431493e-01 -8.72342408e-01 -4.19134125e-02 -1.85395017e-01 -9.65311527e-02 -1.04239392e+00 2.87767261e-01 -2.67661452e-01 1.23818509e-01 -6.36892617e-01 3.37421954e-01 6.97351396e-01 -3.66519213e-01 8.45394313e-01 -1.86720386e-01 -8.20954889e-03 4.56684858e-01 -1.86585546e+00 1.45532954e+00 -5.13838410e-01 2.66582798e-02 3.84161234e-01 -1.02729619e+00 2.29196280e-01 3.18758577e-01 6.57332540e-01 -4.73175436e-01 9.00943726e-02 2.37579480e-01 -3.22274148e-01 -3.18080634e-01 5.37935674e-01 -3.09457690e-01 -3.48579645e-01 5.54776967e-01 -1.96926370e-01 5.51483929e-01 -6.24363311e-02 1.04254521e-01 8.29414845e-01 -2.92684913e-01 3.50368708e-01 -3.98416370e-01 6.42920285e-02 -2.16943994e-02 7.48943806e-01 1.18278039e+00 7.32250512e-01 4.03763294e-01 6.46315157e-01 -2.67056495e-01 -9.80630457e-01 -1.02842855e+00 -1.62627310e-01 1.31044638e+00 3.11388895e-02 -1.41522750e-01 -1.35294273e-01 -4.67453510e-01 3.71652126e-01 4.59351778e-01 -5.63415289e-01 3.79552215e-01 -2.99319364e-02 -8.51943016e-01 -2.64871746e-01 2.72738516e-01 -1.20408498e-01 -3.73703182e-01 6.96124360e-02 2.66790897e-01 -2.56013364e-01 -7.52210140e-01 -7.76547909e-01 3.73193502e-01 -1.04466605e+00 -1.07481647e+00 -5.39160490e-01 -5.18430471e-01 6.38842583e-01 1.91230908e-01 1.08621359e+00 -3.31397891e-01 3.57670486e-01 3.65624219e-01 -2.84746885e-01 8.73405263e-02 5.41330017e-02 -3.42535794e-01 6.65203869e-01 6.32212937e-01 1.82324767e-01 -5.53863108e-01 -5.38908303e-01 5.21498322e-01 -9.10070300e-01 -1.76707000e-01 6.99833095e-01 1.01988113e+00 8.37143183e-01 5.02364151e-02 8.35655853e-02 -1.43785477e+00 8.96709025e-01 -1.04247570e+00 -9.00952816e-01 3.27576965e-01 -7.77588308e-01 4.02454436e-01 7.12201118e-01 -6.67825103e-01 -7.96101391e-01 1.57038569e-01 2.05583006e-01 -6.87947497e-03 -2.99613941e-02 9.90064740e-01 -4.69413877e-01 1.45088047e-01 3.26106101e-01 -1.81659028e-01 -2.91887611e-01 -9.87405360e-01 5.73089600e-01 8.49226594e-01 2.69324929e-01 -8.82292330e-01 4.55393046e-01 4.99420255e-01 7.33638033e-02 -6.06608152e-01 -9.60832000e-01 -6.51579142e-01 -4.63508636e-01 2.52787501e-01 3.79053831e-01 -8.93408835e-01 -1.02008832e+00 -3.78540307e-01 -6.57022774e-01 -3.23301971e-01 -3.89176697e-01 9.75940406e-01 -8.89887750e-01 3.35309833e-01 -3.98873121e-01 -1.09223294e+00 3.09514850e-01 -8.81458521e-01 3.85240972e-01 -1.95181295e-01 -2.51265675e-01 -1.03848803e+00 2.48624012e-01 4.29189980e-01 6.72205612e-02 -6.72701374e-02 9.31246638e-01 -9.80196953e-01 -5.64254522e-01 -2.85894394e-01 3.11630107e-02 4.91734743e-02 -1.05569407e-01 -5.98907888e-01 -4.73462909e-01 -4.96972024e-01 9.42747891e-02 -3.94334018e-01 5.99781811e-01 5.68087935e-01 1.30739856e+00 -9.68754292e-01 -1.07754789e-01 5.58111429e-01 1.43669140e+00 -1.01268142e-01 2.09395736e-01 -1.04395464e-01 2.90265292e-01 7.80774593e-01 7.57521808e-01 8.62181067e-01 4.89123255e-01 7.39949822e-01 3.26618135e-01 8.32877457e-02 9.52901900e-01 -5.09596288e-01 2.83733577e-01 6.66819930e-01 -1.55307585e-02 -2.72065729e-01 -2.77251482e-01 3.93678516e-01 -2.23006725e+00 -1.04790902e+00 -1.70474142e-01 2.92557859e+00 6.01858556e-01 -2.84838490e-02 3.84925723e-01 1.80904232e-02 5.78530252e-01 -2.66060621e-01 -5.70448995e-01 -3.81743699e-01 -3.39533836e-02 4.68962975e-02 9.30810273e-01 8.48586261e-01 -9.00470138e-01 1.21265672e-01 6.83432150e+00 6.58189952e-01 -5.40292919e-01 1.03619784e-01 7.68007934e-01 -3.42962980e-01 -7.62838006e-01 2.47115463e-01 -6.36509061e-01 5.21315515e-01 7.96522558e-01 -2.41160199e-01 9.91497338e-01 7.38164961e-01 4.21741635e-01 -1.32458314e-01 -1.65725982e+00 1.00215006e+00 -3.38063926e-01 -1.06325984e+00 -3.05600286e-01 8.74267220e-01 9.83799160e-01 -7.48784915e-02 3.73805732e-01 1.67693332e-01 9.47369099e-01 -1.06131411e+00 5.66585839e-01 6.59849167e-01 7.86458075e-01 -7.11304545e-01 4.10743624e-01 7.09320247e-01 -6.48987949e-01 -6.05747581e-01 -6.94325626e-01 -4.18809712e-01 2.50169516e-01 8.62254083e-01 -4.12296951e-01 3.40012789e-01 3.12119186e-01 5.30514836e-01 2.53129601e-01 1.22119403e+00 1.92485601e-01 7.06417859e-01 -5.11903763e-01 -1.72647685e-02 1.02846310e-01 -7.34760523e-01 3.98347080e-01 8.27244997e-01 3.61478716e-01 4.38192666e-01 5.66156566e-01 7.38360584e-01 -4.88601416e-01 3.20647180e-01 -4.12899226e-01 -3.38877998e-02 2.75163442e-01 1.16576934e+00 -2.15831473e-01 -5.05973808e-02 -7.75040805e-01 8.86456490e-01 2.65742391e-01 3.57223779e-01 -3.94789785e-01 3.42254013e-01 9.18521166e-01 3.05462688e-01 2.00519845e-01 -5.10095805e-02 -2.04814821e-01 -1.36305285e+00 1.37429625e-01 -9.48211133e-01 5.57176948e-01 -2.71103829e-01 -1.77755809e+00 1.32107791e-02 9.10496563e-02 -1.24098229e+00 -2.08928555e-01 -6.97345197e-01 -2.67451435e-01 9.30101871e-01 -1.10404670e+00 -5.54669023e-01 3.47536176e-01 6.32359624e-01 2.15647161e-01 -5.86770326e-02 7.27177143e-01 2.78121591e-01 -3.65508981e-02 3.70800823e-01 9.75771964e-01 -2.53535986e-01 5.65537572e-01 -1.58485055e+00 -2.99029052e-01 2.92416632e-01 3.14945251e-01 5.97276270e-01 8.27757061e-01 -4.07067031e-01 -1.72330284e+00 -8.35630178e-01 9.26681876e-01 -4.13867831e-01 7.43190765e-01 -2.98701704e-01 -4.74427730e-01 6.63614988e-01 -4.11888152e-01 -6.22853041e-02 1.09185433e+00 6.88466132e-01 -2.74793625e-01 -8.43604878e-02 -1.19417083e+00 5.55766463e-01 8.64728689e-01 -5.90973437e-01 -3.71733576e-01 4.22922820e-01 2.14594454e-01 6.77575022e-02 -1.08365631e+00 4.82708029e-02 1.03303421e+00 -7.39013910e-01 1.04207289e+00 -1.22430372e+00 3.03997248e-01 5.33641875e-02 -1.04539645e+00 -1.44701886e+00 -7.43590474e-01 -5.69777727e-01 -2.26270631e-02 7.78520167e-01 5.94628930e-01 -4.67596054e-01 9.40787613e-01 1.23901260e+00 5.53210616e-01 -8.86719942e-01 -6.41944349e-01 -5.58063030e-01 -1.64308567e-02 -2.92059213e-01 3.17211479e-01 9.29718852e-01 2.43290961e-01 3.32757711e-01 -1.07999468e+00 2.94889688e-01 1.17475998e+00 4.55988348e-01 5.38706303e-01 -1.61345434e+00 -8.27396631e-01 2.32352922e-03 -2.52665617e-02 -1.70031118e+00 3.66803817e-02 -9.61839080e-01 -1.82569861e-01 -1.23806965e+00 7.63814092e-01 -7.59372354e-01 -5.38181961e-01 5.11011807e-03 2.25251049e-01 -3.54571082e-02 1.23084895e-01 2.74506688e-01 -7.86715388e-01 2.37279922e-01 8.37443292e-01 -8.66903216e-02 -2.18038738e-01 5.79097271e-01 -1.10193217e+00 7.31489778e-01 4.76937145e-01 -8.38141978e-01 -3.42170030e-01 -3.29332769e-01 7.57702172e-01 7.60094106e-01 1.80904627e-01 -2.50662476e-01 4.46174413e-01 -8.37616384e-01 3.70044082e-01 -5.98729074e-01 2.69083232e-01 -1.03015757e+00 4.44974303e-01 1.31077498e-01 -9.99532402e-01 1.77159179e-02 -8.04965615e-01 9.32882905e-01 8.92916620e-02 -4.10183221e-01 4.01208639e-01 -6.38842285e-02 -3.38754475e-01 6.35110319e-01 -4.30022299e-01 -2.41330713e-02 3.12804401e-01 9.49618891e-02 2.92235166e-01 -8.09564650e-01 -1.20213759e+00 4.50555712e-01 4.80095804e-01 1.78643346e-01 5.46278715e-01 -1.56039667e+00 -7.29964614e-01 -1.60958841e-01 -4.30334955e-02 -4.78540421e-01 7.42125674e-04 8.04824233e-01 2.96855509e-01 4.40223724e-01 -1.73234548e-02 -7.07002878e-02 -1.21891177e+00 5.70445061e-01 -1.25016659e-01 -5.75104773e-01 4.50510979e-02 4.65259016e-01 2.85038650e-01 -6.43363416e-01 2.79781193e-01 -1.54361919e-01 -1.91391721e-01 1.84672430e-01 4.76806462e-01 6.14049852e-01 -1.03091352e-01 -4.11253929e-01 -2.41746847e-02 5.50810695e-02 -4.56820987e-02 -5.60841441e-01 1.51327157e+00 -5.65807819e-01 -1.03686973e-01 6.96212351e-01 1.33278680e+00 2.91821957e-01 -1.14670527e+00 -7.07570195e-01 3.29419166e-01 -9.29704487e-01 4.60551642e-02 -6.83314979e-01 -5.98122060e-01 4.18448329e-01 4.61404055e-01 2.78581440e-01 8.46078157e-01 -1.68478340e-01 3.61973524e-01 7.73711383e-01 7.23675251e-01 -1.24602699e+00 -2.10347965e-01 3.96260053e-01 7.21683204e-01 -1.30935824e+00 -3.39099392e-02 -2.19587490e-01 -6.01794302e-01 8.30977201e-01 -2.88132459e-01 -2.19312772e-01 1.05852699e+00 -1.56627029e-01 -3.85640323e-01 1.02696866e-01 -9.82474208e-01 -1.12142883e-01 6.25028610e-01 2.85674095e-01 3.62133652e-01 5.78237057e-01 -5.53728759e-01 1.13409030e+00 -8.42114016e-02 -3.81116150e-03 5.57925105e-01 5.54151416e-01 -4.06732589e-01 -1.39165938e+00 -4.08577353e-01 1.03296220e+00 -7.04821229e-01 -1.26246974e-01 -1.28315106e-01 -5.36596477e-02 -2.17872769e-01 1.20842659e+00 -2.18093097e-01 -4.36228782e-01 2.96442509e-01 -1.33435667e-01 3.61790299e-01 -5.93800485e-01 -9.72641110e-02 2.75317073e-01 2.18508914e-01 -6.21549547e-01 -1.62061200e-01 -9.49403226e-01 -5.68127930e-01 -5.08385122e-01 -4.13552463e-01 5.16359687e-01 5.48852682e-01 8.93422127e-01 5.21375909e-02 -2.94424206e-01 1.21247768e+00 -8.33147407e-01 -1.54314256e+00 -5.76516032e-01 -1.30146372e+00 7.77321160e-01 3.69763821e-01 -4.78212684e-01 -3.35465074e-01 -1.71425432e-01]	[9.384124755859375, 5.450090408325195]
c44bd776-6604-450e-b514-5584eeca1f2f	leveraging-vision-language-models-for	2301.10166	null	https://arxiv.org/abs/2301.10166v1	https://arxiv.org/pdf/2301.10166v1.pdf	Leveraging Vision-Language Models for Granular Market Change Prediction	Predicting future direction of stock markets using the historical data has been a fundamental component in financial forecasting. This historical data contains the information of a stock in each specific time span, such as the opening, closing, lowest, and highest price. Leveraging this data, the future direction of the market is commonly predicted using various time-series models such as Long-Short Term Memory networks. This work proposes modeling and predicting market movements with a fundamentally new approach, namely by utilizing image and byte-based number representation of the stock data processed with the recently introduced Vision-Language models. We conduct a large set of experiments on the hourly stock data of the German share index and evaluate various architectures on stock price prediction using historical stock data. We conduct a comprehensive evaluation of the results with various metrics to accurately depict the actual performance of various approaches. Our evaluation results show that our novel approach based on representation of stock data as text (bytes) and image significantly outperforms strong deep learning-based baselines.	['Navid Rekabsaz', 'Christopher Wimmer']	2023-01-17	null	null	null	null	['stock-price-prediction']	['time-series']	[-5.77607810e-01 -9.43645298e-01 -3.07945490e-01 -3.72615635e-01 -5.11175573e-01 -9.28987622e-01 1.15618038e+00 1.38400704e-01 -5.86398602e-01 4.98234361e-01 5.54317594e-01 -5.28990388e-01 4.20411110e-01 -1.43352902e+00 -8.88723493e-01 -1.91307023e-01 -3.57714981e-01 3.22892278e-01 8.20435882e-02 -4.83726978e-01 8.71118963e-01 3.37059021e-01 -1.28973174e+00 3.53207201e-01 -2.52996206e-01 1.87827265e+00 2.96005160e-02 6.41707420e-01 -8.75051022e-01 1.27661479e+00 -6.82423890e-01 -5.80078602e-01 1.04533827e+00 2.50297785e-01 -2.81029314e-01 -7.54952580e-02 3.72431606e-01 -1.03990161e+00 -6.56933129e-01 1.03220809e+00 6.83956593e-02 -3.71196121e-01 3.01509112e-01 -1.04261684e+00 -1.07492816e+00 8.74667287e-01 -8.66887629e-01 9.39736426e-01 -7.67457485e-02 2.60663360e-01 1.30116653e+00 -9.12540138e-01 4.37023908e-01 9.39222753e-01 5.87021947e-01 -1.04658686e-01 -5.31860530e-01 -8.16752195e-01 3.46583903e-01 2.57185191e-01 -8.75765920e-01 -9.00642946e-03 8.30904663e-01 -5.48009992e-01 1.11323941e+00 -2.07848027e-02 9.80953097e-01 6.94725811e-01 5.37641108e-01 8.50676239e-01 1.09350061e+00 -1.66018695e-01 1.28920123e-01 -2.50021189e-01 4.06442583e-01 4.19461250e-01 2.45167568e-01 2.14157343e-01 -6.08364224e-01 -2.10404053e-01 8.77385437e-01 5.01464546e-01 1.51918828e-01 3.61802369e-01 -1.56655252e+00 9.71823514e-01 4.43909436e-01 3.29215646e-01 -6.55687869e-01 3.98509294e-01 5.82385540e-01 6.64177179e-01 8.16140831e-01 6.34037331e-02 -8.13517690e-01 -3.06298047e-01 -1.36164975e+00 6.39763355e-01 9.72930551e-01 5.68851650e-01 4.80229080e-01 5.39991260e-01 2.19218824e-02 1.46135241e-01 3.26676726e-01 5.81211567e-01 1.09167159e+00 -3.82717401e-01 8.09166551e-01 5.52373469e-01 4.20842558e-01 -1.05310321e+00 -4.46549356e-01 -6.73427284e-01 -6.01837337e-01 4.05460238e-01 3.64623249e-01 -3.69428933e-01 -8.79690528e-01 1.07559955e+00 -1.10980593e-01 5.87331533e-01 3.06421369e-01 6.25189900e-01 6.52784050e-01 1.32454169e+00 -2.38904044e-01 -6.37511909e-02 1.23924744e+00 -8.94767046e-01 -5.58495700e-01 3.62810157e-02 3.69886845e-01 -8.75239909e-01 4.62756991e-01 1.51943877e-01 -9.69447911e-01 -6.33838832e-01 -1.11652994e+00 1.50064319e-01 -9.15597141e-01 -1.60830945e-01 6.19004846e-01 4.20559585e-01 -1.07380760e+00 6.80988252e-01 -6.82761669e-01 4.69468474e-01 7.04066679e-02 -2.03966871e-02 1.82121813e-01 6.40095949e-01 -1.24908769e+00 7.12682784e-01 4.84467447e-01 3.40226293e-02 -5.22343576e-01 -8.03862274e-01 -6.75132513e-01 1.66155443e-01 -1.95281446e-01 -2.83685088e-01 1.53230274e+00 -9.01413560e-01 -1.20959878e+00 5.43044984e-01 3.84787053e-01 -1.48772573e+00 6.76939547e-01 -1.78249896e-01 -6.50979936e-01 -2.25137174e-01 -2.28549451e-01 5.30152261e-01 7.75406182e-01 -3.82701099e-01 -1.02857435e+00 -3.67542654e-01 1.56646818e-01 -4.15768772e-02 -5.06713390e-01 7.12473914e-02 -5.51491231e-02 -1.29719901e+00 -1.60331205e-01 -6.44944906e-01 -6.77471096e-03 -2.56164044e-01 -1.09295072e-02 -2.21564457e-01 8.66155922e-01 -8.75424147e-01 1.20637703e+00 -1.89625692e+00 -7.12482333e-01 1.09911874e-01 3.30145424e-03 -5.71129918e-02 -6.83557987e-02 5.38068593e-01 -2.63349831e-01 1.91320386e-02 1.83956340e-01 -3.46231759e-02 3.97819221e-01 -6.23093992e-02 -1.42757273e+00 3.43545675e-01 -1.80406630e-01 1.37513900e+00 -2.84143478e-01 -3.01835388e-02 9.01071578e-02 1.97327271e-01 -1.99126471e-02 -8.28195885e-02 -6.56677544e-01 -1.28071949e-01 -3.88324618e-01 5.92163146e-01 7.57720649e-01 -4.82053280e-01 -2.93011516e-01 5.23140701e-03 -7.27575004e-01 2.99349338e-01 -1.03038192e+00 1.26178455e+00 -1.49944097e-01 5.61637819e-01 -6.83197021e-01 -7.30797529e-01 1.08546448e+00 2.78561920e-01 5.67364514e-01 -1.18630552e+00 -1.38177693e-01 3.46405864e-01 -2.33636677e-01 1.63090259e-01 8.93345058e-01 -2.73262978e-01 -1.79901659e-01 8.15302014e-01 -4.43214804e-01 1.99271306e-01 3.74713123e-01 -8.35466608e-02 7.39406288e-01 -2.33919732e-02 1.65290296e-01 5.38597554e-02 3.67994457e-01 8.76855180e-02 2.37135708e-01 5.82859874e-01 -1.25574157e-01 3.76699805e-01 4.28652793e-01 -1.42449617e+00 -1.24841607e+00 -7.85171807e-01 -4.74221185e-02 8.59235823e-01 -3.39824796e-01 -3.30782570e-02 -3.13702434e-01 -3.55839580e-01 4.65970576e-01 6.76221430e-01 -6.03194416e-01 4.87305522e-01 -5.24294496e-01 -9.93517816e-01 2.26459816e-01 7.21613586e-01 7.72195578e-01 -1.16541171e+00 -7.88204551e-01 4.05434728e-01 4.18384820e-01 -1.16746247e+00 -4.85151201e-01 -1.48218021e-01 -8.88595104e-01 -7.42435157e-01 -1.35889506e+00 -6.44862473e-01 -1.54596508e-01 3.65297608e-02 1.25711703e+00 -5.77250198e-02 1.39197364e-01 3.21164399e-01 -2.00737774e-01 -8.23936701e-01 -9.12457425e-03 -1.19893573e-01 -2.27162287e-01 2.22178534e-01 4.34810072e-01 -3.23037684e-01 -8.47636223e-01 -3.60508472e-01 -9.75178123e-01 -1.20119579e-01 5.13012588e-01 4.02321517e-01 5.14357448e-01 8.83475021e-02 4.04936731e-01 -4.23316956e-01 6.81174874e-01 -5.81468701e-01 -1.38113368e+00 2.20460907e-01 -7.36839771e-01 1.16170697e-01 4.83307362e-01 -1.41125143e-01 -7.19818115e-01 -1.81555495e-01 2.72809733e-02 -3.37767392e-01 3.66215855e-01 1.04172552e+00 7.90462434e-01 2.54332572e-01 -2.02705339e-01 9.22370911e-01 -2.07492873e-01 -5.36160886e-01 1.81021139e-01 4.86081839e-01 5.66648543e-01 -1.23012569e-02 8.74725878e-01 6.40158892e-01 -2.06322953e-01 -5.04290581e-01 -8.97704899e-01 -4.05428082e-01 -6.01022542e-01 -7.60847181e-02 5.09293497e-01 -1.21875322e+00 -5.73736370e-01 1.09940565e+00 -1.10373461e+00 -4.88160849e-02 -1.16320945e-01 5.21485984e-01 -2.43270546e-01 2.49364138e-01 -1.06518400e+00 -8.41378212e-01 -8.21800768e-01 -7.57335007e-01 9.37028050e-01 -9.34519917e-02 1.89034238e-01 -1.15512598e+00 4.72844213e-01 3.16159129e-02 6.49703145e-01 3.36131305e-01 9.23145950e-01 -1.05818295e+00 -8.68137956e-01 -6.02386475e-01 -3.66581619e-01 2.89315790e-01 -1.71600953e-01 -1.25171319e-01 -6.28939927e-01 -1.73208550e-01 9.92451087e-02 -1.82284310e-01 1.29381955e+00 4.26844388e-01 7.87248254e-01 -4.22164559e-01 3.94002199e-01 6.49418175e-01 1.64826012e+00 3.80254894e-01 5.43946207e-01 1.01454377e+00 4.39474940e-01 1.05342597e-01 3.42167854e-01 1.00754476e+00 6.97558463e-01 7.73848593e-02 4.82193828e-01 1.93585619e-01 3.89183611e-01 -1.11086555e-01 6.62588656e-01 9.07986283e-01 5.01127951e-02 -3.62764634e-02 -8.63476217e-01 2.80190855e-01 -1.60397041e+00 -1.33338511e+00 2.76326537e-01 1.85316539e+00 5.22738159e-01 5.63429296e-01 3.15023154e-01 2.28623599e-02 5.90014398e-01 9.18810308e-01 -7.34419107e-01 -6.19837157e-02 -2.67248571e-01 5.31268455e-02 1.03971899e+00 1.47013348e-02 -1.46489370e+00 7.12843478e-01 6.97566223e+00 4.14813071e-01 -1.66560197e+00 -3.06007147e-01 1.14826882e+00 -1.81276411e-01 -3.49131435e-01 -3.98560733e-01 -1.10172975e+00 7.55879879e-01 1.33096159e+00 -7.07603514e-01 2.19346419e-01 8.08250129e-01 9.91496444e-02 3.21517169e-01 -7.92916298e-01 1.17447472e+00 -1.06341153e-01 -2.17045617e+00 6.16680980e-01 1.76764980e-01 6.68134153e-01 4.82827306e-01 6.77927315e-01 3.57954770e-01 1.20058522e-01 -5.71674526e-01 1.46164894e+00 8.70915115e-01 3.24182361e-01 -8.14934134e-01 7.56341875e-01 3.43289524e-01 -1.40965223e+00 -3.78633738e-01 -5.27186632e-01 -5.25116265e-01 9.37124863e-02 4.74748671e-01 -3.83751154e-01 2.24181250e-01 6.03688240e-01 1.01498461e+00 -4.48489189e-01 7.02924073e-01 5.34296036e-01 3.86290878e-01 -3.32082510e-01 -2.33689085e-01 8.24491560e-01 -2.93395340e-01 6.27949908e-02 8.98466170e-01 7.63088524e-01 4.25759517e-02 1.35318413e-01 7.44843543e-01 -3.42249006e-01 1.24832191e-01 -6.63043618e-01 -5.53373396e-01 1.06087960e-01 8.96122456e-01 -7.66297936e-01 -6.93597198e-01 -1.23509729e+00 5.66036046e-01 -1.53100602e-02 1.19834408e-01 -8.89618278e-01 7.96549842e-02 4.99935985e-01 -2.04450041e-02 8.27876627e-01 -5.98278165e-01 -3.22861731e-01 -1.30674446e+00 1.99513599e-01 -4.93961722e-01 6.22609198e-01 -6.65745616e-01 -1.30167997e+00 5.96679866e-01 -1.79171547e-01 -1.55231571e+00 -4.37905282e-01 -9.19350803e-01 -8.03871334e-01 7.75806785e-01 -2.15811324e+00 -8.91098499e-01 2.63367385e-01 4.32286978e-01 7.38030732e-01 -9.75501239e-01 4.75210100e-01 9.04926881e-02 -3.72005045e-01 2.68338695e-02 5.44115424e-01 7.59124458e-01 6.35494590e-02 -1.33015919e+00 1.43903387e+00 7.95191288e-01 7.47776210e-01 1.67323276e-01 3.18544030e-01 -8.16767216e-01 -1.30618966e+00 -1.17627966e+00 7.91283190e-01 -8.77422392e-02 1.50165069e+00 1.07384712e-01 -6.34692132e-01 1.00358677e+00 3.34399521e-01 4.49525714e-02 3.50817144e-01 -6.59553468e-01 -6.41496003e-01 -3.05480212e-01 -5.64374208e-01 2.50321120e-01 2.70973027e-01 -3.86589676e-01 -8.00926805e-01 2.99048930e-01 7.32651472e-01 -2.84535050e-01 -7.48447895e-01 1.73356105e-02 6.52664125e-01 -1.12515390e+00 1.06101429e+00 -3.36273283e-01 5.51953018e-01 1.02942511e-01 -1.87625274e-01 -9.64856684e-01 -2.32491806e-01 -3.21650714e-01 -2.60552019e-01 8.05241287e-01 5.72093666e-01 -7.58677959e-01 8.25933456e-01 6.29000187e-01 3.13954949e-01 -5.41887045e-01 -8.79503787e-01 -6.93151772e-01 3.29138756e-01 -4.89270061e-01 1.30876315e+00 8.87196541e-01 -3.51619899e-01 -2.13909075e-01 -4.68363434e-01 9.27755982e-02 3.57433319e-01 8.90344024e-01 4.25965190e-01 -1.06462455e+00 -4.63809781e-02 -7.07847357e-01 -6.25438750e-01 -1.05834305e+00 1.93382964e-01 -6.71834648e-01 -7.28463829e-01 -1.16449535e+00 4.78285924e-03 2.30864942e-01 -7.47624457e-01 4.85661440e-02 3.83661896e-01 6.48106486e-02 4.42343205e-01 6.65564716e-01 -4.28639144e-01 3.60099018e-01 1.03137743e+00 -5.05840659e-01 2.49752924e-01 2.91055087e-02 -1.42539740e-01 8.72121155e-01 7.94652283e-01 -1.46771927e-04 1.13423273e-01 -6.18852139e-01 6.02420986e-01 2.43130967e-01 2.97007114e-01 -9.57461596e-01 3.96315753e-01 -2.79121697e-02 8.12068582e-01 -1.30191779e+00 2.97352672e-01 -6.47229612e-01 -1.99519336e-01 8.33263695e-01 -4.49701130e-01 1.18645275e+00 1.48952618e-01 7.98869431e-01 -7.03342259e-01 4.39411141e-02 2.75882423e-01 -5.83289504e-01 -1.48453021e+00 5.65943420e-01 -1.25241101e-01 -2.32480988e-01 1.04726911e+00 -8.74197632e-02 -3.72319937e-01 -7.00372994e-01 -4.20085162e-01 1.28613800e-01 -5.61695248e-02 7.17336893e-01 6.82946265e-01 -1.29575527e+00 -9.37287748e-01 2.12993279e-01 -1.02266915e-01 -7.34273791e-01 -4.38366607e-02 2.76726961e-01 -1.07129657e+00 9.42390382e-01 -2.99185663e-01 -4.68872041e-02 -6.80548549e-01 8.06898654e-01 3.18477541e-01 -5.37286341e-01 -8.06894302e-01 6.32824600e-01 1.13555312e-01 1.72467470e-01 3.31288695e-01 -9.05688047e-01 -6.11055255e-01 4.39262778e-01 1.01532447e+00 1.76934019e-01 7.67004490e-02 -6.80591941e-01 4.76816893e-02 6.29607022e-01 -2.26788223e-01 -2.87310719e-01 1.75360107e+00 1.72574278e-02 -9.81493890e-02 8.82590890e-01 1.17099833e+00 -4.11879689e-01 -1.27769959e+00 -3.47651392e-01 6.28690362e-01 -2.98667967e-01 3.43325138e-01 -5.05000293e-01 -1.48848784e+00 6.65522635e-01 6.92705452e-01 5.17992020e-01 7.28863657e-01 -4.87750113e-01 1.62875295e+00 5.34685493e-01 4.65402156e-01 -1.04226255e+00 -1.84141621e-01 8.77995133e-01 9.06888962e-01 -1.42202818e+00 -1.26782209e-01 4.83450830e-01 -5.82636178e-01 1.60849476e+00 -2.31208038e-02 -4.25311476e-01 1.12957120e+00 4.48790580e-01 5.34020185e-01 -2.94132262e-01 -1.09363782e+00 -7.61911348e-02 2.96386898e-01 -2.52840549e-01 4.88485098e-01 -4.17710328e-03 8.14545620e-03 3.72673720e-01 -6.04086578e-01 3.87870371e-02 5.55545986e-01 9.62331831e-01 -5.03270626e-01 -8.24888885e-01 -5.64221263e-01 5.73991477e-01 -9.13273990e-01 -4.04981554e-01 -1.49217844e-01 7.64844120e-01 -4.71050441e-01 3.74084443e-01 9.41315532e-01 -1.58394203e-01 4.64577079e-02 2.79596835e-01 -4.36485652e-03 -2.78908819e-01 -6.68951333e-01 1.93722874e-01 -3.06855649e-01 -2.26731315e-01 -4.36477661e-01 -6.93423748e-01 -1.26914191e+00 -6.55792534e-01 3.64863396e-01 -1.53705090e-01 6.02869272e-01 9.85606849e-01 7.75043815e-02 1.32724449e-01 9.01277423e-01 -8.09726298e-01 -1.14564288e+00 -7.33531594e-01 -1.10058522e+00 3.35599184e-01 7.61352301e-01 -4.02335785e-02 -1.80049881e-01 2.60189474e-01]	[4.41765022277832, 4.257970809936523]
c70ed3d3-cb62-427c-898b-d9ceccde2aa0	multiple-human-pose-estimation-with	null	null	http://campar.in.tum.de/pub/belagiannis2014eccvChalearn/belagiannis2014eccvChalearn.pdf	http://campar.in.tum.de/pub/belagiannis2014eccvChalearn/belagiannis2014eccvChalearn.pdf	Multiple human pose estimation with temporally consistent 3d pictorial structures	Multiple human 3D pose estimation from multiple camera views is a challenging task in unconstrained environments. Each individual has to be matched across each view and then the body pose has to be estimated. Additionally, the body pose of every individual changes in a consistent manner over time. To address these challenges, we propose a temporally consistent 3D Pictorial Structures model (3DPS) for multiple human pose estimation from multiple camera views. Our model builds on the 3D Pictorial Structures to introduce the notion of temporal consistency between the inferred body poses. We derive this property by relying on multi-view human tracking. Identifying each individual before inference significantly reduces the size of the state space and positively influences the performance as well. To evaluate our method, we use two challenging multiple human datasets in unconstrained environments. We compare our method with the state-of-the-art approaches and achieve better results.	['Vasileios Belagiannis', 'Pascal Fua', 'Xinchao Wang', 'Slobodan Ilic', 'Nassir Navab', 'Bernt Schiele']	2014-09-06	null	null	null	null	['3d-multi-person-pose-estimation']	['computer-vision']	[-1.02311306e-01 -1.90407515e-01 -5.59023283e-02 -1.82775602e-01 -2.68100619e-01 -5.80187798e-01 4.84516740e-01 -1.75935686e-01 -3.46524209e-01 5.22710085e-01 1.28185317e-01 4.03282762e-01 2.01713517e-01 -3.35926265e-01 -7.18866050e-01 -2.69991785e-01 -4.07368727e-02 7.60364771e-01 6.14892006e-01 3.54937352e-02 -1.63085088e-01 3.22053313e-01 -1.36075163e+00 -1.08309850e-01 1.81753606e-01 6.01118565e-01 2.18417030e-02 8.74016404e-01 6.73715115e-01 4.49049979e-01 -3.58328134e-01 -4.14902419e-01 4.32951182e-01 -4.08686966e-01 -5.83740652e-01 5.90322018e-01 8.42500985e-01 -4.94731635e-01 -1.04251906e-01 7.65786409e-01 5.35239100e-01 1.62794635e-01 2.78849602e-01 -1.29181552e+00 1.02964208e-01 -1.14686541e-01 -8.68449748e-01 9.84066818e-03 1.00676417e+00 -8.32752232e-03 7.53964245e-01 -7.36452401e-01 9.19687927e-01 1.41613114e+00 8.36745799e-01 5.95611751e-01 -1.06776059e+00 -4.03517067e-01 5.62576830e-01 7.00365603e-02 -1.33730149e+00 -4.34380352e-01 6.97715938e-01 -4.38552171e-01 5.89275002e-01 1.79398432e-01 1.31978273e+00 1.32908142e+00 3.71888638e-01 7.63196409e-01 8.92382801e-01 -2.48026863e-01 1.82415619e-02 -3.86766434e-01 2.34546550e-02 1.01711166e+00 4.98391181e-01 9.33466628e-02 -7.77999878e-01 -1.92913368e-01 9.08385575e-01 3.31671476e-01 1.62318945e-02 -8.75183463e-01 -1.36056697e+00 3.07710201e-01 -1.34743582e-02 -2.39916787e-01 -2.85687596e-01 3.67908567e-01 2.62441665e-01 -2.15747148e-01 2.06226438e-01 -2.11187154e-01 -4.45163220e-01 -6.31974861e-02 -8.12804580e-01 7.23935604e-01 6.97928965e-01 1.07079566e+00 2.88574517e-01 -3.06991994e-01 -6.07356615e-03 4.79803711e-01 5.71484089e-01 7.52546847e-01 -3.94443661e-04 -1.10706687e+00 5.21599054e-01 5.72302818e-01 4.68186259e-01 -1.25285208e+00 -5.64738989e-01 -1.41322955e-01 -7.96665967e-01 -1.07980277e-02 5.77619016e-01 -2.53114074e-01 -6.45913422e-01 1.92172515e+00 9.18941319e-01 1.59569576e-01 -2.71677494e-01 1.02603245e+00 3.00072014e-01 3.06751132e-01 -1.30183205e-01 -1.90028727e-01 1.58422542e+00 -9.55368876e-01 -7.84184933e-01 -6.84437215e-01 7.88371777e-04 -3.60168725e-01 4.13306177e-01 3.94807100e-01 -1.11558247e+00 -7.75549829e-01 -8.55826318e-01 8.85474458e-02 1.23677209e-01 7.70627111e-02 3.95915389e-01 5.94019771e-01 -6.85133100e-01 4.28819835e-01 -1.42243707e+00 -6.14198923e-01 -2.18528479e-01 3.97747308e-01 -7.54387259e-01 8.85526091e-02 -9.28762913e-01 1.12081945e+00 2.62326479e-01 3.28284740e-01 -7.69629061e-01 -1.38366595e-01 -1.17514086e+00 -3.68044913e-01 7.87824094e-01 -1.29272091e+00 1.24012673e+00 -3.00095201e-01 -1.47729218e+00 8.58387947e-01 -3.47878993e-01 -3.22324276e-01 8.91634643e-01 -8.18585098e-01 -5.44277728e-02 1.42601505e-01 3.29408608e-02 4.39131796e-01 8.68361771e-01 -1.30918360e+00 -5.36005974e-01 -8.89990389e-01 1.70611255e-02 4.49374229e-01 7.76284188e-02 -1.23709671e-01 -1.18282759e+00 -5.74452043e-01 3.44530493e-01 -1.57075095e+00 -3.84014904e-01 2.12303564e-01 -4.60352868e-01 1.84276327e-01 6.99216485e-01 -8.56594324e-01 1.16687012e+00 -1.74859893e+00 7.48601079e-01 7.31597170e-02 2.53364533e-01 -2.01837003e-01 4.92874235e-01 2.25887850e-01 3.25629115e-01 -3.34065109e-01 2.98142917e-02 -8.34040940e-01 4.11251970e-02 3.36819410e-01 2.09479019e-01 8.43375921e-01 -2.68532425e-01 6.91116333e-01 -7.67556429e-01 -8.45857918e-01 3.54076654e-01 4.17671621e-01 -5.94510376e-01 2.94440329e-01 4.74227481e-02 7.28021383e-01 -4.42541152e-01 6.58863723e-01 5.08576512e-01 -4.08494681e-01 6.57825291e-01 -3.10356319e-01 4.67613265e-02 -4.13761050e-01 -1.56600380e+00 2.01952720e+00 7.37647619e-03 -6.46504313e-02 6.70979684e-03 -5.77373683e-01 4.75367129e-01 5.08989155e-01 7.49290287e-01 -1.58021465e-01 7.28125274e-02 -2.45719954e-01 -2.91471362e-01 -2.14384213e-01 3.58504653e-01 -5.22584319e-02 -4.70073402e-01 3.08027327e-01 -1.89501405e-01 1.08285755e-01 1.13257885e-01 9.41950083e-02 8.32904637e-01 6.57749116e-01 9.70202804e-01 3.94199640e-02 5.32500207e-01 -2.81505167e-01 8.68575394e-01 5.88897169e-01 -4.79663640e-01 5.12647271e-01 3.13925706e-02 -6.74977779e-01 -8.33336055e-01 -1.33954430e+00 2.67809629e-01 7.06213951e-01 4.41800356e-01 -7.28075504e-01 -7.35401571e-01 -6.93290472e-01 5.87831438e-02 -4.40433063e-02 -6.08917654e-01 9.46847573e-02 -8.43466163e-01 -3.81717473e-01 2.43283138e-01 8.08411002e-01 5.50424933e-01 -4.14846748e-01 -1.35841060e+00 1.08453833e-01 -4.31572676e-01 -1.59007430e+00 -6.72319055e-01 -3.85777891e-01 -9.13569629e-01 -1.24957478e+00 -8.23728919e-01 -3.15308779e-01 6.98189735e-01 1.99377865e-01 1.17536497e+00 5.46278805e-03 -1.21890277e-01 7.64742136e-01 -2.62884557e-01 -2.73695707e-01 -5.59688732e-02 -2.11182907e-01 6.76668167e-01 -1.63761340e-02 -3.51731442e-02 -5.83223581e-01 -5.94972312e-01 5.77460170e-01 -2.89202511e-01 4.64327931e-01 2.49546200e-01 5.72010756e-01 8.23472440e-01 -1.61578238e-01 -1.82879135e-01 -7.61863053e-01 -2.80200858e-02 3.48142348e-02 -8.10231209e-01 4.35480148e-01 -1.43296346e-01 -4.19122837e-02 2.93547213e-01 -5.38264990e-01 -9.41847563e-01 7.71061897e-01 2.68719912e-01 -4.85039055e-01 -1.72993317e-01 1.39007553e-01 -1.40360340e-01 1.27310485e-01 8.00491795e-02 -4.96900678e-02 1.75001356e-03 -4.95306909e-01 2.88677305e-01 2.26408318e-02 8.92892122e-01 -6.49957180e-01 8.70025456e-01 7.70222485e-01 3.16262305e-01 -6.13913119e-01 -1.03044808e+00 -5.90269387e-01 -1.37243724e+00 -6.48553014e-01 1.02358019e+00 -1.17574739e+00 -1.18686700e+00 5.21771312e-01 -1.25914264e+00 -5.25113940e-02 2.04099342e-01 5.00365734e-01 -6.84065759e-01 6.92600667e-01 -6.34676516e-01 -1.02934849e+00 -2.94974357e-01 -1.00738955e+00 1.56084526e+00 -1.95092931e-02 -6.91053212e-01 -8.97768676e-01 4.34140831e-01 3.32326859e-01 -4.34210032e-01 8.26920986e-01 1.90296426e-01 -1.07517838e-01 -4.13208187e-01 -3.12032104e-01 4.11727607e-01 -2.78678983e-01 6.02293126e-02 -1.37090966e-01 -4.27521765e-01 -5.17853856e-01 1.37639297e-02 -4.53156158e-02 3.26491773e-01 5.09372115e-01 6.60400629e-01 -1.67817757e-01 -6.22115850e-01 3.59953731e-01 1.08094656e+00 -1.95608988e-01 1.92555636e-01 2.16831818e-01 8.95946920e-01 5.92228115e-01 7.72430241e-01 8.04394543e-01 8.08969438e-01 1.27148926e+00 2.88738251e-01 2.11848646e-01 4.75298017e-02 -3.97409409e-01 6.63442671e-01 8.68126690e-01 -4.48305726e-01 -2.31859118e-01 -1.01796722e+00 2.76052445e-01 -2.15884352e+00 -1.10976958e+00 -6.20548949e-02 2.26209760e+00 5.16126215e-01 1.33357942e-01 5.52700818e-01 -4.25674245e-02 7.34692574e-01 1.79719627e-01 -6.25115573e-01 3.33746642e-01 2.92295039e-01 -2.50274330e-01 2.76229173e-01 5.04411161e-01 -1.14519131e+00 8.69287133e-01 6.81598997e+00 -5.76194450e-02 -5.75360894e-01 5.63153550e-02 2.01483015e-02 -3.22874278e-01 3.61791641e-01 6.00565486e-02 -1.21131241e+00 2.35854104e-01 5.88305473e-01 -2.83575691e-02 6.40472695e-02 6.67487860e-01 1.58824697e-01 -4.49417025e-01 -1.37974191e+00 1.26194775e+00 3.44953597e-01 -7.26158977e-01 -6.32193387e-02 1.66526541e-01 5.75194180e-01 -3.74313563e-01 -1.97362766e-01 -5.72459586e-02 3.36327434e-01 -4.61279631e-01 1.13108635e+00 6.77577913e-01 4.42757428e-01 -5.29776275e-01 3.59803170e-01 7.10769653e-01 -1.69717562e+00 4.83031757e-02 1.51839890e-02 -2.25198343e-01 7.97947109e-01 3.52954268e-01 -3.78452122e-01 7.70336807e-01 8.20814013e-01 1.01394773e+00 -6.84635937e-01 6.97770417e-01 -3.06253314e-01 -2.02217158e-02 -5.18506169e-01 4.27372456e-01 -3.84043515e-01 -1.15327261e-01 6.45541370e-01 7.94662654e-01 3.00174326e-01 1.22827232e-01 8.51324379e-01 5.10456443e-01 3.26168448e-01 -3.80211145e-01 -5.89215159e-01 3.92044663e-01 2.98164845e-01 1.02956283e+00 -6.92465544e-01 -4.57020909e-01 -4.19703335e-01 1.37094676e+00 2.46219069e-01 5.56268962e-04 -1.10034251e+00 5.17407298e-01 4.67563629e-01 1.22927792e-01 2.12196901e-01 -6.34126782e-01 1.15138158e-01 -1.58128834e+00 3.74734372e-01 -9.50910270e-01 7.30609655e-01 -7.32177496e-01 -1.07648528e+00 4.05295551e-01 4.98908401e-01 -1.53371620e+00 -5.89671552e-01 -3.56122464e-01 -6.52100444e-02 4.35092837e-01 -8.41933131e-01 -1.38739479e+00 -4.24867928e-01 7.52003789e-01 6.48991585e-01 3.91104013e-01 7.76724577e-01 2.81640906e-02 -7.80677438e-01 3.17959696e-01 -5.73140800e-01 1.60394564e-01 7.65900195e-01 -1.10970616e+00 7.55883098e-01 1.09355891e+00 7.43853152e-02 7.83899546e-01 7.93284118e-01 -1.08242333e+00 -1.63614988e+00 -8.34684432e-01 6.07435465e-01 -1.11201477e+00 2.08627105e-01 -6.18216217e-01 -3.80104572e-01 1.15352416e+00 -1.32367253e-01 1.57259285e-01 5.63548386e-01 2.41532028e-01 -2.62250781e-01 1.23123646e-01 -8.09407234e-01 6.75780177e-01 1.44561255e+00 -2.78728068e-01 -6.77499890e-01 3.95731591e-02 4.08801109e-01 -9.17293310e-01 -9.98998523e-01 5.10698318e-01 1.08847260e+00 -1.08758879e+00 1.31396496e+00 -4.30546582e-01 -2.53801551e-02 -5.28293729e-01 -1.98598012e-01 -1.04855049e+00 -2.82572538e-01 -6.00210845e-01 -7.58011937e-01 7.45428443e-01 -6.51766509e-02 -2.80803740e-01 9.05272663e-01 8.22988749e-01 3.76066983e-01 -4.41815227e-01 -7.54962265e-01 -9.33256805e-01 -4.66544688e-01 -1.59328192e-01 2.45136023e-01 4.72269535e-01 -4.24411520e-02 4.36301261e-01 -1.03140914e+00 6.58789515e-01 1.01510918e+00 3.00972253e-01 1.34593093e+00 -1.44414198e+00 -5.21486640e-01 2.17571005e-01 -4.73357946e-01 -1.18478501e+00 2.53794361e-02 -1.73017129e-01 6.40352294e-02 -1.26984644e+00 6.82726741e-01 1.79131612e-01 2.68977135e-01 3.99059564e-01 -4.22433138e-01 -2.51963665e-03 5.66362560e-01 2.02689871e-01 -1.12421370e+00 3.66864532e-01 1.15784347e+00 1.45932987e-01 1.20565660e-01 1.91520587e-01 -1.59028158e-01 1.11526012e+00 4.12856847e-01 -4.95091856e-01 -3.23263586e-01 -4.45432872e-01 2.17812330e-01 5.87637782e-01 7.15627909e-01 -1.28247941e+00 4.13051695e-01 -1.66893393e-01 7.68090844e-01 -9.69711363e-01 8.03057373e-01 -1.05363345e+00 8.09821904e-01 7.85585046e-01 1.85208514e-01 6.51740015e-01 4.02971879e-02 8.94276321e-01 2.71759108e-02 3.65208685e-01 7.09599495e-01 -5.54696918e-01 -7.89916635e-01 5.70408821e-01 -6.98357821e-02 -3.43076102e-02 1.15351999e+00 -4.76096690e-01 2.34208509e-01 -4.48171139e-01 -9.90138769e-01 3.63539070e-01 8.02432954e-01 4.71357226e-01 5.60287535e-01 -1.35581005e+00 -3.09034765e-01 1.69591919e-01 1.18038110e-01 9.95919853e-02 3.41852725e-01 9.04699981e-01 -3.87686610e-01 2.91728705e-01 -3.20695519e-01 -9.33258414e-01 -1.74302101e+00 4.79439974e-01 2.43422866e-01 -4.69555676e-01 -8.05081010e-01 3.35700005e-01 8.12678486e-02 -4.82194602e-01 1.49905622e-01 -1.38958767e-01 -1.42465234e-01 -1.93157285e-01 5.38004339e-01 5.30540824e-01 -4.13420528e-01 -1.07568133e+00 -6.06772363e-01 1.19725990e+00 1.40254930e-01 -3.47661883e-01 1.18139780e+00 -5.46824455e-01 8.87887254e-02 5.99535644e-01 8.50841105e-01 1.12068921e-01 -1.51623380e+00 -7.69109428e-02 -1.42131940e-01 -6.04075730e-01 -4.95866686e-01 -4.76291865e-01 -8.24581921e-01 6.05743945e-01 5.84238529e-01 -2.93847144e-01 8.31646919e-01 -6.15357384e-02 8.47245872e-01 3.56351852e-01 9.05117273e-01 -9.54164147e-01 2.29316413e-01 5.02556622e-01 7.65280128e-01 -1.19372344e+00 4.97869462e-01 -6.99189782e-01 -7.31226206e-01 8.63899410e-01 8.05041015e-01 8.46520290e-02 5.32509685e-01 2.77314007e-01 -1.53781593e-01 -4.01919484e-01 -6.47750795e-01 -1.04576990e-01 4.25109059e-01 4.23369944e-01 2.60148287e-01 3.08199730e-02 4.27738875e-02 4.55341429e-01 -9.75255519e-02 -8.97504613e-02 1.54474095e-01 1.12567568e+00 -1.46591902e-01 -1.23666739e+00 -7.53625333e-01 -2.42866814e-01 -3.22209179e-01 5.54382861e-01 -5.23728251e-01 9.57477629e-01 -5.40605513e-03 7.68594444e-01 -1.65091917e-01 -4.13866013e-01 6.99680686e-01 4.25916761e-02 9.28560615e-01 -6.50345087e-01 -2.62632132e-01 2.97373086e-01 4.27992363e-03 -9.02707696e-01 -8.08741570e-01 -1.08368719e+00 -1.14292836e+00 -3.47568721e-01 -2.97048450e-01 -2.68722981e-01 1.70853361e-01 9.53520894e-01 2.26014450e-01 2.75133997e-01 1.92052037e-01 -1.20187736e+00 -4.11125183e-01 -5.74587524e-01 -3.71931791e-01 7.62386560e-01 2.21214131e-01 -1.10455048e+00 4.03425917e-02 4.52152163e-01]	[7.042168140411377, -0.9822030663490295]
53dd2bc6-0087-4640-a29b-1385cc2746e6	hierarchical-supervision-and-shuffle-data	2304.01464	null	https://arxiv.org/abs/2304.01464v1	https://arxiv.org/pdf/2304.01464v1.pdf	Hierarchical Supervision and Shuffle Data Augmentation for 3D Semi-Supervised Object Detection	State-of-the-art 3D object detectors are usually trained on large-scale datasets with high-quality 3D annotations. However, such 3D annotations are often expensive and time-consuming, which may not be practical for real applications. A natural remedy is to adopt semi-supervised learning (SSL) by leveraging a limited amount of labeled samples and abundant unlabeled samples. Current pseudolabeling-based SSL object detection methods mainly adopt a teacher-student framework, with a single fixed threshold strategy to generate supervision signals, which inevitably brings confused supervision when guiding the student network training. Besides, the data augmentation of the point cloud in the typical teacher-student framework is too weak, and only contains basic down sampling and flip-and-shift (i.e., rotate and scaling), which hinders the effective learning of feature information. Hence, we address these issues by introducing a novel approach of Hierarchical Supervision and Shuffle Data Augmentation (HSSDA), which is a simple yet effective teacher-student framework. The teacher network generates more reasonable supervision for the student network by designing a dynamic dual-threshold strategy. Besides, the shuffle data augmentation strategy is designed to strengthen the feature representation ability of the student network. Extensive experiments show that HSSDA consistently outperforms the recent state-of-the-art methods on different datasets. The code will be released at https://github.com/azhuantou/HSSDA.	['Xinbo Gao', 'Deyu Meng', 'Pengcheng Li', 'Fangcen Liu', 'Chenqiang Gao', 'Chuandong Liu']	2023-04-04	null	http://openaccess.thecvf.com//content/CVPR2023/html/Liu_Hierarchical_Supervision_and_Shuffle_Data_Augmentation_for_3D_Semi-Supervised_Object_CVPR_2023_paper.html	http://openaccess.thecvf.com//content/CVPR2023/papers/Liu_Hierarchical_Supervision_and_Shuffle_Data_Augmentation_for_3D_Semi-Supervised_Object_CVPR_2023_paper.pdf	cvpr-2023-1	['semi-supervised-object-detection']	['computer-vision']	[-1.09328076e-01 2.31268667e-02 -3.75750929e-01 -4.07385528e-01 -4.64095265e-01 -3.01295787e-01 6.22222900e-01 -8.21197685e-03 -2.16690019e-01 2.35617116e-01 -1.84165165e-01 -3.29905272e-01 2.04115376e-01 -6.84836805e-01 -6.29062593e-01 -1.00234616e+00 4.28769827e-01 3.70887101e-01 7.67535269e-01 -1.33719072e-01 -2.69078761e-02 6.10582530e-01 -1.71634603e+00 -1.08468004e-01 8.83823872e-01 1.12669921e+00 4.79450792e-01 2.99606957e-02 -4.71361071e-01 3.40003073e-01 -4.18963730e-01 4.80603464e-02 4.71029162e-01 -2.47757822e-01 -3.54986966e-01 3.36036325e-01 2.26812765e-01 -5.12256086e-01 -3.83064091e-01 1.15830505e+00 5.68694055e-01 -3.77082191e-02 4.95392561e-01 -1.36931801e+00 -4.70843554e-01 2.17151135e-01 -8.73990476e-01 1.55659085e-02 2.04501897e-02 3.24829429e-01 7.49786854e-01 -1.20161247e+00 3.15693557e-01 1.17296779e+00 5.26709616e-01 6.12376809e-01 -1.01123834e+00 -9.44598973e-01 3.23449135e-01 -1.35869458e-02 -1.47790670e+00 -2.09882483e-01 1.18903017e+00 -3.00753444e-01 4.28121179e-01 1.13221407e-01 7.66546607e-01 9.61216927e-01 -5.80334187e-01 1.17886972e+00 1.00978923e+00 -3.86346310e-01 1.84545591e-01 3.49949330e-01 9.58461985e-02 7.15326726e-01 4.13786292e-01 9.23506990e-02 -3.68191808e-01 1.49601744e-02 1.13015246e+00 3.90967309e-01 -1.62708685e-01 -9.65848804e-01 -1.12088537e+00 7.04877138e-01 7.56422222e-01 1.20994203e-01 -7.52474517e-02 -4.22160506e-01 4.45146471e-01 2.21358165e-02 5.73660493e-01 6.80108890e-02 -5.05493999e-01 2.99161047e-01 -7.46705353e-01 1.36904269e-01 2.00021282e-01 1.10654318e+00 1.05470455e+00 -1.46352887e-01 2.15417314e-02 7.40427375e-01 5.63904643e-01 5.17320752e-01 3.76626045e-01 -4.84041452e-01 4.82255548e-01 1.11973941e+00 1.01902947e-01 -7.75057256e-01 -4.40956533e-01 -5.54772258e-01 -8.81731749e-01 3.27003121e-01 4.35723275e-01 6.98982030e-02 -1.14393556e+00 1.46223557e+00 9.17108774e-01 2.29657650e-01 -2.50292510e-01 1.22591114e+00 1.10918581e+00 4.32918102e-01 -8.09882954e-02 -2.19071224e-01 1.08765626e+00 -9.77352500e-01 -4.45157230e-01 -2.76661456e-01 8.44704807e-01 -6.38488173e-01 1.35228169e+00 1.28629029e-01 -7.19332099e-01 -7.81343937e-01 -1.10367143e+00 -5.72623052e-02 -2.42312089e-01 5.05621314e-01 6.10365629e-01 4.22022313e-01 -4.53572601e-01 1.64653987e-01 -8.80774140e-01 -2.15712681e-01 7.66203582e-01 2.97515959e-01 -3.08601081e-01 -2.20996380e-01 -9.17361856e-01 5.95755398e-01 4.08789217e-01 2.43336663e-01 -8.09949994e-01 -5.46583414e-01 -8.86462390e-01 -2.38767490e-01 6.59956694e-01 -3.34884614e-01 1.34777725e+00 -6.43007994e-01 -1.40942395e+00 9.76237476e-01 1.09460041e-01 7.12819397e-02 6.52629554e-01 -1.65704623e-01 -8.48489627e-02 5.57350926e-02 2.68486589e-01 7.29743361e-01 9.89356458e-01 -1.40309823e+00 -5.56105852e-01 -6.72538221e-01 -1.84453145e-01 2.92015105e-01 -5.29496312e-01 -9.24543217e-02 -4.95240629e-01 -5.93962610e-01 6.35017455e-01 -8.41180742e-01 -4.01199788e-01 2.37955123e-01 -4.69951719e-01 -4.69314992e-01 1.13021398e+00 3.85496137e-03 9.61726665e-01 -2.29579830e+00 -1.96254224e-01 5.47156855e-02 3.15802127e-01 5.86786985e-01 -7.42710978e-02 -6.39758781e-02 -1.05357409e-01 -1.82714775e-01 -1.12317659e-01 -4.02864784e-01 -1.96610361e-01 2.24263102e-01 -3.24622691e-01 5.06257594e-01 2.67355353e-01 7.28189409e-01 -1.06229520e+00 -7.03115344e-01 4.40346181e-01 3.11255127e-01 -4.23589528e-01 3.99552345e-01 -3.80916506e-01 5.85941851e-01 -9.18969512e-01 9.53277707e-01 9.12492871e-01 -4.45465237e-01 -2.37096861e-01 -3.41558427e-01 -3.01670849e-01 3.26258808e-01 -1.41793036e+00 1.75106788e+00 -5.95914721e-02 8.47976655e-02 -9.24217049e-03 -1.12692809e+00 1.36073995e+00 1.33012012e-01 4.08726066e-01 -2.95836627e-01 3.98689568e-01 3.92208159e-01 -2.07377225e-01 -4.03202057e-01 2.00434942e-02 -1.04512289e-01 9.07426327e-02 3.45011503e-01 1.83156524e-02 -2.18464807e-01 -1.38075277e-01 1.53364405e-01 7.72863388e-01 4.21630830e-01 1.09738909e-01 -7.74164125e-02 5.35765231e-01 6.11678325e-02 8.89547527e-01 5.04759312e-01 -3.42232615e-01 5.84097266e-01 2.48238727e-01 -3.92046869e-01 -9.45947647e-01 -8.15917194e-01 -2.24929944e-01 7.99263716e-01 4.86314952e-01 -2.98332930e-01 -6.04427218e-01 -1.13609886e+00 9.63989496e-02 2.42813185e-01 -4.32581067e-01 -2.57090747e-01 -4.34470534e-01 -7.90060520e-01 1.83969811e-01 7.08948910e-01 7.40351498e-01 -9.51520562e-01 -4.42773342e-01 6.49692565e-02 1.79297403e-01 -1.05578327e+00 -4.19609189e-01 3.29718411e-01 -1.07297206e+00 -1.11804581e+00 -7.45241404e-01 -8.91074240e-01 1.07403290e+00 8.86541963e-01 6.88631356e-01 2.34366357e-01 -1.42223537e-01 -2.40663707e-01 -4.97119635e-01 -5.83091140e-01 -4.22419347e-02 2.49280959e-01 3.89330089e-01 -1.15785010e-01 6.87442005e-01 -6.02565587e-01 -6.43335819e-01 5.60889065e-01 -7.26130962e-01 2.28977233e-01 9.27043259e-01 9.15012836e-01 7.39172995e-01 -8.39620456e-02 6.26294672e-01 -6.88777447e-01 6.20173216e-02 -2.22209468e-01 -7.53078818e-01 -5.00863791e-02 -6.73985541e-01 -2.11851165e-01 6.10821128e-01 -6.90377891e-01 -1.05270493e+00 5.33777654e-01 -2.18039036e-01 -9.42027867e-01 -5.05608201e-01 4.07807156e-02 -6.13693953e-01 -2.77438223e-01 6.33528650e-01 2.18176976e-01 6.83288053e-02 -7.37451196e-01 1.86439261e-01 8.32140982e-01 3.60742033e-01 -4.84490901e-01 1.21174109e+00 5.80285013e-01 8.14537238e-03 -6.14569724e-01 -1.35237288e+00 -6.34543180e-01 -8.87245297e-01 -1.33669659e-01 5.38578510e-01 -8.93456340e-01 -3.51699412e-01 7.46247351e-01 -1.00974834e+00 -3.29193413e-01 -4.22714800e-01 5.44524610e-01 -2.17221171e-01 2.40380242e-01 -4.57269788e-01 -7.41443276e-01 -1.15045600e-01 -1.15629637e+00 1.14504337e+00 4.90998536e-01 3.27584505e-01 -4.16533560e-01 -2.47579649e-01 4.09890145e-01 2.87733749e-02 -9.04412568e-03 5.37295043e-01 -7.48391330e-01 -7.40815341e-01 -2.88121223e-01 -5.24860322e-01 4.16234225e-01 3.22746783e-01 -1.72493100e-01 -9.57888365e-01 -1.96174845e-01 1.12186179e-01 -5.26397049e-01 6.82607830e-01 8.13673809e-02 1.33382344e+00 1.67154357e-01 -4.25943702e-01 4.49385941e-01 9.52056885e-01 -1.26323730e-01 1.47534132e-01 3.47629458e-01 1.03528631e+00 4.60450351e-01 1.07658112e+00 4.10542130e-01 4.97691661e-01 5.09063065e-01 5.73174298e-01 -3.27673793e-01 -8.23176354e-02 -5.04459918e-01 9.43796039e-02 8.38763833e-01 3.44532542e-02 3.06389272e-01 -9.60901976e-01 4.11511272e-01 -1.80303299e+00 -5.58945358e-01 -2.38415524e-01 2.07409978e+00 9.32872593e-01 5.67041576e-01 2.65376091e-01 4.88672704e-01 7.62681663e-01 9.85792354e-02 -7.53965735e-01 4.21902150e-01 1.60704926e-02 -9.22185630e-02 4.21364605e-01 -4.88234200e-02 -1.20142150e+00 8.93551826e-01 4.62571049e+00 9.38242495e-01 -1.14402187e+00 1.04231127e-01 4.18496281e-01 9.51520950e-02 -6.55569090e-03 3.91184818e-03 -1.27110171e+00 4.45824683e-01 1.72301069e-01 2.89042473e-01 -1.41159743e-01 1.18588841e+00 2.64846355e-01 5.45077026e-03 -9.40763056e-01 1.02743757e+00 -7.21027404e-02 -1.15011549e+00 -7.28623196e-02 6.12029769e-02 6.36358559e-01 1.27565160e-01 -2.69571632e-01 3.37715447e-01 2.52107233e-02 -3.77299309e-01 5.48315942e-01 1.27743306e-02 7.20241368e-01 -5.84367573e-01 7.63346434e-01 8.04355443e-01 -1.23852229e+00 -3.74740325e-02 -4.69352812e-01 2.72036362e-02 -2.76718512e-02 9.43674028e-01 -5.85327566e-01 3.94596696e-01 9.11853254e-01 7.68583477e-01 -6.56307697e-01 1.08643067e+00 -6.22802675e-01 5.51419973e-01 -5.85600913e-01 -8.34853128e-02 2.58805513e-01 -1.46889195e-01 5.94876885e-01 7.33576059e-01 2.45347589e-01 3.16652745e-01 6.63688362e-01 6.14999712e-01 -1.03902377e-01 6.19791672e-02 -4.67946053e-01 3.08491200e-01 7.05519736e-01 1.36878800e+00 -7.96579003e-01 -2.56726027e-01 -5.66917419e-01 6.36655629e-01 2.11286157e-01 4.44973307e-03 -4.94698644e-01 -3.57988209e-01 3.87737751e-01 3.93644422e-01 4.12193209e-01 -1.99781597e-01 -2.91320950e-01 -1.19069910e+00 1.78726688e-01 -5.64631283e-01 3.53808790e-01 -7.06768215e-01 -1.39487183e+00 3.49208087e-01 3.91885191e-02 -1.67395318e+00 2.24965617e-01 -5.45776725e-01 -7.40777612e-01 5.63106537e-01 -1.71421576e+00 -1.24549413e+00 -7.34937310e-01 5.40095925e-01 4.88671362e-01 -6.92430735e-02 4.50854093e-01 4.47978914e-01 -8.49174857e-01 5.46765268e-01 -3.50130498e-01 2.02895641e-01 7.03092575e-01 -1.05009246e+00 2.32265264e-01 6.41232610e-01 -3.70389293e-03 4.36295837e-01 4.95541692e-01 -5.40251851e-01 -1.34509730e+00 -1.21458948e+00 4.51905429e-01 -2.57483482e-01 5.32076776e-01 -5.68340182e-01 -1.25667536e+00 4.58539635e-01 -4.51708794e-01 5.95656455e-01 4.31432188e-01 -7.14205578e-02 -2.89944053e-01 -2.42002815e-01 -1.08190215e+00 4.45756733e-01 1.18564427e+00 -3.37506145e-01 -7.57635832e-01 4.36669827e-01 7.98995674e-01 -5.84590256e-01 -5.47977924e-01 7.93172002e-01 1.93458259e-01 -9.37710881e-01 8.56961071e-01 -2.01637298e-01 1.57908261e-01 -8.22727203e-01 1.05880685e-01 -1.03707540e+00 -1.06899194e-01 -3.72731566e-01 -3.43839198e-01 1.47272897e+00 -2.87935901e-02 -5.41278720e-01 1.16148150e+00 1.63851768e-01 -3.08192790e-01 -1.02075064e+00 -7.83821821e-01 -8.90909910e-01 -1.72496945e-01 -2.52988964e-01 7.88853347e-01 1.11462569e+00 -3.57671559e-01 4.15654242e-01 -5.91665274e-03 3.15382630e-01 7.95019329e-01 3.34900379e-01 1.04959404e+00 -1.48390925e+00 4.60367277e-02 -3.50490481e-01 -4.02372032e-01 -1.41259658e+00 -5.50553910e-02 -7.46499479e-01 1.75503820e-01 -1.23649371e+00 7.58669674e-02 -9.61664855e-01 -2.02952772e-01 6.73369586e-01 -3.07828307e-01 2.50235021e-01 -4.70311977e-02 5.14765322e-01 -5.17294943e-01 9.20115948e-01 1.50695336e+00 1.51726097e-01 -4.25387025e-01 3.81523073e-01 -5.15793443e-01 9.35432017e-01 9.22950327e-01 -5.03281832e-01 -4.07389730e-01 -2.59526670e-01 -2.03813851e-01 -3.86900872e-01 4.83048290e-01 -7.68437207e-01 3.30217957e-01 -1.87882200e-01 5.30710638e-01 -1.05510473e+00 1.74781069e-01 -7.53420532e-01 -4.64331269e-01 2.68564820e-01 1.30882591e-01 -3.21489036e-01 1.21248513e-02 4.56366718e-01 -1.71938837e-01 -1.94958925e-01 8.32719564e-01 -1.69807300e-01 -6.63972318e-01 6.94238424e-01 1.86393812e-01 -9.46051329e-02 1.13092816e+00 -2.59522527e-01 -2.08597541e-01 1.67111889e-01 -3.70208144e-01 6.07413054e-01 5.80312669e-01 3.91483217e-01 5.35613418e-01 -1.53406215e+00 -4.37938958e-01 5.38808405e-01 3.24367851e-01 9.43075955e-01 1.64038450e-01 6.48883641e-01 -8.88353959e-03 1.72987208e-01 -4.92955334e-02 -8.72655630e-01 -1.05561125e+00 5.63597739e-01 1.12496875e-01 -4.04774360e-02 -7.78356314e-01 8.14877152e-01 2.10258380e-01 -8.00935984e-01 3.55113387e-01 -1.41756773e-01 -7.77813569e-02 1.77605569e-01 4.22238827e-01 1.20654523e-01 1.61069050e-01 -3.57409775e-01 -4.02285010e-01 5.42872906e-01 -2.33693242e-01 3.86071503e-01 1.37654805e+00 2.19795108e-02 1.47155106e-01 4.33595061e-01 9.11102593e-01 -3.24804217e-01 -1.45093262e+00 -6.02257192e-01 -2.06909746e-01 -5.36130548e-01 2.43662640e-01 -2.35428751e-01 -1.02690077e+00 1.05258632e+00 4.67295647e-01 2.75247931e-01 1.12932122e+00 2.74683565e-01 8.33410621e-01 4.32859689e-01 3.54222655e-01 -1.02407432e+00 2.67918706e-01 2.99127042e-01 6.15162373e-01 -1.55269980e+00 1.84056163e-01 -7.25329399e-01 -3.27238858e-01 8.24168265e-01 1.16844630e+00 -4.58701886e-02 7.37744629e-01 1.19645655e-01 1.77810505e-01 -1.94831312e-01 -4.04880971e-01 -3.13465327e-01 1.95349947e-01 3.66156161e-01 6.90223053e-02 -1.66464582e-01 -5.61981834e-02 7.58269250e-01 4.54308763e-02 -6.73373267e-02 1.15306117e-01 1.15148258e+00 -5.08492768e-01 -1.22883081e+00 -4.53976631e-01 5.41711211e-01 1.00157253e-01 2.51999974e-01 -1.86663091e-01 8.33575249e-01 3.11329812e-01 5.40780187e-01 -1.64220557e-02 -3.74924749e-01 4.98754621e-01 -6.75281882e-02 3.37502956e-01 -8.09277236e-01 -2.54517943e-01 2.95232177e-01 -4.43892837e-01 -4.28496033e-01 -5.99371970e-01 -8.52539718e-01 -1.43433249e+00 1.11692570e-01 -7.97515869e-01 6.01535663e-02 5.41022241e-01 9.46759164e-01 2.20680103e-01 3.02582443e-01 1.00460994e+00 -1.18600142e+00 -7.87862659e-01 -1.06688333e+00 -4.63449389e-01 2.33500674e-01 3.48621279e-01 -1.17392707e+00 -4.61464703e-01 -2.64895141e-01]	[7.758639335632324, -2.7805960178375244]
fb0b53f5-f222-4042-90a9-e78a79dec0bb	mrgan-multi-rooted-3d-shape-generation-with	2007.12944	null	https://arxiv.org/abs/2007.12944v1	https://arxiv.org/pdf/2007.12944v1.pdf	MRGAN: Multi-Rooted 3D Shape Generation with Unsupervised Part Disentanglement	We present MRGAN, a multi-rooted adversarial network which generates part-disentangled 3D point-cloud shapes without part-based shape supervision. The network fuses multiple branches of tree-structured graph convolution layers which produce point clouds, with learnable constant inputs at the tree roots. Each branch learns to grow a different shape part, offering control over the shape generation at the part level. Our network encourages disentangled generation of semantic parts via two key ingredients: a root-mixing training strategy which helps decorrelate the different branches to facilitate disentanglement, and a set of loss terms designed with part disentanglement and shape semantics in mind. Of these, a novel convexity loss incentivizes the generation of parts that are more convex, as semantic parts tend to be. In addition, a root-dropping loss further ensures that each root seeds a single part, preventing the degeneration or over-growth of the point-producing branches. We evaluate the performance of our network on a number of 3D shape classes, and offer qualitative and quantitative comparisons to previous works and baseline approaches. We demonstrate the controllability offered by our part-disentangled generation through two applications for shape modeling: part mixing and individual part variation, without receiving segmented shapes as input.	['Daniel Cohen-Or', 'Hao Zhang', 'Rinon Gal', 'Amit Bermano']	2020-07-25	null	null	null	null	['3d-shape-generation']	['computer-vision']	[ 3.04072827e-01 9.69075501e-01 -3.54275629e-02 7.63342232e-02 -5.27860403e-01 -1.20798814e+00 6.35865033e-01 -1.25086337e-01 4.45245504e-01 4.77463156e-01 8.41508955e-02 -1.93519443e-01 2.09846318e-01 -1.22848797e+00 -1.10730946e+00 -9.84698236e-01 -7.32044280e-02 8.33414376e-01 -1.32065400e-01 -3.39053333e-01 -3.03605527e-01 1.07453942e+00 -1.00750566e+00 1.74681798e-01 7.72009611e-01 7.73607969e-01 -3.28572690e-01 7.51040995e-01 7.78847188e-03 4.01515037e-01 -4.75127131e-01 -8.19776654e-01 8.45454991e-01 -1.44200191e-01 -3.37550402e-01 4.74022239e-01 6.20036781e-01 -1.55146256e-01 -3.57610375e-01 7.73027480e-01 2.20940188e-01 -2.63950080e-01 8.59342039e-01 -1.70236397e+00 -1.04501975e+00 6.36499166e-01 -5.98156869e-01 -5.69658220e-01 -1.35773987e-01 4.91835296e-01 1.19439840e+00 -5.82085609e-01 6.54234111e-01 1.28612149e+00 7.28592038e-01 7.85218775e-01 -1.64865577e+00 -6.95750177e-01 1.25067949e-01 -1.09728539e+00 -9.75228906e-01 -1.73981130e-01 1.21292627e+00 -5.74433565e-01 5.68507612e-01 4.30824161e-01 8.96401703e-01 1.00269675e+00 1.84922814e-01 8.53208780e-01 6.09867096e-01 1.53355345e-01 2.73413025e-02 -1.10300593e-01 -4.79878306e-01 8.90461802e-01 5.31829000e-01 2.64626235e-01 -1.02369428e-01 -3.48588467e-01 1.41493058e+00 1.69202134e-01 -2.96668530e-01 -1.10593379e+00 -1.01468670e+00 8.34162056e-01 9.25438881e-01 -6.20841831e-02 -2.11919546e-01 4.45981175e-01 -3.02336365e-02 1.04857102e-01 5.50273955e-01 7.14470625e-01 -5.25545239e-01 5.14896631e-01 -6.31593227e-01 6.19721413e-01 7.87214935e-01 1.37846518e+00 8.10914159e-01 3.20492744e-01 -3.29127491e-01 3.08651716e-01 3.51030976e-01 7.72255361e-01 -2.81519145e-01 -1.09504318e+00 5.04891038e-01 1.13069415e+00 -1.74976364e-01 -7.10547090e-01 -4.01097238e-02 -5.12998581e-01 -9.39418435e-01 8.65597785e-01 3.84363711e-01 -2.72860080e-01 -1.44074047e+00 2.02688599e+00 3.12215239e-01 2.22062767e-02 -1.24627598e-01 8.10846627e-01 6.73696518e-01 3.14181983e-01 -1.85287714e-01 4.33030695e-01 1.14025807e+00 -8.02131653e-01 -8.79485998e-03 -3.28254923e-02 3.07781175e-02 -5.14560997e-01 7.51421273e-01 -6.51650876e-02 -1.53262222e+00 -3.90525460e-02 -1.01401651e+00 -2.71283150e-01 -2.21580997e-01 -1.06876053e-01 7.06366062e-01 4.42636311e-01 -1.05182052e+00 7.34762967e-01 -1.03361571e+00 3.70021939e-01 1.02921355e+00 3.18022341e-01 -4.74567890e-01 2.38681152e-01 -5.97343385e-01 3.63946259e-01 -3.66859615e-01 -2.14661822e-01 -1.06021500e+00 -1.31027424e+00 -1.15515018e+00 1.55232444e-01 7.39666447e-02 -1.36811650e+00 1.01481187e+00 -9.11161244e-01 -1.34856093e+00 1.11996698e+00 2.37314835e-01 -2.77682781e-01 8.37811112e-01 2.85418361e-01 2.50123888e-01 1.06625766e-01 6.82116970e-02 9.80795920e-01 1.19106209e+00 -1.77457547e+00 2.77810674e-02 -6.04019165e-01 2.28920594e-01 3.46260190e-01 3.37717742e-01 -8.24430823e-01 -2.66877621e-01 -9.76761937e-01 2.20665216e-01 -1.04072034e+00 -3.43463033e-01 5.14078081e-01 -8.32807481e-01 2.22507000e-01 8.10453176e-01 -4.02580082e-01 3.20929259e-01 -2.05084634e+00 6.32480264e-01 2.15137973e-01 8.87489915e-01 -1.95500717e-01 -3.34735513e-01 4.44750875e-01 -7.59578496e-02 5.83061695e-01 -6.51942372e-01 -7.42266119e-01 5.25708050e-02 3.65904808e-01 -4.80452091e-01 3.60716879e-01 6.63239002e-01 1.28839040e+00 -8.40398788e-01 -2.94456445e-02 4.62506339e-02 6.33112848e-01 -7.82133996e-01 8.19343477e-02 -6.10363662e-01 5.41805089e-01 -4.71464783e-01 6.80429935e-01 9.38756287e-01 -2.05290124e-01 -1.35920450e-01 -2.35991329e-02 1.69279799e-01 1.72250211e-01 -7.60031939e-01 1.63956344e+00 -4.55960125e-01 2.08292678e-01 7.03409910e-01 -4.93600100e-01 6.88107073e-01 3.25614244e-01 5.39762616e-01 -4.81118038e-02 5.63657619e-02 1.81601122e-01 -7.34848902e-02 6.47523850e-02 3.39066774e-01 -3.93343776e-01 -1.13164254e-01 3.92073274e-01 -5.94584905e-02 -1.03094959e+00 -3.71489495e-01 5.61228096e-01 1.12554979e+00 2.84525126e-01 -3.13865989e-01 -2.73993403e-01 -8.45846534e-02 -8.25697407e-02 4.88158166e-01 2.49640778e-01 1.38789415e-01 1.07411754e+00 8.88423622e-01 -1.81063920e-01 -1.44102740e+00 -1.57385445e+00 1.83766201e-01 4.48275506e-01 1.63684055e-01 2.33575795e-02 -6.28698468e-01 -8.47597420e-01 5.86012363e-01 6.33062243e-01 -9.00965393e-01 -2.81601250e-01 -6.70114398e-01 -2.67911375e-01 6.90732300e-01 4.70490128e-01 -3.55647691e-02 -9.98067796e-01 -2.42103800e-01 -1.68714255e-01 2.43979499e-01 -7.85052478e-01 -7.48775423e-01 1.91430241e-01 -1.00142896e+00 -1.22743154e+00 -8.70698810e-01 -7.19433248e-01 1.20369077e+00 1.30667761e-01 1.47544479e+00 1.68654904e-01 -1.75790429e-01 3.01367491e-01 1.05932325e-01 -4.51973677e-01 -7.41118908e-01 1.51039451e-01 -4.03686106e-01 -7.94266909e-02 -5.20547390e-01 -1.19126546e+00 -6.32023752e-01 1.19840451e-01 -1.04909456e+00 3.96457762e-01 4.63592798e-01 6.80593610e-01 6.28500581e-01 -3.07384700e-01 3.21036987e-02 -7.51322031e-01 3.64608169e-01 -5.17542541e-01 -7.42924511e-01 1.55725896e-01 -2.51809806e-01 3.93373877e-01 5.41238546e-01 -5.10248661e-01 -5.99477112e-01 7.09995404e-02 1.12644032e-01 -9.19139564e-01 8.44545066e-02 -3.45390081e-01 -6.38862967e-01 -1.27987653e-01 5.34393489e-01 1.29977942e-01 4.51094449e-01 -2.46973425e-01 9.61223722e-01 -9.15866941e-02 5.01433194e-01 -9.19621170e-01 1.36028838e+00 7.20023274e-01 3.35529387e-01 -5.13068736e-01 -3.36285800e-01 1.61131561e-01 -3.88716817e-01 1.19689263e-01 1.02233791e+00 -9.39279139e-01 -6.63497090e-01 6.41016543e-01 -1.19095612e+00 -7.00825691e-01 -9.56747353e-01 -4.15181190e-01 -6.46276832e-01 -1.35292441e-01 -6.05610192e-01 -5.77134490e-01 -5.07398307e-01 -1.04308569e+00 1.59603250e+00 9.45138857e-02 1.11947104e-01 -9.15699899e-01 -8.31659958e-02 8.76158923e-02 1.90331683e-01 9.23941553e-01 1.09882224e+00 -4.15595055e-01 -1.04789805e+00 -1.56995848e-01 8.00795630e-02 2.37792552e-01 2.44177997e-01 2.86828011e-01 -7.77819514e-01 -2.20962569e-01 -2.12098807e-01 -1.91546574e-01 7.37911403e-01 2.93372631e-01 8.10908675e-01 -7.45861590e-01 -4.65225875e-01 1.02147853e+00 1.20566738e+00 -2.69699991e-01 4.77222651e-01 -3.01830590e-01 1.23587132e+00 3.52719009e-01 -4.34825838e-01 1.36394352e-01 4.03819203e-01 2.53345847e-01 9.88306105e-01 -3.32775772e-01 -4.39716101e-01 -7.42515087e-01 -1.20183751e-01 4.19431329e-01 -9.89291146e-02 -4.90192652e-01 -5.84061146e-01 4.93992895e-01 -1.42078030e+00 -7.35991657e-01 -1.51916876e-01 2.02934742e+00 7.58178234e-01 1.05732724e-01 1.93619981e-01 -1.15007550e-01 5.20132780e-01 3.46356928e-01 -9.47316706e-01 -8.87786746e-02 -2.19737217e-01 4.13189977e-01 6.80756688e-01 6.56549096e-01 -7.96491385e-01 8.29304159e-01 5.74388170e+00 4.79180783e-01 -1.11945355e+00 -1.26918271e-01 6.81263030e-01 -3.04285198e-01 -1.29668462e+00 -4.75891167e-03 -2.92255044e-01 1.45966426e-01 -1.15401402e-01 -8.63928646e-02 6.30982578e-01 6.82431340e-01 -2.35226229e-01 6.45637572e-01 -1.00258744e+00 4.85843956e-01 -3.00556540e-01 -1.57844377e+00 3.67273301e-01 2.77393639e-01 9.94521976e-01 1.19430937e-01 3.26399118e-01 -9.23325792e-02 9.50945914e-01 -1.13976240e+00 1.13005924e+00 3.45350653e-01 8.28584433e-01 -7.52404213e-01 -1.94714874e-01 3.94401193e-01 -1.11565077e+00 2.87667841e-01 3.18620214e-03 1.51342347e-01 1.74431309e-01 6.74262285e-01 -8.16000223e-01 7.10293949e-01 8.16630796e-02 3.75935495e-01 -1.23431101e-01 5.11616945e-01 -6.57749116e-01 2.93170214e-01 -4.20875758e-01 3.60867053e-01 4.53312434e-02 -5.64792931e-01 1.18978977e+00 5.25668085e-01 1.18365288e-01 1.62801802e-01 -4.80613559e-02 1.77926421e+00 -5.22289336e-01 -5.59104264e-01 -8.25704932e-01 -2.78223932e-01 5.74844539e-01 1.31772768e+00 -6.31686628e-01 -9.09239277e-02 -9.18655023e-02 1.03932476e+00 3.42008144e-01 3.66540074e-01 -6.78698719e-01 -1.33953065e-01 1.29085624e+00 4.86408532e-01 4.09458637e-01 -3.45461279e-01 -9.93534744e-01 -1.20712876e+00 8.70983973e-02 -6.76788568e-01 -1.02722861e-01 -8.75950515e-01 -1.40553463e+00 5.33535421e-01 -3.01044196e-01 -1.09333146e+00 1.21766955e-01 -4.54280138e-01 -8.03512931e-01 1.06190467e+00 -1.12442386e+00 -1.84710693e+00 -2.64708567e-02 2.73435205e-01 2.21266493e-01 4.14918810e-02 5.76995492e-01 -1.47010341e-01 -4.28340673e-01 6.80494905e-01 -4.17157054e-01 2.65537083e-01 1.32203400e-01 -1.51670372e+00 1.03468442e+00 8.81209612e-01 2.44435817e-01 6.89755380e-01 4.82032776e-01 -8.86476576e-01 -1.43490338e+00 -1.20505047e+00 4.22859609e-01 -9.15163457e-01 3.38249803e-01 -7.47632086e-01 -7.10558236e-01 9.00674641e-01 -4.15919237e-02 2.18870774e-01 1.37061015e-01 -3.64816129e-01 -7.76411951e-01 2.82796949e-01 -1.48398697e+00 7.29711354e-01 1.38727534e+00 -3.39662462e-01 -3.59074503e-01 5.17112426e-02 1.31026149e+00 -7.52375543e-01 -7.61886895e-01 6.40328288e-01 4.90605861e-01 -8.64271164e-01 1.26262808e+00 -6.41705930e-01 8.09990942e-01 -2.02937245e-01 5.79660013e-02 -1.46831298e+00 -3.27362657e-01 -8.38438690e-01 -3.61212790e-01 1.05284870e+00 6.24473512e-01 -6.46705627e-01 1.10522759e+00 7.22362459e-01 -4.90729928e-01 -1.02353609e+00 -5.34260213e-01 -7.20469117e-01 7.24810779e-01 -9.89035070e-02 1.18833709e+00 8.73730958e-01 -5.61693311e-01 2.68225968e-01 9.84120220e-02 2.97318339e-01 7.02844024e-01 3.16469610e-01 7.64841616e-01 -1.11127532e+00 -4.16931987e-01 -8.47559333e-01 -2.05592394e-01 -1.03607154e+00 -9.46337208e-02 -1.21640134e+00 -1.62820637e-01 -1.51659715e+00 -1.24099158e-01 -8.01495731e-01 3.42991292e-01 7.89834082e-01 -9.61670950e-02 1.83141470e-01 4.16134655e-01 7.44992420e-02 3.00521791e-01 7.70791352e-01 2.20212865e+00 -2.79955328e-01 -2.83333898e-01 2.63593018e-01 -1.02811229e+00 6.28022730e-01 4.88881320e-01 -4.40057009e-01 -6.44198298e-01 -6.06070578e-01 2.34372064e-01 1.30569249e-01 8.02127481e-01 -5.55930316e-01 -2.91368812e-01 -4.40773256e-02 4.05115843e-01 -4.18268234e-01 4.74193096e-01 -8.19561660e-01 4.87546504e-01 4.36396211e-01 -9.31651890e-02 4.22878563e-02 2.16108471e-01 5.90814233e-01 3.61956596e-01 3.86397749e-01 8.49764168e-01 -3.61275643e-01 3.55274141e-01 8.76240909e-01 2.08012938e-01 2.87595123e-01 1.00485897e+00 -2.43799448e-01 -2.30423838e-01 -2.22408593e-01 -5.95873177e-01 2.28981659e-01 1.09594727e+00 5.49197495e-01 3.99360865e-01 -1.53288531e+00 -7.62596607e-01 6.86823189e-01 -1.07983194e-01 1.03617740e+00 5.87709919e-02 3.30039978e-01 -6.66065454e-01 -1.85452387e-01 -1.32196575e-01 -5.80676496e-01 -8.09474051e-01 4.64392722e-01 5.25556564e-01 -2.56974488e-01 -7.45218813e-01 1.13564622e+00 5.82514703e-01 -9.65063095e-01 -3.07965931e-02 -6.18114769e-01 5.03506958e-01 -3.23905885e-01 -1.87908798e-01 3.18435103e-01 -1.67339697e-01 -4.21026260e-01 -1.29996583e-01 5.10144293e-01 4.26540136e-01 -2.50547439e-01 1.40318716e+00 3.94075185e-01 -1.91449210e-01 7.66072050e-02 1.08651376e+00 5.30790448e-01 -1.85471690e+00 2.05871284e-01 -7.67819166e-01 -1.52435586e-01 -3.65788013e-01 -8.57275963e-01 -1.55938864e+00 6.45299554e-01 -2.23770872e-01 4.13667649e-01 7.52481699e-01 2.65546441e-01 9.03717220e-01 -3.25208455e-01 3.57604086e-01 -2.17297271e-01 1.78074077e-01 2.96650559e-01 1.28032613e+00 -7.35471070e-01 -2.50246853e-01 -7.94568241e-01 -5.06141782e-01 8.14962506e-01 5.41238427e-01 -6.00076318e-01 7.24630237e-01 7.56562829e-01 -9.31123123e-02 -4.92509753e-01 -4.86822903e-01 8.54806155e-02 4.87231880e-01 5.83078980e-01 3.84636559e-02 4.69943494e-01 2.84213156e-01 5.31679809e-01 -5.69328427e-01 -4.87508029e-01 2.36965701e-01 6.73068941e-01 8.30455590e-03 -1.24525046e+00 -2.03434572e-01 3.31982315e-01 -2.53804147e-01 -4.44089957e-02 -7.22311199e-01 9.06623960e-01 2.94579148e-01 3.96488488e-01 3.45006317e-01 -2.69998878e-01 3.88371497e-01 -1.04811281e-01 8.11037719e-01 -7.40067959e-01 -5.33893764e-01 -1.47800177e-01 -2.48335093e-01 -4.26971853e-01 1.01918869e-01 -5.97637236e-01 -1.36964440e+00 -4.53227818e-01 -1.30555421e-01 -1.66615292e-01 5.13655066e-01 4.40302759e-01 7.33138263e-01 5.92378736e-01 7.45563090e-01 -1.08601618e+00 -6.22459471e-01 -5.36542177e-01 -4.63512421e-01 6.01772487e-01 6.81339443e-01 -3.46994191e-01 -5.16701639e-01 -6.36847019e-02]	[8.9763765335083, -3.6267504692077637]
4a10e263-bb9a-481b-90ed-59e43129be2c	a-sequence-modelling-approach-to-question	null	null	https://aclanthology.org/2022.wordplay-1.4	https://aclanthology.org/2022.wordplay-1.4.pdf	A Sequence Modelling Approach to Question Answering in Text-Based Games	Interactive Question Answering (IQA) requires an intelligent agent to interact with a dynamic environment in order to gather information necessary to answer a question. IQA tasks have been proposed as means of training systems to develop language or visual comprehension abilities. To this end, the Question Answering with Interactive Text (QAit) task was created to produce and benchmark interactive agents capable of seeking information and answering questions in unseen environments. While prior work has exclusively focused on IQA as a reinforcement learning problem, such methods suffer from low sample efficiency and poor accuracy in zero-shot evaluation. In this paper, we propose the use of the recently proposed Decision Transformer architecture to provide improvements upon prior baselines. By utilising a causally masked GPT-2 Transformer for command generation and a BERT model for question answer prediction, we show that the Decision Transformer achieves performance greater than or equal to current state-of-the-art RL baselines on the QAit task in a sample efficient manner. In addition, these results are achievable by training on sub-optimal random trajectories, therefore not requiring the use of online agents to gather data.	['Jan Buys', 'Jonathan Shock', 'Edan Toledo', 'Gregory Furman']	null	null	null	null	naacl-wordplay-2022-7	['text-based-games']	['playing-games']	[ 4.49264824e-01 4.99501199e-01 4.86963093e-01 -3.82888049e-01 -1.10554528e+00 -7.91705251e-01 1.06503201e+00 1.41859353e-01 -3.32797676e-01 4.55522060e-01 1.55401275e-01 -8.98929298e-01 -4.69895899e-02 -7.96310842e-01 -6.95590436e-01 -2.41180003e-01 -1.37340084e-01 1.10560513e+00 5.17329931e-01 -5.12355268e-01 2.14950904e-01 -6.55822828e-02 -1.74988747e+00 5.62013566e-01 1.01132858e+00 8.49707603e-01 4.66478258e-01 1.28890324e+00 -1.32892564e-01 1.58509147e+00 -8.03358436e-01 -3.45115483e-01 -5.77620305e-02 -1.05155337e+00 -1.42027092e+00 -2.14975625e-01 3.56938243e-01 -8.60027790e-01 -2.00311691e-01 3.11794758e-01 5.07163107e-01 5.72433949e-01 7.23789752e-01 -1.25283802e+00 -6.32438421e-01 4.17603225e-01 2.56310463e-01 2.57489502e-01 8.84593368e-01 6.96808279e-01 1.30331683e+00 -6.60831273e-01 6.89789116e-01 1.36627018e+00 6.98885247e-02 8.54480267e-01 -1.31348765e+00 -1.16333619e-01 2.96402872e-01 5.94205916e-01 -6.68806016e-01 -6.22536063e-01 4.69903231e-01 -2.03889266e-01 1.48442364e+00 2.87386209e-01 5.47019422e-01 1.32306015e+00 -5.28487936e-03 1.39442658e+00 1.11136830e+00 -5.76255858e-01 4.57828403e-01 2.42151283e-02 1.18561856e-01 9.15047407e-01 -5.74011981e-01 2.05767035e-01 -7.26633728e-01 2.43714191e-02 3.46535712e-01 -7.05000639e-01 -2.05131367e-01 -3.07686388e-01 -1.16262126e+00 9.30227816e-01 4.00507808e-01 5.38669601e-02 -4.50303614e-01 3.48588347e-01 2.13744670e-01 7.33553827e-01 2.08251655e-01 8.60189617e-01 -3.21061999e-01 -5.44153273e-01 -4.52063769e-01 6.14836097e-01 1.19802892e+00 8.67223322e-01 3.67537171e-01 1.46143749e-01 -7.39638984e-01 5.11722147e-01 3.70355368e-01 4.07098144e-01 2.53791213e-01 -1.26859295e+00 6.33953750e-01 6.23170078e-01 4.24741447e-01 -5.40551782e-01 -4.28235382e-01 2.67892759e-02 -1.55314505e-01 3.90552133e-01 8.73410940e-01 -1.70101166e-01 -9.27726507e-01 1.73800564e+00 3.03698391e-01 -1.14441067e-01 4.10125315e-01 7.49594569e-01 7.61800349e-01 1.01783860e+00 2.28797868e-01 5.26361503e-02 1.44633043e+00 -1.10294688e+00 -6.44521892e-01 -2.72867471e-01 6.05627298e-01 -2.60357678e-01 1.72643912e+00 4.15788710e-01 -1.29267335e+00 -4.61026490e-01 -8.13229620e-01 -3.51513475e-01 -2.30660960e-01 -3.36080432e-01 3.72640640e-01 3.87237102e-01 -1.21027040e+00 2.65769064e-01 -7.47244418e-01 -4.46543515e-01 3.32263619e-01 4.54527549e-02 2.39354312e-01 -1.10926941e-01 -1.15411568e+00 1.10883808e+00 2.14357406e-01 -2.03525633e-01 -1.53254580e+00 -6.02027297e-01 -8.00956666e-01 -2.83177687e-05 6.12564266e-01 -1.07408476e+00 2.14257073e+00 -8.12915385e-01 -2.11447573e+00 3.86423826e-01 -2.43340030e-01 -8.02011847e-01 5.47831416e-01 -5.18742800e-01 8.69281664e-02 4.69630331e-01 -1.55287579e-01 9.36425388e-01 7.91163266e-01 -1.13155794e+00 -6.60220385e-01 -2.65524656e-01 6.97359800e-01 5.65463960e-01 3.01021844e-01 2.84460448e-02 -1.34882540e-01 -1.76691830e-01 -5.16191959e-01 -7.18360484e-01 -4.71268110e-02 -3.62582266e-01 -1.43862978e-01 -6.87402487e-01 8.56517792e-01 -9.25686955e-01 8.36504042e-01 -1.53242171e+00 2.44709432e-01 -2.83137321e-01 1.48894489e-01 2.88174033e-01 -3.77896875e-01 7.61254132e-01 4.14540499e-01 -2.20874678e-02 -2.75557846e-01 -3.34256500e-01 3.07109773e-01 1.99128747e-01 -7.03264594e-01 -9.27466974e-02 5.20614505e-01 1.29904747e+00 -1.11075509e+00 -3.78058404e-01 2.17021361e-01 6.78093880e-02 -6.84821904e-01 9.25786316e-01 -1.30585802e+00 6.22554541e-01 -4.93003696e-01 2.99077362e-01 -2.89380223e-01 -4.52796847e-01 -3.83294523e-02 4.53428477e-01 6.57855272e-02 6.55938148e-01 -5.92993021e-01 1.89115620e+00 -6.36800528e-01 7.90923238e-01 -6.84581771e-02 -7.57932782e-01 7.05843091e-01 4.20001596e-01 -8.23493451e-02 -1.14412928e+00 1.69421390e-01 -6.89128041e-02 2.24900439e-01 -8.32842708e-01 5.55557132e-01 7.40242973e-02 2.93122623e-02 7.99800456e-01 1.98642269e-01 -4.93804455e-01 3.40255529e-01 5.72123349e-01 1.33811820e+00 5.30492842e-01 6.16695769e-02 -6.49043173e-02 3.50305647e-01 2.96517819e-01 -2.39654705e-01 1.16424179e+00 -2.06748694e-01 4.05855179e-01 5.17273843e-01 -2.02642873e-01 -1.07960522e+00 -9.70467687e-01 4.29524541e-01 1.53607953e+00 1.06439823e-02 -1.39083505e-01 -9.56096590e-01 -7.29490101e-01 -4.38620538e-01 1.54139280e+00 -4.97224599e-01 -1.13536589e-01 -6.04102492e-01 -2.10694551e-01 5.76573193e-01 4.25425023e-01 7.46788621e-01 -1.51994383e+00 -1.34689891e+00 4.31686610e-01 -4.20457661e-01 -9.33086395e-01 -1.12667494e-01 1.76562797e-02 -4.60930854e-01 -1.16179311e+00 -7.05915809e-01 -6.86700284e-01 2.15758041e-01 -5.43975495e-02 1.59698677e+00 -3.94109190e-02 4.23818408e-03 9.69520152e-01 -6.35172546e-01 -4.65613395e-01 -6.39434934e-01 2.46566772e-01 -4.35612589e-01 -2.81867087e-01 1.02114797e-01 -3.72944832e-01 -7.59778798e-01 3.11825275e-01 -7.84907103e-01 5.00133276e-01 3.83378029e-01 9.37495530e-01 1.22017751e-03 -5.02471864e-01 8.68146896e-01 -7.57758558e-01 1.15389776e+00 -5.16919255e-01 -6.94825590e-01 5.04443824e-01 -5.57885766e-01 3.67980719e-01 7.11268663e-01 -3.20607871e-01 -1.48488224e+00 -2.52181411e-01 -3.06903034e-01 9.89826471e-02 -3.81544530e-01 3.45570743e-01 5.02886400e-02 3.50766808e-01 9.23530936e-01 4.36906606e-01 -2.85661705e-02 -4.38236482e-02 7.76835740e-01 5.32931387e-01 5.09035349e-01 -6.20270252e-01 5.17969370e-01 5.82613721e-02 -3.13887149e-01 -6.43003762e-01 -8.16962659e-01 -3.10401350e-01 -3.29862744e-01 -4.60845560e-01 9.58751738e-01 -4.91056055e-01 -1.09022212e+00 3.53445292e-01 -1.21879542e+00 -1.23318934e+00 -3.09091479e-01 -1.90481618e-02 -1.16234946e+00 -3.72701650e-03 -5.24005234e-01 -1.26506674e+00 -4.81208086e-01 -1.08803415e+00 9.48459268e-01 2.57847250e-01 -3.64301562e-01 -8.32868218e-01 2.69683123e-01 6.68727815e-01 5.73955059e-01 1.03976410e-02 1.18171501e+00 -9.75500941e-01 -9.86198306e-01 3.21607023e-01 -1.56886160e-01 -8.08381811e-02 -4.25251156e-01 -4.04251397e-01 -1.13082182e+00 -1.26948968e-01 -1.35915680e-02 -1.05190361e+00 5.59973478e-01 -9.29422900e-02 8.52002382e-01 -5.06925821e-01 1.44003153e-01 5.65709472e-02 7.92024612e-01 5.74554324e-01 5.37513733e-01 2.58503586e-01 2.63923913e-01 9.12775338e-01 5.74180782e-01 3.13068181e-01 9.10467982e-01 5.67485809e-01 5.19214511e-01 3.29541564e-01 -1.86235249e-01 -6.17932260e-01 3.97772014e-01 4.61309612e-01 2.63010323e-01 -7.01202571e-01 -1.20742583e+00 8.52842093e-01 -1.95633411e+00 -9.68058825e-01 -6.62427619e-02 1.92429066e+00 9.87144113e-01 -1.21388458e-01 2.12229565e-01 -1.00143671e-01 -1.73270795e-02 2.34736949e-01 -9.58189130e-01 -4.84511852e-01 2.40605757e-01 2.58512139e-01 -2.77057081e-01 7.98401535e-01 -7.37161040e-01 1.12114775e+00 6.13914490e+00 4.15317506e-01 -3.96670729e-01 -5.93136670e-03 5.86482763e-01 1.45534649e-01 -2.67496705e-01 2.34491602e-02 -3.99907559e-01 8.18863790e-03 1.39206302e+00 -6.27070367e-02 7.72070169e-01 5.83749115e-01 1.37246355e-01 -4.70746785e-01 -1.39913285e+00 6.01266086e-01 1.67428285e-01 -9.86906350e-01 1.03777118e-01 -4.34091389e-01 3.99205446e-01 -1.15480073e-01 2.13683918e-01 8.57244849e-01 9.99293864e-01 -1.14844167e+00 7.46520460e-01 7.15773046e-01 4.32908595e-01 -5.90907276e-01 3.62826884e-01 7.36749887e-01 -7.16504574e-01 -2.16796860e-01 9.35379565e-02 -3.22184235e-01 3.29728395e-01 -3.89363140e-01 -1.44934702e+00 3.58138859e-01 4.78080213e-01 1.01192795e-01 -6.90326631e-01 8.84129524e-01 -5.96281886e-01 9.98885512e-01 -3.00800353e-01 -4.96322960e-01 4.14901972e-01 1.26290381e-01 5.32420695e-01 7.49682486e-01 -1.88746606e-03 5.84703684e-01 1.06383957e-01 9.64845181e-01 9.29561735e-04 8.16653073e-02 -6.25217259e-01 -1.34074390e-01 2.65860289e-01 7.03275561e-01 -2.95024484e-01 -4.80215073e-01 -3.36574167e-01 1.12950015e+00 4.65606838e-01 6.39107883e-01 -6.37977898e-01 -5.08456767e-01 2.05727607e-01 -3.92777026e-02 2.08356634e-01 -3.33646208e-01 2.82984376e-02 -9.80111897e-01 -1.53213039e-01 -1.31536400e+00 4.77443248e-01 -1.11729932e+00 -9.24308836e-01 6.28212929e-01 6.18566573e-02 -5.29132128e-01 -1.12059546e+00 -5.60753167e-01 -5.70225179e-01 7.96326041e-01 -1.47826958e+00 -1.17537785e+00 -3.47811967e-01 6.66077733e-01 1.01608002e+00 -1.37190133e-01 9.37348306e-01 -3.55450779e-01 -3.13920140e-01 2.52894431e-01 -1.79487363e-01 -1.06911592e-01 4.45854992e-01 -1.65062892e+00 7.96802342e-01 7.95333743e-01 3.44694823e-01 2.14550510e-01 9.21823502e-01 -4.42215651e-01 -1.60555506e+00 -7.45882690e-01 7.04314768e-01 -1.01722872e+00 5.77451825e-01 -4.86029536e-01 -9.93824422e-01 8.09699118e-01 7.90987372e-01 -6.45503998e-01 4.79630589e-01 6.15481064e-02 -2.50796407e-01 2.50873178e-01 -8.60867620e-01 8.79757345e-01 7.94267178e-01 -6.68141544e-01 -1.06002176e+00 3.35854828e-01 1.02045488e+00 -3.82412046e-01 -5.12334704e-01 8.61702487e-02 2.10369930e-01 -1.00557792e+00 8.00230265e-01 -8.41266572e-01 4.61847454e-01 -2.41725206e-01 -4.06529643e-02 -1.22676194e+00 1.23047873e-01 -9.15412366e-01 -3.80449563e-01 9.58819866e-01 5.80011666e-01 -4.80541378e-01 5.58894098e-01 9.07609165e-01 -1.77799985e-01 -5.24344862e-01 -7.98591554e-01 -4.02572304e-01 1.63541064e-01 -5.91342866e-01 4.53283906e-01 3.45416754e-01 1.81998268e-01 9.79786396e-01 -2.03676462e-01 -4.89210263e-02 4.18043077e-01 1.02847256e-01 1.02613676e+00 -8.97439241e-01 -5.54543257e-01 -3.98569733e-01 3.43039215e-01 -1.54670525e+00 1.40292928e-01 -6.53383017e-01 5.10948181e-01 -1.84439540e+00 -1.18100181e-01 -1.09521806e-01 2.46356085e-01 3.13028425e-01 -1.88898221e-01 -2.82986164e-01 3.98915470e-01 1.49747059e-01 -1.08520973e+00 9.08949137e-01 1.23144937e+00 1.86034525e-03 -3.27447176e-01 4.54732180e-02 -4.77877855e-01 5.79747260e-01 8.76988530e-01 5.31707332e-02 -8.83591175e-01 -6.73945308e-01 4.47985232e-01 4.71866786e-01 5.32470584e-01 -1.00685835e+00 4.05014783e-01 6.40164465e-02 -4.06881459e-02 -6.30048513e-01 5.32135844e-01 -3.98801148e-01 -4.80381608e-01 3.51120800e-01 -8.79565835e-01 1.65355772e-01 2.61296332e-01 6.57968879e-01 -8.87650065e-03 -4.00324255e-01 4.06267643e-01 -1.90461308e-01 -7.89846122e-01 -9.16052237e-02 -9.04377878e-01 4.74846929e-01 9.39967990e-01 9.84700695e-02 -5.76613367e-01 -1.01034617e+00 -4.67386007e-01 7.69189835e-01 1.85861811e-01 3.47441196e-01 7.60284543e-01 -8.40203583e-01 -7.73501933e-01 -2.73042321e-01 2.27692395e-01 1.43721521e-01 1.63912639e-01 3.46412122e-01 -4.66416895e-01 7.20662177e-01 -6.91689900e-04 -4.10739034e-01 -1.00325048e+00 5.46101928e-01 5.21387935e-01 -5.15047550e-01 -5.37906110e-01 9.08845603e-01 2.58225977e-01 -5.95089138e-01 4.14432645e-01 -3.50086063e-01 -4.82324481e-01 -8.13818257e-03 7.34075904e-01 2.59827942e-01 -1.92343876e-01 -3.51499990e-02 9.95203406e-02 -4.63476405e-02 -1.26208544e-01 -8.18874300e-01 9.53997672e-01 -1.38453782e-01 5.31606376e-01 6.08188748e-01 6.28429294e-01 -5.63686132e-01 -1.47717309e+00 -2.48246059e-01 1.58239648e-01 -1.77738950e-01 -1.61666334e-01 -1.44607913e+00 -1.28938571e-01 1.13549626e+00 3.03362489e-01 5.33186316e-01 8.93420875e-01 1.56657264e-01 6.98787153e-01 9.53646421e-01 2.89349288e-01 -9.03200030e-01 6.80816531e-01 9.09925282e-01 1.12756419e+00 -1.32530713e+00 -6.68594480e-01 2.35914648e-01 -9.05256510e-01 8.28257024e-01 8.14500332e-01 8.80926549e-02 -1.82668328e-01 -2.43794501e-01 2.31596772e-02 -2.96698868e-01 -1.38522792e+00 -5.55047691e-01 2.42715627e-01 9.23214495e-01 2.31331721e-01 -2.42373183e-01 2.53648072e-01 2.62715638e-01 -3.56847525e-01 -8.92566442e-02 3.69078696e-01 1.00427818e+00 -5.81640482e-01 -9.08856094e-01 -7.81918541e-02 4.23520148e-01 -2.00341731e-01 -2.56644547e-01 -6.12190366e-01 6.81459963e-01 -6.91917300e-01 1.43784046e+00 -1.93875488e-02 -1.35804862e-01 4.80242670e-01 5.11854291e-01 5.20614028e-01 -4.82326120e-01 -7.33505070e-01 -5.82673609e-01 5.83263338e-01 -7.00530529e-01 -1.13262735e-01 -7.22580791e-01 -1.11261570e+00 1.07908875e-01 -1.06333688e-01 1.33441418e-01 3.45662028e-01 9.94298041e-01 5.07717729e-01 6.48247123e-01 2.18518749e-01 -5.43434441e-01 -7.52911389e-01 -1.18811309e+00 2.48314902e-01 2.75533438e-01 3.19213808e-01 -3.41509610e-01 1.47526488e-02 7.21623078e-02]	[11.47983169555664, 7.899476051330566]
56d0163c-b7dd-413e-8f9c-bf42ad8385ce	from-saliency-to-dino-saliency-guided-vision	2304.03140	null	https://arxiv.org/abs/2304.03140v1	https://arxiv.org/pdf/2304.03140v1.pdf	From Saliency to DINO: Saliency-guided Vision Transformer for Few-shot Keypoint Detection	Unlike current deep keypoint detectors that are trained to recognize limited number of body parts, few-shot keypoint detection (FSKD) attempts to localize any keypoints, including novel or base keypoints, depending on the reference samples. FSKD requires the semantically meaningful relations for keypoint similarity learning to overcome the ubiquitous noise and ambiguous local patterns. One rescue comes with vision transformer (ViT) as it captures long-range relations well. However, ViT may model irrelevant features outside of the region of interest due to the global attention matrix, thus degrading similarity learning between support and query features. In this paper, we present a novel saliency-guided vision transformer, dubbed SalViT, for few-shot keypoint detection. Our SalViT enjoys a uniquely designed masked self-attention and a morphology learner, where the former introduces saliency map as a soft mask to constrain the self-attention on foregrounds, while the latter leverages the so-called power normalization to adjust morphology of saliency map, realizing ``dynamically changing receptive field''. Moreover, as salinecy detectors add computations, we show that attentive masks of DINO transformer can replace saliency. On top of SalViT, we also investigate i) transductive FSKD that enhances keypoint representations with unlabelled data and ii) FSKD under occlusions. We show that our model performs well on five public datasets and achieves ~10% PCK higher than the normally trained model under severe occlusions.	['Piotr Koniusz', 'Hao Zhu', 'Changsheng Lu']	2023-04-06	null	null	null	null	['keypoint-detection']	['computer-vision']	[ 2.90891498e-01 2.06158787e-01 -2.98934847e-01 -1.25640899e-01 -7.02431560e-01 -3.14929664e-01 5.93985498e-01 1.47387415e-01 -3.38395029e-01 2.67191768e-01 3.31426084e-01 3.25372398e-01 -6.57505989e-02 -5.36078453e-01 -1.00989389e+00 -5.17115533e-01 6.09517805e-02 7.13958368e-02 9.72894430e-01 -3.91851276e-01 1.24422975e-01 3.96062136e-01 -1.87377667e+00 2.05002084e-01 8.17066908e-01 1.11391592e+00 2.46611729e-01 6.78481638e-01 7.22335204e-02 7.97025561e-01 -5.22114336e-01 -4.14259166e-01 3.69033605e-01 -3.08126032e-01 -5.21368265e-01 -2.61079192e-01 1.03855038e+00 -1.94972873e-01 -6.53938591e-01 1.01301742e+00 6.99236512e-01 1.75580174e-01 4.90303874e-01 -1.51828015e+00 -8.58457327e-01 1.71002790e-01 -6.12189114e-01 7.76773095e-01 3.48869473e-01 3.56706351e-01 1.11293828e+00 -1.24933338e+00 6.08068824e-01 1.18061101e+00 8.05900037e-01 5.65502465e-01 -1.02544308e+00 -4.62375730e-01 2.25823760e-01 3.64575356e-01 -1.36531329e+00 -5.63949823e-01 7.92493343e-01 -8.50002747e-03 9.02295470e-01 4.09451813e-01 8.26996982e-01 1.00534499e+00 2.66607910e-01 1.15863836e+00 8.04911613e-01 -3.26663256e-01 2.69560486e-01 -6.50789589e-02 1.34411529e-01 8.53550971e-01 1.47489294e-01 1.26875132e-01 -1.02667642e+00 -6.83798566e-02 8.92399371e-01 2.74097741e-01 -1.89097717e-01 -6.81426466e-01 -1.34107530e+00 5.44988632e-01 9.27466810e-01 1.85480285e-02 -2.80170768e-01 3.32837701e-01 1.03515558e-01 1.06690384e-01 2.62284577e-01 5.71647525e-01 -4.67392474e-01 2.84959227e-01 -9.36896324e-01 2.65206188e-01 3.43692690e-01 8.94185662e-01 9.10934269e-01 -1.78486630e-01 -6.02043152e-01 5.76239228e-01 1.38247982e-01 7.95550644e-01 5.51648140e-01 -6.59539521e-01 1.51564419e-01 8.39995086e-01 -2.16169357e-01 -1.12097538e+00 -4.58082259e-01 -4.96913433e-01 -5.78289032e-01 6.64428771e-02 3.29164982e-01 2.17798337e-01 -1.50778711e+00 1.69414747e+00 5.96260786e-01 4.46435571e-01 -1.69569343e-01 1.12557423e+00 1.19839060e+00 2.28433818e-01 9.99233648e-02 2.51568109e-01 1.42912400e+00 -9.29285288e-01 -4.10388201e-01 -6.89752221e-01 2.40711078e-01 -4.88314301e-01 1.29919887e+00 -4.52894382e-02 -1.15367341e+00 -5.36530375e-01 -1.08511186e+00 -4.12262440e-01 -4.92273986e-01 -5.75167537e-02 9.18079376e-01 3.02320540e-01 -1.11854422e+00 3.00226897e-01 -7.61925459e-01 -6.49022460e-01 6.87133670e-01 4.02396262e-01 -4.11022872e-01 4.39641550e-02 -1.32014775e+00 8.73428166e-01 -4.88249175e-02 -1.21561192e-01 -9.33305919e-01 -1.09662867e+00 -1.05401158e+00 5.92001826e-02 5.93947113e-01 -9.23518598e-01 9.74150777e-01 -7.76820600e-01 -1.18522191e+00 1.08814609e+00 -1.76534668e-01 -5.01761436e-01 3.39696914e-01 -5.08887231e-01 -1.67900831e-01 4.56591368e-01 3.88695270e-01 1.12099755e+00 1.34047294e+00 -8.99011791e-01 -7.01671898e-01 -3.64647478e-01 5.54120094e-02 4.20330316e-01 -1.29236639e-01 -9.30561349e-02 -8.86134267e-01 -8.59281719e-01 3.37475866e-01 -6.63976669e-01 -1.18386410e-01 3.65070254e-01 -5.33988833e-01 -1.73377231e-01 7.23539293e-01 -2.58844227e-01 9.50907946e-01 -2.29637003e+00 9.03206021e-02 1.03294447e-01 4.39174801e-01 1.56718507e-01 -8.82088393e-02 2.98390482e-02 1.35574387e-02 -2.97012001e-01 -1.64499700e-01 -1.54585987e-01 -6.09776787e-02 1.74351513e-01 -4.88368988e-01 4.86161828e-01 6.50240779e-01 1.53275216e+00 -9.82743740e-01 -5.93496799e-01 3.69682968e-01 3.42820466e-01 -5.41088402e-01 -1.20029323e-01 -7.76818991e-02 -1.81220293e-01 -4.54145789e-01 1.07077110e+00 5.51409483e-01 -4.35588688e-01 -4.96848077e-01 -4.24826652e-01 1.27862498e-01 1.75393268e-03 -1.12504387e+00 1.95327306e+00 1.92989811e-01 4.72023636e-01 -4.26050276e-02 -7.86463439e-01 7.26318896e-01 -1.80392802e-01 3.32433730e-01 -9.59272027e-01 3.78051400e-02 -5.13982326e-02 -2.79737532e-01 -1.59000307e-01 5.41802406e-01 4.58103120e-02 -1.04653731e-01 5.85260019e-02 3.89559239e-01 -5.44391107e-03 -1.68214634e-01 5.84474444e-01 1.41707349e+00 8.55166167e-02 2.74912715e-01 -4.03138429e-01 2.44334713e-01 2.84304880e-02 6.07829154e-01 9.82914507e-01 -5.35887361e-01 9.84480262e-01 3.24501693e-01 -3.95260721e-01 -5.54456234e-01 -1.55888021e+00 3.01574022e-02 1.51493263e+00 7.66974866e-01 -2.97448814e-01 -5.61294615e-01 -8.98869574e-01 3.00102651e-01 1.04132049e-01 -8.87466133e-01 -5.76985359e-01 -4.53666389e-01 -5.79411268e-01 6.01695061e-01 6.46834016e-01 4.59155113e-01 -1.05501616e+00 -9.76269305e-01 -1.13678545e-01 1.36190489e-01 -8.87460947e-01 -7.06336379e-01 4.67326909e-01 -5.01993120e-01 -1.21951044e+00 -9.03731167e-01 -7.58464634e-01 6.85588479e-01 5.89853227e-01 1.00953555e+00 1.48839697e-01 -6.92858100e-01 6.98227167e-01 -2.49526873e-01 -5.34564793e-01 3.40822726e-01 8.27097669e-02 1.28802568e-01 -1.50787592e-01 4.47319955e-01 -4.17458057e-01 -9.13367927e-01 3.49380821e-01 -7.40846515e-01 -2.34380681e-02 8.55943203e-01 8.09865355e-01 6.90157890e-01 -3.83389801e-01 5.34031510e-01 -4.33662832e-01 1.46091878e-01 -2.49915898e-01 -2.03547835e-01 3.44810486e-01 -2.79342860e-01 1.28533691e-01 1.37896314e-01 -5.87731302e-01 -7.39231527e-01 1.36768907e-01 2.15161934e-01 -8.17386806e-01 9.45870578e-02 -3.10984049e-02 -1.83815926e-01 -3.27925354e-01 8.80360782e-01 3.62578481e-01 -1.47419006e-01 -2.91911572e-01 5.68450093e-01 4.59328592e-02 1.15061641e+00 -3.54922175e-01 1.10182762e+00 8.88488770e-01 -3.31298485e-02 -6.85693383e-01 -1.01097584e+00 -6.30032420e-01 -6.50013149e-01 -6.51014149e-02 7.98756242e-01 -1.03455937e+00 -4.45147842e-01 4.64577019e-01 -7.34085619e-01 -1.52194515e-01 -7.43498445e-01 3.62667024e-01 -3.96462917e-01 2.04302892e-02 -4.36602622e-01 -4.76412445e-01 -5.09534597e-01 -8.13538194e-01 1.46184564e+00 5.31907856e-01 -1.73483282e-01 -7.04197764e-01 -1.36709988e-01 1.77421831e-02 2.24931464e-01 2.30041146e-01 6.09030426e-01 -5.92869759e-01 -7.28113592e-01 -9.84854177e-02 -4.56490070e-01 3.56700793e-02 9.98262838e-02 -2.39602804e-01 -1.21750712e+00 -2.11012602e-01 -2.82747924e-01 -3.91548663e-01 1.25350010e+00 6.56426907e-01 6.62652791e-01 6.63595200e-02 -5.65012038e-01 9.90151763e-01 1.03821588e+00 -2.03023627e-01 5.69058299e-01 2.44369715e-01 9.13999915e-01 4.46332633e-01 7.19439983e-01 1.93513572e-01 4.71991062e-01 4.25408304e-01 3.69943768e-01 -4.81791824e-01 -4.04507756e-01 -5.33870876e-01 2.97117740e-01 1.69626668e-01 2.07795843e-01 2.40511045e-01 -7.29722917e-01 6.77440107e-01 -1.81861222e+00 -8.21428955e-01 2.33192265e-01 1.98172903e+00 7.69407451e-01 3.25567454e-01 1.05845645e-01 3.37683782e-02 7.55290151e-01 2.50249624e-01 -9.02293861e-01 1.04929626e-01 -4.06447083e-01 5.24436951e-01 5.89495599e-01 2.68724799e-01 -1.38842261e+00 1.27747321e+00 6.24034882e+00 7.41997898e-01 -1.16163218e+00 1.00610666e-01 3.89107317e-01 -3.40636432e-01 -1.83632895e-01 -8.07697475e-02 -8.59132469e-01 2.35131234e-01 2.04067156e-01 5.16799279e-03 7.98503682e-02 8.92268300e-01 -2.59679168e-01 -2.99471974e-01 -1.09684420e+00 1.12136769e+00 3.19511831e-01 -1.33872092e+00 4.04398367e-02 -2.40689978e-01 5.64844191e-01 1.93508402e-01 2.76607752e-01 3.10100824e-01 1.67154744e-01 -9.31550920e-01 8.05590272e-01 5.96847415e-01 6.86899424e-01 -5.98657489e-01 2.69748688e-01 1.16809625e-02 -1.23287964e+00 1.10303634e-04 -5.41707218e-01 4.20517521e-03 -2.65675783e-01 4.20650482e-01 -7.69719183e-01 1.92653552e-01 9.05640602e-01 7.75714040e-01 -1.04114866e+00 1.11833000e+00 -3.61338705e-01 3.53965104e-01 -5.43838859e-01 1.21238112e-01 1.40168220e-01 5.22307515e-01 7.30813742e-01 1.01036012e+00 7.12588022e-04 7.40973651e-02 6.78098127e-02 8.33088219e-01 6.17630295e-02 -1.02398217e-01 -3.73244077e-01 3.34187210e-01 5.28077543e-01 1.20598340e+00 -1.02133369e+00 -3.17015439e-01 -2.73664266e-01 1.24792600e+00 1.68814942e-01 3.93290281e-01 -7.84362972e-01 -4.79561061e-01 9.08420444e-01 2.54727066e-01 4.98396575e-01 2.16301709e-01 -2.84626752e-01 -1.14040971e+00 9.46368799e-02 -5.01387954e-01 6.74283564e-01 -8.50729048e-01 -1.22841394e+00 1.37012228e-01 -1.80450492e-02 -1.01245403e+00 1.78411558e-01 -4.51269984e-01 -7.47176945e-01 6.34180069e-01 -1.58353043e+00 -1.51721692e+00 -3.99436265e-01 8.32206607e-01 4.66960490e-01 7.88411945e-02 3.15473467e-01 5.82904890e-02 -3.53874683e-01 8.03182423e-01 -4.02786732e-01 1.04844272e-01 9.21289623e-01 -1.38673866e+00 7.12885618e-01 1.01693475e+00 2.00080231e-01 7.46564627e-01 6.34991705e-01 -7.50707507e-01 -1.63944960e+00 -1.11540818e+00 5.65855801e-01 -7.39882469e-01 5.84183156e-01 -5.45036316e-01 -9.66627717e-01 5.08863389e-01 -2.74176210e-01 7.21957564e-01 3.40363383e-01 -3.29797305e-02 -5.24855256e-01 -2.24275798e-01 -1.01438844e+00 5.89932084e-01 1.45720053e+00 -6.01936996e-01 -1.11170399e+00 9.95316207e-02 9.57893372e-01 -5.20725131e-01 -4.01649177e-01 6.10076249e-01 5.52052319e-01 -9.22036648e-01 1.49522579e+00 -5.40313184e-01 1.37099653e-01 -4.64082241e-01 -1.14064321e-01 -9.93350148e-01 -6.32220030e-01 -5.72059810e-01 -3.84033173e-01 1.00980771e+00 3.59802842e-02 -3.33782107e-01 9.78413343e-01 3.45455855e-01 -3.21807742e-01 -7.09664941e-01 -1.10032392e+00 -6.20787978e-01 -3.51802528e-01 -4.60413635e-01 5.23738861e-01 7.71581411e-01 -5.06068505e-02 3.94242406e-01 -2.37669095e-01 3.54256928e-01 6.20170295e-01 -2.99537890e-02 7.25388348e-01 -1.12867475e+00 -2.38328829e-01 -2.34495521e-01 -8.52105200e-01 -9.47017848e-01 -8.05796832e-02 -8.29002142e-01 9.37327966e-02 -1.32631969e+00 2.22402394e-01 -2.81847298e-01 -6.54296696e-01 6.91667378e-01 -6.72272027e-01 7.16447353e-01 2.06771538e-01 5.94882965e-02 -9.91874695e-01 4.96948332e-01 1.36358190e+00 -3.38913947e-01 -4.53224599e-01 -1.71066776e-01 -9.47119653e-01 8.83458138e-01 3.30022782e-01 -2.80897766e-01 -5.39332807e-01 -2.49944434e-01 2.43798569e-01 -4.49438930e-01 9.89752769e-01 -1.19270635e+00 5.40372014e-01 -8.04315135e-02 7.99773872e-01 -7.12567389e-01 4.19839591e-01 -3.24908912e-01 -3.43532801e-01 3.68468523e-01 -2.16740474e-01 -1.84075281e-01 2.03484684e-01 6.52503788e-01 2.60403380e-02 2.79986799e-01 6.98761404e-01 -8.27347040e-02 -1.26003540e+00 3.93795490e-01 4.27479930e-02 4.31330025e-01 9.88780260e-01 -4.69520867e-01 -4.87602741e-01 -6.09656610e-02 -6.20005012e-01 3.55210662e-01 6.06671214e-01 5.34887910e-01 1.05164146e+00 -1.27038884e+00 -3.80000144e-01 4.65511829e-01 4.73315567e-01 2.18392611e-01 3.46700400e-01 8.90320003e-01 -1.93140432e-01 2.05209166e-01 -1.50549933e-01 -7.85943806e-01 -1.13349712e+00 7.03737736e-01 2.71599293e-01 2.05191180e-01 -1.01507938e+00 1.35856223e+00 7.36654580e-01 -7.97646791e-02 2.38101244e-01 -6.78189874e-01 -1.77861433e-02 1.29065305e-01 6.51081681e-01 2.36944973e-01 6.99820891e-02 -6.42605424e-01 -7.56218851e-01 6.54051065e-01 -1.65857345e-01 1.55294701e-01 1.04393125e+00 -9.87422168e-02 7.60593340e-02 2.52365828e-01 9.22531724e-01 -1.18369013e-01 -1.42676032e+00 -5.05127370e-01 9.08409711e-03 -4.73449051e-01 1.10922657e-01 -7.05311120e-01 -7.99442530e-01 6.18465304e-01 6.95887327e-01 -2.63778776e-01 1.21424508e+00 4.89695877e-01 8.89029622e-01 3.72765571e-01 1.92676842e-01 -9.83275533e-01 3.26126397e-01 3.68047237e-01 6.98398411e-01 -1.31512320e+00 -3.60666625e-02 -4.22615260e-01 -6.79474056e-01 5.59406102e-01 8.20661306e-01 -3.61164272e-01 4.60696667e-01 2.06529289e-01 2.70837769e-02 -5.36218524e-01 -7.34988749e-01 -7.35242903e-01 7.10930288e-01 9.39465880e-01 -2.39833936e-01 -2.93178499e-01 2.06280395e-01 6.44994617e-01 -2.40253687e-01 -1.25403076e-01 -5.26417745e-03 9.41671014e-01 -5.96062720e-01 -3.69290829e-01 -5.42571306e-01 6.10484064e-01 -3.01590592e-01 -2.80891836e-01 -5.02294362e-01 7.52631783e-01 3.47023606e-01 5.79758108e-01 2.56048501e-01 -3.30202252e-01 4.69845325e-01 -7.80404881e-02 5.14275670e-01 -7.26061642e-01 -5.41633785e-01 -7.97432065e-02 -4.50392693e-01 -8.21489096e-01 -2.30679542e-01 -5.28785110e-01 -1.44451571e+00 1.97415769e-01 -2.79397398e-01 -1.42016172e-01 1.88700214e-01 9.25132215e-01 3.93977493e-01 5.73774397e-01 2.57049561e-01 -8.60443056e-01 -2.37091765e-01 -6.06750727e-01 -5.77444553e-01 4.95324373e-01 6.11991346e-01 -9.66359138e-01 -2.36793026e-01 -3.01701669e-02]	[9.64699649810791, -0.4776627719402313]
ad6e746c-efa1-47ce-8dfd-7186fe4075a3	unsupervised-structure-consistent-image-to	2208.11546	null	https://arxiv.org/abs/2208.11546v1	https://arxiv.org/pdf/2208.11546v1.pdf	Unsupervised Structure-Consistent Image-to-Image Translation	The Swapping Autoencoder achieved state-of-the-art performance in deep image manipulation and image-to-image translation. We improve this work by introducing a simple yet effective auxiliary module based on gradient reversal layers. The auxiliary module's loss forces the generator to learn to reconstruct an image with an all-zero texture code, encouraging better disentanglement between the structure and texture information. The proposed attribute-based transfer method enables refined control in style transfer while preserving structural information without using a semantic mask. To manipulate an image, we encode both the geometry of the objects and the general style of the input images into two latent codes with an additional constraint that enforces structure consistency. Moreover, due to the auxiliary loss, training time is significantly reduced. The superiority of the proposed model is demonstrated in complex domains such as satellite images where state-of-the-art are known to fail. Lastly, we show that our model improves the quality metrics for a wide range of datasets while achieving comparable results with multi-modal image generation techniques.	['Charalambos Poullis', 'Shima Shahfar']	2022-08-24	null	null	null	null	['image-manipulation']	['computer-vision']	[ 4.75904942e-01 1.74663976e-01 1.63607910e-01 -2.97067851e-01 -6.07927740e-01 -6.26493573e-01 8.29044223e-01 -3.47786248e-01 -3.70912969e-01 6.13283396e-01 -7.45875910e-02 5.90473972e-03 -3.13803740e-02 -1.00307393e+00 -1.04106987e+00 -1.10492957e+00 2.81881213e-01 3.35693568e-01 -3.25998254e-02 -4.03976470e-01 2.68034905e-01 6.63332820e-01 -1.70489061e+00 1.79943845e-01 9.80765462e-01 1.16190529e+00 5.78877270e-01 4.21116233e-01 3.95079702e-02 6.65827215e-01 -3.92686516e-01 -4.52125758e-01 6.54129744e-01 -4.69070375e-01 -6.27686679e-01 2.41009817e-01 7.91434824e-01 -3.68914664e-01 -3.53021294e-01 9.39886749e-01 3.69421691e-01 7.62838125e-02 7.34736621e-01 -1.01533425e+00 -9.64244962e-01 3.28678228e-02 -4.41913605e-01 -4.42663521e-01 -1.02791354e-01 2.36815661e-01 1.03606427e+00 -7.02132344e-01 7.12751865e-01 9.44157600e-01 4.11306202e-01 3.94947588e-01 -1.67956972e+00 -5.26238739e-01 -1.12979114e-01 1.59362540e-01 -1.29651177e+00 -3.22584689e-01 1.00985157e+00 -5.44576585e-01 5.28132796e-01 2.62771904e-01 5.38359642e-01 9.97791231e-01 1.54811710e-01 3.21151853e-01 1.35041797e+00 -4.59492564e-01 1.24204248e-01 2.59119868e-01 -6.16056144e-01 9.27446246e-01 4.05704156e-02 2.77464271e-01 -4.87547159e-01 2.37404048e-01 1.08541811e+00 -1.32614747e-01 -3.81483644e-01 -7.83235252e-01 -1.28505516e+00 8.82597864e-01 7.29218364e-01 1.89210311e-01 -2.69493520e-01 2.00910773e-02 1.38547299e-02 3.46018881e-01 4.25524443e-01 6.12141430e-01 -3.15207720e-01 2.56539732e-01 -8.48016441e-01 1.90960377e-01 4.49232429e-01 1.06765354e+00 9.86400545e-01 2.46901199e-01 -2.71246135e-01 7.98129976e-01 4.90174480e-02 7.00474381e-01 2.55129158e-01 -1.19060457e+00 5.09548426e-01 4.91363376e-01 1.47337422e-01 -1.14039660e+00 -4.40543182e-02 -5.96577346e-01 -1.18424237e+00 6.04452908e-01 3.17703038e-01 2.81565040e-01 -9.21446681e-01 1.99065042e+00 1.69715017e-01 -1.62530124e-01 -8.64184946e-02 8.77262115e-01 3.09031934e-01 4.89302516e-01 -2.13499144e-01 1.11722678e-01 1.30840027e+00 -9.21698868e-01 -4.55234528e-01 -2.52113968e-01 2.81510472e-01 -7.93356121e-01 1.26568115e+00 1.87913314e-01 -1.12576020e+00 -6.53486848e-01 -1.09536731e+00 -3.66599888e-01 -1.86560839e-01 3.64427328e-01 6.15521193e-01 3.27168167e-01 -1.08112407e+00 6.87985897e-01 -7.51363814e-01 -2.36777246e-01 4.73939091e-01 3.52764487e-01 -6.79256499e-01 1.44932717e-01 -8.44535947e-01 8.80527139e-01 2.53106922e-01 -4.34143543e-02 -8.12875152e-01 -8.81153762e-01 -9.95439470e-01 1.76220268e-01 6.11123219e-02 -9.76114929e-01 8.05975795e-01 -1.01440108e+00 -1.94077098e+00 9.84984636e-01 5.39658554e-02 -8.72436911e-02 7.55850732e-01 4.51082736e-02 1.55106513e-02 3.12957197e-01 1.56377479e-01 1.08288276e+00 1.24943411e+00 -1.39793253e+00 -3.51439893e-01 -2.21027344e-01 2.00525194e-01 2.33798519e-01 -5.62108755e-01 -5.48541784e-01 -2.59690166e-01 -1.01373768e+00 1.04001336e-01 -1.16099358e+00 -1.69951399e-03 5.64218938e-01 -2.31092542e-01 4.16953325e-01 8.15675974e-01 -5.96120596e-01 7.14583337e-01 -2.23656106e+00 7.88680375e-01 1.91267058e-02 -8.57562851e-03 1.67739466e-02 -4.77531761e-01 2.86496788e-01 -6.58214912e-02 -4.51756967e-03 -5.98415256e-01 -5.22249639e-01 -1.26400024e-01 3.84475648e-01 -3.81911516e-01 4.78633642e-01 4.97944862e-01 1.00521243e+00 -5.83435714e-01 -2.63842762e-01 3.19977611e-01 7.36634612e-01 -8.43573451e-01 4.39960271e-01 -2.54698306e-01 9.43463743e-01 -2.38159850e-01 2.65628427e-01 7.91410744e-01 -2.97048956e-01 4.37267497e-02 -4.20751095e-01 -7.64073431e-02 1.43314421e-01 -7.73077786e-01 2.09085107e+00 -8.03493142e-01 5.26015103e-01 1.99587211e-01 -1.12074625e+00 8.03758681e-01 8.10023993e-02 1.90029040e-01 -9.18698847e-01 -1.53052304e-02 6.61279187e-02 -3.62867601e-02 -3.29451829e-01 4.17283475e-01 -2.78987199e-01 1.04809098e-01 3.62593174e-01 2.47074813e-01 -6.06276214e-01 -7.14256987e-02 5.09083346e-02 5.10287583e-01 4.99655962e-01 -6.32293299e-02 -5.41670144e-01 5.80636322e-01 -3.22631717e-01 3.02114069e-01 6.27068281e-01 3.77295017e-01 7.90030003e-01 3.36153060e-01 -3.78329694e-01 -1.40298438e+00 -1.04641891e+00 -7.53730610e-02 8.18803608e-01 1.98857468e-02 3.09514496e-02 -7.58422852e-01 -5.10349095e-01 -1.38684675e-01 6.58456147e-01 -8.12728763e-01 -1.46159559e-01 -5.69380760e-01 -6.18519545e-01 4.46536481e-01 3.32687378e-01 8.26806247e-01 -8.93627048e-01 -5.77654302e-01 -1.47895971e-02 -3.95155013e-01 -1.23630881e+00 -5.01050234e-01 -5.69692208e-03 -9.48324263e-01 -7.68822432e-01 -8.20115745e-01 -8.26949000e-01 1.05116463e+00 1.66033551e-01 9.01538372e-01 6.43868297e-02 -2.99939305e-01 -2.80484613e-02 -1.36827037e-01 1.09473430e-01 -5.29483199e-01 2.60336399e-01 -2.29988798e-01 3.34868819e-01 -5.26780605e-01 -9.37410474e-01 -8.33504677e-01 4.14733887e-01 -1.28162313e+00 4.99372810e-01 6.40568554e-01 1.16483521e+00 4.83791202e-01 9.71085625e-04 1.18538693e-01 -4.95561898e-01 1.31310546e-03 2.33003441e-02 -7.12987781e-01 4.01458442e-02 -6.47905648e-01 6.12332344e-01 7.27381349e-01 -4.18666333e-01 -1.33979499e+00 1.45401239e-01 -9.05355215e-02 -1.97865546e-01 -1.08187929e-01 8.64384249e-02 -2.40917310e-01 -4.54047441e-01 4.89623666e-01 5.15168190e-01 2.56841183e-01 -5.03261149e-01 4.99228746e-01 2.29785413e-01 6.09352827e-01 -7.96062410e-01 1.04623795e+00 8.97848904e-01 4.07900482e-01 -7.23463655e-01 -7.12793529e-01 1.12634152e-01 -7.77313232e-01 1.59657210e-01 8.34307909e-01 -9.08040583e-01 -6.32653892e-01 7.93732166e-01 -1.11373055e+00 -5.21341503e-01 -2.49878183e-01 2.59631097e-01 -6.97756648e-01 3.26852590e-01 -6.39497280e-01 -2.90702283e-01 -1.87580749e-01 -1.25419903e+00 1.34520078e+00 -8.08403790e-02 2.01728791e-01 -8.49629104e-01 -2.94067085e-01 2.94456780e-01 6.80082798e-01 1.71719357e-01 1.24413490e+00 9.78193432e-02 -9.83258784e-01 1.86334714e-01 -3.05802464e-01 4.77782905e-01 3.46275449e-01 -2.75247425e-01 -9.70332742e-01 -4.27478164e-01 2.13954210e-01 -2.62427747e-01 9.86510873e-01 3.59246768e-02 1.24703610e+00 -4.31975961e-01 1.12662241e-01 9.96989429e-01 1.49493384e+00 -6.31857738e-02 8.20879817e-01 4.67956960e-01 9.80723560e-01 7.77894318e-01 2.07492322e-01 2.71488398e-01 3.59824628e-01 9.94519711e-01 4.80622590e-01 -3.93270373e-01 -2.55646855e-01 -1.93781853e-01 2.21566886e-01 7.78772414e-01 -1.75335899e-01 -1.85017675e-01 -5.34521639e-01 4.29725766e-01 -1.59596336e+00 -8.10361266e-01 1.87064424e-01 2.29128861e+00 9.57214773e-01 -3.02047059e-02 -3.86989892e-01 2.37662997e-02 4.76080984e-01 2.73118645e-01 -4.01154786e-01 -5.62254526e-02 -4.00145471e-01 3.28823149e-01 4.44719493e-01 7.46288359e-01 -9.66306031e-01 9.92695212e-01 5.58916569e+00 8.17392290e-01 -1.52048612e+00 -2.07776446e-02 6.23178720e-01 6.66611046e-02 -5.05441844e-01 -1.69216972e-02 -3.76440883e-01 3.64397734e-01 3.78748059e-01 2.48981208e-01 6.86151385e-01 5.25434196e-01 9.53795165e-02 5.07040471e-02 -1.00035989e+00 7.82344937e-01 2.50206649e-01 -1.37335181e+00 4.65914130e-01 2.37506971e-01 8.71500492e-01 -1.80402219e-01 3.46842498e-01 -9.89996567e-02 -7.53231123e-02 -8.62323701e-01 1.10164380e+00 5.62257230e-01 1.20852292e+00 -6.27192974e-01 3.59142870e-01 5.91150597e-02 -9.25604701e-01 -2.42179632e-02 -2.10657373e-01 -7.05934688e-02 5.49406856e-02 5.62168002e-01 -3.32512110e-01 6.58530593e-01 4.83907372e-01 5.88353217e-01 -6.94294572e-01 4.43236232e-01 -5.16113341e-01 1.88605547e-01 -2.56915927e-01 5.41908860e-01 1.10469915e-01 -5.87690651e-01 5.34908593e-01 8.68746459e-01 4.94865924e-01 -7.43302852e-02 -1.20073721e-01 1.20924926e+00 -1.19402789e-01 -1.09677844e-01 -6.61581457e-01 8.70098546e-02 1.52424484e-01 1.28804100e+00 -5.94867110e-01 -1.76878154e-01 -1.50269374e-01 1.46195579e+00 3.02380353e-01 4.45309848e-01 -7.85471082e-01 -3.62468362e-01 6.70435905e-01 2.41427138e-01 6.75342441e-01 -3.93205822e-01 -4.81111765e-01 -1.31328642e+00 2.51485109e-01 -8.46761882e-01 -2.26472065e-01 -8.03966582e-01 -1.03200185e+00 7.89957941e-01 -7.61292353e-02 -1.27649355e+00 -1.86498642e-01 -7.08603323e-01 -2.78866053e-01 8.68362844e-01 -1.67204976e+00 -1.56347632e+00 -4.94455665e-01 4.97701347e-01 2.57517695e-01 -2.08600372e-01 9.91231501e-01 3.62562805e-01 -2.77853578e-01 6.31050885e-01 7.11214170e-02 1.19894318e-01 7.36343980e-01 -1.10820174e+00 3.71961802e-01 7.58472383e-01 1.82189625e-02 6.46962702e-01 5.16836762e-01 -2.68670708e-01 -1.40744007e+00 -1.14816451e+00 5.99152684e-01 -2.16546476e-01 4.31781262e-01 -6.93891287e-01 -8.08356225e-01 4.68442440e-01 2.69672692e-01 1.75853986e-02 2.28877380e-01 -3.34223986e-01 -6.32986546e-01 -3.31921488e-01 -1.05490625e+00 6.09550476e-01 1.03071868e+00 -7.19645083e-01 -2.78255224e-01 6.02444820e-02 6.90171897e-01 -3.95571917e-01 -7.07486868e-01 4.09064531e-01 7.07412601e-01 -1.02750337e+00 9.63757813e-01 -2.15818897e-01 8.76048446e-01 -4.64517623e-01 -2.77792096e-01 -1.46017230e+00 -5.38527012e-01 -5.11231422e-01 1.43261835e-01 1.22160935e+00 3.43230039e-01 -6.87552571e-01 5.34105062e-01 3.30297172e-01 -2.10330896e-02 -6.42897546e-01 -8.57182980e-01 -7.74983883e-01 1.47996813e-01 7.38209172e-04 5.66587687e-01 9.31101024e-01 -6.01455986e-01 3.77730459e-01 -5.68652034e-01 9.97714847e-02 7.37097681e-01 2.62808591e-01 6.89386308e-01 -1.04306459e+00 -5.65986991e-01 -3.88541520e-01 -3.60052913e-01 -1.14601278e+00 3.50449890e-01 -8.73991191e-01 -4.66284379e-02 -9.94703889e-01 1.50240615e-01 -5.22200465e-01 -2.41582803e-02 5.95856607e-01 1.64572410e-02 6.22877717e-01 3.71529520e-01 1.54608369e-01 -2.86812466e-02 1.00681067e+00 1.51716483e+00 -2.73249060e-01 3.51303741e-02 -4.42154080e-01 -6.81792080e-01 3.56211632e-01 7.40755737e-01 -4.46462601e-01 -4.03965175e-01 -8.78010213e-01 1.73132390e-01 -7.55043849e-02 5.98241866e-01 -8.72276008e-01 -5.50417639e-02 -1.15269393e-01 3.54260236e-01 8.09931830e-02 4.52607363e-01 -8.92040193e-01 2.65649974e-01 3.03530037e-01 -4.63583380e-01 -9.46106669e-03 2.63766259e-01 4.44249898e-01 -1.79534465e-01 1.12885565e-01 1.14194584e+00 3.57822366e-02 -3.24706227e-01 2.42297947e-01 -4.66269106e-02 -1.42048076e-01 7.02841997e-01 -5.26160821e-02 -2.61029422e-01 -3.50090772e-01 -2.68524140e-01 -2.67579317e-01 8.50773871e-01 4.63226318e-01 3.80920917e-01 -1.46942949e+00 -7.40203440e-01 6.41997457e-01 2.07363680e-01 1.50959015e-01 2.82560647e-01 7.36822009e-01 -7.27317214e-01 2.65853822e-01 -6.52492881e-01 -7.84039855e-01 -9.89099026e-01 3.97380888e-01 3.68495286e-01 -1.88384652e-01 -6.75232530e-01 6.02528930e-01 7.25611508e-01 -5.67977369e-01 -7.32730925e-02 -2.32498631e-01 2.19553247e-01 -1.57503918e-01 3.16724032e-01 6.76760674e-02 1.51329905e-01 -7.38217175e-01 -7.54377097e-02 8.62189412e-01 9.81889013e-03 -3.19881409e-01 1.45201206e+00 -9.42789093e-02 -2.45317459e-01 1.64677009e-01 1.28045535e+00 -2.00400148e-02 -1.68654037e+00 -6.95048198e-02 -5.35704136e-01 -6.13927424e-01 2.79369473e-01 -7.04431474e-01 -1.25710630e+00 9.98515666e-01 6.57225847e-01 -2.44418919e-01 1.27551198e+00 -2.18387201e-01 7.21625984e-01 2.88271576e-01 4.65837002e-01 -8.20553958e-01 2.28305414e-01 3.93008143e-01 1.23455274e+00 -1.29342389e+00 -1.62532613e-01 -3.91578317e-01 -5.73306084e-01 8.79331172e-01 4.19543713e-01 -1.32032409e-01 3.97299737e-01 1.21954225e-01 6.64897710e-02 4.45949705e-03 -4.43954110e-01 -4.60440264e-04 4.01034147e-01 5.69728196e-01 2.07181066e-01 -4.55911867e-02 -6.33160323e-02 -1.94809698e-02 -3.45329344e-01 -1.89173758e-01 3.22909474e-01 6.50031269e-01 6.81401864e-02 -1.33433652e+00 -2.09689155e-01 6.24649972e-02 -3.22425306e-01 -3.10774326e-01 -8.19761902e-02 7.51927078e-01 1.17941856e-01 4.97307181e-01 2.81076014e-01 -2.08755419e-01 2.26769477e-01 -1.11922354e-01 8.14433098e-01 -3.57556909e-01 -2.31873930e-01 -4.59976457e-02 -2.96877652e-01 -6.03973985e-01 -4.97908831e-01 -3.70023727e-01 -8.00607026e-01 -2.57274687e-01 -2.13256821e-01 -1.33887619e-01 7.87984908e-01 8.38951349e-01 6.66742384e-01 4.10737336e-01 7.00808406e-01 -1.14793658e+00 -4.62360322e-01 -8.11166286e-01 -5.28244078e-01 7.71857798e-01 4.91536915e-01 -7.98187852e-01 -2.83132613e-01 4.25213933e-01]	[11.489838600158691, -0.6151317954063416]
ecf3fe35-dd70-4ce7-869b-4c6e7902da22	understanding-and-mitigating-overfitting-in	2211.02219	null	https://arxiv.org/abs/2211.02219v3	https://arxiv.org/pdf/2211.02219v3.pdf	Understanding and Mitigating Overfitting in Prompt Tuning for Vision-Language Models	Pretrained vision-language models (VLMs) such as CLIP have shown impressive generalization capability in downstream vision tasks with appropriate text prompts. Instead of designing prompts manually, Context Optimization (CoOp) has been recently proposed to learn continuous prompts using taskspecific training data. Despite the performance improvements on downstream tasks, several studies have reported that CoOp suffers from the overfitting issue in two aspects: (i) the test accuracy on base classes first improves and then worsens during training;(ii) the test accuracy on novel classes keeps decreasing. However, none of the existing studies can understand and mitigate such overfitting problems. In this study, we first explore the cause of overfitting by analyzing the gradient flow. Comparative experiments reveal that CoOp favors generalizable and spurious features in the early and later training stages, respectively, leading to the non-overfitting and overfitting phenomena. Given those observations, we propose Subspace Prompt Tuning (SubPT) to project the gradients in back-propagation onto the low-rank subspace spanned by the early-stage gradient flow eigenvectors during the entire training process and successfully eliminate the overfitting problem. In addition, we equip CoOp with a Novel Feature Learner (NFL) to enhance the generalization ability of the learned prompts onto novel categories beyond the training set, needless of image training data. Extensive experiments on 11 classification datasets demonstrate that SubPT+NFL consistently boost the performance of CoOp and outperform the state-of-the-art CoCoOp approach. Experiments on more challenging vision downstream tasks, including open-vocabulary object detection and zero-shot semantic segmentation, also verify the effectiveness of the proposed method. Codes can be found at https://tinyurl.com/mpe64f89.	['Changsheng Xu', 'WeiMing Dong', 'Lingxi Xie', 'Jiankang Deng', 'Yang Liu', 'Chengcheng Ma']	2022-11-04	null	null	null	null	['open-vocabulary-object-detection']	['computer-vision']	[ 2.28188634e-01 -3.24482620e-01 -1.34271309e-01 -3.89492363e-01 -4.91359055e-01 -5.17410398e-01 5.94086528e-01 -1.17388614e-01 -6.30226493e-01 3.35148901e-01 2.71130770e-01 -2.18235657e-01 -1.07062973e-01 -3.26054633e-01 -6.02872789e-01 -7.54557848e-01 4.42326605e-01 -1.90038085e-02 2.72705436e-01 -6.53513446e-02 2.31650427e-01 1.53387725e-01 -1.50500906e+00 2.09564164e-01 1.28083646e+00 8.55382800e-01 5.07529736e-01 3.47552299e-01 -2.06858352e-01 4.82354015e-01 -2.66280860e-01 -2.79830039e-01 4.38404262e-01 -1.24150760e-01 -4.21449810e-01 1.01151675e-01 7.34316170e-01 -3.49216372e-01 -5.66746950e-01 1.25749731e+00 3.97915274e-01 2.84013003e-01 5.26329696e-01 -1.32019126e+00 -1.10612929e+00 4.27877843e-01 -6.51660442e-01 2.02875406e-01 -5.00395559e-02 5.07711947e-01 9.73475397e-01 -1.40481126e+00 4.19036001e-01 9.85935509e-01 6.11494184e-01 6.99274004e-01 -1.19896770e+00 -8.42730105e-01 6.37781858e-01 3.18449765e-01 -1.36619210e+00 -2.92665154e-01 6.83262467e-01 -5.16314685e-01 7.74908543e-01 1.69817388e-01 6.48061216e-01 1.39377141e+00 -7.68552572e-02 9.79729950e-01 1.20354807e+00 -2.24874407e-01 9.95900631e-02 3.23255062e-01 5.40327728e-01 7.95881093e-01 2.22541481e-01 2.45640159e-01 -6.87672555e-01 1.45364041e-02 7.06102252e-01 1.57635152e-01 -5.19555926e-01 -2.03039780e-01 -1.09340274e+00 6.91926360e-01 5.73485792e-01 1.90266415e-01 -2.55841672e-01 -4.16265093e-02 3.16073686e-01 -1.21443840e-02 3.03155780e-01 3.54557633e-01 -5.28280854e-01 6.55302703e-02 -8.27292323e-01 -1.83660742e-02 4.84720737e-01 1.17684841e+00 6.19498432e-01 1.01503938e-01 -6.64672017e-01 1.07064342e+00 2.62330681e-01 4.70894694e-01 6.00739300e-01 -8.16245317e-01 5.25753498e-01 6.85458124e-01 -2.37147242e-01 -9.06413078e-01 -3.70285928e-01 -8.49947393e-01 -8.37442756e-01 -1.15768820e-01 4.01450306e-01 -2.26535976e-01 -1.24996030e+00 1.90256488e+00 2.30013371e-01 3.63474935e-01 -1.48575474e-03 1.16825283e+00 1.04799819e+00 6.47547066e-01 3.41272265e-01 -2.42031217e-01 1.25247312e+00 -1.28230691e+00 -6.29435182e-01 -4.59278166e-01 3.41090083e-01 -6.81863606e-01 1.58737481e+00 2.54491568e-01 -4.80410635e-01 -7.23994613e-01 -9.42779779e-01 -1.66168466e-01 -1.38868406e-01 4.19643432e-01 8.90581012e-01 3.10765088e-01 -8.64063442e-01 4.16502237e-01 -6.43613458e-01 -4.19768602e-01 5.90680242e-01 9.06648114e-02 -1.58845648e-01 -3.13370496e-01 -9.59724247e-01 4.71918970e-01 2.55960822e-01 2.11153746e-01 -9.38767195e-01 -9.53736901e-01 -5.91086268e-01 8.21758658e-02 7.16036499e-01 -7.00736880e-01 1.09334421e+00 -1.08605695e+00 -1.35498571e+00 4.75827754e-01 -1.63354218e-01 -1.04826063e-01 5.02000332e-01 -5.71319401e-01 -1.41105369e-01 2.67844391e-03 2.71568626e-01 9.14461195e-01 1.06018841e+00 -1.13758743e+00 -6.94928944e-01 -4.24630255e-01 4.83473763e-02 3.03950995e-01 -7.29252160e-01 -3.28292251e-01 -8.87635827e-01 -6.69217765e-01 1.48745209e-01 -1.03945351e+00 -2.81026661e-01 -1.18282579e-01 -4.38652784e-01 -4.01617259e-01 8.33197355e-01 -4.62770790e-01 1.31268728e+00 -2.31285048e+00 3.51331569e-02 -4.20320518e-02 2.25878760e-01 4.56545711e-01 -3.84655714e-01 3.51029411e-02 1.43671921e-02 9.56327468e-03 -1.31583527e-01 -1.31449565e-01 -2.18876049e-01 1.87101558e-01 -4.65488732e-01 2.31706902e-01 3.37669879e-01 1.07397377e+00 -9.72471178e-01 -3.32843781e-01 1.64798260e-01 4.64378089e-01 -6.08365953e-01 1.67743843e-02 -1.86434507e-01 4.35615301e-01 -4.44845438e-01 5.68801939e-01 5.03900170e-01 -4.72880155e-01 -1.44026190e-01 -3.33172977e-01 -2.03604817e-01 -7.75721520e-02 -9.85012531e-01 1.80711520e+00 -1.84601426e-01 6.85491264e-01 -1.90344363e-01 -8.88348818e-01 6.50229812e-01 1.03329748e-01 3.12878430e-01 -6.65893972e-01 3.12842778e-03 4.34184931e-02 1.80956274e-02 -7.45803535e-01 3.10954809e-01 5.84345311e-02 2.50476986e-01 -8.47510621e-02 1.70728415e-01 3.32289189e-01 7.51909688e-02 3.53089839e-01 8.68107855e-01 2.42768630e-01 6.45420775e-02 -7.95346349e-02 5.33118963e-01 1.53981537e-01 7.58222878e-01 9.42441165e-01 -4.45956796e-01 5.57251811e-01 1.88369319e-01 -8.38637650e-02 -7.21785665e-01 -9.63358164e-01 -1.88946798e-01 1.36980069e+00 2.66393214e-01 -4.93603259e-01 -7.04687476e-01 -8.47371519e-01 1.22629497e-02 7.64884472e-01 -5.12395740e-01 -2.40314111e-01 -1.50580779e-01 -8.74355435e-01 3.31809431e-01 7.79537261e-01 7.04318464e-01 -9.34037685e-01 -3.56489241e-01 9.91025046e-02 -2.47532845e-01 -1.27537596e+00 -7.35801518e-01 7.67928660e-02 -9.90549326e-01 -1.00270021e+00 -7.94054687e-01 -8.11175644e-01 9.02225912e-01 7.76446164e-01 5.61920583e-01 6.72524124e-02 -3.55791628e-01 6.26593471e-01 -3.00742447e-01 -3.45201552e-01 1.19311780e-01 -1.54460266e-01 1.77121550e-01 1.68462068e-01 6.01368606e-01 -2.99930841e-01 -7.54019737e-01 1.76653951e-01 -6.82757139e-01 2.31195539e-01 7.13592768e-01 1.13771367e+00 7.51982331e-01 -2.65725881e-01 4.96777743e-01 -7.06036389e-01 6.21222258e-01 -3.65774065e-01 -6.31954730e-01 2.30557933e-01 -9.69031453e-01 2.73176376e-03 6.97602451e-01 -7.34517097e-01 -1.16709912e+00 1.47850484e-01 1.82936981e-01 -8.40181768e-01 -1.30744368e-01 4.26811546e-01 -5.90363098e-03 -3.59633490e-02 7.06396639e-01 3.38461727e-01 -2.15184689e-01 -3.88375282e-01 4.25880700e-01 6.26639545e-01 5.80948114e-01 -4.41698104e-01 7.69108653e-01 4.14230734e-01 -3.14663976e-01 -9.52743828e-01 -1.21175921e+00 -6.35476291e-01 -3.23501080e-01 -3.09719235e-01 8.63789916e-01 -1.14721084e+00 -5.20441175e-01 4.30125833e-01 -9.25243437e-01 -2.71969169e-01 -1.70403630e-01 5.18691003e-01 -2.72383630e-01 3.51701200e-01 -4.33433473e-01 -6.89434886e-01 -4.48168695e-01 -1.11073804e+00 7.28082538e-01 6.99690104e-01 2.84925997e-02 -7.26243973e-01 -2.97077388e-01 5.81674993e-01 2.14897439e-01 -2.73850739e-01 8.43204379e-01 -6.54883564e-01 -6.74891353e-01 6.77214861e-02 -5.04710436e-01 5.57496071e-01 -7.72749260e-02 -7.44542778e-02 -1.17562842e+00 -2.97379524e-01 1.50219658e-02 -2.02317297e-01 1.15136850e+00 5.21140754e-01 1.25162482e+00 -1.32161558e-01 -3.60488236e-01 9.40850377e-01 1.36635590e+00 1.32641584e-01 2.72207499e-01 3.65727901e-01 9.69211102e-01 5.62922716e-01 7.28840232e-01 1.78943843e-01 4.12571013e-01 5.23981631e-01 1.87585473e-01 -2.09500752e-02 -2.96149969e-01 -4.69087929e-01 4.42842901e-01 7.00074494e-01 -8.64513889e-02 1.95828289e-01 -8.98442805e-01 5.02234817e-01 -2.03520489e+00 -6.26884758e-01 -2.99652398e-01 2.00299287e+00 7.38074839e-01 9.00929123e-02 -2.29713246e-01 -3.71161133e-01 5.42233467e-01 -5.41667407e-03 -9.88968134e-01 6.28591552e-02 -8.28735381e-02 -1.23696521e-01 4.25295234e-01 3.48242313e-01 -1.03833020e+00 1.28382480e+00 4.91819000e+00 7.94339538e-01 -1.30703413e+00 3.57267320e-01 4.37395453e-01 -4.05371904e-01 -2.36780927e-01 1.09519981e-01 -9.60491061e-01 3.38841558e-01 3.53414148e-01 -3.66101973e-02 6.03187978e-01 9.57494497e-01 2.30945945e-01 6.22482076e-02 -9.75527823e-01 1.22770667e+00 1.12605326e-01 -1.17212439e+00 1.61555529e-01 -2.64232010e-01 7.99439847e-01 3.37041676e-01 2.38874704e-01 7.39488363e-01 -1.20021746e-01 -7.90192604e-01 6.33272588e-01 4.72514242e-01 7.91162252e-01 -3.66178006e-01 3.56526971e-01 4.69904929e-01 -9.26999867e-01 -4.56625998e-01 -4.90581155e-01 1.92348599e-01 -1.77443221e-01 4.90021616e-01 -7.25899220e-01 2.67963260e-01 8.78757715e-01 7.62058854e-01 -8.25301051e-01 1.04704571e+00 -4.55063909e-01 7.36993134e-01 -1.91190392e-01 -1.89648904e-02 3.92853349e-01 -3.02059263e-01 7.51624405e-01 1.19849133e+00 7.57404938e-02 3.43738467e-01 2.52181590e-01 1.06993687e+00 3.86018418e-02 2.17539102e-01 -4.68388766e-01 -1.36533171e-01 3.52975816e-01 1.40415144e+00 -5.05110025e-01 -2.67042309e-01 -6.86622977e-01 9.99440908e-01 3.58476251e-01 9.24486458e-01 -7.30965137e-01 -2.84697920e-01 6.98296249e-01 -3.48423235e-02 2.88779557e-01 -2.17162833e-01 -5.92228055e-01 -1.38291514e+00 1.40019208e-01 -7.45721877e-01 2.61709630e-01 -6.18094921e-01 -1.29470289e+00 3.81229699e-01 -1.08600512e-01 -1.14558268e+00 2.68300802e-01 -6.25689805e-01 -6.16892576e-01 6.94823742e-01 -1.55424869e+00 -1.07186139e+00 -5.71500540e-01 7.08865821e-01 9.68655109e-01 -3.13020617e-01 4.26467925e-01 1.64880678e-01 -8.66035938e-01 7.35185981e-01 -6.08741455e-02 6.10890463e-02 7.62890100e-01 -9.60024953e-01 2.05770619e-02 1.00114167e+00 2.44204327e-01 9.03767169e-01 5.00182390e-01 -7.53717303e-01 -1.29923868e+00 -1.23400986e+00 4.87433046e-01 -3.37841660e-01 6.26022577e-01 -2.97336251e-01 -1.11406648e+00 5.85444689e-01 -1.44223589e-02 8.16527754e-02 5.87316513e-01 2.63218492e-01 -5.28280854e-01 -2.10568666e-01 -6.06852293e-01 8.98714304e-01 1.39562714e+00 -4.90212888e-01 -5.41557014e-01 4.34502482e-01 7.36387193e-01 -2.80971229e-01 -4.53308314e-01 5.51218271e-01 4.93644446e-01 -8.51446390e-01 8.41014087e-01 -6.81784034e-01 4.59369332e-01 -3.12936932e-01 -1.21874087e-01 -1.12788534e+00 -6.41122937e-01 -3.94980550e-01 -1.57742411e-01 1.30085289e+00 3.98056686e-01 -6.22655749e-01 5.52057147e-01 5.22940218e-01 -2.20193416e-01 -1.03880239e+00 -6.82895362e-01 -7.08427906e-01 -1.29046947e-01 -5.82623959e-01 7.59172589e-02 9.50244904e-01 -2.92222917e-01 6.92912340e-01 -1.26755938e-01 3.18809181e-01 7.26568282e-01 9.50852931e-02 6.10579312e-01 -1.08666325e+00 -2.38496125e-01 -4.24412340e-01 -4.40111645e-02 -1.22586024e+00 1.12845600e-01 -1.11988056e+00 1.23147950e-01 -1.43974400e+00 5.54619014e-01 -4.52863157e-01 -4.46206033e-01 7.40910828e-01 -7.66894877e-01 -1.12708338e-01 5.15018344e-01 4.08540547e-01 -6.58731580e-01 7.20437407e-01 1.41669512e+00 -2.09733218e-01 -3.89117092e-01 -1.98814973e-01 -9.97722268e-01 8.31594050e-01 6.43586755e-01 -3.86999458e-01 -7.17490315e-01 -5.76521933e-01 2.47945301e-02 -3.80834073e-01 5.36039352e-01 -8.30339611e-01 3.23045194e-01 -2.09054440e-01 5.71857750e-01 -3.63960624e-01 1.23369172e-01 -7.49269307e-01 -3.09473515e-01 2.21802399e-01 -3.05887818e-01 -3.39464843e-01 3.20858240e-01 8.80516887e-01 4.49677510e-03 -2.16121063e-01 5.43053329e-01 4.46101613e-02 -1.07792926e+00 4.60876346e-01 -8.35695043e-02 1.09937601e-01 9.55384910e-01 -2.24677250e-01 -4.22340125e-01 -1.46890104e-01 -6.36426806e-01 6.19148076e-01 3.13963473e-01 8.46119344e-01 7.77466655e-01 -1.20284569e+00 -5.69564760e-01 3.54061425e-01 3.31212193e-01 6.30590916e-02 3.89412254e-01 1.14362383e+00 1.04467578e-01 3.89658570e-01 1.34233609e-01 -9.69845355e-01 -1.11018026e+00 5.45282066e-01 2.40935028e-01 3.74681205e-02 -7.55588055e-01 1.11956048e+00 6.58855140e-01 -2.40596220e-01 4.73918825e-01 -2.07263529e-01 -2.99874246e-01 1.36118352e-01 4.27507341e-01 1.56708494e-01 -2.45340481e-01 -2.91104198e-01 -3.37043703e-01 6.63851738e-01 -3.80788207e-01 1.09950311e-01 1.26325190e+00 -1.97616085e-01 2.36768067e-01 3.47133458e-01 9.57196295e-01 -2.02261135e-01 -1.71255505e+00 -4.12428051e-01 -1.48888037e-03 -4.46923137e-01 2.15582535e-01 -8.85004103e-01 -1.07976139e+00 8.48482549e-01 8.24729383e-01 -2.31296122e-01 1.20113540e+00 -9.84322801e-02 7.11649895e-01 3.43154877e-01 2.21300155e-01 -1.30798697e+00 2.88397193e-01 6.30787492e-01 9.90810513e-01 -1.51236105e+00 -2.56407648e-01 -4.04423714e-01 -1.03461218e+00 9.67415214e-01 9.64362860e-01 -2.16526110e-02 5.05926669e-01 -9.37823951e-02 1.39171436e-01 -1.79236159e-01 -7.99109757e-01 -2.42794439e-01 4.67503995e-01 5.43109059e-01 2.18032941e-01 -6.75859675e-02 -4.33394045e-01 9.50773060e-01 -1.99047234e-02 8.22855756e-02 2.72193551e-01 6.51771724e-01 -3.25216353e-01 -6.60629988e-01 -2.41438106e-01 5.44582665e-01 -1.70008317e-01 -3.08495969e-01 -3.56283873e-01 5.87185562e-01 2.36976996e-01 9.97378588e-01 -1.43673703e-01 -4.55783367e-01 5.20354688e-01 1.85795441e-01 3.45055550e-01 -7.06956148e-01 -4.36705768e-01 1.82494774e-01 -2.16782138e-01 -7.63992667e-01 -1.51347265e-01 -6.11591637e-01 -1.37369394e+00 2.71115422e-01 -4.28033531e-01 -1.50374994e-01 5.81606686e-01 9.20551658e-01 4.77083951e-01 7.10483849e-01 3.56651336e-01 -5.84802568e-01 -8.67428660e-01 -9.55065072e-01 -2.71086901e-01 6.00868046e-01 3.68719041e-01 -7.99698412e-01 -3.80293548e-01 -8.09597969e-03]	[10.146772384643555, 1.9059910774230957]
63abc144-ecff-49f0-b1db-5179ff4301bd	jointly-fine-tuning-bert-like-self-supervised-1	null	null	https://arxiv.org/abs/2008.06682	https://arxiv.org/abs/2008.06682	Jointly Fine-Tuning “BERT-like” Self Supervised Models to Improve Multimodal Speech Emotion Recognition	Multimodal emotion recognition from speech is an important area in affective computing. Fusing multiple data modalities and learning representations with limited amounts of labeled data is a challenging task. In this paper, we explore the use of modality-specific "BERT-like" pretrained Self Supervised Learning (SSL) architectures to represent both speech and text modalities for the task of multimodal speech emotion recognition. By conducting experiments on three publicly available datasets (IEMOCAP, CMU-MOSEI, and CMU-MOSI), we show that jointly fine-tuning "BERT-like" SSL architectures achieve state-of-the-art (SOTA) results. We also evaluate two methods of fusing speech and text modalities and show that a simple fusion mechanism can outperform more complex ones when using SSL models that have similar architectural properties to BERT.	['Suranga Nanayakkara', 'Shamane Siriwardhana', 'Rivindu Weerasekera', 'Andrew Reis']	2020-08-15	null	null	null	interspeech-2020-8	['multimodal-emotion-recognition', 'multimodal-emotion-recognition']	['computer-vision', 'speech']	[ 5.79253249e-02 6.91073611e-02 -4.43638442e-03 -8.92959118e-01 -1.02435362e+00 -4.17532712e-01 7.95202255e-01 9.96032357e-03 -5.28410316e-01 5.96917689e-01 4.86938894e-01 7.72934556e-02 2.87987620e-01 -5.25498092e-02 -5.53713083e-01 -4.65333909e-01 1.20276429e-01 2.73707390e-01 -3.21228534e-01 -4.96566772e-01 -3.75687778e-01 3.40606123e-01 -1.75987852e+00 1.06885743e+00 5.09890556e-01 1.52758288e+00 -1.97697446e-01 8.00945520e-01 -4.08536851e-01 1.18428361e+00 -3.50676209e-01 -5.38842678e-01 -3.61932546e-01 -2.32168630e-01 -1.04910803e+00 6.84882924e-02 1.76194534e-02 1.08143166e-01 -3.26722354e-01 6.69585526e-01 7.69622505e-01 4.70278740e-01 7.67805934e-01 -1.60319495e+00 -4.39267516e-01 7.42749333e-01 -3.08043818e-04 -1.72896177e-01 5.29527605e-01 -1.69150323e-01 7.93415904e-01 -1.02255285e+00 4.50056732e-01 1.34138143e+00 5.49758255e-01 1.16557908e+00 -1.06660080e+00 -3.77926290e-01 2.66956002e-01 4.00507450e-01 -1.17387843e+00 -1.07799911e+00 1.00282049e+00 -5.30282632e-02 1.50894260e+00 4.94167179e-01 6.72802180e-02 1.86734438e+00 -1.66960016e-01 1.46916366e+00 1.28824282e+00 -4.49378610e-01 4.81635869e-01 4.32140112e-01 1.22373626e-01 5.97946286e-01 -7.53921986e-01 -2.41839603e-01 -1.09574676e+00 -1.52081728e-01 -9.65365116e-03 -2.41529390e-01 -3.82995419e-02 -1.60446644e-01 -1.06283438e+00 6.97638333e-01 2.04011351e-01 5.54579318e-01 -4.16532487e-01 1.17349133e-01 9.44583595e-01 5.69121063e-01 5.77088416e-01 1.75683245e-01 -6.59928918e-01 -6.52875960e-01 -6.69138014e-01 -6.07341409e-01 9.50355828e-01 7.31063545e-01 6.09658539e-01 3.58833194e-01 1.50577962e-01 1.43417668e+00 2.34752566e-01 3.87281537e-01 7.26674497e-01 -8.72637570e-01 3.29677820e-01 5.95917583e-01 -1.67879671e-01 -3.19433570e-01 -6.54230297e-01 1.86406478e-01 -9.46974814e-01 1.92531559e-03 -2.40880400e-01 -5.75543940e-01 -9.47338700e-01 1.93687510e+00 -1.86012194e-01 -9.68498457e-03 6.95686221e-01 7.95215905e-01 1.48762643e+00 7.58112133e-01 6.38764799e-01 -2.59643435e-01 1.12986147e+00 -1.12394035e+00 -8.95381451e-01 -4.37848121e-01 6.80870712e-01 -3.25739294e-01 9.44810390e-01 3.74726862e-01 -1.04833364e+00 -3.41928065e-01 -8.08546722e-01 -4.85282578e-02 -9.05096412e-01 4.07509118e-01 7.73181856e-01 7.86914408e-01 -1.20981574e+00 3.17670032e-03 -6.95600808e-01 -6.58272803e-01 2.70776331e-01 3.43345255e-01 -8.68555248e-01 3.96471530e-01 -1.23483527e+00 1.03049564e+00 2.93000668e-01 5.53921759e-02 -9.29529190e-01 -7.16878697e-02 -1.18383956e+00 3.26230764e-01 2.07430631e-01 -3.76482427e-01 1.28267074e+00 -1.50640583e+00 -2.00495362e+00 9.83866334e-01 -4.72345203e-01 -3.35081071e-01 -7.12641105e-02 -6.45618290e-02 -1.05984592e+00 3.37298483e-01 -6.05239809e-01 1.04860282e+00 8.17304373e-01 -1.44206381e+00 -2.33129665e-01 -4.38656658e-01 -1.74699932e-01 2.97390789e-01 -8.59088302e-01 2.88084358e-01 -1.60928428e-01 -3.35401416e-01 -2.55900294e-01 -8.15334916e-01 -3.77749503e-02 -4.45321470e-01 -2.49674335e-01 -4.76204276e-01 7.97063589e-01 -3.04858893e-01 1.06109822e+00 -2.34752488e+00 4.21817631e-01 4.38647456e-02 -1.77548781e-01 2.81702399e-01 -6.63142443e-01 6.49178028e-01 -2.69686878e-01 6.87529072e-02 -5.50817363e-02 -1.11937690e+00 4.65308785e-01 3.86881709e-01 -1.32751375e-01 -1.90870032e-01 3.52228761e-01 1.12472165e+00 -4.84731704e-01 -4.18007523e-01 3.86808783e-01 5.27390242e-01 -7.53773525e-02 5.28728545e-01 -1.67809486e-01 2.26268277e-01 -4.08879630e-02 9.01801348e-01 1.70660943e-01 -1.13638170e-01 2.83643663e-01 -4.95475590e-01 2.61058569e-01 4.80007343e-02 -8.41682553e-01 2.07819247e+00 -8.53824317e-01 7.44626164e-01 3.97967696e-01 -9.86433864e-01 9.80996609e-01 9.44361627e-01 5.25662124e-01 -7.62252986e-01 5.92985868e-01 -1.50342565e-02 -5.52392244e-01 -7.42595434e-01 4.03617233e-01 -3.84120315e-01 -6.22653842e-01 3.59828144e-01 1.01506150e+00 3.40627171e-02 -3.75740491e-02 1.10192314e-01 1.06728387e+00 -1.47215232e-01 2.14942172e-01 2.83439219e-01 6.12559259e-01 -4.32672560e-01 2.48088077e-01 5.06726444e-01 -5.34963787e-01 3.32888335e-01 4.04284120e-01 -6.31100312e-02 -5.58046401e-01 -1.02587557e+00 1.31826654e-01 1.70282161e+00 -3.07606488e-01 -5.06304681e-01 -6.52894735e-01 -8.97039056e-01 -3.47642392e-01 7.53708422e-01 -7.37200320e-01 -3.83827835e-01 2.12974578e-01 -5.71256459e-01 9.97124195e-01 7.89768279e-01 3.30363274e-01 -1.25020504e+00 -3.71266186e-01 3.39628980e-02 -4.70761150e-01 -1.47764444e+00 7.83599764e-02 6.49874508e-01 -2.92368352e-01 -4.07900780e-01 -5.75244784e-01 -6.78277612e-01 1.97308794e-01 -1.37145191e-01 1.10698533e+00 -4.22974259e-01 7.60343894e-02 9.69379365e-01 -6.28354669e-01 -5.79325080e-01 -4.94307935e-01 1.28305823e-01 2.31820717e-01 5.44208229e-01 4.79066014e-01 -4.35689777e-01 1.38028944e-02 8.39340985e-02 -1.01878965e+00 3.90547328e-02 3.89005780e-01 8.73793542e-01 6.00968860e-02 -2.93463796e-01 1.08307195e+00 -3.48072678e-01 6.96416736e-01 -5.64400554e-01 2.55808294e-01 4.74446028e-01 5.53108789e-02 4.30244720e-03 4.12671357e-01 -5.63074648e-01 -1.43347037e+00 3.50357026e-01 -5.97164214e-01 -7.26828098e-01 -8.24592710e-01 6.57227218e-01 -4.88508999e-01 -2.44066343e-02 3.98006260e-01 1.29548728e-01 1.29687432e-02 -3.93981755e-01 6.38273299e-01 1.25289321e+00 5.62396228e-01 -5.65607846e-01 -2.12665677e-01 3.14558744e-01 -4.53618437e-01 -8.84437859e-01 -5.90142667e-01 -4.69628185e-01 -3.88317168e-01 -4.60639268e-01 1.02029455e+00 -1.06257975e+00 -9.46519136e-01 4.65268672e-01 -9.56920564e-01 -5.07184923e-01 -2.49471575e-01 4.75231320e-01 -8.38487446e-01 1.53407022e-01 -9.14964855e-01 -1.41266215e+00 -3.86399835e-01 -1.02596116e+00 1.35068047e+00 2.90211052e-01 -4.35255438e-01 -1.17404354e+00 7.48146102e-02 5.74172676e-01 4.26492840e-01 -1.33554354e-01 6.23607755e-01 -1.12139332e+00 3.69824678e-01 -1.31460309e-01 -1.12448521e-01 6.42465591e-01 -2.74806526e-02 -1.19944133e-01 -1.63201976e+00 8.82223919e-02 -9.86724645e-02 -1.21442544e+00 9.65572417e-01 -8.12105313e-02 9.97174084e-01 -1.08418867e-01 -4.51614149e-03 2.06924170e-01 9.77616251e-01 1.86435744e-01 4.72073108e-01 -5.25394343e-02 5.51061273e-01 8.41140628e-01 3.18304569e-01 5.94392538e-01 7.24792838e-01 3.87011975e-01 3.39575559e-01 -1.25847340e-01 7.82359615e-02 8.73310044e-02 7.66232014e-01 1.09821165e+00 2.85073489e-01 -4.74607676e-01 -8.40354741e-01 5.62033951e-01 -2.17942762e+00 -9.97446418e-01 4.69119847e-01 1.53787577e+00 8.14306617e-01 -3.76885444e-01 2.12539762e-01 -1.46486256e-02 2.07328618e-01 1.95890501e-01 -4.40902382e-01 -9.46165800e-01 -6.20268404e-01 2.32912093e-01 -2.47055992e-01 3.10504586e-01 -1.35214972e+00 1.02609372e+00 6.57344103e+00 8.12567055e-01 -1.35662413e+00 3.93157452e-01 7.58355439e-01 -3.39424402e-01 -2.07123727e-01 -5.68263233e-01 -2.78103143e-01 3.11103880e-01 1.67322779e+00 2.37222582e-01 5.29565215e-01 7.24154830e-01 -1.69014767e-01 -6.92631081e-02 -1.18717146e+00 1.44344604e+00 4.36638117e-01 -1.10608244e+00 -1.35611564e-01 -5.33174455e-01 6.05397582e-01 3.22733432e-01 -1.33670093e-02 8.19273174e-01 2.46648580e-01 -1.11506629e+00 6.01814628e-01 6.69394791e-01 7.13239253e-01 -6.62382782e-01 8.02210391e-01 2.31343031e-01 -9.98069465e-01 -2.98718840e-01 1.72744110e-01 2.19266191e-01 2.91392654e-01 6.98815435e-02 -4.43181664e-01 6.51418209e-01 8.55919778e-01 5.08606851e-01 -4.98302221e-01 3.05643231e-01 1.59480572e-01 5.85766137e-01 -3.96276772e-01 -1.69118091e-01 1.93592429e-01 2.06845045e-01 1.87804461e-01 1.70075715e+00 1.35342881e-01 -3.26451510e-02 -7.00095221e-02 4.28350300e-01 -3.19564372e-01 2.23163426e-01 -7.00885892e-01 -5.28101444e-01 1.25494793e-01 1.54603684e+00 -1.82306677e-01 -5.85048735e-01 -5.25449216e-01 1.32276416e+00 6.27007306e-01 5.22986412e-01 -6.22425854e-01 4.75669764e-02 6.61803722e-01 -8.99494827e-01 4.66140695e-02 -6.07153848e-02 -2.48144940e-01 -1.42471814e+00 -3.28763723e-01 -9.02136028e-01 6.06567681e-01 -1.32432556e+00 -1.65261948e+00 1.04489136e+00 -2.93934107e-01 -9.49754894e-01 -6.10670865e-01 -7.39243090e-01 -4.56957906e-01 4.96737123e-01 -1.19206703e+00 -1.54413795e+00 -2.03726292e-01 9.31762874e-01 6.05729938e-01 -5.07161200e-01 1.43146658e+00 3.23183477e-01 -6.55406415e-01 7.28575349e-01 -6.61667436e-02 9.88969803e-02 8.52841437e-01 -1.08459163e+00 -3.25401902e-01 2.74228632e-01 3.48678291e-01 1.00626327e-01 4.08573836e-01 -2.37643257e-01 -1.61288166e+00 -7.14296281e-01 8.89668822e-01 -4.80745465e-01 7.10993230e-01 -4.89620656e-01 -8.25006366e-01 7.23056912e-01 8.57254565e-01 -1.90440014e-01 1.19754255e+00 3.32147688e-01 -5.62610507e-01 -1.25559524e-01 -1.21284211e+00 5.90405822e-01 7.07023501e-01 -1.19719815e+00 -5.02195299e-01 6.87883608e-03 5.35626292e-01 9.99623239e-02 -1.04536998e+00 5.93502462e-01 4.96167898e-01 -8.75665128e-01 7.92875588e-01 -9.86224174e-01 3.41961414e-01 2.65021861e-01 -6.90615952e-01 -1.64826083e+00 2.61655927e-01 -5.31771243e-01 -2.66020954e-01 1.44722950e+00 4.26239103e-01 -3.00528795e-01 5.35091877e-01 7.67132044e-01 -2.29740500e-01 -6.41881764e-01 -1.28294492e+00 -4.55992609e-01 -2.29471773e-01 -8.27633321e-01 3.39799672e-01 1.06709003e+00 8.05993140e-01 7.43310630e-01 -4.35502797e-01 -9.07012373e-02 1.81557953e-01 -3.15446645e-01 5.58446109e-01 -9.02655900e-01 1.38051793e-01 -6.77622020e-01 -3.75000417e-01 -5.07651091e-01 9.34931695e-01 -8.96836400e-01 1.00417510e-01 -1.35826921e+00 1.01443909e-01 5.31776138e-02 -7.49558210e-01 9.09252942e-01 1.25943959e-01 2.75076151e-01 2.23289579e-01 -3.28506231e-01 -1.12553561e+00 1.08218217e+00 3.30540985e-01 -2.65643030e-01 -5.57334460e-02 -4.79914039e-01 -3.90841037e-01 8.02087247e-01 7.60288715e-01 1.60554260e-01 -3.36305380e-01 -2.20118225e-01 5.87376878e-02 3.97016317e-01 2.55415916e-01 -9.53680575e-01 3.34874153e-01 5.47605241e-03 2.33320862e-01 -3.92581582e-01 1.29170525e+00 -9.88413095e-01 -8.98925588e-02 -4.38534558e-01 -7.60500729e-01 -1.68962494e-01 6.25047147e-01 2.01498568e-01 -5.26345074e-01 -6.40078187e-02 5.69049895e-01 1.32634953e-01 -1.10310805e+00 -5.66135459e-02 -8.85792017e-01 -4.48959619e-01 7.82509148e-01 3.31464320e-01 -4.29768413e-01 -8.10214400e-01 -1.31467021e+00 2.14079127e-01 3.34054939e-02 8.26999366e-01 8.50802541e-01 -1.52603400e+00 -4.69467402e-01 7.25448728e-02 6.12194598e-01 -9.49275672e-01 6.30993009e-01 8.78515959e-01 3.31148654e-01 4.00883943e-01 -3.65368932e-01 -3.84860307e-01 -1.45529926e+00 5.43391168e-01 4.71838593e-01 -5.97069860e-02 2.57068694e-01 8.53228569e-01 -7.22256154e-02 -9.57955241e-01 5.67880332e-01 1.12318143e-01 -2.20821783e-01 3.71147424e-01 3.80501866e-01 8.73262808e-02 2.19878256e-01 -8.91640782e-01 -6.60273194e-01 -1.78840712e-01 2.48936638e-01 -6.01477325e-01 1.29709387e+00 -5.05190134e-01 -5.95925748e-03 1.07655489e+00 1.35019636e+00 -3.82409304e-01 -7.37995148e-01 -3.86449337e-01 3.34669165e-02 3.35727066e-01 7.16985762e-02 -1.30873322e+00 -9.16051447e-01 1.19211268e+00 8.18787336e-01 1.55779362e-01 1.36819386e+00 3.56805861e-01 5.99040151e-01 7.51903832e-01 1.74805298e-01 -1.55832338e+00 3.24712247e-01 6.29095078e-01 9.89889622e-01 -1.54690516e+00 -8.05123210e-01 -6.57264516e-02 -1.39867532e+00 9.18045402e-01 7.09639668e-01 4.49723989e-01 5.45391917e-01 3.77574652e-01 5.22954941e-01 -1.36106893e-01 -1.34131384e+00 -6.79729164e-01 4.18651074e-01 4.81193811e-01 5.82157850e-01 2.22191066e-01 2.09752470e-01 1.20085585e+00 2.18505502e-01 1.34516479e-02 1.89920500e-01 1.09849679e+00 -1.35039255e-01 -1.05776238e+00 -1.82897881e-01 2.56035239e-01 -2.90186971e-01 -1.22192092e-01 -8.68740022e-01 3.72668296e-01 -2.49256298e-01 1.43608558e+00 1.08974524e-01 -8.00344646e-01 2.08006516e-01 8.58045340e-01 4.25343096e-01 -2.51490712e-01 -7.95565963e-01 1.53251261e-01 8.09833825e-01 -6.34646237e-01 -8.00632417e-01 -5.00179589e-01 -1.32750535e+00 9.77955875e-04 7.73439631e-02 5.68101704e-02 9.65764642e-01 1.10840416e+00 6.96240485e-01 5.14984310e-01 6.30208910e-01 -9.63433683e-01 -2.24753305e-01 -1.08837318e+00 -4.64639515e-01 5.99055231e-01 2.48346299e-01 -5.41215062e-01 -1.44912973e-01 2.23111492e-02]	[13.297468185424805, 5.344238758087158]
72566a6d-1a23-4345-8da2-5c24e92ae462	a-deep-representation-for-depth-images-from	1609.09713	null	http://arxiv.org/abs/1609.09713v1	http://arxiv.org/pdf/1609.09713v1.pdf	A deep representation for depth images from synthetic data	Convolutional Neural Networks (CNNs) trained on large scale RGB databases have become the secret sauce in the majority of recent approaches for object categorization from RGB-D data. Thanks to colorization techniques, these methods exploit the filters learned from 2D images to extract meaningful representations in 2.5D. Still, the perceptual signature of these two kind of images is very different, with the first usually strongly characterized by textures, and the second mostly by silhouettes of objects. Ideally, one would like to have two CNNs, one for RGB and one for depth, each trained on a suitable data collection, able to capture the perceptual properties of each channel for the task at hand. This has not been possible so far, due to the lack of a suitable depth database. This paper addresses this issue, proposing to opt for synthetically generated images rather than collecting by hand a 2.5D large scale database. While being clearly a proxy for real data, synthetic images allow to trade quality for quantity, making it possible to generate a virtually infinite amount of data. We show that the filters learned from such data collection, using the very same architecture typically used on visual data, learns very different filters, resulting in depth features (a) able to better characterize the different facets of depth images, and (b) complementary with respect to those derived from CNNs pre-trained on 2D datasets. Experiments on two publicly available databases show the power of our approach.	['Barbara Caputo', 'Paolo Russo', 'Fabio Maria Carlucci']	2016-09-30	null	null	null	null	['object-categorization']	['computer-vision']	[ 1.15880489e-01 3.35303515e-01 2.52701938e-01 -3.05520087e-01 -3.27578127e-01 -6.76076531e-01 8.38902593e-01 2.71321416e-01 -5.40180326e-01 5.37483037e-01 -7.24620894e-02 -6.17246814e-02 -2.11206917e-03 -1.25638294e+00 -6.83459699e-01 -8.95016313e-01 -5.81167452e-02 5.62943637e-01 3.87710094e-01 -3.27298164e-01 1.17175184e-01 9.80183184e-01 -2.07351995e+00 4.39341486e-01 3.68273050e-01 1.55687916e+00 2.44153932e-01 5.54341853e-01 -4.66403246e-01 4.52538460e-01 -6.40217483e-01 -4.55137491e-01 7.66172469e-01 -2.42493659e-01 -4.92955744e-01 3.57398897e-01 5.98482430e-01 -1.31537929e-01 -2.94276655e-01 1.01140034e+00 2.65349358e-01 -1.96518019e-01 6.55065060e-01 -1.05624092e+00 -5.24016976e-01 2.51963139e-01 -1.47182599e-01 -1.84660181e-01 4.10908133e-01 2.53797889e-01 8.06667805e-01 -6.21502936e-01 8.27343404e-01 1.33762026e+00 5.36263525e-01 6.75267518e-01 -1.49108350e+00 -1.53291419e-01 -2.99255073e-01 -1.62060216e-01 -1.24380136e+00 -1.23672731e-01 9.69242752e-01 -5.15485823e-01 4.03174669e-01 2.31940567e-01 1.11662519e+00 1.40622306e+00 -3.39462757e-01 6.23903990e-01 1.72443950e+00 -4.45762426e-01 5.06642997e-01 5.21223724e-01 -2.42568269e-01 3.77343148e-01 4.42099839e-01 3.11182022e-01 -2.70743042e-01 1.56861424e-01 9.01549280e-01 5.67552857e-02 -5.90631306e-01 -9.52497542e-01 -1.14937997e+00 6.72640145e-01 8.58556032e-01 4.58010733e-01 -4.28853899e-01 1.07474951e-02 2.06341356e-01 5.14043808e-01 2.36338735e-01 3.93738627e-01 -4.94785279e-01 8.61702114e-02 -7.28592277e-01 3.19952041e-01 7.68335879e-01 7.70797431e-01 1.05368626e+00 -9.21359137e-02 9.63807032e-02 5.00908494e-01 1.50862500e-01 5.43053448e-01 5.71784437e-01 -6.78538620e-01 2.67282873e-01 9.55662549e-01 1.26650959e-01 -1.02745330e+00 -3.96229416e-01 -2.12948903e-01 -9.34087873e-01 8.29720676e-01 1.21254015e+00 3.26327622e-01 -9.94695425e-01 1.46804166e+00 2.06782907e-01 -4.78896379e-01 1.45342276e-01 1.06940544e+00 5.29106498e-01 3.38116229e-01 -3.38230580e-01 3.55023533e-01 1.29636419e+00 -2.83415377e-01 -6.43541738e-02 -1.03816219e-01 2.95442492e-01 -5.70748568e-01 1.18354273e+00 8.07507396e-01 -7.96098471e-01 -7.51915216e-01 -1.17430103e+00 3.22832130e-02 -8.84699702e-01 7.18192523e-03 5.08412957e-01 9.26950157e-01 -1.14959466e+00 7.28684783e-01 -4.33343887e-01 -4.23099697e-01 4.88490611e-01 1.40141159e-01 -6.56848729e-01 -2.08939523e-01 -1.01973498e+00 6.67350352e-01 6.57348752e-01 2.78805435e-01 -8.33983004e-01 -2.03981653e-01 -5.95686913e-01 -8.76368880e-02 5.24402484e-02 -1.90708935e-01 7.32445300e-01 -1.31697035e+00 -1.32085049e+00 1.34658897e+00 5.18561482e-01 -4.27613646e-01 9.97864723e-01 1.36144176e-01 -9.16771367e-02 3.84115905e-01 -2.84427792e-01 7.80191898e-01 1.16487718e+00 -1.85580778e+00 -4.28817660e-01 -6.15582585e-01 3.96665245e-01 -2.78712422e-01 -3.75830919e-01 -5.70769846e-01 -4.28032547e-01 -4.22359526e-01 2.45979071e-01 -8.65892470e-01 -8.36667046e-02 4.68979090e-01 -3.47932518e-01 2.50358045e-01 5.83179891e-01 -3.47852081e-01 4.05215889e-01 -2.12075329e+00 2.50243038e-01 4.09634739e-01 3.66125733e-01 3.30661446e-01 -8.61602128e-02 3.28553557e-01 -1.34803802e-01 3.89195196e-02 -2.37360269e-01 -3.15481782e-01 -6.39237510e-03 3.59927684e-01 -1.44962430e-01 6.18917704e-01 4.47546810e-01 7.45710552e-01 -8.11603308e-01 -1.93675235e-01 3.75164241e-01 5.95523715e-01 -1.90801874e-01 2.72409052e-01 -3.95524770e-01 5.68419576e-01 -4.87728834e-01 4.51949298e-01 9.06175911e-01 1.16329610e-01 5.38082980e-02 -2.85058230e-01 8.50071386e-02 -1.32269368e-01 -1.22173655e+00 1.77900660e+00 -5.19787133e-01 7.08638012e-01 4.05339152e-02 -1.10197687e+00 1.38899541e+00 1.58769727e-01 3.76196116e-01 -1.04675961e+00 3.61953646e-01 5.79584897e-01 -1.70727685e-01 -3.58720928e-01 1.55826181e-01 -3.12337995e-01 8.74392129e-03 3.05071414e-01 1.65710881e-01 -6.24102235e-01 1.06564559e-01 -1.76768526e-01 9.14360523e-01 1.46848723e-01 -1.28935441e-01 -3.61894146e-02 5.26531994e-01 -1.34669945e-01 1.24049075e-01 6.86095476e-01 -1.75361813e-03 1.05247545e+00 8.00091922e-01 -8.06841373e-01 -1.37043524e+00 -1.15439534e+00 -2.89436191e-01 1.69294029e-01 1.56813994e-01 7.96824321e-02 -6.51314974e-01 -6.43449366e-01 2.96986908e-01 1.09026991e-01 -1.06993914e+00 -1.39976248e-01 -4.22020435e-01 -3.86732191e-01 5.78743935e-01 8.82099718e-02 4.88858372e-01 -1.01496410e+00 -1.27020502e+00 -5.25146164e-02 4.42152679e-01 -1.08766854e+00 4.81270880e-01 5.85815012e-01 -9.12475586e-01 -1.23067534e+00 -9.36060488e-01 -1.93964407e-01 5.52616596e-01 2.22907975e-01 1.28750062e+00 8.02833363e-02 -4.60232645e-01 3.09451103e-01 -6.32029533e-01 -3.66335243e-01 -4.98313755e-01 -8.75530392e-02 -1.46178320e-01 3.71133059e-01 2.23531246e-01 -7.19462156e-01 -6.57048762e-01 1.31205767e-01 -1.47784865e+00 -1.45614222e-01 9.01301146e-01 5.84637582e-01 3.29163343e-01 -1.64953381e-01 9.66559052e-02 -7.89026082e-01 3.26824576e-01 -1.51668042e-01 -6.52897000e-01 8.62391442e-02 -4.90631253e-01 2.86817998e-01 8.06446850e-01 -4.26555932e-01 -6.72979116e-01 1.08814158e-01 -1.00085437e-01 -6.18788064e-01 -7.67327964e-01 -7.46789947e-02 -3.16486180e-01 -2.14543119e-01 9.32631850e-01 1.71257779e-01 2.05595240e-01 -6.79614723e-01 4.44599599e-01 6.26963675e-01 3.56558502e-01 -5.65016985e-01 1.06389666e+00 8.54059458e-01 1.64549068e-01 -7.86010206e-01 -4.67571437e-01 -3.25577706e-01 -8.45092893e-01 -1.61841556e-01 9.15603101e-01 -5.94027877e-01 -5.66467822e-01 5.08935750e-01 -1.07874060e+00 -3.79137427e-01 -8.71529698e-01 3.21560621e-01 -5.80422640e-01 2.20659971e-02 -2.69436389e-01 -7.99319446e-01 3.08286905e-01 -1.07536519e+00 1.27809775e+00 9.54549909e-02 2.67310947e-01 -8.43490899e-01 1.44895360e-01 4.71318513e-02 4.45299000e-01 6.54514432e-01 9.61196721e-01 -4.02862817e-01 -7.04165876e-01 -5.19003034e-01 -4.40364093e-01 6.59987390e-01 1.57251790e-01 -8.14726204e-02 -1.51230526e+00 -5.56032099e-02 9.29747894e-02 -5.34834385e-01 1.01633620e+00 -6.54049590e-02 1.11374712e+00 9.15850792e-03 8.17468315e-02 5.70138872e-01 1.75239038e+00 -9.86708179e-02 8.42221737e-01 4.33506250e-01 5.50017357e-01 8.71701598e-01 2.30758205e-01 3.12328637e-01 -2.57054418e-02 8.62685978e-01 8.83673429e-01 -3.05582255e-01 -3.45517516e-01 -1.33887440e-01 8.46594274e-02 5.67974485e-02 -2.13750347e-01 -3.71560454e-02 -9.29838717e-01 4.86945421e-01 -1.41684246e+00 -5.85086226e-01 -1.46782771e-01 2.43060756e+00 5.58085561e-01 5.07090688e-01 3.23737949e-01 5.90407610e-01 3.25222403e-01 1.31684199e-01 -3.17381680e-01 -3.35101992e-01 -6.01754487e-01 4.40571785e-01 5.65184295e-01 -1.03982259e-02 -8.81636620e-01 3.15490663e-01 4.74548244e+00 6.28681600e-01 -1.39458287e+00 -2.78103501e-01 5.53359509e-01 5.60788997e-02 -4.19117481e-01 -5.01201823e-02 -3.30272377e-01 3.64667892e-01 6.78724408e-01 3.29796940e-01 3.03775281e-01 8.93412769e-01 -1.80402026e-02 -2.55627483e-01 -1.15290225e+00 1.18902588e+00 -1.03616312e-01 -1.16586518e+00 2.51211882e-01 3.17070067e-01 4.96460587e-01 -9.58726406e-02 1.08538553e-01 -1.31457731e-01 -9.32828411e-02 -1.15859878e+00 1.15880239e+00 7.17708468e-01 6.25392556e-01 -4.45937485e-01 6.98894024e-01 2.79962510e-01 -8.44880283e-01 -1.62770420e-01 -5.61449528e-01 7.06151128e-02 -2.17141271e-01 8.33515882e-01 -5.29401422e-01 5.37395477e-01 8.50639224e-01 5.58089674e-01 -8.86951268e-01 9.48026299e-01 -1.42558143e-01 7.57749677e-02 -3.22755426e-01 -1.32926986e-01 2.88717210e-01 -2.25627199e-01 3.59419644e-01 1.00663543e+00 2.85235047e-01 -5.04842758e-01 -2.15115264e-01 1.02807081e+00 -6.77238330e-02 -4.88504730e-02 -9.10612345e-01 -1.14461705e-02 6.41359463e-02 1.37174439e+00 -9.09258604e-01 -1.43124193e-01 -3.35858077e-01 9.64685559e-01 1.89760461e-01 2.72984385e-01 -3.72271061e-01 -3.98442477e-01 6.78412139e-01 2.96567023e-01 4.94061530e-01 -4.18399036e-01 -1.23819493e-01 -1.16258979e+00 1.43596947e-01 -8.35610449e-01 -1.27561271e-01 -8.21880221e-01 -1.40165830e+00 8.90932143e-01 -1.66169927e-01 -1.46782780e+00 -1.08104974e-01 -1.26341927e+00 -1.78382516e-01 1.00312686e+00 -1.50794005e+00 -1.03006268e+00 -6.78180277e-01 6.51952267e-01 9.62090343e-02 1.44008681e-01 8.36969256e-01 1.25988021e-01 5.54568274e-03 3.95886041e-02 1.38655663e-01 1.02945171e-01 4.53795135e-01 -1.56153548e+00 2.30554774e-01 4.44437772e-01 4.65553135e-01 3.02044213e-01 6.21950924e-01 3.36450897e-02 -1.54609728e+00 -6.93647504e-01 1.87294975e-01 -4.55147922e-01 3.59224647e-01 -6.63654268e-01 -9.55731452e-01 -8.86743050e-03 -1.31370455e-01 2.11808801e-01 1.41856000e-01 -3.37153584e-01 -6.43461347e-01 -3.52872729e-01 -1.22816122e+00 2.34705031e-01 8.27176034e-01 -6.47726774e-01 -4.69653964e-01 9.13430005e-02 2.65545934e-01 -2.16619924e-01 -8.31960976e-01 1.25139356e-01 6.66755497e-01 -1.79111397e+00 1.22136450e+00 -2.77312338e-01 5.74512184e-01 -2.21786648e-01 -3.46828133e-01 -1.23324645e+00 3.23519856e-01 5.19386977e-02 2.44191334e-01 9.78339970e-01 2.12520942e-01 -5.57444751e-01 9.67370987e-01 3.42360586e-01 1.81070343e-01 -6.29554570e-01 -9.77866054e-01 -9.06778991e-01 1.26529142e-01 -5.71113884e-01 7.83497512e-01 6.79013252e-01 -7.29696095e-01 -9.74600166e-02 -5.37665514e-03 -1.69303566e-01 6.24962568e-01 3.93048048e-01 1.02876627e+00 -1.60833979e+00 -4.52590644e-01 -6.46982729e-01 -8.43930185e-01 -8.22283506e-01 -9.45573896e-02 -6.76595688e-01 -5.14488108e-02 -1.14830697e+00 -2.75669754e-01 -8.09955716e-01 -8.27493221e-02 2.58163124e-01 3.31947595e-01 6.68546796e-01 2.61003256e-01 2.29052439e-01 4.86203879e-02 5.09032786e-01 1.29858029e+00 -2.85145670e-01 -4.23599267e-03 -8.48432109e-02 -4.14242297e-01 6.61742091e-01 6.86975598e-01 -2.40276784e-01 -3.03072840e-01 -3.42180580e-01 3.55422497e-01 -2.25261331e-01 7.55164981e-01 -1.34618866e+00 -2.42266074e-01 1.05395801e-01 7.58656263e-01 -9.82391834e-02 5.75186551e-01 -1.15655017e+00 1.08561806e-01 4.67191428e-01 -2.29626819e-01 -3.42814296e-01 4.02694242e-03 5.84453642e-01 -4.20552373e-01 -2.73340374e-01 8.58206272e-01 -5.71517348e-01 -7.92250156e-01 1.96115404e-01 -3.27125154e-02 -2.03302562e-01 9.02795792e-01 -7.13539600e-01 -4.62433696e-02 -2.65765905e-01 -6.89388573e-01 -4.39806670e-01 1.02351272e+00 3.77688348e-01 6.52737260e-01 -1.24548376e+00 -4.03100818e-01 5.05315244e-01 4.93610889e-01 2.02832088e-01 2.60045771e-02 3.71659011e-01 -7.77631283e-01 2.53129274e-01 -6.97881460e-01 -7.09686995e-01 -7.47157633e-01 8.27902257e-01 4.37744617e-01 -7.60094747e-02 -6.16034269e-01 6.66561544e-01 1.06591657e-01 -3.89533073e-01 9.48271826e-02 -6.77774131e-01 -4.90499334e-03 3.60436112e-01 3.46567452e-01 -1.33758694e-01 4.03377056e-01 -5.25973797e-01 -1.92054406e-01 7.59814739e-01 4.47416455e-01 -1.46879554e-01 1.48380256e+00 -9.94854514e-03 -9.49640051e-02 5.84511697e-01 1.39321220e+00 1.10016800e-01 -1.48699176e+00 -9.89899188e-02 1.27538487e-01 -8.37322235e-01 -1.05673589e-01 -5.30145526e-01 -1.29497564e+00 1.28497088e+00 8.47026825e-01 9.13842738e-01 1.21314490e+00 -9.99241136e-03 4.56261158e-01 2.70938963e-01 8.07328939e-01 -7.50201702e-01 2.62605399e-01 4.59132753e-02 9.00981963e-01 -1.22094822e+00 -3.57617736e-01 -1.19592212e-01 -2.84013689e-01 1.54460180e+00 1.96865812e-01 -3.21657240e-01 3.33684117e-01 2.51461868e-03 2.65576005e-01 -2.95842648e-01 -2.46566460e-01 -5.97276628e-01 1.88887894e-01 8.70685577e-01 1.25830054e-01 -4.17442434e-02 1.55074820e-01 -1.16611337e-02 -1.00632377e-01 -1.29665285e-01 5.49935043e-01 7.84385324e-01 -2.84340888e-01 -1.48986137e+00 -6.46716177e-01 2.54278541e-01 -4.80261967e-02 2.89045185e-01 -7.66183257e-01 1.06998467e+00 5.38076997e-01 5.77675521e-01 7.80649558e-02 -4.31105107e-01 6.86788261e-01 2.66722273e-02 7.99978673e-01 -3.95631194e-01 -4.66780275e-01 -2.83470452e-01 -1.91599831e-01 -6.67824030e-01 -5.75414360e-01 -4.28751498e-01 -6.88356042e-01 -9.88631845e-02 1.05720088e-01 -4.03568260e-02 1.07912111e+00 6.02910340e-01 -5.99844679e-02 2.45469466e-01 7.85853565e-01 -1.37720430e+00 -4.61595714e-01 -8.39098155e-01 -9.65767503e-01 9.64114428e-01 5.49022675e-01 -8.04499805e-01 -5.07058799e-01 -1.14278883e-01]	[8.426189422607422, -2.769665002822876]
aa509eac-4bd0-4845-8544-549252c2ae56	da-bev-depth-aware-bev-transformer-for-3d	2302.13002	null	https://arxiv.org/abs/2302.13002v2	https://arxiv.org/pdf/2302.13002v2.pdf	Introducing Depth into Transformer-based 3D Object Detection	In this paper, we present DAT, a Depth-Aware Transformer framework designed for camera-based 3D detection. Our model is based on observing two major issues in existing methods: large depth translation errors and duplicate predictions along depth axes. To mitigate these issues, we propose two key solutions within DAT. To address the first issue, we introduce a Depth-Aware Spatial Cross-Attention (DA-SCA) module that incorporates depth information into spatial cross-attention when lifting image features to 3D space. To address the second issue, we introduce an auxiliary learning task called Depth-aware Negative Suppression loss. First, based on their reference points, we organize features as a Bird's-Eye-View (BEV) feature map. Then, we sample positive and negative features along each object ray that connects an object and a camera and train the model to distinguish between them. The proposed DA-SCA and DNS methods effectively alleviate these two problems. We show that DAT is a versatile method that enhances the performance of all three popular models, BEVFormer, DETR3D, and PETR. Our evaluation on BEVFormer demonstrates that DAT achieves a significant improvement of +2.8 NDS on nuScenes val under the same settings. Moreover, when using pre-trained VoVNet-99 as the backbone, DAT achieves strong results of 60.0 NDS and 51.5 mAP on nuScenes test. Our code will be soon.	['Xingyu Liao', 'Ailing Zeng', 'Lei Zhang', 'Shilong Liu', 'Feng Li', 'Hongyang Li', 'Hao Zhang']	2023-02-25	null	null	null	null	['auxiliary-learning']	['methodology']	[-3.96433985e-03 -2.08259463e-01 -4.22026545e-01 -2.47968823e-01 -9.22186196e-01 -6.07286990e-01 6.47636294e-01 -3.75781834e-01 -2.88825780e-01 2.54426181e-01 2.50420898e-01 -2.69304395e-01 1.74147353e-01 -8.22970986e-01 -7.35739291e-01 -6.08226061e-01 2.74306566e-01 1.27458185e-01 7.56720603e-01 -2.13696241e-01 5.05215228e-01 7.53430784e-01 -1.50808644e+00 2.47837752e-01 7.33597994e-01 1.25566816e+00 3.38328481e-01 5.24122000e-01 -3.11260484e-02 5.06384015e-01 -3.65181059e-01 -3.71219188e-01 5.63066781e-01 1.24554388e-01 -4.63068306e-01 7.60112703e-02 8.04678917e-01 -7.65135348e-01 -5.01476288e-01 8.77393425e-01 7.50976145e-01 -1.29509613e-01 7.44561672e-01 -1.17275393e+00 -5.52452505e-01 2.46728156e-02 -1.19289958e+00 3.39955449e-01 2.92155266e-01 2.79919416e-01 8.87890816e-01 -1.32294607e+00 5.63199222e-01 1.33534741e+00 7.22535610e-01 5.72417736e-01 -8.27761352e-01 -7.06800580e-01 2.80045241e-01 1.04006015e-01 -1.37276483e+00 -1.27396062e-01 9.34337258e-01 -5.47799885e-01 1.24230444e+00 2.35055923e-01 4.83133495e-01 1.14426517e+00 2.48098344e-01 1.05767930e+00 8.57952714e-01 -3.49612027e-01 -1.63852405e-02 1.97325677e-01 1.30884245e-01 6.93394303e-01 -2.44309586e-02 4.71990645e-01 -5.04648626e-01 2.95903206e-01 8.64281654e-01 -8.74961242e-02 -2.23724827e-01 -4.49063897e-01 -1.02268279e+00 7.59323180e-01 7.82630563e-01 -1.27950355e-01 -9.45003331e-02 -2.21602041e-02 2.07449704e-01 -7.75516778e-02 8.00662816e-01 3.09106231e-01 -4.00159776e-01 2.17958614e-01 -6.69238925e-01 3.39865118e-01 1.23104341e-01 1.06603265e+00 6.38807476e-01 -1.87573150e-01 -2.18467757e-01 7.33429968e-01 3.47910434e-01 6.14322186e-01 2.77145743e-01 -6.82097614e-01 6.77079499e-01 7.63708055e-01 -5.91140613e-02 -1.04536998e+00 -4.46373284e-01 -5.36198080e-01 -6.98283076e-01 5.16129732e-01 8.23187549e-03 4.93955091e-02 -1.09003806e+00 1.56492794e+00 4.89093542e-01 6.59981519e-02 -1.48868784e-01 1.08000302e+00 1.02444541e+00 5.96042931e-01 -1.67015567e-01 2.38210514e-01 1.02642238e+00 -1.24568987e+00 -2.46446207e-01 -3.68345708e-01 6.66694462e-01 -6.80220842e-01 1.06475770e+00 4.22132939e-01 -1.08457220e+00 -5.77105284e-01 -1.13989604e+00 -4.85085994e-01 -3.09981495e-01 3.09060246e-01 6.22121096e-01 5.05339324e-01 -1.09764373e+00 3.34294647e-01 -6.44376874e-01 -3.65516216e-01 4.39022988e-01 3.63055468e-01 -2.97721982e-01 -5.08935377e-02 -8.51657152e-01 7.26487815e-01 6.71142712e-02 -1.18532062e-01 -9.61769223e-01 -6.89444244e-01 -8.44913483e-01 -2.11854890e-01 3.15665662e-01 -7.89608896e-01 1.13319504e+00 -6.18826926e-01 -1.21204984e+00 1.18635809e+00 -2.87991852e-01 -2.00458020e-01 6.75488770e-01 -4.95120913e-01 -2.64932215e-01 5.22994027e-02 3.76845747e-01 9.28204119e-01 6.75070882e-01 -1.32046127e+00 -1.00417209e+00 -6.77990854e-01 7.13005960e-02 4.03602123e-01 -3.26343596e-01 -1.34259060e-01 -9.55623746e-01 -6.65575266e-01 3.53159606e-01 -7.70685256e-01 -5.61561398e-02 3.19330603e-01 -7.66609311e-01 -2.07051054e-01 1.02684116e+00 -2.89250225e-01 1.16519523e+00 -2.29259706e+00 2.29485258e-01 -1.64152810e-03 3.66003841e-01 2.14316547e-01 -3.06649894e-01 7.76159391e-03 -6.24340586e-02 1.49563864e-01 -4.24349587e-03 -5.65210581e-01 -2.97801495e-01 -7.65354186e-02 -3.42086881e-01 4.17449951e-01 4.02553797e-01 8.23298454e-01 -6.79499805e-01 -2.39263445e-01 5.41278780e-01 4.07122701e-01 -7.12716699e-01 1.33752882e-01 -7.35187829e-02 2.12647438e-01 -4.95170355e-01 8.23023558e-01 1.14020097e+00 -1.35904685e-01 -4.39236075e-01 -4.12367702e-01 -3.82392079e-01 2.67501146e-01 -9.56401646e-01 1.68933225e+00 -3.67653847e-01 5.76281965e-01 -2.72883981e-01 -4.99690682e-01 8.57197464e-01 -2.22329587e-01 2.97537595e-01 -8.87389004e-01 1.60439268e-01 1.49964496e-01 -4.50411201e-01 -3.95425290e-01 5.97477198e-01 1.11729175e-01 1.50093958e-01 5.58772199e-02 -5.28340302e-02 -1.59324020e-01 -3.07393283e-01 2.00317129e-01 1.05047500e+00 1.46894470e-01 1.06156871e-01 -7.62367025e-02 4.74704653e-01 -5.46811260e-02 4.54844475e-01 5.68489015e-01 -2.33007550e-01 1.06362569e+00 5.26958406e-01 -4.15968001e-01 -9.90519881e-01 -1.22032154e+00 -1.60629839e-01 7.15593755e-01 5.29302716e-01 -4.68581647e-01 -4.53490704e-01 -1.07901192e+00 6.50226176e-02 6.13542438e-01 -7.18375504e-01 -1.46891266e-01 -3.77829283e-01 -6.17100298e-01 2.52073586e-01 9.00714219e-01 6.53739274e-01 -4.50174838e-01 -3.62581015e-01 -1.87703818e-01 -3.51966266e-03 -1.07853758e+00 -4.90159571e-01 1.92122638e-01 -7.73192644e-01 -9.71628189e-01 -7.48995304e-01 -7.38962471e-01 4.72106934e-01 8.73106897e-01 1.00025392e+00 -1.66862443e-01 -1.38162807e-01 4.90787439e-02 -3.46376121e-01 -3.80653709e-01 1.92758754e-01 2.49968290e-01 -1.16793439e-02 -3.51912469e-01 4.38893139e-01 -4.45498854e-01 -7.44450271e-01 5.77692449e-01 -6.53776944e-01 2.84711838e-01 5.46158612e-01 6.57475829e-01 8.51721406e-01 -2.16519862e-01 1.31041840e-01 -4.76456255e-01 4.80150208e-02 -4.22686905e-01 -8.25777531e-01 -5.46780936e-02 -5.19223928e-01 -4.44899425e-02 1.84663311e-01 -1.80176765e-01 -8.80544484e-01 2.47805938e-01 -4.19620574e-01 -7.39684880e-01 2.17584856e-02 2.18719039e-02 -3.97874326e-01 -1.54670596e-01 5.83538711e-01 1.91690862e-01 -3.53319257e-01 -5.09589732e-01 1.81087792e-01 7.22665370e-01 4.87917006e-01 -1.71496898e-01 1.03240061e+00 7.22371876e-01 6.67212578e-03 -8.47122371e-01 -1.14355314e+00 -5.95449507e-01 -6.36143804e-01 -3.57952267e-02 1.01474547e+00 -1.25731111e+00 -4.93301451e-01 6.56936705e-01 -1.23640275e+00 -2.77129948e-01 -9.51259285e-02 3.75532985e-01 -2.81059295e-01 3.03370267e-01 -3.70456368e-01 -6.68534279e-01 -2.65784889e-01 -1.21256936e+00 1.55373883e+00 1.26247257e-01 1.05240732e-01 -8.22025716e-01 5.42822480e-02 2.85660923e-01 1.37188047e-01 1.83693111e-01 7.74805903e-01 -2.45276347e-01 -7.90628076e-01 -1.35128036e-01 -6.72539055e-01 4.50433463e-01 -1.73528995e-02 1.60008688e-02 -1.31657732e+00 -2.17426702e-01 -7.31844530e-02 -2.76460916e-01 1.05168509e+00 4.78386402e-01 1.31982410e+00 1.80017471e-01 -5.81365645e-01 1.07109857e+00 1.44107497e+00 1.74574465e-01 7.15142846e-01 5.27761757e-01 9.63230133e-01 4.20452356e-01 7.42002249e-01 3.51703316e-01 4.79240894e-01 9.10734951e-01 8.47885728e-01 -4.24008191e-01 -3.27046543e-01 -4.71953213e-01 3.47890407e-01 2.99420059e-01 3.21323983e-02 -5.28661668e-01 -8.28146696e-01 5.06605446e-01 -1.49698460e+00 -7.40107656e-01 -4.08527315e-01 1.96346247e+00 3.13049108e-01 3.86500716e-01 8.19099396e-02 5.05338982e-02 5.85654438e-01 3.02258134e-01 -6.69575989e-01 -4.24760841e-02 -3.29082787e-01 -1.26409074e-02 6.57984018e-01 4.91927266e-01 -1.34112692e+00 1.06346667e+00 5.85099220e+00 8.48996878e-01 -1.34473872e+00 -4.56729904e-03 7.49416649e-01 -3.75247389e-01 -3.96566927e-01 -3.18816155e-01 -1.28180885e+00 3.31213951e-01 2.71620214e-01 1.87440336e-01 -2.78822947e-02 1.06716204e+00 2.06776142e-01 -1.16609477e-01 -1.16731393e+00 1.18601537e+00 2.98147261e-01 -1.26055324e+00 1.08548917e-01 1.38009205e-01 7.51889825e-01 4.06044781e-01 3.03313136e-01 2.17221558e-01 3.62123139e-02 -9.12710190e-01 8.82844150e-01 -1.25644905e-02 9.99813199e-01 -7.56769419e-01 6.76031709e-01 2.83950239e-01 -1.14223456e+00 1.85249746e-02 -4.52828616e-01 -7.39116035e-03 1.35978594e-01 7.14471877e-01 -6.12511694e-01 5.93807697e-01 8.41935158e-01 1.00488341e+00 -6.30487502e-01 1.12376893e+00 -4.22392309e-01 2.06632420e-01 -4.91218954e-01 2.12452665e-01 4.53309238e-01 2.02364758e-01 6.36195838e-01 1.05209625e+00 5.68550766e-01 -3.86145040e-02 -9.39000323e-02 8.05391967e-01 -3.59201059e-02 -1.87631145e-01 -8.01607609e-01 6.24758184e-01 3.30598414e-01 1.10927510e+00 -4.46909487e-01 -8.73137042e-02 -5.48438907e-01 1.24239767e+00 3.01323622e-01 1.77726403e-01 -1.10671723e+00 -1.93834841e-01 8.02286148e-01 4.33012456e-01 4.29516971e-01 -2.89666671e-02 -4.32958841e-01 -1.15600562e+00 2.09404081e-01 -5.02361953e-01 1.80573329e-01 -1.03148746e+00 -1.25955141e+00 6.59329772e-01 -2.15115517e-01 -1.53995943e+00 2.08545357e-01 -8.92433167e-01 -5.37372887e-01 8.02003205e-01 -1.85060513e+00 -1.23350227e+00 -6.08930647e-01 5.25612175e-01 7.00015903e-01 1.81353949e-02 4.34922755e-01 4.49329019e-01 -6.72777653e-01 7.11204529e-01 -1.01622477e-01 9.63388884e-04 7.55028546e-01 -1.15935552e+00 7.75993645e-01 8.29522371e-01 -9.52398628e-02 1.75071329e-01 2.37855420e-01 -5.99771559e-01 -1.28368354e+00 -1.33900189e+00 7.14038432e-01 -7.10701227e-01 3.67432624e-01 -4.33289438e-01 -6.28677309e-01 7.35844076e-01 -1.25391483e-01 1.07345156e-01 2.52816737e-01 -7.83857852e-02 -4.99482960e-01 -1.49717137e-01 -1.06569791e+00 5.35850108e-01 1.41403317e+00 -4.04423177e-01 -4.16924804e-01 3.24560851e-01 9.48499382e-01 -7.25583792e-01 -4.47526008e-01 5.89259028e-01 4.28970158e-01 -1.37882745e+00 1.21989703e+00 -2.29272202e-01 7.81548440e-01 -2.75674134e-01 -4.73520666e-01 -1.12518740e+00 -3.96067828e-01 -2.72150755e-01 7.94979110e-02 1.10112548e+00 4.36428994e-01 -4.19548452e-01 1.06741130e+00 1.39261544e-01 -4.37034518e-01 -9.89164710e-01 -9.11080897e-01 -7.35457778e-01 2.05397382e-01 -6.88928425e-01 4.91448522e-01 8.06935787e-01 -4.11026567e-01 5.55116117e-01 -3.58891129e-01 4.15432423e-01 5.32628179e-01 -7.99349248e-02 9.07717824e-01 -1.04419684e+00 -1.56231269e-01 -4.03272271e-01 -5.37135184e-01 -1.67066431e+00 -1.40285403e-01 -7.32139289e-01 -2.10506096e-01 -1.47144532e+00 2.79592693e-01 -4.21553314e-01 -9.85800475e-02 2.61733353e-01 -1.01748109e-01 3.39932919e-01 1.63400806e-02 2.36552972e-02 -4.63814437e-01 7.18398988e-01 1.30369365e+00 -1.01045825e-01 -2.39073679e-01 -4.06812616e-02 -6.52537644e-01 9.34727371e-01 6.51321828e-01 -2.90729016e-01 -4.74211872e-01 -9.00871873e-01 4.48792614e-02 -2.28339449e-01 5.66678047e-01 -1.03705287e+00 6.52255937e-02 -6.40906096e-02 6.54655337e-01 -1.29396057e+00 5.47670484e-01 -7.22854555e-01 -3.53244543e-01 3.25445980e-01 8.92195255e-02 1.62092492e-01 3.12498957e-01 4.75865662e-01 -1.71465591e-01 1.52588844e-01 9.64252830e-01 5.60440458e-02 -1.01612842e+00 4.83569533e-01 1.02363579e-01 -4.21879031e-02 1.26544082e+00 -4.14943933e-01 -6.04186714e-01 -1.36319563e-01 -3.44618857e-01 4.46336657e-01 6.24573290e-01 4.64051753e-01 9.34833884e-01 -1.42386472e+00 -5.64461529e-01 5.44795692e-01 4.06794071e-01 3.35498244e-01 4.16796654e-01 7.28223801e-01 -5.13641179e-01 5.48750877e-01 9.42376349e-03 -9.44122612e-01 -1.23338866e+00 5.44698834e-01 3.93750548e-01 -1.15892909e-01 -6.74598932e-01 1.23067403e+00 8.08577895e-01 -4.05012339e-01 4.59824681e-01 -4.83665943e-01 -9.02782753e-02 1.85422902e-03 5.72387815e-01 2.94842213e-01 2.18394205e-01 -5.30763209e-01 -5.50363183e-01 1.07676280e+00 -2.55740106e-01 4.40804064e-02 1.35705101e+00 -2.61203468e-01 2.96004713e-01 7.88858160e-02 1.38679719e+00 1.75797746e-01 -1.55484903e+00 -1.37564391e-01 -4.06846434e-01 -7.35945165e-01 4.21896636e-01 -8.24405909e-01 -1.24452460e+00 1.03289294e+00 9.25025880e-01 -1.04924038e-01 1.24403358e+00 1.32176265e-01 7.71445572e-01 6.05936535e-02 1.21409014e-01 -6.38809919e-01 2.70325154e-01 6.46138847e-01 8.99153352e-01 -1.46125054e+00 -9.05004442e-02 -7.55794525e-01 -5.99725366e-01 7.82220542e-01 1.06102169e+00 -7.24760741e-02 5.22911489e-01 2.79072970e-01 2.93820724e-02 -2.93261409e-01 -6.20972633e-01 -1.90790102e-01 4.04031485e-01 6.22741878e-01 1.98101982e-01 -2.95537859e-01 1.35445759e-01 4.22195166e-01 -1.08025946e-01 -1.43445268e-01 2.46557668e-01 6.46255016e-01 -4.81285840e-01 -8.57801974e-01 -2.75225818e-01 1.81007981e-01 -2.37019300e-01 -1.78060472e-01 -5.22351801e-01 9.45418000e-01 2.22271413e-01 7.40745127e-01 2.32656911e-01 -9.46329832e-01 6.59129322e-01 -4.86161202e-01 4.88327384e-01 -6.14479542e-01 -2.93101102e-01 1.89101398e-01 -8.82950649e-02 -7.83609092e-01 -1.30892321e-01 -4.25400615e-01 -9.94110703e-01 -4.53446984e-01 -4.40209329e-01 -4.67132777e-01 5.78894258e-01 6.00041091e-01 3.83848995e-01 5.10153174e-01 8.99381757e-01 -1.03960955e+00 -4.59226757e-01 -6.12166703e-01 -4.90752757e-01 1.61242858e-01 5.15458107e-01 -8.68292153e-01 -5.55674613e-01 -4.60614949e-01]	[8.105399131774902, -2.5303804874420166]
f414e603-db76-4127-886c-75c0cc8133d7	contour-detection-from-deep-patch-level	1705.03159	null	http://arxiv.org/abs/1705.03159v1	http://arxiv.org/pdf/1705.03159v1.pdf	Contour Detection from Deep Patch-level Boundary Prediction	In this paper, we present a novel approach for contour detection with Convolutional Neural Networks. A multi-scale CNN learning framework is designed to automatically learn the most relevant features for contour patch detection. Our method uses patch-level measurements to create contour maps with overlapping patches. We show the proposed CNN is able to to detect large-scale contours in an image efficienly. We further propose a guided filtering method to refine the contour maps produced from large-scale contours. Experimental results on the major contour benchmark databases demonstrate the effectiveness of the proposed technique. We show our method can achieve good detection of both fine-scale and large-scale contours.	['Li Shen', 'Teck Wee Chua']	2017-05-09	null	null	null	null	['contour-detection']	['computer-vision']	[ 1.84343025e-01 -8.01987574e-03 7.29939193e-02 -1.66880503e-01 -9.60084081e-01 -6.84627593e-01 1.51555657e-01 4.26896006e-01 -2.54379064e-01 1.72355354e-01 -2.44207382e-01 -9.70354378e-02 2.06670299e-01 -1.27615845e+00 -6.69986665e-01 -4.65984941e-01 -2.71132141e-01 -6.80350587e-02 9.57038760e-01 -3.28280896e-01 4.43990529e-01 9.98077869e-01 -1.33484411e+00 3.60870600e-01 5.75245798e-01 1.39809334e+00 7.27473050e-02 9.26073730e-01 -5.16078882e-02 2.85561383e-01 -6.28090143e-01 -1.88104883e-01 5.54761827e-01 -2.89908648e-01 -3.48632395e-01 4.30319399e-01 1.08674848e+00 -1.99126437e-01 2.17299938e-01 1.14725637e+00 4.84762639e-01 -3.71277243e-01 5.95722020e-01 -6.60433233e-01 -3.73244524e-01 1.70149699e-01 -8.83210421e-01 1.78767160e-01 -1.97135046e-01 -6.35097146e-01 7.59350955e-01 -1.19869184e+00 7.01129735e-01 1.15374184e+00 1.18162715e+00 -7.64241368e-02 -9.56561148e-01 -4.78619754e-01 -2.04259813e-01 -4.40580547e-01 -1.55844581e+00 5.94425425e-02 1.32973540e+00 -2.91159481e-01 5.23321271e-01 9.04292539e-02 7.33211219e-01 -1.83070287e-01 8.30171049e-01 8.16442549e-01 1.23160636e+00 -6.96278155e-01 1.35418549e-01 -2.72953510e-01 4.84739728e-02 1.21701729e+00 4.05344546e-01 2.89173067e-01 -3.56163919e-01 -9.47622731e-02 1.56314838e+00 -3.40192705e-01 -1.67691141e-01 -5.22749543e-01 -8.32499981e-01 1.02279568e+00 6.80321336e-01 4.25607026e-01 -3.31917167e-01 2.14854985e-01 1.23370305e-01 1.65353373e-01 8.08536708e-01 2.35614106e-01 -4.32665139e-01 6.40463114e-01 -1.26359510e+00 5.21939993e-02 5.12082636e-01 7.93472648e-01 1.00689578e+00 -1.09281324e-01 -6.17296733e-02 8.07035506e-01 1.94533005e-01 3.93116653e-01 6.91890642e-02 -9.50383902e-01 -2.05121666e-01 6.39564276e-01 3.37846391e-02 -1.36551845e+00 -6.74508452e-01 -4.00060564e-01 -9.17475402e-01 8.98288965e-01 5.21832049e-01 -2.68078744e-01 -1.27146399e+00 7.99094915e-01 4.54067349e-01 1.69376895e-01 -8.81426409e-02 7.13080823e-01 9.15484250e-01 5.44476569e-01 -3.45710933e-01 -1.42510667e-01 1.26901734e+00 -7.96959758e-01 -7.11092114e-01 -1.14072978e-01 1.18159838e-01 -1.24035800e+00 4.04998511e-01 3.65778536e-01 -1.27148545e+00 -8.55985105e-01 -1.27404070e+00 1.71515867e-01 -3.26504111e-01 5.57175756e-01 8.14806879e-01 7.04567254e-01 -1.28767955e+00 7.60823071e-01 -7.04627752e-01 -1.02258660e-01 8.18183661e-01 1.41020775e-01 -1.93810359e-01 3.80469978e-01 -6.92045569e-01 5.45466125e-01 4.91903931e-01 4.98943329e-01 -4.06286627e-01 -4.02775854e-01 -1.03713751e+00 7.55992606e-02 3.94643173e-02 -4.65258285e-02 9.78127897e-01 -8.32369387e-01 -1.17552352e+00 9.60140288e-01 -5.03462739e-03 -3.77627343e-01 3.86918753e-01 2.72633955e-02 -1.30726367e-01 7.11099148e-01 7.32605085e-02 7.94700980e-01 1.00085235e+00 -1.43479872e+00 -1.14870095e+00 -3.95886861e-02 -4.11826253e-01 -1.82961687e-01 -1.39733016e-01 8.32273662e-02 -6.38232827e-01 -1.10405576e+00 6.76219523e-01 -3.48029315e-01 -3.20413172e-01 5.83621442e-01 -2.34996766e-01 -2.26784304e-01 9.77761328e-01 -3.32325846e-01 7.67439842e-01 -1.82061255e+00 -5.54243088e-01 5.37187874e-01 1.75476640e-01 2.47784927e-01 -2.73217767e-01 8.95548761e-02 1.98473424e-01 7.23264515e-02 -4.14018780e-01 -2.37807870e-01 -4.93836164e-01 -1.25534907e-01 -7.93877468e-02 5.72118759e-01 5.61011672e-01 1.08603644e+00 -3.59027505e-01 -8.70175421e-01 2.94230253e-01 5.36085725e-01 -1.22738533e-01 1.43248335e-01 -4.13499996e-02 -2.00285479e-01 -2.46362016e-01 1.07976115e+00 1.25635469e+00 -1.92888021e-01 -2.11263210e-01 -6.17289603e-01 -3.66434216e-01 -7.12068379e-01 -1.40529990e+00 1.47486448e+00 -1.91126660e-01 8.48777056e-01 4.01443779e-01 -6.46355033e-01 1.44354784e+00 5.44984937e-02 3.67236197e-01 -5.84506571e-01 3.07411253e-01 3.01178098e-01 -3.29432189e-01 -5.46266772e-02 4.01034653e-01 -2.05655605e-01 1.76973641e-01 2.60072023e-01 1.09244689e-01 -5.95308304e-01 2.65275300e-01 -2.65714258e-01 5.81878662e-01 3.74125242e-02 4.46061164e-01 -7.74208188e-01 5.61719358e-01 3.12560081e-01 6.44149303e-01 7.63220191e-01 -4.89639819e-01 9.93477106e-01 2.06030712e-01 -9.68080044e-01 -7.90096164e-01 -1.01059175e+00 -4.92527843e-01 8.81730556e-01 5.55601478e-01 -1.08346544e-01 -8.20398510e-01 -6.93705916e-01 5.53551614e-02 -2.77717769e-01 -9.47016716e-01 3.84964347e-01 -1.05746841e+00 -7.67719626e-01 3.19235981e-01 7.68352091e-01 6.79470122e-01 -1.40409660e+00 -9.62518275e-01 4.34481055e-01 1.68915436e-01 -8.38769019e-01 -7.62083292e-01 3.03618073e-01 -1.08563614e+00 -1.13480163e+00 -9.18823540e-01 -1.68628228e+00 8.44204247e-01 2.71447212e-01 1.13214326e+00 4.02817100e-01 -8.32424581e-01 -1.08627928e-02 -3.70506346e-01 -3.73005271e-01 -2.81622201e-01 -4.05692875e-01 -9.88074660e-01 1.77292034e-01 -1.36864200e-01 -1.94099978e-01 -8.19047272e-01 4.06528801e-01 -9.33704436e-01 -2.19880760e-01 9.90275204e-01 8.16149652e-01 9.15749133e-01 4.33689833e-01 6.45193458e-01 -7.40887880e-01 5.79983234e-01 3.46876293e-01 -1.04855156e+00 4.65187430e-01 -5.91424644e-01 -9.96960104e-02 4.52856198e-02 -4.75247651e-01 -9.39213812e-01 5.79405785e-01 -2.65953809e-01 -1.53091595e-01 -4.57209021e-01 4.56697315e-01 3.09681982e-01 -8.65249157e-01 5.60445130e-01 2.75264800e-01 -8.62388760e-02 -3.77065003e-01 4.43679929e-01 3.29467684e-01 6.88742459e-01 -3.91026586e-01 8.36533666e-01 1.02848530e+00 6.86070994e-02 -8.78022373e-01 -6.80761814e-01 -6.97348833e-01 -9.63570833e-01 -4.68512744e-01 8.88760328e-01 -7.20151365e-01 -2.96776533e-01 7.94308305e-01 -1.27668560e+00 -4.19065267e-01 -1.66058272e-01 -1.25064477e-01 -4.92199600e-01 3.99982274e-01 -1.00612020e+00 -7.38800526e-01 -7.95530140e-01 -7.52851844e-01 1.49483228e+00 6.35033607e-01 4.42263037e-01 -1.28090000e+00 2.11280555e-01 -2.03682736e-01 5.93073666e-01 6.69607162e-01 4.83861625e-01 -4.90099192e-02 -6.90182745e-01 -2.97474027e-01 -5.11475921e-01 8.40398073e-02 2.10532874e-01 2.34280348e-01 -8.05081844e-01 -3.99254501e-01 -3.15502077e-01 -3.59727025e-01 1.16532767e+00 9.01543081e-01 8.61979187e-01 3.90630662e-01 -4.12917137e-01 7.43734837e-01 1.75531149e+00 1.80397302e-01 5.94304025e-01 1.48701519e-01 2.92298406e-01 3.53653848e-01 9.93071556e-01 1.15288518e-01 -1.88967869e-01 4.00298804e-01 1.64465338e-01 -9.48190570e-01 -5.25742412e-01 -3.17113474e-02 -1.49951413e-01 3.25726122e-01 5.28506823e-02 1.70425683e-01 -6.51388407e-01 7.08974063e-01 -1.52778196e+00 -4.39717829e-01 -3.36100191e-01 1.57810760e+00 7.75531828e-01 5.12174010e-01 -1.65104166e-01 -4.35284190e-02 8.76984477e-01 5.87027781e-02 -2.00599223e-01 -3.32754314e-01 -4.06476110e-01 9.38731372e-01 6.21449947e-01 7.28659749e-01 -1.86164010e+00 1.35923910e+00 7.61146593e+00 9.78069961e-01 -1.27800035e+00 -2.38425270e-01 6.40169799e-01 9.19639170e-01 1.55639231e-01 -3.53656948e-01 -6.82787180e-01 -3.33307564e-01 3.89943421e-02 -3.41909379e-02 -4.80987489e-01 7.77940392e-01 -1.66206568e-01 -2.02037305e-01 -2.45988131e-01 6.61184669e-01 1.51571035e-01 -1.79951358e+00 -1.88465416e-01 -3.20999026e-01 8.81908596e-01 -1.26359299e-01 -2.71624953e-01 -2.46028274e-01 1.53858259e-01 -8.66537988e-01 8.49688947e-01 2.06647635e-01 8.37758958e-01 -8.73653829e-01 8.30691278e-01 2.07683202e-02 -1.98636472e+00 2.93897659e-01 -7.94329643e-01 1.53708532e-01 1.01777852e-01 7.13951230e-01 -6.66086793e-01 2.53864229e-01 6.24265790e-01 6.14224136e-01 -8.67056966e-01 1.50758290e+00 -2.23700181e-01 4.16022629e-01 -3.69379610e-01 -4.82188016e-02 4.79822382e-02 -1.51659578e-01 5.50492518e-02 1.60843253e+00 1.86396852e-01 2.03197211e-01 4.29359615e-01 9.17454541e-01 1.50863498e-01 4.99768525e-01 -3.72866452e-01 3.91925395e-01 1.89642265e-01 1.83769381e+00 -1.78248513e+00 -2.18282148e-01 -3.40896010e-01 8.90053332e-01 2.74766266e-01 1.12697609e-01 -3.35923582e-01 -5.97917557e-01 1.20726936e-01 -1.26033336e-01 7.52988338e-01 -2.28956550e-01 -3.85734320e-01 -7.35321939e-01 -1.15070477e-01 -2.02955484e-01 4.26857501e-01 -6.56492889e-01 -1.15526509e+00 7.85268605e-01 -4.28579301e-01 -1.04594719e+00 4.23665196e-01 -8.02496076e-01 -1.04666567e+00 6.89196527e-01 -1.77644253e+00 -1.61683178e+00 -4.04200852e-01 4.93335307e-01 7.14193404e-01 1.66019052e-01 8.15904140e-01 -8.36000741e-02 4.37534489e-02 4.20451939e-01 -1.40401810e-01 7.71603286e-01 5.90206027e-01 -1.59740424e+00 7.26106107e-01 1.00621700e+00 3.70533347e-01 3.17021221e-01 3.56096506e-01 -1.03478789e+00 -8.71096730e-01 -1.24650013e+00 5.08477926e-01 2.88075674e-03 3.30474496e-01 -4.61960942e-01 -9.25715327e-01 3.33227605e-01 2.93524802e-01 5.50691366e-01 2.77486533e-01 -3.48812461e-01 -2.05629900e-01 -3.68770882e-02 -1.20323396e+00 2.38280483e-02 4.41983223e-01 -8.39549676e-02 -6.02021515e-01 1.29149362e-01 3.77723575e-01 -5.27498901e-01 -1.05825245e+00 7.87111163e-01 7.06932664e-01 -1.16670108e+00 1.10290992e+00 1.15675956e-01 1.85546428e-01 -3.49869013e-01 3.44303310e-01 -1.18269515e+00 -3.33783418e-01 -3.89875412e-01 2.02718556e-01 7.02929795e-01 4.56420183e-01 -4.43582624e-01 1.32032764e+00 -2.91886091e-01 -1.48884133e-01 -8.13299656e-01 -8.55079532e-01 -4.97401416e-01 5.01363695e-01 -1.59859553e-01 2.28820443e-01 5.86561143e-01 -3.96471292e-01 -1.51463464e-01 7.09804744e-02 5.05808711e-01 1.00134373e+00 1.02650189e+00 4.22368556e-01 -1.52294457e+00 1.08989730e-01 -5.32418728e-01 -4.90381926e-01 -1.06299567e+00 -2.80946493e-01 -3.98273528e-01 4.45394188e-01 -1.58321810e+00 1.42599791e-01 -3.71756405e-01 -3.64336818e-01 3.78935248e-01 -3.59297484e-01 9.87503529e-01 7.61092454e-02 -1.29930735e-01 -4.21079218e-01 5.15379459e-02 1.68716538e+00 -1.13499366e-01 -1.37379542e-01 1.21459551e-01 -3.51751626e-01 8.97498608e-01 9.65056777e-01 -3.07541013e-01 1.04235783e-01 1.78900614e-01 -3.33649479e-02 5.41043803e-02 2.53376007e-01 -1.08744299e+00 4.95069534e-01 -3.85631807e-02 8.96170735e-01 -1.19118404e+00 5.01489118e-02 -6.85760081e-01 -4.27134693e-01 7.11883903e-01 -1.87583655e-01 -3.77207190e-01 5.37679076e-01 4.99202549e-01 -4.85468864e-01 -2.53925547e-02 1.27560365e+00 -3.58212702e-02 -8.06779027e-01 2.04576612e-01 -2.72910297e-01 -2.29178578e-01 1.13527071e+00 -6.46737292e-02 -1.81326061e-01 -3.26554626e-01 -1.00157118e+00 -4.75972556e-02 2.85955876e-01 3.81014407e-01 1.21405602e+00 -1.31845450e+00 -8.23278248e-01 6.74148262e-01 7.52709433e-02 -1.82799455e-02 -1.78253412e-01 3.72308522e-01 -1.16752541e+00 1.83923647e-01 -3.98655117e-01 -7.00539231e-01 -1.41394508e+00 2.71689415e-01 8.38589132e-01 -7.41899386e-03 -7.74950325e-01 1.12410927e+00 1.11761503e-01 -2.35507056e-01 9.31704268e-02 -7.71680355e-01 -2.95902073e-01 2.54648123e-02 5.53786159e-01 -7.04446435e-02 2.43649021e-01 -5.77243388e-01 -1.44867212e-01 1.39914393e+00 1.01285018e-01 1.18738905e-01 1.24384224e+00 1.27121210e-01 -2.58270085e-01 -6.19996674e-02 1.21100259e+00 2.06492931e-01 -1.49378526e+00 -6.01055697e-02 5.64861670e-02 -4.53714490e-01 4.01980221e-01 -6.65944278e-01 -1.50747919e+00 8.11608136e-01 1.10533655e+00 2.60748714e-01 1.28596914e+00 -2.70488877e-02 8.82609665e-01 3.65777202e-02 4.37421471e-01 -1.08482754e+00 2.86400825e-01 1.10208131e-01 9.43465114e-01 -1.47859001e+00 3.74338962e-02 -9.99975920e-01 -5.87975122e-02 1.78127134e+00 4.54449594e-01 -6.89106524e-01 1.18256485e+00 8.07740033e-01 7.30714858e-01 -6.90344214e-01 -1.90940410e-01 -5.55368602e-01 6.23485148e-01 4.75472897e-01 3.46056134e-01 4.87209037e-02 -3.74592900e-01 9.81135443e-02 2.55605638e-01 -1.91147327e-01 6.49977267e-01 1.12431645e+00 -1.20321476e+00 -1.04375708e+00 -7.78766572e-01 2.54180908e-01 -4.77125734e-01 7.87901059e-02 -5.02482235e-01 8.83845091e-01 2.14496106e-01 9.51015472e-01 2.87431508e-01 5.38613014e-02 3.72569978e-01 -2.91105777e-01 3.90547097e-01 -4.71294194e-01 -5.57835639e-01 8.70718837e-01 -3.29426199e-01 -6.77766979e-01 -4.62978244e-01 -3.55785400e-01 -1.32469058e+00 4.38956827e-01 -5.43524742e-01 -1.51379630e-02 4.79722500e-01 4.36940759e-01 8.43239203e-03 3.45774889e-01 5.47248483e-01 -1.03985834e+00 -7.24309236e-02 -8.28860462e-01 -9.25217152e-01 4.06218350e-01 4.73615915e-01 -4.46597248e-01 -2.49510512e-01 3.59876364e-01]	[9.497518539428711, 0.217572420835495]
96baa127-f1db-4b24-84ca-6e42d9597695	4d-panoptic-segmentation-as-invariant-and	2303.15651	null	https://arxiv.org/abs/2303.15651v1	https://arxiv.org/pdf/2303.15651v1.pdf	4D Panoptic Segmentation as Invariant and Equivariant Field Prediction	In this paper, we develop rotation-equivariant neural networks for 4D panoptic segmentation. 4D panoptic segmentation is a recently established benchmark task for autonomous driving, which requires recognizing semantic classes and object instances on the road based on LiDAR scans, as well as assigning temporally consistent IDs to instances across time. We observe that the driving scenario is symmetric to rotations on the ground plane. Therefore, rotation-equivariance could provide better generalization and more robust feature learning. Specifically, we review the object instance clustering strategies, and restate the centerness-based approach and the offset-based approach as the prediction of invariant scalar fields and equivariant vector fields. Other sub-tasks are also unified from this perspective, and different invariant and equivariant layers are designed to facilitate their predictions. Through evaluation on the standard 4D panoptic segmentation benchmark of SemanticKITTI, we show that our equivariant models achieve higher accuracy with lower computational costs compared to their non-equivariant counterparts. Moreover, our method sets the new state-of-the-art performance and achieves 1st place on the SemanticKITTI 4D Panoptic Segmentation leaderboard.	['Fatih Porikli', 'Maani Ghaffari Jadidi', 'Shubhankar Borse', 'Hong Cai', 'Shizong Han', 'Minghan Zhu']	2023-03-28	null	null	null	null	['panoptic-segmentation']	['computer-vision']	[-1.03812359e-01 -4.67785120e-01 -3.27668488e-01 -7.34223604e-01 -2.83732653e-01 -7.82940745e-01 8.55815291e-01 -2.28027776e-01 -3.18724573e-01 -6.37328774e-02 -4.53520238e-01 -4.10389811e-01 -3.85507494e-01 -8.61875355e-01 -6.98160648e-01 -8.05599511e-01 -2.61925578e-01 9.85743940e-01 4.12929088e-01 -3.07484508e-01 1.28714696e-01 9.22553897e-01 -1.91068840e+00 -3.64558965e-01 7.28557587e-01 1.04365826e+00 -5.82100824e-02 6.63271964e-01 -1.15721405e-01 1.64121523e-01 -1.77516818e-01 6.92848712e-02 8.94402444e-01 2.58142024e-01 -7.30008721e-01 2.28812814e-01 9.85584795e-01 -1.40967697e-01 -4.99045402e-01 1.05764890e+00 -2.73361467e-02 4.63942230e-01 9.27637160e-01 -1.40166366e+00 -3.32536012e-01 2.41085723e-01 -5.95189631e-01 2.41274312e-01 -6.97186112e-01 1.47844523e-01 1.18663323e+00 -8.91920328e-01 6.17290795e-01 1.30522728e+00 5.17141223e-01 1.63277030e-01 -1.19713724e+00 -7.09117174e-01 5.71108639e-01 3.66145223e-01 -1.29968476e+00 -1.05141975e-01 7.01526046e-01 -5.98963082e-01 7.80322492e-01 2.27117434e-01 7.43412316e-01 5.06657958e-01 2.40413785e-01 8.86226177e-01 1.07293606e+00 2.20775187e-01 2.30810180e-01 -9.85468253e-02 5.68670511e-01 5.05378067e-01 1.47432372e-01 2.34939933e-01 -1.21249594e-01 3.61767501e-01 4.96000946e-01 1.51281223e-01 2.55171329e-01 -9.36685681e-01 -1.33698583e+00 1.02229905e+00 7.86478400e-01 -2.69712415e-02 -1.28008142e-01 1.49361521e-01 2.81261176e-01 1.03726819e-01 7.11377621e-01 4.65574920e-01 -4.87215191e-01 4.04612035e-01 -8.52130175e-01 6.42936051e-01 6.56219900e-01 1.04591918e+00 1.28478348e+00 1.93246603e-01 -5.54804876e-02 7.12003231e-01 3.46722156e-01 9.66317832e-01 -7.87161291e-02 -1.34553802e+00 3.06445301e-01 6.30096555e-01 -1.60368495e-02 -9.21130955e-01 -8.42600882e-01 -5.64291239e-01 -7.92750180e-01 2.71956861e-01 2.02183485e-01 8.39968622e-02 -1.37436271e+00 1.59908807e+00 5.39124787e-01 6.95888758e-01 1.35304615e-01 1.00799954e+00 6.24039650e-01 7.67429173e-01 -2.25228682e-01 2.04150870e-01 1.18396246e+00 -8.91937494e-01 -3.28636229e-01 -2.09921896e-01 4.56339717e-01 -5.14225125e-01 6.72294080e-01 -2.92993281e-02 -5.21751463e-01 -6.99951708e-01 -1.19023597e+00 -1.00024911e-02 -8.74328613e-01 -1.28479317e-01 7.96088278e-01 2.84786075e-01 -1.08815217e+00 5.68429947e-01 -1.02106118e+00 -4.06911552e-01 1.93215996e-01 1.90470606e-01 -7.35116750e-02 2.39484668e-01 -1.15457797e+00 7.74460971e-01 3.26689214e-01 2.19900385e-01 -9.07332540e-01 -9.00155604e-01 -9.42642927e-01 -4.13959652e-01 2.49451622e-01 -5.12089729e-01 1.31693208e+00 -1.43435702e-01 -1.40880418e+00 1.15394294e+00 -8.54025707e-02 -8.81914318e-01 2.90234625e-01 -2.02742487e-01 -3.68490964e-01 1.55099913e-01 3.66551727e-01 1.28997719e+00 9.94990885e-01 -1.30779314e+00 -1.15337050e+00 -5.98809600e-01 8.19770470e-02 3.21479201e-01 2.48068571e-01 -5.72877944e-01 -7.53309667e-01 -2.15998590e-01 6.35611475e-01 -1.48612332e+00 -4.63213384e-01 -1.90389097e-01 -4.79625911e-01 -5.20513535e-01 1.39949965e+00 -1.42822862e-01 5.44996560e-01 -2.00727940e+00 9.37382057e-02 3.44443053e-01 4.77400929e-01 1.24589540e-01 -8.94360393e-02 -2.68210381e-01 -1.31801441e-02 1.95110124e-02 -6.84683263e-01 -2.23156989e-01 2.60025114e-01 5.96203029e-01 -6.96710467e-01 7.41308868e-01 3.54816258e-01 8.03755879e-01 -7.27737427e-01 -2.41119534e-01 7.13104904e-01 1.91123232e-01 -4.23880696e-01 -7.23019391e-02 -4.77547497e-01 4.85941857e-01 -5.46243429e-01 7.40520716e-01 1.15360606e+00 -7.02906847e-02 -3.69351208e-01 -2.17968628e-01 -6.22938037e-01 2.13492721e-01 -1.10261750e+00 1.61932290e+00 -2.96308309e-01 8.99527550e-01 -1.39017150e-01 -1.23452878e+00 1.10749960e+00 -2.72693813e-01 7.87926257e-01 -7.21092761e-01 3.16630155e-02 1.88522384e-01 -1.81491271e-01 -3.07293117e-01 9.06336665e-01 2.84776300e-01 -2.91229248e-01 2.68719107e-01 -1.49116307e-01 -9.48061347e-01 2.93619573e-01 5.12734354e-02 3.79573941e-01 2.54847378e-01 -2.60294467e-01 -5.67775428e-01 5.03663421e-01 3.73435944e-01 5.95651209e-01 8.58150482e-01 -4.47846383e-01 4.54970628e-01 2.28030249e-01 -7.13133991e-01 -9.08031762e-01 -1.27421498e+00 -8.01498592e-01 6.93022907e-01 7.34696090e-01 -9.64604914e-02 -6.28518760e-01 -7.35898018e-01 3.08169037e-01 7.88029075e-01 -5.66976786e-01 1.74020857e-01 -6.63199663e-01 -9.93930817e-01 4.43527877e-01 4.76322800e-01 8.88368428e-01 -4.07612979e-01 -6.53649271e-01 8.19938928e-02 1.06735444e-02 -1.41105723e+00 -7.95257539e-02 3.91191095e-01 -8.77566695e-01 -1.04467070e+00 -3.87693465e-01 -7.46595800e-01 2.25966707e-01 7.98125446e-01 1.06926537e+00 -5.98716497e-01 -2.46875554e-01 5.36892951e-01 -2.84393039e-02 -5.46445191e-01 4.37501855e-02 3.22594851e-01 2.78237522e-01 8.04831237e-02 6.58840239e-01 -5.14347136e-01 -6.78660035e-01 6.70978367e-01 -6.15997434e-01 4.87188622e-02 2.43192479e-01 3.26657027e-01 1.17167234e+00 -9.00100768e-02 -5.94953597e-02 -4.67019379e-01 -1.13113798e-01 -4.02502000e-01 -1.16525185e+00 -1.24524236e-01 -6.16411030e-01 9.49311331e-02 3.30109686e-01 8.29174966e-02 -7.47701406e-01 2.31397644e-01 -5.12467064e-02 -7.93651581e-01 -3.10020447e-01 1.69243231e-01 2.55765289e-01 -8.73292685e-02 4.14960533e-01 1.41250208e-01 -1.27271786e-01 -3.41196030e-01 7.90240228e-01 4.06341732e-01 7.51465917e-01 -4.31098014e-01 1.33095133e+00 1.08870947e+00 4.78143722e-01 -1.03849220e+00 -9.58252370e-01 -7.21217513e-01 -1.10283542e+00 -2.45929837e-01 1.24291408e+00 -1.14147210e+00 -6.34402990e-01 6.23441875e-01 -9.22934532e-01 -4.13655728e-01 -2.68030405e-01 6.23562157e-01 -5.95120251e-01 9.38364416e-02 -1.36513442e-01 -4.62477297e-01 -3.60034630e-02 -1.46895552e+00 1.29281735e+00 2.46563211e-01 2.75866479e-01 -9.03518319e-01 3.06023866e-01 1.67145997e-01 2.60414630e-01 3.00095469e-01 8.74935687e-01 -5.77809334e-01 -9.89865184e-01 3.11931595e-02 -4.20329720e-01 2.71675587e-01 -1.45313904e-01 3.77532572e-01 -1.05848861e+00 -8.53202939e-02 -1.16491935e-03 6.54196963e-02 1.35041273e+00 7.05931246e-01 1.26721954e+00 1.60335392e-01 -4.62832659e-01 1.13562977e+00 1.18127108e+00 1.17048137e-01 9.48565602e-02 4.11651701e-01 8.72742236e-01 9.24861491e-01 8.28306675e-01 2.95065325e-02 8.13463509e-01 6.65241838e-01 8.61729980e-01 -1.60620481e-01 1.04385704e-01 -1.95723288e-02 7.84465224e-02 7.14799464e-01 6.64867694e-03 -2.30609804e-01 -9.91669297e-01 6.54393435e-01 -1.96868694e+00 -6.29721344e-01 -5.40760756e-01 1.89113176e+00 4.18215916e-02 5.25085293e-02 -4.25428152e-02 -1.11430198e-01 6.92435980e-01 7.28784084e-01 -9.33468580e-01 -3.25021356e-01 -2.41162539e-01 1.42821774e-01 1.07600546e+00 5.01965463e-01 -1.76551759e+00 1.41281903e+00 6.26230764e+00 6.21134460e-01 -1.48286760e+00 -1.23906322e-02 4.90247071e-01 2.02667937e-01 -1.38777733e-01 4.82420884e-02 -1.23191750e+00 4.74840142e-02 8.19745421e-01 -5.67714870e-03 2.16416433e-01 9.76409793e-01 1.61228597e-01 2.51823038e-01 -9.80486870e-01 7.99273074e-01 -1.80387720e-01 -1.47178030e+00 3.11590075e-01 6.78974241e-02 8.87957036e-01 8.96750987e-01 4.23734516e-01 3.04361314e-01 4.96917516e-01 -7.29141176e-01 6.14388466e-01 2.15271428e-01 4.92857963e-01 -7.38664627e-01 3.64463508e-01 -8.23650509e-02 -1.61084855e+00 1.27949551e-01 -5.12875617e-01 2.28479162e-01 1.14512719e-01 3.88861090e-01 -7.30418622e-01 5.56916416e-01 1.09092903e+00 1.46365130e+00 -4.45638090e-01 8.21696520e-01 -2.37377480e-01 5.87512612e-01 -5.40546417e-01 3.13951552e-01 1.02460408e+00 -7.27534533e-01 1.00946450e+00 1.06170642e+00 2.60782093e-01 -1.99541658e-01 2.66853243e-01 9.92087483e-01 1.35476759e-03 -1.30814880e-01 -8.69436443e-01 3.21475655e-01 2.43933678e-01 1.45287585e+00 -7.28564799e-01 -3.41320068e-01 -2.54965693e-01 2.78309405e-01 -1.86370954e-01 5.02002716e-01 -9.49870408e-01 -3.50940526e-01 1.31691253e+00 -2.11480930e-01 4.83772963e-01 -7.28428364e-01 -4.44932371e-01 -1.08522058e+00 -3.48040462e-01 -3.34981918e-01 2.86018431e-01 -5.77543437e-01 -1.00182009e+00 3.93176675e-01 4.34902251e-01 -1.46531427e+00 -1.94939032e-01 -9.42186177e-01 -7.01098561e-01 4.54678029e-01 -2.04103804e+00 -1.06113613e+00 -5.53615570e-01 5.64131558e-01 7.04880476e-01 -7.09365904e-02 3.12168747e-01 -8.56607314e-03 -7.22466171e-01 1.40059605e-01 2.46183336e-01 -7.84015879e-02 5.05253136e-01 -1.46729040e+00 8.74994874e-01 1.06895268e+00 2.95649350e-01 2.49806657e-01 5.57658553e-01 -2.86906391e-01 -1.38719606e+00 -1.74465322e+00 5.59772313e-01 -6.93373859e-01 9.51649725e-01 -3.49432081e-01 -1.01088154e+00 8.52262199e-01 -1.09974265e-01 -3.27983052e-02 5.11013642e-02 -1.06999695e-01 -3.61305326e-01 -5.58856606e-01 -7.80718505e-01 7.88702607e-01 1.17067206e+00 -3.91656607e-01 -6.19127393e-01 5.22452891e-01 1.17501187e+00 -5.95622301e-01 -7.50625491e-01 9.20217812e-01 2.52278239e-01 -7.71729946e-01 1.16935062e+00 -3.35331649e-01 7.29465261e-02 -7.79229581e-01 -3.87482196e-01 -1.21613276e+00 -3.16804111e-01 -4.81513262e-01 2.32108548e-01 6.77032530e-01 4.75193501e-01 -8.52070332e-01 6.83390558e-01 2.89023072e-02 -6.43632472e-01 -3.02713960e-01 -1.10770023e+00 -1.03649020e+00 3.39672416e-01 -9.69923019e-01 8.60349894e-01 8.74456346e-01 -8.66038024e-01 2.86659539e-01 -1.10639945e-01 7.40943253e-01 9.04773593e-01 6.51414096e-01 1.02347600e+00 -1.57753766e+00 4.00220543e-01 -6.33253276e-01 -4.68195856e-01 -1.49785209e+00 6.31508231e-01 -1.25176024e+00 2.52226770e-01 -1.23617625e+00 -4.38005984e-01 -7.26016700e-01 -2.49387771e-01 3.33553821e-01 2.41745561e-01 3.37037593e-01 -8.59589055e-02 4.29508030e-01 -5.74103296e-01 5.87415576e-01 1.17432749e+00 -5.82774282e-01 -5.06290913e-01 2.63462722e-01 -1.29098758e-01 7.52072275e-01 8.71845484e-01 -2.96472907e-01 -4.95172411e-01 -6.08723819e-01 1.45614948e-02 -4.52386409e-01 5.62277913e-01 -1.10627401e+00 2.21037477e-01 -2.65385211e-01 -4.94165048e-02 -1.40826678e+00 5.06215572e-01 -6.71520352e-01 -1.97550088e-01 3.96640241e-01 -1.74127221e-02 1.02359995e-01 3.44381720e-01 6.70120060e-01 -3.39921355e-01 2.18035415e-01 9.51042116e-01 1.60223432e-02 -1.30942345e+00 9.43202138e-01 -3.68986189e-01 5.26035540e-02 9.83096719e-01 -1.71905816e-01 -4.23743606e-01 7.03539774e-02 -4.11447406e-01 6.24725223e-01 3.14988226e-01 8.23264360e-01 3.97154838e-01 -1.00402546e+00 -6.98063433e-01 3.95819932e-01 4.04904068e-01 5.05191803e-01 1.92981839e-01 9.60046291e-01 -6.93367183e-01 7.81404316e-01 -1.33355409e-01 -1.65629756e+00 -7.92534709e-01 2.20657229e-01 6.60791576e-01 8.04118067e-02 -9.04573381e-01 5.36678255e-01 6.24971330e-01 -9.44634259e-01 3.99552546e-02 -7.23031580e-01 -3.86697143e-01 2.62695134e-01 5.64295501e-02 3.01473618e-01 1.17747791e-01 -8.91226232e-01 -4.42063898e-01 1.21547198e+00 -3.32473107e-02 -4.32901978e-02 1.23737788e+00 -1.51636630e-01 -9.79332905e-03 5.52080214e-01 1.23175216e+00 -5.78404963e-01 -1.54699421e+00 -4.28031921e-01 -1.55814365e-01 -1.62679076e-01 2.32089683e-01 -1.00668244e-01 -1.29399538e+00 1.25477815e+00 8.31075490e-01 2.82596380e-01 6.11743212e-01 2.56313174e-03 8.13935518e-01 8.65999103e-01 7.88320750e-02 -1.12897623e+00 -3.79381388e-01 1.09997571e+00 6.46184206e-01 -1.37520242e+00 -1.90649197e-01 -2.63797790e-01 -4.70146000e-01 1.03787649e+00 6.17137849e-01 -3.49131107e-01 1.00230086e+00 -2.54703462e-01 2.89767891e-01 -3.64328027e-01 -4.97907311e-01 -5.21152854e-01 5.03932714e-01 5.40566862e-01 -3.70981961e-01 3.99924964e-01 3.12325150e-01 -2.09981889e-01 -4.01063800e-01 -7.70511389e-01 2.06674382e-01 4.88805145e-01 -5.51082492e-01 -5.61117470e-01 -1.54608056e-01 3.05898368e-01 2.04702184e-01 2.21502259e-01 -1.38269812e-01 1.23905838e+00 2.76886493e-01 8.44552517e-01 7.83871233e-01 -6.37114406e-01 4.16010171e-01 -2.59996857e-03 -2.26269457e-02 -2.00502574e-01 -9.77572650e-02 -1.19111858e-01 -2.49667853e-01 -6.65418565e-01 -7.47274995e-01 -7.86455631e-01 -1.27368379e+00 -2.79311806e-01 1.11645654e-01 5.62807824e-03 1.13121831e+00 9.81267810e-01 4.31420803e-01 2.45225221e-01 1.10764456e+00 -1.14117098e+00 -2.77321488e-01 -5.42655170e-01 -5.61819375e-01 7.80036226e-02 5.22961199e-01 -8.83031011e-01 -7.19911516e-01 -2.62076765e-01]	[8.103056907653809, -2.7090184688568115]
68db2228-d8db-4824-aaf8-4f82896b891a	co-saliency-detection-with-co-attention-fully	2008.08909	null	https://arxiv.org/abs/2008.08909v1	https://arxiv.org/pdf/2008.08909v1.pdf	Co-Saliency Detection with Co-Attention Fully Convolutional Network	Co-saliency detection aims to detect common salient objects from a group of relevant images. Some attempts have been made with the Fully Convolutional Network (FCN) framework and achieve satisfactory detection results. However, due to stacking convolution layers and pooling operation, the boundary details tend to be lost. In addition, existing models often utilize the extracted features without discrimination, leading to redundancy in representation since actually not all features are helpful to the final prediction and some even bring distraction. In this paper, we propose a co-attention module embedded FCN framework, called as Co-Attention FCN (CA-FCN). Specifically, the co-attention module is plugged into the high-level convolution layers of FCN, which can assign larger attention weights on the common salient objects and smaller ones on the background and uncommon distractors to boost final detection performance. Extensive experiments on three popular co-saliency benchmark datasets demonstrate the superiority of the proposed CA-FCN, which outperforms state-of-the-arts in most cases. Besides, the effectiveness of our new co-attention module is also validated with ablation studies.	['Guangshuai Gao', 'Qingjie Liu', 'Yunhong Wang', 'Wenting Zhao']	2020-08-20	null	null	null	null	['co-saliency-detection']	['computer-vision']	[ 1.58067599e-01 -5.72028756e-02 2.23499741e-02 2.12610010e-02 -2.76658326e-01 2.16647372e-01 5.19276977e-01 6.13250136e-02 -3.05954784e-01 4.84466344e-01 3.99835646e-01 2.58649886e-01 1.79673299e-01 -5.21747530e-01 -6.63288653e-01 -5.73466003e-01 1.09763779e-01 -4.78035212e-01 8.90675068e-01 -1.67403862e-01 3.82068217e-01 2.19017237e-01 -1.86046541e+00 3.11130077e-01 1.15405631e+00 1.08521211e+00 7.92441905e-01 1.34472623e-01 -1.23922657e-02 9.14275646e-01 -4.55257416e-01 -1.94502398e-01 1.65227890e-01 -1.93900630e-01 -4.92209494e-01 -1.33645266e-01 2.77091712e-01 -4.19479519e-01 -3.41635495e-01 1.23518312e+00 5.07465780e-01 7.48625621e-02 3.10568482e-01 -1.27873397e+00 -9.18527603e-01 3.85434479e-01 -1.00657868e+00 7.35847890e-01 9.96376947e-02 1.22200668e-01 1.04130733e+00 -1.28663170e+00 1.62338421e-01 1.32802939e+00 4.36508596e-01 2.67558098e-01 -8.38447869e-01 -6.74408793e-01 3.89560938e-01 5.28021455e-01 -1.37687314e+00 -2.58542329e-01 9.10035610e-01 -3.54593918e-02 7.60281086e-01 2.57194251e-01 6.61625326e-01 7.05838323e-01 2.37369370e-02 1.38862717e+00 8.65844786e-01 -3.37709606e-01 -1.80904046e-01 1.62978515e-01 1.65552143e-02 5.14366686e-01 4.63338226e-01 2.65085278e-03 -4.40947384e-01 1.13226674e-01 7.33410001e-01 5.29020727e-01 -4.32939827e-01 -2.82145500e-01 -1.19548905e+00 4.93393898e-01 1.12545455e+00 3.53196353e-01 -5.55137157e-01 1.17183244e-02 3.11625481e-01 -2.89150745e-01 4.22856271e-01 3.41659039e-01 -1.65387988e-01 4.42125946e-01 -8.94240081e-01 1.39126584e-01 -1.07313566e-01 9.51840401e-01 7.67630160e-01 -8.58119130e-03 -7.48087049e-01 7.50779808e-01 1.13658562e-01 1.46001443e-01 7.22396910e-01 -4.00875390e-01 3.07195097e-01 9.56753731e-01 1.46214679e-01 -1.31468689e+00 -3.85279655e-01 -1.02296364e+00 -6.75301552e-01 1.24993697e-01 -1.04581140e-01 -9.98420194e-02 -8.02734375e-01 1.61778831e+00 2.56480396e-01 4.21602309e-01 -4.83328328e-02 1.34005451e+00 9.71074939e-01 3.89201909e-01 3.29954386e-01 6.21047094e-02 1.24194777e+00 -1.46371055e+00 -6.62315011e-01 -5.62874138e-01 2.93703765e-01 -9.19662893e-01 1.07332337e+00 -1.66140601e-01 -1.05274189e+00 -9.47884560e-01 -1.10777009e+00 -2.82589972e-01 -3.10876578e-01 5.36124885e-01 7.34623909e-01 -9.29559115e-03 -9.04652297e-01 3.30125570e-01 -5.01041412e-01 -2.20813274e-01 8.41486037e-01 2.51792341e-01 -4.16057780e-02 -3.29688378e-02 -1.14788711e+00 8.82810771e-01 5.08059680e-01 2.89472848e-01 -9.55560863e-01 -6.74921632e-01 -6.77580237e-01 5.09176910e-01 5.35234153e-01 -5.06634772e-01 1.06747282e+00 -1.26178467e+00 -9.24801290e-01 4.36009943e-01 -2.68385082e-01 -4.48138773e-01 4.33646291e-01 -5.92703879e-01 -4.35737878e-01 8.93062875e-02 3.39429587e-01 9.76789594e-01 9.06850278e-01 -1.02627337e+00 -1.07043946e+00 -1.19712330e-01 2.32804820e-01 3.56513232e-01 -4.79781091e-01 2.03246012e-01 -4.71537113e-01 -8.07516754e-01 1.14927493e-01 -5.30391276e-01 -1.26036301e-01 -3.50546627e-03 -4.04289186e-01 -2.56930619e-01 1.21331346e+00 -5.22008538e-01 1.42424309e+00 -2.43590975e+00 -1.43114990e-02 -3.81340832e-01 4.48649615e-01 6.08228087e-01 -8.84933248e-02 -2.75313873e-02 -1.41575620e-01 -1.14629984e-01 9.43174027e-03 -2.16904670e-01 -2.98237264e-01 -2.54944801e-01 -2.40219578e-01 3.69394720e-01 7.23148882e-01 1.02710986e+00 -1.05774307e+00 -6.21428132e-01 3.13335210e-01 4.23013896e-01 -4.61956739e-01 1.20707259e-01 4.13948111e-02 4.95600179e-02 -5.80951273e-01 7.32036233e-01 8.28284562e-01 -3.82640034e-01 -3.97525519e-01 -2.89124578e-01 -3.81989539e-01 1.32799953e-01 -9.02332604e-01 1.45594978e+00 -1.69814482e-01 6.22547030e-01 1.27811478e-02 -8.32998693e-01 7.89915621e-01 -4.12949212e-02 9.02674347e-02 -8.60800743e-01 2.59945542e-01 1.83033735e-01 2.79390961e-01 -3.48968506e-01 6.44911408e-01 2.03873321e-01 2.72912592e-01 3.39509435e-02 4.26216833e-02 6.50111675e-01 -1.01923812e-02 2.82707721e-01 6.70159638e-01 -6.11082418e-03 3.43171388e-01 -4.86385375e-01 8.57399046e-01 -1.41818285e-01 7.73300827e-01 5.77768028e-01 -6.12160385e-01 7.31826007e-01 3.38168442e-01 -2.05389738e-01 -7.66718984e-01 -6.27331436e-01 8.88250321e-02 1.25437355e+00 6.79692149e-01 -1.78653792e-01 -7.02380836e-01 -6.99216485e-01 -2.17459518e-02 4.88203764e-01 -9.05436456e-01 -4.93144602e-01 -3.24877769e-01 -6.48051202e-01 1.43880531e-01 8.75899434e-01 9.14185226e-01 -1.32928395e+00 -1.00533569e+00 1.40570447e-01 -1.03741750e-01 -9.69745517e-01 -7.02179492e-01 -1.88387737e-01 -5.98055542e-01 -1.07379699e+00 -9.86846149e-01 -1.00523019e+00 6.36160254e-01 1.19111741e+00 7.34573007e-01 3.65847200e-01 -4.27365869e-01 -2.26626322e-01 -4.03060377e-01 -8.55152249e-01 3.52891147e-01 2.87850592e-02 -2.05751285e-01 3.29342782e-01 5.57726800e-01 -2.99450576e-01 -9.69248950e-01 3.65042448e-01 -7.19860077e-01 3.16837311e-01 9.85316157e-01 8.52057457e-01 4.32652265e-01 -9.42889303e-02 6.96882665e-01 -5.59809387e-01 4.25900191e-01 -4.44586098e-01 -3.41537893e-01 2.19750479e-01 -2.51544952e-01 -1.63389236e-01 4.86553252e-01 -2.92718351e-01 -1.14096630e+00 -5.62478378e-02 2.67447144e-01 -8.02743912e-01 8.36765096e-02 2.38783345e-01 -3.14628810e-01 -1.56185836e-01 3.49466264e-01 3.69595528e-01 -1.46402523e-01 -4.12870318e-01 6.83362707e-02 5.77873588e-01 5.43851256e-01 -1.10240811e-02 4.33384210e-01 5.00323176e-01 -2.44032785e-01 -5.42209625e-01 -1.21182561e+00 -5.80654621e-01 -4.82647210e-01 -1.39537498e-01 6.84513748e-01 -1.20993650e+00 -2.76662618e-01 5.28679729e-01 -1.06746614e+00 1.44375801e-01 -8.08016285e-02 4.25565749e-01 -1.17639065e-01 1.94027871e-01 -3.30582201e-01 -8.09651256e-01 -5.31795681e-01 -1.22631204e+00 1.07721531e+00 8.24171901e-01 2.13891193e-01 -5.55066586e-01 -5.08444607e-01 3.28946300e-02 7.01044559e-01 8.93979594e-02 5.52386165e-01 -5.04153490e-01 -5.70101678e-01 -1.60923764e-01 -9.90850329e-01 2.44808286e-01 2.24702686e-01 -7.68281147e-02 -1.10964847e+00 -2.77191401e-01 -2.06276104e-01 -6.00995794e-02 1.19891274e+00 5.24805248e-01 1.34376085e+00 -1.37228426e-02 -6.38468444e-01 3.64402950e-01 1.43322635e+00 -1.11110687e-01 6.55152202e-01 3.88709158e-01 7.28255749e-01 4.54412699e-01 8.48169208e-01 2.46996909e-01 2.85798848e-01 5.77587545e-01 7.39280641e-01 -5.25469184e-01 -3.88573319e-01 -2.41490766e-01 1.14173368e-01 3.77553791e-01 -7.18926564e-02 1.61351353e-01 -5.32805562e-01 9.75562096e-01 -1.91284680e+00 -8.86152685e-01 -1.85305223e-01 2.01334834e+00 4.83386457e-01 3.59198362e-01 1.74075380e-01 2.01933216e-02 1.20433724e+00 2.80050814e-01 -6.00532115e-01 3.52251790e-02 -3.92064869e-01 1.24903424e-02 3.16803932e-01 5.90332747e-02 -1.38282359e+00 9.78143156e-01 5.68647146e+00 9.90001142e-01 -1.29380512e+00 3.87725592e-01 5.53528309e-01 -2.53610224e-01 9.91093963e-02 -4.99800481e-02 -7.83072174e-01 6.99789166e-01 2.08944634e-01 -3.96777481e-01 4.05054353e-02 1.15486979e+00 5.31743951e-02 -2.45344967e-01 -5.88615835e-01 8.41412961e-01 1.13852739e-01 -1.12705874e+00 1.11175634e-01 -1.69825792e-01 8.41420710e-01 -3.62958424e-02 1.84221819e-01 5.03124893e-01 -1.28237560e-01 -7.03132570e-01 8.61684144e-01 3.93606961e-01 5.39387882e-01 -9.07290578e-01 9.33516204e-01 2.25304604e-01 -1.31561744e+00 -4.68789428e-01 -8.26230466e-01 -2.79173285e-01 -1.41595796e-01 6.55189693e-01 -4.64570731e-01 5.67491353e-01 9.20239151e-01 9.24750209e-01 -9.47145343e-01 1.48233211e+00 -2.20751092e-01 3.08001339e-01 -6.03621341e-02 -1.42609343e-01 5.10536015e-01 3.08326095e-01 5.11932731e-01 1.23100376e+00 2.02583700e-01 4.55761887e-02 1.63765863e-01 9.59812999e-01 -4.58720140e-02 1.66100278e-01 -1.71531424e-01 3.11798066e-01 2.61466682e-01 1.50582421e+00 -6.37064159e-01 -4.44344252e-01 -6.67465746e-01 9.38141406e-01 5.16075671e-01 2.86766678e-01 -9.14273620e-01 -6.66965067e-01 6.36433601e-01 8.96341428e-02 7.54501820e-01 2.84587890e-01 -1.84714153e-01 -1.10665190e+00 1.65092095e-03 -4.89788115e-01 2.74185419e-01 -8.95788789e-01 -9.75897372e-01 4.79180336e-01 -2.75475055e-01 -1.31133187e+00 5.50832689e-01 -4.09639895e-01 -9.29524004e-01 1.01153827e+00 -1.89491713e+00 -1.17389011e+00 -5.61741471e-01 4.77126658e-01 7.48325348e-01 -5.55154756e-02 2.12803498e-01 4.01253641e-01 -9.20317650e-01 5.92102826e-01 -1.94500014e-01 1.06471762e-01 6.34056568e-01 -9.70110714e-01 1.22201934e-01 1.18589044e+00 -2.35177159e-01 6.90016806e-01 5.42227864e-01 -5.86670697e-01 -8.36540878e-01 -1.42399871e+00 6.46545708e-01 -2.70887140e-05 3.70418578e-01 -1.95503786e-01 -1.13744617e+00 4.39569324e-01 2.77124643e-01 4.60709184e-01 2.37271950e-01 -1.55458957e-01 -1.85807228e-01 -6.29976019e-02 -8.73917580e-01 6.34547770e-01 1.04236329e+00 -3.69806021e-01 -7.47968972e-01 8.36046040e-03 8.20532858e-01 -2.24902138e-01 -1.12505011e-01 5.74022412e-01 4.70856637e-01 -1.24163508e+00 1.01798928e+00 -4.76466030e-01 6.84538841e-01 -6.04985714e-01 6.17510118e-02 -1.17922854e+00 -8.74750733e-01 -1.07996784e-01 -3.32719982e-01 1.19482708e+00 1.11703001e-01 -4.05582249e-01 4.29752767e-01 1.47559196e-01 -3.52231503e-01 -8.75361025e-01 -5.74471831e-01 -4.64619666e-01 -3.79189640e-01 -2.38368846e-02 7.58733451e-01 7.32069612e-01 -1.71856180e-01 3.82463515e-01 -4.03211176e-01 3.20413113e-01 4.78659064e-01 3.25799614e-01 4.33607101e-01 -1.17827594e+00 -6.44864561e-03 -7.59295583e-01 -5.17012417e-01 -1.04926479e+00 -1.48696780e-01 -5.55037439e-01 1.22621059e-01 -1.37326729e+00 5.75693727e-01 -2.92571276e-01 -8.55024338e-01 4.70688343e-01 -9.75905120e-01 2.91253805e-01 3.09621513e-01 2.64561653e-01 -1.00037801e+00 7.70904660e-01 1.48565328e+00 -1.68500975e-01 -1.02383099e-01 -2.46378392e-01 -1.04405797e+00 7.03672290e-01 7.35029519e-01 -1.37383148e-01 -3.29352051e-01 -5.02293766e-01 -3.93275768e-01 -5.40735602e-01 6.49639130e-01 -1.35749304e+00 3.23334008e-01 4.55153221e-03 8.07603717e-01 -7.67209053e-01 5.90200126e-02 -6.00851059e-01 -3.37308705e-01 4.97941911e-01 -1.94092825e-01 -2.20356390e-01 4.53969747e-01 6.25472248e-01 -3.64686340e-01 -2.36230008e-02 8.37910354e-01 -9.26322937e-02 -1.04517210e+00 1.94780007e-01 3.06469593e-02 -1.37643009e-01 1.22065139e+00 -9.18690786e-02 -3.85058820e-01 -1.02711491e-01 -3.62865835e-01 4.03580189e-01 3.47519815e-01 6.59753501e-01 6.64243996e-01 -1.36757445e+00 -6.25573099e-01 3.80317897e-01 3.20179075e-01 5.29284216e-02 6.01106107e-01 1.07365167e+00 -1.41343296e-01 6.37479782e-01 -4.76376534e-01 -5.59319913e-01 -1.10034597e+00 1.06971049e+00 2.19658345e-01 6.40981123e-02 -4.93763030e-01 9.77404118e-01 8.70864570e-01 1.66911840e-01 1.51686430e-01 -3.41454774e-01 -5.14461100e-01 -7.95978457e-02 9.14072335e-01 3.45302910e-01 -1.37709141e-01 -8.48132014e-01 -5.20912766e-01 2.87417680e-01 -4.15106356e-01 5.82144737e-01 1.12030351e+00 -2.18259916e-01 -2.69968677e-02 6.99458364e-03 8.84918988e-01 -1.07794940e-01 -1.50117195e+00 -4.91659462e-01 -1.76548153e-01 -6.51936650e-01 3.58765841e-01 -5.22642910e-01 -1.31422293e+00 1.13639164e+00 6.61693573e-01 2.01214049e-02 1.29967046e+00 -8.25414881e-02 6.22639656e-01 -1.23511560e-01 1.76259354e-01 -8.86918068e-01 2.20350608e-01 1.93902612e-01 1.01802766e+00 -1.32074130e+00 -3.57230827e-02 -4.93398219e-01 -7.36243904e-01 6.63119316e-01 1.13973689e+00 -4.88278598e-01 5.17192543e-01 -1.81590691e-01 -2.65759885e-01 -1.08106323e-01 -5.99552691e-01 -6.60788059e-01 3.80132616e-01 3.11848670e-01 2.96784818e-01 -2.84455679e-02 -4.36288774e-01 8.28334630e-01 3.03621709e-01 -2.25317180e-02 3.87602597e-01 8.67190421e-01 -7.80934334e-01 -4.99933511e-01 -3.08434516e-01 5.47481835e-01 -5.68531692e-01 -2.02500343e-01 -2.79759824e-01 6.25629723e-01 4.15075213e-01 9.00280297e-01 1.40152082e-01 -3.82413089e-01 3.44687194e-01 -2.88440019e-01 -1.33871036e-02 -6.10556066e-01 -5.58200479e-01 1.87947378e-01 -4.76775318e-01 -4.94280607e-01 -5.15018523e-01 -6.73983514e-01 -1.13347149e+00 8.32067356e-02 -7.13163793e-01 1.00263461e-01 1.59784213e-01 7.87340939e-01 6.74602866e-01 9.78906512e-01 5.29676139e-01 -1.10105634e+00 -2.17336431e-01 -1.29137135e+00 -4.87538338e-01 3.51309240e-01 5.43288529e-01 -1.17049229e+00 -1.02006637e-01 -2.88744599e-01]	[9.715469360351562, -0.400240957736969]
bde31ea4-d20a-4b32-9b97-d79edd4cd9fe	stochastic-online-convex-optimization	2102.00729	null	https://arxiv.org/abs/2102.00729v3	https://arxiv.org/pdf/2102.00729v3.pdf	Stochastic Online Convex Optimization. Application to probabilistic time series forecasting	We introduce a general framework of stochastic online convex optimization to obtain fast-rate stochastic regret bounds. We prove that algorithms such as online newton steps and a scale-free 10 version of Bernstein online aggregation achieve best-known rates in unbounded stochastic settings. We apply our approach to calibrate parametric probabilistic forecasters of non-stationary sub-gaussian time series. Our fast-rate stochastic regret bounds are any-time valid. Our proofs combine self-bounded and Poissonnian inequalities for martingales and sub-gaussian random variables, respectively, under a stochastic exp-concavity assumption.	['Olivier Wintenberger']	2021-02-01	null	null	null	null	['probabilistic-time-series-forecasting']	['time-series']	[-5.59472919e-01 7.90104344e-02 -4.91117388e-02 -6.31641567e-01 -1.49849200e+00 -1.01927626e+00 -2.15734705e-01 -2.73737404e-02 -5.16660810e-01 1.20238864e+00 6.99719489e-02 -6.95560396e-01 -4.70809668e-01 -6.74791753e-01 -1.12211096e+00 -7.52402842e-01 -6.21305048e-01 8.30473065e-01 -1.63854614e-01 1.32009625e-01 3.81517559e-02 6.71630204e-01 -8.85664344e-01 -1.65573686e-01 7.66806841e-01 1.53134727e+00 -3.02940249e-01 1.32226455e+00 -4.79173921e-02 4.55574185e-01 -4.03827339e-01 -8.93265843e-01 7.69598365e-01 -7.61576295e-02 -4.78629321e-01 -2.02301025e-01 2.58513421e-01 -5.14988065e-01 -6.34010062e-02 1.11434293e+00 5.61918616e-01 3.45027536e-01 5.94364107e-01 -1.23652458e+00 -6.77999318e-01 9.23718929e-01 -5.86555362e-01 4.09405410e-01 2.68393427e-01 -2.66151398e-01 1.31467938e+00 -7.12513745e-01 3.73926401e-01 1.29846072e+00 9.56874371e-01 2.97167361e-01 -9.82367456e-01 -5.41103184e-01 3.61552209e-01 -2.32451379e-01 -1.20187032e+00 -2.37235844e-01 2.19354600e-01 -2.33705580e-01 4.60781932e-01 7.91865587e-01 4.93181974e-01 6.38669312e-01 2.54714310e-01 9.05220151e-01 1.14778078e+00 -8.22724774e-02 5.98211527e-01 6.40645921e-02 -1.02281861e-01 5.97458780e-01 4.61500466e-01 2.04718322e-01 -3.96185070e-01 -7.35491633e-01 6.86132729e-01 2.67857581e-01 -4.27934937e-02 2.16840412e-02 -7.23289490e-01 9.47755456e-01 3.80917825e-02 -6.45028800e-02 -5.98468721e-01 5.98749816e-01 2.95453340e-01 5.03770530e-01 9.61130440e-01 5.28446138e-02 -9.11359549e-01 -4.08106983e-01 -9.98616517e-01 5.11272490e-01 1.56504953e+00 1.33651912e+00 2.21487463e-01 1.12546310e-01 -6.22839391e-01 1.86650425e-01 1.53821617e-01 1.41527963e+00 -3.51469070e-01 -1.28120995e+00 7.02231526e-01 -6.91522956e-01 9.59693253e-01 -7.93433726e-01 -2.09782749e-01 -8.33386481e-01 -5.16790867e-01 -9.68472064e-02 7.95922220e-01 -7.59441316e-01 -2.77277499e-01 1.64866590e+00 4.71707106e-01 1.64112989e-02 -1.96865320e-01 8.55450928e-01 -3.87570024e-01 6.24108732e-01 -5.64152598e-01 -7.63915002e-01 6.28802836e-01 -5.59554458e-01 -7.08441257e-01 3.65474761e-01 3.39334726e-01 -5.30639648e-01 8.21135283e-01 5.19209921e-01 -1.45704293e+00 3.39129895e-01 -4.10016388e-01 4.17875983e-02 -6.79054260e-02 -4.98342305e-01 9.37627077e-01 9.41727102e-01 -7.86099732e-01 8.82325292e-01 -8.35346699e-01 5.03315806e-01 4.72316206e-01 7.57005066e-02 1.28557980e-01 2.22603947e-01 -6.72310948e-01 2.82682717e-01 -5.41584753e-02 3.04869324e-01 -1.11809659e+00 -1.06335783e+00 -3.48682851e-01 1.08939156e-01 6.01154268e-01 -8.65846097e-01 1.41511405e+00 -6.09504044e-01 -1.47804415e+00 4.41777945e-01 -3.91184539e-01 -9.64028299e-01 1.18808126e+00 -3.28183621e-01 1.08383127e-01 3.11319903e-02 -9.74850133e-02 -6.12720966e-01 7.02717006e-01 -9.66784477e-01 -7.76039362e-01 -5.08354068e-01 4.23927844e-01 2.31266856e-01 -1.10216208e-01 1.89962089e-01 2.76001692e-01 -3.96453530e-01 -1.25977397e-01 -7.58217275e-01 -6.75376713e-01 3.47832739e-01 -2.00048894e-01 1.93382248e-01 -1.27195105e-01 -7.31600881e-01 1.04798412e+00 -1.77756619e+00 -3.05551589e-01 3.17388177e-01 7.80693218e-02 -6.73416555e-01 3.20025593e-01 4.32497263e-01 4.94865716e-01 4.74770725e-01 4.23891991e-02 -5.73482037e-01 6.67827308e-01 2.00886920e-01 -8.10423791e-01 8.90097678e-01 -6.63276434e-01 6.83936000e-01 -1.10824704e+00 -1.80481449e-01 -1.72254384e-01 -1.49457827e-01 -5.66821754e-01 6.11131303e-02 -4.56288069e-01 6.95628300e-02 -5.26728272e-01 6.03105962e-01 9.79106963e-01 -3.25609595e-01 -3.81443083e-01 3.47275615e-01 -4.29270156e-02 7.11955363e-03 -1.29195774e+00 1.10815334e+00 -1.03360856e+00 3.13249975e-01 6.91049457e-01 -5.59414208e-01 2.95253605e-01 2.88160563e-01 4.48964536e-01 1.63259894e-01 3.46869230e-01 5.15527606e-01 -7.34570622e-01 -1.35911882e-01 5.37361801e-01 -7.21464276e-01 -2.05666840e-01 6.42066956e-01 -2.71767437e-01 -1.47206873e-01 1.09562650e-01 1.95792794e-01 7.99527526e-01 -2.31784895e-01 8.43472220e-03 -4.57757205e-01 1.02614351e-01 -4.49960470e-01 3.73649925e-01 1.50334644e+00 -1.77085578e-01 3.30847800e-01 6.30377948e-01 -2.76092768e-01 -1.18060958e+00 -1.32797325e+00 1.12504400e-01 1.63857853e+00 -2.32210606e-01 -2.37063561e-02 -2.68877417e-01 -5.35588682e-01 6.33580327e-01 1.04863250e+00 -9.26310420e-01 5.55876851e-01 1.89656958e-01 -6.04489923e-01 1.17779106e-01 5.24558425e-01 -1.91597492e-01 -1.73423603e-01 -1.16230063e-01 4.07648563e-01 2.81157017e-01 -1.09475958e+00 -1.01476049e+00 1.10656641e-01 -8.47009838e-01 -7.50262260e-01 -1.00635397e+00 9.97939482e-02 5.13558805e-01 2.31535718e-01 9.73718822e-01 -6.08083129e-01 1.77341238e-01 8.56875837e-01 2.66741607e-02 -9.27294970e-01 -4.86107208e-02 -4.20340985e-01 1.12980619e-01 1.75800741e-01 -2.00078666e-01 -5.56625485e-01 -6.64210677e-01 7.74166137e-02 -5.47438145e-01 -5.53972125e-01 1.04508929e-01 6.72227740e-01 9.41983461e-01 -1.29498586e-01 2.18950257e-01 -8.47815812e-01 9.13564265e-01 -6.45159185e-01 -1.43557465e+00 3.99220288e-01 -5.81124365e-01 9.83364657e-02 9.52908099e-01 -5.36589444e-01 -8.87080491e-01 -2.69269764e-01 1.85605243e-01 -5.85100055e-01 5.87394238e-01 6.14481032e-01 2.47710496e-01 -2.52117813e-01 5.82200468e-01 3.15830201e-01 -3.67782623e-01 -2.29487196e-01 6.66869342e-01 6.03697062e-01 6.66833043e-01 -1.19966590e+00 9.09459233e-01 1.17490220e+00 2.82760918e-01 -2.13359222e-01 -1.69770253e+00 -5.35904944e-01 4.98424955e-02 -1.89158648e-01 3.00378829e-01 -6.93201780e-01 -1.24747014e+00 8.92302543e-02 -1.15168321e+00 -4.48862165e-01 -8.71029079e-01 4.54346508e-01 -1.06410515e+00 1.57784298e-01 -6.45290434e-01 -2.17957735e+00 -5.87881684e-01 -3.32699269e-01 1.00718594e+00 1.74195468e-01 5.15634716e-01 -1.33885896e+00 4.82759923e-02 1.70718119e-01 7.92535186e-01 4.75870073e-01 -2.06713274e-01 -6.49581313e-01 -6.59179568e-01 -3.94226849e-01 -2.71177918e-01 4.18087810e-01 -4.74675596e-01 6.86918348e-02 -5.72536647e-01 -2.81424403e-01 3.41730088e-01 -2.44036708e-02 4.72916901e-01 7.18097746e-01 1.51440251e+00 -1.09972572e+00 -6.87835813e-02 9.92324591e-01 1.42177868e+00 -4.49856251e-01 -1.71502121e-02 1.84839889e-01 1.71147108e-01 -3.70455533e-02 6.62865758e-01 1.45752990e+00 3.33918780e-01 -1.70034662e-01 3.31788272e-01 4.71800774e-01 8.89790595e-01 -2.71760255e-01 3.31558108e-01 6.73782885e-01 -1.65424034e-01 -1.82607442e-01 -4.80195761e-01 7.75356412e-01 -1.81041265e+00 -1.16516757e+00 -1.67265043e-01 2.51622224e+00 1.08578753e+00 2.01649666e-01 3.50995213e-01 -4.04025316e-01 5.98073363e-01 -1.63762584e-01 -7.99644053e-01 -6.25074744e-01 -4.60054636e-01 2.75454432e-01 1.71115685e+00 7.53788173e-01 -8.33100438e-01 4.47413683e-01 7.70530987e+00 1.09371126e+00 -4.79841828e-01 6.50194764e-01 6.66685164e-01 -6.72566473e-01 -6.74244106e-01 5.50550297e-02 -8.14015031e-01 6.48757219e-01 1.28990746e+00 -1.10291970e+00 8.59467208e-01 1.28784180e+00 3.40319484e-01 1.12836346e-01 -9.00534570e-01 9.50799584e-01 -4.29569751e-01 -1.42961943e+00 -6.23011231e-01 2.46782824e-01 1.22660851e+00 4.55402844e-02 1.34750724e-01 9.52694789e-02 1.02957964e+00 -8.46289277e-01 8.65943193e-01 9.98175800e-01 5.35299718e-01 -1.04527032e+00 8.88853669e-01 1.85051605e-01 -9.28018212e-01 -3.59861672e-01 -5.77305019e-01 -1.30731419e-01 7.11508751e-01 1.20880413e+00 -2.64246404e-01 5.08968174e-01 4.65495855e-01 -9.49433912e-03 3.63548994e-01 1.44079077e+00 -1.67790264e-01 8.94075871e-01 -1.18936241e+00 -4.21069711e-01 1.95245519e-01 -4.30315048e-01 7.54634917e-01 1.23535419e+00 4.32593256e-01 1.76929146e-01 8.47819522e-02 7.39984095e-01 -1.51615888e-01 3.04103464e-01 -7.65541419e-02 -3.37942317e-02 5.77533424e-01 1.07352686e+00 -3.09501380e-01 -3.99882823e-01 -3.06513757e-01 9.72698212e-01 3.64243239e-02 3.59396666e-01 -1.20210111e+00 -3.81251991e-01 6.21292889e-01 -1.97275102e-01 7.60524452e-01 -6.11756861e-01 -6.09212399e-01 -1.08510447e+00 4.00324702e-01 -3.49876732e-01 7.00679719e-01 -4.28130656e-01 -1.83009732e+00 -5.91028817e-02 -1.05492532e-01 -5.42296290e-01 -1.61201015e-01 -4.96420830e-01 -5.07411718e-01 9.99834776e-01 -1.12965417e+00 -9.08119977e-01 2.78948516e-01 6.14289761e-01 -9.38508064e-02 1.31125823e-01 4.89502013e-01 -7.57084973e-03 -9.16111171e-02 8.69079471e-01 1.11835670e+00 -2.73363858e-01 2.38522097e-01 -1.73232901e+00 1.54144615e-01 8.59764040e-01 2.02424731e-02 4.02455062e-01 1.30601108e+00 -5.19865990e-01 -1.72089267e+00 -8.04326415e-01 5.93492448e-01 -3.30460459e-01 1.55568063e+00 -3.52655560e-01 -3.90878469e-01 5.96683621e-01 -3.97224426e-01 3.39368135e-01 7.36619055e-01 4.69213068e-01 -2.69773215e-01 -4.75544959e-01 -1.42391658e+00 3.67712498e-01 1.07937074e+00 -5.77935159e-01 -1.60550550e-01 1.22901809e+00 9.22114670e-01 -8.14728200e-01 -1.25202775e+00 -1.49304017e-01 6.90853953e-01 -6.36015832e-01 5.91458023e-01 -1.02335393e+00 -1.85326114e-01 3.17359008e-02 -3.43045712e-01 -9.17885303e-01 -1.09474510e-02 -1.77762258e+00 -6.78978920e-01 7.79615164e-01 5.56614220e-01 -1.07932925e+00 7.04305887e-01 9.26704645e-01 1.88590199e-01 -7.76710749e-01 -1.25323999e+00 -1.54921937e+00 2.68410593e-01 -8.22256327e-01 5.27381897e-01 3.35714728e-01 -1.52072221e-01 -4.52101886e-01 -5.28201580e-01 4.29785997e-01 9.82259810e-01 4.20622617e-01 7.16765463e-01 -7.49605656e-01 -9.50042188e-01 -4.16131049e-01 -6.04249723e-03 -1.29219365e+00 8.41183960e-02 -6.12928569e-01 5.28084002e-02 -1.05925453e+00 2.30489925e-01 -4.52517301e-01 -4.06698644e-01 -4.70931232e-02 4.30434830e-02 -3.23603034e-01 1.59400329e-01 -3.86017635e-02 -1.23499513e+00 7.51492023e-01 1.07558429e+00 2.56812572e-01 6.09707683e-02 6.33757353e-01 -7.07012773e-01 5.02831340e-01 6.54710948e-01 -6.10885382e-01 -1.64384544e-01 -2.02815071e-01 7.10381567e-01 5.43060005e-01 5.02262190e-02 -5.53321183e-01 3.32639247e-01 -6.32248580e-01 -7.66926035e-02 -8.82051051e-01 8.71376917e-02 -5.24626493e-01 -2.72205984e-03 3.98487747e-01 -4.54701871e-01 1.23329656e-02 -9.54844281e-02 1.14750493e+00 4.55031157e-01 -2.82425582e-01 5.97239554e-01 4.55099670e-03 4.65762317e-01 8.70028019e-01 -1.62285745e-01 6.92468584e-01 9.45219338e-01 3.44591111e-01 -5.72020896e-02 -1.26271653e+00 -9.86481369e-01 5.33508122e-01 5.32089248e-02 -4.82539415e-01 3.22480798e-01 -1.08393812e+00 -9.55773234e-01 -5.44935703e-01 -6.67934716e-01 -1.18120268e-01 3.52141142e-01 1.05560064e+00 -3.83575320e-01 4.08413172e-01 4.92263734e-01 -2.00729817e-01 -7.46471882e-01 8.03590655e-01 4.05413151e-01 -3.32563609e-01 -3.85644473e-02 1.25059748e+00 -4.04674917e-01 -1.01372845e-01 3.72196645e-01 -3.96318972e-01 9.92373645e-01 -3.12410712e-01 6.92233860e-01 8.56423259e-01 -3.12491119e-01 2.60431260e-01 -1.83500141e-01 -2.18744040e-01 1.07057244e-01 -7.68055975e-01 1.38118410e+00 -5.37426174e-01 -1.16457529e-01 8.28030288e-01 1.23452604e+00 5.67820430e-01 -1.48703027e+00 -1.76272571e-01 1.12830487e-03 -6.07488871e-01 -2.82684058e-01 -7.24743128e-01 -8.61645281e-01 5.01733363e-01 3.34326595e-01 4.41388249e-01 8.69711399e-01 -1.36103583e-02 8.44273090e-01 5.37773728e-01 8.42959762e-01 -1.32105017e+00 -7.38484681e-01 5.78545392e-01 9.29610491e-01 -9.14110363e-01 1.27421424e-01 -1.78491771e-01 -2.43462935e-01 1.26561165e+00 -3.19364399e-01 -4.85077679e-01 9.31432962e-01 5.47114730e-01 -4.12565202e-01 3.11255872e-01 -1.00127840e+00 -3.70856136e-01 2.66768485e-01 2.04038471e-01 1.52914628e-01 5.03494382e-01 -5.09915590e-01 9.62599933e-01 -6.33996248e-01 1.93250030e-02 6.40627742e-01 5.27603030e-01 -3.71496677e-01 -4.70564306e-01 -3.90970320e-01 6.26913607e-01 -1.09258473e+00 -1.51819676e-01 1.73517808e-01 3.73040229e-01 -7.26298511e-01 8.61559749e-01 -1.11585647e-01 1.98044986e-01 -8.24986324e-02 -2.56961495e-01 7.32926607e-01 -1.19938217e-01 -3.18177491e-01 -1.87194310e-02 8.08089599e-02 -6.38236523e-01 1.15093783e-01 -1.01518631e+00 -8.24685872e-01 -8.85220945e-01 -6.11029267e-01 3.74950409e-01 1.06900227e+00 9.62750554e-01 1.43717736e-01 7.34312134e-03 1.11369753e+00 -3.41310114e-01 -1.84128535e+00 -7.09829569e-01 -8.31144214e-01 9.54761431e-02 1.81942001e-01 -1.97878569e-01 -1.04240513e+00 -4.33123678e-01]	[4.781369686126709, 3.429440498352051]
b95e3c74-5e74-4a05-8b58-267a3bae4502	unihd-at-tsar-2022-shared-task-is-compute-all	2301.01764	null	https://arxiv.org/abs/2301.01764v2	https://arxiv.org/pdf/2301.01764v2.pdf	UniHD at TSAR-2022 Shared Task: Is Compute All We Need for Lexical Simplification?	Previous state-of-the-art models for lexical simplification consist of complex pipelines with several components, each of which requires deep technical knowledge and fine-tuned interaction to achieve its full potential. As an alternative, we describe a frustratingly simple pipeline based on prompted GPT-3 responses, beating competing approaches by a wide margin in settings with few training instances. Our best-performing submission to the English language track of the TSAR-2022 shared task consists of an ``ensemble'' of six different prompt templates with varying context levels. As a late-breaking result, we further detail a language transfer technique that allows simplification in languages other than English. Applied to the Spanish and Portuguese subset, we achieve state-of-the-art results with only minor modification to the original prompts. Aside from detailing the implementation and setup, we spend the remainder of this work discussing the particularities of prompting and implications for future work. Code for the experiments is available online at https://github.com/dennlinger/TSAR-2022-Shared-Task	['Michael Gertz', 'Dennis Aumiller']	2023-01-04	null	null	null	null	['lexical-simplification']	['natural-language-processing']	[ 1.13958173e-01 2.02076674e-01 -3.58548202e-03 -5.88485599e-01 -1.44521523e+00 -8.73294294e-01 6.46324277e-01 1.95954442e-01 -7.76966631e-01 6.81761324e-01 4.80817497e-01 -4.72975194e-01 -1.79038737e-02 -2.39122704e-01 -5.48293293e-01 -1.12548910e-01 3.53401214e-01 8.38938236e-01 2.03326866e-01 -8.61864209e-01 7.91998953e-02 2.48637095e-01 -1.18406594e+00 6.49850011e-01 7.28362262e-01 2.49251455e-01 1.78391352e-01 7.34231770e-01 -2.17336789e-01 6.32119179e-01 -7.38224983e-01 -6.14889681e-01 2.79996097e-01 -6.22760728e-02 -1.11626971e+00 -5.78060627e-01 5.00436664e-01 -8.43438059e-02 -3.57957304e-01 5.96193850e-01 8.69572818e-01 4.35514838e-01 3.35573614e-01 -7.81488359e-01 -6.63521469e-01 9.72729266e-01 -1.95926428e-02 2.54889965e-01 8.46131921e-01 3.37861329e-01 1.07978630e+00 -9.54079509e-01 6.38146520e-01 1.33302367e+00 9.53205228e-01 7.42249370e-01 -1.40269029e+00 -6.58577621e-01 5.22801995e-01 2.74923652e-01 -1.38182807e+00 -6.87065780e-01 4.36688215e-01 -1.98705524e-01 1.76752484e+00 5.12322009e-01 3.45495492e-01 1.40411162e+00 -1.95241213e-01 7.76538312e-01 1.11468136e+00 -6.68228209e-01 -4.14889663e-01 3.46781948e-04 3.21181923e-01 2.44627029e-01 9.58102848e-03 -6.20592907e-02 -8.30896378e-01 -2.70445459e-02 2.09275261e-01 -5.87559581e-01 -1.58728942e-01 4.24112320e-01 -1.13189960e+00 7.34194875e-01 1.94166065e-03 2.26604819e-01 -2.63567656e-01 -2.07457036e-01 5.55173278e-01 4.44405973e-01 5.24858534e-01 7.18780041e-01 -9.21226919e-01 -3.98542732e-01 -1.02785218e+00 6.58227623e-01 1.06550968e+00 1.00421786e+00 5.24525762e-01 -8.17071721e-02 -5.51368833e-01 9.91464913e-01 -1.94848299e-01 4.30583775e-01 2.13091478e-01 -8.64517212e-01 6.28635883e-01 1.34222165e-01 2.57362396e-01 -4.13534075e-01 -6.36687696e-01 -2.55430013e-01 -3.55677098e-01 -2.87785381e-01 5.75324953e-01 -3.81973118e-01 -8.14994812e-01 1.85541165e+00 1.31228343e-01 -3.18116426e-01 -2.09531814e-01 6.44803047e-01 1.21671700e+00 5.31350195e-01 4.83137757e-01 -1.06811203e-01 1.52477396e+00 -1.19618118e+00 -8.80854547e-01 -3.58358175e-01 7.66473889e-01 -1.59530473e+00 1.57228875e+00 4.04630393e-01 -1.30713487e+00 -4.70825225e-01 -5.50096393e-01 -7.17542648e-01 -4.46496159e-01 2.76729204e-02 8.13264728e-01 4.05688971e-01 -1.21161425e+00 5.56809247e-01 -7.14226305e-01 -7.26504743e-01 -5.23949079e-02 3.32245618e-01 -3.61485690e-01 -8.41946378e-02 -1.47126830e+00 1.29983342e+00 1.76899195e-01 -2.29729339e-01 -3.38051647e-01 -1.11615217e+00 -7.80343056e-01 -4.53196689e-02 5.48191071e-01 -7.31584489e-01 2.15722275e+00 -5.91925502e-01 -1.83163369e+00 1.08194149e+00 -4.18983519e-01 -3.07819903e-01 5.25929391e-01 -8.16789985e-01 -3.30665916e-01 -2.63016671e-01 2.60887831e-01 9.19347644e-01 2.79301733e-01 -6.20051980e-01 -8.25416028e-01 3.09509963e-01 2.39290237e-01 4.94174659e-01 2.41333712e-02 7.67099917e-01 -4.85968173e-01 -8.82907391e-01 -5.51844478e-01 -1.00071871e+00 -1.77041665e-01 -8.12693596e-01 -4.04906660e-01 -6.65891945e-01 1.93137020e-01 -9.64518666e-01 1.55058730e+00 -1.94485772e+00 1.65723994e-01 -4.18953478e-01 -9.59833488e-02 5.87965012e-01 -3.77716601e-01 9.88604724e-01 -3.10737669e-01 4.11492616e-01 -1.08272843e-01 -7.32854664e-01 1.53704926e-01 -4.84910943e-02 -3.47694814e-01 5.39490320e-02 3.93011153e-01 1.10974300e+00 -9.06490743e-01 -1.72410801e-01 2.94518739e-01 4.96415794e-01 -6.75099015e-01 1.76479846e-01 -2.71999776e-01 5.37486672e-01 -6.80861846e-02 3.85860950e-01 3.92804682e-01 1.61625966e-01 1.05935205e-02 -7.75600448e-02 -4.69138563e-01 1.27402437e+00 -8.59747112e-01 2.01937222e+00 -5.46194851e-01 5.10856152e-01 1.85122013e-01 -4.47116047e-01 4.91398126e-01 3.44886929e-01 1.52056381e-01 -7.89775372e-01 -1.98012823e-03 4.93321091e-01 1.64341316e-01 -3.45504910e-01 7.64702022e-01 -2.58380156e-02 -4.45104390e-01 4.08624917e-01 2.36974031e-01 -5.02675653e-01 5.58277190e-01 3.23597282e-01 1.21958137e+00 3.32074463e-01 7.15751767e-01 -3.51784885e-01 3.69459093e-01 1.69373408e-01 5.01232088e-01 9.67489898e-01 -1.19523846e-01 7.60324061e-01 2.07554236e-01 -4.37520295e-01 -9.93851125e-01 -8.20282996e-01 -3.58805098e-02 1.53556335e+00 -5.24447322e-01 -1.00195909e+00 -7.18263566e-01 -5.04968524e-01 -1.43213242e-01 1.23370612e+00 -4.12963331e-01 1.31197214e-01 -1.11097169e+00 -6.08910739e-01 8.57530892e-01 2.17679620e-01 6.34842068e-02 -1.35779297e+00 -3.87681663e-01 3.64131361e-01 -7.71096051e-01 -1.27859783e+00 -5.29854894e-01 4.08226967e-01 -3.88550848e-01 -8.75127256e-01 -5.02569556e-01 -6.71274304e-01 -7.28563592e-02 2.68022150e-01 1.49303699e+00 1.57464501e-02 4.73476723e-02 2.85909802e-01 -5.70937812e-01 -6.40202403e-01 -3.75915706e-01 7.08650589e-01 -1.39180019e-01 -7.29907572e-01 6.18077397e-01 -5.17178178e-01 -3.68539393e-01 -1.68001339e-01 -5.83818793e-01 1.99867740e-01 4.74585623e-01 7.64281273e-01 4.71578449e-01 -5.68171501e-01 2.54322886e-01 -1.29161847e+00 9.22523797e-01 -4.14653331e-01 -2.31542721e-01 1.40957339e-02 -2.27758050e-01 -1.76946178e-01 8.60802472e-01 -3.50106776e-01 -1.18186760e+00 -3.79000828e-02 -6.16811574e-01 1.77032068e-01 -5.15588164e-01 4.96836036e-01 -9.43424776e-02 2.10319310e-01 8.32802117e-01 -2.28299826e-01 -6.21475041e-01 -9.62665975e-01 7.05700874e-01 5.82213581e-01 6.68456376e-01 -8.62735689e-01 7.06837356e-01 5.42695783e-02 -4.04606491e-01 -7.17564762e-01 -1.29994833e+00 -4.83438402e-01 -4.92727220e-01 2.98791260e-01 5.67138374e-01 -9.17918205e-01 -4.32845443e-01 5.81528485e-01 -1.64826751e+00 -8.40859830e-01 -4.18598711e-01 3.31136227e-01 -4.33443964e-01 1.54626608e-01 -9.21500623e-01 -2.63592482e-01 -5.29317319e-01 -1.06862557e+00 1.14372313e+00 9.94673818e-02 -9.32310224e-01 -9.28654313e-01 3.02921683e-01 4.23127830e-01 5.86952090e-01 -2.53445834e-01 9.91264999e-01 -1.21285713e+00 -1.21602878e-01 -1.67967871e-01 -8.34671594e-03 2.18329936e-01 -4.12859675e-03 4.11511101e-02 -9.93937790e-01 -3.30657437e-02 3.44431438e-02 -3.51963758e-01 7.17837870e-01 3.56500298e-01 7.88240731e-01 -2.29424194e-01 -2.00764537e-02 7.73592055e-01 7.99848974e-01 -8.66696239e-02 4.22676504e-01 7.68537343e-01 7.29140162e-01 7.26068914e-01 7.98738956e-01 5.61960079e-02 7.09240377e-01 9.39934671e-01 -1.21591724e-01 -1.84759587e-01 -6.15329325e-01 -2.44701698e-01 3.54608595e-01 1.29751718e+00 6.77341074e-02 -3.14967304e-01 -9.92981374e-01 5.51361382e-01 -1.82132208e+00 -7.83559024e-01 -2.99960047e-01 2.14442229e+00 1.32487786e+00 1.41055688e-01 2.48199731e-01 -8.25134888e-02 5.78232586e-01 2.24707156e-01 -1.50075510e-01 -9.72920716e-01 -3.71775836e-01 7.15560913e-01 2.37396449e-01 1.14320493e+00 -1.14701188e+00 1.75223863e+00 6.68978786e+00 1.05863452e+00 -1.05878949e+00 3.69989336e-01 1.73642740e-01 -4.67226923e-01 -3.92454207e-01 2.55558759e-01 -1.10802543e+00 2.10845098e-01 1.17081559e+00 -2.02914014e-01 9.37256336e-01 3.62181038e-01 4.42243367e-01 -4.19403836e-02 -1.10266137e+00 7.47466564e-01 6.87551266e-03 -9.86741602e-01 3.47452946e-02 -5.54225504e-01 5.31262934e-01 5.20929635e-01 -3.67565043e-02 8.32361877e-01 5.23905218e-01 -1.08515990e+00 9.34562981e-01 2.59035468e-01 8.51470649e-01 -5.58324754e-01 4.94392395e-01 2.98123509e-01 -1.00053155e+00 2.43210480e-01 -7.19712302e-02 -5.84591687e-01 4.90390718e-01 4.14767683e-01 -6.60503983e-01 7.20272660e-01 8.64645064e-01 5.70021093e-01 -7.20253646e-01 8.26239467e-01 -8.25155139e-01 8.06330264e-01 -6.59563243e-01 2.16999874e-01 1.89206198e-01 2.02653095e-01 8.32452357e-01 1.87759483e+00 -6.15898892e-02 2.68962443e-01 2.36915797e-01 2.58371800e-01 -1.35181114e-01 4.90508139e-01 -3.40116709e-01 1.70461282e-01 8.82769883e-01 1.37860024e+00 -4.15510148e-01 -3.31748009e-01 -5.05239308e-01 8.22519600e-01 8.36710334e-01 4.18047398e-01 -7.78225601e-01 -5.53229809e-01 6.44796669e-01 4.04161923e-02 1.73864812e-01 -1.61515206e-01 -5.98704994e-01 -1.23642480e+00 -1.61031336e-02 -1.46217287e+00 3.20280582e-01 -6.24766529e-01 -1.29495120e+00 6.90591931e-01 3.05378467e-01 -6.99034274e-01 -3.28629285e-01 -5.16496778e-01 -4.74289745e-01 1.34891379e+00 -1.60334229e+00 -1.28076422e+00 -7.68814459e-02 4.32552487e-01 8.46098900e-01 1.22169413e-01 1.25877440e+00 4.32255596e-01 -4.86839503e-01 7.88019717e-01 -2.78696358e-01 -2.13823527e-01 1.49893320e+00 -1.24980509e+00 1.09872782e+00 9.99901772e-01 1.53731659e-01 7.99161792e-01 1.04742491e+00 -6.88097000e-01 -9.39037025e-01 -9.14451778e-01 1.75490248e+00 -8.33887041e-01 8.99716198e-01 -5.02617121e-01 -9.55001831e-01 1.08502710e+00 7.09111869e-01 -5.49609244e-01 6.84288561e-01 6.62783742e-01 -3.48678052e-01 1.38359100e-01 -6.97767079e-01 1.06485915e+00 1.32694018e+00 -5.48056483e-01 -8.79180372e-01 6.36424720e-01 9.93537009e-01 -9.10245001e-01 -6.78263485e-01 4.52560365e-01 3.18425179e-01 -8.19947064e-01 7.55256534e-01 -7.20117033e-01 5.42699434e-02 -1.59322154e-02 -6.44753650e-02 -1.43817329e+00 -6.63831234e-01 -1.29114807e+00 2.76207983e-01 1.32773888e+00 5.45184135e-01 -5.57621658e-01 1.80051655e-01 4.28814709e-01 -6.34380460e-01 -5.96953034e-01 -7.72278607e-01 -6.28753841e-01 4.28593248e-01 -7.55278468e-01 3.98364365e-01 8.48113835e-01 -5.61164021e-02 9.34729576e-01 -2.11367205e-01 -1.31217062e-01 1.76367179e-01 -4.20865744e-01 9.11612213e-01 -9.73466873e-01 -3.88771117e-01 -5.59459686e-01 3.94661725e-01 -1.02781355e+00 2.73960918e-01 -1.08239043e+00 -8.68271962e-02 -1.57296765e+00 1.25585407e-01 -2.78295249e-01 -8.92728195e-02 8.95491064e-01 -4.94492084e-01 2.79405743e-01 6.34783149e-01 1.86918393e-01 -6.86814904e-01 2.17787355e-01 8.93578053e-01 2.20269874e-01 -4.13949758e-01 -1.01537123e-01 -1.09076357e+00 7.80973434e-01 1.05824983e+00 -7.65255630e-01 -4.80709188e-02 -1.11833012e+00 3.17425787e-01 -4.21406776e-01 3.15454379e-02 -5.93076885e-01 3.18220109e-01 -1.21408356e-02 -1.27812088e-01 -5.43862641e-01 4.97924775e-01 -1.43357664e-01 1.19244628e-01 1.41320243e-01 -3.51372689e-01 4.25103188e-01 7.64525533e-01 -2.32309654e-01 -8.21779743e-02 -2.39468962e-01 6.26846313e-01 -1.41032219e-01 -5.44402778e-01 -8.57127756e-02 -6.70651138e-01 5.32891214e-01 5.61995506e-01 1.39567763e-01 -5.33614159e-01 -4.63252485e-01 -6.21804893e-01 2.26905227e-01 3.82585853e-01 6.50119305e-01 -3.72471288e-02 -8.40841591e-01 -1.05967212e+00 -5.98771162e-02 -3.33524831e-02 7.97812268e-02 2.14099795e-01 8.54540527e-01 -5.23799300e-01 9.81556058e-01 1.08061004e-02 -2.03042254e-01 -1.12405837e+00 3.81188124e-01 1.78729609e-01 -7.18395531e-01 -4.31336731e-01 9.58200455e-01 5.63448593e-02 -9.87872541e-01 4.16689105e-02 -2.02030629e-01 -4.06433381e-02 2.22214032e-02 3.89895439e-01 4.22737539e-01 6.27782702e-01 -5.44062972e-01 -5.40300965e-01 2.78182000e-01 -5.44622540e-01 -4.44903374e-01 1.21728814e+00 -1.57019213e-01 -9.12680179e-02 3.71723235e-01 7.32192039e-01 4.76943702e-01 -8.55820477e-01 -2.88383901e-01 1.74186200e-01 -1.46667538e-02 -2.60599822e-01 -1.46288276e+00 -3.62708032e-01 8.42981040e-01 -1.40474930e-01 -1.66725684e-02 1.07895005e+00 -1.20477013e-01 7.97325134e-01 5.09362817e-01 1.77087709e-01 -1.09978271e+00 -4.51264113e-01 1.29196489e+00 1.00757682e+00 -1.10178626e+00 -1.00205675e-01 -5.56597173e-01 -6.47558749e-01 9.23626006e-01 5.16404927e-01 -1.51856109e-01 3.03731501e-01 3.84222597e-01 4.08611834e-01 1.26769155e-01 -1.01013947e+00 -1.64106444e-01 2.67355233e-01 7.80108452e-01 9.85335767e-01 2.10027903e-01 -7.66167641e-01 1.08528614e+00 -8.01115990e-01 -2.52347767e-01 2.69688427e-01 8.26775253e-01 -1.59469917e-01 -1.64362347e+00 -1.15126796e-01 1.57839194e-01 -8.84340405e-01 -8.87194693e-01 -5.91300845e-01 1.15587294e+00 -7.03893080e-02 1.12158418e+00 -3.08780521e-01 -1.67671844e-01 8.86952698e-01 1.95183203e-01 5.47551811e-01 -1.16082931e+00 -1.29120553e+00 3.04096133e-01 5.93248665e-01 -7.55160451e-01 1.55382813e-03 -8.40998232e-01 -9.81719136e-01 -7.32794940e-01 1.35699347e-01 -5.12735173e-02 3.72656584e-01 9.46791351e-01 3.87042314e-01 5.47255456e-01 2.93895304e-02 -1.28004241e+00 -6.12847984e-01 -1.26776791e+00 1.11918293e-01 4.14724410e-01 1.05369762e-02 -3.65322858e-01 -2.58519024e-01 -1.14085041e-01]	[10.943750381469727, 10.269387245178223]
7e227572-0943-453f-9850-5f5516a1c343	cear-cross-entity-aware-reranker-for	2104.08741	null	https://arxiv.org/abs/2104.08741v2	https://arxiv.org/pdf/2104.08741v2.pdf	CEAR: Cross-Entity Aware Reranker for Knowledge Base Completion	Pre-trained language models (LMs) like BERT have shown to store factual knowledge about the world. This knowledge can be used to augment the information present in Knowledge Bases, which tend to be incomplete. However, prior attempts at using BERT for task of Knowledge Base Completion (KBC) resulted in performance worse than embedding based techniques that rely only on the graph structure. In this work we develop a novel model, Cross-Entity Aware Reranker (CEAR), that uses BERT to re-rank the output of existing KBC models with cross-entity attention. Unlike prior work that scores each entity independently, CEAR uses BERT to score the entities together, which is effective for exploiting its factual knowledge. CEAR achieves a new state of art for the OLPBench dataset.	['Mausam', 'Parag Singla', 'Yatin Nandwani', 'Mayank Singh Chauhan', 'Keshav Kolluru']	2021-04-18	null	null	null	null	['knowledge-base-completion', 'knowledge-base-completion']	['graphs', 'knowledge-base']	[-5.19281805e-01 4.67157632e-01 -5.38901269e-01 -3.26663740e-02 -7.65081346e-01 -5.45775771e-01 8.34717751e-01 6.43228710e-01 -8.44301999e-01 9.10379708e-01 9.26861882e-01 -2.34796464e-01 -2.41424158e-01 -9.38579321e-01 -8.34612727e-01 8.67338851e-02 -3.18499297e-01 8.11655939e-01 5.10619938e-01 -4.80504185e-01 1.13723725e-02 3.05311918e-01 -1.02921069e+00 4.66017485e-01 6.68840706e-01 3.52773279e-01 2.22927868e-01 6.62702024e-01 -4.38361019e-01 1.36739671e+00 -4.22202826e-01 -9.27844107e-01 -5.05150072e-02 2.00199217e-01 -1.20780516e+00 -7.99169779e-01 5.91919959e-01 -4.60904717e-01 -7.19621360e-01 7.48656213e-01 2.02624261e-01 2.72309482e-01 7.24334180e-01 -1.08562946e+00 -9.91942763e-01 1.16370296e+00 -1.51554778e-01 2.52184957e-01 4.18933570e-01 -4.47800726e-01 1.57300282e+00 -1.21859598e+00 1.02630579e+00 1.21840084e+00 6.49926007e-01 4.86162305e-01 -1.10249627e+00 -4.72445369e-01 2.49752194e-01 6.84621036e-01 -1.61481845e+00 -3.17678213e-01 6.66265965e-01 -2.63617814e-01 1.82404017e+00 -3.54769416e-02 6.95626855e-01 8.12768042e-01 -2.88350314e-01 8.71600330e-01 7.13226378e-01 -6.92259908e-01 -9.28931534e-02 3.55621696e-01 5.96796811e-01 9.69880283e-01 8.27854335e-01 -2.29330838e-01 -1.00244284e+00 -2.26308972e-01 4.48859036e-01 -3.05718482e-01 -1.53814167e-01 -4.75652903e-01 -1.29806650e+00 6.47497118e-01 7.06636012e-01 3.32363009e-01 -5.53410769e-01 4.22271013e-01 2.67003119e-01 3.01746100e-01 3.79128933e-01 1.00212419e+00 -5.66490710e-01 -7.76080266e-02 -7.29038656e-01 2.96696611e-02 1.14780235e+00 9.84168887e-01 9.44235742e-01 -5.06113768e-02 -8.13896209e-02 7.58800626e-01 3.67287099e-01 2.68894643e-01 4.04805243e-01 -8.18274856e-01 7.78993905e-01 7.57215321e-01 2.50262350e-01 -9.59370852e-01 -2.53635466e-01 -4.76105928e-01 -2.69017458e-01 -1.09943464e-01 1.39080882e-01 8.06613714e-02 -1.02720249e+00 1.67552876e+00 4.68804836e-02 4.25147593e-01 4.95341420e-01 6.44488394e-01 1.07447839e+00 6.11793458e-01 3.79072547e-01 4.83796775e-01 1.10409355e+00 -9.55489993e-01 -7.23746419e-01 -3.61448586e-01 9.82830763e-01 -3.81671846e-01 8.72880757e-01 3.55892405e-02 -7.56852150e-01 -6.93737417e-02 -1.33055806e+00 -4.62689996e-01 -1.07112873e+00 1.09912148e-02 8.97065699e-01 5.01086950e-01 -1.39938545e+00 6.03071332e-01 -8.55689108e-01 -3.42583656e-01 3.77584159e-01 1.84117943e-01 -6.24113202e-01 -4.57505316e-01 -1.56323588e+00 1.53477919e+00 1.15983284e+00 -1.22262813e-01 -8.60809982e-01 -7.50627875e-01 -1.07180786e+00 1.95644692e-01 7.22766936e-01 -7.40220726e-01 8.26071024e-01 -1.80635408e-01 -1.18676662e+00 5.09155929e-01 -1.20044261e-01 -7.41575837e-01 -1.20188355e-01 -6.45139873e-01 -5.91586769e-01 1.74427763e-01 3.64328995e-02 7.06204176e-01 4.34428006e-01 -1.39547825e+00 -5.06484389e-01 -1.15550533e-01 6.02532208e-01 3.26124340e-01 -7.06594527e-01 -1.84832856e-01 -8.90348911e-01 -5.25104105e-01 -2.88786411e-01 -6.58172429e-01 -8.77535418e-02 -4.16719466e-01 -3.91839296e-01 -3.19933802e-01 5.87679565e-01 -1.11381280e+00 1.59004605e+00 -1.71466088e+00 2.93021858e-01 3.96101505e-01 4.39739645e-01 5.48991919e-01 -4.59610879e-01 7.75174022e-01 1.86980724e-01 3.41540366e-01 2.12290883e-01 -3.79958689e-01 6.15416095e-02 4.79140103e-01 -3.98778677e-01 -2.83485670e-02 4.24414635e-01 1.35437453e+00 -1.28281724e+00 -6.54397428e-01 -5.60487024e-02 4.16582942e-01 -6.48512065e-01 -3.31755906e-01 -3.76912326e-01 -4.16277975e-01 -4.38112676e-01 7.07041979e-01 1.27996638e-01 -4.23941225e-01 3.63373846e-01 -3.47600907e-01 2.88445532e-01 5.37088454e-01 -1.04464948e+00 1.70782733e+00 -7.13856876e-01 6.52731836e-01 -3.59033793e-01 -6.23212159e-01 4.57515895e-01 3.50705981e-01 1.02302670e-01 -3.95451665e-01 -4.39633012e-01 3.45273733e-01 -2.15224087e-01 -5.31161353e-02 9.82563376e-01 2.78985985e-02 8.99983495e-02 2.07753241e-01 7.16109216e-01 1.49115128e-02 5.68161786e-01 1.12421644e+00 1.49611127e+00 1.16153657e-01 4.86209750e-01 -2.78418940e-02 3.20723504e-01 3.95587593e-01 3.14348966e-01 7.71090508e-01 3.16309005e-01 -4.90826415e-03 3.48733366e-01 -1.64100066e-01 -9.60715830e-01 -1.14943600e+00 2.58783907e-01 8.87501955e-01 9.04708803e-02 -1.11858034e+00 2.49400381e-02 -1.01917338e+00 4.63879079e-01 1.19487751e+00 -4.58513141e-01 -2.36139849e-01 -4.16389704e-01 -3.26805055e-01 7.23701835e-01 9.17144656e-01 2.66392231e-01 -8.12915087e-01 -3.42832133e-02 3.91756415e-01 -1.54504314e-01 -1.27593839e+00 -1.91448301e-01 1.37261674e-01 -7.45048463e-01 -1.02158165e+00 -4.84344065e-01 -4.53211248e-01 5.18235207e-01 3.54413912e-02 1.53970730e+00 8.23515728e-02 -7.76517317e-02 8.69469702e-01 -5.42387962e-01 -5.12669444e-01 -2.04375446e-01 3.41645658e-01 6.22405410e-02 -4.35617328e-01 4.90281761e-01 -4.33481157e-01 -2.83955544e-01 -1.97095692e-01 -8.65240812e-01 9.60848331e-02 7.58278310e-01 9.17153418e-01 3.41698170e-01 1.49960503e-01 7.94611394e-01 -1.19280040e+00 6.17627084e-01 -5.07647514e-01 -3.72351319e-01 5.75655758e-01 -8.60485971e-01 5.39909184e-01 3.60767871e-01 -1.81551486e-01 -1.15845716e+00 -2.02575207e-01 1.61266729e-01 -3.90049845e-01 4.10232365e-01 1.38238585e+00 1.56521693e-01 -1.51049038e-02 5.81791937e-01 -1.28066614e-01 -5.87394595e-01 -7.16406226e-01 8.24612439e-01 3.05089712e-01 5.31481564e-01 -7.23583281e-01 8.68759990e-01 3.11285943e-01 -3.97086181e-02 -7.14091301e-01 -1.21389389e+00 -1.00831556e+00 -7.03228056e-01 -1.64121613e-02 7.31272519e-01 -1.38040161e+00 -3.06430191e-01 -1.16690785e-01 -1.20347953e+00 -2.39263788e-01 -4.26939785e-01 8.20033073e-01 1.53546559e-03 3.20921004e-01 -6.57622278e-01 -6.15851343e-01 -1.74759522e-01 -1.98826075e-01 8.83624613e-01 -7.37883821e-02 -1.98632509e-01 -1.41159773e+00 2.47999966e-01 4.19684470e-01 4.44706231e-01 -8.57607201e-02 1.24477530e+00 -9.56090987e-01 -9.32211995e-01 -4.40208942e-01 -2.59075463e-01 3.15164447e-01 -1.80476084e-01 -2.73612827e-01 -7.66771376e-01 -2.94481575e-01 -9.39470172e-01 -5.14961958e-01 1.39863420e+00 -4.80133861e-01 5.26852846e-01 -4.41167891e-01 -5.06403506e-01 3.03240865e-01 1.70892453e+00 -3.78497720e-01 6.58323050e-01 5.15011311e-01 9.34713602e-01 3.18781883e-01 3.02106738e-01 1.63819924e-01 7.52233505e-01 5.86709261e-01 2.61833817e-01 1.72566704e-03 -4.96366143e-01 -7.99945533e-01 5.58348835e-01 1.08086562e+00 -4.13961470e-01 -2.90299088e-01 -1.23252594e+00 1.00941038e+00 -1.76243663e+00 -9.80354369e-01 5.54394200e-02 1.97833252e+00 1.07666004e+00 1.99953705e-01 -3.70354444e-01 -3.45229477e-01 2.35194206e-01 1.39535591e-01 -3.92356694e-01 -1.20643273e-01 -3.31083864e-01 4.63669777e-01 1.02509904e+00 7.40106821e-01 -9.25822735e-01 1.33584535e+00 6.03439808e+00 9.48784471e-01 -4.49888051e-01 1.55722588e-01 -2.49549150e-01 -2.68311854e-02 -6.24822438e-01 4.87551153e-01 -1.03734636e+00 -5.47333956e-02 9.00813401e-01 -4.65073764e-01 4.72577959e-01 7.46249139e-01 -6.91988826e-01 -2.41875544e-01 -1.14626610e+00 7.14399278e-01 2.83113360e-01 -1.59528494e+00 4.81656343e-01 -8.91288146e-02 8.01829815e-01 4.39309120e-01 -2.70747632e-01 8.85872245e-01 1.23893809e+00 -8.00392509e-01 3.64282787e-01 7.88761020e-01 7.41242826e-01 -5.13643503e-01 9.59946513e-01 3.03439975e-01 -1.16543353e+00 -7.77072161e-02 -5.22912920e-01 1.87327087e-01 1.80204421e-01 4.79317337e-01 -1.39128196e+00 1.08004177e+00 4.99298275e-01 7.92098999e-01 -1.03341484e+00 8.65837157e-01 -5.01656830e-01 7.64019489e-01 -4.73735362e-01 -1.06207840e-01 2.69870400e-01 2.26449102e-01 6.94548905e-01 1.31940758e+00 6.25784472e-02 1.35767888e-02 1.32240787e-01 5.70348918e-01 -5.45375347e-01 1.47674069e-01 -7.68251598e-01 -5.76346099e-01 5.03544331e-01 1.08318722e+00 -2.63009816e-01 -7.96355247e-01 -6.72518134e-01 9.78396773e-01 1.02732277e+00 5.41144073e-01 -3.82266045e-01 -4.99570280e-01 2.42395058e-01 -4.97702248e-02 4.06375825e-01 -6.15241945e-01 1.94809571e-01 -1.43025506e+00 6.69826120e-02 -3.14087063e-01 7.55437136e-01 -9.16980803e-01 -1.23034286e+00 2.04985529e-01 2.50098199e-01 -6.25990093e-01 -1.93152815e-01 -8.10661674e-01 -2.25859314e-01 9.06009436e-01 -1.99348176e+00 -1.43919766e+00 5.85701764e-02 5.91683328e-01 1.27318263e-01 -3.44464272e-01 1.06257844e+00 3.06667775e-01 -2.20228776e-01 3.42364758e-01 2.25191176e-01 4.61138338e-01 8.08651388e-01 -1.64685547e+00 3.50639522e-01 8.43624890e-01 7.96699286e-01 1.18798804e+00 2.64133155e-01 -1.12814879e+00 -1.51771235e+00 -1.15731001e+00 1.29429972e+00 -9.62010920e-01 1.03924358e+00 -2.79816866e-01 -8.63019109e-01 1.01184475e+00 1.76743969e-01 -5.19907027e-02 7.44449854e-01 8.14604223e-01 -9.88062799e-01 1.67316925e-02 -6.16128087e-01 6.98917627e-01 1.04913092e+00 -9.10136938e-01 -1.06487620e+00 1.97470859e-01 9.82169747e-01 -1.08529858e-01 -1.04282546e+00 3.43104392e-01 4.35337096e-01 -1.47023767e-01 1.19354165e+00 -1.14327621e+00 2.76393771e-01 -3.55931818e-01 -2.83454835e-01 -1.52843571e+00 -3.80124211e-01 -3.20363492e-01 -7.57997394e-01 1.15892458e+00 9.15187955e-01 -5.01212537e-01 7.54269898e-01 6.49133444e-01 -1.06890954e-01 -3.40959787e-01 -8.23541462e-01 -1.08818078e+00 -1.36140674e-01 -3.35105509e-01 4.28265482e-01 1.18790150e+00 3.65449846e-01 6.77656591e-01 -2.62784123e-01 4.10659701e-01 4.01744872e-01 -2.01711744e-01 5.90258777e-01 -1.32645535e+00 -1.89726844e-01 -9.96530876e-02 -6.00351274e-01 -7.49579787e-01 3.90274167e-01 -1.51655269e+00 -3.88807058e-01 -2.22807765e+00 2.70675659e-01 -4.72667664e-01 -7.18859911e-01 8.58633101e-01 -3.24685514e-01 1.14171021e-01 3.30107242e-01 1.36972383e-01 -1.16137981e+00 5.49367607e-01 7.54329622e-01 -3.56272668e-01 -4.44718227e-02 -6.72806442e-01 -5.58760762e-01 6.45185113e-01 6.07514024e-01 -4.75553095e-01 -3.49914402e-01 -6.83673859e-01 8.24306726e-01 -1.62624940e-01 5.03003001e-01 -8.34403455e-01 7.46651411e-01 2.08076924e-01 3.79599929e-01 -6.05443656e-01 6.62459314e-01 -9.29720461e-01 -2.38500200e-02 9.70785692e-02 -4.05376256e-01 -1.45501513e-02 2.81493008e-01 8.90010953e-01 -4.98512030e-01 -4.58488703e-01 -8.59852284e-02 -2.02571988e-01 -1.31323314e+00 1.88156739e-01 -7.10783675e-02 2.39552140e-01 4.73251969e-01 1.21485703e-01 -6.35651469e-01 -4.95834291e-01 -7.92414367e-01 5.64557016e-01 2.08843797e-01 3.47840756e-01 7.68062115e-01 -1.52498007e+00 -7.05595732e-01 -4.48762208e-01 5.72114408e-01 -2.12595597e-01 -1.52494162e-02 5.14078140e-01 -4.19342995e-01 7.41937101e-01 3.93723510e-02 1.04024209e-01 -1.08278131e+00 7.16067672e-01 -1.27991095e-01 -9.36340988e-01 -5.55851996e-01 8.08740854e-01 -4.17348206e-01 -4.40937340e-01 1.31864473e-01 -1.34766504e-01 -4.68247980e-01 2.22113147e-01 2.87654579e-01 3.85850310e-01 2.77532548e-01 -3.36618274e-01 -3.11753929e-01 1.96089506e-01 -4.69542712e-01 -4.20401812e-01 1.41618454e+00 8.02584179e-03 -2.11343661e-01 3.97385627e-01 1.09092760e+00 4.06900495e-01 -6.10936284e-01 -7.75988758e-01 6.86675668e-01 -3.84863079e-01 2.21238837e-01 -1.16924345e+00 -5.82108378e-01 6.51988864e-01 -4.08329777e-02 -2.65295953e-01 6.02162123e-01 1.66616008e-01 5.89848161e-01 1.36545634e+00 7.08798945e-01 -1.25170970e+00 6.26312122e-02 8.92696977e-01 8.77424896e-01 -1.07561815e+00 2.85133451e-01 -3.19424301e-01 -8.06551754e-01 8.88928831e-01 5.52753627e-01 5.95562384e-02 7.64787138e-01 8.81391391e-03 -3.04826230e-01 -4.01289672e-01 -1.04125023e+00 -6.97975397e-01 5.86656928e-01 7.19605505e-01 2.46585488e-01 6.21853173e-02 -1.00328349e-01 5.54698825e-01 -9.09988433e-02 -3.45094800e-02 3.99078667e-01 9.71007288e-01 -3.26495856e-01 -1.14366400e+00 9.80950445e-02 6.54875517e-01 -4.62524503e-01 -6.61610484e-01 -6.61238134e-01 9.36047733e-01 -2.11671844e-01 7.09248900e-01 -3.67634088e-01 -4.09153759e-01 2.87468821e-01 5.51689267e-01 3.75615418e-01 -9.17119563e-01 -2.68815279e-01 -5.18336117e-01 8.03788185e-01 -5.87133467e-01 -2.42445529e-01 -1.74532935e-01 -1.19789684e+00 1.89209245e-02 -4.86970782e-01 3.23885411e-01 5.24669111e-01 5.99923253e-01 5.10483444e-01 4.76335883e-01 -1.23687930e-01 -2.65964270e-01 -4.48912919e-01 -8.42535734e-01 -5.49947500e-01 4.53555554e-01 1.82894900e-01 -8.61250401e-01 2.62489635e-02 8.26291218e-02]	[9.1321439743042, 8.116639137268066]
8260a34b-973d-4493-a7ee-bfe985e23e54	video-based-camera-localization-using-anchor	2107.03068	null	https://arxiv.org/abs/2107.03068v1	https://arxiv.org/pdf/2107.03068v1.pdf	Video-Based Camera Localization Using Anchor View Detection and Recursive 3D Reconstruction	In this paper we introduce a new camera localization strategy designed for image sequences captured in challenging industrial situations such as industrial parts inspection. To deal with peculiar appearances that hurt standard 3D reconstruction pipeline, we exploit pre-knowledge of the scene by selecting key frames in the sequence (called as anchors) which are roughly connected to a certain location. Our method then seek the location of each frame in time-order, while recursively updating an augmented 3D model which can provide current camera location and surrounding 3D structure. In an experiment on a practical industrial situation, our method can localize over 99% frames in the input sequence, whereas standard localization methods fail to reconstruct a complete camera trajectory.	['Masatoshi Okutomi', 'Naoyuki Miyashita', 'Koki Onbe', 'Hajime Taira']	2021-07-07	null	null	null	null	['camera-localization']	['computer-vision']	[ 2.73854017e-01 -9.29284543e-02 -5.59736937e-02 -8.44858140e-02 -6.74359322e-01 -9.25741553e-01 2.50168592e-01 1.71410456e-01 -5.84498420e-02 2.83260018e-01 -2.76118636e-01 -1.14644542e-01 8.86019319e-02 -2.87428886e-01 -8.16243172e-01 -5.15325129e-01 -1.39821991e-02 4.86215383e-01 6.08499467e-01 1.88496992e-01 5.08597255e-01 9.61675704e-01 -1.00352752e+00 -3.75606865e-02 3.23635302e-02 8.48273158e-01 8.70764434e-01 8.49592686e-01 2.22508058e-01 6.48998499e-01 -8.17988217e-01 -1.80570018e-02 3.94924372e-01 -3.41563523e-01 -6.62387192e-01 9.02003407e-01 4.39646870e-01 -4.96471286e-01 -4.53788489e-01 1.23333037e+00 7.97551423e-02 1.85348108e-01 1.58823416e-01 -8.97871733e-01 3.69408205e-02 3.05697590e-01 -6.64573014e-01 3.42259765e-01 9.85212147e-01 1.18468925e-01 3.87677938e-01 -7.91919589e-01 1.06210673e+00 1.05792677e+00 8.77476335e-01 2.08168522e-01 -1.04973161e+00 -2.64623780e-02 3.92503262e-01 1.62553862e-01 -1.35710621e+00 -3.35982144e-01 1.10038996e+00 -2.46800750e-01 6.01807356e-01 -8.69235843e-02 5.55764198e-01 9.42239463e-01 4.02400821e-01 6.62716329e-01 7.34283149e-01 -5.79672456e-01 1.68173417e-01 -1.84017897e-01 -2.27042306e-02 7.75242269e-01 3.76951359e-02 -2.57030483e-02 -4.66635793e-01 4.35477681e-03 1.48903131e+00 2.97669500e-01 -4.03829724e-01 -9.11824048e-01 -1.87470663e+00 3.73107761e-01 3.35894376e-01 2.47484282e-01 -5.79652488e-01 1.52890101e-01 2.74895370e-01 1.38374612e-01 1.39743999e-01 3.32485765e-01 -4.97016460e-01 -1.99372351e-01 -6.83963656e-01 -9.55736712e-02 6.38762772e-01 1.36515903e+00 1.00093794e+00 -2.12322667e-01 3.80341172e-01 3.02186221e-01 1.16094969e-01 3.56775254e-01 1.15215994e-01 -1.46053374e+00 4.27323431e-01 3.52853656e-01 3.18314463e-01 -1.05711699e+00 -2.10534319e-01 -1.76666319e-01 -4.97061551e-01 2.32863456e-01 6.24817371e-01 9.04741585e-02 -6.50533199e-01 1.04179132e+00 7.36640871e-01 6.74196601e-01 1.26452791e-02 9.86139774e-01 1.50677025e-01 5.98411441e-01 -6.58082783e-01 -3.27390075e-01 1.28141117e+00 -9.99661148e-01 -5.74794531e-01 -3.75435352e-01 4.15176511e-01 -1.12263298e+00 6.05374694e-01 5.67338228e-01 -9.32247281e-01 -8.82097542e-01 -1.01151955e+00 2.95959502e-01 1.26781603e-02 2.94081658e-01 3.19188505e-01 2.43312284e-01 -7.59316742e-01 6.45731091e-01 -9.88807440e-01 -4.42700952e-01 -9.78961810e-02 1.23945117e-01 -6.54445648e-01 -4.57843930e-01 -4.10359532e-01 7.94238269e-01 3.73650849e-01 2.67104715e-01 -1.32905519e+00 -3.75911057e-01 -8.92360270e-01 -4.71988022e-01 8.34527373e-01 -4.46106255e-01 1.32190979e+00 -5.34360766e-01 -1.53163695e+00 7.89749801e-01 -3.13424945e-01 -3.02898884e-01 3.69731426e-01 -5.32979310e-01 -1.66717604e-01 4.75572199e-01 2.27891997e-01 2.78531075e-01 1.03358829e+00 -1.59728777e+00 -8.42092872e-01 -2.83267170e-01 3.65869790e-01 1.09794304e-01 3.18865061e-01 -1.54953256e-01 -8.89838099e-01 -4.34149325e-01 7.13103056e-01 -1.07363784e+00 -6.99261665e-01 -9.73477513e-02 -4.17732447e-01 1.92780390e-01 1.19714034e+00 -5.86851180e-01 6.17462456e-01 -2.18598032e+00 1.05208695e-01 1.69518203e-01 -1.06161892e-01 1.05236601e-02 1.22042194e-01 4.81396258e-01 -8.66900757e-02 -5.95840573e-01 1.39013186e-01 -4.23379451e-01 -2.13940755e-01 2.23006710e-01 -1.90042913e-01 9.60389733e-01 -2.49783680e-01 5.12575328e-01 -1.20168149e+00 -5.87674916e-01 9.52768803e-01 3.31441492e-01 -2.37134829e-01 3.29548866e-01 -5.24981096e-02 8.06360006e-01 -6.13806367e-01 9.03709829e-01 6.84618592e-01 -1.63636893e-01 1.55641139e-01 -3.91500354e-01 -9.99121591e-02 -2.10519750e-02 -1.49554300e+00 2.28511786e+00 -4.45161551e-01 5.47992468e-01 3.42392206e-01 -8.16322982e-01 9.50916588e-01 3.42144042e-01 7.89680839e-01 -2.87504196e-02 5.70893101e-02 -1.25546992e-01 -5.50126374e-01 -3.85993272e-01 5.29807031e-01 2.82854021e-01 -2.16857910e-01 2.44572610e-01 -3.72844329e-03 -3.90575647e-01 9.92499739e-02 3.26182768e-02 1.30357015e+00 4.50748801e-01 3.70902568e-01 -3.44687998e-02 5.94740927e-01 2.59756237e-01 3.91905338e-01 6.20188713e-01 -3.17031771e-01 8.63575935e-01 2.43175894e-01 -7.00722277e-01 -1.06798506e+00 -9.01490867e-01 2.16691524e-01 2.49411941e-01 7.52319455e-01 -5.62659025e-01 -7.47132242e-01 -1.04799545e+00 -2.01112643e-01 2.94910312e-01 -2.14857936e-01 2.71657612e-02 -8.98189306e-01 8.00809935e-02 -8.10227543e-02 3.84543568e-01 3.14307064e-01 -7.34109581e-01 -1.18684494e+00 3.01647872e-01 -1.81357026e-01 -1.65645254e+00 -7.95097888e-01 3.63198012e-01 -1.09247780e+00 -1.46281850e+00 -5.03594697e-01 -1.05211306e+00 1.05410063e+00 7.55745709e-01 1.20952260e+00 -9.06225592e-02 -3.22984219e-01 5.92884362e-01 -4.33438301e-01 9.56063271e-02 -5.44643760e-01 -1.52440369e-01 -5.55573264e-03 -2.11833462e-01 8.00070688e-02 -2.99796134e-01 -5.40746987e-01 5.75978756e-01 -5.10433018e-01 -3.25280696e-01 4.52324063e-01 6.49803460e-01 7.53393233e-01 5.56351900e-01 -1.91001162e-01 -6.86506748e-01 3.48344482e-02 -5.17688841e-02 -9.82884884e-01 4.57865633e-02 -9.74816456e-02 -3.92637908e-01 3.91613394e-01 -5.17739356e-01 -6.64319754e-01 9.90065336e-01 1.03740254e-02 -1.08883989e+00 -6.26411498e-01 2.56588608e-01 -2.47191563e-01 -1.97824161e-03 2.34072760e-01 7.78226331e-02 -3.76405530e-02 -5.81846952e-01 2.20848799e-01 1.84952676e-01 9.36242998e-01 -4.79858607e-01 1.03524458e+00 3.86361390e-01 1.04747936e-01 -8.05663109e-01 -7.47659087e-01 -8.55708957e-01 -1.41731501e+00 -5.92615962e-01 6.21463180e-01 -1.00200582e+00 -8.04764092e-01 2.19075829e-01 -1.49243712e+00 -9.40043479e-02 -4.89802778e-01 7.21967101e-01 -6.74121022e-01 7.38067329e-01 -6.50276005e-01 -7.71713197e-01 2.89810181e-01 -1.37349260e+00 1.61476994e+00 -1.16770037e-01 -1.55656427e-01 -1.06871486e+00 -5.13996556e-02 -1.77335402e-03 -2.48194262e-01 5.23199260e-01 3.90190095e-01 -1.33874863e-01 -8.94051492e-01 -5.68681002e-01 3.39640349e-01 2.41088152e-01 3.14855009e-01 -1.34461224e-01 -8.10959399e-01 -2.77800381e-01 2.37685502e-01 3.35626036e-01 2.55407065e-01 4.54883784e-01 7.96525478e-01 1.66637033e-01 -6.87424123e-01 4.06414896e-01 1.49994946e+00 4.28925663e-01 3.27427894e-01 3.97968382e-01 7.50546753e-01 3.78735095e-01 1.14515996e+00 4.48500186e-01 1.23243826e-02 9.13425982e-01 8.16968679e-01 -1.02526248e-01 -5.06560877e-03 -3.71351898e-01 5.44758022e-01 6.91952288e-01 3.17456685e-02 -1.36515394e-01 -5.16405046e-01 3.65186691e-01 -1.80595088e+00 -8.08650911e-01 -2.83506602e-01 2.13379145e+00 2.94324756e-01 3.38377386e-01 -1.67577546e-02 1.90419585e-01 9.84152853e-01 1.81511536e-01 -4.85948414e-01 3.01013201e-01 1.44438714e-01 -2.77574331e-01 8.89142692e-01 6.07627332e-01 -1.23590350e+00 7.77656138e-01 7.05021763e+00 2.91609526e-01 -6.59452915e-01 7.37613859e-03 1.73478439e-01 3.03348809e-01 4.24994618e-01 2.75476485e-01 -8.46114814e-01 2.81628817e-01 5.96329927e-01 2.65421391e-01 2.27802932e-01 1.18604898e+00 6.46677762e-02 -2.71600008e-01 -1.29983211e+00 1.28171408e+00 2.68377542e-01 -1.26523221e+00 -4.41710532e-01 2.14453891e-01 7.01037586e-01 -1.88367873e-01 -3.26109946e-01 -4.05957907e-01 3.57116550e-01 -3.35538208e-01 1.09215105e+00 4.24747676e-01 4.22147095e-01 -6.64571166e-01 8.25731337e-01 5.31962216e-01 -1.39018452e+00 -2.21456289e-02 -5.13594508e-01 2.41354644e-01 8.79070818e-01 8.36035252e-01 -1.11756289e+00 5.51580489e-01 6.76768303e-01 1.04015374e+00 -4.04208660e-01 1.12919617e+00 -3.65216881e-01 1.52787924e-01 -2.96672493e-01 4.06665325e-01 3.08014363e-01 -2.43314907e-01 8.37250590e-01 8.86056423e-01 7.52928615e-01 -6.40340196e-03 5.74515760e-01 4.92779762e-01 3.03040087e-01 -5.81897199e-01 -1.09537315e+00 4.61086273e-01 4.16767061e-01 1.16288757e+00 -1.34042251e+00 -1.97530225e-01 -3.44525188e-01 1.50190926e+00 -3.78131509e-01 2.33164564e-01 -8.74490201e-01 -1.67655036e-01 5.04549980e-01 1.32341161e-01 6.73485398e-01 -8.48649919e-01 3.96462321e-01 -1.07781589e+00 2.11446602e-02 -6.59402668e-01 2.24538401e-01 -1.06692696e+00 -8.76178861e-01 6.09799981e-01 1.45730972e-01 -1.72033536e+00 -3.77191395e-01 -8.00182045e-01 -3.40348780e-01 3.92715722e-01 -1.16937077e+00 -1.07862592e+00 -3.33548188e-01 8.77558351e-01 1.12080085e+00 2.96994168e-02 5.88092804e-01 1.26040220e-01 -3.75658572e-01 -1.84795544e-01 4.19486240e-02 -2.52239425e-02 5.96493304e-01 -1.28440154e+00 4.68610615e-01 1.01532042e+00 5.17676950e-01 6.07801139e-01 7.74990916e-01 -5.96898794e-01 -1.75180924e+00 -8.31556499e-01 5.25052190e-01 -9.14467931e-01 5.13209581e-01 -3.17329288e-01 -3.57441276e-01 9.71852899e-01 1.94442809e-01 1.53994396e-01 -6.90099224e-02 -4.32924002e-01 1.91740468e-01 -1.18738450e-01 -9.91126597e-01 2.68690884e-01 1.11540532e+00 -5.30951500e-01 -4.93182421e-01 5.43803871e-01 6.36304677e-01 -9.74757493e-01 -9.57766533e-01 1.00447059e-01 2.22843394e-01 -8.53278041e-01 1.36613703e+00 1.82036564e-01 -3.72540276e-03 -8.15001547e-01 -2.56833762e-01 -1.22430420e+00 -1.13615729e-01 -9.97940063e-01 -1.37655810e-01 1.11711538e+00 -2.04673365e-01 -1.79056957e-01 1.05078018e+00 1.61912888e-01 -3.84419560e-01 -9.82147679e-02 -9.78257298e-01 -6.61693275e-01 -7.69948006e-01 -3.63913834e-01 5.39420366e-01 6.82318985e-01 -1.35359213e-01 1.68916970e-01 -3.08965206e-01 7.04743922e-01 8.07051361e-01 2.75427222e-01 1.03876233e+00 -1.03097034e+00 2.52265818e-02 1.58066675e-01 -7.83695877e-01 -1.64326715e+00 1.79552794e-01 -3.22614402e-01 3.57923120e-01 -1.21911943e+00 -7.10520223e-02 -1.00317791e-01 -1.56953745e-02 7.01658502e-02 3.39378297e-01 3.27199370e-01 1.93048447e-01 1.28154114e-01 -8.08976829e-01 -8.52054954e-02 1.10031867e+00 -2.12523490e-01 6.96562529e-02 3.40392977e-01 2.23441515e-02 9.08808291e-01 3.76126140e-01 -2.89870381e-01 -4.48622406e-01 -5.01887083e-01 -3.46630812e-01 5.65852165e-01 5.96864402e-01 -1.19216585e+00 3.42577279e-01 2.97305244e-03 6.66922390e-01 -1.22472286e+00 4.69548970e-01 -1.70841539e+00 5.25302052e-01 4.03403401e-01 1.54240420e-02 7.22605586e-01 -1.43042840e-02 7.08398283e-01 -2.36526087e-01 -6.56616092e-01 4.61311400e-01 -5.70640683e-01 -1.08876753e+00 8.70973095e-02 -4.52564687e-01 -7.27533400e-01 1.36504817e+00 -5.30796111e-01 1.62283763e-01 -1.81100637e-01 -9.11017418e-01 -8.35251287e-02 1.01814353e+00 3.89894307e-01 7.93174207e-01 -1.30213273e+00 -1.48849472e-01 4.37340945e-01 4.26870584e-02 3.73186260e-01 1.22199498e-01 6.97904110e-01 -1.02204680e+00 3.04939896e-01 9.57985446e-02 -1.18734956e+00 -1.23710322e+00 9.87778783e-01 1.59958825e-01 -1.29218295e-01 -1.14741337e+00 4.40343112e-01 -2.46211048e-02 -1.18985340e-01 3.94391924e-01 -6.61621451e-01 1.18505284e-01 -3.06390166e-01 5.33217311e-01 3.74511063e-01 2.39691570e-01 -9.70498741e-01 -5.08226752e-01 1.02483606e+00 3.16135436e-01 1.40578607e-02 1.25877953e+00 -6.63115621e-01 5.10969348e-02 5.50957680e-01 1.16927910e+00 1.20915115e-01 -1.79746985e+00 -5.65194488e-02 1.73372671e-01 -9.50664103e-01 -6.07205974e-03 -2.92614132e-01 -1.29016089e+00 6.06637239e-01 7.14171350e-01 2.14974567e-01 1.02899849e+00 2.38739148e-01 7.36087203e-01 4.20093238e-01 9.72092092e-01 -7.28232324e-01 1.59457609e-01 4.25120980e-01 3.38164240e-01 -9.41381812e-01 1.52692646e-01 -6.03446007e-01 -5.51450372e-01 1.46398747e+00 3.36455017e-01 -2.22705156e-01 6.16889179e-01 5.50212264e-01 1.25881329e-01 -3.66431415e-01 -3.69648278e-01 3.65470978e-03 -2.36804634e-01 8.95572364e-01 -6.33474737e-02 -4.45535064e-01 6.30969882e-01 -1.41101673e-01 1.83073252e-01 -3.24338973e-01 6.09211564e-01 1.26909876e+00 -3.15007061e-01 -1.01208520e+00 -7.86453724e-01 -4.26997870e-01 -2.02601984e-01 5.29720306e-01 -8.32761079e-02 8.42236757e-01 -4.86953137e-03 9.47259545e-01 2.00921714e-01 -3.56669903e-01 7.69787133e-01 -2.32612252e-01 6.06874108e-01 -5.86842537e-01 -2.92290211e-01 6.27256215e-01 -1.51611060e-01 -8.92315567e-01 -7.06098974e-01 -1.09141147e+00 -1.06524336e+00 -8.14982876e-02 -5.93258858e-01 1.43828228e-01 5.82283974e-01 6.19253218e-01 2.03837771e-02 5.56606531e-01 8.19753170e-01 -1.49652386e+00 -2.19027683e-01 -6.45271361e-01 -8.39922488e-01 3.27772141e-01 6.30373240e-01 -6.28640294e-01 -2.74227500e-01 7.38191009e-01]	[7.827647686004639, -2.151437282562256]
43644347-1dd6-4b1f-975c-1e29febd90c8	moso-decomposing-motion-scene-and-object-for	2303.03684	null	https://arxiv.org/abs/2303.03684v2	https://arxiv.org/pdf/2303.03684v2.pdf	MOSO: Decomposing MOtion, Scene and Object for Video Prediction	Motion, scene and object are three primary visual components of a video. In particular, objects represent the foreground, scenes represent the background, and motion traces their dynamics. Based on this insight, we propose a two-stage MOtion, Scene and Object decomposition framework (MOSO) for video prediction, consisting of MOSO-VQVAE and MOSO-Transformer. In the first stage, MOSO-VQVAE decomposes a previous video clip into the motion, scene and object components, and represents them as distinct groups of discrete tokens. Then, in the second stage, MOSO-Transformer predicts the object and scene tokens of the subsequent video clip based on the previous tokens and adds dynamic motion at the token level to the generated object and scene tokens. Our framework can be easily extended to unconditional video generation and video frame interpolation tasks. Experimental results demonstrate that our method achieves new state-of-the-art performance on five challenging benchmarks for video prediction and unconditional video generation: BAIR, RoboNet, KTH, KITTI and UCF101. In addition, MOSO can produce realistic videos by combining objects and scenes from different videos.	['Jing Liu', 'Xinxin Zhu', 'Weining Wang', 'Mingzhen Sun']	2023-03-07	null	http://openaccess.thecvf.com//content/CVPR2023/html/Sun_MOSO_Decomposing_MOtion_Scene_and_Object_for_Video_Prediction_CVPR_2023_paper.html	http://openaccess.thecvf.com//content/CVPR2023/papers/Sun_MOSO_Decomposing_MOtion_Scene_and_Object_for_Video_Prediction_CVPR_2023_paper.pdf	cvpr-2023-1	['video-generation', 'video-prediction', 'unconditional-video-generation', 'video-frame-interpolation']	['computer-vision', 'computer-vision', 'computer-vision', 'computer-vision']	[ 2.37113476e-01 -3.98563147e-01 -1.78829491e-01 1.67578608e-02 -4.07516539e-01 -2.77667493e-01 5.84054768e-01 -3.44012767e-01 1.37903377e-01 4.77359653e-01 2.15342090e-01 7.15938285e-02 2.55891353e-01 -5.03847718e-01 -1.07079208e+00 -6.92463279e-01 1.07667316e-02 2.46195689e-01 8.55590165e-01 1.56147346e-01 -4.42908844e-03 7.38752782e-02 -1.58697116e+00 9.31505620e-01 5.66497147e-01 1.20197427e+00 5.45755923e-01 1.03031683e+00 -1.91850230e-01 1.84666348e+00 -4.26475465e-01 -2.71249950e-01 1.59520313e-01 -5.28473496e-01 -8.97756755e-01 6.62025332e-01 5.30598998e-01 -8.40051472e-01 -9.37116861e-01 6.88415706e-01 1.54485539e-01 2.85987139e-01 4.88637000e-01 -1.58111441e+00 -4.99729604e-01 7.29884326e-01 -6.23733044e-01 2.24455982e-01 5.79757869e-01 4.39833075e-01 6.00771666e-01 -1.32306254e+00 9.24311936e-01 1.52605259e+00 3.33706230e-01 7.01609254e-01 -1.07911479e+00 -4.65740114e-01 5.68162203e-01 5.97839475e-01 -1.28028607e+00 -5.68688035e-01 7.25808680e-01 -7.85343647e-01 8.75428021e-01 2.90516883e-01 9.29928839e-01 1.02918673e+00 1.32423351e-02 1.38723111e+00 4.17235464e-01 -2.60954853e-02 1.09330714e-01 -1.40095577e-01 -1.56703323e-01 6.29727483e-01 -2.53903210e-01 -1.58568278e-01 -7.84146190e-01 -2.72716116e-02 7.70998716e-01 1.17065221e-01 -4.49341089e-01 -2.22022280e-01 -1.57666922e+00 2.37526536e-01 1.08268466e-02 -1.63966268e-02 -3.95070970e-01 5.74065447e-01 4.97498810e-01 -7.59359226e-02 2.52177596e-01 -4.75999236e-01 -1.40864760e-01 -2.57335097e-01 -1.19485450e+00 5.70015371e-01 5.69196463e-01 1.42918909e+00 4.85628128e-01 4.25668716e-01 -7.14066863e-01 7.22460985e-01 3.26061636e-01 3.44146341e-01 4.15214717e-01 -1.31688869e+00 7.38186657e-01 2.24867284e-01 4.39477712e-01 -1.04201293e+00 2.08111092e-01 2.24008501e-01 -8.29089463e-01 -2.91517645e-01 1.09741524e-01 1.31891966e-01 -1.12493002e+00 1.24309921e+00 5.35239279e-01 1.00749660e+00 -5.39296344e-02 9.73205447e-01 1.28453958e+00 1.40843236e+00 1.25293985e-01 -3.09982300e-01 1.16173482e+00 -1.62339044e+00 -5.70864439e-01 -8.53760839e-02 3.07587534e-01 -7.41948307e-01 7.48009384e-01 3.93493265e-01 -1.47760797e+00 -1.00759923e+00 -6.31584048e-01 -1.05747126e-01 3.56749743e-01 3.24689269e-01 3.15789849e-01 -3.76297571e-02 -8.92448187e-01 3.97549301e-01 -8.85010421e-01 6.96243569e-02 5.09529650e-01 -6.34475797e-02 6.10012421e-03 -3.19328636e-01 -9.14004028e-01 1.71885520e-01 6.73696697e-01 3.76487494e-01 -1.71808231e+00 -9.19114828e-01 -8.68514895e-01 1.77048072e-01 4.14647192e-01 -8.27878356e-01 1.13153243e+00 -1.20290685e+00 -1.21929646e+00 5.16697466e-01 -4.73570108e-01 -4.71928418e-01 6.99996173e-01 -2.46057346e-01 -3.39040637e-01 3.80092174e-01 4.62650172e-02 1.04985189e+00 1.18288684e+00 -1.42181265e+00 -1.17250597e+00 2.46910185e-01 -3.53956632e-02 4.55433503e-02 -7.79771945e-03 1.64188921e-01 -1.16058588e+00 -1.07423174e+00 -1.06511027e-01 -7.32613325e-01 -1.24941841e-01 -5.00547886e-02 -5.29553831e-01 -1.64523512e-01 9.92267370e-01 -9.16072667e-01 1.46426630e+00 -2.44698119e+00 4.36732322e-01 -2.19915077e-01 3.11413586e-01 1.80516601e-01 -2.62015283e-01 2.22679555e-01 3.60758156e-02 -5.99363223e-02 1.07275374e-01 -3.87715042e-01 -2.64365785e-02 9.83681455e-02 -6.43339515e-01 8.15414861e-02 2.59040177e-01 8.19728196e-01 -9.94793415e-01 -8.41783106e-01 4.33570653e-01 3.62645566e-01 -8.12650859e-01 2.96846449e-01 -5.49006283e-01 4.30672079e-01 -3.13099831e-01 7.63195455e-01 6.69024229e-01 -2.68352419e-01 1.36784896e-01 -2.99707174e-01 9.95251089e-02 1.66660011e-01 -1.27982545e+00 1.48231590e+00 -2.09442507e-02 7.71656036e-01 -2.11394370e-01 -7.43134737e-01 5.27537763e-01 4.63299721e-01 7.11066961e-01 -3.41392636e-01 -1.93259299e-01 -2.30593737e-02 -3.51234883e-01 -7.71515727e-01 8.93687665e-01 3.47325027e-01 2.07761079e-01 -1.29425555e-01 1.67528346e-01 2.17555106e-01 6.07104480e-01 5.56087255e-01 8.78217638e-01 6.80132270e-01 -1.51118366e-02 1.22900888e-01 6.61433637e-01 -9.42335464e-03 9.86941218e-01 6.23388886e-01 -2.58508384e-01 6.97331667e-01 5.37656546e-01 -5.00881732e-01 -1.23627782e+00 -1.15940118e+00 3.36667776e-01 9.67049539e-01 5.03049791e-01 -6.24597371e-01 -8.96271408e-01 -5.19623637e-01 -1.22590937e-01 5.43340802e-01 -4.54376966e-01 5.89341819e-02 -8.01528156e-01 -3.56022656e-01 2.17525631e-01 6.98481023e-01 3.85542780e-01 -1.33002555e+00 -3.19657236e-01 5.39114237e-01 -7.30714619e-01 -1.55028021e+00 -9.27633762e-01 -6.20574057e-01 -6.86999142e-01 -8.55406225e-01 -8.62738431e-01 -1.21138740e+00 5.65931499e-01 7.13774979e-01 1.09739161e+00 2.12486655e-01 -2.53215402e-01 3.28835517e-01 -5.64810276e-01 2.20972031e-01 -6.26853228e-01 -5.68454921e-01 -1.05369993e-01 6.47282183e-01 -1.19636282e-01 -3.97674620e-01 -7.50306070e-01 4.70525056e-01 -9.73478258e-01 6.89543903e-01 1.75894842e-01 6.89653635e-01 8.80005658e-01 4.01529185e-02 7.06795370e-03 -6.48544371e-01 -4.36079144e-01 -7.57510722e-01 -4.87616569e-01 4.29937899e-01 2.69330531e-01 -4.06989485e-01 7.19211161e-01 -7.27187514e-01 -1.26077986e+00 3.45957458e-01 1.06725872e-01 -1.05804265e+00 -2.82652415e-02 2.05760524e-01 -3.20216417e-01 2.96922028e-01 1.96587563e-01 7.78574169e-01 -6.16283476e-01 -3.65271926e-01 2.71036476e-01 3.96266907e-01 8.96202624e-01 -5.46235263e-01 7.96413600e-01 6.19998515e-01 -3.29991162e-01 -8.29275012e-01 -5.82993448e-01 -3.94954801e-01 -4.77781355e-01 -7.30747461e-01 1.06955373e+00 -1.27125025e+00 -5.88168979e-01 7.93024659e-01 -1.30312300e+00 -6.75264955e-01 -2.48415262e-01 2.85728544e-01 -6.77116156e-01 5.65435648e-01 -8.35509360e-01 -6.19637430e-01 -1.35748044e-01 -1.46880090e+00 1.27904224e+00 9.93008763e-02 1.12484306e-01 -7.84352183e-01 -4.26100522e-01 5.55879951e-01 -1.80359215e-01 3.63603145e-01 5.34393311e-01 6.03438728e-02 -1.31881893e+00 1.60720065e-01 -8.22321102e-02 2.69832820e-01 -4.25395630e-02 5.44283926e-01 -6.44336581e-01 -3.01343292e-01 -2.57592589e-01 -2.33650967e-01 1.03750026e+00 4.56877470e-01 1.49174631e+00 -5.41842401e-01 -3.53151858e-01 8.77048075e-01 1.30203879e+00 4.99308527e-01 9.23332036e-01 6.33137450e-02 1.20097148e+00 3.67462099e-01 8.41112733e-01 6.85783207e-01 6.95856690e-01 9.51846004e-01 2.62398362e-01 2.01496914e-01 -4.97168899e-01 -4.55937803e-01 8.69503140e-01 1.00353885e+00 -2.02940911e-01 -4.27356452e-01 -6.43548250e-01 7.30063617e-01 -2.13749647e+00 -1.60579002e+00 -4.60275829e-01 1.92436314e+00 5.47024608e-01 1.32856727e-01 3.49771082e-01 8.24837363e-04 7.74079025e-01 3.30323309e-01 -4.13977861e-01 1.06330283e-01 -2.13431790e-01 -3.30056489e-01 1.17814034e-01 2.84664035e-01 -1.13955760e+00 1.12698185e+00 5.65132284e+00 1.22334754e+00 -7.79389977e-01 6.18474633e-02 8.55175555e-01 -2.88877547e-01 -1.73917517e-01 1.33165330e-01 -9.12546515e-01 9.87253547e-01 5.79350531e-01 -1.46085009e-01 6.35524631e-01 8.70994329e-01 4.99147326e-01 -8.58353302e-02 -1.27776873e+00 1.11017025e+00 1.20723076e-01 -1.75405645e+00 5.56750894e-01 -3.60195547e-01 9.80815291e-01 -3.52649778e-01 -2.05880124e-02 3.22691023e-01 4.69020829e-02 -7.21488595e-01 1.64314425e+00 6.56583428e-01 5.97678483e-01 -5.26401579e-01 3.09709013e-01 2.51452893e-01 -1.84949064e+00 -1.68689787e-01 -4.33371007e-01 2.93542892e-01 5.01077116e-01 3.16968560e-01 -4.86602187e-01 5.54696858e-01 9.13799405e-01 1.17860723e+00 -3.66684884e-01 9.50407863e-01 1.04647055e-01 8.84535193e-01 1.15513451e-01 3.87311220e-01 1.95457712e-01 -5.09206913e-02 6.77432597e-01 1.30867290e+00 2.40840688e-01 8.51733908e-02 5.71596324e-01 8.73244762e-01 -2.30271220e-02 -1.98906481e-01 -8.45793709e-02 1.12820886e-01 5.70476115e-01 1.14725542e+00 -7.19749451e-01 -9.05920029e-01 -5.91096640e-01 1.12586999e+00 -5.95118664e-02 5.54341674e-01 -1.40620291e+00 7.11464044e-03 5.48366129e-01 1.64769366e-01 8.65497112e-01 -2.46390961e-02 3.26994598e-01 -1.51044881e+00 2.75194436e-01 -9.68545079e-01 1.83879554e-01 -9.35542464e-01 -7.82861471e-01 4.11141306e-01 2.88313013e-02 -1.59189451e+00 -1.44630075e-01 -3.09506744e-01 -5.54923236e-01 3.98845345e-01 -1.13753712e+00 -1.30864453e+00 -6.28327847e-01 7.44464993e-01 1.21250844e+00 -1.60556629e-01 7.61726350e-02 5.59683979e-01 -7.57342398e-01 3.59258980e-01 1.41575843e-01 3.23261499e-01 2.82585263e-01 -9.43277717e-01 7.24693358e-01 1.04209888e+00 2.17185870e-01 -7.29255602e-02 4.71222371e-01 -7.31965959e-01 -1.51933813e+00 -1.68815124e+00 5.52672923e-01 -2.90304571e-01 5.70412755e-01 -3.48209172e-01 -8.00996602e-01 7.51405835e-01 -8.60531405e-02 2.33294487e-01 2.04070553e-01 -9.14001226e-01 -8.01179633e-02 -1.18901141e-01 -5.86068988e-01 6.96745813e-01 1.24272418e+00 -2.84043550e-01 -2.32796714e-01 4.78274405e-01 1.05033863e+00 -8.09697092e-01 -8.30127954e-01 1.26116961e-01 5.59957862e-01 -8.74475300e-01 1.20287979e+00 -6.82661414e-01 1.02997446e+00 -6.26090884e-01 -3.55484426e-01 -7.72113681e-01 -4.81441379e-01 -7.87849963e-01 -6.72809362e-01 1.53861058e+00 5.01919873e-02 2.44099349e-01 7.42393911e-01 3.36927801e-01 -3.16545933e-01 -7.52102256e-01 -6.81845188e-01 -7.42652655e-01 -5.22103846e-01 -5.09155035e-01 5.68927526e-01 6.76288545e-01 -4.78138924e-01 -1.33637816e-01 -1.07168388e+00 1.38703987e-01 6.47294402e-01 2.46190369e-01 1.09776342e+00 -5.00602841e-01 -6.40738666e-01 -1.84405744e-01 -4.87496346e-01 -1.60513067e+00 1.00271918e-01 -7.44491994e-01 2.56389171e-01 -1.47588718e+00 5.66190183e-01 -1.78364977e-01 -5.77244721e-02 2.84932494e-01 -5.87317288e-01 3.02241564e-01 8.14232111e-01 4.66402799e-01 -1.03091729e+00 5.39125085e-01 1.46504998e+00 -4.42034215e-01 -2.59326935e-01 -1.87345475e-01 -2.77179442e-02 8.25318038e-01 1.41965300e-01 -2.90412515e-01 -3.20756346e-01 -6.06312573e-01 -1.96686253e-01 5.29148936e-01 6.92921579e-01 -1.08665323e+00 5.34086153e-02 -5.09746492e-01 5.35868049e-01 -8.54904115e-01 4.53213573e-01 -6.12789392e-01 7.02780366e-01 5.87440372e-01 -1.97412118e-01 -1.33692056e-01 9.59627852e-02 7.10187972e-01 -3.31088483e-01 5.21412119e-02 6.20016634e-01 -5.11703417e-02 -1.27845287e+00 7.60995984e-01 -5.10159731e-01 1.62571236e-01 1.28063107e+00 -4.63327855e-01 -3.01643908e-01 -2.34371573e-01 -8.81743133e-01 4.69545424e-01 3.85186285e-01 6.73529088e-01 1.05799985e+00 -1.59076607e+00 -8.58605623e-01 -1.52226970e-01 -6.35978505e-02 1.59042627e-01 8.19084823e-01 8.72311711e-01 -9.64768767e-01 1.45978466e-01 -6.89287018e-03 -8.90639842e-01 -1.44929349e+00 8.96327078e-01 1.22480236e-01 -6.91843703e-02 -8.38150442e-01 9.02013958e-01 7.74878085e-01 3.72286171e-01 3.14885765e-01 -4.47678298e-01 4.68210019e-02 -1.64912388e-01 7.69184232e-01 5.50569296e-01 -5.45618951e-01 -1.13670111e+00 -2.68509209e-01 3.77078563e-01 -1.13823391e-01 1.12491310e-01 1.05456626e+00 -1.42704055e-01 -7.85158798e-02 3.97846490e-01 9.91607845e-01 -1.89016864e-01 -1.73241270e+00 -7.67877884e-03 -3.13825548e-01 -7.94151366e-01 -4.60173011e-01 -1.69671342e-01 -1.25106263e+00 7.08295047e-01 1.48085803e-01 -6.35209633e-03 1.31787157e+00 -2.16812138e-02 1.17970359e+00 -1.12825736e-01 4.68478650e-01 -8.83843660e-01 3.68411750e-01 3.57512861e-01 8.53543222e-01 -7.85312176e-01 -1.49018809e-01 -6.70313060e-01 -8.99564683e-01 9.11140203e-01 8.48913252e-01 -1.43522158e-01 4.21658754e-01 8.68813246e-02 -4.68581766e-01 3.48623812e-01 -1.14458573e+00 1.55809104e-01 4.98677671e-01 4.32462633e-01 1.54269710e-01 5.61417500e-03 8.17743018e-02 5.50476909e-01 1.82190448e-01 2.13042542e-01 5.85948944e-01 8.44647586e-01 -3.80029708e-01 -6.98814750e-01 -4.25403029e-01 3.94757241e-01 -4.89821792e-01 -1.37671173e-01 -5.03904745e-02 4.33835059e-01 4.20803994e-01 7.38286495e-01 8.99791047e-02 -5.77069461e-01 5.59019074e-02 -2.53972679e-01 4.03596342e-01 -3.77048194e-01 -3.66873980e-01 4.06473726e-01 -9.98607278e-03 -9.45801318e-01 -5.84985852e-01 -7.60703027e-01 -1.15760350e+00 -3.94999743e-01 -8.76354724e-02 -1.40618570e-02 7.59251714e-02 7.43316650e-01 3.54806334e-01 8.65155339e-01 6.76785886e-01 -1.29155779e+00 -1.22564711e-01 -7.26129770e-01 -3.80385041e-01 7.93686807e-01 3.50313395e-01 -4.51687396e-01 -3.56834172e-03 1.06165349e+00]	[9.583178520202637, -0.033877067267894745]
b226f553-1c83-46ce-998d-4c84565f0209	online-video-streaming-super-resolution-with	2303.00334	null	https://arxiv.org/abs/2303.00334v3	https://arxiv.org/pdf/2303.00334v3.pdf	Online Streaming Video Super-Resolution with Convolutional Look-Up Table	Online video streaming has fundamental limitations on the transmission bandwidth and computational capacity and super-resolution is a promising potential solution. However, applying existing video super-resolution methods to online streaming is non-trivial. Existing video codecs and streaming protocols (\eg, WebRTC) dynamically change the video quality both spatially and temporally, which leads to diverse and dynamic degradations. Furthermore, online streaming has a strict requirement for latency that most existing methods are less applicable. As a result, this paper focuses on the rarely exploited problem setting of online streaming video super resolution. To facilitate the research on this problem, a new benchmark dataset named LDV-WebRTC is constructed based on a real-world online streaming system. Leveraging the new benchmark dataset, we proposed a novel method specifically for online video streaming, which contains a convolution and Look-Up Table (LUT) hybrid model to achieve better performance-latency trade-off. To tackle the changing degradations, we propose a mixture-of-expert-LUT module, where a set of LUT specialized in different degradations are built and adaptively combined to handle different degradations. Experiments show our method achieves 720P video SR around 100 FPS, while significantly outperforms existing LUT-based methods and offers competitive performance compared to efficient CNN-based methods.	['Lili Qiu', 'Dongsheng Li', 'Yuqing Yang', 'Huan Yang', 'Ningxin Zheng', 'Zhenhua Han', 'Shan Jiang', 'Xinyang Jiang', 'Zefan Qu', 'Guanghao Yin']	2023-03-01	null	null	null	null	['video-super-resolution']	['computer-vision']	[ 1.95318058e-01 -7.55967796e-01 -3.07061136e-01 -3.45894605e-01 -6.85409486e-01 -3.13180447e-01 5.65270474e-03 -2.15673611e-01 -2.23289698e-01 5.90686500e-01 1.98791444e-01 -9.00928117e-03 -6.62986888e-03 -7.96440303e-01 -7.20398903e-01 -5.89731276e-01 -4.20641243e-01 -3.34511250e-01 9.15736079e-01 -4.02469516e-01 3.98917526e-01 3.49332333e-01 -1.82005048e+00 8.11200857e-01 7.44054914e-01 1.40748763e+00 7.05242515e-01 8.60918105e-01 -3.02925795e-01 9.31623459e-01 -5.00339508e-01 -2.93415159e-01 3.84975851e-01 -4.06687021e-01 -4.15184051e-01 2.15061102e-03 3.96405220e-01 -8.84981811e-01 -7.36980796e-01 1.02975380e+00 7.62130439e-01 6.45964071e-02 -2.07463074e-02 -9.40956414e-01 -5.58458865e-01 7.61172295e-01 -6.63547218e-01 1.04788685e+00 3.72746080e-01 1.32483765e-01 7.46829033e-01 -6.48872793e-01 3.91451031e-01 1.15903962e+00 4.36996222e-01 5.21811783e-01 -8.04094732e-01 -8.17370832e-01 2.52719402e-01 8.47195327e-01 -1.38862085e+00 -6.62622273e-01 5.43567479e-01 7.62563571e-02 9.00444984e-01 7.83431679e-02 5.59040070e-01 1.00996602e+00 2.27943480e-01 5.76212347e-01 6.91783130e-01 1.87519208e-01 4.50469464e-01 -2.47821093e-01 -4.48923588e-01 1.97945967e-01 -8.48222747e-02 -1.11717030e-01 -9.74276125e-01 2.27746457e-01 1.16179371e+00 -8.04598108e-02 -6.17499828e-01 1.24120206e-01 -1.06555498e+00 2.31152475e-01 2.67839253e-01 1.54197067e-01 -2.53077298e-01 2.33164996e-01 7.86133170e-01 5.40003359e-01 4.46120262e-01 -3.10319424e-01 -4.40624356e-01 -5.32731295e-01 -1.29366457e+00 1.72428742e-01 4.84457254e-01 1.17470312e+00 3.71500403e-01 3.18764299e-01 -2.37897322e-01 9.15617168e-01 -6.29359707e-02 2.11704671e-01 6.60647869e-01 -1.23293078e+00 9.49313700e-01 2.18879148e-01 8.23521540e-02 -1.15000582e+00 -1.48365363e-01 -3.74161363e-01 -9.67302740e-01 -9.29111056e-03 -7.79533088e-02 1.28655046e-01 -5.82159758e-01 1.36215627e+00 2.78148025e-01 6.90132916e-01 7.12478608e-02 1.29229212e+00 7.79511213e-01 9.66738820e-01 -2.29489833e-01 -7.51130939e-01 1.19016171e+00 -1.00789881e+00 -9.15131569e-01 2.32484758e-01 8.55836831e-03 -5.98040342e-01 1.02797771e+00 6.83950841e-01 -1.52409041e+00 -6.62525654e-01 -1.27901113e+00 -6.87745661e-02 8.35418627e-02 -2.79967040e-01 1.98447913e-01 4.08699512e-01 -1.18115139e+00 7.86755502e-01 -8.11257303e-01 6.58941492e-02 4.04665321e-01 2.57224888e-01 8.71418789e-02 -4.31497879e-02 -1.37838519e+00 1.69523478e-01 4.14051324e-01 8.02314281e-02 -1.00873852e+00 -7.18353570e-01 -4.02389526e-01 3.14756900e-01 6.28892004e-01 -3.48066419e-01 1.29648173e+00 -1.20598292e+00 -1.90863764e+00 2.05967367e-01 -3.91087867e-02 -6.32959008e-01 6.42343342e-01 -9.72753242e-02 -7.80743599e-01 5.24485171e-01 -2.41220042e-01 1.91213161e-01 1.11665738e+00 -9.27345276e-01 -1.08014977e+00 -2.07335040e-01 2.78388292e-01 2.43213281e-01 -7.21766233e-01 4.01724488e-01 -7.36724794e-01 -6.81325376e-01 1.44357868e-02 -2.21159726e-01 4.89277616e-02 1.08587071e-01 2.72078484e-01 1.47893205e-01 9.89209771e-01 -6.68269277e-01 1.78637207e+00 -2.15867043e+00 1.38016388e-01 -2.75164992e-01 2.09772915e-01 4.24993396e-01 -1.87199134e-02 2.83262879e-01 1.68068200e-01 9.55333114e-02 -3.47959958e-02 -5.76518066e-02 -3.87415439e-01 1.28988981e-01 -2.57799774e-01 1.47142023e-01 -4.89082634e-02 2.31514528e-01 -8.64174724e-01 -7.51283348e-01 -2.10089162e-02 6.18421018e-01 -6.11319780e-01 2.63216674e-01 -1.59132540e-01 4.02268559e-01 -3.30556482e-01 8.43716800e-01 9.82364357e-01 -2.80728757e-01 5.62121905e-02 -2.85540819e-01 -4.55514193e-01 -3.93374823e-02 -1.41276383e+00 1.83172047e+00 -6.73738122e-01 5.25043666e-01 3.30400497e-01 -8.35015833e-01 7.85598159e-01 5.26510179e-01 4.45881695e-01 -1.03973508e+00 6.91648349e-02 3.37324739e-01 -4.00197268e-01 -9.09952641e-01 8.81528556e-01 2.14996830e-01 5.15124142e-01 -1.45008832e-01 -2.26747870e-01 4.85495478e-01 3.04422170e-01 1.22444034e-01 1.34383368e+00 1.74247518e-01 -1.02955066e-02 -4.77566980e-02 6.67326748e-01 -4.05680656e-01 9.17387366e-01 3.92595440e-01 -4.63312358e-01 7.89118409e-01 5.19898355e-01 -4.35724199e-01 -1.24296319e+00 -1.02624559e+00 -2.60234997e-02 1.13125455e+00 5.25637925e-01 -3.73333395e-01 -7.97861874e-01 -8.98165107e-02 -4.15377855e-01 2.20119640e-01 -1.68409213e-01 1.17494442e-01 -8.48064780e-01 -4.43497866e-01 2.76014835e-01 5.23763955e-01 1.03802490e+00 -8.46191943e-01 -8.81269693e-01 4.05355394e-01 -2.80390233e-01 -1.59513068e+00 -6.39792740e-01 -4.78540301e-01 -1.04382181e+00 -7.73704827e-01 -7.94598341e-01 -6.55465841e-01 -4.38181907e-02 6.98280036e-01 9.82713342e-01 2.63988748e-02 -2.71528922e-02 7.34080076e-02 -6.91852570e-01 2.41265029e-01 -3.19778025e-01 7.40701109e-02 -8.00310969e-02 3.07982355e-01 -2.37083033e-01 -9.41934407e-01 -1.15061867e+00 3.16975802e-01 -1.26911402e+00 1.76212296e-01 3.46824974e-01 5.20000935e-01 7.40081072e-01 3.76327127e-01 7.86136687e-01 -5.06244719e-01 3.55487376e-01 -8.30940247e-01 -7.46126115e-01 7.06614330e-02 -5.39652824e-01 -3.17609191e-01 1.28760052e+00 -6.23793840e-01 -1.32639086e+00 -2.86018163e-01 -1.72737089e-03 -8.33704352e-01 2.59973049e-01 2.94821113e-01 -2.73271561e-01 6.84868246e-02 3.37639123e-01 5.03180623e-01 -2.64749587e-01 -4.85950351e-01 -1.62295569e-02 9.23729181e-01 7.54098654e-01 -2.86883891e-01 5.23419321e-01 7.05424547e-01 -1.30611897e-01 -6.55458450e-01 -3.20220023e-01 -2.27240145e-01 -9.39982906e-02 -3.82886171e-01 6.06716454e-01 -1.20046389e+00 -6.49016142e-01 5.38139164e-01 -1.06418610e+00 -3.09931397e-01 5.10702543e-02 3.04268748e-01 -7.17165530e-01 5.64074159e-01 -9.80304837e-01 -6.28929496e-01 -4.68124330e-01 -1.28483379e+00 9.29766595e-01 4.53970045e-01 6.29852653e-01 -4.94994640e-01 -3.18824053e-01 2.83154368e-01 9.51605618e-01 4.67707813e-02 3.75931948e-01 -3.55214775e-02 -1.03998661e+00 2.46657073e-01 -6.57628298e-01 4.36320394e-01 2.95360908e-02 1.12871133e-01 -6.65063620e-01 -5.21643460e-01 9.71669331e-02 -3.32946852e-02 7.38837481e-01 1.82956055e-01 1.70517170e+00 -4.45543200e-01 1.72969177e-01 1.07655585e+00 1.75554895e+00 2.99973518e-01 9.36705947e-01 4.70507175e-01 5.14464438e-01 1.07407093e-01 7.99433827e-01 1.11178207e+00 4.33228463e-01 7.97205806e-01 6.69478595e-01 5.12033045e-01 -1.29743099e-01 -4.39767726e-02 6.03256404e-01 1.07274413e+00 -4.02840734e-01 -4.72262681e-01 -3.71700943e-01 4.34037715e-01 -1.81672335e+00 -1.09227479e+00 8.45924020e-02 2.26648593e+00 8.82583678e-01 1.96239695e-01 5.76878190e-02 2.16411084e-01 9.06691253e-01 4.32638943e-01 -7.46015131e-01 -3.05853158e-01 -1.44452438e-01 8.07658061e-02 6.42600417e-01 -1.11100793e-01 -7.57195234e-01 7.13232815e-01 5.82022333e+00 1.29868364e+00 -1.32750535e+00 2.85672128e-01 5.02884805e-01 -5.08220911e-01 -1.37864351e-01 -2.79522777e-01 -7.48082519e-01 9.22184944e-01 1.28521359e+00 -2.99666554e-01 8.14728260e-01 6.73805416e-01 7.43670881e-01 1.05853394e-01 -7.92950034e-01 1.37192690e+00 -6.10010475e-02 -1.41781509e+00 1.18300296e-01 -3.39214921e-01 6.65903509e-01 -5.40590100e-02 9.93108973e-02 1.21454589e-01 -3.59984696e-01 -6.94421768e-01 7.84531355e-01 4.85850424e-01 1.14138865e+00 -8.08186293e-01 6.76476657e-01 1.69762045e-01 -1.54842222e+00 -4.42856640e-01 -5.69145441e-01 1.74241755e-02 3.24382216e-01 4.29814428e-01 3.41076404e-02 5.64197898e-01 1.33224392e+00 7.01572180e-01 -2.67862618e-01 1.00118351e+00 3.59878778e-01 6.31275177e-01 -1.35377854e-01 2.41930157e-01 -3.98516208e-02 1.08009718e-01 5.90884268e-01 1.32911885e+00 6.50440931e-01 4.20978844e-01 -5.72456196e-02 4.91407365e-01 -1.25572518e-01 1.43328175e-01 -4.64003235e-02 2.69291729e-01 6.69383764e-01 1.18217647e+00 -6.08740330e-01 -4.81425285e-01 -8.08840513e-01 1.16325057e+00 4.21963409e-02 3.76909554e-01 -1.21870029e+00 -2.34474897e-01 7.14284003e-01 2.38899276e-01 7.14592695e-01 -3.17463875e-01 9.55825448e-02 -1.38177836e+00 3.27618748e-01 -1.02506757e+00 2.43304595e-01 -6.67626500e-01 -9.51341152e-01 5.63508153e-01 4.50221598e-02 -1.58661878e+00 8.86640251e-02 -1.67386770e-01 -4.29639429e-01 3.78779024e-01 -1.83033490e+00 -6.66384518e-01 -8.07088733e-01 7.69588292e-01 1.26089609e+00 -1.59926891e-01 1.08745776e-01 8.98478806e-01 -7.62300014e-01 7.16520846e-01 3.41058075e-01 -3.56814712e-01 6.04394078e-01 -6.89695776e-01 2.61905134e-01 8.94807518e-01 -4.78795141e-01 5.15118092e-02 7.93635368e-01 -4.08708751e-01 -1.76643133e+00 -1.10670936e+00 1.12462327e-01 5.02031207e-01 8.30487072e-01 -7.87946582e-02 -1.15849805e+00 1.08563006e-01 1.59237564e-01 4.57756609e-01 1.94428355e-01 -7.53140509e-01 -2.70462900e-01 -5.80137432e-01 -1.20834506e+00 5.64168751e-01 1.42647862e+00 -3.09561133e-01 1.79752544e-01 8.93019140e-02 1.25073969e+00 -6.46559954e-01 -1.07475436e+00 3.99576217e-01 4.80199575e-01 -1.42210639e+00 9.36635852e-01 -2.50641793e-01 7.94197917e-01 -2.62600303e-01 -4.55516368e-01 -8.70252907e-01 -2.01553419e-01 -6.96229458e-01 -5.71221769e-01 1.26101601e+00 -1.06958516e-01 -3.40687543e-01 6.70518696e-01 3.06910604e-01 -2.45304123e-01 -9.13509965e-01 -1.15890539e+00 -8.54709268e-01 -4.28908288e-01 -4.14056748e-01 7.18287766e-01 5.43218493e-01 -2.86987484e-01 -1.41360193e-01 -4.80150610e-01 4.07933682e-01 6.56007171e-01 -6.40143603e-02 4.40053582e-01 -5.03578484e-01 -5.60978651e-01 -3.62424612e-01 -5.27484417e-01 -1.31093073e+00 -3.39124560e-01 -2.93859839e-01 -2.18684897e-01 -1.22730350e+00 1.50854796e-01 -1.35028854e-01 -5.13840556e-01 -1.47020683e-01 6.79035559e-02 1.38544574e-01 1.30224362e-01 3.72585803e-01 -9.71708477e-01 5.51551700e-01 1.29160798e+00 1.26622602e-01 -2.72342205e-01 -4.29919548e-02 -1.87656462e-01 4.62942928e-01 7.75767028e-01 -8.52628723e-02 -5.30182123e-01 -6.91855609e-01 4.01400805e-01 7.32877553e-01 1.51657701e-01 -1.13491857e+00 4.15212810e-01 -3.29761654e-01 -7.76571780e-02 -5.36846399e-01 2.23047316e-01 -7.91351736e-01 1.48337752e-01 1.44659758e-01 -2.68662751e-01 1.66126996e-01 3.40787768e-02 7.86478937e-01 -3.53451729e-01 -4.74823564e-02 8.37375760e-01 -7.86666423e-02 -1.04226577e+00 6.51339591e-01 -2.80772567e-01 8.78173113e-02 8.25799167e-01 -4.21121836e-01 -3.49499553e-01 -4.38454479e-01 -4.16735083e-01 3.34865302e-01 5.17589808e-01 3.72714698e-01 9.44849491e-01 -1.18954802e+00 -8.77387345e-01 -1.82318501e-02 -2.28846267e-01 5.99070862e-02 8.76358449e-01 6.80036187e-01 -9.53957736e-01 -3.63883413e-02 -3.01172644e-01 -4.58068192e-01 -1.23813295e+00 8.02256823e-01 1.91803753e-01 -6.70083761e-02 -8.26197624e-01 6.19781852e-01 -7.86511227e-02 5.75959265e-01 3.56765747e-01 -2.02689484e-01 -3.70940655e-01 -1.01246096e-01 9.47843254e-01 7.84149766e-01 -4.23227698e-02 -5.19564688e-01 -6.99248016e-02 3.90259892e-01 -1.68474019e-02 1.11602329e-01 1.25599670e+00 -7.68787324e-01 2.06281040e-02 3.84988427e-01 1.20116603e+00 -3.11974794e-01 -1.61982417e+00 -2.93539882e-01 -2.94382989e-01 -9.51205134e-01 1.64213181e-01 -2.54925311e-01 -1.56441605e+00 6.59213364e-01 9.13790345e-01 3.35063994e-01 1.78068888e+00 -4.72057581e-01 1.46147621e+00 -4.06304225e-02 6.96234226e-01 -1.35740125e+00 1.25560626e-01 3.17323208e-01 6.96246028e-01 -1.22182882e+00 -2.67803948e-02 -5.01908779e-01 -3.44908953e-01 1.40059757e+00 7.33492553e-01 -8.90218243e-02 3.95715714e-01 5.58569551e-01 -2.84762919e-01 3.24751109e-01 -1.11794078e+00 1.19215310e-01 -3.00607622e-01 3.06627274e-01 3.16782385e-01 -1.51801497e-01 -6.00789249e-01 6.97784543e-01 1.11851528e-01 5.19385040e-01 7.55909681e-01 7.35067189e-01 -5.87100327e-01 -7.11670041e-01 -3.18814814e-01 2.33375400e-01 -7.68790781e-01 -7.94571862e-02 5.88055849e-01 2.62807667e-01 2.28209451e-01 8.58301044e-01 2.29462132e-01 -5.64681888e-01 1.66634679e-01 -6.26335144e-01 2.82979876e-01 -5.80300204e-02 -4.39563751e-01 7.42760897e-02 -1.16008326e-01 -1.04082978e+00 -5.03107667e-01 -5.06224930e-01 -1.35340512e+00 -5.61627090e-01 -1.75261542e-01 -2.29608089e-01 7.17983067e-01 6.47415340e-01 4.08374429e-01 8.30732703e-01 9.93466794e-01 -9.23193693e-01 -6.62708163e-01 -5.62017202e-01 -5.92120647e-01 1.69629380e-01 3.50638956e-01 -3.75106990e-01 -4.09824908e-01 2.21471161e-01]	[11.153438568115234, -1.7062418460845947]
51e216b5-ca23-4981-ac3d-40ba653cc624	light-coreference-resolution-for-russian-with	2306.01465	null	https://arxiv.org/abs/2306.01465v1	https://arxiv.org/pdf/2306.01465v1.pdf	Light Coreference Resolution for Russian with Hierarchical Discourse Features	Coreference resolution is the task of identifying and grouping mentions referring to the same real-world entity. Previous neural models have mainly focused on learning span representations and pairwise scores for coreference decisions. However, current methods do not explicitly capture the referential choice in the hierarchical discourse, an important factor in coreference resolution. In this study, we propose a new approach that incorporates rhetorical information into neural coreference resolution models. We collect rhetorical features from automated discourse parses and examine their impact. As a base model, we implement an end-to-end span-based coreference resolver using a partially fine-tuned multilingual entity-aware language model LUKE. We evaluate our method on the RuCoCo-23 Shared Task for coreference resolution in Russian. Our best model employing rhetorical distance between mentions has ranked 1st on the development set (74.6% F1) and 2nd on the test set (73.3% F1) of the Shared Task. We hope that our work will inspire further research on incorporating discourse information in neural coreference resolution models.	['Ivan Smirnov', 'Elena Chistova']	2023-06-02	null	null	null	null	['coreference-resolution']	['natural-language-processing']	[ 1.56118974e-01 7.45280921e-01 -3.61421794e-01 -2.39334345e-01 -1.02128148e+00 -6.92910910e-01 7.71270096e-01 2.63055801e-01 -7.69393146e-01 1.02373171e+00 1.04307044e+00 -2.53003299e-01 -2.45394170e-01 -5.95103085e-01 -6.93074286e-01 -2.61968464e-01 1.41666792e-02 1.05045271e+00 1.56205133e-01 -7.95907438e-01 3.44042152e-01 6.02345951e-02 -1.25987554e+00 8.90669107e-01 1.06572902e+00 2.22096086e-01 3.32067668e-01 6.17881417e-01 -1.01249799e-01 1.17017019e+00 -8.46249640e-01 -7.03598082e-01 -3.88928801e-01 -4.92382735e-01 -1.51743042e+00 -1.10237288e+00 8.08681309e-01 1.76325440e-01 -2.28487715e-01 8.82910967e-01 6.42544210e-01 3.41838419e-01 6.16428375e-01 -6.02950692e-01 -4.66821074e-01 1.70527220e+00 -3.43442559e-01 6.74291849e-01 3.37035656e-01 -4.67374206e-01 1.55870557e+00 -6.19292021e-01 1.22188056e+00 1.80940783e+00 5.89002252e-01 7.29886770e-01 -1.30553281e+00 -8.33715141e-01 1.04825400e-01 6.90642893e-01 -8.22703183e-01 -7.18913019e-01 7.95923352e-01 -2.90698737e-01 1.46636200e+00 4.39575344e-01 1.19079322e-01 1.26539326e+00 -1.17726192e-01 8.03698421e-01 8.18042397e-01 -5.32532394e-01 -2.63924956e-01 -4.54550534e-01 6.60977364e-01 3.00147176e-01 8.25404674e-02 2.46105149e-01 -5.83965659e-01 -7.00919405e-02 2.40004748e-01 -1.01150119e+00 -5.05393565e-01 4.29807752e-02 -1.03147674e+00 1.09918928e+00 7.42435634e-01 7.29742050e-01 -3.07550818e-01 -1.67805791e-01 5.45391321e-01 3.16663206e-01 1.61999375e-01 1.04851353e+00 -3.94853920e-01 -7.18749464e-02 -7.37810612e-01 4.54579115e-01 9.08067405e-01 5.03204405e-01 1.84904203e-01 -4.61153179e-01 -6.13654256e-01 1.22288966e+00 -5.94116971e-02 2.35401571e-01 4.00701284e-01 -1.45181441e+00 8.26128364e-01 3.91708583e-01 -7.99565986e-02 -1.04427755e+00 -8.18794370e-01 -4.86655056e-01 -4.83530134e-01 -2.67442107e-01 5.03813565e-01 -2.44579613e-01 -3.74615658e-03 2.35803819e+00 2.16986239e-02 8.90745372e-02 4.71517116e-01 8.45652819e-01 1.43489313e+00 4.74282742e-01 4.09040809e-01 -5.16479313e-01 1.52832115e+00 -8.06517184e-01 -8.62532020e-01 -9.10809338e-02 8.39212358e-01 -8.29785824e-01 5.64164579e-01 1.09899223e-01 -1.43989241e+00 -3.93092930e-01 -9.21706259e-01 -3.98656249e-01 1.22173429e-01 -8.75453800e-02 2.70092100e-01 2.63003279e-02 -8.21141183e-01 6.68643892e-01 -3.87869656e-01 -2.11641550e-01 -4.70120832e-02 2.65935510e-01 -2.78230727e-01 2.67858893e-01 -2.07059932e+00 1.59385920e+00 8.60358238e-01 -3.62191975e-01 -3.22270513e-01 -8.40117931e-01 -8.22577000e-01 2.04414204e-01 2.27240682e-01 -9.02485490e-01 1.59375274e+00 -8.52235496e-01 -1.10316539e+00 1.34534395e+00 -1.85493052e-01 -7.91866481e-01 1.27647489e-01 -5.26242495e-01 -4.81325001e-01 7.26257451e-03 2.06291795e-01 8.23863268e-01 -1.03746660e-01 -1.06740177e+00 -8.30366075e-01 -2.60280132e-01 2.05262855e-01 6.99617088e-01 1.15532272e-01 2.94407070e-01 1.60477415e-01 -4.43707645e-01 -2.20482096e-01 -8.56131434e-01 2.68678159e-01 -8.70491445e-01 -4.53924954e-01 -9.62432802e-01 4.27803308e-01 -7.33743012e-01 1.45556045e+00 -1.63425088e+00 5.51254451e-01 -2.98827797e-01 9.69323665e-02 5.37927449e-01 -2.93770224e-01 2.78334558e-01 -3.89261067e-01 6.22503348e-02 3.32047790e-02 -1.84504643e-01 -2.52875761e-04 -2.58061215e-02 -4.83203322e-01 6.90921843e-02 2.17405006e-01 9.24969494e-01 -1.14246857e+00 -6.38116956e-01 -1.55270934e-01 4.38870162e-01 -6.51373029e-01 -1.12551786e-02 -3.36173028e-01 3.08215261e-01 -1.24403626e-01 -2.26559922e-01 2.73266912e-01 9.37641263e-02 5.29790223e-01 -5.17380655e-01 -3.51929605e-01 1.31758475e+00 -7.59448528e-01 1.84535205e+00 -3.71203959e-01 7.19745874e-01 2.05319360e-01 -7.36934841e-01 9.10597980e-01 5.10647714e-01 -1.38480105e-02 -1.10563695e+00 2.53767937e-01 7.53072128e-02 5.94618857e-01 -2.92021424e-01 8.62344980e-01 -6.75648525e-02 -3.65726948e-01 3.88662219e-01 9.64923427e-02 4.20955867e-01 3.95109326e-01 2.94840425e-01 1.08677995e+00 7.81479552e-02 4.33499426e-01 -5.40728271e-01 6.40266240e-01 1.75302699e-01 7.71147370e-01 4.88885015e-01 -1.85576394e-01 5.43525875e-01 5.91788113e-01 -2.18933612e-01 -7.56760776e-01 -1.07708049e+00 -3.12444121e-01 1.51716840e+00 -7.39227980e-02 -4.54082578e-01 -9.29653823e-01 -5.76607347e-01 -2.62790114e-01 1.33955812e+00 -6.72228277e-01 3.42401005e-02 -1.42752349e+00 -4.66069132e-01 1.07377756e+00 4.10071880e-01 3.76271963e-01 -1.48951066e+00 -7.04511702e-01 3.37100029e-01 -1.07818043e+00 -8.38681221e-01 -3.49800050e-01 9.04443860e-02 -6.23984098e-01 -1.48665714e+00 -2.63837695e-01 -1.03469419e+00 -2.12140158e-01 -1.48037955e-01 1.65670931e+00 1.81290209e-01 2.46704742e-01 -1.64454281e-01 -1.10174261e-01 -2.02199474e-01 -5.37377477e-01 6.32191181e-01 -3.82612981e-02 -7.72771537e-01 7.04752326e-01 -4.21408892e-01 -2.19323725e-01 -9.24044177e-02 -1.35723084e-01 1.98557749e-01 1.65515825e-01 9.25229430e-01 2.68850893e-01 -7.19141424e-01 8.80908310e-01 -1.29782832e+00 1.03818572e+00 -5.08219123e-01 -2.91261077e-01 3.18064839e-01 -2.74952471e-01 3.81692976e-01 2.07916304e-01 -2.51945406e-01 -1.45403790e+00 -4.92372036e-01 -2.90658265e-01 -6.37663603e-02 -2.99849324e-02 5.39419651e-01 -2.45477095e-01 7.31095135e-01 1.02577305e+00 -6.31193876e-01 -2.94858664e-01 -4.26710606e-01 7.68861890e-01 4.38871115e-01 1.11717558e+00 -1.00697434e+00 1.99188441e-01 -2.26002097e-01 -3.19441766e-01 -6.97493494e-01 -1.27903926e+00 -3.75439256e-01 -7.30795205e-01 1.52930841e-01 9.51615930e-01 -8.86632323e-01 -9.57906067e-01 -1.97055578e-01 -1.88423884e+00 -2.21955091e-01 2.56377496e-02 3.88289869e-01 -6.18165016e-01 2.11392701e-01 -8.07149529e-01 -3.27090174e-01 -8.66464972e-01 -7.44835615e-01 6.08060181e-01 2.68236190e-01 -1.34648001e+00 -9.89917934e-01 8.32004547e-01 5.59592783e-01 2.16657460e-01 2.96257079e-01 1.30211926e+00 -9.05858696e-01 3.36370654e-02 7.60827184e-01 -2.18821049e-01 -4.49618101e-01 -2.19677195e-01 -2.10438266e-01 -1.03505480e+00 -8.32276270e-02 -2.52721071e-01 -3.47065628e-01 1.14233208e+00 4.88385707e-01 3.03710252e-01 -4.10411298e-01 -5.47596753e-01 3.72926295e-01 8.83543313e-01 7.02540129e-02 6.62370324e-01 6.36521518e-01 4.45838779e-01 9.58478749e-01 7.00838506e-01 -1.39706910e-01 6.75384402e-01 9.02815104e-01 2.39997223e-01 3.16101879e-01 -5.35850465e-01 -3.23402703e-01 2.26994559e-01 8.81839573e-01 -6.24533594e-02 -8.39763507e-02 -1.00997388e+00 8.04459512e-01 -1.95989394e+00 -1.54839623e+00 -3.67094368e-01 1.77399623e+00 1.44669580e+00 1.04338072e-01 -3.46306823e-02 -5.80901764e-02 1.11241925e+00 1.74169615e-01 -2.69967914e-01 -6.26742125e-01 -4.40224886e-01 3.06444377e-01 -1.14259329e-02 1.05984116e+00 -1.30251443e+00 1.32672274e+00 5.48006725e+00 3.99715155e-01 -7.37103522e-01 8.98821279e-02 1.74231738e-01 -2.43433341e-01 -3.86428684e-01 -2.15358268e-02 -1.00041652e+00 1.39900863e-01 1.24591637e+00 -1.97048545e-01 1.25884801e-01 3.03587705e-01 -1.10170677e-01 1.28718242e-01 -1.33105898e+00 7.18016088e-01 1.91112980e-01 -1.46903741e+00 -2.13494115e-02 -3.37121576e-01 5.55050850e-01 1.72083631e-01 -3.37117910e-01 6.01563454e-01 8.81842732e-01 -1.23831701e+00 4.59017187e-01 5.02272785e-01 5.41407943e-01 -9.91968930e-01 7.75675774e-01 3.98866445e-01 -8.23293567e-01 7.51490518e-02 -2.64904439e-01 -2.03487068e-01 4.36348230e-01 8.04546624e-02 -9.93545055e-01 4.26553041e-01 5.67017257e-01 5.06035388e-01 -3.22221756e-01 7.35310733e-01 -6.74980342e-01 5.99167824e-01 -5.30142561e-02 -1.32634034e-02 -9.37897805e-03 4.50804025e-01 9.86979485e-01 1.64175236e+00 -1.28337413e-01 4.60624456e-01 -8.91291499e-02 8.84544253e-01 -4.14462507e-01 1.78933412e-01 -4.09772336e-01 3.67913663e-01 1.17483735e+00 1.03156567e+00 -8.96988437e-02 -2.47194935e-02 -2.53697634e-02 4.28739995e-01 1.20078814e+00 -2.74702851e-02 -5.53142011e-01 -2.70272195e-01 7.24511504e-01 -2.66739637e-01 1.19884744e-01 2.52439022e-01 -3.35022658e-01 -7.74468184e-01 -5.74168026e-01 -1.02587950e+00 9.37859714e-01 -6.40866697e-01 -1.28674436e+00 6.95613444e-01 2.11517304e-01 -3.73782814e-01 -7.65654683e-01 -4.56338882e-01 -8.25599611e-01 9.92133677e-01 -1.40554357e+00 -9.59646821e-01 1.96913049e-01 3.68250400e-01 4.50785547e-01 -2.93394983e-01 1.26046252e+00 7.36423060e-02 -3.56174290e-01 7.37546504e-01 -3.26636791e-01 7.06907034e-01 1.08260703e+00 -1.20163977e+00 3.41371596e-01 6.61647499e-01 1.49277121e-01 9.39252913e-01 1.02643847e+00 -6.39086783e-01 -5.02504766e-01 -8.45528305e-01 1.61347032e+00 -7.50065267e-01 5.53665578e-01 2.14319795e-01 -1.12692225e+00 7.99617589e-01 9.07059729e-01 -9.01483953e-01 8.48952353e-01 1.03190780e+00 -5.86439490e-01 3.73815417e-01 -8.46817374e-01 6.22060955e-01 1.19822156e+00 -4.69977587e-01 -1.82030284e+00 -1.72648072e-01 1.07113981e+00 -7.04769611e-01 -1.12677276e+00 7.01591790e-01 1.68386698e-01 -6.32899463e-01 1.01403749e+00 -1.00491095e+00 6.88309610e-01 2.44200304e-02 -2.01903760e-01 -1.47395599e+00 -7.16872275e-01 -3.42532277e-01 -3.75920534e-01 1.55109262e+00 6.52319551e-01 -6.37752414e-02 3.15479249e-01 3.21336865e-01 -2.79751241e-01 -2.52348065e-01 -1.14710391e+00 -2.99645931e-01 8.11016440e-01 4.69395742e-02 3.60815078e-01 1.37415969e+00 5.64747572e-01 1.37128401e+00 8.35556537e-02 -9.54552367e-02 6.07500553e-01 3.61989200e-01 3.37988615e-01 -1.83325362e+00 -1.11311547e-01 -1.00366342e+00 2.80992001e-01 -5.32903075e-01 9.09544289e-01 -1.34827936e+00 -3.65522802e-02 -1.44505847e+00 3.77306461e-01 -4.40526664e-01 -3.14314097e-01 3.88261169e-01 -2.71462649e-01 -1.73720792e-01 4.13459182e-01 2.02755362e-01 -6.94870770e-01 4.25101489e-01 9.34086502e-01 -1.83820426e-01 -2.04564855e-01 -4.40255046e-01 -9.63566601e-01 7.96830356e-01 9.19699550e-01 -4.49273616e-01 -2.32545301e-01 -6.63828671e-01 2.76090711e-01 2.79174328e-01 5.28486595e-02 -5.66080511e-01 5.12991190e-01 -4.60749380e-02 8.57560635e-02 -6.52790785e-01 2.41090491e-01 -8.32508653e-02 -1.69399247e-01 4.16360468e-01 -9.58861768e-01 2.19635293e-01 3.10440809e-01 -6.30828068e-02 -2.26195171e-01 -4.16436255e-01 6.21297956e-01 -1.31738901e-01 -6.93598926e-01 -4.51576829e-01 4.49681692e-02 9.30635452e-01 4.40401703e-01 4.48710740e-01 -1.07591915e+00 -1.20074056e-01 -7.96144187e-01 5.07998168e-01 2.63617277e-01 9.25962210e-01 2.96386063e-01 -1.27200210e+00 -1.39642465e+00 -4.61653292e-01 -1.47569448e-01 -6.91484064e-02 2.16640905e-02 5.42615294e-01 -3.47000510e-02 8.24637711e-01 -3.62461269e-01 -3.48191470e-01 -1.73653543e+00 4.53433961e-01 5.23726463e-01 -7.19251454e-01 -3.93674284e-01 8.95905852e-01 1.76852152e-01 -7.28442073e-01 3.86520594e-01 1.54273450e-01 -1.08898437e+00 5.58464348e-01 7.22183883e-01 4.60265845e-01 -5.24372049e-02 -1.07318950e+00 -3.58744949e-01 3.47296715e-01 -3.81837964e-01 -2.81746626e-01 1.18805802e+00 1.18420990e-02 -2.92262435e-01 4.60034579e-01 9.04515564e-01 1.46488100e-01 -7.01302350e-01 -4.50748891e-01 5.39563894e-01 3.54186058e-01 -1.36860773e-01 -1.15711391e+00 -6.49666190e-01 8.50667715e-01 1.49010241e-01 -1.80172011e-01 5.87527335e-01 2.01289356e-01 5.40649593e-01 6.36584759e-01 1.30841613e-01 -1.12641716e+00 -3.69737685e-01 1.33724070e+00 1.26329684e+00 -8.64242554e-01 -2.61527181e-01 -2.33600572e-01 -6.72395170e-01 9.17489290e-01 9.02805984e-01 -2.15056032e-01 8.57871026e-02 2.03888461e-01 1.13776684e-01 -2.93157429e-01 -1.20030355e+00 -9.16924849e-02 3.48225653e-01 5.06706536e-01 1.28448772e+00 2.27869168e-01 -8.57371986e-01 1.03123248e+00 -8.36568654e-01 -4.57702577e-01 2.98155487e-01 6.07178845e-02 -6.15014911e-01 -1.08691764e+00 -3.89685929e-01 -6.80221841e-02 -6.06405318e-01 -4.31313068e-01 -7.76221573e-01 7.42393792e-01 1.42352972e-02 1.03939545e+00 4.34044093e-01 -1.63147211e-01 4.99454558e-01 2.73987591e-01 5.23287714e-01 -4.97498959e-01 -8.80856931e-01 -4.66534108e-01 9.65205371e-01 -4.36995238e-01 -5.41215599e-01 -9.78033900e-01 -1.68002474e+00 -4.16711301e-01 -1.93190843e-01 5.14429510e-01 -8.43043104e-02 9.77915287e-01 1.99213728e-01 7.54120290e-01 -5.22870421e-02 -5.47405183e-01 -3.29463035e-01 -1.28450680e+00 1.84997469e-01 6.24978185e-01 2.30043873e-01 -7.57289112e-01 -8.90590367e-04 -2.77939439e-01]	[9.343486785888672, 9.539780616760254]
f997da2d-b1ca-4ebb-bfa1-96e54409d32e	translation-of-algorithmic-descriptions-of	1805.07239	null	https://arxiv.org/abs/1805.07239v5	https://arxiv.org/pdf/1805.07239v5.pdf	Translation of Algorithmic Descriptions of Discrete Functions to SAT with Applications to Cryptanalysis Problems	In the present paper, we propose a technology for translating algorithmic descriptions of discrete functions to SAT. The proposed technology is aimed at applications in algebraic cryptanalysis. We describe how cryptanalysis problems are reduced to SAT in such a way that it should be perceived as natural by the cryptographic community. In~the theoretical part of the paper we justify the main principles of general reduction to SAT for discrete functions from a class containing the majority of functions employed in cryptography. Then, we describe the Transalg software tool developed based on these principles with SAT-based cryptanalysis specifics in mind. We demonstrate the results of applications of Transalg to construction of a number of attacks on various cryptographic functions. Some of the corresponding attacks are state of the art. We compare the functional capabilities of the proposed tool with that of other domain-specific software tools which can be used to reduce cryptanalysis problems to SAT, and also with the CBMC system widely employed in symbolic verification. The paper also presents vast experimental data, obtained using the SAT solvers that took first places at the SAT competitions in the recent several years.	['Oleg Zaikin', 'Ilya Otpuschennikov', 'Alexander Semenov', 'Stepan Kochemazov', 'Irina Gribanova']	2018-05-17	null	null	null	null	['cryptanalysis']	['miscellaneous']	[ 2.21779376e-01 1.67117000e-01 4.68941629e-01 -5.60794115e-01 -5.11461318e-01 -9.22082663e-01 5.92939734e-01 -1.33056894e-01 5.78793744e-03 8.91593397e-01 -3.46959472e-01 -8.76513124e-01 -3.85384291e-01 -1.11066258e+00 -5.60845971e-01 -5.01640618e-01 -1.92026213e-01 9.19234812e-01 2.14906499e-01 -1.01973379e+00 4.82826442e-01 6.55215085e-01 -1.30080605e+00 5.81718862e-01 5.58785439e-01 6.34897590e-01 -4.01849359e-01 7.92794228e-01 -3.58463436e-01 5.66206574e-01 -7.50098646e-01 -6.15333974e-01 7.50587881e-01 -6.54930949e-01 -1.41494274e+00 -2.35987559e-01 1.94927096e-01 1.25517875e-01 -2.90017843e-01 1.42067266e+00 -2.00966187e-02 -3.95641536e-01 3.10865015e-01 -1.69140518e+00 1.43655226e-01 1.20214403e+00 1.11256719e-01 -5.79706244e-02 7.99639940e-01 1.25196517e-01 8.28337252e-01 -1.25872254e-01 7.86320210e-01 1.27614558e+00 4.85286713e-01 3.34796488e-01 -1.17031717e+00 -7.98825324e-01 -2.84206778e-01 4.02617365e-01 -1.65904236e+00 -3.15784514e-01 5.53182662e-01 -8.20126161e-02 1.14775050e+00 8.16490114e-01 8.24265897e-01 2.90440202e-01 6.51729107e-01 1.89168498e-01 1.45179415e+00 -7.82541573e-01 3.13771695e-01 5.48357904e-01 3.88035744e-01 5.79583764e-01 8.30045104e-01 -5.00927679e-02 4.30953428e-02 -4.04184312e-01 4.25465822e-01 -5.51288784e-01 -1.71476364e-01 -2.41439790e-01 -1.07285297e+00 9.07634377e-01 1.53267518e-01 6.82190895e-01 -1.20389871e-02 3.67220372e-01 5.73582709e-01 8.17591429e-01 -1.70611769e-01 6.56295002e-01 -4.19504195e-01 -5.04742451e-02 -7.94071376e-01 5.81075609e-01 1.57945406e+00 1.14212978e+00 4.86188799e-01 2.83250287e-02 6.28105581e-01 -4.51027602e-01 4.38805461e-01 4.20706064e-01 5.21435663e-02 -7.12512493e-01 4.75634068e-01 6.96014285e-01 -1.23411536e-01 -7.42562175e-01 -1.95277650e-02 -2.11834401e-01 -2.29331523e-01 5.66470146e-01 4.82654512e-01 -1.25117019e-01 -4.68561769e-01 1.31384110e+00 2.16160625e-01 -1.99190918e-02 5.47407746e-01 4.48818892e-01 6.95427418e-01 8.56946349e-01 -6.15732491e-01 -4.00899589e-01 1.13947451e+00 -4.79998082e-01 -8.48287702e-01 5.03927350e-01 8.72471333e-01 -9.47534502e-01 2.71433175e-01 1.21642244e+00 -1.41187274e+00 -1.24267489e-01 -1.39877570e+00 2.87969828e-01 -4.93921250e-01 -5.26886880e-01 1.09884655e+00 1.39143145e+00 -1.24988365e+00 5.17852306e-01 -7.04382718e-01 -9.82613415e-02 -1.18248895e-01 1.16290760e+00 -3.50019068e-01 -4.84266542e-02 -1.12507594e+00 1.01477730e+00 6.73741162e-01 2.03771800e-01 -4.09714311e-01 -8.78000930e-02 -7.68306553e-01 3.19575429e-01 4.43738490e-01 -3.69261026e-01 1.38511455e+00 -8.63229036e-01 -1.73520231e+00 6.91250622e-01 1.11488305e-01 -7.86471546e-01 5.01420200e-01 3.94710392e-01 -4.81235713e-01 3.86576466e-02 -2.95707762e-01 1.54531121e-01 1.09212510e-01 -7.97634661e-01 -3.73351365e-01 5.86158782e-03 6.02839172e-01 -2.59405792e-01 3.60267878e-01 2.45793834e-01 1.74715072e-01 1.31437063e-01 3.04535955e-01 -9.03099895e-01 -3.85710359e-01 -5.03734946e-01 -2.08849713e-01 5.26949875e-02 6.70922756e-01 -8.28638226e-02 1.02083302e+00 -1.79321373e+00 3.46303344e-01 9.88859832e-01 1.48073621e-02 3.14263284e-01 3.29471737e-01 1.04560447e+00 -4.43651140e-01 7.65348971e-02 -2.00436115e-02 2.78793991e-01 5.46788812e-01 2.78571010e-01 -5.08286834e-01 9.33496356e-01 -1.36832938e-01 6.24299407e-01 -5.45025229e-01 -6.10787153e-01 2.20139757e-01 4.27729599e-02 -7.71699667e-01 -2.17182398e-01 -4.91637200e-01 -5.76148108e-02 -6.02899194e-01 5.26117861e-01 1.10656464e+00 2.93427527e-01 5.95552981e-01 2.54269809e-01 -4.00992960e-01 7.05376208e-01 -1.71954918e+00 1.54343581e+00 -1.42648175e-01 5.95282793e-01 2.37001494e-01 -1.21274805e+00 9.93986368e-01 6.34127975e-01 1.08391464e-01 -1.40896931e-01 7.19662309e-01 5.43790877e-01 6.03017926e-01 8.73324797e-02 2.74612904e-01 -1.75690830e-01 -1.94335073e-01 4.38132972e-01 -8.29292014e-02 -1.00818443e+00 7.04036355e-01 1.40390396e-01 9.02406991e-01 4.74595502e-02 5.81066489e-01 -5.94775438e-01 1.35467863e+00 6.55333757e-01 2.20542610e-01 4.16747034e-01 1.25747502e-01 3.71220857e-02 9.33895469e-01 -5.91788113e-01 -8.35675716e-01 -4.37743157e-01 -1.54194072e-01 1.64287612e-01 1.11180976e-01 -9.71441090e-01 -8.76426995e-01 -2.49541327e-01 -2.86993444e-01 8.38024974e-01 -1.87763616e-01 -2.58105807e-02 -7.38596857e-01 -1.96612716e-01 8.30572426e-01 -9.41980705e-02 5.02245009e-01 -8.27748299e-01 -3.42971027e-01 2.19051808e-01 3.02524745e-01 -9.48106289e-01 3.85552645e-01 6.06920898e-01 -6.94024801e-01 -1.19621718e+00 2.59737462e-01 -8.70057940e-01 5.87449670e-01 7.70185068e-02 8.86089802e-01 3.96241218e-01 -2.40802646e-01 -1.08623160e-02 -4.65847880e-01 -5.97134948e-01 -9.53204691e-01 -2.20613349e-02 -1.68458760e-01 -5.34291029e-01 2.28394359e-01 -4.39827025e-01 2.23486751e-01 1.49141833e-01 -1.00568628e+00 -6.36930317e-02 3.19688618e-01 6.71269178e-01 1.08609304e-01 6.65206373e-01 -1.91027120e-01 -1.21759677e+00 2.57282734e-01 -3.26698311e-02 -1.54987669e+00 9.40720811e-02 -3.38028640e-01 2.98050106e-01 9.30260181e-01 -8.59234706e-02 -5.24458468e-01 7.56688714e-02 -4.24400359e-01 3.24047744e-01 8.82293209e-02 4.74213421e-01 -6.05314314e-01 -9.01130497e-01 6.24821901e-01 2.58055508e-01 -2.59894226e-02 1.19076349e-01 -5.75838350e-02 5.44429660e-01 2.13382304e-01 -8.33294630e-01 1.25298238e+00 1.42146736e-01 8.29750240e-01 -4.35324937e-01 -4.35600467e-02 -2.14326724e-01 -1.33746818e-01 3.62610757e-01 1.97810367e-01 -5.24001896e-01 -1.13400602e+00 1.55662358e-01 -1.17518008e+00 -8.99126288e-03 -2.93249846e-01 5.00672758e-01 -6.62264884e-01 4.13054526e-01 -3.03276271e-01 -8.35343540e-01 -3.42996359e-01 -1.59447575e+00 6.30829871e-01 1.19604208e-01 -2.04767764e-01 -8.68831158e-01 9.36082527e-02 4.80783023e-02 2.46224359e-01 3.88645172e-01 7.82255054e-01 -1.02814305e+00 -8.98327410e-01 -4.57553804e-01 2.36944184e-01 3.95455182e-01 -2.36747250e-01 2.33730748e-01 -6.81406617e-01 -3.13473731e-01 3.69319975e-01 -1.13065548e-01 -3.70037034e-02 -2.00709235e-02 7.42037714e-01 -9.08696577e-02 -1.96993604e-01 6.21472836e-01 1.76215434e+00 3.93592566e-01 1.13060570e+00 6.47911370e-01 -2.55443990e-01 1.32530332e-01 9.11824167e-01 5.14204837e-02 -1.42455816e-01 7.15764940e-01 5.07759273e-01 2.49419823e-01 5.09084463e-01 4.16851103e-01 4.94094789e-01 5.85740089e-01 -1.76259398e-01 -2.27544666e-03 -8.80854130e-01 8.26795772e-02 -1.46897745e+00 -1.11571646e+00 -7.45920122e-01 2.08319211e+00 7.90541410e-01 6.97357178e-01 -7.02923983e-02 8.25357199e-01 5.34600377e-01 -3.57752502e-01 7.41802573e-01 -1.37418401e+00 1.52552143e-01 1.24376392e+00 6.17236793e-01 9.22611952e-01 -8.26527119e-01 9.67351675e-01 6.45958853e+00 6.70861423e-01 -1.25871110e+00 -1.81416735e-01 -1.49134994e-01 4.22865361e-01 -1.92330658e-01 8.24768245e-01 -7.81717539e-01 -3.49532925e-02 1.38747966e+00 -6.25269413e-01 6.66850269e-01 8.78081322e-01 -2.38644302e-01 -1.08725026e-01 -1.36476195e+00 5.13601303e-01 -1.93137124e-01 -1.43208897e+00 -1.49448499e-01 1.26954049e-01 4.62138265e-01 -4.60626662e-01 -1.42518610e-01 3.91451329e-01 1.34263441e-01 -1.16446972e+00 7.01684594e-01 -4.59921248e-02 4.94988620e-01 -1.02610970e+00 1.30059588e+00 2.60055035e-01 -1.12628758e+00 1.68886438e-01 -4.45885152e-01 -7.19680667e-01 -2.99841851e-01 -2.22428337e-01 -1.32655191e+00 1.04902875e+00 -2.58659180e-02 -3.67529206e-02 -1.89377606e-01 1.35614598e+00 -9.95058939e-02 4.95026886e-01 -5.37782311e-01 -5.28903425e-01 5.54445207e-01 -5.09027183e-01 4.17918921e-01 1.31502688e+00 3.32885385e-01 3.14170271e-01 -5.75831309e-02 8.84279430e-01 6.36982977e-01 9.89035293e-02 -5.13955951e-01 -1.47612005e-01 1.99371427e-01 1.00749826e+00 -9.38192248e-01 -4.72596765e-01 -3.08760494e-01 3.97218466e-01 -6.48406982e-01 -2.43841141e-01 -1.08531094e+00 -8.25200021e-01 4.59180325e-01 1.04794726e-01 2.75668979e-01 -2.32177287e-01 -3.16983789e-01 -1.04612553e+00 -1.48889512e-01 -1.41662335e+00 3.20455544e-02 -5.80629885e-01 -3.91365230e-01 8.13217282e-01 5.99170089e-01 -1.23235846e+00 -6.61070764e-01 -1.05401468e+00 -6.83701813e-01 1.17618310e+00 -1.21105862e+00 -1.02155459e+00 -4.86817732e-02 6.89269602e-01 1.54271737e-01 -2.12757871e-01 1.16908324e+00 1.26077369e-01 -3.04192360e-02 4.59468484e-01 -1.11212246e-01 -1.85914367e-01 3.52034897e-01 -1.39648890e+00 5.96138537e-01 1.07106221e+00 1.37447640e-01 1.10251606e+00 1.18384230e+00 -4.82052833e-01 -2.15240240e+00 -3.31308454e-01 1.14268279e+00 -2.13988591e-02 8.53706658e-01 -5.66031218e-01 -4.72582132e-01 8.75136852e-01 1.67393401e-01 -1.10859804e-01 3.32241833e-01 -2.01883107e-01 -1.78078353e-01 -1.98996335e-01 -1.49462879e+00 1.40353054e-01 5.61938941e-01 -2.34917566e-01 -9.49373782e-01 6.76588237e-01 4.73993331e-01 -9.61848021e-01 -5.88991761e-01 -1.25388980e-01 3.38651001e-01 -1.06659460e+00 7.75867522e-01 -4.69526559e-01 -6.96027577e-02 -6.50880635e-01 -3.10075790e-01 -8.67531359e-01 -4.38396670e-02 -1.30835474e+00 4.31103379e-01 8.78541827e-01 2.35108286e-01 -1.11598122e+00 7.35493422e-01 1.74122036e-01 -3.06182593e-01 -1.42068639e-01 -1.13512671e+00 -6.01095498e-01 2.79339343e-01 -3.56325090e-01 8.40687811e-01 7.44165897e-01 5.90777457e-01 -1.21043421e-01 1.99769214e-01 3.49163890e-01 1.90579534e-01 3.12308490e-01 1.14297760e+00 -1.32326925e+00 -7.62982249e-01 -3.29674453e-01 -1.14983249e+00 -1.80596128e-01 6.60226448e-03 -9.48096037e-01 -3.79405677e-01 -1.15637970e+00 -2.53201663e-01 -1.73683405e-01 3.40841599e-02 3.96399885e-01 6.58965051e-01 2.02263653e-01 2.68492401e-01 -3.56651574e-01 -2.83781976e-01 -3.90248656e-01 1.00269723e+00 -1.33171111e-01 2.63399072e-02 2.49662891e-01 -7.10941195e-01 5.92372835e-01 4.56677377e-01 -5.83225429e-01 -3.77579093e-01 2.02063888e-01 4.73567665e-01 2.65451789e-01 5.47481999e-02 -1.36783564e+00 2.74341911e-01 -4.33410287e-01 -1.56472057e-01 -5.21504223e-01 3.69015962e-01 -1.31153691e+00 8.67195427e-01 1.20219648e+00 1.42079279e-01 2.62824625e-01 3.82662177e-01 -4.21680450e-01 -4.56041038e-01 -7.29892135e-01 5.40268242e-01 -2.22059280e-01 -4.18400884e-01 -3.59959066e-01 -3.49724323e-01 -2.90173471e-01 1.27601981e+00 -4.51952636e-01 -2.96284586e-01 -4.92116839e-01 -5.46786308e-01 2.98406817e-02 6.34755611e-01 -5.63727856e-01 3.32188010e-01 -7.29956746e-01 -6.32586420e-01 2.65852332e-01 -2.90610999e-01 -2.21491367e-01 -2.74129361e-01 9.25236642e-01 -1.70838952e+00 9.77304399e-01 -6.12922668e-01 -3.18718642e-01 -1.74300373e+00 5.57672322e-01 3.32050115e-01 -3.97448093e-01 -5.44428587e-01 5.30009687e-01 -2.90150166e-01 -1.17929867e-02 5.20702451e-02 -9.67976034e-01 1.83169767e-01 -8.02710652e-01 4.80363458e-01 3.41886580e-02 4.56276834e-01 -3.94598007e-01 -5.76799452e-01 3.44215870e-01 6.48330003e-02 -2.74553925e-01 1.60396779e+00 4.23517823e-01 -7.21335888e-01 -1.32664710e-01 8.24875593e-01 5.06740808e-01 -8.16647783e-02 3.03262532e-01 -1.28666699e-01 -4.07308847e-01 -1.18148185e-01 -7.07388401e-01 -5.82873106e-01 6.07222319e-01 2.80312359e-01 5.47026277e-01 1.19084823e+00 -5.86418569e-01 4.74814594e-01 1.05201721e+00 1.09901059e+00 -4.53507721e-01 -9.73290324e-01 6.49299622e-01 3.96349192e-01 -4.90575939e-01 4.81931478e-01 -1.03440869e+00 -3.00889194e-01 1.61680460e+00 1.30020350e-01 -6.68692589e-01 2.39704043e-01 8.71012330e-01 -2.39673018e-01 -1.73408821e-01 -6.97132170e-01 1.78002450e-03 -1.52715474e-01 4.76655871e-01 3.68957371e-01 4.51550493e-03 -9.95206535e-01 5.13768137e-01 -7.88095117e-01 1.93816513e-01 1.29771483e+00 1.53091490e+00 -4.48332220e-01 -1.98398280e+00 -1.08469784e+00 -2.72636950e-01 -5.30928195e-01 -1.09054349e-01 -6.90939486e-01 1.66623080e+00 3.90487075e-01 7.62595594e-01 -4.23874706e-01 -3.57676148e-01 8.37744623e-02 6.32901266e-02 1.01341486e+00 -6.19308352e-01 -1.14823425e+00 -3.14009516e-03 5.07342517e-01 -3.23102564e-01 -1.20014571e-01 -5.91527879e-01 -1.37600434e+00 -9.28876579e-01 -5.51624000e-01 1.00119805e+00 9.73118842e-01 9.19533074e-01 -5.23667455e-01 4.02242839e-01 4.81220543e-01 -5.66381335e-01 -8.01722705e-01 -4.32992280e-01 -7.34052598e-01 -2.52028227e-01 -1.00475609e-01 -3.24390054e-01 -2.85852104e-01 -2.43303508e-01]	[5.783853054046631, 4.727225303649902]
35354b98-b355-4d0d-a9b9-d22d284cb289	learning-trajectory-word-alignments-for-video	2301.01953	null	https://arxiv.org/abs/2301.01953v3	https://arxiv.org/pdf/2301.01953v3.pdf	Learning Trajectory-Word Alignments for Video-Language Tasks	In a video, an object usually appears as the trajectory, i.e., it spans over a few spatial but longer temporal patches, that contains abundant spatiotemporal contexts. However, modern Video-Language BERTs (VDL-BERTs) neglect this trajectory characteristic that they usually follow image-language BERTs (IL-BERTs) to deploy the patch-to-word (P2W) attention that may over-exploit trivial spatial contexts and neglect significant temporal contexts. To amend this, we propose a novel TW-BERT to learn Trajectory-Word alignment by a newly designed trajectory-to-word (T2W) attention for solving video-language tasks. Moreover, previous VDL-BERTs usually uniformly sample a few frames into the model while different trajectories have diverse graininess, i.e., some trajectories span longer frames and some span shorter, and using a few frames will lose certain useful temporal contexts. However, simply sampling more frames will also make pre-training infeasible due to the largely increased training burdens. To alleviate the problem, during the fine-tuning stage, we insert a novel Hierarchical Frame-Selector (HFS) module into the video encoder. HFS gradually selects the suitable frames conditioned on the text context for the later cross-modal encoder to learn better trajectory-word alignments. By the proposed T2W attention and HFS, our TW-BERT achieves SOTA performances on text-to-video retrieval tasks, and comparable performances on video question-answering tasks with some VDL-BERTs trained on much more data. The code will be available in the supplementary material.	['Songfang Huang', 'Fei Huang', 'Yu Zhang', 'Ming Yan', 'Chenliang Li', 'Qinghao Ye', 'Hanwang Zhang', 'Haiyang Xu', 'Zhangzikang Li', 'Xu Yang']	2023-01-05	null	null	null	null	['video-question-answering', 'video-retrieval', 'word-alignment']	['computer-vision', 'computer-vision', 'natural-language-processing']	[-3.65088023e-02 -3.79403144e-01 -4.92568314e-01 -3.40644330e-01 -8.80988121e-01 -4.47454184e-01 4.85051602e-01 -2.31569245e-01 -5.83758295e-01 3.96852851e-01 2.44316012e-01 -3.13553005e-01 -5.65276481e-02 -4.86678153e-01 -8.75747323e-01 -6.72979891e-01 9.13501680e-02 1.51980102e-01 7.25780487e-01 -1.52455196e-01 1.07438579e-01 5.56821525e-02 -1.25419128e+00 4.89284754e-01 6.62173510e-01 9.09611881e-01 8.24656069e-01 4.36660677e-01 -3.58283639e-01 9.04047728e-01 -4.12751675e-01 -3.16747367e-01 1.90405846e-01 -3.81000787e-01 -5.37039280e-01 2.28814911e-02 4.54253167e-01 -5.72765887e-01 -8.13040376e-01 8.54728162e-01 3.59156400e-01 4.30701464e-01 2.57853150e-01 -1.17749691e+00 -6.93695188e-01 4.56924081e-01 -7.21343458e-01 5.14194965e-01 2.88718432e-01 4.53596622e-01 1.12311411e+00 -1.00545263e+00 6.00816965e-01 1.30786622e+00 4.14016783e-01 4.52575415e-01 -8.47005606e-01 -7.57996261e-01 7.43994296e-01 5.98878860e-01 -1.64702499e+00 -3.95644248e-01 6.91713095e-01 -2.90731251e-01 9.38860655e-01 2.76579559e-01 6.90883756e-01 1.17360067e+00 7.62739703e-02 1.13421166e+00 3.11858773e-01 -1.27696674e-02 -1.71888918e-02 -3.00601810e-01 -6.41255826e-02 6.20388746e-01 -1.84555396e-01 -1.54760391e-01 -6.14045799e-01 1.47577807e-01 7.74826169e-01 3.73898208e-01 -3.96050423e-01 -5.67965582e-02 -1.47124183e+00 6.82326794e-01 3.31966460e-01 4.15697575e-01 -1.64986163e-01 3.43417108e-01 6.30909681e-01 2.97563553e-01 3.70742440e-01 1.91548169e-01 -4.00016546e-01 -3.40557158e-01 -1.12843323e+00 3.44373763e-01 1.10555008e-01 1.37386322e+00 9.82653439e-01 -9.17066634e-02 -5.84965527e-01 8.00975502e-01 9.66549590e-02 4.45927322e-01 6.00596726e-01 -8.04798126e-01 7.12997854e-01 3.45636100e-01 8.10444504e-02 -9.84664440e-01 -8.16387087e-02 -1.52947724e-01 -6.78801298e-01 -5.66433668e-01 1.20371088e-01 1.15881748e-01 -9.64931011e-01 1.75941074e+00 2.91474640e-01 5.40909648e-01 -3.87506723e-01 1.21471262e+00 6.81731701e-01 1.17963612e+00 5.32406494e-02 -4.53533918e-01 1.30811751e+00 -1.38053441e+00 -8.39297771e-01 -1.52090982e-01 6.52601242e-01 -6.65756762e-01 1.59667349e+00 1.21220730e-01 -1.08936584e+00 -7.08176792e-01 -7.41036713e-01 -2.63109207e-01 -4.51360196e-02 1.71067089e-01 3.77629966e-01 6.68186545e-02 -7.89672315e-01 3.19043517e-01 -8.03493679e-01 -3.04766446e-01 4.46075946e-01 9.85073671e-02 -1.64248213e-01 -4.13096458e-01 -1.30375445e+00 3.12589943e-01 3.08910191e-01 8.40967074e-02 -9.97285306e-01 -6.37949944e-01 -7.62522936e-01 4.88938987e-02 8.31107259e-01 -5.91682732e-01 1.19583666e+00 -1.17508912e+00 -1.41951931e+00 4.93253231e-01 -6.18066370e-01 -3.00578117e-01 5.51889241e-01 -4.08323139e-01 -4.03927177e-01 5.58350503e-01 1.71035007e-01 8.07356596e-01 1.27447748e+00 -8.64339709e-01 -6.11021459e-01 5.92346527e-02 3.81962985e-01 1.90156102e-01 -4.40601617e-01 1.32124245e-01 -1.16539431e+00 -9.66623008e-01 -8.57117996e-02 -7.81298220e-01 -8.41140281e-03 4.93292809e-02 -1.62293032e-01 -4.11722541e-01 1.02558899e+00 -5.69782138e-01 1.74941552e+00 -2.35921192e+00 3.60718876e-01 -2.16990635e-01 1.02822803e-01 4.97142792e-01 -6.20082140e-01 4.97702569e-01 -6.57942444e-02 1.34698316e-01 -6.79099485e-02 -3.11392516e-01 -2.30162650e-01 4.11485255e-01 -4.60219115e-01 2.65434831e-01 3.25266749e-01 1.11691380e+00 -1.15282762e+00 -6.81545138e-01 1.72979936e-01 1.21131696e-01 -7.22487926e-01 3.56183231e-01 -6.30153358e-01 3.39083880e-01 -7.72877812e-01 4.52324986e-01 5.73015451e-01 -2.56524205e-01 -6.08619824e-02 -4.10808206e-01 -2.94315040e-01 2.04339370e-01 -6.53139055e-01 2.01585579e+00 -4.78586107e-01 6.46924734e-01 -1.85253710e-01 -1.08273840e+00 4.56333339e-01 3.52801383e-01 6.54788435e-01 -7.72899926e-01 -7.58225247e-02 9.72469002e-02 -5.61098717e-02 -1.06524336e+00 4.94744837e-01 1.12811670e-01 1.65221334e-01 2.20285371e-01 2.26633903e-02 2.47226894e-01 2.63909310e-01 3.86205435e-01 1.05179203e+00 3.68724376e-01 9.13837180e-02 -8.45639333e-02 7.56643236e-01 -1.20396890e-01 8.99263918e-01 6.03436708e-01 -3.63453776e-01 8.30520093e-01 5.30545533e-01 -4.92290497e-01 -1.09804833e+00 -8.31336975e-01 7.44941235e-02 1.20061433e+00 5.64258099e-01 -8.15539837e-01 -5.85212588e-01 -7.16761529e-01 -2.37699091e-01 5.58355808e-01 -3.86874855e-01 -2.90714592e-01 -8.46596479e-01 -2.66496420e-01 3.66078734e-01 4.62196410e-01 4.05235976e-01 -1.04649627e+00 -4.62168425e-01 3.51236165e-01 -3.94999534e-01 -1.23398781e+00 -1.26654375e+00 -4.19243425e-01 -6.13379359e-01 -8.73403251e-01 -1.00790548e+00 -7.59551287e-01 5.70822656e-01 8.18587720e-01 9.30359542e-01 3.11953485e-01 2.70619951e-02 2.04297528e-01 -9.34237778e-01 7.00920150e-02 2.51103789e-01 2.49989256e-01 -1.85718581e-01 9.79190022e-02 5.04775047e-01 -4.10696924e-01 -7.70022392e-01 6.03864372e-01 -1.17061257e+00 3.91773403e-01 2.97090799e-01 1.14446855e+00 6.26831532e-01 -2.37467270e-02 5.12548029e-01 -5.05456865e-01 2.16599807e-01 -6.03207588e-01 -4.25349951e-01 3.45573515e-01 -7.82166868e-02 -2.04426795e-01 8.40678334e-01 -8.94252419e-01 -7.61091650e-01 -3.03003937e-01 -1.81534484e-01 -1.13125014e+00 1.62798464e-01 5.81955194e-01 -3.36914659e-01 2.87959397e-01 1.02347642e-01 4.28929120e-01 -3.03735405e-01 -2.93321013e-01 1.65945396e-01 5.36258459e-01 2.47586995e-01 -6.95564926e-01 8.02231193e-01 3.03886354e-01 -5.49970508e-01 -8.43337536e-01 -8.49501252e-01 -6.31926954e-01 -4.42972332e-01 -2.24497303e-01 1.00739610e+00 -1.01957178e+00 -4.30507392e-01 3.14440370e-01 -1.21958196e+00 -5.61313093e-01 -1.49199441e-01 5.15329242e-01 -4.20857817e-01 5.41572034e-01 -5.16845584e-01 -6.67585969e-01 -8.51591304e-02 -1.47159433e+00 1.31476915e+00 1.13652349e-01 7.61640817e-02 -5.76733589e-01 -3.13658327e-01 3.27126145e-01 2.27368981e-01 -4.09882903e-01 8.93969774e-01 -3.16229343e-01 -8.79371524e-01 9.98495370e-02 -3.82901996e-01 1.43318683e-01 8.29675570e-02 4.00887541e-02 -5.47592282e-01 -4.93517011e-01 -3.63255516e-02 -2.09880963e-01 1.05000520e+00 3.64179343e-01 1.77562690e+00 -4.21746403e-01 -2.12368101e-01 7.30809331e-01 1.17208648e+00 4.75012481e-01 6.85667336e-01 2.96721011e-01 9.54551339e-01 4.80786383e-01 9.99593198e-01 2.46151090e-01 4.11750704e-01 9.34014559e-01 2.55660266e-01 9.49009433e-02 -9.01791975e-02 -4.38093871e-01 6.70287251e-01 1.06558454e+00 1.41015137e-02 -4.13488060e-01 -6.28566563e-01 6.56398475e-01 -2.13040304e+00 -1.09280968e+00 1.20058708e-01 2.03647232e+00 7.94543564e-01 2.87162848e-02 2.32336044e-01 -2.28357136e-01 8.07530284e-01 6.16616905e-01 -7.63491094e-01 8.45124945e-02 -1.44432755e-02 -2.39402860e-01 3.79048198e-01 4.52526182e-01 -8.66956115e-01 1.19101572e+00 4.66723013e+00 1.47335291e+00 -1.20965934e+00 2.93614924e-01 4.68065947e-01 -4.98824507e-01 -6.66065097e-01 1.43395767e-01 -6.56259239e-01 8.38083148e-01 7.18737960e-01 -1.16050661e-01 5.51958621e-01 5.60773194e-01 6.33368134e-01 3.34825516e-02 -9.02299583e-01 1.13967228e+00 -1.30981933e-02 -1.32264435e+00 3.03751141e-01 -2.31269181e-01 5.35423636e-01 -1.29415030e-02 -5.30800074e-02 4.57921922e-01 -2.91280806e-01 -7.76106298e-01 9.78888512e-01 4.03467983e-01 1.00398886e+00 -5.47596455e-01 4.53038841e-01 4.24546778e-01 -1.55568171e+00 -1.21734738e-01 -7.12129891e-01 2.86081910e-01 4.87161577e-01 5.59997439e-01 -1.66513532e-01 7.09444463e-01 6.91524327e-01 1.10491252e+00 -2.28540018e-01 9.36353922e-01 1.76967289e-02 6.91084266e-01 -1.79544151e-01 2.05052551e-02 6.27966046e-01 -3.37585509e-01 7.32302368e-01 1.16705000e+00 6.18510544e-01 2.45180309e-01 3.54184955e-01 6.68521583e-01 5.90392426e-02 1.37510732e-01 -3.82167488e-01 -1.45812646e-01 5.11869729e-01 9.59030867e-01 -4.85827714e-01 -4.53546047e-01 -8.00022781e-01 1.03089857e+00 2.14195997e-01 6.68172836e-01 -1.19144869e+00 -2.05231205e-01 5.82340896e-01 4.28136408e-01 4.91926044e-01 -3.21300685e-01 4.15489554e-01 -1.51702619e+00 1.98879331e-01 -9.06870008e-01 3.38554233e-01 -9.90356445e-01 -1.23004770e+00 6.86433315e-01 2.39329040e-01 -1.48939586e+00 1.59548208e-01 -3.07239890e-01 -5.70086181e-01 6.57994509e-01 -1.56368291e+00 -1.18644130e+00 -2.66542792e-01 7.91415513e-01 1.25228310e+00 -2.86326259e-01 1.56685010e-01 6.04959190e-01 -8.91452789e-01 6.82654560e-01 -1.87271331e-02 -5.99770201e-03 9.48828161e-01 -8.31021309e-01 3.28744292e-01 8.75760138e-01 1.99283227e-01 7.24583626e-01 3.88140321e-01 -5.47662199e-01 -1.40192485e+00 -1.36165273e+00 7.81848252e-01 -1.36741586e-02 7.96430647e-01 -3.51485014e-01 -1.15495181e+00 6.88654006e-01 2.26477042e-01 9.27376598e-02 3.47375870e-01 -2.93044895e-01 -4.41331267e-01 -2.62900144e-01 -5.67904294e-01 8.72577548e-01 1.28034091e+00 -7.90364683e-01 -5.90839088e-01 4.44260865e-01 1.23214793e+00 -3.56088132e-01 -3.63419443e-01 3.10452700e-01 4.22597080e-01 -8.83179247e-01 9.58256245e-01 -5.62153518e-01 5.88174760e-01 -5.79932332e-01 -2.37120703e-01 -9.18347597e-01 -3.53382021e-01 -9.43292916e-01 -2.24357769e-01 1.20711994e+00 1.28481621e-02 -2.89385736e-01 5.50882638e-01 2.25575104e-01 -3.29098046e-01 -1.04445255e+00 -1.08203566e+00 -7.84635603e-01 -2.41711596e-03 -5.65741479e-01 6.06946409e-01 7.08407223e-01 -2.78335780e-01 2.66569227e-01 -6.39594555e-01 5.27356602e-02 9.14044157e-02 1.97517633e-01 6.41120374e-01 -5.37464559e-01 -3.47454846e-01 -3.86623025e-01 -1.50915936e-01 -1.91702080e+00 1.77049041e-01 -6.67074561e-01 1.67825624e-01 -1.16716957e+00 2.25453168e-01 -3.71829331e-01 -2.73774207e-01 3.14038336e-01 -5.81066191e-01 8.80021788e-03 3.88156533e-01 5.41500449e-01 -9.25915062e-01 8.56402397e-01 1.58818460e+00 -1.63538516e-01 -4.53755260e-02 -1.96314141e-01 -2.44847924e-01 5.24445415e-01 3.67677957e-01 -3.97407800e-01 -7.61238456e-01 -8.39659631e-01 2.67630935e-01 2.38217697e-01 2.66863674e-01 -7.00050831e-01 1.86021298e-01 -3.97856444e-01 -3.01305950e-02 -8.17658961e-01 2.99046963e-01 -8.25436771e-01 7.59070087e-03 2.47574598e-01 -3.66803408e-01 2.53168911e-01 9.14898422e-03 6.93081856e-01 -4.26410824e-01 -2.68436730e-01 5.03267825e-01 -1.00099482e-01 -9.30316925e-01 8.50937247e-01 -4.04027998e-01 7.00387359e-02 1.10112107e+00 -3.23614627e-01 -1.31036475e-01 -4.94409978e-01 -3.73120666e-01 6.58187985e-01 4.28645343e-01 6.01683438e-01 7.31806397e-01 -1.42077279e+00 -4.34428662e-01 1.66775703e-01 2.21477583e-01 3.57170016e-01 7.02125311e-01 9.86276388e-01 -3.39990556e-01 4.46784467e-01 1.71899453e-01 -6.81570590e-01 -1.06602275e+00 8.09767485e-01 1.48950249e-01 -1.58244863e-01 -9.53992963e-01 9.75024700e-01 8.10839891e-01 7.38910958e-02 1.87261418e-01 -3.63703728e-01 -6.05625771e-02 3.71959396e-02 6.33467972e-01 -3.08086677e-03 -3.51536810e-01 -6.59156501e-01 -2.19996408e-01 6.61485672e-01 -2.62515426e-01 2.50199735e-02 1.01756430e+00 -4.74281251e-01 6.83779120e-02 4.96026754e-01 1.40772212e+00 -5.23646511e-02 -1.66587770e+00 -3.63513976e-01 -1.87473252e-01 -8.67524803e-01 8.21343530e-03 -1.30120769e-01 -1.23705125e+00 9.96391118e-01 5.78599088e-02 -7.72719532e-02 1.25871217e+00 -7.08275512e-02 1.32740593e+00 2.22939327e-01 3.14993709e-01 -9.49155390e-01 3.79215777e-01 6.17201030e-01 8.26584578e-01 -9.62867439e-01 -1.58447757e-01 -3.72399092e-01 -7.17171729e-01 9.82587218e-01 7.85278320e-01 -5.19153774e-02 4.94611591e-01 -5.59555888e-02 -2.16228157e-01 7.95547441e-02 -9.94839191e-01 -1.84951708e-01 2.39155024e-01 2.98532605e-01 2.82968313e-01 -3.50990772e-01 -4.79626805e-01 6.31551862e-01 2.30852127e-01 -3.73059665e-05 2.67049432e-01 5.94140172e-01 -3.22243214e-01 -1.12541473e+00 -4.98088151e-02 3.94630075e-01 -3.17637503e-01 -3.13229531e-01 2.23762423e-01 6.28856003e-01 3.40979218e-01 7.01269686e-01 2.67087340e-01 -4.24045801e-01 2.72282571e-01 -1.09370708e-01 2.41795048e-01 -4.73778844e-01 -4.82964873e-01 5.72901428e-01 -2.10679919e-01 -9.03335750e-01 -5.77285886e-01 -6.39934063e-01 -1.17453277e+00 -3.41431677e-01 -3.74453694e-01 1.75037518e-01 -4.14189287e-02 1.05758274e+00 4.21584934e-01 7.26808667e-01 6.49904132e-01 -8.41422558e-01 -3.19397241e-01 -8.45244050e-01 -4.25683737e-01 4.42089170e-01 4.29630041e-01 -7.04939008e-01 -8.90828148e-02 1.69657230e-01]	[9.99427604675293, 0.6183263063430786]
cdc9810c-9cae-4eee-978f-b0ccb3bc8be2	on-estimating-total-time-to-solve-sat-in	1308.0761	null	http://arxiv.org/abs/1308.0761v1	http://arxiv.org/pdf/1308.0761v1.pdf	On estimating total time to solve SAT in distributed computing environments: Application to the SAT@home project	This paper proposes a method to estimate the total time required to solve SAT in distributed environments via partitioning approach. It is based on the observation that for some simple forms of problem partitioning one can use the Monte Carlo approach to estimate the time required to solve an original problem. The method proposed is based on an algorithm for searching for partitioning with an optimal solving time estimation. We applied this method to estimate the time required to perform logical cryptanalysis of the widely known stream ciphers A5/1 and Bivium. The paper also describes a volunteer computing project SAT@home aimed at solving hard combinatorial problems reduced to SAT. In this project during several months there were solved 10 problems of logical cryptanalysis of the A5/1 cipher thatcould not be solved using known rainbow tables.	['Oleg Zaikin', 'Alexander Semenov']	2013-08-04	null	null	null	null	['cryptanalysis']	['miscellaneous']	[ 4.72180843e-02 2.64905274e-01 5.32433450e-01 -3.21511626e-01 -1.06087959e+00 -9.68839049e-01 1.99357033e-01 5.02121866e-01 -4.20097202e-01 1.10784602e+00 -5.30846059e-01 -7.13025212e-01 -5.06630480e-01 -1.14444363e+00 -5.41379571e-01 -6.89508915e-01 -1.15955524e-01 1.22318268e+00 2.52244294e-01 -2.34869376e-01 3.97657424e-01 3.77868652e-01 -1.16163433e+00 4.25271332e-01 6.03497446e-01 6.59469545e-01 4.21984494e-02 1.12325430e+00 -2.37840861e-01 5.94294846e-01 -1.05436265e+00 -3.61746073e-01 6.53161168e-01 -8.53609681e-01 -1.34619570e+00 -2.87231982e-01 -1.17551304e-01 8.99366736e-02 -1.32046536e-01 1.14457381e+00 2.45135114e-01 -6.23813421e-02 4.11321551e-01 -1.49330795e+00 6.34808481e-01 1.19322002e+00 -5.18069446e-01 4.22134787e-01 5.78450978e-01 -3.71647358e-01 8.35389078e-01 1.57237664e-01 5.47067463e-01 9.25819933e-01 4.98122185e-01 -3.75033505e-02 -1.06452131e+00 -8.40788364e-01 -4.28867131e-01 7.59250402e-01 -1.77178526e+00 3.47861610e-02 4.31748897e-01 3.40569437e-01 1.15402305e+00 7.36489952e-01 7.64609694e-01 -1.13239728e-01 4.62310016e-01 1.00278549e-01 1.73940825e+00 -6.76673532e-01 7.80705512e-01 1.26656920e-01 2.97656029e-01 2.70032585e-01 1.01800251e+00 -1.48564264e-01 -1.28026411e-01 -4.85361278e-01 7.97573403e-02 -4.10646290e-01 9.54714343e-02 7.98083395e-02 -8.71787488e-01 8.73433888e-01 -5.31381145e-02 3.70256841e-01 -1.70079663e-01 3.24517697e-01 4.79783267e-01 5.84610283e-01 -1.01900063e-02 5.09074807e-01 -6.84178174e-01 -3.62365037e-01 -1.19749856e+00 4.49351847e-01 1.55101740e+00 8.42958510e-01 6.91696048e-01 -3.24958205e-01 5.06268322e-01 -3.21120203e-01 3.03758353e-01 5.86470485e-01 -1.57286618e-02 -6.30586565e-01 7.02963054e-01 4.27318603e-01 1.18543513e-01 -5.77667654e-01 -3.37081194e-01 5.64257130e-02 -3.08182746e-01 3.75345170e-01 8.31372380e-01 -4.50623572e-01 -5.23864448e-01 1.01987505e+00 5.92428029e-01 -9.90796611e-02 3.48567635e-01 3.97950351e-01 6.42209768e-01 8.42322469e-01 -5.03327131e-01 -3.58636200e-01 1.33894849e+00 -5.76248825e-01 -7.54194438e-01 4.08336520e-01 9.05292213e-01 -9.38203156e-01 7.05150664e-02 1.17673469e+00 -1.46063161e+00 2.99109938e-03 -1.49428761e+00 3.79223675e-01 -3.93243194e-01 -5.69538295e-01 7.32031167e-01 1.47124112e+00 -1.06626999e+00 5.41677177e-01 -8.37311685e-01 -1.97163850e-01 -2.34336555e-02 9.23700988e-01 -8.40540007e-02 -2.75190294e-01 -7.66911805e-01 8.86308730e-01 5.48604369e-01 8.28870311e-02 -6.70102417e-01 -1.55384004e-01 -4.45338458e-01 1.70788720e-01 4.11602914e-01 -2.48087943e-01 8.68930578e-01 -4.10941392e-01 -1.35747600e+00 5.20916998e-01 -6.34330958e-02 -7.48916924e-01 2.98813999e-01 6.24528348e-01 6.62779137e-02 2.50662386e-01 -2.07056344e-01 -6.69914931e-02 -4.54006940e-02 -8.63348424e-01 -4.15163368e-01 -2.66219407e-01 4.70551789e-01 1.10482283e-01 3.28765154e-01 1.41989619e-01 1.74639463e-01 5.29711172e-02 4.21529293e-01 -8.77187133e-01 -5.67401826e-01 -8.49123716e-01 -2.67686158e-01 -7.07508773e-02 3.15509230e-01 -4.61224109e-01 1.00289798e+00 -1.64882195e+00 2.07070276e-01 9.30187821e-01 1.74047828e-01 -5.64423651e-02 1.53791994e-01 1.12346089e+00 -8.50599855e-02 1.94547668e-01 3.41721661e-02 2.08175361e-01 3.65555644e-01 1.76519841e-01 -5.32286093e-02 8.95799875e-01 -4.84604955e-01 4.56157237e-01 -4.79870439e-01 -8.21747780e-01 -3.04654450e-03 -5.01383655e-02 -6.62736416e-01 -1.47026688e-01 -2.55038321e-01 -2.13542432e-01 -3.62195700e-01 4.22495157e-01 1.28638649e+00 1.91369534e-01 7.72233784e-01 1.99495345e-01 -1.86831594e-01 3.42263520e-01 -1.90268159e+00 1.36248362e+00 -2.04406783e-01 6.60344124e-01 2.02078268e-01 -1.17927384e+00 8.04750264e-01 5.16740918e-01 4.12831396e-01 -2.58948952e-01 6.89945400e-01 3.20161164e-01 2.69396096e-01 -9.41840038e-02 4.11189675e-01 -3.14535648e-01 -2.76613891e-01 8.54444981e-01 -2.59490758e-01 -8.19361389e-01 5.53509891e-01 2.80583680e-01 1.55613291e+00 -1.55860856e-01 3.06443691e-01 -6.58193231e-01 7.32770979e-01 5.25521696e-01 2.91317970e-01 6.96167648e-01 2.97147818e-02 -1.16524100e-02 7.63987958e-01 -5.11753798e-01 -8.48113835e-01 -6.66546822e-01 -1.78270280e-01 1.17189199e-01 3.73817235e-01 -9.05324936e-01 -9.18746054e-01 -3.16611528e-01 -5.79025745e-01 4.21156853e-01 -2.79992461e-01 3.88705611e-01 -4.01951462e-01 -7.08183229e-01 6.13329768e-01 2.82149501e-02 5.90352833e-01 -5.02428412e-01 -7.32893944e-01 1.70184597e-01 -1.18038058e-01 -8.83654773e-01 3.99933875e-01 7.92475462e-01 -9.18565810e-01 -1.44530320e+00 -8.07999521e-02 -6.63382471e-01 8.57243419e-01 2.51972616e-01 8.41832817e-01 3.42768580e-01 -5.11496305e-01 7.23602995e-02 -5.59985995e-01 -8.55976164e-01 -3.37391883e-01 6.08211085e-02 -2.59293735e-01 -7.47829437e-01 5.97453713e-01 -3.80337805e-01 -1.52617827e-01 1.68471172e-01 -8.44905078e-01 -3.81272614e-01 2.37135187e-01 4.98066008e-01 4.23035115e-01 1.18947709e+00 -1.86287433e-01 -9.88961577e-01 4.06711459e-01 -3.44263852e-01 -1.31774592e+00 2.33084008e-01 -1.73915714e-01 1.70167744e-01 8.23209226e-01 -1.34179324e-01 -6.16280615e-01 2.97243118e-01 8.20996687e-02 3.39498520e-01 8.53471532e-02 4.60555047e-01 -3.39084834e-01 -5.99433601e-01 5.10766149e-01 -1.48635460e-02 -4.38283116e-01 6.20250218e-02 -1.37055069e-01 4.86737400e-01 -8.43595564e-02 -6.98859036e-01 1.15709376e+00 3.32607716e-01 7.73382664e-01 -6.30716383e-01 -2.36242741e-01 -3.52639705e-01 -1.00645147e-01 1.24226481e-01 5.25334716e-01 -6.55060112e-01 -1.32742524e+00 6.42786175e-02 -1.11483574e+00 -2.76508093e-01 -2.48945981e-01 4.59454179e-01 -4.78599250e-01 5.90479076e-01 -2.09043905e-01 -1.06116307e+00 -2.09848508e-01 -1.02993906e+00 5.05495429e-01 1.50675967e-01 -3.12788248e-01 -9.19261158e-01 2.83680111e-01 5.15901983e-01 1.76198334e-01 4.18218732e-01 7.17694521e-01 -7.63481081e-01 -8.96610320e-01 -5.75516522e-01 3.66480127e-02 -9.40151438e-02 -3.68828118e-01 -1.01230249e-01 -5.86834490e-01 -4.93234038e-01 3.75958562e-01 -2.43561789e-01 1.65187225e-01 3.84788752e-01 8.12654316e-01 -1.60559520e-01 -2.28163064e-01 4.28929567e-01 2.03216863e+00 3.76349270e-01 1.07517183e+00 5.71717203e-01 -1.08458780e-01 2.38490015e-01 9.25517082e-01 6.73542321e-01 2.67078847e-01 3.49362671e-01 4.14706051e-01 3.11928511e-01 6.10929608e-01 3.51028234e-01 2.43486334e-02 4.92717385e-01 -1.77989274e-01 -4.43446249e-01 -9.60453987e-01 3.30375075e-01 -1.36205673e+00 -9.76868629e-01 -7.80121267e-01 2.21419692e+00 7.21873403e-01 4.21306044e-01 -2.13383194e-02 8.80154967e-01 5.09328306e-01 -4.04747963e-01 3.59378129e-01 -1.11665726e+00 2.36312717e-01 1.14027250e+00 1.00962782e+00 8.28977764e-01 -5.94964683e-01 6.00089729e-01 6.79187965e+00 8.49528491e-01 -5.79258621e-01 1.23070471e-01 3.09137523e-01 -1.29899219e-01 -1.39773786e-01 7.35774219e-01 -7.66135931e-01 1.91780865e-01 1.37447631e+00 -5.81396043e-01 6.62054539e-01 5.07749796e-01 -2.17810005e-01 -9.20165896e-01 -9.61273015e-01 7.16803610e-01 -8.00586641e-02 -1.20054281e+00 -5.21052957e-01 4.11640912e-01 8.84518981e-01 -3.03932220e-01 -4.73961115e-01 1.16608150e-01 3.94869059e-01 -9.87786531e-01 3.89577597e-01 -2.29833759e-02 3.71796250e-01 -1.07836926e+00 1.31989872e+00 3.86850893e-01 -1.01950157e+00 1.18958734e-01 -6.30051017e-01 -6.84045374e-01 -1.86077267e-01 1.79312557e-01 -1.07370186e+00 7.12056935e-01 5.53811252e-01 -4.11668420e-01 -1.61111802e-01 1.57080448e+00 6.08705170e-02 5.55089056e-01 -1.06434548e+00 -5.59950531e-01 3.63831729e-01 -3.10508788e-01 2.26916179e-01 6.64703965e-01 5.16155243e-01 4.10001814e-01 -2.36300036e-01 3.60402882e-01 3.40838253e-01 2.62072563e-01 -3.11444044e-01 7.95914605e-02 5.03649652e-01 1.03522754e+00 -1.44756341e+00 -1.72125816e-01 -6.83526928e-03 4.96534854e-01 -4.15968686e-01 -1.26370490e-01 -1.05007744e+00 -1.09384954e+00 -5.70659935e-02 1.62515804e-01 5.70235014e-01 -4.53026265e-01 -5.44234812e-01 -6.17165267e-01 -1.25577584e-01 -1.04467905e+00 4.75876838e-01 -4.17947084e-01 -3.73364151e-01 7.71907628e-01 5.65523207e-01 -9.19039607e-01 -4.94240820e-02 -5.76581419e-01 -8.69191170e-01 9.39613461e-01 -1.13904345e+00 -6.10741854e-01 -2.77066916e-01 5.09271562e-01 2.17643633e-01 -7.45103508e-02 9.82915878e-01 -1.95030451e-01 -2.87726313e-01 3.94363731e-01 1.94936305e-01 -5.14356375e-01 3.07901114e-01 -1.36234069e+00 -1.89729720e-01 9.28693414e-01 5.94134554e-02 4.57064867e-01 1.25791371e+00 -5.34066617e-01 -1.81773698e+00 -1.95077345e-01 1.18231940e+00 -1.98289052e-01 5.84464073e-01 -2.37395898e-01 -1.31173283e-01 3.55710655e-01 5.47466040e-01 -3.27136159e-01 8.24470460e-01 -1.01712704e-01 9.71835554e-02 -3.62507463e-01 -1.58910453e+00 -1.20756298e-01 1.98996499e-01 -7.24501908e-02 -4.81925577e-01 9.10489619e-01 -3.38209346e-02 -7.22512364e-01 -6.84168160e-01 -3.23360771e-01 6.19377270e-02 -9.95805562e-01 6.04594588e-01 1.63826607e-02 8.19763318e-02 -3.91981602e-01 -1.92475498e-01 -7.06256330e-01 2.11861819e-01 -1.22077870e+00 3.20339084e-01 8.42737198e-01 3.11569512e-01 -8.66042435e-01 1.23235393e+00 5.43787956e-01 3.34839255e-01 -4.48916107e-01 -1.33308566e+00 -6.22003496e-01 1.39916182e-01 -2.98448801e-01 9.30708408e-01 5.96506715e-01 3.94747466e-01 -2.90184226e-02 1.71827033e-01 3.19572985e-01 7.51155615e-01 2.05480024e-01 8.70863318e-01 -1.20672631e+00 -7.70896375e-01 4.60016951e-02 -7.88003027e-01 -2.81535268e-01 -1.69701621e-01 -7.08294630e-01 -3.07956815e-01 -1.49492228e+00 1.13549037e-02 -5.12440860e-01 -4.38710488e-02 2.98907250e-01 5.72752178e-01 4.66546714e-01 1.66929618e-01 -2.61705875e-01 -7.11549699e-01 -4.52851593e-01 8.36555600e-01 2.21925359e-02 9.47265103e-02 2.42233425e-01 -5.46910822e-01 3.37638199e-01 7.84518421e-01 -1.02210546e+00 -4.17520761e-01 -5.35703637e-02 9.30708408e-01 5.40758967e-01 -9.85139534e-02 -1.28388321e+00 4.70757365e-01 -3.43252063e-01 2.53331978e-02 -1.07781136e+00 2.54388213e-01 -1.28121686e+00 6.07283950e-01 1.01402140e+00 1.01647332e-01 3.51585835e-01 4.00009632e-01 3.24625283e-01 -2.34243885e-01 -1.01897132e+00 4.84438986e-01 -3.55349958e-01 -1.41227514e-01 -2.32198119e-01 -7.15001225e-01 -1.09684974e-01 1.68798947e+00 -4.54037130e-01 -2.03567088e-01 -5.61495721e-01 -6.33921802e-01 3.38449389e-01 5.05618811e-01 -8.33478391e-01 3.21903676e-01 -7.19384015e-01 -6.26410604e-01 -9.00959298e-02 -4.91611660e-01 -4.25906032e-02 1.54179856e-01 8.68649244e-01 -1.84069276e+00 6.04383111e-01 -3.11972827e-01 -1.65389135e-01 -1.83835757e+00 5.09852409e-01 2.60031849e-01 -7.45179176e-01 -2.55130321e-01 9.54757392e-01 -5.97646892e-01 9.72633958e-02 7.36317262e-02 -4.67168242e-01 1.80808216e-01 -1.86726272e-01 4.81120020e-01 5.77974379e-01 3.61007154e-01 -3.12967561e-02 -5.90273440e-01 3.64147156e-01 1.58679679e-01 -3.19625169e-01 1.59215665e+00 5.25385514e-02 -7.33407497e-01 -1.28176883e-01 1.00532079e+00 2.37529531e-01 -2.27573633e-01 3.35032970e-01 -1.02878451e-01 -4.11223978e-01 2.16973275e-01 -6.16705298e-01 -6.81915700e-01 5.64612031e-01 3.70852768e-01 6.58716261e-01 1.36452937e+00 -4.07989889e-01 6.42857909e-01 1.00962877e+00 9.81805861e-01 -1.10485971e+00 -6.47549748e-01 4.20825899e-01 1.95763215e-01 -7.32506037e-01 7.76191831e-01 -6.48202419e-01 -1.91766903e-01 1.56226170e+00 8.34327564e-02 -4.77571756e-01 5.87064922e-01 9.29517031e-01 -2.88638026e-01 -2.18189389e-01 -7.73613274e-01 3.35717164e-02 -4.75743920e-01 3.54461730e-01 2.66179796e-02 1.56592280e-01 -9.37666595e-01 5.15570462e-01 -6.45079195e-01 -2.22514085e-02 1.16909635e+00 1.33551610e+00 -2.21575409e-01 -1.71431196e+00 -1.00741053e+00 1.77655399e-01 -6.37584031e-01 -7.94243813e-02 -4.52955931e-01 9.77233112e-01 2.05143541e-01 1.20656598e+00 -2.29011640e-01 -2.27918789e-01 -2.66924918e-01 -1.47335559e-01 1.07759118e+00 -4.65442836e-01 -9.40653980e-01 -7.20621571e-02 3.81760627e-01 -1.67815015e-01 -2.81835467e-01 -6.53612614e-01 -1.27631593e+00 -7.20432401e-01 -7.66654134e-01 1.06183481e+00 1.06062925e+00 9.11110044e-01 -5.68167686e-01 2.82275438e-01 6.45613790e-01 -5.48313677e-01 -6.31902039e-01 -5.37066519e-01 -9.75869715e-01 -4.04322982e-01 -4.40038204e-01 -4.62397113e-02 -5.11865973e-01 -3.88006151e-01]	[5.751817226409912, 4.699889183044434]
a6fba741-f7c2-4ce5-9261-bb5f4ea86b32	an-analytical-framework-for-downlink-leo	2212.03549	null	https://arxiv.org/abs/2212.03549v2	https://arxiv.org/pdf/2212.03549v2.pdf	Modeling and Analysis of LEO Satellite Networks Leveraging Cox Point Processes	This work develops an analytical framework for downlink low earth orbit (LEO) satellite communications, leveraging tools from stochastic geometry. We propose a tractable approach to the analysis of such satellite communication systems accounting for the fact that satellites are located on circular orbits. We accurately characterize this geometric property of such LEO satellite constellations by developing a Cox point process model that jointly produces orbits and satellites on these orbits. Our work differs from existing modeling studies that have assumed satellites' locations as completely random binomial point processes. For this newly developed Cox model, we derive the outage probability of the proposed network and the distribution of the signal-to-interference-plus-noise ratio (SINR) of an arbitrarily located user in the network. By presenting various key network performance metrics as functions of key network parameters, this work allows one to assess the statistical properties of downlink LEO satellite communications and thus can be used as a system-level design tool.	['François Baccelli', 'Chang-Sik Choi']	2022-12-07	null	null	null	null	['point-processes']	['methodology']	[-3.47286522e-01 1.92845181e-01 -2.22066134e-01 1.92903593e-01 -9.86719057e-02 -7.73724616e-01 5.61754823e-01 -3.16187888e-01 2.70994250e-02 1.04518151e+00 8.98026377e-02 -8.50183010e-01 -5.52321076e-01 -7.80899048e-01 -3.70433003e-01 -1.12826967e+00 -1.10473025e+00 6.78574324e-01 -1.42321229e-01 -2.95312643e-01 -2.78352231e-01 8.53083551e-01 -1.03232396e+00 -9.28622961e-01 6.50304615e-01 1.03810024e+00 -9.65172648e-02 7.36893177e-01 5.53547144e-01 3.43132943e-01 -6.84095442e-01 -2.06695516e-02 1.87025145e-01 -3.85232091e-01 -4.18782204e-01 1.61515042e-01 -7.61679411e-01 -1.06000036e-01 -7.04052627e-01 8.23419571e-01 5.48835039e-01 -6.38950989e-02 7.99615860e-01 -1.43735301e+00 2.07964152e-01 3.44637632e-01 -2.53293991e-01 6.45578146e-01 -5.29250689e-02 -1.72854553e-03 8.66764843e-01 -1.99062943e-01 7.08065629e-02 1.10329390e+00 1.01691198e+00 -2.96401739e-01 -1.05629742e+00 -3.50449741e-01 -5.74807882e-01 -3.29150945e-01 -1.75110185e+00 -3.38243425e-01 -1.38550594e-01 -3.81735981e-01 5.68553567e-01 6.12293065e-01 8.58190536e-01 2.92440176e-01 6.00585878e-01 3.44090819e-01 1.01946461e+00 -3.70294571e-01 2.53057480e-01 -2.54926711e-01 -7.21950009e-02 9.50361416e-02 6.36777699e-01 2.44094133e-01 -9.58973467e-02 -4.44456220e-01 7.52480984e-01 -6.70547009e-01 -4.66332793e-01 -1.10320181e-01 -9.46512222e-01 4.89294380e-01 2.66653091e-01 4.69923586e-01 -7.79532015e-01 4.84645814e-01 -1.73709810e-01 4.08675075e-01 4.25969005e-01 -8.04824233e-02 1.02856591e-01 -6.29419163e-02 -1.18308091e+00 4.53068949e-02 1.27428985e+00 1.26425350e+00 1.39348611e-01 2.69170791e-01 -2.98838884e-01 -3.05503849e-02 8.68676484e-01 8.20301592e-01 -2.50355482e-01 -5.97041667e-01 2.74791181e-01 -4.44368035e-01 4.98640865e-01 -6.12338781e-01 -6.02473497e-01 -1.51875639e+00 -6.96633756e-01 -2.48562634e-01 6.28019869e-02 -7.40155220e-01 -3.95676643e-01 1.21308017e+00 9.00773853e-02 -1.27185822e-01 4.38209176e-01 5.72749257e-01 2.59580323e-03 8.23673129e-01 -1.68624699e-01 -5.41062713e-01 1.25113833e+00 -4.54525858e-01 -6.09030604e-01 2.36086711e-01 8.52168798e-01 -4.45239216e-01 -1.01721529e-02 1.83783427e-01 -1.07991350e+00 5.26774824e-01 -9.90047336e-01 6.28398895e-01 5.17839372e-01 1.18214324e-01 4.36865121e-01 1.24350059e+00 -1.62631309e+00 3.98992777e-01 -8.39531541e-01 -6.84411764e-01 2.52703696e-01 4.37441587e-01 4.38628674e-01 9.56905708e-02 -1.17848790e+00 6.60319090e-01 3.64933759e-01 2.08934024e-01 -1.12646759e+00 -1.72528163e-01 -2.15490863e-01 5.80385566e-01 1.72904477e-01 -1.12648892e+00 1.46877086e+00 -4.84973609e-01 -1.06851256e+00 2.21135691e-01 -9.41195562e-02 -8.85536790e-01 1.87344939e-01 5.60215592e-01 -4.55831856e-01 2.92947710e-01 5.54260314e-02 -3.83092016e-01 1.98283702e-01 -1.25041556e+00 -6.34256184e-01 -2.90450424e-01 8.15062746e-02 4.19525951e-01 -6.39206544e-02 3.37073952e-01 -3.30626428e-01 -5.86912930e-01 3.87062043e-01 -1.33690929e+00 -3.54319602e-01 -2.02553898e-01 -4.97042924e-01 6.44169077e-02 4.55394655e-01 -5.17753661e-01 1.32635105e+00 -2.06541967e+00 -6.68923780e-02 6.18031323e-01 -2.12765500e-01 -2.04483256e-01 4.44298863e-01 1.02649355e+00 -4.36271122e-03 3.07515681e-01 6.34109154e-02 -5.30420065e-01 3.68589768e-03 3.21533054e-01 -1.42665431e-01 9.89423692e-01 -5.96911609e-01 3.75108451e-01 -9.83307779e-01 -1.20411813e-02 -3.65847021e-01 2.27153391e-01 -1.79416314e-01 -3.99854541e-01 -1.20991029e-01 6.53287709e-01 -6.05147719e-01 7.31321216e-01 6.00157201e-01 -3.28550041e-01 3.14782321e-01 2.78663397e-01 -4.93182659e-01 2.12203488e-01 -6.09649003e-01 7.26482689e-01 -5.54820001e-01 6.30801618e-01 3.37538421e-01 -4.67387140e-01 5.28467119e-01 7.08393216e-01 3.60870928e-01 3.00443242e-03 5.57545006e-01 3.38503063e-01 4.06854004e-02 -1.00792237e-01 5.94810069e-01 -7.92102814e-01 -1.29569760e-02 5.04009664e-01 6.49747699e-02 -4.87461872e-02 3.17896977e-02 2.93084681e-01 7.51092494e-01 -4.81927484e-01 3.80426943e-01 -8.73047292e-01 6.83056787e-02 -1.03966475e-01 1.83480486e-01 6.67898357e-01 2.33767346e-01 -4.44729552e-02 9.23990428e-01 3.52798462e-01 -1.10689032e+00 -7.79321134e-01 -5.47683775e-01 2.27204431e-02 1.78972751e-01 -1.69838965e-01 -4.73407328e-01 2.87154436e-01 -2.71015503e-02 1.09319019e+00 -3.63380194e-01 3.32455367e-01 3.61652792e-01 -1.41693735e+00 8.58584821e-01 -1.11692622e-01 3.73156220e-01 1.96819320e-01 3.05859149e-01 8.35703239e-02 -2.21094787e-01 -1.02696013e+00 -5.32982545e-03 4.43870611e-02 -1.12721837e+00 -6.94994450e-01 -7.78068185e-01 -1.81737885e-01 5.41667342e-01 5.22896349e-01 7.73286462e-01 1.04387552e-01 5.09203315e-01 6.81453466e-01 -2.09667280e-01 -1.88226104e-01 -4.75541174e-01 -5.07885888e-02 5.33786774e-01 2.54864335e-01 -4.03470695e-01 -6.57997668e-01 -5.55994868e-01 7.11207986e-01 -7.98457205e-01 -5.81321239e-01 2.08354488e-01 2.09959060e-01 2.97408968e-01 8.19692016e-01 1.90045789e-01 1.79722771e-01 4.31955636e-01 -1.43062949e+00 -8.70116532e-01 1.32153422e-01 -3.41126174e-01 -1.31130874e-01 1.37344792e-01 5.06454408e-01 -6.78481400e-01 -4.69922245e-01 1.08530685e-01 1.47641987e-01 5.03770709e-01 1.07706714e+00 -1.79073825e-01 -6.09320641e-01 2.40355834e-01 2.78643817e-01 1.79009095e-01 -2.60215789e-01 -1.51352689e-01 9.76649821e-01 1.49988040e-01 -4.77336794e-01 1.03411126e+00 1.06346512e+00 7.28438914e-01 -1.45627010e+00 -4.47843105e-01 -5.27258873e-01 -1.80584580e-01 -5.21292150e-01 2.58478612e-01 -1.49110484e+00 -8.71711254e-01 4.53746438e-01 -8.57446015e-01 -1.52205318e-01 8.98752064e-02 8.93527448e-01 -6.40736759e-01 4.42336738e-01 -2.81598181e-01 -1.53589785e+00 1.48220539e-01 -7.08752990e-01 8.14889073e-01 9.51687396e-02 1.31463349e-01 -1.47846508e+00 -1.52138025e-01 -1.45008519e-01 5.11349857e-01 3.32595676e-01 3.80581766e-01 -2.12730035e-01 -1.17652345e+00 -3.48185748e-01 -8.41917470e-02 1.38844788e-01 -3.22719187e-01 -1.51484981e-01 -4.76872236e-01 -7.30924785e-01 4.69679475e-01 4.97261345e-01 1.90018371e-01 7.08888590e-01 5.02019227e-01 -4.73790348e-01 -8.12984228e-01 8.02448153e-01 1.61209011e+00 -8.48706067e-02 9.61215436e-01 6.57269120e-01 -1.06932454e-01 3.74684423e-01 5.70669234e-01 8.54734600e-01 4.10831302e-01 7.19236553e-01 7.01707304e-01 3.78827512e-01 4.48490322e-01 1.19093955e-01 3.60598028e-01 3.96331698e-01 -4.08899039e-01 -1.12088406e+00 -1.02265155e+00 7.94234931e-01 -1.26272726e+00 -8.59847844e-01 -7.54437447e-01 2.33562589e+00 1.63178578e-01 -6.26551285e-02 2.56285697e-01 -8.39363262e-02 7.33903706e-01 -3.53935026e-02 3.29948187e-01 2.43087843e-01 -5.00484705e-01 -4.91511941e-01 1.67763305e+00 5.05230963e-01 -7.54656136e-01 1.08649328e-01 6.80159664e+00 9.07963574e-01 -7.55208254e-01 2.46120244e-01 2.79028714e-01 -1.92774191e-01 -5.40604711e-01 5.13944268e-01 -1.04429448e+00 6.88878000e-01 1.42373073e+00 -7.39054441e-01 7.86539391e-02 3.43649954e-01 1.03019381e+00 -4.69790548e-01 -3.14749807e-01 4.58025843e-01 -1.46987930e-01 -9.84925270e-01 -3.31307240e-02 8.83482575e-01 6.82892084e-01 4.15559322e-01 1.06777720e-01 -2.80760139e-01 1.63951546e-01 -8.25241148e-01 7.81345725e-01 9.53584373e-01 6.14741266e-01 -7.11734533e-01 6.81400657e-01 3.48960847e-01 -7.18083203e-01 -3.50836784e-01 2.36471370e-01 -2.03822255e-01 1.08704913e+00 9.37607884e-01 -8.05827737e-01 1.24077916e+00 5.35261035e-01 2.54897714e-01 1.65704712e-01 2.01543999e+00 -1.30243391e-01 1.02775455e+00 -1.07573247e+00 -1.94045193e-02 6.98409319e-01 -2.36051381e-01 1.42357290e+00 4.68149036e-01 1.26399922e+00 1.62439764e-01 -3.42310220e-01 2.07259506e-01 1.10956281e-01 -1.38890713e-01 -4.38734263e-01 -7.42723867e-02 6.25348806e-01 8.26308131e-01 -8.37148726e-01 -4.82700169e-02 -2.24532753e-01 4.14353609e-01 -9.53462005e-01 5.20625234e-01 -5.92106402e-01 -1.71025276e-01 6.62006974e-01 4.81938899e-01 5.03498733e-01 -9.06945229e-01 -3.95244688e-01 -8.26121271e-01 -1.63944736e-01 -3.28446895e-01 -9.74114761e-02 -7.57685184e-01 -4.88887280e-01 4.44938451e-01 2.61899620e-01 -1.69242787e+00 -7.67228231e-02 -1.44986898e-01 -5.70880771e-01 1.21611750e+00 -1.50592339e+00 -9.48132336e-01 1.16468199e-01 -7.41331726e-02 -5.38773060e-01 -2.70807266e-01 2.11931273e-01 4.09557492e-01 -3.33219230e-01 4.03259359e-02 1.34521532e+00 -4.73209649e-01 -2.67265350e-01 -7.95374811e-01 4.53964800e-01 8.76114309e-01 -4.32917714e-01 6.45533681e-01 1.26394641e+00 -7.96354055e-01 -1.06805396e+00 -8.89130712e-01 9.74505246e-01 -1.77809685e-01 1.01015532e+00 -3.78724863e-03 -1.28388017e-01 8.57431412e-01 1.35871202e-01 -2.93724030e-01 8.73489439e-01 1.11720368e-01 8.43365252e-01 3.80048007e-01 -9.94090617e-01 5.82041264e-01 6.56153798e-01 -6.80018514e-02 1.10882204e-02 7.57885516e-01 2.23614335e-01 -3.10162872e-01 -1.05683613e+00 1.50714263e-01 2.15839490e-01 -6.93519294e-01 6.58129692e-01 -1.72884777e-01 -4.69659567e-01 -7.08782732e-01 -1.02492899e-01 -1.39377832e+00 -1.44483238e-01 -1.31273091e+00 1.51864499e-01 9.95769143e-01 4.21733171e-01 -8.08470428e-01 4.86362070e-01 1.15347184e-01 -2.88396508e-01 -3.94760668e-01 -1.45548332e+00 -1.56437361e+00 8.48715939e-03 -1.78552821e-01 5.30316949e-01 5.10979176e-01 -2.56231397e-01 4.37620953e-02 -3.86509031e-01 1.10859156e+00 9.54935193e-01 -3.81015688e-01 4.67116833e-01 -1.33284104e+00 -2.81354278e-01 1.30707696e-02 -5.98293185e-01 -1.43559241e+00 -2.82974064e-01 -7.85846293e-01 -1.54498547e-01 -1.42735410e+00 -3.20454180e-01 -9.25257146e-01 4.25711393e-01 -3.09624523e-01 7.18300819e-01 2.74190366e-01 -4.30242047e-02 7.09122300e-01 -2.17174098e-01 1.00443804e+00 8.32472861e-01 5.89893937e-01 1.20894693e-01 8.45042646e-01 -5.75914145e-01 5.45570850e-01 7.74479985e-01 -6.07748330e-01 -1.39371589e-01 -1.87550291e-01 7.61833549e-01 6.10830367e-01 6.84424818e-01 -1.30846870e+00 2.50250190e-01 3.93890589e-01 -3.74136865e-01 -7.63519943e-01 2.56503582e-01 -8.96914065e-01 8.55483592e-01 5.13489842e-01 6.21929407e-01 -4.12327826e-01 1.03078425e-01 1.05805779e+00 -1.73786595e-01 -5.48137665e-01 5.36142170e-01 2.65112072e-01 -4.19220477e-02 2.84635574e-01 -9.66930211e-01 -4.20281321e-01 1.48811996e+00 -1.84646487e-01 -1.22614875e-01 -1.14617753e+00 -1.16634166e+00 8.29669476e-01 6.56975985e-01 -3.77635539e-01 -3.50739092e-01 -1.10611761e+00 -5.48155129e-01 -4.16516840e-01 -2.47022927e-01 -4.51847464e-01 2.61669993e-01 1.48485398e+00 -8.79440069e-01 1.15656126e+00 1.35615557e-01 -6.23233974e-01 -1.08921945e+00 -4.67508994e-02 8.90789986e-01 1.52231306e-01 1.82846095e-02 3.97876918e-01 -2.34222427e-01 6.32499158e-02 -1.78267211e-01 -1.00153752e-01 5.30468673e-02 3.06504279e-01 7.36753568e-02 4.13991541e-01 2.86973864e-02 -9.99094188e-01 -1.01005889e-01 -7.14254612e-03 5.17754018e-01 -4.97379661e-01 1.09044659e+00 -9.88891125e-01 -4.97919828e-01 2.87891865e-01 6.54855549e-01 9.21866894e-02 -9.31987882e-01 -1.64262548e-01 -3.37656699e-02 -3.30799013e-01 4.13782060e-01 -2.23006740e-01 -5.11134803e-01 3.81851941e-01 1.90255851e-01 6.78646684e-01 7.39685297e-01 1.16757721e-01 2.84996629e-01 1.60336792e-01 9.28804994e-01 -6.80023670e-01 -8.16317201e-01 7.98470974e-01 7.56167412e-01 -2.50893384e-01 1.32244021e-01 -7.01183617e-01 -1.36423618e-01 9.20468390e-01 -5.48930526e-01 2.19876587e-01 9.41464841e-01 1.53034180e-01 -5.60310602e-01 -2.76271909e-01 -3.98936808e-01 -7.12482810e-01 -1.25696748e-01 3.53647500e-01 2.92650342e-01 1.25143290e-01 -7.88558722e-01 3.82040620e-01 -4.71125245e-01 8.74733329e-02 1.19232428e+00 8.40340674e-01 -6.05453432e-01 -1.02280188e+00 -8.89931738e-01 4.88184094e-01 -7.09417105e-01 -5.27059436e-01 3.09747666e-01 7.48064935e-01 -6.00724995e-01 7.80246615e-01 1.48199365e-01 8.16595331e-02 1.16830608e-02 -3.38273674e-01 2.65712589e-01 -5.36148906e-01 1.23829477e-01 -8.19271356e-02 7.41274416e-01 2.82628566e-01 -4.52670902e-01 -1.19050622e+00 -8.04886758e-01 -5.25007784e-01 -5.60084760e-01 8.28537285e-01 8.00968945e-01 1.03349042e+00 2.27436781e-01 5.68376899e-01 1.02119184e+00 -7.65941262e-01 -7.33688414e-01 -8.69136393e-01 -1.53947115e+00 -8.47304046e-01 5.35759568e-01 -7.82191694e-01 -8.12934577e-01 -4.97322828e-01]	[6.136445045471191, 1.5562694072723389]
6af0502a-469b-413f-8856-f730e4174e06	optimal-learning-rates-for-kernel-conjugate	null	null	http://papers.nips.cc/paper/4077-optimal-learning-rates-for-kernel-conjugate-gradient-regression	http://papers.nips.cc/paper/4077-optimal-learning-rates-for-kernel-conjugate-gradient-regression.pdf	Optimal learning rates for Kernel Conjugate Gradient regression	We prove rates of convergence in the statistical sense for kernel-based least squares regression using a conjugate gradient algorithm, where regularization against overfitting is obtained by early stopping. This method is directly related to Kernel Partial Least Squares, a regression method that combines supervised dimensionality reduction with least squares projection. The rates depend on two key quantities: first, on the regularity of the target regression function and second, on the effective dimensionality of the data mapped into the kernel space. Lower bounds on attainable rates depending on these two quantities were established in earlier literature, and we obtain upper bounds for the considered method that match these lower bounds (up to a log factor) if the true regression function belongs to the reproducing kernel Hilbert space. If the latter assumption is not fulfilled, we obtain similar convergence rates provided additional unlabeled data are available. The order of the learning rates in these two situations match state-of-the-art results that were recently obtained for the least squares support vector machine and for linear regularization operators.	['Nicole Krämer', 'Gilles Blanchard']	2010-12-01	null	null	null	neurips-2010-12	['supervised-dimensionality-reduction']	['computer-vision']	[ 1.92427143e-01 2.22978741e-01 -2.44219214e-01 -2.17762411e-01 -7.69400537e-01 -3.19028646e-01 4.21238661e-01 2.80643851e-01 -8.58597517e-01 9.03482974e-01 -2.55377203e-01 -1.36172920e-01 -3.25662106e-01 -3.40856105e-01 -7.06519008e-01 -1.01897633e+00 -2.58549359e-02 3.91475230e-01 1.74743086e-01 -2.17321813e-01 3.29327703e-01 7.97772527e-01 -1.35654163e+00 -4.95720506e-01 9.29358840e-01 9.23545778e-01 -3.09512049e-01 6.35645866e-01 1.72436520e-01 5.67558110e-01 1.81439012e-01 3.68727520e-02 3.92366052e-01 -4.21274453e-01 -5.01221895e-01 2.16491416e-01 3.07549417e-01 2.81524748e-01 -2.82653898e-01 1.20270658e+00 1.88012540e-01 1.91449091e-01 1.19888520e+00 -8.68755341e-01 -4.49604481e-01 2.41402984e-01 -5.35628080e-01 7.11420923e-02 8.17002654e-02 -4.20140535e-01 7.00361311e-01 -1.17150354e+00 3.43144923e-01 6.08252168e-01 8.91734183e-01 5.27996242e-01 -1.64872134e+00 -3.27939630e-01 -4.18591917e-01 -4.54366058e-02 -1.37842989e+00 -6.05598032e-01 6.54734671e-01 -8.27107549e-01 5.63407004e-01 2.01502949e-01 1.09818608e-01 5.13386250e-01 -1.48045212e-01 2.26449162e-01 1.20000684e+00 -1.00208604e+00 4.78521228e-01 9.19676065e-01 4.17967081e-01 1.05421805e+00 3.82156372e-01 1.12592719e-01 -1.66422307e-01 -3.44947577e-01 7.71412492e-01 -1.76627159e-01 -4.66758341e-01 -9.68481958e-01 -1.11545897e+00 1.24441862e+00 -3.72867361e-02 6.22857571e-01 -4.03097063e-01 -2.10381210e-01 3.96421313e-01 4.34833169e-01 7.47363269e-01 3.23704362e-01 -3.72597039e-01 2.33168185e-01 -9.02615190e-01 -2.05413282e-01 1.07022548e+00 6.58688366e-01 8.51599753e-01 3.11138537e-02 1.99653924e-01 7.90867448e-01 1.98865354e-01 6.42052770e-01 3.34875643e-01 -6.85243666e-01 4.88706321e-01 3.64714116e-01 3.53003919e-01 -7.92465568e-01 -5.79866886e-01 -1.58364773e-01 -8.60208333e-01 3.80041987e-01 1.03935254e+00 -2.45432943e-01 -3.44217598e-01 1.63650084e+00 3.25890422e-01 2.62886405e-01 2.39347950e-01 9.10381079e-01 2.12445436e-03 3.53907019e-01 -1.28155410e-01 -8.01959991e-01 9.00949419e-01 -6.23076856e-01 -6.51974320e-01 2.90783420e-02 1.09907866e+00 -6.10667586e-01 9.17048633e-01 4.32908744e-01 -7.66032577e-01 -1.35705218e-01 -1.09568942e+00 2.44815648e-01 -5.82314581e-02 6.35045767e-01 4.48720545e-01 8.66610646e-01 -7.97452211e-01 1.12812209e+00 -7.54365981e-01 -2.94883311e-01 -1.93304077e-01 3.51804912e-01 -7.23687828e-01 2.76677221e-01 -8.97997618e-01 8.98189306e-01 1.44084543e-01 3.46364051e-01 -1.28458589e-01 -5.87724090e-01 -8.30760479e-01 -1.78328127e-01 9.52809900e-02 1.10273287e-02 6.51637435e-01 -9.10961390e-01 -1.54458666e+00 1.09816313e+00 1.66085875e-03 -3.44846547e-01 8.26534867e-01 -1.16403237e-01 -2.63550580e-01 1.11802384e-01 -1.07718885e-01 -2.95596689e-01 1.26272237e+00 -6.48065031e-01 -1.84349328e-01 -7.41896868e-01 -4.46101457e-01 -4.19522561e-02 -4.63680834e-01 -1.39177501e-01 1.16024201e-03 -2.91976571e-01 2.04592586e-01 -1.21146715e+00 -4.14949983e-01 2.21280023e-01 -1.18594818e-01 -2.18048573e-01 4.96795207e-01 -6.26882255e-01 1.17742538e+00 -2.41923213e+00 4.13667381e-01 4.22874004e-01 -4.89857495e-02 1.15065537e-01 2.05781206e-01 2.71365911e-01 -6.59890652e-01 -2.96079487e-01 -5.32048404e-01 -2.69701868e-01 5.22852782e-03 -2.30454788e-01 -2.72861093e-01 1.40213823e+00 7.60612413e-02 2.48382792e-01 -4.95190799e-01 -4.13720161e-01 2.03375012e-01 4.06277090e-01 -2.89053023e-01 1.37396142e-01 2.82718569e-01 3.77656013e-01 -4.41070735e-01 -1.44200504e-01 5.62066138e-01 -1.88669160e-01 -4.15043794e-02 -1.84905007e-01 -2.48845622e-01 -3.89162749e-01 -1.57865536e+00 1.39177096e+00 -4.08529192e-01 4.85602021e-01 2.48052672e-01 -1.58859372e+00 1.11274183e+00 4.25918967e-01 6.62710726e-01 -1.22993775e-01 -6.70960844e-02 7.03329921e-01 -5.29595971e-01 -4.70679790e-01 -5.21078845e-03 -5.70315957e-01 1.94769248e-01 2.17586473e-01 -1.06368139e-01 1.76613003e-01 3.03826809e-01 -1.91013321e-01 8.14680636e-01 1.07319452e-01 5.71210384e-01 -7.14012980e-01 1.03762400e+00 -1.93609938e-01 3.07309926e-01 4.41742092e-01 3.35932560e-02 3.06461573e-01 7.91280985e-01 -2.51071572e-01 -1.12781620e+00 -7.96254277e-01 -6.94920182e-01 8.73895288e-01 -2.42483154e-01 1.57202020e-01 -7.33581245e-01 -6.07720494e-01 1.08788513e-01 5.75041592e-01 -8.49067390e-01 -2.26812139e-01 -4.42630619e-01 -7.42119431e-01 3.85457307e-01 2.65388221e-01 4.32983786e-02 -7.29526997e-01 -2.26200402e-01 8.43043327e-02 3.95323783e-01 -1.04981661e+00 -2.97240466e-01 5.25570989e-01 -1.13472044e+00 -1.33138549e+00 -1.02116501e+00 -7.00418413e-01 7.67886341e-01 -3.90718393e-02 5.55267811e-01 -2.53831983e-01 -2.32645079e-01 4.21794862e-01 6.26847818e-02 9.04664248e-02 -5.35524964e-01 -5.08146398e-02 3.72545391e-01 2.46248811e-01 3.33216220e-01 -4.09697533e-01 -2.54818261e-01 4.67907250e-01 -8.03921521e-01 -4.68887985e-01 5.02592385e-01 7.59298086e-01 6.11214459e-01 -1.31907806e-01 3.12617183e-01 -1.13028264e+00 3.44323397e-01 -3.83986831e-01 -1.14255857e+00 2.54593074e-01 -8.54843140e-01 7.19938815e-01 8.81096363e-01 -5.96038520e-01 -7.88396597e-01 6.60529256e-01 3.21424603e-01 -3.61641705e-01 -1.42521620e-01 3.02287012e-01 2.99166590e-01 -5.19685149e-01 1.17006862e+00 2.49773800e-01 -2.27967687e-02 -5.00941336e-01 3.75042886e-01 5.85040510e-01 3.74725580e-01 -4.24120545e-01 8.35989535e-01 5.85076988e-01 5.06114244e-01 -1.15519798e+00 -7.23866343e-01 -7.89346874e-01 -8.91481161e-01 1.15841411e-01 7.45401919e-01 -6.45220339e-01 -7.13853002e-01 2.48530269e-01 -7.95742214e-01 -2.31061280e-01 -5.06535351e-01 9.90618467e-01 -1.18205798e+00 7.24330664e-01 -5.70420444e-01 -1.21899736e+00 -6.76284358e-02 -7.94147849e-01 5.99420309e-01 -2.47541443e-01 1.47428781e-01 -1.30424106e+00 4.43996429e-01 -6.34497181e-02 1.85547903e-01 7.96609372e-02 8.79427254e-01 -6.84242725e-01 2.39455029e-01 -6.19996190e-01 -2.13150710e-01 6.51614904e-01 -7.04738870e-02 -3.49162608e-01 -7.02296317e-01 -3.03079426e-01 4.64272141e-01 -8.28408077e-02 9.08863306e-01 3.59637529e-01 5.50437987e-01 -3.54109883e-01 -2.03255802e-01 5.13912261e-01 1.66496241e+00 -4.21765476e-01 3.84716183e-01 1.29149780e-01 6.12522602e-01 7.66757905e-01 5.44227302e-01 5.00695705e-01 -3.42334092e-01 7.15165555e-01 -6.13368861e-02 -1.52067199e-01 4.53278720e-01 1.86275810e-01 4.17677164e-01 7.99350739e-01 -2.23744139e-01 5.63901126e-01 -6.82833731e-01 3.94453824e-01 -2.04893923e+00 -8.47267747e-01 -6.09449923e-01 2.95365453e+00 9.67730761e-01 1.21853136e-01 2.68784493e-01 3.14290762e-01 8.61325860e-01 -1.80542201e-01 -5.21979153e-01 -3.57709467e-01 -2.91723385e-02 1.44452512e-01 1.05601478e+00 8.77666175e-01 -1.19590962e+00 5.25536716e-01 6.20454359e+00 8.53344858e-01 -1.05791485e+00 5.09021133e-02 2.32995376e-01 1.67773232e-01 9.12709236e-02 1.45520836e-01 -8.04129958e-01 2.30083793e-01 1.02902806e+00 3.84000912e-02 6.12309277e-01 1.03653491e+00 1.16905123e-01 -8.13453570e-02 -1.03624344e+00 9.37621474e-01 -3.61778438e-02 -8.41022015e-01 -7.73994923e-01 2.78811574e-01 6.40245795e-01 -2.16381609e-01 -5.43087348e-02 1.58249468e-01 -3.84656906e-01 -7.31729269e-01 1.71045899e-01 7.75807619e-01 8.20957720e-01 -7.55345464e-01 4.97174531e-01 5.90166092e-01 -9.75377500e-01 2.64984798e-02 -5.21047592e-01 1.49942130e-01 -1.04337350e-01 1.04737544e+00 -4.61766750e-01 3.74457911e-02 2.77227741e-02 5.09730101e-01 -2.66460389e-01 9.35235143e-01 -1.37751997e-01 6.51783168e-01 -7.16312706e-01 2.55441815e-02 1.67319864e-01 -8.55880737e-01 6.22119367e-01 1.29715455e+00 2.17648849e-01 -3.20366658e-02 -1.03755616e-01 6.74905241e-01 3.60974759e-01 6.81901455e-01 -7.07543790e-01 -5.54129742e-02 -9.02769789e-02 1.26917052e+00 -6.01232350e-01 2.74140824e-04 -6.56976521e-01 1.16467977e+00 5.44405580e-01 4.37116176e-01 -6.07533276e-01 -5.79808950e-01 2.15968817e-01 3.00743401e-01 5.00430167e-01 -4.19287801e-01 -1.69318262e-02 -1.22716701e+00 2.90499657e-01 -2.76209861e-01 2.66116321e-01 -8.08242485e-02 -1.16662049e+00 3.24324727e-01 -8.08112621e-02 -1.07898915e+00 -4.02341545e-01 -8.70720088e-01 -1.39453456e-01 9.48639810e-01 -1.29970932e+00 -6.20252132e-01 2.73277551e-01 6.34484828e-01 -7.45836496e-02 -1.75237462e-01 1.12887716e+00 2.06319526e-01 -5.96405506e-01 4.57128525e-01 7.05294967e-01 -8.59438330e-02 7.19341576e-01 -1.42557764e+00 -3.08505565e-01 6.96748376e-01 1.21867612e-01 5.37962496e-01 9.71668720e-01 -4.18572307e-01 -1.28048098e+00 -6.60021842e-01 9.48813617e-01 -1.35371953e-01 9.49313521e-01 -2.00638443e-01 -1.07990503e+00 4.15185571e-01 -4.59196061e-01 4.91278231e-01 7.05571830e-01 2.26101577e-01 -4.39342737e-01 4.15789112e-02 -1.05094290e+00 1.87246814e-01 4.12406892e-01 -7.10399449e-01 -2.24653512e-01 6.66545689e-01 1.14101380e-01 -1.01872329e-02 -1.03797352e+00 2.35828489e-01 3.90181273e-01 -8.20984244e-01 6.80567980e-01 -8.86061251e-01 -2.72460841e-02 -2.75810272e-01 -1.66250452e-01 -9.42481220e-01 -3.41059804e-01 -7.17148960e-01 -2.28330463e-01 7.54373550e-01 5.66235781e-01 -6.33580625e-01 8.66348982e-01 6.86820090e-01 3.69786769e-01 -7.62571096e-01 -1.17062807e+00 -9.87393141e-01 1.36624828e-01 -1.94966987e-01 -4.50926185e-01 1.12993121e+00 5.08446515e-01 4.67250377e-01 -4.92362827e-01 1.37357697e-01 9.20660734e-01 -2.04517588e-01 3.25220495e-01 -1.54105198e+00 -5.98411322e-01 -2.73381799e-01 -6.32953823e-01 -7.23751426e-01 5.74102640e-01 -8.46532285e-01 5.40344045e-03 -8.13807249e-01 -1.11447200e-01 -5.50709069e-01 -3.15396696e-01 1.80220947e-01 -4.60968688e-02 1.88200802e-01 -2.60689586e-01 4.17484045e-01 -3.19568366e-01 3.28128964e-01 6.69087052e-01 3.09287369e-01 -5.06959021e-01 4.54172671e-01 -1.66243553e-01 9.20076311e-01 5.84335148e-01 -4.59236711e-01 -1.27610058e-01 2.89002836e-01 4.60093677e-01 3.28902274e-01 1.61130413e-01 -9.71450508e-01 2.78568536e-01 -1.47198895e-02 1.56836867e-01 4.41697612e-02 2.37071499e-01 -9.36084926e-01 -1.10402390e-01 3.75972778e-01 -6.62617624e-01 -4.94487464e-01 -9.04439911e-02 8.43748808e-01 7.00910240e-02 -8.65267098e-01 1.23442471e+00 2.69879252e-01 -4.53001022e-01 1.02099432e-02 -3.04647446e-01 -1.27067015e-01 1.01453531e+00 -2.17165753e-01 3.08393121e-01 -3.78218055e-01 -1.07873809e+00 -3.80489796e-01 2.81664580e-01 -1.73576877e-01 3.58718634e-01 -1.32411969e+00 -7.85667598e-01 2.73018420e-01 1.57055065e-01 -6.77722454e-01 1.20119015e-02 1.52620041e+00 -4.14417505e-01 5.27309060e-01 2.41962582e-01 -3.78929913e-01 -1.16538453e+00 9.49192822e-01 3.63557726e-01 -1.71824992e-01 -5.52927494e-01 3.56699079e-01 -5.03208674e-02 -2.90457517e-01 2.44995043e-01 -6.61227107e-02 -9.18300375e-02 5.29251108e-03 4.54948962e-01 7.66903639e-01 1.33808821e-01 -6.75920546e-01 -3.48045796e-01 7.78614640e-01 4.71571237e-02 -1.89495757e-01 1.29246747e+00 -5.99230193e-02 -6.11455813e-02 7.88490891e-01 1.67357540e+00 3.04831624e-01 -1.28609347e+00 -5.69478154e-01 1.80645004e-01 -4.30415879e-04 2.29001924e-01 -7.04365894e-02 -7.68277287e-01 6.35515988e-01 8.29116821e-01 3.41332793e-01 7.68448770e-01 -1.20915413e-01 9.70915034e-02 6.93024755e-01 1.07904509e-01 -1.26625896e+00 -3.93763453e-01 2.89562941e-01 7.96979964e-01 -1.23955011e+00 3.57363671e-01 -6.37528002e-01 -4.64878261e-01 1.46084452e+00 -4.95817773e-02 -7.17059255e-01 8.99700344e-01 -6.85841963e-02 -1.34598300e-01 8.06254074e-02 -1.86196789e-01 -1.70595422e-01 6.62277222e-01 3.53240937e-01 6.30850017e-01 -7.48406947e-02 -8.56306136e-01 1.37301221e-01 2.76069731e-01 -7.39183202e-02 2.27102011e-01 4.82258707e-01 -6.36882424e-01 -9.16879117e-01 -3.79962415e-01 1.71599671e-01 -2.81304508e-01 -2.36400738e-02 -1.42837852e-01 7.35419333e-01 -4.91470039e-01 6.24610424e-01 -2.72782236e-01 3.17946404e-01 3.99733990e-01 1.72498137e-01 6.07024550e-01 -3.77625108e-01 -1.46927107e-02 1.26336783e-01 -3.76097634e-02 -4.60749447e-01 -2.81802654e-01 -5.65146625e-01 -1.32109404e+00 -3.60772610e-02 -8.61177266e-01 5.17770529e-01 9.27209020e-01 1.12253308e+00 -1.53317615e-01 -3.02848577e-01 8.60523582e-01 -7.60455847e-01 -1.08800280e+00 -8.21393907e-01 -1.30222917e+00 3.22506815e-01 4.00102973e-01 -4.24738318e-01 -9.16682899e-01 -3.60110849e-02]	[7.622591495513916, 4.153568267822266]
ab7e874f-bd51-4853-a904-d36d2300c233	transsionadd-a-multi-frame-reinforcement	2306.15212	null	https://arxiv.org/abs/2306.15212v1	https://arxiv.org/pdf/2306.15212v1.pdf	TranssionADD: A multi-frame reinforcement based sequence tagging model for audio deepfake detection	Thanks to recent advancements in end-to-end speech modeling technology, it has become increasingly feasible to imitate and clone a user`s voice. This leads to a significant challenge in differentiating between authentic and fabricated audio segments. To address the issue of user voice abuse and misuse, the second Audio Deepfake Detection Challenge (ADD 2023) aims to detect and analyze deepfake speech utterances. Specifically, Track 2, named the Manipulation Region Location (RL), aims to pinpoint the location of manipulated regions in audio, which can be present in both real and generated audio segments. We propose our novel TranssionADD system as a solution to the challenging problem of model robustness and audio segment outliers in the trace competition. Our system provides three unique contributions: 1) we adapt sequence tagging task for audio deepfake detection; 2) we improve model generalization by various data augmentation techniques; 3) we incorporate multi-frame detection (MFD) module to overcome limited representation provided by a single frame and use isolated-frame penalty (IFP) loss to handle outliers in segments. Our best submission achieved 2nd place in Track 2, demonstrating the effectiveness and robustness of our proposed system.	['Fengjie Zhu', 'Benlai Tang', 'Jiangli Hong', 'Quanxiu Wang', 'Caiyan Wan', 'Hui Huang', 'Zhiba Su', 'Jie Liu']	2023-06-27	null	null	null	null	['deepfake-detection', 'face-swapping']	['computer-vision', 'computer-vision']	[ 1.04431927e-01 -1.57075047e-01 9.88768190e-02 -5.67585789e-02 -1.52343953e+00 -6.38144195e-01 2.21963391e-01 1.75785616e-01 -2.18336523e-01 2.79711097e-01 4.00464326e-01 -2.17910066e-01 1.66832343e-01 1.23266332e-01 -8.26166511e-01 -1.68736473e-01 -5.63305337e-03 -2.01285351e-02 3.19565862e-01 7.37601295e-02 9.32427868e-02 4.96331245e-01 -1.59981608e+00 5.45866132e-01 4.83038366e-01 1.15858495e+00 1.38581529e-01 1.09869385e+00 1.95506424e-01 6.72762334e-01 -9.09432709e-01 -7.43574351e-02 1.04268789e-01 -3.46677989e-01 -8.70888650e-01 1.33625239e-01 5.96492529e-01 -7.41520166e-01 -3.84427398e-01 8.92674744e-01 1.12015676e+00 2.04568565e-01 2.82707870e-01 -1.52958858e+00 2.21421607e-02 7.62972772e-01 -3.78529012e-01 5.45222402e-01 6.41640067e-01 4.67502326e-02 9.19988215e-01 -1.18260312e+00 4.49347794e-01 1.26914334e+00 1.01657104e+00 4.45101142e-01 -1.03192782e+00 -7.45010734e-01 -1.17593132e-01 4.08222824e-01 -1.58903027e+00 -1.29489505e+00 7.32226133e-01 -3.84531051e-01 1.14877081e+00 5.55209756e-01 1.77455127e-01 1.31889772e+00 -4.95660275e-01 9.72499490e-01 4.39541280e-01 -6.01205528e-01 1.24148153e-01 -1.31065160e-01 1.18244231e-01 3.13860506e-01 -5.55369318e-01 -8.25148821e-03 -8.22954237e-01 -3.85004580e-01 2.43970037e-01 -4.98097628e-01 -4.65792567e-01 3.08873266e-01 -9.58044827e-01 4.69388515e-01 -3.21250677e-01 2.17309058e-01 -1.96568742e-01 8.63366108e-03 8.18804860e-01 3.15413415e-01 3.22719455e-01 2.48416126e-01 -5.62776625e-01 -8.61749530e-01 -1.23867357e+00 2.76370555e-01 5.30600369e-01 1.12214470e+00 2.51526713e-01 3.69677782e-01 -3.23268265e-01 1.33249831e+00 2.10967451e-01 6.94257617e-02 6.96826994e-01 -9.35194194e-01 6.94696426e-01 -7.18644559e-02 7.76608884e-02 -6.92870736e-01 -2.54212350e-01 -4.24538374e-01 -2.85090566e-01 -2.14247942e-01 1.85716331e-01 -1.42633796e-01 -7.14746296e-01 1.66908860e+00 3.08946848e-01 7.65242577e-01 -2.25084350e-01 6.40825629e-01 7.35259771e-01 5.43322384e-01 -1.26798302e-01 -9.99529511e-02 1.30935717e+00 -8.49366665e-01 -8.17985892e-01 1.33113489e-01 7.98451304e-01 -1.18336725e+00 1.17743242e+00 6.40902877e-01 -1.11418307e+00 -7.10784554e-01 -9.86044466e-01 -7.58736804e-02 -1.13696955e-01 2.81382263e-01 -1.78776905e-01 9.12879050e-01 -8.27535689e-01 4.40717220e-01 -6.14702225e-01 -3.60418893e-02 1.99958652e-01 4.09442961e-01 -4.69763190e-01 3.46365839e-01 -1.13704872e+00 1.67841062e-01 3.22887212e-01 -2.63735875e-02 -1.01441872e+00 -9.76744235e-01 -8.63369107e-01 -4.28347662e-02 5.11842668e-01 -1.19497649e-01 1.66335654e+00 -5.70344269e-01 -1.44105661e+00 6.64944112e-01 -3.89503181e-01 -5.97544193e-01 6.25964046e-01 -6.02105021e-01 -9.57774818e-01 1.37456596e-01 -7.31093064e-02 3.99283260e-01 1.16253567e+00 -1.05171382e+00 -7.52819538e-01 -2.74783410e-02 -4.98321742e-01 -1.63713664e-01 -3.72408688e-01 6.08364940e-01 -4.23510671e-01 -1.12135303e+00 8.17015674e-03 -8.30823958e-01 4.60540533e-01 -4.19980139e-01 -9.00094330e-01 -2.13792384e-01 1.12053990e+00 -1.25359702e+00 1.74663758e+00 -2.59439993e+00 -4.02305514e-01 1.12633213e-01 2.06950456e-02 7.08315253e-01 -1.03395075e-01 3.43636185e-01 -3.13607842e-01 1.97243035e-01 6.41731992e-02 -8.53231072e-01 1.96593508e-01 -6.48349226e-02 -7.99435675e-01 3.24867398e-01 7.05648065e-02 4.41589653e-01 -6.60566926e-01 -3.39064956e-01 1.35223046e-01 2.53000408e-01 -8.65744233e-01 4.02106643e-01 9.71969124e-03 2.86586195e-01 2.05457345e-01 6.76011324e-01 6.72490418e-01 5.88815033e-01 -4.03144628e-01 -1.56106338e-01 -1.98954880e-01 7.93489635e-01 -1.56291890e+00 1.71121156e+00 -3.46382588e-01 6.79542243e-01 3.94693047e-01 -4.17868704e-01 6.99553430e-01 9.13857102e-01 2.63980597e-01 -2.06227526e-01 7.14821815e-02 2.48392254e-01 -6.54110014e-02 -6.78228855e-01 8.11958671e-01 2.14751571e-01 -5.37486970e-02 3.60535294e-01 2.01396585e-01 -5.32932207e-03 -1.51657522e-01 2.02182814e-01 1.36072814e+00 -1.89111114e-01 -6.52030408e-02 2.44572550e-01 5.66804826e-01 -6.93979323e-01 6.04664028e-01 7.09856749e-01 -5.29154599e-01 1.04005718e+00 3.39480013e-01 2.27935344e-01 -9.36737835e-01 -1.09091222e+00 1.80542350e-01 1.39174438e+00 -5.81426859e-01 -7.76904285e-01 -9.19001281e-01 -6.63158000e-01 -1.01382591e-01 6.61546946e-01 -1.61700025e-01 -1.75523624e-01 -7.69073725e-01 -3.84118147e-02 1.53615558e+00 5.40023625e-01 1.65553272e-01 -8.95982087e-01 9.54391807e-02 2.88621992e-01 -6.33673847e-01 -1.42174613e+00 -1.05075288e+00 9.28709656e-02 -3.40341479e-01 -6.82134211e-01 -4.56166983e-01 -7.33203053e-01 -1.86256226e-02 2.06554011e-01 6.15723550e-01 -8.61606225e-02 -3.74984622e-01 4.32355225e-01 -5.30073106e-01 -2.93697238e-01 -8.24514508e-01 1.29021227e-01 4.53328371e-01 1.58886760e-01 1.49336666e-01 -5.55800200e-01 -2.90703893e-01 4.94177878e-01 -7.12060928e-01 -5.39218068e-01 6.05723495e-03 6.13045871e-01 4.25139517e-01 6.46308511e-02 9.97142613e-01 -4.31439042e-01 8.83247554e-01 -3.83059144e-01 -1.92779735e-01 -3.19469646e-02 -1.63180172e-01 -3.20610195e-01 5.14023602e-01 -5.62393904e-01 -7.04510331e-01 -1.34445308e-02 -9.30512190e-01 -6.54473662e-01 -4.93863493e-01 1.27372906e-01 -6.10626101e-01 2.38751307e-01 4.44773942e-01 5.80347180e-02 -1.98773459e-01 -1.01087379e+00 2.38579839e-01 1.41589892e+00 1.06241167e+00 -5.09765565e-01 4.51134533e-01 -6.07175454e-02 -5.99536836e-01 -1.22011137e+00 -3.68313938e-01 -8.36533785e-01 -4.76735920e-01 -1.07349575e-01 3.30400467e-01 -1.11842620e+00 -7.95107245e-01 6.44356489e-01 -1.25596964e+00 -8.91843904e-03 -1.72086149e-01 4.87566024e-01 -5.54079533e-01 9.34520304e-01 -7.21717477e-01 -1.07336760e+00 -3.31373096e-01 -1.21743238e+00 1.14932060e+00 -3.96276355e-01 -7.46652842e-01 -2.68826991e-01 -8.80744457e-02 5.22912085e-01 1.69636801e-01 -1.94391310e-01 7.63322771e-01 -1.00858903e+00 -1.53421879e-01 -3.51444095e-01 2.11039364e-01 6.99521720e-01 6.51018694e-02 8.97029936e-02 -1.48191583e+00 -3.48977864e-01 -4.89937812e-02 -1.96568847e-01 5.64279914e-01 1.78275183e-01 1.19409585e+00 -2.40092069e-01 -1.36266779e-02 3.75651896e-01 4.89890695e-01 2.80990750e-01 6.76536202e-01 8.78206640e-03 6.06482923e-01 3.83852005e-01 6.23240411e-01 6.80912197e-01 -2.24454738e-02 1.12074804e+00 2.67190367e-01 1.74967095e-01 -5.51337540e-01 -5.75206220e-01 6.71152532e-01 1.20881069e+00 4.66188997e-01 -4.62509662e-01 -7.16302454e-01 8.02452385e-01 -1.58667254e+00 -1.02176571e+00 -1.82652682e-01 2.27418661e+00 7.42231011e-01 2.54025370e-01 6.82753146e-01 9.06259179e-01 1.04736423e+00 -1.16117202e-01 -2.30695695e-01 -4.30117607e-01 7.46883675e-02 2.44261533e-01 1.16319276e-01 5.40502310e-01 -1.18432403e+00 8.17766070e-01 6.04832983e+00 1.23138654e+00 -1.03099406e+00 3.73130649e-01 3.18850279e-01 -2.91261017e-01 5.77442758e-02 -4.34440821e-01 -9.30129588e-01 7.79416919e-01 1.31251252e+00 2.59653687e-01 1.80213854e-01 7.10866094e-01 6.18787944e-01 2.62025595e-01 -1.20854855e+00 1.21754515e+00 2.66774911e-02 -9.51934576e-01 -1.25288844e-01 -1.19236186e-02 1.67364985e-01 -9.43339095e-02 2.11335033e-01 4.19184208e-01 -2.44723111e-01 -8.33412528e-01 1.11426079e+00 1.27791017e-01 8.59837592e-01 -1.06569707e+00 4.48053122e-01 3.53844553e-01 -1.31087363e+00 -8.42746720e-02 1.20648034e-02 1.92442775e-01 2.62277186e-01 3.11775416e-01 -1.57287955e+00 3.81254971e-01 6.37084007e-01 2.49156624e-01 -2.36131147e-01 1.35550129e+00 -3.24112363e-02 1.10754240e+00 -4.33053792e-01 5.23523390e-01 -2.36697689e-01 6.56307220e-01 9.40304577e-01 1.50794697e+00 5.77738881e-01 -4.66667920e-01 7.84202665e-02 4.40031946e-01 -2.97816813e-01 1.19016416e-01 -1.88919187e-01 -9.90378186e-02 1.03877819e+00 7.14409649e-01 -2.93119192e-01 8.97395834e-02 -1.57384098e-01 1.12203765e+00 7.08452538e-02 1.86936930e-01 -8.80696416e-01 -6.20932698e-01 9.84600663e-01 2.15752125e-01 4.44046050e-01 -1.24902420e-01 4.58277464e-02 -8.10603023e-01 2.27858007e-01 -1.05835235e+00 3.78671557e-01 -7.34402657e-01 -9.21281755e-01 6.01863623e-01 -4.05062705e-01 -1.37713993e+00 -5.78011453e-01 -2.04080388e-01 -4.04612064e-01 7.65427947e-01 -1.04741096e+00 -9.89215851e-01 1.09342285e-01 6.60422683e-01 7.63935506e-01 -1.96601391e-01 8.88761759e-01 1.04632652e+00 -6.52543068e-01 1.39815664e+00 -6.63359929e-03 3.44894826e-01 1.07583404e+00 -9.74274039e-01 7.64921308e-01 1.14146757e+00 3.38360369e-01 4.77503479e-01 7.31836617e-01 -5.62189341e-01 -1.02388465e+00 -1.18101537e+00 9.58207905e-01 -4.28992093e-01 6.25261724e-01 -7.43351400e-01 -9.52540636e-01 7.65246689e-01 -2.12407440e-01 -9.41977184e-03 7.94021130e-01 -1.00184223e-02 -3.76724064e-01 1.08470999e-01 -1.01790893e+00 3.84080857e-01 8.68358612e-01 -1.01372552e+00 -5.30126512e-01 -1.57920718e-01 1.01234877e+00 -4.76586699e-01 -6.61683738e-01 2.94319570e-01 3.24315161e-01 -8.16923201e-01 9.81981337e-01 -5.83615184e-01 -8.69275779e-02 -4.43705052e-01 -2.75506079e-01 -1.13439381e+00 1.14653215e-01 -1.42273045e+00 -4.24173623e-01 2.09025931e+00 3.19045275e-01 -1.42244801e-01 5.64355731e-01 3.09732378e-01 -7.16411233e-01 -3.13186020e-01 -1.31921196e+00 -9.88148928e-01 -3.97155881e-01 -1.13685489e+00 6.67360783e-01 8.11310291e-01 6.73435256e-02 -2.88834833e-02 -8.48901927e-01 3.98235023e-01 7.29060248e-02 -5.92535853e-01 9.13679838e-01 -9.70131159e-01 -4.81043160e-01 -1.15579642e-01 -5.60837984e-01 -1.36306524e+00 6.96544126e-02 -6.16225600e-01 3.55249375e-01 -6.49900913e-01 -3.46904635e-01 -2.50435710e-01 -2.63690710e-01 3.16856265e-01 4.91165332e-02 2.84030318e-01 3.25700253e-01 8.73963684e-02 -6.38081849e-01 3.94401699e-01 3.57695311e-01 1.28083810e-01 -3.89286578e-01 4.50896382e-01 -3.39769095e-01 7.46012807e-01 5.43428481e-01 -6.51374340e-01 -1.35658890e-01 -2.54450947e-01 -1.01802386e-01 1.45741999e-01 3.41975540e-01 -1.20558286e+00 1.91183507e-01 4.88577068e-01 -8.88269469e-02 -9.19073462e-01 6.11631691e-01 -7.41798878e-01 -3.87640856e-02 -3.62255331e-03 -4.40721124e-01 2.01911759e-03 4.59080189e-01 4.60558683e-01 -3.06810737e-01 -4.52292413e-01 6.50165677e-01 3.64216924e-01 -5.87146163e-01 2.73460075e-02 -7.77441263e-01 1.41473070e-01 6.64830327e-01 -1.17350012e-01 -2.58428872e-01 -5.16363561e-01 -9.69347775e-01 -2.04226106e-01 1.71672076e-01 6.94203019e-01 5.00095546e-01 -1.21518230e+00 -5.75196326e-01 5.16679287e-01 1.65157750e-01 -3.30437273e-01 4.27068740e-01 7.74182320e-01 -2.59521753e-01 9.88483205e-02 3.16276401e-01 -4.91243124e-01 -1.85937905e+00 2.84924388e-01 1.98871702e-01 2.42784277e-01 -5.02406597e-01 1.20280075e+00 -3.36455405e-01 -2.24437788e-01 9.11422133e-01 -4.37287509e-01 2.14174569e-01 1.07209489e-01 8.23774159e-01 8.95788133e-01 6.44314766e-01 -9.50803459e-01 -4.10478085e-01 -6.17144331e-02 -3.54568541e-01 -2.76569128e-01 9.92011726e-01 -4.79827553e-01 2.91454852e-01 4.48542565e-01 1.38087773e+00 6.08353019e-01 -1.04735088e+00 -1.72229588e-01 7.66701624e-02 -4.39886451e-01 1.31541491e-01 -8.29877555e-01 -4.84388053e-01 1.02520418e+00 7.93321073e-01 2.37664089e-01 8.62139285e-01 -6.47180900e-02 1.52989006e+00 -1.50635876e-02 -2.70001665e-02 -1.20186615e+00 9.95124876e-02 8.41619551e-01 8.25343847e-01 -8.63319874e-01 -5.84556758e-01 -3.87237549e-01 -4.48963344e-01 9.45809782e-01 3.27103794e-01 4.85607058e-01 5.73887408e-01 4.28644449e-01 1.87067196e-01 4.02496010e-01 -5.07184267e-01 -3.20831761e-02 2.69729555e-01 6.62631989e-01 3.73318791e-01 -1.56435445e-01 3.53489071e-01 1.02128983e+00 -5.53560019e-01 -2.56017655e-01 4.61934268e-01 8.93299282e-01 -3.83107662e-01 -1.12475407e+00 -6.10085785e-01 1.41163871e-01 -9.35342371e-01 -2.05561638e-01 -4.55302864e-01 2.21960485e-01 1.59663633e-01 1.44911063e+00 2.53571887e-02 -6.95374906e-01 5.21962821e-01 4.70558226e-01 7.19879866e-02 -5.60066938e-01 -7.99751341e-01 5.09999692e-01 2.30989829e-01 -3.41162205e-01 1.83185786e-01 -8.39216411e-01 -1.26907301e+00 -1.44024417e-01 -5.79773903e-01 1.21112466e-01 7.16088653e-01 9.20240700e-01 7.36690462e-01 6.38744533e-01 5.12500703e-01 -7.37794995e-01 -6.02213264e-01 -1.19601738e+00 -8.12271893e-01 4.28344190e-01 8.78080904e-01 -4.87254798e-01 -4.78684396e-01 1.86215311e-01]	[14.472946166992188, 5.9093918800354]
b005ab84-4687-4039-9a60-e113c26944f5	fmodetect-robust-detection-and-trajectory	2012.08216	null	https://arxiv.org/abs/2012.08216v2	https://arxiv.org/pdf/2012.08216v2.pdf	FMODetect: Robust Detection of Fast Moving Objects	We propose the first learning-based approach for fast moving objects detection. Such objects are highly blurred and move over large distances within one video frame. Fast moving objects are associated with a deblurring and matting problem, also called deblatting. We show that the separation of deblatting into consecutive matting and deblurring allows achieving real-time performance, i.e. an order of magnitude speed-up, and thus enabling new classes of application. The proposed method detects fast moving objects as a truncated distance function to the trajectory by learning from synthetic data. For the sharp appearance estimation and accurate trajectory estimation, we propose a matting and fitting network that estimates the blurred appearance without background, followed by an energy minimization based deblurring. The state-of-the-art methods are outperformed in terms of recall, precision, trajectory estimation, and sharp appearance reconstruction. Compared to other methods, such as deblatting, the inference is of several orders of magnitude faster and allows applications such as real-time fast moving object detection and retrieval in large video collections.	['Martin R. Oswald', 'Marc Pollefeys', 'Filip Sroubek', 'Jiri Matas', 'Denys Rozumnyi']	2020-12-15	null	http://openaccess.thecvf.com//content/ICCV2021/html/Rozumnyi_FMODetect_Robust_Detection_of_Fast_Moving_Objects_ICCV_2021_paper.html	http://openaccess.thecvf.com//content/ICCV2021/papers/Rozumnyi_FMODetect_Robust_Detection_of_Fast_Moving_Objects_ICCV_2021_paper.pdf	iccv-2021-1	['moving-object-detection']	['computer-vision']	[ 2.18784437e-01 -5.96435070e-01 1.02282517e-01 9.43651889e-03 -7.23856866e-01 -5.48148334e-01 5.41483939e-01 -2.50796080e-01 -4.27849472e-01 5.91700435e-01 -5.70926405e-02 5.49402051e-02 -1.03736751e-01 -2.82497853e-01 -1.06381929e+00 -8.44816744e-01 -3.10441554e-01 3.59600276e-01 6.44543767e-01 5.03065407e-01 1.70684367e-01 4.00261611e-01 -1.52344620e+00 1.30515829e-01 8.48246038e-01 9.84862447e-01 4.51399654e-01 1.04892230e+00 2.54752904e-01 8.40611160e-01 -5.20724237e-01 -2.46841952e-01 3.17095011e-01 -1.43043995e-01 -4.89005327e-01 2.37329081e-01 1.18320310e+00 -8.10663879e-01 -6.72416687e-01 1.16289032e+00 2.23327532e-01 4.78656799e-01 6.03993535e-01 -1.10146511e+00 -7.49543488e-01 1.54238656e-01 -9.94546473e-01 4.77466643e-01 4.20481622e-01 2.54320297e-02 4.11274850e-01 -1.11456168e+00 6.31052554e-01 1.36602342e+00 8.06279719e-01 4.33935404e-01 -1.08654225e+00 -4.61836547e-01 9.85311493e-02 5.65226078e-01 -1.58343399e+00 -7.16952026e-01 4.47971582e-01 -6.90484107e-01 3.23991716e-01 5.10140598e-01 4.82622623e-01 8.35707545e-01 1.62797958e-01 1.17636216e+00 8.61221015e-01 -2.61130810e-01 1.48055956e-01 -1.14085749e-01 1.73733190e-01 7.71163523e-01 4.27511632e-01 1.42257884e-01 -3.00836295e-01 -5.87937161e-02 9.27074730e-01 3.67208958e-01 -7.16447711e-01 -2.41255656e-01 -1.55684114e+00 2.42490247e-01 5.33926077e-02 5.57939373e-02 -3.04419905e-01 4.72695917e-01 2.77301192e-01 9.78238583e-02 7.88722515e-01 1.62243098e-01 -1.38451904e-01 -3.19025248e-01 -1.64939713e+00 2.59231776e-01 4.07959878e-01 1.10941303e+00 4.60177094e-01 7.00490996e-02 -4.43206459e-01 6.80194914e-01 4.85306308e-02 9.53195274e-01 3.10634077e-01 -1.09332073e+00 2.58911878e-01 -2.15765629e-02 7.58046150e-01 -1.08682632e+00 -1.58152670e-01 -2.07643852e-01 -1.03349459e+00 2.16579363e-01 6.77375376e-01 8.05856213e-02 -9.09441590e-01 1.53678381e+00 4.36295837e-01 8.82895529e-01 -2.50978500e-01 1.28835273e+00 3.83975118e-01 9.49789882e-01 -3.68293971e-01 -5.75779557e-01 1.34712493e+00 -1.16931725e+00 -8.91181648e-01 -4.07422520e-02 1.51346281e-01 -9.87635076e-01 7.71268845e-01 6.73023880e-01 -1.39853477e+00 -7.13694155e-01 -8.50805461e-01 1.14534654e-01 8.84679556e-02 2.74913400e-01 2.76907146e-01 4.80688632e-01 -1.01483226e+00 6.43795311e-01 -1.00246382e+00 -3.63581270e-01 3.98679107e-01 1.24580570e-01 -1.60832182e-01 -4.28423136e-01 -7.63596952e-01 1.00990999e+00 1.44969344e-01 9.90446806e-02 -1.10581994e+00 -8.40418518e-01 -6.30546927e-01 -1.78692669e-01 2.45756224e-01 -8.65595162e-01 1.10840666e+00 -1.04452956e+00 -1.25978816e+00 8.28372180e-01 -3.61399174e-01 -5.61718822e-01 1.03349698e+00 -8.58695924e-01 -5.11870980e-01 3.00712645e-01 -7.86143914e-02 4.13888872e-01 1.67746615e+00 -1.07530189e+00 -7.05668569e-01 -1.62305683e-01 -3.16184193e-01 1.37329161e-01 -1.92531079e-01 2.62027919e-01 -7.70338476e-01 -9.51551974e-01 -1.27561063e-01 -9.35315073e-01 5.93484752e-02 3.92662019e-01 -1.24337658e-01 2.75237381e-01 1.29849815e+00 -1.13561845e+00 1.33649719e+00 -2.07647276e+00 3.93055439e-01 -4.62459177e-01 5.49309313e-01 3.83184016e-01 -8.74078795e-02 -8.51727799e-02 6.01008199e-02 -4.34179813e-01 -7.97496364e-02 -4.35225010e-01 -3.52441847e-01 -2.86203921e-01 -3.08315217e-01 1.07701898e+00 -2.63686508e-01 9.03134227e-01 -1.23907602e+00 -2.67204821e-01 5.27102292e-01 3.29227924e-01 -3.14309709e-02 4.06888455e-01 4.50290032e-02 2.86063671e-01 -5.35066053e-02 5.34644425e-01 1.16033673e+00 -2.70483226e-01 -3.70709211e-01 -4.79222029e-01 -1.57494575e-01 -5.00616670e-01 -1.11897373e+00 1.52011871e+00 -4.45513457e-01 1.13683009e+00 2.01834887e-01 -6.01180434e-01 5.10587513e-01 1.62053183e-01 4.87154752e-01 -2.57353317e-02 -6.51114210e-02 1.00188985e-01 -4.78979319e-01 -7.93502271e-01 8.79703522e-01 1.20775476e-01 4.08655435e-01 3.41836452e-01 -3.83909911e-01 1.60944462e-01 5.80486618e-02 1.68308020e-01 1.13770986e+00 1.58650503e-01 1.30957171e-01 -1.95307001e-01 4.37593728e-01 -2.77889758e-01 2.63193816e-01 7.58398473e-01 -1.00093946e-01 7.85684466e-01 -3.89075816e-01 -6.82673037e-01 -1.21376443e+00 -1.21023965e+00 -8.14950988e-02 7.83485830e-01 7.34799087e-01 -5.30825257e-02 -9.80520546e-01 -4.47275877e-01 -7.40643358e-03 6.30480409e-01 -5.03166795e-01 -1.70782700e-01 -7.43214309e-01 -8.31844032e-01 2.84772485e-01 4.60174322e-01 5.03353417e-01 -5.59353888e-01 -6.45472348e-01 9.03260186e-02 -4.98755932e-01 -1.35969913e+00 -1.15287209e+00 -5.33485711e-01 -9.76013958e-01 -9.22341764e-01 -1.48619819e+00 -5.52125812e-01 8.62617075e-01 8.80183101e-01 9.10462320e-01 -1.01130657e-01 -3.72000515e-01 4.20364231e-01 -1.73791721e-01 2.73497313e-01 -5.20656228e-01 -4.98473108e-01 4.28475052e-01 4.01496440e-01 -5.32950945e-02 -2.83688456e-01 -8.94917488e-01 5.26925743e-01 -9.82352436e-01 2.96298832e-01 6.91551745e-01 6.32661819e-01 3.11059624e-01 1.59676641e-01 1.12605043e-01 -3.63345146e-01 2.49467164e-01 -3.00592244e-01 -9.16326642e-01 3.90281260e-01 -5.38892806e-01 -1.19385161e-01 4.31648254e-01 -1.10518610e+00 -1.10720658e+00 6.35048524e-02 4.24361616e-01 -1.00709617e+00 -6.40090704e-02 -9.83629897e-02 2.60259867e-01 -3.09274763e-01 5.43663561e-01 5.37120402e-01 -1.12653434e-01 -4.79027808e-01 6.17003322e-01 6.84809685e-01 8.29991043e-01 -1.68759152e-01 9.84407306e-01 7.20693171e-01 -1.64040089e-01 -9.64300811e-01 -7.17039108e-01 -7.68806994e-01 -5.46458423e-01 -6.77274287e-01 8.72339070e-01 -9.79046941e-01 -5.51583230e-01 1.00580943e+00 -1.34664750e+00 -3.36601466e-01 7.59297190e-03 6.15962863e-01 -6.68617427e-01 9.64717984e-01 -9.02094364e-01 -8.62152100e-01 -4.27542120e-01 -8.81031871e-01 1.11710620e+00 9.50795412e-02 -4.34967875e-03 -8.07326734e-01 -1.14676636e-02 3.62957805e-01 4.45372373e-01 1.24999985e-01 3.34958822e-01 6.45328835e-02 -9.75093305e-01 -3.89055103e-01 -5.58193684e-01 2.19534114e-01 2.57273287e-01 -4.43313420e-02 -8.99403930e-01 -6.66496575e-01 1.61359996e-01 2.03429207e-01 1.02639234e+00 9.35968697e-01 1.05242705e+00 -5.75513542e-01 -5.96000016e-01 7.99149394e-01 1.22243297e+00 1.02546677e-01 7.68085122e-01 1.71787843e-01 9.23201561e-01 1.89766243e-01 9.18775082e-01 2.85029709e-01 -2.65157782e-02 1.06224132e+00 3.11070770e-01 6.49346560e-02 -4.72007066e-01 1.84027821e-01 5.87813914e-01 7.45086491e-01 -1.96962610e-01 -3.29426199e-01 -5.60441196e-01 7.32075691e-01 -2.19379735e+00 -1.27338541e+00 -3.94932806e-01 2.70154619e+00 6.55161560e-01 -3.87616232e-02 8.91356096e-02 -3.32797706e-01 1.14270127e+00 1.90866709e-01 -5.54414809e-01 3.54743928e-01 1.44335460e-02 -4.87460196e-01 7.42171407e-01 7.10524440e-01 -1.19045556e+00 9.47084725e-01 6.39901304e+00 1.10300374e+00 -9.84940410e-01 4.14304286e-01 6.87554896e-01 -9.81981903e-02 2.21034527e-01 -3.52628767e-01 -7.26129889e-01 7.50306964e-01 7.21627593e-01 -4.93676960e-02 8.32194030e-01 6.99904740e-01 5.86686730e-01 -2.35107273e-01 -1.03390765e+00 1.42431569e+00 2.98451483e-01 -1.39725173e+00 -9.25744548e-02 -2.09706515e-01 1.05475581e+00 -6.60359636e-02 1.08291432e-01 -1.70810580e-01 -4.95687127e-02 -6.72891498e-01 1.10979283e+00 8.27705085e-01 1.02890623e+00 -3.89061779e-01 5.55975676e-01 3.93057495e-01 -1.24063218e+00 8.92175548e-03 -5.24031639e-01 2.48137549e-01 4.93836462e-01 9.13781226e-01 -4.27547514e-01 2.84802139e-01 7.09360480e-01 8.72128606e-01 -4.03623730e-01 1.65014160e+00 2.29122072e-01 3.95583212e-01 -1.14518605e-01 1.83443308e-01 -9.94867682e-02 -2.93819636e-01 9.43473697e-01 1.53614914e+00 6.86343610e-01 -6.12835214e-02 5.51533774e-02 7.69489884e-01 4.10131179e-02 -3.39477390e-01 -1.83708504e-01 2.25701079e-01 4.07511890e-01 1.26422787e+00 -7.89273739e-01 -8.07745516e-01 -2.20683768e-01 1.70534921e+00 1.49559779e-02 4.97605115e-01 -1.22890854e+00 -2.36751676e-01 5.90373814e-01 2.54371405e-01 5.08194625e-01 -3.04941118e-01 3.17189932e-01 -1.51247680e+00 6.52081743e-02 -6.20860219e-01 5.53792119e-02 -1.02345383e+00 -1.25019610e+00 4.86494005e-01 2.20054314e-01 -1.54208779e+00 -3.35159781e-03 -5.69627404e-01 -6.18892729e-01 5.22532225e-01 -1.22809291e+00 -9.87395883e-01 -7.38264620e-01 3.67830306e-01 1.04322183e+00 1.95821360e-01 2.35204384e-01 5.49819767e-01 -5.01073420e-01 2.83567876e-01 5.27587533e-01 -1.53697699e-01 9.83771026e-01 -1.15443408e+00 5.32700598e-01 1.18812180e+00 3.86093333e-02 3.13300550e-01 9.65640903e-01 -7.18316197e-01 -1.53135467e+00 -1.41258371e+00 1.98076323e-01 -5.11557758e-01 5.62256277e-01 -1.73663184e-01 -9.46948707e-01 4.23361510e-01 9.74133611e-02 2.03282356e-01 -1.67230129e-01 -5.21196544e-01 -7.00225234e-02 -2.58289397e-01 -9.50561345e-01 4.99854952e-01 1.04162753e+00 -1.31748587e-01 -3.45087856e-01 5.58287621e-01 6.73389852e-01 -6.30819440e-01 -5.31895638e-01 6.54273853e-02 7.05060363e-01 -7.11710811e-01 1.25183296e+00 -3.10548812e-01 1.52874574e-01 -6.62116349e-01 1.16759077e-01 -1.09699726e+00 -6.68976724e-01 -1.12585509e+00 -1.00299454e+00 8.86326671e-01 -1.64394185e-01 -1.35942325e-01 6.64302528e-01 2.48514891e-01 -5.85595053e-03 -3.77338469e-01 -7.71940351e-01 -1.12749755e+00 -5.41640460e-01 -3.18692356e-01 1.49966136e-01 8.63830030e-01 -3.68205279e-01 -1.04818739e-01 -1.03618217e+00 4.15701061e-01 1.20232797e+00 2.66643018e-01 7.26188600e-01 -9.46714222e-01 -3.49282712e-01 -2.95315653e-01 -4.63931620e-01 -1.80798280e+00 -2.10055694e-01 -2.74835974e-01 4.06709194e-01 -1.34620607e+00 5.30769825e-01 -1.65349483e-01 -1.38918355e-01 -3.05506319e-01 -4.55175817e-01 3.78492773e-01 6.67973384e-02 5.46071708e-01 -8.56983185e-01 2.65875161e-01 1.17974913e+00 -3.14440668e-01 -5.26319668e-02 2.52795070e-01 3.33912000e-02 9.00753021e-01 3.22984070e-01 -4.52748150e-01 -1.01173654e-01 -6.09003186e-01 -7.06178993e-02 2.45021075e-01 5.40237069e-01 -1.15052450e+00 3.43912750e-01 -2.25137159e-01 5.02628982e-01 -7.40331113e-01 5.61262965e-01 -6.80481613e-01 3.96301210e-01 4.61066097e-01 -1.08640537e-01 8.89005661e-02 1.85701966e-01 9.95668530e-01 8.71681720e-02 -3.33827257e-01 9.74472165e-01 7.38215446e-03 -8.27513874e-01 4.31537777e-01 -6.15170598e-01 -3.11873052e-02 1.21414948e+00 -3.03915441e-01 -4.17069107e-01 -7.01447964e-01 -5.76409042e-01 -2.44866103e-01 6.27226830e-01 4.64602619e-01 9.06952977e-01 -1.30016983e+00 -7.74224460e-01 -1.65928125e-01 -2.32075661e-01 -2.98622787e-01 4.43769515e-01 1.26817405e+00 -7.34126270e-01 1.39214694e-01 -2.01002788e-02 -7.96723664e-01 -1.59107959e+00 9.81755674e-01 2.98018724e-01 2.37710495e-02 -8.47691536e-01 8.79158854e-01 5.16564727e-01 4.85661447e-01 3.02536219e-01 -4.83984917e-01 3.23619932e-01 -2.66484529e-01 1.02849257e+00 8.72941971e-01 -1.66359410e-01 -6.59832299e-01 -2.17079833e-01 8.49250317e-01 -1.46754533e-01 3.78644392e-02 1.05282712e+00 -5.59875667e-01 -7.53493160e-02 3.34682465e-01 1.03169858e+00 1.35745376e-01 -1.76755428e+00 -1.81424037e-01 -1.72049388e-01 -1.02568424e+00 2.21299246e-01 -5.90478122e-01 -1.05825543e+00 7.10672021e-01 9.63630617e-01 7.09707662e-02 1.08617020e+00 4.34763543e-02 1.10400224e+00 1.82972699e-01 2.59620845e-01 -7.68889487e-01 2.59474099e-01 1.24500237e-01 7.86891520e-01 -1.08236873e+00 1.95467010e-01 -4.58504498e-01 -2.56856024e-01 1.10235512e+00 4.34498519e-01 -1.33404240e-01 1.97345600e-01 3.71473134e-01 -1.61738560e-01 8.72416645e-02 -5.12493730e-01 1.60504341e-01 7.05418706e-01 5.03072619e-01 -1.26425073e-01 -4.92484234e-02 1.23269811e-01 -5.66169433e-02 4.77911800e-01 -1.09107956e-01 5.35760701e-01 4.59849089e-01 -6.62760675e-01 -4.78623629e-01 -8.74764562e-01 2.85135597e-01 -4.12404388e-01 -1.47566646e-01 3.91624048e-02 4.12792534e-01 9.78178531e-03 7.65506685e-01 1.65792152e-01 -2.37383395e-01 -1.01729035e-01 -2.44502336e-01 7.83119321e-01 -8.90610442e-02 4.85503562e-02 8.60403255e-02 -1.40322313e-01 -5.95014572e-01 -4.28066462e-01 -7.22042918e-01 -6.79201543e-01 -4.69186068e-01 -7.14194417e-01 -1.49552599e-01 3.70106012e-01 7.53776968e-01 3.12305033e-01 2.91213661e-01 4.39302176e-01 -1.24845779e+00 -4.61625129e-01 -1.02472997e+00 -4.57950294e-01 4.97681946e-01 7.73945212e-01 -6.23526573e-01 -4.32903856e-01 5.93908310e-01]	[11.463141441345215, -2.54290509223938]
6d23317c-db07-4477-bbae-b6bd56e4b22f	embeddings-evaluation-using-a-novel-measure	null	null	https://link.springer.com/article/10.1007/s12559-021-09987-7	https://link.springer.com/epdf/10.1007/s12559-021-09987-7?sharing_token=k8n-ML0c5GR2MNCx1njYEPe4RwlQNchNByi7wbcMAY6OYcAV1IB_c4V9xMrd-RK6fL_MvX93E8KnZ0fIlJ9W-PFNTzF5q4CKJ1aaVhvbV71JZEvoWEbRsnSKq-q51QeowZBeBcJBoAcMLoXGlGG6GYKm5hc9TXk8vMyPSA4-pD0%3D	Embeddings Evaluation Using a Novel Measure of Semantic Similarity	Lexical taxonomies and distributional representations are largely used to support a wide range of NLP applications, including semantic similarity measurements. Recently, several scholars have proposed new approaches to combine those resources into unified representation preserving distributional and knowledge-based lexical features. In this paper, we propose and implement TaxoVec, a novel approach to selecting word embeddings based on their ability to preserve taxonomic similarity. In TaxoVec, we first compute the pairwise semantic similarity between taxonomic words through a new measure we previously developed, the Hierarchical Semantic Similarity (HSS), which we show outperforms previous measures on several benchmark tasks. Then, we train several embedding models on a text corpus and select the best model, that is, the model that maximizes the correlation between the HSS and the cosine similarity of the pair of words that are in both the taxonomy and the corpus. To evaluate TaxoVec, we repeat the embedding selection process using three other semantic similarity benchmark measures. We use the vectors of the four selected embeddings as machine learning model features to perform several NLP tasks. The performances of those tasks constitute an extrinsic evaluation of the criteria for the selection of the best embedding (i.e. the adopted semantic similarity measure). Experimental results show that (i) HSS outperforms state-of-the-art measures for measuring semantic similarity in taxonomy on a benchmark intrinsic evaluation and (ii) the embedding selected through TaxoVec achieves a clear victory against embeddings selected by the competing measures on benchmark NLP tasks. We implemented the HSS, together with other benchmark measures of semantic similarity, as a fully-fledged Python package called TaxoSS, whose documentation is available at https:// pypi.org/project/TaxoSS.	['Navid Nobani', 'Mario Mezzanzanica', 'Fabio Mercorio', 'Lorenzo Malandri', 'Anna Giabelli']	2022-01-08	null	null	null	cognitive-computation-2022-1	['embeddings-evaluation']	['natural-language-processing']	[-8.29221010e-02 -3.77384812e-01 -3.60219628e-01 -3.37274849e-01 -5.65870106e-01 -6.67808115e-01 7.79254258e-01 7.56526113e-01 -8.77207100e-01 3.55371922e-01 5.37843227e-01 -2.54917424e-02 -5.75681150e-01 -8.84355724e-01 9.48605835e-02 -5.59343338e-01 1.61408305e-01 4.74385411e-01 2.60202616e-01 -1.62504926e-01 7.24650443e-01 3.59666288e-01 -1.92435002e+00 3.84947620e-02 1.07724786e+00 1.08145893e+00 3.90277892e-01 8.87181461e-02 -4.54935282e-01 -2.51769684e-02 -3.82827967e-01 -4.76963669e-01 1.02766871e-01 -3.52354437e-01 -9.19870138e-01 -6.36150062e-01 1.19654559e-01 4.29542452e-01 6.74757967e-03 1.26190150e+00 5.13971508e-01 3.81089836e-01 9.19301629e-01 -1.24513841e+00 -8.72362316e-01 5.14213800e-01 -5.04797250e-02 3.22407663e-01 5.88093579e-01 -2.13678494e-01 1.64415157e+00 -1.31243515e+00 7.24413633e-01 1.28302145e+00 6.09054863e-01 2.44120449e-01 -1.25081575e+00 -5.29911458e-01 -1.85104892e-01 5.49085081e-01 -1.68982494e+00 -7.22195879e-02 5.74352145e-01 -5.37429810e-01 1.15562987e+00 2.21009478e-01 3.09259862e-01 9.53862965e-01 1.81941926e-01 1.86454907e-01 1.05674148e+00 -6.13240182e-01 3.15895170e-01 4.48580235e-01 4.70215470e-01 4.30869967e-01 3.61853898e-01 -7.30533525e-02 -5.41186810e-01 -4.86274809e-01 3.14886533e-02 -9.12207142e-02 -2.47798264e-01 -3.76143008e-01 -1.26344454e+00 1.27394187e+00 3.66612345e-01 7.81979501e-01 -3.16612870e-01 -3.15906972e-01 4.97470677e-01 3.24286222e-01 4.85193491e-01 8.72650385e-01 -3.99131000e-01 -9.23484713e-02 -6.06329978e-01 2.94151992e-01 7.39257991e-01 5.96346021e-01 7.64338255e-01 -4.04860616e-01 -1.37079000e-01 1.15788877e+00 2.89541125e-01 1.31926596e-01 1.16119993e+00 -6.44672275e-01 1.32416040e-01 8.95148993e-01 -1.31874293e-01 -1.22312760e+00 -4.42306310e-01 -1.81735560e-01 -3.79306078e-01 -2.03272596e-01 2.77521312e-02 2.63169944e-01 -3.87028486e-01 1.72723699e+00 4.01239008e-01 -1.39373969e-02 2.49572664e-01 6.98633015e-01 1.05177462e+00 6.39704645e-01 2.24172205e-01 -5.51905409e-02 1.54155898e+00 -6.01480842e-01 -6.01665735e-01 -1.36217279e-02 9.64775145e-01 -9.17177200e-01 1.34903538e+00 -8.70365370e-03 -6.05291188e-01 -3.90039742e-01 -1.00747085e+00 -7.98313543e-02 -9.71948028e-01 -4.35579978e-02 3.82015884e-01 5.01779258e-01 -8.91350925e-01 9.35176432e-01 -4.22357410e-01 -6.83142543e-01 -1.17070787e-01 3.47091481e-02 -5.23319006e-01 5.82433343e-02 -1.60312867e+00 1.29434121e+00 9.62757885e-01 -6.01013124e-01 -2.16813445e-01 -6.78365350e-01 -8.87284040e-01 3.14047724e-01 1.38193399e-01 -6.88405752e-01 6.31963909e-01 -5.89890957e-01 -1.16541636e+00 1.13271427e+00 -3.12945829e-03 -3.11463267e-01 -9.28279907e-02 -3.94547470e-02 -5.41177154e-01 5.67090549e-02 5.34828663e-01 5.94925463e-01 3.46782684e-01 -9.44184601e-01 -7.77431726e-01 -3.93724829e-01 -2.12954003e-02 3.37724358e-01 -9.25831735e-01 2.59133607e-01 -1.48990646e-01 -7.55267859e-01 3.94283570e-02 -7.67521381e-01 1.02180943e-01 -1.32626101e-01 -3.45835313e-02 -8.84572268e-01 4.08500403e-01 -3.86911005e-01 1.55297899e+00 -2.21872783e+00 4.20940191e-01 2.84654409e-01 3.22682336e-02 4.42577362e-01 -3.28779548e-01 7.12533712e-01 -1.02423809e-01 2.59120166e-01 -3.38954121e-01 -2.81514406e-01 3.21911573e-01 1.81387797e-01 -1.85575917e-01 3.21842194e-01 -1.21062659e-01 6.30602121e-01 -1.02258289e+00 -5.49569249e-01 1.33804232e-01 3.98682863e-01 -4.59237754e-01 1.33586228e-01 5.91877028e-02 -1.31650314e-01 -2.97184736e-01 3.72644633e-01 4.40606266e-01 2.97719575e-02 3.24610561e-01 -2.42671177e-01 -7.82006606e-02 6.12369478e-01 -1.11327147e+00 1.64847028e+00 -4.99793559e-01 3.32474887e-01 -6.99795485e-01 -1.03755188e+00 1.28468621e+00 2.19994813e-01 4.25392866e-01 -6.17867768e-01 1.74223915e-01 5.34892023e-01 5.01236580e-02 -2.68627673e-01 7.06542492e-01 -3.17986101e-01 -2.94696718e-01 5.08474171e-01 4.80018824e-01 -1.41608939e-01 3.52043688e-01 1.38107330e-01 9.83247936e-01 -2.62762010e-01 9.18423533e-01 -6.02415085e-01 7.78329968e-01 -1.14654884e-01 5.29062748e-01 3.00630957e-01 -2.64827907e-01 4.09489155e-01 4.62969273e-01 -1.47738904e-01 -8.78115714e-01 -1.19433868e+00 -4.43290710e-01 1.17140949e+00 2.76742220e-01 -8.81457090e-01 -5.40397406e-01 -6.70651078e-01 2.11893365e-01 1.10386252e+00 -6.95866287e-01 -3.77962887e-01 -1.56249344e-01 -5.95834494e-01 5.90683758e-01 4.12307829e-01 8.21639746e-02 -1.21819115e+00 -3.58978361e-01 2.78554205e-02 -2.32852831e-01 -8.63513827e-01 -4.37940657e-01 9.77663025e-02 -5.89824259e-01 -1.01125181e+00 -4.10832196e-01 -8.47604334e-01 5.49526699e-02 3.86688948e-01 9.52775776e-01 -3.69940177e-02 -1.81659400e-01 2.92380333e-01 -8.89771044e-01 -2.04441071e-01 -1.66239500e-01 2.36486942e-01 3.23806882e-01 -5.86965382e-02 8.57743263e-01 -6.59726977e-01 -3.83078992e-01 2.24254563e-01 -9.17823851e-01 -5.68585336e-01 1.04685023e-01 8.74171734e-01 6.29236281e-01 2.41797045e-02 6.62237406e-01 -5.27655959e-01 1.08774483e+00 -6.82676911e-01 -3.70828956e-01 2.66106933e-01 -8.51239264e-01 2.26546407e-01 6.49925351e-01 -3.12401175e-01 -6.98845387e-01 -5.12275398e-01 -1.22163638e-01 -1.93364710e-01 1.80321429e-02 7.76154816e-01 -1.38062015e-01 4.35243435e-02 5.37239671e-01 2.20738292e-01 -2.73111582e-01 -5.76557755e-01 5.01800358e-01 8.88532698e-01 2.22752184e-01 -5.85877717e-01 6.10966384e-01 1.89483643e-01 -2.87484169e-01 -8.80846560e-01 -8.73542368e-01 -9.58710968e-01 -7.30542958e-01 3.65815192e-01 8.83003592e-01 -4.51232821e-01 -4.06005770e-01 -9.21619236e-02 -1.34306562e+00 4.21329409e-01 -2.30536535e-01 8.89624774e-01 -4.59076554e-01 4.73264754e-01 -3.02795917e-02 -3.60805541e-01 -5.19669533e-01 -8.87375951e-01 1.03218997e+00 2.43496839e-02 -7.90909886e-01 -1.35518169e+00 5.29989719e-01 7.56288469e-02 2.88791358e-01 1.42915919e-01 1.41244948e+00 -1.25382710e+00 2.10628465e-01 -2.05641389e-01 -3.08050871e-01 4.84337419e-01 1.86010137e-01 -1.47700876e-01 -7.52690971e-01 -2.65259355e-01 4.36390825e-02 -1.35361820e-01 8.78656030e-01 3.08466088e-02 9.27761197e-01 -3.84900607e-02 -4.02868718e-01 6.36954784e-01 1.64951444e+00 2.65058458e-01 4.45445985e-01 6.87873125e-01 4.08734202e-01 7.68363535e-01 8.34406197e-01 3.98281217e-01 2.67031789e-01 9.79736507e-01 1.65154576e-01 4.45891470e-01 9.08539072e-02 -3.22310746e-01 3.17577779e-01 1.17498159e+00 7.54330382e-02 -2.00168967e-01 -9.48808432e-01 6.36790037e-01 -1.74050534e+00 -8.81361187e-01 1.27847210e-01 2.28730130e+00 7.62159228e-01 -1.03400923e-01 -2.87492853e-02 1.50277331e-01 6.59927964e-01 2.07335368e-01 -1.78809315e-01 -7.21449375e-01 -1.36885807e-01 5.58645308e-01 1.72252223e-01 5.06792724e-01 -1.00035608e+00 1.10082424e+00 5.23003149e+00 1.15455592e+00 -8.08543086e-01 3.62282097e-01 -2.33667657e-01 9.19619501e-02 -4.68680948e-01 5.09942919e-02 -8.31675410e-01 5.37797570e-01 1.00583196e+00 -8.11068118e-01 2.86672097e-02 8.93378615e-01 -4.29563969e-02 1.11871138e-01 -1.10518479e+00 8.59869242e-01 2.85521597e-01 -1.10310662e+00 3.82750332e-01 2.83295587e-02 4.73359853e-01 -5.51320016e-02 7.29804207e-03 3.35342377e-01 1.92829564e-01 -1.00669992e+00 4.63588148e-01 1.98444933e-01 5.84169209e-01 -8.45650434e-01 1.01746237e+00 1.34810850e-01 -1.45337033e+00 -3.68314944e-02 -7.80926228e-01 1.57794476e-01 -5.06383646e-03 4.95388478e-01 -6.17474079e-01 8.02839279e-01 5.77551961e-01 8.12665403e-01 -5.69156885e-01 9.88771975e-01 -3.30267072e-01 2.72348166e-01 -7.52650201e-02 -2.48655081e-01 4.97349650e-01 -3.40117544e-01 7.20784664e-01 1.42231739e+00 5.71584404e-01 -1.00442730e-01 5.54403998e-02 8.42851639e-01 5.10583520e-02 7.51509488e-01 -6.89085364e-01 -1.18951775e-01 1.05125427e+00 1.14418256e+00 -5.87154448e-01 -2.62019992e-01 -2.80436933e-01 8.50802541e-01 5.10374010e-01 -9.62866843e-02 -8.26558173e-01 -9.05378580e-01 1.18192601e+00 -2.60163724e-01 2.49171421e-01 -7.80503824e-03 -2.32993111e-01 -1.14888668e+00 -3.61395180e-02 -6.15894258e-01 6.60238743e-01 -5.45995355e-01 -1.60547793e+00 7.88425326e-01 1.65814936e-01 -1.22803950e+00 -1.23193435e-01 -7.08281696e-01 -6.49872005e-01 9.30886626e-01 -1.32355213e+00 -6.99829340e-01 -2.06212983e-01 4.43511575e-01 3.00820082e-01 -3.50268155e-01 1.29883492e+00 2.57775128e-01 -3.84866923e-01 5.87989271e-01 3.25171620e-01 -8.62895921e-02 9.19024348e-01 -1.24130380e+00 2.93644965e-01 4.67827380e-01 4.20165390e-01 8.04653823e-01 6.24192595e-01 -4.28208977e-01 -7.62360513e-01 -1.02009761e+00 1.51307869e+00 -3.21869284e-01 1.03940272e+00 -2.23501861e-01 -1.10720742e+00 4.12666619e-01 9.47648659e-02 -3.46755415e-01 1.05754399e+00 1.58489794e-01 -6.99040651e-01 8.70000124e-02 -1.04439378e+00 4.25709158e-01 1.02809262e+00 -6.08895242e-01 -1.25538075e+00 2.61781752e-01 1.02251256e+00 3.83735657e-01 -1.15358233e+00 3.20150584e-01 6.07380152e-01 -8.87670040e-01 1.17220247e+00 -5.31331539e-01 3.19011688e-01 -1.25003174e-01 -6.60677195e-01 -1.73665166e+00 -4.92828906e-01 -3.70888114e-02 3.72038841e-01 1.37663651e+00 3.22900802e-01 -9.80401754e-01 2.72219777e-01 -4.77998406e-02 3.94543260e-02 -8.40531945e-01 -1.01250112e+00 -1.23586392e+00 4.77563351e-01 -1.44820109e-01 7.38860130e-01 1.22648692e+00 2.15585321e-01 5.22169113e-01 1.43797010e-01 -5.76201901e-02 4.48802590e-01 1.31797045e-01 3.13495100e-01 -1.58862090e+00 2.36051187e-01 -7.51166344e-01 -8.58851790e-01 -3.76241744e-01 6.46748483e-01 -1.44384694e+00 -2.52858639e-01 -1.51353455e+00 1.51514441e-01 -2.62420237e-01 -5.56672275e-01 2.73340523e-01 -3.14165056e-01 -8.12500045e-02 2.99688280e-01 3.16084355e-01 -3.28741282e-01 9.35850501e-01 6.15646422e-01 6.32985728e-03 -1.49172917e-01 -4.39019471e-01 -7.62327015e-01 8.31526220e-01 1.01455128e+00 -7.50577390e-01 -3.58835846e-01 -1.92532882e-01 9.76673141e-02 -6.44109726e-01 1.41690791e-01 -8.54737878e-01 1.46043748e-01 -2.12987214e-01 -1.75121695e-01 -2.92477518e-01 3.52936804e-01 -6.85756385e-01 -1.35384828e-01 5.17211199e-01 -3.53388280e-01 2.45033160e-01 -4.71708737e-02 3.55901957e-01 -5.32219648e-01 -7.48715639e-01 8.12728167e-01 1.74074546e-01 -9.06093180e-01 7.47361705e-02 -8.66041146e-03 2.78625816e-01 1.01154613e+00 -2.27720782e-01 -2.40404502e-01 8.91019255e-02 -4.02046829e-01 6.21067993e-02 5.33447742e-01 7.24019587e-01 7.29145885e-01 -1.57889509e+00 -7.88060486e-01 -1.34523973e-01 6.73643291e-01 -7.95061648e-01 -1.90488994e-01 5.73438644e-01 -3.44623804e-01 6.82445586e-01 -2.49627188e-01 -1.93369508e-01 -1.45538247e+00 6.01798117e-01 9.13256928e-02 -4.68406796e-01 -4.12968338e-01 6.37329102e-01 2.42715284e-01 -8.16600680e-01 -1.44652007e-02 -2.35300958e-01 -6.12150908e-01 5.67628264e-01 5.26716709e-01 4.50179040e-01 2.09136575e-01 -8.62740338e-01 -6.68272793e-01 8.51243317e-01 9.81235951e-02 -1.03454590e-01 1.38305879e+00 -2.25335415e-02 -3.09124082e-01 4.80958432e-01 1.66037333e+00 -1.14020355e-01 -2.41003722e-01 -5.45574069e-01 4.74567294e-01 -6.00188971e-01 -1.50711136e-02 -5.03448129e-01 -5.55153549e-01 8.06145072e-01 5.69808185e-01 2.04669908e-01 8.17929864e-01 1.64082330e-02 8.07175756e-01 2.43712306e-01 3.30212921e-01 -1.02310789e+00 8.33873078e-02 7.57908463e-01 9.36760545e-01 -9.45420384e-01 -1.88009255e-02 -4.02594566e-01 -7.30797768e-01 1.07093537e+00 3.32579613e-01 -1.45234302e-01 7.85919607e-01 -2.97565997e-01 -1.96694344e-01 -3.29449773e-01 -7.26417184e-01 -4.97843593e-01 5.32081902e-01 5.97590327e-01 6.53752625e-01 3.01787943e-01 -1.05444074e+00 6.74278736e-01 -5.76406837e-01 -4.33517158e-01 8.60657021e-02 4.86245334e-01 -7.78004289e-01 -1.20248425e+00 -1.33219242e-01 3.19621652e-01 -1.29914597e-01 -3.12819928e-01 -6.24817789e-01 7.04052866e-01 2.76753813e-01 8.05173397e-01 1.26664236e-01 -5.13282001e-01 3.72657150e-01 2.27667451e-01 1.89522341e-01 -8.68743896e-01 -5.66050529e-01 -5.49713731e-01 7.47865289e-02 -5.22706807e-01 -3.64150971e-01 -6.10163212e-01 -1.11191475e+00 -3.34381878e-01 -3.07034105e-01 4.25109506e-01 6.53453946e-01 8.77889276e-01 3.27279836e-01 1.63691550e-01 6.01782501e-01 -5.81645072e-01 -7.45904267e-01 -1.07634878e+00 -7.03310668e-01 7.27870524e-01 -3.81930470e-01 -1.13198364e+00 -6.36689782e-01 -5.04231393e-01]	[10.371586799621582, 8.825360298156738]
3ab35518-ec8b-4b9d-b743-dcf519b5ecdd	fine-tuning-bert-with-focus-words-for	null	null	https://aclanthology.org/2020.starsem-1.13	https://aclanthology.org/2020.starsem-1.13.pdf	Fine-tuning BERT with Focus Words for Explanation Regeneration	Explanation generation introduced as the world tree corpus (Jansen et al., 2018) is an emerging NLP task involving multi-hop inference for explaining the correct answer in multiple-choice QA. It is a challenging task evidenced by low state-of-the-art performances(below 60{\%} in F-score) demonstrated on the task. Of the state-of-the-art approaches, fine-tuned transformer-based (Vaswani et al., 2017) BERT models have shown great promise toward continued system performance improvements compared with approaches relying on surface-level cues alone that demonstrate performance saturation. In this work, we take a novel direction by addressing a particular linguistic characteristic of the data {---} we introduce a novel and lightweight focus feature in the transformer-based model and examine task improvements. Our evaluations reveal a significantly positive impact of this lightweight focus feature achieving the highest scores, second only to a significantly computationally intensive system.	['S{\\"o}ren Auer', "Jennifer D{'}Souza", "Isaiah Onando Mulang{'}"]	2020-12-01	null	null	null	joint-conference-on-lexical-and-computational	['multiple-choice-qa']	['natural-language-processing']	[ 3.75168949e-01 8.49425316e-01 -7.17006698e-02 -3.82250428e-01 -1.75647438e+00 -5.62548339e-01 9.62187767e-01 2.38683477e-01 -2.50478923e-01 1.01236892e+00 5.80920279e-01 -7.06401587e-01 -2.52913952e-01 -2.38231868e-01 -7.27308512e-01 -2.36741871e-01 1.35622442e-01 8.81505430e-01 2.38300219e-01 -7.00784683e-01 4.17592525e-01 -8.72439146e-02 -1.63112342e+00 6.32515132e-01 1.21535552e+00 8.28080475e-01 3.47223468e-02 6.41585588e-01 -1.94631860e-01 7.98243463e-01 -5.97405910e-01 -6.48951471e-01 -6.22187965e-02 -3.84158432e-01 -1.23987782e+00 -5.95030248e-01 8.45473289e-01 7.22587481e-02 1.85026154e-01 6.53469861e-01 4.57093626e-01 5.66136762e-02 5.75145721e-01 -1.02443492e+00 -5.93840837e-01 7.04923332e-01 -1.07049525e-01 3.47511202e-01 9.56146419e-01 2.99926370e-01 1.60080349e+00 -1.11276770e+00 8.08360934e-01 1.22950888e+00 6.80873275e-01 6.78763509e-01 -1.35510695e+00 -3.91113311e-01 1.83367997e-01 6.80894434e-01 -9.54434454e-01 -5.56706846e-01 3.96572292e-01 -2.72639364e-01 1.53734398e+00 3.09853464e-01 1.23123676e-01 1.13918030e+00 1.09989606e-01 8.57114673e-01 1.41448879e+00 -6.93138301e-01 4.85161915e-02 -1.83643132e-01 2.62297332e-01 7.54682004e-01 4.06318270e-02 6.22484535e-02 -1.03230596e+00 -6.25559911e-02 2.68980805e-02 -7.91829824e-01 -2.63876647e-01 -7.94021338e-02 -1.35753560e+00 8.99244249e-01 4.70685840e-01 2.59227842e-01 -4.63180453e-01 1.87768757e-01 1.15421832e-01 2.21208662e-01 3.55133921e-01 9.82835174e-01 -8.02498221e-01 -7.46821105e-01 -8.95181954e-01 7.91561663e-01 8.47455144e-01 7.57930994e-01 4.38665003e-01 -1.26108408e-01 -6.12236321e-01 5.41027188e-01 1.29039317e-01 2.38760978e-01 2.37686485e-01 -1.06383383e+00 9.14718449e-01 5.51911831e-01 2.50936657e-01 -4.95149821e-01 -6.44236863e-01 -6.56852067e-01 -1.81578949e-01 -1.13806657e-01 8.97197008e-01 9.09488276e-02 -1.02516437e+00 1.89930105e+00 3.71035963e-01 -2.61159748e-01 2.33732298e-01 8.23313475e-01 8.95505071e-01 3.57804358e-01 3.36697102e-01 -2.12915465e-02 1.62979269e+00 -1.11866319e+00 -7.70155191e-01 -4.64569330e-01 6.57291532e-01 -8.10025334e-01 1.52516651e+00 5.46267331e-01 -1.24236548e+00 -4.05911326e-01 -8.52118134e-01 -4.96519029e-01 -3.06011349e-01 -9.52336639e-02 8.00425529e-01 4.10720080e-01 -1.11997342e+00 4.01341528e-01 -4.45151776e-01 -2.15012580e-01 2.76415676e-01 2.59235859e-01 -3.28355163e-01 -1.78017452e-01 -1.36416578e+00 1.39919078e+00 7.11973608e-02 -5.70290349e-02 -6.39764190e-01 -8.50500345e-01 -9.27815497e-01 1.39443755e-01 6.28040373e-01 -8.46896291e-01 1.65980518e+00 -3.19858164e-01 -1.53230858e+00 7.20912516e-01 -7.86178052e-01 -6.46273792e-01 4.58746195e-01 -4.49475259e-01 -3.73816848e-01 9.00289118e-02 5.40866196e-01 7.46804118e-01 4.92064089e-01 -1.29526985e+00 -7.59182096e-01 -1.57055393e-01 2.38372877e-01 2.24772334e-01 1.44034192e-01 -4.03420962e-02 2.42553763e-02 -2.74291098e-01 2.30997950e-01 -9.70147789e-01 -6.96884841e-03 -8.39233100e-01 -4.69513685e-01 -8.11101675e-01 2.35061958e-01 -9.01351988e-01 1.38874042e+00 -1.67823708e+00 2.52554059e-01 -2.34792754e-01 1.32499740e-01 5.97529039e-02 -3.86341773e-02 6.30046964e-01 1.03449583e-01 2.35785857e-01 -3.75802487e-01 -4.54133838e-01 3.96768034e-01 -1.27003770e-02 -4.47376519e-01 -1.04396567e-01 5.59720814e-01 1.18661046e+00 -9.68249917e-01 -4.23989952e-01 8.73452425e-03 2.47388944e-01 -6.20495737e-01 1.02053493e-01 -6.32746279e-01 6.20158434e-01 -1.27406895e-01 6.49971306e-01 1.94585323e-01 -4.21704948e-01 1.03972413e-01 -1.46907285e-01 -2.93738902e-01 1.06379676e+00 -6.93523467e-01 1.90155613e+00 -5.27556181e-01 5.64999223e-01 -2.17733428e-01 -5.25038004e-01 4.47509110e-01 4.93730903e-01 -1.45416707e-01 -9.26967382e-01 -1.07361682e-01 7.78914571e-01 5.34560680e-01 -5.97939074e-01 8.03985298e-01 -8.94675627e-02 -3.19162577e-01 1.19744107e-01 9.81518440e-03 -2.49352008e-01 3.51393610e-01 2.66420007e-01 1.20066798e+00 4.73265886e-01 3.70746553e-01 -5.30171454e-01 4.99486208e-01 5.23376167e-01 2.07073078e-01 9.31194305e-01 -1.17937870e-01 6.54340923e-01 4.16855335e-01 -2.64816314e-01 -8.71217489e-01 -9.08030212e-01 -1.73598498e-01 1.19467640e+00 -7.98382610e-02 -6.05480015e-01 -8.98628533e-01 -6.81939721e-01 -9.01231989e-02 1.23574817e+00 -8.56967866e-01 1.73879966e-01 -7.05050409e-01 -3.87126505e-01 7.38075137e-01 4.57488924e-01 4.12046492e-01 -1.15331268e+00 -7.17979610e-01 3.37802947e-01 -6.91941798e-01 -1.26953220e+00 -8.89717266e-02 1.09973855e-01 -6.73902690e-01 -8.70902956e-01 -3.76286745e-01 -5.28003752e-01 1.23780213e-01 -1.84865460e-01 1.56098974e+00 1.26630515e-01 2.64504366e-02 2.58261234e-01 -3.65484387e-01 -5.18100262e-01 -3.05231303e-01 3.53964776e-01 -2.62153894e-01 -4.12892729e-01 3.86298805e-01 -2.40792990e-01 -5.58787584e-01 -2.97624692e-02 -3.97564143e-01 1.77475154e-01 7.31533408e-01 9.49603140e-01 5.18500209e-01 -6.49723411e-01 8.55058193e-01 -1.03371227e+00 8.32535803e-01 -5.62155008e-01 -1.07270636e-01 2.40720585e-01 -9.33658719e-01 3.53855550e-01 4.87238258e-01 7.90007710e-02 -1.13790524e+00 -3.04210603e-01 -2.38892183e-01 2.51731455e-01 -2.45123714e-01 6.54094934e-01 9.93665457e-02 2.09486604e-01 5.62697172e-01 4.26391512e-02 -2.22064644e-01 -4.81948346e-01 5.67201972e-01 3.30263078e-01 6.74889207e-01 -7.70429790e-01 6.95237279e-01 2.10659191e-01 1.87208787e-01 -2.95246333e-01 -1.36644566e+00 -3.06197464e-01 -4.56532866e-01 7.33746262e-03 1.03437662e+00 -7.70559311e-01 -7.06031263e-01 -4.04610820e-02 -1.46356201e+00 -3.12135428e-01 -1.83080539e-01 3.12253177e-01 -6.04717731e-01 1.44225091e-01 -4.37253326e-01 -8.00253212e-01 -3.48032802e-01 -1.13946581e+00 1.30852759e+00 2.07361624e-01 -6.52137816e-01 -9.65697169e-01 8.59433692e-03 1.00767386e+00 6.78958237e-01 2.45617092e-01 1.16591096e+00 -9.20256317e-01 -6.86371624e-01 1.86373740e-01 -1.79437608e-01 -2.31761694e-01 -3.66408020e-01 -4.75776166e-01 -1.06483507e+00 1.46928383e-02 -5.28015904e-02 -3.32022399e-01 7.37330496e-01 1.66353043e-02 9.29456174e-01 -2.52783716e-01 -1.37096643e-01 1.33824095e-01 1.20201242e+00 -1.03102475e-01 4.91892576e-01 7.00538218e-01 5.63405514e-01 7.47930110e-01 7.38641798e-01 -5.20750098e-02 1.03586531e+00 8.92923295e-01 4.43042397e-01 1.69109538e-01 -4.80185926e-01 -3.98584425e-01 2.08949089e-01 7.50850081e-01 -2.17889920e-01 -3.85456502e-01 -1.12053919e+00 8.20576370e-01 -1.95640552e+00 -8.36690962e-01 -6.31835103e-01 2.14077997e+00 8.13248098e-01 3.55519861e-01 -1.88420087e-01 1.30977452e-01 2.90846735e-01 -3.55198048e-02 -2.64395654e-01 -6.58549011e-01 -2.25490004e-01 7.22394764e-01 -2.89585367e-02 9.48374569e-01 -6.44982278e-01 1.16485083e+00 6.30264235e+00 7.92430103e-01 -7.39089429e-01 2.80039579e-01 2.35282823e-01 5.64412540e-03 -7.77953148e-01 3.03045362e-01 -8.38539004e-01 3.26376230e-01 1.21659398e+00 7.37896115e-02 4.50906843e-01 3.92804652e-01 -7.62114897e-02 -2.44403079e-01 -1.28173733e+00 5.47069788e-01 3.04747939e-01 -1.37419558e+00 6.46429658e-02 4.58757319e-02 6.05257332e-01 1.96365103e-01 5.31027913e-02 5.53580344e-01 1.38042718e-01 -1.41212654e+00 1.00200450e+00 3.32750291e-01 6.77444398e-01 -3.93775970e-01 8.94289315e-01 2.92863846e-01 -9.71202195e-01 -6.39055520e-02 1.63552091e-01 -3.45867872e-01 5.32451987e-01 3.42871100e-01 -8.77075970e-01 8.08082461e-01 5.63357055e-01 1.83027446e-01 -7.92217553e-01 6.45148873e-01 -7.17849076e-01 9.84969139e-01 -1.98788017e-01 -4.27765101e-02 4.38886076e-01 2.23717213e-01 6.53349519e-01 1.25304568e+00 2.51880080e-01 1.63287729e-01 -3.81091014e-02 8.35533261e-01 -9.07112882e-02 1.13717541e-02 -3.50696713e-01 1.75519347e-01 6.14668846e-01 1.11866379e+00 -2.99428791e-01 -2.82447070e-01 -2.10949510e-01 7.40537465e-01 7.32182324e-01 2.11146355e-01 -8.36582422e-01 -1.10599399e-01 3.26560616e-01 3.22365314e-02 3.52349848e-01 -1.84419811e-01 -7.37483501e-01 -9.45739746e-01 2.26199776e-01 -1.19989431e+00 5.42841136e-01 -8.84773433e-01 -1.22773302e+00 7.80368745e-01 -5.92491440e-02 -8.77937019e-01 -5.78353167e-01 -5.87664545e-01 -3.33884597e-01 1.03458083e+00 -1.87412059e+00 -1.38668132e+00 -1.95835546e-01 3.59831005e-01 6.52280807e-01 -3.65785733e-02 1.11994040e+00 9.14590508e-02 -1.71984822e-01 6.93779528e-01 -2.91077793e-01 -4.56120342e-01 5.97325504e-01 -1.64745653e+00 6.79856360e-01 8.94599497e-01 2.87823826e-01 8.79204690e-01 1.04522204e+00 -4.93885845e-01 -1.38112414e+00 -4.21987295e-01 1.68697202e+00 -1.21421528e+00 6.42031133e-01 -2.04350084e-01 -8.79674733e-01 7.85430610e-01 5.52661002e-01 -4.48198378e-01 5.02632439e-01 5.88692129e-01 -5.66534162e-01 3.49193931e-01 -1.09394896e+00 6.36174560e-01 1.06314003e+00 -5.83814561e-01 -1.08940458e+00 3.42038929e-01 9.49828446e-01 -7.62692153e-01 -8.44617188e-01 6.34388328e-01 5.34866869e-01 -1.07427049e+00 8.38210046e-01 -7.27374256e-01 8.08384538e-01 -3.14944655e-01 -2.19901800e-01 -1.27002740e+00 -2.14654535e-01 -7.95106888e-01 -2.46358827e-01 1.08232033e+00 1.00535977e+00 -6.39476597e-01 6.88066781e-01 6.09993279e-01 -6.80796623e-01 -1.01195538e+00 -1.33109653e+00 -4.80641752e-01 2.95420498e-01 -4.76269662e-01 5.81003845e-01 6.50560677e-01 2.48768911e-01 7.52460659e-01 -7.07057640e-02 1.19256534e-01 5.79058588e-01 1.29114836e-02 6.98937118e-01 -1.07510269e+00 -2.41954938e-01 -3.95706564e-01 9.42352116e-02 -9.64806437e-01 1.45183191e-01 -1.00739145e+00 2.80302316e-01 -1.92305171e+00 4.33449969e-02 -3.69865000e-01 -1.08131938e-01 4.15450037e-01 -5.62990367e-01 2.24681377e-01 4.37716037e-01 1.78377479e-01 -7.34169424e-01 3.73327047e-01 1.13463259e+00 2.11514890e-01 2.16010779e-01 -4.00288433e-01 -9.66725826e-01 5.33323646e-01 7.07480550e-01 -2.99866349e-01 -1.94501787e-01 -9.84129131e-01 1.73271298e-01 3.20710875e-02 3.13479483e-01 -9.17686641e-01 2.20398515e-01 2.71330797e-03 -1.17923573e-01 -3.95450532e-01 6.00236177e-01 -3.52332324e-01 -8.32544938e-02 5.01128793e-01 -5.80434680e-01 4.22560155e-01 2.84438670e-01 4.21778649e-01 -2.88844466e-01 -6.52185380e-02 2.61925697e-01 -3.56618986e-02 -6.43675506e-01 -2.84365743e-01 -3.80064733e-02 4.99651432e-01 6.02743149e-01 -2.16569498e-01 -8.77020717e-01 -4.64879185e-01 -5.16006529e-01 3.00906926e-01 2.53655463e-02 4.59864408e-01 3.09160322e-01 -9.95720327e-01 -1.04544067e+00 -2.61803210e-01 2.73883671e-01 -1.45907300e-02 7.29648322e-02 9.98983979e-01 -1.91472918e-01 9.53593493e-01 -1.67880394e-02 -4.92334425e-01 -1.02752674e+00 -3.27109545e-02 1.02515079e-01 -8.10457349e-01 -4.16607320e-01 1.01784015e+00 -1.90412551e-01 -6.25287950e-01 -9.61495414e-02 -2.47304708e-01 -2.45386362e-01 6.42382130e-02 2.78354466e-01 3.96647215e-01 3.87895793e-01 -5.62865436e-01 -5.08297741e-01 2.86387831e-01 -7.13534504e-02 -5.07714868e-01 1.11423707e+00 -1.09048501e-01 2.19530296e-02 4.27449375e-01 5.74788868e-01 3.01355422e-01 -8.77288640e-01 -1.55542716e-01 4.82710123e-01 -2.42645010e-01 -3.01750094e-01 -1.66256213e+00 -1.66512951e-01 9.90334213e-01 2.23099560e-01 3.28881145e-01 5.41988432e-01 1.56580925e-01 1.02091467e+00 2.65488446e-01 5.72570562e-01 -8.95605445e-01 -6.29767179e-02 8.53763819e-01 1.07508421e+00 -1.31031013e+00 -1.47421777e-01 -5.44691980e-01 -6.76218987e-01 8.40193033e-01 7.48784542e-01 2.02908546e-01 -1.16468057e-01 -2.41913319e-01 1.74723461e-01 -4.61814284e-01 -1.14871669e+00 -4.62951809e-01 5.98990500e-01 4.48602259e-01 7.56657362e-01 1.43807009e-01 -6.47493482e-01 7.52281249e-01 -7.93667436e-01 -2.95281976e-01 3.71683240e-01 6.90488696e-01 -3.41479152e-01 -1.13130498e+00 -2.28982702e-01 4.72042918e-01 -4.22577292e-01 -6.06710970e-01 -5.44132471e-01 1.05997956e+00 -6.74560815e-02 1.39216220e+00 -1.61622614e-01 -1.71214253e-01 5.44375181e-01 5.19472420e-01 5.33268571e-01 -7.24632442e-01 -1.12456894e+00 -5.76088011e-01 7.04296887e-01 -6.69772744e-01 -3.20373535e-01 -7.74881423e-01 -1.42689455e+00 -9.88845155e-02 -2.93457508e-01 4.74393010e-01 8.19278598e-01 1.36843431e+00 6.06480598e-01 7.56887436e-01 -6.64804950e-02 -6.71336412e-01 -8.41315389e-01 -1.24282074e+00 2.37164155e-01 5.68198502e-01 3.46593887e-01 -7.25810051e-01 -5.75835109e-01 -1.93270758e-01]	[11.008758544921875, 8.331466674804688]
164febca-7b5d-4622-86aa-637073d8c667	retrieving-supporting-evidence-for-llms	2306.13781	null	https://arxiv.org/abs/2306.13781v1	https://arxiv.org/pdf/2306.13781v1.pdf	Retrieving Supporting Evidence for LLMs Generated Answers	Current large language models (LLMs) can exhibit near-human levels of performance on many natural language tasks, including open-domain question answering. Unfortunately, they also convincingly hallucinate incorrect answers, so that responses to questions must be verified against external sources before they can be accepted at face value. In this paper, we report a simple experiment to automatically verify generated answers against a corpus. After presenting a question to an LLM and receiving a generated answer, we query the corpus with the combination of the question + generated answer. We then present the LLM with the combination of the question + generated answer + retrieved answer, prompting it to indicate if the generated answer can be supported by the retrieved answer. We base our experiment on questions and passages from the MS MARCO (V1) test collection, exploring three retrieval approaches ranging from standard BM25 to a full question answering stack, including a reader based on the LLM. For a large fraction of questions, we find that an LLM is capable of verifying its generated answer if appropriate supporting material is provided. However, with an accuracy of 70-80%, this approach cannot be fully relied upon to detect hallucinations.	['Charles L. A. Clarke', 'Negar Arabzadeh', 'Siqing Huo']	2023-06-23	null	null	null	null	['retrieval', 'question-answering', 'open-domain-question-answering']	['methodology', 'natural-language-processing', 'natural-language-processing']	[ 1.26567528e-01 4.59716678e-01 2.02221215e-01 -8.19481835e-02 -1.80055785e+00 -9.66174364e-01 7.15690792e-01 6.50254786e-01 -5.07992089e-01 7.71160424e-01 3.09903711e-01 -6.93395257e-01 -1.27777800e-01 -5.25717378e-01 -3.84905607e-01 3.35494839e-02 5.41140735e-01 6.46403551e-01 7.76023686e-01 -4.25129175e-01 7.61287391e-01 1.63891047e-01 -1.14260924e+00 7.14952886e-01 9.80003476e-01 8.31262171e-01 1.93807557e-01 9.81013417e-01 -5.58992028e-01 1.53048742e+00 -1.16334283e+00 -7.60259926e-01 -2.89610267e-01 -4.85142976e-01 -1.49376535e+00 -2.64839530e-01 5.34786940e-01 -4.96028304e-01 -7.41851255e-02 1.02332032e+00 5.70851445e-01 1.39993712e-01 6.37079060e-01 -1.26269555e+00 -8.03275108e-01 3.19386214e-01 1.81366876e-01 4.44709361e-01 1.25963926e+00 8.13955218e-02 1.04957628e+00 -1.17007303e+00 7.87068367e-01 1.30340803e+00 2.75532812e-01 5.50333977e-01 -9.02989507e-01 -4.55615640e-01 -2.37103984e-01 3.91591042e-01 -1.10458302e+00 -5.28756142e-01 4.06019777e-01 -1.97271913e-01 1.14742064e+00 4.57493335e-01 -1.43235311e-01 8.52896869e-01 1.50623426e-01 6.00869894e-01 1.05166805e+00 -5.44705212e-01 4.47134614e-01 5.70767939e-01 3.54870647e-01 4.81874615e-01 -9.26878378e-02 -5.34798324e-01 -7.71777749e-01 -7.72266209e-01 1.90516841e-02 -8.24345171e-01 -4.78371859e-01 3.45502704e-01 -1.01189494e+00 8.59983504e-01 3.63729507e-01 2.43928120e-01 -5.92981577e-01 -3.29332292e-01 1.72116816e-01 7.55704999e-01 1.70815647e-01 9.59731996e-01 -2.32432216e-01 -5.58592379e-02 -8.40060413e-01 6.69371247e-01 1.39088321e+00 7.02595592e-01 5.52307665e-01 -5.56668460e-01 -5.02565026e-01 1.03890038e+00 2.37577677e-01 6.43590212e-01 8.27232480e-01 -1.27282381e+00 4.85400081e-01 6.86002493e-01 6.48015320e-01 -1.30750453e+00 -8.42474252e-02 -1.66210696e-01 -9.28407162e-02 -1.14827514e-01 4.66551065e-01 -8.99802707e-03 -1.03566207e-01 1.50105011e+00 3.19379926e-01 -3.36326241e-01 3.38858694e-01 6.85165346e-01 1.30899179e+00 6.69300020e-01 -3.04150842e-02 -2.28915066e-01 1.47946751e+00 -7.38423645e-01 -8.22249293e-01 -4.32726443e-01 8.00152421e-01 -1.27597821e+00 1.15783763e+00 4.64430869e-01 -1.22987044e+00 -5.08054852e-01 -9.78047252e-01 -2.65624672e-01 -2.00408682e-01 -2.25191250e-01 -2.69271851e-01 2.06488892e-01 -1.18099117e+00 2.28379920e-01 6.00824393e-02 -6.34068608e-01 -2.23223269e-01 -1.22680791e-01 -2.00872719e-01 -3.71710032e-01 -1.52362931e+00 1.31419528e+00 -4.67065908e-02 -2.48776779e-01 -6.25208855e-01 -2.70650446e-01 -5.98050416e-01 -8.72895345e-02 3.19728822e-01 -6.37231827e-01 1.96658826e+00 -4.10800576e-01 -9.03632998e-01 9.73097801e-01 -5.80934465e-01 -4.59585905e-01 4.60009396e-01 -2.38141179e-01 -7.89611280e-01 9.19654369e-01 5.65313816e-01 6.48215830e-01 8.18235397e-01 -1.22911596e+00 -3.87981176e-01 -1.17335074e-01 3.89397532e-01 3.30559537e-02 -2.22565621e-01 4.81645852e-01 -1.93939403e-01 -2.68721521e-01 1.67028233e-01 -6.31613612e-01 1.18831262e-01 -1.07593656e-01 -3.37065876e-01 -6.48963094e-01 6.25143766e-01 -9.99070168e-01 1.48749280e+00 -1.90546775e+00 -6.16342843e-01 4.40298468e-02 4.44452465e-01 5.38610935e-01 -5.70817769e-01 9.53844786e-01 2.11255655e-01 3.44183952e-01 -2.25371309e-02 1.27873376e-01 2.14546591e-01 -6.26628026e-02 -8.10586631e-01 -1.14945797e-02 3.21946591e-01 1.07465827e+00 -1.07114887e+00 -6.31551147e-01 -2.98355490e-01 2.47200411e-02 -4.12354678e-01 4.26869601e-01 -4.51286286e-01 -5.21403085e-03 -4.84665364e-01 4.20758665e-01 2.81899810e-01 -6.21038735e-01 -3.48745435e-01 2.15950981e-01 3.93376708e-01 7.45160043e-01 -1.03049171e+00 9.20759082e-01 -4.54634696e-01 7.49033451e-01 1.41153276e-01 -3.82150918e-01 1.14074016e+00 6.70659065e-01 -3.37159216e-01 -8.99164200e-01 -2.54149765e-01 6.51858807e-01 -1.24736629e-01 -8.71877968e-01 7.00029671e-01 -1.57576978e-01 3.22028138e-02 8.38099658e-01 -5.27637191e-02 -5.13149738e-01 4.11338508e-01 8.72152925e-01 1.43320060e+00 -4.64981645e-01 2.89438874e-01 -1.18297208e-02 1.03243947e+00 3.08122873e-01 -2.35453725e-01 1.33842027e+00 -2.34752819e-01 4.65059161e-01 4.74086583e-01 1.58501372e-01 -6.34534955e-01 -9.91003811e-01 7.87121430e-02 1.05539465e+00 1.01654055e-02 -5.27655780e-01 -7.39267170e-01 -4.56632853e-01 -7.16998428e-02 1.13658237e+00 -2.17749029e-01 -2.07877755e-01 -2.01451346e-01 1.50572613e-01 7.83539474e-01 1.71442568e-01 4.10639673e-01 -1.35422587e+00 -6.31409228e-01 3.49588931e-01 -9.11033273e-01 -1.08632874e+00 -2.75171310e-01 -2.00860500e-01 -4.49143916e-01 -1.24498606e+00 -7.65218973e-01 -8.04398477e-01 4.10384983e-01 2.96699822e-01 1.37308311e+00 6.43688381e-01 2.86758691e-01 8.47784340e-01 -5.16840696e-01 -2.44945124e-01 -1.07065153e+00 -2.64080375e-01 -2.56352544e-01 -2.34464735e-01 6.72366321e-01 -2.17966825e-01 -4.84801352e-01 4.34729159e-01 -1.13279212e+00 -4.26688612e-01 3.25013191e-01 3.97029757e-01 1.09127685e-01 -4.99613643e-01 1.33794641e+00 -6.23020113e-01 1.49579072e+00 -6.96428776e-01 -1.02419451e-01 4.85091954e-01 -3.40635985e-01 -5.01548350e-02 4.05745178e-01 -6.20268583e-01 -7.79808342e-01 -7.96576858e-01 -4.98937160e-01 -1.47442082e-02 -1.23789132e-01 7.32704341e-01 2.17627093e-01 3.82102095e-02 1.25006747e+00 5.51759265e-02 -1.42376631e-01 -4.06457692e-01 3.11353624e-01 1.02789235e+00 8.32520962e-01 -4.56736237e-01 8.94935548e-01 -1.12430379e-02 -6.02805257e-01 -6.98018968e-01 -1.05463898e+00 -7.50769436e-01 -8.38056803e-02 -4.28221792e-01 5.25110602e-01 -6.42249227e-01 -4.66721922e-01 -1.86483003e-02 -1.58704174e+00 -4.77159731e-02 -7.60589391e-02 3.39667678e-01 -5.30855000e-01 5.39652705e-01 -6.71393156e-01 -9.43174601e-01 -3.68284583e-01 -7.85336733e-01 9.39470410e-01 4.11343575e-03 -1.12063396e+00 -7.59456217e-01 1.38345763e-01 9.35061812e-01 6.69354022e-01 -2.55050927e-01 1.41050971e+00 -1.38682353e+00 -1.82310313e-01 -6.42781794e-01 -2.45375838e-02 2.03790411e-01 -2.48371407e-01 -3.74566555e-01 -1.03166962e+00 -1.01055719e-01 4.30366427e-01 -9.98677909e-01 4.46686864e-01 -3.71721774e-01 6.06118023e-01 -4.87884969e-01 1.43713638e-01 -4.96644288e-01 9.74267423e-01 -3.61796916e-02 5.83944917e-01 3.95709842e-01 -1.90898240e-01 8.72630417e-01 4.04511839e-01 2.31496319e-01 4.00503397e-01 6.27851486e-02 1.20528184e-01 4.60528642e-01 -9.50612351e-02 -5.27728081e-01 4.41602051e-01 7.96888173e-01 8.44846189e-01 -5.18827260e-01 -7.58163631e-01 6.51672125e-01 -1.46466029e+00 -9.21368957e-01 -4.90954190e-01 2.15773726e+00 1.01548553e+00 2.03860015e-01 -2.84568489e-01 2.20538959e-01 5.59148550e-01 4.23462577e-02 -6.19653344e-01 -5.91839254e-01 -2.17304692e-01 3.60753715e-01 -3.71392161e-01 9.07482684e-01 -2.83752620e-01 6.68562651e-01 6.54270887e+00 7.91211247e-01 -6.72718644e-01 9.39668044e-02 4.23197210e-01 2.47717083e-01 -7.85268247e-01 1.72180414e-01 -5.68530858e-01 9.24160853e-02 1.18299794e+00 -6.62549973e-01 1.43341511e-01 5.86100638e-01 1.49184316e-01 -5.43134153e-01 -1.08077025e+00 7.26393819e-01 4.91997868e-01 -1.08147085e+00 3.07201296e-01 -4.76506323e-01 3.79167885e-01 -1.43146291e-01 -2.83926755e-01 6.54688239e-01 7.55358413e-02 -1.02896035e+00 6.66534126e-01 4.72720861e-01 4.64100838e-01 -3.45537633e-01 7.66578734e-01 9.87573087e-01 -6.17789507e-01 9.44276005e-02 -3.26277971e-01 -2.84741074e-01 2.59849161e-01 3.29583734e-01 -9.72000182e-01 1.19193308e-01 3.49362969e-01 -2.81484842e-01 -1.07786965e+00 1.04839790e+00 -7.14333653e-01 7.07853138e-01 -1.65112644e-01 -5.32969594e-01 1.45688847e-01 3.89318883e-01 5.19852459e-01 9.00467217e-01 2.83116937e-01 2.92185366e-01 -3.14120669e-03 8.58645499e-01 -3.01420927e-01 4.62977976e-01 -5.57489693e-01 -2.09872544e-01 6.30054176e-01 1.06695163e+00 -2.31136829e-01 -7.61700511e-01 -3.46774191e-01 8.93429279e-01 3.69734764e-01 4.89929080e-01 -2.48040900e-01 -6.05362773e-01 -2.10496828e-01 1.06624417e-01 -1.57856271e-01 2.23285347e-01 1.79285541e-01 -9.45929825e-01 3.78142476e-01 -1.50558233e+00 4.93080348e-01 -1.67621922e+00 -1.43457210e+00 8.48057210e-01 -2.39330441e-01 -8.11794579e-01 -7.36987352e-01 -3.75787616e-01 -6.03660941e-01 1.23151505e+00 -1.41124034e+00 -4.31404710e-01 -1.57616109e-01 6.49866581e-01 4.83001083e-01 1.74402148e-01 9.57102776e-01 1.43648475e-01 1.71443567e-01 1.98303506e-01 -3.59298766e-01 9.96659398e-02 1.19335723e+00 -9.84643698e-01 1.18657701e-01 7.22726464e-01 2.24737898e-01 9.26202595e-01 9.80078280e-01 -8.51394236e-01 -1.13899040e+00 -6.76278830e-01 1.91367185e+00 -8.47540975e-01 9.85609174e-01 6.44036233e-02 -1.61885333e+00 3.69442850e-01 5.56721032e-01 -3.81324917e-01 8.28400671e-01 -3.07787091e-01 -5.27211070e-01 3.86559308e-01 -1.25555623e+00 7.12029099e-01 2.57589847e-01 -1.15976453e+00 -1.56040835e+00 7.92275429e-01 8.76555622e-01 -2.32735043e-03 -4.59666491e-01 2.14080006e-01 7.85953104e-02 -7.46677399e-01 5.59713602e-01 -8.98488045e-01 5.83750725e-01 -3.78973216e-01 -2.99863428e-01 -1.04045451e+00 2.53283046e-02 -6.15935266e-01 -1.14504524e-01 1.22449136e+00 6.38636887e-01 -5.32596111e-01 1.90787181e-01 1.07420087e+00 2.59370446e-01 -4.54442680e-01 -8.20674419e-01 -5.58809161e-01 2.89733440e-01 -7.15366244e-01 2.75750190e-01 6.37351692e-01 3.94867957e-01 6.43626571e-01 4.36096527e-02 1.76032603e-01 6.15023673e-02 3.26701999e-02 8.15580785e-01 -9.84525681e-01 -2.22501561e-01 -2.96273530e-01 1.82999700e-01 -1.29546380e+00 7.74522796e-02 -9.27324474e-01 1.14538632e-01 -1.76881993e+00 2.07042009e-01 3.24532986e-01 1.30718783e-01 8.82387534e-02 -4.73072082e-01 2.96010017e-01 1.73085615e-01 2.73113191e-01 -8.94900382e-01 1.87434629e-01 1.03984761e+00 -7.61978179e-02 1.16109140e-01 -1.56262890e-01 -1.04851580e+00 9.02294695e-01 6.74077630e-01 -5.42376161e-01 -3.17471415e-01 -1.62444293e-01 6.54017985e-01 5.67630231e-01 6.16734087e-01 -9.73225653e-01 7.34917223e-01 7.04518557e-02 1.64205030e-01 -6.48920774e-01 2.02972248e-01 -1.97472513e-01 -3.51605028e-01 4.94205952e-01 -9.93889034e-01 4.61814344e-01 1.57114863e-01 2.91921616e-01 -4.86770898e-01 -9.36176121e-01 4.09524828e-01 -3.05122584e-01 -5.01562715e-01 -2.51861304e-01 -8.53063762e-01 5.78600168e-01 3.71354550e-01 1.47935018e-01 -7.23167777e-01 -1.15110695e+00 -4.87008154e-01 5.67857802e-01 2.63652593e-01 4.18876201e-01 9.07095909e-01 -1.13940835e+00 -9.24663961e-01 -2.07594037e-01 4.93773967e-01 -6.80895567e-01 4.43479009e-02 6.46740317e-01 -2.66295522e-01 7.26124406e-01 4.85170692e-01 -2.36578137e-01 -1.10254645e+00 6.14389718e-01 1.18571348e-01 -2.44674817e-01 -1.36141211e-01 7.84479141e-01 -6.00246675e-02 -5.51133275e-01 1.88515171e-01 -1.35149360e-01 -4.09200490e-01 1.51539698e-01 1.17968416e+00 1.63922131e-01 2.83722401e-01 -6.02984428e-01 -2.95662105e-01 1.57870740e-01 1.41043559e-01 -7.50887334e-01 6.54561818e-01 -2.42886394e-01 -2.76801884e-01 5.42456985e-01 1.22338200e+00 4.19444233e-01 -2.59712845e-01 -5.11797786e-01 2.61463493e-01 -4.00590539e-01 -3.78726006e-01 -1.17170644e+00 -9.45292935e-02 7.57149994e-01 -1.04731405e-02 5.57976902e-01 7.71381557e-01 3.85278493e-01 1.06641245e+00 9.89764214e-01 3.14708620e-01 -9.39175069e-01 5.82706034e-01 6.81719124e-01 1.34296262e+00 -1.35317373e+00 -4.29323018e-01 4.21496555e-02 -6.00841343e-01 1.07758665e+00 4.55509901e-01 1.27160951e-01 2.38561332e-01 -2.15853676e-01 6.00278437e-01 -3.67991686e-01 -1.16783822e+00 1.67859327e-02 3.45314503e-01 3.35932016e-01 3.87237728e-01 -3.71339828e-01 -5.37320793e-01 5.53582251e-01 -5.23841441e-01 -9.45473015e-02 6.64320827e-01 8.42914402e-01 -1.00617278e+00 -8.54481101e-01 -7.66212165e-01 5.36985517e-01 -6.22038245e-01 -3.76384348e-01 -9.57324505e-01 3.46691221e-01 -4.83335972e-01 1.99665332e+00 -2.50330418e-01 -6.83364347e-02 3.47084492e-01 5.84838271e-01 4.99644037e-03 -7.98897326e-01 -7.34436870e-01 -2.66229808e-01 4.52757716e-01 -4.35995132e-01 -2.28426129e-01 -3.30131829e-01 -1.23286426e+00 -1.00565046e-01 -3.70281130e-01 7.05284536e-01 3.93700361e-01 1.30003023e+00 2.69080549e-01 -1.42945677e-01 2.92279392e-01 1.11889221e-01 -1.09305382e+00 -1.09519303e+00 -1.47194445e-01 6.30400300e-01 4.48013693e-01 -9.00277495e-02 -7.49474645e-01 -1.57361493e-01]	[11.433693885803223, 8.069167137145996]
20529bfb-e492-402b-8da2-806adb6664cf	learning-universal-shape-dictionary-for	2012.01050	null	https://arxiv.org/abs/2012.01050v1	https://arxiv.org/pdf/2012.01050v1.pdf	Learning Universal Shape Dictionary for Realtime Instance Segmentation	We present a novel explicit shape representation for instance segmentation. Based on how to model the object shape, current instance segmentation systems can be divided into two categories, implicit and explicit models. The implicit methods, which represent the object mask/contour by intractable network parameters, and produce it through pixel-wise classification, are predominant. However, the explicit methods, which parameterize the shape with simple and explainable models, are less explored. Since the operations to generate the final shape are light-weighted, the explicit methods have a clear speed advantage over implicit methods, which is crucial for real-world applications. The proposed USD-Seg adopts a linear model, sparse coding with dictionary, for object shapes. First, it learns a dictionary from a large collection of shape datasets, making any shape being able to be decomposed into a linear combination through the dictionary. Hence the name "Universal Shape Dictionary". Then it adds a simple shape vector regression head to ordinary object detector, giving the detector segmentation ability with minimal overhead. For quantitative evaluation, we use both average precision (AP) and the proposed Efficiency of AP (AP$_E$) metric, which intends to also measure the computational consumption of the framework to cater to the requirements of real-world applications. We report experimental results on the challenging COCO dataset, in which our single model on a single Titan Xp GPU achieves 35.8 AP and 27.8 AP$_E$ at 65 fps with YOLOv4 as base detector, 34.1 AP and 28.6 AP$_E$ at 12 fps with FCOS as base detector.	['Cewu Lu', 'Lixin Yang', 'Ruolin Ye', 'Wenqiang Xu', 'Tutian Tang']	2020-12-02	null	null	null	null	['explainable-models']	['computer-vision']	[ 1.55096158e-01 1.82783097e-01 -2.80654393e-02 -2.55127192e-01 -5.81480265e-01 -5.90173185e-01 2.69447148e-01 -7.01283738e-02 -4.19867426e-01 3.34247023e-01 -5.71376026e-01 -9.60308835e-02 1.15807898e-01 -8.35034430e-01 -7.96054959e-01 -7.77532995e-01 2.30885729e-01 6.37309253e-01 4.86524642e-01 4.43040580e-02 2.68755227e-01 7.86192477e-01 -1.43854189e+00 8.08459073e-02 8.51825297e-01 1.42347157e+00 3.14967960e-01 3.59838307e-01 -3.66840750e-01 3.52894694e-01 -2.48175725e-01 -4.34688807e-01 5.52582145e-01 2.64435202e-01 -6.50009632e-01 2.98922777e-01 3.95604134e-01 -2.62035429e-01 -1.85868666e-01 1.16114867e+00 4.53439265e-01 -7.28368536e-02 6.89197123e-01 -1.08846390e+00 -4.62235242e-01 3.34160984e-01 -7.85153985e-01 -3.06784287e-02 -4.28756438e-02 4.83058363e-01 8.61930013e-01 -1.11115193e+00 6.38942957e-01 1.17438436e+00 6.45297885e-01 3.16230327e-01 -1.30751264e+00 -6.01486325e-01 2.25697935e-01 -1.75967604e-01 -1.54999495e+00 -3.89952332e-01 6.24043882e-01 -4.83042210e-01 7.12659121e-01 2.37256303e-01 5.79939842e-01 3.54001820e-01 -4.79597524e-02 9.48245466e-01 9.52980876e-01 -1.36307091e-01 2.49160349e-01 2.45401546e-01 2.38178298e-01 8.99830282e-01 4.02222335e-01 -7.48234987e-02 -1.33098468e-01 -5.95530458e-02 1.12902510e+00 -1.96770001e-02 -3.10684055e-01 -4.83487546e-01 -8.69172752e-01 7.57149279e-01 7.91763306e-01 1.47660375e-01 -3.08493078e-01 2.62147188e-01 1.39259771e-01 -1.73108205e-01 1.96427613e-01 1.29337355e-01 -3.55052322e-01 1.42614156e-01 -1.06707513e+00 8.78560916e-02 7.02135265e-01 1.19952822e+00 8.89862239e-01 3.69192630e-01 -4.16393764e-02 8.49467695e-01 5.52356839e-01 8.01280737e-01 2.37807944e-01 -8.78518462e-01 2.79836416e-01 8.13438416e-01 -1.89999919e-02 -1.19764960e+00 -3.03846568e-01 -4.81087238e-01 -9.24219131e-01 2.17772841e-01 3.69694024e-01 -6.88321143e-02 -1.21621108e+00 1.35285437e+00 4.09421206e-01 2.97738284e-01 -2.88343370e-01 9.78730023e-01 1.12622321e+00 8.49637806e-01 4.41753715e-02 -2.68668294e-01 1.42244601e+00 -1.05570626e+00 -2.99069345e-01 -1.59873515e-01 4.93557066e-01 -7.84090459e-01 9.15546775e-01 5.85427999e-01 -1.18826365e+00 -6.05844438e-01 -8.55789483e-01 -1.57737479e-01 -1.83111548e-01 4.71782774e-01 7.31774330e-01 4.99124497e-01 -1.06074238e+00 4.49385643e-01 -8.50466788e-01 6.80735111e-02 5.42780399e-01 6.39284790e-01 -2.86644250e-02 5.71345277e-02 -5.88657498e-01 3.43319267e-01 3.89220834e-01 3.07189494e-01 -7.08226204e-01 -7.31860697e-01 -7.33237565e-01 1.82109326e-01 5.62211573e-01 -5.64008951e-01 8.59622717e-01 -9.74769056e-01 -1.50228095e+00 9.90049362e-01 -1.64869037e-02 -4.90515441e-01 3.96209985e-01 -3.17155682e-02 -5.71690965e-04 2.06290603e-01 -3.61876264e-02 9.80093479e-01 1.08457971e+00 -1.40578473e+00 -5.65046966e-01 -2.73772776e-01 3.05239204e-02 1.04256712e-01 8.23702663e-02 -2.12983489e-01 -9.64229226e-01 -6.70292199e-01 4.81583059e-01 -1.04846418e+00 -6.07791305e-01 3.56854200e-01 -3.12039047e-01 -1.63045272e-01 8.58058989e-01 -5.09371579e-01 1.04405153e+00 -2.31584072e+00 -4.81393561e-02 4.04092282e-01 2.23242670e-01 4.45738435e-01 3.09956521e-02 -1.54567137e-01 1.34350911e-01 -3.22038047e-02 -7.60677397e-01 -4.92178261e-01 -7.38305077e-02 3.34160268e-01 -3.36555570e-01 3.86929393e-01 2.07001016e-01 8.44367862e-01 -4.34300125e-01 -5.66575289e-01 8.11702237e-02 5.07976472e-01 -6.83502197e-01 1.71171322e-01 -1.78823248e-01 6.83049560e-02 -5.84005117e-01 9.19699013e-01 1.07821620e+00 -2.46542424e-01 5.79246469e-02 -4.10917908e-01 -3.15717012e-01 -3.24668825e-01 -1.52139091e+00 1.64974225e+00 -2.17981502e-01 3.15565646e-01 5.19179046e-01 -1.02686203e+00 1.19385254e+00 2.26629600e-02 5.69593430e-01 -5.21559417e-01 3.90077084e-01 4.50742096e-01 -1.63842991e-01 -9.81211290e-02 3.41048747e-01 1.16908818e-01 1.21238768e-01 5.95694073e-02 -3.06597557e-02 -4.25495207e-01 1.86058760e-01 1.48294941e-01 7.40047455e-01 2.34896913e-01 1.87308490e-01 -5.19335926e-01 5.99106848e-01 1.05002336e-01 6.68608129e-01 4.57122892e-01 -3.83892171e-02 8.22705865e-01 3.82858127e-01 -5.90075016e-01 -7.46728659e-01 -9.75195825e-01 -3.79222602e-01 6.16885722e-01 5.04250586e-01 1.15830265e-02 -9.34916377e-01 -5.08966208e-01 -3.50586437e-02 3.26802164e-01 -2.52380878e-01 2.49949813e-01 -7.98731983e-01 -8.07400942e-01 1.61676437e-01 5.14320850e-01 5.12132525e-01 -1.36343062e+00 -7.71959066e-01 2.16548190e-01 1.48463607e-01 -1.25492632e+00 -5.63351631e-01 8.34047198e-02 -1.13710475e+00 -1.01662135e+00 -7.86051929e-01 -7.93684781e-01 8.92679274e-01 1.45542964e-01 9.63828564e-01 2.84418911e-01 -5.96102417e-01 2.76478469e-01 -2.30021864e-01 -1.73201278e-01 1.86109230e-01 3.43702622e-02 -1.39291286e-01 4.19684619e-01 -5.05241528e-02 -4.79344130e-01 -7.83618331e-01 2.73302406e-01 -9.49110687e-01 5.42976633e-02 8.71731579e-01 6.34148955e-01 1.25935531e+00 -3.74151468e-01 2.54596382e-01 -9.07333553e-01 7.50285685e-02 -2.86348879e-01 -9.21170473e-01 2.90931780e-02 -4.72827017e-01 5.20347094e-04 5.20057976e-01 -4.77566659e-01 -9.22610819e-01 6.46925271e-01 -2.43814066e-01 -6.08104169e-01 -5.15399389e-02 2.28972971e-01 -2.22430289e-01 -3.43003154e-01 3.69349211e-01 4.96776730e-01 -1.30239353e-01 -5.74296713e-01 4.83795583e-01 4.40914571e-01 6.30158603e-01 -8.34398448e-01 7.64605343e-01 5.71751237e-01 4.24510576e-02 -9.82231259e-01 -4.99316841e-01 -5.72199047e-01 -5.35932958e-01 -4.36944179e-02 8.58108342e-01 -7.72612095e-01 -8.87105763e-01 3.33355188e-01 -1.15373826e+00 -3.04570407e-01 -4.21538919e-01 1.31184489e-01 -7.25331545e-01 4.87109333e-01 -4.46351051e-01 -7.04359710e-01 -6.86941862e-01 -1.48678362e+00 1.25276470e+00 4.46785927e-01 1.32278517e-01 -5.02271414e-01 -3.40460092e-01 2.49098361e-01 2.10706025e-01 2.06735015e-01 7.03868866e-01 -2.95368880e-01 -1.00074935e+00 1.43587003e-02 -7.47566521e-01 3.05611879e-01 -2.61857361e-01 -3.33483219e-02 -8.02449048e-01 -3.48289579e-01 8.43094960e-02 -1.08440742e-01 6.69491589e-01 5.69601834e-01 1.47231734e+00 -2.38028586e-01 -4.19816434e-01 1.02609503e+00 1.83429110e+00 3.66563529e-01 6.90669119e-01 -1.16489073e-02 7.36592412e-01 1.98349446e-01 6.25244021e-01 4.69390601e-01 2.24600405e-01 7.81033278e-01 5.46758771e-01 -3.19503456e-01 -2.58235365e-01 4.21793945e-02 1.71149001e-01 9.01599944e-01 -2.64575750e-01 -9.15854201e-02 -8.45721006e-01 4.98321533e-01 -1.69328892e+00 -4.89945441e-01 -3.88641328e-01 2.15597034e+00 6.41644239e-01 2.63212323e-01 7.41375089e-02 -8.12352151e-02 5.55271685e-01 -4.95895334e-02 -6.06722236e-01 -4.27768081e-01 6.77226037e-02 5.38946271e-01 5.98691165e-01 3.62652630e-01 -1.01035166e+00 1.07323980e+00 5.67353106e+00 1.27812803e+00 -1.47719920e+00 6.33129105e-02 8.42509151e-01 1.40638173e-01 -7.52960443e-02 2.77495421e-02 -9.89031851e-01 4.40669447e-01 4.37880903e-01 1.08653173e-01 2.60220200e-01 1.06043220e+00 1.73624437e-02 -5.85160181e-02 -9.28680122e-01 1.43644381e+00 -2.32091308e-01 -1.23922598e+00 1.06690668e-01 -1.43812913e-02 6.82597041e-01 -3.10606435e-02 8.30301940e-02 2.30981022e-01 -1.61089227e-01 -9.35654819e-01 9.15701807e-01 5.05076468e-01 1.02207255e+00 -5.56489706e-01 5.75367630e-01 5.83092749e-01 -1.51787746e+00 1.42119110e-01 -6.50788128e-01 2.97089607e-01 2.29837745e-01 5.33766031e-01 -4.59687322e-01 4.28049207e-01 6.08935237e-01 4.21709508e-01 -3.50220114e-01 1.11007726e+00 1.77637100e-01 4.17633057e-01 -7.47349203e-01 1.49272755e-01 3.73170882e-01 -5.40314376e-01 4.78619337e-01 1.31795788e+00 4.29911375e-01 6.42802000e-01 4.48370486e-01 1.05073893e+00 -8.58892202e-02 3.84377360e-01 -2.22622052e-01 1.79966837e-01 2.43236676e-01 1.49830854e+00 -1.20130014e+00 -4.46121693e-01 -1.73528627e-01 7.21503854e-01 2.12670222e-01 1.86738610e-01 -1.01478541e+00 -2.12609932e-01 3.54285717e-01 2.79068172e-01 5.24294615e-01 -3.13887507e-01 -5.30673385e-01 -1.05943370e+00 6.49762601e-02 -7.20809639e-01 1.06148005e-01 -6.32637382e-01 -1.03905737e+00 6.76278412e-01 -8.30044225e-02 -1.27723265e+00 4.66095477e-01 -8.17146003e-01 -4.28768337e-01 6.54068291e-01 -1.38600326e+00 -1.04485512e+00 -3.94305199e-01 4.79134738e-01 7.55746603e-01 1.04863882e-01 4.03771639e-01 4.38172847e-01 -6.12075150e-01 6.37320399e-01 -2.18990088e-01 1.21068157e-01 2.02989534e-01 -1.01527178e+00 2.24074945e-01 5.86900473e-01 2.73514837e-01 4.73410666e-01 3.95702422e-01 -4.46839601e-01 -1.48245001e+00 -1.11392200e+00 3.75952959e-01 -9.44176763e-02 2.11534187e-01 -1.13095790e-01 -1.05192149e+00 4.38164413e-01 -1.20834924e-01 5.98127723e-01 1.41330123e-01 -4.89390254e-01 -2.82222722e-02 -1.86078772e-01 -1.39405465e+00 4.04158592e-01 1.06979501e+00 -1.86956953e-02 -3.03914070e-01 7.37335086e-02 6.94854379e-01 -5.74453652e-01 -7.43436754e-01 5.49916446e-01 4.17649925e-01 -9.38982010e-01 1.07186604e+00 -7.89002329e-02 2.31852502e-01 -4.48747516e-01 -1.02108307e-01 -7.21696496e-01 -3.35293442e-01 -3.46724540e-01 -1.90159470e-01 1.13323069e+00 3.87867123e-01 -6.65030420e-01 1.01277888e+00 6.00440443e-01 -3.17658871e-01 -1.19823611e+00 -8.03972006e-01 -6.79596961e-01 -7.66709223e-02 -5.53887188e-01 5.94648480e-01 7.42901206e-01 -6.53714418e-01 7.05532953e-02 4.01980476e-03 2.18809530e-01 7.93362677e-01 4.22083616e-01 6.59467459e-01 -1.16959858e+00 -3.83357167e-01 -6.35713696e-01 -4.10542250e-01 -1.71699357e+00 -1.44703299e-01 -8.40085149e-01 2.81966291e-02 -1.23251498e+00 6.56535253e-02 -1.02883971e+00 3.55277359e-02 5.24894834e-01 6.44111484e-02 5.48678935e-01 3.90597820e-01 2.23454654e-01 -3.89596879e-01 4.92298841e-01 1.36332142e+00 -2.56025940e-01 -3.23382825e-01 9.47941765e-02 -4.89736497e-01 1.07885063e+00 4.76581633e-01 -5.17587662e-01 -3.14677060e-01 -5.50790071e-01 -7.42274001e-02 2.51306564e-01 2.53844231e-01 -1.04245651e+00 3.44636559e-01 -2.36532465e-01 1.28989533e-01 -6.88839734e-01 5.61829865e-01 -1.05119383e+00 1.66938752e-01 6.29678190e-01 2.33206034e-01 -2.49541715e-01 2.31695905e-01 3.98288786e-01 -3.33712071e-01 -5.12696505e-01 1.01053548e+00 -5.88990301e-02 -8.52555811e-01 7.73593307e-01 1.64459676e-01 4.20316719e-02 1.12175405e+00 -6.46172941e-01 1.26731157e-01 -8.28921050e-03 -1.10915434e+00 1.32012919e-01 2.76223034e-01 5.07546999e-02 7.54162371e-01 -1.23122418e+00 -5.23649752e-01 2.75097162e-01 -2.68631697e-01 5.37707627e-01 2.19743729e-01 8.18726599e-01 -9.43685055e-01 4.01691496e-01 1.61905351e-04 -9.01355326e-01 -1.00651574e+00 4.89604145e-01 3.08716893e-01 -1.52213469e-01 -7.22753406e-01 9.13519025e-01 6.91330969e-01 -2.33980432e-01 8.39547291e-02 -5.36702812e-01 -1.58640966e-01 -2.91709900e-02 1.79507717e-01 2.78996140e-01 6.18234798e-02 -7.92013824e-01 -2.64686197e-01 1.12164700e+00 2.12239578e-01 1.90328851e-01 1.21092331e+00 1.42140299e-01 -1.44541711e-01 -2.13020165e-02 9.57864821e-01 -4.38732877e-02 -1.29984951e+00 -2.77405530e-01 -1.41701758e-01 -3.41511160e-01 4.39329520e-02 -4.58025992e-01 -1.47315037e+00 8.60017776e-01 7.18670607e-01 9.11142528e-02 1.12556386e+00 -1.27172589e-01 9.72319245e-01 2.34375805e-01 6.90519214e-01 -8.59860718e-01 -2.40263790e-01 4.96386766e-01 9.94194150e-01 -1.10976648e+00 1.16893440e-01 -1.13400948e+00 -5.60584188e-01 1.14554572e+00 6.55481577e-01 -5.37195683e-01 8.03964138e-01 3.53403211e-01 -1.12971649e-01 -2.89922446e-01 -3.46525997e-01 -2.14096636e-01 4.99894023e-01 2.44458675e-01 1.51850164e-01 2.38112748e-01 -4.12686765e-01 8.03847969e-01 -1.84731916e-01 -2.37913784e-02 2.01195523e-01 4.41718012e-01 -5.52957594e-01 -8.33701074e-01 -3.66272688e-01 4.54310417e-01 -2.63129562e-01 7.17817023e-02 4.91799675e-02 7.03809738e-01 4.89694476e-01 4.89845455e-01 2.30930582e-01 1.10197052e-01 4.62798238e-01 -2.57258177e-01 3.33545953e-01 -7.57781148e-01 -5.03934801e-01 1.50866330e-01 -3.00283760e-01 -7.14399695e-01 -2.05982611e-01 -4.97368127e-01 -1.70445085e+00 -1.53345317e-01 -3.92325372e-01 -5.88864386e-02 6.81392848e-01 8.89603019e-01 2.61869609e-01 2.27562323e-01 4.84192729e-01 -1.03158319e+00 -5.56239307e-01 -4.72580284e-01 -5.08646488e-01 3.02645445e-01 -1.76205665e-01 -7.42528498e-01 -2.08764538e-01 2.57765651e-01]	[9.452642440795898, -0.058503374457359314]
cb1a10ca-d476-42c7-b7cb-d11dc8ea8c6a	multi-aspect-explainable-inductive-relation	2301.01664	null	https://arxiv.org/abs/2301.01664v2	https://arxiv.org/pdf/2301.01664v2.pdf	Multi-Aspect Explainable Inductive Relation Prediction by Sentence Transformer	Recent studies on knowledge graphs (KGs) show that path-based methods empowered by pre-trained language models perform well in the provision of inductive and explainable relation predictions. In this paper, we introduce the concepts of relation path coverage and relation path confidence to filter out unreliable paths prior to model training to elevate the model performance. Moreover, we propose Knowledge Reasoning Sentence Transformer (KRST) to predict inductive relations in KGs. KRST is designed to encode the extracted reliable paths in KGs, allowing us to properly cluster paths and provide multi-aspect explanations. We conduct extensive experiments on three real-world datasets. The experimental results show that compared to SOTA models, KRST achieves the best performance in most transductive and inductive test cases (4 of 6), and in 11 of 12 few-shot test cases.	['Lizhen Cui', 'Chunyan Miao', 'Di Wang', 'Zhixiang Su']	2023-01-04	null	null	null	null	['inductive-relation-prediction']	['graphs']	[-1.53564485e-02 9.44804311e-01 -8.24501157e-01 -3.43536794e-01 -5.57275414e-01 -1.59716025e-01 4.89216566e-01 2.97186464e-01 4.21651334e-01 1.22221828e+00 2.89294243e-01 -6.68328941e-01 -5.73110759e-01 -1.32121944e+00 -8.69124293e-01 -1.04023822e-01 -3.20285201e-01 7.76191413e-01 5.40692031e-01 -2.06594720e-01 -3.12059019e-02 4.95271347e-02 -1.11414492e+00 6.06554568e-01 1.47130811e+00 6.93130016e-01 -3.44713509e-01 4.05572742e-01 -4.04790372e-01 1.33175302e+00 -2.68840015e-01 -1.15920317e+00 -3.98947686e-01 -5.50069213e-01 -1.30587268e+00 -3.41296375e-01 -5.02895534e-01 -1.37638906e-02 -2.80249238e-01 6.31695569e-01 -2.48420350e-02 7.48628527e-02 7.47747600e-01 -1.43604839e+00 -8.83489251e-01 1.61500788e+00 -1.89865768e-01 1.27634451e-01 7.70589232e-01 -3.06459188e-01 1.33206666e+00 -8.92494559e-01 8.98075223e-01 1.28088653e+00 6.96285367e-01 2.56755918e-01 -1.11673045e+00 -3.86068642e-01 3.51352751e-01 7.22921848e-01 -1.35005283e+00 -1.73242271e-01 6.51330173e-01 -4.99350391e-02 1.45986593e+00 3.98015112e-01 8.23170662e-01 1.13722265e+00 2.74389029e-01 1.00817633e+00 7.25040197e-01 -5.97814560e-01 1.87838450e-01 3.10597539e-01 4.11045671e-01 9.10037220e-01 7.53247499e-01 -1.42725185e-01 -1.05870271e+00 -2.49596089e-01 5.51402807e-01 -3.98092568e-01 -2.56399274e-01 -1.27512187e-01 -9.25883472e-01 8.27662289e-01 5.15450776e-01 1.74377277e-01 -3.13868761e-01 -1.56872630e-01 7.09914565e-02 2.22639784e-01 6.74261570e-01 5.06785512e-01 -8.40303302e-01 8.19579989e-04 -2.97360688e-01 1.09155610e-01 1.17996979e+00 1.30471778e+00 6.55591011e-01 -2.63805538e-01 -4.33175713e-01 5.47586322e-01 5.16182125e-01 2.26806641e-01 1.79262068e-02 -5.61870635e-01 6.75285101e-01 1.17126143e+00 -2.38425672e-01 -9.79022324e-01 -1.75953597e-01 -4.97316957e-01 -6.41554534e-01 -6.96948111e-01 -3.88929993e-02 1.01948142e-01 -8.43255699e-01 1.36526203e+00 3.56504023e-01 2.83678561e-01 5.84379852e-01 3.90678316e-01 1.06388974e+00 6.29364312e-01 3.17585975e-01 -4.95988727e-01 1.04969990e+00 -1.06528127e+00 -8.39076161e-01 -2.57432610e-01 1.06509995e+00 -3.31610203e-01 1.01098561e+00 1.92691401e-01 -8.32956731e-01 -1.20446227e-01 -8.23353589e-01 2.08230495e-01 -3.42033714e-01 -8.32823142e-02 1.13469350e+00 3.09468806e-01 -8.01092505e-01 5.84215224e-01 -8.19970727e-01 -3.15282315e-01 4.21973228e-01 1.33735776e-01 -2.94814885e-01 -4.33840990e-01 -1.59239721e+00 8.95459712e-01 1.01975346e+00 -2.12190337e-02 -6.28853381e-01 -7.86825895e-01 -9.32639956e-01 3.68833870e-01 8.12748909e-01 -1.16076970e+00 8.29063237e-01 -1.68786094e-01 -1.17900670e+00 3.22970152e-01 -3.66585165e-01 -5.37854552e-01 4.04285267e-02 -4.07126307e-01 -8.83566439e-01 1.86736971e-01 1.77873895e-01 2.50513494e-01 1.42293900e-01 -1.38209498e+00 -5.07581174e-01 -1.04771987e-01 1.89974263e-01 1.73709989e-01 -3.56954128e-01 -3.22447121e-01 -7.11637139e-01 -2.01200560e-01 4.01663035e-01 -4.80777293e-01 -2.30844885e-01 -7.87907124e-01 -1.26540434e+00 -5.85940480e-01 4.18652713e-01 -5.26672661e-01 1.32917786e+00 -1.40805984e+00 9.78234932e-02 5.31622291e-01 1.43796653e-01 5.06875105e-02 4.96036075e-02 7.27264941e-01 1.74324796e-01 3.69511247e-01 3.36486846e-02 1.88371968e-02 -1.36070520e-01 7.43228257e-01 -4.32433456e-01 -4.31613296e-01 4.48030084e-01 1.32065582e+00 -1.02876103e+00 -8.61291707e-01 -1.98163643e-01 1.84565619e-01 -4.52157825e-01 2.85474241e-01 -5.98716915e-01 3.89137626e-01 -8.78540218e-01 8.92973840e-01 2.35044047e-01 -6.64002299e-01 5.16731799e-01 1.71292603e-01 7.09219456e-01 6.46163344e-01 -8.29617918e-01 1.37556767e+00 -2.11315632e-01 1.04189299e-01 -1.00796723e+00 -7.56772399e-01 1.10819018e+00 2.64071822e-01 -4.42925319e-02 -4.57140982e-01 -1.83476597e-01 8.85725245e-02 -1.17978454e-01 -7.63245523e-01 5.41482925e-01 -1.66336950e-02 1.03688635e-01 2.73997188e-01 1.26751930e-01 -3.87600176e-02 3.24798763e-01 8.75384331e-01 1.30624664e+00 1.94087893e-01 4.96138364e-01 1.17192887e-01 4.81127083e-01 2.73626655e-01 7.34392285e-01 7.38638222e-01 3.95760179e-01 1.44764498e-01 8.73453319e-01 -2.25189000e-01 -4.79353249e-01 -1.07821679e+00 5.66889979e-02 6.62865698e-01 3.37666333e-01 -9.48044300e-01 -3.23601753e-01 -1.22930372e+00 -8.50686207e-02 1.40404260e+00 -4.29150790e-01 -4.82658446e-01 -1.36265218e-01 -7.65115917e-01 5.55688262e-01 7.88134515e-01 3.08970600e-01 -1.11102355e+00 3.06343973e-01 1.76927671e-01 -6.39034331e-01 -1.31303895e+00 4.19146538e-01 8.98382217e-02 -1.01400697e+00 -1.33357847e+00 3.26884985e-01 -4.64449704e-01 8.48359585e-01 -2.93302480e-02 1.44707072e+00 3.08057994e-01 1.42394498e-01 1.40655994e-01 -9.28892791e-01 -2.75951475e-01 -3.27431470e-01 2.12396964e-01 -7.90321380e-02 -5.28065026e-01 5.64500272e-01 -7.66224504e-01 -6.35945275e-02 2.39484355e-01 -4.62475598e-01 3.52130026e-01 5.88823676e-01 8.08412731e-01 7.61077881e-01 4.49488461e-01 7.67620146e-01 -1.35387945e+00 7.04407215e-01 -7.36337125e-01 -6.50193021e-02 8.05316508e-01 -1.38645589e+00 2.16551751e-01 4.63939339e-01 -2.01854892e-02 -1.39088964e+00 -4.98260915e-01 -6.56992942e-02 1.86627939e-01 1.21328801e-01 1.17116690e+00 -1.98171318e-01 2.01398462e-01 7.32105136e-01 1.09907791e-01 -6.90287292e-01 -2.70599723e-01 5.55237412e-01 6.06345348e-02 2.25619778e-01 -7.84704387e-01 7.28680074e-01 6.16606772e-02 3.16323377e-02 -8.84204358e-02 -1.29993951e+00 -6.72504082e-02 -5.28684020e-01 -9.27997902e-02 4.73750830e-01 -7.56295085e-01 -5.32847524e-01 -3.22981209e-01 -1.15610945e+00 -1.24050565e-01 -4.73699152e-01 5.75988889e-01 -3.62687051e-01 8.93640965e-02 -7.98503220e-01 -1.00641906e+00 -5.12518108e-01 -5.70959151e-01 6.22298300e-01 3.21949005e-01 -4.21613634e-01 -1.18413091e+00 1.31532755e-02 5.77347755e-01 -1.76864222e-01 1.95615426e-01 1.39533699e+00 -9.44549561e-01 -8.63533497e-01 -2.62256116e-01 -2.18255699e-01 -3.11407205e-02 -1.38159841e-01 7.83020779e-02 -8.12223911e-01 4.22250688e-01 -6.45647109e-01 -4.46747631e-01 9.16815162e-01 1.61467701e-01 9.79323328e-01 -4.62403953e-01 -8.85929525e-01 1.94307908e-01 1.16268206e+00 1.42393857e-01 9.05776858e-01 1.24627374e-01 8.56540203e-01 5.40855706e-01 1.03220665e+00 1.12970859e-01 8.05616319e-01 2.55680501e-01 1.65592164e-01 2.17302054e-01 -1.12336818e-02 -1.15795088e+00 2.53412843e-01 1.02866232e+00 -4.65210289e-01 -5.97960293e-01 -8.87203693e-01 6.65476799e-01 -2.28878188e+00 -7.88183391e-01 -6.45901382e-01 1.64343369e+00 9.91408288e-01 4.31612521e-01 -3.98868054e-01 3.74651372e-01 4.07680094e-01 -2.72148728e-01 -3.02266419e-01 -4.09777969e-01 -2.30235830e-01 1.19443789e-01 -3.81389521e-02 7.46728778e-01 -4.87733603e-01 1.37071466e+00 6.78591013e+00 9.11971211e-01 -3.97669559e-04 -1.04626618e-01 6.30391538e-01 2.93181449e-01 -1.04860747e+00 4.60106134e-01 -9.12995636e-01 1.82176773e-02 9.82172608e-01 -3.58128667e-01 6.87422454e-02 8.27699184e-01 -1.27079412e-01 -2.83178538e-01 -1.10040224e+00 2.37786978e-01 -2.17790917e-01 -1.52367961e+00 4.97596323e-01 -2.04656675e-01 9.84887183e-01 -5.44405878e-01 -4.03531611e-01 5.60779810e-01 8.24032128e-01 -1.10794377e+00 2.52858251e-01 7.15493262e-01 3.77756089e-01 -8.33923519e-01 1.04067755e+00 4.28129166e-01 -1.15921557e+00 7.13804411e-03 -4.46297050e-01 -1.32365599e-01 2.81927735e-01 1.14362049e+00 -1.57091606e+00 1.65687203e+00 5.89545965e-01 9.00330424e-01 -4.58721757e-01 7.54615247e-01 -1.24311614e+00 1.12906945e+00 -1.06983379e-01 -1.05256498e-01 -1.88329443e-01 2.90787704e-02 3.91912580e-01 1.03087449e+00 3.36846024e-01 4.23888624e-01 -4.86455299e-02 9.55802083e-01 1.02475111e-03 1.73144698e-01 -6.42952502e-01 -1.85068473e-01 6.78956449e-01 1.05077636e+00 -7.77437270e-01 -5.21150887e-01 -1.37960687e-01 6.85596168e-01 7.21772254e-01 4.99337018e-01 -7.40719438e-01 -9.24351513e-02 1.23946406e-01 7.88737647e-03 2.19945118e-01 1.68084249e-01 -5.81222236e-01 -1.20279002e+00 1.11323506e-01 -2.25364387e-01 8.82305622e-01 -9.85834479e-01 -1.29499543e+00 8.13390732e-01 1.96507066e-01 -6.65639818e-01 -5.14634609e-01 -1.40588835e-01 -7.63471603e-01 6.17810488e-01 -1.54120612e+00 -1.50847781e+00 -9.04162601e-02 4.72151309e-01 3.60683501e-02 -8.37667212e-02 1.10231578e+00 -5.63271344e-03 -4.86894935e-01 4.63274240e-01 -5.42207479e-01 -2.05385074e-01 2.40776286e-01 -1.15771592e+00 3.72027606e-01 9.27430809e-01 4.69997644e-01 6.83211923e-01 7.80701935e-01 -1.16806448e+00 -1.13848913e+00 -1.23657131e+00 1.50858903e+00 -5.26347101e-01 5.44659674e-01 8.89107957e-02 -1.07864523e+00 1.10831285e+00 -2.13250995e-01 -5.58776024e-04 9.01409566e-01 9.48996484e-01 -5.21762550e-01 -1.18118175e-03 -1.00465739e+00 4.55617845e-01 1.63839686e+00 -3.08113873e-01 -7.27395296e-01 5.26646674e-01 1.26232016e+00 -3.56580079e-01 -1.02024448e+00 7.52417624e-01 1.70161709e-01 -8.60569119e-01 9.13493812e-01 -1.00453711e+00 9.21571255e-01 -2.96326488e-01 5.93245625e-02 -1.37438726e+00 -3.82215083e-01 -2.75206745e-01 -8.73042583e-01 1.52488732e+00 1.19493759e+00 -6.37981832e-01 9.01080608e-01 6.23862028e-01 -2.23079011e-01 -1.25381446e+00 -5.74086428e-01 -7.89343596e-01 -4.48080897e-01 -5.79747200e-01 7.75373757e-01 9.44608271e-01 6.45395517e-01 6.05236828e-01 -2.26829618e-01 4.10824895e-01 4.55696374e-01 3.70424181e-01 5.48268914e-01 -1.31199050e+00 -3.44103158e-01 1.84381962e-01 -3.07692230e-01 -6.55650139e-01 4.44081068e-01 -1.06919372e+00 -1.13388881e-01 -2.23856425e+00 5.97611785e-01 -6.06639385e-01 -1.84802935e-01 8.34967196e-01 -6.56072915e-01 -1.10838357e-02 -5.11609972e-01 4.30333652e-02 -7.86707699e-01 8.03603232e-01 1.42312753e+00 5.60827972e-03 -3.66862416e-01 6.99338764e-02 -1.16109610e+00 6.19020700e-01 6.86566472e-01 -4.85383064e-01 -1.12053788e+00 -1.92421898e-01 6.09836996e-01 2.00043753e-01 2.30208099e-01 -5.27815461e-01 2.04553038e-01 -3.85584742e-01 3.29357207e-01 -6.11868799e-01 2.25459427e-01 -4.45157677e-01 3.39949608e-01 2.76214838e-01 -4.67005014e-01 -5.76218128e-01 2.40691248e-02 8.61848712e-01 -2.47910753e-01 5.92731684e-02 -5.86644448e-02 2.16656402e-01 -7.97699273e-01 2.28740245e-01 8.21269080e-02 6.25485703e-02 9.53935444e-01 1.16239347e-01 -5.51545918e-01 -4.04513448e-01 -7.63187647e-01 4.61592585e-01 -1.91003963e-01 3.56712431e-01 9.55814838e-01 -1.47207224e+00 -6.07016444e-01 -8.41285735e-02 4.80549008e-01 2.80927777e-01 1.43366337e-01 7.80063570e-01 -9.77227390e-02 6.03930652e-01 3.63169491e-01 -2.25474879e-01 -1.00711727e+00 4.44703311e-01 -1.70812204e-01 -8.35691571e-01 -8.47815633e-01 1.12659979e+00 -3.13087881e-01 -4.00250793e-01 -4.80161496e-02 -4.37167674e-01 -3.59626055e-01 -4.81163204e-01 3.66969019e-01 2.10492685e-01 -6.80383146e-02 4.76014428e-02 -4.33176875e-01 9.56831649e-02 -2.58465767e-01 1.85655728e-01 1.26712644e+00 -3.06426659e-02 -2.62205690e-01 3.35817486e-01 4.87657964e-01 -4.77663949e-02 -6.18045688e-01 -4.98103142e-01 4.37127948e-01 -2.96198547e-01 -2.87362278e-01 -1.24187815e+00 -6.49726391e-01 3.47036362e-01 -5.93068480e-01 3.20032179e-01 8.70184124e-01 5.37855148e-01 7.53577530e-01 5.85536301e-01 6.11233234e-01 -6.93655372e-01 -1.05188198e-01 5.15608549e-01 7.73649514e-01 -1.05006766e+00 1.64611056e-01 -1.36609590e+00 -9.90405738e-01 8.36636722e-01 8.26508641e-01 3.90557945e-01 3.81813467e-01 2.08779410e-01 -3.85287285e-01 -5.02478659e-01 -1.21466517e+00 -1.70287028e-01 4.53656286e-01 6.45558715e-01 3.51767898e-01 3.44629705e-01 -2.51903981e-01 1.16691864e+00 -2.94557810e-01 1.08885989e-01 1.86078712e-01 6.26337349e-01 -5.18021822e-01 -1.13912177e+00 9.88028347e-02 6.01345181e-01 5.29972650e-03 -3.58972818e-01 -7.90598631e-01 6.52536929e-01 -2.63251606e-02 1.14831173e+00 -4.77369159e-01 -8.37406754e-01 2.78434396e-01 4.04986918e-01 4.77704495e-01 -5.35515666e-01 -1.70432866e-01 -5.58825612e-01 9.70259726e-01 -5.01690507e-01 -2.89582819e-01 -2.34602094e-01 -1.72013915e+00 -6.29899681e-01 -6.99759603e-01 5.95620632e-01 1.27953431e-03 1.40944874e+00 3.66275549e-01 8.79339218e-01 3.57702285e-01 2.67890841e-01 -3.62875871e-04 -1.01335001e+00 -6.61821365e-01 1.29476145e-01 -5.09297669e-01 -6.85449839e-01 -2.71602571e-02 1.07349793e-03]	[9.013265609741211, 8.067075729370117]
40becb42-0b8d-4cf7-8af5-5910f46cdc95	kernel-e-greedy-for-contextual-bandits	2306.17329	null	https://arxiv.org/abs/2306.17329v1	https://arxiv.org/pdf/2306.17329v1.pdf	Kernel $ε$-Greedy for Contextual Bandits	We consider a kernelized version of the $\epsilon$-greedy strategy for contextual bandits. More precisely, in a setting with finitely many arms, we consider that the mean reward functions lie in a reproducing kernel Hilbert space (RKHS). We propose an online weighted kernel ridge regression estimator for the reward functions. Under some conditions on the exploration probability sequence, $\{\epsilon_t\}_t$, and choice of the regularization parameter, $\{\lambda_t\}_t$, we show that the proposed estimator is consistent. We also show that for any choice of kernel and the corresponding RKHS, we achieve a sub-linear regret rate depending on the intrinsic dimensionality of the RKHS. Furthermore, we achieve the optimal regret rate of $\sqrt{T}$ under a margin condition for finite-dimensional RKHS.	['Bharath K. Sriperumbudur', 'Sakshi Arya']	2023-06-29	null	null	null	null	['multi-armed-bandits']	['miscellaneous']	[-2.49576584e-01 3.26635629e-01 -4.72331583e-01 -2.23022103e-01 -1.15631402e+00 -5.88860989e-01 -1.35910749e-01 5.68708368e-02 -7.64584720e-01 8.91614497e-01 -1.55591905e-01 -6.30763531e-01 -6.36590421e-01 -6.50090754e-01 -9.56839025e-01 -7.93648541e-01 -6.07322752e-01 1.39833108e-01 -3.68967652e-01 1.84024975e-01 2.59641588e-01 2.26222277e-01 -9.07963991e-01 -4.34057355e-01 1.02818394e+00 1.67919683e+00 -9.88562256e-02 6.34540379e-01 1.68576837e-01 5.39355993e-01 -8.91031623e-02 -5.72089374e-01 7.23177075e-01 -5.50550044e-01 -7.06632495e-01 -4.70441878e-02 -7.15017244e-02 -1.84108272e-01 -4.07118738e-01 1.37785983e+00 1.84666440e-01 5.38540125e-01 7.63695121e-01 -8.40704381e-01 -4.65923101e-01 8.09644401e-01 -7.89156795e-01 1.63428243e-02 -1.95312843e-01 -8.71832520e-02 1.28735685e+00 -6.51652098e-01 2.96030581e-01 6.40454948e-01 6.40764892e-01 3.10290009e-01 -1.47113848e+00 -9.16630983e-01 2.30677634e-01 -2.52304226e-01 -1.20679414e+00 -2.61289626e-01 5.21075130e-01 -4.96545404e-01 2.05790907e-01 3.37750375e-01 4.73794043e-01 5.51767290e-01 -1.70053080e-01 7.76596546e-01 1.28813851e+00 -5.44009089e-01 7.74670362e-01 2.78454363e-01 2.29278117e-01 7.93970823e-01 3.01145494e-01 2.37191275e-01 -2.85601050e-01 -4.79214162e-01 9.78059351e-01 2.15229988e-02 -1.38226211e-01 -6.45734847e-01 -8.34898055e-01 1.26518965e+00 8.99053290e-02 -3.23074833e-02 -4.71349537e-01 5.41724026e-01 1.90356761e-01 4.71709788e-01 6.32890165e-01 2.55411774e-01 -3.25697184e-01 -2.27993995e-01 -7.95250595e-01 2.00889945e-01 7.78459489e-01 1.03797460e+00 7.52928317e-01 1.73225068e-02 -3.50455791e-01 6.56184733e-01 -2.85419580e-02 7.78769016e-01 -8.64198282e-02 -1.24719942e+00 8.91926706e-01 1.22948371e-01 9.47296798e-01 -3.78501207e-01 -2.03004122e-01 -5.18501937e-01 -6.43140197e-01 -1.95896961e-02 8.20167959e-01 -6.12277091e-01 -4.04081404e-01 1.92975640e+00 3.44209433e-01 -9.72539186e-03 -1.88953400e-01 9.42994416e-01 -5.27763784e-01 3.77334774e-01 -2.54073650e-01 -5.64416766e-01 9.74864602e-01 -5.15482903e-01 -3.87549460e-01 -2.46471971e-01 6.90342844e-01 -4.02879357e-01 1.23519146e+00 3.64188343e-01 -1.14297128e+00 1.03548825e-01 -9.46751833e-01 4.80706006e-01 6.73907027e-02 1.39484003e-01 8.75780404e-01 9.03346479e-01 -5.67331731e-01 7.50268459e-01 -6.08396888e-01 3.24497193e-01 4.37546104e-01 3.59861612e-01 -8.62851888e-02 4.76599373e-02 -9.07363713e-01 3.51926833e-01 3.12094629e-01 1.67846560e-01 -7.52178490e-01 -7.23987103e-01 -4.88193840e-01 3.19037437e-02 4.65155393e-01 -3.40440609e-02 1.20579767e+00 -7.17227280e-01 -1.62509942e+00 6.11698151e-01 1.01944245e-01 -6.41103745e-01 7.62811542e-01 -2.14344487e-01 -1.03395410e-01 1.05240390e-01 1.79990549e-02 -2.30733007e-01 1.17554748e+00 -5.10730743e-01 -5.32877266e-01 -6.05577350e-01 2.11513415e-02 1.01668537e-02 -3.08842152e-01 -1.55096501e-01 6.18492886e-02 -6.11786544e-01 8.59508067e-02 -1.25402951e+00 -4.46138233e-01 -9.10742953e-02 -2.44669795e-01 1.41479135e-01 6.05837964e-02 -5.64522088e-01 1.17944288e+00 -2.26463604e+00 1.19787978e-03 7.23073006e-01 -2.09801063e-01 -2.60533929e-01 2.32707322e-01 1.46999940e-01 -2.50504408e-02 -5.84349632e-02 -2.82612532e-01 -8.63060951e-02 3.16522270e-01 -1.87212214e-01 -5.86083293e-01 9.39305544e-01 -5.32312036e-01 5.19998193e-01 -5.22888184e-01 9.09958631e-02 -1.21119782e-01 -6.63660988e-02 -5.18951178e-01 3.66338231e-02 -1.79678261e-01 2.12117121e-01 -8.90919507e-01 1.42385602e-01 6.43875539e-01 -4.91029352e-01 1.24219768e-01 4.62112248e-01 -1.81055404e-02 -7.08489716e-02 -1.60381866e+00 1.60653889e+00 -5.01714170e-01 2.35557660e-01 4.85416293e-01 -1.14017272e+00 7.10913420e-01 7.13281631e-02 4.91570830e-01 -3.88939410e-01 1.73973978e-01 3.77873600e-01 -5.41224957e-01 7.79374465e-02 3.19285959e-01 -4.77406591e-01 -2.54622966e-01 8.85530353e-01 -2.68580943e-01 3.69012445e-01 -9.65379030e-02 -1.12703547e-01 1.10990417e+00 -1.32528290e-01 3.61321926e-01 -4.86740202e-01 1.12245873e-01 -3.59914392e-01 5.45354128e-01 1.13480783e+00 -1.40169755e-01 1.00921936e-01 1.08012497e+00 -2.10257307e-01 -1.02165306e+00 -1.00810599e+00 -2.56725430e-01 1.57886112e+00 -4.52355249e-03 1.34882078e-01 -7.39259243e-01 -6.71938956e-01 4.72169012e-01 8.95476222e-01 -8.87347221e-01 -9.22227558e-03 -2.11254135e-01 -7.97884285e-01 4.25881445e-01 4.29447919e-01 6.01631284e-01 -6.22982323e-01 -7.86389709e-01 2.06602454e-01 2.84709662e-01 -8.38711977e-01 -8.18142653e-01 5.25992095e-01 -9.60928977e-01 -6.40253663e-01 -9.54385757e-01 -2.15212539e-01 6.57775283e-01 6.99931532e-02 4.61373389e-01 -8.72894645e-01 -1.64728001e-01 4.56497371e-01 -9.90867168e-02 -2.24058717e-01 2.38498643e-01 -4.12254892e-02 -3.57455090e-02 2.02293485e-01 -5.76212369e-02 -4.25527126e-01 -9.74639475e-01 2.22249538e-01 -7.18518615e-01 -1.78914875e-01 4.29846704e-01 8.30898345e-01 5.78974843e-01 -1.85011283e-01 5.07826567e-01 -9.73147690e-01 6.49616003e-01 -5.04156828e-01 -1.38633227e+00 3.35751772e-01 -7.72117674e-01 6.28546059e-01 5.66445768e-01 -6.63428009e-01 -8.38082254e-01 7.20841214e-02 6.51300967e-01 -5.07250845e-01 4.24659312e-01 3.89458030e-01 3.03381771e-01 -6.80326745e-02 6.98033094e-01 2.54074246e-01 -7.61078298e-02 -4.49633747e-01 6.27702057e-01 4.34908360e-01 1.06959395e-01 -1.03362405e+00 4.73462045e-01 4.79616761e-01 9.90966856e-02 -5.11625290e-01 -8.79262924e-01 -1.23372138e-01 1.03792816e-01 7.58913085e-02 4.25654411e-01 -7.09447920e-01 -1.47737539e+00 -8.92743319e-02 -4.34807420e-01 -7.18172908e-01 -6.51941299e-01 9.84668434e-01 -1.27592862e+00 1.66519150e-01 -5.47055364e-01 -1.71443164e+00 -3.04369569e-01 -7.10316777e-01 4.01710302e-01 1.47803724e-01 1.93134412e-01 -6.88441277e-01 -2.24598367e-02 3.77309561e-01 4.78491992e-01 -1.69171207e-02 8.77343237e-01 -4.97517258e-01 -4.90639627e-01 -4.96854097e-01 -4.47006971e-01 4.05936450e-01 -9.41557810e-02 -6.93554103e-01 -5.52779138e-01 -3.65012288e-01 8.58475342e-02 -3.66020530e-01 8.27622116e-01 6.98867798e-01 1.40493286e+00 -9.44285989e-01 -5.16020134e-02 5.59518576e-01 1.16704416e+00 2.22259432e-01 2.36753821e-01 1.98792800e-01 3.85812595e-02 4.39786464e-01 7.49940813e-01 1.00228441e+00 -6.98200092e-02 2.41931126e-01 1.35478556e-01 3.22018087e-01 9.82546449e-01 -3.40544909e-01 3.34968179e-01 2.15528905e-02 -3.34820338e-02 4.56298858e-01 -5.44066727e-01 4.21483248e-01 -2.09600067e+00 -6.76463664e-01 6.28858805e-01 2.99880815e+00 1.10673451e+00 7.38850981e-02 4.96737719e-01 -3.44989061e-01 8.43008697e-01 7.75795281e-02 -1.08626914e+00 -4.07569289e-01 4.92348820e-02 6.21332526e-01 1.43357837e+00 4.64503139e-01 -9.19140756e-01 6.72134817e-01 5.89574671e+00 1.03898716e+00 -1.00328135e+00 1.70705795e-01 8.16238880e-01 -4.92723495e-01 -1.14207737e-01 3.27354193e-01 -6.01095319e-01 7.29832411e-01 9.55622852e-01 -1.96697250e-01 9.60719526e-01 1.40172648e+00 4.41707522e-02 -3.69605601e-01 -9.49514687e-01 9.82540429e-01 -5.99697053e-01 -1.26062024e+00 -9.20768380e-01 4.97650057e-01 6.53634489e-01 -2.65309453e-01 5.43549478e-01 2.97961414e-01 7.08779454e-01 -1.09311426e+00 6.39948666e-01 3.11733991e-01 8.78120482e-01 -1.18725288e+00 2.72727251e-01 4.69020456e-01 -9.96221542e-01 -5.66447377e-01 -4.67371851e-01 1.32710204e-01 -2.55635321e-01 7.75851727e-01 -5.21959305e-01 2.40437135e-01 5.38056850e-01 3.59521657e-02 1.70772642e-01 9.10484910e-01 9.85448509e-02 5.25482953e-01 -5.34919560e-01 -1.13356560e-01 2.62551904e-01 -6.83386505e-01 3.34789217e-01 1.07677972e+00 5.17372608e-01 3.59774441e-01 3.01269833e-02 7.69940197e-01 -2.83472270e-01 4.04473871e-01 -3.33766520e-01 -2.33458817e-01 4.98582691e-01 8.34251642e-01 -4.27171558e-01 3.43680531e-02 -6.78443387e-02 9.84921396e-01 6.11443937e-01 6.64456725e-01 -7.43973792e-01 -5.65458953e-01 7.16321945e-01 -6.18190132e-02 7.49313533e-01 -2.74965703e-01 -3.79948705e-01 -1.10807884e+00 4.18728411e-01 -3.55533153e-01 7.09876597e-01 -1.53312415e-01 -1.19581473e+00 -1.62777066e-01 1.54713066e-02 -7.48313308e-01 -2.85900623e-01 -4.86420423e-01 -3.08154881e-01 8.40748072e-01 -9.80267882e-01 -5.07046223e-01 6.26552284e-01 6.58233762e-01 -1.60173133e-01 -1.19994335e-01 7.52962649e-01 -3.06394637e-01 -5.53978443e-01 8.03917527e-01 9.73462224e-01 6.30146563e-02 2.71914065e-01 -1.17164695e+00 -1.82956740e-01 4.39882219e-01 -9.04426351e-02 5.92361689e-01 7.63199508e-01 -3.97205472e-01 -1.56965220e+00 -9.51265454e-01 1.00827500e-01 2.81340718e-01 9.94245589e-01 -4.73599106e-01 -5.34484565e-01 8.82067323e-01 -1.82540447e-01 4.09770221e-01 7.84645021e-01 4.84353065e-01 -6.41788781e-01 -4.56556052e-01 -1.46854889e+00 5.12251735e-01 9.42401409e-01 -6.46181345e-01 8.63419622e-02 4.24872041e-01 4.65729654e-01 -3.27395678e-01 -1.07657647e+00 1.61456510e-01 7.43290842e-01 -8.36956382e-01 6.98242903e-01 -9.30681527e-01 -2.68479615e-01 2.42955193e-01 -4.74168092e-01 -9.53444779e-01 -3.10978770e-01 -1.15500784e+00 -3.42850059e-01 5.14756024e-01 5.36236286e-01 -9.89900172e-01 1.01050973e+00 1.13572741e+00 4.47676450e-01 -9.08918500e-01 -1.48301709e+00 -1.02038157e+00 4.63087738e-01 -5.92660367e-01 1.92645296e-01 4.54223394e-01 3.82715195e-01 -2.27816716e-01 -7.10101426e-01 -6.31001741e-02 9.63773310e-01 3.75875592e-01 5.01830995e-01 -8.88452649e-01 -8.97026062e-01 -3.52994800e-01 1.06267199e-01 -1.04417968e+00 2.49505058e-01 -4.55602169e-01 -1.92461133e-01 -5.92839360e-01 2.82347322e-01 -9.96087790e-01 -7.35721648e-01 3.24476212e-01 1.40691802e-01 -3.79637927e-01 1.18084729e-01 4.46166694e-02 -6.26773357e-01 5.13993382e-01 9.06604648e-01 2.27627888e-01 -3.50885510e-01 3.93580019e-01 -7.98923850e-01 4.61796433e-01 7.53074050e-01 -4.22082007e-01 -3.43282729e-01 9.73845571e-02 4.95917022e-01 8.36421669e-01 2.36760229e-01 -3.79875571e-01 5.41078765e-03 -7.21653044e-01 2.30524361e-01 -3.70288730e-01 4.49440092e-01 -3.79959196e-01 8.86029676e-02 3.14099669e-01 -8.78585160e-01 -4.39295828e-01 -1.80470243e-01 9.95807886e-01 5.47230959e-01 -2.70838439e-01 1.06340933e+00 5.20618260e-02 2.32364923e-01 3.48659545e-01 -1.18902035e-01 2.74430990e-01 1.15912950e+00 4.54623848e-02 7.05899373e-02 -5.90331554e-01 -6.29792511e-01 1.26686320e-01 2.26670533e-01 -2.85472661e-01 3.20709050e-01 -1.29339504e+00 -3.41209084e-01 -1.42998016e-02 1.52535159e-02 -3.05010825e-01 3.57694656e-01 9.12805915e-01 6.76297070e-03 5.71574211e-01 3.48232090e-01 -3.83399092e-02 -5.25777638e-01 5.33203721e-01 2.93325782e-01 -2.48235092e-01 -5.00836074e-01 9.38254714e-01 8.60855430e-02 7.54775926e-02 3.88620049e-01 -3.33444118e-01 5.54756403e-01 -1.72342971e-01 5.03043234e-01 7.40167320e-01 -2.21639588e-01 9.44566205e-02 -7.61484131e-02 1.76273435e-01 -1.49637580e-01 -8.04444194e-01 1.12567294e+00 -9.99711603e-02 1.50549516e-01 4.51028675e-01 1.30875838e+00 7.53682330e-02 -1.68484342e+00 -4.64619905e-01 6.60887286e-02 -6.48156703e-01 -1.66705027e-01 -4.97110903e-01 -9.14337754e-01 4.32328373e-01 7.47035861e-01 2.68957555e-01 7.10892081e-01 6.60254508e-02 4.34076548e-01 4.42823231e-01 6.43617630e-01 -1.60868955e+00 -3.01011592e-01 4.18521255e-01 5.07103741e-01 -1.02950799e+00 -1.55501813e-01 7.65511096e-02 -7.25841165e-01 9.31366622e-01 -1.10876128e-01 -3.73929232e-01 7.81365514e-01 -2.04492271e-01 -5.62196314e-01 2.24051639e-01 -4.43684012e-01 -5.93809783e-02 1.09694257e-01 1.33944839e-01 1.62118301e-01 5.73017359e-01 -4.00519073e-01 8.03735197e-01 -1.37677953e-01 -3.07876877e-02 7.82827213e-02 7.44383991e-01 -7.01186061e-01 -9.18537259e-01 -2.99247056e-01 6.06839359e-01 -6.68173254e-01 6.63911030e-02 8.84621814e-02 5.91041982e-01 -5.26206613e-01 7.59975731e-01 -1.91421553e-01 1.10912912e-01 -2.95067653e-02 3.24280798e-01 7.16089785e-01 -2.20721647e-01 -1.43529177e-01 1.36589259e-01 -1.15255684e-01 -5.88426769e-01 2.26049721e-01 -6.47787809e-01 -1.20511341e+00 -5.35049438e-01 -2.73119003e-01 6.25016868e-01 6.99682117e-01 9.41376626e-01 2.63653427e-01 -4.35823560e-01 1.12961638e+00 -1.52735889e-01 -1.36927211e+00 -8.15185905e-01 -1.33552492e+00 3.10080382e-03 2.90293515e-01 -5.39794981e-01 -5.19879341e-01 -6.32939041e-01]	[4.7273783683776855, 3.4760966300964355]
06f28e73-79cb-48cc-9aea-05a97a63f9a1	asymmetric-feature-maps-with-application-to	1704.03946	null	http://arxiv.org/abs/1704.03946v1	http://arxiv.org/pdf/1704.03946v1.pdf	Asymmetric Feature Maps with Application to Sketch Based Retrieval	We propose a novel concept of asymmetric feature maps (AFM), which allows to evaluate multiple kernels between a query and database entries without increasing the memory requirements. To demonstrate the advantages of the AFM method, we derive a short vector image representation that, due to asymmetric feature maps, supports efficient scale and translation invariant sketch-based image retrieval. Unlike most of the short-code based retrieval systems, the proposed method provides the query localization in the retrieved image. The efficiency of the search is boosted by approximating a 2D translation search via trigonometric polynomial of scores by 1D projections. The projections are a special case of AFM. An order of magnitude speed-up is achieved compared to traditional trigonometric polynomials. The results are boosted by an image-based average query expansion, exceeding significantly the state of the art on standard benchmarks.	['Ondřej Chum', 'Giorgos Tolias']	2017-04-12	asymmetric-feature-maps-with-application-to-1	http://openaccess.thecvf.com/content_cvpr_2017/html/Tolias_Asymmetric_Feature_Maps_CVPR_2017_paper.html	http://openaccess.thecvf.com/content_cvpr_2017/papers/Tolias_Asymmetric_Feature_Maps_CVPR_2017_paper.pdf	cvpr-2017-7	['sketch-based-image-retrieval']	['computer-vision']	[ 1.10088885e-01 -7.23142922e-01 -2.34050214e-01 1.36822537e-01 -8.80592346e-01 -9.87194300e-01 8.91857028e-01 2.83477366e-01 -4.60078120e-01 1.79784387e-01 -1.61506698e-01 -2.55940884e-01 -4.82604265e-01 -6.57912254e-01 -6.89070046e-01 -5.11745274e-01 3.22461650e-02 5.82420468e-01 5.19294441e-01 -3.21251154e-01 7.22507775e-01 1.01807070e+00 -1.60315263e+00 1.02108516e-01 4.62218702e-01 1.39325953e+00 1.48236424e-01 6.65495455e-01 -2.02200010e-01 -3.38016339e-02 -3.91174555e-01 -6.37311578e-01 5.36292911e-01 2.50887632e-01 -3.97464514e-01 -2.76575595e-01 1.10260916e+00 -5.32417834e-01 -5.55263042e-01 8.29593718e-01 5.03793895e-01 -5.43070510e-02 7.54460871e-01 -1.17782080e+00 -8.86526823e-01 -1.37516677e-01 -4.00770485e-01 -1.21112488e-01 5.80976307e-01 -4.55790520e-01 1.30111170e+00 -1.69901550e+00 9.35520470e-01 1.08682847e+00 5.33384144e-01 -1.38406649e-01 -1.40464580e+00 -3.73772442e-01 -2.98730791e-01 4.39980179e-01 -1.94246471e+00 -2.48462453e-01 6.23960197e-01 -2.17794374e-01 9.85804141e-01 6.34421408e-01 6.01342857e-01 5.23683906e-01 2.74663568e-01 6.67573571e-01 7.87398040e-01 -6.37587845e-01 2.57677771e-02 1.75456077e-01 1.78891525e-01 8.68691683e-01 4.20782506e-01 -1.78090204e-02 -7.06827402e-01 -6.13902986e-01 9.62579191e-01 4.54069018e-01 -1.32238999e-01 -1.07393062e+00 -1.34874511e+00 7.66500413e-01 3.34688574e-01 1.19090326e-01 -3.59577060e-01 3.19590211e-01 4.27764088e-01 4.76781607e-01 1.79928273e-01 3.58610660e-01 -4.54402268e-02 -8.97812694e-02 -1.22543967e+00 2.57885337e-01 9.35699761e-01 1.11908913e+00 8.48470092e-01 -3.17640394e-01 -2.83049017e-01 7.60604501e-01 -3.00021600e-02 1.13562799e+00 4.11968559e-01 -5.98670423e-01 1.37819275e-01 7.32652724e-01 7.41358623e-02 -1.27567267e+00 -1.15223182e-02 -2.82394327e-02 -6.27745211e-01 1.12812512e-01 7.06698895e-02 8.95076632e-01 -5.50385058e-01 1.13815689e+00 4.54894453e-02 -1.45334303e-01 -1.34737507e-01 7.25161552e-01 2.94125706e-01 4.70572263e-01 -5.19361556e-01 2.49031540e-02 1.49202132e+00 -1.02011418e+00 -4.10825402e-01 4.14549798e-01 2.75666982e-01 -1.18638825e+00 1.15403521e+00 4.24982458e-01 -1.03440225e+00 -3.94790828e-01 -1.17687774e+00 -2.04161480e-01 -6.43162131e-01 4.32412326e-01 6.93215966e-01 5.63470125e-01 -1.37496161e+00 5.08994579e-01 -5.17608941e-01 -3.99276197e-01 -2.82726679e-02 5.07790625e-01 -7.37851620e-01 -8.08292180e-02 -7.76542008e-01 7.33779907e-01 4.09951760e-03 -2.74014205e-01 -3.19543123e-01 -7.00479627e-01 -6.12073720e-01 3.28360438e-01 1.97666734e-01 -4.05074358e-01 8.62677693e-01 -3.17999512e-01 -1.30910194e+00 9.14155841e-01 -3.28071594e-01 -3.85094583e-01 5.88793159e-01 -3.86474162e-01 -2.24996433e-01 7.61248767e-01 1.40401702e-02 4.81885314e-01 1.27631247e+00 -7.78627634e-01 -1.70757875e-01 -3.79324436e-01 -8.58865976e-02 1.24061823e-01 -7.10397184e-01 7.21222684e-02 -1.01007807e+00 -7.42560744e-01 9.20895189e-02 -1.09387171e+00 -4.61089313e-02 5.54456353e-01 -1.90862715e-02 -1.06515490e-01 1.01538861e+00 -3.85525435e-01 1.34491146e+00 -2.20958424e+00 1.89940333e-01 6.61050498e-01 1.64821744e-01 3.03780288e-01 -2.89621025e-01 1.07365251e+00 1.24896251e-01 -2.37881735e-01 2.39788979e-01 -2.62150764e-01 1.64041296e-01 3.75664607e-02 -9.20587182e-01 5.70480645e-01 1.23987421e-01 1.12442446e+00 -6.13343418e-01 -5.19521356e-01 3.35970163e-01 6.90582275e-01 -3.64307463e-01 -1.37739435e-01 2.26204589e-01 -4.13321912e-01 -3.25643897e-01 8.37360859e-01 8.00574780e-01 -4.10008371e-01 5.88709228e-02 -3.75965625e-01 -7.66151175e-02 -2.24968180e-01 -1.18416762e+00 1.77091348e+00 -4.39630479e-01 6.01688087e-01 -1.30136445e-01 -6.44355476e-01 1.10022080e+00 2.63592359e-02 4.52754259e-01 -8.04848552e-01 -5.12359321e-01 6.12480998e-01 -8.23446274e-01 1.09542973e-01 9.02654290e-01 3.27402800e-01 -5.15116975e-02 4.55554545e-01 7.95711204e-02 -1.61327004e-01 2.21367106e-01 4.08662140e-01 1.03507149e+00 2.05291901e-02 2.52792716e-01 -4.05639201e-01 8.32617819e-01 -4.48558256e-02 -3.17419082e-01 7.84731090e-01 3.08897704e-01 5.28856158e-01 3.29154372e-01 -5.61271191e-01 -1.11514854e+00 -1.37086201e+00 -1.54541850e-01 1.07240272e+00 4.20651793e-01 -7.73529947e-01 -3.09535205e-01 -2.68327296e-01 4.81383145e-01 -3.82167064e-02 -3.02103370e-01 -3.27077024e-02 -5.18970072e-01 4.65292949e-03 6.84240997e-01 4.49049294e-01 3.13237906e-01 -5.76929152e-01 -8.09368253e-01 -1.61015585e-01 3.14600229e-01 -1.10730076e+00 -8.45531046e-01 -1.69247359e-01 -1.04101884e+00 -9.53053713e-01 -1.23217487e+00 -8.09075773e-01 7.68074691e-01 6.83977127e-01 9.76882756e-01 1.14006244e-01 -6.62932754e-01 9.21747625e-01 -2.03699604e-01 8.62008631e-02 6.94561750e-02 3.97019312e-02 2.22387478e-01 -1.86350904e-02 1.01624161e-01 -4.48076129e-01 -9.48469996e-01 4.61651623e-01 -1.25213456e+00 -3.56749177e-01 1.00303435e+00 1.16245615e+00 9.59858954e-01 -3.77089679e-01 1.00853257e-01 -3.91736925e-01 7.05092013e-01 1.24231078e-01 -1.04489005e+00 6.98368907e-01 -1.07442594e+00 4.68890786e-01 5.23715556e-01 -7.62688577e-01 -4.32766110e-01 3.00135523e-01 4.47987914e-01 -9.15566385e-01 4.31728095e-01 2.85024256e-01 4.52841640e-01 -9.21276867e-01 5.06515086e-01 6.71172559e-01 2.46595457e-01 -5.76644063e-01 6.40581846e-01 3.59451354e-01 5.12390316e-01 -6.27891183e-01 9.64302897e-01 6.04398191e-01 4.97756243e-01 -9.26950276e-01 1.10451475e-01 -9.90121901e-01 -7.24534988e-01 6.73399568e-02 1.35784149e-01 -6.58936977e-01 -8.63819957e-01 3.57398652e-02 -1.16040325e+00 3.87755483e-01 -2.31015056e-01 2.26832449e-01 -5.70608616e-01 7.33642757e-01 -5.15309870e-01 -7.38797963e-01 -6.43276334e-01 -1.13413286e+00 1.58178985e+00 -1.25946194e-01 9.65253711e-02 -5.65988362e-01 2.57237434e-01 -2.04620615e-01 6.67587042e-01 -1.87916487e-01 9.55315113e-01 -7.62960017e-01 -9.28509593e-01 -8.09812367e-01 -4.99967337e-01 1.01774976e-01 -3.84745061e-01 -4.76497635e-02 -5.80666780e-01 -4.93189514e-01 -4.29007888e-01 -1.31462708e-01 9.34298456e-01 -3.43771391e-02 6.82985008e-01 -1.14585742e-01 -5.55796981e-01 5.47456980e-01 1.74801672e+00 -1.31025031e-01 5.91049492e-01 1.99817166e-01 2.09651291e-01 4.24022377e-02 6.10620022e-01 4.24525797e-01 -8.39956179e-02 1.24874282e+00 1.48400530e-01 9.93035957e-02 -9.36212987e-02 -4.29404408e-01 2.68605500e-01 8.09777379e-01 -1.17777333e-01 4.53298986e-01 -6.01905942e-01 5.50559521e-01 -1.86598372e+00 -8.63925874e-01 1.33659795e-01 2.71132588e+00 4.03600097e-01 -3.05328131e-01 5.19453995e-02 -3.57349850e-02 3.92711788e-01 1.86491817e-01 -3.27627599e-01 -4.74201709e-01 -2.06384405e-01 7.68998265e-01 7.56475806e-01 4.67682242e-01 -9.33388472e-01 8.79304409e-01 7.07775068e+00 1.29230988e+00 -1.16416693e+00 -7.14319665e-03 -8.36784095e-02 -4.06009257e-02 -3.77507776e-01 9.64644030e-02 -7.71279097e-01 -5.23696840e-02 5.54231048e-01 -4.25801009e-01 6.25240445e-01 1.13431919e+00 -5.55606246e-01 -1.26230856e-02 -1.04140759e+00 1.32969809e+00 3.10197413e-01 -1.46864450e+00 4.99607533e-01 2.78235316e-01 5.01304805e-01 -2.32308149e-01 4.24908012e-01 1.60844639e-01 -5.49121976e-01 -7.91920543e-01 6.07117534e-01 6.94810867e-01 1.31639159e+00 -7.07750857e-01 2.98209965e-01 -1.13432556e-01 -1.50850201e+00 6.20808043e-02 -7.14768469e-01 2.80711085e-01 -1.45583391e-01 2.62954384e-01 -8.30462635e-01 5.37048340e-01 3.67170364e-01 1.90546572e-01 -8.70156169e-01 1.13651073e+00 3.32637727e-01 -1.57663956e-01 -5.63961864e-01 -2.75483608e-01 4.63585779e-02 -4.75184262e-01 5.93677163e-01 1.26627111e+00 6.30691051e-01 -4.65203404e-01 -6.22803122e-02 6.36152029e-01 -3.76564823e-03 6.30252719e-01 -1.10208285e+00 -1.77238896e-01 4.84667867e-01 1.39748847e+00 -7.12624669e-01 -3.22867453e-01 -2.82009721e-01 1.65954983e+00 1.93845361e-01 3.82127196e-01 -4.81008202e-01 -8.07955742e-01 4.09814060e-01 2.42259838e-02 7.77465999e-01 -4.93398100e-01 2.52108157e-01 -1.11596489e+00 4.33987647e-01 -6.41568780e-01 6.28329217e-02 -5.46877027e-01 -1.01580763e+00 5.08530021e-01 7.65257925e-02 -1.31322062e+00 -2.82172203e-01 -9.73464787e-01 -1.15881547e-01 7.86679089e-01 -1.33536184e+00 -1.32642531e+00 -7.10935146e-02 7.22317994e-01 2.44852543e-01 -2.45445281e-01 1.28708589e+00 4.66905981e-01 2.11940467e-01 9.10721123e-01 5.94069362e-01 -1.89443901e-01 1.00878179e+00 -1.06191206e+00 4.82784510e-01 3.47600341e-01 7.86609769e-01 1.00295126e+00 2.81315476e-01 -3.94784272e-01 -2.10380030e+00 -4.48008507e-01 8.99449348e-01 -3.89412820e-01 8.70054543e-01 -4.10051376e-01 -5.67776144e-01 2.75354326e-01 -1.17613532e-01 2.82167882e-01 5.01149058e-01 -1.13186397e-01 -1.00794208e+00 -4.48088259e-01 -1.00076640e+00 5.17293394e-01 8.36509407e-01 -1.06708014e+00 -2.11913943e-01 4.02835608e-01 6.86822653e-01 -3.02639514e-01 -9.92919922e-01 2.88680404e-01 1.18396986e+00 -8.62154067e-01 1.61248386e+00 -3.81963164e-01 4.99434629e-03 -2.09888592e-01 -4.75734025e-01 -6.51858568e-01 -2.84415841e-01 -6.45492136e-01 -2.82586008e-01 5.66700697e-01 1.25531465e-01 -6.68382406e-01 6.67547703e-01 3.52589607e-01 4.19078678e-01 -8.75846982e-01 -1.11638510e+00 -1.04152310e+00 -2.66793102e-01 -1.86272487e-01 4.95206773e-01 3.66230518e-01 5.70730157e-02 1.02219053e-01 -3.29001784e-01 -6.98621422e-02 5.96395254e-01 6.07263565e-01 6.83171570e-01 -1.30620444e+00 -3.68017346e-01 -6.10130668e-01 -1.00710845e+00 -1.30607581e+00 3.72848869e-03 -8.73857379e-01 -5.94705105e-01 -9.70400631e-01 2.60553956e-01 -2.31909290e-01 -4.21212673e-01 1.64467230e-01 2.53546417e-01 7.24324644e-01 4.56718445e-01 5.91268122e-01 -7.42654800e-01 2.85170019e-01 9.31721568e-01 -1.33170098e-01 5.52012436e-02 -6.44222349e-02 -1.82198375e-01 1.90402761e-01 2.34481767e-01 -7.32311606e-02 -3.40945065e-01 -1.37186468e-01 4.90888715e-01 7.82024190e-02 5.75302005e-01 -9.29980814e-01 5.76890767e-01 2.16422230e-01 2.50930786e-01 -9.04962361e-01 8.11691642e-01 -9.41657186e-01 1.78451091e-01 5.97931206e-01 -1.39526322e-01 6.09250009e-01 1.69761002e-01 8.27736437e-01 -4.08594340e-01 -2.57373869e-01 1.46249652e-01 2.41415769e-01 -6.14877939e-01 3.28965008e-01 -5.06356172e-02 -5.52627444e-01 7.99654067e-01 -1.72318295e-01 -2.90186435e-01 -3.80878210e-01 -3.27722073e-01 -3.32098931e-01 7.17957020e-01 2.13288426e-01 1.08697605e+00 -1.79143679e+00 -2.26084232e-01 4.27749813e-01 4.79081899e-01 -7.21475959e-01 -1.08742453e-01 9.53698754e-01 -7.96459079e-01 1.04063022e+00 -1.74190089e-01 -6.28615856e-01 -1.55613995e+00 8.25172305e-01 -2.01303408e-01 -5.73992431e-01 -5.45656562e-01 4.91497278e-01 5.11098746e-03 -1.02623031e-01 2.44105220e-01 -8.16335082e-02 2.74950683e-01 -5.54507747e-02 6.19259536e-01 3.59633535e-01 3.22134167e-01 -5.43248773e-01 -4.63024497e-01 1.04889226e+00 -9.67424810e-02 -3.97474259e-01 1.05319309e+00 1.73524737e-01 -3.68292600e-01 2.27656931e-01 1.56927979e+00 3.85317802e-01 -6.32056177e-01 -3.67280751e-01 -2.49406863e-02 -7.40498424e-01 -2.36986458e-01 -4.39709246e-01 -6.75162613e-01 8.85277808e-01 7.71520376e-01 1.37667045e-01 1.03331769e+00 -9.90897268e-02 6.98337972e-01 1.02327275e+00 7.76762307e-01 -7.90916324e-01 1.63734093e-01 4.47208583e-01 1.16802442e+00 -9.45737422e-01 2.33592138e-01 -2.73047149e-01 -2.79990613e-01 1.48450172e+00 -1.13941856e-01 -5.06779492e-01 7.00261652e-01 1.58247560e-01 -2.82987207e-01 -3.04771483e-01 -4.79215413e-01 2.55887192e-02 8.22825372e-01 3.45092833e-01 1.34630566e-02 -2.98220944e-02 -5.58370948e-01 6.49932399e-02 1.80585340e-01 -1.14829950e-01 -1.91622525e-01 8.33918571e-01 -3.33059192e-01 -1.52697754e+00 -4.71826226e-01 2.49135002e-01 -2.13888377e-01 -3.87679428e-01 -4.99475151e-01 6.92186177e-01 -7.44432271e-01 3.18541616e-01 -1.79139227e-02 -3.39323878e-01 4.54619199e-01 2.45955169e-01 7.31806457e-01 -1.06615864e-01 -4.32187885e-01 2.93028802e-02 -3.95840108e-01 -9.60402608e-01 -2.00602952e-02 -1.79900885e-01 -8.09433103e-01 -1.59750074e-01 -4.66188014e-01 -1.95337050e-02 9.71325696e-01 3.54982138e-01 8.65203977e-01 -3.71750563e-01 5.55061042e-01 -9.61094916e-01 -1.12245083e+00 -4.85821575e-01 -5.57930529e-01 4.74289209e-01 1.73880428e-01 -7.87325859e-01 -1.13353014e-01 -2.49279588e-01]	[10.684209823608398, 0.3617570698261261]
2b8f3b84-6a17-4b26-815b-ca14ed3d24f4	bci-breast-cancer-immunohistochemical-image	2204.11425	null	https://arxiv.org/abs/2204.11425v2	https://arxiv.org/pdf/2204.11425v2.pdf	BCI: Breast Cancer Immunohistochemical Image Generation through Pyramid Pix2pix	The evaluation of human epidermal growth factor receptor 2 (HER2) expression is essential to formulate a precise treatment for breast cancer. The routine evaluation of HER2 is conducted with immunohistochemical techniques (IHC), which is very expensive. Therefore, for the first time, we propose a breast cancer immunohistochemical (BCI) benchmark attempting to synthesize IHC data directly with the paired hematoxylin and eosin (HE) stained images. The dataset contains 4870 registered image pairs, covering a variety of HER2 expression levels. Based on BCI, as a minor contribution, we further build a pyramid pix2pix image generation method, which achieves better HE to IHC translation results than the other current popular algorithms. Extensive experiments demonstrate that BCI poses new challenges to the existing image translation research. Besides, BCI also opens the door for future pathology studies in HER2 expression evaluation based on the synthesized IHC images. BCI dataset can be downloaded from https://bupt-ai-cz.github.io/BCI.	['Mulan Jin', 'Zhongyue Shi', 'Xinyu Jia', 'Feng Xu', 'Chuang Zhu', 'ShengJie Liu']	2022-04-25	null	null	null	null	['breast-cancer-detection', 'breast-cancer-detection', 'breast-cancer-histology-image-classification', 'medical-image-generation', 'classification-of-breast-cancer-histology']	['knowledge-base', 'medical', 'medical', 'medical', 'medical']	[ 2.83171326e-01 5.40460497e-02 -5.45526028e-01 -3.64973426e-01 -1.11782157e+00 -3.89454782e-01 1.24892503e-01 2.42681310e-01 -2.48630315e-01 7.59093523e-01 -1.14259213e-01 -4.74660695e-01 3.78039777e-01 -9.41043794e-01 -5.77912688e-01 -9.67934489e-01 3.57958853e-01 5.30976355e-01 -8.10298324e-03 -2.79900819e-01 -2.87486047e-01 7.73870587e-01 -1.07848871e+00 4.34724092e-01 3.85324299e-01 9.71048832e-01 1.61771804e-01 6.65533841e-01 -8.96193366e-03 5.96313775e-01 -1.96356013e-01 -5.89208245e-01 1.51992470e-01 -4.42082644e-01 -7.40851045e-01 -3.37724030e-01 1.70205593e-01 -2.00776368e-01 -9.74582657e-02 1.27328002e+00 5.50139189e-01 -7.14083731e-01 6.67783082e-01 -1.41215301e+00 -3.54996830e-01 3.11815709e-01 -8.82829845e-01 -4.73562889e-02 3.52164470e-02 1.33900955e-01 8.59419882e-01 -9.21744943e-01 9.89403307e-01 6.85327053e-01 5.81952274e-01 6.41015232e-01 -1.32625484e+00 -8.76102626e-01 -7.15738297e-01 4.46938246e-01 -1.76156187e+00 -1.20662220e-01 6.22083664e-01 -4.15145099e-01 2.84775406e-01 6.27634108e-01 9.33255851e-01 7.47006178e-01 1.66409150e-01 8.88277054e-01 1.51644659e+00 -5.10174692e-01 3.13638039e-02 2.03388542e-01 -7.26617351e-02 8.96730304e-01 1.28170997e-01 -1.15653411e-01 -2.57823139e-01 8.78227726e-02 8.48831713e-01 1.98167473e-01 -3.00379753e-01 -2.76095401e-02 -1.22700191e+00 6.34541810e-01 6.16406798e-01 6.49268746e-01 -8.55299607e-02 1.31794512e-02 4.28644747e-01 3.54993075e-01 1.07724674e-01 8.40453207e-02 -1.60775229e-01 3.46002012e-01 -7.43153393e-01 7.77797624e-02 2.18912601e-01 7.43300021e-01 1.12721634e+00 -8.21346402e-01 -1.46142676e-01 7.31323302e-01 2.20195130e-01 7.56179154e-01 4.03767169e-01 -6.34240806e-01 -3.63639258e-02 6.89135730e-01 -2.53692508e-01 -1.18768418e+00 -5.21144629e-01 2.27284618e-02 -1.23670542e+00 2.24174872e-01 4.68645662e-01 2.86145389e-01 -5.20950854e-01 1.48258817e+00 3.47652525e-01 -1.04289345e-01 8.55992064e-02 9.29023504e-01 9.89283085e-01 2.49375939e-01 2.63045996e-01 -4.14577037e-01 1.67866313e+00 -7.28132665e-01 -8.87043536e-01 4.10259575e-01 9.78235483e-01 -7.21748173e-01 1.00940728e+00 1.05874702e-01 -7.31976032e-01 -8.11790973e-02 -1.02691662e+00 -2.69034356e-01 -5.72401822e-01 3.23631704e-01 9.11194384e-01 4.32071656e-01 -1.08924937e+00 4.74429578e-02 -8.67045879e-01 -5.82407415e-01 7.16657400e-01 2.97671288e-01 -8.31813216e-01 -1.17701203e-01 -9.77896571e-01 7.15043783e-01 2.20367134e-01 6.80190846e-02 -3.67581367e-01 -9.36198831e-01 -5.46605289e-01 -3.66811365e-01 -1.78122640e-01 -5.79888701e-01 1.12815034e+00 -9.08638656e-01 -1.39414716e+00 1.42247522e+00 -3.48101556e-01 -8.76727924e-02 4.04651701e-01 7.39890695e-01 -2.74343640e-01 3.82841885e-01 6.30744472e-02 1.02956700e+00 8.19259137e-02 -1.04665124e+00 -7.33777463e-01 -6.80543184e-01 -4.86151814e-01 -2.21547186e-02 -4.04081345e-01 -9.08564106e-02 -7.10659385e-01 -5.30673027e-01 -1.26561835e-01 -9.89095330e-01 6.11618534e-03 4.15671319e-01 -3.41015697e-01 -3.94761935e-02 5.53930998e-01 -7.05972731e-01 1.09194076e+00 -2.25907254e+00 -2.87999194e-02 5.47035038e-01 2.49691233e-01 7.32318088e-02 -1.93018224e-02 7.99824968e-02 -2.59019852e-01 3.34530532e-01 -6.51299953e-02 -2.65996661e-02 4.18464467e-03 -9.72649604e-02 7.55907446e-02 6.03606462e-01 -2.70443745e-02 1.47295892e+00 -7.81126440e-01 -1.06262970e+00 7.83355441e-03 4.44635302e-01 -1.05704166e-01 1.34829804e-02 1.78852618e-01 4.77875561e-01 -4.94332343e-01 1.07197559e+00 7.56120086e-01 -4.81451184e-01 3.68901432e-01 -6.05737329e-01 1.46114528e-01 -5.35627484e-01 -4.43217099e-01 1.48689592e+00 -6.26772568e-02 6.40911758e-01 1.19205073e-01 -7.83784270e-01 8.84823442e-01 2.67901152e-01 6.99714184e-01 -9.43178833e-01 2.40107253e-01 4.21136409e-01 -3.06502223e-01 -4.46688443e-01 -1.46422982e-01 -5.93656659e-01 1.91971228e-01 -8.75378698e-02 -3.66412044e-01 -2.34453917e-01 3.44925106e-01 -9.78268962e-03 1.22029901e+00 -3.97166550e-01 5.25663972e-01 -1.29448488e-01 6.46407366e-01 3.12893033e-01 5.08670628e-01 4.94884774e-02 -5.32432675e-01 5.76759696e-01 5.87622762e-01 -4.43365425e-01 -1.05829024e+00 -9.05744076e-01 -5.13203323e-01 3.99917245e-01 2.38075927e-01 -4.16955352e-02 -6.05073631e-01 -5.68760514e-01 -5.04035130e-02 9.00539011e-02 -9.63251948e-01 3.67447346e-01 -1.75276637e-01 -1.00426435e+00 8.64609301e-01 4.03601974e-01 4.74818170e-01 -8.29356909e-01 -2.00075835e-01 -1.62475049e-01 -4.12603229e-01 -7.86879957e-01 -2.66144484e-01 -3.52883376e-02 -7.24864006e-01 -1.32627821e+00 -9.48865950e-01 -1.08240783e+00 9.75107193e-01 1.78636923e-01 8.87054861e-01 4.79186356e-01 -7.21672833e-01 -4.11494523e-02 -4.35244858e-01 -5.01409471e-01 -5.85155368e-01 -1.99320521e-02 -3.30613792e-01 -5.04016057e-02 6.11406088e-01 -3.19254845e-01 -7.35876977e-01 3.07153881e-01 -1.11656392e+00 6.16541386e-01 1.12332416e+00 8.89502943e-01 1.41056848e+00 1.33208498e-01 3.65032464e-01 -1.05701041e+00 1.08124681e-01 -1.70899808e-01 -5.00593781e-01 5.32826960e-01 -4.10044044e-01 -4.28911030e-01 3.95655006e-01 -2.79150326e-02 -8.32218647e-01 2.29098290e-01 -3.44715148e-01 -1.85351625e-01 -2.60506541e-01 5.24076462e-01 -9.14716870e-02 -2.41356075e-01 5.52471757e-01 -5.73063120e-02 4.08038527e-01 -2.25307718e-02 -2.63178386e-02 6.80353343e-01 7.32969701e-01 -2.29710694e-02 6.98440611e-01 8.15441489e-01 3.51066023e-01 -7.20468283e-01 -4.96188313e-01 -5.81068397e-01 -3.04653704e-01 -1.82848543e-01 9.02168512e-01 -8.65470111e-01 -6.25235498e-01 3.62233579e-01 -8.22560728e-01 -5.99929690e-01 6.81855679e-02 4.39761877e-01 -4.86124784e-01 1.82472393e-01 -9.32856083e-01 -5.06152282e-04 -4.45045650e-01 -1.12841165e+00 1.16984010e+00 1.89461663e-01 -1.86598957e-01 -8.30424428e-01 3.66405100e-01 4.97429311e-01 1.48358688e-01 5.38436472e-01 1.05689538e+00 -1.28112346e-01 -2.36593291e-01 -4.31488395e-01 -6.62598073e-01 -8.12046975e-02 1.81721523e-01 3.82912815e-01 -8.28489661e-01 -8.42081308e-02 -3.01454216e-01 -2.72979677e-01 5.85630238e-01 2.91699797e-01 1.23714280e+00 -1.71128258e-01 -6.61381841e-01 7.87331998e-01 1.69300115e+00 -1.86610352e-02 1.25015306e+00 5.21334529e-01 4.56260413e-01 5.07481217e-01 9.27276850e-01 1.76950544e-01 2.91855693e-01 6.66583359e-01 2.58132607e-01 -8.99614692e-01 -3.23690891e-01 -2.17881724e-01 -1.05243497e-01 5.74657559e-01 -1.19971327e-01 8.51482078e-02 -9.62189615e-01 4.94012505e-01 -1.53380060e+00 -8.69949877e-01 -3.55599225e-01 1.67540669e+00 1.26876652e+00 -5.26662707e-01 -1.11621268e-01 1.27838150e-01 7.39271283e-01 -5.24973452e-01 -3.84235620e-01 1.80002019e-01 -4.00691539e-01 3.07090878e-01 5.95265627e-01 3.53557557e-01 -9.59674239e-01 6.47540092e-01 5.79883623e+00 1.16747642e+00 -1.57013619e+00 1.26709312e-01 1.17269576e+00 1.18457213e-01 -3.36810112e-01 -2.56693721e-01 -5.42786658e-01 2.49415904e-01 4.94930804e-01 -2.26037815e-01 7.83524066e-02 5.99037290e-01 1.99532226e-01 1.96848344e-02 -9.72773671e-01 1.23876536e+00 -1.12381727e-01 -1.49654281e+00 7.60951824e-03 3.86758119e-01 6.30793452e-01 -4.62035015e-02 4.86247577e-02 9.23041031e-02 -9.96223092e-02 -1.11414742e+00 1.70712218e-01 6.10985875e-01 1.52273571e+00 -6.06461108e-01 1.40935838e+00 -6.59607947e-02 -9.61454511e-01 4.37772006e-01 -2.80509531e-01 5.01587570e-01 -1.72469065e-01 7.11965322e-01 -1.27304971e+00 3.91733438e-01 7.57537007e-01 6.04488194e-01 -8.95270944e-01 6.19514346e-01 -1.96596488e-01 3.52824301e-01 -1.94467992e-01 -7.62436464e-02 -2.31664836e-01 -8.37304518e-02 -9.37829688e-02 1.20285547e+00 4.97027814e-01 4.01419014e-01 -6.73241094e-02 6.37702107e-01 -1.30497262e-01 6.88791633e-01 -4.48492110e-01 -1.28768608e-01 2.67713934e-01 1.83422410e+00 -9.43047464e-01 -1.98674560e-01 -3.86893004e-01 8.36291969e-01 2.20839575e-01 3.21483493e-01 -9.57877755e-01 -2.83567190e-01 2.80782104e-01 3.50972712e-01 -1.13407962e-01 5.29537320e-01 -4.17413473e-01 -8.57187629e-01 -2.94062227e-01 -9.46656883e-01 4.61347342e-01 -7.67974854e-01 -1.26479876e+00 2.58175075e-01 -3.64341766e-01 -1.26277423e+00 1.95908591e-01 -8.14671695e-01 -2.84489036e-01 4.97511983e-01 -1.72652972e+00 -1.58904850e+00 -7.08661199e-01 5.00656426e-01 -4.91801091e-02 -2.04767380e-02 1.07059407e+00 4.70377862e-01 -6.99173868e-01 8.76023173e-01 -6.43303022e-02 1.80999428e-01 1.00673890e+00 -1.07074273e+00 -4.47970271e-01 1.91015303e-01 -3.81816655e-01 3.69816542e-01 4.32652682e-01 -5.29030681e-01 -1.43440485e+00 -1.33065915e+00 8.77376914e-01 -2.93781936e-01 7.24650860e-01 1.04909778e-01 -6.18553221e-01 6.70169830e-01 1.06655173e-01 1.99365795e-01 1.26046944e+00 -5.23159862e-01 -6.06008992e-02 -3.18502903e-01 -1.45821154e+00 7.47763395e-01 3.99387091e-01 -3.73034090e-01 2.81232774e-01 4.29276705e-01 2.60603189e-01 -2.36984283e-01 -1.43188226e+00 5.95579147e-01 7.42121160e-01 -8.48569095e-01 6.94014788e-01 6.41296655e-02 5.73068202e-01 -7.54790962e-01 -2.95640111e-01 -7.64157474e-01 -2.74835885e-01 -1.65542752e-01 4.66713130e-01 1.02045465e+00 5.66128135e-01 -5.40719569e-01 9.40053463e-01 3.40882450e-01 1.67828441e-01 -8.91432822e-01 -8.49834621e-01 -2.99756438e-01 3.55533846e-02 -1.33007854e-01 6.65129304e-01 9.02306437e-01 4.50479686e-01 4.00670655e-02 2.17779174e-01 2.60976776e-02 5.38208842e-01 -3.04428190e-02 9.47387338e-01 -8.82673502e-01 -2.72442669e-01 -7.76501775e-01 -8.41124892e-01 -3.06191504e-01 7.27608334e-03 -1.24050772e+00 6.83951937e-03 -1.17973971e+00 8.41176629e-01 -5.44184506e-01 -4.08366263e-01 9.06641543e-01 -8.46548900e-02 1.14725029e+00 -1.57584444e-01 4.49839652e-01 -4.69722986e-01 1.81095108e-01 1.55379140e+00 -4.95060325e-01 2.02760786e-01 -5.23829222e-01 -7.63567209e-01 5.14408052e-01 1.00509465e+00 -2.61262745e-01 4.67632785e-02 5.84888533e-02 2.77654260e-01 1.46371514e-01 4.09077823e-01 -7.02142119e-01 1.15016833e-01 -3.44855964e-01 4.99898195e-01 -7.67932117e-01 1.85133517e-02 -8.44562888e-01 6.58607066e-01 5.82190931e-01 -2.33490869e-01 -1.19300157e-01 1.13580309e-01 2.45497033e-01 -4.81062710e-01 3.81874293e-02 9.32293355e-01 -1.03458306e-02 -4.97967929e-01 4.30588990e-01 -1.11249678e-01 -5.55331886e-01 1.38675380e+00 -2.30805770e-01 -3.91397446e-01 -5.09409346e-02 -5.57177067e-01 2.58322984e-01 9.90751386e-01 -2.23471165e-01 5.67770123e-01 -1.60816467e+00 -7.53658116e-01 7.56361112e-02 6.74048960e-01 -2.15322729e-02 3.84475142e-01 1.44907427e+00 -1.08570516e+00 3.12430382e-01 -3.51618588e-01 -5.89778662e-01 -1.72566676e+00 5.35648882e-01 1.93694785e-01 -5.81177115e-01 -4.05609488e-01 6.69535816e-01 4.79936987e-01 -4.16793883e-01 -2.27048378e-02 4.85288464e-02 -2.33367428e-01 -1.22416787e-01 6.56065047e-01 1.20324865e-01 1.82731390e-01 -6.59520686e-01 -2.17293516e-01 5.03594577e-01 -2.35753716e-03 -5.12928329e-02 1.23845136e+00 -7.36920908e-03 -6.69829905e-01 1.36094227e-01 1.49429810e+00 -8.91957954e-02 -4.91587996e-01 2.16733992e-01 -2.69525379e-01 -3.98287088e-01 6.47028014e-02 -7.67161250e-01 -1.24303138e+00 4.77259994e-01 9.32632744e-01 -2.61076838e-01 1.44659603e+00 1.15752749e-01 7.34223127e-01 1.32441759e-01 3.71851087e-01 -8.47079813e-01 -2.12517098e-01 -2.58437008e-01 8.40538800e-01 -1.32463455e+00 -2.42399219e-02 -6.99432373e-01 -4.67702955e-01 1.19152665e+00 4.04933840e-01 2.29222327e-01 5.21412194e-01 5.41025579e-01 5.19776285e-01 -3.31991106e-01 -5.83009958e-01 -7.43515119e-02 -1.30626380e-01 5.85101724e-01 6.64685607e-01 4.17863816e-01 -7.28002965e-01 4.58218455e-01 -1.98022559e-01 3.95515263e-01 3.28658015e-01 6.34581625e-01 -3.32469493e-02 -1.23021877e+00 -3.42863739e-01 3.66329759e-01 -5.70653796e-01 1.47551060e-01 -5.03500283e-01 9.67877090e-01 1.57377675e-01 4.25406635e-01 -1.79539323e-02 -3.14404905e-01 1.07883438e-01 -1.86243623e-01 4.46954042e-01 -1.88269794e-01 -9.27400663e-02 2.47728080e-01 -3.20704877e-01 -3.19367230e-01 -6.24016881e-01 -5.23859441e-01 -1.59115851e+00 -3.04579645e-01 -1.93752766e-01 5.36142588e-02 8.51792097e-01 4.34153795e-01 8.20678324e-02 4.03705537e-01 5.28639376e-01 -4.87927049e-01 8.88512135e-02 -8.21782053e-01 -8.80628049e-01 6.94237113e-01 7.21794739e-02 -3.52361619e-01 -1.84693560e-01 4.87339556e-01]	[15.040985107421875, -3.0660312175750732]
89ac9376-3696-40fd-9632-32ee825d5eac	convolutional-proteinunetlm-competitive-with	null	null	https://doi.org/10.1002/prot.26452	https://onlinelibrary.wiley.com/doi/epdf/10.1002/prot.26452	Convolutional ProteinUnetLM competitive with long short-term memory-based protein secondary structure predictors	The protein secondary structure (SS) prediction plays an important role in the characterization of general protein structure and function. In recent years, a new generation of algorithms for SS prediction based on embeddings from protein language models (pLMs) is emerging. These algorithms reach state-of-the-art accuracy without the need for time-consuming multiple sequence alignment (MSA) calculations. Long short-term memory (LSTM)-based SPOT-1D-LM and NetSurfP-3.0 are the latest examples of such predictors. We present the ProteinUnetLM model using a convolutional Attention U-Net architecture that provides prediction quality and inference times at least as good as the best LSTM-based models for 8-class SS prediction (SS8). Additionally, we address the issue of the heavily imbalanced nature of the SS8 problem by extending the loss function with the Matthews correlation coefficient, and by proper assessment using previously introduced adjusted geometric mean (AGM) metric. ProteinUnetLM achieved better AGM and sequence overlap score than LSTM-based predictors, especially for the rare structures 310-helix (G), beta-bridge (B), and high curvature loop (S). It is also competitive on challenging datasets without homologs, free-modeling targets, and chameleon sequences. Moreover, ProteinUnetLM outperformed its previous MSA-based version ProteinUnet2, and provided better AGM than AlphaFold2 for 1/3 of proteins from the CASP14 dataset, proving its potential for making a significant step forward in the domain. To facilitate the usage of our solution by protein scientists, we provide an easy-to-use web interface under https://biolib.com/SUT/ProteinUnetLM/.	['Katarzyna Stapor', 'Irena Roterman', 'Piotr Fabian', 'Krzysztof Kotowski']	2022-11-30	null	null	null	proteins-2022-11	['unet-segmentation', 'protein-secondary-structure-prediction', 'multiple-sequence-alignment']	['computer-vision', 'medical', 'medical']	[ 3.08761269e-01 -3.76349911e-02 -3.09400141e-01 -3.39367390e-01 -7.43862092e-01 -3.00565690e-01 1.52071163e-01 7.55251884e-01 -5.24419785e-01 1.11446500e+00 -1.40370384e-01 -8.11840296e-01 1.21037915e-01 -2.88474828e-01 -1.30718887e+00 -1.00816512e+00 -2.09190205e-01 6.18259251e-01 9.12846848e-02 -2.96860278e-01 1.26345515e-01 3.94294858e-01 -1.16017079e+00 4.85978097e-01 1.02894163e+00 8.58940125e-01 5.97102344e-01 3.87140751e-01 -3.45454931e-01 5.28520703e-01 -1.99309751e-01 -3.62208694e-01 -1.25851631e-01 -2.78056473e-01 -8.78851473e-01 -6.41269743e-01 3.07592303e-01 1.13060899e-01 1.88033655e-01 7.68766105e-01 6.79150164e-01 -6.90548792e-02 6.85866237e-01 -8.10549140e-01 -7.71584392e-01 1.63152561e-01 -7.13286579e-01 3.61041352e-02 3.13708156e-01 3.76661003e-01 1.30168915e+00 -1.20189548e+00 8.68793428e-01 1.11501384e+00 8.77325416e-01 4.48372722e-01 -1.54368222e+00 -3.87329936e-01 -1.87363431e-01 6.96360886e-01 -9.05221462e-01 2.28778124e-01 1.91447109e-01 -4.08287942e-01 1.95456862e+00 -6.19390383e-02 4.02374893e-01 1.39004660e+00 8.30961108e-01 8.74089539e-01 8.03806543e-01 -2.71994829e-01 8.03880300e-03 -3.89423281e-01 3.65751594e-01 7.45731592e-01 8.70163664e-02 -1.10250898e-01 -5.14903486e-01 -6.79474533e-01 3.08997273e-01 1.22736879e-01 -3.62050503e-01 -6.08475983e-01 -1.25069189e+00 8.69391561e-01 3.69880974e-01 2.08650678e-01 -5.62351823e-01 -1.55388594e-01 5.79764962e-01 3.37446809e-01 3.96785319e-01 5.48710525e-01 -1.13651025e+00 -2.96218395e-01 -6.43017471e-01 2.03839034e-01 6.28505170e-01 4.36592728e-01 5.62132835e-01 -2.92129368e-01 2.47380018e-01 8.67377877e-01 2.31662452e-01 6.30801693e-02 7.15152323e-01 -3.49233538e-01 1.77404538e-01 6.57770813e-01 1.85588941e-01 -6.08204067e-01 -5.56096077e-01 -2.83230335e-01 -8.63930523e-01 1.68615326e-01 4.15569395e-01 2.07048625e-01 -7.47772813e-01 1.70477629e+00 2.03841910e-01 -4.99418899e-02 8.13613534e-02 7.29838371e-01 6.88554525e-01 8.73256445e-01 3.51450741e-01 -3.27363700e-01 1.20394170e+00 -8.86604726e-01 -4.13842648e-01 8.41347352e-02 9.25202012e-01 -6.91804826e-01 1.04183030e+00 4.18499410e-01 -7.34038770e-01 -5.63264966e-01 -1.23637974e+00 -2.89525360e-01 -5.17336428e-01 5.85921633e-04 3.61231983e-01 -6.11204579e-02 -7.38810897e-01 1.12346685e+00 -9.73515630e-01 -5.70553601e-01 1.81437552e-01 5.58670461e-01 -7.33680367e-01 1.98156625e-01 -1.31974232e+00 1.50531363e+00 5.07339954e-01 -2.33600229e-01 -5.36615908e-01 -8.99644554e-01 -6.47360563e-01 7.56837055e-02 -2.06421241e-01 -4.19631362e-01 9.70081091e-01 -7.84858942e-01 -1.17719316e+00 8.65496159e-01 -4.24430251e-01 -8.97542119e-01 2.23510996e-01 -5.29182434e-01 -9.25067663e-02 -1.20655254e-01 -2.65889287e-01 6.49977267e-01 2.06665337e-01 -7.03919232e-01 -2.51281291e-01 -3.89922172e-01 -3.77200395e-01 4.32979986e-02 1.50019363e-01 1.27267197e-01 3.98587257e-01 -3.61870408e-01 -6.29701838e-02 -7.62443006e-01 -4.08270866e-01 2.03200147e-01 -1.84288189e-01 -5.19889057e-01 6.07868195e-01 -9.01592314e-01 8.64538431e-01 -1.73689818e+00 3.35098356e-01 -4.53325361e-02 2.89348811e-01 8.78514588e-01 -3.38446498e-01 8.43288958e-01 -6.24886692e-01 -8.84288773e-02 -3.86876971e-01 7.75517970e-02 -2.26705059e-01 -5.01333084e-03 -1.38958365e-01 6.53952360e-01 4.10821915e-01 1.01920307e+00 -7.95809865e-01 -7.76816979e-02 2.68681198e-01 7.55644143e-01 -3.14674884e-01 2.77312011e-01 -6.14899516e-01 2.87533939e-01 -2.78221481e-02 4.89184141e-01 7.13608444e-01 -5.48655629e-01 5.35416663e-01 -2.07485095e-01 -1.52187437e-01 3.06462437e-01 -4.09823447e-01 1.54693294e+00 -2.24155374e-02 4.17922556e-01 -3.17079991e-01 -9.99358773e-01 1.16277707e+00 1.79192901e-01 6.31960928e-01 -4.16146338e-01 -1.89956687e-02 3.88975143e-01 4.84849125e-01 -3.40180516e-01 8.75937380e-03 -4.00831029e-02 5.35205603e-01 2.87754953e-01 2.71087319e-01 6.64009094e-01 1.30099058e-03 3.45798060e-02 1.07388937e+00 6.47557080e-01 6.18772388e-01 -3.47543538e-01 4.92925763e-01 -2.48080511e-02 7.55472720e-01 1.77771106e-01 -2.25600228e-01 4.42697555e-01 5.98034441e-01 -8.92860234e-01 -1.29700184e+00 -8.47115576e-01 -3.32154304e-01 1.12674916e+00 -2.24564001e-01 -5.81993639e-01 -6.76511168e-01 -5.49074531e-01 1.32986426e-01 5.54766655e-01 -3.98236334e-01 -2.33081564e-01 -5.35518110e-01 -1.27125859e+00 4.92496163e-01 3.75687659e-01 4.51174639e-02 -1.22980583e+00 -4.84133482e-01 6.41856313e-01 -3.14562112e-01 -5.65409839e-01 -2.76776314e-01 8.97375643e-01 -7.99109578e-01 -1.38305998e+00 -8.61539543e-01 -9.00118768e-01 2.32499480e-01 -8.46428052e-02 9.34985697e-01 -2.60458887e-01 -2.41314650e-01 -5.41643083e-01 -2.28812024e-01 -4.18439597e-01 -4.59924221e-01 -3.98493372e-02 5.32095432e-01 -3.09235126e-01 1.11173034e+00 -7.03256905e-01 -5.80441594e-01 2.42986321e-01 -5.57459116e-01 2.12697968e-01 7.38260746e-01 1.22620320e+00 8.86403322e-01 -6.77466810e-01 7.62753606e-01 -5.56163132e-01 4.97256339e-01 -5.73792040e-01 -4.24153566e-01 4.00717914e-01 -8.88989151e-01 2.98733979e-01 8.16538215e-01 -4.53759044e-01 -5.87108135e-01 1.83713868e-01 -5.90345263e-01 -4.18754995e-01 -1.00543521e-01 5.82276165e-01 -1.72011644e-01 -9.48092118e-02 6.56265438e-01 5.71161628e-01 4.51516718e-01 -8.48515511e-01 1.04023434e-01 6.20244622e-01 1.33090869e-01 -4.48873818e-01 8.61077309e-02 -1.74051568e-01 -4.67592068e-02 -8.56640875e-01 -5.30923724e-01 -3.90619099e-01 -6.08903646e-01 3.32688510e-01 8.08211446e-01 -6.29942298e-01 -1.11144185e+00 5.78556716e-01 -1.28095484e+00 -2.39117980e-01 3.28681558e-01 3.59523088e-01 -6.65340662e-01 8.81764650e-01 -9.47116077e-01 -3.99369210e-01 -8.00080299e-01 -1.35204422e+00 7.12504923e-01 -2.35082582e-01 -4.77559745e-01 -6.44768000e-01 2.20115602e-01 3.88103157e-01 2.93016255e-01 3.83883446e-01 1.45645142e+00 -1.16226089e+00 -2.59844899e-01 1.22619756e-01 -2.17501625e-01 4.60460156e-01 -8.91853124e-02 8.62629786e-02 -5.63309312e-01 -3.38800222e-01 -3.51285815e-01 -4.50467944e-01 1.05360472e+00 5.57421803e-01 9.52063859e-01 -1.65400743e-01 -3.37552249e-01 5.98951161e-01 1.40798438e+00 2.65894175e-01 5.41557968e-01 6.38943374e-01 6.80438638e-01 4.29136992e-01 7.05871522e-01 1.75050706e-01 5.03088813e-03 6.71488881e-01 6.20342433e-01 -1.44405231e-01 3.68884325e-01 -3.25886726e-01 6.55536592e-01 6.61897182e-01 -2.84271501e-02 -2.45992303e-01 -9.81295526e-01 3.17606002e-01 -2.06615806e+00 -8.46678853e-01 -3.64481777e-01 2.09143066e+00 9.64870870e-01 1.11275114e-01 8.37709606e-02 -1.76806837e-01 5.69748104e-01 -6.52053766e-03 -9.89224076e-01 -8.65580082e-01 -3.84141237e-01 3.24336112e-01 5.23490846e-01 3.19067985e-01 -1.05489624e+00 9.38942134e-01 5.90740967e+00 1.02258909e+00 -1.08150661e+00 -6.87097386e-02 6.89415097e-01 1.02993496e-01 -3.98974568e-02 -1.49111450e-01 -8.91219914e-01 6.84518397e-01 1.37155509e+00 1.45079523e-01 -6.86474591e-02 9.23656404e-01 4.22573686e-01 1.42217785e-01 -1.02477062e+00 7.28175521e-01 -3.41415226e-01 -1.61244452e+00 -2.34889295e-02 1.91815019e-01 2.62708694e-01 7.64327645e-01 -1.15613870e-01 1.47796467e-01 2.14448661e-01 -1.21584487e+00 2.32589375e-02 3.28768224e-01 9.06734526e-01 -8.80374968e-01 1.04781878e+00 3.50505173e-01 -8.00485671e-01 2.06901565e-01 -7.31035352e-01 1.04564838e-01 1.32408708e-01 7.26688325e-01 -8.92811060e-01 3.82871509e-01 6.45883083e-01 7.95620024e-01 -1.28067017e-01 7.97243953e-01 1.04050837e-01 6.12956166e-01 -2.68998682e-01 -2.30655462e-01 3.60811234e-01 -5.06457508e-01 4.87225354e-01 1.14568973e+00 -4.02090419e-03 -1.06347166e-01 1.22696757e-01 7.14779556e-01 -2.44993735e-02 4.78093505e-01 -2.50738770e-01 -1.81168392e-01 2.01779902e-02 8.98797452e-01 -1.83652222e-01 -1.42798632e-01 -4.59938884e-01 9.79401469e-01 5.71920633e-01 1.72981143e-01 -8.67834926e-01 -3.53580475e-01 1.28335464e+00 1.17319480e-01 3.17238837e-01 -8.09375942e-03 1.99090704e-01 -1.03921700e+00 -6.22978210e-02 -1.16823876e+00 2.41454616e-01 -5.75366259e-01 -1.42660284e+00 6.29413247e-01 -5.55838466e-01 -1.05797040e+00 1.07741263e-02 -1.04114842e+00 -3.60678107e-01 9.56835747e-01 -1.43333745e+00 -1.16416204e+00 6.07287169e-01 -1.53590620e-01 5.64660490e-01 -3.03615957e-01 1.25466108e+00 1.57059386e-01 -5.61917782e-01 5.49684703e-01 9.03677821e-01 -3.56736332e-01 9.56629455e-01 -1.24197185e+00 8.36391211e-01 3.60732526e-01 -2.18424901e-01 9.09167826e-01 1.05276644e+00 -6.86349273e-01 -1.08052373e+00 -1.11425400e+00 1.25952077e+00 -4.03590828e-01 7.49334395e-01 -1.78565562e-01 -1.29568148e+00 6.10064805e-01 1.98532231e-02 -3.75892103e-01 1.08525836e+00 5.53704686e-02 -4.61355835e-01 3.76380950e-01 -1.17187667e+00 3.37594956e-01 7.87094474e-01 -2.68840969e-01 -6.90187156e-01 4.89846408e-01 9.28116202e-01 -1.50273666e-01 -1.07221174e+00 6.39913201e-01 6.73729122e-01 -1.00327241e+00 1.02297616e+00 -1.07975042e+00 5.96239448e-01 -2.26863444e-01 -2.09497303e-01 -1.06519079e+00 -5.32626987e-01 -2.70844609e-01 -3.93254340e-01 5.60644507e-01 6.47880495e-01 -7.50703752e-01 8.04987013e-01 8.56693089e-02 -2.35846102e-01 -1.50665212e+00 -9.50403333e-01 -5.63429415e-01 4.08857256e-01 -3.51598971e-02 3.07949483e-01 7.11740136e-01 2.20186844e-01 2.32248843e-01 -5.39420426e-01 -2.60457695e-01 4.09421653e-01 2.02662259e-01 4.31112200e-01 -1.25034463e+00 -3.80368203e-01 -2.15997428e-01 -4.20798987e-01 -1.09909868e+00 2.21403584e-01 -9.91651237e-01 -2.54087418e-01 -1.47956514e+00 3.88217419e-01 2.66842455e-01 -7.81285405e-01 6.28470123e-01 -1.43914908e-01 -7.35113695e-02 -1.41704664e-01 1.31308869e-01 -5.02468944e-01 6.56444490e-01 8.69205058e-01 -2.14708641e-01 1.58299096e-02 -7.82802626e-02 -3.96935999e-01 5.21245718e-01 9.87587035e-01 -2.93977618e-01 1.14253424e-01 1.01369515e-01 1.35694012e-01 -1.51052326e-01 -6.40332280e-03 -7.81705618e-01 -2.70412475e-01 -9.26678330e-02 4.87731248e-01 -8.15017045e-01 5.04485488e-01 -5.05482316e-01 1.63728088e-01 9.48028564e-01 -3.53294194e-01 2.67485112e-01 1.72354221e-01 5.51027656e-01 -1.26206717e-02 8.05956945e-02 1.01534867e+00 -1.11082651e-01 -5.10758698e-01 4.11533177e-01 -6.90034032e-01 -4.74462986e-01 8.22065115e-01 -2.69142926e-01 -2.80161440e-01 -9.36680939e-03 -8.02125096e-01 2.10796893e-01 6.61141932e-01 3.97405535e-01 6.20243132e-01 -8.30006063e-01 -6.56188190e-01 3.04546535e-01 3.23916674e-01 -3.60729992e-01 1.45412877e-01 8.93969119e-01 -8.19031596e-01 7.29985833e-01 -4.39754456e-01 -5.19449532e-01 -1.63040602e+00 5.65235198e-01 4.17910635e-01 -4.03083324e-01 -5.39218068e-01 8.75494301e-01 1.20005310e-01 -7.82734334e-01 1.54181644e-01 -2.58671016e-01 -7.91625977e-02 -1.66318357e-01 5.06964624e-01 2.22968459e-01 2.25221902e-01 -7.56888688e-01 -4.71292615e-01 3.06059003e-01 -3.99104983e-01 8.07261169e-01 1.68379211e+00 2.28974655e-01 -3.65608990e-01 4.18059289e-01 1.33815336e+00 -6.43450677e-01 -1.05924356e+00 -1.78078294e-01 3.50368589e-01 1.29387468e-01 -4.86025363e-01 -1.03878057e+00 -4.43062544e-01 9.86933887e-01 9.20986176e-01 -6.12126887e-01 4.82124507e-01 -1.60254285e-01 1.32317364e+00 5.75681448e-01 4.44641888e-01 -8.06920648e-01 -2.72006154e-01 7.28662133e-01 6.48881674e-01 -1.31831396e+00 -1.84894174e-01 1.85860679e-01 -5.09762943e-01 1.16649079e+00 6.68322146e-01 8.78638923e-02 4.16138977e-01 -2.59904098e-02 1.06313109e-01 -1.38314083e-01 -1.12868309e+00 1.27183527e-01 1.94694102e-01 5.46545148e-01 9.11050260e-01 7.79297426e-02 -7.63326764e-01 6.42500401e-01 4.67106551e-02 -1.07204907e-01 2.47087494e-01 6.59693897e-01 -8.18652153e-01 -1.46847427e+00 5.81209734e-03 6.07051849e-01 -8.26480925e-01 -3.58899355e-01 -5.81125200e-01 2.98419267e-01 -1.99603327e-02 5.05230725e-01 -3.02438527e-01 -3.65182281e-01 -1.08619690e-01 5.04213572e-01 2.97175825e-01 -2.97061503e-01 -4.81887549e-01 -1.18973434e-01 -2.12994684e-02 -6.22386098e-01 -1.95177063e-01 -3.62826318e-01 -1.43585956e+00 -4.73331273e-01 -4.22293246e-01 3.14391226e-01 6.96543097e-01 8.23763967e-01 8.44635665e-01 1.31739751e-01 6.75497353e-02 -7.06355751e-01 -6.46263599e-01 -1.27014220e+00 -3.77937049e-01 2.92434275e-01 3.38581204e-01 -6.16908193e-01 -1.57545626e-01 -2.55444258e-01]	[4.706903457641602, 5.617234230041504]
cdfd860d-1585-462c-8bae-39d9df652bf9	semantic-data-set-construction-from-human	null	null	https://aclanthology.org/2021.cl-1.4	https://aclanthology.org/2021.cl-1.4.pdf	Semantic Data Set Construction from Human Clustering and Spatial Arrangement	Abstract Research into representation learning models of lexical semantics usually utilizes some form of intrinsic evaluation to ensure that the learned representations reflect human semantic judgments. Lexical semantic similarity estimation is a widely used evaluation method, but efforts have typically focused on pairwise judgments of words in isolation, or are limited to specific contexts and lexical stimuli. There are limitations with these approaches that either do not provide any context for judgments, and thereby ignore ambiguity, or provide very specific sentential contexts that cannot then be used to generate a larger lexical resource. Furthermore, similarity between more than two items is not considered. We provide a full description and analysis of our recently proposed methodology for large-scale data set construction that produces a semantic classification of a large sample of verbs in the first phase, as well as multi-way similarity judgments made within the resultant semantic classes in the second phase. The methodology uses a spatial multi-arrangement approach proposed in the field of cognitive neuroscience for capturing multi-way similarity judgments of visual stimuli. We have adapted this method to handle polysemous linguistic stimuli and much larger samples than previous work. We specifically target verbs, but the method can equally be applied to other parts of speech. We perform cluster analysis on the data from the first phase and demonstrate how this might be useful in the construction of a comprehensive verb resource. We also analyze the semantic information captured by the second phase and discuss the potential of the spatially induced similarity judgments to better reflect human notions of word similarity. We demonstrate how the resultant data set can be used for fine-grained analyses and evaluation of representation learning models on the intrinsic tasks of semantic clustering and semantic similarity. In particular, we find that stronger static word embedding methods still outperform lexical representations emerging from more recent pre-training methods, both on word-level similarity and clustering. Moreover, thanks to the data set’s vast coverage, we are able to compare the benefits of specializing vector representations for a particular type of external knowledge by evaluating FrameNet- and VerbNet-retrofitted models on specific semantic domains such as “Heat” or “Motion.”	['Anna Korhonen', 'Ivan Vulić', 'Nikolaus Kriegeskorte', 'Jasper J. F. van den Bosch', 'Diana McCarthy', 'Olga Majewska']	null	null	null	null	cl-acl-2021-3	['word-similarity']	['natural-language-processing']	[ 3.06545883e-01 -1.58353671e-01 -6.53799772e-02 -4.74748433e-01 -6.95951521e-01 -7.81570792e-01 9.32746291e-01 7.05094278e-01 -9.00075436e-01 3.44763041e-01 5.61265230e-01 -2.09007174e-01 -3.64045948e-01 -8.87343764e-01 -2.36752003e-01 -6.40227318e-01 2.33076215e-01 5.25133014e-01 3.79954189e-01 -4.59688425e-01 4.67902780e-01 4.90633458e-01 -1.89151609e+00 4.75807995e-01 3.72837782e-01 7.47210085e-01 5.97564638e-01 1.09566838e-01 -3.32040519e-01 3.08035165e-01 -4.99622911e-01 -1.09245956e-01 3.15216742e-02 -5.59589803e-01 -8.78315330e-01 -1.58906356e-01 4.22914773e-01 2.68715113e-01 5.11981733e-02 9.36773896e-01 6.76872969e-01 5.74259877e-01 8.14413905e-01 -8.10322881e-01 -7.98856199e-01 4.47006255e-01 -1.47125404e-03 4.03787255e-01 6.33335054e-01 -8.85073170e-02 1.33321917e+00 -1.02154768e+00 8.38517427e-01 1.36829793e+00 5.25321484e-01 5.15704751e-01 -1.54154420e+00 -4.15516853e-01 6.49548993e-02 5.30160189e-01 -1.30955732e+00 -4.29466516e-01 8.38289618e-01 -4.89667535e-01 1.20510018e+00 2.05789372e-01 8.01814556e-01 1.23804486e+00 -2.02985495e-01 2.52256602e-01 1.35298145e+00 -7.86493719e-01 4.66052592e-01 3.30344707e-01 3.42743367e-01 2.15309829e-01 3.16636860e-01 2.32712626e-01 -4.48917866e-01 -2.26394385e-02 5.75234413e-01 -1.73852056e-01 -2.30881661e-01 -5.85580945e-01 -1.45411158e+00 1.14859903e+00 5.46815515e-01 1.09677422e+00 -3.77758205e-01 -7.79353902e-02 7.23032594e-01 2.84232467e-01 4.41291422e-01 6.98577344e-01 -2.27498218e-01 -4.31644637e-03 -9.03613567e-01 2.69505560e-01 4.75378722e-01 5.16946673e-01 9.37322319e-01 -3.86938937e-02 -1.96511760e-01 1.02232993e+00 2.89944202e-01 2.08774105e-01 1.05503476e+00 -6.95375025e-01 2.08460093e-01 3.98513526e-01 -1.21937506e-01 -1.37406635e+00 -5.26828110e-01 4.43901531e-02 -1.32055864e-01 2.55652249e-01 3.35371852e-01 4.01102871e-01 -6.96367323e-01 2.21319079e+00 1.71615675e-01 -1.46903872e-01 1.05016336e-01 9.17892933e-01 8.56771171e-01 3.16900700e-01 5.28586566e-01 -2.39397600e-01 1.74145389e+00 -3.98960918e-01 -6.32845521e-01 -3.23029339e-01 9.60713446e-01 -6.07364297e-01 1.34668660e+00 -3.50559764e-02 -8.46397042e-01 -8.46078873e-01 -9.82016742e-01 -1.55990019e-01 -9.98511910e-01 -2.24872783e-01 6.70473814e-01 6.33855700e-01 -1.27489328e+00 4.80705082e-01 -4.99792486e-01 -9.37459111e-01 1.50900319e-01 6.29185364e-02 -5.62743545e-01 3.82451452e-02 -1.46673167e+00 1.46918118e+00 7.16794014e-01 -2.77649820e-01 -4.76836890e-01 -4.45643693e-01 -1.22002256e+00 5.04927859e-02 1.60253733e-01 -5.44467747e-01 7.60295987e-01 -1.09586442e+00 -9.78179574e-01 1.43171847e+00 -2.83778280e-01 -2.05153465e-01 -8.48813131e-02 2.94851273e-01 -4.79058236e-01 2.73507297e-01 5.46026170e-01 7.24298000e-01 5.34331977e-01 -1.22918570e+00 -2.43906572e-01 -5.58220029e-01 3.09943594e-02 4.12250191e-01 -4.87892210e-01 1.51664317e-01 1.96096837e-03 -9.74181592e-01 1.03737190e-01 -7.58359909e-01 4.58603688e-02 -2.45979235e-01 2.38015890e-01 -4.86642540e-01 4.32355613e-01 -4.46753860e-01 1.11247838e+00 -2.37269449e+00 3.11109722e-01 2.32309341e-01 -2.64243968e-03 2.00786844e-01 -4.76082981e-01 6.53923750e-01 -4.64064121e-01 1.31374687e-01 -2.40820944e-01 -1.33533314e-01 2.81063557e-01 2.15163067e-01 -2.75138319e-01 3.56243759e-01 -2.57260539e-02 9.84538078e-01 -9.72474992e-01 -5.02465367e-01 4.81607318e-01 4.77285057e-01 -5.20570636e-01 -8.60554948e-02 3.67151797e-02 6.89540654e-02 -1.50614634e-01 1.33241743e-01 2.33652204e-01 1.55952141e-01 3.35641146e-01 -4.96469766e-01 1.40743181e-01 6.17651224e-01 -1.11293972e+00 1.90242350e+00 -7.13274777e-01 7.45756567e-01 -3.87676179e-01 -1.44700587e+00 8.98159444e-01 3.28264654e-01 3.34051728e-01 -9.71705437e-01 1.53469369e-01 1.47267714e-01 2.56529421e-01 -4.27114576e-01 5.21546721e-01 -6.85361445e-01 -1.43061072e-01 6.30993366e-01 3.15672129e-01 -2.76353389e-01 3.56200278e-01 4.97122556e-02 7.69558430e-01 2.21812531e-01 5.87651968e-01 -5.66586137e-01 4.79495943e-01 1.23395987e-01 1.89432755e-01 4.20600981e-01 -2.80917943e-01 4.85546470e-01 2.18672931e-01 -2.52889276e-01 -1.02088749e+00 -1.25952530e+00 -5.35347939e-01 1.38676918e+00 1.54974788e-01 -6.60415530e-01 -6.20008767e-01 -4.69462126e-01 -1.92256823e-01 1.15972865e+00 -9.15978611e-01 -4.25835311e-01 -3.73247266e-01 -6.62611783e-01 4.77204353e-01 7.10084021e-01 -3.97669040e-02 -1.59152782e+00 -9.20966744e-01 2.27109239e-01 -5.95314316e-02 -1.02196431e+00 1.28221530e-02 2.35221580e-01 -6.42889380e-01 -1.11829436e+00 -3.51255178e-01 -1.04409564e+00 3.05440634e-01 2.48222008e-01 1.14568186e+00 -3.36704850e-02 -3.52625459e-01 6.64879620e-01 -5.55171251e-01 -1.64586172e-01 -3.25011462e-01 -3.14647406e-01 2.08434761e-01 -2.40280181e-01 9.76419032e-01 -4.70477849e-01 -2.98831493e-01 1.59985423e-01 -1.16173410e+00 -4.09404129e-01 2.80465037e-01 9.66636777e-01 4.57977742e-01 -8.01227912e-02 7.71988511e-01 -6.66336894e-01 9.14359868e-01 -5.33355474e-01 -7.49645978e-02 1.53902844e-01 -5.54444253e-01 1.07866101e-01 3.15593570e-01 -5.48607945e-01 -9.89334166e-01 -3.79385293e-01 -7.78396502e-02 -1.95404515e-01 -5.09605706e-01 5.62379301e-01 -3.19676697e-02 1.16125077e-01 9.82416332e-01 1.61711514e-01 1.46822393e-01 -2.29770124e-01 6.65889025e-01 5.15828967e-01 2.01839671e-01 -7.21633673e-01 4.75074053e-01 5.01749635e-01 -1.24906152e-01 -9.09388959e-01 -5.25721610e-01 -5.98381937e-01 -7.74185538e-01 1.32557884e-01 1.23397148e+00 -7.67386496e-01 -4.33474988e-01 -1.03212886e-01 -9.91627872e-01 -2.08427519e-01 -6.66236103e-01 7.52360880e-01 -8.75105143e-01 4.44286257e-01 -3.44298035e-01 -3.29989582e-01 1.43750533e-01 -1.16251004e+00 1.15999603e+00 -3.06322455e-01 -8.59026611e-01 -1.58698463e+00 3.05447668e-01 1.52251616e-01 4.21627045e-01 4.91183922e-02 1.26657128e+00 -1.17602813e+00 6.76962808e-02 7.37066418e-02 -1.59971714e-01 1.91770643e-01 3.53731275e-01 -4.90385711e-01 -1.01007807e+00 -2.39836022e-01 3.34952861e-01 -5.66373646e-01 9.29651558e-01 1.67094365e-01 8.13350499e-01 1.80348009e-01 -3.09640825e-01 2.72709012e-01 1.52291512e+00 1.03607655e-01 6.40581608e-01 4.92541164e-01 6.51450038e-01 1.09021676e+00 5.53497970e-01 1.07847145e-02 2.22483754e-01 1.00724697e+00 1.89052671e-02 1.68181673e-01 -3.12565744e-01 -6.42593578e-02 3.17056775e-01 7.41671562e-01 4.53702398e-02 1.31174371e-01 -8.00296068e-01 8.64109874e-01 -1.63644195e+00 -1.32156885e+00 1.25915185e-01 2.30389285e+00 7.41150618e-01 -1.76691145e-01 6.02026843e-02 3.02462190e-01 7.34007955e-01 3.58450979e-01 -3.84980701e-02 -6.14211380e-01 -2.25518525e-01 5.91400206e-01 -7.69613460e-02 4.78926629e-01 -8.19905579e-01 1.26858974e+00 6.39674330e+00 9.47253704e-01 -8.90092552e-01 2.03327194e-01 -6.52005011e-03 6.54366612e-02 -6.70901954e-01 1.25991523e-01 -3.08063507e-01 2.33934075e-01 9.94834900e-01 -1.39651746e-01 2.70705253e-01 6.24017239e-01 1.03709869e-01 -1.84414595e-01 -1.42491615e+00 1.00960243e+00 3.99395525e-01 -1.17803752e+00 2.70033032e-01 -5.13280854e-02 2.06407025e-01 -1.98156059e-01 -1.53447047e-01 2.91582376e-01 -6.11749552e-02 -1.23940718e+00 6.63091660e-01 1.76672637e-01 8.50237548e-01 -3.88150066e-01 5.68227351e-01 -3.06901745e-02 -1.22318053e+00 5.16734384e-02 -6.58482850e-01 -1.91307202e-01 1.64480373e-01 1.23367675e-01 -6.12592757e-01 3.22724193e-01 5.56577384e-01 7.50272810e-01 -7.89036930e-01 5.52029550e-01 -2.26323485e-01 2.83955276e-01 -2.16230601e-01 -2.66617000e-01 2.35104248e-01 -1.79588929e-01 4.48226482e-01 1.34705484e+00 2.53657043e-01 -8.59687924e-02 9.05546360e-03 1.00126815e+00 3.64582032e-01 6.36475027e-01 -9.94021773e-01 3.46027464e-02 6.17572784e-01 1.18421113e+00 -9.25781012e-01 -3.72599632e-01 -5.92555463e-01 8.46576095e-01 5.55860341e-01 3.33605915e-01 -5.74809551e-01 -3.12061220e-01 8.35287869e-01 3.41795501e-03 3.11752915e-01 -1.53537378e-01 -3.42336535e-01 -1.21313059e+00 -5.40018966e-03 -7.29855835e-01 3.71396214e-01 -8.32866430e-01 -1.44055057e+00 6.17136776e-01 3.49107891e-01 -8.81634295e-01 -4.61040199e-01 -8.57774496e-01 -4.62503523e-01 1.04747856e+00 -1.07296169e+00 -9.69884217e-01 -1.03461951e-01 7.90683806e-01 5.98172009e-01 -2.17664346e-01 1.28135312e+00 -2.40158793e-02 2.24293247e-02 3.16513240e-01 -1.86432332e-01 -5.00423834e-02 8.39378536e-01 -1.16120052e+00 1.19891420e-01 5.55786312e-01 6.50189459e-01 9.13554072e-01 6.01373971e-01 -5.05918622e-01 -6.71275854e-01 -5.45541227e-01 1.10619605e+00 -8.02893639e-01 8.63830447e-01 -5.12975752e-01 -9.95245159e-01 6.44982815e-01 2.17680231e-01 -9.70092416e-02 9.45345104e-01 2.78238714e-01 -5.79108715e-01 2.68849999e-01 -1.06759250e+00 6.86368287e-01 1.21363008e+00 -1.01825035e+00 -1.44358468e+00 2.23204568e-01 6.40141726e-01 2.44904414e-01 -8.32688391e-01 1.58912674e-01 3.48284185e-01 -1.08403957e+00 1.20717943e+00 -7.57312298e-01 2.12504193e-01 -1.28809363e-01 -6.53017461e-01 -1.50917673e+00 -6.12778068e-01 9.21144485e-02 5.76431811e-01 1.20781386e+00 3.15812320e-01 -6.58260763e-01 3.22955042e-01 2.43527517e-01 -2.88860761e-02 -3.43500763e-01 -9.58449006e-01 -6.81588113e-01 3.71567547e-01 -7.13256717e-01 3.69120598e-01 1.17924762e+00 2.69076884e-01 5.63576519e-01 3.02716583e-01 -3.91046733e-01 3.58707041e-01 7.79973194e-02 2.62143075e-01 -1.27372169e+00 -1.19470380e-01 -5.94361842e-01 -8.65969837e-01 -2.71104604e-01 6.84164107e-01 -1.33753836e+00 -6.07723519e-02 -1.52088594e+00 1.28100947e-01 -3.61152887e-01 -3.97485048e-01 3.72495323e-01 -1.28150091e-01 3.54836702e-01 3.94813031e-01 3.00724775e-01 -1.57344222e-01 4.44893867e-01 8.46356869e-01 1.38442650e-01 -2.84499545e-02 -5.95704675e-01 -8.18118095e-01 6.84370160e-01 7.31378257e-01 -3.37562591e-01 -6.59158528e-01 -2.14161322e-01 2.15667784e-01 -5.11989415e-01 6.49276674e-01 -9.18812633e-01 -1.61661908e-01 -1.05361931e-01 3.45626146e-01 1.18200563e-01 3.50950748e-01 -8.90834749e-01 -5.17486036e-02 2.10784048e-01 -5.31255603e-01 1.96771279e-01 2.28116363e-01 4.98951256e-01 -4.77415740e-01 -4.42080289e-01 7.81349123e-01 -2.29702532e-01 -1.25696266e+00 -2.59380072e-01 -3.95358622e-01 2.88671225e-01 1.18498099e+00 -6.05601013e-01 -1.36545613e-01 -1.41793773e-01 -1.06726253e+00 -3.24108750e-01 7.25945115e-01 6.84579849e-01 5.75692654e-01 -1.63942695e+00 -5.47931850e-01 1.35210007e-01 5.15806258e-01 -6.45771861e-01 -2.11810898e-02 6.39531016e-01 -2.24213526e-01 5.20502687e-01 -4.03562307e-01 -4.93317008e-01 -9.70915079e-01 1.07663202e+00 2.65641343e-02 3.21183316e-02 -5.64587712e-01 4.67287391e-01 5.65599322e-01 -4.48163897e-01 -1.50555894e-01 -1.85155183e-01 -6.20228350e-01 6.95916533e-01 3.15255225e-01 1.58738308e-02 7.03107864e-02 -1.20808780e+00 -5.49972892e-01 8.60603929e-01 2.32071325e-01 -5.41657150e-01 1.32976103e+00 -2.40576431e-01 -1.95558921e-01 8.68972838e-01 1.39270341e+00 5.93741313e-02 -6.33929193e-01 -2.21537977e-01 2.19908789e-01 -3.22225928e-01 -1.60676122e-01 -5.35903513e-01 -6.21346176e-01 9.46573794e-01 6.19383216e-01 2.28372470e-01 9.68247056e-01 3.54029387e-01 3.05204272e-01 2.62683839e-01 3.30882162e-01 -1.19223511e+00 1.57088518e-01 3.24530870e-01 8.67113769e-01 -1.09679091e+00 -1.76170826e-01 -2.49824986e-01 -7.21428096e-01 1.01787019e+00 3.18623394e-01 -2.77543843e-01 5.04472315e-01 -1.17073648e-01 1.04956903e-01 -3.90197247e-01 -4.90517050e-01 -6.31937921e-01 4.03486729e-01 9.31437969e-01 7.24960744e-01 1.39450096e-02 -8.02511275e-01 4.47877020e-01 -2.80825317e-01 -4.00729716e-01 2.57231832e-01 8.19730878e-01 -4.99108881e-01 -1.25562084e+00 -3.46667737e-01 1.78884849e-01 -2.03396648e-01 -4.48277205e-01 -3.49013507e-01 1.07083297e+00 1.60683811e-01 8.56102705e-01 3.16953212e-01 -2.40271941e-01 2.55410522e-01 4.92201626e-01 6.65124655e-01 -1.00659859e+00 -5.45089066e-01 -1.40356347e-01 1.92491040e-01 -7.39018619e-01 -8.97844493e-01 -7.48784423e-01 -1.22604299e+00 1.43975511e-01 -8.25112388e-02 2.26339743e-01 6.52697921e-01 1.10015476e+00 1.59650579e-01 3.71224165e-01 4.49621901e-02 -9.91189957e-01 -3.37604970e-01 -1.01850832e+00 -5.90994895e-01 1.23595452e+00 -1.19417146e-01 -1.02471185e+00 -4.87070411e-01 3.83882001e-02]	[10.388326644897461, 8.988271713256836]
e065b908-40ac-4745-84e0-3334730e44c1	cdec-net-composite-deformable-cascade-network	2008.10831	null	https://arxiv.org/abs/2008.10831v1	https://arxiv.org/pdf/2008.10831v1.pdf	CDeC-Net: Composite Deformable Cascade Network for Table Detection in Document Images	Localizing page elements/objects such as tables, figures, equations, etc. is the primary step in extracting information from document images. We propose a novel end-to-end trainable deep network, (CDeC-Net) for detecting tables present in the documents. The proposed network consists of a multistage extension of Mask R-CNN with a dual backbone having deformable convolution for detecting tables varying in scale with high detection accuracy at higher IoU threshold. We empirically evaluate CDeC-Net on all the publicly available benchmark datasets - ICDAR-2013, ICDAR-2017, ICDAR-2019,UNLV, Marmot, PubLayNet, and TableBank - with extensive experiments. Our solution has three important properties: (i) a single trained model CDeC-Net{\ddag} performs well across all the popular benchmark datasets; (ii) we report excellent performances across multiple, including higher, thresholds of IoU; (iii) by following the same protocol of the recent papers for each of the benchmarks, we consistently demonstrate the superior quantitative performance. Our code and models will be publicly released for enabling the reproducibility of the results.	['Madhav Agarwal', 'C. V. Jawahar', 'Ajoy Mondal']	2020-08-25	null	null	null	null	['table-detection']	['miscellaneous']	[-2.80950189e-01 -4.07384522e-02 2.29190998e-02 -2.17565522e-01 -1.29203749e+00 -9.65505838e-01 6.49474561e-01 1.58472672e-01 -2.56240904e-01 6.75940633e-01 4.09327865e-01 -2.74652511e-01 -3.49368304e-02 -7.36378133e-01 -1.09451544e+00 -2.76745230e-01 -1.48398414e-01 5.23092091e-01 4.20593768e-01 -1.56033248e-01 1.46441966e-01 5.32499671e-01 -1.04164231e+00 8.77570689e-01 5.95647156e-01 1.36485553e+00 -4.35824357e-02 7.36963391e-01 -8.48692730e-02 1.07131112e+00 -6.90163732e-01 -8.28494370e-01 3.61306846e-01 -1.12567984e-01 -8.79779041e-01 3.10921092e-02 7.14838624e-01 -5.87817192e-01 -5.82090735e-01 9.83279943e-01 3.96010578e-01 -2.17295989e-01 5.87563872e-01 -8.81133616e-01 -1.18666029e+00 1.08876967e+00 -6.27290785e-01 4.26938474e-01 2.34732032e-01 -1.05319053e-01 1.25656235e+00 -1.08627892e+00 9.45519269e-01 1.43514669e+00 7.94250667e-01 2.53351301e-01 -8.45661104e-01 -4.00639355e-01 1.76785827e-01 1.24967232e-01 -1.50470543e+00 -3.00427467e-01 3.57689291e-01 -2.28975356e-01 1.02860463e+00 3.48835200e-01 -1.08525328e-01 1.27702177e+00 2.77739108e-01 1.00864708e+00 8.63013685e-01 -3.56078386e-01 2.53430512e-02 2.24409267e-01 3.40173006e-01 9.39298451e-01 5.24924994e-01 -7.35111833e-01 -3.78083944e-01 1.09525710e-01 7.33574331e-01 -1.77443758e-01 -2.15114355e-01 -1.88365251e-01 -1.42570841e+00 5.66180885e-01 4.74230826e-01 5.14878213e-01 -7.29871988e-02 8.13594460e-02 6.69028342e-01 2.81796992e-01 4.91014183e-01 2.12634012e-01 -6.20287061e-01 1.25228809e-02 -7.78581858e-01 4.16870713e-01 6.80996060e-01 1.35683322e+00 2.62112349e-01 -3.14683795e-01 -6.36104405e-01 8.65824282e-01 6.34389222e-02 2.46949941e-01 3.36392641e-01 -4.72047538e-01 1.28242183e+00 8.42752218e-01 2.08126903e-01 -1.12107515e+00 -5.06945729e-01 -5.04054546e-01 -1.07126319e+00 -1.90877199e-01 5.08097529e-01 6.13861270e-02 -1.08671486e+00 1.24756646e+00 3.86790410e-02 -5.22802413e-01 -1.95985213e-02 7.49882817e-01 1.33962965e+00 8.44212055e-01 -2.57786989e-01 2.20297545e-01 1.73916912e+00 -9.30493712e-01 -8.82261217e-01 1.45614212e-02 4.54757750e-01 -8.20458472e-01 1.33617675e+00 4.14966375e-01 -1.13016450e+00 -6.31626070e-01 -1.29306209e+00 -6.66334331e-01 -8.98398161e-01 7.04845011e-01 4.13728803e-01 4.28197354e-01 -1.13645995e+00 4.31714207e-01 -5.66747129e-01 -5.90669215e-01 5.60378492e-01 2.84372121e-01 -3.77622068e-01 1.30602524e-01 -1.00496340e+00 5.45922637e-01 4.97256458e-01 2.88048625e-01 -7.11565137e-01 -4.95258659e-01 -6.60110474e-01 3.51911455e-01 6.30733192e-01 -3.96468133e-01 9.50841606e-01 -4.01767284e-01 -1.01574421e+00 1.28210509e+00 2.11332127e-01 -6.26353979e-01 9.91821706e-01 -5.16819596e-01 -4.67203021e-01 1.70133933e-01 2.67461121e-01 6.24911785e-01 3.08197707e-01 -1.13366461e+00 -3.77498388e-01 -4.20454472e-01 -5.79868928e-02 -1.76813155e-01 -3.18354368e-01 3.69604319e-01 -1.04084897e+00 -1.04004610e+00 1.06119767e-01 -6.16935670e-01 3.19510132e-01 -5.01047410e-02 -1.10653138e+00 -4.10292417e-01 5.99396646e-01 -9.11096990e-01 1.39191151e+00 -2.00230622e+00 -1.50771379e-01 -6.78793043e-02 4.42646623e-01 1.89803913e-01 -1.57227721e-02 4.48828280e-01 3.57273072e-02 2.57308841e-01 -2.68841445e-01 -3.25007737e-01 3.29957277e-01 -2.19215766e-01 -4.27497953e-01 4.62327659e-01 2.23315760e-01 1.14394486e+00 -3.80387992e-01 -6.41278267e-01 1.60264950e-02 3.54560524e-01 -3.47225189e-01 -2.49958434e-03 -2.82114923e-01 -1.83449805e-01 -2.66026169e-01 8.73872817e-01 8.47869992e-01 -7.58025885e-01 2.44782820e-01 -2.69549072e-01 2.69811437e-03 4.17128146e-01 -1.63701212e+00 1.37519395e+00 4.77931574e-02 7.30942547e-01 1.13136418e-01 -7.27937996e-01 8.32721770e-01 9.19266045e-02 3.88210341e-02 -8.08648348e-01 1.94320112e-01 2.27886587e-01 -3.10043842e-01 -1.89055905e-01 5.72232962e-01 7.61041164e-01 -2.44689479e-01 1.37766197e-01 2.71193624e-01 4.85447913e-01 6.10595584e-01 5.23408353e-01 1.25539076e+00 -1.56447232e-01 1.74107507e-01 -6.03155017e-01 7.22464502e-01 -2.59733409e-01 2.16049299e-01 1.04108787e+00 -2.05681071e-01 9.09458756e-01 8.93440962e-01 -6.45122826e-01 -1.17164063e+00 -1.15830350e+00 -2.73633718e-01 1.06281710e+00 -6.45494163e-02 -5.28837383e-01 -9.01883066e-01 -8.24650764e-01 5.02608679e-02 2.60040522e-01 -8.55412245e-01 4.48074430e-01 -8.34530115e-01 -7.58879185e-01 5.69135010e-01 8.91822815e-01 9.15951550e-01 -1.29237866e+00 7.06042722e-02 1.43975452e-01 -1.17503412e-01 -1.61050618e+00 -6.59459949e-01 3.91636938e-01 -4.76693034e-01 -1.04970634e+00 -7.22134888e-01 -1.16334128e+00 5.59115648e-01 5.45161031e-03 1.37501574e+00 9.82985124e-02 -2.99528658e-01 1.08405404e-01 -3.99231426e-02 -2.96478510e-01 -2.89515287e-01 2.89408475e-01 -2.69533068e-01 -9.03495848e-02 3.02593350e-01 7.16199055e-02 -4.50588018e-01 2.76409954e-01 -7.71413982e-01 -7.34616667e-02 4.31446463e-01 6.21911466e-01 6.70539260e-01 6.00402206e-02 2.76778013e-01 -1.20884883e+00 7.11147308e-01 -3.02562505e-01 -8.99629831e-01 3.68600637e-01 -4.51571256e-01 4.77537559e-03 8.16939533e-01 -4.46174257e-02 -8.51402164e-01 -2.47700810e-01 -9.17521864e-02 -3.64390947e-02 -7.22746477e-02 2.08123595e-01 -6.31881297e-01 4.43497211e-01 6.11248612e-01 2.47438759e-01 -6.56009614e-01 -7.78073907e-01 2.72723347e-01 7.34183729e-01 8.31038117e-01 -3.55995566e-01 9.47646201e-01 5.69840014e-01 -4.05391246e-01 -3.11508834e-01 -9.14386272e-01 -3.66152853e-01 -8.58030140e-01 1.21296078e-01 1.03529239e+00 -1.08142924e+00 -7.38442779e-01 5.39604127e-01 -1.21358168e+00 -1.32668093e-01 1.81565508e-01 5.95248416e-02 -1.90788507e-01 2.59915113e-01 -1.19400477e+00 -4.63398904e-01 -6.43858314e-01 -1.14607382e+00 1.23750150e+00 1.35514036e-01 -2.46684682e-02 -8.67906511e-01 -2.53549218e-01 4.22879845e-01 5.46724238e-02 4.28188533e-01 1.06885505e+00 -9.02217984e-01 -9.06890810e-01 -1.15398020e-01 -8.29490781e-01 2.58472055e-01 1.62700433e-02 2.36469105e-01 -9.39568639e-01 -2.21711054e-01 -5.00722468e-01 -4.68305290e-01 1.34384704e+00 2.72222400e-01 1.46221018e+00 -4.94730234e-01 -3.02886575e-01 5.49380004e-01 1.50042081e+00 2.21736655e-01 8.09051394e-01 8.18015933e-01 8.07393372e-01 1.93417579e-01 3.23532879e-01 3.40441763e-01 4.02707249e-01 3.94896269e-01 3.16268533e-01 -3.10552567e-01 -1.40151739e-01 -1.18902229e-01 3.59460413e-01 7.85748780e-01 2.85100132e-01 -8.29351723e-01 -9.04141843e-01 5.75050235e-01 -1.88348210e+00 -7.74927378e-01 -2.68300503e-01 1.80659485e+00 7.40368605e-01 7.66891062e-01 2.77826548e-01 2.34022275e-01 8.78846109e-01 3.06818068e-01 -3.14731836e-01 -4.18048233e-01 -4.85840917e-01 1.56042144e-01 5.10285914e-01 1.66291725e-02 -1.60566688e+00 9.15727973e-01 6.30358934e+00 7.29818463e-01 -6.64412677e-01 3.21822963e-03 1.15148771e+00 6.34175241e-02 8.36656801e-03 -7.06379235e-01 -1.24616146e+00 4.70658183e-01 7.85069227e-01 2.63686717e-01 2.73029119e-01 8.77949715e-01 -1.51018158e-01 2.45771140e-01 -1.13082862e+00 8.10278296e-01 6.46570250e-02 -1.74148285e+00 2.75219947e-01 -1.05471663e-01 5.43158233e-01 -6.69660866e-02 2.19267532e-01 3.66369605e-01 4.77218926e-01 -9.91457403e-01 9.35515702e-01 1.15051657e-01 1.00282705e+00 -8.03824365e-01 7.35374570e-01 -1.09504431e-01 -1.27941155e+00 6.29087090e-02 -5.31127870e-01 5.16746938e-01 -4.85439181e-01 5.08980393e-01 -5.87680221e-01 5.86614549e-01 1.20592940e+00 6.97549582e-01 -1.02577829e+00 6.68395579e-01 -8.14762861e-02 6.71685219e-01 -9.73605812e-02 -1.27535552e-01 5.08540690e-01 2.77263690e-02 2.36113995e-01 1.69568455e+00 -6.29935833e-03 -3.66882950e-01 -1.84827253e-01 1.04930687e+00 -7.92934835e-01 1.06328093e-01 -3.50056857e-01 -1.94206014e-01 3.38528395e-01 1.52495205e+00 -1.32346952e+00 -4.50831056e-01 -6.55950785e-01 8.65348876e-01 5.88749588e-01 7.56381750e-02 -1.03735757e+00 -6.45747662e-01 2.66228706e-01 1.29586156e-03 1.00966966e+00 -8.47418532e-02 -4.48577553e-01 -1.12392437e+00 5.11570215e-01 -1.14649141e+00 7.74915338e-01 -8.10776830e-01 -1.29156339e+00 9.30829525e-01 -3.29675734e-01 -1.06213391e+00 1.72751188e-01 -1.00395656e+00 -2.91262686e-01 6.72976077e-01 -1.18014324e+00 -1.06694543e+00 -3.04076105e-01 6.69779956e-01 6.63056314e-01 -4.42625344e-01 4.68927741e-01 4.85050678e-01 -9.84139800e-01 1.05981314e+00 4.65452731e-01 9.51791942e-01 8.79291773e-01 -1.66218281e+00 9.29915190e-01 1.05345607e+00 2.97937781e-01 8.48797321e-01 4.04352099e-01 -5.86233497e-01 -1.39037299e+00 -1.41915560e+00 7.57297754e-01 -8.09962392e-01 7.82626212e-01 -1.16646492e+00 -1.07006133e+00 1.08638239e+00 4.65861052e-01 1.21206954e-01 1.77168533e-01 1.84697747e-01 -6.81628406e-01 -3.29693705e-01 -1.15329146e+00 4.36941385e-01 1.05295062e+00 -3.97551268e-01 -4.10036772e-01 6.84786797e-01 8.73730779e-01 -8.03913951e-01 -9.06109691e-01 2.11827010e-01 5.32316864e-01 -9.39497232e-01 9.65896487e-01 -4.18394983e-01 7.95894563e-01 -6.80884197e-02 -2.13754013e-01 -7.01929331e-01 -5.47410905e-01 -5.06946146e-01 -3.40747476e-01 1.61778808e+00 6.32510602e-01 -3.81187707e-01 6.63556814e-01 1.15739210e-02 -1.00205861e-01 -7.90991962e-01 -7.41622984e-01 -8.48897398e-01 2.06984039e-02 -1.35063857e-01 6.82625294e-01 7.07152784e-01 -5.64142466e-01 3.73958647e-01 -2.24038854e-01 2.75347501e-01 4.92163122e-01 7.91997984e-02 7.86165416e-01 -9.51710463e-01 -2.07620412e-01 -3.99657100e-01 -9.92485434e-02 -1.03884113e+00 -1.55230194e-01 -8.20683241e-01 -1.80423602e-01 -1.83924568e+00 5.37737012e-01 -6.50454611e-02 -5.64660132e-01 4.39610004e-01 -2.78696001e-01 3.34483713e-01 2.89246291e-01 3.79082441e-01 -1.15425432e+00 -1.54083408e-02 1.22853529e+00 -2.80079752e-01 8.48097056e-02 -4.40246195e-01 -8.51653636e-01 5.80628276e-01 4.47464824e-01 -3.51160109e-01 3.87823544e-02 -4.20798153e-01 1.98887363e-01 -1.80190161e-01 2.13164493e-01 -1.14768553e+00 -1.75515920e-01 2.41049170e-01 1.07636929e+00 -1.34400618e+00 -2.82570988e-01 -4.52037305e-01 -4.23598468e-01 2.94058710e-01 -4.73431915e-01 4.10082877e-01 4.70824897e-01 4.91196901e-01 -1.15342639e-01 1.02630213e-01 7.57324338e-01 -2.52841383e-01 -6.56521857e-01 1.67148039e-01 -3.19032818e-01 2.91212946e-01 5.36987484e-01 6.36264905e-02 -7.84790635e-01 -1.49598822e-01 -3.75304729e-01 2.03227147e-01 1.88926548e-01 6.40969634e-01 2.43135437e-01 -1.31158340e+00 -7.12543786e-01 -3.59309800e-02 2.85903066e-01 9.45727155e-02 -1.54289305e-01 3.99746269e-01 -9.80256677e-01 9.41148221e-01 -3.83207984e-02 -5.33662617e-01 -1.11513066e+00 8.09965372e-01 2.13060394e-01 -8.01914036e-01 -7.88618982e-01 9.07705188e-01 3.39652568e-01 -4.83122945e-01 5.96296012e-01 -8.06208253e-01 -1.98516771e-01 3.35956477e-02 5.52448332e-01 3.66137862e-01 6.83304071e-01 -1.48178101e-01 -6.68920100e-01 1.27603173e-01 -6.52253091e-01 3.02367359e-01 1.43145931e+00 7.35647306e-02 1.43320905e-02 3.81603301e-01 1.20971334e+00 1.55919820e-01 -1.04356849e+00 -2.18177631e-01 1.73619226e-01 5.85008785e-02 -9.36702639e-02 -1.17299902e+00 -1.01412880e+00 5.43559611e-01 5.19532800e-01 4.11916316e-01 1.07582951e+00 -4.35423059e-03 8.88588428e-01 4.73278701e-01 1.22116752e-01 -1.12172949e+00 1.40756056e-01 5.18285513e-01 1.13086379e+00 -1.25393426e+00 1.36228204e-01 -3.75298589e-01 -3.84683073e-01 1.14520669e+00 7.53221989e-01 -1.05897047e-01 2.76182979e-01 5.38980246e-01 1.11106180e-01 -3.44725758e-01 -7.10161865e-01 -7.69289136e-02 5.80338001e-01 1.23812482e-01 6.28924549e-01 -2.13438079e-01 -1.77943856e-01 7.85360932e-01 -2.05510706e-01 -1.85915157e-01 4.32636917e-01 1.00974393e+00 -1.88819826e-01 -6.23416901e-01 -5.46742678e-01 5.54220557e-01 -8.06774855e-01 -2.04776883e-01 -8.26802254e-01 1.29819787e+00 -8.43231156e-02 5.78669667e-01 2.94586182e-01 9.44897346e-03 3.20666999e-01 -9.69607159e-02 3.29000503e-01 -4.03443545e-01 -8.78582835e-01 1.02115370e-01 1.67342290e-01 -6.14065409e-01 8.84811729e-02 -5.34721315e-01 -1.17941701e+00 -5.02242208e-01 -3.37930173e-02 -1.38567358e-01 2.53705353e-01 5.53491831e-01 6.11808956e-01 8.11747491e-01 3.08761239e-01 -9.27409828e-02 -5.01577199e-01 -1.11655319e+00 -6.50635481e-01 5.09311199e-01 3.45995426e-01 -5.73827088e-01 -2.64444351e-01 1.86139211e-01]	[11.683302879333496, 2.994267702102661]
fa8ff8ed-cced-4b02-87f2-60dea67f957b	from-universal-humanoid-control-to-automatic	2206.09286	null	https://arxiv.org/abs/2206.09286v1	https://arxiv.org/pdf/2206.09286v1.pdf	From Universal Humanoid Control to Automatic Physically Valid Character Creation	Automatically designing virtual humans and humanoids holds great potential in aiding the character creation process in games, movies, and robots. In some cases, a character creator may wish to design a humanoid body customized for certain motions such as karate kicks and parkour jumps. In this work, we propose a humanoid design framework to automatically generate physically valid humanoid bodies conditioned on sequence(s) of pre-specified human motions. First, we learn a generalized humanoid controller trained on a large-scale human motion dataset that features diverse human motion and body shapes. Second, we use a design-and-control framework to optimize a humanoid's physical attributes to find body designs that can better imitate the pre-specified human motion sequence(s). Leveraging the pre-trained humanoid controller and physics simulation as guidance, our method is able to discover new humanoid designs that are customized to perform pre-specified human motions.	['Kris M. Kitani', 'Ye Yuan', 'Zhengyi Luo']	2022-06-18	null	null	null	null	['humanoid-control']	['robots']	[-9.03928503e-02 1.91053599e-01 -4.16438729e-02 1.14954002e-01 1.87026665e-01 -6.43278182e-01 2.97033429e-01 -3.47395241e-01 -1.86835706e-01 4.01911378e-01 1.80286705e-01 9.80797783e-02 -1.10925227e-01 -8.72928262e-01 -6.50931299e-01 -2.22285718e-01 7.41321892e-02 8.46041977e-01 4.02714700e-01 -7.35232830e-01 2.13875502e-01 6.98849618e-01 -1.75757289e+00 -1.46640643e-01 8.96453559e-01 3.96570385e-01 5.70394278e-01 9.66128826e-01 3.87965947e-01 3.21616888e-01 -4.47111249e-01 -2.82200813e-01 3.55046630e-01 -7.24579155e-01 -6.57001972e-01 4.71648782e-01 8.87018591e-02 1.39287934e-01 -1.48214817e-01 5.89668632e-01 6.26094162e-01 6.25178635e-01 8.06568503e-01 -1.22756004e+00 -5.19078732e-01 6.90564156e-01 -2.80974001e-01 -4.90879804e-01 7.75041401e-01 7.76284873e-01 8.39451134e-01 -3.81610423e-01 9.45209324e-01 1.31631923e+00 7.32479274e-01 1.06114113e+00 -1.08677888e+00 -4.63872612e-01 -2.09108189e-01 -1.24738462e-01 -1.30954015e+00 1.40003607e-01 7.76170313e-01 -8.13244760e-01 6.67637706e-01 2.21245095e-01 1.47013748e+00 1.08988070e+00 3.70473117e-01 8.49774241e-01 1.95866004e-01 -3.43352646e-01 4.00977343e-01 -3.99880826e-01 -6.12716258e-01 8.67100775e-01 1.70077935e-01 7.36766085e-02 -2.49320492e-01 -8.44333693e-02 1.23947501e+00 -3.41517150e-01 -3.23485941e-01 -8.14046502e-01 -1.52437043e+00 7.26177454e-01 2.57712066e-01 1.68663144e-01 -5.39456367e-01 4.23064083e-01 2.78064966e-01 -2.82931656e-01 -5.13023257e-01 1.38942647e+00 -4.94721085e-01 -2.98126221e-01 -7.94089973e-01 1.10162985e+00 9.44872856e-01 1.10453618e+00 3.89290184e-01 3.25770944e-01 -3.35108757e-01 4.41521406e-01 -3.35165076e-02 4.23975408e-01 4.73613232e-01 -1.15892291e+00 2.37484321e-01 8.41461658e-01 3.07596833e-01 -8.63415182e-01 -6.86730921e-01 9.31052957e-04 -6.93599224e-01 2.48467922e-01 1.62012100e-01 -2.20921785e-01 -8.97869289e-01 1.49290001e+00 5.04712999e-01 -1.01248369e-01 -6.46062642e-02 1.43313515e+00 5.23763001e-01 4.48431998e-01 9.55945849e-02 3.20808768e-01 1.19534171e+00 -9.06137586e-01 -2.22661942e-01 -1.59500420e-01 6.15631878e-01 -3.41370374e-01 1.55471909e+00 4.21548367e-01 -1.11536825e+00 -7.04157352e-01 -1.07283580e+00 2.38943025e-01 3.59328538e-01 3.24150294e-01 2.95406342e-01 6.35699749e-01 -4.95875925e-01 9.85430121e-01 -7.21183002e-01 -3.88699234e-01 -8.99066254e-02 3.97489071e-01 -1.61253363e-02 3.44686091e-01 -1.06979406e+00 7.77482033e-01 5.92472672e-01 -1.54857591e-01 -9.31817710e-01 -7.21050680e-01 -9.99312222e-01 -1.33364305e-01 5.42412698e-01 -1.33139634e+00 1.18446469e+00 -7.46464074e-01 -1.90644908e+00 7.18134284e-01 6.46297157e-01 -5.48024215e-02 7.82947421e-01 -2.51502007e-01 -1.51176676e-01 -6.13487558e-03 -2.94453464e-02 8.77291799e-01 7.62341142e-01 -1.30080140e+00 -3.59554380e-01 6.03123792e-02 -2.36499503e-01 3.43279272e-01 -1.11882620e-01 -2.20530748e-01 -7.31833577e-01 -9.87009883e-01 -1.26202583e-01 -1.32039034e+00 -5.75589657e-01 2.16571555e-01 -7.84929752e-01 7.99934715e-02 5.19462287e-01 -3.76068920e-01 1.10621536e+00 -1.62245238e+00 1.04413795e+00 4.97213244e-01 -9.93190631e-02 3.40424150e-01 -1.45548582e-01 5.22113025e-01 4.17673260e-01 1.01163924e-01 -8.72936323e-02 8.20698868e-03 4.15120795e-02 1.45863086e-01 1.03499949e-01 5.36329299e-02 -7.36228898e-02 9.35242832e-01 -9.31716919e-01 -5.78423560e-01 7.37059265e-02 2.85874754e-01 -8.97938609e-01 3.40782911e-01 -7.13289261e-01 1.04454076e+00 -8.26608121e-01 4.33956444e-01 -5.49999923e-02 5.14095509e-03 1.58690393e-01 -1.12607814e-01 -6.88084289e-02 -3.99195850e-01 -1.11029899e+00 1.70761538e+00 -2.52895087e-01 6.42904863e-02 -6.25355989e-02 -3.42624635e-01 1.18565285e+00 2.29969457e-01 7.62676120e-01 -1.71403512e-01 3.72837692e-01 1.04232825e-01 -3.72065008e-02 -8.04324627e-01 6.67154729e-01 -2.16490388e-01 -3.75606507e-01 5.27115107e-01 -2.81397045e-01 -8.23726177e-01 1.42309502e-01 -3.36344033e-01 9.18209434e-01 6.03321791e-01 2.85906255e-01 -1.84372067e-01 4.96403813e-01 3.31853479e-01 5.99968195e-01 4.00943488e-01 7.62765631e-02 8.81439030e-01 1.32004648e-01 -4.74795729e-01 -1.68497932e+00 -9.01207566e-01 8.03919911e-01 1.05051863e+00 3.22304726e-01 -3.75043809e-01 -9.20653045e-01 9.52002704e-02 4.34201062e-02 5.25330067e-01 -3.46377462e-01 -4.63677049e-01 -1.04587913e+00 -8.96259993e-02 7.29834378e-01 6.69703364e-01 3.93880010e-01 -1.54878616e+00 -1.49882877e+00 4.03129220e-01 -2.35845610e-01 -1.00621343e+00 -8.80855203e-01 -4.57118183e-01 -5.68193793e-01 -9.38161254e-01 -1.02307308e+00 -9.50538576e-01 3.37529242e-01 -2.78300732e-01 9.53971922e-01 3.09445709e-01 -5.70531428e-01 3.26863468e-01 -5.53584099e-01 -8.55366588e-02 -6.85885906e-01 1.00437470e-01 4.04806703e-01 -3.20968151e-01 -4.55824077e-01 -5.55735111e-01 -6.60965681e-01 6.61247075e-01 -8.15925419e-01 6.04274452e-01 4.31716114e-01 6.35058939e-01 6.37601137e-01 8.02696794e-02 3.32659408e-02 -2.55985200e-01 7.11059093e-01 -2.59119235e-02 -3.17894578e-01 3.78040820e-01 -2.19166368e-01 9.09139812e-02 8.08621764e-01 -1.10762453e+00 -8.66717398e-01 5.10972440e-01 2.42996048e-02 -4.68322307e-01 -9.78023335e-02 2.68080294e-01 -5.14321327e-01 1.22596323e-01 9.88873482e-01 3.88965197e-02 -1.74582273e-01 -3.58576089e-01 6.43623769e-01 6.75860718e-02 8.19467425e-01 -1.26207340e+00 8.24837148e-01 1.39511883e-01 1.47696599e-01 -7.64230192e-01 7.05154315e-02 -2.49890059e-01 -8.70770395e-01 -5.83303988e-01 1.32214582e+00 -5.87409198e-01 -1.27665234e+00 6.18648708e-01 -1.01190376e+00 -7.87465632e-01 -2.79445201e-01 4.79655445e-01 -1.23573148e+00 1.78945467e-01 -4.53765303e-01 -6.54818594e-01 -3.68741840e-01 -1.21936142e+00 1.34074950e+00 4.30234432e-01 -9.57087517e-01 -4.51706439e-01 4.39256668e-01 1.74782425e-01 -2.79419217e-02 6.96090698e-01 8.62675488e-01 1.48196786e-03 -3.24961573e-01 -1.54485404e-01 5.38751841e-01 -2.21085325e-01 3.40038501e-02 2.89799422e-01 -9.32278708e-02 -3.93574446e-01 -6.59005344e-01 -3.77402335e-01 7.29348436e-02 3.28040510e-01 9.05128300e-01 -4.66894001e-01 -3.75125617e-01 7.25880861e-01 1.09686315e+00 2.11792558e-01 5.88033795e-01 3.24545026e-01 9.98386264e-01 9.70985055e-01 7.41934001e-01 8.86865199e-01 3.55737150e-01 1.11210215e+00 1.03389718e-01 2.71832764e-01 -2.17133947e-02 -7.49337137e-01 2.27098733e-01 3.38600487e-01 -5.10596693e-01 -3.42496812e-01 -1.27308416e+00 5.78302920e-01 -1.90241349e+00 -9.71114457e-01 -6.10111579e-02 1.94938207e+00 8.26446950e-01 -3.81475911e-02 7.68309891e-01 -3.90303954e-02 8.98912191e-01 -3.68361801e-01 -7.72553623e-01 -2.20043316e-01 9.02980268e-02 1.81531861e-01 3.17342192e-01 1.11475037e-02 -6.22900128e-01 1.17729199e+00 5.55042124e+00 5.94135582e-01 -1.18608904e+00 -5.08627355e-01 -8.28165188e-02 1.68487415e-01 -4.45924193e-01 -1.32518830e-02 -6.22479260e-01 2.64355391e-01 3.67104411e-01 -3.97821218e-01 3.31442773e-01 7.81885326e-01 4.92565811e-01 1.19100288e-01 -9.85705912e-01 8.73036265e-01 -1.59946874e-01 -1.34659362e+00 2.34393731e-01 1.28075127e-02 9.29582655e-01 -8.54342461e-01 3.43384407e-02 1.03087962e-01 6.51869833e-01 -1.33251417e+00 1.10820472e+00 6.71966851e-01 6.96052969e-01 -8.01511049e-01 8.80473405e-02 7.98445702e-01 -1.44344282e+00 -1.18890181e-01 2.35562716e-02 -1.13285119e-02 5.68144143e-01 -1.13451764e-01 -8.01370084e-01 4.07611758e-01 5.31862378e-01 3.11512709e-01 -3.77377689e-01 1.15749252e+00 -2.86600500e-01 1.68646336e-01 -2.00873166e-01 -4.48496908e-01 1.01965833e-02 -2.62084305e-01 8.07402313e-01 8.03505599e-01 5.89560807e-01 4.88863260e-01 5.46653450e-01 1.12581837e+00 4.41162795e-01 3.13993812e-01 -5.01268685e-01 -2.30887651e-01 3.34297985e-01 9.85348582e-01 -7.98100948e-01 -1.02891400e-01 3.83844763e-01 1.23267078e+00 -1.74151622e-02 2.06901088e-01 -8.40830743e-01 -2.48253375e-01 5.30258358e-01 4.57519054e-01 1.78599343e-01 -5.59421659e-01 -5.08630872e-01 -9.81997967e-01 -2.24162042e-01 -1.00994301e+00 -2.85823159e-02 -1.15205002e+00 -7.97682106e-01 4.37973827e-01 1.34054437e-01 -1.43286228e+00 -6.44849181e-01 -2.80746669e-01 -7.87802339e-01 5.72831392e-01 -3.63756493e-02 -1.24626541e+00 -4.37157184e-01 5.17316520e-01 4.60739940e-01 -3.52008641e-01 4.61509436e-01 -1.76404789e-01 -1.90858707e-01 4.52692300e-01 -6.56623006e-01 7.13570416e-02 4.16156262e-01 -9.42270875e-01 8.40964437e-01 4.97016162e-01 -1.55222282e-01 3.36949646e-01 1.29008865e+00 -1.11506689e+00 -1.23568666e+00 -9.30573285e-01 3.54616702e-01 -3.57138127e-01 3.04396331e-01 6.54879492e-03 -6.85275078e-01 3.71007562e-01 -4.92320150e-01 -5.31080127e-01 3.66586238e-01 -4.93072301e-01 1.87671304e-01 5.18929482e-01 -9.93643701e-01 1.10578048e+00 1.42324054e+00 1.88104603e-02 -5.43761313e-01 -5.10225855e-02 6.38200283e-01 -6.84208035e-01 -9.11581337e-01 6.70572996e-01 9.68766749e-01 -4.53826189e-01 1.19758356e+00 -7.71787703e-01 5.17506480e-01 -4.64552522e-01 9.29500908e-02 -1.39359689e+00 -3.62465352e-01 -9.36961532e-01 2.66343821e-02 6.45920217e-01 4.35575359e-02 1.35333851e-01 1.16615200e+00 7.52588511e-01 -2.12498963e-01 -6.21353090e-01 -4.01096165e-01 -1.13545430e+00 1.69300158e-02 -2.10505769e-01 8.77506614e-01 5.79908252e-01 3.66177708e-02 1.28395678e-02 -7.65755236e-01 -2.98749227e-02 4.76043373e-01 1.19276777e-01 1.41842794e+00 -1.25955379e+00 -6.66814148e-01 -5.91845810e-01 -5.11540174e-01 -1.14269567e+00 1.20446421e-01 -4.87711549e-01 4.27394539e-01 -1.40485096e+00 -9.99834165e-02 -5.22564411e-01 6.92357540e-01 3.29431087e-01 -2.57925510e-01 -2.16960967e-01 5.44892251e-01 1.62449837e-01 -2.62066424e-01 7.49148011e-01 1.99225926e+00 1.30455494e-02 -9.15648758e-01 2.65514284e-01 -3.04300368e-01 6.38179004e-01 9.08054173e-01 -2.54654884e-01 -4.92350340e-01 -2.39414014e-02 4.00795013e-01 4.75239217e-01 4.92063582e-01 -1.31087005e+00 2.38481075e-01 -7.29399860e-01 2.45627686e-01 -3.96968454e-01 1.85734943e-01 -4.09158409e-01 7.98038304e-01 9.55762804e-01 -3.29409689e-01 2.74948090e-01 3.39207277e-02 4.85139638e-01 2.17576474e-01 -5.64700589e-02 7.67368436e-01 -2.64644206e-01 -9.14518416e-01 2.88643718e-01 -5.28679013e-01 1.05289206e-01 1.25913978e+00 -5.10415256e-01 1.35733023e-01 -4.99639720e-01 -8.35193694e-01 4.98287141e-01 9.97117102e-01 7.08740950e-01 7.56941020e-01 -1.33101797e+00 -6.03301823e-01 2.85445303e-02 2.23506629e-01 1.61920220e-01 2.31505036e-01 2.13818327e-01 -1.08796656e+00 -3.06939408e-02 -7.21805096e-01 -5.44448912e-01 -1.11898208e+00 5.63203752e-01 3.01862240e-01 -3.70407254e-02 -9.33117986e-01 6.62230194e-01 7.37460405e-02 -7.60139644e-01 -1.84754089e-01 -4.97345418e-01 -1.55457452e-01 -6.63526535e-01 -8.44340175e-02 5.12316167e-01 -5.95472753e-01 -7.79668808e-01 -2.27627337e-01 9.58877742e-01 7.78914511e-01 -5.85922241e-01 1.19615304e+00 2.74956971e-01 4.12524223e-01 7.15407431e-02 5.68784416e-01 -9.08400863e-02 -1.57279062e+00 3.92579556e-01 6.29560053e-02 -3.24906021e-01 -8.08712900e-01 -5.52040458e-01 -8.97706807e-01 4.76359874e-01 9.17347148e-02 -6.01656497e-01 7.65884340e-01 1.08480543e-01 1.19211292e+00 4.14582878e-01 8.09195697e-01 -1.25355697e+00 8.40550482e-01 4.38056588e-01 1.14779484e+00 -6.19772792e-01 -2.56707110e-02 -4.29004878e-01 -1.26801908e+00 1.07656133e+00 9.64001894e-01 -2.51982540e-01 3.61882180e-01 2.68780351e-01 -1.93995178e-01 -1.53541446e-01 -4.72483575e-01 -1.72915399e-01 8.08225155e-01 9.07067180e-01 7.76428357e-02 3.71456712e-01 -2.92174608e-01 8.43487024e-01 -8.34033847e-01 8.42479169e-02 5.67769170e-01 9.33882177e-01 -6.31626129e-01 -1.09461761e+00 -5.36796570e-01 -8.82942528e-02 3.11018378e-01 5.26507854e-01 -6.15803897e-01 8.61606956e-01 3.38952571e-01 5.61536610e-01 -2.48315245e-01 -8.04708540e-01 8.00202549e-01 -2.37724677e-01 7.24423647e-01 -8.80539954e-01 -7.81209946e-01 -1.88931555e-01 -9.79818255e-02 -5.21019995e-01 -1.07118180e-02 -5.89556217e-01 -1.83961272e+00 -4.37643647e-01 1.70975804e-01 -1.48672566e-01 3.15615237e-01 6.42777443e-01 1.19278505e-01 3.88959587e-01 8.09528306e-02 -1.47116804e+00 -3.46254110e-01 -4.71613765e-01 -4.48679686e-01 8.91312182e-01 -1.48498818e-01 -9.24389005e-01 3.69070470e-01 4.18221533e-01]	[5.146846294403076, 0.6731878519058228]
81f891d0-f027-4d62-9b45-7eac6dc62e6d	high-dimensional-bayesian-optimisation-with	2106.03609	null	https://arxiv.org/abs/2106.03609v3	https://arxiv.org/pdf/2106.03609v3.pdf	High-Dimensional Bayesian Optimisation with Variational Autoencoders and Deep Metric Learning	We introduce a method combining variational autoencoders (VAEs) and deep metric learning to perform Bayesian optimisation (BO) over high-dimensional and structured input spaces. By adapting ideas from deep metric learning, we use label guidance from the blackbox function to structure the VAE latent space, facilitating the Gaussian process fit and yielding improved BO performance. Importantly for BO problem settings, our method operates in semi-supervised regimes where only few labelled data points are available. We run experiments on three real-world tasks, achieving state-of-the-art results on the penalised logP molecule generation benchmark using just 3% of the labelled data required by previous approaches. As a theoretical contribution, we present a proof of vanishing regret for VAE BO.	['Haitham Bou-Ammar', 'Jan Peters', 'Jun Wang', 'Zhitang Chen', 'Wenlong Lyu', 'Lin Zhu', 'Lin Yang', 'Alexander I. Cowen-Rivers', 'Ryan-Rhys Griffiths', 'Alexandre Max Maraval', 'Rasul Tutunov', 'Antoine Grosnit']	2021-06-07	null	null	null	null	['bayesian-optimisation']	['methodology']	[ 2.80485660e-01 3.10769916e-01 6.56459108e-02 -2.97957808e-01 -9.99845982e-01 -6.89203978e-01 9.20929551e-01 7.58746490e-02 -7.46717632e-01 1.05014479e+00 1.56147584e-01 -3.42137367e-01 -4.08025533e-01 -5.57047367e-01 -8.47825527e-01 -9.60386574e-01 2.97119431e-02 7.61141300e-01 -2.13682979e-01 6.28563836e-02 2.31255829e-01 4.68184233e-01 -1.07686186e+00 8.05518478e-02 5.73537529e-01 9.73265290e-01 3.49198096e-02 1.01259840e+00 1.92510128e-01 4.87095267e-01 -3.08959305e-01 -6.58245325e-01 5.06253719e-01 -4.62303460e-01 -8.01670551e-01 -1.92520171e-01 9.80347097e-02 -1.57855764e-01 -3.19828033e-01 8.99604976e-01 7.71597803e-01 4.77396399e-01 1.25983393e+00 -9.35754478e-01 -6.03554428e-01 3.51042271e-01 -2.04909250e-01 3.06552202e-02 5.37687987e-02 3.85171443e-01 1.26305044e+00 -8.97427738e-01 8.13945651e-01 1.19872046e+00 6.42001629e-01 8.04909885e-01 -1.81374967e+00 -2.75255710e-01 -2.24489257e-01 4.99327257e-02 -1.11735475e+00 -6.25572979e-01 5.38822353e-01 -6.08354568e-01 1.11379659e+00 1.90035194e-01 4.60384011e-01 1.55885005e+00 3.47291917e-01 7.13627040e-01 1.01161575e+00 -2.67346382e-01 7.61818051e-01 -6.78699687e-02 -1.93560511e-01 5.91131687e-01 -1.61744595e-01 6.04000807e-01 -4.85508919e-01 -5.18587053e-01 6.05549395e-01 -4.20385376e-02 1.16479574e-02 -9.63104248e-01 -1.29585421e+00 1.42465818e+00 3.73501033e-01 -2.02204347e-01 -5.54182053e-01 4.33460027e-01 3.13071072e-01 1.70027465e-01 4.85550314e-01 7.69090056e-01 -6.86047912e-01 -4.23304319e-01 -9.57855701e-01 4.92453933e-01 1.01385558e+00 8.46525609e-01 4.42030251e-01 -1.62623122e-01 -4.45931494e-01 5.57898760e-01 5.58653653e-01 3.10377479e-01 1.92399070e-01 -1.46005380e+00 2.26098105e-01 -9.87400785e-02 3.85753065e-01 -3.30409348e-01 -3.42328638e-01 -3.93497735e-01 -8.18396747e-01 3.06458682e-01 6.68404639e-01 -4.50703382e-01 -1.04913855e+00 1.70725012e+00 4.69555289e-01 -1.63005799e-01 1.28923044e-01 6.85295284e-01 2.20192552e-01 6.73649848e-01 5.30460337e-03 -2.79015541e-01 1.02283716e+00 -7.86077559e-01 -5.69116473e-01 1.09989591e-01 5.86010993e-01 -5.69821715e-01 8.17242205e-01 7.72975445e-01 -1.12072682e+00 -1.31315649e-01 -1.04570532e+00 -2.72990435e-01 -3.89535517e-01 -2.16894776e-01 8.61216187e-01 7.44280934e-01 -9.20264602e-01 1.27984571e+00 -1.08204091e+00 -8.31601769e-02 8.42341959e-01 4.44324493e-01 -1.72763258e-01 -5.43619953e-02 -1.03369880e+00 7.33406723e-01 5.11467457e-01 1.20636962e-01 -1.64085424e+00 -8.78746808e-01 -8.02336872e-01 -5.07751703e-02 4.37735111e-01 -9.98326361e-01 1.30574763e+00 -2.66742140e-01 -2.01659751e+00 4.64776695e-01 -3.30285579e-02 -7.55176365e-01 9.10649061e-01 -2.29923248e-01 1.27619013e-01 -1.04218327e-01 -3.48361075e-01 8.02612484e-01 9.59778845e-01 -8.15418541e-01 -3.81434798e-01 -2.04929620e-01 -1.11365117e-01 1.73523054e-02 -7.38808513e-02 -4.25280742e-02 -9.24823247e-03 -3.26573133e-01 -2.17548192e-01 -9.15516317e-01 -5.68447232e-01 4.32315171e-02 -3.27337116e-01 -4.15356427e-01 3.86765867e-01 -5.53335547e-01 9.44227397e-01 -1.82374835e+00 8.18891525e-01 2.25913495e-01 4.82316107e-01 -4.63980529e-03 -5.80157675e-02 2.90789425e-01 4.82818596e-02 1.92053542e-01 -6.44123971e-01 -6.28407300e-01 5.60895383e-01 3.26296777e-01 -1.42886683e-01 6.32332861e-01 3.17270160e-01 1.05724680e+00 -1.07581842e+00 -1.54653996e-01 1.99497983e-01 4.27864730e-01 -8.43385577e-01 3.57802480e-01 -6.19612694e-01 5.29465556e-01 -1.71294510e-01 3.63632113e-01 4.35433537e-01 -1.40574813e-01 1.27183810e-01 6.50046691e-02 1.81819305e-01 2.73863107e-01 -1.14412594e+00 2.14836979e+00 -3.23215365e-01 4.36809927e-01 9.75906998e-02 -9.15278912e-01 5.74983776e-01 1.74194485e-01 5.48608959e-01 -2.53973603e-01 2.03815147e-01 3.76835316e-02 9.91040468e-03 -7.36677051e-02 1.53348386e-01 -3.27137083e-01 4.26717252e-02 3.58272612e-01 4.52477515e-01 -4.08401966e-01 1.71349093e-01 2.84392774e-01 1.17750359e+00 5.90405703e-01 3.18262994e-01 -4.44556952e-01 2.32039884e-01 -3.61190945e-01 5.85063338e-01 9.79770660e-01 -4.35827464e-01 5.12061894e-01 8.48425806e-01 -4.82486725e-01 -1.53097057e+00 -1.22611785e+00 -4.30089891e-01 1.21434462e+00 -5.27785897e-01 -6.45689428e-01 -8.72691810e-01 -8.04285049e-01 8.64504129e-02 8.01848054e-01 -8.53696048e-01 -1.37797400e-01 -2.77761668e-01 -1.20042348e+00 4.14149761e-01 3.96264255e-01 -2.37173557e-01 -9.67576504e-01 -4.00391966e-01 4.75103945e-01 4.46861267e-01 -7.01951921e-01 -3.09159130e-01 6.93143845e-01 -5.43903172e-01 -8.43185306e-01 -8.10743272e-01 -1.24604039e-01 1.96696237e-01 -4.84162092e-01 1.15297675e+00 -7.91426301e-01 -3.89925003e-01 6.57667965e-02 5.16347587e-02 -6.15768850e-01 -5.62160313e-01 7.69966617e-02 2.54814416e-01 -2.90179551e-01 3.48561853e-01 -5.60726762e-01 -7.72579789e-01 1.04566447e-01 -5.84586620e-01 -1.69110820e-01 4.15497363e-01 1.25122738e+00 6.78835392e-01 -3.61214221e-01 3.10243249e-01 -7.64160752e-01 7.02993989e-01 -6.11867726e-01 -1.00448453e+00 -6.48923069e-02 -8.44278514e-01 6.57085419e-01 4.86888081e-01 -3.44852805e-01 -8.15285385e-01 -8.67242552e-03 -1.84204131e-01 -5.10500193e-01 -1.90898731e-01 2.43451700e-01 -3.54331061e-02 1.59418687e-01 9.17176485e-01 -7.66688362e-02 1.98600590e-01 -5.18304884e-01 9.41081285e-01 5.77490270e-01 2.52970785e-01 -8.58628511e-01 4.79470193e-01 4.71220672e-01 4.81286108e-01 -2.85163611e-01 -9.30035353e-01 -2.43330657e-01 -7.54793048e-01 2.35458493e-01 1.08661211e+00 -6.04860008e-01 -1.07619250e+00 9.14389938e-02 -1.10617459e+00 -6.46041691e-01 -6.22224748e-01 5.28013706e-01 -1.08836591e+00 2.43783921e-01 -5.26357353e-01 -8.93725574e-01 -4.35916960e-01 -1.27606297e+00 1.09323835e+00 -2.40601569e-01 -7.60645494e-02 -1.02672803e+00 5.15128136e-01 2.06449553e-01 3.48098189e-01 4.09941316e-01 6.49875700e-01 -8.08874547e-01 -4.08376515e-01 8.95445608e-03 -2.03339234e-02 5.81067741e-01 -1.88815057e-01 -1.08748935e-01 -1.16770399e+00 -4.36027884e-01 -2.46356484e-02 -5.49173772e-01 1.09193158e+00 5.36419332e-01 1.22769761e+00 -2.10068434e-01 -1.60375163e-01 9.87354577e-01 1.06278169e+00 -1.11427560e-01 3.53640586e-01 9.02379975e-02 8.46763492e-01 5.02945065e-01 4.46327507e-01 6.94765031e-01 1.95873920e-02 5.80141187e-01 4.51271206e-01 1.71904057e-01 2.98285723e-01 -1.76450580e-01 3.45529765e-01 4.44942176e-01 1.08446345e-01 -2.24828362e-01 -8.02789092e-01 3.63821894e-01 -1.96929812e+00 -8.52185726e-01 1.22770689e-01 2.31043935e+00 1.14082575e+00 1.60420641e-01 3.16106945e-01 -9.13095251e-02 3.04149330e-01 1.93130359e-01 -8.36506128e-01 -6.66648388e-01 1.61271065e-01 5.40447891e-01 8.63274038e-01 6.62994385e-01 -1.30599475e+00 9.03904855e-01 7.40505028e+00 1.08004618e+00 -4.96895045e-01 3.35228175e-01 6.38012230e-01 -6.67406261e-01 -2.34601468e-01 -2.01404244e-02 -7.13569045e-01 4.34282213e-01 1.52258527e+00 3.44939195e-02 9.23779845e-01 9.02512610e-01 1.53699249e-01 1.28470868e-01 -1.51435935e+00 1.04679894e+00 -3.77242863e-01 -1.47667539e+00 -3.71511191e-01 4.91818577e-01 8.13589990e-01 4.62449431e-01 2.36735389e-01 3.68332148e-01 1.01265430e+00 -1.33010161e+00 4.74978507e-01 6.29720211e-01 7.79715300e-01 -1.05650949e+00 5.33054531e-01 3.04641515e-01 -5.14745295e-01 -1.22467957e-01 -7.10639060e-01 1.30337313e-01 1.62753060e-01 7.46999443e-01 -9.29374099e-01 4.76623178e-01 4.02651340e-01 7.46286750e-01 -1.93118066e-01 7.43482113e-01 -3.08630258e-01 6.91806793e-01 -5.21059453e-01 -1.08035490e-01 5.01789868e-01 -5.54290891e-01 7.25674093e-01 1.13694382e+00 1.95726696e-02 -3.11490417e-01 -1.51775956e-01 1.23541880e+00 -3.86645466e-01 -1.85398400e-01 -5.62230289e-01 -1.78009674e-01 1.97074905e-01 1.15309322e+00 -2.28563458e-01 -1.82288870e-01 1.19002210e-02 1.03271103e+00 4.94267166e-01 3.56505752e-01 -7.00588465e-01 -4.24321175e-01 9.02674615e-01 -4.16644037e-01 6.78913593e-01 -3.71841490e-01 -1.92993879e-01 -9.34408069e-01 -2.45562181e-01 -7.80040920e-01 5.08219957e-01 -3.58630031e-01 -1.49904656e+00 4.48378138e-02 -7.05221668e-02 -5.60295463e-01 -5.44863641e-01 -9.63626742e-01 -2.85997152e-01 1.00465620e+00 -1.28368092e+00 -7.73496747e-01 4.13079917e-01 2.59181470e-01 2.84658164e-01 -1.85994312e-01 1.03905821e+00 -3.74387018e-02 -5.73363245e-01 5.16422272e-01 9.77362096e-01 -3.00112635e-01 6.02275252e-01 -1.94155216e+00 6.40227616e-01 5.89433193e-01 3.15219253e-01 6.77995205e-01 8.61598372e-01 -4.75809485e-01 -1.41696990e+00 -8.70328009e-01 3.31258863e-01 -9.96908009e-01 8.53730440e-01 -9.09398139e-01 -5.97746909e-01 5.02045393e-01 1.80822425e-02 1.33132592e-01 1.03630221e+00 2.82871068e-01 -2.38914236e-01 2.13635474e-01 -1.30807555e+00 3.44775498e-01 1.11363459e+00 -6.06193423e-01 -5.53383291e-01 7.22469509e-01 7.93937087e-01 -4.93822962e-01 -1.21559441e+00 1.37117669e-01 5.57329595e-01 -6.76930249e-01 1.04855490e+00 -1.16601336e+00 3.70174199e-01 -8.82352665e-02 -2.83504963e-01 -1.38524663e+00 -4.11318690e-01 -1.27477789e+00 -8.04059505e-01 7.64524043e-01 4.48092461e-01 -4.52179849e-01 8.34372342e-01 7.26636350e-01 6.95619583e-02 -1.02851021e+00 -1.21679211e+00 -6.46863759e-01 5.76386929e-01 -4.50893193e-01 4.39633280e-01 4.31300908e-01 -2.28309184e-01 3.13145787e-01 -6.31988525e-01 -2.41361424e-01 9.72237051e-01 -3.49672616e-01 5.98249197e-01 -1.13551021e+00 -7.58736372e-01 -3.86618704e-01 -3.02816302e-01 -9.98715937e-01 2.72900879e-01 -9.28298533e-01 1.54511910e-02 -1.20059907e+00 1.83609366e-01 2.02419125e-02 -6.47964954e-01 2.33914524e-01 -1.85154513e-01 5.13763502e-02 2.20676251e-02 -4.07501645e-02 -7.33955264e-01 1.02894628e+00 8.91150534e-01 -1.19785033e-01 -1.21080928e-01 -6.55799285e-02 -5.31579912e-01 3.93179417e-01 6.13461137e-01 -7.24133313e-01 -2.81761885e-01 -2.12884098e-01 3.73095810e-01 -3.18274558e-01 2.62022406e-01 -6.96299732e-01 -2.31124517e-02 -2.41911963e-01 5.53404450e-01 -3.46232921e-01 3.23915273e-01 -4.77688938e-01 -1.08214840e-01 3.17052335e-01 -5.76229393e-01 -5.53543687e-01 2.71250252e-02 9.86971200e-01 4.59053248e-01 -4.49396342e-01 8.36554050e-01 -4.75567020e-02 -8.68729949e-02 5.53553045e-01 -3.58465016e-01 1.45097688e-01 8.58205676e-01 2.50617743e-01 -1.30702350e-02 -1.19419575e-01 -9.49871957e-01 3.13186109e-01 4.77138221e-01 -2.83742677e-02 4.02990818e-01 -1.20895636e+00 -7.53249288e-01 2.24401262e-02 -8.83764494e-03 3.77389401e-01 -3.20518129e-02 6.96559608e-01 -3.71758431e-01 6.37815416e-01 7.31473714e-02 -5.31041980e-01 -7.59759605e-01 8.73302281e-01 3.79985660e-01 -3.79128218e-01 -5.22840679e-01 1.04243886e+00 1.55768096e-02 -6.18564427e-01 2.05412731e-01 -2.05064744e-01 2.35948563e-01 5.51887639e-02 3.91651452e-01 5.08782744e-01 -8.11307039e-03 -4.64277230e-02 -1.32762611e-01 1.63828209e-02 -2.74532318e-01 -5.45438111e-01 1.45490539e+00 1.32318392e-01 1.97905242e-01 3.76671463e-01 1.36638987e+00 -2.75525153e-01 -2.01687121e+00 -3.15607607e-01 1.31318912e-01 -1.95764810e-01 5.11641264e-01 -7.88921177e-01 -3.73492628e-01 1.07256937e+00 6.95522547e-01 -9.92858708e-02 4.40910429e-01 3.75904106e-02 6.10105515e-01 7.68094122e-01 1.68665841e-01 -1.25943291e+00 -3.69309820e-02 4.79493439e-01 7.81185508e-01 -1.30294430e+00 9.48307291e-02 2.95160055e-01 -7.10813522e-01 8.91351998e-01 -1.56084895e-01 -3.97112779e-02 5.89342415e-01 5.17524220e-03 -3.71548772e-01 -3.15063864e-01 -8.82000446e-01 -1.65068323e-03 3.93980086e-01 7.80573308e-01 1.92501456e-01 8.14985484e-02 -3.89226899e-02 5.36338151e-01 -1.61567107e-01 -1.66782424e-01 2.77599454e-01 9.21189785e-01 -2.86124259e-01 -1.29062378e+00 -4.73441370e-02 5.84762752e-01 -6.20017648e-01 -3.68555218e-01 -2.73045629e-01 3.30347717e-01 -2.95246840e-01 7.92882264e-01 -9.13202614e-02 1.38487406e-02 -2.80677211e-02 3.86711150e-01 6.73002005e-01 -4.70373213e-01 -2.40938440e-01 1.91390947e-01 -8.60870108e-02 -8.84896934e-01 -1.75783813e-01 -8.03037345e-01 -8.33591461e-01 -2.94309616e-01 -2.31164530e-01 2.19122857e-01 9.28894818e-01 9.20173466e-01 4.77414489e-01 6.01640999e-01 6.41583920e-01 -1.04809225e+00 -1.31762731e+00 -1.18243098e+00 -4.63602215e-01 3.56765866e-01 4.64626670e-01 -6.77497685e-01 -5.11044145e-01 -2.39954904e-01]	[6.75409460067749, 4.031822681427002]
5590f7a0-519d-4ce7-8d4a-75ca994d99b2	ovenet-offset-vector-network-for-semantic	2303.14516	null	https://arxiv.org/abs/2303.14516v1	https://arxiv.org/pdf/2303.14516v1.pdf	OVeNet: Offset Vector Network for Semantic Segmentation	Semantic segmentation is a fundamental task in visual scene understanding. We focus on the supervised setting, where ground-truth semantic annotations are available. Based on knowledge about the high regularity of real-world scenes, we propose a method for improving class predictions by learning to selectively exploit information from neighboring pixels. In particular, our method is based on the prior that for each pixel, there is a seed pixel in its close neighborhood sharing the same prediction with the former. Motivated by this prior, we design a novel two-head network, named Offset Vector Network (OVeNet), which generates both standard semantic predictions and a dense 2D offset vector field indicating the offset from each pixel to the respective seed pixel, which is used to compute an alternative, seed-based semantic prediction. The two predictions are adaptively fused at each pixel using a learnt dense confidence map for the predicted offset vector field. We supervise offset vectors indirectly via optimizing the seed-based prediction and via a novel loss on the confidence map. Compared to the baseline state-of-the-art architectures HRNet and HRNet+OCR on which OVeNet is built, the latter achieves significant performance gains on two prominent benchmarks for semantic segmentation of driving scenes, namely Cityscapes and ACDC. Code is available at https://github.com/stamatisalex/OVeNet	['Petros Maragos', 'Christos Sakaridis', 'Stamatis Alexandropoulos']	2023-03-25	null	null	null	null	['optical-character-recognition']	['computer-vision']	[ 5.14363170e-01 4.69222397e-01 -3.50972295e-01 -7.72054732e-01 -6.10366404e-01 -3.32030743e-01 4.88224119e-01 2.73069352e-01 -6.10958755e-01 4.75112915e-01 6.43885054e-04 -5.79598993e-02 2.60365635e-01 -8.28319013e-01 -1.01659369e+00 -7.01899946e-01 3.16338032e-01 4.07696784e-01 6.45450592e-01 -1.33957639e-01 2.58054763e-01 1.73650384e-01 -1.47301650e+00 2.73621947e-01 1.01405263e+00 1.33128166e+00 5.52424848e-01 4.04556602e-01 -1.27351388e-01 5.33490777e-01 -2.14564204e-01 -4.19836998e-01 3.76612872e-01 -2.28938282e-01 -8.79600585e-01 1.60585910e-01 5.12793481e-01 2.56350115e-02 -2.62647778e-01 1.32052720e+00 7.34786540e-02 3.68424565e-01 4.86002803e-01 -1.12607777e+00 -2.94216841e-01 5.35478354e-01 -6.65472984e-01 1.39009520e-01 -3.06238443e-01 1.66907713e-01 1.23960924e+00 -6.87284827e-01 5.22047400e-01 9.00282145e-01 6.44336104e-01 3.77953440e-01 -1.21806347e+00 -5.64307332e-01 6.31903648e-01 4.07606602e-01 -1.42649150e+00 -1.25209391e-01 8.06936085e-01 -4.33919877e-01 6.38186574e-01 9.02127847e-03 5.67294836e-01 8.13731313e-01 -1.00305498e-01 1.00340927e+00 9.42620516e-01 -1.39107063e-01 3.75694454e-01 2.72373170e-01 1.99532270e-01 7.19898224e-01 -1.33950666e-01 1.22131795e-01 -4.98300880e-01 2.95264095e-01 5.48386335e-01 -3.20542301e-03 -3.84507328e-01 -5.32503724e-01 -9.17003632e-01 9.05529618e-01 1.02873445e+00 3.44486162e-02 -3.98211896e-01 1.92958981e-01 1.28508106e-01 -2.66776472e-01 5.34171760e-01 1.43014833e-01 -6.42025948e-01 2.43336469e-01 -1.09122658e+00 1.77771792e-01 4.68293697e-01 7.48032510e-01 1.32025135e+00 -2.86050230e-01 -3.31284702e-01 8.42178762e-01 3.33377212e-01 2.46376202e-01 4.10353899e-01 -1.01933682e+00 5.63716412e-01 5.76562583e-01 1.69084817e-01 -9.05412912e-01 -3.10123742e-01 -7.05079496e-01 -6.48804307e-01 2.68758059e-01 3.72376621e-01 4.49084975e-02 -1.42742360e+00 1.66452622e+00 5.20330310e-01 8.29913080e-01 -2.07917020e-03 1.12234497e+00 5.78498542e-01 7.50242352e-01 1.71324745e-01 3.35833788e-01 1.00474977e+00 -1.38585424e+00 -1.49809718e-01 -6.98297262e-01 5.56892157e-01 -4.80366796e-01 9.55119550e-01 2.70381808e-01 -5.79873383e-01 -7.52256095e-01 -1.03252912e+00 -1.29400790e-01 -4.34481055e-01 2.22067446e-01 3.60531062e-01 3.40308487e-01 -1.03046238e+00 7.23636150e-01 -9.96013582e-01 -1.28370941e-01 7.25692213e-01 8.48450959e-02 -1.56048730e-01 -8.45813677e-02 -9.85150635e-01 6.80196404e-01 6.13794506e-01 2.72077411e-01 -7.47439504e-01 -7.54810452e-01 -1.06833887e+00 -5.28240502e-02 4.28948909e-01 -4.71426308e-01 1.13416243e+00 -1.31889045e+00 -1.34800637e+00 9.65501785e-01 -3.27314109e-01 -9.27700043e-01 6.37989223e-01 -2.84166694e-01 -1.30494386e-01 1.39777690e-01 5.26588976e-01 1.17066193e+00 9.29884732e-01 -1.42084670e+00 -9.52618480e-01 -3.27605009e-01 -1.34468928e-01 2.82001078e-01 8.91204178e-02 -5.00583649e-01 -9.26270306e-01 -6.17093205e-01 2.11406007e-01 -8.23615074e-01 -6.26546144e-01 1.12522498e-01 -7.33983755e-01 -6.92766234e-02 8.62347484e-01 -7.31666148e-01 9.06311274e-01 -2.27096343e+00 5.64720407e-02 4.33181971e-01 2.03824863e-01 3.88049006e-01 -4.19886224e-02 -2.72879004e-01 1.17907241e-01 -1.25989988e-01 -8.14236164e-01 -5.92997909e-01 -1.64274678e-01 2.88042009e-01 -3.59450698e-01 4.51640934e-01 3.25249076e-01 8.54043663e-01 -9.38660204e-01 -3.06328267e-01 4.74519491e-01 5.19453824e-01 -5.81268311e-01 8.18517655e-02 -5.51087260e-01 5.91103792e-01 -4.83035564e-01 2.88042217e-01 8.00509930e-01 -4.24788922e-01 -1.24592513e-01 -2.10100532e-01 -1.61762625e-01 1.84220389e-01 -1.12896669e+00 1.73683977e+00 -3.72454852e-01 6.34624720e-01 -1.21058583e-01 -1.27048516e+00 1.00781989e+00 -1.61998138e-01 2.02207491e-01 -7.73874462e-01 9.01151523e-02 1.37273341e-01 -4.42358911e-01 -1.01283051e-01 5.93223751e-01 5.36845438e-02 1.25128245e-02 -1.39173791e-01 4.06880639e-02 -1.66819111e-01 -2.87665967e-02 1.55752197e-01 7.32038856e-01 3.01757991e-01 -1.26152374e-02 -2.34007895e-01 5.30495644e-01 1.77522197e-01 8.53883862e-01 7.56149411e-01 -2.52428055e-01 9.65208650e-01 4.09901232e-01 -3.86544555e-01 -9.09393191e-01 -1.10523081e+00 -2.32413173e-01 9.42568481e-01 7.58561313e-01 -2.82100439e-01 -8.95517528e-01 -9.12547708e-01 1.64684858e-02 8.80030811e-01 -7.44576275e-01 -2.41479743e-02 -4.88019407e-01 -5.00594974e-01 1.00187309e-01 7.58911371e-01 9.13332641e-01 -8.76366615e-01 -5.38475394e-01 1.94961876e-01 -2.04878345e-01 -1.27637005e+00 -5.36103964e-01 3.89214694e-01 -6.32635832e-01 -9.10678029e-01 -5.65165758e-01 -6.77572787e-01 7.00032711e-01 2.20001891e-01 1.01569605e+00 4.05528471e-02 -1.55826226e-01 -2.24100496e-03 -2.93533713e-01 -2.66286135e-01 -6.57678396e-02 1.24072023e-01 -5.51687241e-01 4.28767741e-01 2.15842202e-01 -1.97014630e-01 -8.58640254e-01 3.20423722e-01 -6.21879995e-01 4.21342343e-01 4.57139730e-01 8.02387238e-01 1.16028678e+00 1.32793859e-01 2.26014555e-01 -1.11474931e+00 -3.05971235e-01 -5.15859425e-01 -9.20558751e-01 3.92563194e-02 -4.50788617e-01 2.06345409e-01 5.52671015e-01 -8.29426944e-03 -1.03742552e+00 5.14941096e-01 -3.39987993e-01 -4.45249826e-01 -3.62522870e-01 3.10261756e-01 -1.88349798e-01 8.51185620e-02 3.86777997e-01 1.07035905e-01 -3.12307775e-01 -4.54708844e-01 4.75956589e-01 4.33195263e-01 9.15251553e-01 -2.47766763e-01 7.04678774e-01 6.50061488e-01 -2.46827230e-01 -6.54817224e-01 -1.32644403e+00 -7.20770299e-01 -6.30809009e-01 -5.56452498e-02 1.29619980e+00 -1.03173351e+00 -1.75776929e-01 5.99538624e-01 -1.03987646e+00 -7.91623950e-01 -2.85690725e-01 3.72531801e-01 -5.58283210e-01 8.28425139e-02 -2.45760754e-01 -4.48465765e-01 -1.33281931e-01 -1.24064434e+00 1.29081166e+00 4.40955281e-01 2.27221549e-02 -9.69004393e-01 -3.02082181e-01 5.10163665e-01 8.53973404e-02 2.07758024e-01 6.80094600e-01 -5.15590966e-01 -7.55776823e-01 -5.51701449e-02 -5.43707371e-01 6.43369496e-01 -1.70338795e-01 -2.31659546e-01 -1.09211135e+00 2.15650573e-02 -2.76021212e-01 -5.07494882e-02 1.45756626e+00 6.39301062e-01 1.59823811e+00 -1.70531701e-02 -5.70512176e-01 8.82428885e-01 1.53079116e+00 -1.47058949e-01 5.46502829e-01 2.78769225e-01 1.08033991e+00 5.45647860e-01 8.78915429e-01 3.07852417e-01 5.63241422e-01 7.55019546e-01 7.47151732e-01 -2.47385398e-01 -1.27547458e-01 -4.33480710e-01 -7.12484494e-02 6.52912781e-02 1.28925294e-01 -2.12957069e-01 -8.47114444e-01 7.51278698e-01 -2.03302312e+00 -5.24233937e-01 -1.34909496e-01 2.45170331e+00 7.69330561e-01 4.51892853e-01 -9.93948057e-02 -3.05453464e-02 8.91239345e-01 3.79603148e-01 -8.08974683e-01 6.61555231e-02 -3.72015201e-02 3.38837206e-01 9.45714116e-01 7.60062218e-01 -1.40911829e+00 1.39821959e+00 4.55914640e+00 8.88006389e-01 -1.30879450e+00 2.25086853e-01 1.11587274e+00 1.12868048e-01 -2.40694344e-01 1.31690025e-01 -1.13787019e+00 5.95329046e-01 6.51228547e-01 2.73416102e-01 6.78651258e-02 9.16369081e-01 2.30491608e-01 -2.94915646e-01 -8.21887195e-01 6.54305577e-01 -1.32130593e-01 -1.56582737e+00 -8.99846181e-02 -2.21789837e-01 8.52272570e-01 4.48879033e-01 1.05160080e-01 2.53925994e-02 4.62044090e-01 -9.36206222e-01 9.89010334e-01 2.71783173e-01 5.84967792e-01 -8.02327275e-01 6.61357582e-01 2.52356440e-01 -1.27289236e+00 -2.36436259e-02 -3.23117912e-01 2.36367270e-01 1.51397318e-01 8.31679165e-01 -7.73424923e-01 3.14674944e-01 9.19485390e-01 1.11195087e+00 -5.05168438e-01 1.12625861e+00 -6.78988278e-01 7.89523065e-01 -3.30575347e-01 4.42872316e-01 6.02503777e-01 -2.05592066e-01 4.31979448e-01 1.15188503e+00 7.49521330e-02 -1.30954713e-01 3.50153446e-01 9.75821257e-01 -1.24099284e-01 -1.77395701e-01 -2.08645985e-01 4.69726115e-01 4.84756798e-01 1.24653530e+00 -1.06334245e+00 -4.63610530e-01 -3.54814798e-01 1.24478269e+00 4.09736633e-01 4.95224565e-01 -9.49644208e-01 -2.77357340e-01 8.90617788e-01 1.64000973e-01 8.05924594e-01 2.18564067e-02 -5.67443192e-01 -9.12211537e-01 2.78762579e-02 -2.51881063e-01 3.43771935e-01 -7.37772226e-01 -1.01951945e+00 6.07195497e-01 -1.48346454e-01 -1.02093017e+00 1.11777440e-01 -5.37992656e-01 -8.17310512e-01 9.22182322e-01 -1.91590965e+00 -1.02239072e+00 -4.93976951e-01 4.66813087e-01 6.78576827e-01 2.24709272e-01 2.92171985e-01 7.13924319e-02 -6.09360993e-01 4.50434834e-01 -4.36570458e-02 1.07770137e-01 4.36348140e-01 -1.36790478e+00 5.47543585e-01 9.81050491e-01 2.97286868e-01 -6.14024028e-02 7.00299025e-01 -5.54198325e-01 -5.40237725e-01 -1.76547611e+00 7.70591021e-01 -2.46137306e-01 5.26033342e-01 -3.64470989e-01 -9.98814285e-01 7.53304839e-01 -1.61521971e-01 4.75013524e-01 2.61878930e-02 -2.43563905e-01 -3.84788007e-01 -1.98893920e-01 -9.89479244e-01 4.55409169e-01 1.00489199e+00 -3.66655678e-01 -2.50717372e-01 3.68762404e-01 9.87469971e-01 -5.44626176e-01 -3.26804042e-01 4.61058199e-01 1.22519754e-01 -9.82080817e-01 1.06911671e+00 -1.57439128e-01 4.68893230e-01 -5.74609160e-01 -8.41218159e-02 -1.38854551e+00 -6.18488081e-02 -2.66606122e-01 3.79333556e-01 9.05949950e-01 6.57988727e-01 -6.29215777e-01 1.11953843e+00 5.23177207e-01 -5.14082491e-01 -8.66115808e-01 -8.58285367e-01 -6.34129226e-01 -3.44389267e-02 -6.84554935e-01 5.14031649e-01 6.78237557e-01 -7.22363114e-01 2.37014219e-01 -1.35260105e-01 6.27460778e-01 6.54725015e-01 1.37341261e-01 5.47693491e-01 -1.11430442e+00 -4.04846907e-01 -4.04654503e-01 -6.95656002e-01 -1.37114465e+00 4.24778193e-01 -9.41188633e-01 4.56369728e-01 -1.55224299e+00 -2.07172167e-02 -6.44555509e-01 -3.58179122e-01 6.42788172e-01 -3.98175657e-01 5.75550437e-01 1.18539684e-01 -1.33120930e-02 -7.37443388e-01 6.97851837e-01 1.04552352e+00 -2.18550861e-01 -3.95107687e-01 2.66723305e-01 -5.94933271e-01 9.41710413e-01 8.65037620e-01 -4.91270870e-01 -4.26448375e-01 -4.27196473e-01 -1.02099419e-01 -2.63445765e-01 7.58533776e-01 -1.13526094e+00 2.04087064e-01 -8.53948668e-02 3.14768165e-01 -6.84607685e-01 3.69047523e-01 -6.49865806e-01 -1.78691834e-01 2.31170535e-01 -5.00311255e-01 -4.86093283e-01 6.65080994e-02 8.00876200e-01 -3.97990197e-01 -2.72705138e-01 1.08475327e+00 4.71345633e-02 -1.34189916e+00 5.00392139e-01 3.35799232e-02 2.04400361e-01 1.00762010e+00 -2.76199132e-01 -2.48821586e-01 -1.24966197e-01 -7.37469792e-01 4.23707157e-01 4.33047354e-01 3.66238058e-01 6.06626332e-01 -9.50590253e-01 -5.15155971e-01 2.53659129e-01 2.82336384e-01 5.01372099e-01 3.36107284e-01 7.39344716e-01 -4.15119290e-01 1.78326845e-01 4.50542159e-02 -8.60515714e-01 -1.02488184e+00 3.33557874e-01 4.95382607e-01 -7.14241564e-02 -8.17903817e-01 1.27618694e+00 6.94990933e-01 -3.55548233e-01 1.66050687e-01 -4.45439070e-01 -2.05773935e-01 -2.68485188e-01 3.83167058e-01 -1.42215658e-02 1.34808317e-01 -8.81558299e-01 -2.20295072e-01 5.52309394e-01 -8.50685313e-02 -4.54661176e-02 1.24132681e+00 -2.36216336e-01 1.33458391e-01 3.26688558e-01 1.27748668e+00 -3.61345470e-01 -1.94250238e+00 -5.69710314e-01 7.55021051e-02 -6.09969020e-01 4.21200931e-01 -7.09854543e-01 -1.55134308e+00 7.70396411e-01 5.74943960e-01 -2.68326640e-01 1.06420159e+00 3.53078663e-01 8.90135586e-01 4.53054905e-02 1.21470600e-01 -1.20712364e+00 -6.23719171e-02 4.87226605e-01 5.67937016e-01 -1.55011010e+00 -3.22536260e-01 -6.91157758e-01 -9.60305095e-01 6.99424267e-01 7.11847365e-01 -2.06012189e-01 7.44733274e-01 -2.28004865e-02 1.17045216e-01 -1.13317892e-01 -3.90321553e-01 -5.31955659e-01 4.12842155e-01 5.84861040e-01 1.15063034e-01 2.94189245e-01 1.29569173e-01 5.06673455e-01 -3.13392878e-01 -1.75949559e-01 1.90307751e-01 6.38217926e-01 -6.41096056e-01 -8.56530070e-01 -1.03519611e-01 5.83484530e-01 -1.70647964e-01 -2.38634989e-01 -7.20159709e-02 4.20446873e-01 3.35173458e-01 7.47209787e-01 4.32282984e-01 -3.76365095e-01 1.05637930e-01 -1.59661934e-01 6.78965300e-02 -7.45040476e-01 -2.20490888e-01 -1.10603265e-01 -1.09316163e-01 -8.67572188e-01 -2.81021059e-01 -6.26458466e-01 -1.57346654e+00 1.70878336e-01 -1.87689841e-01 6.99096248e-02 6.99009061e-01 1.12570179e+00 2.90482104e-01 3.91314924e-01 4.76509660e-01 -1.00913858e+00 -1.61815003e-01 -5.92938185e-01 -5.91770709e-01 3.62749577e-01 3.66172552e-01 -6.42087162e-01 -3.28190356e-01 1.61823243e-01]	[9.54838752746582, 0.43084728717803955]
29a67c63-0083-4660-ae75-60b0d4280246	encoder-decoder-approach-to-automated-essay	null	null	https://aclanthology.org/2021.icon-main.48	https://aclanthology.org/2021.icon-main.48.pdf	Encoder Decoder Approach to Automated Essay Scoring For Deeper Semantic Analysis	Descriptive or essay type of answers have always played a major role in education. They clearly capture the student’s grasp on knowledge and presentation skills. Manual essay scoring can be a daunting process to human evaluators; assessing descriptive answers can present a huge overhead owing to limited numbers of evaluators and an out of proportional number of essays to be graded hence leading to an inefficient or an inaccurate score. There has been a major shift in paradigm from traditional classroom education to online education engendered by COVID-19 pandemic; it seems plausible to infer that future assessment of education shall be online, making the solution of automatic essay scorer not only relevant, but of paramount importance. We explore several neural architecture models for the task of automated essay scoring system. Results and Experimental analysis exhibit that our model based on recurrent encoder-decoder provides for a deeper semantic analysis hence, outperforming a strong baseline in terms of quadratic weighted kappa score.	['Srujana Inturi', 'Sreedeep Rayavarapu', 'Priyatam Naravajhula']	null	null	null	null	icon-2021-12	['automated-essay-scoring']	['natural-language-processing']	[ 4.05702740e-02 3.05071026e-01 1.68703288e-01 -4.79043990e-01 -9.23029304e-01 -7.04259098e-01 4.54096317e-01 6.95807755e-01 -6.09516621e-01 6.62797868e-01 2.71765351e-01 -6.22342646e-01 -4.72855657e-01 -8.76001716e-01 -1.96400076e-01 -1.62528828e-01 6.21237040e-01 5.41126728e-01 2.92800218e-01 -4.20699030e-01 7.64591694e-01 -5.58360443e-02 -1.69396651e+00 2.05649227e-01 7.70418227e-01 7.98002839e-01 2.12362051e-01 8.36171389e-01 -5.85223675e-01 1.33447456e+00 -1.00327301e+00 -1.08342707e+00 -1.92761093e-01 -6.32129669e-01 -9.80508387e-01 -4.43706304e-01 7.29718447e-01 -4.78066623e-01 1.97052047e-01 1.19720018e+00 4.76195157e-01 3.28132600e-01 7.45331705e-01 -6.90651774e-01 -1.02928042e+00 6.62769079e-01 -2.81235397e-01 2.24136025e-01 7.83731401e-01 -1.88864589e-01 1.04904389e+00 -7.71838903e-01 3.70619118e-01 6.89695060e-01 6.45182490e-01 5.33131123e-01 -8.72596681e-01 -5.09474099e-01 -2.44942382e-01 4.07396764e-01 -7.41033375e-01 -2.15818807e-01 6.35504842e-01 -7.06234574e-01 4.61582720e-01 3.02288800e-01 8.57208073e-01 8.08125436e-01 7.04594925e-02 6.71814263e-01 1.42503941e+00 -3.93830806e-01 1.56257644e-01 4.81142908e-01 5.90335727e-01 7.61583209e-01 2.12270975e-01 -2.38944441e-01 -8.06775987e-01 3.73528779e-01 4.05227751e-01 -5.75982779e-02 -1.23102829e-01 -5.99460974e-02 -7.39062667e-01 8.27877104e-01 1.34826839e-01 3.75696868e-02 -7.01464415e-02 -5.37351333e-02 6.14199281e-01 1.08726704e+00 3.53814997e-02 7.77530909e-01 -2.65918195e-01 -7.25823283e-01 -1.32765543e+00 3.25397432e-01 9.39391613e-01 4.70803767e-01 2.99786866e-01 -5.91607355e-02 -1.48194805e-01 7.38709569e-01 2.23303840e-01 1.19270690e-01 8.41491103e-01 -7.51490951e-01 4.14109945e-01 1.10841298e+00 -6.00155778e-02 -1.05994356e+00 -1.33191392e-01 -5.10573745e-01 -4.21122342e-01 5.87089479e-01 7.15292156e-01 -1.05252555e-02 -5.20488679e-01 1.48254585e+00 2.70133931e-02 -3.49007696e-01 -1.44236758e-01 1.02347386e+00 1.07969272e+00 3.03990901e-01 1.44624665e-01 7.76619911e-02 1.42965567e+00 -7.41481066e-01 -7.73925304e-01 8.22595656e-02 6.33608878e-01 -1.05657864e+00 1.29185367e+00 6.30103588e-01 -1.51394069e+00 -5.96414506e-01 -1.11834621e+00 -4.56716895e-01 -5.71198940e-01 1.38594642e-01 3.31752062e-01 9.88204360e-01 -1.08311045e+00 7.44716465e-01 -3.21720749e-01 -1.89943656e-01 4.47295129e-01 3.68737966e-01 -1.68283090e-01 1.04269147e-01 -1.01033640e+00 1.21152210e+00 -6.02828078e-02 -1.60611495e-01 -5.00476360e-01 -8.55390429e-01 -5.54921687e-01 4.84158218e-01 1.35244653e-01 -4.95884925e-01 1.47885549e+00 -7.69790947e-01 -1.80667114e+00 1.02907777e+00 1.78102836e-01 -2.37000003e-01 7.21083283e-01 -2.60894716e-01 1.26973623e-02 -5.69092669e-02 -1.32556900e-01 3.81467074e-01 4.32836562e-01 -4.00129348e-01 -3.93379688e-01 -4.18575287e-01 2.90614218e-01 4.13765758e-01 -8.20325494e-01 2.16968089e-01 4.08343494e-01 -3.42452437e-01 1.50903329e-01 -5.06752729e-01 1.37446687e-01 -1.32520288e-01 1.85628042e-01 -5.95993876e-01 4.23222005e-01 -9.09841597e-01 1.56725454e+00 -1.55923009e+00 1.03372969e-02 -1.56899959e-01 3.28470349e-01 3.82264286e-01 1.65960416e-01 4.72516775e-01 1.29687160e-01 -3.68980095e-02 5.76451570e-02 -1.70390353e-01 4.43724781e-01 -4.47759420e-01 -2.62082666e-01 2.24954300e-02 -8.55609849e-02 7.95942128e-01 -1.04845858e+00 -4.94684339e-01 6.03982545e-02 2.49483883e-01 -4.48398679e-01 4.77418602e-01 3.15834209e-02 -2.35134512e-02 -4.54049945e-01 5.29978454e-01 1.39483690e-01 -3.35142165e-01 -4.85798791e-02 4.20120358e-01 -1.10790946e-01 8.68561506e-01 -1.03725171e+00 1.63684630e+00 -6.42959654e-01 7.48671174e-01 -1.60642862e-01 -8.55354011e-01 1.19931412e+00 5.00382185e-01 -8.82495940e-02 -8.06654334e-01 9.53569189e-02 3.83115560e-01 1.55562326e-01 -4.71461564e-01 8.90983760e-01 -2.57841766e-01 -5.99781051e-02 1.02268708e+00 1.61706746e-01 -4.16559756e-01 2.82642841e-01 2.84524173e-01 1.16530621e+00 4.29282159e-01 1.74900442e-01 -3.78696889e-01 6.42192423e-01 -2.15863377e-01 -9.34178662e-03 5.30942738e-01 -4.91540223e-01 4.57982242e-01 5.99975765e-01 -5.40713966e-01 -8.32140148e-01 -8.52267265e-01 4.03384566e-02 1.24705887e+00 -4.15469319e-01 -2.12844580e-01 -9.16920364e-01 -5.83074868e-01 -3.31356078e-01 4.62020069e-01 -5.22978485e-01 -1.15517087e-01 -3.47373515e-01 -2.98418462e-01 4.27128524e-01 3.51928145e-01 4.58622217e-01 -1.02365065e+00 -1.16275108e+00 2.86307335e-01 -7.40325078e-02 -6.83950484e-01 -3.01087290e-01 2.90303499e-01 -6.97487354e-01 -8.90910208e-01 -9.03036892e-01 -7.41710007e-01 4.42522824e-01 4.81724292e-02 1.44593894e+00 2.48147160e-01 -8.19607377e-02 4.77342188e-01 -1.75927967e-01 -5.58821440e-01 -3.12596947e-01 2.32726246e-01 -2.26659343e-01 -4.92486835e-01 1.03597677e+00 -5.29703259e-01 -9.53857601e-01 -3.78980339e-02 -6.50735497e-01 8.34989399e-02 4.17477250e-01 6.88866794e-01 -5.05107269e-02 -5.17726958e-01 1.03217530e+00 -9.26168263e-01 1.16538763e+00 -3.12893271e-01 -5.14343798e-01 1.99196354e-01 -8.21559966e-01 1.65105425e-02 7.42716134e-01 -1.44588396e-01 -1.14774048e+00 -3.43985468e-01 -2.74749219e-01 7.36599192e-02 -3.83755714e-02 7.75144875e-01 4.60371315e-01 -1.56385496e-01 6.29886210e-01 1.70381933e-01 -2.46163905e-02 -2.93355644e-01 -1.02660758e-03 6.44765317e-01 3.45298439e-01 -6.91228271e-01 3.96090090e-01 -1.02042563e-01 -9.06393863e-03 -6.05265677e-01 -1.24199283e+00 -5.38443148e-01 -4.28096890e-01 -7.28747725e-01 6.75129652e-01 -6.38699532e-01 -1.11756909e+00 7.21367300e-02 -1.01782644e+00 -2.16416031e-01 -2.26691231e-01 4.16144580e-01 -3.35600436e-01 2.99872965e-01 -7.83582807e-01 -7.83601344e-01 -2.58415103e-01 -1.22410643e+00 4.25548404e-01 4.61291820e-01 -8.00056458e-01 -1.14109802e+00 3.61448646e-01 1.18879306e+00 6.66304171e-01 -8.74013603e-02 1.01321399e+00 -6.63896382e-01 -4.47040081e-01 -3.48968536e-01 -2.99528539e-01 2.53792852e-01 -3.79269868e-01 -1.63933232e-01 -1.24566185e+00 9.33001116e-02 1.59815684e-01 -1.00610697e+00 6.69578552e-01 -1.18535519e-01 1.01166213e+00 -1.37855440e-01 5.73764741e-01 1.44134304e-02 1.42612374e+00 -1.07202530e-01 4.18691039e-01 2.80188143e-01 4.03085113e-01 1.01859486e+00 2.95671821e-01 1.90399900e-01 6.79076314e-01 5.86682439e-01 4.48764451e-02 5.03832281e-01 -4.20924842e-01 -2.78123975e-01 4.60854501e-01 1.71905327e+00 -2.00977489e-01 1.64594024e-01 -9.46321607e-01 7.94527590e-01 -1.47590470e+00 -1.10704792e+00 -4.58254784e-01 2.50287104e+00 1.16518271e+00 5.99853322e-02 6.74879774e-02 4.04601961e-01 2.61609048e-01 -4.05402780e-02 2.24330295e-02 -1.12013280e+00 3.44757378e-01 6.71830654e-01 3.27706747e-02 6.18506968e-01 -4.71212685e-01 4.87277865e-01 5.89431810e+00 6.69110656e-01 -9.37828958e-01 8.86564553e-02 3.38083237e-01 -2.01664850e-01 -3.55028838e-01 -2.32410192e-01 -7.01304018e-01 7.10611463e-01 1.35772431e+00 -2.53370702e-01 1.22360922e-01 5.96742809e-01 -3.98584493e-02 -3.06982636e-01 -1.15278232e+00 7.60163724e-01 1.09311558e-01 -1.03580737e+00 1.53749064e-02 -9.56380293e-02 8.59517872e-01 -3.69165540e-01 2.60763913e-01 5.83235919e-01 5.17987907e-01 -1.30966222e+00 6.16929948e-01 5.61290681e-01 5.14192700e-01 -7.63039947e-01 8.17405462e-01 3.31577778e-01 -7.86450624e-01 -7.62849301e-02 -5.36697626e-01 -7.51691937e-01 -2.93008685e-01 3.38573247e-01 -7.57742226e-01 -4.59780321e-02 4.13477689e-01 2.01566085e-01 -6.42717540e-01 9.41833436e-01 -4.75253165e-01 8.15983713e-01 5.57201654e-02 -6.71256959e-01 2.37337694e-01 -3.41274559e-01 8.39944780e-02 1.07482874e+00 4.93235230e-01 1.92016825e-01 -1.12045795e-01 7.19906688e-01 -2.16526911e-01 3.10288817e-01 -4.33541000e-01 -2.92595811e-02 5.26875019e-01 1.28245866e+00 -7.20289767e-01 -1.87107459e-01 -6.11091673e-01 9.07127082e-01 6.02682829e-01 2.58372091e-02 -4.40199137e-01 -4.61856037e-01 3.19120049e-01 2.17365652e-01 2.05805991e-02 1.88406277e-02 -6.64343894e-01 -9.06068087e-01 5.16512357e-02 -8.74535322e-01 4.23350543e-01 -6.31112933e-01 -8.87764156e-01 2.59348601e-01 -6.78444505e-01 -8.69224072e-01 -3.11156303e-01 -9.10633147e-01 -6.65870190e-01 8.85550618e-01 -1.48432171e+00 -6.23038590e-01 -4.15150732e-01 2.05913022e-01 6.23462498e-01 -1.71052694e-01 1.00457013e+00 5.81181705e-01 -2.63055027e-01 8.38710725e-01 -9.28070992e-02 9.46584642e-02 8.23423982e-01 -1.73097193e+00 -2.41442055e-01 6.11521602e-01 1.18996151e-01 5.81415534e-01 7.53591478e-01 -2.08854511e-01 -1.22361481e+00 -4.66530651e-01 1.46824646e+00 -9.77881432e-01 6.49840832e-01 -8.63303547e-04 -1.01835167e+00 1.97544500e-01 3.98596913e-01 -7.66407967e-01 1.12812674e+00 1.88323542e-01 -2.94974506e-01 -2.04621367e-02 -8.55910778e-01 5.96428990e-01 5.71187615e-01 -9.11796570e-01 -9.52481806e-01 1.75185144e-01 1.80958927e-01 -2.87968427e-01 -1.03732991e+00 -1.95445508e-01 7.16863453e-01 -1.27115595e+00 6.68006897e-01 -5.00280917e-01 1.35588121e+00 -5.58911124e-03 3.84541005e-01 -1.19207954e+00 4.25055921e-02 -4.06681478e-01 -1.58506960e-01 1.09963071e+00 4.13082272e-01 -1.72782958e-01 1.10483134e+00 8.14462364e-01 -1.20173432e-01 -9.59349930e-01 -7.06946909e-01 -3.79888445e-01 4.41262662e-01 -2.67439306e-01 5.05350590e-01 1.18375206e+00 5.61615884e-01 6.41133189e-01 1.08589500e-01 -4.88704205e-01 5.30042946e-01 3.14336747e-01 4.60445493e-01 -1.53410041e+00 -1.93647608e-01 -1.05780292e+00 -5.18454790e-01 -8.52359354e-01 -5.58815598e-02 -9.96809304e-01 -2.12393209e-01 -1.47849739e+00 3.62392306e-01 -1.64798480e-02 -3.80650014e-01 1.24022096e-01 -2.13449553e-01 4.76586282e-01 -1.27682149e-01 -1.97408333e-01 -8.13331783e-01 1.90022752e-01 1.33392954e+00 7.31577724e-02 2.41798908e-01 -3.71900983e-02 -7.78341591e-01 7.33548403e-01 6.19821370e-01 -3.43352288e-01 -5.91076732e-01 -5.16155183e-01 1.24203789e+00 2.78869718e-01 2.11369529e-01 -9.10576165e-01 5.76906323e-01 9.47609730e-03 3.72549087e-01 -1.88136935e-01 1.91605035e-02 -5.91491401e-01 -5.34935176e-01 2.62997419e-01 -7.65228927e-01 2.77717113e-01 -1.24576569e-01 3.99296954e-02 -6.00006580e-01 -9.34088528e-01 6.65460408e-01 -3.55830848e-01 -4.20765221e-01 -2.01266065e-01 -3.12040508e-01 6.44094869e-02 6.18042290e-01 -6.71311319e-01 -2.18490258e-01 -4.64039296e-01 -5.31577170e-01 1.48990065e-01 2.62483805e-01 2.71855384e-01 5.87876201e-01 -1.05284154e+00 -7.64169872e-01 -1.59406990e-01 -6.94543347e-02 -1.99772850e-01 2.18399048e-01 8.39570940e-01 -6.70110881e-01 7.40935385e-01 -3.54799300e-01 -2.22490393e-02 -1.31429100e+00 4.50531431e-02 1.10736176e-01 -4.74153548e-01 -4.05357361e-01 1.16849029e+00 -1.34263545e-01 -6.39760613e-01 4.11895066e-01 -2.61481106e-01 -7.15637147e-01 5.59959829e-01 7.30639338e-01 8.03227842e-01 4.09479588e-01 -1.12706147e-01 2.54349113e-01 3.91056955e-01 -2.08916321e-01 -1.11491606e-02 1.25397944e+00 -2.30158344e-02 3.87567468e-02 4.63465333e-01 7.76129723e-01 1.27655685e-01 -8.94639015e-01 1.33776575e-01 3.25399339e-01 -3.28454882e-01 5.20880073e-02 -1.29770851e+00 -5.59657454e-01 1.37023890e+00 4.15348053e-01 2.97361732e-01 6.86028481e-01 -5.52525282e-01 6.96615160e-01 5.39439559e-01 2.44393423e-01 -1.40658689e+00 3.02416980e-01 6.63824856e-01 5.52721441e-01 -1.39785218e+00 3.75099070e-02 1.38969213e-01 -2.84580708e-01 1.33408725e+00 6.89478040e-01 -4.63009067e-02 1.58402681e-01 1.57622859e-01 2.87176549e-01 -4.32620674e-01 -8.01233649e-01 2.47171685e-01 3.91479254e-01 1.42644733e-01 1.35397637e+00 1.48980141e-01 -6.27920866e-01 8.66153121e-01 -6.35520756e-01 3.52447703e-02 1.01382565e+00 6.94077373e-01 -6.44196689e-01 -1.03675759e+00 -3.82957458e-01 6.07928157e-01 -8.96336496e-01 -1.86941832e-01 -6.17577195e-01 3.25130194e-01 -2.56180823e-01 8.01553130e-01 -1.71414927e-01 8.31926125e-04 1.55126020e-01 6.31356120e-01 7.29537547e-01 -6.91319704e-01 -1.09740722e+00 -8.11821938e-01 -2.14929841e-02 -1.78569794e-01 -1.32571653e-01 -5.07416487e-01 -9.20960784e-01 -5.48132002e-01 -4.11620051e-01 3.72760087e-01 7.59209633e-01 1.03755534e+00 -4.96587381e-02 8.21899116e-01 6.46153390e-02 -8.22018161e-02 -1.23159921e+00 -1.02415371e+00 -2.65686005e-01 1.41138911e-01 1.42903268e-01 -3.94834876e-01 -2.01750815e-01 5.31873330e-02]	[11.290714263916016, 9.286169052124023]
ce1175ed-6ac3-4811-abfa-9d2afe10ccd1	findings-of-the-wmt-2020-shared-task-on-1	null	null	https://aclanthology.org/2020.wmt-1.75	https://aclanthology.org/2020.wmt-1.75.pdf	Findings of the WMT 2020 Shared Task on Automatic Post-Editing	We present the results of the 6th round of the WMT task on MT Automatic Post-Editing. The task consists in automatically correcting the output of a “black-box” machine translation system by learning from existing human corrections of different sentences. This year, the challenge consisted of fixing the errors present in English Wikipedia pages translated into German and Chinese by state-ofthe-art, not domain-adapted neural MT (NMT) systems unknown to participants. Six teams participated in the English-German task, submitting a total of 11 runs. Two teams participated in the English-Chinese task submitting 2 runs each. Due to i) the different source/domain of data compared to the past (Wikipedia vs Information Technology), ii) the different quality of the initial translations to be corrected and iii) the introduction of a new language pair (English-Chinese), this year’s results are not directly comparable with last year’s round. However, on both language directions, participants’ submissions show considerable improvements over the baseline results. On English-German, the top ranked system improves over the baseline by -11.35 TER and +16.68 BLEU points, while on EnglishChinese the improvements are respectively up to -12.13 TER and +14.57 BLEU points. Overall, coherent gains are also highlighted by the outcomes of human evaluation, which confirms the effectiveness of APE to improve MT quality, especially in the new generic domain selected for this year’s round.	['Marco Turchi', 'Matteo Negri', 'Markus Freitag', 'Rajen Chatterjee']	null	null	null	null	wmt-emnlp-2020-11	['automatic-post-editing', 'automatic-post-editing']	['computer-vision', 'natural-language-processing']	[ 2.33373940e-01 2.55538434e-01 3.88661996e-02 -1.63737565e-01 -1.33182847e+00 -6.05217755e-01 8.57468247e-01 1.81533530e-01 -9.40899909e-01 1.34995854e+00 5.31314075e-01 -4.32681799e-01 3.41683090e-01 -2.25828275e-01 -1.06233215e+00 -1.04731910e-01 4.98682916e-01 7.17306256e-01 -5.26116155e-02 -6.85068846e-01 3.53374869e-01 -1.29678294e-01 -8.70665073e-01 5.57234585e-01 1.23778880e+00 7.91999474e-02 3.70325297e-01 7.96277285e-01 -8.47617071e-03 3.24132055e-01 -6.38418376e-01 -1.07691920e+00 2.45985836e-01 -4.41611528e-01 -9.98020172e-01 -4.48473066e-01 7.88363039e-01 2.20735781e-02 -1.77544370e-01 1.06090963e+00 6.16546810e-01 2.44414248e-02 5.65652132e-01 -5.51936865e-01 -1.02770281e+00 8.59235168e-01 -3.05826902e-01 2.69526780e-01 4.29514438e-01 1.16739929e-01 7.53186822e-01 -1.40770519e+00 1.00145698e+00 9.70738649e-01 8.45030010e-01 9.60761726e-01 -1.09744620e+00 -4.10509646e-01 -1.63667113e-01 1.19790576e-01 -1.11001360e+00 -6.22575164e-01 -2.22307406e-02 -4.77245688e-01 1.52274561e+00 3.83857280e-01 5.73365241e-02 1.29987383e+00 3.67981791e-01 3.82400721e-01 1.07405055e+00 -8.48050356e-01 -2.24659771e-01 5.71816564e-01 -1.39279291e-01 1.97512433e-01 3.27207655e-01 3.93191440e-04 -5.97365499e-01 2.99747944e-01 4.47958633e-02 -5.39700985e-01 -2.50104189e-01 2.51638979e-01 -1.79856670e+00 4.92853880e-01 1.59194529e-01 5.63966215e-01 -3.82843465e-01 -7.47629702e-02 6.34183228e-01 7.71924376e-01 7.49030471e-01 7.99938619e-01 -8.17517459e-01 -5.02447009e-01 -1.14761555e+00 3.37914407e-01 6.30228579e-01 1.06456530e+00 5.08813262e-01 -1.50021553e-01 -5.02485096e-01 9.74548936e-01 -1.63206965e-01 7.83232808e-01 7.03774691e-01 -5.00442863e-01 1.42892540e+00 1.83861256e-01 5.51429927e-01 -7.07108200e-01 -2.30075479e-01 -6.06582463e-01 -7.81964242e-01 -2.13800624e-01 6.39231145e-01 -5.12310624e-01 -8.64119053e-01 1.81355870e+00 -2.18555525e-01 -6.73501015e-01 9.41754952e-02 7.52930284e-01 5.08486331e-01 7.11884022e-01 2.61650868e-02 -2.59899080e-01 9.78314161e-01 -1.33933079e+00 -7.93819189e-01 -4.25774723e-01 9.07738328e-01 -1.34194887e+00 1.35857320e+00 1.41136676e-01 -1.48132491e+00 -7.97426939e-01 -1.00275648e+00 -2.61299193e-01 -4.84036088e-01 5.53366482e-01 -2.15784848e-01 4.71576333e-01 -1.44190073e+00 8.09464812e-01 -4.54821736e-01 -7.10183144e-01 -1.36701480e-01 2.63295263e-01 -5.57338774e-01 -1.38467744e-01 -1.49976349e+00 1.73436320e+00 2.66121626e-01 3.52644101e-02 -4.81582254e-01 -8.51263940e-01 -5.65962553e-01 -2.35702440e-01 -7.63241351e-02 -5.98210990e-01 1.50596142e+00 -1.07399833e+00 -1.49136436e+00 9.66333807e-01 -4.11797613e-01 -5.46551526e-01 1.22175705e+00 -4.93904710e-01 -6.34727299e-01 -4.19351429e-01 4.45724577e-01 6.51466548e-01 3.42301071e-01 -6.74651682e-01 -7.94705033e-01 -1.75257791e-02 -2.10508943e-01 4.73249823e-01 -2.09848374e-01 3.01286906e-01 -2.41245866e-01 -7.78690696e-01 -2.89148539e-01 -9.95246887e-01 -1.21991538e-01 -7.68837810e-01 -3.65945965e-01 -1.03154136e-02 1.45303398e-01 -1.44854128e+00 1.27297390e+00 -1.99122822e+00 3.53470951e-01 -2.18813524e-01 -1.06046475e-01 4.91840661e-01 -4.09086615e-01 7.65910923e-01 -1.48581654e-01 3.74846965e-01 -1.63633630e-01 -7.08262742e-01 5.90540171e-02 -2.67465800e-01 -2.35301062e-01 1.74786568e-01 2.88833857e-01 1.07365012e+00 -9.14872646e-01 -3.06335557e-02 -1.79477170e-01 1.00835398e-01 -1.03395648e-01 -9.55639482e-02 -5.48035838e-02 6.06707335e-01 1.76646277e-01 9.12148878e-02 6.98090792e-01 2.77673364e-01 -9.30180773e-03 1.98955461e-01 -4.85019177e-01 8.26337993e-01 -6.60680592e-01 1.80579412e+00 -6.40912235e-01 9.16352510e-01 -2.44133905e-01 -2.98717618e-01 7.24743307e-01 6.35771632e-01 -2.21576184e-01 -9.74092782e-01 -2.09560275e-01 7.84844398e-01 2.77307868e-01 -2.74286717e-01 1.01744044e+00 4.18490134e-02 -1.57304369e-02 5.87616146e-01 1.57817647e-01 -8.78981203e-02 6.38983786e-01 2.11366758e-01 9.48177516e-01 1.68212220e-01 2.02662055e-03 -5.12812078e-01 5.28973103e-01 3.12686056e-01 3.45717967e-01 8.03862810e-01 -1.24866337e-01 8.39925766e-01 -4.82064812e-03 -4.14315939e-01 -1.46915209e+00 -9.47166741e-01 1.92290112e-01 1.01698279e+00 -3.17191154e-01 -5.25286674e-01 -1.13155711e+00 -6.30375266e-01 -3.65980387e-01 9.84566748e-01 -6.16559565e-01 -1.38070494e-01 -9.17768061e-01 -5.78334749e-01 7.12462366e-01 3.49420607e-01 5.04992008e-01 -1.04820836e+00 -1.62934899e-01 3.62529963e-01 -9.19610262e-01 -1.10298383e+00 -1.08405185e+00 -3.75989713e-02 -8.99944723e-01 -4.33135599e-01 -1.07016158e+00 -8.35548699e-01 4.14237589e-01 1.29825950e-01 1.22866058e+00 6.08761497e-02 3.70018154e-01 2.80591883e-02 -3.24077159e-01 -5.44027150e-01 -8.53015423e-01 7.27867544e-01 3.10693175e-01 -4.76862848e-01 3.13262016e-01 -7.27017671e-02 -1.40521288e-01 1.71233956e-02 -3.80219966e-01 1.94646850e-01 8.01801682e-01 8.58771682e-01 2.14954481e-01 -6.52931035e-01 5.55087030e-01 -9.13519621e-01 9.24570382e-01 -2.39425331e-01 -2.19394401e-01 4.97100472e-01 -9.15902793e-01 -2.16616709e-02 6.05575562e-01 -3.14709872e-01 -1.09657109e+00 -2.98373669e-01 4.01590355e-02 3.71619433e-01 9.47308075e-03 5.97529233e-01 1.60908662e-02 1.88290283e-01 9.49023008e-01 1.21593095e-01 -3.34571272e-01 -4.52818125e-01 2.96156406e-01 7.20722258e-01 6.81109011e-01 -3.95855963e-01 8.52871358e-01 -4.70196933e-01 -5.76657295e-01 -3.75664920e-01 -5.19403696e-01 -2.24607028e-02 -9.37869608e-01 9.93036851e-02 8.58038962e-01 -1.05578268e+00 1.18018724e-01 5.64875722e-01 -1.59072959e+00 -5.45670569e-01 -1.36170864e-01 6.17955983e-01 -3.00297141e-01 3.65645230e-01 -9.08570528e-01 -2.27782607e-01 -7.54317820e-01 -1.06974530e+00 6.40735209e-01 5.29598556e-02 -6.65167332e-01 -9.42930937e-01 3.20119619e-01 4.55102414e-01 6.63471520e-01 -1.46732524e-01 8.75283182e-01 -5.60053468e-01 -4.30346318e-02 -1.62085488e-01 -1.32282928e-01 3.77743334e-01 9.49108452e-02 -4.47848476e-02 -6.77548945e-01 -3.41740876e-01 -3.67548525e-01 -5.93030900e-02 5.45849979e-01 3.14043858e-03 1.84226260e-01 -3.90406042e-01 -1.80395931e-01 2.52564326e-02 1.01486886e+00 -1.05613209e-01 6.63678586e-01 7.03172207e-01 5.14743924e-01 5.32395542e-01 6.64109826e-01 -1.59017757e-01 5.63082933e-01 8.93911541e-01 -1.03343561e-01 -3.10595091e-02 -5.16068399e-01 -4.08990562e-01 9.23188984e-01 1.40306997e+00 -1.66908249e-01 -3.58712852e-01 -8.99878860e-01 8.14541578e-01 -1.72649419e+00 -8.17314863e-01 -7.24487185e-01 2.51224637e+00 1.17565799e+00 2.47230485e-01 -4.08612043e-02 -3.51484567e-01 9.60900664e-01 -3.50351632e-01 -1.40362769e-01 -1.04870188e+00 -3.99006039e-01 3.72274548e-01 5.93371212e-01 8.31087291e-01 -8.43326569e-01 1.27115297e+00 6.27731800e+00 6.26681209e-01 -1.12310553e+00 5.28433979e-01 3.64235222e-01 -6.33706152e-02 -1.39253914e-01 -4.81263101e-02 -7.70320475e-01 6.91166997e-01 1.53611898e+00 -4.91643846e-01 6.89060390e-01 1.82038054e-01 3.84763986e-01 -1.35407245e-04 -1.15844786e+00 6.43935561e-01 2.79713869e-02 -1.18746912e+00 1.26977950e-01 1.49901668e-02 1.24614835e+00 5.65600157e-01 -4.38510329e-02 5.15361071e-01 2.66137242e-01 -1.00522995e+00 1.18156254e+00 5.30170321e-01 9.40005243e-01 -6.56680584e-01 1.04750180e+00 5.43076396e-01 -5.04375637e-01 3.50147814e-01 -5.62267542e-01 -3.98951828e-01 1.37895808e-01 6.14455462e-01 -1.00954497e+00 8.07552516e-01 6.91993892e-01 5.09386182e-01 -8.05583894e-01 8.05576503e-01 -5.68463802e-01 6.59478664e-01 4.47615273e-02 -6.23782799e-02 2.35794514e-01 -7.50028342e-02 5.64989090e-01 1.79527521e+00 6.30363405e-01 -3.79141480e-01 -4.20246691e-01 6.46463096e-01 -5.47273457e-01 2.67979711e-01 -3.50445956e-01 -7.13486075e-02 5.73084176e-01 1.10538566e+00 -1.61870882e-01 -4.56438154e-01 -1.68749303e-01 1.48922193e+00 5.93525410e-01 3.52778196e-01 -8.05866778e-01 -6.76801562e-01 4.24753785e-01 2.86261383e-02 1.26685932e-01 -2.91577429e-01 -5.98025322e-01 -1.14061022e+00 4.61019814e-01 -1.09611475e+00 -8.34968966e-03 -7.74877846e-01 -1.03745043e+00 9.12837565e-01 -3.50249588e-01 -1.17520130e+00 -2.83529371e-01 -4.04187053e-01 -5.75314820e-01 1.59107947e+00 -1.17075276e+00 -9.30878878e-01 1.23260096e-01 1.69474840e-01 7.40372002e-01 -2.14120835e-01 8.87927115e-01 5.60974717e-01 -3.43210757e-01 1.10981023e+00 5.53652465e-01 6.73747584e-02 1.48133373e+00 -1.31221199e+00 1.18254054e+00 1.36042523e+00 2.27971271e-01 7.79339731e-01 8.50185812e-01 -7.92172968e-01 -7.91241825e-01 -1.13648045e+00 2.20463634e+00 -9.66568410e-01 5.59919059e-01 -4.25685436e-01 -7.08444834e-01 6.23816609e-01 8.91680241e-01 -6.58745944e-01 1.64338559e-01 1.97342914e-02 -1.71769992e-01 1.08466849e-01 -1.10230803e+00 8.47475529e-01 9.99004543e-01 -6.08744144e-01 -7.23917186e-01 4.55599725e-01 9.01956677e-01 -7.25335598e-01 -7.52429843e-01 1.88554659e-01 2.84653366e-01 -4.04710114e-01 4.32123899e-01 -7.27149367e-01 8.03522170e-01 -2.90424258e-01 -4.18903269e-02 -1.99548781e+00 -6.12600744e-01 -7.09353626e-01 4.13348198e-01 1.37371254e+00 1.11990499e+00 -5.72677910e-01 2.17756972e-01 4.97328311e-01 -4.26820368e-01 -2.50320584e-01 -9.75184500e-01 -6.89642489e-01 7.59630024e-01 -2.98374761e-02 3.30831677e-01 9.69242513e-01 1.14571571e-01 6.78643584e-01 -4.45451647e-01 -1.79142430e-01 1.58135772e-01 -5.61574936e-01 7.92833984e-01 -9.44580495e-01 -1.48185253e-01 -5.54561317e-01 1.68422848e-01 -7.38222599e-01 -1.41830295e-01 -1.11921608e+00 1.81899309e-01 -1.68009377e+00 2.28486255e-01 1.41219180e-02 -7.06756860e-02 4.55123246e-01 -5.61037481e-01 3.83916467e-01 4.34045702e-01 2.91025519e-01 -3.01086932e-01 1.83063179e-01 1.06189525e+00 -2.15566710e-01 -5.65531142e-02 -1.38611540e-01 -8.12375903e-01 1.53869912e-01 9.25746202e-01 -6.49559796e-01 4.79994416e-02 -1.28442621e+00 3.99688363e-01 -2.87611753e-01 -3.85069922e-02 -9.85300362e-01 2.12937862e-01 1.07100010e-01 3.56887668e-01 -3.21891695e-01 -7.43247941e-02 -2.37684473e-01 1.28063127e-01 5.39421678e-01 -4.66585547e-01 8.04018497e-01 4.34239030e-01 2.85927802e-02 -1.16972744e-01 -2.66181350e-01 6.79586828e-01 -2.24771649e-01 -3.66239607e-01 -2.35565588e-01 -5.69583654e-01 2.76831627e-01 4.63545501e-01 -1.73456714e-01 -5.10641098e-01 -4.96629000e-01 -6.18174911e-01 4.24363725e-02 4.60118353e-01 6.65387094e-01 2.37869173e-02 -1.24683762e+00 -1.36254096e+00 -2.26330712e-01 1.94174331e-02 -5.78319550e-01 5.14123142e-02 1.16924131e+00 -4.75003302e-01 6.45121634e-01 -2.74049401e-01 -2.02620760e-01 -1.07712507e+00 1.88687474e-01 2.43794158e-01 -6.48972034e-01 -1.78432882e-01 6.99605227e-01 -3.63392621e-01 -9.64083195e-01 -6.62902445e-02 -1.28954694e-01 1.40326217e-01 -1.41082093e-01 5.62603354e-01 6.41799927e-01 8.35080385e-01 -6.08017564e-01 -3.00602257e-01 3.42617035e-01 -4.88433480e-01 -5.89987099e-01 1.07461250e+00 -4.07849133e-01 -2.08225057e-01 3.83651108e-01 7.99044371e-01 4.74858373e-01 -6.41969562e-01 -2.81271011e-01 4.37335759e-01 -1.37020811e-01 -4.95020866e-01 -1.70972884e+00 -5.07570624e-01 1.10805333e+00 4.90025967e-01 -4.99954283e-01 5.90916991e-01 -4.65314776e-01 8.37856948e-01 4.79046017e-01 3.98989677e-01 -1.49870265e+00 -4.02963281e-01 1.20054495e+00 1.14337993e+00 -1.12000525e+00 -3.00287962e-01 -2.53252611e-02 -7.38800824e-01 1.15245962e+00 6.69311702e-01 2.23108694e-01 -1.21439040e-01 -1.70882836e-01 3.20948750e-01 5.02548695e-01 -6.57455564e-01 3.02652985e-01 4.27788585e-01 6.09046817e-01 9.29076970e-01 2.41159633e-01 -9.38449681e-01 6.20300591e-01 -5.43476939e-01 -7.41798282e-02 7.58067191e-01 4.69894022e-01 -2.51479089e-01 -1.41101992e+00 -2.21280351e-01 1.06995523e-01 -5.14698386e-01 -5.63000858e-01 -8.40745330e-01 7.97630012e-01 2.78596226e-02 1.25594711e+00 -2.24165723e-01 -5.22388160e-01 6.14155948e-01 2.55872369e-01 5.10244489e-01 -6.93157196e-01 -1.15853608e+00 -3.16607744e-01 3.22435439e-01 -2.00306848e-01 3.24261584e-03 -8.16950023e-01 -1.01739764e+00 -5.86757481e-01 -1.12988986e-01 4.57010269e-01 7.28008091e-01 1.03492904e+00 5.23575008e-01 2.18573615e-01 3.10516357e-01 -7.38772154e-01 -7.67531574e-01 -1.58912838e+00 1.64871886e-01 3.51992905e-01 6.85203969e-02 -1.39289021e-01 -9.63817388e-02 1.46577835e-01]	[11.628674507141113, 10.312995910644531]
ae4e9415-7f57-4468-ace5-e1b9544f9f8e	deeparchitect-automatically-designing-and	1704.08792	null	http://arxiv.org/abs/1704.08792v1	http://arxiv.org/pdf/1704.08792v1.pdf	DeepArchitect: Automatically Designing and Training Deep Architectures	In deep learning, performance is strongly affected by the choice of architecture and hyperparameters. While there has been extensive work on automatic hyperparameter optimization for simple spaces, complex spaces such as the space of deep architectures remain largely unexplored. As a result, the choice of architecture is done manually by the human expert through a slow trial and error process guided mainly by intuition. In this paper we describe a framework for automatically designing and training deep models. We propose an extensible and modular language that allows the human expert to compactly represent complex search spaces over architectures and their hyperparameters. The resulting search spaces are tree-structured and therefore easy to traverse. Models can be automatically compiled to computational graphs once values for all hyperparameters have been chosen. We can leverage the structure of the search space to introduce different model search algorithms, such as random search, Monte Carlo tree search (MCTS), and sequential model-based optimization (SMBO). We present experiments comparing the different algorithms on CIFAR-10 and show that MCTS and SMBO outperform random search. In addition, these experiments show that our framework can be used effectively for model discovery, as it is possible to describe expressive search spaces and discover competitive models without much effort from the human expert. Code for our framework and experiments has been made publicly available.	['Renato Negrinho', 'Geoff Gordon']	2017-04-28	deeparchitect-automatically-designing-and-1	https://openreview.net/forum?id=rkTBjG-AZ	https://openreview.net/pdf?id=rkTBjG-AZ	iclr-2018-1	['model-discovery']	['miscellaneous']	[-2.89212346e-01 -5.43207936e-02 -2.95352876e-01 -4.70102608e-01 -7.27036476e-01 -7.36193895e-01 5.52492797e-01 1.05387624e-02 -6.41347826e-01 5.94379067e-01 -1.12390667e-01 -7.20943868e-01 -2.54304588e-01 -6.22799575e-01 -5.05514681e-01 -6.52355671e-01 -2.77821580e-03 9.75016952e-01 2.49331996e-01 -7.20822588e-02 1.55815333e-01 4.95413870e-01 -1.33942854e+00 9.83170606e-03 7.00776517e-01 8.70784223e-01 1.85806021e-01 6.78390682e-01 -1.33413970e-01 4.25347298e-01 -4.64842290e-01 -4.72596496e-01 4.45807487e-01 -2.97566801e-01 -1.04709280e+00 -1.94748566e-01 1.61863908e-01 -6.59306571e-02 3.76743264e-02 8.76376629e-01 6.49937510e-01 1.80797771e-01 4.44184422e-01 -1.12746823e+00 -5.58086485e-02 8.89457524e-01 -2.39445031e-01 1.57346934e-01 -2.59445608e-01 3.00092697e-01 1.35137999e+00 -7.85368264e-01 4.17807102e-01 1.17499840e+00 4.75508660e-01 5.81971169e-01 -1.61695063e+00 -7.66228676e-01 3.63389283e-01 2.59186268e-01 -1.52199745e+00 -4.89272833e-01 6.58205509e-01 -5.18997967e-01 1.15839481e+00 2.81011224e-01 9.17652488e-01 9.76650357e-01 -1.59791633e-01 8.36089015e-01 6.22902453e-01 -6.22798920e-01 7.79365182e-01 1.63187370e-01 2.98943430e-01 8.36986423e-01 2.73582160e-01 7.79079348e-02 -1.96155667e-01 -5.06040633e-01 8.47484231e-01 -2.56484419e-01 -1.92357928e-01 -8.99066150e-01 -9.58204806e-01 1.17667294e+00 4.55754489e-01 7.73451999e-02 -2.88246214e-01 4.62194920e-01 2.42038950e-01 2.47829646e-01 -3.99882495e-02 1.01718342e+00 -6.58159435e-01 -1.44254416e-01 -9.14118528e-01 5.76224923e-01 1.14237154e+00 7.25753784e-01 7.17735767e-01 -1.26555696e-01 9.99232158e-02 9.93627846e-01 3.90515178e-01 1.45214060e-02 4.90034848e-01 -9.78224874e-01 3.02039444e-01 5.12096524e-01 1.85479105e-01 -7.31842637e-01 -6.62488759e-01 -8.93283844e-01 -5.83840907e-01 1.44956589e-01 3.46766025e-01 -2.90322334e-01 -9.96905029e-01 1.65052497e+00 4.76693183e-01 -2.67463505e-01 -2.24287942e-01 8.69494975e-01 4.25151676e-01 3.30654621e-01 1.88611716e-01 1.52431414e-01 1.32249761e+00 -1.13251042e+00 -2.90952951e-01 -5.03416002e-01 1.02584612e+00 -3.72385889e-01 1.13839400e+00 4.36959088e-01 -1.15842235e+00 -9.92227271e-02 -1.02014875e+00 1.11194983e-01 -1.88032314e-01 1.79349899e-01 8.86099279e-01 7.50685692e-01 -1.07148290e+00 4.02768493e-01 -1.19868267e+00 -1.15121983e-01 7.07499027e-01 7.54188955e-01 1.00565497e-02 1.35090917e-01 -1.00702250e+00 8.40075970e-01 7.93756068e-01 5.55019230e-02 -8.39134872e-01 -5.28384924e-01 -5.32649815e-01 2.66299188e-01 5.24336636e-01 -1.04863775e+00 1.68593264e+00 -8.58085215e-01 -1.48837197e+00 6.01338208e-01 5.63363582e-02 -7.31755555e-01 3.77342433e-01 1.23020932e-02 1.20353073e-01 -1.05519399e-01 -4.94090289e-01 6.90574765e-01 6.05325818e-01 -1.04705298e+00 -4.27355379e-01 -2.08283797e-01 3.61343861e-01 1.94149151e-01 -4.20791626e-01 8.70258063e-02 -7.58575797e-01 -5.08702159e-01 1.13690086e-01 -1.24545336e+00 -6.74823523e-01 -1.61410853e-01 -4.41069663e-01 -2.36169323e-01 6.28239140e-02 -4.70480584e-02 1.69686699e+00 -1.82647097e+00 1.55058160e-01 5.32260954e-01 3.15126240e-01 2.21531481e-01 -1.55046791e-01 2.58245677e-01 1.06696986e-01 4.60403919e-01 -1.25244349e-01 -1.78974137e-01 2.75073856e-01 2.45611429e-01 2.23776326e-03 1.62343636e-01 -1.69967398e-01 9.96058464e-01 -6.34952545e-01 -5.61436057e-01 -6.62542805e-02 3.04829031e-01 -1.04724693e+00 -1.56806037e-02 -5.71347654e-01 8.51655453e-02 -7.63816297e-01 4.98565853e-01 3.02987158e-01 -7.40677297e-01 3.26054722e-01 -7.73734897e-02 1.44917399e-01 6.01921797e-01 -1.27050793e+00 1.38861477e+00 -6.17113411e-01 3.71727079e-01 9.28060263e-02 -8.29335332e-01 8.41462374e-01 -2.25773267e-03 8.39688256e-02 -4.32990402e-01 1.94186270e-01 2.38822997e-01 2.44308844e-01 -9.35241506e-02 1.31191894e-01 -7.70061649e-03 1.75453365e-01 8.02650154e-01 -7.15181231e-02 -5.64243235e-02 2.64955968e-01 -6.55632094e-02 1.17113423e+00 -1.72656149e-01 3.32726151e-01 -2.92308897e-01 2.70065844e-01 2.46101663e-01 4.21646655e-01 9.49084282e-01 1.51704445e-01 2.61377156e-01 5.78282416e-01 -8.28808188e-01 -1.10603821e+00 -7.63263941e-01 -2.06782788e-01 1.35858750e+00 -2.34874025e-01 -6.90625072e-01 -9.33977902e-01 -5.64958990e-01 -2.67366886e-01 6.46950722e-01 -5.25901675e-01 -1.05398726e-02 -7.83255756e-01 -9.72266197e-01 3.01525295e-01 5.77002943e-01 3.03078830e-01 -9.77677226e-01 -8.64565551e-01 3.09828728e-01 1.03631958e-01 -8.37980211e-01 -3.55718344e-01 5.47711492e-01 -1.06156397e+00 -8.52780282e-01 -3.83368880e-01 -6.84228957e-01 5.99417806e-01 -1.83918148e-01 1.33489656e+00 3.13589573e-01 -3.63345921e-01 4.40865271e-02 -9.68190804e-02 -4.24674064e-01 -3.85622323e-01 7.00157404e-01 -3.35699737e-01 -3.20270091e-01 3.25670779e-01 -4.55103844e-01 -7.78975129e-01 3.84105206e-01 -7.77260303e-01 3.45937610e-01 6.54466331e-01 9.38811660e-01 6.76883578e-01 8.31437334e-02 2.76800573e-01 -9.85574305e-01 6.98840320e-01 -3.53439510e-01 -1.03390551e+00 2.74381697e-01 -8.40140224e-01 6.38726652e-01 3.89711916e-01 -5.77779114e-01 -5.63754737e-01 3.21699679e-01 -1.68295279e-01 -4.48679745e-01 -7.01527148e-02 6.52189791e-01 -1.17386758e-01 -1.21599659e-01 8.21836233e-01 -2.40374774e-01 -2.30211690e-01 -7.19161510e-01 2.67222971e-01 3.32204312e-01 3.54233496e-02 -7.54915893e-01 5.91660023e-01 1.12089865e-01 -1.21760599e-01 -5.47187626e-01 -7.63131917e-01 -1.15647919e-01 -3.77276629e-01 2.36789882e-01 6.07980311e-01 -7.13503897e-01 -6.29104853e-01 6.37225062e-02 -8.96937311e-01 -6.99283719e-01 8.25310778e-03 4.63538080e-01 -3.76231313e-01 -1.03164993e-01 -3.05503100e-01 -5.76873600e-01 -5.03101051e-01 -1.44556201e+00 7.84930408e-01 1.56005248e-01 -5.23074746e-01 -1.30757535e+00 1.88718736e-02 2.51789659e-01 4.76592004e-01 -8.64368528e-02 1.33782792e+00 -1.15228045e+00 -8.97763789e-01 -2.53251851e-01 1.14101857e-01 -4.99201529e-02 -4.35866207e-01 -7.16664866e-02 -8.42503726e-01 -2.76586473e-01 -1.80429697e-01 -3.78825486e-01 7.42129445e-01 5.66149294e-01 1.41388750e+00 -5.78348100e-01 -5.21322548e-01 8.60617220e-01 1.27402103e+00 5.52264154e-02 2.30434701e-01 8.60489964e-01 5.57742655e-01 3.43865871e-01 9.18954387e-02 4.14122134e-01 3.94716382e-01 8.61002028e-01 4.10095543e-01 -6.31472468e-02 4.37346101e-01 -1.30902544e-01 -1.70080364e-01 3.39091331e-01 1.17394023e-01 -2.34102666e-01 -1.34241676e+00 4.04385537e-01 -1.83170593e+00 -6.26931131e-01 2.46587142e-01 2.19145727e+00 1.11034334e+00 4.04207468e-01 1.96993709e-01 5.45487832e-03 4.03144598e-01 -6.12854492e-03 -6.44360244e-01 -4.83992606e-01 1.87746108e-01 3.26041281e-01 6.60512149e-01 6.04212701e-01 -1.02462018e+00 1.09659815e+00 6.45914650e+00 9.09133077e-01 -9.92644787e-01 -1.56594619e-01 7.66985118e-01 -5.33180952e-01 -4.39808100e-01 3.08204174e-01 -9.77470040e-01 7.47944638e-02 9.87638175e-01 -9.85706374e-02 6.40285254e-01 1.05195451e+00 1.02768883e-01 1.69636324e-01 -1.26506186e+00 1.12435389e+00 -5.57661176e-01 -1.52112281e+00 4.91301715e-03 2.33544275e-01 5.89828551e-01 2.83560365e-01 -2.22876295e-02 4.41133469e-01 7.72631347e-01 -1.29686773e+00 6.29771888e-01 1.11999661e-01 2.65779704e-01 -8.02906573e-01 3.79934162e-01 4.87370014e-01 -9.01183903e-01 -4.65133637e-01 -3.61982554e-01 2.57646561e-01 -1.92601323e-01 4.15732831e-01 -1.01880574e+00 -5.22623546e-02 5.79928339e-01 1.80674251e-02 -7.25958705e-01 1.30786359e+00 -1.99598402e-01 7.87383795e-01 -6.04227364e-01 -3.96981150e-01 5.19952416e-01 1.14660174e-01 3.65542233e-01 1.08302033e+00 4.55307327e-02 -5.44048958e-02 2.97784477e-01 9.98698354e-01 -6.98321834e-02 1.64099336e-01 -1.45729274e-01 -5.22852093e-02 7.25546956e-01 1.24635291e+00 -7.83466101e-01 -1.13945477e-01 -1.19811691e-01 2.29919940e-01 5.69815278e-01 5.24924219e-01 -7.48461306e-01 -2.33445331e-01 5.30802667e-01 1.74023941e-01 5.82237720e-01 -2.10743472e-01 -4.45931822e-01 -9.74706471e-01 -1.41513333e-01 -1.19098067e+00 5.50690770e-01 -4.58290040e-01 -8.40754330e-01 8.88921261e-01 2.57812142e-01 -6.12783730e-01 -5.54611683e-01 -5.58575273e-01 -5.10506809e-01 7.56360829e-01 -1.09722185e+00 -5.63563347e-01 -4.74786758e-02 3.29792202e-01 5.61402500e-01 -1.77267343e-01 7.84414887e-01 -6.30908608e-02 -7.85526156e-01 8.11409831e-01 2.99311448e-02 4.71599996e-02 1.31537810e-01 -1.13279319e+00 6.35313809e-01 3.63658309e-01 3.89999896e-01 8.35278213e-01 8.84172261e-01 -2.07623720e-01 -1.32535279e+00 -7.46487677e-01 6.60109222e-01 -2.02992365e-01 7.69221604e-01 -5.76513231e-01 -8.84348631e-01 5.87481976e-01 -4.55158949e-02 -1.44176841e-01 6.98782027e-01 4.81336802e-01 -3.42127055e-01 -3.57619561e-02 -6.83809698e-01 8.38554978e-01 8.88974428e-01 -1.59735680e-01 -1.12294599e-01 4.27866787e-01 5.21230400e-01 -4.17743802e-01 -6.94655120e-01 4.03955430e-01 5.84048092e-01 -8.48463774e-01 9.75253224e-01 -7.13399470e-01 -1.20458253e-01 -2.22990550e-02 2.06633769e-02 -1.21361315e+00 -2.93776214e-01 -8.64694178e-01 -2.99185783e-01 6.72651708e-01 8.92936528e-01 -6.81056499e-01 8.93592477e-01 1.08651030e+00 1.28635302e-01 -1.38133466e+00 -7.30515480e-01 -5.82792222e-01 2.29596838e-01 -4.85854417e-01 9.90349770e-01 5.77337086e-01 -2.85278499e-01 6.20406568e-01 1.75465822e-01 3.42344604e-02 4.44108397e-01 3.10322046e-01 8.72086585e-01 -1.42257857e+00 -7.28499651e-01 -1.00196838e+00 -8.38966519e-02 -1.11842620e+00 8.86757746e-02 -9.32240427e-01 -1.40345559e-01 -1.37545240e+00 2.14205682e-01 -9.55152094e-01 -1.88230023e-01 7.62536049e-01 -6.05841801e-02 -2.45643795e-01 3.81430984e-02 4.20598924e-01 -6.34605467e-01 4.33531374e-01 1.07141256e+00 1.51157632e-01 -3.81729066e-01 1.71560109e-01 -8.06125283e-01 8.82400692e-01 1.06709433e+00 -6.09952867e-01 -6.35594249e-01 -6.22089863e-01 4.92787212e-01 -8.24699551e-02 3.62889498e-01 -8.49817038e-01 2.11886227e-01 -3.06219250e-01 2.52392948e-01 -2.69214332e-01 2.14980647e-01 -6.27795041e-01 2.32443094e-01 4.51190472e-01 -6.92472160e-01 1.94797710e-01 3.64243686e-01 3.98664355e-01 9.88716111e-02 -5.23938179e-01 8.72898698e-01 -2.05830589e-01 -3.42142969e-01 3.63095373e-01 -9.68983173e-02 2.90043801e-01 8.48329246e-01 -1.21830128e-01 -3.80906463e-02 -2.66160339e-01 -8.66814017e-01 5.06598830e-01 5.45576394e-01 1.63492918e-01 3.27969074e-01 -1.00741899e+00 -3.85487944e-01 2.06505701e-01 3.91344503e-02 1.96273595e-01 -1.61265090e-01 5.86158156e-01 -5.74361801e-01 5.01689255e-01 2.52579212e-01 -6.08248770e-01 -1.14059711e+00 3.87922913e-01 7.02995420e-01 -5.36629438e-01 -4.77145672e-01 9.42061424e-01 2.57593274e-01 -4.19353127e-01 5.88607490e-01 -3.27582687e-01 -3.38084958e-02 -1.75953835e-01 4.55609798e-01 5.49455211e-02 1.32753730e-01 -1.21058635e-01 -3.65114689e-01 3.07585120e-01 -3.46849144e-01 -2.34012365e-01 1.37121630e+00 1.20017409e-01 1.65983126e-01 1.00324877e-01 1.16087019e+00 -4.95024651e-01 -1.24477553e+00 -2.56298810e-01 2.72708595e-01 -1.63236335e-01 4.07206863e-01 -8.96403730e-01 -9.70496655e-01 8.37238848e-01 3.55474144e-01 1.66317850e-01 1.10293770e+00 8.35542232e-02 4.43409443e-01 8.23047280e-01 2.78199434e-01 -1.07110834e+00 -3.68955955e-02 5.76332211e-01 6.54868960e-01 -1.00442779e+00 -2.91183516e-02 7.17148976e-03 -6.28777683e-01 1.06784880e+00 5.96007705e-01 1.20273240e-01 7.28643835e-01 3.17000180e-01 -5.36461882e-02 -3.69333178e-01 -1.33588767e+00 -6.79901168e-02 3.08477044e-01 1.92463726e-01 3.70933205e-01 -7.04927146e-02 6.84234500e-02 5.90499878e-01 -6.25453234e-01 -1.34723201e-01 4.89291772e-02 8.58698726e-01 -6.44288480e-01 -1.35516131e+00 -2.34772503e-01 4.78833109e-01 -5.00983834e-01 -2.88539708e-01 -2.51592994e-01 8.69336128e-01 -2.69963235e-01 6.97470367e-01 -4.20740023e-02 -1.28255665e-01 1.27180591e-01 2.31244057e-01 5.54805994e-01 -6.69424355e-01 -6.64775312e-01 1.77938387e-01 2.67753631e-01 -4.78749067e-01 2.06938893e-01 -4.44604844e-01 -1.21657956e+00 -5.75538278e-02 -4.78863448e-01 3.83250207e-01 9.11354065e-01 8.74776423e-01 3.94758046e-01 2.01503143e-01 3.20349842e-01 -7.13021874e-01 -1.00773787e+00 -6.25300109e-01 -1.21064119e-01 -8.09732378e-02 2.61158586e-01 -6.65898561e-01 -2.46027231e-01 -1.88047960e-01]	[8.462992668151855, 3.604593276977539]
ceb44ad3-3154-4a4a-85c7-04dc5ceb609f	on-the-use-of-modality-specific-large-scale	2210.15937	null	https://arxiv.org/abs/2210.15937v1	https://arxiv.org/pdf/2210.15937v1.pdf	On the Use of Modality-Specific Large-Scale Pre-Trained Encoders for Multimodal Sentiment Analysis	This paper investigates the effectiveness and implementation of modality-specific large-scale pre-trained encoders for multimodal sentiment analysis~(MSA). Although the effectiveness of pre-trained encoders in various fields has been reported, conventional MSA methods employ them for only linguistic modality, and their application has not been investigated. This paper compares the features yielded by large-scale pre-trained encoders with conventional heuristic features. One each of the largest pre-trained encoders publicly available for each modality are used; CLIP-ViT, WavLM, and BERT for visual, acoustic, and linguistic modalities, respectively. Experiments on two datasets reveal that methods with domain-specific pre-trained encoders attain better performance than those with conventional features in both unimodal and multimodal scenarios. We also find it better to use the outputs of the intermediate layers of the encoders than those of the output layer. The codes are available at https://github.com/ando-hub/MSA_Pretrain.	['Hiroshi Sato', 'Takanori Ashihara', 'Takafumi Moriya', 'Keita Suzuki', 'Naoki Makishima', 'Satoshi Suzuki', 'Akihiko Takashima', 'Ryo Masumura', 'Atsushi Ando']	2022-10-28	null	null	null	null	['multimodal-sentiment-analysis', 'multimodal-sentiment-analysis']	['computer-vision', 'natural-language-processing']	[ 1.98851675e-01 3.47025208e-02 -7.48304725e-02 -7.58717239e-01 -1.13932455e+00 -6.17443740e-01 5.64008296e-01 1.23493504e-02 -6.73028409e-01 6.28790736e-01 3.74927044e-01 -2.83739120e-01 4.47359651e-01 -3.82026911e-01 -9.16855812e-01 -6.17181718e-01 1.28671691e-01 2.69066185e-01 -1.61493167e-01 -3.76786321e-01 1.24827502e-02 -1.09261684e-01 -1.62325275e+00 8.13172877e-01 4.65131044e-01 1.17772961e+00 2.43600205e-01 9.70076919e-01 -1.11673459e-01 7.92164207e-01 -3.39634836e-01 -7.64472485e-01 -1.66450128e-01 -5.44593215e-01 -9.19374526e-01 8.48529264e-02 3.83514941e-01 -2.24748805e-01 -4.88326490e-01 8.09293985e-01 6.98951721e-01 6.19517304e-02 5.92469275e-01 -1.31753206e+00 -7.63552785e-01 7.82578588e-01 -3.85613561e-01 -7.67709315e-02 7.33504653e-01 7.46699795e-02 1.14556170e+00 -8.84145558e-01 7.62024701e-01 1.27002311e+00 6.98015094e-01 6.64812565e-01 -7.06835747e-01 -4.89098608e-01 -1.34017214e-01 3.67448926e-01 -1.18617618e+00 -6.60028994e-01 7.03794539e-01 -1.24249227e-01 1.12714505e+00 7.39096180e-02 1.88864276e-01 1.49398351e+00 9.82500911e-02 1.22969854e+00 1.05204558e+00 -5.75829387e-01 -5.79901040e-02 5.95784247e-01 -8.54309127e-02 1.03848457e+00 -4.64067042e-01 -2.07926407e-01 -9.64818478e-01 6.06414936e-02 2.51162261e-01 -2.80683488e-01 -2.38783821e-01 9.35400371e-03 -1.38786590e+00 8.69981110e-01 3.42896879e-01 3.80943328e-01 -3.14013451e-01 7.83617273e-02 7.68845797e-01 5.48488200e-01 1.95209578e-01 3.17944646e-01 -7.89023399e-01 -5.00952065e-01 -9.37262714e-01 -2.19239846e-01 6.03724778e-01 1.09194767e+00 7.66111255e-01 -8.86935815e-02 -1.31338403e-01 9.84905303e-01 4.00950968e-01 6.54895425e-01 6.18905246e-01 -8.55344296e-01 6.16813183e-01 6.15907967e-01 -2.07709774e-01 -6.02233529e-01 -6.21691704e-01 -4.62286128e-03 -7.29097605e-01 -1.76141784e-01 3.76394451e-01 -6.45409107e-01 -1.01698899e+00 1.76057899e+00 -2.59347390e-02 1.95596158e-03 4.63835090e-01 8.09841931e-01 1.28450823e+00 9.25251663e-01 1.37967736e-01 5.24698161e-02 1.38951588e+00 -1.17335880e+00 -8.39332879e-01 -1.54678285e-01 6.27329648e-01 -8.42613220e-01 1.15415764e+00 1.62906930e-01 -1.21903563e+00 -5.90244651e-01 -9.41975474e-01 -3.46432269e-01 -8.90588224e-01 4.02724534e-01 6.17104948e-01 4.71308917e-01 -1.14383996e+00 1.83734447e-01 -6.89377427e-01 -5.59914768e-01 2.83938438e-01 4.76893187e-01 -6.29523933e-01 -5.76329716e-02 -1.39527965e+00 1.20762861e+00 2.40148470e-01 2.79079318e-01 -8.06394994e-01 -1.20867543e-01 -1.25669646e+00 -3.57778594e-02 -8.06340873e-02 -6.56498015e-01 1.38521647e+00 -1.42111170e+00 -1.61066425e+00 8.57443690e-01 -5.25224924e-01 -2.62467802e-01 -8.78381729e-03 -3.58761139e-02 -5.87120831e-01 3.49229246e-01 -1.58140644e-01 1.22747278e+00 8.60541701e-01 -1.34759486e+00 -8.00611317e-01 1.08139768e-01 3.77515078e-01 4.61821437e-01 -6.32395267e-01 2.02454805e-01 -4.83154595e-01 -1.89634264e-01 -5.10776401e-01 -8.73161316e-01 -6.55424073e-02 -4.64569092e-01 -4.53362644e-01 -2.36300036e-01 6.60453975e-01 -7.68853128e-01 1.06622100e+00 -2.45622969e+00 3.20924729e-01 7.19762072e-02 -2.94151634e-01 -2.88146809e-02 -6.25634074e-01 7.69571483e-01 7.85847306e-02 6.28229789e-03 -2.70570993e-01 -6.97158039e-01 3.35196853e-01 1.44245371e-01 -7.64199048e-02 2.52670258e-01 4.28671002e-01 9.90047932e-01 -7.23758399e-01 -5.78023136e-01 4.05085832e-01 7.11427867e-01 -2.85980940e-01 2.36093968e-01 -4.19218764e-02 3.00504416e-01 -1.32180631e-01 9.66753781e-01 4.29693252e-01 -3.34865749e-01 1.61396563e-01 -4.40280467e-01 3.79230119e-02 3.00156176e-01 -8.99582505e-01 1.92088103e+00 -6.70332193e-01 1.00350153e+00 3.12065724e-02 -8.26816738e-01 4.95473593e-01 7.02717364e-01 2.91824758e-01 -8.84232700e-01 5.26311696e-01 3.76134545e-01 -1.48194879e-01 -6.29906654e-01 8.31757367e-01 -1.83725476e-01 -4.07169789e-01 2.89077342e-01 8.34046662e-01 3.42414938e-02 4.49398935e-01 2.62948215e-01 8.45073342e-01 1.18245229e-01 -5.53455055e-02 3.99061739e-01 6.63793683e-01 -1.35431262e-02 5.64029813e-02 5.54474533e-01 -1.79816037e-01 9.16087270e-01 4.18040454e-01 3.88262048e-02 -8.44193637e-01 -7.22958267e-01 -1.92951009e-01 1.73464155e+00 2.16387846e-02 -5.13749719e-01 -6.48179471e-01 -6.31098032e-01 -3.09816990e-02 6.54551625e-01 -8.00900698e-01 -1.20208815e-01 -9.96176898e-02 -6.41756415e-01 8.05493772e-01 6.19048297e-01 2.70945042e-01 -1.28238606e+00 -5.66690981e-01 -1.77537948e-02 -4.62672800e-01 -1.24596632e+00 -1.96075171e-01 5.06793439e-01 -5.75516999e-01 -9.92879212e-01 -7.44672120e-01 -8.63132298e-01 5.48245728e-01 -1.80967823e-01 9.00872231e-01 -1.17716432e-01 2.16320828e-01 7.02362418e-01 -7.91754186e-01 -5.66725254e-01 -3.03943157e-01 3.59329373e-01 -2.34746397e-01 2.47825995e-01 5.32187760e-01 9.29007120e-03 -2.09592149e-01 1.50043622e-01 -1.01233518e+00 2.82404590e-02 7.48137236e-01 1.02162623e+00 3.10658187e-01 -1.94399476e-01 5.94540179e-01 -7.04972088e-01 6.98327303e-01 -5.32938719e-01 -4.43828888e-02 2.51510859e-01 -7.51220882e-02 3.50819416e-02 3.92626107e-01 -2.96499729e-01 -1.02700722e+00 3.04172665e-01 -4.78163093e-01 -2.31923014e-01 -5.15501261e-01 8.83725941e-01 -3.61995399e-02 1.43641695e-01 2.71454662e-01 2.54270017e-01 -1.12009905e-01 -2.71940500e-01 4.96533781e-01 1.00856769e+00 4.86108631e-01 -4.41076726e-01 3.45458567e-01 2.05081552e-01 -4.13086832e-01 -7.66642034e-01 -6.29943669e-01 -4.31616127e-01 -5.09379327e-01 -3.65537882e-01 1.02289629e+00 -1.13502514e+00 -4.74745184e-01 5.05184412e-01 -8.84713709e-01 -3.13562959e-01 -1.49416998e-01 5.71139753e-01 -5.75674772e-01 1.49211362e-01 -7.96621382e-01 -5.77687383e-01 -2.70976841e-01 -1.41831517e+00 1.30524206e+00 4.65882182e-01 -1.97826117e-01 -1.21511829e+00 8.03797133e-03 5.40926337e-01 5.14382601e-01 -2.75486168e-02 7.36338496e-01 -5.38413465e-01 9.01251957e-02 -2.57887274e-01 -8.91702771e-02 4.08127815e-01 -1.19316494e-02 2.61932462e-01 -1.56282628e+00 -3.68280381e-01 -5.63281178e-01 -9.69730020e-01 1.02790070e+00 5.49957991e-01 8.20379496e-01 9.39174369e-02 -1.95788607e-01 4.46396440e-01 1.17393017e+00 2.34082080e-02 6.34330750e-01 5.05992651e-01 4.94232029e-01 5.28902709e-01 5.76542556e-01 3.93130273e-01 8.11040819e-01 3.43889117e-01 5.41320264e-01 -2.03196526e-01 1.44367339e-02 -1.77480001e-02 8.57640564e-01 1.11438227e+00 -2.43634343e-01 -4.55568701e-01 -6.70950770e-01 8.24563801e-01 -1.79374695e+00 -8.34318936e-01 1.43174097e-01 1.69396257e+00 1.01265883e+00 -1.38203025e-01 1.34787753e-01 3.12351603e-02 5.09884536e-01 1.97724655e-01 -2.65824407e-01 -9.06646729e-01 -4.95157719e-01 1.05201088e-01 2.30138108e-01 5.21670043e-01 -1.36888659e+00 9.69850898e-01 6.08673096e+00 5.16005337e-01 -1.27272904e+00 2.03716144e-01 4.32375044e-01 -1.73154339e-01 -3.65434527e-01 -1.86391875e-01 -6.37482345e-01 4.20744687e-01 1.50762665e+00 3.45047534e-01 2.62828529e-01 6.46165133e-01 -1.75659627e-01 -2.85159409e-01 -1.14027929e+00 9.43396747e-01 2.00282261e-01 -1.02351832e+00 -8.78777504e-02 -2.84172297e-01 6.24012053e-01 3.21161002e-01 3.27591628e-01 5.40121377e-01 -1.02785707e-01 -1.13813031e+00 9.08960104e-01 5.12116253e-01 9.81017470e-01 -1.03724194e+00 1.37220538e+00 -8.68256241e-02 -1.01973450e+00 -1.32821396e-01 -3.57938379e-01 1.89463675e-01 3.48967284e-01 1.75662622e-01 -7.19889045e-01 5.83160102e-01 5.93862772e-01 7.91401863e-01 -8.26877236e-01 7.21084416e-01 -1.76376000e-01 7.80809879e-01 -2.76930541e-01 -2.06435516e-01 5.14152586e-01 2.59066641e-01 1.92760468e-01 1.74698627e+00 2.11523205e-01 -3.75608265e-01 -1.06777251e-01 7.33659938e-02 -2.11940184e-01 1.61701158e-01 -5.13441920e-01 -3.83343190e-01 1.73382550e-01 1.32237720e+00 -2.12645650e-01 -4.02784169e-01 -8.84724796e-01 1.02996635e+00 2.80974418e-01 4.86953199e-01 -9.80425954e-01 -5.28493404e-01 4.37484503e-01 -5.11014581e-01 4.21948522e-01 2.33787559e-02 -2.91867584e-01 -1.25182688e+00 -1.93353310e-01 -1.12571084e+00 6.60697877e-01 -1.07034910e+00 -1.25946712e+00 8.75954568e-01 -1.78491548e-01 -1.15718448e+00 -5.57210088e-01 -9.79294658e-01 -2.41463587e-01 8.01481068e-01 -1.64104354e+00 -1.54204011e+00 -1.48510113e-02 9.72613633e-01 5.60152471e-01 -3.03391725e-01 1.14653122e+00 4.27050352e-01 -4.89076436e-01 8.59360874e-01 1.60147384e-01 2.86652446e-01 9.99680638e-01 -1.33615804e+00 -1.85445413e-01 5.39803386e-01 1.80219010e-01 4.85686362e-01 5.59259653e-01 -2.46400118e-01 -1.75479519e+00 -5.63570201e-01 1.19676185e+00 -4.90126729e-01 6.46392584e-01 -3.22825074e-01 -6.47691190e-01 8.19595039e-01 1.13860118e+00 -3.12245190e-01 9.17336285e-01 3.83559883e-01 -3.19961578e-01 -1.98313389e-02 -9.69268143e-01 2.61644274e-01 3.49249601e-01 -9.10539865e-01 -5.91853261e-01 1.37777120e-01 3.37902606e-01 -5.46983838e-01 -9.35523391e-01 3.04833293e-01 7.04524159e-01 -6.49889350e-01 6.96783245e-01 -6.80637598e-01 9.06723022e-01 -9.15796980e-02 -4.71042544e-01 -1.39729464e+00 -1.49352297e-01 -1.45306289e-01 -2.44262055e-01 1.29043710e+00 1.01089644e+00 -3.80544782e-01 3.46868962e-01 3.07699233e-01 -3.02313596e-01 -6.79529905e-01 -7.85917997e-01 -1.57793194e-01 4.34150584e-02 -6.58001482e-01 3.69806021e-01 1.16529465e+00 3.99369866e-01 6.46111131e-01 -3.76643986e-01 2.72816867e-01 3.82245295e-02 1.44456655e-01 6.25558436e-01 -5.54880381e-01 4.88589220e-02 -5.04070818e-01 -2.51824170e-01 -8.10883462e-01 2.47408435e-01 -8.43148112e-01 1.68924749e-01 -1.73635292e+00 2.79085338e-01 2.35389383e-03 -6.11169100e-01 9.51439738e-01 -1.29326001e-01 4.68536228e-01 1.95612401e-01 -3.67538363e-01 -8.20614696e-01 4.36644524e-01 1.09258580e+00 -3.32095802e-01 -1.20711215e-01 -3.56117845e-01 -5.83389163e-01 5.28150260e-01 9.55783665e-01 -1.74748853e-01 -2.99433023e-01 -6.54772639e-01 4.00923401e-01 -1.70815587e-01 1.39848173e-01 -8.60676587e-01 1.90427706e-01 -3.90580110e-02 4.92298722e-01 -6.24199212e-01 6.52774274e-01 -8.19558084e-01 -2.24334598e-01 -1.08975448e-01 -5.31350493e-01 2.55809635e-01 5.59055269e-01 9.18917730e-02 -7.26086199e-01 -3.37702274e-01 5.88378906e-01 -3.14860679e-02 -1.02976453e+00 -1.55255616e-01 -5.87974966e-01 -1.50077388e-01 5.24560094e-01 8.51092637e-02 -3.85564774e-01 -7.46771038e-01 -7.31131077e-01 3.29228133e-01 2.81317502e-01 5.70650876e-01 7.87648380e-01 -1.26687801e+00 -6.69628203e-01 -2.27515232e-02 3.21174741e-01 -3.12363297e-01 2.41596088e-01 1.12331736e+00 -3.89058053e-01 4.96572077e-01 -3.37577730e-01 -4.84576732e-01 -1.29370332e+00 3.27515036e-01 2.79166430e-01 -2.44364031e-02 -1.27812505e-01 1.04184139e+00 -1.93665162e-01 -7.30749547e-01 3.50357592e-01 -1.63740501e-01 -3.63847256e-01 3.56444120e-01 3.97381663e-01 1.48588913e-02 -2.37381700e-02 -1.12938035e+00 -5.37680805e-01 3.66013408e-01 -3.50285098e-02 -3.56800765e-01 1.36509013e+00 -3.68298173e-01 1.95666496e-02 7.13222146e-01 1.36842048e+00 -1.00540124e-01 -9.59639132e-01 -1.32264555e-01 -2.64104217e-01 -1.14733158e-02 1.30684361e-01 -1.09153593e+00 -1.13738883e+00 1.23288763e+00 5.89940906e-01 9.34657305e-02 1.42236984e+00 7.16391504e-02 8.02831590e-01 4.91437316e-01 8.41657594e-02 -1.31471789e+00 -9.61355492e-02 8.81497264e-01 7.50475109e-01 -1.37868631e+00 -3.08724821e-01 5.65266795e-02 -1.29940093e+00 1.17092264e+00 5.88521004e-01 3.02763313e-01 4.40581590e-01 2.47591957e-01 5.01559377e-01 -1.08674295e-01 -8.65536869e-01 -5.48686147e-01 4.77569252e-01 3.79640192e-01 7.95938671e-01 -4.87737469e-02 -6.95536137e-02 6.94523394e-01 -1.65851906e-01 -5.51889688e-02 4.60746020e-01 1.28148460e+00 -1.13063753e-01 -9.51917768e-01 -1.47802085e-01 4.04498369e-01 -4.88104731e-01 -3.00151110e-01 -4.61788267e-01 7.06930578e-01 2.35454276e-01 1.09824765e+00 -7.37723261e-02 -6.64918661e-01 2.89851069e-01 5.13500392e-01 5.49654365e-01 -3.76602739e-01 -7.32267857e-01 7.20791379e-03 5.47658980e-01 -3.98322523e-01 -9.28635299e-01 -7.47447610e-01 -1.35889506e+00 -2.24862233e-01 -2.74170458e-01 4.45283204e-02 8.45733821e-01 9.75088477e-01 3.26631844e-01 4.22837436e-01 5.92958987e-01 -9.19846773e-01 -3.00105177e-02 -1.31985950e+00 -3.38196814e-01 4.48676527e-01 5.56738138e-01 -5.69583476e-01 -2.23166183e-01 4.22119051e-01]	[13.210585594177246, 5.09957218170166]
07eccf41-3150-496b-aba7-4dfa97b74a77	automatic-segmentation-of-the-placenta-in	2208.02895	null	https://arxiv.org/abs/2208.02895v1	https://arxiv.org/pdf/2208.02895v1.pdf	Automatic Segmentation of the Placenta in BOLD MRI Time Series	Blood oxygen level dependent (BOLD) MRI with maternal hyperoxia can assess oxygen transport within the placenta and has emerged as a promising tool to study placental function. Measuring signal changes over time requires segmenting the placenta in each volume of the time series. Due to the large number of volumes in the BOLD time series, existing studies rely on registration to map all volumes to a manually segmented template. As the placenta can undergo large deformation due to fetal motion, maternal motion, and contractions, this approach often results in a large number of discarded volumes, where the registration approach fails. In this work, we propose a machine learning model based on a U-Net neural network architecture to automatically segment the placenta in BOLD MRI and apply it to segmenting each volume in a time series. We use a boundary-weighted loss function to accurately capture the placental shape. Our model is trained and tested on a cohort of 91 subjects containing healthy fetuses, fetuses with fetal growth restriction, and mothers with high BMI. We achieve a Dice score of 0.83+/-0.04 when matching with ground truth labels and our model performs reliably in segmenting volumes in both normoxic and hyperoxic points in the BOLD time series. Our code and trained model are available at https://github.com/mabulnaga/automatic-placenta-segmentation.	['Polina Golland', 'Esra Abaci Turk', 'P. Ellen Grant', 'Clinton J. Wang', 'Eileen Pan', 'Katherine Hobgood', 'Sean I. Young', 'S. Mazdak Abulnaga']	2022-08-04	null	null	null	null	['placenta-segmentation']	['medical']	[ 2.19051227e-01 2.63791859e-01 1.42283991e-01 -7.85101593e-01 -3.57856125e-01 -4.95259285e-01 9.85054020e-03 3.00091833e-01 -2.95417607e-01 4.29100901e-01 -1.11985557e-01 -5.92847914e-02 -2.19494067e-02 -6.95476651e-01 -6.77284658e-01 -6.95864379e-01 -6.90943003e-01 4.86906171e-01 1.92299083e-01 3.15196723e-01 1.13364868e-01 7.27630854e-01 -6.55098438e-01 2.20782995e-01 1.20230293e+00 8.60880852e-01 -2.12212786e-01 6.79419994e-01 -5.60677350e-02 4.08197135e-01 -4.49079163e-02 -6.16451323e-01 4.38207746e-01 -7.34664023e-01 -7.17658103e-01 -5.05498767e-01 6.23999715e-01 -5.05694330e-01 -3.74115556e-01 9.98328686e-01 8.99409533e-01 8.06921795e-02 5.66281021e-01 -9.97424424e-01 -1.30352199e-01 8.83018613e-01 -8.32292318e-01 7.57340670e-01 -2.82748103e-01 -1.56078115e-01 1.82068244e-01 -7.98062265e-01 6.41908288e-01 7.28884995e-01 9.48520184e-01 5.52212536e-01 -1.31004381e+00 -9.37102616e-01 -5.07906199e-01 -4.68259566e-02 -1.25369525e+00 -7.94919014e-01 7.29732752e-01 -8.82458270e-01 7.02654541e-01 8.68601799e-02 9.02859986e-01 5.08736134e-01 4.77672815e-01 1.15243830e-01 7.99137831e-01 -6.11698702e-02 5.98081341e-03 -5.96882463e-01 -2.04607084e-01 8.78957272e-01 3.38843107e-01 -1.70663476e-01 -1.73579335e-01 -1.10098019e-01 8.60539854e-01 -1.89227462e-01 -1.56403363e-01 -4.50765729e-01 -9.32404101e-01 5.67381263e-01 3.99642020e-01 5.54790556e-01 -5.24356008e-01 2.72567987e-01 5.79128623e-01 1.02002427e-01 6.94868565e-01 3.26993704e-01 -2.69183993e-01 -6.14494011e-02 -1.27048397e+00 -9.38813239e-02 4.84296471e-01 5.04498780e-01 5.70014477e-01 7.26516545e-02 -1.13610826e-01 9.46126401e-01 2.89643675e-01 5.06575465e-01 4.95088637e-01 -1.51984143e+00 2.80880779e-01 2.68051654e-01 -2.09315360e-01 -1.06845963e+00 -8.91259491e-01 -1.04332037e-01 -8.69644761e-01 1.12306498e-01 6.01699710e-01 -6.04750514e-01 -8.19659531e-01 1.74538219e+00 6.31430686e-01 5.40797949e-01 -2.85117388e-01 9.87101614e-01 9.30491209e-01 2.46468280e-02 1.40686110e-01 -5.57835340e-01 9.07527983e-01 -4.99329329e-01 -6.08648062e-01 8.23349655e-02 6.82695150e-01 -3.07469755e-01 2.82247782e-01 -1.05948627e-01 -1.62191725e+00 4.39835489e-02 -8.06110919e-01 6.94315135e-02 3.57066281e-03 -5.11564970e-01 5.31176269e-01 7.41145611e-01 -1.09084010e+00 1.09652138e+00 -1.57019436e+00 -1.07638024e-01 9.79160309e-01 4.51136440e-01 -5.70810437e-01 3.72063480e-02 -9.80944335e-01 1.11963940e+00 4.27233316e-02 3.65957379e-01 -7.63471484e-01 -1.35627115e+00 -1.01019537e+00 1.01490229e-01 -3.39790106e-01 -1.12673372e-01 9.20185685e-01 -7.72967756e-01 -1.06722260e+00 9.94562507e-01 -1.87349901e-01 -3.38845372e-01 9.31247711e-01 3.37389380e-01 -1.44482896e-01 6.99574232e-01 1.87402621e-01 6.65722251e-01 3.38362157e-01 -8.79851699e-01 3.96225378e-02 -5.19275308e-01 -6.43096805e-01 -8.63317028e-02 1.03676341e-01 3.64193410e-01 3.77387069e-02 -2.23392621e-01 6.52373731e-01 -7.56295800e-01 -1.26634568e-01 3.16994429e-01 2.84125209e-01 3.78856719e-01 1.53213844e-01 -1.07700980e+00 8.45236480e-01 -1.85807133e+00 -4.72428948e-01 3.22452724e-01 6.17790759e-01 1.19582176e-01 -2.74732113e-01 -1.30096987e-01 -3.38411868e-01 3.52353185e-01 -5.83215356e-01 1.28156975e-01 -3.51328641e-01 -4.88644056e-02 6.34619832e-01 1.23195171e+00 3.95788290e-02 1.22571349e+00 -1.25538278e+00 -7.78135359e-01 -1.02758043e-01 5.52232862e-01 -4.53502178e-01 1.52469516e-01 4.67344820e-01 6.21419013e-01 -2.24183843e-01 4.33819383e-01 1.03362954e+00 1.87079296e-01 3.20850283e-01 -1.42294019e-01 -3.24022025e-01 -6.09265231e-02 -4.17134017e-01 1.74121964e+00 3.97094488e-02 8.94327879e-01 2.08344698e-01 -1.12980497e+00 8.15075159e-01 7.60769367e-01 1.13456702e+00 -8.50910664e-01 5.46661496e-01 4.62014318e-01 8.22688401e-01 -1.14379704e+00 -5.92027485e-01 -5.85300565e-01 6.46565020e-01 4.11511481e-01 2.23759059e-02 -1.43746987e-01 4.98183697e-01 -3.50366771e-01 1.13170373e+00 1.40091479e-01 -1.87575117e-01 -5.64288735e-01 1.79725781e-01 -5.08431852e-01 8.92099500e-01 4.14084285e-01 -7.09387958e-01 1.04225743e+00 9.08645272e-01 -5.75226486e-01 -1.15587234e+00 -8.72513294e-01 -6.93702340e-01 6.50428355e-01 -1.79700851e-01 6.01971447e-01 -8.83569360e-01 -5.74363649e-01 1.14963949e-01 4.00146753e-01 -9.49461639e-01 -7.31609464e-02 -1.00000870e+00 -9.80920911e-01 1.03178394e+00 4.26790416e-01 1.32206634e-01 -1.07490671e+00 -8.93843949e-01 3.68510962e-01 -2.43996397e-01 -8.97649288e-01 -4.72825319e-01 -7.77262822e-02 -9.34013128e-01 -1.19042814e+00 -1.02475524e+00 -7.54559994e-01 9.73184049e-01 -2.49106795e-01 8.96618724e-01 2.37915456e-01 -4.69021797e-01 -8.71055424e-02 -4.18216549e-02 -4.79535729e-01 -5.26447654e-01 -2.01995835e-01 -6.72087371e-02 7.33919665e-02 5.29209909e-04 -1.02761555e+00 -1.02858496e+00 7.17620552e-02 -6.57070160e-01 1.43790944e-02 1.11403450e-01 3.67934465e-01 3.21518987e-01 -8.28281522e-01 6.45889819e-01 -7.27055967e-01 1.30699709e-01 -8.57750297e-01 -7.05859423e-01 1.61070257e-01 -4.74259615e-01 -4.32707727e-01 8.52769241e-02 -5.14057159e-01 -8.34939897e-01 -2.32071280e-01 -6.76847622e-02 -4.23503458e-01 1.44499876e-02 5.88069141e-01 3.93041670e-01 -4.51088727e-01 6.84442401e-01 -2.17441857e-01 3.61961722e-01 -2.08552569e-01 -1.56852365e-01 2.45512992e-01 4.52750325e-01 -3.74018908e-01 4.20168191e-01 4.81828749e-01 5.64745367e-01 -5.75357974e-01 -4.55463856e-01 -9.41248909e-02 -1.19328237e+00 -5.11181653e-01 8.97312880e-01 -3.72769237e-01 -7.49476969e-01 1.84297681e-01 -1.08857119e+00 -8.38434458e-01 1.26772895e-02 9.65474963e-01 -4.24547821e-01 3.18210632e-01 -6.71887577e-01 -4.07244384e-01 -6.95604086e-01 -9.84074175e-01 2.16965452e-01 3.48846614e-01 -7.13373795e-02 -1.25553465e+00 2.94449061e-01 2.24986807e-01 7.15666234e-01 1.02196217e+00 8.84100437e-01 -7.09263086e-01 9.22464654e-02 -2.28928879e-01 -2.71443427e-01 2.00683028e-01 1.30029455e-01 1.09479025e-01 -7.14745343e-01 -7.56507218e-02 1.33384869e-01 5.57647683e-02 4.84704643e-01 8.15255523e-01 9.09812450e-01 -2.68322323e-02 -3.71736139e-02 9.46073890e-01 1.14769673e+00 5.01996756e-01 4.56260383e-01 -2.54482388e-01 6.09570444e-01 8.99825633e-01 2.21585974e-01 3.53394568e-01 4.22402203e-01 -4.20483351e-02 4.02853817e-01 -2.39282414e-01 -1.82623461e-01 2.79080510e-01 -2.79296517e-01 7.40886927e-01 -3.67079347e-01 -4.42496017e-02 -1.51050854e+00 8.24044287e-01 -1.41753244e+00 -8.55088949e-01 -4.39777613e-01 2.15882325e+00 1.01462996e+00 -2.53270805e-01 -9.93650556e-02 -4.59766239e-01 8.84559035e-01 -2.03751832e-01 -6.14534855e-01 -5.75135052e-01 1.92054540e-01 3.04463565e-01 4.15220588e-01 5.81953704e-01 -7.94542551e-01 5.19194365e-01 6.38304520e+00 -2.43200302e-01 -1.56018174e+00 4.39293444e-01 8.72813225e-01 -2.80838698e-01 -1.24034598e-01 -2.88438529e-01 2.18812048e-01 4.87157702e-01 1.12216759e+00 -3.66826147e-01 3.78233790e-01 1.67877540e-01 6.57901227e-01 -7.16656819e-03 -1.08308506e+00 5.40327072e-01 2.13546269e-02 -1.17074430e+00 -7.43128121e-01 -2.24048674e-01 8.52436543e-01 5.38950086e-01 -3.69886130e-01 -2.83200592e-01 -4.94080424e-01 -1.25675941e+00 4.24056441e-01 6.70843422e-01 1.28574884e+00 -2.67495394e-01 7.25308299e-01 6.07423577e-03 -7.61101067e-01 4.04750198e-01 -2.52504677e-01 7.26097748e-02 -2.59224847e-02 4.06897247e-01 -9.25112247e-01 2.52231164e-03 5.49454749e-01 4.49849695e-01 -1.40864298e-01 1.62336707e+00 7.38770366e-02 7.60995865e-01 -3.49637270e-01 4.34126973e-01 -7.75491223e-02 -4.23495591e-01 3.36695999e-01 1.35451603e+00 4.89629179e-01 5.26293457e-01 -3.06666434e-01 1.03042006e+00 -2.88012832e-01 3.23568910e-01 -3.71564895e-01 1.29923850e-01 3.41438800e-01 1.41451013e+00 -9.55070972e-01 -2.26530313e-01 -5.24129868e-01 3.38010311e-01 2.81776220e-01 3.73269796e-01 -7.22454965e-01 -4.27621394e-01 2.31624752e-01 3.98516983e-01 -2.04039842e-01 -8.32043961e-02 -6.32691443e-01 -8.40185702e-01 -5.69526702e-02 -8.04266632e-02 1.53947741e-01 -5.21148264e-01 -7.95663178e-01 3.52717400e-01 1.34236678e-01 -7.22647190e-01 1.34107456e-01 8.58374238e-02 -8.94118786e-01 1.11278510e+00 -1.68158317e+00 -6.55835092e-01 -5.17256439e-01 -1.52167186e-01 -1.21725816e-02 3.71382296e-01 4.92132574e-01 7.97043920e-01 -5.51187336e-01 7.08929718e-01 -3.16800743e-01 5.61326325e-01 5.34775913e-01 -1.18516183e+00 1.00523971e-01 8.41857910e-01 -6.36321366e-01 7.03549206e-01 4.53634948e-01 -9.58278656e-01 -1.03154087e+00 -1.07420063e+00 1.03287482e+00 -4.01007272e-02 6.77064419e-01 -7.56288245e-02 -1.32837200e+00 8.27953696e-01 1.22957990e-01 9.12761331e-01 7.73107886e-01 -6.65251493e-01 6.78893402e-02 -2.50813425e-01 -1.64359641e+00 1.92946374e-01 8.63163710e-01 1.64317250e-01 -2.92298883e-01 2.28075191e-01 2.27529094e-01 -8.74372065e-01 -1.59146273e+00 6.69939041e-01 1.03656030e+00 -7.29078293e-01 5.49471676e-01 -5.22729516e-01 6.13141239e-01 2.26631500e-02 5.65303385e-01 -9.78198647e-01 -2.52753586e-01 -7.99579918e-01 1.03485964e-01 1.03605139e+00 4.87572342e-01 -7.51883447e-01 7.37346530e-01 1.12933505e+00 -3.21605802e-01 -6.95900977e-01 -1.22054231e+00 -2.63693869e-01 5.60141206e-01 -1.68100238e-01 5.08943975e-01 1.23681235e+00 1.31329849e-01 -4.94506449e-01 3.12315702e-01 1.43953636e-01 7.69741178e-01 -1.21677987e-01 -1.81048483e-01 -1.31166005e+00 7.17881262e-01 -6.15368783e-01 -2.57115424e-01 -1.29138485e-01 3.05081695e-01 -1.39981472e+00 3.03684473e-01 -1.57115972e+00 3.20736974e-01 -7.89070904e-01 -4.96112138e-01 8.61437500e-01 -6.41136393e-02 5.91996372e-01 -1.83113292e-01 -2.80613393e-01 -1.01901591e-01 1.50883675e-01 1.49642229e+00 1.53120682e-01 -2.17565894e-01 -1.89986572e-01 -3.29536557e-01 7.14173675e-01 9.92767572e-01 -8.17324162e-01 -1.06107555e-01 -5.24572372e-01 -1.38126547e-02 6.48592651e-01 6.10706732e-02 -5.79331338e-01 -4.76094149e-02 -2.57394552e-01 7.29921281e-01 -8.47079754e-02 -1.48897737e-01 -6.38475478e-01 9.47824344e-02 7.40491748e-01 -3.67356062e-01 -8.16346630e-02 8.16448331e-02 -5.15013993e-01 2.21514955e-01 -4.47712332e-01 1.23937762e+00 -1.80234566e-01 2.60291457e-01 7.76105165e-01 -2.99127191e-01 4.30726886e-01 7.86536157e-01 -6.00402020e-02 -6.14751168e-02 -3.29686478e-02 -8.13564777e-01 4.52987641e-01 3.19899827e-01 1.36522427e-01 5.61099708e-01 -1.04600751e+00 -1.19680810e+00 1.46386653e-01 -4.60945368e-01 -5.91117563e-03 4.26686734e-01 1.73125935e+00 -1.28434813e+00 -8.71639624e-02 -4.88581210e-01 -5.78743994e-01 -9.87591922e-01 9.67642665e-02 1.05734539e+00 3.95129323e-01 -7.36877799e-01 8.06742191e-01 1.67042725e-02 -2.26576731e-01 -1.81315735e-01 -2.37384915e-01 -3.48319322e-01 1.58882931e-01 3.85274023e-01 7.57203817e-01 8.59433636e-02 -8.93803835e-01 -2.46310174e-01 6.53158665e-01 4.32899088e-01 1.14584751e-01 1.42624533e+00 -2.12072432e-01 -6.38791144e-01 2.03226954e-01 1.46082497e+00 -1.10863827e-01 -1.23491955e+00 8.66767839e-02 -2.85818018e-02 -4.83908594e-01 -1.46921948e-02 -6.10573828e-01 -1.58358383e+00 1.03939545e+00 8.97653341e-01 2.65851170e-01 7.90112853e-01 -2.57934064e-01 6.50306821e-01 -2.83922762e-01 -1.61691383e-01 -5.29175282e-01 -4.47426021e-01 2.07640007e-01 9.75486755e-01 -1.22366071e+00 -5.30045442e-02 -2.48878255e-01 -3.68112922e-01 1.19989109e+00 6.11788034e-01 -1.14933364e-01 5.98899484e-01 5.74108243e-01 5.74733377e-01 -2.42964685e-01 -3.52626652e-01 2.41828412e-01 3.20549965e-01 5.96732199e-01 8.28678787e-01 -6.27394989e-02 -7.22405374e-01 3.03962171e-01 -1.00664742e-01 -6.38572872e-02 7.13873327e-01 7.98289359e-01 -3.11868787e-01 -5.99567533e-01 9.03203487e-02 7.14059591e-01 -1.12701440e+00 -8.71660411e-02 2.75289416e-01 3.66835356e-01 3.03335100e-01 6.65390909e-01 1.42836213e-01 4.75709021e-01 2.97734499e-01 3.57599169e-01 7.36156583e-01 -5.60135782e-01 -6.78687096e-01 2.33328789e-01 -1.45507693e-01 -5.51340044e-01 -2.76566029e-01 -1.20016205e+00 -1.87979531e+00 5.04772104e-02 -2.31553819e-02 4.53160703e-02 8.93500924e-01 6.91328585e-01 1.61646172e-01 4.27949071e-01 6.75366342e-01 -9.16069508e-01 1.23957098e-01 -1.16031277e+00 -6.43742383e-01 4.45656717e-01 6.01999044e-01 -5.62790394e-01 -1.64855078e-01 8.12521726e-02]	[14.054388999938965, -2.393808126449585]
b179926d-f0d4-4b40-922a-b2bcb216a9b7	semi-supervised-and-long-tailed-object	2305.14813	null	https://arxiv.org/abs/2305.14813v1	https://arxiv.org/pdf/2305.14813v1.pdf	Semi-Supervised and Long-Tailed Object Detection with CascadeMatch	This paper focuses on long-tailed object detection in the semi-supervised learning setting, which poses realistic challenges, but has rarely been studied in the literature. We propose a novel pseudo-labeling-based detector called CascadeMatch. Our detector features a cascade network architecture, which has multi-stage detection heads with progressive confidence thresholds. To avoid manually tuning the thresholds, we design a new adaptive pseudo-label mining mechanism to automatically identify suitable values from data. To mitigate confirmation bias, where a model is negatively reinforced by incorrect pseudo-labels produced by itself, each detection head is trained by the ensemble pseudo-labels of all detection heads. Experiments on two long-tailed datasets, i.e., LVIS and COCO-LT, demonstrate that CascadeMatch surpasses existing state-of-the-art semi-supervised approaches -- across a wide range of detection architectures -- in handling long-tailed object detection. For instance, CascadeMatch outperforms Unbiased Teacher by 1.9 AP Fix on LVIS when using a ResNet50-based Cascade R-CNN structure, and by 1.7 AP Fix when using Sparse R-CNN with a Transformer encoder. We also show that CascadeMatch can even handle the challenging sparsely annotated object detection problem.	['Chen Change Loy', 'Chen Huang', 'Kaiyang Zhou', 'Yuhang Zang']	2023-05-24	null	null	null	null	['pseudo-label']	['miscellaneous']	[ 2.04430401e-01 3.89403522e-01 -3.46794695e-01 -3.77491742e-01 -9.32541788e-01 -5.32136142e-01 3.73121738e-01 -1.21523459e-02 -6.74261570e-01 4.76616085e-01 -3.28009695e-01 -2.33852714e-01 3.02663088e-01 -3.73633772e-01 -1.00218606e+00 -6.28410876e-01 1.48280933e-01 6.67363524e-01 7.53648818e-01 1.78091168e-01 -1.57552257e-01 2.20804811e-01 -1.69861245e+00 3.90167952e-01 4.05330300e-01 1.16404843e+00 4.69532572e-02 7.08845079e-01 1.87557951e-01 1.02215791e+00 -7.12863266e-01 -5.93124032e-01 3.16642791e-01 -3.43097091e-01 -4.69537765e-01 2.26309910e-01 8.78517270e-01 -4.68556374e-01 -1.95594072e-01 1.20205927e+00 5.61209142e-01 -4.57082421e-01 7.06675708e-01 -1.30153835e+00 -5.94759583e-01 1.01828921e+00 -9.72848594e-01 4.06285137e-01 -4.52906303e-02 1.34869531e-01 1.27113509e+00 -1.44116235e+00 4.15085852e-01 1.23553014e+00 9.34675753e-01 6.39150202e-01 -1.44778717e+00 -1.07641578e+00 1.19351424e-01 8.02204162e-02 -1.71750009e+00 -1.94995418e-01 2.91098177e-01 -4.44136351e-01 7.17078328e-01 -2.34232768e-01 4.35962617e-01 1.09969115e+00 -2.76618004e-01 1.29905844e+00 1.19641638e+00 -4.64244217e-01 1.04330843e-02 3.94063741e-01 2.51820177e-01 1.07793903e+00 4.88013953e-01 2.05760688e-01 -6.88643754e-01 2.36213356e-02 5.17660320e-01 -1.60031080e-01 1.01135239e-01 -3.79423320e-01 -1.09857130e+00 8.68654847e-01 6.15078211e-01 7.62148667e-03 4.41167131e-03 1.25490174e-01 3.53848517e-01 2.61261940e-01 6.28825128e-01 2.99039543e-01 -6.81404948e-01 4.29340154e-01 -1.35409820e+00 -5.42562120e-02 6.30814016e-01 1.30410421e+00 6.90722048e-01 1.11118138e-01 -5.34584105e-01 6.32052660e-01 4.24652815e-01 4.88374442e-01 3.32417846e-01 -5.71164548e-01 1.24330916e-01 7.57950962e-01 1.81138724e-01 -2.00773686e-01 -4.87912446e-01 -1.02228558e+00 -5.92995465e-01 3.11321914e-01 4.68130022e-01 -1.39213741e-01 -1.13689697e+00 1.68334126e+00 5.90822935e-01 2.94579178e-01 -2.19507158e-01 8.02463770e-01 1.11789477e+00 2.77584553e-01 3.69268358e-02 -2.09248856e-01 1.55041599e+00 -1.41550756e+00 -2.94977874e-01 -2.38924518e-01 7.97975302e-01 -5.85712373e-01 9.05037463e-01 4.97747302e-01 -1.06649256e+00 -6.97364986e-01 -1.09865952e+00 5.28131844e-03 -1.37258634e-01 7.62365162e-01 1.60898283e-01 4.79080647e-01 -1.09583950e+00 1.72298074e-01 -5.91461599e-01 -2.03057855e-01 7.23037899e-01 2.87935555e-01 -2.16986775e-01 4.45629433e-02 -7.85963953e-01 9.73609984e-01 4.88693237e-01 -9.74494368e-02 -1.47637475e+00 -6.27089441e-01 -6.40868008e-01 7.44302571e-02 7.50804126e-01 -4.54843789e-01 1.91060889e+00 -9.36140954e-01 -9.28117454e-01 1.41821837e+00 1.42143276e-02 -8.53549957e-01 7.87303269e-01 -3.68017048e-01 -1.88903272e-01 8.64034593e-02 5.25316358e-01 1.20882988e+00 1.28660285e+00 -1.03550982e+00 -1.12791193e+00 -1.72567382e-01 -1.43134058e-01 5.18183783e-02 -4.22002822e-01 3.94097954e-01 -4.54360306e-01 -5.58410525e-01 -6.57910630e-02 -9.89496827e-01 -2.57042140e-01 3.22228014e-01 -6.24504209e-01 -7.49660194e-01 8.18250477e-01 1.47840068e-01 1.04228985e+00 -2.25846457e+00 -1.90511242e-01 -1.36438072e-01 6.32005990e-01 6.86542690e-01 -1.49449155e-01 -8.95229578e-02 -8.17724392e-02 -1.40584633e-01 -1.07233651e-01 -8.28177333e-01 -1.93688691e-01 2.16290534e-01 -3.99862617e-01 6.15970194e-01 4.48044062e-01 9.38798964e-01 -1.01142812e+00 -9.62021887e-01 -1.67859215e-02 2.65978158e-01 -2.97167569e-01 4.06440526e-01 -2.90041447e-01 2.43986631e-03 -2.07277372e-01 7.30017066e-01 5.89215577e-01 -7.61681795e-01 -1.92097545e-01 -7.41794929e-02 -1.31393716e-01 1.84379425e-02 -1.10275531e+00 1.20645177e+00 -1.50285721e-01 6.55435205e-01 1.95230931e-01 -8.61965299e-01 8.76712441e-01 1.10062934e-01 9.27179009e-02 -2.63991952e-01 2.18123496e-01 2.22953275e-01 -8.46925750e-02 -9.96448845e-02 3.21941435e-01 2.59786457e-01 1.18402295e-01 5.34334660e-01 5.25680423e-01 1.56774875e-02 4.31971937e-01 6.50758147e-01 1.19743097e+00 -2.46050879e-02 3.32146198e-01 -9.86717194e-02 4.14383292e-01 1.77987561e-01 6.96354628e-01 1.35512745e+00 -3.65059167e-01 6.48717403e-01 4.15540099e-01 -3.23920161e-01 -8.88225079e-01 -1.00419176e+00 -3.61459702e-01 1.72464335e+00 1.93938578e-03 -4.23339605e-01 -7.62542665e-01 -1.21524167e+00 2.43875608e-01 5.34087539e-01 -8.13832819e-01 -1.49821714e-01 -2.68345684e-01 -7.01312244e-01 8.56170654e-01 9.06176209e-01 4.24893945e-01 -1.11146808e+00 -6.78933501e-01 2.82974422e-01 1.90687925e-01 -1.32858753e+00 -5.29382110e-01 7.91795313e-01 -5.65803945e-01 -1.22855008e+00 -6.92927957e-01 -1.03743160e+00 9.30615187e-01 5.44561028e-01 1.33975732e+00 3.07928205e-01 -5.43596447e-01 2.93043256e-01 -4.04864728e-01 -6.70169830e-01 -6.51669860e-01 1.24254838e-01 1.29763886e-01 2.00486351e-02 6.64877951e-01 9.18881893e-02 -5.52343309e-01 7.37652838e-01 -4.73963588e-01 -1.17054567e-01 8.05184901e-01 1.09190142e+00 7.29721487e-01 -2.39158362e-01 6.51886463e-01 -1.00850475e+00 6.31268397e-02 -4.19292122e-01 -8.80035698e-01 2.99163431e-01 -8.59044135e-01 -1.19752161e-01 3.88171315e-01 -8.45379770e-01 -8.74672592e-01 5.15605032e-01 1.80059206e-02 -7.33204186e-01 -1.79559842e-01 -1.44699261e-01 3.82096976e-01 -1.90823406e-01 1.07368279e+00 1.73201591e-01 -3.46880406e-01 -4.22532201e-01 4.82976764e-01 7.16744363e-01 7.65327454e-01 -1.18339404e-01 9.81833160e-01 5.20704985e-01 -2.14157104e-01 -4.25992131e-01 -1.89589012e+00 -9.50604141e-01 -6.54681385e-01 -1.87413052e-01 7.96693742e-01 -1.48730707e+00 -5.19583762e-01 4.34349000e-01 -1.09832168e+00 -3.12857151e-01 -4.82934505e-01 2.59731650e-01 -2.71035522e-01 6.43697232e-02 -8.85918617e-01 -8.46290648e-01 -4.80898350e-01 -1.01112127e+00 1.26104832e+00 2.28726849e-01 1.52930794e-02 -3.44816089e-01 -1.46247640e-01 3.00797462e-01 1.77891716e-01 -3.44507426e-01 9.20601785e-02 -9.72127736e-01 -3.87050927e-01 -3.30774367e-01 -6.31209552e-01 6.29047394e-01 -4.71877426e-01 6.21316135e-02 -1.12996864e+00 -4.60765690e-01 -3.41431558e-01 -1.07264233e+00 1.41578615e+00 2.40962774e-01 1.15890622e+00 -3.87524962e-02 -5.33118784e-01 3.36810350e-01 8.97600830e-01 -5.21182895e-01 -5.06676026e-02 -1.07153216e-02 5.77938199e-01 3.24408293e-01 8.65311325e-01 5.16054392e-01 3.15293640e-01 4.69965070e-01 7.00476646e-01 -3.19950044e-01 -5.06029606e-01 -5.85249364e-01 3.73885036e-01 3.29456866e-01 3.78653616e-01 -1.80241331e-01 -6.87663317e-01 6.59416199e-01 -1.91790593e+00 -7.36900449e-01 -4.02056456e-01 1.85822296e+00 1.01318038e+00 5.60816467e-01 4.05469865e-01 1.97297484e-01 9.77931499e-01 -2.34190583e-01 -6.72583520e-01 1.46230146e-01 -2.10049406e-01 1.31224141e-01 6.69875324e-01 7.82340765e-03 -1.43622279e+00 1.30491674e+00 6.58238602e+00 9.85715508e-01 -6.97028041e-01 5.19892275e-01 5.37286758e-01 -6.52317330e-02 4.76994365e-01 -1.47821859e-01 -1.57822049e+00 1.97170645e-01 6.00999296e-01 2.27051958e-01 -2.47903243e-01 1.50029314e+00 -3.05766702e-01 -1.18536055e-01 -1.31686223e+00 8.88555050e-01 2.80951858e-01 -1.13269687e+00 -2.62616873e-01 -3.04719597e-01 9.39815760e-01 5.67960143e-01 2.08594322e-01 6.55231893e-01 8.08618426e-01 -7.28919148e-01 1.23552537e+00 -9.69702676e-02 1.07345676e+00 -3.72251868e-01 6.29182816e-01 6.92628503e-01 -1.20320058e+00 -3.32397223e-01 -4.91092533e-01 1.26886413e-01 -8.47135633e-02 7.67815828e-01 -1.25124705e+00 -1.91113696e-01 1.00050938e+00 6.41404986e-01 -9.18103516e-01 1.26424468e+00 -9.48799968e-01 1.22040439e+00 -4.93130088e-01 -1.31406739e-01 1.97793037e-01 5.21954119e-01 2.86111116e-01 1.68264472e+00 -4.60417904e-02 -2.34689713e-01 4.33871210e-01 1.02364850e+00 -3.42298806e-01 -2.25620255e-01 -4.14931238e-01 4.57388133e-01 6.17299736e-01 1.67608571e+00 -8.22393298e-01 -7.03437030e-01 -5.86001277e-01 6.60569668e-01 7.10537672e-01 -1.79063994e-02 -9.22466278e-01 4.78766710e-02 -6.21798299e-02 7.82466531e-02 6.68460250e-01 1.63793162e-01 -1.13073010e-02 -1.05050623e+00 -2.98449069e-01 -6.31427109e-01 6.99449897e-01 -8.29331040e-01 -1.41591811e+00 4.15958673e-01 -5.68887070e-02 -1.13436127e+00 -8.34627822e-02 -6.35051489e-01 -5.43472350e-01 1.24507859e-01 -1.57960987e+00 -1.27923226e+00 -1.62648335e-01 6.09436035e-01 6.79716170e-01 -8.99110213e-02 4.44969684e-01 2.44482249e-01 -7.42752790e-01 8.86401236e-01 -2.86577165e-01 3.41424465e-01 8.29134345e-01 -1.53610742e+00 4.66621369e-01 8.24993432e-01 6.26025975e-01 1.77944139e-01 5.96516192e-01 -4.68724221e-01 -8.70645583e-01 -1.52541852e+00 6.97339296e-01 -5.72691321e-01 7.70251811e-01 -5.70386529e-01 -8.53950679e-01 8.65433753e-01 -5.92968427e-02 8.02739501e-01 4.38423604e-01 -4.82436530e-02 -7.81027734e-01 5.48172742e-02 -1.08402622e+00 2.85026997e-01 1.24204576e+00 -4.58500445e-01 -6.88697815e-01 6.87214375e-01 7.23110914e-01 -4.33019608e-01 -4.11478847e-01 5.50690174e-01 2.75415838e-01 -8.94922733e-01 7.67668128e-01 -4.83191222e-01 4.04602885e-02 -5.02742290e-01 1.92867264e-01 -9.60515261e-01 -4.85194534e-01 -4.18984920e-01 -5.23525059e-01 1.04816604e+00 6.37648284e-01 -3.18809628e-01 9.97625828e-01 -6.71907067e-02 -1.06901281e-01 -6.54833019e-01 -8.62358510e-01 -8.85868549e-01 -2.14153975e-01 -3.75176281e-01 8.39390159e-02 6.08374953e-01 -4.02868360e-01 5.60096800e-01 -3.44463468e-01 2.70014167e-01 1.04112494e+00 4.20073867e-02 9.46125031e-01 -1.27253032e+00 -3.94998908e-01 -3.28413814e-01 -1.83024034e-01 -1.42122149e+00 2.05572218e-01 -8.92943501e-01 4.35405403e-01 -9.97688353e-01 5.50727665e-01 -4.54746455e-01 -4.84572291e-01 8.66079211e-01 -2.03726068e-01 8.85938823e-01 8.73523112e-03 2.74900317e-01 -1.37198007e+00 2.98370242e-01 9.68270600e-01 -1.22770458e-01 7.64162764e-02 2.09170952e-01 -4.76994038e-01 1.06088316e+00 4.90838766e-01 -1.01323509e+00 -1.66066270e-02 -2.08968699e-01 3.89180750e-01 -1.71513781e-01 5.77876329e-01 -9.91080105e-01 6.08067751e-01 3.36682886e-01 4.55387324e-01 -1.04136169e+00 1.88486967e-02 -5.24754882e-01 -6.19071782e-01 5.60311735e-01 -5.54090798e-01 -9.97762680e-02 -2.08469331e-01 7.44800866e-01 -2.50153765e-02 -3.91592473e-01 1.05549455e+00 -6.17598630e-02 -6.88407004e-01 3.40575218e-01 -3.43135893e-01 3.04982394e-01 1.20567048e+00 -3.63567658e-02 -3.38971436e-01 -4.64809276e-02 -7.21248150e-01 6.15212619e-01 1.23293251e-01 3.48534435e-01 5.79257965e-01 -9.93016124e-01 -9.27780211e-01 2.31135592e-01 5.33274472e-01 3.11851054e-01 -2.96968043e-01 7.82197952e-01 8.04648921e-02 -1.68664905e-04 3.03815424e-01 -1.08810544e+00 -1.38269186e+00 6.15136206e-01 2.61991978e-01 -3.08573753e-01 -6.10470891e-01 1.42606962e+00 3.79957885e-01 -3.13144803e-01 6.43384576e-01 -1.62039459e-01 5.02876611e-03 1.18829779e-01 6.19354606e-01 2.43964255e-01 -6.41521215e-02 -6.65360987e-01 -3.17511052e-01 4.05919522e-01 -5.26068449e-01 1.64023101e-01 9.74603415e-01 7.93628320e-02 2.79183060e-01 2.89935321e-01 7.84978390e-01 -2.52774060e-01 -1.54744768e+00 -7.62985766e-01 1.65922850e-01 -7.24427775e-02 2.01814502e-01 -8.30641210e-01 -1.00780475e+00 7.26171494e-01 5.10375738e-01 9.41646025e-02 7.43155897e-01 5.50963163e-01 3.36317480e-01 5.52228749e-01 3.10094863e-01 -1.12614357e+00 6.16567314e-01 4.50013757e-01 5.52536428e-01 -1.57425451e+00 -3.05785742e-02 -3.08427870e-01 -6.04698956e-01 7.88507700e-01 1.13423812e+00 -1.58388510e-01 6.02494478e-01 5.49488962e-01 -2.31267381e-02 -2.62883812e-01 -1.00943291e+00 -7.03069568e-01 2.23016232e-01 2.83032298e-01 1.34368718e-01 9.73866135e-02 1.15131937e-01 3.80654931e-01 5.74712381e-02 -6.99575245e-02 4.42714930e-01 7.92257667e-01 -7.71529853e-01 -6.46953881e-01 -4.62434232e-01 7.27827966e-01 -5.58108032e-01 -8.78122747e-02 -5.51321089e-01 6.56522989e-01 4.57770556e-01 1.04365134e+00 5.86908534e-02 -2.30726823e-01 1.99022919e-01 -3.58136036e-02 3.30474406e-01 -1.05814683e+00 -8.02490592e-01 5.09842932e-02 -6.71025217e-02 -4.23840582e-01 -5.33300459e-01 -5.48945546e-01 -1.10940087e+00 2.50806957e-01 -1.15656602e+00 -7.56061375e-02 5.15078545e-01 6.93276107e-01 1.20754629e-01 2.67795563e-01 5.51820040e-01 -7.90935814e-01 -1.10194695e+00 -1.18329167e+00 -5.62253237e-01 2.82381386e-01 4.75877255e-01 -9.02537465e-01 -5.51402509e-01 -2.13218316e-01]	[9.220699310302734, 1.3178293704986572]
b1f2ff78-5902-43f1-aec7-5d01ccaa4bd7	modeling-personalized-item-frequency	2006.00556	null	https://arxiv.org/abs/2006.00556v1	https://arxiv.org/pdf/2006.00556v1.pdf	Modeling Personalized Item Frequency Information for Next-basket Recommendation	Next-basket recommendation (NBR) is prevalent in e-commerce and retail industry. In this scenario, a user purchases a set of items (a basket) at a time. NBR performs sequential modeling and recommendation based on a sequence of baskets. NBR is in general more complex than the widely studied sequential (session-based) recommendation which recommends the next item based on a sequence of items. Recurrent neural network (RNN) has proved to be very effective for sequential modeling and thus been adapted for NBR. However, we argue that existing RNNs cannot directly capture item frequency information in the recommendation scenario. Through careful analysis of real-world datasets, we find that {\em personalized item frequency} (PIF) information (which records the number of times that each item is purchased by a user) provides two critical signals for NBR. But, this has been largely ignored by existing methods. Even though existing methods such as RNN based methods have strong representation ability, our empirical results show that they fail to learn and capture PIF. As a result, existing methods cannot fully exploit the critical signals contained in PIF. Given this inherent limitation of RNNs, we propose a simple item frequency based k-nearest neighbors (kNN) method to directly utilize these critical signals. We evaluate our method on four public real-world datasets. Despite its relative simplicity, our method frequently outperforms the state-of-the-art NBR methods -- including deep learning based methods using RNNs -- when patterns associated with PIF play an important role in the data.	['Zhi-Li Zhang', 'Jinyang Gao', 'Haoji Hu', 'Xiangnan He']	2020-05-31	null	null	null	null	['next-basket-recommendation']	['miscellaneous']	[ 5.96677698e-02 -7.08421528e-01 -7.88273454e-01 -3.48263562e-01 -3.45543176e-01 -5.29865026e-01 1.71959385e-01 -1.46949530e-01 -2.75203705e-01 3.71846229e-01 7.15737104e-01 -5.09605944e-01 -4.53777909e-01 -1.02078450e+00 -8.47939014e-01 -4.42181468e-01 -2.06992373e-01 2.65944362e-01 -3.19625199e-01 -6.38075709e-01 2.93625057e-01 2.75341719e-01 -1.42971015e+00 7.53298700e-01 3.15885127e-01 1.01408505e+00 5.45550466e-01 1.12656944e-01 -2.07905799e-01 9.62475538e-01 -3.59409213e-01 -2.74995595e-01 4.35625017e-01 -4.73770380e-01 -5.36879182e-01 -3.74424785e-01 -1.93894550e-01 -5.81937313e-01 -7.38944471e-01 8.09950590e-01 4.50168312e-01 6.54163003e-01 6.22181594e-01 -6.45291626e-01 -1.15280366e+00 1.52547240e+00 -5.60965478e-01 4.59644437e-01 2.98061877e-01 -2.87711799e-01 1.47717321e+00 -9.34335291e-01 4.45713282e-01 7.30556905e-01 1.00818682e+00 5.26074171e-01 -1.12533498e+00 -7.90941119e-01 6.38740897e-01 3.61374080e-01 -1.19295001e+00 -3.95215303e-02 9.93621230e-01 -4.55000764e-03 1.07700932e+00 2.17785135e-01 8.73907149e-01 1.77015054e+00 -2.93647293e-02 1.28033113e+00 6.62997603e-01 -2.47035041e-01 8.61230865e-03 1.63464174e-02 4.31024671e-01 -1.04647137e-01 5.16513921e-02 2.38886997e-01 -6.25159919e-01 -6.89095855e-02 8.88620436e-01 9.12696838e-01 -3.06928843e-01 3.21392059e-01 -1.15954709e+00 1.02609622e+00 4.81012374e-01 6.57264948e-01 -8.01092982e-01 6.54735416e-02 3.89391899e-01 4.63659137e-01 3.97654891e-01 3.50407362e-01 -7.15861142e-01 -1.71935916e-01 -1.03253651e+00 2.50384390e-01 7.57639825e-01 8.02921832e-01 3.12887281e-01 3.38733554e-01 -1.34364426e-01 1.16221499e+00 2.87758201e-01 2.62005776e-01 9.27928150e-01 -2.79615253e-01 4.58777308e-01 2.46172115e-01 2.01274931e-01 -1.25388575e+00 -4.47709560e-01 -7.82383144e-01 -1.14699149e+00 -6.56962156e-01 3.67479801e-01 3.70401517e-02 -7.42961109e-01 1.50026941e+00 -1.64956108e-01 5.68853617e-01 -1.59538016e-02 9.71573949e-01 8.64829481e-01 8.66701961e-01 -1.73190489e-01 -3.34837109e-01 1.04478371e+00 -7.87412405e-01 -6.59439623e-01 1.68979988e-01 3.35930854e-01 -6.55974329e-01 1.04296148e+00 7.63613760e-01 -7.41427541e-01 -7.09176421e-01 -8.97337079e-01 1.96365476e-01 -3.16584527e-01 6.23308234e-02 9.81813073e-01 5.91160595e-01 -6.29799604e-01 9.13707018e-01 -5.13078630e-01 -1.87736660e-01 1.60339758e-01 2.61297345e-01 6.49954453e-02 -2.31224611e-01 -1.52629542e+00 4.50914800e-01 1.03986137e-01 4.61620241e-01 -7.21522510e-01 -9.13002193e-01 -3.97754550e-01 2.34623447e-01 4.71355170e-01 -3.71357113e-01 1.29996228e+00 -9.97343004e-01 -1.65682495e+00 5.13300188e-02 -1.24099016e-01 -8.97802591e-01 1.98397383e-01 -5.56351483e-01 -9.57943857e-01 -4.36448276e-01 -1.90388694e-01 8.41147900e-02 6.31876528e-01 -1.14283383e+00 -8.65910113e-01 -2.24975869e-01 3.39538336e-01 7.27268681e-02 -5.87379336e-01 -1.83233693e-01 -4.20095950e-01 -1.29609621e+00 1.29659504e-01 -8.93392861e-01 -3.81528914e-01 -1.06610322e+00 -4.19247270e-01 -3.95923853e-01 2.74109989e-01 -4.86662447e-01 1.84564269e+00 -2.12451220e+00 -2.60704160e-01 4.60209370e-01 -1.17883822e-02 2.24230587e-01 -2.43706509e-01 7.15653062e-01 -1.97546169e-01 7.11236298e-02 6.34518802e-01 -9.92072523e-02 1.76532775e-01 1.67695805e-01 -9.86244440e-01 3.58402044e-01 -7.02504575e-01 1.01763487e+00 -8.86268497e-01 3.79653454e-01 2.73267239e-01 7.50130296e-01 -6.14849508e-01 3.50016430e-02 -2.48609304e-01 1.28148675e-01 -3.88549745e-01 5.28167129e-01 4.64833379e-01 -5.07661045e-01 4.89172816e-01 -5.10646164e-01 1.04219049e-01 8.55857491e-01 -1.23473203e+00 1.66401327e+00 -5.52922189e-01 1.45977870e-01 -7.14291096e-01 -1.15570498e+00 8.57856750e-01 3.96357208e-01 7.50115693e-01 -1.03902292e+00 2.05762219e-03 1.09874383e-01 1.26692995e-01 -1.68950856e-01 8.09096873e-01 -3.89492549e-02 5.81601681e-03 7.10161626e-01 -4.90349196e-02 1.13516569e+00 -1.14697376e-02 6.88743517e-02 1.01509142e+00 1.96842160e-02 3.43902677e-01 -4.33674455e-02 2.27305237e-02 -3.64724874e-01 6.97488964e-01 1.36341763e+00 2.76109189e-01 6.26081586e-01 -2.28907883e-01 -7.83235013e-01 -9.62036133e-01 -7.76967645e-01 -2.31510010e-02 1.50285256e+00 8.81396830e-02 -5.36841214e-01 -7.82302991e-02 -6.92281544e-01 2.17495635e-01 9.60433424e-01 -8.86472702e-01 -1.27341866e-01 -6.84087098e-01 -9.91441607e-01 3.01483631e-01 7.71123350e-01 9.75686163e-02 -1.37068081e+00 -9.69193503e-02 6.88565910e-01 -3.92194569e-01 -7.03681648e-01 -7.08022654e-01 3.46355200e-01 -8.52506816e-01 -6.27298176e-01 -7.63470054e-01 -4.52928394e-01 1.51215985e-01 7.85707653e-01 1.36188376e+00 -7.27187693e-02 2.31468469e-01 2.04400554e-01 -9.42487001e-01 -1.67448282e-01 -2.29589809e-02 3.18615824e-01 3.57654452e-01 2.21316144e-01 9.43406701e-01 -9.14054215e-01 -9.59922314e-01 6.14059508e-01 -8.39946806e-01 -1.62435010e-01 5.39780080e-01 7.67909884e-01 7.33152032e-01 3.29985678e-01 1.07101166e+00 -1.32583058e+00 7.37408042e-01 -9.86212134e-01 -2.08414376e-01 9.84727293e-02 -7.25144327e-01 -4.09058481e-01 1.08154690e+00 -9.59330022e-01 -8.01642418e-01 -4.75214154e-01 -4.18724835e-01 -5.19400001e-01 -1.82164326e-01 9.74841356e-01 1.89408123e-01 5.83705962e-01 5.50260842e-01 5.86303711e-01 -6.49599850e-01 -9.80998933e-01 2.31128037e-01 6.44228935e-01 1.92868665e-01 -4.09556836e-01 3.29834640e-01 4.18689430e-01 -5.00736296e-01 -3.20693105e-01 -1.27136993e+00 -8.07031393e-01 -1.66863158e-01 -5.92995659e-02 1.33914649e-01 -1.00639343e+00 -1.00810885e+00 2.81214640e-02 -6.43639982e-01 -1.40783727e-01 -3.92121404e-01 8.97101462e-01 -3.31251502e-01 5.45756035e-02 -9.04050648e-01 -8.64527404e-01 -5.69122910e-01 -7.33884990e-01 4.44138378e-01 -5.45668304e-02 -3.22299987e-01 -8.72220993e-01 -4.25141193e-02 1.80132210e-01 6.29928827e-01 -3.52885246e-01 7.13927031e-01 -9.78490233e-01 -2.52156109e-01 -1.46781266e-01 5.35351634e-02 1.45164326e-01 5.02005219e-01 -5.71457505e-01 -5.74390411e-01 -3.36164296e-01 2.56659873e-02 8.26157406e-02 1.04913116e+00 7.65668750e-01 1.38773918e+00 -4.79640126e-01 -2.97947526e-01 5.30167043e-01 1.64470053e+00 5.52803397e-01 6.44355059e-01 3.94714683e-01 7.08887696e-01 5.01920097e-02 4.52420235e-01 5.78936636e-01 3.66638005e-01 7.31118679e-01 1.66847199e-01 1.88666388e-01 7.97138140e-02 -6.88874066e-01 5.15426874e-01 1.10540354e+00 -4.61858004e-01 -6.58911109e-01 -3.58538330e-01 6.32774770e-01 -2.16799736e+00 -1.37996364e+00 1.60539314e-01 2.17818522e+00 7.25740314e-01 6.70538843e-02 4.03223336e-01 1.05101436e-01 5.05468428e-01 1.97567344e-02 -7.87584126e-01 -8.44448283e-02 -7.40979165e-02 2.62904055e-02 5.56387901e-01 -1.47034049e-01 -9.69911277e-01 6.97298944e-01 5.98502970e+00 1.17298567e+00 -1.03303242e+00 8.31227377e-02 5.68785787e-01 -4.55132991e-01 -4.94406164e-01 -5.71381807e-01 -1.21609068e+00 8.52995038e-01 1.07182896e+00 1.93894267e-01 9.38049495e-01 8.30316722e-01 4.11320210e-01 4.31334794e-01 -1.21139288e+00 1.14539766e+00 8.62860382e-02 -1.59141886e+00 3.68293226e-01 1.34232059e-01 7.95324624e-01 2.27305278e-01 4.33731318e-01 6.63932204e-01 6.30332291e-01 -1.13636386e+00 5.27106762e-01 5.97509265e-01 4.59437013e-01 -1.01136279e+00 8.17504525e-01 4.13911998e-01 -1.29817772e+00 -4.40389752e-01 -7.51834691e-01 -3.45006436e-02 2.65212595e-01 8.99303019e-01 -3.85180295e-01 7.29647815e-01 1.05590212e+00 1.25641406e+00 1.24010533e-01 7.77724147e-01 1.15074292e-01 1.21687317e+00 -3.46865177e-01 -1.16654746e-01 3.67081463e-01 -3.45587790e-01 3.09108973e-01 1.20922053e+00 5.36046207e-01 2.01447412e-01 1.17724776e-01 6.70565724e-01 -3.94688308e-01 3.18864256e-01 -5.94952106e-01 3.15663218e-02 3.45231682e-01 1.02320087e+00 -8.19095016e-01 -1.40709534e-01 -6.95901811e-01 8.08026373e-01 2.84897745e-01 4.91488963e-01 -8.76662970e-01 -1.03151519e-02 7.43610263e-01 1.72273248e-01 9.15666640e-01 9.43954438e-02 4.94598225e-02 -1.46301949e+00 -4.79272269e-02 -1.18118906e+00 4.60275531e-01 -3.77251863e-01 -1.98263848e+00 5.09164751e-01 -2.64332712e-01 -1.45117700e+00 -2.69840509e-01 -4.06519443e-01 -2.58007854e-01 5.55673063e-01 -1.36761129e+00 -1.07178664e+00 4.46394414e-01 8.82704854e-01 8.51504326e-01 -4.06961411e-01 7.97177732e-01 8.43596935e-01 -3.74572277e-01 9.58063543e-01 6.87851667e-01 3.37721199e-01 3.32933784e-01 -1.04702675e+00 3.98463458e-01 4.83623236e-01 7.79507339e-01 1.27608216e+00 4.65576261e-01 -6.75076663e-01 -1.69205010e+00 -1.18094039e+00 6.25961185e-01 -2.76851803e-01 6.63822651e-01 -4.21359897e-01 -7.68303871e-01 1.00664330e+00 -8.43478963e-02 -1.65863484e-01 1.16334069e+00 8.15344870e-01 -5.16588509e-01 -2.90003896e-01 -6.95390165e-01 6.23456717e-01 1.42077470e+00 -4.67535526e-01 -5.75376391e-01 2.42460892e-01 8.48194718e-01 3.04146041e-03 -8.64419639e-01 5.28297126e-01 9.13033187e-01 -8.31890941e-01 1.23985946e+00 -8.45837355e-01 3.41684222e-01 -1.29053280e-01 -6.06382012e-01 -1.35099065e+00 -7.65658677e-01 -5.43185830e-01 -5.26797473e-01 1.20640421e+00 4.87379283e-01 -4.97041136e-01 8.54131281e-01 6.88888431e-02 1.21801980e-01 -8.61991346e-01 -5.18598914e-01 -7.72722363e-01 -1.98061317e-01 -8.60525131e-01 9.83535409e-01 1.12480962e+00 9.58407763e-03 2.82963216e-01 -1.08992290e+00 1.22256130e-01 2.11171418e-01 6.56332791e-01 4.73416418e-01 -1.06785059e+00 -7.00260460e-01 -4.12570953e-01 6.18556365e-02 -1.65884125e+00 -1.08949527e-01 -9.51308191e-01 -3.68682027e-01 -1.37573040e+00 2.85059810e-01 -6.17736518e-01 -1.21870804e+00 3.29428405e-01 2.24054739e-01 3.99885625e-01 1.82565898e-01 4.10157382e-01 -7.03255236e-01 2.71855265e-01 9.77317750e-01 -5.83699681e-02 -4.87172872e-01 5.17913461e-01 -1.10184443e+00 6.61221266e-01 6.74791276e-01 -4.59962875e-01 -5.05853653e-01 -2.12616578e-01 9.31825161e-01 6.94331825e-02 -1.09284662e-01 -6.12900913e-01 1.09295830e-01 -5.32582588e-02 6.38759792e-01 -9.62825418e-01 9.77313146e-02 -9.99109566e-01 5.13278246e-01 4.10907902e-02 -5.78208566e-01 8.66568014e-02 -1.13114245e-01 9.03130174e-01 -7.22716525e-02 -1.80781260e-01 7.87587985e-02 -3.05935264e-01 -5.72090149e-01 4.85851347e-01 -4.95849669e-01 -3.62413466e-01 3.03052098e-01 -1.46262959e-01 1.08987734e-01 -6.28584266e-01 -8.97225857e-01 -1.49448439e-01 -2.28840575e-01 6.98881507e-01 6.75321937e-01 -1.48433506e+00 -5.65952063e-01 6.41523376e-02 -1.24941818e-01 -4.80398506e-01 5.54810286e-01 4.72713679e-01 2.01915830e-01 5.44677675e-01 1.87788606e-01 -1.38710409e-01 -6.79959118e-01 9.18614924e-01 1.10110939e-01 -7.17629433e-01 -1.04195762e+00 6.56773388e-01 2.14896843e-01 -4.47566390e-01 3.76358330e-01 -4.31889534e-01 -7.65928268e-01 2.11511537e-01 8.06594968e-01 2.73640752e-01 1.80541724e-01 -4.38267112e-01 5.30710854e-02 3.32724035e-01 -6.47595584e-01 3.48523945e-01 1.83208215e+00 -2.74524719e-01 2.43689686e-01 8.66476357e-01 1.04524624e+00 7.15860650e-02 -8.64324749e-01 -6.21566176e-01 -2.36548021e-01 -4.78786767e-01 6.49915561e-02 -1.06094766e+00 -1.40192008e+00 4.71595317e-01 3.99525851e-01 5.97049773e-01 1.15614152e+00 -2.71487415e-01 1.34682822e+00 5.85594893e-01 7.91228890e-01 -1.03789210e+00 -1.24519609e-01 5.41800022e-01 5.66502690e-01 -1.07664442e+00 -1.41510561e-01 2.86232252e-02 -5.04551947e-01 9.95266616e-01 2.87615538e-01 -4.10391271e-01 1.13041103e+00 -3.51225473e-02 -3.66431773e-02 1.45998150e-02 -7.78257966e-01 2.53980756e-02 3.27635050e-01 2.01286465e-01 5.57733715e-01 2.00557962e-01 -2.17654929e-01 1.35111511e+00 -3.72079432e-01 3.33127350e-01 3.25742245e-01 4.72385406e-01 -4.05313596e-02 -9.75436687e-01 5.41228876e-02 9.89943624e-01 -9.24047709e-01 -3.43558401e-01 4.27981645e-01 5.42177737e-01 1.55710978e-02 1.03354657e+00 2.45820824e-02 -6.95608199e-01 2.93044508e-01 -2.13407815e-01 1.82100326e-01 -4.95269686e-01 -9.91663158e-01 2.75418818e-01 -2.35321209e-01 -6.32753551e-01 -5.41530550e-01 -5.36165118e-01 -8.76920819e-01 -5.64901769e-01 -5.09398520e-01 1.73338935e-01 2.95098454e-01 9.25332904e-01 3.83597970e-01 6.53542161e-01 7.97971368e-01 -8.86833131e-01 -5.21815598e-01 -1.09511161e+00 -1.10728228e+00 5.97871423e-01 2.21174002e-01 -5.97035944e-01 -2.09987819e-01 -1.68608621e-01]	[10.118081092834473, 5.632514476776123]
ade098fb-8571-4034-9c5e-e45dc399fdd1	classification-of-ecg-based-on-hybrid	2206.07648	null	https://arxiv.org/abs/2206.07648v1	https://arxiv.org/pdf/2206.07648v1.pdf	Classification of ECG based on Hybrid Features using CNNs for Wearable Applications	Sudden cardiac death and arrhythmia account for a large percentage of all deaths worldwide. Electrocardiography (ECG) is the most widely used screening tool for cardiovascular diseases. Traditionally, ECG signals are classified manually, requiring experience and great skill, while being time-consuming and prone to error. Thus machine learning algorithms have been widely adopted because of their ability to perform complex data analysis. Features derived from the points of interest in ECG - mainly Q, R, and S, are widely used for arrhythmia detection. In this work, we demonstrate improved performance for ECG classification using hybrid features and three different models, building on a 1-D convolutional neural network (CNN) model that we had proposed in the past. An RR interval features based model proposed in this work achieved an accuracy of 98.98%, which is an improvement over the baseline model. To make the model immune to noise, we updated the model using frequency features and achieved good sustained performance in presence of noise with a slightly lower accuracy of 98.69%. Further, another model combining the frequency features and the RR interval features was developed, which achieved a high accuracy of 99% with good sustained performance in noisy environments. Due to its high accuracy and noise immunity, the proposed model which combines multiple hybrid features, is well suited for ambulatory wearable sensing applications.	['Deepu John', 'Barry Cardiff', 'Rajesh C. Panicker', 'Fang Xiang', 'Li Xiaolin']	2022-06-14	null	null	null	null	['arrhythmia-detection', 'ecg-classification', 'electrocardiography-ecg']	['medical', 'medical', 'methodology']	[ 1.05258152e-01 -3.50406855e-01 9.53474566e-02 -1.90683812e-01 -7.62541533e-01 -2.33312994e-01 -1.18582338e-01 5.00918210e-01 -5.30230165e-01 8.01544070e-01 -2.37900957e-01 -1.52883962e-01 -1.81130588e-01 -6.66749775e-01 1.20693957e-02 -6.28992558e-01 -5.07687926e-01 -8.99512768e-02 -6.96523562e-02 -5.32678291e-02 1.63960096e-04 6.75486803e-01 -1.35430467e+00 8.16177577e-02 9.05901134e-01 1.29233134e+00 -2.24799410e-01 7.03804731e-01 4.66332704e-01 3.18132520e-01 -9.61593390e-01 -7.53799155e-02 1.23928366e-02 -6.31257057e-01 -3.06251675e-01 -4.28788453e-01 -9.59707946e-02 -1.52972504e-01 -9.30413380e-02 4.72223967e-01 1.11909521e+00 -2.05636829e-01 3.90306950e-01 -6.77473128e-01 -3.08555186e-01 3.75442088e-01 -5.03373384e-01 5.45274794e-01 3.57587695e-01 -1.52374476e-01 5.69945455e-01 -7.21733689e-01 6.16510697e-02 2.83839613e-01 1.38082445e+00 3.15803200e-01 -1.17881274e+00 -6.96771801e-01 -4.84377146e-01 1.76647991e-01 -1.70549774e+00 -9.66594666e-02 9.93655145e-01 -2.44492918e-01 1.02189016e+00 4.69044417e-01 1.10977983e+00 7.97729552e-01 4.97432739e-01 1.89195767e-01 1.05215704e+00 -4.91068363e-01 1.09622039e-01 9.92732272e-02 1.28089324e-01 3.84338230e-01 3.43530029e-01 1.09797575e-01 -3.02423835e-01 -5.19943833e-01 8.09356391e-01 2.78857321e-01 -3.88274074e-01 9.46370214e-02 -1.23441207e+00 5.97454429e-01 3.21226418e-01 6.60296082e-01 -4.64269608e-01 4.42809984e-03 4.63822156e-01 2.96489090e-01 2.11023718e-01 5.84860146e-01 -4.96760368e-01 -5.12087941e-01 -9.69520092e-01 2.22877890e-01 6.23704731e-01 2.71774530e-01 2.12703824e-01 1.45152554e-01 -3.65388423e-01 9.49144602e-01 1.09978810e-01 4.20252383e-01 6.89192951e-01 -5.71219981e-01 8.53777453e-02 5.76097369e-01 7.19105303e-02 -1.19333017e+00 -9.77453589e-01 -1.07082176e+00 -1.42234612e+00 -1.30279854e-01 3.40975285e-01 -3.59154701e-01 -6.72386825e-01 1.48463523e+00 5.82402796e-02 2.22540766e-01 -5.82710914e-02 9.06311631e-01 8.18360090e-01 2.83541262e-01 7.41387978e-02 -4.23671335e-01 1.27656829e+00 -1.48885280e-01 -7.67423868e-01 2.37704888e-01 5.29984593e-01 -5.64481378e-01 6.03277743e-01 6.70180082e-01 -9.19460535e-01 -8.35985601e-01 -1.31968212e+00 3.61553967e-01 9.36844293e-03 3.61327082e-01 4.64212894e-01 1.03672385e+00 -1.00082040e+00 9.21275496e-01 -8.07021022e-01 -4.73541081e-01 5.11213422e-01 4.85172063e-01 -2.14447334e-01 5.00642359e-01 -1.41563284e+00 6.79090500e-01 7.08487257e-02 3.89559865e-01 -1.95736796e-01 -4.88682568e-01 -6.55171931e-01 7.80291781e-02 -1.87490284e-01 -6.25240982e-01 7.97465682e-01 -5.03716171e-01 -1.57284224e+00 5.81577718e-01 -9.04266760e-02 -5.94710112e-01 5.38364589e-01 -2.98718303e-01 -8.41864645e-01 2.87083447e-01 -1.19371809e-01 -8.53558704e-02 7.29207158e-01 -5.50218284e-01 -3.32236379e-01 -5.60674012e-01 -3.48172486e-01 -1.73332334e-01 -2.01115668e-01 -2.46709958e-02 2.20798329e-02 -7.89263964e-01 3.77554744e-01 -9.52734113e-01 -1.49350569e-01 -2.21329674e-01 -2.23612580e-02 6.67720959e-02 4.57773000e-01 -7.37744689e-01 1.57340705e+00 -2.21215296e+00 -3.02383065e-01 5.57377160e-01 4.32136953e-01 6.57431662e-01 3.72443229e-01 4.13545758e-01 -1.89115420e-01 2.29618683e-01 -3.67810905e-01 1.03219301e-01 -5.54845512e-01 -9.31819901e-02 1.67128578e-01 4.68988031e-01 3.74339730e-01 9.23305392e-01 -6.26222432e-01 -2.01168269e-01 3.65780145e-01 7.34356999e-01 -1.46889836e-01 1.36044860e-01 6.84616506e-01 7.29640424e-01 -3.45261246e-01 4.37912226e-01 4.75563794e-01 -2.37872243e-01 1.59783199e-01 -2.71328777e-01 4.36114818e-02 -7.09780976e-02 -1.19252145e+00 1.55784273e+00 -2.57427424e-01 3.99235249e-01 -4.33371037e-01 -1.16581643e+00 1.51138675e+00 7.63079762e-01 8.52280021e-01 -6.66197121e-01 1.92078456e-01 3.97600353e-01 3.91307920e-01 -7.14587033e-01 -1.86878577e-01 -1.64348334e-01 6.10146113e-02 1.66386813e-01 -2.58941263e-01 -4.98015946e-03 -2.99477249e-01 -1.75420716e-01 1.19049633e+00 4.27236520e-02 6.48025870e-01 -3.97247195e-01 6.79462373e-01 -4.15462762e-01 6.86576545e-01 8.83964956e-01 -4.59640026e-01 1.00534475e+00 2.62495309e-01 -1.01482511e+00 -5.67740500e-01 -8.13206613e-01 -5.44063330e-01 6.88385516e-02 -3.55525948e-02 -4.93859738e-01 -4.32933271e-01 -2.81164408e-01 -7.12102950e-02 -3.41541208e-02 -5.46901703e-01 -2.48378083e-01 -7.20594704e-01 -1.18555701e+00 1.06739414e+00 7.50714064e-01 7.03294098e-01 -8.33989263e-01 -1.08674824e+00 6.30245864e-01 -2.63777286e-01 -8.69326413e-01 1.32063748e-02 2.05576673e-01 -1.12019062e+00 -1.03816628e+00 -9.60512280e-01 -4.27422285e-01 7.82893375e-02 -1.74309999e-01 9.15641069e-01 2.62728602e-01 -7.66219258e-01 3.97065021e-02 -3.67507726e-01 -7.50058949e-01 -1.20461926e-01 2.20897589e-02 1.80179253e-01 1.88811526e-01 3.11057001e-01 -7.73123205e-01 -9.57806170e-01 1.17017642e-01 -5.69106281e-01 -4.49560106e-01 5.27267396e-01 9.72858250e-01 4.78259116e-01 -1.81168988e-01 1.28950799e+00 -7.19005346e-01 7.45352805e-01 -2.23252758e-01 2.71581728e-02 -1.25582382e-01 -7.57563233e-01 -4.14330244e-01 7.54542708e-01 -2.97070146e-01 -4.73816723e-01 2.75757369e-02 -5.98431408e-01 -5.13786776e-03 -3.62190008e-01 5.50198853e-01 1.00023888e-01 -7.21374974e-02 8.46420586e-01 9.73665640e-02 2.66562998e-01 -3.24726194e-01 -5.09958148e-01 8.24332237e-01 3.67947966e-01 -3.19973558e-01 4.10523236e-01 2.32485533e-01 1.30088687e-01 -1.06441402e+00 -5.62058866e-01 -6.16840124e-01 -6.42224610e-01 -1.50613025e-01 7.75205731e-01 -9.34590697e-01 -5.85705459e-01 6.77336574e-01 -9.65499043e-01 2.08733618e-01 -2.48323634e-01 8.22498441e-01 -2.29736894e-01 5.05173981e-01 -4.01106983e-01 -9.85166132e-01 -7.62479961e-01 -6.48328781e-01 7.75320172e-01 2.10070759e-01 -5.76452732e-01 -7.41799414e-01 1.34023041e-01 -1.33974791e-01 9.11002576e-01 9.19103682e-01 8.11280310e-01 -6.49648309e-01 1.91516206e-01 -5.12264132e-01 -5.60835376e-02 3.90728831e-01 5.03876805e-01 -1.59586713e-01 -1.14208913e+00 -4.52281922e-01 2.96166658e-01 1.19091049e-02 6.43124998e-01 4.88983512e-01 1.24422681e+00 2.99209595e-01 -3.35670441e-01 6.11323118e-01 1.33906019e+00 5.85940182e-01 9.17216778e-01 2.00962976e-01 3.76913399e-01 -7.40031451e-02 2.26225957e-01 5.93871713e-01 6.24612793e-02 6.12113774e-01 8.84551704e-02 -6.06143773e-01 1.57065868e-01 4.22259629e-01 -1.73283890e-01 8.03013563e-01 -7.06748843e-01 1.51374280e-01 -9.49253261e-01 3.90107483e-01 -1.75751901e+00 -9.02902603e-01 -3.30191433e-01 2.34054399e+00 6.25065267e-01 1.82038486e-01 3.57169122e-01 8.36662054e-01 5.20556271e-01 -1.78666711e-01 -4.92806822e-01 -5.70762455e-01 -1.59198612e-01 7.51851439e-01 6.84131905e-02 -1.69792235e-01 -1.13385820e+00 -1.28099054e-01 6.44046402e+00 1.85808107e-01 -1.43760705e+00 -4.81095649e-02 7.08433151e-01 2.21155897e-01 4.17943060e-01 -5.15140533e-01 -3.34721506e-01 4.94797349e-01 1.08286369e+00 2.70405207e-02 -1.58589825e-01 4.84938890e-01 4.20008451e-01 -1.36898151e-02 -8.44700575e-01 1.40746546e+00 6.29507452e-02 -1.07031000e+00 -4.06849861e-01 -1.76203430e-01 3.19069386e-01 -2.28976101e-01 -2.11701915e-01 1.24975950e-01 -7.91095853e-01 -1.06910348e+00 2.57916272e-01 6.77488387e-01 1.03433836e+00 -8.56655777e-01 1.30321336e+00 1.84451476e-01 -1.21977222e+00 -2.16491431e-01 -7.45520666e-02 -4.58806157e-01 3.02560460e-02 1.02711940e+00 -5.13333023e-01 7.21426487e-01 9.65363503e-01 8.93142223e-01 -4.76236820e-01 1.23971534e+00 1.54477894e-01 8.91015232e-01 -3.80517960e-01 6.33499473e-02 -3.13889503e-01 1.40641272e-01 4.58939046e-01 1.18386328e+00 5.85692883e-01 1.79830223e-01 2.64443249e-01 5.87254763e-01 3.48410428e-01 2.41823271e-01 -6.22770727e-01 4.60676044e-01 3.43510330e-01 1.23094726e+00 -5.99835813e-01 -2.88927078e-01 -2.84106493e-01 5.51852465e-01 -5.04534645e-03 2.12859109e-01 -1.02583492e+00 -9.79739964e-01 3.19398463e-01 2.70915538e-01 -7.40320757e-02 -1.45320535e-01 -6.11092269e-01 -9.40783143e-01 3.08262527e-01 -7.86380708e-01 2.90892452e-01 -1.60127535e-01 -1.05806148e+00 9.45568144e-01 -3.27875555e-01 -1.56288469e+00 -2.09385127e-01 -3.03744763e-01 -6.69059694e-01 1.24742961e+00 -1.24167192e+00 -5.63643396e-01 -5.30270994e-01 4.84115213e-01 1.13817737e-01 -4.33114022e-02 1.34348774e+00 5.88800907e-01 -5.62010825e-01 7.84916878e-01 -1.23843580e-01 2.17078298e-01 7.13661015e-01 -1.17662084e+00 2.67969340e-01 5.40416062e-01 -5.28308116e-02 8.50410461e-01 2.93807179e-01 -4.43367302e-01 -9.48167682e-01 -1.07359588e+00 9.56599534e-01 -1.57204211e-01 7.58142397e-02 5.73749430e-02 -9.67821062e-01 -2.27105916e-02 -1.25086859e-01 7.70064592e-02 1.03740597e+00 2.78819762e-02 9.60071236e-02 -5.04960120e-01 -1.14070177e+00 1.08698227e-01 6.48835599e-01 -3.40516329e-01 -5.46707749e-01 -1.23015732e-01 -1.32924998e-02 -3.89818460e-01 -1.26105976e+00 8.88759017e-01 1.10460520e+00 -1.11555183e+00 9.47014749e-01 -1.75893128e-01 -7.97748752e-03 -2.16163054e-01 2.96196669e-01 -1.21539581e+00 -6.10780180e-01 -7.06034422e-01 -1.48952454e-01 9.88692403e-01 4.55958605e-01 -9.55415666e-01 4.00206745e-01 2.98298895e-01 -1.61626411e-03 -1.06819749e+00 -9.23030019e-01 -6.57319427e-01 -3.53244692e-01 -3.77321303e-01 4.88094896e-01 7.25814581e-01 7.31198117e-02 2.59733707e-01 -4.20889944e-01 -1.32222716e-02 3.97647768e-01 5.79888150e-02 3.64248902e-01 -1.61830544e+00 -2.47794122e-01 -2.10518688e-01 -7.55208254e-01 -3.17662477e-01 -6.03886425e-01 -7.00586438e-01 -3.04425716e-01 -1.33568454e+00 8.74223486e-02 -5.72433352e-01 -9.42526996e-01 3.50836605e-01 -4.03208107e-01 1.03513849e+00 -1.61454707e-01 2.16328457e-01 -7.25578144e-02 8.69817659e-02 7.89146245e-01 8.77345260e-03 -6.88917339e-01 3.81283432e-01 -7.31800675e-01 6.91117346e-01 1.14937556e+00 -3.46287012e-01 -1.75864220e-01 -3.28617282e-02 7.50923306e-02 1.55927584e-01 2.48637274e-01 -1.50893295e+00 -1.85034513e-01 5.33248365e-01 1.04093170e+00 -4.10026968e-01 3.01796019e-01 -6.74724221e-01 5.69904566e-01 8.04947078e-01 -1.29463509e-01 2.72688121e-01 2.89835662e-01 3.65778565e-01 -4.16838437e-01 1.81380793e-01 7.04752445e-01 -1.49378821e-01 -8.80546868e-02 1.89765766e-01 -4.74742115e-01 -2.54232198e-01 9.71247256e-01 -5.79283178e-01 1.38468593e-01 -1.87195495e-01 -1.13237929e+00 -2.52052069e-01 -3.41104455e-02 2.77984560e-01 8.39597106e-01 -1.22571290e+00 -8.21885824e-01 5.06696761e-01 2.09054202e-01 -2.56268293e-01 3.35317969e-01 1.31040084e+00 -6.59096062e-01 3.00695837e-01 -3.35859179e-01 -9.59985793e-01 -1.21555090e+00 8.21784362e-02 4.20899719e-01 -1.58088416e-01 -1.04512560e+00 4.46983010e-01 -5.28469622e-01 1.31606132e-01 1.49295747e-01 -5.20182014e-01 -5.49333394e-01 -2.01466940e-02 7.11234272e-01 6.56941414e-01 5.56911230e-01 -3.77691418e-01 -5.85563540e-01 9.33478117e-01 1.73541188e-01 2.76451588e-01 1.21659744e+00 -5.72854318e-02 -2.06216760e-02 5.49126983e-01 9.50294018e-01 5.54122888e-02 -4.63332176e-01 1.03541143e-01 -7.15652555e-02 -2.27447212e-01 -3.19264643e-02 -9.65214849e-01 -1.03194797e+00 1.06559432e+00 1.14624846e+00 4.29279685e-01 1.38764441e+00 -5.57947099e-01 8.32227528e-01 1.59083784e-01 4.53618705e-01 -8.46950352e-01 -2.17400670e-01 9.63830426e-02 5.08352458e-01 -9.99080300e-01 -2.87795607e-02 -1.61345944e-01 -3.95403147e-01 1.39862251e+00 1.83880165e-01 -3.08729172e-01 9.32796657e-01 1.99831396e-01 3.85730833e-01 -1.09786205e-02 -1.69853404e-01 1.27065122e-01 3.23031366e-01 7.59360731e-01 8.00357401e-01 1.05248801e-01 -7.81635523e-01 9.13509369e-01 8.68054703e-02 5.34659207e-01 2.99749523e-01 1.01655269e+00 -1.77572086e-01 -1.02409172e+00 -2.66494215e-01 8.70492697e-01 -1.04297245e+00 9.40327942e-02 -1.53403193e-01 5.57903230e-01 3.60959232e-01 1.17163849e+00 -2.96169132e-01 -4.23497796e-01 4.65020567e-01 2.91498244e-01 3.40557665e-01 -4.17979598e-01 -1.01903439e+00 1.59045026e-01 -4.59774993e-02 -5.43607295e-01 -6.00613654e-01 -4.38904971e-01 -1.11627936e+00 3.31844956e-01 -4.00087208e-01 2.59173691e-01 5.33642471e-01 8.24576974e-01 6.55727565e-01 8.26882243e-01 6.87940657e-01 -5.21627307e-01 -4.42226440e-01 -1.13192201e+00 -8.45149279e-01 4.13026839e-01 4.31803256e-01 -5.06455541e-01 -8.04105550e-02 -1.22875031e-02]	[14.248613357543945, 3.26104998588562]
2ed50e99-0645-42d7-905c-5fe3b191f4ef	scalable-multiagent-driving-policies-for	2103.00058	null	https://arxiv.org/abs/2103.00058v2	https://arxiv.org/pdf/2103.00058v2.pdf	Scalable Multiagent Driving Policies For Reducing Traffic Congestion	Traffic congestion is a major challenge in modern urban settings. The industry-wide development of autonomous and automated vehicles (AVs) motivates the question of how can AVs contribute to congestion reduction. Past research has shown that in small scale mixed traffic scenarios with both AVs and human-driven vehicles, a small fraction of AVs executing a controlled multiagent driving policy can mitigate congestion. In this paper, we scale up existing approaches and develop new multiagent driving policies for AVs in scenarios with greater complexity. We start by showing that a congestion metric used by past research is manipulable in open road network scenarios where vehicles dynamically join and leave the road. We then propose using a different metric that is robust to manipulation and reflects open network traffic efficiency. Next, we propose a modular transfer reinforcement learning approach, and use it to scale up a multiagent driving policy to outperform human-like traffic and existing approaches in a simulated realistic scenario, which is an order of magnitude larger than past scenarios (hundreds instead of tens of vehicles). Additionally, our modular transfer learning approach saves up to 80% of the training time in our experiments, by focusing its data collection on key locations in the network. Finally, we show for the first time a distributed multiagent policy that improves congestion over human-driven traffic. The distributed approach is more realistic and practical, as it relies solely on existing sensing and actuation capabilities, and does not require adding new communication infrastructure.	['Daniel Urieli', 'Peter Stone', 'Harel Yedidsion', 'William Macke', 'Jiaxun Cui']	2021-02-26	null	null	null	null	['transfer-reinforcement-learning']	['methodology']	[-2.72436917e-01 4.22924340e-01 -3.86015475e-01 -1.52030904e-02 -2.91042089e-01 -5.74595213e-01 5.70846379e-01 -1.62739724e-01 -5.60980201e-01 1.22542143e+00 -2.94195592e-01 -8.29995453e-01 -1.98545560e-01 -1.31091130e+00 -7.24303246e-01 -3.58507454e-01 -4.26848590e-01 6.70703173e-01 9.21972394e-01 -9.26286817e-01 -5.52257188e-02 6.23950660e-01 -1.61668777e+00 -5.82236588e-01 9.27690625e-01 5.23993194e-01 1.60449669e-01 1.00541472e+00 2.14350581e-01 7.88635850e-01 -5.24083078e-01 5.97206876e-02 6.14366055e-01 1.57621741e-01 -7.76972055e-01 -1.52410433e-01 3.70615512e-01 -5.32370687e-01 -7.59592891e-01 2.83144772e-01 4.34523463e-01 3.60387325e-01 5.73023021e-01 -2.24664998e+00 -6.51514456e-02 4.20256168e-01 -5.03496647e-01 1.77594095e-01 -7.93267041e-02 7.18754172e-01 8.39515984e-01 -1.44509329e-02 5.50569057e-01 1.34464121e+00 5.50550103e-01 7.08047390e-01 -1.29527080e+00 -9.21173096e-01 2.85028487e-01 3.29081804e-01 -1.06229424e+00 -4.69950080e-01 4.87616777e-01 -3.03713381e-01 9.46194589e-01 1.09536126e-01 6.74907982e-01 8.25953722e-01 1.73398420e-01 6.21042132e-01 8.50328326e-01 -4.91755977e-02 4.44741994e-01 1.22486815e-01 -3.20797801e-01 5.56523263e-01 5.42271435e-01 2.66814560e-01 7.11979792e-02 3.17225195e-02 5.57020307e-01 -3.58965665e-01 3.40659201e-01 -6.57879770e-01 -9.89790261e-01 1.07715404e+00 5.01484036e-01 -1.46582067e-01 -3.67089689e-01 8.26991081e-01 5.04629612e-01 5.14781356e-01 4.50859964e-02 3.66072267e-01 -4.40290213e-01 -4.53760237e-01 -5.59299707e-01 4.65644300e-01 1.03730643e+00 1.12567103e+00 1.28703904e+00 2.80218989e-01 2.71571934e-01 4.34154063e-01 1.26048058e-01 1.05964220e+00 -2.39102885e-01 -1.78163624e+00 5.53141594e-01 2.22278118e-01 3.75671685e-01 -9.38331783e-01 -5.96440315e-01 2.56354630e-01 -5.51982582e-01 8.67251754e-01 3.67919177e-01 -1.01356757e+00 -5.16884029e-01 1.93567598e+00 5.00417888e-01 3.30991149e-01 1.96733624e-01 4.66918439e-01 1.09436000e-02 7.52103090e-01 2.17265621e-01 2.26128325e-01 8.02581906e-01 -1.06986892e+00 -3.72003675e-01 -3.36688817e-01 7.66798735e-01 -3.06580335e-01 6.68764710e-01 -1.39301449e-01 -9.83885705e-01 -3.61458123e-01 -1.09291363e+00 3.44751775e-01 -8.01585197e-01 -7.39584446e-01 6.06901109e-01 8.17165554e-01 -1.41662478e+00 3.97568434e-01 -6.18737757e-01 -7.58408427e-01 5.22389472e-01 5.72279274e-01 -1.37760743e-01 1.46425590e-01 -1.41880965e+00 1.31373632e+00 -1.18322223e-01 -3.50801528e-01 -1.00411737e+00 -1.00487781e+00 -7.54989862e-01 -1.39892772e-01 6.41163290e-01 -9.08186018e-01 1.38867855e+00 -5.08938193e-01 -1.70663249e+00 8.41578171e-02 4.20774035e-02 -7.92429566e-01 6.82673335e-01 1.09722190e-01 -3.35270584e-01 2.75911987e-01 3.54907751e-01 1.30178893e+00 5.22384882e-01 -1.43065560e+00 -1.13458824e+00 2.85697758e-01 6.75599575e-01 4.40884456e-02 -2.56825745e-01 -4.52946633e-01 1.87889963e-01 1.03961810e-01 -9.77981508e-01 -1.30450773e+00 -6.48509681e-01 4.00899410e-01 -2.03355029e-02 -3.65363032e-01 1.43937540e+00 2.59555638e-01 7.86400259e-01 -1.67035687e+00 -2.12121248e-01 5.29026210e-01 4.64560658e-01 5.23106158e-01 -4.29191202e-01 9.34899211e-01 6.11455083e-01 3.49056780e-01 2.74801608e-02 -2.77059972e-02 3.04246396e-01 7.35268235e-01 -2.28341162e-01 2.48276353e-01 3.38936299e-02 1.05375099e+00 -1.08970809e+00 -5.65162539e-01 6.88838303e-01 2.45726153e-01 -6.04449809e-01 -1.19207442e-01 -2.07917020e-01 4.00657952e-01 -6.43490136e-01 2.36266047e-01 7.01284051e-01 1.41721815e-01 -4.04280052e-02 2.65787035e-01 -5.31591952e-01 -1.54430404e-01 -1.10823596e+00 9.29241300e-01 -8.15663397e-01 1.10375249e+00 4.01917547e-01 -1.13512647e+00 6.67582989e-01 1.37835264e-01 9.30933416e-01 -9.83865678e-01 4.69497591e-02 2.05929279e-01 2.09041107e-02 -6.69801891e-01 4.94496047e-01 2.06640303e-01 -1.91027567e-01 5.59652686e-01 -4.82484877e-01 -3.54308724e-01 5.63691437e-01 3.31614524e-01 1.53498650e+00 -2.45148897e-01 -1.49766982e-01 -2.09862635e-01 4.08358932e-01 3.48141134e-01 4.42598313e-01 8.68250012e-01 -8.62909853e-01 -4.82271791e-01 4.88050342e-01 -3.08424801e-01 -1.23822427e+00 -1.13667917e+00 2.52258092e-01 1.23350382e+00 5.79267263e-01 -1.26715181e-02 -5.59714019e-01 -4.38361615e-01 6.49211466e-01 6.56237006e-01 -5.50199628e-01 -1.75088137e-01 -9.05660868e-01 -8.92489627e-02 8.22563291e-01 5.15322924e-01 6.11297071e-01 -8.80885124e-01 -7.63463974e-01 3.88695836e-01 -5.47217093e-02 -1.32855737e+00 -3.11663747e-01 -2.61725366e-01 -3.68603766e-01 -1.24099660e+00 -4.37826097e-01 -5.85628986e-01 2.76196331e-01 8.26499879e-01 7.33159065e-01 1.00123711e-01 -6.99255094e-02 7.47281730e-01 -8.79778937e-02 -5.20422518e-01 -6.20245159e-01 5.17859042e-01 1.70103535e-01 -2.84679264e-01 1.91261709e-01 -9.18891966e-01 -7.94824958e-01 5.53318918e-01 -4.63804007e-01 -9.11351740e-02 6.26743555e-01 3.35322320e-01 -1.33880302e-01 -5.27679808e-02 1.10043597e+00 -5.27764976e-01 6.05729043e-01 -7.84923077e-01 -7.07947731e-01 -2.52441257e-01 -5.47722340e-01 -6.23277910e-02 6.95431352e-01 -3.07616264e-01 -6.84702516e-01 -7.37915710e-02 1.16130285e-01 -1.80091828e-01 -2.91897088e-01 -6.50464892e-02 2.49109834e-01 -5.34308732e-01 6.22731686e-01 -3.43935877e-01 6.08033299e-01 3.22749674e-01 6.75794661e-01 7.25849509e-01 -2.05781497e-02 -5.95450044e-01 1.30277514e+00 7.92781174e-01 5.68791807e-01 -1.05204260e+00 -1.66725934e-01 -2.07427412e-01 -4.68760878e-01 -6.61730051e-01 5.65967739e-01 -7.86821008e-01 -1.65732062e+00 9.84171256e-02 -8.31266105e-01 -1.12736976e+00 -4.03974265e-01 2.84960687e-01 -9.75523889e-01 6.40691370e-02 -4.08428669e-01 -6.51976526e-01 6.80596828e-02 -1.15552950e+00 6.81403756e-01 2.66491979e-01 -6.28938526e-02 -1.12316144e+00 3.65554094e-01 2.73685217e-01 1.12039578e+00 3.31698179e-01 5.83334744e-01 -1.72253713e-01 -8.08463216e-01 1.27333626e-01 -3.82425457e-01 -5.14022745e-02 1.34703666e-02 1.86400339e-01 -5.78991592e-01 -3.97125512e-01 -1.23099327e+00 -2.93770492e-01 5.95516562e-01 3.00127089e-01 6.68548167e-01 -1.37378544e-01 -7.64006019e-01 -1.37938827e-01 1.28259218e+00 1.86231583e-01 6.75983310e-01 6.77835822e-01 5.73328435e-01 7.16559887e-01 7.02322245e-01 2.31848136e-01 1.12264550e+00 5.34404874e-01 6.80368006e-01 -4.73515123e-01 -2.70702839e-01 -2.02859342e-01 3.94364744e-01 2.61691839e-01 -2.19073910e-02 -2.87578315e-01 -8.61216366e-01 7.46467233e-01 -2.19642282e+00 -1.22540438e+00 -6.10873699e-02 1.88633668e+00 4.40034926e-01 2.86788642e-01 6.65664375e-01 -4.87299403e-03 7.20840275e-01 -3.17004547e-02 -7.61746883e-01 -8.15756977e-01 2.95832366e-01 1.47597551e-01 1.29203141e+00 8.46341193e-01 -9.68931437e-01 1.12600970e+00 7.29685831e+00 7.41816580e-01 -9.28065062e-01 -6.94511458e-02 1.63508341e-01 4.87940982e-02 -1.73063233e-01 1.04170986e-01 -7.02665865e-01 4.24485624e-01 1.45539320e+00 -5.18089294e-01 5.26607811e-01 7.55266130e-01 1.01900315e+00 -3.55682462e-01 -9.38676119e-01 2.17085376e-01 -5.56902289e-01 -1.30979824e+00 -1.41506940e-01 4.90413725e-01 7.37649441e-01 4.18814480e-01 -1.06102273e-01 7.35009074e-01 1.03651595e+00 -8.95665586e-01 3.87123972e-01 3.11431348e-01 4.65043038e-01 -9.58579302e-01 3.72945577e-01 3.14614058e-01 -1.55959117e+00 -4.84406978e-01 -1.72822550e-01 -2.22231030e-01 5.33556342e-01 4.90748994e-02 -7.09090114e-01 1.02113344e-01 3.89222533e-01 5.42235732e-01 -3.07629526e-01 1.04221654e+00 9.59534720e-02 5.79126716e-01 -5.45712650e-01 -4.01800960e-01 6.61683142e-01 -2.42441613e-02 7.90382445e-01 1.00642121e+00 -7.62680545e-02 -3.76440078e-01 5.97211778e-01 6.56027734e-01 2.15406995e-02 -2.94308126e-01 -1.10231698e+00 6.27934813e-01 8.64095509e-01 1.38203573e+00 -2.01297745e-01 -3.23181868e-01 -5.03008604e-01 2.00837851e-01 2.67963082e-01 5.69402695e-01 -1.20800149e+00 -9.00654674e-01 1.24435270e+00 3.19899231e-01 5.19604802e-01 -5.06471574e-01 1.08219340e-01 -4.18126851e-01 -2.89136797e-01 -3.14064771e-01 -1.23092212e-01 -3.99753600e-01 -9.19675648e-01 4.13698889e-02 2.46364057e-01 -1.29374659e+00 -3.49578291e-01 -3.59065771e-01 -8.18042457e-01 2.54928499e-01 -2.19759440e+00 -1.18550551e+00 -2.30794191e-01 6.28616154e-01 3.46300811e-01 -1.22837000e-01 4.14155871e-01 6.27700746e-01 -5.53463578e-01 4.20496434e-01 -5.02126664e-02 -7.62272105e-02 6.55590355e-01 -1.06040609e+00 3.59201461e-01 4.78612810e-01 -7.95081317e-01 3.89895327e-02 8.86057675e-01 -3.97229642e-01 -1.47338164e+00 -1.37492394e+00 4.53467160e-01 -1.88977420e-01 1.06220508e+00 -3.18369001e-01 -2.64180928e-01 6.64520562e-01 5.44662178e-01 9.68648493e-02 2.12251127e-01 -1.52417079e-01 -3.53198200e-02 -5.94565332e-01 -1.22181785e+00 9.25388396e-01 1.05215752e+00 -2.65222136e-02 -3.63469357e-03 2.77279943e-01 8.43268692e-01 -1.50587140e-02 -7.29704499e-01 1.96197972e-01 5.06186187e-01 -4.92085129e-01 7.13268697e-01 -6.03946447e-01 -1.48252413e-01 -5.56626797e-01 -4.36528325e-02 -1.55404639e+00 -1.99956194e-01 -1.01743948e+00 -3.99542749e-02 8.42300951e-01 5.33837616e-01 -1.11120915e+00 8.86160910e-01 6.98310316e-01 -1.34175479e-01 -5.13024330e-01 -1.15578973e+00 -9.67754364e-01 6.07235789e-01 -2.56811112e-01 4.83057022e-01 4.30825174e-01 9.62557942e-02 2.58955568e-01 -2.61019826e-01 1.75245106e-01 7.31826961e-01 -2.18911856e-01 1.42981410e+00 -1.20042169e+00 3.14649403e-01 -4.91251975e-01 -5.90217590e-01 -1.17606115e+00 4.65535253e-01 -5.78219414e-01 5.68999089e-02 -1.64767992e+00 -3.31041425e-01 -8.76897633e-01 1.51096925e-01 7.20854461e-01 4.69636887e-01 1.37785435e-01 2.08890617e-01 -1.25781149e-01 -8.90771866e-01 6.45505369e-01 1.28790820e+00 -2.58521318e-01 -2.91090161e-01 4.66548800e-02 -6.11090779e-01 4.76302415e-01 1.36276567e+00 -1.90807894e-01 -6.34131193e-01 -2.42031679e-01 -5.28077334e-02 -7.49647915e-02 6.07469797e-01 -1.11724687e+00 5.74352086e-01 -6.64140463e-01 -3.43546987e-01 -3.74643832e-01 3.45289171e-01 -1.09876180e+00 -2.81992704e-01 7.70656109e-01 -7.50254169e-02 -2.09207106e-02 4.41015512e-01 7.46107101e-01 3.14380467e-01 3.96490097e-01 8.89355242e-01 1.54056594e-01 -9.00087118e-01 5.01457632e-01 -1.28225780e+00 2.15253055e-01 1.83688474e+00 -4.04411316e-01 -7.78822780e-01 -6.60687625e-01 -3.84277493e-01 1.12891459e+00 1.24475695e-01 5.38582802e-01 2.34941661e-01 -1.16339087e+00 -6.50651932e-01 -1.85299143e-01 -6.19984791e-02 -4.04159278e-01 5.83208241e-02 8.03535700e-01 -5.21789014e-01 4.38218683e-01 -4.10250366e-01 -5.26758373e-01 -9.36323047e-01 4.27076697e-01 5.23898005e-01 8.77945274e-02 -5.23222864e-01 -7.58916587e-02 -3.68143380e-01 -6.17108345e-01 -3.47470539e-03 -1.05383741e-02 6.87549040e-02 -2.73778826e-01 1.63193509e-01 1.01692128e+00 -3.69221598e-01 -7.25279808e-01 -2.00978249e-01 6.88680112e-01 1.31457359e-01 -2.30322123e-01 1.18153524e+00 -6.52959645e-01 3.69987011e-01 7.36794770e-02 9.97820854e-01 -1.20073296e-01 -1.48607457e+00 7.52315596e-02 -4.13741380e-01 -2.64652342e-01 -5.97135015e-02 -3.37988108e-01 -1.14679158e+00 5.85822701e-01 3.57157260e-01 4.89706397e-01 5.07434309e-01 -1.17663726e-01 1.19557762e+00 8.23089302e-01 5.23723960e-01 -1.50509405e+00 -2.44857892e-02 5.42898297e-01 4.61753726e-01 -1.34973025e+00 -3.00862700e-01 -4.67352033e-01 -5.34154832e-01 9.22538459e-01 1.02072191e+00 -4.20225173e-01 9.03100848e-01 5.42263150e-01 6.22996055e-02 -4.43094335e-02 -1.07769835e+00 -6.76116407e-01 -6.81954026e-01 1.14390278e+00 -4.33012962e-01 3.48465681e-01 5.80358654e-02 -4.67709213e-01 -1.78700939e-01 1.78650972e-02 9.58649933e-01 1.11161852e+00 -1.08801782e+00 -1.14847302e+00 -2.48309970e-02 3.75130624e-01 4.02056098e-01 5.51043153e-01 3.70276310e-02 1.25512338e+00 3.61505263e-02 1.53289902e+00 2.07598552e-01 -3.62742096e-01 6.31557107e-01 -5.41805029e-01 6.06208071e-02 -3.83676924e-02 -5.80479950e-02 -6.18516982e-01 2.91148424e-01 -8.09981704e-01 -4.65496480e-01 -6.65784180e-01 -1.53135765e+00 -1.20180893e+00 -8.67467746e-02 -2.13567354e-02 5.46012282e-01 8.48689139e-01 7.11795509e-01 4.38791543e-01 1.20854700e+00 -1.02713084e+00 -2.92928934e-01 -4.12663460e-01 -2.31307358e-01 1.71377715e-02 5.74638128e-01 -9.31599796e-01 -4.61762190e-01 -4.53543276e-01]	[5.270834445953369, 1.4127401113510132]
a01bca1f-21ed-4e6a-84c6-e636d4a8a425	a-multi-task-deep-learning-approach-for	2303.11100	null	https://arxiv.org/abs/2303.11100v1	https://arxiv.org/pdf/2303.11100v1.pdf	A Multi-Task Deep Learning Approach for Sensor-based Human Activity Recognition and Segmentation	Sensor-based human activity segmentation and recognition are two important and challenging problems in many real-world applications and they have drawn increasing attention from the deep learning community in recent years. Most of the existing deep learning works were designed based on pre-segmented sensor streams and they have treated activity segmentation and recognition as two separate tasks. In practice, performing data stream segmentation is very challenging. We believe that both activity segmentation and recognition may convey unique information which can complement each other to improve the performance of the two tasks. In this paper, we firstly proposes a new multitask deep neural network to solve the two tasks simultaneously. The proposed neural network adopts selective convolution and features multiscale windows to segment activities of long or short time durations. First, multiple windows of different scales are generated to center on each unit of the feature sequence. Then, the model is trained to predict, for each window, the activity class and the offset to the true activity boundaries. Finally, overlapping windows are filtered out by non-maximum suppression, and adjacent windows of the same activity are concatenated to complete the segmentation task. Extensive experiments were conducted on eight popular benchmarking datasets, and the results show that our proposed method outperforms the state-of-the-art methods both for activity recognition and segmentation.	['Yaping Wan', 'Huansheng Ning', 'Liming Chen', 'Jinqiang Wang', 'Tao Zhu', 'Furong Duan']	2023-03-20	null	null	null	null	['human-activity-recognition', 'human-activity-recognition']	['computer-vision', 'time-series']	[ 6.63083375e-01 -4.78060573e-01 -3.73599291e-01 -3.23711574e-01 -5.22990406e-01 -1.51246414e-01 4.13722217e-01 2.02210382e-01 -6.49577200e-01 6.63546503e-01 2.74076223e-01 1.22146174e-01 1.16820103e-02 -5.71588695e-01 -4.95717436e-01 -9.50152218e-01 -1.90448910e-01 -4.41078618e-02 5.48542261e-01 4.23702359e-01 2.44743735e-01 2.29957640e-01 -1.47251296e+00 4.28105444e-01 8.50527942e-01 1.55084252e+00 8.10269490e-02 4.76703376e-01 -2.08486795e-01 9.02040243e-01 -9.66347933e-01 4.35592651e-01 1.13418154e-01 -6.44026995e-01 -6.11300766e-01 3.80971611e-01 -1.97232753e-01 -2.73349345e-01 -2.79537380e-01 6.93140984e-01 4.94050384e-01 4.25607681e-01 4.08574045e-01 -1.08636785e+00 -1.94242317e-02 5.02022028e-01 -7.59944618e-01 6.61894917e-01 3.02278906e-01 -2.88051292e-02 6.02591753e-01 -6.49325788e-01 -1.03288488e-02 6.54708803e-01 4.92778778e-01 1.72348142e-01 -9.12561238e-01 -6.36476159e-01 4.13753450e-01 3.55698228e-01 -1.35980356e+00 -2.58757651e-01 9.37392235e-01 -4.92146790e-01 9.64101315e-01 1.38698712e-01 1.00805986e+00 1.01138401e+00 5.48377559e-02 1.18023133e+00 8.08275878e-01 -3.02132946e-02 4.09351379e-01 -3.20612341e-01 1.40845895e-01 3.01965743e-01 9.78994444e-02 -4.30773318e-01 -5.77079117e-01 9.98549089e-02 7.79009640e-01 4.12560344e-01 -4.53200668e-01 -9.74941850e-02 -1.49198890e+00 6.20209515e-01 2.25017339e-01 6.40088558e-01 -5.84313393e-01 1.11897632e-01 4.75179344e-01 -1.25344560e-01 5.13797224e-01 -2.63495035e-02 -4.20281857e-01 -5.39750338e-01 -1.12151122e+00 1.39159262e-01 6.95386708e-01 4.48901832e-01 5.54504037e-01 -9.34264138e-02 -4.14163172e-01 8.50810230e-01 2.24206746e-01 1.01245217e-01 8.38525295e-01 -6.46198392e-01 5.36619663e-01 8.02566409e-01 1.25277579e-01 -7.59035766e-01 -5.11632264e-01 -4.30689305e-01 -8.67549598e-01 -2.38802359e-01 6.14733636e-01 -4.26913351e-01 -8.97406042e-01 1.30925882e+00 3.26880932e-01 7.09047616e-01 -4.25592018e-03 9.72194374e-01 6.63200974e-01 9.86066580e-01 2.02817887e-01 -5.04542947e-01 1.28045058e+00 -1.11691558e+00 -8.77985299e-01 -4.60008353e-01 4.44232047e-01 -4.71380442e-01 8.47041368e-01 5.38154960e-01 -9.74785924e-01 -7.63008952e-01 -1.41637456e+00 2.19523281e-01 -2.38599539e-01 1.26937181e-01 5.67947090e-01 5.01528740e-01 -3.22618812e-01 7.02095687e-01 -1.24697971e+00 -1.30276024e-01 7.67078757e-01 2.13932216e-01 1.11778796e-01 2.99394399e-01 -1.17003858e+00 2.70478845e-01 5.41850865e-01 3.35263014e-01 -8.46854866e-01 -2.10164964e-01 -6.91186309e-01 1.67802259e-01 5.93898237e-01 -9.40257460e-02 1.23300946e+00 -1.13424098e+00 -1.34646749e+00 4.66152728e-01 -2.79253930e-01 -5.61295271e-01 5.48266470e-01 -3.75209242e-01 -4.18828368e-01 7.00673610e-02 6.15917705e-02 2.96871483e-01 7.96649754e-01 -7.79572248e-01 -1.00466144e+00 -4.14714992e-01 -2.84648538e-01 4.02158856e-01 -4.57971722e-01 -4.27984633e-03 -6.46808803e-01 -7.86586225e-01 1.27480909e-01 -5.63327312e-01 -1.47991806e-01 -2.96351075e-01 -2.52530515e-01 -4.47810829e-01 1.03877187e+00 -7.47219980e-01 1.49484181e+00 -2.17985415e+00 -1.24695553e-02 6.97344020e-02 1.23125635e-01 4.13630545e-01 6.47169203e-02 1.35525391e-01 4.00962122e-02 -2.43887275e-01 -4.31619465e-01 -2.87316561e-01 -2.57440329e-01 2.23962754e-01 3.07438206e-02 6.14207685e-01 5.07360846e-02 6.97848737e-01 -7.47667730e-01 -6.24540448e-01 2.63405085e-01 3.98404837e-01 1.21806636e-02 3.95885587e-01 -1.99937642e-01 6.57383502e-01 -5.00121891e-01 6.63410783e-01 4.44832891e-01 -2.49194250e-01 7.81175941e-02 -1.45855779e-02 -1.99643865e-01 2.52305329e-01 -1.53055704e+00 1.73177636e+00 -1.44970745e-01 5.52383304e-01 -2.19950870e-01 -1.59857464e+00 8.41457903e-01 2.99672216e-01 1.25010872e+00 -8.50645661e-01 3.01592737e-01 6.94559067e-02 2.03438494e-02 -6.28985703e-01 2.11949036e-01 1.70799524e-01 -1.93903327e-01 6.24795198e-01 -1.45265713e-01 3.72681856e-01 4.66688216e-01 -4.37166274e-01 1.09662652e+00 2.55368620e-01 2.81948894e-01 1.63647503e-01 7.44765103e-01 -3.58485520e-01 9.71299589e-01 6.27685606e-01 -5.03823161e-01 5.54651678e-01 3.84116918e-01 -5.25685132e-01 -5.25041103e-01 -1.01288497e+00 3.97868007e-02 1.22642732e+00 4.38539773e-01 -1.60773918e-01 -9.55117166e-01 -7.40886331e-01 -2.88341880e-01 2.24346280e-01 -5.87537706e-01 1.81367137e-02 -8.75546694e-01 -9.94417965e-01 6.17951632e-01 8.86249483e-01 9.92604792e-01 -1.48640084e+00 -1.14250350e+00 6.32467985e-01 -4.97526795e-01 -1.25184858e+00 -5.86910546e-01 3.97454083e-01 -9.67490077e-01 -1.27115738e+00 -9.14837301e-01 -8.43756437e-01 3.11598510e-01 3.16433847e-01 6.97901666e-01 2.92644044e-03 -1.87557638e-01 -1.08071685e-01 -4.30686891e-01 -5.02209306e-01 8.77377838e-02 2.98216194e-01 -3.12620312e-01 4.64975268e-01 5.76362133e-01 -6.32222712e-01 -8.02300453e-01 5.17475784e-01 -1.07640254e+00 -7.66812190e-02 6.14166975e-01 5.35661578e-01 7.97926545e-01 8.68230984e-02 5.53295135e-01 -4.92713779e-01 8.60663176e-01 -4.17684406e-01 -4.10333842e-01 1.55616730e-01 -3.27958256e-01 -1.09367773e-01 4.79685158e-01 -7.52539694e-01 -9.16176438e-01 2.82690048e-01 -1.04793340e-01 -2.29690433e-01 -3.91888201e-01 6.01281583e-01 -4.31381375e-01 3.72862130e-01 3.30585897e-01 6.11938655e-01 -2.02837303e-01 -4.93070215e-01 -8.54576305e-02 7.16814458e-01 4.76460218e-01 -2.31497139e-01 1.52770296e-01 5.04577637e-01 -2.71735311e-01 -8.67327750e-01 -7.54274487e-01 -6.10894442e-01 -6.67372465e-01 -3.62437487e-01 1.13077724e+00 -7.37174213e-01 -5.14124691e-01 1.20510435e+00 -1.11885428e+00 -4.27811265e-01 -3.11079741e-01 4.71831471e-01 -3.40289354e-01 4.40205127e-01 -4.35854554e-01 -8.66415858e-01 -4.70077395e-01 -1.09034824e+00 9.53371942e-01 6.54111803e-01 -2.43705869e-01 -8.20261300e-01 2.22236872e-01 4.71478343e-01 2.23114759e-01 3.71223927e-01 3.89309049e-01 -9.64025140e-01 -3.04479629e-01 -3.05498213e-01 -8.22917745e-02 4.20013964e-01 4.78139043e-01 -2.37115651e-01 -9.04747307e-01 -4.87972870e-02 2.56170481e-01 -1.46839947e-01 9.11337137e-01 5.59289694e-01 1.49148822e+00 -7.84872919e-02 -5.14124453e-01 4.69891578e-01 9.29009438e-01 7.87205219e-01 7.09249854e-01 3.01321447e-01 6.47128582e-01 3.28531772e-01 6.23502195e-01 7.38625169e-01 1.35415748e-01 6.19655311e-01 2.94779509e-01 -2.56746501e-01 3.38869303e-01 1.20358199e-01 3.10935050e-01 5.90349376e-01 -2.26943880e-01 -1.99547619e-01 -7.36878872e-01 6.48493707e-01 -2.06691670e+00 -1.11126316e+00 -1.10490330e-01 2.19238067e+00 7.16996789e-01 3.60954374e-01 3.73362780e-01 7.00574100e-01 6.98759019e-01 7.02946424e-01 -8.50857854e-01 1.62647329e-02 2.60831439e-03 3.77494156e-01 2.97600448e-01 -1.63573418e-02 -1.59048390e+00 4.14040208e-01 5.67177534e+00 7.60657787e-01 -1.12720156e+00 1.38301730e-01 6.61168873e-01 -3.64922464e-01 4.20171350e-01 -3.02110285e-01 -5.52818298e-01 7.72300839e-01 8.28631282e-01 2.42700838e-02 2.55845875e-01 6.17616057e-01 5.23140728e-01 -6.36291981e-01 -1.01096714e+00 1.09853733e+00 3.57829742e-02 -1.02367449e+00 -3.93724233e-01 -1.33420387e-02 6.50804937e-01 -1.27393439e-01 -3.52319032e-01 3.00033927e-01 -2.92508304e-01 -9.34342265e-01 5.60311437e-01 4.80320692e-01 1.31186813e-01 -7.99807370e-01 8.03048730e-01 5.48320293e-01 -1.71113944e+00 -2.58077979e-01 1.76849663e-01 -2.44051799e-01 3.72418433e-01 7.91998446e-01 -3.52978081e-01 3.65916908e-01 8.24524522e-01 8.53700161e-01 -4.27736074e-01 1.18037689e+00 -8.16884562e-02 7.17966020e-01 -4.07026976e-01 -1.63935632e-01 2.59722650e-01 -2.29992226e-01 2.50453115e-01 1.30156183e+00 2.96911985e-01 -5.84809072e-02 5.07613301e-01 5.74533999e-01 -7.80814812e-02 -2.41573621e-02 -6.16642982e-02 -2.51009166e-01 2.49745116e-01 1.03478205e+00 -1.03552270e+00 -6.12197280e-01 -6.58882618e-01 9.40748096e-01 -7.67573938e-02 1.95301831e-01 -1.20491970e+00 -6.89046502e-01 6.42627954e-01 5.28227128e-02 3.51038247e-01 -2.64621615e-01 -4.77876633e-01 -1.10640073e+00 3.21114659e-01 -6.27742350e-01 6.50427520e-01 -2.44088441e-01 -7.64041841e-01 2.61283457e-01 -8.27842876e-02 -1.26168871e+00 -1.60751089e-01 -1.79489270e-01 -8.99068773e-01 6.10021174e-01 -1.27007735e+00 -7.10514128e-01 -7.55661130e-01 6.95753396e-01 9.84587193e-01 2.53035665e-01 4.35556918e-01 5.66923201e-01 -8.66520345e-01 3.03002864e-01 -2.69926563e-02 5.93452275e-01 3.00428718e-01 -9.88918245e-01 3.57175440e-01 1.11752236e+00 1.55695260e-01 1.54216677e-01 3.54769856e-01 -6.40966713e-01 -9.53388691e-01 -1.07219791e+00 5.02878726e-01 1.17724732e-01 3.75459880e-01 -3.02796513e-01 -1.09438109e+00 4.39694345e-01 1.71140224e-01 1.19324155e-01 7.46791184e-01 -4.10721809e-01 2.47781590e-01 -2.02778652e-01 -8.28210473e-01 1.87987521e-01 9.39118266e-01 -1.73622519e-01 -6.35204971e-01 8.92300829e-02 3.53465378e-01 -4.22238708e-01 -7.93255031e-01 5.26808143e-01 5.70507169e-01 -1.04287446e+00 6.80322051e-01 -2.12301373e-01 1.50693268e-01 -4.03603464e-01 1.93520933e-01 -1.10336351e+00 6.83815777e-02 -4.08060104e-01 -5.32668710e-01 1.02591312e+00 8.06583837e-02 -3.79794687e-01 9.85359907e-01 2.45809317e-01 -2.44326606e-01 -9.76614714e-01 -9.57293332e-01 -6.01259410e-01 -5.15733957e-01 -5.04408777e-01 5.74341536e-01 6.80259109e-01 -3.53194661e-02 2.66044617e-01 -3.08721185e-01 -1.05633587e-01 3.82382482e-01 3.01021338e-01 6.05269790e-01 -1.24642348e+00 -1.98386043e-01 -6.02937996e-01 -2.09117249e-01 -1.42246699e+00 -2.18993798e-01 -3.52430016e-01 3.41297746e-01 -1.69194829e+00 8.13474655e-02 6.77506113e-03 -6.27281785e-01 5.50985157e-01 -2.84766257e-01 1.47116005e-01 -1.89461455e-01 1.66312814e-01 -7.91828811e-01 5.89443684e-01 1.00408447e+00 -3.38023305e-01 -7.56833136e-01 3.92639637e-01 -3.95477325e-01 7.68455148e-01 9.30501282e-01 -3.79317343e-01 -4.95647252e-01 -2.71604210e-01 -4.38309848e-01 3.19323987e-02 1.02872953e-01 -1.56646752e+00 2.04385579e-01 -3.48970324e-01 7.43105948e-01 -7.43960440e-01 3.34894836e-01 -6.82652056e-01 -6.80373162e-02 4.87005919e-01 -3.16343695e-01 -3.17947239e-01 8.57889131e-02 6.02137983e-01 -3.46096039e-01 9.91748180e-03 7.56611109e-01 -9.77367610e-02 -8.31921875e-01 4.55224246e-01 -7.62654006e-01 5.31565994e-02 1.27967012e+00 -5.70836008e-01 1.10276856e-01 -1.25558883e-01 -7.52385318e-01 2.60981977e-01 -1.15217283e-01 5.68909287e-01 6.65649056e-01 -1.32708347e+00 -3.49631488e-01 2.43802473e-01 7.65121579e-02 3.92217845e-01 2.57048249e-01 1.01748919e+00 -1.59395978e-01 2.08243474e-01 -2.25829050e-01 -6.37607217e-01 -1.20543349e+00 3.15409958e-01 5.27365386e-01 -3.88150156e-01 -4.26286876e-01 6.35773659e-01 -4.95963842e-02 2.33349338e-01 6.75269008e-01 -7.16407537e-01 -5.09154677e-01 4.18592155e-01 6.90683186e-01 5.71908236e-01 2.73900293e-02 -5.23805678e-01 -4.10417706e-01 4.59534824e-01 1.18771449e-01 -5.87997921e-02 1.31266320e+00 6.70908345e-03 1.18601449e-01 5.05708873e-01 1.08229721e+00 -4.11897033e-01 -1.58622742e+00 -2.85561889e-01 1.76079214e-01 -2.58748084e-01 -3.09952218e-02 -4.83109176e-01 -1.32551146e+00 1.01663160e+00 7.39524007e-01 4.33448553e-01 1.51743019e+00 -1.95953518e-01 1.21591580e+00 1.22897699e-01 -1.31413573e-02 -1.48396456e+00 4.73760039e-01 3.60326380e-01 3.81652325e-01 -9.43483889e-01 -6.53244853e-02 -9.29787531e-02 -5.32891035e-01 9.72477496e-01 8.31304789e-01 1.96362305e-02 5.24493456e-01 2.46137753e-01 -3.09414156e-02 5.81218675e-02 -3.05891484e-01 -2.42187142e-01 1.15149304e-01 4.96929705e-01 4.43330467e-01 -1.62189394e-01 -4.33574617e-01 9.43511426e-01 4.29601192e-01 4.59326446e-01 2.45786831e-01 1.38885140e+00 -7.43372977e-01 -8.78226519e-01 -4.55042511e-01 6.38725281e-01 -5.87588012e-01 3.76008868e-01 -1.81923732e-01 3.48925948e-01 4.29870248e-01 1.08068085e+00 2.68734753e-01 -3.74910891e-01 4.73983526e-01 1.44398063e-01 2.32763901e-01 -3.50711673e-01 -5.51544130e-01 3.67080748e-01 -1.44708782e-01 -6.50483787e-01 -6.03242636e-01 -7.15462685e-01 -1.61236250e+00 1.05819747e-01 -4.84776720e-02 1.02384828e-01 4.66647655e-01 1.27432835e+00 1.72132194e-01 8.66367340e-01 5.64561188e-01 -1.00599396e+00 -2.51816809e-01 -1.14765370e+00 -5.64039350e-01 4.86355126e-01 2.03722417e-01 -6.70419574e-01 -1.23466991e-01 2.57176965e-01]	[7.910459041595459, 0.7207064032554626]
35fa5bbe-4ddb-4eb3-ad54-5126acde84bd	texture-relative-superpixel-generation-with	null	null	https://ieeexplore.ieee.org/document/8626535	https://ieeexplore.ieee.org/document/8626535	Texture Relative Superpixel Generation With Adaptive Parameters	Abstract—Superpixel generation, which is an essential step in many image processing applications, has attracted increasing attention from researchers. In this paper, we present an efficient flooding-based superpixel generation algorithm that generates compact and highly boundary-adherent superpixels. In particular, by considering various superpixel properties , we measure the similarities between image pixels by proposing a new distance metric that combines various image features (e.g., colors, spatial locations, neighbor information and texture features). To control the relative significance of these image features, we acquire the weights of image features (e.g., colors, texture features and neighbor information of pixels) through a neural network. Then, the final superpixels are obtained through a greedy optimization that considers both the current superpixel and its neighboring superpixels. We perform extensive experiments on two datasets to verify the efficacy of our algorithm. The results show that our algorithm has considerable advantages over existing stateof-the-art methods, particularly regarding the compactness of the resulting superpixels.	['Caiming Zhang', 'Zhonggui Chen', 'Yuanfeng Zhou', 'Xiao Pan']	2019-01-25	null	null	null	ieee-2019-1	['superpixels']	['computer-vision']	[ 3.12273204e-01 -1.29885033e-01 -1.11653887e-01 -2.93823749e-01 -2.85161972e-01 -2.76948839e-01 3.02017421e-01 2.67724812e-01 -4.96050239e-01 8.34582746e-01 4.61402461e-02 1.69250891e-01 -1.62565801e-02 -9.74220812e-01 -5.49248755e-01 -1.03475189e+00 5.76991960e-02 -9.69244316e-02 7.81209409e-01 1.12153560e-01 5.86061239e-01 4.38826859e-01 -1.54760218e+00 -1.32586539e-01 1.26041889e+00 1.03686774e+00 4.30525422e-01 1.73424125e-01 -3.43590230e-02 4.58030164e-01 -3.51411670e-01 5.25615290e-02 2.64753640e-01 -4.01153266e-01 -6.92257941e-01 4.73297298e-01 3.34995985e-01 -3.71869653e-01 -5.30039985e-03 1.35280132e+00 1.95216313e-01 1.69861138e-01 4.40160573e-01 -8.49175215e-01 -6.70530498e-01 5.72186172e-01 -8.77027869e-01 3.40544999e-01 -8.19656849e-02 9.09363776e-02 8.47173691e-01 -7.29754746e-01 5.45694530e-01 9.31085885e-01 3.25322598e-01 -8.01433921e-02 -1.10299981e+00 -4.31808531e-01 8.31382945e-02 2.05580950e-01 -1.56508994e+00 -6.83656409e-02 8.01318765e-01 -2.65209377e-01 5.61162792e-02 2.99221426e-01 6.45969748e-01 9.67625901e-02 -1.85545310e-02 9.06441152e-01 1.17370427e+00 -4.40866858e-01 3.80810052e-01 2.68251687e-01 7.65432268e-02 7.84343362e-01 2.75440842e-01 -1.92771569e-01 -2.45757878e-01 2.77498760e-03 9.74237442e-01 1.45250916e-01 -4.10463750e-01 -3.53514194e-01 -1.30790484e+00 5.13914585e-01 7.96386302e-01 5.20907402e-01 -5.09267330e-01 1.48346812e-01 -1.74137339e-01 -3.74886334e-01 6.38487875e-01 6.46694144e-03 6.78491145e-02 1.70929521e-01 -1.08852446e+00 -5.92726246e-02 2.20415026e-01 7.44187176e-01 1.04227662e+00 -3.41508448e-01 -4.21966404e-01 8.93258810e-01 2.67832428e-01 3.05889428e-01 2.71126479e-01 -1.14152467e+00 3.06056291e-01 7.99458325e-01 2.52297133e-01 -1.39397168e+00 -9.40771326e-02 -2.45878965e-01 -1.08678794e+00 3.30392234e-02 2.27886871e-01 -6.76744282e-02 -6.93191588e-01 1.38228309e+00 6.06938064e-01 4.97277141e-01 -1.35588974e-01 1.06275010e+00 6.04125679e-01 8.13892245e-01 -1.60045903e-02 -4.66263831e-01 1.15867376e+00 -9.55668747e-01 -3.25447410e-01 -9.84683186e-02 9.81934443e-02 -7.75400043e-01 6.76507711e-01 -3.19826007e-02 -1.18743575e+00 -6.03882909e-01 -9.92298186e-01 1.85235128e-01 -2.00455844e-01 3.04055125e-01 5.69394708e-01 3.84637892e-01 -9.83778834e-01 6.82630122e-01 -6.96274459e-01 -4.16296422e-01 3.47093433e-01 2.41141289e-01 7.43453726e-02 6.61295727e-02 -9.09898579e-01 1.33099690e-01 8.00116777e-01 8.29922035e-02 -3.38635445e-01 -3.59013319e-01 -5.56239367e-01 1.86645687e-01 2.23848805e-01 -4.56108749e-01 7.12955236e-01 -8.83513808e-01 -1.13834870e+00 6.34343863e-01 -1.87666133e-01 -8.57082158e-02 3.89379323e-01 3.20743501e-01 -1.95477959e-02 4.30396438e-01 2.17460483e-01 8.64923477e-01 7.50904739e-01 -1.25626302e+00 -1.16922319e+00 -2.31853500e-01 3.57840036e-04 2.94178396e-01 -7.29893088e-01 -2.33162995e-02 -8.14592719e-01 -6.78421915e-01 4.67597306e-01 -8.24534059e-01 -4.51964825e-01 2.48037204e-01 -6.06363952e-01 -3.36327702e-01 6.22246265e-01 -3.45127970e-01 1.33084702e+00 -2.35644269e+00 1.27673849e-01 4.51083958e-01 2.87161529e-01 2.62652248e-01 3.00737005e-02 -8.25822800e-02 3.08282912e-01 2.36533552e-01 -5.18811047e-01 -2.38020401e-02 -3.50433350e-01 -1.22193553e-01 1.50945082e-01 2.67347395e-01 1.27630504e-02 6.10462070e-01 -1.04267693e+00 -1.01858389e+00 4.07517582e-01 2.84805566e-01 -1.98258311e-01 1.18987188e-01 -1.07026719e-01 3.04582536e-01 -9.55195546e-01 6.10718131e-01 9.76407707e-01 -3.58720332e-01 -5.11166640e-02 -4.08397257e-01 -5.30882537e-01 -3.65327895e-01 -1.18013024e+00 1.26671112e+00 -7.81040639e-02 4.44544554e-01 -1.00152805e-01 -8.65320802e-01 9.26219821e-01 -3.18166837e-02 6.95154011e-01 -5.32486558e-01 1.51947007e-01 2.93102086e-01 -4.55248684e-01 -4.08440143e-01 8.17708969e-01 2.80618727e-01 1.25758827e-01 6.48024976e-01 -5.49593568e-01 -4.80521396e-02 6.45350993e-01 2.26409167e-01 6.70191109e-01 -1.29057303e-01 1.06628895e-01 -3.68445337e-01 6.62861228e-01 6.40683323e-02 5.68533421e-01 4.47347760e-01 -2.39394084e-01 8.81109357e-01 2.69748569e-01 -4.19638604e-01 -1.02529049e+00 -8.52514088e-01 -8.98942351e-02 6.13315284e-01 9.06482041e-01 -1.00981973e-01 -1.06338382e+00 -4.38935488e-01 -5.87708503e-02 3.51966411e-01 -5.38631201e-01 -9.03016403e-02 -5.12783408e-01 -1.03980613e+00 1.40767604e-01 5.46018302e-01 9.88367081e-01 -8.42531323e-01 -7.17791080e-01 2.74460346e-01 -4.49439198e-01 -1.04677832e+00 -7.94535577e-01 -2.60815024e-01 -1.14166045e+00 -8.25601220e-01 -9.92979527e-01 -1.08037603e+00 1.13742864e+00 9.43287849e-01 7.89602637e-01 3.35896969e-01 -4.58635032e-01 -1.78632110e-01 -4.92139459e-01 2.80446112e-01 -1.40056089e-01 -5.38529828e-03 -2.44001105e-01 3.30147147e-01 -4.91821170e-02 -5.51304102e-01 -1.02068222e+00 4.15264636e-01 -1.06414175e+00 4.06776816e-01 9.38924193e-01 5.20618081e-01 1.06262553e+00 3.99140209e-01 -2.73456406e-02 -8.10217083e-01 5.57780743e-01 -3.58397126e-01 -7.70300448e-01 5.30517995e-01 -4.13010418e-01 1.54301986e-01 3.19341838e-01 -1.94252491e-01 -1.14446592e+00 8.16291273e-02 4.59206879e-01 -1.23391569e-01 5.32109030e-02 5.00307977e-01 -1.02933601e-01 -3.00694555e-01 2.04673514e-01 6.12291038e-01 -2.33603343e-01 -3.13235849e-01 3.59257162e-01 7.98666477e-01 4.36536372e-01 -3.33474636e-01 7.39912331e-01 7.05850124e-01 -1.34883329e-01 -8.38195384e-01 -5.57159960e-01 -4.35721517e-01 -7.25966632e-01 -1.83985144e-01 9.03475463e-01 -5.65871418e-01 -5.61764240e-01 7.33582437e-01 -1.10950887e+00 3.97988930e-02 1.60015654e-02 3.76685888e-01 -2.89969563e-01 6.78488553e-01 -5.50625801e-01 -6.23081744e-01 -3.21525306e-01 -9.81631458e-01 9.79857624e-01 7.52126038e-01 1.36083290e-01 -8.84334862e-01 -1.59716889e-01 3.65829408e-01 3.19658637e-01 4.01795745e-01 8.60209584e-01 4.98063713e-02 -1.16843486e+00 7.47449100e-02 -8.44143569e-01 3.76198947e-01 4.09176707e-01 2.76544720e-01 -3.47537905e-01 -7.46885017e-02 -3.34863186e-01 2.17366088e-02 9.34305847e-01 6.31051898e-01 1.21716785e+00 -3.59130889e-01 -6.11692786e-01 7.10241079e-01 1.61177266e+00 1.42739952e-01 5.44390500e-01 3.45172763e-01 6.29517794e-01 3.00563097e-01 9.15953934e-01 5.69417119e-01 2.08974779e-01 5.72167158e-01 4.23325211e-01 -2.82582581e-01 -1.12748481e-01 -1.48800546e-02 6.18442520e-03 6.67364240e-01 -3.38266611e-01 -2.27340236e-01 -4.69660789e-01 7.25508273e-01 -1.82126117e+00 -7.70635068e-01 -1.63348615e-01 2.15461206e+00 8.98973167e-01 5.92756048e-02 -1.51462749e-01 -9.36086550e-02 1.21994340e+00 3.38763148e-01 -6.51418269e-01 1.01267159e-01 -1.87773123e-01 -1.87732309e-01 6.72098756e-01 2.57277131e-01 -1.33273983e+00 1.08926105e+00 5.73213196e+00 1.03763247e+00 -1.01105881e+00 -2.19684064e-01 1.08030176e+00 6.23261333e-02 -3.44151706e-01 -2.67870650e-02 -6.41005576e-01 7.96435058e-01 -5.92362806e-02 -2.93611467e-01 3.27961981e-01 6.39957488e-01 4.41003472e-01 -5.76559961e-01 -5.45619488e-01 7.63231754e-01 1.39210448e-01 -1.38912880e+00 6.89331293e-02 -3.23679894e-02 1.11975384e+00 -1.89005345e-01 8.28766152e-02 -6.20893121e-01 3.47346693e-01 -6.49132133e-01 7.84240067e-01 3.94585788e-01 5.79323530e-01 -6.51824296e-01 7.46349692e-01 1.78405344e-01 -1.41368842e+00 6.79291189e-02 -5.16143739e-01 3.16905379e-01 2.75289595e-01 9.99004900e-01 -4.64315504e-01 4.91895318e-01 7.09191918e-01 8.22905481e-01 -5.06835282e-01 1.27981877e+00 -2.03010201e-01 4.30156112e-01 -4.83256072e-01 -2.83402294e-01 3.93784881e-01 -6.45367622e-01 3.05487812e-01 8.78576994e-01 4.50394809e-01 3.35025162e-01 1.57384560e-01 1.19577014e+00 -6.74360767e-02 3.09928358e-01 -1.29915819e-01 1.17153093e-01 7.83309281e-01 1.46277714e+00 -1.19233525e+00 -4.33281600e-01 -2.12316051e-01 1.01306331e+00 3.13558757e-01 3.65458161e-01 -7.34843254e-01 -5.09948075e-01 4.27879453e-01 1.91804897e-02 3.34905297e-01 -1.21731460e-01 -3.04863006e-01 -9.21748400e-01 2.73223251e-01 -2.12261394e-01 1.14327125e-01 -7.37953901e-01 -7.50197947e-01 4.88802105e-01 -6.17801361e-02 -1.39400971e+00 9.72621888e-02 -6.47503734e-02 -8.76392841e-01 5.93835950e-01 -1.54804289e+00 -9.93622422e-01 -6.70857847e-01 3.90138388e-01 3.71746659e-01 2.76289761e-01 1.63287252e-01 1.25538707e-01 -7.53167987e-01 2.19636172e-01 4.07777488e-01 1.93210661e-01 3.27961266e-01 -9.89425421e-01 2.08233342e-01 9.68748927e-01 -4.59681042e-02 4.28241640e-01 4.47571725e-01 -4.85802412e-01 -7.65662253e-01 -1.29056704e+00 7.68304110e-01 3.90128911e-01 3.85791838e-01 2.15250298e-01 -6.27330363e-01 1.53110966e-01 7.24105015e-02 1.06687867e-03 3.36159229e-01 -5.64596534e-01 2.64019936e-01 -3.75322193e-01 -1.07514071e+00 6.46871030e-01 8.87845397e-01 9.44046602e-02 -1.60503253e-01 3.55054349e-01 5.11379719e-01 -1.13292903e-01 -6.27806723e-01 3.43555272e-01 4.05448884e-01 -1.29553783e+00 1.00602388e+00 6.57945573e-02 4.66475099e-01 -5.97174466e-01 1.78123474e-01 -1.24391866e+00 -5.08472323e-01 -2.45415717e-01 3.88046622e-01 1.33233523e+00 2.11057156e-01 -5.31641304e-01 9.43502009e-01 5.60337782e-01 1.20211974e-01 -7.75712430e-01 -7.22255528e-01 -6.20993316e-01 -4.36588556e-01 1.71585873e-01 6.58913672e-01 5.74114323e-01 2.21722592e-02 -4.29348201e-02 -2.32832178e-01 1.57732293e-01 8.76136422e-01 6.95286930e-01 3.72188926e-01 -1.07617521e+00 6.09876998e-02 -5.57498395e-01 -4.20496315e-01 -1.23540914e+00 -1.15868829e-01 -5.80842435e-01 2.51209140e-01 -1.61301136e+00 7.27513790e-01 -8.22243273e-01 -4.08903509e-01 3.69460911e-01 -6.58824503e-01 6.10717058e-01 5.86420074e-02 4.25358087e-01 -7.91292906e-01 5.08561492e-01 1.52275527e+00 -2.58518875e-01 -3.48526865e-01 -6.50350079e-02 -6.35390639e-01 7.12321818e-01 9.26180124e-01 -3.13960612e-01 -3.16784859e-01 -6.71250761e-01 -2.04940721e-01 -1.45085379e-01 3.36315870e-01 -1.18165815e+00 5.03556073e-01 -4.63379920e-01 2.53018349e-01 -5.56128919e-01 1.09248720e-01 -6.18921101e-01 -9.36326757e-03 5.46608031e-01 -2.05934256e-01 -4.12524581e-01 -2.89101481e-01 5.57102203e-01 -3.88225347e-01 -2.35559568e-01 9.36086416e-01 -2.51353055e-01 -9.04133379e-01 4.72626030e-01 -1.83785409e-01 -1.13684461e-01 1.44860923e+00 -5.78713000e-01 -1.45832717e-01 -2.00176552e-01 -2.32001781e-01 3.64578843e-01 7.00045228e-01 1.11876786e-01 8.17206979e-01 -1.30677879e+00 -7.59810150e-01 1.86086804e-01 1.16270512e-01 2.72072703e-01 1.35517552e-01 8.03180754e-01 -9.08961535e-01 2.67801553e-01 -2.24375933e-01 -6.36095047e-01 -1.53049743e+00 3.72879475e-01 2.13726666e-02 1.48827359e-02 -3.24430615e-01 9.45546567e-01 2.43153259e-01 2.18740985e-01 -2.36003734e-02 -5.99431574e-01 -2.77551025e-01 -1.14911497e-01 3.65460634e-01 5.36971450e-01 -3.72957379e-01 -7.02813089e-01 -2.71046400e-01 9.76205349e-01 -5.63335754e-02 6.86827153e-02 1.36487460e+00 -3.54193509e-01 -5.14334559e-01 -7.04844743e-02 1.06704819e+00 -1.08837508e-01 -1.43140757e+00 -4.05451447e-01 -8.57274756e-02 -9.23443377e-01 2.09602103e-01 -1.29291758e-01 -1.64224136e+00 6.41566515e-01 5.18498063e-01 2.25127086e-01 1.25590992e+00 1.57384783e-01 1.12111175e+00 -6.97976500e-02 4.93229330e-01 -1.25602961e+00 1.07802160e-01 4.23582569e-02 3.63029838e-01 -1.18050122e+00 2.88117737e-01 -8.47802877e-01 -3.30508679e-01 1.02174878e+00 4.81713206e-01 -4.49840873e-02 6.25084817e-01 -7.30603635e-02 2.78769471e-02 9.61260423e-02 -2.00079635e-01 -1.59640849e-01 4.46406752e-02 2.40589783e-01 1.81369871e-01 1.03163429e-01 -5.49601316e-01 1.41835615e-01 1.75555319e-01 1.83319286e-01 4.00155842e-01 7.79977739e-01 -8.59704852e-01 -9.49549019e-01 -3.73124301e-01 3.69114131e-01 -1.43427417e-01 -1.85822591e-01 -1.71234846e-01 3.39594036e-01 2.98900962e-01 9.22515750e-01 2.47319743e-01 -2.95969963e-01 -1.07826263e-01 -7.32518077e-01 2.31267899e-01 -4.18351948e-01 -1.93363130e-02 2.24913359e-02 -4.61975098e-01 -4.08516377e-01 -8.09849918e-01 -5.82660675e-01 -1.26726329e+00 -2.85534829e-01 -4.36462134e-01 2.08463088e-01 4.25177187e-01 8.29630792e-01 3.71983647e-01 2.42988333e-01 9.74318087e-01 -9.24933195e-01 -2.58222550e-01 -7.42488325e-01 -8.38846207e-01 4.96697903e-01 -3.43173891e-02 -3.85762841e-01 -2.90173024e-01 3.07900310e-01]	[9.636926651000977, -0.5314019322395325]
d44966d5-0a1f-4bbe-966e-9adb79e2540c	semantic-composition-in-visually-grounded	2305.16328	null	https://arxiv.org/abs/2305.16328v1	https://arxiv.org/pdf/2305.16328v1.pdf	Semantic Composition in Visually Grounded Language Models	What is sentence meaning and its ideal representation? Much of the expressive power of human language derives from semantic composition, the mind's ability to represent meaning hierarchically & relationally over constituents. At the same time, much sentential meaning is outside the text and requires grounding in sensory, motor, and experiential modalities to be adequately learned. Although large language models display considerable compositional ability, recent work shows that visually-grounded language models drastically fail to represent compositional structure. In this thesis, we explore whether & how models compose visually grounded semantics, and how we might improve their ability to do so. Specifically, we introduce 1) WinogroundVQA, a new compositional visual question answering benchmark, 2) Syntactic Neural Module Distillation, a measure of compositional ability in sentence embedding models, 3) Causal Tracing for Image Captioning Models to locate neural representations vital for vision-language composition, 4) Syntactic MeanPool to inject a compositional inductive bias into sentence embeddings, and 5) Cross-modal Attention Congruence Regularization, a self-supervised objective function for vision-language relation alignment. We close by discussing connections of our work to neuroscience, psycholinguistics, formal semantics, and philosophy.	['Rohan Pandey']	2023-05-15	null	null	null	null	['image-captioning', 'sentence-embedding', 'sentence-embeddings', 'philosophy', 'sentence-embedding', 'sentence-embeddings', 'semantic-composition']	['computer-vision', 'methodology', 'methodology', 'miscellaneous', 'natural-language-processing', 'natural-language-processing', 'natural-language-processing']	[ 3.44973832e-01 4.35480028e-01 -4.66079712e-02 -4.98130918e-01 -2.13872820e-01 -5.00872552e-01 9.22209740e-01 2.07765862e-01 -2.80400008e-01 2.35089317e-01 1.25723863e+00 -4.28188413e-01 -6.52032578e-03 -6.30145609e-01 -6.06174707e-01 -4.59445238e-01 1.65051416e-01 1.18910894e-01 -1.48627415e-01 -3.80231291e-01 4.91335541e-01 1.82007357e-01 -1.44838893e+00 8.36520135e-01 6.56655550e-01 6.92744613e-01 5.31477451e-01 6.65587008e-01 -6.88284993e-01 1.58270669e+00 -1.72342643e-01 -4.48843300e-01 -1.13174543e-01 -1.15186429e+00 -1.13798225e+00 2.19100099e-02 9.40717757e-01 1.19607367e-01 -4.16558444e-01 1.15812945e+00 3.41131717e-01 -2.52118587e-01 8.49106252e-01 -1.06542540e+00 -1.59619379e+00 8.19776714e-01 -1.66707903e-01 4.92862165e-01 3.70998859e-01 5.11873543e-01 1.27307796e+00 -6.36983156e-01 8.76868427e-01 1.73860085e+00 3.14832509e-01 8.21096122e-01 -1.52355206e+00 -2.37832144e-01 2.31968030e-01 1.48881450e-01 -7.32815325e-01 -4.70641911e-01 6.91470563e-01 -9.13216233e-01 1.16826200e+00 1.33643776e-01 9.95796978e-01 1.05408502e+00 4.29750055e-01 5.78814626e-01 1.46473110e+00 -4.84516382e-01 2.48399414e-02 1.70722991e-01 3.27739805e-01 1.03808188e+00 -9.82909575e-02 1.52945623e-01 -9.01239991e-01 6.27280995e-02 7.28299201e-01 -4.14358675e-01 -9.63551775e-02 -1.12859748e-01 -1.54475129e+00 8.68823826e-01 8.61219585e-01 3.39593589e-01 -3.38412344e-01 7.66569138e-01 3.73492151e-01 2.98393250e-01 6.52350634e-02 8.30980241e-01 -2.10504588e-02 4.24827456e-01 -4.27794784e-01 1.43195078e-01 4.01813984e-01 4.35373008e-01 4.61238384e-01 3.48443598e-01 -2.49925166e-01 5.42882442e-01 5.36973178e-01 5.63537836e-01 6.08564138e-01 -1.13006425e+00 4.45023477e-02 7.44777977e-01 -5.86383641e-01 -1.02334547e+00 -2.46121228e-01 1.49203083e-02 -3.99301678e-01 2.82465965e-01 1.67790309e-01 3.53803009e-01 -6.31206989e-01 2.16450357e+00 -2.23986357e-01 -2.42952809e-01 1.51207373e-01 1.10722053e+00 1.20189369e+00 5.06657064e-01 7.31245100e-01 1.75127722e-02 1.82699978e+00 -4.43734169e-01 -5.38555622e-01 -5.81389546e-01 4.25735712e-01 -6.16127372e-01 1.56346643e+00 -1.13382727e-01 -1.32463861e+00 -5.40816128e-01 -9.77912486e-01 -6.97373807e-01 -2.60438293e-01 -2.14931726e-01 9.00510311e-01 1.05173007e-01 -1.30735612e+00 1.80173308e-01 -4.07782018e-01 -6.44876838e-01 6.82514369e-01 2.18158104e-02 -4.48014766e-01 1.07690483e-01 -1.00257385e+00 1.33294582e+00 3.12621742e-01 -2.09791601e-01 -9.64789987e-01 -6.55358791e-01 -1.09931052e+00 -1.43089801e-01 -1.26402035e-01 -1.54231000e+00 1.05333078e+00 -1.56078339e+00 -1.00796425e+00 1.50395346e+00 -5.36012530e-01 -5.58090627e-01 -3.60879213e-01 1.10343061e-01 -4.16468382e-02 4.19123858e-01 2.02696487e-01 1.19290638e+00 7.87139893e-01 -1.41979158e+00 -1.95816889e-01 -6.04002297e-01 2.55698740e-01 3.77215773e-01 -6.34748628e-03 2.48141885e-01 1.91128775e-01 -6.48108900e-01 2.37144694e-01 -4.50006187e-01 -8.68907943e-02 3.48480105e-01 -1.77612022e-01 -3.85024846e-01 2.25478947e-01 -6.92679226e-01 8.73104751e-01 -2.16860986e+00 6.62110150e-01 -1.63318947e-01 6.12160206e-01 -4.44145769e-01 -3.87538671e-01 3.95838410e-01 -3.82016212e-01 2.98313469e-01 -3.93466473e-01 4.29504849e-02 1.53571740e-01 3.46213132e-01 -5.79480946e-01 4.32119399e-01 5.40959775e-01 1.49256241e+00 -1.00516939e+00 -7.41715968e-01 2.76716292e-01 3.73688281e-01 -7.13561654e-01 -1.46255583e-01 -4.61063236e-01 8.69540423e-02 -1.40957743e-01 4.05943990e-01 -1.51775315e-01 -3.63838404e-01 1.46332070e-01 -3.42401922e-01 -1.10909501e-02 6.71198905e-01 -5.03031075e-01 1.72416639e+00 -2.43804261e-01 1.14765990e+00 -9.62172151e-02 -1.11877120e+00 5.37883461e-01 1.45844683e-01 -1.21266313e-01 -1.19446707e+00 2.02409893e-01 -1.37859806e-01 4.38510954e-01 -8.81871700e-01 2.99496293e-01 -8.41639817e-01 -6.23501502e-02 4.03319359e-01 7.58773088e-02 -5.00656724e-01 -1.26388192e-01 5.03709674e-01 8.50296259e-01 1.88720584e-01 3.29277843e-01 -5.35211980e-01 4.28467035e-01 3.15535814e-01 1.41285002e-01 2.54034668e-01 -2.35832691e-01 4.02902037e-01 7.70120561e-01 -3.49348575e-01 -1.14080715e+00 -1.57229340e+00 -1.18450753e-01 1.24453402e+00 7.83783291e-03 -3.73993665e-01 -7.52438605e-01 -3.34597453e-02 1.90275311e-02 1.39894450e+00 -9.19112086e-01 -3.76530677e-01 -3.85062009e-01 -3.85979533e-01 5.10038435e-01 5.84470987e-01 6.25276286e-03 -1.66376734e+00 -8.07051778e-01 -1.72927663e-01 -2.34936178e-02 -9.10323501e-01 -2.74269760e-01 3.68341468e-02 -7.27150023e-01 -1.04733944e+00 -1.93612218e-01 -1.10685444e+00 7.75357604e-01 1.31751582e-01 1.47965348e+00 -9.43387598e-02 -5.55409133e-01 8.92725289e-01 -1.49044311e-02 -5.00996649e-01 -6.92569733e-01 -7.76257038e-01 -6.35426044e-02 -4.53420222e-01 5.31765103e-01 -2.66575128e-01 -5.13633788e-01 -1.46828398e-01 -8.44062388e-01 3.48614872e-01 4.10070568e-01 7.05286562e-01 5.48376501e-01 -6.63728476e-01 3.58124107e-01 -5.64666748e-01 9.32607830e-01 -3.47932667e-01 1.84365697e-02 1.50719404e-01 -2.09183842e-01 3.37201029e-01 2.84368396e-01 -3.19206148e-01 -8.22843969e-01 -2.43885443e-01 1.00711592e-01 -1.22527018e-01 -4.77201603e-02 4.50503230e-01 3.37612666e-02 2.40168199e-01 1.14247048e+00 3.28298986e-01 6.56278491e-01 1.15658090e-01 1.02133000e+00 1.38537258e-01 5.97021401e-01 -7.59198070e-01 5.72448075e-01 6.98651016e-01 1.01089636e-02 -1.17036104e+00 -9.63539362e-01 -1.61082670e-01 -5.53869069e-01 3.18117291e-02 1.58475828e+00 -8.57637167e-01 -8.91504467e-01 -9.29692462e-02 -1.29596579e+00 -2.59699434e-01 -7.87934482e-01 4.35718626e-01 -8.93328905e-01 4.06413466e-01 -6.38194799e-01 -5.63801646e-01 -3.23713750e-01 -1.03597927e+00 8.48850191e-01 3.53985988e-02 -7.10583866e-01 -9.84677136e-01 -2.32927240e-02 4.45395619e-01 1.68791965e-01 3.52530450e-01 1.49755728e+00 -1.03243001e-01 -4.38098073e-01 4.41223472e-01 -6.70031011e-01 3.63151401e-01 -2.04480410e-01 -2.49875113e-01 -8.59735966e-01 1.82344079e-01 -1.01056637e-03 -6.78111970e-01 1.03178144e+00 5.78558564e-01 9.45931733e-01 -2.71745622e-01 1.07651986e-01 3.92830014e-01 1.24277079e+00 -1.32178545e-01 7.14469314e-01 1.03542425e-01 6.87430203e-01 1.27197540e+00 -1.32738933e-01 -2.63077855e-01 7.58079946e-01 2.01563939e-01 4.27026093e-01 -1.65293425e-01 -6.89136565e-01 -2.90330559e-01 6.42589748e-01 9.30691481e-01 1.95788182e-02 3.48905832e-01 -9.73098695e-01 5.33713877e-01 -1.52463305e+00 -1.42931700e+00 -3.17058086e-01 1.60227203e+00 9.79866624e-01 -4.94841672e-02 -5.18065058e-02 -3.14047247e-01 2.16488406e-01 2.01977715e-01 -3.61355990e-01 -8.64075899e-01 -4.98302609e-01 2.18017384e-01 7.96997845e-02 5.62752903e-01 -6.08138263e-01 1.19843054e+00 7.04345989e+00 2.08698586e-01 -8.44237626e-01 1.27249151e-01 3.37303460e-01 -6.72036335e-02 -1.00467730e+00 1.27379641e-01 -1.41691402e-01 8.22994336e-02 9.21203613e-01 1.44398054e-02 5.92486739e-01 4.02992368e-01 2.43589446e-01 -1.35052159e-01 -1.40483487e+00 9.92250681e-01 6.55533254e-01 -1.69765091e+00 4.40519601e-01 -5.16558215e-02 2.59038627e-01 7.93641508e-02 6.21425807e-02 4.21330072e-02 2.83351034e-01 -1.50558507e+00 1.03827369e+00 8.01385045e-01 6.57968998e-01 -2.96021342e-01 9.68991444e-02 -4.10755351e-02 -1.00450194e+00 -1.33267507e-01 -4.97586966e-01 -3.47676754e-01 1.02248326e-01 -1.05413541e-01 -3.52085322e-01 -1.28675401e-01 4.36596572e-01 4.81163174e-01 -8.40559661e-01 3.85055423e-01 -4.48822349e-01 4.92445141e-01 2.16194302e-01 -3.65909666e-01 2.59782344e-01 1.05059586e-01 4.68195707e-01 1.20115125e+00 -2.78901219e-01 3.13172460e-01 -2.01826870e-01 1.45921588e+00 2.25073099e-01 1.06651798e-01 -9.34969187e-01 -5.40967882e-01 7.71277398e-02 8.34564984e-01 -4.59805250e-01 -2.36439288e-01 -5.72970808e-01 1.03696311e+00 4.05016094e-01 4.00041938e-01 -6.26349032e-01 1.26023561e-01 9.39581215e-01 2.37962112e-01 -2.95639277e-01 -3.29527736e-01 -5.99729240e-01 -1.03955305e+00 -3.18339497e-01 -8.00204039e-01 1.45349815e-01 -1.42797554e+00 -1.62243223e+00 2.24549279e-01 -7.02300593e-02 -6.19141936e-01 -1.59054384e-01 -1.05185938e+00 -4.95952696e-01 1.21938396e+00 -1.15231609e+00 -1.44775295e+00 -3.06734256e-02 5.65247416e-01 5.76064408e-01 -2.13066429e-01 1.06767964e+00 -3.31167191e-01 -8.35089236e-02 1.75713882e-01 -7.15057135e-01 2.23649278e-01 4.26805168e-01 -1.47218215e+00 3.06901276e-01 6.03308141e-01 4.69219238e-01 9.80107427e-01 7.74492085e-01 -3.73027116e-01 -1.50541842e+00 -5.71847498e-01 1.00402927e+00 -9.60504830e-01 1.06819808e+00 -3.64948720e-01 -7.53643870e-01 6.52815700e-01 8.15491915e-01 -2.00591475e-01 1.01649249e+00 1.08873248e-01 -8.55215192e-01 1.85939506e-01 -8.43216538e-01 9.90028501e-01 1.14552903e+00 -1.14293599e+00 -1.48679018e+00 2.35304043e-01 1.07817900e+00 1.52635664e-01 -5.87893903e-01 -5.63109703e-02 5.92570543e-01 -7.52608776e-01 1.16168034e+00 -1.08026373e+00 9.65981483e-01 -4.58998382e-01 -4.77509648e-01 -9.21523333e-01 -6.55974507e-01 -1.17581904e-01 1.69227362e-01 8.52520823e-01 3.27460587e-01 -4.55156624e-01 2.57489562e-01 4.97960836e-01 -2.07313895e-01 -5.51174939e-01 -7.83510089e-01 -1.91944391e-01 4.87481147e-01 -6.30921662e-01 4.15871292e-02 1.20215118e+00 1.09046675e-01 1.00199711e+00 3.53047192e-01 -1.70456380e-01 6.55201793e-01 2.03463174e-02 4.07713950e-01 -1.11045885e+00 -3.63131255e-01 -9.38945830e-01 -6.55628741e-01 -6.89746380e-01 5.24163485e-01 -1.48642230e+00 -2.17849985e-01 -2.10718417e+00 5.33189833e-01 3.55339468e-01 -6.22088388e-02 5.14444768e-01 5.31012677e-02 2.83679575e-01 5.84456801e-01 8.72218162e-02 -2.44784594e-01 3.64800096e-01 1.58494866e+00 -2.61753440e-01 2.22805560e-01 -1.05029488e+00 -1.23627245e+00 1.12397575e+00 5.80420136e-01 -5.31320944e-02 -6.86699510e-01 -6.05496943e-01 7.37474859e-01 -1.29321113e-01 1.08928585e+00 -3.89827549e-01 -9.26012993e-02 -4.08046603e-01 6.47619426e-01 -1.67983428e-01 1.78657159e-01 -5.24191678e-01 -3.44817609e-01 5.66675186e-01 -6.87634885e-01 1.32671624e-01 3.12542558e-01 2.73494929e-01 -1.52312949e-01 -2.03337163e-01 7.41196275e-01 -4.92268175e-01 -1.00002277e+00 -2.77528822e-01 -5.48992872e-01 5.12492657e-01 7.12408245e-01 -3.99177730e-01 -7.10920334e-01 -3.17237782e-03 -9.12254691e-01 1.39143154e-01 5.57752907e-01 5.39795160e-01 9.60018694e-01 -1.39232564e+00 -7.62444317e-01 -1.71191588e-01 3.98100197e-01 -4.98803854e-01 5.15864305e-02 8.63390148e-01 -6.27403140e-01 2.29109675e-01 -2.86278814e-01 -5.44510186e-01 -1.01266420e+00 6.67402983e-01 4.38047796e-01 6.28510952e-01 -6.62539124e-01 1.11351526e+00 8.17130685e-01 -2.72410512e-01 -9.45233777e-02 -6.08334243e-01 -2.50719339e-01 2.02560514e-01 5.11088014e-01 -1.53271630e-01 -8.13950658e-01 -1.04734707e+00 -4.92076308e-01 8.65147352e-01 4.17866498e-01 -3.30448538e-01 9.76505160e-01 -1.48413673e-01 -8.04603934e-01 8.06016862e-01 1.10954571e+00 -8.04779083e-02 -8.13255489e-01 -2.29503512e-02 1.84298810e-02 -3.15556265e-02 -8.56364220e-02 -7.86891460e-01 -4.03762788e-01 1.20632946e+00 3.85250628e-01 1.27544954e-01 8.93139243e-01 6.67996347e-01 3.17818344e-01 2.18849674e-01 -9.97630507e-02 -8.08943450e-01 4.58919674e-01 4.57494974e-01 1.31352174e+00 -1.13775730e+00 5.91732115e-02 -3.15275267e-02 -9.87646222e-01 1.04140234e+00 5.78152299e-01 -3.71777683e-01 4.53117192e-01 3.06767263e-02 -3.62863787e-03 -7.26563871e-01 -7.95494139e-01 -6.11505926e-01 6.17500246e-01 5.91048121e-01 8.21955383e-01 1.76645592e-01 -2.36394852e-01 2.80572653e-01 -5.74411213e-01 -3.38114202e-01 1.88792557e-01 4.96624529e-01 -6.74549460e-01 -4.82779086e-01 -3.05610448e-01 4.52921629e-01 -1.46835208e-01 -5.16228616e-01 -7.21198916e-01 6.42615378e-01 4.15599793e-01 6.70655072e-01 4.02827144e-01 -8.80166143e-02 1.99039251e-01 3.93709630e-01 1.02528119e+00 -9.52746749e-01 -6.04250431e-01 -1.52696341e-01 1.80922046e-01 -6.37778521e-01 -7.02098906e-01 -5.63842833e-01 -1.76369214e+00 -2.16461778e-01 3.24799895e-01 -3.58939409e-01 5.60924292e-01 9.68117237e-01 1.48445070e-01 7.61906147e-01 -2.54907161e-01 -3.63763571e-01 -2.14917749e-01 -7.44898260e-01 -2.20822349e-01 6.54066563e-01 2.92334586e-01 -3.10469121e-01 -2.88692802e-01 5.41518986e-01]	[10.728156089782715, 1.9423229694366455]
9f790380-14c4-4fdb-8f54-d58c2975af15	mask-selection-and-propagation-for	null	null	https://openaccess.thecvf.com/content/WACV2021/html/Garg_Mask_Selection_and_Propagation_for_Unsupervised_Video_Object_Segmentation_WACV_2021_paper.html	https://openaccess.thecvf.com/content/WACV2021/papers/Garg_Mask_Selection_and_Propagation_for_Unsupervised_Video_Object_Segmentation_WACV_2021_paper.pdf	Mask Selection and Propagation for Unsupervised Video Object Segmentation	In this work we present a novel approach for Unsupervised Video Object Segmentation, that is automatically generating instance level segmentation masks for salient objects and tracking them in a video. We efficiently handle problems present in existing methods such as drift while temporal propagation, tracking and addition of new objects. To this end, we propose a novel idea of improving masks in an online manner using ensemble of criteria whose task is to inspect the quality of masks. We introduce a novel idea of assessing mask quality using a neural network called Selector Net. The proposed network is trained is such way that it is generalizes across various datasets. Our proposed method is able to limit the noise accumulated along the video, giving state of the art result on Davis 2019 Unsupervised challenge dataset with J &F mean 61.6%. We also tested on datasets such as FBMS and SegTrack V2 and performed better or on par compared to the other methods.	['Vidit Goel', 'Shubhika Garg']	2021-01-05	null	null	null	null	['unsupervised-video-object-segmentation']	['computer-vision']	[ 5.00840247e-01 6.89188614e-02 6.81673288e-02 -4.17192280e-01 -5.48914135e-01 -6.02960467e-01 4.72145379e-01 8.75062570e-02 -8.82915020e-01 6.69883549e-01 -6.92492276e-02 2.50576109e-01 -1.19674392e-01 -4.11139995e-01 -8.18460524e-01 -5.13040841e-01 -2.00351268e-01 5.44868290e-01 1.02653837e+00 -7.84562156e-02 4.11207676e-01 5.11058509e-01 -1.68358719e+00 3.53439897e-01 8.47966731e-01 8.36444855e-01 2.29147539e-01 9.29988146e-01 7.51808286e-02 7.71725714e-01 -6.01491392e-01 -2.63801336e-01 4.60851640e-01 -1.53565466e-01 -1.07769680e+00 4.96942192e-01 6.90630794e-01 -1.45199642e-01 7.45030791e-02 1.09990311e+00 3.24525505e-01 3.85171801e-01 7.38913715e-01 -1.09251583e+00 1.33617118e-03 9.74256456e-01 -6.47955179e-01 7.30945826e-01 -9.97139812e-02 2.28228465e-01 5.92898011e-01 -6.20544732e-01 1.01207435e+00 1.08263493e+00 7.17171192e-01 7.32479632e-01 -1.06513429e+00 -3.66476238e-01 4.92813408e-01 1.12723283e-01 -1.32425046e+00 -4.56987530e-01 5.73158562e-01 -7.08086848e-01 7.57820725e-01 2.91011006e-01 4.02675390e-01 6.87639534e-01 -7.21710324e-02 1.03563547e+00 1.02399158e+00 -2.79300988e-01 2.19931275e-01 2.33779654e-01 2.29931056e-01 5.96251905e-01 1.04039967e-01 -1.82165369e-01 -4.73861396e-01 3.39368939e-01 4.23604429e-01 -3.02136838e-01 -2.31465977e-02 -2.93304235e-01 -1.04817724e+00 3.68401438e-01 3.02848041e-01 5.22950768e-01 -4.61148739e-01 2.32309461e-01 4.63763297e-01 -1.30768627e-01 6.18524790e-01 4.33569342e-01 -6.98055685e-01 -1.51274381e-02 -1.62156653e+00 4.40185905e-01 3.21821749e-01 8.38489950e-01 5.10655403e-01 -1.21164117e-02 -6.02426708e-01 5.39402068e-01 3.09453666e-01 -1.61454640e-02 3.88822496e-01 -9.82747674e-01 3.11787903e-01 5.94127119e-01 2.37953559e-01 -8.34603965e-01 -5.90788364e-01 -6.30757153e-01 -4.48722064e-01 4.92908537e-01 4.70487595e-01 -3.74976337e-01 -1.70101619e+00 1.62443089e+00 5.08128703e-01 3.38036060e-01 -1.07442230e-01 7.85101295e-01 7.38487422e-01 6.13221645e-01 1.95092857e-01 -4.39483196e-01 9.47174072e-01 -1.13303244e+00 -9.83745277e-01 -9.90922675e-02 3.41716290e-01 -7.34879792e-01 4.71543074e-01 7.38979101e-01 -1.34987497e+00 -8.42297792e-01 -1.08660102e+00 3.31110001e-01 -4.87674445e-01 1.97239727e-01 3.56030434e-01 7.68957496e-01 -1.29159439e+00 1.04029214e+00 -9.50775623e-01 -4.09597009e-01 7.06299007e-01 7.08755076e-01 -2.71570496e-02 4.68724757e-01 -7.29248405e-01 7.12130189e-01 1.03777421e+00 1.79923028e-01 -1.09643352e+00 -5.66530108e-01 -5.39564669e-01 -1.69156149e-01 6.30107462e-01 -4.13807422e-01 1.26767063e+00 -1.39295566e+00 -1.27888381e+00 8.96481633e-01 -3.05218883e-02 -8.73792350e-01 9.27294254e-01 -4.51810718e-01 -3.86277139e-01 9.00779292e-02 4.04762700e-02 1.14151967e+00 1.03929043e+00 -1.43186271e+00 -1.08785093e+00 -3.07475954e-01 -2.79747784e-01 1.46760449e-01 -2.20439807e-01 6.72660843e-02 -7.94271767e-01 -7.21790314e-01 8.84053633e-02 -7.22807109e-01 -4.45944011e-01 -3.63882244e-01 -5.08380651e-01 -2.77442366e-01 1.12841523e+00 -7.53116429e-01 1.39189386e+00 -1.95347869e+00 2.52575129e-01 2.02932864e-01 1.86038762e-01 6.10186040e-01 9.76091251e-02 -6.85642734e-02 2.37914324e-02 3.19052339e-01 -6.52814865e-01 -4.95702058e-01 -2.54855186e-01 8.47725719e-02 1.54756993e-01 4.42590654e-01 5.55126846e-01 5.74889660e-01 -7.49001443e-01 -7.56190777e-01 2.02561244e-01 2.91793048e-01 -5.38707674e-01 -2.90327519e-02 -6.01625919e-01 3.91159385e-01 -1.94492295e-01 6.01563156e-01 8.59342813e-01 1.00295313e-01 -2.62302518e-01 -7.88512155e-02 -3.03921670e-01 -3.48510414e-01 -1.72184730e+00 1.72325373e+00 3.77308458e-01 6.87735915e-01 1.80017110e-02 -8.85029376e-01 8.67280066e-01 3.19723666e-01 5.25948882e-01 -2.86310852e-01 3.40704262e-01 4.46385294e-02 -3.03602424e-02 -6.61437511e-01 6.78185999e-01 2.13259205e-01 3.52066219e-01 -1.57807358e-02 4.25877780e-01 1.61946967e-01 8.34146559e-01 1.17626235e-01 9.95432019e-01 4.63699102e-01 -5.18115237e-02 -5.48060715e-01 6.18143499e-01 1.35104343e-01 5.71347058e-01 9.44097579e-01 -5.83409369e-01 8.01836193e-01 2.83980340e-01 -2.77390659e-01 -9.29185510e-01 -7.77909219e-01 -6.22273423e-02 9.87901688e-01 1.82196721e-01 -1.54811978e-01 -1.21082163e+00 -1.09018385e+00 -2.45505095e-01 5.06854355e-01 -8.09482694e-01 2.32716396e-01 -7.56100237e-01 -7.83406019e-01 3.80882829e-01 4.85641688e-01 4.93293524e-01 -1.27392459e+00 -7.28841662e-01 3.77401412e-01 1.00593910e-01 -1.19111264e+00 -4.47808743e-01 3.38682264e-01 -1.10830164e+00 -1.00516272e+00 -7.36522257e-01 -8.70096982e-01 6.63500249e-01 -2.05041215e-01 1.11291456e+00 2.01127920e-02 -4.21318829e-01 3.66910845e-02 -3.72428775e-01 -5.59943199e-01 -3.06371033e-01 2.99118549e-01 -4.88584116e-02 3.09297383e-01 9.22331288e-02 -2.34719232e-01 -6.52211010e-01 2.92820841e-01 -1.11996210e+00 -1.10583641e-01 4.68106151e-01 3.61826181e-01 7.18481123e-01 2.98226118e-01 6.01961613e-01 -1.19567358e+00 3.17254573e-01 -2.75671631e-01 -7.49318480e-01 1.06203474e-01 -6.13522172e-01 1.29368201e-01 3.32879752e-01 -3.24210048e-01 -1.10214663e+00 4.71024424e-01 -2.10440129e-01 -2.98498929e-01 -4.21172678e-01 -1.25521943e-01 -3.77683863e-02 -1.26833051e-01 7.41539061e-01 -1.06261462e-01 -3.99937391e-01 -5.51979244e-01 3.38780999e-01 3.92036885e-01 6.73040032e-01 -1.55720338e-01 7.97906160e-01 5.04147887e-01 -1.72420844e-01 -5.31379879e-01 -6.29504621e-01 -6.72845066e-01 -8.72809231e-01 -6.23709500e-01 1.09166169e+00 -5.46686411e-01 -4.33056355e-01 5.79391956e-01 -1.09772420e+00 -3.76173198e-01 -3.45692873e-01 1.87179089e-01 -4.72284973e-01 1.06581613e-01 -3.65476668e-01 -1.08279157e+00 -3.18932980e-01 -1.12002182e+00 8.65747511e-01 5.67429900e-01 -1.89343140e-01 -7.11581349e-01 7.08959438e-03 3.30964088e-01 2.35192701e-01 6.47476256e-01 1.65665716e-01 -8.00648928e-01 -6.59751177e-01 5.79367802e-02 -7.02585652e-02 5.31428218e-01 2.02630069e-02 3.97324175e-01 -1.04395998e+00 -2.48007938e-01 -1.30879670e-01 -3.60247903e-02 1.28804457e+00 7.66200364e-01 1.22205329e+00 -7.66369626e-02 -4.61283505e-01 5.38085341e-01 1.59312320e+00 5.68278611e-01 5.72738230e-01 4.25655961e-01 7.17684686e-01 6.01935208e-01 8.20761085e-01 2.56961793e-01 -1.40229657e-01 5.97691357e-01 4.62902993e-01 -2.27933317e-01 -1.60134584e-01 1.63668826e-01 1.13283247e-01 3.69210184e-01 -1.67835087e-01 -4.89110380e-01 -9.11188424e-01 9.01855648e-01 -2.03945279e+00 -8.75811458e-01 -2.85049886e-01 1.93330073e+00 7.02145934e-01 8.19093227e-01 4.62359339e-01 2.32431903e-01 9.45307493e-01 7.49676079e-02 -4.16844845e-01 -3.18858325e-01 -1.40889302e-01 2.15017557e-01 7.62821496e-01 6.23187602e-01 -1.52434063e+00 1.18279481e+00 6.14798880e+00 7.12857962e-01 -8.70103061e-01 2.63080776e-01 9.51777220e-01 -1.52139306e-01 2.01179296e-01 -2.48196363e-01 -9.48348582e-01 4.05349821e-01 5.04636049e-01 3.22120965e-01 4.10595350e-02 5.36369741e-01 3.87789518e-01 -3.77868980e-01 -9.61952806e-01 7.30628610e-01 1.63278595e-01 -1.38692331e+00 -7.61241019e-02 -5.92124045e-01 1.15234590e+00 9.05921757e-02 1.39462473e-02 4.17538993e-02 4.43838835e-02 -9.52682912e-01 9.87053454e-01 6.69690788e-01 2.19790772e-01 -8.38221788e-01 7.81556010e-01 2.76727557e-01 -9.92924273e-01 9.65401307e-02 9.53282565e-02 1.14586577e-01 2.98303068e-01 3.27172905e-01 -9.52907920e-01 4.47717011e-01 8.59816074e-01 5.66838384e-01 -8.81030440e-01 1.67091429e+00 1.03818670e-01 7.12671161e-01 -3.64317477e-01 5.10543957e-03 5.65380931e-01 2.27743506e-01 7.38332987e-01 1.67601335e+00 3.89595516e-02 -7.58312121e-02 1.59657538e-01 6.31772757e-01 7.80178932e-03 2.69142658e-01 -2.91873336e-01 2.00303778e-01 1.80186331e-02 1.38396096e+00 -1.58215332e+00 -5.34464300e-01 1.76103577e-01 9.35668945e-01 6.86783269e-02 2.97595203e-01 -8.99637282e-01 -1.12740077e-01 -1.95368845e-02 1.84902951e-01 6.77800179e-01 2.87094321e-02 -3.05201679e-01 -6.23069942e-01 1.88760787e-01 -6.81671500e-01 4.76574212e-01 -5.86698711e-01 -8.12075377e-01 8.31148446e-01 1.08079150e-01 -9.31076527e-01 1.05749778e-01 -5.09509861e-01 -4.16889995e-01 4.56551254e-01 -1.31645060e+00 -7.97228992e-01 -2.93264091e-01 4.28568006e-01 1.05613697e+00 -4.15257700e-02 1.49067992e-03 4.56878841e-01 -6.25432611e-01 3.28594178e-01 -1.20825425e-01 -2.05681263e-03 6.47042215e-01 -1.48031878e+00 3.12876016e-01 1.39798391e+00 2.96617866e-01 1.24741010e-01 1.01681578e+00 -9.15781319e-01 -5.85186958e-01 -1.38791442e+00 4.85190898e-01 -3.71554226e-01 2.34164119e-01 -2.21840858e-01 -8.69198680e-01 4.87782657e-01 4.87953186e-01 -1.71224345e-02 -3.61940525e-02 -2.56940156e-01 3.32830846e-01 -1.73685312e-01 -1.35370696e+00 3.96766216e-01 1.11135566e+00 2.79087961e-01 -3.76836419e-01 3.90709549e-01 7.46452808e-01 -6.26509905e-01 -3.73703420e-01 8.15667808e-01 6.77374899e-02 -1.09136367e+00 6.41354382e-01 -6.09835625e-01 3.11266094e-01 -7.85761654e-01 2.45582327e-01 -9.72924531e-01 -3.04718196e-01 -9.57023680e-01 -1.41097799e-01 1.46916473e+00 6.05273187e-01 -7.02074682e-03 1.05182958e+00 3.49512219e-01 -1.75591379e-01 -7.58383751e-01 -7.99721897e-01 -6.04583323e-01 -4.15411592e-01 -4.10254717e-01 2.08420381e-01 7.01225758e-01 -7.25709677e-01 -5.28689772e-02 -3.91875654e-01 2.93398917e-01 6.39672995e-01 -6.06920302e-01 4.95275795e-01 -1.07516432e+00 -1.40654668e-01 -5.50812662e-01 -6.37490332e-01 -6.95397556e-01 -1.90966666e-01 -6.63798511e-01 3.43162656e-01 -1.50187457e+00 2.25844115e-01 -3.38355422e-01 -6.17585480e-01 1.70328006e-01 -4.12803113e-01 4.93012577e-01 3.22226614e-01 -1.81688324e-01 -1.09691966e+00 -3.46273892e-02 1.15947306e+00 -2.44202882e-01 -4.81681198e-01 1.35181054e-01 -3.36633503e-01 8.93579245e-01 8.87627125e-01 -7.12295890e-01 -2.64059335e-01 -3.83423597e-01 -1.25897244e-01 -3.11296701e-01 3.13397706e-01 -1.47304702e+00 3.32928151e-01 3.89908031e-02 5.87556183e-01 -1.07887018e+00 -8.26610718e-03 -7.94668436e-01 1.66724384e-01 4.69502091e-01 -3.14424038e-01 1.92836806e-01 4.55217153e-01 4.89656925e-01 -1.29281223e-01 -5.97231388e-01 9.50821519e-01 -1.30384058e-01 -1.15308011e+00 2.73790359e-01 -2.26717606e-01 2.26321325e-01 1.33507037e+00 -3.15693676e-01 6.18618615e-02 1.66540463e-02 -1.07682681e+00 3.82697612e-01 1.73955679e-01 5.72699368e-01 3.93544883e-01 -9.02201056e-01 -6.70781732e-01 -1.20606035e-01 -1.83025911e-01 3.88899297e-02 2.65262455e-01 7.60888815e-01 -7.58863270e-01 5.79526275e-02 -2.09795117e-01 -8.93420935e-01 -1.50240970e+00 5.96675098e-01 4.61426854e-01 -3.05048555e-01 -4.44310606e-01 1.27417755e+00 -1.27559662e-01 8.23821202e-02 5.93339145e-01 -4.85456079e-01 -6.83015943e-01 2.48180926e-01 3.53427649e-01 3.98878217e-01 1.58590764e-01 -6.05429173e-01 -4.93482858e-01 5.66725969e-01 -3.04373920e-01 -2.57968932e-01 1.26381755e+00 -9.83988680e-03 1.25530303e-01 5.45493782e-01 9.26627338e-01 -1.66507632e-01 -1.61377585e+00 -5.21375239e-02 4.47574794e-01 -3.14889520e-01 -3.58214378e-02 -9.56979096e-01 -1.20684659e+00 3.89101416e-01 1.23617291e+00 3.07329923e-01 1.25592387e+00 -1.06716163e-01 4.66915518e-01 1.00701205e-01 -5.75115420e-02 -1.59803367e+00 -1.81245804e-02 2.73512602e-01 4.96630609e-01 -1.44443166e+00 -6.08004145e-02 -3.49029124e-01 -5.98599315e-01 9.85874832e-01 8.14078033e-01 -2.53528893e-01 3.81706953e-01 4.49628681e-01 8.56617317e-02 -2.66031325e-01 -4.60983694e-01 -4.98827964e-01 4.93060172e-01 3.81684244e-01 2.15609372e-01 -1.80158302e-01 -5.58781207e-01 2.33461961e-01 1.55774653e-01 2.03153491e-01 4.35079962e-01 1.00307512e+00 -6.85791194e-01 -8.29358459e-01 -3.72666836e-01 4.89025414e-01 -8.62484515e-01 1.51878983e-01 -3.10187727e-01 7.94729531e-01 7.71499217e-01 8.19799125e-01 -4.78072651e-02 -2.16877788e-01 3.53794396e-01 1.76340993e-02 3.67162436e-01 -5.44367433e-01 -9.11683142e-01 2.45149687e-01 9.98521969e-02 -5.10862172e-01 -9.13910747e-01 -8.31424057e-01 -1.12414181e+00 4.58030254e-01 -4.58163440e-01 9.01744701e-03 5.48181713e-01 1.06770813e+00 -5.79261780e-03 1.11779070e+00 3.48057687e-01 -8.60667288e-01 -1.84031180e-03 -8.51074219e-01 -3.13948184e-01 6.48443460e-01 4.43157792e-01 -6.56236708e-01 -4.20248322e-02 6.06748521e-01]	[9.070505142211914, -0.2355431318283081]
371cdef6-3a82-4ed2-9fe6-88c82abd475f	a-multilingual-modeling-method-for-span	2105.14880	null	https://arxiv.org/abs/2105.14880v1	https://arxiv.org/pdf/2105.14880v1.pdf	A Multilingual Modeling Method for Span-Extraction Reading Comprehension	Span-extraction reading comprehension models have made tremendous advances enabled by the availability of large-scale, high-quality training datasets. Despite such rapid progress and widespread application, extractive reading comprehension datasets in languages other than English remain scarce, and creating such a sufficient amount of training data for each language is costly and even impossible. An alternative to creating large-scale high-quality monolingual span-extraction training datasets is to develop multilingual modeling approaches and systems which can transfer to the target language without requiring training data in that language. In this paper, in order to solve the scarce availability of extractive reading comprehension training data in the target language, we propose a multilingual extractive reading comprehension approach called XLRC by simultaneously modeling the existing extractive reading comprehension training data in a multilingual environment using self-adaptive attention and multilingual attention. Specifically, we firstly construct multilingual parallel corpora by translating the existing extractive reading comprehension datasets (i.e., CMRC 2018) from the target language (i.e., Chinese) into different language families (i.e., English). Secondly, to enhance the final target representation, we adopt self-adaptive attention (SAA) to combine self-attention and inter-attention to extract the semantic relations from each pair of the target and source languages. Furthermore, we propose multilingual attention (MLA) to learn the rich knowledge from various language families. Experimental results show that our model outperforms the state-of-the-art baseline (i.e., RoBERTa_Large) on the CMRC 2018 task, which demonstrate the effectiveness of our proposed multi-lingual modeling approach and show the potentials in multilingual NLP tasks.	['Bangchang Liu', 'Dejie Chang', 'Bin Xu', 'Gaochen Wu']	2021-05-31	null	null	null	null	['multilingual-nlp']	['natural-language-processing']	[ 3.65307927e-01 6.84571266e-02 -1.39431298e-01 -3.63684505e-01 -1.21123254e+00 -6.65987849e-01 3.43393326e-01 9.33919623e-02 -6.97008550e-01 7.05520451e-01 5.38161457e-01 -8.32050085e-01 1.30925089e-01 -8.66870821e-01 -1.12557936e+00 -1.49927676e-01 6.06304169e-01 5.67837656e-01 -2.74328232e-01 -5.53246796e-01 -9.90617201e-02 -2.11449802e-01 -1.15689170e+00 1.75620258e-01 1.94310641e+00 6.99471056e-01 7.26759970e-01 4.67506140e-01 -3.79266620e-01 7.05238879e-01 -4.98721063e-01 -4.94770855e-01 -1.30216151e-01 -7.22623527e-01 -1.12473655e+00 -1.77326709e-01 5.03776968e-01 -2.82680839e-01 -1.27914414e-01 9.46362853e-01 1.92470223e-01 -9.53148678e-02 4.24198240e-01 -7.77836263e-01 -1.30993402e+00 1.03900719e+00 -5.91873527e-01 1.05685078e-01 5.57982624e-01 -6.81001097e-02 1.13484693e+00 -1.04945898e+00 3.69040996e-01 1.29335272e+00 1.80963710e-01 5.44262826e-01 -9.48232770e-01 -8.22575927e-01 4.61356848e-01 4.54748720e-01 -1.16689289e+00 -4.90869492e-01 8.35170686e-01 7.48961940e-02 1.15551615e+00 -9.45108756e-03 4.42748189e-01 1.28568280e+00 6.72896579e-02 1.24874437e+00 1.31408989e+00 -8.52686703e-01 -4.02568460e-01 -1.41826719e-01 3.38099658e-01 5.32226384e-01 -1.60426851e-02 -1.98233500e-01 -5.62539041e-01 5.27197301e-01 3.72053295e-01 -3.16785842e-01 -5.93916357e-01 2.10784808e-01 -1.48641777e+00 7.56451607e-01 2.67328680e-01 2.84204125e-01 -1.75916418e-01 -2.65849411e-01 5.15619516e-01 5.82051277e-01 4.45410967e-01 4.84274626e-01 -9.00688350e-01 2.43787635e-02 -4.84299362e-01 -1.16589321e-02 6.58355653e-01 1.50480115e+00 5.67725182e-01 -1.11833893e-01 4.94037159e-02 1.01785100e+00 2.84944743e-01 1.09525573e+00 7.45085955e-01 -4.77607846e-01 1.19531858e+00 6.43329382e-01 -3.20540220e-01 -5.68723261e-01 -3.21503133e-01 -5.85767806e-01 -8.57562602e-01 -5.50899506e-01 3.84168923e-01 -2.74101704e-01 -6.09660566e-01 2.07062626e+00 -7.57839810e-03 -2.97678530e-01 6.77588105e-01 5.94240189e-01 8.03288460e-01 9.48700190e-01 2.10344553e-01 -2.10354030e-01 1.48512113e+00 -1.39526832e+00 -6.99483275e-01 -4.94071156e-01 6.64217234e-01 -6.72604382e-01 1.58520353e+00 1.93714365e-01 -1.14754093e+00 -8.12973380e-01 -1.12183011e+00 -6.69028103e-01 -4.81424719e-01 2.18592003e-01 3.77206445e-01 1.65827334e-01 -6.08444214e-01 -1.56283095e-01 -7.67217934e-01 -3.27181667e-01 1.53983608e-01 -4.94433157e-02 -3.65836024e-01 -5.96246064e-01 -1.48693824e+00 1.03310847e+00 7.48760223e-01 1.00004971e-01 -8.34044814e-01 -7.32460320e-01 -1.18811810e+00 2.06195235e-01 5.69882274e-01 -5.31182945e-01 1.13626552e+00 -1.13850582e+00 -1.30332828e+00 7.69398093e-01 -2.75255650e-01 -2.33321369e-01 1.11554153e-01 -9.34306145e-01 -4.24032509e-01 -7.38381222e-02 2.94430852e-01 6.88487232e-01 5.45833349e-01 -1.06367218e+00 -7.44234502e-01 -4.76700366e-01 1.54270932e-01 6.44598305e-01 -3.71668667e-01 8.10647532e-02 -4.74404097e-01 -6.36874318e-01 -5.34987487e-02 -5.48525810e-01 1.98050514e-01 -9.11109030e-01 -3.77899796e-01 -5.53902388e-01 4.90200400e-01 -1.42743397e+00 1.15030086e+00 -1.69695473e+00 5.76185882e-01 -1.60145611e-01 6.70032054e-02 2.91563421e-01 -5.98003626e-01 4.28869992e-01 5.15182354e-02 1.75385535e-01 -3.76094609e-01 -3.47544909e-01 -3.64272110e-02 2.57245094e-01 -3.44624609e-01 -8.26911256e-02 5.48154354e-01 1.26952279e+00 -1.19189370e+00 -4.60388541e-01 -4.53825742e-02 2.13574782e-01 -3.72416735e-01 5.68577528e-01 -3.94585848e-01 7.96102643e-01 -4.32177216e-01 6.28698885e-01 5.45218825e-01 -6.58427179e-02 1.91624925e-01 4.64699827e-02 -7.85561129e-02 6.57954931e-01 -3.60742807e-01 2.17635894e+00 -1.18628526e+00 4.10825610e-01 -2.53729671e-01 -9.16905046e-01 8.86686444e-01 2.21256852e-01 -1.66391768e-02 -1.11522245e+00 1.28631622e-01 5.23907483e-01 2.19456732e-01 -4.59726810e-01 5.35795212e-01 1.21360486e-02 -5.44373512e-01 6.48743570e-01 3.27068299e-01 5.37972227e-02 3.31257761e-01 3.27561110e-01 8.18055630e-01 4.21272993e-01 3.36462080e-01 -2.62233078e-01 9.22117352e-01 -1.69510171e-02 5.65818012e-01 3.91479403e-01 1.20851949e-01 2.13587701e-01 2.64930159e-01 9.33768675e-02 -8.98479819e-01 -1.02836275e+00 3.66923958e-02 1.49386668e+00 1.30754367e-01 -5.15581369e-01 -1.06689978e+00 -8.83335114e-01 -3.97264451e-01 9.54414189e-01 -2.47313768e-01 -1.21211559e-01 -1.01341355e+00 -4.96126473e-01 6.56993926e-01 8.43947411e-01 8.58746350e-01 -1.28277469e+00 -3.10782462e-01 4.26920593e-01 -7.75659621e-01 -1.35268807e+00 -5.87258339e-01 -8.42543021e-02 -4.89431471e-01 -1.05013835e+00 -5.58115780e-01 -1.24735498e+00 5.57412505e-01 3.18496376e-02 1.56223309e+00 3.50163877e-02 2.79598802e-01 3.09415430e-01 -6.85757458e-01 -5.25959432e-01 -4.48752642e-01 8.21209550e-01 -1.62909269e-01 -2.55656034e-01 6.15759850e-01 -2.49738738e-01 -8.37935433e-02 -1.15745207e-02 -8.57264340e-01 5.55358589e-01 8.13884735e-01 8.92098248e-01 7.08720326e-01 -3.04367423e-01 1.21138728e+00 -9.57117677e-01 8.97588313e-01 -6.26243293e-01 -5.67057610e-01 9.19863045e-01 -4.96239066e-01 4.03771736e-02 1.14519215e+00 -3.09907705e-01 -1.37707043e+00 -2.95705229e-01 -3.53460938e-01 -3.74070108e-02 -8.22429657e-02 1.09755826e+00 -8.38974357e-01 5.83485305e-01 2.35113949e-01 5.35809457e-01 -2.66762137e-01 -6.69167817e-01 6.84902072e-01 8.12655628e-01 8.45424414e-01 -9.28242862e-01 6.21236682e-01 -4.24151093e-01 -6.07377231e-01 -7.47787476e-01 -1.27757347e+00 -2.02958837e-01 -8.87211740e-01 1.74429432e-01 1.03326952e+00 -1.12403524e+00 -2.72555202e-01 9.97165442e-01 -1.40119827e+00 -3.26738417e-01 3.46748233e-02 3.29519272e-01 -4.98321891e-01 1.55147821e-01 -6.15815997e-01 -3.43963891e-01 -6.25636280e-01 -1.21256828e+00 9.90670145e-01 2.60060549e-01 4.80439179e-02 -1.25128984e+00 -4.20598052e-02 7.67610490e-01 4.67668146e-01 -1.08225003e-01 1.52747226e+00 -7.32109547e-01 -7.00518191e-01 1.46492615e-01 -4.99355137e-01 6.08362854e-01 9.51862633e-02 -6.00238264e-01 -6.20275080e-01 -3.01731914e-01 -1.52353114e-02 -9.37309504e-01 5.82498252e-01 -9.39209312e-02 1.10740674e+00 -2.92956710e-01 6.03918470e-02 5.64972520e-01 1.30834651e+00 -3.42876352e-02 4.88475144e-01 2.21967638e-01 1.08052397e+00 7.93843806e-01 5.87320924e-01 -3.36232901e-01 1.20534253e+00 2.65772045e-01 7.28092715e-02 -1.51775301e-01 -3.04577351e-01 -7.37000585e-01 4.66928959e-01 1.91445374e+00 6.01010174e-02 -3.67033571e-01 -1.03474510e+00 7.01459229e-01 -1.62233961e+00 -2.96788603e-01 -8.66744388e-03 2.00823522e+00 1.24995053e+00 -1.97249740e-01 -4.41929072e-01 -3.45575452e-01 3.04940253e-01 1.87630467e-02 -7.57898211e-01 -3.24676424e-01 -3.89631361e-01 5.53703666e-01 2.02563733e-01 6.20963275e-01 -8.91076744e-01 1.53073967e+00 4.61649179e+00 8.54200423e-01 -9.16911840e-01 3.31868500e-01 3.82186174e-01 3.44383925e-01 -6.57104313e-01 1.71596054e-02 -9.37933624e-01 3.70138198e-01 1.01444924e+00 -3.69972885e-01 5.24554849e-01 3.91158521e-01 -1.33799553e-01 8.66037607e-02 -1.14461315e+00 7.09048629e-01 4.45598513e-01 -6.55812442e-01 2.51192391e-01 -2.58245915e-01 7.20252931e-01 1.14139810e-01 -1.67990357e-01 9.72105324e-01 2.63511330e-01 -1.19748771e+00 5.47487795e-01 3.40491533e-01 1.04811537e+00 -9.30550754e-01 8.56491029e-01 7.31461465e-01 -1.16530359e+00 1.40690103e-01 -3.79425943e-01 9.53729004e-02 3.89364421e-01 8.05240571e-02 -2.17904598e-01 8.94051135e-01 5.41901767e-01 8.94375384e-01 -7.67496884e-01 3.51417184e-01 -8.85482788e-01 8.64093363e-01 -6.36069030e-02 1.94639303e-02 2.47549385e-01 -3.09525013e-01 3.15984428e-01 9.84215379e-01 2.43807152e-01 -5.72304334e-03 2.95211524e-01 8.68560612e-01 -5.52492917e-01 7.30041444e-01 -3.83985639e-01 -1.64718777e-01 4.30268317e-01 9.23737645e-01 4.70752493e-02 -4.02099699e-01 -9.32779789e-01 9.87467110e-01 9.59454179e-01 3.83663267e-01 -6.52972758e-01 -5.77153981e-01 9.20162797e-02 -2.54499108e-01 -6.44358173e-02 -3.24281573e-01 -1.35513932e-01 -1.66523170e+00 3.38334173e-01 -1.50771248e+00 3.45300317e-01 -6.69681370e-01 -1.39485013e+00 8.14745545e-01 6.25523180e-02 -8.24033320e-01 -3.45233828e-01 -5.52699864e-01 -4.80101526e-01 1.39220929e+00 -1.77431440e+00 -1.77927446e+00 -1.80981964e-01 6.92168713e-01 1.05053556e+00 -3.13725561e-01 8.97717297e-01 1.85822919e-01 -5.87800682e-01 8.75569701e-01 -3.10947783e-02 3.50291818e-01 7.91749358e-01 -1.23154986e+00 5.07929325e-01 1.02722180e+00 1.25125781e-01 8.45006287e-01 1.34149231e-02 -6.05176687e-01 -1.45069194e+00 -1.13877439e+00 1.44647515e+00 -4.21366304e-01 9.30375099e-01 -5.17676353e-01 -1.30026782e+00 1.14522231e+00 8.77333701e-01 -7.12727785e-01 6.94095671e-01 5.32943964e-01 -4.26178604e-01 -4.61718477e-02 -4.80278581e-01 7.38424718e-01 1.02904606e+00 -7.39877641e-01 -1.12624097e+00 2.53717810e-01 1.15967059e+00 -4.20406699e-01 -1.07728589e+00 6.03875041e-01 3.24250400e-01 -3.26877117e-01 5.83684683e-01 -6.36949837e-01 7.02849805e-01 -3.19412574e-02 -1.27145082e-01 -1.70416224e+00 1.38257474e-01 -2.40861356e-01 2.35849321e-02 1.47329259e+00 7.70516336e-01 -7.30603755e-01 1.14599820e-02 2.54623413e-01 -4.87806290e-01 -8.30960631e-01 -6.79893672e-01 -6.14362955e-01 9.21692789e-01 -4.41691458e-01 7.53820002e-01 8.59334886e-01 1.29857212e-01 1.13888323e+00 -2.17313319e-01 9.50907990e-02 4.85455364e-01 3.56112719e-01 6.45330727e-01 -7.79741764e-01 -2.29983434e-01 -2.48644471e-01 3.27603757e-01 -1.54466641e+00 9.02223468e-01 -1.21258616e+00 1.82732448e-01 -1.43724513e+00 2.85823464e-01 -4.91590798e-01 -3.52582693e-01 5.77235103e-01 -7.76666880e-01 -3.38453114e-01 1.52302235e-01 4.65107076e-02 -6.84093893e-01 9.34611857e-01 1.50011051e+00 -2.58096725e-01 -7.67363310e-02 -3.94909531e-01 -9.27397251e-01 5.53770006e-01 9.36528027e-01 -1.22016683e-01 -5.59103251e-01 -1.16555762e+00 2.40570426e-01 1.32743686e-01 -3.66464369e-02 -6.17359221e-01 1.68665081e-01 1.01006322e-01 1.85072020e-01 -7.33128786e-01 -1.73941135e-01 -5.17375767e-01 -6.64855897e-01 1.87272206e-02 -5.07813275e-01 3.48641813e-01 1.93923846e-01 2.41767347e-01 -5.53016007e-01 -3.06167215e-01 4.52397674e-01 -2.03299627e-01 -7.30129659e-01 1.43080994e-01 -9.35368016e-02 6.78744316e-01 6.37413800e-01 4.17101026e-01 -6.78734064e-01 -3.57657701e-01 -3.55358958e-01 8.18193913e-01 3.55270028e-01 9.50658560e-01 5.10083497e-01 -1.24038589e+00 -1.20121610e+00 4.26609963e-01 4.75158334e-01 4.16233867e-01 1.82588294e-01 6.61763668e-01 -2.77104676e-01 6.15914822e-01 -2.61519283e-01 -3.02713484e-01 -8.92957628e-01 5.56003094e-01 9.56912562e-02 -6.99059665e-01 -4.01267558e-01 5.92551649e-01 4.07120645e-01 -1.06228340e+00 -9.44500715e-02 -3.13369572e-01 -3.54840517e-01 -2.72683740e-01 5.38560808e-01 1.74677327e-01 3.57100740e-02 -7.70654023e-01 -1.09549247e-01 6.20855391e-01 -4.67929035e-01 -1.25328228e-01 1.06322157e+00 -4.15378749e-01 -3.05050254e-01 2.95419782e-01 1.13094532e+00 3.62363756e-01 -9.25104022e-01 -4.95090574e-01 1.32400736e-01 7.53669739e-02 -2.03688070e-01 -1.09289265e+00 -8.77122283e-01 1.16535497e+00 6.12412021e-03 -4.59430248e-01 1.35649586e+00 2.15156749e-01 1.32545769e+00 4.31137264e-01 4.09816384e-01 -1.13456213e+00 -2.12698311e-01 1.13588905e+00 1.13214886e+00 -1.46748090e+00 -4.86080080e-01 -2.97226876e-01 -6.65730596e-01 7.84346938e-01 9.85557199e-01 1.15374945e-01 2.76650369e-01 -1.13965437e-01 3.42516810e-01 3.93434754e-03 -6.67313457e-01 -2.74320036e-01 4.80792522e-01 5.45956969e-01 7.44270861e-01 2.93364316e-01 -4.67401862e-01 1.00781131e+00 -5.49002111e-01 -3.69029731e-01 2.09917113e-01 7.14577675e-01 -1.38216287e-01 -1.38737440e+00 -1.42615378e-01 3.34631652e-01 -4.65069085e-01 -7.17552423e-01 -2.76512623e-01 9.00139391e-01 1.17823489e-01 1.15043724e+00 -1.05607674e-01 -8.93542841e-02 4.03267443e-01 2.20184460e-01 5.39975464e-01 -6.90738261e-01 -4.19192642e-01 -4.20618020e-02 6.62763864e-02 -3.00274312e-01 -2.92312235e-01 -5.58412969e-01 -1.19310224e+00 8.06772783e-02 -3.19079787e-01 9.61808711e-02 5.55276632e-01 1.26684034e+00 1.58957154e-01 6.99500978e-01 4.10912901e-01 -2.14095712e-01 -4.48371619e-01 -1.35487103e+00 -1.51173964e-01 3.90106380e-01 6.91598728e-02 -1.91872880e-01 -7.98272863e-02 1.64860576e-01]	[11.007255554199219, 9.190930366516113]
1c9b8bac-8b56-43c6-b99a-1d1aa15581e3	punctuation-restoration-using-transformer	null	null	https://aclanthology.org/2020.wnut-1.18	https://aclanthology.org/2020.wnut-1.18.pdf	Punctuation Restoration using Transformer Models for High-and Low-Resource Languages	Punctuation restoration is a common post-processing problem for Automatic Speech Recognition (ASR) systems. It is important to improve the readability of the transcribed text for the human reader and facilitate NLP tasks. Current state-of-art address this problem using different deep learning models. Recently, transformer models have proven their success in downstream NLP tasks, and these models have been explored very little for the punctuation restoration problem. In this work, we explore different transformer based models and propose an augmentation strategy for this task, focusing on high-resource (English) and low-resource (Bangla) languages. For English, we obtain comparable state-of-the-art results, while for Bangla, it is the first reported work, which can serve as a strong baseline for future work. We have made our developed Bangla dataset publicly available for the research community.	['Firoj Alam', 'Akib Khan', 'Tanvirul Alam']	null	null	null	null	emnlp-wnut-2020-11	['punctuation-restoration']	['natural-language-processing']	[ 1.61385298e-01 8.50641280e-02 6.37188042e-03 -3.54956120e-01 -1.16911507e+00 -4.53582346e-01 6.12714291e-01 6.41011745e-02 -6.72157645e-01 7.22373724e-01 6.49423003e-01 -7.11613774e-01 3.73365849e-01 -3.44185859e-01 -5.08958220e-01 -4.74999845e-01 3.04350555e-01 4.59771603e-01 7.08416989e-03 -3.01325023e-01 7.87582472e-02 6.05946660e-01 -1.10521603e+00 7.15693951e-01 7.79450834e-01 4.53998268e-01 5.95125675e-01 4.63297427e-01 -2.71180272e-01 7.99322724e-01 -7.63554752e-01 -5.11066794e-01 -5.73055074e-02 -2.71324039e-01 -1.24167550e+00 -1.77951828e-02 7.10073113e-02 -2.19187997e-02 -6.13274574e-01 7.54385352e-01 6.96314156e-01 6.28174767e-02 3.63016069e-01 -6.64968491e-01 -8.65838110e-01 1.14907050e+00 -1.91921443e-01 4.56825644e-01 1.72843009e-01 -1.32024661e-01 1.02886140e+00 -1.09447765e+00 5.54647803e-01 1.10908771e+00 4.91778165e-01 7.40102351e-01 -7.36710787e-01 -3.92918795e-01 -3.28205116e-02 4.07606483e-01 -1.33546519e+00 -8.50747168e-01 8.00698102e-01 1.69217438e-01 1.47572863e+00 2.61949271e-01 6.97042122e-02 1.35736728e+00 -1.45126134e-01 1.33637929e+00 1.36695123e+00 -7.84727037e-01 -7.95231164e-02 1.02016762e-01 -5.85622415e-02 3.30150425e-01 -3.13301951e-01 -3.82479280e-01 -6.91440880e-01 4.16569263e-01 3.77641410e-01 -3.07300240e-01 -4.75191981e-01 4.93826956e-01 -1.28429794e+00 6.66321337e-01 4.55877138e-03 6.66868687e-01 -5.09380162e-01 -3.16810787e-01 4.15834993e-01 4.74792957e-01 7.52241254e-01 4.46467102e-01 -7.21389234e-01 -5.82195342e-01 -1.20027101e+00 6.66987598e-02 7.97655940e-01 9.94991064e-01 4.40142423e-01 2.48678491e-01 -3.45706314e-01 1.41292000e+00 2.69483626e-01 3.84375811e-01 7.79230356e-01 -3.49919558e-01 8.27938616e-01 2.07826808e-01 -7.23587871e-02 -3.08728486e-01 -2.48782292e-01 -4.38769877e-01 -8.02823126e-01 -3.12598109e-01 4.37028766e-01 -2.05525309e-01 -1.16844249e+00 1.34320235e+00 -1.16918229e-01 -5.58037274e-02 4.18529391e-01 7.60430336e-01 8.81862521e-01 1.18108869e+00 -7.68847913e-02 -2.47703955e-01 1.45829082e+00 -1.34971428e+00 -1.05680573e+00 -4.37385470e-01 6.28871441e-01 -1.39363956e+00 1.28893518e+00 5.27492881e-01 -1.36364400e+00 -4.20111746e-01 -9.36867952e-01 -4.99084473e-01 -6.32990777e-01 7.09689736e-01 1.69338316e-01 7.32993662e-01 -1.35262430e+00 6.09166920e-01 -8.79462779e-01 -6.06129467e-01 1.83426455e-01 -1.55292163e-02 -3.09539765e-01 -1.67805687e-01 -1.42816603e+00 1.07335162e+00 2.28190705e-01 2.38183677e-01 -7.29118705e-01 -4.73080337e-01 -7.26434708e-01 1.43272663e-02 2.66010731e-01 -1.29366487e-01 1.64529467e+00 -6.05143726e-01 -1.84762847e+00 1.07900786e+00 -5.24223804e-01 -6.63421154e-01 4.15504366e-01 -4.65746105e-01 -6.05070412e-01 -1.54577538e-01 -2.87655830e-01 3.36061150e-01 6.47493362e-01 -9.36986506e-01 -6.76018715e-01 -3.52607250e-01 -3.44838262e-01 2.06001908e-01 -3.26365054e-01 7.23604202e-01 -5.23766875e-01 -1.04476070e+00 -6.18477203e-02 -8.89157236e-01 -1.44139543e-01 -6.52621984e-01 -3.36105078e-01 -5.88208497e-01 7.77869225e-01 -1.30263340e+00 1.44901264e+00 -2.08748055e+00 -1.12318359e-01 -2.86662877e-01 -4.56631362e-01 9.05358553e-01 -3.33758831e-01 7.84288585e-01 -1.27746519e-02 4.03609127e-01 -3.12232167e-01 -8.21566641e-01 -1.00361988e-01 2.09971398e-01 -5.14109075e-01 2.30729237e-01 7.13881135e-01 8.68919134e-01 -6.97627962e-01 -4.91214134e-02 1.80044800e-01 6.90861762e-01 2.02115085e-02 3.01194280e-01 -6.07708190e-03 1.84319764e-01 -7.66101703e-02 4.96874213e-01 5.11509776e-01 2.83332437e-01 -8.88615251e-02 1.95436031e-01 -4.48998690e-01 1.34288406e+00 -8.00044358e-01 1.59926307e+00 -8.71120155e-01 8.18654776e-01 1.25501648e-01 -9.24565494e-01 1.05693269e+00 6.78342223e-01 9.46713053e-03 -7.74616003e-01 5.50826229e-02 4.30527538e-01 5.74705303e-02 -2.59922802e-01 9.70049560e-01 -1.12412326e-01 2.03136429e-01 1.86055779e-01 1.18891768e-01 -2.28967384e-01 1.80277109e-01 2.18147263e-02 1.07463765e+00 3.40168327e-02 2.91161776e-01 -1.43206924e-01 5.90768039e-01 -3.59752700e-02 2.47268543e-01 6.43479168e-01 -1.34934798e-01 1.09729850e+00 9.87409279e-02 6.37146831e-02 -1.28331554e+00 -8.44665468e-01 -5.94597608e-02 9.33895469e-01 -6.15076363e-01 -5.25145054e-01 -8.89044106e-01 -5.93271136e-01 -4.56033647e-01 1.01759338e+00 -1.75652370e-01 1.45931140e-01 -9.02636349e-01 -5.99305511e-01 9.53670561e-01 5.78408062e-01 3.94366622e-01 -1.44003987e+00 2.29996607e-01 4.86459792e-01 -7.07567275e-01 -1.52141893e+00 -5.04113972e-01 3.40397447e-01 -6.53762639e-01 -4.28140193e-01 -1.06957424e+00 -1.20367694e+00 8.38332400e-02 3.82902324e-01 9.04312849e-01 -1.37652218e-01 1.54642507e-01 -1.10166203e-02 -7.65320420e-01 -6.21813416e-01 -8.21592331e-01 3.77452821e-01 -1.91724245e-02 1.12514468e-02 3.97767752e-01 -2.73821890e-01 -1.41360745e-01 1.37718096e-01 -8.32100511e-01 -1.06798910e-01 6.07147694e-01 6.02975965e-01 6.39522791e-01 -1.47264928e-01 7.33984709e-01 -7.66189754e-01 7.43756235e-01 -2.12629125e-01 -2.32954606e-01 1.75687194e-01 -2.58961320e-01 7.95273781e-02 9.84814584e-01 -2.14292735e-01 -1.33286464e+00 -1.92971408e-01 -9.15390432e-01 -1.85212806e-01 -2.97103941e-01 8.76845121e-01 -4.27457482e-01 2.61613518e-01 1.88015074e-01 5.20410061e-01 -4.01539534e-01 -1.04522121e+00 3.66185129e-01 1.38663709e+00 5.80314875e-01 -2.80563712e-01 4.23899025e-01 1.06267318e-01 -4.33164626e-01 -1.46604419e+00 -8.78064811e-01 -9.11619306e-01 -6.67183936e-01 2.27050141e-01 6.58565998e-01 -1.00666523e+00 -3.91879678e-02 8.92895460e-01 -1.39886403e+00 -4.34353858e-01 -1.65355593e-01 3.54644299e-01 -3.27201366e-01 6.83744609e-01 -7.64434218e-01 -7.37025559e-01 -4.65676337e-01 -1.12966990e+00 1.09566724e+00 7.18330666e-02 -7.90208131e-02 -8.76520574e-01 4.03240249e-02 6.15452707e-01 5.28877079e-01 -7.04160392e-01 5.71284533e-01 -8.95360053e-01 -2.46495798e-01 3.62488702e-02 -6.11520559e-02 7.05246389e-01 2.50561953e-01 -1.17369480e-01 -1.23859155e+00 -2.09301859e-01 -6.16112947e-02 -2.67974675e-01 1.03359520e+00 1.22714072e-01 1.13745856e+00 -4.32823539e-01 2.00886503e-02 3.65249634e-01 8.61117780e-01 9.22148824e-02 1.14839053e+00 4.86543149e-01 5.76246858e-01 5.30829251e-01 6.48237407e-01 2.83248365e-01 5.97220957e-01 6.04827166e-01 -1.44992381e-01 -2.74855077e-01 -7.17249811e-01 -3.04318637e-01 6.49906397e-01 1.57949793e+00 1.93107471e-01 -7.98071027e-01 -1.17070270e+00 1.12050736e+00 -1.61824965e+00 -7.84480155e-01 -2.51899302e-01 2.03697991e+00 1.28780162e+00 -7.54231140e-02 -8.83870572e-02 2.27848202e-01 6.73462093e-01 3.71953040e-01 2.27094665e-01 -8.06381404e-01 -3.30190867e-01 5.59992611e-01 1.75804883e-01 5.27322233e-01 -1.22920549e+00 1.50434887e+00 5.88773775e+00 1.17545033e+00 -1.32080364e+00 2.44989038e-01 5.41941047e-01 2.72232473e-01 -1.99954540e-01 1.44664114e-02 -1.10415900e+00 4.24065828e-01 1.48844862e+00 -5.26184663e-02 3.38883489e-01 6.47840798e-01 5.96560597e-01 1.58776715e-01 -9.68516588e-01 9.94333267e-01 9.27505940e-02 -1.13437474e+00 5.30368276e-02 -2.16666251e-01 5.39644063e-01 6.04616940e-01 -4.23773676e-02 4.90167588e-01 1.86524898e-01 -1.07193375e+00 5.91394663e-01 1.19863249e-01 5.65814435e-01 -8.29947591e-01 9.39197898e-01 2.16516107e-01 -9.89536345e-01 4.23105896e-01 -3.31168801e-01 -3.22060585e-02 3.65648091e-01 5.39126337e-01 -1.13933051e+00 5.84624112e-01 6.54335737e-01 5.92992961e-01 -6.06235802e-01 1.01177537e+00 -6.16703093e-01 1.36642277e+00 -3.17879140e-01 -1.45136416e-01 6.35969937e-01 -6.85654581e-02 6.08118653e-01 1.80428410e+00 2.96862215e-01 -2.20270474e-02 4.72077951e-02 4.92317706e-01 -3.99978369e-01 5.07728815e-01 -2.58228272e-01 -5.09413898e-01 3.84636015e-01 1.21493459e+00 -6.47443354e-01 -2.11801082e-01 -5.56523979e-01 1.22161090e+00 6.81543231e-01 2.93545038e-01 -4.28367317e-01 -6.24660015e-01 6.97977722e-01 -8.18276405e-03 4.52247113e-01 -5.98032594e-01 -3.59127760e-01 -1.26077223e+00 2.59468794e-01 -1.18493414e+00 1.91958606e-01 -7.95158446e-01 -1.28457415e+00 8.08691561e-01 -4.48427945e-01 -8.44580054e-01 -1.21681817e-01 -6.38098300e-01 -5.54045439e-01 1.11228931e+00 -2.17617655e+00 -1.35724795e+00 2.88830966e-01 4.10457790e-01 1.08510756e+00 -2.98385978e-01 7.05926538e-01 3.80528837e-01 -5.67459166e-01 6.43574655e-01 4.86539513e-01 3.04164499e-01 8.46729577e-01 -1.30128920e+00 1.00129974e+00 1.38052917e+00 5.61767817e-01 5.49563169e-01 5.06610811e-01 -5.12950897e-01 -1.22489738e+00 -1.26486778e+00 1.77373087e+00 -3.17811310e-01 8.85619402e-01 -4.62037623e-01 -1.25046682e+00 6.21679783e-01 5.83216548e-01 -3.47773969e-01 4.73880112e-01 2.44309410e-01 -3.73872705e-02 4.36820239e-02 -7.94814706e-01 7.89736688e-01 8.79252553e-01 -7.82187879e-01 -8.42851639e-01 2.25328460e-01 7.82735348e-01 -3.35850656e-01 -6.13496304e-01 1.37810796e-01 2.05289517e-02 -5.42392612e-01 4.92325604e-01 -4.09489155e-01 3.35777968e-01 -2.35617131e-01 4.91272472e-02 -1.69902050e+00 -1.96104437e-01 -9.65310276e-01 1.20761059e-01 1.74156201e+00 6.54573321e-01 -4.49521005e-01 4.22239035e-01 1.66763499e-01 -7.56076694e-01 -5.30037761e-01 -9.84273255e-01 -9.26874399e-01 3.40291560e-01 -6.02262437e-01 4.32592422e-01 7.99884081e-01 1.40067205e-01 5.18806458e-01 -4.27475750e-01 2.32707202e-01 -1.04830470e-02 -2.20461383e-01 4.02300805e-01 -6.51220560e-01 6.82019889e-02 -3.43889058e-01 -9.90975872e-02 -1.13170683e+00 3.90454710e-01 -1.04504323e+00 3.26748550e-01 -1.77342105e+00 7.72284269e-02 -3.23355228e-01 -1.65691435e-01 9.29868102e-01 -2.45578468e-01 2.01517090e-01 2.68711358e-01 -6.43788576e-02 -2.74336785e-01 7.68227756e-01 8.80325496e-01 -7.71442279e-02 -2.45863840e-01 1.43051013e-01 -5.27414680e-01 3.96912575e-01 1.23394310e+00 -4.44168180e-01 -4.69639152e-02 -7.91322172e-01 -1.16882473e-01 -1.86404198e-01 -6.68066666e-02 -7.91500330e-01 5.43690808e-02 6.82126805e-02 6.23723902e-02 -8.86508226e-01 2.21195504e-01 -4.31793094e-01 -5.56995928e-01 -5.29820612e-03 -3.04227024e-01 2.65583932e-01 4.54385996e-01 -4.89651598e-02 -5.04939914e-01 -3.72702271e-01 8.21739197e-01 -5.31983860e-02 -6.96172416e-01 1.96120575e-01 -1.02441990e+00 3.23475897e-01 3.80633652e-01 5.21301217e-02 -1.39173180e-01 -4.98036683e-01 -4.51184124e-01 5.92585420e-03 1.88702166e-01 8.44491363e-01 7.29190588e-01 -8.47058475e-01 -1.21865189e+00 1.56908303e-01 -7.36661330e-02 -2.05898970e-01 -1.11488551e-02 6.47324383e-01 -4.81151789e-01 7.25402296e-01 9.08375680e-02 -9.82218012e-02 -1.36268461e+00 4.92937714e-01 1.04728928e-02 -4.20292318e-01 -4.52890724e-01 7.82062829e-01 -2.82251060e-01 -5.77483296e-01 3.77783269e-01 -3.63628864e-01 -3.39236021e-01 3.11432909e-02 7.57407367e-01 3.82793248e-01 7.78989315e-01 -8.38303983e-01 -4.48919475e-01 -5.00134341e-02 -5.70690095e-01 -3.26208711e-01 1.44536412e+00 -4.04126167e-01 -8.04396123e-02 3.60461295e-01 1.10175705e+00 3.69790226e-01 -6.40874863e-01 -3.32526952e-01 2.61399984e-01 -1.02565981e-01 4.05097872e-01 -1.00829756e+00 -7.34800875e-01 1.26710522e+00 2.13705763e-01 1.25796676e-01 1.01921391e+00 -7.24417120e-02 9.63163674e-01 5.61932504e-01 2.56752353e-02 -1.41256726e+00 -1.53531045e-01 1.27736712e+00 1.07850301e+00 -1.29159665e+00 -3.87493104e-01 -3.26923370e-01 -7.85657227e-01 1.13918471e+00 2.57629722e-01 2.73507893e-01 4.11103159e-01 4.02424663e-01 5.07235944e-01 2.91591436e-01 -5.68021953e-01 -5.56456327e-01 2.93408245e-01 7.05875039e-01 1.00877666e+00 -2.13540671e-03 -6.11007094e-01 6.68547630e-01 -5.13867676e-01 -1.90446272e-01 7.43971407e-01 8.70506048e-01 -2.23839000e-01 -1.48585629e+00 -2.24394053e-01 2.81422697e-02 -9.95834708e-01 -7.99304843e-01 -7.04321086e-01 5.46172559e-01 -4.88845795e-01 1.39382243e+00 -1.79014578e-01 7.50585496e-02 3.37937206e-01 3.02752405e-01 2.83701658e-01 -7.62323201e-01 -7.01225877e-01 1.04713544e-01 4.72538382e-01 -2.80402452e-01 -1.82802483e-01 -8.71600449e-01 -1.18755174e+00 -5.45603409e-02 -2.56547809e-01 1.32615253e-01 8.59305203e-01 9.53015447e-01 3.53647053e-01 4.18949842e-01 6.71935678e-01 -5.47014415e-01 -5.36481440e-01 -1.29438710e+00 -4.98147786e-01 2.22406670e-01 3.31516653e-01 1.38661731e-02 -1.37188852e-01 1.52943507e-01]	[14.284995079040527, 7.126537799835205]
ccbe1b10-5ee3-4348-bfd6-78f4d53bc565	leveraging-triplet-loss-for-unsupervised	2304.06403	null	https://arxiv.org/abs/2304.06403v1	https://arxiv.org/pdf/2304.06403v1.pdf	Leveraging triplet loss for unsupervised action segmentation	In this paper, we propose a novel fully unsupervised framework that learns action representations suitable for the action segmentation task from the single input video itself, without requiring any training data. Our method is a deep metric learning approach rooted in a shallow network with a triplet loss operating on similarity distributions and a novel triplet selection strategy that effectively models temporal and semantic priors to discover actions in the new representational space. Under these circumstances, we successfully recover temporal boundaries in the learned action representations with higher quality compared with existing unsupervised approaches. The proposed method is evaluated on two widely used benchmark datasets for the action segmentation task and it achieves competitive performance by applying a generic clustering algorithm on the learned representations.	['M. Dimiccoli', 'B. Tura', 'E. Bueno-Benito']	2023-04-13	null	null	null	null	['metric-learning', 'video-understanding', 'action-segmentation', 'metric-learning']	['computer-vision', 'computer-vision', 'computer-vision', 'methodology']	[ 8.46807837e-01 -2.79790815e-02 -4.94564265e-01 -5.15624106e-01 -8.86236668e-01 -2.83048838e-01 6.11349761e-01 -2.96503287e-02 -6.88319206e-01 5.46989083e-01 1.56494528e-01 3.27358544e-01 -5.36283195e-01 -3.92800182e-01 -5.15096486e-01 -7.90577352e-01 4.62302975e-02 6.41259134e-01 5.47006488e-01 2.68511802e-01 4.87324119e-01 2.87297904e-01 -1.69052231e+00 3.38812947e-01 9.02491808e-01 1.32746565e+00 -4.04228605e-02 4.47951972e-01 1.94477141e-01 1.07228720e+00 -3.34487110e-01 -7.79714212e-02 5.21455526e-01 -6.74667776e-01 -1.15905213e+00 6.75966859e-01 4.65986758e-01 -1.27333283e-01 -3.74273777e-01 7.38673449e-01 6.22602813e-02 6.49314940e-01 8.59728217e-01 -1.15887451e+00 -1.69706479e-01 2.88848013e-01 -4.51726854e-01 2.06212416e-01 2.91098654e-01 1.40545622e-01 1.21638250e+00 -5.44688344e-01 6.71098351e-01 1.09192228e+00 3.80928844e-01 5.18458962e-01 -1.31552863e+00 -1.77259833e-01 8.48786458e-02 6.27559543e-01 -1.18085897e+00 -3.69615972e-01 7.17054784e-01 -5.81890702e-01 1.01418006e+00 -8.53035599e-02 5.25739133e-01 1.31508219e+00 -1.16000101e-02 1.05590260e+00 9.41383541e-01 -2.49810711e-01 6.32979929e-01 -3.66820902e-01 6.81978539e-02 4.86538708e-01 -2.35819846e-01 1.32293791e-01 -6.62256420e-01 3.30818677e-03 6.73229337e-01 2.51065105e-01 1.63160581e-02 -9.83724952e-01 -1.22195780e+00 8.18582177e-01 2.64548063e-01 5.33761919e-01 -5.07730186e-01 3.42229605e-01 6.36324286e-01 1.29913270e-01 3.95614237e-01 1.82545185e-01 -4.19205636e-01 -4.62479472e-01 -1.16286397e+00 -5.90195647e-03 3.97728503e-01 8.36698949e-01 7.43744791e-01 -1.44522727e-01 -4.90753055e-01 6.59679413e-01 1.87411278e-01 3.63126136e-02 6.84167922e-01 -1.43460882e+00 4.19300139e-01 8.10603201e-01 1.08437479e-01 -8.68666351e-01 -3.11966121e-01 9.87430960e-02 -3.91131192e-01 1.74408838e-01 3.75983298e-01 1.20399646e-01 -1.14156926e+00 1.68040228e+00 3.50665241e-01 6.73901021e-01 1.21087395e-01 8.58957410e-01 2.74520785e-01 2.42957383e-01 1.62800476e-01 -1.72742546e-01 7.78659761e-01 -1.10965657e+00 -6.28235102e-01 -6.51891455e-02 7.81911671e-01 -3.24366391e-01 8.42331111e-01 4.91305888e-01 -8.74573529e-01 -6.42657816e-01 -1.04839277e+00 5.57956770e-02 -2.21181944e-01 3.29234838e-01 6.23458326e-01 4.62301224e-01 -1.09818649e+00 1.05617547e+00 -1.23134828e+00 -7.82271981e-01 7.64372885e-01 4.27256167e-01 -4.04797882e-01 -5.74366339e-02 -8.12088788e-01 6.08345449e-01 8.98972452e-01 5.37049696e-02 -1.21117604e+00 -3.21804404e-01 -9.55113947e-01 -1.55274570e-01 5.61515510e-01 -4.49609488e-01 1.16276383e+00 -1.37776864e+00 -1.89117551e+00 8.33177984e-01 1.24496490e-01 -1.00977755e+00 6.06157839e-01 -3.52278084e-01 -1.34916380e-01 6.79110646e-01 2.07847521e-01 7.40763545e-01 1.11683369e+00 -7.36136556e-01 -8.10302913e-01 -5.69115758e-01 1.07807122e-01 2.70842731e-01 -3.99584830e-01 -5.92045188e-02 -2.58126378e-01 -7.16337979e-01 1.47408754e-01 -8.46838415e-01 -4.16313559e-01 -1.67402420e-02 -9.85191017e-02 -4.26833153e-01 9.77735758e-01 -4.28354949e-01 9.27116990e-01 -2.18964601e+00 5.98078072e-01 7.29024857e-02 -1.01844750e-01 3.23028773e-01 -2.39821196e-01 3.63635749e-01 4.95711062e-03 -2.95566350e-01 -7.80886531e-01 -5.04805923e-01 5.33003546e-02 4.47247088e-01 -3.66893709e-02 7.23629653e-01 1.86197519e-01 6.79654241e-01 -1.00441790e+00 -5.40618658e-01 3.73212904e-01 3.10930997e-01 -4.59768981e-01 2.81457812e-01 -3.83487433e-01 6.20820999e-01 -6.02341115e-01 4.48604971e-01 2.42736474e-01 -1.18440920e-02 2.20136493e-01 7.13555440e-02 1.36686251e-01 1.63658887e-01 -1.11311626e+00 2.54991913e+00 -7.87694007e-02 4.76561815e-01 -3.51501882e-01 -1.73075473e+00 1.02836490e+00 2.70602852e-01 1.08526230e+00 -5.29506564e-01 2.58639991e-01 6.23906367e-02 -2.07309842e-01 -5.66872358e-01 1.04862973e-01 -6.05999269e-02 -6.68199956e-02 5.99892855e-01 6.28849387e-01 1.79982275e-01 4.75433111e-01 1.28168017e-01 1.29557359e+00 8.06726754e-01 3.08951419e-02 -1.38868466e-01 6.58885717e-01 4.41009514e-02 9.08367276e-01 5.61476648e-01 -4.73993689e-01 5.39444327e-01 3.68669331e-01 -4.87899035e-01 -6.43391609e-01 -1.06868756e+00 9.15658548e-02 1.14844584e+00 1.24873266e-01 -4.06565309e-01 -8.14574301e-01 -1.24217308e+00 -2.79929042e-01 6.45524442e-01 -7.80847728e-01 -4.77190316e-01 -4.75362718e-01 -2.57022679e-01 3.84918600e-01 8.37234497e-01 6.62696898e-01 -1.20764220e+00 -9.79279041e-01 2.47158021e-01 -1.94169760e-01 -1.41244125e+00 -3.30080897e-01 2.07280844e-01 -1.17914832e+00 -1.55857670e+00 -5.35334706e-01 -8.50589454e-01 5.91857076e-01 1.87817559e-01 7.97321260e-01 -2.50762612e-01 -4.44156349e-01 8.46625090e-01 -7.31214046e-01 1.28676176e-01 -9.35557634e-02 -9.39509943e-02 -2.64126118e-02 6.50039077e-01 7.07646608e-01 -6.41709983e-01 -6.08691514e-01 3.78229469e-01 -9.32687342e-01 -2.37510175e-01 4.02194411e-01 6.91017866e-01 6.84931636e-01 1.03992067e-01 3.27947944e-01 -5.73449016e-01 -3.89532861e-03 -3.34177852e-01 -4.78033960e-01 1.13270015e-01 -4.06061232e-01 6.55144304e-02 4.25495923e-01 -2.45526910e-01 -1.12250948e+00 5.48665762e-01 5.59042275e-01 -6.60803676e-01 -5.08588850e-01 1.98649094e-01 -2.01516837e-01 1.75388724e-01 5.15012503e-01 2.65124738e-01 -5.96409142e-02 -7.27094591e-01 4.11934584e-01 4.49218392e-01 5.65868914e-01 -6.01576626e-01 5.59960604e-01 8.30224693e-01 -7.57633373e-02 -4.51594144e-01 -8.63738120e-01 -7.65323460e-01 -1.48707569e+00 -2.01519176e-01 1.24714923e+00 -6.40118837e-01 -2.58703947e-01 5.02456129e-01 -8.62826407e-01 -5.15824378e-01 -5.90303421e-01 9.08487320e-01 -1.32969689e+00 6.65657282e-01 -2.87335008e-01 -6.25524580e-01 5.35614491e-02 -9.85015750e-01 1.10376084e+00 8.87584090e-02 -1.14760943e-01 -1.00220656e+00 3.42250407e-01 6.30621135e-01 -9.19222236e-02 5.93847632e-01 5.95941722e-01 -8.70016396e-01 -5.23294628e-01 -8.07441473e-02 2.79097091e-02 6.99428201e-01 4.17669117e-01 7.89869651e-02 -6.91711247e-01 -2.77997524e-01 4.98683490e-02 -6.50415361e-01 1.27147245e+00 3.48134935e-01 1.32571316e+00 -5.05285263e-02 -2.53847122e-01 5.00692308e-01 1.24868667e+00 3.20996791e-01 6.55326247e-01 3.71677667e-01 6.45228922e-01 7.22165465e-01 9.72328722e-01 6.92484617e-01 9.79638323e-02 6.91094995e-01 4.18290257e-01 1.54052511e-01 1.55404121e-01 -2.46576518e-01 5.05460978e-01 4.41930652e-01 -2.23992571e-01 5.72026633e-02 -6.67710245e-01 7.48980284e-01 -2.38170743e+00 -1.01518786e+00 3.55006367e-01 2.25616241e+00 7.23274708e-01 1.27169654e-01 3.49325329e-01 3.60619396e-01 6.81181729e-01 3.30174357e-01 -7.33832240e-01 -4.30006623e-01 1.82406932e-01 4.77469951e-01 2.95716703e-01 1.33214504e-01 -1.62157071e+00 1.15588331e+00 6.32163095e+00 7.58085370e-01 -6.48183107e-01 9.06752869e-02 3.69405717e-01 -1.86611593e-01 2.99276680e-01 7.59775266e-02 -3.62738937e-01 2.06914634e-01 9.12959039e-01 -7.47200176e-02 1.09183282e-01 7.25694895e-01 4.19683486e-01 -1.74033195e-01 -1.42380440e+00 9.51307714e-01 2.16632962e-01 -9.82670426e-01 1.66424453e-01 -1.60214677e-01 8.62769246e-01 -1.80491328e-01 7.70782828e-02 1.02423944e-01 3.01576614e-01 -9.78725195e-01 4.44554985e-01 5.76843262e-01 4.08981711e-01 -7.96971440e-01 3.37057054e-01 1.47138640e-01 -1.12174881e+00 -3.38157982e-01 -2.98378438e-01 -1.42217144e-01 1.14891930e-02 9.16478038e-02 -5.28415620e-01 5.55475414e-01 7.18441784e-01 1.55519545e+00 -5.20720959e-01 1.29662597e+00 -3.78888667e-01 5.73180795e-01 -1.15078300e-01 4.64649707e-01 6.71109796e-01 -4.90030468e-01 4.71807480e-01 9.37861264e-01 1.24616936e-01 -6.06018081e-02 4.94747818e-01 6.15759969e-01 6.84236661e-02 1.83060914e-01 -7.06734121e-01 -2.19758064e-01 -8.64171088e-02 1.11078954e+00 -9.53193843e-01 -2.70406246e-01 -3.08722705e-01 1.14613914e+00 2.35735342e-01 4.40095484e-01 -9.72391546e-01 -2.16754436e-01 6.76928282e-01 -2.58550316e-01 8.04997981e-01 -2.30145454e-01 8.17153156e-02 -1.07610750e+00 8.31961855e-02 -6.91396177e-01 8.69371891e-01 -2.39050180e-01 -1.15469480e+00 3.15080613e-01 2.03832999e-01 -1.63927543e+00 -5.09963930e-01 -5.19792259e-01 -5.46030045e-01 4.73139100e-02 -1.28864217e+00 -8.85856092e-01 -9.14004222e-02 1.07880700e+00 7.04785585e-01 -2.80881077e-01 6.57494426e-01 1.89104661e-01 -5.78249454e-01 4.14677471e-01 3.15323174e-01 2.17296481e-01 4.95453566e-01 -1.40760744e+00 -5.16063906e-03 8.71164262e-01 5.86491525e-01 1.44558862e-01 2.47364014e-01 -3.39201480e-01 -9.64625597e-01 -1.22121620e+00 5.47377706e-01 -3.40521753e-01 5.76153815e-01 -2.56697565e-01 -7.40475118e-01 7.60294795e-01 2.11888522e-01 5.69602326e-02 8.96715999e-01 -1.42303288e-01 -3.96266073e-01 -1.82690263e-01 -1.23527837e+00 2.75157660e-01 1.26462924e+00 -4.30142313e-01 -8.29151094e-01 6.34588540e-01 5.15970170e-01 -1.54589862e-02 -9.72090185e-01 5.11013031e-01 2.68912256e-01 -1.03210676e+00 7.76420772e-01 -9.67945695e-01 3.12712640e-01 -3.92585933e-01 -2.05266848e-01 -9.19778943e-01 -1.51636198e-01 -8.15273941e-01 -1.11093871e-01 1.00131774e+00 1.04757354e-01 -1.90123156e-01 9.52581584e-01 4.80422348e-01 -8.95854160e-02 -5.96275032e-01 -1.20964670e+00 -9.56129193e-01 -2.26753935e-01 -3.87726277e-01 1.95056930e-01 6.99978173e-01 -1.14460289e-01 8.44682828e-02 -3.09049547e-01 -2.33261228e-01 8.81397724e-01 2.49080181e-01 5.08140624e-01 -1.22210968e+00 -3.14960092e-01 -3.12985003e-01 -9.93596494e-01 -9.25700426e-01 5.54210305e-01 -8.16918492e-01 1.22378811e-01 -1.48963332e+00 2.00787827e-01 8.43993872e-02 -8.30234826e-01 5.81751287e-01 2.47309193e-01 3.02805424e-01 1.42698824e-01 9.53818858e-02 -1.27397382e+00 9.84522283e-01 8.99286747e-01 -4.16110575e-01 -2.29000911e-01 6.49629831e-02 -2.48717472e-01 1.05550134e+00 7.78644919e-01 -4.85766023e-01 -7.76006341e-01 -3.64632219e-01 -5.25400579e-01 -1.83766618e-01 3.67379636e-01 -1.41002858e+00 4.69539538e-02 -1.97319925e-01 2.32982695e-01 -5.23179412e-01 3.28085184e-01 -7.62662888e-01 -1.90435737e-01 2.67668486e-01 -7.52144337e-01 -4.94776964e-01 -1.35266110e-01 9.99369323e-01 -4.07849848e-01 -1.71317205e-01 9.91297960e-01 -8.76163095e-02 -1.12132227e+00 5.11805415e-01 -2.81113297e-01 2.64962971e-01 1.48560953e+00 -5.29299140e-01 2.87021965e-01 -5.45890245e-04 -1.20921659e+00 2.01215670e-01 4.00010109e-01 6.15248978e-01 7.32256532e-01 -1.62686121e+00 -4.77411896e-01 -5.86050376e-02 2.43172571e-01 -2.37399861e-01 2.11738944e-02 9.92753983e-01 -1.68063775e-01 2.66828299e-01 -5.06910205e-01 -8.12716544e-01 -1.08101964e+00 6.11276150e-01 3.48446995e-01 -3.02014053e-01 -8.42617631e-01 7.40713477e-01 1.03687063e-01 -2.08074525e-01 5.08314908e-01 -3.75539660e-01 -4.65852529e-01 1.47704735e-01 4.02424127e-01 5.87508798e-01 -1.95274830e-01 -8.81267428e-01 -4.16203350e-01 5.51240504e-01 -2.98375301e-02 -1.67554349e-01 1.30708301e+00 -1.12719290e-01 1.02022342e-01 6.19120300e-01 1.42729223e+00 -9.39066827e-01 -1.76887381e+00 -4.46687996e-01 5.48911452e-01 -6.96500003e-01 -1.05861031e-01 -4.51726049e-01 -1.29260814e+00 8.78760397e-01 6.79752767e-01 -1.81126967e-01 1.28576183e+00 -2.59102043e-02 8.69857669e-01 4.80783761e-01 3.56203616e-01 -1.71840596e+00 6.04485929e-01 3.72135311e-01 6.15849495e-01 -1.23765862e+00 -1.71506196e-01 3.80622335e-02 -7.83945918e-01 1.09119344e+00 6.36518776e-01 -4.45572406e-01 4.35591698e-01 -4.57895875e-01 -1.60857514e-01 -1.73506975e-01 -5.82125247e-01 -6.06478572e-01 2.68951148e-01 8.24885368e-01 3.92755643e-02 -2.23144710e-01 -5.20107090e-01 3.78958523e-01 4.60538805e-01 1.75368249e-01 2.73058563e-01 1.03484809e+00 -3.82063240e-01 -1.20825851e+00 1.90690979e-01 2.84635901e-01 -3.28740150e-01 3.64666969e-01 -5.91167212e-01 4.41547155e-01 1.75531030e-01 1.01919830e+00 1.01373672e-01 -4.14927542e-01 2.85695970e-01 1.43358126e-01 5.78601539e-01 -7.64649868e-01 -3.40014100e-01 1.89238176e-01 -1.01211652e-01 -1.26789355e+00 -1.26827836e+00 -1.02485216e+00 -1.47285044e+00 4.85403836e-01 5.80807514e-02 1.37134045e-01 1.56463906e-01 1.21614707e+00 3.37641060e-01 2.15727955e-01 9.51160431e-01 -7.99088776e-01 -5.96829534e-01 -8.23946357e-01 -8.46755087e-01 7.06568599e-01 -8.55116826e-03 -1.10827386e+00 -2.90993452e-01 4.12929267e-01]	[8.528399467468262, 0.7484104037284851]
2d59d51d-c885-4ab3-9b8e-e1c0d0eacd40	shadowdiffusion-when-degradation-prior-meets	2212.04711	null	https://arxiv.org/abs/2212.04711v2	https://arxiv.org/pdf/2212.04711v2.pdf	ShadowDiffusion: When Degradation Prior Meets Diffusion Model for Shadow Removal	Recent deep learning methods have achieved promising results in image shadow removal. However, their restored images still suffer from unsatisfactory boundary artifacts, due to the lack of degradation prior embedding and the deficiency in modeling capacity. Our work addresses these issues by proposing a unified diffusion framework that integrates both the image and degradation priors for highly effective shadow removal. In detail, we first propose a shadow degradation model, which inspires us to build a novel unrolling diffusion model, dubbed ShandowDiffusion. It remarkably improves the model's capacity in shadow removal via progressively refining the desired output with both degradation prior and diffusive generative prior, which by nature can serve as a new strong baseline for image restoration. Furthermore, ShadowDiffusion progressively refines the estimated shadow mask as an auxiliary task of the diffusion generator, which leads to more accurate and robust shadow-free image generation. We conduct extensive experiments on three popular public datasets, including ISTD, ISTD+, and SRD, to validate our method's effectiveness. Compared to the state-of-the-art methods, our model achieves a significant improvement in terms of PSNR, increasing from 31.69dB to 34.73dB over SRD dataset.	['Bihan Wen', 'Hanspeter Pfister', 'YuFei Wang', 'Siyu Huang', 'Wenhan Yang', 'Chong Wang', 'Lanqing Guo']	2022-12-09	null	http://openaccess.thecvf.com//content/CVPR2023/html/Guo_ShadowDiffusion_When_Degradation_Prior_Meets_Diffusion_Model_for_Shadow_Removal_CVPR_2023_paper.html	http://openaccess.thecvf.com//content/CVPR2023/papers/Guo_ShadowDiffusion_When_Degradation_Prior_Meets_Diffusion_Model_for_Shadow_Removal_CVPR_2023_paper.pdf	cvpr-2023-1	['shadow-removal', 'image-shadow-removal']	['computer-vision', 'computer-vision']	[ 3.81932944e-01 -2.69473851e-01 1.82352766e-01 -4.24626470e-02 -4.54481691e-01 -9.35942680e-02 4.79304045e-01 -5.11738479e-01 3.70152364e-03 8.26497257e-01 6.27043903e-01 -4.59389359e-01 3.33033264e-01 -6.28255606e-01 -4.61474150e-01 -1.18294787e+00 2.82730162e-01 -3.30807000e-01 5.51963747e-01 -2.14291826e-01 -1.02241132e-02 3.38781834e-01 -1.09850216e+00 4.28896472e-02 1.37953615e+00 8.93489003e-01 5.65114737e-01 3.62269282e-01 8.40603709e-02 8.60931456e-01 -7.71973014e-01 -1.07344061e-01 2.86131501e-01 -4.72096533e-01 -6.16158620e-02 3.39774102e-01 3.50362919e-02 -9.42425489e-01 -8.45707238e-01 7.99332857e-01 8.53170455e-01 -8.60898644e-02 5.72994411e-01 -1.12444413e+00 -1.15854275e+00 8.84995684e-02 -1.07385993e+00 1.62946895e-01 7.17995167e-02 3.52891117e-01 5.85099220e-01 -9.10336614e-01 4.25604224e-01 1.39896691e+00 6.19441688e-01 4.14898306e-01 -1.20855975e+00 -7.08549261e-01 1.26201823e-01 1.82964250e-01 -1.21549761e+00 -4.78866875e-01 1.01888180e+00 -9.64794606e-02 4.50132012e-01 2.00708315e-01 5.28242648e-01 1.06902063e+00 2.79461861e-01 1.21158803e+00 1.47244811e+00 -2.69959748e-01 -4.40377695e-03 -1.32587656e-01 -2.21317589e-01 6.44412279e-01 2.88922817e-01 2.90582150e-01 -5.05525291e-01 3.54754254e-02 1.06868935e+00 -9.85240862e-02 -8.46104622e-01 -2.71560550e-01 -1.11388147e+00 4.67469245e-01 6.51349306e-01 8.68119113e-03 -3.71805251e-01 2.05269337e-01 -2.92308182e-02 -2.87384894e-02 7.70784497e-01 -2.19802007e-01 -1.05561823e-01 3.68598223e-01 -9.74204600e-01 7.19609782e-02 3.70567679e-01 7.07993865e-01 4.18516219e-01 3.84412795e-01 -5.95522404e-01 9.92650628e-01 5.74926496e-01 8.08095694e-01 9.16769207e-02 -1.09349585e+00 3.21576208e-01 2.53161490e-01 3.79928023e-01 -1.15129852e+00 2.34347023e-03 -7.92856812e-01 -1.36524212e+00 4.14279163e-01 -1.89419948e-02 4.14668291e-04 -1.13307750e+00 1.64422178e+00 3.92831624e-01 4.76167351e-01 8.44759718e-02 1.00235271e+00 8.67649257e-01 1.05014408e+00 -1.66007146e-01 -4.31193829e-01 1.02574396e+00 -1.32333922e+00 -9.77207243e-01 -2.14203730e-01 -2.43125092e-02 -9.24805105e-01 1.27673697e+00 4.73357230e-01 -1.01831305e+00 -5.63300848e-01 -1.28503120e+00 -8.84904191e-02 2.81822175e-01 3.08459163e-01 6.59372032e-01 6.11756802e-01 -1.26912904e+00 2.83445120e-01 -5.93104303e-01 4.70970524e-03 7.60699272e-01 -7.23323599e-02 2.87372470e-01 -3.67040992e-01 -1.03767276e+00 5.72121024e-01 -2.25030527e-01 2.32533082e-01 -1.34386218e+00 -7.06637681e-01 -5.48377454e-01 -1.02297440e-01 3.47347140e-01 -8.76102805e-01 9.38879371e-01 -6.10313952e-01 -1.52651811e+00 5.13786793e-01 -2.57666498e-01 -2.90255278e-01 8.29878390e-01 -2.97294527e-01 -2.93039083e-01 5.92603236e-02 5.06867701e-03 5.36046267e-01 1.10632122e+00 -1.97591186e+00 -3.40761870e-01 -1.26577362e-01 3.69684845e-02 3.46980751e-01 -5.81177771e-01 -1.68059677e-01 -9.63556111e-01 -1.18232691e+00 8.08391646e-02 -7.35929072e-01 -2.02029124e-01 4.25386429e-01 -5.12749195e-01 1.07808046e-01 1.16121435e+00 -9.78016078e-01 1.42768514e+00 -2.22092342e+00 9.42577049e-02 -2.47358859e-01 4.85964507e-01 3.85344386e-01 -1.85274020e-01 2.54039019e-01 2.71015525e-01 -1.50366515e-01 -7.08352447e-01 -7.65899181e-01 -3.19852872e-04 3.69815558e-01 -7.49537051e-01 5.46075225e-01 -1.74968600e-01 7.88655877e-01 -7.45960057e-01 -5.13344705e-01 2.03506917e-01 9.61558938e-01 -1.33816123e-01 3.38465512e-01 2.25717928e-02 4.59056884e-01 -3.90667677e-01 8.14616084e-01 1.24395359e+00 -2.78003156e-01 -4.41734716e-02 -3.22794557e-01 -1.24898795e-02 -3.06115020e-02 -9.49693084e-01 1.57002485e+00 -5.97773135e-01 7.86526859e-01 2.47910932e-01 -4.45992231e-01 9.35750902e-01 2.52478253e-02 3.18720102e-01 -1.02085531e+00 -5.30210212e-02 2.33275861e-01 -4.01856035e-01 -3.04175287e-01 4.70934778e-01 -3.40189338e-02 3.93781304e-01 5.76943040e-01 -5.63580513e-01 -1.94212273e-02 -1.54850975e-01 4.06305224e-01 1.02099729e+00 2.23069698e-01 -3.64817172e-01 -3.01961243e-01 6.74978256e-01 -6.02971613e-01 7.57847726e-01 5.23731470e-01 -2.52558649e-01 8.04401338e-01 2.29773268e-01 -1.20549880e-01 -9.57319081e-01 -1.28398120e+00 1.57288779e-02 6.98558211e-01 7.53747463e-01 -1.02114789e-01 -7.18829572e-01 -5.52238941e-01 -1.04567610e-01 6.74226224e-01 -5.56268096e-01 -1.66340783e-01 -5.91008723e-01 -1.33180285e+00 5.73050082e-01 4.66524750e-01 1.15703154e+00 -7.88474083e-01 -3.14446926e-01 3.62043804e-03 -5.25024712e-01 -9.32931244e-01 -6.89542353e-01 -2.33735994e-01 -7.70235181e-01 -7.43677318e-01 -1.37074339e+00 -7.39592850e-01 7.13344216e-01 9.69663024e-01 9.39265728e-01 4.65502530e-01 -1.72962919e-01 -2.15354692e-02 -2.04025626e-01 -2.81436265e-01 -3.55731279e-01 -3.42167795e-01 -1.97750017e-01 1.34806737e-01 -4.76328939e-01 -8.10451627e-01 -1.25493526e+00 4.41058397e-01 -1.18555951e+00 4.25047696e-01 9.20358300e-01 8.64942014e-01 3.56364518e-01 2.84822911e-01 3.40690166e-01 -5.80022752e-01 8.55799675e-01 -3.36624205e-01 -4.50128317e-01 1.36427820e-01 -9.79282260e-01 -2.41699368e-02 4.47758496e-01 -3.09009016e-01 -1.76077950e+00 -2.43568927e-01 -1.05430357e-01 -3.83589506e-01 2.23679110e-01 -1.02575787e-03 -3.58922988e-01 -2.19407633e-01 4.42099035e-01 6.17220700e-01 -2.26810753e-01 -7.15366066e-01 2.87294418e-01 5.71218967e-01 6.22871161e-01 -4.25114512e-01 1.19435132e+00 9.35151517e-01 -1.42837718e-01 -6.48186922e-01 -7.81948984e-01 -4.13723066e-02 -2.38982975e-01 -2.93503523e-01 5.68382144e-01 -1.04681158e+00 -5.26278257e-01 1.16615951e+00 -1.13624680e+00 -8.16306531e-01 5.32076322e-02 -7.08794966e-02 -1.39887705e-01 8.73127162e-01 -8.38363469e-01 -8.54840457e-01 -5.32839715e-01 -1.14775276e+00 1.17114162e+00 2.57329643e-01 5.36542892e-01 -9.13267016e-01 -1.51991069e-01 4.13250744e-01 6.92288876e-01 2.33161300e-01 8.33283901e-01 5.82074523e-01 -1.02936542e+00 3.27319384e-01 -7.97390521e-01 6.93911016e-01 3.12431812e-01 -2.52627134e-01 -1.02546835e+00 -4.20383751e-01 1.80617750e-01 2.51053774e-04 1.45866370e+00 4.34669167e-01 1.11667943e+00 -1.74205482e-01 -4.52505916e-01 8.65417778e-01 1.48776913e+00 1.10862203e-01 1.09019089e+00 3.81161124e-01 7.36260116e-01 2.38285869e-01 6.33492887e-01 5.33031225e-01 4.57435966e-01 5.70459545e-01 5.14592111e-01 -7.14665949e-01 -9.99325693e-01 -2.13614151e-01 3.90381098e-01 7.52531350e-01 -5.04673012e-02 -6.80927575e-01 -5.93261600e-01 4.85704780e-01 -1.74680638e+00 -5.21349847e-01 -3.18848282e-01 1.87018466e+00 9.28536177e-01 1.60015002e-01 -2.91956604e-01 2.97882259e-01 4.72245902e-01 7.97808766e-01 -7.26177394e-01 2.94743001e-01 -5.53043664e-01 -9.50282067e-02 5.08148015e-01 6.65208220e-01 -8.26429307e-01 9.08983052e-01 6.22142029e+00 1.08995974e+00 -8.44627738e-01 2.33638078e-01 8.22768807e-01 3.59859228e-01 -6.30497813e-01 -1.36142597e-02 -4.84475732e-01 7.27188706e-01 1.68283239e-01 8.29433799e-02 4.69389677e-01 3.75786543e-01 5.73629797e-01 -3.91131490e-01 -1.86272264e-01 9.17725921e-01 2.72543997e-01 -1.24113333e+00 7.38052204e-02 2.08948150e-01 9.77324545e-01 -2.75292665e-01 3.32123488e-01 -5.21355942e-02 4.10297811e-01 -8.40295374e-01 6.91828191e-01 5.91682792e-01 8.00743878e-01 -5.88739395e-01 5.17017663e-01 2.68496215e-01 -1.15240479e+00 -1.44853041e-01 -3.87240261e-01 1.17183223e-01 4.23256487e-01 1.09409666e+00 -4.77561295e-01 5.70829511e-01 7.97372282e-01 8.12713802e-01 -4.24882382e-01 9.92362022e-01 -7.49254346e-01 7.30302811e-01 -2.40251776e-02 4.33667004e-01 2.35842504e-02 -3.22122276e-01 4.82090741e-01 1.10864699e+00 2.61627406e-01 2.08609492e-01 1.57916863e-02 8.34926069e-01 -2.77850717e-01 -2.88927078e-01 -1.19730391e-01 4.05627817e-01 4.10131961e-01 1.08293486e+00 -7.23468661e-01 -3.63249362e-01 -2.46439889e-01 1.53457975e+00 -1.54282197e-01 7.46824861e-01 -1.21340668e+00 -2.16721147e-01 6.88617885e-01 -4.02587801e-02 1.88864484e-01 -4.68741268e-01 -4.11968410e-01 -1.12840629e+00 7.98177198e-02 -8.25077772e-01 -1.75532743e-01 -1.07690048e+00 -1.06112862e+00 4.64050829e-01 -3.13438356e-01 -1.02781749e+00 6.26645863e-01 -1.89124301e-01 -6.51934922e-01 9.52118516e-01 -2.13582706e+00 -1.30782664e+00 -8.41747344e-01 6.66413784e-01 6.24378800e-01 9.65612084e-02 2.20137700e-01 6.49304271e-01 -6.76125050e-01 3.63906115e-01 3.65501016e-01 -2.28734091e-01 9.53653574e-01 -1.02955663e+00 5.54401338e-01 1.19266295e+00 -2.72482902e-01 1.34120971e-01 8.14239562e-01 -8.18428874e-01 -1.33122790e+00 -1.11433959e+00 5.37506580e-01 -1.31425381e-01 4.40053016e-01 -1.72824293e-01 -1.00597441e+00 1.30013108e-01 3.44024211e-01 -1.29733369e-01 1.73565313e-01 -5.84412515e-01 -2.18707234e-01 -4.28229898e-01 -8.80602121e-01 8.50750446e-01 1.34722757e+00 -3.38828981e-01 -1.47999465e-01 3.99197996e-01 1.06119907e+00 -4.78680640e-01 -3.67105573e-01 4.40538973e-01 3.24153990e-01 -1.35157228e+00 1.29470062e+00 4.41021264e-01 5.03883123e-01 -5.74645996e-01 -1.14319265e-01 -1.13966620e+00 -3.56856048e-01 -8.43235612e-01 -5.11260927e-01 1.51853490e+00 -2.61638667e-02 -5.93297243e-01 5.74353993e-01 1.00915276e-01 -4.10871953e-01 -1.02667820e+00 -5.16022503e-01 -8.36708546e-01 -3.39538723e-01 -3.17314208e-01 6.99821174e-01 5.25055647e-01 -9.87759948e-01 1.92752570e-01 -9.97181594e-01 3.09654266e-01 1.10823965e+00 2.58922845e-01 9.07605529e-01 -7.20855832e-01 -3.60553354e-01 -2.67512858e-01 2.46071354e-01 -1.84980381e+00 -1.11431539e-01 -3.04827362e-01 2.74597198e-01 -1.98073161e+00 4.09829855e-01 -5.64833879e-01 -3.01896721e-01 3.53316009e-01 -4.05973911e-01 5.97374737e-01 1.07884392e-01 5.81236780e-01 -3.00966561e-01 1.05380499e+00 1.86293757e+00 -1.37286022e-01 -2.57536769e-01 -1.55946864e-02 -9.26785171e-01 6.52257025e-01 5.67568958e-01 -2.29354799e-01 -5.96993089e-01 -9.45768595e-01 -1.57865033e-01 -4.62702587e-02 5.71025908e-01 -8.99863958e-01 3.72304544e-02 -2.02145532e-01 4.60942805e-01 -7.12941885e-01 6.72916532e-01 -5.05967021e-01 -1.85395626e-03 6.03083849e-01 -4.79482766e-03 -4.58352804e-01 8.83189887e-02 9.12581205e-01 -6.04263358e-02 4.48539257e-01 9.04298365e-01 1.87644422e-01 -6.37513578e-01 3.83668900e-01 -7.79251829e-02 -1.61470532e-01 7.67458916e-01 -2.02127129e-01 -6.09290779e-01 -6.40138090e-01 -1.03436135e-01 1.77949965e-01 6.81168377e-01 3.16818595e-01 9.12091136e-01 -1.23961186e+00 -8.97764862e-01 1.64461717e-01 -3.70716214e-01 -1.45254254e-01 4.94814724e-01 1.01319504e+00 -5.34289956e-01 6.68643638e-02 6.38660938e-02 -4.45311636e-01 -1.21735871e+00 4.30813372e-01 1.45391569e-01 -2.04888597e-01 -9.36627924e-01 9.59211826e-01 7.54122674e-01 9.03864279e-02 4.43270862e-01 -1.71192944e-01 2.00161636e-01 -3.15187067e-01 4.89610165e-01 4.87122774e-01 -2.35727847e-01 -5.14571249e-01 -1.44897446e-01 3.93744439e-01 2.07455486e-01 4.15459238e-02 1.38940394e+00 -6.32247746e-01 -5.20863310e-02 -1.29455969e-01 8.70211005e-01 2.47530714e-01 -1.91093349e+00 -4.34927732e-01 -5.77236533e-01 -9.73524511e-01 3.64272177e-01 -9.71774101e-01 -1.39569342e+00 9.36487257e-01 7.51136124e-01 3.44411563e-03 1.66737378e+00 -1.48822010e-01 1.55644858e+00 -1.86243996e-01 1.92175493e-01 -7.57906079e-01 5.99121332e-01 2.60333538e-01 1.10470188e+00 -1.16155601e+00 3.53675634e-01 -5.73515415e-01 -5.24273038e-01 6.91879213e-01 4.01151240e-01 -2.82730330e-02 4.37866241e-01 4.52502549e-01 1.56955644e-01 1.81056783e-01 -4.78287131e-01 -4.29860838e-02 2.74941862e-01 7.06740677e-01 8.28902498e-02 -1.47358090e-01 -3.96778077e-01 4.09060717e-01 2.94180930e-01 -3.42852809e-02 4.55976039e-01 8.01784277e-01 -5.73553205e-01 -1.05132937e+00 -5.52496076e-01 4.66191880e-02 -7.25328401e-02 -3.64654154e-01 -2.51799852e-01 5.19863546e-01 8.04972351e-02 1.05242491e+00 -5.02315104e-01 -3.10928404e-01 1.07390560e-01 -7.23783493e-01 2.96898514e-01 -1.56947732e-01 1.32848039e-01 4.26229686e-01 -1.03415601e-01 -4.92421001e-01 -4.18073535e-01 -3.84100229e-01 -1.08899987e+00 -4.03438270e-01 -2.73179352e-01 -2.48799294e-01 4.41006780e-01 8.00750554e-01 3.72068644e-01 7.97523320e-01 8.57919097e-01 -9.54943061e-01 -3.35010737e-01 -6.84159219e-01 -6.00430965e-01 1.03019692e-01 5.81436217e-01 -7.84176707e-01 -3.96274626e-01 1.47684380e-01]	[10.830586433410645, -3.978990077972412]
ee363ec1-74a9-45d5-8619-46efa291a76a	understanding-and-generalizing-monotonic	2107.13052	null	https://arxiv.org/abs/2107.13052v1	https://arxiv.org/pdf/2107.13052v1.pdf	Understanding and Generalizing Monotonic Proximity Graphs for Approximate Nearest Neighbor Search	Graph-based algorithms have shown great empirical potential for the approximate nearest neighbor (ANN) search problem. Currently, graph-based ANN search algorithms are designed mainly using heuristics, whereas theoretical analysis of such algorithms is quite lacking. In this paper, we study a fundamental model of proximity graphs used in graph-based ANN search, called Monotonic Relative Neighborhood Graph (MRNG), from a theoretical perspective. We use mathematical proofs to explain why proximity graphs that are built based on MRNG tend to have good searching performance. We also run experiments on MRNG and graphs generalizing MRNG to obtain a deeper understanding of the model. Our experiments give guidance on how to approximate and generalize MRNG to build proximity graphs on a large scale. In addition, we discover and study a hidden structure of MRNG called conflicting nodes, and we give theoretical evidence how conflicting nodes could be used to improve ANN search methods that are based on MRNG.	['Minjia Zhang', 'Dantong Zhu']	2021-07-27	null	null	null	null	['mathematical-proofs']	['miscellaneous']	[-2.05576211e-01 2.23771200e-01 -6.19792163e-01 -3.15024137e-01 -3.27641457e-01 -6.38635218e-01 2.48697370e-01 4.32488024e-01 3.87827978e-02 6.75207615e-01 2.30113968e-01 -6.21250272e-01 -9.80107546e-01 -1.57499754e+00 -6.31159544e-01 -3.17465514e-01 -4.46630418e-01 7.62416244e-01 2.95752168e-01 -4.18045819e-01 5.94655573e-01 9.25096631e-01 -1.42447925e+00 2.54485589e-02 7.74364352e-01 5.78269899e-01 1.68863181e-02 5.74763954e-01 -1.29342496e-01 6.58517122e-01 -4.18615341e-01 -5.09545207e-01 4.64922726e-01 -5.19146621e-01 -1.02112496e+00 -4.20645535e-01 3.87594104e-01 9.24517214e-02 -5.61761856e-01 9.89917099e-01 5.52432716e-01 3.85953128e-01 5.09602606e-01 -1.52222931e+00 -1.00475633e+00 1.18058228e+00 -3.87111634e-01 1.59887299e-01 6.38282657e-01 -4.59204048e-01 1.42358398e+00 -5.23457348e-01 5.47501862e-01 1.14749289e+00 9.96432543e-01 3.26292902e-01 -9.90875185e-01 -3.03362638e-01 -1.53720200e-01 5.88592112e-01 -1.74726462e+00 3.04806931e-03 8.24704409e-01 3.21491867e-01 9.56309497e-01 7.86433995e-01 9.16916490e-01 3.88545305e-01 1.11252829e-01 3.67935836e-01 7.38729239e-01 -9.08752918e-01 3.01511377e-01 -8.84232968e-02 4.12587821e-01 9.86151695e-01 8.34748626e-01 3.69555384e-01 -6.11716747e-01 -5.77478170e-01 7.70578206e-01 -1.46699324e-01 -3.52958918e-01 -7.67751753e-01 -6.94756329e-01 1.24681854e+00 1.02966785e+00 6.91121817e-01 -1.32649496e-01 4.50396776e-01 -1.28349900e-01 3.25089127e-01 2.29013816e-01 7.93479443e-01 -1.11682624e-01 2.00209454e-01 -7.10568368e-01 2.89794221e-03 1.24641991e+00 1.20987785e+00 8.73887837e-01 -5.10198534e-01 1.18776210e-01 6.85674310e-01 1.18565731e-01 2.28209138e-01 1.69601753e-01 -8.59983563e-01 3.22663933e-01 7.46494412e-01 -2.46103257e-01 -1.69401336e+00 -6.13370299e-01 -4.51944023e-01 -5.31759262e-01 -3.92195553e-01 3.18652719e-01 3.65290105e-01 -5.53524554e-01 1.44187498e+00 1.68412596e-01 -2.66506430e-02 -7.92915560e-03 6.28411412e-01 5.03388345e-01 5.38654089e-01 -4.97548163e-01 -2.63667464e-01 9.41836894e-01 -1.00349569e+00 -2.91038126e-01 1.50878699e-02 1.45441329e+00 -3.66840601e-01 9.81314421e-01 5.88974217e-03 -1.03103662e+00 -1.29813835e-01 -9.98950660e-01 4.26816978e-02 -6.49644315e-01 -5.76322317e-01 7.81389356e-01 7.02888966e-01 -1.50068450e+00 9.53972161e-01 -5.48904419e-01 -8.83609235e-01 6.29472136e-02 4.07311499e-01 -6.96498230e-02 -3.85734022e-01 -1.23583949e+00 9.48965073e-01 5.57106972e-01 5.77778332e-02 1.20698467e-01 -3.00253391e-01 -7.39091516e-01 1.98961422e-02 4.83016998e-01 -7.21025765e-01 5.93592763e-01 -2.35219538e-01 -4.98381913e-01 6.02126360e-01 -2.48907879e-01 -6.46049619e-01 5.18471636e-02 4.44922715e-01 -6.30684197e-01 1.89629033e-01 4.40814495e-02 3.41556221e-01 1.40764832e-01 -1.12187290e+00 -4.32838649e-01 -6.45648897e-01 2.81449348e-01 1.37879953e-01 -6.67430401e-01 -3.89062494e-01 -1.82273760e-01 -5.87627828e-01 4.19940978e-01 -1.02474165e+00 -4.96087223e-01 -1.08559147e-01 -3.92210215e-01 -5.36238253e-01 4.52005357e-01 5.07741906e-02 1.90656388e+00 -1.51825953e+00 1.79399177e-02 1.21095645e+00 5.57407260e-01 -9.31911767e-02 -4.37698364e-01 9.77977216e-01 1.79727480e-01 3.62126976e-01 3.19933072e-02 6.27230644e-01 -2.14429479e-02 6.07806563e-01 -6.29695058e-02 4.01429921e-01 -8.69448304e-01 1.26112843e+00 -1.20298433e+00 -5.05514860e-01 -1.06394149e-01 -1.20800883e-02 -4.00984734e-01 -3.23800713e-01 -1.72594756e-01 -5.80216706e-01 -6.54015064e-01 9.64508057e-01 5.20762086e-01 -6.18235826e-01 3.76123428e-01 -2.59878099e-01 1.17002465e-01 -9.39991474e-02 -7.85111070e-01 1.37513673e+00 -3.44027996e-01 4.95379269e-01 -1.70511335e-01 -1.06192160e+00 9.84875381e-01 -2.40527064e-01 2.82468140e-01 -6.99651897e-01 -1.35958731e-01 4.33064908e-01 -8.77775345e-03 -2.39674002e-01 6.33230448e-01 3.51084411e-01 -6.34246022e-02 7.90075481e-01 -6.83355808e-01 5.37065975e-02 2.98160672e-01 3.99372160e-01 1.47216439e+00 -5.64629793e-01 4.49425966e-01 -6.70354724e-01 3.33472788e-01 4.43436444e-01 -4.22992408e-02 1.25278294e+00 1.37351066e-01 2.12261036e-01 2.43372783e-01 -7.16885388e-01 -8.56301844e-01 -7.71639228e-01 -1.26054332e-01 8.81663084e-01 5.27803779e-01 -9.95865107e-01 -5.93970239e-01 -5.66023231e-01 1.53105795e-01 7.18532443e-01 -6.41024768e-01 -3.04337323e-01 -5.47923148e-01 -6.90316617e-01 5.28643489e-01 6.35987163e-01 1.84267253e-01 -8.85121882e-01 -4.06744063e-01 1.32476166e-01 -4.32673879e-02 -8.26341450e-01 -5.77964067e-01 6.51052073e-02 -1.08295715e+00 -1.38814998e+00 -2.36609712e-01 -7.28994131e-01 8.34712982e-01 4.32721645e-01 1.32908893e+00 6.32835448e-01 -2.87226319e-01 6.82748139e-01 -6.13390982e-01 -4.94482480e-02 -3.02268058e-01 4.60533172e-01 2.29374528e-01 -6.95455730e-01 5.94762564e-01 -8.69588375e-01 -4.69695628e-01 6.64582968e-01 -8.46979618e-01 -4.73530650e-01 7.25096166e-01 6.65216923e-01 5.72585762e-01 4.58793342e-01 3.39625120e-01 -9.01080608e-01 1.01237869e+00 -5.26773453e-01 -8.88193548e-01 6.61135912e-01 -1.34458005e+00 5.45510113e-01 4.55206901e-01 -2.24213749e-01 3.41362841e-02 -7.02565387e-02 -2.60597859e-02 -3.52934659e-01 4.59610313e-01 8.37379754e-01 1.03920251e-01 -8.25656176e-01 1.01196122e+00 1.45212680e-01 -1.01934597e-01 -2.53536761e-01 6.26177728e-01 4.22747165e-01 3.05489659e-01 -5.93815267e-01 8.70432794e-01 3.12013447e-01 6.30763471e-01 -6.96035147e-01 -7.39095747e-01 -3.56426537e-01 -4.10063237e-01 1.70228094e-01 4.27056432e-01 -2.73331821e-01 -7.09717274e-01 -3.95262778e-01 -9.35706973e-01 -2.67457396e-01 -2.48582050e-01 1.81307077e-01 -5.06785154e-01 3.69076639e-01 -4.09877270e-01 -8.17559719e-01 -2.25628451e-01 -5.14594138e-01 5.83469391e-01 -1.55631574e-02 -4.78641689e-01 -1.37310517e+00 2.55665332e-01 1.09071270e-01 3.40775609e-01 1.22872636e-01 1.16144252e+00 -1.00494158e+00 -6.99895859e-01 -2.04731032e-01 -1.56598896e-01 -5.61707139e-01 2.32345521e-01 -2.37452447e-01 -2.20638171e-01 -2.94758677e-01 -5.03074765e-01 1.92336217e-01 5.97683251e-01 3.83039951e-01 1.31075788e+00 -7.62028575e-01 -1.14063048e+00 6.97755218e-01 1.64976037e+00 1.42506108e-01 6.31130934e-01 4.19546843e-01 7.44862199e-01 6.01887226e-01 5.58820963e-01 1.72757313e-01 4.38312441e-01 6.96235538e-01 4.86740887e-01 1.44654304e-01 1.65646434e-01 -5.06844461e-01 -3.06863755e-01 7.36642599e-01 -1.35231256e-01 -5.86996913e-01 -9.64092493e-01 6.58656061e-01 -2.00562119e+00 -9.27367866e-01 -4.38206226e-01 2.18727708e+00 5.05807042e-01 -1.41365603e-01 3.88259768e-01 2.48952433e-01 1.07561541e+00 2.34502498e-02 -2.21271574e-01 -4.97795880e-01 -1.55631229e-01 2.61492997e-01 8.58708978e-01 6.80767179e-01 -4.62639183e-01 7.88097858e-01 7.35343742e+00 9.14365649e-01 -4.05579805e-01 -3.76233965e-01 3.47768366e-01 1.98078334e-01 -7.57683158e-01 2.99627662e-01 -6.72143400e-01 1.87167168e-01 9.14057136e-01 -4.88825917e-01 8.25169206e-01 1.00236058e+00 -3.26192200e-01 2.52802670e-01 -1.31962955e+00 1.13519716e+00 -1.22696333e-01 -1.67663014e+00 2.52106816e-01 4.52576697e-01 8.42081308e-01 -1.73922390e-01 -2.69402504e-01 -1.42087594e-01 4.32522953e-01 -1.21769381e+00 -2.25353446e-02 4.26198244e-01 4.71293032e-01 -1.15548980e+00 6.47233009e-01 2.59883016e-01 -1.50752664e+00 -1.77600548e-01 -6.88806355e-01 -3.66235562e-02 -1.88672379e-01 7.09484100e-01 -7.24481344e-01 7.88810670e-01 6.66374147e-01 4.21950161e-01 -6.89884126e-01 9.94411707e-01 1.14852108e-01 2.02107474e-01 -6.63432956e-01 -6.99152708e-01 2.74924606e-01 -2.77363241e-01 6.11427009e-01 6.13975823e-01 7.32070327e-01 3.63720000e-01 -1.76650614e-01 9.76434886e-01 2.60799844e-02 9.90201905e-02 -1.15154564e+00 -2.10935414e-01 1.04705620e+00 9.93907034e-01 -9.84977841e-01 -5.01524471e-02 -7.39712417e-02 8.47398877e-01 4.69704866e-01 1.10961564e-01 -6.97707474e-01 -7.76325405e-01 2.34509274e-01 5.81054091e-01 2.22390711e-01 -3.78708430e-02 -2.34357700e-01 -7.21788943e-01 1.24969922e-01 -6.00880623e-01 9.31564569e-01 -8.31122637e-01 -1.48062539e+00 7.35874057e-01 2.49848261e-01 -8.90268445e-01 -5.08952558e-01 -3.90877843e-01 -5.53575814e-01 5.27695954e-01 -8.91473472e-01 -6.71211064e-01 -2.57381678e-01 7.00905442e-01 -1.57059535e-01 1.50998726e-01 7.18490362e-01 -8.56591314e-02 1.08914644e-01 8.64417851e-01 7.96943530e-02 -3.69856991e-02 1.93717629e-01 -1.01586187e+00 4.04083222e-01 5.88072896e-01 6.37376547e-01 7.34901071e-01 5.13794720e-01 -8.62043083e-01 -1.70976019e+00 -9.31259096e-01 1.17751205e+00 -5.89883506e-01 8.09189081e-01 3.82980034e-02 -6.95934832e-01 6.52496338e-01 -2.65150815e-01 3.75770599e-01 4.94024098e-01 4.59377915e-01 -2.77681530e-01 -2.16827631e-01 -1.36213124e+00 8.09203923e-01 1.94779599e+00 -3.52775007e-01 -4.14022267e-01 6.37113929e-01 7.65582085e-01 -1.51852801e-01 -9.92808878e-01 4.94820446e-01 5.91070533e-01 -1.16138518e+00 1.20195663e+00 -3.44861895e-01 -5.69082871e-02 -2.45560676e-01 -3.71327579e-01 -1.22904205e+00 -5.66453159e-01 -6.05812788e-01 -3.78960013e-01 7.62667716e-01 5.55096030e-01 -1.05746329e+00 1.12857103e+00 5.63810527e-01 2.03906953e-01 -1.06695855e+00 -8.64485562e-01 -1.28722739e+00 -3.03819608e-02 -1.33512706e-01 9.36873913e-01 1.08814895e+00 5.04955828e-01 2.91119307e-01 -1.99005291e-01 1.81797862e-01 7.68744469e-01 4.31635290e-01 3.24662894e-01 -1.53326893e+00 -1.78172722e-01 -5.23971796e-01 -7.48240411e-01 -8.13726246e-01 -1.84371956e-02 -1.15091300e+00 -4.25294518e-01 -1.87674427e+00 5.69252903e-03 -8.35177064e-01 -6.80243149e-02 4.08257008e-01 2.00572953e-01 3.42509747e-01 -2.99059421e-01 1.63216233e-01 -6.85323894e-01 1.33137971e-01 9.82637286e-01 -2.98856739e-02 -3.82640332e-01 7.27626607e-02 -8.42903972e-01 4.59971994e-01 8.69190872e-01 -5.88777900e-01 -7.27900624e-01 -2.30282545e-01 7.65772045e-01 1.80336148e-01 1.53756514e-01 -8.08918357e-01 7.51985133e-01 -2.30931252e-01 1.06026746e-01 -7.10250676e-01 5.64001873e-02 -1.12867188e+00 4.02743131e-01 6.78603947e-01 -3.91194969e-01 5.47603905e-01 -1.88617498e-01 7.47637033e-01 7.43423253e-02 -4.17641938e-01 2.51269430e-01 -7.92996585e-02 -5.90480685e-01 3.78854543e-01 7.86158964e-02 -9.20038298e-02 9.36285019e-01 -5.07002294e-01 -3.84990335e-01 -7.94754088e-01 -6.30357683e-01 3.16704065e-01 6.25627935e-01 -1.00619964e-01 8.04796338e-01 -1.72066987e+00 -3.98781747e-01 -1.68738768e-01 3.17282706e-01 -3.36947173e-01 -3.97501826e-01 5.91135859e-01 -7.90198982e-01 4.95186478e-01 2.56184369e-01 -3.53601843e-01 -1.14789844e+00 1.12621284e+00 1.66164890e-01 -2.81978339e-01 -4.37230796e-01 7.39820361e-01 -4.59090650e-01 -4.53795731e-01 1.99545220e-01 -2.10130736e-01 2.61711150e-01 -3.20564777e-01 1.58799201e-01 8.97664607e-01 -4.65188809e-02 -2.84112006e-01 -6.86009049e-01 9.05152678e-01 1.86438113e-01 1.25122532e-01 1.18175185e+00 -2.44903848e-01 -6.21567726e-01 -4.42291610e-02 1.23199594e+00 3.71328086e-01 -4.14563194e-02 -1.08572364e-01 3.44083816e-01 -5.69820940e-01 -1.92249835e-01 -3.79864812e-01 -9.70937014e-01 1.33046344e-01 1.20891124e-01 8.37347209e-01 1.12704659e+00 3.92988116e-01 8.09627116e-01 1.04686046e+00 7.35320389e-01 -9.32732880e-01 -2.68168360e-01 2.44289994e-01 7.03403890e-01 -9.03315842e-01 4.05561060e-01 -5.93499720e-01 1.30212471e-01 1.35124815e+00 3.99344206e-01 -2.23580673e-01 8.56318772e-01 1.27254993e-01 -5.33415437e-01 -4.77713764e-01 -5.88805914e-01 -2.31614172e-01 4.69322205e-01 7.25709975e-01 -2.86918730e-02 -6.52229190e-02 -5.52537203e-01 3.51809591e-01 -5.58444202e-01 -2.28335410e-01 2.68478364e-01 8.31104875e-01 -5.40758371e-01 -1.39044309e+00 -3.01746190e-01 6.48208320e-01 4.04983759e-02 -5.16411364e-01 -8.66747022e-01 9.56359804e-01 -3.05458874e-01 1.11082721e+00 6.54445216e-03 -7.88114190e-01 9.43328291e-02 -3.70679885e-01 8.02412093e-01 -1.04666024e-01 -1.88973784e-01 -6.00933850e-01 2.62781233e-01 -7.62798429e-01 -2.25986466e-01 -3.60394716e-02 -1.21490848e+00 -9.95297015e-01 -5.57833612e-01 9.76066470e-01 3.85448873e-01 6.67546868e-01 5.69104314e-01 -1.45138100e-01 5.63983142e-01 -2.17916608e-01 -4.11858171e-01 -3.90576690e-01 -7.82294273e-01 -2.32319489e-01 -1.25405518e-02 -2.48986110e-01 -6.75894201e-01 -8.09292257e-01]	[7.213820457458496, 5.977259159088135]
d2fc70cd-9e7e-46e2-a13d-e5b37f6db6d9	maximum-correntropy-value-decomposition-for	2208.03663	null	https://arxiv.org/abs/2208.03663v1	https://arxiv.org/pdf/2208.03663v1.pdf	Maximum Correntropy Value Decomposition for Multi-agent Deep Reinforcemen Learning	We explore value decomposition solutions for multi-agent deep reinforcement learning in the popular paradigm of centralized training with decentralized execution(CTDE). As the recognized best solution to CTDE, Weighted QMIX is cutting-edge on StarCraft Multi-agent Challenge (SMAC), with a weighting scheme implemented on QMIX to place more emphasis on the optimal joint actions. However, the fixed weight requires manual tuning according to the application scenarios, which painfully prevents Weighted QMIX from being used in broader engineering applications. In this paper, we first demonstrate the flaw of Weighted QMIX using an ordinary One-Step Matrix Game (OMG), that no matter how the weight is chosen, Weighted QMIX struggles to deal with non-monotonic value decomposition problems with a large variance of reward distributions. Then we characterize the problem of value decomposition as an Underfitting One-edged Robust Regression problem and make the first attempt to give a solution to the value decomposition problem from the perspective of information-theoretical learning. We introduce the Maximum Correntropy Criterion (MCC) as a cost function to dynamically adapt the weight to eliminate the effects of minimum in reward distributions. We simplify the implementation and propose a new algorithm called MCVD. A preliminary experiment conducted on OMG shows that MCVD could deal with non-monotonic value decomposition problems with a large tolerance of kernel bandwidth selection. Further experiments are carried out on Cooperative-Navigation and multiple SMAC scenarios, where MCVD exhibits unprecedented ease of implementation, broad applicability, and stability.	['Lingjiang Kong', 'Tianxian Zhang', 'Kai Liu']	2022-08-07	null	null	null	null	['smac-1', 'starcraft', 'smac']	['playing-games', 'playing-games', 'playing-games']	[-2.48617560e-01 -5.35745472e-02 -7.96019733e-02 2.47048423e-01 -7.35082388e-01 -3.97734523e-01 2.31406093e-01 -6.11156635e-02 -8.90841246e-01 1.10139823e+00 -1.06221057e-01 -3.31309557e-01 -9.37320292e-01 -5.39502859e-01 -5.49726009e-01 -9.08344805e-01 -5.32356083e-01 3.42613935e-01 1.52477533e-01 -6.10787094e-01 4.17056441e-01 2.59608090e-01 -1.39348829e+00 -1.17007360e-01 8.25661361e-01 8.77118707e-01 4.13320124e-01 7.63952315e-01 3.27226847e-01 9.17034924e-01 -9.38023388e-01 -2.03561932e-01 7.25521863e-01 -4.55119640e-01 -5.95880151e-01 -1.33382767e-01 -4.05910350e-02 -4.17526662e-01 1.13266192e-01 9.15964305e-01 8.01016510e-01 5.60942709e-01 5.32617807e-01 -1.56957567e+00 -2.12773204e-01 6.47044957e-01 -8.36534321e-01 2.42095843e-01 -1.27187267e-01 2.50544190e-01 1.09634185e+00 -2.58435726e-01 3.34764093e-01 1.03940260e+00 6.67453229e-01 4.99243528e-01 -1.20881712e+00 -3.25400144e-01 1.80584475e-01 2.52343863e-01 -1.06117105e+00 -3.90873030e-02 5.54859877e-01 -2.31592804e-01 1.03282356e+00 1.37260407e-01 6.73550069e-01 7.41945446e-01 3.89433503e-01 6.80249631e-01 1.01016891e+00 -3.25872868e-01 7.78333664e-01 -7.07443133e-02 -2.79425234e-01 7.05650330e-01 3.52693886e-01 3.73643309e-01 -4.07318115e-01 -3.07636559e-01 7.90141821e-01 -3.91343415e-01 -1.11762263e-01 -9.06362474e-01 -1.15407813e+00 1.21287084e+00 2.46305242e-01 1.29267257e-02 -6.42476976e-01 5.95856607e-01 4.85198438e-01 8.70466769e-01 1.52908444e-01 8.61116230e-01 -7.18065381e-01 -4.57147717e-01 -5.84559500e-01 6.46803141e-01 7.82234371e-01 5.83008885e-01 4.94411886e-01 4.77891952e-01 1.23294666e-01 8.52678061e-01 2.78245479e-01 5.77766001e-01 6.36138916e-01 -1.46973157e+00 5.40090621e-01 2.61641294e-01 3.09376389e-01 -7.89656341e-01 -7.05365479e-01 -7.11071610e-01 -3.83192956e-01 9.02998507e-01 6.80902898e-01 -7.95395076e-01 -2.04941884e-01 1.84549785e+00 3.16709280e-01 -2.72794515e-01 3.98665041e-01 1.08551562e+00 1.32774830e-01 3.21116060e-01 -2.10962370e-01 -3.36270005e-01 1.05303991e+00 -8.35342586e-01 -5.33512831e-01 -2.83001304e-01 6.87364221e-01 -2.58753628e-01 9.76463020e-01 6.79208279e-01 -8.72448921e-01 -1.42880529e-01 -1.32813966e+00 6.49651587e-01 2.14423030e-03 -1.50364578e-01 7.16645420e-01 7.53924549e-01 -1.01267028e+00 8.00346553e-01 -8.16000998e-01 -4.79686037e-02 -2.87041869e-02 5.30718207e-01 -1.25383601e-01 3.78943264e-01 -1.16279101e+00 1.15716422e+00 4.15771574e-01 -1.90972030e-01 -9.71667886e-01 -5.89481294e-01 -5.99605858e-01 1.62816957e-01 8.54522407e-01 -4.63145882e-01 1.31807625e+00 -1.01126695e+00 -1.97051978e+00 -1.58198941e-02 6.84317768e-01 -7.03673303e-01 5.49116135e-01 -1.77179545e-01 4.90390323e-02 1.16237074e-01 -1.13820685e-02 3.87408793e-01 9.10374463e-01 -1.09182489e+00 -6.85258687e-01 -1.23951159e-01 2.96573162e-01 5.32486081e-01 -2.24647596e-01 -3.19720745e-01 2.39828736e-01 -4.69582886e-01 -2.12292954e-01 -9.74560440e-01 -5.51841557e-01 -4.75596756e-01 1.74475491e-01 3.13235782e-02 3.63112926e-01 -2.36550584e-01 1.01916909e+00 -1.97149801e+00 5.03576458e-01 2.57439137e-01 2.06653178e-01 -1.47466287e-01 -4.57246989e-01 6.00802720e-01 1.04131617e-01 -2.74586588e-01 -1.55630037e-01 8.23840126e-02 2.52048135e-01 3.67377639e-01 -1.26891613e-01 5.59736311e-01 -1.36168778e-01 4.95704114e-01 -9.43256497e-01 -1.75309822e-01 -8.19604993e-02 4.63973321e-02 -1.04633474e+00 -2.32947208e-02 -3.34261805e-01 3.33667211e-02 -5.01782119e-01 3.96002084e-01 3.74705374e-01 -4.57410477e-02 4.34641063e-01 1.72330618e-01 -2.44245842e-01 -3.36799324e-01 -1.50801957e+00 1.49511027e+00 -3.67050141e-01 4.09264624e-01 4.70714033e-01 -1.16375899e+00 8.13680053e-01 1.46896066e-02 9.79628086e-01 -6.80964172e-01 1.07488014e-01 1.38529420e-01 3.88922572e-01 -3.92777443e-01 8.75207722e-01 -8.56517032e-02 -2.10040227e-01 4.60105628e-01 1.59349337e-01 -1.14903897e-01 2.96459556e-01 1.53211623e-01 1.28863966e+00 2.43172914e-01 4.11598593e-01 -4.87042874e-01 2.17194349e-01 9.25665796e-02 6.49393260e-01 8.82000208e-01 -5.59127510e-01 2.81611502e-01 8.23824644e-01 -1.86592609e-01 -8.70217621e-01 -9.06716347e-01 1.85638160e-01 1.51410568e+00 1.31049171e-01 -2.72188634e-01 -7.58942544e-01 -6.52631342e-01 2.93856144e-01 6.32686853e-01 -4.57562566e-01 -8.17481279e-02 -3.66976142e-01 -1.18367314e+00 3.61075491e-01 1.25087649e-01 4.64268327e-01 -9.98999298e-01 -1.39424896e+00 4.19038713e-01 -3.56940366e-02 -6.57209277e-01 -3.78938556e-01 6.45541489e-01 -6.58659875e-01 -1.04509783e+00 -7.68895984e-01 -2.95615762e-01 2.25241691e-01 2.08899170e-01 8.16693008e-01 -1.73208818e-01 -7.47876689e-02 8.97474527e-01 -4.56393987e-01 -3.58564794e-01 -2.11112946e-01 3.42598855e-02 3.42603654e-01 -2.94433832e-01 -1.63916990e-01 -4.29578125e-01 -5.61544538e-01 3.76251847e-01 -7.06254363e-01 -2.57159054e-01 7.15623200e-01 1.16912436e+00 9.38760191e-02 1.05742231e-01 1.01943028e+00 -4.67872381e-01 1.22935426e+00 -6.01596534e-01 -9.33366239e-01 1.91982225e-01 -8.81415904e-01 4.19345945e-01 5.49244940e-01 -6.16874814e-01 -7.56038129e-01 -9.37087387e-02 1.97788045e-01 -3.72658670e-01 4.92258728e-01 7.29725540e-01 2.59191304e-01 -2.17647836e-01 7.83483207e-01 1.34851366e-01 5.34503996e-01 8.97035748e-03 4.13405865e-01 2.65379161e-01 1.40599594e-01 -7.49093771e-01 3.94157350e-01 -1.17462471e-01 1.18474714e-01 -7.67900229e-01 -2.78634191e-01 -5.75701036e-02 -1.44420817e-01 -4.22475547e-01 5.73396862e-01 -7.63389647e-01 -1.21938038e+00 2.22482070e-01 -7.48129249e-01 -8.03234160e-01 -3.41703266e-01 7.06984699e-01 -1.09782445e+00 3.40850562e-01 -4.63448852e-01 -9.97537911e-01 -1.82009056e-01 -1.09160984e+00 3.70329499e-01 1.57103464e-01 1.69340372e-01 -8.88500988e-01 3.89006138e-01 2.37973273e-01 6.31492674e-01 2.05887899e-01 6.95396066e-01 -6.11112058e-01 -3.44970554e-01 1.02145210e-01 2.43825272e-01 3.22630078e-01 -3.07493865e-01 -2.76905447e-01 -4.82480347e-01 -7.49940395e-01 7.13779777e-02 -5.53656757e-01 5.40871263e-01 4.44797218e-01 5.15529633e-01 -2.22629592e-01 3.81604195e-01 3.92654836e-01 1.56817472e+00 3.64432812e-01 3.25356543e-01 9.44245994e-01 1.58300787e-01 4.20685679e-01 8.54515493e-01 1.08377981e+00 3.41474861e-01 6.27957702e-01 9.69819307e-01 3.39380622e-01 4.42939341e-01 4.61790003e-02 7.82040656e-01 5.97963452e-01 -2.08891034e-01 -3.17918994e-02 -6.93547249e-01 1.88187361e-01 -2.28282619e+00 -1.11413908e+00 2.40025133e-01 2.36979985e+00 6.91301167e-01 9.14841071e-02 4.39474732e-01 -1.31178036e-01 4.16463673e-01 1.04132257e-01 -7.12530673e-01 -4.74207044e-01 6.25388846e-02 -1.20272376e-01 7.86633313e-01 4.81134087e-01 -9.25719202e-01 6.84528470e-01 6.43953514e+00 9.53864217e-01 -8.55971217e-01 2.30413720e-01 9.65095386e-02 -5.01267135e-01 5.47311036e-03 -9.87143517e-02 -4.44116205e-01 4.96532500e-01 8.63927603e-01 -3.31983268e-01 9.08113062e-01 1.10224974e+00 2.42787600e-01 -4.15234923e-01 -6.47881091e-01 1.01091254e+00 -2.58888513e-01 -9.55310762e-01 -5.51846266e-01 2.60362059e-01 4.96583670e-01 3.49221498e-01 1.89946089e-02 7.72603631e-01 7.98011243e-01 -6.88210130e-01 7.95134068e-01 2.66534418e-01 3.41013014e-01 -9.84738827e-01 6.85854673e-01 5.70021152e-01 -8.83868337e-01 -6.77902043e-01 -6.94012403e-01 -2.87389427e-01 -2.31869251e-01 1.19488277e-01 -5.80075443e-01 5.54836869e-01 5.14503658e-01 1.87884077e-01 -2.39054248e-01 9.91905332e-01 9.85136330e-02 1.85316443e-01 -2.57746011e-01 -4.53233927e-01 4.57679361e-01 -2.92829931e-01 5.85889399e-01 7.33119965e-01 3.41074497e-01 -5.07939644e-02 2.92050481e-01 5.06316960e-01 4.28453445e-01 1.06483206e-01 -3.84646356e-01 1.06834002e-01 3.79990369e-01 1.29451144e+00 -6.05481386e-01 6.29690215e-02 -3.98328424e-01 6.13160372e-01 5.83821654e-01 3.25460941e-01 -9.59910095e-01 -4.28944677e-01 7.33985901e-01 -3.35535258e-01 4.52748120e-01 -4.04493183e-01 -4.41556722e-02 -1.12812543e+00 -2.38403067e-01 -1.13865089e+00 4.71463650e-01 -3.96536469e-01 -1.09502554e+00 4.69238073e-01 1.34683549e-01 -1.46214545e+00 -6.30892098e-01 -5.91255486e-01 -3.98518234e-01 4.10712779e-01 -1.26771867e+00 -4.82167631e-01 1.25618890e-01 5.61645925e-01 4.55532581e-01 -7.20151901e-01 6.85496390e-01 5.93132973e-02 -5.02180338e-01 4.95263964e-01 5.73679566e-01 -3.67005736e-01 5.24383843e-01 -1.48732102e+00 -3.23606938e-01 5.71489453e-01 -1.91542149e-01 2.49032766e-01 1.03419030e+00 -3.84249210e-01 -1.59609973e+00 -5.70733190e-01 -1.54549539e-01 -1.10971428e-01 8.53117883e-01 -5.51062897e-02 -5.32915294e-01 2.37699538e-01 3.38679731e-01 -2.38439366e-01 4.94259506e-01 2.04351500e-01 2.37822533e-02 -2.37488449e-01 -1.12429810e+00 5.13175905e-01 5.02914250e-01 -2.51658708e-02 -4.32346880e-01 1.69094548e-01 5.13126135e-01 -2.29336694e-01 -9.35013592e-01 2.19953388e-01 5.37112832e-01 -1.03525412e+00 7.14579105e-01 -5.17834961e-01 2.15367004e-01 -1.91132098e-01 -3.17815304e-01 -1.87049556e+00 -3.98809135e-01 -7.65069485e-01 -1.61837250e-01 7.28129387e-01 2.91709214e-01 -6.46403790e-01 8.37019861e-01 5.02009153e-01 -1.06159873e-01 -8.81766617e-01 -1.13872230e+00 -1.00369656e+00 1.60955980e-01 -2.44511887e-01 1.54064611e-01 7.90947616e-01 3.81976128e-01 2.00417235e-01 -7.33423173e-01 -1.71230175e-03 8.17827940e-01 -3.08418516e-02 8.88289571e-01 -9.64166164e-01 -8.94209862e-01 -4.86148059e-01 -1.49277031e-01 -7.73179173e-01 -1.30638152e-01 -4.80169922e-01 1.67443320e-01 -1.15086865e+00 -3.85862552e-02 -2.77121753e-01 -4.47673112e-01 3.93600553e-01 1.01177007e-01 -2.93269187e-01 7.11580396e-01 1.64283201e-01 -8.86208236e-01 8.61682236e-01 1.15509713e+00 -9.15952921e-02 -4.77501243e-01 8.60291868e-02 -6.26223087e-01 4.84722644e-01 1.02616715e+00 -4.82794791e-01 -7.82063365e-01 -2.88514525e-01 4.09553707e-01 5.51073492e-01 1.48626015e-01 -1.11054623e+00 2.15212420e-01 -4.96609747e-01 8.99114981e-02 -9.03068259e-02 2.49518543e-01 -7.84698665e-01 -2.38577072e-02 7.37559259e-01 -4.05493647e-01 4.07635391e-01 1.69295996e-01 6.24417901e-01 3.90303768e-02 -5.31773329e-01 7.54218400e-01 -1.94683492e-01 -7.94120908e-01 -7.99015760e-02 -7.19803214e-01 1.61618665e-01 1.16316211e+00 6.47039860e-02 -4.16751802e-01 -5.59670150e-01 -5.87477863e-01 6.76633120e-01 2.08901972e-01 2.22821444e-01 6.46434069e-01 -1.05149007e+00 -5.80248415e-01 3.67903872e-03 -2.06595168e-01 -4.98156726e-01 2.99685180e-01 8.09463739e-01 -5.08697331e-01 8.09572786e-02 -5.21515071e-01 -2.91880786e-01 -1.00507224e+00 5.16110003e-01 6.91366911e-01 -5.62074661e-01 -4.07379448e-01 4.13803935e-01 -2.32677445e-01 -4.55326170e-01 3.42812479e-01 -1.30388901e-01 -1.81170061e-01 1.57498926e-01 2.87143409e-01 4.14403111e-01 -1.12845272e-01 1.74900927e-02 -1.73229724e-01 2.23118871e-01 -2.97573470e-02 -5.24400651e-01 1.48462152e+00 -1.51470333e-01 2.69229233e-01 1.34846777e-01 6.28169835e-01 -1.59731314e-01 -1.76321959e+00 -5.14927283e-02 2.19351724e-01 -2.61527747e-01 2.02020049e-01 -8.48793685e-01 -1.06539750e+00 4.26967382e-01 7.09716976e-01 2.90603369e-01 9.27576900e-01 -5.41543245e-01 3.35202038e-01 8.60171139e-01 6.85556293e-01 -1.60890388e+00 4.97176528e-01 8.25079024e-01 9.26063895e-01 -1.07707000e+00 1.85509473e-01 4.04151171e-01 -1.22049582e+00 1.18937624e+00 8.08771253e-01 -2.76500076e-01 4.89136457e-01 2.98480064e-01 8.59641507e-02 -7.29000196e-02 -1.04778337e+00 -2.91319132e-01 -2.68269420e-01 5.57988644e-01 -9.34490412e-02 6.31295098e-03 -5.11053026e-01 5.38048804e-01 -1.77978322e-01 -2.15218723e-01 7.79580355e-01 1.03753984e+00 -7.79914618e-01 -9.07718539e-01 -4.38425928e-01 2.78041989e-01 -3.36617619e-01 1.86750904e-01 1.13459580e-01 1.00711691e+00 -1.85011700e-01 9.53973293e-01 -1.87423974e-01 -5.23024142e-01 3.06864351e-01 -1.54746056e-01 4.37457770e-01 -1.13316946e-01 -7.39349186e-01 1.76586494e-01 -2.44542193e-02 -6.44718468e-01 -2.93134212e-01 -6.16464972e-01 -1.27887893e+00 -3.42939645e-01 -1.97278515e-01 3.74915689e-01 7.08553612e-01 7.77220547e-01 2.78939933e-01 6.66269600e-01 6.31369710e-01 -7.91187704e-01 -1.45663750e+00 -8.67293954e-01 -9.09960866e-01 1.62282974e-01 4.49783176e-01 -9.77935076e-01 -3.71232241e-01 -7.44799733e-01]	[3.8630733489990234, 2.118299961090088]
850c247a-eb08-4bef-a8ff-9e130fa331dd	matching-adversarial-networks	null	null	http://openaccess.thecvf.com/content_cvpr_2018/html/Mattyus_Matching_Adversarial_Networks_CVPR_2018_paper.html	http://openaccess.thecvf.com/content_cvpr_2018/papers/Mattyus_Matching_Adversarial_Networks_CVPR_2018_paper.pdf	Matching Adversarial Networks	Generative Adversarial Nets (GANs) and Conditonal GANs (CGANs) show that using a trained network as loss function (discriminator) enables to synthesize highly structured outputs (e.g. natural images). However, applying a discriminator network as a universal loss function for common supervised tasks (e.g. semantic segmentation, line detection, depth estimation) is considerably less successful. We argue that the main difficulty of applying CGANs to supervised tasks is that the generator training consists of optimizing a loss function that does not depend directly on the ground truth labels. To overcome this, we propose to replace the discriminator with a matching network taking into account both the ground truth outputs as well as the generated examples. As a consequence, the generator loss function also depends on the targets of the training examples, thus facilitating learning. We demonstrate on three computer vision tasks that this approach can significantly outperform CGANs achieving comparable or superior results to task-specific solutions and results in stable training. Importantly, this is a general approach that does not require the use of task-specific loss functions.	['GellÃ©rt MÃ¡ttyus', 'Raquel Urtasun']	2018-06-01	null	null	null	cvpr-2018-6	['line-detection']	['computer-vision']	[ 5.74604213e-01 4.69906092e-01 1.57575071e-01 -3.07327062e-01 -9.07408416e-01 -6.95279062e-01 7.95888901e-01 -5.33678681e-02 -4.17191505e-01 9.30942595e-01 -2.89530933e-01 -7.54913017e-02 2.34329298e-01 -9.86919582e-01 -9.26278949e-01 -8.59760284e-01 5.34525871e-01 6.79715097e-01 2.64186472e-01 -9.22066867e-02 -2.70909872e-02 5.75510800e-01 -1.33238947e+00 -3.14511731e-02 1.01735079e+00 9.50295150e-01 1.48598626e-01 6.25774264e-01 1.45335430e-02 6.47220314e-01 -9.75753009e-01 -5.23899496e-01 3.96895468e-01 -9.37085032e-01 -7.46069372e-01 1.78904802e-01 5.62351346e-01 -8.49795714e-02 -4.64230217e-02 9.50977921e-01 3.84458363e-01 1.09641112e-01 1.02260005e+00 -1.20733500e+00 -4.15536374e-01 4.03212667e-01 -1.97068706e-01 -2.89119571e-01 -1.83955710e-02 1.48873046e-01 9.61336553e-01 -4.83447343e-01 6.56898320e-01 1.01947081e+00 6.77308559e-01 7.68437088e-01 -1.55224276e+00 -3.10843080e-01 -1.31783411e-01 -3.38342547e-01 -1.04381633e+00 -2.35316187e-01 8.03598821e-01 -4.44958478e-01 3.47051919e-01 1.11959942e-01 3.47957909e-01 1.25860739e+00 5.43616712e-03 7.56921411e-01 1.04412091e+00 -5.73515475e-01 5.22569299e-01 3.94498050e-01 -4.58994389e-01 6.34786069e-01 1.61913514e-01 1.24883778e-01 -2.33082604e-02 1.83460742e-01 9.34700668e-01 -4.53578144e-01 -3.84310603e-01 -5.97045183e-01 -9.09379065e-01 1.02487469e+00 7.01057196e-01 4.37392175e-01 -2.95517176e-01 3.26683700e-01 2.93290734e-01 2.68464923e-01 3.23453039e-01 8.54254067e-01 -1.01974428e-01 2.50774235e-01 -9.31566000e-01 2.27461725e-01 7.14748323e-01 7.12198257e-01 9.80668306e-01 3.88289779e-01 -3.29821199e-01 6.87270641e-01 -9.87173337e-03 3.40457618e-01 5.68103731e-01 -1.00257003e+00 3.23717147e-01 5.33972621e-01 -8.16203281e-03 -6.87973738e-01 -1.15081824e-01 -7.40469515e-01 -8.20281684e-01 7.08446324e-01 9.08785462e-01 -3.02245498e-01 -1.11087537e+00 2.11558127e+00 2.06489936e-01 -3.08824945e-02 1.76059410e-01 6.99215710e-01 4.02833611e-01 5.22110581e-01 -1.11869182e-02 1.19522139e-01 8.39371979e-01 -9.35353994e-01 -2.80768245e-01 -5.34089148e-01 5.87269068e-01 -5.20525932e-01 1.15966380e+00 3.51426959e-01 -1.09767699e+00 -5.58557272e-01 -1.06384778e+00 -1.01011261e-01 -3.93808216e-01 3.87809366e-01 3.90804917e-01 6.58680201e-01 -1.01616013e+00 7.14621902e-01 -5.86206436e-01 -2.45067298e-01 4.73127961e-01 2.85289556e-01 -2.14219913e-01 1.08487912e-01 -8.66070867e-01 8.48536432e-01 6.16384327e-01 8.77088867e-03 -1.07558846e+00 -6.77608728e-01 -8.61940980e-01 8.37240964e-02 2.92555660e-01 -9.57203805e-01 1.14307702e+00 -1.59063256e+00 -1.73287511e+00 9.89951015e-01 2.83973843e-01 -7.24756420e-01 1.16071355e+00 9.80965421e-02 2.73414999e-01 1.44120231e-01 7.87947923e-02 1.01392400e+00 1.03832984e+00 -1.37577045e+00 -3.82655114e-01 -1.18166737e-01 1.33209020e-01 2.88184248e-02 2.30857804e-02 -5.12246847e-01 -6.71412870e-02 -7.74219453e-01 -2.07478374e-01 -8.79323542e-01 -2.84670591e-01 2.13807866e-01 -5.65359414e-01 2.73286388e-03 6.21844769e-01 -5.06375849e-01 5.06966650e-01 -2.15467429e+00 3.66799474e-01 2.66821444e-01 -1.82215255e-02 3.61796260e-01 -1.92322463e-01 2.20008492e-01 6.10973202e-02 -6.34976029e-02 -7.58675158e-01 -6.03673816e-01 -3.21252048e-02 3.15813720e-01 -1.68515369e-01 3.67636770e-01 4.69257683e-01 1.09086359e+00 -8.72269928e-01 -2.20006347e-01 3.13929737e-01 5.33521950e-01 -4.40808415e-01 3.58836234e-01 -5.90874851e-01 8.95289361e-01 -3.36000413e-01 6.10832721e-02 3.36190790e-01 -1.95318073e-01 5.63451312e-02 -1.41913807e-02 1.50408998e-01 2.24749640e-01 -8.92497957e-01 1.71800959e+00 -7.79646218e-01 7.44776368e-01 -2.30017975e-02 -1.34962893e+00 1.14402235e+00 2.02644646e-01 1.91333532e-01 -5.62471211e-01 2.02676356e-01 4.39463556e-01 -3.34073231e-02 -1.06004570e-02 -1.10936295e-02 -3.95041168e-01 -2.41406392e-02 3.88949811e-01 4.26103771e-01 -5.15764117e-01 1.85638532e-01 -3.99428122e-02 8.59528005e-01 3.79992515e-01 1.64867550e-01 -1.79664955e-01 6.12615407e-01 -8.37601572e-02 4.93678451e-01 6.52372956e-01 2.38354862e-01 1.12946987e+00 8.09140265e-01 -1.28452927e-01 -1.36457229e+00 -9.13774312e-01 5.40648773e-02 6.05982065e-01 -1.39088795e-01 4.57977317e-02 -1.25521922e+00 -8.66400242e-01 -1.93705946e-01 7.44767547e-01 -8.15750003e-01 -2.33364284e-01 -6.13776147e-01 -5.69310784e-01 5.25336146e-01 6.83865368e-01 5.38111687e-01 -1.23258722e+00 -6.22668386e-01 1.39187321e-01 1.06407061e-01 -1.25533164e+00 -2.60829180e-01 3.81550163e-01 -8.88507247e-01 -8.79922748e-01 -9.81001675e-01 -6.33514285e-01 9.39576566e-01 -3.07344705e-01 1.23360968e+00 -1.67656675e-01 -2.09059045e-01 3.16679329e-01 -1.37895226e-01 -2.05273047e-01 -7.01479077e-01 3.24939787e-01 -5.96003175e-01 2.71160036e-01 -2.25385085e-01 -7.21696556e-01 -4.91966963e-01 1.85503051e-01 -9.72004116e-01 1.65433794e-01 6.28925025e-01 1.02099466e+00 6.05498731e-01 -1.41710967e-01 6.42799973e-01 -1.33794498e+00 4.10796970e-01 -7.82642290e-02 -8.03163409e-01 2.39860222e-01 -6.21192873e-01 2.99784213e-01 9.24202502e-01 -3.25212687e-01 -1.03566265e+00 2.66562700e-01 -3.35614234e-01 -3.13366055e-01 -3.42159688e-01 1.12263396e-01 -3.34246129e-01 -9.39230919e-02 9.57096756e-01 2.32341409e-01 1.57820240e-01 -3.17358971e-01 3.10375363e-01 1.17580451e-01 6.35485351e-01 -5.37439346e-01 9.33066785e-01 4.44418848e-01 3.57401431e-01 -6.34356737e-01 -6.40256464e-01 1.24242688e-02 -5.45675159e-01 5.34496875e-03 1.02425754e+00 -6.58749461e-01 -2.88492203e-01 4.19373393e-01 -1.08879054e+00 -8.12821329e-01 -8.39859426e-01 3.45247686e-01 -9.87149298e-01 3.28234732e-02 -4.56014842e-01 -6.27539575e-01 -2.85882831e-01 -9.65318203e-01 1.01550758e+00 2.25286290e-01 -6.68152422e-02 -1.31370306e+00 -1.22816317e-01 1.19139448e-01 3.33810389e-01 8.10375631e-01 9.67499495e-01 -5.29572189e-01 -7.23308563e-01 -2.37354890e-01 -1.81286156e-01 8.61736357e-01 6.83926120e-02 -7.46625662e-02 -1.04809320e+00 -1.42926276e-01 1.05107479e-01 -4.89690214e-01 8.58744323e-01 3.93669099e-01 9.92809057e-01 -2.67346144e-01 -8.18870682e-03 7.38443255e-01 1.55845571e+00 9.08368547e-03 8.75793874e-01 2.47970447e-01 8.51394713e-01 6.73332512e-01 2.05512598e-01 -5.97945154e-02 4.47532684e-02 8.30785275e-01 5.38953424e-01 -3.91499788e-01 -4.55805987e-01 -4.93193954e-01 2.66472727e-01 1.97041146e-02 -6.16990253e-02 -3.66048723e-01 -7.37596214e-01 4.26492721e-01 -1.75132871e+00 -6.83869183e-01 -3.39944288e-02 2.27048659e+00 7.70707548e-01 2.36956775e-01 2.77086914e-01 2.79458523e-01 5.89494169e-01 -4.78002280e-02 -6.51024520e-01 -4.10988122e-01 -2.91417509e-01 5.02569079e-01 5.13795257e-01 5.20414829e-01 -9.37532187e-01 7.93990314e-01 5.74703550e+00 7.34904647e-01 -1.31702030e+00 2.24867836e-01 7.75525749e-01 3.70685965e-01 -4.19977874e-01 -1.36886448e-01 -4.55269307e-01 4.99763459e-01 6.34761453e-01 2.32534315e-02 2.87633538e-01 9.16609764e-01 -8.05838481e-02 2.60323156e-02 -1.22532320e+00 7.22254395e-01 5.50136901e-02 -1.09809971e+00 1.24091357e-01 -1.84062123e-02 9.56540048e-01 -3.06109875e-01 2.04069838e-01 6.60958365e-02 2.42084712e-01 -1.12801802e+00 9.11150336e-01 2.20702484e-01 8.21907103e-01 -5.98985493e-01 7.15050220e-01 3.82309020e-01 -6.60896599e-01 2.39737183e-01 -2.46029913e-01 2.08396375e-01 7.02584088e-02 5.72712958e-01 -9.28393483e-01 6.23476744e-01 1.27151370e-01 4.96239364e-01 -4.32184279e-01 9.50478673e-01 -7.06147492e-01 5.32362282e-01 -1.94578245e-01 3.07510763e-01 5.05250096e-01 -2.98219442e-01 4.10322160e-01 1.14605510e+00 2.85630435e-01 -4.83370483e-01 -7.27729425e-02 1.22794008e+00 -1.18993320e-01 4.57354188e-02 -6.33686066e-01 4.52655070e-02 -5.47697768e-02 1.15464473e+00 -8.27078581e-01 -7.71351084e-02 -1.54637009e-01 1.10613632e+00 3.02420646e-01 4.65277344e-01 -6.87547922e-01 -3.44325721e-01 3.43670875e-01 2.58750111e-01 4.34109241e-01 -3.71604078e-02 -4.47807014e-01 -9.78753388e-01 3.92558686e-02 -7.51813233e-01 9.27348584e-02 -6.32871091e-01 -1.25668430e+00 8.29616845e-01 -2.70825773e-01 -1.25835776e+00 -7.61728525e-01 -6.79951131e-01 -8.71657312e-01 9.76650894e-01 -1.56650054e+00 -1.24217916e+00 -4.14961666e-01 5.54734349e-01 3.45192552e-01 -3.03894542e-02 7.10839450e-01 1.31579846e-01 -3.43745828e-01 6.65019333e-01 -8.84544179e-02 2.67828584e-01 4.76128548e-01 -1.59399414e+00 2.31375605e-01 8.90422404e-01 2.16084823e-01 1.73949778e-01 8.22827041e-01 -1.76885441e-01 -7.14112639e-01 -1.05785179e+00 6.16652548e-01 -2.87138432e-01 3.52811694e-01 -5.57352781e-01 -8.40344727e-01 6.56205416e-01 1.48698524e-01 6.92988336e-02 3.01003546e-01 -2.52567619e-01 -1.75011426e-01 -1.51946917e-01 -1.27309823e+00 2.84619391e-01 8.70511591e-01 -3.99476141e-01 -2.19697312e-01 2.65750766e-01 3.48241568e-01 -3.97645354e-01 -3.99145633e-01 4.76063728e-01 1.83747739e-01 -1.18771446e+00 7.59827495e-01 -4.00746852e-01 6.52200937e-01 -1.96733564e-01 1.25312492e-01 -1.36306942e+00 1.62491888e-01 -4.09906805e-01 1.26113728e-01 1.27510834e+00 4.53560889e-01 -7.89363682e-01 9.30043638e-01 3.39397579e-01 -1.90251932e-01 -7.41043806e-01 -7.68311381e-01 -8.99162531e-01 3.23277742e-01 6.86183050e-02 3.12713563e-01 7.50414550e-01 -7.46041477e-01 5.01436830e-01 -2.56142378e-01 -1.64794967e-01 6.33202434e-01 1.07408129e-01 9.01624739e-01 -1.25875449e+00 -5.30649304e-01 -4.98152614e-01 -4.08226669e-01 -1.06142199e+00 3.75951290e-01 -8.10731232e-01 1.67164579e-01 -1.40343821e+00 -3.56536031e-01 -8.14727902e-01 1.67582650e-02 4.20738161e-01 -4.71641757e-02 4.62005705e-01 1.45790115e-01 -1.42466361e-02 -7.96107277e-02 5.88669002e-01 1.49035347e+00 -6.41302094e-02 -9.99546275e-02 3.28099877e-01 -5.44599593e-01 6.36826515e-01 8.46979320e-01 -4.26848680e-01 -6.38318777e-01 -4.69737947e-01 1.32272601e-01 -1.13474905e-01 6.43969774e-01 -1.11012053e+00 4.70091142e-02 8.09488744e-02 3.21446538e-01 1.29232779e-01 3.30410302e-01 -6.84923530e-01 2.42890164e-01 4.50744063e-01 -3.37144464e-01 -3.96667421e-01 -4.57328465e-03 3.98730516e-01 -4.07511652e-01 -6.56742096e-01 9.69282031e-01 -2.46993154e-01 -3.22529793e-01 1.36693031e-01 1.14318887e-02 1.84885666e-01 9.56955016e-01 -3.98319006e-01 -1.39250666e-01 -5.29067039e-01 -6.86997294e-01 -1.40030906e-01 6.75486267e-01 4.93103229e-02 2.34477296e-01 -1.12700164e+00 -6.85674608e-01 1.88495934e-01 -2.22215861e-01 4.19769496e-01 -1.17490314e-01 4.66176927e-01 -6.11279786e-01 3.37748915e-01 -3.82221133e-01 -5.49320221e-01 -7.69207180e-01 3.48812878e-01 6.87102735e-01 -4.11742061e-01 -5.25683463e-01 9.01538253e-01 5.56970358e-01 -3.05729598e-01 1.40501827e-01 -1.86388969e-01 9.92451981e-02 -6.96196854e-02 1.09302744e-01 1.15494847e-01 1.37650758e-01 -2.98115373e-01 -5.53435739e-03 5.37461877e-01 1.60287738e-01 -1.15107849e-01 1.26625121e+00 2.09449261e-01 2.33236119e-01 3.27558905e-01 1.05638719e+00 -9.74268615e-02 -1.70085514e+00 -1.02689043e-01 -1.44029409e-01 -2.85814971e-01 -1.17942154e-01 -6.98613763e-01 -1.37879777e+00 9.32809412e-01 3.22891653e-01 2.72882015e-01 1.24774766e+00 -3.13446112e-02 5.65571487e-01 -3.92091423e-02 3.99570435e-01 -7.86569655e-01 2.91902214e-01 1.53315589e-01 8.26772511e-01 -1.13270533e+00 -4.32023972e-01 -5.41938663e-01 -5.69379687e-01 1.00753057e+00 4.82100725e-01 -5.19387782e-01 1.32640809e-01 1.21014744e-01 1.90812752e-01 1.21399142e-01 -2.99648762e-01 -3.95237356e-01 3.21728498e-01 7.81390131e-01 2.84382552e-01 -1.00400262e-01 -4.31658924e-01 1.25170827e-01 -3.37417394e-01 -1.42229721e-01 3.68858993e-01 6.45514965e-01 -6.69174343e-02 -1.51329684e+00 -8.18982124e-02 2.30108634e-01 -3.93884152e-01 7.64498934e-02 -4.57587034e-01 7.86745071e-01 2.98962981e-01 4.64671969e-01 6.88243508e-02 -3.75770545e-03 1.54769808e-01 7.63262436e-02 7.28725791e-01 -8.18014979e-01 -6.17769003e-01 -1.61192954e-01 -2.40244821e-01 -3.19180220e-01 -3.65119487e-01 -4.46826696e-01 -9.87184286e-01 -1.86548755e-02 -2.45068371e-01 2.01841146e-01 7.58131742e-01 7.71231949e-01 5.46643175e-02 6.44457877e-01 6.00678861e-01 -6.99663401e-01 -6.86239898e-01 -7.80908823e-01 -4.73515332e-01 6.45020247e-01 3.16600531e-01 -5.21577179e-01 -4.78375405e-01 1.45449206e-01]	[11.597972869873047, -0.38413962721824646]
9c2dab79-f1f9-45c2-ad18-b898a175e7ff	a-principled-design-of-image-representation	2203.00913	null	https://arxiv.org/abs/2203.00913v4	https://arxiv.org/pdf/2203.00913v4.pdf	A Principled Design of Image Representation: Towards Forensic Tasks	Image forensics is a rising topic as the trustworthy multimedia content is critical for modern society. Like other vision-related applications, forensic analysis relies heavily on the proper image representation. Despite the importance, current theoretical understanding for such representation remains limited, with varying degrees of neglect for its key role. For this gap, we attempt to investigate the forensic-oriented image representation as a distinct problem, from the perspectives of theory, implementation, and application. Our work starts from the abstraction of basic principles that the representation for forensics should satisfy, especially revealing the criticality of robustness, interpretability, and coverage. At the theoretical level, we propose a new representation framework for forensics, called Dense Invariant Representation (DIR), which is characterized by stable description with mathematical guarantees. At the implementation level, the discrete calculation problems of DIR are discussed, and the corresponding accurate and fast solutions are designed with generic nature and constant complexity. We demonstrate the above arguments on the dense-domain pattern detection and matching experiments, providing comparison results with state-of-the-art descriptors. Also, at the application level, the proposed DIR is initially explored in passive and active forensics, namely copy-move forgery detection and perceptual hashing, exhibiting the benefits in fulfilling the requirements of such forensic tasks.	['Xiaochun Cao', 'Jiantao Zhou', 'Chao Wang', 'Yushu Zhang', 'Shuren Qi']	2022-03-02	null	null	null	null	['image-forensics']	['computer-vision']	[ 3.33190441e-01 -3.23886424e-01 1.87556110e-02 1.49501666e-01 -7.83803582e-01 -4.86420363e-01 3.77445519e-01 1.51582167e-01 -3.24467510e-01 3.95041227e-01 -1.86410055e-01 -1.71850860e-01 -5.35395324e-01 -6.45177007e-01 -3.03282171e-01 -1.04567063e+00 -1.01242535e-01 6.61360100e-02 2.99971282e-01 -1.14130482e-01 6.08906865e-01 9.16202486e-01 -1.57696640e+00 -2.01570569e-03 5.67854822e-01 1.24107444e+00 2.29769628e-02 2.78553545e-01 4.92074788e-02 6.26733601e-01 -5.80622554e-01 -9.69441473e-01 3.98306608e-01 -2.42826119e-01 -5.49820960e-01 2.24249035e-01 1.77345723e-02 -5.35127878e-01 -7.18156338e-01 1.36900795e+00 6.35383904e-01 -5.98858297e-02 4.88509089e-01 -1.53441739e+00 -8.29746366e-01 1.11754090e-01 -7.37100959e-01 4.44015384e-01 3.65681142e-01 2.04004183e-01 6.66539609e-01 -8.39547038e-01 7.03497231e-01 1.06822407e+00 5.04567385e-01 2.24647298e-01 -6.72280312e-01 -4.91816819e-01 -2.45837390e-01 9.19246614e-01 -1.66596234e+00 -4.82876241e-01 8.99514914e-01 -3.31856221e-01 2.41022065e-01 2.26536050e-01 3.67546171e-01 8.37580264e-01 1.24197572e-01 6.28589869e-01 1.17144394e+00 -4.72635031e-01 1.75329223e-01 3.26809317e-01 5.36188543e-01 4.53016728e-01 8.95807326e-01 8.05103928e-02 -7.36612797e-01 -3.45607847e-01 5.36406994e-01 1.58018485e-01 -4.97910559e-01 -2.64789224e-01 -7.58503675e-01 7.61790097e-01 -8.10481682e-02 3.71188879e-01 -2.89711446e-01 -7.58942589e-02 5.84095180e-01 1.90482706e-01 2.38936052e-01 -3.46407667e-02 5.30391991e-01 -1.31287828e-01 -1.08804917e+00 1.90583631e-01 6.42976582e-01 7.83118069e-01 5.91850042e-01 7.49857351e-02 -1.98300183e-02 2.96033353e-01 2.92958260e-01 5.51384151e-01 3.22601378e-01 -8.69986653e-01 2.52650261e-01 3.07114929e-01 -4.79838774e-02 -1.60407484e+00 2.36918762e-01 -2.88255900e-01 -8.02979231e-01 2.24303827e-01 2.56746799e-01 6.76294863e-01 -1.03559323e-01 1.39954233e+00 4.41012830e-01 1.59079835e-01 -2.13386212e-02 7.54131198e-01 5.68914115e-01 4.71800417e-01 -1.44155994e-01 -4.74954128e-01 1.53743958e+00 -2.70348966e-01 -8.72264564e-01 3.43728662e-01 6.89134970e-02 -8.96499872e-01 7.44370341e-01 7.91379750e-01 -1.09647477e+00 -4.42612708e-01 -1.05221021e+00 -1.13394847e-02 -2.92609483e-01 -1.03898473e-01 4.15590733e-01 1.03590000e+00 -1.01165819e+00 4.99734342e-01 -4.05588388e-01 -4.12840098e-01 2.77632415e-01 5.87702952e-02 -5.26037216e-01 -2.68372059e-01 -1.07890034e+00 8.82351458e-01 2.12567270e-01 2.29400158e-01 -5.94486296e-01 -3.90332252e-01 -7.04554856e-01 1.21486023e-01 9.55831185e-02 -3.11050087e-01 6.36042953e-01 -3.35352600e-01 -1.07569957e+00 1.00978541e+00 -1.56094376e-02 -5.51471651e-01 6.31192267e-01 9.98169836e-03 -5.49003601e-01 9.69684362e-01 1.11512713e-01 -9.11729410e-02 1.26616085e+00 -1.22609472e+00 -1.79527193e-01 -5.32295942e-01 -9.24692899e-02 -3.46172899e-01 -6.95026994e-01 3.93965781e-01 -4.30907607e-01 -5.82794309e-01 2.83397790e-02 -5.11877954e-01 7.09277540e-02 4.62725550e-01 -2.48545110e-01 1.16175212e-01 1.15591621e+00 -7.93500423e-01 1.34969234e+00 -2.56964779e+00 -1.29229516e-01 2.93923050e-01 3.20644975e-01 4.89512920e-01 1.96421921e-01 7.50356078e-01 -6.63325936e-02 -1.85677290e-01 -4.38694507e-01 -4.08564091e-01 7.82123357e-02 -1.53762549e-02 -7.59397030e-01 1.33676016e+00 9.44488198e-02 5.35745144e-01 -8.52800071e-01 -9.02054727e-01 4.38396901e-01 5.72115064e-01 -2.83039808e-01 1.84996992e-01 5.31348348e-01 1.48302123e-01 -5.14315724e-01 1.01878321e+00 1.16685796e+00 -8.20287541e-02 1.35074064e-01 -4.15939629e-01 -1.57148972e-01 -2.58648783e-01 -1.44273281e+00 1.26635385e+00 -5.25513738e-02 6.35978699e-01 5.15441895e-01 -1.35421121e+00 1.00222278e+00 2.46302664e-01 6.07921660e-01 -7.78764725e-01 1.40108928e-01 3.29455882e-01 -3.10992211e-01 -8.00570428e-01 8.68877232e-01 -6.75026327e-02 5.05576953e-02 6.81407332e-01 -9.74270031e-02 3.34587961e-01 2.31741533e-01 3.29669058e-01 1.05317390e+00 -1.11995935e-02 4.97179419e-01 -1.23053633e-01 7.17073679e-01 -2.47466683e-01 3.24672878e-01 4.70197350e-01 -4.73998457e-01 7.17469454e-01 3.58453274e-01 -1.62307426e-01 -9.26940978e-01 -9.78772223e-01 -5.76494753e-01 2.84932315e-01 6.21279001e-01 -3.26454610e-01 -7.85037816e-01 -3.13755155e-01 2.27437466e-02 3.34891498e-01 -4.14593577e-01 -1.46033004e-01 -6.48510695e-01 -7.42679775e-01 7.78767347e-01 8.90405010e-03 6.25417590e-01 -9.09460604e-01 -8.86328042e-01 -3.80066074e-02 -1.92074031e-01 -1.44302356e+00 -7.18803853e-02 -2.72668630e-01 -7.10523963e-01 -1.60350120e+00 -6.50532246e-01 -2.43591592e-01 4.96536732e-01 1.03143358e+00 6.90201461e-01 4.69773382e-01 -5.43266296e-01 7.93680072e-01 -5.17576814e-01 5.99943995e-02 -3.67489427e-01 -4.94453102e-01 -7.39683281e-04 5.87848663e-01 1.53103203e-01 -9.44293022e-01 -6.80692375e-01 2.69853204e-01 -1.49086022e+00 -5.04896045e-01 6.10692978e-01 6.30480766e-01 3.92984152e-01 1.40934363e-01 3.14335942e-01 -5.61373293e-01 5.43739259e-01 -6.41317725e-01 -2.98086107e-01 4.27869380e-01 -4.98594940e-01 -3.07246417e-01 5.59635818e-01 -1.88130975e-01 -8.79300058e-01 -3.60610753e-01 -6.66721314e-02 -6.63222551e-01 3.97872366e-02 5.27869128e-02 -4.16675746e-01 -4.86221820e-01 4.51998860e-01 8.24578822e-01 2.10559040e-01 -5.48428178e-01 3.44645470e-01 8.27337265e-01 8.74943793e-01 -8.37966979e-01 1.20614254e+00 1.09253287e+00 2.78323263e-01 -1.14252043e+00 -2.18980417e-01 -8.05456877e-01 -5.36710918e-01 -4.31781232e-01 3.71000081e-01 -4.31338042e-01 -9.69996214e-01 4.37651992e-01 -1.51255381e+00 6.44905031e-01 -3.79799157e-01 2.38065064e-01 -5.86864352e-01 1.50764823e+00 -4.93550986e-01 -1.30466068e+00 -4.59050864e-01 -1.30200624e+00 1.14317620e+00 -1.66171402e-01 2.25494325e-01 -6.62950337e-01 -7.99911693e-02 3.33035618e-01 2.07079932e-01 5.18334568e-01 7.11821079e-01 -5.10577679e-01 -7.29795396e-01 -4.22438234e-01 -4.97331172e-01 5.42176485e-01 -2.27202296e-01 -1.16535001e-01 -1.11253130e+00 -3.93308759e-01 6.07644916e-01 2.02400070e-02 6.11682415e-01 -4.34147827e-02 1.16493344e+00 -2.18277618e-01 -3.99409607e-02 8.26286614e-01 1.63222754e+00 6.37805462e-02 1.15940809e+00 5.91845632e-01 2.18822643e-01 8.74784350e-01 5.57849586e-01 7.54811585e-01 4.24932390e-02 8.54645252e-01 6.74657226e-01 3.07220042e-01 -3.17203760e-01 1.70555152e-02 5.24468422e-01 8.55536520e-01 -8.49020202e-03 -2.46658385e-01 -4.03614879e-01 4.70156461e-01 -1.75514853e+00 -1.39919412e+00 -2.33897477e-01 2.40341783e+00 2.45280236e-01 -8.27569589e-02 1.08466603e-01 7.63465703e-01 1.22419727e+00 3.63949805e-01 -3.11663121e-01 -1.89578667e-01 -3.29786569e-01 2.71960020e-01 3.26412261e-01 -5.47110662e-02 -7.74913549e-01 4.37058955e-01 5.97695875e+00 1.47221410e+00 -1.06220305e+00 3.87148768e-01 3.32934111e-01 3.41419071e-01 -1.75260961e-01 1.57790020e-01 -7.08736062e-01 8.36906433e-01 6.98949516e-01 -1.24524608e-01 1.59607321e-01 8.63325775e-01 1.77662805e-01 -1.22827776e-01 -6.56911492e-01 1.41265190e+00 4.24543291e-01 -1.24441850e+00 2.56313622e-01 5.10056436e-01 -1.62450038e-03 -8.75540197e-01 3.14111352e-01 -2.38678053e-01 -6.10224068e-01 -6.81751251e-01 1.06757748e+00 3.85456741e-01 8.67311537e-01 -6.77121639e-01 5.23923934e-01 1.64339840e-01 -1.13144803e+00 4.16076593e-02 -6.02017045e-01 8.61185715e-02 5.36896110e-01 9.18093204e-01 -2.42943510e-01 1.02633905e+00 4.90205139e-01 5.99793613e-01 -3.79017115e-01 1.17908418e+00 -8.21505580e-03 5.20556390e-01 -1.58486664e-01 2.50765055e-01 2.68862564e-02 -4.55663204e-01 7.37748325e-01 1.32048976e+00 6.80572629e-01 1.90730970e-02 -2.66948700e-01 7.96052516e-01 2.17773557e-01 1.49356157e-01 -8.34041595e-01 1.54098094e-01 7.06617773e-01 1.31417990e+00 -8.83506417e-01 -5.65894432e-02 -2.08626062e-01 8.99084568e-01 -1.20824892e-02 2.30796486e-02 -8.89571726e-01 -3.79712224e-01 5.55553019e-01 3.61449957e-01 2.82789946e-01 -4.03107971e-01 -2.68103510e-01 -1.24317610e+00 3.90782952e-01 -8.86756837e-01 4.63951468e-01 -3.53718817e-01 -1.30975747e+00 4.82318699e-01 1.59677252e-01 -1.71422708e+00 -2.96176523e-02 -7.89429426e-01 -6.50135040e-01 4.60819423e-01 -1.84958184e+00 -1.18675089e+00 -2.13795125e-01 9.06496704e-01 1.55199155e-01 -2.00799167e-01 5.33837616e-01 6.87586069e-01 -5.21911025e-01 6.41546488e-01 2.91071028e-01 5.08604869e-02 6.25108898e-01 -5.88889599e-01 9.69954655e-02 1.21617210e+00 1.81326568e-02 9.20889556e-01 7.23626792e-01 -3.86597067e-01 -1.92839253e+00 -5.26166975e-01 6.75905883e-01 -3.02180741e-02 7.83421218e-01 1.62154087e-04 -8.88870895e-01 1.67264901e-02 -7.52481967e-02 9.62232128e-02 7.63210773e-01 -5.96608520e-01 -7.23293245e-01 -1.41194820e-01 -1.54051185e+00 2.30799109e-01 9.66153622e-01 -7.63920784e-01 -6.04786038e-01 2.98082203e-01 2.99567193e-01 1.07000671e-01 -6.39024734e-01 -1.01752900e-01 5.08655250e-01 -1.45333874e+00 1.46994328e+00 -1.03876106e-01 2.01779857e-01 -4.25259739e-01 -5.15454948e-01 -2.53776759e-01 8.84561893e-03 -8.37893784e-01 -4.20196652e-01 1.31911564e+00 -6.94167674e-01 -5.40570796e-01 6.06792808e-01 1.45887360e-01 -2.80213188e-02 -4.00025815e-01 -1.43533981e+00 -1.21639800e+00 -3.97778779e-01 -5.77703893e-01 5.25588751e-01 7.52163887e-01 4.07065377e-02 -3.15824956e-01 -7.30468035e-01 1.57211125e-01 1.29159939e+00 2.47796670e-01 6.60209060e-01 -1.05594611e+00 -4.04836714e-01 -3.37050319e-01 -1.08107686e+00 -8.42055798e-01 6.90286979e-02 -6.34412348e-01 -5.17684221e-01 -9.14364159e-01 3.88590127e-01 -1.26462922e-01 -7.95008317e-02 -2.07053170e-01 1.03870571e-01 4.96421307e-01 5.53713083e-01 8.19303572e-01 -6.21358991e-01 5.14539897e-01 9.12516594e-01 -1.24182239e-01 5.39647818e-01 -1.73873514e-01 -6.85267985e-01 5.12354374e-01 2.66831189e-01 -4.85527307e-01 -2.23230287e-01 -8.38220790e-02 5.84457517e-02 8.45371410e-02 7.09404290e-01 -1.09156072e+00 4.91096407e-01 -1.19709512e-02 -2.11280301e-01 -4.80041027e-01 6.59379542e-01 -8.96575749e-01 2.11185798e-01 6.24610901e-01 1.54637218e-01 2.24610090e-01 -2.51962513e-01 7.03982174e-01 -4.45916325e-01 -6.94808304e-01 8.53031933e-01 -7.40473121e-02 -8.25997710e-01 3.24070811e-01 -4.63085808e-02 -5.72125055e-02 1.09501827e+00 -8.90798330e-01 -4.31266814e-01 -3.09380889e-01 -3.04371893e-01 -5.52693784e-01 5.64524531e-01 -2.24072486e-01 9.82565999e-01 -1.15816259e+00 -6.36736810e-01 -8.15371126e-02 7.62948915e-02 -5.70162714e-01 5.84445298e-01 7.71350980e-01 -6.73021257e-01 2.77138650e-01 -4.11533922e-01 -4.41355288e-01 -1.05448699e+00 9.65341330e-01 -1.72409698e-01 -2.77933270e-01 -6.96035981e-01 1.18659653e-01 1.00777872e-01 3.33898962e-01 1.35596260e-01 4.32291210e-01 -3.36704105e-02 8.79357904e-02 9.06507075e-01 8.59259605e-01 1.53604701e-01 -9.58545208e-01 -3.26209038e-01 7.53612816e-01 2.18971312e-01 -1.01421848e-01 1.30330825e+00 -2.52763838e-01 -3.69170487e-01 -2.24511907e-01 1.31578290e+00 1.68992251e-01 -7.93895185e-01 -1.25543907e-01 6.30597770e-02 -9.95047808e-01 -1.37266055e-01 1.41835093e-01 -1.08179462e+00 1.16048706e+00 5.23803115e-01 4.54833746e-01 1.01735616e+00 -2.46382639e-01 1.00038111e+00 1.83929890e-01 7.96634495e-01 -7.77243376e-01 1.63777322e-01 1.44319644e-03 7.20035374e-01 -8.19979250e-01 3.23782146e-01 -5.78559458e-01 -2.68803477e-01 1.34712315e+00 -6.89628795e-02 -2.52251089e-01 5.17771721e-01 2.20946118e-01 -5.11812806e-01 -2.73766518e-01 -5.33732288e-02 -1.28014624e-01 -1.41083941e-01 1.02069116e+00 -3.20190825e-02 -3.25689256e-01 -5.97014368e-01 4.82002735e-01 2.92586535e-02 -2.23308265e-01 4.38899100e-01 8.76910925e-01 -4.58886564e-01 -1.27002800e+00 -1.02647603e+00 -2.09685400e-01 -5.93412638e-01 1.36300651e-02 -1.31515190e-01 1.02245975e+00 1.00138925e-01 8.99369180e-01 -2.97754884e-01 -2.60968953e-01 1.82519689e-01 -2.99541414e-01 4.39172357e-01 -1.05839998e-01 -4.70172167e-01 -3.33864182e-01 -3.09737682e-01 -5.37606359e-01 -6.02215409e-01 -9.02661443e-01 -6.97141111e-01 -7.90317953e-01 -2.06244975e-01 3.73015374e-01 7.48509407e-01 7.38831162e-01 2.79705644e-01 -1.46906972e-01 8.79990280e-01 -7.49224782e-01 -9.52683806e-01 -3.94495249e-01 -9.74768400e-01 5.62358618e-01 1.33516654e-01 -7.82015979e-01 -5.19963801e-01 1.98697727e-02]	[12.387300491333008, 0.9560402035713196]
689ee461-fba4-4cdf-8e1a-4497d4c18338	cramer-rao-bound-informed-training-of-neural	2109.10535	null	https://arxiv.org/abs/2109.10535v2	https://arxiv.org/pdf/2109.10535v2.pdf	Cramér-Rao bound-informed training of neural networks for quantitative MRI	Neural networks are increasingly used to estimate parameters in quantitative MRI, in particular in magnetic resonance fingerprinting. Their advantages over the gold standard non-linear least square fitting are their superior speed and their immunity to the non-convexity of many fitting problems. We find, however, that in heterogeneous parameter spaces, i.e. in spaces in which the variance of the estimated parameters varies considerably, good performance is hard to achieve and requires arduous tweaking of the loss function, hyper parameters, and the distribution of the training data in parameter space. Here, we address these issues with a theoretically well-founded loss function: the Cram\'er-Rao bound (CRB) provides a theoretical lower bound for the variance of an unbiased estimator and we propose to normalize the squared error with respective CRB. With this normalization, we balance the contributions of hard-to-estimate and not-so-hard-to-estimate parameters and areas in parameter space, and avoid a dominance of the former in the overall training loss. Further, the CRB-based loss function equals one for a maximally-efficient unbiased estimator, which we consider the ideal estimator. Hence, the proposed CRB-based loss function provides an absolute evaluation metric. We compare a network trained with the CRB-based loss with a network trained with the commonly used means squared error loss and demonstrate the advantages of the former in numerical, phantom, and in vivo experiments.	['Jakob Assländer', 'Carlos Fernandez-Granda', 'Cem Gultekin', 'Sebastian Flassbeck', 'Kangning Liu', 'Quentin Duchemin', 'Xiaoxia Zhang']	2021-09-22	null	null	null	null	['magnetic-resonance-fingerprinting']	['medical']	[ 3.26184362e-01 1.68137595e-01 -1.27310961e-01 -5.08043110e-01 -9.78214145e-01 -6.42902106e-02 -2.33644973e-02 2.63833731e-01 -9.98299181e-01 9.90429461e-01 -1.74002007e-01 -2.13028416e-01 -5.62393367e-01 -3.99599820e-01 -8.18353415e-01 -1.08690023e+00 -1.17387488e-01 2.96171010e-01 9.11094993e-02 1.39843047e-01 5.26248571e-03 4.89463031e-01 -9.47084188e-01 -5.87776423e-01 1.04239535e+00 1.32027531e+00 1.12074219e-01 3.35076638e-02 3.51411372e-01 2.62163907e-01 -3.62010509e-01 -2.97521681e-01 2.14622796e-01 -2.63847649e-01 -4.75522488e-01 -4.45532322e-01 2.50931680e-01 -2.89716423e-01 -1.42189547e-01 1.09382737e+00 7.36698151e-01 2.07371816e-01 1.00948548e+00 -8.91354263e-01 5.29049756e-03 4.19243693e-01 -6.30550206e-01 1.02665506e-01 -5.57539910e-02 -2.63400793e-01 6.89357579e-01 -6.43265069e-01 4.00861561e-01 5.99463284e-01 1.18616331e+00 3.35155845e-01 -1.29680240e+00 -5.18399656e-01 -2.25191563e-01 -2.09724292e-01 -1.77663100e+00 -4.48869437e-01 6.71191394e-01 -5.43519616e-01 2.98197269e-01 3.14854056e-01 1.64962038e-01 6.90294564e-01 4.95856106e-01 2.71001369e-01 9.03326511e-01 -4.48900789e-01 1.66799203e-01 4.55281168e-01 -8.30461308e-02 7.26091206e-01 5.21340907e-01 7.32240677e-02 -6.72156364e-02 -2.56436735e-01 1.03806543e+00 -1.71145514e-01 -5.30103028e-01 -8.46892655e-01 -9.69202101e-01 8.72677267e-01 5.12081504e-01 4.60008562e-01 -4.61004555e-01 3.97963226e-01 3.75036210e-01 -1.76189672e-02 3.55279118e-01 4.04793322e-01 -1.98504895e-01 6.77747279e-02 -1.09980249e+00 1.30458012e-01 5.43916881e-01 4.59925920e-01 3.62454116e-01 2.05735676e-02 -3.26322354e-02 1.03344762e+00 3.43807012e-01 3.94078374e-01 4.56496418e-01 -8.51092577e-01 4.82664227e-01 2.47086976e-02 2.22057402e-01 -1.19041443e+00 -7.01232791e-01 -8.99788439e-01 -9.81724679e-01 8.43628775e-03 7.29194105e-01 -4.19090569e-01 -3.84727746e-01 1.97629654e+00 4.05988038e-01 -2.10165784e-01 -3.84997457e-01 1.02841353e+00 3.66829783e-01 1.28694311e-01 1.28931731e-01 -4.65040177e-01 1.09179103e+00 -4.12011236e-01 -6.12107873e-01 -2.33300596e-01 3.87973279e-01 -4.67321813e-01 9.01052773e-01 2.68652231e-01 -1.13276172e+00 -2.12476835e-01 -1.26661015e+00 3.46418470e-01 2.69741882e-02 9.90306288e-02 4.22309548e-01 7.48492777e-01 -6.48398936e-01 8.34393620e-01 -8.94446611e-01 1.05710387e-01 3.32549274e-01 5.05659223e-01 -4.43573415e-01 3.16356659e-01 -9.99944627e-01 1.07084417e+00 2.28387937e-01 4.31964427e-01 -4.14061069e-01 -9.94968295e-01 -6.41901314e-01 1.62322059e-01 4.02298510e-01 -5.48880339e-01 8.93862903e-01 -6.58727884e-01 -1.40456712e+00 5.42108297e-01 2.34949857e-01 -3.93140852e-01 8.76306295e-01 -8.46656188e-02 -1.31579891e-01 -2.85050441e-02 8.36164225e-03 2.01511070e-01 7.09490597e-01 -1.11715603e+00 1.08692110e-01 -4.41527158e-01 -3.21554035e-01 3.00974641e-02 -1.73375189e-01 -8.47324803e-02 -1.82196781e-01 -6.28029048e-01 4.92450625e-01 -8.40557158e-01 -3.82878810e-01 2.60730147e-01 -3.09366167e-01 1.80382326e-01 -2.00765152e-02 -7.06374645e-01 1.22102880e+00 -2.28028083e+00 -3.26469511e-01 6.12564087e-01 2.32356012e-01 1.47781670e-02 1.47533372e-01 4.78494447e-03 -1.02021925e-01 -1.76934272e-01 -5.05457222e-01 -1.88921078e-03 3.31039564e-03 -7.11847842e-02 2.29713887e-01 9.74189341e-01 -9.21799615e-02 5.33529878e-01 -8.32405031e-01 -5.50989032e-01 3.62964459e-02 8.01396549e-01 -3.41094404e-01 -9.98374298e-02 3.84785473e-01 4.41409826e-01 -4.69439417e-01 1.12142034e-01 5.73567510e-01 -3.02015752e-01 1.35328382e-01 -4.35420543e-01 8.01384896e-02 -9.02685821e-02 -1.31918347e+00 1.37034810e+00 -6.86339915e-01 4.94265854e-01 1.79039061e-01 -1.24927866e+00 9.22513783e-01 3.35809976e-01 7.44384825e-01 -6.16760790e-01 3.67742658e-01 6.90004408e-01 -1.57204382e-02 -3.54453713e-01 7.64171034e-02 -5.39943755e-01 3.40964645e-02 1.79383084e-01 -1.07315853e-02 1.41068876e-01 -5.31479232e-02 -3.27967286e-01 7.19580710e-01 -3.29890996e-02 5.14663815e-01 -5.99177957e-01 5.85583568e-01 -5.89563966e-01 5.85493326e-01 8.00593615e-01 -2.66498208e-01 7.61716366e-01 5.36259234e-01 -1.22693509e-01 -9.04509127e-01 -8.57896328e-01 -8.62262428e-01 6.98957682e-01 3.66479084e-02 3.15740943e-01 -7.49766707e-01 -6.03719175e-01 1.23413660e-01 5.93327522e-01 -4.72393394e-01 -1.04524106e-01 -6.14496350e-01 -1.06174111e+00 5.34635127e-01 3.67921978e-01 1.69062614e-01 -4.60084945e-01 -6.55085206e-01 3.14692974e-01 -1.57712385e-01 -1.01719153e+00 -5.67053735e-01 3.77699465e-01 -8.79224718e-01 -9.40254152e-01 -1.06806552e+00 -4.06899095e-01 8.71398091e-01 -3.23499858e-01 8.70690584e-01 -1.41868040e-01 -1.33960187e-01 1.66619271e-01 1.22878455e-01 -2.81616151e-01 -2.54667103e-01 -1.88642424e-02 2.68666968e-02 -6.18432164e-02 -1.77562639e-01 -5.62294781e-01 -8.85318220e-01 5.20430744e-01 -6.50865376e-01 -4.84563977e-01 3.88937682e-01 9.68818188e-01 8.01238358e-01 -1.87187530e-02 7.82051623e-01 -8.41126263e-01 5.23180366e-01 -3.79563093e-01 -6.69330001e-01 2.92441994e-01 -9.02684689e-01 2.03069225e-01 5.63022494e-01 -4.76485312e-01 -8.01169336e-01 -1.77482422e-02 -1.71055555e-01 -1.49341002e-02 2.42485046e-01 5.64339340e-01 1.12708546e-01 -5.75388074e-01 7.73437202e-01 -3.43169346e-02 1.59330308e-01 -3.90034080e-01 2.75318176e-02 4.19931084e-01 4.88898695e-01 -6.07850730e-01 4.25632387e-01 3.12563390e-01 4.77827877e-01 -7.77587414e-01 -6.36824906e-01 -3.87515455e-01 -2.55577415e-01 -2.35581487e-01 5.16102076e-01 -3.94667268e-01 -8.48336875e-01 1.60486579e-01 -8.82578015e-01 -7.85266086e-02 -4.03491974e-01 8.17844331e-01 -7.32218087e-01 6.49669230e-01 -3.07338357e-01 -9.55178797e-01 -5.42231083e-01 -1.29387772e+00 6.31792426e-01 5.48646487e-02 -2.22022250e-01 -1.12764299e+00 -1.20570827e-02 5.99110872e-02 7.04763710e-01 5.53290725e-01 9.07854617e-01 -6.51566505e-01 2.66194232e-02 -5.53866148e-01 -3.77673000e-01 5.77625513e-01 -1.13085426e-01 -3.63115251e-01 -8.08174133e-01 -3.10529321e-01 3.57911170e-01 4.46347101e-03 6.36492491e-01 8.62946808e-01 1.30068314e+00 -2.82212824e-01 -1.79594487e-01 7.23812938e-01 1.72457600e+00 1.24871522e-01 4.28230822e-01 1.91174388e-01 3.68724972e-01 5.28602540e-01 2.95183331e-01 4.25569475e-01 1.09751604e-01 8.75956059e-01 1.62357241e-01 -1.67989850e-01 1.42052844e-01 2.61225458e-02 -4.29770797e-02 7.96302676e-01 -1.51285231e-01 1.14895571e-02 -7.98357248e-01 2.50205010e-01 -1.62380874e+00 -4.37274456e-01 1.14118829e-01 2.69481277e+00 9.54571605e-01 2.37811267e-01 8.96834508e-02 1.49409056e-01 5.97730458e-01 1.91200618e-02 -4.16906208e-01 -1.85544431e-01 5.88539615e-02 1.23532735e-01 8.98620069e-01 6.02003157e-01 -1.08880520e+00 4.76278849e-02 6.31067085e+00 1.22372973e+00 -1.27360344e+00 2.96433985e-01 6.73643231e-01 1.89368371e-02 -1.04393683e-01 -2.39573732e-01 -4.69350755e-01 5.83198547e-01 8.86506081e-01 -9.72446725e-02 3.28564733e-01 7.57848561e-01 2.75826097e-01 -1.50446311e-01 -8.64896655e-01 8.54010105e-01 -1.76310211e-01 -9.50075030e-01 -5.44853210e-01 2.59447321e-02 4.18272674e-01 -2.70031188e-02 1.28299845e-02 -3.26708774e-03 -4.60590482e-01 -1.03386378e+00 6.54569149e-01 6.66112661e-01 9.05479848e-01 -5.62860906e-01 1.03838813e+00 2.57568508e-01 -7.85454750e-01 1.29605845e-01 -3.44145000e-01 4.37247902e-01 2.04699695e-01 9.24460471e-01 -6.72222733e-01 3.62899393e-01 2.37561181e-01 -3.91443111e-02 -8.45355168e-02 1.39269888e+00 8.34408775e-02 2.73866087e-01 -6.37427628e-01 -1.57972768e-01 1.39685094e-01 -4.98785704e-01 4.88573939e-01 1.16036224e+00 3.76093119e-01 -1.65104613e-01 -1.93749472e-01 9.01946366e-01 1.02909848e-01 4.72633243e-01 -8.60709473e-02 2.82318145e-01 4.27350968e-01 1.12044668e+00 -6.58912361e-01 8.43109637e-02 -2.26338699e-01 4.58125114e-01 1.41139463e-01 4.23355222e-01 -8.63386691e-01 -5.93460321e-01 1.86947450e-01 5.06598711e-01 1.86187357e-01 4.98290844e-02 -3.46517652e-01 -7.63230264e-01 3.88968706e-01 -6.61664426e-01 2.07308263e-01 -3.76290083e-01 -1.20527947e+00 6.34070337e-01 2.76851773e-01 -1.07134354e+00 -3.18295032e-01 -6.32349610e-01 -2.21178353e-01 8.93658698e-01 -1.34212172e+00 -5.64668298e-01 -5.98451272e-02 1.35934621e-01 -2.18636930e-01 1.28591195e-01 5.43813884e-01 8.04349840e-01 -5.67602456e-01 9.75279093e-01 4.29572254e-01 -4.87721637e-02 5.03450036e-01 -1.11819196e+00 -3.24194789e-01 4.62338805e-01 -3.79792213e-01 8.16151142e-01 9.80511248e-01 -3.02364498e-01 -8.94658029e-01 -5.00781894e-01 7.40387738e-01 7.95886964e-02 6.21394932e-01 -6.43769354e-02 -8.56024981e-01 2.93493330e-01 -6.20411813e-01 2.42126808e-01 4.54726070e-01 1.03272023e-02 -1.01626702e-01 -5.16775429e-01 -1.39984131e+00 3.28880370e-01 4.68335956e-01 -3.73394847e-01 -4.07193005e-02 3.63956422e-01 2.70206124e-01 -4.18760538e-01 -1.22285295e+00 7.01031625e-01 1.04401577e+00 -8.59910131e-01 1.09798574e+00 -2.78755575e-01 1.12815991e-01 3.90172787e-02 -6.66522682e-02 -9.76282895e-01 -1.12312853e-01 -4.17553753e-01 2.54818380e-01 8.80070329e-01 5.38734257e-01 -9.13953960e-01 8.10109854e-01 7.07334876e-01 1.53782219e-01 -1.24016666e+00 -1.33839929e+00 -8.69157135e-01 2.59960175e-01 -3.39471459e-01 3.36768925e-01 7.43601143e-01 -6.32135868e-02 -8.52435529e-02 -3.42180967e-01 -1.78540468e-01 9.09636617e-01 -4.92316484e-01 2.97668815e-01 -1.13512278e+00 -3.52179676e-01 -5.10418892e-01 -3.50264668e-01 -8.97270918e-01 -1.16337828e-01 -7.00980484e-01 3.55868846e-01 -1.08817983e+00 1.27671972e-01 -9.83457565e-01 -5.15007019e-01 3.31448093e-02 -8.15646425e-02 5.86836450e-02 -4.62677106e-02 9.36625749e-02 -2.54228622e-01 4.10709232e-01 1.19080842e+00 1.46892428e-01 -1.56468049e-01 2.41490766e-01 -4.03297782e-01 8.73662770e-01 4.59192812e-01 -8.00668538e-01 -2.88360864e-01 -1.65730327e-01 3.56734157e-01 3.37892920e-01 3.77769828e-01 -9.19630766e-01 1.67934503e-02 6.51842505e-02 2.73811877e-01 -3.36105973e-02 3.13795388e-01 -1.10075414e+00 3.55920523e-01 4.20227408e-01 -4.11415637e-01 -4.37930614e-01 -1.07947648e-01 3.35972011e-01 -4.46928889e-02 -7.50813484e-01 1.07914817e+00 1.44519955e-01 3.69565822e-02 1.36339024e-01 6.74169734e-02 7.06999451e-02 5.30551374e-01 -3.86634856e-01 -3.90099399e-02 -4.67087954e-01 -5.71763694e-01 -2.63917059e-01 5.07424735e-02 -2.08735719e-01 3.49237531e-01 -1.14675057e+00 -5.97545087e-01 1.23547666e-01 -1.47637546e-01 -2.74888277e-01 3.84839684e-01 1.39950526e+00 -6.47150338e-01 4.38928545e-01 6.59720525e-02 -4.79098022e-01 -7.03482509e-01 1.81663856e-01 7.33756840e-01 -3.79017621e-01 -3.34969759e-01 7.17006624e-01 1.73196509e-01 -4.19278353e-01 4.10236001e-01 -2.96711653e-01 -9.35847685e-03 -7.68362135e-02 2.24137247e-01 6.69381857e-01 3.25620055e-01 -5.41968882e-01 -4.24910724e-01 5.27768135e-01 2.79273599e-01 3.21295448e-02 1.27135277e+00 -9.10827145e-02 -3.20084617e-02 3.98509324e-01 1.35493457e+00 3.48986275e-02 -1.07216024e+00 -2.01395109e-01 -1.63273718e-02 -1.65242225e-01 4.02833104e-01 -5.92463732e-01 -1.02670419e+00 6.09262764e-01 9.27845657e-01 1.61060989e-01 9.24179435e-01 -3.83863360e-01 6.32228434e-01 1.90419480e-01 2.32755765e-01 -1.05030584e+00 -4.28776413e-01 2.57684086e-02 8.36829960e-01 -1.14149666e+00 4.08629358e-01 -4.08359498e-01 -3.81535023e-01 9.65713680e-01 1.60150751e-01 -1.19967118e-01 8.19006741e-01 1.76729396e-01 6.25891238e-02 -5.67994965e-03 9.05520748e-03 3.14786047e-01 6.32972598e-01 3.22856635e-01 5.00459969e-01 1.76146314e-01 -8.15967202e-01 7.38789737e-01 -7.01850057e-02 1.09067336e-01 9.44891870e-02 4.59227979e-01 -2.26082608e-01 -7.45948136e-01 -3.22394729e-01 4.74014640e-01 -6.62718296e-01 4.59938534e-02 4.42585170e-01 9.87603366e-01 -4.07011881e-02 4.74949479e-01 -8.12881887e-02 8.87364149e-02 3.61677647e-01 -1.51894033e-01 5.97874463e-01 2.73232739e-02 -3.87995034e-01 1.86815813e-01 1.44264409e-02 -5.05154431e-01 -2.93189794e-01 -5.84231079e-01 -1.06246531e+00 -8.15893859e-02 -6.98076725e-01 1.36038870e-01 1.08877158e+00 9.84050632e-01 -7.62938261e-02 3.03710103e-01 5.07666707e-01 -5.21452010e-01 -1.08145273e+00 -7.93666959e-01 -7.41205692e-01 2.35337660e-01 3.46176833e-01 -8.27270031e-01 -5.79549372e-01 -5.20544648e-01]	[7.102792263031006, 3.9549379348754883]
069336bf-0937-4c92-9a09-a36d9726c5f5	fine-grained-image-captioning-with-clip	2205.13115	null	https://arxiv.org/abs/2205.13115v2	https://arxiv.org/pdf/2205.13115v2.pdf	Fine-grained Image Captioning with CLIP Reward	Modern image captioning models are usually trained with text similarity objectives. However, since reference captions in public datasets often describe the most salient common objects, models trained with text similarity objectives tend to ignore specific and detailed aspects of an image that distinguish it from others. Toward more descriptive and distinctive caption generation, we propose using CLIP, a multimodal encoder trained on huge image-text pairs from web, to calculate multimodal similarity and use it as a reward function. We also propose a simple finetuning strategy of the CLIP text encoder to improve grammar that does not require extra text annotation. This completely eliminates the need for reference captions during the reward computation. To comprehensively evaluate descriptive captions, we introduce FineCapEval, a new dataset for caption evaluation with fine-grained criteria: overall, background, object, relations. In our experiments on text-to-image retrieval and FineCapEval, the proposed CLIP-guided model generates more distinctive captions than the CIDEr-optimized model. We also show that our unsupervised grammar finetuning of the CLIP text encoder alleviates the degeneration problem of the naive CLIP reward. Lastly, we show human analysis where the annotators strongly prefer the CLIP reward to the CIDEr and MLE objectives according to various criteria. Code and Data: https://github.com/j-min/CLIP-Caption-Reward	['Mohit Bansal', 'Trung Bui', 'Franck Dernoncourt', 'Ajinkya Kale', 'Seunghyun Yoon', 'Jaemin Cho']	2022-05-26	null	https://aclanthology.org/2022.findings-naacl.39	https://aclanthology.org/2022.findings-naacl.39.pdf	findings-naacl-2022-7	['text-annotation']	['natural-language-processing']	[ 3.05347890e-01 3.83578688e-02 -8.66859108e-02 -6.53537810e-01 -1.40555441e+00 -8.58427227e-01 6.87349677e-01 1.58468127e-01 -3.93895715e-01 5.62541425e-01 3.50073218e-01 1.14467934e-01 2.00043812e-01 -2.39127576e-01 -1.03535616e+00 -5.61039388e-01 4.48287696e-01 6.25865519e-01 8.76631774e-03 -1.05530873e-01 1.92039505e-01 3.62467952e-02 -1.61582148e+00 5.39616287e-01 9.66757834e-01 1.21069264e+00 5.04899621e-01 7.09521711e-01 -4.96830344e-02 5.24178386e-01 -4.52577293e-01 -1.07981479e+00 1.33184060e-01 -5.93128562e-01 -7.11873829e-01 8.89939666e-02 1.12788546e+00 -2.89146274e-01 -2.65342861e-01 1.15433288e+00 6.72809124e-01 9.92780775e-02 7.25012422e-01 -1.46902525e+00 -1.17053974e+00 7.54564047e-01 -6.20701313e-01 -3.51145910e-03 5.16555727e-01 2.51542836e-01 1.31464183e+00 -9.62686598e-01 7.55440950e-01 1.34301364e+00 2.57479489e-01 7.60440707e-01 -1.26728952e+00 -4.34645712e-01 4.26549874e-02 2.16780707e-01 -1.35047936e+00 -4.65875089e-01 6.99275911e-01 -2.51927376e-01 3.01621169e-01 5.28333902e-01 5.99150686e-03 1.56363034e+00 -2.08627224e-01 1.14241469e+00 7.44178236e-01 -2.07010701e-01 5.40360203e-03 3.47257107e-01 -1.58263743e-01 5.51166236e-01 1.67667738e-03 -1.34535700e-01 -3.46131295e-01 -8.93404856e-02 4.61437225e-01 -1.51164934e-01 -4.18290645e-01 -5.13275981e-01 -1.35741782e+00 8.29068840e-01 3.86629701e-01 1.73520789e-01 -2.51557678e-01 4.96568918e-01 5.57671130e-01 -3.86455432e-02 7.66987056e-02 5.27439117e-01 -1.17044322e-01 -1.53418168e-01 -9.59249198e-01 1.53563663e-01 4.30310071e-01 1.29627657e+00 7.27276146e-01 -3.05090874e-01 -8.98274302e-01 1.13283968e+00 1.48816302e-01 9.35626268e-01 4.26330298e-01 -1.14469004e+00 5.97731054e-01 3.20583582e-01 4.34968732e-02 -8.83105278e-01 4.21316177e-02 -2.97459841e-01 -7.04951704e-01 -2.60525972e-01 2.54030287e-01 6.51029218e-03 -9.71615613e-01 1.96212840e+00 -9.43811834e-02 -2.76488513e-01 6.94998950e-02 1.35688269e+00 1.04138267e+00 6.58544540e-01 4.10900742e-01 1.16151221e-01 1.45722604e+00 -1.17658150e+00 -6.73469067e-01 -1.84668571e-01 6.31590366e-01 -9.52318907e-01 1.49054956e+00 -6.11059032e-02 -1.21438825e+00 -5.15086710e-01 -6.84496164e-01 -2.28140682e-01 -3.72078419e-01 4.65098768e-01 4.20538992e-01 3.29521209e-01 -9.48734164e-01 2.58297443e-01 -1.65653631e-01 -4.54328299e-01 4.31651324e-01 2.20786542e-01 -4.22526896e-01 -1.97189867e-01 -1.20196438e+00 7.83443689e-01 4.28811163e-01 -1.83382347e-01 -1.01070321e+00 -4.17891502e-01 -9.89529550e-01 1.87797055e-01 3.50623250e-01 -8.01053643e-01 1.37309813e+00 -1.50652456e+00 -1.07131112e+00 1.06024885e+00 -1.03307925e-01 -3.38297486e-01 3.10130894e-01 -1.09162942e-01 -2.59376109e-01 5.69656074e-01 2.96951234e-01 1.44842982e+00 1.02984655e+00 -1.55822349e+00 -3.67069364e-01 6.37162402e-02 1.17027730e-01 3.50026637e-01 -3.45293969e-01 4.55341898e-02 -8.90796840e-01 -7.80224383e-01 -4.87213790e-01 -9.78041768e-01 9.99499038e-02 2.54726615e-02 -4.47992176e-01 -2.22320616e-01 5.98595083e-01 -4.33322728e-01 1.07976699e+00 -2.22440028e+00 3.39040719e-02 9.71938856e-03 -4.92438935e-02 1.10115334e-01 -7.44398355e-01 3.89307737e-01 -1.05637550e-01 2.56805003e-01 -3.78632665e-01 -4.61990207e-01 3.29283386e-01 2.31355250e-01 -3.32290947e-01 7.14973174e-03 3.75825077e-01 1.24313343e+00 -1.02894354e+00 -1.07218671e+00 7.13927746e-02 4.12102997e-01 -4.13795590e-01 3.46396565e-01 -5.39740860e-01 7.05095604e-02 -4.32534724e-01 8.05152535e-01 6.23572350e-01 -5.00075042e-01 -1.85786456e-01 -6.11976504e-01 2.96145260e-01 -2.86546707e-01 -6.70734584e-01 1.93700850e+00 -3.88533533e-01 7.30876327e-01 -3.28253433e-02 -7.45981276e-01 8.57508063e-01 2.76243597e-01 3.87456179e-01 -9.66172934e-01 2.49842286e-01 2.05687076e-01 -3.29693735e-01 -7.33365536e-01 7.66006052e-01 1.76913336e-01 -4.03497189e-01 3.13098818e-01 2.47380286e-01 -1.13817461e-01 5.91900051e-01 5.73082149e-01 7.69988716e-01 3.24016213e-01 -1.67661369e-01 7.17125237e-02 5.33462763e-01 -4.48283218e-02 1.14001215e-01 8.26506615e-01 -2.46906027e-01 1.16004372e+00 4.94100302e-01 -4.00875807e-02 -1.24778962e+00 -1.17102110e+00 -9.58534852e-02 1.47166216e+00 3.86242598e-01 -3.22115898e-01 -8.63666713e-01 -8.60319197e-01 -1.70874745e-01 7.05506980e-01 -7.53158867e-01 -6.84781447e-02 -3.45087260e-01 -5.51115453e-01 6.66059554e-01 3.50519955e-01 3.60895127e-01 -1.10105264e+00 -2.75213659e-01 -1.76628470e-01 -7.70753562e-01 -1.33765292e+00 -1.25442898e+00 -1.00584991e-01 -3.75401080e-01 -8.66326213e-01 -1.18080735e+00 -1.02515459e+00 8.35657299e-01 5.21266833e-02 1.35217762e+00 -1.03992045e-01 -2.40315840e-01 7.88247883e-01 -6.86039925e-01 -1.04087122e-01 -4.31273580e-01 5.31499051e-02 -3.92611861e-01 2.19981328e-01 3.11812490e-01 -1.68453559e-01 -7.03550994e-01 4.06557053e-01 -1.14865470e+00 1.90319881e-01 8.72516632e-01 8.92040849e-01 6.60620332e-01 -7.38480866e-01 5.78528821e-01 -4.24965352e-01 7.32072055e-01 -3.74068797e-01 -3.51984203e-01 6.55625224e-01 -3.13693196e-01 3.66202325e-01 5.20663381e-01 -6.96921229e-01 -8.59532893e-01 4.17181998e-01 1.02612592e-01 -8.58063281e-01 -9.98531356e-02 1.42590389e-01 -6.78251311e-02 4.18245234e-02 6.44840658e-01 4.42709446e-01 -1.43881962e-01 -2.44659528e-01 6.67354703e-01 6.76287532e-01 8.24256718e-01 -8.82678270e-01 6.84165776e-01 1.75563455e-01 -3.26525450e-01 -3.87594968e-01 -9.61149275e-01 -4.01927084e-01 -2.27939516e-01 -3.00691873e-01 1.03296936e+00 -9.26244080e-01 -9.15151536e-01 -1.92467511e-01 -1.33561695e+00 -8.42105374e-02 -1.67689681e-01 3.68947566e-01 -9.37133133e-01 4.84893918e-01 -3.54213387e-01 -5.67258596e-01 -6.74146473e-01 -1.19824409e+00 1.62765837e+00 2.88102210e-01 -1.39444143e-01 -6.95912600e-01 2.61520334e-02 4.64804292e-01 2.90720612e-01 3.48129243e-01 6.15019381e-01 -6.42488658e-01 -5.97835124e-01 -1.33749656e-02 -6.02817595e-01 3.27383548e-01 -1.66377068e-01 5.63777983e-02 -9.17777538e-01 -1.04748778e-01 -5.29582322e-01 -7.81876087e-01 8.79981279e-01 2.17773899e-01 1.33265400e+00 -4.71536309e-01 -6.30004928e-02 4.79244530e-01 1.45268214e+00 1.28132850e-02 6.80268288e-01 2.05782264e-01 7.16445863e-01 5.93200445e-01 7.80651093e-01 3.36100489e-01 3.26681644e-01 8.35045755e-01 6.29540026e-01 -1.89360544e-01 -2.44365409e-01 -3.79039317e-01 5.35302877e-01 4.66578573e-01 2.10692450e-01 -4.96082097e-01 -6.79548204e-01 7.74281442e-01 -1.97289932e+00 -1.18408573e+00 1.80676594e-01 2.03112626e+00 1.04798806e+00 -2.00723872e-01 -5.87868970e-03 -5.11835754e-01 1.08514118e+00 -7.68335024e-03 -4.19833302e-01 -5.09139180e-01 -4.38732594e-01 -1.71779707e-01 4.54348385e-01 4.24271286e-01 -1.11685324e+00 1.05841970e+00 5.61309862e+00 1.22001088e+00 -1.02105844e+00 1.66802257e-02 6.93629980e-01 -2.67429709e-01 -5.98007739e-01 -2.23552093e-01 -5.79779983e-01 6.97918296e-01 7.07494080e-01 -2.98243284e-01 4.21274960e-01 9.15134609e-01 1.57916784e-01 -7.38592520e-02 -1.42323315e+00 1.38448060e+00 4.95250195e-01 -1.19589174e+00 5.74167252e-01 -3.19468766e-01 7.52006352e-01 -6.34101927e-02 3.39471728e-01 2.58542746e-01 5.30828945e-02 -8.96565259e-01 8.67540538e-01 6.46839499e-01 9.74910259e-01 -5.08300781e-01 7.54109800e-01 -2.52802789e-01 -8.97656322e-01 1.42476112e-01 -3.42832088e-01 7.79814243e-01 2.75523961e-01 3.27056736e-01 -8.25382888e-01 3.79434794e-01 5.99831104e-01 5.34169436e-01 -9.93348241e-01 1.08108938e+00 3.70669141e-02 1.42587095e-01 -7.45092034e-02 -2.27975607e-01 4.93207306e-01 -4.14762199e-02 5.88934183e-01 1.44920981e+00 3.79508883e-01 -2.18271852e-01 2.49351934e-02 1.06701493e+00 -4.52443212e-01 3.04296046e-01 -4.69224036e-01 -3.03254962e-01 2.58403778e-01 1.38929880e+00 -3.83819431e-01 -4.33582455e-01 -2.96526134e-01 1.21243656e+00 1.28699884e-01 4.91574675e-01 -1.15214205e+00 -5.31786621e-01 4.13540155e-01 -1.01804040e-01 4.34906751e-01 2.23585680e-01 3.16714728e-03 -1.07427835e+00 2.07906142e-01 -8.91665280e-01 3.46826494e-01 -1.50918090e+00 -1.41738379e+00 9.09306824e-01 5.73191978e-02 -1.31828129e+00 -3.71624351e-01 -4.88477498e-01 -3.40254068e-01 6.35537326e-01 -1.34916019e+00 -1.29831064e+00 -4.57512617e-01 8.35866511e-01 6.12289667e-01 6.12292811e-02 5.68919897e-01 4.27969098e-01 -3.71546835e-01 1.01440060e+00 3.89237106e-02 2.25631014e-01 1.26367724e+00 -1.14272428e+00 -2.28211228e-02 5.28005123e-01 1.15920916e-01 3.31055313e-01 9.36409116e-01 -4.17902112e-01 -1.10843599e+00 -1.20737815e+00 7.98975289e-01 -5.58457255e-01 5.31843662e-01 -4.69374418e-01 -7.43315339e-01 3.46631229e-01 6.18357062e-01 -1.41193345e-01 5.28455079e-01 -3.04315776e-01 -5.45739889e-01 -2.15906337e-01 -9.70696211e-01 7.70471156e-01 7.06772625e-01 -5.32158732e-01 -5.55688560e-01 4.78494048e-01 9.86691296e-01 -1.83548063e-01 -8.30586791e-01 2.45233119e-01 6.09393835e-01 -6.94275975e-01 7.80053139e-01 -3.97027165e-01 7.71493614e-01 -3.17779452e-01 -3.83939624e-01 -1.06701899e+00 -1.78289786e-01 -5.30445755e-01 2.02388421e-01 1.39612532e+00 6.79295540e-01 -1.99306607e-02 6.85705900e-01 5.74630022e-01 -2.64612526e-01 -5.27120054e-01 -7.32754052e-01 -8.14454556e-01 -1.77375585e-01 -1.72462702e-01 4.57055181e-01 8.70965123e-01 3.27394791e-02 4.86798286e-01 -5.39922833e-01 -9.71676558e-02 7.27818668e-01 2.97807992e-01 6.43156648e-01 -7.55154371e-01 -1.53086483e-01 -5.85828185e-01 -1.72912017e-01 -9.90407765e-01 1.43717825e-01 -9.73040819e-01 2.48943016e-01 -1.54107058e+00 6.59677804e-01 -1.62616834e-01 -1.72205657e-01 5.74579775e-01 -1.88756317e-01 5.19659162e-01 4.67358857e-01 2.45395303e-01 -1.29748023e+00 7.70456433e-01 1.58023381e+00 -4.68087226e-01 1.45862296e-01 -5.22263408e-01 -7.21225381e-01 2.82972962e-01 7.17071533e-01 -4.63501126e-01 -3.92441094e-01 -6.87376082e-01 2.86822647e-01 -4.26243544e-02 5.76979458e-01 -7.70669818e-01 1.19940415e-01 -2.08352238e-01 3.12894583e-01 -5.69666564e-01 4.50937450e-01 -8.40704918e-01 2.91500404e-03 1.47574246e-02 -8.52325857e-01 5.04884198e-02 1.82931885e-01 3.20481777e-01 -4.02704984e-01 -4.17461842e-01 7.37292528e-01 4.09217738e-02 -5.43043375e-01 4.57840979e-01 -6.77660108e-02 3.31696689e-01 8.65669191e-01 -7.46918395e-02 -5.45814574e-01 -7.17197359e-01 -7.32945085e-01 5.49144626e-01 5.95720053e-01 7.75708199e-01 6.74767137e-01 -1.70691717e+00 -9.38273609e-01 -3.12889308e-01 7.58294642e-01 -4.75922823e-01 3.36854249e-01 7.42224514e-01 -2.51234949e-01 4.70454663e-01 -3.22662950e-01 -7.27341294e-01 -1.22112274e+00 8.69439423e-01 2.00550884e-01 5.39586581e-02 -1.95026815e-01 5.69138467e-01 6.16109729e-01 -7.32168974e-03 4.15906936e-01 -1.62341952e-01 -8.39227289e-02 2.49160454e-02 4.59514827e-01 -3.61580253e-02 -2.53170371e-01 -8.41916203e-01 -2.13823602e-01 7.06743717e-01 -1.61613196e-01 -1.99104920e-01 8.51141453e-01 -4.46280122e-01 1.25666946e-01 3.67135480e-02 1.43930626e+00 -2.49851435e-01 -1.23858273e+00 -9.68958288e-02 -1.90095246e-01 -3.25756609e-01 -2.05934957e-01 -1.05381250e+00 -9.73182142e-01 8.56343150e-01 7.64187336e-01 3.88398767e-02 1.30973589e+00 3.33986253e-01 8.20563018e-01 5.07766366e-01 -5.99141866e-02 -1.20124066e+00 3.93473387e-01 3.41560900e-01 1.29488206e+00 -1.59254789e+00 -5.05822837e-01 -1.92183182e-01 -1.14358258e+00 9.14466858e-01 7.52340972e-01 2.31357098e-01 -8.17814544e-02 -2.07017854e-01 2.21299175e-02 -5.64771006e-03 -8.23210001e-01 -5.46489239e-01 4.70182925e-01 6.61209941e-01 3.49864632e-01 -1.55347005e-01 -2.69474477e-01 5.36769152e-01 -3.05020157e-02 -2.05612138e-01 4.69944179e-01 3.73442471e-01 -3.74009758e-01 -1.02790022e+00 -4.48981822e-01 2.58600324e-01 -4.48562831e-01 -4.69728827e-01 -5.08038223e-01 4.31884557e-01 1.09739015e-02 7.98462987e-01 1.96972936e-01 -2.74772555e-01 2.26413548e-01 -2.19944753e-02 4.55406040e-01 -3.00712675e-01 -4.52516764e-01 3.74647230e-02 1.12452447e-01 -5.58531880e-01 -3.93186450e-01 -4.38717186e-01 -1.09027946e+00 -5.83276860e-02 -1.17719375e-01 3.93480808e-01 7.72050321e-01 4.61639285e-01 4.70788956e-01 2.39651814e-01 6.71419322e-01 -7.66639233e-01 -2.99457759e-01 -8.72823000e-01 -2.53763705e-01 9.32227015e-01 1.89992025e-01 -4.57295239e-01 -3.60525548e-01 5.24679959e-01]	[11.031025886535645, 1.051996111869812]
a486fdae-ea70-46a5-ad92-e7131098560a	contrastive-regression-for-domain-adaptation	null	null	http://openaccess.thecvf.com//content/CVPR2022/html/Wang_Contrastive_Regression_for_Domain_Adaptation_on_Gaze_Estimation_CVPR_2022_paper.html	http://openaccess.thecvf.com//content/CVPR2022/papers/Wang_Contrastive_Regression_for_Domain_Adaptation_on_Gaze_Estimation_CVPR_2022_paper.pdf	Contrastive Regression for Domain Adaptation on Gaze Estimation	Appearance-based Gaze Estimation leverages deep neural networks to regress the gaze direction from monocular images and achieve impressive performance. However, its success depends on expensive and cumbersome annotation capture. When lacking precise annotation, the large domain gap hinders the performance of trained models on new domains. In this paper, we propose a novel gaze adaptation approach, namely Contrastive Regression Gaze Adaptation (CRGA), for generalizing gaze estimation on the target domain in an unsupervised manner. CRGA leverages the Contrastive Domain Generalization (CDG) module to learn the stable representation from the source domain and leverages the Contrastive Self-training Adaptation (CSA) module to learn from the pseudo labels on the target domain. The core of both CDG and CSA is the Contrastive Regression (CR) loss, a novel contrastive loss for regression by pulling features with closer gaze directions closer together while pushing features with farther gaze directions farther apart. Experimentally, we choose ETH-XGAZE and Gaze-360 as the source domain and test the domain generalization and adaptation performance on MPIIGAZE, RT-GENE, GazeCapture, EyeDiap respectively. The results demonstrate that our CRGA achieves remarkable performance improvement compared with the baseline models and also outperforms the state-of-the-art domain adaptation approaches on gaze adaptation tasks.	['Teng Li', 'Hongkai Xiong', 'Chenglin Li', 'Wenrui Dai', 'Bingbing Ni', 'Jin Li', 'Yangzhou Jiang', 'Yaoming Wang']	2022-01-01	null	null	null	cvpr-2022-1	['gaze-estimation']	['computer-vision']	[ 2.34898385e-02 -1.08491443e-01 -2.16526106e-01 -6.82295263e-01 -1.62252977e-01 -2.96369523e-01 3.74419779e-01 -5.82776904e-01 -3.71362507e-01 8.74160111e-01 -7.58244768e-02 2.56594401e-02 6.42598420e-02 2.67871153e-02 -8.43163252e-01 -8.99695277e-01 2.94156820e-01 -6.37531653e-03 1.63041726e-01 -2.43658423e-01 4.27928686e-01 6.08718544e-02 -1.94312394e+00 -2.27549285e-01 1.11642444e+00 1.19338059e+00 9.06186923e-02 4.33656365e-01 1.63193047e-01 6.24090374e-01 -4.04151529e-01 -9.72714275e-02 2.92172194e-01 -5.93814909e-01 -6.58022225e-01 -2.62583137e-01 9.46244597e-01 -3.59313607e-01 2.17897687e-02 9.77685273e-01 5.00352144e-01 3.43419045e-01 8.55901778e-01 -1.74778783e+00 -1.05734336e+00 -2.75097668e-01 -1.24799681e+00 4.62495267e-01 4.21405345e-01 3.83650988e-01 7.70697951e-01 -7.16674447e-01 7.03464746e-01 1.04819894e+00 5.20926058e-01 1.17270744e+00 -1.07515359e+00 -1.30482996e+00 5.62526524e-01 3.31443191e-01 -1.25484335e+00 -6.34753764e-01 7.64009416e-01 -4.90998745e-01 6.48087263e-01 -9.64677185e-02 1.20918743e-01 1.39341033e+00 -9.14491639e-02 7.89467812e-01 1.46203494e+00 -4.53364134e-01 -2.64677429e-03 3.18863928e-01 1.90765470e-01 5.11452675e-01 1.19900703e-02 3.78012925e-01 -9.42616403e-01 5.14249325e-01 5.04984021e-01 1.88810397e-02 -5.12435079e-01 -6.36587620e-01 -6.52270675e-01 5.51676512e-01 1.10145438e+00 -2.38545448e-01 -1.51317164e-01 -5.82501329e-02 1.26950398e-01 4.05022711e-01 8.13276350e-01 2.50794888e-01 -5.59299886e-01 -1.28154531e-01 -7.64194429e-01 2.51486391e-01 2.98958182e-01 1.19678879e+00 1.00484920e+00 -2.34687537e-01 -1.70108408e-01 8.13343406e-01 6.94291949e-01 7.10366666e-01 5.57841241e-01 -6.77661777e-01 4.81018573e-01 8.24266195e-01 -1.02642654e-02 -5.31751096e-01 -5.94332218e-01 -1.91168904e-01 -3.84413749e-01 7.56023169e-01 6.25663042e-01 -3.97930413e-01 -1.07211661e+00 2.29442167e+00 5.56790948e-01 2.35403940e-01 -1.68108672e-01 1.21689105e+00 9.34011459e-01 2.14273036e-01 4.33752745e-01 -8.85109380e-02 1.04088211e+00 -1.16850519e+00 -4.95052785e-01 -4.78263021e-01 5.46653271e-01 -4.53480631e-01 1.51785266e+00 1.56427503e-01 -9.50155854e-01 -6.07806027e-01 -1.09026277e+00 -4.54518616e-01 -2.82156497e-01 1.22372232e-01 4.29620683e-01 4.84771967e-01 -1.07346022e+00 2.01243699e-01 -6.48219109e-01 -6.78423941e-01 8.96869302e-01 6.31458402e-01 -4.83740181e-01 6.60810247e-02 -8.04050684e-01 8.55948627e-01 -1.51760057e-01 -8.63623992e-02 -6.02379620e-01 -9.08190489e-01 -9.41370428e-01 1.83453560e-02 2.28995934e-01 -7.29155421e-01 1.22983623e+00 -1.64842153e+00 -1.88677609e+00 1.24436426e+00 -5.85493684e-01 -3.20706129e-01 3.77402306e-01 -5.23599088e-01 -3.64592254e-01 -1.38934478e-01 6.73573539e-02 1.10274959e+00 1.36267316e+00 -1.13792205e+00 -7.92792439e-01 -7.12271094e-01 -1.08938583e-03 5.90237498e-01 -1.49218246e-01 5.15726283e-02 -5.54071963e-02 -1.38682634e-01 -1.69104666e-01 -1.14576209e+00 6.13967478e-01 3.43693018e-01 -2.97593296e-01 -6.16262615e-01 1.02126932e+00 -3.90018582e-01 1.33450425e+00 -2.31708026e+00 4.09978420e-01 -1.56816766e-01 6.53153181e-01 1.89632937e-01 -4.41038497e-02 -2.02955410e-01 -4.06825423e-01 -2.58053839e-01 -5.83922938e-02 -5.29217541e-01 -3.34340073e-02 -1.92479670e-01 -2.92595178e-01 5.36415815e-01 4.28109050e-01 9.77225542e-01 -8.78265023e-01 -2.08735839e-01 -1.43536881e-01 2.53573686e-01 -5.27494729e-01 3.26238513e-01 -1.39090136e-01 7.47553766e-01 -1.88100666e-01 5.52102089e-01 9.07979906e-01 -5.43567002e-01 -3.08205485e-01 8.93639307e-03 -1.37556046e-01 3.02107811e-01 -4.62614715e-01 1.83300161e+00 -2.52495855e-01 1.06832671e+00 -2.01440647e-01 -4.76309478e-01 9.32217717e-01 -2.54973114e-01 -5.47569133e-02 -1.02989185e+00 3.65759641e-01 -8.81533399e-02 4.87896577e-02 -5.96631110e-01 4.09305215e-01 -7.65779763e-02 1.26454785e-01 5.82436323e-01 3.00243050e-01 3.97313446e-01 -2.48642027e-01 -1.49334565e-01 5.95714688e-01 8.60820293e-01 2.39188582e-01 -1.93741992e-01 7.00694799e-01 -2.39920393e-01 3.92309368e-01 2.79512387e-02 -7.86895216e-01 6.31634653e-01 5.54159820e-01 -2.31062934e-01 -7.23641694e-01 -9.42707837e-01 -2.03339979e-02 1.84913671e+00 4.71667707e-01 2.90120114e-02 -9.51626420e-01 -1.39020848e+00 7.50620663e-02 5.98055720e-01 -1.23586774e+00 -4.35886979e-01 -3.62458080e-01 -3.81308496e-01 1.78468287e-01 5.45345187e-01 7.21310258e-01 -9.87392128e-01 -4.65761602e-01 -7.74188101e-01 1.45466834e-01 -6.57033980e-01 -6.78133190e-01 -8.53222683e-02 -6.20002091e-01 -1.16692185e+00 -9.45337474e-01 -7.68308938e-01 8.59753668e-01 4.65582162e-01 9.28792298e-01 -1.27639174e-01 2.33989701e-01 1.22675933e-01 -1.64031178e-01 -7.57059872e-01 1.85946718e-01 3.69652241e-01 2.06620142e-01 1.64002463e-01 1.19053519e+00 -4.81409013e-01 -8.99738312e-01 5.45870423e-01 -2.61748821e-01 -1.19767122e-01 3.70195121e-01 8.57104421e-01 3.01245064e-01 -7.61190832e-01 5.73356211e-01 -1.02786124e+00 5.67932129e-01 -6.96089625e-01 -7.39221632e-01 6.88127354e-02 -1.00436068e+00 1.38800174e-01 8.85387063e-02 -4.72373933e-01 -1.35585642e+00 -2.10216343e-01 3.47263485e-01 -7.31857121e-01 -2.90296644e-01 1.12284664e-02 3.40966359e-02 -2.58976996e-01 1.18655956e+00 -3.97497229e-02 2.13435397e-01 -4.18819517e-01 2.67147303e-01 8.55742276e-01 4.19552684e-01 -1.71281561e-01 8.12941730e-01 3.61194789e-01 3.35821956e-02 -4.88751143e-01 -1.04756737e+00 -4.29072917e-01 -7.49571741e-01 -2.06334978e-01 8.73834074e-01 -9.82479036e-01 -9.53028679e-01 6.96408510e-01 -7.23087311e-01 -4.72429544e-01 5.93906976e-02 2.50383407e-01 -4.40953195e-01 -9.36433598e-02 2.85565734e-01 -6.63869441e-01 -3.56476188e-01 -8.33173454e-01 1.16881955e+00 9.89457607e-01 -2.90118277e-01 -8.93473387e-01 3.65027517e-01 1.53962985e-01 3.93839210e-01 1.91856205e-01 5.04970968e-01 -5.60993969e-01 -5.48083127e-01 1.86448753e-01 -6.91793740e-01 3.51918310e-01 1.52972415e-01 -1.98545251e-02 -1.42282140e+00 -3.03289503e-01 -3.60950202e-01 -6.47415578e-01 7.97868252e-01 5.22876143e-01 8.68737280e-01 3.95797193e-02 -6.26245141e-01 1.17948782e+00 1.12563276e+00 -6.50805905e-02 6.06924117e-01 6.10673070e-01 8.36534500e-01 7.04113543e-01 7.57181466e-01 5.27088270e-02 7.75201857e-01 6.90363765e-01 4.11996752e-01 -2.86043789e-02 -3.32985133e-01 -2.99899578e-01 2.94233888e-01 8.93465057e-02 -2.97763169e-01 -7.23174214e-02 -8.12852800e-01 3.68089139e-01 -1.75393414e+00 -7.58747280e-01 1.54429734e-01 2.25434279e+00 9.25584197e-01 -7.46018020e-03 4.98267859e-01 -3.22833478e-01 6.23918355e-01 -1.09185435e-01 -1.13150823e+00 -3.22586209e-01 -3.11499033e-02 1.11019455e-01 3.27206761e-01 1.87245339e-01 -1.08766794e+00 1.09185910e+00 5.66824913e+00 1.95874393e-01 -1.70365834e+00 2.32130468e-01 3.29095215e-01 -5.43017209e-01 2.33516768e-01 -2.90834874e-01 -1.05457449e+00 7.11952567e-01 8.86614025e-01 -9.06511396e-03 5.21947682e-01 9.39376533e-01 -1.06605522e-01 -2.71062553e-01 -1.29008067e+00 1.19861209e+00 3.68059337e-01 -5.70579112e-01 -5.56524575e-01 5.01551433e-04 7.56114483e-01 3.74672472e-01 7.02957809e-01 4.95872647e-01 6.14846125e-02 -1.00963354e+00 5.29569089e-01 7.30608821e-01 1.19650054e+00 -3.87439340e-01 4.12748873e-01 -1.30023994e-02 -5.75438499e-01 -3.18065882e-01 -2.49930367e-01 -2.30008528e-01 -2.50528693e-01 -5.25992632e-01 -8.65385830e-01 3.30603011e-02 1.07036602e+00 1.13127708e+00 -8.73921275e-01 1.03575861e+00 -5.26202500e-01 4.07376409e-01 -1.21933371e-01 -2.56078746e-02 4.78526801e-02 -1.40388301e-02 5.70809722e-01 6.62173688e-01 3.24097387e-02 -2.59107202e-01 -6.96575403e-01 1.09973216e+00 -4.25641507e-01 -2.10123762e-01 -4.19188023e-01 3.84801805e-01 5.89966178e-01 1.10830426e+00 1.39262736e-01 8.29395354e-02 -4.72174823e-01 8.79052699e-01 8.23573887e-01 7.19104350e-01 -8.07670832e-01 -4.58256006e-01 1.15041375e+00 2.43875057e-01 3.50660414e-01 1.54454842e-01 -5.17460823e-01 -1.01730525e+00 -9.44430288e-03 -7.23675549e-01 3.34513873e-01 -1.29078352e+00 -1.39765072e+00 7.28909254e-01 1.79130118e-02 -1.51590383e+00 -4.37824279e-01 -6.33309364e-01 -5.32680869e-01 1.43148768e+00 -2.18160343e+00 -1.41188288e+00 -9.95736003e-01 9.82963085e-01 2.43017882e-01 -3.75696868e-01 5.07255554e-01 -3.71707678e-02 -8.13991487e-01 1.17872918e+00 6.39364170e-03 -1.67811528e-01 1.52695370e+00 -1.34528553e+00 1.50484875e-01 5.76230407e-01 -3.79060954e-01 7.93815017e-01 4.87035871e-01 -1.87482521e-01 -6.66820705e-01 -8.87971997e-01 7.39238560e-01 -9.87947047e-01 3.59543353e-01 -4.20515984e-01 -1.11127388e+00 9.62499022e-01 4.84069884e-01 7.44931325e-02 6.55340075e-01 4.40350831e-01 -7.24701941e-01 -3.95023584e-01 -1.07006419e+00 4.94172186e-01 1.17572248e+00 -5.84183991e-01 -7.12924480e-01 -3.54899541e-02 5.38535595e-01 -6.34970129e-01 -4.56466317e-01 1.04821019e-01 6.96521997e-01 -1.05358613e+00 6.04550242e-01 -8.51306796e-01 5.44698119e-01 -3.83377939e-01 3.39651734e-01 -1.31368506e+00 -2.30757892e-01 -8.09253037e-01 -2.85902619e-01 1.09852767e+00 3.12850326e-01 -7.34479189e-01 7.48417914e-01 6.79236174e-01 7.93877803e-03 -4.49102134e-01 -7.42473841e-01 -4.67215568e-01 2.40503237e-01 1.47557080e-01 7.72704840e-01 9.20526862e-01 -3.06543615e-02 8.53899777e-01 -2.88701713e-01 4.75132130e-02 5.48970222e-01 -8.72147381e-02 1.22614014e+00 -1.50027323e+00 1.82917446e-01 -5.51862836e-01 -3.49185467e-01 -1.29746652e+00 3.39982748e-01 -4.75159079e-01 -1.05510950e-02 -8.88621032e-01 -6.65040836e-02 -4.30946201e-01 -4.47267830e-01 5.09909809e-01 -5.02038419e-01 3.55954677e-01 1.48476020e-01 4.46001381e-01 -8.26736271e-01 6.41592920e-01 1.31480682e+00 1.12667076e-01 -6.03202224e-01 1.55994415e-01 -9.58053231e-01 6.46222949e-01 4.36715275e-01 -3.76404732e-01 -6.43855155e-01 -5.97416937e-01 3.60647380e-01 -7.40995586e-01 2.89999664e-01 -8.29504371e-01 4.64164793e-01 4.64553200e-02 5.46588182e-01 -4.50950086e-01 2.19697297e-01 -6.98949456e-01 -5.72047591e-01 -2.16715544e-01 -2.65025944e-01 -1.16818599e-01 4.29870546e-01 6.17307425e-01 9.28070769e-03 1.07016735e-01 9.38629925e-01 4.35407788e-01 -9.32938457e-01 3.08031857e-01 4.45572585e-01 3.72672558e-01 1.13305438e+00 -5.84103823e-01 -7.42891848e-01 -2.32907400e-01 -7.72510171e-01 3.68770510e-01 8.71470809e-01 8.68873417e-01 3.83145928e-01 -1.20104659e+00 -4.87043172e-01 7.14578331e-01 6.89153075e-01 1.38716221e-01 3.14896882e-01 1.13511038e+00 -1.76833585e-01 2.61721313e-01 -6.64105654e-01 -9.58650887e-01 -1.33683431e+00 7.08944619e-01 4.10181969e-01 3.49445671e-01 -1.16523534e-01 1.40875423e+00 8.36476326e-01 -2.35831693e-01 2.32832775e-01 -2.67073810e-01 -5.91822803e-01 -7.19827339e-02 6.22351408e-01 4.24382299e-01 -2.03099802e-01 -8.75561893e-01 -4.15020674e-01 9.02519763e-01 -3.42863142e-01 3.67298603e-01 1.26053739e+00 -7.51423717e-01 9.60663185e-02 3.11032683e-01 1.12740278e+00 -2.89350122e-01 -2.00258684e+00 -4.59030718e-01 -1.05876267e-01 -5.61242163e-01 9.80108008e-02 -1.18638062e+00 -8.82389307e-01 1.12772894e+00 1.00790691e+00 -2.27116182e-01 1.61035776e+00 2.98252385e-02 4.80185628e-01 1.95619967e-02 -5.50379492e-02 -8.93452823e-01 2.10491456e-02 4.11478102e-01 8.70056570e-01 -1.73195660e+00 -8.93241391e-02 -1.18870251e-01 -9.55942452e-01 7.68739402e-01 1.25250316e+00 -1.96689323e-01 7.57871449e-01 -5.11137307e-01 2.44220525e-01 -2.39103511e-01 -6.40827060e-01 -3.51068944e-01 8.29090297e-01 1.00240326e+00 3.45596969e-01 -3.05866361e-01 1.33705676e-01 5.69014132e-01 -2.43674546e-01 2.98623353e-01 1.91800013e-01 5.79707742e-01 -1.89861104e-01 -7.66395926e-01 -1.61296532e-01 1.99793041e-01 -2.15839624e-01 -7.60982856e-02 -4.76998687e-01 1.04060185e+00 1.95003152e-01 6.91270828e-01 3.09001625e-01 -4.07372802e-01 3.85865539e-01 1.82812646e-01 6.48066938e-01 -5.88432312e-01 -3.38748723e-01 -1.32821932e-01 -2.88396329e-01 -6.55975461e-01 -6.79709792e-01 -7.31718063e-01 -9.15050626e-01 -3.29749703e-01 -3.93751055e-01 -3.76635253e-01 5.70760369e-01 8.92229974e-01 7.80146480e-01 3.86047840e-01 6.29693091e-01 -8.44967306e-01 -4.09388900e-01 -1.30988443e+00 -4.25916284e-01 5.16146660e-01 9.71526504e-01 -1.42919636e+00 -3.74347478e-01 1.61793172e-01]	[14.152359008789062, 0.010211425833404064]
3482a6c3-df45-408a-ad24-22180baed117	visualizing-color-wise-saliency-of-black-box	2010.02468	null	https://arxiv.org/abs/2010.02468v1	https://arxiv.org/pdf/2010.02468v1.pdf	Visualizing Color-wise Saliency of Black-Box Image Classification Models	Image classification based on machine learning is being commonly used. However, a classification result given by an advanced method, including deep learning, is often hard to interpret. This problem of interpretability is one of the major obstacles in deploying a trained model in safety-critical systems. Several techniques have been proposed to address this problem; one of which is RISE, which explains a classification result by a heatmap, called a saliency map, which explains the significance of each pixel. We propose MC-RISE (Multi-Color RISE), which is an enhancement of RISE to take color information into account in an explanation. Our method not only shows the saliency of each pixel in a given image as the original RISE does, but the significance of color components of each pixel; a saliency map with color information is useful especially in the domain where the color information matters (e.g., traffic-sign recognition). We implemented MC-RISE and evaluate them using two datasets (GTSRB and ImageNet) to demonstrate the effectiveness of our methods in comparison with existing techniques for interpreting image classification results.	['Kohei Suenaga', 'Yoshinori Konishi', 'Hiroki Sakuma', 'Yuhki Hatakeyama']	2020-10-06	null	null	null	null	['traffic-sign-recognition']	['computer-vision']	[ 4.82768744e-01 6.88402429e-02 -2.44824722e-01 -6.79258883e-01 -1.21474840e-01 -1.46315873e-01 3.82347941e-01 1.37541100e-01 -1.07867874e-01 6.55934513e-01 6.53722137e-02 -5.28740227e-01 -6.11363165e-02 -6.26022756e-01 -8.09679508e-01 -6.23936057e-01 2.97034979e-01 9.05491505e-03 4.38024163e-01 -4.12180543e-01 5.89420676e-01 3.12627733e-01 -1.94690275e+00 8.05431426e-01 1.05856097e+00 1.11284423e+00 2.32170716e-01 4.96768951e-01 -4.84693378e-01 1.09887159e+00 -7.98112214e-01 -3.59357119e-01 -4.32166010e-02 -7.30985165e-01 -8.63263607e-01 2.42023915e-01 4.44340855e-01 -2.61273533e-01 3.03094417e-01 1.26968026e+00 -1.33937284e-01 -1.05813220e-01 6.33604944e-01 -1.83032763e+00 -8.29181314e-01 4.99048173e-01 -7.08405137e-01 2.89277762e-01 -4.83624153e-02 7.25260153e-02 8.97480011e-01 -6.96289062e-01 2.45055839e-01 1.32536840e+00 2.67629176e-01 6.40206099e-01 -8.22118819e-01 -4.29713428e-01 1.83437213e-01 8.85746300e-01 -9.29360569e-01 1.60094827e-01 7.89668560e-01 -3.15673411e-01 2.08478928e-01 6.78473234e-01 6.46910071e-01 5.82587183e-01 3.12203228e-01 1.04736948e+00 1.38212359e+00 -5.26016712e-01 4.10102487e-01 2.47185841e-01 3.97819996e-01 6.16309285e-01 4.62029576e-01 -7.40179792e-02 -5.00203073e-01 3.72131377e-01 6.37503445e-01 3.21796209e-01 -2.45185509e-01 -3.05988044e-01 -1.04978538e+00 5.04857421e-01 1.10261095e+00 6.43851981e-02 -2.49243915e-01 2.71140665e-01 1.94369316e-01 -8.89689252e-02 2.32252479e-01 2.37146899e-01 -3.16189378e-01 1.24528976e-02 -5.58710754e-01 -3.32421362e-02 3.13485712e-01 5.00664651e-01 8.98286581e-01 1.40745819e-01 -1.58219382e-01 5.50441861e-01 3.39258224e-01 5.50693154e-01 3.84137511e-01 -7.07265198e-01 2.73110569e-01 8.49936604e-01 2.77123637e-02 -1.08173144e+00 -3.13682765e-01 -4.17894989e-01 -8.48445356e-01 8.97388935e-01 3.21866184e-01 1.93902627e-01 -1.24837053e+00 1.40713906e+00 5.35189621e-02 2.04073727e-01 1.08309634e-01 1.49580121e+00 1.05570436e+00 6.16186619e-01 3.59699100e-01 4.54243690e-01 1.47166860e+00 -1.16553819e+00 -7.50941038e-01 -4.37938690e-01 2.97258556e-01 -3.62128854e-01 1.42960811e+00 2.82673120e-01 -4.69590634e-01 -7.26868987e-01 -1.26423395e+00 6.19370304e-02 -8.14149082e-01 1.92625433e-01 7.78359830e-01 6.10201478e-01 -8.62927794e-01 4.44348216e-01 -3.83095801e-01 -3.66635352e-01 4.59045202e-01 7.33562037e-02 -1.43610388e-01 -2.57289838e-02 -1.17089391e+00 1.10894251e+00 5.21373332e-01 2.31335849e-01 -4.79678541e-01 -4.00582403e-01 -7.15474069e-01 2.51439422e-01 1.54864356e-01 -3.33256006e-01 1.04109764e+00 -1.52541518e+00 -1.08186507e+00 6.75136447e-01 -2.79189587e-01 -3.63874495e-01 3.71735185e-01 -2.72980183e-01 -6.42586589e-01 1.48798093e-01 -6.22239932e-02 8.91794145e-01 9.85702038e-01 -1.59123063e+00 -8.73323321e-01 -1.98881850e-01 3.20950240e-01 9.43537429e-02 -1.59425393e-01 -4.97852340e-02 -2.11848930e-01 -3.85813236e-01 1.13636173e-01 -6.26606166e-01 -8.84007514e-02 2.88796872e-01 -3.85832429e-01 -1.15914755e-01 1.23725522e+00 -7.21635044e-01 1.07852733e+00 -2.10650897e+00 -3.01036209e-01 2.75593162e-01 3.94630522e-01 4.71503228e-01 7.57635832e-02 -1.58883721e-01 -3.74857515e-01 3.62450719e-01 -4.08086300e-01 1.50190324e-01 -1.46108210e-01 3.28816533e-01 -1.82396263e-01 -1.66712925e-02 5.56825459e-01 8.38698089e-01 -1.01758564e+00 -4.63855952e-01 4.27298814e-01 4.60447133e-01 -1.74210310e-01 -3.35533521e-03 -1.82152346e-01 4.12865102e-01 -4.45170850e-01 5.33537865e-01 6.90236330e-01 -4.83980209e-01 -2.33242244e-01 -3.53435755e-01 -1.83259528e-02 -1.82360291e-01 -7.38008142e-01 1.06095219e+00 -1.75951242e-01 1.02215862e+00 -6.35250986e-01 -9.22896206e-01 1.06075680e+00 -6.32635280e-02 -3.41183655e-02 -9.56673980e-01 7.28794411e-02 2.69218743e-01 1.64591923e-01 -7.48890460e-01 6.22224331e-01 1.15275998e-02 1.39133543e-01 4.41064656e-01 -6.20885670e-01 2.32755840e-01 1.77431926e-01 2.64356494e-01 6.47721946e-01 2.16094479e-01 1.93867728e-01 -1.49121955e-01 7.13282943e-01 3.16006392e-01 5.90021670e-01 5.74588895e-01 -3.76016527e-01 9.87552583e-01 6.16660178e-01 -8.56790841e-01 -1.02656925e+00 -7.95649946e-01 1.39998168e-01 1.04290891e+00 7.43969977e-01 -6.38218746e-02 -9.49703157e-01 -8.78727317e-01 -3.49344127e-02 1.01772642e+00 -8.12790036e-01 -3.26639354e-01 -4.56438422e-01 -7.97812164e-01 5.99890314e-02 6.64001048e-01 1.06334126e+00 -1.24941432e+00 -1.10131609e+00 -2.48849630e-01 -2.85586685e-01 -9.49012816e-01 -3.21752638e-01 -3.62465866e-02 -7.84322262e-01 -1.24017167e+00 -4.74076301e-01 -5.66248894e-01 1.19249654e+00 5.18269777e-01 1.04619014e+00 6.86159730e-01 -3.03096712e-01 -7.16680586e-02 -7.03898549e-01 -7.10916996e-01 -4.45318192e-01 -2.13025108e-01 -4.52584922e-01 4.82495606e-01 3.95523906e-01 2.02406272e-01 -7.93304920e-01 4.14437950e-01 -1.18525434e+00 7.73191214e-01 9.25675869e-01 6.71131730e-01 2.68939823e-01 -5.82625577e-03 4.76102740e-01 -9.72974300e-01 4.82500345e-01 -1.54345304e-01 -3.08014393e-01 6.58258200e-01 -5.80639780e-01 4.62451220e-01 4.70649093e-01 3.61167081e-02 -1.08181787e+00 -1.64200649e-01 3.05558145e-02 -3.60925198e-02 -3.30105394e-01 3.99385959e-01 -8.40275139e-02 -5.14724366e-02 5.96272528e-01 1.89169973e-01 1.62302345e-01 -3.96510810e-01 2.80067176e-01 7.76682198e-01 6.27640188e-01 -8.56375769e-02 6.10255003e-01 2.92680383e-01 1.29974559e-01 -4.71989095e-01 -9.26890314e-01 -8.63186345e-02 -5.62555134e-01 -6.63584828e-01 9.02491510e-01 -3.42419028e-01 -5.48356116e-01 4.21760470e-01 -1.15229547e+00 -6.75527304e-02 -8.33058879e-02 3.38686168e-01 -2.49191985e-01 2.14692965e-01 -1.33836493e-01 -8.07882667e-01 -2.73447931e-01 -1.07284486e+00 9.24550295e-01 7.12099671e-01 -9.64682326e-02 -9.40960288e-01 -6.41153872e-01 4.87160087e-01 6.71475172e-01 4.85980719e-01 1.31242394e+00 -3.54179382e-01 -5.90255320e-01 -5.78820184e-02 -7.93202639e-01 4.16744590e-01 1.71218961e-01 1.18543603e-01 -9.35681105e-01 3.00473511e-01 -3.22774917e-01 6.46424741e-02 9.67627347e-01 3.10182750e-01 1.57203114e+00 -2.47066587e-01 -1.74712062e-01 2.04152733e-01 1.44035363e+00 3.71440679e-01 1.00186753e+00 6.48395181e-01 7.26144016e-01 6.15163505e-01 8.81892323e-01 1.71285495e-02 4.38954055e-01 5.81863105e-01 8.04784417e-01 -8.53306234e-01 -4.62333202e-01 -2.44144529e-01 1.78349242e-01 2.55142570e-01 -2.30951861e-01 -1.16311595e-01 -9.12926674e-01 2.86619216e-01 -2.09641695e+00 -7.47817457e-01 -6.55556440e-01 1.94554281e+00 4.64556307e-01 1.75165460e-01 7.40907714e-03 5.87250352e-01 1.03674066e+00 -6.79258704e-02 -5.47191679e-01 -6.25676215e-01 -1.30052760e-01 -8.76397416e-02 3.63146394e-01 2.60762721e-01 -9.67303216e-01 8.21911812e-01 6.11285019e+00 5.88494897e-01 -1.49135363e+00 2.73171570e-02 1.02590239e+00 5.79795420e-01 -3.07478130e-01 3.59660275e-02 -3.57092589e-01 5.73753655e-01 4.96439427e-01 3.62114012e-02 1.42376930e-01 9.89216626e-01 3.75681847e-01 -4.63695794e-01 -9.64300811e-01 1.07396936e+00 3.47452253e-01 -1.12284636e+00 3.75849873e-01 -3.26129258e-01 5.73262215e-01 -5.54075897e-01 3.62064809e-01 -1.96101107e-02 -3.28360475e-03 -9.99997616e-01 1.03786719e+00 3.63493234e-01 6.33607686e-01 -5.33969939e-01 9.77843761e-01 1.06721759e-01 -1.07644033e+00 -1.32044166e-01 -2.69319594e-01 -1.86532214e-01 1.62467901e-02 5.75228691e-01 -9.57441568e-01 3.85323048e-01 7.57093072e-01 6.77483022e-01 -1.15223312e+00 1.40244973e+00 -8.77178669e-01 5.98246157e-01 1.38653308e-01 -3.81616354e-01 2.25993484e-01 8.60159006e-03 2.50450969e-01 9.88364875e-01 1.92504957e-01 -4.89626788e-02 -2.74705917e-01 1.17187989e+00 3.34129274e-01 -1.69199347e-01 -1.70459867e-01 3.35526973e-01 -6.27557263e-02 1.17480469e+00 -1.10814548e+00 -6.80676699e-01 -2.23166987e-01 1.01374638e+00 -1.64674565e-01 6.04615569e-01 -1.11089587e+00 -5.45023382e-01 4.54826176e-01 1.58328310e-01 7.00759664e-02 1.73431367e-01 -8.71688008e-01 -8.38440776e-01 9.04377848e-02 -6.72974110e-01 4.26601201e-01 -1.41521573e+00 -7.76991248e-01 7.93233335e-01 3.21650580e-02 -1.48023903e+00 7.59166703e-02 -9.09384847e-01 -7.57984340e-01 7.28267729e-01 -1.77040875e+00 -1.08259439e+00 -9.14795935e-01 3.41779262e-01 7.09148228e-01 7.12254420e-02 3.76910180e-01 1.63479373e-01 -5.98789275e-01 2.51987934e-01 -3.26320499e-01 1.17064856e-01 3.25703353e-01 -1.41399217e+00 3.38732123e-01 1.02520323e+00 1.12256691e-01 2.29233339e-01 7.91141689e-01 -4.55151021e-01 -8.42873037e-01 -9.88328755e-01 8.22259545e-01 -3.31270993e-01 8.37919042e-02 -8.75484794e-02 -1.04779613e+00 2.42927641e-01 1.45006984e-01 -3.21588516e-02 3.83118868e-01 -1.30178347e-01 -3.21136653e-01 -4.94047672e-01 -1.13746715e+00 6.43431783e-01 6.72923684e-01 -1.33946553e-01 -7.88431644e-01 -7.64045268e-02 4.39935446e-01 -2.72294492e-01 -5.96570957e-04 3.70760262e-01 5.85793912e-01 -1.17970037e+00 7.01875985e-01 -6.15883410e-01 6.13732934e-01 -9.41711128e-01 3.49453121e-01 -1.45495296e+00 -3.02397430e-01 1.02307566e-01 2.13482559e-01 8.60715985e-01 7.35749245e-01 -6.01938128e-01 6.81602895e-01 8.84617686e-01 -1.70666710e-01 -4.36898828e-01 -3.70929539e-01 -4.90852207e-01 -5.43379188e-01 -2.70952880e-01 9.56063986e-01 6.14736497e-01 -8.45395625e-02 1.06439091e-01 -3.68909538e-01 1.58175066e-01 6.72166646e-01 2.03762263e-01 6.58363760e-01 -1.18794906e+00 2.16249347e-01 -4.86370265e-01 -8.55136931e-01 -6.21209562e-01 -2.55698383e-01 -5.29313147e-01 2.24660426e-01 -1.86502492e+00 5.08909047e-01 -2.69523591e-01 -5.83862841e-01 7.96530426e-01 -6.01702750e-01 3.23897868e-01 4.05411541e-01 1.35547101e-01 -7.77377725e-01 3.45401764e-01 1.38323307e+00 -3.54834318e-01 9.74451751e-02 -1.33429006e-01 -8.82174253e-01 7.76537716e-01 9.13426220e-01 -2.95824558e-01 -2.76715726e-01 -5.51472843e-01 -4.00265045e-02 -4.80737925e-01 7.54330873e-01 -1.16261649e+00 1.97907522e-01 -2.57862359e-01 5.80237448e-01 -5.91839373e-01 -7.39566833e-02 -9.80250835e-01 1.11269422e-01 7.68604338e-01 -5.39797783e-01 7.34037757e-02 1.10294133e-01 3.32409531e-01 -5.18952429e-01 -1.44807398e-01 8.94889891e-01 7.23843114e-04 -1.65374601e+00 -1.10183336e-01 -1.45817190e-01 -3.49316835e-01 9.52351809e-01 -5.09574413e-01 -6.59613013e-01 -5.80179036e-01 -3.98534805e-01 1.88414156e-01 1.29287660e-01 8.61617267e-01 9.74710286e-01 -1.47281921e+00 -5.70133090e-01 3.44109356e-01 4.05662179e-01 -2.12185159e-01 2.19936728e-01 6.98961437e-01 -9.53747690e-01 3.47614914e-01 -6.12931967e-01 -7.86723316e-01 -1.45475340e+00 3.47816557e-01 2.66226172e-01 1.70231014e-01 -5.91143131e-01 5.86447656e-01 4.85017747e-01 1.70126900e-01 1.68791726e-01 -7.02575266e-01 -6.24627709e-01 -4.17511791e-01 8.58984232e-01 1.77717626e-01 -5.49403876e-02 -6.44513607e-01 -4.28606808e-01 7.43196964e-01 1.00634098e-01 8.61923099e-02 1.05035472e+00 -1.06871031e-01 -2.09143624e-01 3.81421953e-01 9.30595398e-01 -6.60834730e-01 -1.27522004e+00 -9.39528048e-02 1.97425947e-01 -7.94355631e-01 1.13813533e-02 -1.15826070e+00 -1.41988552e+00 1.09918308e+00 8.01078796e-01 2.30941743e-01 1.34333181e+00 -1.51206627e-01 5.58274686e-01 1.71239331e-01 3.67553115e-01 -1.05464745e+00 2.33857691e-01 1.56317726e-01 1.02736068e+00 -1.58387697e+00 -1.74838945e-01 -5.25357723e-01 -1.16525471e+00 1.39419556e+00 9.37780082e-01 2.23174557e-01 2.93515563e-01 -1.74035475e-01 3.61282319e-01 -2.35125542e-01 -3.64740610e-01 -4.54642326e-01 5.66273868e-01 4.97488111e-01 1.91679418e-01 1.22985423e-01 -3.13953131e-01 3.81005138e-01 -1.54132838e-03 5.92604764e-02 4.96304899e-01 9.26899433e-01 -6.08763933e-01 -8.58266115e-01 -7.39921153e-01 2.45718285e-01 -2.86161005e-01 9.46593750e-03 -6.64874077e-01 5.91594100e-01 3.98637682e-01 9.82755542e-01 7.61385188e-02 -6.19910836e-01 2.47341454e-01 -1.05408460e-01 -5.15443645e-02 -4.28621680e-01 -2.59086132e-01 -4.93121415e-01 -1.27328753e-01 -4.67489034e-01 -5.03271341e-01 -2.45263711e-01 -1.60957110e+00 -1.78144693e-01 -4.40294780e-02 1.51768014e-01 1.05868852e+00 1.06077302e+00 1.49902567e-01 8.84252012e-01 4.70483214e-01 -5.41681349e-01 -2.53144372e-02 -6.80644095e-01 -5.90827465e-01 9.30644572e-01 4.93463904e-01 -7.40756094e-01 -4.40197468e-01 3.13603550e-01]	[9.975016593933105, 2.0305206775665283]
417adad0-69a2-49a3-b5ea-793ac107c678	geoglue-a-geographic-language-understanding	2305.06545	null	https://arxiv.org/abs/2305.06545v1	https://arxiv.org/pdf/2305.06545v1.pdf	GeoGLUE: A GeoGraphic Language Understanding Evaluation Benchmark	With a fast developing pace of geographic applications, automatable and intelligent models are essential to be designed to handle the large volume of information. However, few researchers focus on geographic natural language processing, and there has never been a benchmark to build a unified standard. In this work, we propose a GeoGraphic Language Understanding Evaluation benchmark, named GeoGLUE. We collect data from open-released geographic resources and introduce six natural language understanding tasks, including geographic textual similarity on recall, geographic textual similarity on rerank, geographic elements tagging, geographic composition analysis, geographic where what cut, and geographic entity alignment. We also pro vide evaluation experiments and analysis of general baselines, indicating the effectiveness and significance of the GeoGLUE benchmark.	['Xiaofeng He', 'Fei Huang', 'Ning Guo', 'Xin Li', 'Yao Xu', 'Pengjun Xie', 'Boli Chen', 'Zheng Li', 'Qiang Zhang', 'Ruixue Ding', 'Dongyang Li']	2023-05-11	null	null	null	null	['entity-alignment', 'entity-alignment']	['knowledge-base', 'natural-language-processing']	[-5.74598610e-01 -2.23682821e-01 -5.47434211e-01 -6.84936345e-01 -8.61207426e-01 -8.54610503e-01 1.06665254e+00 6.80828869e-01 -6.17689550e-01 6.14825845e-01 9.02132630e-01 -4.04213250e-01 -2.40332797e-01 -8.47940505e-01 -2.61108011e-01 -4.79738712e-02 -1.35636136e-01 6.49114251e-01 8.65588114e-02 -2.13867307e-01 8.29716504e-01 2.13824868e-01 -1.17689419e+00 1.26055330e-01 1.48056233e+00 7.32926071e-01 9.03298855e-02 3.90336961e-01 -5.01005709e-01 7.77395427e-01 -7.08281875e-01 -1.90904289e-01 4.42387611e-02 -9.06760469e-02 -1.13032568e+00 -6.45911694e-01 5.72210789e-01 -1.96044803e-01 -2.06231192e-01 9.06047404e-01 2.55448341e-01 3.41339886e-01 8.64332080e-01 -1.28802562e+00 -1.23717141e+00 9.75638568e-01 -7.80689001e-01 5.05028725e-01 5.69345534e-01 -9.45782885e-02 1.20959091e+00 -9.14638460e-01 8.67137074e-01 1.04378879e+00 7.01995254e-01 -3.04314811e-02 -5.97815692e-01 -4.20865119e-01 2.65428156e-01 1.35309249e-01 -1.59775078e+00 -2.36392796e-01 1.42314151e-01 -4.39743459e-01 1.01277971e+00 2.00621724e-01 1.16796114e-01 5.79532623e-01 -2.72897948e-02 6.61924481e-01 8.56699049e-01 -3.16872180e-01 1.10920712e-01 2.17773858e-02 5.81130147e-01 5.82945645e-01 2.29157701e-01 -3.02534491e-01 -5.31340420e-01 -2.73381501e-01 2.77108610e-01 -2.15135813e-02 -2.68070698e-01 5.49665012e-04 -1.40666401e+00 8.60858142e-01 7.76461720e-01 4.37730342e-01 -1.78575009e-01 1.25602797e-01 4.07573760e-01 1.57432929e-01 7.94165015e-01 8.33343625e-01 -3.10050637e-01 -2.56334156e-01 -9.74274516e-01 2.84671485e-01 9.00614142e-01 1.09183836e+00 8.15552115e-01 -4.67193484e-01 2.90825635e-01 9.48625684e-01 3.11158657e-01 7.96313465e-01 4.91087407e-01 -6.26368403e-01 7.34617591e-01 8.53336215e-01 1.92018569e-01 -1.46309662e+00 -6.74208105e-01 2.24337112e-02 -6.70458674e-01 -5.08171618e-01 4.56980437e-01 -2.08887249e-01 -8.65607917e-01 1.68855035e+00 1.82110339e-01 2.50598311e-01 1.28711656e-01 7.90806890e-01 1.02916908e+00 1.03197563e+00 3.52193415e-01 5.82446575e-01 1.43639112e+00 -8.59859705e-01 -4.74922597e-01 -5.29261410e-01 1.09245574e+00 -7.85157204e-01 1.13841391e+00 -2.07742348e-01 -5.57508707e-01 -2.02527091e-01 -7.65564203e-01 -3.86021823e-01 -1.14000297e+00 1.29404455e-01 1.01392341e+00 3.57005298e-01 -1.04083908e+00 2.90997736e-02 -5.64171016e-01 -9.70075727e-01 -7.84695446e-02 -1.87676977e-02 -4.91320580e-01 -1.60307884e-01 -1.53520656e+00 1.06963515e+00 5.45668066e-01 -8.41729715e-02 -2.89414883e-01 -1.11919153e+00 -1.12826228e+00 2.08387002e-02 5.79303205e-02 -4.96717423e-01 9.79819238e-01 -3.10348481e-01 -4.70172435e-01 9.97389495e-01 -1.31448805e-01 -4.56676722e-01 6.72360286e-02 -1.96361274e-01 -8.76927555e-01 -5.33433333e-02 8.41887295e-01 6.95176303e-01 -2.53082544e-01 -9.07586753e-01 -1.18563259e+00 -4.74488467e-01 -5.31926975e-02 2.63841897e-01 -2.32723996e-01 1.94627777e-01 -5.91263592e-01 -7.44347692e-01 1.60949416e-02 -7.12875843e-01 -3.20499092e-01 -5.35050809e-01 -3.55026215e-01 -4.30229872e-01 5.00069797e-01 -9.03775573e-01 1.66457391e+00 -2.15841103e+00 -3.99238169e-01 5.16049385e-01 2.49051794e-01 -2.48363912e-01 -3.03280890e-01 7.12595344e-01 2.60083497e-01 6.89676225e-01 -1.66198790e-01 1.39782876e-01 2.30557457e-01 -1.88332126e-01 -8.14462185e-01 2.78204173e-01 -2.25639418e-01 7.46773660e-01 -1.20820487e+00 -5.09827197e-01 -7.83665292e-03 7.06488267e-02 -4.59162593e-01 1.23504186e-02 -1.83677167e-01 -2.04451203e-01 -7.76122570e-01 9.29768205e-01 3.48870993e-01 -5.18022656e-01 -8.84929523e-02 6.79307058e-02 -4.73107696e-01 5.47048330e-01 -9.43292439e-01 1.63805616e+00 -7.09586322e-01 9.15260911e-01 -4.12033081e-01 -2.75906235e-01 8.63701224e-01 -3.33845645e-01 3.46518189e-01 -8.85220468e-01 -2.96727717e-01 3.79327416e-01 -4.32400733e-01 -4.80751932e-01 1.27617824e+00 4.72373426e-01 -9.04959083e-01 8.62897992e-01 -2.61819482e-01 -1.20891780e-01 4.14464891e-01 7.23486245e-01 9.52565789e-01 -2.46594504e-01 5.42680740e-01 -7.07861304e-01 2.60428727e-01 7.59524822e-01 1.72942221e-01 9.25552130e-01 -6.36853874e-02 6.22286379e-01 3.16061109e-01 -7.57522702e-01 -8.38204503e-01 -1.07261610e+00 -1.11084640e-01 1.32947850e+00 6.58382893e-01 -6.97958469e-01 -5.76943219e-01 -5.24971902e-01 1.59368604e-01 8.42699587e-01 -7.06886232e-01 1.15288071e-01 -4.75818068e-01 -9.12401378e-01 9.08966899e-01 5.75955391e-01 6.89100087e-01 -8.48465800e-01 -3.14039767e-01 1.36714518e-01 -4.42537516e-01 -9.07222152e-01 -8.64464760e-01 -3.92324299e-01 -3.29355448e-01 -9.47981417e-01 -4.68780577e-01 -8.21513772e-01 3.20924073e-01 4.90198970e-01 1.35886121e+00 2.41043776e-01 -9.55401435e-02 3.20719600e-01 -3.97484154e-01 -4.95473295e-01 1.61523432e-01 8.06058824e-01 -1.29093397e-02 -3.64194483e-01 8.81084800e-01 -1.31026030e-01 -6.92879617e-01 3.87214810e-01 -7.90067315e-01 -8.62007514e-02 4.17455018e-01 6.16674662e-01 1.62975863e-01 -2.85908937e-01 7.62207150e-01 -7.51251876e-01 7.41445065e-01 -1.09574115e+00 -5.08970201e-01 7.70476222e-01 -5.32158196e-01 1.20011427e-01 5.42569578e-01 1.11283176e-01 -1.25609863e+00 -3.97940814e-01 1.58202723e-02 5.84875584e-01 -2.41834179e-01 1.17276192e+00 2.55601387e-02 9.83539373e-02 7.31930256e-01 1.29852638e-01 -5.71597695e-01 -4.84533995e-01 8.45414460e-01 1.01078951e+00 7.59275675e-01 -7.88156390e-01 6.34201705e-01 6.52716696e-01 -7.27574289e-01 -1.11148906e+00 -8.21861088e-01 -9.48129594e-01 -6.09605432e-01 7.38195851e-02 9.03145850e-01 -1.03554094e+00 -3.57323527e-01 4.56929594e-01 -1.07985783e+00 -2.38978386e-01 2.01836869e-01 5.84100425e-01 -5.29297814e-02 2.65791595e-01 -4.35433805e-01 -2.83028424e-01 -2.31922656e-01 -6.37856185e-01 1.04058492e+00 5.70009947e-01 -4.97807622e-01 -1.47350824e+00 5.61076462e-01 2.49793217e-01 6.35804236e-01 1.39586195e-01 9.24142778e-01 -9.80784237e-01 -6.57656729e-01 -1.79762882e-03 -5.23683667e-01 -5.31365275e-01 1.20961390e-01 1.14145607e-01 -5.32849371e-01 8.90825465e-02 -6.32927716e-01 -2.26993769e-01 9.87587988e-01 3.48787546e-01 8.36037040e-01 -3.81579787e-01 -7.87188172e-01 7.81145930e-01 1.48452544e+00 1.99871510e-01 5.79223394e-01 7.33891249e-01 6.35489464e-01 8.10593188e-01 8.77910733e-01 3.18897337e-01 1.18039572e+00 3.17982972e-01 -2.06817359e-01 -2.59293824e-01 1.87371120e-01 -4.82795984e-01 -1.13599211e-01 6.52036071e-01 1.66114971e-01 -6.18645847e-01 -1.68953109e+00 1.15072334e+00 -1.71306026e+00 -1.02456176e+00 -8.42398778e-02 1.82017171e+00 6.82360411e-01 -3.75698507e-01 -1.80527255e-01 -7.62883306e-01 6.86010003e-01 4.47368294e-01 -3.81585926e-01 -9.96678770e-02 -3.02491158e-01 -3.03502083e-01 6.66204095e-01 5.79815269e-01 -1.29615116e+00 1.45981145e+00 6.62615442e+00 7.01869965e-01 -1.11469936e+00 -2.93287367e-01 8.15351307e-01 2.27981940e-01 -6.05588973e-01 1.71037316e-01 -8.49028409e-01 4.30861980e-01 8.67899060e-01 -7.65672982e-01 1.13051925e-02 1.11485517e+00 7.40378648e-02 -4.06653881e-01 -9.40657616e-01 8.31874251e-01 2.62808233e-01 -1.52751398e+00 1.10053167e-01 -9.19669867e-02 8.02744687e-01 5.08838773e-01 -5.50188608e-02 1.93553135e-01 9.20760393e-01 -1.17175579e+00 5.10750055e-01 3.44588608e-01 5.79819381e-01 -6.33324802e-01 8.11674476e-01 -5.31386621e-02 -1.46579206e+00 1.93675235e-01 -3.16636145e-01 1.90456718e-01 4.37848240e-01 3.95347238e-01 -7.81251550e-01 5.34309447e-01 9.03202236e-01 8.81430447e-01 -8.88365030e-01 1.19389498e+00 -1.71911135e-01 4.23563182e-01 -4.66068178e-01 -2.83053130e-01 7.33042121e-01 -4.17656958e-01 2.46134207e-01 1.39574158e+00 4.74148989e-01 3.50892633e-01 2.35050857e-01 5.32267809e-01 -2.78459996e-01 3.37814331e-01 -8.54846716e-01 -1.34242639e-01 1.01369405e+00 1.03137755e+00 -7.63535500e-01 -5.52052855e-01 -4.96357977e-01 4.74953115e-01 3.56992215e-01 2.93952137e-01 -8.09033334e-01 -7.72669911e-01 9.07499433e-01 -1.13228500e-01 -4.03730571e-01 -3.03117573e-01 -6.36337519e-01 -1.20345354e+00 -2.11373672e-01 -7.27832258e-01 9.48860109e-01 -9.94989872e-01 -1.44468296e+00 5.92860281e-01 2.42108211e-01 -8.95901203e-01 -3.86819780e-01 -2.10271850e-01 -7.08466709e-01 7.58440077e-01 -1.42105770e+00 -1.13865089e+00 -5.64967811e-01 3.85364473e-01 3.83910060e-01 -1.54057726e-01 4.94341403e-01 3.75989139e-01 -3.88857722e-01 4.16253567e-01 3.17945629e-01 4.28744078e-01 1.08349860e+00 -1.42169464e+00 9.56363440e-01 9.61222708e-01 2.83912599e-01 1.12301815e+00 4.45692599e-01 -8.16647351e-01 -1.11261809e+00 -1.16545057e+00 1.49373913e+00 -7.56059587e-01 1.19134116e+00 -2.47868404e-01 -9.77501571e-01 9.80789304e-01 3.37845415e-01 -5.13376713e-01 7.54059315e-01 4.47001606e-01 -6.51217759e-01 -9.56232622e-02 -9.11175132e-01 8.33018064e-01 9.70920861e-01 -5.90837240e-01 -1.09390867e+00 4.02012736e-01 7.21397579e-01 -1.66830048e-01 -8.50916564e-01 1.10036120e-01 3.54189336e-01 -3.78941298e-01 9.58805263e-01 -5.98625898e-01 3.09634179e-01 -5.36060095e-01 2.01186687e-02 -1.54061913e+00 -2.32527316e-01 -1.29866078e-01 7.70561039e-01 1.51678455e+00 8.86142075e-01 -7.71992564e-01 6.01695955e-01 8.05139720e-01 -1.77442968e-01 -3.10102329e-02 -4.32709575e-01 -6.51430666e-01 1.59594610e-01 -1.44801721e-01 1.03279340e+00 1.48291790e+00 4.19896901e-01 1.00251444e-01 -2.25272238e-01 4.35846776e-01 4.33028281e-01 4.05175626e-01 7.74173498e-01 -1.12513494e+00 5.06391168e-01 -6.39924705e-01 -2.11732179e-01 -1.07990777e+00 2.01755837e-01 -7.01599240e-01 2.29625627e-02 -1.91672766e+00 2.44638219e-01 -9.35152888e-01 -1.06733274e-02 6.30710125e-01 -3.23239386e-01 1.51185706e-01 -1.35248587e-01 4.38665301e-01 -9.89407539e-01 2.73230493e-01 3.08978230e-01 -2.91673422e-01 -2.96628952e-01 -6.29204273e-01 -9.45038915e-01 7.20609307e-01 8.58490229e-01 -3.63845706e-01 -3.60873371e-01 -8.83577466e-01 3.95389736e-01 -2.30446130e-01 -8.74631165e-04 -7.65885472e-01 8.01766813e-01 -5.83837628e-01 1.07738562e-01 -8.20724070e-01 -3.50086123e-01 -3.18269908e-01 -8.93931687e-02 7.48247877e-02 -4.49341446e-01 6.76633120e-01 7.88370222e-02 2.58533150e-01 -6.76545858e-01 8.27804878e-02 3.85476381e-01 2.85401382e-02 -1.70980251e+00 3.07508588e-01 -4.00176108e-01 6.93003893e-01 9.21967626e-01 -1.75071843e-02 -8.15606296e-01 -3.34590137e-01 -1.29543409e-01 8.36274803e-01 8.35149050e-01 7.66612709e-01 3.72784615e-01 -1.14272535e+00 -8.17038834e-01 -2.32439980e-01 7.23182440e-01 -9.19322371e-02 3.76266778e-01 3.49494874e-01 -1.03491974e+00 7.69742429e-01 -1.27230391e-01 -2.54379481e-01 -9.03417110e-01 3.52660030e-01 2.30569452e-01 -2.76215434e-01 -3.32684875e-01 5.31238735e-01 4.40683991e-01 -7.73719311e-01 6.07205406e-02 -3.27086449e-01 -3.42060596e-01 1.47553548e-01 9.80426669e-01 2.46740177e-01 -7.29753673e-02 -6.52067542e-01 -6.11349404e-01 6.66034043e-01 4.10893075e-02 -1.17962860e-01 1.45777881e+00 -3.57428402e-01 -4.29957092e-01 2.51589745e-01 1.01971602e+00 2.16852948e-01 -5.07448733e-01 -2.33420953e-01 8.94847453e-01 -5.10139704e-01 -1.89257875e-01 -9.74213302e-01 -8.32464159e-01 5.00366628e-01 1.40913054e-01 2.37031892e-01 8.17751944e-01 3.22017431e-01 7.61557996e-01 5.72130084e-01 5.42110920e-01 -1.21706498e+00 -3.86197835e-01 9.37109053e-01 7.58189321e-01 -1.28004777e+00 -8.30650330e-02 -3.16034377e-01 -7.01546311e-01 6.27429903e-01 7.92540789e-01 2.12675348e-01 7.48659611e-01 2.18413442e-01 2.55193263e-01 -5.02385855e-01 -4.93867397e-01 -2.43608937e-01 4.64960545e-01 6.68556809e-01 4.51070577e-01 -2.85203531e-02 -2.13941485e-01 3.78116727e-01 -4.96641636e-01 -3.27245831e-01 4.42489773e-01 5.09311080e-01 -5.54979503e-01 -6.79253697e-01 -2.99667686e-01 4.66041088e-01 -4.21819329e-01 -5.72662592e-01 -6.39419675e-01 1.30321038e+00 -3.23849440e-01 9.77036238e-01 3.20941865e-01 -9.81890783e-02 1.33703604e-01 -3.61277342e-01 -2.79092342e-01 -6.00756407e-01 -5.37156224e-01 -6.25843644e-01 3.03743333e-01 -4.72884178e-01 -1.81593552e-01 -5.12807608e-01 -1.38867581e+00 -6.69762015e-01 5.82462102e-02 5.95281243e-01 7.45973587e-01 7.78446436e-01 5.85330546e-01 -2.60496199e-01 3.15004438e-01 -2.12610856e-01 7.96010867e-02 -9.68112409e-01 -5.42278111e-01 5.72285891e-01 -5.05547710e-02 -4.02224779e-01 -4.51971292e-01 -2.29857683e-01]	[9.318225860595703, 9.077142715454102]
f25d3364-9050-4b5a-8342-5e9585395761	is-automated-topic-model-evaluation-broken-1	null	null	https://openreview.net/forum?id=tjdHCnPqoo	https://openreview.net/pdf?id=tjdHCnPqoo	Is Automated Topic Model Evaluation Broken? The Incoherence of Coherence	Topic model evaluation, like evaluation of other unsupervised methods, can be contentious. However, the field has coalesced around automated estimates of topic coherence, which rely on the frequency of word co-occurrences in a reference corpus. Contemporary neural topic models surpass classical ones according to these metrics. At the same time, topic model evaluation suffers from a validation gap: automated coherence, developed for classical models, has not been validated using human experimentation for neural models. In addition, a meta-analysis of topic modeling literature reveals a substantial standardization gap in automated topic modeling benchmarks. To address the validation gap, we compare automated coherence with the two most widely accepted human judgment tasks: topic rating and word intrusion. To address the standardization gap, we systematically evaluate a dominant classical model and two state-of-the-art neural models on two commonly used datasets. Automated evaluations declare a winning model when corresponding human evaluations do not, calling into question the validity of fully automatic evaluations independent of human judgments.	['Philip Resnik', 'Jordan Lee Boyd-Graber', 'Denis Peskov', 'Andrew Hian-Cheong', 'Pranav Goel', 'Alexander Hoyle']	2021-05-21	null	null	null	neurips-2021-12	['topic-models']	['natural-language-processing']	[ 7.76749849e-02 5.41195989e-01 -2.49946520e-01 -3.61219347e-01 -8.90577614e-01 -4.50600296e-01 1.04297936e+00 7.11372733e-01 -5.64805984e-01 4.93312567e-01 4.12513644e-01 -5.41102998e-02 -3.65960389e-01 -7.92051971e-01 -2.00261697e-01 -4.70587283e-01 1.77978754e-01 9.15626764e-01 2.99416244e-01 -5.78117138e-03 5.83591759e-01 -4.24028188e-01 -1.87781250e+00 1.15194589e-01 1.03028250e+00 8.44325840e-01 6.99179843e-02 1.33399516e-01 -4.24587339e-01 2.86102682e-01 -8.78460944e-01 -6.02198422e-01 -4.46030200e-01 -2.13356867e-01 -6.53573692e-01 -2.23332778e-01 6.50556028e-01 6.98798941e-03 6.96953163e-02 1.11591542e+00 3.31030726e-01 1.06421269e-01 1.05847335e+00 -1.25578713e+00 -6.00815773e-01 1.13269365e+00 2.10298821e-02 3.56470905e-02 1.73082411e-01 -1.42492950e-01 1.40315998e+00 -8.57360661e-01 9.53453124e-01 1.03340709e+00 9.17266667e-01 1.06350154e-01 -1.44993150e+00 -6.23486519e-01 1.51403740e-01 1.20750196e-01 -1.40105033e+00 -3.15770917e-02 7.42172599e-01 -8.78762007e-01 9.89872038e-01 1.71587486e-02 7.20782816e-01 1.38452089e+00 1.81759685e-01 2.77251363e-01 1.32280827e+00 -3.06298375e-01 5.38058877e-01 6.21350348e-01 9.68858361e-01 -1.47364782e-02 7.94885933e-01 -1.44033760e-01 -6.69263601e-01 -2.59600937e-01 1.27742425e-01 -4.84404057e-01 -2.14837939e-01 -2.07838759e-01 -1.17094481e+00 1.18934190e+00 -9.99872014e-02 6.62918091e-01 -5.25041997e-01 -1.95186377e-01 5.17921925e-01 3.68830264e-02 1.02286780e+00 9.10814703e-01 -1.35891795e-01 -3.03704113e-01 -1.75766027e+00 6.61891282e-01 1.02123177e+00 6.22175336e-01 8.16982985e-01 2.07254943e-02 -2.33355150e-01 9.41100776e-01 5.11409998e-01 3.44575047e-01 8.48104000e-01 -6.80404127e-01 1.28189802e-01 6.51620567e-01 -1.35732353e-01 -1.34839451e+00 -5.14987350e-01 -7.30625451e-01 -7.41427362e-01 1.57602001e-02 2.03024745e-01 1.05384462e-01 -5.60999393e-01 1.70248830e+00 -1.83607657e-02 -1.96322203e-01 -7.28601366e-02 5.50579250e-01 1.01275671e+00 5.87380171e-01 4.10279751e-01 -3.59960228e-01 1.37240970e+00 -6.37605667e-01 -9.70040679e-01 -1.42188042e-01 4.66889769e-01 -5.35700023e-01 1.20179784e+00 7.68621564e-01 -9.95803833e-01 -4.88296986e-01 -1.24277115e+00 1.46041857e-02 -8.72374117e-01 3.24314530e-03 6.21955514e-01 9.39175546e-01 -1.13524294e+00 5.46054244e-01 -6.37187481e-01 -6.06097817e-01 3.12485516e-01 2.00489983e-01 -1.90708339e-01 5.33414066e-01 -1.51237381e+00 1.22980917e+00 6.40404701e-01 -4.03365016e-01 -8.74794185e-01 -9.10983443e-01 -6.04336202e-01 1.69898078e-01 3.33367407e-01 -5.24908602e-01 1.32025027e+00 -7.09162414e-01 -1.37461579e+00 1.05273318e+00 -2.79252697e-02 -9.22551572e-01 1.59607410e-01 -4.97008115e-01 -4.02510136e-01 -4.59197946e-02 2.78149605e-01 7.55757809e-01 5.25698602e-01 -1.28811693e+00 -4.93579209e-01 -2.07490861e-01 -4.36390579e-01 1.51059523e-01 -7.15943873e-01 -4.72677760e-02 -8.51413086e-02 -7.40935981e-01 -1.00523857e-02 -6.64466560e-01 -2.80334726e-02 -5.27610481e-01 -4.89737779e-01 -7.77994394e-01 5.46827435e-01 -2.75656968e-01 1.68195665e+00 -1.91308033e+00 -1.86683267e-01 8.89719725e-02 5.56331694e-01 -9.09216106e-02 1.93348855e-01 5.08648813e-01 -6.69619516e-02 3.08798313e-01 -2.78670192e-01 -5.97416341e-01 3.42156917e-01 -3.16981196e-01 -1.00034308e+00 3.34056109e-01 -1.21403307e-01 6.66571558e-01 -6.68230951e-01 -6.86175883e-01 4.38826047e-02 4.98894542e-01 -5.58709323e-01 -3.19313109e-02 -4.00711179e-01 -4.40173149e-01 -1.47609830e-01 3.08508486e-01 2.43410751e-01 -3.99464697e-01 1.41510323e-01 -9.81234759e-02 -2.86526829e-01 7.82192826e-01 -8.58714998e-01 1.72759128e+00 1.28403872e-01 1.19588327e+00 -5.88510215e-01 -8.56747806e-01 1.09995198e+00 5.42515159e-01 4.97140259e-01 -3.58825833e-01 2.26591110e-01 2.33737141e-01 -6.34219917e-03 1.29364729e-01 1.26281726e+00 -3.12854536e-04 -1.22180283e-01 7.46342182e-01 5.63199878e-01 -4.64211702e-01 1.95260525e-01 2.55377233e-01 6.74225450e-01 -5.94536476e-02 3.66349190e-01 -7.02103913e-01 -4.56493311e-02 4.16510612e-01 7.15717748e-02 9.07689512e-01 -1.04724862e-01 8.04454923e-01 5.96124947e-01 -3.00494999e-01 -1.02954257e+00 -1.14606333e+00 -5.27255535e-01 1.20489502e+00 -4.80939113e-02 -9.63417590e-01 -1.16334152e+00 -2.73272514e-01 -2.57616848e-01 1.17135823e+00 -1.08389187e+00 -1.41048178e-01 1.17225470e-02 -9.66135144e-01 6.57759249e-01 9.96995047e-02 1.01002239e-01 -1.01308525e+00 -9.78679955e-01 3.72418374e-01 -2.32300550e-01 -9.27106917e-01 2.57843167e-01 2.80312002e-01 -1.04528821e+00 -7.02202857e-01 -7.84098446e-01 -2.30961114e-01 3.05998977e-02 1.28440663e-01 1.61957347e+00 -3.35794896e-01 1.12691514e-01 3.11381072e-01 -2.86324114e-01 -7.61929810e-01 -2.72142619e-01 6.33574724e-01 1.59648880e-01 -4.75931257e-01 9.63186860e-01 -6.24343216e-01 -4.54888612e-01 2.67911226e-01 -1.02972293e+00 -8.83901268e-02 3.28623384e-01 6.46119654e-01 3.70433778e-01 -1.30344909e-02 7.30368078e-01 -9.89648819e-01 1.13074625e+00 -6.83959365e-01 -3.40712398e-01 6.16832785e-02 -1.33373404e+00 -1.77241430e-01 -1.41913861e-01 -4.23403263e-01 -8.24119925e-01 -7.50365734e-01 2.87097573e-01 -4.79882024e-02 -3.35951000e-01 9.87156868e-01 3.43906313e-01 6.56264782e-01 9.87137437e-01 9.75169763e-02 -8.69459584e-02 -2.91991800e-01 4.89325523e-01 5.50282478e-01 5.63639104e-01 -5.63966095e-01 4.38899338e-01 3.67047399e-01 -6.86092496e-01 -9.07990575e-01 -8.96508574e-01 -5.06375194e-01 -3.91157210e-01 -2.45393157e-01 8.21197212e-01 -9.83422041e-01 -1.84437796e-01 2.89069146e-01 -1.43209934e+00 -1.31400704e-01 -4.16012377e-01 6.44324481e-01 -6.29806399e-01 1.18089482e-01 -1.70168862e-01 -6.90825284e-01 -6.11975014e-01 -9.60172534e-01 9.65714574e-01 1.04170427e-01 -1.13579297e+00 -1.01994181e+00 8.11458111e-01 -3.51258777e-02 8.13989341e-01 -2.00775843e-02 9.80872571e-01 -1.36995888e+00 -6.01794533e-02 -2.54311532e-01 -7.38689750e-02 -5.20577878e-02 -1.68257594e-01 1.69544354e-01 -1.26936710e+00 1.44306228e-01 1.27126366e-01 -2.49664515e-01 1.03993416e+00 5.90312362e-01 6.63707554e-01 -9.13902596e-02 -4.66388196e-01 7.99672678e-03 1.30480230e+00 3.33379656e-02 4.54867631e-01 6.25370145e-01 4.48582619e-02 9.85802829e-01 2.97324508e-01 2.56272376e-01 4.80340719e-01 6.74697280e-01 8.97164270e-02 2.82018572e-01 3.46519537e-02 -1.68290481e-01 1.92711845e-01 1.04411757e+00 -1.18501067e-01 -3.22725207e-01 -1.34581423e+00 6.38471842e-01 -1.66738272e+00 -9.10261571e-01 -7.13078603e-02 2.18545294e+00 1.00014949e+00 5.67182243e-01 1.29827276e-01 2.68298119e-01 5.51990151e-01 2.98989087e-01 -1.21882178e-01 -1.54811516e-01 -2.63448030e-01 2.82510370e-03 -1.41160339e-01 2.58557111e-01 -1.18702710e+00 1.05556083e+00 7.23132801e+00 9.94475245e-01 -1.02884579e+00 4.24018294e-01 6.45653248e-01 -4.85921353e-02 -5.15539289e-01 -5.57288639e-02 -8.18955421e-01 3.88208240e-01 1.27941263e+00 -5.51447809e-01 -3.55825186e-01 1.32476437e+00 -1.08081922e-01 -3.46423596e-01 -1.17624068e+00 7.34648705e-01 4.20834929e-01 -1.29328430e+00 2.13695869e-01 2.15352535e-01 8.48636746e-01 1.27297118e-01 3.63094360e-01 6.94808185e-01 2.34879345e-01 -1.02786767e+00 8.06523860e-01 6.03036284e-01 4.36084479e-01 -3.47687006e-01 8.71539414e-01 -1.23504484e-02 -5.26548743e-01 4.53224450e-01 -4.77424622e-01 8.19587111e-02 1.71059236e-01 8.79896820e-01 -7.11495996e-01 1.14011593e-01 6.22437239e-01 4.50685650e-01 -7.44760513e-01 1.04022884e+00 1.00160830e-01 1.01223564e+00 -3.80949348e-01 -2.14232087e-01 3.85862857e-01 -1.10128425e-01 7.56309748e-01 1.52727544e+00 1.79707542e-01 -3.93327504e-01 -1.93233043e-01 1.23584187e+00 2.46216729e-01 4.96070683e-01 -7.51617432e-01 -3.26811343e-01 5.88445008e-01 1.12187493e+00 -1.03422773e+00 -5.72549820e-01 -1.51378378e-01 3.47071171e-01 2.34529644e-01 2.76009440e-01 -6.81107163e-01 -3.01751286e-01 4.11245286e-01 4.74676453e-02 -1.66127607e-01 -1.94338515e-01 -8.07359040e-01 -9.13425744e-01 -2.29870155e-01 -7.53907084e-01 1.46060020e-01 -6.74981296e-01 -1.23273456e+00 1.10029912e+00 5.25486469e-01 -1.01339388e+00 -4.16818708e-01 -2.95019716e-01 -6.94275081e-01 5.72482765e-01 -9.86956954e-01 -8.82261157e-01 -4.27308679e-01 8.87234062e-02 6.21566176e-01 -3.68209839e-01 1.03334320e+00 4.66896407e-02 -3.73571485e-01 4.29955482e-01 -1.07365772e-01 -3.12405348e-01 8.59420598e-01 -1.37334788e+00 5.04300535e-01 3.27842951e-01 3.09921890e-01 9.95517910e-01 1.04818714e+00 -7.40145147e-01 -5.57744384e-01 -7.72300363e-01 1.05933523e+00 -7.61559665e-01 7.63727427e-01 -4.02944714e-01 -1.19024885e+00 3.85144711e-01 5.23783028e-01 -9.38639343e-01 1.05711687e+00 7.92961955e-01 -6.66427135e-01 2.51541287e-01 -4.58535075e-01 5.74741304e-01 3.85872424e-01 -5.98070979e-01 -1.14745772e+00 2.38552675e-01 8.77135754e-01 1.93213180e-01 -8.75052691e-01 3.73698086e-01 7.27195024e-01 -9.57796633e-01 7.62311459e-01 -5.00451207e-01 7.72751868e-01 7.01364055e-02 -4.04678643e-01 -1.27363431e+00 -6.23532385e-02 -3.65003645e-01 -7.22921500e-03 1.41430819e+00 6.74253345e-01 -5.11689067e-01 7.58299708e-01 5.66088915e-01 -1.79378316e-01 -7.21426129e-01 -8.58276546e-01 -6.05037868e-01 4.61681873e-01 -8.72753859e-01 3.02154064e-01 1.04472518e+00 5.10807633e-01 6.08758032e-01 1.98367089e-02 -3.10943544e-01 5.56679845e-01 -1.06191941e-01 6.82416439e-01 -1.87969542e+00 1.67364866e-01 -1.21604478e+00 -3.75076264e-01 -3.49463731e-01 1.09122366e-01 -6.58547461e-01 2.20353499e-01 -1.59523058e+00 4.18746710e-01 -1.09162688e-01 -1.97347730e-01 7.25526735e-02 -2.46619672e-01 2.58845657e-01 -3.31258052e-03 5.40171981e-01 -8.11745822e-01 6.52636290e-01 3.46735984e-01 -7.20119923e-02 -1.41114339e-01 -4.52365845e-01 -7.79674888e-01 1.00039399e+00 8.08862090e-01 -5.99986076e-01 -5.89425802e-01 -1.32722020e-01 7.59821892e-01 -4.92319554e-01 2.49458238e-01 -1.42501497e+00 5.02007663e-01 2.40604818e-01 -1.33305803e-01 -9.27635729e-01 3.24973136e-01 -4.92210865e-01 6.08497001e-02 1.59601733e-01 -7.88946688e-01 -4.38575773e-03 1.74105555e-01 5.44391513e-01 -4.65365469e-01 -2.52684385e-01 2.96197534e-01 9.43643376e-02 -3.53360265e-01 2.35165376e-02 -7.88962841e-01 2.74580687e-01 6.32494748e-01 -3.41375202e-01 -5.89546561e-01 -5.13438165e-01 -5.47479987e-01 -1.92712381e-01 2.73811847e-01 5.71594238e-01 3.16240638e-01 -1.01986456e+00 -7.64890790e-01 -4.05855358e-01 3.39783907e-01 -2.92472959e-01 1.84769988e-01 8.45231473e-01 -1.41071379e-01 1.07747650e+00 9.90258809e-03 -7.18860507e-01 -7.50164628e-01 5.63514054e-01 -2.63787135e-02 -4.95943546e-01 -4.26367640e-01 4.01216269e-01 4.45840597e-01 -1.95461988e-01 4.96329784e-01 -5.11674702e-01 -4.71315444e-01 6.18037760e-01 4.48611528e-01 3.87514055e-01 1.54068798e-01 -3.40776682e-01 3.04062502e-03 4.36913699e-01 -2.54292041e-01 -6.40867472e-01 1.15450096e+00 5.62130772e-02 1.39999520e-02 1.19374967e+00 8.32860172e-01 -3.80851269e-01 -5.62357187e-01 -2.29217872e-01 4.64892924e-01 1.80158347e-01 1.24673344e-01 -8.57868254e-01 -2.82345772e-01 9.71448958e-01 5.03849626e-01 7.03285933e-01 6.44841254e-01 -7.73824677e-02 2.13311478e-01 5.10708153e-01 1.77541316e-01 -1.35982597e+00 6.07086048e-02 6.37653708e-01 9.37157393e-01 -1.22386324e+00 2.21796364e-01 -3.65939677e-01 -6.30303383e-01 8.80026937e-01 5.00567794e-01 -2.19273686e-01 9.93193984e-01 -5.01089692e-02 -2.71544303e-03 -7.13894904e-01 -9.76174355e-01 2.39936560e-02 6.88319981e-01 4.82759893e-01 7.62117088e-01 7.85862580e-02 -8.01101208e-01 1.21093082e+00 -8.69150341e-01 -1.51579633e-01 3.25380027e-01 3.11637610e-01 -6.57167673e-01 -5.58349609e-01 -1.56634584e-01 5.02863824e-01 -5.09413779e-01 -3.84744972e-01 -5.19618273e-01 9.95517015e-01 -1.03925668e-01 9.69814777e-01 3.57335031e-01 -4.19517338e-01 -1.05808496e-01 5.24563551e-01 -1.60925165e-01 -7.78341532e-01 -7.96171844e-01 7.71784931e-02 1.22969881e-01 -3.66434425e-01 -5.47998011e-01 -6.26433074e-01 -5.66068292e-01 1.13338567e-01 -5.46715915e-01 5.85257947e-01 9.93546903e-01 1.01390219e+00 1.73287973e-01 4.35490996e-01 -6.55848458e-02 -6.74589217e-01 -4.68709469e-01 -1.68488371e+00 -4.33084577e-01 2.06393540e-01 -2.79003084e-01 -8.99926603e-01 -4.60895896e-01 1.64491788e-01]	[10.382389068603516, 7.1030354499816895]
18637b52-4c8f-4f32-9e48-678bcc7ec4d3	mostgan-v-video-generation-with-temporal	2304.02777	null	https://arxiv.org/abs/2304.02777v1	https://arxiv.org/pdf/2304.02777v1.pdf	MoStGAN-V: Video Generation with Temporal Motion Styles	Video generation remains a challenging task due to spatiotemporal complexity and the requirement of synthesizing diverse motions with temporal consistency. Previous works attempt to generate videos in arbitrary lengths either in an autoregressive manner or regarding time as a continuous signal. However, they struggle to synthesize detailed and diverse motions with temporal coherence and tend to generate repetitive scenes after a few time steps. In this work, we argue that a single time-agnostic latent vector of style-based generator is insufficient to model various and temporally-consistent motions. Hence, we introduce additional time-dependent motion styles to model diverse motion patterns. In addition, a Motion Style Attention modulation mechanism, dubbed as MoStAtt, is proposed to augment frames with vivid dynamics for each specific scale (i.e., layer), which assigns attention score for each motion style w.r.t deconvolution filter weights in the target synthesis layer and softly attends different motion styles for weight modulation. Experimental results show our model achieves state-of-the-art performance on four unconditional $256^2$ video synthesis benchmarks trained with only 3 frames per clip and produces better qualitative results with respect to dynamic motions. Code and videos have been made available at https://github.com/xiaoqian-shen/MoStGAN-V.	['Mohamed Elhoseiny', 'Xiang Li', 'Xiaoqian Shen']	2023-04-05	null	http://openaccess.thecvf.com//content/CVPR2023/html/Shen_MoStGAN-V_Video_Generation_With_Temporal_Motion_Styles_CVPR_2023_paper.html	http://openaccess.thecvf.com//content/CVPR2023/papers/Shen_MoStGAN-V_Video_Generation_With_Temporal_Motion_Styles_CVPR_2023_paper.pdf	cvpr-2023-1	['video-generation']	['computer-vision']	[ 1.64203539e-01 -2.67228812e-01 -2.09566668e-01 -1.65890425e-01 -5.48567891e-01 -5.42295635e-01 7.16613710e-01 -8.49537373e-01 -2.51353290e-02 6.60052001e-01 5.53723752e-01 8.86614993e-03 3.30985665e-01 -7.19074249e-01 -7.73297846e-01 -8.44768167e-01 1.14008822e-01 -1.85708866e-01 1.27807468e-01 -2.80910164e-01 -2.52594631e-02 2.87983268e-01 -1.24832833e+00 5.20640194e-01 6.66719019e-01 6.97122991e-01 4.39317197e-01 9.73459959e-01 1.56565413e-01 1.04202139e+00 -6.30084813e-01 -3.35896522e-01 4.11457419e-01 -1.09529340e+00 -5.33421099e-01 2.99130499e-01 6.86670184e-01 -5.71743786e-01 -8.83640647e-01 9.33118761e-01 5.39488137e-01 3.48263413e-01 4.11978602e-01 -1.10743582e+00 -9.54564750e-01 6.16654396e-01 -5.55059195e-01 3.87788326e-01 3.09713006e-01 7.96522379e-01 8.36309552e-01 -9.52772677e-01 9.46487188e-01 1.34280646e+00 2.83151239e-01 1.06015265e+00 -1.28980219e+00 -7.80080855e-01 4.01365668e-01 2.86950320e-01 -1.25134909e+00 -6.36006773e-01 1.03967214e+00 -3.11441392e-01 8.78311455e-01 3.42065483e-01 7.58627653e-01 1.73072982e+00 3.28157395e-01 7.94568002e-01 7.38173664e-01 -1.36758303e-02 1.38348028e-01 -5.50499380e-01 -7.17891634e-01 4.71228212e-01 -3.51462401e-02 2.04016760e-01 -7.84376800e-01 3.44041646e-01 1.22305393e+00 -1.87647209e-01 -4.17620778e-01 -4.58082892e-02 -1.60183084e+00 6.87918365e-01 3.67197633e-01 2.90191561e-01 -5.71003318e-01 7.17401624e-01 3.62463653e-01 2.11314857e-01 4.12838995e-01 4.38660651e-01 -1.03412509e-01 -3.62294346e-01 -1.11925268e+00 4.52224553e-01 3.35832834e-01 1.26916659e+00 4.16815281e-01 9.52624500e-01 -7.44744420e-01 6.46718562e-01 -2.52543613e-02 5.66027939e-01 5.64760864e-01 -1.45831835e+00 4.69023257e-01 -8.22903495e-03 2.39303440e-01 -1.10392308e+00 2.88878623e-02 -4.18775290e-01 -1.17973685e+00 1.20775998e-01 2.71594107e-01 -3.81989866e-01 -9.38619792e-01 2.02892995e+00 1.38869792e-01 6.13446951e-01 -6.24745116e-02 1.09011984e+00 7.99076319e-01 1.06103265e+00 1.76032484e-01 -2.09034845e-01 1.15986407e+00 -1.25710690e+00 -9.91904676e-01 -2.94557989e-01 2.86330003e-02 -9.82743382e-01 1.14881694e+00 2.79877037e-01 -1.65439284e+00 -9.84109879e-01 -8.78265083e-01 -7.39262402e-02 1.84529543e-01 7.34752864e-02 5.03189087e-01 2.21934751e-01 -1.08782566e+00 5.11024177e-01 -9.63555634e-01 -1.06869459e-01 2.74691641e-01 -6.98323548e-02 -1.30819246e-01 -4.60025631e-02 -1.27632689e+00 5.27176082e-01 2.13297695e-01 1.34136677e-01 -1.03600824e+00 -5.71547210e-01 -1.03926134e+00 -3.40680964e-02 6.67189956e-02 -1.12027502e+00 1.25165045e+00 -1.25912833e+00 -1.78108859e+00 3.60577285e-01 -4.47405994e-01 -4.95677799e-01 7.24778414e-01 -2.07262114e-01 -5.14413834e-01 2.99066573e-01 1.26162112e-01 1.11683917e+00 1.18431354e+00 -1.24976075e+00 -5.95867038e-01 3.78655225e-01 9.64732189e-03 2.08073243e-01 -2.88540572e-01 -3.32857296e-02 -8.39357913e-01 -1.51780713e+00 -1.41171262e-01 -1.04397702e+00 -3.72021377e-01 -8.99322256e-02 -2.38261253e-01 1.55603468e-01 8.18651855e-01 -7.78395057e-01 1.62531173e+00 -2.05082607e+00 7.15061009e-01 -2.86355823e-01 6.47192150e-02 1.66738182e-01 -4.79923159e-01 2.35872388e-01 -4.45040129e-02 8.28251988e-02 -2.47623786e-01 -4.76364434e-01 -3.60868536e-02 1.20937534e-01 -4.56837714e-01 1.91400230e-01 5.08390307e-01 1.02450442e+00 -1.11979222e+00 -2.31629878e-01 3.65313292e-01 6.90635443e-01 -8.09890270e-01 2.93079197e-01 -3.17850351e-01 9.27403808e-01 -3.55578989e-01 5.84965706e-01 4.84849185e-01 -3.26768935e-01 1.10052943e-01 -4.09395725e-01 -1.22335464e-01 8.68476778e-02 -1.09206915e+00 1.98153985e+00 -5.82595229e-01 8.11954796e-01 -2.20062405e-01 -6.32528663e-01 6.39390826e-01 5.30692995e-01 4.71382558e-01 -6.77177787e-01 1.75421890e-02 8.54114965e-02 3.31419781e-02 -4.68552768e-01 6.81862116e-01 -3.26668285e-02 -1.35052577e-01 9.68730301e-02 1.14478767e-01 -2.30306149e-01 4.36181873e-01 5.94790019e-02 1.03823578e+00 5.41229129e-01 5.36944121e-02 5.90204895e-02 3.93143862e-01 -2.65520096e-01 7.82604754e-01 5.62353134e-01 -1.42697319e-01 1.06362712e+00 1.94869041e-01 -4.63359654e-01 -1.42338181e+00 -1.08782530e+00 3.86946768e-01 9.31304753e-01 1.82250276e-01 -2.99248844e-01 -5.88313282e-01 -2.45844945e-01 -4.49256152e-01 8.31538498e-01 -6.47646606e-01 -2.79499501e-01 -1.10124540e+00 -4.54827458e-01 4.53008175e-01 5.43745518e-01 5.74027300e-01 -1.36226571e+00 -7.11860538e-01 5.16106606e-01 -5.32712221e-01 -1.15908277e+00 -1.07020569e+00 -4.31241661e-01 -6.40843391e-01 -5.77348292e-01 -1.24386382e+00 -7.46668696e-01 4.68823522e-01 2.91798413e-01 1.30749738e+00 -2.57835865e-01 -5.83772548e-02 2.18728468e-01 -4.88831311e-01 1.34019088e-03 -4.22260195e-01 -1.02299720e-01 -1.17134750e-01 2.51489192e-01 -6.49951771e-02 -6.10949814e-01 -1.10057855e+00 2.79293984e-01 -1.16915894e+00 5.31841815e-01 4.68558997e-01 9.93571997e-01 5.57802856e-01 -2.49981076e-01 4.59569961e-01 -5.46027839e-01 4.41802055e-01 -6.01881087e-01 -3.24106336e-01 -1.84229203e-02 4.47271504e-02 -4.21560481e-02 8.29259276e-01 -8.82904828e-01 -1.16176510e+00 3.00338659e-02 -8.14296603e-02 -9.31996465e-01 -2.22064946e-02 1.45189613e-01 -6.08484037e-02 2.70849854e-01 6.66693807e-01 5.98590672e-01 -2.80569375e-01 -7.99846500e-02 6.13306344e-01 2.78928727e-02 6.64267898e-01 -5.64083159e-01 1.04115796e+00 5.98196268e-01 -2.83498764e-02 -7.27723420e-01 -6.24449492e-01 -8.41215104e-02 -2.01125577e-01 -5.26640475e-01 1.08620548e+00 -1.23397613e+00 -2.47573882e-01 6.15525424e-01 -1.19739020e+00 -8.22568476e-01 -3.49076003e-01 4.46257561e-01 -8.85365546e-01 2.50283241e-01 -9.28173959e-01 -4.32165891e-01 -3.43011856e-01 -1.23055363e+00 1.01209712e+00 2.56686717e-01 -5.87544441e-01 -8.07628274e-01 -3.84965762e-02 3.78141291e-02 7.08033562e-01 5.70795655e-01 3.47034544e-01 3.93012136e-01 -8.04211795e-01 1.44002870e-01 8.12743306e-02 2.52919912e-01 3.69297564e-01 1.68362036e-01 -7.44648337e-01 -3.56001645e-01 3.66869126e-03 -9.35184732e-02 8.90604615e-01 5.95615864e-01 1.34675169e+00 -6.04076385e-01 1.96455956e-01 9.71771181e-01 1.20977688e+00 4.27186638e-01 8.51658344e-01 1.12298824e-01 1.01736999e+00 3.98793012e-01 5.50108671e-01 6.14458203e-01 2.07911119e-01 8.42342556e-01 2.66633034e-01 -1.38253137e-01 -6.10422611e-01 -2.52681702e-01 5.75882792e-01 8.12277377e-01 -4.41318005e-01 -5.53249121e-01 -5.61740994e-01 6.75991893e-01 -1.83893335e+00 -1.60599184e+00 7.82839358e-02 1.73182094e+00 9.99159276e-01 1.13920763e-01 7.93075562e-02 -1.02897123e-01 6.50948882e-01 6.61446869e-01 -6.41361475e-01 -2.24171564e-01 -5.01102448e-01 2.67575324e-01 2.91374207e-01 5.21260142e-01 -9.91244733e-01 1.27386546e+00 5.43232393e+00 8.96437526e-01 -1.44401777e+00 -2.10434198e-03 8.83360565e-01 -4.91905510e-01 -5.13383448e-01 -2.30386198e-01 -5.38411617e-01 8.38732421e-01 7.80403137e-01 -1.98057130e-01 6.55836940e-01 5.53453505e-01 7.56682694e-01 3.98304999e-01 -8.81608725e-01 1.02595925e+00 -3.79667655e-02 -1.76861596e+00 5.32754719e-01 -2.95202583e-01 1.24300361e+00 -2.81156391e-01 4.89602536e-01 2.30999976e-01 2.43404835e-01 -1.08047819e+00 1.21321905e+00 6.48531616e-01 9.76108670e-01 -6.07940614e-01 1.71787873e-01 -6.48119226e-02 -1.47102559e+00 9.45362151e-02 -8.57889131e-02 4.42097895e-02 4.62721109e-01 1.74399301e-01 -1.03495032e-01 4.40732867e-01 6.57592595e-01 9.64130878e-01 -2.23959148e-01 7.00361967e-01 -2.84742922e-01 6.28483534e-01 1.08752176e-01 2.50260800e-01 5.13870239e-01 -9.93327200e-02 6.38425410e-01 1.36198413e+00 7.77180612e-01 2.67073125e-01 -3.09446566e-02 8.81164789e-01 -1.13390582e-02 -3.16786319e-01 -5.77086568e-01 -3.30483392e-02 4.18598384e-01 1.05598783e+00 -4.74825621e-01 -4.99368578e-01 -5.13200104e-01 1.36448514e+00 -1.22926757e-01 7.57157028e-01 -1.18608809e+00 -1.11444168e-01 7.75072157e-01 5.17230062e-03 5.26213169e-01 -5.16055405e-01 -1.42980888e-01 -1.40325034e+00 3.21606994e-02 -1.06429279e+00 1.47526160e-01 -8.95510674e-01 -1.13740921e+00 7.92816818e-01 -1.34112954e-01 -1.78477836e+00 -6.05552554e-01 -2.66732782e-01 -7.44296730e-01 9.54770744e-01 -1.25699139e+00 -1.13041258e+00 -4.23003167e-01 6.99504852e-01 1.22902274e+00 -1.09352134e-01 4.95794743e-01 3.99331838e-01 -4.66731101e-01 6.17410004e-01 -2.23300502e-01 7.31428787e-02 8.22370887e-01 -1.07041240e+00 7.97714591e-01 1.25243378e+00 1.03249000e-02 5.26833117e-01 1.02756572e+00 -5.86348236e-01 -1.52264452e+00 -1.45175469e+00 7.69741178e-01 -2.74260014e-01 5.70472956e-01 -1.46897793e-01 -8.83661985e-01 5.18572152e-01 5.51300406e-01 1.41788214e-01 4.55013335e-01 -7.76005149e-01 -1.73998445e-01 7.78649822e-02 -7.27940917e-01 1.21608293e+00 1.38837349e+00 -3.62590343e-01 -1.60281494e-01 5.58905303e-02 8.38853896e-01 -6.28975868e-01 -8.32510650e-01 3.98343563e-01 5.12408733e-01 -9.50504363e-01 1.08139122e+00 -5.03041863e-01 1.01028347e+00 -5.89198232e-01 -1.99990794e-01 -1.24876082e+00 -7.08224475e-01 -9.97717559e-01 -5.07462382e-01 1.03781939e+00 2.38027677e-01 -6.36613145e-02 7.21228182e-01 2.34157309e-01 -1.87270209e-01 -6.82938099e-01 -6.89298034e-01 -6.91791117e-01 1.56235602e-03 -3.16959202e-01 4.85762030e-01 1.03492296e+00 -5.45533419e-01 2.47639835e-01 -9.61332500e-01 9.33761895e-02 3.84236932e-01 1.51773095e-01 8.17001343e-01 -1.66805416e-01 -5.81301630e-01 -7.21275866e-01 -1.13851398e-01 -1.31500077e+00 1.14797406e-01 -5.40280879e-01 1.16415255e-01 -1.38450491e+00 -1.59516960e-01 9.94975329e-04 -1.20456956e-01 1.22691177e-01 -4.42156017e-01 5.15296519e-01 5.08202255e-01 1.43149808e-01 -4.27544773e-01 8.75205219e-01 1.78006911e+00 -2.16261268e-01 -7.08150789e-02 -1.73316985e-01 -5.30094624e-01 6.17096186e-01 8.63708556e-01 -1.48742432e-02 -6.82679832e-01 -7.82949805e-01 -5.15159741e-02 4.07316834e-01 4.11438704e-01 -1.06322610e+00 3.87530327e-02 -4.88528430e-01 6.27067864e-01 -3.15505922e-01 5.41720688e-01 -4.80437487e-01 6.26273572e-01 5.59867263e-01 -4.14322883e-01 2.53285497e-01 1.64564058e-01 5.73605359e-01 -3.91139090e-01 4.06086624e-01 8.21963310e-01 -2.57542193e-01 -9.15761173e-01 7.06751823e-01 -3.64230096e-01 7.38056973e-02 7.80853093e-01 -1.48155645e-01 -1.50159195e-01 -7.11366832e-01 -6.81572735e-01 8.82857069e-02 5.29348314e-01 7.77599514e-01 7.28101552e-01 -1.68728495e+00 -9.60373521e-01 -1.81921534e-02 -4.88850549e-02 -1.05092607e-01 7.12261140e-01 3.75253290e-01 -6.49192274e-01 1.82939857e-01 -4.59135413e-01 -5.47368646e-01 -8.98243785e-01 5.20037472e-01 1.32610708e-01 -5.76586612e-02 -6.95932269e-01 8.73182952e-01 5.11308014e-01 4.33127075e-01 6.38752729e-02 -3.89698505e-01 -7.71914721e-02 -1.96872711e-01 5.90638638e-01 1.09683402e-01 -5.09372592e-01 -8.91089439e-01 5.11720106e-02 6.61465228e-01 2.67072231e-01 -2.82825857e-01 1.02722120e+00 -1.48548886e-01 3.50181043e-01 3.78393978e-01 9.92501736e-01 1.00066075e-02 -2.04740667e+00 -4.48527522e-02 -4.24924016e-01 -6.58677340e-01 -3.67440045e-01 -4.48022515e-01 -1.30255187e+00 6.72959089e-01 3.63696843e-01 2.03317944e-02 1.41846311e+00 -3.28020304e-01 9.67263162e-01 -1.06730364e-01 1.58789635e-01 -8.06760013e-01 4.40010935e-01 4.80111748e-01 1.31758082e+00 -1.13650954e+00 -3.07927668e-01 -2.63079256e-01 -8.69477749e-01 9.36944723e-01 8.31715524e-01 -3.69941354e-01 1.74852416e-01 1.37940407e-01 -9.14267637e-03 3.18485856e-01 -1.04008424e+00 4.20431271e-02 4.97854918e-01 3.72374445e-01 6.31175876e-01 8.29644725e-02 -2.72751123e-01 2.31294185e-01 -2.86702514e-01 -7.37975761e-02 4.24586415e-01 6.82273090e-01 -4.30958718e-02 -9.89409924e-01 -2.62068957e-01 9.13664550e-02 -4.66243535e-01 -1.53683528e-01 1.13610946e-01 6.20202839e-01 1.61492333e-01 8.43883932e-01 8.00289586e-02 -4.49970752e-01 2.03385144e-01 -2.87454277e-01 5.24405003e-01 -3.63205850e-01 -4.90112782e-01 4.56636399e-01 4.72090207e-02 -8.56899798e-01 -5.32461822e-01 -5.80020905e-01 -9.52337146e-01 -3.91582936e-01 4.48902428e-01 -3.02841365e-01 1.30791634e-01 3.76491606e-01 4.06206727e-01 1.00138772e+00 6.68578625e-01 -1.26736856e+00 -1.60659641e-01 -8.62282872e-01 -1.04283564e-01 7.47613013e-01 3.78812611e-01 -3.46208811e-01 -2.45435491e-01 7.00078607e-01]	[10.858342170715332, -0.6657297015190125]
0542f1e5-5237-4243-8ca8-2d6d0bc4d5f5	a-probabilistic-formulation-of-unsupervised-1	2002.03912	null	https://arxiv.org/abs/2002.03912v3	https://arxiv.org/pdf/2002.03912v3.pdf	A Probabilistic Formulation of Unsupervised Text Style Transfer	We present a deep generative model for unsupervised text style transfer that unifies previously proposed non-generative techniques. Our probabilistic approach models non-parallel data from two domains as a partially observed parallel corpus. By hypothesizing a parallel latent sequence that generates each observed sequence, our model learns to transform sequences from one domain to another in a completely unsupervised fashion. In contrast with traditional generative sequence models (e.g. the HMM), our model makes few assumptions about the data it generates: it uses a recurrent language model as a prior and an encoder-decoder as a transduction distribution. While computation of marginal data likelihood is intractable in this model class, we show that amortized variational inference admits a practical surrogate. Further, by drawing connections between our variational objective and other recent unsupervised style transfer and machine translation techniques, we show how our probabilistic view can unify some known non-generative objectives such as backtranslation and adversarial loss. Finally, we demonstrate the effectiveness of our method on a wide range of unsupervised style transfer tasks, including sentiment transfer, formality transfer, word decipherment, author imitation, and related language translation. Across all style transfer tasks, our approach yields substantial gains over state-of-the-art non-generative baselines, including the state-of-the-art unsupervised machine translation techniques that our approach generalizes. Further, we conduct experiments on a standard unsupervised machine translation task and find that our unified approach matches the current state-of-the-art.	['Taylor Berg-Kirkpatrick', 'Xinyi Wang', 'Junxian He', 'Graham Neubig']	2020-02-10	null	https://openreview.net/forum?id=HJlA0C4tPS	https://openreview.net/pdf?id=HJlA0C4tPS	iclr-2020-1	['decipherment', 'unsupervised-machine-translation']	['natural-language-processing', 'natural-language-processing']	[ 7.67806888e-01 4.38904017e-01 -2.43621573e-01 -2.83220500e-01 -1.26116383e+00 -9.63697791e-01 1.09635460e+00 -6.07900262e-01 -6.93406239e-02 1.04225528e+00 4.12919551e-01 -4.56300795e-01 4.90735292e-01 -7.47365713e-01 -1.21898937e+00 -6.60350442e-01 6.96843505e-01 1.16077101e+00 -2.47152701e-01 -3.78776014e-01 6.35380670e-02 5.19982502e-02 -8.14250886e-01 2.76534796e-01 9.30241823e-01 3.29470605e-01 1.12758197e-01 5.81541121e-01 -2.82665640e-01 4.72154588e-01 -4.70784217e-01 -8.13474238e-01 1.14289440e-01 -1.00688875e+00 -1.18317056e+00 2.70742804e-01 1.65566012e-01 -3.24742049e-01 -2.47251019e-01 1.00456083e+00 2.61205226e-01 -4.29177918e-02 1.27425206e+00 -1.16646445e+00 -1.41229391e+00 5.92048883e-01 -4.27019566e-01 -2.57732242e-01 2.67572492e-01 8.88476670e-02 1.01918280e+00 -9.37333107e-01 8.75419378e-01 1.39707899e+00 4.84649092e-01 9.32400465e-01 -1.87602830e+00 -5.08659959e-01 -3.03660613e-02 -4.69751447e-01 -8.77916276e-01 -3.70191544e-01 7.63909042e-01 -5.39016426e-01 7.79247701e-01 -4.49206457e-02 3.03602427e-01 1.95404077e+00 3.68422568e-01 1.08554626e+00 1.32662094e+00 -7.35842645e-01 1.30945936e-01 2.22566932e-01 -3.54796559e-01 5.76098025e-01 -9.83868688e-02 1.37084246e-01 -5.60886800e-01 -3.38219106e-01 9.30025637e-01 -5.42908832e-02 3.36983316e-02 -3.65780950e-01 -1.29094493e+00 1.05367851e+00 -1.51214585e-01 1.10711880e-01 -2.42896210e-02 2.90318251e-01 2.92801648e-01 5.76547027e-01 7.86088705e-01 3.29739302e-01 -3.15644562e-01 -2.47060642e-01 -9.87318158e-01 2.53564477e-01 1.08593464e+00 1.40145969e+00 8.13429952e-01 5.56148887e-02 -2.98087358e-01 5.93336523e-01 3.92385602e-01 8.28947246e-01 3.49949867e-01 -8.87172163e-01 5.81250131e-01 1.75038483e-02 1.17033347e-01 -3.33175480e-01 5.29128551e-01 -1.24487795e-01 -7.43758440e-01 -1.45388216e-01 2.72075653e-01 -4.12756890e-01 -8.95007849e-01 2.09311867e+00 -1.61007971e-01 8.69814456e-02 2.01250389e-01 5.13219655e-01 1.60047725e-01 7.69460618e-01 -1.53630748e-01 -2.36583397e-01 1.15016139e+00 -9.79734123e-01 -6.39711976e-01 -2.37431750e-01 3.53117198e-01 -1.01381326e+00 1.34302318e+00 1.40258908e-01 -1.19734383e+00 -3.67036790e-01 -7.33260930e-01 -2.47613609e-01 -7.19324499e-02 -1.21306136e-01 5.28143167e-01 4.94268894e-01 -1.14576697e+00 7.30969727e-01 -1.03721094e+00 -5.71436703e-01 2.91929096e-01 2.59211630e-01 -1.24951884e-01 1.30733907e-01 -1.11711597e+00 8.31697345e-01 1.50720328e-01 -4.03693169e-01 -1.24839473e+00 -6.35322094e-01 -7.61081100e-01 -6.59141764e-02 -1.46058854e-02 -1.47822547e+00 1.60683608e+00 -1.34369874e+00 -2.29290700e+00 1.03436327e+00 -6.14463329e-01 -4.17159021e-01 6.51169181e-01 -5.42376757e-01 -8.96173641e-02 -8.76310021e-02 1.49997100e-01 8.00117671e-01 1.10526872e+00 -1.27854490e+00 -2.80337811e-01 -4.39599622e-03 -1.94931373e-01 3.29856783e-01 -3.57727289e-01 1.43836454e-01 -1.84643865e-01 -1.00295424e+00 -4.60517853e-01 -1.27643895e+00 -3.80410291e-02 -2.44603276e-01 -5.34657717e-01 -2.55735457e-01 8.31484735e-01 -5.08166432e-01 6.30302668e-01 -1.84286726e+00 9.60244775e-01 -1.75980568e-01 -5.86328916e-02 -8.48226771e-02 -1.41901016e-01 8.08072329e-01 9.07215774e-02 2.80917585e-01 -7.29910254e-01 -9.09241915e-01 3.35088432e-01 4.98599976e-01 -9.61448550e-01 1.25870913e-01 4.07193214e-01 1.35091972e+00 -9.68184888e-01 -2.39919335e-01 -2.19137534e-01 4.25876826e-01 -7.41270244e-01 4.75007921e-01 -6.59413755e-01 7.58116245e-01 -4.51539606e-01 2.38237962e-01 2.55877405e-01 -4.42501456e-01 3.08638275e-01 3.78526270e-01 3.40641141e-01 6.10487223e-01 -5.00452399e-01 2.28665209e+00 -5.30778050e-01 5.92842162e-01 -3.17096353e-01 -9.93996382e-01 8.21075022e-01 5.89018285e-01 -2.57038102e-02 3.48782167e-02 1.70525201e-02 2.32241869e-01 -4.19200838e-01 -1.17341205e-01 4.84526277e-01 -5.56733668e-01 -2.72238135e-01 1.00908029e+00 5.82926452e-01 -3.57056230e-01 -1.61813900e-01 4.07523245e-01 8.13781679e-01 8.25790286e-01 5.54987341e-02 -4.11725700e-01 1.30188495e-01 -3.48252133e-02 3.30517143e-01 7.98960686e-01 1.99878678e-01 6.76481724e-01 4.91528898e-01 1.03873737e-01 -1.41307390e+00 -1.59025252e+00 2.20857292e-01 1.12782025e+00 -2.25328296e-01 -2.98397988e-01 -1.02362752e+00 -9.11883175e-01 -2.55454630e-01 9.29959834e-01 -6.11102462e-01 -1.30056500e-01 -6.29004657e-01 -6.70226634e-01 7.75318861e-01 5.36068857e-01 1.98477820e-01 -1.12413967e+00 2.26630807e-01 1.73719153e-01 -4.47810799e-01 -9.85534668e-01 -9.88054633e-01 -7.56775495e-03 -1.13643813e+00 -3.78373057e-01 -8.40837121e-01 -1.07361233e+00 6.43542051e-01 -2.01012358e-01 1.56169975e+00 -6.00731492e-01 2.49972552e-01 5.17622411e-01 -9.21600237e-02 -4.73415762e-01 -1.14600456e+00 4.34275955e-01 1.90821484e-01 1.79788787e-02 4.77089852e-01 -1.00378454e+00 -3.41073513e-01 3.50036239e-03 -1.01295888e+00 2.33450890e-01 6.17361546e-01 1.12366188e+00 6.81748867e-01 -7.01726973e-01 5.61789215e-01 -1.29538918e+00 8.59279513e-01 -4.98096615e-01 -4.34299678e-01 2.52176583e-01 -6.17681682e-01 5.17546892e-01 7.81810522e-01 -6.26016974e-01 -1.35381365e+00 -6.83416128e-02 2.64506694e-02 -5.47835886e-01 -4.35288340e-01 1.99330255e-01 -2.60056764e-01 4.09803838e-01 6.25555336e-01 7.91680694e-01 2.51387000e-01 -4.91517335e-01 9.17317688e-01 5.86090147e-01 5.78711808e-01 -1.07623267e+00 1.27519727e+00 4.80055064e-01 -9.76665691e-02 -5.37329853e-01 -7.41558075e-01 8.67230892e-02 -7.52938747e-01 4.28362846e-01 9.97924984e-01 -1.02739263e+00 -4.03434187e-01 3.12087029e-01 -1.55941725e+00 -5.32472610e-01 -4.54385877e-01 1.53040648e-01 -1.11592746e+00 4.91339445e-01 -9.93835270e-01 -5.94460845e-01 -5.28602481e-01 -1.23649466e+00 1.50646186e+00 -1.03961341e-01 -6.38647258e-01 -1.47247088e+00 4.70717221e-01 2.29753047e-01 4.02238995e-01 1.02987744e-01 1.03520656e+00 -6.42383814e-01 -7.10982740e-01 2.96068519e-01 2.20876619e-01 4.50979114e-01 2.98653483e-01 -1.96247220e-01 -7.30991840e-01 -3.60135406e-01 3.44704390e-02 -5.35647631e-01 8.28554094e-01 7.07857031e-03 6.90648377e-01 -5.50248623e-01 -3.47821742e-01 7.13846385e-01 1.15910852e+00 -3.58813740e-02 6.74627960e-01 8.83651450e-02 8.62100720e-01 3.50939900e-01 3.69220152e-02 -9.95076969e-02 2.80737787e-01 5.63800454e-01 -7.27304742e-02 -1.35836303e-01 1.18062340e-01 -7.22971201e-01 9.14033711e-01 1.09436655e+00 -1.13020830e-01 -3.65245432e-01 -4.58542466e-01 6.54372036e-01 -1.88728464e+00 -9.59137261e-01 2.64809191e-01 1.85003102e+00 1.26781225e+00 -1.29752485e-02 2.13961810e-01 -4.66407508e-01 5.35060346e-01 6.55554309e-02 -4.72816199e-01 -6.86558008e-01 -1.16950825e-01 7.07213044e-01 3.49414647e-01 6.30518436e-01 -9.53965187e-01 1.39909339e+00 6.91263247e+00 8.24560106e-01 -9.17261124e-01 4.41997468e-01 4.84633684e-01 5.43227307e-02 -8.90887499e-01 1.74120128e-01 -7.76080132e-01 5.06426156e-01 1.07645559e+00 -3.10108900e-01 8.88022840e-01 7.01794028e-01 -3.74139845e-02 6.58598006e-01 -1.74713302e+00 6.13448918e-01 2.65081227e-01 -1.43134773e+00 4.54933703e-01 3.42271358e-01 1.13599539e+00 1.10105194e-01 3.97999167e-01 3.08165193e-01 1.02652609e+00 -1.11469769e+00 7.17052221e-01 4.45663512e-01 1.05420280e+00 -6.69739664e-01 2.23987952e-01 3.62617254e-01 -5.81059098e-01 5.09132445e-01 -1.80926636e-01 -2.01297551e-02 4.46002275e-01 3.12887609e-01 -6.49830759e-01 7.22572625e-01 1.34864807e-01 7.85040319e-01 7.75464401e-02 -1.17761999e-01 -6.90748870e-01 9.84306931e-01 -1.38091236e-01 1.94694847e-01 1.93874776e-01 -6.04318976e-01 7.96812236e-01 1.36050332e+00 3.26453149e-01 -1.22225933e-01 9.22333375e-02 1.47057462e+00 -5.83659232e-01 -9.29200947e-02 -9.44988787e-01 -2.67231643e-01 1.58151180e-01 8.50434065e-01 -3.19662392e-01 -5.01886427e-01 -4.41328406e-01 1.77119184e+00 3.65338594e-01 7.18960583e-01 -7.45605469e-01 -1.72435418e-01 8.35910320e-01 -1.98748022e-01 4.50973928e-01 -3.51586729e-01 -4.46757346e-01 -1.70670128e+00 -6.73853140e-03 -1.09444833e+00 -2.50709001e-02 -6.10249639e-01 -1.65979815e+00 6.42006218e-01 -2.27385364e-03 -1.10048759e+00 -9.26875353e-01 -4.55767751e-01 -7.57747710e-01 1.28904772e+00 -1.20317316e+00 -1.42371559e+00 4.95272696e-01 5.46418428e-01 8.58188570e-01 -4.35541928e-01 1.23244703e+00 -1.27602145e-01 -1.93395212e-01 8.16427410e-01 4.30999309e-01 2.48764545e-01 9.69223082e-01 -1.44787085e+00 1.06801248e+00 1.03800726e+00 4.42215681e-01 9.41222548e-01 7.59441793e-01 -7.02511907e-01 -1.52897453e+00 -1.23210645e+00 9.76119339e-01 -9.45661783e-01 8.38304043e-01 -8.44096780e-01 -7.53953338e-01 1.36339653e+00 7.44001508e-01 -5.04462302e-01 7.73778975e-01 -1.40593247e-02 -3.93486559e-01 4.32952225e-01 -8.05570483e-01 9.84424174e-01 1.23573768e+00 -8.80427301e-01 -8.96132290e-01 5.19536614e-01 1.01532531e+00 -3.13009292e-01 -8.18259776e-01 -3.78538184e-02 4.57846403e-01 -4.21486795e-01 8.02292824e-01 -1.06277537e+00 1.12713850e+00 1.56068485e-02 -2.19337747e-01 -1.64571011e+00 -3.13074410e-01 -1.34098125e+00 -2.43492424e-01 1.42728961e+00 5.79956651e-01 -7.36388922e-01 7.56497860e-01 2.16435492e-01 -8.72156546e-02 -4.29109871e-01 -6.07853591e-01 -9.12119925e-01 8.93963814e-01 -6.11388125e-02 4.31349844e-01 9.55137134e-01 -1.51897818e-01 9.38321650e-01 -6.95558012e-01 -1.03660479e-01 8.63819301e-01 2.52110451e-01 9.58587170e-01 -9.67424750e-01 -9.76585150e-01 -3.41986716e-01 1.73630491e-01 -1.54977429e+00 6.76258802e-01 -1.29595363e+00 5.61724929e-03 -1.36898816e+00 3.97292197e-01 1.14458822e-01 5.22174314e-02 2.62248725e-01 -1.82336912e-01 2.45563596e-01 1.35417208e-01 5.18501520e-01 -2.72010773e-01 8.62099648e-01 1.27672899e+00 -2.03665107e-01 -3.83277349e-02 -1.54449835e-01 -8.99868429e-01 5.93130887e-01 6.00519776e-01 -7.09747016e-01 -5.05194068e-01 -6.10822618e-01 -1.35885747e-02 1.53114438e-01 2.04080909e-01 -1.63558543e-01 -9.74753723e-02 -2.43502229e-01 2.14766681e-01 -1.38291597e-01 1.67876184e-01 -3.68151635e-01 9.87635851e-02 1.89351395e-01 -5.37332594e-01 4.23225621e-03 2.60206833e-02 7.36326158e-01 -5.43136075e-02 7.09096417e-02 5.94256282e-01 -1.80856884e-01 -8.28819349e-03 2.65627772e-01 -5.63017786e-01 4.06281531e-01 5.68420887e-01 1.52073860e-01 -1.29181698e-01 -5.46061635e-01 -7.31943071e-01 -1.70037910e-01 7.91961968e-01 5.91889441e-01 2.62918144e-01 -1.46422219e+00 -9.90905762e-01 8.74123424e-02 -7.39271566e-02 -1.40593782e-01 -3.05175513e-01 5.80951035e-01 -1.25630215e-01 3.26291919e-01 -5.07559665e-02 -7.20308185e-01 -6.89125597e-01 5.63324392e-01 8.14746916e-02 -6.80207074e-01 -5.22447526e-01 5.33420026e-01 7.23107040e-01 -6.00997031e-01 -2.12228328e-01 -1.49419317e-02 5.21552444e-01 -5.57451427e-01 2.48386830e-01 -6.96673468e-02 -2.59287328e-01 -4.04631674e-01 -9.94125009e-03 4.03620362e-01 -2.97272563e-01 -6.86749160e-01 1.16551983e+00 -1.84893832e-01 -2.32089505e-01 7.22196162e-01 1.27122104e+00 -7.95162469e-02 -1.33010292e+00 -4.63494480e-01 -2.44608209e-01 7.98263326e-02 -5.57727635e-01 -6.46632552e-01 -4.78790373e-01 1.02732587e+00 1.09983668e-01 -1.96223378e-01 9.11268175e-01 1.17595330e-01 1.00361085e+00 4.21519518e-01 3.86933953e-01 -8.03031504e-01 1.01547547e-01 7.42437184e-01 8.20955575e-01 -1.13513422e+00 -4.98899251e-01 -3.01592946e-01 -7.31073260e-01 8.87110770e-01 1.70262381e-01 -3.61446083e-01 4.59775001e-01 4.42264766e-01 -7.49345403e-03 2.40765214e-01 -7.79559374e-01 1.82047158e-01 2.64819711e-01 6.92319751e-01 5.97245336e-01 1.91528693e-01 -2.25025237e-01 5.23801267e-01 -4.92477983e-01 1.92612797e-01 4.60469157e-01 7.83465624e-01 2.31914800e-02 -1.72647691e+00 -2.61887163e-01 -4.14950959e-02 -7.00466573e-01 -4.41914409e-01 -5.90963244e-01 5.50270438e-01 -4.32209134e-01 6.79256380e-01 1.41525328e-01 -2.21105993e-01 -6.68517360e-03 6.06217802e-01 7.43016422e-01 -8.74124587e-01 -4.03241336e-01 2.94638842e-01 -1.18248977e-01 -1.85758263e-01 -3.17194998e-01 -7.66249359e-01 -8.20026696e-01 -3.86602312e-01 7.68933073e-02 2.25306019e-01 4.03500408e-01 1.23948574e+00 5.54446101e-01 4.71985430e-01 5.18777013e-01 -8.63489747e-01 -9.07056034e-01 -1.11106968e+00 -2.83313215e-01 6.15406156e-01 8.40822831e-02 -3.16116005e-01 -8.93034562e-02 7.94508398e-01]	[11.667075157165527, 9.588464736938477]
32f90a39-6abb-4da6-ae4c-0c022beeeb71	detecting-hands-in-egocentric-videos-towards	1709.02780	null	http://arxiv.org/abs/1709.02780v1	http://arxiv.org/pdf/1709.02780v1.pdf	Detecting Hands in Egocentric Videos: Towards Action Recognition	Recently, there has been a growing interest in analyzing human daily activities from data collected by wearable cameras. Since the hands are involved in a vast set of daily tasks, detecting hands in egocentric images is an important step towards the recognition of a variety of egocentric actions. However, besides extreme illumination changes in egocentric images, hand detection is not a trivial task because of the intrinsic large variability of hand appearance. We propose a hand detector that exploits skin modeling for fast hand proposal generation and Convolutional Neural Networks for hand recognition. We tested our method on UNIGE-HANDS dataset and we showed that the proposed approach achieves competitive hand detection results.	['Petia Radeva', 'Mariella Dimiccoli', 'Alejandro Cartas']	2017-09-08	null	null	null	null	['hand-detection']	['computer-vision']	[-3.29114497e-03 -4.82179582e-01 -1.37343913e-01 -1.65772021e-01 -8.03687871e-02 -7.34339416e-01 4.31253076e-01 -5.81285715e-01 -4.75904226e-01 5.67780018e-01 4.24800456e-01 4.26152200e-01 9.66945589e-02 -3.19530398e-01 -3.60480577e-01 -7.95983255e-01 5.29413186e-02 1.25140503e-01 2.30782166e-01 2.70321399e-01 1.38551950e-01 1.04814041e+00 -1.30686581e+00 6.72961846e-02 9.58548561e-02 8.18384647e-01 9.78685766e-02 1.07701576e+00 3.43366772e-01 6.38734877e-01 -3.85770619e-01 -3.99443835e-01 2.71467805e-01 -5.10320842e-01 -7.32255220e-01 4.59396482e-01 6.08834624e-01 -6.13260508e-01 -6.51413441e-01 9.10017252e-01 1.04530013e+00 2.83743501e-01 6.65044248e-01 -8.33414614e-01 -1.24600090e-01 1.34911641e-01 -7.57208824e-01 3.33828419e-01 6.05372012e-01 2.47340918e-01 7.10376740e-01 -8.61227095e-01 6.76433146e-01 1.13656461e+00 2.21689448e-01 6.81659877e-01 -1.02634311e+00 -5.10481954e-01 9.67936069e-02 3.56149554e-01 -1.36836541e+00 -4.82274860e-01 9.69546080e-01 -5.22548199e-01 9.13116693e-01 4.43100557e-02 9.38010812e-01 1.49879134e+00 1.58549055e-01 1.14220345e+00 1.15264177e+00 -5.64025521e-01 1.48215711e-01 4.07358743e-02 2.27249023e-02 6.10577226e-01 2.62672901e-01 -3.60836655e-01 -4.21786159e-01 1.33364037e-01 1.21849132e+00 3.62606615e-01 -3.07544619e-01 -5.86915731e-01 -1.35729110e+00 7.54474550e-02 2.89110065e-01 3.31176490e-01 -7.69296706e-01 3.77730042e-01 3.70106906e-01 -1.65627688e-01 1.29448712e-01 1.20830752e-01 -2.66295075e-02 -3.48424703e-01 -6.51696026e-01 1.02251738e-01 6.18026376e-01 9.33556736e-01 -6.17254153e-02 -1.68823004e-01 -2.76963383e-01 5.48117101e-01 -1.29260167e-01 6.67521894e-01 2.21961588e-01 -7.57113397e-01 4.02755022e-01 6.31612778e-01 3.20149302e-01 -7.05969274e-01 -4.91887778e-01 -3.72043043e-01 -8.07918787e-01 1.64893955e-01 1.01267302e+00 -1.44898459e-01 -6.76426530e-01 1.45591378e+00 3.27784687e-01 -4.81620938e-01 -4.33355391e-01 1.20121288e+00 3.23205233e-01 -1.74501881e-01 4.22470450e-01 -6.73675612e-02 1.50793338e+00 -4.96346593e-01 -7.61278749e-01 -1.66971400e-01 1.16711460e-01 -7.65583277e-01 8.88355017e-01 6.21357262e-01 -9.23479319e-01 -4.23061013e-01 -6.60057962e-01 1.87573224e-01 -1.22725315e-01 3.92316014e-01 8.03167224e-01 8.55466962e-01 -5.72632730e-01 3.76535624e-01 -7.93731153e-01 -1.12807393e+00 4.59102303e-01 3.31646949e-01 -6.21527970e-01 -1.82725452e-02 -7.04136550e-01 6.06594801e-01 3.39390010e-01 2.02940717e-01 -4.72113609e-01 1.65785789e-01 -3.66552204e-01 4.97102737e-02 4.91389811e-01 -4.61222112e-01 9.39688623e-01 -8.52612615e-01 -1.32025111e+00 9.03382182e-01 -3.06015790e-01 9.18933824e-02 8.14607441e-01 -3.48413914e-01 -2.97928452e-01 2.39175081e-01 -5.50035954e-01 2.91789412e-01 1.20342743e+00 -8.29598069e-01 -2.48919636e-01 -9.95323241e-01 3.57360579e-02 3.41577500e-01 -3.68839264e-01 2.67647088e-01 -4.92982149e-01 -6.13685369e-01 2.49668323e-02 -1.16393638e+00 2.58069903e-01 1.05311058e-01 -4.22370911e-01 -4.13473427e-01 7.60441482e-01 -8.46989453e-01 9.96760786e-01 -1.98950768e+00 3.15264612e-01 1.66980237e-01 2.63067842e-01 3.95413488e-01 1.41679510e-01 2.26894855e-01 -3.42101743e-03 -5.03816724e-01 3.04979831e-01 -1.40487969e-01 -1.69806898e-01 -1.18606977e-01 -1.54361993e-01 6.10101521e-01 -1.90036803e-01 8.50272834e-01 -1.01386786e+00 -6.75771296e-01 5.40341496e-01 6.39570594e-01 -1.79204211e-01 3.14757168e-01 1.54195100e-01 6.09707594e-01 -4.20529872e-01 9.38254297e-01 2.89802492e-01 -2.27558300e-01 2.93022394e-01 -1.95313036e-01 2.90016621e-01 -3.56691003e-01 -1.21601081e+00 1.69937670e+00 -3.14014226e-01 5.41194141e-01 -1.22348540e-01 -4.29273576e-01 4.81869131e-01 6.84398711e-01 3.30819488e-01 -3.99746597e-01 5.79197347e-01 -1.29493684e-01 1.43037349e-01 -6.59807920e-01 2.23516271e-01 2.40727067e-01 3.83461475e-01 7.06697762e-01 1.00113300e-03 1.83784723e-01 2.15171739e-01 -2.97481455e-02 9.74564135e-01 3.46339703e-01 4.87494677e-01 1.65651798e-01 4.40220743e-01 -3.93054128e-01 2.15461090e-01 4.91499543e-01 -7.12176681e-01 7.31496751e-01 3.44038606e-01 -5.95480859e-01 -9.24156129e-01 -1.20671546e+00 1.76885188e-01 1.08277106e+00 -1.83848307e-01 -3.37382294e-02 -9.22645330e-01 -9.57507014e-01 -9.12084803e-02 2.58603215e-01 -6.63262784e-01 1.94807157e-01 -5.15924513e-01 -4.08957839e-01 3.80966455e-01 1.17996693e+00 6.80155456e-01 -1.36322248e+00 -8.29455495e-01 -5.78080267e-02 -9.22675952e-02 -1.25877345e+00 -7.23106802e-01 -5.55595718e-02 -7.13980138e-01 -1.45707381e+00 -1.37678826e+00 -5.43416798e-01 7.49828279e-01 5.17168522e-01 7.96893537e-01 -4.36230779e-01 -1.01884520e+00 9.45198774e-01 -2.22723827e-01 -3.66065204e-01 1.70231700e-01 1.65663406e-01 4.17496383e-01 1.93178028e-01 7.44044602e-01 -5.46231866e-01 -1.12066960e+00 2.85414755e-01 -3.27380955e-01 -3.09635729e-01 6.82512045e-01 4.96394247e-01 2.09342957e-01 -4.95384365e-01 2.40062490e-01 -4.00207698e-01 4.75265503e-01 3.79322208e-02 -3.69408190e-01 4.38967019e-01 1.78428385e-02 -1.26447782e-01 3.47942054e-01 -5.50369620e-01 -1.24718213e+00 6.38846338e-01 1.09251633e-01 -4.47441459e-01 -4.74415481e-01 -4.54672396e-01 -3.10867310e-01 -6.63455054e-02 6.80084467e-01 2.45337591e-01 -1.23003341e-01 -3.48286301e-01 4.04479861e-01 9.67832208e-01 3.54561776e-01 -4.03351963e-01 4.80543643e-01 7.53250003e-01 1.97981615e-02 -1.41182804e+00 -5.07032871e-01 -7.60836363e-01 -1.25746608e+00 -6.59380078e-01 1.15848935e+00 -6.40656054e-01 -1.11475372e+00 7.49428689e-01 -1.35909951e+00 -2.41426453e-02 -1.27213702e-01 7.15468168e-01 -5.63910484e-01 6.29904866e-01 -2.70910114e-01 -1.21825278e+00 -5.89078069e-01 -9.33930695e-01 1.01803696e+00 2.27532163e-01 -7.16973066e-01 -8.51736248e-01 2.37821005e-02 2.63685644e-01 9.59167704e-02 9.89918411e-02 5.74234247e-01 -2.32088998e-01 -6.58183455e-01 -6.53223693e-01 -3.78774494e-01 3.48833919e-01 5.16227007e-01 -2.26151451e-01 -1.32617736e+00 -2.43583933e-01 -4.24525082e-01 -3.42916429e-01 4.75837946e-01 3.58301729e-01 1.18489492e+00 2.66639411e-01 -1.89873084e-01 3.97267669e-01 8.81608903e-01 2.13266477e-01 5.91641486e-01 -1.23295978e-01 7.79500902e-01 6.86192214e-01 2.16645047e-01 7.28097141e-01 -1.44535273e-01 7.06788957e-01 2.35782877e-01 8.43778253e-02 -3.92219722e-01 1.40716165e-01 -2.96613411e-03 1.92505196e-01 -1.05376649e+00 -4.47299406e-02 -6.99810386e-01 7.00457275e-01 -1.59975088e+00 -1.17609334e+00 2.57769488e-02 2.31966329e+00 3.44701350e-01 -2.51210898e-01 6.32378340e-01 3.27338219e-01 7.40357399e-01 -2.11578561e-03 -6.74104571e-01 -2.30152328e-02 1.56393662e-01 2.80205250e-01 3.83968204e-01 -1.78403482e-01 -1.23025894e+00 9.14456248e-01 6.42968369e+00 1.34008631e-01 -1.02407384e+00 -3.65347303e-02 1.36785032e-02 -2.29277760e-01 7.74653256e-01 -6.75452411e-01 -6.07884943e-01 2.10355178e-01 -1.25222519e-01 1.81512803e-01 4.67791200e-01 1.10235775e+00 5.27722351e-02 -2.11517274e-01 -1.19398940e+00 1.49912989e+00 2.05131888e-01 -4.03862745e-01 -1.74715549e-01 -7.48927658e-03 5.73602974e-01 -4.10689324e-01 5.34965470e-02 -2.95843244e-01 -7.40091642e-03 -7.94094920e-01 1.73479542e-01 5.45605838e-01 7.95108140e-01 -7.17882752e-01 5.17439127e-01 1.88755780e-01 -1.25020397e+00 -1.23960242e-01 -1.80242077e-01 -1.72881588e-01 -2.63275169e-02 2.80523837e-01 -8.71310711e-01 -1.40117049e-01 6.77746117e-01 5.28744698e-01 -4.30904388e-01 9.29881692e-01 -5.17406940e-01 2.05930650e-01 -1.55188590e-01 -1.78756326e-01 -3.70015472e-01 6.23413511e-02 5.75821042e-01 1.31297541e+00 -7.86549896e-02 1.84552342e-01 -2.33334586e-01 6.33176565e-01 -1.15066841e-02 2.88690049e-02 -6.98169053e-01 -3.48378390e-01 -7.19611496e-02 1.37141466e+00 -9.88034666e-01 -5.98653145e-02 -3.72560263e-01 1.54438138e+00 2.29373902e-01 4.90409672e-01 -4.03107673e-01 -6.11377418e-01 8.90533507e-01 1.69514298e-01 7.89334923e-02 -5.76156974e-01 8.54682699e-02 -1.28502917e+00 2.18488708e-01 -5.64204276e-01 2.57316113e-01 -7.47095704e-01 -1.01316118e+00 4.42130983e-01 -1.68379784e-01 -1.05578399e+00 -4.71744806e-01 -7.87972629e-01 -5.59055388e-01 8.56768548e-01 -8.57856333e-01 -1.23485291e+00 -1.14605844e+00 8.74441326e-01 7.09129989e-01 -6.20815018e-03 7.79998958e-01 1.11650996e-01 -4.98029709e-01 5.00697315e-01 -1.66538283e-01 5.43953001e-01 1.02648938e+00 -1.24652874e+00 4.78770703e-01 7.42608607e-01 7.95711577e-02 9.55999792e-01 4.64605331e-01 -6.48966253e-01 -1.74137592e+00 -7.36760318e-01 5.58686376e-01 -7.53747344e-01 3.01802844e-01 -3.87578040e-01 -4.09694582e-01 7.87119567e-01 3.64747010e-02 8.18946287e-02 4.94162172e-01 1.38546586e-01 -2.48866498e-01 -4.79940511e-02 -9.88877416e-01 6.82792246e-01 1.40688026e+00 -6.86883390e-01 -5.87680340e-01 3.82042527e-01 -5.12867033e-01 -1.90043867e-01 -5.13905466e-01 -9.71949548e-02 1.35383308e+00 -8.15860152e-01 1.00650823e+00 -6.25795066e-01 8.25001970e-02 -8.96178335e-02 -7.06720725e-02 -1.04163671e+00 -3.61490071e-01 -2.11381420e-01 -3.08613747e-01 9.67314422e-01 -5.61683476e-01 -4.64453310e-01 1.18561637e+00 6.71565950e-01 7.82747984e-01 -2.17141658e-01 -5.38598657e-01 -8.47749770e-01 -5.52938402e-01 -2.40260556e-01 9.52254012e-02 5.11505008e-01 2.44700328e-01 1.15880020e-01 -4.16067034e-01 -1.15465105e-01 1.15444779e+00 1.73540879e-02 1.04898715e+00 -1.23887622e+00 -2.66093135e-01 -1.88192248e-01 -6.89322531e-01 -8.83331895e-01 2.26214100e-02 -5.07832706e-01 -2.07234606e-01 -1.44850838e+00 6.35153234e-01 1.12258859e-01 -4.44782823e-01 1.28717795e-01 -7.42415711e-02 5.54938674e-01 2.69222796e-01 1.37640029e-01 -4.24772203e-01 1.14090502e-01 1.36835229e+00 5.32265753e-03 -2.03657195e-01 2.20537797e-01 -1.53589686e-02 8.43539894e-01 6.12634599e-01 2.64704339e-02 -1.97036475e-01 -1.50332540e-01 1.79234877e-01 -8.03623199e-02 5.43127477e-01 -1.17166913e+00 2.32410997e-01 -5.06665632e-02 8.87007833e-01 -7.03995347e-01 3.93430322e-01 -9.17275190e-01 -8.77798274e-02 6.12915456e-01 -2.00359989e-02 -3.09488505e-01 -2.40375280e-01 5.17960608e-01 2.08047405e-01 1.66963503e-01 7.94636190e-01 -4.68403965e-01 -8.77343118e-01 1.24268286e-01 -2.74765611e-01 -2.40452290e-01 1.08841860e+00 -4.71575707e-01 1.15074605e-01 -3.52628082e-01 -9.39161420e-01 -1.90789044e-01 4.40299809e-01 4.73461598e-01 4.92718160e-01 -9.47246492e-01 -2.86422312e-01 2.62464523e-01 1.46769494e-01 -3.80517930e-01 2.49969035e-01 9.98753011e-01 -3.85806799e-01 5.74929059e-01 -5.41869998e-01 -6.19094253e-01 -1.76734400e+00 4.98245955e-01 3.30430686e-01 1.37790382e-01 -5.76220632e-01 7.68305421e-01 4.05275106e-01 1.67031959e-01 5.92070341e-01 -1.35197893e-01 -8.40274692e-02 7.35066012e-02 7.29986012e-01 9.64256108e-01 -1.17266163e-01 -4.80650425e-01 -4.00055051e-01 8.01179767e-01 -4.77696136e-02 1.61134675e-01 9.84069705e-01 6.49731979e-02 1.39622673e-01 2.95744181e-01 7.72732377e-01 -9.82420444e-02 -1.30287576e+00 -1.30380124e-01 -2.82018274e-01 -7.80421734e-01 -8.80678743e-02 -7.14760303e-01 -1.04942620e+00 1.24519467e+00 9.92630482e-01 -4.70402479e-01 1.06006205e+00 1.11296371e-01 7.20250249e-01 8.16414952e-01 8.08491647e-01 -1.30745661e+00 4.43294168e-01 3.12839597e-01 8.82548511e-01 -1.34791505e+00 1.44515663e-01 -3.62940639e-01 -7.18396664e-01 1.20928597e+00 5.39551973e-01 -8.31904411e-02 3.86454374e-01 2.80087907e-02 -1.03728183e-01 -7.30081722e-02 1.42704383e-01 -4.77564394e-01 3.09388757e-01 6.55292511e-01 4.78871793e-01 1.57292902e-01 -1.24302916e-01 2.50708461e-01 3.24757606e-01 3.58836591e-01 8.68304819e-02 1.09871316e+00 -1.56688839e-01 -9.05794978e-01 -5.29733241e-01 3.42905760e-01 -4.60392445e-01 4.36433733e-01 -7.67077446e-01 8.59641373e-01 -1.39682829e-01 5.50595522e-01 -6.91238791e-02 -2.82669961e-01 4.61337954e-01 6.39379621e-01 1.32129526e+00 -4.50681686e-01 -4.01729852e-01 -1.02460742e-01 -3.34031016e-01 -7.12700069e-01 -4.25585002e-01 -7.56261826e-01 -8.34402382e-01 -1.66750774e-01 -1.38395473e-01 -5.29497147e-01 8.54975104e-01 1.01481581e+00 -1.15918834e-03 4.30249572e-01 3.98833662e-01 -1.19427180e+00 -5.40177107e-01 -1.03659964e+00 -1.15697551e+00 5.40202260e-01 3.62436324e-02 -9.21930969e-01 -1.87696382e-01 1.67629898e-01]	[6.628370761871338, -0.5933261513710022]
61ac820e-67e1-4500-95b1-81d1ad645f03	implicit-discourse-relation-classification-we	null	null	https://aclanthology.org/2020.acl-main.480	https://aclanthology.org/2020.acl-main.480.pdf	Implicit Discourse Relation Classification: We Need to Talk about Evaluation	Implicit relation classification on Penn Discourse TreeBank (PDTB) 2.0 is a common benchmark task for evaluating the understanding of discourse relations. However, the lack of consistency in preprocessing and evaluation poses challenges to fair comparison of results in the literature. In this work, we highlight these inconsistencies and propose an improved evaluation protocol. Paired with this protocol, we report strong baseline results from pretrained sentence encoders, which set the new state-of-the-art for PDTB 2.0. Furthermore, this work is the first to explore fine-grained relation classification on PDTB 3.0. We expect our work to serve as a point of comparison for future work, and also as an initiative to discuss models of larger context and possible data augmentations for downstream transferability.	['Luis Lastras', 'Song Feng', 'Chulaka Gunasekara', 'Najoung Kim']	2020-07-01	null	null	null	acl-2020-6	['implicit-discourse-relation-classification']	['natural-language-processing']	[ 1.75684929e-01 1.08789527e+00 -4.35009509e-01 -6.90568745e-01 -1.01833236e+00 -6.66643858e-01 9.27998304e-01 5.76688647e-01 -4.41901475e-01 1.11696661e+00 8.52226794e-01 -8.66314888e-01 -6.42805770e-02 -7.27003872e-01 -5.70528209e-01 -1.01892196e-01 -2.75284111e-01 7.09058940e-01 2.70925552e-01 -7.64321089e-01 7.45620281e-02 1.75055355e-01 -8.48120928e-01 9.39348221e-01 4.98416007e-01 5.40730476e-01 2.39072833e-03 7.16610491e-01 1.63764507e-02 1.51795733e+00 -9.64921653e-01 -8.53308797e-01 -3.95171732e-01 -3.12870771e-01 -1.96606910e+00 -3.86161923e-01 5.90024769e-01 -3.57999086e-01 -4.44595248e-01 5.95030606e-01 4.33303446e-01 3.46194655e-01 5.05920112e-01 -8.32898200e-01 -9.63512123e-01 1.38922739e+00 3.53161572e-03 7.11906672e-01 5.24661303e-01 2.36920118e-02 1.52368391e+00 -3.27120572e-01 1.21167016e+00 1.42198277e+00 7.08906233e-01 7.91270912e-01 -1.12120259e+00 -2.63030618e-01 3.81856710e-01 6.15006089e-01 -8.16712976e-01 -9.28795218e-01 5.27942955e-01 -3.08685184e-01 1.89024234e+00 5.65331399e-01 1.93917617e-01 1.35816443e+00 -3.05139512e-01 8.02949369e-01 1.13465965e+00 -6.44404471e-01 -3.55098456e-01 -1.49153590e-01 8.36942017e-01 4.74296659e-01 9.63470563e-02 -5.54865822e-02 -7.35698879e-01 1.99091882e-01 2.89546341e-01 -1.15479946e+00 -5.50137043e-01 2.73862481e-01 -1.08195734e+00 7.99902081e-01 5.36347687e-01 4.93172109e-01 9.44443569e-02 -6.48124516e-02 8.12629163e-01 4.47043508e-01 6.31024122e-01 7.70900905e-01 -5.49396157e-01 -6.62207842e-01 -4.56232607e-01 3.80813986e-01 1.23669744e+00 1.05710053e+00 2.05111876e-01 -4.84576434e-01 -4.12163824e-01 8.94851685e-01 3.35720003e-01 -3.39507073e-01 2.14959338e-01 -1.25854790e+00 1.14930165e+00 4.83829051e-01 -1.42265201e-01 -8.36424232e-01 -3.78175735e-01 1.98518425e-01 -3.37600142e-01 -1.61597103e-01 8.01616013e-01 -9.83007699e-02 -9.06130075e-02 1.78570211e+00 -7.39525119e-03 -1.47764996e-01 5.98917663e-01 9.22992051e-01 1.47861040e+00 5.72439253e-01 3.39063346e-01 -2.72509962e-01 1.61171818e+00 -1.19931996e+00 -1.14206922e+00 -1.63081452e-01 1.19213939e+00 -8.76569748e-01 9.74542379e-01 -7.44697377e-02 -1.41724455e+00 -4.69598264e-01 -1.15746367e+00 -7.60208964e-01 -2.54255027e-01 -1.20161310e-01 9.11888480e-01 5.74492335e-01 -1.03714371e+00 6.73436582e-01 -1.03361309e+00 -4.99980181e-01 4.56387490e-01 1.01386141e-02 -3.46055239e-01 3.03119928e-01 -1.61744642e+00 1.53631687e+00 5.88796854e-01 1.77372754e-01 -3.67555469e-01 -6.67765915e-01 -1.19751871e+00 -1.38854355e-01 3.13912630e-01 -3.51563781e-01 1.89191580e+00 -3.97629201e-01 -1.52338219e+00 1.47738779e+00 -2.50949174e-01 -9.16846871e-01 1.31988257e-01 -5.11529982e-01 -2.34037861e-01 -1.04301415e-01 1.52223453e-01 5.27297974e-01 -3.74334425e-01 -8.36101592e-01 -4.75974619e-01 -1.00730538e-01 7.01143265e-01 4.23054665e-01 -1.56610273e-02 5.39026201e-01 3.08699608e-02 -4.85739172e-01 -2.70318627e-01 -5.28924704e-01 3.33578140e-02 -6.50462508e-01 -4.96007144e-01 -9.17942822e-01 6.50959015e-01 -8.82241488e-01 1.36100209e+00 -1.82061040e+00 1.47129670e-01 -7.23520458e-01 3.03254519e-02 5.18758237e-01 -3.72425877e-02 7.11638391e-01 -2.21746475e-01 4.58942115e-01 -1.85323015e-01 -6.48656487e-01 -4.63682460e-03 3.94125879e-01 -3.55617136e-01 2.67534018e-01 8.79863977e-01 1.24682486e+00 -1.09733665e+00 -5.53682148e-01 -4.22207378e-02 1.30520687e-01 -1.00939676e-01 3.94743711e-01 -4.06482756e-01 4.59462523e-01 -2.05709010e-01 2.68591732e-01 2.01758623e-01 -6.77537769e-02 5.83202600e-01 -3.42969686e-01 -2.86524653e-01 1.82584977e+00 -4.92767006e-01 1.77847981e+00 -2.20601931e-01 1.23604250e+00 6.38819709e-02 -1.22370577e+00 7.68294573e-01 7.71811843e-01 -1.32282898e-01 -4.70888644e-01 1.65641680e-01 1.16501972e-01 5.64567447e-01 -6.15539670e-01 9.77398396e-01 -1.61922067e-01 -1.36137739e-01 4.61276740e-01 2.65586287e-01 -1.39424592e-01 4.85669345e-01 2.61905789e-01 1.17735219e+00 3.35936815e-01 5.64553440e-01 -3.96139413e-01 5.55926025e-01 2.38033265e-01 2.06658542e-01 5.52194178e-01 -3.37575674e-01 4.74447668e-01 8.36847544e-01 -3.66209924e-01 -7.52050459e-01 -7.31653154e-01 -3.88646364e-01 8.73440742e-01 -1.39697865e-01 -6.69115961e-01 -5.35797834e-01 -9.72265005e-01 -2.78146327e-01 1.04179537e+00 -7.33096123e-01 2.93595165e-01 -1.34378552e+00 -6.46945477e-01 1.16271496e+00 7.10545480e-01 4.25047785e-01 -1.32164884e+00 -3.18124354e-01 3.80895674e-01 -5.39050102e-01 -1.34317434e+00 3.92967127e-02 2.53770083e-01 -7.17671514e-01 -1.34544110e+00 -1.44539267e-01 -1.11191082e+00 6.74838852e-03 -4.71211970e-02 1.75835371e+00 2.19801128e-01 4.08827960e-01 3.02184671e-02 -7.75448859e-01 -3.71422350e-01 -8.25345993e-01 6.00152433e-01 -5.98378539e-01 -9.92424846e-01 4.24399525e-01 -1.77550822e-01 -3.18885386e-01 -1.08154438e-01 -2.76648700e-01 1.60313755e-01 -9.46551412e-02 9.54700470e-01 1.28500149e-01 -4.05849516e-01 6.93458915e-01 -1.44323516e+00 1.01632905e+00 -5.05310833e-01 -1.64078593e-01 2.21615270e-01 -2.44059801e-01 -1.43607438e-01 3.71309370e-02 7.16529489e-02 -1.50292242e+00 -7.95311391e-01 -4.58963990e-01 5.60819149e-01 -3.27417374e-01 7.25190938e-01 -1.74101368e-01 4.27084208e-01 7.65480638e-01 -7.54055977e-01 -1.43786058e-01 -5.59009790e-01 7.53012717e-01 7.42573261e-01 5.18480062e-01 -9.35408592e-01 -1.35261575e-02 -3.74287181e-02 -2.58054644e-01 -7.63945460e-01 -1.38259745e+00 -3.45055908e-01 -8.67586315e-01 1.62962914e-01 1.23306692e+00 -8.10725570e-01 -7.36296654e-01 8.01139697e-02 -1.81182802e+00 -1.06075847e+00 -3.34779471e-01 1.09231532e-01 -5.00754774e-01 2.67281681e-01 -1.26749229e+00 -6.52167857e-01 -2.82988161e-01 -1.10581791e+00 9.11373794e-01 4.10361215e-02 -1.05161202e+00 -1.38490033e+00 1.94180369e-01 8.82734597e-01 -1.14172906e-01 2.39499941e-01 6.93649590e-01 -8.73654664e-01 -5.19186080e-01 4.21890348e-01 -2.56613523e-01 3.29187244e-01 2.01283604e-01 -7.36411437e-02 -1.21093869e+00 2.01977100e-02 -5.10547943e-02 -8.34111452e-01 7.19053268e-01 1.72966033e-01 7.33406305e-01 -3.74777436e-01 -3.03143054e-01 2.67428905e-01 7.17918038e-01 1.10453866e-01 6.09404743e-01 7.60880589e-01 5.90847015e-01 9.93003130e-01 9.21840429e-01 -3.54554355e-01 9.96012747e-01 7.71692634e-01 -3.81560531e-03 5.01825586e-02 -8.08232427e-01 1.69330999e-01 2.21743599e-01 1.13510668e+00 -3.29842418e-01 -3.50551039e-01 -1.15806806e+00 8.20075095e-01 -1.83635366e+00 -9.18351471e-01 -5.78895509e-01 1.55635273e+00 1.50342655e+00 2.55678207e-01 -6.68347776e-02 1.10987216e-01 4.57945526e-01 4.96808231e-01 2.93259621e-01 -9.59227920e-01 -3.08891714e-01 7.02767313e-01 -1.35819043e-03 9.62574899e-01 -1.30929101e+00 1.27796447e+00 6.15766335e+00 4.32439059e-01 -6.95483267e-01 3.41708332e-01 7.73740828e-01 2.41616160e-01 -8.78822505e-02 2.34697163e-01 -8.19732666e-01 -5.52311651e-02 1.23810697e+00 1.58547293e-02 5.89033291e-02 3.51608336e-01 -2.56797403e-01 -6.68808743e-02 -1.66329014e+00 1.87126115e-01 -9.45956334e-02 -1.61762476e+00 -4.44460213e-01 -2.36599788e-01 3.57308358e-01 2.35286683e-01 -4.13653284e-01 6.63508117e-01 5.26305556e-01 -1.27601194e+00 5.79389274e-01 1.03136972e-01 6.50079191e-01 -3.41964871e-01 9.97980058e-01 4.70046327e-02 -9.26782489e-01 4.44552153e-01 -1.05277978e-01 -7.33837605e-01 6.21118128e-01 -6.25442564e-02 -1.14449787e+00 7.18280137e-01 5.39055645e-01 8.55486035e-01 -3.47710460e-01 3.63536656e-01 -7.69145191e-01 1.07080841e+00 1.29635066e-01 -2.56409645e-01 2.33471707e-01 2.20653027e-01 6.74765527e-01 1.67187464e+00 -3.67018789e-01 5.29059529e-01 3.61060388e-02 4.55395967e-01 -2.15044305e-01 3.13874558e-02 -5.91223240e-01 -4.39716689e-02 7.30686426e-01 9.91809487e-01 -4.34423447e-01 -4.20182705e-01 -6.21864080e-01 5.54126561e-01 1.30119622e+00 1.17026992e-01 -4.00091350e-01 -5.25279567e-02 7.88443267e-01 -2.34769240e-01 4.77476753e-02 -4.17216182e-01 -6.04771852e-01 -9.21347737e-01 -1.32754818e-01 -9.72127080e-01 7.28624344e-01 -5.14807165e-01 -1.55173230e+00 7.56156802e-01 3.10664088e-01 -4.74077195e-01 -2.72679925e-01 -8.94032419e-01 -4.49027002e-01 1.09451020e+00 -1.57074642e+00 -1.12221968e+00 1.58660524e-02 -8.65309015e-02 6.81500793e-01 5.98567501e-02 1.33892047e+00 6.68606907e-02 -4.48306203e-01 6.71401918e-01 -5.68163037e-01 5.51953375e-01 9.45068061e-01 -1.58475268e+00 9.23086882e-01 4.96709675e-01 2.64453143e-01 8.35462868e-01 6.62372470e-01 -5.76853871e-01 -7.73804545e-01 -7.64342785e-01 1.40589118e+00 -1.12434649e+00 1.05514455e+00 -2.17963979e-01 -1.23381066e+00 1.24795377e+00 1.11928594e+00 -3.88441443e-01 1.00640059e+00 1.08188879e+00 -2.44067237e-01 2.11603254e-01 -7.23887801e-01 6.47820115e-01 1.14322340e+00 -5.59004605e-01 -1.43948352e+00 3.00134957e-01 1.15672922e+00 -1.07208884e+00 -1.52259707e+00 6.08945668e-01 2.55988836e-01 -9.60057855e-01 8.22701395e-01 -8.21452022e-01 8.67240667e-01 1.52614355e-01 -1.23103857e-01 -1.20600724e+00 -1.58183575e-01 -6.05616093e-01 -2.95314848e-01 1.86615014e+00 7.91026771e-01 -5.04947901e-01 5.36395371e-01 5.11802852e-01 -6.74434364e-01 -9.31574523e-01 -1.09949732e+00 -4.98925447e-01 6.44664109e-01 -5.33564508e-01 3.85854006e-01 1.18306923e+00 7.31453300e-01 1.03294301e+00 2.20174134e-01 -9.33906212e-02 1.61978289e-01 5.32964617e-02 6.96492314e-01 -8.26494157e-01 -3.93008381e-01 -6.47701085e-01 -1.14197664e-01 -1.07580590e+00 6.48058891e-01 -1.10189879e+00 -6.88704997e-02 -1.65924072e+00 -7.80719370e-02 -6.80243850e-01 3.17618176e-02 5.11731565e-01 -5.45273304e-01 7.33488649e-02 3.54578346e-01 5.60436361e-02 -5.50424516e-01 4.64146495e-01 1.33671546e+00 -9.82194319e-02 -1.48842424e-01 -1.78114429e-01 -8.70860338e-01 5.44168949e-01 1.33264256e+00 -2.26665422e-01 -4.31486070e-01 -8.49189639e-01 -1.43019065e-01 1.24682441e-01 -3.19525935e-02 -4.24451202e-01 -1.14547849e-01 -4.11393270e-02 -2.41892010e-01 -4.69590098e-01 5.28928220e-01 -9.87584591e-02 -5.70089877e-01 5.86227849e-02 -8.46573353e-01 -6.35745078e-02 5.62614024e-01 -4.79713753e-02 -4.58921283e-01 -5.85417569e-01 1.20047152e-01 -2.75423497e-01 -5.20554364e-01 -4.36815858e-01 -5.15693963e-01 6.28042340e-01 6.36673570e-01 2.21994460e-01 -1.09242499e+00 -3.48893911e-01 -5.37113607e-01 3.12938750e-01 1.41742349e-01 4.07980949e-01 1.29163697e-01 -1.17139530e+00 -1.16140759e+00 -4.64996576e-01 9.58512723e-03 4.45001841e-01 -2.13121012e-01 6.63414598e-01 -4.91247028e-01 7.27643073e-01 6.56634420e-02 -1.84616253e-01 -1.61059129e+00 1.91443697e-01 1.01090014e-01 -7.71903932e-01 -5.87369502e-01 1.20722055e+00 -1.33795306e-01 -4.94800389e-01 1.46665081e-01 -5.11973262e-01 -6.20659888e-01 1.90629914e-01 4.71441925e-01 3.10004622e-01 2.15361878e-01 -7.91114748e-01 -5.04174292e-01 -2.44670123e-01 -2.32744947e-01 -3.56645346e-01 1.31305099e+00 -1.48651227e-01 -4.92086440e-01 8.25045824e-01 8.68260145e-01 1.44436941e-01 -7.97499180e-01 -1.10942051e-01 6.53673589e-01 4.62418050e-02 -4.18370456e-01 -8.87773991e-01 -3.80849510e-01 8.47056329e-01 -2.20911860e-01 6.01950228e-01 7.84704030e-01 3.63440961e-01 6.45979166e-01 5.74246109e-01 9.03844237e-02 -8.40137005e-01 -1.84054464e-01 1.22242808e+00 1.18303061e+00 -1.20304585e+00 1.80855498e-01 -9.24624085e-01 -6.64339960e-01 1.00526965e+00 9.53125119e-01 -1.15801364e-01 4.05045986e-01 5.43945909e-01 4.48868185e-01 -4.33179259e-01 -1.07247412e+00 -3.13141108e-01 3.20438564e-01 8.59445274e-01 1.65562201e+00 2.14714989e-01 -8.45322490e-01 7.26936162e-01 -8.82911801e-01 -3.53583753e-01 3.86424690e-01 9.22361851e-01 6.54921830e-02 -1.76292050e+00 -6.19580746e-02 8.66576955e-02 -6.14509881e-01 -4.92792875e-01 -6.78365767e-01 1.19526505e+00 -1.33559585e-01 1.34574795e+00 3.02692771e-01 -9.15975422e-02 6.20394766e-01 2.50461608e-01 9.26455140e-01 -1.04909515e+00 -9.46809411e-01 -4.26738679e-01 1.59671187e+00 -3.98633182e-01 -9.21673536e-01 -9.47624147e-01 -1.25041103e+00 -1.51286677e-01 -2.79853284e-01 2.97542840e-01 -4.59234667e-04 9.38843548e-01 -8.77383817e-03 9.24316883e-01 -2.24956781e-01 -5.26616633e-01 -2.73828506e-01 -1.63618636e+00 1.45324752e-01 5.71135581e-01 1.78247154e-01 -5.07556260e-01 1.86122462e-01 2.33028620e-01]	[10.851173400878906, 9.353347778320312]
96a724e2-839a-4cf0-bf4a-c1349795a137	high-order-recurrent-neural-networks-for	1802.08314	null	http://arxiv.org/abs/1802.08314v1	http://arxiv.org/pdf/1802.08314v1.pdf	High Order Recurrent Neural Networks for Acoustic Modelling	Vanishing long-term gradients are a major issue in training standard recurrent neural networks (RNNs), which can be alleviated by long short-term memory (LSTM) models with memory cells. However, the extra parameters associated with the memory cells mean an LSTM layer has four times as many parameters as an RNN with the same hidden vector size. This paper addresses the vanishing gradient problem using a high order RNN (HORNN) which has additional connections from multiple previous time steps. Speech recognition experiments using British English multi-genre broadcast (MGB3) data showed that the proposed HORNN architectures for rectified linear unit and sigmoid activation functions reduced word error rates (WER) by 4.2% and 6.3% over the corresponding RNNs, and gave similar WERs to a (projected) LSTM while using only 20%--50% of the recurrent layer parameters and computation.	['Philip Woodland', 'Chao Zhang']	2018-02-22	null	null	null	null	['acoustic-modelling']	['speech']	[ 2.49497369e-01 5.15289128e-01 1.35042280e-01 -2.73084849e-01 -5.33144295e-01 -5.71155921e-02 4.53234851e-01 -5.33684611e-01 -8.75092566e-01 1.01260245e+00 4.20832217e-01 -7.44098723e-01 4.31194812e-01 -5.89891076e-01 -7.49370337e-01 -7.96889007e-01 2.05704883e-01 9.65710953e-02 2.06636295e-01 -5.28320730e-01 3.20832580e-01 4.52767491e-01 -1.16538596e+00 4.84545290e-01 5.25176406e-01 8.24323356e-01 5.87892830e-01 8.82247031e-01 -3.01552713e-01 1.30817974e+00 -9.21916187e-01 5.29424995e-02 -2.65423477e-01 -2.82210261e-01 -9.89841521e-01 -3.28506976e-01 1.48783013e-01 -7.66433999e-02 -6.26363635e-01 6.69628680e-01 7.12657511e-01 6.69462681e-01 1.90485954e-01 -3.17691624e-01 -6.85901463e-01 9.65427220e-01 -1.38694778e-01 4.00927573e-01 -2.55412981e-02 -2.74393231e-01 6.80779219e-01 -1.14477313e+00 2.45056733e-01 1.23685038e+00 8.24581146e-01 1.03423309e+00 -8.51040244e-01 -7.31781840e-01 2.16057479e-01 7.02709379e-03 -1.23143041e+00 -7.95495450e-01 1.51971832e-01 2.32741445e-01 2.20652890e+00 3.19529891e-01 4.34156418e-01 1.14370954e+00 2.89371669e-01 6.11771047e-01 6.61658764e-01 -5.97280204e-01 -2.85148442e-01 1.48924038e-01 3.89418066e-01 7.19511688e-01 -1.98071972e-01 2.32305136e-02 -5.42591035e-01 2.77479231e-01 8.23941886e-01 -9.21759531e-02 -3.27897012e-01 6.71235681e-01 -9.64383006e-01 7.74506629e-01 6.49264216e-01 7.19726205e-01 -4.48717445e-01 3.42714787e-01 4.85483378e-01 5.68298817e-01 7.14721859e-01 3.41398865e-01 -5.45516312e-01 -3.81923169e-01 -8.60733628e-01 -5.23510098e-01 7.83387303e-01 8.92879009e-01 4.86661762e-01 1.00028741e+00 1.15009025e-02 1.40421474e+00 1.07851595e-01 5.36288798e-01 1.20713472e+00 -3.98909450e-01 6.66283369e-01 2.59390146e-01 -2.03123823e-01 -7.17224777e-01 -4.57693636e-01 -5.83288968e-01 -1.11877942e+00 -1.89527646e-01 2.68976726e-02 -3.62285912e-01 -1.39946342e+00 1.59376514e+00 -5.16235888e-01 1.59814522e-01 4.74567592e-01 5.88121831e-01 8.49093556e-01 1.55026555e+00 -1.11042298e-01 -4.34364736e-01 9.01213288e-01 -1.32652271e+00 -1.03842759e+00 -6.83172584e-01 9.94441688e-01 -7.88591385e-01 9.58308756e-01 1.27627730e-01 -1.30934477e+00 -5.59544504e-01 -1.07266247e+00 -2.55469978e-01 -5.80866814e-01 1.57726943e-01 2.01492354e-01 6.41722918e-01 -1.45813155e+00 7.77701795e-01 -7.73583710e-01 -2.87778735e-01 -3.32832903e-01 7.04847157e-01 -1.03198387e-01 3.43369842e-01 -1.56680012e+00 1.36645615e+00 5.00781178e-01 7.87350118e-01 -6.15352929e-01 -4.74279910e-01 -8.05602372e-01 1.47203177e-01 -1.58479080e-01 -4.21190374e-02 1.33981502e+00 -8.33988309e-01 -2.11267161e+00 4.40103263e-01 -5.65534532e-01 -7.48811126e-01 -6.24431334e-02 -3.76975656e-01 -7.16203988e-01 -4.04779464e-01 -6.32375300e-01 5.74167669e-01 6.25213861e-01 -4.80437040e-01 -5.16276240e-01 5.50679257e-03 -4.42660064e-01 1.95673272e-01 -4.67005700e-01 1.65706947e-01 1.33368745e-01 -7.05933690e-01 3.84496540e-01 -1.03426945e+00 -2.06445724e-01 -9.11344230e-01 -1.56611174e-01 -2.26298481e-01 8.86119783e-01 -8.63689542e-01 1.51830029e+00 -2.07387471e+00 1.67898133e-01 7.32870027e-02 -4.06899005e-01 7.23390996e-01 -1.91214815e-01 1.47152692e-01 -2.47566074e-01 1.72047421e-01 5.67366928e-02 -3.53300363e-01 -2.95911789e-01 6.06403112e-01 -6.75121784e-01 2.27516174e-01 3.90782225e-04 8.75891089e-01 -6.92472279e-01 1.16971865e-01 1.41297936e-01 9.08579469e-01 -2.46574916e-02 -1.40940100e-02 3.14648569e-01 -2.19408441e-02 3.68315816e-01 1.67584538e-01 9.07881260e-02 1.05975859e-01 6.16235584e-02 1.26515254e-01 -3.95256937e-01 9.72827315e-01 -7.93573678e-01 1.34318554e+00 -1.03456926e+00 1.26069248e+00 -2.03894109e-01 -8.62621903e-01 1.25490785e+00 8.97346318e-01 -3.67338181e-01 -1.08708441e+00 5.75405620e-02 7.70049214e-01 4.96396236e-02 -3.04698553e-02 8.01800787e-01 -3.32053632e-01 1.96984857e-01 2.69017041e-01 2.94456661e-01 2.46392623e-01 -2.24893987e-01 -2.62213200e-01 8.38639140e-01 -4.17668551e-01 -1.99447677e-01 -1.61821276e-01 5.58487654e-01 -6.41663730e-01 7.54245877e-01 6.85636878e-01 3.51803184e-01 5.63256681e-01 1.31887585e-01 -6.84180081e-01 -1.06897247e+00 -4.93000746e-01 1.76043864e-02 1.35884786e+00 -5.50192416e-01 -1.48480877e-01 -6.05819881e-01 1.80736646e-01 -5.82998931e-01 1.02674079e+00 -3.27821255e-01 -1.29321977e-01 -1.24402046e+00 -9.39900517e-01 1.05094302e+00 7.08337903e-01 5.96804738e-01 -1.53198111e+00 -4.27442968e-01 5.97657561e-01 -1.96056798e-01 -1.00886428e+00 -4.67719853e-01 7.17279077e-01 -1.25054765e+00 -1.86538339e-01 -1.10441434e+00 -1.05379212e+00 6.48055971e-01 -9.35150757e-02 8.09988320e-01 -4.21325155e-02 4.51251268e-02 -4.08099413e-01 -2.32117385e-01 -1.09826237e-01 -5.52878201e-01 6.10301495e-01 2.09688172e-01 -2.41977915e-01 2.37428039e-01 -4.42367107e-01 -1.06964119e-01 2.71850884e-01 -4.26025033e-01 1.04409501e-01 7.27972388e-01 1.18046510e+00 3.11753601e-01 -3.44254553e-01 6.92112863e-01 -7.94890523e-01 7.71397412e-01 9.81174293e-04 -5.04305363e-01 2.63341516e-01 -7.21367061e-01 4.04800177e-01 6.08320653e-01 -5.39609671e-01 -1.36645567e+00 -2.57968098e-01 -4.24581140e-01 -1.11509888e-02 2.28018910e-01 4.44473058e-01 3.79811645e-01 -1.36818644e-02 5.02478004e-01 3.86437654e-01 -3.45734626e-01 -5.74245691e-01 1.97865337e-01 8.37478876e-01 2.83639818e-01 4.06356379e-02 1.00908674e-01 -2.35227495e-01 -5.13043284e-01 -1.24828637e+00 -7.10533321e-01 -2.41494015e-01 -4.88828182e-01 2.64102574e-02 6.80212915e-01 -7.82245398e-01 -5.58051944e-01 7.55059302e-01 -1.60911405e+00 -6.31810188e-01 -2.94485509e-01 7.09275424e-01 -2.60313272e-01 -1.92145497e-01 -1.37167585e+00 -1.00014794e+00 -7.70078897e-01 -9.56930816e-01 2.75990009e-01 1.76890850e-01 -2.45452717e-01 -1.15655077e+00 -1.05515823e-01 1.47269806e-02 1.01649582e+00 -5.16310573e-01 8.44393373e-01 -4.96060193e-01 -4.64447215e-02 -1.26329064e-01 -3.86118926e-02 7.75404334e-01 -1.07358657e-01 -3.64111215e-01 -1.35465121e+00 -3.38437408e-01 2.14718387e-01 -1.73099086e-01 1.28832102e+00 4.27425146e-01 6.92006648e-01 -5.65575540e-01 -8.67730603e-02 5.86807072e-01 1.16556644e+00 5.75764179e-01 1.11197937e+00 3.08916837e-01 7.12711811e-01 2.98324943e-01 2.05975380e-02 -2.05848329e-02 -2.28441224e-01 2.16738418e-01 -2.40855888e-02 -2.59320199e-01 -1.57744601e-01 -1.85582384e-01 8.96895111e-01 1.89870799e+00 -3.49500269e-01 -3.25006545e-01 -8.41862857e-01 6.36509657e-01 -1.61599016e+00 -1.12277508e+00 -2.08463803e-01 1.99051762e+00 9.82082129e-01 4.03041691e-01 -4.07453537e-01 3.26282203e-01 8.75533402e-01 5.07600665e-01 -3.60066772e-01 -1.42554915e+00 -3.56042743e-01 4.43149388e-01 8.59882116e-01 9.79200721e-01 -5.28674245e-01 1.27321231e+00 6.92855644e+00 8.45467091e-01 -1.48109114e+00 1.65224388e-01 6.14376903e-01 -3.06169510e-01 -6.56666905e-02 -1.95938066e-01 -1.19204700e+00 1.32367253e-01 2.01606202e+00 2.64419556e-01 5.24627447e-01 5.70766687e-01 2.57434607e-01 7.24725276e-02 -6.65635407e-01 9.40167964e-01 5.55705093e-02 -1.17656589e+00 1.10897273e-01 5.27049322e-03 7.05686510e-01 6.11061394e-01 2.39047959e-01 6.74737394e-01 4.22744043e-02 -1.38384032e+00 3.77204448e-01 5.60048461e-01 7.55776167e-01 -1.04923201e+00 9.00132775e-01 2.69711852e-01 -1.02527761e+00 -2.37137914e-01 -9.14535582e-01 -3.75729859e-01 1.10158436e-01 5.05595267e-01 -1.01425707e+00 -1.60190031e-01 5.34066260e-01 2.93427557e-01 -1.34808511e-01 5.24351060e-01 -1.97002396e-01 8.81326735e-01 -3.94617945e-01 -3.89820576e-01 7.28569388e-01 -3.26282866e-02 1.96769178e-01 1.58679926e+00 5.31783879e-01 5.31242378e-02 -6.95118010e-01 4.85269338e-01 -3.81436020e-01 -6.47127405e-02 -3.86334836e-01 -7.39449114e-02 3.10478568e-01 8.45237672e-01 -3.86702836e-01 -3.15196007e-01 -2.96949953e-01 1.27479160e+00 3.16006511e-01 7.04910278e-01 -7.00870633e-01 -9.52439606e-01 2.15769827e-01 2.49713995e-02 3.41356188e-01 -1.25518784e-01 -2.21890509e-01 -8.85321558e-01 -4.84385416e-02 -4.92061168e-01 -7.51833916e-02 -8.26433778e-01 -6.68319523e-01 1.25586212e+00 -6.34167254e-01 -6.72308922e-01 -5.90681732e-01 -7.51731098e-01 -4.97534454e-01 1.29089510e+00 -1.72184038e+00 -8.26459229e-01 2.57661134e-01 4.37306792e-01 8.56857896e-01 -3.48982364e-01 1.24334836e+00 3.90707731e-01 -8.01026225e-01 7.10503817e-01 3.07930738e-01 2.37694934e-01 1.86711252e-01 -1.02522612e+00 6.80336356e-01 6.33964837e-01 -3.30744162e-02 7.77855635e-01 5.16784012e-01 -2.00188801e-01 -9.88709629e-01 -1.02168977e+00 1.76923418e+00 1.82661191e-01 5.37774563e-01 -3.99608225e-01 -1.14963400e+00 9.78822589e-01 4.31301624e-01 -2.21952666e-02 3.41937214e-01 2.42675290e-01 -1.04274876e-01 -2.86933798e-02 -6.78811908e-01 5.63006520e-01 8.19407046e-01 -8.45984697e-01 -7.07828283e-01 1.65458396e-02 9.07840788e-01 -4.47130531e-01 -5.40865600e-01 3.18451792e-01 5.45727611e-01 -6.73720896e-01 5.96552968e-01 -3.52115542e-01 -1.84534296e-01 1.31263286e-01 -2.08841443e-01 -1.49565709e+00 -6.87613562e-02 -7.05545604e-01 -4.30427790e-02 9.84403789e-01 1.08937836e+00 -9.22963381e-01 5.50170839e-01 5.59437573e-01 -4.86793399e-01 -6.77892327e-01 -1.13241971e+00 -8.01149130e-01 1.37018830e-01 -3.81576300e-01 6.46928921e-02 6.05266571e-01 1.41198859e-01 6.07145131e-01 -7.03743041e-01 -1.45267382e-01 -2.12314010e-01 -6.14355326e-01 -1.23661228e-01 -9.47182178e-01 8.16599578e-02 -4.13491249e-01 -1.41531840e-01 -1.43379676e+00 3.51546407e-01 -7.31405377e-01 3.15410554e-01 -1.45625031e+00 -3.52762371e-01 -3.07356298e-01 -6.74650013e-01 8.00178885e-01 2.91695446e-01 1.46016806e-01 -4.50040363e-02 -4.26687077e-02 -1.34812951e-01 4.92388517e-01 7.62482166e-01 -5.82255535e-02 -4.90927547e-01 1.40225142e-01 -3.13702524e-02 8.07686031e-01 9.34965789e-01 -5.83442748e-01 -3.48601997e-01 -8.63436162e-01 3.09297085e-01 3.37017000e-01 -1.05823889e-01 -9.70950544e-01 6.78988576e-01 3.86608958e-01 3.33750159e-01 -6.64147258e-01 6.76194131e-01 -3.56012046e-01 -7.27370530e-02 6.95681572e-01 -6.02129817e-01 4.47969228e-01 4.69493628e-01 1.31709725e-01 -4.81338054e-01 -4.69779462e-01 7.87392914e-01 -3.23987722e-01 -6.35909617e-01 -3.45921338e-01 -1.02560735e+00 -3.47548991e-01 2.89460212e-01 -4.01093692e-01 -2.26364717e-01 -4.13863361e-01 -8.09412360e-01 -1.13231473e-01 -3.57305259e-01 4.02123243e-01 1.01350868e+00 -1.16374052e+00 -6.06966436e-01 5.18058360e-01 -6.21515214e-01 -1.39040276e-01 8.58540013e-02 6.82475805e-01 -5.85095882e-01 1.01014400e+00 7.74260843e-03 -2.96877492e-02 -1.19949841e+00 -1.70376468e-02 8.04283500e-01 -2.98389733e-01 -7.06155479e-01 1.30989969e+00 -2.19445303e-01 -5.53577542e-01 6.22635365e-01 -5.93502939e-01 -2.48683125e-01 2.10633472e-01 5.89226246e-01 6.79269254e-01 5.57907820e-01 -6.72811210e-01 -1.74342260e-01 3.60065371e-01 -3.15617770e-01 -4.03906345e-01 1.44871891e+00 -1.87180694e-02 -1.70791879e-01 1.04737103e+00 1.28920376e+00 -3.80084932e-01 -5.06209195e-01 -2.66795874e-01 3.58278424e-01 3.16404998e-01 5.04003763e-01 -7.25989759e-01 -1.16858983e+00 1.23650277e+00 5.98560810e-01 5.59772402e-02 8.45678389e-01 -6.09497964e-01 1.24789464e+00 9.60990608e-01 1.04333475e-01 -1.54747772e+00 -2.68988401e-01 1.29747069e+00 9.31328714e-01 -7.37041891e-01 -5.11951089e-01 3.25252503e-01 -3.93945575e-01 1.46164894e+00 3.69593203e-01 -2.31775433e-01 6.17628813e-01 2.93238819e-01 1.82453141e-01 1.54808328e-01 -1.12694108e+00 1.56974137e-01 -3.01238839e-02 -3.39294150e-02 9.00014520e-01 -1.40218195e-02 -2.74168313e-01 1.36909440e-01 -2.32509941e-01 -3.07959676e-01 6.49884999e-01 7.16713428e-01 -8.44531655e-01 -9.19481754e-01 -2.90416956e-01 1.83822900e-01 -6.53682947e-01 -5.59054971e-01 9.85947438e-03 4.18380231e-01 -4.47041571e-01 1.05753136e+00 3.17892551e-01 -4.34698850e-01 4.13674176e-01 4.44252342e-01 1.33402258e-01 -6.08708501e-01 -9.70326543e-01 2.27941260e-01 3.02039355e-01 -3.04414153e-01 -1.34317115e-01 -1.21236071e-01 -1.60536814e+00 -4.60394442e-01 -7.47033536e-01 2.58244902e-01 9.79446888e-01 9.01496172e-01 1.46196991e-01 9.58436310e-01 3.92399967e-01 -6.48890972e-01 -4.94260728e-01 -1.49738145e+00 -6.85474694e-01 -2.38195732e-01 5.38377643e-01 1.31395860e-02 -5.58782220e-01 -2.32293904e-01]	[10.865675926208496, 6.345398426055908]
02bbc0b7-61b5-4724-9a3f-b05aaf01c837	word-alignment-in-the-era-of-deep-learning-a	2212.00138	null	https://arxiv.org/abs/2212.00138v1	https://arxiv.org/pdf/2212.00138v1.pdf	Word Alignment in the Era of Deep Learning: A Tutorial	The word alignment task, despite its prominence in the era of statistical machine translation (SMT), is niche and under-explored today. In this two-part tutorial, we argue for the continued relevance for word alignment. The first part provides a historical background to word alignment as a core component of the traditional SMT pipeline. We zero-in on GIZA++, an unsupervised, statistical word aligner with surprising longevity. Jumping forward to the era of neural machine translation (NMT), we show how insights from word alignment inspired the attention mechanism fundamental to present-day NMT. The second part shifts to a survey approach. We cover neural word aligners, showing the slow but steady progress towards surpassing GIZA++ performance. Finally, we cover the present-day applications of word alignment, from cross-lingual annotation projection, to improving translation.	['Bryan Li']	2022-11-30	null	null	null	null	['nmt', 'word-alignment']	['computer-code', 'natural-language-processing']	[ 7.16499567e-01 2.76788712e-01 -3.40117484e-01 -3.16072375e-01 -1.24105024e+00 -5.73566496e-01 6.41322792e-01 -1.20019667e-01 -5.62409163e-01 8.06922436e-01 4.36907440e-01 -1.06911898e+00 2.44549915e-01 -3.04830492e-01 -6.61419034e-01 -5.03507733e-01 1.41528487e-01 9.16639805e-01 -7.22143769e-01 -7.63902783e-01 3.30873638e-01 5.15087973e-03 -7.81062484e-01 2.97749460e-01 8.09247255e-01 6.93387538e-02 9.35700163e-02 6.25388324e-01 -5.40965319e-01 -3.25021967e-02 -4.97751296e-01 -8.01810265e-01 3.52636725e-01 -6.55221283e-01 -1.18787420e+00 -4.45166558e-01 9.56993341e-01 1.09489083e-01 2.96946280e-02 9.96944547e-01 7.24810421e-01 -1.97536066e-01 2.86989123e-01 -1.02200675e+00 -1.05904651e+00 9.03932512e-01 -5.10010064e-01 5.24255574e-01 2.20205039e-01 1.99529096e-01 1.53603804e+00 -1.06376088e+00 7.36192763e-01 1.04790783e+00 1.07804847e+00 7.26292551e-01 -1.13489187e+00 -5.09236455e-01 3.05174023e-01 1.30070910e-01 -9.36855912e-01 -6.52741730e-01 1.47620499e-01 -3.94280314e-01 1.86729336e+00 3.96825045e-01 6.66873395e-01 1.24866605e+00 4.65885192e-01 8.01800489e-01 9.92565036e-01 -8.53811204e-01 -4.02836382e-01 -5.04294336e-01 2.92177856e-01 2.56519198e-01 2.29864806e-01 4.13469709e-02 -8.77364039e-01 -4.26379815e-02 4.46623862e-01 -6.80370092e-01 7.56497979e-02 6.82717189e-02 -1.79825187e+00 6.72480702e-01 1.00812882e-01 6.16286755e-01 -3.31579745e-01 1.97337896e-01 5.87239802e-01 6.33268476e-01 8.29152644e-01 9.98394430e-01 -5.60147643e-01 -4.43279803e-01 -1.11720085e+00 1.71868071e-01 4.94296253e-01 1.00437486e+00 4.44944024e-01 3.53428215e-01 -1.82234451e-01 7.49335945e-01 -6.86846748e-02 6.14023507e-01 7.47291207e-01 -5.46679139e-01 7.45728552e-01 -3.65472361e-02 -1.36154115e-01 -4.98145938e-01 -2.29423732e-01 -6.76464677e-01 -6.15318775e-01 -2.49422699e-01 2.79643327e-01 -2.23314106e-01 -7.92098939e-01 1.95289636e+00 -1.20407596e-01 -9.32114497e-02 -1.38515723e-03 7.99114943e-01 5.38755119e-01 6.23848438e-01 9.60704610e-02 -2.49760523e-01 1.26160395e+00 -1.28761303e+00 -8.00856769e-01 -7.82683969e-01 9.18288112e-01 -1.28605545e+00 1.07708180e+00 1.98215678e-01 -1.44970226e+00 -4.36326504e-01 -1.13002217e+00 -3.72308046e-01 -4.80603367e-01 -1.93606824e-01 7.09495783e-01 4.08036202e-01 -1.47013021e+00 7.18727410e-01 -9.45165992e-01 -9.12496507e-01 1.12031586e-01 5.02414405e-01 -4.94511425e-01 1.93175256e-01 -1.36680353e+00 1.72933400e+00 7.53244981e-02 1.81725815e-01 -1.70201913e-01 -3.99522096e-01 -6.62537813e-01 -3.14786613e-01 -4.38008346e-02 -9.12038386e-01 1.71377885e+00 -1.15500486e+00 -1.54484081e+00 1.35117245e+00 -5.93177259e-01 -5.68567276e-01 2.52108306e-01 -6.00743532e-01 -4.10229176e-01 -5.75732410e-01 2.68263876e-01 7.66837716e-01 2.92819500e-01 -6.77390695e-01 -5.65944374e-01 -2.11070701e-01 -5.39140940e-01 4.95411485e-01 -1.78155467e-01 6.07227445e-01 -1.11105658e-01 -6.27225280e-01 1.49110302e-01 -1.01867974e+00 -4.21805739e-01 -6.07188344e-01 -3.65647554e-01 -2.34510556e-01 6.78249747e-02 -1.02872670e+00 1.34584606e+00 -1.59471953e+00 5.84594131e-01 -2.04376474e-01 5.87482713e-02 3.08615595e-01 -5.70653200e-01 9.93093669e-01 -5.22959650e-01 7.80676827e-02 -5.26109636e-01 -5.56399167e-01 1.18691735e-01 2.51793385e-01 -5.72365999e-01 3.51342797e-01 4.23846006e-01 1.67447376e+00 -1.00803435e+00 -2.53112227e-01 -7.75038898e-02 1.82095394e-01 -2.30039179e-01 -1.29874036e-01 -5.78483343e-02 4.22977120e-01 -2.38118675e-02 6.20022416e-01 3.32718521e-01 -2.77722981e-02 3.74112725e-01 1.42380416e-01 -6.22325182e-01 9.89916742e-01 -2.30991304e-01 2.23950720e+00 -4.92550761e-01 1.05707729e+00 1.09153360e-01 -9.99043524e-01 8.91598940e-01 4.60057795e-01 4.27938402e-01 -8.46980393e-01 1.37855127e-01 8.39803457e-01 5.56701064e-01 -3.02772105e-01 9.07844663e-01 -4.92413566e-02 3.80618195e-03 7.53741860e-01 2.97664076e-01 -2.58572012e-01 1.53940812e-01 -5.82775287e-02 8.26483965e-01 7.70870984e-01 6.16044283e-01 -7.39172935e-01 1.42310604e-01 5.55783451e-01 2.53699720e-01 4.94147450e-01 4.69514355e-02 5.97559869e-01 -6.44339174e-02 -8.57947409e-01 -1.56714845e+00 -7.75752962e-01 -1.10548176e-01 1.16822910e+00 -4.52012360e-01 -5.60724318e-01 -9.57877457e-01 -3.07179481e-01 -2.99635857e-01 7.90182352e-01 -5.06128252e-01 1.03351466e-01 -1.37583959e+00 -1.14085424e+00 6.88889027e-01 4.42290336e-01 -7.02892616e-02 -1.19445705e+00 -3.40763211e-01 3.42562705e-01 -5.92887580e-01 -9.63505805e-01 -6.38129950e-01 5.13832569e-01 -1.05254364e+00 -3.90847683e-01 -7.33895123e-01 -1.01533544e+00 3.57675374e-01 1.80715367e-01 1.80851769e+00 6.25482574e-02 3.79479080e-02 9.49921161e-02 -7.30014220e-02 -4.93227929e-01 -6.63206339e-01 9.05083060e-01 3.48164469e-01 -6.45117640e-01 9.99527931e-01 -8.80412042e-01 -3.01405251e-01 1.43731600e-02 -3.98055851e-01 3.87478918e-01 8.41220021e-01 8.35599124e-01 3.39710146e-01 -1.36526394e+00 4.85611588e-01 -8.95853817e-01 8.41511905e-01 -3.90771180e-01 -1.87541515e-01 1.36312753e-01 -8.90657961e-01 -2.67631263e-01 2.57168144e-01 -2.23585516e-01 -3.67369682e-01 -3.59025985e-01 -3.95594150e-01 2.94283062e-01 1.27069458e-01 5.74369371e-01 5.34299836e-02 2.69960761e-01 7.82305837e-01 2.67684072e-01 3.04350644e-01 -3.61792892e-01 6.80596173e-01 6.85546994e-01 8.64473462e-01 -7.08868384e-01 9.55112517e-01 6.57881238e-03 -2.21463218e-01 -4.05746281e-01 -5.87685525e-01 -2.04753831e-01 -1.11436307e+00 2.20698655e-01 8.99867952e-01 -7.13949203e-01 -1.58928216e-01 2.29506671e-01 -1.59701455e+00 -5.33982217e-01 -2.15683535e-01 3.49243343e-01 -7.02085614e-01 3.38218033e-01 -7.25188732e-01 -3.07681888e-01 -1.06805694e+00 -1.19616723e+00 1.03971481e+00 -8.82264599e-02 -1.09383380e+00 -1.10006404e+00 5.42556286e-01 3.14714760e-01 8.40373099e-01 -1.79802150e-01 8.92017663e-01 -1.10797203e+00 -2.84314036e-01 -1.91139936e-01 1.15728199e-01 2.11740479e-01 2.23455012e-01 -1.40714303e-01 -7.09328651e-01 -1.71618447e-01 -1.92241758e-01 1.76878974e-01 4.46338534e-01 2.69847035e-01 4.33278292e-01 -2.44029954e-01 -2.64788687e-01 6.47304356e-01 1.07157481e+00 5.09887077e-02 5.24674356e-01 9.09426212e-01 6.98885918e-01 6.76158369e-01 4.48451817e-01 -5.12058377e-01 2.17637420e-01 9.05062497e-01 1.47774041e-01 -3.52316529e-01 -3.85585815e-01 -6.31039515e-02 5.86862028e-01 1.90504301e+00 -3.10035825e-01 -2.32171744e-01 -1.29786086e+00 7.54200995e-01 -1.88438082e+00 -8.33541512e-01 -3.75766516e-01 2.14144182e+00 1.06192470e+00 3.25493962e-02 1.57420579e-02 -4.81606424e-01 7.61191845e-01 1.82656109e-01 -3.41744661e-01 -1.17992747e+00 -4.52978313e-01 6.36178553e-01 6.17203295e-01 8.12131345e-01 -7.79095054e-01 1.46977162e+00 7.88275766e+00 6.55516326e-01 -1.08323336e+00 2.20693365e-01 5.26934564e-01 -5.71904844e-03 -4.56863165e-01 3.77632380e-02 -7.43371308e-01 2.59527892e-01 1.27388752e+00 -3.14343542e-01 6.82308912e-01 6.08661413e-01 -2.33511180e-02 4.74585354e-01 -1.35451412e+00 7.27045894e-01 3.02235395e-01 -1.37555194e+00 4.65744287e-02 1.87498227e-01 8.23061883e-01 8.27301502e-01 1.41087711e-01 2.88702935e-01 6.46808267e-01 -1.41965127e+00 4.42159504e-01 -8.72598866e-06 9.67685759e-01 -4.26156402e-01 8.69906068e-01 -4.96623069e-02 -6.22386038e-01 4.43831414e-01 -2.94569492e-01 -3.46807837e-01 6.75014436e-01 2.53161013e-01 -8.00753236e-01 6.12202108e-01 3.02047819e-01 7.62486756e-01 -1.30564794e-01 6.97874725e-01 -2.49093413e-01 5.89707017e-01 -3.22466120e-02 1.63517907e-01 6.64518654e-01 -4.96529669e-01 8.13600600e-01 1.66249180e+00 5.87192416e-01 -2.62778610e-01 -1.66678682e-01 3.46065402e-01 6.07142784e-03 4.44282144e-01 -7.14925110e-01 -3.48019779e-01 3.39825541e-01 1.10468256e+00 -3.50074142e-01 -3.93157363e-01 -5.49245954e-01 1.28436923e+00 4.34009194e-01 2.09848061e-01 -7.73907900e-01 -8.23551193e-02 1.17228997e+00 -2.07076266e-01 -1.63511798e-01 -6.19238019e-01 -8.49364877e-01 -1.04373813e+00 1.61298644e-02 -1.56143713e+00 2.13146098e-02 -6.85623109e-01 -1.30407906e+00 9.96045291e-01 -2.07401901e-01 -1.06029165e+00 -4.78721827e-01 -7.32936978e-01 -6.92692399e-01 1.44904900e+00 -1.27627385e+00 -1.26378059e+00 1.38486877e-01 -2.72941619e-01 9.64811027e-01 -4.01360810e-01 1.28447115e+00 3.85719121e-01 -5.24978280e-01 7.58763790e-01 1.82581946e-01 1.53484702e-01 1.06828010e+00 -1.29235351e+00 1.99385834e+00 9.19044852e-01 7.27793634e-01 1.03480923e+00 8.35895598e-01 -5.64291179e-01 -1.27882206e+00 -6.48419857e-01 1.74042594e+00 -1.07250285e+00 9.95910883e-01 -3.97887558e-01 -8.51656258e-01 1.10462379e+00 1.03323042e+00 -7.62605846e-01 8.46709013e-01 3.42079878e-01 -3.76342654e-01 3.50275576e-01 -4.49806213e-01 1.23238921e+00 1.36491871e+00 -5.47208667e-01 -8.30926061e-01 5.48572958e-01 1.04476774e+00 -6.71486616e-01 -6.00360572e-01 3.95041794e-01 8.31386685e-01 -2.80171454e-01 8.98924530e-01 -1.25548506e+00 7.76624799e-01 -8.73353183e-02 -2.61907578e-01 -1.68869841e+00 -5.53777874e-01 -1.37062895e+00 3.39660823e-01 8.64071488e-01 8.74356627e-01 -8.03057432e-01 6.38357699e-01 1.72066987e-01 -8.04468811e-01 -8.16744447e-01 -1.20664752e+00 -5.29766798e-01 9.34934139e-01 -3.91412228e-01 7.50811398e-01 1.40148675e+00 3.10389429e-01 5.56848824e-01 -4.80884820e-01 -3.23334366e-01 2.46040642e-01 -3.67594026e-02 8.51561129e-01 -8.44868541e-01 -4.61174637e-01 -9.43697512e-01 -1.72887459e-01 -1.17952538e+00 -9.11932811e-02 -1.38247061e+00 2.48350263e-01 -1.70053697e+00 1.35004640e-01 1.10815249e-01 -4.43970919e-01 5.92605829e-01 -4.54484463e-01 6.90759599e-01 4.58547808e-02 3.92003804e-01 4.91459109e-02 2.67412830e-02 9.89423275e-01 -8.95520970e-02 1.70926556e-01 -2.52826333e-01 -8.28062594e-01 3.08342218e-01 1.08674538e+00 -4.38512176e-01 1.23066686e-01 -1.36143959e+00 7.04858005e-01 -4.01952416e-01 -3.19492042e-01 -4.72798228e-01 3.83874662e-02 -1.70847967e-01 -1.69336572e-01 -4.17346179e-01 1.35618478e-01 -2.74709046e-01 -2.57995110e-02 3.59179974e-01 -4.48722363e-01 1.20684111e+00 5.20845473e-01 -3.55749547e-01 -1.28582984e-01 -5.46501018e-02 3.59855443e-01 -2.87453562e-01 -4.59667057e-01 6.69290870e-02 -4.52039003e-01 -5.99711500e-02 3.56583327e-01 -3.27562690e-01 -4.88140255e-01 -1.59311980e-01 -4.88627642e-01 3.29002380e-01 3.91159475e-01 6.02302015e-01 -5.71661256e-02 -1.27924204e+00 -1.23658597e+00 -1.11447543e-01 1.12670980e-01 -1.58793703e-01 -1.80159777e-01 1.15738678e+00 -4.37135011e-01 8.38573456e-01 -4.37388808e-01 -4.62899148e-01 -1.28461349e+00 3.48247349e-01 2.60259032e-01 -1.97666824e-01 -6.00689471e-01 9.64836001e-01 -2.06379041e-01 -7.25598991e-01 -1.93770766e-01 -1.87507436e-01 2.04421699e-01 -1.47870094e-01 3.28547806e-01 3.84951234e-02 3.80512416e-01 -7.51437783e-01 -2.43781924e-01 6.93481147e-01 -2.54759312e-01 -4.02628660e-01 1.31998467e+00 -2.27534041e-01 -6.52414918e-01 6.22630596e-01 7.48483956e-01 7.77973235e-02 -4.33213532e-01 -3.09692889e-01 4.35642987e-01 -4.83361073e-03 -5.31086564e-01 -1.00099027e+00 -4.27880734e-01 1.20569813e+00 1.93412662e-01 -2.23585099e-01 6.13329053e-01 -2.15397000e-01 1.29152715e+00 4.26208675e-01 1.47279188e-01 -1.02683139e+00 -4.65647072e-01 8.92198443e-01 7.43908107e-01 -1.00574267e+00 8.92258994e-03 -5.32476567e-02 -3.95976692e-01 1.16158462e+00 5.24075091e-01 -9.89529863e-02 -2.04814240e-01 3.17058891e-01 5.64617574e-01 -9.74546000e-02 -7.77255893e-01 1.09253794e-01 2.80590534e-01 6.72684371e-01 1.17600834e+00 5.97910723e-03 -6.18980229e-01 2.16788188e-01 -9.58782554e-01 -4.78333205e-01 2.84184724e-01 5.14937878e-01 -3.70236307e-01 -1.72356570e+00 -2.23773330e-01 1.90580282e-02 -6.44323885e-01 -9.52700853e-01 -7.05507576e-01 8.35683048e-01 -7.98979774e-02 5.43645382e-01 3.04941744e-01 -2.53682733e-01 8.54457542e-02 8.27129781e-01 5.76665163e-01 -7.85838962e-01 -1.06487048e+00 -9.96352434e-02 4.42710400e-01 -3.95533621e-01 -2.38100469e-01 -8.53829145e-01 -6.11936033e-01 -4.75201935e-01 -1.14959747e-01 3.39545012e-01 1.01318109e+00 1.12713921e+00 3.09016943e-01 3.93075138e-01 -1.96927339e-02 -8.99183333e-01 -5.58555126e-01 -1.32726336e+00 4.60368156e-01 -1.45090073e-01 1.99443907e-01 5.29810339e-02 2.48480099e-03 -1.18489280e-01]	[11.576485633850098, 10.29174518585205]
1704dd60-55b3-4838-8a8c-69559a14d176	deeplsd-line-segment-detection-and-refinement	2212.07766	null	https://arxiv.org/abs/2212.07766v3	https://arxiv.org/pdf/2212.07766v3.pdf	DeepLSD: Line Segment Detection and Refinement with Deep Image Gradients	Line segments are ubiquitous in our human-made world and are increasingly used in vision tasks. They are complementary to feature points thanks to their spatial extent and the structural information they provide. Traditional line detectors based on the image gradient are extremely fast and accurate, but lack robustness in noisy images and challenging conditions. Their learned counterparts are more repeatable and can handle challenging images, but at the cost of a lower accuracy and a bias towards wireframe lines. We propose to combine traditional and learned approaches to get the best of both worlds: an accurate and robust line detector that can be trained in the wild without ground truth lines. Our new line segment detector, DeepLSD, processes images with a deep network to generate a line attraction field, before converting it to a surrogate image gradient magnitude and angle, which is then fed to any existing handcrafted line detector. Additionally, we propose a new optimization tool to refine line segments based on the attraction field and vanishing points. This refinement improves the accuracy of current deep detectors by a large margin. We demonstrate the performance of our method on low-level line detection metrics, as well as on several downstream tasks using multiple challenging datasets. The source code and models are available at https://github.com/cvg/DeepLSD.	['Marc Pollefeys', 'Martin R. Oswald', 'Viktor Larsson', 'Daniel Barath', 'Rémi Pautrat']	2022-12-15	null	http://openaccess.thecvf.com//content/CVPR2023/html/Pautrat_DeepLSD_Line_Segment_Detection_and_Refinement_With_Deep_Image_Gradients_CVPR_2023_paper.html	http://openaccess.thecvf.com//content/CVPR2023/papers/Pautrat_DeepLSD_Line_Segment_Detection_and_Refinement_With_Deep_Image_Gradients_CVPR_2023_paper.pdf	cvpr-2023-1	['line-segment-detection', 'line-detection']	['computer-vision', 'computer-vision']	[-1.02483705e-01 -3.34815562e-01 -2.01412112e-01 -4.45923179e-01 -6.65081918e-01 -7.37490714e-01 5.77962577e-01 2.79552191e-01 -4.96466547e-01 3.51866454e-01 -1.14971288e-01 -1.39550522e-01 3.50814134e-01 -6.81602299e-01 -9.01701808e-01 -2.71480501e-01 1.81804910e-01 2.71780044e-01 6.75567746e-01 -2.44732082e-01 4.99006420e-01 7.32793391e-01 -1.04798937e+00 -4.52869944e-02 7.31619895e-01 1.09832978e+00 -1.34274885e-01 5.55769920e-01 1.44686744e-01 4.57443863e-01 -2.36979753e-01 -3.65277320e-01 6.02278352e-01 -2.00451642e-01 -4.43125248e-01 2.44610474e-01 1.03278053e+00 -6.29442990e-01 -5.06817877e-01 9.00694489e-01 4.96878415e-01 -1.05324775e-01 5.08027554e-01 -1.14958048e+00 -4.53077197e-01 2.29519591e-01 -1.08535111e+00 3.97364013e-02 4.77018505e-01 2.98298448e-01 1.03713965e+00 -1.15078497e+00 7.49183536e-01 9.18662548e-01 1.18962955e+00 1.96241677e-01 -1.27224743e+00 -3.17694783e-01 3.21400352e-02 -9.55083966e-02 -1.37418211e+00 -3.83139670e-01 7.46925712e-01 -5.46197534e-01 7.90403128e-01 -1.89819574e-01 6.53824449e-01 8.63855004e-01 3.28228436e-02 8.52842212e-01 7.64461398e-01 -4.29105043e-01 -2.50563622e-02 -1.07513867e-01 7.60532729e-03 1.01784015e+00 3.24573517e-01 -2.58238856e-02 -5.03812194e-01 1.50686026e-01 1.41942430e+00 6.86327890e-02 -5.89266539e-01 -9.29077029e-01 -1.25891256e+00 7.46984184e-01 8.99991572e-01 -8.01810324e-02 -1.51886687e-01 2.67335087e-01 6.70750253e-03 -3.51605966e-04 2.41040632e-01 5.23766279e-01 -2.65942931e-01 9.85232815e-02 -1.01315987e+00 3.21813196e-01 7.01018333e-01 1.00988829e+00 8.90701354e-01 -2.68373340e-01 -9.58531722e-02 7.02794611e-01 3.26304227e-01 5.66191792e-01 3.44514340e-01 -8.40638041e-01 5.12848079e-01 5.91368198e-01 2.02174112e-01 -1.16511309e+00 -8.18184495e-01 -8.29157472e-01 -4.22579199e-01 4.68373328e-01 6.74825191e-01 -1.61748558e-01 -8.14191282e-01 1.31218219e+00 3.65618497e-01 3.88572225e-03 -4.56971169e-01 1.09663880e+00 4.47828859e-01 4.13261503e-01 -4.57953244e-01 3.45712841e-01 1.17120123e+00 -1.20525360e+00 -8.41408744e-02 -6.06522501e-01 7.81353176e-01 -9.67322826e-01 1.15310431e+00 3.09088081e-01 -9.81405318e-01 -3.73456091e-01 -1.26823771e+00 -4.94782716e-01 -8.25139508e-02 4.76837903e-01 5.27156889e-01 2.88935900e-01 -1.02177715e+00 5.24790645e-01 -8.94020319e-01 -4.40764785e-01 4.97432411e-01 1.74290195e-01 -3.28927636e-01 1.06647454e-01 -5.15094757e-01 6.64202154e-01 5.60322553e-02 2.10375115e-01 -2.50951171e-01 -7.38228202e-01 -1.01660299e+00 -3.52481157e-02 4.95582938e-01 -8.87621343e-01 1.36149991e+00 -7.39342690e-01 -1.40374732e+00 8.59743416e-01 -7.87854567e-02 -6.47468567e-01 1.06618452e+00 -7.01983392e-01 1.19584844e-01 3.43133122e-01 2.65322655e-01 8.64384711e-01 6.86440468e-01 -1.15545022e+00 -7.26696670e-01 -2.21329540e-01 -1.24116294e-01 2.18775675e-01 1.57518890e-02 -4.52505499e-01 -7.44993329e-01 -5.70126057e-01 2.90655375e-01 -9.51281846e-01 -2.37687692e-01 6.48912907e-01 -6.79871976e-01 -4.22424898e-02 8.38206112e-01 -3.43818963e-01 9.11808312e-01 -2.01512098e+00 -2.85999328e-01 3.57598394e-01 3.14832091e-01 2.58573562e-01 -6.17818832e-02 4.12735134e-01 1.47278488e-01 -5.16278483e-02 -2.77044922e-01 -3.47719461e-01 -1.78130895e-01 -3.73137593e-01 -2.94263661e-01 7.68931866e-01 3.99896771e-01 1.01444221e+00 -9.32252467e-01 -2.91200429e-01 4.59393561e-01 4.86442447e-01 -5.17565966e-01 -6.80539906e-02 -7.64042661e-02 2.85883009e-01 -3.42379153e-01 4.06991869e-01 5.92114389e-01 -3.64179045e-01 -3.09650183e-01 -4.71858084e-01 -3.96292359e-01 2.97355000e-03 -1.15462136e+00 1.82401574e+00 -2.60867923e-01 1.07179415e+00 -3.29955697e-01 -5.21918416e-01 1.17916203e+00 -1.92678705e-01 3.97567809e-01 -7.65430093e-01 2.28747800e-01 4.28736001e-01 -1.98873922e-01 -1.44013301e-01 4.14134473e-01 4.78974968e-01 1.08924165e-01 2.40949854e-01 -6.11780807e-02 -5.78110039e-01 2.29002997e-01 2.54883528e-01 9.82009172e-01 4.24808502e-01 2.94968992e-01 -1.00183599e-01 3.45181793e-01 5.25263287e-02 3.95074338e-01 7.15700328e-01 -1.20125532e-01 1.01898527e+00 4.28915739e-01 -5.99352896e-01 -1.47143662e+00 -1.12621224e+00 -2.31060028e-01 5.80010653e-01 4.97957081e-01 -4.15738106e-01 -7.01164961e-01 -4.96156543e-01 1.71948686e-01 2.47119963e-01 -3.35643232e-01 7.79681932e-03 -6.18747532e-01 -9.36437696e-02 4.28229034e-01 7.62152135e-01 7.53581345e-01 -5.10472357e-01 -6.45747483e-01 1.64089292e-01 1.77244186e-01 -1.39615786e+00 -7.82794416e-01 -8.31466839e-02 -6.65698469e-01 -1.17554832e+00 -8.01337361e-01 -9.09941792e-01 8.13559055e-01 5.68475842e-01 9.17557955e-01 1.44133732e-01 -4.78057921e-01 2.69454718e-01 -1.07974976e-01 -4.08535838e-01 1.76009059e-01 8.70621204e-02 -2.53803041e-02 5.21338023e-02 8.32048282e-02 -2.49330401e-01 -1.05098808e+00 3.34432065e-01 -5.64294457e-01 2.65461922e-01 5.25511861e-01 8.11584592e-01 6.12768948e-01 -5.02734065e-01 -8.94225575e-03 -7.69504726e-01 5.14043450e-01 7.70721957e-02 -1.02896166e+00 -7.74128735e-02 -4.27184582e-01 1.54746726e-01 5.00998914e-01 -2.76336297e-02 -6.47202253e-01 4.99054551e-01 -2.29827568e-01 -3.87265712e-01 -5.13758287e-02 2.42263973e-01 3.70679170e-01 -4.53219414e-01 9.09586608e-01 -9.26901922e-02 -1.95046380e-01 -3.68894547e-01 5.69527864e-01 1.95939764e-01 9.33747351e-01 -2.79148281e-01 9.62717414e-01 8.64882410e-01 1.12544209e-01 -9.57084656e-01 -9.73384857e-01 -7.16458261e-01 -1.04615593e+00 -2.00103611e-01 5.58203101e-01 -8.24695647e-01 -4.45785612e-01 7.64710069e-01 -1.13754249e+00 -5.35341263e-01 -2.95450538e-01 4.93352443e-01 -6.24305010e-01 5.09469390e-01 -5.70571721e-01 -3.10333788e-01 -3.00717562e-01 -9.94181693e-01 1.25733948e+00 5.18861711e-01 -2.26725210e-02 -1.00058448e+00 8.58050659e-02 -3.71007361e-02 1.34561077e-01 4.72962588e-01 4.32606399e-01 -2.74341851e-01 -7.24271178e-01 -4.74436074e-01 -5.32763362e-01 1.84578121e-01 -1.19685933e-01 4.00670737e-01 -7.36868203e-01 -1.89935714e-01 -4.04695958e-01 -3.04792672e-01 1.03562939e+00 3.87516737e-01 8.66036892e-01 -1.32795259e-01 -4.40371245e-01 1.00596428e+00 1.59154344e+00 -1.86802760e-01 4.30956274e-01 6.90872908e-01 9.13258851e-01 3.48253161e-01 4.16496724e-01 3.77736837e-01 4.58960205e-01 8.00941944e-01 1.32238910e-01 -5.33226669e-01 -3.55970949e-01 -4.95502055e-01 5.68515249e-02 3.65213901e-01 4.48469706e-02 -2.00432494e-01 -1.10809135e+00 3.61759633e-01 -1.98455763e+00 -8.29173863e-01 -4.04819787e-01 2.31363344e+00 6.55369997e-01 5.66788554e-01 2.79249698e-01 9.01738852e-02 4.64042783e-01 -1.19758233e-01 -6.54591203e-01 -1.51805878e-01 -2.90612608e-01 -7.86826238e-02 8.84129941e-01 5.61231434e-01 -1.31808460e+00 1.13540316e+00 5.93714905e+00 4.00655419e-01 -1.63997281e+00 -4.57828283e-01 5.58462620e-01 1.62254363e-01 -1.38565660e-01 6.03102781e-02 -7.60773182e-01 1.12692766e-01 1.45564765e-01 -8.35195463e-03 -2.32052803e-03 8.55437398e-01 3.12086672e-01 -1.03348926e-01 -1.19344652e+00 1.26583648e+00 1.18189946e-01 -1.51856613e+00 -1.97587937e-01 -1.82578534e-01 8.69338453e-01 4.29322392e-01 1.07165612e-01 -3.41108054e-01 1.81628630e-01 -7.49373078e-01 8.79492640e-01 5.29939771e-01 5.98620832e-01 -4.48304564e-01 4.69993621e-01 1.86255500e-01 -1.09101117e+00 9.16401148e-02 -1.81936607e-01 -4.63984422e-02 2.82009006e-01 5.34726024e-01 -1.09626722e+00 2.05589741e-01 4.92003173e-01 9.18815255e-01 -8.12756181e-01 1.70822644e+00 -6.36504710e-01 3.92097235e-01 -7.39113510e-01 8.96413475e-02 4.36217040e-01 -2.41200835e-01 3.90502632e-01 1.30755711e+00 2.81569690e-01 -4.09979045e-01 2.52281129e-01 8.91394377e-01 8.21261108e-03 1.22409999e-01 -6.64572239e-01 3.52525622e-01 5.42512834e-01 1.40760589e+00 -9.33686852e-01 -3.38637047e-02 -6.32848680e-01 1.12002587e+00 5.33186793e-01 4.38103497e-01 -6.58072472e-01 -6.68111086e-01 3.91428024e-01 3.68448108e-01 3.36454868e-01 -7.42802620e-01 -3.60942960e-01 -1.26629984e+00 3.14261049e-01 -4.23982650e-01 9.16321762e-03 -7.31774211e-01 -1.14144003e+00 5.82895577e-01 -4.63239849e-01 -1.41759169e+00 -2.58841693e-01 -7.86018908e-01 -6.85070813e-01 4.76081878e-01 -1.60336816e+00 -1.23417556e+00 -6.65779829e-01 6.29271746e-01 4.83718336e-01 1.63944334e-01 2.23528117e-01 1.55252129e-01 -5.33466101e-01 6.94106340e-01 1.71070322e-01 5.73505640e-01 8.94205809e-01 -1.30593777e+00 9.46369946e-01 9.80277002e-01 4.21760470e-01 4.91041094e-01 6.72957718e-01 -3.28218013e-01 -1.22759843e+00 -9.99633431e-01 5.17149806e-01 -4.08837199e-01 7.59296596e-01 -5.00244498e-01 -7.82979429e-01 6.39668047e-01 -1.48766980e-01 3.29158247e-01 1.15264140e-01 -1.07019722e-01 -6.11875355e-01 -9.92864072e-02 -7.38457918e-01 8.03759515e-01 1.02502418e+00 -5.30890703e-01 -2.49079451e-01 3.73932660e-01 4.93398577e-01 -5.94948292e-01 -5.36541998e-01 2.39131913e-01 6.47501647e-01 -1.24492157e+00 1.06098044e+00 -7.11034164e-02 2.51246244e-01 -4.92457390e-01 3.86136830e-01 -1.26502287e+00 -3.63799870e-01 -7.03678071e-01 2.46674731e-01 9.52597022e-01 5.86230457e-01 -7.51367450e-01 9.83637631e-01 3.42388690e-01 -2.53048003e-01 -9.91294861e-01 -7.25401640e-01 -7.68944144e-01 -1.87863246e-01 -3.22253466e-01 2.97603518e-01 6.78495646e-01 -1.21090397e-01 4.95985061e-01 -2.57379144e-01 1.46462128e-01 8.34181070e-01 2.30870053e-01 1.05581260e+00 -9.57784593e-01 -8.83414224e-02 -7.21810639e-01 -8.35039139e-01 -1.59277081e+00 -3.19289684e-01 -6.90842807e-01 1.81065172e-01 -1.68143117e+00 -2.65534908e-01 -4.96495873e-01 2.29671806e-01 4.01034832e-01 -8.21812153e-02 6.17212772e-01 3.89125377e-01 4.50629234e-01 -4.94281232e-01 3.51702511e-01 1.12203300e+00 9.27355289e-02 -3.49474162e-01 -1.01826042e-02 -4.07542288e-01 1.15571678e+00 8.18770349e-01 -5.82638271e-02 -1.26789182e-01 -7.55493104e-01 2.70303160e-01 -3.48214120e-01 5.96954823e-01 -1.28731251e+00 6.16095126e-01 2.24339560e-01 7.02643335e-01 -5.75728178e-01 3.84774387e-01 -4.83633965e-01 -5.29806972e-01 3.67728204e-01 -1.99646577e-01 9.83244702e-02 5.29919341e-02 3.57136488e-01 6.11315249e-03 -1.16082080e-01 8.21532726e-01 1.81199178e-01 -9.95129526e-01 5.65862954e-01 1.96851850e-01 -2.90420186e-02 1.17530215e+00 -4.94330317e-01 -5.58655262e-01 -5.32751381e-01 -2.27072522e-01 3.21943969e-01 8.51264596e-01 3.40607941e-01 7.75967300e-01 -1.21188819e+00 -6.98850632e-01 3.50261271e-01 1.96029529e-01 2.59365946e-01 -1.39594734e-01 8.92292082e-01 -1.25965786e+00 4.08644110e-01 -2.47507006e-01 -8.99397731e-01 -8.46068501e-01 3.97071004e-01 5.96043050e-01 2.32623488e-01 -9.46722746e-01 9.27341223e-01 3.54361266e-01 -1.24748833e-01 2.46912763e-01 -5.50934017e-01 7.23000020e-02 -2.18789071e-01 3.49989623e-01 2.46358782e-01 1.55078451e-04 -5.94855547e-01 -3.25864345e-01 1.16338599e+00 -6.93563446e-02 -4.57918383e-02 1.21992123e+00 -1.44362822e-01 3.05762500e-01 3.01628560e-01 1.23746645e+00 3.65999728e-01 -1.75355315e+00 -3.20383400e-01 1.04570210e-01 -6.33125484e-01 -4.66743819e-02 -3.83153796e-01 -9.28620696e-01 8.35377514e-01 4.38941509e-01 1.15965083e-01 7.60581613e-01 -2.79700048e-02 9.15391505e-01 3.36053401e-01 2.81912655e-01 -9.01695013e-01 2.81575024e-01 3.73890102e-01 8.42782974e-01 -1.33213699e+00 1.17123522e-01 -6.58264995e-01 -5.03910899e-01 1.54839075e+00 6.16392136e-01 -5.35892308e-01 3.82727623e-01 4.20571327e-01 3.85432720e-01 -1.20758295e-01 -3.22754651e-01 -2.08229616e-01 3.85127544e-01 5.93708575e-01 4.75914240e-01 -2.44774655e-01 -1.46361455e-01 3.15260105e-02 -3.48561674e-01 -2.18010359e-02 5.08416295e-01 8.17925274e-01 -6.16304576e-01 -9.17109370e-01 -3.05507660e-01 2.68934399e-01 -6.08543791e-02 7.08900839e-02 -4.41445768e-01 7.90054798e-01 -1.75287724e-01 6.17205083e-01 1.40105233e-01 -2.41365224e-01 5.90938389e-01 -6.17713571e-01 4.33041722e-01 -3.84910196e-01 -1.76970601e-01 1.07279457e-01 -1.03155665e-01 -8.27140629e-01 -6.14594556e-02 -6.27839923e-01 -1.36957443e+00 -9.30821672e-02 -3.06633949e-01 -1.92124590e-01 6.92471802e-01 6.54287159e-01 3.71350765e-01 6.06044419e-02 6.10460579e-01 -1.08566964e+00 -4.72010612e-01 -4.68430549e-01 -2.70073295e-01 4.83596712e-01 4.78747278e-01 -4.68958139e-01 -1.69597268e-01 -4.75465357e-02]	[8.234997749328613, -1.8226003646850586]
1afe8a97-6882-4763-a594-40045214162d	a-multi-scale-video-denoising-algorithm-for	2209.01740	null	https://arxiv.org/abs/2209.01740v1	https://arxiv.org/pdf/2209.01740v1.pdf	A Multi-scale Video Denoising Algorithm for Raw Image	Video denoising for raw image has always been the difficulty of camera image processing. On the one hand, image denoising performance largely determines the image quality, moreover denoising effect in raw image will affect the accuracy of the following operations of ISP processing flow. On the other hand, compared with image, video have motion information in time sequence, thus motion estimation which is complex and computationally expensive is needed in video denoising. In view of the above problems, this paper proposes a video denoising algorithm for raw image, performing multiple cascading processing stages on raw-RGB image based on convolutional neural network, and carries out implicit motion estimation in the network. The denoising performance is far superior to that of traditional algorithms with minimal computation and bandwidth, and has computational advantages compared with most deep learning algorithms.	['Kai Li', 'Xianxian Lv', 'Yueli Hu', 'Bin Ma']	2022-09-05	null	null	null	null	['video-denoising']	['computer-vision']	[ 2.18253940e-01 -7.68227458e-01 2.70518929e-01 -2.03154594e-01 8.64532441e-02 -1.13909692e-01 4.62779924e-02 -3.61872166e-01 -9.95213449e-01 4.33845431e-01 6.03701174e-02 -1.29091293e-01 -1.42899733e-02 -8.42056572e-01 -2.62641519e-01 -1.06851220e+00 9.74752307e-02 -5.31059563e-01 3.12959611e-01 -2.26300418e-01 3.33116502e-01 3.68782222e-01 -1.20059907e+00 1.86351359e-01 5.18092096e-01 1.27435732e+00 3.58266711e-01 7.86199450e-01 -2.05172777e-01 1.12922990e+00 -5.91069937e-01 -2.12387189e-01 2.00505674e-01 -3.40713531e-01 -3.50616932e-01 1.73533514e-01 -1.65910736e-01 -9.51207876e-01 -8.64566147e-01 1.38284624e+00 4.38855618e-01 1.73807353e-01 3.19004089e-01 -9.28093612e-01 -6.30520225e-01 2.10354388e-01 -6.23909533e-01 6.34834945e-01 -2.52249222e-02 6.85709417e-02 1.30162641e-01 -7.09654868e-01 2.79267043e-01 1.21644127e+00 9.18487549e-01 4.19313729e-01 -7.26734042e-01 -5.68348765e-01 1.04968727e-01 5.38518667e-01 -1.20397592e+00 -2.54720777e-01 8.20462883e-01 -6.33400232e-02 5.88201582e-01 -2.21272558e-02 9.14136946e-01 8.16856623e-01 5.36509573e-01 5.53789437e-01 6.66230798e-01 -7.71351457e-02 4.61050831e-02 -4.83840853e-01 1.48229882e-01 3.06640357e-01 3.08184087e-01 -1.14009276e-01 -1.76800355e-01 5.76297224e-01 1.11401880e+00 3.97527039e-01 -4.64822769e-01 4.77161914e-01 -9.09379542e-01 4.78529900e-01 1.88606441e-01 3.70833457e-01 -4.82786804e-01 4.64339912e-01 6.63657784e-01 5.58834374e-01 2.38328159e-01 -4.98283178e-01 -4.37698960e-01 -4.31461364e-01 -1.11331689e+00 -1.13073744e-01 3.96931201e-01 6.75757647e-01 5.97712636e-01 4.63154107e-01 3.39444876e-01 5.36303163e-01 1.70588747e-01 4.33587760e-01 5.52702308e-01 -1.22159386e+00 3.62079203e-01 1.89737558e-01 -1.35569394e-01 -1.37606025e+00 -3.18551987e-01 -7.70760104e-02 -1.69084311e+00 5.44865072e-01 3.99292439e-01 -1.87520519e-01 -1.02833760e+00 1.22007501e+00 -2.28279568e-02 1.63134396e-01 -2.09956691e-02 1.07609046e+00 9.43167210e-01 1.14948332e+00 1.94993719e-01 -7.46760011e-01 1.03575611e+00 -8.03913891e-01 -1.43926859e+00 5.58950491e-02 1.05544418e-01 -9.01130259e-01 6.33599043e-01 1.08704102e+00 -1.11029446e+00 -9.83952701e-01 -1.06915891e+00 -2.93819487e-01 -1.54919073e-01 5.18268459e-02 7.40855753e-01 5.99325240e-01 -1.07762694e+00 8.61635864e-01 -8.92106950e-01 1.92721584e-03 2.83164918e-01 5.00613809e-01 -4.23445821e-01 -2.40929484e-01 -1.15303326e+00 7.40187407e-01 5.95462263e-01 9.42803979e-01 -6.83027804e-01 -2.62324005e-01 -6.78455055e-01 -5.47573622e-03 9.25916210e-02 -5.31759262e-01 1.13940334e+00 -1.52915037e+00 -1.37897801e+00 1.05742447e-01 -2.85638779e-01 -2.10016355e-01 7.73195744e-01 -3.06834221e-01 -7.34443903e-01 3.44416112e-01 -2.79970020e-01 4.75349188e-01 1.25051153e+00 -9.50538516e-01 -6.23078942e-01 -2.15496406e-01 -2.34311208e-01 8.94178972e-02 -3.94230664e-01 -6.20895624e-02 -7.27421820e-01 -7.03999996e-01 5.32923222e-01 -2.03548685e-01 -3.60062093e-01 8.00496861e-02 2.07750708e-01 -6.27443567e-03 1.38803196e+00 -1.27029276e+00 1.55479121e+00 -2.37217450e+00 5.45700826e-02 6.89009717e-03 1.03730060e-01 4.41600293e-01 -6.92772642e-02 7.49366954e-02 -2.93784946e-01 6.93629608e-02 -7.96840042e-02 1.09123670e-01 -6.27242744e-01 3.46229553e-01 4.21995483e-02 6.56259358e-01 1.06090285e-01 3.76876324e-01 -6.16838992e-01 -5.93198180e-01 5.22816598e-01 7.66629040e-01 -3.26349854e-01 1.56205460e-01 2.87822366e-01 6.02199018e-01 -5.23610473e-01 4.44891840e-01 1.11717749e+00 2.20915303e-01 -1.66848749e-01 -6.82778835e-01 -2.40678236e-01 -3.58661443e-01 -1.52186537e+00 1.54582024e+00 -1.80662185e-01 7.23887086e-01 5.57448030e-01 -1.23705697e+00 6.67114139e-01 4.59956110e-01 5.84092855e-01 -9.20330048e-01 4.79757547e-01 5.93423829e-05 -1.32232502e-01 -1.13778079e+00 4.80055690e-01 -1.56183317e-01 4.77610201e-01 3.83947715e-02 -1.90159753e-01 4.77893092e-02 2.17653036e-01 9.05905217e-02 1.00795746e+00 4.08484817e-01 1.73455819e-01 7.23152012e-02 7.29648054e-01 3.39127779e-02 9.07118559e-01 5.21921754e-01 -4.06571329e-01 7.57189810e-01 3.15883845e-01 -8.58135998e-01 -1.21810961e+00 -8.23793650e-01 1.01103492e-01 5.80812633e-01 5.70804596e-01 -1.29486611e-02 -9.28031802e-01 -1.92862656e-02 -3.69051337e-01 1.69249445e-01 -3.35226446e-01 -1.07179135e-01 -1.00557530e+00 -8.67276073e-01 3.63867342e-01 5.41824043e-01 1.17509866e+00 -9.82096136e-01 -3.80653739e-01 4.63805109e-01 -3.14290792e-01 -1.06162250e+00 -4.11270946e-01 6.38428256e-02 -1.43639135e+00 -1.06139946e+00 -4.77074653e-01 -1.00495565e+00 6.77779496e-01 7.09615588e-01 9.35975790e-01 6.25698209e-01 -2.33993411e-01 9.49229747e-02 -4.43826288e-01 -1.72608480e-01 -9.70772132e-02 -5.24518132e-01 -7.94284567e-02 -1.46326080e-01 4.21067268e-01 -6.94802761e-01 -9.74508762e-01 -5.03116660e-02 -1.45495236e+00 2.93015260e-02 6.44531906e-01 8.57199371e-01 3.22663635e-01 1.08264947e+00 2.39860430e-01 -3.30725282e-01 4.24232423e-01 -1.69597894e-01 -7.28913605e-01 -1.08302847e-01 -2.68465519e-01 -4.07695770e-01 8.49366724e-01 -2.11282670e-01 -1.33971536e+00 5.83645962e-02 -4.54152405e-01 -3.89803648e-01 -5.00995815e-02 4.19127196e-01 -2.79882580e-01 -6.90813735e-02 6.82042027e-03 4.71934557e-01 3.24531555e-01 -5.60178459e-01 5.95712587e-02 7.51676321e-01 8.03557098e-01 7.38286376e-02 6.78060651e-01 5.30428886e-01 5.69152348e-02 -1.26767898e+00 -2.10587502e-01 -3.43702734e-01 -8.40796709e-01 -4.96519238e-01 1.30366087e+00 -1.10220385e+00 -9.60064411e-01 1.01666832e+00 -1.54439926e+00 7.22751990e-02 3.50630522e-01 6.41843021e-01 -3.42768990e-02 9.84468341e-01 -1.19881928e+00 -7.37454534e-01 -3.30148965e-01 -1.20443130e+00 4.10580367e-01 3.69134128e-01 2.39456967e-01 -1.02909958e+00 -5.35585642e-01 2.71321833e-01 3.76744807e-01 5.98450340e-02 6.67626560e-01 2.70069510e-01 -8.21144104e-01 -3.55020970e-01 -3.03140402e-01 9.94894326e-01 4.53853793e-02 2.62778819e-01 -7.80152380e-01 -1.34131089e-01 7.44521201e-01 1.60190955e-01 1.00015676e+00 1.03481519e+00 1.54434860e+00 -9.06382315e-03 1.86144471e-01 8.59234393e-01 1.86494923e+00 5.62128842e-01 1.37005711e+00 7.34917402e-01 7.22359836e-01 2.05722108e-01 6.42382145e-01 3.46783310e-01 -8.37913528e-02 -6.85957968e-02 6.04978383e-01 -3.11959207e-01 -6.58376142e-03 3.11577708e-01 4.86192167e-01 1.23472643e+00 -5.55350840e-01 -3.04395407e-01 -4.19891953e-01 3.55867743e-01 -1.64729095e+00 -1.20414531e+00 -6.44014299e-01 1.90045476e+00 4.83359247e-01 1.13492511e-01 -3.71138781e-01 4.85355794e-01 8.48198056e-01 2.29000911e-01 -3.98623317e-01 -3.16350490e-01 -2.61659443e-01 2.83764657e-02 6.90670371e-01 1.55151397e-01 -1.14276159e+00 5.21233618e-01 6.63249731e+00 8.03036332e-01 -1.16176212e+00 2.25241259e-01 7.57349670e-01 -3.73450331e-02 1.60873666e-01 -2.02239871e-01 -3.17598850e-01 6.71909153e-01 5.93965173e-01 2.97675490e-01 6.43465638e-01 5.36770821e-01 9.06976342e-01 -4.81358886e-01 -7.08353519e-01 1.43581581e+00 -1.43714294e-01 -1.02210665e+00 8.71522576e-02 -2.65074462e-01 4.38963413e-01 -2.87356555e-01 -1.30343691e-01 -1.99943054e-02 -1.97091356e-01 -1.06468749e+00 7.21569836e-01 4.53853250e-01 4.73841995e-01 -9.83748317e-01 1.17693985e+00 3.36924762e-01 -1.15386951e+00 -2.43586630e-01 -7.31411755e-01 -5.54885149e-01 4.99936998e-01 8.39158356e-01 2.70648479e-01 5.45899689e-01 1.16535079e+00 8.28920424e-01 -2.07888275e-01 9.28416014e-01 -2.76896954e-01 6.80755913e-01 -5.13841994e-02 2.94411153e-01 3.83812100e-01 -8.68729353e-01 2.11771667e-01 1.23248041e+00 6.03266835e-01 4.89120483e-01 -9.88235921e-02 2.00521931e-01 -4.94067222e-02 -1.53732583e-01 -3.76969099e-01 -3.09008900e-02 9.32645798e-02 1.29909348e+00 -7.92237282e-01 -5.23689985e-01 -7.64917791e-01 1.22507930e+00 -4.05211627e-01 5.75020492e-01 -8.79342139e-01 -5.04253924e-01 5.12747169e-01 -1.36987552e-01 3.75793248e-01 -5.14260471e-01 -4.62728083e-01 -1.12323749e+00 1.72522962e-01 -6.66688502e-01 2.55273461e-01 -8.82879019e-01 -1.04907405e+00 3.50212932e-01 -5.04871786e-01 -1.32243633e+00 2.52757996e-01 -8.85157406e-01 -8.27913344e-01 8.70510101e-01 -1.50987613e+00 -7.34205306e-01 -7.30238736e-01 7.40206957e-01 8.36919188e-01 5.16083688e-02 2.15959013e-01 7.81169832e-01 -9.61144030e-01 -3.67662385e-02 5.04603028e-01 3.94070983e-01 4.46240485e-01 -7.50505030e-01 6.37373701e-02 1.41160500e+00 -6.53125107e-01 4.70885098e-01 7.09460557e-01 -7.23097146e-01 -1.68059313e+00 -8.82321060e-01 3.75692219e-01 2.57040679e-01 4.85927463e-01 2.57556379e-01 -1.00266826e+00 4.51946825e-01 6.95252717e-01 -2.00181574e-01 2.13669419e-01 -5.35195291e-01 1.80073529e-01 -3.83643329e-01 -1.09775877e+00 5.40554106e-01 8.04406941e-01 -2.12356716e-01 -4.77332920e-01 -3.76148969e-02 6.04989290e-01 -1.83307052e-01 -9.61620867e-01 2.58590162e-01 5.19060075e-01 -1.00719011e+00 1.13827515e+00 -2.18557149e-01 6.46643579e-01 -7.32359171e-01 -6.13075793e-02 -8.36265147e-01 -6.86705530e-01 -2.76956558e-01 1.56641141e-01 1.23447144e+00 -1.91514105e-01 -2.56445795e-01 7.30563998e-01 4.46839452e-01 -1.09361783e-01 -2.47056633e-01 -8.09888899e-01 -5.70381105e-01 -2.12222740e-01 -6.61105156e-01 2.15681612e-01 7.42248952e-01 -5.99959075e-01 1.22716032e-01 -7.76072443e-01 3.41673702e-01 8.53001475e-01 -5.14754117e-01 4.13547933e-01 -8.67835045e-01 4.11412641e-02 -2.76227146e-01 -5.10658622e-01 -1.15597081e+00 -1.92305073e-01 -1.07714795e-01 5.66622280e-02 -1.83836234e+00 4.16960269e-02 1.98434934e-01 -2.93956697e-01 1.38291884e-02 -1.97625786e-01 4.88088489e-01 -2.98537277e-02 2.77859092e-01 -4.28891152e-01 5.09972870e-01 1.50577700e+00 -2.93870211e-01 6.59365505e-02 -3.26775134e-01 -2.61029154e-01 1.00293863e+00 6.32450819e-01 -2.63545752e-01 -3.08290899e-01 -1.06050706e+00 9.48486850e-02 1.44474134e-01 2.69145161e-01 -1.10963285e+00 5.15589356e-01 -6.30849451e-02 9.86140311e-01 -8.76821280e-01 2.38023207e-01 -1.32888329e+00 3.06915909e-01 8.53001475e-01 2.38442004e-01 4.45302844e-01 4.87542786e-02 6.12786651e-01 -7.56564319e-01 -4.88909811e-01 8.07782352e-01 -5.09197593e-01 -1.52192152e+00 3.19253355e-01 -7.89289355e-01 -3.51615071e-01 7.85614491e-01 -7.59744883e-01 1.97755657e-02 -4.79305238e-01 -6.41821444e-01 -6.85863644e-02 3.01142603e-01 5.34118712e-02 8.26273084e-01 -1.20361388e+00 -3.76962572e-01 3.33834328e-02 -7.72987783e-01 2.21113414e-01 6.17610395e-01 9.47565854e-01 -1.45692122e+00 -1.74718767e-01 -3.24186832e-01 -5.24119198e-01 -1.30082440e+00 6.67810857e-01 2.15814427e-01 1.81809291e-01 -6.89887762e-01 7.27518559e-01 -1.38961226e-01 3.39801103e-01 2.01418430e-01 -1.79345191e-01 -3.83497000e-01 4.89159673e-02 9.97720659e-01 6.72178447e-01 5.29849865e-02 -5.04989803e-01 1.40138073e-02 6.70913756e-01 -1.98939368e-02 1.50414303e-01 1.58899534e+00 -5.77880442e-01 -6.91054106e-01 8.49404112e-02 1.37113607e+00 -5.30857503e-01 -1.28122854e+00 1.92698404e-01 -2.67635822e-01 -6.46185696e-01 6.48474991e-01 -2.20114306e-01 -1.56118119e+00 9.71337974e-01 8.76291513e-01 1.70647815e-01 1.87944400e+00 -9.18372095e-01 1.03232622e+00 4.48936522e-01 2.69267678e-01 -1.63540137e+00 8.56529176e-02 4.41080242e-01 3.38152736e-01 -1.29421282e+00 2.02390641e-01 -5.29637933e-01 -3.03293496e-01 1.69370139e+00 5.58855951e-01 -1.93673402e-01 6.00971758e-01 4.53494847e-01 3.42679024e-01 9.87117141e-02 -3.41933101e-01 2.46323168e-01 -3.31569523e-01 5.11344135e-01 5.48619986e-01 -6.41114295e-01 -6.27409756e-01 3.72764587e-01 3.66928607e-01 4.50904697e-01 6.13638341e-01 9.35332477e-01 -6.22494876e-01 -8.40977430e-01 -7.56022871e-01 2.65536755e-01 -8.75280440e-01 -1.61571860e-01 4.31457132e-01 6.91707015e-01 3.68331134e-01 1.20741987e+00 1.88048482e-01 -4.44661498e-01 2.31918111e-01 -3.46158564e-01 2.74328321e-01 2.27482140e-01 -3.98283035e-01 2.83574671e-01 -1.06271200e-01 -6.04937375e-01 -6.33782685e-01 -4.41820115e-01 -1.19472790e+00 -7.95805871e-01 -3.66016775e-02 2.51092855e-02 7.66922116e-01 9.85950291e-01 -1.93103209e-01 8.05949450e-01 5.07482469e-01 -9.30784166e-01 -3.76842290e-01 -8.53829682e-01 -7.80823708e-01 6.48516774e-01 3.53490472e-01 -2.44452193e-01 -3.89773130e-01 4.92870390e-01]	[11.358030319213867, -2.2590293884277344]
46ed51f9-7198-4da7-a911-92b83eda8b00	generative-and-pseudo-relevant-feedback-for	2305.07477	null	https://arxiv.org/abs/2305.07477v1	https://arxiv.org/pdf/2305.07477v1.pdf	Generative and Pseudo-Relevant Feedback for Sparse, Dense and Learned Sparse Retrieval	Pseudo-relevance feedback (PRF) is a classical approach to address lexical mismatch by enriching the query using first-pass retrieval. Moreover, recent work on generative-relevance feedback (GRF) shows that query expansion models using text generated from large language models can improve sparse retrieval without depending on first-pass retrieval effectiveness. This work extends GRF to dense and learned sparse retrieval paradigms with experiments over six standard document ranking benchmarks. We find that GRF improves over comparable PRF techniques by around 10% on both precision and recall-oriented measures. Nonetheless, query analysis shows that GRF and PRF have contrasting benefits, with GRF providing external context not present in first-pass retrieval, whereas PRF grounds the query to the information contained within the target corpus. Thus, we propose combining generative and pseudo-relevance feedback ranking signals to achieve the benefits of both feedback classes, which significantly increases recall over PRF methods on 95% of experiments.	['Jeffrey Dalton', 'Shubham Chatterjee', 'Iain Mackie']	2023-05-12	null	null	null	null	['document-ranking']	['natural-language-processing']	[ 2.78232008e-01 -8.20939913e-02 -7.17796326e-01 -1.08592369e-01 -1.58735049e+00 -7.13753879e-01 1.09661233e+00 5.19611597e-01 -5.86378157e-01 6.87226951e-01 8.15459788e-01 -1.61493286e-01 -4.55457360e-01 -6.82201326e-01 -5.07229388e-01 1.08979391e-02 4.11906280e-02 7.17644989e-01 5.37272215e-01 -8.51094961e-01 7.47430861e-01 1.50217116e-01 -1.64384556e+00 6.76680088e-01 6.70635045e-01 7.50351369e-01 3.30372810e-01 6.51849806e-01 -2.57594258e-01 4.70745802e-01 -6.71907127e-01 -2.12692857e-01 3.38632375e-01 -2.10966572e-01 -9.44684505e-01 -7.14980423e-01 6.72032297e-01 -5.02579033e-01 -4.55277473e-01 7.26771355e-01 5.90314806e-01 3.96801233e-01 7.43874073e-01 -6.81462049e-01 -1.08031797e+00 8.99381876e-01 -4.93759245e-01 5.51324666e-01 9.74407375e-01 -3.48057032e-01 1.62277508e+00 -1.43944061e+00 8.10301125e-01 1.43500006e+00 3.12687516e-01 4.57778811e-01 -1.16619873e+00 -4.37272221e-01 2.34300852e-01 -1.97622135e-01 -1.33991206e+00 -3.27634901e-01 3.50581318e-01 -1.58171222e-01 1.48293316e+00 8.19087088e-01 4.13466424e-01 7.28906929e-01 -9.59783569e-02 8.03715527e-01 7.11026371e-01 -9.21964705e-01 2.49549584e-03 2.02020824e-01 5.59820652e-01 8.15770775e-02 4.05211300e-01 4.21108931e-01 -7.54786670e-01 -7.80736268e-01 5.18999159e-01 1.17742352e-01 -3.52268428e-01 3.04461662e-02 -8.50030184e-01 1.01114619e+00 4.17512923e-01 6.24404192e-01 -4.55989450e-01 1.93741724e-01 2.71123648e-01 5.96583605e-01 5.31413734e-01 9.82308447e-01 -3.35139781e-01 1.71493933e-01 -1.06666446e+00 6.12314284e-01 6.74094260e-01 9.68066216e-01 1.03407371e+00 -3.19927096e-01 -1.00856173e+00 1.11624420e+00 5.50492108e-01 8.11925828e-01 9.55084801e-01 -7.57024229e-01 2.95212716e-01 6.00487411e-01 2.26411775e-01 -9.55931187e-01 -2.48226225e-02 -4.84662384e-01 -1.75623342e-01 -5.64579189e-01 -2.59508818e-01 3.47259015e-01 -8.33065510e-01 1.56283319e+00 -1.34808138e-01 -5.90860546e-01 -5.10731973e-02 9.14735377e-01 9.61068332e-01 7.01360881e-01 8.25432912e-02 -3.06906074e-01 1.22559488e+00 -9.66743708e-01 -6.37430847e-01 -6.11842394e-01 9.07524109e-01 -1.04053223e+00 1.44567668e+00 -1.49320960e-02 -8.71443510e-01 -4.94449288e-01 -8.18757713e-01 -1.24327786e-01 -2.32206181e-01 1.61889300e-01 7.83346474e-01 5.60416400e-01 -1.65048122e+00 3.33495736e-01 -1.44429818e-01 -3.30567688e-01 -1.47292122e-01 3.60905558e-01 -2.02166557e-01 -5.04821062e-01 -1.63894033e+00 7.36083508e-01 3.43498945e-01 -3.21153820e-01 -8.26666176e-01 -8.87135506e-01 -6.15927160e-01 2.55515248e-01 3.78964186e-01 -9.53148842e-01 1.46935809e+00 -4.42112595e-01 -9.89744127e-01 5.45818448e-01 -4.53211904e-01 -4.49130535e-01 -8.15130100e-02 -8.78897429e-01 -1.99784636e-01 1.48972481e-01 3.04859459e-01 7.45826423e-01 8.06258976e-01 -1.06537604e+00 -4.35110956e-01 -2.44315386e-01 1.82525992e-01 6.40121639e-01 -4.48300451e-01 7.35349432e-02 -6.04052842e-01 -5.45569360e-01 -6.19103154e-03 -6.67970181e-01 -2.43800282e-01 -7.87696838e-01 -1.43768087e-01 -6.38249218e-01 6.69442177e-01 -3.00914377e-01 1.86785662e+00 -1.94148409e+00 -3.95248622e-01 4.72939461e-01 2.75084049e-01 2.07639918e-01 -5.67889035e-01 9.96229589e-01 9.75958705e-02 4.28957194e-01 3.00978452e-01 8.80077202e-03 7.88016990e-02 1.33368773e-02 -9.79426622e-01 -2.01608494e-01 4.19063009e-02 1.46576023e+00 -1.15640736e+00 -3.55773598e-01 -2.00996161e-01 2.83912301e-01 -7.93041885e-01 6.43327460e-02 -2.86587536e-01 -1.90330267e-01 -4.50659126e-01 5.59447706e-01 -3.38552706e-02 -5.33880293e-01 1.76294819e-01 2.08478525e-01 4.56525475e-01 7.87503719e-01 -7.38051713e-01 1.67794001e+00 -4.91560727e-01 3.48848343e-01 -4.82149512e-01 -4.08701092e-01 9.57817018e-01 2.25006223e-01 3.30444723e-01 -1.28810072e+00 -3.91596466e-01 5.50617695e-01 -3.40701103e-01 4.15362753e-02 1.29692912e+00 8.27446207e-02 -1.15180582e-01 7.43759811e-01 1.00789100e-01 -1.76478460e-01 6.16492212e-01 9.83216584e-01 1.30793417e+00 -1.66659132e-01 2.98256844e-01 -2.02255890e-01 3.58465344e-01 -1.62739351e-01 -1.11056574e-01 1.52327251e+00 3.39517146e-01 5.44880331e-01 -1.19866990e-01 6.46404400e-02 -5.75619996e-01 -8.64646435e-01 1.66593969e-01 1.81576252e+00 2.70515680e-01 -9.56969738e-01 -6.60621524e-02 -7.38763332e-01 2.26776823e-01 5.18120587e-01 -6.92068756e-01 -4.99692768e-01 -4.90237832e-01 -6.54280126e-01 3.04884046e-01 4.97053236e-01 -1.80488288e-01 -9.59523499e-01 -8.39323327e-02 1.36981413e-01 -3.15593421e-01 -5.09803474e-01 -8.41981351e-01 2.59511799e-01 -1.05736244e+00 -6.86021268e-01 -1.04725993e+00 -4.08468962e-01 3.33830327e-01 8.33314180e-01 1.74492693e+00 4.38596100e-01 1.99511107e-02 8.62635434e-01 -8.25647354e-01 3.54958363e-02 -2.03560710e-01 4.96231079e-01 -2.36610129e-01 -6.52062833e-01 5.50954998e-01 -2.30129436e-01 -7.59666622e-01 1.56830028e-01 -1.02038145e+00 -7.06960022e-01 8.67223322e-01 1.08566713e+00 7.20031738e-01 -6.32729411e-01 9.54227626e-01 -1.00500309e+00 1.30960321e+00 -5.14657497e-01 -3.25165272e-01 4.91195172e-01 -1.41245711e+00 3.28047067e-01 -8.13561454e-02 -4.61187959e-01 -7.58703411e-01 -5.80318868e-01 -5.43834642e-03 -2.35760406e-01 4.47182745e-01 8.19200814e-01 4.88242388e-01 4.95356359e-02 1.30493367e+00 2.35588551e-01 -4.91249889e-01 -4.45628554e-01 5.80050349e-01 4.00044531e-01 1.96533859e-01 -5.70065022e-01 5.32863140e-01 5.63569367e-02 -3.35907340e-01 -3.75348121e-01 -9.21090782e-01 -1.29517257e+00 -5.22126071e-02 4.63234372e-02 3.10682748e-02 -9.76765811e-01 -9.93516222e-02 -5.44091821e-01 -7.83090889e-01 6.77720085e-02 -8.24886262e-01 5.08673429e-01 -1.55896679e-01 5.06940067e-01 -5.70429504e-01 -9.78928328e-01 -8.54771793e-01 -7.53330767e-01 1.76231325e+00 -1.64492745e-02 -5.07296264e-01 -8.46339643e-01 5.84745586e-01 7.05051050e-02 7.75509357e-01 -8.02646339e-01 9.20044422e-01 -1.13760281e+00 -4.79556471e-01 -4.98193175e-01 -2.15834737e-01 -1.10465825e-01 5.91571592e-02 -5.27540505e-01 -1.07406294e+00 -4.79584605e-01 -3.43306631e-01 -5.90254247e-01 1.35623777e+00 2.42074132e-01 4.28784519e-01 -4.23299700e-01 -5.34041166e-01 -1.36907443e-01 1.47362268e+00 9.17233452e-02 7.49080479e-01 2.59839624e-01 2.34866872e-01 4.69371945e-01 9.33112442e-01 2.81720072e-01 3.26747857e-02 7.81831086e-01 2.20910739e-02 1.42017275e-01 -4.58024353e-01 -6.50738835e-01 3.97689611e-01 5.40012896e-01 2.55102009e-01 -3.33034635e-01 -6.72763109e-01 5.41099370e-01 -1.68541002e+00 -9.36937332e-01 1.48621812e-01 2.47510171e+00 1.02625561e+00 2.39920840e-02 4.41934876e-02 1.19641803e-01 3.89684349e-01 -1.39144333e-02 -1.94756344e-01 -1.45558700e-01 -2.46479705e-01 5.90063393e-01 2.81102508e-01 8.81937206e-01 -7.38968194e-01 9.72620070e-01 7.44292974e+00 1.09680760e+00 -9.31894958e-01 1.33945001e-02 2.77024597e-01 -1.64407626e-01 -1.07814074e+00 2.09739164e-01 -1.23606634e+00 -6.03696406e-02 9.98909652e-01 -5.67826748e-01 3.42228323e-01 8.92113149e-01 -3.95805925e-01 -2.52747387e-01 -1.09054863e+00 7.83560574e-01 2.49697253e-01 -1.11679566e+00 6.06555879e-01 4.98799495e-02 8.43189418e-01 2.33728811e-01 1.62183121e-01 9.39609826e-01 3.66186261e-01 -8.95710170e-01 3.74826580e-01 4.90140349e-01 8.44574869e-01 -2.80112058e-01 8.81670296e-01 3.16892833e-01 -1.00815868e+00 -1.26336023e-01 -4.16534215e-01 2.76435707e-02 -1.15731932e-01 6.59042418e-01 -1.13245964e+00 7.96568394e-01 5.94310224e-01 4.23603088e-01 -9.46742594e-01 9.04068172e-01 -6.36683032e-02 5.82596719e-01 -4.66570854e-01 -1.95132986e-01 1.73682287e-01 2.91004658e-01 5.18837512e-01 1.30849731e+00 2.59605169e-01 -2.94174939e-01 3.01852465e-01 5.44874370e-01 -1.83488846e-01 6.13735497e-01 -7.20195293e-01 -2.41228715e-01 6.30480945e-01 9.58830059e-01 -4.39974159e-01 -5.72919309e-01 -1.51027456e-01 6.18349612e-01 1.39196143e-01 4.39714342e-01 -9.97409597e-02 -3.79192919e-01 5.38302353e-03 2.69146949e-01 1.16644822e-01 3.88085961e-01 1.15753800e-01 -1.17823470e+00 1.72723815e-01 -8.72160733e-01 5.75184286e-01 -7.18565166e-01 -1.25603497e+00 6.29360437e-01 2.63817877e-01 -9.63257611e-01 -1.01902437e+00 7.13968277e-02 2.25985229e-01 1.23822987e+00 -1.64075744e+00 -8.92687619e-01 9.42683313e-03 3.62564176e-01 5.61919630e-01 -1.19461166e-02 1.00800347e+00 3.72037411e-01 3.11428398e-01 7.10602045e-01 1.46984264e-01 -3.48388195e-01 1.03872466e+00 -1.23346293e+00 3.06143373e-01 5.30057251e-01 7.43252754e-01 1.31281936e+00 3.08773369e-01 -8.93850744e-01 -1.22946537e+00 -6.40859663e-01 1.31510067e+00 -5.93282640e-01 3.73047203e-01 -6.70743212e-02 -9.84981835e-01 4.09512639e-01 1.49401575e-01 -2.25704536e-01 6.92672431e-01 6.98506057e-01 -8.49432647e-01 1.66471861e-03 -7.05723286e-01 3.64050061e-01 9.13327515e-01 -9.19212520e-01 -9.56199169e-01 4.53909248e-01 9.60574925e-01 -8.60751569e-02 -5.20091712e-01 7.39329517e-01 6.27059937e-01 -6.91924751e-01 1.19342148e+00 -4.70715195e-01 1.64680898e-01 -1.30140975e-01 -6.25250995e-01 -1.06597292e+00 -4.76417929e-01 -6.37353659e-01 -3.61664087e-01 8.48761499e-01 6.15784585e-01 -5.12378454e-01 5.43708086e-01 3.15900594e-01 -8.90097544e-02 -7.42158890e-01 -4.95064020e-01 -8.31381977e-01 9.63844657e-02 -2.78748900e-01 3.59041423e-01 4.27787751e-01 1.48137197e-01 8.33339751e-01 -1.67925373e-01 -4.46722597e-01 -1.06945634e-01 1.71331048e-01 6.26328409e-01 -1.23219097e+00 -5.46686828e-01 -5.41946709e-01 3.42953317e-02 -1.54282558e+00 -1.86361805e-01 -1.12089980e+00 -1.36094855e-03 -1.64827979e+00 4.36145425e-01 -4.12310928e-01 -4.63745266e-01 3.96037817e-01 -6.11344159e-01 5.63397467e-01 8.58743265e-02 7.48394728e-01 -8.16138089e-01 5.57989240e-01 1.27628851e+00 -3.58512968e-01 -3.31787795e-01 -2.10271046e-01 -1.18825543e+00 -2.58353967e-02 2.31383100e-01 -4.14926201e-01 -8.51501167e-01 -3.12739491e-01 8.21070969e-01 4.58545499e-02 1.26200125e-01 -4.17703390e-01 1.90308213e-01 1.29049703e-01 2.21697643e-01 -6.43883646e-01 2.98993707e-01 -4.33726907e-01 -8.82765651e-02 4.30764914e-01 -8.80722702e-01 2.37183407e-01 1.29946455e-01 7.35297859e-01 -6.09662592e-01 -4.77090746e-01 -8.01445823e-03 -1.68103382e-01 -6.74417734e-01 -9.15254131e-02 -1.09809153e-01 3.66975576e-01 6.99019209e-02 1.22031838e-01 -4.89150405e-01 -7.67447054e-01 -4.73716050e-01 1.13278069e-01 1.72131538e-01 4.85693455e-01 7.77366459e-01 -1.33235431e+00 -6.92570865e-01 1.57566980e-01 6.05205536e-01 -4.50306505e-01 -2.92428080e-02 5.27144432e-01 1.94993779e-01 1.26642454e+00 6.22057617e-01 -5.29061139e-01 -1.16892493e+00 6.73568547e-01 -1.17724985e-01 -1.07674718e+00 -3.09939295e-01 9.67195094e-01 3.79807860e-01 -3.78684253e-01 1.48023590e-01 -2.13055491e-01 -2.77949810e-01 1.31391868e-01 7.85492182e-01 2.32120324e-02 6.01726472e-01 -1.81394473e-01 -2.79548913e-01 3.06765556e-01 -6.17891014e-01 -6.40980065e-01 7.36666977e-01 -2.73254484e-01 -3.95478271e-02 3.65642875e-01 1.14070284e+00 5.11131346e-01 -1.50855944e-01 -6.08179748e-01 3.46522421e-01 -6.01558030e-01 1.76958725e-01 -1.23553884e+00 -5.18033683e-01 5.16875207e-01 4.58665133e-01 2.99350023e-01 1.19196343e+00 2.90092468e-01 5.56937456e-01 8.75387251e-01 6.64842904e-01 -6.96208775e-01 3.98758680e-01 7.68056631e-01 1.18570864e+00 -1.12853396e+00 1.15485705e-01 -1.25958249e-01 -2.93712109e-01 6.31408751e-01 4.31118965e-01 -1.37881979e-01 5.21122754e-01 -1.66454241e-01 1.04672797e-02 -3.52243513e-01 -1.05393946e+00 -4.10328746e-01 9.63379323e-01 2.96030939e-01 9.01327431e-01 -3.12336683e-01 -7.37559497e-01 3.53572130e-01 -2.96933621e-01 -9.43753868e-03 -1.98145822e-01 1.07554317e+00 -5.97355425e-01 -1.42330015e+00 -2.77015239e-01 8.74241412e-01 -4.97107744e-01 -7.85712421e-01 -7.36618996e-01 7.13073194e-01 -7.53406465e-01 8.82237732e-01 -2.29691267e-01 -4.45400983e-01 1.14517808e-01 2.72358924e-01 4.30617899e-01 -1.12907469e+00 -1.07729733e+00 5.81694663e-01 4.86087538e-02 -5.94063699e-01 -3.47061425e-01 -3.34668547e-01 -8.08355629e-01 2.91453183e-01 -9.20454502e-01 8.86853874e-01 3.28167468e-01 6.53662682e-01 5.96770287e-01 1.89388275e-01 5.58774054e-01 -4.75568622e-01 -9.81766045e-01 -1.45014060e+00 -3.67216349e-01 2.78756857e-01 2.29433730e-01 -5.62778831e-01 -4.76637453e-01 -3.89207512e-01]	[11.477747917175293, 7.613698482513428]
df6d4193-666b-4c99-9bb9-2773dd2b7061	self-supervised-deep-learning-to-enhance	2203.08812	null	https://arxiv.org/abs/2203.08812v1	https://arxiv.org/pdf/2203.08812v1.pdf	Self-Supervised Deep Learning to Enhance Breast Cancer Detection on Screening Mammography	A major limitation in applying deep learning to artificial intelligence (AI) systems is the scarcity of high-quality curated datasets. We investigate strong augmentation based self-supervised learning (SSL) techniques to address this problem. Using breast cancer detection as an example, we first identify a mammogram-specific transformation paradigm and then systematically compare four recent SSL methods representing a diversity of approaches. We develop a method to convert a pretrained model from making predictions on uniformly tiled patches to whole images, and an attention-based pooling method that improves the classification performance. We found that the best SSL model substantially outperformed the baseline supervised model. The best SSL model also improved the data efficiency of sample labeling by nearly 4-fold and was highly transferrable from one dataset to another. SSL represents a major breakthrough in computer vision and may help the AI for medical imaging field to shift away from supervised learning and dependency on scarce labels.	['Li Shen', 'Weiva Sieh', 'Laurie R. Margolies', 'Albert X. Pu', 'Vignesh A. Arasu', 'John D. Miller']	2022-03-16	null	null	null	null	['breast-cancer-detection', 'breast-cancer-detection']	['knowledge-base', 'medical']	[ 8.92835259e-01 6.31099761e-01 -4.08803046e-01 -5.98108172e-01 -1.14908242e+00 -1.72691971e-01 6.74786866e-01 1.55285522e-02 -4.89908248e-01 6.71703815e-01 1.85814649e-01 -2.17752740e-01 -8.94760992e-03 -6.35193408e-01 -9.20974433e-01 -8.40601563e-01 1.89615488e-01 5.18151104e-01 1.99176371e-01 8.89140368e-02 -3.48278843e-02 6.06183529e-01 -1.04136276e+00 6.36872768e-01 5.62678039e-01 1.22568429e+00 1.89124674e-01 4.73736346e-01 -1.47482768e-01 9.65210736e-01 -3.86242062e-01 -1.27457455e-01 1.80414602e-01 -4.71338362e-01 -1.03852701e+00 2.42994040e-01 6.00097358e-01 2.39466410e-02 8.58265255e-03 9.03609455e-01 5.09061396e-01 -4.01410878e-01 8.05667579e-01 -9.38520074e-01 -8.30830157e-01 4.95097458e-01 -6.19774461e-01 2.33740553e-01 -3.02030742e-01 -3.01746815e-03 7.58519113e-01 -7.31151879e-01 7.99898148e-01 9.79226828e-01 9.34133291e-01 7.70679116e-01 -1.42027509e+00 -4.45463687e-01 -6.02523610e-02 1.35621149e-02 -1.12805283e+00 -4.41250056e-01 8.17117453e-01 -5.51471293e-01 6.49781048e-01 2.65830278e-01 5.41060865e-01 1.06647193e+00 3.64337534e-01 1.03322804e+00 1.39412153e+00 -7.60257423e-01 3.50679308e-01 2.82518715e-01 2.89773881e-01 9.83571351e-01 3.43474150e-01 -9.41508356e-03 -4.12213176e-01 -1.13177441e-01 7.74395823e-01 -1.52239976e-02 -3.78877446e-02 -5.50285876e-01 -1.50037467e+00 8.02550614e-01 8.36938620e-01 4.88903701e-01 -2.61211276e-01 8.37998465e-02 2.68473774e-01 2.38211639e-02 7.89189458e-01 1.01821351e+00 -4.84661192e-01 5.53441167e-01 -9.59819436e-01 -7.99247772e-02 5.16941845e-01 5.91352761e-01 6.95168614e-01 -1.37439370e-01 -4.05099124e-01 8.84199500e-01 -8.96369740e-02 3.87022406e-01 6.27178013e-01 -9.14918423e-01 -3.88858058e-02 7.81733155e-01 -1.83621079e-01 -8.77507746e-01 -7.31554389e-01 -7.41881847e-01 -9.75651205e-01 1.43478900e-01 3.04476798e-01 -3.33732516e-02 -1.13519621e+00 1.56606579e+00 1.31207764e-01 -9.07767098e-03 9.27318782e-02 5.94384432e-01 8.91824961e-01 2.97306657e-01 2.33445838e-01 -3.58135439e-02 1.25397909e+00 -1.14602005e+00 -6.37457013e-01 -4.63915378e-01 1.08852232e+00 -2.95095861e-01 9.42218006e-01 2.77373403e-01 -8.41162920e-01 -5.10458529e-01 -1.13717663e+00 -4.58329357e-02 -4.71339971e-01 3.17044914e-01 8.71810913e-01 5.91753006e-01 -1.10996628e+00 4.09647882e-01 -1.01088715e+00 -5.62042892e-01 1.27887523e+00 4.37140733e-01 -5.70754290e-01 -2.79817164e-01 -8.00519764e-01 9.94267523e-01 2.18645915e-01 -3.45559567e-02 -7.47326255e-01 -9.18930829e-01 -8.52429748e-01 -1.51591480e-01 1.60472944e-01 -6.71131551e-01 1.22308421e+00 -1.35393071e+00 -1.26469135e+00 1.37842953e+00 -1.14790715e-01 -8.63116324e-01 3.44356805e-01 -2.81171547e-03 -3.64413075e-02 1.29614457e-01 1.80010602e-01 1.14138949e+00 6.94187701e-01 -1.28339672e+00 -4.73752886e-01 -3.71266007e-01 -5.04079342e-01 4.60417792e-02 -6.38522804e-01 -1.83141366e-01 -2.26336583e-01 -6.85018659e-01 1.19268045e-01 -9.62529242e-01 -5.24892807e-01 3.42521787e-01 -4.02390301e-01 -1.25676081e-01 6.18076384e-01 -4.17219698e-01 7.40910769e-01 -2.01728916e+00 1.31846040e-01 -1.05447046e-01 3.44143361e-01 3.13569754e-01 -2.06755787e-01 -1.40825659e-01 -1.52880207e-01 7.74425939e-02 -6.34218514e-01 -3.01357508e-01 -3.72045934e-01 3.21211964e-01 -5.44919260e-02 4.55093592e-01 6.97117507e-01 1.51928759e+00 -8.64005089e-01 -7.20723629e-01 5.40851578e-02 3.20551634e-01 -4.81000870e-01 1.68423101e-01 -2.01136023e-01 5.41322649e-01 -3.12712491e-01 7.36808240e-01 4.91556466e-01 -7.15158403e-01 1.33206490e-02 -4.30357754e-01 2.30447143e-01 -6.03596196e-02 -3.29140753e-01 1.93115890e+00 -4.52530116e-01 6.21638358e-01 -2.18311012e-01 -1.43839943e+00 9.42431629e-01 1.70189902e-01 8.07157636e-01 -7.47229159e-01 1.78507075e-01 2.54502743e-01 1.43780693e-01 -7.03935325e-01 -1.22837991e-01 -2.36048251e-01 -1.07822761e-01 4.14289862e-01 4.11333859e-01 -9.27328691e-02 -2.27660537e-01 -1.09765828e-01 1.33358562e+00 7.51986057e-02 5.09653926e-01 -6.86639726e-01 2.62524545e-01 4.64071810e-01 5.75126588e-01 8.20358276e-01 -3.60940278e-01 7.37666428e-01 3.06655169e-01 -7.33665943e-01 -9.37916338e-01 -8.57795298e-01 -4.88541305e-01 1.03382695e+00 -1.86171234e-01 1.24369629e-01 -9.16673541e-01 -1.00566578e+00 -4.54387581e-03 3.48367870e-01 -1.08402467e+00 -3.82613361e-01 -5.96070647e-01 -1.31843591e+00 4.71503407e-01 7.47773170e-01 7.77800798e-01 -1.20947123e+00 -5.78520656e-01 1.18794940e-01 -9.03390571e-02 -1.12420213e+00 -1.13075629e-01 5.76497495e-01 -8.66319358e-01 -8.40547264e-01 -1.00976467e+00 -1.11961257e+00 1.07201004e+00 -4.39255647e-02 1.13259566e+00 -3.72277685e-02 -7.54846573e-01 4.26660776e-02 -2.79453278e-01 -9.22656178e-01 -4.66757029e-01 4.35147017e-01 -2.15596765e-01 2.00676262e-01 5.57286620e-01 -8.54969844e-02 -5.34319758e-01 1.48237780e-01 -8.88284445e-01 3.02930444e-01 8.85533392e-01 1.14786649e+00 6.97653651e-01 -3.34390461e-01 8.29995334e-01 -1.45119083e+00 1.89645752e-01 -5.23893714e-01 -1.69727907e-01 3.46364200e-01 -4.73110586e-01 2.09334865e-01 3.37186545e-01 -4.05004174e-01 -9.94997382e-01 5.03066897e-01 -5.81358001e-02 -1.71330467e-01 -3.18074137e-01 4.82230902e-01 -3.87214050e-02 -2.70319492e-01 1.04132712e+00 6.82240203e-02 1.80405527e-01 -3.60130668e-01 1.44434676e-01 6.35962784e-01 7.28160799e-01 -3.20689231e-01 3.92934650e-01 8.71226728e-01 7.30838552e-02 -6.16973698e-01 -1.30914617e+00 -2.57015914e-01 -9.35851872e-01 4.26959172e-02 1.10452235e+00 -6.75645828e-01 -1.16048880e-01 3.99785399e-01 -9.29188311e-01 -7.16743410e-01 -6.81629717e-01 3.63870770e-01 -5.84927797e-01 -1.63026378e-02 -5.58691919e-01 -3.87782156e-01 -3.51832420e-01 -9.95182335e-01 1.16694701e+00 -1.79014411e-02 -2.99457520e-01 -1.06886053e+00 1.02442048e-01 4.64263260e-01 6.13004506e-01 3.63486052e-01 9.51865733e-01 -7.98748791e-01 -2.99215287e-01 -3.56602371e-01 -4.77317661e-01 3.57655346e-01 4.43716109e-01 -6.61538482e-01 -1.20491588e+00 -2.76381403e-01 3.93617190e-02 -6.16628766e-01 1.24109292e+00 6.00264072e-01 1.52674389e+00 -8.64280388e-02 -7.18533635e-01 8.18791926e-01 1.18353558e+00 1.58100784e-01 4.55771625e-01 4.63199645e-01 6.27573729e-01 5.88159978e-01 2.47084290e-01 -2.70053953e-01 1.22773491e-01 5.31594098e-01 1.06598571e-01 -9.78009820e-01 -6.51612222e-01 2.84101684e-02 -1.07370831e-01 7.09869444e-01 1.69643819e-01 5.74370883e-02 -1.06383348e+00 6.71440065e-01 -1.78801978e+00 -4.62719083e-01 3.04351032e-01 1.88092327e+00 9.67013538e-01 9.74525660e-02 -1.66443199e-01 4.50043418e-02 3.71740490e-01 -1.76111326e-01 -7.48576462e-01 -8.03349167e-02 -2.24624217e-01 4.31925565e-01 6.95337951e-01 2.30647251e-01 -1.49922168e+00 1.00376678e+00 7.41578054e+00 7.98877299e-01 -1.26356745e+00 3.03974807e-01 1.18297517e+00 2.00007111e-01 -5.36728799e-02 -5.55532336e-01 -5.87113380e-01 2.00564370e-01 8.78620982e-01 2.27554560e-01 -1.28967345e-01 8.48000646e-01 -2.91047305e-01 -6.43933052e-03 -1.37121677e+00 8.42100501e-01 3.90899420e-01 -1.65697563e+00 8.87983292e-02 5.80046996e-02 1.07044756e+00 2.34221205e-01 1.00332588e-01 7.20072910e-02 3.59754264e-01 -1.35926962e+00 1.31835580e-01 6.24163270e-01 1.04835021e+00 -3.33390176e-01 9.64203775e-01 1.67941958e-01 -7.00709641e-01 -1.33674070e-02 -3.66394520e-01 2.14938149e-01 -2.53668338e-01 5.37009001e-01 -1.07774878e+00 2.56495357e-01 8.13027620e-01 7.80514359e-01 -8.26611638e-01 1.00056076e+00 1.56787425e-01 7.88640738e-01 -2.16774032e-01 7.47144222e-02 1.99894950e-01 2.92526513e-01 1.10350758e-01 1.47325540e+00 -6.37151022e-03 -1.50539756e-01 1.89334974e-01 7.89586127e-01 -2.81522781e-01 1.36900321e-02 -6.40426815e-01 1.35320336e-01 5.62821440e-02 1.24963975e+00 -9.72168446e-01 -4.45826083e-01 -3.15519720e-01 9.34833229e-01 4.56302762e-01 2.36694813e-01 -5.09090245e-01 -1.71220437e-01 9.88860987e-03 1.61175892e-01 2.33889192e-01 2.59906203e-01 -6.00707829e-01 -9.62795794e-01 -1.34244189e-01 -7.20473945e-01 3.05487812e-01 -6.39612198e-01 -1.47430158e+00 8.73773098e-01 -1.15398385e-01 -9.79323924e-01 -2.86873858e-02 -7.90967464e-01 -2.19149813e-01 4.47709501e-01 -1.78074455e+00 -1.47216094e+00 -4.31705207e-01 2.41460338e-01 5.66859245e-01 -4.98876363e-01 1.23713052e+00 1.19757585e-01 -5.52702188e-01 6.84346676e-01 7.48434737e-02 3.11477691e-01 7.55495489e-01 -1.14886236e+00 2.77320683e-01 5.81360519e-01 2.29009330e-01 1.72000781e-01 1.95758656e-01 -3.96342069e-01 -1.02127588e+00 -1.42994094e+00 6.73623562e-01 -5.79549909e-01 4.77689952e-01 -3.86616945e-01 -9.32929337e-01 7.09272921e-01 1.06263630e-01 6.52767003e-01 8.01141202e-01 -1.52901143e-01 -2.82200813e-01 -3.25135857e-01 -1.19640827e+00 3.57644171e-01 9.36879456e-01 -3.89755905e-01 -4.83590305e-01 5.19687295e-01 5.76022983e-01 -1.70612782e-01 -7.75189638e-01 8.01301003e-01 4.70048994e-01 -5.54454267e-01 8.77148449e-01 -7.40840018e-01 5.85404515e-01 -4.70342953e-03 -8.26076865e-02 -1.21967053e+00 -5.72760761e-01 -2.89039642e-01 1.33485079e-01 7.56687999e-01 6.00216627e-01 -6.29571676e-01 1.10176671e+00 4.91598964e-01 -1.87054127e-01 -1.07802176e+00 -8.36765349e-01 -6.55371785e-01 4.64907110e-01 -1.63137078e-01 2.82622248e-01 1.09644067e+00 -2.60684434e-02 1.65007979e-01 -2.13805571e-01 -1.29459709e-01 6.75418913e-01 9.98633355e-02 4.78576183e-01 -1.30565941e+00 -2.97647804e-01 -3.59848112e-01 -5.42601883e-01 -7.13750243e-01 1.59492508e-01 -1.37921369e+00 1.83152288e-01 -1.53623593e+00 6.18613183e-01 -6.36863708e-01 -6.86167240e-01 8.82360339e-01 -1.34789690e-01 9.04208541e-01 -1.01431333e-01 2.73060203e-01 -5.73290586e-01 1.18461408e-01 1.24656880e+00 -4.55451727e-01 7.76655972e-02 -1.39822975e-01 -9.97886360e-01 7.48070478e-01 8.41770291e-01 -6.15179658e-01 -1.75451204e-01 -5.58276713e-01 -9.02806967e-02 -3.52329373e-01 3.80605251e-01 -1.07264984e+00 1.97179094e-01 5.40322363e-02 6.69574082e-01 -1.24193862e-01 9.66621488e-02 -7.36373901e-01 -2.79198527e-01 6.68998301e-01 -8.68732631e-01 -6.25550091e-01 2.95293838e-01 3.76237839e-01 -9.25246999e-02 -7.53455982e-02 9.72592473e-01 -3.51587951e-01 -5.22796333e-01 1.83793202e-01 -2.92594403e-01 -2.01082632e-01 1.08394694e+00 -1.51872575e-01 -2.33059213e-01 1.40350118e-01 -7.28917062e-01 -1.95474084e-02 1.11916654e-01 2.11173534e-01 3.96073192e-01 -1.34696281e+00 -9.03831065e-01 4.56741244e-01 3.96684736e-01 1.71658471e-01 -2.32482031e-02 8.46505523e-01 -4.85421687e-01 6.08919919e-01 -3.11480761e-01 -8.54027510e-01 -1.04004192e+00 5.08897185e-01 5.29265165e-01 -4.93301243e-01 -6.70552969e-01 1.02693856e+00 4.30637568e-01 -6.07040226e-01 2.53206819e-01 -4.26218450e-01 -6.51524439e-02 -2.79635519e-01 6.12561226e-01 -1.23400487e-01 3.52990955e-01 -4.38853830e-01 -3.16765875e-01 5.96752882e-01 -2.84452707e-01 1.80073380e-01 1.47024953e+00 3.08528870e-01 1.77988578e-02 5.68746328e-01 1.37484586e+00 -5.89288414e-01 -1.23394442e+00 -5.42520642e-01 2.20493317e-01 -1.70912892e-01 3.53983432e-01 -9.89124656e-01 -1.18997371e+00 7.86484003e-01 7.58305252e-01 -1.00255199e-02 1.15056038e+00 3.77364308e-01 5.60465813e-01 6.70002699e-01 3.52898210e-01 -8.35685492e-01 2.52682418e-01 9.05709788e-02 8.99304450e-01 -1.65862620e+00 1.34049520e-01 -3.28475863e-01 -6.12596273e-01 9.46110606e-01 5.45252860e-01 -3.41250211e-01 8.24856341e-01 6.60143495e-01 2.37699166e-01 -3.85482937e-01 -5.64052761e-01 -1.88855737e-01 5.05817115e-01 7.65936434e-01 5.40270984e-01 -9.36458781e-02 -6.31905645e-02 6.05023086e-01 -3.13791260e-02 4.03014183e-01 1.01916492e-01 9.93434489e-01 -2.28614569e-01 -9.65454280e-01 -2.35567883e-01 8.10320735e-01 -5.01412988e-01 2.04317085e-02 -4.48345751e-01 6.93722844e-01 2.94439852e-01 5.07480562e-01 4.22637045e-01 -2.16283455e-01 1.86827686e-02 1.75470889e-01 5.97053587e-01 -8.10201526e-01 -5.26563704e-01 1.46639779e-01 -2.33175755e-01 -3.04212034e-01 -8.22433710e-01 -4.27617878e-01 -1.04886472e+00 4.39883590e-01 -3.64465564e-01 -2.64584303e-01 4.95942563e-01 9.69244063e-01 4.59889621e-01 7.33398974e-01 4.89929855e-01 -8.82010341e-01 -1.71290502e-01 -1.00905418e+00 -3.43626320e-01 4.97461110e-01 4.23707336e-01 -5.19776165e-01 -6.38540238e-02 4.30507630e-01]	[14.950340270996094, -2.4859132766723633]
55adb12f-d969-4e30-9dd1-d57a9cdf6d2c	sim-to-real-segmentation-in-robot-assisted	2305.11686	null	https://arxiv.org/abs/2305.11686v2	https://arxiv.org/pdf/2305.11686v2.pdf	Domain Adaptive Sim-to-Real Segmentation of Oropharyngeal Organs Towards Robot-assisted Intubation	Robotic-assisted tracheal intubation requires the robot to distinguish anatomical features like an experienced physician using deep-learning techniques. However, real datasets of oropharyngeal organs are limited due to patient privacy issues, making it challenging to train deep-learning models for accurate image segmentation. We hereby consider generating a new data modality through a virtual environment to assist the training process. Specifically, this work introduces a virtual dataset generated by the Simulation Open Framework Architecture (SOFA) framework to overcome the limited availability of actual endoscopic images. We also propose a domain adaptive Sim-to-Real method for oropharyngeal organ image segmentation, which employs an image blending strategy called IoU-Ranking Blend (IRB) and style-transfer techniques to address discrepancies between datasets. Experimental results demonstrate the superior performance of the proposed approach with domain adaptive models, improving segmentation accuracy and training stability. In the practical application, the trained segmentation model holds great promise for robot-assisted intubation surgery and intelligent surgical navigation.	['Hongliang Ren', 'Long Bai', 'JIEWEN LAI', 'Tian-Ao Ren', 'Guankun Wang']	2023-05-19	null	null	null	null	['style-transfer']	['computer-vision']	[-1.48899823e-01 6.05320871e-01 -2.31517747e-01 -2.64270484e-01 -8.78091693e-01 -8.15106571e-01 2.46121988e-01 -1.06493466e-01 -6.32921338e-01 4.17367607e-01 -9.54354480e-02 -8.04964185e-01 -9.61719230e-02 -4.22114730e-01 -7.34595060e-01 -6.25998616e-01 2.55025089e-01 5.88561118e-01 7.69451186e-02 -1.61092520e-01 -1.07291684e-01 5.37845492e-01 -1.29418445e+00 1.90336108e-01 1.29896593e+00 8.78732324e-01 6.04321718e-01 6.03603542e-01 -2.75946736e-01 4.98215318e-01 -3.21795702e-01 -2.24546134e-01 6.80147648e-01 -2.80071676e-01 -9.86085415e-01 -1.20569915e-01 2.16517031e-01 -5.78788459e-01 -9.76815745e-02 9.81550574e-01 7.23135829e-01 6.17704801e-02 4.63942021e-01 -9.30769682e-01 -1.17280833e-01 3.22561473e-01 -7.61421323e-02 -2.09481210e-01 2.95486420e-01 1.37994006e-01 3.87689024e-01 -4.57666069e-01 1.07939208e+00 8.46649826e-01 7.77390778e-01 9.03735995e-01 -7.64599502e-01 -4.62529987e-01 -2.05907330e-01 -2.44768903e-01 -8.22395623e-01 2.57353425e-01 5.98783851e-01 -5.94098330e-01 4.22750741e-01 2.20781073e-01 7.17822075e-01 1.01861095e+00 1.49150088e-01 9.34519291e-01 9.45221126e-01 -2.62952238e-01 1.58345029e-01 4.90402102e-01 -1.22810736e-01 9.14137602e-01 1.47966057e-01 3.37252229e-01 1.45295039e-01 1.34380080e-03 9.53403890e-01 1.72954082e-01 -5.00428379e-01 -9.58200037e-01 -1.50190306e+00 6.63241684e-01 9.66804266e-01 2.26480260e-01 -5.25191247e-01 -2.90480316e-01 7.78286636e-01 1.93058372e-01 -1.36771783e-01 5.87350488e-01 -5.69539666e-01 -5.66966943e-02 -5.00870109e-01 -1.63679197e-01 9.41027164e-01 1.31826758e+00 3.62459034e-01 -2.42922127e-01 -2.89513618e-02 9.11332905e-01 3.39968204e-01 1.70538481e-02 9.53128755e-01 -1.01450515e+00 2.89512217e-01 6.05480254e-01 1.01363016e-02 -1.51375338e-01 -6.88063860e-01 -2.16694042e-01 -6.64955616e-01 3.04944515e-01 5.05525112e-01 -2.33359814e-01 -1.24906754e+00 1.28659761e+00 6.81380570e-01 1.67604893e-01 4.89564687e-01 1.17181969e+00 1.29547310e+00 1.60536855e-01 3.53350282e-01 5.03323190e-02 1.37811661e+00 -1.25002718e+00 -5.85438848e-01 1.31068513e-01 1.09935033e+00 -7.96828806e-01 1.06458426e+00 2.21091166e-01 -9.63605583e-01 -6.56372547e-01 -8.81587207e-01 -5.36068603e-02 -3.50843161e-01 1.80936947e-01 5.80667138e-01 6.40888095e-01 -9.06491756e-01 4.63226527e-01 -9.36302722e-01 -5.17465651e-01 4.69901651e-01 7.92289972e-01 -5.58203995e-01 3.70982662e-02 -8.22701812e-01 8.79684389e-01 6.64337635e-01 -1.38643563e-01 -7.51761138e-01 -7.08338082e-01 -1.09768462e+00 -3.30992341e-01 7.83832595e-02 -1.01246834e+00 1.54285359e+00 -9.38426197e-01 -1.84518778e+00 1.09361398e+00 3.31240237e-01 -5.44496894e-01 8.37594271e-01 -2.48615108e-02 6.44864067e-02 3.19565058e-01 -2.14415789e-01 9.25999880e-01 5.69309533e-01 -1.50423980e+00 -5.68215966e-01 -2.84353495e-01 2.84008831e-01 4.27332073e-01 -2.22862959e-01 -2.99421787e-01 -5.23948848e-01 -4.03356403e-01 1.43128112e-01 -1.36678421e+00 -7.44125664e-01 4.19896811e-01 -1.61159992e-01 1.62180141e-01 8.46563935e-01 -7.08577335e-01 5.33287823e-01 -2.18863487e+00 -1.36019373e-02 1.11118846e-01 7.53685534e-02 6.59085751e-01 -1.74352154e-01 -4.79013808e-02 -1.14725633e-02 -7.17612281e-02 -2.03360140e-01 -2.65500009e-01 -4.42173094e-01 5.85326076e-01 2.62091279e-01 3.30337435e-01 -3.52424115e-01 8.49166095e-01 -1.08901548e+00 -1.03291667e+00 7.04277098e-01 4.73680407e-01 -6.33577645e-01 6.12815201e-01 -8.31576884e-02 1.16675401e+00 -6.93736494e-01 5.26185811e-01 8.21917832e-01 7.32193068e-02 2.20149949e-01 -3.64976734e-01 -4.98729497e-02 8.20436999e-02 -6.98184788e-01 2.39771175e+00 -9.51487601e-01 4.56384830e-02 5.85820436e-01 -7.13255882e-01 9.19616401e-01 6.30266190e-01 8.03478301e-01 -6.30640030e-01 5.74063778e-01 5.22251606e-01 8.27078745e-02 -8.98279190e-01 3.61808002e-01 -1.33870453e-01 1.88837215e-01 -6.79585487e-02 2.78287500e-01 -7.38362134e-01 -1.97204754e-01 -4.02024128e-02 5.58013558e-01 3.52958947e-01 1.68677464e-01 -2.51606196e-01 5.86061120e-01 4.92477804e-01 4.01641071e-01 5.64357936e-01 -7.55179346e-01 9.54514503e-01 1.39088407e-02 -4.00455445e-01 -8.89715433e-01 -1.02221942e+00 -3.56908023e-01 5.89392066e-01 4.98023033e-01 2.44088337e-01 -1.09840834e+00 -1.18151689e+00 -1.04840659e-01 5.81965685e-01 -4.26687717e-01 -5.67359328e-02 -5.94557047e-01 -2.46236548e-01 3.39928001e-01 5.00279665e-01 3.25693578e-01 -1.13420630e+00 -6.91026688e-01 2.92718679e-01 -2.26053417e-01 -1.19498527e+00 -3.70663166e-01 2.38728195e-01 -1.23245931e+00 -1.26212835e+00 -1.05796826e+00 -1.42188668e+00 8.45093429e-01 2.77370781e-01 8.23072135e-01 -2.91755479e-02 -4.80193704e-01 6.14815533e-01 -4.90023822e-01 -3.92841995e-01 -1.13105142e+00 2.00811371e-01 -2.09432930e-01 -4.23251718e-01 -1.77360922e-01 -8.89883563e-02 -1.10640144e+00 6.19643509e-01 -9.22101498e-01 2.47104336e-02 6.74454749e-01 1.00468671e+00 6.87369108e-01 -6.16399050e-01 4.52241421e-01 -8.98283899e-01 6.29353225e-01 -4.84927922e-01 -5.41701734e-01 2.27105990e-01 -4.12568569e-01 -1.26681492e-01 7.73271143e-01 -4.76834416e-01 -1.14428437e+00 5.74708760e-01 -3.99715543e-01 -7.53938496e-01 -3.52355242e-01 4.07273680e-01 1.56731129e-01 -3.77983212e-01 8.13218892e-01 -3.85009460e-02 7.33599782e-01 -2.26357356e-01 3.44527960e-01 9.03976619e-01 6.85060620e-01 -4.74068761e-01 3.54296505e-01 4.14947510e-01 -1.56450182e-01 -4.97388870e-01 -3.67349923e-01 -7.93210924e-01 -9.63922262e-01 -2.28526086e-01 1.03861213e+00 -1.02527273e+00 -5.83112240e-01 1.06326133e-01 -9.95792985e-01 -4.72119808e-01 -3.67812157e-01 9.48064983e-01 -6.18125319e-01 5.46763420e-01 -6.37981057e-01 -3.61993343e-01 -6.11684799e-01 -1.80838931e+00 1.07660604e+00 3.21456671e-01 -1.58915743e-02 -1.21905136e+00 7.46823475e-02 5.77620208e-01 2.46171460e-01 3.47459644e-01 6.19492292e-01 -8.37866962e-01 -6.20542884e-01 -3.82487237e-01 -7.01194555e-02 4.68551368e-01 1.17997713e-01 -2.40403816e-01 -8.19845617e-01 -3.52521002e-01 1.53487131e-01 -4.00278926e-01 2.62576103e-01 4.83599633e-01 1.39281487e+00 6.53462633e-02 -3.93363863e-01 8.79959524e-01 1.11693418e+00 4.06338364e-01 5.42488456e-01 4.50551778e-01 7.76340425e-01 6.72778249e-01 9.98806834e-01 1.19613007e-01 4.71596748e-01 5.22081375e-01 7.11307347e-01 -5.68104625e-01 -2.00109497e-01 -1.42240852e-01 -2.15844691e-01 9.41055596e-01 -3.24273860e-05 1.02444954e-01 -8.86940658e-01 5.68430662e-01 -1.66132462e+00 -3.72629762e-01 -1.77158803e-01 2.21568203e+00 6.02875292e-01 -4.50633556e-01 -9.13666785e-02 -4.59799379e-01 6.00055099e-01 -4.04404640e-01 -5.84003210e-01 -6.18579447e-01 3.72019976e-01 -1.58525561e-03 6.30059063e-01 2.41640776e-01 -1.09106541e+00 7.20470488e-01 5.73250866e+00 6.70522690e-01 -1.39912844e+00 1.53824210e-01 4.17397678e-01 4.81379032e-01 -1.57236487e-01 -4.19275939e-01 -1.82198241e-01 1.80230409e-01 6.59320474e-01 2.52666593e-01 2.06977338e-01 1.27416551e+00 1.42190009e-01 1.83097228e-01 -1.02001274e+00 9.50946152e-01 -1.76584795e-01 -1.17508328e+00 -1.22861706e-01 6.30659889e-03 7.10693836e-01 3.98249060e-01 3.09982859e-02 3.22073072e-01 6.52432963e-02 -6.44232392e-01 3.16609949e-01 1.63947836e-01 8.88686895e-01 -2.53642887e-01 9.38860953e-01 3.72421652e-01 -9.04595137e-01 -1.03499196e-01 -1.91082746e-01 6.93896830e-01 1.33631993e-02 -1.85424179e-01 -1.68337405e+00 7.71468103e-01 6.23288393e-01 3.98965031e-01 -9.94214341e-02 1.38806248e+00 2.18419030e-01 2.07811847e-01 -3.26591462e-01 -3.57171260e-02 4.13008511e-01 -4.16433990e-01 5.08012414e-01 9.81550395e-01 4.46129739e-01 -9.18929130e-02 1.72417611e-01 5.19420743e-01 -1.69148475e-01 4.38058287e-01 -8.17271113e-01 3.78621608e-01 1.92377761e-01 1.48469782e+00 -7.39932358e-01 -1.55061036e-01 -4.79732305e-01 9.51258421e-01 2.04153210e-02 1.00130968e-01 -7.63991416e-01 3.61269899e-02 3.60875398e-01 -1.81812406e-01 4.89767902e-02 7.01467087e-03 -4.43760991e-01 -9.90189314e-01 -1.46668091e-01 -7.88571477e-01 4.33052719e-01 -7.76836991e-01 -8.67533267e-01 9.96419847e-01 -1.71552107e-01 -1.90484679e+00 -2.80737519e-01 -8.30723226e-01 -5.30388772e-01 6.04541183e-01 -1.50003850e+00 -1.30651546e+00 -6.97705150e-01 6.31296039e-01 7.45907664e-01 -1.26164600e-01 9.75825489e-01 2.01931506e-01 -3.27211618e-01 5.40829122e-01 4.48373824e-01 -6.12583607e-02 8.29081595e-01 -1.30513215e+00 -1.16115376e-01 2.58121252e-01 -3.01710904e-01 6.17525339e-01 4.98415440e-01 -4.39631045e-01 -1.44771612e+00 -1.22101116e+00 4.79836613e-02 -4.19516057e-01 2.56739855e-01 3.87892273e-04 -8.25651288e-01 6.53498828e-01 3.22427809e-01 2.03992471e-01 7.45310366e-01 -2.59353518e-01 1.29711345e-01 3.48080099e-02 -1.76290405e+00 6.43804252e-01 8.26667726e-01 -1.97374418e-01 -6.51972830e-01 5.62184691e-01 7.62923181e-01 -1.13311219e+00 -1.36465549e+00 7.08580911e-01 5.91116130e-01 -6.85380697e-01 1.14640582e+00 -2.60274917e-01 2.13182822e-01 -2.15141118e-01 4.13906932e-01 -1.26441514e+00 3.21431160e-01 -6.11135900e-01 3.19593281e-01 6.45885468e-01 4.95696574e-01 -8.16655755e-01 1.01548326e+00 9.25923467e-01 -7.30486512e-01 -8.32112491e-01 -9.82629061e-01 -5.71952879e-01 2.93657750e-01 -1.02092803e-01 5.74175656e-01 1.03221881e+00 9.15856287e-02 -2.34252885e-01 2.64420658e-01 1.63333341e-01 2.78479666e-01 -4.17704247e-02 9.61169362e-01 -9.91987228e-01 -1.07518189e-01 -3.24823856e-01 -2.27753475e-01 -1.01634407e+00 -1.06892316e-02 -1.08846915e+00 2.82701969e-01 -1.73122478e+00 -9.32943374e-02 -8.61044347e-01 -3.41190070e-01 1.27791420e-01 -3.76432911e-02 -5.13436422e-02 1.89839259e-01 4.29112703e-01 -4.06674743e-01 5.12046993e-01 2.08569646e+00 2.89152991e-02 -4.26628113e-01 3.73742491e-01 -3.15504909e-01 6.49476171e-01 8.09758365e-01 -2.64371336e-01 -5.85441411e-01 -3.74155313e-01 -6.88749671e-01 4.44981992e-01 2.70524204e-01 -9.29036379e-01 2.36192346e-01 3.45938772e-01 -5.67890443e-02 -5.63020766e-01 1.24222107e-01 -1.39516854e+00 1.22211250e-02 8.39286745e-01 -2.19083399e-01 -2.42404819e-01 4.68908340e-01 4.44840819e-01 -4.52582121e-01 -4.59589064e-01 9.06450510e-01 -3.07935983e-01 -5.37318230e-01 2.73567855e-01 -2.32097313e-01 -2.57849067e-01 1.14185846e+00 -3.82846713e-01 -3.54117937e-02 -1.33580148e-01 -8.56194079e-01 2.89054930e-01 5.43811858e-01 4.95957673e-01 6.60668135e-01 -7.49563932e-01 -3.27466190e-01 2.51985639e-01 3.65033269e-01 6.64569497e-01 5.18740952e-01 9.96310234e-01 -1.33441818e+00 3.15806985e-01 -2.45872304e-01 -9.38065112e-01 -1.22270191e+00 6.89237297e-01 7.36311138e-01 -2.55183131e-01 -7.10329354e-01 7.72587180e-01 4.11260068e-01 -1.32548022e+00 3.47592294e-01 -6.22278690e-01 -1.71903476e-01 -4.68936473e-01 -1.34366304e-01 3.99187058e-02 2.21963346e-01 -4.51487035e-01 -6.99692592e-02 5.05813777e-01 -1.54292211e-01 2.31634200e-01 8.56144249e-01 -1.77985489e-01 1.56327844e-01 -2.31169224e-01 1.28581488e+00 -3.73977423e-01 -1.33179796e+00 9.27393436e-02 -3.25838625e-01 -2.85119206e-01 -5.51022403e-02 -9.80626285e-01 -1.07939613e+00 8.89632881e-01 1.08056223e+00 -1.22207999e-01 1.07206285e+00 -1.34789899e-01 1.03566277e+00 4.13489342e-01 5.94697833e-01 -9.78457749e-01 -2.44504079e-01 3.16748142e-01 7.98187733e-01 -1.71890473e+00 -4.62217867e-01 -5.79273105e-01 -9.95579183e-01 1.15455055e+00 8.15792859e-01 4.82018664e-02 6.78511620e-01 4.23138514e-02 9.37979937e-01 6.37152195e-02 6.01286329e-02 -5.90087324e-02 2.63546884e-01 7.83841312e-01 4.06777859e-01 1.51739046e-01 -5.05184770e-01 3.06306571e-01 -1.04789868e-01 2.25932881e-01 3.25598210e-01 1.00684178e+00 7.11841956e-02 -1.31138504e+00 -3.01580489e-01 3.29517990e-01 -4.50006872e-01 1.68252140e-01 9.73010510e-02 1.07113612e+00 2.70469524e-02 6.10110700e-01 -1.54104486e-01 -1.31826058e-01 4.28863555e-01 -1.98592588e-01 4.39182311e-01 -4.57664579e-01 -1.13071012e+00 1.38202071e-01 -4.38853689e-02 -4.75280851e-01 -3.47203046e-01 -4.99703348e-01 -1.52853143e+00 4.03699309e-01 -3.99401367e-01 2.69762635e-01 1.29009438e+00 5.40795982e-01 4.99802649e-01 6.82147563e-01 7.78964877e-01 -1.03946340e+00 -4.74128395e-01 -7.74374366e-01 -2.68536568e-01 7.18594551e-01 2.35135898e-01 -5.02532363e-01 -1.68404102e-01 5.30658998e-02]	[14.484713554382324, -2.517747163772583]
88b2bd7c-8cf3-486a-99bc-8b65b2251a2d	cpnet-a-context-preserver-convolutional	1810.05778	null	http://arxiv.org/abs/1810.05778v1	http://arxiv.org/pdf/1810.05778v1.pdf	CPNet: A Context Preserver Convolutional Neural Network for Detecting Shadows in Single RGB Images	Automatic detection of shadow regions in an image is a difficult task due to the lack of prior information about the illumination source and the dynamic of the scene objects. To address this problem, in this paper, a deep-learning based segmentation method is proposed that identifies shadow regions at the pixel-level in a single RGB image. We exploit a novel Convolutional Neural Network (CNN) architecture to identify and extract shadow features in an end-to-end manner. This network preserves learned contexts during the training and observes the entire image to detect global and local shadow patterns simultaneously. The proposed method is evaluated on two publicly available datasets of SBU and UCF. We have improved the state-of-the-art Balanced Error Rate (BER) on these datasets by 22\% and 14\%, respectively.	['Parvaneh Saeedi', 'Sorour Mohajerani']	2018-10-13	null	null	null	null	['detecting-shadows']	['computer-vision']	[ 7.60092676e-01 -1.13569871e-01 1.17893338e-01 -6.58054888e-01 -6.81429267e-01 -3.78001571e-01 1.71901226e-01 -7.71851763e-02 -4.23358649e-01 6.63699687e-01 -2.24602550e-01 -2.67329663e-01 3.19223672e-01 -5.80376387e-01 -8.16729903e-01 -9.07195270e-01 1.35258496e-01 -2.34282643e-01 6.49674714e-01 5.07166758e-02 2.03362480e-01 5.61828613e-01 -1.36803067e+00 2.24397689e-01 7.96996713e-01 1.40053320e+00 4.17746454e-01 8.22766781e-01 2.41651118e-01 8.29135656e-01 -5.71526110e-01 3.19025777e-02 1.98770553e-01 -4.55985755e-01 -3.71857285e-01 2.11776257e-01 7.30811954e-01 -5.34346104e-01 -6.20222628e-01 7.95338035e-01 5.76726437e-01 -1.07068848e-03 3.55222642e-01 -8.18626702e-01 -1.42454460e-01 1.07209690e-01 -6.73621774e-01 3.44788313e-01 -1.79391891e-01 7.59933591e-02 6.35972977e-01 -7.42965043e-01 3.32387179e-01 7.84044921e-01 6.25089705e-01 1.36408418e-01 -1.10232019e+00 -7.93051720e-01 -5.29680140e-02 2.56904215e-01 -1.21447170e+00 -3.21601689e-01 9.10452008e-01 5.12534082e-02 5.07613122e-01 2.62785137e-01 4.30315942e-01 9.66598511e-01 1.98947251e-01 8.21901023e-01 1.67994106e+00 -5.80261111e-01 2.19952360e-01 -2.61397004e-01 2.97039419e-01 1.01443756e+00 3.92732285e-02 2.83616424e-01 -8.00034344e-01 2.53143638e-01 5.57866454e-01 -1.24055319e-01 -5.06803632e-01 -9.94982570e-02 -8.17308605e-01 4.08166945e-01 8.26557100e-01 2.33811513e-02 -5.34467362e-02 5.21251082e-01 6.48778081e-02 -1.68062016e-01 4.06513482e-01 5.56374304e-02 -5.08047044e-01 4.62959427e-03 -1.07097077e+00 -1.32886246e-01 5.52399814e-01 6.47343397e-01 9.00810540e-01 -9.82185081e-03 -3.65961790e-01 6.00452304e-01 2.65540689e-01 8.43317568e-01 -1.49157777e-01 -9.00952756e-01 9.72011238e-02 4.00452793e-01 7.82992244e-02 -8.89707804e-01 -4.69491661e-01 -5.79382837e-01 -6.52315795e-01 4.90832448e-01 3.50021452e-01 -8.17206576e-02 -1.40872025e+00 1.45497382e+00 3.53469223e-01 4.54787970e-01 -4.61518541e-02 9.65034485e-01 8.24301004e-01 7.01320767e-01 -2.86557317e-01 -1.68520883e-01 1.20250821e+00 -1.01178741e+00 -6.58479929e-01 -4.01488125e-01 -8.64292532e-02 -9.35167074e-01 1.04432619e+00 5.01177669e-01 -7.24885941e-01 -5.77928007e-01 -1.35514438e+00 -1.91951141e-01 -1.96029842e-01 5.30323565e-01 5.48150241e-01 8.61371338e-01 -6.42925680e-01 4.60775197e-01 -8.42771173e-01 -1.76677808e-01 7.61207879e-01 2.93310583e-01 2.89688736e-01 -2.36898676e-01 -7.55798578e-01 4.21361566e-01 2.15792015e-01 4.71150249e-01 -1.13758767e+00 -4.98246849e-01 -6.83735013e-01 2.23425314e-01 6.01439416e-01 7.86720775e-03 1.12962651e+00 -1.17582929e+00 -1.55706596e+00 5.49912512e-01 -3.38644624e-01 -4.70934033e-01 4.76951033e-01 -5.20625532e-01 -1.19169660e-01 1.19733207e-01 -1.90206558e-01 1.33022845e-01 8.55550289e-01 -1.63171458e+00 -6.66557431e-01 -3.16789120e-01 -5.43607958e-02 -3.32227089e-02 -1.37245148e-01 -4.02536094e-02 -8.99186671e-01 -5.97439826e-01 4.67857383e-02 -9.15114403e-01 5.80240078e-02 2.61873364e-01 -6.72284067e-01 2.14750320e-01 1.41879892e+00 -6.79866433e-01 8.98286700e-01 -2.15610385e+00 -4.03797954e-01 2.14348435e-01 7.53221065e-02 2.88381130e-01 1.32597566e-01 -1.12275332e-01 3.32084984e-01 -4.45744067e-01 -6.67298973e-01 -5.08034170e-01 -3.01218152e-01 2.95735806e-01 -2.83529639e-01 7.42532074e-01 -4.56021726e-02 5.89986801e-01 -6.55864000e-01 -4.01282072e-01 4.55664366e-01 6.31742418e-01 -8.76696929e-02 4.29727167e-01 -2.22850993e-01 5.02326846e-01 -1.27197266e-01 9.75351214e-01 9.70725536e-01 -2.60939542e-02 2.38602981e-01 -4.05894488e-01 -1.21168397e-01 4.20539465e-04 -1.10667026e+00 1.66641045e+00 -6.67839348e-01 1.29086196e+00 7.23520070e-02 -5.20201027e-01 7.04390466e-01 -4.29025702e-02 3.33816141e-01 -9.56473827e-01 2.91391551e-01 3.18714410e-01 -9.53040272e-02 -2.10907176e-01 4.13940698e-01 1.53320283e-01 -5.56994677e-02 2.58091867e-01 -1.64141178e-01 2.02812672e-01 -9.86016318e-02 -9.00315493e-02 1.09447718e+00 1.67659193e-01 -3.58454399e-02 -3.71921122e-01 2.41796613e-01 -2.32490823e-01 7.79942334e-01 8.98612142e-01 -2.48945117e-01 8.77324998e-01 2.38062099e-01 -3.06583673e-01 -7.08427548e-01 -1.25689089e+00 -3.03063273e-01 8.12610984e-01 5.46431243e-01 -9.56837535e-02 -7.17449129e-01 -7.04853594e-01 -8.67541283e-02 6.74401343e-01 -6.27317667e-01 1.76092442e-02 -7.16074646e-01 -8.67730677e-01 4.73620743e-01 6.93807244e-01 1.16795969e+00 -8.50213885e-01 -1.34510291e+00 -1.40051963e-02 -1.57628044e-01 -1.53461242e+00 -3.01666528e-01 6.91915095e-01 -3.99402618e-01 -1.27017283e+00 -4.01103973e-01 -6.14010632e-01 6.56058967e-01 3.92589301e-01 1.05356550e+00 -3.45631428e-02 -6.52261972e-01 -8.24319497e-02 -3.35153759e-01 -5.90364039e-01 -4.56654765e-02 -1.45542234e-01 -4.27048862e-01 2.43347406e-01 -1.79350629e-01 -3.97906840e-01 -1.08652055e+00 4.37246263e-01 -7.40536809e-01 2.34173581e-01 7.24261642e-01 8.55263412e-01 6.80816352e-01 1.81907430e-01 -3.99009511e-02 -1.00555158e+00 -5.87730184e-02 7.16117844e-02 -9.79075313e-01 1.75817862e-01 -5.80708981e-01 -1.66572124e-01 4.82027590e-01 1.90603174e-02 -1.41816425e+00 5.23126125e-01 2.11152211e-01 -3.08292389e-01 -2.69167066e-01 -8.49563852e-02 -3.48893613e-01 -4.05333996e-01 5.29082894e-01 2.70222604e-01 -5.67326128e-01 -3.53067368e-01 2.49222532e-01 6.59800231e-01 9.00954664e-01 -3.13857198e-01 7.81647623e-01 1.01851916e+00 1.72662005e-01 -9.58814442e-01 -1.11271727e+00 -5.47881842e-01 -7.26769328e-01 -4.66179371e-01 9.06004250e-01 -9.02205646e-01 -5.08439779e-01 8.24528277e-01 -1.02771997e+00 -9.87199128e-01 6.39933273e-02 -1.47853955e-03 -3.60060394e-01 3.72674316e-01 -2.43457183e-01 -8.33619773e-01 -4.65035528e-01 -1.08881032e+00 1.40201831e+00 6.20601237e-01 3.47729683e-01 -4.92072672e-01 -3.43514204e-01 3.69280726e-01 3.84485662e-01 6.32335782e-01 6.91193879e-01 7.99113289e-02 -1.04911828e+00 1.45889269e-02 -7.51376569e-01 5.67454934e-01 2.09633291e-01 -2.09047630e-01 -1.48659885e+00 -1.71761438e-01 -1.17778614e-01 -2.92673111e-01 1.30385900e+00 4.23503935e-01 1.64845634e+00 6.02158569e-02 -3.41128737e-01 8.77928555e-01 1.90610909e+00 1.39229789e-01 6.86812520e-01 7.39925653e-02 8.96452546e-01 3.09441179e-01 6.71909571e-01 4.74991173e-01 -1.72312021e-01 6.53011084e-01 4.93877351e-01 -5.54372251e-01 -6.17595553e-01 3.14393342e-01 2.00578898e-01 -1.21022284e-01 1.66576415e-01 -4.73944545e-01 -9.01484370e-01 5.45963049e-01 -1.65927994e+00 -4.38689798e-01 -2.75516272e-01 2.02117801e+00 7.34688282e-01 3.44958603e-01 -4.56095666e-01 1.60542503e-01 4.02726740e-01 4.33079958e-01 -7.40911424e-01 -9.64670926e-02 -4.17010635e-01 5.31823933e-01 1.10171938e+00 4.86435771e-01 -1.24173391e+00 1.09034586e+00 5.74899769e+00 8.40205073e-01 -1.43408048e+00 1.85790330e-01 1.03535748e+00 -9.47155356e-02 1.89927921e-01 -6.24829158e-02 -5.65566897e-01 4.61437851e-01 6.74003541e-01 7.04597950e-01 4.66358691e-01 5.76211929e-01 2.95993984e-01 -9.40684199e-01 -7.41815865e-01 9.58148241e-01 3.01388443e-01 -1.14300394e+00 -6.71692431e-01 -9.89764407e-02 9.04015839e-01 2.43707165e-01 9.93478671e-02 -1.70154095e-01 2.48594597e-01 -9.57530618e-01 8.02391410e-01 4.69900727e-01 9.15678799e-01 -9.35839176e-01 6.82547569e-01 -5.30509613e-02 -1.15366125e+00 -1.64834037e-01 -6.65245876e-02 6.60385117e-02 4.39137109e-02 9.77184832e-01 -8.30449581e-01 3.50979805e-01 1.01592755e+00 3.09524357e-01 -7.39477158e-01 9.75755274e-01 -5.72764575e-01 1.03620911e+00 -4.70779032e-01 2.13382632e-01 2.17498615e-01 9.24373865e-02 2.43920356e-01 1.32762933e+00 4.52761538e-02 9.86924022e-02 1.66960075e-01 7.90126681e-01 -2.83888817e-01 -4.83075738e-01 -1.72984257e-01 3.45413536e-01 1.02919333e-01 1.43269122e+00 -1.09477246e+00 -2.16772053e-02 -3.12524617e-01 1.48707652e+00 -1.74515359e-02 4.82356787e-01 -1.11431730e+00 -6.17000997e-01 4.25882876e-01 -1.42428666e-01 4.67295408e-01 -3.62975746e-01 -7.61769414e-01 -7.84744143e-01 1.45238966e-01 -4.96346712e-01 8.71062055e-02 -8.64546597e-01 -6.69383049e-01 4.53639746e-01 -2.32558846e-01 -7.53781319e-01 4.18406725e-01 -6.71994030e-01 -7.28766620e-01 6.94520295e-01 -2.17417550e+00 -1.29250157e+00 -1.06907332e+00 5.38078368e-01 5.30023754e-01 2.33929247e-01 5.19414723e-01 2.63081044e-01 -6.97057188e-01 4.70789135e-01 5.97565174e-01 4.16724920e-01 7.87370920e-01 -1.32425845e+00 2.61616319e-01 1.31736338e+00 4.38454002e-02 -8.97248834e-03 7.42678165e-01 -4.47292060e-01 -1.32151282e+00 -1.33047414e+00 4.10695076e-01 -1.19959563e-01 1.98080018e-01 -6.91427529e-01 -5.75805128e-01 1.97514266e-01 3.51284504e-01 4.61479604e-01 4.45824146e-01 -4.35968488e-02 -2.46326119e-01 -6.00757480e-01 -9.85416651e-01 4.30683285e-01 9.98488426e-01 -5.25845647e-01 1.44531786e-01 4.09696639e-01 4.78204668e-01 -8.47038746e-01 -3.63266319e-01 6.35667324e-01 5.76207519e-01 -1.33728409e+00 8.30974042e-01 2.22761810e-01 3.95790905e-01 -4.70421255e-01 -4.57222044e-01 -8.88469815e-01 1.42330006e-01 -6.62613988e-01 -2.66689688e-01 1.07176375e+00 1.30197078e-01 -3.27050775e-01 1.02075899e+00 2.35406056e-01 -3.68019164e-01 -1.02889812e+00 -1.01133978e+00 -5.60850143e-01 -5.38643241e-01 -5.07175148e-01 1.15950525e-01 1.62739694e-01 -8.91912401e-01 7.62959495e-02 -4.20759380e-01 6.47371650e-01 7.99325824e-01 7.50799656e-01 8.08587372e-01 -9.30840790e-01 -3.13595533e-01 1.81421861e-02 -1.14112079e-01 -8.80440772e-01 1.56650349e-01 -3.35281432e-01 6.40675843e-01 -1.41544914e+00 2.72331297e-01 -4.57216710e-01 -4.80582714e-01 4.43427056e-01 -2.82941997e-01 6.75787032e-01 1.72257215e-01 2.91892253e-02 -6.58378720e-01 6.57616138e-01 8.43943655e-01 -2.01150253e-01 -1.84650123e-02 -1.08526044e-01 -3.08967739e-01 6.81390166e-01 7.88767874e-01 -4.85052705e-01 -1.37521148e-01 -5.79117775e-01 -3.20753336e-01 -3.68528247e-01 6.19439065e-01 -1.50276375e+00 -4.41597849e-02 -3.98887955e-02 8.26875865e-01 -8.29926074e-01 5.41408420e-01 -8.03441763e-01 -5.33493161e-01 6.18561566e-01 -8.73871520e-03 -4.62949038e-01 3.58390331e-01 9.00874972e-01 1.12552922e-02 2.68177480e-01 1.25226927e+00 1.67453572e-01 -1.01233566e+00 1.36199355e-01 -1.13800570e-01 -2.58238316e-01 1.01767933e+00 -1.00556880e-01 -3.75627995e-01 -1.91768572e-01 -5.59400395e-02 8.22475180e-02 3.74596208e-01 -2.34059524e-02 6.82613015e-01 -8.66396487e-01 -3.46845120e-01 2.31073454e-01 1.39460981e-01 2.19782472e-01 2.05928609e-01 8.66607666e-01 -8.04196596e-01 2.15029970e-01 -1.17019631e-01 -6.63123012e-01 -1.60222614e+00 6.83975145e-02 6.07476711e-01 -1.14785060e-02 -6.30566537e-01 1.04078424e+00 2.13918597e-01 2.01119855e-02 5.27993917e-01 -5.67749619e-01 1.72525287e-01 -3.57354283e-01 3.57795447e-01 2.22481266e-01 1.23427942e-01 -5.99165022e-01 -1.53742060e-01 4.43235457e-01 2.33054981e-01 7.92037249e-02 1.17584646e+00 -2.11374208e-01 -7.06322026e-03 3.22744429e-01 1.32822752e+00 5.45078963e-02 -1.79564512e+00 -3.35322499e-01 -3.35727632e-01 -7.14255035e-01 5.96200287e-01 -1.18212080e+00 -1.31065238e+00 9.19984221e-01 1.18590260e+00 -3.07732552e-01 1.62184441e+00 -2.02118486e-01 1.25748229e+00 3.46946269e-01 4.77958731e-02 -1.17937624e+00 2.92829692e-01 4.80061203e-01 5.20261705e-01 -1.41310680e+00 1.59232453e-01 -6.71627402e-01 -2.16427177e-01 1.35496807e+00 6.47238493e-01 -6.02498837e-02 6.73927665e-01 4.07209575e-01 2.42837504e-01 -1.10988513e-01 -1.42121062e-01 -3.77815962e-01 2.33273566e-01 4.32480574e-01 5.78458048e-02 -8.74587614e-03 1.62543021e-02 2.60460943e-01 1.88762307e-01 -1.85291395e-01 4.26018089e-01 1.10369766e+00 -5.40899217e-01 -7.62130022e-01 -5.14633954e-01 2.80947357e-01 -4.47714627e-01 -1.01410121e-01 -5.52163661e-01 5.66379905e-01 5.04819751e-01 1.09254050e+00 -4.80206534e-02 -2.81577885e-01 3.86794955e-02 -3.68562251e-01 5.37484884e-01 -3.73833299e-01 -2.86583006e-01 2.42998451e-01 4.36529070e-02 -9.25427675e-01 -4.49159682e-01 -5.93945146e-01 -1.41232359e+00 1.22384317e-01 -3.62843752e-01 -4.70713109e-01 8.56744170e-01 1.02053761e+00 1.81497380e-01 8.89897943e-01 8.06683183e-01 -8.41559112e-01 -1.35753945e-01 -6.97251141e-01 -5.75932026e-01 1.18549101e-01 5.17221451e-01 -5.32602787e-01 1.05844047e-02 8.25050995e-02]	[10.851176261901855, -4.113837718963623]
9898cc34-707c-479b-a8e7-369f87e1d2d3	speaking-the-language-of-your-listener	2305.19933	null	https://arxiv.org/abs/2305.19933v1	https://arxiv.org/pdf/2305.19933v1.pdf	Speaking the Language of Your Listener: Audience-Aware Adaptation via Plug-and-Play Theory of Mind	Dialogue participants may have varying levels of knowledge about the topic under discussion. In such cases, it is essential for speakers to adapt their utterances by taking their audience into account. Yet, it is an open question how such adaptation can be modelled in computational agents. In this paper, we model a visually grounded referential game between a knowledgeable speaker and a listener with more limited visual and linguistic experience. Inspired by psycholinguistic theories, we endow our speaker with the ability to adapt its referring expressions via a simulation module that monitors the effectiveness of planned utterances from the listener's perspective. We propose an adaptation mechanism building on plug-and-play approaches to controlled language generation, where utterance generation is steered on the fly by the simulator without finetuning the speaker's underlying language model. Our results and analyses show that our approach is effective: the speaker's utterances become closer to the listener's domain of expertise, which leads to higher communicative success.	['Raquel Fernández', 'Sandro Pezzelle', 'Mario Giulianelli', "Nicolo' Brandizzi", 'Ece Takmaz']	2023-05-31	null	null	null	null	['open-question']	['natural-language-processing']	[-2.28703216e-01 1.11125624e+00 4.16961133e-01 -2.25626528e-01 -2.88822591e-01 -7.66988039e-01 8.50339472e-01 3.87217253e-01 -4.73120749e-01 5.06301939e-01 4.83137369e-01 -2.90973485e-01 2.53841937e-01 -8.85048509e-01 -4.29971308e-01 -2.32942656e-01 1.66168973e-01 6.13628924e-01 3.25176984e-01 -6.62373364e-01 2.16591716e-01 3.74554485e-01 -1.57429647e+00 2.92150199e-01 8.08221579e-01 1.35027081e-01 4.59358245e-01 9.72810566e-01 -2.97616124e-01 1.13591743e+00 -7.68373609e-01 -2.98379004e-01 -1.13215841e-01 -6.91993058e-01 -1.11416781e+00 3.38076234e-01 -2.79629707e-01 -1.40588731e-01 3.88291836e-01 8.96650076e-01 4.25867200e-01 3.18284869e-01 4.07517523e-01 -9.31080282e-01 -2.87446260e-01 1.02900255e+00 1.51521519e-01 -1.66216195e-01 1.00692284e+00 5.05067050e-01 7.06849217e-01 -4.99247164e-01 8.88145864e-01 1.59801900e+00 2.42213562e-01 9.99146879e-01 -1.13616407e+00 3.71556431e-02 5.86080432e-01 5.14376573e-02 -1.28979111e+00 -5.60128450e-01 8.13779235e-01 -4.88018185e-01 7.91260779e-01 3.53251696e-01 9.57215607e-01 9.22306836e-01 -1.90989494e-01 4.65869039e-01 9.26736772e-01 -1.09874892e+00 6.18273497e-01 1.08484399e+00 -2.74157912e-01 4.63037103e-01 -3.00468922e-01 1.58809289e-01 -7.79670119e-01 -5.40740490e-02 8.87022913e-01 -7.01049507e-01 -4.18854952e-01 -7.56662428e-01 -1.08023620e+00 9.26974833e-01 1.07778616e-01 6.22161388e-01 -6.72668159e-01 -2.88869012e-02 9.29476544e-02 6.25735164e-01 3.63977969e-01 7.04752684e-01 5.22317402e-02 -3.97048295e-01 -2.95305431e-01 3.85055691e-01 1.21814442e+00 6.98854327e-01 5.74166834e-01 7.53486203e-03 -2.58464009e-01 6.68750286e-01 7.78709292e-01 2.06975281e-01 3.24031472e-01 -1.01682186e+00 -1.37244329e-01 8.83932173e-01 5.14317453e-01 -7.10480988e-01 -3.66381794e-01 -1.15432069e-01 4.52095233e-02 8.10706913e-01 5.02011776e-01 -4.61202055e-01 -4.34652954e-01 1.98016322e+00 9.25543129e-01 -2.62774646e-01 4.43537444e-01 1.22124124e+00 7.90079057e-01 6.24768853e-01 4.22600567e-01 -2.63846070e-01 1.48435080e+00 -6.21424675e-01 -6.82674944e-01 -2.25192592e-01 7.13491678e-01 -4.88547623e-01 1.39805734e+00 1.35686263e-01 -1.39003515e+00 -3.32306415e-01 -7.22402334e-01 1.33641049e-01 -1.21994987e-01 -2.67547131e-01 2.75250763e-01 6.45018816e-01 -1.06596541e+00 1.91112369e-01 -5.82006216e-01 -9.05196905e-01 -2.38585368e-01 2.49107510e-01 -3.58486213e-02 5.28346479e-01 -1.31914699e+00 1.22632074e+00 3.00380260e-01 3.96749526e-02 -9.12715673e-01 -1.47321075e-01 -9.00784075e-01 7.27502480e-02 2.79769659e-01 -1.03396404e+00 1.85186768e+00 -1.55766618e+00 -2.40903115e+00 8.32729042e-01 2.90825009e-01 -2.50184774e-01 7.79261947e-01 2.62599647e-01 7.73634762e-02 1.68948993e-01 -1.22261398e-01 7.26984024e-01 5.30099332e-01 -1.60736609e+00 -5.17098010e-01 -2.59117037e-01 8.63848865e-01 8.80752206e-01 1.05078449e-04 1.73566341e-01 -1.29093215e-01 -2.99306452e-01 -2.23715112e-01 -7.11138248e-01 -3.99277478e-01 3.36443959e-03 -1.79893941e-01 -2.49166861e-01 4.36280549e-01 -1.26539081e-01 8.72254133e-01 -2.09561419e+00 4.20794874e-01 2.77267814e-01 1.83037564e-01 2.28976250e-01 -1.44436121e-01 8.59628260e-01 1.46779180e-01 -8.50903243e-02 1.04906544e-01 -4.30359542e-01 3.21400523e-01 4.38295654e-04 3.35015394e-02 1.49210259e-01 -1.28247783e-01 5.75753570e-01 -8.37301195e-01 -5.59827685e-01 3.41992170e-01 4.44723248e-01 -6.29421771e-01 8.10756385e-01 -6.29853368e-01 8.04362118e-01 -7.75078475e-01 -9.96886343e-02 1.73153490e-01 4.97427508e-02 4.13606524e-01 4.05408084e-01 -4.72521722e-01 5.20648479e-01 -1.20287681e+00 1.50682092e+00 -7.93337822e-01 3.45012039e-01 6.32244349e-01 -5.10433972e-01 7.15574265e-01 7.35781908e-01 -4.28068399e-01 -5.65096855e-01 5.80666214e-02 -2.57806659e-01 3.33758861e-01 -6.91538393e-01 2.12758973e-01 -4.84611481e-01 -8.15580711e-02 9.40717220e-01 -4.00936753e-01 -4.30817991e-01 -1.06423818e-01 4.75539148e-01 6.07023656e-01 3.03784847e-01 6.35945380e-01 -2.24663317e-01 6.13534153e-01 5.92268594e-02 3.71948257e-02 6.71478748e-01 1.13929417e-02 -8.07903782e-02 5.73189855e-01 -4.80866522e-01 -6.69233620e-01 -8.32986057e-01 5.39539218e-01 1.50925565e+00 9.36385989e-02 -1.54288188e-01 -1.49752748e+00 -2.70887852e-01 -6.44884109e-01 1.22922003e+00 -5.72167635e-01 -1.78433254e-01 -5.93902647e-01 1.65805221e-01 2.69482791e-01 -1.00868158e-02 2.52785653e-01 -1.76509941e+00 -1.45515919e+00 4.51850623e-01 -8.85907561e-02 -6.35441065e-01 -2.84434974e-01 -2.26543412e-01 -5.46137750e-01 -5.59831202e-01 -3.11782330e-01 -6.01244867e-01 7.51813889e-01 -8.08733404e-02 1.06121719e+00 1.25014499e-01 1.53296947e-01 9.64051366e-01 -5.57881296e-01 -6.69970930e-01 -1.38453424e+00 2.45268568e-02 -3.65808666e-01 2.94445013e-03 -3.95480655e-02 -5.26393771e-01 -4.59914088e-01 2.62881547e-01 -8.07409346e-01 5.59095800e-01 1.04459696e-01 2.72817284e-01 -4.47679893e-04 -4.67863411e-01 3.97622436e-01 -8.28838706e-01 1.30245745e+00 -8.92394334e-02 -6.37616277e-01 3.02837908e-01 -1.98489934e-01 1.31288067e-01 4.29964244e-01 -6.76039338e-01 -1.43454897e+00 9.69071761e-02 1.62781030e-01 3.33618112e-02 -4.58474874e-01 4.56786960e-01 -4.19071555e-01 -6.35630116e-02 8.46981466e-01 3.64573561e-02 2.41945148e-01 -1.34510219e-01 7.02882230e-01 5.80747187e-01 1.91657051e-01 -7.82280266e-01 5.52374363e-01 -1.89344939e-02 -6.41329229e-01 -9.09391284e-01 -2.16720536e-01 4.25521620e-02 -4.34167296e-01 -6.59014404e-01 6.78086698e-01 -6.08021915e-01 -1.31902969e+00 1.82272777e-01 -1.31905758e+00 -8.32362533e-01 -4.85340178e-01 3.29603463e-01 -8.15672696e-01 7.47762918e-02 -1.02659419e-01 -1.14905274e+00 -1.59706742e-01 -1.18789148e+00 7.71772206e-01 4.28095847e-01 -8.98228049e-01 -1.25573623e+00 2.16275290e-01 1.82586238e-01 5.99073052e-01 -1.18766308e-01 8.26945484e-01 -1.00146699e+00 -6.44730210e-01 6.85887858e-02 3.89033705e-01 -2.61832148e-01 6.74053421e-03 3.06177847e-02 -1.06634092e+00 1.35983273e-01 1.53799951e-01 -1.41666561e-01 -5.66430502e-02 1.34085283e-01 3.63267154e-01 -6.26968920e-01 -1.67178854e-01 -4.98759411e-02 8.07897449e-01 4.17345375e-01 5.19612491e-01 3.73536885e-01 2.85678625e-01 1.14633262e+00 6.34172916e-01 4.35546339e-01 8.64761710e-01 1.06373060e+00 3.80226493e-01 1.98110216e-03 5.86914755e-02 -3.31512541e-01 4.33230758e-01 1.80665657e-01 -2.29171902e-01 -4.09979492e-01 -8.48486245e-01 3.33669543e-01 -2.00499845e+00 -1.03521168e+00 2.71184534e-01 2.24195766e+00 9.60892558e-01 -6.74220324e-02 3.46935511e-01 -1.57284215e-01 8.49364102e-01 4.15144674e-02 -2.37514362e-01 -9.75116253e-01 4.73449886e-01 -1.56872258e-01 -3.79515290e-01 1.16385794e+00 -3.95751268e-01 1.30732095e+00 5.96222639e+00 2.28855461e-01 -1.03751087e+00 2.04648152e-01 4.10306126e-01 -6.33787876e-03 -5.68818629e-01 1.11440435e-01 -4.36891437e-01 4.09712531e-02 9.11007941e-01 -4.00483519e-01 5.80197573e-01 4.14763093e-01 7.19636142e-01 -3.72018158e-01 -1.28309941e+00 2.14419395e-01 3.67539539e-03 -1.07728827e+00 -9.53840092e-02 -1.08810253e-01 -1.25669196e-01 -6.35030985e-01 -1.56586066e-01 2.65622586e-01 5.73549151e-01 -8.98346424e-01 1.11896145e+00 5.18619120e-01 2.96007484e-01 -5.71282089e-01 4.50849861e-01 8.68337691e-01 -8.01450491e-01 1.66166455e-01 2.75714815e-01 -4.01787311e-01 3.78459007e-01 -4.08837497e-01 -1.39244449e+00 -1.67252496e-01 2.61328697e-01 -3.86658400e-01 -1.15582749e-01 5.28381705e-01 -4.87887233e-01 4.52963263e-01 -2.14240238e-01 -5.36559284e-01 6.46622851e-02 -1.70369640e-01 8.60477567e-01 1.02570879e+00 -1.05371745e-02 5.89730024e-01 2.99609572e-01 1.12005520e+00 3.81354541e-01 4.12521660e-01 -5.82389534e-01 1.54597342e-01 4.54656839e-01 1.08291221e+00 -6.03178322e-01 -3.18687946e-01 -7.17389286e-02 6.96870446e-01 3.42387140e-01 4.44886446e-01 -5.33621490e-01 -6.33868109e-03 4.85705107e-01 4.29242134e-01 -2.16423869e-02 2.21378386e-01 1.51479945e-01 -7.24779487e-01 -2.73996741e-01 -1.14929450e+00 6.42346144e-02 -1.12595999e+00 -6.08263552e-01 7.09232748e-01 1.13263413e-01 -6.54421449e-01 -8.27402472e-01 -2.25358486e-01 -7.99714684e-01 9.19243217e-01 -9.17780280e-01 -1.26195252e+00 -1.67602882e-01 2.89437532e-01 6.56146228e-01 2.23614890e-02 1.01897275e+00 -5.30824840e-01 -2.70497739e-01 2.23990738e-01 -9.73580897e-01 -2.07752988e-01 3.73092651e-01 -1.15554047e+00 2.74681062e-01 4.48469162e-01 -1.80895925e-02 5.46982408e-01 1.48097599e+00 -3.54017764e-01 -1.17562795e+00 -5.61285913e-01 9.76475298e-01 -2.85500705e-01 5.85381627e-01 -3.56530666e-01 -1.17281544e+00 7.32201695e-01 6.78154826e-01 -4.76749629e-01 5.93651235e-01 -5.11553921e-02 7.55730048e-02 1.60298899e-01 -1.27840137e+00 1.04158187e+00 8.62175882e-01 -4.80155766e-01 -9.58819985e-01 2.27800146e-01 6.47429228e-01 -6.04513347e-01 -3.93150449e-01 -2.03294545e-01 3.64082426e-01 -1.14180398e+00 6.81110203e-01 -4.23395306e-01 -1.27190024e-01 -4.36975688e-01 1.94679782e-01 -1.49859893e+00 -5.38351536e-02 -1.00539064e+00 3.39900672e-01 1.24647462e+00 5.18548787e-01 -8.25683057e-01 4.76411939e-01 9.66112018e-01 3.15350574e-03 -1.06886975e-01 -9.69082117e-01 -3.17734145e-02 1.96691435e-02 -2.75057554e-01 6.20794296e-01 5.53135455e-01 9.45309579e-01 6.03814602e-01 4.07921374e-02 4.76064682e-01 -1.52358669e-03 4.47137718e-05 1.07814240e+00 -1.23986983e+00 -4.61435467e-01 -5.09542644e-01 -7.78673496e-03 -7.73046374e-01 2.74156451e-01 -3.97470653e-01 4.18631613e-01 -1.58301151e+00 -4.65140641e-02 -9.40029100e-02 3.07026267e-01 3.70298088e-01 2.09027939e-02 -5.22535563e-01 3.56232166e-01 -9.58432406e-02 -6.79460227e-01 5.24640560e-01 1.39422953e+00 2.25183278e-01 -7.36451566e-01 2.30652869e-01 -7.60042012e-01 1.01892054e+00 7.02320516e-01 -3.02588016e-01 -6.58753037e-01 -2.06891462e-01 4.25746709e-01 4.60620582e-01 4.74834859e-01 -5.93756914e-01 5.78261375e-01 -5.70993900e-01 -3.07611555e-01 1.92490086e-01 3.85061741e-01 -6.66369736e-01 3.84908468e-01 4.86089408e-01 -7.58173048e-01 6.99014068e-02 3.44615310e-01 2.72132814e-01 -5.46378642e-02 -3.59674096e-01 6.70371056e-01 -3.36686760e-01 -4.13764715e-01 -4.31284100e-01 -1.26123905e+00 -2.54742503e-01 1.17667770e+00 -5.13858616e-01 -3.11721563e-02 -9.78076577e-01 -1.10720932e+00 1.99198037e-01 6.63905203e-01 1.55034423e-01 6.02218151e-01 -7.18845427e-01 -6.53284073e-01 1.59605667e-02 2.65243232e-01 7.12796450e-02 1.09608181e-01 3.73370171e-01 -5.52161872e-01 1.91652223e-01 1.44232977e-02 -3.77848357e-01 -1.39749396e+00 4.18489635e-01 8.57438445e-01 1.48343891e-01 -3.55496407e-01 7.52046943e-01 5.93181074e-01 -6.33722544e-01 1.52538151e-01 -2.42523596e-01 -7.28625894e-01 1.38174161e-01 6.18249655e-01 6.10047467e-02 -3.21467757e-01 -8.19772303e-01 -1.67476594e-01 4.56902742e-01 3.37541290e-02 -7.14904726e-01 1.00259066e+00 -5.15064180e-01 4.73012552e-02 6.39372766e-01 1.96345702e-01 3.56903583e-01 -1.23213708e+00 -6.43037707e-02 -1.64630488e-01 -2.11531162e-01 -2.66705751e-01 -1.02747571e+00 -4.17429656e-01 6.32933199e-01 3.86550814e-01 6.52298510e-01 8.37876081e-01 3.49280030e-01 -2.38258183e-01 3.99590671e-01 4.65594292e-01 -1.12613392e+00 1.93046033e-01 2.44635135e-01 1.17530608e+00 -9.26666260e-01 -3.08022559e-01 -5.05708098e-01 -1.06845129e+00 9.17430282e-01 7.11450279e-01 3.27761292e-01 1.56873226e-01 1.24574266e-01 7.24286258e-01 -3.50004941e-01 -1.24656928e+00 -2.63445228e-01 -6.23271763e-02 8.67621660e-01 4.55904186e-01 2.30252594e-01 -2.35829234e-01 3.35191369e-01 -5.00927329e-01 -9.36931968e-02 7.76442349e-01 6.52517557e-01 -7.77710259e-01 -1.12407410e+00 -4.97175992e-01 -6.30096197e-01 -3.80904675e-02 1.24280453e-01 -7.04194069e-01 9.01845872e-01 -6.95340857e-02 1.23784637e+00 9.54020321e-02 3.62971425e-01 6.56012833e-01 7.31317177e-02 3.68239671e-01 -1.09296882e+00 -9.60362792e-01 2.41036098e-02 3.78979057e-01 -4.00858700e-01 -4.55460370e-01 -7.50045836e-01 -1.53096390e+00 -8.40329230e-02 -1.02999359e-01 4.39045012e-01 5.81521571e-01 9.48459089e-01 2.76836604e-01 3.19414079e-01 5.73973656e-01 -8.04904222e-01 -4.30045724e-01 -9.80057120e-01 -1.78995088e-01 2.53870934e-01 2.67395169e-01 -2.77796894e-01 -3.60994756e-01 6.33814484e-02]	[12.94544506072998, 7.893821716308594]
526d7a6b-cfcd-46f1-99c8-aa43f250999b	a-teacher-student-framework-for-semi	2010.12219	null	https://arxiv.org/abs/2010.12219v1	https://arxiv.org/pdf/2010.12219v1.pdf	A Teacher-Student Framework for Semi-supervised Medical Image Segmentation From Mixed Supervision	Standard segmentation of medical images based on full-supervised convolutional networks demands accurate dense annotations. Such learning framework is built on laborious manual annotation with restrict demands for expertise, leading to insufficient high-quality labels. To overcome such limitation and exploit massive weakly labeled data, we relaxed the rigid labeling requirement and developed a semi-supervised learning framework based on a teacher-student fashion for organ and lesion segmentation with partial dense-labeled supervision and supplementary loose bounding-box supervision which are easier to acquire. Observing the geometrical relation of an organ and its inner lesions in most cases, we propose a hierarchical organ-to-lesion (O2L) attention module in a teacher segmentor to produce pseudo-labels. Then a student segmentor is trained with combinations of manual-labeled and pseudo-labeled annotations. We further proposed a localization branch realized via an aggregation of high-level features in a deep decoder to predict locations of organ and lesion, which enriches student segmentor with precise localization information. We validated each design in our model on LiTS challenge datasets by ablation study and showed its state-of-the-art performance compared with recent methods. We show our model is robust to the quality of bounding box and achieves comparable performance compared with full-supervised learning methods.	['Yizhou Yu', 'Guisheng Wang', 'Yue Huang', 'Xinghao Ding', 'Jianxiong Wu', 'Liyan Sun']	2020-10-23	null	null	null	null	['semi-supervised-medical-image-segmentation']	['computer-vision']	[ 2.21652940e-01 9.24966097e-01 -3.43926281e-01 -6.22260034e-01 -1.06763315e+00 -5.90462089e-01 3.90046805e-01 2.30963781e-01 -4.31969613e-01 5.57653487e-01 -9.85612497e-02 -2.81409949e-01 1.54955566e-01 -4.36971843e-01 -9.61600184e-01 -7.14592040e-01 1.60154596e-01 7.32936442e-01 6.31410420e-01 1.47283077e-01 -2.65135348e-01 3.47450316e-01 -1.01425588e+00 3.16756189e-01 1.12944424e+00 9.50385392e-01 4.48493391e-01 4.39953059e-01 -1.23005383e-01 8.01298022e-01 -2.77013719e-01 -3.62363845e-01 3.37060571e-01 -4.78126049e-01 -1.12128329e+00 4.42140013e-01 3.81123930e-01 -3.33527744e-01 -1.15157798e-01 1.03622591e+00 3.42163265e-01 -4.65445608e-01 8.39904308e-01 -8.19770038e-01 -6.14275753e-01 7.21581042e-01 -6.40554309e-01 1.84759945e-02 -8.44520330e-02 8.22774172e-02 8.20421398e-01 -6.89729273e-01 5.01795352e-01 6.25407219e-01 8.73497069e-01 5.89181066e-01 -1.09173298e+00 -3.45080227e-01 1.18970789e-01 -3.02574158e-01 -1.34937692e+00 -2.34298438e-01 5.04235506e-01 -5.69320202e-01 4.33968782e-01 1.59811825e-01 6.09371901e-01 9.13945138e-01 -5.56584895e-02 9.78288233e-01 8.95114422e-01 -3.53372961e-01 -1.77496416e-03 4.35667604e-01 1.94311380e-01 1.21603072e+00 5.13470694e-02 -6.03862219e-02 1.07752159e-01 1.41475901e-01 9.89563346e-01 1.47196025e-01 -2.93787539e-01 -6.57655597e-01 -1.22073543e+00 7.21290767e-01 1.01799834e+00 4.10356402e-01 -1.46185532e-01 6.47924542e-02 2.38968179e-01 -2.71699935e-01 4.59605575e-01 3.40766221e-01 -6.35434330e-01 4.90921021e-01 -1.13117087e+00 -4.32385832e-01 5.88923037e-01 9.59608316e-01 6.55944526e-01 -3.61684501e-01 -5.22898018e-01 6.73735261e-01 5.28862178e-01 1.43515334e-01 5.04251897e-01 -5.90559721e-01 2.39255652e-01 1.06532621e+00 -1.44415393e-01 -3.99531662e-01 -8.68672848e-01 -1.04962397e+00 -9.89370406e-01 9.87957232e-03 6.68983281e-01 -1.31172717e-01 -1.40671825e+00 1.49011934e+00 6.18131042e-01 2.72628009e-01 -1.44957051e-01 1.08969760e+00 9.05136943e-01 2.20213473e-01 1.53437078e-01 1.26180099e-02 1.48907018e+00 -1.57241356e+00 -4.84562695e-01 -5.75883575e-02 1.31960475e+00 -4.27166998e-01 1.11925375e+00 1.33188382e-01 -1.01815951e+00 -4.45628643e-01 -6.84985995e-01 -2.25160971e-01 -1.51393652e-01 6.56849802e-01 7.02161968e-01 5.56508064e-01 -1.14308059e+00 5.09150505e-01 -1.11473155e+00 -2.01666161e-01 9.61369097e-01 5.36964595e-01 -3.83500755e-01 1.40361860e-01 -7.82135367e-01 8.85709226e-01 4.47741896e-01 2.76262045e-01 -1.17548871e+00 -8.54069531e-01 -1.01570833e+00 4.07406986e-02 5.19663393e-01 -8.23189974e-01 1.23521185e+00 -1.06202292e+00 -1.50065589e+00 1.25205767e+00 1.73550069e-01 -3.99557352e-01 6.55926883e-01 -4.01029885e-02 6.47530854e-02 3.67036164e-01 2.49407858e-01 9.16589439e-01 8.07833135e-01 -1.22368932e+00 -4.87719864e-01 -3.94326627e-01 -1.78261455e-02 5.44633754e-02 -1.31315768e-01 -3.85022789e-01 -6.30150497e-01 -5.41689396e-01 2.95412809e-01 -9.15657401e-01 -5.56134760e-01 2.05206230e-01 -7.39643693e-01 -2.64837950e-01 7.40725696e-01 -4.48459357e-01 9.25448537e-01 -2.05579662e+00 6.11439571e-02 5.91232181e-02 3.99621904e-01 3.59474689e-01 -1.02539686e-02 -1.80181146e-01 -1.24955773e-01 -3.56855989e-02 -4.46992904e-01 -6.12930000e-01 -1.90369517e-01 1.92640319e-01 2.17365362e-02 7.41340756e-01 2.64967233e-01 1.17214370e+00 -9.93255258e-01 -9.58530843e-01 2.15867162e-01 4.31731552e-01 -6.69440627e-01 4.47664380e-01 -2.66404808e-01 9.02262509e-01 -5.63353598e-01 8.72154653e-01 3.46845061e-01 -8.76329660e-01 -8.48276392e-02 -4.13135201e-01 7.33464211e-02 5.88671118e-02 -7.09648609e-01 2.17931700e+00 -5.62429488e-01 1.05979048e-01 2.38359615e-01 -1.15463722e+00 5.87632775e-01 4.73181129e-01 5.13557136e-01 -1.80976167e-01 4.12666649e-01 3.16550106e-01 -5.68248704e-02 -7.42749631e-01 -2.16187775e-01 7.17049912e-02 -3.46060060e-02 1.33194566e-01 4.79813039e-01 -2.93388277e-01 1.39166452e-02 2.75849104e-01 1.06034541e+00 5.39883316e-01 1.67248115e-01 -4.88092601e-01 6.58892572e-01 1.12211769e-02 4.53643560e-01 5.25619447e-01 -3.44282150e-01 8.57575536e-01 4.95006263e-01 -4.92904782e-01 -7.68983603e-01 -8.32801342e-01 -4.93086249e-01 1.10709977e+00 2.60203004e-01 -2.19154596e-01 -1.10722578e+00 -1.43875086e+00 -3.06025028e-01 3.84069800e-01 -9.25038517e-01 2.78727375e-02 -4.09469157e-01 -8.35687220e-01 6.06829107e-01 5.93728483e-01 3.61908197e-01 -9.57688987e-01 -5.96595943e-01 2.78832875e-02 4.63241339e-02 -1.26310110e+00 -6.54725790e-01 5.71635008e-01 -8.87211502e-01 -1.29034531e+00 -1.04414916e+00 -1.23242593e+00 1.25229120e+00 -2.65031278e-01 9.86527979e-01 2.14108765e-01 -4.54079747e-01 -4.84848693e-02 -2.65800655e-01 -2.43735194e-01 -2.33749181e-01 4.15102094e-01 -3.77913356e-01 9.43578631e-02 -1.46578297e-01 -4.23868626e-01 -8.27548444e-01 4.90299702e-01 -8.86987746e-01 4.19737220e-01 1.04980242e+00 9.73338604e-01 7.00889647e-01 -3.32970709e-01 5.17910063e-01 -1.20298696e+00 -1.78976133e-01 -3.15192133e-01 -7.17118323e-01 2.41549402e-01 -3.23900968e-01 2.22472578e-01 6.15751386e-01 -3.50725502e-01 -9.13241982e-01 8.02802324e-01 -3.04892063e-01 -3.07696253e-01 -4.19823855e-01 3.04526478e-01 -7.27799386e-02 -1.44533187e-01 6.45906687e-01 7.37830773e-02 -1.45597339e-01 -4.86160934e-01 4.02390599e-01 5.61229050e-01 6.64727092e-01 -3.61932874e-01 6.79017186e-01 5.30940354e-01 -7.72937685e-02 -3.08943629e-01 -1.45484364e+00 -4.05589938e-01 -1.18498039e+00 9.33706984e-02 1.21266651e+00 -8.85359108e-01 -5.23494303e-01 2.45564774e-01 -1.07627749e+00 -6.79703414e-01 -4.84522223e-01 5.71797729e-01 -3.99885982e-01 2.43754461e-01 -8.56057644e-01 -2.30419964e-01 -2.52385437e-01 -1.45528233e+00 1.46136332e+00 1.86869159e-01 1.44358858e-01 -1.06416571e+00 -1.51185878e-02 5.27216256e-01 3.17476988e-01 1.74195275e-01 6.08582377e-01 -9.31652546e-01 -6.82649791e-01 -2.64323533e-01 -5.04052520e-01 2.09505349e-01 7.89991394e-02 -3.11420351e-01 -9.73125219e-01 -1.94730729e-01 -1.67872146e-01 -5.82953751e-01 9.79890227e-01 5.28432012e-01 1.50034451e+00 -1.19878493e-01 -7.77912974e-01 1.05539429e+00 1.14595509e+00 -3.96709561e-01 2.74921566e-01 -8.59666057e-03 1.06035340e+00 6.01383388e-01 3.60055506e-01 8.14741328e-02 4.52235281e-01 5.51367700e-01 6.45134151e-01 -7.18709171e-01 -3.42874229e-01 -1.34134814e-01 3.01537719e-02 7.38884270e-01 1.15655862e-01 -9.05273631e-02 -9.31636631e-01 7.65170991e-01 -1.71387589e+00 -2.53763974e-01 -2.44420215e-01 1.77290630e+00 1.18598783e+00 1.51882038e-01 4.83199097e-02 -1.41854540e-01 5.80454350e-01 -1.74818739e-01 -5.55035472e-01 1.17498405e-01 2.48963103e-01 2.30628364e-02 6.75239027e-01 4.53946024e-01 -1.43974340e+00 1.07917058e+00 5.61431170e+00 9.37853575e-01 -1.14391923e+00 3.34255636e-01 8.79357159e-01 2.86331356e-01 3.68849337e-02 -1.71823829e-01 -8.90271783e-01 3.40566218e-01 6.82092607e-01 6.20784283e-01 -2.23386690e-01 9.55290794e-01 -7.37092569e-02 8.67766142e-03 -1.38966620e+00 5.75563431e-01 3.80923375e-02 -1.16687953e+00 -2.28009373e-01 2.04111598e-02 8.44176650e-01 2.52971143e-01 -1.55806229e-01 2.48545527e-01 3.29073399e-01 -1.16353750e+00 4.83629376e-01 2.52004385e-01 9.24121439e-01 -2.29446173e-01 9.53720629e-01 5.67288220e-01 -1.06807482e+00 1.05306856e-01 -1.40547499e-01 3.37946743e-01 2.41488934e-01 6.23611391e-01 -1.05360782e+00 5.35204887e-01 4.97440994e-01 6.99075341e-01 -6.86287820e-01 1.03902447e+00 -4.36176062e-01 7.10058928e-01 -4.61858571e-01 3.85856658e-01 4.25332159e-01 5.54323904e-02 1.60364866e-01 1.28466415e+00 5.16804494e-02 1.35678396e-01 5.22634268e-01 9.15658355e-01 -1.60795256e-01 1.37681514e-01 -1.97328404e-01 2.58837610e-01 5.02166059e-03 1.62344909e+00 -1.19274199e+00 -5.18051863e-01 -3.27083260e-01 1.12949622e+00 4.89588618e-01 6.23773411e-02 -9.55719113e-01 -1.31583810e-01 -2.70108551e-01 2.71783471e-01 3.34365666e-01 2.55331725e-01 -3.80625367e-01 -1.24703312e+00 -2.20252961e-01 -3.08998674e-01 4.62596238e-01 -5.37081659e-01 -1.10808134e+00 9.07171905e-01 -1.88400015e-01 -1.10194325e+00 7.49841239e-03 -5.51918507e-01 -5.06508112e-01 6.00249887e-01 -1.72680056e+00 -1.61118412e+00 -4.70790952e-01 6.15618825e-01 4.15177584e-01 7.80631155e-02 9.00925577e-01 5.34346759e-01 -7.68034339e-01 7.61854827e-01 -3.88569474e-01 4.91236955e-01 7.18783557e-01 -1.47911382e+00 -8.67941678e-02 5.70503354e-01 1.90611407e-01 2.58673698e-01 1.29362270e-01 -4.99931127e-01 -8.99256051e-01 -1.43526995e+00 7.08716810e-01 -6.36946023e-01 5.45601606e-01 -3.52959722e-01 -8.26309025e-01 8.88610840e-01 9.49824452e-02 8.25781167e-01 5.81007659e-01 -2.07085833e-01 -1.10673122e-01 3.17748152e-02 -1.25089264e+00 2.96211630e-01 9.05400813e-01 -3.34114999e-01 -6.07082367e-01 7.98337877e-01 9.78348672e-01 -7.86028326e-01 -7.60143280e-01 6.51503265e-01 1.28044277e-01 -8.41660738e-01 8.53573740e-01 -3.02462578e-01 3.94379050e-01 -4.97294545e-01 3.21910292e-01 -1.05054951e+00 -2.05580354e-01 -3.94231528e-01 8.13868120e-02 9.30552542e-01 6.02772713e-01 -2.68327951e-01 1.02093518e+00 3.74617279e-01 -7.59682178e-01 -1.22417164e+00 -6.28166735e-01 -3.49566907e-01 4.51975083e-03 -1.26616552e-01 1.99394256e-01 1.00681829e+00 1.92832798e-02 4.61952239e-01 -3.43262292e-02 4.88177270e-01 5.73463738e-01 6.15692958e-02 3.85456175e-01 -1.13747871e+00 -3.61009866e-01 -3.44504982e-01 -2.64173627e-01 -1.28835845e+00 1.26922756e-01 -1.31762338e+00 6.90091178e-02 -1.60744607e+00 3.12744290e-01 -7.02483416e-01 -2.97770530e-01 8.58308136e-01 -1.85149118e-01 7.79407024e-01 -2.56310552e-01 2.51695842e-01 -1.04931402e+00 4.17947948e-01 1.60309446e+00 -7.40840212e-02 -2.34374329e-01 1.46831870e-01 -4.57973152e-01 1.06454182e+00 4.70142484e-01 -5.96946597e-01 -2.35739619e-01 -5.56571126e-01 -1.69586912e-01 2.07953170e-01 4.76242304e-01 -7.86040604e-01 3.60237420e-01 3.29709500e-01 4.58543211e-01 -5.84598839e-01 1.32440524e-02 -1.13538730e+00 -3.90234023e-01 5.71115851e-01 -5.68320215e-01 -5.75427532e-01 -3.17719355e-02 3.78760338e-01 -1.18656754e-01 -2.99565077e-01 9.14972961e-01 -1.84679255e-01 -1.21863693e-01 5.23315430e-01 9.96083021e-02 -7.93201923e-02 1.29310071e+00 -7.24376217e-02 8.44013393e-02 1.15086503e-01 -1.18100572e+00 4.97779965e-01 2.46368587e-01 -9.29691419e-02 2.16979653e-01 -9.90527511e-01 -5.80403447e-01 2.08275720e-01 8.77017751e-02 8.43126774e-01 1.51043788e-01 1.34029317e+00 -6.51936889e-01 5.62914908e-01 1.68795913e-01 -1.12962627e+00 -8.61683726e-01 6.62856638e-01 6.80356383e-01 -6.27991319e-01 -7.95720518e-01 1.16318631e+00 7.51997828e-01 -8.51491868e-01 5.05362093e-01 -7.33034372e-01 -1.29819721e-01 -1.65776998e-01 2.58212417e-01 -1.77584544e-01 1.37188062e-01 -6.17322683e-01 -3.57666433e-01 7.12039709e-01 -1.34878963e-01 3.76480520e-01 1.16860569e+00 -1.01850927e-01 -6.48740232e-02 7.73245916e-02 1.19339752e+00 -1.89825669e-01 -1.53535569e+00 -3.42802823e-01 1.47945836e-01 1.67968236e-02 2.03155309e-01 -1.09502673e+00 -1.35131907e+00 9.71322000e-01 5.68295419e-01 -7.62735009e-02 1.00899792e+00 3.65100831e-01 7.95800030e-01 1.64231166e-01 1.74421862e-01 -6.51566744e-01 1.17463693e-01 2.17916474e-01 4.90602434e-01 -1.62622273e+00 -1.86080933e-01 -7.06204355e-01 -6.00041568e-01 9.76655900e-01 7.87243128e-01 -1.79100022e-01 8.79156172e-01 4.71762151e-01 2.78740019e-01 -3.12696248e-01 -4.33566421e-01 -2.44000867e-01 7.26127207e-01 3.44220906e-01 5.34405231e-01 -1.00873031e-01 -7.25202635e-02 9.26093638e-01 1.44740362e-02 1.26276150e-01 2.01771826e-01 6.60041451e-01 -3.35008293e-01 -9.93632197e-01 -1.33503854e-01 3.67120385e-01 -6.49074256e-01 -1.36638775e-01 -6.49118125e-02 7.28554606e-01 4.63653803e-01 4.39357460e-01 1.71751510e-02 8.93076882e-02 1.74853820e-02 -1.49107859e-01 4.88607168e-01 -1.00033402e+00 -7.51654565e-01 3.86101693e-01 -2.94970006e-01 -5.65192759e-01 -3.99772555e-01 -2.58186668e-01 -1.62903559e+00 4.93288338e-01 -6.61942482e-01 1.60698831e-01 4.61860538e-01 1.20025671e+00 2.19345450e-01 8.34990740e-01 4.65814114e-01 -9.31154251e-01 -4.58257526e-01 -9.78917599e-01 -5.39886892e-01 3.93547893e-01 4.51692730e-01 -5.04734814e-01 -1.46078140e-01 2.82264620e-01]	[14.710126876831055, -2.412985324859619]
6863ed7f-e723-4c5f-90c1-143aa9b56f58	facing-off-world-model-backbones-rnns	2307.02064	null	https://arxiv.org/abs/2307.02064v1	https://arxiv.org/pdf/2307.02064v1.pdf	Facing off World Model Backbones: RNNs, Transformers, and S4	World models are a fundamental component in model-based reinforcement learning (MBRL) agents. To perform temporally extended and consistent simulations of the future in partially observable environments, world models need to possess long-term memory. However, state-of-the-art MBRL agents, such as Dreamer, predominantly employ recurrent neural networks (RNNs) as their world model backbone, which have limited memory capacity. In this paper, we seek to explore alternative world model backbones for improving long-term memory. In particular, we investigate the effectiveness of Transformers and Structured State Space Sequence (S4) models, motivated by their remarkable ability to capture long-range dependencies in low-dimensional sequences and their complementary strengths. We propose S4WM, the first S4-based world model that can generate high-dimensional image sequences through latent imagination. Furthermore, we extensively compare RNN-, Transformer-, and S4-based world models across four sets of environments, which we have specifically tailored to assess crucial memory capabilities of world models, including long-term imagination, context-dependent recall, reward prediction, and memory-based reasoning. Our findings demonstrate that S4WM outperforms Transformer-based world models in terms of long-term memory, while exhibiting greater efficiency during training and imagination. These results pave the way for the development of stronger MBRL agents.	['Sungjin Ahn', 'Junyeong Park', 'Fei Deng']	2023-07-05	null	null	null	null	['model-based-reinforcement-learning']	['reasoning']	[-3.39182556e-01 4.79285903e-02 -2.62234598e-01 1.88301906e-01 -1.82469040e-01 -3.59760314e-01 1.04377604e+00 -3.65770340e-01 -4.46881086e-01 1.04377496e+00 4.17685211e-01 -1.54187292e-01 -4.61364865e-01 -1.17459214e+00 -6.88811898e-01 -4.82242078e-01 -3.21065724e-01 7.95867562e-01 1.64367676e-01 -7.44619131e-01 2.74505258e-01 4.45740342e-01 -1.43094182e+00 1.71494156e-01 8.68811250e-01 4.68168288e-01 7.98331082e-01 5.52594543e-01 -8.36710781e-02 1.54909444e+00 -4.85435307e-01 -1.59922048e-01 -9.47650149e-02 -5.39130151e-01 -8.35219264e-01 -1.44322067e-01 -5.17041624e-01 -4.53081936e-01 -8.46148908e-01 6.39489472e-01 5.17621636e-01 4.74844247e-01 3.04032058e-01 -9.87832487e-01 -9.00998831e-01 5.50389946e-01 1.30087703e-01 1.63766459e-01 4.99848843e-01 8.01543891e-01 6.16539121e-01 -5.63018620e-01 9.63510633e-01 1.18221283e+00 3.37245613e-01 1.07941949e+00 -1.15415728e+00 -3.61353517e-01 1.96560591e-01 6.29298449e-01 -9.17201102e-01 -4.15910602e-01 9.10049200e-01 7.03651160e-02 1.65447378e+00 -1.53346226e-01 1.26610124e+00 1.64495015e+00 6.53783739e-01 1.01168823e+00 1.46190608e+00 -2.75681049e-01 5.20095587e-01 -4.21681732e-01 -2.33995616e-01 8.52147162e-01 -1.43799990e-01 8.19476247e-01 -1.05810130e+00 2.34798849e-01 1.31087971e+00 -1.31517708e-01 -2.91910442e-03 -4.67560798e-01 -1.56998634e+00 6.20169759e-01 4.75515515e-01 3.35386813e-01 -7.10105717e-01 4.74292576e-01 5.69792278e-02 4.82142657e-01 3.01110029e-01 9.74250436e-01 -9.03377980e-02 -5.36610425e-01 -5.95016599e-01 3.09847891e-01 6.54789865e-01 6.79689586e-01 4.29336578e-01 5.04327476e-01 -2.39761189e-01 5.98846853e-01 2.53651410e-01 6.19279623e-01 1.12242377e+00 -1.06379986e+00 3.07292432e-01 3.51954490e-01 3.69823873e-01 -6.17510080e-01 -6.09216690e-01 -7.27522135e-01 -6.80251837e-01 1.13998413e-01 -1.16234653e-01 5.34561686e-02 -1.02658999e+00 1.99976003e+00 -1.50614530e-01 6.51170492e-01 6.32333875e-01 8.58316958e-01 6.22833371e-01 8.89595211e-01 1.20396525e-01 -3.64563197e-01 7.69702256e-01 -1.27811158e+00 -6.46833420e-01 -6.64553344e-01 7.15775430e-01 9.59150344e-02 1.22023702e+00 1.96759790e-01 -1.56237292e+00 -5.87092578e-01 -1.20429301e+00 2.76967555e-01 -4.13886398e-01 -4.13438141e-01 8.15958500e-01 2.91013420e-01 -1.44185078e+00 8.38680446e-01 -9.08630013e-01 -2.93538302e-01 4.18791957e-02 2.10714504e-01 -2.59758294e-01 9.83740017e-03 -1.65990031e+00 1.59421790e+00 5.51839113e-01 5.55499233e-02 -1.53977346e+00 -2.14987889e-01 -9.28352714e-01 4.52278294e-02 2.92379737e-01 -1.07306337e+00 1.23412609e+00 -4.56913680e-01 -1.98768079e+00 6.19691670e-01 5.07531762e-02 -9.33402300e-01 9.69341248e-02 -2.98656635e-02 -6.41282856e-01 6.08168952e-02 -2.93500721e-01 7.74737775e-01 7.46490836e-01 -1.08648467e+00 4.68492210e-02 -7.20785409e-02 1.99655026e-01 4.69649553e-01 -2.53968596e-01 -4.37264174e-01 1.53185083e-02 -6.52027488e-01 -6.89752996e-02 -9.73330081e-01 -5.16725183e-01 -4.52081412e-01 2.43160158e-01 -2.03169852e-01 4.21841949e-01 -3.35308731e-01 1.11669731e+00 -1.83914733e+00 4.09683347e-01 -1.13779925e-01 1.89363007e-02 4.84768271e-01 -8.15276265e-01 8.36825907e-01 2.47909933e-01 -8.18343684e-02 6.50010677e-03 -2.57749498e-01 2.00236648e-01 7.06200063e-01 -7.02770412e-01 9.74521507e-03 1.67145818e-01 1.64777398e+00 -1.17804813e+00 -4.13914919e-01 3.42802048e-01 2.29974076e-01 -3.15300435e-01 2.67315149e-01 -8.00239205e-01 5.00542879e-01 -4.11218852e-01 5.01559190e-02 8.51769298e-02 -5.66805899e-01 4.34892535e-01 4.27464545e-01 -2.87620407e-02 5.24592400e-01 -6.67151570e-01 2.16980076e+00 -8.63644123e-01 3.39028835e-01 -5.46229303e-01 -7.74014235e-01 9.96958017e-01 3.64710569e-01 1.66157722e-01 -1.58098972e+00 -2.48349637e-01 1.46563008e-01 5.51863834e-02 -4.22034502e-01 6.61913216e-01 -4.28014845e-01 2.12507658e-02 7.47315168e-01 9.84029323e-02 -5.05243480e-01 4.98391509e-01 1.99732631e-01 1.01296425e+00 4.82123345e-01 1.20806471e-01 1.12866648e-02 1.80742845e-01 -1.68696880e-01 3.58832121e-01 9.79493082e-01 -1.87758625e-01 2.85082787e-01 1.54474109e-01 -4.64534581e-01 -9.09836948e-01 -1.25466132e+00 4.58674580e-01 9.51961935e-01 3.65143418e-01 -2.45551020e-01 -4.41923469e-01 -2.48099506e-01 -4.22640055e-01 1.13434947e+00 -5.35534024e-01 -5.78068614e-01 -8.22856128e-01 -6.26599729e-01 5.05535126e-01 5.38248599e-01 7.91321695e-01 -1.96494567e+00 -1.18180227e+00 6.83235645e-01 -1.94075704e-01 -7.80028760e-01 -1.44020822e-02 1.86514765e-01 -1.13002992e+00 -6.38478994e-01 -7.58284032e-01 -5.63638270e-01 2.24699080e-01 7.11924061e-02 1.43360651e+00 2.32387241e-02 2.41654709e-01 5.23686409e-01 -3.09771419e-01 7.29744434e-02 -4.42366332e-01 -8.98438692e-03 3.42880130e-01 -5.41568816e-01 -2.39417344e-01 -1.02463818e+00 -5.61799645e-01 2.25627542e-01 -9.07245457e-01 5.37999451e-01 8.96064043e-01 1.13907874e+00 3.84320170e-01 -1.56415194e-01 8.86103988e-01 -3.47763240e-01 1.07096064e+00 -4.88114953e-01 -4.70783144e-01 5.26061237e-01 -7.19406545e-01 5.62048852e-01 6.08607054e-01 -5.58097959e-01 -1.46005869e+00 -5.70692778e-01 -3.35600853e-01 -1.79784685e-01 1.63857415e-01 9.22758043e-01 2.51928151e-01 8.99948701e-02 6.82079196e-01 9.82385755e-01 9.85299498e-02 -1.20099224e-01 4.74039793e-01 -3.51699479e-02 4.81637031e-01 -8.10780346e-01 4.49009806e-01 1.92851096e-01 3.30012701e-02 -4.92297173e-01 -6.20442212e-01 7.91103691e-02 -1.97148547e-01 -2.76413411e-01 6.31395161e-01 -7.89556682e-01 -6.23515904e-01 6.78395271e-01 -1.07477486e+00 -1.14477694e+00 -5.99833548e-01 4.58822191e-01 -1.34447646e+00 4.40644585e-02 -9.47960079e-01 -8.95157099e-01 -1.71091810e-01 -1.08937204e+00 5.97338140e-01 2.26378575e-01 -2.17300937e-01 -1.11506450e+00 5.36100328e-01 1.16791442e-01 7.76867151e-01 -8.58140960e-02 9.62144792e-01 -3.04765671e-01 -7.87378013e-01 3.00165951e-01 1.90763086e-01 -2.11770341e-01 -3.78735781e-01 -7.23232687e-01 -6.42515361e-01 -2.72552222e-01 5.44719212e-02 -8.45680118e-01 9.85294163e-01 2.05906272e-01 8.32575858e-01 -1.69910178e-01 -1.23058371e-01 4.18641627e-01 1.12373257e+00 3.94516945e-01 1.07558084e+00 6.50447369e-01 1.37976646e-01 4.03651327e-01 6.92878485e-01 5.03358662e-01 4.68428463e-01 5.60092390e-01 6.05433762e-01 2.86519796e-01 -2.10166886e-01 -8.77631307e-01 7.13312387e-01 1.20446539e+00 -1.49467915e-01 -3.49253476e-01 -8.65583479e-01 5.94481409e-01 -2.04221869e+00 -1.34711015e+00 4.96798068e-01 1.81726992e+00 9.41867828e-01 3.79571468e-01 -1.25998929e-01 -1.98018447e-01 1.25397816e-01 8.00226092e-01 -1.00525475e+00 -3.59252274e-01 -4.14440453e-01 3.03342283e-01 -9.87307355e-02 3.16236138e-01 -5.34860134e-01 1.27665377e+00 6.71708298e+00 9.34035063e-01 -9.25086379e-01 2.08142176e-02 4.96587306e-01 -2.02984646e-01 -5.05043328e-01 -1.13358468e-01 -3.56838703e-01 3.38769853e-01 1.43572819e+00 -1.78379595e-01 9.61697817e-01 6.61413670e-01 1.59445256e-01 -2.65657343e-02 -8.90587747e-01 9.62103963e-01 -1.29241079e-01 -1.86231828e+00 1.72212288e-01 -5.05754258e-03 8.44642818e-01 2.22099766e-01 5.60832441e-01 8.66971314e-01 7.02415109e-01 -1.30130911e+00 8.82352889e-01 8.51620793e-01 5.86813450e-01 -6.81599796e-01 3.69638443e-01 8.64001572e-01 -9.24636841e-01 -3.34676504e-01 -3.03449720e-01 -4.68136400e-01 4.54215109e-01 4.39965315e-02 -3.66324127e-01 2.67185599e-01 2.23900452e-01 7.44250178e-01 -4.62951124e-01 5.53167582e-01 -3.83760780e-01 3.46629530e-01 -1.28380843e-02 -3.14300656e-01 4.07091081e-01 -1.22100182e-01 3.67606074e-01 6.56693995e-01 4.63948756e-01 3.22146475e-01 -4.03678827e-02 9.39224899e-01 -3.76652032e-02 -4.47656929e-01 -8.17121744e-01 -2.76211828e-01 4.94161606e-01 6.32498085e-01 -5.65892100e-01 -3.25486392e-01 1.69795677e-02 9.93866205e-01 7.05535829e-01 4.86541539e-01 -8.92424464e-01 2.02016592e-01 4.62470829e-01 -2.32294038e-01 9.79893059e-02 -7.40928590e-01 -1.54440880e-01 -1.26242888e+00 -2.37498090e-01 -1.09171772e+00 3.03686298e-02 -1.23063624e+00 -9.57651019e-01 6.77548766e-01 -8.13695788e-02 -9.21783030e-01 -8.71199369e-01 -4.28317904e-01 -5.42930186e-01 5.33529997e-01 -1.68910277e+00 -1.28291678e+00 1.18031397e-01 7.51324654e-01 7.43888855e-01 -2.59013683e-01 1.16237128e+00 -3.96205395e-01 -3.02570343e-01 9.54272225e-02 1.95556059e-01 -2.32254967e-01 1.74061149e-01 -9.76728916e-01 8.00069094e-01 6.03290200e-01 4.71009225e-01 8.40537548e-01 6.93286479e-01 -7.65376687e-01 -1.74568832e+00 -9.82644916e-01 5.59608519e-01 -3.19104403e-01 7.32892036e-01 -1.00139782e-01 -7.01491952e-01 7.00255930e-01 2.32506871e-01 -1.66054040e-01 2.35025331e-01 3.16631943e-02 -1.90721020e-01 2.53398925e-01 -9.82841671e-01 1.05279255e+00 1.44346535e+00 -7.81311989e-01 -8.18332970e-01 1.71943009e-01 7.29554951e-01 -2.43948191e-01 -7.05976665e-01 2.49012083e-01 6.02173209e-01 -1.26318395e+00 1.18410265e+00 -6.07572734e-01 4.81089860e-01 -5.58592081e-02 1.03889897e-01 -1.79549420e+00 -3.96625668e-01 -7.40551591e-01 -7.70843387e-01 4.95702356e-01 3.31680954e-01 -9.14408684e-01 8.22956324e-01 4.56051379e-01 -2.09974781e-01 -8.20874870e-01 -9.50409830e-01 -1.19843149e+00 1.70295879e-01 -4.93699878e-01 8.67111862e-01 5.70970237e-01 2.99754649e-01 2.60045379e-01 -7.88885891e-01 -2.97521740e-01 2.50007182e-01 2.94396818e-01 4.97739077e-01 -8.81968856e-01 -5.59830785e-01 -3.72252136e-01 6.35456368e-02 -1.37517238e+00 5.73721707e-01 -6.27435029e-01 1.36706019e-02 -1.80008769e+00 1.40871540e-01 -4.27774489e-01 -5.09534657e-01 2.63025522e-01 1.22962922e-01 -5.99704981e-02 4.75587249e-01 2.48246282e-01 -1.07426536e+00 1.14406931e+00 1.70179987e+00 -6.65128045e-03 -1.87137991e-01 -3.35456759e-01 -2.08726600e-01 5.58039844e-01 1.07446110e+00 -8.73513818e-02 -9.03642774e-01 -6.59674108e-01 6.74412489e-01 6.89039826e-01 4.69464630e-01 -1.20969975e+00 4.28709298e-01 -5.88227391e-01 3.55717570e-01 -3.67299914e-01 8.60970855e-01 -4.06096250e-01 3.57139796e-01 8.22148919e-01 -5.92597365e-01 3.35240513e-01 1.73455060e-01 7.41480768e-01 -1.45741060e-01 -9.89243239e-02 5.10023713e-01 -6.98317826e-01 -1.26482761e+00 2.77050644e-01 -7.88101435e-01 -8.02338794e-02 9.50945616e-01 -1.85066313e-01 -6.17029905e-01 -6.68083429e-01 -7.67167985e-01 2.57030845e-01 4.38318610e-01 4.82266098e-01 1.18151200e+00 -1.32500100e+00 -5.21863639e-01 1.76413044e-01 -5.92737421e-02 -4.83179808e-01 4.73199844e-01 4.92379218e-01 -3.53691280e-01 7.72084177e-01 -6.28625572e-01 -4.05337801e-03 -5.01491487e-01 8.42010379e-01 4.27039832e-01 -9.76690292e-01 -6.57815695e-01 6.83883667e-01 1.61042698e-02 -4.85253215e-01 -1.71243936e-01 -1.22456118e-01 -2.93386161e-01 -3.11650068e-01 3.22602183e-01 3.36400062e-01 -3.33554864e-01 -2.97243685e-01 -3.26693244e-02 1.52751133e-01 9.86186266e-02 -6.04006112e-01 1.41609514e+00 -4.88449596e-02 7.13359490e-02 5.11272728e-01 4.67988044e-01 -6.62283838e-01 -1.45544815e+00 -5.71885984e-03 1.65218875e-01 -1.09420456e-01 -6.46696910e-02 -9.81556177e-01 -6.46648347e-01 8.61724317e-01 4.09930825e-01 -4.01301339e-04 1.03070247e+00 -1.92622617e-01 1.08636081e+00 7.82751739e-01 1.10111928e+00 -1.10678864e+00 7.66159713e-01 9.89723027e-01 9.95908678e-01 -9.31992114e-01 -3.17684710e-01 5.50945580e-01 -7.06361592e-01 7.83014417e-01 8.06640923e-01 -2.39225477e-01 2.22415492e-01 9.93263442e-03 -2.61844218e-01 -2.48201445e-01 -1.43944108e+00 -4.23660904e-01 -1.19289560e-02 8.97136033e-01 3.76394540e-02 2.61090361e-02 -1.85292996e-02 3.47999692e-01 -2.90892750e-01 2.40882650e-01 4.67379123e-01 1.04078174e+00 -5.00896871e-01 -1.14452469e+00 -8.67201313e-02 2.57223636e-01 1.31578162e-01 -2.30743513e-01 1.75658949e-02 5.83557189e-01 -2.62889713e-01 9.38180029e-01 -1.93130240e-01 -2.71948546e-01 -2.37741657e-02 2.15823174e-01 9.05181229e-01 -4.34702098e-01 -4.61069554e-01 -5.29636323e-01 2.74764150e-01 -7.28429258e-01 -3.58392298e-01 -2.06914425e-01 -1.31783926e+00 -3.82709742e-01 -1.93399563e-02 8.41364563e-02 4.25754339e-01 9.39999819e-01 5.54580450e-01 5.69596112e-01 1.92590639e-01 -9.95425463e-01 -6.69245005e-01 -8.64525855e-01 -5.88765860e-01 2.80268729e-01 2.30554983e-01 -8.11485767e-01 2.28331640e-01 -3.06709051e-01]	[4.15914249420166, 1.5507582426071167]
15407c91-7286-4c08-82a4-ec92673ac5bc	unsupervised-meta-learning-through-latent	2006.10236	null	https://arxiv.org/abs/2006.10236v1	https://arxiv.org/pdf/2006.10236v1.pdf	Unsupervised Meta-Learning through Latent-Space Interpolation in Generative Models	Unsupervised meta-learning approaches rely on synthetic meta-tasks that are created using techniques such as random selection, clustering and/or augmentation. Unfortunately, clustering and augmentation are domain-dependent, and thus they require either manual tweaking or expensive learning. In this work, we describe an approach that generates meta-tasks using generative models. A critical component is a novel approach of sampling from the latent space that generates objects grouped into synthetic classes forming the training and validation data of a meta-task. We find that the proposed approach, LAtent Space Interpolation Unsupervised Meta-learning (LASIUM), outperforms or is competitive with current unsupervised learning baselines on few-shot classification tasks on the most widely used benchmark datasets. In addition, the approach promises to be applicable without manual tweaking over a wider range of domains than previous approaches.	['Ladislau Bölöni', 'Sharare Zehtabian', 'Saeed Vahidian', 'Weijia Wang', 'Siavash Khodadadeh', 'Bill Lin']	2020-06-18	null	https://openreview.net/forum?id=XOjv2HxIF6i	https://openreview.net/pdf?id=XOjv2HxIF6i	iclr-2021-1	['unsupervised-few-shot-image-classification']	['computer-vision']	[ 6.76207185e-01 2.80542284e-01 -4.36496496e-01 -3.63327980e-01 -1.24571860e+00 -3.63347828e-01 1.33436620e+00 1.93047464e-01 -4.38739717e-01 1.01434863e+00 3.55124980e-01 8.31409916e-02 -2.48092432e-02 -7.63458073e-01 -6.92903101e-01 -7.06720531e-01 4.98671979e-01 9.55797613e-01 1.61269590e-01 1.20311461e-01 3.83183062e-01 -1.70007855e-01 -2.08721662e+00 7.99970806e-01 1.09856355e+00 4.10042077e-01 2.03846350e-01 4.97906506e-01 -3.62031043e-01 9.04104054e-01 -8.26664567e-01 -2.06042826e-01 -7.51833841e-02 -6.62398756e-01 -8.24006438e-01 4.01177377e-01 2.04245329e-01 8.80833492e-02 2.59477735e-01 6.26766860e-01 4.67629582e-01 5.57007134e-01 1.20426273e+00 -1.41005206e+00 -4.74990666e-01 8.21482360e-01 -2.16540962e-01 -2.63537467e-01 2.23142561e-03 1.43555790e-01 6.99846983e-01 -1.14792824e+00 8.95722628e-01 1.03699422e+00 7.29437470e-01 8.08845878e-01 -1.76688063e+00 -6.40231311e-01 -3.74149173e-01 7.25634322e-02 -1.10618436e+00 -7.71103024e-01 9.10123885e-01 -6.63802505e-01 8.87244344e-01 6.09861612e-02 2.05180913e-01 1.75326610e+00 -1.66891977e-01 8.29105079e-01 1.52995443e+00 -9.58499849e-01 9.17080998e-01 6.34003460e-01 4.90723215e-02 1.90448582e-01 1.62143126e-01 -2.57540960e-02 -5.36995053e-01 -4.96409595e-01 3.31323177e-01 1.57837853e-01 3.50783646e-01 -6.36387765e-01 -1.35811400e+00 9.96705234e-01 -1.91510573e-01 3.35445821e-01 -4.48479444e-01 2.59664178e-01 5.54451048e-01 1.95514858e-01 9.59570646e-01 7.49750853e-01 -3.10825169e-01 -2.33721405e-01 -1.52577019e+00 3.23315501e-01 5.13586700e-01 1.21927297e+00 8.58950973e-01 2.15445280e-01 -4.82897431e-01 1.01310182e+00 1.92296728e-01 6.34726658e-02 9.86998975e-01 -1.07812941e+00 4.67653215e-01 6.32556498e-01 3.44919622e-01 -2.26693973e-01 -9.22897756e-02 -3.98360193e-01 -6.15819454e-01 2.68113434e-01 1.64419621e-01 -1.95355654e-01 -1.20189321e+00 1.62443852e+00 3.15913200e-01 3.01628888e-01 3.68599176e-01 1.83230974e-02 8.05300176e-01 5.18044829e-01 3.98566276e-01 -3.38365316e-01 8.74771118e-01 -1.29625046e+00 -6.98437691e-01 -2.67863691e-01 7.07020402e-01 -6.54268920e-01 1.31120360e+00 3.24745834e-01 -1.22323000e+00 -9.51187491e-01 -1.05519891e+00 2.20492512e-01 -6.25612199e-01 2.41884157e-01 6.08414054e-01 8.17089319e-01 -8.21475506e-01 6.86505854e-01 -7.92734921e-01 -4.26652938e-01 4.21267092e-01 2.62738019e-02 -4.84969020e-02 9.70220789e-02 -9.63610709e-01 5.55535674e-01 8.69446278e-01 -7.75265038e-01 -1.21044457e+00 -7.52954960e-01 -7.59660602e-01 -2.23622054e-01 2.65301585e-01 -7.06063449e-01 1.52399635e+00 -1.04905188e+00 -1.42001784e+00 9.12838578e-01 -1.98451057e-01 -6.69802487e-01 7.33709157e-01 -1.48460850e-01 -2.37134844e-01 -2.17655048e-01 3.12932730e-01 8.72743368e-01 1.30906677e+00 -1.54313755e+00 -4.99172270e-01 -8.52259900e-03 -4.13329810e-01 1.33102804e-01 -5.19260526e-01 -1.63341150e-01 -5.47237620e-02 -1.00182557e+00 -2.02440739e-01 -9.12588477e-01 -4.18115556e-01 -7.79593706e-01 -4.19671118e-01 -3.13585937e-01 8.20098698e-01 -4.77974154e-02 1.09549224e+00 -2.08311176e+00 9.15151031e-04 -1.68591142e-01 1.20819978e-01 2.86855459e-01 -7.37863183e-02 8.08281064e-01 -2.93836236e-01 2.35786259e-01 -2.76299149e-01 -9.88830745e-01 9.45390202e-03 7.20607638e-02 -5.70787787e-01 2.11918261e-02 -1.92902181e-02 9.04055417e-01 -1.14676046e+00 -6.03557765e-01 3.05941224e-01 1.83147937e-01 -4.10059929e-01 1.39829159e-01 -7.27522671e-01 4.50857699e-01 -8.14322010e-02 2.97904223e-01 3.48621607e-01 -3.44010144e-01 1.29206106e-01 1.17920987e-01 3.12812142e-02 4.33052897e-01 -1.12215567e+00 2.05631042e+00 -5.78439057e-01 6.85913444e-01 -8.64949703e-01 -1.00618672e+00 9.95286345e-01 5.55264056e-01 4.30622250e-01 -1.39717236e-01 -1.00183122e-01 2.13246956e-01 -5.61953068e-01 -3.05874377e-01 5.97385824e-01 -1.99019998e-01 -1.32043138e-01 1.00264037e+00 6.10730469e-01 -5.20790704e-02 4.66267973e-01 2.44050041e-01 9.51153874e-01 5.75051665e-01 3.15799892e-01 -4.29624207e-02 4.78386804e-02 4.28086430e-01 2.81689197e-01 1.08455372e+00 2.25316271e-01 6.92970872e-01 1.21841453e-01 -2.89799809e-01 -1.54849207e+00 -1.20557427e+00 -3.67667992e-03 1.38066339e+00 -4.40424174e-01 -5.92030883e-01 -8.55732024e-01 -9.09430802e-01 -3.93602818e-01 1.37698781e+00 -7.62614608e-01 -3.64415973e-01 -9.59236771e-02 -8.45977068e-01 5.07447779e-01 5.12668669e-01 3.44311774e-01 -1.41570663e+00 -5.74789405e-01 4.77495223e-01 -2.46312574e-01 -9.54994321e-01 -6.13704287e-02 3.60315770e-01 -1.18232179e+00 -7.12821245e-01 -7.76620150e-01 -7.63794243e-01 6.38579845e-01 2.67959952e-01 1.29129469e+00 -5.24949431e-01 -1.75020456e-01 2.89445460e-01 -6.04546726e-01 -6.44305050e-01 -9.11398172e-01 2.68608630e-01 1.06900193e-01 6.80767447e-02 5.52756131e-01 -7.74372220e-01 -2.33365923e-01 1.36962637e-01 -9.98689592e-01 4.59517688e-01 5.76901376e-01 1.01838565e+00 6.05161667e-01 2.51035959e-01 8.88095737e-01 -1.60305238e+00 7.14493275e-01 -6.13992691e-01 -1.00778617e-01 3.25977772e-01 -8.01617563e-01 1.27222732e-01 5.26675761e-01 -7.45149612e-01 -1.52728510e+00 1.99885547e-01 4.58696216e-01 -5.08521140e-01 -6.96183681e-01 2.32490405e-01 9.21100006e-02 5.75810373e-01 1.36909747e+00 5.00714660e-01 -2.95807451e-01 -6.09286308e-01 6.48072004e-01 8.36159408e-01 2.45871723e-01 -4.56163257e-01 8.70933831e-01 6.66475713e-01 -3.98143023e-01 -8.33893836e-01 -8.64212751e-01 -4.64323401e-01 -9.10119295e-01 -8.65335837e-02 7.11041093e-01 -9.45104778e-01 3.79546106e-01 2.15784803e-01 -9.20068383e-01 -7.21485436e-01 -8.13948691e-01 4.38229442e-01 -9.63346183e-01 -2.88092978e-02 -2.70148665e-01 -9.09228504e-01 -2.72884309e-01 -8.71987879e-01 1.17826843e+00 -4.14991602e-02 -6.80709839e-01 -1.10977376e+00 3.68112594e-01 4.71405953e-01 2.62045264e-01 4.25433517e-01 7.43506491e-01 -9.50094044e-01 -2.62139112e-01 -2.46264651e-01 3.03109884e-01 2.49540210e-01 3.62600535e-01 1.86892673e-02 -1.32000148e+00 -2.81403035e-01 -9.28785354e-02 -8.22478890e-01 1.00208807e+00 2.65540063e-01 1.11677563e+00 -3.61206025e-01 -4.41436678e-01 2.82846659e-01 1.14941108e+00 1.44475311e-01 5.56297660e-01 4.03022677e-01 3.08800012e-01 7.22786248e-01 8.27583313e-01 5.11743367e-01 1.66650981e-01 5.30136585e-01 -1.03923842e-01 -4.99830246e-02 -2.20607042e-01 -4.53654200e-01 3.55431885e-01 5.55575848e-01 -8.86970088e-02 2.25367658e-02 -9.98493969e-01 7.04163134e-01 -2.06887889e+00 -1.31844234e+00 -1.41717005e-03 2.33461428e+00 1.17526543e+00 3.62652361e-01 3.20622325e-01 2.83695489e-01 6.21015131e-01 -1.26819694e-02 -4.21151757e-01 -2.11946517e-01 1.75123394e-01 4.39470440e-01 1.02505520e-01 5.67455627e-02 -1.29866123e+00 1.08434403e+00 6.38292885e+00 1.13624918e+00 -6.62401617e-01 6.58722997e-01 5.89373589e-01 -1.84308678e-01 -2.38273874e-01 1.25924498e-01 -8.48455250e-01 5.18610597e-01 1.22565007e+00 -1.29279420e-01 6.08735718e-02 1.14941585e+00 1.12666145e-01 -1.13675326e-01 -1.26924539e+00 6.45133376e-01 1.39286473e-01 -1.52496159e+00 1.25581771e-01 -4.94489335e-02 1.46545136e+00 -1.15679845e-01 1.96348578e-01 7.14830041e-01 7.79365242e-01 -8.45668375e-01 6.74809933e-01 4.64834005e-01 9.81662273e-01 -7.77888477e-01 4.46383268e-01 7.75854707e-01 -7.66877592e-01 8.71625319e-02 -3.25266838e-01 1.22643553e-01 -1.14593826e-01 4.23324317e-01 -1.18940997e+00 2.35669911e-01 4.87751961e-01 4.92046863e-01 -8.27572167e-01 1.00829399e+00 -2.29761377e-01 8.89700294e-01 8.06903839e-02 1.27416700e-01 1.44654706e-01 1.93644002e-01 3.41411740e-01 1.30504489e+00 1.55637309e-01 -5.50973475e-01 2.06717715e-01 9.72804725e-01 3.98840569e-02 -1.93684502e-03 -8.47032666e-01 -3.90167981e-01 5.94450235e-01 1.20520592e+00 -8.63909066e-01 -8.47001255e-01 -9.39519256e-02 9.19853389e-01 1.49659559e-01 4.82507735e-01 -7.46693969e-01 -4.54873979e-01 2.29387227e-02 8.23914036e-02 5.04453555e-02 -1.07856750e-01 -4.54835087e-01 -1.19646263e+00 -3.56899828e-01 -9.12993252e-01 3.33317965e-01 -6.81982458e-01 -1.32561350e+00 5.28146982e-01 3.40366423e-01 -1.65654063e+00 -8.53258967e-01 -2.18944228e-03 -7.64745235e-01 5.87006271e-01 -9.59512889e-01 -1.47921598e+00 -4.40859079e-01 3.24770033e-01 1.16148436e+00 -6.24865890e-01 9.93459404e-01 -9.54340249e-02 -1.90151021e-01 4.91500020e-01 4.19474959e-01 -1.59044683e-01 9.13816333e-01 -1.42979014e+00 6.97680950e-01 5.20043790e-01 5.58705330e-01 6.48249924e-01 8.10613990e-01 -7.01688766e-01 -6.96410775e-01 -1.50588071e+00 7.72775412e-01 -8.38123560e-01 4.06138688e-01 -7.48893440e-01 -7.10710227e-01 8.19257081e-01 2.66434640e-01 -3.71788323e-01 1.23576093e+00 9.69757736e-02 -3.20874512e-01 2.07618430e-01 -1.07279992e+00 6.28713012e-01 8.55212271e-01 -4.95154321e-01 -8.92862916e-01 7.05775142e-01 6.88087940e-01 3.20547186e-02 -5.69147587e-01 2.77524412e-01 3.02794755e-01 -9.21032190e-01 1.01007307e+00 -9.35060084e-01 7.57372200e-01 -1.32139875e-02 -3.47582176e-02 -1.58282650e+00 -2.71574706e-01 -6.85637414e-01 -4.63324070e-01 1.61358774e+00 5.52775025e-01 -2.52415806e-01 1.07364154e+00 3.48355800e-01 2.99371704e-02 -3.19146901e-01 -5.35543859e-01 -9.66179550e-01 7.28623271e-02 -5.88968754e-01 4.30360436e-01 1.23033440e+00 -4.38761264e-02 5.66042781e-01 -4.06983584e-01 -4.86863047e-01 1.20828485e+00 4.37839329e-02 1.14860940e+00 -1.36211038e+00 -4.62248832e-01 -3.12100172e-01 1.32482737e-01 -2.94259369e-01 1.97089925e-01 -8.97492945e-01 -7.70680755e-02 -1.52981222e+00 2.92723566e-01 -6.50046229e-01 -2.25472271e-01 4.85427409e-01 -8.34124237e-02 4.55618441e-01 -1.33070648e-01 5.87426424e-01 -8.11053336e-01 6.46870911e-01 7.31650591e-01 -1.82343006e-01 -6.72608972e-01 2.56938487e-01 -5.36853790e-01 1.01168656e+00 1.08027649e+00 -7.64601290e-01 -7.60254622e-01 -5.52527085e-02 1.50156552e-02 -2.26050273e-01 2.36168921e-01 -1.29116106e+00 1.95736706e-01 -2.62204707e-01 5.47590792e-01 -5.86640894e-01 5.75600505e-01 -4.65071976e-01 3.04177284e-01 3.16161871e-01 -7.63065577e-01 -3.18704456e-01 -3.44633050e-02 6.97951078e-01 -9.65471566e-03 -5.93431354e-01 8.07107866e-01 -4.80102986e-01 -6.51133418e-01 -8.27787295e-02 -4.63627458e-01 3.16322058e-01 1.27424777e+00 -3.41835469e-01 -9.36794877e-02 -2.60524988e-01 -9.91322160e-01 -3.00589859e-01 5.69096506e-01 4.81517375e-01 4.47739244e-01 -1.42502344e+00 -7.31906533e-01 1.30161658e-01 4.81676221e-01 -1.10910706e-01 4.29087132e-02 4.05242324e-01 -7.54161039e-03 2.51937479e-01 -1.03757732e-01 -5.83202839e-01 -9.80817974e-01 5.91538906e-01 -2.09786564e-01 -4.28569198e-01 -4.58678246e-01 5.50880909e-01 -1.07120067e-01 -5.38253486e-01 2.71180451e-01 1.57625303e-01 -2.55351782e-01 3.41363788e-01 5.19496679e-01 6.17210031e-01 -4.01142389e-02 -3.11495543e-01 2.57475257e-01 -7.68233985e-02 -2.20593974e-01 -5.40155053e-01 1.36193740e+00 2.37102553e-01 2.63443947e-01 1.08497560e+00 7.71141052e-01 -3.15670878e-01 -1.12920737e+00 -4.95248049e-01 4.75039005e-01 -3.49202991e-01 -1.83110803e-01 -6.90196693e-01 -2.22901449e-01 1.02201879e+00 5.08324325e-01 2.35401019e-01 8.32882345e-01 1.17917128e-01 3.90606284e-01 5.95416009e-01 4.23234195e-01 -1.54292870e+00 6.22566521e-01 1.88901007e-01 5.75825095e-01 -1.31759036e+00 3.26266997e-02 3.64502706e-02 -8.54261458e-01 7.71481037e-01 7.69921720e-01 -6.81263730e-02 4.03622091e-01 1.64268658e-01 -1.62233576e-01 3.83807421e-02 -1.11853909e+00 -1.80579379e-01 4.07426119e-01 8.43568027e-01 5.96720338e-01 2.12347228e-02 -1.72063366e-01 4.86826807e-01 -9.42453817e-02 2.16521934e-01 4.61477757e-01 1.24096107e+00 -6.12647772e-01 -1.36514461e+00 -2.61020094e-01 6.52369559e-01 -2.17219234e-01 -1.38636172e-01 -4.15126294e-01 5.35781026e-01 2.70300508e-01 9.38655436e-01 -3.61716561e-02 -2.69434720e-01 -2.59401761e-02 6.99728966e-01 2.95080423e-01 -1.33717966e+00 -4.87705678e-01 1.43848345e-01 5.20914681e-02 -8.61106515e-02 -6.06063366e-01 -7.31169939e-01 -8.51788044e-01 1.67959854e-01 -3.85684371e-01 3.62807274e-01 6.18072212e-01 9.65201616e-01 3.56620729e-01 5.20965278e-01 5.32207072e-01 -1.12822294e+00 -5.33406019e-01 -1.44643486e+00 -4.26722437e-01 5.77353597e-01 -1.46533728e-01 -9.20994759e-01 -1.31065041e-01 5.82567811e-01]	[9.958028793334961, 3.08443021774292]
f754dc45-4a3f-41f9-b45b-15833e10fc73	dont-search-for-a-search-method-simple	null	null	https://aclanthology.org/2021.emnlp-main.647	https://aclanthology.org/2021.emnlp-main.647.pdf	Don’t Search for a Search Method — Simple Heuristics Suffice for Adversarial Text Attacks	Recently more attention has been given to adversarial attacks on neural networks for natural language processing (NLP). A central research topic has been the investigation of search algorithms and search constraints, accompanied by benchmark algorithms and tasks. We implement an algorithm inspired by zeroth order optimization-based attacks and compare with the benchmark results in the TextAttack framework. Surprisingly, we find that optimization-based methods do not yield any improvement in a constrained setup and slightly benefit from approximate gradient information only in unconstrained setups where search spaces are larger. In contrast, simple heuristics exploiting nearest neighbors without querying the target function yield substantial success rates in constrained setups, and nearly full success rate in unconstrained setups, at an order of magnitude fewer queries. We conclude from these results that current TextAttack benchmark tasks are too easy and constraints are too strict, preventing meaningful research on black-box adversarial text attacks.	['Artem Sokolov', 'Sebastian Ebert', 'Stefan Riezler', 'Nathaniel Berger']	null	null	null	null	emnlp-2021-11	['adversarial-text']	['adversarial']	[ 3.23758513e-01 -2.86113820e-03 -3.34503055e-01 -3.55665594e-01 -1.06989610e+00 -1.18973219e+00 7.45294452e-01 1.27661675e-01 -1.19931757e+00 7.94779301e-01 6.31415769e-02 -6.13543153e-01 -8.09914693e-02 -5.95742047e-01 -8.68941784e-01 -6.82527125e-01 -2.23669708e-02 6.29720688e-01 1.92149758e-01 -2.15125382e-01 3.30960453e-01 6.53758347e-01 -8.24395716e-01 1.05383836e-01 4.69302267e-01 8.69602323e-01 -3.92251194e-01 9.59967196e-01 -2.31851824e-02 3.57671738e-01 -8.48256588e-01 -9.16397333e-01 7.56500483e-01 -7.98352212e-02 -9.97122347e-01 -7.82764554e-01 1.15367544e+00 -2.13042811e-01 -7.69691229e-01 1.52512968e+00 8.09760392e-01 4.41785246e-01 3.29230279e-01 -1.21831703e+00 -6.62944317e-01 8.37728918e-01 -1.07407838e-01 4.09086376e-01 2.24243790e-01 3.24894249e-01 1.26961827e+00 -6.09637260e-01 5.65258384e-01 1.37465501e+00 5.60355902e-01 8.94404471e-01 -1.27971387e+00 -7.43299127e-01 3.13402683e-01 2.24555820e-01 -9.96465385e-01 -5.47459364e-01 4.86768931e-01 1.82804346e-01 1.22065306e+00 8.07665408e-01 1.34929925e-01 1.47096789e+00 2.41670117e-01 9.56802607e-01 7.40951061e-01 -2.63401926e-01 4.39300150e-01 4.35561202e-02 1.84477583e-01 6.73736393e-01 4.29542243e-01 5.43713987e-01 -5.40455878e-01 -4.77597684e-01 1.90777794e-01 -1.80778340e-01 -2.54638493e-01 -3.96984965e-01 -1.14350426e+00 1.26262105e+00 5.41506529e-01 2.48441160e-01 3.84206534e-03 6.00364208e-01 7.55110502e-01 6.72102869e-01 3.80455226e-01 1.14837909e+00 -8.78705263e-01 -3.22781131e-02 -1.00345051e+00 8.17741036e-01 1.25196552e+00 7.96367824e-01 3.49135548e-01 2.63210833e-01 -2.83935219e-01 5.04052877e-01 -9.76156667e-02 7.45539546e-01 4.41407770e-01 -9.83196974e-01 8.57912838e-01 -1.65316045e-01 -6.47691265e-02 -1.21536267e+00 -5.51949918e-01 -4.08932209e-01 -7.77488768e-01 3.20645511e-01 7.93245614e-01 -5.74800253e-01 -9.96872962e-01 1.86234295e+00 1.53893977e-01 -1.95726678e-01 1.36458308e-01 8.56916904e-01 5.25134265e-01 4.76438195e-01 -8.01598281e-03 -1.11892801e-02 1.27882564e+00 -1.00596571e+00 -4.83927935e-01 -6.37719393e-01 7.16922998e-01 -7.65304804e-01 1.06959105e+00 4.37680483e-01 -1.14147687e+00 1.36412546e-01 -1.02577543e+00 3.41309607e-02 -1.02265918e+00 -6.25953913e-01 8.43802929e-01 1.02786875e+00 -1.02090836e+00 8.05442750e-01 -8.31775725e-01 -1.73722893e-01 5.43031216e-01 4.62716788e-01 -3.01081359e-01 -8.23340416e-02 -1.50726080e+00 9.69117165e-01 5.23190677e-01 -3.94407995e-02 -8.41031134e-01 -8.14951837e-01 -7.30786204e-01 1.69140458e-01 7.77904510e-01 -4.95429367e-01 1.14308226e+00 -5.73033571e-01 -1.33005011e+00 8.19847584e-01 1.09875128e-01 -1.12159312e+00 7.43681848e-01 -3.80286276e-01 -1.18500218e-01 8.56154859e-02 -2.77791113e-01 6.30158246e-01 8.57287228e-01 -8.48017037e-01 -3.70954573e-01 -4.16144788e-01 3.29941809e-01 1.63940549e-01 -4.29058105e-01 3.16657275e-01 -2.39203632e-01 -9.22650099e-01 -1.56866565e-01 -1.06466150e+00 -6.12158895e-01 1.38847440e-01 -8.97351146e-01 -1.94171797e-02 6.11110449e-01 -2.74492055e-01 1.01887751e+00 -1.81550980e+00 8.64246674e-03 3.26305807e-01 3.03923458e-01 5.52233934e-01 -4.09458727e-01 3.42960209e-01 -1.16210379e-01 6.23804390e-01 -2.08036393e-01 -1.27044827e-01 5.96669316e-01 1.78619340e-01 -7.99464285e-01 5.08951426e-01 -1.85129613e-01 1.00162303e+00 -7.47829974e-01 -2.98935622e-01 -1.08465120e-01 -8.25931877e-02 -8.56360078e-01 -3.02405179e-01 -5.65919876e-01 -1.58207729e-01 -4.74444866e-01 6.50841415e-01 6.02129281e-01 -2.71787439e-02 -2.23136559e-01 1.43667743e-01 3.47353369e-01 2.19258174e-01 -9.29996908e-01 1.63030100e+00 -3.12945873e-01 1.03613687e+00 2.47364059e-01 -1.05535960e+00 3.02582175e-01 8.61518458e-02 5.94995096e-02 -5.28964043e-01 7.33128712e-02 7.92919546e-02 3.99677232e-02 -2.54917413e-01 5.14296949e-01 3.94206226e-01 -1.74046054e-01 4.14460838e-01 -2.76145548e-01 -3.57390612e-01 1.67595938e-01 3.21500927e-01 1.58423448e+00 -5.39991677e-01 -9.58273858e-02 -1.22618414e-01 3.00424635e-01 1.83844298e-01 1.82540342e-01 1.52674472e+00 -4.13456619e-01 3.93780589e-01 5.92842042e-01 -7.91677237e-01 -8.52110624e-01 -1.04141033e+00 -1.46235436e-01 1.53767574e+00 8.23312551e-02 -5.15788198e-01 -9.33398664e-01 -1.08601475e+00 1.44707963e-01 7.39397585e-01 -7.41880715e-01 -3.32513839e-01 -6.39403701e-01 -9.87557948e-01 1.22496021e+00 3.69797677e-01 6.23523295e-01 -1.23824716e+00 -4.03994530e-01 -5.67082129e-02 2.22503290e-01 -1.20132005e+00 -8.05025578e-01 5.70685983e-01 -6.97343111e-01 -9.80468273e-01 -6.57532394e-01 -7.45208383e-01 5.07850647e-01 -1.74131781e-01 1.08026564e+00 9.09232302e-04 -3.67242217e-01 2.22952485e-01 -4.39092191e-03 -5.13028204e-01 -2.96738267e-01 3.98955822e-01 2.28922982e-02 -5.75094342e-01 3.97300303e-01 -3.36944371e-01 -5.21549702e-01 1.77905977e-01 -1.04702532e+00 -7.38438666e-01 5.52023053e-01 1.17859805e+00 3.04182112e-01 -4.38690651e-03 2.49031320e-01 -1.15023673e+00 9.97672558e-01 -1.69601962e-01 -9.92952108e-01 3.45588893e-01 -7.03373134e-01 4.21231568e-01 8.98151875e-01 -8.17884982e-01 -4.97303873e-01 1.01106770e-01 -1.40850365e-01 -5.41999102e-01 -1.43354267e-01 2.66451716e-01 -6.09740987e-02 -4.89737064e-01 1.23418534e+00 1.76249251e-01 -3.67573351e-01 -1.04925968e-01 4.75449622e-01 2.54949685e-02 5.63134789e-01 -8.43024492e-01 1.13216174e+00 3.94644380e-01 1.29042953e-01 -5.55304170e-01 -8.09520125e-01 -1.77118838e-01 -1.48517057e-01 5.53168774e-01 8.28066349e-01 -3.67674679e-01 -8.84046853e-01 2.82851845e-01 -1.23723185e+00 -5.38354993e-01 -2.49467537e-01 3.44746709e-01 -4.67790544e-01 5.23649514e-01 -7.33463883e-01 -5.11694193e-01 -6.69707716e-01 -1.17161202e+00 8.32529485e-01 -7.26472661e-02 -8.02522153e-02 -1.12079895e+00 -5.71405292e-02 2.79696733e-01 5.71278930e-01 1.74752057e-01 8.48831773e-01 -1.51939034e+00 -5.76519608e-01 -6.25022411e-01 -1.26442254e-01 1.25227883e-01 -4.09299135e-01 -3.56386423e-01 -1.01126111e+00 -4.49712753e-01 3.17304395e-02 -5.97521126e-01 1.04198980e+00 2.89782584e-01 1.37359726e+00 -9.17124867e-01 -2.91295648e-01 1.00439966e+00 1.25645530e+00 6.09131791e-02 3.61181438e-01 5.02413869e-01 5.36863804e-01 2.47682959e-01 3.55718702e-01 2.20121264e-01 -3.98020029e-01 4.38963205e-01 6.62779272e-01 -8.46569911e-02 4.09067154e-01 -1.34672299e-01 3.64633590e-01 -1.19621105e-01 5.94234467e-01 -7.04425097e-01 -9.62773800e-01 3.14183891e-01 -1.66870511e+00 -1.05523002e+00 3.39511722e-01 2.06621575e+00 1.12001538e+00 6.27904654e-01 -2.03581661e-01 -1.61676124e-01 5.41108131e-01 5.02579987e-01 -9.40999925e-01 -6.77523851e-01 -1.60239980e-01 4.02842373e-01 1.04261386e+00 7.46138930e-01 -1.44720387e+00 1.45671070e+00 7.54369450e+00 1.18101978e+00 -9.78476584e-01 -1.77326694e-01 6.66632831e-01 -5.81373453e-01 -2.11895287e-01 -1.55833140e-01 -7.72057950e-01 2.91184872e-01 8.84411037e-01 -3.21889728e-01 9.67852414e-01 1.05228102e+00 -3.44423711e-01 2.24776700e-01 -1.19868314e+00 8.73839736e-01 1.29091274e-02 -1.46680796e+00 -4.10775729e-02 4.44347523e-02 7.67903209e-01 4.22951132e-01 5.25550485e-01 2.47313485e-01 9.58862484e-01 -1.25123036e+00 3.09967369e-01 -1.08637519e-01 5.77145398e-01 -9.21708941e-01 5.66009581e-01 3.49247396e-01 -5.43778121e-01 -6.54701740e-02 -7.03421116e-01 3.37670505e-01 -2.99851507e-01 3.18208724e-01 -7.70968258e-01 -6.78778514e-02 6.77877843e-01 -5.28297722e-02 -5.99720955e-01 9.68508840e-01 -2.21213773e-01 5.85524261e-01 -7.72587895e-01 -5.41341722e-01 7.01231956e-01 2.94574767e-01 8.65816712e-01 1.36689031e+00 -3.34871799e-01 -2.46165812e-01 2.44437918e-01 8.17359030e-01 -4.75984573e-01 3.26773338e-03 -8.43323946e-01 -1.97505549e-01 6.24197602e-01 1.04153538e+00 -6.39191508e-01 -1.76465705e-01 -1.45688318e-02 9.87662137e-01 3.84399533e-01 6.32868409e-01 -9.18341875e-01 -7.81568348e-01 7.75012434e-01 -4.57784832e-01 2.79871643e-01 -1.43360972e-01 -5.08260489e-01 -1.14484668e+00 7.03430101e-02 -1.30271411e+00 6.14088893e-01 -2.69991726e-01 -1.31029618e+00 5.65059543e-01 -4.22732998e-03 -3.78794938e-01 -4.05735910e-01 -9.43424344e-01 -7.07216859e-01 5.46956956e-01 -1.05604577e+00 -6.19677663e-01 3.80172044e-01 7.80566454e-01 5.37393510e-01 -4.01411712e-01 9.25052464e-01 7.18679884e-03 -5.68101168e-01 1.32290280e+00 5.56421697e-01 4.75017190e-01 6.69012249e-01 -1.47334754e+00 8.77517104e-01 1.05844855e+00 7.00372577e-01 7.80053854e-01 1.03168952e+00 -4.09112990e-01 -1.53622401e+00 -7.65674055e-01 7.33485937e-01 -7.07474530e-01 9.77939665e-01 -7.12067127e-01 -6.79714799e-01 5.96379459e-01 1.65870681e-01 2.17955321e-01 3.19710851e-01 2.13066474e-01 -7.27107704e-01 9.34507698e-02 -1.32729924e+00 1.06931758e+00 1.05944633e+00 -6.60217226e-01 -3.89420897e-01 1.05709147e+00 1.12420261e+00 -8.03687692e-01 -5.02331197e-01 5.53589277e-02 4.04761016e-01 -4.76151228e-01 1.28843951e+00 -1.17920911e+00 1.38234839e-01 3.90336305e-01 -2.02122331e-01 -1.22902763e+00 1.08366221e-01 -1.14328885e+00 6.67711571e-02 6.53783143e-01 7.34471679e-01 -8.73027921e-01 1.32716060e+00 1.03154528e+00 7.42608830e-02 -7.24238932e-01 -1.22098458e+00 -8.86687756e-01 6.72807992e-01 -6.11362517e-01 3.28465551e-01 1.01947343e+00 -4.41062450e-02 1.48035854e-01 -3.00055325e-01 2.04031006e-01 5.49640954e-01 -1.55062675e-01 6.24577224e-01 -7.88328588e-01 -4.22568977e-01 -7.15351164e-01 -2.84212828e-01 -9.89034235e-01 6.12518668e-01 -9.24845338e-01 1.21830389e-01 -8.82895947e-01 -3.24951172e-01 -3.00237924e-01 -1.21376291e-01 6.96582615e-01 -9.07124877e-02 1.40944839e-01 3.69211227e-01 -1.53298900e-01 -6.91807806e-01 1.84361413e-01 9.70065653e-01 -4.89480704e-01 5.72033785e-02 2.32802033e-01 -6.88371837e-01 8.01119626e-01 9.64775622e-01 -7.87490070e-01 -4.35302407e-01 -6.94243014e-01 3.66375297e-01 -1.47219509e-01 3.23466420e-01 -9.09498632e-01 8.30261230e-01 -3.04567695e-01 3.75123054e-01 -3.21383893e-01 4.35454518e-01 -9.28118348e-01 -6.38327479e-01 6.61849380e-01 -1.03946006e+00 3.25062752e-01 3.96426767e-01 6.88519299e-01 1.33187294e-01 -4.86639887e-01 7.55525649e-01 -1.40164658e-01 -2.20760196e-01 5.56381583e-01 -3.10465425e-01 6.51653528e-01 7.60963142e-01 1.01204425e-01 -5.62264621e-01 -5.07838368e-01 -9.28805709e-01 3.84236723e-01 1.63053855e-01 2.23258421e-01 5.23133755e-01 -9.57822442e-01 -6.42612755e-01 1.51268780e-01 -2.37008989e-01 -7.10880905e-02 -1.99946225e-01 2.65749902e-01 -7.04920709e-01 6.99729204e-01 2.11920515e-01 -1.35750532e-01 -1.30268645e+00 9.10949647e-01 4.90091234e-01 -6.86010301e-01 -2.93103516e-01 1.11839485e+00 6.35872083e-03 -5.95353365e-01 9.50123489e-01 -2.35570237e-01 3.68294150e-01 -2.17802137e-01 5.36619008e-01 2.48357147e-01 6.96466565e-02 9.95919947e-03 -2.56851643e-01 6.45969063e-02 -5.45451880e-01 -2.83232450e-01 1.13486803e+00 5.20745695e-01 1.96823075e-01 -2.55045474e-01 1.32740200e+00 2.52252042e-01 -8.25966179e-01 -4.12936717e-01 2.04126388e-01 -4.06865656e-01 4.78432253e-02 -9.76425409e-01 -1.27487016e+00 8.84474456e-01 4.82887566e-01 4.25607681e-01 8.41492891e-01 -2.99696952e-01 8.98948312e-01 1.55738890e+00 1.51292101e-01 -1.06093633e+00 -3.21193933e-02 8.01530540e-01 8.96420836e-01 -1.19572926e+00 1.07444085e-01 -1.39970362e-01 -3.69259745e-01 9.71674442e-01 6.25380516e-01 -3.39293182e-01 6.21733427e-01 4.46226299e-01 -5.14277108e-02 -1.42661572e-01 -8.90897155e-01 1.10220514e-01 1.74990699e-01 6.33834898e-01 -2.17128843e-01 -4.08045888e-01 2.38141958e-02 1.70865729e-01 -5.27435124e-01 -6.74619734e-01 1.67925693e-02 7.27407396e-01 -3.83971483e-01 -1.13068163e+00 -4.68268454e-01 4.30393785e-01 -1.14860666e+00 -6.54157162e-01 -8.09886515e-01 1.03025925e+00 -5.70403934e-01 8.75913322e-01 -1.80000186e-01 -3.32832217e-01 8.80841911e-02 9.94508862e-02 2.79930949e-01 -3.25519115e-01 -1.02348399e+00 -4.05690581e-01 1.91846684e-01 -8.41907501e-01 3.96763086e-01 -4.24229652e-01 -1.03900170e+00 -6.13478959e-01 -4.80982989e-01 2.55786210e-01 8.53680789e-01 7.92738914e-01 2.35714912e-01 8.94934908e-02 3.49467486e-01 -5.91285706e-01 -1.32638776e+00 -5.64623594e-01 -1.14069134e-01 1.15232497e-01 4.66767281e-01 2.79830247e-02 -7.94136107e-01 -3.26409698e-01]	[5.99072790145874, 8.103569984436035]
b2c8262d-d41a-4728-915b-db15be538537	differentiable-tree-operations-promote	2306.00751	null	https://arxiv.org/abs/2306.00751v1	https://arxiv.org/pdf/2306.00751v1.pdf	Differentiable Tree Operations Promote Compositional Generalization	In the context of structure-to-structure transformation tasks, learning sequences of discrete symbolic operations poses significant challenges due to their non-differentiability. To facilitate the learning of these symbolic sequences, we introduce a differentiable tree interpreter that compiles high-level symbolic tree operations into subsymbolic matrix operations on tensors. We present a novel Differentiable Tree Machine (DTM) architecture that integrates our interpreter with an external memory and an agent that learns to sequentially select tree operations to execute the target transformation in an end-to-end manner. With respect to out-of-distribution compositional generalization on synthetic semantic parsing and language generation tasks, DTM achieves 100% while existing baselines such as Transformer, Tree Transformer, LSTM, and Tree2Tree LSTM achieve less than 30%. DTM remains highly interpretable in addition to its perfect performance.	['Jianfeng Gao', 'Paul Smolensky', 'Roland Fernandez', 'Yunmo Chen', 'Kate McCurdy', 'Edward Hu', 'Paul Soulos']	2023-06-01	null	null	null	null	['semantic-parsing']	['natural-language-processing']	[ 8.12257648e-01 7.56800413e-01 -7.55354390e-02 -6.96219802e-01 -1.07349503e+00 -7.92256236e-01 7.35416591e-01 -2.51857996e-01 -2.60292124e-02 5.22260606e-01 2.37862408e-01 -9.14824188e-01 3.46637338e-01 -1.03678644e+00 -1.20325232e+00 -2.87282795e-01 -2.13756412e-01 8.89566362e-01 -1.06619811e-02 -2.24108338e-01 -1.73156098e-01 3.54914039e-01 -1.08600569e+00 6.96122527e-01 1.03795946e+00 9.53038096e-01 1.07600160e-01 8.08939099e-01 -5.27340472e-01 1.24091589e+00 -4.68871504e-01 -6.68509781e-01 1.55901611e-01 -5.22644281e-01 -1.23458219e+00 -1.87129170e-01 5.23438215e-01 -4.81807500e-01 -2.77125627e-01 1.04227161e+00 -3.80772352e-02 4.05463437e-03 8.12882066e-01 -1.25392520e+00 -8.96470070e-01 1.30322146e+00 -1.79788187e-01 -8.37478861e-02 2.68720090e-01 3.97843570e-01 1.35569561e+00 -7.55506814e-01 5.79473734e-01 1.82747400e+00 4.65647638e-01 5.47871351e-01 -1.71280134e+00 -4.54450816e-01 3.09294045e-01 -3.30928206e-01 -6.19627953e-01 -1.69054493e-01 4.38879281e-01 -4.37547743e-01 1.29793131e+00 1.24174282e-01 2.87218153e-01 1.53804886e+00 1.90836504e-01 9.00710106e-01 8.53124678e-01 -2.56933093e-01 8.13761428e-02 -4.70171481e-01 1.64483592e-01 1.05912173e+00 -1.15920931e-01 1.49770126e-01 -3.94987017e-01 -7.63766840e-02 9.78727996e-01 -3.70831728e-01 2.47177497e-01 -2.73727298e-01 -1.49313188e+00 8.20564687e-01 7.41427660e-01 -1.35900415e-02 -1.56445429e-01 8.63256812e-01 7.25016713e-01 4.63482141e-01 1.50134787e-01 6.75004840e-01 -5.25642872e-01 -1.95929527e-01 -5.40178537e-01 3.79217684e-01 8.25993419e-01 1.20471084e+00 4.68813002e-01 6.82355940e-01 -3.42485696e-01 4.21848238e-01 6.45137951e-02 7.15902448e-01 3.73473763e-01 -1.07200325e+00 8.16685557e-01 6.28404975e-01 -2.75846958e-01 -4.61179405e-01 -1.15182154e-01 -4.03445870e-01 -7.88939536e-01 1.60436213e-01 3.88374299e-01 -2.12285910e-02 -1.26170266e+00 1.81581414e+00 1.64977342e-01 8.39468837e-03 2.26722717e-01 4.90953326e-01 5.90964615e-01 8.67665231e-01 4.73619163e-01 4.91323590e-01 1.25030398e+00 -1.11766016e+00 -2.40129545e-01 -3.75954479e-01 8.80593538e-01 -2.44696409e-01 1.57416499e+00 1.27678826e-01 -1.28447413e+00 -4.20757443e-01 -8.44922543e-01 -5.27002811e-01 -1.86700672e-01 2.90603414e-02 9.84181821e-01 3.26639861e-01 -1.04940176e+00 7.72180200e-01 -1.31317735e+00 5.28069809e-02 3.98491114e-01 4.42513704e-01 -1.59854874e-01 3.00376505e-01 -9.02218282e-01 8.99384260e-01 7.63141215e-01 1.96327530e-02 -1.35392225e+00 -1.05395627e+00 -1.22981465e+00 1.47903070e-01 1.17075600e-01 -1.24213302e+00 1.87180567e+00 -1.12222779e+00 -1.76569498e+00 8.02064776e-01 -1.97119325e-01 -9.24193621e-01 4.79763329e-01 -1.35735646e-01 -3.55328172e-02 2.57691718e-03 2.04875931e-01 8.70937765e-01 8.51002216e-01 -8.55431378e-01 -6.29120111e-01 1.87726188e-02 1.85271665e-01 3.56804319e-02 7.52375424e-02 6.92161098e-02 1.78385064e-01 -9.51779723e-01 3.36483009e-02 -8.40593100e-01 -2.50469595e-01 -3.41825843e-01 -6.45091116e-01 -2.08758086e-01 7.07542717e-01 -8.39126647e-01 8.53350341e-01 -1.84811831e+00 6.72227085e-01 -1.50005650e-02 2.02711359e-01 -2.06565529e-01 -3.80111903e-01 3.99041384e-01 -2.68195093e-01 1.21926047e-01 -8.49671006e-01 -3.26411426e-01 3.96980584e-01 3.98798794e-01 -9.02953923e-01 -2.17003465e-01 6.09942794e-01 1.40700674e+00 -1.14689755e+00 -3.71853739e-01 -1.11528179e-02 1.16916612e-01 -7.30252922e-01 2.76961386e-01 -9.71467137e-01 4.07391369e-01 -4.92495358e-01 7.43701935e-01 1.52332455e-01 -2.75603384e-01 3.27814966e-01 -4.29060720e-02 3.20575118e-01 9.96123731e-01 -4.59435165e-01 1.77973175e+00 -8.77959430e-01 4.09712225e-01 2.96822097e-02 -8.87050629e-01 6.53156579e-01 1.08365491e-01 -3.26951966e-02 -5.74667335e-01 -1.10742226e-01 4.78767008e-01 8.97429511e-02 7.15867952e-02 4.89312857e-01 -3.13765824e-01 -3.72425795e-01 4.27986592e-01 1.31893307e-01 -4.59120512e-01 3.40521224e-02 2.24528313e-01 1.32063842e+00 5.59665620e-01 -1.21939607e-01 -1.50026515e-01 2.86072433e-01 3.20290923e-02 1.36762813e-01 5.49048066e-01 5.36886632e-01 3.33424389e-01 8.09340894e-01 -4.93337393e-01 -1.17321372e+00 -1.67765141e+00 2.99867749e-01 1.41856003e+00 -5.05597949e-01 -1.91005901e-01 -1.01752937e+00 -7.31334329e-01 1.29599124e-01 1.37383699e+00 -5.01934588e-01 -2.78485179e-01 -1.12567472e+00 -4.62398648e-01 9.56741393e-01 7.92601049e-01 4.58752334e-01 -1.26939547e+00 -4.74581301e-01 3.92981201e-01 -5.91992177e-02 -1.23449576e+00 -7.45053232e-01 4.04402494e-01 -1.27345479e+00 -6.43948257e-01 -2.16927245e-01 -1.01142776e+00 8.16871941e-01 -4.48874235e-01 1.52612984e+00 -3.23736101e-01 1.80353224e-01 -1.26025409e-01 8.29223469e-02 -3.16989534e-02 -1.21537828e+00 6.64924860e-01 -3.79298687e-01 -1.77863032e-01 -1.72365829e-01 -9.04078245e-01 -1.43578157e-01 -1.59601927e-01 -8.29330862e-01 6.14571750e-01 4.42408502e-01 1.03584993e+00 4.30090964e-01 -2.22832128e-01 3.21931392e-01 -1.11884224e+00 6.02839172e-01 -1.48802638e-01 -9.73260880e-01 2.71461189e-01 -2.95671642e-01 6.52435422e-01 1.04020023e+00 -3.13466489e-01 -1.13039684e+00 1.11406162e-01 -2.18813065e-02 -2.76578844e-01 2.57804543e-01 4.87259895e-01 -1.21400088e-01 2.26356909e-01 7.24026263e-01 2.42840573e-01 2.82556973e-02 -4.52823758e-01 7.46910870e-01 9.66943651e-02 9.26674306e-01 -1.34310579e+00 1.11630404e+00 1.15094237e-01 1.80222198e-01 -2.14108557e-01 -9.78908360e-01 5.38238168e-01 -5.73671162e-01 4.72483665e-01 9.41460788e-01 -9.06443834e-01 -8.04970026e-01 5.47797382e-01 -1.40923500e+00 -1.01154780e+00 -5.19164264e-01 -7.52623156e-02 -8.13310564e-01 1.46855623e-01 -1.04117012e+00 -4.17496085e-01 -7.08830476e-01 -1.33193350e+00 1.49356663e+00 -3.75117689e-01 -4.64065462e-01 -1.21098864e+00 -5.84568143e-01 2.94915557e-01 4.69364822e-01 6.80260360e-01 1.70009017e+00 -3.99330050e-01 -1.13501048e+00 1.96749911e-01 -7.79947340e-02 4.81019586e-01 1.20655686e-01 -2.15286408e-02 -6.44239783e-01 -1.39984682e-01 -2.58184224e-01 -4.05402124e-01 7.71400928e-01 7.92215616e-02 1.27598715e+00 -6.30309343e-01 -1.23159498e-01 9.89418685e-01 8.13407481e-01 1.31353050e-01 3.80219340e-01 2.71651387e-01 1.06003344e+00 3.91758472e-01 8.93483311e-02 -1.12099141e-01 5.51098049e-01 4.05058205e-01 4.53307360e-01 -8.18740129e-02 -3.15160185e-01 -9.04835463e-01 9.04150188e-01 7.43610859e-01 3.12657624e-01 1.47188932e-01 -9.78043675e-01 2.37021133e-01 -1.55531287e+00 -6.90268099e-01 5.77846207e-02 1.91487813e+00 1.23586679e+00 5.24909377e-01 3.73608470e-02 -1.38858780e-01 3.72311115e-01 1.85468405e-01 -7.20785141e-01 -8.33967865e-01 -1.71455648e-02 7.40567684e-01 6.08518064e-01 7.23025203e-01 -1.04008818e+00 1.47033668e+00 6.81304026e+00 6.79037869e-01 -1.25347590e+00 -8.10354352e-02 6.50815666e-01 1.32711038e-01 -7.75114536e-01 1.08531117e-01 -6.13676727e-01 2.15920940e-01 1.34177351e+00 -2.22110689e-01 8.88305843e-01 9.24450457e-01 -1.80667862e-01 5.00357091e-01 -1.73554230e+00 4.82015431e-01 -6.42760813e-01 -1.47209430e+00 6.75903499e-01 -2.13505194e-01 5.26278317e-01 1.73742607e-01 2.24016249e-01 6.82950616e-01 1.16525424e+00 -1.53983581e+00 1.06904149e+00 1.44049048e-01 1.08247709e+00 -4.86033201e-01 -3.41543667e-02 2.57797658e-01 -1.14848447e+00 -1.46577246e-02 3.85979936e-02 -1.27479926e-01 3.76171887e-01 3.06607842e-01 -1.26002121e+00 4.47682261e-01 2.51850516e-01 6.28986299e-01 -4.62691337e-01 -1.12029061e-01 -6.49116993e-01 8.89878631e-01 -3.85212541e-01 1.39690280e-01 6.27079606e-01 -2.83638179e-01 3.74890596e-01 1.26550913e+00 3.27730983e-01 -2.29680061e-01 2.05841735e-01 1.44564807e+00 -3.92499387e-01 -4.30191934e-01 -6.30200148e-01 -4.95745480e-01 3.73028815e-01 9.03194964e-01 -3.75823528e-01 -6.57848537e-01 -1.57982424e-01 1.05883908e+00 5.55045724e-01 4.91021216e-01 -9.77698147e-01 -3.16780925e-01 7.00963974e-01 1.79421250e-02 2.17855439e-01 -5.35895169e-01 -6.97539508e-01 -1.18081939e+00 2.07115620e-01 -1.28117883e+00 4.19694096e-01 -8.25668097e-01 -9.77203012e-01 7.15332448e-01 2.14612409e-01 -5.91659486e-01 -7.92821229e-01 -7.20149279e-01 -4.70958918e-01 1.09590673e+00 -1.16816700e+00 -1.59527051e+00 2.30890647e-01 1.51188880e-01 6.00526869e-01 -2.64734507e-01 9.91769910e-01 -1.15033882e-02 -3.90263349e-01 7.03315914e-01 -2.75061935e-01 4.73405838e-01 -2.14277878e-01 -1.66894352e+00 1.65888143e+00 8.87085676e-01 4.47198115e-02 7.27554321e-01 5.22313952e-01 -4.82486039e-01 -1.67927146e+00 -1.48664010e+00 8.41242373e-01 -7.71827400e-01 1.05283821e+00 -7.69472539e-01 -8.40173364e-01 1.40077949e+00 1.12871751e-01 -1.34261981e-01 -1.33184910e-01 -6.82398677e-02 -8.41156900e-01 -1.03443801e-01 -8.99148047e-01 1.01753604e+00 1.50086629e+00 -7.74242818e-01 -6.86803877e-01 4.53164041e-01 1.35264790e+00 -1.00371730e+00 -1.06824875e+00 4.74583238e-01 2.54642695e-01 -5.09148896e-01 1.12918437e+00 -1.05764306e+00 8.29835117e-01 -2.17508823e-01 -2.54445285e-01 -1.26602674e+00 -1.13838375e-01 -9.82186556e-01 -1.20412491e-01 1.02666295e+00 7.08802223e-01 -9.23861682e-01 8.01083922e-01 5.72253048e-01 -4.49615926e-01 -7.14671910e-01 -9.07791078e-01 -9.66182947e-01 4.79924619e-01 -2.94008583e-01 8.68198395e-01 5.55774629e-01 -3.34325910e-01 7.18533516e-01 -6.38107061e-02 1.45813316e-01 5.72989583e-01 4.23657238e-01 7.66944289e-01 -7.20995367e-01 -6.33067191e-01 -6.57796919e-01 -5.25828674e-02 -1.32721269e+00 7.64528334e-01 -1.72137523e+00 -2.38691494e-01 -1.48754120e+00 -1.31641611e-01 -2.29393780e-01 4.57389839e-02 7.50279129e-01 4.42084149e-02 -4.62002933e-01 1.79298162e-01 -1.25219196e-01 -1.25114024e-01 7.88634121e-01 1.39933562e+00 -4.15434420e-01 9.46494192e-02 -1.73022687e-01 -6.13085747e-01 6.53641641e-01 5.49237907e-01 -5.01391947e-01 -5.79737782e-01 -9.97783482e-01 6.17948454e-03 3.43108326e-01 4.93692189e-01 -7.09966362e-01 -2.34795347e-01 -2.60594070e-01 -1.02232233e-01 -3.26716334e-01 3.14681590e-01 -4.53482687e-01 2.63313293e-01 6.93966508e-01 -7.44982660e-01 5.22054911e-01 1.13483764e-01 5.90073727e-02 -2.94269592e-01 -7.39619583e-02 7.18866646e-01 -2.83141971e-01 -4.20294285e-01 2.15096325e-01 -1.57839417e-01 5.31433284e-01 5.33841848e-01 1.56529807e-02 -3.01621854e-01 -2.89522856e-01 -6.39624536e-01 2.92052060e-01 2.64300108e-01 4.55032080e-01 2.91944236e-01 -1.25567973e+00 -7.32241333e-01 -3.12669128e-02 -2.15472847e-01 5.03068984e-01 -3.83661509e-01 4.02610481e-01 -8.08881462e-01 4.24172610e-01 1.27777815e-01 -4.93317217e-01 -8.22758913e-01 4.47962523e-01 6.12290740e-01 -4.82461035e-01 -8.37634683e-01 8.11228812e-01 6.75489426e-01 -9.20999646e-01 2.56164283e-01 -1.29077280e+00 7.31518805e-01 -5.37185609e-01 5.39110042e-02 7.82579407e-02 -4.14811783e-02 -1.84056103e-01 -1.95158646e-01 2.74283350e-01 5.25574386e-02 -2.18451425e-01 1.13827300e+00 6.66974127e-01 -4.02997732e-01 3.38472605e-01 1.28025901e+00 -3.46335649e-01 -1.21486342e+00 -1.78600937e-01 4.14192468e-01 1.50002092e-01 -5.33073306e-01 -9.43036199e-01 -8.54855537e-01 8.99587810e-01 -1.99688226e-01 -4.79818918e-02 8.97925138e-01 -9.18056071e-02 1.19836426e+00 7.24990785e-01 3.23763996e-01 -5.48826098e-01 4.76548672e-02 9.90126252e-01 1.15442085e+00 -8.06888521e-01 -5.33235013e-01 -5.79447269e-01 -5.41801512e-01 1.15664232e+00 4.81317788e-01 -1.73537835e-01 2.72440404e-01 6.31910741e-01 -1.14423327e-01 3.50722857e-02 -1.13307869e+00 9.77225080e-02 1.86197773e-01 6.32893085e-01 5.71438551e-01 3.50894272e-01 2.33931556e-01 3.04072231e-01 -8.96761537e-01 -9.83030722e-02 2.72562385e-01 6.11316442e-01 -5.27067222e-02 -1.24512017e+00 -4.10757437e-02 3.18417370e-01 -3.36657077e-01 -4.48748589e-01 -1.97721109e-01 8.14135373e-01 -5.18413723e-01 4.00437027e-01 3.20819877e-02 -2.38477424e-01 5.11020303e-01 3.98743331e-01 6.67395413e-01 -8.34968448e-01 -9.50481594e-01 -4.43779290e-01 6.82888210e-01 -4.81646895e-01 2.55941361e-01 -5.63819468e-01 -1.81515384e+00 -3.46451342e-01 4.58805025e-01 -3.60518768e-02 4.01282430e-01 5.82305968e-01 3.97461772e-01 8.99493098e-01 3.47135544e-01 -6.24415100e-01 -1.39367282e+00 -8.05705845e-01 -1.01995260e-01 5.82487166e-01 2.80126512e-01 -2.00107768e-01 5.32536879e-02 3.80044132e-01]	[10.467679023742676, 8.685667991638184]
3bb0b21a-4c66-44ae-ad6e-68886f079173	unpaired-image-to-image-translation-via-1	2305.15086	null	https://arxiv.org/abs/2305.15086v1	https://arxiv.org/pdf/2305.15086v1.pdf	Unpaired Image-to-Image Translation via Neural Schrödinger Bridge	Diffusion models are a powerful class of generative models which simulate stochastic differential equations (SDEs) to generate data from noise. Although diffusion models have achieved remarkable progress in recent years, they have limitations in the unpaired image-to-image translation tasks due to the Gaussian prior assumption. Schr\"odinger Bridge (SB), which learns an SDE to translate between two arbitrary distributions, have risen as an attractive solution to this problem. However, none of SB models so far have been successful at unpaired translation between high-resolution images. In this work, we propose the Unpaired Neural Schr\"odinger Bridge (UNSB), which combines SB with adversarial training and regularization to learn a SB between unpaired data. We demonstrate that UNSB is scalable, and that it successfully solves various unpaired image-to-image translation tasks. Code: \url{https://github.com/cyclomon/UNSB}	['Jong Chul Ye', 'Kwanyoung Kim', 'Gihyun Kwon', 'Beomsu Kim']	2023-05-24	null	null	null	null	['image-to-image-translation', 'image-to-image-translation']	['computer-vision', 'miscellaneous']	[ 4.02458906e-01 -4.66260426e-02 2.56769419e-01 -2.85551786e-01 -1.16430163e+00 -5.80619276e-01 7.57377088e-01 -7.72504389e-01 -2.17265785e-01 1.04666090e+00 -9.37586054e-02 -3.27892989e-01 1.99990600e-01 -7.54696012e-01 -9.42576766e-01 -1.07720208e+00 4.94968742e-01 5.63041210e-01 -1.87142678e-02 -2.32453763e-01 1.16234254e-02 2.88858950e-01 -1.03387690e+00 1.07958041e-01 1.08733630e+00 6.96204722e-01 2.10192978e-01 7.60583341e-01 1.22642837e-01 8.06085408e-01 -4.46724951e-01 -3.60466272e-01 3.37421507e-01 -1.11155069e+00 -5.68149984e-01 -4.10714477e-01 3.65083218e-01 -2.60698020e-01 -7.08288610e-01 1.24079907e+00 9.41205144e-01 6.38720542e-02 8.91878843e-01 -1.35727477e+00 -1.42859340e+00 3.54247361e-01 -6.58746481e-01 3.54118377e-01 2.59636670e-01 3.14840287e-01 4.45542216e-01 -1.17218375e+00 1.08635247e+00 9.74527597e-01 6.12954021e-01 1.16749644e+00 -1.68275750e+00 -7.80117750e-01 -5.75461924e-01 -1.37092963e-01 -1.45184040e+00 -5.08798242e-01 7.14538813e-01 -3.92158151e-01 6.40092552e-01 1.01346418e-01 3.46743524e-01 1.65402210e+00 3.95525426e-01 6.94052339e-01 1.67875123e+00 -2.09761336e-01 1.94716975e-01 -8.06712881e-02 -5.02553225e-01 5.92789948e-01 -1.27122356e-02 1.47818774e-01 -7.23579466e-01 -1.89442158e-01 1.12002218e+00 -3.00855428e-01 -3.29822451e-01 -2.26252645e-01 -1.11947906e+00 8.74742627e-01 4.78549093e-01 8.84214193e-02 -1.67929202e-01 3.16097826e-01 -6.38715625e-02 3.53660971e-01 5.80560148e-01 2.81762302e-01 9.67431515e-02 -3.59672159e-02 -7.62180567e-01 4.86153245e-01 5.54005504e-01 9.51588273e-01 6.10567927e-01 3.73349488e-01 -7.28206858e-02 6.14319086e-01 2.62322035e-02 9.16225910e-01 4.17513251e-01 -1.17761803e+00 3.17354798e-01 -7.76881874e-02 5.90711758e-02 -7.23213732e-01 -6.60173316e-03 -2.57546574e-01 -1.29080713e+00 3.80764931e-01 5.48848987e-01 -3.52166146e-01 -9.92064476e-01 2.14493370e+00 1.50054425e-01 3.29539150e-01 2.34342262e-01 1.02412617e+00 8.38689983e-01 7.70228148e-01 -2.31638879e-01 2.05064360e-02 7.84130096e-01 -5.91394663e-01 -5.28620005e-01 -1.19377494e-01 3.52796704e-01 -8.60643148e-01 1.03535628e+00 2.20383093e-01 -1.39977992e+00 -2.96731114e-01 -8.26420784e-01 -3.30897748e-01 -2.08242163e-01 -1.43130079e-01 3.84009629e-01 4.06211257e-01 -1.41219985e+00 6.68262362e-01 -1.00043678e+00 -1.65101930e-01 5.87104142e-01 3.50620836e-01 -3.62755328e-01 -5.86183853e-02 -1.28651643e+00 7.16035068e-01 -3.15405190e-01 -6.22478165e-02 -1.10881245e+00 -6.30067050e-01 -6.73956573e-01 -3.31764013e-01 4.72680777e-02 -1.27084458e+00 1.04440176e+00 -8.43515992e-01 -1.83911455e+00 1.09573925e+00 -2.65072644e-01 -5.54037154e-01 8.27885032e-01 2.44621515e-01 -8.26993287e-02 8.10266435e-02 2.02043623e-01 9.08102334e-01 1.00576472e+00 -1.32273495e+00 1.19343713e-01 -4.71708208e-01 -4.43775594e-01 6.72358349e-02 1.06292874e-01 1.90219015e-01 -5.01107611e-02 -9.40388441e-01 1.55715540e-01 -1.20034659e+00 -2.05743343e-01 1.54054046e-01 -4.28395629e-01 1.89759344e-01 9.02556479e-01 -5.58671534e-01 6.66462839e-01 -1.95339715e+00 5.18331051e-01 -1.61255449e-01 2.99038678e-01 3.41043770e-01 -2.00961664e-01 3.59050363e-01 -1.04364187e-01 2.60551155e-01 -4.81159478e-01 -6.88145816e-01 -6.48382679e-02 2.36701384e-01 -5.05193174e-01 5.38507819e-01 1.44218221e-01 1.34590256e+00 -7.95593739e-01 -3.88185978e-01 -1.66346461e-01 1.01579022e+00 -4.96871084e-01 2.43783966e-01 -1.57587320e-01 1.20518494e+00 -3.03338408e-01 3.12636614e-01 7.81764984e-01 -3.60906392e-01 -1.49192959e-01 1.36755124e-01 9.06832442e-02 -1.46756545e-01 -6.72355711e-01 1.73676944e+00 -2.56204814e-01 7.33214855e-01 2.54822105e-01 -8.16063821e-01 9.05828416e-01 2.44327039e-01 2.46523440e-01 -6.45582378e-01 2.57104456e-01 4.56520051e-01 -1.13050982e-01 -2.59552032e-01 1.85644060e-01 -6.37937725e-01 -8.63429978e-02 5.01502872e-01 1.55558601e-01 -7.42537320e-01 6.96465001e-02 3.60966653e-01 1.00014794e+00 4.61294651e-01 -2.30967849e-01 -9.39128548e-02 2.99712121e-01 -1.58938557e-01 6.12431288e-01 7.88284063e-01 -1.80686727e-01 1.33186865e+00 3.90874654e-01 -2.63399959e-01 -1.38264537e+00 -1.68796730e+00 1.14833593e-01 3.56656939e-01 1.67758346e-01 -3.49477008e-02 -9.57545638e-01 -2.58051604e-01 -2.14116007e-01 5.01854956e-01 -5.13073504e-01 -2.59409010e-01 -4.91868973e-01 -7.41721272e-01 9.97220814e-01 4.86348808e-01 7.07171619e-01 -9.11342621e-01 2.25932281e-02 -2.25663353e-02 -3.09963435e-01 -9.75646794e-01 -6.59145474e-01 1.11703619e-01 -6.88936055e-01 -6.32349849e-01 -1.25322306e+00 -7.39580750e-01 8.67272973e-01 1.19810127e-01 9.66825664e-01 -3.06205302e-01 -4.50482041e-01 1.80119440e-01 7.29023591e-02 -1.98068783e-01 -7.92105973e-01 -5.95632009e-02 2.78305143e-01 -9.04586986e-02 1.19056515e-01 -9.02954638e-01 -7.42190719e-01 4.39643413e-01 -1.13407183e+00 1.02801077e-01 3.32548052e-01 9.66672003e-01 9.24658597e-01 -3.31158817e-01 6.37188911e-01 -7.59551704e-01 8.18776250e-01 -3.48831385e-01 -5.95656097e-01 -8.67407471e-02 -3.39231312e-01 1.59420282e-01 8.82785499e-01 -3.71279001e-01 -1.11162221e+00 -1.01134613e-01 -3.51149142e-01 -6.90498233e-01 -2.12572634e-01 1.75059080e-01 7.06499889e-02 -2.80142248e-01 8.81861091e-01 5.73490262e-01 4.98772681e-01 -3.31668317e-01 3.31852764e-01 5.20666540e-01 6.85344696e-01 -7.73796916e-01 9.45826352e-01 7.95087993e-01 4.80911992e-02 -6.99858367e-01 -8.07197452e-01 1.17724843e-01 -4.73076791e-01 9.22278464e-02 1.15267062e+00 -8.66975725e-01 -6.07865870e-01 7.84622371e-01 -1.01032794e+00 -7.30126441e-01 -3.06983739e-01 2.73343742e-01 -1.07297587e+00 2.67535001e-01 -9.94117439e-01 -5.97250998e-01 -3.23397726e-01 -1.33853805e+00 9.54589188e-01 3.65132064e-01 -1.73986107e-01 -9.44474936e-01 8.04169029e-02 3.47567499e-01 8.42182517e-01 6.00636959e-01 6.42408669e-01 -5.64191677e-02 -7.52843559e-01 -3.81991565e-02 -3.25657353e-02 5.16717732e-01 -8.56858492e-02 2.72479728e-02 -5.95503390e-01 -3.91714245e-01 2.16092020e-01 -5.04133224e-01 8.35977912e-01 5.18487751e-01 9.88305271e-01 -6.90735579e-02 -2.37631232e-01 1.00858772e+00 1.29724574e+00 -6.87637459e-03 8.66071284e-01 1.15545243e-02 5.31665742e-01 -5.09902369e-03 -3.47751826e-02 1.45837560e-01 2.02335373e-01 5.89540601e-01 2.22307369e-01 -1.79000333e-01 -4.58931535e-01 -4.20450181e-01 5.04053950e-01 7.31265008e-01 -2.26685435e-01 -3.50452036e-01 -8.17192674e-01 2.39868596e-01 -1.55545413e+00 -1.20334375e+00 -1.96877569e-01 1.93710947e+00 8.34360182e-01 -6.51140092e-03 -1.29955024e-01 -4.01338995e-01 7.79004455e-01 2.33396456e-01 -7.78049946e-01 -2.51023650e-01 -6.29628539e-01 4.59126443e-01 3.10906053e-01 5.16049385e-01 -7.45574474e-01 9.42405343e-01 6.19927740e+00 8.26736033e-01 -1.21218300e+00 3.88427585e-01 8.06624949e-01 -2.36638300e-02 -5.16242504e-01 -7.77992280e-03 -6.45614445e-01 5.85654378e-01 6.45647168e-01 -3.39976460e-01 6.90337956e-01 5.12057900e-01 1.46043703e-01 -8.40865169e-03 -7.26791441e-01 1.06062138e+00 2.73105770e-01 -1.40129364e+00 7.04254806e-02 2.05496568e-02 1.22852349e+00 3.10597807e-01 7.99934983e-01 2.64487937e-02 5.48228383e-01 -1.19255078e+00 5.86331844e-01 6.38458967e-01 1.02529609e+00 -5.98593652e-01 4.31356311e-01 5.34484565e-01 -6.18644059e-01 4.34452802e-01 -5.13988197e-01 1.77500263e-01 3.56011927e-01 3.60711694e-01 -2.83793330e-01 2.21252874e-01 5.10277569e-01 4.29747969e-01 -2.79097974e-01 5.46251893e-01 -4.99820322e-01 5.60476303e-01 -3.27951074e-01 2.20326528e-01 9.02668685e-02 -7.33613670e-01 7.28757858e-01 6.24824941e-01 7.44218111e-01 1.07190438e-01 -1.73543781e-01 1.51541984e+00 -3.55992824e-01 -2.45806113e-01 -9.00613427e-01 -6.42327368e-02 4.07357104e-02 9.39749658e-01 -6.11317158e-01 -1.04256526e-01 -2.27454856e-01 1.51487505e+00 3.52396131e-01 6.45906806e-01 -1.11937177e+00 -2.59799033e-01 8.48279834e-01 2.16935441e-01 1.91671565e-01 -5.15293241e-01 -1.71863675e-01 -1.46439767e+00 -8.35983679e-02 -7.62760341e-01 1.53271435e-03 -1.26781321e+00 -1.41557956e+00 6.71129823e-01 -3.51745456e-01 -1.08318281e+00 -3.84985395e-02 -4.54715908e-01 -5.68634033e-01 1.10657609e+00 -1.10719776e+00 -1.15595031e+00 -1.56169638e-01 7.18648195e-01 1.70744374e-01 -1.82204749e-02 8.64033580e-01 1.86761394e-01 -5.35623372e-01 5.24728000e-01 5.86532772e-01 1.32040262e-01 9.15775299e-01 -1.16360712e+00 6.35862589e-01 8.46737504e-01 8.88203606e-02 7.39245176e-01 8.75825107e-01 -6.05309904e-01 -1.40399933e+00 -8.21739852e-01 5.62296748e-01 -6.91783726e-01 5.44989228e-01 -5.51145494e-01 -7.64870703e-01 7.58460701e-01 3.59412432e-01 4.43152189e-01 6.40094101e-01 -7.98342586e-01 -4.17927682e-01 1.78027287e-01 -1.15774858e+00 8.33629191e-01 1.13558602e+00 -7.80528307e-01 -2.34552771e-01 2.56517917e-01 5.45153737e-01 -7.88515687e-01 -7.91038811e-01 1.11589983e-01 3.15883428e-01 -1.02714133e+00 1.12279880e+00 -5.16055703e-01 8.96643043e-01 -3.73317301e-01 -1.06130671e-02 -1.58121574e+00 -5.86978644e-02 -1.15350842e+00 7.96930641e-02 1.07144761e+00 5.29385209e-01 -8.52544904e-01 8.36476564e-01 5.47760785e-01 -7.35034421e-02 -6.69215798e-01 -1.11575890e+00 -9.24380302e-01 7.15233147e-01 -6.05820492e-03 4.30950314e-01 7.54337072e-01 -4.11408782e-01 2.14823514e-01 -6.15438819e-01 -5.78753985e-02 7.97869861e-01 7.99122229e-02 8.05895209e-01 -7.51083374e-01 -4.14372802e-01 -3.77623141e-01 -2.70237237e-01 -1.22848058e+00 3.78697366e-01 -1.16627181e+00 1.00859888e-02 -1.47128248e+00 1.15777962e-01 -2.77407408e-01 1.17700636e-01 1.44015953e-01 -1.49829641e-01 8.64630640e-01 1.09998003e-01 3.65276635e-01 -2.61587709e-01 8.45533192e-01 1.62303758e+00 -1.41736880e-01 5.79747260e-02 1.99163165e-02 -6.66350424e-01 5.08470833e-01 1.08565378e+00 -6.60539746e-01 -3.44971210e-01 -6.31494403e-01 3.24059665e-01 2.52553940e-01 5.30826688e-01 -9.01869237e-01 2.48107657e-01 -2.69238837e-02 4.54825640e-01 -2.30527163e-01 3.93918753e-01 -2.29441091e-01 5.13510406e-01 3.46618801e-01 -2.13294461e-01 1.52497172e-01 -6.14962615e-02 5.52414715e-01 -3.03292930e-01 5.32749631e-02 1.19097555e+00 -3.10739577e-01 -2.09102288e-01 5.69994867e-01 -4.62510675e-01 3.87618214e-01 9.37642217e-01 6.47942126e-02 -4.61087286e-01 -7.19596207e-01 -1.04129934e+00 -1.60256058e-01 8.45632970e-01 1.25633955e-01 6.20572805e-01 -1.43482208e+00 -8.24147403e-01 3.26322287e-01 -1.99955285e-01 1.20097496e-01 3.38678300e-01 1.01614153e+00 -7.68632829e-01 2.97839735e-02 -1.53981969e-01 -5.22319615e-01 -9.90152597e-01 4.98331875e-01 4.77515727e-01 -5.97709976e-02 -6.33535385e-01 1.13207579e+00 1.17406167e-01 -5.45262277e-01 -2.95041740e-01 2.01661408e-01 5.66102922e-01 -4.35501724e-01 3.04785371e-01 1.88687816e-01 -2.96327800e-01 -9.18432415e-01 -6.30747601e-02 6.57861471e-01 8.17852020e-02 -3.21776748e-01 1.41558552e+00 -3.06822091e-01 -1.65478453e-01 7.40778074e-02 1.31218696e+00 -1.06895544e-01 -1.47662199e+00 -1.13664031e-01 -5.70677221e-01 -3.62116396e-01 -2.25229457e-01 -4.91810948e-01 -1.03857017e+00 1.13808465e+00 3.27394098e-01 -6.98563010e-02 9.06583846e-01 2.58595832e-02 1.08650064e+00 3.11921025e-03 4.55029458e-01 -7.25641370e-01 2.12514266e-01 4.63815898e-01 9.08721864e-01 -1.17288935e+00 -2.95396328e-01 -3.26359361e-01 -6.39807403e-01 8.25516880e-01 4.93767172e-01 -4.20852095e-01 5.68552375e-01 6.40435755e-01 2.14690223e-01 -1.21692950e-02 -5.76914966e-01 -1.03398107e-01 -1.83980614e-01 8.95468354e-01 3.41469496e-01 -1.41643599e-01 -2.62927711e-01 2.66205102e-01 -1.15551934e-01 2.45815411e-01 6.27909064e-01 7.62118340e-01 9.15933475e-02 -1.32954252e+00 -2.84739465e-01 1.36133507e-01 -5.72349012e-01 -1.02225073e-01 -4.48471785e-01 4.26372737e-01 -7.41073191e-02 5.58281839e-01 -9.48658958e-02 -1.37815058e-01 -7.40197813e-03 9.97899845e-02 8.16415668e-01 -3.91665608e-01 -2.35832170e-01 -8.40544775e-02 -4.43494856e-01 -4.32338446e-01 -3.15333575e-01 -5.43678582e-01 -1.16150784e+00 -5.64690173e-01 -1.29950464e-01 1.36153951e-01 5.03732562e-01 6.18782938e-01 6.69295132e-01 2.39671797e-01 4.91055101e-01 -8.65684211e-01 -7.43249476e-01 -6.49297833e-01 -5.94351172e-01 5.72632968e-01 1.41331255e-01 -3.88534814e-01 -5.24934590e-01 1.17294595e-01]	[11.603638648986816, -0.38879454135894775]
9d7cd246-d932-4951-82ad-f0e511f4a299	a-bayesian-topic-model-for-human-evaluated	null	null	https://aclanthology.org/2022.lrec-1.674	https://aclanthology.org/2022.lrec-1.674.pdf	A Bayesian Topic Model for Human-Evaluated Interpretability	One desiderata of topic modeling is to produce interpretable topics. Given a cluster of document-tokens comprising a topic, we can order the topic by counting each word. It is natural to think that each topic could easily be labeled by looking at the words with the highest word count. However, this is not always the case. A human evaluator can often have difficulty identifying a single label that accurately describes the topic as many top words seem unrelated. This paper aims to improve interpretability in topic modeling by providing a novel, outperforming interpretable topic model Our approach combines two previously established subdomains in topic modeling: nonparametric and weakly-supervised topic models. Given a nonparametric topic model, we can include weakly-supervised input using novel modifications to the nonparametric generative model. These modifications lay the groundwork for a compelling setting—one in which most corpora, without any previous supervised or weakly-supervised input, can discover interpretable topics. This setting also presents various challenging sub-problems of which we provide resolutions. Combining nonparametric topic models with weakly-supervised topic models leads to an exciting discovery—a complete, self-contained and outperforming topic model for interpretability.	['Wei Wang', 'Corey Arnold', 'Justin Wood']	null	null	null	null	lrec-2022-6	['topic-models']	['natural-language-processing']	[ 6.94615915e-02 8.20195973e-01 -3.38063776e-01 -6.21339858e-01 -8.53274822e-01 -8.32579255e-01 8.61444533e-01 4.18267578e-01 1.12226129e-01 6.66657865e-01 4.50633079e-01 -4.22145605e-01 -1.67062715e-01 -7.47108638e-01 -6.12461567e-01 -5.71623445e-01 -1.57220438e-02 1.27576542e+00 6.52109534e-02 -2.42766812e-02 2.27844685e-01 -1.43194184e-01 -1.47339392e+00 3.67226154e-01 9.56653535e-01 4.86192286e-01 3.26389998e-01 5.70355296e-01 -6.12098932e-01 3.48289251e-01 -7.26638556e-01 -3.75124335e-01 -1.07397556e-01 -2.89993435e-01 -1.08026445e+00 5.47196984e-01 3.55811387e-01 -7.98203573e-02 3.28749508e-01 7.48778641e-01 -2.71479905e-01 1.97033510e-01 1.19766867e+00 -1.55519521e+00 -5.60082734e-01 1.08328164e+00 -4.88514572e-01 -6.68339431e-02 2.13552311e-01 -4.99070823e-01 1.48364747e+00 -9.05334234e-01 6.31122530e-01 1.55769372e+00 5.19280672e-01 3.49490136e-01 -1.45176244e+00 -2.99875796e-01 6.76612854e-01 -2.37871751e-01 -9.60955203e-01 3.74945775e-02 7.48459637e-01 -5.64640224e-01 6.03971124e-01 3.58582675e-01 4.94248569e-01 1.24347627e+00 -8.43933821e-02 9.49768841e-01 1.16005957e+00 -4.76030439e-01 4.18706805e-01 7.04539299e-01 7.70581067e-01 2.28586748e-01 4.36122626e-01 -4.37617481e-01 -5.03927708e-01 -6.92750454e-01 5.33662081e-01 1.74274877e-01 -1.81021810e-01 -3.57747883e-01 -1.27528703e+00 1.23583901e+00 -1.48421869e-01 2.38666698e-01 -1.28723323e-01 8.09416547e-02 3.63806576e-01 2.78414488e-01 1.11227238e+00 8.22992563e-01 -4.24096227e-01 -1.11717403e-01 -1.30555749e+00 6.00976646e-01 1.25633240e+00 1.03336942e+00 8.17167759e-01 -3.22333246e-01 -3.27404565e-03 6.25396967e-01 5.37886262e-01 3.06345254e-01 4.03404951e-01 -7.99112976e-01 3.15259062e-02 3.39760035e-01 3.34338754e-01 -8.95928621e-01 -3.19924176e-01 -4.42784190e-01 -5.35154402e-01 -8.42477009e-02 4.99784619e-01 -2.95622018e-03 -8.41972888e-01 1.69469798e+00 2.17083111e-01 -3.99299227e-02 1.77169386e-02 6.73918664e-01 5.06998718e-01 9.63429987e-01 2.79649496e-01 -3.56323570e-01 1.73950589e+00 -8.72242391e-01 -8.14720273e-01 -1.04323730e-01 4.92748111e-01 -6.75391138e-01 1.27790844e+00 6.81777000e-01 -9.09944057e-01 -3.17076534e-01 -8.60759676e-01 -1.94453284e-01 -5.32700777e-01 -2.81527415e-02 1.07165205e+00 6.97944462e-01 -1.03943753e+00 2.81430990e-01 -1.01150429e+00 -7.29386330e-01 6.53029699e-03 2.50372261e-01 -4.16739546e-02 1.26293078e-01 -1.12491417e+00 7.40503609e-01 5.58326781e-01 -4.29025501e-01 -8.14222872e-01 -7.25210786e-01 -7.59408474e-01 1.44557238e-01 5.85794747e-01 -7.17082083e-01 1.51394808e+00 -8.26973975e-01 -1.07822335e+00 7.80699134e-01 -7.34846294e-01 -6.13164783e-01 2.43805885e-01 -3.08778644e-01 -4.41356264e-02 4.98090610e-02 5.42805612e-01 7.36299694e-01 7.57365644e-01 -1.60849202e+00 -6.01690650e-01 -2.40050882e-01 1.24078818e-01 9.73202363e-02 -5.86513102e-01 4.29804027e-02 -2.39147440e-01 -9.04630899e-01 2.44062915e-01 -7.96300352e-01 -1.16324387e-01 -3.24290246e-01 -7.22669303e-01 -9.06337917e-01 1.12689984e+00 -3.27565819e-01 1.03797519e+00 -1.87625408e+00 4.09292150e-03 1.70553610e-01 6.07589900e-01 -4.51416999e-01 3.55412126e-01 5.13502419e-01 -2.09074214e-01 6.11015856e-01 -3.08301300e-01 -8.09456646e-01 4.13672566e-01 2.99093664e-01 -1.25065756e+00 2.22077444e-01 2.21394360e-01 6.10805333e-01 -9.28846061e-01 -5.93456745e-01 2.71053035e-02 2.63927989e-02 -6.27241433e-01 1.42706737e-01 -7.94349909e-01 -1.30570848e-02 -4.95796144e-01 3.68614107e-01 4.23290372e-01 -5.93091667e-01 6.34709969e-02 2.32582584e-01 -3.23394910e-02 5.07656634e-01 -1.28381813e+00 1.48729360e+00 -3.49550784e-01 9.42268133e-01 -2.58820772e-01 -1.15168738e+00 1.02077746e+00 5.31673849e-01 4.10977244e-01 4.55795050e-01 -3.29504818e-01 1.49045050e-01 -4.70875055e-01 -2.08323807e-01 9.52760220e-01 -7.57388353e-01 -3.19562554e-01 1.12234831e+00 2.88073450e-01 -4.17622000e-01 1.53077915e-01 5.64951658e-01 6.06396139e-01 -9.67099071e-02 3.58343244e-01 -7.91882873e-01 -1.19977899e-01 3.41960549e-01 1.27340406e-01 1.08884621e+00 2.50504583e-01 7.81546295e-01 8.82754982e-01 -4.99670446e-01 -1.13725483e+00 -1.35469759e+00 -4.36466336e-01 1.47868240e+00 -1.64040979e-02 -7.56634057e-01 -6.42720282e-01 -5.75380802e-01 -2.19070092e-01 8.98526847e-01 -9.29691195e-01 2.36149669e-01 4.29407880e-02 -7.43201375e-01 3.32285702e-01 4.24764305e-01 -1.43839449e-01 -6.37390912e-01 -3.37285161e-01 1.50244147e-01 -4.52214599e-01 -7.58923292e-01 -2.07179502e-01 6.21723115e-01 -1.02317178e+00 -6.44650340e-01 -8.39178681e-01 -6.97320998e-01 5.99386930e-01 2.07427338e-01 1.43230772e+00 -1.74276248e-01 1.52686656e-01 3.21519434e-01 -3.13930213e-01 -9.13531601e-01 -5.34139395e-01 4.16961521e-01 1.35224178e-01 -3.17482591e-01 6.28499091e-01 -4.83476937e-01 -5.46855330e-02 2.40423843e-01 -1.15463066e+00 8.12427793e-03 1.90029249e-01 8.59649420e-01 3.15063238e-01 1.87517986e-01 4.62358505e-01 -1.39325964e+00 7.95981705e-01 -8.78362358e-01 -1.77192435e-01 2.05519661e-01 -5.59047699e-01 3.44850063e-01 5.51542580e-01 -4.65695977e-01 -1.00740170e+00 -4.62707281e-01 2.78976828e-01 -1.01487666e-01 -5.12259364e-01 5.67633927e-01 -1.48252383e-01 9.18977380e-01 5.76167345e-01 4.62095328e-02 -2.47539222e-01 -4.76671249e-01 6.31655693e-01 7.32097030e-01 4.07828122e-01 -9.72660840e-01 7.82532513e-01 8.22239697e-01 -3.72641802e-01 -9.76684809e-01 -1.15962350e+00 -9.13745403e-01 -4.12979215e-01 1.47257596e-01 6.64545655e-01 -9.58270311e-01 -2.21821472e-01 -3.64105403e-02 -1.59138453e+00 3.31800478e-03 -5.26981771e-01 4.30294067e-01 -7.80349493e-01 2.62478799e-01 -1.36455014e-01 -8.94704521e-01 -2.76895184e-02 -9.42091227e-01 1.40871918e+00 -6.44069090e-02 -9.55580056e-01 -1.56195402e+00 1.52997583e-01 -9.68571194e-03 2.26985797e-01 1.62739649e-01 1.18424344e+00 -1.41373658e+00 -2.39638895e-01 -8.03715810e-02 2.06719954e-02 2.95172986e-02 9.89238173e-02 -1.52299404e-02 -1.18389225e+00 -1.02808392e-02 2.65595108e-01 -3.28832924e-01 1.07025075e+00 5.08158386e-01 1.05614817e+00 -4.19886619e-01 -5.31767309e-01 1.55082598e-01 1.00914979e+00 -1.58792272e-01 2.43273109e-01 3.51957470e-01 4.49718118e-01 8.38313222e-01 4.19034988e-01 3.81899416e-01 5.96197307e-01 6.81057990e-01 3.79361631e-03 -2.53891230e-01 4.09949273e-01 -2.95158833e-01 2.87247181e-01 5.99391460e-01 5.29487198e-03 -5.19098401e-01 -1.07952607e+00 7.99386859e-01 -1.89071190e+00 -1.02511084e+00 -2.77028888e-01 1.90925634e+00 9.66622055e-01 2.82909662e-01 3.73358756e-01 9.54133570e-02 6.77946329e-01 8.65915269e-02 -2.79856920e-01 -4.64680523e-01 -9.94314775e-02 2.91962009e-02 -1.12942539e-01 7.22854376e-01 -1.25280678e+00 8.67785215e-01 6.72206163e+00 8.42795551e-01 -7.15808630e-01 2.97267914e-01 8.66544843e-01 1.35183215e-01 -9.58850384e-01 2.90908903e-01 -1.04250622e+00 3.13282698e-01 1.00976527e+00 -5.74172318e-01 -2.61470437e-01 1.37028253e+00 3.16727936e-01 -2.17740178e-01 -1.45714831e+00 3.97812009e-01 2.53697693e-01 -1.15048790e+00 4.02228206e-01 2.62992412e-01 8.27716589e-01 -3.50063413e-01 1.72300696e-01 3.39432091e-01 6.61545992e-01 -1.20652819e+00 7.60976374e-01 2.21605599e-01 3.62901986e-01 -5.11001885e-01 6.65047526e-01 4.19285864e-01 -7.60411203e-01 2.97955066e-01 -5.75919628e-01 -1.97758466e-01 2.21969083e-01 8.96939695e-01 -1.07370555e+00 2.60263503e-01 5.35819471e-01 6.16760731e-01 -2.97345132e-01 8.66056800e-01 -9.87504572e-02 1.05884826e+00 -5.04612803e-01 -9.62385237e-02 3.99621457e-01 -4.20828350e-02 8.92223358e-01 1.42556274e+00 1.97116643e-01 -2.24887934e-02 3.52388144e-01 1.21274209e+00 1.16511673e-01 -5.18957572e-03 -7.15491831e-01 7.13651776e-02 2.56132096e-01 1.11679626e+00 -1.23844922e+00 -7.16934085e-01 -1.66952357e-01 6.07434154e-01 -9.40472167e-03 3.69618297e-01 -5.18117607e-01 -1.54755592e-01 5.55414557e-01 2.86519855e-01 5.62547240e-03 -3.37490588e-01 -7.22197175e-01 -1.23052716e+00 -1.15262836e-01 -7.56190777e-01 3.81487340e-01 -7.99394488e-01 -1.34808433e+00 5.58899701e-01 6.89486146e-01 -8.91848087e-01 -5.97180665e-01 -4.52851534e-01 -8.07462037e-01 8.49762142e-01 -1.16117942e+00 -1.11225307e+00 -1.45649418e-01 1.23204164e-01 1.08135343e+00 1.78385481e-01 9.52820003e-01 -5.51353991e-01 4.24309410e-02 1.16882093e-01 7.57805035e-02 -3.03289056e-01 7.84124851e-01 -1.97218502e+00 4.50265408e-01 6.09345019e-01 5.04871428e-01 1.12604070e+00 1.29941964e+00 -5.63324988e-01 -7.53701806e-01 -1.09233773e+00 1.21610844e+00 -9.02674735e-01 8.51574898e-01 -8.50122273e-01 -1.12772536e+00 9.46608603e-01 2.50924379e-01 -6.91408038e-01 9.51496959e-01 9.13524389e-01 -3.81121993e-01 3.68259221e-01 -5.83758354e-01 4.77476865e-01 5.28725505e-01 -5.53544283e-01 -1.24470186e+00 8.01858783e-01 1.16105139e+00 -5.00347987e-02 -5.31461418e-01 -2.36872807e-01 2.44786382e-01 -5.94088972e-01 7.13945150e-01 -9.20920551e-01 6.60121024e-01 -1.16302364e-01 3.13663520e-02 -1.43348241e+00 -2.15391815e-02 -8.25996697e-01 -1.18710794e-01 1.41375995e+00 6.63710654e-01 -5.39992213e-01 8.68619859e-01 8.55060756e-01 -9.03682858e-02 -5.50699353e-01 -5.42261839e-01 -8.21932435e-01 4.65887904e-01 -7.91644752e-01 4.37463850e-01 9.55633283e-01 5.09896040e-01 4.06200290e-01 -7.49344751e-02 1.07460521e-01 6.66224122e-01 3.32836598e-01 7.37808049e-01 -1.61975098e+00 -2.80662239e-01 -3.38380039e-01 -2.58829415e-01 -1.23804843e+00 2.67409414e-01 -6.69817448e-01 2.51670390e-01 -1.62285280e+00 5.85817993e-01 -4.27371860e-01 1.92711025e-01 5.20566106e-01 -2.89221913e-01 1.81719977e-02 -2.39476399e-03 5.06332159e-01 -8.05099487e-01 3.67761463e-01 7.77204573e-01 -8.90843011e-03 -1.94785610e-01 2.37272039e-01 -1.16128409e+00 9.60408986e-01 7.08608449e-01 -6.32252216e-01 -5.93445599e-01 -2.58995384e-01 3.45629752e-01 -2.34680504e-01 4.73838478e-01 -5.40987134e-01 2.60157824e-01 -1.64655164e-01 1.21719195e-02 -8.52773786e-01 2.54823476e-01 -5.95243037e-01 -9.47403386e-02 -3.12942155e-02 -8.34228396e-01 -1.40263051e-01 5.58795445e-02 9.31232154e-01 -4.18071836e-01 -3.14657301e-01 2.08198324e-01 -3.66404563e-01 -4.92593527e-01 -1.02099823e-02 -6.16319835e-01 2.29098216e-01 8.51476371e-01 -2.96632081e-01 -1.29363179e-01 -6.71256125e-01 -1.01418006e+00 1.27518609e-01 4.76492196e-01 4.51378912e-01 2.36897901e-01 -8.94465566e-01 -6.82344854e-01 -1.72449306e-01 2.33425722e-01 2.91646898e-01 -1.78717643e-01 4.99401629e-01 7.98087418e-02 8.28662932e-01 4.27790225e-01 -8.12505186e-01 -8.65310550e-01 4.00803894e-01 -1.00646585e-01 -4.46266949e-01 -4.28135127e-01 7.41106689e-01 8.43239248e-01 -3.38532597e-01 2.75406480e-01 -6.26516819e-01 -1.09061934e-01 2.53465503e-01 5.38730145e-01 1.16250731e-01 -1.15665637e-01 -1.95945933e-01 -9.45488513e-02 2.61506736e-01 -3.78713191e-01 -5.53234696e-01 1.40932584e+00 -2.75880933e-01 -1.43803716e-01 1.25647712e+00 1.09350407e+00 -2.17863023e-01 -9.34446037e-01 -1.73831522e-01 3.69479299e-01 -6.06864616e-02 -7.73344710e-02 -5.47436714e-01 -1.79855511e-01 9.94278193e-01 -1.38317138e-01 8.84080589e-01 7.76357234e-01 6.82160616e-01 3.33950549e-01 4.85348523e-01 8.36341307e-02 -8.22468102e-01 8.26471373e-02 4.51838940e-01 8.27804685e-01 -1.38538361e+00 6.32778034e-02 -6.52155101e-01 -6.69636011e-01 1.37044072e+00 2.25298509e-01 -3.13649438e-02 8.05901945e-01 1.73447341e-01 -1.09883904e-01 -6.28900707e-01 -9.38443065e-01 8.15271437e-02 4.46858317e-01 6.30498230e-01 6.55315340e-01 1.83193266e-01 -1.29727527e-01 9.35333073e-01 -7.46291220e-01 -4.35393035e-01 6.75080240e-01 5.57843506e-01 -7.33588278e-01 -9.96259153e-01 -5.27025104e-01 5.76373160e-01 -4.85100269e-01 -3.15604746e-01 -4.62038815e-01 8.88648748e-01 -1.83119208e-01 1.15508318e+00 4.56388861e-01 9.43489075e-02 -2.67481685e-01 3.43625784e-01 -2.10359126e-01 -1.20007217e+00 -2.55298495e-01 2.77039796e-01 3.23486142e-02 -4.96288091e-02 -5.17799616e-01 -8.23324680e-01 -1.04340851e+00 -9.82841700e-02 -4.95658278e-01 8.59586298e-01 7.55761385e-01 1.28641343e+00 1.92470718e-02 1.21265091e-01 4.86136109e-01 -5.96802294e-01 -4.01721120e-01 -1.16947520e+00 -8.19295704e-01 2.75296479e-01 3.61254334e-01 -7.18278050e-01 -6.48026526e-01 7.09795892e-01]	[10.404953002929688, 6.964382648468018]
95166af7-faa1-487c-9e06-61da20ea2d9a	filmy-cloud-removal-on-satellite-imagery-with	1710.04835	null	http://arxiv.org/abs/1710.04835v1	http://arxiv.org/pdf/1710.04835v1.pdf	Filmy Cloud Removal on Satellite Imagery with Multispectral Conditional Generative Adversarial Nets	In this paper, we propose a method for cloud removal from visible light RGB satellite images by extending the conditional Generative Adversarial Networks (cGANs) from RGB images to multispectral images. Satellite images have been widely utilized for various purposes, such as natural environment monitoring (pollution, forest or rivers), transportation improvement and prompt emergency response to disasters. However, the obscurity caused by clouds makes it unstable to monitor the situation on the ground with the visible light camera. Images captured by a longer wavelength are introduced to reduce the effects of clouds. Synthetic Aperture Radar (SAR) is such an example that improves visibility even the clouds exist. On the other hand, the spatial resolution decreases as the wavelength increases. Furthermore, the images captured by long wavelengths differs considerably from those captured by visible light in terms of their appearance. Therefore, we propose a network that can remove clouds and generate visible light images from the multispectral images taken as inputs. This is achieved by extending the input channels of cGANs to be compatible with multispectral images. The networks are trained to output images that are close to the ground truth using the images synthesized with clouds over the ground truth as inputs. In the available dataset, the proportion of images of the forest or the sea is very high, which will introduce bias in the training dataset if uniformly sampled from the original dataset. Thus, we utilize the t-Distributed Stochastic Neighbor Embedding (t-SNE) to improve the problem of bias in the training dataset. Finally, we confirm the feasibility of the proposed network on the dataset of four bands images, which include three visible light bands and one near-infrared (NIR) band.	['Nobuo Kawaguchi', 'Weimin WANG', 'Hiroshi Fukui', 'Masashi Matsuoka', 'Ken Sakurada', 'Kenji Enomoto', 'Ryosuke Nakamura']	2017-10-13	null	null	null	null	['cloud-removal']	['computer-vision']	[ 5.62068582e-01 -3.99399906e-01 3.06798518e-01 -1.68148935e-01 -4.01115805e-01 -7.60918081e-01 3.98202866e-01 -7.30220020e-01 -4.91522253e-01 1.03696644e+00 -2.09309697e-01 -2.66125947e-01 9.76548418e-02 -1.36364365e+00 -7.49152243e-01 -1.39801097e+00 3.20227236e-01 -2.90945888e-01 -8.34881067e-02 -2.83995539e-01 -1.87021434e-01 8.29058528e-01 -1.72710276e+00 2.00115696e-01 1.10766232e+00 8.06873858e-01 4.07890677e-01 3.87863487e-01 -2.57480741e-01 3.60641837e-01 -5.75335979e-01 -1.96404364e-02 7.36784339e-01 -4.55550283e-01 -2.29126349e-01 3.01842630e-01 2.37308472e-01 -5.21086931e-01 -1.61337957e-01 1.56098425e+00 4.00070786e-01 9.66868773e-02 4.91487622e-01 -1.25099730e+00 -8.77392173e-01 2.18135819e-01 -7.59078920e-01 -5.63597307e-02 -2.56972402e-01 5.79517543e-01 3.72305930e-01 -7.36049771e-01 4.08872753e-01 1.12449050e+00 3.66349846e-01 5.69972575e-01 -9.50047791e-01 -8.99963379e-01 6.60344213e-02 -1.26966327e-01 -1.50468922e+00 1.61641221e-02 8.14540684e-01 -2.75418252e-01 1.66662857e-01 6.10153556e-01 8.77979100e-01 8.91318083e-01 2.48650953e-01 2.02312797e-01 1.78350365e+00 -3.68081391e-01 1.16740800e-01 2.89962709e-01 -2.49346063e-01 2.78778076e-01 5.67294776e-01 5.38565874e-01 1.25096709e-01 3.49420384e-02 6.64956689e-01 5.43604016e-01 -7.05533564e-01 1.07905798e-01 -7.84300745e-01 8.30950856e-01 9.78290200e-01 2.59463817e-01 -4.55537260e-01 -1.62719384e-01 -2.64535785e-01 1.07136115e-01 4.33211207e-01 8.22618604e-02 -1.14652239e-01 7.29948163e-01 -8.29519868e-01 -6.50084242e-02 1.35050476e-01 5.73912442e-01 1.15005207e+00 5.17350078e-01 2.37961233e-01 5.39177656e-01 4.11124617e-01 1.14831209e+00 3.97926629e-01 -5.89465201e-01 3.26140702e-01 4.83226120e-01 2.82903522e-01 -9.83820081e-01 -8.42687413e-02 -6.99358821e-01 -1.22743881e+00 6.86956823e-01 9.78213772e-02 -2.86104232e-01 -1.14584756e+00 1.50446308e+00 2.51425743e-01 4.10892516e-02 5.14312446e-01 1.40035939e+00 8.69864106e-01 1.06529570e+00 -4.15963158e-02 -4.42329556e-01 1.04103470e+00 -5.91143489e-01 -7.37822771e-01 -4.30625051e-01 6.91199675e-02 -6.74967706e-01 1.11452043e+00 2.43066609e-01 -3.08970809e-01 -5.97424805e-01 -9.45985198e-01 4.05166119e-01 -5.46451747e-01 1.76155806e-01 6.56139851e-01 7.07950890e-01 -7.43743479e-01 1.80309430e-01 -5.43751776e-01 -5.65916151e-02 2.28903368e-01 -6.18561581e-02 -3.26803654e-01 -2.06424281e-01 -1.43042171e+00 6.13167226e-01 3.57001930e-01 7.09434628e-01 -7.11927831e-01 -2.38871500e-01 -5.03451169e-01 -1.22145705e-01 1.02178983e-01 -2.20557079e-01 2.26145998e-01 -1.60428321e+00 -1.21631086e+00 6.83945954e-01 1.49541497e-01 -6.62592724e-02 4.65294510e-01 6.98278844e-02 -6.12290800e-01 4.44113202e-02 3.64954807e-02 3.23852390e-01 7.52573788e-01 -1.77370989e+00 -3.81573439e-01 -5.91271579e-01 8.55611488e-02 1.51651636e-01 4.53619659e-03 -2.09755123e-01 7.41563216e-02 -4.40040201e-01 3.39268565e-01 -1.33140683e+00 -2.68740028e-01 4.61307876e-02 -5.18343449e-01 5.62016010e-01 9.33960795e-01 -6.24301791e-01 4.17899966e-01 -2.34200525e+00 -2.67271549e-01 2.70847172e-01 -2.70053059e-01 2.03317672e-01 -1.20449483e-01 2.39546329e-01 -2.56189317e-01 2.94636309e-01 -5.90835035e-01 2.02388152e-01 -5.36953926e-01 4.03169483e-01 -4.45424616e-01 5.55994034e-01 8.99065360e-02 5.52247405e-01 -5.90452969e-01 -2.97022462e-01 9.70595330e-02 5.39121628e-01 5.86688556e-02 2.86137134e-01 -1.07574277e-01 8.03797305e-01 -3.01053584e-01 5.17836511e-01 1.35855687e+00 1.96880654e-01 -7.45489523e-02 -2.52430022e-01 -3.10648739e-01 -5.62947750e-01 -1.23150170e+00 1.10984349e+00 -4.84464049e-01 4.43690270e-01 1.22823246e-01 -5.52663743e-01 1.16850829e+00 2.58438975e-01 1.70802161e-01 -5.54709077e-01 -9.92837101e-02 1.23964660e-01 -9.60917771e-02 -6.86240017e-01 3.53767365e-01 -5.67533255e-01 4.60833192e-01 2.96898961e-01 -6.22998774e-01 -4.53985542e-01 -4.15608466e-01 -2.39514679e-01 1.95007369e-01 4.50614356e-02 -1.23859465e-01 1.81135818e-01 4.77699757e-01 6.21478967e-02 7.47386098e-01 4.24514115e-01 9.37444046e-02 6.66613758e-01 -1.81342423e-01 -4.34150398e-01 -7.45938838e-01 -1.05026996e+00 -2.50427842e-01 4.82751220e-01 3.58945638e-01 6.37624145e-01 -4.76957023e-01 -4.46000069e-01 -7.27404207e-02 6.49864614e-01 -4.26180840e-01 -1.17736414e-01 -1.15834862e-01 -1.27518654e+00 4.41353977e-01 2.13550091e-01 1.04746675e+00 -1.12142253e+00 -4.46983308e-01 -1.09978981e-01 -3.84111494e-01 -1.03653574e+00 1.85121357e-01 -4.31441814e-02 -8.99950683e-01 -9.69039023e-01 -7.39026189e-01 -3.99093688e-01 7.66253173e-01 6.44639373e-01 6.04222298e-01 4.18444775e-04 -3.04301292e-01 -2.69728601e-02 -3.47156435e-01 -4.92899537e-01 -3.74441057e-01 -4.75198060e-01 -2.90134698e-02 4.37153518e-01 2.25559339e-01 -4.14721519e-01 -5.26417196e-01 5.05722500e-02 -1.42894769e+00 6.33281916e-02 6.06408358e-01 6.99270010e-01 6.45496666e-01 6.67103291e-01 1.83193773e-01 -7.76600242e-01 2.16526091e-01 -3.42763990e-01 -8.14611495e-01 1.40895426e-01 -4.22594607e-01 -3.02836359e-01 7.62244940e-01 -2.94341624e-01 -1.28734601e+00 2.31261611e-01 8.81439373e-02 -4.45206285e-01 -4.20588523e-01 4.35975581e-01 -3.64994794e-01 -3.44799578e-01 7.18125641e-01 6.95083737e-01 -8.29925910e-02 -2.42259040e-01 1.65121570e-01 9.00619805e-01 3.97675753e-01 2.44490169e-02 1.30991268e+00 8.49835694e-01 1.38974264e-01 -1.10805357e+00 -6.14131510e-01 -6.37707636e-02 -3.01193804e-01 -3.70051831e-01 9.86733258e-01 -1.19236672e+00 -2.79253989e-01 5.96571267e-01 -9.82411742e-01 2.82363221e-02 1.66632310e-02 7.25760400e-01 -8.84544179e-02 3.82786602e-01 -4.06349838e-01 -1.22740543e+00 -4.30896640e-01 -1.11735296e+00 6.83834493e-01 5.51730037e-01 7.06394136e-01 -6.49688005e-01 -2.91491151e-01 3.55753034e-01 3.26866001e-01 7.00851321e-01 7.97680140e-01 2.37418026e-01 -1.00628936e+00 1.22957502e-03 -2.55705863e-01 7.82980740e-01 5.21537304e-01 4.25197273e-01 -1.19770944e+00 -3.76363605e-01 2.17739642e-01 -1.30300492e-01 9.20652926e-01 4.13248211e-01 1.08870232e+00 -3.55787337e-01 8.95971246e-03 8.96055579e-01 2.09903669e+00 3.29544216e-01 8.74655187e-01 5.80754519e-01 6.85921609e-01 7.07818329e-01 7.10993767e-01 1.58145145e-01 -2.46210426e-01 8.48115161e-02 1.00392282e+00 -6.68774724e-01 1.75568253e-01 1.28255814e-01 2.99031466e-01 2.61770517e-01 -4.66770321e-01 -2.87203103e-01 -6.29969358e-01 5.15411794e-01 -1.33075035e+00 -1.20958805e+00 -6.69328570e-01 2.34583735e+00 6.72524452e-01 -1.57604277e-01 -4.59769070e-01 1.68085381e-01 9.58559632e-01 2.99241722e-01 -3.74817610e-01 2.95423549e-02 -6.18976533e-01 1.33297071e-01 9.52619016e-01 4.19779301e-01 -8.93470407e-01 6.63177669e-01 5.06322098e+00 4.25713569e-01 -1.71556342e+00 4.06774953e-02 4.56751645e-01 1.72140732e-01 -6.54606581e-01 8.55585933e-02 -4.91226763e-01 7.31812358e-01 2.78659731e-01 3.62588823e-01 7.36437380e-01 5.18487990e-01 6.32367551e-01 -2.16038227e-01 -2.04866618e-01 9.49930549e-01 -1.43827066e-01 -8.01071465e-01 2.60165453e-01 8.96771923e-02 9.37794745e-01 1.23770051e-01 3.11427474e-01 -2.31632844e-01 1.84812710e-01 -8.90594363e-01 3.92049819e-01 6.31917655e-01 1.01201963e+00 -8.13850343e-01 8.85914981e-01 6.75742805e-01 -8.34542513e-01 -1.30518496e-01 -7.04329550e-01 7.23111108e-02 -1.76505208e-01 8.77178252e-01 -4.85208839e-01 7.59530127e-01 7.14440405e-01 1.73538491e-01 -4.72759753e-01 7.42928028e-01 -4.69521642e-01 5.90689778e-01 -2.48200893e-01 2.43833840e-01 3.52458954e-01 -8.22562516e-01 3.79768431e-01 6.85473800e-01 5.19180834e-01 3.68667454e-01 8.29476491e-02 1.13687909e+00 1.33876517e-01 1.54826604e-02 -9.78938878e-01 -3.68963853e-02 2.44787112e-01 1.21894765e+00 -5.08501887e-01 -6.81753457e-02 -4.18137103e-01 8.64475608e-01 -4.80131775e-01 8.33437622e-01 -8.64903569e-01 -3.18964601e-01 5.00207424e-01 -5.69940545e-02 -5.19074835e-02 -9.00233611e-02 -1.40147269e-01 -1.11459816e+00 2.97194123e-02 -8.51896822e-01 -7.22430125e-02 -1.28405225e+00 -1.19014597e+00 8.58529329e-01 -2.24175945e-01 -1.59210849e+00 1.76862150e-01 -3.95776421e-01 -5.72662890e-01 1.45501566e+00 -2.06238651e+00 -1.14967310e+00 -1.01100647e+00 7.28192627e-01 1.69535592e-01 2.40946691e-02 7.72088826e-01 7.93167725e-02 -3.27248961e-01 -9.67755448e-03 5.07747173e-01 1.59210265e-01 5.45428514e-01 -7.55098403e-01 -1.51773244e-01 1.12670720e+00 -2.56350134e-02 3.67677927e-01 6.19527578e-01 -5.73694885e-01 -1.10387552e+00 -1.40247381e+00 2.67832905e-01 2.08271176e-01 1.06183425e-01 7.84455985e-02 -8.28539908e-01 5.81762373e-01 8.10856074e-02 3.33770990e-01 5.95434785e-01 -5.75400651e-01 -9.71745625e-02 -4.97441679e-01 -1.41478932e+00 4.06730920e-01 4.73935485e-01 -5.45773804e-01 -2.40761057e-01 6.09400272e-01 5.73756635e-01 -7.55440593e-02 -3.92255425e-01 4.14683223e-01 4.03851986e-01 -1.19378459e+00 8.25511396e-01 -3.16931546e-01 5.37581086e-01 -6.89251006e-01 -3.31765771e-01 -1.53239512e+00 -3.65853876e-01 8.39887634e-02 8.62038970e-01 1.10016644e+00 2.30866790e-01 -9.71638680e-01 6.51246488e-01 2.60868460e-01 1.70650288e-01 -1.05760396e-01 -6.88569784e-01 -8.41847241e-01 -1.63916983e-02 -9.76796914e-03 8.26714754e-01 1.15641284e+00 -9.53953028e-01 -6.36708438e-02 -3.54989260e-01 9.27093387e-01 8.56904745e-01 6.40924811e-01 7.60163069e-01 -1.27047503e+00 -1.86314151e-01 1.92770526e-01 -1.36499003e-01 -4.40865636e-01 8.70713070e-02 -7.02180207e-01 -4.33008419e-03 -1.48180687e+00 1.85521901e-01 -6.12214684e-01 -1.01769902e-01 4.11378801e-01 -2.17569202e-01 5.91318846e-01 3.63183439e-01 4.47350621e-01 5.40054739e-01 6.06060684e-01 1.59800160e+00 -5.09163320e-01 -2.16570720e-01 2.33174905e-01 -3.93066317e-01 6.62062466e-01 9.25852597e-01 -5.72982252e-01 -3.40239823e-01 -5.36782146e-01 6.93633407e-02 1.20449640e-01 5.61065197e-01 -9.80757356e-01 -2.22984090e-01 -6.09843075e-01 6.12731636e-01 -4.13930148e-01 3.99296045e-01 -1.31610978e+00 7.02658474e-01 5.21602035e-01 1.14124887e-01 -2.43837252e-01 6.28834516e-02 4.62069929e-01 -2.11239442e-01 -3.30010831e-01 1.02456009e+00 -5.33347905e-01 -5.53383052e-01 3.08343887e-01 -3.68848413e-01 -4.15341526e-01 9.62291718e-01 -3.70589703e-01 -5.27835190e-01 -5.90947986e-01 -5.02482951e-01 -1.69723094e-01 6.29518270e-01 1.62130266e-01 6.27363563e-01 -1.18661880e+00 -7.39387453e-01 4.85490143e-01 1.03278585e-01 9.76743773e-02 4.14361358e-01 4.48933572e-01 -8.34968328e-01 -7.51836505e-03 -3.83351594e-01 -4.45393324e-01 -1.27560365e+00 5.31262279e-01 5.79327822e-01 2.43399769e-01 -3.35411429e-01 5.49189448e-01 4.09089297e-01 -5.00319898e-01 -3.32809031e-01 -2.23457724e-01 -3.29647958e-01 -1.97404355e-01 3.35079342e-01 9.89027545e-02 -3.10432576e-02 -7.01115429e-01 -1.97539017e-01 5.66494465e-01 5.55730879e-01 -8.26127306e-02 1.38960147e+00 -9.40450951e-02 -4.30976391e-01 4.29616243e-01 9.57323492e-01 3.70734334e-01 -1.14038467e+00 -1.15731671e-01 -8.37635815e-01 -7.04553306e-01 3.67449164e-01 -7.77755618e-01 -1.51182163e+00 9.83742177e-01 1.03494811e+00 3.29914033e-01 1.43029630e+00 -5.36391318e-01 4.94696915e-01 2.51479805e-01 4.63020355e-01 -8.23520839e-01 -4.61775810e-01 6.06576726e-02 8.07973623e-01 -1.27417791e+00 -1.55951856e-02 -5.43469906e-01 -6.37020588e-01 1.12902725e+00 5.04840612e-01 -4.32361476e-02 3.75262678e-01 1.56569287e-01 5.58957279e-01 3.59546356e-02 -1.00308307e-01 -4.39208150e-01 -2.84789443e-01 7.07501531e-01 5.53247854e-02 2.46359050e-01 -2.36829966e-01 1.37383550e-01 2.89425347e-02 -8.25634748e-02 8.63892496e-01 6.32988393e-01 -4.02264357e-01 -6.31104827e-01 -1.09852624e+00 2.54827261e-01 -3.21583688e-01 -2.37758994e-01 -4.00615275e-01 6.83534384e-01 4.72526520e-01 1.06291282e+00 -1.78698283e-02 -3.23145956e-01 -2.10640170e-02 -1.16510289e-02 1.71013042e-01 -3.70995164e-01 -2.50713259e-01 -4.16300334e-02 -4.87530708e-01 -1.14744686e-01 -7.25245655e-01 -3.24423283e-01 -1.07415533e+00 -2.49724150e-01 -6.77213430e-01 1.77509829e-01 8.26696873e-01 6.76775396e-01 -8.28721076e-02 3.51885676e-01 1.22589839e+00 -7.30903268e-01 -4.21443433e-01 -9.08970952e-01 -1.16812837e+00 3.10428172e-01 4.01732475e-01 -3.74299049e-01 -8.59194279e-01 7.25731626e-02]	[10.01423454284668, -1.9199985265731812]
1dcff945-b52f-4723-8e68-369b084ba55c	training-data-set-assessment-for-decision	2004.05380	null	https://arxiv.org/abs/2004.05380v1	https://arxiv.org/pdf/2004.05380v1.pdf	Training Data Set Assessment for Decision-Making in a Multiagent Landmine Detection Platform	Real-world problems such as landmine detection require multiple sources of information to reduce the uncertainty of decision-making. A novel approach to solve these problems includes distributed systems, as presented in this work based on hardware and software multi-agent systems. To achieve a high rate of landmine detection, we evaluate the performance of a trained system over the distribution of samples between training and validation sets. Additionally, a general explanation of the data set is provided, presenting the samples gathered by a cooperative multi-agent system developed for detecting improvised explosive devices. The results show that input samples affect the performance of the output decisions, and a decision-making system can be less sensitive to sensor noise with intelligent systems obtained from a diverse and suitably organised training set.	['Johana Florez-Lozano', 'Fabio Caraffini', 'Mario Gongora', 'Carlos Parra']	2020-04-11	null	null	null	null	['landmine']	['computer-vision']	[ 2.11227074e-01 3.46145868e-01 -8.76692757e-02 -2.87580937e-01 -5.70246816e-01 -1.53029695e-01 4.37793344e-01 3.20670128e-01 -6.94897294e-01 1.07935417e+00 -3.88581812e-01 -2.73916095e-01 -5.39831996e-01 -1.39716768e+00 -4.56897616e-01 -8.91672194e-01 -2.80443639e-01 9.65417027e-01 3.92617226e-01 -4.41154867e-01 1.70265526e-01 5.97429872e-01 -2.00878096e+00 2.87529886e-01 9.08703744e-01 1.00857460e+00 4.41862106e-01 7.28819132e-01 2.56821603e-01 8.48880172e-01 -1.12010300e+00 3.64458174e-01 3.20727438e-01 -3.54440570e-01 -4.52490002e-01 -2.37991381e-02 -4.94778812e-01 -4.75605577e-01 2.08914176e-01 1.16090035e+00 5.84538639e-01 1.60964236e-01 1.11297691e+00 -1.46581197e+00 -1.45853549e-01 7.62578011e-01 -3.45412076e-01 4.88670915e-02 1.91748664e-01 1.76534608e-01 2.74073243e-01 -2.68225580e-01 3.75891536e-01 1.19354045e+00 4.11198109e-01 4.81854469e-01 -9.38142478e-01 -3.83594096e-01 -1.06438883e-01 4.08919334e-01 -1.50330663e+00 -2.61712581e-01 6.71563923e-01 -3.38937610e-01 8.86588097e-01 5.36252201e-01 4.65386540e-01 6.50243938e-01 6.54615521e-01 6.96829379e-01 9.79030848e-01 -8.02859783e-01 1.11643934e+00 3.70524645e-01 4.72449996e-02 4.53490227e-01 8.96155655e-01 2.60294646e-01 6.12423271e-02 -5.83280206e-01 1.88884482e-01 -6.24224395e-02 3.00121665e-01 -8.63054320e-02 -4.25920367e-01 9.20902550e-01 1.53345615e-01 7.77364314e-01 -1.09102452e+00 -1.41173005e-01 2.78078258e-01 3.11267704e-01 4.90159392e-01 3.02997112e-01 -4.41877067e-01 2.77269989e-01 -5.91138005e-01 2.29693398e-01 1.03267694e+00 5.95609903e-01 9.72436249e-01 5.84043600e-02 1.10401399e-01 4.37920362e-01 6.99614465e-01 1.04171836e+00 5.52223384e-01 -7.07462311e-01 1.44366190e-01 6.94705963e-01 6.17328025e-02 -1.01856470e+00 -6.60324395e-01 1.32626534e-01 -6.55405521e-01 9.06547785e-01 1.96913350e-02 -8.00738633e-01 -8.74147713e-01 1.03626204e+00 4.34555203e-01 -2.91944116e-01 6.79967403e-01 6.45750761e-01 3.24840158e-01 5.92154980e-01 1.06385663e-01 -2.78651506e-01 9.00216520e-01 -2.49616981e-01 -7.71145880e-01 -3.99856478e-01 4.08498466e-01 1.06077522e-01 7.42602497e-02 4.61977392e-01 -7.60160089e-01 -2.32405320e-01 -1.19049430e+00 9.86194968e-01 -5.25504231e-01 -1.66893065e-01 5.67789137e-01 8.43373239e-01 -8.06916773e-01 4.55845207e-01 -8.26690614e-01 -3.38388890e-01 2.90533811e-01 3.97207856e-01 4.11756001e-02 8.52091610e-02 -1.26429701e+00 1.46466839e+00 5.09242713e-01 3.16463053e-01 -9.04689014e-01 -1.86129799e-03 -6.97443247e-01 -2.12816596e-01 1.18820265e-01 -2.62165844e-01 1.17680025e+00 -8.09917212e-01 -1.37797058e+00 5.14823139e-01 3.97172928e-01 -6.25654757e-01 7.40496039e-01 3.26056540e-01 -5.36067307e-01 2.62752831e-01 3.49690020e-01 -8.87238011e-02 7.69296110e-01 -1.51394153e+00 -1.14744627e+00 -5.38425922e-01 -5.45903802e-01 1.90069467e-01 1.31570920e-02 -3.56374010e-02 5.34074962e-01 2.13622123e-01 -2.09530845e-01 -7.45991647e-01 -9.46610987e-01 -5.00370920e-01 -3.30978185e-02 -3.27892065e-01 9.38361645e-01 -2.20913500e-01 6.92169964e-01 -1.91568887e+00 -3.57453912e-01 8.11539292e-01 -1.67706683e-01 2.80484229e-01 1.40410528e-01 5.49530029e-01 4.12314117e-01 2.31481120e-02 -3.19846570e-01 9.56714600e-02 9.12954137e-02 6.76291645e-01 1.93095446e-01 4.64321524e-01 2.70967990e-01 2.84359843e-01 -9.79622841e-01 -5.07395506e-01 5.05988061e-01 -1.43334076e-01 8.10381100e-02 1.63680404e-01 -4.45696115e-01 -2.89229164e-03 -1.12602091e+00 7.38098562e-01 6.65947497e-01 2.97251493e-01 3.21280062e-01 3.78363729e-01 -6.48390055e-02 -6.33299291e-01 -1.30236602e+00 1.05475616e+00 -4.45919752e-01 3.41572583e-01 2.96367735e-01 -1.10953093e+00 1.27004087e+00 4.16047454e-01 8.02738965e-01 -9.86921310e-01 6.17340267e-01 3.73453081e-01 6.81868345e-02 -1.11277795e+00 6.20944440e-01 -9.00935233e-02 -2.04630956e-01 7.80754268e-01 -2.87667274e-01 -3.07604820e-01 6.10495619e-02 -2.64682293e-01 1.48849559e+00 -4.34311509e-01 3.16402256e-01 -3.44699383e-01 2.26546042e-02 4.26198125e-01 4.58798230e-01 1.11216187e+00 -3.33166331e-01 -4.47755866e-03 7.19635189e-02 -5.66100955e-01 -7.23361492e-01 -7.77085066e-01 -1.73528820e-01 9.10666585e-01 3.81266087e-01 4.31715012e-01 -7.12276101e-01 -4.89270270e-01 3.35891128e-01 1.01931071e+00 -6.21125996e-01 -2.60073900e-01 -1.91811010e-01 -1.29505908e+00 5.85011125e-01 2.30026618e-01 4.63646382e-01 -1.54286122e+00 -1.40169346e+00 5.01257002e-01 9.99669358e-02 -5.28058410e-01 7.09634304e-01 6.61918819e-01 -5.72671354e-01 -1.12179649e+00 -1.21983387e-01 -4.67549443e-01 7.48280942e-01 -1.70321345e-01 8.30606818e-01 2.69758284e-01 -3.64686698e-01 5.73120236e-01 -5.38458586e-01 -1.34854841e+00 -1.07432222e+00 -3.49223390e-02 9.49532092e-02 -7.00408965e-03 5.76837420e-01 -3.50006193e-01 -1.91444620e-01 3.54903221e-01 -9.65674698e-01 -5.37162066e-01 9.14174199e-01 4.42750812e-01 1.79455981e-01 5.99672854e-01 1.01473439e+00 -7.56733596e-01 1.08939028e+00 -8.64365041e-01 -4.87667412e-01 4.35960770e-01 -7.79908299e-01 1.08391628e-01 2.74099469e-01 -2.62836874e-01 -1.08012998e+00 2.02464581e-01 7.75703266e-02 5.36618987e-03 -4.67492133e-01 6.06902599e-01 -1.77984148e-01 2.81996131e-02 1.39053726e+00 -1.00705691e-01 4.71697092e-01 5.49347401e-02 -7.60443583e-02 1.30215037e+00 2.86775529e-01 -5.44868410e-01 3.55303198e-01 4.02611494e-01 4.17048149e-02 -9.78917062e-01 -9.31224301e-02 -2.26266176e-01 -2.69310832e-01 -7.50170767e-01 4.24998373e-01 -5.95001519e-01 -5.95433652e-01 5.32846570e-01 -1.06478763e+00 -4.77146834e-01 -5.03500700e-01 4.19787824e-01 -6.01786613e-01 -1.90009996e-01 4.17853566e-03 -1.57410240e+00 -4.27145869e-01 -7.39769161e-01 8.26044500e-01 3.29723597e-01 -1.97012443e-02 -9.31926847e-01 4.58284259e-01 -9.02349725e-02 6.26215518e-01 6.55656695e-01 4.21231598e-01 -9.27556872e-01 -3.55688155e-01 -7.40042865e-01 3.91191125e-01 2.47015208e-01 4.28661883e-01 1.67566538e-01 -1.03716636e+00 -7.66711012e-02 3.20669353e-01 -2.07042247e-01 5.96706390e-01 4.88773733e-01 3.41529101e-01 -2.09126323e-01 -7.87337482e-01 -3.75451535e-01 1.31227386e+00 5.44434786e-01 7.90317893e-01 8.13752711e-01 -5.16300090e-02 7.10504770e-01 1.21725488e+00 7.08887696e-01 1.42035544e-01 9.64649543e-02 8.60342503e-01 -1.30918697e-01 5.36542118e-01 3.29038858e-01 9.55216959e-02 -9.71786156e-02 -2.50278205e-01 -3.54256630e-01 -1.03879368e+00 4.35775667e-01 -2.30800581e+00 -1.19933546e+00 1.46178544e-01 1.83965111e+00 4.47389990e-01 -9.38811526e-03 1.80393413e-01 4.17041153e-01 1.05321300e+00 -1.72881216e-01 -7.20059276e-01 -3.70245844e-01 -1.01624019e-01 -1.74616948e-01 8.62894118e-01 4.90710795e-01 -1.01801538e+00 3.66965115e-01 7.65836668e+00 7.21739233e-01 -9.26628888e-01 -1.51293904e-01 4.09074217e-01 3.46983403e-01 -2.36932486e-01 -3.96659434e-01 -3.53663862e-01 4.75957632e-01 1.22866106e+00 -5.52483439e-01 1.28105819e-01 1.01863384e+00 3.58289063e-01 -9.32930946e-01 -7.16298997e-01 3.82039040e-01 -5.50449900e-02 -1.03902054e+00 -3.35595369e-01 7.68217295e-02 5.07899940e-01 3.71056020e-01 -4.55431372e-01 6.91935942e-02 8.01316082e-01 -5.88052511e-01 7.78528869e-01 7.31768012e-01 8.28059316e-02 -1.02567244e+00 1.14501989e+00 8.57948661e-01 -6.78792596e-01 -3.46585482e-01 -3.91954154e-01 -3.73735279e-01 -3.67298014e-02 7.14976609e-01 -1.17507863e+00 7.36004889e-01 5.94654739e-01 9.42667797e-02 -3.99090588e-01 9.85650957e-01 9.21037495e-02 2.13614419e-01 -8.17979813e-01 -8.52451921e-01 -3.07863895e-02 -1.70739740e-01 4.24063981e-01 1.07741296e+00 1.21121198e-01 2.18777001e-01 3.13302159e-01 5.39325893e-01 7.13312507e-01 -2.08159506e-01 -1.02265239e+00 5.05886018e-01 5.37315011e-01 1.14745224e+00 -7.79767990e-01 -2.35218853e-01 1.58697873e-01 2.86193937e-01 -8.15485939e-02 9.56152156e-02 -5.40274739e-01 -5.48001528e-01 2.90870577e-01 -2.42191344e-01 -1.24645896e-01 2.52399981e-01 -2.73485571e-01 -4.86265332e-01 3.78150828e-02 -5.62952101e-01 3.99005800e-01 -5.39456308e-01 -1.23050523e+00 9.12285507e-01 1.47816032e-01 -1.25392020e+00 -9.60123479e-01 -4.42021281e-01 -6.00054920e-01 6.86081111e-01 -1.14043856e+00 -8.82141113e-01 -1.82606369e-01 5.05380332e-01 7.49081448e-02 -6.57247126e-01 1.09441757e+00 -6.81860326e-03 -3.76915663e-01 -5.52447811e-02 4.08801317e-01 -1.09399162e-01 2.16842696e-01 -1.01500702e+00 -1.36753887e-01 7.54781067e-01 -4.41949546e-01 -3.40366095e-01 9.84294593e-01 -9.29853678e-01 -1.23407042e+00 -1.18196630e+00 2.64536619e-01 -3.56384188e-01 5.60449898e-01 1.31835759e-01 -5.87895989e-01 3.66121799e-01 1.83974355e-01 -1.83874696e-01 4.04608220e-01 -1.43622803e-02 5.23698092e-01 -1.83536395e-01 -1.84923494e+00 2.13352531e-01 3.96470398e-01 -1.31665636e-02 -5.60555458e-01 4.40334648e-01 2.31353417e-02 -1.87978357e-01 -5.65323055e-01 2.67589808e-01 1.56566694e-01 -8.65879357e-01 3.96791786e-01 -3.82687300e-01 -1.09001920e-01 -1.67585060e-01 -1.79352835e-01 -1.51123047e+00 -2.23995879e-01 -2.12796181e-01 1.76129609e-01 9.08054650e-01 6.80270195e-01 -8.92533064e-01 9.55500007e-01 9.99652028e-01 6.97016716e-02 -2.43068814e-01 -1.16372466e+00 -7.01545060e-01 -1.39460310e-01 -4.18311566e-01 4.66914952e-01 6.58341169e-01 2.92368442e-01 9.92476195e-03 -1.03625972e-02 6.45545840e-01 9.09905910e-01 -5.08187898e-02 5.33516407e-01 -1.45378447e+00 8.93118903e-02 -4.48342226e-03 -8.31856549e-01 1.72528759e-01 1.13823600e-01 -5.63071430e-01 5.37389219e-01 -1.68504202e+00 -1.34795278e-01 -8.11453700e-01 -1.18229896e-01 7.36848593e-01 1.61993906e-01 -8.05733576e-02 -9.97018162e-03 1.29356265e-01 -6.42598033e-01 3.82644057e-01 6.09803021e-01 -4.22947317e-01 -4.66212094e-01 4.81428355e-01 -5.32897830e-01 7.57842422e-01 1.02801013e+00 -1.02186739e+00 -1.60428271e-01 -3.58085871e-01 8.70590955e-02 2.04552770e-01 1.51651517e-01 -1.20626056e+00 6.28635347e-01 -5.32545626e-01 3.08286041e-01 -4.86316502e-01 3.78448926e-02 -1.35342801e+00 3.73080134e-01 9.82316554e-01 6.97054863e-02 -1.66260570e-01 1.60939217e-01 7.22578168e-01 -1.28586814e-01 -7.41228402e-01 5.89681625e-01 -2.62617379e-01 -7.40497053e-01 -1.80671290e-01 -1.26355255e+00 -6.26752496e-01 1.68220234e+00 -2.45255068e-01 -3.61764401e-01 -5.14462031e-02 -7.44888246e-01 6.21830285e-01 1.50533646e-01 3.18673313e-01 4.43318933e-01 -9.24796283e-01 -9.50778127e-01 8.76365975e-02 6.49561919e-03 2.08256692e-01 4.20846837e-03 7.53952265e-02 -5.53170323e-01 -2.43732870e-01 -5.66525519e-01 -8.12682748e-01 -1.07479334e+00 2.73263991e-01 6.85019493e-01 3.99325266e-02 -3.58502209e-01 8.68605971e-02 -7.93788075e-01 -4.03561860e-01 4.49494570e-02 -8.85226950e-02 -2.85667419e-01 2.40818128e-01 7.01504171e-01 5.25915205e-01 4.09380615e-01 -1.68558136e-01 -4.66810673e-01 1.39199898e-01 1.70815200e-01 -4.03379440e-01 1.72276950e+00 2.10176092e-02 -1.23696797e-01 4.19198871e-01 4.86834168e-01 -4.38255608e-01 -9.84932721e-01 -4.78012264e-02 8.64387676e-02 -1.53443038e-01 8.05338845e-02 -9.57994640e-01 -8.35904300e-01 -4.48199697e-02 8.65900457e-01 9.89291310e-01 1.17120481e+00 -9.68484208e-03 -3.09838682e-01 9.67479050e-01 9.66244936e-01 -1.45890081e+00 -3.07256848e-01 6.81573600e-02 9.53224361e-01 -1.35635531e+00 6.59919605e-02 5.65251373e-02 -7.22698033e-01 9.90726650e-01 3.40043724e-01 -3.34504157e-01 8.39701533e-01 7.59054422e-01 2.75276333e-01 -3.29724133e-01 -6.35407269e-01 -2.05420926e-01 -5.95863819e-01 1.05475533e+00 -6.38758123e-01 3.82914484e-01 -1.32210374e-01 7.15916157e-01 2.99323618e-01 1.45538077e-01 7.51894772e-01 1.49882412e+00 -9.97108281e-01 -7.85054326e-01 -9.28600550e-01 6.13446951e-01 -6.10736907e-02 5.99204242e-01 -4.53709334e-01 8.49154890e-01 1.86750993e-01 1.40919507e+00 1.62881121e-01 -2.93304443e-01 6.52170420e-01 -4.55514461e-01 1.68253496e-01 -5.75479686e-01 -6.34791553e-01 -3.11666787e-01 2.80278236e-01 -9.24376100e-02 -6.26535237e-01 -9.02020216e-01 -1.52205777e+00 -2.06081852e-01 -5.80004275e-01 6.16761804e-01 1.07549334e+00 9.88249660e-01 2.72330731e-01 6.26570463e-01 1.09017766e+00 -1.23289299e+00 -7.20552027e-01 -1.02474940e+00 -1.12269068e+00 3.25764678e-02 1.76370904e-01 -4.47034746e-01 -6.34129941e-01 -3.20642292e-01]	[4.5448784828186035, 2.0691747665405273]
a12668cf-60b0-4eaa-9cf1-80ce27ca8a40	contextual-text-block-detection-towards-scene	2207.12955	null	https://arxiv.org/abs/2207.12955v1	https://arxiv.org/pdf/2207.12955v1.pdf	Contextual Text Block Detection towards Scene Text Understanding	Most existing scene text detectors focus on detecting characters or words that only capture partial text messages due to missing contextual information. For a better understanding of text in scenes, it is more desired to detect contextual text blocks (CTBs) which consist of one or multiple integral text units (e.g., characters, words, or phrases) in natural reading order and transmit certain complete text messages. This paper presents contextual text detection, a new setup that detects CTBs for better understanding of texts in scenes. We formulate the new setup by a dual detection task which first detects integral text units and then groups them into a CTB. To this end, we design a novel scene text clustering technique that treats integral text units as tokens and groups them (belonging to the same CTB) into an ordered token sequence. In addition, we create two datasets SCUT-CTW-Context and ReCTS-Context to facilitate future research, where each CTB is well annotated by an ordered sequence of integral text units. Further, we introduce three metrics that measure contextual text detection in local accuracy, continuity, and global accuracy. Extensive experiments show that our method accurately detects CTBs which effectively facilitates downstream tasks such as text classification and translation. The project is available at https://sg-vilab.github.io/publication/xue2022contextual/.	['Song Bai', 'Changhu Wang', 'Shijian Lu', 'Jiaxing Huang', 'Chuhui Xue']	2022-07-26	null	null	null	null	['text-clustering']	['natural-language-processing']	[ 5.52319229e-01 -5.04704118e-01 -1.41493440e-01 -2.77576566e-01 -6.44852102e-01 -5.97707510e-01 8.65973353e-01 4.55407202e-01 -2.25703046e-01 1.71570331e-01 4.91590917e-01 -5.06780624e-01 4.01396155e-01 -6.76872671e-01 -5.74959993e-01 -7.34864295e-01 4.72631007e-01 2.46141076e-01 5.81753731e-01 -1.20341117e-02 5.14363587e-01 5.89152165e-02 -1.21204460e+00 9.32348192e-01 7.52809167e-01 7.13238895e-01 7.95031965e-01 9.40599561e-01 -6.96778774e-01 9.07858253e-01 -7.77439952e-01 -1.05865374e-01 -2.83867922e-02 -7.29394794e-01 -7.59645104e-01 2.00622022e-01 3.36561322e-01 -3.06945086e-01 -2.25787044e-01 9.99076724e-01 3.08761865e-01 6.75875694e-02 5.57701111e-01 -9.93028581e-01 -3.00196975e-01 9.12898004e-01 -6.38522863e-01 2.77016342e-01 4.91970062e-01 -2.22134009e-01 1.11267400e+00 -1.11969662e+00 4.54529971e-01 1.15959132e+00 3.72983098e-01 2.32075438e-01 -7.76338220e-01 -4.60061878e-01 4.37444001e-01 2.70817667e-01 -1.29547226e+00 -2.15490147e-01 4.44860578e-01 -4.13713843e-01 7.69881308e-01 8.12999547e-01 4.51198280e-01 1.08068228e+00 1.72651231e-01 1.35595906e+00 7.81259120e-01 -6.85031772e-01 2.35800609e-01 7.02237114e-02 4.40618247e-01 4.06181812e-01 1.04098702e-02 -7.08864331e-01 -6.47078753e-01 2.70000488e-01 3.83074969e-01 2.26414636e-01 -3.09846967e-01 8.17156136e-02 -1.56426358e+00 7.14606047e-01 7.01388791e-02 6.03995025e-01 -3.35893035e-02 7.15954453e-02 5.51869214e-01 2.95723453e-02 5.34585416e-01 -2.15815291e-01 -1.34315774e-01 -1.55970305e-01 -9.40806210e-01 -1.35551840e-01 5.36093891e-01 1.48004138e+00 7.87585974e-01 -3.57953131e-01 -4.32921946e-01 9.14307833e-01 1.45110443e-01 5.69205701e-01 6.11259639e-01 -1.64241090e-01 9.67277586e-01 8.64059746e-01 -6.68075755e-02 -1.12483394e+00 -3.67573678e-01 -1.41405582e-01 -7.93794155e-01 -6.27573550e-01 5.80313429e-02 4.21788618e-02 -8.34872603e-01 1.07340443e+00 4.74062711e-01 2.33694285e-01 -2.98825413e-01 8.81696403e-01 8.80071998e-01 1.16576219e+00 -2.16032520e-01 -8.42874274e-02 1.67378342e+00 -1.02961087e+00 -7.64255583e-01 -3.24930698e-01 8.72924566e-01 -1.23225009e+00 1.31329525e+00 2.83952624e-01 -6.18481219e-01 -4.86915797e-01 -8.26125443e-01 -3.12078863e-01 -5.66069782e-01 5.48916698e-01 -4.37213816e-02 4.89186674e-01 -7.64114439e-01 -5.71138635e-02 -7.44721651e-01 -6.44587338e-01 5.91182783e-02 -1.65767372e-01 1.44198596e-01 -1.09174542e-01 -9.05586243e-01 1.68338567e-01 4.36206609e-01 1.19851995e-02 -5.93819380e-01 -1.12316899e-01 -8.45286846e-01 -5.43347374e-02 6.67608798e-01 -2.32463852e-01 1.14051628e+00 -1.02601564e+00 -9.07622516e-01 8.25702369e-01 -6.88214540e-01 -1.45729885e-01 6.68361664e-01 -2.79446155e-01 -4.51363474e-01 3.82508308e-01 3.89463484e-01 3.94879103e-01 1.05796623e+00 -1.19129860e+00 -1.11310685e+00 -2.66591340e-01 -2.57678717e-01 5.03867447e-01 -5.40033937e-01 3.76876235e-01 -9.69042182e-01 -1.04063106e+00 3.08242738e-01 -6.84389174e-01 2.08966181e-01 -3.57046664e-01 -9.89224851e-01 -3.47354084e-01 1.25405478e+00 -6.78251982e-01 1.53629375e+00 -2.19655752e+00 -1.46642342e-01 3.03667895e-02 2.62591332e-01 9.56181586e-02 1.21656675e-02 7.56095767e-01 4.13508415e-02 3.31794947e-01 -1.86897919e-01 -5.44836581e-01 5.32473698e-02 1.43584132e-01 -5.79324186e-01 3.23466361e-01 -1.86151311e-01 8.99276257e-01 -8.01026046e-01 -7.70026267e-01 4.71237898e-01 -1.64604189e-05 -7.31181204e-02 -1.08768463e-01 -3.75601053e-01 2.68926322e-01 -6.44243479e-01 5.08813560e-01 7.08501101e-01 -1.56182811e-01 7.47167990e-02 -6.42795339e-02 -4.48304117e-01 4.01330501e-01 -1.24997342e+00 1.32604074e+00 -7.65660778e-02 1.16037428e+00 -9.87447426e-02 -1.14808178e+00 6.35934353e-01 1.61811113e-01 4.02526021e-01 -6.65173888e-01 2.85382569e-01 -9.30559635e-02 -5.24871588e-01 -6.12368166e-01 1.04458630e+00 2.48749286e-01 -2.71334499e-01 5.09819567e-01 -4.58166450e-01 9.51019302e-02 3.35440636e-01 5.57845592e-01 1.01065564e+00 -1.09753728e-01 3.09832722e-01 -1.47734895e-01 4.67644721e-01 1.52414933e-01 3.51165056e-01 9.78597522e-01 -1.37722701e-01 7.91384459e-01 6.05365574e-01 -5.58405593e-02 -1.16676903e+00 -7.27551937e-01 -1.49284005e-01 1.37463951e+00 4.49031383e-01 -8.69971395e-01 -7.19347477e-01 -5.62984288e-01 -3.32928240e-01 7.25505412e-01 -5.93368292e-01 1.90679878e-01 -6.88349128e-01 -6.03215337e-01 5.15535176e-01 4.83733922e-01 7.08852232e-01 -9.55792367e-01 -5.42318285e-01 -9.57995206e-02 -7.71846712e-01 -1.37765217e+00 -8.82857800e-01 9.49213728e-02 -3.99006277e-01 -7.65010774e-01 -7.13957548e-01 -1.26584184e+00 7.36582518e-01 9.78304565e-01 8.21113288e-01 2.61299729e-01 -3.96551132e-01 3.94288987e-01 -1.16991782e+00 -2.18252540e-01 -4.47238982e-01 5.42541035e-02 -3.84898126e-01 1.93845987e-01 4.00920808e-01 1.77912354e-01 -6.71714008e-01 6.98413551e-01 -1.12400544e+00 6.12825572e-01 3.86276335e-01 5.88897765e-01 5.27510762e-01 2.34542042e-01 -2.09079072e-01 -8.23783636e-01 5.41904807e-01 -4.60780293e-01 -3.01854700e-01 4.06713009e-01 -7.67086521e-02 -4.09930080e-01 6.97588146e-01 -3.39447379e-01 -1.02660716e+00 -1.90719023e-01 1.09383613e-01 -5.83123006e-02 -3.67887229e-01 6.11298263e-01 -7.80594572e-02 6.01824999e-01 4.22603667e-01 7.18083620e-01 -7.53251851e-01 -2.98488706e-01 2.51884639e-01 1.29777765e+00 3.07822764e-01 -5.28091550e-01 6.50141239e-01 5.87141454e-01 -4.45969313e-01 -1.47446954e+00 -5.65870941e-01 -1.16132820e+00 -7.84360111e-01 -3.28071445e-01 9.80019510e-01 -7.97007143e-01 -3.36749643e-01 5.89849293e-01 -9.61477935e-01 -5.86483955e-01 2.30271071e-01 3.16987038e-01 -1.19163372e-01 9.88476872e-01 -6.75981939e-01 -7.20749199e-01 -1.67472601e-01 -9.52246785e-01 1.58825576e+00 -5.49677685e-02 -1.35189071e-01 -8.69918466e-01 -3.50332975e-01 3.75230312e-01 -1.42503694e-01 -6.57062382e-02 7.25116968e-01 -5.72814345e-01 -4.82441634e-01 -2.12604150e-01 -3.53212625e-01 4.66501229e-02 1.31030157e-01 1.67227983e-01 -6.59019232e-01 -1.81022108e-01 -5.86103126e-02 6.35138676e-02 8.75537336e-01 3.36183369e-01 1.13669944e+00 -2.85477072e-01 -5.77687740e-01 2.91693449e-01 1.16208255e+00 2.97038525e-01 5.57506502e-01 4.89044070e-01 8.74284625e-01 5.57132363e-01 8.52853596e-01 5.66025496e-01 2.78693527e-01 7.29604721e-01 2.14503825e-01 -1.82487994e-01 -4.06111926e-02 -2.20942482e-01 4.91178215e-01 1.02107358e+00 4.87276793e-01 -8.87890875e-01 -1.16819334e+00 6.45608485e-01 -2.03152037e+00 -9.06598270e-01 -8.58586013e-01 1.81760216e+00 6.76430762e-01 2.16546908e-01 -3.26269045e-02 2.27174848e-01 1.23852551e+00 1.76154166e-01 -3.08676690e-01 -2.55542725e-01 -1.62551850e-01 -2.48796269e-01 4.11260933e-01 3.00844848e-01 -1.08446157e+00 1.28377664e+00 4.74957705e+00 1.30101371e+00 -9.84665394e-01 -8.41415953e-03 6.39618158e-01 3.77697647e-02 -7.66224414e-02 3.99352610e-02 -8.89264941e-01 7.13613808e-01 3.67076874e-01 1.02502137e-01 2.09115610e-01 6.09109640e-01 7.30094790e-01 -5.12195706e-01 -7.63424098e-01 9.14569795e-01 2.52185494e-01 -1.02445340e+00 2.55432069e-01 -2.20207497e-01 7.34811366e-01 -1.09080613e-01 -3.96572500e-02 1.32449150e-01 1.37013718e-01 -6.00186527e-01 1.05649209e+00 5.36491051e-02 8.32712054e-01 -3.30624163e-01 5.29909670e-01 6.19511187e-01 -1.60769093e+00 1.03342384e-01 -5.01543701e-01 9.83784050e-02 -5.12231924e-02 6.39969528e-01 -9.56097722e-01 6.14591718e-01 7.27620780e-01 9.51587796e-01 -6.81907535e-01 7.96020389e-01 -3.63953352e-01 9.87284362e-01 -1.83877543e-01 -5.29218614e-01 4.51183379e-01 -3.07264715e-01 5.69691837e-01 1.72573173e+00 3.25664401e-01 1.32055789e-01 3.88334185e-01 5.71536064e-01 3.47933285e-02 5.62017143e-01 -2.96339273e-01 1.47180468e-01 5.35375774e-01 1.32777917e+00 -1.46322608e+00 -7.49827206e-01 -5.28719008e-01 1.35420895e+00 -1.12210177e-01 3.94334108e-01 -1.01435637e+00 -4.87763733e-01 2.26637244e-01 -1.03133351e-01 4.03800458e-01 -3.00467968e-01 -5.87034404e-01 -1.40773690e+00 2.84056395e-01 -6.97747588e-01 2.97060221e-01 -1.03646302e+00 -7.95358539e-01 2.36164466e-01 -6.42821416e-02 -1.39068615e+00 2.86681920e-01 -5.64064801e-01 -6.56062067e-01 4.90124941e-01 -1.06117678e+00 -1.17328286e+00 -8.26985002e-01 7.97983110e-01 1.19983387e+00 2.31721058e-01 2.22285688e-01 1.78251609e-01 -9.16812956e-01 4.97109473e-01 5.62177718e-01 6.10253990e-01 7.93757379e-01 -1.21032012e+00 5.67348063e-01 1.18323636e+00 1.93637386e-01 5.47649920e-01 6.21932566e-01 -8.82981718e-01 -1.18412483e+00 -1.28244174e+00 8.64974380e-01 -3.31236541e-01 7.10052669e-01 -9.74343717e-01 -6.92626953e-01 8.38189721e-01 2.43559241e-01 -6.18470669e-01 3.75095308e-01 -2.26868242e-01 4.56177108e-02 1.05706595e-01 -5.75038373e-01 9.82369721e-01 1.06357086e+00 -5.34051120e-01 -5.63012064e-01 5.63946187e-01 7.78342366e-01 -4.50788587e-01 -1.70612469e-01 -1.37380324e-02 2.03830153e-01 -8.74905348e-01 5.48938513e-01 3.64369899e-02 5.09136379e-01 -3.92204076e-01 -1.99823573e-01 -9.34938967e-01 2.39390973e-02 -4.54503626e-01 3.12893450e-01 1.43900633e+00 1.21381558e-01 -4.92772102e-01 6.30973399e-01 -5.39150164e-02 -3.58962685e-01 -2.11418241e-01 -7.66578853e-01 -5.95901787e-01 -1.16001844e-01 -8.38155568e-01 4.26700830e-01 1.22339404e+00 3.58554363e-01 4.89602596e-01 -3.37871760e-01 3.74497510e-02 1.06272385e-01 2.38017693e-01 8.81963432e-01 -6.31118476e-01 1.97699592e-01 -6.58201456e-01 -1.93929866e-01 -1.81075490e+00 -3.33119899e-01 -8.37654173e-01 4.44202542e-01 -1.52583313e+00 5.00416338e-01 -3.25753212e-01 1.67406738e-01 4.79875654e-01 -3.80278617e-01 1.48635611e-01 1.68103203e-02 5.10834098e-01 -8.74764264e-01 3.51107359e-01 1.01256883e+00 -2.09206909e-01 -1.45729929e-01 -3.31171267e-02 -3.05099189e-01 5.64640105e-01 1.04708707e+00 -2.97737747e-01 -2.57209957e-01 -5.52920818e-01 3.10961068e-01 -2.19236091e-01 2.65614927e-01 -8.12643349e-01 3.98879558e-01 -1.86451748e-01 2.42984727e-01 -1.24877262e+00 -7.25164413e-02 -6.95962906e-01 -3.95696580e-01 2.54250109e-01 -4.49526787e-01 6.41137883e-02 1.10906009e-02 5.61121464e-01 -1.11601733e-01 -3.63873780e-01 3.13655287e-01 -9.65260901e-03 -8.23230863e-01 -1.03556283e-01 -9.99112785e-01 -4.92770858e-02 1.09510076e+00 -3.58889282e-01 -5.16555011e-01 -3.69869888e-01 -3.75846833e-01 3.88701439e-01 4.03310895e-01 4.77646589e-01 7.29330361e-01 -9.82278824e-01 -7.13355660e-01 1.81086659e-02 3.70098650e-01 1.80283427e-01 1.33340955e-01 8.63391519e-01 -8.81830692e-01 6.65565133e-01 4.75124389e-01 -9.85907316e-01 -1.79442346e+00 4.36961174e-01 8.40507522e-02 -6.35219365e-02 -8.70597720e-01 5.00258446e-01 5.63422203e-01 -2.08194673e-01 2.78512567e-01 -8.57834041e-01 -2.27269709e-01 1.85238793e-01 5.45073271e-01 2.44281068e-01 7.16395453e-02 -8.48930895e-01 -7.42848739e-02 6.76210523e-01 -9.43386108e-02 3.07419486e-02 7.70881891e-01 -6.03446007e-01 -2.41260260e-01 6.71728492e-01 1.13649571e+00 2.48430729e-01 -8.96332979e-01 -3.86211723e-01 1.33542806e-01 -5.24197102e-01 -5.72009198e-02 -5.94986379e-01 -5.64589143e-01 8.44617605e-01 2.69411117e-01 5.91960490e-01 1.13329589e+00 1.92398429e-01 8.75260592e-01 2.36613214e-01 -1.15427990e-02 -1.36211586e+00 4.28009421e-01 7.14340389e-01 5.92421353e-01 -1.10321414e+00 -6.40069172e-02 -6.69743419e-01 -5.83972871e-01 1.21737683e+00 4.02941883e-01 4.49989498e-01 2.37084880e-01 1.52824596e-01 3.96210477e-02 -1.40141562e-01 -4.56787884e-01 -4.34998304e-01 2.04626739e-01 1.59901679e-01 2.93032169e-01 2.80730098e-01 -3.70312989e-01 2.24411711e-01 -2.97344357e-01 -5.76866329e-01 5.12897491e-01 1.13428974e+00 -8.64629030e-01 -8.84587705e-01 -6.71695888e-01 4.54144657e-01 -3.50959092e-01 -4.93528485e-01 -7.02062011e-01 6.89721525e-01 1.09177247e-01 1.36031866e+00 2.29923189e-01 -4.95030642e-01 6.49828538e-02 4.14946303e-02 -8.98298472e-02 -7.19515502e-01 -3.56366694e-01 6.80499673e-01 -1.39815614e-01 -1.13594681e-01 -2.66404808e-01 -9.43702698e-01 -1.45379376e+00 -4.34160203e-01 -4.68020678e-01 -4.04758528e-02 6.15665495e-01 9.58620250e-01 -2.10706443e-02 5.41063547e-01 7.45436907e-01 -4.21682000e-01 2.76075244e-01 -1.13694274e+00 -3.90962541e-01 4.72542167e-01 2.72529453e-01 -8.03040415e-02 -2.52268285e-01 4.01071489e-01]	[11.961259841918945, 2.236856460571289]
ba2d0c0d-733b-429a-b57a-6080fa67f79c	autoregressive-denoising-diffusion-models-for	2101.12072	null	https://arxiv.org/abs/2101.12072v2	https://arxiv.org/pdf/2101.12072v2.pdf	Autoregressive Denoising Diffusion Models for Multivariate Probabilistic Time Series Forecasting	In this work, we propose \texttt{TimeGrad}, an autoregressive model for multivariate probabilistic time series forecasting which samples from the data distribution at each time step by estimating its gradient. To this end, we use diffusion probabilistic models, a class of latent variable models closely connected to score matching and energy-based methods. Our model learns gradients by optimizing a variational bound on the data likelihood and at inference time converts white noise into a sample of the distribution of interest through a Markov chain using Langevin sampling. We demonstrate experimentally that the proposed autoregressive denoising diffusion model is the new state-of-the-art multivariate probabilistic forecasting method on real-world data sets with thousands of correlated dimensions. We hope that this method is a useful tool for practitioners and lays the foundation for future research in this area.	['Roland Vollgraf', 'Ingmar Schuster', 'Calvin Seward', 'Kashif Rasul']	2021-01-28	null	null	null	null	['probabilistic-time-series-forecasting']	['time-series']	[-7.69395456e-02 -7.89562315e-02 -4.69496883e-02 -4.50135946e-01 -9.67550397e-01 -3.99449944e-01 8.09802592e-01 -2.21687600e-01 -1.96601868e-01 5.21287143e-01 4.38226491e-01 3.26763093e-02 -1.14755780e-01 -8.81564975e-01 -6.83200002e-01 -1.15355957e+00 -2.31253162e-01 9.78299797e-01 7.62589648e-02 2.72609860e-01 2.65737027e-01 4.13966060e-01 -1.20176327e+00 -1.37547255e-01 7.67043352e-01 9.36595559e-01 6.90868404e-03 5.59935212e-01 -1.14650406e-01 9.20464158e-01 -1.10611714e-01 -4.78310317e-01 -7.44467229e-02 -4.91782993e-01 -3.58484685e-01 -2.22980037e-01 -1.30491629e-02 -2.54275858e-01 -5.92652202e-01 1.11265254e+00 3.03515583e-01 5.95350504e-01 1.33385122e+00 -9.57116485e-01 -7.60229468e-01 5.29292464e-01 -5.18060029e-01 2.99231321e-01 -1.00249663e-01 -1.64439194e-02 7.64347911e-01 -8.58520508e-01 6.02191508e-01 1.36711669e+00 7.87412882e-01 5.20629406e-01 -1.43826151e+00 -4.39915568e-01 1.13834858e-01 1.52906284e-01 -1.02538550e+00 -3.93117189e-01 1.34683430e+00 -7.74383664e-01 6.48498893e-01 -3.23623009e-02 5.10182023e-01 1.32420838e+00 6.98605955e-01 9.30263162e-01 1.10620260e+00 -1.90478474e-01 8.21584046e-01 -6.91063553e-02 2.34504864e-01 7.09113121e-01 -2.55811751e-01 3.65394175e-01 -6.43262029e-01 -6.88406825e-01 5.65049052e-01 3.61406535e-01 -4.68902476e-02 -6.01527333e-01 -8.32705617e-01 1.06119502e+00 -1.04852490e-01 3.36891055e-01 -8.36704314e-01 7.28916228e-01 1.13935387e-02 7.24488720e-02 1.33936965e+00 -3.94779563e-01 -1.29184306e-01 -5.19359350e-01 -1.49036407e+00 2.12436512e-01 8.60797882e-01 4.71183538e-01 4.27834302e-01 3.52243632e-01 -1.76494628e-01 5.10248125e-01 8.50277424e-01 9.48047042e-01 3.23009968e-01 -1.37025023e+00 7.47302845e-02 -1.17011042e-02 3.58982623e-01 -8.62675548e-01 -7.01912791e-02 -3.61247241e-01 -9.99603987e-01 2.76709110e-01 3.98830444e-01 -2.17864603e-01 -7.18215585e-01 1.73058236e+00 3.02828044e-01 7.83249021e-01 -5.20100333e-02 7.71889985e-01 9.47372913e-02 9.65815663e-01 1.00656047e-01 -5.09728312e-01 8.64699841e-01 -6.91011131e-01 -7.86819339e-01 2.78938264e-01 4.84957881e-02 -6.61370635e-01 7.23777711e-01 5.99176049e-01 -1.26265562e+00 -2.86219418e-01 -5.77127635e-01 2.00867370e-01 8.31699893e-02 -1.04692668e-01 6.17224097e-01 5.60565352e-01 -1.11477947e+00 1.10375345e+00 -1.58095050e+00 -3.08512282e-02 1.75674289e-01 -2.51295090e-01 3.09474647e-01 -1.16381384e-02 -9.72810864e-01 6.60031140e-01 -5.81560791e-01 1.84477001e-01 -1.31951356e+00 -5.87483287e-01 -4.53820020e-01 4.43400964e-02 -2.44583830e-01 -7.64306068e-01 1.25351214e+00 -7.35093296e-01 -1.75712645e+00 4.49112684e-01 -6.33009851e-01 -6.53343439e-01 6.92116857e-01 -1.62487388e-01 -3.16242993e-01 -6.15252480e-02 -2.82363802e-01 1.64275363e-01 1.60518003e+00 -9.44157243e-01 -2.09726412e-02 -6.15783393e-01 -7.31776237e-01 -3.16275477e-01 -2.13577434e-01 -2.26691291e-01 -1.46093279e-01 -9.34426308e-01 3.25075448e-01 -9.22341168e-01 -4.28963542e-01 1.17346309e-01 7.29820803e-02 -5.08849323e-01 6.19972706e-01 -8.52700353e-01 9.50461268e-01 -1.86042905e+00 6.65227294e-01 3.45363081e-01 3.31966877e-01 -4.56726223e-01 7.02143013e-02 5.55439234e-01 3.36295605e-01 -7.70409927e-02 -5.02182305e-01 -9.13893521e-01 1.17287815e-01 5.22059128e-02 -8.46718371e-01 8.97646844e-01 -2.24888802e-01 7.59484649e-01 -1.03860569e+00 -2.18569785e-01 1.66827992e-01 9.00052130e-01 -2.10898653e-01 1.56328142e-01 -4.03442115e-01 5.41692019e-01 -7.06091881e-01 3.46585184e-01 6.49700999e-01 -1.58803180e-01 -1.29153296e-01 -1.01007111e-02 -2.67018434e-02 -1.82352886e-01 -1.08645189e+00 1.79713953e+00 -2.89066046e-01 5.96457899e-01 -1.66862771e-01 -8.95233512e-01 9.95582819e-01 3.62822533e-01 9.04236495e-01 -2.18191132e-01 -6.77708387e-02 6.39898404e-02 -6.52383864e-01 -3.10151756e-01 5.34443259e-01 -3.89505744e-01 1.08755797e-01 6.52203381e-01 -9.11226775e-03 -2.71430582e-01 -1.00224592e-01 1.83069721e-01 1.03686965e+00 3.30292583e-01 -4.13908660e-01 -2.27907494e-01 2.05204654e-02 -2.12720647e-01 4.44722593e-01 9.22844470e-01 -1.57425463e-01 4.58521575e-01 2.72715032e-01 -4.54522014e-01 -1.23279321e+00 -1.56815982e+00 -5.46181574e-02 6.75483167e-01 -3.49865437e-01 -8.79338458e-02 -6.88419104e-01 -4.16069716e-01 4.06380557e-03 1.04924393e+00 -6.13543689e-01 -6.35221228e-02 -3.13422084e-01 -7.77491927e-01 1.35305837e-01 4.68852520e-01 2.65943296e-02 -8.52375925e-01 -1.94693044e-01 6.04036570e-01 -3.47791985e-02 -3.58721405e-01 -5.27215123e-01 -4.49531414e-02 -1.27152967e+00 -4.53442514e-01 -1.21441090e+00 -3.64164919e-01 4.38460618e-01 -7.93522373e-02 1.17441106e+00 -6.41511023e-01 3.37220579e-02 7.54624903e-01 1.76655035e-02 -1.54852271e-01 -4.81185436e-01 -2.85535306e-01 1.51390523e-01 2.68361866e-01 4.27785784e-01 -9.56073344e-01 -7.02932537e-01 1.19464643e-01 -8.40004742e-01 -2.25837365e-01 -1.93132944e-02 6.19693339e-01 9.48431969e-01 4.08284925e-02 1.60951748e-01 -6.06038511e-01 8.59570503e-01 -8.18121552e-01 -1.05742157e+00 2.05303401e-01 -8.35852265e-01 5.59442222e-01 4.12651479e-01 -7.24428415e-01 -1.33549523e+00 1.37682348e-01 6.33194149e-02 -7.52670109e-01 3.47732306e-02 6.36754513e-01 5.32110393e-01 1.77475959e-01 4.14523125e-01 5.20473421e-01 -1.38485342e-01 -9.39184368e-01 7.04215467e-01 4.05750841e-01 4.33514297e-01 -6.92014873e-01 6.64234996e-01 1.06467485e+00 2.45905146e-01 -4.75132704e-01 -6.40012264e-01 -4.60454106e-01 -3.57060850e-01 -4.53243554e-01 9.69673574e-01 -7.60994673e-01 -6.64277673e-01 7.40931869e-01 -1.16992629e+00 -5.07956147e-01 -4.44466412e-01 6.55459523e-01 -7.15784550e-01 4.63212460e-01 -8.24452639e-01 -1.46332717e+00 -1.76659793e-01 -7.20871925e-01 1.23435593e+00 6.20726719e-02 -4.60661538e-02 -1.48980010e+00 8.36107314e-01 3.88178080e-02 6.06979609e-01 9.97873247e-02 5.20623982e-01 -8.91329721e-02 -5.83835423e-01 -3.07208806e-01 4.80441004e-01 3.72496307e-01 -3.26914281e-01 3.49224210e-01 -8.81501734e-01 -1.90098584e-01 6.16858065e-01 -8.52061883e-02 1.28091085e+00 9.39474761e-01 1.16813016e+00 -6.92907050e-02 -3.92977715e-01 4.77375418e-01 1.26966548e+00 5.60545847e-02 5.21707773e-01 -2.56736845e-01 4.05353576e-01 4.70750600e-01 2.16858029e-01 7.38010228e-01 4.98672575e-01 3.49520147e-01 2.39210967e-02 2.40243524e-01 1.63285390e-01 -4.78917778e-01 6.91207409e-01 1.28677225e+00 -1.01892971e-01 -3.96218598e-01 -9.69348490e-01 5.34880459e-01 -1.99949765e+00 -1.41077924e+00 -2.95383811e-01 2.09696889e+00 5.89861333e-01 -3.09279673e-02 -2.64409147e-02 -1.90086544e-01 6.38930619e-01 3.57841551e-01 -8.79717767e-01 -7.69613357e-03 -2.36642025e-02 1.50021970e-01 4.18095827e-01 7.23241508e-01 -9.15400863e-01 6.69462144e-01 6.61793756e+00 9.51267838e-01 -9.52388525e-01 5.00699401e-01 7.10038900e-01 4.12995629e-02 -6.88260019e-01 -3.59597919e-03 -7.19638646e-01 7.36343920e-01 1.24489605e+00 -2.15001747e-01 7.52851367e-01 8.09293985e-01 5.44425845e-01 3.04382034e-02 -8.64292562e-01 7.91784942e-01 -1.06901899e-01 -1.25339222e+00 -3.48547369e-01 6.38689324e-02 8.83194327e-01 4.83576655e-01 3.23330700e-01 1.06539890e-01 8.66502225e-01 -7.10420191e-01 8.41095507e-01 1.57125080e+00 1.69820756e-01 -4.68670934e-01 3.18927079e-01 4.69952941e-01 -1.00249505e+00 2.81654537e-01 -4.82680619e-01 3.13503370e-02 7.31819272e-01 1.33221495e+00 -1.32780880e-01 -9.65822712e-02 5.18425107e-01 8.27560067e-01 9.61904898e-02 9.58808124e-01 -2.11310431e-01 1.33872914e+00 -5.62375605e-01 -1.19699933e-01 4.35825363e-02 -7.55074739e-01 8.62793922e-01 8.15722942e-01 6.25792205e-01 -5.02127856e-02 1.09092861e-01 1.15160739e+00 -1.72435846e-02 -2.81451523e-01 -3.71244282e-01 -1.78959742e-01 2.80039877e-01 1.03393233e+00 -6.92522526e-01 -4.71406013e-01 -1.11929320e-01 1.14921510e+00 2.92134196e-01 7.87858784e-01 -8.51404011e-01 1.27762258e-01 3.34765524e-01 -1.91002592e-01 3.95727158e-01 -6.77876234e-01 -1.46713883e-01 -1.48565543e+00 7.05758408e-02 -2.35878631e-01 3.23626071e-01 -8.95826101e-01 -1.93858397e+00 4.49353129e-01 -1.26818076e-01 -1.08784735e+00 -6.44111335e-01 -1.65852293e-01 -6.41616344e-01 1.05138350e+00 -1.42965162e+00 -7.34185934e-01 1.12059392e-01 4.80786920e-01 5.64795017e-01 -3.22994143e-02 7.79386759e-01 -2.22172707e-01 -1.39962152e-01 2.61586104e-02 1.15333331e+00 -3.44040871e-01 3.00469697e-01 -1.23643315e+00 6.85968220e-01 6.10052407e-01 4.07231599e-01 4.57371950e-01 1.05065787e+00 -8.25721502e-01 -1.42207491e+00 -8.52615058e-01 8.28748465e-01 -5.19662380e-01 9.97607768e-01 -2.79433221e-01 -1.01110399e+00 3.84403676e-01 6.60996735e-02 -9.47833657e-02 5.08309364e-01 -1.24143265e-01 -5.06431460e-02 1.76612008e-02 -1.41639376e+00 2.00019002e-01 6.63345218e-01 -6.19365692e-01 -4.56393063e-01 5.22091329e-01 5.48738360e-01 -1.07576326e-01 -1.15058875e+00 -1.53269455e-01 5.32985449e-01 -6.49288714e-01 8.57623935e-01 -6.00642979e-01 2.99156398e-01 7.09255487e-02 -2.15754971e-01 -1.49480796e+00 -4.38825935e-01 -9.39497590e-01 -8.09196055e-01 1.24046743e+00 2.20272601e-01 -5.16421676e-01 9.07220542e-01 9.84116077e-01 3.08204591e-01 -5.72604120e-01 -1.23397589e+00 -8.73186529e-01 4.92637694e-01 -8.68706822e-01 2.84716219e-01 6.67856216e-01 -3.49232435e-01 -1.53420687e-01 -6.32821023e-01 1.12771019e-01 1.40825295e+00 6.35602698e-02 1.80420846e-01 -1.38816881e+00 -5.44525862e-01 -3.22213918e-01 -7.98110440e-02 -1.28218389e+00 4.67373341e-01 -6.10872746e-01 4.49786633e-02 -1.31114256e+00 1.94584772e-01 -3.36830735e-01 -3.86519700e-01 -2.87396878e-01 1.23511843e-01 -5.20208254e-02 -1.23183941e-02 5.47572017e-01 -3.47562999e-01 1.02176559e+00 9.22494411e-01 -3.51347655e-01 -1.34835511e-01 3.88283283e-01 -2.62057595e-02 7.45961726e-01 7.25846052e-01 -1.04068267e+00 -3.46980095e-01 -3.01570535e-01 1.28068075e-01 5.40445924e-01 4.30816323e-01 -6.45531535e-01 4.11959112e-01 -3.25538903e-01 2.92486161e-01 -7.83790648e-01 6.50027394e-01 -7.09114194e-01 3.42665017e-01 3.26908469e-01 -5.77919424e-01 9.94971618e-02 -3.96291643e-01 1.11108959e+00 1.30736781e-02 -3.37474436e-01 5.52761078e-01 -3.32183242e-02 -3.68444413e-01 5.06602824e-01 -7.23160326e-01 1.48273632e-01 5.85538030e-01 3.39433014e-01 7.53014833e-02 -8.27554226e-01 -1.11540401e+00 -1.19288556e-01 4.32594806e-01 1.01672940e-01 6.52158380e-01 -1.36358273e+00 -7.48711348e-01 -1.05242021e-01 -4.42071527e-01 -5.05943716e-01 3.03361624e-01 7.66158402e-01 5.38900159e-02 7.05059096e-02 4.08599168e-01 -6.76646352e-01 -7.44610012e-01 5.16630054e-01 3.32503915e-01 -4.30931419e-01 -4.64966744e-01 8.13579977e-01 -3.50391179e-01 -2.53777921e-01 2.91992754e-01 -3.35886300e-01 4.92832698e-02 -7.68950954e-02 4.12154555e-01 7.92352319e-01 -3.17800730e-01 -4.90980029e-01 -1.41252428e-01 4.76047665e-01 3.22433650e-01 -7.87174642e-01 1.68924630e+00 -2.70075023e-01 -2.19264910e-01 1.06787777e+00 1.28735888e+00 -8.01534876e-02 -1.76972699e+00 -3.25378031e-01 4.61192764e-02 -3.07383209e-01 6.40514433e-01 -5.57067573e-01 -1.01692677e+00 1.10819459e+00 1.09392011e+00 3.03311735e-01 7.99203217e-01 1.50776699e-01 8.51016998e-01 2.03533843e-01 3.69648099e-01 -1.12099218e+00 -2.06938423e-02 3.13255817e-01 7.10209072e-01 -1.14216292e+00 -1.55719832e-01 1.83285624e-01 -5.06012261e-01 1.20676458e+00 -2.15493634e-01 -5.53112686e-01 1.18517804e+00 3.70886356e-01 -3.38620663e-01 -2.95059413e-01 -8.96572053e-01 1.20245941e-01 5.13681829e-01 6.20059848e-01 4.50559527e-01 3.24544311e-01 -2.44046852e-01 5.80222309e-01 1.18818343e-01 3.76736879e-01 1.59043774e-01 6.10456526e-01 -4.54394728e-01 -1.07005668e+00 -3.19203645e-01 4.29513961e-01 -3.12056065e-01 -1.71099737e-01 1.12127792e-02 -8.30279514e-02 -4.71403599e-01 8.41888785e-01 4.36957031e-02 1.14055716e-01 -5.67408949e-02 3.14786136e-01 3.87580097e-01 1.03102319e-01 -8.52199867e-02 3.73506814e-01 -3.88940424e-01 -4.03850526e-01 -3.88877004e-01 -1.21024108e+00 -1.00407422e+00 -4.01520967e-01 -1.04964167e-01 2.41397321e-01 1.02627456e+00 8.52493823e-01 3.90765429e-01 2.33574882e-01 6.76522315e-01 -1.05825460e+00 -1.04540014e+00 -8.75628471e-01 -7.00837135e-01 2.66412705e-01 1.85325697e-01 -6.29024148e-01 -5.86760819e-01 2.65168369e-01]	[6.9056925773620605, 3.7397594451904297]
823b0f42-02b4-4a99-b569-f2a025da783a	rotation-constrained-cross-view-feature	2305.12704	null	https://arxiv.org/abs/2305.12704v1	https://arxiv.org/pdf/2305.12704v1.pdf	Rotation-Constrained Cross-View Feature Fusion for Multi-View Appearance-based Gaze Estimation	Appearance-based gaze estimation has been actively studied in recent years. However, its generalization performance for unseen head poses is still a significant limitation for existing methods. This work proposes a generalizable multi-view gaze estimation task and a cross-view feature fusion method to address this issue. In addition to paired images, our method takes the relative rotation matrix between two cameras as additional input. The proposed network learns to extract rotatable feature representation by using relative rotation as a constraint and adaptively fuses the rotatable features via stacked fusion modules. This simple yet efficient approach significantly improves generalization performance under unseen head poses without significantly increasing computational cost. The model can be trained with random combinations of cameras without fixing the positioning and can generalize to unseen camera pairs during inference. Through experiments using multiple datasets, we demonstrate the advantage of the proposed method over baseline methods, including state-of-the-art domain generalization approaches.	['Yusuke Sugano', 'Jiawei Qin', 'Tianyi Wu', 'Yoichiro Hisadome']	2023-05-22	null	null	null	null	['gaze-estimation']	['computer-vision']	[-8.04735254e-03 -2.78009791e-02 -1.87162399e-01 -5.98075092e-01 -6.04814589e-01 -4.47265714e-01 3.16899776e-01 -5.53800523e-01 -5.45368016e-01 5.79625905e-01 -5.40788956e-02 1.82323948e-01 6.82050511e-02 -1.09371714e-01 -8.68677199e-01 -8.42492163e-01 4.20411140e-01 -3.78248170e-02 1.45287737e-01 -3.44013348e-02 2.54380792e-01 3.29020292e-01 -1.95918965e+00 -2.01700911e-01 9.28699732e-01 8.39794993e-01 1.56031385e-01 4.20346141e-01 7.41529524e-01 2.79772341e-01 -5.25898576e-01 -3.93102974e-01 2.97918469e-01 -3.76336500e-02 -5.03646672e-01 3.05492103e-01 1.32538295e+00 -6.38279200e-01 -1.75549716e-01 1.00503147e+00 6.76855147e-01 3.71301144e-01 3.91577721e-01 -1.36681569e+00 -8.76867712e-01 -1.04942463e-01 -1.06268704e+00 2.39024639e-01 7.05766857e-01 -6.68053105e-02 7.44408965e-01 -1.04240596e+00 4.92053658e-01 1.21540606e+00 3.25245500e-01 8.76242816e-01 -1.06673694e+00 -1.06632149e+00 6.60423756e-01 4.36761975e-01 -1.43592834e+00 -6.31154418e-01 7.71770298e-01 -1.45582080e-01 7.55929351e-01 4.25353087e-02 5.44250607e-01 1.34995151e+00 -6.57360852e-02 8.96949947e-01 1.10308826e+00 -3.20219219e-01 1.86001491e-02 6.86724484e-02 1.35482043e-01 6.91650331e-01 4.12277639e-01 -1.46275878e-01 -9.74542856e-01 -7.05019087e-02 6.10397935e-01 2.09419891e-01 -7.84881413e-01 -6.64536893e-01 -1.04129040e+00 5.02864778e-01 6.37060940e-01 -1.40919581e-01 -3.29721011e-02 -1.44741416e-01 -9.13136974e-02 1.79462694e-02 6.10813797e-01 1.11420073e-01 -4.73278224e-01 3.10685877e-02 -9.34804320e-01 2.28049651e-01 3.86354446e-01 1.25537407e+00 7.60140538e-01 -1.22388631e-01 2.22544461e-01 7.95604885e-01 4.99953061e-01 6.67083561e-01 4.38233078e-01 -8.53797734e-01 6.10399008e-01 5.94161987e-01 2.73185614e-02 -9.11595523e-01 -5.39095938e-01 -3.15079510e-01 -5.22240460e-01 7.04379901e-02 3.65726292e-01 -2.92205781e-01 -9.50309694e-01 2.15329909e+00 7.06150532e-01 3.72878790e-01 -5.14141805e-02 1.10768700e+00 7.75411546e-01 1.37967974e-01 -2.11983964e-01 -3.38196039e-01 1.42067122e+00 -1.05715024e+00 -6.98350668e-01 -3.92974436e-01 3.31526756e-01 -5.90315104e-01 7.18669057e-01 5.96939683e-01 -1.10577619e+00 -4.66240466e-01 -1.17203736e+00 -3.44314128e-01 -1.52940691e-01 3.62174481e-01 5.49671233e-01 6.88782215e-01 -1.40663695e+00 8.72919038e-02 -1.04806471e+00 -6.62355363e-01 3.63823861e-01 8.70325685e-01 -7.35797048e-01 -1.41659930e-01 -7.95458496e-01 7.77011931e-01 1.74869150e-01 2.73713619e-01 -4.90163296e-01 -4.79809970e-01 -1.17233443e+00 -9.05641094e-02 4.41007912e-01 -9.70133305e-01 1.24678290e+00 -8.63743842e-01 -1.79764402e+00 7.70776689e-01 -7.91143656e-01 -4.41562105e-03 2.65687525e-01 -7.13621557e-01 -4.08037305e-01 2.06940114e-01 -1.37288272e-01 9.86825347e-01 1.11935115e+00 -1.15696895e+00 -5.96679449e-01 -1.00079501e+00 1.64356723e-01 6.67708278e-01 -5.48308671e-01 7.65034184e-03 -6.37104630e-01 -3.98715228e-01 4.82324421e-01 -1.31597638e+00 2.89719760e-01 1.41864315e-01 -1.45492449e-01 -3.63950163e-01 1.02213323e+00 -3.69952410e-01 1.04785800e+00 -2.02646279e+00 4.67716306e-01 -1.61114380e-01 3.30910921e-01 -3.18596065e-02 8.94659087e-02 -7.06813857e-02 -1.63469508e-01 -2.22342402e-01 7.94057027e-02 -7.39778757e-01 -2.13116735e-01 -1.17786303e-01 -1.37310317e-02 8.29911530e-01 5.72743304e-02 5.93637824e-01 -5.95147133e-01 -3.42960805e-01 8.69270191e-02 8.11055243e-01 -7.81409562e-01 2.89245188e-01 2.39173084e-01 6.55052304e-01 -2.23029017e-01 5.07366002e-01 1.10286283e+00 -3.45390856e-01 1.46248773e-01 -3.24690402e-01 2.32280195e-01 -4.24741209e-02 -1.02520490e+00 2.02865243e+00 -3.08384687e-01 6.95588171e-01 -5.44217415e-02 -7.38797784e-01 7.05397904e-01 2.01646701e-01 1.25272363e-01 -3.66542161e-01 3.62896293e-01 -1.44537196e-01 -1.24527566e-01 -5.74522018e-01 6.20749354e-01 2.13744894e-01 6.00848161e-02 3.91063511e-01 3.66431952e-01 2.34572738e-01 -1.21381804e-01 3.83657217e-03 5.45639336e-01 6.01114333e-01 4.66072470e-01 7.91695639e-02 7.25287139e-01 -7.17950761e-01 6.88921154e-01 1.57080024e-01 -3.86667699e-01 7.60058761e-01 1.09026186e-01 -2.70896673e-01 -4.96590078e-01 -9.91703689e-01 -1.51271358e-01 1.30478084e+00 4.74219829e-01 -3.17809731e-01 -9.46260035e-01 -8.10261428e-01 -8.02700594e-02 5.04760861e-01 -8.67382228e-01 -7.45285898e-02 -6.76543415e-01 -6.30806625e-01 1.52088508e-01 7.02866077e-01 6.79769456e-01 -3.80863905e-01 -5.21117032e-01 -5.88746190e-01 -3.15263033e-01 -1.39536130e+00 -7.05466092e-01 -2.98433334e-01 -8.22187185e-01 -1.00592637e+00 -9.27240133e-01 -7.63844311e-01 1.08506060e+00 7.76838899e-01 6.19119465e-01 -1.84351727e-01 1.23967789e-01 6.38740003e-01 -2.34551474e-01 -3.58771205e-01 3.32834482e-01 1.64894626e-01 4.10144627e-01 2.62526661e-01 6.58310413e-01 -5.63149869e-01 -9.31805670e-01 6.15943849e-01 -4.91546631e-01 -4.16757576e-02 3.93762082e-01 7.65162170e-01 3.72863948e-01 -7.00503707e-01 3.09035093e-01 -5.67584932e-01 3.78172338e-01 -3.90902936e-01 -7.14996696e-01 3.63884598e-01 -4.19978023e-01 1.01661287e-01 2.28337228e-01 -4.98733282e-01 -1.40021479e+00 8.67344737e-02 4.80223447e-01 -6.37929082e-01 -3.45528305e-01 9.58692282e-02 -3.72895867e-01 -3.56944650e-01 4.78827268e-01 3.83859649e-02 5.73311895e-02 -4.42956805e-01 3.34461987e-01 7.97473252e-01 4.41942334e-01 -3.62442434e-01 7.11926520e-01 6.62546098e-01 -5.27456999e-02 -7.90931344e-01 -9.22205567e-01 -5.34688771e-01 -9.53291833e-01 -2.05549076e-01 9.05612290e-01 -1.31334245e+00 -1.15930676e+00 6.49868667e-01 -1.01913798e+00 2.10835040e-01 7.32344568e-01 7.48009861e-01 -3.60288173e-01 5.18354058e-01 -1.27634600e-01 -7.01454818e-01 -2.76567549e-01 -1.34430861e+00 1.41795945e+00 6.73993051e-01 -1.25540107e-01 -8.22266757e-01 -8.64451304e-02 7.22782075e-01 1.37868095e-02 3.53010744e-02 2.02028319e-01 -3.72771770e-01 -6.66739047e-01 -1.26909465e-01 -1.61326125e-01 1.19854480e-01 2.02882662e-01 -9.40492675e-02 -1.28308392e+00 -8.05018783e-01 6.50051385e-02 -4.52712387e-01 7.31777549e-01 4.29607302e-01 1.09647870e+00 -1.19602598e-01 -6.34843767e-01 9.97593105e-01 1.02769887e+00 4.43677604e-02 3.27891171e-01 2.86767811e-01 8.91887426e-01 3.80492091e-01 5.19358814e-01 2.85156488e-01 8.99264157e-01 8.50938380e-01 4.79549825e-01 2.01711178e-01 1.85446069e-01 -8.67097750e-02 4.51566130e-01 6.65823340e-01 -3.27364951e-01 -4.06075925e-01 -5.72664797e-01 3.84643227e-01 -1.71845794e+00 -8.63607764e-01 4.07374591e-01 2.28940058e+00 5.15838563e-01 -3.29257518e-01 1.87264666e-01 -1.21549748e-01 8.61905694e-01 1.21197402e-01 -8.63953114e-01 -1.31081000e-01 4.99923229e-02 7.22652897e-02 2.34424591e-01 2.23605007e-01 -1.17186368e+00 8.82832587e-01 5.91135168e+00 1.23809278e-01 -1.46213925e+00 8.79101008e-02 1.25401661e-01 -4.48310137e-01 1.53923571e-01 -1.22160867e-01 -1.10905814e+00 2.31301636e-01 5.26842475e-01 2.38453910e-01 4.52314615e-01 8.75693142e-01 -1.33495495e-01 -2.59945691e-01 -1.18981469e+00 1.32928288e+00 9.87917304e-01 -5.73825479e-01 -1.98394567e-01 1.09319478e-01 6.84627891e-01 7.39066899e-02 5.55377066e-01 7.83780217e-02 -1.00647800e-01 -7.42769837e-01 5.81404507e-01 4.45552707e-01 8.42748880e-01 -5.45948923e-01 3.49681616e-01 2.60970950e-01 -9.07443941e-01 -2.90767103e-01 -2.09747061e-01 -2.82845180e-02 6.61709607e-02 -1.93940997e-01 -7.37837434e-01 5.64316690e-01 9.45350230e-01 9.35800612e-01 -8.55061710e-01 8.99853408e-01 -3.45937252e-01 3.20501417e-01 -4.67115074e-01 1.54924080e-01 -2.02235714e-01 7.35760704e-02 5.89324892e-01 5.76199889e-01 3.58965099e-01 -2.95492131e-02 -1.13215365e-01 4.43359584e-01 -1.33992046e-01 -9.68423113e-02 -4.61971849e-01 6.02013409e-01 4.91882980e-01 1.24450672e+00 -2.20975459e-01 -2.62009818e-02 -7.46041179e-01 1.10728180e+00 7.71499276e-01 7.68685460e-01 -8.62336278e-01 -1.80935472e-01 7.17384338e-01 -5.39559387e-02 5.87750256e-01 -2.36585572e-01 1.58533588e-01 -1.83763576e+00 3.63427371e-01 -7.79346108e-01 3.79183769e-01 -1.01581228e+00 -9.42045808e-01 5.55853307e-01 3.72052163e-01 -1.48012614e+00 -3.65583420e-01 -6.60396397e-01 -2.56987482e-01 7.99944699e-01 -1.49915564e+00 -1.50771511e+00 -7.75901139e-01 8.33642542e-01 4.37843233e-01 -1.95353389e-01 6.64753497e-01 4.44026031e-02 -9.28648412e-01 1.19702160e+00 -1.83085397e-01 -7.61687458e-02 1.30413651e+00 -1.14444804e+00 1.33724093e-01 8.67178798e-01 -6.08609654e-02 1.15221632e+00 5.71555495e-01 -2.55656958e-01 -1.27546537e+00 -7.73718953e-01 5.17647445e-01 -5.56757867e-01 1.62327573e-01 -6.09467864e-01 -8.39868486e-01 1.06184494e+00 5.72170675e-01 2.17069536e-01 8.64580631e-01 3.67986113e-01 -5.86999834e-01 -3.80728185e-01 -1.04934371e+00 5.59098661e-01 1.17575681e+00 -3.80926907e-01 -6.76028788e-01 -2.62437691e-03 5.42112172e-01 -9.40333486e-01 -7.55172908e-01 4.02027458e-01 9.03224468e-01 -1.02909911e+00 7.07489014e-01 -4.39635932e-01 -1.16164476e-01 -4.80024636e-01 5.42487875e-02 -1.28479743e+00 -1.01243593e-01 -7.90186703e-01 -1.69723421e-01 9.99285102e-01 2.01610774e-01 -9.06592071e-01 7.56010056e-01 7.82136083e-01 4.65717306e-03 -6.30185306e-01 -9.05283749e-01 -5.15503466e-01 -7.38030002e-02 -1.33299455e-01 6.16267920e-01 6.89822435e-01 1.59891546e-01 4.92970347e-01 -4.98540044e-01 5.81221461e-01 7.46332765e-01 3.00629549e-02 1.21248066e+00 -1.28930163e+00 -1.18833005e-01 -3.75002846e-02 -7.58574128e-01 -1.41700792e+00 5.40589929e-01 -5.79798758e-01 -2.30381638e-01 -9.17516470e-01 3.46760273e-01 1.46250710e-01 -2.04313532e-01 4.20582443e-01 -4.70941216e-01 2.35398561e-01 1.61251113e-01 1.65035367e-01 -7.61875331e-01 5.58237076e-01 1.35247672e+00 1.62129968e-01 -1.86046407e-01 6.86777905e-02 -8.03225160e-01 9.08338547e-01 6.41406894e-01 -2.78406173e-01 -7.29811072e-01 -7.78370261e-01 2.39812940e-01 1.32479880e-03 4.57505971e-01 -1.06108379e+00 5.46336234e-01 3.61626819e-02 7.05752492e-01 -6.67037010e-01 7.77104855e-01 -8.12842667e-01 -2.42909715e-01 -1.69641823e-01 -6.28085956e-02 3.81156504e-01 3.09680969e-01 8.18583608e-01 -1.18435651e-01 1.85521916e-01 7.59527683e-01 3.09528649e-01 -5.06675959e-01 3.72092456e-01 1.70559406e-01 -2.77670361e-02 1.17659986e+00 -5.11479139e-01 -4.22193825e-01 -5.20846903e-01 -8.18418741e-01 2.23638579e-01 8.17756653e-01 6.64302826e-01 6.57341480e-01 -1.05036390e+00 -2.84418225e-01 4.59599197e-01 2.86946297e-01 1.03088729e-01 4.66907740e-01 1.03371823e+00 -1.25736579e-01 3.91890287e-01 -1.73229411e-01 -9.90206242e-01 -1.59541500e+00 6.23262346e-01 3.62879455e-01 3.39411706e-01 -2.46790931e-01 9.89035785e-01 4.94766563e-01 -3.30824941e-01 2.24058941e-01 -2.99233675e-01 -4.49100912e-01 -8.29067081e-03 6.13110602e-01 2.51176387e-01 -5.02124242e-03 -1.08850861e+00 -5.25074184e-01 1.18295312e+00 -5.42019308e-01 8.15690961e-03 1.26818216e+00 -5.64257503e-01 6.74256235e-02 3.08377922e-01 1.37398231e+00 -1.26245752e-01 -1.42569768e+00 -2.81682789e-01 -5.07844031e-01 -5.79533517e-01 7.67344236e-02 -6.09380245e-01 -1.22662008e+00 1.08091938e+00 8.67841959e-01 -5.63424885e-01 1.31868541e+00 -1.03049256e-01 5.57281077e-01 5.26562572e-01 3.86091232e-01 -9.04663086e-01 1.44927725e-01 1.77495852e-01 7.96711743e-01 -1.68634868e+00 2.41233036e-01 -4.73344594e-01 -7.26829767e-01 1.02208030e+00 1.12771106e+00 1.27118127e-02 6.82613790e-01 -1.77600160e-01 2.71246899e-02 -8.29246864e-02 -7.01106369e-01 -6.62193894e-02 6.22743905e-01 6.54654980e-01 4.24629599e-01 -1.81155995e-01 -1.32679865e-01 2.39743829e-01 -2.07256690e-01 5.92812821e-02 3.49910915e-01 8.70708704e-01 -1.11766875e-01 -8.92770946e-01 -4.52400863e-01 6.80024177e-02 -5.01805425e-01 8.98353085e-02 -2.77419955e-01 9.97091532e-01 5.76240495e-02 1.00713515e+00 1.22295089e-01 -4.49979693e-01 2.73041844e-01 9.76049379e-02 8.90351832e-01 -4.98532325e-01 -2.59500891e-01 1.21337332e-01 -3.98772478e-01 -6.42015755e-01 -8.25406134e-01 -9.88902569e-01 -8.03258002e-01 -9.66282338e-02 -8.37128580e-01 -2.12949961e-01 4.85639185e-01 9.36188340e-01 7.76778042e-01 2.21651848e-02 5.86376131e-01 -1.09396648e+00 -4.09908116e-01 -1.14890742e+00 -3.70078802e-01 3.75675410e-01 7.05005050e-01 -1.32668149e+00 -4.85511273e-01 1.45025715e-01]	[14.11101245880127, 0.039550408720970154]

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.